diff options
506 files changed, 44738 insertions, 8304 deletions
diff --git a/Documentation/ABI/testing/sysfs-devices-cache_disable b/Documentation/ABI/testing/sysfs-devices-cache_disable new file mode 100644 index 00000000000..175bb4f7051 --- /dev/null +++ b/Documentation/ABI/testing/sysfs-devices-cache_disable @@ -0,0 +1,18 @@ +What: /sys/devices/system/cpu/cpu*/cache/index*/cache_disable_X +Date: August 2008 +KernelVersion: 2.6.27 +Contact: mark.langsdorf@amd.com +Description: These files exist in every cpu's cache index directories. + There are currently 2 cache_disable_# files in each + directory. Reading from these files on a supported + processor will return that cache disable index value + for that processor and node. Writing to one of these + files will cause the specificed cache index to be disabled. + + Currently, only AMD Family 10h Processors support cache index + disable, and only for their L3 caches. See the BIOS and + Kernel Developer's Guide at + http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/31116-Public-GH-BKDG_3.20_2-4-09.pdf + for formatting information and other details on the + cache index disable. +Users: joachim.deguara@amd.com diff --git a/Documentation/DMA-API.txt b/Documentation/DMA-API.txt index d9aa43d78bc..25fb8bcf32a 100644 --- a/Documentation/DMA-API.txt +++ b/Documentation/DMA-API.txt @@ -704,12 +704,24 @@ this directory the following files can currently be found: The current number of free dma_debug_entries in the allocator. + dma-api/driver-filter + You can write a name of a driver into this file + to limit the debug output to requests from that + particular driver. Write an empty string to + that file to disable the filter and see + all errors again. + If you have this code compiled into your kernel it will be enabled by default. If you want to boot without the bookkeeping anyway you can provide 'dma_debug=off' as a boot parameter. This will disable DMA-API debugging. Notice that you can not enable it again at runtime. You have to reboot to do so. +If you want to see debug messages only for a special device driver you can +specify the dma_debug_driver=<drivername> parameter. This will enable the +driver filter at boot time. The debug code will only print errors for that +driver afterwards. This filter can be disabled or changed later using debugfs. + When the code disables itself at runtime this is most likely because it ran out of dma_debug_entries. These entries are preallocated at boot. The number of preallocated entries is defined per architecture. If it is too low for you diff --git a/Documentation/DocBook/Makefile b/Documentation/DocBook/Makefile index b1eb661e630..9632444f6c6 100644 --- a/Documentation/DocBook/Makefile +++ b/Documentation/DocBook/Makefile @@ -13,7 +13,8 @@ DOCBOOKS := z8530book.xml mcabook.xml device-drivers.xml \ gadget.xml libata.xml mtdnand.xml librs.xml rapidio.xml \ genericirq.xml s390-drivers.xml uio-howto.xml scsi.xml \ mac80211.xml debugobjects.xml sh.xml regulator.xml \ - alsa-driver-api.xml writing-an-alsa-driver.xml + alsa-driver-api.xml writing-an-alsa-driver.xml \ + tracepoint.xml ### # The build process is as follows (targets): diff --git a/Documentation/DocBook/tracepoint.tmpl b/Documentation/DocBook/tracepoint.tmpl new file mode 100644 index 00000000000..b0756d0fd57 --- /dev/null +++ b/Documentation/DocBook/tracepoint.tmpl @@ -0,0 +1,89 @@ +<?xml version="1.0" encoding="UTF-8"?> +<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN" + "http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd" []> + +<book id="Tracepoints"> + <bookinfo> + <title>The Linux Kernel Tracepoint API</title> + + <authorgroup> + <author> + <firstname>Jason</firstname> + <surname>Baron</surname> + <affiliation> + <address> + <email>jbaron@redhat.com</email> + </address> + </affiliation> + </author> + </authorgroup> + + <legalnotice> + <para> + This documentation is free software; you can redistribute + it and/or modify it under the terms of the GNU General Public + License as published by the Free Software Foundation; either + version 2 of the License, or (at your option) any later + version. + </para> + + <para> + This program is distributed in the hope that it will be + useful, but WITHOUT ANY WARRANTY; without even the implied + warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. + See the GNU General Public License for more details. + </para> + + <para> + You should have received a copy of the GNU General Public + License along with this program; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, + MA 02111-1307 USA + </para> + + <para> + For more details see the file COPYING in the source + distribution of Linux. + </para> + </legalnotice> + </bookinfo> + + <toc></toc> + <chapter id="intro"> + <title>Introduction</title> + <para> + Tracepoints are static probe points that are located in strategic points + throughout the kernel. 'Probes' register/unregister with tracepoints + via a callback mechanism. The 'probes' are strictly typed functions that + are passed a unique set of parameters defined by each tracepoint. + </para> + + <para> + From this simple callback mechanism, 'probes' can be used to profile, debug, + and understand kernel behavior. There are a number of tools that provide a + framework for using 'probes'. These tools include Systemtap, ftrace, and + LTTng. + </para> + + <para> + Tracepoints are defined in a number of header files via various macros. Thus, + the purpose of this document is to provide a clear accounting of the available + tracepoints. The intention is to understand not only what tracepoints are + available but also to understand where future tracepoints might be added. + </para> + + <para> + The API presented has functions of the form: + <function>trace_tracepointname(function parameters)</function>. These are the + tracepoints callbacks that are found throughout the code. Registering and + unregistering probes with these callback sites is covered in the + <filename>Documentation/trace/*</filename> directory. + </para> + </chapter> + + <chapter id="irq"> + <title>IRQ</title> +!Iinclude/trace/events/irq.h + </chapter> + +</book> diff --git a/Documentation/RCU/trace.txt b/Documentation/RCU/trace.txt index 068848240a8..02cced183b2 100644 --- a/Documentation/RCU/trace.txt +++ b/Documentation/RCU/trace.txt @@ -192,23 +192,24 @@ rcu/rcuhier (which displays the struct rcu_node hierarchy). The output of "cat rcu/rcudata" looks as follows: rcu: - 0 c=4011 g=4012 pq=1 pqc=4011 qp=0 rpfq=1 rp=3c2a dt=23301/73 dn=2 df=1882 of=0 ri=2126 ql=2 b=10 - 1 c=4011 g=4012 pq=1 pqc=4011 qp=0 rpfq=3 rp=39a6 dt=78073/1 dn=2 df=1402 of=0 ri=1875 ql=46 b=10 - 2 c=4010 g=4010 pq=1 pqc=4010 qp=0 rpfq=-5 rp=1d12 dt=16646/0 dn=2 df=3140 of=0 ri=2080 ql=0 b=10 - 3 c=4012 g=4013 pq=1 pqc=4012 qp=1 rpfq=3 rp=2b50 dt=21159/1 dn=2 df=2230 of=0 ri=1923 ql=72 b=10 - 4 c=4012 g=4013 pq=1 pqc=4012 qp=1 rpfq=3 rp=1644 dt=5783/1 dn=2 df=3348 of=0 ri=2805 ql=7 b=10 - 5 c=4012 g=4013 pq=0 pqc=4011 qp=1 rpfq=3 rp=1aac dt=5879/1 dn=2 df=3140 of=0 ri=2066 ql=10 b=10 - 6 c=4012 g=4013 pq=1 pqc=4012 qp=1 rpfq=3 rp=ed8 dt=5847/1 dn=2 df=3797 of=0 ri=1266 ql=10 b=10 - 7 c=4012 g=4013 pq=1 pqc=4012 qp=1 rpfq=3 rp=1fa2 dt=6199/1 dn=2 df=2795 of=0 ri=2162 ql=28 b=10 +rcu: + 0 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=10951/1 dn=0 df=1101 of=0 ri=36 ql=0 b=10 + 1 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=16117/1 dn=0 df=1015 of=0 ri=0 ql=0 b=10 + 2 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=1445/1 dn=0 df=1839 of=0 ri=0 ql=0 b=10 + 3 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=6681/1 dn=0 df=1545 of=0 ri=0 ql=0 b=10 + 4 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=1003/1 dn=0 df=1992 of=0 ri=0 ql=0 b=10 + 5 c=17829 g=17830 pq=1 pqc=17829 qp=1 dt=3887/1 dn=0 df=3331 of=0 ri=4 ql=2 b=10 + 6 c=17829 g=17829 pq=1 pqc=17829 qp=0 dt=859/1 dn=0 df=3224 of=0 ri=0 ql=0 b=10 + 7 c=17829 g=17830 pq=0 pqc=17829 qp=1 dt=3761/1 dn=0 df=1818 of=0 ri=0 ql=2 b=10 rcu_bh: - 0 c=-268 g=-268 pq=1 pqc=-268 qp=0 rpfq=-145 rp=21d6 dt=23301/73 dn=2 df=0 of=0 ri=0 ql=0 b=10 - 1 c=-268 g=-268 pq=1 pqc=-268 qp=1 rpfq=-170 rp=20ce dt=78073/1 dn=2 df=26 of=0 ri=5 ql=0 b=10 - 2 c=-268 g=-268 pq=1 pqc=-268 qp=1 rpfq=-83 rp=fbd dt=16646/0 dn=2 df=28 of=0 ri=4 ql=0 b=10 - 3 c=-268 g=-268 pq=1 pqc=-268 qp=0 rpfq=-105 rp=178c dt=21159/1 dn=2 df=28 of=0 ri=2 ql=0 b=10 - 4 c=-268 g=-268 pq=1 pqc=-268 qp=1 rpfq=-30 rp=b54 dt=5783/1 dn=2 df=32 of=0 ri=0 ql=0 b=10 - 5 c=-268 g=-268 pq=1 pqc=-268 qp=1 rpfq=-29 rp=df5 dt=5879/1 dn=2 df=30 of=0 ri=3 ql=0 b=10 - 6 c=-268 g=-268 pq=1 pqc=-268 qp=1 rpfq=-28 rp=788 dt=5847/1 dn=2 df=32 of=0 ri=0 ql=0 b=10 - 7 c=-268 g=-268 pq=1 pqc=-268 qp=1 rpfq=-53 rp=1098 dt=6199/1 dn=2 df=30 of=0 ri=3 ql=0 b=10 + 0 c=-275 g=-275 pq=1 pqc=-275 qp=0 dt=10951/1 dn=0 df=0 of=0 ri=0 ql=0 b=10 + 1 c=-275 g=-275 pq=1 pqc=-275 qp=0 dt=16117/1 dn=0 df=13 of=0 ri=0 ql=0 b=10 + 2 c=-275 g=-275 pq=1 pqc=-275 qp=0 dt=1445/1 dn=0 df=15 of=0 ri=0 ql=0 b=10 + 3 c=-275 g=-275 pq=1 pqc=-275 qp=0 dt=6681/1 dn=0 df=9 of=0 ri=0 ql=0 b=10 + 4 c=-275 g=-275 pq=1 pqc=-275 qp=0 dt=1003/1 dn=0 df=15 of=0 ri=0 ql=0 b=10 + 5 c=-275 g=-275 pq=1 pqc=-275 qp=0 dt=3887/1 dn=0 df=15 of=0 ri=0 ql=0 b=10 + 6 c=-275 g=-275 pq=1 pqc=-275 qp=0 dt=859/1 dn=0 df=15 of=0 ri=0 ql=0 b=10 + 7 c=-275 g=-275 pq=1 pqc=-275 qp=0 dt=3761/1 dn=0 df=15 of=0 ri=0 ql=0 b=10 The first section lists the rcu_data structures for rcu, the second for rcu_bh. Each section has one line per CPU, or eight for this 8-CPU system. @@ -253,12 +254,6 @@ o "pqc" indicates which grace period the last-observed quiescent o "qp" indicates that RCU still expects a quiescent state from this CPU. -o "rpfq" is the number of rcu_pending() calls on this CPU required - to induce this CPU to invoke force_quiescent_state(). - -o "rp" is low-order four hex digits of the count of how many times - rcu_pending() has been invoked on this CPU. - o "dt" is the current value of the dyntick counter that is incremented when entering or leaving dynticks idle state, either by the scheduler or by irq. The number after the "/" is the interrupt @@ -305,6 +300,9 @@ o "b" is the batch limit for this CPU. If more than this number of RCU callbacks is ready to invoke, then the remainder will be deferred. +There is also an rcu/rcudata.csv file with the same information in +comma-separated-variable spreadsheet format. + The output of "cat rcu/rcugp" looks as follows: @@ -411,3 +409,63 @@ o Each element of the form "1/1 0:127 ^0" represents one struct For example, the first entry at the lowest level shows "^0", indicating that it corresponds to bit zero in the first entry at the middle level. + + +The output of "cat rcu/rcu_pending" looks as follows: + +rcu: + 0 np=255892 qsp=53936 cbr=0 cng=14417 gpc=10033 gps=24320 nf=6445 nn=146741 + 1 np=261224 qsp=54638 cbr=0 cng=25723 gpc=16310 gps=2849 nf=5912 nn=155792 + 2 np=237496 qsp=49664 cbr=0 cng=2762 gpc=45478 gps=1762 nf=1201 nn=136629 + 3 np=236249 qsp=48766 cbr=0 cng=286 gpc=48049 gps=1218 nf=207 nn=137723 + 4 np=221310 qsp=46850 cbr=0 cng=26 gpc=43161 gps=4634 nf=3529 nn=123110 + 5 np=237332 qsp=48449 cbr=0 cng=54 gpc=47920 gps=3252 nf=201 nn=137456 + 6 np=219995 qsp=46718 cbr=0 cng=50 gpc=42098 gps=6093 nf=4202 nn=120834 + 7 np=249893 qsp=49390 cbr=0 cng=72 gpc=38400 gps=17102 nf=41 nn=144888 +rcu_bh: + 0 np=146741 qsp=1419 cbr=0 cng=6 gpc=0 gps=0 nf=2 nn=145314 + 1 np=155792 qsp=12597 cbr=0 cng=0 gpc=4 gps=8 nf=3 nn=143180 + 2 np=136629 qsp=18680 cbr=0 cng=0 gpc=7 gps=6 nf=0 nn=117936 + 3 np=137723 qsp=2843 cbr=0 cng=0 gpc=10 gps=7 nf=0 nn=134863 + 4 np=123110 qsp=12433 cbr=0 cng=0 gpc=4 gps=2 nf=0 nn=110671 + 5 np=137456 qsp=4210 cbr=0 cng=0 gpc=6 gps=5 nf=0 nn=133235 + 6 np=120834 qsp=9902 cbr=0 cng=0 gpc=6 gps=3 nf=2 nn=110921 + 7 np=144888 qsp=26336 cbr=0 cng=0 gpc=8 gps=2 nf=0 nn=118542 + +As always, this is once again split into "rcu" and "rcu_bh" portions. +The fields are as follows: + +o "np" is the number of times that __rcu_pending() has been invoked + for the corresponding flavor of RCU. + +o "qsp" is the number of times that the RCU was waiting for a + quiescent state from this CPU. + +o "cbr" is the number of times that this CPU had RCU callbacks + that had passed through a grace period, and were thus ready + to be invoked. + +o "cng" is the number of times that this CPU needed another + grace period while RCU was idle. + +o "gpc" is the number of times that an old grace period had + completed, but this CPU was not yet aware of it. + +o "gps" is the number of times that a new grace period had started, + but this CPU was not yet aware of it. + +o "nf" is the number of times that this CPU suspected that the + current grace period had run for too long, and thus needed to + be forced. + + Please note that "forcing" consists of sending resched IPIs + to holdout CPUs. If that CPU really still is in an old RCU + read-side critical section, then we really do have to wait for it. + The assumption behing "forcing" is that the CPU is not still in + an old RCU read-side critical section, but has not yet responded + for some other reason. + +o "nn" is the number of times that this CPU needed nothing. Alert + readers will note that the rcu "nn" number for a given CPU very + closely matches the rcu_bh "np" number for that same CPU. This + is due to short-circuit evaluation in rcu_pending(). diff --git a/Documentation/futex-requeue-pi.txt b/Documentation/futex-requeue-pi.txt new file mode 100644 index 00000000000..9dc1ff4fd53 --- /dev/null +++ b/Documentation/futex-requeue-pi.txt @@ -0,0 +1,131 @@ +Futex Requeue PI +---------------- + +Requeueing of tasks from a non-PI futex to a PI futex requires +special handling in order to ensure the underlying rt_mutex is never +left without an owner if it has waiters; doing so would break the PI +boosting logic [see rt-mutex-desgin.txt] For the purposes of +brevity, this action will be referred to as "requeue_pi" throughout +this document. Priority inheritance is abbreviated throughout as +"PI". + +Motivation +---------- + +Without requeue_pi, the glibc implementation of +pthread_cond_broadcast() must resort to waking all the tasks waiting +on a pthread_condvar and letting them try to sort out which task +gets to run first in classic thundering-herd formation. An ideal +implementation would wake the highest-priority waiter, and leave the +rest to the natural wakeup inherent in unlocking the mutex +associated with the condvar. + +Consider the simplified glibc calls: + +/* caller must lock mutex */ +pthread_cond_wait(cond, mutex) +{ + lock(cond->__data.__lock); + unlock(mutex); + do { + unlock(cond->__data.__lock); + futex_wait(cond->__data.__futex); + lock(cond->__data.__lock); + } while(...) + unlock(cond->__data.__lock); + lock(mutex); +} + +pthread_cond_broadcast(cond) +{ + lock(cond->__data.__lock); + unlock(cond->__data.__lock); + futex_requeue(cond->data.__futex, cond->mutex); +} + +Once pthread_cond_broadcast() requeues the tasks, the cond->mutex +has waiters. Note that pthread_cond_wait() attempts to lock the +mutex only after it has returned to user space. This will leave the +underlying rt_mutex with waiters, and no owner, breaking the +previously mentioned PI-boosting algorithms. + +In order to support PI-aware pthread_condvar's, the kernel needs to +be able to requeue tasks to PI futexes. This support implies that +upon a successful futex_wait system call, the caller would return to +user space already holding the PI futex. The glibc implementation +would be modified as follows: + + +/* caller must lock mutex */ +pthread_cond_wait_pi(cond, mutex) +{ + lock(cond->__data.__lock); + unlock(mutex); + do { + unlock(cond->__data.__lock); + futex_wait_requeue_pi(cond->__data.__futex); + lock(cond->__data.__lock); + } while(...) + unlock(cond->__data.__lock); + /* the kernel acquired the the mutex for us */ +} + +pthread_cond_broadcast_pi(cond) +{ + lock(cond->__data.__lock); + unlock(cond->__data.__lock); + futex_requeue_pi(cond->data.__futex, cond->mutex); +} + +The actual glibc implementation will likely test for PI and make the +necessary changes inside the existing calls rather than creating new +calls for the PI cases. Similar changes are needed for +pthread_cond_timedwait() and pthread_cond_signal(). + +Implementation +-------------- + +In order to ensure the rt_mutex has an owner if it has waiters, it +is necessary for both the requeue code, as well as the waiting code, +to be able to acquire the rt_mutex before returning to user space. +The requeue code cannot simply wake the waiter and leave it to +acquire the rt_mutex as it would open a race window between the +requeue call returning to user space and the waiter waking and +starting to run. This is especially true in the uncontended case. + +The solution involves two new rt_mutex helper routines, +rt_mutex_start_proxy_lock() and rt_mutex_finish_proxy_lock(), which +allow the requeue code to acquire an uncontended rt_mutex on behalf +of the waiter and to enqueue the waiter on a contended rt_mutex. +Two new system calls provide the kernel<->user interface to +requeue_pi: FUTEX_WAIT_REQUEUE_PI and FUTEX_REQUEUE_CMP_PI. + +FUTEX_WAIT_REQUEUE_PI is called by the waiter (pthread_cond_wait() +and pthread_cond_timedwait()) to block on the initial futex and wait +to be requeued to a PI-aware futex. The implementation is the +result of a high-speed collision between futex_wait() and +futex_lock_pi(), with some extra logic to check for the additional +wake-up scenarios. + +FUTEX_REQUEUE_CMP_PI is called by the waker +(pthread_cond_broadcast() and pthread_cond_signal()) to requeue and +possibly wake the waiting tasks. Internally, this system call is +still handled by futex_requeue (by passing requeue_pi=1). Before +requeueing, futex_requeue() attempts to acquire the requeue target +PI futex on behalf of the top waiter. If it can, this waiter is +woken. futex_requeue() then proceeds to requeue the remaining +nr_wake+nr_requeue tasks to the PI futex, calling +rt_mutex_start_proxy_lock() prior to each requeue to prepare the +task as a waiter on the underlying rt_mutex. It is possible that +the lock can be acquired at this stage as well, if so, the next +waiter is woken to finish the acquisition of the lock. + +FUTEX_REQUEUE_PI accepts nr_wake and nr_requeue as arguments, but +their sum is all that really matters. futex_requeue() will wake or +requeue up to nr_wake + nr_requeue tasks. It will wake only as many +tasks as it can acquire the lock for, which in the majority of cases +should be 0 as good programming practice dictates that the caller of +either pthread_cond_broadcast() or pthread_cond_signal() acquire the +mutex prior to making the call. FUTEX_REQUEUE_PI requires that +nr_wake=1. nr_requeue should be INT_MAX for broadcast and 0 for +signal. diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 19d83e1fc01..4a3c2209a12 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -56,7 +56,6 @@ parameter is applicable: ISAPNP ISA PnP code is enabled. ISDN Appropriate ISDN support is enabled. JOY Appropriate joystick support is enabled. - KMEMTRACE kmemtrace is enabled. LIBATA Libata driver is enabled LP Printer support is enabled. LOOP Loopback device support is enabled. @@ -329,11 +328,6 @@ and is between 256 and 4096 characters. It is defined in the file flushed before they will be reused, which is a lot of faster - amd_iommu_size= [HW,X86-64] - Define the size of the aperture for the AMD IOMMU - driver. Possible values are: - '32M', '64M' (default), '128M', '256M', '512M', '1G' - amijoy.map= [HW,JOY] Amiga joystick support Map of devices attached to JOY0DAT and JOY1DAT Format: <a>,<b> @@ -646,6 +640,13 @@ and is between 256 and 4096 characters. It is defined in the file DMA-API debugging code disables itself because the architectural default is too low. + dma_debug_driver=<driver_name> + With this option the DMA-API debugging driver + filter feature can be enabled at boot time. Just + pass the driver to filter for as the parameter. + The filter can be disabled or changed to another + driver later using sysfs. + dscc4.setup= [NET] dtc3181e= [HW,SCSI] @@ -752,12 +753,25 @@ and is between 256 and 4096 characters. It is defined in the file ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. ftrace=[tracer] - [ftrace] will set and start the specified tracer + [FTRACE] will set and start the specified tracer as early as possible in order to facilitate early boot debugging. ftrace_dump_on_oops - [ftrace] will dump the trace buffers on oops. + [FTRACE] will dump the trace buffers on oops. + + ftrace_filter=[function-list] + [FTRACE] Limit the functions traced by the function + tracer at boot up. function-list is a comma separated + list of functions. This list can be changed at run + time by the set_ftrace_filter file in the debugfs + tracing directory. + + ftrace_notrace=[function-list] + [FTRACE] Do not trace the functions specified in + function-list. This list can be changed at run time + by the set_ftrace_notrace file in the debugfs + tracing directory. gamecon.map[2|3]= [HW,JOY] Multisystem joystick and NES/SNES/PSX pad @@ -1054,15 +1068,6 @@ and is between 256 and 4096 characters. It is defined in the file use the HighMem zone if it exists, and the Normal zone if it does not. - kmemtrace.enable= [KNL,KMEMTRACE] Format: { yes | no } - Controls whether kmemtrace is enabled - at boot-time. - - kmemtrace.subbufs=n [KNL,KMEMTRACE] Overrides the number of - subbufs kmemtrace's relay channel has. Set this - higher than default (KMEMTRACE_N_SUBBUFS in code) if - you experience buffer overruns. - kgdboc= [HW] kgdb over consoles. Requires a tty driver that supports console polling. (only serial suported for now) @@ -1575,6 +1580,9 @@ and is between 256 and 4096 characters. It is defined in the file noinitrd [RAM] Tells the kernel not to load any configured initial RAM disk. + nointremap [X86-64, Intel-IOMMU] Do not enable interrupt + remapping. + nointroute [IA-64] nojitter [IA64] Disables jitter checking for ITC timers. diff --git a/Documentation/memory-barriers.txt b/Documentation/memory-barriers.txt index f5b7127f54a..7f5809eddee 100644 --- a/Documentation/memory-barriers.txt +++ b/Documentation/memory-barriers.txt @@ -31,6 +31,7 @@ Contents: - Locking functions. - Interrupt disabling functions. + - Sleep and wake-up functions. - Miscellaneous functions. (*) Inter-CPU locking barrier effects. @@ -1217,6 +1218,132 @@ barriers are required in such a situation, they must be provided from some other means. +SLEEP AND WAKE-UP FUNCTIONS +--------------------------- + +Sleeping and waking on an event flagged in global data can be viewed as an +interaction between two pieces of data: the task state of the task waiting for +the event and the global data used to indicate the event. To make sure that +these appear to happen in the right order, the primitives to begin the process +of going to sleep, and the primitives to initiate a wake up imply certain +barriers. + +Firstly, the sleeper normally follows something like this sequence of events: + + for (;;) { + set_current_state(TASK_UNINTERRUPTIBLE); + if (event_indicated) + break; + schedule(); + } + +A general memory barrier is interpolated automatically by set_current_state() +after it has altered the task state: + + CPU 1 + =============================== + set_current_state(); + set_mb(); + STORE current->state + <general barrier> + LOAD event_indicated + +set_current_state() may be wrapped by: + + prepare_to_wait(); + prepare_to_wait_exclusive(); + +which therefore also imply a general memory barrier after setting the state. +The whole sequence above is available in various canned forms, all of which +interpolate the memory barrier in the right place: + + wait_event(); + wait_event_interruptible(); + wait_event_interruptible_exclusive(); + wait_event_interruptible_timeout(); + wait_event_killable(); + wait_event_timeout(); + wait_on_bit(); + wait_on_bit_lock(); + + +Secondly, code that performs a wake up normally follows something like this: + + event_indicated = 1; + wake_up(&event_wait_queue); + +or: + + event_indicated = 1; + wake_up_process(event_daemon); + +A write memory barrier is implied by wake_up() and co. if and only if they wake +something up. The barrier occurs before the task state is cleared, and so sits +between the STORE to indicate the event and the STORE to set TASK_RUNNING: + + CPU 1 CPU 2 + =============================== =============================== + set_current_state(); STORE event_indicated + set_mb(); wake_up(); + STORE current->state <write barrier> + <general barrier> STORE current->state + LOAD event_indicated + +The available waker functions include: + + complete(); + wake_up(); + wake_up_all(); + wake_up_bit(); + wake_up_interruptible(); + wake_up_interruptible_all(); + wake_up_interruptible_nr(); + wake_up_interruptible_poll(); + wake_up_interruptible_sync(); + wake_up_interruptible_sync_poll(); + wake_up_locked(); + wake_up_locked_poll(); + wake_up_nr(); + wake_up_poll(); + wake_up_process(); + + +[!] Note that the memory barriers implied by the sleeper and the waker do _not_ +order multiple stores before the wake-up with respect to loads of those stored +values after the sleeper has called set_current_state(). For instance, if the +sleeper does: + + set_current_state(TASK_INTERRUPTIBLE); + if (event_indicated) + break; + __set_current_state(TASK_RUNNING); + do_something(my_data); + +and the waker does: + + my_data = value; + event_indicated = 1; + wake_up(&event_wait_queue); + +there's no guarantee that the change to event_indicated will be perceived by +the sleeper as coming after the change to my_data. In such a circumstance, the +code on both sides must interpolate its own memory barriers between the +separate data accesses. Thus the above sleeper ought to do: + + set_current_state(TASK_INTERRUPTIBLE); + if (event_indicated) { + smp_rmb(); + do_something(my_data); + } + +and the waker should do: + + my_data = value; + smp_wmb(); + event_indicated = 1; + wake_up(&event_wait_queue); + + MISCELLANEOUS FUNCTIONS ----------------------- @@ -1366,7 +1493,7 @@ WHERE ARE MEMORY BARRIERS NEEDED? Under normal operation, memory operation reordering is generally not going to be a problem as a single-threaded linear piece of code will still appear to -work correctly, even if it's in an SMP kernel. There are, however, three +work correctly, even if it's in an SMP kernel. There are, however, four circumstances in which reordering definitely _could_ be a problem: (*) Interprocessor interaction. diff --git a/Documentation/scheduler/sched-rt-group.txt b/Documentation/scheduler/sched-rt-group.txt index 5ba4d3fc625..1df7f9cdab0 100644 --- a/Documentation/scheduler/sched-rt-group.txt +++ b/Documentation/scheduler/sched-rt-group.txt @@ -4,6 +4,7 @@ CONTENTS ======== +0. WARNING 1. Overview 1.1 The problem 1.2 The solution @@ -14,6 +15,23 @@ CONTENTS 3. Future plans +0. WARNING +========== + + Fiddling with these settings can result in an unstable system, the knobs are + root only and assumes root knows what he is doing. + +Most notable: + + * very small values in sched_rt_period_us can result in an unstable + system when the period is smaller than either the available hrtimer + resolution, or the time it takes to handle the budget refresh itself. + + * very small values in sched_rt_runtime_us can result in an unstable + system when the runtime is so small the system has difficulty making + forward progress (NOTE: the migration thread and kstopmachine both + are real-time processes). + 1. Overview =========== @@ -169,7 +187,7 @@ get their allocated time. Implementing SCHED_EDF might take a while to complete. Priority Inheritance is the biggest challenge as the current linux PI infrastructure is geared towards -the limited static priority levels 0-139. With deadline scheduling you need to +the limited static priority levels 0-99. With deadline scheduling you need to do deadline inheritance (since priority is inversely proportional to the deadline delta (deadline - now). diff --git a/Documentation/trace/events.txt b/Documentation/trace/events.txt new file mode 100644 index 00000000000..f157d7594ea --- /dev/null +++ b/Documentation/trace/events.txt @@ -0,0 +1,90 @@ + Event Tracing + + Documentation written by Theodore Ts'o + Updated by Li Zefan + +1. Introduction +=============== + +Tracepoints (see Documentation/trace/tracepoints.txt) can be used +without creating custom kernel modules to register probe functions +using the event tracing infrastructure. + +Not all tracepoints can be traced using the event tracing system; +the kernel developer must provide code snippets which define how the +tracing information is saved into the tracing buffer, and how the +tracing information should be printed. + +2. Using Event Tracing +====================== + +2.1 Via the 'set_event' interface +--------------------------------- + +The events which are available for tracing can be found in the file +/debug/tracing/available_events. + +To enable a particular event, such as 'sched_wakeup', simply echo it +to /debug/tracing/set_event. For example: + + # echo sched_wakeup >> /debug/tracing/set_event + +[ Note: '>>' is necessary, otherwise it will firstly disable + all the events. ] + +To disable an event, echo the event name to the set_event file prefixed +with an exclamation point: + + # echo '!sched_wakeup' >> /debug/tracing/set_event + +To disable all events, echo an empty line to the set_event file: + + # echo > /debug/tracing/set_event + +To enable all events, echo '*:*' or '*:' to the set_event file: + + # echo *:* > /debug/tracing/set_event + +The events are organized into subsystems, such as ext4, irq, sched, +etc., and a full event name looks like this: <subsystem>:<event>. The +subsystem name is optional, but it is displayed in the available_events +file. All of the events in a subsystem can be specified via the syntax +"<subsystem>:*"; for example, to enable all irq events, you can use the +command: + + # echo 'irq:*' > /debug/tracing/set_event + +2.2 Via the 'enable' toggle +--------------------------- + +The events available are also listed in /debug/tracing/events/ hierarchy +of directories. + +To enable event 'sched_wakeup': + + # echo 1 > /debug/tracing/events/sched/sched_wakeup/enable + +To disable it: + + # echo 0 > /debug/tracing/events/sched/sched_wakeup/enable + +To enable all events in sched subsystem: + + # echo 1 > /debug/tracing/events/sched/enable + +To eanble all events: + + # echo 1 > /debug/tracing/events/enable + +When reading one of these enable files, there are four results: + + 0 - all events this file affects are disabled + 1 - all events this file affects are enabled + X - there is a mixture of events enabled and disabled + ? - this file does not affect any event + +3. Defining an event-enabled tracepoint +======================================= + +See The example provided in samples/trace_events + diff --git a/Documentation/trace/ftrace.txt b/Documentation/trace/ftrace.txt index fd9a3e69381..2a82d860294 100644 --- a/Documentation/trace/ftrace.txt +++ b/Documentation/trace/ftrace.txt @@ -179,7 +179,7 @@ Here is the list of current tracers that may be configured. Function call tracer to trace all kernel functions. - "function_graph_tracer" + "function_graph" Similar to the function tracer except that the function tracer probes the functions on their entry @@ -518,9 +518,18 @@ priority with zero (0) being the highest priority and the nice values starting at 100 (nice -20). Below is a quick chart to map the kernel priority to user land priorities. - Kernel priority: 0 to 99 ==> user RT priority 99 to 0 - Kernel priority: 100 to 139 ==> user nice -20 to 19 - Kernel priority: 140 ==> idle task priority + Kernel Space User Space + =============================================================== + 0(high) to 98(low) user RT priority 99(high) to 1(low) + with SCHED_RR or SCHED_FIFO + --------------------------------------------------------------- + 99 sched_priority is not used in scheduling + decisions(it must be specified as 0) + --------------------------------------------------------------- + 100(high) to 139(low) user nice -20(high) to 19(low) + --------------------------------------------------------------- + 140 idle task priority + --------------------------------------------------------------- The task states are: diff --git a/Documentation/trace/power.txt b/Documentation/trace/power.txt new file mode 100644 index 00000000000..cd805e16dc2 --- /dev/null +++ b/Documentation/trace/power.txt @@ -0,0 +1,17 @@ +The power tracer collects detailed information about C-state and P-state +transitions, instead of just looking at the high-level "average" +information. + +There is a helper script found in scrips/tracing/power.pl in the kernel +sources which can be used to parse this information and create a +Scalable Vector Graphics (SVG) picture from the trace data. + +To use this tracer: + + echo 0 > /sys/kernel/debug/tracing/tracing_enabled + echo power > /sys/kernel/debug/tracing/current_tracer + echo 1 > /sys/kernel/debug/tracing/tracing_enabled + sleep 1 + echo 0 > /sys/kernel/debug/tracing/tracing_enabled + cat /sys/kernel/debug/tracing/trace | \ + perl scripts/tracing/power.pl > out.sv diff --git a/Documentation/x86/boot.txt b/Documentation/x86/boot.txt index e0203662f9e..8da3a795083 100644 --- a/Documentation/x86/boot.txt +++ b/Documentation/x86/boot.txt @@ -50,6 +50,10 @@ Protocol 2.08: (Kernel 2.6.26) Added crc32 checksum and ELF format Protocol 2.09: (Kernel 2.6.26) Added a field of 64-bit physical pointer to single linked list of struct setup_data. +Protocol 2.10: (Kernel 2.6.31) Added a protocol for relaxed alignment + beyond the kernel_alignment added, new init_size and + pref_address fields. Added extended boot loader IDs. + **** MEMORY LAYOUT The traditional memory map for the kernel loader, used for Image or @@ -168,12 +172,13 @@ Offset Proto Name Meaning 021C/4 2.00+ ramdisk_size initrd size (set by boot loader) 0220/4 2.00+ bootsect_kludge DO NOT USE - for bootsect.S use only 0224/2 2.01+ heap_end_ptr Free memory after setup end -0226/2 N/A pad1 Unused +0226/1 2.02+(3 ext_loader_ver Extended boot loader version +0227/1 2.02+(3 ext_loader_type Extended boot loader ID 0228/4 2.02+ cmd_line_ptr 32-bit pointer to the kernel command line 022C/4 2.03+ ramdisk_max Highest legal initrd address 0230/4 2.05+ kernel_alignment Physical addr alignment required for kernel 0234/1 2.05+ relocatable_kernel Whether kernel is relocatable or not -0235/1 N/A pad2 Unused +0235/1 2.10+ min_alignment Minimum alignment, as a power of two 0236/2 N/A pad3 Unused 0238/4 2.06+ cmdline_size Maximum size of the kernel command line 023C/4 2.07+ hardware_subarch Hardware subarchitecture @@ -182,6 +187,8 @@ Offset Proto Name Meaning 024C/4 2.08+ payload_length Length of kernel payload 0250/8 2.09+ setup_data 64-bit physical pointer to linked list of struct setup_data +0258/8 2.10+ pref_address Preferred loading address +0260/4 2.10+ init_size Linear memory required during initialization (1) For backwards compatibility, if the setup_sects field contains 0, the real value is 4. @@ -190,6 +197,8 @@ Offset Proto Name Meaning field are unusable, which means the size of a bzImage kernel cannot be determined. +(3) Ignored, but safe to set, for boot protocols 2.02-2.09. + If the "HdrS" (0x53726448) magic number is not found at offset 0x202, the boot protocol version is "old". Loading an old kernel, the following parameters should be assumed: @@ -343,18 +352,32 @@ Protocol: 2.00+ 0xTV here, where T is an identifier for the boot loader and V is a version number. Otherwise, enter 0xFF here. + For boot loader IDs above T = 0xD, write T = 0xE to this field and + write the extended ID minus 0x10 to the ext_loader_type field. + Similarly, the ext_loader_ver field can be used to provide more than + four bits for the bootloader version. + + For example, for T = 0x15, V = 0x234, write: + + type_of_loader <- 0xE4 + ext_loader_type <- 0x05 + ext_loader_ver <- 0x23 + Assigned boot loader ids: 0 LILO (0x00 reserved for pre-2.00 bootloader) 1 Loadlin 2 bootsect-loader (0x20, all other values reserved) - 3 SYSLINUX - 4 EtherBoot + 3 Syslinux + 4 Etherboot/gPXE 5 ELILO 7 GRUB - 8 U-BOOT + 8 U-Boot 9 Xen A Gujin B Qemu + C Arcturus Networks uCbootloader + E Extended (see ext_loader_type) + F Special (0xFF = undefined) Please contact <hpa@zytor.com> if you need a bootloader ID value assigned. @@ -453,6 +476,35 @@ Protocol: 2.01+ Set this field to the offset (from the beginning of the real-mode code) of the end of the setup stack/heap, minus 0x0200. +Field name: ext_loader_ver +Type: write (optional) +Offset/size: 0x226/1 +Protocol: 2.02+ + + This field is used as an extension of the version number in the + type_of_loader field. The total version number is considered to be + (type_of_loader & 0x0f) + (ext_loader_ver << 4). + + The use of this field is boot loader specific. If not written, it + is zero. + + Kernels prior to 2.6.31 did not recognize this field, but it is safe + to write for protocol version 2.02 or higher. + +Field name: ext_loader_type +Type: write (obligatory if (type_of_loader & 0xf0) == 0xe0) +Offset/size: 0x227/1 +Protocol: 2.02+ + + This field is used as an extension of the type number in + type_of_loader field. If the type in type_of_loader is 0xE, then + the actual type is (ext_loader_type + 0x10). + + This field is ignored if the type in type_of_loader is not 0xE. + + Kernels prior to 2.6.31 did not recognize this field, but it is safe + to write for protocol version 2.02 or higher. + Field name: cmd_line_ptr Type: write (obligatory) Offset/size: 0x228/4 @@ -482,11 +534,19 @@ Protocol: 2.03+ 0x37FFFFFF, you can start your ramdisk at 0x37FE0000.) Field name: kernel_alignment -Type: read (reloc) +Type: read/modify (reloc) Offset/size: 0x230/4 -Protocol: 2.05+ +Protocol: 2.05+ (read), 2.10+ (modify) + + Alignment unit required by the kernel (if relocatable_kernel is + true.) A relocatable kernel that is loaded at an alignment + incompatible with the value in this field will be realigned during + kernel initialization. - Alignment unit required by the kernel (if relocatable_kernel is true.) + Starting with protocol version 2.10, this reflects the kernel + alignment preferred for optimal performance; it is possible for the + loader to modify this field to permit a lesser alignment. See the + min_alignment and pref_address field below. Field name: relocatable_kernel Type: read (reloc) @@ -498,6 +558,22 @@ Protocol: 2.05+ After loading, the boot loader must set the code32_start field to point to the loaded code, or to a boot loader hook. +Field name: min_alignment +Type: read (reloc) +Offset/size: 0x235/1 +Protocol: 2.10+ + + This field, if nonzero, indicates as a power of two the minimum + alignment required, as opposed to preferred, by the kernel to boot. + If a boot loader makes use of this field, it should update the + kernel_alignment field with the alignment unit desired; typically: + + kernel_alignment = 1 << min_alignment + + There may be a considerable performance cost with an excessively + misaligned kernel. Therefore, a loader should typically try each + power-of-two alignment from kernel_alignment down to this alignment. + Field name: cmdline_size Type: read Offset/size: 0x238/4 @@ -582,6 +658,36 @@ Protocol: 2.09+ sure to consider the case where the linked list already contains entries. +Field name: pref_address +Type: read (reloc) +Offset/size: 0x258/8 +Protocol: 2.10+ + + This field, if nonzero, represents a preferred load address for the + kernel. A relocating bootloader should attempt to load at this + address if possible. + + A non-relocatable kernel will unconditionally move itself and to run + at this address. + +Field name: init_size +Type: read +Offset/size: 0x25c/4 + + This field indicates the amount of linear contiguous memory starting + at the kernel runtime start address that the kernel needs before it + is capable of examining its memory map. This is not the same thing + as the total amount of memory the kernel needs to boot, but it can + be used by a relocating boot loader to help select a safe load + address for the kernel. + + The kernel runtime start address is determined by the following algorithm: + + if (relocatable_kernel) + runtime_start = align_up(load_address, kernel_alignment) + else + runtime_start = pref_address + **** THE IMAGE CHECKSUM diff --git a/Documentation/x86/x86_64/boot-options.txt b/Documentation/x86/x86_64/boot-options.txt index 34c13040a71..2db5893d6c9 100644 --- a/Documentation/x86/x86_64/boot-options.txt +++ b/Documentation/x86/x86_64/boot-options.txt @@ -150,11 +150,6 @@ NUMA Otherwise, the remaining system RAM is allocated to an additional node. - numa=hotadd=percent - Only allow hotadd memory to preallocate page structures upto - percent of already available memory. - numa=hotadd=0 will disable hotadd memory. - ACPI acpi=off Don't enable ACPI diff --git a/Documentation/x86/x86_64/mm.txt b/Documentation/x86/x86_64/mm.txt index 29b52b14d0b..d6498e3cd71 100644 --- a/Documentation/x86/x86_64/mm.txt +++ b/Documentation/x86/x86_64/mm.txt @@ -6,10 +6,11 @@ Virtual memory map with 4 level page tables: 0000000000000000 - 00007fffffffffff (=47 bits) user space, different per mm hole caused by [48:63] sign extension ffff800000000000 - ffff80ffffffffff (=40 bits) guard hole -ffff880000000000 - ffffc0ffffffffff (=57 TB) direct mapping of all phys. memory -ffffc10000000000 - ffffc1ffffffffff (=40 bits) hole -ffffc20000000000 - ffffe1ffffffffff (=45 bits) vmalloc/ioremap space -ffffe20000000000 - ffffe2ffffffffff (=40 bits) virtual memory map (1TB) +ffff880000000000 - ffffc7ffffffffff (=64 TB) direct mapping of all phys. memory +ffffc80000000000 - ffffc8ffffffffff (=40 bits) hole +ffffc90000000000 - ffffe8ffffffffff (=45 bits) vmalloc/ioremap space +ffffe90000000000 - ffffe9ffffffffff (=40 bits) hole +ffffea0000000000 - ffffeaffffffffff (=40 bits) virtual memory map (1TB) ... unused hole ... ffffffff80000000 - ffffffffa0000000 (=512 MB) kernel text mapping, from phys 0 ffffffffa0000000 - fffffffffff00000 (=1536 MB) module mapping space diff --git a/MAINTAINERS b/MAINTAINERS index cf4abddfc8a..fd24af2ff32 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4392,6 +4392,16 @@ S: Maintained F: include/linux/delayacct.h F: kernel/delayacct.c +PERFORMANCE COUNTER SUBSYSTEM +P: Peter Zijlstra +M: a.p.zijlstra@chello.nl +P: Paul Mackerras +M: paulus@samba.org +P: Ingo Molnar +M: mingo@elte.hu +L: linux-kernel@vger.kernel.org +S: Supported + PERSONALITY HANDLING P: Christoph Hellwig M: hch@infradead.org diff --git a/arch/alpha/kernel/sys_dp264.c b/arch/alpha/kernel/sys_dp264.c index 9c9d1fd4155..5bd5259324b 100644 --- a/arch/alpha/kernel/sys_dp264.c +++ b/arch/alpha/kernel/sys_dp264.c @@ -176,22 +176,26 @@ cpu_set_irq_affinity(unsigned int irq, cpumask_t affinity) } } -static void +static int dp264_set_affinity(unsigned int irq, const struct cpumask *affinity) { spin_lock(&dp264_irq_lock); cpu_set_irq_affinity(irq, *affinity); tsunami_update_irq_hw(cached_irq_mask); spin_unlock(&dp264_irq_lock); + + return 0; } -static void +static int clipper_set_affinity(unsigned int irq, const struct cpumask *affinity) { spin_lock(&dp264_irq_lock); cpu_set_irq_affinity(irq - 16, *affinity); tsunami_update_irq_hw(cached_irq_mask); spin_unlock(&dp264_irq_lock); + + return 0; } static struct hw_interrupt_type dp264_irq_type = { diff --git a/arch/alpha/kernel/sys_titan.c b/arch/alpha/kernel/sys_titan.c index 27f840a4ad3..8dd239ebdb9 100644 --- a/arch/alpha/kernel/sys_titan.c +++ b/arch/alpha/kernel/sys_titan.c @@ -157,13 +157,15 @@ titan_cpu_set_irq_affinity(unsigned int irq, cpumask_t affinity) } -static void +static int titan_set_irq_affinity(unsigned int irq, const struct cpumask *affinity) { spin_lock(&titan_irq_lock); titan_cpu_set_irq_affinity(irq - 16, *affinity); titan_update_irq_hw(titan_cached_irq_mask); spin_unlock(&titan_irq_lock); + + return 0; } static void diff --git a/arch/arm/common/gic.c b/arch/arm/common/gic.c index 3e1714c6523..664c7b8b1ba 100644 --- a/arch/arm/common/gic.c +++ b/arch/arm/common/gic.c @@ -109,7 +109,7 @@ static void gic_unmask_irq(unsigned int irq) } #ifdef CONFIG_SMP -static void gic_set_cpu(unsigned int irq, const struct cpumask *mask_val) +static int gic_set_cpu(unsigned int irq, const struct cpumask *mask_val) { void __iomem *reg = gic_dist_base(irq) + GIC_DIST_TARGET + (gic_irq(irq) & ~3); unsigned int shift = (irq % 4) * 8; @@ -122,6 +122,8 @@ static void gic_set_cpu(unsigned int irq, const struct cpumask *mask_val) val |= 1 << (cpu + shift); writel(val, reg); spin_unlock(&irq_controller_lock); + + return 0; } #endif diff --git a/arch/cris/arch-v32/kernel/irq.c b/arch/cris/arch-v32/kernel/irq.c index df3925cb1c7..d70b445f4a8 100644 --- a/arch/cris/arch-v32/kernel/irq.c +++ b/arch/cris/arch-v32/kernel/irq.c @@ -325,12 +325,14 @@ static void end_crisv32_irq(unsigned int irq) { } -void set_affinity_crisv32_irq(unsigned int irq, const struct cpumask *dest) +int set_affinity_crisv32_irq(unsigned int irq, const struct cpumask *dest) { unsigned long flags; spin_lock_irqsave(&irq_lock, flags); irq_allocations[irq - FIRST_IRQ].mask = *dest; spin_unlock_irqrestore(&irq_lock, flags); + + return 0; } static struct irq_chip crisv32_irq_type = { diff --git a/arch/ia64/hp/sim/hpsim_irq.c b/arch/ia64/hp/sim/hpsim_irq.c index cc0a3182db3..acb5047ab57 100644 --- a/arch/ia64/hp/sim/hpsim_irq.c +++ b/arch/ia64/hp/sim/hpsim_irq.c @@ -21,9 +21,10 @@ hpsim_irq_noop (unsigned int irq) { } -static void +static int hpsim_set_affinity_noop(unsigned int a, const struct cpumask *b) { + return 0; } static struct hw_interrupt_type irq_type_hp_sim = { diff --git a/arch/ia64/kernel/acpi.c b/arch/ia64/kernel/acpi.c index 5510317db37..baec6f00f7f 100644 --- a/arch/ia64/kernel/acpi.c +++ b/arch/ia64/kernel/acpi.c @@ -636,7 +636,7 @@ void __init acpi_numa_arch_fixup(void) * success: return IRQ number (>=0) * failure: return < 0 */ -int acpi_register_gsi(u32 gsi, int triggering, int polarity) +int acpi_register_gsi(struct device *dev, u32 gsi, int triggering, int polarity) { if (acpi_irq_model == ACPI_IRQ_MODEL_PLATFORM) return gsi; @@ -678,7 +678,8 @@ static int __init acpi_parse_fadt(struct acpi_table_header *table) fadt = (struct acpi_table_fadt *)fadt_header; - acpi_register_gsi(fadt->sci_interrupt, ACPI_LEVEL_SENSITIVE, ACPI_ACTIVE_LOW); + acpi_register_gsi(NULL, fadt->sci_interrupt, ACPI_LEVEL_SENSITIVE, + ACPI_ACTIVE_LOW); return 0; } diff --git a/arch/ia64/kernel/iosapic.c b/arch/ia64/kernel/iosapic.c index 166e0d839fa..f92cef47bf8 100644 --- a/arch/ia64/kernel/iosapic.c +++ b/arch/ia64/kernel/iosapic.c @@ -329,7 +329,7 @@ unmask_irq (unsigned int irq) } -static void +static int iosapic_set_affinity(unsigned int irq, const struct cpumask *mask) { #ifdef CONFIG_SMP @@ -343,15 +343,15 @@ iosapic_set_affinity(unsigned int irq, const struct cpumask *mask) cpu = cpumask_first_and(cpu_online_mask, mask); if (cpu >= nr_cpu_ids) - return; + return -1; if (irq_prepare_move(irq, cpu)) - return; + return -1; dest = cpu_physical_id(cpu); if (!iosapic_intr_info[irq].count) - return; /* not an IOSAPIC interrupt */ + return -1; /* not an IOSAPIC interrupt */ set_irq_affinity_info(irq, dest, redir); @@ -376,7 +376,9 @@ iosapic_set_affinity(unsigned int irq, const struct cpumask *mask) iosapic_write(iosapic, IOSAPIC_RTE_HIGH(rte_index), high32); iosapic_write(iosapic, IOSAPIC_RTE_LOW(rte_index), low32); } + #endif + return 0; } /* diff --git a/arch/ia64/kernel/msi_ia64.c b/arch/ia64/kernel/msi_ia64.c index 2b15e233f7f..0f8ade9331b 100644 --- a/arch/ia64/kernel/msi_ia64.c +++ b/arch/ia64/kernel/msi_ia64.c @@ -12,7 +12,7 @@ static struct irq_chip ia64_msi_chip; #ifdef CONFIG_SMP -static void ia64_set_msi_irq_affinity(unsigned int irq, +static int ia64_set_msi_irq_affinity(unsigned int irq, const cpumask_t *cpu_mask) { struct msi_msg msg; @@ -20,10 +20,10 @@ static void ia64_set_msi_irq_affinity(unsigned int irq, int cpu = first_cpu(*cpu_mask); if (!cpu_online(cpu)) - return; + return -1; if (irq_prepare_move(irq, cpu)) - return; + return -1; read_msi_msg(irq, &msg); @@ -39,6 +39,8 @@ static void ia64_set_msi_irq_affinity(unsigned int irq, write_msi_msg(irq, &msg); cpumask_copy(irq_desc[irq].affinity, cpumask_of(cpu)); + + return 0; } #endif /* CONFIG_SMP */ @@ -130,17 +132,17 @@ void arch_teardown_msi_irq(unsigned int irq) #ifdef CONFIG_DMAR #ifdef CONFIG_SMP -static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) +static int dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) { struct irq_cfg *cfg = irq_cfg + irq; struct msi_msg msg; int cpu = cpumask_first(mask); if (!cpu_online(cpu)) - return; + return -1; if (irq_prepare_move(irq, cpu)) - return; + return -1; dmar_msi_read(irq, &msg); @@ -151,6 +153,8 @@ static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) dmar_msi_write(irq, &msg); cpumask_copy(irq_desc[irq].affinity, mask); + + return 0; } #endif /* CONFIG_SMP */ diff --git a/arch/ia64/sn/kernel/irq.c b/arch/ia64/sn/kernel/irq.c index 66fd705e82c..764f26abac0 100644 --- a/arch/ia64/sn/kernel/irq.c +++ b/arch/ia64/sn/kernel/irq.c @@ -227,7 +227,7 @@ finish_up: return new_irq_info; } -static void sn_set_affinity_irq(unsigned int irq, const struct cpumask *mask) +static int sn_set_affinity_irq(unsigned int irq, const struct cpumask *mask) { struct sn_irq_info *sn_irq_info, *sn_irq_info_safe; nasid_t nasid; @@ -239,6 +239,8 @@ static void sn_set_affinity_irq(unsigned int irq, const struct cpumask *mask) list_for_each_entry_safe(sn_irq_info, sn_irq_info_safe, sn_irq_lh[irq], list) (void)sn_retarget_vector(sn_irq_info, nasid, slice); + + return 0; } #ifdef CONFIG_SMP diff --git a/arch/ia64/sn/kernel/msi_sn.c b/arch/ia64/sn/kernel/msi_sn.c index 81e428943d7..fbbfb970120 100644 --- a/arch/ia64/sn/kernel/msi_sn.c +++ b/arch/ia64/sn/kernel/msi_sn.c @@ -151,7 +151,7 @@ int sn_setup_msi_irq(struct pci_dev *pdev, struct msi_desc *entry) } #ifdef CONFIG_SMP -static void sn_set_msi_irq_affinity(unsigned int irq, +static int sn_set_msi_irq_affinity(unsigned int irq, const struct cpumask *cpu_mask) { struct msi_msg msg; @@ -168,7 +168,7 @@ static void sn_set_msi_irq_affinity(unsigned int irq, cpu = cpumask_first(cpu_mask); sn_irq_info = sn_msi_info[irq].sn_irq_info; if (sn_irq_info == NULL || sn_irq_info->irq_int_bit >= 0) - return; + return -1; /* * Release XIO resources for the old MSI PCI address @@ -189,7 +189,7 @@ static void sn_set_msi_irq_affinity(unsigned int irq, new_irq_info = sn_retarget_vector(sn_irq_info, nasid, slice); sn_msi_info[irq].sn_irq_info = new_irq_info; if (new_irq_info == NULL) - return; + return -1; /* * Map the xio address into bus space @@ -206,6 +206,8 @@ static void sn_set_msi_irq_affinity(unsigned int irq, write_msi_msg(irq, &msg); cpumask_copy(irq_desc[irq].affinity, cpu_mask); + + return 0; } #endif /* CONFIG_SMP */ diff --git a/arch/mips/cavium-octeon/octeon-irq.c b/arch/mips/cavium-octeon/octeon-irq.c index 1c19af8daa6..d3a0c8154be 100644 --- a/arch/mips/cavium-octeon/octeon-irq.c +++ b/arch/mips/cavium-octeon/octeon-irq.c @@ -177,7 +177,7 @@ static void octeon_irq_ciu0_disable(unsigned int irq) } #ifdef CONFIG_SMP -static void octeon_irq_ciu0_set_affinity(unsigned int irq, const struct cpumask *dest) +static int octeon_irq_ciu0_set_affinity(unsigned int irq, const struct cpumask *dest) { int cpu; int bit = irq - OCTEON_IRQ_WORKQ0; /* Bit 0-63 of EN0 */ @@ -199,6 +199,8 @@ static void octeon_irq_ciu0_set_affinity(unsigned int irq, const struct cpumask */ cvmx_read_csr(CVMX_CIU_INTX_EN0(cvmx_get_core_num() * 2)); write_unlock(&octeon_irq_ciu0_rwlock); + + return 0; } #endif @@ -292,7 +294,7 @@ static void octeon_irq_ciu1_disable(unsigned int irq) } #ifdef CONFIG_SMP -static void octeon_irq_ciu1_set_affinity(unsigned int irq, const struct cpumask *dest) +static int octeon_irq_ciu1_set_affinity(unsigned int irq, const struct cpumask *dest) { int cpu; int bit = irq - OCTEON_IRQ_WDOG0; /* Bit 0-63 of EN1 */ @@ -315,6 +317,8 @@ static void octeon_irq_ciu1_set_affinity(unsigned int irq, const struct cpumask */ cvmx_read_csr(CVMX_CIU_INTX_EN1(cvmx_get_core_num() * 2 + 1)); write_unlock(&octeon_irq_ciu1_rwlock); + + return 0; } #endif diff --git a/arch/mips/include/asm/irq.h b/arch/mips/include/asm/irq.h index 3214ade02d1..4f1eed107b0 100644 --- a/arch/mips/include/asm/irq.h +++ b/arch/mips/include/asm/irq.h @@ -49,7 +49,7 @@ static inline void smtc_im_ack_irq(unsigned int irq) #ifdef CONFIG_MIPS_MT_SMTC_IRQAFF #include <linux/cpumask.h> -extern void plat_set_irq_affinity(unsigned int irq, +extern int plat_set_irq_affinity(unsigned int irq, const struct cpumask *affinity); extern void smtc_forward_irq(unsigned int irq); diff --git a/arch/mips/kernel/irq-gic.c b/arch/mips/kernel/irq-gic.c index 87deb8f6c45..3f43c2e3aa5 100644 --- a/arch/mips/kernel/irq-gic.c +++ b/arch/mips/kernel/irq-gic.c @@ -155,7 +155,7 @@ static void gic_unmask_irq(unsigned int irq) static DEFINE_SPINLOCK(gic_lock); -static void gic_set_affinity(unsigned int irq, const struct cpumask *cpumask) +static int gic_set_affinity(unsigned int irq, const struct cpumask *cpumask) { cpumask_t tmp = CPU_MASK_NONE; unsigned long flags; @@ -166,7 +166,7 @@ static void gic_set_affinity(unsigned int irq, const struct cpumask *cpumask) cpumask_and(&tmp, cpumask, cpu_online_mask); if (cpus_empty(tmp)) - return; + return -1; /* Assumption : cpumask refers to a single CPU */ spin_lock_irqsave(&gic_lock, flags); @@ -190,6 +190,7 @@ static void gic_set_affinity(unsigned int irq, const struct cpumask *cpumask) cpumask_copy(irq_desc[irq].affinity, cpumask); spin_unlock_irqrestore(&gic_lock, flags); + return 0; } #endif diff --git a/arch/mips/mti-malta/malta-smtc.c b/arch/mips/mti-malta/malta-smtc.c index 5ba31888fef..499ffe5475d 100644 --- a/arch/mips/mti-malta/malta-smtc.c +++ b/arch/mips/mti-malta/malta-smtc.c @@ -114,7 +114,7 @@ struct plat_smp_ops msmtc_smp_ops = { */ -void plat_set_irq_affinity(unsigned int irq, const struct cpumask *affinity) +int plat_set_irq_affinity(unsigned int irq, const struct cpumask *affinity) { cpumask_t tmask; int cpu = 0; @@ -156,5 +156,7 @@ void plat_set_irq_affinity(unsigned int irq, const struct cpumask *affinity) /* Do any generic SMTC IRQ affinity setup */ smtc_set_irq_affinity(irq, tmask); + + return 0; } #endif /* CONFIG_MIPS_MT_SMTC_IRQAFF */ diff --git a/arch/mips/sibyte/bcm1480/irq.c b/arch/mips/sibyte/bcm1480/irq.c index c147c4b35d3..690de06bde9 100644 --- a/arch/mips/sibyte/bcm1480/irq.c +++ b/arch/mips/sibyte/bcm1480/irq.c @@ -50,7 +50,7 @@ static void enable_bcm1480_irq(unsigned int irq); static void disable_bcm1480_irq(unsigned int irq); static void ack_bcm1480_irq(unsigned int irq); #ifdef CONFIG_SMP -static void bcm1480_set_affinity(unsigned int irq, const struct cpumask *mask); +static int bcm1480_set_affinity(unsigned int irq, const struct cpumask *mask); #endif #ifdef CONFIG_PCI @@ -109,7 +109,7 @@ void bcm1480_unmask_irq(int cpu, int irq) } #ifdef CONFIG_SMP -static void bcm1480_set_affinity(unsigned int irq, const struct cpumask *mask) +static int bcm1480_set_affinity(unsigned int irq, const struct cpumask *mask) { int i = 0, old_cpu, cpu, int_on, k; u64 cur_ints; @@ -118,7 +118,7 @@ static void bcm1480_set_affinity(unsigned int irq, const struct cpumask *mask) if (cpumask_weight(mask) != 1) { printk("attempted to set irq affinity for irq %d to multiple CPUs\n", irq); - return; + return -1; } i = cpumask_first(mask); @@ -152,6 +152,8 @@ static void bcm1480_set_affinity(unsigned int irq, const struct cpumask *mask) } } spin_unlock_irqrestore(&bcm1480_imr_lock, flags); + + return 0; } #endif diff --git a/arch/mips/sibyte/sb1250/irq.c b/arch/mips/sibyte/sb1250/irq.c index 38cb998ade2..409dec79886 100644 --- a/arch/mips/sibyte/sb1250/irq.c +++ b/arch/mips/sibyte/sb1250/irq.c @@ -50,7 +50,7 @@ static void enable_sb1250_irq(unsigned int irq); static void disable_sb1250_irq(unsigned int irq); static void ack_sb1250_irq(unsigned int irq); #ifdef CONFIG_SMP -static void sb1250_set_affinity(unsigned int irq, const struct cpumask *mask); +static int sb1250_set_affinity(unsigned int irq, const struct cpumask *mask); #endif #ifdef CONFIG_SIBYTE_HAS_LDT @@ -103,7 +103,7 @@ void sb1250_unmask_irq(int cpu, int irq) } #ifdef CONFIG_SMP -static void sb1250_set_affinity(unsigned int irq, const struct cpumask *mask) +static int sb1250_set_affinity(unsigned int irq, const struct cpumask *mask) { int i = 0, old_cpu, cpu, int_on; u64 cur_ints; @@ -113,7 +113,7 @@ static void sb1250_set_affinity(unsigned int irq, const struct cpumask *mask) if (cpumask_weight(mask) > 1) { printk("attempted to set irq affinity for irq %d to multiple CPUs\n", irq); - return; + return -1; } /* Convert logical CPU to physical CPU */ @@ -143,6 +143,8 @@ static void sb1250_set_affinity(unsigned int irq, const struct cpumask *mask) R_IMR_INTERRUPT_MASK)); } spin_unlock_irqrestore(&sb1250_imr_lock, flags); + + return 0; } #endif diff --git a/arch/parisc/kernel/irq.c b/arch/parisc/kernel/irq.c index 4ea4229d765..8007f1e6572 100644 --- a/arch/parisc/kernel/irq.c +++ b/arch/parisc/kernel/irq.c @@ -130,15 +130,17 @@ int cpu_check_affinity(unsigned int irq, const struct cpumask *dest) return cpu_dest; } -static void cpu_set_affinity_irq(unsigned int irq, const struct cpumask *dest) +static int cpu_set_affinity_irq(unsigned int irq, const struct cpumask *dest) { int cpu_dest; cpu_dest = cpu_check_affinity(irq, dest); if (cpu_dest < 0) - return; + return -1; cpumask_copy(&irq_desc[irq].affinity, dest); + + return 0; } #endif diff --git a/arch/powerpc/include/asm/hw_irq.h b/arch/powerpc/include/asm/hw_irq.h index b7e034b0a6d..20a44d0c9fd 100644 --- a/arch/powerpc/include/asm/hw_irq.h +++ b/arch/powerpc/include/asm/hw_irq.h @@ -131,5 +131,44 @@ static inline int irqs_disabled_flags(unsigned long flags) */ struct irq_chip; +#ifdef CONFIG_PERF_COUNTERS +static inline unsigned long test_perf_counter_pending(void) +{ + unsigned long x; + + asm volatile("lbz %0,%1(13)" + : "=r" (x) + : "i" (offsetof(struct paca_struct, perf_counter_pending))); + return x; +} + +static inline void set_perf_counter_pending(void) +{ + asm volatile("stb %0,%1(13)" : : + "r" (1), + "i" (offsetof(struct paca_struct, perf_counter_pending))); +} + +static inline void clear_perf_counter_pending(void) +{ + asm volatile("stb %0,%1(13)" : : + "r" (0), + "i" (offsetof(struct paca_struct, perf_counter_pending))); +} + +extern void perf_counter_do_pending(void); + +#else + +static inline unsigned long test_perf_counter_pending(void) +{ + return 0; +} + +static inline void set_perf_counter_pending(void) {} +static inline void clear_perf_counter_pending(void) {} +static inline void perf_counter_do_pending(void) {} +#endif /* CONFIG_PERF_COUNTERS */ + #endif /* __KERNEL__ */ #endif /* _ASM_POWERPC_HW_IRQ_H */ diff --git a/arch/powerpc/include/asm/paca.h b/arch/powerpc/include/asm/paca.h index 082b3aedf14..6ef05572301 100644 --- a/arch/powerpc/include/asm/paca.h +++ b/arch/powerpc/include/asm/paca.h @@ -99,6 +99,7 @@ struct paca_struct { u8 soft_enabled; /* irq soft-enable flag */ u8 hard_enabled; /* set if irqs are enabled in MSR */ u8 io_sync; /* writel() needs spin_unlock sync */ + u8 perf_counter_pending; /* PM interrupt while soft-disabled */ /* Stuff for accurate time accounting */ u64 user_time; /* accumulated usermode TB ticks */ diff --git a/arch/powerpc/include/asm/perf_counter.h b/arch/powerpc/include/asm/perf_counter.h new file mode 100644 index 00000000000..cc7c887705b --- /dev/null +++ b/arch/powerpc/include/asm/perf_counter.h @@ -0,0 +1,98 @@ +/* + * Performance counter support - PowerPC-specific definitions. + * + * Copyright 2008-2009 Paul Mackerras, IBM Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/types.h> + +#define MAX_HWCOUNTERS 8 +#define MAX_EVENT_ALTERNATIVES 8 +#define MAX_LIMITED_HWCOUNTERS 2 + +/* + * This struct provides the constants and functions needed to + * describe the PMU on a particular POWER-family CPU. + */ +struct power_pmu { + int n_counter; + int max_alternatives; + u64 add_fields; + u64 test_adder; + int (*compute_mmcr)(u64 events[], int n_ev, + unsigned int hwc[], u64 mmcr[]); + int (*get_constraint)(u64 event, u64 *mskp, u64 *valp); + int (*get_alternatives)(u64 event, unsigned int flags, + u64 alt[]); + void (*disable_pmc)(unsigned int pmc, u64 mmcr[]); + int (*limited_pmc_event)(u64 event); + u32 flags; + int n_generic; + int *generic_events; + int (*cache_events)[PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX]; +}; + +extern struct power_pmu *ppmu; + +/* + * Values for power_pmu.flags + */ +#define PPMU_LIMITED_PMC5_6 1 /* PMC5/6 have limited function */ +#define PPMU_ALT_SIPR 2 /* uses alternate posn for SIPR/HV */ + +/* + * Values for flags to get_alternatives() + */ +#define PPMU_LIMITED_PMC_OK 1 /* can put this on a limited PMC */ +#define PPMU_LIMITED_PMC_REQD 2 /* have to put this on a limited PMC */ +#define PPMU_ONLY_COUNT_RUN 4 /* only counting in run state */ + +struct pt_regs; +extern unsigned long perf_misc_flags(struct pt_regs *regs); +#define perf_misc_flags(regs) perf_misc_flags(regs) + +extern unsigned long perf_instruction_pointer(struct pt_regs *regs); + +/* + * The power_pmu.get_constraint function returns a 64-bit value and + * a 64-bit mask that express the constraints between this event and + * other events. + * + * The value and mask are divided up into (non-overlapping) bitfields + * of three different types: + * + * Select field: this expresses the constraint that some set of bits + * in MMCR* needs to be set to a specific value for this event. For a + * select field, the mask contains 1s in every bit of the field, and + * the value contains a unique value for each possible setting of the + * MMCR* bits. The constraint checking code will ensure that two events + * that set the same field in their masks have the same value in their + * value dwords. + * + * Add field: this expresses the constraint that there can be at most + * N events in a particular class. A field of k bits can be used for + * N <= 2^(k-1) - 1. The mask has the most significant bit of the field + * set (and the other bits 0), and the value has only the least significant + * bit of the field set. In addition, the 'add_fields' and 'test_adder' + * in the struct power_pmu for this processor come into play. The + * add_fields value contains 1 in the LSB of the field, and the + * test_adder contains 2^(k-1) - 1 - N in the field. + * + * NAND field: this expresses the constraint that you may not have events + * in all of a set of classes. (For example, on PPC970, you can't select + * events from the FPU, ISU and IDU simultaneously, although any two are + * possible.) For N classes, the field is N+1 bits wide, and each class + * is assigned one bit from the least-significant N bits. The mask has + * only the most-significant bit set, and the value has only the bit + * for the event's class set. The test_adder has the least significant + * bit set in the field. + * + * If an event is not subject to the constraint expressed by a particular + * field, then it will have 0 in both the mask and value for that field. + */ diff --git a/arch/powerpc/include/asm/reg.h b/arch/powerpc/include/asm/reg.h index e8018d540e8..fb359b0a693 100644 --- a/arch/powerpc/include/asm/reg.h +++ b/arch/powerpc/include/asm/reg.h @@ -492,11 +492,13 @@ #define MMCR0_FCHV 0x00000001UL /* freeze conditions in hypervisor mode */ #define SPRN_MMCR1 798 #define SPRN_MMCRA 0x312 +#define MMCRA_SDSYNC 0x80000000UL /* SDAR synced with SIAR */ #define MMCRA_SIHV 0x10000000UL /* state of MSR HV when SIAR set */ #define MMCRA_SIPR 0x08000000UL /* state of MSR PR when SIAR set */ #define MMCRA_SLOT 0x07000000UL /* SLOT bits (37-39) */ #define MMCRA_SLOT_SHIFT 24 #define MMCRA_SAMPLE_ENABLE 0x00000001UL /* enable sampling */ +#define POWER6_MMCRA_SDSYNC 0x0000080000000000ULL /* SDAR/SIAR synced */ #define POWER6_MMCRA_SIHV 0x0000040000000000ULL #define POWER6_MMCRA_SIPR 0x0000020000000000ULL #define POWER6_MMCRA_THRM 0x00000020UL diff --git a/arch/powerpc/include/asm/systbl.h b/arch/powerpc/include/asm/systbl.h index d98a30dfd41..a0b92de51c7 100644 --- a/arch/powerpc/include/asm/systbl.h +++ b/arch/powerpc/include/asm/systbl.h @@ -322,6 +322,6 @@ SYSCALL_SPU(epoll_create1) SYSCALL_SPU(dup3) SYSCALL_SPU(pipe2) SYSCALL(inotify_init1) -SYSCALL(ni_syscall) +SYSCALL_SPU(perf_counter_open) COMPAT_SYS_SPU(preadv) COMPAT_SYS_SPU(pwritev) diff --git a/arch/powerpc/include/asm/unistd.h b/arch/powerpc/include/asm/unistd.h index 3f06f8ec81c..4badac2d11d 100644 --- a/arch/powerpc/include/asm/unistd.h +++ b/arch/powerpc/include/asm/unistd.h @@ -341,6 +341,7 @@ #define __NR_dup3 316 #define __NR_pipe2 317 #define __NR_inotify_init1 318 +#define __NR_perf_counter_open 319 #define __NR_preadv 320 #define __NR_pwritev 321 diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile index 71901fbda4a..a2c683403c2 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -94,6 +94,9 @@ obj64-$(CONFIG_AUDIT) += compat_audit.o obj-$(CONFIG_DYNAMIC_FTRACE) += ftrace.o obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o +obj-$(CONFIG_PERF_COUNTERS) += perf_counter.o power4-pmu.o ppc970-pmu.o \ + power5-pmu.o power5+-pmu.o power6-pmu.o \ + power7-pmu.o obj-$(CONFIG_8XX_MINIMAL_FPEMU) += softemu8xx.o diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c index 1e40bc05394..e981d1ce191 100644 --- a/arch/powerpc/kernel/asm-offsets.c +++ b/arch/powerpc/kernel/asm-offsets.c @@ -131,6 +131,7 @@ int main(void) DEFINE(PACAKMSR, offsetof(struct paca_struct, kernel_msr)); DEFINE(PACASOFTIRQEN, offsetof(struct paca_struct, soft_enabled)); DEFINE(PACAHARDIRQEN, offsetof(struct paca_struct, hard_enabled)); + DEFINE(PACAPERFPEND, offsetof(struct paca_struct, perf_counter_pending)); DEFINE(PACASLBCACHE, offsetof(struct paca_struct, slb_cache)); DEFINE(PACASLBCACHEPTR, offsetof(struct paca_struct, slb_cache_ptr)); DEFINE(PACACONTEXTID, offsetof(struct paca_struct, context.id)); diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S index abfc3233047..43e073477c3 100644 --- a/arch/powerpc/kernel/entry_64.S +++ b/arch/powerpc/kernel/entry_64.S @@ -526,6 +526,15 @@ ALT_FW_FTR_SECTION_END_IFCLR(FW_FEATURE_ISERIES) 2: TRACE_AND_RESTORE_IRQ(r5); +#ifdef CONFIG_PERF_COUNTERS + /* check paca->perf_counter_pending if we're enabling ints */ + lbz r3,PACAPERFPEND(r13) + and. r3,r3,r5 + beq 27f + bl .perf_counter_do_pending +27: +#endif /* CONFIG_PERF_COUNTERS */ + /* extract EE bit and use it to restore paca->hard_enabled */ ld r3,_MSR(r1) rldicl r4,r3,49,63 /* r0 = (r3 >> 15) & 1 */ diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c index 8c1a4966867..feff792ed0f 100644 --- a/arch/powerpc/kernel/irq.c +++ b/arch/powerpc/kernel/irq.c @@ -135,6 +135,11 @@ notrace void raw_local_irq_restore(unsigned long en) iseries_handle_interrupts(); } + if (test_perf_counter_pending()) { + clear_perf_counter_pending(); + perf_counter_do_pending(); + } + /* * if (get_paca()->hard_enabled) return; * But again we need to take care that gcc gets hard_enabled directly diff --git a/arch/powerpc/kernel/perf_counter.c b/arch/powerpc/kernel/perf_counter.c new file mode 100644 index 00000000000..bb202388170 --- /dev/null +++ b/arch/powerpc/kernel/perf_counter.c @@ -0,0 +1,1263 @@ +/* + * Performance counter support - powerpc architecture code + * + * Copyright 2008-2009 Paul Mackerras, IBM Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/perf_counter.h> +#include <linux/percpu.h> +#include <linux/hardirq.h> +#include <asm/reg.h> +#include <asm/pmc.h> +#include <asm/machdep.h> +#include <asm/firmware.h> +#include <asm/ptrace.h> + +struct cpu_hw_counters { + int n_counters; + int n_percpu; + int disabled; + int n_added; + int n_limited; + u8 pmcs_enabled; + struct perf_counter *counter[MAX_HWCOUNTERS]; + u64 events[MAX_HWCOUNTERS]; + unsigned int flags[MAX_HWCOUNTERS]; + u64 mmcr[3]; + struct perf_counter *limited_counter[MAX_LIMITED_HWCOUNTERS]; + u8 limited_hwidx[MAX_LIMITED_HWCOUNTERS]; +}; +DEFINE_PER_CPU(struct cpu_hw_counters, cpu_hw_counters); + +struct power_pmu *ppmu; + +/* + * Normally, to ignore kernel events we set the FCS (freeze counters + * in supervisor mode) bit in MMCR0, but if the kernel runs with the + * hypervisor bit set in the MSR, or if we are running on a processor + * where the hypervisor bit is forced to 1 (as on Apple G5 processors), + * then we need to use the FCHV bit to ignore kernel events. + */ +static unsigned int freeze_counters_kernel = MMCR0_FCS; + +static void perf_counter_interrupt(struct pt_regs *regs); + +void perf_counter_print_debug(void) +{ +} + +/* + * Read one performance monitor counter (PMC). + */ +static unsigned long read_pmc(int idx) +{ + unsigned long val; + + switch (idx) { + case 1: + val = mfspr(SPRN_PMC1); + break; + case 2: + val = mfspr(SPRN_PMC2); + break; + case 3: + val = mfspr(SPRN_PMC3); + break; + case 4: + val = mfspr(SPRN_PMC4); + break; + case 5: + val = mfspr(SPRN_PMC5); + break; + case 6: + val = mfspr(SPRN_PMC6); + break; + case 7: + val = mfspr(SPRN_PMC7); + break; + case 8: + val = mfspr(SPRN_PMC8); + break; + default: + printk(KERN_ERR "oops trying to read PMC%d\n", idx); + val = 0; + } + return val; +} + +/* + * Write one PMC. + */ +static void write_pmc(int idx, unsigned long val) +{ + switch (idx) { + case 1: + mtspr(SPRN_PMC1, val); + break; + case 2: + mtspr(SPRN_PMC2, val); + break; + case 3: + mtspr(SPRN_PMC3, val); + break; + case 4: + mtspr(SPRN_PMC4, val); + break; + case 5: + mtspr(SPRN_PMC5, val); + break; + case 6: + mtspr(SPRN_PMC6, val); + break; + case 7: + mtspr(SPRN_PMC7, val); + break; + case 8: + mtspr(SPRN_PMC8, val); + break; + default: + printk(KERN_ERR "oops trying to write PMC%d\n", idx); + } +} + +/* + * Check if a set of events can all go on the PMU at once. + * If they can't, this will look at alternative codes for the events + * and see if any combination of alternative codes is feasible. + * The feasible set is returned in event[]. + */ +static int power_check_constraints(u64 event[], unsigned int cflags[], + int n_ev) +{ + u64 mask, value, nv; + u64 alternatives[MAX_HWCOUNTERS][MAX_EVENT_ALTERNATIVES]; + u64 amasks[MAX_HWCOUNTERS][MAX_EVENT_ALTERNATIVES]; + u64 avalues[MAX_HWCOUNTERS][MAX_EVENT_ALTERNATIVES]; + u64 smasks[MAX_HWCOUNTERS], svalues[MAX_HWCOUNTERS]; + int n_alt[MAX_HWCOUNTERS], choice[MAX_HWCOUNTERS]; + int i, j; + u64 addf = ppmu->add_fields; + u64 tadd = ppmu->test_adder; + + if (n_ev > ppmu->n_counter) + return -1; + + /* First see if the events will go on as-is */ + for (i = 0; i < n_ev; ++i) { + if ((cflags[i] & PPMU_LIMITED_PMC_REQD) + && !ppmu->limited_pmc_event(event[i])) { + ppmu->get_alternatives(event[i], cflags[i], + alternatives[i]); + event[i] = alternatives[i][0]; + } + if (ppmu->get_constraint(event[i], &amasks[i][0], + &avalues[i][0])) + return -1; + } + value = mask = 0; + for (i = 0; i < n_ev; ++i) { + nv = (value | avalues[i][0]) + (value & avalues[i][0] & addf); + if ((((nv + tadd) ^ value) & mask) != 0 || + (((nv + tadd) ^ avalues[i][0]) & amasks[i][0]) != 0) + break; + value = nv; + mask |= amasks[i][0]; + } + if (i == n_ev) + return 0; /* all OK */ + + /* doesn't work, gather alternatives... */ + if (!ppmu->get_alternatives) + return -1; + for (i = 0; i < n_ev; ++i) { + choice[i] = 0; + n_alt[i] = ppmu->get_alternatives(event[i], cflags[i], + alternatives[i]); + for (j = 1; j < n_alt[i]; ++j) + ppmu->get_constraint(alternatives[i][j], + &amasks[i][j], &avalues[i][j]); + } + + /* enumerate all possibilities and see if any will work */ + i = 0; + j = -1; + value = mask = nv = 0; + while (i < n_ev) { + if (j >= 0) { + /* we're backtracking, restore context */ + value = svalues[i]; + mask = smasks[i]; + j = choice[i]; + } + /* + * See if any alternative k for event i, + * where k > j, will satisfy the constraints. + */ + while (++j < n_alt[i]) { + nv = (value | avalues[i][j]) + + (value & avalues[i][j] & addf); + if ((((nv + tadd) ^ value) & mask) == 0 && + (((nv + tadd) ^ avalues[i][j]) + & amasks[i][j]) == 0) + break; + } + if (j >= n_alt[i]) { + /* + * No feasible alternative, backtrack + * to event i-1 and continue enumerating its + * alternatives from where we got up to. + */ + if (--i < 0) + return -1; + } else { + /* + * Found a feasible alternative for event i, + * remember where we got up to with this event, + * go on to the next event, and start with + * the first alternative for it. + */ + choice[i] = j; + svalues[i] = value; + smasks[i] = mask; + value = nv; + mask |= amasks[i][j]; + ++i; + j = -1; + } + } + + /* OK, we have a feasible combination, tell the caller the solution */ + for (i = 0; i < n_ev; ++i) + event[i] = alternatives[i][choice[i]]; + return 0; +} + +/* + * Check if newly-added counters have consistent settings for + * exclude_{user,kernel,hv} with each other and any previously + * added counters. + */ +static int check_excludes(struct perf_counter **ctrs, unsigned int cflags[], + int n_prev, int n_new) +{ + int eu = 0, ek = 0, eh = 0; + int i, n, first; + struct perf_counter *counter; + + n = n_prev + n_new; + if (n <= 1) + return 0; + + first = 1; + for (i = 0; i < n; ++i) { + if (cflags[i] & PPMU_LIMITED_PMC_OK) { + cflags[i] &= ~PPMU_LIMITED_PMC_REQD; + continue; + } + counter = ctrs[i]; + if (first) { + eu = counter->attr.exclude_user; + ek = counter->attr.exclude_kernel; + eh = counter->attr.exclude_hv; + first = 0; + } else if (counter->attr.exclude_user != eu || + counter->attr.exclude_kernel != ek || + counter->attr.exclude_hv != eh) { + return -EAGAIN; + } + } + + if (eu || ek || eh) + for (i = 0; i < n; ++i) + if (cflags[i] & PPMU_LIMITED_PMC_OK) + cflags[i] |= PPMU_LIMITED_PMC_REQD; + + return 0; +} + +static void power_pmu_read(struct perf_counter *counter) +{ + long val, delta, prev; + + if (!counter->hw.idx) + return; + /* + * Performance monitor interrupts come even when interrupts + * are soft-disabled, as long as interrupts are hard-enabled. + * Therefore we treat them like NMIs. + */ + do { + prev = atomic64_read(&counter->hw.prev_count); + barrier(); + val = read_pmc(counter->hw.idx); + } while (atomic64_cmpxchg(&counter->hw.prev_count, prev, val) != prev); + + /* The counters are only 32 bits wide */ + delta = (val - prev) & 0xfffffffful; + atomic64_add(delta, &counter->count); + atomic64_sub(delta, &counter->hw.period_left); +} + +/* + * On some machines, PMC5 and PMC6 can't be written, don't respect + * the freeze conditions, and don't generate interrupts. This tells + * us if `counter' is using such a PMC. + */ +static int is_limited_pmc(int pmcnum) +{ + return (ppmu->flags & PPMU_LIMITED_PMC5_6) + && (pmcnum == 5 || pmcnum == 6); +} + +static void freeze_limited_counters(struct cpu_hw_counters *cpuhw, + unsigned long pmc5, unsigned long pmc6) +{ + struct perf_counter *counter; + u64 val, prev, delta; + int i; + + for (i = 0; i < cpuhw->n_limited; ++i) { + counter = cpuhw->limited_counter[i]; + if (!counter->hw.idx) + continue; + val = (counter->hw.idx == 5) ? pmc5 : pmc6; + prev = atomic64_read(&counter->hw.prev_count); + counter->hw.idx = 0; + delta = (val - prev) & 0xfffffffful; + atomic64_add(delta, &counter->count); + } +} + +static void thaw_limited_counters(struct cpu_hw_counters *cpuhw, + unsigned long pmc5, unsigned long pmc6) +{ + struct perf_counter *counter; + u64 val; + int i; + + for (i = 0; i < cpuhw->n_limited; ++i) { + counter = cpuhw->limited_counter[i]; + counter->hw.idx = cpuhw->limited_hwidx[i]; + val = (counter->hw.idx == 5) ? pmc5 : pmc6; + atomic64_set(&counter->hw.prev_count, val); + perf_counter_update_userpage(counter); + } +} + +/* + * Since limited counters don't respect the freeze conditions, we + * have to read them immediately after freezing or unfreezing the + * other counters. We try to keep the values from the limited + * counters as consistent as possible by keeping the delay (in + * cycles and instructions) between freezing/unfreezing and reading + * the limited counters as small and consistent as possible. + * Therefore, if any limited counters are in use, we read them + * both, and always in the same order, to minimize variability, + * and do it inside the same asm that writes MMCR0. + */ +static void write_mmcr0(struct cpu_hw_counters *cpuhw, unsigned long mmcr0) +{ + unsigned long pmc5, pmc6; + + if (!cpuhw->n_limited) { + mtspr(SPRN_MMCR0, mmcr0); + return; + } + + /* + * Write MMCR0, then read PMC5 and PMC6 immediately. + * To ensure we don't get a performance monitor interrupt + * between writing MMCR0 and freezing/thawing the limited + * counters, we first write MMCR0 with the counter overflow + * interrupt enable bits turned off. + */ + asm volatile("mtspr %3,%2; mfspr %0,%4; mfspr %1,%5" + : "=&r" (pmc5), "=&r" (pmc6) + : "r" (mmcr0 & ~(MMCR0_PMC1CE | MMCR0_PMCjCE)), + "i" (SPRN_MMCR0), + "i" (SPRN_PMC5), "i" (SPRN_PMC6)); + + if (mmcr0 & MMCR0_FC) + freeze_limited_counters(cpuhw, pmc5, pmc6); + else + thaw_limited_counters(cpuhw, pmc5, pmc6); + + /* + * Write the full MMCR0 including the counter overflow interrupt + * enable bits, if necessary. + */ + if (mmcr0 & (MMCR0_PMC1CE | MMCR0_PMCjCE)) + mtspr(SPRN_MMCR0, mmcr0); +} + +/* + * Disable all counters to prevent PMU interrupts and to allow + * counters to be added or removed. + */ +void hw_perf_disable(void) +{ + struct cpu_hw_counters *cpuhw; + unsigned long ret; + unsigned long flags; + + local_irq_save(flags); + cpuhw = &__get_cpu_var(cpu_hw_counters); + + ret = cpuhw->disabled; + if (!ret) { + cpuhw->disabled = 1; + cpuhw->n_added = 0; + + /* + * Check if we ever enabled the PMU on this cpu. + */ + if (!cpuhw->pmcs_enabled) { + if (ppc_md.enable_pmcs) + ppc_md.enable_pmcs(); + cpuhw->pmcs_enabled = 1; + } + + /* + * Disable instruction sampling if it was enabled + */ + if (cpuhw->mmcr[2] & MMCRA_SAMPLE_ENABLE) { + mtspr(SPRN_MMCRA, + cpuhw->mmcr[2] & ~MMCRA_SAMPLE_ENABLE); + mb(); + } + + /* + * Set the 'freeze counters' bit. + * The barrier is to make sure the mtspr has been + * executed and the PMU has frozen the counters + * before we return. + */ + write_mmcr0(cpuhw, mfspr(SPRN_MMCR0) | MMCR0_FC); + mb(); + } + local_irq_restore(flags); +} + +/* + * Re-enable all counters if disable == 0. + * If we were previously disabled and counters were added, then + * put the new config on the PMU. + */ +void hw_perf_enable(void) +{ + struct perf_counter *counter; + struct cpu_hw_counters *cpuhw; + unsigned long flags; + long i; + unsigned long val; + s64 left; + unsigned int hwc_index[MAX_HWCOUNTERS]; + int n_lim; + int idx; + + local_irq_save(flags); + cpuhw = &__get_cpu_var(cpu_hw_counters); + if (!cpuhw->disabled) { + local_irq_restore(flags); + return; + } + cpuhw->disabled = 0; + + /* + * If we didn't change anything, or only removed counters, + * no need to recalculate MMCR* settings and reset the PMCs. + * Just reenable the PMU with the current MMCR* settings + * (possibly updated for removal of counters). + */ + if (!cpuhw->n_added) { + mtspr(SPRN_MMCRA, cpuhw->mmcr[2] & ~MMCRA_SAMPLE_ENABLE); + mtspr(SPRN_MMCR1, cpuhw->mmcr[1]); + if (cpuhw->n_counters == 0) + get_lppaca()->pmcregs_in_use = 0; + goto out_enable; + } + + /* + * Compute MMCR* values for the new set of counters + */ + if (ppmu->compute_mmcr(cpuhw->events, cpuhw->n_counters, hwc_index, + cpuhw->mmcr)) { + /* shouldn't ever get here */ + printk(KERN_ERR "oops compute_mmcr failed\n"); + goto out; + } + + /* + * Add in MMCR0 freeze bits corresponding to the + * attr.exclude_* bits for the first counter. + * We have already checked that all counters have the + * same values for these bits as the first counter. + */ + counter = cpuhw->counter[0]; + if (counter->attr.exclude_user) + cpuhw->mmcr[0] |= MMCR0_FCP; + if (counter->attr.exclude_kernel) + cpuhw->mmcr[0] |= freeze_counters_kernel; + if (counter->attr.exclude_hv) + cpuhw->mmcr[0] |= MMCR0_FCHV; + + /* + * Write the new configuration to MMCR* with the freeze + * bit set and set the hardware counters to their initial values. + * Then unfreeze the counters. + */ + get_lppaca()->pmcregs_in_use = 1; + mtspr(SPRN_MMCRA, cpuhw->mmcr[2] & ~MMCRA_SAMPLE_ENABLE); + mtspr(SPRN_MMCR1, cpuhw->mmcr[1]); + mtspr(SPRN_MMCR0, (cpuhw->mmcr[0] & ~(MMCR0_PMC1CE | MMCR0_PMCjCE)) + | MMCR0_FC); + + /* + * Read off any pre-existing counters that need to move + * to another PMC. + */ + for (i = 0; i < cpuhw->n_counters; ++i) { + counter = cpuhw->counter[i]; + if (counter->hw.idx && counter->hw.idx != hwc_index[i] + 1) { + power_pmu_read(counter); + write_pmc(counter->hw.idx, 0); + counter->hw.idx = 0; + } + } + + /* + * Initialize the PMCs for all the new and moved counters. + */ + cpuhw->n_limited = n_lim = 0; + for (i = 0; i < cpuhw->n_counters; ++i) { + counter = cpuhw->counter[i]; + if (counter->hw.idx) + continue; + idx = hwc_index[i] + 1; + if (is_limited_pmc(idx)) { + cpuhw->limited_counter[n_lim] = counter; + cpuhw->limited_hwidx[n_lim] = idx; + ++n_lim; + continue; + } + val = 0; + if (counter->hw.sample_period) { + left = atomic64_read(&counter->hw.period_left); + if (left < 0x80000000L) + val = 0x80000000L - left; + } + atomic64_set(&counter->hw.prev_count, val); + counter->hw.idx = idx; + write_pmc(idx, val); + perf_counter_update_userpage(counter); + } + cpuhw->n_limited = n_lim; + cpuhw->mmcr[0] |= MMCR0_PMXE | MMCR0_FCECE; + + out_enable: + mb(); + write_mmcr0(cpuhw, cpuhw->mmcr[0]); + + /* + * Enable instruction sampling if necessary + */ + if (cpuhw->mmcr[2] & MMCRA_SAMPLE_ENABLE) { + mb(); + mtspr(SPRN_MMCRA, cpuhw->mmcr[2]); + } + + out: + local_irq_restore(flags); +} + +static int collect_events(struct perf_counter *group, int max_count, + struct perf_counter *ctrs[], u64 *events, + unsigned int *flags) +{ + int n = 0; + struct perf_counter *counter; + + if (!is_software_counter(group)) { + if (n >= max_count) + return -1; + ctrs[n] = group; + flags[n] = group->hw.counter_base; + events[n++] = group->hw.config; + } + list_for_each_entry(counter, &group->sibling_list, list_entry) { + if (!is_software_counter(counter) && + counter->state != PERF_COUNTER_STATE_OFF) { + if (n >= max_count) + return -1; + ctrs[n] = counter; + flags[n] = counter->hw.counter_base; + events[n++] = counter->hw.config; + } + } + return n; +} + +static void counter_sched_in(struct perf_counter *counter, int cpu) +{ + counter->state = PERF_COUNTER_STATE_ACTIVE; + counter->oncpu = cpu; + counter->tstamp_running += counter->ctx->time - counter->tstamp_stopped; + if (is_software_counter(counter)) + counter->pmu->enable(counter); +} + +/* + * Called to enable a whole group of counters. + * Returns 1 if the group was enabled, or -EAGAIN if it could not be. + * Assumes the caller has disabled interrupts and has + * frozen the PMU with hw_perf_save_disable. + */ +int hw_perf_group_sched_in(struct perf_counter *group_leader, + struct perf_cpu_context *cpuctx, + struct perf_counter_context *ctx, int cpu) +{ + struct cpu_hw_counters *cpuhw; + long i, n, n0; + struct perf_counter *sub; + + cpuhw = &__get_cpu_var(cpu_hw_counters); + n0 = cpuhw->n_counters; + n = collect_events(group_leader, ppmu->n_counter - n0, + &cpuhw->counter[n0], &cpuhw->events[n0], + &cpuhw->flags[n0]); + if (n < 0) + return -EAGAIN; + if (check_excludes(cpuhw->counter, cpuhw->flags, n0, n)) + return -EAGAIN; + i = power_check_constraints(cpuhw->events, cpuhw->flags, n + n0); + if (i < 0) + return -EAGAIN; + cpuhw->n_counters = n0 + n; + cpuhw->n_added += n; + + /* + * OK, this group can go on; update counter states etc., + * and enable any software counters + */ + for (i = n0; i < n0 + n; ++i) + cpuhw->counter[i]->hw.config = cpuhw->events[i]; + cpuctx->active_oncpu += n; + n = 1; + counter_sched_in(group_leader, cpu); + list_for_each_entry(sub, &group_leader->sibling_list, list_entry) { + if (sub->state != PERF_COUNTER_STATE_OFF) { + counter_sched_in(sub, cpu); + ++n; + } + } + ctx->nr_active += n; + + return 1; +} + +/* + * Add a counter to the PMU. + * If all counters are not already frozen, then we disable and + * re-enable the PMU in order to get hw_perf_enable to do the + * actual work of reconfiguring the PMU. + */ +static int power_pmu_enable(struct perf_counter *counter) +{ + struct cpu_hw_counters *cpuhw; + unsigned long flags; + int n0; + int ret = -EAGAIN; + + local_irq_save(flags); + perf_disable(); + + /* + * Add the counter to the list (if there is room) + * and check whether the total set is still feasible. + */ + cpuhw = &__get_cpu_var(cpu_hw_counters); + n0 = cpuhw->n_counters; + if (n0 >= ppmu->n_counter) + goto out; + cpuhw->counter[n0] = counter; + cpuhw->events[n0] = counter->hw.config; + cpuhw->flags[n0] = counter->hw.counter_base; + if (check_excludes(cpuhw->counter, cpuhw->flags, n0, 1)) + goto out; + if (power_check_constraints(cpuhw->events, cpuhw->flags, n0 + 1)) + goto out; + + counter->hw.config = cpuhw->events[n0]; + ++cpuhw->n_counters; + ++cpuhw->n_added; + + ret = 0; + out: + perf_enable(); + local_irq_restore(flags); + return ret; +} + +/* + * Remove a counter from the PMU. + */ +static void power_pmu_disable(struct perf_counter *counter) +{ + struct cpu_hw_counters *cpuhw; + long i; + unsigned long flags; + + local_irq_save(flags); + perf_disable(); + + power_pmu_read(counter); + + cpuhw = &__get_cpu_var(cpu_hw_counters); + for (i = 0; i < cpuhw->n_counters; ++i) { + if (counter == cpuhw->counter[i]) { + while (++i < cpuhw->n_counters) + cpuhw->counter[i-1] = cpuhw->counter[i]; + --cpuhw->n_counters; + ppmu->disable_pmc(counter->hw.idx - 1, cpuhw->mmcr); + if (counter->hw.idx) { + write_pmc(counter->hw.idx, 0); + counter->hw.idx = 0; + } + perf_counter_update_userpage(counter); + break; + } + } + for (i = 0; i < cpuhw->n_limited; ++i) + if (counter == cpuhw->limited_counter[i]) + break; + if (i < cpuhw->n_limited) { + while (++i < cpuhw->n_limited) { + cpuhw->limited_counter[i-1] = cpuhw->limited_counter[i]; + cpuhw->limited_hwidx[i-1] = cpuhw->limited_hwidx[i]; + } + --cpuhw->n_limited; + } + if (cpuhw->n_counters == 0) { + /* disable exceptions if no counters are running */ + cpuhw->mmcr[0] &= ~(MMCR0_PMXE | MMCR0_FCECE); + } + + perf_enable(); + local_irq_restore(flags); +} + +/* + * Re-enable interrupts on a counter after they were throttled + * because they were coming too fast. + */ +static void power_pmu_unthrottle(struct perf_counter *counter) +{ + s64 val, left; + unsigned long flags; + + if (!counter->hw.idx || !counter->hw.sample_period) + return; + local_irq_save(flags); + perf_disable(); + power_pmu_read(counter); + left = counter->hw.sample_period; + counter->hw.last_period = left; + val = 0; + if (left < 0x80000000L) + val = 0x80000000L - left; + write_pmc(counter->hw.idx, val); + atomic64_set(&counter->hw.prev_count, val); + atomic64_set(&counter->hw.period_left, left); + perf_counter_update_userpage(counter); + perf_enable(); + local_irq_restore(flags); +} + +struct pmu power_pmu = { + .enable = power_pmu_enable, + .disable = power_pmu_disable, + .read = power_pmu_read, + .unthrottle = power_pmu_unthrottle, +}; + +/* + * Return 1 if we might be able to put counter on a limited PMC, + * or 0 if not. + * A counter can only go on a limited PMC if it counts something + * that a limited PMC can count, doesn't require interrupts, and + * doesn't exclude any processor mode. + */ +static int can_go_on_limited_pmc(struct perf_counter *counter, u64 ev, + unsigned int flags) +{ + int n; + u64 alt[MAX_EVENT_ALTERNATIVES]; + + if (counter->attr.exclude_user + || counter->attr.exclude_kernel + || counter->attr.exclude_hv + || counter->attr.sample_period) + return 0; + + if (ppmu->limited_pmc_event(ev)) + return 1; + + /* + * The requested event isn't on a limited PMC already; + * see if any alternative code goes on a limited PMC. + */ + if (!ppmu->get_alternatives) + return 0; + + flags |= PPMU_LIMITED_PMC_OK | PPMU_LIMITED_PMC_REQD; + n = ppmu->get_alternatives(ev, flags, alt); + + return n > 0; +} + +/* + * Find an alternative event that goes on a normal PMC, if possible, + * and return the event code, or 0 if there is no such alternative. + * (Note: event code 0 is "don't count" on all machines.) + */ +static u64 normal_pmc_alternative(u64 ev, unsigned long flags) +{ + u64 alt[MAX_EVENT_ALTERNATIVES]; + int n; + + flags &= ~(PPMU_LIMITED_PMC_OK | PPMU_LIMITED_PMC_REQD); + n = ppmu->get_alternatives(ev, flags, alt); + if (!n) + return 0; + return alt[0]; +} + +/* Number of perf_counters counting hardware events */ +static atomic_t num_counters; +/* Used to avoid races in calling reserve/release_pmc_hardware */ +static DEFINE_MUTEX(pmc_reserve_mutex); + +/* + * Release the PMU if this is the last perf_counter. + */ +static void hw_perf_counter_destroy(struct perf_counter *counter) +{ + if (!atomic_add_unless(&num_counters, -1, 1)) { + mutex_lock(&pmc_reserve_mutex); + if (atomic_dec_return(&num_counters) == 0) + release_pmc_hardware(); + mutex_unlock(&pmc_reserve_mutex); + } +} + +/* + * Translate a generic cache event config to a raw event code. + */ +static int hw_perf_cache_event(u64 config, u64 *eventp) +{ + unsigned long type, op, result; + int ev; + + if (!ppmu->cache_events) + return -EINVAL; + + /* unpack config */ + type = config & 0xff; + op = (config >> 8) & 0xff; + result = (config >> 16) & 0xff; + + if (type >= PERF_COUNT_HW_CACHE_MAX || + op >= PERF_COUNT_HW_CACHE_OP_MAX || + result >= PERF_COUNT_HW_CACHE_RESULT_MAX) + return -EINVAL; + + ev = (*ppmu->cache_events)[type][op][result]; + if (ev == 0) + return -EOPNOTSUPP; + if (ev == -1) + return -EINVAL; + *eventp = ev; + return 0; +} + +const struct pmu *hw_perf_counter_init(struct perf_counter *counter) +{ + u64 ev; + unsigned long flags; + struct perf_counter *ctrs[MAX_HWCOUNTERS]; + u64 events[MAX_HWCOUNTERS]; + unsigned int cflags[MAX_HWCOUNTERS]; + int n; + int err; + + if (!ppmu) + return ERR_PTR(-ENXIO); + switch (counter->attr.type) { + case PERF_TYPE_HARDWARE: + ev = counter->attr.config; + if (ev >= ppmu->n_generic || ppmu->generic_events[ev] == 0) + return ERR_PTR(-EOPNOTSUPP); + ev = ppmu->generic_events[ev]; + break; + case PERF_TYPE_HW_CACHE: + err = hw_perf_cache_event(counter->attr.config, &ev); + if (err) + return ERR_PTR(err); + break; + case PERF_TYPE_RAW: + ev = counter->attr.config; + break; + } + counter->hw.config_base = ev; + counter->hw.idx = 0; + + /* + * If we are not running on a hypervisor, force the + * exclude_hv bit to 0 so that we don't care what + * the user set it to. + */ + if (!firmware_has_feature(FW_FEATURE_LPAR)) + counter->attr.exclude_hv = 0; + + /* + * If this is a per-task counter, then we can use + * PM_RUN_* events interchangeably with their non RUN_* + * equivalents, e.g. PM_RUN_CYC instead of PM_CYC. + * XXX we should check if the task is an idle task. + */ + flags = 0; + if (counter->ctx->task) + flags |= PPMU_ONLY_COUNT_RUN; + + /* + * If this machine has limited counters, check whether this + * event could go on a limited counter. + */ + if (ppmu->flags & PPMU_LIMITED_PMC5_6) { + if (can_go_on_limited_pmc(counter, ev, flags)) { + flags |= PPMU_LIMITED_PMC_OK; + } else if (ppmu->limited_pmc_event(ev)) { + /* + * The requested event is on a limited PMC, + * but we can't use a limited PMC; see if any + * alternative goes on a normal PMC. + */ + ev = normal_pmc_alternative(ev, flags); + if (!ev) + return ERR_PTR(-EINVAL); + } + } + + /* + * If this is in a group, check if it can go on with all the + * other hardware counters in the group. We assume the counter + * hasn't been linked into its leader's sibling list at this point. + */ + n = 0; + if (counter->group_leader != counter) { + n = collect_events(counter->group_leader, ppmu->n_counter - 1, + ctrs, events, cflags); + if (n < 0) + return ERR_PTR(-EINVAL); + } + events[n] = ev; + ctrs[n] = counter; + cflags[n] = flags; + if (check_excludes(ctrs, cflags, n, 1)) + return ERR_PTR(-EINVAL); + if (power_check_constraints(events, cflags, n + 1)) + return ERR_PTR(-EINVAL); + + counter->hw.config = events[n]; + counter->hw.counter_base = cflags[n]; + counter->hw.last_period = counter->hw.sample_period; + atomic64_set(&counter->hw.period_left, counter->hw.last_period); + + /* + * See if we need to reserve the PMU. + * If no counters are currently in use, then we have to take a + * mutex to ensure that we don't race with another task doing + * reserve_pmc_hardware or release_pmc_hardware. + */ + err = 0; + if (!atomic_inc_not_zero(&num_counters)) { + mutex_lock(&pmc_reserve_mutex); + if (atomic_read(&num_counters) == 0 && + reserve_pmc_hardware(perf_counter_interrupt)) + err = -EBUSY; + else + atomic_inc(&num_counters); + mutex_unlock(&pmc_reserve_mutex); + } + counter->destroy = hw_perf_counter_destroy; + + if (err) + return ERR_PTR(err); + return &power_pmu; +} + +/* + * A counter has overflowed; update its count and record + * things if requested. Note that interrupts are hard-disabled + * here so there is no possibility of being interrupted. + */ +static void record_and_restart(struct perf_counter *counter, long val, + struct pt_regs *regs, int nmi) +{ + u64 period = counter->hw.sample_period; + s64 prev, delta, left; + int record = 0; + u64 addr, mmcra, sdsync; + + /* we don't have to worry about interrupts here */ + prev = atomic64_read(&counter->hw.prev_count); + delta = (val - prev) & 0xfffffffful; + atomic64_add(delta, &counter->count); + + /* + * See if the total period for this counter has expired, + * and update for the next period. + */ + val = 0; + left = atomic64_read(&counter->hw.period_left) - delta; + if (period) { + if (left <= 0) { + left += period; + if (left <= 0) + left = period; + record = 1; + } + if (left < 0x80000000L) + val = 0x80000000L - left; + } + + /* + * Finally record data if requested. + */ + if (record) { + struct perf_sample_data data = { + .regs = regs, + .addr = 0, + .period = counter->hw.last_period, + }; + + if (counter->attr.sample_type & PERF_SAMPLE_ADDR) { + /* + * The user wants a data address recorded. + * If we're not doing instruction sampling, + * give them the SDAR (sampled data address). + * If we are doing instruction sampling, then only + * give them the SDAR if it corresponds to the + * instruction pointed to by SIAR; this is indicated + * by the [POWER6_]MMCRA_SDSYNC bit in MMCRA. + */ + mmcra = regs->dsisr; + sdsync = (ppmu->flags & PPMU_ALT_SIPR) ? + POWER6_MMCRA_SDSYNC : MMCRA_SDSYNC; + if (!(mmcra & MMCRA_SAMPLE_ENABLE) || (mmcra & sdsync)) + data.addr = mfspr(SPRN_SDAR); + } + if (perf_counter_overflow(counter, nmi, &data)) { + /* + * Interrupts are coming too fast - throttle them + * by setting the counter to 0, so it will be + * at least 2^30 cycles until the next interrupt + * (assuming each counter counts at most 2 counts + * per cycle). + */ + val = 0; + left = ~0ULL >> 1; + } + } + + write_pmc(counter->hw.idx, val); + atomic64_set(&counter->hw.prev_count, val); + atomic64_set(&counter->hw.period_left, left); + perf_counter_update_userpage(counter); +} + +/* + * Called from generic code to get the misc flags (i.e. processor mode) + * for an event. + */ +unsigned long perf_misc_flags(struct pt_regs *regs) +{ + unsigned long mmcra; + + if (TRAP(regs) != 0xf00) { + /* not a PMU interrupt */ + return user_mode(regs) ? PERF_EVENT_MISC_USER : + PERF_EVENT_MISC_KERNEL; + } + + mmcra = regs->dsisr; + if (ppmu->flags & PPMU_ALT_SIPR) { + if (mmcra & POWER6_MMCRA_SIHV) + return PERF_EVENT_MISC_HYPERVISOR; + return (mmcra & POWER6_MMCRA_SIPR) ? PERF_EVENT_MISC_USER : + PERF_EVENT_MISC_KERNEL; + } + if (mmcra & MMCRA_SIHV) + return PERF_EVENT_MISC_HYPERVISOR; + return (mmcra & MMCRA_SIPR) ? PERF_EVENT_MISC_USER : + PERF_EVENT_MISC_KERNEL; +} + +/* + * Called from generic code to get the instruction pointer + * for an event. + */ +unsigned long perf_instruction_pointer(struct pt_regs *regs) +{ + unsigned long mmcra; + unsigned long ip; + unsigned long slot; + + if (TRAP(regs) != 0xf00) + return regs->nip; /* not a PMU interrupt */ + + ip = mfspr(SPRN_SIAR); + mmcra = regs->dsisr; + if ((mmcra & MMCRA_SAMPLE_ENABLE) && !(ppmu->flags & PPMU_ALT_SIPR)) { + slot = (mmcra & MMCRA_SLOT) >> MMCRA_SLOT_SHIFT; + if (slot > 1) + ip += 4 * (slot - 1); + } + return ip; +} + +/* + * Performance monitor interrupt stuff + */ +static void perf_counter_interrupt(struct pt_regs *regs) +{ + int i; + struct cpu_hw_counters *cpuhw = &__get_cpu_var(cpu_hw_counters); + struct perf_counter *counter; + long val; + int found = 0; + int nmi; + + if (cpuhw->n_limited) + freeze_limited_counters(cpuhw, mfspr(SPRN_PMC5), + mfspr(SPRN_PMC6)); + + /* + * Overload regs->dsisr to store MMCRA so we only need to read it once. + */ + regs->dsisr = mfspr(SPRN_MMCRA); + + /* + * If interrupts were soft-disabled when this PMU interrupt + * occurred, treat it as an NMI. + */ + nmi = !regs->softe; + if (nmi) + nmi_enter(); + else + irq_enter(); + + for (i = 0; i < cpuhw->n_counters; ++i) { + counter = cpuhw->counter[i]; + if (!counter->hw.idx || is_limited_pmc(counter->hw.idx)) + continue; + val = read_pmc(counter->hw.idx); + if ((int)val < 0) { + /* counter has overflowed */ + found = 1; + record_and_restart(counter, val, regs, nmi); + } + } + + /* + * In case we didn't find and reset the counter that caused + * the interrupt, scan all counters and reset any that are + * negative, to avoid getting continual interrupts. + * Any that we processed in the previous loop will not be negative. + */ + if (!found) { + for (i = 0; i < ppmu->n_counter; ++i) { + if (is_limited_pmc(i + 1)) + continue; + val = read_pmc(i + 1); + if ((int)val < 0) + write_pmc(i + 1, 0); + } + } + + /* + * Reset MMCR0 to its normal value. This will set PMXE and + * clear FC (freeze counters) and PMAO (perf mon alert occurred) + * and thus allow interrupts to occur again. + * XXX might want to use MSR.PM to keep the counters frozen until + * we get back out of this interrupt. + */ + write_mmcr0(cpuhw, cpuhw->mmcr[0]); + + if (nmi) + nmi_exit(); + else + irq_exit(); +} + +void hw_perf_counter_setup(int cpu) +{ + struct cpu_hw_counters *cpuhw = &per_cpu(cpu_hw_counters, cpu); + + memset(cpuhw, 0, sizeof(*cpuhw)); + cpuhw->mmcr[0] = MMCR0_FC; +} + +extern struct power_pmu power4_pmu; +extern struct power_pmu ppc970_pmu; +extern struct power_pmu power5_pmu; +extern struct power_pmu power5p_pmu; +extern struct power_pmu power6_pmu; +extern struct power_pmu power7_pmu; + +static int init_perf_counters(void) +{ + unsigned long pvr; + + /* XXX should get this from cputable */ + pvr = mfspr(SPRN_PVR); + switch (PVR_VER(pvr)) { + case PV_POWER4: + case PV_POWER4p: + ppmu = &power4_pmu; + break; + case PV_970: + case PV_970FX: + case PV_970MP: + ppmu = &ppc970_pmu; + break; + case PV_POWER5: + ppmu = &power5_pmu; + break; + case PV_POWER5p: + ppmu = &power5p_pmu; + break; + case 0x3e: + ppmu = &power6_pmu; + break; + case 0x3f: + ppmu = &power7_pmu; + break; + } + + /* + * Use FCHV to ignore kernel events if MSR.HV is set. + */ + if (mfmsr() & MSR_HV) + freeze_counters_kernel = MMCR0_FCHV; + + return 0; +} + +arch_initcall(init_perf_counters); diff --git a/arch/powerpc/kernel/power4-pmu.c b/arch/powerpc/kernel/power4-pmu.c new file mode 100644 index 00000000000..07bd308a5fa --- /dev/null +++ b/arch/powerpc/kernel/power4-pmu.c @@ -0,0 +1,598 @@ +/* + * Performance counter support for POWER4 (GP) and POWER4+ (GQ) processors. + * + * Copyright 2009 Paul Mackerras, IBM Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/kernel.h> +#include <linux/perf_counter.h> +#include <asm/reg.h> + +/* + * Bits in event code for POWER4 + */ +#define PM_PMC_SH 12 /* PMC number (1-based) for direct events */ +#define PM_PMC_MSK 0xf +#define PM_UNIT_SH 8 /* TTMMUX number and setting - unit select */ +#define PM_UNIT_MSK 0xf +#define PM_LOWER_SH 6 +#define PM_LOWER_MSK 1 +#define PM_LOWER_MSKS 0x40 +#define PM_BYTE_SH 4 /* Byte number of event bus to use */ +#define PM_BYTE_MSK 3 +#define PM_PMCSEL_MSK 7 + +/* + * Unit code values + */ +#define PM_FPU 1 +#define PM_ISU1 2 +#define PM_IFU 3 +#define PM_IDU0 4 +#define PM_ISU1_ALT 6 +#define PM_ISU2 7 +#define PM_IFU_ALT 8 +#define PM_LSU0 9 +#define PM_LSU1 0xc +#define PM_GPS 0xf + +/* + * Bits in MMCR0 for POWER4 + */ +#define MMCR0_PMC1SEL_SH 8 +#define MMCR0_PMC2SEL_SH 1 +#define MMCR_PMCSEL_MSK 0x1f + +/* + * Bits in MMCR1 for POWER4 + */ +#define MMCR1_TTM0SEL_SH 62 +#define MMCR1_TTC0SEL_SH 61 +#define MMCR1_TTM1SEL_SH 59 +#define MMCR1_TTC1SEL_SH 58 +#define MMCR1_TTM2SEL_SH 56 +#define MMCR1_TTC2SEL_SH 55 +#define MMCR1_TTM3SEL_SH 53 +#define MMCR1_TTC3SEL_SH 52 +#define MMCR1_TTMSEL_MSK 3 +#define MMCR1_TD_CP_DBG0SEL_SH 50 +#define MMCR1_TD_CP_DBG1SEL_SH 48 +#define MMCR1_TD_CP_DBG2SEL_SH 46 +#define MMCR1_TD_CP_DBG3SEL_SH 44 +#define MMCR1_DEBUG0SEL_SH 43 +#define MMCR1_DEBUG1SEL_SH 42 +#define MMCR1_DEBUG2SEL_SH 41 +#define MMCR1_DEBUG3SEL_SH 40 +#define MMCR1_PMC1_ADDER_SEL_SH 39 +#define MMCR1_PMC2_ADDER_SEL_SH 38 +#define MMCR1_PMC6_ADDER_SEL_SH 37 +#define MMCR1_PMC5_ADDER_SEL_SH 36 +#define MMCR1_PMC8_ADDER_SEL_SH 35 +#define MMCR1_PMC7_ADDER_SEL_SH 34 +#define MMCR1_PMC3_ADDER_SEL_SH 33 +#define MMCR1_PMC4_ADDER_SEL_SH 32 +#define MMCR1_PMC3SEL_SH 27 +#define MMCR1_PMC4SEL_SH 22 +#define MMCR1_PMC5SEL_SH 17 +#define MMCR1_PMC6SEL_SH 12 +#define MMCR1_PMC7SEL_SH 7 +#define MMCR1_PMC8SEL_SH 2 /* note bit 0 is in MMCRA for GP */ + +static short mmcr1_adder_bits[8] = { + MMCR1_PMC1_ADDER_SEL_SH, + MMCR1_PMC2_ADDER_SEL_SH, + MMCR1_PMC3_ADDER_SEL_SH, + MMCR1_PMC4_ADDER_SEL_SH, + MMCR1_PMC5_ADDER_SEL_SH, + MMCR1_PMC6_ADDER_SEL_SH, + MMCR1_PMC7_ADDER_SEL_SH, + MMCR1_PMC8_ADDER_SEL_SH +}; + +/* + * Bits in MMCRA + */ +#define MMCRA_PMC8SEL0_SH 17 /* PMC8SEL bit 0 for GP */ + +/* + * Layout of constraint bits: + * 6666555555555544444444443333333333222222222211111111110000000000 + * 3210987654321098765432109876543210987654321098765432109876543210 + * |[ >[ >[ >|||[ >[ >< >< >< >< ><><><><><><><><> + * | UC1 UC2 UC3 ||| PS1 PS2 B0 B1 B2 B3 P1P2P3P4P5P6P7P8 + * \SMPL ||\TTC3SEL + * |\TTC_IFU_SEL + * \TTM2SEL0 + * + * SMPL - SAMPLE_ENABLE constraint + * 56: SAMPLE_ENABLE value 0x0100_0000_0000_0000 + * + * UC1 - unit constraint 1: can't have all three of FPU/ISU1/IDU0|ISU2 + * 55: UC1 error 0x0080_0000_0000_0000 + * 54: FPU events needed 0x0040_0000_0000_0000 + * 53: ISU1 events needed 0x0020_0000_0000_0000 + * 52: IDU0|ISU2 events needed 0x0010_0000_0000_0000 + * + * UC2 - unit constraint 2: can't have all three of FPU/IFU/LSU0 + * 51: UC2 error 0x0008_0000_0000_0000 + * 50: FPU events needed 0x0004_0000_0000_0000 + * 49: IFU events needed 0x0002_0000_0000_0000 + * 48: LSU0 events needed 0x0001_0000_0000_0000 + * + * UC3 - unit constraint 3: can't have all four of LSU0/IFU/IDU0|ISU2/ISU1 + * 47: UC3 error 0x8000_0000_0000 + * 46: LSU0 events needed 0x4000_0000_0000 + * 45: IFU events needed 0x2000_0000_0000 + * 44: IDU0|ISU2 events needed 0x1000_0000_0000 + * 43: ISU1 events needed 0x0800_0000_0000 + * + * TTM2SEL0 + * 42: 0 = IDU0 events needed + * 1 = ISU2 events needed 0x0400_0000_0000 + * + * TTC_IFU_SEL + * 41: 0 = IFU.U events needed + * 1 = IFU.L events needed 0x0200_0000_0000 + * + * TTC3SEL + * 40: 0 = LSU1.U events needed + * 1 = LSU1.L events needed 0x0100_0000_0000 + * + * PS1 + * 39: PS1 error 0x0080_0000_0000 + * 36-38: count of events needing PMC1/2/5/6 0x0070_0000_0000 + * + * PS2 + * 35: PS2 error 0x0008_0000_0000 + * 32-34: count of events needing PMC3/4/7/8 0x0007_0000_0000 + * + * B0 + * 28-31: Byte 0 event source 0xf000_0000 + * 1 = FPU + * 2 = ISU1 + * 3 = IFU + * 4 = IDU0 + * 7 = ISU2 + * 9 = LSU0 + * c = LSU1 + * f = GPS + * + * B1, B2, B3 + * 24-27, 20-23, 16-19: Byte 1, 2, 3 event sources + * + * P8 + * 15: P8 error 0x8000 + * 14-15: Count of events needing PMC8 + * + * P1..P7 + * 0-13: Count of events needing PMC1..PMC7 + * + * Note: this doesn't allow events using IFU.U to be combined with events + * using IFU.L, though that is feasible (using TTM0 and TTM2). However + * there are no listed events for IFU.L (they are debug events not + * verified for performance monitoring) so this shouldn't cause a + * problem. + */ + +static struct unitinfo { + u64 value, mask; + int unit; + int lowerbit; +} p4_unitinfo[16] = { + [PM_FPU] = { 0x44000000000000ull, 0x88000000000000ull, PM_FPU, 0 }, + [PM_ISU1] = { 0x20080000000000ull, 0x88000000000000ull, PM_ISU1, 0 }, + [PM_ISU1_ALT] = + { 0x20080000000000ull, 0x88000000000000ull, PM_ISU1, 0 }, + [PM_IFU] = { 0x02200000000000ull, 0x08820000000000ull, PM_IFU, 41 }, + [PM_IFU_ALT] = + { 0x02200000000000ull, 0x08820000000000ull, PM_IFU, 41 }, + [PM_IDU0] = { 0x10100000000000ull, 0x80840000000000ull, PM_IDU0, 1 }, + [PM_ISU2] = { 0x10140000000000ull, 0x80840000000000ull, PM_ISU2, 0 }, + [PM_LSU0] = { 0x01400000000000ull, 0x08800000000000ull, PM_LSU0, 0 }, + [PM_LSU1] = { 0x00000000000000ull, 0x00010000000000ull, PM_LSU1, 40 }, + [PM_GPS] = { 0x00000000000000ull, 0x00000000000000ull, PM_GPS, 0 } +}; + +static unsigned char direct_marked_event[8] = { + (1<<2) | (1<<3), /* PMC1: PM_MRK_GRP_DISP, PM_MRK_ST_CMPL */ + (1<<3) | (1<<5), /* PMC2: PM_THRESH_TIMEO, PM_MRK_BRU_FIN */ + (1<<3), /* PMC3: PM_MRK_ST_CMPL_INT */ + (1<<4) | (1<<5), /* PMC4: PM_MRK_GRP_CMPL, PM_MRK_CRU_FIN */ + (1<<4) | (1<<5), /* PMC5: PM_MRK_GRP_TIMEO */ + (1<<3) | (1<<4) | (1<<5), + /* PMC6: PM_MRK_ST_GPS, PM_MRK_FXU_FIN, PM_MRK_GRP_ISSUED */ + (1<<4) | (1<<5), /* PMC7: PM_MRK_FPU_FIN, PM_MRK_INST_FIN */ + (1<<4), /* PMC8: PM_MRK_LSU_FIN */ +}; + +/* + * Returns 1 if event counts things relating to marked instructions + * and thus needs the MMCRA_SAMPLE_ENABLE bit set, or 0 if not. + */ +static int p4_marked_instr_event(u64 event) +{ + int pmc, psel, unit, byte, bit; + unsigned int mask; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + psel = event & PM_PMCSEL_MSK; + if (pmc) { + if (direct_marked_event[pmc - 1] & (1 << psel)) + return 1; + if (psel == 0) /* add events */ + bit = (pmc <= 4)? pmc - 1: 8 - pmc; + else if (psel == 6) /* decode events */ + bit = 4; + else + return 0; + } else + bit = psel; + + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + mask = 0; + switch (unit) { + case PM_LSU1: + if (event & PM_LOWER_MSKS) + mask = 1 << 28; /* byte 7 bit 4 */ + else + mask = 6 << 24; /* byte 3 bits 1 and 2 */ + break; + case PM_LSU0: + /* byte 3, bit 3; byte 2 bits 0,2,3,4,5; byte 1 */ + mask = 0x083dff00; + } + return (mask >> (byte * 8 + bit)) & 1; +} + +static int p4_get_constraint(u64 event, u64 *maskp, u64 *valp) +{ + int pmc, byte, unit, lower, sh; + u64 mask = 0, value = 0; + int grp = -1; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 8) + return -1; + sh = (pmc - 1) * 2; + mask |= 2 << sh; + value |= 1 << sh; + grp = ((pmc - 1) >> 1) & 1; + } + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + if (unit) { + lower = (event >> PM_LOWER_SH) & PM_LOWER_MSK; + + /* + * Bus events on bytes 0 and 2 can be counted + * on PMC1/2/5/6; bytes 1 and 3 on PMC3/4/7/8. + */ + if (!pmc) + grp = byte & 1; + + if (!p4_unitinfo[unit].unit) + return -1; + mask |= p4_unitinfo[unit].mask; + value |= p4_unitinfo[unit].value; + sh = p4_unitinfo[unit].lowerbit; + if (sh > 1) + value |= (u64)lower << sh; + else if (lower != sh) + return -1; + unit = p4_unitinfo[unit].unit; + + /* Set byte lane select field */ + mask |= 0xfULL << (28 - 4 * byte); + value |= (u64)unit << (28 - 4 * byte); + } + if (grp == 0) { + /* increment PMC1/2/5/6 field */ + mask |= 0x8000000000ull; + value |= 0x1000000000ull; + } else { + /* increment PMC3/4/7/8 field */ + mask |= 0x800000000ull; + value |= 0x100000000ull; + } + + /* Marked instruction events need sample_enable set */ + if (p4_marked_instr_event(event)) { + mask |= 1ull << 56; + value |= 1ull << 56; + } + + /* PMCSEL=6 decode events on byte 2 need sample_enable clear */ + if (pmc && (event & PM_PMCSEL_MSK) == 6 && byte == 2) + mask |= 1ull << 56; + + *maskp = mask; + *valp = value; + return 0; +} + +static unsigned int ppc_inst_cmpl[] = { + 0x1001, 0x4001, 0x6001, 0x7001, 0x8001 +}; + +static int p4_get_alternatives(u64 event, unsigned int flags, u64 alt[]) +{ + int i, j, na; + + alt[0] = event; + na = 1; + + /* 2 possibilities for PM_GRP_DISP_REJECT */ + if (event == 0x8003 || event == 0x0224) { + alt[1] = event ^ (0x8003 ^ 0x0224); + return 2; + } + + /* 2 possibilities for PM_ST_MISS_L1 */ + if (event == 0x0c13 || event == 0x0c23) { + alt[1] = event ^ (0x0c13 ^ 0x0c23); + return 2; + } + + /* several possibilities for PM_INST_CMPL */ + for (i = 0; i < ARRAY_SIZE(ppc_inst_cmpl); ++i) { + if (event == ppc_inst_cmpl[i]) { + for (j = 0; j < ARRAY_SIZE(ppc_inst_cmpl); ++j) + if (j != i) + alt[na++] = ppc_inst_cmpl[j]; + break; + } + } + + return na; +} + +static int p4_compute_mmcr(u64 event[], int n_ev, + unsigned int hwc[], u64 mmcr[]) +{ + u64 mmcr0 = 0, mmcr1 = 0, mmcra = 0; + unsigned int pmc, unit, byte, psel, lower; + unsigned int ttm, grp; + unsigned int pmc_inuse = 0; + unsigned int pmc_grp_use[2]; + unsigned char busbyte[4]; + unsigned char unituse[16]; + unsigned int unitlower = 0; + int i; + + if (n_ev > 8) + return -1; + + /* First pass to count resource use */ + pmc_grp_use[0] = pmc_grp_use[1] = 0; + memset(busbyte, 0, sizeof(busbyte)); + memset(unituse, 0, sizeof(unituse)); + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc_inuse & (1 << (pmc - 1))) + return -1; + pmc_inuse |= 1 << (pmc - 1); + /* count 1/2/5/6 vs 3/4/7/8 use */ + ++pmc_grp_use[((pmc - 1) >> 1) & 1]; + } + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event[i] >> PM_BYTE_SH) & PM_BYTE_MSK; + lower = (event[i] >> PM_LOWER_SH) & PM_LOWER_MSK; + if (unit) { + if (!pmc) + ++pmc_grp_use[byte & 1]; + if (unit == 6 || unit == 8) + /* map alt ISU1/IFU codes: 6->2, 8->3 */ + unit = (unit >> 1) - 1; + if (busbyte[byte] && busbyte[byte] != unit) + return -1; + busbyte[byte] = unit; + lower <<= unit; + if (unituse[unit] && lower != (unitlower & lower)) + return -1; + unituse[unit] = 1; + unitlower |= lower; + } + } + if (pmc_grp_use[0] > 4 || pmc_grp_use[1] > 4) + return -1; + + /* + * Assign resources and set multiplexer selects. + * + * Units 1,2,3 are on TTM0, 4,6,7 on TTM1, 8,10 on TTM2. + * Each TTMx can only select one unit, but since + * units 2 and 6 are both ISU1, and 3 and 8 are both IFU, + * we have some choices. + */ + if (unituse[2] & (unituse[1] | (unituse[3] & unituse[9]))) { + unituse[6] = 1; /* Move 2 to 6 */ + unituse[2] = 0; + } + if (unituse[3] & (unituse[1] | unituse[2])) { + unituse[8] = 1; /* Move 3 to 8 */ + unituse[3] = 0; + unitlower = (unitlower & ~8) | ((unitlower & 8) << 5); + } + /* Check only one unit per TTMx */ + if (unituse[1] + unituse[2] + unituse[3] > 1 || + unituse[4] + unituse[6] + unituse[7] > 1 || + unituse[8] + unituse[9] > 1 || + (unituse[5] | unituse[10] | unituse[11] | + unituse[13] | unituse[14])) + return -1; + + /* Set TTMxSEL fields. Note, units 1-3 => TTM0SEL codes 0-2 */ + mmcr1 |= (u64)(unituse[3] * 2 + unituse[2]) << MMCR1_TTM0SEL_SH; + mmcr1 |= (u64)(unituse[7] * 3 + unituse[6] * 2) << MMCR1_TTM1SEL_SH; + mmcr1 |= (u64)unituse[9] << MMCR1_TTM2SEL_SH; + + /* Set TTCxSEL fields. */ + if (unitlower & 0xe) + mmcr1 |= 1ull << MMCR1_TTC0SEL_SH; + if (unitlower & 0xf0) + mmcr1 |= 1ull << MMCR1_TTC1SEL_SH; + if (unitlower & 0xf00) + mmcr1 |= 1ull << MMCR1_TTC2SEL_SH; + if (unitlower & 0x7000) + mmcr1 |= 1ull << MMCR1_TTC3SEL_SH; + + /* Set byte lane select fields. */ + for (byte = 0; byte < 4; ++byte) { + unit = busbyte[byte]; + if (!unit) + continue; + if (unit == 0xf) { + /* special case for GPS */ + mmcr1 |= 1ull << (MMCR1_DEBUG0SEL_SH - byte); + } else { + if (!unituse[unit]) + ttm = unit - 1; /* 2->1, 3->2 */ + else + ttm = unit >> 2; + mmcr1 |= (u64)ttm << (MMCR1_TD_CP_DBG0SEL_SH - 2*byte); + } + } + + /* Second pass: assign PMCs, set PMCxSEL and PMCx_ADDER_SEL fields */ + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event[i] >> PM_BYTE_SH) & PM_BYTE_MSK; + psel = event[i] & PM_PMCSEL_MSK; + if (!pmc) { + /* Bus event or 00xxx direct event (off or cycles) */ + if (unit) + psel |= 0x10 | ((byte & 2) << 2); + for (pmc = 0; pmc < 8; ++pmc) { + if (pmc_inuse & (1 << pmc)) + continue; + grp = (pmc >> 1) & 1; + if (unit) { + if (grp == (byte & 1)) + break; + } else if (pmc_grp_use[grp] < 4) { + ++pmc_grp_use[grp]; + break; + } + } + pmc_inuse |= 1 << pmc; + } else { + /* Direct event */ + --pmc; + if (psel == 0 && (byte & 2)) + /* add events on higher-numbered bus */ + mmcr1 |= 1ull << mmcr1_adder_bits[pmc]; + else if (psel == 6 && byte == 3) + /* seem to need to set sample_enable here */ + mmcra |= MMCRA_SAMPLE_ENABLE; + psel |= 8; + } + if (pmc <= 1) + mmcr0 |= psel << (MMCR0_PMC1SEL_SH - 7 * pmc); + else + mmcr1 |= psel << (MMCR1_PMC3SEL_SH - 5 * (pmc - 2)); + if (pmc == 7) /* PMC8 */ + mmcra |= (psel & 1) << MMCRA_PMC8SEL0_SH; + hwc[i] = pmc; + if (p4_marked_instr_event(event[i])) + mmcra |= MMCRA_SAMPLE_ENABLE; + } + + if (pmc_inuse & 1) + mmcr0 |= MMCR0_PMC1CE; + if (pmc_inuse & 0xfe) + mmcr0 |= MMCR0_PMCjCE; + + mmcra |= 0x2000; /* mark only one IOP per PPC instruction */ + + /* Return MMCRx values */ + mmcr[0] = mmcr0; + mmcr[1] = mmcr1; + mmcr[2] = mmcra; + return 0; +} + +static void p4_disable_pmc(unsigned int pmc, u64 mmcr[]) +{ + /* + * Setting the PMCxSEL field to 0 disables PMC x. + * (Note that pmc is 0-based here, not 1-based.) + */ + if (pmc <= 1) { + mmcr[0] &= ~(0x1fUL << (MMCR0_PMC1SEL_SH - 7 * pmc)); + } else { + mmcr[1] &= ~(0x1fUL << (MMCR1_PMC3SEL_SH - 5 * (pmc - 2))); + if (pmc == 7) + mmcr[2] &= ~(1UL << MMCRA_PMC8SEL0_SH); + } +} + +static int p4_generic_events[] = { + [PERF_COUNT_HW_CPU_CYCLES] = 7, + [PERF_COUNT_HW_INSTRUCTIONS] = 0x1001, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x8c10, /* PM_LD_REF_L1 */ + [PERF_COUNT_HW_CACHE_MISSES] = 0x3c10, /* PM_LD_MISS_L1 */ + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x330, /* PM_BR_ISSUED */ + [PERF_COUNT_HW_BRANCH_MISSES] = 0x331, /* PM_BR_MPRED_CR */ +}; + +#define C(x) PERF_COUNT_HW_CACHE_##x + +/* + * Table of generalized cache-related events. + * 0 means not supported, -1 means nonsensical, other values + * are event codes. + */ +static int power4_cache_events[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = { + [C(L1D)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x8c10, 0x3c10 }, + [C(OP_WRITE)] = { 0x7c10, 0xc13 }, + [C(OP_PREFETCH)] = { 0xc35, 0 }, + }, + [C(L1I)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { 0, 0 }, + }, + [C(LL)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0 }, + [C(OP_WRITE)] = { 0, 0 }, + [C(OP_PREFETCH)] = { 0xc34, 0 }, + }, + [C(DTLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x904 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(ITLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x900 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(BPU)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x330, 0x331 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, +}; + +struct power_pmu power4_pmu = { + .n_counter = 8, + .max_alternatives = 5, + .add_fields = 0x0000001100005555ull, + .test_adder = 0x0011083300000000ull, + .compute_mmcr = p4_compute_mmcr, + .get_constraint = p4_get_constraint, + .get_alternatives = p4_get_alternatives, + .disable_pmc = p4_disable_pmc, + .n_generic = ARRAY_SIZE(p4_generic_events), + .generic_events = p4_generic_events, + .cache_events = &power4_cache_events, +}; diff --git a/arch/powerpc/kernel/power5+-pmu.c b/arch/powerpc/kernel/power5+-pmu.c new file mode 100644 index 00000000000..41e5d2d958d --- /dev/null +++ b/arch/powerpc/kernel/power5+-pmu.c @@ -0,0 +1,671 @@ +/* + * Performance counter support for POWER5+/++ (not POWER5) processors. + * + * Copyright 2009 Paul Mackerras, IBM Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/kernel.h> +#include <linux/perf_counter.h> +#include <asm/reg.h> + +/* + * Bits in event code for POWER5+ (POWER5 GS) and POWER5++ (POWER5 GS DD3) + */ +#define PM_PMC_SH 20 /* PMC number (1-based) for direct events */ +#define PM_PMC_MSK 0xf +#define PM_PMC_MSKS (PM_PMC_MSK << PM_PMC_SH) +#define PM_UNIT_SH 16 /* TTMMUX number and setting - unit select */ +#define PM_UNIT_MSK 0xf +#define PM_BYTE_SH 12 /* Byte number of event bus to use */ +#define PM_BYTE_MSK 7 +#define PM_GRS_SH 8 /* Storage subsystem mux select */ +#define PM_GRS_MSK 7 +#define PM_BUSEVENT_MSK 0x80 /* Set if event uses event bus */ +#define PM_PMCSEL_MSK 0x7f + +/* Values in PM_UNIT field */ +#define PM_FPU 0 +#define PM_ISU0 1 +#define PM_IFU 2 +#define PM_ISU1 3 +#define PM_IDU 4 +#define PM_ISU0_ALT 6 +#define PM_GRS 7 +#define PM_LSU0 8 +#define PM_LSU1 0xc +#define PM_LASTUNIT 0xc + +/* + * Bits in MMCR1 for POWER5+ + */ +#define MMCR1_TTM0SEL_SH 62 +#define MMCR1_TTM1SEL_SH 60 +#define MMCR1_TTM2SEL_SH 58 +#define MMCR1_TTM3SEL_SH 56 +#define MMCR1_TTMSEL_MSK 3 +#define MMCR1_TD_CP_DBG0SEL_SH 54 +#define MMCR1_TD_CP_DBG1SEL_SH 52 +#define MMCR1_TD_CP_DBG2SEL_SH 50 +#define MMCR1_TD_CP_DBG3SEL_SH 48 +#define MMCR1_GRS_L2SEL_SH 46 +#define MMCR1_GRS_L2SEL_MSK 3 +#define MMCR1_GRS_L3SEL_SH 44 +#define MMCR1_GRS_L3SEL_MSK 3 +#define MMCR1_GRS_MCSEL_SH 41 +#define MMCR1_GRS_MCSEL_MSK 7 +#define MMCR1_GRS_FABSEL_SH 39 +#define MMCR1_GRS_FABSEL_MSK 3 +#define MMCR1_PMC1_ADDER_SEL_SH 35 +#define MMCR1_PMC2_ADDER_SEL_SH 34 +#define MMCR1_PMC3_ADDER_SEL_SH 33 +#define MMCR1_PMC4_ADDER_SEL_SH 32 +#define MMCR1_PMC1SEL_SH 25 +#define MMCR1_PMC2SEL_SH 17 +#define MMCR1_PMC3SEL_SH 9 +#define MMCR1_PMC4SEL_SH 1 +#define MMCR1_PMCSEL_SH(n) (MMCR1_PMC1SEL_SH - (n) * 8) +#define MMCR1_PMCSEL_MSK 0x7f + +/* + * Bits in MMCRA + */ + +/* + * Layout of constraint bits: + * 6666555555555544444444443333333333222222222211111111110000000000 + * 3210987654321098765432109876543210987654321098765432109876543210 + * [ ><><>< ><> <><>[ > < >< >< >< ><><><><><><> + * NC G0G1G2 G3 T0T1 UC B0 B1 B2 B3 P6P5P4P3P2P1 + * + * NC - number of counters + * 51: NC error 0x0008_0000_0000_0000 + * 48-50: number of events needing PMC1-4 0x0007_0000_0000_0000 + * + * G0..G3 - GRS mux constraints + * 46-47: GRS_L2SEL value + * 44-45: GRS_L3SEL value + * 41-44: GRS_MCSEL value + * 39-40: GRS_FABSEL value + * Note that these match up with their bit positions in MMCR1 + * + * T0 - TTM0 constraint + * 36-37: TTM0SEL value (0=FPU, 2=IFU, 3=ISU1) 0x30_0000_0000 + * + * T1 - TTM1 constraint + * 34-35: TTM1SEL value (0=IDU, 3=GRS) 0x0c_0000_0000 + * + * UC - unit constraint: can't have all three of FPU|IFU|ISU1, ISU0, IDU|GRS + * 33: UC3 error 0x02_0000_0000 + * 32: FPU|IFU|ISU1 events needed 0x01_0000_0000 + * 31: ISU0 events needed 0x01_8000_0000 + * 30: IDU|GRS events needed 0x00_4000_0000 + * + * B0 + * 24-27: Byte 0 event source 0x0f00_0000 + * Encoding as for the event code + * + * B1, B2, B3 + * 20-23, 16-19, 12-15: Byte 1, 2, 3 event sources + * + * P6 + * 11: P6 error 0x800 + * 10-11: Count of events needing PMC6 + * + * P1..P5 + * 0-9: Count of events needing PMC1..PMC5 + */ + +static const int grsel_shift[8] = { + MMCR1_GRS_L2SEL_SH, MMCR1_GRS_L2SEL_SH, MMCR1_GRS_L2SEL_SH, + MMCR1_GRS_L3SEL_SH, MMCR1_GRS_L3SEL_SH, MMCR1_GRS_L3SEL_SH, + MMCR1_GRS_MCSEL_SH, MMCR1_GRS_FABSEL_SH +}; + +/* Masks and values for using events from the various units */ +static u64 unit_cons[PM_LASTUNIT+1][2] = { + [PM_FPU] = { 0x3200000000ull, 0x0100000000ull }, + [PM_ISU0] = { 0x0200000000ull, 0x0080000000ull }, + [PM_ISU1] = { 0x3200000000ull, 0x3100000000ull }, + [PM_IFU] = { 0x3200000000ull, 0x2100000000ull }, + [PM_IDU] = { 0x0e00000000ull, 0x0040000000ull }, + [PM_GRS] = { 0x0e00000000ull, 0x0c40000000ull }, +}; + +static int power5p_get_constraint(u64 event, u64 *maskp, u64 *valp) +{ + int pmc, byte, unit, sh; + int bit, fmask; + u64 mask = 0, value = 0; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 6) + return -1; + sh = (pmc - 1) * 2; + mask |= 2 << sh; + value |= 1 << sh; + if (pmc >= 5 && !(event == 0x500009 || event == 0x600005)) + return -1; + } + if (event & PM_BUSEVENT_MSK) { + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + if (unit > PM_LASTUNIT) + return -1; + if (unit == PM_ISU0_ALT) + unit = PM_ISU0; + mask |= unit_cons[unit][0]; + value |= unit_cons[unit][1]; + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + if (byte >= 4) { + if (unit != PM_LSU1) + return -1; + /* Map LSU1 low word (bytes 4-7) to unit LSU1+1 */ + ++unit; + byte &= 3; + } + if (unit == PM_GRS) { + bit = event & 7; + fmask = (bit == 6)? 7: 3; + sh = grsel_shift[bit]; + mask |= (u64)fmask << sh; + value |= (u64)((event >> PM_GRS_SH) & fmask) << sh; + } + /* Set byte lane select field */ + mask |= 0xfULL << (24 - 4 * byte); + value |= (u64)unit << (24 - 4 * byte); + } + if (pmc < 5) { + /* need a counter from PMC1-4 set */ + mask |= 0x8000000000000ull; + value |= 0x1000000000000ull; + } + *maskp = mask; + *valp = value; + return 0; +} + +static int power5p_limited_pmc_event(u64 event) +{ + int pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + + return pmc == 5 || pmc == 6; +} + +#define MAX_ALT 3 /* at most 3 alternatives for any event */ + +static const unsigned int event_alternatives[][MAX_ALT] = { + { 0x100c0, 0x40001f }, /* PM_GCT_FULL_CYC */ + { 0x120e4, 0x400002 }, /* PM_GRP_DISP_REJECT */ + { 0x230e2, 0x323087 }, /* PM_BR_PRED_CR */ + { 0x230e3, 0x223087, 0x3230a0 }, /* PM_BR_PRED_TA */ + { 0x410c7, 0x441084 }, /* PM_THRD_L2MISS_BOTH_CYC */ + { 0x800c4, 0xc20e0 }, /* PM_DTLB_MISS */ + { 0xc50c6, 0xc60e0 }, /* PM_MRK_DTLB_MISS */ + { 0x100005, 0x600005 }, /* PM_RUN_CYC */ + { 0x100009, 0x200009 }, /* PM_INST_CMPL */ + { 0x200015, 0x300015 }, /* PM_LSU_LMQ_SRQ_EMPTY_CYC */ + { 0x300009, 0x400009 }, /* PM_INST_DISP */ +}; + +/* + * Scan the alternatives table for a match and return the + * index into the alternatives table if found, else -1. + */ +static int find_alternative(unsigned int event) +{ + int i, j; + + for (i = 0; i < ARRAY_SIZE(event_alternatives); ++i) { + if (event < event_alternatives[i][0]) + break; + for (j = 0; j < MAX_ALT && event_alternatives[i][j]; ++j) + if (event == event_alternatives[i][j]) + return i; + } + return -1; +} + +static const unsigned char bytedecode_alternatives[4][4] = { + /* PMC 1 */ { 0x21, 0x23, 0x25, 0x27 }, + /* PMC 2 */ { 0x07, 0x17, 0x0e, 0x1e }, + /* PMC 3 */ { 0x20, 0x22, 0x24, 0x26 }, + /* PMC 4 */ { 0x07, 0x17, 0x0e, 0x1e } +}; + +/* + * Some direct events for decodes of event bus byte 3 have alternative + * PMCSEL values on other counters. This returns the alternative + * event code for those that do, or -1 otherwise. This also handles + * alternative PCMSEL values for add events. + */ +static s64 find_alternative_bdecode(u64 event) +{ + int pmc, altpmc, pp, j; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc == 0 || pmc > 4) + return -1; + altpmc = 5 - pmc; /* 1 <-> 4, 2 <-> 3 */ + pp = event & PM_PMCSEL_MSK; + for (j = 0; j < 4; ++j) { + if (bytedecode_alternatives[pmc - 1][j] == pp) { + return (event & ~(PM_PMC_MSKS | PM_PMCSEL_MSK)) | + (altpmc << PM_PMC_SH) | + bytedecode_alternatives[altpmc - 1][j]; + } + } + + /* new decode alternatives for power5+ */ + if (pmc == 1 && (pp == 0x0d || pp == 0x0e)) + return event + (2 << PM_PMC_SH) + (0x2e - 0x0d); + if (pmc == 3 && (pp == 0x2e || pp == 0x2f)) + return event - (2 << PM_PMC_SH) - (0x2e - 0x0d); + + /* alternative add event encodings */ + if (pp == 0x10 || pp == 0x28) + return ((event ^ (0x10 ^ 0x28)) & ~PM_PMC_MSKS) | + (altpmc << PM_PMC_SH); + + return -1; +} + +static int power5p_get_alternatives(u64 event, unsigned int flags, u64 alt[]) +{ + int i, j, nalt = 1; + int nlim; + s64 ae; + + alt[0] = event; + nalt = 1; + nlim = power5p_limited_pmc_event(event); + i = find_alternative(event); + if (i >= 0) { + for (j = 0; j < MAX_ALT; ++j) { + ae = event_alternatives[i][j]; + if (ae && ae != event) + alt[nalt++] = ae; + nlim += power5p_limited_pmc_event(ae); + } + } else { + ae = find_alternative_bdecode(event); + if (ae > 0) + alt[nalt++] = ae; + } + + if (flags & PPMU_ONLY_COUNT_RUN) { + /* + * We're only counting in RUN state, + * so PM_CYC is equivalent to PM_RUN_CYC + * and PM_INST_CMPL === PM_RUN_INST_CMPL. + * This doesn't include alternatives that don't provide + * any extra flexibility in assigning PMCs (e.g. + * 0x100005 for PM_RUN_CYC vs. 0xf for PM_CYC). + * Note that even with these additional alternatives + * we never end up with more than 3 alternatives for any event. + */ + j = nalt; + for (i = 0; i < nalt; ++i) { + switch (alt[i]) { + case 0xf: /* PM_CYC */ + alt[j++] = 0x600005; /* PM_RUN_CYC */ + ++nlim; + break; + case 0x600005: /* PM_RUN_CYC */ + alt[j++] = 0xf; + break; + case 0x100009: /* PM_INST_CMPL */ + alt[j++] = 0x500009; /* PM_RUN_INST_CMPL */ + ++nlim; + break; + case 0x500009: /* PM_RUN_INST_CMPL */ + alt[j++] = 0x100009; /* PM_INST_CMPL */ + alt[j++] = 0x200009; + break; + } + } + nalt = j; + } + + if (!(flags & PPMU_LIMITED_PMC_OK) && nlim) { + /* remove the limited PMC events */ + j = 0; + for (i = 0; i < nalt; ++i) { + if (!power5p_limited_pmc_event(alt[i])) { + alt[j] = alt[i]; + ++j; + } + } + nalt = j; + } else if ((flags & PPMU_LIMITED_PMC_REQD) && nlim < nalt) { + /* remove all but the limited PMC events */ + j = 0; + for (i = 0; i < nalt; ++i) { + if (power5p_limited_pmc_event(alt[i])) { + alt[j] = alt[i]; + ++j; + } + } + nalt = j; + } + + return nalt; +} + +/* + * Map of which direct events on which PMCs are marked instruction events. + * Indexed by PMCSEL value, bit i (LE) set if PMC i is a marked event. + * Bit 0 is set if it is marked for all PMCs. + * The 0x80 bit indicates a byte decode PMCSEL value. + */ +static unsigned char direct_event_is_marked[0x28] = { + 0, /* 00 */ + 0x1f, /* 01 PM_IOPS_CMPL */ + 0x2, /* 02 PM_MRK_GRP_DISP */ + 0xe, /* 03 PM_MRK_ST_CMPL, PM_MRK_ST_GPS, PM_MRK_ST_CMPL_INT */ + 0, /* 04 */ + 0x1c, /* 05 PM_MRK_BRU_FIN, PM_MRK_INST_FIN, PM_MRK_CRU_FIN */ + 0x80, /* 06 */ + 0x80, /* 07 */ + 0, 0, 0,/* 08 - 0a */ + 0x18, /* 0b PM_THRESH_TIMEO, PM_MRK_GRP_TIMEO */ + 0, /* 0c */ + 0x80, /* 0d */ + 0x80, /* 0e */ + 0, /* 0f */ + 0, /* 10 */ + 0x14, /* 11 PM_MRK_GRP_BR_REDIR, PM_MRK_GRP_IC_MISS */ + 0, /* 12 */ + 0x10, /* 13 PM_MRK_GRP_CMPL */ + 0x1f, /* 14 PM_GRP_MRK, PM_MRK_{FXU,FPU,LSU}_FIN */ + 0x2, /* 15 PM_MRK_GRP_ISSUED */ + 0x80, /* 16 */ + 0x80, /* 17 */ + 0, 0, 0, 0, 0, + 0x80, /* 1d */ + 0x80, /* 1e */ + 0, /* 1f */ + 0x80, /* 20 */ + 0x80, /* 21 */ + 0x80, /* 22 */ + 0x80, /* 23 */ + 0x80, /* 24 */ + 0x80, /* 25 */ + 0x80, /* 26 */ + 0x80, /* 27 */ +}; + +/* + * Returns 1 if event counts things relating to marked instructions + * and thus needs the MMCRA_SAMPLE_ENABLE bit set, or 0 if not. + */ +static int power5p_marked_instr_event(u64 event) +{ + int pmc, psel; + int bit, byte, unit; + u32 mask; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + psel = event & PM_PMCSEL_MSK; + if (pmc >= 5) + return 0; + + bit = -1; + if (psel < sizeof(direct_event_is_marked)) { + if (direct_event_is_marked[psel] & (1 << pmc)) + return 1; + if (direct_event_is_marked[psel] & 0x80) + bit = 4; + else if (psel == 0x08) + bit = pmc - 1; + else if (psel == 0x10) + bit = 4 - pmc; + else if (psel == 0x1b && (pmc == 1 || pmc == 3)) + bit = 4; + } else if ((psel & 0x48) == 0x40) { + bit = psel & 7; + } else if (psel == 0x28) { + bit = pmc - 1; + } else if (pmc == 3 && (psel == 0x2e || psel == 0x2f)) { + bit = 4; + } + + if (!(event & PM_BUSEVENT_MSK) || bit == -1) + return 0; + + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + if (unit == PM_LSU0) { + /* byte 1 bits 0-7, byte 2 bits 0,2-4,6 */ + mask = 0x5dff00; + } else if (unit == PM_LSU1 && byte >= 4) { + byte -= 4; + /* byte 5 bits 6-7, byte 6 bits 0,4, byte 7 bits 0-4,6 */ + mask = 0x5f11c000; + } else + return 0; + + return (mask >> (byte * 8 + bit)) & 1; +} + +static int power5p_compute_mmcr(u64 event[], int n_ev, + unsigned int hwc[], u64 mmcr[]) +{ + u64 mmcr1 = 0; + u64 mmcra = 0; + unsigned int pmc, unit, byte, psel; + unsigned int ttm; + int i, isbus, bit, grsel; + unsigned int pmc_inuse = 0; + unsigned char busbyte[4]; + unsigned char unituse[16]; + int ttmuse; + + if (n_ev > 6) + return -1; + + /* First pass to count resource use */ + memset(busbyte, 0, sizeof(busbyte)); + memset(unituse, 0, sizeof(unituse)); + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 6) + return -1; + if (pmc_inuse & (1 << (pmc - 1))) + return -1; + pmc_inuse |= 1 << (pmc - 1); + } + if (event[i] & PM_BUSEVENT_MSK) { + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event[i] >> PM_BYTE_SH) & PM_BYTE_MSK; + if (unit > PM_LASTUNIT) + return -1; + if (unit == PM_ISU0_ALT) + unit = PM_ISU0; + if (byte >= 4) { + if (unit != PM_LSU1) + return -1; + ++unit; + byte &= 3; + } + if (busbyte[byte] && busbyte[byte] != unit) + return -1; + busbyte[byte] = unit; + unituse[unit] = 1; + } + } + + /* + * Assign resources and set multiplexer selects. + * + * PM_ISU0 can go either on TTM0 or TTM1, but that's the only + * choice we have to deal with. + */ + if (unituse[PM_ISU0] & + (unituse[PM_FPU] | unituse[PM_IFU] | unituse[PM_ISU1])) { + unituse[PM_ISU0_ALT] = 1; /* move ISU to TTM1 */ + unituse[PM_ISU0] = 0; + } + /* Set TTM[01]SEL fields. */ + ttmuse = 0; + for (i = PM_FPU; i <= PM_ISU1; ++i) { + if (!unituse[i]) + continue; + if (ttmuse++) + return -1; + mmcr1 |= (u64)i << MMCR1_TTM0SEL_SH; + } + ttmuse = 0; + for (; i <= PM_GRS; ++i) { + if (!unituse[i]) + continue; + if (ttmuse++) + return -1; + mmcr1 |= (u64)(i & 3) << MMCR1_TTM1SEL_SH; + } + if (ttmuse > 1) + return -1; + + /* Set byte lane select fields, TTM[23]SEL and GRS_*SEL. */ + for (byte = 0; byte < 4; ++byte) { + unit = busbyte[byte]; + if (!unit) + continue; + if (unit == PM_ISU0 && unituse[PM_ISU0_ALT]) { + /* get ISU0 through TTM1 rather than TTM0 */ + unit = PM_ISU0_ALT; + } else if (unit == PM_LSU1 + 1) { + /* select lower word of LSU1 for this byte */ + mmcr1 |= 1ull << (MMCR1_TTM3SEL_SH + 3 - byte); + } + ttm = unit >> 2; + mmcr1 |= (u64)ttm << (MMCR1_TD_CP_DBG0SEL_SH - 2 * byte); + } + + /* Second pass: assign PMCs, set PMCxSEL and PMCx_ADDER_SEL fields */ + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event[i] >> PM_BYTE_SH) & PM_BYTE_MSK; + psel = event[i] & PM_PMCSEL_MSK; + isbus = event[i] & PM_BUSEVENT_MSK; + if (!pmc) { + /* Bus event or any-PMC direct event */ + for (pmc = 0; pmc < 4; ++pmc) { + if (!(pmc_inuse & (1 << pmc))) + break; + } + if (pmc >= 4) + return -1; + pmc_inuse |= 1 << pmc; + } else if (pmc <= 4) { + /* Direct event */ + --pmc; + if (isbus && (byte & 2) && + (psel == 8 || psel == 0x10 || psel == 0x28)) + /* add events on higher-numbered bus */ + mmcr1 |= 1ull << (MMCR1_PMC1_ADDER_SEL_SH - pmc); + } else { + /* Instructions or run cycles on PMC5/6 */ + --pmc; + } + if (isbus && unit == PM_GRS) { + bit = psel & 7; + grsel = (event[i] >> PM_GRS_SH) & PM_GRS_MSK; + mmcr1 |= (u64)grsel << grsel_shift[bit]; + } + if (power5p_marked_instr_event(event[i])) + mmcra |= MMCRA_SAMPLE_ENABLE; + if ((psel & 0x58) == 0x40 && (byte & 1) != ((pmc >> 1) & 1)) + /* select alternate byte lane */ + psel |= 0x10; + if (pmc <= 3) + mmcr1 |= psel << MMCR1_PMCSEL_SH(pmc); + hwc[i] = pmc; + } + + /* Return MMCRx values */ + mmcr[0] = 0; + if (pmc_inuse & 1) + mmcr[0] = MMCR0_PMC1CE; + if (pmc_inuse & 0x3e) + mmcr[0] |= MMCR0_PMCjCE; + mmcr[1] = mmcr1; + mmcr[2] = mmcra; + return 0; +} + +static void power5p_disable_pmc(unsigned int pmc, u64 mmcr[]) +{ + if (pmc <= 3) + mmcr[1] &= ~(0x7fUL << MMCR1_PMCSEL_SH(pmc)); +} + +static int power5p_generic_events[] = { + [PERF_COUNT_HW_CPU_CYCLES] = 0xf, + [PERF_COUNT_HW_INSTRUCTIONS] = 0x100009, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x1c10a8, /* LD_REF_L1 */ + [PERF_COUNT_HW_CACHE_MISSES] = 0x3c1088, /* LD_MISS_L1 */ + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x230e4, /* BR_ISSUED */ + [PERF_COUNT_HW_BRANCH_MISSES] = 0x230e5, /* BR_MPRED_CR */ +}; + +#define C(x) PERF_COUNT_HW_CACHE_##x + +/* + * Table of generalized cache-related events. + * 0 means not supported, -1 means nonsensical, other values + * are event codes. + */ +static int power5p_cache_events[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = { + [C(L1D)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x1c10a8, 0x3c1088 }, + [C(OP_WRITE)] = { 0x2c10a8, 0xc10c3 }, + [C(OP_PREFETCH)] = { 0xc70e7, -1 }, + }, + [C(L1I)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { 0, 0 }, + }, + [C(LL)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0 }, + [C(OP_WRITE)] = { 0, 0 }, + [C(OP_PREFETCH)] = { 0xc50c3, 0 }, + }, + [C(DTLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0xc20e4, 0x800c4 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(ITLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x800c0 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(BPU)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x230e4, 0x230e5 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, +}; + +struct power_pmu power5p_pmu = { + .n_counter = 6, + .max_alternatives = MAX_ALT, + .add_fields = 0x7000000000055ull, + .test_adder = 0x3000040000000ull, + .compute_mmcr = power5p_compute_mmcr, + .get_constraint = power5p_get_constraint, + .get_alternatives = power5p_get_alternatives, + .disable_pmc = power5p_disable_pmc, + .limited_pmc_event = power5p_limited_pmc_event, + .flags = PPMU_LIMITED_PMC5_6, + .n_generic = ARRAY_SIZE(power5p_generic_events), + .generic_events = power5p_generic_events, + .cache_events = &power5p_cache_events, +}; diff --git a/arch/powerpc/kernel/power5-pmu.c b/arch/powerpc/kernel/power5-pmu.c new file mode 100644 index 00000000000..05600b66221 --- /dev/null +++ b/arch/powerpc/kernel/power5-pmu.c @@ -0,0 +1,611 @@ +/* + * Performance counter support for POWER5 (not POWER5++) processors. + * + * Copyright 2009 Paul Mackerras, IBM Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/kernel.h> +#include <linux/perf_counter.h> +#include <asm/reg.h> + +/* + * Bits in event code for POWER5 (not POWER5++) + */ +#define PM_PMC_SH 20 /* PMC number (1-based) for direct events */ +#define PM_PMC_MSK 0xf +#define PM_PMC_MSKS (PM_PMC_MSK << PM_PMC_SH) +#define PM_UNIT_SH 16 /* TTMMUX number and setting - unit select */ +#define PM_UNIT_MSK 0xf +#define PM_BYTE_SH 12 /* Byte number of event bus to use */ +#define PM_BYTE_MSK 7 +#define PM_GRS_SH 8 /* Storage subsystem mux select */ +#define PM_GRS_MSK 7 +#define PM_BUSEVENT_MSK 0x80 /* Set if event uses event bus */ +#define PM_PMCSEL_MSK 0x7f + +/* Values in PM_UNIT field */ +#define PM_FPU 0 +#define PM_ISU0 1 +#define PM_IFU 2 +#define PM_ISU1 3 +#define PM_IDU 4 +#define PM_ISU0_ALT 6 +#define PM_GRS 7 +#define PM_LSU0 8 +#define PM_LSU1 0xc +#define PM_LASTUNIT 0xc + +/* + * Bits in MMCR1 for POWER5 + */ +#define MMCR1_TTM0SEL_SH 62 +#define MMCR1_TTM1SEL_SH 60 +#define MMCR1_TTM2SEL_SH 58 +#define MMCR1_TTM3SEL_SH 56 +#define MMCR1_TTMSEL_MSK 3 +#define MMCR1_TD_CP_DBG0SEL_SH 54 +#define MMCR1_TD_CP_DBG1SEL_SH 52 +#define MMCR1_TD_CP_DBG2SEL_SH 50 +#define MMCR1_TD_CP_DBG3SEL_SH 48 +#define MMCR1_GRS_L2SEL_SH 46 +#define MMCR1_GRS_L2SEL_MSK 3 +#define MMCR1_GRS_L3SEL_SH 44 +#define MMCR1_GRS_L3SEL_MSK 3 +#define MMCR1_GRS_MCSEL_SH 41 +#define MMCR1_GRS_MCSEL_MSK 7 +#define MMCR1_GRS_FABSEL_SH 39 +#define MMCR1_GRS_FABSEL_MSK 3 +#define MMCR1_PMC1_ADDER_SEL_SH 35 +#define MMCR1_PMC2_ADDER_SEL_SH 34 +#define MMCR1_PMC3_ADDER_SEL_SH 33 +#define MMCR1_PMC4_ADDER_SEL_SH 32 +#define MMCR1_PMC1SEL_SH 25 +#define MMCR1_PMC2SEL_SH 17 +#define MMCR1_PMC3SEL_SH 9 +#define MMCR1_PMC4SEL_SH 1 +#define MMCR1_PMCSEL_SH(n) (MMCR1_PMC1SEL_SH - (n) * 8) +#define MMCR1_PMCSEL_MSK 0x7f + +/* + * Bits in MMCRA + */ + +/* + * Layout of constraint bits: + * 6666555555555544444444443333333333222222222211111111110000000000 + * 3210987654321098765432109876543210987654321098765432109876543210 + * <><>[ ><><>< ><> [ >[ >[ >< >< >< >< ><><><><><><> + * T0T1 NC G0G1G2 G3 UC PS1PS2 B0 B1 B2 B3 P6P5P4P3P2P1 + * + * T0 - TTM0 constraint + * 54-55: TTM0SEL value (0=FPU, 2=IFU, 3=ISU1) 0xc0_0000_0000_0000 + * + * T1 - TTM1 constraint + * 52-53: TTM1SEL value (0=IDU, 3=GRS) 0x30_0000_0000_0000 + * + * NC - number of counters + * 51: NC error 0x0008_0000_0000_0000 + * 48-50: number of events needing PMC1-4 0x0007_0000_0000_0000 + * + * G0..G3 - GRS mux constraints + * 46-47: GRS_L2SEL value + * 44-45: GRS_L3SEL value + * 41-44: GRS_MCSEL value + * 39-40: GRS_FABSEL value + * Note that these match up with their bit positions in MMCR1 + * + * UC - unit constraint: can't have all three of FPU|IFU|ISU1, ISU0, IDU|GRS + * 37: UC3 error 0x20_0000_0000 + * 36: FPU|IFU|ISU1 events needed 0x10_0000_0000 + * 35: ISU0 events needed 0x08_0000_0000 + * 34: IDU|GRS events needed 0x04_0000_0000 + * + * PS1 + * 33: PS1 error 0x2_0000_0000 + * 31-32: count of events needing PMC1/2 0x1_8000_0000 + * + * PS2 + * 30: PS2 error 0x4000_0000 + * 28-29: count of events needing PMC3/4 0x3000_0000 + * + * B0 + * 24-27: Byte 0 event source 0x0f00_0000 + * Encoding as for the event code + * + * B1, B2, B3 + * 20-23, 16-19, 12-15: Byte 1, 2, 3 event sources + * + * P1..P6 + * 0-11: Count of events needing PMC1..PMC6 + */ + +static const int grsel_shift[8] = { + MMCR1_GRS_L2SEL_SH, MMCR1_GRS_L2SEL_SH, MMCR1_GRS_L2SEL_SH, + MMCR1_GRS_L3SEL_SH, MMCR1_GRS_L3SEL_SH, MMCR1_GRS_L3SEL_SH, + MMCR1_GRS_MCSEL_SH, MMCR1_GRS_FABSEL_SH +}; + +/* Masks and values for using events from the various units */ +static u64 unit_cons[PM_LASTUNIT+1][2] = { + [PM_FPU] = { 0xc0002000000000ull, 0x00001000000000ull }, + [PM_ISU0] = { 0x00002000000000ull, 0x00000800000000ull }, + [PM_ISU1] = { 0xc0002000000000ull, 0xc0001000000000ull }, + [PM_IFU] = { 0xc0002000000000ull, 0x80001000000000ull }, + [PM_IDU] = { 0x30002000000000ull, 0x00000400000000ull }, + [PM_GRS] = { 0x30002000000000ull, 0x30000400000000ull }, +}; + +static int power5_get_constraint(u64 event, u64 *maskp, u64 *valp) +{ + int pmc, byte, unit, sh; + int bit, fmask; + u64 mask = 0, value = 0; + int grp = -1; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 6) + return -1; + sh = (pmc - 1) * 2; + mask |= 2 << sh; + value |= 1 << sh; + if (pmc <= 4) + grp = (pmc - 1) >> 1; + else if (event != 0x500009 && event != 0x600005) + return -1; + } + if (event & PM_BUSEVENT_MSK) { + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + if (unit > PM_LASTUNIT) + return -1; + if (unit == PM_ISU0_ALT) + unit = PM_ISU0; + mask |= unit_cons[unit][0]; + value |= unit_cons[unit][1]; + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + if (byte >= 4) { + if (unit != PM_LSU1) + return -1; + /* Map LSU1 low word (bytes 4-7) to unit LSU1+1 */ + ++unit; + byte &= 3; + } + if (unit == PM_GRS) { + bit = event & 7; + fmask = (bit == 6)? 7: 3; + sh = grsel_shift[bit]; + mask |= (u64)fmask << sh; + value |= (u64)((event >> PM_GRS_SH) & fmask) << sh; + } + /* + * Bus events on bytes 0 and 2 can be counted + * on PMC1/2; bytes 1 and 3 on PMC3/4. + */ + if (!pmc) + grp = byte & 1; + /* Set byte lane select field */ + mask |= 0xfULL << (24 - 4 * byte); + value |= (u64)unit << (24 - 4 * byte); + } + if (grp == 0) { + /* increment PMC1/2 field */ + mask |= 0x200000000ull; + value |= 0x080000000ull; + } else if (grp == 1) { + /* increment PMC3/4 field */ + mask |= 0x40000000ull; + value |= 0x10000000ull; + } + if (pmc < 5) { + /* need a counter from PMC1-4 set */ + mask |= 0x8000000000000ull; + value |= 0x1000000000000ull; + } + *maskp = mask; + *valp = value; + return 0; +} + +#define MAX_ALT 3 /* at most 3 alternatives for any event */ + +static const unsigned int event_alternatives[][MAX_ALT] = { + { 0x120e4, 0x400002 }, /* PM_GRP_DISP_REJECT */ + { 0x410c7, 0x441084 }, /* PM_THRD_L2MISS_BOTH_CYC */ + { 0x100005, 0x600005 }, /* PM_RUN_CYC */ + { 0x100009, 0x200009, 0x500009 }, /* PM_INST_CMPL */ + { 0x300009, 0x400009 }, /* PM_INST_DISP */ +}; + +/* + * Scan the alternatives table for a match and return the + * index into the alternatives table if found, else -1. + */ +static int find_alternative(u64 event) +{ + int i, j; + + for (i = 0; i < ARRAY_SIZE(event_alternatives); ++i) { + if (event < event_alternatives[i][0]) + break; + for (j = 0; j < MAX_ALT && event_alternatives[i][j]; ++j) + if (event == event_alternatives[i][j]) + return i; + } + return -1; +} + +static const unsigned char bytedecode_alternatives[4][4] = { + /* PMC 1 */ { 0x21, 0x23, 0x25, 0x27 }, + /* PMC 2 */ { 0x07, 0x17, 0x0e, 0x1e }, + /* PMC 3 */ { 0x20, 0x22, 0x24, 0x26 }, + /* PMC 4 */ { 0x07, 0x17, 0x0e, 0x1e } +}; + +/* + * Some direct events for decodes of event bus byte 3 have alternative + * PMCSEL values on other counters. This returns the alternative + * event code for those that do, or -1 otherwise. + */ +static s64 find_alternative_bdecode(u64 event) +{ + int pmc, altpmc, pp, j; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc == 0 || pmc > 4) + return -1; + altpmc = 5 - pmc; /* 1 <-> 4, 2 <-> 3 */ + pp = event & PM_PMCSEL_MSK; + for (j = 0; j < 4; ++j) { + if (bytedecode_alternatives[pmc - 1][j] == pp) { + return (event & ~(PM_PMC_MSKS | PM_PMCSEL_MSK)) | + (altpmc << PM_PMC_SH) | + bytedecode_alternatives[altpmc - 1][j]; + } + } + return -1; +} + +static int power5_get_alternatives(u64 event, unsigned int flags, u64 alt[]) +{ + int i, j, nalt = 1; + s64 ae; + + alt[0] = event; + nalt = 1; + i = find_alternative(event); + if (i >= 0) { + for (j = 0; j < MAX_ALT; ++j) { + ae = event_alternatives[i][j]; + if (ae && ae != event) + alt[nalt++] = ae; + } + } else { + ae = find_alternative_bdecode(event); + if (ae > 0) + alt[nalt++] = ae; + } + return nalt; +} + +/* + * Map of which direct events on which PMCs are marked instruction events. + * Indexed by PMCSEL value, bit i (LE) set if PMC i is a marked event. + * Bit 0 is set if it is marked for all PMCs. + * The 0x80 bit indicates a byte decode PMCSEL value. + */ +static unsigned char direct_event_is_marked[0x28] = { + 0, /* 00 */ + 0x1f, /* 01 PM_IOPS_CMPL */ + 0x2, /* 02 PM_MRK_GRP_DISP */ + 0xe, /* 03 PM_MRK_ST_CMPL, PM_MRK_ST_GPS, PM_MRK_ST_CMPL_INT */ + 0, /* 04 */ + 0x1c, /* 05 PM_MRK_BRU_FIN, PM_MRK_INST_FIN, PM_MRK_CRU_FIN */ + 0x80, /* 06 */ + 0x80, /* 07 */ + 0, 0, 0,/* 08 - 0a */ + 0x18, /* 0b PM_THRESH_TIMEO, PM_MRK_GRP_TIMEO */ + 0, /* 0c */ + 0x80, /* 0d */ + 0x80, /* 0e */ + 0, /* 0f */ + 0, /* 10 */ + 0x14, /* 11 PM_MRK_GRP_BR_REDIR, PM_MRK_GRP_IC_MISS */ + 0, /* 12 */ + 0x10, /* 13 PM_MRK_GRP_CMPL */ + 0x1f, /* 14 PM_GRP_MRK, PM_MRK_{FXU,FPU,LSU}_FIN */ + 0x2, /* 15 PM_MRK_GRP_ISSUED */ + 0x80, /* 16 */ + 0x80, /* 17 */ + 0, 0, 0, 0, 0, + 0x80, /* 1d */ + 0x80, /* 1e */ + 0, /* 1f */ + 0x80, /* 20 */ + 0x80, /* 21 */ + 0x80, /* 22 */ + 0x80, /* 23 */ + 0x80, /* 24 */ + 0x80, /* 25 */ + 0x80, /* 26 */ + 0x80, /* 27 */ +}; + +/* + * Returns 1 if event counts things relating to marked instructions + * and thus needs the MMCRA_SAMPLE_ENABLE bit set, or 0 if not. + */ +static int power5_marked_instr_event(u64 event) +{ + int pmc, psel; + int bit, byte, unit; + u32 mask; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + psel = event & PM_PMCSEL_MSK; + if (pmc >= 5) + return 0; + + bit = -1; + if (psel < sizeof(direct_event_is_marked)) { + if (direct_event_is_marked[psel] & (1 << pmc)) + return 1; + if (direct_event_is_marked[psel] & 0x80) + bit = 4; + else if (psel == 0x08) + bit = pmc - 1; + else if (psel == 0x10) + bit = 4 - pmc; + else if (psel == 0x1b && (pmc == 1 || pmc == 3)) + bit = 4; + } else if ((psel & 0x58) == 0x40) + bit = psel & 7; + + if (!(event & PM_BUSEVENT_MSK)) + return 0; + + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + if (unit == PM_LSU0) { + /* byte 1 bits 0-7, byte 2 bits 0,2-4,6 */ + mask = 0x5dff00; + } else if (unit == PM_LSU1 && byte >= 4) { + byte -= 4; + /* byte 4 bits 1,3,5,7, byte 5 bits 6-7, byte 7 bits 0-4,6 */ + mask = 0x5f00c0aa; + } else + return 0; + + return (mask >> (byte * 8 + bit)) & 1; +} + +static int power5_compute_mmcr(u64 event[], int n_ev, + unsigned int hwc[], u64 mmcr[]) +{ + u64 mmcr1 = 0; + u64 mmcra = 0; + unsigned int pmc, unit, byte, psel; + unsigned int ttm, grp; + int i, isbus, bit, grsel; + unsigned int pmc_inuse = 0; + unsigned int pmc_grp_use[2]; + unsigned char busbyte[4]; + unsigned char unituse[16]; + int ttmuse; + + if (n_ev > 6) + return -1; + + /* First pass to count resource use */ + pmc_grp_use[0] = pmc_grp_use[1] = 0; + memset(busbyte, 0, sizeof(busbyte)); + memset(unituse, 0, sizeof(unituse)); + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 6) + return -1; + if (pmc_inuse & (1 << (pmc - 1))) + return -1; + pmc_inuse |= 1 << (pmc - 1); + /* count 1/2 vs 3/4 use */ + if (pmc <= 4) + ++pmc_grp_use[(pmc - 1) >> 1]; + } + if (event[i] & PM_BUSEVENT_MSK) { + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event[i] >> PM_BYTE_SH) & PM_BYTE_MSK; + if (unit > PM_LASTUNIT) + return -1; + if (unit == PM_ISU0_ALT) + unit = PM_ISU0; + if (byte >= 4) { + if (unit != PM_LSU1) + return -1; + ++unit; + byte &= 3; + } + if (!pmc) + ++pmc_grp_use[byte & 1]; + if (busbyte[byte] && busbyte[byte] != unit) + return -1; + busbyte[byte] = unit; + unituse[unit] = 1; + } + } + if (pmc_grp_use[0] > 2 || pmc_grp_use[1] > 2) + return -1; + + /* + * Assign resources and set multiplexer selects. + * + * PM_ISU0 can go either on TTM0 or TTM1, but that's the only + * choice we have to deal with. + */ + if (unituse[PM_ISU0] & + (unituse[PM_FPU] | unituse[PM_IFU] | unituse[PM_ISU1])) { + unituse[PM_ISU0_ALT] = 1; /* move ISU to TTM1 */ + unituse[PM_ISU0] = 0; + } + /* Set TTM[01]SEL fields. */ + ttmuse = 0; + for (i = PM_FPU; i <= PM_ISU1; ++i) { + if (!unituse[i]) + continue; + if (ttmuse++) + return -1; + mmcr1 |= (u64)i << MMCR1_TTM0SEL_SH; + } + ttmuse = 0; + for (; i <= PM_GRS; ++i) { + if (!unituse[i]) + continue; + if (ttmuse++) + return -1; + mmcr1 |= (u64)(i & 3) << MMCR1_TTM1SEL_SH; + } + if (ttmuse > 1) + return -1; + + /* Set byte lane select fields, TTM[23]SEL and GRS_*SEL. */ + for (byte = 0; byte < 4; ++byte) { + unit = busbyte[byte]; + if (!unit) + continue; + if (unit == PM_ISU0 && unituse[PM_ISU0_ALT]) { + /* get ISU0 through TTM1 rather than TTM0 */ + unit = PM_ISU0_ALT; + } else if (unit == PM_LSU1 + 1) { + /* select lower word of LSU1 for this byte */ + mmcr1 |= 1ull << (MMCR1_TTM3SEL_SH + 3 - byte); + } + ttm = unit >> 2; + mmcr1 |= (u64)ttm << (MMCR1_TD_CP_DBG0SEL_SH - 2 * byte); + } + + /* Second pass: assign PMCs, set PMCxSEL and PMCx_ADDER_SEL fields */ + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event[i] >> PM_BYTE_SH) & PM_BYTE_MSK; + psel = event[i] & PM_PMCSEL_MSK; + isbus = event[i] & PM_BUSEVENT_MSK; + if (!pmc) { + /* Bus event or any-PMC direct event */ + for (pmc = 0; pmc < 4; ++pmc) { + if (pmc_inuse & (1 << pmc)) + continue; + grp = (pmc >> 1) & 1; + if (isbus) { + if (grp == (byte & 1)) + break; + } else if (pmc_grp_use[grp] < 2) { + ++pmc_grp_use[grp]; + break; + } + } + pmc_inuse |= 1 << pmc; + } else if (pmc <= 4) { + /* Direct event */ + --pmc; + if ((psel == 8 || psel == 0x10) && isbus && (byte & 2)) + /* add events on higher-numbered bus */ + mmcr1 |= 1ull << (MMCR1_PMC1_ADDER_SEL_SH - pmc); + } else { + /* Instructions or run cycles on PMC5/6 */ + --pmc; + } + if (isbus && unit == PM_GRS) { + bit = psel & 7; + grsel = (event[i] >> PM_GRS_SH) & PM_GRS_MSK; + mmcr1 |= (u64)grsel << grsel_shift[bit]; + } + if (power5_marked_instr_event(event[i])) + mmcra |= MMCRA_SAMPLE_ENABLE; + if (pmc <= 3) + mmcr1 |= psel << MMCR1_PMCSEL_SH(pmc); + hwc[i] = pmc; + } + + /* Return MMCRx values */ + mmcr[0] = 0; + if (pmc_inuse & 1) + mmcr[0] = MMCR0_PMC1CE; + if (pmc_inuse & 0x3e) + mmcr[0] |= MMCR0_PMCjCE; + mmcr[1] = mmcr1; + mmcr[2] = mmcra; + return 0; +} + +static void power5_disable_pmc(unsigned int pmc, u64 mmcr[]) +{ + if (pmc <= 3) + mmcr[1] &= ~(0x7fUL << MMCR1_PMCSEL_SH(pmc)); +} + +static int power5_generic_events[] = { + [PERF_COUNT_HW_CPU_CYCLES] = 0xf, + [PERF_COUNT_HW_INSTRUCTIONS] = 0x100009, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x4c1090, /* LD_REF_L1 */ + [PERF_COUNT_HW_CACHE_MISSES] = 0x3c1088, /* LD_MISS_L1 */ + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x230e4, /* BR_ISSUED */ + [PERF_COUNT_HW_BRANCH_MISSES] = 0x230e5, /* BR_MPRED_CR */ +}; + +#define C(x) PERF_COUNT_HW_CACHE_##x + +/* + * Table of generalized cache-related events. + * 0 means not supported, -1 means nonsensical, other values + * are event codes. + */ +static int power5_cache_events[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = { + [C(L1D)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x4c1090, 0x3c1088 }, + [C(OP_WRITE)] = { 0x3c1090, 0xc10c3 }, + [C(OP_PREFETCH)] = { 0xc70e7, 0 }, + }, + [C(L1I)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { 0, 0 }, + }, + [C(LL)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x3c309b }, + [C(OP_WRITE)] = { 0, 0 }, + [C(OP_PREFETCH)] = { 0xc50c3, 0 }, + }, + [C(DTLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x2c4090, 0x800c4 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(ITLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x800c0 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(BPU)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x230e4, 0x230e5 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, +}; + +struct power_pmu power5_pmu = { + .n_counter = 6, + .max_alternatives = MAX_ALT, + .add_fields = 0x7000090000555ull, + .test_adder = 0x3000490000000ull, + .compute_mmcr = power5_compute_mmcr, + .get_constraint = power5_get_constraint, + .get_alternatives = power5_get_alternatives, + .disable_pmc = power5_disable_pmc, + .n_generic = ARRAY_SIZE(power5_generic_events), + .generic_events = power5_generic_events, + .cache_events = &power5_cache_events, +}; diff --git a/arch/powerpc/kernel/power6-pmu.c b/arch/powerpc/kernel/power6-pmu.c new file mode 100644 index 00000000000..46f74bebcfd --- /dev/null +++ b/arch/powerpc/kernel/power6-pmu.c @@ -0,0 +1,532 @@ +/* + * Performance counter support for POWER6 processors. + * + * Copyright 2008-2009 Paul Mackerras, IBM Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/kernel.h> +#include <linux/perf_counter.h> +#include <asm/reg.h> + +/* + * Bits in event code for POWER6 + */ +#define PM_PMC_SH 20 /* PMC number (1-based) for direct events */ +#define PM_PMC_MSK 0x7 +#define PM_PMC_MSKS (PM_PMC_MSK << PM_PMC_SH) +#define PM_UNIT_SH 16 /* Unit event comes (TTMxSEL encoding) */ +#define PM_UNIT_MSK 0xf +#define PM_UNIT_MSKS (PM_UNIT_MSK << PM_UNIT_SH) +#define PM_LLAV 0x8000 /* Load lookahead match value */ +#define PM_LLA 0x4000 /* Load lookahead match enable */ +#define PM_BYTE_SH 12 /* Byte of event bus to use */ +#define PM_BYTE_MSK 3 +#define PM_SUBUNIT_SH 8 /* Subunit event comes from (NEST_SEL enc.) */ +#define PM_SUBUNIT_MSK 7 +#define PM_SUBUNIT_MSKS (PM_SUBUNIT_MSK << PM_SUBUNIT_SH) +#define PM_PMCSEL_MSK 0xff /* PMCxSEL value */ +#define PM_BUSEVENT_MSK 0xf3700 + +/* + * Bits in MMCR1 for POWER6 + */ +#define MMCR1_TTM0SEL_SH 60 +#define MMCR1_TTMSEL_SH(n) (MMCR1_TTM0SEL_SH - (n) * 4) +#define MMCR1_TTMSEL_MSK 0xf +#define MMCR1_TTMSEL(m, n) (((m) >> MMCR1_TTMSEL_SH(n)) & MMCR1_TTMSEL_MSK) +#define MMCR1_NESTSEL_SH 45 +#define MMCR1_NESTSEL_MSK 0x7 +#define MMCR1_NESTSEL(m) (((m) >> MMCR1_NESTSEL_SH) & MMCR1_NESTSEL_MSK) +#define MMCR1_PMC1_LLA ((u64)1 << 44) +#define MMCR1_PMC1_LLA_VALUE ((u64)1 << 39) +#define MMCR1_PMC1_ADDR_SEL ((u64)1 << 35) +#define MMCR1_PMC1SEL_SH 24 +#define MMCR1_PMCSEL_SH(n) (MMCR1_PMC1SEL_SH - (n) * 8) +#define MMCR1_PMCSEL_MSK 0xff + +/* + * Map of which direct events on which PMCs are marked instruction events. + * Indexed by PMCSEL value >> 1. + * Bottom 4 bits are a map of which PMCs are interesting, + * top 4 bits say what sort of event: + * 0 = direct marked event, + * 1 = byte decode event, + * 4 = add/and event (PMC1 -> bits 0 & 4), + * 5 = add/and event (PMC1 -> bits 1 & 5), + * 6 = add/and event (PMC1 -> bits 2 & 6), + * 7 = add/and event (PMC1 -> bits 3 & 7). + */ +static unsigned char direct_event_is_marked[0x60 >> 1] = { + 0, /* 00 */ + 0, /* 02 */ + 0, /* 04 */ + 0x07, /* 06 PM_MRK_ST_CMPL, PM_MRK_ST_GPS, PM_MRK_ST_CMPL_INT */ + 0x04, /* 08 PM_MRK_DFU_FIN */ + 0x06, /* 0a PM_MRK_IFU_FIN, PM_MRK_INST_FIN */ + 0, /* 0c */ + 0, /* 0e */ + 0x02, /* 10 PM_MRK_INST_DISP */ + 0x08, /* 12 PM_MRK_LSU_DERAT_MISS */ + 0, /* 14 */ + 0, /* 16 */ + 0x0c, /* 18 PM_THRESH_TIMEO, PM_MRK_INST_FIN */ + 0x0f, /* 1a PM_MRK_INST_DISP, PM_MRK_{FXU,FPU,LSU}_FIN */ + 0x01, /* 1c PM_MRK_INST_ISSUED */ + 0, /* 1e */ + 0, /* 20 */ + 0, /* 22 */ + 0, /* 24 */ + 0, /* 26 */ + 0x15, /* 28 PM_MRK_DATA_FROM_L2MISS, PM_MRK_DATA_FROM_L3MISS */ + 0, /* 2a */ + 0, /* 2c */ + 0, /* 2e */ + 0x4f, /* 30 */ + 0x7f, /* 32 */ + 0x4f, /* 34 */ + 0x5f, /* 36 */ + 0x6f, /* 38 */ + 0x4f, /* 3a */ + 0, /* 3c */ + 0x08, /* 3e PM_MRK_INST_TIMEO */ + 0x1f, /* 40 */ + 0x1f, /* 42 */ + 0x1f, /* 44 */ + 0x1f, /* 46 */ + 0x1f, /* 48 */ + 0x1f, /* 4a */ + 0x1f, /* 4c */ + 0x1f, /* 4e */ + 0, /* 50 */ + 0x05, /* 52 PM_MRK_BR_TAKEN, PM_MRK_BR_MPRED */ + 0x1c, /* 54 PM_MRK_PTEG_FROM_L3MISS, PM_MRK_PTEG_FROM_L2MISS */ + 0x02, /* 56 PM_MRK_LD_MISS_L1 */ + 0, /* 58 */ + 0, /* 5a */ + 0, /* 5c */ + 0, /* 5e */ +}; + +/* + * Masks showing for each unit which bits are marked events. + * These masks are in LE order, i.e. 0x00000001 is byte 0, bit 0. + */ +static u32 marked_bus_events[16] = { + 0x01000000, /* direct events set 1: byte 3 bit 0 */ + 0x00010000, /* direct events set 2: byte 2 bit 0 */ + 0, 0, 0, 0, /* IDU, IFU, nest: nothing */ + 0x00000088, /* VMX set 1: byte 0 bits 3, 7 */ + 0x000000c0, /* VMX set 2: byte 0 bits 4-7 */ + 0x04010000, /* LSU set 1: byte 2 bit 0, byte 3 bit 2 */ + 0xff010000u, /* LSU set 2: byte 2 bit 0, all of byte 3 */ + 0, /* LSU set 3 */ + 0x00000010, /* VMX set 3: byte 0 bit 4 */ + 0, /* BFP set 1 */ + 0x00000022, /* BFP set 2: byte 0 bits 1, 5 */ + 0, 0 +}; + +/* + * Returns 1 if event counts things relating to marked instructions + * and thus needs the MMCRA_SAMPLE_ENABLE bit set, or 0 if not. + */ +static int power6_marked_instr_event(u64 event) +{ + int pmc, psel, ptype; + int bit, byte, unit; + u32 mask; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + psel = (event & PM_PMCSEL_MSK) >> 1; /* drop edge/level bit */ + if (pmc >= 5) + return 0; + + bit = -1; + if (psel < sizeof(direct_event_is_marked)) { + ptype = direct_event_is_marked[psel]; + if (pmc == 0 || !(ptype & (1 << (pmc - 1)))) + return 0; + ptype >>= 4; + if (ptype == 0) + return 1; + if (ptype == 1) + bit = 0; + else + bit = ptype ^ (pmc - 1); + } else if ((psel & 0x48) == 0x40) + bit = psel & 7; + + if (!(event & PM_BUSEVENT_MSK) || bit == -1) + return 0; + + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + mask = marked_bus_events[unit]; + return (mask >> (byte * 8 + bit)) & 1; +} + +/* + * Assign PMC numbers and compute MMCR1 value for a set of events + */ +static int p6_compute_mmcr(u64 event[], int n_ev, + unsigned int hwc[], u64 mmcr[]) +{ + u64 mmcr1 = 0; + u64 mmcra = 0; + int i; + unsigned int pmc, ev, b, u, s, psel; + unsigned int ttmset = 0; + unsigned int pmc_inuse = 0; + + if (n_ev > 6) + return -1; + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc_inuse & (1 << (pmc - 1))) + return -1; /* collision! */ + pmc_inuse |= 1 << (pmc - 1); + } + } + for (i = 0; i < n_ev; ++i) { + ev = event[i]; + pmc = (ev >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + --pmc; + } else { + /* can go on any PMC; find a free one */ + for (pmc = 0; pmc < 4; ++pmc) + if (!(pmc_inuse & (1 << pmc))) + break; + if (pmc >= 4) + return -1; + pmc_inuse |= 1 << pmc; + } + hwc[i] = pmc; + psel = ev & PM_PMCSEL_MSK; + if (ev & PM_BUSEVENT_MSK) { + /* this event uses the event bus */ + b = (ev >> PM_BYTE_SH) & PM_BYTE_MSK; + u = (ev >> PM_UNIT_SH) & PM_UNIT_MSK; + /* check for conflict on this byte of event bus */ + if ((ttmset & (1 << b)) && MMCR1_TTMSEL(mmcr1, b) != u) + return -1; + mmcr1 |= (u64)u << MMCR1_TTMSEL_SH(b); + ttmset |= 1 << b; + if (u == 5) { + /* Nest events have a further mux */ + s = (ev >> PM_SUBUNIT_SH) & PM_SUBUNIT_MSK; + if ((ttmset & 0x10) && + MMCR1_NESTSEL(mmcr1) != s) + return -1; + ttmset |= 0x10; + mmcr1 |= (u64)s << MMCR1_NESTSEL_SH; + } + if (0x30 <= psel && psel <= 0x3d) { + /* these need the PMCx_ADDR_SEL bits */ + if (b >= 2) + mmcr1 |= MMCR1_PMC1_ADDR_SEL >> pmc; + } + /* bus select values are different for PMC3/4 */ + if (pmc >= 2 && (psel & 0x90) == 0x80) + psel ^= 0x20; + } + if (ev & PM_LLA) { + mmcr1 |= MMCR1_PMC1_LLA >> pmc; + if (ev & PM_LLAV) + mmcr1 |= MMCR1_PMC1_LLA_VALUE >> pmc; + } + if (power6_marked_instr_event(event[i])) + mmcra |= MMCRA_SAMPLE_ENABLE; + if (pmc < 4) + mmcr1 |= (u64)psel << MMCR1_PMCSEL_SH(pmc); + } + mmcr[0] = 0; + if (pmc_inuse & 1) + mmcr[0] = MMCR0_PMC1CE; + if (pmc_inuse & 0xe) + mmcr[0] |= MMCR0_PMCjCE; + mmcr[1] = mmcr1; + mmcr[2] = mmcra; + return 0; +} + +/* + * Layout of constraint bits: + * + * 0-1 add field: number of uses of PMC1 (max 1) + * 2-3, 4-5, 6-7, 8-9, 10-11: ditto for PMC2, 3, 4, 5, 6 + * 12-15 add field: number of uses of PMC1-4 (max 4) + * 16-19 select field: unit on byte 0 of event bus + * 20-23, 24-27, 28-31 ditto for bytes 1, 2, 3 + * 32-34 select field: nest (subunit) event selector + */ +static int p6_get_constraint(u64 event, u64 *maskp, u64 *valp) +{ + int pmc, byte, sh, subunit; + u64 mask = 0, value = 0; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 4 && !(event == 0x500009 || event == 0x600005)) + return -1; + sh = (pmc - 1) * 2; + mask |= 2 << sh; + value |= 1 << sh; + } + if (event & PM_BUSEVENT_MSK) { + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + sh = byte * 4 + (16 - PM_UNIT_SH); + mask |= PM_UNIT_MSKS << sh; + value |= (u64)(event & PM_UNIT_MSKS) << sh; + if ((event & PM_UNIT_MSKS) == (5 << PM_UNIT_SH)) { + subunit = (event >> PM_SUBUNIT_SH) & PM_SUBUNIT_MSK; + mask |= (u64)PM_SUBUNIT_MSK << 32; + value |= (u64)subunit << 32; + } + } + if (pmc <= 4) { + mask |= 0x8000; /* add field for count of PMC1-4 uses */ + value |= 0x1000; + } + *maskp = mask; + *valp = value; + return 0; +} + +static int p6_limited_pmc_event(u64 event) +{ + int pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + + return pmc == 5 || pmc == 6; +} + +#define MAX_ALT 4 /* at most 4 alternatives for any event */ + +static const unsigned int event_alternatives[][MAX_ALT] = { + { 0x0130e8, 0x2000f6, 0x3000fc }, /* PM_PTEG_RELOAD_VALID */ + { 0x080080, 0x10000d, 0x30000c, 0x4000f0 }, /* PM_LD_MISS_L1 */ + { 0x080088, 0x200054, 0x3000f0 }, /* PM_ST_MISS_L1 */ + { 0x10000a, 0x2000f4, 0x600005 }, /* PM_RUN_CYC */ + { 0x10000b, 0x2000f5 }, /* PM_RUN_COUNT */ + { 0x10000e, 0x400010 }, /* PM_PURR */ + { 0x100010, 0x4000f8 }, /* PM_FLUSH */ + { 0x10001a, 0x200010 }, /* PM_MRK_INST_DISP */ + { 0x100026, 0x3000f8 }, /* PM_TB_BIT_TRANS */ + { 0x100054, 0x2000f0 }, /* PM_ST_FIN */ + { 0x100056, 0x2000fc }, /* PM_L1_ICACHE_MISS */ + { 0x1000f0, 0x40000a }, /* PM_INST_IMC_MATCH_CMPL */ + { 0x1000f8, 0x200008 }, /* PM_GCT_EMPTY_CYC */ + { 0x1000fc, 0x400006 }, /* PM_LSU_DERAT_MISS_CYC */ + { 0x20000e, 0x400007 }, /* PM_LSU_DERAT_MISS */ + { 0x200012, 0x300012 }, /* PM_INST_DISP */ + { 0x2000f2, 0x3000f2 }, /* PM_INST_DISP */ + { 0x2000f8, 0x300010 }, /* PM_EXT_INT */ + { 0x2000fe, 0x300056 }, /* PM_DATA_FROM_L2MISS */ + { 0x2d0030, 0x30001a }, /* PM_MRK_FPU_FIN */ + { 0x30000a, 0x400018 }, /* PM_MRK_INST_FIN */ + { 0x3000f6, 0x40000e }, /* PM_L1_DCACHE_RELOAD_VALID */ + { 0x3000fe, 0x400056 }, /* PM_DATA_FROM_L3MISS */ +}; + +/* + * This could be made more efficient with a binary search on + * a presorted list, if necessary + */ +static int find_alternatives_list(u64 event) +{ + int i, j; + unsigned int alt; + + for (i = 0; i < ARRAY_SIZE(event_alternatives); ++i) { + if (event < event_alternatives[i][0]) + return -1; + for (j = 0; j < MAX_ALT; ++j) { + alt = event_alternatives[i][j]; + if (!alt || event < alt) + break; + if (event == alt) + return i; + } + } + return -1; +} + +static int p6_get_alternatives(u64 event, unsigned int flags, u64 alt[]) +{ + int i, j, nlim; + unsigned int psel, pmc; + unsigned int nalt = 1; + u64 aevent; + + alt[0] = event; + nlim = p6_limited_pmc_event(event); + + /* check the alternatives table */ + i = find_alternatives_list(event); + if (i >= 0) { + /* copy out alternatives from list */ + for (j = 0; j < MAX_ALT; ++j) { + aevent = event_alternatives[i][j]; + if (!aevent) + break; + if (aevent != event) + alt[nalt++] = aevent; + nlim += p6_limited_pmc_event(aevent); + } + + } else { + /* Check for alternative ways of computing sum events */ + /* PMCSEL 0x32 counter N == PMCSEL 0x34 counter 5-N */ + psel = event & (PM_PMCSEL_MSK & ~1); /* ignore edge bit */ + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc && (psel == 0x32 || psel == 0x34)) + alt[nalt++] = ((event ^ 0x6) & ~PM_PMC_MSKS) | + ((5 - pmc) << PM_PMC_SH); + + /* PMCSEL 0x38 counter N == PMCSEL 0x3a counter N+/-2 */ + if (pmc && (psel == 0x38 || psel == 0x3a)) + alt[nalt++] = ((event ^ 0x2) & ~PM_PMC_MSKS) | + ((pmc > 2? pmc - 2: pmc + 2) << PM_PMC_SH); + } + + if (flags & PPMU_ONLY_COUNT_RUN) { + /* + * We're only counting in RUN state, + * so PM_CYC is equivalent to PM_RUN_CYC, + * PM_INST_CMPL === PM_RUN_INST_CMPL, PM_PURR === PM_RUN_PURR. + * This doesn't include alternatives that don't provide + * any extra flexibility in assigning PMCs (e.g. + * 0x10000a for PM_RUN_CYC vs. 0x1e for PM_CYC). + * Note that even with these additional alternatives + * we never end up with more than 4 alternatives for any event. + */ + j = nalt; + for (i = 0; i < nalt; ++i) { + switch (alt[i]) { + case 0x1e: /* PM_CYC */ + alt[j++] = 0x600005; /* PM_RUN_CYC */ + ++nlim; + break; + case 0x10000a: /* PM_RUN_CYC */ + alt[j++] = 0x1e; /* PM_CYC */ + break; + case 2: /* PM_INST_CMPL */ + alt[j++] = 0x500009; /* PM_RUN_INST_CMPL */ + ++nlim; + break; + case 0x500009: /* PM_RUN_INST_CMPL */ + alt[j++] = 2; /* PM_INST_CMPL */ + break; + case 0x10000e: /* PM_PURR */ + alt[j++] = 0x4000f4; /* PM_RUN_PURR */ + break; + case 0x4000f4: /* PM_RUN_PURR */ + alt[j++] = 0x10000e; /* PM_PURR */ + break; + } + } + nalt = j; + } + + if (!(flags & PPMU_LIMITED_PMC_OK) && nlim) { + /* remove the limited PMC events */ + j = 0; + for (i = 0; i < nalt; ++i) { + if (!p6_limited_pmc_event(alt[i])) { + alt[j] = alt[i]; + ++j; + } + } + nalt = j; + } else if ((flags & PPMU_LIMITED_PMC_REQD) && nlim < nalt) { + /* remove all but the limited PMC events */ + j = 0; + for (i = 0; i < nalt; ++i) { + if (p6_limited_pmc_event(alt[i])) { + alt[j] = alt[i]; + ++j; + } + } + nalt = j; + } + + return nalt; +} + +static void p6_disable_pmc(unsigned int pmc, u64 mmcr[]) +{ + /* Set PMCxSEL to 0 to disable PMCx */ + if (pmc <= 3) + mmcr[1] &= ~(0xffUL << MMCR1_PMCSEL_SH(pmc)); +} + +static int power6_generic_events[] = { + [PERF_COUNT_HW_CPU_CYCLES] = 0x1e, + [PERF_COUNT_HW_INSTRUCTIONS] = 2, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x280030, /* LD_REF_L1 */ + [PERF_COUNT_HW_CACHE_MISSES] = 0x30000c, /* LD_MISS_L1 */ + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x410a0, /* BR_PRED */ + [PERF_COUNT_HW_BRANCH_MISSES] = 0x400052, /* BR_MPRED */ +}; + +#define C(x) PERF_COUNT_HW_CACHE_##x + +/* + * Table of generalized cache-related events. + * 0 means not supported, -1 means nonsensical, other values + * are event codes. + * The "DTLB" and "ITLB" events relate to the DERAT and IERAT. + */ +static int power6_cache_events[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = { + [C(L1D)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x80082, 0x80080 }, + [C(OP_WRITE)] = { 0x80086, 0x80088 }, + [C(OP_PREFETCH)] = { 0x810a4, 0 }, + }, + [C(L1I)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x100056 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { 0x4008c, 0 }, + }, + [C(LL)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x150730, 0x250532 }, + [C(OP_WRITE)] = { 0x250432, 0x150432 }, + [C(OP_PREFETCH)] = { 0x810a6, 0 }, + }, + [C(DTLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x20000e }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(ITLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x420ce }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(BPU)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x430e6, 0x400052 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, +}; + +struct power_pmu power6_pmu = { + .n_counter = 6, + .max_alternatives = MAX_ALT, + .add_fields = 0x1555, + .test_adder = 0x3000, + .compute_mmcr = p6_compute_mmcr, + .get_constraint = p6_get_constraint, + .get_alternatives = p6_get_alternatives, + .disable_pmc = p6_disable_pmc, + .limited_pmc_event = p6_limited_pmc_event, + .flags = PPMU_LIMITED_PMC5_6 | PPMU_ALT_SIPR, + .n_generic = ARRAY_SIZE(power6_generic_events), + .generic_events = power6_generic_events, + .cache_events = &power6_cache_events, +}; diff --git a/arch/powerpc/kernel/power7-pmu.c b/arch/powerpc/kernel/power7-pmu.c new file mode 100644 index 00000000000..b3f7d1216ba --- /dev/null +++ b/arch/powerpc/kernel/power7-pmu.c @@ -0,0 +1,357 @@ +/* + * Performance counter support for POWER7 processors. + * + * Copyright 2009 Paul Mackerras, IBM Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/kernel.h> +#include <linux/perf_counter.h> +#include <asm/reg.h> + +/* + * Bits in event code for POWER7 + */ +#define PM_PMC_SH 16 /* PMC number (1-based) for direct events */ +#define PM_PMC_MSK 0xf +#define PM_PMC_MSKS (PM_PMC_MSK << PM_PMC_SH) +#define PM_UNIT_SH 12 /* TTMMUX number and setting - unit select */ +#define PM_UNIT_MSK 0xf +#define PM_COMBINE_SH 11 /* Combined event bit */ +#define PM_COMBINE_MSK 1 +#define PM_COMBINE_MSKS 0x800 +#define PM_L2SEL_SH 8 /* L2 event select */ +#define PM_L2SEL_MSK 7 +#define PM_PMCSEL_MSK 0xff + +/* + * Bits in MMCR1 for POWER7 + */ +#define MMCR1_TTM0SEL_SH 60 +#define MMCR1_TTM1SEL_SH 56 +#define MMCR1_TTM2SEL_SH 52 +#define MMCR1_TTM3SEL_SH 48 +#define MMCR1_TTMSEL_MSK 0xf +#define MMCR1_L2SEL_SH 45 +#define MMCR1_L2SEL_MSK 7 +#define MMCR1_PMC1_COMBINE_SH 35 +#define MMCR1_PMC2_COMBINE_SH 34 +#define MMCR1_PMC3_COMBINE_SH 33 +#define MMCR1_PMC4_COMBINE_SH 32 +#define MMCR1_PMC1SEL_SH 24 +#define MMCR1_PMC2SEL_SH 16 +#define MMCR1_PMC3SEL_SH 8 +#define MMCR1_PMC4SEL_SH 0 +#define MMCR1_PMCSEL_SH(n) (MMCR1_PMC1SEL_SH - (n) * 8) +#define MMCR1_PMCSEL_MSK 0xff + +/* + * Bits in MMCRA + */ + +/* + * Layout of constraint bits: + * 6666555555555544444444443333333333222222222211111111110000000000 + * 3210987654321098765432109876543210987654321098765432109876543210 + * [ ><><><><><><> + * NC P6P5P4P3P2P1 + * + * NC - number of counters + * 15: NC error 0x8000 + * 12-14: number of events needing PMC1-4 0x7000 + * + * P6 + * 11: P6 error 0x800 + * 10-11: Count of events needing PMC6 + * + * P1..P5 + * 0-9: Count of events needing PMC1..PMC5 + */ + +static int power7_get_constraint(u64 event, u64 *maskp, u64 *valp) +{ + int pmc, sh; + u64 mask = 0, value = 0; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 6) + return -1; + sh = (pmc - 1) * 2; + mask |= 2 << sh; + value |= 1 << sh; + if (pmc >= 5 && !(event == 0x500fa || event == 0x600f4)) + return -1; + } + if (pmc < 5) { + /* need a counter from PMC1-4 set */ + mask |= 0x8000; + value |= 0x1000; + } + *maskp = mask; + *valp = value; + return 0; +} + +#define MAX_ALT 2 /* at most 2 alternatives for any event */ + +static const unsigned int event_alternatives[][MAX_ALT] = { + { 0x200f2, 0x300f2 }, /* PM_INST_DISP */ + { 0x200f4, 0x600f4 }, /* PM_RUN_CYC */ + { 0x400fa, 0x500fa }, /* PM_RUN_INST_CMPL */ +}; + +/* + * Scan the alternatives table for a match and return the + * index into the alternatives table if found, else -1. + */ +static int find_alternative(u64 event) +{ + int i, j; + + for (i = 0; i < ARRAY_SIZE(event_alternatives); ++i) { + if (event < event_alternatives[i][0]) + break; + for (j = 0; j < MAX_ALT && event_alternatives[i][j]; ++j) + if (event == event_alternatives[i][j]) + return i; + } + return -1; +} + +static s64 find_alternative_decode(u64 event) +{ + int pmc, psel; + + /* this only handles the 4x decode events */ + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + psel = event & PM_PMCSEL_MSK; + if ((pmc == 2 || pmc == 4) && (psel & ~7) == 0x40) + return event - (1 << PM_PMC_SH) + 8; + if ((pmc == 1 || pmc == 3) && (psel & ~7) == 0x48) + return event + (1 << PM_PMC_SH) - 8; + return -1; +} + +static int power7_get_alternatives(u64 event, unsigned int flags, u64 alt[]) +{ + int i, j, nalt = 1; + s64 ae; + + alt[0] = event; + nalt = 1; + i = find_alternative(event); + if (i >= 0) { + for (j = 0; j < MAX_ALT; ++j) { + ae = event_alternatives[i][j]; + if (ae && ae != event) + alt[nalt++] = ae; + } + } else { + ae = find_alternative_decode(event); + if (ae > 0) + alt[nalt++] = ae; + } + + if (flags & PPMU_ONLY_COUNT_RUN) { + /* + * We're only counting in RUN state, + * so PM_CYC is equivalent to PM_RUN_CYC + * and PM_INST_CMPL === PM_RUN_INST_CMPL. + * This doesn't include alternatives that don't provide + * any extra flexibility in assigning PMCs. + */ + j = nalt; + for (i = 0; i < nalt; ++i) { + switch (alt[i]) { + case 0x1e: /* PM_CYC */ + alt[j++] = 0x600f4; /* PM_RUN_CYC */ + break; + case 0x600f4: /* PM_RUN_CYC */ + alt[j++] = 0x1e; + break; + case 0x2: /* PM_PPC_CMPL */ + alt[j++] = 0x500fa; /* PM_RUN_INST_CMPL */ + break; + case 0x500fa: /* PM_RUN_INST_CMPL */ + alt[j++] = 0x2; /* PM_PPC_CMPL */ + break; + } + } + nalt = j; + } + + return nalt; +} + +/* + * Returns 1 if event counts things relating to marked instructions + * and thus needs the MMCRA_SAMPLE_ENABLE bit set, or 0 if not. + */ +static int power7_marked_instr_event(u64 event) +{ + int pmc, psel; + int unit; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + psel = event & PM_PMCSEL_MSK & ~1; /* trim off edge/level bit */ + if (pmc >= 5) + return 0; + + switch (psel >> 4) { + case 2: + return pmc == 2 || pmc == 4; + case 3: + if (psel == 0x3c) + return pmc == 1; + if (psel == 0x3e) + return pmc != 2; + return 1; + case 4: + case 5: + return unit == 0xd; + case 6: + if (psel == 0x64) + return pmc >= 3; + case 8: + return unit == 0xd; + } + return 0; +} + +static int power7_compute_mmcr(u64 event[], int n_ev, + unsigned int hwc[], u64 mmcr[]) +{ + u64 mmcr1 = 0; + u64 mmcra = 0; + unsigned int pmc, unit, combine, l2sel, psel; + unsigned int pmc_inuse = 0; + int i; + + /* First pass to count resource use */ + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 6) + return -1; + if (pmc_inuse & (1 << (pmc - 1))) + return -1; + pmc_inuse |= 1 << (pmc - 1); + } + } + + /* Second pass: assign PMCs, set all MMCR1 fields */ + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + combine = (event[i] >> PM_COMBINE_SH) & PM_COMBINE_MSK; + l2sel = (event[i] >> PM_L2SEL_SH) & PM_L2SEL_MSK; + psel = event[i] & PM_PMCSEL_MSK; + if (!pmc) { + /* Bus event or any-PMC direct event */ + for (pmc = 0; pmc < 4; ++pmc) { + if (!(pmc_inuse & (1 << pmc))) + break; + } + if (pmc >= 4) + return -1; + pmc_inuse |= 1 << pmc; + } else { + /* Direct or decoded event */ + --pmc; + } + if (pmc <= 3) { + mmcr1 |= (u64) unit << (MMCR1_TTM0SEL_SH - 4 * pmc); + mmcr1 |= (u64) combine << (MMCR1_PMC1_COMBINE_SH - pmc); + mmcr1 |= psel << MMCR1_PMCSEL_SH(pmc); + if (unit == 6) /* L2 events */ + mmcr1 |= (u64) l2sel << MMCR1_L2SEL_SH; + } + if (power7_marked_instr_event(event[i])) + mmcra |= MMCRA_SAMPLE_ENABLE; + hwc[i] = pmc; + } + + /* Return MMCRx values */ + mmcr[0] = 0; + if (pmc_inuse & 1) + mmcr[0] = MMCR0_PMC1CE; + if (pmc_inuse & 0x3e) + mmcr[0] |= MMCR0_PMCjCE; + mmcr[1] = mmcr1; + mmcr[2] = mmcra; + return 0; +} + +static void power7_disable_pmc(unsigned int pmc, u64 mmcr[]) +{ + if (pmc <= 3) + mmcr[1] &= ~(0xffULL << MMCR1_PMCSEL_SH(pmc)); +} + +static int power7_generic_events[] = { + [PERF_COUNT_CPU_CYCLES] = 0x1e, + [PERF_COUNT_INSTRUCTIONS] = 2, + [PERF_COUNT_CACHE_REFERENCES] = 0xc880, /* LD_REF_L1_LSU */ + [PERF_COUNT_CACHE_MISSES] = 0x400f0, /* LD_MISS_L1 */ + [PERF_COUNT_BRANCH_INSTRUCTIONS] = 0x10068, /* BRU_FIN */ + [PERF_COUNT_BRANCH_MISSES] = 0x400f6, /* BR_MPRED */ +}; + +#define C(x) PERF_COUNT_HW_CACHE_##x + +/* + * Table of generalized cache-related events. + * 0 means not supported, -1 means nonsensical, other values + * are event codes. + */ +static int power7_cache_events[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = { + [C(L1D)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x400f0, 0xc880 }, + [C(OP_WRITE)] = { 0, 0x300f0 }, + [C(OP_PREFETCH)] = { 0xd8b8, 0 }, + }, + [C(L1I)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x200fc }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { 0x408a, 0 }, + }, + [C(LL)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x6080, 0x6084 }, + [C(OP_WRITE)] = { 0x6082, 0x6086 }, + [C(OP_PREFETCH)] = { 0, 0 }, + }, + [C(DTLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x300fc }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(ITLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x400fc }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(BPU)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x10068, 0x400f6 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, +}; + +struct power_pmu power7_pmu = { + .n_counter = 6, + .max_alternatives = MAX_ALT + 1, + .add_fields = 0x1555ull, + .test_adder = 0x3000ull, + .compute_mmcr = power7_compute_mmcr, + .get_constraint = power7_get_constraint, + .get_alternatives = power7_get_alternatives, + .disable_pmc = power7_disable_pmc, + .n_generic = ARRAY_SIZE(power7_generic_events), + .generic_events = power7_generic_events, + .cache_events = &power7_cache_events, +}; diff --git a/arch/powerpc/kernel/ppc970-pmu.c b/arch/powerpc/kernel/ppc970-pmu.c new file mode 100644 index 00000000000..ba0a357a89f --- /dev/null +++ b/arch/powerpc/kernel/ppc970-pmu.c @@ -0,0 +1,482 @@ +/* + * Performance counter support for PPC970-family processors. + * + * Copyright 2008-2009 Paul Mackerras, IBM Corporation. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/string.h> +#include <linux/perf_counter.h> +#include <asm/reg.h> + +/* + * Bits in event code for PPC970 + */ +#define PM_PMC_SH 12 /* PMC number (1-based) for direct events */ +#define PM_PMC_MSK 0xf +#define PM_UNIT_SH 8 /* TTMMUX number and setting - unit select */ +#define PM_UNIT_MSK 0xf +#define PM_SPCSEL_SH 6 +#define PM_SPCSEL_MSK 3 +#define PM_BYTE_SH 4 /* Byte number of event bus to use */ +#define PM_BYTE_MSK 3 +#define PM_PMCSEL_MSK 0xf + +/* Values in PM_UNIT field */ +#define PM_NONE 0 +#define PM_FPU 1 +#define PM_VPU 2 +#define PM_ISU 3 +#define PM_IFU 4 +#define PM_IDU 5 +#define PM_STS 6 +#define PM_LSU0 7 +#define PM_LSU1U 8 +#define PM_LSU1L 9 +#define PM_LASTUNIT 9 + +/* + * Bits in MMCR0 for PPC970 + */ +#define MMCR0_PMC1SEL_SH 8 +#define MMCR0_PMC2SEL_SH 1 +#define MMCR_PMCSEL_MSK 0x1f + +/* + * Bits in MMCR1 for PPC970 + */ +#define MMCR1_TTM0SEL_SH 62 +#define MMCR1_TTM1SEL_SH 59 +#define MMCR1_TTM3SEL_SH 53 +#define MMCR1_TTMSEL_MSK 3 +#define MMCR1_TD_CP_DBG0SEL_SH 50 +#define MMCR1_TD_CP_DBG1SEL_SH 48 +#define MMCR1_TD_CP_DBG2SEL_SH 46 +#define MMCR1_TD_CP_DBG3SEL_SH 44 +#define MMCR1_PMC1_ADDER_SEL_SH 39 +#define MMCR1_PMC2_ADDER_SEL_SH 38 +#define MMCR1_PMC6_ADDER_SEL_SH 37 +#define MMCR1_PMC5_ADDER_SEL_SH 36 +#define MMCR1_PMC8_ADDER_SEL_SH 35 +#define MMCR1_PMC7_ADDER_SEL_SH 34 +#define MMCR1_PMC3_ADDER_SEL_SH 33 +#define MMCR1_PMC4_ADDER_SEL_SH 32 +#define MMCR1_PMC3SEL_SH 27 +#define MMCR1_PMC4SEL_SH 22 +#define MMCR1_PMC5SEL_SH 17 +#define MMCR1_PMC6SEL_SH 12 +#define MMCR1_PMC7SEL_SH 7 +#define MMCR1_PMC8SEL_SH 2 + +static short mmcr1_adder_bits[8] = { + MMCR1_PMC1_ADDER_SEL_SH, + MMCR1_PMC2_ADDER_SEL_SH, + MMCR1_PMC3_ADDER_SEL_SH, + MMCR1_PMC4_ADDER_SEL_SH, + MMCR1_PMC5_ADDER_SEL_SH, + MMCR1_PMC6_ADDER_SEL_SH, + MMCR1_PMC7_ADDER_SEL_SH, + MMCR1_PMC8_ADDER_SEL_SH +}; + +/* + * Bits in MMCRA + */ + +/* + * Layout of constraint bits: + * 6666555555555544444444443333333333222222222211111111110000000000 + * 3210987654321098765432109876543210987654321098765432109876543210 + * <><><>[ >[ >[ >< >< >< >< ><><><><><><><><> + * SPT0T1 UC PS1 PS2 B0 B1 B2 B3 P1P2P3P4P5P6P7P8 + * + * SP - SPCSEL constraint + * 48-49: SPCSEL value 0x3_0000_0000_0000 + * + * T0 - TTM0 constraint + * 46-47: TTM0SEL value (0=FPU, 2=IFU, 3=VPU) 0xC000_0000_0000 + * + * T1 - TTM1 constraint + * 44-45: TTM1SEL value (0=IDU, 3=STS) 0x3000_0000_0000 + * + * UC - unit constraint: can't have all three of FPU|IFU|VPU, ISU, IDU|STS + * 43: UC3 error 0x0800_0000_0000 + * 42: FPU|IFU|VPU events needed 0x0400_0000_0000 + * 41: ISU events needed 0x0200_0000_0000 + * 40: IDU|STS events needed 0x0100_0000_0000 + * + * PS1 + * 39: PS1 error 0x0080_0000_0000 + * 36-38: count of events needing PMC1/2/5/6 0x0070_0000_0000 + * + * PS2 + * 35: PS2 error 0x0008_0000_0000 + * 32-34: count of events needing PMC3/4/7/8 0x0007_0000_0000 + * + * B0 + * 28-31: Byte 0 event source 0xf000_0000 + * Encoding as for the event code + * + * B1, B2, B3 + * 24-27, 20-23, 16-19: Byte 1, 2, 3 event sources + * + * P1 + * 15: P1 error 0x8000 + * 14-15: Count of events needing PMC1 + * + * P2..P8 + * 0-13: Count of events needing PMC2..PMC8 + */ + +static unsigned char direct_marked_event[8] = { + (1<<2) | (1<<3), /* PMC1: PM_MRK_GRP_DISP, PM_MRK_ST_CMPL */ + (1<<3) | (1<<5), /* PMC2: PM_THRESH_TIMEO, PM_MRK_BRU_FIN */ + (1<<3) | (1<<5), /* PMC3: PM_MRK_ST_CMPL_INT, PM_MRK_VMX_FIN */ + (1<<4) | (1<<5), /* PMC4: PM_MRK_GRP_CMPL, PM_MRK_CRU_FIN */ + (1<<4) | (1<<5), /* PMC5: PM_GRP_MRK, PM_MRK_GRP_TIMEO */ + (1<<3) | (1<<4) | (1<<5), + /* PMC6: PM_MRK_ST_STS, PM_MRK_FXU_FIN, PM_MRK_GRP_ISSUED */ + (1<<4) | (1<<5), /* PMC7: PM_MRK_FPU_FIN, PM_MRK_INST_FIN */ + (1<<4) /* PMC8: PM_MRK_LSU_FIN */ +}; + +/* + * Returns 1 if event counts things relating to marked instructions + * and thus needs the MMCRA_SAMPLE_ENABLE bit set, or 0 if not. + */ +static int p970_marked_instr_event(u64 event) +{ + int pmc, psel, unit, byte, bit; + unsigned int mask; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + psel = event & PM_PMCSEL_MSK; + if (pmc) { + if (direct_marked_event[pmc - 1] & (1 << psel)) + return 1; + if (psel == 0) /* add events */ + bit = (pmc <= 4)? pmc - 1: 8 - pmc; + else if (psel == 7 || psel == 13) /* decode events */ + bit = 4; + else + return 0; + } else + bit = psel; + + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + mask = 0; + switch (unit) { + case PM_VPU: + mask = 0x4c; /* byte 0 bits 2,3,6 */ + case PM_LSU0: + /* byte 2 bits 0,2,3,4,6; all of byte 1 */ + mask = 0x085dff00; + case PM_LSU1L: + mask = 0x50 << 24; /* byte 3 bits 4,6 */ + break; + } + return (mask >> (byte * 8 + bit)) & 1; +} + +/* Masks and values for using events from the various units */ +static u64 unit_cons[PM_LASTUNIT+1][2] = { + [PM_FPU] = { 0xc80000000000ull, 0x040000000000ull }, + [PM_VPU] = { 0xc80000000000ull, 0xc40000000000ull }, + [PM_ISU] = { 0x080000000000ull, 0x020000000000ull }, + [PM_IFU] = { 0xc80000000000ull, 0x840000000000ull }, + [PM_IDU] = { 0x380000000000ull, 0x010000000000ull }, + [PM_STS] = { 0x380000000000ull, 0x310000000000ull }, +}; + +static int p970_get_constraint(u64 event, u64 *maskp, u64 *valp) +{ + int pmc, byte, unit, sh, spcsel; + u64 mask = 0, value = 0; + int grp = -1; + + pmc = (event >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc > 8) + return -1; + sh = (pmc - 1) * 2; + mask |= 2 << sh; + value |= 1 << sh; + grp = ((pmc - 1) >> 1) & 1; + } + unit = (event >> PM_UNIT_SH) & PM_UNIT_MSK; + if (unit) { + if (unit > PM_LASTUNIT) + return -1; + mask |= unit_cons[unit][0]; + value |= unit_cons[unit][1]; + byte = (event >> PM_BYTE_SH) & PM_BYTE_MSK; + /* + * Bus events on bytes 0 and 2 can be counted + * on PMC1/2/5/6; bytes 1 and 3 on PMC3/4/7/8. + */ + if (!pmc) + grp = byte & 1; + /* Set byte lane select field */ + mask |= 0xfULL << (28 - 4 * byte); + value |= (u64)unit << (28 - 4 * byte); + } + if (grp == 0) { + /* increment PMC1/2/5/6 field */ + mask |= 0x8000000000ull; + value |= 0x1000000000ull; + } else if (grp == 1) { + /* increment PMC3/4/7/8 field */ + mask |= 0x800000000ull; + value |= 0x100000000ull; + } + spcsel = (event >> PM_SPCSEL_SH) & PM_SPCSEL_MSK; + if (spcsel) { + mask |= 3ull << 48; + value |= (u64)spcsel << 48; + } + *maskp = mask; + *valp = value; + return 0; +} + +static int p970_get_alternatives(u64 event, unsigned int flags, u64 alt[]) +{ + alt[0] = event; + + /* 2 alternatives for LSU empty */ + if (event == 0x2002 || event == 0x3002) { + alt[1] = event ^ 0x1000; + return 2; + } + + return 1; +} + +static int p970_compute_mmcr(u64 event[], int n_ev, + unsigned int hwc[], u64 mmcr[]) +{ + u64 mmcr0 = 0, mmcr1 = 0, mmcra = 0; + unsigned int pmc, unit, byte, psel; + unsigned int ttm, grp; + unsigned int pmc_inuse = 0; + unsigned int pmc_grp_use[2]; + unsigned char busbyte[4]; + unsigned char unituse[16]; + unsigned char unitmap[] = { 0, 0<<3, 3<<3, 1<<3, 2<<3, 0|4, 3|4 }; + unsigned char ttmuse[2]; + unsigned char pmcsel[8]; + int i; + int spcsel; + + if (n_ev > 8) + return -1; + + /* First pass to count resource use */ + pmc_grp_use[0] = pmc_grp_use[1] = 0; + memset(busbyte, 0, sizeof(busbyte)); + memset(unituse, 0, sizeof(unituse)); + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + if (pmc) { + if (pmc_inuse & (1 << (pmc - 1))) + return -1; + pmc_inuse |= 1 << (pmc - 1); + /* count 1/2/5/6 vs 3/4/7/8 use */ + ++pmc_grp_use[((pmc - 1) >> 1) & 1]; + } + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event[i] >> PM_BYTE_SH) & PM_BYTE_MSK; + if (unit) { + if (unit > PM_LASTUNIT) + return -1; + if (!pmc) + ++pmc_grp_use[byte & 1]; + if (busbyte[byte] && busbyte[byte] != unit) + return -1; + busbyte[byte] = unit; + unituse[unit] = 1; + } + } + if (pmc_grp_use[0] > 4 || pmc_grp_use[1] > 4) + return -1; + + /* + * Assign resources and set multiplexer selects. + * + * PM_ISU can go either on TTM0 or TTM1, but that's the only + * choice we have to deal with. + */ + if (unituse[PM_ISU] & + (unituse[PM_FPU] | unituse[PM_IFU] | unituse[PM_VPU])) + unitmap[PM_ISU] = 2 | 4; /* move ISU to TTM1 */ + /* Set TTM[01]SEL fields. */ + ttmuse[0] = ttmuse[1] = 0; + for (i = PM_FPU; i <= PM_STS; ++i) { + if (!unituse[i]) + continue; + ttm = unitmap[i]; + ++ttmuse[(ttm >> 2) & 1]; + mmcr1 |= (u64)(ttm & ~4) << MMCR1_TTM1SEL_SH; + } + /* Check only one unit per TTMx */ + if (ttmuse[0] > 1 || ttmuse[1] > 1) + return -1; + + /* Set byte lane select fields and TTM3SEL. */ + for (byte = 0; byte < 4; ++byte) { + unit = busbyte[byte]; + if (!unit) + continue; + if (unit <= PM_STS) + ttm = (unitmap[unit] >> 2) & 1; + else if (unit == PM_LSU0) + ttm = 2; + else { + ttm = 3; + if (unit == PM_LSU1L && byte >= 2) + mmcr1 |= 1ull << (MMCR1_TTM3SEL_SH + 3 - byte); + } + mmcr1 |= (u64)ttm << (MMCR1_TD_CP_DBG0SEL_SH - 2 * byte); + } + + /* Second pass: assign PMCs, set PMCxSEL and PMCx_ADDER_SEL fields */ + memset(pmcsel, 0x8, sizeof(pmcsel)); /* 8 means don't count */ + for (i = 0; i < n_ev; ++i) { + pmc = (event[i] >> PM_PMC_SH) & PM_PMC_MSK; + unit = (event[i] >> PM_UNIT_SH) & PM_UNIT_MSK; + byte = (event[i] >> PM_BYTE_SH) & PM_BYTE_MSK; + psel = event[i] & PM_PMCSEL_MSK; + if (!pmc) { + /* Bus event or any-PMC direct event */ + if (unit) + psel |= 0x10 | ((byte & 2) << 2); + else + psel |= 8; + for (pmc = 0; pmc < 8; ++pmc) { + if (pmc_inuse & (1 << pmc)) + continue; + grp = (pmc >> 1) & 1; + if (unit) { + if (grp == (byte & 1)) + break; + } else if (pmc_grp_use[grp] < 4) { + ++pmc_grp_use[grp]; + break; + } + } + pmc_inuse |= 1 << pmc; + } else { + /* Direct event */ + --pmc; + if (psel == 0 && (byte & 2)) + /* add events on higher-numbered bus */ + mmcr1 |= 1ull << mmcr1_adder_bits[pmc]; + } + pmcsel[pmc] = psel; + hwc[i] = pmc; + spcsel = (event[i] >> PM_SPCSEL_SH) & PM_SPCSEL_MSK; + mmcr1 |= spcsel; + if (p970_marked_instr_event(event[i])) + mmcra |= MMCRA_SAMPLE_ENABLE; + } + for (pmc = 0; pmc < 2; ++pmc) + mmcr0 |= pmcsel[pmc] << (MMCR0_PMC1SEL_SH - 7 * pmc); + for (; pmc < 8; ++pmc) + mmcr1 |= (u64)pmcsel[pmc] << (MMCR1_PMC3SEL_SH - 5 * (pmc - 2)); + if (pmc_inuse & 1) + mmcr0 |= MMCR0_PMC1CE; + if (pmc_inuse & 0xfe) + mmcr0 |= MMCR0_PMCjCE; + + mmcra |= 0x2000; /* mark only one IOP per PPC instruction */ + + /* Return MMCRx values */ + mmcr[0] = mmcr0; + mmcr[1] = mmcr1; + mmcr[2] = mmcra; + return 0; +} + +static void p970_disable_pmc(unsigned int pmc, u64 mmcr[]) +{ + int shift, i; + + if (pmc <= 1) { + shift = MMCR0_PMC1SEL_SH - 7 * pmc; + i = 0; + } else { + shift = MMCR1_PMC3SEL_SH - 5 * (pmc - 2); + i = 1; + } + /* + * Setting the PMCxSEL field to 0x08 disables PMC x. + */ + mmcr[i] = (mmcr[i] & ~(0x1fUL << shift)) | (0x08UL << shift); +} + +static int ppc970_generic_events[] = { + [PERF_COUNT_HW_CPU_CYCLES] = 7, + [PERF_COUNT_HW_INSTRUCTIONS] = 1, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x8810, /* PM_LD_REF_L1 */ + [PERF_COUNT_HW_CACHE_MISSES] = 0x3810, /* PM_LD_MISS_L1 */ + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x431, /* PM_BR_ISSUED */ + [PERF_COUNT_HW_BRANCH_MISSES] = 0x327, /* PM_GRP_BR_MPRED */ +}; + +#define C(x) PERF_COUNT_HW_CACHE_##x + +/* + * Table of generalized cache-related events. + * 0 means not supported, -1 means nonsensical, other values + * are event codes. + */ +static int ppc970_cache_events[C(MAX)][C(OP_MAX)][C(RESULT_MAX)] = { + [C(L1D)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x8810, 0x3810 }, + [C(OP_WRITE)] = { 0x7810, 0x813 }, + [C(OP_PREFETCH)] = { 0x731, 0 }, + }, + [C(L1I)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { 0, 0 }, + }, + [C(LL)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0 }, + [C(OP_WRITE)] = { 0, 0 }, + [C(OP_PREFETCH)] = { 0x733, 0 }, + }, + [C(DTLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x704 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(ITLB)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0, 0x700 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, + [C(BPU)] = { /* RESULT_ACCESS RESULT_MISS */ + [C(OP_READ)] = { 0x431, 0x327 }, + [C(OP_WRITE)] = { -1, -1 }, + [C(OP_PREFETCH)] = { -1, -1 }, + }, +}; + +struct power_pmu ppc970_pmu = { + .n_counter = 8, + .max_alternatives = 2, + .add_fields = 0x001100005555ull, + .test_adder = 0x013300000000ull, + .compute_mmcr = p970_compute_mmcr, + .get_constraint = p970_get_constraint, + .get_alternatives = p970_get_alternatives, + .disable_pmc = p970_disable_pmc, + .n_generic = ARRAY_SIZE(ppc970_generic_events), + .generic_events = ppc970_generic_events, + .cache_events = &ppc970_cache_events, +}; diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c index 76993941cac..5beffc8f481 100644 --- a/arch/powerpc/mm/fault.c +++ b/arch/powerpc/mm/fault.c @@ -29,6 +29,7 @@ #include <linux/module.h> #include <linux/kprobes.h> #include <linux/kdebug.h> +#include <linux/perf_counter.h> #include <asm/firmware.h> #include <asm/page.h> @@ -170,6 +171,8 @@ int __kprobes do_page_fault(struct pt_regs *regs, unsigned long address, die("Weird page fault", regs, SIGSEGV); } + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address); + /* When running in the kernel we expect faults to occur only to * addresses in user space. All other faults represent errors in the * kernel and should generate an OOPS. Unfortunately, in the case of an @@ -309,6 +312,8 @@ good_area: } if (ret & VM_FAULT_MAJOR) { current->maj_flt++; + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, 0, + regs, address); #ifdef CONFIG_PPC_SMLPAR if (firmware_has_feature(FW_FEATURE_CMO)) { preempt_disable(); @@ -316,8 +321,11 @@ good_area: preempt_enable(); } #endif - } else + } else { current->min_flt++; + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, 0, + regs, address); + } up_read(&mm->mmap_sem); return 0; diff --git a/arch/powerpc/platforms/Kconfig.cputype b/arch/powerpc/platforms/Kconfig.cputype index 9da795e4933..732ee93a8e9 100644 --- a/arch/powerpc/platforms/Kconfig.cputype +++ b/arch/powerpc/platforms/Kconfig.cputype @@ -1,6 +1,7 @@ config PPC64 bool "64-bit kernel" default n + select HAVE_PERF_COUNTERS help This option selects whether a 32-bit or a 64-bit kernel will be built. diff --git a/arch/powerpc/platforms/pseries/xics.c b/arch/powerpc/platforms/pseries/xics.c index 80b513449f4..be3581a8c29 100644 --- a/arch/powerpc/platforms/pseries/xics.c +++ b/arch/powerpc/platforms/pseries/xics.c @@ -333,7 +333,7 @@ static void xics_eoi_lpar(unsigned int virq) lpar_xirr_info_set((0xff << 24) | irq); } -static void xics_set_affinity(unsigned int virq, const struct cpumask *cpumask) +static int xics_set_affinity(unsigned int virq, const struct cpumask *cpumask) { unsigned int irq; int status; @@ -342,14 +342,14 @@ static void xics_set_affinity(unsigned int virq, const struct cpumask *cpumask) irq = (unsigned int)irq_map[virq].hwirq; if (irq == XICS_IPI || irq == XICS_IRQ_SPURIOUS) - return; + return -1; status = rtas_call(ibm_get_xive, 1, 3, xics_status, irq); if (status) { printk(KERN_ERR "%s: ibm,get-xive irq=%u returns %d\n", __func__, irq, status); - return; + return -1; } /* @@ -363,7 +363,7 @@ static void xics_set_affinity(unsigned int virq, const struct cpumask *cpumask) printk(KERN_WARNING "%s: No online cpus in the mask %s for irq %d\n", __func__, cpulist, virq); - return; + return -1; } status = rtas_call(ibm_set_xive, 3, 1, NULL, @@ -372,8 +372,10 @@ static void xics_set_affinity(unsigned int virq, const struct cpumask *cpumask) if (status) { printk(KERN_ERR "%s: ibm,set-xive irq=%u returns %d\n", __func__, irq, status); - return; + return -1; } + + return 0; } static struct irq_chip xics_pic_direct = { diff --git a/arch/powerpc/sysdev/mpic.c b/arch/powerpc/sysdev/mpic.c index 0efc12d1a3d..352d8c3ef52 100644 --- a/arch/powerpc/sysdev/mpic.c +++ b/arch/powerpc/sysdev/mpic.c @@ -807,7 +807,7 @@ static void mpic_end_ipi(unsigned int irq) #endif /* CONFIG_SMP */ -void mpic_set_affinity(unsigned int irq, const struct cpumask *cpumask) +int mpic_set_affinity(unsigned int irq, const struct cpumask *cpumask) { struct mpic *mpic = mpic_from_irq(irq); unsigned int src = mpic_irq_to_hw(irq); @@ -824,6 +824,8 @@ void mpic_set_affinity(unsigned int irq, const struct cpumask *cpumask) mpic_irq_write(src, MPIC_INFO(IRQ_DESTINATION), mpic_physmask(cpus_addr(tmp)[0])); } + + return 0; } static unsigned int mpic_type_to_vecpri(struct mpic *mpic, unsigned int type) diff --git a/arch/powerpc/sysdev/mpic.h b/arch/powerpc/sysdev/mpic.h index 3cef2af10f4..eff433c322a 100644 --- a/arch/powerpc/sysdev/mpic.h +++ b/arch/powerpc/sysdev/mpic.h @@ -36,6 +36,6 @@ static inline int mpic_pasemi_msi_init(struct mpic *mpic) extern int mpic_set_irq_type(unsigned int virq, unsigned int flow_type); extern void mpic_set_vector(unsigned int virq, unsigned int vector); -extern void mpic_set_affinity(unsigned int irq, const struct cpumask *cpumask); +extern int mpic_set_affinity(unsigned int irq, const struct cpumask *cpumask); #endif /* _POWERPC_SYSDEV_MPIC_H */ diff --git a/arch/sparc/include/asm/thread_info_64.h b/arch/sparc/include/asm/thread_info_64.h index 639ac805448..65865726b28 100644 --- a/arch/sparc/include/asm/thread_info_64.h +++ b/arch/sparc/include/asm/thread_info_64.h @@ -102,8 +102,8 @@ struct thread_info { #define TI_KERN_CNTD1 0x00000488 #define TI_PCR 0x00000490 #define TI_RESTART_BLOCK 0x00000498 -#define TI_KUNA_REGS 0x000004c0 -#define TI_KUNA_INSN 0x000004c8 +#define TI_KUNA_REGS 0x000004c8 +#define TI_KUNA_INSN 0x000004d0 #define TI_FPREGS 0x00000500 /* We embed this in the uppermost byte of thread_info->flags */ diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c index 5deabe921a4..e5e78f9cfc9 100644 --- a/arch/sparc/kernel/irq_64.c +++ b/arch/sparc/kernel/irq_64.c @@ -318,10 +318,12 @@ static void sun4u_irq_enable(unsigned int virt_irq) } } -static void sun4u_set_affinity(unsigned int virt_irq, +static int sun4u_set_affinity(unsigned int virt_irq, const struct cpumask *mask) { sun4u_irq_enable(virt_irq); + + return 0; } /* Don't do anything. The desc->status check for IRQ_DISABLED in @@ -377,7 +379,7 @@ static void sun4v_irq_enable(unsigned int virt_irq) ino, err); } -static void sun4v_set_affinity(unsigned int virt_irq, +static int sun4v_set_affinity(unsigned int virt_irq, const struct cpumask *mask) { unsigned int ino = virt_irq_table[virt_irq].dev_ino; @@ -388,6 +390,8 @@ static void sun4v_set_affinity(unsigned int virt_irq, if (err != HV_EOK) printk(KERN_ERR "sun4v_intr_settarget(%x,%lu): " "err(%d)\n", ino, cpuid, err); + + return 0; } static void sun4v_irq_disable(unsigned int virt_irq) @@ -445,7 +449,7 @@ static void sun4v_virq_enable(unsigned int virt_irq) dev_handle, dev_ino, err); } -static void sun4v_virt_set_affinity(unsigned int virt_irq, +static int sun4v_virt_set_affinity(unsigned int virt_irq, const struct cpumask *mask) { unsigned long cpuid, dev_handle, dev_ino; @@ -461,6 +465,8 @@ static void sun4v_virt_set_affinity(unsigned int virt_irq, printk(KERN_ERR "sun4v_vintr_set_target(%lx,%lx,%lu): " "err(%d)\n", dev_handle, dev_ino, cpuid, err); + + return 0; } static void sun4v_virq_disable(unsigned int virt_irq) diff --git a/arch/x86/Kbuild b/arch/x86/Kbuild new file mode 100644 index 00000000000..ad8ec356fb3 --- /dev/null +++ b/arch/x86/Kbuild @@ -0,0 +1,16 @@ + +obj-$(CONFIG_KVM) += kvm/ + +# Xen paravirtualization support +obj-$(CONFIG_XEN) += xen/ + +# lguest paravirtualization support +obj-$(CONFIG_LGUEST_GUEST) += lguest/ + +obj-y += kernel/ +obj-y += mm/ + +obj-y += crypto/ +obj-y += vdso/ +obj-$(CONFIG_IA32_EMULATION) += ia32/ + diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index a6efe0a2e9a..68f5578fe38 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -47,6 +47,11 @@ config X86 select HAVE_KERNEL_BZIP2 select HAVE_KERNEL_LZMA +config OUTPUT_FORMAT + string + default "elf32-i386" if X86_32 + default "elf64-x86-64" if X86_64 + config ARCH_DEFCONFIG string default "arch/x86/configs/i386_defconfig" if X86_32 @@ -274,15 +279,9 @@ config SPARSE_IRQ If you don't know what to do here, say N. -config NUMA_MIGRATE_IRQ_DESC - bool "Move irq desc when changing irq smp_affinity" +config NUMA_IRQ_DESC + def_bool y depends on SPARSE_IRQ && NUMA - depends on BROKEN - default n - ---help--- - This enables moving irq_desc to cpu/node that irq will use handled. - - If you don't know what to do here, say N. config X86_MPPARSE bool "Enable MPS table" if ACPI @@ -355,7 +354,7 @@ config X86_UV depends on X86_64 depends on X86_EXTENDED_PLATFORM depends on NUMA - select X86_X2APIC + depends on X86_X2APIC ---help--- This option is needed in order to support SGI Ultraviolet systems. If you don't have one of these, you should say N here. @@ -740,6 +739,7 @@ config X86_UP_IOAPIC config X86_LOCAL_APIC def_bool y depends on X86_64 || SMP || X86_32_NON_STANDARD || X86_UP_APIC + select HAVE_PERF_COUNTERS if (!M386 && !M486) config X86_IO_APIC def_bool y @@ -1466,9 +1466,7 @@ config KEXEC_JUMP config PHYSICAL_START hex "Physical address where the kernel is loaded" if (EMBEDDED || CRASH_DUMP) - default "0x1000000" if X86_NUMAQ - default "0x200000" if X86_64 - default "0x100000" + default "0x1000000" ---help--- This gives the physical address where the kernel is loaded. @@ -1487,15 +1485,15 @@ config PHYSICAL_START to be specifically compiled to run from a specific memory area (normally a reserved region) and this option comes handy. - So if you are using bzImage for capturing the crash dump, leave - the value here unchanged to 0x100000 and set CONFIG_RELOCATABLE=y. - Otherwise if you plan to use vmlinux for capturing the crash dump - change this value to start of the reserved region (Typically 16MB - 0x1000000). In other words, it can be set based on the "X" value as - specified in the "crashkernel=YM@XM" command line boot parameter - passed to the panic-ed kernel. Typically this parameter is set as - crashkernel=64M@16M. Please take a look at - Documentation/kdump/kdump.txt for more details about crash dumps. + So if you are using bzImage for capturing the crash dump, + leave the value here unchanged to 0x1000000 and set + CONFIG_RELOCATABLE=y. Otherwise if you plan to use vmlinux + for capturing the crash dump change this value to start of + the reserved region. In other words, it can be set based on + the "X" value as specified in the "crashkernel=YM@XM" + command line boot parameter passed to the panic-ed + kernel. Please take a look at Documentation/kdump/kdump.txt + for more details about crash dumps. Usage of bzImage for capturing the crash dump is recommended as one does not have to build two kernels. Same kernel can be used @@ -1508,8 +1506,8 @@ config PHYSICAL_START Don't change this unless you know what you are doing. config RELOCATABLE - bool "Build a relocatable kernel (EXPERIMENTAL)" - depends on EXPERIMENTAL + bool "Build a relocatable kernel" + default y ---help--- This builds a kernel image that retains relocation information so it can be loaded someplace besides the default 1MB. @@ -1524,12 +1522,16 @@ config RELOCATABLE it has been loaded at and the compile time physical address (CONFIG_PHYSICAL_START) is ignored. +# Relocation on x86-32 needs some additional build support +config X86_NEED_RELOCS + def_bool y + depends on X86_32 && RELOCATABLE + config PHYSICAL_ALIGN hex prompt "Alignment value to which kernel should be aligned" if X86_32 - default "0x100000" if X86_32 - default "0x200000" if X86_64 - range 0x2000 0x400000 + default "0x1000000" + range 0x2000 0x1000000 ---help--- This value puts the alignment restrictions on physical address where kernel is loaded and run from. Kernel is compiled for an diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug index d8359e73317..d105f29bb6b 100644 --- a/arch/x86/Kconfig.debug +++ b/arch/x86/Kconfig.debug @@ -159,14 +159,30 @@ config IOMMU_DEBUG options. See Documentation/x86_64/boot-options.txt for more details. +config IOMMU_STRESS + bool "Enable IOMMU stress-test mode" + ---help--- + This option disables various optimizations in IOMMU related + code to do real stress testing of the IOMMU code. This option + will cause a performance drop and should only be enabled for + testing. + config IOMMU_LEAK bool "IOMMU leak tracing" - depends on DEBUG_KERNEL - depends on IOMMU_DEBUG + depends on IOMMU_DEBUG && DMA_API_DEBUG ---help--- Add a simple leak tracer to the IOMMU code. This is useful when you are debugging a buggy device driver that leaks IOMMU mappings. +config X86_DS_SELFTEST + bool "DS selftest" + default y + depends on DEBUG_KERNEL + depends on X86_DS + ---help--- + Perform Debug Store selftests at boot time. + If in doubt, say "N". + config HAVE_MMIOTRACE_SUPPORT def_bool y diff --git a/arch/x86/Makefile b/arch/x86/Makefile index 8c86b72afdc..edbd0ca6206 100644 --- a/arch/x86/Makefile +++ b/arch/x86/Makefile @@ -7,8 +7,6 @@ else KBUILD_DEFCONFIG := $(ARCH)_defconfig endif -core-$(CONFIG_KVM) += arch/x86/kvm/ - # BITS is used as extension for files which are available in a 32 bit # and a 64 bit version to simplify shared Makefiles. # e.g.: obj-y += foo_$(BITS).o @@ -118,21 +116,8 @@ head-y += arch/x86/kernel/init_task.o libs-y += arch/x86/lib/ -# Sub architecture files that needs linking first -core-y += $(fcore-y) - -# Xen paravirtualization support -core-$(CONFIG_XEN) += arch/x86/xen/ - -# lguest paravirtualization support -core-$(CONFIG_LGUEST_GUEST) += arch/x86/lguest/ - -core-y += arch/x86/kernel/ -core-y += arch/x86/mm/ - -core-y += arch/x86/crypto/ -core-y += arch/x86/vdso/ -core-$(CONFIG_IA32_EMULATION) += arch/x86/ia32/ +# See arch/x86/Kbuild for content of core part of the kernel +core-y += arch/x86/ # drivers-y are linked after core-y drivers-$(CONFIG_MATH_EMULATION) += arch/x86/math-emu/ diff --git a/arch/x86/boot/.gitignore b/arch/x86/boot/.gitignore index 172cf8a98bd..851fe936d24 100644 --- a/arch/x86/boot/.gitignore +++ b/arch/x86/boot/.gitignore @@ -3,6 +3,8 @@ bzImage cpustr.h mkcpustr offsets.h +voffset.h +zoffset.h setup setup.bin setup.elf diff --git a/arch/x86/boot/Makefile b/arch/x86/boot/Makefile index 6633b6e7505..8d16ada2504 100644 --- a/arch/x86/boot/Makefile +++ b/arch/x86/boot/Makefile @@ -26,9 +26,10 @@ targets := vmlinux.bin setup.bin setup.elf bzImage targets += fdimage fdimage144 fdimage288 image.iso mtools.conf subdir- := compressed -setup-y += a20.o cmdline.o copy.o cpu.o cpucheck.o edd.o +setup-y += a20.o bioscall.o cmdline.o copy.o cpu.o cpucheck.o edd.o setup-y += header.o main.o mca.o memory.o pm.o pmjump.o -setup-y += printf.o string.o tty.o video.o video-mode.o version.o +setup-y += printf.o regs.o string.o tty.o video.o video-mode.o +setup-y += version.o setup-$(CONFIG_X86_APM_BOOT) += apm.o # The link order of the video-*.o modules can matter. In particular, @@ -86,19 +87,27 @@ $(obj)/vmlinux.bin: $(obj)/compressed/vmlinux FORCE SETUP_OBJS = $(addprefix $(obj)/,$(setup-y)) -sed-offsets := -e 's/^00*/0/' \ - -e 's/^\([0-9a-fA-F]*\) . \(input_data\|input_data_end\)$$/\#define \2 0x\1/p' +sed-voffset := -e 's/^\([0-9a-fA-F]*\) . \(_text\|_end\)$$/\#define VO_\2 0x\1/p' -quiet_cmd_offsets = OFFSETS $@ - cmd_offsets = $(NM) $< | sed -n $(sed-offsets) > $@ +quiet_cmd_voffset = VOFFSET $@ + cmd_voffset = $(NM) $< | sed -n $(sed-voffset) > $@ -$(obj)/offsets.h: $(obj)/compressed/vmlinux FORCE - $(call if_changed,offsets) +targets += voffset.h +$(obj)/voffset.h: vmlinux FORCE + $(call if_changed,voffset) + +sed-zoffset := -e 's/^\([0-9a-fA-F]*\) . \(startup_32\|input_data\|_end\|z_.*\)$$/\#define ZO_\2 0x\1/p' + +quiet_cmd_zoffset = ZOFFSET $@ + cmd_zoffset = $(NM) $< | sed -n $(sed-zoffset) > $@ + +targets += zoffset.h +$(obj)/zoffset.h: $(obj)/compressed/vmlinux FORCE + $(call if_changed,zoffset) -targets += offsets.h AFLAGS_header.o += -I$(obj) -$(obj)/header.o: $(obj)/offsets.h +$(obj)/header.o: $(obj)/voffset.h $(obj)/zoffset.h LDFLAGS_setup.elf := -T $(obj)/setup.elf: $(src)/setup.ld $(SETUP_OBJS) FORCE diff --git a/arch/x86/boot/a20.c b/arch/x86/boot/a20.c index 7c19ce8c244..64a31a6d751 100644 --- a/arch/x86/boot/a20.c +++ b/arch/x86/boot/a20.c @@ -2,7 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007-2008 rPath, Inc. - All Rights Reserved - * Copyright 2009 Intel Corporation + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -90,8 +90,11 @@ static int a20_test_long(void) static void enable_a20_bios(void) { - asm volatile("pushfl; int $0x15; popfl" - : : "a" ((u16)0x2401)); + struct biosregs ireg; + + initregs(&ireg); + ireg.ax = 0x2401; + intcall(0x15, &ireg, NULL); } static void enable_a20_kbc(void) diff --git a/arch/x86/boot/apm.c b/arch/x86/boot/apm.c index 7aa6033001f..ee274834ea8 100644 --- a/arch/x86/boot/apm.c +++ b/arch/x86/boot/apm.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * Original APM BIOS checking by Stephen Rothwell, May 1994 * (sfr@canb.auug.org.au) @@ -19,75 +20,56 @@ int query_apm_bios(void) { - u16 ax, bx, cx, dx, di; - u32 ebx, esi; - u8 err; + struct biosregs ireg, oreg; /* APM BIOS installation check */ - ax = 0x5300; - bx = cx = 0; - asm volatile("pushl %%ebp ; int $0x15 ; popl %%ebp ; setc %0" - : "=d" (err), "+a" (ax), "+b" (bx), "+c" (cx) - : : "esi", "edi"); + initregs(&ireg); + ireg.ah = 0x53; + intcall(0x15, &ireg, &oreg); - if (err) + if (oreg.flags & X86_EFLAGS_CF) return -1; /* No APM BIOS */ - if (bx != 0x504d) /* "PM" signature */ + if (oreg.bx != 0x504d) /* "PM" signature */ return -1; - if (!(cx & 0x02)) /* 32 bits supported? */ + if (!(oreg.cx & 0x02)) /* 32 bits supported? */ return -1; /* Disconnect first, just in case */ - ax = 0x5304; - bx = 0; - asm volatile("pushl %%ebp ; int $0x15 ; popl %%ebp" - : "+a" (ax), "+b" (bx) - : : "ecx", "edx", "esi", "edi"); - - /* Paranoia */ - ebx = esi = 0; - cx = dx = di = 0; + ireg.al = 0x04; + intcall(0x15, &ireg, NULL); /* 32-bit connect */ - asm volatile("pushl %%ebp ; int $0x15 ; popl %%ebp ; setc %6" - : "=a" (ax), "+b" (ebx), "+c" (cx), "+d" (dx), - "+S" (esi), "+D" (di), "=m" (err) - : "a" (0x5303)); - - boot_params.apm_bios_info.cseg = ax; - boot_params.apm_bios_info.offset = ebx; - boot_params.apm_bios_info.cseg_16 = cx; - boot_params.apm_bios_info.dseg = dx; - boot_params.apm_bios_info.cseg_len = (u16)esi; - boot_params.apm_bios_info.cseg_16_len = esi >> 16; - boot_params.apm_bios_info.dseg_len = di; - - if (err) + ireg.al = 0x03; + intcall(0x15, &ireg, &oreg); + + boot_params.apm_bios_info.cseg = oreg.ax; + boot_params.apm_bios_info.offset = oreg.ebx; + boot_params.apm_bios_info.cseg_16 = oreg.cx; + boot_params.apm_bios_info.dseg = oreg.dx; + boot_params.apm_bios_info.cseg_len = oreg.si; + boot_params.apm_bios_info.cseg_16_len = oreg.hsi; + boot_params.apm_bios_info.dseg_len = oreg.di; + + if (oreg.flags & X86_EFLAGS_CF) return -1; /* Redo the installation check as the 32-bit connect; some BIOSes return different flags this way... */ - ax = 0x5300; - bx = cx = 0; - asm volatile("pushl %%ebp ; int $0x15 ; popl %%ebp ; setc %0" - : "=d" (err), "+a" (ax), "+b" (bx), "+c" (cx) - : : "esi", "edi"); + ireg.al = 0x00; + intcall(0x15, &ireg, &oreg); - if (err || bx != 0x504d) { + if ((oreg.eflags & X86_EFLAGS_CF) || oreg.bx != 0x504d) { /* Failure with 32-bit connect, try to disconect and ignore */ - ax = 0x5304; - bx = 0; - asm volatile("pushl %%ebp ; int $0x15 ; popl %%ebp" - : "+a" (ax), "+b" (bx) - : : "ecx", "edx", "esi", "edi"); + ireg.al = 0x04; + intcall(0x15, &ireg, NULL); return -1; } - boot_params.apm_bios_info.version = ax; - boot_params.apm_bios_info.flags = cx; + boot_params.apm_bios_info.version = oreg.ax; + boot_params.apm_bios_info.flags = oreg.cx; return 0; } diff --git a/arch/x86/boot/bioscall.S b/arch/x86/boot/bioscall.S new file mode 100644 index 00000000000..507793739ea --- /dev/null +++ b/arch/x86/boot/bioscall.S @@ -0,0 +1,82 @@ +/* ----------------------------------------------------------------------- + * + * Copyright 2009 Intel Corporation; author H. Peter Anvin + * + * This file is part of the Linux kernel, and is made available under + * the terms of the GNU General Public License version 2 or (at your + * option) any later version; incorporated herein by reference. + * + * ----------------------------------------------------------------------- */ + +/* + * "Glove box" for BIOS calls. Avoids the constant problems with BIOSes + * touching registers they shouldn't be. + */ + + .code16 + .text + .globl intcall + .type intcall, @function +intcall: + /* Self-modify the INT instruction. Ugly, but works. */ + cmpb %al, 3f + je 1f + movb %al, 3f + jmp 1f /* Synchronize pipeline */ +1: + /* Save state */ + pushfl + pushw %fs + pushw %gs + pushal + + /* Copy input state to stack frame */ + subw $44, %sp + movw %dx, %si + movw %sp, %di + movw $11, %cx + rep; movsd + + /* Pop full state from the stack */ + popal + popw %gs + popw %fs + popw %es + popw %ds + popfl + + /* Actual INT */ + .byte 0xcd /* INT opcode */ +3: .byte 0 + + /* Push full state to the stack */ + pushfl + pushw %ds + pushw %es + pushw %fs + pushw %gs + pushal + + /* Re-establish C environment invariants */ + cld + movzwl %sp, %esp + movw %cs, %ax + movw %ax, %ds + movw %ax, %es + + /* Copy output state from stack frame */ + movw 68(%esp), %di /* Original %cx == 3rd argument */ + andw %di, %di + jz 4f + movw %sp, %si + movw $11, %cx + rep; movsd +4: addw $44, %sp + + /* Restore state and return */ + popal + popw %gs + popw %fs + popfl + retl + .size intcall, .-intcall diff --git a/arch/x86/boot/boot.h b/arch/x86/boot/boot.h index 7b2692e897e..98239d2658f 100644 --- a/arch/x86/boot/boot.h +++ b/arch/x86/boot/boot.h @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -26,6 +27,7 @@ #include <asm/setup.h> #include "bitops.h" #include <asm/cpufeature.h> +#include <asm/processor-flags.h> /* Useful macros */ #define BUILD_BUG_ON(condition) ((void)sizeof(char[1 - 2*!!(condition)])) @@ -241,6 +243,49 @@ int enable_a20(void); /* apm.c */ int query_apm_bios(void); +/* bioscall.c */ +struct biosregs { + union { + struct { + u32 edi; + u32 esi; + u32 ebp; + u32 _esp; + u32 ebx; + u32 edx; + u32 ecx; + u32 eax; + u32 _fsgs; + u32 _dses; + u32 eflags; + }; + struct { + u16 di, hdi; + u16 si, hsi; + u16 bp, hbp; + u16 _sp, _hsp; + u16 bx, hbx; + u16 dx, hdx; + u16 cx, hcx; + u16 ax, hax; + u16 gs, fs; + u16 es, ds; + u16 flags, hflags; + }; + struct { + u8 dil, dih, edi2, edi3; + u8 sil, sih, esi2, esi3; + u8 bpl, bph, ebp2, ebp3; + u8 _spl, _sph, _esp2, _esp3; + u8 bl, bh, ebx2, ebx3; + u8 dl, dh, edx2, edx3; + u8 cl, ch, ecx2, ecx3; + u8 al, ah, eax2, eax3; + }; + }; +}; +void intcall(u8 int_no, const struct biosregs *ireg, struct biosregs *oreg); + /* cmdline.c */ int cmdline_find_option(const char *option, char *buffer, int bufsize); int cmdline_find_option_bool(const char *option); @@ -279,6 +324,9 @@ int sprintf(char *buf, const char *fmt, ...); int vsprintf(char *buf, const char *fmt, va_list args); int printf(const char *fmt, ...); +/* regs.c */ +void initregs(struct biosregs *regs); + /* string.c */ int strcmp(const char *str1, const char *str2); size_t strnlen(const char *s, size_t maxlen); diff --git a/arch/x86/boot/compressed/.gitignore b/arch/x86/boot/compressed/.gitignore index 63eff3b04d0..4a46fab7162 100644 --- a/arch/x86/boot/compressed/.gitignore +++ b/arch/x86/boot/compressed/.gitignore @@ -1,3 +1,6 @@ relocs vmlinux.bin.all vmlinux.relocs +vmlinux.lds +mkpiggy +piggy.S diff --git a/arch/x86/boot/compressed/Makefile b/arch/x86/boot/compressed/Makefile index 65551c9f857..49c8a4c37d7 100644 --- a/arch/x86/boot/compressed/Makefile +++ b/arch/x86/boot/compressed/Makefile @@ -19,7 +19,9 @@ KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ LDFLAGS := -m elf_$(UTS_MACHINE) LDFLAGS_vmlinux := -T -$(obj)/vmlinux: $(src)/vmlinux_$(BITS).lds $(obj)/head_$(BITS).o $(obj)/misc.o $(obj)/piggy.o FORCE +hostprogs-y := mkpiggy + +$(obj)/vmlinux: $(obj)/vmlinux.lds $(obj)/head_$(BITS).o $(obj)/misc.o $(obj)/piggy.o FORCE $(call if_changed,ld) @: @@ -29,7 +31,7 @@ $(obj)/vmlinux.bin: vmlinux FORCE targets += vmlinux.bin.all vmlinux.relocs relocs -hostprogs-$(CONFIG_X86_32) += relocs +hostprogs-$(CONFIG_X86_NEED_RELOCS) += relocs quiet_cmd_relocs = RELOCS $@ cmd_relocs = $(obj)/relocs $< > $@;$(obj)/relocs --abs-relocs $< @@ -37,46 +39,22 @@ $(obj)/vmlinux.relocs: vmlinux $(obj)/relocs FORCE $(call if_changed,relocs) vmlinux.bin.all-y := $(obj)/vmlinux.bin -vmlinux.bin.all-$(CONFIG_RELOCATABLE) += $(obj)/vmlinux.relocs -quiet_cmd_relocbin = BUILD $@ - cmd_relocbin = cat $(filter-out FORCE,$^) > $@ -$(obj)/vmlinux.bin.all: $(vmlinux.bin.all-y) FORCE - $(call if_changed,relocbin) - -ifeq ($(CONFIG_X86_32),y) +vmlinux.bin.all-$(CONFIG_X86_NEED_RELOCS) += $(obj)/vmlinux.relocs -ifdef CONFIG_RELOCATABLE -$(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin.all FORCE - $(call if_changed,gzip) -$(obj)/vmlinux.bin.bz2: $(obj)/vmlinux.bin.all FORCE - $(call if_changed,bzip2) -$(obj)/vmlinux.bin.lzma: $(obj)/vmlinux.bin.all FORCE - $(call if_changed,lzma) -else -$(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE +$(obj)/vmlinux.bin.gz: $(vmlinux.bin.all-y) FORCE $(call if_changed,gzip) -$(obj)/vmlinux.bin.bz2: $(obj)/vmlinux.bin FORCE +$(obj)/vmlinux.bin.bz2: $(vmlinux.bin.all-y) FORCE $(call if_changed,bzip2) -$(obj)/vmlinux.bin.lzma: $(obj)/vmlinux.bin FORCE +$(obj)/vmlinux.bin.lzma: $(vmlinux.bin.all-y) FORCE $(call if_changed,lzma) -endif -LDFLAGS_piggy.o := -r --format binary --oformat elf32-i386 -T -else +suffix-$(CONFIG_KERNEL_GZIP) := gz +suffix-$(CONFIG_KERNEL_BZIP2) := bz2 +suffix-$(CONFIG_KERNEL_LZMA) := lzma -$(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE - $(call if_changed,gzip) -$(obj)/vmlinux.bin.bz2: $(obj)/vmlinux.bin FORCE - $(call if_changed,bzip2) -$(obj)/vmlinux.bin.lzma: $(obj)/vmlinux.bin FORCE - $(call if_changed,lzma) - -LDFLAGS_piggy.o := -r --format binary --oformat elf64-x86-64 -T -endif +quiet_cmd_mkpiggy = MKPIGGY $@ + cmd_mkpiggy = $(obj)/mkpiggy $< > $@ || ( rm -f $@ ; false ) -suffix_$(CONFIG_KERNEL_GZIP) = gz -suffix_$(CONFIG_KERNEL_BZIP2) = bz2 -suffix_$(CONFIG_KERNEL_LZMA) = lzma - -$(obj)/piggy.o: $(obj)/vmlinux.scr $(obj)/vmlinux.bin.$(suffix_y) FORCE - $(call if_changed,ld) +targets += piggy.S +$(obj)/piggy.S: $(obj)/vmlinux.bin.$(suffix-y) $(obj)/mkpiggy FORCE + $(call if_changed,mkpiggy) diff --git a/arch/x86/boot/compressed/head_32.S b/arch/x86/boot/compressed/head_32.S index 3a8a866fb2e..75e4f001e70 100644 --- a/arch/x86/boot/compressed/head_32.S +++ b/arch/x86/boot/compressed/head_32.S @@ -12,16 +12,16 @@ * the page directory. [According to comments etc elsewhere on a compressed * kernel it will end up at 0x1000 + 1Mb I hope so as I assume this. - AC] * - * Page 0 is deliberately kept safe, since System Management Mode code in + * Page 0 is deliberately kept safe, since System Management Mode code in * laptops may need to access the BIOS data stored there. This is also - * useful for future device drivers that either access the BIOS via VM86 + * useful for future device drivers that either access the BIOS via VM86 * mode. */ /* * High loaded stuff by Hans Lermen & Werner Almesberger, Feb. 1996 */ -.text + .text #include <linux/linkage.h> #include <asm/segment.h> @@ -29,161 +29,151 @@ #include <asm/boot.h> #include <asm/asm-offsets.h> -.section ".text.head","ax",@progbits + .section ".text.head","ax",@progbits ENTRY(startup_32) cld - /* test KEEP_SEGMENTS flag to see if the bootloader is asking - * us to not reload segments */ - testb $(1<<6), BP_loadflags(%esi) - jnz 1f + /* + * Test KEEP_SEGMENTS flag to see if the bootloader is asking + * us to not reload segments + */ + testb $(1<<6), BP_loadflags(%esi) + jnz 1f cli - movl $(__BOOT_DS),%eax - movl %eax,%ds - movl %eax,%es - movl %eax,%fs - movl %eax,%gs - movl %eax,%ss + movl $__BOOT_DS, %eax + movl %eax, %ds + movl %eax, %es + movl %eax, %fs + movl %eax, %gs + movl %eax, %ss 1: -/* Calculate the delta between where we were compiled to run +/* + * Calculate the delta between where we were compiled to run * at and where we were actually loaded at. This can only be done * with a short local call on x86. Nothing else will tell us what * address we are running at. The reserved chunk of the real-mode * data at 0x1e4 (defined as a scratch field) are used as the stack * for this calculation. Only 4 bytes are needed. */ - leal (0x1e4+4)(%esi), %esp - call 1f -1: popl %ebp - subl $1b, %ebp + leal (BP_scratch+4)(%esi), %esp + call 1f +1: popl %ebp + subl $1b, %ebp -/* %ebp contains the address we are loaded at by the boot loader and %ebx +/* + * %ebp contains the address we are loaded at by the boot loader and %ebx * contains the address where we should move the kernel image temporarily * for safe in-place decompression. */ #ifdef CONFIG_RELOCATABLE - movl %ebp, %ebx - addl $(CONFIG_PHYSICAL_ALIGN - 1), %ebx - andl $(~(CONFIG_PHYSICAL_ALIGN - 1)), %ebx + movl %ebp, %ebx + movl BP_kernel_alignment(%esi), %eax + decl %eax + addl %eax, %ebx + notl %eax + andl %eax, %ebx #else - movl $LOAD_PHYSICAL_ADDR, %ebx + movl $LOAD_PHYSICAL_ADDR, %ebx #endif - /* Replace the compressed data size with the uncompressed size */ - subl input_len(%ebp), %ebx - movl output_len(%ebp), %eax - addl %eax, %ebx - /* Add 8 bytes for every 32K input block */ - shrl $12, %eax - addl %eax, %ebx - /* Add 32K + 18 bytes of extra slack */ - addl $(32768 + 18), %ebx - /* Align on a 4K boundary */ - addl $4095, %ebx - andl $~4095, %ebx - -/* Copy the compressed kernel to the end of our buffer + /* Target address to relocate to for decompression */ + addl $z_extract_offset, %ebx + + /* Set up the stack */ + leal boot_stack_end(%ebx), %esp + + /* Zero EFLAGS */ + pushl $0 + popfl + +/* + * Copy the compressed kernel to the end of our buffer * where decompression in place becomes safe. */ - pushl %esi - leal _end(%ebp), %esi - leal _end(%ebx), %edi - movl $(_end - startup_32), %ecx + pushl %esi + leal (_bss-4)(%ebp), %esi + leal (_bss-4)(%ebx), %edi + movl $(_bss - startup_32), %ecx + shrl $2, %ecx std - rep - movsb + rep movsl cld - popl %esi - -/* Compute the kernel start address. - */ -#ifdef CONFIG_RELOCATABLE - addl $(CONFIG_PHYSICAL_ALIGN - 1), %ebp - andl $(~(CONFIG_PHYSICAL_ALIGN - 1)), %ebp -#else - movl $LOAD_PHYSICAL_ADDR, %ebp -#endif + popl %esi /* * Jump to the relocated address. */ - leal relocated(%ebx), %eax - jmp *%eax + leal relocated(%ebx), %eax + jmp *%eax ENDPROC(startup_32) -.section ".text" + .text relocated: /* - * Clear BSS - */ - xorl %eax,%eax - leal _edata(%ebx),%edi - leal _end(%ebx), %ecx - subl %edi,%ecx - cld - rep - stosb - -/* - * Setup the stack for the decompressor + * Clear BSS (stack is currently empty) */ - leal boot_stack_end(%ebx), %esp + xorl %eax, %eax + leal _bss(%ebx), %edi + leal _ebss(%ebx), %ecx + subl %edi, %ecx + shrl $2, %ecx + rep stosl /* * Do the decompression, and jump to the new kernel.. */ - movl output_len(%ebx), %eax - pushl %eax - # push arguments for decompress_kernel: - pushl %ebp # output address - movl input_len(%ebx), %eax - pushl %eax # input_len - leal input_data(%ebx), %eax - pushl %eax # input_data - leal boot_heap(%ebx), %eax - pushl %eax # heap area - pushl %esi # real mode pointer - call decompress_kernel - addl $20, %esp - popl %ecx + leal z_extract_offset_negative(%ebx), %ebp + /* push arguments for decompress_kernel: */ + pushl %ebp /* output address */ + pushl $z_input_len /* input_len */ + leal input_data(%ebx), %eax + pushl %eax /* input_data */ + leal boot_heap(%ebx), %eax + pushl %eax /* heap area */ + pushl %esi /* real mode pointer */ + call decompress_kernel + addl $20, %esp #if CONFIG_RELOCATABLE -/* Find the address of the relocations. +/* + * Find the address of the relocations. */ - movl %ebp, %edi - addl %ecx, %edi + leal z_output_len(%ebp), %edi -/* Calculate the delta between where vmlinux was compiled to run +/* + * Calculate the delta between where vmlinux was compiled to run * and where it was actually loaded. */ - movl %ebp, %ebx - subl $LOAD_PHYSICAL_ADDR, %ebx - jz 2f /* Nothing to be done if loaded at compiled addr. */ + movl %ebp, %ebx + subl $LOAD_PHYSICAL_ADDR, %ebx + jz 2f /* Nothing to be done if loaded at compiled addr. */ /* * Process relocations. */ -1: subl $4, %edi - movl 0(%edi), %ecx - testl %ecx, %ecx - jz 2f - addl %ebx, -__PAGE_OFFSET(%ebx, %ecx) - jmp 1b +1: subl $4, %edi + movl (%edi), %ecx + testl %ecx, %ecx + jz 2f + addl %ebx, -__PAGE_OFFSET(%ebx, %ecx) + jmp 1b 2: #endif /* * Jump to the decompressed kernel. */ - xorl %ebx,%ebx - jmp *%ebp + xorl %ebx, %ebx + jmp *%ebp -.bss -/* Stack and heap for uncompression */ -.balign 4 +/* + * Stack and heap for uncompression + */ + .bss + .balign 4 boot_heap: .fill BOOT_HEAP_SIZE, 1, 0 boot_stack: diff --git a/arch/x86/boot/compressed/head_64.S b/arch/x86/boot/compressed/head_64.S index ed4a8294800..f62c284db9e 100644 --- a/arch/x86/boot/compressed/head_64.S +++ b/arch/x86/boot/compressed/head_64.S @@ -21,8 +21,8 @@ /* * High loaded stuff by Hans Lermen & Werner Almesberger, Feb. 1996 */ -.code32 -.text + .code32 + .text #include <linux/linkage.h> #include <asm/segment.h> @@ -33,12 +33,14 @@ #include <asm/processor-flags.h> #include <asm/asm-offsets.h> -.section ".text.head" + .section ".text.head" .code32 ENTRY(startup_32) cld - /* test KEEP_SEGMENTS flag to see if the bootloader is asking - * us to not reload segments */ + /* + * Test KEEP_SEGMENTS flag to see if the bootloader is asking + * us to not reload segments + */ testb $(1<<6), BP_loadflags(%esi) jnz 1f @@ -49,14 +51,15 @@ ENTRY(startup_32) movl %eax, %ss 1: -/* Calculate the delta between where we were compiled to run +/* + * Calculate the delta between where we were compiled to run * at and where we were actually loaded at. This can only be done * with a short local call on x86. Nothing else will tell us what * address we are running at. The reserved chunk of the real-mode * data at 0x1e4 (defined as a scratch field) are used as the stack * for this calculation. Only 4 bytes are needed. */ - leal (0x1e4+4)(%esi), %esp + leal (BP_scratch+4)(%esi), %esp call 1f 1: popl %ebp subl $1b, %ebp @@ -70,32 +73,28 @@ ENTRY(startup_32) testl %eax, %eax jnz no_longmode -/* Compute the delta between where we were compiled to run at +/* + * Compute the delta between where we were compiled to run at * and where the code will actually run at. - */ -/* %ebp contains the address we are loaded at by the boot loader and %ebx + * + * %ebp contains the address we are loaded at by the boot loader and %ebx * contains the address where we should move the kernel image temporarily * for safe in-place decompression. */ #ifdef CONFIG_RELOCATABLE movl %ebp, %ebx - addl $(PMD_PAGE_SIZE -1), %ebx - andl $PMD_PAGE_MASK, %ebx + movl BP_kernel_alignment(%esi), %eax + decl %eax + addl %eax, %ebx + notl %eax + andl %eax, %ebx #else - movl $CONFIG_PHYSICAL_START, %ebx + movl $LOAD_PHYSICAL_ADDR, %ebx #endif - /* Replace the compressed data size with the uncompressed size */ - subl input_len(%ebp), %ebx - movl output_len(%ebp), %eax - addl %eax, %ebx - /* Add 8 bytes for every 32K input block */ - shrl $12, %eax - addl %eax, %ebx - /* Add 32K + 18 bytes of extra slack and align on a 4K boundary */ - addl $(32768 + 18 + 4095), %ebx - andl $~4095, %ebx + /* Target address to relocate to for decompression */ + addl $z_extract_offset, %ebx /* * Prepare for entering 64 bit mode @@ -114,7 +113,7 @@ ENTRY(startup_32) /* * Build early 4G boot pagetable */ - /* Initialize Page tables to 0*/ + /* Initialize Page tables to 0 */ leal pgtable(%ebx), %edi xorl %eax, %eax movl $((4096*6)/4), %ecx @@ -155,7 +154,8 @@ ENTRY(startup_32) btsl $_EFER_LME, %eax wrmsr - /* Setup for the jump to 64bit mode + /* + * Setup for the jump to 64bit mode * * When the jump is performend we will be in long mode but * in 32bit compatibility mode with EFER.LME = 1, CS.L = 0, CS.D = 1 @@ -184,7 +184,8 @@ no_longmode: #include "../../kernel/verify_cpu_64.S" - /* Be careful here startup_64 needs to be at a predictable + /* + * Be careful here startup_64 needs to be at a predictable * address so I can export it in an ELF header. Bootloaders * should look at the ELF header to find this address, as * it may change in the future. @@ -192,7 +193,8 @@ no_longmode: .code64 .org 0x200 ENTRY(startup_64) - /* We come here either from startup_32 or directly from a + /* + * We come here either from startup_32 or directly from a * 64bit bootloader. If we come here from a bootloader we depend on * an identity mapped page table being provied that maps our * entire text+data+bss and hopefully all of memory. @@ -209,50 +211,54 @@ ENTRY(startup_64) movl $0x20, %eax ltr %ax - /* Compute the decompressed kernel start address. It is where + /* + * Compute the decompressed kernel start address. It is where * we were loaded at aligned to a 2M boundary. %rbp contains the * decompressed kernel start address. * * If it is a relocatable kernel then decompress and run the kernel * from load address aligned to 2MB addr, otherwise decompress and - * run the kernel from CONFIG_PHYSICAL_START + * run the kernel from LOAD_PHYSICAL_ADDR + * + * We cannot rely on the calculation done in 32-bit mode, since we + * may have been invoked via the 64-bit entry point. */ /* Start with the delta to where the kernel will run at. */ #ifdef CONFIG_RELOCATABLE leaq startup_32(%rip) /* - $startup_32 */, %rbp - addq $(PMD_PAGE_SIZE - 1), %rbp - andq $PMD_PAGE_MASK, %rbp - movq %rbp, %rbx + movl BP_kernel_alignment(%rsi), %eax + decl %eax + addq %rax, %rbp + notq %rax + andq %rax, %rbp #else - movq $CONFIG_PHYSICAL_START, %rbp - movq %rbp, %rbx + movq $LOAD_PHYSICAL_ADDR, %rbp #endif - /* Replace the compressed data size with the uncompressed size */ - movl input_len(%rip), %eax - subq %rax, %rbx - movl output_len(%rip), %eax - addq %rax, %rbx - /* Add 8 bytes for every 32K input block */ - shrq $12, %rax - addq %rax, %rbx - /* Add 32K + 18 bytes of extra slack and align on a 4K boundary */ - addq $(32768 + 18 + 4095), %rbx - andq $~4095, %rbx - -/* Copy the compressed kernel to the end of our buffer + /* Target address to relocate to for decompression */ + leaq z_extract_offset(%rbp), %rbx + + /* Set up the stack */ + leaq boot_stack_end(%rbx), %rsp + + /* Zero EFLAGS */ + pushq $0 + popfq + +/* + * Copy the compressed kernel to the end of our buffer * where decompression in place becomes safe. */ - leaq _end_before_pgt(%rip), %r8 - leaq _end_before_pgt(%rbx), %r9 - movq $_end_before_pgt /* - $startup_32 */, %rcx -1: subq $8, %r8 - subq $8, %r9 - movq 0(%r8), %rax - movq %rax, 0(%r9) - subq $8, %rcx - jnz 1b + pushq %rsi + leaq (_bss-8)(%rip), %rsi + leaq (_bss-8)(%rbx), %rdi + movq $_bss /* - $startup_32 */, %rcx + shrq $3, %rcx + std + rep movsq + cld + popq %rsi /* * Jump to the relocated address. @@ -260,37 +266,28 @@ ENTRY(startup_64) leaq relocated(%rbx), %rax jmp *%rax -.section ".text" + .text relocated: /* - * Clear BSS + * Clear BSS (stack is currently empty) */ - xorq %rax, %rax - leaq _edata(%rbx), %rdi - leaq _end_before_pgt(%rbx), %rcx + xorl %eax, %eax + leaq _bss(%rip), %rdi + leaq _ebss(%rip), %rcx subq %rdi, %rcx - cld - rep - stosb - - /* Setup the stack */ - leaq boot_stack_end(%rip), %rsp - - /* zero EFLAGS after setting rsp */ - pushq $0 - popfq + shrq $3, %rcx + rep stosq /* * Do the decompression, and jump to the new kernel.. */ - pushq %rsi # Save the real mode argument - movq %rsi, %rdi # real mode address - leaq boot_heap(%rip), %rsi # malloc area for uncompression - leaq input_data(%rip), %rdx # input_data - movl input_len(%rip), %eax - movq %rax, %rcx # input_len - movq %rbp, %r8 # output + pushq %rsi /* Save the real mode argument */ + movq %rsi, %rdi /* real mode address */ + leaq boot_heap(%rip), %rsi /* malloc area for uncompression */ + leaq input_data(%rip), %rdx /* input_data */ + movl $z_input_len, %ecx /* input_len */ + movq %rbp, %r8 /* output target address */ call decompress_kernel popq %rsi @@ -311,11 +308,21 @@ gdt: .quad 0x0000000000000000 /* TS continued */ gdt_end: -.bss -/* Stack and heap for uncompression */ -.balign 4 +/* + * Stack and heap for uncompression + */ + .bss + .balign 4 boot_heap: .fill BOOT_HEAP_SIZE, 1, 0 boot_stack: .fill BOOT_STACK_SIZE, 1, 0 boot_stack_end: + +/* + * Space for page tables (not in .bss so not zeroed) + */ + .section ".pgtable","a",@nobits + .balign 4096 +pgtable: + .fill 6*4096, 1, 0 diff --git a/arch/x86/boot/compressed/misc.c b/arch/x86/boot/compressed/misc.c index e45be73684f..842b2a36174 100644 --- a/arch/x86/boot/compressed/misc.c +++ b/arch/x86/boot/compressed/misc.c @@ -325,21 +325,19 @@ asmlinkage void decompress_kernel(void *rmode, memptr heap, free_mem_ptr = heap; /* Heap */ free_mem_end_ptr = heap + BOOT_HEAP_SIZE; + if ((unsigned long)output & (MIN_KERNEL_ALIGN - 1)) + error("Destination address inappropriately aligned"); #ifdef CONFIG_X86_64 - if ((unsigned long)output & (__KERNEL_ALIGN - 1)) - error("Destination address not 2M aligned"); - if ((unsigned long)output >= 0xffffffffffUL) + if (heap > 0x3fffffffffffUL) error("Destination address too large"); #else - if ((u32)output & (CONFIG_PHYSICAL_ALIGN - 1)) - error("Destination address not CONFIG_PHYSICAL_ALIGN aligned"); if (heap > ((-__PAGE_OFFSET-(512<<20)-1) & 0x7fffffff)) error("Destination address too large"); +#endif #ifndef CONFIG_RELOCATABLE - if ((u32)output != LOAD_PHYSICAL_ADDR) + if ((unsigned long)output != LOAD_PHYSICAL_ADDR) error("Wrong destination address"); #endif -#endif if (!quiet) putstr("\nDecompressing Linux... "); diff --git a/arch/x86/boot/compressed/mkpiggy.c b/arch/x86/boot/compressed/mkpiggy.c new file mode 100644 index 00000000000..bcbd36c4143 --- /dev/null +++ b/arch/x86/boot/compressed/mkpiggy.c @@ -0,0 +1,97 @@ +/* ----------------------------------------------------------------------- * + * + * Copyright (C) 2009 Intel Corporation. All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version + * 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA + * 02110-1301, USA. + * + * H. Peter Anvin <hpa@linux.intel.com> + * + * ----------------------------------------------------------------------- */ + +/* + * Compute the desired load offset from a compressed program; outputs + * a small assembly wrapper with the appropriate symbols defined. + */ + +#include <stdlib.h> +#include <stdio.h> +#include <string.h> +#include <inttypes.h> + +static uint32_t getle32(const void *p) +{ + const uint8_t *cp = p; + + return (uint32_t)cp[0] + ((uint32_t)cp[1] << 8) + + ((uint32_t)cp[2] << 16) + ((uint32_t)cp[3] << 24); +} + +int main(int argc, char *argv[]) +{ + uint32_t olen; + long ilen; + unsigned long offs; + FILE *f; + + if (argc < 2) { + fprintf(stderr, "Usage: %s compressed_file\n", argv[0]); + return 1; + } + + /* Get the information for the compressed kernel image first */ + + f = fopen(argv[1], "r"); + if (!f) { + perror(argv[1]); + return 1; + } + + + if (fseek(f, -4L, SEEK_END)) { + perror(argv[1]); + } + fread(&olen, sizeof olen, 1, f); + ilen = ftell(f); + olen = getle32(&olen); + fclose(f); + + /* + * Now we have the input (compressed) and output (uncompressed) + * sizes, compute the necessary decompression offset... + */ + + offs = (olen > ilen) ? olen - ilen : 0; + offs += olen >> 12; /* Add 8 bytes for each 32K block */ + offs += 32*1024 + 18; /* Add 32K + 18 bytes slack */ + offs = (offs+4095) & ~4095; /* Round to a 4K boundary */ + + printf(".section \".rodata.compressed\",\"a\",@progbits\n"); + printf(".globl z_input_len\n"); + printf("z_input_len = %lu\n", ilen); + printf(".globl z_output_len\n"); + printf("z_output_len = %lu\n", (unsigned long)olen); + printf(".globl z_extract_offset\n"); + printf("z_extract_offset = 0x%lx\n", offs); + /* z_extract_offset_negative allows simplification of head_32.S */ + printf(".globl z_extract_offset_negative\n"); + printf("z_extract_offset_negative = -0x%lx\n", offs); + + printf(".globl input_data, input_data_end\n"); + printf("input_data:\n"); + printf(".incbin \"%s\"\n", argv[1]); + printf("input_data_end:\n"); + + return 0; +} diff --git a/arch/x86/boot/compressed/vmlinux_64.lds b/arch/x86/boot/compressed/vmlinux.lds.S index bef1ac891bc..cc353e1b3ff 100644 --- a/arch/x86/boot/compressed/vmlinux_64.lds +++ b/arch/x86/boot/compressed/vmlinux.lds.S @@ -1,6 +1,17 @@ -OUTPUT_FORMAT("elf64-x86-64", "elf64-x86-64", "elf64-x86-64") +OUTPUT_FORMAT(CONFIG_OUTPUT_FORMAT, CONFIG_OUTPUT_FORMAT, CONFIG_OUTPUT_FORMAT) + +#undef i386 + +#include <asm/page_types.h> + +#ifdef CONFIG_X86_64 OUTPUT_ARCH(i386:x86-64) ENTRY(startup_64) +#else +OUTPUT_ARCH(i386) +ENTRY(startup_32) +#endif + SECTIONS { /* Be careful parts of head_64.S assume startup_32 is at @@ -33,16 +44,22 @@ SECTIONS *(.data.*) _edata = . ; } + . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); .bss : { _bss = . ; *(.bss) *(.bss.*) *(COMMON) - . = ALIGN(8); - _end_before_pgt = . ; - . = ALIGN(4096); - pgtable = . ; - . = . + 4096 * 6; + . = ALIGN(8); /* For convenience during zeroing */ _ebss = .; } +#ifdef CONFIG_X86_64 + . = ALIGN(PAGE_SIZE); + .pgtable : { + _pgtable = . ; + *(.pgtable) + _epgtable = . ; + } +#endif + _end = .; } diff --git a/arch/x86/boot/compressed/vmlinux.scr b/arch/x86/boot/compressed/vmlinux.scr deleted file mode 100644 index f02382ae5c4..00000000000 --- a/arch/x86/boot/compressed/vmlinux.scr +++ /dev/null @@ -1,10 +0,0 @@ -SECTIONS -{ - .rodata.compressed : { - input_len = .; - LONG(input_data_end - input_data) input_data = .; - *(.data) - output_len = . - 4; - input_data_end = .; - } -} diff --git a/arch/x86/boot/compressed/vmlinux_32.lds b/arch/x86/boot/compressed/vmlinux_32.lds deleted file mode 100644 index bb3c48379c4..00000000000 --- a/arch/x86/boot/compressed/vmlinux_32.lds +++ /dev/null @@ -1,43 +0,0 @@ -OUTPUT_FORMAT("elf32-i386", "elf32-i386", "elf32-i386") -OUTPUT_ARCH(i386) -ENTRY(startup_32) -SECTIONS -{ - /* Be careful parts of head_32.S assume startup_32 is at - * address 0. - */ - . = 0; - .text.head : { - _head = . ; - *(.text.head) - _ehead = . ; - } - .rodata.compressed : { - *(.rodata.compressed) - } - .text : { - _text = .; /* Text */ - *(.text) - *(.text.*) - _etext = . ; - } - .rodata : { - _rodata = . ; - *(.rodata) /* read-only data */ - *(.rodata.*) - _erodata = . ; - } - .data : { - _data = . ; - *(.data) - *(.data.*) - _edata = . ; - } - .bss : { - _bss = . ; - *(.bss) - *(.bss.*) - *(COMMON) - _end = . ; - } -} diff --git a/arch/x86/boot/edd.c b/arch/x86/boot/edd.c index 1aae8f3e5ca..c501a5b466f 100644 --- a/arch/x86/boot/edd.c +++ b/arch/x86/boot/edd.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -22,17 +23,17 @@ */ static int read_mbr(u8 devno, void *buf) { - u16 ax, bx, cx, dx; + struct biosregs ireg, oreg; - ax = 0x0201; /* Legacy Read, one sector */ - cx = 0x0001; /* Sector 0-0-1 */ - dx = devno; - bx = (size_t)buf; - asm volatile("pushfl; stc; int $0x13; setc %%al; popfl" - : "+a" (ax), "+c" (cx), "+d" (dx), "+b" (bx) - : : "esi", "edi", "memory"); + initregs(&ireg); + ireg.ax = 0x0201; /* Legacy Read, one sector */ + ireg.cx = 0x0001; /* Sector 0-0-1 */ + ireg.dl = devno; + ireg.bx = (size_t)buf; - return -(u8)ax; /* 0 or -1 */ + intcall(0x13, &ireg, &oreg); + + return -(oreg.eflags & X86_EFLAGS_CF); /* 0 or -1 */ } static u32 read_mbr_sig(u8 devno, struct edd_info *ei, u32 *mbrsig) @@ -72,56 +73,46 @@ static u32 read_mbr_sig(u8 devno, struct edd_info *ei, u32 *mbrsig) static int get_edd_info(u8 devno, struct edd_info *ei) { - u16 ax, bx, cx, dx, di; + struct biosregs ireg, oreg; memset(ei, 0, sizeof *ei); /* Check Extensions Present */ - ax = 0x4100; - bx = EDDMAGIC1; - dx = devno; - asm("pushfl; stc; int $0x13; setc %%al; popfl" - : "+a" (ax), "+b" (bx), "=c" (cx), "+d" (dx) - : : "esi", "edi"); + initregs(&ireg); + ireg.ah = 0x41; + ireg.bx = EDDMAGIC1; + ireg.dl = devno; + intcall(0x13, &ireg, &oreg); - if ((u8)ax) + if (oreg.eflags & X86_EFLAGS_CF) return -1; /* No extended information */ - if (bx != EDDMAGIC2) + if (oreg.bx != EDDMAGIC2) return -1; ei->device = devno; - ei->version = ax >> 8; /* EDD version number */ - ei->interface_support = cx; /* EDD functionality subsets */ + ei->version = oreg.ah; /* EDD version number */ + ei->interface_support = oreg.cx; /* EDD functionality subsets */ /* Extended Get Device Parameters */ ei->params.length = sizeof(ei->params); - ax = 0x4800; - dx = devno; - asm("pushfl; int $0x13; popfl" - : "+a" (ax), "+d" (dx), "=m" (ei->params) - : "S" (&ei->params) - : "ebx", "ecx", "edi"); + ireg.ah = 0x48; + ireg.si = (size_t)&ei->params; + intcall(0x13, &ireg, &oreg); /* Get legacy CHS parameters */ /* Ralf Brown recommends setting ES:DI to 0:0 */ - ax = 0x0800; - dx = devno; - di = 0; - asm("pushw %%es; " - "movw %%di,%%es; " - "pushfl; stc; int $0x13; setc %%al; popfl; " - "popw %%es" - : "+a" (ax), "=b" (bx), "=c" (cx), "+d" (dx), "+D" (di) - : : "esi"); - - if ((u8)ax == 0) { - ei->legacy_max_cylinder = (cx >> 8) + ((cx & 0xc0) << 2); - ei->legacy_max_head = dx >> 8; - ei->legacy_sectors_per_track = cx & 0x3f; + ireg.ah = 0x08; + ireg.es = 0; + intcall(0x13, &ireg, &oreg); + + if (!(oreg.eflags & X86_EFLAGS_CF)) { + ei->legacy_max_cylinder = oreg.ch + ((oreg.cl & 0xc0) << 2); + ei->legacy_max_head = oreg.dh; + ei->legacy_sectors_per_track = oreg.cl & 0x3f; } return 0; diff --git a/arch/x86/boot/header.S b/arch/x86/boot/header.S index 5d84d1c74e4..b31cc54b464 100644 --- a/arch/x86/boot/header.S +++ b/arch/x86/boot/header.S @@ -22,7 +22,8 @@ #include <asm/page_types.h> #include <asm/setup.h> #include "boot.h" -#include "offsets.h" +#include "voffset.h" +#include "zoffset.h" BOOTSEG = 0x07C0 /* original address of boot-sector */ SYSSEG = 0x1000 /* historical load address >> 4 */ @@ -115,7 +116,7 @@ _start: # Part 2 of the header, from the old setup.S .ascii "HdrS" # header signature - .word 0x0209 # header version number (>= 0x0105) + .word 0x020a # header version number (>= 0x0105) # or else old loadlin-1.5 will fail) .globl realmode_swtch realmode_swtch: .word 0, 0 # default_switch, SETUPSEG @@ -168,7 +169,11 @@ heap_end_ptr: .word _end+STACK_SIZE-512 # end of setup code can be used by setup # for local heap purposes. -pad1: .word 0 +ext_loader_ver: + .byte 0 # Extended boot loader version +ext_loader_type: + .byte 0 # Extended boot loader type + cmd_line_ptr: .long 0 # (Header version 0x0202 or later) # If nonzero, a 32-bit pointer # to the kernel command line. @@ -200,7 +205,7 @@ relocatable_kernel: .byte 1 #else relocatable_kernel: .byte 0 #endif -pad2: .byte 0 +min_alignment: .byte MIN_KERNEL_ALIGN_LG2 # minimum alignment pad3: .word 0 cmdline_size: .long COMMAND_LINE_SIZE-1 #length of the command line, @@ -212,16 +217,27 @@ hardware_subarch: .long 0 # subarchitecture, added with 2.07 hardware_subarch_data: .quad 0 -payload_offset: .long input_data -payload_length: .long input_data_end-input_data +payload_offset: .long ZO_input_data +payload_length: .long ZO_z_input_len setup_data: .quad 0 # 64-bit physical pointer to # single linked list of # struct setup_data +pref_address: .quad LOAD_PHYSICAL_ADDR # preferred load addr + +#define ZO_INIT_SIZE (ZO__end - ZO_startup_32 + ZO_z_extract_offset) +#define VO_INIT_SIZE (VO__end - VO__text) +#if ZO_INIT_SIZE > VO_INIT_SIZE +#define INIT_SIZE ZO_INIT_SIZE +#else +#define INIT_SIZE VO_INIT_SIZE +#endif +init_size: .long INIT_SIZE # kernel initialization size + # End of setup header ##################################################### - .section ".inittext", "ax" + .section ".entrytext", "ax" start_of_setup: #ifdef SAFE_RESET_DISK_CONTROLLER # Reset the disk controller. diff --git a/arch/x86/boot/main.c b/arch/x86/boot/main.c index 58f0415d3ae..140172b895b 100644 --- a/arch/x86/boot/main.c +++ b/arch/x86/boot/main.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -61,11 +62,10 @@ static void copy_boot_params(void) */ static void keyboard_set_repeat(void) { - u16 ax = 0x0305; - u16 bx = 0; - asm volatile("int $0x16" - : "+a" (ax), "+b" (bx) - : : "ecx", "edx", "esi", "edi"); + struct biosregs ireg; + initregs(&ireg); + ireg.ax = 0x0305; + intcall(0x16, &ireg, NULL); } /* @@ -73,18 +73,22 @@ static void keyboard_set_repeat(void) */ static void query_ist(void) { + struct biosregs ireg, oreg; + /* Some older BIOSes apparently crash on this call, so filter it from machines too old to have SpeedStep at all. */ if (cpu.level < 6) return; - asm("int $0x15" - : "=a" (boot_params.ist_info.signature), - "=b" (boot_params.ist_info.command), - "=c" (boot_params.ist_info.event), - "=d" (boot_params.ist_info.perf_level) - : "a" (0x0000e980), /* IST Support */ - "d" (0x47534943)); /* Request value */ + initregs(&ireg); + ireg.ax = 0xe980; /* IST Support */ + ireg.edx = 0x47534943; /* Request value */ + intcall(0x15, &ireg, &oreg); + + boot_params.ist_info.signature = oreg.eax; + boot_params.ist_info.command = oreg.ebx; + boot_params.ist_info.event = oreg.ecx; + boot_params.ist_info.perf_level = oreg.edx; } /* @@ -93,13 +97,12 @@ static void query_ist(void) static void set_bios_mode(void) { #ifdef CONFIG_X86_64 - u32 eax, ebx; + struct biosregs ireg; - eax = 0xec00; - ebx = 2; - asm volatile("int $0x15" - : "+a" (eax), "+b" (ebx) - : : "ecx", "edx", "esi", "edi"); + initregs(&ireg); + ireg.ax = 0xec00; + ireg.bx = 2; + intcall(0x15, &ireg, NULL); #endif } diff --git a/arch/x86/boot/mca.c b/arch/x86/boot/mca.c index 911eaae5d69..a95a531148e 100644 --- a/arch/x86/boot/mca.c +++ b/arch/x86/boot/mca.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -16,26 +17,22 @@ int query_mca(void) { - u8 err; - u16 es, bx, len; - - asm("pushw %%es ; " - "int $0x15 ; " - "setc %0 ; " - "movw %%es, %1 ; " - "popw %%es" - : "=acd" (err), "=acdSD" (es), "=b" (bx) - : "a" (0xc000)); - - if (err) + struct biosregs ireg, oreg; + u16 len; + + initregs(&ireg); + ireg.ah = 0xc0; + intcall(0x15, &ireg, &oreg); + + if (oreg.eflags & X86_EFLAGS_CF) return -1; /* No MCA present */ - set_fs(es); - len = rdfs16(bx); + set_fs(oreg.es); + len = rdfs16(oreg.bx); if (len > sizeof(boot_params.sys_desc_table)) len = sizeof(boot_params.sys_desc_table); - copy_from_fs(&boot_params.sys_desc_table, bx, len); + copy_from_fs(&boot_params.sys_desc_table, oreg.bx, len); return 0; } diff --git a/arch/x86/boot/memory.c b/arch/x86/boot/memory.c index 74b3d2ba84e..cae3feb1035 100644 --- a/arch/x86/boot/memory.c +++ b/arch/x86/boot/memory.c @@ -20,12 +20,16 @@ static int detect_memory_e820(void) { int count = 0; - u32 next = 0; - u32 size, id, edi; - u8 err; + struct biosregs ireg, oreg; struct e820entry *desc = boot_params.e820_map; static struct e820entry buf; /* static so it is zeroed */ + initregs(&ireg); + ireg.ax = 0xe820; + ireg.cx = sizeof buf; + ireg.edx = SMAP; + ireg.di = (size_t)&buf; + /* * Note: at least one BIOS is known which assumes that the * buffer pointed to by one e820 call is the same one as @@ -41,22 +45,13 @@ static int detect_memory_e820(void) */ do { - size = sizeof buf; - - /* Important: %edx and %esi are clobbered by some BIOSes, - so they must be either used for the error output - or explicitly marked clobbered. Given that, assume there - is something out there clobbering %ebp and %edi, too. */ - asm("pushl %%ebp; int $0x15; popl %%ebp; setc %0" - : "=d" (err), "+b" (next), "=a" (id), "+c" (size), - "=D" (edi), "+m" (buf) - : "D" (&buf), "d" (SMAP), "a" (0xe820) - : "esi"); + intcall(0x15, &ireg, &oreg); + ireg.ebx = oreg.ebx; /* for next iteration... */ /* BIOSes which terminate the chain with CF = 1 as opposed to %ebx = 0 don't always report the SMAP signature on the final, failing, probe. */ - if (err) + if (oreg.eflags & X86_EFLAGS_CF) break; /* Some BIOSes stop returning SMAP in the middle of @@ -64,60 +59,64 @@ static int detect_memory_e820(void) screwed up the map at that point, we might have a partial map, the full map, or complete garbage, so just return failure. */ - if (id != SMAP) { + if (oreg.eax != SMAP) { count = 0; break; } *desc++ = buf; count++; - } while (next && count < ARRAY_SIZE(boot_params.e820_map)); + } while (ireg.ebx && count < ARRAY_SIZE(boot_params.e820_map)); return boot_params.e820_entries = count; } static int detect_memory_e801(void) { - u16 ax, bx, cx, dx; - u8 err; + struct biosregs ireg, oreg; - bx = cx = dx = 0; - ax = 0xe801; - asm("stc; int $0x15; setc %0" - : "=m" (err), "+a" (ax), "+b" (bx), "+c" (cx), "+d" (dx)); + initregs(&ireg); + ireg.ax = 0xe801; + intcall(0x15, &ireg, &oreg); - if (err) + if (oreg.eflags & X86_EFLAGS_CF) return -1; /* Do we really need to do this? */ - if (cx || dx) { - ax = cx; - bx = dx; + if (oreg.cx || oreg.dx) { + oreg.ax = oreg.cx; + oreg.bx = oreg.dx; } - if (ax > 15*1024) + if (oreg.ax > 15*1024) { return -1; /* Bogus! */ - - /* This ignores memory above 16MB if we have a memory hole - there. If someone actually finds a machine with a memory - hole at 16MB and no support for 0E820h they should probably - generate a fake e820 map. */ - boot_params.alt_mem_k = (ax == 15*1024) ? (dx << 6)+ax : ax; + } else if (oreg.ax == 15*1024) { + boot_params.alt_mem_k = (oreg.dx << 6) + oreg.ax; + } else { + /* + * This ignores memory above 16MB if we have a memory + * hole there. If someone actually finds a machine + * with a memory hole at 16MB and no support for + * 0E820h they should probably generate a fake e820 + * map. + */ + boot_params.alt_mem_k = oreg.ax; + } return 0; } static int detect_memory_88(void) { - u16 ax; - u8 err; + struct biosregs ireg, oreg; - ax = 0x8800; - asm("stc; int $0x15; setc %0" : "=bcdm" (err), "+a" (ax)); + initregs(&ireg); + ireg.ah = 0x88; + intcall(0x15, &ireg, &oreg); - boot_params.screen_info.ext_mem_k = ax; + boot_params.screen_info.ext_mem_k = oreg.ax; - return -err; + return -(oreg.eflags & X86_EFLAGS_CF); /* 0 or -1 */ } int detect_memory(void) diff --git a/arch/x86/boot/regs.c b/arch/x86/boot/regs.c new file mode 100644 index 00000000000..958019b1cfa --- /dev/null +++ b/arch/x86/boot/regs.c @@ -0,0 +1,29 @@ +/* ----------------------------------------------------------------------- + * + * Copyright 2009 Intel Corporation; author H. Peter Anvin + * + * This file is part of the Linux kernel, and is made available under + * the terms of the GNU General Public License version 2 or (at your + * option) any later version; incorporated herein by reference. + * + * ----------------------------------------------------------------------- */ + +/* + * Simple helper function for initializing a register set. + * + * Note that this sets EFLAGS_CF in the input register set; this + * makes it easier to catch functions which do nothing but don't + * explicitly set CF. + */ + +#include "boot.h" + +void initregs(struct biosregs *reg) +{ + memset(reg, 0, sizeof *reg); + reg->eflags |= X86_EFLAGS_CF; + reg->ds = ds(); + reg->es = ds(); + reg->fs = fs(); + reg->gs = gs(); +} diff --git a/arch/x86/boot/setup.ld b/arch/x86/boot/setup.ld index bb8dc2de796..0f6ec455a2b 100644 --- a/arch/x86/boot/setup.ld +++ b/arch/x86/boot/setup.ld @@ -15,8 +15,11 @@ SECTIONS . = 497; .header : { *(.header) } + .entrytext : { *(.entrytext) } .inittext : { *(.inittext) } .initdata : { *(.initdata) } + __end_init = .; + .text : { *(.text) } .text32 : { *(.text32) } @@ -52,4 +55,7 @@ SECTIONS . = ASSERT(_end <= 0x8000, "Setup too big!"); . = ASSERT(hdr == 0x1f1, "The setup header has the wrong offset!"); + /* Necessary for the very-old-loader check to work... */ + . = ASSERT(__end_init <= 5*512, "init sections too big!"); + } diff --git a/arch/x86/boot/tty.c b/arch/x86/boot/tty.c index 7e8e8b25f5f..01ec69c901c 100644 --- a/arch/x86/boot/tty.c +++ b/arch/x86/boot/tty.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -22,24 +23,23 @@ void __attribute__((section(".inittext"))) putchar(int ch) { - unsigned char c = ch; + struct biosregs ireg; - if (c == '\n') + if (ch == '\n') putchar('\r'); /* \n -> \r\n */ - /* int $0x10 is known to have bugs involving touching registers - it shouldn't. Be extra conservative... */ - asm volatile("pushal; pushw %%ds; int $0x10; popw %%ds; popal" - : : "b" (0x0007), "c" (0x0001), "a" (0x0e00|ch)); + initregs(&ireg); + ireg.bx = 0x0007; + ireg.cx = 0x0001; + ireg.ah = 0x0e; + ireg.al = ch; + intcall(0x10, &ireg, NULL); } void __attribute__((section(".inittext"))) puts(const char *str) { - int n = 0; - while (*str) { + while (*str) putchar(*str++); - n++; - } } /* @@ -49,14 +49,13 @@ void __attribute__((section(".inittext"))) puts(const char *str) static u8 gettime(void) { - u16 ax = 0x0200; - u16 cx, dx; + struct biosregs ireg, oreg; - asm volatile("int $0x1a" - : "+a" (ax), "=c" (cx), "=d" (dx) - : : "ebx", "esi", "edi"); + initregs(&ireg); + ireg.ah = 0x02; + intcall(0x1a, &ireg, &oreg); - return dx >> 8; + return oreg.dh; } /* @@ -64,19 +63,24 @@ static u8 gettime(void) */ int getchar(void) { - u16 ax = 0; - asm volatile("int $0x16" : "+a" (ax)); + struct biosregs ireg, oreg; + + initregs(&ireg); + /* ireg.ah = 0x00; */ + intcall(0x16, &ireg, &oreg); - return ax & 0xff; + return oreg.al; } static int kbd_pending(void) { - u8 pending; - asm volatile("int $0x16; setnz %0" - : "=qm" (pending) - : "a" (0x0100)); - return pending; + struct biosregs ireg, oreg; + + initregs(&ireg); + ireg.ah = 0x01; + intcall(0x16, &ireg, &oreg); + + return !(oreg.eflags & X86_EFLAGS_ZF); } void kbd_flush(void) diff --git a/arch/x86/boot/video-bios.c b/arch/x86/boot/video-bios.c index 3fa979c9c36..d660be49236 100644 --- a/arch/x86/boot/video-bios.c +++ b/arch/x86/boot/video-bios.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -29,21 +30,21 @@ static int bios_set_mode(struct mode_info *mi) static int set_bios_mode(u8 mode) { - u16 ax; + struct biosregs ireg, oreg; u8 new_mode; - ax = mode; /* AH=0x00 Set Video Mode */ - asm volatile(INT10 - : "+a" (ax) - : : "ebx", "ecx", "edx", "esi", "edi"); + initregs(&ireg); + ireg.al = mode; /* AH=0x00 Set Video Mode */ + intcall(0x10, &ireg, NULL); - ax = 0x0f00; /* Get Current Video Mode */ - asm volatile(INT10 - : "+a" (ax) - : : "ebx", "ecx", "edx", "esi", "edi"); + + ireg.ah = 0x0f; /* Get Current Video Mode */ + intcall(0x10, &ireg, &oreg); do_restore = 1; /* Assume video contents were lost */ - new_mode = ax & 0x7f; /* Not all BIOSes are clean with the top bit */ + + /* Not all BIOSes are clean with the top bit */ + new_mode = ireg.al & 0x7f; if (new_mode == mode) return 0; /* Mode change OK */ @@ -53,10 +54,8 @@ static int set_bios_mode(u8 mode) /* Mode setting failed, but we didn't end up where we started. That's bad. Try to revert to the original video mode. */ - ax = boot_params.screen_info.orig_video_mode; - asm volatile(INT10 - : "+a" (ax) - : : "ebx", "ecx", "edx", "esi", "edi"); + ireg.ax = boot_params.screen_info.orig_video_mode; + intcall(0x10, &ireg, NULL); } #endif return -1; diff --git a/arch/x86/boot/video-vesa.c b/arch/x86/boot/video-vesa.c index 4a58c8ce3f6..c700147d6ff 100644 --- a/arch/x86/boot/video-vesa.c +++ b/arch/x86/boot/video-vesa.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -31,7 +32,7 @@ static inline void vesa_store_mode_params_graphics(void) {} static int vesa_probe(void) { #if defined(CONFIG_VIDEO_VESA) || defined(CONFIG_FIRMWARE_EDID) - u16 ax, cx, di; + struct biosregs ireg, oreg; u16 mode; addr_t mode_ptr; struct mode_info *mi; @@ -39,13 +40,12 @@ static int vesa_probe(void) video_vesa.modes = GET_HEAP(struct mode_info, 0); - ax = 0x4f00; - di = (size_t)&vginfo; - asm(INT10 - : "+a" (ax), "+D" (di), "=m" (vginfo) - : : "ebx", "ecx", "edx", "esi"); + initregs(&ireg); + ireg.ax = 0x4f00; + ireg.di = (size_t)&vginfo; + intcall(0x10, &ireg, &oreg); - if (ax != 0x004f || + if (ireg.ax != 0x004f || vginfo.signature != VESA_MAGIC || vginfo.version < 0x0102) return 0; /* Not present */ @@ -65,14 +65,12 @@ static int vesa_probe(void) memset(&vminfo, 0, sizeof vminfo); /* Just in case... */ - ax = 0x4f01; - cx = mode; - di = (size_t)&vminfo; - asm(INT10 - : "+a" (ax), "+c" (cx), "+D" (di), "=m" (vminfo) - : : "ebx", "edx", "esi"); + ireg.ax = 0x4f01; + ireg.cx = mode; + ireg.di = (size_t)&vminfo; + intcall(0x10, &ireg, &oreg); - if (ax != 0x004f) + if (ireg.ax != 0x004f) continue; if ((vminfo.mode_attr & 0x15) == 0x05) { @@ -111,20 +109,19 @@ static int vesa_probe(void) static int vesa_set_mode(struct mode_info *mode) { - u16 ax, bx, cx, di; + struct biosregs ireg, oreg; int is_graphic; u16 vesa_mode = mode->mode - VIDEO_FIRST_VESA; memset(&vminfo, 0, sizeof vminfo); /* Just in case... */ - ax = 0x4f01; - cx = vesa_mode; - di = (size_t)&vminfo; - asm(INT10 - : "+a" (ax), "+c" (cx), "+D" (di), "=m" (vminfo) - : : "ebx", "edx", "esi"); + initregs(&ireg); + ireg.ax = 0x4f01; + ireg.cx = vesa_mode; + ireg.di = (size_t)&vminfo; + intcall(0x10, &ireg, &oreg); - if (ax != 0x004f) + if (oreg.ax != 0x004f) return -1; if ((vminfo.mode_attr & 0x15) == 0x05) { @@ -141,14 +138,12 @@ static int vesa_set_mode(struct mode_info *mode) } - ax = 0x4f02; - bx = vesa_mode; - di = 0; - asm volatile(INT10 - : "+a" (ax), "+b" (bx), "+D" (di) - : : "ecx", "edx", "esi"); + initregs(&ireg); + ireg.ax = 0x4f02; + ireg.bx = vesa_mode; + intcall(0x10, &ireg, &oreg); - if (ax != 0x004f) + if (oreg.ax != 0x004f) return -1; graphic_mode = is_graphic; @@ -171,50 +166,45 @@ static int vesa_set_mode(struct mode_info *mode) /* Switch DAC to 8-bit mode */ static void vesa_dac_set_8bits(void) { + struct biosregs ireg, oreg; u8 dac_size = 6; /* If possible, switch the DAC to 8-bit mode */ if (vginfo.capabilities & 1) { - u16 ax, bx; - - ax = 0x4f08; - bx = 0x0800; - asm volatile(INT10 - : "+a" (ax), "+b" (bx) - : : "ecx", "edx", "esi", "edi"); - - if (ax == 0x004f) - dac_size = bx >> 8; + initregs(&ireg); + ireg.ax = 0x4f08; + ireg.bh = 0x08; + intcall(0x10, &ireg, &oreg); + if (oreg.ax == 0x004f) + dac_size = oreg.bh; } /* Set the color sizes to the DAC size, and offsets to 0 */ - boot_params.screen_info.red_size = dac_size; + boot_params.screen_info.red_size = dac_size; boot_params.screen_info.green_size = dac_size; - boot_params.screen_info.blue_size = dac_size; - boot_params.screen_info.rsvd_size = dac_size; + boot_params.screen_info.blue_size = dac_size; + boot_params.screen_info.rsvd_size = dac_size; - boot_params.screen_info.red_pos = 0; - boot_params.screen_info.green_pos = 0; - boot_params.screen_info.blue_pos = 0; - boot_params.screen_info.rsvd_pos = 0; + boot_params.screen_info.red_pos = 0; + boot_params.screen_info.green_pos = 0; + boot_params.screen_info.blue_pos = 0; + boot_params.screen_info.rsvd_pos = 0; } /* Save the VESA protected mode info */ static void vesa_store_pm_info(void) { - u16 ax, bx, di, es; + struct biosregs ireg, oreg; - ax = 0x4f0a; - bx = di = 0; - asm("pushw %%es; "INT10"; movw %%es,%0; popw %%es" - : "=d" (es), "+a" (ax), "+b" (bx), "+D" (di) - : : "ecx", "esi"); + initregs(&ireg); + ireg.ax = 0x4f0a; + intcall(0x10, &ireg, &oreg); - if (ax != 0x004f) + if (oreg.ax != 0x004f) return; - boot_params.screen_info.vesapm_seg = es; - boot_params.screen_info.vesapm_off = di; + boot_params.screen_info.vesapm_seg = oreg.es; + boot_params.screen_info.vesapm_off = oreg.di; } /* @@ -252,7 +242,7 @@ static void vesa_store_mode_params_graphics(void) void vesa_store_edid(void) { #ifdef CONFIG_FIRMWARE_EDID - u16 ax, bx, cx, dx, di; + struct biosregs ireg, oreg; /* Apparently used as a nonsense token... */ memset(&boot_params.edid_info, 0x13, sizeof boot_params.edid_info); @@ -260,33 +250,26 @@ void vesa_store_edid(void) if (vginfo.version < 0x0200) return; /* EDID requires VBE 2.0+ */ - ax = 0x4f15; /* VBE DDC */ - bx = 0x0000; /* Report DDC capabilities */ - cx = 0; /* Controller 0 */ - di = 0; /* ES:DI must be 0 by spec */ - - /* Note: The VBE DDC spec is different from the main VESA spec; - we genuinely have to assume all registers are destroyed here. */ - - asm("pushw %%es; movw %2,%%es; "INT10"; popw %%es" - : "+a" (ax), "+b" (bx), "+c" (cx), "+D" (di) - : : "esi", "edx"); + initregs(&ireg); + ireg.ax = 0x4f15; /* VBE DDC */ + /* ireg.bx = 0x0000; */ /* Report DDC capabilities */ + /* ireg.cx = 0; */ /* Controller 0 */ + ireg.es = 0; /* ES:DI must be 0 by spec */ + intcall(0x10, &ireg, &oreg); - if (ax != 0x004f) + if (oreg.ax != 0x004f) return; /* No EDID */ /* BH = time in seconds to transfer EDD information */ /* BL = DDC level supported */ - ax = 0x4f15; /* VBE DDC */ - bx = 0x0001; /* Read EDID */ - cx = 0; /* Controller 0 */ - dx = 0; /* EDID block number */ - di =(size_t) &boot_params.edid_info; /* (ES:)Pointer to block */ - asm(INT10 - : "+a" (ax), "+b" (bx), "+d" (dx), "=m" (boot_params.edid_info), - "+c" (cx), "+D" (di) - : : "esi"); + ireg.ax = 0x4f15; /* VBE DDC */ + ireg.bx = 0x0001; /* Read EDID */ + /* ireg.cx = 0; */ /* Controller 0 */ + /* ireg.dx = 0; */ /* EDID block number */ + ireg.es = ds(); + ireg.di =(size_t)&boot_params.edid_info; /* (ES:)Pointer to block */ + intcall(0x10, &ireg, &oreg); #endif /* CONFIG_FIRMWARE_EDID */ } diff --git a/arch/x86/boot/video-vga.c b/arch/x86/boot/video-vga.c index 9e0587a3776..8f8d827e254 100644 --- a/arch/x86/boot/video-vga.c +++ b/arch/x86/boot/video-vga.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -39,30 +40,30 @@ static __videocard video_vga; /* Set basic 80x25 mode */ static u8 vga_set_basic_mode(void) { + struct biosregs ireg, oreg; u16 ax; u8 rows; u8 mode; + initregs(&ireg); + #ifdef CONFIG_VIDEO_400_HACK if (adapter >= ADAPTER_VGA) { - asm volatile(INT10 - : : "a" (0x1202), "b" (0x0030) - : "ecx", "edx", "esi", "edi"); + ireg.ax = 0x1202; + ireg.bx = 0x0030; + intcall(0x10, &ireg, NULL); } #endif ax = 0x0f00; - asm volatile(INT10 - : "+a" (ax) - : : "ebx", "ecx", "edx", "esi", "edi"); - - mode = (u8)ax; + intcall(0x10, &ireg, &oreg); + mode = oreg.al; set_fs(0); rows = rdfs8(0x484); /* rows minus one */ #ifndef CONFIG_VIDEO_400_HACK - if ((ax == 0x5003 || ax == 0x5007) && + if ((oreg.ax == 0x5003 || oreg.ax == 0x5007) && (rows == 0 || rows == 24)) return mode; #endif @@ -71,10 +72,8 @@ static u8 vga_set_basic_mode(void) mode = 3; /* Set the mode */ - ax = mode; - asm volatile(INT10 - : "+a" (ax) - : : "ebx", "ecx", "edx", "esi", "edi"); + ireg.ax = mode; /* AH=0: set mode */ + intcall(0x10, &ireg, NULL); do_restore = 1; return mode; } @@ -82,43 +81,69 @@ static u8 vga_set_basic_mode(void) static void vga_set_8font(void) { /* Set 8x8 font - 80x43 on EGA, 80x50 on VGA */ + struct biosregs ireg; + + initregs(&ireg); /* Set 8x8 font */ - asm volatile(INT10 : : "a" (0x1112), "b" (0)); + ireg.ax = 0x1112; + /* ireg.bl = 0; */ + intcall(0x10, &ireg, NULL); /* Use alternate print screen */ - asm volatile(INT10 : : "a" (0x1200), "b" (0x20)); + ireg.ax = 0x1200; + ireg.bl = 0x20; + intcall(0x10, &ireg, NULL); /* Turn off cursor emulation */ - asm volatile(INT10 : : "a" (0x1201), "b" (0x34)); + ireg.ax = 0x1201; + ireg.bl = 0x34; + intcall(0x10, &ireg, NULL); /* Cursor is scan lines 6-7 */ - asm volatile(INT10 : : "a" (0x0100), "c" (0x0607)); + ireg.ax = 0x0100; + ireg.cx = 0x0607; + intcall(0x10, &ireg, NULL); } static void vga_set_14font(void) { /* Set 9x14 font - 80x28 on VGA */ + struct biosregs ireg; + + initregs(&ireg); /* Set 9x14 font */ - asm volatile(INT10 : : "a" (0x1111), "b" (0)); + ireg.ax = 0x1111; + /* ireg.bl = 0; */ + intcall(0x10, &ireg, NULL); /* Turn off cursor emulation */ - asm volatile(INT10 : : "a" (0x1201), "b" (0x34)); + ireg.ax = 0x1201; + ireg.bl = 0x34; + intcall(0x10, &ireg, NULL); /* Cursor is scan lines 11-12 */ - asm volatile(INT10 : : "a" (0x0100), "c" (0x0b0c)); + ireg.ax = 0x0100; + ireg.cx = 0x0b0c; + intcall(0x10, &ireg, NULL); } static void vga_set_80x43(void) { /* Set 80x43 mode on VGA (not EGA) */ + struct biosregs ireg; + + initregs(&ireg); /* Set 350 scans */ - asm volatile(INT10 : : "a" (0x1201), "b" (0x30)); + ireg.ax = 0x1201; + ireg.bl = 0x30; + intcall(0x10, &ireg, NULL); /* Reset video mode */ - asm volatile(INT10 : : "a" (0x0003)); + ireg.ax = 0x0003; + intcall(0x10, &ireg, NULL); vga_set_8font(); } @@ -225,8 +250,6 @@ static int vga_set_mode(struct mode_info *mode) */ static int vga_probe(void) { - u16 ega_bx; - static const char *card_name[] = { "CGA/MDA/HGC", "EGA", "VGA" }; @@ -240,26 +263,26 @@ static int vga_probe(void) sizeof(ega_modes)/sizeof(struct mode_info), sizeof(vga_modes)/sizeof(struct mode_info), }; - u8 vga_flag; - asm(INT10 - : "=b" (ega_bx) - : "a" (0x1200), "b" (0x10) /* Check EGA/VGA */ - : "ecx", "edx", "esi", "edi"); + struct biosregs ireg, oreg; + + initregs(&ireg); + + ireg.ax = 0x1200; + ireg.bl = 0x10; /* Check EGA/VGA */ + intcall(0x10, &ireg, &oreg); #ifndef _WAKEUP - boot_params.screen_info.orig_video_ega_bx = ega_bx; + boot_params.screen_info.orig_video_ega_bx = oreg.bx; #endif /* If we have MDA/CGA/HGC then BL will be unchanged at 0x10 */ - if ((u8)ega_bx != 0x10) { + if (oreg.bl != 0x10) { /* EGA/VGA */ - asm(INT10 - : "=a" (vga_flag) - : "a" (0x1a00) - : "ebx", "ecx", "edx", "esi", "edi"); + ireg.ax = 0x1a00; + intcall(0x10, &ireg, &oreg); - if (vga_flag == 0x1a) { + if (oreg.al == 0x1a) { adapter = ADAPTER_VGA; #ifndef _WAKEUP boot_params.screen_info.orig_video_isVGA = 1; diff --git a/arch/x86/boot/video.c b/arch/x86/boot/video.c index 3bef2c1febe..bad728b76fc 100644 --- a/arch/x86/boot/video.c +++ b/arch/x86/boot/video.c @@ -2,6 +2,7 @@ * * Copyright (C) 1991, 1992 Linus Torvalds * Copyright 2007 rPath, Inc. - All Rights Reserved + * Copyright 2009 Intel Corporation; author H. Peter Anvin * * This file is part of the Linux kernel, and is made available under * the terms of the GNU General Public License version 2. @@ -18,33 +19,29 @@ static void store_cursor_position(void) { - u16 curpos; - u16 ax, bx; + struct biosregs ireg, oreg; - ax = 0x0300; - bx = 0; - asm(INT10 - : "=d" (curpos), "+a" (ax), "+b" (bx) - : : "ecx", "esi", "edi"); + initregs(&ireg); + ireg.ah = 0x03; + intcall(0x10, &ireg, &oreg); - boot_params.screen_info.orig_x = curpos; - boot_params.screen_info.orig_y = curpos >> 8; + boot_params.screen_info.orig_x = oreg.dl; + boot_params.screen_info.orig_y = oreg.dh; } static void store_video_mode(void) { - u16 ax, page; + struct biosregs ireg, oreg; /* N.B.: the saving of the video page here is a bit silly, since we pretty much assume page 0 everywhere. */ - ax = 0x0f00; - asm(INT10 - : "+a" (ax), "=b" (page) - : : "ecx", "edx", "esi", "edi"); + initregs(&ireg); + ireg.ah = 0x0f; + intcall(0x10, &ireg, &oreg); /* Not all BIOSes are clean with respect to the top bit */ - boot_params.screen_info.orig_video_mode = ax & 0x7f; - boot_params.screen_info.orig_video_page = page >> 8; + boot_params.screen_info.orig_video_mode = oreg.al & 0x7f; + boot_params.screen_info.orig_video_page = oreg.bh; } /* @@ -257,7 +254,7 @@ static void restore_screen(void) int y; addr_t dst = 0; u16 *src = saved.data; - u16 ax, bx, dx; + struct biosregs ireg; if (graphic_mode) return; /* Can't restore onto a graphic mode */ @@ -296,12 +293,11 @@ static void restore_screen(void) } /* Restore cursor position */ - ax = 0x0200; /* Set cursor position */ - bx = 0; /* Page number (<< 8) */ - dx = (saved.cury << 8)+saved.curx; - asm volatile(INT10 - : "+a" (ax), "+b" (bx), "+d" (dx) - : : "ecx", "esi", "edi"); + initregs(&ireg); + ireg.ah = 0x02; /* Set cursor position */ + ireg.dh = saved.cury; + ireg.dl = saved.curx; + intcall(0x10, &ireg, NULL); } #else #define save_screen() ((void)0) diff --git a/arch/x86/boot/video.h b/arch/x86/boot/video.h index ee63f5d1446..5bb174a997f 100644 --- a/arch/x86/boot/video.h +++ b/arch/x86/boot/video.h @@ -112,20 +112,6 @@ extern int force_x, force_y; /* Don't query the BIOS for cols/rows */ extern int do_restore; /* Restore screen contents */ extern int graphic_mode; /* Graphics mode with linear frame buffer */ -/* - * int $0x10 is notorious for touching registers it shouldn't. - * gcc doesn't like %ebp being clobbered, so define it as a push/pop - * sequence here. - * - * A number of systems, including the original PC can clobber %bp in - * certain circumstances, like when scrolling. There exists at least - * one Trident video card which could clobber DS under a set of - * circumstances that we are unlikely to encounter (scrolling when - * using an extended graphics mode of more than 800x600 pixels), but - * it's cheap insurance to deal with that here. - */ -#define INT10 "pushl %%ebp; pushw %%ds; int $0x10; popw %%ds; popl %%ebp" - /* Accessing VGA indexed registers */ static inline u8 in_idx(u16 port, u8 index) { diff --git a/arch/x86/configs/i386_defconfig b/arch/x86/configs/i386_defconfig index 235b81d0f6f..edb992ebef9 100644 --- a/arch/x86/configs/i386_defconfig +++ b/arch/x86/configs/i386_defconfig @@ -1,12 +1,13 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.29-rc4 -# Tue Feb 24 15:50:58 2009 +# Linux kernel version: 2.6.30-rc2 +# Mon May 11 16:21:55 2009 # # CONFIG_64BIT is not set CONFIG_X86_32=y # CONFIG_X86_64 is not set CONFIG_X86=y +CONFIG_OUTPUT_FORMAT="elf32-i386" CONFIG_ARCH_DEFCONFIG="arch/x86/configs/i386_defconfig" CONFIG_GENERIC_TIME=y CONFIG_GENERIC_CMOS_UPDATE=y @@ -33,6 +34,7 @@ CONFIG_ARCH_HAS_CPU_RELAX=y CONFIG_ARCH_HAS_DEFAULT_IDLE=y CONFIG_ARCH_HAS_CACHE_LINE_SIZE=y CONFIG_HAVE_SETUP_PER_CPU_AREA=y +CONFIG_HAVE_DYNAMIC_PER_CPU_AREA=y # CONFIG_HAVE_CPUMASK_OF_CPU_MAP is not set CONFIG_ARCH_HIBERNATION_POSSIBLE=y CONFIG_ARCH_SUSPEND_POSSIBLE=y @@ -40,15 +42,16 @@ CONFIG_ARCH_SUSPEND_POSSIBLE=y CONFIG_ARCH_POPULATES_NODE_MAP=y # CONFIG_AUDIT_ARCH is not set CONFIG_ARCH_SUPPORTS_OPTIMIZED_INLINING=y +CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC=y CONFIG_GENERIC_HARDIRQS=y +CONFIG_GENERIC_HARDIRQS_NO__DO_IRQ=y CONFIG_GENERIC_IRQ_PROBE=y CONFIG_GENERIC_PENDING_IRQ=y -CONFIG_X86_SMP=y CONFIG_USE_GENERIC_SMP_HELPERS=y CONFIG_X86_32_SMP=y CONFIG_X86_HT=y -CONFIG_X86_BIOS_REBOOT=y CONFIG_X86_TRAMPOLINE=y +CONFIG_X86_32_LAZY_GS=y CONFIG_KTIME_SCALAR=y CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config" @@ -60,10 +63,17 @@ CONFIG_LOCK_KERNEL=y CONFIG_INIT_ENV_ARG_LIMIT=32 CONFIG_LOCALVERSION="" # CONFIG_LOCALVERSION_AUTO is not set +CONFIG_HAVE_KERNEL_GZIP=y +CONFIG_HAVE_KERNEL_BZIP2=y +CONFIG_HAVE_KERNEL_LZMA=y +CONFIG_KERNEL_GZIP=y +# CONFIG_KERNEL_BZIP2 is not set +# CONFIG_KERNEL_LZMA is not set CONFIG_SWAP=y CONFIG_SYSVIPC=y CONFIG_SYSVIPC_SYSCTL=y CONFIG_POSIX_MQUEUE=y +CONFIG_POSIX_MQUEUE_SYSCTL=y CONFIG_BSD_PROCESS_ACCT=y # CONFIG_BSD_PROCESS_ACCT_V3 is not set CONFIG_TASKSTATS=y @@ -113,23 +123,26 @@ CONFIG_PID_NS=y CONFIG_NET_NS=y CONFIG_BLK_DEV_INITRD=y CONFIG_INITRAMFS_SOURCE="" +CONFIG_RD_GZIP=y +CONFIG_RD_BZIP2=y +CONFIG_RD_LZMA=y CONFIG_CC_OPTIMIZE_FOR_SIZE=y CONFIG_SYSCTL=y +CONFIG_ANON_INODES=y # CONFIG_EMBEDDED is not set CONFIG_UID16=y CONFIG_SYSCTL_SYSCALL=y CONFIG_KALLSYMS=y CONFIG_KALLSYMS_ALL=y CONFIG_KALLSYMS_EXTRA_PASS=y +# CONFIG_STRIP_ASM_SYMS is not set CONFIG_HOTPLUG=y CONFIG_PRINTK=y CONFIG_BUG=y CONFIG_ELF_CORE=y CONFIG_PCSPKR_PLATFORM=y -# CONFIG_COMPAT_BRK is not set CONFIG_BASE_FULL=y CONFIG_FUTEX=y -CONFIG_ANON_INODES=y CONFIG_EPOLL=y CONFIG_SIGNALFD=y CONFIG_TIMERFD=y @@ -139,6 +152,7 @@ CONFIG_AIO=y CONFIG_VM_EVENT_COUNTERS=y CONFIG_PCI_QUIRKS=y CONFIG_SLUB_DEBUG=y +# CONFIG_COMPAT_BRK is not set # CONFIG_SLAB is not set CONFIG_SLUB=y # CONFIG_SLOB is not set @@ -154,6 +168,8 @@ CONFIG_HAVE_IOREMAP_PROT=y CONFIG_HAVE_KPROBES=y CONFIG_HAVE_KRETPROBES=y CONFIG_HAVE_ARCH_TRACEHOOK=y +CONFIG_HAVE_DMA_API_DEBUG=y +# CONFIG_SLOW_WORK is not set CONFIG_HAVE_GENERIC_DMA_COHERENT=y CONFIG_SLABINFO=y CONFIG_RT_MUTEXES=y @@ -167,7 +183,6 @@ CONFIG_MODULE_FORCE_UNLOAD=y CONFIG_STOP_MACHINE=y CONFIG_BLOCK=y # CONFIG_LBD is not set -CONFIG_BLK_DEV_IO_TRACE=y CONFIG_BLK_DEV_BSG=y # CONFIG_BLK_DEV_INTEGRITY is not set @@ -194,12 +209,12 @@ CONFIG_HIGH_RES_TIMERS=y CONFIG_GENERIC_CLOCKEVENTS_BUILD=y CONFIG_SMP=y CONFIG_SPARSE_IRQ=y -CONFIG_X86_FIND_SMP_CONFIG=y CONFIG_X86_MPPARSE=y +# CONFIG_X86_BIGSMP is not set +CONFIG_X86_EXTENDED_PLATFORM=y # CONFIG_X86_ELAN is not set -# CONFIG_X86_GENERICARCH is not set -# CONFIG_X86_VSMP is not set # CONFIG_X86_RDC321X is not set +# CONFIG_X86_32_NON_STANDARD is not set CONFIG_SCHED_OMIT_FRAME_POINTER=y # CONFIG_PARAVIRT_GUEST is not set # CONFIG_MEMTEST is not set @@ -230,8 +245,10 @@ CONFIG_M686=y # CONFIG_GENERIC_CPU is not set CONFIG_X86_GENERIC=y CONFIG_X86_CPU=y +CONFIG_X86_L1_CACHE_BYTES=64 +CONFIG_X86_INTERNODE_CACHE_BYTES=64 CONFIG_X86_CMPXCHG=y -CONFIG_X86_L1_CACHE_SHIFT=7 +CONFIG_X86_L1_CACHE_SHIFT=5 CONFIG_X86_XADD=y # CONFIG_X86_PPRO_FENCE is not set CONFIG_X86_WP_WORKS_OK=y @@ -247,7 +264,7 @@ CONFIG_X86_DEBUGCTLMSR=y CONFIG_CPU_SUP_INTEL=y CONFIG_CPU_SUP_CYRIX_32=y CONFIG_CPU_SUP_AMD=y -CONFIG_CPU_SUP_CENTAUR_32=y +CONFIG_CPU_SUP_CENTAUR=y CONFIG_CPU_SUP_TRANSMETA_32=y CONFIG_CPU_SUP_UMC_32=y CONFIG_X86_DS=y @@ -279,6 +296,7 @@ CONFIG_MICROCODE_AMD=y CONFIG_MICROCODE_OLD_INTERFACE=y CONFIG_X86_MSR=y CONFIG_X86_CPUID=y +# CONFIG_X86_CPU_DEBUG is not set # CONFIG_NOHIGHMEM is not set CONFIG_HIGHMEM4G=y # CONFIG_HIGHMEM64G is not set @@ -302,6 +320,8 @@ CONFIG_ZONE_DMA_FLAG=1 CONFIG_BOUNCE=y CONFIG_VIRT_TO_BUS=y CONFIG_UNEVICTABLE_LRU=y +CONFIG_HAVE_MLOCK=y +CONFIG_HAVE_MLOCKED_PAGE_BIT=y CONFIG_HIGHPTE=y CONFIG_X86_CHECK_BIOS_CORRUPTION=y CONFIG_X86_BOOTPARAM_MEMORY_CORRUPTION_CHECK=y @@ -312,6 +332,7 @@ CONFIG_MTRR=y CONFIG_X86_PAT=y CONFIG_EFI=y CONFIG_SECCOMP=y +# CONFIG_CC_STACKPROTECTOR is not set # CONFIG_HZ_100 is not set # CONFIG_HZ_250 is not set # CONFIG_HZ_300 is not set @@ -322,8 +343,9 @@ CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y # CONFIG_KEXEC_JUMP is not set CONFIG_PHYSICAL_START=0x1000000 -# CONFIG_RELOCATABLE is not set -CONFIG_PHYSICAL_ALIGN=0x200000 +CONFIG_RELOCATABLE=y +CONFIG_X86_NEED_RELOCS=y +CONFIG_PHYSICAL_ALIGN=0x1000000 CONFIG_HOTPLUG_CPU=y # CONFIG_COMPAT_VDSO is not set # CONFIG_CMDLINE_BOOL is not set @@ -363,7 +385,6 @@ CONFIG_ACPI_THERMAL=y CONFIG_ACPI_BLACKLIST_YEAR=0 # CONFIG_ACPI_DEBUG is not set # CONFIG_ACPI_PCI_SLOT is not set -CONFIG_ACPI_SYSTEM=y CONFIG_X86_PM_TIMER=y CONFIG_ACPI_CONTAINER=y # CONFIG_ACPI_SBS is not set @@ -425,6 +446,7 @@ CONFIG_PCI_BIOS=y CONFIG_PCI_DIRECT=y CONFIG_PCI_MMCONFIG=y CONFIG_PCI_DOMAINS=y +# CONFIG_DMAR is not set CONFIG_PCIEPORTBUS=y # CONFIG_HOTPLUG_PCI_PCIE is not set CONFIG_PCIEAER=y @@ -435,6 +457,7 @@ CONFIG_PCI_MSI=y # CONFIG_PCI_DEBUG is not set # CONFIG_PCI_STUB is not set CONFIG_HT_IRQ=y +# CONFIG_PCI_IOV is not set CONFIG_ISA_DMA_API=y # CONFIG_ISA is not set # CONFIG_MCA is not set @@ -481,7 +504,6 @@ CONFIG_NET=y # # Networking options # -CONFIG_COMPAT_NET_DEV_OPS=y CONFIG_PACKET=y CONFIG_PACKET_MMAP=y CONFIG_UNIX=y @@ -639,6 +661,7 @@ CONFIG_LLC=y # CONFIG_LAPB is not set # CONFIG_ECONET is not set # CONFIG_WAN_ROUTER is not set +# CONFIG_PHONET is not set CONFIG_NET_SCHED=y # @@ -696,6 +719,7 @@ CONFIG_NET_SCH_FIFO=y # # CONFIG_NET_PKTGEN is not set # CONFIG_NET_TCPPROBE is not set +# CONFIG_NET_DROP_MONITOR is not set CONFIG_HAMRADIO=y # @@ -706,12 +730,10 @@ CONFIG_HAMRADIO=y # CONFIG_IRDA is not set # CONFIG_BT is not set # CONFIG_AF_RXRPC is not set -# CONFIG_PHONET is not set CONFIG_FIB_RULES=y CONFIG_WIRELESS=y CONFIG_CFG80211=y # CONFIG_CFG80211_REG_DEBUG is not set -CONFIG_NL80211=y CONFIG_WIRELESS_OLD_REGULATORY=y CONFIG_WIRELESS_EXT=y CONFIG_WIRELESS_EXT_SYSFS=y @@ -789,6 +811,7 @@ CONFIG_MISC_DEVICES=y # CONFIG_ICS932S401 is not set # CONFIG_ENCLOSURE_SERVICES is not set # CONFIG_HP_ILO is not set +# CONFIG_ISL29003 is not set # CONFIG_C2PORT is not set # @@ -842,6 +865,7 @@ CONFIG_SCSI_SPI_ATTRS=y # CONFIG_SCSI_LOWLEVEL is not set # CONFIG_SCSI_LOWLEVEL_PCMCIA is not set # CONFIG_SCSI_DH is not set +# CONFIG_SCSI_OSD_INITIATOR is not set CONFIG_ATA=y # CONFIG_ATA_NONSTANDARD is not set CONFIG_ATA_ACPI=y @@ -940,6 +964,7 @@ CONFIG_DM_ZERO=y CONFIG_MACINTOSH_DRIVERS=y CONFIG_MAC_EMUMOUSEBTN=y CONFIG_NETDEVICES=y +CONFIG_COMPAT_NET_DEV_OPS=y # CONFIG_IFB is not set # CONFIG_DUMMY is not set # CONFIG_BONDING is not set @@ -977,6 +1002,8 @@ CONFIG_MII=y CONFIG_NET_VENDOR_3COM=y # CONFIG_VORTEX is not set # CONFIG_TYPHOON is not set +# CONFIG_ETHOC is not set +# CONFIG_DNET is not set CONFIG_NET_TULIP=y # CONFIG_DE2104X is not set # CONFIG_TULIP is not set @@ -1026,6 +1053,7 @@ CONFIG_E1000=y CONFIG_E1000E=y # CONFIG_IP1000 is not set # CONFIG_IGB is not set +# CONFIG_IGBVF is not set # CONFIG_NS83820 is not set # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set @@ -1040,6 +1068,7 @@ CONFIG_BNX2=y # CONFIG_QLA3XXX is not set # CONFIG_ATL1 is not set # CONFIG_ATL1E is not set +# CONFIG_ATL1C is not set # CONFIG_JME is not set CONFIG_NETDEV_10000=y # CONFIG_CHELSIO_T1 is not set @@ -1049,6 +1078,7 @@ CONFIG_CHELSIO_T3_DEPENDS=y # CONFIG_IXGBE is not set # CONFIG_IXGB is not set # CONFIG_S2IO is not set +# CONFIG_VXGE is not set # CONFIG_MYRI10GE is not set # CONFIG_NETXEN_NIC is not set # CONFIG_NIU is not set @@ -1058,6 +1088,7 @@ CONFIG_CHELSIO_T3_DEPENDS=y # CONFIG_BNX2X is not set # CONFIG_QLGE is not set # CONFIG_SFC is not set +# CONFIG_BE2NET is not set CONFIG_TR=y # CONFIG_IBMOL is not set # CONFIG_IBMLS is not set @@ -1073,8 +1104,8 @@ CONFIG_WLAN_80211=y # CONFIG_LIBERTAS is not set # CONFIG_LIBERTAS_THINFIRM is not set # CONFIG_AIRO is not set -# CONFIG_HERMES is not set # CONFIG_ATMEL is not set +# CONFIG_AT76C50X_USB is not set # CONFIG_AIRO_CS is not set # CONFIG_PCMCIA_WL3501 is not set # CONFIG_PRISM54 is not set @@ -1084,21 +1115,21 @@ CONFIG_WLAN_80211=y # CONFIG_RTL8187 is not set # CONFIG_ADM8211 is not set # CONFIG_MAC80211_HWSIM is not set +# CONFIG_MWL8K is not set # CONFIG_P54_COMMON is not set CONFIG_ATH5K=y # CONFIG_ATH5K_DEBUG is not set # CONFIG_ATH9K is not set +# CONFIG_AR9170_USB is not set # CONFIG_IPW2100 is not set # CONFIG_IPW2200 is not set -# CONFIG_IWLCORE is not set -# CONFIG_IWLWIFI_LEDS is not set -# CONFIG_IWLAGN is not set -# CONFIG_IWL3945 is not set +# CONFIG_IWLWIFI is not set # CONFIG_HOSTAP is not set # CONFIG_B43 is not set # CONFIG_B43LEGACY is not set # CONFIG_ZD1211RW is not set # CONFIG_RT2X00 is not set +# CONFIG_HERMES is not set # # Enable WiMAX (Networking options) to see the WiMAX drivers @@ -1209,6 +1240,8 @@ CONFIG_INPUT_TABLET=y # CONFIG_TABLET_USB_KBTAB is not set # CONFIG_TABLET_USB_WACOM is not set CONFIG_INPUT_TOUCHSCREEN=y +# CONFIG_TOUCHSCREEN_AD7879_I2C is not set +# CONFIG_TOUCHSCREEN_AD7879 is not set # CONFIG_TOUCHSCREEN_FUJITSU is not set # CONFIG_TOUCHSCREEN_GUNZE is not set # CONFIG_TOUCHSCREEN_ELO is not set @@ -1303,6 +1336,7 @@ CONFIG_UNIX98_PTYS=y # CONFIG_LEGACY_PTYS is not set # CONFIG_IPMI_HANDLER is not set CONFIG_HW_RANDOM=y +# CONFIG_HW_RANDOM_TIMERIOMEM is not set CONFIG_HW_RANDOM_INTEL=y CONFIG_HW_RANDOM_AMD=y CONFIG_HW_RANDOM_GEODE=y @@ -1390,7 +1424,6 @@ CONFIG_I2C_I801=y # CONFIG_SENSORS_PCF8574 is not set # CONFIG_PCF8575 is not set # CONFIG_SENSORS_PCA9539 is not set -# CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_MAX6875 is not set # CONFIG_SENSORS_TSL2550 is not set # CONFIG_I2C_DEBUG_CORE is not set @@ -1424,6 +1457,7 @@ CONFIG_HWMON=y # CONFIG_SENSORS_ADT7475 is not set # CONFIG_SENSORS_K8TEMP is not set # CONFIG_SENSORS_ASB100 is not set +# CONFIG_SENSORS_ATK0110 is not set # CONFIG_SENSORS_ATXP1 is not set # CONFIG_SENSORS_DS1621 is not set # CONFIG_SENSORS_I5K_AMB is not set @@ -1433,6 +1467,7 @@ CONFIG_HWMON=y # CONFIG_SENSORS_FSCHER is not set # CONFIG_SENSORS_FSCPOS is not set # CONFIG_SENSORS_FSCHMD is not set +# CONFIG_SENSORS_G760A is not set # CONFIG_SENSORS_GL518SM is not set # CONFIG_SENSORS_GL520SM is not set # CONFIG_SENSORS_CORETEMP is not set @@ -1448,11 +1483,14 @@ CONFIG_HWMON=y # CONFIG_SENSORS_LM90 is not set # CONFIG_SENSORS_LM92 is not set # CONFIG_SENSORS_LM93 is not set +# CONFIG_SENSORS_LTC4215 is not set # CONFIG_SENSORS_LTC4245 is not set +# CONFIG_SENSORS_LM95241 is not set # CONFIG_SENSORS_MAX1619 is not set # CONFIG_SENSORS_MAX6650 is not set # CONFIG_SENSORS_PC87360 is not set # CONFIG_SENSORS_PC87427 is not set +# CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_SIS5595 is not set # CONFIG_SENSORS_DME1737 is not set # CONFIG_SENSORS_SMSC47M1 is not set @@ -1643,7 +1681,6 @@ CONFIG_FB_EFI=y # CONFIG_FB_3DFX is not set # CONFIG_FB_VOODOO1 is not set # CONFIG_FB_VT8623 is not set -# CONFIG_FB_CYBLA is not set # CONFIG_FB_TRIDENT is not set # CONFIG_FB_ARK is not set # CONFIG_FB_PM3 is not set @@ -1652,6 +1689,7 @@ CONFIG_FB_EFI=y # CONFIG_FB_VIRTUAL is not set # CONFIG_FB_METRONOME is not set # CONFIG_FB_MB862XX is not set +# CONFIG_FB_BROADSHEET is not set CONFIG_BACKLIGHT_LCD_SUPPORT=y # CONFIG_LCD_CLASS_DEVICE is not set CONFIG_BACKLIGHT_CLASS_DEVICE=y @@ -1738,6 +1776,8 @@ CONFIG_SND_PCI=y # CONFIG_SND_INDIGO is not set # CONFIG_SND_INDIGOIO is not set # CONFIG_SND_INDIGODJ is not set +# CONFIG_SND_INDIGOIOX is not set +# CONFIG_SND_INDIGODJX is not set # CONFIG_SND_EMU10K1 is not set # CONFIG_SND_EMU10K1X is not set # CONFIG_SND_ENS1370 is not set @@ -1811,15 +1851,17 @@ CONFIG_USB_HIDDEV=y # # Special HID drivers # -CONFIG_HID_COMPAT=y CONFIG_HID_A4TECH=y CONFIG_HID_APPLE=y CONFIG_HID_BELKIN=y CONFIG_HID_CHERRY=y CONFIG_HID_CHICONY=y CONFIG_HID_CYPRESS=y +# CONFIG_DRAGONRISE_FF is not set CONFIG_HID_EZKEY=y +CONFIG_HID_KYE=y CONFIG_HID_GYRATION=y +CONFIG_HID_KENSINGTON=y CONFIG_HID_LOGITECH=y CONFIG_LOGITECH_FF=y # CONFIG_LOGIRUMBLEPAD2_FF is not set @@ -1885,11 +1927,11 @@ CONFIG_USB_PRINTER=y # CONFIG_USB_TMC is not set # -# NOTE: USB_STORAGE depends on SCSI but BLK_DEV_SD may also be needed; +# NOTE: USB_STORAGE depends on SCSI but BLK_DEV_SD may # # -# see USB_STORAGE Help for more information +# also be needed; see USB_STORAGE Help for more info # CONFIG_USB_STORAGE=y # CONFIG_USB_STORAGE_DEBUG is not set @@ -1931,7 +1973,6 @@ CONFIG_USB_LIBUSUAL=y # CONFIG_USB_LED is not set # CONFIG_USB_CYPRESS_CY7C63 is not set # CONFIG_USB_CYTHERM is not set -# CONFIG_USB_PHIDGET is not set # CONFIG_USB_IDMOUSE is not set # CONFIG_USB_FTDI_ELAN is not set # CONFIG_USB_APPLEDISPLAY is not set @@ -1947,6 +1988,7 @@ CONFIG_USB_LIBUSUAL=y # # OTG and related infrastructure # +# CONFIG_NOP_USB_XCEIV is not set # CONFIG_UWB is not set # CONFIG_MMC is not set # CONFIG_MEMSTICK is not set @@ -1958,8 +2000,10 @@ CONFIG_LEDS_CLASS=y # # CONFIG_LEDS_ALIX2 is not set # CONFIG_LEDS_PCA9532 is not set +# CONFIG_LEDS_LP5521 is not set # CONFIG_LEDS_CLEVO_MAIL is not set # CONFIG_LEDS_PCA955X is not set +# CONFIG_LEDS_BD2802 is not set # # LED Triggers @@ -1969,6 +2013,10 @@ CONFIG_LEDS_TRIGGERS=y # CONFIG_LEDS_TRIGGER_HEARTBEAT is not set # CONFIG_LEDS_TRIGGER_BACKLIGHT is not set # CONFIG_LEDS_TRIGGER_DEFAULT_ON is not set + +# +# iptables trigger is under Netfilter config (LED target) +# # CONFIG_ACCESSIBILITY is not set # CONFIG_INFINIBAND is not set CONFIG_EDAC=y @@ -2037,6 +2085,7 @@ CONFIG_DMADEVICES=y # DMA Devices # # CONFIG_INTEL_IOATDMA is not set +# CONFIG_AUXDISPLAY is not set # CONFIG_UIO is not set # CONFIG_STAGING is not set CONFIG_X86_PLATFORM_DEVICES=y @@ -2071,6 +2120,7 @@ CONFIG_DMIID=y # # CONFIG_EXT2_FS is not set CONFIG_EXT3_FS=y +# CONFIG_EXT3_DEFAULTS_TO_ORDERED is not set CONFIG_EXT3_FS_XATTR=y CONFIG_EXT3_FS_POSIX_ACL=y CONFIG_EXT3_FS_SECURITY=y @@ -2101,6 +2151,11 @@ CONFIG_AUTOFS4_FS=y CONFIG_GENERIC_ACL=y # +# Caches +# +# CONFIG_FSCACHE is not set + +# # CD-ROM/DVD Filesystems # CONFIG_ISO9660_FS=y @@ -2151,6 +2206,7 @@ CONFIG_MISC_FILESYSTEMS=y # CONFIG_ROMFS_FS is not set # CONFIG_SYSV_FS is not set # CONFIG_UFS_FS is not set +# CONFIG_NILFS2_FS is not set CONFIG_NETWORK_FILESYSTEMS=y CONFIG_NFS_FS=y CONFIG_NFS_V3=y @@ -2164,7 +2220,6 @@ CONFIG_NFS_ACL_SUPPORT=y CONFIG_NFS_COMMON=y CONFIG_SUNRPC=y CONFIG_SUNRPC_GSS=y -# CONFIG_SUNRPC_REGISTER_V4 is not set CONFIG_RPCSEC_GSS_KRB5=y # CONFIG_RPCSEC_GSS_SPKM3 is not set # CONFIG_SMB_FS is not set @@ -2251,6 +2306,7 @@ CONFIG_DEBUG_FS=y CONFIG_DEBUG_KERNEL=y # CONFIG_DEBUG_SHIRQ is not set # CONFIG_DETECT_SOFTLOCKUP is not set +# CONFIG_DETECT_HUNG_TASK is not set # CONFIG_SCHED_DEBUG is not set CONFIG_SCHEDSTATS=y CONFIG_TIMER_STATS=y @@ -2266,6 +2322,7 @@ CONFIG_TIMER_STATS=y # CONFIG_LOCK_STAT is not set # CONFIG_DEBUG_SPINLOCK_SLEEP is not set # CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set +CONFIG_STACKTRACE=y # CONFIG_DEBUG_KOBJECT is not set # CONFIG_DEBUG_HIGHMEM is not set CONFIG_DEBUG_BUGVERBOSE=y @@ -2289,13 +2346,19 @@ CONFIG_FRAME_POINTER=y # CONFIG_FAULT_INJECTION is not set # CONFIG_LATENCYTOP is not set CONFIG_SYSCTL_SYSCALL_CHECK=y +# CONFIG_DEBUG_PAGEALLOC is not set CONFIG_USER_STACKTRACE_SUPPORT=y +CONFIG_NOP_TRACER=y CONFIG_HAVE_FUNCTION_TRACER=y CONFIG_HAVE_FUNCTION_GRAPH_TRACER=y CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST=y CONFIG_HAVE_DYNAMIC_FTRACE=y CONFIG_HAVE_FTRACE_MCOUNT_RECORD=y CONFIG_HAVE_HW_BRANCH_TRACER=y +CONFIG_HAVE_FTRACE_SYSCALLS=y +CONFIG_RING_BUFFER=y +CONFIG_TRACING=y +CONFIG_TRACING_SUPPORT=y # # Tracers @@ -2305,13 +2368,21 @@ CONFIG_HAVE_HW_BRANCH_TRACER=y # CONFIG_SYSPROF_TRACER is not set # CONFIG_SCHED_TRACER is not set # CONFIG_CONTEXT_SWITCH_TRACER is not set +# CONFIG_EVENT_TRACER is not set +# CONFIG_FTRACE_SYSCALLS is not set # CONFIG_BOOT_TRACER is not set # CONFIG_TRACE_BRANCH_PROFILING is not set # CONFIG_POWER_TRACER is not set # CONFIG_STACK_TRACER is not set # CONFIG_HW_BRANCH_TRACER is not set +# CONFIG_KMEMTRACE is not set +# CONFIG_WORKQUEUE_TRACER is not set +CONFIG_BLK_DEV_IO_TRACE=y +# CONFIG_FTRACE_STARTUP_TEST is not set +# CONFIG_MMIOTRACE is not set CONFIG_PROVIDE_OHCI1394_DMA_INIT=y -# CONFIG_DYNAMIC_PRINTK_DEBUG is not set +# CONFIG_DYNAMIC_DEBUG is not set +# CONFIG_DMA_API_DEBUG is not set # CONFIG_SAMPLES is not set CONFIG_HAVE_ARCH_KGDB=y # CONFIG_KGDB is not set @@ -2321,7 +2392,6 @@ CONFIG_EARLY_PRINTK=y CONFIG_EARLY_PRINTK_DBGP=y CONFIG_DEBUG_STACKOVERFLOW=y CONFIG_DEBUG_STACK_USAGE=y -# CONFIG_DEBUG_PAGEALLOC is not set # CONFIG_DEBUG_PER_CPU_MAPS is not set # CONFIG_X86_PTDUMP is not set CONFIG_DEBUG_RODATA=y @@ -2329,7 +2399,7 @@ CONFIG_DEBUG_RODATA=y CONFIG_DEBUG_NX_TEST=m # CONFIG_4KSTACKS is not set CONFIG_DOUBLEFAULT=y -# CONFIG_MMIOTRACE is not set +CONFIG_HAVE_MMIOTRACE_SUPPORT=y CONFIG_IO_DELAY_TYPE_0X80=0 CONFIG_IO_DELAY_TYPE_0XED=1 CONFIG_IO_DELAY_TYPE_UDELAY=2 @@ -2365,6 +2435,8 @@ CONFIG_SECURITY_SELINUX_AVC_STATS=y CONFIG_SECURITY_SELINUX_CHECKREQPROT_VALUE=1 # CONFIG_SECURITY_SELINUX_POLICYDB_VERSION_MAX is not set # CONFIG_SECURITY_SMACK is not set +# CONFIG_SECURITY_TOMOYO is not set +# CONFIG_IMA is not set CONFIG_CRYPTO=y # @@ -2380,10 +2452,12 @@ CONFIG_CRYPTO_BLKCIPHER2=y CONFIG_CRYPTO_HASH=y CONFIG_CRYPTO_HASH2=y CONFIG_CRYPTO_RNG2=y +CONFIG_CRYPTO_PCOMP=y CONFIG_CRYPTO_MANAGER=y CONFIG_CRYPTO_MANAGER2=y # CONFIG_CRYPTO_GF128MUL is not set # CONFIG_CRYPTO_NULL is not set +CONFIG_CRYPTO_WORKQUEUE=y # CONFIG_CRYPTO_CRYPTD is not set CONFIG_CRYPTO_AUTHENC=y # CONFIG_CRYPTO_TEST is not set @@ -2456,6 +2530,7 @@ CONFIG_CRYPTO_DES=y # Compression # # CONFIG_CRYPTO_DEFLATE is not set +# CONFIG_CRYPTO_ZLIB is not set # CONFIG_CRYPTO_LZO is not set # @@ -2467,11 +2542,13 @@ CONFIG_CRYPTO_HW=y # CONFIG_CRYPTO_DEV_GEODE is not set # CONFIG_CRYPTO_DEV_HIFN_795X is not set CONFIG_HAVE_KVM=y +CONFIG_HAVE_KVM_IRQCHIP=y CONFIG_VIRTUALIZATION=y # CONFIG_KVM is not set # CONFIG_LGUEST is not set # CONFIG_VIRTIO_PCI is not set # CONFIG_VIRTIO_BALLOON is not set +CONFIG_BINARY_PRINTF=y # # Library routines @@ -2489,7 +2566,10 @@ CONFIG_CRC32=y # CONFIG_LIBCRC32C is not set CONFIG_AUDIT_GENERIC=y CONFIG_ZLIB_INFLATE=y -CONFIG_PLIST=y +CONFIG_DECOMPRESS_GZIP=y +CONFIG_DECOMPRESS_BZIP2=y +CONFIG_DECOMPRESS_LZMA=y CONFIG_HAS_IOMEM=y CONFIG_HAS_IOPORT=y CONFIG_HAS_DMA=y +CONFIG_NLATTR=y diff --git a/arch/x86/configs/x86_64_defconfig b/arch/x86/configs/x86_64_defconfig index 9fe5d212ab4..cee1dd2e69b 100644 --- a/arch/x86/configs/x86_64_defconfig +++ b/arch/x86/configs/x86_64_defconfig @@ -1,12 +1,13 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.29-rc4 -# Tue Feb 24 15:44:16 2009 +# Linux kernel version: 2.6.30-rc2 +# Mon May 11 16:22:00 2009 # CONFIG_64BIT=y # CONFIG_X86_32 is not set CONFIG_X86_64=y CONFIG_X86=y +CONFIG_OUTPUT_FORMAT="elf64-x86-64" CONFIG_ARCH_DEFCONFIG="arch/x86/configs/x86_64_defconfig" CONFIG_GENERIC_TIME=y CONFIG_GENERIC_CMOS_UPDATE=y @@ -34,6 +35,7 @@ CONFIG_ARCH_HAS_CPU_RELAX=y CONFIG_ARCH_HAS_DEFAULT_IDLE=y CONFIG_ARCH_HAS_CACHE_LINE_SIZE=y CONFIG_HAVE_SETUP_PER_CPU_AREA=y +CONFIG_HAVE_DYNAMIC_PER_CPU_AREA=y CONFIG_HAVE_CPUMASK_OF_CPU_MAP=y CONFIG_ARCH_HIBERNATION_POSSIBLE=y CONFIG_ARCH_SUSPEND_POSSIBLE=y @@ -41,14 +43,14 @@ CONFIG_ZONE_DMA32=y CONFIG_ARCH_POPULATES_NODE_MAP=y CONFIG_AUDIT_ARCH=y CONFIG_ARCH_SUPPORTS_OPTIMIZED_INLINING=y +CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC=y CONFIG_GENERIC_HARDIRQS=y +CONFIG_GENERIC_HARDIRQS_NO__DO_IRQ=y CONFIG_GENERIC_IRQ_PROBE=y CONFIG_GENERIC_PENDING_IRQ=y -CONFIG_X86_SMP=y CONFIG_USE_GENERIC_SMP_HELPERS=y CONFIG_X86_64_SMP=y CONFIG_X86_HT=y -CONFIG_X86_BIOS_REBOOT=y CONFIG_X86_TRAMPOLINE=y # CONFIG_KTIME_SCALAR is not set CONFIG_DEFCONFIG_LIST="/lib/modules/$UNAME_RELEASE/.config" @@ -61,10 +63,17 @@ CONFIG_LOCK_KERNEL=y CONFIG_INIT_ENV_ARG_LIMIT=32 CONFIG_LOCALVERSION="" # CONFIG_LOCALVERSION_AUTO is not set +CONFIG_HAVE_KERNEL_GZIP=y +CONFIG_HAVE_KERNEL_BZIP2=y +CONFIG_HAVE_KERNEL_LZMA=y +CONFIG_KERNEL_GZIP=y +# CONFIG_KERNEL_BZIP2 is not set +# CONFIG_KERNEL_LZMA is not set CONFIG_SWAP=y CONFIG_SYSVIPC=y CONFIG_SYSVIPC_SYSCTL=y CONFIG_POSIX_MQUEUE=y +CONFIG_POSIX_MQUEUE_SYSCTL=y CONFIG_BSD_PROCESS_ACCT=y # CONFIG_BSD_PROCESS_ACCT_V3 is not set CONFIG_TASKSTATS=y @@ -114,23 +123,26 @@ CONFIG_PID_NS=y CONFIG_NET_NS=y CONFIG_BLK_DEV_INITRD=y CONFIG_INITRAMFS_SOURCE="" +CONFIG_RD_GZIP=y +CONFIG_RD_BZIP2=y +CONFIG_RD_LZMA=y CONFIG_CC_OPTIMIZE_FOR_SIZE=y CONFIG_SYSCTL=y +CONFIG_ANON_INODES=y # CONFIG_EMBEDDED is not set CONFIG_UID16=y CONFIG_SYSCTL_SYSCALL=y CONFIG_KALLSYMS=y CONFIG_KALLSYMS_ALL=y CONFIG_KALLSYMS_EXTRA_PASS=y +# CONFIG_STRIP_ASM_SYMS is not set CONFIG_HOTPLUG=y CONFIG_PRINTK=y CONFIG_BUG=y CONFIG_ELF_CORE=y CONFIG_PCSPKR_PLATFORM=y -# CONFIG_COMPAT_BRK is not set CONFIG_BASE_FULL=y CONFIG_FUTEX=y -CONFIG_ANON_INODES=y CONFIG_EPOLL=y CONFIG_SIGNALFD=y CONFIG_TIMERFD=y @@ -140,6 +152,7 @@ CONFIG_AIO=y CONFIG_VM_EVENT_COUNTERS=y CONFIG_PCI_QUIRKS=y CONFIG_SLUB_DEBUG=y +# CONFIG_COMPAT_BRK is not set # CONFIG_SLAB is not set CONFIG_SLUB=y # CONFIG_SLOB is not set @@ -155,6 +168,8 @@ CONFIG_HAVE_IOREMAP_PROT=y CONFIG_HAVE_KPROBES=y CONFIG_HAVE_KRETPROBES=y CONFIG_HAVE_ARCH_TRACEHOOK=y +CONFIG_HAVE_DMA_API_DEBUG=y +# CONFIG_SLOW_WORK is not set # CONFIG_HAVE_GENERIC_DMA_COHERENT is not set CONFIG_SLABINFO=y CONFIG_RT_MUTEXES=y @@ -167,7 +182,6 @@ CONFIG_MODULE_FORCE_UNLOAD=y # CONFIG_MODULE_SRCVERSION_ALL is not set CONFIG_STOP_MACHINE=y CONFIG_BLOCK=y -CONFIG_BLK_DEV_IO_TRACE=y CONFIG_BLK_DEV_BSG=y # CONFIG_BLK_DEV_INTEGRITY is not set CONFIG_BLOCK_COMPAT=y @@ -195,12 +209,10 @@ CONFIG_HIGH_RES_TIMERS=y CONFIG_GENERIC_CLOCKEVENTS_BUILD=y CONFIG_SMP=y CONFIG_SPARSE_IRQ=y -# CONFIG_NUMA_MIGRATE_IRQ_DESC is not set -CONFIG_X86_FIND_SMP_CONFIG=y CONFIG_X86_MPPARSE=y -# CONFIG_X86_ELAN is not set -# CONFIG_X86_GENERICARCH is not set +CONFIG_X86_EXTENDED_PLATFORM=y # CONFIG_X86_VSMP is not set +# CONFIG_X86_UV is not set CONFIG_SCHED_OMIT_FRAME_POINTER=y # CONFIG_PARAVIRT_GUEST is not set # CONFIG_MEMTEST is not set @@ -230,10 +242,10 @@ CONFIG_SCHED_OMIT_FRAME_POINTER=y # CONFIG_MCORE2 is not set CONFIG_GENERIC_CPU=y CONFIG_X86_CPU=y -CONFIG_X86_L1_CACHE_BYTES=128 -CONFIG_X86_INTERNODE_CACHE_BYTES=128 +CONFIG_X86_L1_CACHE_BYTES=64 +CONFIG_X86_INTERNODE_CACHE_BYTES=64 CONFIG_X86_CMPXCHG=y -CONFIG_X86_L1_CACHE_SHIFT=7 +CONFIG_X86_L1_CACHE_SHIFT=6 CONFIG_X86_WP_WORKS_OK=y CONFIG_X86_TSC=y CONFIG_X86_CMPXCHG64=y @@ -242,7 +254,7 @@ CONFIG_X86_MINIMUM_CPU_FAMILY=64 CONFIG_X86_DEBUGCTLMSR=y CONFIG_CPU_SUP_INTEL=y CONFIG_CPU_SUP_AMD=y -CONFIG_CPU_SUP_CENTAUR_64=y +CONFIG_CPU_SUP_CENTAUR=y CONFIG_X86_DS=y CONFIG_X86_PTRACE_BTS=y CONFIG_HPET_TIMER=y @@ -269,6 +281,7 @@ CONFIG_X86_REROUTE_FOR_BROKEN_BOOT_IRQS=y CONFIG_X86_MCE=y CONFIG_X86_MCE_INTEL=y CONFIG_X86_MCE_AMD=y +CONFIG_X86_MCE_THRESHOLD=y # CONFIG_I8K is not set CONFIG_MICROCODE=y CONFIG_MICROCODE_INTEL=y @@ -276,6 +289,7 @@ CONFIG_MICROCODE_AMD=y CONFIG_MICROCODE_OLD_INTERFACE=y CONFIG_X86_MSR=y CONFIG_X86_CPUID=y +# CONFIG_X86_CPU_DEBUG is not set CONFIG_ARCH_PHYS_ADDR_T_64BIT=y CONFIG_DIRECT_GBPAGES=y CONFIG_NUMA=y @@ -309,6 +323,8 @@ CONFIG_ZONE_DMA_FLAG=1 CONFIG_BOUNCE=y CONFIG_VIRT_TO_BUS=y CONFIG_UNEVICTABLE_LRU=y +CONFIG_HAVE_MLOCK=y +CONFIG_HAVE_MLOCKED_PAGE_BIT=y CONFIG_X86_CHECK_BIOS_CORRUPTION=y CONFIG_X86_BOOTPARAM_MEMORY_CORRUPTION_CHECK=y CONFIG_X86_RESERVE_LOW_64K=y @@ -317,6 +333,7 @@ CONFIG_MTRR=y CONFIG_X86_PAT=y CONFIG_EFI=y CONFIG_SECCOMP=y +# CONFIG_CC_STACKPROTECTOR is not set # CONFIG_HZ_100 is not set # CONFIG_HZ_250 is not set # CONFIG_HZ_300 is not set @@ -325,9 +342,10 @@ CONFIG_HZ=1000 CONFIG_SCHED_HRTICK=y CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y +# CONFIG_KEXEC_JUMP is not set CONFIG_PHYSICAL_START=0x1000000 -# CONFIG_RELOCATABLE is not set -CONFIG_PHYSICAL_ALIGN=0x200000 +CONFIG_RELOCATABLE=y +CONFIG_PHYSICAL_ALIGN=0x1000000 CONFIG_HOTPLUG_CPU=y # CONFIG_COMPAT_VDSO is not set # CONFIG_CMDLINE_BOOL is not set @@ -370,7 +388,6 @@ CONFIG_ACPI_NUMA=y CONFIG_ACPI_BLACKLIST_YEAR=0 # CONFIG_ACPI_DEBUG is not set # CONFIG_ACPI_PCI_SLOT is not set -CONFIG_ACPI_SYSTEM=y CONFIG_X86_PM_TIMER=y CONFIG_ACPI_CONTAINER=y # CONFIG_ACPI_SBS is not set @@ -436,6 +453,7 @@ CONFIG_PCI_MSI=y # CONFIG_PCI_DEBUG is not set # CONFIG_PCI_STUB is not set CONFIG_HT_IRQ=y +# CONFIG_PCI_IOV is not set CONFIG_ISA_DMA_API=y CONFIG_K8_NB=y CONFIG_PCCARD=y @@ -481,7 +499,6 @@ CONFIG_NET=y # # Networking options # -CONFIG_COMPAT_NET_DEV_OPS=y CONFIG_PACKET=y CONFIG_PACKET_MMAP=y CONFIG_UNIX=y @@ -639,6 +656,7 @@ CONFIG_LLC=y # CONFIG_LAPB is not set # CONFIG_ECONET is not set # CONFIG_WAN_ROUTER is not set +# CONFIG_PHONET is not set CONFIG_NET_SCHED=y # @@ -696,6 +714,7 @@ CONFIG_NET_SCH_FIFO=y # # CONFIG_NET_PKTGEN is not set # CONFIG_NET_TCPPROBE is not set +# CONFIG_NET_DROP_MONITOR is not set CONFIG_HAMRADIO=y # @@ -706,12 +725,10 @@ CONFIG_HAMRADIO=y # CONFIG_IRDA is not set # CONFIG_BT is not set # CONFIG_AF_RXRPC is not set -# CONFIG_PHONET is not set CONFIG_FIB_RULES=y CONFIG_WIRELESS=y CONFIG_CFG80211=y # CONFIG_CFG80211_REG_DEBUG is not set -CONFIG_NL80211=y CONFIG_WIRELESS_OLD_REGULATORY=y CONFIG_WIRELESS_EXT=y CONFIG_WIRELESS_EXT_SYSFS=y @@ -788,9 +805,8 @@ CONFIG_MISC_DEVICES=y # CONFIG_TIFM_CORE is not set # CONFIG_ICS932S401 is not set # CONFIG_ENCLOSURE_SERVICES is not set -# CONFIG_SGI_XP is not set # CONFIG_HP_ILO is not set -# CONFIG_SGI_GRU is not set +# CONFIG_ISL29003 is not set # CONFIG_C2PORT is not set # @@ -844,6 +860,7 @@ CONFIG_SCSI_SPI_ATTRS=y # CONFIG_SCSI_LOWLEVEL is not set # CONFIG_SCSI_LOWLEVEL_PCMCIA is not set # CONFIG_SCSI_DH is not set +# CONFIG_SCSI_OSD_INITIATOR is not set CONFIG_ATA=y # CONFIG_ATA_NONSTANDARD is not set CONFIG_ATA_ACPI=y @@ -940,6 +957,7 @@ CONFIG_DM_ZERO=y CONFIG_MACINTOSH_DRIVERS=y CONFIG_MAC_EMUMOUSEBTN=y CONFIG_NETDEVICES=y +CONFIG_COMPAT_NET_DEV_OPS=y # CONFIG_IFB is not set # CONFIG_DUMMY is not set # CONFIG_BONDING is not set @@ -977,6 +995,8 @@ CONFIG_MII=y CONFIG_NET_VENDOR_3COM=y # CONFIG_VORTEX is not set # CONFIG_TYPHOON is not set +# CONFIG_ETHOC is not set +# CONFIG_DNET is not set CONFIG_NET_TULIP=y # CONFIG_DE2104X is not set # CONFIG_TULIP is not set @@ -1026,6 +1046,7 @@ CONFIG_E1000=y # CONFIG_E1000E is not set # CONFIG_IP1000 is not set # CONFIG_IGB is not set +# CONFIG_IGBVF is not set # CONFIG_NS83820 is not set # CONFIG_HAMACHI is not set # CONFIG_YELLOWFIN is not set @@ -1040,6 +1061,7 @@ CONFIG_TIGON3=y # CONFIG_QLA3XXX is not set # CONFIG_ATL1 is not set # CONFIG_ATL1E is not set +# CONFIG_ATL1C is not set # CONFIG_JME is not set CONFIG_NETDEV_10000=y # CONFIG_CHELSIO_T1 is not set @@ -1049,6 +1071,7 @@ CONFIG_CHELSIO_T3_DEPENDS=y # CONFIG_IXGBE is not set # CONFIG_IXGB is not set # CONFIG_S2IO is not set +# CONFIG_VXGE is not set # CONFIG_MYRI10GE is not set # CONFIG_NETXEN_NIC is not set # CONFIG_NIU is not set @@ -1058,6 +1081,7 @@ CONFIG_CHELSIO_T3_DEPENDS=y # CONFIG_BNX2X is not set # CONFIG_QLGE is not set # CONFIG_SFC is not set +# CONFIG_BE2NET is not set CONFIG_TR=y # CONFIG_IBMOL is not set # CONFIG_3C359 is not set @@ -1072,8 +1096,8 @@ CONFIG_WLAN_80211=y # CONFIG_LIBERTAS is not set # CONFIG_LIBERTAS_THINFIRM is not set # CONFIG_AIRO is not set -# CONFIG_HERMES is not set # CONFIG_ATMEL is not set +# CONFIG_AT76C50X_USB is not set # CONFIG_AIRO_CS is not set # CONFIG_PCMCIA_WL3501 is not set # CONFIG_PRISM54 is not set @@ -1083,21 +1107,21 @@ CONFIG_WLAN_80211=y # CONFIG_RTL8187 is not set # CONFIG_ADM8211 is not set # CONFIG_MAC80211_HWSIM is not set +# CONFIG_MWL8K is not set # CONFIG_P54_COMMON is not set CONFIG_ATH5K=y # CONFIG_ATH5K_DEBUG is not set # CONFIG_ATH9K is not set +# CONFIG_AR9170_USB is not set # CONFIG_IPW2100 is not set # CONFIG_IPW2200 is not set -# CONFIG_IWLCORE is not set -# CONFIG_IWLWIFI_LEDS is not set -# CONFIG_IWLAGN is not set -# CONFIG_IWL3945 is not set +# CONFIG_IWLWIFI is not set # CONFIG_HOSTAP is not set # CONFIG_B43 is not set # CONFIG_B43LEGACY is not set # CONFIG_ZD1211RW is not set # CONFIG_RT2X00 is not set +# CONFIG_HERMES is not set # # Enable WiMAX (Networking options) to see the WiMAX drivers @@ -1208,6 +1232,8 @@ CONFIG_INPUT_TABLET=y # CONFIG_TABLET_USB_KBTAB is not set # CONFIG_TABLET_USB_WACOM is not set CONFIG_INPUT_TOUCHSCREEN=y +# CONFIG_TOUCHSCREEN_AD7879_I2C is not set +# CONFIG_TOUCHSCREEN_AD7879 is not set # CONFIG_TOUCHSCREEN_FUJITSU is not set # CONFIG_TOUCHSCREEN_GUNZE is not set # CONFIG_TOUCHSCREEN_ELO is not set @@ -1301,6 +1327,7 @@ CONFIG_UNIX98_PTYS=y # CONFIG_LEGACY_PTYS is not set # CONFIG_IPMI_HANDLER is not set CONFIG_HW_RANDOM=y +# CONFIG_HW_RANDOM_TIMERIOMEM is not set # CONFIG_HW_RANDOM_INTEL is not set # CONFIG_HW_RANDOM_AMD is not set CONFIG_NVRAM=y @@ -1382,7 +1409,6 @@ CONFIG_I2C_I801=y # CONFIG_SENSORS_PCF8574 is not set # CONFIG_PCF8575 is not set # CONFIG_SENSORS_PCA9539 is not set -# CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_MAX6875 is not set # CONFIG_SENSORS_TSL2550 is not set # CONFIG_I2C_DEBUG_CORE is not set @@ -1416,6 +1442,7 @@ CONFIG_HWMON=y # CONFIG_SENSORS_ADT7475 is not set # CONFIG_SENSORS_K8TEMP is not set # CONFIG_SENSORS_ASB100 is not set +# CONFIG_SENSORS_ATK0110 is not set # CONFIG_SENSORS_ATXP1 is not set # CONFIG_SENSORS_DS1621 is not set # CONFIG_SENSORS_I5K_AMB is not set @@ -1425,6 +1452,7 @@ CONFIG_HWMON=y # CONFIG_SENSORS_FSCHER is not set # CONFIG_SENSORS_FSCPOS is not set # CONFIG_SENSORS_FSCHMD is not set +# CONFIG_SENSORS_G760A is not set # CONFIG_SENSORS_GL518SM is not set # CONFIG_SENSORS_GL520SM is not set # CONFIG_SENSORS_CORETEMP is not set @@ -1440,11 +1468,14 @@ CONFIG_HWMON=y # CONFIG_SENSORS_LM90 is not set # CONFIG_SENSORS_LM92 is not set # CONFIG_SENSORS_LM93 is not set +# CONFIG_SENSORS_LTC4215 is not set # CONFIG_SENSORS_LTC4245 is not set +# CONFIG_SENSORS_LM95241 is not set # CONFIG_SENSORS_MAX1619 is not set # CONFIG_SENSORS_MAX6650 is not set # CONFIG_SENSORS_PC87360 is not set # CONFIG_SENSORS_PC87427 is not set +# CONFIG_SENSORS_PCF8591 is not set # CONFIG_SENSORS_SIS5595 is not set # CONFIG_SENSORS_DME1737 is not set # CONFIG_SENSORS_SMSC47M1 is not set @@ -1635,6 +1666,7 @@ CONFIG_FB_EFI=y # CONFIG_FB_VIRTUAL is not set # CONFIG_FB_METRONOME is not set # CONFIG_FB_MB862XX is not set +# CONFIG_FB_BROADSHEET is not set CONFIG_BACKLIGHT_LCD_SUPPORT=y # CONFIG_LCD_CLASS_DEVICE is not set CONFIG_BACKLIGHT_CLASS_DEVICE=y @@ -1720,6 +1752,8 @@ CONFIG_SND_PCI=y # CONFIG_SND_INDIGO is not set # CONFIG_SND_INDIGOIO is not set # CONFIG_SND_INDIGODJ is not set +# CONFIG_SND_INDIGOIOX is not set +# CONFIG_SND_INDIGODJX is not set # CONFIG_SND_EMU10K1 is not set # CONFIG_SND_EMU10K1X is not set # CONFIG_SND_ENS1370 is not set @@ -1792,15 +1826,17 @@ CONFIG_USB_HIDDEV=y # # Special HID drivers # -CONFIG_HID_COMPAT=y CONFIG_HID_A4TECH=y CONFIG_HID_APPLE=y CONFIG_HID_BELKIN=y CONFIG_HID_CHERRY=y CONFIG_HID_CHICONY=y CONFIG_HID_CYPRESS=y +# CONFIG_DRAGONRISE_FF is not set CONFIG_HID_EZKEY=y +CONFIG_HID_KYE=y CONFIG_HID_GYRATION=y +CONFIG_HID_KENSINGTON=y CONFIG_HID_LOGITECH=y CONFIG_LOGITECH_FF=y # CONFIG_LOGIRUMBLEPAD2_FF is not set @@ -1866,11 +1902,11 @@ CONFIG_USB_PRINTER=y # CONFIG_USB_TMC is not set # -# NOTE: USB_STORAGE depends on SCSI but BLK_DEV_SD may also be needed; +# NOTE: USB_STORAGE depends on SCSI but BLK_DEV_SD may # # -# see USB_STORAGE Help for more information +# also be needed; see USB_STORAGE Help for more info # CONFIG_USB_STORAGE=y # CONFIG_USB_STORAGE_DEBUG is not set @@ -1912,7 +1948,6 @@ CONFIG_USB_LIBUSUAL=y # CONFIG_USB_LED is not set # CONFIG_USB_CYPRESS_CY7C63 is not set # CONFIG_USB_CYTHERM is not set -# CONFIG_USB_PHIDGET is not set # CONFIG_USB_IDMOUSE is not set # CONFIG_USB_FTDI_ELAN is not set # CONFIG_USB_APPLEDISPLAY is not set @@ -1928,6 +1963,7 @@ CONFIG_USB_LIBUSUAL=y # # OTG and related infrastructure # +# CONFIG_NOP_USB_XCEIV is not set # CONFIG_UWB is not set # CONFIG_MMC is not set # CONFIG_MEMSTICK is not set @@ -1939,8 +1975,10 @@ CONFIG_LEDS_CLASS=y # # CONFIG_LEDS_ALIX2 is not set # CONFIG_LEDS_PCA9532 is not set +# CONFIG_LEDS_LP5521 is not set # CONFIG_LEDS_CLEVO_MAIL is not set # CONFIG_LEDS_PCA955X is not set +# CONFIG_LEDS_BD2802 is not set # # LED Triggers @@ -1950,6 +1988,10 @@ CONFIG_LEDS_TRIGGERS=y # CONFIG_LEDS_TRIGGER_HEARTBEAT is not set # CONFIG_LEDS_TRIGGER_BACKLIGHT is not set # CONFIG_LEDS_TRIGGER_DEFAULT_ON is not set + +# +# iptables trigger is under Netfilter config (LED target) +# # CONFIG_ACCESSIBILITY is not set # CONFIG_INFINIBAND is not set CONFIG_EDAC=y @@ -2018,6 +2060,7 @@ CONFIG_DMADEVICES=y # DMA Devices # # CONFIG_INTEL_IOATDMA is not set +# CONFIG_AUXDISPLAY is not set # CONFIG_UIO is not set # CONFIG_STAGING is not set CONFIG_X86_PLATFORM_DEVICES=y @@ -2051,6 +2094,7 @@ CONFIG_DMIID=y # # CONFIG_EXT2_FS is not set CONFIG_EXT3_FS=y +# CONFIG_EXT3_DEFAULTS_TO_ORDERED is not set CONFIG_EXT3_FS_XATTR=y CONFIG_EXT3_FS_POSIX_ACL=y CONFIG_EXT3_FS_SECURITY=y @@ -2082,6 +2126,11 @@ CONFIG_AUTOFS4_FS=y CONFIG_GENERIC_ACL=y # +# Caches +# +# CONFIG_FSCACHE is not set + +# # CD-ROM/DVD Filesystems # CONFIG_ISO9660_FS=y @@ -2132,6 +2181,7 @@ CONFIG_MISC_FILESYSTEMS=y # CONFIG_ROMFS_FS is not set # CONFIG_SYSV_FS is not set # CONFIG_UFS_FS is not set +# CONFIG_NILFS2_FS is not set CONFIG_NETWORK_FILESYSTEMS=y CONFIG_NFS_FS=y CONFIG_NFS_V3=y @@ -2145,7 +2195,6 @@ CONFIG_NFS_ACL_SUPPORT=y CONFIG_NFS_COMMON=y CONFIG_SUNRPC=y CONFIG_SUNRPC_GSS=y -# CONFIG_SUNRPC_REGISTER_V4 is not set CONFIG_RPCSEC_GSS_KRB5=y # CONFIG_RPCSEC_GSS_SPKM3 is not set # CONFIG_SMB_FS is not set @@ -2232,6 +2281,7 @@ CONFIG_DEBUG_FS=y CONFIG_DEBUG_KERNEL=y # CONFIG_DEBUG_SHIRQ is not set # CONFIG_DETECT_SOFTLOCKUP is not set +# CONFIG_DETECT_HUNG_TASK is not set # CONFIG_SCHED_DEBUG is not set CONFIG_SCHEDSTATS=y CONFIG_TIMER_STATS=y @@ -2247,6 +2297,7 @@ CONFIG_TIMER_STATS=y # CONFIG_LOCK_STAT is not set # CONFIG_DEBUG_SPINLOCK_SLEEP is not set # CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set +CONFIG_STACKTRACE=y # CONFIG_DEBUG_KOBJECT is not set CONFIG_DEBUG_BUGVERBOSE=y # CONFIG_DEBUG_INFO is not set @@ -2269,13 +2320,19 @@ CONFIG_FRAME_POINTER=y # CONFIG_FAULT_INJECTION is not set # CONFIG_LATENCYTOP is not set CONFIG_SYSCTL_SYSCALL_CHECK=y +# CONFIG_DEBUG_PAGEALLOC is not set CONFIG_USER_STACKTRACE_SUPPORT=y +CONFIG_NOP_TRACER=y CONFIG_HAVE_FUNCTION_TRACER=y CONFIG_HAVE_FUNCTION_GRAPH_TRACER=y CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST=y CONFIG_HAVE_DYNAMIC_FTRACE=y CONFIG_HAVE_FTRACE_MCOUNT_RECORD=y CONFIG_HAVE_HW_BRANCH_TRACER=y +CONFIG_HAVE_FTRACE_SYSCALLS=y +CONFIG_RING_BUFFER=y +CONFIG_TRACING=y +CONFIG_TRACING_SUPPORT=y # # Tracers @@ -2285,13 +2342,21 @@ CONFIG_HAVE_HW_BRANCH_TRACER=y # CONFIG_SYSPROF_TRACER is not set # CONFIG_SCHED_TRACER is not set # CONFIG_CONTEXT_SWITCH_TRACER is not set +# CONFIG_EVENT_TRACER is not set +# CONFIG_FTRACE_SYSCALLS is not set # CONFIG_BOOT_TRACER is not set # CONFIG_TRACE_BRANCH_PROFILING is not set # CONFIG_POWER_TRACER is not set # CONFIG_STACK_TRACER is not set # CONFIG_HW_BRANCH_TRACER is not set +# CONFIG_KMEMTRACE is not set +# CONFIG_WORKQUEUE_TRACER is not set +CONFIG_BLK_DEV_IO_TRACE=y +# CONFIG_FTRACE_STARTUP_TEST is not set +# CONFIG_MMIOTRACE is not set CONFIG_PROVIDE_OHCI1394_DMA_INIT=y -# CONFIG_DYNAMIC_PRINTK_DEBUG is not set +# CONFIG_DYNAMIC_DEBUG is not set +# CONFIG_DMA_API_DEBUG is not set # CONFIG_SAMPLES is not set CONFIG_HAVE_ARCH_KGDB=y # CONFIG_KGDB is not set @@ -2301,14 +2366,13 @@ CONFIG_EARLY_PRINTK=y CONFIG_EARLY_PRINTK_DBGP=y CONFIG_DEBUG_STACKOVERFLOW=y CONFIG_DEBUG_STACK_USAGE=y -# CONFIG_DEBUG_PAGEALLOC is not set # CONFIG_DEBUG_PER_CPU_MAPS is not set # CONFIG_X86_PTDUMP is not set CONFIG_DEBUG_RODATA=y # CONFIG_DEBUG_RODATA_TEST is not set CONFIG_DEBUG_NX_TEST=m # CONFIG_IOMMU_DEBUG is not set -# CONFIG_MMIOTRACE is not set +CONFIG_HAVE_MMIOTRACE_SUPPORT=y CONFIG_IO_DELAY_TYPE_0X80=0 CONFIG_IO_DELAY_TYPE_0XED=1 CONFIG_IO_DELAY_TYPE_UDELAY=2 @@ -2344,6 +2408,8 @@ CONFIG_SECURITY_SELINUX_AVC_STATS=y CONFIG_SECURITY_SELINUX_CHECKREQPROT_VALUE=1 # CONFIG_SECURITY_SELINUX_POLICYDB_VERSION_MAX is not set # CONFIG_SECURITY_SMACK is not set +# CONFIG_SECURITY_TOMOYO is not set +# CONFIG_IMA is not set CONFIG_CRYPTO=y # @@ -2359,10 +2425,12 @@ CONFIG_CRYPTO_BLKCIPHER2=y CONFIG_CRYPTO_HASH=y CONFIG_CRYPTO_HASH2=y CONFIG_CRYPTO_RNG2=y +CONFIG_CRYPTO_PCOMP=y CONFIG_CRYPTO_MANAGER=y CONFIG_CRYPTO_MANAGER2=y # CONFIG_CRYPTO_GF128MUL is not set # CONFIG_CRYPTO_NULL is not set +CONFIG_CRYPTO_WORKQUEUE=y # CONFIG_CRYPTO_CRYPTD is not set CONFIG_CRYPTO_AUTHENC=y # CONFIG_CRYPTO_TEST is not set @@ -2414,6 +2482,7 @@ CONFIG_CRYPTO_SHA1=y # CONFIG_CRYPTO_AES=y # CONFIG_CRYPTO_AES_X86_64 is not set +# CONFIG_CRYPTO_AES_NI_INTEL is not set # CONFIG_CRYPTO_ANUBIS is not set CONFIG_CRYPTO_ARC4=y # CONFIG_CRYPTO_BLOWFISH is not set @@ -2435,6 +2504,7 @@ CONFIG_CRYPTO_DES=y # Compression # # CONFIG_CRYPTO_DEFLATE is not set +# CONFIG_CRYPTO_ZLIB is not set # CONFIG_CRYPTO_LZO is not set # @@ -2444,10 +2514,12 @@ CONFIG_CRYPTO_DES=y CONFIG_CRYPTO_HW=y # CONFIG_CRYPTO_DEV_HIFN_795X is not set CONFIG_HAVE_KVM=y +CONFIG_HAVE_KVM_IRQCHIP=y CONFIG_VIRTUALIZATION=y # CONFIG_KVM is not set # CONFIG_VIRTIO_PCI is not set # CONFIG_VIRTIO_BALLOON is not set +CONFIG_BINARY_PRINTF=y # # Library routines @@ -2464,7 +2536,10 @@ CONFIG_CRC32=y # CONFIG_CRC7 is not set # CONFIG_LIBCRC32C is not set CONFIG_ZLIB_INFLATE=y -CONFIG_PLIST=y +CONFIG_DECOMPRESS_GZIP=y +CONFIG_DECOMPRESS_BZIP2=y +CONFIG_DECOMPRESS_LZMA=y CONFIG_HAS_IOMEM=y CONFIG_HAS_IOPORT=y CONFIG_HAS_DMA=y +CONFIG_NLATTR=y diff --git a/arch/x86/ia32/ia32entry.S b/arch/x86/ia32/ia32entry.S index a505202086e..e590261ba05 100644 --- a/arch/x86/ia32/ia32entry.S +++ b/arch/x86/ia32/ia32entry.S @@ -825,9 +825,11 @@ ia32_sys_call_table: .quad compat_sys_signalfd4 .quad sys_eventfd2 .quad sys_epoll_create1 - .quad sys_dup3 /* 330 */ + .quad sys_dup3 /* 330 */ .quad sys_pipe2 .quad sys_inotify_init1 .quad compat_sys_preadv .quad compat_sys_pwritev + .quad compat_sys_rt_tgsigqueueinfo /* 335 */ + .quad sys_perf_counter_open ia32_syscall_end: diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h index f6aa18eadf7..1a37bcdc860 100644 --- a/arch/x86/include/asm/alternative.h +++ b/arch/x86/include/asm/alternative.h @@ -3,6 +3,7 @@ #include <linux/types.h> #include <linux/stddef.h> +#include <linux/stringify.h> #include <asm/asm.h> /* @@ -74,6 +75,22 @@ static inline void alternatives_smp_switch(int smp) {} const unsigned char *const *find_nop_table(void); +/* alternative assembly primitive: */ +#define ALTERNATIVE(oldinstr, newinstr, feature) \ + \ + "661:\n\t" oldinstr "\n662:\n" \ + ".section .altinstructions,\"a\"\n" \ + _ASM_ALIGN "\n" \ + _ASM_PTR "661b\n" /* label */ \ + _ASM_PTR "663f\n" /* new instruction */ \ + " .byte " __stringify(feature) "\n" /* feature bit */ \ + " .byte 662b-661b\n" /* sourcelen */ \ + " .byte 664f-663f\n" /* replacementlen */ \ + ".previous\n" \ + ".section .altinstr_replacement, \"ax\"\n" \ + "663:\n\t" newinstr "\n664:\n" /* replacement */ \ + ".previous" + /* * Alternative instructions for different CPU types or capabilities. * @@ -87,18 +104,7 @@ const unsigned char *const *find_nop_table(void); * without volatile and memory clobber. */ #define alternative(oldinstr, newinstr, feature) \ - asm volatile ("661:\n\t" oldinstr "\n662:\n" \ - ".section .altinstructions,\"a\"\n" \ - _ASM_ALIGN "\n" \ - _ASM_PTR "661b\n" /* label */ \ - _ASM_PTR "663f\n" /* new instruction */ \ - " .byte %c0\n" /* feature bit */ \ - " .byte 662b-661b\n" /* sourcelen */ \ - " .byte 664f-663f\n" /* replacementlen */ \ - ".previous\n" \ - ".section .altinstr_replacement,\"ax\"\n" \ - "663:\n\t" newinstr "\n664:\n" /* replacement */ \ - ".previous" :: "i" (feature) : "memory") + asm volatile (ALTERNATIVE(oldinstr, newinstr, feature) : : : "memory") /* * Alternative inline assembly with input. @@ -109,35 +115,16 @@ const unsigned char *const *find_nop_table(void); * Best is to use constraints that are fixed size (like (%1) ... "r") * If you use variable sized constraints like "m" or "g" in the * replacement make sure to pad to the worst case length. + * Leaving an unused argument 0 to keep API compatibility. */ #define alternative_input(oldinstr, newinstr, feature, input...) \ - asm volatile ("661:\n\t" oldinstr "\n662:\n" \ - ".section .altinstructions,\"a\"\n" \ - _ASM_ALIGN "\n" \ - _ASM_PTR "661b\n" /* label */ \ - _ASM_PTR "663f\n" /* new instruction */ \ - " .byte %c0\n" /* feature bit */ \ - " .byte 662b-661b\n" /* sourcelen */ \ - " .byte 664f-663f\n" /* replacementlen */ \ - ".previous\n" \ - ".section .altinstr_replacement,\"ax\"\n" \ - "663:\n\t" newinstr "\n664:\n" /* replacement */ \ - ".previous" :: "i" (feature), ##input) + asm volatile (ALTERNATIVE(oldinstr, newinstr, feature) \ + : : "i" (0), ## input) /* Like alternative_input, but with a single output argument */ #define alternative_io(oldinstr, newinstr, feature, output, input...) \ - asm volatile ("661:\n\t" oldinstr "\n662:\n" \ - ".section .altinstructions,\"a\"\n" \ - _ASM_ALIGN "\n" \ - _ASM_PTR "661b\n" /* label */ \ - _ASM_PTR "663f\n" /* new instruction */ \ - " .byte %c[feat]\n" /* feature bit */ \ - " .byte 662b-661b\n" /* sourcelen */ \ - " .byte 664f-663f\n" /* replacementlen */ \ - ".previous\n" \ - ".section .altinstr_replacement,\"ax\"\n" \ - "663:\n\t" newinstr "\n664:\n" /* replacement */ \ - ".previous" : output : [feat] "i" (feature), ##input) + asm volatile (ALTERNATIVE(oldinstr, newinstr, feature) \ + : output : "i" (0), ## input) /* * use this macro(s) if you need more than one output parameter diff --git a/arch/x86/include/asm/amd_iommu.h b/arch/x86/include/asm/amd_iommu.h index f712344329b..262e0282004 100644 --- a/arch/x86/include/asm/amd_iommu.h +++ b/arch/x86/include/asm/amd_iommu.h @@ -27,6 +27,8 @@ extern int amd_iommu_init(void); extern int amd_iommu_init_dma_ops(void); extern void amd_iommu_detect(void); extern irqreturn_t amd_iommu_int_handler(int irq, void *data); +extern void amd_iommu_flush_all_domains(void); +extern void amd_iommu_flush_all_devices(void); #else static inline int amd_iommu_init(void) { return -ENODEV; } static inline void amd_iommu_detect(void) { } diff --git a/arch/x86/include/asm/amd_iommu_types.h b/arch/x86/include/asm/amd_iommu_types.h index 95c8cd9d22b..0c878caaa0a 100644 --- a/arch/x86/include/asm/amd_iommu_types.h +++ b/arch/x86/include/asm/amd_iommu_types.h @@ -194,6 +194,27 @@ #define PD_DMA_OPS_MASK (1UL << 0) /* domain used for dma_ops */ #define PD_DEFAULT_MASK (1UL << 1) /* domain is a default dma_ops domain for an IOMMU */ +extern bool amd_iommu_dump; +#define DUMP_printk(format, arg...) \ + do { \ + if (amd_iommu_dump) \ + printk(KERN_INFO "AMD IOMMU: " format, ## arg); \ + } while(0); + +/* + * Make iterating over all IOMMUs easier + */ +#define for_each_iommu(iommu) \ + list_for_each_entry((iommu), &amd_iommu_list, list) +#define for_each_iommu_safe(iommu, next) \ + list_for_each_entry_safe((iommu), (next), &amd_iommu_list, list) + +#define APERTURE_RANGE_SHIFT 27 /* 128 MB */ +#define APERTURE_RANGE_SIZE (1ULL << APERTURE_RANGE_SHIFT) +#define APERTURE_RANGE_PAGES (APERTURE_RANGE_SIZE >> PAGE_SHIFT) +#define APERTURE_MAX_RANGES 32 /* allows 4GB of DMA address space */ +#define APERTURE_RANGE_INDEX(a) ((a) >> APERTURE_RANGE_SHIFT) +#define APERTURE_PAGE_INDEX(a) (((a) >> 21) & 0x3fULL) /* * This structure contains generic data for IOMMU protection domains @@ -210,6 +231,26 @@ struct protection_domain { }; /* + * For dynamic growth the aperture size is split into ranges of 128MB of + * DMA address space each. This struct represents one such range. + */ +struct aperture_range { + + /* address allocation bitmap */ + unsigned long *bitmap; + + /* + * Array of PTE pages for the aperture. In this array we save all the + * leaf pages of the domain page table used for the aperture. This way + * we don't need to walk the page table to find a specific PTE. We can + * just calculate its address in constant time. + */ + u64 *pte_pages[64]; + + unsigned long offset; +}; + +/* * Data container for a dma_ops specific protection domain */ struct dma_ops_domain { @@ -222,18 +263,10 @@ struct dma_ops_domain { unsigned long aperture_size; /* address we start to search for free addresses */ - unsigned long next_bit; - - /* address allocation bitmap */ - unsigned long *bitmap; + unsigned long next_address; - /* - * Array of PTE pages for the aperture. In this array we save all the - * leaf pages of the domain page table used for the aperture. This way - * we don't need to walk the page table to find a specific PTE. We can - * just calculate its address in constant time. - */ - u64 **pte_pages; + /* address space relevant data */ + struct aperture_range *aperture[APERTURE_MAX_RANGES]; /* This will be set to true when TLB needs to be flushed */ bool need_flush; diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index 42f2f837742..bb7d4792584 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -107,8 +107,7 @@ extern u32 native_safe_apic_wait_icr_idle(void); extern void native_apic_icr_write(u32 low, u32 id); extern u64 native_apic_icr_read(void); -#define EIM_8BIT_APIC_ID 0 -#define EIM_32BIT_APIC_ID 1 +extern int x2apic_mode; #ifdef CONFIG_X86_X2APIC /* @@ -166,10 +165,9 @@ static inline u64 native_x2apic_icr_read(void) return val; } -extern int x2apic, x2apic_phys; +extern int x2apic_phys; extern void check_x2apic(void); extern void enable_x2apic(void); -extern void enable_IR_x2apic(void); extern void x2apic_icr_write(u32 low, u32 id); static inline int x2apic_enabled(void) { @@ -183,6 +181,8 @@ static inline int x2apic_enabled(void) return 1; return 0; } + +#define x2apic_supported() (cpu_has_x2apic) #else static inline void check_x2apic(void) { @@ -190,28 +190,20 @@ static inline void check_x2apic(void) static inline void enable_x2apic(void) { } -static inline void enable_IR_x2apic(void) -{ -} static inline int x2apic_enabled(void) { return 0; } -#define x2apic 0 - +#define x2apic_preenabled 0 +#define x2apic_supported() 0 #endif -extern int get_physical_broadcast(void); +extern void enable_IR_x2apic(void); -#ifdef CONFIG_X86_X2APIC -static inline void ack_x2APIC_irq(void) -{ - /* Docs say use 0 for future compatibility */ - native_apic_msr_write(APIC_EOI, 0); -} -#endif +extern int get_physical_broadcast(void); +extern void apic_disable(void); extern int lapic_get_maxlvt(void); extern void clear_local_APIC(void); extern void connect_bsp_APIC(void); @@ -252,7 +244,7 @@ static inline void lapic_shutdown(void) { } #define local_apic_timer_c2_ok 1 static inline void init_apic_mappings(void) { } static inline void disable_local_APIC(void) { } - +static inline void apic_disable(void) { } #endif /* !CONFIG_X86_LOCAL_APIC */ #ifdef CONFIG_X86_64 @@ -410,7 +402,7 @@ static inline unsigned default_get_apic_id(unsigned long x) { unsigned int ver = GET_APIC_VERSION(apic_read(APIC_LVR)); - if (APIC_XAPIC(ver)) + if (APIC_XAPIC(ver) || boot_cpu_has(X86_FEATURE_EXTD_APICID)) return (x >> 24) & 0xFF; else return (x >> 24) & 0x0F; @@ -478,6 +470,9 @@ static inline unsigned int read_apic_id(void) extern void default_setup_apic_routing(void); #ifdef CONFIG_X86_32 + +extern struct apic apic_default; + /* * Set up the logical destination ID. * diff --git a/arch/x86/include/asm/apicdef.h b/arch/x86/include/asm/apicdef.h index bc9514fb3b1..7ddb36ab933 100644 --- a/arch/x86/include/asm/apicdef.h +++ b/arch/x86/include/asm/apicdef.h @@ -22,6 +22,7 @@ # define APIC_INTEGRATED(x) (1) #endif #define APIC_XAPIC(x) ((x) >= 0x14) +#define APIC_EXT_SPACE(x) ((x) & 0x80000000) #define APIC_TASKPRI 0x80 #define APIC_TPRI_MASK 0xFFu #define APIC_ARBPRI 0x90 @@ -116,7 +117,9 @@ #define APIC_TDR_DIV_32 0x8 #define APIC_TDR_DIV_64 0x9 #define APIC_TDR_DIV_128 0xA -#define APIC_EILVT0 0x500 +#define APIC_EFEAT 0x400 +#define APIC_ECTRL 0x410 +#define APIC_EILVTn(n) (0x500 + 0x10 * n) #define APIC_EILVT_NR_AMD_K8 1 /* # of extended interrupts */ #define APIC_EILVT_NR_AMD_10H 4 #define APIC_EILVT_LVTOFF(x) (((x) >> 4) & 0xF) @@ -125,9 +128,6 @@ #define APIC_EILVT_MSG_NMI 0x4 #define APIC_EILVT_MSG_EXT 0x7 #define APIC_EILVT_MASKED (1 << 16) -#define APIC_EILVT1 0x510 -#define APIC_EILVT2 0x520 -#define APIC_EILVT3 0x530 #define APIC_BASE (fix_to_virt(FIX_APIC_BASE)) #define APIC_BASE_MSR 0x800 diff --git a/arch/x86/include/asm/atomic_32.h b/arch/x86/include/asm/atomic_32.h index 85b46fba422..aff9f1fcdcd 100644 --- a/arch/x86/include/asm/atomic_32.h +++ b/arch/x86/include/asm/atomic_32.h @@ -247,5 +247,241 @@ static inline int atomic_add_unless(atomic_t *v, int a, int u) #define smp_mb__before_atomic_inc() barrier() #define smp_mb__after_atomic_inc() barrier() +/* An 64bit atomic type */ + +typedef struct { + unsigned long long counter; +} atomic64_t; + +#define ATOMIC64_INIT(val) { (val) } + +/** + * atomic64_read - read atomic64 variable + * @v: pointer of type atomic64_t + * + * Atomically reads the value of @v. + * Doesn't imply a read memory barrier. + */ +#define __atomic64_read(ptr) ((ptr)->counter) + +static inline unsigned long long +cmpxchg8b(unsigned long long *ptr, unsigned long long old, unsigned long long new) +{ + asm volatile( + + LOCK_PREFIX "cmpxchg8b (%[ptr])\n" + + : "=A" (old) + + : [ptr] "D" (ptr), + "A" (old), + "b" (ll_low(new)), + "c" (ll_high(new)) + + : "memory"); + + return old; +} + +static inline unsigned long long +atomic64_cmpxchg(atomic64_t *ptr, unsigned long long old_val, + unsigned long long new_val) +{ + return cmpxchg8b(&ptr->counter, old_val, new_val); +} + +/** + * atomic64_xchg - xchg atomic64 variable + * @ptr: pointer to type atomic64_t + * @new_val: value to assign + * @old_val: old value that was there + * + * Atomically xchgs the value of @ptr to @new_val and returns + * the old value. + */ + +static inline unsigned long long +atomic64_xchg(atomic64_t *ptr, unsigned long long new_val) +{ + unsigned long long old_val; + + do { + old_val = atomic_read(ptr); + } while (atomic64_cmpxchg(ptr, old_val, new_val) != old_val); + + return old_val; +} + +/** + * atomic64_set - set atomic64 variable + * @ptr: pointer to type atomic64_t + * @new_val: value to assign + * + * Atomically sets the value of @ptr to @new_val. + */ +static inline void atomic64_set(atomic64_t *ptr, unsigned long long new_val) +{ + atomic64_xchg(ptr, new_val); +} + +/** + * atomic64_read - read atomic64 variable + * @ptr: pointer to type atomic64_t + * + * Atomically reads the value of @ptr and returns it. + */ +static inline unsigned long long atomic64_read(atomic64_t *ptr) +{ + unsigned long long curr_val; + + do { + curr_val = __atomic64_read(ptr); + } while (atomic64_cmpxchg(ptr, curr_val, curr_val) != curr_val); + + return curr_val; +} + +/** + * atomic64_add_return - add and return + * @delta: integer value to add + * @ptr: pointer to type atomic64_t + * + * Atomically adds @delta to @ptr and returns @delta + *@ptr + */ +static inline unsigned long long +atomic64_add_return(unsigned long long delta, atomic64_t *ptr) +{ + unsigned long long old_val, new_val; + + do { + old_val = atomic_read(ptr); + new_val = old_val + delta; + + } while (atomic64_cmpxchg(ptr, old_val, new_val) != old_val); + + return new_val; +} + +static inline long atomic64_sub_return(unsigned long long delta, atomic64_t *ptr) +{ + return atomic64_add_return(-delta, ptr); +} + +static inline long atomic64_inc_return(atomic64_t *ptr) +{ + return atomic64_add_return(1, ptr); +} + +static inline long atomic64_dec_return(atomic64_t *ptr) +{ + return atomic64_sub_return(1, ptr); +} + +/** + * atomic64_add - add integer to atomic64 variable + * @delta: integer value to add + * @ptr: pointer to type atomic64_t + * + * Atomically adds @delta to @ptr. + */ +static inline void atomic64_add(unsigned long long delta, atomic64_t *ptr) +{ + atomic64_add_return(delta, ptr); +} + +/** + * atomic64_sub - subtract the atomic64 variable + * @delta: integer value to subtract + * @ptr: pointer to type atomic64_t + * + * Atomically subtracts @delta from @ptr. + */ +static inline void atomic64_sub(unsigned long long delta, atomic64_t *ptr) +{ + atomic64_add(-delta, ptr); +} + +/** + * atomic64_sub_and_test - subtract value from variable and test result + * @delta: integer value to subtract + * @ptr: pointer to type atomic64_t + * + * Atomically subtracts @delta from @ptr and returns + * true if the result is zero, or false for all + * other cases. + */ +static inline int +atomic64_sub_and_test(unsigned long long delta, atomic64_t *ptr) +{ + unsigned long long old_val = atomic64_sub_return(delta, ptr); + + return old_val == 0; +} + +/** + * atomic64_inc - increment atomic64 variable + * @ptr: pointer to type atomic64_t + * + * Atomically increments @ptr by 1. + */ +static inline void atomic64_inc(atomic64_t *ptr) +{ + atomic64_add(1, ptr); +} + +/** + * atomic64_dec - decrement atomic64 variable + * @ptr: pointer to type atomic64_t + * + * Atomically decrements @ptr by 1. + */ +static inline void atomic64_dec(atomic64_t *ptr) +{ + atomic64_sub(1, ptr); +} + +/** + * atomic64_dec_and_test - decrement and test + * @ptr: pointer to type atomic64_t + * + * Atomically decrements @ptr by 1 and + * returns true if the result is 0, or false for all other + * cases. + */ +static inline int atomic64_dec_and_test(atomic64_t *ptr) +{ + return atomic64_sub_and_test(1, ptr); +} + +/** + * atomic64_inc_and_test - increment and test + * @ptr: pointer to type atomic64_t + * + * Atomically increments @ptr by 1 + * and returns true if the result is zero, or false for all + * other cases. + */ +static inline int atomic64_inc_and_test(atomic64_t *ptr) +{ + return atomic64_sub_and_test(-1, ptr); +} + +/** + * atomic64_add_negative - add and test if negative + * @delta: integer value to add + * @ptr: pointer to type atomic64_t + * + * Atomically adds @delta to @ptr and returns true + * if the result is negative, or false when + * result is greater than or equal to zero. + */ +static inline int +atomic64_add_negative(unsigned long long delta, atomic64_t *ptr) +{ + long long old_val = atomic64_add_return(delta, ptr); + + return old_val < 0; +} + #include <asm-generic/atomic.h> #endif /* _ASM_X86_ATOMIC_32_H */ diff --git a/arch/x86/include/asm/boot.h b/arch/x86/include/asm/boot.h index 6ba23dd9fc9..418e632d4a8 100644 --- a/arch/x86/include/asm/boot.h +++ b/arch/x86/include/asm/boot.h @@ -8,11 +8,26 @@ #ifdef __KERNEL__ +#include <asm/page_types.h> + /* Physical address where kernel should be loaded. */ #define LOAD_PHYSICAL_ADDR ((CONFIG_PHYSICAL_START \ + (CONFIG_PHYSICAL_ALIGN - 1)) \ & ~(CONFIG_PHYSICAL_ALIGN - 1)) +/* Minimum kernel alignment, as a power of two */ +#ifdef CONFIG_x86_64 +#define MIN_KERNEL_ALIGN_LG2 PMD_SHIFT +#else +#define MIN_KERNEL_ALIGN_LG2 (PAGE_SHIFT+1) +#endif +#define MIN_KERNEL_ALIGN (_AC(1, UL) << MIN_KERNEL_ALIGN_LG2) + +#if (CONFIG_PHYSICAL_ALIGN & (CONFIG_PHYSICAL_ALIGN-1)) || \ + (CONFIG_PHYSICAL_ALIGN < (_AC(1, UL) << MIN_KERNEL_ALIGN_LG2)) +#error "Invalid value for CONFIG_PHYSICAL_ALIGN" +#endif + #ifdef CONFIG_KERNEL_BZIP2 #define BOOT_HEAP_SIZE 0x400000 #else /* !CONFIG_KERNEL_BZIP2 */ diff --git a/arch/x86/include/asm/bootparam.h b/arch/x86/include/asm/bootparam.h index 433adaebf9b..1724e8de317 100644 --- a/arch/x86/include/asm/bootparam.h +++ b/arch/x86/include/asm/bootparam.h @@ -50,7 +50,8 @@ struct setup_header { __u32 ramdisk_size; __u32 bootsect_kludge; __u16 heap_end_ptr; - __u16 _pad1; + __u8 ext_loader_ver; + __u8 ext_loader_type; __u32 cmd_line_ptr; __u32 initrd_addr_max; __u32 kernel_alignment; diff --git a/arch/x86/include/asm/cpu_debug.h b/arch/x86/include/asm/cpu_debug.h index 222802029fa..d96c1ee3a95 100644 --- a/arch/x86/include/asm/cpu_debug.h +++ b/arch/x86/include/asm/cpu_debug.h @@ -86,105 +86,7 @@ enum cpu_file_bit { CPU_VALUE_BIT, /* value */ }; -#define CPU_FILE_VALUE (1 << CPU_VALUE_BIT) - -/* - * DisplayFamily_DisplayModel Processor Families/Processor Number Series - * -------------------------- ------------------------------------------ - * 05_01, 05_02, 05_04 Pentium, Pentium with MMX - * - * 06_01 Pentium Pro - * 06_03, 06_05 Pentium II Xeon, Pentium II - * 06_07, 06_08, 06_0A, 06_0B Pentium III Xeon, Pentum III - * - * 06_09, 060D Pentium M - * - * 06_0E Core Duo, Core Solo - * - * 06_0F Xeon 3000, 3200, 5100, 5300, 7300 series, - * Core 2 Quad, Core 2 Extreme, Core 2 Duo, - * Pentium dual-core - * 06_17 Xeon 5200, 5400 series, Core 2 Quad Q9650 - * - * 06_1C Atom - * - * 0F_00, 0F_01, 0F_02 Xeon, Xeon MP, Pentium 4 - * 0F_03, 0F_04 Xeon, Xeon MP, Pentium 4, Pentium D - * - * 0F_06 Xeon 7100, 5000 Series, Xeon MP, - * Pentium 4, Pentium D - */ - -/* Register processors bits */ -enum cpu_processor_bit { - CPU_NONE, -/* Intel */ - CPU_INTEL_PENTIUM_BIT, - CPU_INTEL_P6_BIT, - CPU_INTEL_PENTIUM_M_BIT, - CPU_INTEL_CORE_BIT, - CPU_INTEL_CORE2_BIT, - CPU_INTEL_ATOM_BIT, - CPU_INTEL_XEON_P4_BIT, - CPU_INTEL_XEON_MP_BIT, -/* AMD */ - CPU_AMD_K6_BIT, - CPU_AMD_K7_BIT, - CPU_AMD_K8_BIT, - CPU_AMD_0F_BIT, - CPU_AMD_10_BIT, - CPU_AMD_11_BIT, -}; - -#define CPU_INTEL_PENTIUM (1 << CPU_INTEL_PENTIUM_BIT) -#define CPU_INTEL_P6 (1 << CPU_INTEL_P6_BIT) -#define CPU_INTEL_PENTIUM_M (1 << CPU_INTEL_PENTIUM_M_BIT) -#define CPU_INTEL_CORE (1 << CPU_INTEL_CORE_BIT) -#define CPU_INTEL_CORE2 (1 << CPU_INTEL_CORE2_BIT) -#define CPU_INTEL_ATOM (1 << CPU_INTEL_ATOM_BIT) -#define CPU_INTEL_XEON_P4 (1 << CPU_INTEL_XEON_P4_BIT) -#define CPU_INTEL_XEON_MP (1 << CPU_INTEL_XEON_MP_BIT) - -#define CPU_INTEL_PX (CPU_INTEL_P6 | CPU_INTEL_PENTIUM_M) -#define CPU_INTEL_COREX (CPU_INTEL_CORE | CPU_INTEL_CORE2) -#define CPU_INTEL_XEON (CPU_INTEL_XEON_P4 | CPU_INTEL_XEON_MP) -#define CPU_CO_AT (CPU_INTEL_CORE | CPU_INTEL_ATOM) -#define CPU_C2_AT (CPU_INTEL_CORE2 | CPU_INTEL_ATOM) -#define CPU_CX_AT (CPU_INTEL_COREX | CPU_INTEL_ATOM) -#define CPU_CX_XE (CPU_INTEL_COREX | CPU_INTEL_XEON) -#define CPU_P6_XE (CPU_INTEL_P6 | CPU_INTEL_XEON) -#define CPU_PM_CO_AT (CPU_INTEL_PENTIUM_M | CPU_CO_AT) -#define CPU_C2_AT_XE (CPU_C2_AT | CPU_INTEL_XEON) -#define CPU_CX_AT_XE (CPU_CX_AT | CPU_INTEL_XEON) -#define CPU_P6_CX_AT (CPU_INTEL_P6 | CPU_CX_AT) -#define CPU_P6_CX_XE (CPU_P6_XE | CPU_INTEL_COREX) -#define CPU_P6_CX_AT_XE (CPU_INTEL_P6 | CPU_CX_AT_XE) -#define CPU_PM_CX_AT_XE (CPU_INTEL_PENTIUM_M | CPU_CX_AT_XE) -#define CPU_PM_CX_AT (CPU_INTEL_PENTIUM_M | CPU_CX_AT) -#define CPU_PM_CX_XE (CPU_INTEL_PENTIUM_M | CPU_CX_XE) -#define CPU_PX_CX_AT (CPU_INTEL_PX | CPU_CX_AT) -#define CPU_PX_CX_AT_XE (CPU_INTEL_PX | CPU_CX_AT_XE) - -/* Select all supported Intel CPUs */ -#define CPU_INTEL_ALL (CPU_INTEL_PENTIUM | CPU_PX_CX_AT_XE) - -#define CPU_AMD_K6 (1 << CPU_AMD_K6_BIT) -#define CPU_AMD_K7 (1 << CPU_AMD_K7_BIT) -#define CPU_AMD_K8 (1 << CPU_AMD_K8_BIT) -#define CPU_AMD_0F (1 << CPU_AMD_0F_BIT) -#define CPU_AMD_10 (1 << CPU_AMD_10_BIT) -#define CPU_AMD_11 (1 << CPU_AMD_11_BIT) - -#define CPU_K10_PLUS (CPU_AMD_10 | CPU_AMD_11) -#define CPU_K0F_PLUS (CPU_AMD_0F | CPU_K10_PLUS) -#define CPU_K8_PLUS (CPU_AMD_K8 | CPU_K0F_PLUS) -#define CPU_K7_PLUS (CPU_AMD_K7 | CPU_K8_PLUS) - -/* Select all supported AMD CPUs */ -#define CPU_AMD_ALL (CPU_AMD_K6 | CPU_K7_PLUS) - -/* Select all supported CPUs */ -#define CPU_ALL (CPU_INTEL_ALL | CPU_AMD_ALL) +#define CPU_FILE_VALUE (1 << CPU_VALUE_BIT) #define MAX_CPU_FILES 512 @@ -220,7 +122,6 @@ struct cpu_debug_range { unsigned min; /* Register range min */ unsigned max; /* Register range max */ unsigned flag; /* Supported flags */ - unsigned model; /* Supported models */ }; #endif /* _ASM_X86_CPU_DEBUG_H */ diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h index bb83b1c397a..19af42138f7 100644 --- a/arch/x86/include/asm/cpufeature.h +++ b/arch/x86/include/asm/cpufeature.h @@ -22,7 +22,7 @@ #define X86_FEATURE_TSC (0*32+ 4) /* Time Stamp Counter */ #define X86_FEATURE_MSR (0*32+ 5) /* Model-Specific Registers */ #define X86_FEATURE_PAE (0*32+ 6) /* Physical Address Extensions */ -#define X86_FEATURE_MCE (0*32+ 7) /* Machine Check Architecture */ +#define X86_FEATURE_MCE (0*32+ 7) /* Machine Check Exception */ #define X86_FEATURE_CX8 (0*32+ 8) /* CMPXCHG8 instruction */ #define X86_FEATURE_APIC (0*32+ 9) /* Onboard APIC */ #define X86_FEATURE_SEP (0*32+11) /* SYSENTER/SYSEXIT */ @@ -94,6 +94,7 @@ #define X86_FEATURE_TSC_RELIABLE (3*32+23) /* TSC is known to be reliable */ #define X86_FEATURE_NONSTOP_TSC (3*32+24) /* TSC does not stop in C states */ #define X86_FEATURE_CLFLUSH_MONITOR (3*32+25) /* "" clflush reqd with monitor */ +#define X86_FEATURE_EXTD_APICID (3*32+26) /* has extended APICID (8 bits) */ /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */ #define X86_FEATURE_XMM3 (4*32+ 0) /* "pni" SSE-3 */ @@ -192,11 +193,11 @@ extern const char * const x86_power_flags[32]; #define clear_cpu_cap(c, bit) clear_bit(bit, (unsigned long *)((c)->x86_capability)) #define setup_clear_cpu_cap(bit) do { \ clear_cpu_cap(&boot_cpu_data, bit); \ - set_bit(bit, (unsigned long *)cleared_cpu_caps); \ + set_bit(bit, (unsigned long *)cpu_caps_cleared); \ } while (0) #define setup_force_cpu_cap(bit) do { \ set_cpu_cap(&boot_cpu_data, bit); \ - clear_bit(bit, (unsigned long *)cleared_cpu_caps); \ + set_bit(bit, (unsigned long *)cpu_caps_set); \ } while (0) #define cpu_has_fpu boot_cpu_has(X86_FEATURE_FPU) diff --git a/arch/x86/include/asm/ds.h b/arch/x86/include/asm/ds.h index a8f672ba100..70dac199b09 100644 --- a/arch/x86/include/asm/ds.h +++ b/arch/x86/include/asm/ds.h @@ -15,8 +15,8 @@ * - buffer allocation (memory accounting) * * - * Copyright (C) 2007-2008 Intel Corporation. - * Markus Metzger <markus.t.metzger@intel.com>, 2007-2008 + * Copyright (C) 2007-2009 Intel Corporation. + * Markus Metzger <markus.t.metzger@intel.com>, 2007-2009 */ #ifndef _ASM_X86_DS_H @@ -83,8 +83,10 @@ enum ds_feature { * The interrupt threshold is independent from the overflow callback * to allow users to use their own overflow interrupt handling mechanism. * - * task: the task to request recording for; - * NULL for per-cpu recording on the current cpu + * The function might sleep. + * + * task: the task to request recording for + * cpu: the cpu to request recording for * base: the base pointer for the (non-pageable) buffer; * size: the size of the provided buffer in bytes * ovfl: pointer to a function to be called on buffer overflow; @@ -93,19 +95,28 @@ enum ds_feature { * -1 if no interrupt threshold is requested. * flags: a bit-mask of the above flags */ -extern struct bts_tracer *ds_request_bts(struct task_struct *task, - void *base, size_t size, - bts_ovfl_callback_t ovfl, - size_t th, unsigned int flags); -extern struct pebs_tracer *ds_request_pebs(struct task_struct *task, - void *base, size_t size, - pebs_ovfl_callback_t ovfl, - size_t th, unsigned int flags); +extern struct bts_tracer *ds_request_bts_task(struct task_struct *task, + void *base, size_t size, + bts_ovfl_callback_t ovfl, + size_t th, unsigned int flags); +extern struct bts_tracer *ds_request_bts_cpu(int cpu, void *base, size_t size, + bts_ovfl_callback_t ovfl, + size_t th, unsigned int flags); +extern struct pebs_tracer *ds_request_pebs_task(struct task_struct *task, + void *base, size_t size, + pebs_ovfl_callback_t ovfl, + size_t th, unsigned int flags); +extern struct pebs_tracer *ds_request_pebs_cpu(int cpu, + void *base, size_t size, + pebs_ovfl_callback_t ovfl, + size_t th, unsigned int flags); /* * Release BTS or PEBS resources * Suspend and resume BTS or PEBS tracing * + * Must be called with irq's enabled. + * * tracer: the tracer handle returned from ds_request_~() */ extern void ds_release_bts(struct bts_tracer *tracer); @@ -115,6 +126,28 @@ extern void ds_release_pebs(struct pebs_tracer *tracer); extern void ds_suspend_pebs(struct pebs_tracer *tracer); extern void ds_resume_pebs(struct pebs_tracer *tracer); +/* + * Release BTS or PEBS resources + * Suspend and resume BTS or PEBS tracing + * + * Cpu tracers must call this on the traced cpu. + * Task tracers must call ds_release_~_noirq() for themselves. + * + * May be called with irq's disabled. + * + * Returns 0 if successful; + * -EPERM if the cpu tracer does not trace the current cpu. + * -EPERM if the task tracer does not trace itself. + * + * tracer: the tracer handle returned from ds_request_~() + */ +extern int ds_release_bts_noirq(struct bts_tracer *tracer); +extern int ds_suspend_bts_noirq(struct bts_tracer *tracer); +extern int ds_resume_bts_noirq(struct bts_tracer *tracer); +extern int ds_release_pebs_noirq(struct pebs_tracer *tracer); +extern int ds_suspend_pebs_noirq(struct pebs_tracer *tracer); +extern int ds_resume_pebs_noirq(struct pebs_tracer *tracer); + /* * The raw DS buffer state as it is used for BTS and PEBS recording. @@ -170,9 +203,9 @@ struct bts_struct { } lbr; /* BTS_TASK_ARRIVES or BTS_TASK_DEPARTS */ struct { - __u64 jiffies; + __u64 clock; pid_t pid; - } timestamp; + } event; } variant; }; @@ -201,8 +234,12 @@ struct bts_trace { struct pebs_trace { struct ds_trace ds; - /* the PEBS reset value */ - unsigned long long reset_value; + /* the number of valid counters in the below array */ + unsigned int counters; + +#define MAX_PEBS_COUNTERS 4 + /* the counter reset value */ + unsigned long long counter_reset[MAX_PEBS_COUNTERS]; }; @@ -237,9 +274,11 @@ extern int ds_reset_pebs(struct pebs_tracer *tracer); * Returns 0 on success; -Eerrno on error * * tracer: the tracer handle returned from ds_request_pebs() + * counter: the index of the counter * value: the new counter reset value */ -extern int ds_set_pebs_reset(struct pebs_tracer *tracer, u64 value); +extern int ds_set_pebs_reset(struct pebs_tracer *tracer, + unsigned int counter, u64 value); /* * Initialization @@ -252,21 +291,12 @@ extern void __cpuinit ds_init_intel(struct cpuinfo_x86 *); */ extern void ds_switch_to(struct task_struct *prev, struct task_struct *next); -/* - * Task clone/init and cleanup work - */ -extern void ds_copy_thread(struct task_struct *tsk, struct task_struct *father); -extern void ds_exit_thread(struct task_struct *tsk); - #else /* CONFIG_X86_DS */ struct cpuinfo_x86; static inline void __cpuinit ds_init_intel(struct cpuinfo_x86 *ignored) {} static inline void ds_switch_to(struct task_struct *prev, struct task_struct *next) {} -static inline void ds_copy_thread(struct task_struct *tsk, - struct task_struct *father) {} -static inline void ds_exit_thread(struct task_struct *tsk) {} #endif /* CONFIG_X86_DS */ #endif /* _ASM_X86_DS_H */ diff --git a/arch/x86/include/asm/entry_arch.h b/arch/x86/include/asm/entry_arch.h index c2e6bedaf25..d750a10ccad 100644 --- a/arch/x86/include/asm/entry_arch.h +++ b/arch/x86/include/asm/entry_arch.h @@ -49,7 +49,7 @@ BUILD_INTERRUPT(error_interrupt,ERROR_APIC_VECTOR) BUILD_INTERRUPT(spurious_interrupt,SPURIOUS_APIC_VECTOR) #ifdef CONFIG_PERF_COUNTERS -BUILD_INTERRUPT(perf_counter_interrupt, LOCAL_PERF_VECTOR) +BUILD_INTERRUPT(perf_pending_interrupt, LOCAL_PENDING_VECTOR) #endif #ifdef CONFIG_X86_MCE_P4THERMAL diff --git a/arch/x86/include/asm/hardirq.h b/arch/x86/include/asm/hardirq.h index 37555e52f98..9ebc5c25503 100644 --- a/arch/x86/include/asm/hardirq.h +++ b/arch/x86/include/asm/hardirq.h @@ -13,6 +13,8 @@ typedef struct { unsigned int irq_spurious_count; #endif unsigned int generic_irqs; /* arch dependent */ + unsigned int apic_perf_irqs; + unsigned int apic_pending_irqs; #ifdef CONFIG_SMP unsigned int irq_resched_count; unsigned int irq_call_count; diff --git a/arch/x86/include/asm/hw_irq.h b/arch/x86/include/asm/hw_irq.h index b762ea49bd7..6df45f63966 100644 --- a/arch/x86/include/asm/hw_irq.h +++ b/arch/x86/include/asm/hw_irq.h @@ -29,6 +29,8 @@ extern void apic_timer_interrupt(void); extern void generic_interrupt(void); extern void error_interrupt(void); +extern void perf_pending_interrupt(void); + extern void spurious_interrupt(void); extern void thermal_interrupt(void); extern void reschedule_interrupt(void); @@ -63,7 +65,26 @@ extern unsigned long io_apic_irqs; extern void init_VISWS_APIC_irqs(void); extern void setup_IO_APIC(void); extern void disable_IO_APIC(void); -extern int IO_APIC_get_PCI_irq_vector(int bus, int slot, int fn); + +struct io_apic_irq_attr { + int ioapic; + int ioapic_pin; + int trigger; + int polarity; +}; + +static inline void set_io_apic_irq_attr(struct io_apic_irq_attr *irq_attr, + int ioapic, int ioapic_pin, + int trigger, int polarity) +{ + irq_attr->ioapic = ioapic; + irq_attr->ioapic_pin = ioapic_pin; + irq_attr->trigger = trigger; + irq_attr->polarity = polarity; +} + +extern int IO_APIC_get_PCI_irq_vector(int bus, int devfn, int pin, + struct io_apic_irq_attr *irq_attr); extern void setup_ioapic_dest(void); extern void enable_IO_APIC(void); @@ -78,7 +99,11 @@ extern void eisa_set_level_irq(unsigned int irq); /* SMP */ extern void smp_apic_timer_interrupt(struct pt_regs *); extern void smp_spurious_interrupt(struct pt_regs *); +extern void smp_generic_interrupt(struct pt_regs *); extern void smp_error_interrupt(struct pt_regs *); +#ifdef CONFIG_X86_IO_APIC +extern asmlinkage void smp_irq_move_cleanup_interrupt(void); +#endif #ifdef CONFIG_SMP extern void smp_reschedule_interrupt(struct pt_regs *); extern void smp_call_function_interrupt(struct pt_regs *); diff --git a/arch/x86/include/asm/i387.h b/arch/x86/include/asm/i387.h index 71c9e518398..175adf58dd4 100644 --- a/arch/x86/include/asm/i387.h +++ b/arch/x86/include/asm/i387.h @@ -67,7 +67,7 @@ static inline int fxrstor_checking(struct i387_fxsave_struct *fx) ".previous\n" _ASM_EXTABLE(1b, 3b) : [err] "=r" (err) -#if 0 /* See comment in __save_init_fpu() below. */ +#if 0 /* See comment in fxsave() below. */ : [fx] "r" (fx), "m" (*fx), "0" (0)); #else : [fx] "cdaSDb" (fx), "m" (*fx), "0" (0)); @@ -75,14 +75,6 @@ static inline int fxrstor_checking(struct i387_fxsave_struct *fx) return err; } -static inline int restore_fpu_checking(struct task_struct *tsk) -{ - if (task_thread_info(tsk)->status & TS_XSAVE) - return xrstor_checking(&tsk->thread.xstate->xsave); - else - return fxrstor_checking(&tsk->thread.xstate->fxsave); -} - /* AMD CPUs don't save/restore FDP/FIP/FOP unless an exception is pending. Clear the x87 state here by setting it to fixed values. The kernel data segment can be sometimes 0 and sometimes @@ -120,7 +112,7 @@ static inline int fxsave_user(struct i387_fxsave_struct __user *fx) ".previous\n" _ASM_EXTABLE(1b, 3b) : [err] "=r" (err), "=m" (*fx) -#if 0 /* See comment in __fxsave_clear() below. */ +#if 0 /* See comment in fxsave() below. */ : [fx] "r" (fx), "0" (0)); #else : [fx] "cdaSDb" (fx), "0" (0)); @@ -185,12 +177,9 @@ static inline void tolerant_fwait(void) asm volatile("fnclex ; fwait"); } -static inline void restore_fpu(struct task_struct *tsk) +/* perform fxrstor iff the processor has extended states, otherwise frstor */ +static inline int fxrstor_checking(struct i387_fxsave_struct *fx) { - if (task_thread_info(tsk)->status & TS_XSAVE) { - xrstor_checking(&tsk->thread.xstate->xsave); - return; - } /* * The "nop" is needed to make the instructions the same * length. @@ -199,7 +188,9 @@ static inline void restore_fpu(struct task_struct *tsk) "nop ; frstor %1", "fxrstor %1", X86_FEATURE_FXSR, - "m" (tsk->thread.xstate->fxsave)); + "m" (*fx)); + + return 0; } /* We need a safe address that is cheap to find and that is already @@ -262,6 +253,14 @@ end: #endif /* CONFIG_X86_64 */ +static inline int restore_fpu_checking(struct task_struct *tsk) +{ + if (task_thread_info(tsk)->status & TS_XSAVE) + return xrstor_checking(&tsk->thread.xstate->xsave); + else + return fxrstor_checking(&tsk->thread.xstate->fxsave); +} + /* * Signal frame handlers... */ @@ -305,18 +304,18 @@ static inline void kernel_fpu_end(void) /* * Some instructions like VIA's padlock instructions generate a spurious * DNA fault but don't modify SSE registers. And these instructions - * get used from interrupt context aswell. To prevent these kernel instructions - * in interrupt context interact wrongly with other user/kernel fpu usage, we + * get used from interrupt context as well. To prevent these kernel instructions + * in interrupt context interacting wrongly with other user/kernel fpu usage, we * should use them only in the context of irq_ts_save/restore() */ static inline int irq_ts_save(void) { /* - * If we are in process context, we are ok to take a spurious DNA fault. - * Otherwise, doing clts() in process context require pre-emption to - * be disabled or some heavy lifting like kernel_fpu_begin() + * If in process context and not atomic, we can take a spurious DNA fault. + * Otherwise, doing clts() in process context requires disabling preemption + * or some heavy lifting like kernel_fpu_begin() */ - if (!in_interrupt()) + if (!in_atomic()) return 0; if (read_cr0() & X86_CR0_TS) { diff --git a/arch/x86/include/asm/i8259.h b/arch/x86/include/asm/i8259.h index 1a99e6c092a..58d7091eeb1 100644 --- a/arch/x86/include/asm/i8259.h +++ b/arch/x86/include/asm/i8259.h @@ -60,8 +60,4 @@ extern struct irq_chip i8259A_chip; extern void mask_8259A(void); extern void unmask_8259A(void); -#ifdef CONFIG_X86_32 -extern void init_ISA_irqs(void); -#endif - #endif /* _ASM_X86_I8259_H */ diff --git a/arch/x86/include/asm/intel_arch_perfmon.h b/arch/x86/include/asm/intel_arch_perfmon.h deleted file mode 100644 index fa0fd068bc2..00000000000 --- a/arch/x86/include/asm/intel_arch_perfmon.h +++ /dev/null @@ -1,31 +0,0 @@ -#ifndef _ASM_X86_INTEL_ARCH_PERFMON_H -#define _ASM_X86_INTEL_ARCH_PERFMON_H - -#define MSR_ARCH_PERFMON_PERFCTR0 0xc1 -#define MSR_ARCH_PERFMON_PERFCTR1 0xc2 - -#define MSR_ARCH_PERFMON_EVENTSEL0 0x186 -#define MSR_ARCH_PERFMON_EVENTSEL1 0x187 - -#define ARCH_PERFMON_EVENTSEL0_ENABLE (1 << 22) -#define ARCH_PERFMON_EVENTSEL_INT (1 << 20) -#define ARCH_PERFMON_EVENTSEL_OS (1 << 17) -#define ARCH_PERFMON_EVENTSEL_USR (1 << 16) - -#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_SEL (0x3c) -#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_UMASK (0x00 << 8) -#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_INDEX (0) -#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_PRESENT \ - (1 << (ARCH_PERFMON_UNHALTED_CORE_CYCLES_INDEX)) - -union cpuid10_eax { - struct { - unsigned int version_id:8; - unsigned int num_counters:8; - unsigned int bit_width:8; - unsigned int mask_length:8; - } split; - unsigned int full; -}; - -#endif /* _ASM_X86_INTEL_ARCH_PERFMON_H */ diff --git a/arch/x86/include/asm/io_apic.h b/arch/x86/include/asm/io_apic.h index 9d826e43601..daf866ed061 100644 --- a/arch/x86/include/asm/io_apic.h +++ b/arch/x86/include/asm/io_apic.h @@ -154,22 +154,19 @@ extern int timer_through_8259; extern int io_apic_get_unique_id(int ioapic, int apic_id); extern int io_apic_get_version(int ioapic); extern int io_apic_get_redir_entries(int ioapic); -extern int io_apic_set_pci_routing(int ioapic, int pin, int irq, - int edge_level, int active_high_low); #endif /* CONFIG_ACPI */ +struct io_apic_irq_attr; +extern int io_apic_set_pci_routing(struct device *dev, int irq, + struct io_apic_irq_attr *irq_attr); extern int (*ioapic_renumber_irq)(int ioapic, int irq); extern void ioapic_init_mappings(void); -#ifdef CONFIG_X86_64 extern struct IO_APIC_route_entry **alloc_ioapic_entries(void); extern void free_ioapic_entries(struct IO_APIC_route_entry **ioapic_entries); extern int save_IO_APIC_setup(struct IO_APIC_route_entry **ioapic_entries); extern void mask_IO_APIC_setup(struct IO_APIC_route_entry **ioapic_entries); extern int restore_IO_APIC_setup(struct IO_APIC_route_entry **ioapic_entries); -extern void reinit_intr_remapped_IO_APIC(int intr_remapping, - struct IO_APIC_route_entry **ioapic_entries); -#endif extern void probe_nr_irqs_gsi(void); diff --git a/arch/x86/include/asm/iomap.h b/arch/x86/include/asm/iomap.h index 86af26091d6..0e9fe1d9d97 100644 --- a/arch/x86/include/asm/iomap.h +++ b/arch/x86/include/asm/iomap.h @@ -1,3 +1,6 @@ +#ifndef _ASM_X86_IOMAP_H +#define _ASM_X86_IOMAP_H + /* * Copyright © 2008 Ingo Molnar * @@ -31,3 +34,5 @@ iomap_atomic_prot_pfn(unsigned long pfn, enum km_type type, pgprot_t prot); void iounmap_atomic(void *kvaddr, enum km_type type); + +#endif /* _ASM_X86_IOMAP_H */ diff --git a/arch/x86/include/asm/irq_remapping.h b/arch/x86/include/asm/irq_remapping.h index 0396760fccb..f275e224450 100644 --- a/arch/x86/include/asm/irq_remapping.h +++ b/arch/x86/include/asm/irq_remapping.h @@ -1,6 +1,6 @@ #ifndef _ASM_X86_IRQ_REMAPPING_H #define _ASM_X86_IRQ_REMAPPING_H -#define IRTE_DEST(dest) ((x2apic) ? dest : dest << 8) +#define IRTE_DEST(dest) ((x2apic_mode) ? dest : dest << 8) #endif /* _ASM_X86_IRQ_REMAPPING_H */ diff --git a/arch/x86/include/asm/irq_vectors.h b/arch/x86/include/asm/irq_vectors.h index 3cbd79bbb47..e997be98c9b 100644 --- a/arch/x86/include/asm/irq_vectors.h +++ b/arch/x86/include/asm/irq_vectors.h @@ -34,6 +34,7 @@ #ifdef CONFIG_X86_32 # define SYSCALL_VECTOR 0x80 +# define IA32_SYSCALL_VECTOR 0x80 #else # define IA32_SYSCALL_VECTOR 0x80 #endif @@ -107,14 +108,14 @@ #define LOCAL_TIMER_VECTOR 0xef /* - * Performance monitoring interrupt vector: + * Generic system vector for platform specific use */ -#define LOCAL_PERF_VECTOR 0xee +#define GENERIC_INTERRUPT_VECTOR 0xed /* - * Generic system vector for platform specific use + * Performance monitoring pending work vector: */ -#define GENERIC_INTERRUPT_VECTOR 0xed +#define LOCAL_PENDING_VECTOR 0xec /* * First APIC vector available to drivers: (vectors 0x30-0xee) we diff --git a/arch/x86/include/asm/k8.h b/arch/x86/include/asm/k8.h index 54c8cc53b24..c2d1f3b58e5 100644 --- a/arch/x86/include/asm/k8.h +++ b/arch/x86/include/asm/k8.h @@ -12,4 +12,17 @@ extern int cache_k8_northbridges(void); extern void k8_flush_garts(void); extern int k8_scan_nodes(unsigned long start, unsigned long end); +#ifdef CONFIG_K8_NB +static inline struct pci_dev *node_to_k8_nb_misc(int node) +{ + return (node < num_k8_northbridges) ? k8_northbridges[node] : NULL; +} +#else +static inline struct pci_dev *node_to_k8_nb_misc(int node) +{ + return NULL; +} +#endif + + #endif /* _ASM_X86_K8_H */ diff --git a/arch/x86/include/asm/microcode.h b/arch/x86/include/asm/microcode.h index c882664716c..ef51b501e22 100644 --- a/arch/x86/include/asm/microcode.h +++ b/arch/x86/include/asm/microcode.h @@ -9,20 +9,31 @@ struct cpu_signature { struct device; +enum ucode_state { UCODE_ERROR, UCODE_OK, UCODE_NFOUND }; + struct microcode_ops { - int (*request_microcode_user) (int cpu, const void __user *buf, size_t size); - int (*request_microcode_fw) (int cpu, struct device *device); + enum ucode_state (*request_microcode_user) (int cpu, + const void __user *buf, size_t size); - void (*apply_microcode) (int cpu); + enum ucode_state (*request_microcode_fw) (int cpu, + struct device *device); - int (*collect_cpu_info) (int cpu, struct cpu_signature *csig); void (*microcode_fini_cpu) (int cpu); + + /* + * The generic 'microcode_core' part guarantees that + * the callbacks below run on a target cpu when they + * are being called. + * See also the "Synchronization" section in microcode_core.c. + */ + int (*apply_microcode) (int cpu); + int (*collect_cpu_info) (int cpu, struct cpu_signature *csig); }; struct ucode_cpu_info { - struct cpu_signature cpu_sig; - int valid; - void *mc; + struct cpu_signature cpu_sig; + int valid; + void *mc; }; extern struct ucode_cpu_info ucode_cpu_info[]; diff --git a/arch/x86/include/asm/mpspec.h b/arch/x86/include/asm/mpspec.h index 642fc7fc8cd..e2a1bb6d71e 100644 --- a/arch/x86/include/asm/mpspec.h +++ b/arch/x86/include/asm/mpspec.h @@ -61,9 +61,11 @@ extern void get_smp_config(void); #ifdef CONFIG_X86_MPPARSE extern void find_smp_config(void); extern void early_reserve_e820_mpc_new(void); +extern int enable_update_mptable; #else static inline void find_smp_config(void) { } static inline void early_reserve_e820_mpc_new(void) { } +#define enable_update_mptable 0 #endif void __cpuinit generic_processor_info(int apicid, int version); @@ -72,20 +74,13 @@ extern void mp_register_ioapic(int id, u32 address, u32 gsi_base); extern void mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi); extern void mp_config_acpi_legacy_irqs(void); -extern int mp_register_gsi(u32 gsi, int edge_level, int active_high_low); +struct device; +extern int mp_register_gsi(struct device *dev, u32 gsi, int edge_level, + int active_high_low); extern int acpi_probe_gsi(void); #ifdef CONFIG_X86_IO_APIC -extern int mp_config_acpi_gsi(unsigned char number, unsigned int devfn, u8 pin, - u32 gsi, int triggering, int polarity); extern int mp_find_ioapic(int gsi); extern int mp_find_ioapic_pin(int ioapic, int gsi); -#else -static inline int -mp_config_acpi_gsi(unsigned char number, unsigned int devfn, u8 pin, - u32 gsi, int triggering, int polarity) -{ - return 0; -} #endif #else /* !CONFIG_ACPI: */ static inline int acpi_probe_gsi(void) diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index ec41fc16c16..4d58d04fca8 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -121,7 +121,6 @@ #define MSR_K8_TOP_MEM1 0xc001001a #define MSR_K8_TOP_MEM2 0xc001001d #define MSR_K8_SYSCFG 0xc0010010 -#define MSR_K8_HWCR 0xc0010015 #define MSR_K8_INT_PENDING_MSG 0xc0010055 /* C1E active bits in int pending message */ #define K8_INTP_C1E_ACTIVE_MASK 0x18000000 diff --git a/arch/x86/include/asm/nmi.h b/arch/x86/include/asm/nmi.h index c45a0a568df..c9726440993 100644 --- a/arch/x86/include/asm/nmi.h +++ b/arch/x86/include/asm/nmi.h @@ -64,7 +64,7 @@ static inline int nmi_watchdog_active(void) * but since they are power of two we could use a * cheaper way --cvg */ - return nmi_watchdog & 0x3; + return nmi_watchdog & (NMI_LOCAL_APIC | NMI_IO_APIC); } #endif diff --git a/arch/x86/include/asm/numa_64.h b/arch/x86/include/asm/numa_64.h index 064ed6df4cb..c4ae822e415 100644 --- a/arch/x86/include/asm/numa_64.h +++ b/arch/x86/include/asm/numa_64.h @@ -17,9 +17,6 @@ extern int compute_hash_shift(struct bootnode *nodes, int numblks, extern void numa_init_array(void); extern int numa_off; -extern void srat_reserve_add_area(int nodeid); -extern int hotadd_percent; - extern s16 apicid_to_node[MAX_LOCAL_APIC]; extern unsigned long numa_free_all_bootmem(void); @@ -27,6 +24,13 @@ extern void setup_node_bootmem(int nodeid, unsigned long start, unsigned long end); #ifdef CONFIG_NUMA +/* + * Too small node sizes may confuse the VM badly. Usually they + * result from BIOS bugs. So dont recognize nodes as standalone + * NUMA entities that have less than this amount of RAM listed: + */ +#define NODE_MIN_SIZE (4*1024*1024) + extern void __init init_cpu_to_node(void); extern void __cpuinit numa_set_node(int cpu, int node); extern void __cpuinit numa_clear_node(int cpu); diff --git a/arch/x86/include/asm/page_32_types.h b/arch/x86/include/asm/page_32_types.h index 0f915ae649a..6f1b7331313 100644 --- a/arch/x86/include/asm/page_32_types.h +++ b/arch/x86/include/asm/page_32_types.h @@ -54,10 +54,6 @@ extern unsigned int __VMALLOC_RESERVE; extern int sysctl_legacy_va_layout; extern void find_low_pfn_range(void); -extern unsigned long init_memory_mapping(unsigned long start, - unsigned long end); -extern void initmem_init(unsigned long, unsigned long); -extern void free_initmem(void); extern void setup_bootmem_allocator(void); #endif /* !__ASSEMBLY__ */ diff --git a/arch/x86/include/asm/page_64_types.h b/arch/x86/include/asm/page_64_types.h index d38c91b7024..8d382d3abf3 100644 --- a/arch/x86/include/asm/page_64_types.h +++ b/arch/x86/include/asm/page_64_types.h @@ -32,22 +32,14 @@ */ #define __PAGE_OFFSET _AC(0xffff880000000000, UL) -#define __PHYSICAL_START CONFIG_PHYSICAL_START -#define __KERNEL_ALIGN 0x200000 - -/* - * Make sure kernel is aligned to 2MB address. Catching it at compile - * time is better. Change your config file and compile the kernel - * for a 2MB aligned address (CONFIG_PHYSICAL_START) - */ -#if (CONFIG_PHYSICAL_START % __KERNEL_ALIGN) != 0 -#error "CONFIG_PHYSICAL_START must be a multiple of 2MB" -#endif +#define __PHYSICAL_START ((CONFIG_PHYSICAL_START + \ + (CONFIG_PHYSICAL_ALIGN - 1)) & \ + ~(CONFIG_PHYSICAL_ALIGN - 1)) #define __START_KERNEL (__START_KERNEL_map + __PHYSICAL_START) #define __START_KERNEL_map _AC(0xffffffff80000000, UL) -/* See Documentation/x86_64/mm.txt for a description of the memory map. */ +/* See Documentation/x86/x86_64/mm.txt for a description of the memory map. */ #define __PHYSICAL_MASK_SHIFT 46 #define __VIRTUAL_MASK_SHIFT 48 @@ -71,12 +63,6 @@ extern unsigned long __phys_addr(unsigned long); #define vmemmap ((struct page *)VMEMMAP_START) -extern unsigned long init_memory_mapping(unsigned long start, - unsigned long end); - -extern void initmem_init(unsigned long start_pfn, unsigned long end_pfn); -extern void free_initmem(void); - extern void init_extra_mapping_uc(unsigned long phys, unsigned long size); extern void init_extra_mapping_wb(unsigned long phys, unsigned long size); diff --git a/arch/x86/include/asm/page_types.h b/arch/x86/include/asm/page_types.h index 826ad37006a..6473f5ccff8 100644 --- a/arch/x86/include/asm/page_types.h +++ b/arch/x86/include/asm/page_types.h @@ -46,6 +46,12 @@ extern int devmem_is_allowed(unsigned long pagenr); extern unsigned long max_low_pfn_mapped; extern unsigned long max_pfn_mapped; +extern unsigned long init_memory_mapping(unsigned long start, + unsigned long end); + +extern void initmem_init(unsigned long start_pfn, unsigned long end_pfn); +extern void free_initmem(void); + #endif /* !__ASSEMBLY__ */ #endif /* _ASM_X86_PAGE_DEFS_H */ diff --git a/arch/x86/include/asm/paravirt.h b/arch/x86/include/asm/paravirt.h index a53da004e08..4fb37c8a083 100644 --- a/arch/x86/include/asm/paravirt.h +++ b/arch/x86/include/asm/paravirt.h @@ -56,6 +56,7 @@ struct desc_ptr; struct tss_struct; struct mm_struct; struct desc_struct; +struct task_struct; /* * Wrapper type for pointers to code which uses the non-standard @@ -203,7 +204,8 @@ struct pv_cpu_ops { void (*swapgs)(void); - struct pv_lazy_ops lazy_mode; + void (*start_context_switch)(struct task_struct *prev); + void (*end_context_switch)(struct task_struct *next); }; struct pv_irq_ops { @@ -1399,25 +1401,23 @@ enum paravirt_lazy_mode { }; enum paravirt_lazy_mode paravirt_get_lazy_mode(void); -void paravirt_enter_lazy_cpu(void); -void paravirt_leave_lazy_cpu(void); +void paravirt_start_context_switch(struct task_struct *prev); +void paravirt_end_context_switch(struct task_struct *next); + void paravirt_enter_lazy_mmu(void); void paravirt_leave_lazy_mmu(void); -void paravirt_leave_lazy(enum paravirt_lazy_mode mode); -#define __HAVE_ARCH_ENTER_LAZY_CPU_MODE -static inline void arch_enter_lazy_cpu_mode(void) +#define __HAVE_ARCH_START_CONTEXT_SWITCH +static inline void arch_start_context_switch(struct task_struct *prev) { - PVOP_VCALL0(pv_cpu_ops.lazy_mode.enter); + PVOP_VCALL1(pv_cpu_ops.start_context_switch, prev); } -static inline void arch_leave_lazy_cpu_mode(void) +static inline void arch_end_context_switch(struct task_struct *next) { - PVOP_VCALL0(pv_cpu_ops.lazy_mode.leave); + PVOP_VCALL1(pv_cpu_ops.end_context_switch, next); } -void arch_flush_lazy_cpu_mode(void); - #define __HAVE_ARCH_ENTER_LAZY_MMU_MODE static inline void arch_enter_lazy_mmu_mode(void) { diff --git a/arch/x86/include/asm/perf_counter.h b/arch/x86/include/asm/perf_counter.h new file mode 100644 index 00000000000..876ed97147b --- /dev/null +++ b/arch/x86/include/asm/perf_counter.h @@ -0,0 +1,100 @@ +#ifndef _ASM_X86_PERF_COUNTER_H +#define _ASM_X86_PERF_COUNTER_H + +/* + * Performance counter hw details: + */ + +#define X86_PMC_MAX_GENERIC 8 +#define X86_PMC_MAX_FIXED 3 + +#define X86_PMC_IDX_GENERIC 0 +#define X86_PMC_IDX_FIXED 32 +#define X86_PMC_IDX_MAX 64 + +#define MSR_ARCH_PERFMON_PERFCTR0 0xc1 +#define MSR_ARCH_PERFMON_PERFCTR1 0xc2 + +#define MSR_ARCH_PERFMON_EVENTSEL0 0x186 +#define MSR_ARCH_PERFMON_EVENTSEL1 0x187 + +#define ARCH_PERFMON_EVENTSEL0_ENABLE (1 << 22) +#define ARCH_PERFMON_EVENTSEL_INT (1 << 20) +#define ARCH_PERFMON_EVENTSEL_OS (1 << 17) +#define ARCH_PERFMON_EVENTSEL_USR (1 << 16) + +/* + * Includes eventsel and unit mask as well: + */ +#define ARCH_PERFMON_EVENT_MASK 0xffff + +#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_SEL 0x3c +#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_UMASK (0x00 << 8) +#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_INDEX 0 +#define ARCH_PERFMON_UNHALTED_CORE_CYCLES_PRESENT \ + (1 << (ARCH_PERFMON_UNHALTED_CORE_CYCLES_INDEX)) + +#define ARCH_PERFMON_BRANCH_MISSES_RETIRED 6 + +/* + * Intel "Architectural Performance Monitoring" CPUID + * detection/enumeration details: + */ +union cpuid10_eax { + struct { + unsigned int version_id:8; + unsigned int num_counters:8; + unsigned int bit_width:8; + unsigned int mask_length:8; + } split; + unsigned int full; +}; + +union cpuid10_edx { + struct { + unsigned int num_counters_fixed:4; + unsigned int reserved:28; + } split; + unsigned int full; +}; + + +/* + * Fixed-purpose performance counters: + */ + +/* + * All 3 fixed-mode PMCs are configured via this single MSR: + */ +#define MSR_ARCH_PERFMON_FIXED_CTR_CTRL 0x38d + +/* + * The counts are available in three separate MSRs: + */ + +/* Instr_Retired.Any: */ +#define MSR_ARCH_PERFMON_FIXED_CTR0 0x309 +#define X86_PMC_IDX_FIXED_INSTRUCTIONS (X86_PMC_IDX_FIXED + 0) + +/* CPU_CLK_Unhalted.Core: */ +#define MSR_ARCH_PERFMON_FIXED_CTR1 0x30a +#define X86_PMC_IDX_FIXED_CPU_CYCLES (X86_PMC_IDX_FIXED + 1) + +/* CPU_CLK_Unhalted.Ref: */ +#define MSR_ARCH_PERFMON_FIXED_CTR2 0x30b +#define X86_PMC_IDX_FIXED_BUS_CYCLES (X86_PMC_IDX_FIXED + 2) + +extern void set_perf_counter_pending(void); + +#define clear_perf_counter_pending() do { } while (0) +#define test_perf_counter_pending() (0) + +#ifdef CONFIG_PERF_COUNTERS +extern void init_hw_perf_counters(void); +extern void perf_counters_lapic_init(void); +#else +static inline void init_hw_perf_counters(void) { } +static inline void perf_counters_lapic_init(void) { } +#endif + +#endif /* _ASM_X86_PERF_COUNTER_H */ diff --git a/arch/x86/include/asm/pgtable.h b/arch/x86/include/asm/pgtable.h index 29d96d168bc..18ef7ebf263 100644 --- a/arch/x86/include/asm/pgtable.h +++ b/arch/x86/include/asm/pgtable.h @@ -81,6 +81,8 @@ static inline void __init paravirt_pagetable_setup_done(pgd_t *base) #define pte_val(x) native_pte_val(x) #define __pte(x) native_make_pte(x) +#define arch_end_context_switch(prev) do {} while(0) + #endif /* CONFIG_PARAVIRT */ /* @@ -503,6 +505,8 @@ static inline int pgd_none(pgd_t pgd) #ifndef __ASSEMBLY__ +extern int direct_gbpages; + /* local pte updates need not use xchg for locking */ static inline pte_t native_local_ptep_get_and_clear(pte_t *ptep) { diff --git a/arch/x86/include/asm/pgtable_64.h b/arch/x86/include/asm/pgtable_64.h index 6b87bc6d501..abde308fdb0 100644 --- a/arch/x86/include/asm/pgtable_64.h +++ b/arch/x86/include/asm/pgtable_64.h @@ -25,10 +25,6 @@ extern pgd_t init_level4_pgt[]; extern void paging_init(void); -#endif /* !__ASSEMBLY__ */ - -#ifndef __ASSEMBLY__ - #define pte_ERROR(e) \ printk("%s:%d: bad pte %p(%016lx).\n", \ __FILE__, __LINE__, &(e), pte_val(e)) @@ -135,8 +131,6 @@ static inline int pgd_large(pgd_t pgd) { return 0; } #define update_mmu_cache(vma, address, pte) do { } while (0) -extern int direct_gbpages; - /* Encode and de-code a swap entry */ #if _PAGE_BIT_FILE < _PAGE_BIT_PROTNONE #define SWP_TYPE_BITS (_PAGE_BIT_FILE - _PAGE_BIT_PRESENT - 1) diff --git a/arch/x86/include/asm/pgtable_64_types.h b/arch/x86/include/asm/pgtable_64_types.h index fbf42b8e038..766ea16fbbb 100644 --- a/arch/x86/include/asm/pgtable_64_types.h +++ b/arch/x86/include/asm/pgtable_64_types.h @@ -51,11 +51,11 @@ typedef struct { pteval_t pte; } pte_t; #define PGDIR_SIZE (_AC(1, UL) << PGDIR_SHIFT) #define PGDIR_MASK (~(PGDIR_SIZE - 1)) - +/* See Documentation/x86/x86_64/mm.txt for a description of the memory map. */ #define MAXMEM _AC(__AC(1, UL) << MAX_PHYSMEM_BITS, UL) -#define VMALLOC_START _AC(0xffffc20000000000, UL) -#define VMALLOC_END _AC(0xffffe1ffffffffff, UL) -#define VMEMMAP_START _AC(0xffffe20000000000, UL) +#define VMALLOC_START _AC(0xffffc90000000000, UL) +#define VMALLOC_END _AC(0xffffe8ffffffffff, UL) +#define VMEMMAP_START _AC(0xffffea0000000000, UL) #define MODULES_VADDR _AC(0xffffffffa0000000, UL) #define MODULES_END _AC(0xffffffffff000000, UL) #define MODULES_LEN (MODULES_END - MODULES_VADDR) diff --git a/arch/x86/include/asm/pgtable_types.h b/arch/x86/include/asm/pgtable_types.h index b8238dc8786..4d258ad76a0 100644 --- a/arch/x86/include/asm/pgtable_types.h +++ b/arch/x86/include/asm/pgtable_types.h @@ -273,7 +273,6 @@ typedef struct page *pgtable_t; extern pteval_t __supported_pte_mask; extern int nx_enabled; -extern void set_nx(void); #define pgprot_writecombine pgprot_writecombine extern pgprot_t pgprot_writecombine(pgprot_t prot); diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index c2cceae709c..c7768269b1c 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -135,7 +135,8 @@ extern struct cpuinfo_x86 boot_cpu_data; extern struct cpuinfo_x86 new_cpu_data; extern struct tss_struct doublefault_tss; -extern __u32 cleared_cpu_caps[NCAPINTS]; +extern __u32 cpu_caps_cleared[NCAPINTS]; +extern __u32 cpu_caps_set[NCAPINTS]; #ifdef CONFIG_SMP DECLARE_PER_CPU_SHARED_ALIGNED(struct cpuinfo_x86, cpu_info); @@ -409,9 +410,6 @@ DECLARE_PER_CPU(unsigned long, stack_canary); extern unsigned int xstate_size; extern void free_thread_xstate(struct task_struct *); extern struct kmem_cache *task_xstate_cachep; -extern void init_scattered_cpuid_features(struct cpuinfo_x86 *c); -extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c); -extern unsigned short num_cache_leaves; struct thread_struct { /* Cached TLS descriptors: */ @@ -427,8 +425,12 @@ struct thread_struct { unsigned short fsindex; unsigned short gsindex; #endif +#ifdef CONFIG_X86_32 unsigned long ip; +#endif +#ifdef CONFIG_X86_64 unsigned long fs; +#endif unsigned long gs; /* Hardware debugging registers: */ unsigned long debugreg0; @@ -460,14 +462,8 @@ struct thread_struct { unsigned io_bitmap_max; /* MSR_IA32_DEBUGCTLMSR value to switch in if TIF_DEBUGCTLMSR is set. */ unsigned long debugctlmsr; -#ifdef CONFIG_X86_DS -/* Debug Store context; see include/asm-x86/ds.h; goes into MSR_IA32_DS_AREA */ + /* Debug Store context; see asm/ds.h */ struct ds_context *ds_ctx; -#endif /* CONFIG_X86_DS */ -#ifdef CONFIG_X86_PTRACE_BTS -/* the signal to send on a bts buffer overflow */ - unsigned int bts_ovfl_signal; -#endif /* CONFIG_X86_PTRACE_BTS */ }; static inline unsigned long native_get_debugreg(int regno) @@ -795,6 +791,21 @@ static inline unsigned long get_debugctlmsr(void) return debugctlmsr; } +static inline unsigned long get_debugctlmsr_on_cpu(int cpu) +{ + u64 debugctlmsr = 0; + u32 val1, val2; + +#ifndef CONFIG_X86_DEBUGCTLMSR + if (boot_cpu_data.x86 < 6) + return 0; +#endif + rdmsr_on_cpu(cpu, MSR_IA32_DEBUGCTLMSR, &val1, &val2); + debugctlmsr = val1 | ((u64)val2 << 32); + + return debugctlmsr; +} + static inline void update_debugctlmsr(unsigned long debugctlmsr) { #ifndef CONFIG_X86_DEBUGCTLMSR @@ -804,6 +815,18 @@ static inline void update_debugctlmsr(unsigned long debugctlmsr) wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctlmsr); } +static inline void update_debugctlmsr_on_cpu(int cpu, + unsigned long debugctlmsr) +{ +#ifndef CONFIG_X86_DEBUGCTLMSR + if (boot_cpu_data.x86 < 6) + return; +#endif + wrmsr_on_cpu(cpu, MSR_IA32_DEBUGCTLMSR, + (u32)((u64)debugctlmsr), + (u32)((u64)debugctlmsr >> 32)); +} + /* * from system description table in BIOS. Mostly for MCA use, but * others may find it useful: @@ -814,6 +837,7 @@ extern unsigned int BIOS_revision; /* Boot loader type from the setup header: */ extern int bootloader_type; +extern int bootloader_version; extern char ignore_fpu_irq; @@ -874,7 +898,6 @@ static inline void spin_lock_prefetch(const void *x) .vm86_info = NULL, \ .sysenter_cs = __KERNEL_CS, \ .io_bitmap_ptr = NULL, \ - .fs = __KERNEL_PERCPU, \ } /* diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h index 624f133943e..0f0d908349a 100644 --- a/arch/x86/include/asm/ptrace.h +++ b/arch/x86/include/asm/ptrace.h @@ -236,12 +236,11 @@ extern int do_get_thread_area(struct task_struct *p, int idx, extern int do_set_thread_area(struct task_struct *p, int idx, struct user_desc __user *info, int can_allocate); -extern void x86_ptrace_untrace(struct task_struct *); -extern void x86_ptrace_fork(struct task_struct *child, - unsigned long clone_flags); +#ifdef CONFIG_X86_PTRACE_BTS +extern void ptrace_bts_untrace(struct task_struct *tsk); -#define arch_ptrace_untrace(tsk) x86_ptrace_untrace(tsk) -#define arch_ptrace_fork(child, flags) x86_ptrace_fork(child, flags) +#define arch_ptrace_untrace(tsk) ptrace_bts_untrace(tsk) +#endif /* CONFIG_X86_PTRACE_BTS */ #endif /* __KERNEL__ */ diff --git a/arch/x86/include/asm/required-features.h b/arch/x86/include/asm/required-features.h index a4737dddfd5..64cf2d24fad 100644 --- a/arch/x86/include/asm/required-features.h +++ b/arch/x86/include/asm/required-features.h @@ -48,9 +48,15 @@ #endif #ifdef CONFIG_X86_64 +#ifdef CONFIG_PARAVIRT +/* Paravirtualized systems may not have PSE or PGE available */ #define NEED_PSE 0 -#define NEED_MSR (1<<(X86_FEATURE_MSR & 31)) #define NEED_PGE 0 +#else +#define NEED_PSE (1<<(X86_FEATURE_PSE) & 31) +#define NEED_PGE (1<<(X86_FEATURE_PGE) & 31) +#endif +#define NEED_MSR (1<<(X86_FEATURE_MSR & 31)) #define NEED_FXSR (1<<(X86_FEATURE_FXSR & 31)) #define NEED_XMM (1<<(X86_FEATURE_XMM & 31)) #define NEED_XMM2 (1<<(X86_FEATURE_XMM2 & 31)) diff --git a/arch/x86/include/asm/setup.h b/arch/x86/include/asm/setup.h index bdc2ada05ae..4093d1ed6db 100644 --- a/arch/x86/include/asm/setup.h +++ b/arch/x86/include/asm/setup.h @@ -33,7 +33,6 @@ struct x86_quirks { int (*setup_ioapic_ids)(void); }; -extern void x86_quirk_pre_intr_init(void); extern void x86_quirk_intr_init(void); extern void x86_quirk_trap_init(void); diff --git a/arch/x86/include/asm/smp.h b/arch/x86/include/asm/smp.h index 19e0d88b966..6a84ed166ae 100644 --- a/arch/x86/include/asm/smp.h +++ b/arch/x86/include/asm/smp.h @@ -180,7 +180,7 @@ extern int safe_smp_processor_id(void); static inline int logical_smp_processor_id(void) { /* we don't want to mark this access volatile - bad code generation */ - return GET_APIC_LOGICAL_ID(*(u32 *)(APIC_BASE + APIC_LDR)); + return GET_APIC_LOGICAL_ID(apic_read(APIC_LDR)); } #endif diff --git a/arch/x86/include/asm/sparsemem.h b/arch/x86/include/asm/sparsemem.h index e3cc3c063ec..4517d6b9318 100644 --- a/arch/x86/include/asm/sparsemem.h +++ b/arch/x86/include/asm/sparsemem.h @@ -27,7 +27,7 @@ #else /* CONFIG_X86_32 */ # define SECTION_SIZE_BITS 27 /* matt - 128 is convenient right now */ # define MAX_PHYSADDR_BITS 44 -# define MAX_PHYSMEM_BITS 44 /* Can be max 45 bits */ +# define MAX_PHYSMEM_BITS 46 #endif #endif /* CONFIG_SPARSEMEM */ diff --git a/arch/x86/include/asm/syscalls.h b/arch/x86/include/asm/syscalls.h index 7043408f690..372b76edd63 100644 --- a/arch/x86/include/asm/syscalls.h +++ b/arch/x86/include/asm/syscalls.h @@ -1,7 +1,7 @@ /* * syscalls.h - Linux syscall interfaces (arch-specific) * - * Copyright (c) 2008 Jaswinder Singh + * Copyright (c) 2008 Jaswinder Singh Rajput * * This file is released under the GPLv2. * See the file COPYING for more details. @@ -12,50 +12,55 @@ #include <linux/compiler.h> #include <linux/linkage.h> -#include <linux/types.h> #include <linux/signal.h> +#include <linux/types.h> /* Common in X86_32 and X86_64 */ /* kernel/ioport.c */ asmlinkage long sys_ioperm(unsigned long, unsigned long, int); +/* kernel/process.c */ +int sys_fork(struct pt_regs *); +int sys_vfork(struct pt_regs *); + /* kernel/ldt.c */ asmlinkage int sys_modify_ldt(int, void __user *, unsigned long); +/* kernel/signal.c */ +long sys_rt_sigreturn(struct pt_regs *); + /* kernel/tls.c */ asmlinkage int sys_set_thread_area(struct user_desc __user *); asmlinkage int sys_get_thread_area(struct user_desc __user *); /* X86_32 only */ #ifdef CONFIG_X86_32 +/* kernel/ioport.c */ +long sys_iopl(struct pt_regs *); + /* kernel/process_32.c */ -int sys_fork(struct pt_regs *); int sys_clone(struct pt_regs *); -int sys_vfork(struct pt_regs *); int sys_execve(struct pt_regs *); -/* kernel/signal_32.c */ +/* kernel/signal.c */ asmlinkage int sys_sigsuspend(int, int, old_sigset_t); asmlinkage int sys_sigaction(int, const struct old_sigaction __user *, struct old_sigaction __user *); int sys_sigaltstack(struct pt_regs *); unsigned long sys_sigreturn(struct pt_regs *); -long sys_rt_sigreturn(struct pt_regs *); - -/* kernel/ioport.c */ -long sys_iopl(struct pt_regs *); /* kernel/sys_i386_32.c */ +struct mmap_arg_struct; +struct sel_arg_struct; +struct oldold_utsname; +struct old_utsname; + asmlinkage long sys_mmap2(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long); -struct mmap_arg_struct; asmlinkage int old_mmap(struct mmap_arg_struct __user *); -struct sel_arg_struct; asmlinkage int old_select(struct sel_arg_struct __user *); asmlinkage int sys_ipc(uint, int, int, int, void __user *, long); -struct old_utsname; asmlinkage int sys_uname(struct old_utsname __user *); -struct oldold_utsname; asmlinkage int sys_olduname(struct oldold_utsname __user *); /* kernel/vm86_32.c */ @@ -65,29 +70,27 @@ int sys_vm86(struct pt_regs *); #else /* CONFIG_X86_32 */ /* X86_64 only */ +/* kernel/ioport.c */ +asmlinkage long sys_iopl(unsigned int, struct pt_regs *); + /* kernel/process_64.c */ -asmlinkage long sys_fork(struct pt_regs *); asmlinkage long sys_clone(unsigned long, unsigned long, void __user *, void __user *, struct pt_regs *); -asmlinkage long sys_vfork(struct pt_regs *); asmlinkage long sys_execve(char __user *, char __user * __user *, char __user * __user *, struct pt_regs *); long sys_arch_prctl(int, unsigned long); -/* kernel/ioport.c */ -asmlinkage long sys_iopl(unsigned int, struct pt_regs *); - -/* kernel/signal_64.c */ +/* kernel/signal.c */ asmlinkage long sys_sigaltstack(const stack_t __user *, stack_t __user *, struct pt_regs *); -long sys_rt_sigreturn(struct pt_regs *); /* kernel/sys_x86_64.c */ +struct new_utsname; + asmlinkage long sys_mmap(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long); -struct new_utsname; asmlinkage long sys_uname(struct new_utsname __user *); #endif /* CONFIG_X86_32 */ diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index 8820a73ae09..602c769fc98 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -94,7 +94,8 @@ struct thread_info { #define TIF_FORCED_TF 24 /* true if TF in eflags artificially */ #define TIF_DEBUGCTLMSR 25 /* uses thread_struct.debugctlmsr */ #define TIF_DS_AREA_MSR 26 /* uses thread_struct.ds_area_msr */ -#define TIF_SYSCALL_FTRACE 27 /* for ftrace syscall instrumentation */ +#define TIF_LAZY_MMU_UPDATES 27 /* task is updating the mmu lazily */ +#define TIF_SYSCALL_FTRACE 28 /* for ftrace syscall instrumentation */ #define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE) #define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME) @@ -116,6 +117,7 @@ struct thread_info { #define _TIF_FORCED_TF (1 << TIF_FORCED_TF) #define _TIF_DEBUGCTLMSR (1 << TIF_DEBUGCTLMSR) #define _TIF_DS_AREA_MSR (1 << TIF_DS_AREA_MSR) +#define _TIF_LAZY_MMU_UPDATES (1 << TIF_LAZY_MMU_UPDATES) #define _TIF_SYSCALL_FTRACE (1 << TIF_SYSCALL_FTRACE) /* work to do in syscall_trace_enter() */ diff --git a/arch/x86/include/asm/tlbflush.h b/arch/x86/include/asm/tlbflush.h index 16a5c84b032..a5ecc9c33e9 100644 --- a/arch/x86/include/asm/tlbflush.h +++ b/arch/x86/include/asm/tlbflush.h @@ -17,7 +17,7 @@ static inline void __native_flush_tlb(void) { - write_cr3(read_cr3()); + native_write_cr3(native_read_cr3()); } static inline void __native_flush_tlb_global(void) @@ -32,11 +32,11 @@ static inline void __native_flush_tlb_global(void) */ raw_local_irq_save(flags); - cr4 = read_cr4(); + cr4 = native_read_cr4(); /* clear PGE */ - write_cr4(cr4 & ~X86_CR4_PGE); + native_write_cr4(cr4 & ~X86_CR4_PGE); /* write old PGE again and flush TLBs */ - write_cr4(cr4); + native_write_cr4(cr4); raw_local_irq_restore(flags); } diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index f44b49abca4..066ef590d7e 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -203,7 +203,8 @@ struct pci_bus; void x86_pci_root_bus_res_quirks(struct pci_bus *b); #ifdef CONFIG_SMP -#define mc_capable() (cpumask_weight(cpu_core_mask(0)) != nr_cpu_ids) +#define mc_capable() ((boot_cpu_data.x86_max_cores > 1) && \ + (cpumask_weight(cpu_core_mask(0)) != nr_cpu_ids)) #define smt_capable() (smp_num_siblings > 1) #endif diff --git a/arch/x86/include/asm/traps.h b/arch/x86/include/asm/traps.h index 0d5342515b8..bfd74c032fc 100644 --- a/arch/x86/include/asm/traps.h +++ b/arch/x86/include/asm/traps.h @@ -2,6 +2,7 @@ #define _ASM_X86_TRAPS_H #include <asm/debugreg.h> +#include <asm/siginfo.h> /* TRAP_TRACE, ... */ #ifdef CONFIG_X86_32 #define dotraplinkage @@ -13,6 +14,9 @@ asmlinkage void divide_error(void); asmlinkage void debug(void); asmlinkage void nmi(void); asmlinkage void int3(void); +asmlinkage void xen_debug(void); +asmlinkage void xen_int3(void); +asmlinkage void xen_stack_segment(void); asmlinkage void overflow(void); asmlinkage void bounds(void); asmlinkage void invalid_op(void); @@ -74,7 +78,6 @@ static inline int get_si_code(unsigned long condition) } extern int panic_on_unrecovered_nmi; -extern int kstack_depth_to_print; void math_error(void __user *); void math_emulate(struct math_emu_info *); diff --git a/arch/x86/include/asm/unistd_32.h b/arch/x86/include/asm/unistd_32.h index 6e72d74cf8d..732a3070615 100644 --- a/arch/x86/include/asm/unistd_32.h +++ b/arch/x86/include/asm/unistd_32.h @@ -340,6 +340,8 @@ #define __NR_inotify_init1 332 #define __NR_preadv 333 #define __NR_pwritev 334 +#define __NR_rt_tgsigqueueinfo 335 +#define __NR_perf_counter_open 336 #ifdef __KERNEL__ diff --git a/arch/x86/include/asm/unistd_64.h b/arch/x86/include/asm/unistd_64.h index f8182946232..900e1617e67 100644 --- a/arch/x86/include/asm/unistd_64.h +++ b/arch/x86/include/asm/unistd_64.h @@ -657,7 +657,10 @@ __SYSCALL(__NR_inotify_init1, sys_inotify_init1) __SYSCALL(__NR_preadv, sys_preadv) #define __NR_pwritev 296 __SYSCALL(__NR_pwritev, sys_pwritev) - +#define __NR_rt_tgsigqueueinfo 297 +__SYSCALL(__NR_rt_tgsigqueueinfo, sys_rt_tgsigqueueinfo) +#define __NR_perf_counter_open 298 +__SYSCALL(__NR_perf_counter_open, sys_perf_counter_open) #ifndef __NO_STUBS #define __ARCH_WANT_OLD_READDIR diff --git a/arch/x86/include/asm/uv/uv_bau.h b/arch/x86/include/asm/uv/uv_bau.h index 9b0e61bf7a8..bddd44f2f0a 100644 --- a/arch/x86/include/asm/uv/uv_bau.h +++ b/arch/x86/include/asm/uv/uv_bau.h @@ -37,7 +37,7 @@ #define UV_CPUS_PER_ACT_STATUS 32 #define UV_ACT_STATUS_MASK 0x3 #define UV_ACT_STATUS_SIZE 2 -#define UV_ACTIVATION_DESCRIPTOR_SIZE 32 +#define UV_ADP_SIZE 32 #define UV_DISTRIBUTION_SIZE 256 #define UV_SW_ACK_NPENDING 8 #define UV_NET_ENDPOINT_INTD 0x38 diff --git a/arch/x86/include/asm/uv/uv_hub.h b/arch/x86/include/asm/uv/uv_hub.h index d3a98ea1062..341070f7ad5 100644 --- a/arch/x86/include/asm/uv/uv_hub.h +++ b/arch/x86/include/asm/uv/uv_hub.h @@ -133,6 +133,7 @@ struct uv_scir_s { struct uv_hub_info_s { unsigned long global_mmr_base; unsigned long gpa_mask; + unsigned int gnode_extra; unsigned long gnode_upper; unsigned long lowmem_remap_top; unsigned long lowmem_remap_base; @@ -159,7 +160,8 @@ DECLARE_PER_CPU(struct uv_hub_info_s, __uv_hub_info); * p - PNODE (local part of nsids, right shifted 1) */ #define UV_NASID_TO_PNODE(n) (((n) >> 1) & uv_hub_info->pnode_mask) -#define UV_PNODE_TO_NASID(p) (((p) << 1) | uv_hub_info->gnode_upper) +#define UV_PNODE_TO_GNODE(p) ((p) |uv_hub_info->gnode_extra) +#define UV_PNODE_TO_NASID(p) (UV_PNODE_TO_GNODE(p) << 1) #define UV_LOCAL_MMR_BASE 0xf4000000UL #define UV_GLOBAL_MMR32_BASE 0xf8000000UL @@ -173,7 +175,7 @@ DECLARE_PER_CPU(struct uv_hub_info_s, __uv_hub_info); #define UV_GLOBAL_MMR32_PNODE_BITS(p) ((p) << (UV_GLOBAL_MMR32_PNODE_SHIFT)) #define UV_GLOBAL_MMR64_PNODE_BITS(p) \ - ((unsigned long)(p) << UV_GLOBAL_MMR64_PNODE_SHIFT) + ((unsigned long)(UV_PNODE_TO_GNODE(p)) << UV_GLOBAL_MMR64_PNODE_SHIFT) #define UV_APIC_PNODE_SHIFT 6 diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 88d1bfc847d..4f78bd68212 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -28,7 +28,7 @@ CFLAGS_paravirt.o := $(nostackp) obj-y := process_$(BITS).o signal.o entry_$(BITS).o obj-y += traps.o irq.o irq_$(BITS).o dumpstack_$(BITS).o obj-y += time_$(BITS).o ioport.o ldt.o dumpstack.o -obj-y += setup.o i8259.o irqinit_$(BITS).o +obj-y += setup.o i8259.o irqinit.o obj-$(CONFIG_X86_VISWS) += visws_quirks.o obj-$(CONFIG_X86_32) += probe_roms_32.o obj-$(CONFIG_X86_32) += sys_i386_32.o i386_ksyms_32.o @@ -44,6 +44,7 @@ obj-y += process.o obj-y += i387.o xsave.o obj-y += ptrace.o obj-$(CONFIG_X86_DS) += ds.o +obj-$(CONFIG_X86_DS_SELFTEST) += ds_selftest.o obj-$(CONFIG_X86_32) += tls.o obj-$(CONFIG_IA32_EMULATION) += tls.o obj-y += step.o diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 723989d7f80..631086159c5 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -33,6 +33,7 @@ #include <linux/irq.h> #include <linux/bootmem.h> #include <linux/ioport.h> +#include <linux/pci.h> #include <asm/pgtable.h> #include <asm/io_apic.h> @@ -522,7 +523,7 @@ int acpi_gsi_to_irq(u32 gsi, unsigned int *irq) * success: return IRQ number (>=0) * failure: return < 0 */ -int acpi_register_gsi(u32 gsi, int triggering, int polarity) +int acpi_register_gsi(struct device *dev, u32 gsi, int trigger, int polarity) { unsigned int irq; unsigned int plat_gsi = gsi; @@ -532,14 +533,14 @@ int acpi_register_gsi(u32 gsi, int triggering, int polarity) * Make sure all (legacy) PCI IRQs are set as level-triggered. */ if (acpi_irq_model == ACPI_IRQ_MODEL_PIC) { - if (triggering == ACPI_LEVEL_SENSITIVE) + if (trigger == ACPI_LEVEL_SENSITIVE) eisa_set_level_irq(gsi); } #endif #ifdef CONFIG_X86_IO_APIC if (acpi_irq_model == ACPI_IRQ_MODEL_IOAPIC) { - plat_gsi = mp_register_gsi(gsi, triggering, polarity); + plat_gsi = mp_register_gsi(dev, gsi, trigger, polarity); } #endif acpi_gsi_to_irq(plat_gsi, &irq); @@ -903,10 +904,8 @@ extern int es7000_plat; #endif static struct { - int apic_id; int gsi_base; int gsi_end; - DECLARE_BITMAP(pin_programmed, MP_MAX_IOAPIC_PIN + 1); } mp_ioapic_routing[MAX_IO_APICS]; int mp_find_ioapic(int gsi) @@ -986,16 +985,12 @@ void __init mp_register_ioapic(int id, u32 address, u32 gsi_base) set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address); mp_ioapics[idx].apicid = uniq_ioapic_id(id); -#ifdef CONFIG_X86_32 mp_ioapics[idx].apicver = io_apic_get_version(idx); -#else - mp_ioapics[idx].apicver = 0; -#endif + /* * Build basic GSI lookup table to facilitate gsi->io_apic lookups * and to prevent reprogramming of IOAPIC pins (PCI GSIs). */ - mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].apicid; mp_ioapic_routing[idx].gsi_base = gsi_base; mp_ioapic_routing[idx].gsi_end = gsi_base + io_apic_get_redir_entries(idx); @@ -1158,26 +1153,52 @@ void __init mp_config_acpi_legacy_irqs(void) } } -int mp_register_gsi(u32 gsi, int triggering, int polarity) +static int mp_config_acpi_gsi(struct device *dev, u32 gsi, int trigger, + int polarity) { +#ifdef CONFIG_X86_MPPARSE + struct mpc_intsrc mp_irq; + struct pci_dev *pdev; + unsigned char number; + unsigned int devfn; int ioapic; - int ioapic_pin; -#ifdef CONFIG_X86_32 -#define MAX_GSI_NUM 4096 -#define IRQ_COMPRESSION_START 64 + u8 pin; - static int pci_irq = IRQ_COMPRESSION_START; - /* - * Mapping between Global System Interrupts, which - * represent all possible interrupts, and IRQs - * assigned to actual devices. - */ - static int gsi_to_irq[MAX_GSI_NUM]; -#else + if (!acpi_ioapic) + return 0; + if (!dev) + return 0; + if (dev->bus != &pci_bus_type) + return 0; + + pdev = to_pci_dev(dev); + number = pdev->bus->number; + devfn = pdev->devfn; + pin = pdev->pin; + /* print the entry should happen on mptable identically */ + mp_irq.type = MP_INTSRC; + mp_irq.irqtype = mp_INT; + mp_irq.irqflag = (trigger == ACPI_EDGE_SENSITIVE ? 4 : 0x0c) | + (polarity == ACPI_ACTIVE_HIGH ? 1 : 3); + mp_irq.srcbus = number; + mp_irq.srcbusirq = (((devfn >> 3) & 0x1f) << 2) | ((pin - 1) & 3); + ioapic = mp_find_ioapic(gsi); + mp_irq.dstapic = mp_ioapics[ioapic].apicid; + mp_irq.dstirq = mp_find_ioapic_pin(ioapic, gsi); + + save_mp_irq(&mp_irq); +#endif + return 0; +} + +int mp_register_gsi(struct device *dev, u32 gsi, int trigger, int polarity) +{ + int ioapic; + int ioapic_pin; + struct io_apic_irq_attr irq_attr; if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC) return gsi; -#endif /* Don't set up the ACPI SCI because it's already set up */ if (acpi_gbl_FADT.sci_interrupt == gsi) @@ -1196,93 +1217,22 @@ int mp_register_gsi(u32 gsi, int triggering, int polarity) gsi = ioapic_renumber_irq(ioapic, gsi); #endif - /* - * Avoid pin reprogramming. PRTs typically include entries - * with redundant pin->gsi mappings (but unique PCI devices); - * we only program the IOAPIC on the first. - */ if (ioapic_pin > MP_MAX_IOAPIC_PIN) { printk(KERN_ERR "Invalid reference to IOAPIC pin " - "%d-%d\n", mp_ioapic_routing[ioapic].apic_id, + "%d-%d\n", mp_ioapics[ioapic].apicid, ioapic_pin); return gsi; } - if (test_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed)) { - pr_debug("Pin %d-%d already programmed\n", - mp_ioapic_routing[ioapic].apic_id, ioapic_pin); -#ifdef CONFIG_X86_32 - return (gsi < IRQ_COMPRESSION_START ? gsi : gsi_to_irq[gsi]); -#else - return gsi; -#endif - } - - set_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed); -#ifdef CONFIG_X86_32 - /* - * For GSI >= 64, use IRQ compression - */ - if ((gsi >= IRQ_COMPRESSION_START) - && (triggering == ACPI_LEVEL_SENSITIVE)) { - /* - * For PCI devices assign IRQs in order, avoiding gaps - * due to unused I/O APIC pins. - */ - int irq = gsi; - if (gsi < MAX_GSI_NUM) { - /* - * Retain the VIA chipset work-around (gsi > 15), but - * avoid a problem where the 8254 timer (IRQ0) is setup - * via an override (so it's not on pin 0 of the ioapic), - * and at the same time, the pin 0 interrupt is a PCI - * type. The gsi > 15 test could cause these two pins - * to be shared as IRQ0, and they are not shareable. - * So test for this condition, and if necessary, avoid - * the pin collision. - */ - gsi = pci_irq++; - /* - * Don't assign IRQ used by ACPI SCI - */ - if (gsi == acpi_gbl_FADT.sci_interrupt) - gsi = pci_irq++; - gsi_to_irq[irq] = gsi; - } else { - printk(KERN_ERR "GSI %u is too high\n", gsi); - return gsi; - } - } -#endif - io_apic_set_pci_routing(ioapic, ioapic_pin, gsi, - triggering == ACPI_EDGE_SENSITIVE ? 0 : 1, - polarity == ACPI_ACTIVE_HIGH ? 0 : 1); - return gsi; -} -int mp_config_acpi_gsi(unsigned char number, unsigned int devfn, u8 pin, - u32 gsi, int triggering, int polarity) -{ -#ifdef CONFIG_X86_MPPARSE - struct mpc_intsrc mp_irq; - int ioapic; + if (enable_update_mptable) + mp_config_acpi_gsi(dev, gsi, trigger, polarity); - if (!acpi_ioapic) - return 0; + set_io_apic_irq_attr(&irq_attr, ioapic, ioapic_pin, + trigger == ACPI_EDGE_SENSITIVE ? 0 : 1, + polarity == ACPI_ACTIVE_HIGH ? 0 : 1); + io_apic_set_pci_routing(dev, gsi, &irq_attr); - /* print the entry should happen on mptable identically */ - mp_irq.type = MP_INTSRC; - mp_irq.irqtype = mp_INT; - mp_irq.irqflag = (triggering == ACPI_EDGE_SENSITIVE ? 4 : 0x0c) | - (polarity == ACPI_ACTIVE_HIGH ? 1 : 3); - mp_irq.srcbus = number; - mp_irq.srcbusirq = (((devfn >> 3) & 0x1f) << 2) | ((pin - 1) & 3); - ioapic = mp_find_ioapic(gsi); - mp_irq.dstapic = mp_ioapic_routing[ioapic].apic_id; - mp_irq.dstirq = mp_find_ioapic_pin(ioapic, gsi); - - save_mp_irq(&mp_irq); -#endif - return 0; + return gsi; } /* diff --git a/arch/x86/kernel/acpi/realmode/Makefile b/arch/x86/kernel/acpi/realmode/Makefile index 1c31cc0e9de..167bc16ce0e 100644 --- a/arch/x86/kernel/acpi/realmode/Makefile +++ b/arch/x86/kernel/acpi/realmode/Makefile @@ -9,7 +9,7 @@ always := wakeup.bin targets := wakeup.elf wakeup.lds -wakeup-y += wakeup.o wakemain.o video-mode.o copy.o +wakeup-y += wakeup.o wakemain.o video-mode.o copy.o bioscall.o regs.o # The link order of the video-*.o modules can matter. In particular, # video-vga.o *must* be listed first, followed by video-vesa.o. diff --git a/arch/x86/kernel/acpi/realmode/bioscall.S b/arch/x86/kernel/acpi/realmode/bioscall.S new file mode 100644 index 00000000000..f51eb0bb56c --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/bioscall.S @@ -0,0 +1 @@ +#include "../../../boot/bioscall.S" diff --git a/arch/x86/kernel/acpi/realmode/regs.c b/arch/x86/kernel/acpi/realmode/regs.c new file mode 100644 index 00000000000..6206033ba20 --- /dev/null +++ b/arch/x86/kernel/acpi/realmode/regs.c @@ -0,0 +1 @@ +#include "../../../boot/regs.c" diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index a97db99dad5..1c60554537c 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -55,7 +55,16 @@ struct iommu_cmd { static int dma_ops_unity_map(struct dma_ops_domain *dma_dom, struct unity_map_entry *e); static struct dma_ops_domain *find_protection_domain(u16 devid); +static u64* alloc_pte(struct protection_domain *dom, + unsigned long address, u64 + **pte_page, gfp_t gfp); +static void dma_ops_reserve_addresses(struct dma_ops_domain *dom, + unsigned long start_page, + unsigned int pages); +#ifndef BUS_NOTIFY_UNBOUND_DRIVER +#define BUS_NOTIFY_UNBOUND_DRIVER 0x0005 +#endif #ifdef CONFIG_AMD_IOMMU_STATS @@ -213,7 +222,7 @@ irqreturn_t amd_iommu_int_handler(int irq, void *data) { struct amd_iommu *iommu; - list_for_each_entry(iommu, &amd_iommu_list, list) + for_each_iommu(iommu) iommu_poll_events(iommu); return IRQ_HANDLED; @@ -440,7 +449,7 @@ static void iommu_flush_domain(u16 domid) __iommu_build_inv_iommu_pages(&cmd, CMD_INV_IOMMU_ALL_PAGES_ADDRESS, domid, 1, 1); - list_for_each_entry(iommu, &amd_iommu_list, list) { + for_each_iommu(iommu) { spin_lock_irqsave(&iommu->lock, flags); __iommu_queue_command(iommu, &cmd); __iommu_completion_wait(iommu); @@ -449,6 +458,35 @@ static void iommu_flush_domain(u16 domid) } } +void amd_iommu_flush_all_domains(void) +{ + int i; + + for (i = 1; i < MAX_DOMAIN_ID; ++i) { + if (!test_bit(i, amd_iommu_pd_alloc_bitmap)) + continue; + iommu_flush_domain(i); + } +} + +void amd_iommu_flush_all_devices(void) +{ + struct amd_iommu *iommu; + int i; + + for (i = 0; i <= amd_iommu_last_bdf; ++i) { + if (amd_iommu_pd_table[i] == NULL) + continue; + + iommu = amd_iommu_rlookup_table[i]; + if (!iommu) + continue; + + iommu_queue_inv_dev_entry(iommu, i); + iommu_completion_wait(iommu); + } +} + /**************************************************************************** * * The functions below are used the create the page table mappings for @@ -468,7 +506,7 @@ static int iommu_map_page(struct protection_domain *dom, unsigned long phys_addr, int prot) { - u64 __pte, *pte, *page; + u64 __pte, *pte; bus_addr = PAGE_ALIGN(bus_addr); phys_addr = PAGE_ALIGN(phys_addr); @@ -477,27 +515,7 @@ static int iommu_map_page(struct protection_domain *dom, if (bus_addr > IOMMU_MAP_SIZE_L3 || !(prot & IOMMU_PROT_MASK)) return -EINVAL; - pte = &dom->pt_root[IOMMU_PTE_L2_INDEX(bus_addr)]; - - if (!IOMMU_PTE_PRESENT(*pte)) { - page = (u64 *)get_zeroed_page(GFP_KERNEL); - if (!page) - return -ENOMEM; - *pte = IOMMU_L2_PDE(virt_to_phys(page)); - } - - pte = IOMMU_PTE_PAGE(*pte); - pte = &pte[IOMMU_PTE_L1_INDEX(bus_addr)]; - - if (!IOMMU_PTE_PRESENT(*pte)) { - page = (u64 *)get_zeroed_page(GFP_KERNEL); - if (!page) - return -ENOMEM; - *pte = IOMMU_L1_PDE(virt_to_phys(page)); - } - - pte = IOMMU_PTE_PAGE(*pte); - pte = &pte[IOMMU_PTE_L0_INDEX(bus_addr)]; + pte = alloc_pte(dom, bus_addr, NULL, GFP_KERNEL); if (IOMMU_PTE_PRESENT(*pte)) return -EBUSY; @@ -595,7 +613,8 @@ static int dma_ops_unity_map(struct dma_ops_domain *dma_dom, * as allocated in the aperture */ if (addr < dma_dom->aperture_size) - __set_bit(addr >> PAGE_SHIFT, dma_dom->bitmap); + __set_bit(addr >> PAGE_SHIFT, + dma_dom->aperture[0]->bitmap); } return 0; @@ -632,42 +651,191 @@ static int init_unity_mappings_for_device(struct dma_ops_domain *dma_dom, ****************************************************************************/ /* - * The address allocator core function. + * The address allocator core functions. * * called with domain->lock held */ + +/* + * This function checks if there is a PTE for a given dma address. If + * there is one, it returns the pointer to it. + */ +static u64* fetch_pte(struct protection_domain *domain, + unsigned long address) +{ + u64 *pte; + + pte = &domain->pt_root[IOMMU_PTE_L2_INDEX(address)]; + + if (!IOMMU_PTE_PRESENT(*pte)) + return NULL; + + pte = IOMMU_PTE_PAGE(*pte); + pte = &pte[IOMMU_PTE_L1_INDEX(address)]; + + if (!IOMMU_PTE_PRESENT(*pte)) + return NULL; + + pte = IOMMU_PTE_PAGE(*pte); + pte = &pte[IOMMU_PTE_L0_INDEX(address)]; + + return pte; +} + +/* + * This function is used to add a new aperture range to an existing + * aperture in case of dma_ops domain allocation or address allocation + * failure. + */ +static int alloc_new_range(struct amd_iommu *iommu, + struct dma_ops_domain *dma_dom, + bool populate, gfp_t gfp) +{ + int index = dma_dom->aperture_size >> APERTURE_RANGE_SHIFT; + int i; + +#ifdef CONFIG_IOMMU_STRESS + populate = false; +#endif + + if (index >= APERTURE_MAX_RANGES) + return -ENOMEM; + + dma_dom->aperture[index] = kzalloc(sizeof(struct aperture_range), gfp); + if (!dma_dom->aperture[index]) + return -ENOMEM; + + dma_dom->aperture[index]->bitmap = (void *)get_zeroed_page(gfp); + if (!dma_dom->aperture[index]->bitmap) + goto out_free; + + dma_dom->aperture[index]->offset = dma_dom->aperture_size; + + if (populate) { + unsigned long address = dma_dom->aperture_size; + int i, num_ptes = APERTURE_RANGE_PAGES / 512; + u64 *pte, *pte_page; + + for (i = 0; i < num_ptes; ++i) { + pte = alloc_pte(&dma_dom->domain, address, + &pte_page, gfp); + if (!pte) + goto out_free; + + dma_dom->aperture[index]->pte_pages[i] = pte_page; + + address += APERTURE_RANGE_SIZE / 64; + } + } + + dma_dom->aperture_size += APERTURE_RANGE_SIZE; + + /* Intialize the exclusion range if necessary */ + if (iommu->exclusion_start && + iommu->exclusion_start >= dma_dom->aperture[index]->offset && + iommu->exclusion_start < dma_dom->aperture_size) { + unsigned long startpage = iommu->exclusion_start >> PAGE_SHIFT; + int pages = iommu_num_pages(iommu->exclusion_start, + iommu->exclusion_length, + PAGE_SIZE); + dma_ops_reserve_addresses(dma_dom, startpage, pages); + } + + /* + * Check for areas already mapped as present in the new aperture + * range and mark those pages as reserved in the allocator. Such + * mappings may already exist as a result of requested unity + * mappings for devices. + */ + for (i = dma_dom->aperture[index]->offset; + i < dma_dom->aperture_size; + i += PAGE_SIZE) { + u64 *pte = fetch_pte(&dma_dom->domain, i); + if (!pte || !IOMMU_PTE_PRESENT(*pte)) + continue; + + dma_ops_reserve_addresses(dma_dom, i << PAGE_SHIFT, 1); + } + + return 0; + +out_free: + free_page((unsigned long)dma_dom->aperture[index]->bitmap); + + kfree(dma_dom->aperture[index]); + dma_dom->aperture[index] = NULL; + + return -ENOMEM; +} + +static unsigned long dma_ops_area_alloc(struct device *dev, + struct dma_ops_domain *dom, + unsigned int pages, + unsigned long align_mask, + u64 dma_mask, + unsigned long start) +{ + unsigned long next_bit = dom->next_address % APERTURE_RANGE_SIZE; + int max_index = dom->aperture_size >> APERTURE_RANGE_SHIFT; + int i = start >> APERTURE_RANGE_SHIFT; + unsigned long boundary_size; + unsigned long address = -1; + unsigned long limit; + + next_bit >>= PAGE_SHIFT; + + boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1, + PAGE_SIZE) >> PAGE_SHIFT; + + for (;i < max_index; ++i) { + unsigned long offset = dom->aperture[i]->offset >> PAGE_SHIFT; + + if (dom->aperture[i]->offset >= dma_mask) + break; + + limit = iommu_device_max_index(APERTURE_RANGE_PAGES, offset, + dma_mask >> PAGE_SHIFT); + + address = iommu_area_alloc(dom->aperture[i]->bitmap, + limit, next_bit, pages, 0, + boundary_size, align_mask); + if (address != -1) { + address = dom->aperture[i]->offset + + (address << PAGE_SHIFT); + dom->next_address = address + (pages << PAGE_SHIFT); + break; + } + + next_bit = 0; + } + + return address; +} + static unsigned long dma_ops_alloc_addresses(struct device *dev, struct dma_ops_domain *dom, unsigned int pages, unsigned long align_mask, u64 dma_mask) { - unsigned long limit; unsigned long address; - unsigned long boundary_size; - boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1, - PAGE_SIZE) >> PAGE_SHIFT; - limit = iommu_device_max_index(dom->aperture_size >> PAGE_SHIFT, 0, - dma_mask >> PAGE_SHIFT); +#ifdef CONFIG_IOMMU_STRESS + dom->next_address = 0; + dom->need_flush = true; +#endif - if (dom->next_bit >= limit) { - dom->next_bit = 0; - dom->need_flush = true; - } + address = dma_ops_area_alloc(dev, dom, pages, align_mask, + dma_mask, dom->next_address); - address = iommu_area_alloc(dom->bitmap, limit, dom->next_bit, pages, - 0 , boundary_size, align_mask); if (address == -1) { - address = iommu_area_alloc(dom->bitmap, limit, 0, pages, - 0, boundary_size, align_mask); + dom->next_address = 0; + address = dma_ops_area_alloc(dev, dom, pages, align_mask, + dma_mask, 0); dom->need_flush = true; } - if (likely(address != -1)) { - dom->next_bit = address + pages; - address <<= PAGE_SHIFT; - } else + if (unlikely(address == -1)) address = bad_dma_address; WARN_ON((address + (PAGE_SIZE*pages)) > dom->aperture_size); @@ -684,11 +852,23 @@ static void dma_ops_free_addresses(struct dma_ops_domain *dom, unsigned long address, unsigned int pages) { - address >>= PAGE_SHIFT; - iommu_area_free(dom->bitmap, address, pages); + unsigned i = address >> APERTURE_RANGE_SHIFT; + struct aperture_range *range = dom->aperture[i]; - if (address >= dom->next_bit) + BUG_ON(i >= APERTURE_MAX_RANGES || range == NULL); + +#ifdef CONFIG_IOMMU_STRESS + if (i < 4) + return; +#endif + + if (address >= dom->next_address) dom->need_flush = true; + + address = (address % APERTURE_RANGE_SIZE) >> PAGE_SHIFT; + + iommu_area_free(range->bitmap, address, pages); + } /**************************************************************************** @@ -736,12 +916,16 @@ static void dma_ops_reserve_addresses(struct dma_ops_domain *dom, unsigned long start_page, unsigned int pages) { - unsigned int last_page = dom->aperture_size >> PAGE_SHIFT; + unsigned int i, last_page = dom->aperture_size >> PAGE_SHIFT; if (start_page + pages > last_page) pages = last_page - start_page; - iommu_area_reserve(dom->bitmap, start_page, pages); + for (i = start_page; i < start_page + pages; ++i) { + int index = i / APERTURE_RANGE_PAGES; + int page = i % APERTURE_RANGE_PAGES; + __set_bit(page, dom->aperture[index]->bitmap); + } } static void free_pagetable(struct protection_domain *domain) @@ -780,14 +964,19 @@ static void free_pagetable(struct protection_domain *domain) */ static void dma_ops_domain_free(struct dma_ops_domain *dom) { + int i; + if (!dom) return; free_pagetable(&dom->domain); - kfree(dom->pte_pages); - - kfree(dom->bitmap); + for (i = 0; i < APERTURE_MAX_RANGES; ++i) { + if (!dom->aperture[i]) + continue; + free_page((unsigned long)dom->aperture[i]->bitmap); + kfree(dom->aperture[i]); + } kfree(dom); } @@ -797,19 +986,9 @@ static void dma_ops_domain_free(struct dma_ops_domain *dom) * It also intializes the page table and the address allocator data * structures required for the dma_ops interface */ -static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu, - unsigned order) +static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu) { struct dma_ops_domain *dma_dom; - unsigned i, num_pte_pages; - u64 *l2_pde; - u64 address; - - /* - * Currently the DMA aperture must be between 32 MB and 1GB in size - */ - if ((order < 25) || (order > 30)) - return NULL; dma_dom = kzalloc(sizeof(struct dma_ops_domain), GFP_KERNEL); if (!dma_dom) @@ -826,55 +1005,20 @@ static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu, dma_dom->domain.priv = dma_dom; if (!dma_dom->domain.pt_root) goto free_dma_dom; - dma_dom->aperture_size = (1ULL << order); - dma_dom->bitmap = kzalloc(dma_dom->aperture_size / (PAGE_SIZE * 8), - GFP_KERNEL); - if (!dma_dom->bitmap) - goto free_dma_dom; - /* - * mark the first page as allocated so we never return 0 as - * a valid dma-address. So we can use 0 as error value - */ - dma_dom->bitmap[0] = 1; - dma_dom->next_bit = 0; dma_dom->need_flush = false; dma_dom->target_dev = 0xffff; - /* Intialize the exclusion range if necessary */ - if (iommu->exclusion_start && - iommu->exclusion_start < dma_dom->aperture_size) { - unsigned long startpage = iommu->exclusion_start >> PAGE_SHIFT; - int pages = iommu_num_pages(iommu->exclusion_start, - iommu->exclusion_length, - PAGE_SIZE); - dma_ops_reserve_addresses(dma_dom, startpage, pages); - } + if (alloc_new_range(iommu, dma_dom, true, GFP_KERNEL)) + goto free_dma_dom; /* - * At the last step, build the page tables so we don't need to - * allocate page table pages in the dma_ops mapping/unmapping - * path. + * mark the first page as allocated so we never return 0 as + * a valid dma-address. So we can use 0 as error value */ - num_pte_pages = dma_dom->aperture_size / (PAGE_SIZE * 512); - dma_dom->pte_pages = kzalloc(num_pte_pages * sizeof(void *), - GFP_KERNEL); - if (!dma_dom->pte_pages) - goto free_dma_dom; - - l2_pde = (u64 *)get_zeroed_page(GFP_KERNEL); - if (l2_pde == NULL) - goto free_dma_dom; + dma_dom->aperture[0]->bitmap[0] = 1; + dma_dom->next_address = 0; - dma_dom->domain.pt_root[0] = IOMMU_L2_PDE(virt_to_phys(l2_pde)); - - for (i = 0; i < num_pte_pages; ++i) { - dma_dom->pte_pages[i] = (u64 *)get_zeroed_page(GFP_KERNEL); - if (!dma_dom->pte_pages[i]) - goto free_dma_dom; - address = virt_to_phys(dma_dom->pte_pages[i]); - l2_pde[i] = IOMMU_L1_PDE(address); - } return dma_dom; @@ -983,7 +1127,6 @@ static int device_change_notifier(struct notifier_block *nb, struct protection_domain *domain; struct dma_ops_domain *dma_domain; struct amd_iommu *iommu; - int order = amd_iommu_aperture_order; unsigned long flags; if (devid > amd_iommu_last_bdf) @@ -1002,17 +1145,7 @@ static int device_change_notifier(struct notifier_block *nb, "to a non-dma-ops domain\n", dev_name(dev)); switch (action) { - case BUS_NOTIFY_BOUND_DRIVER: - if (domain) - goto out; - dma_domain = find_protection_domain(devid); - if (!dma_domain) - dma_domain = iommu->default_dom; - attach_device(iommu, &dma_domain->domain, devid); - printk(KERN_INFO "AMD IOMMU: Using protection domain %d for " - "device %s\n", dma_domain->domain.id, dev_name(dev)); - break; - case BUS_NOTIFY_UNBIND_DRIVER: + case BUS_NOTIFY_UNBOUND_DRIVER: if (!domain) goto out; detach_device(domain, devid); @@ -1022,7 +1155,7 @@ static int device_change_notifier(struct notifier_block *nb, dma_domain = find_protection_domain(devid); if (dma_domain) goto out; - dma_domain = dma_ops_domain_alloc(iommu, order); + dma_domain = dma_ops_domain_alloc(iommu); if (!dma_domain) goto out; dma_domain->target_dev = devid; @@ -1133,8 +1266,8 @@ static int get_device_resources(struct device *dev, dma_dom = (*iommu)->default_dom; *domain = &dma_dom->domain; attach_device(*iommu, *domain, *bdf); - printk(KERN_INFO "AMD IOMMU: Using protection domain %d for " - "device %s\n", (*domain)->id, dev_name(dev)); + DUMP_printk("Using protection domain %d for device %s\n", + (*domain)->id, dev_name(dev)); } if (domain_for_device(_bdf) == NULL) @@ -1144,6 +1277,66 @@ static int get_device_resources(struct device *dev, } /* + * If the pte_page is not yet allocated this function is called + */ +static u64* alloc_pte(struct protection_domain *dom, + unsigned long address, u64 **pte_page, gfp_t gfp) +{ + u64 *pte, *page; + + pte = &dom->pt_root[IOMMU_PTE_L2_INDEX(address)]; + + if (!IOMMU_PTE_PRESENT(*pte)) { + page = (u64 *)get_zeroed_page(gfp); + if (!page) + return NULL; + *pte = IOMMU_L2_PDE(virt_to_phys(page)); + } + + pte = IOMMU_PTE_PAGE(*pte); + pte = &pte[IOMMU_PTE_L1_INDEX(address)]; + + if (!IOMMU_PTE_PRESENT(*pte)) { + page = (u64 *)get_zeroed_page(gfp); + if (!page) + return NULL; + *pte = IOMMU_L1_PDE(virt_to_phys(page)); + } + + pte = IOMMU_PTE_PAGE(*pte); + + if (pte_page) + *pte_page = pte; + + pte = &pte[IOMMU_PTE_L0_INDEX(address)]; + + return pte; +} + +/* + * This function fetches the PTE for a given address in the aperture + */ +static u64* dma_ops_get_pte(struct dma_ops_domain *dom, + unsigned long address) +{ + struct aperture_range *aperture; + u64 *pte, *pte_page; + + aperture = dom->aperture[APERTURE_RANGE_INDEX(address)]; + if (!aperture) + return NULL; + + pte = aperture->pte_pages[APERTURE_PAGE_INDEX(address)]; + if (!pte) { + pte = alloc_pte(&dom->domain, address, &pte_page, GFP_ATOMIC); + aperture->pte_pages[APERTURE_PAGE_INDEX(address)] = pte_page; + } else + pte += IOMMU_PTE_L0_INDEX(address); + + return pte; +} + +/* * This is the generic map function. It maps one 4kb page at paddr to * the given address in the DMA address space for the domain. */ @@ -1159,8 +1352,9 @@ static dma_addr_t dma_ops_domain_map(struct amd_iommu *iommu, paddr &= PAGE_MASK; - pte = dom->pte_pages[IOMMU_PTE_L1_INDEX(address)]; - pte += IOMMU_PTE_L0_INDEX(address); + pte = dma_ops_get_pte(dom, address); + if (!pte) + return bad_dma_address; __pte = paddr | IOMMU_PTE_P | IOMMU_PTE_FC; @@ -1185,14 +1379,20 @@ static void dma_ops_domain_unmap(struct amd_iommu *iommu, struct dma_ops_domain *dom, unsigned long address) { + struct aperture_range *aperture; u64 *pte; if (address >= dom->aperture_size) return; - WARN_ON(address & ~PAGE_MASK || address >= dom->aperture_size); + aperture = dom->aperture[APERTURE_RANGE_INDEX(address)]; + if (!aperture) + return; + + pte = aperture->pte_pages[APERTURE_PAGE_INDEX(address)]; + if (!pte) + return; - pte = dom->pte_pages[IOMMU_PTE_L1_INDEX(address)]; pte += IOMMU_PTE_L0_INDEX(address); WARN_ON(!*pte); @@ -1216,7 +1416,7 @@ static dma_addr_t __map_single(struct device *dev, u64 dma_mask) { dma_addr_t offset = paddr & ~PAGE_MASK; - dma_addr_t address, start; + dma_addr_t address, start, ret; unsigned int pages; unsigned long align_mask = 0; int i; @@ -1232,14 +1432,33 @@ static dma_addr_t __map_single(struct device *dev, if (align) align_mask = (1UL << get_order(size)) - 1; +retry: address = dma_ops_alloc_addresses(dev, dma_dom, pages, align_mask, dma_mask); - if (unlikely(address == bad_dma_address)) - goto out; + if (unlikely(address == bad_dma_address)) { + /* + * setting next_address here will let the address + * allocator only scan the new allocated range in the + * first run. This is a small optimization. + */ + dma_dom->next_address = dma_dom->aperture_size; + + if (alloc_new_range(iommu, dma_dom, false, GFP_ATOMIC)) + goto out; + + /* + * aperture was sucessfully enlarged by 128 MB, try + * allocation again + */ + goto retry; + } start = address; for (i = 0; i < pages; ++i) { - dma_ops_domain_map(iommu, dma_dom, start, paddr, dir); + ret = dma_ops_domain_map(iommu, dma_dom, start, paddr, dir); + if (ret == bad_dma_address) + goto out_unmap; + paddr += PAGE_SIZE; start += PAGE_SIZE; } @@ -1255,6 +1474,17 @@ static dma_addr_t __map_single(struct device *dev, out: return address; + +out_unmap: + + for (--i; i >= 0; --i) { + start -= PAGE_SIZE; + dma_ops_domain_unmap(iommu, dma_dom, start); + } + + dma_ops_free_addresses(dma_dom, address, pages); + + return bad_dma_address; } /* @@ -1537,8 +1767,10 @@ static void *alloc_coherent(struct device *dev, size_t size, *dma_addr = __map_single(dev, iommu, domain->priv, paddr, size, DMA_BIDIRECTIONAL, true, dma_mask); - if (*dma_addr == bad_dma_address) + if (*dma_addr == bad_dma_address) { + spin_unlock_irqrestore(&domain->lock, flags); goto out_free; + } iommu_completion_wait(iommu); @@ -1625,7 +1857,6 @@ static void prealloc_protection_domains(void) struct pci_dev *dev = NULL; struct dma_ops_domain *dma_dom; struct amd_iommu *iommu; - int order = amd_iommu_aperture_order; u16 devid; while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) { @@ -1638,7 +1869,7 @@ static void prealloc_protection_domains(void) iommu = amd_iommu_rlookup_table[devid]; if (!iommu) continue; - dma_dom = dma_ops_domain_alloc(iommu, order); + dma_dom = dma_ops_domain_alloc(iommu); if (!dma_dom) continue; init_unity_mappings_for_device(dma_dom, devid); @@ -1664,7 +1895,6 @@ static struct dma_map_ops amd_iommu_dma_ops = { int __init amd_iommu_init_dma_ops(void) { struct amd_iommu *iommu; - int order = amd_iommu_aperture_order; int ret; /* @@ -1672,8 +1902,8 @@ int __init amd_iommu_init_dma_ops(void) * found in the system. Devices not assigned to any other * protection domain will be assigned to the default one. */ - list_for_each_entry(iommu, &amd_iommu_list, list) { - iommu->default_dom = dma_ops_domain_alloc(iommu, order); + for_each_iommu(iommu) { + iommu->default_dom = dma_ops_domain_alloc(iommu); if (iommu->default_dom == NULL) return -ENOMEM; iommu->default_dom->domain.flags |= PD_DEFAULT_MASK; @@ -1710,7 +1940,7 @@ int __init amd_iommu_init_dma_ops(void) free_domains: - list_for_each_entry(iommu, &amd_iommu_list, list) { + for_each_iommu(iommu) { if (iommu->default_dom) dma_ops_domain_free(iommu->default_dom); } @@ -1842,7 +2072,7 @@ static int amd_iommu_attach_device(struct iommu_domain *dom, old_domain = domain_for_device(devid); if (old_domain) - return -EBUSY; + detach_device(old_domain, devid); attach_device(iommu, domain, devid); diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c index 8c0be0902da..238989ec077 100644 --- a/arch/x86/kernel/amd_iommu_init.c +++ b/arch/x86/kernel/amd_iommu_init.c @@ -115,15 +115,21 @@ struct ivmd_header { u64 range_length; } __attribute__((packed)); +bool amd_iommu_dump; + static int __initdata amd_iommu_detected; u16 amd_iommu_last_bdf; /* largest PCI device id we have to handle */ LIST_HEAD(amd_iommu_unity_map); /* a list of required unity mappings we find in ACPI */ -unsigned amd_iommu_aperture_order = 26; /* size of aperture in power of 2 */ +#ifdef CONFIG_IOMMU_STRESS +bool amd_iommu_isolate = false; +#else bool amd_iommu_isolate = true; /* if true, device isolation is enabled */ +#endif + bool amd_iommu_unmap_flush; /* if true, flush on every unmap */ LIST_HEAD(amd_iommu_list); /* list of all AMD IOMMUs in the @@ -175,7 +181,7 @@ static inline void update_last_devid(u16 devid) static inline unsigned long tbl_size(int entry_size) { unsigned shift = PAGE_SHIFT + - get_order(amd_iommu_last_bdf * entry_size); + get_order(((int)amd_iommu_last_bdf + 1) * entry_size); return 1UL << shift; } @@ -193,7 +199,7 @@ static inline unsigned long tbl_size(int entry_size) * This function set the exclusion range in the IOMMU. DMA accesses to the * exclusion range are passed through untranslated */ -static void __init iommu_set_exclusion_range(struct amd_iommu *iommu) +static void iommu_set_exclusion_range(struct amd_iommu *iommu) { u64 start = iommu->exclusion_start & PAGE_MASK; u64 limit = (start + iommu->exclusion_length) & PAGE_MASK; @@ -225,7 +231,7 @@ static void __init iommu_set_device_table(struct amd_iommu *iommu) } /* Generic functions to enable/disable certain features of the IOMMU. */ -static void __init iommu_feature_enable(struct amd_iommu *iommu, u8 bit) +static void iommu_feature_enable(struct amd_iommu *iommu, u8 bit) { u32 ctrl; @@ -244,7 +250,7 @@ static void __init iommu_feature_disable(struct amd_iommu *iommu, u8 bit) } /* Function to enable the hardware */ -static void __init iommu_enable(struct amd_iommu *iommu) +static void iommu_enable(struct amd_iommu *iommu) { printk(KERN_INFO "AMD IOMMU: Enabling IOMMU at %s cap 0x%hx\n", dev_name(&iommu->dev->dev), iommu->cap_ptr); @@ -252,11 +258,9 @@ static void __init iommu_enable(struct amd_iommu *iommu) iommu_feature_enable(iommu, CONTROL_IOMMU_EN); } -/* Function to enable IOMMU event logging and event interrupts */ -static void __init iommu_enable_event_logging(struct amd_iommu *iommu) +static void iommu_disable(struct amd_iommu *iommu) { - iommu_feature_enable(iommu, CONTROL_EVT_LOG_EN); - iommu_feature_enable(iommu, CONTROL_EVT_INT_EN); + iommu_feature_disable(iommu, CONTROL_IOMMU_EN); } /* @@ -413,25 +417,36 @@ static u8 * __init alloc_command_buffer(struct amd_iommu *iommu) { u8 *cmd_buf = (u8 *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, get_order(CMD_BUFFER_SIZE)); - u64 entry; if (cmd_buf == NULL) return NULL; iommu->cmd_buf_size = CMD_BUFFER_SIZE; - entry = (u64)virt_to_phys(cmd_buf); + return cmd_buf; +} + +/* + * This function writes the command buffer address to the hardware and + * enables it. + */ +static void iommu_enable_command_buffer(struct amd_iommu *iommu) +{ + u64 entry; + + BUG_ON(iommu->cmd_buf == NULL); + + entry = (u64)virt_to_phys(iommu->cmd_buf); entry |= MMIO_CMD_SIZE_512; + memcpy_toio(iommu->mmio_base + MMIO_CMD_BUF_OFFSET, - &entry, sizeof(entry)); + &entry, sizeof(entry)); /* set head and tail to zero manually */ writel(0x00, iommu->mmio_base + MMIO_CMD_HEAD_OFFSET); writel(0x00, iommu->mmio_base + MMIO_CMD_TAIL_OFFSET); iommu_feature_enable(iommu, CONTROL_CMDBUF_EN); - - return cmd_buf; } static void __init free_command_buffer(struct amd_iommu *iommu) @@ -443,20 +458,27 @@ static void __init free_command_buffer(struct amd_iommu *iommu) /* allocates the memory where the IOMMU will log its events to */ static u8 * __init alloc_event_buffer(struct amd_iommu *iommu) { - u64 entry; iommu->evt_buf = (u8 *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, get_order(EVT_BUFFER_SIZE)); if (iommu->evt_buf == NULL) return NULL; + return iommu->evt_buf; +} + +static void iommu_enable_event_buffer(struct amd_iommu *iommu) +{ + u64 entry; + + BUG_ON(iommu->evt_buf == NULL); + entry = (u64)virt_to_phys(iommu->evt_buf) | EVT_LEN_MASK; + memcpy_toio(iommu->mmio_base + MMIO_EVT_BUF_OFFSET, &entry, sizeof(entry)); - iommu->evt_buf_size = EVT_BUFFER_SIZE; - - return iommu->evt_buf; + iommu_feature_enable(iommu, CONTROL_EVT_LOG_EN); } static void __init free_event_buffer(struct amd_iommu *iommu) @@ -596,32 +618,83 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu, p += sizeof(struct ivhd_header); end += h->length; + while (p < end) { e = (struct ivhd_entry *)p; switch (e->type) { case IVHD_DEV_ALL: + + DUMP_printk(" DEV_ALL\t\t\t first devid: %02x:%02x.%x" + " last device %02x:%02x.%x flags: %02x\n", + PCI_BUS(iommu->first_device), + PCI_SLOT(iommu->first_device), + PCI_FUNC(iommu->first_device), + PCI_BUS(iommu->last_device), + PCI_SLOT(iommu->last_device), + PCI_FUNC(iommu->last_device), + e->flags); + for (dev_i = iommu->first_device; dev_i <= iommu->last_device; ++dev_i) set_dev_entry_from_acpi(iommu, dev_i, e->flags, 0); break; case IVHD_DEV_SELECT: + + DUMP_printk(" DEV_SELECT\t\t\t devid: %02x:%02x.%x " + "flags: %02x\n", + PCI_BUS(e->devid), + PCI_SLOT(e->devid), + PCI_FUNC(e->devid), + e->flags); + devid = e->devid; set_dev_entry_from_acpi(iommu, devid, e->flags, 0); break; case IVHD_DEV_SELECT_RANGE_START: + + DUMP_printk(" DEV_SELECT_RANGE_START\t " + "devid: %02x:%02x.%x flags: %02x\n", + PCI_BUS(e->devid), + PCI_SLOT(e->devid), + PCI_FUNC(e->devid), + e->flags); + devid_start = e->devid; flags = e->flags; ext_flags = 0; alias = false; break; case IVHD_DEV_ALIAS: + + DUMP_printk(" DEV_ALIAS\t\t\t devid: %02x:%02x.%x " + "flags: %02x devid_to: %02x:%02x.%x\n", + PCI_BUS(e->devid), + PCI_SLOT(e->devid), + PCI_FUNC(e->devid), + e->flags, + PCI_BUS(e->ext >> 8), + PCI_SLOT(e->ext >> 8), + PCI_FUNC(e->ext >> 8)); + devid = e->devid; devid_to = e->ext >> 8; - set_dev_entry_from_acpi(iommu, devid, e->flags, 0); + set_dev_entry_from_acpi(iommu, devid_to, e->flags, 0); amd_iommu_alias_table[devid] = devid_to; break; case IVHD_DEV_ALIAS_RANGE: + + DUMP_printk(" DEV_ALIAS_RANGE\t\t " + "devid: %02x:%02x.%x flags: %02x " + "devid_to: %02x:%02x.%x\n", + PCI_BUS(e->devid), + PCI_SLOT(e->devid), + PCI_FUNC(e->devid), + e->flags, + PCI_BUS(e->ext >> 8), + PCI_SLOT(e->ext >> 8), + PCI_FUNC(e->ext >> 8)); + devid_start = e->devid; flags = e->flags; devid_to = e->ext >> 8; @@ -629,17 +702,39 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu, alias = true; break; case IVHD_DEV_EXT_SELECT: + + DUMP_printk(" DEV_EXT_SELECT\t\t devid: %02x:%02x.%x " + "flags: %02x ext: %08x\n", + PCI_BUS(e->devid), + PCI_SLOT(e->devid), + PCI_FUNC(e->devid), + e->flags, e->ext); + devid = e->devid; set_dev_entry_from_acpi(iommu, devid, e->flags, e->ext); break; case IVHD_DEV_EXT_SELECT_RANGE: + + DUMP_printk(" DEV_EXT_SELECT_RANGE\t devid: " + "%02x:%02x.%x flags: %02x ext: %08x\n", + PCI_BUS(e->devid), + PCI_SLOT(e->devid), + PCI_FUNC(e->devid), + e->flags, e->ext); + devid_start = e->devid; flags = e->flags; ext_flags = e->ext; alias = false; break; case IVHD_DEV_RANGE_END: + + DUMP_printk(" DEV_RANGE_END\t\t devid: %02x:%02x.%x\n", + PCI_BUS(e->devid), + PCI_SLOT(e->devid), + PCI_FUNC(e->devid)); + devid = e->devid; for (dev_i = devid_start; dev_i <= devid; ++dev_i) { if (alias) @@ -679,7 +774,7 @@ static void __init free_iommu_all(void) { struct amd_iommu *iommu, *next; - list_for_each_entry_safe(iommu, next, &amd_iommu_list, list) { + for_each_iommu_safe(iommu, next) { list_del(&iommu->list); free_iommu_one(iommu); kfree(iommu); @@ -710,7 +805,6 @@ static int __init init_iommu_one(struct amd_iommu *iommu, struct ivhd_header *h) if (!iommu->mmio_base) return -ENOMEM; - iommu_set_device_table(iommu); iommu->cmd_buf = alloc_command_buffer(iommu); if (!iommu->cmd_buf) return -ENOMEM; @@ -746,6 +840,15 @@ static int __init init_iommu_all(struct acpi_table_header *table) h = (struct ivhd_header *)p; switch (*p) { case ACPI_IVHD_TYPE: + + DUMP_printk("IOMMU: device: %02x:%02x.%01x cap: %04x " + "seg: %d flags: %01x info %04x\n", + PCI_BUS(h->devid), PCI_SLOT(h->devid), + PCI_FUNC(h->devid), h->cap_ptr, + h->pci_seg, h->flags, h->info); + DUMP_printk(" mmio-addr: %016llx\n", + h->mmio_phys); + iommu = kzalloc(sizeof(struct amd_iommu), GFP_KERNEL); if (iommu == NULL) return -ENOMEM; @@ -773,56 +876,9 @@ static int __init init_iommu_all(struct acpi_table_header *table) * ****************************************************************************/ -static int __init iommu_setup_msix(struct amd_iommu *iommu) -{ - struct amd_iommu *curr; - struct msix_entry entries[32]; /* only 32 supported by AMD IOMMU */ - int nvec = 0, i; - - list_for_each_entry(curr, &amd_iommu_list, list) { - if (curr->dev == iommu->dev) { - entries[nvec].entry = curr->evt_msi_num; - entries[nvec].vector = 0; - curr->int_enabled = true; - nvec++; - } - } - - if (pci_enable_msix(iommu->dev, entries, nvec)) { - pci_disable_msix(iommu->dev); - return 1; - } - - for (i = 0; i < nvec; ++i) { - int r = request_irq(entries->vector, amd_iommu_int_handler, - IRQF_SAMPLE_RANDOM, - "AMD IOMMU", - NULL); - if (r) - goto out_free; - } - - return 0; - -out_free: - for (i -= 1; i >= 0; --i) - free_irq(entries->vector, NULL); - - pci_disable_msix(iommu->dev); - - return 1; -} - static int __init iommu_setup_msi(struct amd_iommu *iommu) { int r; - struct amd_iommu *curr; - - list_for_each_entry(curr, &amd_iommu_list, list) { - if (curr->dev == iommu->dev) - curr->int_enabled = true; - } - if (pci_enable_msi(iommu->dev)) return 1; @@ -837,17 +893,18 @@ static int __init iommu_setup_msi(struct amd_iommu *iommu) return 1; } + iommu->int_enabled = true; + iommu_feature_enable(iommu, CONTROL_EVT_INT_EN); + return 0; } -static int __init iommu_init_msi(struct amd_iommu *iommu) +static int iommu_init_msi(struct amd_iommu *iommu) { if (iommu->int_enabled) return 0; - if (pci_find_capability(iommu->dev, PCI_CAP_ID_MSIX)) - return iommu_setup_msix(iommu); - else if (pci_find_capability(iommu->dev, PCI_CAP_ID_MSI)) + if (pci_find_capability(iommu->dev, PCI_CAP_ID_MSI)) return iommu_setup_msi(iommu); return 1; @@ -899,6 +956,7 @@ static int __init init_exclusion_range(struct ivmd_header *m) static int __init init_unity_map_range(struct ivmd_header *m) { struct unity_map_entry *e = 0; + char *s; e = kzalloc(sizeof(*e), GFP_KERNEL); if (e == NULL) @@ -906,14 +964,19 @@ static int __init init_unity_map_range(struct ivmd_header *m) switch (m->type) { default: + kfree(e); + return 0; case ACPI_IVMD_TYPE: + s = "IVMD_TYPEi\t\t\t"; e->devid_start = e->devid_end = m->devid; break; case ACPI_IVMD_TYPE_ALL: + s = "IVMD_TYPE_ALL\t\t"; e->devid_start = 0; e->devid_end = amd_iommu_last_bdf; break; case ACPI_IVMD_TYPE_RANGE: + s = "IVMD_TYPE_RANGE\t\t"; e->devid_start = m->devid; e->devid_end = m->aux; break; @@ -922,6 +985,13 @@ static int __init init_unity_map_range(struct ivmd_header *m) e->address_end = e->address_start + PAGE_ALIGN(m->range_length); e->prot = m->flags >> 1; + DUMP_printk("%s devid_start: %02x:%02x.%x devid_end: %02x:%02x.%x" + " range_start: %016llx range_end: %016llx flags: %x\n", s, + PCI_BUS(e->devid_start), PCI_SLOT(e->devid_start), + PCI_FUNC(e->devid_start), PCI_BUS(e->devid_end), + PCI_SLOT(e->devid_end), PCI_FUNC(e->devid_end), + e->address_start, e->address_end, m->flags); + list_add_tail(&e->list, &amd_iommu_unity_map); return 0; @@ -967,18 +1037,28 @@ static void init_device_table(void) * This function finally enables all IOMMUs found in the system after * they have been initialized */ -static void __init enable_iommus(void) +static void enable_iommus(void) { struct amd_iommu *iommu; - list_for_each_entry(iommu, &amd_iommu_list, list) { + for_each_iommu(iommu) { + iommu_set_device_table(iommu); + iommu_enable_command_buffer(iommu); + iommu_enable_event_buffer(iommu); iommu_set_exclusion_range(iommu); iommu_init_msi(iommu); - iommu_enable_event_logging(iommu); iommu_enable(iommu); } } +static void disable_iommus(void) +{ + struct amd_iommu *iommu; + + for_each_iommu(iommu) + iommu_disable(iommu); +} + /* * Suspend/Resume support * disable suspend until real resume implemented @@ -986,12 +1066,31 @@ static void __init enable_iommus(void) static int amd_iommu_resume(struct sys_device *dev) { + /* + * Disable IOMMUs before reprogramming the hardware registers. + * IOMMU is still enabled from the resume kernel. + */ + disable_iommus(); + + /* re-load the hardware */ + enable_iommus(); + + /* + * we have to flush after the IOMMUs are enabled because a + * disabled IOMMU will never execute the commands we send + */ + amd_iommu_flush_all_domains(); + amd_iommu_flush_all_devices(); + return 0; } static int amd_iommu_suspend(struct sys_device *dev, pm_message_t state) { - return -EINVAL; + /* disable IOMMUs to go out of the way for BIOS */ + disable_iommus(); + + return 0; } static struct sysdev_class amd_iommu_sysdev_class = { @@ -1137,9 +1236,6 @@ int __init amd_iommu_init(void) enable_iommus(); - printk(KERN_INFO "AMD IOMMU: aperture size is %d MB\n", - (1 << (amd_iommu_aperture_order-20))); - printk(KERN_INFO "AMD IOMMU: device isolation "); if (amd_iommu_isolate) printk("enabled\n"); @@ -1211,6 +1307,13 @@ void __init amd_iommu_detect(void) * ****************************************************************************/ +static int __init parse_amd_iommu_dump(char *str) +{ + amd_iommu_dump = true; + + return 1; +} + static int __init parse_amd_iommu_options(char *str) { for (; *str; ++str) { @@ -1225,15 +1328,5 @@ static int __init parse_amd_iommu_options(char *str) return 1; } -static int __init parse_amd_iommu_size_options(char *str) -{ - unsigned order = PAGE_SHIFT + get_order(memparse(str, &str)); - - if ((order > 24) && (order < 31)) - amd_iommu_aperture_order = order; - - return 1; -} - +__setup("amd_iommu_dump", parse_amd_iommu_dump); __setup("amd_iommu=", parse_amd_iommu_options); -__setup("amd_iommu_size=", parse_amd_iommu_size_options); diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index f2870920f24..076d3881f3d 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -14,6 +14,7 @@ * Mikael Pettersson : PM converted to driver model. */ +#include <linux/perf_counter.h> #include <linux/kernel_stat.h> #include <linux/mc146818rtc.h> #include <linux/acpi_pmtmr.h> @@ -34,6 +35,7 @@ #include <linux/smp.h> #include <linux/mm.h> +#include <asm/perf_counter.h> #include <asm/pgalloc.h> #include <asm/atomic.h> #include <asm/mpspec.h> @@ -98,6 +100,29 @@ early_param("lapic", parse_lapic); /* Local APIC was disabled by the BIOS and enabled by the kernel */ static int enabled_via_apicbase; +/* + * Handle interrupt mode configuration register (IMCR). + * This register controls whether the interrupt signals + * that reach the BSP come from the master PIC or from the + * local APIC. Before entering Symmetric I/O Mode, either + * the BIOS or the operating system must switch out of + * PIC Mode by changing the IMCR. + */ +static inline void imcr_pic_to_apic(void) +{ + /* select IMCR register */ + outb(0x70, 0x22); + /* NMI and 8259 INTR go through APIC */ + outb(0x01, 0x23); +} + +static inline void imcr_apic_to_pic(void) +{ + /* select IMCR register */ + outb(0x70, 0x22); + /* NMI and 8259 INTR go directly to BSP */ + outb(0x00, 0x23); +} #endif #ifdef CONFIG_X86_64 @@ -111,13 +136,19 @@ static __init int setup_apicpmtimer(char *s) __setup("apicpmtimer", setup_apicpmtimer); #endif +int x2apic_mode; #ifdef CONFIG_X86_X2APIC -int x2apic; /* x2apic enabled before OS handover */ static int x2apic_preenabled; static int disable_x2apic; static __init int setup_nox2apic(char *str) { + if (x2apic_enabled()) { + pr_warning("Bios already enabled x2apic, " + "can't enforce nox2apic"); + return 0; + } + disable_x2apic = 1; setup_clear_cpu_cap(X86_FEATURE_X2APIC); return 0; @@ -209,6 +240,31 @@ static int modern_apic(void) return lapic_get_version() >= 0x14; } +/* + * bare function to substitute write operation + * and it's _that_ fast :) + */ +static void native_apic_write_dummy(u32 reg, u32 v) +{ + WARN_ON_ONCE((cpu_has_apic || !disable_apic)); +} + +static u32 native_apic_read_dummy(u32 reg) +{ + WARN_ON_ONCE((cpu_has_apic && !disable_apic)); + return 0; +} + +/* + * right after this call apic->write/read doesn't do anything + * note that there is no restore operation it works one way + */ +void apic_disable(void) +{ + apic->read = native_apic_read_dummy; + apic->write = native_apic_write_dummy; +} + void native_apic_wait_icr_idle(void) { while (apic_read(APIC_ICR) & APIC_ICR_BUSY) @@ -348,7 +404,7 @@ static void __setup_APIC_LVTT(unsigned int clocks, int oneshot, int irqen) static void setup_APIC_eilvt(u8 lvt_off, u8 vector, u8 msg_type, u8 mask) { - unsigned long reg = (lvt_off << 4) + APIC_EILVT0; + unsigned long reg = (lvt_off << 4) + APIC_EILVTn(0); unsigned int v = (mask << 16) | (msg_type << 8) | vector; apic_write(reg, v); @@ -815,7 +871,7 @@ void clear_local_APIC(void) u32 v; /* APIC hasn't been mapped yet */ - if (!x2apic && !apic_phys) + if (!x2apic_mode && !apic_phys) return; maxlvt = lapic_get_maxlvt(); @@ -1133,6 +1189,7 @@ void __cpuinit setup_local_APIC(void) apic_write(APIC_ESR, 0); } #endif + perf_counters_lapic_init(); preempt_disable(); @@ -1287,7 +1344,7 @@ void check_x2apic(void) { if (x2apic_enabled()) { pr_info("x2apic enabled by BIOS, switching to x2apic ops\n"); - x2apic_preenabled = x2apic = 1; + x2apic_preenabled = x2apic_mode = 1; } } @@ -1295,7 +1352,7 @@ void enable_x2apic(void) { int msr, msr2; - if (!x2apic) + if (!x2apic_mode) return; rdmsr(MSR_IA32_APICBASE, msr, msr2); @@ -1304,6 +1361,7 @@ void enable_x2apic(void) wrmsr(MSR_IA32_APICBASE, msr | X2APIC_ENABLE, 0); } } +#endif /* CONFIG_X86_X2APIC */ void __init enable_IR_x2apic(void) { @@ -1312,32 +1370,21 @@ void __init enable_IR_x2apic(void) unsigned long flags; struct IO_APIC_route_entry **ioapic_entries = NULL; - if (!cpu_has_x2apic) - return; - - if (!x2apic_preenabled && disable_x2apic) { - pr_info("Skipped enabling x2apic and Interrupt-remapping " - "because of nox2apic\n"); - return; + ret = dmar_table_init(); + if (ret) { + pr_debug("dmar_table_init() failed with %d:\n", ret); + goto ir_failed; } - if (x2apic_preenabled && disable_x2apic) - panic("Bios already enabled x2apic, can't enforce nox2apic"); - - if (!x2apic_preenabled && skip_ioapic_setup) { - pr_info("Skipped enabling x2apic and Interrupt-remapping " - "because of skipping io-apic setup\n"); - return; + if (!intr_remapping_supported()) { + pr_debug("intr-remapping not supported\n"); + goto ir_failed; } - ret = dmar_table_init(); - if (ret) { - pr_info("dmar_table_init() failed with %d:\n", ret); - if (x2apic_preenabled) - panic("x2apic enabled by bios. But IR enabling failed"); - else - pr_info("Not enabling x2apic,Intr-remapping\n"); + if (!x2apic_preenabled && skip_ioapic_setup) { + pr_info("Skipped enabling intr-remap because of skipping " + "io-apic setup\n"); return; } @@ -1357,19 +1404,16 @@ void __init enable_IR_x2apic(void) mask_IO_APIC_setup(ioapic_entries); mask_8259A(); - ret = enable_intr_remapping(EIM_32BIT_APIC_ID); - - if (ret && x2apic_preenabled) { - local_irq_restore(flags); - panic("x2apic enabled by bios. But IR enabling failed"); - } - + ret = enable_intr_remapping(x2apic_supported()); if (ret) goto end_restore; - if (!x2apic) { - x2apic = 1; + pr_info("Enabled Interrupt-remapping\n"); + + if (x2apic_supported() && !x2apic_mode) { + x2apic_mode = 1; enable_x2apic(); + pr_info("Enabled x2apic\n"); } end_restore: @@ -1378,37 +1422,34 @@ end_restore: * IR enabling failed */ restore_IO_APIC_setup(ioapic_entries); - else - reinit_intr_remapped_IO_APIC(x2apic_preenabled, ioapic_entries); unmask_8259A(); local_irq_restore(flags); end: - if (!ret) { - if (!x2apic_preenabled) - pr_info("Enabled x2apic and interrupt-remapping\n"); - else - pr_info("Enabled Interrupt-remapping\n"); - } else - pr_err("Failed to enable Interrupt-remapping and x2apic\n"); if (ioapic_entries) free_ioapic_entries(ioapic_entries); + + if (!ret) + return; + +ir_failed: + if (x2apic_preenabled) + panic("x2apic enabled by bios. But IR enabling failed"); + else if (cpu_has_x2apic) + pr_info("Not enabling x2apic,Intr-remapping\n"); #else if (!cpu_has_x2apic) return; if (x2apic_preenabled) panic("x2apic enabled prior OS handover," - " enable CONFIG_INTR_REMAP"); - - pr_info("Enable CONFIG_INTR_REMAP for enabling intr-remapping " - " and x2apic\n"); + " enable CONFIG_X86_X2APIC, CONFIG_INTR_REMAP"); #endif return; } -#endif /* CONFIG_X86_X2APIC */ + #ifdef CONFIG_X86_64 /* @@ -1425,7 +1466,6 @@ static int __init detect_init_APIC(void) } mp_lapic_addr = APIC_DEFAULT_PHYS_BASE; - boot_cpu_physical_apicid = 0; return 0; } #else @@ -1539,32 +1579,49 @@ void __init early_init_lapic_mapping(void) */ void __init init_apic_mappings(void) { - if (x2apic) { + unsigned int new_apicid; + + if (x2apic_mode) { boot_cpu_physical_apicid = read_apic_id(); return; } - /* - * If no local APIC can be found then set up a fake all - * zeroes page to simulate the local APIC and another - * one for the IO-APIC. - */ + /* If no local APIC can be found return early */ if (!smp_found_config && detect_init_APIC()) { - apic_phys = (unsigned long) alloc_bootmem_pages(PAGE_SIZE); - apic_phys = __pa(apic_phys); - } else + /* lets NOP'ify apic operations */ + pr_info("APIC: disable apic facility\n"); + apic_disable(); + } else { apic_phys = mp_lapic_addr; - set_fixmap_nocache(FIX_APIC_BASE, apic_phys); - apic_printk(APIC_VERBOSE, "mapped APIC to %08lx (%08lx)\n", - APIC_BASE, apic_phys); + /* + * acpi lapic path already maps that address in + * acpi_register_lapic_address() + */ + if (!acpi_lapic) + set_fixmap_nocache(FIX_APIC_BASE, apic_phys); + + apic_printk(APIC_VERBOSE, "mapped APIC to %08lx (%08lx)\n", + APIC_BASE, apic_phys); + } /* * Fetch the APIC ID of the BSP in case we have a * default configuration (or the MP table is broken). */ - if (boot_cpu_physical_apicid == -1U) - boot_cpu_physical_apicid = read_apic_id(); + new_apicid = read_apic_id(); + if (boot_cpu_physical_apicid != new_apicid) { + boot_cpu_physical_apicid = new_apicid; + /* + * yeah -- we lie about apic_version + * in case if apic was disabled via boot option + * but it's not a problem for SMP compiled kernel + * since smp_sanity_check is prepared for such a case + * and disable smp mode + */ + apic_version[new_apicid] = + GET_APIC_VERSION(apic_read(APIC_LVR)); + } } /* @@ -1733,8 +1790,7 @@ void __init connect_bsp_APIC(void) */ apic_printk(APIC_VERBOSE, "leaving PIC mode, " "enabling APIC mode.\n"); - outb(0x70, 0x22); - outb(0x01, 0x23); + imcr_pic_to_apic(); } #endif if (apic->enable_apic_mode) @@ -1762,8 +1818,7 @@ void disconnect_bsp_APIC(int virt_wire_setup) */ apic_printk(APIC_VERBOSE, "disabling APIC mode, " "entering PIC mode.\n"); - outb(0x70, 0x22); - outb(0x00, 0x23); + imcr_apic_to_pic(); return; } #endif @@ -1969,10 +2024,10 @@ static int lapic_suspend(struct sys_device *dev, pm_message_t state) local_irq_save(flags); disable_local_APIC(); -#ifdef CONFIG_INTR_REMAP + if (intr_remapping_enabled) disable_intr_remapping(); -#endif + local_irq_restore(flags); return 0; } @@ -1982,42 +2037,34 @@ static int lapic_resume(struct sys_device *dev) unsigned int l, h; unsigned long flags; int maxlvt; - -#ifdef CONFIG_INTR_REMAP - int ret; + int ret = 0; struct IO_APIC_route_entry **ioapic_entries = NULL; if (!apic_pm_state.active) return 0; local_irq_save(flags); - if (x2apic) { + if (intr_remapping_enabled) { ioapic_entries = alloc_ioapic_entries(); if (!ioapic_entries) { WARN(1, "Alloc ioapic_entries in lapic resume failed."); - return -ENOMEM; + ret = -ENOMEM; + goto restore; } ret = save_IO_APIC_setup(ioapic_entries); if (ret) { WARN(1, "Saving IO-APIC state failed: %d\n", ret); free_ioapic_entries(ioapic_entries); - return ret; + goto restore; } mask_IO_APIC_setup(ioapic_entries); mask_8259A(); - enable_x2apic(); } -#else - if (!apic_pm_state.active) - return 0; - local_irq_save(flags); - if (x2apic) + if (x2apic_mode) enable_x2apic(); -#endif - else { /* * Make sure the APICBASE points to the right address @@ -2055,21 +2102,16 @@ static int lapic_resume(struct sys_device *dev) apic_write(APIC_ESR, 0); apic_read(APIC_ESR); -#ifdef CONFIG_INTR_REMAP - if (intr_remapping_enabled) - reenable_intr_remapping(EIM_32BIT_APIC_ID); - - if (x2apic) { + if (intr_remapping_enabled) { + reenable_intr_remapping(x2apic_mode); unmask_8259A(); restore_IO_APIC_setup(ioapic_entries); free_ioapic_entries(ioapic_entries); } -#endif - +restore: local_irq_restore(flags); - - return 0; + return ret; } /* @@ -2117,31 +2159,14 @@ static void apic_pm_activate(void) { } #endif /* CONFIG_PM */ #ifdef CONFIG_X86_64 -/* - * apic_is_clustered_box() -- Check if we can expect good TSC - * - * Thus far, the major user of this is IBM's Summit2 series: - * - * Clustered boxes may have unsynced TSC problems if they are - * multi-chassis. Use available data to take a good guess. - * If in doubt, go HPET. - */ -__cpuinit int apic_is_clustered_box(void) + +static int __cpuinit apic_cluster_num(void) { int i, clusters, zeros; unsigned id; u16 *bios_cpu_apicid; DECLARE_BITMAP(clustermap, NUM_APIC_CLUSTERS); - /* - * there is not this kind of box with AMD CPU yet. - * Some AMD box with quadcore cpu and 8 sockets apicid - * will be [4, 0x23] or [8, 0x27] could be thought to - * vsmp box still need checking... - */ - if ((boot_cpu_data.x86_vendor == X86_VENDOR_AMD) && !is_vsmp_box()) - return 0; - bios_cpu_apicid = early_per_cpu_ptr(x86_bios_cpu_apicid); bitmap_zero(clustermap, NUM_APIC_CLUSTERS); @@ -2177,18 +2202,67 @@ __cpuinit int apic_is_clustered_box(void) ++zeros; } - /* ScaleMP vSMPowered boxes have one cluster per board and TSCs are - * not guaranteed to be synced between boards - */ - if (is_vsmp_box() && clusters > 1) + return clusters; +} + +static int __cpuinitdata multi_checked; +static int __cpuinitdata multi; + +static int __cpuinit set_multi(const struct dmi_system_id *d) +{ + if (multi) + return 0; + pr_info("APIC: %s detected, Multi Chassis\n", d->ident); + multi = 1; + return 0; +} + +static const __cpuinitconst struct dmi_system_id multi_dmi_table[] = { + { + .callback = set_multi, + .ident = "IBM System Summit2", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "IBM"), + DMI_MATCH(DMI_PRODUCT_NAME, "Summit2"), + }, + }, + {} +}; + +static void __cpuinit dmi_check_multi(void) +{ + if (multi_checked) + return; + + dmi_check_system(multi_dmi_table); + multi_checked = 1; +} + +/* + * apic_is_clustered_box() -- Check if we can expect good TSC + * + * Thus far, the major user of this is IBM's Summit2 series: + * Clustered boxes may have unsynced TSC problems if they are + * multi-chassis. + * Use DMI to check them + */ +__cpuinit int apic_is_clustered_box(void) +{ + dmi_check_multi(); + if (multi) return 1; + if (!is_vsmp_box()) + return 0; + /* - * If clusters > 2, then should be multi-chassis. - * May have to revisit this when multi-core + hyperthreaded CPUs come - * out, but AFAIK this will work even for them. + * ScaleMP vSMPowered boxes have one cluster per board and TSCs are + * not guaranteed to be synced between boards */ - return (clusters > 2); + if (apic_cluster_num() > 1) + return 1; + + return 0; } #endif diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c index 306e5e88fb6..d0c99abc26c 100644 --- a/arch/x86/kernel/apic/apic_flat_64.c +++ b/arch/x86/kernel/apic/apic_flat_64.c @@ -161,7 +161,7 @@ static int flat_apic_id_registered(void) static int flat_phys_pkg_id(int initial_apic_id, int index_msb) { - return hard_smp_processor_id() >> index_msb; + return initial_apic_id >> index_msb; } struct apic apic_flat = { @@ -235,7 +235,7 @@ static int physflat_acpi_madt_oem_check(char *oem_id, char *oem_table_id) * regardless of how many processors are present (x86_64 ES7000 * is an example). */ - if (acpi_gbl_FADT.header.revision > FADT2_REVISION_ID && + if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && (acpi_gbl_FADT.flags & ACPI_FADT_APIC_PHYSICAL)) { printk(KERN_DEBUG "system APIC only can use physical flat"); return 1; diff --git a/arch/x86/kernel/apic/es7000_32.c b/arch/x86/kernel/apic/es7000_32.c index 30294777557..69328ac8de9 100644 --- a/arch/x86/kernel/apic/es7000_32.c +++ b/arch/x86/kernel/apic/es7000_32.c @@ -145,7 +145,7 @@ es7000_rename_gsi(int ioapic, int gsi) return gsi; } -static int wakeup_secondary_cpu_via_mip(int cpu, unsigned long eip) +static int __cpuinit wakeup_secondary_cpu_via_mip(int cpu, unsigned long eip) { unsigned long vect = 0, psaival = 0; diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 30da617d18e..1946fac42ab 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -59,6 +59,7 @@ #include <asm/setup.h> #include <asm/irq_remapping.h> #include <asm/hpet.h> +#include <asm/hw_irq.h> #include <asm/uv/uv_hub.h> #include <asm/uv/uv_irq.h> @@ -129,12 +130,9 @@ struct irq_pin_list { struct irq_pin_list *next; }; -static struct irq_pin_list *get_one_free_irq_2_pin(int cpu) +static struct irq_pin_list *get_one_free_irq_2_pin(int node) { struct irq_pin_list *pin; - int node; - - node = cpu_to_node(cpu); pin = kzalloc_node(sizeof(*pin), GFP_ATOMIC, node); @@ -148,9 +146,6 @@ struct irq_cfg { unsigned move_cleanup_count; u8 vector; u8 move_in_progress : 1; -#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC - u8 move_desc_pending : 1; -#endif }; /* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */ @@ -212,12 +207,9 @@ static struct irq_cfg *irq_cfg(unsigned int irq) return cfg; } -static struct irq_cfg *get_one_free_irq_cfg(int cpu) +static struct irq_cfg *get_one_free_irq_cfg(int node) { struct irq_cfg *cfg; - int node; - - node = cpu_to_node(cpu); cfg = kzalloc_node(sizeof(*cfg), GFP_ATOMIC, node); if (cfg) { @@ -238,13 +230,13 @@ static struct irq_cfg *get_one_free_irq_cfg(int cpu) return cfg; } -int arch_init_chip_data(struct irq_desc *desc, int cpu) +int arch_init_chip_data(struct irq_desc *desc, int node) { struct irq_cfg *cfg; cfg = desc->chip_data; if (!cfg) { - desc->chip_data = get_one_free_irq_cfg(cpu); + desc->chip_data = get_one_free_irq_cfg(node); if (!desc->chip_data) { printk(KERN_ERR "can not alloc irq_cfg\n"); BUG_ON(1); @@ -254,10 +246,9 @@ int arch_init_chip_data(struct irq_desc *desc, int cpu) return 0; } -#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC - +/* for move_irq_desc */ static void -init_copy_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg, int cpu) +init_copy_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg, int node) { struct irq_pin_list *old_entry, *head, *tail, *entry; @@ -266,7 +257,7 @@ init_copy_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg, int cpu) if (!old_entry) return; - entry = get_one_free_irq_2_pin(cpu); + entry = get_one_free_irq_2_pin(node); if (!entry) return; @@ -276,7 +267,7 @@ init_copy_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg, int cpu) tail = entry; old_entry = old_entry->next; while (old_entry) { - entry = get_one_free_irq_2_pin(cpu); + entry = get_one_free_irq_2_pin(node); if (!entry) { entry = head; while (entry) { @@ -316,12 +307,12 @@ static void free_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg) } void arch_init_copy_chip_data(struct irq_desc *old_desc, - struct irq_desc *desc, int cpu) + struct irq_desc *desc, int node) { struct irq_cfg *cfg; struct irq_cfg *old_cfg; - cfg = get_one_free_irq_cfg(cpu); + cfg = get_one_free_irq_cfg(node); if (!cfg) return; @@ -332,7 +323,7 @@ void arch_init_copy_chip_data(struct irq_desc *old_desc, memcpy(cfg, old_cfg, sizeof(struct irq_cfg)); - init_copy_irq_2_pin(old_cfg, cfg, cpu); + init_copy_irq_2_pin(old_cfg, cfg, node); } static void free_irq_cfg(struct irq_cfg *old_cfg) @@ -356,19 +347,7 @@ void arch_free_chip_data(struct irq_desc *old_desc, struct irq_desc *desc) old_desc->chip_data = NULL; } } - -static void -set_extra_move_desc(struct irq_desc *desc, const struct cpumask *mask) -{ - struct irq_cfg *cfg = desc->chip_data; - - if (!cfg->move_in_progress) { - /* it means that domain is not changed */ - if (!cpumask_intersects(desc->affinity, mask)) - cfg->move_desc_pending = 1; - } -} -#endif +/* end for move_irq_desc */ #else static struct irq_cfg *irq_cfg(unsigned int irq) @@ -378,13 +357,6 @@ static struct irq_cfg *irq_cfg(unsigned int irq) #endif -#ifndef CONFIG_NUMA_MIGRATE_IRQ_DESC -static inline void -set_extra_move_desc(struct irq_desc *desc, const struct cpumask *mask) -{ -} -#endif - struct io_apic { unsigned int index; unsigned int unused[3]; @@ -518,132 +490,18 @@ static void ioapic_mask_entry(int apic, int pin) spin_unlock_irqrestore(&ioapic_lock, flags); } -#ifdef CONFIG_SMP -static void send_cleanup_vector(struct irq_cfg *cfg) -{ - cpumask_var_t cleanup_mask; - - if (unlikely(!alloc_cpumask_var(&cleanup_mask, GFP_ATOMIC))) { - unsigned int i; - cfg->move_cleanup_count = 0; - for_each_cpu_and(i, cfg->old_domain, cpu_online_mask) - cfg->move_cleanup_count++; - for_each_cpu_and(i, cfg->old_domain, cpu_online_mask) - apic->send_IPI_mask(cpumask_of(i), IRQ_MOVE_CLEANUP_VECTOR); - } else { - cpumask_and(cleanup_mask, cfg->old_domain, cpu_online_mask); - cfg->move_cleanup_count = cpumask_weight(cleanup_mask); - apic->send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR); - free_cpumask_var(cleanup_mask); - } - cfg->move_in_progress = 0; -} - -static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq_cfg *cfg) -{ - int apic, pin; - struct irq_pin_list *entry; - u8 vector = cfg->vector; - - entry = cfg->irq_2_pin; - for (;;) { - unsigned int reg; - - if (!entry) - break; - - apic = entry->apic; - pin = entry->pin; - /* - * With interrupt-remapping, destination information comes - * from interrupt-remapping table entry. - */ - if (!irq_remapped(irq)) - io_apic_write(apic, 0x11 + pin*2, dest); - reg = io_apic_read(apic, 0x10 + pin*2); - reg &= ~IO_APIC_REDIR_VECTOR_MASK; - reg |= vector; - io_apic_modify(apic, 0x10 + pin*2, reg); - if (!entry->next) - break; - entry = entry->next; - } -} - -static int -assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask); - -/* - * Either sets desc->affinity to a valid value, and returns - * ->cpu_mask_to_apicid of that, or returns BAD_APICID and - * leaves desc->affinity untouched. - */ -static unsigned int -set_desc_affinity(struct irq_desc *desc, const struct cpumask *mask) -{ - struct irq_cfg *cfg; - unsigned int irq; - - if (!cpumask_intersects(mask, cpu_online_mask)) - return BAD_APICID; - - irq = desc->irq; - cfg = desc->chip_data; - if (assign_irq_vector(irq, cfg, mask)) - return BAD_APICID; - - /* check that before desc->addinity get updated */ - set_extra_move_desc(desc, mask); - - cpumask_copy(desc->affinity, mask); - - return apic->cpu_mask_to_apicid_and(desc->affinity, cfg->domain); -} - -static void -set_ioapic_affinity_irq_desc(struct irq_desc *desc, const struct cpumask *mask) -{ - struct irq_cfg *cfg; - unsigned long flags; - unsigned int dest; - unsigned int irq; - - irq = desc->irq; - cfg = desc->chip_data; - - spin_lock_irqsave(&ioapic_lock, flags); - dest = set_desc_affinity(desc, mask); - if (dest != BAD_APICID) { - /* Only the high 8 bits are valid. */ - dest = SET_APIC_LOGICAL_ID(dest); - __target_IO_APIC_irq(irq, dest, cfg); - } - spin_unlock_irqrestore(&ioapic_lock, flags); -} - -static void -set_ioapic_affinity_irq(unsigned int irq, const struct cpumask *mask) -{ - struct irq_desc *desc; - - desc = irq_to_desc(irq); - - set_ioapic_affinity_irq_desc(desc, mask); -} -#endif /* CONFIG_SMP */ - /* * The common case is 1:1 IRQ<->pin mappings. Sometimes there are * shared ISA-space IRQs, so we have to support them. We are super * fast in the common case, and fast for shared ISA-space IRQs. */ -static void add_pin_to_irq_cpu(struct irq_cfg *cfg, int cpu, int apic, int pin) +static void add_pin_to_irq_node(struct irq_cfg *cfg, int node, int apic, int pin) { struct irq_pin_list *entry; entry = cfg->irq_2_pin; if (!entry) { - entry = get_one_free_irq_2_pin(cpu); + entry = get_one_free_irq_2_pin(node); if (!entry) { printk(KERN_ERR "can not alloc irq_2_pin to add %d - %d\n", apic, pin); @@ -663,7 +521,7 @@ static void add_pin_to_irq_cpu(struct irq_cfg *cfg, int cpu, int apic, int pin) entry = entry->next; } - entry->next = get_one_free_irq_2_pin(cpu); + entry->next = get_one_free_irq_2_pin(node); entry = entry->next; entry->apic = apic; entry->pin = pin; @@ -672,7 +530,7 @@ static void add_pin_to_irq_cpu(struct irq_cfg *cfg, int cpu, int apic, int pin) /* * Reroute an IRQ to a different pin. */ -static void __init replace_pin_at_irq_cpu(struct irq_cfg *cfg, int cpu, +static void __init replace_pin_at_irq_node(struct irq_cfg *cfg, int node, int oldapic, int oldpin, int newapic, int newpin) { @@ -692,7 +550,7 @@ static void __init replace_pin_at_irq_cpu(struct irq_cfg *cfg, int cpu, /* why? call replace before add? */ if (!replaced) - add_pin_to_irq_cpu(cfg, cpu, newapic, newpin); + add_pin_to_irq_node(cfg, node, newapic, newpin); } static inline void io_apic_modify_irq(struct irq_cfg *cfg, @@ -850,7 +708,6 @@ static int __init ioapic_pirq_setup(char *str) __setup("pirq=", ioapic_pirq_setup); #endif /* CONFIG_X86_32 */ -#ifdef CONFIG_INTR_REMAP struct IO_APIC_route_entry **alloc_ioapic_entries(void) { int apic; @@ -948,20 +805,6 @@ int restore_IO_APIC_setup(struct IO_APIC_route_entry **ioapic_entries) return 0; } -void reinit_intr_remapped_IO_APIC(int intr_remapping, - struct IO_APIC_route_entry **ioapic_entries) - -{ - /* - * for now plain restore of previous settings. - * TBD: In the case of OS enabling interrupt-remapping, - * IO-APIC RTE's need to be setup to point to interrupt-remapping - * table entries. for now, do a plain restore, and wait for - * the setup_IO_APIC_irqs() to do proper initialization. - */ - restore_IO_APIC_setup(ioapic_entries); -} - void free_ioapic_entries(struct IO_APIC_route_entry **ioapic_entries) { int apic; @@ -971,7 +814,6 @@ void free_ioapic_entries(struct IO_APIC_route_entry **ioapic_entries) kfree(ioapic_entries); } -#endif /* * Find the IRQ entry number of a certain pin. @@ -1032,54 +874,6 @@ static int __init find_isa_irq_apic(int irq, int type) return -1; } -/* - * Find a specific PCI IRQ entry. - * Not an __init, possibly needed by modules - */ -static int pin_2_irq(int idx, int apic, int pin); - -int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) -{ - int apic, i, best_guess = -1; - - apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", - bus, slot, pin); - if (test_bit(bus, mp_bus_not_pci)) { - apic_printk(APIC_VERBOSE, "PCI BIOS passed nonexistent PCI bus %d!\n", bus); - return -1; - } - for (i = 0; i < mp_irq_entries; i++) { - int lbus = mp_irqs[i].srcbus; - - for (apic = 0; apic < nr_ioapics; apic++) - if (mp_ioapics[apic].apicid == mp_irqs[i].dstapic || - mp_irqs[i].dstapic == MP_APIC_ALL) - break; - - if (!test_bit(lbus, mp_bus_not_pci) && - !mp_irqs[i].irqtype && - (bus == lbus) && - (slot == ((mp_irqs[i].srcbusirq >> 2) & 0x1f))) { - int irq = pin_2_irq(i, apic, mp_irqs[i].dstirq); - - if (!(apic || IO_APIC_IRQ(irq))) - continue; - - if (pin == (mp_irqs[i].srcbusirq & 3)) - return irq; - /* - * Use the first all-but-pin matching entry as a - * best-guess fuzzy result for broken mptables. - */ - if (best_guess < 0) - best_guess = irq; - } - } - return best_guess; -} - -EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector); - #if defined(CONFIG_EISA) || defined(CONFIG_MCA) /* * EISA Edge/Level control register, ELCR @@ -1298,6 +1092,64 @@ static int pin_2_irq(int idx, int apic, int pin) return irq; } +/* + * Find a specific PCI IRQ entry. + * Not an __init, possibly needed by modules + */ +int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin, + struct io_apic_irq_attr *irq_attr) +{ + int apic, i, best_guess = -1; + + apic_printk(APIC_DEBUG, + "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", + bus, slot, pin); + if (test_bit(bus, mp_bus_not_pci)) { + apic_printk(APIC_VERBOSE, + "PCI BIOS passed nonexistent PCI bus %d!\n", bus); + return -1; + } + for (i = 0; i < mp_irq_entries; i++) { + int lbus = mp_irqs[i].srcbus; + + for (apic = 0; apic < nr_ioapics; apic++) + if (mp_ioapics[apic].apicid == mp_irqs[i].dstapic || + mp_irqs[i].dstapic == MP_APIC_ALL) + break; + + if (!test_bit(lbus, mp_bus_not_pci) && + !mp_irqs[i].irqtype && + (bus == lbus) && + (slot == ((mp_irqs[i].srcbusirq >> 2) & 0x1f))) { + int irq = pin_2_irq(i, apic, mp_irqs[i].dstirq); + + if (!(apic || IO_APIC_IRQ(irq))) + continue; + + if (pin == (mp_irqs[i].srcbusirq & 3)) { + set_io_apic_irq_attr(irq_attr, apic, + mp_irqs[i].dstirq, + irq_trigger(i), + irq_polarity(i)); + return irq; + } + /* + * Use the first all-but-pin matching entry as a + * best-guess fuzzy result for broken mptables. + */ + if (best_guess < 0) { + set_io_apic_irq_attr(irq_attr, apic, + mp_irqs[i].dstirq, + irq_trigger(i), + irq_polarity(i)); + best_guess = irq; + } + } + } + return best_guess; +} +EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector); + void lock_vector_lock(void) { /* Used to the online set of cpus does not change @@ -1628,58 +1480,70 @@ static void setup_IO_APIC_irq(int apic_id, int pin, unsigned int irq, struct irq ioapic_write_entry(apic_id, pin, entry); } +static struct { + DECLARE_BITMAP(pin_programmed, MP_MAX_IOAPIC_PIN + 1); +} mp_ioapic_routing[MAX_IO_APICS]; + static void __init setup_IO_APIC_irqs(void) { - int apic_id, pin, idx, irq; + int apic_id = 0, pin, idx, irq; int notcon = 0; struct irq_desc *desc; struct irq_cfg *cfg; - int cpu = boot_cpu_id; + int node = cpu_to_node(boot_cpu_id); apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n"); - for (apic_id = 0; apic_id < nr_ioapics; apic_id++) { - for (pin = 0; pin < nr_ioapic_registers[apic_id]; pin++) { - - idx = find_irq_entry(apic_id, pin, mp_INT); - if (idx == -1) { - if (!notcon) { - notcon = 1; - apic_printk(APIC_VERBOSE, - KERN_DEBUG " %d-%d", - mp_ioapics[apic_id].apicid, pin); - } else - apic_printk(APIC_VERBOSE, " %d-%d", - mp_ioapics[apic_id].apicid, pin); - continue; - } - if (notcon) { - apic_printk(APIC_VERBOSE, - " (apicid-pin) not connected\n"); - notcon = 0; - } +#ifdef CONFIG_ACPI + if (!acpi_disabled && acpi_ioapic) { + apic_id = mp_find_ioapic(0); + if (apic_id < 0) + apic_id = 0; + } +#endif - irq = pin_2_irq(idx, apic_id, pin); + for (pin = 0; pin < nr_ioapic_registers[apic_id]; pin++) { + idx = find_irq_entry(apic_id, pin, mp_INT); + if (idx == -1) { + if (!notcon) { + notcon = 1; + apic_printk(APIC_VERBOSE, + KERN_DEBUG " %d-%d", + mp_ioapics[apic_id].apicid, pin); + } else + apic_printk(APIC_VERBOSE, " %d-%d", + mp_ioapics[apic_id].apicid, pin); + continue; + } + if (notcon) { + apic_printk(APIC_VERBOSE, + " (apicid-pin) not connected\n"); + notcon = 0; + } - /* - * Skip the timer IRQ if there's a quirk handler - * installed and if it returns 1: - */ - if (apic->multi_timer_check && - apic->multi_timer_check(apic_id, irq)) - continue; + irq = pin_2_irq(idx, apic_id, pin); - desc = irq_to_desc_alloc_cpu(irq, cpu); - if (!desc) { - printk(KERN_INFO "can not get irq_desc for %d\n", irq); - continue; - } - cfg = desc->chip_data; - add_pin_to_irq_cpu(cfg, cpu, apic_id, pin); + /* + * Skip the timer IRQ if there's a quirk handler + * installed and if it returns 1: + */ + if (apic->multi_timer_check && + apic->multi_timer_check(apic_id, irq)) + continue; - setup_IO_APIC_irq(apic_id, pin, irq, desc, - irq_trigger(idx), irq_polarity(idx)); + desc = irq_to_desc_alloc_node(irq, node); + if (!desc) { + printk(KERN_INFO "can not get irq_desc for %d\n", irq); + continue; } + cfg = desc->chip_data; + add_pin_to_irq_node(cfg, node, apic_id, pin); + /* + * don't mark it in pin_programmed, so later acpi could + * set it correctly when irq < 16 + */ + setup_IO_APIC_irq(apic_id, pin, irq, desc, + irq_trigger(idx), irq_polarity(idx)); } if (notcon) @@ -1869,7 +1733,7 @@ __apicdebuginit(void) print_APIC_bitfield(int base) __apicdebuginit(void) print_local_APIC(void *dummy) { - unsigned int v, ver, maxlvt; + unsigned int i, v, ver, maxlvt; u64 icr; if (apic_verbosity == APIC_QUIET) @@ -1957,6 +1821,18 @@ __apicdebuginit(void) print_local_APIC(void *dummy) printk(KERN_DEBUG "... APIC TMCCT: %08x\n", v); v = apic_read(APIC_TDCR); printk(KERN_DEBUG "... APIC TDCR: %08x\n", v); + + if (boot_cpu_has(X86_FEATURE_EXTAPIC)) { + v = apic_read(APIC_EFEAT); + maxlvt = (v >> 16) & 0xff; + printk(KERN_DEBUG "... APIC EFEAT: %08x\n", v); + v = apic_read(APIC_ECTRL); + printk(KERN_DEBUG "... APIC ECTRL: %08x\n", v); + for (i = 0; i < maxlvt; i++) { + v = apic_read(APIC_EILVTn(i)); + printk(KERN_DEBUG "... APIC EILVT%d: %08x\n", i, v); + } + } printk("\n"); } @@ -2005,6 +1881,11 @@ __apicdebuginit(void) print_PIC(void) __apicdebuginit(int) print_all_ICs(void) { print_PIC(); + + /* don't print out if apic is not there */ + if (!cpu_has_apic || disable_apic) + return 0; + print_all_local_APICs(); print_IO_APIC(); @@ -2360,6 +2241,118 @@ static int ioapic_retrigger_irq(unsigned int irq) */ #ifdef CONFIG_SMP +static void send_cleanup_vector(struct irq_cfg *cfg) +{ + cpumask_var_t cleanup_mask; + + if (unlikely(!alloc_cpumask_var(&cleanup_mask, GFP_ATOMIC))) { + unsigned int i; + cfg->move_cleanup_count = 0; + for_each_cpu_and(i, cfg->old_domain, cpu_online_mask) + cfg->move_cleanup_count++; + for_each_cpu_and(i, cfg->old_domain, cpu_online_mask) + apic->send_IPI_mask(cpumask_of(i), IRQ_MOVE_CLEANUP_VECTOR); + } else { + cpumask_and(cleanup_mask, cfg->old_domain, cpu_online_mask); + cfg->move_cleanup_count = cpumask_weight(cleanup_mask); + apic->send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR); + free_cpumask_var(cleanup_mask); + } + cfg->move_in_progress = 0; +} + +static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq_cfg *cfg) +{ + int apic, pin; + struct irq_pin_list *entry; + u8 vector = cfg->vector; + + entry = cfg->irq_2_pin; + for (;;) { + unsigned int reg; + + if (!entry) + break; + + apic = entry->apic; + pin = entry->pin; + /* + * With interrupt-remapping, destination information comes + * from interrupt-remapping table entry. + */ + if (!irq_remapped(irq)) + io_apic_write(apic, 0x11 + pin*2, dest); + reg = io_apic_read(apic, 0x10 + pin*2); + reg &= ~IO_APIC_REDIR_VECTOR_MASK; + reg |= vector; + io_apic_modify(apic, 0x10 + pin*2, reg); + if (!entry->next) + break; + entry = entry->next; + } +} + +static int +assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask); + +/* + * Either sets desc->affinity to a valid value, and returns + * ->cpu_mask_to_apicid of that, or returns BAD_APICID and + * leaves desc->affinity untouched. + */ +static unsigned int +set_desc_affinity(struct irq_desc *desc, const struct cpumask *mask) +{ + struct irq_cfg *cfg; + unsigned int irq; + + if (!cpumask_intersects(mask, cpu_online_mask)) + return BAD_APICID; + + irq = desc->irq; + cfg = desc->chip_data; + if (assign_irq_vector(irq, cfg, mask)) + return BAD_APICID; + + cpumask_copy(desc->affinity, mask); + + return apic->cpu_mask_to_apicid_and(desc->affinity, cfg->domain); +} + +static int +set_ioapic_affinity_irq_desc(struct irq_desc *desc, const struct cpumask *mask) +{ + struct irq_cfg *cfg; + unsigned long flags; + unsigned int dest; + unsigned int irq; + int ret = -1; + + irq = desc->irq; + cfg = desc->chip_data; + + spin_lock_irqsave(&ioapic_lock, flags); + dest = set_desc_affinity(desc, mask); + if (dest != BAD_APICID) { + /* Only the high 8 bits are valid. */ + dest = SET_APIC_LOGICAL_ID(dest); + __target_IO_APIC_irq(irq, dest, cfg); + ret = 0; + } + spin_unlock_irqrestore(&ioapic_lock, flags); + + return ret; +} + +static int +set_ioapic_affinity_irq(unsigned int irq, const struct cpumask *mask) +{ + struct irq_desc *desc; + + desc = irq_to_desc(irq); + + return set_ioapic_affinity_irq_desc(desc, mask); +} #ifdef CONFIG_INTR_REMAP @@ -2374,26 +2367,25 @@ static int ioapic_retrigger_irq(unsigned int irq) * Real vector that is used for interrupting cpu will be coming from * the interrupt-remapping table entry. */ -static void +static int migrate_ioapic_irq_desc(struct irq_desc *desc, const struct cpumask *mask) { struct irq_cfg *cfg; struct irte irte; unsigned int dest; unsigned int irq; + int ret = -1; if (!cpumask_intersects(mask, cpu_online_mask)) - return; + return ret; irq = desc->irq; if (get_irte(irq, &irte)) - return; + return ret; cfg = desc->chip_data; if (assign_irq_vector(irq, cfg, mask)) - return; - - set_extra_move_desc(desc, mask); + return ret; dest = apic->cpu_mask_to_apicid_and(cfg->domain, mask); @@ -2409,27 +2401,30 @@ migrate_ioapic_irq_desc(struct irq_desc *desc, const struct cpumask *mask) send_cleanup_vector(cfg); cpumask_copy(desc->affinity, mask); + + return 0; } /* * Migrates the IRQ destination in the process context. */ -static void set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc, +static int set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc, const struct cpumask *mask) { - migrate_ioapic_irq_desc(desc, mask); + return migrate_ioapic_irq_desc(desc, mask); } -static void set_ir_ioapic_affinity_irq(unsigned int irq, +static int set_ir_ioapic_affinity_irq(unsigned int irq, const struct cpumask *mask) { struct irq_desc *desc = irq_to_desc(irq); - set_ir_ioapic_affinity_irq_desc(desc, mask); + return set_ir_ioapic_affinity_irq_desc(desc, mask); } #else -static inline void set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc, +static inline int set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc, const struct cpumask *mask) { + return 0; } #endif @@ -2491,86 +2486,19 @@ static void irq_complete_move(struct irq_desc **descp) struct irq_cfg *cfg = desc->chip_data; unsigned vector, me; - if (likely(!cfg->move_in_progress)) { -#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC - if (likely(!cfg->move_desc_pending)) - return; - - /* domain has not changed, but affinity did */ - me = smp_processor_id(); - if (cpumask_test_cpu(me, desc->affinity)) { - *descp = desc = move_irq_desc(desc, me); - /* get the new one */ - cfg = desc->chip_data; - cfg->move_desc_pending = 0; - } -#endif + if (likely(!cfg->move_in_progress)) return; - } vector = ~get_irq_regs()->orig_ax; me = smp_processor_id(); - if (vector == cfg->vector && cpumask_test_cpu(me, cfg->domain)) { -#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC - *descp = desc = move_irq_desc(desc, me); - /* get the new one */ - cfg = desc->chip_data; -#endif + if (vector == cfg->vector && cpumask_test_cpu(me, cfg->domain)) send_cleanup_vector(cfg); - } } #else static inline void irq_complete_move(struct irq_desc **descp) {} #endif -static void __eoi_ioapic_irq(unsigned int irq, struct irq_cfg *cfg) -{ - int apic, pin; - struct irq_pin_list *entry; - - entry = cfg->irq_2_pin; - for (;;) { - - if (!entry) - break; - - apic = entry->apic; - pin = entry->pin; - io_apic_eoi(apic, pin); - entry = entry->next; - } -} - -static void -eoi_ioapic_irq(struct irq_desc *desc) -{ - struct irq_cfg *cfg; - unsigned long flags; - unsigned int irq; - - irq = desc->irq; - cfg = desc->chip_data; - - spin_lock_irqsave(&ioapic_lock, flags); - __eoi_ioapic_irq(irq, cfg); - spin_unlock_irqrestore(&ioapic_lock, flags); -} - -#ifdef CONFIG_X86_X2APIC -static void ack_x2apic_level(unsigned int irq) -{ - struct irq_desc *desc = irq_to_desc(irq); - ack_x2APIC_irq(); - eoi_ioapic_irq(desc); -} - -static void ack_x2apic_edge(unsigned int irq) -{ - ack_x2APIC_irq(); -} -#endif - static void ack_apic_edge(unsigned int irq) { struct irq_desc *desc = irq_to_desc(irq); @@ -2634,9 +2562,6 @@ static void ack_apic_level(unsigned int irq) */ ack_APIC_irq(); - if (irq_remapped(irq)) - eoi_ioapic_irq(desc); - /* Now we can move and renable the irq */ if (unlikely(do_unmask_irq)) { /* Only migrate the irq if the ack has been received. @@ -2683,22 +2608,50 @@ static void ack_apic_level(unsigned int irq) } #ifdef CONFIG_INTR_REMAP +static void __eoi_ioapic_irq(unsigned int irq, struct irq_cfg *cfg) +{ + int apic, pin; + struct irq_pin_list *entry; + + entry = cfg->irq_2_pin; + for (;;) { + + if (!entry) + break; + + apic = entry->apic; + pin = entry->pin; + io_apic_eoi(apic, pin); + entry = entry->next; + } +} + +static void +eoi_ioapic_irq(struct irq_desc *desc) +{ + struct irq_cfg *cfg; + unsigned long flags; + unsigned int irq; + + irq = desc->irq; + cfg = desc->chip_data; + + spin_lock_irqsave(&ioapic_lock, flags); + __eoi_ioapic_irq(irq, cfg); + spin_unlock_irqrestore(&ioapic_lock, flags); +} + static void ir_ack_apic_edge(unsigned int irq) { -#ifdef CONFIG_X86_X2APIC - if (x2apic_enabled()) - return ack_x2apic_edge(irq); -#endif - return ack_apic_edge(irq); + ack_APIC_irq(); } static void ir_ack_apic_level(unsigned int irq) { -#ifdef CONFIG_X86_X2APIC - if (x2apic_enabled()) - return ack_x2apic_level(irq); -#endif - return ack_apic_level(irq); + struct irq_desc *desc = irq_to_desc(irq); + + ack_APIC_irq(); + eoi_ioapic_irq(desc); } #endif /* CONFIG_INTR_REMAP */ @@ -2903,7 +2856,7 @@ static inline void __init check_timer(void) { struct irq_desc *desc = irq_to_desc(0); struct irq_cfg *cfg = desc->chip_data; - int cpu = boot_cpu_id; + int node = cpu_to_node(boot_cpu_id); int apic1, pin1, apic2, pin2; unsigned long flags; int no_pin1 = 0; @@ -2969,7 +2922,7 @@ static inline void __init check_timer(void) * Ok, does IRQ0 through the IOAPIC work? */ if (no_pin1) { - add_pin_to_irq_cpu(cfg, cpu, apic1, pin1); + add_pin_to_irq_node(cfg, node, apic1, pin1); setup_timer_IRQ0_pin(apic1, pin1, cfg->vector); } else { /* for edge trigger, setup_IO_APIC_irq already @@ -3006,7 +2959,7 @@ static inline void __init check_timer(void) /* * legacy devices should be connected to IO APIC #0 */ - replace_pin_at_irq_cpu(cfg, cpu, apic1, pin1, apic2, pin2); + replace_pin_at_irq_node(cfg, node, apic1, pin1, apic2, pin2); setup_timer_IRQ0_pin(apic2, pin2, cfg->vector); enable_8259A_irq(0); if (timer_irq_works()) { @@ -3218,14 +3171,13 @@ static int nr_irqs_gsi = NR_IRQS_LEGACY; /* * Dynamic irq allocate and deallocation */ -unsigned int create_irq_nr(unsigned int irq_want) +unsigned int create_irq_nr(unsigned int irq_want, int node) { /* Allocate an unused irq */ unsigned int irq; unsigned int new; unsigned long flags; struct irq_cfg *cfg_new = NULL; - int cpu = boot_cpu_id; struct irq_desc *desc_new = NULL; irq = 0; @@ -3234,7 +3186,7 @@ unsigned int create_irq_nr(unsigned int irq_want) spin_lock_irqsave(&vector_lock, flags); for (new = irq_want; new < nr_irqs; new++) { - desc_new = irq_to_desc_alloc_cpu(new, cpu); + desc_new = irq_to_desc_alloc_node(new, node); if (!desc_new) { printk(KERN_INFO "can not get irq_desc for %d\n", new); continue; @@ -3243,6 +3195,9 @@ unsigned int create_irq_nr(unsigned int irq_want) if (cfg_new->vector != 0) continue; + + desc_new = move_irq_desc(desc_new, node); + if (__assign_irq_vector(new, cfg_new, apic->target_cpus()) == 0) irq = new; break; @@ -3260,11 +3215,12 @@ unsigned int create_irq_nr(unsigned int irq_want) int create_irq(void) { + int node = cpu_to_node(boot_cpu_id); unsigned int irq_want; int irq; irq_want = nr_irqs_gsi; - irq = create_irq_nr(irq_want); + irq = create_irq_nr(irq_want, node); if (irq == 0) irq = -1; @@ -3366,7 +3322,7 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms } #ifdef CONFIG_SMP -static void set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask) +static int set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask) { struct irq_desc *desc = irq_to_desc(irq); struct irq_cfg *cfg; @@ -3375,7 +3331,7 @@ static void set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask) dest = set_desc_affinity(desc, mask); if (dest == BAD_APICID) - return; + return -1; cfg = desc->chip_data; @@ -3387,13 +3343,15 @@ static void set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask) msg.address_lo |= MSI_ADDR_DEST_ID(dest); write_msi_msg_desc(desc, &msg); + + return 0; } #ifdef CONFIG_INTR_REMAP /* * Migrate the MSI irq to another cpumask. This migration is * done in the process context using interrupt-remapping hardware. */ -static void +static int ir_set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask) { struct irq_desc *desc = irq_to_desc(irq); @@ -3402,11 +3360,11 @@ ir_set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask) struct irte irte; if (get_irte(irq, &irte)) - return; + return -1; dest = set_desc_affinity(desc, mask); if (dest == BAD_APICID) - return; + return -1; irte.vector = cfg->vector; irte.dest_id = IRTE_DEST(dest); @@ -3423,6 +3381,8 @@ ir_set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask) */ if (cfg->move_in_progress) send_cleanup_vector(cfg); + + return 0; } #endif @@ -3518,15 +3478,17 @@ int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type) unsigned int irq_want; struct intel_iommu *iommu = NULL; int index = 0; + int node; /* x86 doesn't support multiple MSI yet */ if (type == PCI_CAP_ID_MSI && nvec > 1) return 1; + node = dev_to_node(&dev->dev); irq_want = nr_irqs_gsi; sub_handle = 0; list_for_each_entry(msidesc, &dev->msi_list, list) { - irq = create_irq_nr(irq_want); + irq = create_irq_nr(irq_want, node); if (irq == 0) return -1; irq_want = irq + 1; @@ -3576,7 +3538,7 @@ void arch_teardown_msi_irq(unsigned int irq) #if defined (CONFIG_DMAR) || defined (CONFIG_INTR_REMAP) #ifdef CONFIG_SMP -static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) +static int dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) { struct irq_desc *desc = irq_to_desc(irq); struct irq_cfg *cfg; @@ -3585,7 +3547,7 @@ static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) dest = set_desc_affinity(desc, mask); if (dest == BAD_APICID) - return; + return -1; cfg = desc->chip_data; @@ -3597,6 +3559,8 @@ static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) msg.address_lo |= MSI_ADDR_DEST_ID(dest); dmar_msi_write(irq, &msg); + + return 0; } #endif /* CONFIG_SMP */ @@ -3630,7 +3594,7 @@ int arch_setup_dmar_msi(unsigned int irq) #ifdef CONFIG_HPET_TIMER #ifdef CONFIG_SMP -static void hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask) +static int hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask) { struct irq_desc *desc = irq_to_desc(irq); struct irq_cfg *cfg; @@ -3639,7 +3603,7 @@ static void hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask) dest = set_desc_affinity(desc, mask); if (dest == BAD_APICID) - return; + return -1; cfg = desc->chip_data; @@ -3651,6 +3615,8 @@ static void hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask) msg.address_lo |= MSI_ADDR_DEST_ID(dest); hpet_msi_write(irq, &msg); + + return 0; } #endif /* CONFIG_SMP */ @@ -3707,7 +3673,7 @@ static void target_ht_irq(unsigned int irq, unsigned int dest, u8 vector) write_ht_irq_msg(irq, &msg); } -static void set_ht_irq_affinity(unsigned int irq, const struct cpumask *mask) +static int set_ht_irq_affinity(unsigned int irq, const struct cpumask *mask) { struct irq_desc *desc = irq_to_desc(irq); struct irq_cfg *cfg; @@ -3715,11 +3681,13 @@ static void set_ht_irq_affinity(unsigned int irq, const struct cpumask *mask) dest = set_desc_affinity(desc, mask); if (dest == BAD_APICID) - return; + return -1; cfg = desc->chip_data; target_ht_irq(irq, dest, cfg->vector); + + return 0; } #endif @@ -3794,6 +3762,8 @@ int arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade, unsigned long flags; int err; + BUILD_BUG_ON(sizeof(struct uv_IO_APIC_route_entry) != sizeof(unsigned long)); + cfg = irq_cfg(irq); err = assign_irq_vector(irq, cfg, eligible_cpu); @@ -3807,15 +3777,13 @@ int arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade, mmr_value = 0; entry = (struct uv_IO_APIC_route_entry *)&mmr_value; - BUG_ON(sizeof(struct uv_IO_APIC_route_entry) != sizeof(unsigned long)); - - entry->vector = cfg->vector; - entry->delivery_mode = apic->irq_delivery_mode; - entry->dest_mode = apic->irq_dest_mode; - entry->polarity = 0; - entry->trigger = 0; - entry->mask = 0; - entry->dest = apic->cpu_mask_to_apicid(eligible_cpu); + entry->vector = cfg->vector; + entry->delivery_mode = apic->irq_delivery_mode; + entry->dest_mode = apic->irq_dest_mode; + entry->polarity = 0; + entry->trigger = 0; + entry->mask = 0; + entry->dest = apic->cpu_mask_to_apicid(eligible_cpu); mmr_pnode = uv_blade_to_pnode(mmr_blade); uv_write_global_mmr64(mmr_pnode, mmr_offset, mmr_value); @@ -3833,10 +3801,10 @@ void arch_disable_uv_irq(int mmr_blade, unsigned long mmr_offset) struct uv_IO_APIC_route_entry *entry; int mmr_pnode; + BUILD_BUG_ON(sizeof(struct uv_IO_APIC_route_entry) != sizeof(unsigned long)); + mmr_value = 0; entry = (struct uv_IO_APIC_route_entry *)&mmr_value; - BUG_ON(sizeof(struct uv_IO_APIC_route_entry) != sizeof(unsigned long)); - entry->mask = 1; mmr_pnode = uv_blade_to_pnode(mmr_blade); @@ -3900,6 +3868,71 @@ int __init arch_probe_nr_irqs(void) } #endif +static int __io_apic_set_pci_routing(struct device *dev, int irq, + struct io_apic_irq_attr *irq_attr) +{ + struct irq_desc *desc; + struct irq_cfg *cfg; + int node; + int ioapic, pin; + int trigger, polarity; + + ioapic = irq_attr->ioapic; + if (!IO_APIC_IRQ(irq)) { + apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n", + ioapic); + return -EINVAL; + } + + if (dev) + node = dev_to_node(dev); + else + node = cpu_to_node(boot_cpu_id); + + desc = irq_to_desc_alloc_node(irq, node); + if (!desc) { + printk(KERN_INFO "can not get irq_desc %d\n", irq); + return 0; + } + + pin = irq_attr->ioapic_pin; + trigger = irq_attr->trigger; + polarity = irq_attr->polarity; + + /* + * IRQs < 16 are already in the irq_2_pin[] map + */ + if (irq >= NR_IRQS_LEGACY) { + cfg = desc->chip_data; + add_pin_to_irq_node(cfg, node, ioapic, pin); + } + + setup_IO_APIC_irq(ioapic, pin, irq, desc, trigger, polarity); + + return 0; +} + +int io_apic_set_pci_routing(struct device *dev, int irq, + struct io_apic_irq_attr *irq_attr) +{ + int ioapic, pin; + /* + * Avoid pin reprogramming. PRTs typically include entries + * with redundant pin->gsi mappings (but unique PCI devices); + * we only program the IOAPIC on the first. + */ + ioapic = irq_attr->ioapic; + pin = irq_attr->ioapic_pin; + if (test_bit(pin, mp_ioapic_routing[ioapic].pin_programmed)) { + pr_debug("Pin %d-%d already programmed\n", + mp_ioapics[ioapic].apicid, pin); + return 0; + } + set_bit(pin, mp_ioapic_routing[ioapic].pin_programmed); + + return __io_apic_set_pci_routing(dev, irq, irq_attr); +} + /* -------------------------------------------------------------------------- ACPI-based IOAPIC Configuration -------------------------------------------------------------------------- */ @@ -3980,6 +4013,7 @@ int __init io_apic_get_unique_id(int ioapic, int apic_id) return apic_id; } +#endif int __init io_apic_get_version(int ioapic) { @@ -3992,39 +4026,6 @@ int __init io_apic_get_version(int ioapic) return reg_01.bits.version; } -#endif - -int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity) -{ - struct irq_desc *desc; - struct irq_cfg *cfg; - int cpu = boot_cpu_id; - - if (!IO_APIC_IRQ(irq)) { - apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n", - ioapic); - return -EINVAL; - } - - desc = irq_to_desc_alloc_cpu(irq, cpu); - if (!desc) { - printk(KERN_INFO "can not get irq_desc %d\n", irq); - return 0; - } - - /* - * IRQs < 16 are already in the irq_2_pin[] map - */ - if (irq >= NR_IRQS_LEGACY) { - cfg = desc->chip_data; - add_pin_to_irq_cpu(cfg, cpu, ioapic, pin); - } - - setup_IO_APIC_irq(ioapic, pin, irq, desc, triggering, polarity); - - return 0; -} - int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity) { @@ -4055,51 +4056,44 @@ int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity) #ifdef CONFIG_SMP void __init setup_ioapic_dest(void) { - int pin, ioapic, irq, irq_entry; + int pin, ioapic = 0, irq, irq_entry; struct irq_desc *desc; - struct irq_cfg *cfg; const struct cpumask *mask; if (skip_ioapic_setup == 1) return; - for (ioapic = 0; ioapic < nr_ioapics; ioapic++) { - for (pin = 0; pin < nr_ioapic_registers[ioapic]; pin++) { - irq_entry = find_irq_entry(ioapic, pin, mp_INT); - if (irq_entry == -1) - continue; - irq = pin_2_irq(irq_entry, ioapic, pin); - - /* setup_IO_APIC_irqs could fail to get vector for some device - * when you have too many devices, because at that time only boot - * cpu is online. - */ - desc = irq_to_desc(irq); - cfg = desc->chip_data; - if (!cfg->vector) { - setup_IO_APIC_irq(ioapic, pin, irq, desc, - irq_trigger(irq_entry), - irq_polarity(irq_entry)); - continue; +#ifdef CONFIG_ACPI + if (!acpi_disabled && acpi_ioapic) { + ioapic = mp_find_ioapic(0); + if (ioapic < 0) + ioapic = 0; + } +#endif - } + for (pin = 0; pin < nr_ioapic_registers[ioapic]; pin++) { + irq_entry = find_irq_entry(ioapic, pin, mp_INT); + if (irq_entry == -1) + continue; + irq = pin_2_irq(irq_entry, ioapic, pin); - /* - * Honour affinities which have been set in early boot - */ - if (desc->status & - (IRQ_NO_BALANCING | IRQ_AFFINITY_SET)) - mask = desc->affinity; - else - mask = apic->target_cpus(); + desc = irq_to_desc(irq); - if (intr_remapping_enabled) - set_ir_ioapic_affinity_irq_desc(desc, mask); - else - set_ioapic_affinity_irq_desc(desc, mask); - } + /* + * Honour affinities which have been set in early boot + */ + if (desc->status & + (IRQ_NO_BALANCING | IRQ_AFFINITY_SET)) + mask = desc->affinity; + else + mask = apic->target_cpus(); + if (intr_remapping_enabled) + set_ir_ioapic_affinity_irq_desc(desc, mask); + else + set_ioapic_affinity_irq_desc(desc, mask); } + } #endif diff --git a/arch/x86/kernel/apic/nmi.c b/arch/x86/kernel/apic/nmi.c index ce4fbfa315a..a691302dc3f 100644 --- a/arch/x86/kernel/apic/nmi.c +++ b/arch/x86/kernel/apic/nmi.c @@ -104,7 +104,7 @@ static __init void nmi_cpu_busy(void *data) } #endif -static void report_broken_nmi(int cpu, int *prev_nmi_count) +static void report_broken_nmi(int cpu, unsigned int *prev_nmi_count) { printk(KERN_CONT "\n"); diff --git a/arch/x86/kernel/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c index 01eda2ac65e..440a8bccd91 100644 --- a/arch/x86/kernel/apic/probe_32.c +++ b/arch/x86/kernel/apic/probe_32.c @@ -160,7 +160,6 @@ extern struct apic apic_summit; extern struct apic apic_bigsmp; extern struct apic apic_es7000; extern struct apic apic_es7000_cluster; -extern struct apic apic_default; struct apic *apic = &apic_default; EXPORT_SYMBOL_GPL(apic); diff --git a/arch/x86/kernel/apic/probe_64.c b/arch/x86/kernel/apic/probe_64.c index 1783652bb0e..bc3e880f9b8 100644 --- a/arch/x86/kernel/apic/probe_64.c +++ b/arch/x86/kernel/apic/probe_64.c @@ -50,7 +50,7 @@ static struct apic *apic_probe[] __initdata = { void __init default_setup_apic_routing(void) { #ifdef CONFIG_X86_X2APIC - if (x2apic && (apic != &apic_x2apic_phys && + if (x2apic_mode && (apic != &apic_x2apic_phys && #ifdef CONFIG_X86_UV apic != &apic_x2apic_uv_x && #endif diff --git a/arch/x86/kernel/apic/summit_32.c b/arch/x86/kernel/apic/summit_32.c index 9cfe1f415d8..344eee4ac0a 100644 --- a/arch/x86/kernel/apic/summit_32.c +++ b/arch/x86/kernel/apic/summit_32.c @@ -173,13 +173,6 @@ static inline int is_WPEG(struct rio_detail *rio){ rio->type == LookOutAWPEG || rio->type == LookOutBWPEG); } - -/* In clustered mode, the high nibble of APIC ID is a cluster number. - * The low nibble is a 4-bit bitmap. */ -#define XAPIC_DEST_CPUS_SHIFT 4 -#define XAPIC_DEST_CPUS_MASK ((1u << XAPIC_DEST_CPUS_SHIFT) - 1) -#define XAPIC_DEST_CLUSTER_MASK (XAPIC_DEST_CPUS_MASK << XAPIC_DEST_CPUS_SHIFT) - #define SUMMIT_APIC_DFR_VALUE (APIC_DFR_CLUSTER) static const struct cpumask *summit_target_cpus(void) diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c index 4a903e2f0d1..8e4cbb255c3 100644 --- a/arch/x86/kernel/apic/x2apic_cluster.c +++ b/arch/x86/kernel/apic/x2apic_cluster.c @@ -10,7 +10,7 @@ #include <asm/apic.h> #include <asm/ipi.h> -DEFINE_PER_CPU(u32, x86_cpu_to_logical_apicid); +static DEFINE_PER_CPU(u32, x86_cpu_to_logical_apicid); static int x2apic_acpi_madt_oem_check(char *oem_id, char *oem_table_id) { diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 2bda6935297..ef0ae207a7c 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -105,7 +105,7 @@ static void uv_vector_allocation_domain(int cpu, struct cpumask *retmask) cpumask_set_cpu(cpu, retmask); } -static int uv_wakeup_secondary(int phys_apicid, unsigned long start_rip) +static int __cpuinit uv_wakeup_secondary(int phys_apicid, unsigned long start_rip) { #ifdef CONFIG_SMP unsigned long val; @@ -562,7 +562,7 @@ void __init uv_system_init(void) union uvh_node_id_u node_id; unsigned long gnode_upper, lowmem_redir_base, lowmem_redir_size; int bytes, nid, cpu, lcpu, pnode, blade, i, j, m_val, n_val; - int max_pnode = 0; + int gnode_extra, max_pnode = 0; unsigned long mmr_base, present, paddr; unsigned short pnode_mask; @@ -574,6 +574,13 @@ void __init uv_system_init(void) mmr_base = uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR) & ~UV_MMR_ENABLE; + pnode_mask = (1 << n_val) - 1; + node_id.v = uv_read_local_mmr(UVH_NODE_ID); + gnode_extra = (node_id.s.node_id & ~((1 << n_val) - 1)) >> 1; + gnode_upper = ((unsigned long)gnode_extra << m_val); + printk(KERN_DEBUG "UV: N %d, M %d, gnode_upper 0x%lx, gnode_extra 0x%x\n", + n_val, m_val, gnode_upper, gnode_extra); + printk(KERN_DEBUG "UV: global MMR base 0x%lx\n", mmr_base); for(i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++) @@ -583,15 +590,18 @@ void __init uv_system_init(void) bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades(); uv_blade_info = kmalloc(bytes, GFP_KERNEL); + BUG_ON(!uv_blade_info); get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size); bytes = sizeof(uv_node_to_blade[0]) * num_possible_nodes(); uv_node_to_blade = kmalloc(bytes, GFP_KERNEL); + BUG_ON(!uv_node_to_blade); memset(uv_node_to_blade, 255, bytes); bytes = sizeof(uv_cpu_to_blade[0]) * num_possible_cpus(); uv_cpu_to_blade = kmalloc(bytes, GFP_KERNEL); + BUG_ON(!uv_cpu_to_blade); memset(uv_cpu_to_blade, 255, bytes); blade = 0; @@ -607,11 +617,6 @@ void __init uv_system_init(void) } } - pnode_mask = (1 << n_val) - 1; - node_id.v = uv_read_local_mmr(UVH_NODE_ID); - gnode_upper = (((unsigned long)node_id.s.node_id) & - ~((1 << n_val) - 1)) << m_val; - uv_bios_init(); uv_bios_get_sn_info(0, &uv_type, &sn_partition_id, &sn_coherency_id, &sn_region_size); @@ -634,6 +639,7 @@ void __init uv_system_init(void) uv_cpu_hub_info(cpu)->pnode_mask = pnode_mask; uv_cpu_hub_info(cpu)->gpa_mask = (1 << (m_val + n_val)) - 1; uv_cpu_hub_info(cpu)->gnode_upper = gnode_upper; + uv_cpu_hub_info(cpu)->gnode_extra = gnode_extra; uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base; uv_cpu_hub_info(cpu)->coherency_domain_number = sn_coherency_id; uv_cpu_hub_info(cpu)->scir.offset = SCIR_LOCAL_MMR_BASE + lcpu; diff --git a/arch/x86/kernel/asm-offsets_32.c b/arch/x86/kernel/asm-offsets_32.c index 5a6aa1c1162..1a830cbd701 100644 --- a/arch/x86/kernel/asm-offsets_32.c +++ b/arch/x86/kernel/asm-offsets_32.c @@ -146,4 +146,5 @@ void foo(void) OFFSET(BP_loadflags, boot_params, hdr.loadflags); OFFSET(BP_hardware_subarch, boot_params, hdr.hardware_subarch); OFFSET(BP_version, boot_params, hdr.version); + OFFSET(BP_kernel_alignment, boot_params, hdr.kernel_alignment); } diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c index e72f062fb4b..898ecc47e12 100644 --- a/arch/x86/kernel/asm-offsets_64.c +++ b/arch/x86/kernel/asm-offsets_64.c @@ -125,6 +125,7 @@ int main(void) OFFSET(BP_loadflags, boot_params, hdr.loadflags); OFFSET(BP_hardware_subarch, boot_params, hdr.hardware_subarch); OFFSET(BP_version, boot_params, hdr.version); + OFFSET(BP_kernel_alignment, boot_params, hdr.kernel_alignment); BLANK(); DEFINE(PAGE_SIZE_asm, PAGE_SIZE); diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index 4e242f9a06e..3efcb2b96a1 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -1,5 +1,5 @@ # -# Makefile for x86-compatible CPU details and quirks +# Makefile for x86-compatible CPU details, features and quirks # # Don't trace early stages of a secondary CPU boot @@ -23,11 +23,13 @@ obj-$(CONFIG_CPU_SUP_CENTAUR) += centaur.o obj-$(CONFIG_CPU_SUP_TRANSMETA_32) += transmeta.o obj-$(CONFIG_CPU_SUP_UMC_32) += umc.o -obj-$(CONFIG_X86_MCE) += mcheck/ -obj-$(CONFIG_MTRR) += mtrr/ -obj-$(CONFIG_CPU_FREQ) += cpufreq/ +obj-$(CONFIG_PERF_COUNTERS) += perf_counter.o -obj-$(CONFIG_X86_LOCAL_APIC) += perfctr-watchdog.o +obj-$(CONFIG_X86_MCE) += mcheck/ +obj-$(CONFIG_MTRR) += mtrr/ +obj-$(CONFIG_CPU_FREQ) += cpufreq/ + +obj-$(CONFIG_X86_LOCAL_APIC) += perfctr-watchdog.o quiet_cmd_mkcapflags = MKCAP $@ cmd_mkcapflags = $(PERL) $(srctree)/$(src)/mkcapflags.pl $< $@ diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 7e4a459daa6..e5b27d8f1b4 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -6,6 +6,7 @@ #include <asm/processor.h> #include <asm/apic.h> #include <asm/cpu.h> +#include <asm/pci-direct.h> #ifdef CONFIG_X86_64 # include <asm/numa_64.h> @@ -272,7 +273,7 @@ static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) #if defined(CONFIG_NUMA) && defined(CONFIG_X86_64) int cpu = smp_processor_id(); int node; - unsigned apicid = hard_smp_processor_id(); + unsigned apicid = cpu_has_apic ? hard_smp_processor_id() : c->apicid; node = c->phys_proc_id; if (apicid_to_node[apicid] != NUMA_NO_NODE) @@ -351,6 +352,15 @@ static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) (c->x86_model == 8 && c->x86_mask >= 8)) set_cpu_cap(c, X86_FEATURE_K6_MTRR); #endif +#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_PCI) + /* check CPU config space for extended APIC ID */ + if (c->x86 >= 0xf) { + unsigned int val; + val = read_pci_config(0, 24, 0, 0x68); + if ((val & ((1 << 17) | (1 << 18))) == ((1 << 17) | (1 << 18))) + set_cpu_cap(c, X86_FEATURE_EXTD_APICID); + } +#endif } static void __cpuinit init_amd(struct cpuinfo_x86 *c) diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 77848d9fca6..3ffdcfa9abd 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -13,6 +13,7 @@ #include <linux/io.h> #include <asm/stackprotector.h> +#include <asm/perf_counter.h> #include <asm/mmu_context.h> #include <asm/hypervisor.h> #include <asm/processor.h> @@ -299,7 +300,8 @@ static const char *__cpuinit table_lookup_model(struct cpuinfo_x86 *c) return NULL; /* Not found */ } -__u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata; +__u32 cpu_caps_cleared[NCAPINTS] __cpuinitdata; +__u32 cpu_caps_set[NCAPINTS] __cpuinitdata; void load_percpu_segment(int cpu) { @@ -768,6 +770,12 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) if (this_cpu->c_identify) this_cpu->c_identify(c); + /* Clear/Set all flags overriden by options, after probe */ + for (i = 0; i < NCAPINTS; i++) { + c->x86_capability[i] &= ~cpu_caps_cleared[i]; + c->x86_capability[i] |= cpu_caps_set[i]; + } + #ifdef CONFIG_X86_64 c->apicid = apic->phys_pkg_id(c->initial_apicid, 0); #endif @@ -813,6 +821,16 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) #endif init_hypervisor(c); + + /* + * Clear/Set all flags overriden by options, need do it + * before following smp all cpus cap AND. + */ + for (i = 0; i < NCAPINTS; i++) { + c->x86_capability[i] &= ~cpu_caps_cleared[i]; + c->x86_capability[i] |= cpu_caps_set[i]; + } + /* * On SMP, boot_cpu_data holds the common feature set between * all CPUs; so make sure that we indicate which features are @@ -825,10 +843,6 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) boot_cpu_data.x86_capability[i] &= c->x86_capability[i]; } - /* Clear all flags overriden by options */ - for (i = 0; i < NCAPINTS; i++) - c->x86_capability[i] &= ~cleared_cpu_caps[i]; - #ifdef CONFIG_X86_MCE /* Init Machine Check Exception if available. */ mcheck_init(c); @@ -861,6 +875,7 @@ void __init identify_boot_cpu(void) #else vgetcpu_set_mode(); #endif + init_hw_perf_counters(); } void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c) diff --git a/arch/x86/kernel/cpu/cpu_debug.c b/arch/x86/kernel/cpu/cpu_debug.c index 46e29ab96c6..6b2a52dd040 100644 --- a/arch/x86/kernel/cpu/cpu_debug.c +++ b/arch/x86/kernel/cpu/cpu_debug.c @@ -32,9 +32,7 @@ static DEFINE_PER_CPU(struct cpu_cpuX_base, cpu_arr[CPU_REG_ALL_BIT]); static DEFINE_PER_CPU(struct cpu_private *, priv_arr[MAX_CPU_FILES]); -static DEFINE_PER_CPU(unsigned, cpu_modelflag); static DEFINE_PER_CPU(int, cpu_priv_count); -static DEFINE_PER_CPU(unsigned, cpu_model); static DEFINE_MUTEX(cpu_debug_lock); @@ -80,302 +78,102 @@ static struct cpu_file_base cpu_file[] = { { "value", CPU_REG_ALL, 1 }, }; -/* Intel Registers Range */ -static struct cpu_debug_range cpu_intel_range[] = { - { 0x00000000, 0x00000001, CPU_MC, CPU_INTEL_ALL }, - { 0x00000006, 0x00000007, CPU_MONITOR, CPU_CX_AT_XE }, - { 0x00000010, 0x00000010, CPU_TIME, CPU_INTEL_ALL }, - { 0x00000011, 0x00000013, CPU_PMC, CPU_INTEL_PENTIUM }, - { 0x00000017, 0x00000017, CPU_PLATFORM, CPU_PX_CX_AT_XE }, - { 0x0000001B, 0x0000001B, CPU_APIC, CPU_P6_CX_AT_XE }, - - { 0x0000002A, 0x0000002A, CPU_POWERON, CPU_PX_CX_AT_XE }, - { 0x0000002B, 0x0000002B, CPU_POWERON, CPU_INTEL_XEON }, - { 0x0000002C, 0x0000002C, CPU_FREQ, CPU_INTEL_XEON }, - { 0x0000003A, 0x0000003A, CPU_CONTROL, CPU_CX_AT_XE }, - - { 0x00000040, 0x00000043, CPU_LBRANCH, CPU_PM_CX_AT_XE }, - { 0x00000044, 0x00000047, CPU_LBRANCH, CPU_PM_CO_AT }, - { 0x00000060, 0x00000063, CPU_LBRANCH, CPU_C2_AT }, - { 0x00000064, 0x00000067, CPU_LBRANCH, CPU_INTEL_ATOM }, - - { 0x00000079, 0x00000079, CPU_BIOS, CPU_P6_CX_AT_XE }, - { 0x00000088, 0x0000008A, CPU_CACHE, CPU_INTEL_P6 }, - { 0x0000008B, 0x0000008B, CPU_BIOS, CPU_P6_CX_AT_XE }, - { 0x0000009B, 0x0000009B, CPU_MONITOR, CPU_INTEL_XEON }, - - { 0x000000C1, 0x000000C2, CPU_PMC, CPU_P6_CX_AT }, - { 0x000000CD, 0x000000CD, CPU_FREQ, CPU_CX_AT }, - { 0x000000E7, 0x000000E8, CPU_PERF, CPU_CX_AT }, - { 0x000000FE, 0x000000FE, CPU_MTRR, CPU_P6_CX_XE }, - - { 0x00000116, 0x00000116, CPU_CACHE, CPU_INTEL_P6 }, - { 0x00000118, 0x00000118, CPU_CACHE, CPU_INTEL_P6 }, - { 0x00000119, 0x00000119, CPU_CACHE, CPU_INTEL_PX }, - { 0x0000011A, 0x0000011B, CPU_CACHE, CPU_INTEL_P6 }, - { 0x0000011E, 0x0000011E, CPU_CACHE, CPU_PX_CX_AT }, - - { 0x00000174, 0x00000176, CPU_SYSENTER, CPU_P6_CX_AT_XE }, - { 0x00000179, 0x0000017A, CPU_MC, CPU_PX_CX_AT_XE }, - { 0x0000017B, 0x0000017B, CPU_MC, CPU_P6_XE }, - { 0x00000186, 0x00000187, CPU_PMC, CPU_P6_CX_AT }, - { 0x00000198, 0x00000199, CPU_PERF, CPU_PM_CX_AT_XE }, - { 0x0000019A, 0x0000019A, CPU_TIME, CPU_PM_CX_AT_XE }, - { 0x0000019B, 0x0000019D, CPU_THERM, CPU_PM_CX_AT_XE }, - { 0x000001A0, 0x000001A0, CPU_MISC, CPU_PM_CX_AT_XE }, - - { 0x000001C9, 0x000001C9, CPU_LBRANCH, CPU_PM_CX_AT }, - { 0x000001D7, 0x000001D8, CPU_LBRANCH, CPU_INTEL_XEON }, - { 0x000001D9, 0x000001D9, CPU_DEBUG, CPU_CX_AT_XE }, - { 0x000001DA, 0x000001DA, CPU_LBRANCH, CPU_INTEL_XEON }, - { 0x000001DB, 0x000001DB, CPU_LBRANCH, CPU_P6_XE }, - { 0x000001DC, 0x000001DC, CPU_LBRANCH, CPU_INTEL_P6 }, - { 0x000001DD, 0x000001DE, CPU_LBRANCH, CPU_PX_CX_AT_XE }, - { 0x000001E0, 0x000001E0, CPU_LBRANCH, CPU_INTEL_P6 }, - - { 0x00000200, 0x0000020F, CPU_MTRR, CPU_P6_CX_XE }, - { 0x00000250, 0x00000250, CPU_MTRR, CPU_P6_CX_XE }, - { 0x00000258, 0x00000259, CPU_MTRR, CPU_P6_CX_XE }, - { 0x00000268, 0x0000026F, CPU_MTRR, CPU_P6_CX_XE }, - { 0x00000277, 0x00000277, CPU_PAT, CPU_C2_AT_XE }, - { 0x000002FF, 0x000002FF, CPU_MTRR, CPU_P6_CX_XE }, - - { 0x00000300, 0x00000308, CPU_PMC, CPU_INTEL_XEON }, - { 0x00000309, 0x0000030B, CPU_PMC, CPU_C2_AT_XE }, - { 0x0000030C, 0x00000311, CPU_PMC, CPU_INTEL_XEON }, - { 0x00000345, 0x00000345, CPU_PMC, CPU_C2_AT }, - { 0x00000360, 0x00000371, CPU_PMC, CPU_INTEL_XEON }, - { 0x0000038D, 0x00000390, CPU_PMC, CPU_C2_AT }, - { 0x000003A0, 0x000003BE, CPU_PMC, CPU_INTEL_XEON }, - { 0x000003C0, 0x000003CD, CPU_PMC, CPU_INTEL_XEON }, - { 0x000003E0, 0x000003E1, CPU_PMC, CPU_INTEL_XEON }, - { 0x000003F0, 0x000003F0, CPU_PMC, CPU_INTEL_XEON }, - { 0x000003F1, 0x000003F1, CPU_PMC, CPU_C2_AT_XE }, - { 0x000003F2, 0x000003F2, CPU_PMC, CPU_INTEL_XEON }, - - { 0x00000400, 0x00000402, CPU_MC, CPU_PM_CX_AT_XE }, - { 0x00000403, 0x00000403, CPU_MC, CPU_INTEL_XEON }, - { 0x00000404, 0x00000406, CPU_MC, CPU_PM_CX_AT_XE }, - { 0x00000407, 0x00000407, CPU_MC, CPU_INTEL_XEON }, - { 0x00000408, 0x0000040A, CPU_MC, CPU_PM_CX_AT_XE }, - { 0x0000040B, 0x0000040B, CPU_MC, CPU_INTEL_XEON }, - { 0x0000040C, 0x0000040E, CPU_MC, CPU_PM_CX_XE }, - { 0x0000040F, 0x0000040F, CPU_MC, CPU_INTEL_XEON }, - { 0x00000410, 0x00000412, CPU_MC, CPU_PM_CX_AT_XE }, - { 0x00000413, 0x00000417, CPU_MC, CPU_CX_AT_XE }, - { 0x00000480, 0x0000048B, CPU_VMX, CPU_CX_AT_XE }, - - { 0x00000600, 0x00000600, CPU_DEBUG, CPU_PM_CX_AT_XE }, - { 0x00000680, 0x0000068F, CPU_LBRANCH, CPU_INTEL_XEON }, - { 0x000006C0, 0x000006CF, CPU_LBRANCH, CPU_INTEL_XEON }, - - { 0x000107CC, 0x000107D3, CPU_PMC, CPU_INTEL_XEON_MP }, - - { 0xC0000080, 0xC0000080, CPU_FEATURES, CPU_INTEL_XEON }, - { 0xC0000081, 0xC0000082, CPU_CALL, CPU_INTEL_XEON }, - { 0xC0000084, 0xC0000084, CPU_CALL, CPU_INTEL_XEON }, - { 0xC0000100, 0xC0000102, CPU_BASE, CPU_INTEL_XEON }, +/* CPU Registers Range */ +static struct cpu_debug_range cpu_reg_range[] = { + { 0x00000000, 0x00000001, CPU_MC, }, + { 0x00000006, 0x00000007, CPU_MONITOR, }, + { 0x00000010, 0x00000010, CPU_TIME, }, + { 0x00000011, 0x00000013, CPU_PMC, }, + { 0x00000017, 0x00000017, CPU_PLATFORM, }, + { 0x0000001B, 0x0000001B, CPU_APIC, }, + { 0x0000002A, 0x0000002B, CPU_POWERON, }, + { 0x0000002C, 0x0000002C, CPU_FREQ, }, + { 0x0000003A, 0x0000003A, CPU_CONTROL, }, + { 0x00000040, 0x00000047, CPU_LBRANCH, }, + { 0x00000060, 0x00000067, CPU_LBRANCH, }, + { 0x00000079, 0x00000079, CPU_BIOS, }, + { 0x00000088, 0x0000008A, CPU_CACHE, }, + { 0x0000008B, 0x0000008B, CPU_BIOS, }, + { 0x0000009B, 0x0000009B, CPU_MONITOR, }, + { 0x000000C1, 0x000000C4, CPU_PMC, }, + { 0x000000CD, 0x000000CD, CPU_FREQ, }, + { 0x000000E7, 0x000000E8, CPU_PERF, }, + { 0x000000FE, 0x000000FE, CPU_MTRR, }, + + { 0x00000116, 0x0000011E, CPU_CACHE, }, + { 0x00000174, 0x00000176, CPU_SYSENTER, }, + { 0x00000179, 0x0000017B, CPU_MC, }, + { 0x00000186, 0x00000189, CPU_PMC, }, + { 0x00000198, 0x00000199, CPU_PERF, }, + { 0x0000019A, 0x0000019A, CPU_TIME, }, + { 0x0000019B, 0x0000019D, CPU_THERM, }, + { 0x000001A0, 0x000001A0, CPU_MISC, }, + { 0x000001C9, 0x000001C9, CPU_LBRANCH, }, + { 0x000001D7, 0x000001D8, CPU_LBRANCH, }, + { 0x000001D9, 0x000001D9, CPU_DEBUG, }, + { 0x000001DA, 0x000001E0, CPU_LBRANCH, }, + + { 0x00000200, 0x0000020F, CPU_MTRR, }, + { 0x00000250, 0x00000250, CPU_MTRR, }, + { 0x00000258, 0x00000259, CPU_MTRR, }, + { 0x00000268, 0x0000026F, CPU_MTRR, }, + { 0x00000277, 0x00000277, CPU_PAT, }, + { 0x000002FF, 0x000002FF, CPU_MTRR, }, + + { 0x00000300, 0x00000311, CPU_PMC, }, + { 0x00000345, 0x00000345, CPU_PMC, }, + { 0x00000360, 0x00000371, CPU_PMC, }, + { 0x0000038D, 0x00000390, CPU_PMC, }, + { 0x000003A0, 0x000003BE, CPU_PMC, }, + { 0x000003C0, 0x000003CD, CPU_PMC, }, + { 0x000003E0, 0x000003E1, CPU_PMC, }, + { 0x000003F0, 0x000003F2, CPU_PMC, }, + + { 0x00000400, 0x00000417, CPU_MC, }, + { 0x00000480, 0x0000048B, CPU_VMX, }, + + { 0x00000600, 0x00000600, CPU_DEBUG, }, + { 0x00000680, 0x0000068F, CPU_LBRANCH, }, + { 0x000006C0, 0x000006CF, CPU_LBRANCH, }, + + { 0x000107CC, 0x000107D3, CPU_PMC, }, + + { 0xC0000080, 0xC0000080, CPU_FEATURES, }, + { 0xC0000081, 0xC0000084, CPU_CALL, }, + { 0xC0000100, 0xC0000102, CPU_BASE, }, + { 0xC0000103, 0xC0000103, CPU_TIME, }, + + { 0xC0010000, 0xC0010007, CPU_PMC, }, + { 0xC0010010, 0xC0010010, CPU_CONF, }, + { 0xC0010015, 0xC0010015, CPU_CONF, }, + { 0xC0010016, 0xC001001A, CPU_MTRR, }, + { 0xC001001D, 0xC001001D, CPU_MTRR, }, + { 0xC001001F, 0xC001001F, CPU_CONF, }, + { 0xC0010030, 0xC0010035, CPU_BIOS, }, + { 0xC0010044, 0xC0010048, CPU_MC, }, + { 0xC0010050, 0xC0010056, CPU_SMM, }, + { 0xC0010058, 0xC0010058, CPU_CONF, }, + { 0xC0010060, 0xC0010060, CPU_CACHE, }, + { 0xC0010061, 0xC0010068, CPU_SMM, }, + { 0xC0010069, 0xC001006B, CPU_SMM, }, + { 0xC0010070, 0xC0010071, CPU_SMM, }, + { 0xC0010111, 0xC0010113, CPU_SMM, }, + { 0xC0010114, 0xC0010118, CPU_SVM, }, + { 0xC0010140, 0xC0010141, CPU_OSVM, }, + { 0xC0011022, 0xC0011023, CPU_CONF, }, }; -/* AMD Registers Range */ -static struct cpu_debug_range cpu_amd_range[] = { - { 0x00000000, 0x00000001, CPU_MC, CPU_K10_PLUS, }, - { 0x00000010, 0x00000010, CPU_TIME, CPU_K8_PLUS, }, - { 0x0000001B, 0x0000001B, CPU_APIC, CPU_K8_PLUS, }, - { 0x0000002A, 0x0000002A, CPU_POWERON, CPU_K7_PLUS }, - { 0x0000008B, 0x0000008B, CPU_VER, CPU_K8_PLUS }, - { 0x000000FE, 0x000000FE, CPU_MTRR, CPU_K8_PLUS, }, - - { 0x00000174, 0x00000176, CPU_SYSENTER, CPU_K8_PLUS, }, - { 0x00000179, 0x0000017B, CPU_MC, CPU_K8_PLUS, }, - { 0x000001D9, 0x000001D9, CPU_DEBUG, CPU_K8_PLUS, }, - { 0x000001DB, 0x000001DE, CPU_LBRANCH, CPU_K8_PLUS, }, - - { 0x00000200, 0x0000020F, CPU_MTRR, CPU_K8_PLUS, }, - { 0x00000250, 0x00000250, CPU_MTRR, CPU_K8_PLUS, }, - { 0x00000258, 0x00000259, CPU_MTRR, CPU_K8_PLUS, }, - { 0x00000268, 0x0000026F, CPU_MTRR, CPU_K8_PLUS, }, - { 0x00000277, 0x00000277, CPU_PAT, CPU_K8_PLUS, }, - { 0x000002FF, 0x000002FF, CPU_MTRR, CPU_K8_PLUS, }, - - { 0x00000400, 0x00000413, CPU_MC, CPU_K8_PLUS, }, - - { 0xC0000080, 0xC0000080, CPU_FEATURES, CPU_AMD_ALL, }, - { 0xC0000081, 0xC0000084, CPU_CALL, CPU_K8_PLUS, }, - { 0xC0000100, 0xC0000102, CPU_BASE, CPU_K8_PLUS, }, - { 0xC0000103, 0xC0000103, CPU_TIME, CPU_K10_PLUS, }, - - { 0xC0010000, 0xC0010007, CPU_PMC, CPU_K8_PLUS, }, - { 0xC0010010, 0xC0010010, CPU_CONF, CPU_K7_PLUS, }, - { 0xC0010015, 0xC0010015, CPU_CONF, CPU_K7_PLUS, }, - { 0xC0010016, 0xC001001A, CPU_MTRR, CPU_K8_PLUS, }, - { 0xC001001D, 0xC001001D, CPU_MTRR, CPU_K8_PLUS, }, - { 0xC001001F, 0xC001001F, CPU_CONF, CPU_K8_PLUS, }, - { 0xC0010030, 0xC0010035, CPU_BIOS, CPU_K8_PLUS, }, - { 0xC0010044, 0xC0010048, CPU_MC, CPU_K8_PLUS, }, - { 0xC0010050, 0xC0010056, CPU_SMM, CPU_K0F_PLUS, }, - { 0xC0010058, 0xC0010058, CPU_CONF, CPU_K10_PLUS, }, - { 0xC0010060, 0xC0010060, CPU_CACHE, CPU_AMD_11, }, - { 0xC0010061, 0xC0010068, CPU_SMM, CPU_K10_PLUS, }, - { 0xC0010069, 0xC001006B, CPU_SMM, CPU_AMD_11, }, - { 0xC0010070, 0xC0010071, CPU_SMM, CPU_K10_PLUS, }, - { 0xC0010111, 0xC0010113, CPU_SMM, CPU_K8_PLUS, }, - { 0xC0010114, 0xC0010118, CPU_SVM, CPU_K10_PLUS, }, - { 0xC0010140, 0xC0010141, CPU_OSVM, CPU_K10_PLUS, }, - { 0xC0011022, 0xC0011023, CPU_CONF, CPU_K10_PLUS, }, -}; - - -/* Intel */ -static int get_intel_modelflag(unsigned model) -{ - int flag; - - switch (model) { - case 0x0501: - case 0x0502: - case 0x0504: - flag = CPU_INTEL_PENTIUM; - break; - case 0x0601: - case 0x0603: - case 0x0605: - case 0x0607: - case 0x0608: - case 0x060A: - case 0x060B: - flag = CPU_INTEL_P6; - break; - case 0x0609: - case 0x060D: - flag = CPU_INTEL_PENTIUM_M; - break; - case 0x060E: - flag = CPU_INTEL_CORE; - break; - case 0x060F: - case 0x0617: - flag = CPU_INTEL_CORE2; - break; - case 0x061C: - flag = CPU_INTEL_ATOM; - break; - case 0x0F00: - case 0x0F01: - case 0x0F02: - case 0x0F03: - case 0x0F04: - flag = CPU_INTEL_XEON_P4; - break; - case 0x0F06: - flag = CPU_INTEL_XEON_MP; - break; - default: - flag = CPU_NONE; - break; - } - - return flag; -} - -/* AMD */ -static int get_amd_modelflag(unsigned model) -{ - int flag; - - switch (model >> 8) { - case 0x6: - flag = CPU_AMD_K6; - break; - case 0x7: - flag = CPU_AMD_K7; - break; - case 0x8: - flag = CPU_AMD_K8; - break; - case 0xf: - flag = CPU_AMD_0F; - break; - case 0x10: - flag = CPU_AMD_10; - break; - case 0x11: - flag = CPU_AMD_11; - break; - default: - flag = CPU_NONE; - break; - } - - return flag; -} - -static int get_cpu_modelflag(unsigned cpu) -{ - int flag; - - flag = per_cpu(cpu_model, cpu); - - switch (flag >> 16) { - case X86_VENDOR_INTEL: - flag = get_intel_modelflag(flag); - break; - case X86_VENDOR_AMD: - flag = get_amd_modelflag(flag & 0xffff); - break; - default: - flag = CPU_NONE; - break; - } - - return flag; -} - -static int get_cpu_range_count(unsigned cpu) -{ - int index; - - switch (per_cpu(cpu_model, cpu) >> 16) { - case X86_VENDOR_INTEL: - index = ARRAY_SIZE(cpu_intel_range); - break; - case X86_VENDOR_AMD: - index = ARRAY_SIZE(cpu_amd_range); - break; - default: - index = 0; - break; - } - - return index; -} - static int is_typeflag_valid(unsigned cpu, unsigned flag) { - unsigned vendor, modelflag; - int i, index; + int i; /* Standard Registers should be always valid */ if (flag >= CPU_TSS) return 1; - modelflag = per_cpu(cpu_modelflag, cpu); - vendor = per_cpu(cpu_model, cpu) >> 16; - index = get_cpu_range_count(cpu); - - for (i = 0; i < index; i++) { - switch (vendor) { - case X86_VENDOR_INTEL: - if ((cpu_intel_range[i].model & modelflag) && - (cpu_intel_range[i].flag & flag)) - return 1; - break; - case X86_VENDOR_AMD: - if ((cpu_amd_range[i].model & modelflag) && - (cpu_amd_range[i].flag & flag)) - return 1; - break; - } + for (i = 0; i < ARRAY_SIZE(cpu_reg_range); i++) { + if (cpu_reg_range[i].flag == flag) + return 1; } /* Invalid */ @@ -385,26 +183,11 @@ static int is_typeflag_valid(unsigned cpu, unsigned flag) static unsigned get_cpu_range(unsigned cpu, unsigned *min, unsigned *max, int index, unsigned flag) { - unsigned modelflag; - - modelflag = per_cpu(cpu_modelflag, cpu); - *max = 0; - switch (per_cpu(cpu_model, cpu) >> 16) { - case X86_VENDOR_INTEL: - if ((cpu_intel_range[index].model & modelflag) && - (cpu_intel_range[index].flag & flag)) { - *min = cpu_intel_range[index].min; - *max = cpu_intel_range[index].max; - } - break; - case X86_VENDOR_AMD: - if ((cpu_amd_range[index].model & modelflag) && - (cpu_amd_range[index].flag & flag)) { - *min = cpu_amd_range[index].min; - *max = cpu_amd_range[index].max; - } - break; - } + if (cpu_reg_range[index].flag == flag) { + *min = cpu_reg_range[index].min; + *max = cpu_reg_range[index].max; + } else + *max = 0; return *max; } @@ -434,7 +217,7 @@ static void print_msr(struct seq_file *seq, unsigned cpu, unsigned flag) unsigned msr, msr_min, msr_max; struct cpu_private *priv; u32 low, high; - int i, range; + int i; if (seq) { priv = seq->private; @@ -446,9 +229,7 @@ static void print_msr(struct seq_file *seq, unsigned cpu, unsigned flag) } } - range = get_cpu_range_count(cpu); - - for (i = 0; i < range; i++) { + for (i = 0; i < ARRAY_SIZE(cpu_reg_range); i++) { if (!get_cpu_range(cpu, &msr_min, &msr_max, i, flag)) continue; @@ -588,8 +369,20 @@ static void print_apic(void *arg) seq_printf(seq, " TMICT\t\t: %08x\n", apic_read(APIC_TMICT)); seq_printf(seq, " TMCCT\t\t: %08x\n", apic_read(APIC_TMCCT)); seq_printf(seq, " TDCR\t\t: %08x\n", apic_read(APIC_TDCR)); -#endif /* CONFIG_X86_LOCAL_APIC */ + if (boot_cpu_has(X86_FEATURE_EXTAPIC)) { + unsigned int i, v, maxeilvt; + + v = apic_read(APIC_EFEAT); + maxeilvt = (v >> 16) & 0xff; + seq_printf(seq, " EFEAT\t\t: %08x\n", v); + seq_printf(seq, " ECTRL\t\t: %08x\n", apic_read(APIC_ECTRL)); + for (i = 0; i < maxeilvt; i++) { + v = apic_read(APIC_EILVTn(i)); + seq_printf(seq, " EILVT%d\t\t: %08x\n", i, v); + } + } +#endif /* CONFIG_X86_LOCAL_APIC */ seq_printf(seq, "\n MSR\t:\n"); } @@ -788,13 +581,11 @@ static int cpu_init_msr(unsigned cpu, unsigned type, struct dentry *dentry) { struct dentry *cpu_dentry = NULL; unsigned reg, reg_min, reg_max; - int i, range, err = 0; + int i, err = 0; char reg_dir[12]; u32 low, high; - range = get_cpu_range_count(cpu); - - for (i = 0; i < range; i++) { + for (i = 0; i < ARRAY_SIZE(cpu_reg_range); i++) { if (!get_cpu_range(cpu, ®_min, ®_max, i, cpu_base[type].flag)) continue; @@ -850,10 +641,6 @@ static int cpu_init_cpu(void) cpui = &cpu_data(cpu); if (!cpu_has(cpui, X86_FEATURE_MSR)) continue; - per_cpu(cpu_model, cpu) = ((cpui->x86_vendor << 16) | - (cpui->x86 << 8) | - (cpui->x86_model)); - per_cpu(cpu_modelflag, cpu) = get_cpu_modelflag(cpu); sprintf(cpu_dir, "cpu%d", cpu); cpu_dentry = debugfs_create_dir(cpu_dir, cpu_debugfs_dir); diff --git a/arch/x86/kernel/cpu/cpufreq/Kconfig b/arch/x86/kernel/cpu/cpufreq/Kconfig index 52c83987547..f138c6c389b 100644 --- a/arch/x86/kernel/cpu/cpufreq/Kconfig +++ b/arch/x86/kernel/cpu/cpufreq/Kconfig @@ -220,11 +220,14 @@ config X86_LONGHAUL If in doubt, say N. config X86_E_POWERSAVER - tristate "VIA C7 Enhanced PowerSaver" + tristate "VIA C7 Enhanced PowerSaver (DANGEROUS)" select CPU_FREQ_TABLE - depends on X86_32 + depends on X86_32 && EXPERIMENTAL help - This adds the CPUFreq driver for VIA C7 processors. + This adds the CPUFreq driver for VIA C7 processors. However, this driver + does not have any safeguards to prevent operating the CPU out of spec + and is thus considered dangerous. Please use the regular ACPI cpufreq + driver, enabled by CONFIG_X86_ACPI_CPUFREQ. If in doubt, say N. diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c index 752e8c6b2c7..ae9b503220c 100644 --- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -90,11 +90,7 @@ static int check_est_cpu(unsigned int cpuid) { struct cpuinfo_x86 *cpu = &cpu_data(cpuid); - if (cpu->x86_vendor != X86_VENDOR_INTEL || - !cpu_has(cpu, X86_FEATURE_EST)) - return 0; - - return 1; + return cpu_has(cpu, X86_FEATURE_EST); } static unsigned extract_io(u32 value, struct acpi_cpufreq_data *data) diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 7437fa133c0..daed39ba261 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -229,12 +229,12 @@ static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) } #endif -static void __cpuinit srat_detect_node(void) +static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) { #if defined(CONFIG_NUMA) && defined(CONFIG_X86_64) unsigned node; int cpu = smp_processor_id(); - int apicid = hard_smp_processor_id(); + int apicid = cpu_has_apic ? hard_smp_processor_id() : c->apicid; /* Don't do the funky fallback heuristics the AMD version employs for now. */ @@ -400,7 +400,7 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) } /* Work around errata */ - srat_detect_node(); + srat_detect_node(c); if (cpu_has(c, X86_FEATURE_VMX)) detect_vmx_virtcap(c); diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 483eda96e10..789efe217e1 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -17,6 +17,7 @@ #include <asm/processor.h> #include <asm/smp.h> +#include <asm/k8.h> #define LVL_1_INST 1 #define LVL_1_DATA 2 @@ -159,14 +160,6 @@ struct _cpuid4_info_regs { unsigned long can_disable; }; -#if defined(CONFIG_PCI) && defined(CONFIG_SYSFS) -static struct pci_device_id k8_nb_id[] = { - { PCI_DEVICE(PCI_VENDOR_ID_AMD, 0x1103) }, - { PCI_DEVICE(PCI_VENDOR_ID_AMD, 0x1203) }, - {} -}; -#endif - unsigned short num_cache_leaves; /* AMD doesn't have CPUID4. Emulate it here to report the same @@ -207,10 +200,17 @@ union l3_cache { }; static const unsigned short __cpuinitconst assocs[] = { - [1] = 1, [2] = 2, [4] = 4, [6] = 8, - [8] = 16, [0xa] = 32, [0xb] = 48, + [1] = 1, + [2] = 2, + [4] = 4, + [6] = 8, + [8] = 16, + [0xa] = 32, + [0xb] = 48, [0xc] = 64, - [0xf] = 0xffff // ?? + [0xd] = 96, + [0xe] = 128, + [0xf] = 0xffff /* fully associative - no way to show this currently */ }; static const unsigned char __cpuinitconst levels[] = { 1, 1, 2, 3 }; @@ -271,7 +271,8 @@ amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax, eax->split.type = types[leaf]; eax->split.level = levels[leaf]; if (leaf == 3) - eax->split.num_threads_sharing = current_cpu_data.x86_max_cores - 1; + eax->split.num_threads_sharing = + current_cpu_data.x86_max_cores - 1; else eax->split.num_threads_sharing = 0; eax->split.num_cores_on_die = current_cpu_data.x86_max_cores - 1; @@ -291,6 +292,14 @@ amd_check_l3_disable(int index, struct _cpuid4_info_regs *this_leaf) { if (index < 3) return; + + if (boot_cpu_data.x86 == 0x11) + return; + + /* see erratum #382 */ + if ((boot_cpu_data.x86 == 0x10) && (boot_cpu_data.x86_model < 0x8)) + return; + this_leaf->can_disable = 1; } @@ -696,97 +705,75 @@ static ssize_t show_type(struct _cpuid4_info *this_leaf, char *buf) #define to_object(k) container_of(k, struct _index_kobject, kobj) #define to_attr(a) container_of(a, struct _cache_attr, attr) -#ifdef CONFIG_PCI -static struct pci_dev *get_k8_northbridge(int node) -{ - struct pci_dev *dev = NULL; - int i; - - for (i = 0; i <= node; i++) { - do { - dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev); - if (!dev) - break; - } while (!pci_match_id(&k8_nb_id[0], dev)); - if (!dev) - break; - } - return dev; -} -#else -static struct pci_dev *get_k8_northbridge(int node) -{ - return NULL; -} -#endif - -static ssize_t show_cache_disable(struct _cpuid4_info *this_leaf, char *buf) +static ssize_t show_cache_disable(struct _cpuid4_info *this_leaf, char *buf, + unsigned int index) { - const struct cpumask *mask = to_cpumask(this_leaf->shared_cpu_map); - int node = cpu_to_node(cpumask_first(mask)); - struct pci_dev *dev = NULL; - ssize_t ret = 0; - int i; + int cpu = cpumask_first(to_cpumask(this_leaf->shared_cpu_map)); + int node = cpu_to_node(cpu); + struct pci_dev *dev = node_to_k8_nb_misc(node); + unsigned int reg = 0; if (!this_leaf->can_disable) - return sprintf(buf, "Feature not enabled\n"); - - dev = get_k8_northbridge(node); - if (!dev) { - printk(KERN_ERR "Attempting AMD northbridge operation on a system with no northbridge\n"); return -EINVAL; - } - for (i = 0; i < 2; i++) { - unsigned int reg; + if (!dev) + return -EINVAL; - pci_read_config_dword(dev, 0x1BC + i * 4, ®); + pci_read_config_dword(dev, 0x1BC + index * 4, ®); + return sprintf(buf, "%x\n", reg); +} - ret += sprintf(buf, "%sEntry: %d\n", buf, i); - ret += sprintf(buf, "%sReads: %s\tNew Entries: %s\n", - buf, - reg & 0x80000000 ? "Disabled" : "Allowed", - reg & 0x40000000 ? "Disabled" : "Allowed"); - ret += sprintf(buf, "%sSubCache: %x\tIndex: %x\n", - buf, (reg & 0x30000) >> 16, reg & 0xfff); - } - return ret; +#define SHOW_CACHE_DISABLE(index) \ +static ssize_t \ +show_cache_disable_##index(struct _cpuid4_info *this_leaf, char *buf) \ +{ \ + return show_cache_disable(this_leaf, buf, index); \ } +SHOW_CACHE_DISABLE(0) +SHOW_CACHE_DISABLE(1) -static ssize_t -store_cache_disable(struct _cpuid4_info *this_leaf, const char *buf, - size_t count) +static ssize_t store_cache_disable(struct _cpuid4_info *this_leaf, + const char *buf, size_t count, unsigned int index) { - const struct cpumask *mask = to_cpumask(this_leaf->shared_cpu_map); - int node = cpu_to_node(cpumask_first(mask)); - struct pci_dev *dev = NULL; - unsigned int ret, index, val; + int cpu = cpumask_first(to_cpumask(this_leaf->shared_cpu_map)); + int node = cpu_to_node(cpu); + struct pci_dev *dev = node_to_k8_nb_misc(node); + unsigned long val = 0; + unsigned int scrubber = 0; if (!this_leaf->can_disable) - return 0; - - if (strlen(buf) > 15) return -EINVAL; - ret = sscanf(buf, "%x %x", &index, &val); - if (ret != 2) + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (!dev) return -EINVAL; - if (index > 1) + + if (strict_strtoul(buf, 10, &val) < 0) return -EINVAL; val |= 0xc0000000; - dev = get_k8_northbridge(node); - if (!dev) { - printk(KERN_ERR "Attempting AMD northbridge operation on a system with no northbridge\n"); - return -EINVAL; - } + + pci_read_config_dword(dev, 0x58, &scrubber); + scrubber &= ~0x1f000000; + pci_write_config_dword(dev, 0x58, scrubber); pci_write_config_dword(dev, 0x1BC + index * 4, val & ~0x40000000); wbinvd(); pci_write_config_dword(dev, 0x1BC + index * 4, val); + return count; +} - return 1; +#define STORE_CACHE_DISABLE(index) \ +static ssize_t \ +store_cache_disable_##index(struct _cpuid4_info *this_leaf, \ + const char *buf, size_t count) \ +{ \ + return store_cache_disable(this_leaf, buf, count, index); \ } +STORE_CACHE_DISABLE(0) +STORE_CACHE_DISABLE(1) struct _cache_attr { struct attribute attr; @@ -808,7 +795,10 @@ define_one_ro(size); define_one_ro(shared_cpu_map); define_one_ro(shared_cpu_list); -static struct _cache_attr cache_disable = __ATTR(cache_disable, 0644, show_cache_disable, store_cache_disable); +static struct _cache_attr cache_disable_0 = __ATTR(cache_disable_0, 0644, + show_cache_disable_0, store_cache_disable_0); +static struct _cache_attr cache_disable_1 = __ATTR(cache_disable_1, 0644, + show_cache_disable_1, store_cache_disable_1); static struct attribute * default_attrs[] = { &type.attr, @@ -820,7 +810,8 @@ static struct attribute * default_attrs[] = { &size.attr, &shared_cpu_map.attr, &shared_cpu_list.attr, - &cache_disable.attr, + &cache_disable_0.attr, + &cache_disable_1.attr, NULL }; diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c index cef3ee30744..65a0fceedcd 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c @@ -15,7 +15,6 @@ #include <asm/hw_irq.h> #include <asm/idle.h> #include <asm/therm_throt.h> -#include <asm/apic.h> asmlinkage void smp_thermal_interrupt(void) { diff --git a/arch/x86/kernel/cpu/mtrr/cleanup.c b/arch/x86/kernel/cpu/mtrr/cleanup.c index ce0fe4b5c04..1d584a18a50 100644 --- a/arch/x86/kernel/cpu/mtrr/cleanup.c +++ b/arch/x86/kernel/cpu/mtrr/cleanup.c @@ -808,7 +808,7 @@ int __init mtrr_cleanup(unsigned address_bits) if (!is_cpu(INTEL) || enable_mtrr_cleanup < 1) return 0; - rdmsr(MTRRdefType_MSR, def, dummy); + rdmsr(MSR_MTRRdefType, def, dummy); def &= 0xff; if (def != MTRR_TYPE_UNCACHABLE) return 0; @@ -1003,7 +1003,7 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn) */ if (!is_cpu(INTEL) || disable_mtrr_trim) return 0; - rdmsr(MTRRdefType_MSR, def, dummy); + rdmsr(MSR_MTRRdefType, def, dummy); def &= 0xff; if (def != MTRR_TYPE_UNCACHABLE) return 0; diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c index d21d4fb161f..0543f69f0b2 100644 --- a/arch/x86/kernel/cpu/mtrr/generic.c +++ b/arch/x86/kernel/cpu/mtrr/generic.c @@ -20,9 +20,9 @@ struct fixed_range_block { }; static struct fixed_range_block fixed_range_blocks[] = { - { MTRRfix64K_00000_MSR, 1 }, /* one 64k MTRR */ - { MTRRfix16K_80000_MSR, 2 }, /* two 16k MTRRs */ - { MTRRfix4K_C0000_MSR, 8 }, /* eight 4k MTRRs */ + { MSR_MTRRfix64K_00000, 1 }, /* one 64k MTRR */ + { MSR_MTRRfix16K_80000, 2 }, /* two 16k MTRRs */ + { MSR_MTRRfix4K_C0000, 8 }, /* eight 4k MTRRs */ {} }; @@ -194,12 +194,12 @@ get_fixed_ranges(mtrr_type * frs) k8_check_syscfg_dram_mod_en(); - rdmsr(MTRRfix64K_00000_MSR, p[0], p[1]); + rdmsr(MSR_MTRRfix64K_00000, p[0], p[1]); for (i = 0; i < 2; i++) - rdmsr(MTRRfix16K_80000_MSR + i, p[2 + i * 2], p[3 + i * 2]); + rdmsr(MSR_MTRRfix16K_80000 + i, p[2 + i * 2], p[3 + i * 2]); for (i = 0; i < 8; i++) - rdmsr(MTRRfix4K_C0000_MSR + i, p[6 + i * 2], p[7 + i * 2]); + rdmsr(MSR_MTRRfix4K_C0000 + i, p[6 + i * 2], p[7 + i * 2]); } void mtrr_save_fixed_ranges(void *info) @@ -310,7 +310,7 @@ void __init get_mtrr_state(void) vrs = mtrr_state.var_ranges; - rdmsr(MTRRcap_MSR, lo, dummy); + rdmsr(MSR_MTRRcap, lo, dummy); mtrr_state.have_fixed = (lo >> 8) & 1; for (i = 0; i < num_var_ranges; i++) @@ -318,7 +318,7 @@ void __init get_mtrr_state(void) if (mtrr_state.have_fixed) get_fixed_ranges(mtrr_state.fixed_ranges); - rdmsr(MTRRdefType_MSR, lo, dummy); + rdmsr(MSR_MTRRdefType, lo, dummy); mtrr_state.def_type = (lo & 0xff); mtrr_state.enabled = (lo & 0xc00) >> 10; @@ -583,10 +583,10 @@ static void prepare_set(void) __acquires(set_atomicity_lock) __flush_tlb(); /* Save MTRR state */ - rdmsr(MTRRdefType_MSR, deftype_lo, deftype_hi); + rdmsr(MSR_MTRRdefType, deftype_lo, deftype_hi); /* Disable MTRRs, and set the default type to uncached */ - mtrr_wrmsr(MTRRdefType_MSR, deftype_lo & ~0xcff, deftype_hi); + mtrr_wrmsr(MSR_MTRRdefType, deftype_lo & ~0xcff, deftype_hi); } static void post_set(void) __releases(set_atomicity_lock) @@ -595,7 +595,7 @@ static void post_set(void) __releases(set_atomicity_lock) __flush_tlb(); /* Intel (P6) standard MTRRs */ - mtrr_wrmsr(MTRRdefType_MSR, deftype_lo, deftype_hi); + mtrr_wrmsr(MSR_MTRRdefType, deftype_lo, deftype_hi); /* Enable caches */ write_cr0(read_cr0() & 0xbfffffff); @@ -707,7 +707,7 @@ int generic_validate_add_page(unsigned long base, unsigned long size, unsigned i static int generic_have_wrcomb(void) { unsigned long config, dummy; - rdmsr(MTRRcap_MSR, config, dummy); + rdmsr(MSR_MTRRcap, config, dummy); return (config & (1 << 10)); } diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c index 03cda01f57c..8fc248b5aea 100644 --- a/arch/x86/kernel/cpu/mtrr/main.c +++ b/arch/x86/kernel/cpu/mtrr/main.c @@ -104,7 +104,7 @@ static void __init set_num_var_ranges(void) unsigned long config = 0, dummy; if (use_intel()) { - rdmsr(MTRRcap_MSR, config, dummy); + rdmsr(MSR_MTRRcap, config, dummy); } else if (is_cpu(AMD)) config = 2; else if (is_cpu(CYRIX) || is_cpu(CENTAUR)) diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.h b/arch/x86/kernel/cpu/mtrr/mtrr.h index 77f67f7b347..7538b767f20 100644 --- a/arch/x86/kernel/cpu/mtrr/mtrr.h +++ b/arch/x86/kernel/cpu/mtrr/mtrr.h @@ -5,21 +5,6 @@ #include <linux/types.h> #include <linux/stddef.h> -#define MTRRcap_MSR 0x0fe -#define MTRRdefType_MSR 0x2ff - -#define MTRRfix64K_00000_MSR 0x250 -#define MTRRfix16K_80000_MSR 0x258 -#define MTRRfix16K_A0000_MSR 0x259 -#define MTRRfix4K_C0000_MSR 0x268 -#define MTRRfix4K_C8000_MSR 0x269 -#define MTRRfix4K_D0000_MSR 0x26a -#define MTRRfix4K_D8000_MSR 0x26b -#define MTRRfix4K_E0000_MSR 0x26c -#define MTRRfix4K_E8000_MSR 0x26d -#define MTRRfix4K_F0000_MSR 0x26e -#define MTRRfix4K_F8000_MSR 0x26f - #define MTRR_CHANGE_MASK_FIXED 0x01 #define MTRR_CHANGE_MASK_VARIABLE 0x02 #define MTRR_CHANGE_MASK_DEFTYPE 0x04 diff --git a/arch/x86/kernel/cpu/mtrr/state.c b/arch/x86/kernel/cpu/mtrr/state.c index 7f7e2753685..1f5fb1588d1 100644 --- a/arch/x86/kernel/cpu/mtrr/state.c +++ b/arch/x86/kernel/cpu/mtrr/state.c @@ -35,7 +35,7 @@ void set_mtrr_prepare_save(struct set_mtrr_context *ctxt) if (use_intel()) /* Save MTRR state */ - rdmsr(MTRRdefType_MSR, ctxt->deftype_lo, ctxt->deftype_hi); + rdmsr(MSR_MTRRdefType, ctxt->deftype_lo, ctxt->deftype_hi); else /* Cyrix ARRs - everything else were excluded at the top */ ctxt->ccr3 = getCx86(CX86_CCR3); @@ -46,7 +46,7 @@ void set_mtrr_cache_disable(struct set_mtrr_context *ctxt) { if (use_intel()) /* Disable MTRRs, and set the default type to uncached */ - mtrr_wrmsr(MTRRdefType_MSR, ctxt->deftype_lo & 0xf300UL, + mtrr_wrmsr(MSR_MTRRdefType, ctxt->deftype_lo & 0xf300UL, ctxt->deftype_hi); else if (is_cpu(CYRIX)) /* Cyrix ARRs - everything else were excluded at the top */ @@ -64,7 +64,7 @@ void set_mtrr_done(struct set_mtrr_context *ctxt) /* Restore MTRRdefType */ if (use_intel()) /* Intel (P6) standard MTRRs */ - mtrr_wrmsr(MTRRdefType_MSR, ctxt->deftype_lo, ctxt->deftype_hi); + mtrr_wrmsr(MSR_MTRRdefType, ctxt->deftype_lo, ctxt->deftype_hi); else /* Cyrix ARRs - everything else was excluded at the top */ setCx86(CX86_CCR3, ctxt->ccr3); diff --git a/arch/x86/kernel/cpu/perf_counter.c b/arch/x86/kernel/cpu/perf_counter.c new file mode 100644 index 00000000000..895c82e7845 --- /dev/null +++ b/arch/x86/kernel/cpu/perf_counter.c @@ -0,0 +1,1704 @@ +/* + * Performance counter x86 architecture code + * + * Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de> + * Copyright (C) 2008-2009 Red Hat, Inc., Ingo Molnar + * Copyright (C) 2009 Jaswinder Singh Rajput + * Copyright (C) 2009 Advanced Micro Devices, Inc., Robert Richter + * Copyright (C) 2008-2009 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com> + * + * For licencing details see kernel-base/COPYING + */ + +#include <linux/perf_counter.h> +#include <linux/capability.h> +#include <linux/notifier.h> +#include <linux/hardirq.h> +#include <linux/kprobes.h> +#include <linux/module.h> +#include <linux/kdebug.h> +#include <linux/sched.h> +#include <linux/uaccess.h> + +#include <asm/apic.h> +#include <asm/stacktrace.h> +#include <asm/nmi.h> + +static u64 perf_counter_mask __read_mostly; + +struct cpu_hw_counters { + struct perf_counter *counters[X86_PMC_IDX_MAX]; + unsigned long used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; + unsigned long active_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; + unsigned long interrupts; + int enabled; +}; + +/* + * struct x86_pmu - generic x86 pmu + */ +struct x86_pmu { + const char *name; + int version; + int (*handle_irq)(struct pt_regs *); + void (*disable_all)(void); + void (*enable_all)(void); + void (*enable)(struct hw_perf_counter *, int); + void (*disable)(struct hw_perf_counter *, int); + unsigned eventsel; + unsigned perfctr; + u64 (*event_map)(int); + u64 (*raw_event)(u64); + int max_events; + int num_counters; + int num_counters_fixed; + int counter_bits; + u64 counter_mask; + u64 max_period; + u64 intel_ctrl; +}; + +static struct x86_pmu x86_pmu __read_mostly; + +static DEFINE_PER_CPU(struct cpu_hw_counters, cpu_hw_counters) = { + .enabled = 1, +}; + +/* + * Intel PerfMon v3. Used on Core2 and later. + */ +static const u64 intel_perfmon_event_map[] = +{ + [PERF_COUNT_HW_CPU_CYCLES] = 0x003c, + [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x4f2e, + [PERF_COUNT_HW_CACHE_MISSES] = 0x412e, + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c4, + [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c5, + [PERF_COUNT_HW_BUS_CYCLES] = 0x013c, +}; + +static u64 intel_pmu_event_map(int event) +{ + return intel_perfmon_event_map[event]; +} + +/* + * Generalized hw caching related event table, filled + * in on a per model basis. A value of 0 means + * 'not supported', -1 means 'event makes no sense on + * this CPU', any other value means the raw event + * ID. + */ + +#define C(x) PERF_COUNT_HW_CACHE_##x + +static u64 __read_mostly hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX]; + +static const u64 nehalem_hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(L1D) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0f40, /* L1D_CACHE_LD.MESI */ + [ C(RESULT_MISS) ] = 0x0140, /* L1D_CACHE_LD.I_STATE */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x0f41, /* L1D_CACHE_ST.MESI */ + [ C(RESULT_MISS) ] = 0x0141, /* L1D_CACHE_ST.I_STATE */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x014e, /* L1D_PREFETCH.REQUESTS */ + [ C(RESULT_MISS) ] = 0x024e, /* L1D_PREFETCH.MISS */ + }, + }, + [ C(L1I ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0380, /* L1I.READS */ + [ C(RESULT_MISS) ] = 0x0280, /* L1I.MISSES */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + [ C(RESULT_MISS) ] = 0x0, + }, + }, + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0324, /* L2_RQSTS.LOADS */ + [ C(RESULT_MISS) ] = 0x0224, /* L2_RQSTS.LD_MISS */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x0c24, /* L2_RQSTS.RFOS */ + [ C(RESULT_MISS) ] = 0x0824, /* L2_RQSTS.RFO_MISS */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x4f2e, /* LLC Reference */ + [ C(RESULT_MISS) ] = 0x412e, /* LLC Misses */ + }, + }, + [ C(DTLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0f40, /* L1D_CACHE_LD.MESI (alias) */ + [ C(RESULT_MISS) ] = 0x0108, /* DTLB_LOAD_MISSES.ANY */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x0f41, /* L1D_CACHE_ST.MESI (alias) */ + [ C(RESULT_MISS) ] = 0x010c, /* MEM_STORE_RETIRED.DTLB_MISS */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + [ C(RESULT_MISS) ] = 0x0, + }, + }, + [ C(ITLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x01c0, /* INST_RETIRED.ANY_P */ + [ C(RESULT_MISS) ] = 0x20c8, /* ITLB_MISS_RETIRED */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, + [ C(BPU ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c4, /* BR_INST_RETIRED.ALL_BRANCHES */ + [ C(RESULT_MISS) ] = 0x03e8, /* BPU_CLEARS.ANY */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, +}; + +static const u64 core2_hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(L1D) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0f40, /* L1D_CACHE_LD.MESI */ + [ C(RESULT_MISS) ] = 0x0140, /* L1D_CACHE_LD.I_STATE */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x0f41, /* L1D_CACHE_ST.MESI */ + [ C(RESULT_MISS) ] = 0x0141, /* L1D_CACHE_ST.I_STATE */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x104e, /* L1D_PREFETCH.REQUESTS */ + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(L1I ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0080, /* L1I.READS */ + [ C(RESULT_MISS) ] = 0x0081, /* L1I.MISSES */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x4f29, /* L2_LD.MESI */ + [ C(RESULT_MISS) ] = 0x4129, /* L2_LD.ISTATE */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x4f2A, /* L2_ST.MESI */ + [ C(RESULT_MISS) ] = 0x412A, /* L2_ST.ISTATE */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(DTLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0f40, /* L1D_CACHE_LD.MESI (alias) */ + [ C(RESULT_MISS) ] = 0x0208, /* DTLB_MISSES.MISS_LD */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x0f41, /* L1D_CACHE_ST.MESI (alias) */ + [ C(RESULT_MISS) ] = 0x0808, /* DTLB_MISSES.MISS_ST */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(ITLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c0, /* INST_RETIRED.ANY_P */ + [ C(RESULT_MISS) ] = 0x1282, /* ITLBMISSES */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, + [ C(BPU ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c4, /* BR_INST_RETIRED.ANY */ + [ C(RESULT_MISS) ] = 0x00c5, /* BP_INST_RETIRED.MISPRED */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, +}; + +static const u64 atom_hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(L1D) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x2140, /* L1D_CACHE.LD */ + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x2240, /* L1D_CACHE.ST */ + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(L1I ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0380, /* L1I.READS */ + [ C(RESULT_MISS) ] = 0x0280, /* L1I.MISSES */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x4f29, /* L2_LD.MESI */ + [ C(RESULT_MISS) ] = 0x4129, /* L2_LD.ISTATE */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x4f2A, /* L2_ST.MESI */ + [ C(RESULT_MISS) ] = 0x412A, /* L2_ST.ISTATE */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(DTLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x2140, /* L1D_CACHE_LD.MESI (alias) */ + [ C(RESULT_MISS) ] = 0x0508, /* DTLB_MISSES.MISS_LD */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x2240, /* L1D_CACHE_ST.MESI (alias) */ + [ C(RESULT_MISS) ] = 0x0608, /* DTLB_MISSES.MISS_ST */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(ITLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c0, /* INST_RETIRED.ANY_P */ + [ C(RESULT_MISS) ] = 0x0282, /* ITLB.MISSES */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, + [ C(BPU ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c4, /* BR_INST_RETIRED.ANY */ + [ C(RESULT_MISS) ] = 0x00c5, /* BP_INST_RETIRED.MISPRED */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, +}; + +static u64 intel_pmu_raw_event(u64 event) +{ +#define CORE_EVNTSEL_EVENT_MASK 0x000000FFULL +#define CORE_EVNTSEL_UNIT_MASK 0x0000FF00ULL +#define CORE_EVNTSEL_EDGE_MASK 0x00040000ULL +#define CORE_EVNTSEL_INV_MASK 0x00800000ULL +#define CORE_EVNTSEL_COUNTER_MASK 0xFF000000ULL + +#define CORE_EVNTSEL_MASK \ + (CORE_EVNTSEL_EVENT_MASK | \ + CORE_EVNTSEL_UNIT_MASK | \ + CORE_EVNTSEL_EDGE_MASK | \ + CORE_EVNTSEL_INV_MASK | \ + CORE_EVNTSEL_COUNTER_MASK) + + return event & CORE_EVNTSEL_MASK; +} + +static const u64 amd_0f_hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(L1D) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(L1I ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0080, /* Instruction cache fetches */ + [ C(RESULT_MISS) ] = 0x0081, /* Instruction cache misses */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(DTLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(ITLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0080, /* Instruction fecthes */ + [ C(RESULT_MISS) ] = 0x0085, /* Instr. fetch ITLB misses */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, + [ C(BPU ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c2, /* Retired Branch Instr. */ + [ C(RESULT_MISS) ] = 0x00c3, /* Retired Mispredicted BI */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, +}; + +/* + * AMD Performance Monitor K7 and later. + */ +static const u64 amd_perfmon_event_map[] = +{ + [PERF_COUNT_HW_CPU_CYCLES] = 0x0076, + [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x0080, + [PERF_COUNT_HW_CACHE_MISSES] = 0x0081, + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c4, + [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c5, +}; + +static u64 amd_pmu_event_map(int event) +{ + return amd_perfmon_event_map[event]; +} + +static u64 amd_pmu_raw_event(u64 event) +{ +#define K7_EVNTSEL_EVENT_MASK 0x7000000FFULL +#define K7_EVNTSEL_UNIT_MASK 0x00000FF00ULL +#define K7_EVNTSEL_EDGE_MASK 0x000040000ULL +#define K7_EVNTSEL_INV_MASK 0x000800000ULL +#define K7_EVNTSEL_COUNTER_MASK 0x0FF000000ULL + +#define K7_EVNTSEL_MASK \ + (K7_EVNTSEL_EVENT_MASK | \ + K7_EVNTSEL_UNIT_MASK | \ + K7_EVNTSEL_EDGE_MASK | \ + K7_EVNTSEL_INV_MASK | \ + K7_EVNTSEL_COUNTER_MASK) + + return event & K7_EVNTSEL_MASK; +} + +/* + * Propagate counter elapsed time into the generic counter. + * Can only be executed on the CPU where the counter is active. + * Returns the delta events processed. + */ +static u64 +x86_perf_counter_update(struct perf_counter *counter, + struct hw_perf_counter *hwc, int idx) +{ + int shift = 64 - x86_pmu.counter_bits; + u64 prev_raw_count, new_raw_count; + s64 delta; + + /* + * Careful: an NMI might modify the previous counter value. + * + * Our tactic to handle this is to first atomically read and + * exchange a new raw count - then add that new-prev delta + * count to the generic counter atomically: + */ +again: + prev_raw_count = atomic64_read(&hwc->prev_count); + rdmsrl(hwc->counter_base + idx, new_raw_count); + + if (atomic64_cmpxchg(&hwc->prev_count, prev_raw_count, + new_raw_count) != prev_raw_count) + goto again; + + /* + * Now we have the new raw value and have updated the prev + * timestamp already. We can now calculate the elapsed delta + * (counter-)time and add that to the generic counter. + * + * Careful, not all hw sign-extends above the physical width + * of the count. + */ + delta = (new_raw_count << shift) - (prev_raw_count << shift); + delta >>= shift; + + atomic64_add(delta, &counter->count); + atomic64_sub(delta, &hwc->period_left); + + return new_raw_count; +} + +static atomic_t active_counters; +static DEFINE_MUTEX(pmc_reserve_mutex); + +static bool reserve_pmc_hardware(void) +{ + int i; + + if (nmi_watchdog == NMI_LOCAL_APIC) + disable_lapic_nmi_watchdog(); + + for (i = 0; i < x86_pmu.num_counters; i++) { + if (!reserve_perfctr_nmi(x86_pmu.perfctr + i)) + goto perfctr_fail; + } + + for (i = 0; i < x86_pmu.num_counters; i++) { + if (!reserve_evntsel_nmi(x86_pmu.eventsel + i)) + goto eventsel_fail; + } + + return true; + +eventsel_fail: + for (i--; i >= 0; i--) + release_evntsel_nmi(x86_pmu.eventsel + i); + + i = x86_pmu.num_counters; + +perfctr_fail: + for (i--; i >= 0; i--) + release_perfctr_nmi(x86_pmu.perfctr + i); + + if (nmi_watchdog == NMI_LOCAL_APIC) + enable_lapic_nmi_watchdog(); + + return false; +} + +static void release_pmc_hardware(void) +{ + int i; + + for (i = 0; i < x86_pmu.num_counters; i++) { + release_perfctr_nmi(x86_pmu.perfctr + i); + release_evntsel_nmi(x86_pmu.eventsel + i); + } + + if (nmi_watchdog == NMI_LOCAL_APIC) + enable_lapic_nmi_watchdog(); +} + +static void hw_perf_counter_destroy(struct perf_counter *counter) +{ + if (atomic_dec_and_mutex_lock(&active_counters, &pmc_reserve_mutex)) { + release_pmc_hardware(); + mutex_unlock(&pmc_reserve_mutex); + } +} + +static inline int x86_pmu_initialized(void) +{ + return x86_pmu.handle_irq != NULL; +} + +static inline int +set_ext_hw_attr(struct hw_perf_counter *hwc, struct perf_counter_attr *attr) +{ + unsigned int cache_type, cache_op, cache_result; + u64 config, val; + + config = attr->config; + + cache_type = (config >> 0) & 0xff; + if (cache_type >= PERF_COUNT_HW_CACHE_MAX) + return -EINVAL; + + cache_op = (config >> 8) & 0xff; + if (cache_op >= PERF_COUNT_HW_CACHE_OP_MAX) + return -EINVAL; + + cache_result = (config >> 16) & 0xff; + if (cache_result >= PERF_COUNT_HW_CACHE_RESULT_MAX) + return -EINVAL; + + val = hw_cache_event_ids[cache_type][cache_op][cache_result]; + + if (val == 0) + return -ENOENT; + + if (val == -1) + return -EINVAL; + + hwc->config |= val; + + return 0; +} + +/* + * Setup the hardware configuration for a given attr_type + */ +static int __hw_perf_counter_init(struct perf_counter *counter) +{ + struct perf_counter_attr *attr = &counter->attr; + struct hw_perf_counter *hwc = &counter->hw; + int err; + + if (!x86_pmu_initialized()) + return -ENODEV; + + err = 0; + if (!atomic_inc_not_zero(&active_counters)) { + mutex_lock(&pmc_reserve_mutex); + if (atomic_read(&active_counters) == 0 && !reserve_pmc_hardware()) + err = -EBUSY; + else + atomic_inc(&active_counters); + mutex_unlock(&pmc_reserve_mutex); + } + if (err) + return err; + + /* + * Generate PMC IRQs: + * (keep 'enabled' bit clear for now) + */ + hwc->config = ARCH_PERFMON_EVENTSEL_INT; + + /* + * Count user and OS events unless requested not to. + */ + if (!attr->exclude_user) + hwc->config |= ARCH_PERFMON_EVENTSEL_USR; + if (!attr->exclude_kernel) + hwc->config |= ARCH_PERFMON_EVENTSEL_OS; + + if (!hwc->sample_period) { + hwc->sample_period = x86_pmu.max_period; + hwc->last_period = hwc->sample_period; + atomic64_set(&hwc->period_left, hwc->sample_period); + } + + counter->destroy = hw_perf_counter_destroy; + + /* + * Raw event type provide the config in the event structure + */ + if (attr->type == PERF_TYPE_RAW) { + hwc->config |= x86_pmu.raw_event(attr->config); + return 0; + } + + if (attr->type == PERF_TYPE_HW_CACHE) + return set_ext_hw_attr(hwc, attr); + + if (attr->config >= x86_pmu.max_events) + return -EINVAL; + /* + * The generic map: + */ + hwc->config |= x86_pmu.event_map(attr->config); + + return 0; +} + +static void intel_pmu_disable_all(void) +{ + wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0); +} + +static void amd_pmu_disable_all(void) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + int idx; + + if (!cpuc->enabled) + return; + + cpuc->enabled = 0; + /* + * ensure we write the disable before we start disabling the + * counters proper, so that amd_pmu_enable_counter() does the + * right thing. + */ + barrier(); + + for (idx = 0; idx < x86_pmu.num_counters; idx++) { + u64 val; + + if (!test_bit(idx, cpuc->active_mask)) + continue; + rdmsrl(MSR_K7_EVNTSEL0 + idx, val); + if (!(val & ARCH_PERFMON_EVENTSEL0_ENABLE)) + continue; + val &= ~ARCH_PERFMON_EVENTSEL0_ENABLE; + wrmsrl(MSR_K7_EVNTSEL0 + idx, val); + } +} + +void hw_perf_disable(void) +{ + if (!x86_pmu_initialized()) + return; + return x86_pmu.disable_all(); +} + +static void intel_pmu_enable_all(void) +{ + wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, x86_pmu.intel_ctrl); +} + +static void amd_pmu_enable_all(void) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + int idx; + + if (cpuc->enabled) + return; + + cpuc->enabled = 1; + barrier(); + + for (idx = 0; idx < x86_pmu.num_counters; idx++) { + u64 val; + + if (!test_bit(idx, cpuc->active_mask)) + continue; + rdmsrl(MSR_K7_EVNTSEL0 + idx, val); + if (val & ARCH_PERFMON_EVENTSEL0_ENABLE) + continue; + val |= ARCH_PERFMON_EVENTSEL0_ENABLE; + wrmsrl(MSR_K7_EVNTSEL0 + idx, val); + } +} + +void hw_perf_enable(void) +{ + if (!x86_pmu_initialized()) + return; + x86_pmu.enable_all(); +} + +static inline u64 intel_pmu_get_status(void) +{ + u64 status; + + rdmsrl(MSR_CORE_PERF_GLOBAL_STATUS, status); + + return status; +} + +static inline void intel_pmu_ack_status(u64 ack) +{ + wrmsrl(MSR_CORE_PERF_GLOBAL_OVF_CTRL, ack); +} + +static inline void x86_pmu_enable_counter(struct hw_perf_counter *hwc, int idx) +{ + int err; + err = checking_wrmsrl(hwc->config_base + idx, + hwc->config | ARCH_PERFMON_EVENTSEL0_ENABLE); +} + +static inline void x86_pmu_disable_counter(struct hw_perf_counter *hwc, int idx) +{ + int err; + err = checking_wrmsrl(hwc->config_base + idx, + hwc->config); +} + +static inline void +intel_pmu_disable_fixed(struct hw_perf_counter *hwc, int __idx) +{ + int idx = __idx - X86_PMC_IDX_FIXED; + u64 ctrl_val, mask; + int err; + + mask = 0xfULL << (idx * 4); + + rdmsrl(hwc->config_base, ctrl_val); + ctrl_val &= ~mask; + err = checking_wrmsrl(hwc->config_base, ctrl_val); +} + +static inline void +intel_pmu_disable_counter(struct hw_perf_counter *hwc, int idx) +{ + if (unlikely(hwc->config_base == MSR_ARCH_PERFMON_FIXED_CTR_CTRL)) { + intel_pmu_disable_fixed(hwc, idx); + return; + } + + x86_pmu_disable_counter(hwc, idx); +} + +static inline void +amd_pmu_disable_counter(struct hw_perf_counter *hwc, int idx) +{ + x86_pmu_disable_counter(hwc, idx); +} + +static DEFINE_PER_CPU(u64, prev_left[X86_PMC_IDX_MAX]); + +/* + * Set the next IRQ period, based on the hwc->period_left value. + * To be called with the counter disabled in hw: + */ +static int +x86_perf_counter_set_period(struct perf_counter *counter, + struct hw_perf_counter *hwc, int idx) +{ + s64 left = atomic64_read(&hwc->period_left); + s64 period = hwc->sample_period; + int err, ret = 0; + + /* + * If we are way outside a reasoable range then just skip forward: + */ + if (unlikely(left <= -period)) { + left = period; + atomic64_set(&hwc->period_left, left); + hwc->last_period = period; + ret = 1; + } + + if (unlikely(left <= 0)) { + left += period; + atomic64_set(&hwc->period_left, left); + hwc->last_period = period; + ret = 1; + } + /* + * Quirk: certain CPUs dont like it if just 1 event is left: + */ + if (unlikely(left < 2)) + left = 2; + + if (left > x86_pmu.max_period) + left = x86_pmu.max_period; + + per_cpu(prev_left[idx], smp_processor_id()) = left; + + /* + * The hw counter starts counting from this counter offset, + * mark it to be able to extra future deltas: + */ + atomic64_set(&hwc->prev_count, (u64)-left); + + err = checking_wrmsrl(hwc->counter_base + idx, + (u64)(-left) & x86_pmu.counter_mask); + + return ret; +} + +static inline void +intel_pmu_enable_fixed(struct hw_perf_counter *hwc, int __idx) +{ + int idx = __idx - X86_PMC_IDX_FIXED; + u64 ctrl_val, bits, mask; + int err; + + /* + * Enable IRQ generation (0x8), + * and enable ring-3 counting (0x2) and ring-0 counting (0x1) + * if requested: + */ + bits = 0x8ULL; + if (hwc->config & ARCH_PERFMON_EVENTSEL_USR) + bits |= 0x2; + if (hwc->config & ARCH_PERFMON_EVENTSEL_OS) + bits |= 0x1; + bits <<= (idx * 4); + mask = 0xfULL << (idx * 4); + + rdmsrl(hwc->config_base, ctrl_val); + ctrl_val &= ~mask; + ctrl_val |= bits; + err = checking_wrmsrl(hwc->config_base, ctrl_val); +} + +static void intel_pmu_enable_counter(struct hw_perf_counter *hwc, int idx) +{ + if (unlikely(hwc->config_base == MSR_ARCH_PERFMON_FIXED_CTR_CTRL)) { + intel_pmu_enable_fixed(hwc, idx); + return; + } + + x86_pmu_enable_counter(hwc, idx); +} + +static void amd_pmu_enable_counter(struct hw_perf_counter *hwc, int idx) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + + if (cpuc->enabled) + x86_pmu_enable_counter(hwc, idx); + else + x86_pmu_disable_counter(hwc, idx); +} + +static int +fixed_mode_idx(struct perf_counter *counter, struct hw_perf_counter *hwc) +{ + unsigned int event; + + if (!x86_pmu.num_counters_fixed) + return -1; + + event = hwc->config & ARCH_PERFMON_EVENT_MASK; + + if (unlikely(event == x86_pmu.event_map(PERF_COUNT_HW_INSTRUCTIONS))) + return X86_PMC_IDX_FIXED_INSTRUCTIONS; + if (unlikely(event == x86_pmu.event_map(PERF_COUNT_HW_CPU_CYCLES))) + return X86_PMC_IDX_FIXED_CPU_CYCLES; + if (unlikely(event == x86_pmu.event_map(PERF_COUNT_HW_BUS_CYCLES))) + return X86_PMC_IDX_FIXED_BUS_CYCLES; + + return -1; +} + +/* + * Find a PMC slot for the freshly enabled / scheduled in counter: + */ +static int x86_pmu_enable(struct perf_counter *counter) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + struct hw_perf_counter *hwc = &counter->hw; + int idx; + + idx = fixed_mode_idx(counter, hwc); + if (idx >= 0) { + /* + * Try to get the fixed counter, if that is already taken + * then try to get a generic counter: + */ + if (test_and_set_bit(idx, cpuc->used_mask)) + goto try_generic; + + hwc->config_base = MSR_ARCH_PERFMON_FIXED_CTR_CTRL; + /* + * We set it so that counter_base + idx in wrmsr/rdmsr maps to + * MSR_ARCH_PERFMON_FIXED_CTR0 ... CTR2: + */ + hwc->counter_base = + MSR_ARCH_PERFMON_FIXED_CTR0 - X86_PMC_IDX_FIXED; + hwc->idx = idx; + } else { + idx = hwc->idx; + /* Try to get the previous generic counter again */ + if (test_and_set_bit(idx, cpuc->used_mask)) { +try_generic: + idx = find_first_zero_bit(cpuc->used_mask, + x86_pmu.num_counters); + if (idx == x86_pmu.num_counters) + return -EAGAIN; + + set_bit(idx, cpuc->used_mask); + hwc->idx = idx; + } + hwc->config_base = x86_pmu.eventsel; + hwc->counter_base = x86_pmu.perfctr; + } + + perf_counters_lapic_init(); + + x86_pmu.disable(hwc, idx); + + cpuc->counters[idx] = counter; + set_bit(idx, cpuc->active_mask); + + x86_perf_counter_set_period(counter, hwc, idx); + x86_pmu.enable(hwc, idx); + + return 0; +} + +static void x86_pmu_unthrottle(struct perf_counter *counter) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + struct hw_perf_counter *hwc = &counter->hw; + + if (WARN_ON_ONCE(hwc->idx >= X86_PMC_IDX_MAX || + cpuc->counters[hwc->idx] != counter)) + return; + + x86_pmu.enable(hwc, hwc->idx); +} + +void perf_counter_print_debug(void) +{ + u64 ctrl, status, overflow, pmc_ctrl, pmc_count, prev_left, fixed; + struct cpu_hw_counters *cpuc; + unsigned long flags; + int cpu, idx; + + if (!x86_pmu.num_counters) + return; + + local_irq_save(flags); + + cpu = smp_processor_id(); + cpuc = &per_cpu(cpu_hw_counters, cpu); + + if (x86_pmu.version >= 2) { + rdmsrl(MSR_CORE_PERF_GLOBAL_CTRL, ctrl); + rdmsrl(MSR_CORE_PERF_GLOBAL_STATUS, status); + rdmsrl(MSR_CORE_PERF_GLOBAL_OVF_CTRL, overflow); + rdmsrl(MSR_ARCH_PERFMON_FIXED_CTR_CTRL, fixed); + + pr_info("\n"); + pr_info("CPU#%d: ctrl: %016llx\n", cpu, ctrl); + pr_info("CPU#%d: status: %016llx\n", cpu, status); + pr_info("CPU#%d: overflow: %016llx\n", cpu, overflow); + pr_info("CPU#%d: fixed: %016llx\n", cpu, fixed); + } + pr_info("CPU#%d: used: %016llx\n", cpu, *(u64 *)cpuc->used_mask); + + for (idx = 0; idx < x86_pmu.num_counters; idx++) { + rdmsrl(x86_pmu.eventsel + idx, pmc_ctrl); + rdmsrl(x86_pmu.perfctr + idx, pmc_count); + + prev_left = per_cpu(prev_left[idx], cpu); + + pr_info("CPU#%d: gen-PMC%d ctrl: %016llx\n", + cpu, idx, pmc_ctrl); + pr_info("CPU#%d: gen-PMC%d count: %016llx\n", + cpu, idx, pmc_count); + pr_info("CPU#%d: gen-PMC%d left: %016llx\n", + cpu, idx, prev_left); + } + for (idx = 0; idx < x86_pmu.num_counters_fixed; idx++) { + rdmsrl(MSR_ARCH_PERFMON_FIXED_CTR0 + idx, pmc_count); + + pr_info("CPU#%d: fixed-PMC%d count: %016llx\n", + cpu, idx, pmc_count); + } + local_irq_restore(flags); +} + +static void x86_pmu_disable(struct perf_counter *counter) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + struct hw_perf_counter *hwc = &counter->hw; + int idx = hwc->idx; + + /* + * Must be done before we disable, otherwise the nmi handler + * could reenable again: + */ + clear_bit(idx, cpuc->active_mask); + x86_pmu.disable(hwc, idx); + + /* + * Make sure the cleared pointer becomes visible before we + * (potentially) free the counter: + */ + barrier(); + + /* + * Drain the remaining delta count out of a counter + * that we are disabling: + */ + x86_perf_counter_update(counter, hwc, idx); + cpuc->counters[idx] = NULL; + clear_bit(idx, cpuc->used_mask); +} + +/* + * Save and restart an expired counter. Called by NMI contexts, + * so it has to be careful about preempting normal counter ops: + */ +static int intel_pmu_save_and_restart(struct perf_counter *counter) +{ + struct hw_perf_counter *hwc = &counter->hw; + int idx = hwc->idx; + int ret; + + x86_perf_counter_update(counter, hwc, idx); + ret = x86_perf_counter_set_period(counter, hwc, idx); + + if (counter->state == PERF_COUNTER_STATE_ACTIVE) + intel_pmu_enable_counter(hwc, idx); + + return ret; +} + +static void intel_pmu_reset(void) +{ + unsigned long flags; + int idx; + + if (!x86_pmu.num_counters) + return; + + local_irq_save(flags); + + printk("clearing PMU state on CPU#%d\n", smp_processor_id()); + + for (idx = 0; idx < x86_pmu.num_counters; idx++) { + checking_wrmsrl(x86_pmu.eventsel + idx, 0ull); + checking_wrmsrl(x86_pmu.perfctr + idx, 0ull); + } + for (idx = 0; idx < x86_pmu.num_counters_fixed; idx++) { + checking_wrmsrl(MSR_ARCH_PERFMON_FIXED_CTR0 + idx, 0ull); + } + + local_irq_restore(flags); +} + + +/* + * This handler is triggered by the local APIC, so the APIC IRQ handling + * rules apply: + */ +static int intel_pmu_handle_irq(struct pt_regs *regs) +{ + struct perf_sample_data data; + struct cpu_hw_counters *cpuc; + int bit, cpu, loops; + u64 ack, status; + + data.regs = regs; + data.addr = 0; + + cpu = smp_processor_id(); + cpuc = &per_cpu(cpu_hw_counters, cpu); + + perf_disable(); + status = intel_pmu_get_status(); + if (!status) { + perf_enable(); + return 0; + } + + loops = 0; +again: + if (++loops > 100) { + WARN_ONCE(1, "perfcounters: irq loop stuck!\n"); + perf_counter_print_debug(); + intel_pmu_reset(); + perf_enable(); + return 1; + } + + inc_irq_stat(apic_perf_irqs); + ack = status; + for_each_bit(bit, (unsigned long *)&status, X86_PMC_IDX_MAX) { + struct perf_counter *counter = cpuc->counters[bit]; + + clear_bit(bit, (unsigned long *) &status); + if (!test_bit(bit, cpuc->active_mask)) + continue; + + if (!intel_pmu_save_and_restart(counter)) + continue; + + if (perf_counter_overflow(counter, 1, &data)) + intel_pmu_disable_counter(&counter->hw, bit); + } + + intel_pmu_ack_status(ack); + + /* + * Repeat if there is more work to be done: + */ + status = intel_pmu_get_status(); + if (status) + goto again; + + perf_enable(); + + return 1; +} + +static int amd_pmu_handle_irq(struct pt_regs *regs) +{ + struct perf_sample_data data; + struct cpu_hw_counters *cpuc; + struct perf_counter *counter; + struct hw_perf_counter *hwc; + int cpu, idx, handled = 0; + u64 val; + + data.regs = regs; + data.addr = 0; + + cpu = smp_processor_id(); + cpuc = &per_cpu(cpu_hw_counters, cpu); + + for (idx = 0; idx < x86_pmu.num_counters; idx++) { + if (!test_bit(idx, cpuc->active_mask)) + continue; + + counter = cpuc->counters[idx]; + hwc = &counter->hw; + + val = x86_perf_counter_update(counter, hwc, idx); + if (val & (1ULL << (x86_pmu.counter_bits - 1))) + continue; + + /* + * counter overflow + */ + handled = 1; + data.period = counter->hw.last_period; + + if (!x86_perf_counter_set_period(counter, hwc, idx)) + continue; + + if (perf_counter_overflow(counter, 1, &data)) + amd_pmu_disable_counter(hwc, idx); + } + + if (handled) + inc_irq_stat(apic_perf_irqs); + + return handled; +} + +void smp_perf_pending_interrupt(struct pt_regs *regs) +{ + irq_enter(); + ack_APIC_irq(); + inc_irq_stat(apic_pending_irqs); + perf_counter_do_pending(); + irq_exit(); +} + +void set_perf_counter_pending(void) +{ + apic->send_IPI_self(LOCAL_PENDING_VECTOR); +} + +void perf_counters_lapic_init(void) +{ + if (!x86_pmu_initialized()) + return; + + /* + * Always use NMI for PMU + */ + apic_write(APIC_LVTPC, APIC_DM_NMI); +} + +static int __kprobes +perf_counter_nmi_handler(struct notifier_block *self, + unsigned long cmd, void *__args) +{ + struct die_args *args = __args; + struct pt_regs *regs; + + if (!atomic_read(&active_counters)) + return NOTIFY_DONE; + + switch (cmd) { + case DIE_NMI: + case DIE_NMI_IPI: + break; + + default: + return NOTIFY_DONE; + } + + regs = args->regs; + + apic_write(APIC_LVTPC, APIC_DM_NMI); + /* + * Can't rely on the handled return value to say it was our NMI, two + * counters could trigger 'simultaneously' raising two back-to-back NMIs. + * + * If the first NMI handles both, the latter will be empty and daze + * the CPU. + */ + x86_pmu.handle_irq(regs); + + return NOTIFY_STOP; +} + +static __read_mostly struct notifier_block perf_counter_nmi_notifier = { + .notifier_call = perf_counter_nmi_handler, + .next = NULL, + .priority = 1 +}; + +static struct x86_pmu intel_pmu = { + .name = "Intel", + .handle_irq = intel_pmu_handle_irq, + .disable_all = intel_pmu_disable_all, + .enable_all = intel_pmu_enable_all, + .enable = intel_pmu_enable_counter, + .disable = intel_pmu_disable_counter, + .eventsel = MSR_ARCH_PERFMON_EVENTSEL0, + .perfctr = MSR_ARCH_PERFMON_PERFCTR0, + .event_map = intel_pmu_event_map, + .raw_event = intel_pmu_raw_event, + .max_events = ARRAY_SIZE(intel_perfmon_event_map), + /* + * Intel PMCs cannot be accessed sanely above 32 bit width, + * so we install an artificial 1<<31 period regardless of + * the generic counter period: + */ + .max_period = (1ULL << 31) - 1, +}; + +static struct x86_pmu amd_pmu = { + .name = "AMD", + .handle_irq = amd_pmu_handle_irq, + .disable_all = amd_pmu_disable_all, + .enable_all = amd_pmu_enable_all, + .enable = amd_pmu_enable_counter, + .disable = amd_pmu_disable_counter, + .eventsel = MSR_K7_EVNTSEL0, + .perfctr = MSR_K7_PERFCTR0, + .event_map = amd_pmu_event_map, + .raw_event = amd_pmu_raw_event, + .max_events = ARRAY_SIZE(amd_perfmon_event_map), + .num_counters = 4, + .counter_bits = 48, + .counter_mask = (1ULL << 48) - 1, + /* use highest bit to detect overflow */ + .max_period = (1ULL << 47) - 1, +}; + +static int intel_pmu_init(void) +{ + union cpuid10_edx edx; + union cpuid10_eax eax; + unsigned int unused; + unsigned int ebx; + int version; + + if (!cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) + return -ENODEV; + + /* + * Check whether the Architectural PerfMon supports + * Branch Misses Retired Event or not. + */ + cpuid(10, &eax.full, &ebx, &unused, &edx.full); + if (eax.split.mask_length <= ARCH_PERFMON_BRANCH_MISSES_RETIRED) + return -ENODEV; + + version = eax.split.version_id; + if (version < 2) + return -ENODEV; + + x86_pmu = intel_pmu; + x86_pmu.version = version; + x86_pmu.num_counters = eax.split.num_counters; + x86_pmu.counter_bits = eax.split.bit_width; + x86_pmu.counter_mask = (1ULL << eax.split.bit_width) - 1; + + /* + * Quirk: v2 perfmon does not report fixed-purpose counters, so + * assume at least 3 counters: + */ + x86_pmu.num_counters_fixed = max((int)edx.split.num_counters_fixed, 3); + + rdmsrl(MSR_CORE_PERF_GLOBAL_CTRL, x86_pmu.intel_ctrl); + + /* + * Install the hw-cache-events table: + */ + switch (boot_cpu_data.x86_model) { + case 15: /* original 65 nm celeron/pentium/core2/xeon, "Merom"/"Conroe" */ + case 22: /* single-core 65 nm celeron/core2solo "Merom-L"/"Conroe-L" */ + case 23: /* current 45 nm celeron/core2/xeon "Penryn"/"Wolfdale" */ + case 29: /* six-core 45 nm xeon "Dunnington" */ + memcpy(hw_cache_event_ids, core2_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); + + pr_cont("Core2 events, "); + break; + default: + case 26: + memcpy(hw_cache_event_ids, nehalem_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); + + pr_cont("Nehalem/Corei7 events, "); + break; + case 28: + memcpy(hw_cache_event_ids, atom_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); + + pr_cont("Atom events, "); + break; + } + return 0; +} + +static int amd_pmu_init(void) +{ + x86_pmu = amd_pmu; + + switch (boot_cpu_data.x86) { + case 0x0f: + case 0x10: + case 0x11: + memcpy(hw_cache_event_ids, amd_0f_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); + + pr_cont("AMD Family 0f/10/11 events, "); + break; + } + return 0; +} + +void __init init_hw_perf_counters(void) +{ + int err; + + pr_info("Performance Counters: "); + + switch (boot_cpu_data.x86_vendor) { + case X86_VENDOR_INTEL: + err = intel_pmu_init(); + break; + case X86_VENDOR_AMD: + err = amd_pmu_init(); + break; + default: + return; + } + if (err != 0) { + pr_cont("no PMU driver, software counters only.\n"); + return; + } + + pr_cont("%s PMU driver.\n", x86_pmu.name); + + if (x86_pmu.num_counters > X86_PMC_MAX_GENERIC) { + x86_pmu.num_counters = X86_PMC_MAX_GENERIC; + WARN(1, KERN_ERR "hw perf counters %d > max(%d), clipping!", + x86_pmu.num_counters, X86_PMC_MAX_GENERIC); + } + perf_counter_mask = (1 << x86_pmu.num_counters) - 1; + perf_max_counters = x86_pmu.num_counters; + + if (x86_pmu.num_counters_fixed > X86_PMC_MAX_FIXED) { + x86_pmu.num_counters_fixed = X86_PMC_MAX_FIXED; + WARN(1, KERN_ERR "hw perf counters fixed %d > max(%d), clipping!", + x86_pmu.num_counters_fixed, X86_PMC_MAX_FIXED); + } + + perf_counter_mask |= + ((1LL << x86_pmu.num_counters_fixed)-1) << X86_PMC_IDX_FIXED; + + perf_counters_lapic_init(); + register_die_notifier(&perf_counter_nmi_notifier); + + pr_info("... version: %d\n", x86_pmu.version); + pr_info("... bit width: %d\n", x86_pmu.counter_bits); + pr_info("... generic counters: %d\n", x86_pmu.num_counters); + pr_info("... value mask: %016Lx\n", x86_pmu.counter_mask); + pr_info("... max period: %016Lx\n", x86_pmu.max_period); + pr_info("... fixed-purpose counters: %d\n", x86_pmu.num_counters_fixed); + pr_info("... counter mask: %016Lx\n", perf_counter_mask); +} + +static inline void x86_pmu_read(struct perf_counter *counter) +{ + x86_perf_counter_update(counter, &counter->hw, counter->hw.idx); +} + +static const struct pmu pmu = { + .enable = x86_pmu_enable, + .disable = x86_pmu_disable, + .read = x86_pmu_read, + .unthrottle = x86_pmu_unthrottle, +}; + +const struct pmu *hw_perf_counter_init(struct perf_counter *counter) +{ + int err; + + err = __hw_perf_counter_init(counter); + if (err) + return ERR_PTR(err); + + return &pmu; +} + +/* + * callchain support + */ + +static inline +void callchain_store(struct perf_callchain_entry *entry, unsigned long ip) +{ + if (entry->nr < MAX_STACK_DEPTH) + entry->ip[entry->nr++] = ip; +} + +static DEFINE_PER_CPU(struct perf_callchain_entry, irq_entry); +static DEFINE_PER_CPU(struct perf_callchain_entry, nmi_entry); + + +static void +backtrace_warning_symbol(void *data, char *msg, unsigned long symbol) +{ + /* Ignore warnings */ +} + +static void backtrace_warning(void *data, char *msg) +{ + /* Ignore warnings */ +} + +static int backtrace_stack(void *data, char *name) +{ + /* Don't bother with IRQ stacks for now */ + return -1; +} + +static void backtrace_address(void *data, unsigned long addr, int reliable) +{ + struct perf_callchain_entry *entry = data; + + if (reliable) + callchain_store(entry, addr); +} + +static const struct stacktrace_ops backtrace_ops = { + .warning = backtrace_warning, + .warning_symbol = backtrace_warning_symbol, + .stack = backtrace_stack, + .address = backtrace_address, +}; + +static void +perf_callchain_kernel(struct pt_regs *regs, struct perf_callchain_entry *entry) +{ + unsigned long bp; + char *stack; + int nr = entry->nr; + + callchain_store(entry, instruction_pointer(regs)); + + stack = ((char *)regs + sizeof(struct pt_regs)); +#ifdef CONFIG_FRAME_POINTER + bp = frame_pointer(regs); +#else + bp = 0; +#endif + + dump_trace(NULL, regs, (void *)stack, bp, &backtrace_ops, entry); + + entry->kernel = entry->nr - nr; +} + + +struct stack_frame { + const void __user *next_fp; + unsigned long return_address; +}; + +static int copy_stack_frame(const void __user *fp, struct stack_frame *frame) +{ + int ret; + + if (!access_ok(VERIFY_READ, fp, sizeof(*frame))) + return 0; + + ret = 1; + pagefault_disable(); + if (__copy_from_user_inatomic(frame, fp, sizeof(*frame))) + ret = 0; + pagefault_enable(); + + return ret; +} + +static void +perf_callchain_user(struct pt_regs *regs, struct perf_callchain_entry *entry) +{ + struct stack_frame frame; + const void __user *fp; + int nr = entry->nr; + + regs = (struct pt_regs *)current->thread.sp0 - 1; + fp = (void __user *)regs->bp; + + callchain_store(entry, regs->ip); + + while (entry->nr < MAX_STACK_DEPTH) { + frame.next_fp = NULL; + frame.return_address = 0; + + if (!copy_stack_frame(fp, &frame)) + break; + + if ((unsigned long)fp < user_stack_pointer(regs)) + break; + + callchain_store(entry, frame.return_address); + fp = frame.next_fp; + } + + entry->user = entry->nr - nr; +} + +static void +perf_do_callchain(struct pt_regs *regs, struct perf_callchain_entry *entry) +{ + int is_user; + + if (!regs) + return; + + is_user = user_mode(regs); + + if (!current || current->pid == 0) + return; + + if (is_user && current->state != TASK_RUNNING) + return; + + if (!is_user) + perf_callchain_kernel(regs, entry); + + if (current->mm) + perf_callchain_user(regs, entry); +} + +struct perf_callchain_entry *perf_callchain(struct pt_regs *regs) +{ + struct perf_callchain_entry *entry; + + if (in_nmi()) + entry = &__get_cpu_var(nmi_entry); + else + entry = &__get_cpu_var(irq_entry); + + entry->nr = 0; + entry->hv = 0; + entry->kernel = 0; + entry->user = 0; + + perf_do_callchain(regs, entry); + + return entry; +} diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c index f6c70a164e3..d6f5b9fbde3 100644 --- a/arch/x86/kernel/cpu/perfctr-watchdog.c +++ b/arch/x86/kernel/cpu/perfctr-watchdog.c @@ -19,8 +19,8 @@ #include <linux/nmi.h> #include <linux/kprobes.h> -#include <asm/genapic.h> -#include <asm/intel_arch_perfmon.h> +#include <asm/apic.h> +#include <asm/perf_counter.h> struct nmi_watchdog_ctlblk { unsigned int cccr_msr; diff --git a/arch/x86/kernel/ds.c b/arch/x86/kernel/ds.c index 87b67e3a765..48bfe138603 100644 --- a/arch/x86/kernel/ds.c +++ b/arch/x86/kernel/ds.c @@ -19,45 +19,61 @@ * Markus Metzger <markus.t.metzger@intel.com>, 2007-2009 */ - -#include <asm/ds.h> - -#include <linux/errno.h> +#include <linux/kernel.h> #include <linux/string.h> -#include <linux/slab.h> +#include <linux/errno.h> #include <linux/sched.h> +#include <linux/slab.h> #include <linux/mm.h> -#include <linux/kernel.h> +#include <linux/trace_clock.h> + +#include <asm/ds.h> +#include "ds_selftest.h" /* - * The configuration for a particular DS hardware implementation. + * The configuration for a particular DS hardware implementation: */ struct ds_configuration { - /* the name of the configuration */ - const char *name; - /* the size of one pointer-typed field in the DS structure and - in the BTS and PEBS buffers in bytes; - this covers the first 8 DS fields related to buffer management. */ - unsigned char sizeof_field; - /* the size of a BTS/PEBS record in bytes */ - unsigned char sizeof_rec[2]; - /* a series of bit-masks to control various features indexed - * by enum ds_feature */ - unsigned long ctl[dsf_ctl_max]; + /* The name of the configuration: */ + const char *name; + + /* The size of pointer-typed fields in DS, BTS, and PEBS: */ + unsigned char sizeof_ptr_field; + + /* The size of a BTS/PEBS record in bytes: */ + unsigned char sizeof_rec[2]; + + /* The number of pebs counter reset values in the DS structure. */ + unsigned char nr_counter_reset; + + /* Control bit-masks indexed by enum ds_feature: */ + unsigned long ctl[dsf_ctl_max]; }; -static DEFINE_PER_CPU(struct ds_configuration, ds_cfg_array); +static struct ds_configuration ds_cfg __read_mostly; + + +/* Maximal size of a DS configuration: */ +#define MAX_SIZEOF_DS 0x80 -#define ds_cfg per_cpu(ds_cfg_array, smp_processor_id()) +/* Maximal size of a BTS record: */ +#define MAX_SIZEOF_BTS (3 * 8) -#define MAX_SIZEOF_DS (12 * 8) /* maximal size of a DS configuration */ -#define MAX_SIZEOF_BTS (3 * 8) /* maximal size of a BTS record */ -#define DS_ALIGNMENT (1 << 3) /* BTS and PEBS buffer alignment */ +/* BTS and PEBS buffer alignment: */ +#define DS_ALIGNMENT (1 << 3) -#define BTS_CONTROL \ - (ds_cfg.ctl[dsf_bts] | ds_cfg.ctl[dsf_bts_kernel] | ds_cfg.ctl[dsf_bts_user] |\ - ds_cfg.ctl[dsf_bts_overflow]) +/* Number of buffer pointers in DS: */ +#define NUM_DS_PTR_FIELDS 8 +/* Size of a pebs reset value in DS: */ +#define PEBS_RESET_FIELD_SIZE 8 + +/* Mask of control bits in the DS MSR register: */ +#define BTS_CONTROL \ + ( ds_cfg.ctl[dsf_bts] | \ + ds_cfg.ctl[dsf_bts_kernel] | \ + ds_cfg.ctl[dsf_bts_user] | \ + ds_cfg.ctl[dsf_bts_overflow] ) /* * A BTS or PEBS tracer. @@ -66,29 +82,36 @@ static DEFINE_PER_CPU(struct ds_configuration, ds_cfg_array); * to identify tracers. */ struct ds_tracer { - /* the DS context (partially) owned by this tracer */ - struct ds_context *context; - /* the buffer provided on ds_request() and its size in bytes */ - void *buffer; - size_t size; + /* The DS context (partially) owned by this tracer. */ + struct ds_context *context; + /* The buffer provided on ds_request() and its size in bytes. */ + void *buffer; + size_t size; }; struct bts_tracer { - /* the common DS part */ - struct ds_tracer ds; - /* the trace including the DS configuration */ - struct bts_trace trace; - /* buffer overflow notification function */ - bts_ovfl_callback_t ovfl; + /* The common DS part: */ + struct ds_tracer ds; + + /* The trace including the DS configuration: */ + struct bts_trace trace; + + /* Buffer overflow notification function: */ + bts_ovfl_callback_t ovfl; + + /* Active flags affecting trace collection. */ + unsigned int flags; }; struct pebs_tracer { - /* the common DS part */ - struct ds_tracer ds; - /* the trace including the DS configuration */ - struct pebs_trace trace; - /* buffer overflow notification function */ - pebs_ovfl_callback_t ovfl; + /* The common DS part: */ + struct ds_tracer ds; + + /* The trace including the DS configuration: */ + struct pebs_trace trace; + + /* Buffer overflow notification function: */ + pebs_ovfl_callback_t ovfl; }; /* @@ -97,6 +120,7 @@ struct pebs_tracer { * * The DS configuration consists of the following fields; different * architetures vary in the size of those fields. + * * - double-word aligned base linear address of the BTS buffer * - write pointer into the BTS buffer * - end linear address of the BTS buffer (one byte beyond the end of @@ -135,21 +159,22 @@ enum ds_field { }; enum ds_qualifier { - ds_bts = 0, + ds_bts = 0, ds_pebs }; -static inline unsigned long ds_get(const unsigned char *base, - enum ds_qualifier qual, enum ds_field field) +static inline unsigned long +ds_get(const unsigned char *base, enum ds_qualifier qual, enum ds_field field) { - base += (ds_cfg.sizeof_field * (field + (4 * qual))); + base += (ds_cfg.sizeof_ptr_field * (field + (4 * qual))); return *(unsigned long *)base; } -static inline void ds_set(unsigned char *base, enum ds_qualifier qual, - enum ds_field field, unsigned long value) +static inline void +ds_set(unsigned char *base, enum ds_qualifier qual, enum ds_field field, + unsigned long value) { - base += (ds_cfg.sizeof_field * (field + (4 * qual))); + base += (ds_cfg.sizeof_ptr_field * (field + (4 * qual))); (*(unsigned long *)base) = value; } @@ -159,7 +184,6 @@ static inline void ds_set(unsigned char *base, enum ds_qualifier qual, */ static DEFINE_SPINLOCK(ds_lock); - /* * We either support (system-wide) per-cpu or per-thread allocation. * We distinguish the two based on the task_struct pointer, where a @@ -178,12 +202,28 @@ static DEFINE_SPINLOCK(ds_lock); */ static atomic_t tracers = ATOMIC_INIT(0); -static inline void get_tracer(struct task_struct *task) +static inline int get_tracer(struct task_struct *task) { - if (task) + int error; + + spin_lock_irq(&ds_lock); + + if (task) { + error = -EPERM; + if (atomic_read(&tracers) < 0) + goto out; atomic_inc(&tracers); - else + } else { + error = -EPERM; + if (atomic_read(&tracers) > 0) + goto out; atomic_dec(&tracers); + } + + error = 0; +out: + spin_unlock_irq(&ds_lock); + return error; } static inline void put_tracer(struct task_struct *task) @@ -194,14 +234,6 @@ static inline void put_tracer(struct task_struct *task) atomic_inc(&tracers); } -static inline int check_tracer(struct task_struct *task) -{ - return task ? - (atomic_read(&tracers) >= 0) : - (atomic_read(&tracers) <= 0); -} - - /* * The DS context is either attached to a thread or to a cpu: * - in the former case, the thread_struct contains a pointer to the @@ -213,61 +245,58 @@ static inline int check_tracer(struct task_struct *task) * deallocated when the last user puts the context. */ struct ds_context { - /* pointer to the DS configuration; goes into MSR_IA32_DS_AREA */ - unsigned char ds[MAX_SIZEOF_DS]; - /* the owner of the BTS and PEBS configuration, respectively */ - struct bts_tracer *bts_master; - struct pebs_tracer *pebs_master; - /* use count */ - unsigned long count; - /* a pointer to the context location inside the thread_struct - * or the per_cpu context array */ - struct ds_context **this; - /* a pointer to the task owning this context, or NULL, if the - * context is owned by a cpu */ - struct task_struct *task; -}; + /* The DS configuration; goes into MSR_IA32_DS_AREA: */ + unsigned char ds[MAX_SIZEOF_DS]; + + /* The owner of the BTS and PEBS configuration, respectively: */ + struct bts_tracer *bts_master; + struct pebs_tracer *pebs_master; -static DEFINE_PER_CPU(struct ds_context *, system_context_array); + /* Use count: */ + unsigned long count; -#define system_context per_cpu(system_context_array, smp_processor_id()) + /* Pointer to the context pointer field: */ + struct ds_context **this; + + /* The traced task; NULL for cpu tracing: */ + struct task_struct *task; + + /* The traced cpu; only valid if task is NULL: */ + int cpu; +}; +static DEFINE_PER_CPU(struct ds_context *, cpu_context); -static inline struct ds_context *ds_get_context(struct task_struct *task) + +static struct ds_context *ds_get_context(struct task_struct *task, int cpu) { struct ds_context **p_context = - (task ? &task->thread.ds_ctx : &system_context); + (task ? &task->thread.ds_ctx : &per_cpu(cpu_context, cpu)); struct ds_context *context = NULL; struct ds_context *new_context = NULL; - unsigned long irq; /* Chances are small that we already have a context. */ new_context = kzalloc(sizeof(*new_context), GFP_KERNEL); if (!new_context) return NULL; - spin_lock_irqsave(&ds_lock, irq); + spin_lock_irq(&ds_lock); context = *p_context; - if (!context) { + if (likely(!context)) { context = new_context; context->this = p_context; context->task = task; + context->cpu = cpu; context->count = 0; - if (task) - set_tsk_thread_flag(task, TIF_DS_AREA_MSR); - - if (!task || (task == current)) - wrmsrl(MSR_IA32_DS_AREA, (unsigned long)context->ds); - *p_context = context; } context->count++; - spin_unlock_irqrestore(&ds_lock, irq); + spin_unlock_irq(&ds_lock); if (context != new_context) kfree(new_context); @@ -275,8 +304,9 @@ static inline struct ds_context *ds_get_context(struct task_struct *task) return context; } -static inline void ds_put_context(struct ds_context *context) +static void ds_put_context(struct ds_context *context) { + struct task_struct *task; unsigned long irq; if (!context) @@ -291,17 +321,55 @@ static inline void ds_put_context(struct ds_context *context) *(context->this) = NULL; - if (context->task) - clear_tsk_thread_flag(context->task, TIF_DS_AREA_MSR); + task = context->task; + + if (task) + clear_tsk_thread_flag(task, TIF_DS_AREA_MSR); - if (!context->task || (context->task == current)) - wrmsrl(MSR_IA32_DS_AREA, 0); + /* + * We leave the (now dangling) pointer to the DS configuration in + * the DS_AREA msr. This is as good or as bad as replacing it with + * NULL - the hardware would crash if we enabled tracing. + * + * This saves us some problems with having to write an msr on a + * different cpu while preventing others from doing the same for the + * next context for that same cpu. + */ spin_unlock_irqrestore(&ds_lock, irq); + /* The context might still be in use for context switching. */ + if (task && (task != current)) + wait_task_context_switch(task); + kfree(context); } +static void ds_install_ds_area(struct ds_context *context) +{ + unsigned long ds; + + ds = (unsigned long)context->ds; + + /* + * There is a race between the bts master and the pebs master. + * + * The thread/cpu access is synchronized via get/put_cpu() for + * task tracing and via wrmsr_on_cpu for cpu tracing. + * + * If bts and pebs are collected for the same task or same cpu, + * the same confiuration is written twice. + */ + if (context->task) { + get_cpu(); + if (context->task == current) + wrmsrl(MSR_IA32_DS_AREA, ds); + set_tsk_thread_flag(context->task, TIF_DS_AREA_MSR); + put_cpu(); + } else + wrmsr_on_cpu(context->cpu, MSR_IA32_DS_AREA, + (u32)((u64)ds), (u32)((u64)ds >> 32)); +} /* * Call the tracer's callback on a buffer overflow. @@ -332,9 +400,9 @@ static void ds_overflow(struct ds_context *context, enum ds_qualifier qual) * The remainder of any partially written record is zeroed out. * * context: the DS context - * qual: the buffer type - * record: the data to write - * size: the size of the data + * qual: the buffer type + * record: the data to write + * size: the size of the data */ static int ds_write(struct ds_context *context, enum ds_qualifier qual, const void *record, size_t size) @@ -349,14 +417,14 @@ static int ds_write(struct ds_context *context, enum ds_qualifier qual, unsigned long write_size, adj_write_size; /* - * write as much as possible without producing an + * Write as much as possible without producing an * overflow interrupt. * - * interrupt_threshold must either be + * Interrupt_threshold must either be * - bigger than absolute_maximum or * - point to a record between buffer_base and absolute_maximum * - * index points to a valid record. + * Index points to a valid record. */ base = ds_get(context->ds, qual, ds_buffer_base); index = ds_get(context->ds, qual, ds_index); @@ -365,8 +433,10 @@ static int ds_write(struct ds_context *context, enum ds_qualifier qual, write_end = min(end, int_th); - /* if we are already beyond the interrupt threshold, - * we fill the entire buffer */ + /* + * If we are already beyond the interrupt threshold, + * we fill the entire buffer. + */ if (write_end <= index) write_end = end; @@ -383,7 +453,7 @@ static int ds_write(struct ds_context *context, enum ds_qualifier qual, adj_write_size = write_size / ds_cfg.sizeof_rec[qual]; adj_write_size *= ds_cfg.sizeof_rec[qual]; - /* zero out trailing bytes */ + /* Zero out trailing bytes. */ memset((char *)index + write_size, 0, adj_write_size - write_size); index += adj_write_size; @@ -410,7 +480,7 @@ static int ds_write(struct ds_context *context, enum ds_qualifier qual, * Later architectures use 64bit pointers throughout, whereas earlier * architectures use 32bit pointers in 32bit mode. * - * We compute the base address for the first 8 fields based on: + * We compute the base address for the fields based on: * - the field size stored in the DS configuration * - the relative field position * @@ -431,23 +501,23 @@ enum bts_field { bts_to, bts_flags, - bts_qual = bts_from, - bts_jiffies = bts_to, - bts_pid = bts_flags, + bts_qual = bts_from, + bts_clock = bts_to, + bts_pid = bts_flags, - bts_qual_mask = (bts_qual_max - 1), - bts_escape = ((unsigned long)-1 & ~bts_qual_mask) + bts_qual_mask = (bts_qual_max - 1), + bts_escape = ((unsigned long)-1 & ~bts_qual_mask) }; static inline unsigned long bts_get(const char *base, enum bts_field field) { - base += (ds_cfg.sizeof_field * field); + base += (ds_cfg.sizeof_ptr_field * field); return *(unsigned long *)base; } static inline void bts_set(char *base, enum bts_field field, unsigned long val) { - base += (ds_cfg.sizeof_field * field);; + base += (ds_cfg.sizeof_ptr_field * field);; (*(unsigned long *)base) = val; } @@ -463,8 +533,8 @@ static inline void bts_set(char *base, enum bts_field field, unsigned long val) * * return: bytes read/written on success; -Eerrno, otherwise */ -static int bts_read(struct bts_tracer *tracer, const void *at, - struct bts_struct *out) +static int +bts_read(struct bts_tracer *tracer, const void *at, struct bts_struct *out) { if (!tracer) return -EINVAL; @@ -478,8 +548,8 @@ static int bts_read(struct bts_tracer *tracer, const void *at, memset(out, 0, sizeof(*out)); if ((bts_get(at, bts_qual) & ~bts_qual_mask) == bts_escape) { out->qualifier = (bts_get(at, bts_qual) & bts_qual_mask); - out->variant.timestamp.jiffies = bts_get(at, bts_jiffies); - out->variant.timestamp.pid = bts_get(at, bts_pid); + out->variant.event.clock = bts_get(at, bts_clock); + out->variant.event.pid = bts_get(at, bts_pid); } else { out->qualifier = bts_branch; out->variant.lbr.from = bts_get(at, bts_from); @@ -516,8 +586,8 @@ static int bts_write(struct bts_tracer *tracer, const struct bts_struct *in) case bts_task_arrives: case bts_task_departs: bts_set(raw, bts_qual, (bts_escape | in->qualifier)); - bts_set(raw, bts_jiffies, in->variant.timestamp.jiffies); - bts_set(raw, bts_pid, in->variant.timestamp.pid); + bts_set(raw, bts_clock, in->variant.event.clock); + bts_set(raw, bts_pid, in->variant.event.pid); break; default: return -EINVAL; @@ -555,7 +625,8 @@ static void ds_init_ds_trace(struct ds_trace *trace, enum ds_qualifier qual, unsigned int flags) { unsigned long buffer, adj; - /* adjust the buffer address and size to meet alignment + /* + * Adjust the buffer address and size to meet alignment * constraints: * - buffer is double-word aligned * - size is multiple of record size @@ -577,9 +648,11 @@ static void ds_init_ds_trace(struct ds_trace *trace, enum ds_qualifier qual, trace->begin = (void *)buffer; trace->top = trace->begin; trace->end = (void *)(buffer + size); - /* The value for 'no threshold' is -1, which will set the + /* + * The value for 'no threshold' is -1, which will set the * threshold outside of the buffer, just like we want it. */ + ith *= ds_cfg.sizeof_rec[qual]; trace->ith = (void *)(buffer + size - ith); trace->flags = flags; @@ -588,18 +661,27 @@ static void ds_init_ds_trace(struct ds_trace *trace, enum ds_qualifier qual, static int ds_request(struct ds_tracer *tracer, struct ds_trace *trace, enum ds_qualifier qual, struct task_struct *task, - void *base, size_t size, size_t th, unsigned int flags) + int cpu, void *base, size_t size, size_t th) { struct ds_context *context; int error; + size_t req_size; + + error = -EOPNOTSUPP; + if (!ds_cfg.sizeof_rec[qual]) + goto out; error = -EINVAL; if (!base) goto out; - /* we require some space to do alignment adjustments below */ + req_size = ds_cfg.sizeof_rec[qual]; + /* We might need space for alignment adjustments. */ + if (!IS_ALIGNED((unsigned long)base, DS_ALIGNMENT)) + req_size += DS_ALIGNMENT; + error = -EINVAL; - if (size < (DS_ALIGNMENT + ds_cfg.sizeof_rec[qual])) + if (size < req_size) goto out; if (th != (size_t)-1) { @@ -614,182 +696,318 @@ static int ds_request(struct ds_tracer *tracer, struct ds_trace *trace, tracer->size = size; error = -ENOMEM; - context = ds_get_context(task); + context = ds_get_context(task, cpu); if (!context) goto out; tracer->context = context; - ds_init_ds_trace(trace, qual, base, size, th, flags); + /* + * Defer any tracer-specific initialization work for the context until + * context ownership has been clarified. + */ error = 0; out: return error; } -struct bts_tracer *ds_request_bts(struct task_struct *task, - void *base, size_t size, - bts_ovfl_callback_t ovfl, size_t th, - unsigned int flags) +static struct bts_tracer *ds_request_bts(struct task_struct *task, int cpu, + void *base, size_t size, + bts_ovfl_callback_t ovfl, size_t th, + unsigned int flags) { struct bts_tracer *tracer; - unsigned long irq; int error; + /* Buffer overflow notification is not yet implemented. */ error = -EOPNOTSUPP; - if (!ds_cfg.ctl[dsf_bts]) + if (ovfl) goto out; - /* buffer overflow notification is not yet implemented */ - error = -EOPNOTSUPP; - if (ovfl) + error = get_tracer(task); + if (error < 0) goto out; error = -ENOMEM; tracer = kzalloc(sizeof(*tracer), GFP_KERNEL); if (!tracer) - goto out; + goto out_put_tracer; tracer->ovfl = ovfl; + /* Do some more error checking and acquire a tracing context. */ error = ds_request(&tracer->ds, &tracer->trace.ds, - ds_bts, task, base, size, th, flags); + ds_bts, task, cpu, base, size, th); if (error < 0) goto out_tracer; - - spin_lock_irqsave(&ds_lock, irq); - - error = -EPERM; - if (!check_tracer(task)) - goto out_unlock; - get_tracer(task); + /* Claim the bts part of the tracing context we acquired above. */ + spin_lock_irq(&ds_lock); error = -EPERM; if (tracer->ds.context->bts_master) - goto out_put_tracer; + goto out_unlock; tracer->ds.context->bts_master = tracer; - spin_unlock_irqrestore(&ds_lock, irq); + spin_unlock_irq(&ds_lock); + /* + * Now that we own the bts part of the context, let's complete the + * initialization for that part. + */ + ds_init_ds_trace(&tracer->trace.ds, ds_bts, base, size, th, flags); + ds_write_config(tracer->ds.context, &tracer->trace.ds, ds_bts); + ds_install_ds_area(tracer->ds.context); tracer->trace.read = bts_read; tracer->trace.write = bts_write; - ds_write_config(tracer->ds.context, &tracer->trace.ds, ds_bts); + /* Start tracing. */ ds_resume_bts(tracer); return tracer; - out_put_tracer: - put_tracer(task); out_unlock: - spin_unlock_irqrestore(&ds_lock, irq); + spin_unlock_irq(&ds_lock); ds_put_context(tracer->ds.context); out_tracer: kfree(tracer); + out_put_tracer: + put_tracer(task); out: return ERR_PTR(error); } -struct pebs_tracer *ds_request_pebs(struct task_struct *task, - void *base, size_t size, - pebs_ovfl_callback_t ovfl, size_t th, - unsigned int flags) +struct bts_tracer *ds_request_bts_task(struct task_struct *task, + void *base, size_t size, + bts_ovfl_callback_t ovfl, + size_t th, unsigned int flags) +{ + return ds_request_bts(task, 0, base, size, ovfl, th, flags); +} + +struct bts_tracer *ds_request_bts_cpu(int cpu, void *base, size_t size, + bts_ovfl_callback_t ovfl, + size_t th, unsigned int flags) +{ + return ds_request_bts(NULL, cpu, base, size, ovfl, th, flags); +} + +static struct pebs_tracer *ds_request_pebs(struct task_struct *task, int cpu, + void *base, size_t size, + pebs_ovfl_callback_t ovfl, size_t th, + unsigned int flags) { struct pebs_tracer *tracer; - unsigned long irq; int error; - /* buffer overflow notification is not yet implemented */ + /* Buffer overflow notification is not yet implemented. */ error = -EOPNOTSUPP; if (ovfl) goto out; + error = get_tracer(task); + if (error < 0) + goto out; + error = -ENOMEM; tracer = kzalloc(sizeof(*tracer), GFP_KERNEL); if (!tracer) - goto out; + goto out_put_tracer; tracer->ovfl = ovfl; + /* Do some more error checking and acquire a tracing context. */ error = ds_request(&tracer->ds, &tracer->trace.ds, - ds_pebs, task, base, size, th, flags); + ds_pebs, task, cpu, base, size, th); if (error < 0) goto out_tracer; - spin_lock_irqsave(&ds_lock, irq); - - error = -EPERM; - if (!check_tracer(task)) - goto out_unlock; - get_tracer(task); + /* Claim the pebs part of the tracing context we acquired above. */ + spin_lock_irq(&ds_lock); error = -EPERM; if (tracer->ds.context->pebs_master) - goto out_put_tracer; + goto out_unlock; tracer->ds.context->pebs_master = tracer; - spin_unlock_irqrestore(&ds_lock, irq); + spin_unlock_irq(&ds_lock); + /* + * Now that we own the pebs part of the context, let's complete the + * initialization for that part. + */ + ds_init_ds_trace(&tracer->trace.ds, ds_pebs, base, size, th, flags); ds_write_config(tracer->ds.context, &tracer->trace.ds, ds_pebs); + ds_install_ds_area(tracer->ds.context); + + /* Start tracing. */ ds_resume_pebs(tracer); return tracer; - out_put_tracer: - put_tracer(task); out_unlock: - spin_unlock_irqrestore(&ds_lock, irq); + spin_unlock_irq(&ds_lock); ds_put_context(tracer->ds.context); out_tracer: kfree(tracer); + out_put_tracer: + put_tracer(task); out: return ERR_PTR(error); } -void ds_release_bts(struct bts_tracer *tracer) +struct pebs_tracer *ds_request_pebs_task(struct task_struct *task, + void *base, size_t size, + pebs_ovfl_callback_t ovfl, + size_t th, unsigned int flags) { - if (!tracer) - return; + return ds_request_pebs(task, 0, base, size, ovfl, th, flags); +} - ds_suspend_bts(tracer); +struct pebs_tracer *ds_request_pebs_cpu(int cpu, void *base, size_t size, + pebs_ovfl_callback_t ovfl, + size_t th, unsigned int flags) +{ + return ds_request_pebs(NULL, cpu, base, size, ovfl, th, flags); +} + +static void ds_free_bts(struct bts_tracer *tracer) +{ + struct task_struct *task; + + task = tracer->ds.context->task; WARN_ON_ONCE(tracer->ds.context->bts_master != tracer); tracer->ds.context->bts_master = NULL; - put_tracer(tracer->ds.context->task); + /* Make sure tracing stopped and the tracer is not in use. */ + if (task && (task != current)) + wait_task_context_switch(task); + ds_put_context(tracer->ds.context); + put_tracer(task); kfree(tracer); } +void ds_release_bts(struct bts_tracer *tracer) +{ + might_sleep(); + + if (!tracer) + return; + + ds_suspend_bts(tracer); + ds_free_bts(tracer); +} + +int ds_release_bts_noirq(struct bts_tracer *tracer) +{ + struct task_struct *task; + unsigned long irq; + int error; + + if (!tracer) + return 0; + + task = tracer->ds.context->task; + + local_irq_save(irq); + + error = -EPERM; + if (!task && + (tracer->ds.context->cpu != smp_processor_id())) + goto out; + + error = -EPERM; + if (task && (task != current)) + goto out; + + ds_suspend_bts_noirq(tracer); + ds_free_bts(tracer); + + error = 0; + out: + local_irq_restore(irq); + return error; +} + +static void update_task_debugctlmsr(struct task_struct *task, + unsigned long debugctlmsr) +{ + task->thread.debugctlmsr = debugctlmsr; + + get_cpu(); + if (task == current) + update_debugctlmsr(debugctlmsr); + put_cpu(); +} + void ds_suspend_bts(struct bts_tracer *tracer) { struct task_struct *task; + unsigned long debugctlmsr; + int cpu; if (!tracer) return; + tracer->flags = 0; + task = tracer->ds.context->task; + cpu = tracer->ds.context->cpu; - if (!task || (task == current)) - update_debugctlmsr(get_debugctlmsr() & ~BTS_CONTROL); + WARN_ON(!task && irqs_disabled()); - if (task) { - task->thread.debugctlmsr &= ~BTS_CONTROL; + debugctlmsr = (task ? + task->thread.debugctlmsr : + get_debugctlmsr_on_cpu(cpu)); + debugctlmsr &= ~BTS_CONTROL; - if (!task->thread.debugctlmsr) - clear_tsk_thread_flag(task, TIF_DEBUGCTLMSR); - } + if (task) + update_task_debugctlmsr(task, debugctlmsr); + else + update_debugctlmsr_on_cpu(cpu, debugctlmsr); } -void ds_resume_bts(struct bts_tracer *tracer) +int ds_suspend_bts_noirq(struct bts_tracer *tracer) { struct task_struct *task; - unsigned long control; + unsigned long debugctlmsr, irq; + int cpu, error = 0; if (!tracer) - return; + return 0; + + tracer->flags = 0; task = tracer->ds.context->task; + cpu = tracer->ds.context->cpu; + + local_irq_save(irq); + + error = -EPERM; + if (!task && (cpu != smp_processor_id())) + goto out; + + debugctlmsr = (task ? + task->thread.debugctlmsr : + get_debugctlmsr()); + debugctlmsr &= ~BTS_CONTROL; + + if (task) + update_task_debugctlmsr(task, debugctlmsr); + else + update_debugctlmsr(debugctlmsr); + + error = 0; + out: + local_irq_restore(irq); + return error; +} + +static unsigned long ds_bts_control(struct bts_tracer *tracer) +{ + unsigned long control; control = ds_cfg.ctl[dsf_bts]; if (!(tracer->trace.ds.flags & BTS_KERNEL)) @@ -797,41 +1015,149 @@ void ds_resume_bts(struct bts_tracer *tracer) if (!(tracer->trace.ds.flags & BTS_USER)) control |= ds_cfg.ctl[dsf_bts_user]; - if (task) { - task->thread.debugctlmsr |= control; - set_tsk_thread_flag(task, TIF_DEBUGCTLMSR); - } - - if (!task || (task == current)) - update_debugctlmsr(get_debugctlmsr() | control); + return control; } -void ds_release_pebs(struct pebs_tracer *tracer) +void ds_resume_bts(struct bts_tracer *tracer) { + struct task_struct *task; + unsigned long debugctlmsr; + int cpu; + if (!tracer) return; - ds_suspend_pebs(tracer); + tracer->flags = tracer->trace.ds.flags; + + task = tracer->ds.context->task; + cpu = tracer->ds.context->cpu; + + WARN_ON(!task && irqs_disabled()); + + debugctlmsr = (task ? + task->thread.debugctlmsr : + get_debugctlmsr_on_cpu(cpu)); + debugctlmsr |= ds_bts_control(tracer); + + if (task) + update_task_debugctlmsr(task, debugctlmsr); + else + update_debugctlmsr_on_cpu(cpu, debugctlmsr); +} + +int ds_resume_bts_noirq(struct bts_tracer *tracer) +{ + struct task_struct *task; + unsigned long debugctlmsr, irq; + int cpu, error = 0; + + if (!tracer) + return 0; + + tracer->flags = tracer->trace.ds.flags; + + task = tracer->ds.context->task; + cpu = tracer->ds.context->cpu; + + local_irq_save(irq); + + error = -EPERM; + if (!task && (cpu != smp_processor_id())) + goto out; + + debugctlmsr = (task ? + task->thread.debugctlmsr : + get_debugctlmsr()); + debugctlmsr |= ds_bts_control(tracer); + + if (task) + update_task_debugctlmsr(task, debugctlmsr); + else + update_debugctlmsr(debugctlmsr); + + error = 0; + out: + local_irq_restore(irq); + return error; +} + +static void ds_free_pebs(struct pebs_tracer *tracer) +{ + struct task_struct *task; + + task = tracer->ds.context->task; WARN_ON_ONCE(tracer->ds.context->pebs_master != tracer); tracer->ds.context->pebs_master = NULL; - put_tracer(tracer->ds.context->task); ds_put_context(tracer->ds.context); + put_tracer(task); kfree(tracer); } +void ds_release_pebs(struct pebs_tracer *tracer) +{ + might_sleep(); + + if (!tracer) + return; + + ds_suspend_pebs(tracer); + ds_free_pebs(tracer); +} + +int ds_release_pebs_noirq(struct pebs_tracer *tracer) +{ + struct task_struct *task; + unsigned long irq; + int error; + + if (!tracer) + return 0; + + task = tracer->ds.context->task; + + local_irq_save(irq); + + error = -EPERM; + if (!task && + (tracer->ds.context->cpu != smp_processor_id())) + goto out; + + error = -EPERM; + if (task && (task != current)) + goto out; + + ds_suspend_pebs_noirq(tracer); + ds_free_pebs(tracer); + + error = 0; + out: + local_irq_restore(irq); + return error; +} + void ds_suspend_pebs(struct pebs_tracer *tracer) { } +int ds_suspend_pebs_noirq(struct pebs_tracer *tracer) +{ + return 0; +} + void ds_resume_pebs(struct pebs_tracer *tracer) { } +int ds_resume_pebs_noirq(struct pebs_tracer *tracer) +{ + return 0; +} + const struct bts_trace *ds_read_bts(struct bts_tracer *tracer) { if (!tracer) @@ -847,8 +1173,12 @@ const struct pebs_trace *ds_read_pebs(struct pebs_tracer *tracer) return NULL; ds_read_config(tracer->ds.context, &tracer->trace.ds, ds_pebs); - tracer->trace.reset_value = - *(u64 *)(tracer->ds.context->ds + (ds_cfg.sizeof_field * 8)); + + tracer->trace.counters = ds_cfg.nr_counter_reset; + memcpy(tracer->trace.counter_reset, + tracer->ds.context->ds + + (NUM_DS_PTR_FIELDS * ds_cfg.sizeof_ptr_field), + ds_cfg.nr_counter_reset * PEBS_RESET_FIELD_SIZE); return &tracer->trace; } @@ -873,18 +1203,24 @@ int ds_reset_pebs(struct pebs_tracer *tracer) tracer->trace.ds.top = tracer->trace.ds.begin; - ds_set(tracer->ds.context->ds, ds_bts, ds_index, + ds_set(tracer->ds.context->ds, ds_pebs, ds_index, (unsigned long)tracer->trace.ds.top); return 0; } -int ds_set_pebs_reset(struct pebs_tracer *tracer, u64 value) +int ds_set_pebs_reset(struct pebs_tracer *tracer, + unsigned int counter, u64 value) { if (!tracer) return -EINVAL; - *(u64 *)(tracer->ds.context->ds + (ds_cfg.sizeof_field * 8)) = value; + if (ds_cfg.nr_counter_reset < counter) + return -EINVAL; + + *(u64 *)(tracer->ds.context->ds + + (NUM_DS_PTR_FIELDS * ds_cfg.sizeof_ptr_field) + + (counter * PEBS_RESET_FIELD_SIZE)) = value; return 0; } @@ -894,73 +1230,117 @@ static const struct ds_configuration ds_cfg_netburst = { .ctl[dsf_bts] = (1 << 2) | (1 << 3), .ctl[dsf_bts_kernel] = (1 << 5), .ctl[dsf_bts_user] = (1 << 6), - - .sizeof_field = sizeof(long), - .sizeof_rec[ds_bts] = sizeof(long) * 3, -#ifdef __i386__ - .sizeof_rec[ds_pebs] = sizeof(long) * 10, -#else - .sizeof_rec[ds_pebs] = sizeof(long) * 18, -#endif + .nr_counter_reset = 1, }; static const struct ds_configuration ds_cfg_pentium_m = { .name = "Pentium M", .ctl[dsf_bts] = (1 << 6) | (1 << 7), - - .sizeof_field = sizeof(long), - .sizeof_rec[ds_bts] = sizeof(long) * 3, -#ifdef __i386__ - .sizeof_rec[ds_pebs] = sizeof(long) * 10, -#else - .sizeof_rec[ds_pebs] = sizeof(long) * 18, -#endif + .nr_counter_reset = 1, }; static const struct ds_configuration ds_cfg_core2_atom = { .name = "Core 2/Atom", .ctl[dsf_bts] = (1 << 6) | (1 << 7), .ctl[dsf_bts_kernel] = (1 << 9), .ctl[dsf_bts_user] = (1 << 10), - - .sizeof_field = 8, - .sizeof_rec[ds_bts] = 8 * 3, - .sizeof_rec[ds_pebs] = 8 * 18, + .nr_counter_reset = 1, +}; +static const struct ds_configuration ds_cfg_core_i7 = { + .name = "Core i7", + .ctl[dsf_bts] = (1 << 6) | (1 << 7), + .ctl[dsf_bts_kernel] = (1 << 9), + .ctl[dsf_bts_user] = (1 << 10), + .nr_counter_reset = 4, }; static void -ds_configure(const struct ds_configuration *cfg) +ds_configure(const struct ds_configuration *cfg, + struct cpuinfo_x86 *cpu) { + unsigned long nr_pebs_fields = 0; + + printk(KERN_INFO "[ds] using %s configuration\n", cfg->name); + +#ifdef __i386__ + nr_pebs_fields = 10; +#else + nr_pebs_fields = 18; +#endif + + /* + * Starting with version 2, architectural performance + * monitoring supports a format specifier. + */ + if ((cpuid_eax(0xa) & 0xff) > 1) { + unsigned long perf_capabilities, format; + + rdmsrl(MSR_IA32_PERF_CAPABILITIES, perf_capabilities); + + format = (perf_capabilities >> 8) & 0xf; + + switch (format) { + case 0: + nr_pebs_fields = 18; + break; + case 1: + nr_pebs_fields = 22; + break; + default: + printk(KERN_INFO + "[ds] unknown PEBS format: %lu\n", format); + nr_pebs_fields = 0; + break; + } + } + memset(&ds_cfg, 0, sizeof(ds_cfg)); ds_cfg = *cfg; - printk(KERN_INFO "[ds] using %s configuration\n", ds_cfg.name); + ds_cfg.sizeof_ptr_field = + (cpu_has(cpu, X86_FEATURE_DTES64) ? 8 : 4); + + ds_cfg.sizeof_rec[ds_bts] = ds_cfg.sizeof_ptr_field * 3; + ds_cfg.sizeof_rec[ds_pebs] = ds_cfg.sizeof_ptr_field * nr_pebs_fields; - if (!cpu_has_bts) { - ds_cfg.ctl[dsf_bts] = 0; + if (!cpu_has(cpu, X86_FEATURE_BTS)) { + ds_cfg.sizeof_rec[ds_bts] = 0; printk(KERN_INFO "[ds] bts not available\n"); } - if (!cpu_has_pebs) + if (!cpu_has(cpu, X86_FEATURE_PEBS)) { + ds_cfg.sizeof_rec[ds_pebs] = 0; printk(KERN_INFO "[ds] pebs not available\n"); + } + + printk(KERN_INFO "[ds] sizes: address: %u bit, ", + 8 * ds_cfg.sizeof_ptr_field); + printk("bts/pebs record: %u/%u bytes\n", + ds_cfg.sizeof_rec[ds_bts], ds_cfg.sizeof_rec[ds_pebs]); - WARN_ON_ONCE(MAX_SIZEOF_DS < (12 * ds_cfg.sizeof_field)); + WARN_ON_ONCE(MAX_PEBS_COUNTERS < ds_cfg.nr_counter_reset); } void __cpuinit ds_init_intel(struct cpuinfo_x86 *c) { + /* Only configure the first cpu. Others are identical. */ + if (ds_cfg.name) + return; + switch (c->x86) { case 0x6: switch (c->x86_model) { case 0x9: case 0xd: /* Pentium M */ - ds_configure(&ds_cfg_pentium_m); + ds_configure(&ds_cfg_pentium_m, c); break; case 0xf: case 0x17: /* Core2 */ case 0x1c: /* Atom */ - ds_configure(&ds_cfg_core2_atom); + ds_configure(&ds_cfg_core2_atom, c); + break; + case 0x1a: /* Core i7 */ + ds_configure(&ds_cfg_core_i7, c); break; - case 0x1a: /* i7 */ default: - /* sorry, don't know about them */ + /* Sorry, don't know about them. */ break; } break; @@ -969,64 +1349,89 @@ void __cpuinit ds_init_intel(struct cpuinfo_x86 *c) case 0x0: case 0x1: case 0x2: /* Netburst */ - ds_configure(&ds_cfg_netburst); + ds_configure(&ds_cfg_netburst, c); break; default: - /* sorry, don't know about them */ + /* Sorry, don't know about them. */ break; } break; default: - /* sorry, don't know about them */ + /* Sorry, don't know about them. */ break; } } +static inline void ds_take_timestamp(struct ds_context *context, + enum bts_qualifier qualifier, + struct task_struct *task) +{ + struct bts_tracer *tracer = context->bts_master; + struct bts_struct ts; + + /* Prevent compilers from reading the tracer pointer twice. */ + barrier(); + + if (!tracer || !(tracer->flags & BTS_TIMESTAMPS)) + return; + + memset(&ts, 0, sizeof(ts)); + ts.qualifier = qualifier; + ts.variant.event.clock = trace_clock_global(); + ts.variant.event.pid = task->pid; + + bts_write(tracer, &ts); +} + /* * Change the DS configuration from tracing prev to tracing next. */ void ds_switch_to(struct task_struct *prev, struct task_struct *next) { - struct ds_context *prev_ctx = prev->thread.ds_ctx; - struct ds_context *next_ctx = next->thread.ds_ctx; + struct ds_context *prev_ctx = prev->thread.ds_ctx; + struct ds_context *next_ctx = next->thread.ds_ctx; + unsigned long debugctlmsr = next->thread.debugctlmsr; + + /* Make sure all data is read before we start. */ + barrier(); if (prev_ctx) { update_debugctlmsr(0); - if (prev_ctx->bts_master && - (prev_ctx->bts_master->trace.ds.flags & BTS_TIMESTAMPS)) { - struct bts_struct ts = { - .qualifier = bts_task_departs, - .variant.timestamp.jiffies = jiffies_64, - .variant.timestamp.pid = prev->pid - }; - bts_write(prev_ctx->bts_master, &ts); - } + ds_take_timestamp(prev_ctx, bts_task_departs, prev); } if (next_ctx) { - if (next_ctx->bts_master && - (next_ctx->bts_master->trace.ds.flags & BTS_TIMESTAMPS)) { - struct bts_struct ts = { - .qualifier = bts_task_arrives, - .variant.timestamp.jiffies = jiffies_64, - .variant.timestamp.pid = next->pid - }; - bts_write(next_ctx->bts_master, &ts); - } + ds_take_timestamp(next_ctx, bts_task_arrives, next); wrmsrl(MSR_IA32_DS_AREA, (unsigned long)next_ctx->ds); } - update_debugctlmsr(next->thread.debugctlmsr); + update_debugctlmsr(debugctlmsr); } -void ds_copy_thread(struct task_struct *tsk, struct task_struct *father) +static __init int ds_selftest(void) { - clear_tsk_thread_flag(tsk, TIF_DS_AREA_MSR); - tsk->thread.ds_ctx = NULL; -} + if (ds_cfg.sizeof_rec[ds_bts]) { + int error; -void ds_exit_thread(struct task_struct *tsk) -{ + error = ds_selftest_bts(); + if (error) { + WARN(1, "[ds] selftest failed. disabling bts.\n"); + ds_cfg.sizeof_rec[ds_bts] = 0; + } + } + + if (ds_cfg.sizeof_rec[ds_pebs]) { + int error; + + error = ds_selftest_pebs(); + if (error) { + WARN(1, "[ds] selftest failed. disabling pebs.\n"); + ds_cfg.sizeof_rec[ds_pebs] = 0; + } + } + + return 0; } +device_initcall(ds_selftest); diff --git a/arch/x86/kernel/ds_selftest.c b/arch/x86/kernel/ds_selftest.c new file mode 100644 index 00000000000..6bc7c199ab9 --- /dev/null +++ b/arch/x86/kernel/ds_selftest.c @@ -0,0 +1,408 @@ +/* + * Debug Store support - selftest + * + * + * Copyright (C) 2009 Intel Corporation. + * Markus Metzger <markus.t.metzger@intel.com>, 2009 + */ + +#include "ds_selftest.h" + +#include <linux/kernel.h> +#include <linux/string.h> +#include <linux/smp.h> +#include <linux/cpu.h> + +#include <asm/ds.h> + + +#define BUFFER_SIZE 521 /* Intentionally chose an odd size. */ +#define SMALL_BUFFER_SIZE 24 /* A single bts entry. */ + +struct ds_selftest_bts_conf { + struct bts_tracer *tracer; + int error; + int (*suspend)(struct bts_tracer *); + int (*resume)(struct bts_tracer *); +}; + +static int ds_selftest_bts_consistency(const struct bts_trace *trace) +{ + int error = 0; + + if (!trace) { + printk(KERN_CONT "failed to access trace..."); + /* Bail out. Other tests are pointless. */ + return -1; + } + + if (!trace->read) { + printk(KERN_CONT "bts read not available..."); + error = -1; + } + + /* Do some sanity checks on the trace configuration. */ + if (!trace->ds.n) { + printk(KERN_CONT "empty bts buffer..."); + error = -1; + } + if (!trace->ds.size) { + printk(KERN_CONT "bad bts trace setup..."); + error = -1; + } + if (trace->ds.end != + (char *)trace->ds.begin + (trace->ds.n * trace->ds.size)) { + printk(KERN_CONT "bad bts buffer setup..."); + error = -1; + } + /* + * We allow top in [begin; end], since its not clear when the + * overflow adjustment happens: after the increment or before the + * write. + */ + if ((trace->ds.top < trace->ds.begin) || + (trace->ds.end < trace->ds.top)) { + printk(KERN_CONT "bts top out of bounds..."); + error = -1; + } + + return error; +} + +static int ds_selftest_bts_read(struct bts_tracer *tracer, + const struct bts_trace *trace, + const void *from, const void *to) +{ + const unsigned char *at; + + /* + * Check a few things which do not belong to this test. + * They should be covered by other tests. + */ + if (!trace) + return -1; + + if (!trace->read) + return -1; + + if (to < from) + return -1; + + if (from < trace->ds.begin) + return -1; + + if (trace->ds.end < to) + return -1; + + if (!trace->ds.size) + return -1; + + /* Now to the test itself. */ + for (at = from; (void *)at < to; at += trace->ds.size) { + struct bts_struct bts; + unsigned long index; + int error; + + if (((void *)at - trace->ds.begin) % trace->ds.size) { + printk(KERN_CONT + "read from non-integer index..."); + return -1; + } + index = ((void *)at - trace->ds.begin) / trace->ds.size; + + memset(&bts, 0, sizeof(bts)); + error = trace->read(tracer, at, &bts); + if (error < 0) { + printk(KERN_CONT + "error reading bts trace at [%lu] (0x%p)...", + index, at); + return error; + } + + switch (bts.qualifier) { + case BTS_BRANCH: + break; + default: + printk(KERN_CONT + "unexpected bts entry %llu at [%lu] (0x%p)...", + bts.qualifier, index, at); + return -1; + } + } + + return 0; +} + +static void ds_selftest_bts_cpu(void *arg) +{ + struct ds_selftest_bts_conf *conf = arg; + const struct bts_trace *trace; + void *top; + + if (IS_ERR(conf->tracer)) { + conf->error = PTR_ERR(conf->tracer); + conf->tracer = NULL; + + printk(KERN_CONT + "initialization failed (err: %d)...", conf->error); + return; + } + + /* We should meanwhile have enough trace. */ + conf->error = conf->suspend(conf->tracer); + if (conf->error < 0) + return; + + /* Let's see if we can access the trace. */ + trace = ds_read_bts(conf->tracer); + + conf->error = ds_selftest_bts_consistency(trace); + if (conf->error < 0) + return; + + /* If everything went well, we should have a few trace entries. */ + if (trace->ds.top == trace->ds.begin) { + /* + * It is possible but highly unlikely that we got a + * buffer overflow and end up at exactly the same + * position we started from. + * Let's issue a warning, but continue. + */ + printk(KERN_CONT "no trace/overflow..."); + } + + /* Let's try to read the trace we collected. */ + conf->error = + ds_selftest_bts_read(conf->tracer, trace, + trace->ds.begin, trace->ds.top); + if (conf->error < 0) + return; + + /* + * Let's read the trace again. + * Since we suspended tracing, we should get the same result. + */ + top = trace->ds.top; + + trace = ds_read_bts(conf->tracer); + conf->error = ds_selftest_bts_consistency(trace); + if (conf->error < 0) + return; + + if (top != trace->ds.top) { + printk(KERN_CONT "suspend not working..."); + conf->error = -1; + return; + } + + /* Let's collect some more trace - see if resume is working. */ + conf->error = conf->resume(conf->tracer); + if (conf->error < 0) + return; + + conf->error = conf->suspend(conf->tracer); + if (conf->error < 0) + return; + + trace = ds_read_bts(conf->tracer); + + conf->error = ds_selftest_bts_consistency(trace); + if (conf->error < 0) + return; + + if (trace->ds.top == top) { + /* + * It is possible but highly unlikely that we got a + * buffer overflow and end up at exactly the same + * position we started from. + * Let's issue a warning and check the full trace. + */ + printk(KERN_CONT + "no resume progress/overflow..."); + + conf->error = + ds_selftest_bts_read(conf->tracer, trace, + trace->ds.begin, trace->ds.end); + } else if (trace->ds.top < top) { + /* + * We had a buffer overflow - the entire buffer should + * contain trace records. + */ + conf->error = + ds_selftest_bts_read(conf->tracer, trace, + trace->ds.begin, trace->ds.end); + } else { + /* + * It is quite likely that the buffer did not overflow. + * Let's just check the delta trace. + */ + conf->error = + ds_selftest_bts_read(conf->tracer, trace, top, + trace->ds.top); + } + if (conf->error < 0) + return; + + conf->error = 0; +} + +static int ds_suspend_bts_wrap(struct bts_tracer *tracer) +{ + ds_suspend_bts(tracer); + return 0; +} + +static int ds_resume_bts_wrap(struct bts_tracer *tracer) +{ + ds_resume_bts(tracer); + return 0; +} + +static void ds_release_bts_noirq_wrap(void *tracer) +{ + (void)ds_release_bts_noirq(tracer); +} + +static int ds_selftest_bts_bad_release_noirq(int cpu, + struct bts_tracer *tracer) +{ + int error = -EPERM; + + /* Try to release the tracer on the wrong cpu. */ + get_cpu(); + if (cpu != smp_processor_id()) { + error = ds_release_bts_noirq(tracer); + if (error != -EPERM) + printk(KERN_CONT "release on wrong cpu..."); + } + put_cpu(); + + return error ? 0 : -1; +} + +static int ds_selftest_bts_bad_request_cpu(int cpu, void *buffer) +{ + struct bts_tracer *tracer; + int error; + + /* Try to request cpu tracing while task tracing is active. */ + tracer = ds_request_bts_cpu(cpu, buffer, BUFFER_SIZE, NULL, + (size_t)-1, BTS_KERNEL); + error = PTR_ERR(tracer); + if (!IS_ERR(tracer)) { + ds_release_bts(tracer); + error = 0; + } + + if (error != -EPERM) + printk(KERN_CONT "cpu/task tracing overlap..."); + + return error ? 0 : -1; +} + +static int ds_selftest_bts_bad_request_task(void *buffer) +{ + struct bts_tracer *tracer; + int error; + + /* Try to request cpu tracing while task tracing is active. */ + tracer = ds_request_bts_task(current, buffer, BUFFER_SIZE, NULL, + (size_t)-1, BTS_KERNEL); + error = PTR_ERR(tracer); + if (!IS_ERR(tracer)) { + error = 0; + ds_release_bts(tracer); + } + + if (error != -EPERM) + printk(KERN_CONT "task/cpu tracing overlap..."); + + return error ? 0 : -1; +} + +int ds_selftest_bts(void) +{ + struct ds_selftest_bts_conf conf; + unsigned char buffer[BUFFER_SIZE], *small_buffer; + unsigned long irq; + int cpu; + + printk(KERN_INFO "[ds] bts selftest..."); + conf.error = 0; + + small_buffer = (unsigned char *)ALIGN((unsigned long)buffer, 8) + 8; + + get_online_cpus(); + for_each_online_cpu(cpu) { + conf.suspend = ds_suspend_bts_wrap; + conf.resume = ds_resume_bts_wrap; + conf.tracer = + ds_request_bts_cpu(cpu, buffer, BUFFER_SIZE, + NULL, (size_t)-1, BTS_KERNEL); + ds_selftest_bts_cpu(&conf); + if (conf.error >= 0) + conf.error = ds_selftest_bts_bad_request_task(buffer); + ds_release_bts(conf.tracer); + if (conf.error < 0) + goto out; + + conf.suspend = ds_suspend_bts_noirq; + conf.resume = ds_resume_bts_noirq; + conf.tracer = + ds_request_bts_cpu(cpu, buffer, BUFFER_SIZE, + NULL, (size_t)-1, BTS_KERNEL); + smp_call_function_single(cpu, ds_selftest_bts_cpu, &conf, 1); + if (conf.error >= 0) { + conf.error = + ds_selftest_bts_bad_release_noirq(cpu, + conf.tracer); + /* We must not release the tracer twice. */ + if (conf.error < 0) + conf.tracer = NULL; + } + if (conf.error >= 0) + conf.error = ds_selftest_bts_bad_request_task(buffer); + smp_call_function_single(cpu, ds_release_bts_noirq_wrap, + conf.tracer, 1); + if (conf.error < 0) + goto out; + } + + conf.suspend = ds_suspend_bts_wrap; + conf.resume = ds_resume_bts_wrap; + conf.tracer = + ds_request_bts_task(current, buffer, BUFFER_SIZE, + NULL, (size_t)-1, BTS_KERNEL); + ds_selftest_bts_cpu(&conf); + if (conf.error >= 0) + conf.error = ds_selftest_bts_bad_request_cpu(0, buffer); + ds_release_bts(conf.tracer); + if (conf.error < 0) + goto out; + + conf.suspend = ds_suspend_bts_noirq; + conf.resume = ds_resume_bts_noirq; + conf.tracer = + ds_request_bts_task(current, small_buffer, SMALL_BUFFER_SIZE, + NULL, (size_t)-1, BTS_KERNEL); + local_irq_save(irq); + ds_selftest_bts_cpu(&conf); + if (conf.error >= 0) + conf.error = ds_selftest_bts_bad_request_cpu(0, buffer); + ds_release_bts_noirq(conf.tracer); + local_irq_restore(irq); + if (conf.error < 0) + goto out; + + conf.error = 0; + out: + put_online_cpus(); + printk(KERN_CONT "%s.\n", (conf.error ? "failed" : "passed")); + + return conf.error; +} + +int ds_selftest_pebs(void) +{ + return 0; +} diff --git a/arch/x86/kernel/ds_selftest.h b/arch/x86/kernel/ds_selftest.h new file mode 100644 index 00000000000..2ba8745c666 --- /dev/null +++ b/arch/x86/kernel/ds_selftest.h @@ -0,0 +1,15 @@ +/* + * Debug Store support - selftest + * + * + * Copyright (C) 2009 Intel Corporation. + * Markus Metzger <markus.t.metzger@intel.com>, 2009 + */ + +#ifdef CONFIG_X86_DS_SELFTEST +extern int ds_selftest_bts(void); +extern int ds_selftest_pebs(void); +#else +static inline int ds_selftest_bts(void) { return 0; } +static inline int ds_selftest_pebs(void) { return 0; } +#endif diff --git a/arch/x86/kernel/dumpstack.h b/arch/x86/kernel/dumpstack.h index da87590b869..81086c227ab 100644 --- a/arch/x86/kernel/dumpstack.h +++ b/arch/x86/kernel/dumpstack.h @@ -29,7 +29,6 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, unsigned long *sp, unsigned long bp, char *log_lvl); extern unsigned int code_bytes; -extern int kstack_depth_to_print; /* The form of the top of the frame on the stack */ struct stack_frame { diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 00628130292..7271fa33d79 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -617,7 +617,7 @@ __init int e820_search_gap(unsigned long *gapstart, unsigned long *gapsize, */ __init void e820_setup_gap(void) { - unsigned long gapstart, gapsize, round; + unsigned long gapstart, gapsize; int found; gapstart = 0x10000000; @@ -635,14 +635,9 @@ __init void e820_setup_gap(void) #endif /* - * See how much we want to round up: start off with - * rounding to the next 1MB area. + * e820_reserve_resources_late protect stolen RAM already */ - round = 0x100000; - while ((gapsize >> 4) > round) - round += round; - /* Fun with two's complement */ - pci_mem_start = (gapstart + round) & -round; + pci_mem_start = gapstart; printk(KERN_INFO "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n", @@ -1371,6 +1366,23 @@ void __init e820_reserve_resources(void) } } +/* How much should we pad RAM ending depending on where it is? */ +static unsigned long ram_alignment(resource_size_t pos) +{ + unsigned long mb = pos >> 20; + + /* To 64kB in the first megabyte */ + if (!mb) + return 64*1024; + + /* To 1MB in the first 16MB */ + if (mb < 16) + return 1024*1024; + + /* To 32MB for anything above that */ + return 32*1024*1024; +} + void __init e820_reserve_resources_late(void) { int i; @@ -1382,6 +1394,24 @@ void __init e820_reserve_resources_late(void) insert_resource_expand_to_fit(&iomem_resource, res); res++; } + + /* + * Try to bump up RAM regions to reasonable boundaries to + * avoid stolen RAM: + */ + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *entry = &e820_saved.map[i]; + resource_size_t start, end; + + if (entry->type != E820_RAM) + continue; + start = entry->addr + entry->size; + end = round_up(start, ram_alignment(start)); + if (start == end) + continue; + reserve_region_with_split(&iomem_resource, start, + end - 1, "RAM buffer"); + } } char *__init default_machine_specific_memory_setup(void) diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c index 76b8cd953de..ebdb85cf268 100644 --- a/arch/x86/kernel/early-quirks.c +++ b/arch/x86/kernel/early-quirks.c @@ -97,6 +97,7 @@ static void __init nvidia_bugs(int num, int slot, int func) } #if defined(CONFIG_ACPI) && defined(CONFIG_X86_IO_APIC) +#if defined(CONFIG_ACPI) && defined(CONFIG_X86_IO_APIC) static u32 __init ati_ixp4x0_rev(int num, int slot, int func) { u32 d; @@ -114,6 +115,7 @@ static u32 __init ati_ixp4x0_rev(int num, int slot, int func) d &= 0xff; return d; } +#endif static void __init ati_bugs(int num, int slot, int func) { diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 38946c6e843..a4742a340d8 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -147,27 +147,14 @@ END(ftrace_graph_caller) GLOBAL(return_to_handler) subq $80, %rsp + /* Save the return values */ movq %rax, (%rsp) - movq %rcx, 8(%rsp) - movq %rdx, 16(%rsp) - movq %rsi, 24(%rsp) - movq %rdi, 32(%rsp) - movq %r8, 40(%rsp) - movq %r9, 48(%rsp) - movq %r10, 56(%rsp) - movq %r11, 64(%rsp) + movq %rdx, 8(%rsp) call ftrace_return_to_handler movq %rax, 72(%rsp) - movq 64(%rsp), %r11 - movq 56(%rsp), %r10 - movq 48(%rsp), %r9 - movq 40(%rsp), %r8 - movq 32(%rsp), %rdi - movq 24(%rsp), %rsi - movq 16(%rsp), %rdx - movq 8(%rsp), %rcx + movq 8(%rsp), %rdx movq (%rsp), %rax addq $72, %rsp retq @@ -1025,6 +1012,11 @@ apicinterrupt ERROR_APIC_VECTOR \ apicinterrupt SPURIOUS_APIC_VECTOR \ spurious_interrupt smp_spurious_interrupt +#ifdef CONFIG_PERF_COUNTERS +apicinterrupt LOCAL_PENDING_VECTOR \ + perf_pending_interrupt smp_perf_pending_interrupt +#endif + /* * Exception entry points. */ @@ -1379,6 +1371,11 @@ END(xen_failsafe_callback) paranoidzeroentry_ist debug do_debug DEBUG_STACK paranoidzeroentry_ist int3 do_int3 DEBUG_STACK paranoiderrorentry stack_segment do_stack_segment +#ifdef CONFIG_XEN +zeroentry xen_debug do_debug +zeroentry xen_int3 do_int3 +errorentry xen_stack_segment do_stack_segment +#endif errorentry general_protection do_general_protection errorentry page_fault do_page_fault #ifdef CONFIG_X86_MCE diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index 30683883e0c..dc5ed4bdd88 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -608,13 +608,6 @@ ignore_int: ENTRY(initial_code) .long i386_start_kernel -.section .text -/* - * Real beginning of normal "text" segment - */ -ENTRY(stext) -ENTRY(_stext) - /* * BSS section */ diff --git a/arch/x86/kernel/irq.c b/arch/x86/kernel/irq.c index c3fe010d74c..38287b5f116 100644 --- a/arch/x86/kernel/irq.c +++ b/arch/x86/kernel/irq.c @@ -12,6 +12,7 @@ #include <asm/io_apic.h> #include <asm/irq.h> #include <asm/idle.h> +#include <asm/hw_irq.h> atomic_t irq_err_count; @@ -24,9 +25,9 @@ void (*generic_interrupt_extension)(void) = NULL; */ void ack_bad_irq(unsigned int irq) { - printk(KERN_ERR "unexpected IRQ trap at vector %02x\n", irq); + if (printk_ratelimit()) + pr_err("unexpected IRQ trap at vector %02x\n", irq); -#ifdef CONFIG_X86_LOCAL_APIC /* * Currently unexpected vectors happen only on SMP and APIC. * We _must_ ack these because every local APIC has only N @@ -36,9 +37,7 @@ void ack_bad_irq(unsigned int irq) * completely. * But only ack when the APIC is enabled -AK */ - if (cpu_has_apic) - ack_APIC_irq(); -#endif + ack_APIC_irq(); } #define irq_stats(x) (&per_cpu(irq_stat, x)) @@ -63,6 +62,14 @@ static int show_other_interrupts(struct seq_file *p, int prec) for_each_online_cpu(j) seq_printf(p, "%10u ", irq_stats(j)->irq_spurious_count); seq_printf(p, " Spurious interrupts\n"); + seq_printf(p, "%*s: ", prec, "CNT"); + for_each_online_cpu(j) + seq_printf(p, "%10u ", irq_stats(j)->apic_perf_irqs); + seq_printf(p, " Performance counter interrupts\n"); + seq_printf(p, "%*s: ", prec, "PND"); + for_each_online_cpu(j) + seq_printf(p, "%10u ", irq_stats(j)->apic_pending_irqs); + seq_printf(p, " Performance pending work\n"); #endif if (generic_interrupt_extension) { seq_printf(p, "%*s: ", prec, "PLT"); @@ -166,6 +173,8 @@ u64 arch_irq_stat_cpu(unsigned int cpu) #ifdef CONFIG_X86_LOCAL_APIC sum += irq_stats(cpu)->apic_timer_irqs; sum += irq_stats(cpu)->irq_spurious_count; + sum += irq_stats(cpu)->apic_perf_irqs; + sum += irq_stats(cpu)->apic_pending_irqs; #endif if (generic_interrupt_extension) sum += irq_stats(cpu)->generic_irqs; @@ -178,7 +187,7 @@ u64 arch_irq_stat_cpu(unsigned int cpu) sum += irq_stats(cpu)->irq_thermal_count; # ifdef CONFIG_X86_64 sum += irq_stats(cpu)->irq_threshold_count; -#endif +# endif #endif return sum; } @@ -213,14 +222,11 @@ unsigned int __irq_entry do_IRQ(struct pt_regs *regs) irq = __get_cpu_var(vector_irq)[vector]; if (!handle_irq(irq, regs)) { -#ifdef CONFIG_X86_64 - if (!disable_apic) - ack_APIC_irq(); -#endif + ack_APIC_irq(); if (printk_ratelimit()) - printk(KERN_EMERG "%s: %d.%d No irq handler for vector (irq %d)\n", - __func__, smp_processor_id(), vector, irq); + pr_emerg("%s: %d.%d No irq handler for vector (irq %d)\n", + __func__, smp_processor_id(), vector, irq); } irq_exit(); diff --git a/arch/x86/kernel/irqinit_32.c b/arch/x86/kernel/irqinit.c index 368b0a8836f..267c6624c77 100644 --- a/arch/x86/kernel/irqinit_32.c +++ b/arch/x86/kernel/irqinit.c @@ -1,20 +1,25 @@ +#include <linux/linkage.h> #include <linux/errno.h> #include <linux/signal.h> #include <linux/sched.h> #include <linux/ioport.h> #include <linux/interrupt.h> +#include <linux/timex.h> #include <linux/slab.h> #include <linux/random.h> +#include <linux/kprobes.h> #include <linux/init.h> #include <linux/kernel_stat.h> #include <linux/sysdev.h> #include <linux/bitops.h> +#include <linux/acpi.h> #include <linux/io.h> #include <linux/delay.h> #include <asm/atomic.h> #include <asm/system.h> #include <asm/timer.h> +#include <asm/hw_irq.h> #include <asm/pgtable.h> #include <asm/desc.h> #include <asm/apic.h> @@ -22,7 +27,23 @@ #include <asm/i8259.h> #include <asm/traps.h> +/* + * ISA PIC or low IO-APIC triggered (INTA-cycle or APIC) interrupts: + * (these are usually mapped to vectors 0x30-0x3f) + */ + +/* + * The IO-APIC gives us many more interrupt sources. Most of these + * are unused but an SMP system is supposed to have enough memory ... + * sometimes (mostly wrt. hw bugs) we get corrupted vectors all + * across the spectrum, so we really want to be prepared to get all + * of these. Plus, more powerful systems might have more than 64 + * IO-APIC registers. + * + * (these are usually mapped into the 0x30-0xff vector range) + */ +#ifdef CONFIG_X86_32 /* * Note that on a 486, we don't want to do a SIGFPE on an irq13 * as the irq is unreliable, and exception 16 works correctly @@ -52,30 +73,7 @@ static struct irqaction fpu_irq = { .handler = math_error_irq, .name = "fpu", }; - -void __init init_ISA_irqs(void) -{ - int i; - -#ifdef CONFIG_X86_LOCAL_APIC - init_bsp_APIC(); #endif - init_8259A(0); - - /* - * 16 old-style INTA-cycle interrupts: - */ - for (i = 0; i < NR_IRQS_LEGACY; i++) { - struct irq_desc *desc = irq_to_desc(i); - - desc->status = IRQ_DISABLED; - desc->action = NULL; - desc->depth = 1; - - set_irq_chip_and_handler_name(i, &i8259A_chip, - handle_level_irq, "XT"); - } -} /* * IRQ2 is cascade interrupt to second interrupt controller @@ -118,29 +116,37 @@ int vector_used_by_percpu_irq(unsigned int vector) return 0; } -/* Overridden in paravirt.c */ -void init_IRQ(void) __attribute__((weak, alias("native_init_IRQ"))); - -void __init native_init_IRQ(void) +static void __init init_ISA_irqs(void) { int i; - /* Execute any quirks before the call gates are initialised: */ - x86_quirk_pre_intr_init(); +#if defined(CONFIG_X86_64) || defined(CONFIG_X86_LOCAL_APIC) + init_bsp_APIC(); +#endif + init_8259A(0); /* - * Cover the whole vector space, no vector can escape - * us. (some of these will be overridden and become - * 'special' SMP interrupts) + * 16 old-style INTA-cycle interrupts: */ - for (i = FIRST_EXTERNAL_VECTOR; i < NR_VECTORS; i++) { - /* SYSCALL_VECTOR was reserved in trap_init. */ - if (i != SYSCALL_VECTOR) - set_intr_gate(i, interrupt[i-FIRST_EXTERNAL_VECTOR]); + for (i = 0; i < NR_IRQS_LEGACY; i++) { + struct irq_desc *desc = irq_to_desc(i); + + desc->status = IRQ_DISABLED; + desc->action = NULL; + desc->depth = 1; + + set_irq_chip_and_handler_name(i, &i8259A_chip, + handle_level_irq, "XT"); } +} +/* Overridden in paravirt.c */ +void init_IRQ(void) __attribute__((weak, alias("native_init_IRQ"))); -#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_SMP) +static void __init smp_intr_init(void) +{ +#ifdef CONFIG_SMP +#if defined(CONFIG_X86_64) || defined(CONFIG_X86_LOCAL_APIC) /* * The reschedule interrupt is a CPU-to-CPU reschedule-helper * IPI, driven by wakeup. @@ -160,16 +166,32 @@ void __init native_init_IRQ(void) /* IPI for generic function call */ alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); - /* IPI for single call function */ + /* IPI for generic single function call */ alloc_intr_gate(CALL_FUNCTION_SINGLE_VECTOR, - call_function_single_interrupt); + call_function_single_interrupt); /* Low priority IPI to cleanup after moving an irq */ set_intr_gate(IRQ_MOVE_CLEANUP_VECTOR, irq_move_cleanup_interrupt); set_bit(IRQ_MOVE_CLEANUP_VECTOR, used_vectors); #endif +#endif /* CONFIG_SMP */ +} + +static void __init apic_intr_init(void) +{ + smp_intr_init(); + +#ifdef CONFIG_X86_THERMAL_VECTOR + alloc_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt); +#endif +#ifdef CONFIG_X86_THRESHOLD + alloc_intr_gate(THRESHOLD_APIC_VECTOR, threshold_interrupt); +#endif +#if defined(CONFIG_X86_NEW_MCE) && defined(CONFIG_X86_LOCAL_APIC) + alloc_intr_gate(MCE_SELF_VECTOR, mce_self_interrupt); +#endif -#ifdef CONFIG_X86_LOCAL_APIC +#if defined(CONFIG_X86_64) || defined(CONFIG_X86_LOCAL_APIC) /* self generated IPI for local APIC timer */ alloc_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt); @@ -179,16 +201,59 @@ void __init native_init_IRQ(void) /* IPI vectors for APIC spurious and error interrupts */ alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); alloc_intr_gate(ERROR_APIC_VECTOR, error_interrupt); + + /* Performance monitoring interrupts: */ +# ifdef CONFIG_PERF_COUNTERS + alloc_intr_gate(LOCAL_PENDING_VECTOR, perf_pending_interrupt); +# endif + #endif +} -#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86_MCE_P4THERMAL) - /* thermal monitor LVT interrupt */ - alloc_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt); +/** + * x86_quirk_pre_intr_init - initialisation prior to setting up interrupt vectors + * + * Description: + * Perform any necessary interrupt initialisation prior to setting up + * the "ordinary" interrupt call gates. For legacy reasons, the ISA + * interrupts should be initialised here if the machine emulates a PC + * in any way. + **/ +static void __init x86_quirk_pre_intr_init(void) +{ +#ifdef CONFIG_X86_32 + if (x86_quirks->arch_pre_intr_init) { + if (x86_quirks->arch_pre_intr_init()) + return; + } #endif + init_ISA_irqs(); +} + +void __init native_init_IRQ(void) +{ + int i; + + /* Execute any quirks before the call gates are initialised: */ + x86_quirk_pre_intr_init(); + + apic_intr_init(); + + /* + * Cover the whole vector space, no vector can escape + * us. (some of these will be overridden and become + * 'special' SMP interrupts) + */ + for (i = FIRST_EXTERNAL_VECTOR; i < NR_VECTORS; i++) { + /* IA32_SYSCALL_VECTOR could be used in trap_init already. */ + if (!test_bit(i, used_vectors)) + set_intr_gate(i, interrupt[i-FIRST_EXTERNAL_VECTOR]); + } if (!acpi_ioapic) setup_irq(2, &irq2); +#ifdef CONFIG_X86_32 /* * Call quirks after call gates are initialised (usually add in * the architecture specific gates): @@ -203,4 +268,5 @@ void __init native_init_IRQ(void) setup_irq(FPU_IRQ, &fpu_irq); irq_ctx_init(smp_processor_id()); +#endif } diff --git a/arch/x86/kernel/irqinit_64.c b/arch/x86/kernel/irqinit_64.c deleted file mode 100644 index 8cd10537fd4..00000000000 --- a/arch/x86/kernel/irqinit_64.c +++ /dev/null @@ -1,177 +0,0 @@ -#include <linux/linkage.h> -#include <linux/errno.h> -#include <linux/signal.h> -#include <linux/sched.h> -#include <linux/ioport.h> -#include <linux/interrupt.h> -#include <linux/timex.h> -#include <linux/slab.h> -#include <linux/random.h> -#include <linux/init.h> -#include <linux/kernel_stat.h> -#include <linux/sysdev.h> -#include <linux/bitops.h> -#include <linux/acpi.h> -#include <linux/io.h> -#include <linux/delay.h> - -#include <asm/atomic.h> -#include <asm/system.h> -#include <asm/hw_irq.h> -#include <asm/pgtable.h> -#include <asm/desc.h> -#include <asm/apic.h> -#include <asm/i8259.h> - -/* - * ISA PIC or low IO-APIC triggered (INTA-cycle or APIC) interrupts: - * (these are usually mapped to vectors 0x30-0x3f) - */ - -/* - * The IO-APIC gives us many more interrupt sources. Most of these - * are unused but an SMP system is supposed to have enough memory ... - * sometimes (mostly wrt. hw bugs) we get corrupted vectors all - * across the spectrum, so we really want to be prepared to get all - * of these. Plus, more powerful systems might have more than 64 - * IO-APIC registers. - * - * (these are usually mapped into the 0x30-0xff vector range) - */ - -/* - * IRQ2 is cascade interrupt to second interrupt controller - */ - -static struct irqaction irq2 = { - .handler = no_action, - .name = "cascade", -}; -DEFINE_PER_CPU(vector_irq_t, vector_irq) = { - [0 ... IRQ0_VECTOR - 1] = -1, - [IRQ0_VECTOR] = 0, - [IRQ1_VECTOR] = 1, - [IRQ2_VECTOR] = 2, - [IRQ3_VECTOR] = 3, - [IRQ4_VECTOR] = 4, - [IRQ5_VECTOR] = 5, - [IRQ6_VECTOR] = 6, - [IRQ7_VECTOR] = 7, - [IRQ8_VECTOR] = 8, - [IRQ9_VECTOR] = 9, - [IRQ10_VECTOR] = 10, - [IRQ11_VECTOR] = 11, - [IRQ12_VECTOR] = 12, - [IRQ13_VECTOR] = 13, - [IRQ14_VECTOR] = 14, - [IRQ15_VECTOR] = 15, - [IRQ15_VECTOR + 1 ... NR_VECTORS - 1] = -1 -}; - -int vector_used_by_percpu_irq(unsigned int vector) -{ - int cpu; - - for_each_online_cpu(cpu) { - if (per_cpu(vector_irq, cpu)[vector] != -1) - return 1; - } - - return 0; -} - -static void __init init_ISA_irqs(void) -{ - int i; - - init_bsp_APIC(); - init_8259A(0); - - for (i = 0; i < NR_IRQS_LEGACY; i++) { - struct irq_desc *desc = irq_to_desc(i); - - desc->status = IRQ_DISABLED; - desc->action = NULL; - desc->depth = 1; - - /* - * 16 old-style INTA-cycle interrupts: - */ - set_irq_chip_and_handler_name(i, &i8259A_chip, - handle_level_irq, "XT"); - } -} - -void init_IRQ(void) __attribute__((weak, alias("native_init_IRQ"))); - -static void __init smp_intr_init(void) -{ -#ifdef CONFIG_SMP - /* - * The reschedule interrupt is a CPU-to-CPU reschedule-helper - * IPI, driven by wakeup. - */ - alloc_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); - - /* IPIs for invalidation */ - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+0, invalidate_interrupt0); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+1, invalidate_interrupt1); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+2, invalidate_interrupt2); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+3, invalidate_interrupt3); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+4, invalidate_interrupt4); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+5, invalidate_interrupt5); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+6, invalidate_interrupt6); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+7, invalidate_interrupt7); - - /* IPI for generic function call */ - alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); - - /* IPI for generic single function call */ - alloc_intr_gate(CALL_FUNCTION_SINGLE_VECTOR, - call_function_single_interrupt); - - /* Low priority IPI to cleanup after moving an irq */ - set_intr_gate(IRQ_MOVE_CLEANUP_VECTOR, irq_move_cleanup_interrupt); - set_bit(IRQ_MOVE_CLEANUP_VECTOR, used_vectors); -#endif -} - -static void __init apic_intr_init(void) -{ - smp_intr_init(); - - alloc_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt); - alloc_intr_gate(THRESHOLD_APIC_VECTOR, threshold_interrupt); - - /* self generated IPI for local APIC timer */ - alloc_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt); - - /* generic IPI for platform specific use */ - alloc_intr_gate(GENERIC_INTERRUPT_VECTOR, generic_interrupt); - - /* IPI vectors for APIC spurious and error interrupts */ - alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); - alloc_intr_gate(ERROR_APIC_VECTOR, error_interrupt); -} - -void __init native_init_IRQ(void) -{ - int i; - - init_ISA_irqs(); - /* - * Cover the whole vector space, no vector can escape - * us. (some of these will be overridden and become - * 'special' SMP interrupts) - */ - for (i = 0; i < (NR_VECTORS - FIRST_EXTERNAL_VECTOR); i++) { - int vector = FIRST_EXTERNAL_VECTOR + i; - if (vector != IA32_SYSCALL_VECTOR) - set_intr_gate(vector, interrupt[i]); - } - - apic_intr_init(); - - if (!acpi_ioapic) - setup_irq(2, &irq2); -} diff --git a/arch/x86/kernel/kgdb.c b/arch/x86/kernel/kgdb.c index b1f4dffb919..8d82a77a3f3 100644 --- a/arch/x86/kernel/kgdb.c +++ b/arch/x86/kernel/kgdb.c @@ -142,7 +142,7 @@ void sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p) gdb_regs32[GDB_PS] = *(unsigned long *)(p->thread.sp + 8); gdb_regs32[GDB_CS] = __KERNEL_CS; gdb_regs32[GDB_SS] = __KERNEL_DS; - gdb_regs[GDB_PC] = p->thread.ip; + gdb_regs[GDB_PC] = 0; gdb_regs[GDB_R8] = 0; gdb_regs[GDB_R9] = 0; gdb_regs[GDB_R10] = 0; diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c index 33019ddb56b..6551dedee20 100644 --- a/arch/x86/kernel/kvm.c +++ b/arch/x86/kernel/kvm.c @@ -195,7 +195,7 @@ static void kvm_leave_lazy_mmu(void) struct kvm_para_state *state = kvm_para_state(); mmu_queue_flush(state); - paravirt_leave_lazy(paravirt_get_lazy_mode()); + paravirt_leave_lazy_mmu(); state->mode = paravirt_get_lazy_mode(); } diff --git a/arch/x86/kernel/microcode_amd.c b/arch/x86/kernel/microcode_amd.c index 453b5795a5c..366baa17991 100644 --- a/arch/x86/kernel/microcode_amd.c +++ b/arch/x86/kernel/microcode_amd.c @@ -13,25 +13,13 @@ * Licensed under the terms of the GNU General Public * License version 2. See file COPYING for details. */ -#include <linux/platform_device.h> -#include <linux/capability.h> -#include <linux/miscdevice.h> #include <linux/firmware.h> -#include <linux/spinlock.h> -#include <linux/cpumask.h> #include <linux/pci_ids.h> #include <linux/uaccess.h> #include <linux/vmalloc.h> #include <linux/kernel.h> #include <linux/module.h> -#include <linux/mutex.h> -#include <linux/sched.h> -#include <linux/init.h> -#include <linux/slab.h> -#include <linux/cpu.h> #include <linux/pci.h> -#include <linux/fs.h> -#include <linux/mm.h> #include <asm/microcode.h> #include <asm/processor.h> @@ -79,9 +67,6 @@ struct microcode_amd { #define UCODE_CONTAINER_SECTION_HDR 8 #define UCODE_CONTAINER_HEADER_SIZE 12 -/* serialize access to the physical write */ -static DEFINE_SPINLOCK(microcode_update_lock); - static struct equiv_cpu_entry *equiv_cpu_table; static int collect_cpu_info_amd(int cpu, struct cpu_signature *csig) @@ -144,9 +129,8 @@ static int get_matching_microcode(int cpu, void *mc, int rev) return 1; } -static void apply_microcode_amd(int cpu) +static int apply_microcode_amd(int cpu) { - unsigned long flags; u32 rev, dummy; int cpu_num = raw_smp_processor_id(); struct ucode_cpu_info *uci = ucode_cpu_info + cpu_num; @@ -156,25 +140,25 @@ static void apply_microcode_amd(int cpu) BUG_ON(cpu_num != cpu); if (mc_amd == NULL) - return; + return 0; - spin_lock_irqsave(µcode_update_lock, flags); wrmsrl(MSR_AMD64_PATCH_LOADER, (u64)(long)&mc_amd->hdr.data_code); /* get patch id after patching */ rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); - spin_unlock_irqrestore(µcode_update_lock, flags); /* check current patch id and patch's id for match */ if (rev != mc_amd->hdr.patch_id) { printk(KERN_ERR "microcode: CPU%d: update failed " "(for patch_level=0x%x)\n", cpu, mc_amd->hdr.patch_id); - return; + return -1; } printk(KERN_INFO "microcode: CPU%d: updated (new patch_level=0x%x)\n", cpu, rev); uci->cpu_sig.rev = rev; + + return 0; } static int get_ucode_data(void *to, const u8 *from, size_t n) @@ -257,13 +241,12 @@ static int install_equiv_cpu_table(const u8 *buf) static void free_equiv_cpu_table(void) { - if (equiv_cpu_table) { - vfree(equiv_cpu_table); - equiv_cpu_table = NULL; - } + vfree(equiv_cpu_table); + equiv_cpu_table = NULL; } -static int generic_load_microcode(int cpu, const u8 *data, size_t size) +static enum ucode_state +generic_load_microcode(int cpu, const u8 *data, size_t size) { struct ucode_cpu_info *uci = ucode_cpu_info + cpu; const u8 *ucode_ptr = data; @@ -272,12 +255,13 @@ static int generic_load_microcode(int cpu, const u8 *data, size_t size) int new_rev = uci->cpu_sig.rev; unsigned int leftover; unsigned long offset; + enum ucode_state state = UCODE_OK; offset = install_equiv_cpu_table(ucode_ptr); if (!offset) { printk(KERN_ERR "microcode: failed to create " "equivalent cpu table\n"); - return -EINVAL; + return UCODE_ERROR; } ucode_ptr += offset; @@ -293,8 +277,7 @@ static int generic_load_microcode(int cpu, const u8 *data, size_t size) mc_header = (struct microcode_header_amd *)mc; if (get_matching_microcode(cpu, mc, new_rev)) { - if (new_mc) - vfree(new_mc); + vfree(new_mc); new_rev = mc_header->patch_id; new_mc = mc; } else @@ -306,34 +289,32 @@ static int generic_load_microcode(int cpu, const u8 *data, size_t size) if (new_mc) { if (!leftover) { - if (uci->mc) - vfree(uci->mc); + vfree(uci->mc); uci->mc = new_mc; pr_debug("microcode: CPU%d found a matching microcode " "update with version 0x%x (current=0x%x)\n", cpu, new_rev, uci->cpu_sig.rev); - } else + } else { vfree(new_mc); - } + state = UCODE_ERROR; + } + } else + state = UCODE_NFOUND; free_equiv_cpu_table(); - return (int)leftover; + return state; } -static int request_microcode_fw(int cpu, struct device *device) +static enum ucode_state request_microcode_fw(int cpu, struct device *device) { const char *fw_name = "amd-ucode/microcode_amd.bin"; const struct firmware *firmware; - int ret; - - /* We should bind the task to the CPU */ - BUG_ON(cpu != raw_smp_processor_id()); + enum ucode_state ret; - ret = request_firmware(&firmware, fw_name, device); - if (ret) { + if (request_firmware(&firmware, fw_name, device)) { printk(KERN_ERR "microcode: failed to load file %s\n", fw_name); - return ret; + return UCODE_NFOUND; } ret = generic_load_microcode(cpu, firmware->data, firmware->size); @@ -343,11 +324,12 @@ static int request_microcode_fw(int cpu, struct device *device) return ret; } -static int request_microcode_user(int cpu, const void __user *buf, size_t size) +static enum ucode_state +request_microcode_user(int cpu, const void __user *buf, size_t size) { printk(KERN_INFO "microcode: AMD microcode update via " "/dev/cpu/microcode not supported\n"); - return -1; + return UCODE_ERROR; } static void microcode_fini_cpu_amd(int cpu) diff --git a/arch/x86/kernel/microcode_core.c b/arch/x86/kernel/microcode_core.c index 98c470c069d..9c4461501fc 100644 --- a/arch/x86/kernel/microcode_core.c +++ b/arch/x86/kernel/microcode_core.c @@ -71,27 +71,18 @@ * Thanks to Stuart Swales for pointing out this bug. */ #include <linux/platform_device.h> -#include <linux/capability.h> #include <linux/miscdevice.h> -#include <linux/firmware.h> +#include <linux/capability.h> #include <linux/smp_lock.h> -#include <linux/spinlock.h> -#include <linux/cpumask.h> -#include <linux/uaccess.h> -#include <linux/vmalloc.h> #include <linux/kernel.h> #include <linux/module.h> #include <linux/mutex.h> -#include <linux/sched.h> -#include <linux/init.h> -#include <linux/slab.h> #include <linux/cpu.h> #include <linux/fs.h> #include <linux/mm.h> #include <asm/microcode.h> #include <asm/processor.h> -#include <asm/msr.h> MODULE_DESCRIPTION("Microcode Update Driver"); MODULE_AUTHOR("Tigran Aivazian <tigran@aivazian.fsnet.co.uk>"); @@ -101,36 +92,110 @@ MODULE_LICENSE("GPL"); static struct microcode_ops *microcode_ops; -/* no concurrent ->write()s are allowed on /dev/cpu/microcode */ +/* + * Synchronization. + * + * All non cpu-hotplug-callback call sites use: + * + * - microcode_mutex to synchronize with each other; + * - get/put_online_cpus() to synchronize with + * the cpu-hotplug-callback call sites. + * + * We guarantee that only a single cpu is being + * updated at any particular moment of time. + */ static DEFINE_MUTEX(microcode_mutex); struct ucode_cpu_info ucode_cpu_info[NR_CPUS]; EXPORT_SYMBOL_GPL(ucode_cpu_info); +/* + * Operations that are run on a target cpu: + */ + +struct cpu_info_ctx { + struct cpu_signature *cpu_sig; + int err; +}; + +static void collect_cpu_info_local(void *arg) +{ + struct cpu_info_ctx *ctx = arg; + + ctx->err = microcode_ops->collect_cpu_info(smp_processor_id(), + ctx->cpu_sig); +} + +static int collect_cpu_info_on_target(int cpu, struct cpu_signature *cpu_sig) +{ + struct cpu_info_ctx ctx = { .cpu_sig = cpu_sig, .err = 0 }; + int ret; + + ret = smp_call_function_single(cpu, collect_cpu_info_local, &ctx, 1); + if (!ret) + ret = ctx.err; + + return ret; +} + +static int collect_cpu_info(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + int ret; + + memset(uci, 0, sizeof(*uci)); + + ret = collect_cpu_info_on_target(cpu, &uci->cpu_sig); + if (!ret) + uci->valid = 1; + + return ret; +} + +struct apply_microcode_ctx { + int err; +}; + +static void apply_microcode_local(void *arg) +{ + struct apply_microcode_ctx *ctx = arg; + + ctx->err = microcode_ops->apply_microcode(smp_processor_id()); +} + +static int apply_microcode_on_target(int cpu) +{ + struct apply_microcode_ctx ctx = { .err = 0 }; + int ret; + + ret = smp_call_function_single(cpu, apply_microcode_local, &ctx, 1); + if (!ret) + ret = ctx.err; + + return ret; +} + #ifdef CONFIG_MICROCODE_OLD_INTERFACE static int do_microcode_update(const void __user *buf, size_t size) { - cpumask_t old; int error = 0; int cpu; - old = current->cpus_allowed; - for_each_online_cpu(cpu) { struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + enum ucode_state ustate; if (!uci->valid) continue; - set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); - error = microcode_ops->request_microcode_user(cpu, buf, size); - if (error < 0) - goto out; - if (!error) - microcode_ops->apply_microcode(cpu); + ustate = microcode_ops->request_microcode_user(cpu, buf, size); + if (ustate == UCODE_ERROR) { + error = -1; + break; + } else if (ustate == UCODE_OK) + apply_microcode_on_target(cpu); } -out: - set_cpus_allowed_ptr(current, &old); + return error; } @@ -143,19 +208,17 @@ static int microcode_open(struct inode *unused1, struct file *unused2) static ssize_t microcode_write(struct file *file, const char __user *buf, size_t len, loff_t *ppos) { - ssize_t ret; + ssize_t ret = -EINVAL; if ((len >> PAGE_SHIFT) > num_physpages) { - printk(KERN_ERR "microcode: too much data (max %ld pages)\n", - num_physpages); - return -EINVAL; + pr_err("microcode: too much data (max %ld pages)\n", num_physpages); + return ret; } get_online_cpus(); mutex_lock(µcode_mutex); - ret = do_microcode_update(buf, len); - if (!ret) + if (do_microcode_update(buf, len) == 0) ret = (ssize_t)len; mutex_unlock(µcode_mutex); @@ -165,15 +228,15 @@ static ssize_t microcode_write(struct file *file, const char __user *buf, } static const struct file_operations microcode_fops = { - .owner = THIS_MODULE, - .write = microcode_write, - .open = microcode_open, + .owner = THIS_MODULE, + .write = microcode_write, + .open = microcode_open, }; static struct miscdevice microcode_dev = { - .minor = MICROCODE_MINOR, - .name = "microcode", - .fops = µcode_fops, + .minor = MICROCODE_MINOR, + .name = "microcode", + .fops = µcode_fops, }; static int __init microcode_dev_init(void) @@ -182,9 +245,7 @@ static int __init microcode_dev_init(void) error = misc_register(µcode_dev); if (error) { - printk(KERN_ERR - "microcode: can't misc_register on minor=%d\n", - MICROCODE_MINOR); + pr_err("microcode: can't misc_register on minor=%d\n", MICROCODE_MINOR); return error; } @@ -205,42 +266,51 @@ MODULE_ALIAS_MISCDEV(MICROCODE_MINOR); /* fake device for request_firmware */ static struct platform_device *microcode_pdev; -static long reload_for_cpu(void *unused) +static int reload_for_cpu(int cpu) { - struct ucode_cpu_info *uci = ucode_cpu_info + smp_processor_id(); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; int err = 0; mutex_lock(µcode_mutex); if (uci->valid) { - err = microcode_ops->request_microcode_fw(smp_processor_id(), - µcode_pdev->dev); - if (!err) - microcode_ops->apply_microcode(smp_processor_id()); + enum ucode_state ustate; + + ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev); + if (ustate == UCODE_OK) + apply_microcode_on_target(cpu); + else + if (ustate == UCODE_ERROR) + err = -EINVAL; } mutex_unlock(µcode_mutex); + return err; } static ssize_t reload_store(struct sys_device *dev, struct sysdev_attribute *attr, - const char *buf, size_t sz) + const char *buf, size_t size) { - char *end; - unsigned long val = simple_strtoul(buf, &end, 0); - int err = 0; + unsigned long val; int cpu = dev->id; + int ret = 0; + char *end; + val = simple_strtoul(buf, &end, 0); if (end == buf) return -EINVAL; + if (val == 1) { get_online_cpus(); if (cpu_online(cpu)) - err = work_on_cpu(cpu, reload_for_cpu, NULL); + ret = reload_for_cpu(cpu); put_online_cpus(); } - if (err) - return err; - return sz; + + if (!ret) + ret = size; + + return ret; } static ssize_t version_show(struct sys_device *dev, @@ -271,11 +341,11 @@ static struct attribute *mc_default_attrs[] = { }; static struct attribute_group mc_attr_group = { - .attrs = mc_default_attrs, - .name = "microcode", + .attrs = mc_default_attrs, + .name = "microcode", }; -static void __microcode_fini_cpu(int cpu) +static void microcode_fini_cpu(int cpu) { struct ucode_cpu_info *uci = ucode_cpu_info + cpu; @@ -283,103 +353,68 @@ static void __microcode_fini_cpu(int cpu) uci->valid = 0; } -static void microcode_fini_cpu(int cpu) -{ - mutex_lock(µcode_mutex); - __microcode_fini_cpu(cpu); - mutex_unlock(µcode_mutex); -} - -static void collect_cpu_info(int cpu) +static enum ucode_state microcode_resume_cpu(int cpu) { struct ucode_cpu_info *uci = ucode_cpu_info + cpu; - memset(uci, 0, sizeof(*uci)); - if (!microcode_ops->collect_cpu_info(cpu, &uci->cpu_sig)) - uci->valid = 1; + if (!uci->mc) + return UCODE_NFOUND; + + pr_debug("microcode: CPU%d updated upon resume\n", cpu); + apply_microcode_on_target(cpu); + + return UCODE_OK; } -static int microcode_resume_cpu(int cpu) +static enum ucode_state microcode_init_cpu(int cpu) { - struct ucode_cpu_info *uci = ucode_cpu_info + cpu; - struct cpu_signature nsig; + enum ucode_state ustate; - pr_debug("microcode: CPU%d resumed\n", cpu); + if (collect_cpu_info(cpu)) + return UCODE_ERROR; - if (!uci->mc) - return 1; + /* --dimm. Trigger a delayed update? */ + if (system_state != SYSTEM_RUNNING) + return UCODE_NFOUND; - /* - * Let's verify that the 'cached' ucode does belong - * to this cpu (a bit of paranoia): - */ - if (microcode_ops->collect_cpu_info(cpu, &nsig)) { - __microcode_fini_cpu(cpu); - printk(KERN_ERR "failed to collect_cpu_info for resuming cpu #%d\n", - cpu); - return -1; - } + ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev); - if ((nsig.sig != uci->cpu_sig.sig) || (nsig.pf != uci->cpu_sig.pf)) { - __microcode_fini_cpu(cpu); - printk(KERN_ERR "cached ucode doesn't match the resuming cpu #%d\n", - cpu); - /* Should we look for a new ucode here? */ - return 1; + if (ustate == UCODE_OK) { + pr_debug("microcode: CPU%d updated upon init\n", cpu); + apply_microcode_on_target(cpu); } - return 0; + return ustate; } -static long microcode_update_cpu(void *unused) +static enum ucode_state microcode_update_cpu(int cpu) { - struct ucode_cpu_info *uci = ucode_cpu_info + smp_processor_id(); - int err = 0; + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + enum ucode_state ustate; - /* - * Check if the system resume is in progress (uci->valid != NULL), - * otherwise just request a firmware: - */ - if (uci->valid) { - err = microcode_resume_cpu(smp_processor_id()); - } else { - collect_cpu_info(smp_processor_id()); - if (uci->valid && system_state == SYSTEM_RUNNING) - err = microcode_ops->request_microcode_fw( - smp_processor_id(), - µcode_pdev->dev); - } - if (!err) - microcode_ops->apply_microcode(smp_processor_id()); - return err; -} + if (uci->valid) + ustate = microcode_resume_cpu(cpu); + else + ustate = microcode_init_cpu(cpu); -static int microcode_init_cpu(int cpu) -{ - int err; - mutex_lock(µcode_mutex); - err = work_on_cpu(cpu, microcode_update_cpu, NULL); - mutex_unlock(µcode_mutex); - - return err; + return ustate; } static int mc_sysdev_add(struct sys_device *sys_dev) { int err, cpu = sys_dev->id; - struct ucode_cpu_info *uci = ucode_cpu_info + cpu; if (!cpu_online(cpu)) return 0; pr_debug("microcode: CPU%d added\n", cpu); - memset(uci, 0, sizeof(*uci)); err = sysfs_create_group(&sys_dev->kobj, &mc_attr_group); if (err) return err; - err = microcode_init_cpu(cpu); + if (microcode_init_cpu(cpu) == UCODE_ERROR) + err = -EINVAL; return err; } @@ -400,19 +435,30 @@ static int mc_sysdev_remove(struct sys_device *sys_dev) static int mc_sysdev_resume(struct sys_device *dev) { int cpu = dev->id; + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; if (!cpu_online(cpu)) return 0; - /* only CPU 0 will apply ucode here */ - microcode_update_cpu(NULL); + /* + * All non-bootup cpus are still disabled, + * so only CPU 0 will apply ucode here. + * + * Moreover, there can be no concurrent + * updates from any other places at this point. + */ + WARN_ON(cpu != 0); + + if (uci->valid && uci->mc) + microcode_ops->apply_microcode(cpu); + return 0; } static struct sysdev_driver mc_sysdev_driver = { - .add = mc_sysdev_add, - .remove = mc_sysdev_remove, - .resume = mc_sysdev_resume, + .add = mc_sysdev_add, + .remove = mc_sysdev_remove, + .resume = mc_sysdev_resume, }; static __cpuinit int @@ -425,15 +471,12 @@ mc_cpu_callback(struct notifier_block *nb, unsigned long action, void *hcpu) switch (action) { case CPU_ONLINE: case CPU_ONLINE_FROZEN: - if (microcode_init_cpu(cpu)) - printk(KERN_ERR "microcode: failed to init CPU%d\n", - cpu); + microcode_update_cpu(cpu); case CPU_DOWN_FAILED: case CPU_DOWN_FAILED_FROZEN: pr_debug("microcode: CPU%d added\n", cpu); if (sysfs_create_group(&sys_dev->kobj, &mc_attr_group)) - printk(KERN_ERR "microcode: Failed to create the sysfs " - "group for CPU%d\n", cpu); + pr_err("microcode: Failed to create group for CPU%d\n", cpu); break; case CPU_DOWN_PREPARE: case CPU_DOWN_PREPARE_FROZEN: @@ -465,13 +508,10 @@ static int __init microcode_init(void) microcode_ops = init_amd_microcode(); if (!microcode_ops) { - printk(KERN_ERR "microcode: no support for this CPU vendor\n"); + pr_err("microcode: no support for this CPU vendor\n"); return -ENODEV; } - error = microcode_dev_init(); - if (error) - return error; microcode_pdev = platform_device_register_simple("microcode", -1, NULL, 0); if (IS_ERR(microcode_pdev)) { @@ -480,23 +520,31 @@ static int __init microcode_init(void) } get_online_cpus(); + mutex_lock(µcode_mutex); + error = sysdev_driver_register(&cpu_sysdev_class, &mc_sysdev_driver); + + mutex_unlock(µcode_mutex); put_online_cpus(); + if (error) { - microcode_dev_exit(); platform_device_unregister(microcode_pdev); return error; } + error = microcode_dev_init(); + if (error) + return error; + register_hotcpu_notifier(&mc_cpu_notifier); - printk(KERN_INFO - "Microcode Update Driver: v" MICROCODE_VERSION + pr_info("Microcode Update Driver: v" MICROCODE_VERSION " <tigran@aivazian.fsnet.co.uk>," " Peter Oruba\n"); return 0; } +module_init(microcode_init); static void __exit microcode_exit(void) { @@ -505,16 +553,17 @@ static void __exit microcode_exit(void) unregister_hotcpu_notifier(&mc_cpu_notifier); get_online_cpus(); + mutex_lock(µcode_mutex); + sysdev_driver_unregister(&cpu_sysdev_class, &mc_sysdev_driver); + + mutex_unlock(µcode_mutex); put_online_cpus(); platform_device_unregister(microcode_pdev); microcode_ops = NULL; - printk(KERN_INFO - "Microcode Update Driver: v" MICROCODE_VERSION " removed.\n"); + pr_info("Microcode Update Driver: v" MICROCODE_VERSION " removed.\n"); } - -module_init(microcode_init); module_exit(microcode_exit); diff --git a/arch/x86/kernel/microcode_intel.c b/arch/x86/kernel/microcode_intel.c index 149b9ec7c1a..0d334ddd0a9 100644 --- a/arch/x86/kernel/microcode_intel.c +++ b/arch/x86/kernel/microcode_intel.c @@ -70,24 +70,11 @@ * Fix sigmatch() macro to handle old CPUs with pf == 0. * Thanks to Stuart Swales for pointing out this bug. */ -#include <linux/platform_device.h> -#include <linux/capability.h> -#include <linux/miscdevice.h> #include <linux/firmware.h> -#include <linux/smp_lock.h> -#include <linux/spinlock.h> -#include <linux/cpumask.h> #include <linux/uaccess.h> -#include <linux/vmalloc.h> #include <linux/kernel.h> #include <linux/module.h> -#include <linux/mutex.h> -#include <linux/sched.h> -#include <linux/init.h> -#include <linux/slab.h> -#include <linux/cpu.h> -#include <linux/fs.h> -#include <linux/mm.h> +#include <linux/vmalloc.h> #include <asm/microcode.h> #include <asm/processor.h> @@ -150,13 +137,9 @@ struct extended_sigtable { #define exttable_size(et) ((et)->count * EXT_SIGNATURE_SIZE + EXT_HEADER_SIZE) -/* serialize access to the physical write to MSR 0x79 */ -static DEFINE_SPINLOCK(microcode_update_lock); - static int collect_cpu_info(int cpu_num, struct cpu_signature *csig) { struct cpuinfo_x86 *c = &cpu_data(cpu_num); - unsigned long flags; unsigned int val[2]; memset(csig, 0, sizeof(*csig)); @@ -176,18 +159,14 @@ static int collect_cpu_info(int cpu_num, struct cpu_signature *csig) csig->pf = 1 << ((val[1] >> 18) & 7); } - /* serialize access to the physical write to MSR 0x79 */ - spin_lock_irqsave(µcode_update_lock, flags); - wrmsr(MSR_IA32_UCODE_REV, 0, 0); /* see notes above for revision 1.07. Apparent chip bug */ sync_core(); /* get the current revision from MSR 0x8B */ rdmsr(MSR_IA32_UCODE_REV, val[0], csig->rev); - spin_unlock_irqrestore(µcode_update_lock, flags); - pr_debug("microcode: collect_cpu_info : sig=0x%x, pf=0x%x, rev=0x%x\n", - csig->sig, csig->pf, csig->rev); + printk(KERN_INFO "microcode: CPU%d sig=0x%x, pf=0x%x, revision=0x%x\n", + cpu_num, csig->sig, csig->pf, csig->rev); return 0; } @@ -318,11 +297,10 @@ get_matching_microcode(struct cpu_signature *cpu_sig, void *mc, int rev) return 0; } -static void apply_microcode(int cpu) +static int apply_microcode(int cpu) { struct microcode_intel *mc_intel; struct ucode_cpu_info *uci; - unsigned long flags; unsigned int val[2]; int cpu_num; @@ -334,10 +312,7 @@ static void apply_microcode(int cpu) BUG_ON(cpu_num != cpu); if (mc_intel == NULL) - return; - - /* serialize access to the physical write to MSR 0x79 */ - spin_lock_irqsave(µcode_update_lock, flags); + return 0; /* write microcode via MSR 0x79 */ wrmsr(MSR_IA32_UCODE_WRITE, @@ -351,30 +326,32 @@ static void apply_microcode(int cpu) /* get the current revision from MSR 0x8B */ rdmsr(MSR_IA32_UCODE_REV, val[0], val[1]); - spin_unlock_irqrestore(µcode_update_lock, flags); if (val[1] != mc_intel->hdr.rev) { - printk(KERN_ERR "microcode: CPU%d update from revision " - "0x%x to 0x%x failed\n", - cpu_num, uci->cpu_sig.rev, val[1]); - return; + printk(KERN_ERR "microcode: CPU%d update " + "to revision 0x%x failed\n", + cpu_num, mc_intel->hdr.rev); + return -1; } - printk(KERN_INFO "microcode: CPU%d updated from revision " - "0x%x to 0x%x, date = %04x-%02x-%02x \n", - cpu_num, uci->cpu_sig.rev, val[1], + printk(KERN_INFO "microcode: CPU%d updated to revision " + "0x%x, date = %04x-%02x-%02x \n", + cpu_num, val[1], mc_intel->hdr.date & 0xffff, mc_intel->hdr.date >> 24, (mc_intel->hdr.date >> 16) & 0xff); uci->cpu_sig.rev = val[1]; + + return 0; } -static int generic_load_microcode(int cpu, void *data, size_t size, - int (*get_ucode_data)(void *, const void *, size_t)) +static enum ucode_state generic_load_microcode(int cpu, void *data, size_t size, + int (*get_ucode_data)(void *, const void *, size_t)) { struct ucode_cpu_info *uci = ucode_cpu_info + cpu; u8 *ucode_ptr = data, *new_mc = NULL, *mc; int new_rev = uci->cpu_sig.rev; unsigned int leftover = size; + enum ucode_state state = UCODE_OK; while (leftover) { struct microcode_header_intel mc_header; @@ -412,11 +389,15 @@ static int generic_load_microcode(int cpu, void *data, size_t size, leftover -= mc_size; } - if (!new_mc) + if (leftover) { + if (new_mc) + vfree(new_mc); + state = UCODE_ERROR; goto out; + } - if (leftover) { - vfree(new_mc); + if (!new_mc) { + state = UCODE_NFOUND; goto out; } @@ -427,9 +408,8 @@ static int generic_load_microcode(int cpu, void *data, size_t size, pr_debug("microcode: CPU%d found a matching microcode update with" " version 0x%x (current=0x%x)\n", cpu, new_rev, uci->cpu_sig.rev); - - out: - return (int)leftover; +out: + return state; } static int get_ucode_fw(void *to, const void *from, size_t n) @@ -438,21 +418,19 @@ static int get_ucode_fw(void *to, const void *from, size_t n) return 0; } -static int request_microcode_fw(int cpu, struct device *device) +static enum ucode_state request_microcode_fw(int cpu, struct device *device) { char name[30]; struct cpuinfo_x86 *c = &cpu_data(cpu); const struct firmware *firmware; - int ret; + enum ucode_state ret; - /* We should bind the task to the CPU */ - BUG_ON(cpu != raw_smp_processor_id()); sprintf(name, "intel-ucode/%02x-%02x-%02x", c->x86, c->x86_model, c->x86_mask); - ret = request_firmware(&firmware, name, device); - if (ret) { + + if (request_firmware(&firmware, name, device)) { pr_debug("microcode: data file %s load failed\n", name); - return ret; + return UCODE_NFOUND; } ret = generic_load_microcode(cpu, (void *)firmware->data, @@ -468,11 +446,9 @@ static int get_ucode_user(void *to, const void *from, size_t n) return copy_from_user(to, from, n); } -static int request_microcode_user(int cpu, const void __user *buf, size_t size) +static enum ucode_state +request_microcode_user(int cpu, const void __user *buf, size_t size) { - /* We should bind the task to the CPU */ - BUG_ON(cpu != raw_smp_processor_id()); - return generic_load_microcode(cpu, (void *)buf, size, &get_ucode_user); } diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index 70fd7e414c1..651c93b2886 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -17,6 +17,7 @@ #include <linux/acpi.h> #include <linux/module.h> #include <linux/smp.h> +#include <linux/pci.h> #include <asm/mtrr.h> #include <asm/mpspec.h> @@ -870,24 +871,17 @@ static inline void __init check_irq_src(struct mpc_intsrc *m, int *nr_m_spare) {} #endif /* CONFIG_X86_IO_APIC */ -static int check_slot(unsigned long mpc_new_phys, unsigned long mpc_new_length, - int count) +static int +check_slot(unsigned long mpc_new_phys, unsigned long mpc_new_length, int count) { - if (!mpc_new_phys) { - pr_info("No spare slots, try to append...take your risk, " - "new mpc_length %x\n", count); - } else { - if (count <= mpc_new_length) - pr_info("No spare slots, try to append..., " - "new mpc_length %x\n", count); - else { - pr_err("mpc_new_length %lx is too small\n", - mpc_new_length); - return -1; - } + int ret = 0; + + if (!mpc_new_phys || count <= mpc_new_length) { + WARN(1, "update_mptable: No spare slots (length: %x)\n", count); + return -1; } - return 0; + return ret; } static int __init replace_intsrc_all(struct mpc_table *mpc, @@ -946,7 +940,7 @@ static int __init replace_intsrc_all(struct mpc_table *mpc, } else { struct mpc_intsrc *m = (struct mpc_intsrc *)mpt; count += sizeof(struct mpc_intsrc); - if (!check_slot(mpc_new_phys, mpc_new_length, count)) + if (check_slot(mpc_new_phys, mpc_new_length, count) < 0) goto out; assign_to_mpc_intsrc(&mp_irqs[i], m); mpc->length = count; @@ -963,11 +957,14 @@ out: return 0; } -static int __initdata enable_update_mptable; +int enable_update_mptable; static int __init update_mptable_setup(char *str) { enable_update_mptable = 1; +#ifdef CONFIG_PCI + pci_routeirq = 1; +#endif return 0; } early_param("update_mptable", update_mptable_setup); @@ -980,6 +977,9 @@ static int __initdata alloc_mptable; static int __init parse_alloc_mptable_opt(char *p) { enable_update_mptable = 1; +#ifdef CONFIG_PCI + pci_routeirq = 1; +#endif alloc_mptable = 1; if (!p) return 0; diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 9faf43bea33..70ec9b951d7 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -248,18 +248,16 @@ static DEFINE_PER_CPU(enum paravirt_lazy_mode, paravirt_lazy_mode) = PARAVIRT_LA static inline void enter_lazy(enum paravirt_lazy_mode mode) { - BUG_ON(__get_cpu_var(paravirt_lazy_mode) != PARAVIRT_LAZY_NONE); - BUG_ON(preemptible()); + BUG_ON(percpu_read(paravirt_lazy_mode) != PARAVIRT_LAZY_NONE); - __get_cpu_var(paravirt_lazy_mode) = mode; + percpu_write(paravirt_lazy_mode, mode); } -void paravirt_leave_lazy(enum paravirt_lazy_mode mode) +static void leave_lazy(enum paravirt_lazy_mode mode) { - BUG_ON(__get_cpu_var(paravirt_lazy_mode) != mode); - BUG_ON(preemptible()); + BUG_ON(percpu_read(paravirt_lazy_mode) != mode); - __get_cpu_var(paravirt_lazy_mode) = PARAVIRT_LAZY_NONE; + percpu_write(paravirt_lazy_mode, PARAVIRT_LAZY_NONE); } void paravirt_enter_lazy_mmu(void) @@ -269,22 +267,36 @@ void paravirt_enter_lazy_mmu(void) void paravirt_leave_lazy_mmu(void) { - paravirt_leave_lazy(PARAVIRT_LAZY_MMU); + leave_lazy(PARAVIRT_LAZY_MMU); } -void paravirt_enter_lazy_cpu(void) +void paravirt_start_context_switch(struct task_struct *prev) { + BUG_ON(preemptible()); + + if (percpu_read(paravirt_lazy_mode) == PARAVIRT_LAZY_MMU) { + arch_leave_lazy_mmu_mode(); + set_ti_thread_flag(task_thread_info(prev), TIF_LAZY_MMU_UPDATES); + } enter_lazy(PARAVIRT_LAZY_CPU); } -void paravirt_leave_lazy_cpu(void) +void paravirt_end_context_switch(struct task_struct *next) { - paravirt_leave_lazy(PARAVIRT_LAZY_CPU); + BUG_ON(preemptible()); + + leave_lazy(PARAVIRT_LAZY_CPU); + + if (test_and_clear_ti_thread_flag(task_thread_info(next), TIF_LAZY_MMU_UPDATES)) + arch_enter_lazy_mmu_mode(); } enum paravirt_lazy_mode paravirt_get_lazy_mode(void) { - return __get_cpu_var(paravirt_lazy_mode); + if (in_interrupt()) + return PARAVIRT_LAZY_NONE; + + return percpu_read(paravirt_lazy_mode); } void arch_flush_lazy_mmu_mode(void) @@ -292,7 +304,6 @@ void arch_flush_lazy_mmu_mode(void) preempt_disable(); if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_MMU) { - WARN_ON(preempt_count() == 1); arch_leave_lazy_mmu_mode(); arch_enter_lazy_mmu_mode(); } @@ -300,19 +311,6 @@ void arch_flush_lazy_mmu_mode(void) preempt_enable(); } -void arch_flush_lazy_cpu_mode(void) -{ - preempt_disable(); - - if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_CPU) { - WARN_ON(preempt_count() == 1); - arch_leave_lazy_cpu_mode(); - arch_enter_lazy_cpu_mode(); - } - - preempt_enable(); -} - struct pv_info pv_info = { .name = "bare hardware", .paravirt_enabled = 0, @@ -404,10 +402,8 @@ struct pv_cpu_ops pv_cpu_ops = { .set_iopl_mask = native_set_iopl_mask, .io_delay = native_io_delay, - .lazy_mode = { - .enter = paravirt_nop, - .leave = paravirt_nop, - }, + .start_context_switch = paravirt_nop, + .end_context_switch = paravirt_nop, }; struct pv_apic_ops pv_apic_ops = { diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index 755c21e906f..971a3bec47a 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -186,37 +186,6 @@ static struct cal_chipset_ops calioc2_chip_ops = { static struct calgary_bus_info bus_info[MAX_PHB_BUS_NUM] = { { NULL, 0, 0 }, }; -/* enable this to stress test the chip's TCE cache */ -#ifdef CONFIG_IOMMU_DEBUG -static int debugging = 1; - -static inline unsigned long verify_bit_range(unsigned long* bitmap, - int expected, unsigned long start, unsigned long end) -{ - unsigned long idx = start; - - BUG_ON(start >= end); - - while (idx < end) { - if (!!test_bit(idx, bitmap) != expected) - return idx; - ++idx; - } - - /* all bits have the expected value */ - return ~0UL; -} -#else /* debugging is disabled */ -static int debugging; - -static inline unsigned long verify_bit_range(unsigned long* bitmap, - int expected, unsigned long start, unsigned long end) -{ - return ~0UL; -} - -#endif /* CONFIG_IOMMU_DEBUG */ - static inline int translation_enabled(struct iommu_table *tbl) { /* only PHBs with translation enabled have an IOMMU table */ @@ -228,7 +197,6 @@ static void iommu_range_reserve(struct iommu_table *tbl, { unsigned long index; unsigned long end; - unsigned long badbit; unsigned long flags; index = start_addr >> PAGE_SHIFT; @@ -243,14 +211,6 @@ static void iommu_range_reserve(struct iommu_table *tbl, spin_lock_irqsave(&tbl->it_lock, flags); - badbit = verify_bit_range(tbl->it_map, 0, index, end); - if (badbit != ~0UL) { - if (printk_ratelimit()) - printk(KERN_ERR "Calgary: entry already allocated at " - "0x%lx tbl %p dma 0x%lx npages %u\n", - badbit, tbl, start_addr, npages); - } - iommu_area_reserve(tbl->it_map, index, npages); spin_unlock_irqrestore(&tbl->it_lock, flags); @@ -326,7 +286,6 @@ static void iommu_free(struct iommu_table *tbl, dma_addr_t dma_addr, unsigned int npages) { unsigned long entry; - unsigned long badbit; unsigned long badend; unsigned long flags; @@ -346,14 +305,6 @@ static void iommu_free(struct iommu_table *tbl, dma_addr_t dma_addr, spin_lock_irqsave(&tbl->it_lock, flags); - badbit = verify_bit_range(tbl->it_map, 1, entry, entry + npages); - if (badbit != ~0UL) { - if (printk_ratelimit()) - printk(KERN_ERR "Calgary: bit is off at 0x%lx " - "tbl %p dma 0x%Lx entry 0x%lx npages %u\n", - badbit, tbl, dma_addr, entry, npages); - } - iommu_area_free(tbl->it_map, entry, npages); spin_unlock_irqrestore(&tbl->it_lock, flags); @@ -1488,9 +1439,8 @@ void __init detect_calgary(void) iommu_detected = 1; calgary_detected = 1; printk(KERN_INFO "PCI-DMA: Calgary IOMMU detected.\n"); - printk(KERN_INFO "PCI-DMA: Calgary TCE table spec is %d, " - "CONFIG_IOMMU_DEBUG is %s.\n", specified_table_size, - debugging ? "enabled" : "disabled"); + printk(KERN_INFO "PCI-DMA: Calgary TCE table spec is %d\n", + specified_table_size); /* swiotlb for devices that aren't behind the Calgary. */ if (max_pfn > MAX_DMA32_PFN) diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index b284b58c035..cfd9f906389 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -144,48 +144,21 @@ static void flush_gart(void) } #ifdef CONFIG_IOMMU_LEAK - -#define SET_LEAK(x) \ - do { \ - if (iommu_leak_tab) \ - iommu_leak_tab[x] = __builtin_return_address(0);\ - } while (0) - -#define CLEAR_LEAK(x) \ - do { \ - if (iommu_leak_tab) \ - iommu_leak_tab[x] = NULL; \ - } while (0) - /* Debugging aid for drivers that don't free their IOMMU tables */ -static void **iommu_leak_tab; static int leak_trace; static int iommu_leak_pages = 20; static void dump_leak(void) { - int i; static int dump; - if (dump || !iommu_leak_tab) + if (dump) return; dump = 1; - show_stack(NULL, NULL); - /* Very crude. dump some from the end of the table too */ - printk(KERN_DEBUG "Dumping %d pages from end of IOMMU:\n", - iommu_leak_pages); - for (i = 0; i < iommu_leak_pages; i += 2) { - printk(KERN_DEBUG "%lu: ", iommu_pages-i); - printk_address((unsigned long) iommu_leak_tab[iommu_pages-i], - 0); - printk(KERN_CONT "%c", (i+1)%2 == 0 ? '\n' : ' '); - } - printk(KERN_DEBUG "\n"); + show_stack(NULL, NULL); + debug_dma_dump_mappings(NULL); } -#else -# define SET_LEAK(x) -# define CLEAR_LEAK(x) #endif static void iommu_full(struct device *dev, size_t size, int dir) @@ -248,7 +221,6 @@ static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem, for (i = 0; i < npages; i++) { iommu_gatt_base[iommu_page + i] = GPTE_ENCODE(phys_mem); - SET_LEAK(iommu_page + i); phys_mem += PAGE_SIZE; } return iommu_bus_base + iommu_page*PAGE_SIZE + (phys_mem & ~PAGE_MASK); @@ -294,7 +266,6 @@ static void gart_unmap_page(struct device *dev, dma_addr_t dma_addr, npages = iommu_num_pages(dma_addr, size, PAGE_SIZE); for (i = 0; i < npages; i++) { iommu_gatt_base[iommu_page + i] = gart_unmapped_entry; - CLEAR_LEAK(iommu_page + i); } free_iommu(iommu_page, npages); } @@ -377,7 +348,6 @@ static int __dma_map_cont(struct device *dev, struct scatterlist *start, pages = iommu_num_pages(s->offset, s->length, PAGE_SIZE); while (pages--) { iommu_gatt_base[iommu_page] = GPTE_ENCODE(addr); - SET_LEAK(iommu_page); addr += PAGE_SIZE; iommu_page++; } @@ -688,8 +658,6 @@ static __init int init_k8_gatt(struct agp_kern_info *info) agp_gatt_table = gatt; - enable_gart_translations(); - error = sysdev_class_register(&gart_sysdev_class); if (!error) error = sysdev_register(&device_gart); @@ -801,11 +769,12 @@ void __init gart_iommu_init(void) #ifdef CONFIG_IOMMU_LEAK if (leak_trace) { - iommu_leak_tab = (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO, - get_order(iommu_pages*sizeof(void *))); - if (!iommu_leak_tab) + int ret; + + ret = dma_debug_resize_entries(iommu_pages); + if (ret) printk(KERN_DEBUG - "PCI-DMA: Cannot allocate leak trace area\n"); + "PCI-DMA: Cannot trace all the entries\n"); } #endif @@ -845,6 +814,14 @@ void __init gart_iommu_init(void) * the pages as Not-Present: */ wbinvd(); + + /* + * Now all caches are flushed and we can safely enable + * GART hardware. Doing it early leaves the possibility + * of stale cache entries that can lead to GART PTE + * errors. + */ + enable_gart_translations(); /* * Try to workaround a bug (thanks to BenH): diff --git a/arch/x86/kernel/pci-swiotlb.c b/arch/x86/kernel/pci-swiotlb.c index 221a3853e26..a1712f2b50f 100644 --- a/arch/x86/kernel/pci-swiotlb.c +++ b/arch/x86/kernel/pci-swiotlb.c @@ -28,7 +28,7 @@ dma_addr_t swiotlb_phys_to_bus(struct device *hwdev, phys_addr_t paddr) return paddr; } -phys_addr_t swiotlb_bus_to_phys(dma_addr_t baddr) +phys_addr_t swiotlb_bus_to_phys(struct device *hwdev, dma_addr_t baddr) { return baddr; } diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index ca989158e84..3bb2be1649b 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -8,12 +8,15 @@ #include <linux/module.h> #include <linux/pm.h> #include <linux/clockchips.h> +#include <linux/random.h> #include <trace/power.h> #include <asm/system.h> #include <asm/apic.h> +#include <asm/syscalls.h> #include <asm/idle.h> #include <asm/uaccess.h> #include <asm/i387.h> +#include <asm/ds.h> unsigned long idle_halt; EXPORT_SYMBOL(idle_halt); @@ -45,6 +48,8 @@ void free_thread_xstate(struct task_struct *tsk) kmem_cache_free(task_xstate_cachep, tsk->thread.xstate); tsk->thread.xstate = NULL; } + + WARN(tsk->thread.ds_ctx, "leaking DS context\n"); } void free_thread_info(struct thread_info *ti) @@ -83,8 +88,6 @@ void exit_thread(void) put_cpu(); kfree(bp); } - - ds_exit_thread(current); } void flush_thread(void) @@ -613,3 +616,16 @@ static int __init idle_setup(char *str) } early_param("idle", idle_setup); +unsigned long arch_align_stack(unsigned long sp) +{ + if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space) + sp -= get_random_int() % 8192; + return sp & ~0xf; +} + +unsigned long arch_randomize_brk(struct mm_struct *mm) +{ + unsigned long range_end = mm->brk + 0x02000000; + return randomize_range(mm->brk, range_end, 0) ? : mm->brk; +} + diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 76f8f84043a..59f4524984a 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -9,8 +9,6 @@ * This file handles the architecture-dependent parts of process handling.. */ -#include <stdarg.h> - #include <linux/stackprotector.h> #include <linux/cpu.h> #include <linux/errno.h> @@ -33,7 +31,6 @@ #include <linux/module.h> #include <linux/kallsyms.h> #include <linux/ptrace.h> -#include <linux/random.h> #include <linux/personality.h> #include <linux/tick.h> #include <linux/percpu.h> @@ -290,7 +287,8 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, p->thread.io_bitmap_max = 0; } - ds_copy_thread(p, current); + clear_tsk_thread_flag(p, TIF_DS_AREA_MSR); + p->thread.ds_ctx = NULL; clear_tsk_thread_flag(p, TIF_DEBUGCTLMSR); p->thread.debugctlmsr = 0; @@ -407,7 +405,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) * done before math_state_restore, so the TS bit is up * to date. */ - arch_leave_lazy_cpu_mode(); + arch_end_context_switch(next_p); /* If the task has used fpu the last 5 timeslices, just do a full * restore of the math state immediately to avoid the trap; the @@ -497,15 +495,3 @@ unsigned long get_wchan(struct task_struct *p) return 0; } -unsigned long arch_align_stack(unsigned long sp) -{ - if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space) - sp -= get_random_int() % 8192; - return sp & ~0xf; -} - -unsigned long arch_randomize_brk(struct mm_struct *mm) -{ - unsigned long range_end = mm->brk + 0x02000000; - return randomize_range(mm->brk, range_end, 0) ? : mm->brk; -} diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index b751a41392b..ebefb5407b9 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -14,8 +14,6 @@ * This file handles the architecture-dependent parts of process handling.. */ -#include <stdarg.h> - #include <linux/stackprotector.h> #include <linux/cpu.h> #include <linux/errno.h> @@ -32,7 +30,6 @@ #include <linux/delay.h> #include <linux/module.h> #include <linux/ptrace.h> -#include <linux/random.h> #include <linux/notifier.h> #include <linux/kprobes.h> #include <linux/kdebug.h> @@ -335,7 +332,8 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, goto out; } - ds_copy_thread(p, me); + clear_tsk_thread_flag(p, TIF_DS_AREA_MSR); + p->thread.ds_ctx = NULL; clear_tsk_thread_flag(p, TIF_DEBUGCTLMSR); p->thread.debugctlmsr = 0; @@ -428,7 +426,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) * done before math_state_restore, so the TS bit is up * to date. */ - arch_leave_lazy_cpu_mode(); + arch_end_context_switch(next_p); /* * Switch FS and GS. @@ -660,15 +658,3 @@ long sys_arch_prctl(int code, unsigned long addr) return do_arch_prctl(current, code, addr); } -unsigned long arch_align_stack(unsigned long sp) -{ - if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space) - sp -= get_random_int() % 8192; - return sp & ~0xf; -} - -unsigned long arch_randomize_brk(struct mm_struct *mm) -{ - unsigned long range_end = mm->brk + 0x02000000; - return randomize_range(mm->brk, range_end, 0) ? : mm->brk; -} diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 23b7c8f017e..09ecbde91c1 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -21,6 +21,7 @@ #include <linux/audit.h> #include <linux/seccomp.h> #include <linux/signal.h> +#include <linux/workqueue.h> #include <asm/uaccess.h> #include <asm/pgtable.h> @@ -578,17 +579,130 @@ static int ioperm_get(struct task_struct *target, } #ifdef CONFIG_X86_PTRACE_BTS +/* + * A branch trace store context. + * + * Contexts may only be installed by ptrace_bts_config() and only for + * ptraced tasks. + * + * Contexts are destroyed when the tracee is detached from the tracer. + * The actual destruction work requires interrupts enabled, so the + * work is deferred and will be scheduled during __ptrace_unlink(). + * + * Contexts hold an additional task_struct reference on the traced + * task, as well as a reference on the tracer's mm. + * + * Ptrace already holds a task_struct for the duration of ptrace operations, + * but since destruction is deferred, it may be executed after both + * tracer and tracee exited. + */ +struct bts_context { + /* The branch trace handle. */ + struct bts_tracer *tracer; + + /* The buffer used to store the branch trace and its size. */ + void *buffer; + unsigned int size; + + /* The mm that paid for the above buffer. */ + struct mm_struct *mm; + + /* The task this context belongs to. */ + struct task_struct *task; + + /* The signal to send on a bts buffer overflow. */ + unsigned int bts_ovfl_signal; + + /* The work struct to destroy a context. */ + struct work_struct work; +}; + +static int alloc_bts_buffer(struct bts_context *context, unsigned int size) +{ + void *buffer = NULL; + int err = -ENOMEM; + + err = account_locked_memory(current->mm, current->signal->rlim, size); + if (err < 0) + return err; + + buffer = kzalloc(size, GFP_KERNEL); + if (!buffer) + goto out_refund; + + context->buffer = buffer; + context->size = size; + context->mm = get_task_mm(current); + + return 0; + + out_refund: + refund_locked_memory(current->mm, size); + return err; +} + +static inline void free_bts_buffer(struct bts_context *context) +{ + if (!context->buffer) + return; + + kfree(context->buffer); + context->buffer = NULL; + + refund_locked_memory(context->mm, context->size); + context->size = 0; + + mmput(context->mm); + context->mm = NULL; +} + +static void free_bts_context_work(struct work_struct *w) +{ + struct bts_context *context; + + context = container_of(w, struct bts_context, work); + + ds_release_bts(context->tracer); + put_task_struct(context->task); + free_bts_buffer(context); + kfree(context); +} + +static inline void free_bts_context(struct bts_context *context) +{ + INIT_WORK(&context->work, free_bts_context_work); + schedule_work(&context->work); +} + +static inline struct bts_context *alloc_bts_context(struct task_struct *task) +{ + struct bts_context *context = kzalloc(sizeof(*context), GFP_KERNEL); + if (context) { + context->task = task; + task->bts = context; + + get_task_struct(task); + } + + return context; +} + static int ptrace_bts_read_record(struct task_struct *child, size_t index, struct bts_struct __user *out) { + struct bts_context *context; const struct bts_trace *trace; struct bts_struct bts; const unsigned char *at; int error; - trace = ds_read_bts(child->bts); + context = child->bts; + if (!context) + return -ESRCH; + + trace = ds_read_bts(context->tracer); if (!trace) - return -EPERM; + return -ESRCH; at = trace->ds.top - ((index + 1) * trace->ds.size); if ((void *)at < trace->ds.begin) @@ -597,7 +711,7 @@ static int ptrace_bts_read_record(struct task_struct *child, size_t index, if (!trace->read) return -EOPNOTSUPP; - error = trace->read(child->bts, at, &bts); + error = trace->read(context->tracer, at, &bts); if (error < 0) return error; @@ -611,13 +725,18 @@ static int ptrace_bts_drain(struct task_struct *child, long size, struct bts_struct __user *out) { + struct bts_context *context; const struct bts_trace *trace; const unsigned char *at; int error, drained = 0; - trace = ds_read_bts(child->bts); + context = child->bts; + if (!context) + return -ESRCH; + + trace = ds_read_bts(context->tracer); if (!trace) - return -EPERM; + return -ESRCH; if (!trace->read) return -EOPNOTSUPP; @@ -628,9 +747,8 @@ static int ptrace_bts_drain(struct task_struct *child, for (at = trace->ds.begin; (void *)at < trace->ds.top; out++, drained++, at += trace->ds.size) { struct bts_struct bts; - int error; - error = trace->read(child->bts, at, &bts); + error = trace->read(context->tracer, at, &bts); if (error < 0) return error; @@ -640,35 +758,18 @@ static int ptrace_bts_drain(struct task_struct *child, memset(trace->ds.begin, 0, trace->ds.n * trace->ds.size); - error = ds_reset_bts(child->bts); + error = ds_reset_bts(context->tracer); if (error < 0) return error; return drained; } -static int ptrace_bts_allocate_buffer(struct task_struct *child, size_t size) -{ - child->bts_buffer = alloc_locked_buffer(size); - if (!child->bts_buffer) - return -ENOMEM; - - child->bts_size = size; - - return 0; -} - -static void ptrace_bts_free_buffer(struct task_struct *child) -{ - free_locked_buffer(child->bts_buffer, child->bts_size); - child->bts_buffer = NULL; - child->bts_size = 0; -} - static int ptrace_bts_config(struct task_struct *child, long cfg_size, const struct ptrace_bts_config __user *ucfg) { + struct bts_context *context; struct ptrace_bts_config cfg; unsigned int flags = 0; @@ -678,28 +779,33 @@ static int ptrace_bts_config(struct task_struct *child, if (copy_from_user(&cfg, ucfg, sizeof(cfg))) return -EFAULT; - if (child->bts) { - ds_release_bts(child->bts); - child->bts = NULL; - } + context = child->bts; + if (!context) + context = alloc_bts_context(child); + if (!context) + return -ENOMEM; if (cfg.flags & PTRACE_BTS_O_SIGNAL) { if (!cfg.signal) return -EINVAL; - child->thread.bts_ovfl_signal = cfg.signal; return -EOPNOTSUPP; + context->bts_ovfl_signal = cfg.signal; } - if ((cfg.flags & PTRACE_BTS_O_ALLOC) && - (cfg.size != child->bts_size)) { - int error; + ds_release_bts(context->tracer); + context->tracer = NULL; - ptrace_bts_free_buffer(child); + if ((cfg.flags & PTRACE_BTS_O_ALLOC) && (cfg.size != context->size)) { + int err; - error = ptrace_bts_allocate_buffer(child, cfg.size); - if (error < 0) - return error; + free_bts_buffer(context); + if (!cfg.size) + return 0; + + err = alloc_bts_buffer(context, cfg.size); + if (err < 0) + return err; } if (cfg.flags & PTRACE_BTS_O_TRACE) @@ -708,15 +814,14 @@ static int ptrace_bts_config(struct task_struct *child, if (cfg.flags & PTRACE_BTS_O_SCHED) flags |= BTS_TIMESTAMPS; - child->bts = ds_request_bts(child, child->bts_buffer, child->bts_size, - /* ovfl = */ NULL, /* th = */ (size_t)-1, - flags); - if (IS_ERR(child->bts)) { - int error = PTR_ERR(child->bts); - - ptrace_bts_free_buffer(child); - child->bts = NULL; + context->tracer = + ds_request_bts_task(child, context->buffer, context->size, + NULL, (size_t)-1, flags); + if (unlikely(IS_ERR(context->tracer))) { + int error = PTR_ERR(context->tracer); + free_bts_buffer(context); + context->tracer = NULL; return error; } @@ -727,20 +832,25 @@ static int ptrace_bts_status(struct task_struct *child, long cfg_size, struct ptrace_bts_config __user *ucfg) { + struct bts_context *context; const struct bts_trace *trace; struct ptrace_bts_config cfg; + context = child->bts; + if (!context) + return -ESRCH; + if (cfg_size < sizeof(cfg)) return -EIO; - trace = ds_read_bts(child->bts); + trace = ds_read_bts(context->tracer); if (!trace) - return -EPERM; + return -ESRCH; memset(&cfg, 0, sizeof(cfg)); - cfg.size = trace->ds.end - trace->ds.begin; - cfg.signal = child->thread.bts_ovfl_signal; - cfg.bts_size = sizeof(struct bts_struct); + cfg.size = trace->ds.end - trace->ds.begin; + cfg.signal = context->bts_ovfl_signal; + cfg.bts_size = sizeof(struct bts_struct); if (cfg.signal) cfg.flags |= PTRACE_BTS_O_SIGNAL; @@ -759,80 +869,51 @@ static int ptrace_bts_status(struct task_struct *child, static int ptrace_bts_clear(struct task_struct *child) { + struct bts_context *context; const struct bts_trace *trace; - trace = ds_read_bts(child->bts); + context = child->bts; + if (!context) + return -ESRCH; + + trace = ds_read_bts(context->tracer); if (!trace) - return -EPERM; + return -ESRCH; memset(trace->ds.begin, 0, trace->ds.n * trace->ds.size); - return ds_reset_bts(child->bts); + return ds_reset_bts(context->tracer); } static int ptrace_bts_size(struct task_struct *child) { + struct bts_context *context; const struct bts_trace *trace; - trace = ds_read_bts(child->bts); + context = child->bts; + if (!context) + return -ESRCH; + + trace = ds_read_bts(context->tracer); if (!trace) - return -EPERM; + return -ESRCH; return (trace->ds.top - trace->ds.begin) / trace->ds.size; } -static void ptrace_bts_fork(struct task_struct *tsk) -{ - tsk->bts = NULL; - tsk->bts_buffer = NULL; - tsk->bts_size = 0; - tsk->thread.bts_ovfl_signal = 0; -} - -static void ptrace_bts_untrace(struct task_struct *child) +/* + * Called from __ptrace_unlink() after the child has been moved back + * to its original parent. + */ +void ptrace_bts_untrace(struct task_struct *child) { if (unlikely(child->bts)) { - ds_release_bts(child->bts); + free_bts_context(child->bts); child->bts = NULL; - - /* We cannot update total_vm and locked_vm since - child's mm is already gone. But we can reclaim the - memory. */ - kfree(child->bts_buffer); - child->bts_buffer = NULL; - child->bts_size = 0; } } - -static void ptrace_bts_detach(struct task_struct *child) -{ - /* - * Ptrace_detach() races with ptrace_untrace() in case - * the child dies and is reaped by another thread. - * - * We only do the memory accounting at this point and - * leave the buffer deallocation and the bts tracer - * release to ptrace_bts_untrace() which will be called - * later on with tasklist_lock held. - */ - release_locked_buffer(child->bts_buffer, child->bts_size); -} -#else -static inline void ptrace_bts_fork(struct task_struct *tsk) {} -static inline void ptrace_bts_detach(struct task_struct *child) {} -static inline void ptrace_bts_untrace(struct task_struct *child) {} #endif /* CONFIG_X86_PTRACE_BTS */ -void x86_ptrace_fork(struct task_struct *child, unsigned long clone_flags) -{ - ptrace_bts_fork(child); -} - -void x86_ptrace_untrace(struct task_struct *child) -{ - ptrace_bts_untrace(child); -} - /* * Called by kernel/ptrace.c when detaching.. * @@ -844,7 +925,6 @@ void ptrace_disable(struct task_struct *child) #ifdef TIF_SYSCALL_EMU clear_tsk_thread_flag(child, TIF_SYSCALL_EMU); #endif - ptrace_bts_detach(child); } #if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION diff --git a/arch/x86/kernel/quirks.c b/arch/x86/kernel/quirks.c index 7563b31b4f0..af71d06624b 100644 --- a/arch/x86/kernel/quirks.c +++ b/arch/x86/kernel/quirks.c @@ -491,5 +491,42 @@ void force_hpet_resume(void) break; } } +#endif + +#if defined(CONFIG_PCI) && defined(CONFIG_NUMA) +/* Set correct numa_node information for AMD NB functions */ +static void __init quirk_amd_nb_node(struct pci_dev *dev) +{ + struct pci_dev *nb_ht; + unsigned int devfn; + u32 val; + + devfn = PCI_DEVFN(PCI_SLOT(dev->devfn), 0); + nb_ht = pci_get_slot(dev->bus, devfn); + if (!nb_ht) + return; + + pci_read_config_dword(nb_ht, 0x60, &val); + set_dev_node(&dev->dev, val & 7); + pci_dev_put(dev); +} +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB, + quirk_amd_nb_node); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_ADDRMAP, + quirk_amd_nb_node); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MEMCTL, + quirk_amd_nb_node); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MISC, + quirk_amd_nb_node); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_HT, + quirk_amd_nb_node); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_MAP, + quirk_amd_nb_node); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_DRAM, + quirk_amd_nb_node); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC, + quirk_amd_nb_node); +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_LINK, + quirk_amd_nb_node); #endif diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index 667188e0b5a..d2d1ce8170f 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -192,6 +192,15 @@ static struct dmi_system_id __initdata reboot_dmi_table[] = { DMI_MATCH(DMI_BOARD_NAME, "0KP561"), }, }, + { /* Handle problems with rebooting on Dell Optiplex 360 with 0T656F */ + .callback = set_bios_reboot, + .ident = "Dell OptiPlex 360", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 360"), + DMI_MATCH(DMI_BOARD_NAME, "0T656F"), + }, + }, { /* Handle problems with rebooting on Dell 2400's */ .callback = set_bios_reboot, .ident = "Dell PowerEdge 2400", diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index b4158439bf6..d1c636bf31a 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -112,6 +112,14 @@ #define ARCH_SETUP #endif +/* + * end_pfn only includes RAM, while max_pfn_mapped includes all e820 entries. + * The direct mapping extends to max_pfn_mapped, so that we can directly access + * apertures, ACPI and other tables without having to play with fixmaps. + */ +unsigned long max_low_pfn_mapped; +unsigned long max_pfn_mapped; + RESERVE_BRK(dmi_alloc, 65536); unsigned int boot_cpu_id __read_mostly; @@ -214,8 +222,8 @@ unsigned long mmu_cr4_features; unsigned long mmu_cr4_features = X86_CR4_PAE; #endif -/* Boot loader ID as an integer, for the benefit of proc_dointvec */ -int bootloader_type; +/* Boot loader ID and version as integers, for the benefit of proc_dointvec */ +int bootloader_type, bootloader_version; /* * Setup options @@ -706,6 +714,12 @@ void __init setup_arch(char **cmdline_p) #endif saved_video_mode = boot_params.hdr.vid_mode; bootloader_type = boot_params.hdr.type_of_loader; + if ((bootloader_type >> 4) == 0xe) { + bootloader_type &= 0xf; + bootloader_type |= (boot_params.hdr.ext_loader_type+0x10) << 4; + } + bootloader_version = bootloader_type & 0xf; + bootloader_version |= boot_params.hdr.ext_loader_ver << 4; #ifdef CONFIG_BLK_DEV_RAM rd_image_start = boot_params.hdr.ram_size & RAMDISK_IMAGE_START_MASK; @@ -854,12 +868,16 @@ void __init setup_arch(char **cmdline_p) max_low_pfn = max_pfn; high_memory = (void *)__va(max_pfn * PAGE_SIZE - 1) + 1; + max_pfn_mapped = KERNEL_IMAGE_SIZE >> PAGE_SHIFT; #endif #ifdef CONFIG_X86_CHECK_BIOS_CORRUPTION setup_bios_corruption_check(); #endif + printk(KERN_DEBUG "initial memory mapped : 0 - %08lx\n", + max_pfn_mapped<<PAGE_SHIFT); + reserve_brk(); /* max_pfn_mapped is updated here */ @@ -997,24 +1015,6 @@ void __init setup_arch(char **cmdline_p) #ifdef CONFIG_X86_32 /** - * x86_quirk_pre_intr_init - initialisation prior to setting up interrupt vectors - * - * Description: - * Perform any necessary interrupt initialisation prior to setting up - * the "ordinary" interrupt call gates. For legacy reasons, the ISA - * interrupts should be initialised here if the machine emulates a PC - * in any way. - **/ -void __init x86_quirk_pre_intr_init(void) -{ - if (x86_quirks->arch_pre_intr_init) { - if (x86_quirks->arch_pre_intr_init()) - return; - } - init_ISA_irqs(); -} - -/** * x86_quirk_intr_init - post gate setup interrupt initialisation * * Description: diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 8f0e13be36b..9c3f0823e6a 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -425,6 +425,14 @@ void __init setup_per_cpu_areas(void) early_per_cpu_ptr(x86_cpu_to_node_map) = NULL; #endif +#if defined(CONFIG_X86_64) && defined(CONFIG_NUMA) + /* + * make sure boot cpu node_number is right, when boot cpu is on the + * node that doesn't have mem installed + */ + per_cpu(node_number, boot_cpu_id) = cpu_to_node(boot_cpu_id); +#endif + /* Setup node to cpumask map */ setup_node_to_cpumask_map(); diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c index 14425166b8e..0a813b17b17 100644 --- a/arch/x86/kernel/signal.c +++ b/arch/x86/kernel/signal.c @@ -6,7 +6,6 @@ * 2000-06-20 Pentium III FXSR, SSE support by Gareth Hughes * 2000-2002 x86-64 support by Andi Kleen */ - #include <linux/sched.h> #include <linux/mm.h> #include <linux/smp.h> diff --git a/arch/x86/kernel/smp.c b/arch/x86/kernel/smp.c index 13f33ea8cca..f6db48c405b 100644 --- a/arch/x86/kernel/smp.c +++ b/arch/x86/kernel/smp.c @@ -193,19 +193,19 @@ void smp_call_function_single_interrupt(struct pt_regs *regs) } struct smp_ops smp_ops = { - .smp_prepare_boot_cpu = native_smp_prepare_boot_cpu, - .smp_prepare_cpus = native_smp_prepare_cpus, - .smp_cpus_done = native_smp_cpus_done, + .smp_prepare_boot_cpu = native_smp_prepare_boot_cpu, + .smp_prepare_cpus = native_smp_prepare_cpus, + .smp_cpus_done = native_smp_cpus_done, - .smp_send_stop = native_smp_send_stop, - .smp_send_reschedule = native_smp_send_reschedule, + .smp_send_stop = native_smp_send_stop, + .smp_send_reschedule = native_smp_send_reschedule, - .cpu_up = native_cpu_up, - .cpu_die = native_cpu_die, - .cpu_disable = native_cpu_disable, - .play_dead = native_play_dead, + .cpu_up = native_cpu_up, + .cpu_die = native_cpu_die, + .cpu_disable = native_cpu_disable, + .play_dead = native_play_dead, - .send_call_func_ipi = native_send_call_func_ipi, + .send_call_func_ipi = native_send_call_func_ipi, .send_call_func_single_ipi = native_send_call_func_single_ipi, }; EXPORT_SYMBOL_GPL(smp_ops); diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 58d24ef917d..7c80007ea5f 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -504,7 +504,7 @@ void __inquire_remote_apic(int apicid) * INIT, INIT, STARTUP sequence will reset the chip hard for us, and this * won't ... remember to clear down the APIC, etc later. */ -int __devinit +int __cpuinit wakeup_secondary_cpu_via_nmi(int logical_apicid, unsigned long start_eip) { unsigned long send_status, accept_status = 0; @@ -538,7 +538,7 @@ wakeup_secondary_cpu_via_nmi(int logical_apicid, unsigned long start_eip) return (send_status | accept_status); } -int __devinit +static int __cpuinit wakeup_secondary_cpu_via_init(int phys_apicid, unsigned long start_eip) { unsigned long send_status, accept_status = 0; @@ -822,10 +822,12 @@ do_rest: /* mark "stuck" area as not stuck */ *((volatile unsigned long *)trampoline_base) = 0; - /* - * Cleanup possible dangling ends... - */ - smpboot_restore_warm_reset_vector(); + if (get_uv_system_type() != UV_NON_UNIQUE_APIC) { + /* + * Cleanup possible dangling ends... + */ + smpboot_restore_warm_reset_vector(); + } return boot_error; } @@ -990,10 +992,12 @@ static int __init smp_sanity_check(unsigned max_cpus) */ if (APIC_INTEGRATED(apic_version[boot_cpu_physical_apicid]) && !cpu_has_apic) { - printk(KERN_ERR "BIOS bug, local APIC #%d not detected!...\n", - boot_cpu_physical_apicid); - printk(KERN_ERR "... forcing use of dummy APIC emulation." + if (!disable_apic) { + pr_err("BIOS bug, local APIC #%d not detected!...\n", + boot_cpu_physical_apicid); + pr_err("... forcing use of dummy APIC emulation." "(tell your hw vendor)\n"); + } smpboot_clear_io_apic(); arch_disable_smp_support(); return -1; diff --git a/arch/x86/kernel/stacktrace.c b/arch/x86/kernel/stacktrace.c index f7bddc2e37d..4aaf7e48394 100644 --- a/arch/x86/kernel/stacktrace.c +++ b/arch/x86/kernel/stacktrace.c @@ -20,7 +20,7 @@ save_stack_warning_symbol(void *data, char *msg, unsigned long symbol) static int save_stack_stack(void *data, char *name) { - return -1; + return 0; } static void save_stack_address(void *data, unsigned long addr, int reliable) diff --git a/arch/x86/kernel/syscall_table_32.S b/arch/x86/kernel/syscall_table_32.S index ff5c8736b49..d51321ddafd 100644 --- a/arch/x86/kernel/syscall_table_32.S +++ b/arch/x86/kernel/syscall_table_32.S @@ -334,3 +334,5 @@ ENTRY(sys_call_table) .long sys_inotify_init1 .long sys_preadv .long sys_pwritev + .long sys_rt_tgsigqueueinfo /* 335 */ + .long sys_perf_counter_open diff --git a/arch/x86/kernel/tlb_uv.c b/arch/x86/kernel/tlb_uv.c index 8c7b03b0cfc..124d40c575d 100644 --- a/arch/x86/kernel/tlb_uv.c +++ b/arch/x86/kernel/tlb_uv.c @@ -715,7 +715,12 @@ uv_activation_descriptor_init(int node, int pnode) struct bau_desc *adp; struct bau_desc *ad2; - adp = (struct bau_desc *)kmalloc_node(16384, GFP_KERNEL, node); + /* + * each bau_desc is 64 bytes; there are 8 (UV_ITEMS_PER_DESCRIPTOR) + * per cpu; and up to 32 (UV_ADP_SIZE) cpu's per blade + */ + adp = (struct bau_desc *)kmalloc_node(sizeof(struct bau_desc)* + UV_ADP_SIZE*UV_ITEMS_PER_DESCRIPTOR, GFP_KERNEL, node); BUG_ON(!adp); pa = uv_gpa(adp); /* need the real nasid*/ @@ -729,7 +734,13 @@ uv_activation_descriptor_init(int node, int pnode) (n << UV_DESC_BASE_PNODE_SHIFT | m)); } - for (i = 0, ad2 = adp; i < UV_ACTIVATION_DESCRIPTOR_SIZE; i++, ad2++) { + /* + * initializing all 8 (UV_ITEMS_PER_DESCRIPTOR) descriptors for each + * cpu even though we only use the first one; one descriptor can + * describe a broadcast to 256 nodes. + */ + for (i = 0, ad2 = adp; i < (UV_ADP_SIZE*UV_ITEMS_PER_DESCRIPTOR); + i++, ad2++) { memset(ad2, 0, sizeof(struct bau_desc)); ad2->header.sw_ack_flag = 1; /* diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index a1d288327ff..07d60c870ce 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -839,9 +839,6 @@ asmlinkage void math_state_restore(void) } clts(); /* Allow maths ops (or we recurse) */ -#ifdef CONFIG_X86_32 - restore_fpu(tsk); -#else /* * Paranoid restore. send a SIGSEGV if we fail to restore the state. */ @@ -850,7 +847,7 @@ asmlinkage void math_state_restore(void) force_sig(SIGSEGV, tsk); return; } -#endif + thread->status |= TS_USEDFPU; /* So we fnsave on switch_to() */ tsk->fpu_counter++; } @@ -945,8 +942,13 @@ void __init trap_init(void) #endif set_intr_gate(19, &simd_coprocessor_error); + /* Reserve all the builtin and the syscall vector: */ + for (i = 0; i < FIRST_EXTERNAL_VECTOR; i++) + set_bit(i, used_vectors); + #ifdef CONFIG_IA32_EMULATION set_system_intr_gate(IA32_SYSCALL_VECTOR, ia32_syscall); + set_bit(IA32_SYSCALL_VECTOR, used_vectors); #endif #ifdef CONFIG_X86_32 @@ -963,17 +965,9 @@ void __init trap_init(void) } set_system_trap_gate(SYSCALL_VECTOR, &system_call); -#endif - - /* Reserve all the builtin and the syscall vector: */ - for (i = 0; i < FIRST_EXTERNAL_VECTOR; i++) - set_bit(i, used_vectors); - -#ifdef CONFIG_X86_64 - set_bit(IA32_SYSCALL_VECTOR, used_vectors); -#else set_bit(SYSCALL_VECTOR, used_vectors); #endif + /* * Should be a barrier for any external CPU state: */ diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index d57de05dc43..3e1c057e98f 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -384,13 +384,13 @@ unsigned long native_calibrate_tsc(void) { u64 tsc1, tsc2, delta, ref1, ref2; unsigned long tsc_pit_min = ULONG_MAX, tsc_ref_min = ULONG_MAX; - unsigned long flags, latch, ms, fast_calibrate, tsc_khz; + unsigned long flags, latch, ms, fast_calibrate, hv_tsc_khz; int hpet = is_hpet_enabled(), i, loopmin; - tsc_khz = get_hypervisor_tsc_freq(); - if (tsc_khz) { + hv_tsc_khz = get_hypervisor_tsc_freq(); + if (hv_tsc_khz) { printk(KERN_INFO "TSC: Frequency read from the hypervisor\n"); - return tsc_khz; + return hv_tsc_khz; } local_irq_save(flags); @@ -710,7 +710,16 @@ static cycle_t read_tsc(struct clocksource *cs) #ifdef CONFIG_X86_64 static cycle_t __vsyscall_fn vread_tsc(void) { - cycle_t ret = (cycle_t)vget_cycles(); + cycle_t ret; + + /* + * Surround the RDTSC by barriers, to make sure it's not + * speculated to outside the seqlock critical section and + * does not cause time warps: + */ + rdtsc_barrier(); + ret = (cycle_t)vget_cycles(); + rdtsc_barrier(); return ret >= __vsyscall_gtod_data.clock.cycle_last ? ret : __vsyscall_gtod_data.clock.cycle_last; diff --git a/arch/x86/kernel/tsc_sync.c b/arch/x86/kernel/tsc_sync.c index bf36328f6ef..027b5b49899 100644 --- a/arch/x86/kernel/tsc_sync.c +++ b/arch/x86/kernel/tsc_sync.c @@ -34,6 +34,7 @@ static __cpuinitdata atomic_t stop_count; * of a critical section, to be able to prove TSC time-warps: */ static __cpuinitdata raw_spinlock_t sync_lock = __RAW_SPIN_LOCK_UNLOCKED; + static __cpuinitdata cycles_t last_tsc; static __cpuinitdata cycles_t max_warp; static __cpuinitdata int nr_warps; @@ -113,13 +114,12 @@ void __cpuinit check_tsc_sync_source(int cpu) return; if (boot_cpu_has(X86_FEATURE_TSC_RELIABLE)) { - printk(KERN_INFO - "Skipping synchronization checks as TSC is reliable.\n"); + pr_info("Skipping synchronization checks as TSC is reliable.\n"); return; } - printk(KERN_INFO "checking TSC synchronization [CPU#%d -> CPU#%d]:", - smp_processor_id(), cpu); + pr_info("checking TSC synchronization [CPU#%d -> CPU#%d]:", + smp_processor_id(), cpu); /* * Reset it - in case this is a second bootup: @@ -143,8 +143,8 @@ void __cpuinit check_tsc_sync_source(int cpu) if (nr_warps) { printk("\n"); - printk(KERN_WARNING "Measured %Ld cycles TSC warp between CPUs," - " turning off TSC clock.\n", max_warp); + pr_warning("Measured %Ld cycles TSC warp between CPUs, " + "turning off TSC clock.\n", max_warp); mark_tsc_unstable("check_tsc_sync_source failed"); } else { printk(" passed.\n"); @@ -195,5 +195,3 @@ void __cpuinit check_tsc_sync_target(void) while (atomic_read(&stop_count) != cpus) cpu_relax(); } -#undef NR_LOOPS - diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c index d7ac84e7fc1..9c4e6253905 100644 --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -287,10 +287,9 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk info->regs.pt.ds = 0; info->regs.pt.es = 0; info->regs.pt.fs = 0; - -/* we are clearing gs later just before "jmp resume_userspace", - * because it is not saved/restored. - */ +#ifndef CONFIG_X86_32_LAZY_GS + info->regs.pt.gs = 0; +#endif /* * The flags register is also special: we cannot trust that the user @@ -318,9 +317,9 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk } /* - * Save old state, set default return value (%ax) to 0 + * Save old state, set default return value (%ax) to 0 (VM86_SIGNAL) */ - info->regs32->ax = 0; + info->regs32->ax = VM86_SIGNAL; tsk->thread.saved_sp0 = tsk->thread.sp0; tsk->thread.saved_fs = info->regs32->fs; tsk->thread.saved_gs = get_user_gs(info->regs32); @@ -343,7 +342,9 @@ static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk __asm__ __volatile__( "movl %0,%%esp\n\t" "movl %1,%%ebp\n\t" +#ifdef CONFIG_X86_32_LAZY_GS "mov %2, %%gs\n\t" +#endif "jmp resume_userspace" : /* no outputs */ :"r" (&info->regs), "r" (task_thread_info(tsk)), "r" (0)); diff --git a/arch/x86/kernel/vmi_32.c b/arch/x86/kernel/vmi_32.c index 95deb9f2211..b263423fbe2 100644 --- a/arch/x86/kernel/vmi_32.c +++ b/arch/x86/kernel/vmi_32.c @@ -462,22 +462,28 @@ vmi_startup_ipi_hook(int phys_apicid, unsigned long start_eip, } #endif -static void vmi_enter_lazy_cpu(void) +static void vmi_start_context_switch(struct task_struct *prev) { - paravirt_enter_lazy_cpu(); + paravirt_start_context_switch(prev); vmi_ops.set_lazy_mode(2); } +static void vmi_end_context_switch(struct task_struct *next) +{ + vmi_ops.set_lazy_mode(0); + paravirt_end_context_switch(next); +} + static void vmi_enter_lazy_mmu(void) { paravirt_enter_lazy_mmu(); vmi_ops.set_lazy_mode(1); } -static void vmi_leave_lazy(void) +static void vmi_leave_lazy_mmu(void) { - paravirt_leave_lazy(paravirt_get_lazy_mode()); vmi_ops.set_lazy_mode(0); + paravirt_leave_lazy_mmu(); } static inline int __init check_vmi_rom(struct vrom_header *rom) @@ -711,14 +717,14 @@ static inline int __init activate_vmi(void) para_fill(pv_cpu_ops.set_iopl_mask, SetIOPLMask); para_fill(pv_cpu_ops.io_delay, IODelay); - para_wrap(pv_cpu_ops.lazy_mode.enter, vmi_enter_lazy_cpu, + para_wrap(pv_cpu_ops.start_context_switch, vmi_start_context_switch, set_lazy_mode, SetLazyMode); - para_wrap(pv_cpu_ops.lazy_mode.leave, vmi_leave_lazy, + para_wrap(pv_cpu_ops.end_context_switch, vmi_end_context_switch, set_lazy_mode, SetLazyMode); para_wrap(pv_mmu_ops.lazy_mode.enter, vmi_enter_lazy_mmu, set_lazy_mode, SetLazyMode); - para_wrap(pv_mmu_ops.lazy_mode.leave, vmi_leave_lazy, + para_wrap(pv_mmu_ops.lazy_mode.leave, vmi_leave_lazy_mmu, set_lazy_mode, SetLazyMode); /* user and kernel flush are just handled with different flags to FlushTLB */ diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S index 849ee611f01..4c85b2e2bb6 100644 --- a/arch/x86/kernel/vmlinux.lds.S +++ b/arch/x86/kernel/vmlinux.lds.S @@ -1,5 +1,431 @@ +/* + * ld script for the x86 kernel + * + * Historic 32-bit version written by Martin Mares <mj@atrey.karlin.mff.cuni.cz> + * + * Modernisation, unification and other changes and fixes: + * Copyright (C) 2007-2009 Sam Ravnborg <sam@ravnborg.org> + * + * + * Don't define absolute symbols until and unless you know that symbol + * value is should remain constant even if kernel image is relocated + * at run time. Absolute symbols are not relocated. If symbol value should + * change if kernel is relocated, make the symbol section relative and + * put it inside the section definition. + */ + #ifdef CONFIG_X86_32 -# include "vmlinux_32.lds.S" +#define LOAD_OFFSET __PAGE_OFFSET #else -# include "vmlinux_64.lds.S" +#define LOAD_OFFSET __START_KERNEL_map #endif + +#include <asm-generic/vmlinux.lds.h> +#include <asm/asm-offsets.h> +#include <asm/thread_info.h> +#include <asm/page_types.h> +#include <asm/cache.h> +#include <asm/boot.h> + +#undef i386 /* in case the preprocessor is a 32bit one */ + +OUTPUT_FORMAT(CONFIG_OUTPUT_FORMAT, CONFIG_OUTPUT_FORMAT, CONFIG_OUTPUT_FORMAT) + +#ifdef CONFIG_X86_32 +OUTPUT_ARCH(i386) +ENTRY(phys_startup_32) +jiffies = jiffies_64; +#else +OUTPUT_ARCH(i386:x86-64) +ENTRY(phys_startup_64) +jiffies_64 = jiffies; +#endif + +PHDRS { + text PT_LOAD FLAGS(5); /* R_E */ + data PT_LOAD FLAGS(7); /* RWE */ +#ifdef CONFIG_X86_64 + user PT_LOAD FLAGS(7); /* RWE */ + data.init PT_LOAD FLAGS(7); /* RWE */ +#ifdef CONFIG_SMP + percpu PT_LOAD FLAGS(7); /* RWE */ +#endif + data.init2 PT_LOAD FLAGS(7); /* RWE */ +#endif + note PT_NOTE FLAGS(0); /* ___ */ +} + +SECTIONS +{ +#ifdef CONFIG_X86_32 + . = LOAD_OFFSET + LOAD_PHYSICAL_ADDR; + phys_startup_32 = startup_32 - LOAD_OFFSET; +#else + . = __START_KERNEL; + phys_startup_64 = startup_64 - LOAD_OFFSET; +#endif + + /* Text and read-only data */ + + /* bootstrapping code */ + .text.head : AT(ADDR(.text.head) - LOAD_OFFSET) { + _text = .; + *(.text.head) + } :text = 0x9090 + + /* The rest of the text */ + .text : AT(ADDR(.text) - LOAD_OFFSET) { +#ifdef CONFIG_X86_32 + /* not really needed, already page aligned */ + . = ALIGN(PAGE_SIZE); + *(.text.page_aligned) +#endif + . = ALIGN(8); + _stext = .; + TEXT_TEXT + SCHED_TEXT + LOCK_TEXT + KPROBES_TEXT + IRQENTRY_TEXT + *(.fixup) + *(.gnu.warning) + /* End of text section */ + _etext = .; + } :text = 0x9090 + + NOTES :text :note + + /* Exception table */ + . = ALIGN(16); + __ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { + __start___ex_table = .; + *(__ex_table) + __stop___ex_table = .; + } :text = 0x9090 + + RODATA + + /* Data */ + . = ALIGN(PAGE_SIZE); + .data : AT(ADDR(.data) - LOAD_OFFSET) { + DATA_DATA + CONSTRUCTORS + +#ifdef CONFIG_X86_64 + /* End of data section */ + _edata = .; +#endif + } :data + +#ifdef CONFIG_X86_32 + /* 32 bit has nosave before _edata */ + . = ALIGN(PAGE_SIZE); + .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { + __nosave_begin = .; + *(.data.nosave) + . = ALIGN(PAGE_SIZE); + __nosave_end = .; + } +#endif + + . = ALIGN(PAGE_SIZE); + .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { + *(.data.page_aligned) + *(.data.idt) + } + +#ifdef CONFIG_X86_32 + . = ALIGN(32); +#else + . = ALIGN(PAGE_SIZE); + . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); +#endif + .data.cacheline_aligned : + AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { + *(.data.cacheline_aligned) + } + + /* rarely changed data like cpu maps */ +#ifdef CONFIG_X86_32 + . = ALIGN(32); +#else + . = ALIGN(CONFIG_X86_INTERNODE_CACHE_BYTES); +#endif + .data.read_mostly : AT(ADDR(.data.read_mostly) - LOAD_OFFSET) { + *(.data.read_mostly) + +#ifdef CONFIG_X86_32 + /* End of data section */ + _edata = .; +#endif + } + +#ifdef CONFIG_X86_64 + +#define VSYSCALL_ADDR (-10*1024*1024) +#define VSYSCALL_PHYS_ADDR ((LOADADDR(.data.read_mostly) + \ + SIZEOF(.data.read_mostly) + 4095) & ~(4095)) +#define VSYSCALL_VIRT_ADDR ((ADDR(.data.read_mostly) + \ + SIZEOF(.data.read_mostly) + 4095) & ~(4095)) + +#define VLOAD_OFFSET (VSYSCALL_ADDR - VSYSCALL_PHYS_ADDR) +#define VLOAD(x) (ADDR(x) - VLOAD_OFFSET) + +#define VVIRT_OFFSET (VSYSCALL_ADDR - VSYSCALL_VIRT_ADDR) +#define VVIRT(x) (ADDR(x) - VVIRT_OFFSET) + + . = VSYSCALL_ADDR; + .vsyscall_0 : AT(VSYSCALL_PHYS_ADDR) { + *(.vsyscall_0) + } :user + + __vsyscall_0 = VSYSCALL_VIRT_ADDR; + + . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); + .vsyscall_fn : AT(VLOAD(.vsyscall_fn)) { + *(.vsyscall_fn) + } + + . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); + .vsyscall_gtod_data : AT(VLOAD(.vsyscall_gtod_data)) { + *(.vsyscall_gtod_data) + } + + vsyscall_gtod_data = VVIRT(.vsyscall_gtod_data); + .vsyscall_clock : AT(VLOAD(.vsyscall_clock)) { + *(.vsyscall_clock) + } + vsyscall_clock = VVIRT(.vsyscall_clock); + + + .vsyscall_1 ADDR(.vsyscall_0) + 1024: AT(VLOAD(.vsyscall_1)) { + *(.vsyscall_1) + } + .vsyscall_2 ADDR(.vsyscall_0) + 2048: AT(VLOAD(.vsyscall_2)) { + *(.vsyscall_2) + } + + .vgetcpu_mode : AT(VLOAD(.vgetcpu_mode)) { + *(.vgetcpu_mode) + } + vgetcpu_mode = VVIRT(.vgetcpu_mode); + + . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); + .jiffies : AT(VLOAD(.jiffies)) { + *(.jiffies) + } + jiffies = VVIRT(.jiffies); + + .vsyscall_3 ADDR(.vsyscall_0) + 3072: AT(VLOAD(.vsyscall_3)) { + *(.vsyscall_3) + } + + . = VSYSCALL_VIRT_ADDR + PAGE_SIZE; + +#undef VSYSCALL_ADDR +#undef VSYSCALL_PHYS_ADDR +#undef VSYSCALL_VIRT_ADDR +#undef VLOAD_OFFSET +#undef VLOAD +#undef VVIRT_OFFSET +#undef VVIRT + +#endif /* CONFIG_X86_64 */ + + /* init_task */ + . = ALIGN(THREAD_SIZE); + .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { + *(.data.init_task) + } +#ifdef CONFIG_X86_64 + :data.init +#endif + + /* + * smp_locks might be freed after init + * start/end must be page aligned + */ + . = ALIGN(PAGE_SIZE); + .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) { + __smp_locks = .; + *(.smp_locks) + __smp_locks_end = .; + . = ALIGN(PAGE_SIZE); + } + + /* Init code and data - will be freed after init */ + . = ALIGN(PAGE_SIZE); + .init.text : AT(ADDR(.init.text) - LOAD_OFFSET) { + __init_begin = .; /* paired with __init_end */ + _sinittext = .; + INIT_TEXT + _einittext = .; + } + + .init.data : AT(ADDR(.init.data) - LOAD_OFFSET) { + INIT_DATA + } + + . = ALIGN(16); + .init.setup : AT(ADDR(.init.setup) - LOAD_OFFSET) { + __setup_start = .; + *(.init.setup) + __setup_end = .; + } + .initcall.init : AT(ADDR(.initcall.init) - LOAD_OFFSET) { + __initcall_start = .; + INITCALLS + __initcall_end = .; + } + + .con_initcall.init : AT(ADDR(.con_initcall.init) - LOAD_OFFSET) { + __con_initcall_start = .; + *(.con_initcall.init) + __con_initcall_end = .; + } + + .x86_cpu_dev.init : AT(ADDR(.x86_cpu_dev.init) - LOAD_OFFSET) { + __x86_cpu_dev_start = .; + *(.x86_cpu_dev.init) + __x86_cpu_dev_end = .; + } + + SECURITY_INIT + + . = ALIGN(8); + .parainstructions : AT(ADDR(.parainstructions) - LOAD_OFFSET) { + __parainstructions = .; + *(.parainstructions) + __parainstructions_end = .; + } + + . = ALIGN(8); + .altinstructions : AT(ADDR(.altinstructions) - LOAD_OFFSET) { + __alt_instructions = .; + *(.altinstructions) + __alt_instructions_end = .; + } + + .altinstr_replacement : AT(ADDR(.altinstr_replacement) - LOAD_OFFSET) { + *(.altinstr_replacement) + } + + /* + * .exit.text is discard at runtime, not link time, to deal with + * references from .altinstructions and .eh_frame + */ + .exit.text : AT(ADDR(.exit.text) - LOAD_OFFSET) { + EXIT_TEXT + } + + .exit.data : AT(ADDR(.exit.data) - LOAD_OFFSET) { + EXIT_DATA + } + +#ifdef CONFIG_BLK_DEV_INITRD + . = ALIGN(PAGE_SIZE); + .init.ramfs : AT(ADDR(.init.ramfs) - LOAD_OFFSET) { + __initramfs_start = .; + *(.init.ramfs) + __initramfs_end = .; + } +#endif + +#if defined(CONFIG_X86_64) && defined(CONFIG_SMP) + /* + * percpu offsets are zero-based on SMP. PERCPU_VADDR() changes the + * output PHDR, so the next output section - __data_nosave - should + * start another section data.init2. Also, pda should be at the head of + * percpu area. Preallocate it and define the percpu offset symbol + * so that it can be accessed as a percpu variable. + */ + . = ALIGN(PAGE_SIZE); + PERCPU_VADDR(0, :percpu) +#else + PERCPU(PAGE_SIZE) +#endif + + . = ALIGN(PAGE_SIZE); + + /* freed after init ends here */ + .init.end : AT(ADDR(.init.end) - LOAD_OFFSET) { + __init_end = .; + } + +#ifdef CONFIG_X86_64 + .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { + . = ALIGN(PAGE_SIZE); + __nosave_begin = .; + *(.data.nosave) + . = ALIGN(PAGE_SIZE); + __nosave_end = .; + } :data.init2 + /* use another section data.init2, see PERCPU_VADDR() above */ +#endif + + /* BSS */ + . = ALIGN(PAGE_SIZE); + .bss : AT(ADDR(.bss) - LOAD_OFFSET) { + __bss_start = .; + *(.bss.page_aligned) + *(.bss) + . = ALIGN(4); + __bss_stop = .; + } + + . = ALIGN(PAGE_SIZE); + .brk : AT(ADDR(.brk) - LOAD_OFFSET) { + __brk_base = .; + . += 64 * 1024; /* 64k alignment slop space */ + *(.brk_reservation) /* areas brk users have reserved */ + __brk_limit = .; + } + + .end : AT(ADDR(.end) - LOAD_OFFSET) { + _end = .; + } + + /* Sections to be discarded */ + /DISCARD/ : { + *(.exitcall.exit) + *(.eh_frame) + *(.discard) + } + + STABS_DEBUG + DWARF_DEBUG +} + + +#ifdef CONFIG_X86_32 +ASSERT((_end - LOAD_OFFSET <= KERNEL_IMAGE_SIZE), + "kernel image bigger than KERNEL_IMAGE_SIZE") +#else +/* + * Per-cpu symbols which need to be offset from __per_cpu_load + * for the boot processor. + */ +#define INIT_PER_CPU(x) init_per_cpu__##x = per_cpu__##x + __per_cpu_load +INIT_PER_CPU(gdt_page); +INIT_PER_CPU(irq_stack_union); + +/* + * Build-time check on the image size: + */ +ASSERT((_end - _text <= KERNEL_IMAGE_SIZE), + "kernel image bigger than KERNEL_IMAGE_SIZE") + +#ifdef CONFIG_SMP +ASSERT((per_cpu__irq_stack_union == 0), + "irq_stack_union is not at start of per-cpu area"); +#endif + +#endif /* CONFIG_X86_32 */ + +#ifdef CONFIG_KEXEC +#include <asm/kexec.h> + +ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE, + "kexec control code size is too big") +#endif + diff --git a/arch/x86/kernel/vmlinux_32.lds.S b/arch/x86/kernel/vmlinux_32.lds.S deleted file mode 100644 index 62ad500d55f..00000000000 --- a/arch/x86/kernel/vmlinux_32.lds.S +++ /dev/null @@ -1,229 +0,0 @@ -/* ld script to make i386 Linux kernel - * Written by Martin Mares <mj@atrey.karlin.mff.cuni.cz>; - * - * Don't define absolute symbols until and unless you know that symbol - * value is should remain constant even if kernel image is relocated - * at run time. Absolute symbols are not relocated. If symbol value should - * change if kernel is relocated, make the symbol section relative and - * put it inside the section definition. - */ - -#define LOAD_OFFSET __PAGE_OFFSET - -#include <asm-generic/vmlinux.lds.h> -#include <asm/thread_info.h> -#include <asm/page_types.h> -#include <asm/cache.h> -#include <asm/boot.h> - -OUTPUT_FORMAT("elf32-i386", "elf32-i386", "elf32-i386") -OUTPUT_ARCH(i386) -ENTRY(phys_startup_32) -jiffies = jiffies_64; - -PHDRS { - text PT_LOAD FLAGS(5); /* R_E */ - data PT_LOAD FLAGS(7); /* RWE */ - note PT_NOTE FLAGS(0); /* ___ */ -} -SECTIONS -{ - . = LOAD_OFFSET + LOAD_PHYSICAL_ADDR; - phys_startup_32 = startup_32 - LOAD_OFFSET; - - .text.head : AT(ADDR(.text.head) - LOAD_OFFSET) { - _text = .; /* Text and read-only data */ - *(.text.head) - } :text = 0x9090 - - /* read-only */ - .text : AT(ADDR(.text) - LOAD_OFFSET) { - . = ALIGN(PAGE_SIZE); /* not really needed, already page aligned */ - *(.text.page_aligned) - TEXT_TEXT - SCHED_TEXT - LOCK_TEXT - KPROBES_TEXT - IRQENTRY_TEXT - *(.fixup) - *(.gnu.warning) - _etext = .; /* End of text section */ - } :text = 0x9090 - - NOTES :text :note - - . = ALIGN(16); /* Exception table */ - __ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { - __start___ex_table = .; - *(__ex_table) - __stop___ex_table = .; - } :text = 0x9090 - - RODATA - - /* writeable */ - . = ALIGN(PAGE_SIZE); - .data : AT(ADDR(.data) - LOAD_OFFSET) { /* Data */ - DATA_DATA - CONSTRUCTORS - } :data - - . = ALIGN(PAGE_SIZE); - .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { - __nosave_begin = .; - *(.data.nosave) - . = ALIGN(PAGE_SIZE); - __nosave_end = .; - } - - . = ALIGN(PAGE_SIZE); - .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { - *(.data.page_aligned) - *(.data.idt) - } - - . = ALIGN(32); - .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { - *(.data.cacheline_aligned) - } - - /* rarely changed data like cpu maps */ - . = ALIGN(32); - .data.read_mostly : AT(ADDR(.data.read_mostly) - LOAD_OFFSET) { - *(.data.read_mostly) - _edata = .; /* End of data section */ - } - - . = ALIGN(THREAD_SIZE); /* init_task */ - .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { - *(.data.init_task) - } - - /* might get freed after init */ - . = ALIGN(PAGE_SIZE); - .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) { - __smp_locks = .; - *(.smp_locks) - __smp_locks_end = .; - } - /* will be freed after init - * Following ALIGN() is required to make sure no other data falls on the - * same page where __smp_alt_end is pointing as that page might be freed - * after boot. Always make sure that ALIGN() directive is present after - * the section which contains __smp_alt_end. - */ - . = ALIGN(PAGE_SIZE); - - /* will be freed after init */ - . = ALIGN(PAGE_SIZE); /* Init code and data */ - .init.text : AT(ADDR(.init.text) - LOAD_OFFSET) { - __init_begin = .; - _sinittext = .; - INIT_TEXT - _einittext = .; - } - .init.data : AT(ADDR(.init.data) - LOAD_OFFSET) { - INIT_DATA - } - . = ALIGN(16); - .init.setup : AT(ADDR(.init.setup) - LOAD_OFFSET) { - __setup_start = .; - *(.init.setup) - __setup_end = .; - } - .initcall.init : AT(ADDR(.initcall.init) - LOAD_OFFSET) { - __initcall_start = .; - INITCALLS - __initcall_end = .; - } - .con_initcall.init : AT(ADDR(.con_initcall.init) - LOAD_OFFSET) { - __con_initcall_start = .; - *(.con_initcall.init) - __con_initcall_end = .; - } - .x86_cpu_dev.init : AT(ADDR(.x86_cpu_dev.init) - LOAD_OFFSET) { - __x86_cpu_dev_start = .; - *(.x86_cpu_dev.init) - __x86_cpu_dev_end = .; - } - SECURITY_INIT - . = ALIGN(4); - .altinstructions : AT(ADDR(.altinstructions) - LOAD_OFFSET) { - __alt_instructions = .; - *(.altinstructions) - __alt_instructions_end = .; - } - .altinstr_replacement : AT(ADDR(.altinstr_replacement) - LOAD_OFFSET) { - *(.altinstr_replacement) - } - . = ALIGN(4); - .parainstructions : AT(ADDR(.parainstructions) - LOAD_OFFSET) { - __parainstructions = .; - *(.parainstructions) - __parainstructions_end = .; - } - /* .exit.text is discard at runtime, not link time, to deal with references - from .altinstructions and .eh_frame */ - .exit.text : AT(ADDR(.exit.text) - LOAD_OFFSET) { - EXIT_TEXT - } - .exit.data : AT(ADDR(.exit.data) - LOAD_OFFSET) { - EXIT_DATA - } -#if defined(CONFIG_BLK_DEV_INITRD) - . = ALIGN(PAGE_SIZE); - .init.ramfs : AT(ADDR(.init.ramfs) - LOAD_OFFSET) { - __initramfs_start = .; - *(.init.ramfs) - __initramfs_end = .; - } -#endif - PERCPU(PAGE_SIZE) - . = ALIGN(PAGE_SIZE); - /* freed after init ends here */ - - .bss : AT(ADDR(.bss) - LOAD_OFFSET) { - __init_end = .; - __bss_start = .; /* BSS */ - *(.bss.page_aligned) - *(.bss) - . = ALIGN(4); - __bss_stop = .; - } - - .brk : AT(ADDR(.brk) - LOAD_OFFSET) { - . = ALIGN(PAGE_SIZE); - __brk_base = . ; - . += 64 * 1024 ; /* 64k alignment slop space */ - *(.brk_reservation) /* areas brk users have reserved */ - __brk_limit = . ; - } - - .end : AT(ADDR(.end) - LOAD_OFFSET) { - _end = . ; - } - - /* Sections to be discarded */ - /DISCARD/ : { - *(.exitcall.exit) - *(.discard) - } - - STABS_DEBUG - - DWARF_DEBUG -} - -/* - * Build-time check on the image size: - */ -ASSERT((_end - LOAD_OFFSET <= KERNEL_IMAGE_SIZE), - "kernel image bigger than KERNEL_IMAGE_SIZE") - -#ifdef CONFIG_KEXEC -/* Link time checks */ -#include <asm/kexec.h> - -ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE, - "kexec control code size is too big") -#endif diff --git a/arch/x86/kernel/vmlinux_64.lds.S b/arch/x86/kernel/vmlinux_64.lds.S deleted file mode 100644 index c8742507b03..00000000000 --- a/arch/x86/kernel/vmlinux_64.lds.S +++ /dev/null @@ -1,298 +0,0 @@ -/* ld script to make x86-64 Linux kernel - * Written by Martin Mares <mj@atrey.karlin.mff.cuni.cz>; - */ - -#define LOAD_OFFSET __START_KERNEL_map - -#include <asm-generic/vmlinux.lds.h> -#include <asm/asm-offsets.h> -#include <asm/page_types.h> - -#undef i386 /* in case the preprocessor is a 32bit one */ - -OUTPUT_FORMAT("elf64-x86-64", "elf64-x86-64", "elf64-x86-64") -OUTPUT_ARCH(i386:x86-64) -ENTRY(phys_startup_64) -jiffies_64 = jiffies; -PHDRS { - text PT_LOAD FLAGS(5); /* R_E */ - data PT_LOAD FLAGS(7); /* RWE */ - user PT_LOAD FLAGS(7); /* RWE */ - data.init PT_LOAD FLAGS(7); /* RWE */ -#ifdef CONFIG_SMP - percpu PT_LOAD FLAGS(7); /* RWE */ -#endif - data.init2 PT_LOAD FLAGS(7); /* RWE */ - note PT_NOTE FLAGS(0); /* ___ */ -} -SECTIONS -{ - . = __START_KERNEL; - phys_startup_64 = startup_64 - LOAD_OFFSET; - .text : AT(ADDR(.text) - LOAD_OFFSET) { - _text = .; /* Text and read-only data */ - /* First the code that has to be first for bootstrapping */ - *(.text.head) - _stext = .; - /* Then the rest */ - TEXT_TEXT - SCHED_TEXT - LOCK_TEXT - KPROBES_TEXT - IRQENTRY_TEXT - *(.fixup) - *(.gnu.warning) - _etext = .; /* End of text section */ - } :text = 0x9090 - - NOTES :text :note - - . = ALIGN(16); /* Exception table */ - __ex_table : AT(ADDR(__ex_table) - LOAD_OFFSET) { - __start___ex_table = .; - *(__ex_table) - __stop___ex_table = .; - } :text = 0x9090 - - RODATA - - . = ALIGN(PAGE_SIZE); /* Align data segment to page size boundary */ - /* Data */ - .data : AT(ADDR(.data) - LOAD_OFFSET) { - DATA_DATA - CONSTRUCTORS - _edata = .; /* End of data section */ - } :data - - - .data.cacheline_aligned : AT(ADDR(.data.cacheline_aligned) - LOAD_OFFSET) { - . = ALIGN(PAGE_SIZE); - . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); - *(.data.cacheline_aligned) - } - . = ALIGN(CONFIG_X86_INTERNODE_CACHE_BYTES); - .data.read_mostly : AT(ADDR(.data.read_mostly) - LOAD_OFFSET) { - *(.data.read_mostly) - } - -#define VSYSCALL_ADDR (-10*1024*1024) -#define VSYSCALL_PHYS_ADDR ((LOADADDR(.data.read_mostly) + SIZEOF(.data.read_mostly) + 4095) & ~(4095)) -#define VSYSCALL_VIRT_ADDR ((ADDR(.data.read_mostly) + SIZEOF(.data.read_mostly) + 4095) & ~(4095)) - -#define VLOAD_OFFSET (VSYSCALL_ADDR - VSYSCALL_PHYS_ADDR) -#define VLOAD(x) (ADDR(x) - VLOAD_OFFSET) - -#define VVIRT_OFFSET (VSYSCALL_ADDR - VSYSCALL_VIRT_ADDR) -#define VVIRT(x) (ADDR(x) - VVIRT_OFFSET) - - . = VSYSCALL_ADDR; - .vsyscall_0 : AT(VSYSCALL_PHYS_ADDR) { *(.vsyscall_0) } :user - __vsyscall_0 = VSYSCALL_VIRT_ADDR; - - . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); - .vsyscall_fn : AT(VLOAD(.vsyscall_fn)) { *(.vsyscall_fn) } - . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); - .vsyscall_gtod_data : AT(VLOAD(.vsyscall_gtod_data)) - { *(.vsyscall_gtod_data) } - vsyscall_gtod_data = VVIRT(.vsyscall_gtod_data); - .vsyscall_clock : AT(VLOAD(.vsyscall_clock)) - { *(.vsyscall_clock) } - vsyscall_clock = VVIRT(.vsyscall_clock); - - - .vsyscall_1 ADDR(.vsyscall_0) + 1024: AT(VLOAD(.vsyscall_1)) - { *(.vsyscall_1) } - .vsyscall_2 ADDR(.vsyscall_0) + 2048: AT(VLOAD(.vsyscall_2)) - { *(.vsyscall_2) } - - .vgetcpu_mode : AT(VLOAD(.vgetcpu_mode)) { *(.vgetcpu_mode) } - vgetcpu_mode = VVIRT(.vgetcpu_mode); - - . = ALIGN(CONFIG_X86_L1_CACHE_BYTES); - .jiffies : AT(VLOAD(.jiffies)) { *(.jiffies) } - jiffies = VVIRT(.jiffies); - - .vsyscall_3 ADDR(.vsyscall_0) + 3072: AT(VLOAD(.vsyscall_3)) - { *(.vsyscall_3) } - - . = VSYSCALL_VIRT_ADDR + PAGE_SIZE; - -#undef VSYSCALL_ADDR -#undef VSYSCALL_PHYS_ADDR -#undef VSYSCALL_VIRT_ADDR -#undef VLOAD_OFFSET -#undef VLOAD -#undef VVIRT_OFFSET -#undef VVIRT - - .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { - . = ALIGN(THREAD_SIZE); /* init_task */ - *(.data.init_task) - }:data.init - - .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { - . = ALIGN(PAGE_SIZE); - *(.data.page_aligned) - } - - .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) { - /* might get freed after init */ - . = ALIGN(PAGE_SIZE); - __smp_alt_begin = .; - __smp_locks = .; - *(.smp_locks) - __smp_locks_end = .; - . = ALIGN(PAGE_SIZE); - __smp_alt_end = .; - } - - . = ALIGN(PAGE_SIZE); /* Init code and data */ - __init_begin = .; /* paired with __init_end */ - .init.text : AT(ADDR(.init.text) - LOAD_OFFSET) { - _sinittext = .; - INIT_TEXT - _einittext = .; - } - .init.data : AT(ADDR(.init.data) - LOAD_OFFSET) { - __initdata_begin = .; - INIT_DATA - __initdata_end = .; - } - - .init.setup : AT(ADDR(.init.setup) - LOAD_OFFSET) { - . = ALIGN(16); - __setup_start = .; - *(.init.setup) - __setup_end = .; - } - .initcall.init : AT(ADDR(.initcall.init) - LOAD_OFFSET) { - __initcall_start = .; - INITCALLS - __initcall_end = .; - } - .con_initcall.init : AT(ADDR(.con_initcall.init) - LOAD_OFFSET) { - __con_initcall_start = .; - *(.con_initcall.init) - __con_initcall_end = .; - } - .x86_cpu_dev.init : AT(ADDR(.x86_cpu_dev.init) - LOAD_OFFSET) { - __x86_cpu_dev_start = .; - *(.x86_cpu_dev.init) - __x86_cpu_dev_end = .; - } - SECURITY_INIT - - . = ALIGN(8); - .parainstructions : AT(ADDR(.parainstructions) - LOAD_OFFSET) { - __parainstructions = .; - *(.parainstructions) - __parainstructions_end = .; - } - - .altinstructions : AT(ADDR(.altinstructions) - LOAD_OFFSET) { - . = ALIGN(8); - __alt_instructions = .; - *(.altinstructions) - __alt_instructions_end = .; - } - .altinstr_replacement : AT(ADDR(.altinstr_replacement) - LOAD_OFFSET) { - *(.altinstr_replacement) - } - /* .exit.text is discard at runtime, not link time, to deal with references - from .altinstructions and .eh_frame */ - .exit.text : AT(ADDR(.exit.text) - LOAD_OFFSET) { - EXIT_TEXT - } - .exit.data : AT(ADDR(.exit.data) - LOAD_OFFSET) { - EXIT_DATA - } - -#ifdef CONFIG_BLK_DEV_INITRD - . = ALIGN(PAGE_SIZE); - .init.ramfs : AT(ADDR(.init.ramfs) - LOAD_OFFSET) { - __initramfs_start = .; - *(.init.ramfs) - __initramfs_end = .; - } -#endif - -#ifdef CONFIG_SMP - /* - * percpu offsets are zero-based on SMP. PERCPU_VADDR() changes the - * output PHDR, so the next output section - __data_nosave - should - * start another section data.init2. Also, pda should be at the head of - * percpu area. Preallocate it and define the percpu offset symbol - * so that it can be accessed as a percpu variable. - */ - . = ALIGN(PAGE_SIZE); - PERCPU_VADDR(0, :percpu) -#else - PERCPU(PAGE_SIZE) -#endif - - . = ALIGN(PAGE_SIZE); - __init_end = .; - - .data_nosave : AT(ADDR(.data_nosave) - LOAD_OFFSET) { - . = ALIGN(PAGE_SIZE); - __nosave_begin = .; - *(.data.nosave) - . = ALIGN(PAGE_SIZE); - __nosave_end = .; - } :data.init2 /* use another section data.init2, see PERCPU_VADDR() above */ - - .bss : AT(ADDR(.bss) - LOAD_OFFSET) { - . = ALIGN(PAGE_SIZE); - __bss_start = .; /* BSS */ - *(.bss.page_aligned) - *(.bss) - __bss_stop = .; - } - - .brk : AT(ADDR(.brk) - LOAD_OFFSET) { - . = ALIGN(PAGE_SIZE); - __brk_base = . ; - . += 64 * 1024 ; /* 64k alignment slop space */ - *(.brk_reservation) /* areas brk users have reserved */ - __brk_limit = . ; - } - - _end = . ; - - /* Sections to be discarded */ - /DISCARD/ : { - *(.exitcall.exit) - *(.eh_frame) - *(.discard) - } - - STABS_DEBUG - - DWARF_DEBUG -} - - /* - * Per-cpu symbols which need to be offset from __per_cpu_load - * for the boot processor. - */ -#define INIT_PER_CPU(x) init_per_cpu__##x = per_cpu__##x + __per_cpu_load -INIT_PER_CPU(gdt_page); -INIT_PER_CPU(irq_stack_union); - -/* - * Build-time check on the image size: - */ -ASSERT((_end - _text <= KERNEL_IMAGE_SIZE), - "kernel image bigger than KERNEL_IMAGE_SIZE") - -#ifdef CONFIG_SMP -ASSERT((per_cpu__irq_stack_union == 0), - "irq_stack_union is not at start of per-cpu area"); -#endif - -#ifdef CONFIG_KEXEC -#include <asm/kexec.h> - -ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE, - "kexec control code size is too big") -#endif diff --git a/arch/x86/kernel/vsyscall_64.c b/arch/x86/kernel/vsyscall_64.c index 44153afc906..25ee06a80aa 100644 --- a/arch/x86/kernel/vsyscall_64.c +++ b/arch/x86/kernel/vsyscall_64.c @@ -132,15 +132,7 @@ static __always_inline void do_vgettimeofday(struct timeval * tv) return; } - /* - * Surround the RDTSC by barriers, to make sure it's not - * speculated to outside the seqlock critical section and - * does not cause time warps: - */ - rdtsc_barrier(); now = vread(); - rdtsc_barrier(); - base = __vsyscall_gtod_data.clock.cycle_last; mask = __vsyscall_gtod_data.clock.mask; mult = __vsyscall_gtod_data.clock.mult; diff --git a/arch/x86/lguest/boot.c b/arch/x86/lguest/boot.c index 33a93b41739..4e0c2655939 100644 --- a/arch/x86/lguest/boot.c +++ b/arch/x86/lguest/boot.c @@ -167,10 +167,16 @@ static void lazy_hcall3(unsigned long call, /* When lazy mode is turned off reset the per-cpu lazy mode variable and then * issue the do-nothing hypercall to flush any stored calls. */ -static void lguest_leave_lazy_mode(void) +static void lguest_leave_lazy_mmu_mode(void) { - paravirt_leave_lazy(paravirt_get_lazy_mode()); kvm_hypercall0(LHCALL_FLUSH_ASYNC); + paravirt_leave_lazy_mmu(); +} + +static void lguest_end_context_switch(struct task_struct *next) +{ + kvm_hypercall0(LHCALL_FLUSH_ASYNC); + paravirt_end_context_switch(next); } /*G:033 @@ -637,7 +643,7 @@ static void __init lguest_init_IRQ(void) void lguest_setup_irq(unsigned int irq) { - irq_to_desc_alloc_cpu(irq, 0); + irq_to_desc_alloc_node(irq, 0); set_irq_chip_and_handler_name(irq, &lguest_irq_controller, handle_level_irq, "level"); } @@ -1054,8 +1060,8 @@ __init void lguest_init(void) pv_cpu_ops.write_gdt_entry = lguest_write_gdt_entry; pv_cpu_ops.write_idt_entry = lguest_write_idt_entry; pv_cpu_ops.wbinvd = lguest_wbinvd; - pv_cpu_ops.lazy_mode.enter = paravirt_enter_lazy_cpu; - pv_cpu_ops.lazy_mode.leave = lguest_leave_lazy_mode; + pv_cpu_ops.start_context_switch = paravirt_start_context_switch; + pv_cpu_ops.end_context_switch = lguest_end_context_switch; /* pagetable management */ pv_mmu_ops.write_cr3 = lguest_write_cr3; @@ -1068,7 +1074,7 @@ __init void lguest_init(void) pv_mmu_ops.read_cr2 = lguest_read_cr2; pv_mmu_ops.read_cr3 = lguest_read_cr3; pv_mmu_ops.lazy_mode.enter = paravirt_enter_lazy_mmu; - pv_mmu_ops.lazy_mode.leave = lguest_leave_lazy_mode; + pv_mmu_ops.lazy_mode.leave = lguest_leave_lazy_mmu_mode; pv_mmu_ops.pte_update = lguest_pte_update; pv_mmu_ops.pte_update_defer = lguest_pte_update; diff --git a/arch/x86/mm/dump_pagetables.c b/arch/x86/mm/dump_pagetables.c index e7277cbcfb4..a725b7f760a 100644 --- a/arch/x86/mm/dump_pagetables.c +++ b/arch/x86/mm/dump_pagetables.c @@ -161,13 +161,14 @@ static void note_page(struct seq_file *m, struct pg_state *st, st->current_address >= st->marker[1].start_address) { const char *unit = units; unsigned long delta; + int width = sizeof(unsigned long) * 2; /* * Now print the actual finished series */ - seq_printf(m, "0x%p-0x%p ", - (void *)st->start_address, - (void *)st->current_address); + seq_printf(m, "0x%0*lx-0x%0*lx ", + width, st->start_address, + width, st->current_address); delta = (st->current_address - st->start_address) >> 10; while (!(delta & 1023) && unit[1]) { diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index a03b7279efa..c6acc632637 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -3,40 +3,17 @@ * Copyright (C) 2001, 2002 Andi Kleen, SuSE Labs. * Copyright (C) 2008-2009, Red Hat Inc., Ingo Molnar */ -#include <linux/interrupt.h> -#include <linux/mmiotrace.h> -#include <linux/bootmem.h> -#include <linux/compiler.h> -#include <linux/highmem.h> -#include <linux/kprobes.h> -#include <linux/uaccess.h> -#include <linux/vmalloc.h> -#include <linux/vt_kern.h> -#include <linux/signal.h> -#include <linux/kernel.h> -#include <linux/ptrace.h> -#include <linux/string.h> -#include <linux/module.h> -#include <linux/kdebug.h> -#include <linux/errno.h> -#include <linux/magic.h> -#include <linux/sched.h> -#include <linux/types.h> -#include <linux/init.h> -#include <linux/mman.h> -#include <linux/tty.h> -#include <linux/smp.h> -#include <linux/mm.h> - -#include <asm-generic/sections.h> - -#include <asm/tlbflush.h> -#include <asm/pgalloc.h> -#include <asm/segment.h> -#include <asm/system.h> -#include <asm/proto.h> -#include <asm/traps.h> -#include <asm/desc.h> +#include <linux/magic.h> /* STACK_END_MAGIC */ +#include <linux/sched.h> /* test_thread_flag(), ... */ +#include <linux/kdebug.h> /* oops_begin/end, ... */ +#include <linux/module.h> /* search_exception_table */ +#include <linux/bootmem.h> /* max_low_pfn */ +#include <linux/kprobes.h> /* __kprobes, ... */ +#include <linux/mmiotrace.h> /* kmmio_handler, ... */ +#include <linux/perf_counter.h> /* perf_swcounter_event */ + +#include <asm/traps.h> /* dotraplinkage, ... */ +#include <asm/pgalloc.h> /* pgd_*(), ... */ /* * Page fault error code bits: @@ -225,12 +202,10 @@ static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address) if (!pmd_present(*pmd_k)) return NULL; - if (!pmd_present(*pmd)) { + if (!pmd_present(*pmd)) set_pmd(pmd, *pmd_k); - arch_flush_lazy_mmu_mode(); - } else { + else BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k)); - } return pmd_k; } @@ -538,8 +513,6 @@ bad: static int is_errata93(struct pt_regs *regs, unsigned long address) { #ifdef CONFIG_X86_64 - static int once; - if (address != regs->ip) return 0; @@ -549,10 +522,7 @@ static int is_errata93(struct pt_regs *regs, unsigned long address) address |= 0xffffffffUL << 32; if ((address >= (u64)_stext && address <= (u64)_etext) || (address >= MODULES_VADDR && address <= MODULES_END)) { - if (!once) { - printk(errata93_warning); - once = 1; - } + printk_once(errata93_warning); regs->ip = address; return 1; } @@ -1044,6 +1014,8 @@ do_page_fault(struct pt_regs *regs, unsigned long error_code) if (unlikely(error_code & PF_RSVD)) pgtable_bad(regs, error_code, address); + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address); + /* * If we're in an interrupt, have no user context or are running * in an atomic region then we must not take the fault: @@ -1137,10 +1109,15 @@ good_area: return; } - if (fault & VM_FAULT_MAJOR) + if (fault & VM_FAULT_MAJOR) { tsk->maj_flt++; - else + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, 0, + regs, address); + } else { tsk->min_flt++; + perf_swcounter_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, 0, + regs, address); + } check_v8086_mode(regs, address, tsk); diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c index 8126e8d1a2a..58f621e8191 100644 --- a/arch/x86/mm/highmem_32.c +++ b/arch/x86/mm/highmem_32.c @@ -44,7 +44,6 @@ void *kmap_atomic_prot(struct page *page, enum km_type type, pgprot_t prot) vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx); BUG_ON(!pte_none(*(kmap_pte-idx))); set_pte(kmap_pte-idx, mk_pte(page, prot)); - arch_flush_lazy_mmu_mode(); return (void *)vaddr; } @@ -74,7 +73,6 @@ void kunmap_atomic(void *kvaddr, enum km_type type) #endif } - arch_flush_lazy_mmu_mode(); pagefault_enable(); } diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c index ae4f7b5d710..34c1bfb64f1 100644 --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -1,3 +1,4 @@ +#include <linux/initrd.h> #include <linux/ioport.h> #include <linux/swap.h> @@ -10,6 +11,9 @@ #include <asm/setup.h> #include <asm/system.h> #include <asm/tlbflush.h> +#include <asm/tlb.h> + +DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); unsigned long __initdata e820_table_start; unsigned long __meminitdata e820_table_end; @@ -23,6 +27,69 @@ int direct_gbpages #endif ; +int nx_enabled; + +#if defined(CONFIG_X86_64) || defined(CONFIG_X86_PAE) +static int disable_nx __cpuinitdata; + +/* + * noexec = on|off + * + * Control non-executable mappings for processes. + * + * on Enable + * off Disable + */ +static int __init noexec_setup(char *str) +{ + if (!str) + return -EINVAL; + if (!strncmp(str, "on", 2)) { + __supported_pte_mask |= _PAGE_NX; + disable_nx = 0; + } else if (!strncmp(str, "off", 3)) { + disable_nx = 1; + __supported_pte_mask &= ~_PAGE_NX; + } + return 0; +} +early_param("noexec", noexec_setup); +#endif + +#ifdef CONFIG_X86_PAE +static void __init set_nx(void) +{ + unsigned int v[4], l, h; + + if (cpu_has_pae && (cpuid_eax(0x80000000) > 0x80000001)) { + cpuid(0x80000001, &v[0], &v[1], &v[2], &v[3]); + + if ((v[3] & (1 << 20)) && !disable_nx) { + rdmsr(MSR_EFER, l, h); + l |= EFER_NX; + wrmsr(MSR_EFER, l, h); + nx_enabled = 1; + __supported_pte_mask |= _PAGE_NX; + } + } +} +#else +static inline void set_nx(void) +{ +} +#endif + +#ifdef CONFIG_X86_64 +void __cpuinit check_efer(void) +{ + unsigned long efer; + + rdmsrl(MSR_EFER, efer); + if (!(efer & EFER_NX) || disable_nx) + __supported_pte_mask &= ~_PAGE_NX; +} +#endif + static void __init find_early_table_space(unsigned long end, int use_pse, int use_gbpages) { @@ -66,12 +133,11 @@ static void __init find_early_table_space(unsigned long end, int use_pse, */ #ifdef CONFIG_X86_32 start = 0x7000; - e820_table_start = find_e820_area(start, max_pfn_mapped<<PAGE_SHIFT, - tables, PAGE_SIZE); -#else /* CONFIG_X86_64 */ +#else start = 0x8000; - e820_table_start = find_e820_area(start, end, tables, PAGE_SIZE); #endif + e820_table_start = find_e820_area(start, max_pfn_mapped<<PAGE_SHIFT, + tables, PAGE_SIZE); if (e820_table_start == -1UL) panic("Cannot find space for the kernel page tables"); @@ -159,12 +225,9 @@ unsigned long __init_refok init_memory_mapping(unsigned long start, use_gbpages = direct_gbpages; #endif -#ifdef CONFIG_X86_32 -#ifdef CONFIG_X86_PAE set_nx(); if (nx_enabled) printk(KERN_INFO "NX (Execute Disable) protection: active\n"); -#endif /* Enable PSE if available */ if (cpu_has_pse) @@ -175,7 +238,6 @@ unsigned long __init_refok init_memory_mapping(unsigned long start, set_in_cr4(X86_CR4_PGE); __supported_pte_mask |= _PAGE_GLOBAL; } -#endif if (use_gbpages) page_size_mask |= 1 << PG_LEVEL_1G; diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index 749559ed80f..949708d7a48 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -49,12 +49,9 @@ #include <asm/paravirt.h> #include <asm/setup.h> #include <asm/cacheflush.h> +#include <asm/page_types.h> #include <asm/init.h> -unsigned long max_low_pfn_mapped; -unsigned long max_pfn_mapped; - -DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); unsigned long highstart_pfn, highend_pfn; static noinline int do_test_wp_bit(void); @@ -587,61 +584,9 @@ void zap_low_mappings(void) flush_tlb_all(); } -int nx_enabled; - pteval_t __supported_pte_mask __read_mostly = ~(_PAGE_NX | _PAGE_GLOBAL | _PAGE_IOMAP); EXPORT_SYMBOL_GPL(__supported_pte_mask); -#ifdef CONFIG_X86_PAE - -static int disable_nx __initdata; - -/* - * noexec = on|off - * - * Control non executable mappings. - * - * on Enable - * off Disable - */ -static int __init noexec_setup(char *str) -{ - if (!str || !strcmp(str, "on")) { - if (cpu_has_nx) { - __supported_pte_mask |= _PAGE_NX; - disable_nx = 0; - } - } else { - if (!strcmp(str, "off")) { - disable_nx = 1; - __supported_pte_mask &= ~_PAGE_NX; - } else { - return -EINVAL; - } - } - - return 0; -} -early_param("noexec", noexec_setup); - -void __init set_nx(void) -{ - unsigned int v[4], l, h; - - if (cpu_has_pae && (cpuid_eax(0x80000000) > 0x80000001)) { - cpuid(0x80000001, &v[0], &v[1], &v[2], &v[3]); - - if ((v[3] & (1 << 20)) && !disable_nx) { - rdmsr(MSR_EFER, l, h); - l |= EFER_NX; - wrmsr(MSR_EFER, l, h); - nx_enabled = 1; - __supported_pte_mask |= _PAGE_NX; - } - } -} -#endif - /* user-defined highmem size */ static unsigned int highmem_pages = -1; @@ -761,15 +706,15 @@ void __init initmem_init(unsigned long start_pfn, highstart_pfn = highend_pfn = max_pfn; if (max_pfn > max_low_pfn) highstart_pfn = max_low_pfn; - memory_present(0, 0, highend_pfn); e820_register_active_regions(0, 0, highend_pfn); + sparse_memory_present_with_active_regions(0); printk(KERN_NOTICE "%ldMB HIGHMEM available.\n", pages_to_mb(highend_pfn - highstart_pfn)); num_physpages = highend_pfn; high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1; #else - memory_present(0, 0, max_low_pfn); e820_register_active_regions(0, 0, max_low_pfn); + sparse_memory_present_with_active_regions(0); num_physpages = max_low_pfn; high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1; #endif diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index 1753e8020df..52bb9519bb8 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -50,18 +50,8 @@ #include <asm/cacheflush.h> #include <asm/init.h> -/* - * end_pfn only includes RAM, while max_pfn_mapped includes all e820 entries. - * The direct mapping extends to max_pfn_mapped, so that we can directly access - * apertures, ACPI and other tables without having to play with fixmaps. - */ -unsigned long max_low_pfn_mapped; -unsigned long max_pfn_mapped; - static unsigned long dma_reserve __initdata; -DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); - static int __init parse_direct_gbpages_off(char *arg) { direct_gbpages = 0; @@ -85,39 +75,6 @@ early_param("gbpages", parse_direct_gbpages_on); pteval_t __supported_pte_mask __read_mostly = ~_PAGE_IOMAP; EXPORT_SYMBOL_GPL(__supported_pte_mask); -static int disable_nx __cpuinitdata; - -/* - * noexec=on|off - * Control non-executable mappings for 64-bit processes. - * - * on Enable (default) - * off Disable - */ -static int __init nonx_setup(char *str) -{ - if (!str) - return -EINVAL; - if (!strncmp(str, "on", 2)) { - __supported_pte_mask |= _PAGE_NX; - disable_nx = 0; - } else if (!strncmp(str, "off", 3)) { - disable_nx = 1; - __supported_pte_mask &= ~_PAGE_NX; - } - return 0; -} -early_param("noexec", nonx_setup); - -void __cpuinit check_efer(void) -{ - unsigned long efer; - - rdmsrl(MSR_EFER, efer); - if (!(efer & EFER_NX) || disable_nx) - __supported_pte_mask &= ~_PAGE_NX; -} - int force_personality32; /* @@ -628,6 +585,7 @@ void __init initmem_init(unsigned long start_pfn, unsigned long end_pfn) early_res_to_bootmem(0, end_pfn<<PAGE_SHIFT); reserve_bootmem(bootmap, bootmap_size, BOOTMEM_DEFAULT); } +#endif void __init paging_init(void) { @@ -638,11 +596,10 @@ void __init paging_init(void) max_zone_pfns[ZONE_DMA32] = MAX_DMA32_PFN; max_zone_pfns[ZONE_NORMAL] = max_pfn; - memory_present(0, 0, max_pfn); + sparse_memory_present_with_active_regions(MAX_NUMNODES); sparse_init(); free_area_init_nodes(max_zone_pfns); } -#endif /* * Memory hotplug specific functions diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c index 8056545e2d3..fe6f84ca121 100644 --- a/arch/x86/mm/iomap_32.c +++ b/arch/x86/mm/iomap_32.c @@ -82,7 +82,6 @@ iounmap_atomic(void *kvaddr, enum km_type type) if (vaddr == __fix_to_virt(FIX_KMAP_BEGIN+idx)) kpte_clear_flush(kmap_pte-idx, vaddr); - arch_flush_lazy_mmu_mode(); pagefault_enable(); } EXPORT_SYMBOL_GPL(iounmap_atomic); diff --git a/arch/x86/mm/kmmio.c b/arch/x86/mm/kmmio.c index 50dc802a1c4..16ccbd77917 100644 --- a/arch/x86/mm/kmmio.c +++ b/arch/x86/mm/kmmio.c @@ -32,7 +32,7 @@ struct kmmio_fault_page { struct list_head list; struct kmmio_fault_page *release_next; unsigned long page; /* location of the fault page */ - bool old_presence; /* page presence prior to arming */ + pteval_t old_presence; /* page presence prior to arming */ bool armed; /* @@ -97,60 +97,62 @@ static struct kmmio_probe *get_kmmio_probe(unsigned long addr) static struct kmmio_fault_page *get_kmmio_fault_page(unsigned long page) { struct list_head *head; - struct kmmio_fault_page *p; + struct kmmio_fault_page *f; page &= PAGE_MASK; head = kmmio_page_list(page); - list_for_each_entry_rcu(p, head, list) { - if (p->page == page) - return p; + list_for_each_entry_rcu(f, head, list) { + if (f->page == page) + return f; } return NULL; } -static void set_pmd_presence(pmd_t *pmd, bool present, bool *old) +static void clear_pmd_presence(pmd_t *pmd, bool clear, pmdval_t *old) { pmdval_t v = pmd_val(*pmd); - *old = !!(v & _PAGE_PRESENT); - v &= ~_PAGE_PRESENT; - if (present) - v |= _PAGE_PRESENT; + if (clear) { + *old = v & _PAGE_PRESENT; + v &= ~_PAGE_PRESENT; + } else /* presume this has been called with clear==true previously */ + v |= *old; set_pmd(pmd, __pmd(v)); } -static void set_pte_presence(pte_t *pte, bool present, bool *old) +static void clear_pte_presence(pte_t *pte, bool clear, pteval_t *old) { pteval_t v = pte_val(*pte); - *old = !!(v & _PAGE_PRESENT); - v &= ~_PAGE_PRESENT; - if (present) - v |= _PAGE_PRESENT; + if (clear) { + *old = v & _PAGE_PRESENT; + v &= ~_PAGE_PRESENT; + } else /* presume this has been called with clear==true previously */ + v |= *old; set_pte_atomic(pte, __pte(v)); } -static int set_page_presence(unsigned long addr, bool present, bool *old) +static int clear_page_presence(struct kmmio_fault_page *f, bool clear) { unsigned int level; - pte_t *pte = lookup_address(addr, &level); + pte_t *pte = lookup_address(f->page, &level); if (!pte) { - pr_err("kmmio: no pte for page 0x%08lx\n", addr); + pr_err("kmmio: no pte for page 0x%08lx\n", f->page); return -1; } switch (level) { case PG_LEVEL_2M: - set_pmd_presence((pmd_t *)pte, present, old); + clear_pmd_presence((pmd_t *)pte, clear, &f->old_presence); break; case PG_LEVEL_4K: - set_pte_presence(pte, present, old); + clear_pte_presence(pte, clear, &f->old_presence); break; default: pr_err("kmmio: unexpected page level 0x%x.\n", level); return -1; } - __flush_tlb_one(addr); + __flush_tlb_one(f->page); return 0; } @@ -171,9 +173,9 @@ static int arm_kmmio_fault_page(struct kmmio_fault_page *f) WARN_ONCE(f->armed, KERN_ERR "kmmio page already armed.\n"); if (f->armed) { pr_warning("kmmio double-arm: page 0x%08lx, ref %d, old %d\n", - f->page, f->count, f->old_presence); + f->page, f->count, !!f->old_presence); } - ret = set_page_presence(f->page, false, &f->old_presence); + ret = clear_page_presence(f, true); WARN_ONCE(ret < 0, KERN_ERR "kmmio arming 0x%08lx failed.\n", f->page); f->armed = true; return ret; @@ -182,8 +184,7 @@ static int arm_kmmio_fault_page(struct kmmio_fault_page *f) /** Restore the given page to saved presence state. */ static void disarm_kmmio_fault_page(struct kmmio_fault_page *f) { - bool tmp; - int ret = set_page_presence(f->page, f->old_presence, &tmp); + int ret = clear_page_presence(f, false); WARN_ONCE(ret < 0, KERN_ERR "kmmio disarming 0x%08lx failed.\n", f->page); f->armed = false; @@ -310,7 +311,12 @@ static int post_kmmio_handler(unsigned long condition, struct pt_regs *regs) struct kmmio_context *ctx = &get_cpu_var(kmmio_ctx); if (!ctx->active) { - pr_debug("kmmio: spurious debug trap on CPU %d.\n", + /* + * debug traps without an active context are due to either + * something external causing them (f.e. using a debugger while + * mmio tracing enabled), or erroneous behaviour + */ + pr_warning("kmmio: unexpected debug trap on CPU %d.\n", smp_processor_id()); goto out; } @@ -439,12 +445,12 @@ static void rcu_free_kmmio_fault_pages(struct rcu_head *head) head, struct kmmio_delayed_release, rcu); - struct kmmio_fault_page *p = dr->release_list; - while (p) { - struct kmmio_fault_page *next = p->release_next; - BUG_ON(p->count); - kfree(p); - p = next; + struct kmmio_fault_page *f = dr->release_list; + while (f) { + struct kmmio_fault_page *next = f->release_next; + BUG_ON(f->count); + kfree(f); + f = next; } kfree(dr); } @@ -453,19 +459,19 @@ static void remove_kmmio_fault_pages(struct rcu_head *head) { struct kmmio_delayed_release *dr = container_of(head, struct kmmio_delayed_release, rcu); - struct kmmio_fault_page *p = dr->release_list; + struct kmmio_fault_page *f = dr->release_list; struct kmmio_fault_page **prevp = &dr->release_list; unsigned long flags; spin_lock_irqsave(&kmmio_lock, flags); - while (p) { - if (!p->count) { - list_del_rcu(&p->list); - prevp = &p->release_next; + while (f) { + if (!f->count) { + list_del_rcu(&f->list); + prevp = &f->release_next; } else { - *prevp = p->release_next; + *prevp = f->release_next; } - p = p->release_next; + f = f->release_next; } spin_unlock_irqrestore(&kmmio_lock, flags); @@ -528,8 +534,8 @@ void unregister_kmmio_probe(struct kmmio_probe *p) } EXPORT_SYMBOL(unregister_kmmio_probe); -static int kmmio_die_notifier(struct notifier_block *nb, unsigned long val, - void *args) +static int +kmmio_die_notifier(struct notifier_block *nb, unsigned long val, void *args) { struct die_args *arg = args; @@ -544,11 +550,23 @@ static struct notifier_block nb_die = { .notifier_call = kmmio_die_notifier }; -static int __init init_kmmio(void) +int kmmio_init(void) { int i; + for (i = 0; i < KMMIO_PAGE_TABLE_SIZE; i++) INIT_LIST_HEAD(&kmmio_page_table[i]); + return register_die_notifier(&nb_die); } -fs_initcall(init_kmmio); /* should be before device_initcall() */ + +void kmmio_cleanup(void) +{ + int i; + + unregister_die_notifier(&nb_die); + for (i = 0; i < KMMIO_PAGE_TABLE_SIZE; i++) { + WARN_ONCE(!list_empty(&kmmio_page_table[i]), + KERN_ERR "kmmio_page_table not empty at cleanup, any further tracing will leak memory.\n"); + } +} diff --git a/arch/x86/mm/memtest.c b/arch/x86/mm/memtest.c index 605c8be0621..c0bedcd10f9 100644 --- a/arch/x86/mm/memtest.c +++ b/arch/x86/mm/memtest.c @@ -40,23 +40,23 @@ static void __init reserve_bad_mem(u64 pattern, u64 start_bad, u64 end_bad) static void __init memtest(u64 pattern, u64 start_phys, u64 size) { - u64 i, count; - u64 *start; + u64 *p; + void *start, *end; u64 start_bad, last_bad; u64 start_phys_aligned; size_t incr; incr = sizeof(pattern); start_phys_aligned = ALIGN(start_phys, incr); - count = (size - (start_phys_aligned - start_phys))/incr; start = __va(start_phys_aligned); + end = start + size - (start_phys_aligned - start_phys); start_bad = 0; last_bad = 0; - for (i = 0; i < count; i++) - start[i] = pattern; - for (i = 0; i < count; i++, start++, start_phys_aligned += incr) { - if (*start == pattern) + for (p = start; p < end; p++) + *p = pattern; + for (p = start; p < end; p++, start_phys_aligned += incr) { + if (*p == pattern) continue; if (start_phys_aligned == last_bad + incr) { last_bad += incr; diff --git a/arch/x86/mm/mmio-mod.c b/arch/x86/mm/mmio-mod.c index c9342ed8b40..132772a8ec5 100644 --- a/arch/x86/mm/mmio-mod.c +++ b/arch/x86/mm/mmio-mod.c @@ -451,6 +451,7 @@ void enable_mmiotrace(void) if (nommiotrace) pr_info(NAME "MMIO tracing disabled.\n"); + kmmio_init(); enter_uniprocessor(); spin_lock_irq(&trace_lock); atomic_inc(&mmiotrace_enabled); @@ -473,6 +474,7 @@ void disable_mmiotrace(void) clear_trace_list(); /* guarantees: no more kmmio callbacks */ leave_uniprocessor(); + kmmio_cleanup(); pr_info(NAME "disabled.\n"); out: mutex_unlock(&mmiotrace_mutex); diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index 2d05a12029d..459913beac7 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -179,18 +179,25 @@ static void * __init early_node_mem(int nodeid, unsigned long start, } /* Initialize bootmem allocator for a node */ -void __init setup_node_bootmem(int nodeid, unsigned long start, - unsigned long end) +void __init +setup_node_bootmem(int nodeid, unsigned long start, unsigned long end) { unsigned long start_pfn, last_pfn, bootmap_pages, bootmap_size; + const int pgdat_size = roundup(sizeof(pg_data_t), PAGE_SIZE); unsigned long bootmap_start, nodedata_phys; void *bootmap; - const int pgdat_size = roundup(sizeof(pg_data_t), PAGE_SIZE); int nid; if (!end) return; + /* + * Don't confuse VM with a node that doesn't have the + * minimum amount of memory: + */ + if (end && (end - start) < NODE_MIN_SIZE) + return; + start = roundup(start, ZONE_ALIGN); printk(KERN_INFO "Bootmem setup node %d %016lx-%016lx\n", nodeid, @@ -272,9 +279,6 @@ void __init setup_node_bootmem(int nodeid, unsigned long start, reserve_bootmem_node(NODE_DATA(nodeid), bootmap_start, bootmap_pages<<PAGE_SHIFT, BOOTMEM_DEFAULT); -#ifdef CONFIG_ACPI_NUMA - srat_reserve_add_area(nodeid); -#endif node_set_online(nodeid); } @@ -578,21 +582,6 @@ unsigned long __init numa_free_all_bootmem(void) return pages; } -void __init paging_init(void) -{ - unsigned long max_zone_pfns[MAX_NR_ZONES]; - - memset(max_zone_pfns, 0, sizeof(max_zone_pfns)); - max_zone_pfns[ZONE_DMA] = MAX_DMA_PFN; - max_zone_pfns[ZONE_DMA32] = MAX_DMA32_PFN; - max_zone_pfns[ZONE_NORMAL] = max_pfn; - - sparse_memory_present_with_active_regions(MAX_NUMNODES); - sparse_init(); - - free_area_init_nodes(max_zone_pfns); -} - static __init int numa_setup(char *opt) { if (!opt) @@ -606,8 +595,6 @@ static __init int numa_setup(char *opt) #ifdef CONFIG_ACPI_NUMA if (!strncmp(opt, "noacpi", 6)) acpi_numa = -1; - if (!strncmp(opt, "hotadd=", 7)) - hotadd_percent = simple_strtoul(opt+7, NULL, 10); #endif return 0; } diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c index e17efed088c..6ce9518fe2a 100644 --- a/arch/x86/mm/pageattr.c +++ b/arch/x86/mm/pageattr.c @@ -839,13 +839,6 @@ static int change_page_attr_set_clr(unsigned long *addr, int numpages, vm_unmap_aliases(); - /* - * If we're called with lazy mmu updates enabled, the - * in-memory pte state may be stale. Flush pending updates to - * bring them up to date. - */ - arch_flush_lazy_mmu_mode(); - cpa.vaddr = addr; cpa.pages = pages; cpa.numpages = numpages; @@ -890,13 +883,6 @@ static int change_page_attr_set_clr(unsigned long *addr, int numpages, } else cpa_flush_all(cache); - /* - * If we've been called with lazy mmu updates enabled, then - * make sure that everything gets flushed out before we - * return. - */ - arch_flush_lazy_mmu_mode(); - out: return ret; } diff --git a/arch/x86/mm/srat_64.c b/arch/x86/mm/srat_64.c index 01765955baa..2dfcbf9df2a 100644 --- a/arch/x86/mm/srat_64.c +++ b/arch/x86/mm/srat_64.c @@ -31,17 +31,11 @@ static nodemask_t nodes_parsed __initdata; static nodemask_t cpu_nodes_parsed __initdata; static struct bootnode nodes[MAX_NUMNODES] __initdata; static struct bootnode nodes_add[MAX_NUMNODES]; -static int found_add_area __initdata; -int hotadd_percent __initdata = 0; static int num_node_memblks __initdata; static struct bootnode node_memblk_range[NR_NODE_MEMBLKS] __initdata; static int memblk_nodeid[NR_NODE_MEMBLKS] __initdata; -/* Too small nodes confuse the VM badly. Usually they result - from BIOS bugs. */ -#define NODE_MIN_SIZE (4*1024*1024) - static __init int setup_node(int pxm) { return acpi_map_pxm_to_node(pxm); @@ -66,9 +60,6 @@ static __init void cutoff_node(int i, unsigned long start, unsigned long end) { struct bootnode *nd = &nodes[i]; - if (found_add_area) - return; - if (nd->start < start) { nd->start = start; if (nd->end < nd->start) @@ -86,7 +77,6 @@ static __init void bad_srat(void) int i; printk(KERN_ERR "SRAT: SRAT not used.\n"); acpi_numa = -1; - found_add_area = 0; for (i = 0; i < MAX_LOCAL_APIC; i++) apicid_to_node[i] = NUMA_NO_NODE; for (i = 0; i < MAX_NUMNODES; i++) @@ -182,24 +172,21 @@ acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa) pxm, apic_id, node); } -static int update_end_of_memory(unsigned long end) {return -1;} -static int hotadd_enough_memory(struct bootnode *nd) {return 1;} #ifdef CONFIG_MEMORY_HOTPLUG_SPARSE static inline int save_add_info(void) {return 1;} #else static inline int save_add_info(void) {return 0;} #endif /* - * Update nodes_add and decide if to include add are in the zone. - * Both SPARSE and RESERVE need nodes_add information. - * This code supports one contiguous hot add area per node. + * Update nodes_add[] + * This code supports one contiguous hot add area per node */ -static int __init -reserve_hotadd(int node, unsigned long start, unsigned long end) +static void __init +update_nodes_add(int node, unsigned long start, unsigned long end) { unsigned long s_pfn = start >> PAGE_SHIFT; unsigned long e_pfn = end >> PAGE_SHIFT; - int ret = 0, changed = 0; + int changed = 0; struct bootnode *nd = &nodes_add[node]; /* I had some trouble with strange memory hotadd regions breaking @@ -210,7 +197,7 @@ reserve_hotadd(int node, unsigned long start, unsigned long end) mistakes */ if ((signed long)(end - start) < NODE_MIN_SIZE) { printk(KERN_ERR "SRAT: Hotplug area too small\n"); - return -1; + return; } /* This check might be a bit too strict, but I'm keeping it for now. */ @@ -218,12 +205,7 @@ reserve_hotadd(int node, unsigned long start, unsigned long end) printk(KERN_ERR "SRAT: Hotplug area %lu -> %lu has existing memory\n", s_pfn, e_pfn); - return -1; - } - - if (!hotadd_enough_memory(&nodes_add[node])) { - printk(KERN_ERR "SRAT: Hotplug area too large\n"); - return -1; + return; } /* Looks good */ @@ -245,11 +227,9 @@ reserve_hotadd(int node, unsigned long start, unsigned long end) printk(KERN_ERR "SRAT: Hotplug zone not continuous. Partly ignored\n"); } - ret = update_end_of_memory(nd->end); - if (changed) - printk(KERN_INFO "SRAT: hot plug zone found %Lx - %Lx\n", nd->start, nd->end); - return ret; + printk(KERN_INFO "SRAT: hot plug zone found %Lx - %Lx\n", + nd->start, nd->end); } /* Callback for parsing of the Proximity Domain <-> Memory Area mappings */ @@ -310,13 +290,10 @@ acpi_numa_memory_affinity_init(struct acpi_srat_mem_affinity *ma) start, end); e820_register_active_regions(node, start >> PAGE_SHIFT, end >> PAGE_SHIFT); - push_node_boundaries(node, nd->start >> PAGE_SHIFT, - nd->end >> PAGE_SHIFT); - if ((ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) && - (reserve_hotadd(node, start, end) < 0)) { - /* Ignore hotadd region. Undo damage */ - printk(KERN_NOTICE "SRAT: Hotplug region ignored\n"); + if (ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) { + update_nodes_add(node, start, end); + /* restore nodes[node] */ *nd = oldnode; if ((nd->start | nd->end) == 0) node_clear(node, nodes_parsed); @@ -345,9 +322,9 @@ static int __init nodes_cover_memory(const struct bootnode *nodes) pxmram = 0; } - e820ram = max_pfn - absent_pages_in_range(0, max_pfn); - /* We seem to lose 3 pages somewhere. Allow a bit of slack. */ - if ((long)(e820ram - pxmram) >= 1*1024*1024) { + e820ram = max_pfn - (e820_hole_size(0, max_pfn<<PAGE_SHIFT)>>PAGE_SHIFT); + /* We seem to lose 3 pages somewhere. Allow 1M of slack. */ + if ((long)(e820ram - pxmram) >= (1<<(20 - PAGE_SHIFT))) { printk(KERN_ERR "SRAT: PXMs only cover %luMB of your %luMB e820 RAM. Not used.\n", (pxmram << PAGE_SHIFT) >> 20, @@ -357,17 +334,6 @@ static int __init nodes_cover_memory(const struct bootnode *nodes) return 1; } -static void __init unparse_node(int node) -{ - int i; - node_clear(node, nodes_parsed); - node_clear(node, cpu_nodes_parsed); - for (i = 0; i < MAX_LOCAL_APIC; i++) { - if (apicid_to_node[i] == node) - apicid_to_node[i] = NUMA_NO_NODE; - } -} - void __init acpi_numa_arch_fixup(void) {} /* Use the information discovered above to actually set up the nodes. */ @@ -379,18 +345,8 @@ int __init acpi_scan_nodes(unsigned long start, unsigned long end) return -1; /* First clean up the node list */ - for (i = 0; i < MAX_NUMNODES; i++) { + for (i = 0; i < MAX_NUMNODES; i++) cutoff_node(i, start, end); - /* - * don't confuse VM with a node that doesn't have the - * minimum memory. - */ - if (nodes[i].end && - (nodes[i].end - nodes[i].start) < NODE_MIN_SIZE) { - unparse_node(i); - node_set_offline(i); - } - } if (!nodes_cover_memory(nodes)) { bad_srat(); @@ -423,7 +379,7 @@ int __init acpi_scan_nodes(unsigned long start, unsigned long end) if (node == NUMA_NO_NODE) continue; - if (!node_isset(node, node_possible_map)) + if (!node_online(node)) numa_clear_node(i); } numa_init_array(); @@ -510,26 +466,6 @@ static int null_slit_node_compare(int a, int b) } #endif /* CONFIG_NUMA_EMU */ -void __init srat_reserve_add_area(int nodeid) -{ - if (found_add_area && nodes_add[nodeid].end) { - u64 total_mb; - - printk(KERN_INFO "SRAT: Reserving hot-add memory space " - "for node %d at %Lx-%Lx\n", - nodeid, nodes_add[nodeid].start, nodes_add[nodeid].end); - total_mb = (nodes_add[nodeid].end - nodes_add[nodeid].start) - >> PAGE_SHIFT; - total_mb *= sizeof(struct page); - total_mb >>= 20; - printk(KERN_INFO "SRAT: This will cost you %Lu MB of " - "pre-allocated memory.\n", (unsigned long long)total_mb); - reserve_bootmem_node(NODE_DATA(nodeid), nodes_add[nodeid].start, - nodes_add[nodeid].end - nodes_add[nodeid].start, - BOOTMEM_DEFAULT); - } -} - int __node_distance(int a, int b) { int index; diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c index 80b63d5db50..7826dfcc842 100644 --- a/arch/x86/oprofile/nmi_int.c +++ b/arch/x86/oprofile/nmi_int.c @@ -60,8 +60,9 @@ static int profile_exceptions_notify(struct notifier_block *self, switch (val) { case DIE_NMI: - if (model->check_ctrs(args->regs, &per_cpu(cpu_msrs, cpu))) - ret = NOTIFY_STOP; + case DIE_NMI_IPI: + model->check_ctrs(args->regs, &per_cpu(cpu_msrs, cpu)); + ret = NOTIFY_STOP; break; default: break; @@ -146,7 +147,7 @@ static void nmi_cpu_setup(void *dummy) static struct notifier_block profile_exceptions_nb = { .notifier_call = profile_exceptions_notify, .next = NULL, - .priority = 0 + .priority = 2 }; static int nmi_setup(void) diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c index 0a261a5c696..cd72d5c73b4 100644 --- a/arch/x86/oprofile/op_model_ppro.c +++ b/arch/x86/oprofile/op_model_ppro.c @@ -118,6 +118,13 @@ static int ppro_check_ctrs(struct pt_regs * const regs, u64 val; int i; + /* + * This can happen if perf counters are in use when + * we steal the die notifier NMI. + */ + if (unlikely(!reset_value)) + goto out; + for (i = 0 ; i < num_counters; ++i) { if (!reset_value[i]) continue; @@ -128,6 +135,7 @@ static int ppro_check_ctrs(struct pt_regs * const regs, wrmsrl(msrs->counters[i].addr, -reset_value[i]); } +out: /* Only P6 based Pentium M need to re-unmask the apic vector but it * doesn't hurt other P6 variant */ apic_write(APIC_LVTPC, apic_read(APIC_LVTPC) & ~APIC_LVT_MASKED); diff --git a/arch/x86/oprofile/op_x86_model.h b/arch/x86/oprofile/op_x86_model.h index fda52b4c1b9..505489873b9 100644 --- a/arch/x86/oprofile/op_x86_model.h +++ b/arch/x86/oprofile/op_x86_model.h @@ -13,7 +13,7 @@ #define OP_X86_MODEL_H #include <asm/types.h> -#include <asm/intel_arch_perfmon.h> +#include <asm/perf_counter.h> struct op_msr { unsigned long addr; diff --git a/arch/x86/pci/irq.c b/arch/x86/pci/irq.c index fecbce6e7d7..0696d506c4a 100644 --- a/arch/x86/pci/irq.c +++ b/arch/x86/pci/irq.c @@ -889,6 +889,9 @@ static int pcibios_lookup_irq(struct pci_dev *dev, int assign) return 0; } + if (io_apic_assign_pci_irqs) + return 0; + /* Find IRQ routing entry */ if (!pirq_table) @@ -1039,56 +1042,15 @@ static void __init pcibios_fixup_irqs(void) pirq_penalty[dev->irq]++; } + if (io_apic_assign_pci_irqs) + return; + dev = NULL; while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) { pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin); if (!pin) continue; -#ifdef CONFIG_X86_IO_APIC - /* - * Recalculate IRQ numbers if we use the I/O APIC. - */ - if (io_apic_assign_pci_irqs) { - int irq; - - /* - * interrupt pins are numbered starting from 1 - */ - irq = IO_APIC_get_PCI_irq_vector(dev->bus->number, - PCI_SLOT(dev->devfn), pin - 1); - /* - * Busses behind bridges are typically not listed in the - * MP-table. In this case we have to look up the IRQ - * based on the parent bus, parent slot, and pin number. - * The SMP code detects such bridged busses itself so we - * should get into this branch reliably. - */ - if (irq < 0 && dev->bus->parent) { - /* go back to the bridge */ - struct pci_dev *bridge = dev->bus->self; - int bus; - - pin = pci_swizzle_interrupt_pin(dev, pin); - bus = bridge->bus->number; - irq = IO_APIC_get_PCI_irq_vector(bus, - PCI_SLOT(bridge->devfn), pin - 1); - if (irq >= 0) - dev_warn(&dev->dev, - "using bridge %s INT %c to " - "get IRQ %d\n", - pci_name(bridge), - 'A' + pin - 1, irq); - } - if (irq >= 0) { - dev_info(&dev->dev, - "PCI->APIC IRQ transform: INT %c " - "-> IRQ %d\n", - 'A' + pin - 1, irq); - dev->irq = irq; - } - } -#endif /* * Still no IRQ? Try to lookup one... */ @@ -1183,6 +1145,19 @@ int __init pcibios_irq_init(void) pcibios_enable_irq = pirq_enable_irq; pcibios_fixup_irqs(); + + if (io_apic_assign_pci_irqs && pci_routeirq) { + struct pci_dev *dev = NULL; + /* + * PCI IRQ routing is set up by pci_enable_device(), but we + * also do it here in case there are still broken drivers that + * don't use pci_enable_device(). + */ + printk(KERN_INFO "PCI: Routing PCI interrupts for all devices because \"pci=routeirq\" specified\n"); + for_each_pci_dev(dev) + pirq_enable_irq(dev); + } + return 0; } @@ -1213,16 +1188,23 @@ void pcibios_penalize_isa_irq(int irq, int active) static int pirq_enable_irq(struct pci_dev *dev) { u8 pin; - struct pci_dev *temp_dev; pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin); - if (pin && !pcibios_lookup_irq(dev, 1) && !dev->irq) { + if (pin && !pcibios_lookup_irq(dev, 1)) { char *msg = ""; + if (!io_apic_assign_pci_irqs && dev->irq) + return 0; + if (io_apic_assign_pci_irqs) { +#ifdef CONFIG_X86_IO_APIC + struct pci_dev *temp_dev; int irq; + struct io_apic_irq_attr irq_attr; - irq = IO_APIC_get_PCI_irq_vector(dev->bus->number, PCI_SLOT(dev->devfn), pin - 1); + irq = IO_APIC_get_PCI_irq_vector(dev->bus->number, + PCI_SLOT(dev->devfn), + pin - 1, &irq_attr); /* * Busses behind bridges are typically not listed in the MP-table. * In this case we have to look up the IRQ based on the parent bus, @@ -1235,7 +1217,8 @@ static int pirq_enable_irq(struct pci_dev *dev) pin = pci_swizzle_interrupt_pin(dev, pin); irq = IO_APIC_get_PCI_irq_vector(bridge->bus->number, - PCI_SLOT(bridge->devfn), pin - 1); + PCI_SLOT(bridge->devfn), + pin - 1, &irq_attr); if (irq >= 0) dev_warn(&dev->dev, "using bridge %s " "INT %c to get IRQ %d\n", @@ -1245,12 +1228,15 @@ static int pirq_enable_irq(struct pci_dev *dev) } dev = temp_dev; if (irq >= 0) { + io_apic_set_pci_routing(&dev->dev, irq, + &irq_attr); + dev->irq = irq; dev_info(&dev->dev, "PCI->APIC IRQ transform: " "INT %c -> IRQ %d\n", 'A' + pin - 1, irq); - dev->irq = irq; return 0; } else msg = "; probably buggy MP table"; +#endif } else if (pci_probe & PCI_BIOS_IRQ_SCAN) msg = ""; else diff --git a/arch/x86/vdso/vdso32-setup.c b/arch/x86/vdso/vdso32-setup.c index 1241f118ab5..58bc00f68b1 100644 --- a/arch/x86/vdso/vdso32-setup.c +++ b/arch/x86/vdso/vdso32-setup.c @@ -338,6 +338,8 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) } } + current->mm->context.vdso = (void *)addr; + if (compat_uses_vma || !compat) { /* * MAYWRITE to allow gdb to COW and set breakpoints @@ -358,11 +360,13 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) goto up_fail; } - current->mm->context.vdso = (void *)addr; current_thread_info()->sysenter_return = VDSO32_SYMBOL(addr, SYSENTER_RETURN); up_fail: + if (ret) + current->mm->context.vdso = NULL; + up_write(&mm->mmap_sem); return ret; diff --git a/arch/x86/vdso/vma.c b/arch/x86/vdso/vma.c index 7133cdf9098..21e1aeb9f3e 100644 --- a/arch/x86/vdso/vma.c +++ b/arch/x86/vdso/vma.c @@ -8,6 +8,7 @@ #include <linux/sched.h> #include <linux/init.h> #include <linux/random.h> +#include <linux/elf.h> #include <asm/vsyscall.h> #include <asm/vgtod.h> #include <asm/proto.h> @@ -115,15 +116,18 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) goto up_fail; } + current->mm->context.vdso = (void *)addr; + ret = install_special_mapping(mm, addr, vdso_size, VM_READ|VM_EXEC| VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC| VM_ALWAYSDUMP, vdso_pages); - if (ret) + if (ret) { + current->mm->context.vdso = NULL; goto up_fail; + } - current->mm->context.vdso = (void *)addr; up_fail: up_write(&mm->mmap_sem); return ret; diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index f09e8c36ee8..0a1700a2be9 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -20,6 +20,7 @@ #include <linux/delay.h> #include <linux/start_kernel.h> #include <linux/sched.h> +#include <linux/kprobes.h> #include <linux/bootmem.h> #include <linux/module.h> #include <linux/mm.h> @@ -44,6 +45,7 @@ #include <asm/processor.h> #include <asm/proto.h> #include <asm/msr-index.h> +#include <asm/traps.h> #include <asm/setup.h> #include <asm/desc.h> #include <asm/pgtable.h> @@ -240,10 +242,10 @@ static unsigned long xen_get_debugreg(int reg) return HYPERVISOR_get_debugreg(reg); } -void xen_leave_lazy(void) +static void xen_end_context_switch(struct task_struct *next) { - paravirt_leave_lazy(paravirt_get_lazy_mode()); xen_mc_flush(); + paravirt_end_context_switch(next); } static unsigned long xen_store_tr(void) @@ -428,11 +430,44 @@ static void xen_write_ldt_entry(struct desc_struct *dt, int entrynum, static int cvt_gate_to_trap(int vector, const gate_desc *val, struct trap_info *info) { + unsigned long addr; + if (val->type != GATE_TRAP && val->type != GATE_INTERRUPT) return 0; info->vector = vector; - info->address = gate_offset(*val); + + addr = gate_offset(*val); +#ifdef CONFIG_X86_64 + /* + * Look for known traps using IST, and substitute them + * appropriately. The debugger ones are the only ones we care + * about. Xen will handle faults like double_fault and + * machine_check, so we should never see them. Warn if + * there's an unexpected IST-using fault handler. + */ + if (addr == (unsigned long)debug) + addr = (unsigned long)xen_debug; + else if (addr == (unsigned long)int3) + addr = (unsigned long)xen_int3; + else if (addr == (unsigned long)stack_segment) + addr = (unsigned long)xen_stack_segment; + else if (addr == (unsigned long)double_fault || + addr == (unsigned long)nmi) { + /* Don't need to handle these */ + return 0; +#ifdef CONFIG_X86_MCE + } else if (addr == (unsigned long)machine_check) { + return 0; +#endif + } else { + /* Some other trap using IST? */ + if (WARN_ON(val->ist != 0)) + return 0; + } +#endif /* CONFIG_X86_64 */ + info->address = addr; + info->cs = gate_segment(*val); info->flags = val->dpl; /* interrupt gates clear IF */ @@ -623,10 +658,26 @@ static void xen_clts(void) xen_mc_issue(PARAVIRT_LAZY_CPU); } +static DEFINE_PER_CPU(unsigned long, xen_cr0_value); + +static unsigned long xen_read_cr0(void) +{ + unsigned long cr0 = percpu_read(xen_cr0_value); + + if (unlikely(cr0 == 0)) { + cr0 = native_read_cr0(); + percpu_write(xen_cr0_value, cr0); + } + + return cr0; +} + static void xen_write_cr0(unsigned long cr0) { struct multicall_space mcs; + percpu_write(xen_cr0_value, cr0); + /* Only pay attention to cr0.TS; everything else is ignored. */ mcs = xen_mc_entry(0); @@ -812,7 +863,7 @@ static const struct pv_cpu_ops xen_cpu_ops __initdata = { .clts = xen_clts, - .read_cr0 = native_read_cr0, + .read_cr0 = xen_read_cr0, .write_cr0 = xen_write_cr0, .read_cr4 = native_read_cr4, @@ -860,10 +911,8 @@ static const struct pv_cpu_ops xen_cpu_ops __initdata = { /* Xen takes care of %gs when switching to usermode for us */ .swapgs = paravirt_nop, - .lazy_mode = { - .enter = paravirt_enter_lazy_cpu, - .leave = xen_leave_lazy, - }, + .start_context_switch = paravirt_start_context_switch, + .end_context_switch = xen_end_context_switch, }; static const struct pv_apic_ops xen_apic_ops __initdata = { diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index fba55b1a402..4ceb2858165 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -452,10 +452,6 @@ void set_pte_mfn(unsigned long vaddr, unsigned long mfn, pgprot_t flags) void xen_set_pte_at(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pte_t pteval) { - /* updates to init_mm may be done without lock */ - if (mm == &init_mm) - preempt_disable(); - ADD_STATS(set_pte_at, 1); // ADD_STATS(set_pte_at_pinned, xen_page_pinned(ptep)); ADD_STATS(set_pte_at_current, mm == current->mm); @@ -476,9 +472,7 @@ void xen_set_pte_at(struct mm_struct *mm, unsigned long addr, } xen_set_pte(ptep, pteval); -out: - if (mm == &init_mm) - preempt_enable(); +out: return; } pte_t xen_ptep_modify_prot_start(struct mm_struct *mm, @@ -1152,10 +1146,8 @@ static void drop_other_mm_ref(void *info) /* If this cpu still has a stale cr3 reference, then make sure it has been flushed. */ - if (percpu_read(xen_current_cr3) == __pa(mm->pgd)) { + if (percpu_read(xen_current_cr3) == __pa(mm->pgd)) load_cr3(swapper_pg_dir); - arch_flush_lazy_cpu_mode(); - } } static void xen_drop_mm_ref(struct mm_struct *mm) @@ -1168,7 +1160,6 @@ static void xen_drop_mm_ref(struct mm_struct *mm) load_cr3(swapper_pg_dir); else leave_mm(smp_processor_id()); - arch_flush_lazy_cpu_mode(); } /* Get the "official" set of cpus referring to our pagetable. */ @@ -1876,6 +1867,14 @@ __init void xen_post_allocator_init(void) xen_mark_init_mm_pinned(); } +static void xen_leave_lazy_mmu(void) +{ + preempt_disable(); + xen_mc_flush(); + paravirt_leave_lazy_mmu(); + preempt_enable(); +} + const struct pv_mmu_ops xen_mmu_ops __initdata = { .pagetable_setup_start = xen_pagetable_setup_start, .pagetable_setup_done = xen_pagetable_setup_done, @@ -1949,7 +1948,7 @@ const struct pv_mmu_ops xen_mmu_ops __initdata = { .lazy_mode = { .enter = paravirt_enter_lazy_mmu, - .leave = xen_leave_lazy, + .leave = xen_leave_lazy_mmu, }, .set_fixmap = xen_set_fixmap, diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c index 15c6c68db6a..ad0047f47cd 100644 --- a/arch/x86/xen/setup.c +++ b/arch/x86/xen/setup.c @@ -61,9 +61,9 @@ char * __init xen_memory_setup(void) * - xen_start_info * See comment above "struct start_info" in <xen/interface/xen.h> */ - e820_add_region(__pa(xen_start_info->mfn_list), - xen_start_info->pt_base - xen_start_info->mfn_list, - E820_RESERVED); + reserve_early(__pa(xen_start_info->mfn_list), + __pa(xen_start_info->pt_base), + "XEN START INFO"); sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h index ca6596b05d5..22494fd4c9b 100644 --- a/arch/x86/xen/xen-ops.h +++ b/arch/x86/xen/xen-ops.h @@ -30,7 +30,6 @@ pgd_t *xen_setup_kernel_pagetable(pgd_t *pgd, unsigned long max_pfn); void xen_ident_map_ISA(void); void xen_reserve_top(void); -void xen_leave_lazy(void); void xen_post_allocator_init(void); char * __init xen_memory_setup(void); diff --git a/block/blk-core.c b/block/blk-core.c index c89883be873..9475bf99b89 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -28,22 +28,14 @@ #include <linux/task_io_accounting_ops.h> #include <linux/blktrace_api.h> #include <linux/fault-inject.h> -#include <trace/block.h> + +#define CREATE_TRACE_POINTS +#include <trace/events/block.h> #include "blk.h" -DEFINE_TRACE(block_plug); -DEFINE_TRACE(block_unplug_io); -DEFINE_TRACE(block_unplug_timer); -DEFINE_TRACE(block_getrq); -DEFINE_TRACE(block_sleeprq); -DEFINE_TRACE(block_rq_requeue); -DEFINE_TRACE(block_bio_backmerge); -DEFINE_TRACE(block_bio_frontmerge); -DEFINE_TRACE(block_bio_queue); -DEFINE_TRACE(block_rq_complete); -DEFINE_TRACE(block_remap); /* Also used in drivers/md/dm.c */ EXPORT_TRACEPOINT_SYMBOL_GPL(block_remap); +EXPORT_TRACEPOINT_SYMBOL_GPL(block_bio_complete); static int __make_request(struct request_queue *q, struct bio *bio); @@ -1277,7 +1269,7 @@ static inline void blk_partition_remap(struct bio *bio) bio->bi_bdev = bdev->bd_contains; trace_block_remap(bdev_get_queue(bio->bi_bdev), bio, - bdev->bd_dev, bio->bi_sector, + bdev->bd_dev, bio->bi_sector - p->start_sect); } } @@ -1446,8 +1438,7 @@ static inline void __generic_make_request(struct bio *bio) goto end_io; if (old_sector != -1) - trace_block_remap(q, bio, old_dev, bio->bi_sector, - old_sector); + trace_block_remap(q, bio, old_dev, old_sector); trace_block_bio_queue(q, bio); diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index 3ff9bba3379..26f9ec28f56 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -383,16 +383,21 @@ struct kobj_type blk_queue_ktype = { int blk_register_queue(struct gendisk *disk) { int ret; + struct device *dev = disk_to_dev(disk); struct request_queue *q = disk->queue; if (WARN_ON(!q)) return -ENXIO; + ret = blk_trace_init_sysfs(dev); + if (ret) + return ret; + if (!q->request_fn) return 0; - ret = kobject_add(&q->kobj, kobject_get(&disk_to_dev(disk)->kobj), + ret = kobject_add(&q->kobj, kobject_get(&dev->kobj), "%s", "queue"); if (ret < 0) return ret; diff --git a/block/compat_ioctl.c b/block/compat_ioctl.c index f87615dea46..f8c218cd08e 100644 --- a/block/compat_ioctl.c +++ b/block/compat_ioctl.c @@ -568,7 +568,7 @@ static int compat_blk_trace_setup(struct block_device *bdev, char __user *arg) memcpy(&buts.name, &cbuts.name, 32); mutex_lock(&bdev->bd_mutex); - ret = do_blk_trace_setup(q, b, bdev->bd_dev, &buts); + ret = do_blk_trace_setup(q, b, bdev->bd_dev, bdev, &buts); mutex_unlock(&bdev->bd_mutex); if (ret) return ret; diff --git a/block/elevator.c b/block/elevator.c index 7073a907257..e220f0c543e 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -33,17 +33,16 @@ #include <linux/compiler.h> #include <linux/delay.h> #include <linux/blktrace_api.h> -#include <trace/block.h> #include <linux/hash.h> #include <linux/uaccess.h> +#include <trace/events/block.h> + #include "blk.h" static DEFINE_SPINLOCK(elv_list_lock); static LIST_HEAD(elv_list); -DEFINE_TRACE(block_rq_abort); - /* * Merge hash stuff. */ @@ -55,9 +54,6 @@ static const int elv_hash_shift = 6; #define rq_hash_key(rq) ((rq)->sector + (rq)->nr_sectors) #define ELV_ON_HASH(rq) (!hlist_unhashed(&(rq)->hash)) -DEFINE_TRACE(block_rq_insert); -DEFINE_TRACE(block_rq_issue); - /* * Query io scheduler to see if the current process issuing bio may be * merged with rq. diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c index 51b9f8280f8..2faa9e2ac89 100644 --- a/drivers/acpi/pci_irq.c +++ b/drivers/acpi/pci_irq.c @@ -401,7 +401,8 @@ int acpi_pci_irq_enable(struct pci_dev *dev) /* Interrupt Line values above 0xF are forbidden */ if (dev->irq > 0 && (dev->irq <= 0xF)) { printk(" - using IRQ %d\n", dev->irq); - acpi_register_gsi(dev->irq, ACPI_LEVEL_SENSITIVE, + acpi_register_gsi(&dev->dev, dev->irq, + ACPI_LEVEL_SENSITIVE, ACPI_ACTIVE_LOW); return 0; } else { @@ -410,7 +411,7 @@ int acpi_pci_irq_enable(struct pci_dev *dev) } } - rc = acpi_register_gsi(gsi, triggering, polarity); + rc = acpi_register_gsi(&dev->dev, gsi, triggering, polarity); if (rc < 0) { dev_warn(&dev->dev, "PCI INT %c: failed to register GSI\n", pin_name(pin)); diff --git a/drivers/char/hpet.c b/drivers/char/hpet.c index 340ba4f9dc5..4a9f3492b92 100644 --- a/drivers/char/hpet.c +++ b/drivers/char/hpet.c @@ -224,7 +224,7 @@ static void hpet_timer_set_irq(struct hpet_dev *devp) break; } - gsi = acpi_register_gsi(irq, ACPI_LEVEL_SENSITIVE, + gsi = acpi_register_gsi(NULL, irq, ACPI_LEVEL_SENSITIVE, ACPI_ACTIVE_LOW); if (gsi > 0) break; @@ -939,7 +939,7 @@ static acpi_status hpet_resources(struct acpi_resource *res, void *data) irqp = &res->data.extended_irq; for (i = 0; i < irqp->interrupt_count; i++) { - irq = acpi_register_gsi(irqp->interrupts[i], + irq = acpi_register_gsi(NULL, irqp->interrupts[i], irqp->triggering, irqp->polarity); if (irq < 0) return AE_ERROR; diff --git a/drivers/char/mem.c b/drivers/char/mem.c index 65e12bca657..f96d0bef855 100644 --- a/drivers/char/mem.c +++ b/drivers/char/mem.c @@ -694,9 +694,8 @@ static ssize_t read_zero(struct file * file, char __user * buf, written += chunk - unwritten; if (unwritten) break; - /* Consider changing this to just 'signal_pending()' with lots of testing */ - if (fatal_signal_pending(current)) - return written ? written : -EINTR; + if (signal_pending(current)) + return written ? written : -ERESTARTSYS; buf += chunk; count -= chunk; cond_resched(); diff --git a/drivers/char/sysrq.c b/drivers/char/sysrq.c index d6a807f4077..39a05b5fa9c 100644 --- a/drivers/char/sysrq.c +++ b/drivers/char/sysrq.c @@ -25,6 +25,7 @@ #include <linux/kbd_kern.h> #include <linux/proc_fs.h> #include <linux/quotaops.h> +#include <linux/perf_counter.h> #include <linux/kernel.h> #include <linux/module.h> #include <linux/suspend.h> @@ -243,6 +244,7 @@ static void sysrq_handle_showregs(int key, struct tty_struct *tty) struct pt_regs *regs = get_irq_regs(); if (regs) show_regs(regs); + perf_counter_print_debug(); } static struct sysrq_key_op sysrq_showregs_op = { .handler = sysrq_handle_showregs, diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 424f7b048c3..3fd8b1e6548 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -20,7 +20,8 @@ #include <linux/idr.h> #include <linux/hdreg.h> #include <linux/blktrace_api.h> -#include <trace/block.h> + +#include <trace/events/block.h> #define DM_MSG_PREFIX "core" @@ -53,8 +54,6 @@ struct dm_target_io { union map_info info; }; -DEFINE_TRACE(block_bio_complete); - /* * For request-based dm. * One of these is allocated per request. @@ -656,8 +655,7 @@ static void __map_bio(struct dm_target *ti, struct bio *clone, /* the bio has been remapped so dispatch it */ trace_block_remap(bdev_get_queue(clone->bi_bdev), clone, - tio->io->bio->bi_bdev->bd_dev, - clone->bi_sector, sector); + tio->io->bio->bi_bdev->bd_dev, sector); generic_make_request(clone); } else if (r < 0 || r == DM_MAPIO_REQUEUE) { diff --git a/drivers/parisc/iosapic.c b/drivers/parisc/iosapic.c index 73348c4047e..4a9cc92d4d1 100644 --- a/drivers/parisc/iosapic.c +++ b/drivers/parisc/iosapic.c @@ -702,7 +702,7 @@ static unsigned int iosapic_startup_irq(unsigned int irq) } #ifdef CONFIG_SMP -static void iosapic_set_affinity_irq(unsigned int irq, +static int iosapic_set_affinity_irq(unsigned int irq, const struct cpumask *dest) { struct vector_info *vi = iosapic_get_vector(irq); @@ -712,7 +712,7 @@ static void iosapic_set_affinity_irq(unsigned int irq, dest_cpu = cpu_check_affinity(irq, dest); if (dest_cpu < 0) - return; + return -1; cpumask_copy(irq_desc[irq].affinity, cpumask_of(dest_cpu)); vi->txn_addr = txn_affinity_addr(irq, dest_cpu); @@ -724,6 +724,8 @@ static void iosapic_set_affinity_irq(unsigned int irq, iosapic_set_irt_data(vi, &dummy_d0, &d1); iosapic_wr_irt_entry(vi, d0, d1); spin_unlock_irqrestore(&iosapic_lock, flags); + + return 0; } #endif diff --git a/drivers/pci/hotplug/ibmphp_core.c b/drivers/pci/hotplug/ibmphp_core.c index dd18f857dfb..42e4260c3b1 100644 --- a/drivers/pci/hotplug/ibmphp_core.c +++ b/drivers/pci/hotplug/ibmphp_core.c @@ -153,45 +153,47 @@ int ibmphp_init_devno(struct slot **cur_slot) return -1; } for (loop = 0; loop < len; loop++) { - if ((*cur_slot)->number == rtable->slots[loop].slot) { - if ((*cur_slot)->bus == rtable->slots[loop].bus) { + if ((*cur_slot)->number == rtable->slots[loop].slot && + (*cur_slot)->bus == rtable->slots[loop].bus) { + struct io_apic_irq_attr irq_attr; + (*cur_slot)->device = PCI_SLOT(rtable->slots[loop].devfn); for (i = 0; i < 4; i++) (*cur_slot)->irq[i] = IO_APIC_get_PCI_irq_vector((int) (*cur_slot)->bus, - (int) (*cur_slot)->device, i); - - debug("(*cur_slot)->irq[0] = %x\n", - (*cur_slot)->irq[0]); - debug("(*cur_slot)->irq[1] = %x\n", - (*cur_slot)->irq[1]); - debug("(*cur_slot)->irq[2] = %x\n", - (*cur_slot)->irq[2]); - debug("(*cur_slot)->irq[3] = %x\n", - (*cur_slot)->irq[3]); - - debug("rtable->exlusive_irqs = %x\n", + (int) (*cur_slot)->device, i, + &irq_attr); + + debug("(*cur_slot)->irq[0] = %x\n", + (*cur_slot)->irq[0]); + debug("(*cur_slot)->irq[1] = %x\n", + (*cur_slot)->irq[1]); + debug("(*cur_slot)->irq[2] = %x\n", + (*cur_slot)->irq[2]); + debug("(*cur_slot)->irq[3] = %x\n", + (*cur_slot)->irq[3]); + + debug("rtable->exlusive_irqs = %x\n", rtable->exclusive_irqs); - debug("rtable->slots[loop].irq[0].bitmap = %x\n", + debug("rtable->slots[loop].irq[0].bitmap = %x\n", rtable->slots[loop].irq[0].bitmap); - debug("rtable->slots[loop].irq[1].bitmap = %x\n", + debug("rtable->slots[loop].irq[1].bitmap = %x\n", rtable->slots[loop].irq[1].bitmap); - debug("rtable->slots[loop].irq[2].bitmap = %x\n", + debug("rtable->slots[loop].irq[2].bitmap = %x\n", rtable->slots[loop].irq[2].bitmap); - debug("rtable->slots[loop].irq[3].bitmap = %x\n", + debug("rtable->slots[loop].irq[3].bitmap = %x\n", rtable->slots[loop].irq[3].bitmap); - debug("rtable->slots[loop].irq[0].link = %x\n", + debug("rtable->slots[loop].irq[0].link = %x\n", rtable->slots[loop].irq[0].link); - debug("rtable->slots[loop].irq[1].link = %x\n", + debug("rtable->slots[loop].irq[1].link = %x\n", rtable->slots[loop].irq[1].link); - debug("rtable->slots[loop].irq[2].link = %x\n", + debug("rtable->slots[loop].irq[2].link = %x\n", rtable->slots[loop].irq[2].link); - debug("rtable->slots[loop].irq[3].link = %x\n", + debug("rtable->slots[loop].irq[3].link = %x\n", rtable->slots[loop].irq[3].link); - debug("end of init_devno\n"); - kfree(rtable); - return 0; - } + debug("end of init_devno\n"); + kfree(rtable); + return 0; } } diff --git a/drivers/pci/htirq.c b/drivers/pci/htirq.c index 6808d8333ec..737a1c44b07 100644 --- a/drivers/pci/htirq.c +++ b/drivers/pci/htirq.c @@ -98,6 +98,7 @@ int __ht_create_irq(struct pci_dev *dev, int idx, ht_irq_update_t *update) int max_irq; int pos; int irq; + int node; pos = pci_find_ht_capability(dev, HT_CAPTYPE_IRQ); if (!pos) @@ -125,7 +126,8 @@ int __ht_create_irq(struct pci_dev *dev, int idx, ht_irq_update_t *update) cfg->msg.address_lo = 0xffffffff; cfg->msg.address_hi = 0xffffffff; - irq = create_irq(); + node = dev_to_node(&dev->dev); + irq = create_irq_nr(0, node); if (irq <= 0) { kfree(cfg); diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index a563fbe559d..cd389162735 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -1972,15 +1972,6 @@ static int __init init_dmars(void) } } -#ifdef CONFIG_INTR_REMAP - if (!intr_remapping_enabled) { - ret = enable_intr_remapping(0); - if (ret) - printk(KERN_ERR - "IOMMU: enable interrupt remapping failed\n"); - } -#endif - /* * For each rmrr * for each dev attached to rmrr diff --git a/drivers/pci/intr_remapping.c b/drivers/pci/intr_remapping.c index f5e0ea724a6..3a0cb0bb059 100644 --- a/drivers/pci/intr_remapping.c +++ b/drivers/pci/intr_remapping.c @@ -15,6 +15,14 @@ static struct ioapic_scope ir_ioapic[MAX_IO_APICS]; static int ir_ioapic_num; int intr_remapping_enabled; +static int disable_intremap; +static __init int setup_nointremap(char *str) +{ + disable_intremap = 1; + return 0; +} +early_param("nointremap", setup_nointremap); + struct irq_2_iommu { struct intel_iommu *iommu; u16 irte_index; @@ -23,15 +31,12 @@ struct irq_2_iommu { }; #ifdef CONFIG_GENERIC_HARDIRQS -static struct irq_2_iommu *get_one_free_irq_2_iommu(int cpu) +static struct irq_2_iommu *get_one_free_irq_2_iommu(int node) { struct irq_2_iommu *iommu; - int node; - - node = cpu_to_node(cpu); iommu = kzalloc_node(sizeof(*iommu), GFP_ATOMIC, node); - printk(KERN_DEBUG "alloc irq_2_iommu on cpu %d node %d\n", cpu, node); + printk(KERN_DEBUG "alloc irq_2_iommu on node %d\n", node); return iommu; } @@ -48,7 +53,7 @@ static struct irq_2_iommu *irq_2_iommu(unsigned int irq) return desc->irq_2_iommu; } -static struct irq_2_iommu *irq_2_iommu_alloc_cpu(unsigned int irq, int cpu) +static struct irq_2_iommu *irq_2_iommu_alloc_node(unsigned int irq, int node) { struct irq_desc *desc; struct irq_2_iommu *irq_iommu; @@ -56,7 +61,7 @@ static struct irq_2_iommu *irq_2_iommu_alloc_cpu(unsigned int irq, int cpu) /* * alloc irq desc if not allocated already. */ - desc = irq_to_desc_alloc_cpu(irq, cpu); + desc = irq_to_desc_alloc_node(irq, node); if (!desc) { printk(KERN_INFO "can not get irq_desc for %d\n", irq); return NULL; @@ -65,14 +70,14 @@ static struct irq_2_iommu *irq_2_iommu_alloc_cpu(unsigned int irq, int cpu) irq_iommu = desc->irq_2_iommu; if (!irq_iommu) - desc->irq_2_iommu = get_one_free_irq_2_iommu(cpu); + desc->irq_2_iommu = get_one_free_irq_2_iommu(node); return desc->irq_2_iommu; } static struct irq_2_iommu *irq_2_iommu_alloc(unsigned int irq) { - return irq_2_iommu_alloc_cpu(irq, boot_cpu_id); + return irq_2_iommu_alloc_node(irq, cpu_to_node(boot_cpu_id)); } #else /* !CONFIG_SPARSE_IRQ */ @@ -423,20 +428,6 @@ static void iommu_set_intr_remapping(struct intel_iommu *iommu, int mode) readl, (sts & DMA_GSTS_IRTPS), sts); spin_unlock_irqrestore(&iommu->register_lock, flags); - if (mode == 0) { - spin_lock_irqsave(&iommu->register_lock, flags); - - /* enable comaptiblity format interrupt pass through */ - cmd = iommu->gcmd | DMA_GCMD_CFI; - iommu->gcmd |= DMA_GCMD_CFI; - writel(cmd, iommu->reg + DMAR_GCMD_REG); - - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, - readl, (sts & DMA_GSTS_CFIS), sts); - - spin_unlock_irqrestore(&iommu->register_lock, flags); - } - /* * global invalidation of interrupt entry cache before enabling * interrupt-remapping. @@ -516,6 +507,23 @@ end: spin_unlock_irqrestore(&iommu->register_lock, flags); } +int __init intr_remapping_supported(void) +{ + struct dmar_drhd_unit *drhd; + + if (disable_intremap) + return 0; + + for_each_drhd_unit(drhd) { + struct intel_iommu *iommu = drhd->iommu; + + if (!ecap_ir_support(iommu->ecap)) + return 0; + } + + return 1; +} + int __init enable_intr_remapping(int eim) { struct dmar_drhd_unit *drhd; diff --git a/drivers/pnp/pnpacpi/rsparser.c b/drivers/pnp/pnpacpi/rsparser.c index adf17856bac..7f207f335be 100644 --- a/drivers/pnp/pnpacpi/rsparser.c +++ b/drivers/pnp/pnpacpi/rsparser.c @@ -123,7 +123,7 @@ static void pnpacpi_parse_allocated_irqresource(struct pnp_dev *dev, } flags = irq_flags(triggering, polarity, shareable); - irq = acpi_register_gsi(gsi, triggering, polarity); + irq = acpi_register_gsi(&dev->dev, gsi, triggering, polarity); if (irq >= 0) pcibios_penalize_isa_irq(irq, 1); else diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index e1716f14cd4..91e316fe652 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -1065,6 +1065,7 @@ sg_ioctl(struct inode *inode, struct file *filp, return blk_trace_setup(sdp->device->request_queue, sdp->disk->disk_name, MKDEV(SCSI_GENERIC_MAJOR, sdp->index), + NULL, (char *)arg); case BLKTRACESTART: return blk_trace_startstop(sdp->device->request_queue, 1); diff --git a/drivers/xen/Kconfig b/drivers/xen/Kconfig index 8ac9cddac57..cab100acf98 100644 --- a/drivers/xen/Kconfig +++ b/drivers/xen/Kconfig @@ -18,6 +18,16 @@ config XEN_SCRUB_PAGES secure, but slightly less efficient. If in doubt, say yes. +config XEN_DEV_EVTCHN + tristate "Xen /dev/xen/evtchn device" + depends on XEN + default y + help + The evtchn driver allows a userspace process to triger event + channels and to receive notification of an event channel + firing. + If in doubt, say yes. + config XENFS tristate "Xen filesystem" depends on XEN @@ -41,3 +51,13 @@ config XEN_COMPAT_XENFS a xen platform. If in doubt, say yes. +config XEN_SYS_HYPERVISOR + bool "Create xen entries under /sys/hypervisor" + depends on XEN && SYSFS + select SYS_HYPERVISOR + default y + help + Create entries under /sys/hypervisor describing the Xen + hypervisor environment. When running native or in another + virtual environment, /sys/hypervisor will still be present, + but will have no xen contents.
\ No newline at end of file diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile index ff8accc9e10..ec2a39b1e26 100644 --- a/drivers/xen/Makefile +++ b/drivers/xen/Makefile @@ -4,4 +4,6 @@ obj-y += xenbus/ obj-$(CONFIG_HOTPLUG_CPU) += cpu_hotplug.o obj-$(CONFIG_XEN_XENCOMM) += xencomm.o obj-$(CONFIG_XEN_BALLOON) += balloon.o -obj-$(CONFIG_XENFS) += xenfs/
\ No newline at end of file +obj-$(CONFIG_XEN_DEV_EVTCHN) += evtchn.o +obj-$(CONFIG_XENFS) += xenfs/ +obj-$(CONFIG_XEN_SYS_HYPERVISOR) += sys-hypervisor.o
\ No newline at end of file diff --git a/drivers/xen/events.c b/drivers/xen/events.c index 30963af5dba..891d2e90753 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -151,6 +151,12 @@ static unsigned int evtchn_from_irq(unsigned irq) return info_for_irq(irq)->evtchn; } +unsigned irq_from_evtchn(unsigned int evtchn) +{ + return evtchn_to_irq[evtchn]; +} +EXPORT_SYMBOL_GPL(irq_from_evtchn); + static enum ipi_vector ipi_from_irq(unsigned irq) { struct irq_info *info = info_for_irq(irq); @@ -335,7 +341,7 @@ static int find_unbound_irq(void) if (irq == nr_irqs) panic("No available IRQ to bind to: increase nr_irqs!\n"); - desc = irq_to_desc_alloc_cpu(irq, 0); + desc = irq_to_desc_alloc_node(irq, 0); if (WARN_ON(desc == NULL)) return -1; @@ -688,13 +694,13 @@ void rebind_evtchn_irq(int evtchn, int irq) } /* Rebind an evtchn so that it gets delivered to a specific cpu */ -static void rebind_irq_to_cpu(unsigned irq, unsigned tcpu) +static int rebind_irq_to_cpu(unsigned irq, unsigned tcpu) { struct evtchn_bind_vcpu bind_vcpu; int evtchn = evtchn_from_irq(irq); if (!VALID_EVTCHN(evtchn)) - return; + return -1; /* Send future instances of this interrupt to other vcpu. */ bind_vcpu.port = evtchn; @@ -707,13 +713,15 @@ static void rebind_irq_to_cpu(unsigned irq, unsigned tcpu) */ if (HYPERVISOR_event_channel_op(EVTCHNOP_bind_vcpu, &bind_vcpu) >= 0) bind_evtchn_to_cpu(evtchn, tcpu); -} + return 0; +} -static void set_affinity_irq(unsigned irq, const struct cpumask *dest) +static int set_affinity_irq(unsigned irq, const struct cpumask *dest) { unsigned tcpu = cpumask_first(dest); - rebind_irq_to_cpu(irq, tcpu); + + return rebind_irq_to_cpu(irq, tcpu); } int resend_irq_on_evtchn(unsigned int irq) diff --git a/drivers/xen/evtchn.c b/drivers/xen/evtchn.c new file mode 100644 index 00000000000..af031950f9b --- /dev/null +++ b/drivers/xen/evtchn.c @@ -0,0 +1,507 @@ +/****************************************************************************** + * evtchn.c + * + * Driver for receiving and demuxing event-channel signals. + * + * Copyright (c) 2004-2005, K A Fraser + * Multi-process extensions Copyright (c) 2004, Steven Smith + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include <linux/module.h> +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/slab.h> +#include <linux/string.h> +#include <linux/errno.h> +#include <linux/fs.h> +#include <linux/errno.h> +#include <linux/miscdevice.h> +#include <linux/major.h> +#include <linux/proc_fs.h> +#include <linux/stat.h> +#include <linux/poll.h> +#include <linux/irq.h> +#include <linux/init.h> +#include <linux/gfp.h> +#include <linux/mutex.h> +#include <linux/cpu.h> +#include <xen/events.h> +#include <xen/evtchn.h> +#include <asm/xen/hypervisor.h> + +struct per_user_data { + struct mutex bind_mutex; /* serialize bind/unbind operations */ + + /* Notification ring, accessed via /dev/xen/evtchn. */ +#define EVTCHN_RING_SIZE (PAGE_SIZE / sizeof(evtchn_port_t)) +#define EVTCHN_RING_MASK(_i) ((_i)&(EVTCHN_RING_SIZE-1)) + evtchn_port_t *ring; + unsigned int ring_cons, ring_prod, ring_overflow; + struct mutex ring_cons_mutex; /* protect against concurrent readers */ + + /* Processes wait on this queue when ring is empty. */ + wait_queue_head_t evtchn_wait; + struct fasync_struct *evtchn_async_queue; + const char *name; +}; + +/* Who's bound to each port? */ +static struct per_user_data *port_user[NR_EVENT_CHANNELS]; +static DEFINE_SPINLOCK(port_user_lock); /* protects port_user[] and ring_prod */ + +irqreturn_t evtchn_interrupt(int irq, void *data) +{ + unsigned int port = (unsigned long)data; + struct per_user_data *u; + + spin_lock(&port_user_lock); + + u = port_user[port]; + + disable_irq_nosync(irq); + + if ((u->ring_prod - u->ring_cons) < EVTCHN_RING_SIZE) { + u->ring[EVTCHN_RING_MASK(u->ring_prod)] = port; + wmb(); /* Ensure ring contents visible */ + if (u->ring_cons == u->ring_prod++) { + wake_up_interruptible(&u->evtchn_wait); + kill_fasync(&u->evtchn_async_queue, + SIGIO, POLL_IN); + } + } else { + u->ring_overflow = 1; + } + + spin_unlock(&port_user_lock); + + return IRQ_HANDLED; +} + +static ssize_t evtchn_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + int rc; + unsigned int c, p, bytes1 = 0, bytes2 = 0; + struct per_user_data *u = file->private_data; + + /* Whole number of ports. */ + count &= ~(sizeof(evtchn_port_t)-1); + + if (count == 0) + return 0; + + if (count > PAGE_SIZE) + count = PAGE_SIZE; + + for (;;) { + mutex_lock(&u->ring_cons_mutex); + + rc = -EFBIG; + if (u->ring_overflow) + goto unlock_out; + + c = u->ring_cons; + p = u->ring_prod; + if (c != p) + break; + + mutex_unlock(&u->ring_cons_mutex); + + if (file->f_flags & O_NONBLOCK) + return -EAGAIN; + + rc = wait_event_interruptible(u->evtchn_wait, + u->ring_cons != u->ring_prod); + if (rc) + return rc; + } + + /* Byte lengths of two chunks. Chunk split (if any) is at ring wrap. */ + if (((c ^ p) & EVTCHN_RING_SIZE) != 0) { + bytes1 = (EVTCHN_RING_SIZE - EVTCHN_RING_MASK(c)) * + sizeof(evtchn_port_t); + bytes2 = EVTCHN_RING_MASK(p) * sizeof(evtchn_port_t); + } else { + bytes1 = (p - c) * sizeof(evtchn_port_t); + bytes2 = 0; + } + + /* Truncate chunks according to caller's maximum byte count. */ + if (bytes1 > count) { + bytes1 = count; + bytes2 = 0; + } else if ((bytes1 + bytes2) > count) { + bytes2 = count - bytes1; + } + + rc = -EFAULT; + rmb(); /* Ensure that we see the port before we copy it. */ + if (copy_to_user(buf, &u->ring[EVTCHN_RING_MASK(c)], bytes1) || + ((bytes2 != 0) && + copy_to_user(&buf[bytes1], &u->ring[0], bytes2))) + goto unlock_out; + + u->ring_cons += (bytes1 + bytes2) / sizeof(evtchn_port_t); + rc = bytes1 + bytes2; + + unlock_out: + mutex_unlock(&u->ring_cons_mutex); + return rc; +} + +static ssize_t evtchn_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) +{ + int rc, i; + evtchn_port_t *kbuf = (evtchn_port_t *)__get_free_page(GFP_KERNEL); + struct per_user_data *u = file->private_data; + + if (kbuf == NULL) + return -ENOMEM; + + /* Whole number of ports. */ + count &= ~(sizeof(evtchn_port_t)-1); + + rc = 0; + if (count == 0) + goto out; + + if (count > PAGE_SIZE) + count = PAGE_SIZE; + + rc = -EFAULT; + if (copy_from_user(kbuf, buf, count) != 0) + goto out; + + spin_lock_irq(&port_user_lock); + for (i = 0; i < (count/sizeof(evtchn_port_t)); i++) + if ((kbuf[i] < NR_EVENT_CHANNELS) && (port_user[kbuf[i]] == u)) + enable_irq(irq_from_evtchn(kbuf[i])); + spin_unlock_irq(&port_user_lock); + + rc = count; + + out: + free_page((unsigned long)kbuf); + return rc; +} + +static int evtchn_bind_to_user(struct per_user_data *u, int port) +{ + int rc = 0; + + /* + * Ports are never reused, so every caller should pass in a + * unique port. + * + * (Locking not necessary because we haven't registered the + * interrupt handler yet, and our caller has already + * serialized bind operations.) + */ + BUG_ON(port_user[port] != NULL); + port_user[port] = u; + + rc = bind_evtchn_to_irqhandler(port, evtchn_interrupt, IRQF_DISABLED, + u->name, (void *)(unsigned long)port); + if (rc >= 0) + rc = 0; + + return rc; +} + +static void evtchn_unbind_from_user(struct per_user_data *u, int port) +{ + int irq = irq_from_evtchn(port); + + unbind_from_irqhandler(irq, (void *)(unsigned long)port); + + /* make sure we unbind the irq handler before clearing the port */ + barrier(); + + port_user[port] = NULL; +} + +static long evtchn_ioctl(struct file *file, + unsigned int cmd, unsigned long arg) +{ + int rc; + struct per_user_data *u = file->private_data; + void __user *uarg = (void __user *) arg; + + /* Prevent bind from racing with unbind */ + mutex_lock(&u->bind_mutex); + + switch (cmd) { + case IOCTL_EVTCHN_BIND_VIRQ: { + struct ioctl_evtchn_bind_virq bind; + struct evtchn_bind_virq bind_virq; + + rc = -EFAULT; + if (copy_from_user(&bind, uarg, sizeof(bind))) + break; + + bind_virq.virq = bind.virq; + bind_virq.vcpu = 0; + rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_virq, + &bind_virq); + if (rc != 0) + break; + + rc = evtchn_bind_to_user(u, bind_virq.port); + if (rc == 0) + rc = bind_virq.port; + break; + } + + case IOCTL_EVTCHN_BIND_INTERDOMAIN: { + struct ioctl_evtchn_bind_interdomain bind; + struct evtchn_bind_interdomain bind_interdomain; + + rc = -EFAULT; + if (copy_from_user(&bind, uarg, sizeof(bind))) + break; + + bind_interdomain.remote_dom = bind.remote_domain; + bind_interdomain.remote_port = bind.remote_port; + rc = HYPERVISOR_event_channel_op(EVTCHNOP_bind_interdomain, + &bind_interdomain); + if (rc != 0) + break; + + rc = evtchn_bind_to_user(u, bind_interdomain.local_port); + if (rc == 0) + rc = bind_interdomain.local_port; + break; + } + + case IOCTL_EVTCHN_BIND_UNBOUND_PORT: { + struct ioctl_evtchn_bind_unbound_port bind; + struct evtchn_alloc_unbound alloc_unbound; + + rc = -EFAULT; + if (copy_from_user(&bind, uarg, sizeof(bind))) + break; + + alloc_unbound.dom = DOMID_SELF; + alloc_unbound.remote_dom = bind.remote_domain; + rc = HYPERVISOR_event_channel_op(EVTCHNOP_alloc_unbound, + &alloc_unbound); + if (rc != 0) + break; + + rc = evtchn_bind_to_user(u, alloc_unbound.port); + if (rc == 0) + rc = alloc_unbound.port; + break; + } + + case IOCTL_EVTCHN_UNBIND: { + struct ioctl_evtchn_unbind unbind; + + rc = -EFAULT; + if (copy_from_user(&unbind, uarg, sizeof(unbind))) + break; + + rc = -EINVAL; + if (unbind.port >= NR_EVENT_CHANNELS) + break; + + spin_lock_irq(&port_user_lock); + + rc = -ENOTCONN; + if (port_user[unbind.port] != u) { + spin_unlock_irq(&port_user_lock); + break; + } + + evtchn_unbind_from_user(u, unbind.port); + + spin_unlock_irq(&port_user_lock); + + rc = 0; + break; + } + + case IOCTL_EVTCHN_NOTIFY: { + struct ioctl_evtchn_notify notify; + + rc = -EFAULT; + if (copy_from_user(¬ify, uarg, sizeof(notify))) + break; + + if (notify.port >= NR_EVENT_CHANNELS) { + rc = -EINVAL; + } else if (port_user[notify.port] != u) { + rc = -ENOTCONN; + } else { + notify_remote_via_evtchn(notify.port); + rc = 0; + } + break; + } + + case IOCTL_EVTCHN_RESET: { + /* Initialise the ring to empty. Clear errors. */ + mutex_lock(&u->ring_cons_mutex); + spin_lock_irq(&port_user_lock); + u->ring_cons = u->ring_prod = u->ring_overflow = 0; + spin_unlock_irq(&port_user_lock); + mutex_unlock(&u->ring_cons_mutex); + rc = 0; + break; + } + + default: + rc = -ENOSYS; + break; + } + mutex_unlock(&u->bind_mutex); + + return rc; +} + +static unsigned int evtchn_poll(struct file *file, poll_table *wait) +{ + unsigned int mask = POLLOUT | POLLWRNORM; + struct per_user_data *u = file->private_data; + + poll_wait(file, &u->evtchn_wait, wait); + if (u->ring_cons != u->ring_prod) + mask |= POLLIN | POLLRDNORM; + if (u->ring_overflow) + mask = POLLERR; + return mask; +} + +static int evtchn_fasync(int fd, struct file *filp, int on) +{ + struct per_user_data *u = filp->private_data; + return fasync_helper(fd, filp, on, &u->evtchn_async_queue); +} + +static int evtchn_open(struct inode *inode, struct file *filp) +{ + struct per_user_data *u; + + u = kzalloc(sizeof(*u), GFP_KERNEL); + if (u == NULL) + return -ENOMEM; + + u->name = kasprintf(GFP_KERNEL, "evtchn:%s", current->comm); + if (u->name == NULL) { + kfree(u); + return -ENOMEM; + } + + init_waitqueue_head(&u->evtchn_wait); + + u->ring = (evtchn_port_t *)__get_free_page(GFP_KERNEL); + if (u->ring == NULL) { + kfree(u->name); + kfree(u); + return -ENOMEM; + } + + mutex_init(&u->bind_mutex); + mutex_init(&u->ring_cons_mutex); + + filp->private_data = u; + + return 0; +} + +static int evtchn_release(struct inode *inode, struct file *filp) +{ + int i; + struct per_user_data *u = filp->private_data; + + spin_lock_irq(&port_user_lock); + + free_page((unsigned long)u->ring); + + for (i = 0; i < NR_EVENT_CHANNELS; i++) { + if (port_user[i] != u) + continue; + + evtchn_unbind_from_user(port_user[i], i); + } + + spin_unlock_irq(&port_user_lock); + + kfree(u->name); + kfree(u); + + return 0; +} + +static const struct file_operations evtchn_fops = { + .owner = THIS_MODULE, + .read = evtchn_read, + .write = evtchn_write, + .unlocked_ioctl = evtchn_ioctl, + .poll = evtchn_poll, + .fasync = evtchn_fasync, + .open = evtchn_open, + .release = evtchn_release, +}; + +static struct miscdevice evtchn_miscdev = { + .minor = MISC_DYNAMIC_MINOR, + .name = "evtchn", + .fops = &evtchn_fops, +}; +static int __init evtchn_init(void) +{ + int err; + + if (!xen_domain()) + return -ENODEV; + + spin_lock_init(&port_user_lock); + memset(port_user, 0, sizeof(port_user)); + + /* Create '/dev/misc/evtchn'. */ + err = misc_register(&evtchn_miscdev); + if (err != 0) { + printk(KERN_ALERT "Could not register /dev/misc/evtchn\n"); + return err; + } + + printk(KERN_INFO "Event-channel device installed.\n"); + + return 0; +} + +static void __exit evtchn_cleanup(void) +{ + misc_deregister(&evtchn_miscdev); +} + +module_init(evtchn_init); +module_exit(evtchn_cleanup); + +MODULE_LICENSE("GPL"); diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c index 4b5b84837ee..fddc2025dec 100644 --- a/drivers/xen/manage.c +++ b/drivers/xen/manage.c @@ -98,9 +98,8 @@ static void do_suspend(void) goto out; } - printk("suspending xenbus...\n"); - /* XXX use normal device tree? */ - xenbus_suspend(); + printk(KERN_DEBUG "suspending xenstore...\n"); + xs_suspend(); err = device_power_down(PMSG_SUSPEND); if (err) { @@ -116,9 +115,9 @@ static void do_suspend(void) if (!cancelled) { xen_arch_resume(); - xenbus_resume(); + xs_resume(); } else - xenbus_suspend_cancel(); + xs_suspend_cancel(); device_power_up(PMSG_RESUME); diff --git a/drivers/xen/sys-hypervisor.c b/drivers/xen/sys-hypervisor.c new file mode 100644 index 00000000000..88a60e03ccf --- /dev/null +++ b/drivers/xen/sys-hypervisor.c @@ -0,0 +1,445 @@ +/* + * copyright (c) 2006 IBM Corporation + * Authored by: Mike D. Day <ncmike@us.ibm.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/kobject.h> + +#include <asm/xen/hypervisor.h> +#include <asm/xen/hypercall.h> + +#include <xen/xenbus.h> +#include <xen/interface/xen.h> +#include <xen/interface/version.h> + +#define HYPERVISOR_ATTR_RO(_name) \ +static struct hyp_sysfs_attr _name##_attr = __ATTR_RO(_name) + +#define HYPERVISOR_ATTR_RW(_name) \ +static struct hyp_sysfs_attr _name##_attr = \ + __ATTR(_name, 0644, _name##_show, _name##_store) + +struct hyp_sysfs_attr { + struct attribute attr; + ssize_t (*show)(struct hyp_sysfs_attr *, char *); + ssize_t (*store)(struct hyp_sysfs_attr *, const char *, size_t); + void *hyp_attr_data; +}; + +static ssize_t type_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + return sprintf(buffer, "xen\n"); +} + +HYPERVISOR_ATTR_RO(type); + +static int __init xen_sysfs_type_init(void) +{ + return sysfs_create_file(hypervisor_kobj, &type_attr.attr); +} + +static void xen_sysfs_type_destroy(void) +{ + sysfs_remove_file(hypervisor_kobj, &type_attr.attr); +} + +/* xen version attributes */ +static ssize_t major_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int version = HYPERVISOR_xen_version(XENVER_version, NULL); + if (version) + return sprintf(buffer, "%d\n", version >> 16); + return -ENODEV; +} + +HYPERVISOR_ATTR_RO(major); + +static ssize_t minor_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int version = HYPERVISOR_xen_version(XENVER_version, NULL); + if (version) + return sprintf(buffer, "%d\n", version & 0xff); + return -ENODEV; +} + +HYPERVISOR_ATTR_RO(minor); + +static ssize_t extra_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int ret = -ENOMEM; + char *extra; + + extra = kmalloc(XEN_EXTRAVERSION_LEN, GFP_KERNEL); + if (extra) { + ret = HYPERVISOR_xen_version(XENVER_extraversion, extra); + if (!ret) + ret = sprintf(buffer, "%s\n", extra); + kfree(extra); + } + + return ret; +} + +HYPERVISOR_ATTR_RO(extra); + +static struct attribute *version_attrs[] = { + &major_attr.attr, + &minor_attr.attr, + &extra_attr.attr, + NULL +}; + +static struct attribute_group version_group = { + .name = "version", + .attrs = version_attrs, +}; + +static int __init xen_sysfs_version_init(void) +{ + return sysfs_create_group(hypervisor_kobj, &version_group); +} + +static void xen_sysfs_version_destroy(void) +{ + sysfs_remove_group(hypervisor_kobj, &version_group); +} + +/* UUID */ + +static ssize_t uuid_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + char *vm, *val; + int ret; + extern int xenstored_ready; + + if (!xenstored_ready) + return -EBUSY; + + vm = xenbus_read(XBT_NIL, "vm", "", NULL); + if (IS_ERR(vm)) + return PTR_ERR(vm); + val = xenbus_read(XBT_NIL, vm, "uuid", NULL); + kfree(vm); + if (IS_ERR(val)) + return PTR_ERR(val); + ret = sprintf(buffer, "%s\n", val); + kfree(val); + return ret; +} + +HYPERVISOR_ATTR_RO(uuid); + +static int __init xen_sysfs_uuid_init(void) +{ + return sysfs_create_file(hypervisor_kobj, &uuid_attr.attr); +} + +static void xen_sysfs_uuid_destroy(void) +{ + sysfs_remove_file(hypervisor_kobj, &uuid_attr.attr); +} + +/* xen compilation attributes */ + +static ssize_t compiler_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int ret = -ENOMEM; + struct xen_compile_info *info; + + info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL); + if (info) { + ret = HYPERVISOR_xen_version(XENVER_compile_info, info); + if (!ret) + ret = sprintf(buffer, "%s\n", info->compiler); + kfree(info); + } + + return ret; +} + +HYPERVISOR_ATTR_RO(compiler); + +static ssize_t compiled_by_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int ret = -ENOMEM; + struct xen_compile_info *info; + + info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL); + if (info) { + ret = HYPERVISOR_xen_version(XENVER_compile_info, info); + if (!ret) + ret = sprintf(buffer, "%s\n", info->compile_by); + kfree(info); + } + + return ret; +} + +HYPERVISOR_ATTR_RO(compiled_by); + +static ssize_t compile_date_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int ret = -ENOMEM; + struct xen_compile_info *info; + + info = kmalloc(sizeof(struct xen_compile_info), GFP_KERNEL); + if (info) { + ret = HYPERVISOR_xen_version(XENVER_compile_info, info); + if (!ret) + ret = sprintf(buffer, "%s\n", info->compile_date); + kfree(info); + } + + return ret; +} + +HYPERVISOR_ATTR_RO(compile_date); + +static struct attribute *xen_compile_attrs[] = { + &compiler_attr.attr, + &compiled_by_attr.attr, + &compile_date_attr.attr, + NULL +}; + +static struct attribute_group xen_compilation_group = { + .name = "compilation", + .attrs = xen_compile_attrs, +}; + +int __init static xen_compilation_init(void) +{ + return sysfs_create_group(hypervisor_kobj, &xen_compilation_group); +} + +static void xen_compilation_destroy(void) +{ + sysfs_remove_group(hypervisor_kobj, &xen_compilation_group); +} + +/* xen properties info */ + +static ssize_t capabilities_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int ret = -ENOMEM; + char *caps; + + caps = kmalloc(XEN_CAPABILITIES_INFO_LEN, GFP_KERNEL); + if (caps) { + ret = HYPERVISOR_xen_version(XENVER_capabilities, caps); + if (!ret) + ret = sprintf(buffer, "%s\n", caps); + kfree(caps); + } + + return ret; +} + +HYPERVISOR_ATTR_RO(capabilities); + +static ssize_t changeset_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int ret = -ENOMEM; + char *cset; + + cset = kmalloc(XEN_CHANGESET_INFO_LEN, GFP_KERNEL); + if (cset) { + ret = HYPERVISOR_xen_version(XENVER_changeset, cset); + if (!ret) + ret = sprintf(buffer, "%s\n", cset); + kfree(cset); + } + + return ret; +} + +HYPERVISOR_ATTR_RO(changeset); + +static ssize_t virtual_start_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int ret = -ENOMEM; + struct xen_platform_parameters *parms; + + parms = kmalloc(sizeof(struct xen_platform_parameters), GFP_KERNEL); + if (parms) { + ret = HYPERVISOR_xen_version(XENVER_platform_parameters, + parms); + if (!ret) + ret = sprintf(buffer, "%lx\n", parms->virt_start); + kfree(parms); + } + + return ret; +} + +HYPERVISOR_ATTR_RO(virtual_start); + +static ssize_t pagesize_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + int ret; + + ret = HYPERVISOR_xen_version(XENVER_pagesize, NULL); + if (ret > 0) + ret = sprintf(buffer, "%x\n", ret); + + return ret; +} + +HYPERVISOR_ATTR_RO(pagesize); + +static ssize_t xen_feature_show(int index, char *buffer) +{ + ssize_t ret; + struct xen_feature_info info; + + info.submap_idx = index; + ret = HYPERVISOR_xen_version(XENVER_get_features, &info); + if (!ret) + ret = sprintf(buffer, "%08x", info.submap); + + return ret; +} + +static ssize_t features_show(struct hyp_sysfs_attr *attr, char *buffer) +{ + ssize_t len; + int i; + + len = 0; + for (i = XENFEAT_NR_SUBMAPS-1; i >= 0; i--) { + int ret = xen_feature_show(i, buffer + len); + if (ret < 0) { + if (len == 0) + len = ret; + break; + } + len += ret; + } + if (len > 0) + buffer[len++] = '\n'; + + return len; +} + +HYPERVISOR_ATTR_RO(features); + +static struct attribute *xen_properties_attrs[] = { + &capabilities_attr.attr, + &changeset_attr.attr, + &virtual_start_attr.attr, + &pagesize_attr.attr, + &features_attr.attr, + NULL +}; + +static struct attribute_group xen_properties_group = { + .name = "properties", + .attrs = xen_properties_attrs, +}; + +static int __init xen_properties_init(void) +{ + return sysfs_create_group(hypervisor_kobj, &xen_properties_group); +} + +static void xen_properties_destroy(void) +{ + sysfs_remove_group(hypervisor_kobj, &xen_properties_group); +} + +static int __init hyper_sysfs_init(void) +{ + int ret; + + if (!xen_domain()) + return -ENODEV; + + ret = xen_sysfs_type_init(); + if (ret) + goto out; + ret = xen_sysfs_version_init(); + if (ret) + goto version_out; + ret = xen_compilation_init(); + if (ret) + goto comp_out; + ret = xen_sysfs_uuid_init(); + if (ret) + goto uuid_out; + ret = xen_properties_init(); + if (ret) + goto prop_out; + + goto out; + +prop_out: + xen_sysfs_uuid_destroy(); +uuid_out: + xen_compilation_destroy(); +comp_out: + xen_sysfs_version_destroy(); +version_out: + xen_sysfs_type_destroy(); +out: + return ret; +} + +static void __exit hyper_sysfs_exit(void) +{ + xen_properties_destroy(); + xen_compilation_destroy(); + xen_sysfs_uuid_destroy(); + xen_sysfs_version_destroy(); + xen_sysfs_type_destroy(); + +} +module_init(hyper_sysfs_init); +module_exit(hyper_sysfs_exit); + +static ssize_t hyp_sysfs_show(struct kobject *kobj, + struct attribute *attr, + char *buffer) +{ + struct hyp_sysfs_attr *hyp_attr; + hyp_attr = container_of(attr, struct hyp_sysfs_attr, attr); + if (hyp_attr->show) + return hyp_attr->show(hyp_attr, buffer); + return 0; +} + +static ssize_t hyp_sysfs_store(struct kobject *kobj, + struct attribute *attr, + const char *buffer, + size_t len) +{ + struct hyp_sysfs_attr *hyp_attr; + hyp_attr = container_of(attr, struct hyp_sysfs_attr, attr); + if (hyp_attr->store) + return hyp_attr->store(hyp_attr, buffer, len); + return 0; +} + +static struct sysfs_ops hyp_sysfs_ops = { + .show = hyp_sysfs_show, + .store = hyp_sysfs_store, +}; + +static struct kobj_type hyp_sysfs_kobj_type = { + .sysfs_ops = &hyp_sysfs_ops, +}; + +static int __init hypervisor_subsys_init(void) +{ + if (!xen_domain()) + return -ENODEV; + + hypervisor_kobj->ktype = &hyp_sysfs_kobj_type; + return 0; +} +device_initcall(hypervisor_subsys_init); diff --git a/drivers/xen/xenbus/xenbus_probe.c b/drivers/xen/xenbus/xenbus_probe.c index 773d1cf2328..d42e25d5968 100644 --- a/drivers/xen/xenbus/xenbus_probe.c +++ b/drivers/xen/xenbus/xenbus_probe.c @@ -71,6 +71,9 @@ static int xenbus_probe_frontend(const char *type, const char *name); static void xenbus_dev_shutdown(struct device *_dev); +static int xenbus_dev_suspend(struct device *dev, pm_message_t state); +static int xenbus_dev_resume(struct device *dev); + /* If something in array of ids matches this device, return it. */ static const struct xenbus_device_id * match_device(const struct xenbus_device_id *arr, struct xenbus_device *dev) @@ -188,6 +191,9 @@ static struct xen_bus_type xenbus_frontend = { .remove = xenbus_dev_remove, .shutdown = xenbus_dev_shutdown, .dev_attrs = xenbus_dev_attrs, + + .suspend = xenbus_dev_suspend, + .resume = xenbus_dev_resume, }, }; @@ -654,6 +660,7 @@ void xenbus_dev_changed(const char *node, struct xen_bus_type *bus) kfree(root); } +EXPORT_SYMBOL_GPL(xenbus_dev_changed); static void frontend_changed(struct xenbus_watch *watch, const char **vec, unsigned int len) @@ -669,7 +676,7 @@ static struct xenbus_watch fe_watch = { .callback = frontend_changed, }; -static int suspend_dev(struct device *dev, void *data) +static int xenbus_dev_suspend(struct device *dev, pm_message_t state) { int err = 0; struct xenbus_driver *drv; @@ -682,35 +689,14 @@ static int suspend_dev(struct device *dev, void *data) drv = to_xenbus_driver(dev->driver); xdev = container_of(dev, struct xenbus_device, dev); if (drv->suspend) - err = drv->suspend(xdev); + err = drv->suspend(xdev, state); if (err) printk(KERN_WARNING "xenbus: suspend %s failed: %i\n", dev_name(dev), err); return 0; } -static int suspend_cancel_dev(struct device *dev, void *data) -{ - int err = 0; - struct xenbus_driver *drv; - struct xenbus_device *xdev; - - DPRINTK(""); - - if (dev->driver == NULL) - return 0; - drv = to_xenbus_driver(dev->driver); - xdev = container_of(dev, struct xenbus_device, dev); - if (drv->suspend_cancel) - err = drv->suspend_cancel(xdev); - if (err) - printk(KERN_WARNING - "xenbus: suspend_cancel %s failed: %i\n", - dev_name(dev), err); - return 0; -} - -static int resume_dev(struct device *dev, void *data) +static int xenbus_dev_resume(struct device *dev) { int err; struct xenbus_driver *drv; @@ -755,33 +741,6 @@ static int resume_dev(struct device *dev, void *data) return 0; } -void xenbus_suspend(void) -{ - DPRINTK(""); - - bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_dev); - xenbus_backend_suspend(suspend_dev); - xs_suspend(); -} -EXPORT_SYMBOL_GPL(xenbus_suspend); - -void xenbus_resume(void) -{ - xb_init_comms(); - xs_resume(); - bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, resume_dev); - xenbus_backend_resume(resume_dev); -} -EXPORT_SYMBOL_GPL(xenbus_resume); - -void xenbus_suspend_cancel(void) -{ - xs_suspend_cancel(); - bus_for_each_dev(&xenbus_frontend.bus, NULL, NULL, suspend_cancel_dev); - xenbus_backend_resume(suspend_cancel_dev); -} -EXPORT_SYMBOL_GPL(xenbus_suspend_cancel); - /* A flag to determine if xenstored is 'ready' (i.e. has started) */ int xenstored_ready = 0; diff --git a/drivers/xen/xenbus/xenbus_xs.c b/drivers/xen/xenbus/xenbus_xs.c index e325eab4724..eab33f1dbdf 100644 --- a/drivers/xen/xenbus/xenbus_xs.c +++ b/drivers/xen/xenbus/xenbus_xs.c @@ -673,6 +673,8 @@ void xs_resume(void) struct xenbus_watch *watch; char token[sizeof(watch) * 2 + 1]; + xb_init_comms(); + mutex_unlock(&xs_state.response_mutex); mutex_unlock(&xs_state.request_mutex); up_write(&xs_state.transaction_mutex); diff --git a/drivers/xen/xenfs/super.c b/drivers/xen/xenfs/super.c index 515741a8e6b..6559e0c752c 100644 --- a/drivers/xen/xenfs/super.c +++ b/drivers/xen/xenfs/super.c @@ -20,10 +20,27 @@ MODULE_DESCRIPTION("Xen filesystem"); MODULE_LICENSE("GPL"); +static ssize_t capabilities_read(struct file *file, char __user *buf, + size_t size, loff_t *off) +{ + char *tmp = ""; + + if (xen_initial_domain()) + tmp = "control_d\n"; + + return simple_read_from_buffer(buf, size, off, tmp, strlen(tmp)); +} + +static const struct file_operations capabilities_file_ops = { + .read = capabilities_read, +}; + static int xenfs_fill_super(struct super_block *sb, void *data, int silent) { static struct tree_descr xenfs_files[] = { - [2] = {"xenbus", &xenbus_file_ops, S_IRUSR|S_IWUSR}, + [1] = {}, + { "xenbus", &xenbus_file_ops, S_IRUSR|S_IWUSR }, + { "capabilities", &capabilities_file_ops, S_IRUGO }, {""}, }; @@ -26,10 +26,9 @@ #include <linux/mempool.h> #include <linux/workqueue.h> #include <linux/blktrace_api.h> -#include <trace/block.h> #include <scsi/sg.h> /* for struct sg_iovec */ -DEFINE_TRACE(block_split); +#include <trace/events/block.h> /* * Test patch to inline a certain number of bi_io_vec's inside the bio diff --git a/fs/exec.c b/fs/exec.c index 895823d0149..ad4f28c2327 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -33,6 +33,7 @@ #include <linux/string.h> #include <linux/init.h> #include <linux/pagemap.h> +#include <linux/perf_counter.h> #include <linux/highmem.h> #include <linux/spinlock.h> #include <linux/key.h> @@ -922,6 +923,7 @@ void set_task_comm(struct task_struct *tsk, char *buf) task_lock(tsk); strlcpy(tsk->comm, buf, sizeof(tsk->comm)); task_unlock(tsk); + perf_counter_comm(tsk); } int flush_old_exec(struct linux_binprm * bprm) @@ -990,6 +992,13 @@ int flush_old_exec(struct linux_binprm * bprm) current->personality &= ~bprm->per_clear; + /* + * Flush performance counters when crossing a + * security domain: + */ + if (!get_dumpable(current->mm)) + perf_counter_exit_task(current); + /* An exec changes our domain. We are no longer part of the thread group */ diff --git a/fs/proc/loadavg.c b/fs/proc/loadavg.c index 9bca39cf99e..1afa4dd4cae 100644 --- a/fs/proc/loadavg.c +++ b/fs/proc/loadavg.c @@ -12,20 +12,14 @@ static int loadavg_proc_show(struct seq_file *m, void *v) { - int a, b, c; - unsigned long seq; + unsigned long avnrun[3]; - do { - seq = read_seqbegin(&xtime_lock); - a = avenrun[0] + (FIXED_1/200); - b = avenrun[1] + (FIXED_1/200); - c = avenrun[2] + (FIXED_1/200); - } while (read_seqretry(&xtime_lock, seq)); + get_avenrun(avnrun, FIXED_1/200, 0); - seq_printf(m, "%d.%02d %d.%02d %d.%02d %ld/%d %d\n", - LOAD_INT(a), LOAD_FRAC(a), - LOAD_INT(b), LOAD_FRAC(b), - LOAD_INT(c), LOAD_FRAC(c), + seq_printf(m, "%lu.%02lu %lu.%02lu %lu.%02lu %ld/%d %d\n", + LOAD_INT(avnrun[0]), LOAD_FRAC(avnrun[0]), + LOAD_INT(avnrun[1]), LOAD_FRAC(avnrun[1]), + LOAD_INT(avnrun[2]), LOAD_FRAC(avnrun[2]), nr_running(), nr_threads, task_active_pid_ns(current)->last_pid); return 0; diff --git a/include/Kbuild b/include/Kbuild index d8c3e3cbf41..fe36accd432 100644 --- a/include/Kbuild +++ b/include/Kbuild @@ -8,3 +8,4 @@ header-y += mtd/ header-y += rdma/ header-y += video/ header-y += drm/ +header-y += xen/ diff --git a/include/asm-generic/atomic.h b/include/asm-generic/atomic.h index 3673a13b670..81d3be459ef 100644 --- a/include/asm-generic/atomic.h +++ b/include/asm-generic/atomic.h @@ -134,7 +134,7 @@ static inline long atomic_long_add_unless(atomic_long_t *l, long a, long u) #define atomic_long_cmpxchg(l, old, new) \ (atomic64_cmpxchg((atomic64_t *)(l), (old), (new))) #define atomic_long_xchg(v, new) \ - (atomic64_xchg((atomic64_t *)(l), (new))) + (atomic64_xchg((atomic64_t *)(v), (new))) #else /* BITS_PER_LONG == 64 */ diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h index 8e6d0ca70ab..e410f602cab 100644 --- a/include/asm-generic/pgtable.h +++ b/include/asm-generic/pgtable.h @@ -280,17 +280,18 @@ static inline void ptep_modify_prot_commit(struct mm_struct *mm, #endif /* - * A facility to provide batching of the reload of page tables with the - * actual context switch code for paravirtualized guests. By convention, - * only one of the lazy modes (CPU, MMU) should be active at any given - * time, entry should never be nested, and entry and exits should always - * be paired. This is for sanity of maintaining and reasoning about the - * kernel code. + * A facility to provide batching of the reload of page tables and + * other process state with the actual context switch code for + * paravirtualized guests. By convention, only one of the batched + * update (lazy) modes (CPU, MMU) should be active at any given time, + * entry should never be nested, and entry and exits should always be + * paired. This is for sanity of maintaining and reasoning about the + * kernel code. In this case, the exit (end of the context switch) is + * in architecture-specific code, and so doesn't need a generic + * definition. */ -#ifndef __HAVE_ARCH_ENTER_LAZY_CPU_MODE -#define arch_enter_lazy_cpu_mode() do {} while (0) -#define arch_leave_lazy_cpu_mode() do {} while (0) -#define arch_flush_lazy_cpu_mode() do {} while (0) +#ifndef __HAVE_ARCH_START_CONTEXT_SWITCH +#define arch_start_context_switch(prev) do {} while (0) #endif #ifndef __HAVE_PFNMAP_TRACKING diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index 89853bcd27a..f1736ca7922 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -63,7 +63,7 @@ #define BRANCH_PROFILE() #endif -#ifdef CONFIG_EVENT_TRACER +#ifdef CONFIG_EVENT_TRACING #define FTRACE_EVENTS() VMLINUX_SYMBOL(__start_ftrace_events) = .; \ *(_ftrace_events) \ VMLINUX_SYMBOL(__stop_ftrace_events) = .; diff --git a/include/linux/acpi.h b/include/linux/acpi.h index 88be890ee3c..51b4b0a5ce8 100644 --- a/include/linux/acpi.h +++ b/include/linux/acpi.h @@ -119,7 +119,7 @@ extern int pci_mmcfg_config_num; extern int sbf_port; extern unsigned long acpi_realmode_flags; -int acpi_register_gsi (u32 gsi, int triggering, int polarity); +int acpi_register_gsi (struct device *dev, u32 gsi, int triggering, int polarity); int acpi_gsi_to_irq (u32 gsi, unsigned int *irq); #ifdef CONFIG_X86_IO_APIC diff --git a/include/linux/blktrace_api.h b/include/linux/blktrace_api.h index d960889e92e..7e4350ece0f 100644 --- a/include/linux/blktrace_api.h +++ b/include/linux/blktrace_api.h @@ -116,9 +116,9 @@ struct blk_io_trace { * The remap event */ struct blk_io_trace_remap { - __be32 device; __be32 device_from; - __be64 sector; + __be32 device_to; + __be64 sector_from; }; enum { @@ -165,8 +165,9 @@ struct blk_trace { extern int blk_trace_ioctl(struct block_device *, unsigned, char __user *); extern void blk_trace_shutdown(struct request_queue *); -extern int do_blk_trace_setup(struct request_queue *q, - char *name, dev_t dev, struct blk_user_trace_setup *buts); +extern int do_blk_trace_setup(struct request_queue *q, char *name, + dev_t dev, struct block_device *bdev, + struct blk_user_trace_setup *buts); extern void __trace_note_message(struct blk_trace *, const char *fmt, ...); /** @@ -193,22 +194,42 @@ extern void __trace_note_message(struct blk_trace *, const char *fmt, ...); extern void blk_add_driver_data(struct request_queue *q, struct request *rq, void *data, size_t len); extern int blk_trace_setup(struct request_queue *q, char *name, dev_t dev, + struct block_device *bdev, char __user *arg); extern int blk_trace_startstop(struct request_queue *q, int start); extern int blk_trace_remove(struct request_queue *q); +extern int blk_trace_init_sysfs(struct device *dev); extern struct attribute_group blk_trace_attr_group; #else /* !CONFIG_BLK_DEV_IO_TRACE */ -#define blk_trace_ioctl(bdev, cmd, arg) (-ENOTTY) -#define blk_trace_shutdown(q) do { } while (0) -#define do_blk_trace_setup(q, name, dev, buts) (-ENOTTY) -#define blk_add_driver_data(q, rq, data, len) do {} while (0) -#define blk_trace_setup(q, name, dev, arg) (-ENOTTY) -#define blk_trace_startstop(q, start) (-ENOTTY) -#define blk_trace_remove(q) (-ENOTTY) -#define blk_add_trace_msg(q, fmt, ...) do { } while (0) +# define blk_trace_ioctl(bdev, cmd, arg) (-ENOTTY) +# define blk_trace_shutdown(q) do { } while (0) +# define do_blk_trace_setup(q, name, dev, bdev, buts) (-ENOTTY) +# define blk_add_driver_data(q, rq, data, len) do {} while (0) +# define blk_trace_setup(q, name, dev, bdev, arg) (-ENOTTY) +# define blk_trace_startstop(q, start) (-ENOTTY) +# define blk_trace_remove(q) (-ENOTTY) +# define blk_add_trace_msg(q, fmt, ...) do { } while (0) +static inline int blk_trace_init_sysfs(struct device *dev) +{ + return 0; +} #endif /* CONFIG_BLK_DEV_IO_TRACE */ + +#if defined(CONFIG_EVENT_TRACING) && defined(CONFIG_BLOCK) + +static inline int blk_cmd_buf_len(struct request *rq) +{ + return blk_pc_request(rq) ? rq->cmd_len * 3 : 1; +} + +extern void blk_dump_cmd(char *buf, struct request *rq); +extern void blk_fill_rwbs(char *rwbs, u32 rw, int bytes); +extern void blk_fill_rwbs_rq(char *rwbs, struct request *rq); + +#endif /* CONFIG_EVENT_TRACING && CONFIG_BLOCK */ + #endif /* __KERNEL__ */ #endif diff --git a/include/linux/compat.h b/include/linux/compat.h index f2ded21f9a3..af931ee43dd 100644 --- a/include/linux/compat.h +++ b/include/linux/compat.h @@ -222,6 +222,8 @@ int copy_siginfo_from_user32(siginfo_t *to, struct compat_siginfo __user *from); int copy_siginfo_to_user32(struct compat_siginfo __user *to, siginfo_t *from); int get_compat_sigevent(struct sigevent *event, const struct compat_sigevent __user *u_event); +long compat_sys_rt_tgsigqueueinfo(compat_pid_t tgid, compat_pid_t pid, int sig, + struct compat_siginfo __user *uinfo); static inline int compat_timeval_compare(struct compat_timeval *lhs, struct compat_timeval *rhs) diff --git a/include/linux/dma-debug.h b/include/linux/dma-debug.h index 28d53cb7b5a..171ad8aedc8 100644 --- a/include/linux/dma-debug.h +++ b/include/linux/dma-debug.h @@ -32,6 +32,8 @@ extern void dma_debug_add_bus(struct bus_type *bus); extern void dma_debug_init(u32 num_entries); +extern int dma_debug_resize_entries(u32 num_entries); + extern void debug_dma_map_page(struct device *dev, struct page *page, size_t offset, size_t size, int direction, dma_addr_t dma_addr, @@ -91,6 +93,11 @@ static inline void dma_debug_init(u32 num_entries) { } +static inline int dma_debug_resize_entries(u32 num_entries) +{ + return 0; +} + static inline void debug_dma_map_page(struct device *dev, struct page *page, size_t offset, size_t size, int direction, dma_addr_t dma_addr, diff --git a/include/linux/dmar.h b/include/linux/dmar.h index e397dc342cd..10ff5c49882 100644 --- a/include/linux/dmar.h +++ b/include/linux/dmar.h @@ -108,6 +108,7 @@ struct irte { }; #ifdef CONFIG_INTR_REMAP extern int intr_remapping_enabled; +extern int intr_remapping_supported(void); extern int enable_intr_remapping(int); extern void disable_intr_remapping(void); extern int reenable_intr_remapping(int); @@ -157,6 +158,8 @@ static inline struct intel_iommu *map_ioapic_to_ir(int apic) } #define irq_remapped(irq) (0) #define enable_intr_remapping(mode) (-1) +#define disable_intr_remapping() (0) +#define reenable_intr_remapping(mode) (0) #define intr_remapping_enabled (0) #endif diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 8a0c2f221e6..39b95c56587 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -233,8 +233,6 @@ extern int ftrace_arch_read_dyn_info(char *buf, int size); extern int skip_trace(unsigned long ip); -extern void ftrace_release(void *start, unsigned long size); - extern void ftrace_disable_daemon(void); extern void ftrace_enable_daemon(void); #else @@ -325,13 +323,8 @@ static inline void __ftrace_enabled_restore(int enabled) #ifdef CONFIG_FTRACE_MCOUNT_RECORD extern void ftrace_init(void); -extern void ftrace_init_module(struct module *mod, - unsigned long *start, unsigned long *end); #else static inline void ftrace_init(void) { } -static inline void -ftrace_init_module(struct module *mod, - unsigned long *start, unsigned long *end) { } #endif /* @@ -368,6 +361,7 @@ struct ftrace_ret_stack { unsigned long ret; unsigned long func; unsigned long long calltime; + unsigned long long subtime; }; /* @@ -379,8 +373,6 @@ extern void return_to_handler(void); extern int ftrace_push_return_trace(unsigned long ret, unsigned long func, int *depth); -extern void -ftrace_pop_return_trace(struct ftrace_graph_ret *trace, unsigned long *ret); /* * Sometimes we don't want to trace a function with the function @@ -496,8 +488,15 @@ static inline int test_tsk_trace_graph(struct task_struct *tsk) extern int ftrace_dump_on_oops; +#ifdef CONFIG_PREEMPT +#define INIT_TRACE_RECURSION .trace_recursion = 0, +#endif + #endif /* CONFIG_TRACING */ +#ifndef INIT_TRACE_RECURSION +#define INIT_TRACE_RECURSION +#endif #ifdef CONFIG_HW_BRANCH_TRACER diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h new file mode 100644 index 00000000000..5c093ffc655 --- /dev/null +++ b/include/linux/ftrace_event.h @@ -0,0 +1,172 @@ +#ifndef _LINUX_FTRACE_EVENT_H +#define _LINUX_FTRACE_EVENT_H + +#include <linux/trace_seq.h> +#include <linux/ring_buffer.h> +#include <linux/percpu.h> + +struct trace_array; +struct tracer; +struct dentry; + +DECLARE_PER_CPU(struct trace_seq, ftrace_event_seq); + +struct trace_print_flags { + unsigned long mask; + const char *name; +}; + +const char *ftrace_print_flags_seq(struct trace_seq *p, const char *delim, + unsigned long flags, + const struct trace_print_flags *flag_array); + +const char *ftrace_print_symbols_seq(struct trace_seq *p, unsigned long val, + const struct trace_print_flags *symbol_array); + +/* + * The trace entry - the most basic unit of tracing. This is what + * is printed in the end as a single line in the trace output, such as: + * + * bash-15816 [01] 235.197585: idle_cpu <- irq_enter + */ +struct trace_entry { + unsigned short type; + unsigned char flags; + unsigned char preempt_count; + int pid; + int tgid; +}; + +#define FTRACE_MAX_EVENT \ + ((1 << (sizeof(((struct trace_entry *)0)->type) * 8)) - 1) + +/* + * Trace iterator - used by printout routines who present trace + * results to users and which routines might sleep, etc: + */ +struct trace_iterator { + struct trace_array *tr; + struct tracer *trace; + void *private; + int cpu_file; + struct mutex mutex; + struct ring_buffer_iter *buffer_iter[NR_CPUS]; + unsigned long iter_flags; + + /* The below is zeroed out in pipe_read */ + struct trace_seq seq; + struct trace_entry *ent; + int cpu; + u64 ts; + + loff_t pos; + long idx; + + cpumask_var_t started; +}; + + +typedef enum print_line_t (*trace_print_func)(struct trace_iterator *iter, + int flags); +struct trace_event { + struct hlist_node node; + struct list_head list; + int type; + trace_print_func trace; + trace_print_func raw; + trace_print_func hex; + trace_print_func binary; +}; + +extern int register_ftrace_event(struct trace_event *event); +extern int unregister_ftrace_event(struct trace_event *event); + +/* Return values for print_line callback */ +enum print_line_t { + TRACE_TYPE_PARTIAL_LINE = 0, /* Retry after flushing the seq */ + TRACE_TYPE_HANDLED = 1, + TRACE_TYPE_UNHANDLED = 2, /* Relay to other output functions */ + TRACE_TYPE_NO_CONSUME = 3 /* Handled but ask to not consume */ +}; + + +struct ring_buffer_event * +trace_current_buffer_lock_reserve(int type, unsigned long len, + unsigned long flags, int pc); +void trace_current_buffer_unlock_commit(struct ring_buffer_event *event, + unsigned long flags, int pc); +void trace_nowake_buffer_unlock_commit(struct ring_buffer_event *event, + unsigned long flags, int pc); +void trace_current_buffer_discard_commit(struct ring_buffer_event *event); + +void tracing_record_cmdline(struct task_struct *tsk); + +struct ftrace_event_call { + struct list_head list; + char *name; + char *system; + struct dentry *dir; + struct trace_event *event; + int enabled; + int (*regfunc)(void); + void (*unregfunc)(void); + int id; + int (*raw_init)(void); + int (*show_format)(struct trace_seq *s); + int (*define_fields)(void); + struct list_head fields; + int filter_active; + void *filter; + void *mod; + +#ifdef CONFIG_EVENT_PROFILE + atomic_t profile_count; + int (*profile_enable)(struct ftrace_event_call *); + void (*profile_disable)(struct ftrace_event_call *); +#endif +}; + +#define MAX_FILTER_PRED 32 +#define MAX_FILTER_STR_VAL 128 + +extern int init_preds(struct ftrace_event_call *call); +extern void destroy_preds(struct ftrace_event_call *call); +extern int filter_match_preds(struct ftrace_event_call *call, void *rec); +extern int filter_current_check_discard(struct ftrace_event_call *call, + void *rec, + struct ring_buffer_event *event); + +extern int trace_define_field(struct ftrace_event_call *call, char *type, + char *name, int offset, int size, int is_signed); + +#define is_signed_type(type) (((type)(-1)) < 0) + +int trace_set_clr_event(const char *system, const char *event, int set); + +/* + * The double __builtin_constant_p is because gcc will give us an error + * if we try to allocate the static variable to fmt if it is not a + * constant. Even with the outer if statement optimizing out. + */ +#define event_trace_printk(ip, fmt, args...) \ +do { \ + __trace_printk_check_format(fmt, ##args); \ + tracing_record_cmdline(current); \ + if (__builtin_constant_p(fmt)) { \ + static const char *trace_printk_fmt \ + __attribute__((section("__trace_printk_fmt"))) = \ + __builtin_constant_p(fmt) ? fmt : NULL; \ + \ + __trace_bprintk(ip, trace_printk_fmt, ##args); \ + } else \ + __trace_printk(ip, fmt, ##args); \ +} while (0) + +#define __common_field(type, item, is_signed) \ + ret = trace_define_field(event_call, #type, "common_" #item, \ + offsetof(typeof(field.ent), item), \ + sizeof(field.ent.item), is_signed); \ + if (ret) \ + return ret; + +#endif /* _LINUX_FTRACE_EVENT_H */ diff --git a/include/linux/futex.h b/include/linux/futex.h index 3bf5bb5a34f..34956c8fdeb 100644 --- a/include/linux/futex.h +++ b/include/linux/futex.h @@ -23,6 +23,8 @@ union ktime; #define FUTEX_TRYLOCK_PI 8 #define FUTEX_WAIT_BITSET 9 #define FUTEX_WAKE_BITSET 10 +#define FUTEX_WAIT_REQUEUE_PI 11 +#define FUTEX_CMP_REQUEUE_PI 12 #define FUTEX_PRIVATE_FLAG 128 #define FUTEX_CLOCK_REALTIME 256 @@ -38,6 +40,10 @@ union ktime; #define FUTEX_TRYLOCK_PI_PRIVATE (FUTEX_TRYLOCK_PI | FUTEX_PRIVATE_FLAG) #define FUTEX_WAIT_BITSET_PRIVATE (FUTEX_WAIT_BITS | FUTEX_PRIVATE_FLAG) #define FUTEX_WAKE_BITSET_PRIVATE (FUTEX_WAKE_BITS | FUTEX_PRIVATE_FLAG) +#define FUTEX_WAIT_REQUEUE_PI_PRIVATE (FUTEX_WAIT_REQUEUE_PI | \ + FUTEX_PRIVATE_FLAG) +#define FUTEX_CMP_REQUEUE_PI_PRIVATE (FUTEX_CMP_REQUEUE_PI | \ + FUTEX_PRIVATE_FLAG) /* * Support for robust futexes: the kernel cleans up held futexes at diff --git a/include/linux/init_task.h b/include/linux/init_task.h index d87247d2641..b6b7cf23c2a 100644 --- a/include/linux/init_task.h +++ b/include/linux/init_task.h @@ -108,6 +108,15 @@ extern struct group_info init_groups; extern struct cred init_cred; +#ifdef CONFIG_PERF_COUNTERS +# define INIT_PERF_COUNTERS(tsk) \ + .perf_counter_mutex = \ + __MUTEX_INITIALIZER(tsk.perf_counter_mutex), \ + .perf_counter_list = LIST_HEAD_INIT(tsk.perf_counter_list), +#else +# define INIT_PERF_COUNTERS(tsk) +#endif + /* * INIT_TASK is used to set up the first task table, touch at * your own risk!. Base=0, limit=0x1fffff (=2MB) @@ -171,9 +180,11 @@ extern struct cred init_cred; }, \ .dirties = INIT_PROP_LOCAL_SINGLE(dirties), \ INIT_IDS \ + INIT_PERF_COUNTERS(tsk) \ INIT_TRACE_IRQFLAGS \ INIT_LOCKDEP \ INIT_FTRACE_GRAPH \ + INIT_TRACE_RECURSION \ } diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h index 91bb76f44f1..ff374ceface 100644 --- a/include/linux/interrupt.h +++ b/include/linux/interrupt.h @@ -566,6 +566,6 @@ struct irq_desc; extern int early_irq_init(void); extern int arch_probe_nr_irqs(void); extern int arch_early_irq_init(void); -extern int arch_init_chip_data(struct irq_desc *desc, int cpu); +extern int arch_init_chip_data(struct irq_desc *desc, int node); #endif diff --git a/include/linux/irq.h b/include/linux/irq.h index b7cbeed972e..eedbb8e5e0c 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -117,7 +117,7 @@ struct irq_chip { void (*eoi)(unsigned int irq); void (*end)(unsigned int irq); - void (*set_affinity)(unsigned int irq, + int (*set_affinity)(unsigned int irq, const struct cpumask *dest); int (*retrigger)(unsigned int irq); int (*set_type)(unsigned int irq, unsigned int flow_type); @@ -187,7 +187,7 @@ struct irq_desc { spinlock_t lock; #ifdef CONFIG_SMP cpumask_var_t affinity; - unsigned int cpu; + unsigned int node; #ifdef CONFIG_GENERIC_PENDING_IRQ cpumask_var_t pending_mask; #endif @@ -201,26 +201,23 @@ struct irq_desc { } ____cacheline_internodealigned_in_smp; extern void arch_init_copy_chip_data(struct irq_desc *old_desc, - struct irq_desc *desc, int cpu); + struct irq_desc *desc, int node); extern void arch_free_chip_data(struct irq_desc *old_desc, struct irq_desc *desc); #ifndef CONFIG_SPARSE_IRQ extern struct irq_desc irq_desc[NR_IRQS]; -#else /* CONFIG_SPARSE_IRQ */ -extern struct irq_desc *move_irq_desc(struct irq_desc *old_desc, int cpu); -#endif /* CONFIG_SPARSE_IRQ */ - -extern struct irq_desc *irq_to_desc_alloc_cpu(unsigned int irq, int cpu); +#endif -static inline struct irq_desc * -irq_remap_to_desc(unsigned int irq, struct irq_desc *desc) -{ -#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC - return irq_to_desc(irq); +#ifdef CONFIG_NUMA_IRQ_DESC +extern struct irq_desc *move_irq_desc(struct irq_desc *old_desc, int node); #else +static inline struct irq_desc *move_irq_desc(struct irq_desc *desc, int node) +{ return desc; -#endif } +#endif + +extern struct irq_desc *irq_to_desc_alloc_node(unsigned int irq, int node); /* * Migration helpers for obsolete names, they will go away: @@ -386,7 +383,7 @@ extern void set_irq_noprobe(unsigned int irq); extern void set_irq_probe(unsigned int irq); /* Handle dynamic irq creation and destruction */ -extern unsigned int create_irq_nr(unsigned int irq_want); +extern unsigned int create_irq_nr(unsigned int irq_want, int node); extern int create_irq(void); extern void destroy_irq(unsigned int irq); @@ -424,47 +421,48 @@ extern int set_irq_msi(unsigned int irq, struct msi_desc *entry); #ifdef CONFIG_SMP /** - * init_alloc_desc_masks - allocate cpumasks for irq_desc + * alloc_desc_masks - allocate cpumasks for irq_desc * @desc: pointer to irq_desc struct * @cpu: cpu which will be handling the cpumasks * @boot: true if need bootmem * * Allocates affinity and pending_mask cpumask if required. * Returns true if successful (or not required). - * Side effect: affinity has all bits set, pending_mask has all bits clear. */ -static inline bool init_alloc_desc_masks(struct irq_desc *desc, int cpu, +static inline bool alloc_desc_masks(struct irq_desc *desc, int node, bool boot) { - int node; - +#ifdef CONFIG_CPUMASK_OFFSTACK if (boot) { alloc_bootmem_cpumask_var(&desc->affinity); - cpumask_setall(desc->affinity); #ifdef CONFIG_GENERIC_PENDING_IRQ alloc_bootmem_cpumask_var(&desc->pending_mask); - cpumask_clear(desc->pending_mask); #endif return true; } - node = cpu_to_node(cpu); - if (!alloc_cpumask_var_node(&desc->affinity, GFP_ATOMIC, node)) return false; - cpumask_setall(desc->affinity); #ifdef CONFIG_GENERIC_PENDING_IRQ if (!alloc_cpumask_var_node(&desc->pending_mask, GFP_ATOMIC, node)) { free_cpumask_var(desc->affinity); return false; } - cpumask_clear(desc->pending_mask); +#endif #endif return true; } +static inline void init_desc_masks(struct irq_desc *desc) +{ + cpumask_setall(desc->affinity); +#ifdef CONFIG_GENERIC_PENDING_IRQ + cpumask_clear(desc->pending_mask); +#endif +} + /** * init_copy_desc_masks - copy cpumasks for irq_desc * @old_desc: pointer to old irq_desc struct @@ -478,7 +476,7 @@ static inline bool init_alloc_desc_masks(struct irq_desc *desc, int cpu, static inline void init_copy_desc_masks(struct irq_desc *old_desc, struct irq_desc *new_desc) { -#ifdef CONFIG_CPUMASKS_OFFSTACK +#ifdef CONFIG_CPUMASK_OFFSTACK cpumask_copy(new_desc->affinity, old_desc->affinity); #ifdef CONFIG_GENERIC_PENDING_IRQ @@ -499,12 +497,16 @@ static inline void free_desc_masks(struct irq_desc *old_desc, #else /* !CONFIG_SMP */ -static inline bool init_alloc_desc_masks(struct irq_desc *desc, int cpu, +static inline bool alloc_desc_masks(struct irq_desc *desc, int node, bool boot) { return true; } +static inline void init_desc_masks(struct irq_desc *desc) +{ +} + static inline void init_copy_desc_masks(struct irq_desc *old_desc, struct irq_desc *new_desc) { diff --git a/include/linux/kernel_stat.h b/include/linux/kernel_stat.h index 0c8b89f28a9..a77c6007dc9 100644 --- a/include/linux/kernel_stat.h +++ b/include/linux/kernel_stat.h @@ -81,7 +81,12 @@ static inline unsigned int kstat_irqs(unsigned int irq) return sum; } + +/* + * Lock/unlock the current runqueue - to extract task statistics: + */ extern unsigned long long task_delta_exec(struct task_struct *); + extern void account_user_time(struct task_struct *, cputime_t, cputime_t); extern void account_system_time(struct task_struct *, int, cputime_t, cputime_t); extern void account_steal_time(cputime_t); diff --git a/include/linux/kmemtrace.h b/include/linux/kmemtrace.h new file mode 100644 index 00000000000..b616d3930c3 --- /dev/null +++ b/include/linux/kmemtrace.h @@ -0,0 +1,25 @@ +/* + * Copyright (C) 2008 Eduard - Gabriel Munteanu + * + * This file is released under GPL version 2. + */ + +#ifndef _LINUX_KMEMTRACE_H +#define _LINUX_KMEMTRACE_H + +#ifdef __KERNEL__ + +#include <trace/events/kmem.h> + +#ifdef CONFIG_KMEMTRACE +extern void kmemtrace_init(void); +#else +static inline void kmemtrace_init(void) +{ +} +#endif + +#endif /* __KERNEL__ */ + +#endif /* _LINUX_KMEMTRACE_H */ + diff --git a/include/linux/mm.h b/include/linux/mm.h index bff1f0d475c..9772d6cbfc8 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -19,6 +19,7 @@ struct anon_vma; struct file_ra_state; struct user_struct; struct writeback_control; +struct rlimit; #ifndef CONFIG_DISCONTIGMEM /* Don't use mapnrs, do it properly */ extern unsigned long max_mapnr; @@ -1031,8 +1032,6 @@ extern void add_active_range(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn); extern void remove_active_range(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn); -extern void push_node_boundaries(unsigned int nid, unsigned long start_pfn, - unsigned long end_pfn); extern void remove_all_active_ranges(void); extern unsigned long absent_pages_in_range(unsigned long start_pfn, unsigned long end_pfn); @@ -1319,8 +1318,8 @@ int vmemmap_populate_basepages(struct page *start_page, int vmemmap_populate(struct page *start_page, unsigned long pages, int node); void vmemmap_populate_print_last(void); -extern void *alloc_locked_buffer(size_t size); -extern void free_locked_buffer(void *buffer, size_t size); -extern void release_locked_buffer(void *buffer, size_t size); +extern int account_locked_memory(struct mm_struct *mm, struct rlimit *rlim, + size_t size); +extern void refund_locked_memory(struct mm_struct *mm, size_t size); #endif /* __KERNEL__ */ #endif /* _LINUX_MM_H */ diff --git a/include/linux/mmiotrace.h b/include/linux/mmiotrace.h index 3d1b7bde128..97491f78b08 100644 --- a/include/linux/mmiotrace.h +++ b/include/linux/mmiotrace.h @@ -30,6 +30,8 @@ extern unsigned int kmmio_count; extern int register_kmmio_probe(struct kmmio_probe *p); extern void unregister_kmmio_probe(struct kmmio_probe *p); +extern int kmmio_init(void); +extern void kmmio_cleanup(void); #ifdef CONFIG_MMIOTRACE /* kmmio is active by some kmmio_probes? */ diff --git a/include/linux/module.h b/include/linux/module.h index 627ac082e2a..a8f2c0aa4c3 100644 --- a/include/linux/module.h +++ b/include/linux/module.h @@ -337,6 +337,14 @@ struct module const char **trace_bprintk_fmt_start; unsigned int num_trace_bprintk_fmt; #endif +#ifdef CONFIG_EVENT_TRACING + struct ftrace_event_call *trace_events; + unsigned int num_trace_events; +#endif +#ifdef CONFIG_FTRACE_MCOUNT_RECORD + unsigned long *ftrace_callsites; + unsigned int num_ftrace_callsites; +#endif #ifdef CONFIG_MODULE_UNLOAD /* What modules depend on me? */ diff --git a/include/linux/mutex.h b/include/linux/mutex.h index 3069ec7e0ab..878cab4f5fc 100644 --- a/include/linux/mutex.h +++ b/include/linux/mutex.h @@ -150,5 +150,6 @@ extern int __must_check mutex_lock_killable(struct mutex *lock); */ extern int mutex_trylock(struct mutex *lock); extern void mutex_unlock(struct mutex *lock); +extern int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock); #endif diff --git a/include/linux/perf_counter.h b/include/linux/perf_counter.h new file mode 100644 index 00000000000..6e133954e2e --- /dev/null +++ b/include/linux/perf_counter.h @@ -0,0 +1,697 @@ +/* + * Performance counters: + * + * Copyright (C) 2008-2009, Thomas Gleixner <tglx@linutronix.de> + * Copyright (C) 2008-2009, Red Hat, Inc., Ingo Molnar + * Copyright (C) 2008-2009, Red Hat, Inc., Peter Zijlstra + * + * Data type definitions, declarations, prototypes. + * + * Started by: Thomas Gleixner and Ingo Molnar + * + * For licencing details see kernel-base/COPYING + */ +#ifndef _LINUX_PERF_COUNTER_H +#define _LINUX_PERF_COUNTER_H + +#include <linux/types.h> +#include <linux/ioctl.h> +#include <asm/byteorder.h> + +/* + * User-space ABI bits: + */ + +/* + * attr.type + */ +enum perf_type_id { + PERF_TYPE_HARDWARE = 0, + PERF_TYPE_SOFTWARE = 1, + PERF_TYPE_TRACEPOINT = 2, + PERF_TYPE_HW_CACHE = 3, + PERF_TYPE_RAW = 4, + + PERF_TYPE_MAX, /* non-ABI */ +}; + +/* + * Generalized performance counter event types, used by the + * attr.event_id parameter of the sys_perf_counter_open() + * syscall: + */ +enum perf_hw_id { + /* + * Common hardware events, generalized by the kernel: + */ + PERF_COUNT_HW_CPU_CYCLES = 0, + PERF_COUNT_HW_INSTRUCTIONS = 1, + PERF_COUNT_HW_CACHE_REFERENCES = 2, + PERF_COUNT_HW_CACHE_MISSES = 3, + PERF_COUNT_HW_BRANCH_INSTRUCTIONS = 4, + PERF_COUNT_HW_BRANCH_MISSES = 5, + PERF_COUNT_HW_BUS_CYCLES = 6, + + PERF_COUNT_HW_MAX, /* non-ABI */ +}; + +/* + * Generalized hardware cache counters: + * + * { L1-D, L1-I, LLC, ITLB, DTLB, BPU } x + * { read, write, prefetch } x + * { accesses, misses } + */ +enum perf_hw_cache_id { + PERF_COUNT_HW_CACHE_L1D = 0, + PERF_COUNT_HW_CACHE_L1I = 1, + PERF_COUNT_HW_CACHE_LL = 2, + PERF_COUNT_HW_CACHE_DTLB = 3, + PERF_COUNT_HW_CACHE_ITLB = 4, + PERF_COUNT_HW_CACHE_BPU = 5, + + PERF_COUNT_HW_CACHE_MAX, /* non-ABI */ +}; + +enum perf_hw_cache_op_id { + PERF_COUNT_HW_CACHE_OP_READ = 0, + PERF_COUNT_HW_CACHE_OP_WRITE = 1, + PERF_COUNT_HW_CACHE_OP_PREFETCH = 2, + + PERF_COUNT_HW_CACHE_OP_MAX, /* non-ABI */ +}; + +enum perf_hw_cache_op_result_id { + PERF_COUNT_HW_CACHE_RESULT_ACCESS = 0, + PERF_COUNT_HW_CACHE_RESULT_MISS = 1, + + PERF_COUNT_HW_CACHE_RESULT_MAX, /* non-ABI */ +}; + +/* + * Special "software" counters provided by the kernel, even if the hardware + * does not support performance counters. These counters measure various + * physical and sw events of the kernel (and allow the profiling of them as + * well): + */ +enum perf_sw_ids { + PERF_COUNT_SW_CPU_CLOCK = 0, + PERF_COUNT_SW_TASK_CLOCK = 1, + PERF_COUNT_SW_PAGE_FAULTS = 2, + PERF_COUNT_SW_CONTEXT_SWITCHES = 3, + PERF_COUNT_SW_CPU_MIGRATIONS = 4, + PERF_COUNT_SW_PAGE_FAULTS_MIN = 5, + PERF_COUNT_SW_PAGE_FAULTS_MAJ = 6, + + PERF_COUNT_SW_MAX, /* non-ABI */ +}; + +/* + * Bits that can be set in attr.sample_type to request information + * in the overflow packets. + */ +enum perf_counter_sample_format { + PERF_SAMPLE_IP = 1U << 0, + PERF_SAMPLE_TID = 1U << 1, + PERF_SAMPLE_TIME = 1U << 2, + PERF_SAMPLE_ADDR = 1U << 3, + PERF_SAMPLE_GROUP = 1U << 4, + PERF_SAMPLE_CALLCHAIN = 1U << 5, + PERF_SAMPLE_ID = 1U << 6, + PERF_SAMPLE_CPU = 1U << 7, + PERF_SAMPLE_PERIOD = 1U << 8, +}; + +/* + * Bits that can be set in attr.read_format to request that + * reads on the counter should return the indicated quantities, + * in increasing order of bit value, after the counter value. + */ +enum perf_counter_read_format { + PERF_FORMAT_TOTAL_TIME_ENABLED = 1U << 0, + PERF_FORMAT_TOTAL_TIME_RUNNING = 1U << 1, + PERF_FORMAT_ID = 1U << 2, +}; + +/* + * Hardware event to monitor via a performance monitoring counter: + */ +struct perf_counter_attr { + /* + * Major type: hardware/software/tracepoint/etc. + */ + __u32 type; + __u32 __reserved_1; + + /* + * Type specific configuration information. + */ + __u64 config; + + union { + __u64 sample_period; + __u64 sample_freq; + }; + + __u64 sample_type; + __u64 read_format; + + __u64 disabled : 1, /* off by default */ + inherit : 1, /* children inherit it */ + pinned : 1, /* must always be on PMU */ + exclusive : 1, /* only group on PMU */ + exclude_user : 1, /* don't count user */ + exclude_kernel : 1, /* ditto kernel */ + exclude_hv : 1, /* ditto hypervisor */ + exclude_idle : 1, /* don't count when idle */ + mmap : 1, /* include mmap data */ + comm : 1, /* include comm data */ + freq : 1, /* use freq, not period */ + + __reserved_2 : 53; + + __u32 wakeup_events; /* wakeup every n events */ + __u32 __reserved_3; + + __u64 __reserved_4; +}; + +/* + * Ioctls that can be done on a perf counter fd: + */ +#define PERF_COUNTER_IOC_ENABLE _IO ('$', 0) +#define PERF_COUNTER_IOC_DISABLE _IO ('$', 1) +#define PERF_COUNTER_IOC_REFRESH _IO ('$', 2) +#define PERF_COUNTER_IOC_RESET _IO ('$', 3) +#define PERF_COUNTER_IOC_PERIOD _IOW('$', 4, u64) + +enum perf_counter_ioc_flags { + PERF_IOC_FLAG_GROUP = 1U << 0, +}; + +/* + * Structure of the page that can be mapped via mmap + */ +struct perf_counter_mmap_page { + __u32 version; /* version number of this structure */ + __u32 compat_version; /* lowest version this is compat with */ + + /* + * Bits needed to read the hw counters in user-space. + * + * u32 seq; + * s64 count; + * + * do { + * seq = pc->lock; + * + * barrier() + * if (pc->index) { + * count = pmc_read(pc->index - 1); + * count += pc->offset; + * } else + * goto regular_read; + * + * barrier(); + * } while (pc->lock != seq); + * + * NOTE: for obvious reason this only works on self-monitoring + * processes. + */ + __u32 lock; /* seqlock for synchronization */ + __u32 index; /* hardware counter identifier */ + __s64 offset; /* add to hardware counter value */ + + /* + * Control data for the mmap() data buffer. + * + * User-space reading this value should issue an rmb(), on SMP capable + * platforms, after reading this value -- see perf_counter_wakeup(). + */ + __u64 data_head; /* head in the data section */ +}; + +#define PERF_EVENT_MISC_CPUMODE_MASK (3 << 0) +#define PERF_EVENT_MISC_CPUMODE_UNKNOWN (0 << 0) +#define PERF_EVENT_MISC_KERNEL (1 << 0) +#define PERF_EVENT_MISC_USER (2 << 0) +#define PERF_EVENT_MISC_HYPERVISOR (3 << 0) +#define PERF_EVENT_MISC_OVERFLOW (1 << 2) + +struct perf_event_header { + __u32 type; + __u16 misc; + __u16 size; +}; + +enum perf_event_type { + + /* + * The MMAP events record the PROT_EXEC mappings so that we can + * correlate userspace IPs to code. They have the following structure: + * + * struct { + * struct perf_event_header header; + * + * u32 pid, tid; + * u64 addr; + * u64 len; + * u64 pgoff; + * char filename[]; + * }; + */ + PERF_EVENT_MMAP = 1, + + /* + * struct { + * struct perf_event_header header; + * + * u32 pid, tid; + * char comm[]; + * }; + */ + PERF_EVENT_COMM = 3, + + /* + * struct { + * struct perf_event_header header; + * u64 time; + * u64 id; + * u64 sample_period; + * }; + */ + PERF_EVENT_PERIOD = 4, + + /* + * struct { + * struct perf_event_header header; + * u64 time; + * u64 id; + * }; + */ + PERF_EVENT_THROTTLE = 5, + PERF_EVENT_UNTHROTTLE = 6, + + /* + * struct { + * struct perf_event_header header; + * u32 pid, ppid; + * }; + */ + PERF_EVENT_FORK = 7, + + /* + * When header.misc & PERF_EVENT_MISC_OVERFLOW the event_type field + * will be PERF_RECORD_* + * + * struct { + * struct perf_event_header header; + * + * { u64 ip; } && PERF_RECORD_IP + * { u32 pid, tid; } && PERF_RECORD_TID + * { u64 time; } && PERF_RECORD_TIME + * { u64 addr; } && PERF_RECORD_ADDR + * { u64 config; } && PERF_RECORD_CONFIG + * { u32 cpu, res; } && PERF_RECORD_CPU + * + * { u64 nr; + * { u64 id, val; } cnt[nr]; } && PERF_RECORD_GROUP + * + * { u16 nr, + * hv, + * kernel, + * user; + * u64 ips[nr]; } && PERF_RECORD_CALLCHAIN + * }; + */ +}; + +#ifdef __KERNEL__ +/* + * Kernel-internal data types and definitions: + */ + +#ifdef CONFIG_PERF_COUNTERS +# include <asm/perf_counter.h> +#endif + +#include <linux/list.h> +#include <linux/mutex.h> +#include <linux/rculist.h> +#include <linux/rcupdate.h> +#include <linux/spinlock.h> +#include <linux/hrtimer.h> +#include <linux/fs.h> +#include <linux/pid_namespace.h> +#include <asm/atomic.h> + +struct task_struct; + +/** + * struct hw_perf_counter - performance counter hardware details: + */ +struct hw_perf_counter { +#ifdef CONFIG_PERF_COUNTERS + union { + struct { /* hardware */ + u64 config; + unsigned long config_base; + unsigned long counter_base; + int idx; + }; + union { /* software */ + atomic64_t count; + struct hrtimer hrtimer; + }; + }; + atomic64_t prev_count; + u64 sample_period; + u64 last_period; + atomic64_t period_left; + u64 interrupts; + + u64 freq_count; + u64 freq_interrupts; + u64 freq_stamp; +#endif +}; + +struct perf_counter; + +/** + * struct pmu - generic performance monitoring unit + */ +struct pmu { + int (*enable) (struct perf_counter *counter); + void (*disable) (struct perf_counter *counter); + void (*read) (struct perf_counter *counter); + void (*unthrottle) (struct perf_counter *counter); +}; + +/** + * enum perf_counter_active_state - the states of a counter + */ +enum perf_counter_active_state { + PERF_COUNTER_STATE_ERROR = -2, + PERF_COUNTER_STATE_OFF = -1, + PERF_COUNTER_STATE_INACTIVE = 0, + PERF_COUNTER_STATE_ACTIVE = 1, +}; + +struct file; + +struct perf_mmap_data { + struct rcu_head rcu_head; + int nr_pages; /* nr of data pages */ + int nr_locked; /* nr pages mlocked */ + + atomic_t poll; /* POLL_ for wakeups */ + atomic_t events; /* event limit */ + + atomic_long_t head; /* write position */ + atomic_long_t done_head; /* completed head */ + + atomic_t lock; /* concurrent writes */ + + atomic_t wakeup; /* needs a wakeup */ + + struct perf_counter_mmap_page *user_page; + void *data_pages[0]; +}; + +struct perf_pending_entry { + struct perf_pending_entry *next; + void (*func)(struct perf_pending_entry *); +}; + +/** + * struct perf_counter - performance counter kernel representation: + */ +struct perf_counter { +#ifdef CONFIG_PERF_COUNTERS + struct list_head list_entry; + struct list_head event_entry; + struct list_head sibling_list; + int nr_siblings; + struct perf_counter *group_leader; + const struct pmu *pmu; + + enum perf_counter_active_state state; + atomic64_t count; + + /* + * These are the total time in nanoseconds that the counter + * has been enabled (i.e. eligible to run, and the task has + * been scheduled in, if this is a per-task counter) + * and running (scheduled onto the CPU), respectively. + * + * They are computed from tstamp_enabled, tstamp_running and + * tstamp_stopped when the counter is in INACTIVE or ACTIVE state. + */ + u64 total_time_enabled; + u64 total_time_running; + + /* + * These are timestamps used for computing total_time_enabled + * and total_time_running when the counter is in INACTIVE or + * ACTIVE state, measured in nanoseconds from an arbitrary point + * in time. + * tstamp_enabled: the notional time when the counter was enabled + * tstamp_running: the notional time when the counter was scheduled on + * tstamp_stopped: in INACTIVE state, the notional time when the + * counter was scheduled off. + */ + u64 tstamp_enabled; + u64 tstamp_running; + u64 tstamp_stopped; + + struct perf_counter_attr attr; + struct hw_perf_counter hw; + + struct perf_counter_context *ctx; + struct file *filp; + + /* + * These accumulate total time (in nanoseconds) that children + * counters have been enabled and running, respectively. + */ + atomic64_t child_total_time_enabled; + atomic64_t child_total_time_running; + + /* + * Protect attach/detach and child_list: + */ + struct mutex child_mutex; + struct list_head child_list; + struct perf_counter *parent; + + int oncpu; + int cpu; + + struct list_head owner_entry; + struct task_struct *owner; + + /* mmap bits */ + struct mutex mmap_mutex; + atomic_t mmap_count; + struct perf_mmap_data *data; + + /* poll related */ + wait_queue_head_t waitq; + struct fasync_struct *fasync; + + /* delayed work for NMIs and such */ + int pending_wakeup; + int pending_kill; + int pending_disable; + struct perf_pending_entry pending; + + atomic_t event_limit; + + void (*destroy)(struct perf_counter *); + struct rcu_head rcu_head; + + struct pid_namespace *ns; + u64 id; +#endif +}; + +/** + * struct perf_counter_context - counter context structure + * + * Used as a container for task counters and CPU counters as well: + */ +struct perf_counter_context { + /* + * Protect the states of the counters in the list, + * nr_active, and the list: + */ + spinlock_t lock; + /* + * Protect the list of counters. Locking either mutex or lock + * is sufficient to ensure the list doesn't change; to change + * the list you need to lock both the mutex and the spinlock. + */ + struct mutex mutex; + + struct list_head counter_list; + struct list_head event_list; + int nr_counters; + int nr_active; + int is_active; + atomic_t refcount; + struct task_struct *task; + + /* + * Context clock, runs when context enabled. + */ + u64 time; + u64 timestamp; + + /* + * These fields let us detect when two contexts have both + * been cloned (inherited) from a common ancestor. + */ + struct perf_counter_context *parent_ctx; + u64 parent_gen; + u64 generation; + int pin_count; + struct rcu_head rcu_head; +}; + +/** + * struct perf_counter_cpu_context - per cpu counter context structure + */ +struct perf_cpu_context { + struct perf_counter_context ctx; + struct perf_counter_context *task_ctx; + int active_oncpu; + int max_pertask; + int exclusive; + + /* + * Recursion avoidance: + * + * task, softirq, irq, nmi context + */ + int recursion[4]; +}; + +#ifdef CONFIG_PERF_COUNTERS + +/* + * Set by architecture code: + */ +extern int perf_max_counters; + +extern const struct pmu *hw_perf_counter_init(struct perf_counter *counter); + +extern void perf_counter_task_sched_in(struct task_struct *task, int cpu); +extern void perf_counter_task_sched_out(struct task_struct *task, + struct task_struct *next, int cpu); +extern void perf_counter_task_tick(struct task_struct *task, int cpu); +extern int perf_counter_init_task(struct task_struct *child); +extern void perf_counter_exit_task(struct task_struct *child); +extern void perf_counter_free_task(struct task_struct *task); +extern void perf_counter_do_pending(void); +extern void perf_counter_print_debug(void); +extern void __perf_disable(void); +extern bool __perf_enable(void); +extern void perf_disable(void); +extern void perf_enable(void); +extern int perf_counter_task_disable(void); +extern int perf_counter_task_enable(void); +extern int hw_perf_group_sched_in(struct perf_counter *group_leader, + struct perf_cpu_context *cpuctx, + struct perf_counter_context *ctx, int cpu); +extern void perf_counter_update_userpage(struct perf_counter *counter); + +struct perf_sample_data { + struct pt_regs *regs; + u64 addr; + u64 period; +}; + +extern int perf_counter_overflow(struct perf_counter *counter, int nmi, + struct perf_sample_data *data); + +/* + * Return 1 for a software counter, 0 for a hardware counter + */ +static inline int is_software_counter(struct perf_counter *counter) +{ + return (counter->attr.type != PERF_TYPE_RAW) && + (counter->attr.type != PERF_TYPE_HARDWARE); +} + +extern void perf_swcounter_event(u32, u64, int, struct pt_regs *, u64); + +extern void __perf_counter_mmap(struct vm_area_struct *vma); + +static inline void perf_counter_mmap(struct vm_area_struct *vma) +{ + if (vma->vm_flags & VM_EXEC) + __perf_counter_mmap(vma); +} + +extern void perf_counter_comm(struct task_struct *tsk); +extern void perf_counter_fork(struct task_struct *tsk); + +extern void perf_counter_task_migration(struct task_struct *task, int cpu); + +#define MAX_STACK_DEPTH 255 + +struct perf_callchain_entry { + u16 nr; + u16 hv; + u16 kernel; + u16 user; + u64 ip[MAX_STACK_DEPTH]; +}; + +extern struct perf_callchain_entry *perf_callchain(struct pt_regs *regs); + +extern int sysctl_perf_counter_paranoid; +extern int sysctl_perf_counter_mlock; +extern int sysctl_perf_counter_sample_rate; + +extern void perf_counter_init(void); + +#ifndef perf_misc_flags +#define perf_misc_flags(regs) (user_mode(regs) ? PERF_EVENT_MISC_USER : \ + PERF_EVENT_MISC_KERNEL) +#define perf_instruction_pointer(regs) instruction_pointer(regs) +#endif + +#else +static inline void +perf_counter_task_sched_in(struct task_struct *task, int cpu) { } +static inline void +perf_counter_task_sched_out(struct task_struct *task, + struct task_struct *next, int cpu) { } +static inline void +perf_counter_task_tick(struct task_struct *task, int cpu) { } +static inline int perf_counter_init_task(struct task_struct *child) { return 0; } +static inline void perf_counter_exit_task(struct task_struct *child) { } +static inline void perf_counter_free_task(struct task_struct *task) { } +static inline void perf_counter_do_pending(void) { } +static inline void perf_counter_print_debug(void) { } +static inline void perf_disable(void) { } +static inline void perf_enable(void) { } +static inline int perf_counter_task_disable(void) { return -EINVAL; } +static inline int perf_counter_task_enable(void) { return -EINVAL; } + +static inline void +perf_swcounter_event(u32 event, u64 nr, int nmi, + struct pt_regs *regs, u64 addr) { } + +static inline void perf_counter_mmap(struct vm_area_struct *vma) { } +static inline void perf_counter_comm(struct task_struct *tsk) { } +static inline void perf_counter_fork(struct task_struct *tsk) { } +static inline void perf_counter_init(void) { } +static inline void perf_counter_task_migration(struct task_struct *task, + int cpu) { } +#endif + +#endif /* __KERNEL__ */ +#endif /* _LINUX_PERF_COUNTER_H */ diff --git a/include/linux/prctl.h b/include/linux/prctl.h index 48d887e3c6e..b00df4c79c6 100644 --- a/include/linux/prctl.h +++ b/include/linux/prctl.h @@ -85,4 +85,7 @@ #define PR_SET_TIMERSLACK 29 #define PR_GET_TIMERSLACK 30 +#define PR_TASK_PERF_COUNTERS_DISABLE 31 +#define PR_TASK_PERF_COUNTERS_ENABLE 32 + #endif /* _LINUX_PRCTL_H */ diff --git a/include/linux/ptrace.h b/include/linux/ptrace.h index 67c15653fc2..59e133d39d5 100644 --- a/include/linux/ptrace.h +++ b/include/linux/ptrace.h @@ -95,7 +95,6 @@ extern void __ptrace_link(struct task_struct *child, struct task_struct *new_parent); extern void __ptrace_unlink(struct task_struct *child); extern void exit_ptrace(struct task_struct *tracer); -extern void ptrace_fork(struct task_struct *task, unsigned long clone_flags); #define PTRACE_MODE_READ 1 #define PTRACE_MODE_ATTACH 2 /* Returns 0 on success, -errno on denial. */ @@ -327,15 +326,6 @@ static inline void user_enable_block_step(struct task_struct *task) #define arch_ptrace_untrace(task) do { } while (0) #endif -#ifndef arch_ptrace_fork -/* - * Do machine-specific work to initialize a new task. - * - * This is called from copy_process(). - */ -#define arch_ptrace_fork(child, clone_flags) do { } while (0) -#endif - extern int task_current_syscall(struct task_struct *target, long *callno, unsigned long args[6], unsigned int maxargs, unsigned long *sp, unsigned long *pc); diff --git a/include/linux/rculist.h b/include/linux/rculist.h index e649bd3f2c9..5710f43bbc9 100644 --- a/include/linux/rculist.h +++ b/include/linux/rculist.h @@ -198,6 +198,32 @@ static inline void list_splice_init_rcu(struct list_head *list, at->prev = last; } +/** + * list_entry_rcu - get the struct for this entry + * @ptr: the &struct list_head pointer. + * @type: the type of the struct this is embedded in. + * @member: the name of the list_struct within the struct. + * + * This primitive may safely run concurrently with the _rcu list-mutation + * primitives such as list_add_rcu() as long as it's guarded by rcu_read_lock(). + */ +#define list_entry_rcu(ptr, type, member) \ + container_of(rcu_dereference(ptr), type, member) + +/** + * list_first_entry_rcu - get the first element from a list + * @ptr: the list head to take the element from. + * @type: the type of the struct this is embedded in. + * @member: the name of the list_struct within the struct. + * + * Note, that list is expected to be not empty. + * + * This primitive may safely run concurrently with the _rcu list-mutation + * primitives such as list_add_rcu() as long as it's guarded by rcu_read_lock(). + */ +#define list_first_entry_rcu(ptr, type, member) \ + list_entry_rcu((ptr)->next, type, member) + #define __list_for_each_rcu(pos, head) \ for (pos = rcu_dereference((head)->next); \ pos != (head); \ @@ -214,9 +240,9 @@ static inline void list_splice_init_rcu(struct list_head *list, * as long as the traversal is guarded by rcu_read_lock(). */ #define list_for_each_entry_rcu(pos, head, member) \ - for (pos = list_entry(rcu_dereference((head)->next), typeof(*pos), member); \ + for (pos = list_entry_rcu((head)->next, typeof(*pos), member); \ prefetch(pos->member.next), &pos->member != (head); \ - pos = list_entry(rcu_dereference(pos->member.next), typeof(*pos), member)) + pos = list_entry_rcu(pos->member.next, typeof(*pos), member)) /** diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index 58b2aa5312b..5a5153806c4 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -161,8 +161,15 @@ struct rcu_data { unsigned long offline_fqs; /* Kicked due to being offline. */ unsigned long resched_ipi; /* Sent a resched IPI. */ - /* 5) For future __rcu_pending statistics. */ + /* 5) __rcu_pending() statistics. */ long n_rcu_pending; /* rcu_pending() calls since boot. */ + long n_rp_qs_pending; + long n_rp_cb_ready; + long n_rp_cpu_needs_gp; + long n_rp_gp_completed; + long n_rp_gp_started; + long n_rp_need_fqs; + long n_rp_need_nothing; int cpu; }; diff --git a/include/linux/ring_buffer.h b/include/linux/ring_buffer.h index e1b7b217388..8670f1575fe 100644 --- a/include/linux/ring_buffer.h +++ b/include/linux/ring_buffer.h @@ -11,7 +11,7 @@ struct ring_buffer_iter; * Don't refer to this struct directly, use functions below. */ struct ring_buffer_event { - u32 type:2, len:3, time_delta:27; + u32 type_len:5, time_delta:27; u32 array[]; }; @@ -24,7 +24,8 @@ struct ring_buffer_event { * size is variable depending on how much * padding is needed * If time_delta is non zero: - * everything else same as RINGBUF_TYPE_DATA + * array[0] holds the actual length + * size = 4 + length (bytes) * * @RINGBUF_TYPE_TIME_EXTEND: Extend the time delta * array[0] = time delta (28 .. 59) @@ -35,22 +36,23 @@ struct ring_buffer_event { * array[1..2] = tv_sec * size = 16 bytes * - * @RINGBUF_TYPE_DATA: Data record - * If len is zero: + * <= @RINGBUF_TYPE_DATA_TYPE_LEN_MAX: + * Data record + * If type_len is zero: * array[0] holds the actual length * array[1..(length+3)/4] holds data - * size = 4 + 4 + length (bytes) + * size = 4 + length (bytes) * else - * length = len << 2 + * length = type_len << 2 * array[0..(length+3)/4-1] holds data * size = 4 + length (bytes) */ enum ring_buffer_type { + RINGBUF_TYPE_DATA_TYPE_LEN_MAX = 28, RINGBUF_TYPE_PADDING, RINGBUF_TYPE_TIME_EXTEND, /* FIXME: RINGBUF_TYPE_TIME_STAMP not implemented */ RINGBUF_TYPE_TIME_STAMP, - RINGBUF_TYPE_DATA, }; unsigned ring_buffer_event_length(struct ring_buffer_event *event); @@ -68,13 +70,54 @@ ring_buffer_event_time_delta(struct ring_buffer_event *event) return event->time_delta; } +/* + * ring_buffer_event_discard can discard any event in the ring buffer. + * it is up to the caller to protect against a reader from + * consuming it or a writer from wrapping and replacing it. + * + * No external protection is needed if this is called before + * the event is commited. But in that case it would be better to + * use ring_buffer_discard_commit. + * + * Note, if an event that has not been committed is discarded + * with ring_buffer_event_discard, it must still be committed. + */ void ring_buffer_event_discard(struct ring_buffer_event *event); /* + * ring_buffer_discard_commit will remove an event that has not + * ben committed yet. If this is used, then ring_buffer_unlock_commit + * must not be called on the discarded event. This function + * will try to remove the event from the ring buffer completely + * if another event has not been written after it. + * + * Example use: + * + * if (some_condition) + * ring_buffer_discard_commit(buffer, event); + * else + * ring_buffer_unlock_commit(buffer, event); + */ +void ring_buffer_discard_commit(struct ring_buffer *buffer, + struct ring_buffer_event *event); + +/* * size is in bytes for each per CPU buffer. */ struct ring_buffer * -ring_buffer_alloc(unsigned long size, unsigned flags); +__ring_buffer_alloc(unsigned long size, unsigned flags, struct lock_class_key *key); + +/* + * Because the ring buffer is generic, if other users of the ring buffer get + * traced by ftrace, it can produce lockdep warnings. We need to keep each + * ring buffer's lock class separate. + */ +#define ring_buffer_alloc(size, flags) \ +({ \ + static struct lock_class_key __key; \ + __ring_buffer_alloc((size), (flags), &__key); \ +}) + void ring_buffer_free(struct ring_buffer *buffer); int ring_buffer_resize(struct ring_buffer *buffer, unsigned long size); @@ -122,6 +165,8 @@ unsigned long ring_buffer_entries(struct ring_buffer *buffer); unsigned long ring_buffer_overruns(struct ring_buffer *buffer); unsigned long ring_buffer_entries_cpu(struct ring_buffer *buffer, int cpu); unsigned long ring_buffer_overrun_cpu(struct ring_buffer *buffer, int cpu); +unsigned long ring_buffer_commit_overrun_cpu(struct ring_buffer *buffer, int cpu); +unsigned long ring_buffer_nmi_dropped_cpu(struct ring_buffer *buffer, int cpu); u64 ring_buffer_time_stamp(struct ring_buffer *buffer, int cpu); void ring_buffer_normalize_time_stamp(struct ring_buffer *buffer, @@ -137,6 +182,11 @@ void ring_buffer_free_read_page(struct ring_buffer *buffer, void *data); int ring_buffer_read_page(struct ring_buffer *buffer, void **data_page, size_t len, int cpu, int full); +struct trace_seq; + +int ring_buffer_print_entry_header(struct trace_seq *s); +int ring_buffer_print_page_header(struct trace_seq *s); + enum ring_buffer_flags { RB_FL_OVERWRITE = 1 << 0, }; diff --git a/include/linux/sched.h b/include/linux/sched.h index b4c38bc8049..28c774ff3cc 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -77,6 +77,7 @@ struct sched_param { #include <linux/proportions.h> #include <linux/seccomp.h> #include <linux/rcupdate.h> +#include <linux/rculist.h> #include <linux/rtmutex.h> #include <linux/time.h> @@ -96,8 +97,9 @@ struct exec_domain; struct futex_pi_state; struct robust_list_head; struct bio; -struct bts_tracer; struct fs_struct; +struct bts_context; +struct perf_counter_context; /* * List of flags we want to share for kernel threads, @@ -116,6 +118,7 @@ struct fs_struct; * 11 bit fractions. */ extern unsigned long avenrun[]; /* Load averages */ +extern void get_avenrun(unsigned long *loads, unsigned long offset, int shift); #define FSHIFT 11 /* nr of bits of precision */ #define FIXED_1 (1<<FSHIFT) /* 1.0 as fixed-point */ @@ -135,8 +138,9 @@ DECLARE_PER_CPU(unsigned long, process_counts); extern int nr_processes(void); extern unsigned long nr_running(void); extern unsigned long nr_uninterruptible(void); -extern unsigned long nr_active(void); extern unsigned long nr_iowait(void); +extern void calc_global_load(void); +extern u64 cpu_nr_migrations(int cpu); extern unsigned long get_parent_ip(unsigned long addr); @@ -672,6 +676,10 @@ struct user_struct { struct work_struct work; #endif #endif + +#ifdef CONFIG_PERF_COUNTERS + atomic_long_t locked_vm; +#endif }; extern int uids_sysfs_init(void); @@ -838,7 +846,17 @@ struct sched_group { */ u32 reciprocal_cpu_power; - unsigned long cpumask[]; + /* + * The CPUs this group covers. + * + * NOTE: this field is variable length. (Allocated dynamically + * by attaching extra space to the end of the structure, + * depending on how many CPUs the kernel has booted up with) + * + * It is also be embedded into static data structures at build + * time. (See 'struct static_sched_group' in kernel/sched.c) + */ + unsigned long cpumask[0]; }; static inline struct cpumask *sched_group_cpus(struct sched_group *sg) @@ -924,8 +942,17 @@ struct sched_domain { char *name; #endif - /* span of all CPUs in this domain */ - unsigned long span[]; + /* + * Span of all CPUs in this domain. + * + * NOTE: this field is variable length. (Allocated dynamically + * by attaching extra space to the end of the structure, + * depending on how many CPUs the kernel has booted up with) + * + * It is also be embedded into static data structures at build + * time. (See 'struct static_sched_domain' in kernel/sched.c) + */ + unsigned long span[0]; }; static inline struct cpumask *sched_domain_span(struct sched_domain *sd) @@ -1052,9 +1079,10 @@ struct sched_entity { u64 last_wakeup; u64 avg_overlap; + u64 nr_migrations; + u64 start_runtime; u64 avg_wakeup; - u64 nr_migrations; #ifdef CONFIG_SCHEDSTATS u64 wait_start; @@ -1209,18 +1237,11 @@ struct task_struct { struct list_head ptraced; struct list_head ptrace_entry; -#ifdef CONFIG_X86_PTRACE_BTS /* * This is the tracer handle for the ptrace BTS extension. * This field actually belongs to the ptracer task. */ - struct bts_tracer *bts; - /* - * The buffer to hold the BTS data. - */ - void *bts_buffer; - size_t bts_size; -#endif /* CONFIG_X86_PTRACE_BTS */ + struct bts_context *bts; /* PID/PID hash table linkage. */ struct pid_link pids[PIDTYPE_MAX]; @@ -1380,6 +1401,11 @@ struct task_struct { struct list_head pi_state_list; struct futex_pi_state *pi_state_cache; #endif +#ifdef CONFIG_PERF_COUNTERS + struct perf_counter_context *perf_counter_ctxp; + struct mutex perf_counter_mutex; + struct list_head perf_counter_list; +#endif #ifdef CONFIG_NUMA struct mempolicy *mempolicy; short il_next; @@ -1428,7 +1454,9 @@ struct task_struct { #ifdef CONFIG_TRACING /* state flags for use by tracers */ unsigned long trace; -#endif + /* bitmask of trace recursion */ + unsigned long trace_recursion; +#endif /* CONFIG_TRACING */ }; /* Future-safe accessor for struct task_struct's cpus_allowed. */ @@ -2001,8 +2029,10 @@ extern void set_task_comm(struct task_struct *tsk, char *from); extern char *get_task_comm(char *to, struct task_struct *tsk); #ifdef CONFIG_SMP +extern void wait_task_context_switch(struct task_struct *p); extern unsigned long wait_task_inactive(struct task_struct *, long match_state); #else +static inline void wait_task_context_switch(struct task_struct *p) {} static inline unsigned long wait_task_inactive(struct task_struct *p, long match_state) { @@ -2010,7 +2040,8 @@ static inline unsigned long wait_task_inactive(struct task_struct *p, } #endif -#define next_task(p) list_entry(rcu_dereference((p)->tasks.next), struct task_struct, tasks) +#define next_task(p) \ + list_entry_rcu((p)->tasks.next, struct task_struct, tasks) #define for_each_process(p) \ for (p = &init_task ; (p = next_task(p)) != &init_task ; ) @@ -2049,8 +2080,8 @@ int same_thread_group(struct task_struct *p1, struct task_struct *p2) static inline struct task_struct *next_thread(const struct task_struct *p) { - return list_entry(rcu_dereference(p->thread_group.next), - struct task_struct, thread_group); + return list_entry_rcu(p->thread_group.next, + struct task_struct, thread_group); } static inline int thread_group_empty(struct task_struct *p) @@ -2388,6 +2419,13 @@ static inline void inc_syscw(struct task_struct *tsk) #define TASK_SIZE_OF(tsk) TASK_SIZE #endif +/* + * Call the function if the target task is executing on a CPU right now: + */ +extern void task_oncpu_function_call(struct task_struct *p, + void (*func) (void *info), void *info); + + #ifdef CONFIG_MM_OWNER extern void mm_update_next_owner(struct mm_struct *mm); extern void mm_init_owner(struct mm_struct *mm, struct task_struct *p); diff --git a/include/linux/signal.h b/include/linux/signal.h index 84f997f8aa5..c7552836bd9 100644 --- a/include/linux/signal.h +++ b/include/linux/signal.h @@ -235,6 +235,8 @@ static inline int valid_signal(unsigned long sig) extern int next_signal(struct sigpending *pending, sigset_t *mask); extern int group_send_sig_info(int sig, struct siginfo *info, struct task_struct *p); extern int __group_send_sig_info(int, struct siginfo *, struct task_struct *); +extern long do_rt_tgsigqueueinfo(pid_t tgid, pid_t pid, int sig, + siginfo_t *info); extern long do_sigpending(void __user *, unsigned long); extern int sigprocmask(int, sigset_t *, sigset_t *); extern int show_unhandled_signals; diff --git a/include/linux/slab_def.h b/include/linux/slab_def.h index 5ac9b0bcaf9..713f841ecaa 100644 --- a/include/linux/slab_def.h +++ b/include/linux/slab_def.h @@ -14,7 +14,7 @@ #include <asm/page.h> /* kmalloc_sizes.h needs PAGE_SIZE */ #include <asm/cache.h> /* kmalloc_sizes.h needs L1_CACHE_BYTES */ #include <linux/compiler.h> -#include <trace/kmemtrace.h> +#include <linux/kmemtrace.h> /* Size description struct for general caches. */ struct cache_sizes { diff --git a/include/linux/slub_def.h b/include/linux/slub_def.h index 5046f90c117..be5d40c43bd 100644 --- a/include/linux/slub_def.h +++ b/include/linux/slub_def.h @@ -10,7 +10,7 @@ #include <linux/gfp.h> #include <linux/workqueue.h> #include <linux/kobject.h> -#include <trace/kmemtrace.h> +#include <linux/kmemtrace.h> enum stat_item { ALLOC_FASTPATH, /* Allocation from cpu slab */ diff --git a/include/linux/spinlock_up.h b/include/linux/spinlock_up.h index 938234c4a99..d4841ed8215 100644 --- a/include/linux/spinlock_up.h +++ b/include/linux/spinlock_up.h @@ -60,6 +60,7 @@ static inline void __raw_spin_unlock(raw_spinlock_t *lock) #define __raw_spin_is_locked(lock) ((void)(lock), 0) /* for sched.c and kernel_lock.c: */ # define __raw_spin_lock(lock) do { (void)(lock); } while (0) +# define __raw_spin_lock_flags(lock, flags) do { (void)(lock); } while (0) # define __raw_spin_unlock(lock) do { (void)(lock); } while (0) # define __raw_spin_trylock(lock) ({ (void)(lock); 1; }) #endif /* DEBUG_SPINLOCK */ diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h index ac9ff54f7cb..cb1a6631b8f 100644 --- a/include/linux/swiotlb.h +++ b/include/linux/swiotlb.h @@ -29,7 +29,8 @@ extern void *swiotlb_alloc(unsigned order, unsigned long nslabs); extern dma_addr_t swiotlb_phys_to_bus(struct device *hwdev, phys_addr_t address); -extern phys_addr_t swiotlb_bus_to_phys(dma_addr_t address); +extern phys_addr_t swiotlb_bus_to_phys(struct device *hwdev, + dma_addr_t address); extern int swiotlb_arch_range_needs_mapping(phys_addr_t paddr, size_t size); diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 30520844b8d..c6c84ad8bd7 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -55,6 +55,7 @@ struct compat_timeval; struct robust_list_head; struct getcpu_cache; struct old_linux_dirent; +struct perf_counter_attr; #include <linux/types.h> #include <linux/aio_abi.h> @@ -755,4 +756,8 @@ asmlinkage long sys_pipe(int __user *); int kernel_execve(const char *filename, char *const argv[], char *const envp[]); + +asmlinkage long sys_perf_counter_open( + const struct perf_counter_attr __user *attr_uptr, + pid_t pid, int cpu, int group_fd, unsigned long flags); #endif diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h index e6b820f8b56..a8cc4e13434 100644 --- a/include/linux/thread_info.h +++ b/include/linux/thread_info.h @@ -21,13 +21,14 @@ struct restart_block { struct { unsigned long arg0, arg1, arg2, arg3; }; - /* For futex_wait */ + /* For futex_wait and futex_wait_requeue_pi */ struct { u32 *uaddr; u32 val; u32 flags; u32 bitset; u64 time; + u32 *uaddr2; } futex; /* For nanosleep */ struct { diff --git a/include/linux/trace_seq.h b/include/linux/trace_seq.h new file mode 100644 index 00000000000..c68bccba207 --- /dev/null +++ b/include/linux/trace_seq.h @@ -0,0 +1,92 @@ +#ifndef _LINUX_TRACE_SEQ_H +#define _LINUX_TRACE_SEQ_H + +#include <linux/fs.h> + +/* + * Trace sequences are used to allow a function to call several other functions + * to create a string of data to use (up to a max of PAGE_SIZE. + */ + +struct trace_seq { + unsigned char buffer[PAGE_SIZE]; + unsigned int len; + unsigned int readpos; +}; + +static inline void +trace_seq_init(struct trace_seq *s) +{ + s->len = 0; + s->readpos = 0; +} + +/* + * Currently only defined when tracing is enabled. + */ +#ifdef CONFIG_TRACING +extern int trace_seq_printf(struct trace_seq *s, const char *fmt, ...) + __attribute__ ((format (printf, 2, 3))); +extern int trace_seq_vprintf(struct trace_seq *s, const char *fmt, va_list args) + __attribute__ ((format (printf, 2, 0))); +extern int +trace_seq_bprintf(struct trace_seq *s, const char *fmt, const u32 *binary); +extern void trace_print_seq(struct seq_file *m, struct trace_seq *s); +extern ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, + size_t cnt); +extern int trace_seq_puts(struct trace_seq *s, const char *str); +extern int trace_seq_putc(struct trace_seq *s, unsigned char c); +extern int trace_seq_putmem(struct trace_seq *s, const void *mem, size_t len); +extern int trace_seq_putmem_hex(struct trace_seq *s, const void *mem, + size_t len); +extern void *trace_seq_reserve(struct trace_seq *s, size_t len); +extern int trace_seq_path(struct trace_seq *s, struct path *path); + +#else /* CONFIG_TRACING */ +static inline int trace_seq_printf(struct trace_seq *s, const char *fmt, ...) +{ + return 0; +} +static inline int +trace_seq_bprintf(struct trace_seq *s, const char *fmt, const u32 *binary) +{ + return 0; +} + +static inline void trace_print_seq(struct seq_file *m, struct trace_seq *s) +{ +} +static inline ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, + size_t cnt) +{ + return 0; +} +static inline int trace_seq_puts(struct trace_seq *s, const char *str) +{ + return 0; +} +static inline int trace_seq_putc(struct trace_seq *s, unsigned char c) +{ + return 0; +} +static inline int +trace_seq_putmem(struct trace_seq *s, const void *mem, size_t len) +{ + return 0; +} +static inline int trace_seq_putmem_hex(struct trace_seq *s, const void *mem, + size_t len) +{ + return 0; +} +static inline void *trace_seq_reserve(struct trace_seq *s, size_t len) +{ + return NULL; +} +static inline int trace_seq_path(struct trace_seq *s, struct path *path) +{ + return 0; +} +#endif /* CONFIG_TRACING */ + +#endif /* _LINUX_TRACE_SEQ_H */ diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h index d35a7ee7611..14df7e635d4 100644 --- a/include/linux/tracepoint.h +++ b/include/linux/tracepoint.h @@ -31,6 +31,8 @@ struct tracepoint { * Keep in sync with vmlinux.lds.h. */ +#ifndef DECLARE_TRACE + #define TP_PROTO(args...) args #define TP_ARGS(args...) args @@ -114,6 +116,7 @@ static inline void tracepoint_update_probe_range(struct tracepoint *begin, struct tracepoint *end) { } #endif /* CONFIG_TRACEPOINTS */ +#endif /* DECLARE_TRACE */ /* * Connect a probe to a tracepoint. @@ -154,10 +157,8 @@ static inline void tracepoint_synchronize_unregister(void) } #define PARAMS(args...) args -#define TRACE_FORMAT(name, proto, args, fmt) \ - DECLARE_TRACE(name, PARAMS(proto), PARAMS(args)) - +#ifndef TRACE_EVENT /* * For use with the TRACE_EVENT macro: * @@ -262,5 +263,6 @@ static inline void tracepoint_synchronize_unregister(void) #define TRACE_EVENT(name, proto, args, struct, assign, print) \ DECLARE_TRACE(name, PARAMS(proto), PARAMS(args)) +#endif #endif diff --git a/include/linux/wait.h b/include/linux/wait.h index bc024632f36..6788e1a4d4c 100644 --- a/include/linux/wait.h +++ b/include/linux/wait.h @@ -132,8 +132,6 @@ static inline void __remove_wait_queue(wait_queue_head_t *head, list_del(&old->task_list); } -void __wake_up_common(wait_queue_head_t *q, unsigned int mode, - int nr_exclusive, int sync, void *key); void __wake_up(wait_queue_head_t *q, unsigned int mode, int nr, void *key); void __wake_up_locked_key(wait_queue_head_t *q, unsigned int mode, void *key); void __wake_up_sync_key(wait_queue_head_t *q, unsigned int mode, int nr, diff --git a/include/trace/block.h b/include/trace/block.h deleted file mode 100644 index 25b7068b819..00000000000 --- a/include/trace/block.h +++ /dev/null @@ -1,76 +0,0 @@ -#ifndef _TRACE_BLOCK_H -#define _TRACE_BLOCK_H - -#include <linux/blkdev.h> -#include <linux/tracepoint.h> - -DECLARE_TRACE(block_rq_abort, - TP_PROTO(struct request_queue *q, struct request *rq), - TP_ARGS(q, rq)); - -DECLARE_TRACE(block_rq_insert, - TP_PROTO(struct request_queue *q, struct request *rq), - TP_ARGS(q, rq)); - -DECLARE_TRACE(block_rq_issue, - TP_PROTO(struct request_queue *q, struct request *rq), - TP_ARGS(q, rq)); - -DECLARE_TRACE(block_rq_requeue, - TP_PROTO(struct request_queue *q, struct request *rq), - TP_ARGS(q, rq)); - -DECLARE_TRACE(block_rq_complete, - TP_PROTO(struct request_queue *q, struct request *rq), - TP_ARGS(q, rq)); - -DECLARE_TRACE(block_bio_bounce, - TP_PROTO(struct request_queue *q, struct bio *bio), - TP_ARGS(q, bio)); - -DECLARE_TRACE(block_bio_complete, - TP_PROTO(struct request_queue *q, struct bio *bio), - TP_ARGS(q, bio)); - -DECLARE_TRACE(block_bio_backmerge, - TP_PROTO(struct request_queue *q, struct bio *bio), - TP_ARGS(q, bio)); - -DECLARE_TRACE(block_bio_frontmerge, - TP_PROTO(struct request_queue *q, struct bio *bio), - TP_ARGS(q, bio)); - -DECLARE_TRACE(block_bio_queue, - TP_PROTO(struct request_queue *q, struct bio *bio), - TP_ARGS(q, bio)); - -DECLARE_TRACE(block_getrq, - TP_PROTO(struct request_queue *q, struct bio *bio, int rw), - TP_ARGS(q, bio, rw)); - -DECLARE_TRACE(block_sleeprq, - TP_PROTO(struct request_queue *q, struct bio *bio, int rw), - TP_ARGS(q, bio, rw)); - -DECLARE_TRACE(block_plug, - TP_PROTO(struct request_queue *q), - TP_ARGS(q)); - -DECLARE_TRACE(block_unplug_timer, - TP_PROTO(struct request_queue *q), - TP_ARGS(q)); - -DECLARE_TRACE(block_unplug_io, - TP_PROTO(struct request_queue *q), - TP_ARGS(q)); - -DECLARE_TRACE(block_split, - TP_PROTO(struct request_queue *q, struct bio *bio, unsigned int pdu), - TP_ARGS(q, bio, pdu)); - -DECLARE_TRACE(block_remap, - TP_PROTO(struct request_queue *q, struct bio *bio, dev_t dev, - sector_t from, sector_t to), - TP_ARGS(q, bio, dev, from, to)); - -#endif diff --git a/include/trace/define_trace.h b/include/trace/define_trace.h new file mode 100644 index 00000000000..f7a7ae1e8f9 --- /dev/null +++ b/include/trace/define_trace.h @@ -0,0 +1,75 @@ +/* + * Trace files that want to automate creationg of all tracepoints defined + * in their file should include this file. The following are macros that the + * trace file may define: + * + * TRACE_SYSTEM defines the system the tracepoint is for + * + * TRACE_INCLUDE_FILE if the file name is something other than TRACE_SYSTEM.h + * This macro may be defined to tell define_trace.h what file to include. + * Note, leave off the ".h". + * + * TRACE_INCLUDE_PATH if the path is something other than core kernel include/trace + * then this macro can define the path to use. Note, the path is relative to + * define_trace.h, not the file including it. Full path names for out of tree + * modules must be used. + */ + +#ifdef CREATE_TRACE_POINTS + +/* Prevent recursion */ +#undef CREATE_TRACE_POINTS + +#include <linux/stringify.h> + +#undef TRACE_EVENT +#define TRACE_EVENT(name, proto, args, tstruct, assign, print) \ + DEFINE_TRACE(name) + +#undef DECLARE_TRACE +#define DECLARE_TRACE(name, proto, args) \ + DEFINE_TRACE(name) + +#undef TRACE_INCLUDE +#undef __TRACE_INCLUDE + +#ifndef TRACE_INCLUDE_FILE +# define TRACE_INCLUDE_FILE TRACE_SYSTEM +# define UNDEF_TRACE_INCLUDE_FILE +#endif + +#ifndef TRACE_INCLUDE_PATH +# define __TRACE_INCLUDE(system) <trace/events/system.h> +# define UNDEF_TRACE_INCLUDE_PATH +#else +# define __TRACE_INCLUDE(system) __stringify(TRACE_INCLUDE_PATH/system.h) +#endif + +# define TRACE_INCLUDE(system) __TRACE_INCLUDE(system) + +/* Let the trace headers be reread */ +#define TRACE_HEADER_MULTI_READ + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + +#ifdef CONFIG_EVENT_TRACING +#include <trace/ftrace.h> +#endif + +#undef TRACE_HEADER_MULTI_READ + +/* Only undef what we defined in this file */ +#ifdef UNDEF_TRACE_INCLUDE_FILE +# undef TRACE_INCLUDE_FILE +# undef UNDEF_TRACE_INCLUDE_FILE +#endif + +#ifdef UNDEF_TRACE_INCLUDE_PATH +# undef TRACE_INCLUDE_PATH +# undef UNDEF_TRACE_INCLUDE_PATH +#endif + +/* We may be processing more files */ +#define CREATE_TRACE_POINTS + +#endif /* CREATE_TRACE_POINTS */ diff --git a/include/trace/events/block.h b/include/trace/events/block.h new file mode 100644 index 00000000000..53effd496a5 --- /dev/null +++ b/include/trace/events/block.h @@ -0,0 +1,498 @@ +#if !defined(_TRACE_BLOCK_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_BLOCK_H + +#include <linux/blktrace_api.h> +#include <linux/blkdev.h> +#include <linux/tracepoint.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM block + +TRACE_EVENT(block_rq_abort, + + TP_PROTO(struct request_queue *q, struct request *rq), + + TP_ARGS(q, rq), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __field( int, errors ) + __array( char, rwbs, 6 ) + __dynamic_array( char, cmd, blk_cmd_buf_len(rq) ) + ), + + TP_fast_assign( + __entry->dev = rq->rq_disk ? disk_devt(rq->rq_disk) : 0; + __entry->sector = blk_pc_request(rq) ? 0 : rq->hard_sector; + __entry->nr_sector = blk_pc_request(rq) ? + 0 : rq->hard_nr_sectors; + __entry->errors = rq->errors; + + blk_fill_rwbs_rq(__entry->rwbs, rq); + blk_dump_cmd(__get_str(cmd), rq); + ), + + TP_printk("%d,%d %s (%s) %llu + %u [%d]", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->rwbs, __get_str(cmd), + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->errors) +); + +TRACE_EVENT(block_rq_insert, + + TP_PROTO(struct request_queue *q, struct request *rq), + + TP_ARGS(q, rq), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __field( unsigned int, bytes ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + __dynamic_array( char, cmd, blk_cmd_buf_len(rq) ) + ), + + TP_fast_assign( + __entry->dev = rq->rq_disk ? disk_devt(rq->rq_disk) : 0; + __entry->sector = blk_pc_request(rq) ? 0 : rq->hard_sector; + __entry->nr_sector = blk_pc_request(rq) ? + 0 : rq->hard_nr_sectors; + __entry->bytes = blk_pc_request(rq) ? rq->data_len : 0; + + blk_fill_rwbs_rq(__entry->rwbs, rq); + blk_dump_cmd(__get_str(cmd), rq); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %u (%s) %llu + %u [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->rwbs, __entry->bytes, __get_str(cmd), + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->comm) +); + +TRACE_EVENT(block_rq_issue, + + TP_PROTO(struct request_queue *q, struct request *rq), + + TP_ARGS(q, rq), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __field( unsigned int, bytes ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + __dynamic_array( char, cmd, blk_cmd_buf_len(rq) ) + ), + + TP_fast_assign( + __entry->dev = rq->rq_disk ? disk_devt(rq->rq_disk) : 0; + __entry->sector = blk_pc_request(rq) ? 0 : rq->hard_sector; + __entry->nr_sector = blk_pc_request(rq) ? + 0 : rq->hard_nr_sectors; + __entry->bytes = blk_pc_request(rq) ? rq->data_len : 0; + + blk_fill_rwbs_rq(__entry->rwbs, rq); + blk_dump_cmd(__get_str(cmd), rq); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %u (%s) %llu + %u [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->rwbs, __entry->bytes, __get_str(cmd), + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->comm) +); + +TRACE_EVENT(block_rq_requeue, + + TP_PROTO(struct request_queue *q, struct request *rq), + + TP_ARGS(q, rq), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __field( int, errors ) + __array( char, rwbs, 6 ) + __dynamic_array( char, cmd, blk_cmd_buf_len(rq) ) + ), + + TP_fast_assign( + __entry->dev = rq->rq_disk ? disk_devt(rq->rq_disk) : 0; + __entry->sector = blk_pc_request(rq) ? 0 : rq->hard_sector; + __entry->nr_sector = blk_pc_request(rq) ? + 0 : rq->hard_nr_sectors; + __entry->errors = rq->errors; + + blk_fill_rwbs_rq(__entry->rwbs, rq); + blk_dump_cmd(__get_str(cmd), rq); + ), + + TP_printk("%d,%d %s (%s) %llu + %u [%d]", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->rwbs, __get_str(cmd), + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->errors) +); + +TRACE_EVENT(block_rq_complete, + + TP_PROTO(struct request_queue *q, struct request *rq), + + TP_ARGS(q, rq), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __field( int, errors ) + __array( char, rwbs, 6 ) + __dynamic_array( char, cmd, blk_cmd_buf_len(rq) ) + ), + + TP_fast_assign( + __entry->dev = rq->rq_disk ? disk_devt(rq->rq_disk) : 0; + __entry->sector = blk_pc_request(rq) ? 0 : rq->hard_sector; + __entry->nr_sector = blk_pc_request(rq) ? + 0 : rq->hard_nr_sectors; + __entry->errors = rq->errors; + + blk_fill_rwbs_rq(__entry->rwbs, rq); + blk_dump_cmd(__get_str(cmd), rq); + ), + + TP_printk("%d,%d %s (%s) %llu + %u [%d]", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->rwbs, __get_str(cmd), + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->errors) +); +TRACE_EVENT(block_bio_bounce, + + TP_PROTO(struct request_queue *q, struct bio *bio), + + TP_ARGS(q, bio), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->dev = bio->bi_bdev->bd_dev; + __entry->sector = bio->bi_sector; + __entry->nr_sector = bio->bi_size >> 9; + blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %llu + %u [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->comm) +); + +TRACE_EVENT(block_bio_complete, + + TP_PROTO(struct request_queue *q, struct bio *bio), + + TP_ARGS(q, bio), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned, nr_sector ) + __field( int, error ) + __array( char, rwbs, 6 ) + ), + + TP_fast_assign( + __entry->dev = bio->bi_bdev->bd_dev; + __entry->sector = bio->bi_sector; + __entry->nr_sector = bio->bi_size >> 9; + blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size); + ), + + TP_printk("%d,%d %s %llu + %u [%d]", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->error) +); + +TRACE_EVENT(block_bio_backmerge, + + TP_PROTO(struct request_queue *q, struct bio *bio), + + TP_ARGS(q, bio), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->dev = bio->bi_bdev->bd_dev; + __entry->sector = bio->bi_sector; + __entry->nr_sector = bio->bi_size >> 9; + blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %llu + %u [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->comm) +); + +TRACE_EVENT(block_bio_frontmerge, + + TP_PROTO(struct request_queue *q, struct bio *bio), + + TP_ARGS(q, bio), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned, nr_sector ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->dev = bio->bi_bdev->bd_dev; + __entry->sector = bio->bi_sector; + __entry->nr_sector = bio->bi_size >> 9; + blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %llu + %u [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->comm) +); + +TRACE_EVENT(block_bio_queue, + + TP_PROTO(struct request_queue *q, struct bio *bio), + + TP_ARGS(q, bio), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->dev = bio->bi_bdev->bd_dev; + __entry->sector = bio->bi_sector; + __entry->nr_sector = bio->bi_size >> 9; + blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %llu + %u [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->comm) +); + +TRACE_EVENT(block_getrq, + + TP_PROTO(struct request_queue *q, struct bio *bio, int rw), + + TP_ARGS(q, bio, rw), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->dev = bio ? bio->bi_bdev->bd_dev : 0; + __entry->sector = bio ? bio->bi_sector : 0; + __entry->nr_sector = bio ? bio->bi_size >> 9 : 0; + blk_fill_rwbs(__entry->rwbs, + bio ? bio->bi_rw : 0, __entry->nr_sector); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %llu + %u [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->comm) +); + +TRACE_EVENT(block_sleeprq, + + TP_PROTO(struct request_queue *q, struct bio *bio, int rw), + + TP_ARGS(q, bio, rw), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->dev = bio ? bio->bi_bdev->bd_dev : 0; + __entry->sector = bio ? bio->bi_sector : 0; + __entry->nr_sector = bio ? bio->bi_size >> 9 : 0; + blk_fill_rwbs(__entry->rwbs, + bio ? bio->bi_rw : 0, __entry->nr_sector); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %llu + %u [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + __entry->nr_sector, __entry->comm) +); + +TRACE_EVENT(block_plug, + + TP_PROTO(struct request_queue *q), + + TP_ARGS(q), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("[%s]", __entry->comm) +); + +TRACE_EVENT(block_unplug_timer, + + TP_PROTO(struct request_queue *q), + + TP_ARGS(q), + + TP_STRUCT__entry( + __field( int, nr_rq ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->nr_rq = q->rq.count[READ] + q->rq.count[WRITE]; + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("[%s] %d", __entry->comm, __entry->nr_rq) +); + +TRACE_EVENT(block_unplug_io, + + TP_PROTO(struct request_queue *q), + + TP_ARGS(q), + + TP_STRUCT__entry( + __field( int, nr_rq ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->nr_rq = q->rq.count[READ] + q->rq.count[WRITE]; + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("[%s] %d", __entry->comm, __entry->nr_rq) +); + +TRACE_EVENT(block_split, + + TP_PROTO(struct request_queue *q, struct bio *bio, + unsigned int new_sector), + + TP_ARGS(q, bio, new_sector), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( sector_t, new_sector ) + __array( char, rwbs, 6 ) + __array( char, comm, TASK_COMM_LEN ) + ), + + TP_fast_assign( + __entry->dev = bio->bi_bdev->bd_dev; + __entry->sector = bio->bi_sector; + __entry->new_sector = new_sector; + blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size); + memcpy(__entry->comm, current->comm, TASK_COMM_LEN); + ), + + TP_printk("%d,%d %s %llu / %llu [%s]", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + (unsigned long long)__entry->new_sector, + __entry->comm) +); + +TRACE_EVENT(block_remap, + + TP_PROTO(struct request_queue *q, struct bio *bio, dev_t dev, + sector_t from), + + TP_ARGS(q, bio, dev, from), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( sector_t, sector ) + __field( unsigned int, nr_sector ) + __field( dev_t, old_dev ) + __field( sector_t, old_sector ) + __array( char, rwbs, 6 ) + ), + + TP_fast_assign( + __entry->dev = bio->bi_bdev->bd_dev; + __entry->sector = bio->bi_sector; + __entry->nr_sector = bio->bi_size >> 9; + __entry->old_dev = dev; + __entry->old_sector = from; + blk_fill_rwbs(__entry->rwbs, bio->bi_rw, bio->bi_size); + ), + + TP_printk("%d,%d %s %llu + %u <- (%d,%d) %llu", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->rwbs, + (unsigned long long)__entry->sector, + __entry->nr_sector, + MAJOR(__entry->old_dev), MINOR(__entry->old_dev), + (unsigned long long)__entry->old_sector) +); + +#endif /* _TRACE_BLOCK_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> + diff --git a/include/trace/events/irq.h b/include/trace/events/irq.h new file mode 100644 index 00000000000..b0c7ede55eb --- /dev/null +++ b/include/trace/events/irq.h @@ -0,0 +1,145 @@ +#if !defined(_TRACE_IRQ_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_IRQ_H + +#include <linux/tracepoint.h> +#include <linux/interrupt.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM irq + +#define softirq_name(sirq) { sirq##_SOFTIRQ, #sirq } +#define show_softirq_name(val) \ + __print_symbolic(val, \ + softirq_name(HI), \ + softirq_name(TIMER), \ + softirq_name(NET_TX), \ + softirq_name(NET_RX), \ + softirq_name(BLOCK), \ + softirq_name(TASKLET), \ + softirq_name(SCHED), \ + softirq_name(HRTIMER), \ + softirq_name(RCU)) + +/** + * irq_handler_entry - called immediately before the irq action handler + * @irq: irq number + * @action: pointer to struct irqaction + * + * The struct irqaction pointed to by @action contains various + * information about the handler, including the device name, + * @action->name, and the device id, @action->dev_id. When used in + * conjunction with the irq_handler_exit tracepoint, we can figure + * out irq handler latencies. + */ +TRACE_EVENT(irq_handler_entry, + + TP_PROTO(int irq, struct irqaction *action), + + TP_ARGS(irq, action), + + TP_STRUCT__entry( + __field( int, irq ) + __string( name, action->name ) + ), + + TP_fast_assign( + __entry->irq = irq; + __assign_str(name, action->name); + ), + + TP_printk("irq=%d handler=%s", __entry->irq, __get_str(name)) +); + +/** + * irq_handler_exit - called immediately after the irq action handler returns + * @irq: irq number + * @action: pointer to struct irqaction + * @ret: return value + * + * If the @ret value is set to IRQ_HANDLED, then we know that the corresponding + * @action->handler scuccessully handled this irq. Otherwise, the irq might be + * a shared irq line, or the irq was not handled successfully. Can be used in + * conjunction with the irq_handler_entry to understand irq handler latencies. + */ +TRACE_EVENT(irq_handler_exit, + + TP_PROTO(int irq, struct irqaction *action, int ret), + + TP_ARGS(irq, action, ret), + + TP_STRUCT__entry( + __field( int, irq ) + __field( int, ret ) + ), + + TP_fast_assign( + __entry->irq = irq; + __entry->ret = ret; + ), + + TP_printk("irq=%d return=%s", + __entry->irq, __entry->ret ? "handled" : "unhandled") +); + +/** + * softirq_entry - called immediately before the softirq handler + * @h: pointer to struct softirq_action + * @vec: pointer to first struct softirq_action in softirq_vec array + * + * The @h parameter, contains a pointer to the struct softirq_action + * which has a pointer to the action handler that is called. By subtracting + * the @vec pointer from the @h pointer, we can determine the softirq + * number. Also, when used in combination with the softirq_exit tracepoint + * we can determine the softirq latency. + */ +TRACE_EVENT(softirq_entry, + + TP_PROTO(struct softirq_action *h, struct softirq_action *vec), + + TP_ARGS(h, vec), + + TP_STRUCT__entry( + __field( int, vec ) + ), + + TP_fast_assign( + __entry->vec = (int)(h - vec); + ), + + TP_printk("softirq=%d action=%s", __entry->vec, + show_softirq_name(__entry->vec)) +); + +/** + * softirq_exit - called immediately after the softirq handler returns + * @h: pointer to struct softirq_action + * @vec: pointer to first struct softirq_action in softirq_vec array + * + * The @h parameter contains a pointer to the struct softirq_action + * that has handled the softirq. By subtracting the @vec pointer from + * the @h pointer, we can determine the softirq number. Also, when used in + * combination with the softirq_entry tracepoint we can determine the softirq + * latency. + */ +TRACE_EVENT(softirq_exit, + + TP_PROTO(struct softirq_action *h, struct softirq_action *vec), + + TP_ARGS(h, vec), + + TP_STRUCT__entry( + __field( int, vec ) + ), + + TP_fast_assign( + __entry->vec = (int)(h - vec); + ), + + TP_printk("softirq=%d action=%s", __entry->vec, + show_softirq_name(__entry->vec)) +); + +#endif /* _TRACE_IRQ_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/include/trace/events/kmem.h b/include/trace/events/kmem.h new file mode 100644 index 00000000000..9baba50d651 --- /dev/null +++ b/include/trace/events/kmem.h @@ -0,0 +1,231 @@ +#if !defined(_TRACE_KMEM_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_KMEM_H + +#include <linux/types.h> +#include <linux/tracepoint.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM kmem + +/* + * The order of these masks is important. Matching masks will be seen + * first and the left over flags will end up showing by themselves. + * + * For example, if we have GFP_KERNEL before GFP_USER we wil get: + * + * GFP_KERNEL|GFP_HARDWALL + * + * Thus most bits set go first. + */ +#define show_gfp_flags(flags) \ + (flags) ? __print_flags(flags, "|", \ + {(unsigned long)GFP_HIGHUSER_MOVABLE, "GFP_HIGHUSER_MOVABLE"}, \ + {(unsigned long)GFP_HIGHUSER, "GFP_HIGHUSER"}, \ + {(unsigned long)GFP_USER, "GFP_USER"}, \ + {(unsigned long)GFP_TEMPORARY, "GFP_TEMPORARY"}, \ + {(unsigned long)GFP_KERNEL, "GFP_KERNEL"}, \ + {(unsigned long)GFP_NOFS, "GFP_NOFS"}, \ + {(unsigned long)GFP_ATOMIC, "GFP_ATOMIC"}, \ + {(unsigned long)GFP_NOIO, "GFP_NOIO"}, \ + {(unsigned long)__GFP_HIGH, "GFP_HIGH"}, \ + {(unsigned long)__GFP_WAIT, "GFP_WAIT"}, \ + {(unsigned long)__GFP_IO, "GFP_IO"}, \ + {(unsigned long)__GFP_COLD, "GFP_COLD"}, \ + {(unsigned long)__GFP_NOWARN, "GFP_NOWARN"}, \ + {(unsigned long)__GFP_REPEAT, "GFP_REPEAT"}, \ + {(unsigned long)__GFP_NOFAIL, "GFP_NOFAIL"}, \ + {(unsigned long)__GFP_NORETRY, "GFP_NORETRY"}, \ + {(unsigned long)__GFP_COMP, "GFP_COMP"}, \ + {(unsigned long)__GFP_ZERO, "GFP_ZERO"}, \ + {(unsigned long)__GFP_NOMEMALLOC, "GFP_NOMEMALLOC"}, \ + {(unsigned long)__GFP_HARDWALL, "GFP_HARDWALL"}, \ + {(unsigned long)__GFP_THISNODE, "GFP_THISNODE"}, \ + {(unsigned long)__GFP_RECLAIMABLE, "GFP_RECLAIMABLE"}, \ + {(unsigned long)__GFP_MOVABLE, "GFP_MOVABLE"} \ + ) : "GFP_NOWAIT" + +TRACE_EVENT(kmalloc, + + TP_PROTO(unsigned long call_site, + const void *ptr, + size_t bytes_req, + size_t bytes_alloc, + gfp_t gfp_flags), + + TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags), + + TP_STRUCT__entry( + __field( unsigned long, call_site ) + __field( const void *, ptr ) + __field( size_t, bytes_req ) + __field( size_t, bytes_alloc ) + __field( gfp_t, gfp_flags ) + ), + + TP_fast_assign( + __entry->call_site = call_site; + __entry->ptr = ptr; + __entry->bytes_req = bytes_req; + __entry->bytes_alloc = bytes_alloc; + __entry->gfp_flags = gfp_flags; + ), + + TP_printk("call_site=%lx ptr=%p bytes_req=%zu bytes_alloc=%zu gfp_flags=%s", + __entry->call_site, + __entry->ptr, + __entry->bytes_req, + __entry->bytes_alloc, + show_gfp_flags(__entry->gfp_flags)) +); + +TRACE_EVENT(kmem_cache_alloc, + + TP_PROTO(unsigned long call_site, + const void *ptr, + size_t bytes_req, + size_t bytes_alloc, + gfp_t gfp_flags), + + TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags), + + TP_STRUCT__entry( + __field( unsigned long, call_site ) + __field( const void *, ptr ) + __field( size_t, bytes_req ) + __field( size_t, bytes_alloc ) + __field( gfp_t, gfp_flags ) + ), + + TP_fast_assign( + __entry->call_site = call_site; + __entry->ptr = ptr; + __entry->bytes_req = bytes_req; + __entry->bytes_alloc = bytes_alloc; + __entry->gfp_flags = gfp_flags; + ), + + TP_printk("call_site=%lx ptr=%p bytes_req=%zu bytes_alloc=%zu gfp_flags=%s", + __entry->call_site, + __entry->ptr, + __entry->bytes_req, + __entry->bytes_alloc, + show_gfp_flags(__entry->gfp_flags)) +); + +TRACE_EVENT(kmalloc_node, + + TP_PROTO(unsigned long call_site, + const void *ptr, + size_t bytes_req, + size_t bytes_alloc, + gfp_t gfp_flags, + int node), + + TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags, node), + + TP_STRUCT__entry( + __field( unsigned long, call_site ) + __field( const void *, ptr ) + __field( size_t, bytes_req ) + __field( size_t, bytes_alloc ) + __field( gfp_t, gfp_flags ) + __field( int, node ) + ), + + TP_fast_assign( + __entry->call_site = call_site; + __entry->ptr = ptr; + __entry->bytes_req = bytes_req; + __entry->bytes_alloc = bytes_alloc; + __entry->gfp_flags = gfp_flags; + __entry->node = node; + ), + + TP_printk("call_site=%lx ptr=%p bytes_req=%zu bytes_alloc=%zu gfp_flags=%s node=%d", + __entry->call_site, + __entry->ptr, + __entry->bytes_req, + __entry->bytes_alloc, + show_gfp_flags(__entry->gfp_flags), + __entry->node) +); + +TRACE_EVENT(kmem_cache_alloc_node, + + TP_PROTO(unsigned long call_site, + const void *ptr, + size_t bytes_req, + size_t bytes_alloc, + gfp_t gfp_flags, + int node), + + TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags, node), + + TP_STRUCT__entry( + __field( unsigned long, call_site ) + __field( const void *, ptr ) + __field( size_t, bytes_req ) + __field( size_t, bytes_alloc ) + __field( gfp_t, gfp_flags ) + __field( int, node ) + ), + + TP_fast_assign( + __entry->call_site = call_site; + __entry->ptr = ptr; + __entry->bytes_req = bytes_req; + __entry->bytes_alloc = bytes_alloc; + __entry->gfp_flags = gfp_flags; + __entry->node = node; + ), + + TP_printk("call_site=%lx ptr=%p bytes_req=%zu bytes_alloc=%zu gfp_flags=%s node=%d", + __entry->call_site, + __entry->ptr, + __entry->bytes_req, + __entry->bytes_alloc, + show_gfp_flags(__entry->gfp_flags), + __entry->node) +); + +TRACE_EVENT(kfree, + + TP_PROTO(unsigned long call_site, const void *ptr), + + TP_ARGS(call_site, ptr), + + TP_STRUCT__entry( + __field( unsigned long, call_site ) + __field( const void *, ptr ) + ), + + TP_fast_assign( + __entry->call_site = call_site; + __entry->ptr = ptr; + ), + + TP_printk("call_site=%lx ptr=%p", __entry->call_site, __entry->ptr) +); + +TRACE_EVENT(kmem_cache_free, + + TP_PROTO(unsigned long call_site, const void *ptr), + + TP_ARGS(call_site, ptr), + + TP_STRUCT__entry( + __field( unsigned long, call_site ) + __field( const void *, ptr ) + ), + + TP_fast_assign( + __entry->call_site = call_site; + __entry->ptr = ptr; + ), + + TP_printk("call_site=%lx ptr=%p", __entry->call_site, __entry->ptr) +); +#endif /* _TRACE_KMEM_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/include/trace/events/lockdep.h b/include/trace/events/lockdep.h new file mode 100644 index 00000000000..0e956c9dfd7 --- /dev/null +++ b/include/trace/events/lockdep.h @@ -0,0 +1,96 @@ +#if !defined(_TRACE_LOCKDEP_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_LOCKDEP_H + +#include <linux/lockdep.h> +#include <linux/tracepoint.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM lockdep + +#ifdef CONFIG_LOCKDEP + +TRACE_EVENT(lock_acquire, + + TP_PROTO(struct lockdep_map *lock, unsigned int subclass, + int trylock, int read, int check, + struct lockdep_map *next_lock, unsigned long ip), + + TP_ARGS(lock, subclass, trylock, read, check, next_lock, ip), + + TP_STRUCT__entry( + __field(unsigned int, flags) + __string(name, lock->name) + ), + + TP_fast_assign( + __entry->flags = (trylock ? 1 : 0) | (read ? 2 : 0); + __assign_str(name, lock->name); + ), + + TP_printk("%s%s%s", (__entry->flags & 1) ? "try " : "", + (__entry->flags & 2) ? "read " : "", + __get_str(name)) +); + +TRACE_EVENT(lock_release, + + TP_PROTO(struct lockdep_map *lock, int nested, unsigned long ip), + + TP_ARGS(lock, nested, ip), + + TP_STRUCT__entry( + __string(name, lock->name) + ), + + TP_fast_assign( + __assign_str(name, lock->name); + ), + + TP_printk("%s", __get_str(name)) +); + +#ifdef CONFIG_LOCK_STAT + +TRACE_EVENT(lock_contended, + + TP_PROTO(struct lockdep_map *lock, unsigned long ip), + + TP_ARGS(lock, ip), + + TP_STRUCT__entry( + __string(name, lock->name) + ), + + TP_fast_assign( + __assign_str(name, lock->name); + ), + + TP_printk("%s", __get_str(name)) +); + +TRACE_EVENT(lock_acquired, + TP_PROTO(struct lockdep_map *lock, unsigned long ip, s64 waittime), + + TP_ARGS(lock, ip, waittime), + + TP_STRUCT__entry( + __string(name, lock->name) + __field(unsigned long, wait_usec) + __field(unsigned long, wait_nsec_rem) + ), + TP_fast_assign( + __assign_str(name, lock->name); + __entry->wait_nsec_rem = do_div(waittime, NSEC_PER_USEC); + __entry->wait_usec = (unsigned long) waittime; + ), + TP_printk("%s (%lu.%03lu us)", __get_str(name), __entry->wait_usec, + __entry->wait_nsec_rem) +); + +#endif +#endif + +#endif /* _TRACE_LOCKDEP_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/include/trace/sched_event_types.h b/include/trace/events/sched.h index 63547dc1125..24ab5bcff7b 100644 --- a/include/trace/sched_event_types.h +++ b/include/trace/events/sched.h @@ -1,9 +1,8 @@ +#if !defined(_TRACE_SCHED_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_SCHED_H -/* use <trace/sched.h> instead */ -#ifndef TRACE_EVENT -# error Do not include this file directly. -# error Unless you know what you are doing. -#endif +#include <linux/sched.h> +#include <linux/tracepoint.h> #undef TRACE_SYSTEM #define TRACE_SYSTEM sched @@ -157,6 +156,7 @@ TRACE_EVENT(sched_switch, __array( char, prev_comm, TASK_COMM_LEN ) __field( pid_t, prev_pid ) __field( int, prev_prio ) + __field( long, prev_state ) __array( char, next_comm, TASK_COMM_LEN ) __field( pid_t, next_pid ) __field( int, next_prio ) @@ -166,13 +166,19 @@ TRACE_EVENT(sched_switch, memcpy(__entry->next_comm, next->comm, TASK_COMM_LEN); __entry->prev_pid = prev->pid; __entry->prev_prio = prev->prio; + __entry->prev_state = prev->state; memcpy(__entry->prev_comm, prev->comm, TASK_COMM_LEN); __entry->next_pid = next->pid; __entry->next_prio = next->prio; ), - TP_printk("task %s:%d [%d] ==> %s:%d [%d]", + TP_printk("task %s:%d [%d] (%s) ==> %s:%d [%d]", __entry->prev_comm, __entry->prev_pid, __entry->prev_prio, + __entry->prev_state ? + __print_flags(__entry->prev_state, "|", + { 1, "S"} , { 2, "D" }, { 4, "T" }, { 8, "t" }, + { 16, "Z" }, { 32, "X" }, { 64, "x" }, + { 128, "W" }) : "R", __entry->next_comm, __entry->next_pid, __entry->next_prio) ); @@ -181,9 +187,9 @@ TRACE_EVENT(sched_switch, */ TRACE_EVENT(sched_migrate_task, - TP_PROTO(struct task_struct *p, int orig_cpu, int dest_cpu), + TP_PROTO(struct task_struct *p, int dest_cpu), - TP_ARGS(p, orig_cpu, dest_cpu), + TP_ARGS(p, dest_cpu), TP_STRUCT__entry( __array( char, comm, TASK_COMM_LEN ) @@ -197,7 +203,7 @@ TRACE_EVENT(sched_migrate_task, memcpy(__entry->comm, p->comm, TASK_COMM_LEN); __entry->pid = p->pid; __entry->prio = p->prio; - __entry->orig_cpu = orig_cpu; + __entry->orig_cpu = task_cpu(p); __entry->dest_cpu = dest_cpu; ), @@ -334,4 +340,7 @@ TRACE_EVENT(sched_signal_send, __entry->sig, __entry->comm, __entry->pid) ); -#undef TRACE_SYSTEM +#endif /* _TRACE_SCHED_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h new file mode 100644 index 00000000000..1e8fabb57c0 --- /dev/null +++ b/include/trace/events/skb.h @@ -0,0 +1,40 @@ +#if !defined(_TRACE_SKB_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_SKB_H + +#include <linux/skbuff.h> +#include <linux/tracepoint.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM skb + +/* + * Tracepoint for free an sk_buff: + */ +TRACE_EVENT(kfree_skb, + + TP_PROTO(struct sk_buff *skb, void *location), + + TP_ARGS(skb, location), + + TP_STRUCT__entry( + __field( void *, skbaddr ) + __field( unsigned short, protocol ) + __field( void *, location ) + ), + + TP_fast_assign( + __entry->skbaddr = skb; + if (skb) { + __entry->protocol = ntohs(skb->protocol); + } + __entry->location = location; + ), + + TP_printk("skbaddr=%p protocol=%u location=%p", + __entry->skbaddr, __entry->protocol, __entry->location) +); + +#endif /* _TRACE_SKB_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/include/trace/events/workqueue.h b/include/trace/events/workqueue.h new file mode 100644 index 00000000000..035f1bff288 --- /dev/null +++ b/include/trace/events/workqueue.h @@ -0,0 +1,100 @@ +#if !defined(_TRACE_WORKQUEUE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_WORKQUEUE_H + +#include <linux/workqueue.h> +#include <linux/sched.h> +#include <linux/tracepoint.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM workqueue + +TRACE_EVENT(workqueue_insertion, + + TP_PROTO(struct task_struct *wq_thread, struct work_struct *work), + + TP_ARGS(wq_thread, work), + + TP_STRUCT__entry( + __array(char, thread_comm, TASK_COMM_LEN) + __field(pid_t, thread_pid) + __field(work_func_t, func) + ), + + TP_fast_assign( + memcpy(__entry->thread_comm, wq_thread->comm, TASK_COMM_LEN); + __entry->thread_pid = wq_thread->pid; + __entry->func = work->func; + ), + + TP_printk("thread=%s:%d func=%pF", __entry->thread_comm, + __entry->thread_pid, __entry->func) +); + +TRACE_EVENT(workqueue_execution, + + TP_PROTO(struct task_struct *wq_thread, struct work_struct *work), + + TP_ARGS(wq_thread, work), + + TP_STRUCT__entry( + __array(char, thread_comm, TASK_COMM_LEN) + __field(pid_t, thread_pid) + __field(work_func_t, func) + ), + + TP_fast_assign( + memcpy(__entry->thread_comm, wq_thread->comm, TASK_COMM_LEN); + __entry->thread_pid = wq_thread->pid; + __entry->func = work->func; + ), + + TP_printk("thread=%s:%d func=%pF", __entry->thread_comm, + __entry->thread_pid, __entry->func) +); + +/* Trace the creation of one workqueue thread on a cpu */ +TRACE_EVENT(workqueue_creation, + + TP_PROTO(struct task_struct *wq_thread, int cpu), + + TP_ARGS(wq_thread, cpu), + + TP_STRUCT__entry( + __array(char, thread_comm, TASK_COMM_LEN) + __field(pid_t, thread_pid) + __field(int, cpu) + ), + + TP_fast_assign( + memcpy(__entry->thread_comm, wq_thread->comm, TASK_COMM_LEN); + __entry->thread_pid = wq_thread->pid; + __entry->cpu = cpu; + ), + + TP_printk("thread=%s:%d cpu=%d", __entry->thread_comm, + __entry->thread_pid, __entry->cpu) +); + +TRACE_EVENT(workqueue_destruction, + + TP_PROTO(struct task_struct *wq_thread), + + TP_ARGS(wq_thread), + + TP_STRUCT__entry( + __array(char, thread_comm, TASK_COMM_LEN) + __field(pid_t, thread_pid) + ), + + TP_fast_assign( + memcpy(__entry->thread_comm, wq_thread->comm, TASK_COMM_LEN); + __entry->thread_pid = wq_thread->pid; + ), + + TP_printk("thread=%s:%d", __entry->thread_comm, __entry->thread_pid) +); + +#endif /* _TRACE_WORKQUEUE_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h new file mode 100644 index 00000000000..1867553c61e --- /dev/null +++ b/include/trace/ftrace.h @@ -0,0 +1,591 @@ +/* + * Stage 1 of the trace events. + * + * Override the macros in <trace/trace_events.h> to include the following: + * + * struct ftrace_raw_<call> { + * struct trace_entry ent; + * <type> <item>; + * <type2> <item2>[<len>]; + * [...] + * }; + * + * The <type> <item> is created by the __field(type, item) macro or + * the __array(type2, item2, len) macro. + * We simply do "type item;", and that will create the fields + * in the structure. + */ + +#include <linux/ftrace_event.h> + +#undef __field +#define __field(type, item) type item; + +#undef __array +#define __array(type, item, len) type item[len]; + +#undef __dynamic_array +#define __dynamic_array(type, item, len) unsigned short __data_loc_##item; + +#undef __string +#define __string(item, src) __dynamic_array(char, item, -1) + +#undef TP_STRUCT__entry +#define TP_STRUCT__entry(args...) args + +#undef TRACE_EVENT +#define TRACE_EVENT(name, proto, args, tstruct, assign, print) \ + struct ftrace_raw_##name { \ + struct trace_entry ent; \ + tstruct \ + char __data[0]; \ + }; \ + static struct ftrace_event_call event_##name + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + + +/* + * Stage 2 of the trace events. + * + * Include the following: + * + * struct ftrace_data_offsets_<call> { + * int <item1>; + * int <item2>; + * [...] + * }; + * + * The __dynamic_array() macro will create each int <item>, this is + * to keep the offset of each array from the beginning of the event. + */ + +#undef __field +#define __field(type, item); + +#undef __array +#define __array(type, item, len) + +#undef __dynamic_array +#define __dynamic_array(type, item, len) int item; + +#undef __string +#define __string(item, src) __dynamic_array(char, item, -1) + +#undef TRACE_EVENT +#define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ + struct ftrace_data_offsets_##call { \ + tstruct; \ + }; + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + +/* + * Setup the showing format of trace point. + * + * int + * ftrace_format_##call(struct trace_seq *s) + * { + * struct ftrace_raw_##call field; + * int ret; + * + * ret = trace_seq_printf(s, #type " " #item ";" + * " offset:%u; size:%u;\n", + * offsetof(struct ftrace_raw_##call, item), + * sizeof(field.type)); + * + * } + */ + +#undef TP_STRUCT__entry +#define TP_STRUCT__entry(args...) args + +#undef __field +#define __field(type, item) \ + ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t" \ + "offset:%u;\tsize:%u;\n", \ + (unsigned int)offsetof(typeof(field), item), \ + (unsigned int)sizeof(field.item)); \ + if (!ret) \ + return 0; + +#undef __array +#define __array(type, item, len) \ + ret = trace_seq_printf(s, "\tfield:" #type " " #item "[" #len "];\t" \ + "offset:%u;\tsize:%u;\n", \ + (unsigned int)offsetof(typeof(field), item), \ + (unsigned int)sizeof(field.item)); \ + if (!ret) \ + return 0; + +#undef __dynamic_array +#define __dynamic_array(type, item, len) \ + ret = trace_seq_printf(s, "\tfield:__data_loc " #item ";\t" \ + "offset:%u;\tsize:%u;\n", \ + (unsigned int)offsetof(typeof(field), \ + __data_loc_##item), \ + (unsigned int)sizeof(field.__data_loc_##item)); \ + if (!ret) \ + return 0; + +#undef __string +#define __string(item, src) __dynamic_array(char, item, -1) + +#undef __entry +#define __entry REC + +#undef __print_symbolic +#undef __get_dynamic_array +#undef __get_str + +#undef TP_printk +#define TP_printk(fmt, args...) "%s, %s\n", #fmt, __stringify(args) + +#undef TP_fast_assign +#define TP_fast_assign(args...) args + +#undef TRACE_EVENT +#define TRACE_EVENT(call, proto, args, tstruct, func, print) \ +static int \ +ftrace_format_##call(struct trace_seq *s) \ +{ \ + struct ftrace_raw_##call field __attribute__((unused)); \ + int ret = 0; \ + \ + tstruct; \ + \ + trace_seq_printf(s, "\nprint fmt: " print); \ + \ + return ret; \ +} + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + +/* + * Stage 3 of the trace events. + * + * Override the macros in <trace/trace_events.h> to include the following: + * + * enum print_line_t + * ftrace_raw_output_<call>(struct trace_iterator *iter, int flags) + * { + * struct trace_seq *s = &iter->seq; + * struct ftrace_raw_<call> *field; <-- defined in stage 1 + * struct trace_entry *entry; + * struct trace_seq *p; + * int ret; + * + * entry = iter->ent; + * + * if (entry->type != event_<call>.id) { + * WARN_ON_ONCE(1); + * return TRACE_TYPE_UNHANDLED; + * } + * + * field = (typeof(field))entry; + * + * p = get_cpu_var(ftrace_event_seq); + * trace_seq_init(p); + * ret = trace_seq_printf(s, <TP_printk> "\n"); + * put_cpu(); + * if (!ret) + * return TRACE_TYPE_PARTIAL_LINE; + * + * return TRACE_TYPE_HANDLED; + * } + * + * This is the method used to print the raw event to the trace + * output format. Note, this is not needed if the data is read + * in binary. + */ + +#undef __entry +#define __entry field + +#undef TP_printk +#define TP_printk(fmt, args...) fmt "\n", args + +#undef __get_dynamic_array +#define __get_dynamic_array(field) \ + ((void *)__entry + __entry->__data_loc_##field) + +#undef __get_str +#define __get_str(field) (char *)__get_dynamic_array(field) + +#undef __print_flags +#define __print_flags(flag, delim, flag_array...) \ + ({ \ + static const struct trace_print_flags flags[] = \ + { flag_array, { -1, NULL }}; \ + ftrace_print_flags_seq(p, delim, flag, flags); \ + }) + +#undef __print_symbolic +#define __print_symbolic(value, symbol_array...) \ + ({ \ + static const struct trace_print_flags symbols[] = \ + { symbol_array, { -1, NULL }}; \ + ftrace_print_symbols_seq(p, value, symbols); \ + }) + +#undef TRACE_EVENT +#define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ +enum print_line_t \ +ftrace_raw_output_##call(struct trace_iterator *iter, int flags) \ +{ \ + struct trace_seq *s = &iter->seq; \ + struct ftrace_raw_##call *field; \ + struct trace_entry *entry; \ + struct trace_seq *p; \ + int ret; \ + \ + entry = iter->ent; \ + \ + if (entry->type != event_##call.id) { \ + WARN_ON_ONCE(1); \ + return TRACE_TYPE_UNHANDLED; \ + } \ + \ + field = (typeof(field))entry; \ + \ + p = &get_cpu_var(ftrace_event_seq); \ + trace_seq_init(p); \ + ret = trace_seq_printf(s, #call ": " print); \ + put_cpu(); \ + if (!ret) \ + return TRACE_TYPE_PARTIAL_LINE; \ + \ + return TRACE_TYPE_HANDLED; \ +} + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + +#undef __field +#define __field(type, item) \ + ret = trace_define_field(event_call, #type, #item, \ + offsetof(typeof(field), item), \ + sizeof(field.item), is_signed_type(type)); \ + if (ret) \ + return ret; + +#undef __array +#define __array(type, item, len) \ + BUILD_BUG_ON(len > MAX_FILTER_STR_VAL); \ + ret = trace_define_field(event_call, #type "[" #len "]", #item, \ + offsetof(typeof(field), item), \ + sizeof(field.item), 0); \ + if (ret) \ + return ret; + +#undef __dynamic_array +#define __dynamic_array(type, item, len) \ + ret = trace_define_field(event_call, "__data_loc" "[" #type "]", #item,\ + offsetof(typeof(field), __data_loc_##item), \ + sizeof(field.__data_loc_##item), 0); + +#undef __string +#define __string(item, src) __dynamic_array(char, item, -1) + +#undef TRACE_EVENT +#define TRACE_EVENT(call, proto, args, tstruct, func, print) \ +int \ +ftrace_define_fields_##call(void) \ +{ \ + struct ftrace_raw_##call field; \ + struct ftrace_event_call *event_call = &event_##call; \ + int ret; \ + \ + __common_field(int, type, 1); \ + __common_field(unsigned char, flags, 0); \ + __common_field(unsigned char, preempt_count, 0); \ + __common_field(int, pid, 1); \ + __common_field(int, tgid, 1); \ + \ + tstruct; \ + \ + return ret; \ +} + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + +/* + * remember the offset of each array from the beginning of the event. + */ + +#undef __entry +#define __entry entry + +#undef __field +#define __field(type, item) + +#undef __array +#define __array(type, item, len) + +#undef __dynamic_array +#define __dynamic_array(type, item, len) \ + __data_offsets->item = __data_size + \ + offsetof(typeof(*entry), __data); \ + __data_size += (len) * sizeof(type); + +#undef __string +#define __string(item, src) __dynamic_array(char, item, strlen(src) + 1) \ + +#undef TRACE_EVENT +#define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ +static inline int ftrace_get_offsets_##call( \ + struct ftrace_data_offsets_##call *__data_offsets, proto) \ +{ \ + int __data_size = 0; \ + struct ftrace_raw_##call __maybe_unused *entry; \ + \ + tstruct; \ + \ + return __data_size; \ +} + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + +/* + * Stage 4 of the trace events. + * + * Override the macros in <trace/trace_events.h> to include the following: + * + * static void ftrace_event_<call>(proto) + * { + * event_trace_printk(_RET_IP_, "<call>: " <fmt>); + * } + * + * static int ftrace_reg_event_<call>(void) + * { + * int ret; + * + * ret = register_trace_<call>(ftrace_event_<call>); + * if (!ret) + * pr_info("event trace: Could not activate trace point " + * "probe to <call>"); + * return ret; + * } + * + * static void ftrace_unreg_event_<call>(void) + * { + * unregister_trace_<call>(ftrace_event_<call>); + * } + * + * + * For those macros defined with TRACE_EVENT: + * + * static struct ftrace_event_call event_<call>; + * + * static void ftrace_raw_event_<call>(proto) + * { + * struct ring_buffer_event *event; + * struct ftrace_raw_<call> *entry; <-- defined in stage 1 + * unsigned long irq_flags; + * int pc; + * + * local_save_flags(irq_flags); + * pc = preempt_count(); + * + * event = trace_current_buffer_lock_reserve(event_<call>.id, + * sizeof(struct ftrace_raw_<call>), + * irq_flags, pc); + * if (!event) + * return; + * entry = ring_buffer_event_data(event); + * + * <assign>; <-- Here we assign the entries by the __field and + * __array macros. + * + * trace_current_buffer_unlock_commit(event, irq_flags, pc); + * } + * + * static int ftrace_raw_reg_event_<call>(void) + * { + * int ret; + * + * ret = register_trace_<call>(ftrace_raw_event_<call>); + * if (!ret) + * pr_info("event trace: Could not activate trace point " + * "probe to <call>"); + * return ret; + * } + * + * static void ftrace_unreg_event_<call>(void) + * { + * unregister_trace_<call>(ftrace_raw_event_<call>); + * } + * + * static struct trace_event ftrace_event_type_<call> = { + * .trace = ftrace_raw_output_<call>, <-- stage 2 + * }; + * + * static int ftrace_raw_init_event_<call>(void) + * { + * int id; + * + * id = register_ftrace_event(&ftrace_event_type_<call>); + * if (!id) + * return -ENODEV; + * event_<call>.id = id; + * return 0; + * } + * + * static struct ftrace_event_call __used + * __attribute__((__aligned__(4))) + * __attribute__((section("_ftrace_events"))) event_<call> = { + * .name = "<call>", + * .system = "<system>", + * .raw_init = ftrace_raw_init_event_<call>, + * .regfunc = ftrace_reg_event_<call>, + * .unregfunc = ftrace_unreg_event_<call>, + * .show_format = ftrace_format_<call>, + * } + * + */ + +#undef TP_FMT +#define TP_FMT(fmt, args...) fmt "\n", ##args + +#ifdef CONFIG_EVENT_PROFILE +#define _TRACE_PROFILE(call, proto, args) \ +static void ftrace_profile_##call(proto) \ +{ \ + extern void perf_tpcounter_event(int); \ + perf_tpcounter_event(event_##call.id); \ +} \ + \ +static int ftrace_profile_enable_##call(struct ftrace_event_call *event_call) \ +{ \ + int ret = 0; \ + \ + if (!atomic_inc_return(&event_call->profile_count)) \ + ret = register_trace_##call(ftrace_profile_##call); \ + \ + return ret; \ +} \ + \ +static void ftrace_profile_disable_##call(struct ftrace_event_call *event_call)\ +{ \ + if (atomic_add_negative(-1, &event_call->profile_count)) \ + unregister_trace_##call(ftrace_profile_##call); \ +} + +#define _TRACE_PROFILE_INIT(call) \ + .profile_count = ATOMIC_INIT(-1), \ + .profile_enable = ftrace_profile_enable_##call, \ + .profile_disable = ftrace_profile_disable_##call, + +#else +#define _TRACE_PROFILE(call, proto, args) +#define _TRACE_PROFILE_INIT(call) +#endif + +#undef __entry +#define __entry entry + +#undef __field +#define __field(type, item) + +#undef __array +#define __array(type, item, len) + +#undef __dynamic_array +#define __dynamic_array(type, item, len) \ + __entry->__data_loc_##item = __data_offsets.item; + +#undef __string +#define __string(item, src) __dynamic_array(char, item, -1) \ + +#undef __assign_str +#define __assign_str(dst, src) \ + strcpy(__get_str(dst), src); + +#undef TRACE_EVENT +#define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ +_TRACE_PROFILE(call, PARAMS(proto), PARAMS(args)) \ + \ +static struct ftrace_event_call event_##call; \ + \ +static void ftrace_raw_event_##call(proto) \ +{ \ + struct ftrace_data_offsets_##call __maybe_unused __data_offsets;\ + struct ftrace_event_call *event_call = &event_##call; \ + struct ring_buffer_event *event; \ + struct ftrace_raw_##call *entry; \ + unsigned long irq_flags; \ + int __data_size; \ + int pc; \ + \ + local_save_flags(irq_flags); \ + pc = preempt_count(); \ + \ + __data_size = ftrace_get_offsets_##call(&__data_offsets, args); \ + \ + event = trace_current_buffer_lock_reserve(event_##call.id, \ + sizeof(*entry) + __data_size, \ + irq_flags, pc); \ + if (!event) \ + return; \ + entry = ring_buffer_event_data(event); \ + \ + \ + tstruct \ + \ + { assign; } \ + \ + if (!filter_current_check_discard(event_call, entry, event)) \ + trace_nowake_buffer_unlock_commit(event, irq_flags, pc); \ +} \ + \ +static int ftrace_raw_reg_event_##call(void) \ +{ \ + int ret; \ + \ + ret = register_trace_##call(ftrace_raw_event_##call); \ + if (ret) \ + pr_info("event trace: Could not activate trace point " \ + "probe to " #call "\n"); \ + return ret; \ +} \ + \ +static void ftrace_raw_unreg_event_##call(void) \ +{ \ + unregister_trace_##call(ftrace_raw_event_##call); \ +} \ + \ +static struct trace_event ftrace_event_type_##call = { \ + .trace = ftrace_raw_output_##call, \ +}; \ + \ +static int ftrace_raw_init_event_##call(void) \ +{ \ + int id; \ + \ + id = register_ftrace_event(&ftrace_event_type_##call); \ + if (!id) \ + return -ENODEV; \ + event_##call.id = id; \ + INIT_LIST_HEAD(&event_##call.fields); \ + init_preds(&event_##call); \ + return 0; \ +} \ + \ +static struct ftrace_event_call __used \ +__attribute__((__aligned__(4))) \ +__attribute__((section("_ftrace_events"))) event_##call = { \ + .name = #call, \ + .system = __stringify(TRACE_SYSTEM), \ + .event = &ftrace_event_type_##call, \ + .raw_init = ftrace_raw_init_event_##call, \ + .regfunc = ftrace_raw_reg_event_##call, \ + .unregfunc = ftrace_raw_unreg_event_##call, \ + .show_format = ftrace_format_##call, \ + .define_fields = ftrace_define_fields_##call, \ + _TRACE_PROFILE_INIT(call) \ +} + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) + +#undef _TRACE_PROFILE +#undef _TRACE_PROFILE_INIT + diff --git a/include/trace/irq.h b/include/trace/irq.h deleted file mode 100644 index ff5d4495dc3..00000000000 --- a/include/trace/irq.h +++ /dev/null @@ -1,9 +0,0 @@ -#ifndef _TRACE_IRQ_H -#define _TRACE_IRQ_H - -#include <linux/interrupt.h> -#include <linux/tracepoint.h> - -#include <trace/irq_event_types.h> - -#endif diff --git a/include/trace/irq_event_types.h b/include/trace/irq_event_types.h deleted file mode 100644 index 85964ebd47e..00000000000 --- a/include/trace/irq_event_types.h +++ /dev/null @@ -1,55 +0,0 @@ - -/* use <trace/irq.h> instead */ -#ifndef TRACE_FORMAT -# error Do not include this file directly. -# error Unless you know what you are doing. -#endif - -#undef TRACE_SYSTEM -#define TRACE_SYSTEM irq - -/* - * Tracepoint for entry of interrupt handler: - */ -TRACE_FORMAT(irq_handler_entry, - TP_PROTO(int irq, struct irqaction *action), - TP_ARGS(irq, action), - TP_FMT("irq=%d handler=%s", irq, action->name) - ); - -/* - * Tracepoint for return of an interrupt handler: - */ -TRACE_EVENT(irq_handler_exit, - - TP_PROTO(int irq, struct irqaction *action, int ret), - - TP_ARGS(irq, action, ret), - - TP_STRUCT__entry( - __field( int, irq ) - __field( int, ret ) - ), - - TP_fast_assign( - __entry->irq = irq; - __entry->ret = ret; - ), - - TP_printk("irq=%d return=%s", - __entry->irq, __entry->ret ? "handled" : "unhandled") -); - -TRACE_FORMAT(softirq_entry, - TP_PROTO(struct softirq_action *h, struct softirq_action *vec), - TP_ARGS(h, vec), - TP_FMT("softirq=%d action=%s", (int)(h - vec), softirq_to_name[h-vec]) - ); - -TRACE_FORMAT(softirq_exit, - TP_PROTO(struct softirq_action *h, struct softirq_action *vec), - TP_ARGS(h, vec), - TP_FMT("softirq=%d action=%s", (int)(h - vec), softirq_to_name[h-vec]) - ); - -#undef TRACE_SYSTEM diff --git a/include/trace/kmemtrace.h b/include/trace/kmemtrace.h deleted file mode 100644 index 28ee69f9cd4..00000000000 --- a/include/trace/kmemtrace.h +++ /dev/null @@ -1,63 +0,0 @@ -/* - * Copyright (C) 2008 Eduard - Gabriel Munteanu - * - * This file is released under GPL version 2. - */ - -#ifndef _LINUX_KMEMTRACE_H -#define _LINUX_KMEMTRACE_H - -#ifdef __KERNEL__ - -#include <linux/tracepoint.h> -#include <linux/types.h> - -#ifdef CONFIG_KMEMTRACE -extern void kmemtrace_init(void); -#else -static inline void kmemtrace_init(void) -{ -} -#endif - -DECLARE_TRACE(kmalloc, - TP_PROTO(unsigned long call_site, - const void *ptr, - size_t bytes_req, - size_t bytes_alloc, - gfp_t gfp_flags), - TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags)); -DECLARE_TRACE(kmem_cache_alloc, - TP_PROTO(unsigned long call_site, - const void *ptr, - size_t bytes_req, - size_t bytes_alloc, - gfp_t gfp_flags), - TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags)); -DECLARE_TRACE(kmalloc_node, - TP_PROTO(unsigned long call_site, - const void *ptr, - size_t bytes_req, - size_t bytes_alloc, - gfp_t gfp_flags, - int node), - TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags, node)); -DECLARE_TRACE(kmem_cache_alloc_node, - TP_PROTO(unsigned long call_site, - const void *ptr, - size_t bytes_req, - size_t bytes_alloc, - gfp_t gfp_flags, - int node), - TP_ARGS(call_site, ptr, bytes_req, bytes_alloc, gfp_flags, node)); -DECLARE_TRACE(kfree, - TP_PROTO(unsigned long call_site, const void *ptr), - TP_ARGS(call_site, ptr)); -DECLARE_TRACE(kmem_cache_free, - TP_PROTO(unsigned long call_site, const void *ptr), - TP_ARGS(call_site, ptr)); - -#endif /* __KERNEL__ */ - -#endif /* _LINUX_KMEMTRACE_H */ - diff --git a/include/trace/lockdep.h b/include/trace/lockdep.h deleted file mode 100644 index 5ca67df87f2..00000000000 --- a/include/trace/lockdep.h +++ /dev/null @@ -1,9 +0,0 @@ -#ifndef _TRACE_LOCKDEP_H -#define _TRACE_LOCKDEP_H - -#include <linux/lockdep.h> -#include <linux/tracepoint.h> - -#include <trace/lockdep_event_types.h> - -#endif diff --git a/include/trace/lockdep_event_types.h b/include/trace/lockdep_event_types.h deleted file mode 100644 index adccfcd2ec8..00000000000 --- a/include/trace/lockdep_event_types.h +++ /dev/null @@ -1,44 +0,0 @@ - -#ifndef TRACE_FORMAT -# error Do not include this file directly. -# error Unless you know what you are doing. -#endif - -#undef TRACE_SYSTEM -#define TRACE_SYSTEM lock - -#ifdef CONFIG_LOCKDEP - -TRACE_FORMAT(lock_acquire, - TP_PROTO(struct lockdep_map *lock, unsigned int subclass, - int trylock, int read, int check, - struct lockdep_map *next_lock, unsigned long ip), - TP_ARGS(lock, subclass, trylock, read, check, next_lock, ip), - TP_FMT("%s%s%s", trylock ? "try " : "", - read ? "read " : "", lock->name) - ); - -TRACE_FORMAT(lock_release, - TP_PROTO(struct lockdep_map *lock, int nested, unsigned long ip), - TP_ARGS(lock, nested, ip), - TP_FMT("%s", lock->name) - ); - -#ifdef CONFIG_LOCK_STAT - -TRACE_FORMAT(lock_contended, - TP_PROTO(struct lockdep_map *lock, unsigned long ip), - TP_ARGS(lock, ip), - TP_FMT("%s", lock->name) - ); - -TRACE_FORMAT(lock_acquired, - TP_PROTO(struct lockdep_map *lock, unsigned long ip), - TP_ARGS(lock, ip), - TP_FMT("%s", lock->name) - ); - -#endif -#endif - -#undef TRACE_SYSTEM diff --git a/include/trace/sched.h b/include/trace/sched.h deleted file mode 100644 index 4e372a1a29b..00000000000 --- a/include/trace/sched.h +++ /dev/null @@ -1,9 +0,0 @@ -#ifndef _TRACE_SCHED_H -#define _TRACE_SCHED_H - -#include <linux/sched.h> -#include <linux/tracepoint.h> - -#include <trace/sched_event_types.h> - -#endif diff --git a/include/trace/skb.h b/include/trace/skb.h deleted file mode 100644 index b66206d9be7..00000000000 --- a/include/trace/skb.h +++ /dev/null @@ -1,11 +0,0 @@ -#ifndef _TRACE_SKB_H_ -#define _TRACE_SKB_H_ - -#include <linux/skbuff.h> -#include <linux/tracepoint.h> - -DECLARE_TRACE(kfree_skb, - TP_PROTO(struct sk_buff *skb, void *location), - TP_ARGS(skb, location)); - -#endif diff --git a/include/trace/trace_event_types.h b/include/trace/trace_event_types.h deleted file mode 100644 index df56f5694be..00000000000 --- a/include/trace/trace_event_types.h +++ /dev/null @@ -1,5 +0,0 @@ -/* trace/<type>_event_types.h here */ - -#include <trace/sched_event_types.h> -#include <trace/irq_event_types.h> -#include <trace/lockdep_event_types.h> diff --git a/include/trace/trace_events.h b/include/trace/trace_events.h deleted file mode 100644 index fd13750ca4b..00000000000 --- a/include/trace/trace_events.h +++ /dev/null @@ -1,5 +0,0 @@ -/* trace/<type>.h here */ - -#include <trace/sched.h> -#include <trace/irq.h> -#include <trace/lockdep.h> diff --git a/include/trace/workqueue.h b/include/trace/workqueue.h deleted file mode 100644 index 7626523deeb..00000000000 --- a/include/trace/workqueue.h +++ /dev/null @@ -1,25 +0,0 @@ -#ifndef __TRACE_WORKQUEUE_H -#define __TRACE_WORKQUEUE_H - -#include <linux/tracepoint.h> -#include <linux/workqueue.h> -#include <linux/sched.h> - -DECLARE_TRACE(workqueue_insertion, - TP_PROTO(struct task_struct *wq_thread, struct work_struct *work), - TP_ARGS(wq_thread, work)); - -DECLARE_TRACE(workqueue_execution, - TP_PROTO(struct task_struct *wq_thread, struct work_struct *work), - TP_ARGS(wq_thread, work)); - -/* Trace the creation of one workqueue thread on a cpu */ -DECLARE_TRACE(workqueue_creation, - TP_PROTO(struct task_struct *wq_thread, int cpu), - TP_ARGS(wq_thread, cpu)); - -DECLARE_TRACE(workqueue_destruction, - TP_PROTO(struct task_struct *wq_thread), - TP_ARGS(wq_thread)); - -#endif /* __TRACE_WORKQUEUE_H */ diff --git a/include/xen/Kbuild b/include/xen/Kbuild new file mode 100644 index 00000000000..4e65c16a445 --- /dev/null +++ b/include/xen/Kbuild @@ -0,0 +1 @@ +header-y += evtchn.h diff --git a/include/xen/events.h b/include/xen/events.h index 0d5f1adc036..e68d59a90ca 100644 --- a/include/xen/events.h +++ b/include/xen/events.h @@ -53,4 +53,7 @@ bool xen_test_irq_pending(int irq); irq will be disabled so it won't deliver an interrupt. */ void xen_poll_irq(int irq); +/* Determine the IRQ which is bound to an event channel */ +unsigned irq_from_evtchn(unsigned int evtchn); + #endif /* _XEN_EVENTS_H */ diff --git a/include/xen/evtchn.h b/include/xen/evtchn.h new file mode 100644 index 00000000000..14e833ee4e0 --- /dev/null +++ b/include/xen/evtchn.h @@ -0,0 +1,88 @@ +/****************************************************************************** + * evtchn.h + * + * Interface to /dev/xen/evtchn. + * + * Copyright (c) 2003-2005, K A Fraser + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License version 2 + * as published by the Free Software Foundation; or, when distributed + * separately from the Linux kernel or incorporated into other + * software packages, subject to the following license: + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this source file (the "Software"), to deal in the Software without + * restriction, including without limitation the rights to use, copy, modify, + * merge, publish, distribute, sublicense, and/or sell copies of the Software, + * and to permit persons to whom the Software is furnished to do so, subject to + * the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#ifndef __LINUX_PUBLIC_EVTCHN_H__ +#define __LINUX_PUBLIC_EVTCHN_H__ + +/* + * Bind a fresh port to VIRQ @virq. + * Return allocated port. + */ +#define IOCTL_EVTCHN_BIND_VIRQ \ + _IOC(_IOC_NONE, 'E', 0, sizeof(struct ioctl_evtchn_bind_virq)) +struct ioctl_evtchn_bind_virq { + unsigned int virq; +}; + +/* + * Bind a fresh port to remote <@remote_domain, @remote_port>. + * Return allocated port. + */ +#define IOCTL_EVTCHN_BIND_INTERDOMAIN \ + _IOC(_IOC_NONE, 'E', 1, sizeof(struct ioctl_evtchn_bind_interdomain)) +struct ioctl_evtchn_bind_interdomain { + unsigned int remote_domain, remote_port; +}; + +/* + * Allocate a fresh port for binding to @remote_domain. + * Return allocated port. + */ +#define IOCTL_EVTCHN_BIND_UNBOUND_PORT \ + _IOC(_IOC_NONE, 'E', 2, sizeof(struct ioctl_evtchn_bind_unbound_port)) +struct ioctl_evtchn_bind_unbound_port { + unsigned int remote_domain; +}; + +/* + * Unbind previously allocated @port. + */ +#define IOCTL_EVTCHN_UNBIND \ + _IOC(_IOC_NONE, 'E', 3, sizeof(struct ioctl_evtchn_unbind)) +struct ioctl_evtchn_unbind { + unsigned int port; +}; + +/* + * Unbind previously allocated @port. + */ +#define IOCTL_EVTCHN_NOTIFY \ + _IOC(_IOC_NONE, 'E', 4, sizeof(struct ioctl_evtchn_notify)) +struct ioctl_evtchn_notify { + unsigned int port; +}; + +/* Clear and reinitialise the event buffer. Clear error condition. */ +#define IOCTL_EVTCHN_RESET \ + _IOC(_IOC_NONE, 'E', 5, 0) + +#endif /* __LINUX_PUBLIC_EVTCHN_H__ */ diff --git a/include/xen/interface/version.h b/include/xen/interface/version.h index 453235e923f..e8b6519d47e 100644 --- a/include/xen/interface/version.h +++ b/include/xen/interface/version.h @@ -57,4 +57,7 @@ struct xen_feature_info { /* Declares the features reported by XENVER_get_features. */ #include "features.h" +/* arg == NULL; returns host memory page size. */ +#define XENVER_pagesize 7 + #endif /* __XEN_PUBLIC_VERSION_H__ */ diff --git a/include/xen/xenbus.h b/include/xen/xenbus.h index f87f9614844..b9763badbd7 100644 --- a/include/xen/xenbus.h +++ b/include/xen/xenbus.h @@ -91,8 +91,7 @@ struct xenbus_driver { void (*otherend_changed)(struct xenbus_device *dev, enum xenbus_state backend_state); int (*remove)(struct xenbus_device *dev); - int (*suspend)(struct xenbus_device *dev); - int (*suspend_cancel)(struct xenbus_device *dev); + int (*suspend)(struct xenbus_device *dev, pm_message_t state); int (*resume)(struct xenbus_device *dev); int (*uevent)(struct xenbus_device *, char **, int, char *, int); struct device_driver driver; diff --git a/init/Kconfig b/init/Kconfig index 7be4d383674..9b68fee8d79 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -308,7 +308,7 @@ menu "RCU Subsystem" choice prompt "RCU Implementation" - default CLASSIC_RCU + default TREE_RCU config CLASSIC_RCU bool "Classic RCU" @@ -933,6 +933,40 @@ config AIO by some high performance threaded applications. Disabling this option saves about 7k. +config HAVE_PERF_COUNTERS + bool + +menu "Performance Counters" + +config PERF_COUNTERS + bool "Kernel Performance Counters" + depends on HAVE_PERF_COUNTERS + select ANON_INODES + help + Enable kernel support for performance counter hardware. + + Performance counters are special hardware registers available + on most modern CPUs. These registers count the number of certain + types of hw events: such as instructions executed, cachemisses + suffered, or branches mis-predicted - without slowing down the + kernel or applications. These registers can also trigger interrupts + when a threshold number of events have passed - and can thus be + used to profile the code that runs on that CPU. + + The Linux Performance Counter subsystem provides an abstraction of + these hardware capabilities, available via a system call. It + provides per task and per CPU counters, and it provides event + capabilities on top of those. + + Say Y if unsure. + +config EVENT_PROFILE + bool "Tracepoint profile sources" + depends on PERF_COUNTERS && EVENT_TRACER + default y + +endmenu + config VM_EVENT_COUNTERS default y bool "Enable VM event counters for /proc/vmstat" if EMBEDDED diff --git a/init/main.c b/init/main.c index d721dad05dd..bb7dc57eee3 100644 --- a/init/main.c +++ b/init/main.c @@ -64,6 +64,7 @@ #include <linux/idr.h> #include <linux/ftrace.h> #include <linux/async.h> +#include <linux/kmemtrace.h> #include <trace/boot.h> #include <asm/io.h> @@ -71,7 +72,6 @@ #include <asm/setup.h> #include <asm/sections.h> #include <asm/cacheflush.h> -#include <trace/kmemtrace.h> #ifdef CONFIG_X86_LOCAL_APIC #include <asm/smp.h> diff --git a/ipc/sem.c b/ipc/sem.c index 16a2189e96f..87c2b641fd7 100644 --- a/ipc/sem.c +++ b/ipc/sem.c @@ -1290,8 +1290,8 @@ void exit_sem(struct task_struct *tsk) int i; rcu_read_lock(); - un = list_entry(rcu_dereference(ulp->list_proc.next), - struct sem_undo, list_proc); + un = list_entry_rcu(ulp->list_proc.next, + struct sem_undo, list_proc); if (&un->list_proc == &ulp->list_proc) semid = -1; else diff --git a/kernel/Makefile b/kernel/Makefile index 42423665660..90b53f6dc22 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -93,8 +93,10 @@ obj-$(CONFIG_LATENCYTOP) += latencytop.o obj-$(CONFIG_HAVE_GENERIC_DMA_COHERENT) += dma-coherent.o obj-$(CONFIG_FUNCTION_TRACER) += trace/ obj-$(CONFIG_TRACING) += trace/ +obj-$(CONFIG_X86_DS) += trace/ obj-$(CONFIG_SMP) += sched_cpupri.o obj-$(CONFIG_SLOW_WORK) += slow-work.o +obj-$(CONFIG_PERF_COUNTERS) += perf_counter.o ifneq ($(CONFIG_SCHED_OMIT_FRAME_POINTER),y) # According to Alan Modra <alan@linuxcare.com.au>, the -fno-omit-frame-pointer is diff --git a/kernel/compat.c b/kernel/compat.c index 42d56544460..f6c204f07ea 100644 --- a/kernel/compat.c +++ b/kernel/compat.c @@ -882,6 +882,17 @@ compat_sys_rt_sigtimedwait (compat_sigset_t __user *uthese, } +asmlinkage long +compat_sys_rt_tgsigqueueinfo(compat_pid_t tgid, compat_pid_t pid, int sig, + struct compat_siginfo __user *uinfo) +{ + siginfo_t info; + + if (copy_siginfo_from_user32(&info, uinfo)) + return -EFAULT; + return do_rt_tgsigqueueinfo(tgid, pid, sig, &info); +} + #ifdef __ARCH_WANT_COMPAT_SYS_TIME /* compat_time_t is a 32 bit "long" and needs to get converted. */ diff --git a/kernel/exit.c b/kernel/exit.c index abf9cf3b95c..49cdf6946f3 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -48,7 +48,8 @@ #include <linux/tracehook.h> #include <linux/fs_struct.h> #include <linux/init_task.h> -#include <trace/sched.h> +#include <linux/perf_counter.h> +#include <trace/events/sched.h> #include <asm/uaccess.h> #include <asm/unistd.h> @@ -56,10 +57,6 @@ #include <asm/mmu_context.h> #include "cred-internals.h" -DEFINE_TRACE(sched_process_free); -DEFINE_TRACE(sched_process_exit); -DEFINE_TRACE(sched_process_wait); - static void exit_mm(struct task_struct * tsk); static void __unhash_process(struct task_struct *p) @@ -158,6 +155,9 @@ static void delayed_put_task_struct(struct rcu_head *rhp) { struct task_struct *tsk = container_of(rhp, struct task_struct, rcu); +#ifdef CONFIG_PERF_COUNTERS + WARN_ON_ONCE(tsk->perf_counter_ctxp); +#endif trace_sched_process_free(tsk); put_task_struct(tsk); } @@ -174,6 +174,7 @@ repeat: atomic_dec(&__task_cred(p)->user->processes); proc_flush_task(p); + write_lock_irq(&tasklist_lock); tracehook_finish_release_task(p); __exit_signal(p); @@ -975,16 +976,19 @@ NORET_TYPE void do_exit(long code) module_put(tsk->binfmt->module); proc_exit_connector(tsk); + + /* + * Flush inherited counters to the parent - before the parent + * gets woken up by child-exit notifications. + */ + perf_counter_exit_task(tsk); + exit_notify(tsk, group_dead); #ifdef CONFIG_NUMA mpol_put(tsk->mempolicy); tsk->mempolicy = NULL; #endif #ifdef CONFIG_FUTEX - /* - * This must happen late, after the PID is not - * hashed anymore: - */ if (unlikely(!list_empty(&tsk->pi_state_list))) exit_pi_state_list(tsk); if (unlikely(current->pi_state_cache)) diff --git a/kernel/fork.c b/kernel/fork.c index 875ffbdd96d..4430eb1376f 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -61,8 +61,8 @@ #include <linux/proc_fs.h> #include <linux/blkdev.h> #include <linux/fs_struct.h> -#include <trace/sched.h> #include <linux/magic.h> +#include <linux/perf_counter.h> #include <asm/pgtable.h> #include <asm/pgalloc.h> @@ -71,6 +71,8 @@ #include <asm/cacheflush.h> #include <asm/tlbflush.h> +#include <trace/events/sched.h> + /* * Protected counters by write_lock_irq(&tasklist_lock) */ @@ -83,8 +85,6 @@ DEFINE_PER_CPU(unsigned long, process_counts) = 0; __cacheline_aligned DEFINE_RWLOCK(tasklist_lock); /* outer */ -DEFINE_TRACE(sched_process_fork); - int nr_processes(void) { int cpu; @@ -982,6 +982,8 @@ static struct task_struct *copy_process(unsigned long clone_flags, if (!p) goto fork_out; + ftrace_graph_init_task(p); + rt_mutex_init_task(p); #ifdef CONFIG_PROVE_LOCKING @@ -1089,12 +1091,16 @@ static struct task_struct *copy_process(unsigned long clone_flags, #ifdef CONFIG_DEBUG_MUTEXES p->blocked_on = NULL; /* not blocked yet */ #endif - if (unlikely(current->ptrace)) - ptrace_fork(p, clone_flags); + + p->bts = NULL; /* Perform scheduler related setup. Assign this task to a CPU. */ sched_fork(p, clone_flags); + retval = perf_counter_init_task(p); + if (retval) + goto bad_fork_cleanup_policy; + if ((retval = audit_alloc(p))) goto bad_fork_cleanup_policy; /* copy all the process information */ @@ -1131,8 +1137,6 @@ static struct task_struct *copy_process(unsigned long clone_flags, } } - ftrace_graph_init_task(p); - p->pid = pid_nr(pid); p->tgid = p->pid; if (clone_flags & CLONE_THREAD) @@ -1141,7 +1145,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, if (current->nsproxy != p->nsproxy) { retval = ns_cgroup_clone(p, pid); if (retval) - goto bad_fork_free_graph; + goto bad_fork_free_pid; } p->set_child_tid = (clone_flags & CLONE_CHILD_SETTID) ? child_tidptr : NULL; @@ -1233,7 +1237,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, spin_unlock(¤t->sighand->siglock); write_unlock_irq(&tasklist_lock); retval = -ERESTARTNOINTR; - goto bad_fork_free_graph; + goto bad_fork_free_pid; } if (clone_flags & CLONE_THREAD) { @@ -1268,8 +1272,6 @@ static struct task_struct *copy_process(unsigned long clone_flags, cgroup_post_fork(p); return p; -bad_fork_free_graph: - ftrace_graph_exit_task(p); bad_fork_free_pid: if (pid != &init_struct_pid) free_pid(pid); @@ -1293,6 +1295,7 @@ bad_fork_cleanup_semundo: bad_fork_cleanup_audit: audit_free(p); bad_fork_cleanup_policy: + perf_counter_free_task(p); #ifdef CONFIG_NUMA mpol_put(p->mempolicy); bad_fork_cleanup_cgroup: @@ -1406,6 +1409,12 @@ long do_fork(unsigned long clone_flags, if (clone_flags & CLONE_VFORK) { p->vfork_done = &vfork; init_completion(&vfork); + } else if (!(clone_flags & CLONE_VM)) { + /* + * vfork will do an exec which will call + * set_task_comm() + */ + perf_counter_fork(p); } audit_finish_fork(p); diff --git a/kernel/futex.c b/kernel/futex.c index d546b2d53a6..80b5ce71659 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -19,6 +19,10 @@ * PRIVATE futexes by Eric Dumazet * Copyright (C) 2007 Eric Dumazet <dada1@cosmosbay.com> * + * Requeue-PI support by Darren Hart <dvhltc@us.ibm.com> + * Copyright (C) IBM Corporation, 2009 + * Thanks to Thomas Gleixner for conceptual design and careful reviews. + * * Thanks to Ben LaHaise for yelling "hashed waitqueues" loudly * enough at me, Linus for the original (flawed) idea, Matthew * Kirkwood for proof-of-concept implementation. @@ -96,8 +100,8 @@ struct futex_pi_state { */ struct futex_q { struct plist_node list; - /* There can only be a single waiter */ - wait_queue_head_t waiter; + /* Waiter reference */ + struct task_struct *task; /* Which hash list lock to use: */ spinlock_t *lock_ptr; @@ -107,7 +111,9 @@ struct futex_q { /* Optional priority inheritance state: */ struct futex_pi_state *pi_state; - struct task_struct *task; + + /* rt_waiter storage for requeue_pi: */ + struct rt_mutex_waiter *rt_waiter; /* Bitset for the optional bitmasked wakeup */ u32 bitset; @@ -278,6 +284,25 @@ void put_futex_key(int fshared, union futex_key *key) drop_futex_key_refs(key); } +/** + * futex_top_waiter() - Return the highest priority waiter on a futex + * @hb: the hash bucket the futex_q's reside in + * @key: the futex key (to distinguish it from other futex futex_q's) + * + * Must be called with the hb lock held. + */ +static struct futex_q *futex_top_waiter(struct futex_hash_bucket *hb, + union futex_key *key) +{ + struct futex_q *this; + + plist_for_each_entry(this, &hb->chain, list) { + if (match_futex(&this->key, key)) + return this; + } + return NULL; +} + static u32 cmpxchg_futex_value_locked(u32 __user *uaddr, u32 uval, u32 newval) { u32 curval; @@ -539,28 +564,160 @@ lookup_pi_state(u32 uval, struct futex_hash_bucket *hb, return 0; } +/** + * futex_lock_pi_atomic() - atomic work required to acquire a pi aware futex + * @uaddr: the pi futex user address + * @hb: the pi futex hash bucket + * @key: the futex key associated with uaddr and hb + * @ps: the pi_state pointer where we store the result of the + * lookup + * @task: the task to perform the atomic lock work for. This will + * be "current" except in the case of requeue pi. + * @set_waiters: force setting the FUTEX_WAITERS bit (1) or not (0) + * + * Returns: + * 0 - ready to wait + * 1 - acquired the lock + * <0 - error + * + * The hb->lock and futex_key refs shall be held by the caller. + */ +static int futex_lock_pi_atomic(u32 __user *uaddr, struct futex_hash_bucket *hb, + union futex_key *key, + struct futex_pi_state **ps, + struct task_struct *task, int set_waiters) +{ + int lock_taken, ret, ownerdied = 0; + u32 uval, newval, curval; + +retry: + ret = lock_taken = 0; + + /* + * To avoid races, we attempt to take the lock here again + * (by doing a 0 -> TID atomic cmpxchg), while holding all + * the locks. It will most likely not succeed. + */ + newval = task_pid_vnr(task); + if (set_waiters) + newval |= FUTEX_WAITERS; + + curval = cmpxchg_futex_value_locked(uaddr, 0, newval); + + if (unlikely(curval == -EFAULT)) + return -EFAULT; + + /* + * Detect deadlocks. + */ + if ((unlikely((curval & FUTEX_TID_MASK) == task_pid_vnr(task)))) + return -EDEADLK; + + /* + * Surprise - we got the lock. Just return to userspace: + */ + if (unlikely(!curval)) + return 1; + + uval = curval; + + /* + * Set the FUTEX_WAITERS flag, so the owner will know it has someone + * to wake at the next unlock. + */ + newval = curval | FUTEX_WAITERS; + + /* + * There are two cases, where a futex might have no owner (the + * owner TID is 0): OWNER_DIED. We take over the futex in this + * case. We also do an unconditional take over, when the owner + * of the futex died. + * + * This is safe as we are protected by the hash bucket lock ! + */ + if (unlikely(ownerdied || !(curval & FUTEX_TID_MASK))) { + /* Keep the OWNER_DIED bit */ + newval = (curval & ~FUTEX_TID_MASK) | task_pid_vnr(task); + ownerdied = 0; + lock_taken = 1; + } + + curval = cmpxchg_futex_value_locked(uaddr, uval, newval); + + if (unlikely(curval == -EFAULT)) + return -EFAULT; + if (unlikely(curval != uval)) + goto retry; + + /* + * We took the lock due to owner died take over. + */ + if (unlikely(lock_taken)) + return 1; + + /* + * We dont have the lock. Look up the PI state (or create it if + * we are the first waiter): + */ + ret = lookup_pi_state(uval, hb, key, ps); + + if (unlikely(ret)) { + switch (ret) { + case -ESRCH: + /* + * No owner found for this futex. Check if the + * OWNER_DIED bit is set to figure out whether + * this is a robust futex or not. + */ + if (get_futex_value_locked(&curval, uaddr)) + return -EFAULT; + + /* + * We simply start over in case of a robust + * futex. The code above will take the futex + * and return happy. + */ + if (curval & FUTEX_OWNER_DIED) { + ownerdied = 1; + goto retry; + } + default: + break; + } + } + + return ret; +} + /* * The hash bucket lock must be held when this is called. * Afterwards, the futex_q must not be accessed. */ static void wake_futex(struct futex_q *q) { - plist_del(&q->list, &q->list.plist); + struct task_struct *p = q->task; + /* - * The lock in wake_up_all() is a crucial memory barrier after the - * plist_del() and also before assigning to q->lock_ptr. + * We set q->lock_ptr = NULL _before_ we wake up the task. If + * a non futex wake up happens on another CPU then the task + * might exit and p would dereference a non existing task + * struct. Prevent this by holding a reference on p across the + * wake up. */ - wake_up(&q->waiter); + get_task_struct(p); + + plist_del(&q->list, &q->list.plist); /* - * The waiting task can free the futex_q as soon as this is written, - * without taking any locks. This must come last. - * - * A memory barrier is required here to prevent the following store to - * lock_ptr from getting ahead of the wakeup. Clearing the lock at the - * end of wake_up() does not prevent this store from moving. + * The waiting task can free the futex_q as soon as + * q->lock_ptr = NULL is written, without taking any locks. A + * memory barrier is required here to prevent the following + * store to lock_ptr from getting ahead of the plist_del. */ smp_wmb(); q->lock_ptr = NULL; + + wake_up_state(p, TASK_NORMAL); + put_task_struct(p); } static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this) @@ -689,7 +846,7 @@ static int futex_wake(u32 __user *uaddr, int fshared, int nr_wake, u32 bitset) plist_for_each_entry_safe(this, next, head, list) { if (match_futex (&this->key, &key)) { - if (this->pi_state) { + if (this->pi_state || this->rt_waiter) { ret = -EINVAL; break; } @@ -802,24 +959,185 @@ out: return ret; } -/* - * Requeue all waiters hashed on one physical page to another - * physical page. +/** + * requeue_futex() - Requeue a futex_q from one hb to another + * @q: the futex_q to requeue + * @hb1: the source hash_bucket + * @hb2: the target hash_bucket + * @key2: the new key for the requeued futex_q + */ +static inline +void requeue_futex(struct futex_q *q, struct futex_hash_bucket *hb1, + struct futex_hash_bucket *hb2, union futex_key *key2) +{ + + /* + * If key1 and key2 hash to the same bucket, no need to + * requeue. + */ + if (likely(&hb1->chain != &hb2->chain)) { + plist_del(&q->list, &hb1->chain); + plist_add(&q->list, &hb2->chain); + q->lock_ptr = &hb2->lock; +#ifdef CONFIG_DEBUG_PI_LIST + q->list.plist.lock = &hb2->lock; +#endif + } + get_futex_key_refs(key2); + q->key = *key2; +} + +/** + * requeue_pi_wake_futex() - Wake a task that acquired the lock during requeue + * q: the futex_q + * key: the key of the requeue target futex + * + * During futex_requeue, with requeue_pi=1, it is possible to acquire the + * target futex if it is uncontended or via a lock steal. Set the futex_q key + * to the requeue target futex so the waiter can detect the wakeup on the right + * futex, but remove it from the hb and NULL the rt_waiter so it can detect + * atomic lock acquisition. Must be called with the q->lock_ptr held. + */ +static inline +void requeue_pi_wake_futex(struct futex_q *q, union futex_key *key) +{ + drop_futex_key_refs(&q->key); + get_futex_key_refs(key); + q->key = *key; + + WARN_ON(plist_node_empty(&q->list)); + plist_del(&q->list, &q->list.plist); + + WARN_ON(!q->rt_waiter); + q->rt_waiter = NULL; + + wake_up_state(q->task, TASK_NORMAL); +} + +/** + * futex_proxy_trylock_atomic() - Attempt an atomic lock for the top waiter + * @pifutex: the user address of the to futex + * @hb1: the from futex hash bucket, must be locked by the caller + * @hb2: the to futex hash bucket, must be locked by the caller + * @key1: the from futex key + * @key2: the to futex key + * @ps: address to store the pi_state pointer + * @set_waiters: force setting the FUTEX_WAITERS bit (1) or not (0) + * + * Try and get the lock on behalf of the top waiter if we can do it atomically. + * Wake the top waiter if we succeed. If the caller specified set_waiters, + * then direct futex_lock_pi_atomic() to force setting the FUTEX_WAITERS bit. + * hb1 and hb2 must be held by the caller. + * + * Returns: + * 0 - failed to acquire the lock atomicly + * 1 - acquired the lock + * <0 - error + */ +static int futex_proxy_trylock_atomic(u32 __user *pifutex, + struct futex_hash_bucket *hb1, + struct futex_hash_bucket *hb2, + union futex_key *key1, union futex_key *key2, + struct futex_pi_state **ps, int set_waiters) +{ + struct futex_q *top_waiter = NULL; + u32 curval; + int ret; + + if (get_futex_value_locked(&curval, pifutex)) + return -EFAULT; + + /* + * Find the top_waiter and determine if there are additional waiters. + * If the caller intends to requeue more than 1 waiter to pifutex, + * force futex_lock_pi_atomic() to set the FUTEX_WAITERS bit now, + * as we have means to handle the possible fault. If not, don't set + * the bit unecessarily as it will force the subsequent unlock to enter + * the kernel. + */ + top_waiter = futex_top_waiter(hb1, key1); + + /* There are no waiters, nothing for us to do. */ + if (!top_waiter) + return 0; + + /* + * Try to take the lock for top_waiter. Set the FUTEX_WAITERS bit in + * the contended case or if set_waiters is 1. The pi_state is returned + * in ps in contended cases. + */ + ret = futex_lock_pi_atomic(pifutex, hb2, key2, ps, top_waiter->task, + set_waiters); + if (ret == 1) + requeue_pi_wake_futex(top_waiter, key2); + + return ret; +} + +/** + * futex_requeue() - Requeue waiters from uaddr1 to uaddr2 + * uaddr1: source futex user address + * uaddr2: target futex user address + * nr_wake: number of waiters to wake (must be 1 for requeue_pi) + * nr_requeue: number of waiters to requeue (0-INT_MAX) + * requeue_pi: if we are attempting to requeue from a non-pi futex to a + * pi futex (pi to pi requeue is not supported) + * + * Requeue waiters on uaddr1 to uaddr2. In the requeue_pi case, try to acquire + * uaddr2 atomically on behalf of the top waiter. + * + * Returns: + * >=0 - on success, the number of tasks requeued or woken + * <0 - on error */ static int futex_requeue(u32 __user *uaddr1, int fshared, u32 __user *uaddr2, - int nr_wake, int nr_requeue, u32 *cmpval) + int nr_wake, int nr_requeue, u32 *cmpval, + int requeue_pi) { union futex_key key1 = FUTEX_KEY_INIT, key2 = FUTEX_KEY_INIT; + int drop_count = 0, task_count = 0, ret; + struct futex_pi_state *pi_state = NULL; struct futex_hash_bucket *hb1, *hb2; struct plist_head *head1; struct futex_q *this, *next; - int ret, drop_count = 0; + u32 curval2; + + if (requeue_pi) { + /* + * requeue_pi requires a pi_state, try to allocate it now + * without any locks in case it fails. + */ + if (refill_pi_state_cache()) + return -ENOMEM; + /* + * requeue_pi must wake as many tasks as it can, up to nr_wake + * + nr_requeue, since it acquires the rt_mutex prior to + * returning to userspace, so as to not leave the rt_mutex with + * waiters and no owner. However, second and third wake-ups + * cannot be predicted as they involve race conditions with the + * first wake and a fault while looking up the pi_state. Both + * pthread_cond_signal() and pthread_cond_broadcast() should + * use nr_wake=1. + */ + if (nr_wake != 1) + return -EINVAL; + } retry: + if (pi_state != NULL) { + /* + * We will have to lookup the pi_state again, so free this one + * to keep the accounting correct. + */ + free_pi_state(pi_state); + pi_state = NULL; + } + ret = get_futex_key(uaddr1, fshared, &key1, VERIFY_READ); if (unlikely(ret != 0)) goto out; - ret = get_futex_key(uaddr2, fshared, &key2, VERIFY_READ); + ret = get_futex_key(uaddr2, fshared, &key2, + requeue_pi ? VERIFY_WRITE : VERIFY_READ); if (unlikely(ret != 0)) goto out_put_key1; @@ -854,32 +1172,99 @@ retry_private: } } + if (requeue_pi && (task_count - nr_wake < nr_requeue)) { + /* + * Attempt to acquire uaddr2 and wake the top waiter. If we + * intend to requeue waiters, force setting the FUTEX_WAITERS + * bit. We force this here where we are able to easily handle + * faults rather in the requeue loop below. + */ + ret = futex_proxy_trylock_atomic(uaddr2, hb1, hb2, &key1, + &key2, &pi_state, nr_requeue); + + /* + * At this point the top_waiter has either taken uaddr2 or is + * waiting on it. If the former, then the pi_state will not + * exist yet, look it up one more time to ensure we have a + * reference to it. + */ + if (ret == 1) { + WARN_ON(pi_state); + task_count++; + ret = get_futex_value_locked(&curval2, uaddr2); + if (!ret) + ret = lookup_pi_state(curval2, hb2, &key2, + &pi_state); + } + + switch (ret) { + case 0: + break; + case -EFAULT: + double_unlock_hb(hb1, hb2); + put_futex_key(fshared, &key2); + put_futex_key(fshared, &key1); + ret = get_user(curval2, uaddr2); + if (!ret) + goto retry; + goto out; + case -EAGAIN: + /* The owner was exiting, try again. */ + double_unlock_hb(hb1, hb2); + put_futex_key(fshared, &key2); + put_futex_key(fshared, &key1); + cond_resched(); + goto retry; + default: + goto out_unlock; + } + } + head1 = &hb1->chain; plist_for_each_entry_safe(this, next, head1, list) { - if (!match_futex (&this->key, &key1)) + if (task_count - nr_wake >= nr_requeue) + break; + + if (!match_futex(&this->key, &key1)) continue; - if (++ret <= nr_wake) { + + WARN_ON(!requeue_pi && this->rt_waiter); + WARN_ON(requeue_pi && !this->rt_waiter); + + /* + * Wake nr_wake waiters. For requeue_pi, if we acquired the + * lock, we already woke the top_waiter. If not, it will be + * woken by futex_unlock_pi(). + */ + if (++task_count <= nr_wake && !requeue_pi) { wake_futex(this); - } else { - /* - * If key1 and key2 hash to the same bucket, no need to - * requeue. - */ - if (likely(head1 != &hb2->chain)) { - plist_del(&this->list, &hb1->chain); - plist_add(&this->list, &hb2->chain); - this->lock_ptr = &hb2->lock; -#ifdef CONFIG_DEBUG_PI_LIST - this->list.plist.lock = &hb2->lock; -#endif - } - this->key = key2; - get_futex_key_refs(&key2); - drop_count++; + continue; + } - if (ret - nr_wake >= nr_requeue) - break; + /* + * Requeue nr_requeue waiters and possibly one more in the case + * of requeue_pi if we couldn't acquire the lock atomically. + */ + if (requeue_pi) { + /* Prepare the waiter to take the rt_mutex. */ + atomic_inc(&pi_state->refcount); + this->pi_state = pi_state; + ret = rt_mutex_start_proxy_lock(&pi_state->pi_mutex, + this->rt_waiter, + this->task, 1); + if (ret == 1) { + /* We got the lock. */ + requeue_pi_wake_futex(this, &key2); + continue; + } else if (ret) { + /* -EDEADLK */ + this->pi_state = NULL; + free_pi_state(pi_state); + goto out_unlock; + } } + requeue_futex(this, hb1, hb2, &key2); + drop_count++; } out_unlock: @@ -899,7 +1284,9 @@ out_put_keys: out_put_key1: put_futex_key(fshared, &key1); out: - return ret; + if (pi_state != NULL) + free_pi_state(pi_state); + return ret ? ret : task_count; } /* The key must be already stored in q->key. */ @@ -907,8 +1294,6 @@ static inline struct futex_hash_bucket *queue_lock(struct futex_q *q) { struct futex_hash_bucket *hb; - init_waitqueue_head(&q->waiter); - get_futex_key_refs(&q->key); hb = hash_futex(&q->key); q->lock_ptr = &hb->lock; @@ -1119,35 +1504,149 @@ handle_fault: */ #define FLAGS_SHARED 0x01 #define FLAGS_CLOCKRT 0x02 +#define FLAGS_HAS_TIMEOUT 0x04 static long futex_wait_restart(struct restart_block *restart); -static int futex_wait(u32 __user *uaddr, int fshared, - u32 val, ktime_t *abs_time, u32 bitset, int clockrt) +/** + * fixup_owner() - Post lock pi_state and corner case management + * @uaddr: user address of the futex + * @fshared: whether the futex is shared (1) or not (0) + * @q: futex_q (contains pi_state and access to the rt_mutex) + * @locked: if the attempt to take the rt_mutex succeeded (1) or not (0) + * + * After attempting to lock an rt_mutex, this function is called to cleanup + * the pi_state owner as well as handle race conditions that may allow us to + * acquire the lock. Must be called with the hb lock held. + * + * Returns: + * 1 - success, lock taken + * 0 - success, lock not taken + * <0 - on error (-EFAULT) + */ +static int fixup_owner(u32 __user *uaddr, int fshared, struct futex_q *q, + int locked) { - struct task_struct *curr = current; - struct restart_block *restart; - DECLARE_WAITQUEUE(wait, curr); - struct futex_hash_bucket *hb; - struct futex_q q; - u32 uval; - int ret; - struct hrtimer_sleeper t; - int rem = 0; + struct task_struct *owner; + int ret = 0; - if (!bitset) - return -EINVAL; + if (locked) { + /* + * Got the lock. We might not be the anticipated owner if we + * did a lock-steal - fix up the PI-state in that case: + */ + if (q->pi_state->owner != current) + ret = fixup_pi_state_owner(uaddr, q, current, fshared); + goto out; + } - q.pi_state = NULL; - q.bitset = bitset; -retry: - q.key = FUTEX_KEY_INIT; - ret = get_futex_key(uaddr, fshared, &q.key, VERIFY_READ); - if (unlikely(ret != 0)) + /* + * Catch the rare case, where the lock was released when we were on the + * way back before we locked the hash bucket. + */ + if (q->pi_state->owner == current) { + /* + * Try to get the rt_mutex now. This might fail as some other + * task acquired the rt_mutex after we removed ourself from the + * rt_mutex waiters list. + */ + if (rt_mutex_trylock(&q->pi_state->pi_mutex)) { + locked = 1; + goto out; + } + + /* + * pi_state is incorrect, some other task did a lock steal and + * we returned due to timeout or signal without taking the + * rt_mutex. Too late. We can access the rt_mutex_owner without + * locking, as the other task is now blocked on the hash bucket + * lock. Fix the state up. + */ + owner = rt_mutex_owner(&q->pi_state->pi_mutex); + ret = fixup_pi_state_owner(uaddr, q, owner, fshared); goto out; + } -retry_private: - hb = queue_lock(&q); + /* + * Paranoia check. If we did not take the lock, then we should not be + * the owner, nor the pending owner, of the rt_mutex. + */ + if (rt_mutex_owner(&q->pi_state->pi_mutex) == current) + printk(KERN_ERR "fixup_owner: ret = %d pi-mutex: %p " + "pi-state %p\n", ret, + q->pi_state->pi_mutex.owner, + q->pi_state->owner); + +out: + return ret ? ret : locked; +} + +/** + * futex_wait_queue_me() - queue_me() and wait for wakeup, timeout, or signal + * @hb: the futex hash bucket, must be locked by the caller + * @q: the futex_q to queue up on + * @timeout: the prepared hrtimer_sleeper, or null for no timeout + */ +static void futex_wait_queue_me(struct futex_hash_bucket *hb, struct futex_q *q, + struct hrtimer_sleeper *timeout) +{ + queue_me(q, hb); + + /* + * There might have been scheduling since the queue_me(), as we + * cannot hold a spinlock across the get_user() in case it + * faults, and we cannot just set TASK_INTERRUPTIBLE state when + * queueing ourselves into the futex hash. This code thus has to + * rely on the futex_wake() code removing us from hash when it + * wakes us up. + */ + set_current_state(TASK_INTERRUPTIBLE); + + /* Arm the timer */ + if (timeout) { + hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); + if (!hrtimer_active(&timeout->timer)) + timeout->task = NULL; + } + + /* + * !plist_node_empty() is safe here without any lock. + * q.lock_ptr != 0 is not safe, because of ordering against wakeup. + */ + if (likely(!plist_node_empty(&q->list))) { + /* + * If the timer has already expired, current will already be + * flagged for rescheduling. Only call schedule if there + * is no timeout, or if it has yet to expire. + */ + if (!timeout || timeout->task) + schedule(); + } + __set_current_state(TASK_RUNNING); +} + +/** + * futex_wait_setup() - Prepare to wait on a futex + * @uaddr: the futex userspace address + * @val: the expected value + * @fshared: whether the futex is shared (1) or not (0) + * @q: the associated futex_q + * @hb: storage for hash_bucket pointer to be returned to caller + * + * Setup the futex_q and locate the hash_bucket. Get the futex value and + * compare it with the expected value. Handle atomic faults internally. + * Return with the hb lock held and a q.key reference on success, and unlocked + * with no q.key reference on failure. + * + * Returns: + * 0 - uaddr contains val and hb has been locked + * <1 - -EFAULT or -EWOULDBLOCK (uaddr does not contain val) and hb is unlcoked + */ +static int futex_wait_setup(u32 __user *uaddr, u32 val, int fshared, + struct futex_q *q, struct futex_hash_bucket **hb) +{ + u32 uval; + int ret; /* * Access the page AFTER the hash-bucket is locked. @@ -1165,95 +1664,83 @@ retry_private: * A consequence is that futex_wait() can return zero and absorb * a wakeup when *uaddr != val on entry to the syscall. This is * rare, but normal. - * - * For shared futexes, we hold the mmap semaphore, so the mapping - * cannot have changed since we looked it up in get_futex_key. */ +retry: + q->key = FUTEX_KEY_INIT; + ret = get_futex_key(uaddr, fshared, &q->key, VERIFY_READ); + if (unlikely(ret != 0)) + return ret; + +retry_private: + *hb = queue_lock(q); + ret = get_futex_value_locked(&uval, uaddr); - if (unlikely(ret)) { - queue_unlock(&q, hb); + if (ret) { + queue_unlock(q, *hb); ret = get_user(uval, uaddr); if (ret) - goto out_put_key; + goto out; if (!fshared) goto retry_private; - put_futex_key(fshared, &q.key); + put_futex_key(fshared, &q->key); goto retry; } - ret = -EWOULDBLOCK; - if (unlikely(uval != val)) { - queue_unlock(&q, hb); - goto out_put_key; - } - /* Only actually queue if *uaddr contained val. */ - queue_me(&q, hb); + if (uval != val) { + queue_unlock(q, *hb); + ret = -EWOULDBLOCK; + } - /* - * There might have been scheduling since the queue_me(), as we - * cannot hold a spinlock across the get_user() in case it - * faults, and we cannot just set TASK_INTERRUPTIBLE state when - * queueing ourselves into the futex hash. This code thus has to - * rely on the futex_wake() code removing us from hash when it - * wakes us up. - */ +out: + if (ret) + put_futex_key(fshared, &q->key); + return ret; +} - /* add_wait_queue is the barrier after __set_current_state. */ - __set_current_state(TASK_INTERRUPTIBLE); - add_wait_queue(&q.waiter, &wait); - /* - * !plist_node_empty() is safe here without any lock. - * q.lock_ptr != 0 is not safe, because of ordering against wakeup. - */ - if (likely(!plist_node_empty(&q.list))) { - if (!abs_time) - schedule(); - else { - hrtimer_init_on_stack(&t.timer, - clockrt ? CLOCK_REALTIME : - CLOCK_MONOTONIC, - HRTIMER_MODE_ABS); - hrtimer_init_sleeper(&t, current); - hrtimer_set_expires_range_ns(&t.timer, *abs_time, - current->timer_slack_ns); - - hrtimer_start_expires(&t.timer, HRTIMER_MODE_ABS); - if (!hrtimer_active(&t.timer)) - t.task = NULL; +static int futex_wait(u32 __user *uaddr, int fshared, + u32 val, ktime_t *abs_time, u32 bitset, int clockrt) +{ + struct hrtimer_sleeper timeout, *to = NULL; + struct restart_block *restart; + struct futex_hash_bucket *hb; + struct futex_q q; + int ret; - /* - * the timer could have already expired, in which - * case current would be flagged for rescheduling. - * Don't bother calling schedule. - */ - if (likely(t.task)) - schedule(); + if (!bitset) + return -EINVAL; - hrtimer_cancel(&t.timer); + q.pi_state = NULL; + q.bitset = bitset; + q.rt_waiter = NULL; - /* Flag if a timeout occured */ - rem = (t.task == NULL); + if (abs_time) { + to = &timeout; - destroy_hrtimer_on_stack(&t.timer); - } + hrtimer_init_on_stack(&to->timer, clockrt ? CLOCK_REALTIME : + CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + hrtimer_init_sleeper(to, current); + hrtimer_set_expires_range_ns(&to->timer, *abs_time, + current->timer_slack_ns); } - __set_current_state(TASK_RUNNING); - /* - * NOTE: we don't remove ourselves from the waitqueue because - * we are the only user of it. - */ + /* Prepare to wait on uaddr. */ + ret = futex_wait_setup(uaddr, val, fshared, &q, &hb); + if (ret) + goto out; + + /* queue_me and wait for wakeup, timeout, or a signal. */ + futex_wait_queue_me(hb, &q, to); /* If we were woken (and unqueued), we succeeded, whatever. */ ret = 0; if (!unqueue_me(&q)) goto out_put_key; ret = -ETIMEDOUT; - if (rem) + if (to && !to->task) goto out_put_key; /* @@ -1270,7 +1757,7 @@ retry_private: restart->futex.val = val; restart->futex.time = abs_time->tv64; restart->futex.bitset = bitset; - restart->futex.flags = 0; + restart->futex.flags = FLAGS_HAS_TIMEOUT; if (fshared) restart->futex.flags |= FLAGS_SHARED; @@ -1282,6 +1769,10 @@ retry_private: out_put_key: put_futex_key(fshared, &q.key); out: + if (to) { + hrtimer_cancel(&to->timer); + destroy_hrtimer_on_stack(&to->timer); + } return ret; } @@ -1290,13 +1781,16 @@ static long futex_wait_restart(struct restart_block *restart) { u32 __user *uaddr = (u32 __user *)restart->futex.uaddr; int fshared = 0; - ktime_t t; + ktime_t t, *tp = NULL; - t.tv64 = restart->futex.time; + if (restart->futex.flags & FLAGS_HAS_TIMEOUT) { + t.tv64 = restart->futex.time; + tp = &t; + } restart->fn = do_no_restart_syscall; if (restart->futex.flags & FLAGS_SHARED) fshared = 1; - return (long)futex_wait(uaddr, fshared, restart->futex.val, &t, + return (long)futex_wait(uaddr, fshared, restart->futex.val, tp, restart->futex.bitset, restart->futex.flags & FLAGS_CLOCKRT); } @@ -1312,11 +1806,10 @@ static int futex_lock_pi(u32 __user *uaddr, int fshared, int detect, ktime_t *time, int trylock) { struct hrtimer_sleeper timeout, *to = NULL; - struct task_struct *curr = current; struct futex_hash_bucket *hb; - u32 uval, newval, curval; + u32 uval; struct futex_q q; - int ret, lock_taken, ownerdied = 0; + int res, ret; if (refill_pi_state_cache()) return -ENOMEM; @@ -1330,6 +1823,7 @@ static int futex_lock_pi(u32 __user *uaddr, int fshared, } q.pi_state = NULL; + q.rt_waiter = NULL; retry: q.key = FUTEX_KEY_INIT; ret = get_futex_key(uaddr, fshared, &q.key, VERIFY_WRITE); @@ -1339,81 +1833,15 @@ retry: retry_private: hb = queue_lock(&q); -retry_locked: - ret = lock_taken = 0; - - /* - * To avoid races, we attempt to take the lock here again - * (by doing a 0 -> TID atomic cmpxchg), while holding all - * the locks. It will most likely not succeed. - */ - newval = task_pid_vnr(current); - - curval = cmpxchg_futex_value_locked(uaddr, 0, newval); - - if (unlikely(curval == -EFAULT)) - goto uaddr_faulted; - - /* - * Detect deadlocks. In case of REQUEUE_PI this is a valid - * situation and we return success to user space. - */ - if (unlikely((curval & FUTEX_TID_MASK) == task_pid_vnr(current))) { - ret = -EDEADLK; - goto out_unlock_put_key; - } - - /* - * Surprise - we got the lock. Just return to userspace: - */ - if (unlikely(!curval)) - goto out_unlock_put_key; - - uval = curval; - - /* - * Set the WAITERS flag, so the owner will know it has someone - * to wake at next unlock - */ - newval = curval | FUTEX_WAITERS; - - /* - * There are two cases, where a futex might have no owner (the - * owner TID is 0): OWNER_DIED. We take over the futex in this - * case. We also do an unconditional take over, when the owner - * of the futex died. - * - * This is safe as we are protected by the hash bucket lock ! - */ - if (unlikely(ownerdied || !(curval & FUTEX_TID_MASK))) { - /* Keep the OWNER_DIED bit */ - newval = (curval & ~FUTEX_TID_MASK) | task_pid_vnr(current); - ownerdied = 0; - lock_taken = 1; - } - - curval = cmpxchg_futex_value_locked(uaddr, uval, newval); - - if (unlikely(curval == -EFAULT)) - goto uaddr_faulted; - if (unlikely(curval != uval)) - goto retry_locked; - - /* - * We took the lock due to owner died take over. - */ - if (unlikely(lock_taken)) - goto out_unlock_put_key; - - /* - * We dont have the lock. Look up the PI state (or create it if - * we are the first waiter): - */ - ret = lookup_pi_state(uval, hb, &q.key, &q.pi_state); - + ret = futex_lock_pi_atomic(uaddr, hb, &q.key, &q.pi_state, current, 0); if (unlikely(ret)) { switch (ret) { - + case 1: + /* We got the lock. */ + ret = 0; + goto out_unlock_put_key; + case -EFAULT: + goto uaddr_faulted; case -EAGAIN: /* * Task is exiting and we just wait for the @@ -1423,25 +1851,6 @@ retry_locked: put_futex_key(fshared, &q.key); cond_resched(); goto retry; - - case -ESRCH: - /* - * No owner found for this futex. Check if the - * OWNER_DIED bit is set to figure out whether - * this is a robust futex or not. - */ - if (get_futex_value_locked(&curval, uaddr)) - goto uaddr_faulted; - - /* - * We simply start over in case of a robust - * futex. The code above will take the futex - * and return happy. - */ - if (curval & FUTEX_OWNER_DIED) { - ownerdied = 1; - goto retry_locked; - } default: goto out_unlock_put_key; } @@ -1465,71 +1874,21 @@ retry_locked: } spin_lock(q.lock_ptr); - - if (!ret) { - /* - * Got the lock. We might not be the anticipated owner - * if we did a lock-steal - fix up the PI-state in - * that case: - */ - if (q.pi_state->owner != curr) - ret = fixup_pi_state_owner(uaddr, &q, curr, fshared); - } else { - /* - * Catch the rare case, where the lock was released - * when we were on the way back before we locked the - * hash bucket. - */ - if (q.pi_state->owner == curr) { - /* - * Try to get the rt_mutex now. This might - * fail as some other task acquired the - * rt_mutex after we removed ourself from the - * rt_mutex waiters list. - */ - if (rt_mutex_trylock(&q.pi_state->pi_mutex)) - ret = 0; - else { - /* - * pi_state is incorrect, some other - * task did a lock steal and we - * returned due to timeout or signal - * without taking the rt_mutex. Too - * late. We can access the - * rt_mutex_owner without locking, as - * the other task is now blocked on - * the hash bucket lock. Fix the state - * up. - */ - struct task_struct *owner; - int res; - - owner = rt_mutex_owner(&q.pi_state->pi_mutex); - res = fixup_pi_state_owner(uaddr, &q, owner, - fshared); - - /* propagate -EFAULT, if the fixup failed */ - if (res) - ret = res; - } - } else { - /* - * Paranoia check. If we did not take the lock - * in the trylock above, then we should not be - * the owner of the rtmutex, neither the real - * nor the pending one: - */ - if (rt_mutex_owner(&q.pi_state->pi_mutex) == curr) - printk(KERN_ERR "futex_lock_pi: ret = %d " - "pi-mutex: %p pi-state %p\n", ret, - q.pi_state->pi_mutex.owner, - q.pi_state->owner); - } - } + /* + * Fixup the pi_state owner and possibly acquire the lock if we + * haven't already. + */ + res = fixup_owner(uaddr, fshared, &q, !ret); + /* + * If fixup_owner() returned an error, proprogate that. If it acquired + * the lock, clear our -ETIMEDOUT or -EINTR. + */ + if (res) + ret = (res < 0) ? res : 0; /* - * If fixup_pi_state_owner() faulted and was unable to handle the - * fault, unlock it and return the fault to userspace. + * If fixup_owner() faulted and was unable to handle the fault, unlock + * it and return the fault to userspace. */ if (ret && (rt_mutex_owner(&q.pi_state->pi_mutex) == current)) rt_mutex_unlock(&q.pi_state->pi_mutex); @@ -1537,9 +1896,7 @@ retry_locked: /* Unqueue and drop the lock */ unqueue_me_pi(&q); - if (to) - destroy_hrtimer_on_stack(&to->timer); - return ret != -EINTR ? ret : -ERESTARTNOINTR; + goto out; out_unlock_put_key: queue_unlock(&q, hb); @@ -1549,7 +1906,7 @@ out_put_key: out: if (to) destroy_hrtimer_on_stack(&to->timer); - return ret; + return ret != -EINTR ? ret : -ERESTARTNOINTR; uaddr_faulted: /* @@ -1572,7 +1929,6 @@ uaddr_faulted: goto retry; } - /* * Userspace attempted a TID -> 0 atomic transition, and failed. * This is the in-kernel slowpath: we look up the PI state (if any), @@ -1674,6 +2030,229 @@ pi_faulted: return ret; } +/** + * handle_early_requeue_pi_wakeup() - Detect early wakeup on the initial futex + * @hb: the hash_bucket futex_q was original enqueued on + * @q: the futex_q woken while waiting to be requeued + * @key2: the futex_key of the requeue target futex + * @timeout: the timeout associated with the wait (NULL if none) + * + * Detect if the task was woken on the initial futex as opposed to the requeue + * target futex. If so, determine if it was a timeout or a signal that caused + * the wakeup and return the appropriate error code to the caller. Must be + * called with the hb lock held. + * + * Returns + * 0 - no early wakeup detected + * <0 - -ETIMEDOUT or -ERESTARTNOINTR + */ +static inline +int handle_early_requeue_pi_wakeup(struct futex_hash_bucket *hb, + struct futex_q *q, union futex_key *key2, + struct hrtimer_sleeper *timeout) +{ + int ret = 0; + + /* + * With the hb lock held, we avoid races while we process the wakeup. + * We only need to hold hb (and not hb2) to ensure atomicity as the + * wakeup code can't change q.key from uaddr to uaddr2 if we hold hb. + * It can't be requeued from uaddr2 to something else since we don't + * support a PI aware source futex for requeue. + */ + if (!match_futex(&q->key, key2)) { + WARN_ON(q->lock_ptr && (&hb->lock != q->lock_ptr)); + /* + * We were woken prior to requeue by a timeout or a signal. + * Unqueue the futex_q and determine which it was. + */ + plist_del(&q->list, &q->list.plist); + drop_futex_key_refs(&q->key); + + if (timeout && !timeout->task) + ret = -ETIMEDOUT; + else + ret = -ERESTARTNOINTR; + } + return ret; +} + +/** + * futex_wait_requeue_pi() - Wait on uaddr and take uaddr2 + * @uaddr: the futex we initialyl wait on (non-pi) + * @fshared: whether the futexes are shared (1) or not (0). They must be + * the same type, no requeueing from private to shared, etc. + * @val: the expected value of uaddr + * @abs_time: absolute timeout + * @bitset: 32 bit wakeup bitset set by userspace, defaults to all. + * @clockrt: whether to use CLOCK_REALTIME (1) or CLOCK_MONOTONIC (0) + * @uaddr2: the pi futex we will take prior to returning to user-space + * + * The caller will wait on uaddr and will be requeued by futex_requeue() to + * uaddr2 which must be PI aware. Normal wakeup will wake on uaddr2 and + * complete the acquisition of the rt_mutex prior to returning to userspace. + * This ensures the rt_mutex maintains an owner when it has waiters; without + * one, the pi logic wouldn't know which task to boost/deboost, if there was a + * need to. + * + * We call schedule in futex_wait_queue_me() when we enqueue and return there + * via the following: + * 1) wakeup on uaddr2 after an atomic lock acquisition by futex_requeue() + * 2) wakeup on uaddr2 after a requeue and subsequent unlock + * 3) signal (before or after requeue) + * 4) timeout (before or after requeue) + * + * If 3, we setup a restart_block with futex_wait_requeue_pi() as the function. + * + * If 2, we may then block on trying to take the rt_mutex and return via: + * 5) successful lock + * 6) signal + * 7) timeout + * 8) other lock acquisition failure + * + * If 6, we setup a restart_block with futex_lock_pi() as the function. + * + * If 4 or 7, we cleanup and return with -ETIMEDOUT. + * + * Returns: + * 0 - On success + * <0 - On error + */ +static int futex_wait_requeue_pi(u32 __user *uaddr, int fshared, + u32 val, ktime_t *abs_time, u32 bitset, + int clockrt, u32 __user *uaddr2) +{ + struct hrtimer_sleeper timeout, *to = NULL; + struct rt_mutex_waiter rt_waiter; + struct rt_mutex *pi_mutex = NULL; + struct futex_hash_bucket *hb; + union futex_key key2; + struct futex_q q; + int res, ret; + + if (!bitset) + return -EINVAL; + + if (abs_time) { + to = &timeout; + hrtimer_init_on_stack(&to->timer, clockrt ? CLOCK_REALTIME : + CLOCK_MONOTONIC, HRTIMER_MODE_ABS); + hrtimer_init_sleeper(to, current); + hrtimer_set_expires_range_ns(&to->timer, *abs_time, + current->timer_slack_ns); + } + + /* + * The waiter is allocated on our stack, manipulated by the requeue + * code while we sleep on uaddr. + */ + debug_rt_mutex_init_waiter(&rt_waiter); + rt_waiter.task = NULL; + + q.pi_state = NULL; + q.bitset = bitset; + q.rt_waiter = &rt_waiter; + + key2 = FUTEX_KEY_INIT; + ret = get_futex_key(uaddr2, fshared, &key2, VERIFY_WRITE); + if (unlikely(ret != 0)) + goto out; + + /* Prepare to wait on uaddr. */ + ret = futex_wait_setup(uaddr, val, fshared, &q, &hb); + if (ret) + goto out_key2; + + /* Queue the futex_q, drop the hb lock, wait for wakeup. */ + futex_wait_queue_me(hb, &q, to); + + spin_lock(&hb->lock); + ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to); + spin_unlock(&hb->lock); + if (ret) + goto out_put_keys; + + /* + * In order for us to be here, we know our q.key == key2, and since + * we took the hb->lock above, we also know that futex_requeue() has + * completed and we no longer have to concern ourselves with a wakeup + * race with the atomic proxy lock acquition by the requeue code. + */ + + /* Check if the requeue code acquired the second futex for us. */ + if (!q.rt_waiter) { + /* + * Got the lock. We might not be the anticipated owner if we + * did a lock-steal - fix up the PI-state in that case. + */ + if (q.pi_state && (q.pi_state->owner != current)) { + spin_lock(q.lock_ptr); + ret = fixup_pi_state_owner(uaddr2, &q, current, + fshared); + spin_unlock(q.lock_ptr); + } + } else { + /* + * We have been woken up by futex_unlock_pi(), a timeout, or a + * signal. futex_unlock_pi() will not destroy the lock_ptr nor + * the pi_state. + */ + WARN_ON(!&q.pi_state); + pi_mutex = &q.pi_state->pi_mutex; + ret = rt_mutex_finish_proxy_lock(pi_mutex, to, &rt_waiter, 1); + debug_rt_mutex_free_waiter(&rt_waiter); + + spin_lock(q.lock_ptr); + /* + * Fixup the pi_state owner and possibly acquire the lock if we + * haven't already. + */ + res = fixup_owner(uaddr2, fshared, &q, !ret); + /* + * If fixup_owner() returned an error, proprogate that. If it + * acquired the lock, clear our -ETIMEDOUT or -EINTR. + */ + if (res) + ret = (res < 0) ? res : 0; + + /* Unqueue and drop the lock. */ + unqueue_me_pi(&q); + } + + /* + * If fixup_pi_state_owner() faulted and was unable to handle the + * fault, unlock the rt_mutex and return the fault to userspace. + */ + if (ret == -EFAULT) { + if (rt_mutex_owner(pi_mutex) == current) + rt_mutex_unlock(pi_mutex); + } else if (ret == -EINTR) { + /* + * We've already been requeued, but we have no way to + * restart by calling futex_lock_pi() directly. We + * could restart the syscall, but that will look at + * the user space value and return right away. So we + * drop back with EWOULDBLOCK to tell user space that + * "val" has been changed. That's the same what the + * restart of the syscall would do in + * futex_wait_setup(). + */ + ret = -EWOULDBLOCK; + } + +out_put_keys: + put_futex_key(fshared, &q.key); +out_key2: + put_futex_key(fshared, &key2); + +out: + if (to) { + hrtimer_cancel(&to->timer); + destroy_hrtimer_on_stack(&to->timer); + } + return ret; +} + /* * Support for robust futexes: the kernel cleans up held futexes at * thread exit time. @@ -1896,7 +2475,7 @@ long do_futex(u32 __user *uaddr, int op, u32 val, ktime_t *timeout, fshared = 1; clockrt = op & FUTEX_CLOCK_REALTIME; - if (clockrt && cmd != FUTEX_WAIT_BITSET) + if (clockrt && cmd != FUTEX_WAIT_BITSET && cmd != FUTEX_WAIT_REQUEUE_PI) return -ENOSYS; switch (cmd) { @@ -1911,10 +2490,11 @@ long do_futex(u32 __user *uaddr, int op, u32 val, ktime_t *timeout, ret = futex_wake(uaddr, fshared, val, val3); break; case FUTEX_REQUEUE: - ret = futex_requeue(uaddr, fshared, uaddr2, val, val2, NULL); + ret = futex_requeue(uaddr, fshared, uaddr2, val, val2, NULL, 0); break; case FUTEX_CMP_REQUEUE: - ret = futex_requeue(uaddr, fshared, uaddr2, val, val2, &val3); + ret = futex_requeue(uaddr, fshared, uaddr2, val, val2, &val3, + 0); break; case FUTEX_WAKE_OP: ret = futex_wake_op(uaddr, fshared, uaddr2, val, val2, val3); @@ -1931,6 +2511,15 @@ long do_futex(u32 __user *uaddr, int op, u32 val, ktime_t *timeout, if (futex_cmpxchg_enabled) ret = futex_lock_pi(uaddr, fshared, 0, timeout, 1); break; + case FUTEX_WAIT_REQUEUE_PI: + val3 = FUTEX_BITSET_MATCH_ANY; + ret = futex_wait_requeue_pi(uaddr, fshared, val, timeout, val3, + clockrt, uaddr2); + break; + case FUTEX_CMP_REQUEUE_PI: + ret = futex_requeue(uaddr, fshared, uaddr2, val, val2, &val3, + 1); + break; default: ret = -ENOSYS; } @@ -1948,7 +2537,8 @@ SYSCALL_DEFINE6(futex, u32 __user *, uaddr, int, op, u32, val, int cmd = op & FUTEX_CMD_MASK; if (utime && (cmd == FUTEX_WAIT || cmd == FUTEX_LOCK_PI || - cmd == FUTEX_WAIT_BITSET)) { + cmd == FUTEX_WAIT_BITSET || + cmd == FUTEX_WAIT_REQUEUE_PI)) { if (copy_from_user(&ts, utime, sizeof(ts)) != 0) return -EFAULT; if (!timespec_valid(&ts)) @@ -1960,11 +2550,11 @@ SYSCALL_DEFINE6(futex, u32 __user *, uaddr, int, op, u32, val, tp = &t; } /* - * requeue parameter in 'utime' if cmd == FUTEX_REQUEUE. + * requeue parameter in 'utime' if cmd == FUTEX_*_REQUEUE_*. * number of waiters to wake in 'utime' if cmd == FUTEX_WAKE_OP. */ if (cmd == FUTEX_REQUEUE || cmd == FUTEX_CMP_REQUEUE || - cmd == FUTEX_WAKE_OP) + cmd == FUTEX_CMP_REQUEUE_PI || cmd == FUTEX_WAKE_OP) val2 = (u32) (unsigned long) utime; return do_futex(uaddr, op, val, tp, uaddr2, val2, val3); diff --git a/kernel/irq/Makefile b/kernel/irq/Makefile index 3394f8f5296..7d047808419 100644 --- a/kernel/irq/Makefile +++ b/kernel/irq/Makefile @@ -3,5 +3,5 @@ obj-y := handle.o manage.o spurious.o resend.o chip.o devres.o obj-$(CONFIG_GENERIC_IRQ_PROBE) += autoprobe.o obj-$(CONFIG_PROC_FS) += proc.o obj-$(CONFIG_GENERIC_PENDING_IRQ) += migration.o -obj-$(CONFIG_NUMA_MIGRATE_IRQ_DESC) += numa_migrate.o +obj-$(CONFIG_NUMA_IRQ_DESC) += numa_migrate.o obj-$(CONFIG_PM_SLEEP) += pm.o diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c index c687ba4363f..13c68e71b72 100644 --- a/kernel/irq/chip.c +++ b/kernel/irq/chip.c @@ -359,7 +359,6 @@ handle_level_irq(unsigned int irq, struct irq_desc *desc) spin_lock(&desc->lock); mask_ack_irq(desc, irq); - desc = irq_remap_to_desc(irq, desc); if (unlikely(desc->status & IRQ_INPROGRESS)) goto out_unlock; @@ -438,7 +437,6 @@ handle_fasteoi_irq(unsigned int irq, struct irq_desc *desc) desc->status &= ~IRQ_INPROGRESS; out: desc->chip->eoi(irq); - desc = irq_remap_to_desc(irq, desc); spin_unlock(&desc->lock); } @@ -475,7 +473,6 @@ handle_edge_irq(unsigned int irq, struct irq_desc *desc) !desc->action)) { desc->status |= (IRQ_PENDING | IRQ_MASKED); mask_ack_irq(desc, irq); - desc = irq_remap_to_desc(irq, desc); goto out_unlock; } kstat_incr_irqs_this_cpu(irq, desc); @@ -483,7 +480,6 @@ handle_edge_irq(unsigned int irq, struct irq_desc *desc) /* Start handling the irq */ if (desc->chip->ack) desc->chip->ack(irq); - desc = irq_remap_to_desc(irq, desc); /* Mark the IRQ currently in progress.*/ desc->status |= IRQ_INPROGRESS; @@ -544,10 +540,8 @@ handle_percpu_irq(unsigned int irq, struct irq_desc *desc) if (!noirqdebug) note_interrupt(irq, desc, action_ret); - if (desc->chip->eoi) { + if (desc->chip->eoi) desc->chip->eoi(irq); - desc = irq_remap_to_desc(irq, desc); - } } void @@ -582,10 +576,8 @@ __set_irq_handler(unsigned int irq, irq_flow_handler_t handle, int is_chained, /* Uninstall? */ if (handle == handle_bad_irq) { - if (desc->chip != &no_irq_chip) { + if (desc->chip != &no_irq_chip) mask_ack_irq(desc, irq); - desc = irq_remap_to_desc(irq, desc); - } desc->status |= IRQ_DISABLED; desc->depth = 1; } diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c index 26e08754744..a60018402f4 100644 --- a/kernel/irq/handle.c +++ b/kernel/irq/handle.c @@ -11,14 +11,15 @@ */ #include <linux/irq.h> +#include <linux/slab.h> #include <linux/module.h> #include <linux/random.h> #include <linux/interrupt.h> #include <linux/kernel_stat.h> #include <linux/rculist.h> #include <linux/hash.h> -#include <trace/irq.h> #include <linux/bootmem.h> +#include <trace/events/irq.h> #include "internals.h" @@ -81,45 +82,48 @@ static struct irq_desc irq_desc_init = { .lock = __SPIN_LOCK_UNLOCKED(irq_desc_init.lock), }; -void init_kstat_irqs(struct irq_desc *desc, int cpu, int nr) +void __ref init_kstat_irqs(struct irq_desc *desc, int node, int nr) { - int node; void *ptr; - node = cpu_to_node(cpu); - ptr = kzalloc_node(nr * sizeof(*desc->kstat_irqs), GFP_ATOMIC, node); + if (slab_is_available()) + ptr = kzalloc_node(nr * sizeof(*desc->kstat_irqs), + GFP_ATOMIC, node); + else + ptr = alloc_bootmem_node(NODE_DATA(node), + nr * sizeof(*desc->kstat_irqs)); /* * don't overwite if can not get new one * init_copy_kstat_irqs() could still use old one */ if (ptr) { - printk(KERN_DEBUG " alloc kstat_irqs on cpu %d node %d\n", - cpu, node); + printk(KERN_DEBUG " alloc kstat_irqs on node %d\n", node); desc->kstat_irqs = ptr; } } -static void init_one_irq_desc(int irq, struct irq_desc *desc, int cpu) +static void init_one_irq_desc(int irq, struct irq_desc *desc, int node) { memcpy(desc, &irq_desc_init, sizeof(struct irq_desc)); spin_lock_init(&desc->lock); desc->irq = irq; #ifdef CONFIG_SMP - desc->cpu = cpu; + desc->node = node; #endif lockdep_set_class(&desc->lock, &irq_desc_lock_class); - init_kstat_irqs(desc, cpu, nr_cpu_ids); + init_kstat_irqs(desc, node, nr_cpu_ids); if (!desc->kstat_irqs) { printk(KERN_ERR "can not alloc kstat_irqs\n"); BUG_ON(1); } - if (!init_alloc_desc_masks(desc, cpu, false)) { + if (!alloc_desc_masks(desc, node, false)) { printk(KERN_ERR "can not alloc irq_desc cpumasks\n"); BUG_ON(1); } - arch_init_chip_data(desc, cpu); + init_desc_masks(desc); + arch_init_chip_data(desc, node); } /* @@ -169,7 +173,8 @@ int __init early_irq_init(void) desc[i].irq = i; desc[i].kstat_irqs = kstat_irqs_legacy + i * nr_cpu_ids; lockdep_set_class(&desc[i].lock, &irq_desc_lock_class); - init_alloc_desc_masks(&desc[i], 0, true); + alloc_desc_masks(&desc[i], 0, true); + init_desc_masks(&desc[i]); irq_desc_ptrs[i] = desc + i; } @@ -187,11 +192,10 @@ struct irq_desc *irq_to_desc(unsigned int irq) return NULL; } -struct irq_desc *irq_to_desc_alloc_cpu(unsigned int irq, int cpu) +struct irq_desc * __ref irq_to_desc_alloc_node(unsigned int irq, int node) { struct irq_desc *desc; unsigned long flags; - int node; if (irq >= nr_irqs) { WARN(1, "irq (%d) >= nr_irqs (%d) in irq_to_desc_alloc\n", @@ -210,15 +214,17 @@ struct irq_desc *irq_to_desc_alloc_cpu(unsigned int irq, int cpu) if (desc) goto out_unlock; - node = cpu_to_node(cpu); - desc = kzalloc_node(sizeof(*desc), GFP_ATOMIC, node); - printk(KERN_DEBUG " alloc irq_desc for %d on cpu %d node %d\n", - irq, cpu, node); + if (slab_is_available()) + desc = kzalloc_node(sizeof(*desc), GFP_ATOMIC, node); + else + desc = alloc_bootmem_node(NODE_DATA(node), sizeof(*desc)); + + printk(KERN_DEBUG " alloc irq_desc for %d on node %d\n", irq, node); if (!desc) { printk(KERN_ERR "can not alloc irq_desc\n"); BUG_ON(1); } - init_one_irq_desc(irq, desc, cpu); + init_one_irq_desc(irq, desc, node); irq_desc_ptrs[irq] = desc; @@ -256,7 +262,8 @@ int __init early_irq_init(void) for (i = 0; i < count; i++) { desc[i].irq = i; - init_alloc_desc_masks(&desc[i], 0, true); + alloc_desc_masks(&desc[i], 0, true); + init_desc_masks(&desc[i]); desc[i].kstat_irqs = kstat_irqs_all[i]; } return arch_early_irq_init(); @@ -267,7 +274,7 @@ struct irq_desc *irq_to_desc(unsigned int irq) return (irq < NR_IRQS) ? irq_desc + irq : NULL; } -struct irq_desc *irq_to_desc_alloc_cpu(unsigned int irq, int cpu) +struct irq_desc *irq_to_desc_alloc_node(unsigned int irq, int node) { return irq_to_desc(irq); } @@ -348,9 +355,6 @@ static void warn_no_thread(unsigned int irq, struct irqaction *action) "but no thread function available.", irq, action->name); } -DEFINE_TRACE(irq_handler_entry); -DEFINE_TRACE(irq_handler_exit); - /** * handle_IRQ_event - irq action chain handler * @irq: the interrupt number @@ -453,11 +457,8 @@ unsigned int __do_IRQ(unsigned int irq) /* * No locking required for CPU-local interrupts: */ - if (desc->chip->ack) { + if (desc->chip->ack) desc->chip->ack(irq); - /* get new one */ - desc = irq_remap_to_desc(irq, desc); - } if (likely(!(desc->status & IRQ_DISABLED))) { action_ret = handle_IRQ_event(irq, desc->action); if (!noirqdebug) @@ -468,10 +469,8 @@ unsigned int __do_IRQ(unsigned int irq) } spin_lock(&desc->lock); - if (desc->chip->ack) { + if (desc->chip->ack) desc->chip->ack(irq); - desc = irq_remap_to_desc(irq, desc); - } /* * REPLAY is when Linux resends an IRQ that was dropped earlier * WAITING is used by probe to mark irqs that are being tested diff --git a/kernel/irq/internals.h b/kernel/irq/internals.h index 01ce20eab38..73468253143 100644 --- a/kernel/irq/internals.h +++ b/kernel/irq/internals.h @@ -16,7 +16,7 @@ extern void __disable_irq(struct irq_desc *desc, unsigned int irq, bool susp); extern void __enable_irq(struct irq_desc *desc, unsigned int irq, bool resume); extern struct lock_class_key irq_desc_lock_class; -extern void init_kstat_irqs(struct irq_desc *desc, int cpu, int nr); +extern void init_kstat_irqs(struct irq_desc *desc, int node, int nr); extern void clear_kstat_irqs(struct irq_desc *desc); extern spinlock_t sparse_irq_lock; @@ -42,6 +42,9 @@ static inline void unregister_handler_proc(unsigned int irq, extern int irq_select_affinity_usr(unsigned int irq); +extern void +irq_set_thread_affinity(struct irq_desc *desc, const struct cpumask *cpumask); + /* * Debugging printout: */ diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index 2734eca5924..aaf5c9d0577 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -80,7 +80,7 @@ int irq_can_set_affinity(unsigned int irq) return 1; } -static void +void irq_set_thread_affinity(struct irq_desc *desc, const struct cpumask *cpumask) { struct irqaction *action = desc->action; @@ -109,17 +109,22 @@ int irq_set_affinity(unsigned int irq, const struct cpumask *cpumask) spin_lock_irqsave(&desc->lock, flags); #ifdef CONFIG_GENERIC_PENDING_IRQ - if (desc->status & IRQ_MOVE_PCNTXT) - desc->chip->set_affinity(irq, cpumask); + if (desc->status & IRQ_MOVE_PCNTXT) { + if (!desc->chip->set_affinity(irq, cpumask)) { + cpumask_copy(desc->affinity, cpumask); + irq_set_thread_affinity(desc, cpumask); + } + } else { desc->status |= IRQ_MOVE_PENDING; cpumask_copy(desc->pending_mask, cpumask); } #else - cpumask_copy(desc->affinity, cpumask); - desc->chip->set_affinity(irq, cpumask); + if (!desc->chip->set_affinity(irq, cpumask)) { + cpumask_copy(desc->affinity, cpumask); + irq_set_thread_affinity(desc, cpumask); + } #endif - irq_set_thread_affinity(desc, cpumask); desc->status |= IRQ_AFFINITY_SET; spin_unlock_irqrestore(&desc->lock, flags); return 0; diff --git a/kernel/irq/migration.c b/kernel/irq/migration.c index e05ad9be43b..cfe767ca154 100644 --- a/kernel/irq/migration.c +++ b/kernel/irq/migration.c @@ -1,5 +1,8 @@ #include <linux/irq.h> +#include <linux/interrupt.h> + +#include "internals.h" void move_masked_irq(int irq) { @@ -39,11 +42,12 @@ void move_masked_irq(int irq) * masking the irqs. */ if (likely(cpumask_any_and(desc->pending_mask, cpu_online_mask) - < nr_cpu_ids)) { - cpumask_and(desc->affinity, - desc->pending_mask, cpu_online_mask); - desc->chip->set_affinity(irq, desc->affinity); - } + < nr_cpu_ids)) + if (!desc->chip->set_affinity(irq, desc->pending_mask)) { + cpumask_copy(desc->affinity, desc->pending_mask); + irq_set_thread_affinity(desc, desc->pending_mask); + } + cpumask_clear(desc->pending_mask); } diff --git a/kernel/irq/numa_migrate.c b/kernel/irq/numa_migrate.c index 44bbdcbaf8d..2f69bee57bf 100644 --- a/kernel/irq/numa_migrate.c +++ b/kernel/irq/numa_migrate.c @@ -15,9 +15,9 @@ static void init_copy_kstat_irqs(struct irq_desc *old_desc, struct irq_desc *desc, - int cpu, int nr) + int node, int nr) { - init_kstat_irqs(desc, cpu, nr); + init_kstat_irqs(desc, node, nr); if (desc->kstat_irqs != old_desc->kstat_irqs) memcpy(desc->kstat_irqs, old_desc->kstat_irqs, @@ -34,20 +34,20 @@ static void free_kstat_irqs(struct irq_desc *old_desc, struct irq_desc *desc) } static bool init_copy_one_irq_desc(int irq, struct irq_desc *old_desc, - struct irq_desc *desc, int cpu) + struct irq_desc *desc, int node) { memcpy(desc, old_desc, sizeof(struct irq_desc)); - if (!init_alloc_desc_masks(desc, cpu, false)) { + if (!alloc_desc_masks(desc, node, false)) { printk(KERN_ERR "irq %d: can not get new irq_desc cpumask " "for migration.\n", irq); return false; } spin_lock_init(&desc->lock); - desc->cpu = cpu; + desc->node = node; lockdep_set_class(&desc->lock, &irq_desc_lock_class); - init_copy_kstat_irqs(old_desc, desc, cpu, nr_cpu_ids); + init_copy_kstat_irqs(old_desc, desc, node, nr_cpu_ids); init_copy_desc_masks(old_desc, desc); - arch_init_copy_chip_data(old_desc, desc, cpu); + arch_init_copy_chip_data(old_desc, desc, node); return true; } @@ -59,12 +59,11 @@ static void free_one_irq_desc(struct irq_desc *old_desc, struct irq_desc *desc) } static struct irq_desc *__real_move_irq_desc(struct irq_desc *old_desc, - int cpu) + int node) { struct irq_desc *desc; unsigned int irq; unsigned long flags; - int node; irq = old_desc->irq; @@ -76,7 +75,6 @@ static struct irq_desc *__real_move_irq_desc(struct irq_desc *old_desc, if (desc && old_desc != desc) goto out_unlock; - node = cpu_to_node(cpu); desc = kzalloc_node(sizeof(*desc), GFP_ATOMIC, node); if (!desc) { printk(KERN_ERR "irq %d: can not get new irq_desc " @@ -85,7 +83,7 @@ static struct irq_desc *__real_move_irq_desc(struct irq_desc *old_desc, desc = old_desc; goto out_unlock; } - if (!init_copy_one_irq_desc(irq, old_desc, desc, cpu)) { + if (!init_copy_one_irq_desc(irq, old_desc, desc, node)) { /* still use old one */ kfree(desc); desc = old_desc; @@ -97,9 +95,7 @@ static struct irq_desc *__real_move_irq_desc(struct irq_desc *old_desc, /* free the old one */ free_one_irq_desc(old_desc, desc); - spin_unlock(&old_desc->lock); kfree(old_desc); - spin_lock(&desc->lock); return desc; @@ -109,24 +105,14 @@ out_unlock: return desc; } -struct irq_desc *move_irq_desc(struct irq_desc *desc, int cpu) +struct irq_desc *move_irq_desc(struct irq_desc *desc, int node) { - int old_cpu; - int node, old_node; - /* those all static, do move them */ if (desc->irq < NR_IRQS_LEGACY) return desc; - old_cpu = desc->cpu; - if (old_cpu != cpu) { - node = cpu_to_node(cpu); - old_node = cpu_to_node(old_cpu); - if (old_node != node) - desc = __real_move_irq_desc(desc, cpu); - else - desc->cpu = cpu; - } + if (desc->node != node) + desc = __real_move_irq_desc(desc, node); return desc; } diff --git a/kernel/kthread.c b/kernel/kthread.c index 4ebaf8519ab..41c88fe4050 100644 --- a/kernel/kthread.c +++ b/kernel/kthread.c @@ -13,7 +13,7 @@ #include <linux/file.h> #include <linux/module.h> #include <linux/mutex.h> -#include <trace/sched.h> +#include <trace/events/sched.h> #define KTHREAD_NICE_LEVEL (-5) @@ -21,9 +21,6 @@ static DEFINE_SPINLOCK(kthread_create_lock); static LIST_HEAD(kthread_create_list); struct task_struct *kthreadd_task; -DEFINE_TRACE(sched_kthread_stop); -DEFINE_TRACE(sched_kthread_stop_ret); - struct kthread_create_info { /* Information passed to kthread() from kthreadd. */ diff --git a/kernel/lockdep.c b/kernel/lockdep.c index accb40cdb12..8bbeef996c7 100644 --- a/kernel/lockdep.c +++ b/kernel/lockdep.c @@ -42,12 +42,14 @@ #include <linux/hash.h> #include <linux/ftrace.h> #include <linux/stringify.h> -#include <trace/lockdep.h> #include <asm/sections.h> #include "lockdep_internals.h" +#define CREATE_TRACE_POINTS +#include <trace/events/lockdep.h> + #ifdef CONFIG_PROVE_LOCKING int prove_locking = 1; module_param(prove_locking, int, 0644); @@ -2935,8 +2937,6 @@ void lock_set_class(struct lockdep_map *lock, const char *name, } EXPORT_SYMBOL_GPL(lock_set_class); -DEFINE_TRACE(lock_acquire); - /* * We are not always called with irqs disabled - do that here, * and also avoid lockdep recursion: @@ -2963,8 +2963,6 @@ void lock_acquire(struct lockdep_map *lock, unsigned int subclass, } EXPORT_SYMBOL_GPL(lock_acquire); -DEFINE_TRACE(lock_release); - void lock_release(struct lockdep_map *lock, int nested, unsigned long ip) { @@ -3105,6 +3103,8 @@ found_it: hlock->holdtime_stamp = now; } + trace_lock_acquired(lock, ip, waittime); + stats = get_lock_stats(hlock_class(hlock)); if (waittime) { if (hlock->read) @@ -3120,8 +3120,6 @@ found_it: lock->ip = ip; } -DEFINE_TRACE(lock_contended); - void lock_contended(struct lockdep_map *lock, unsigned long ip) { unsigned long flags; @@ -3143,14 +3141,10 @@ void lock_contended(struct lockdep_map *lock, unsigned long ip) } EXPORT_SYMBOL_GPL(lock_contended); -DEFINE_TRACE(lock_acquired); - void lock_acquired(struct lockdep_map *lock, unsigned long ip) { unsigned long flags; - trace_lock_acquired(lock, ip); - if (unlikely(!lock_stat)) return; diff --git a/kernel/module.c b/kernel/module.c index e797812a4d9..2383e60fcf3 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -18,6 +18,7 @@ */ #include <linux/module.h> #include <linux/moduleloader.h> +#include <linux/ftrace_event.h> #include <linux/init.h> #include <linux/kallsyms.h> #include <linux/fs.h> @@ -1489,9 +1490,6 @@ static void free_module(struct module *mod) /* Free any allocated parameters. */ destroy_params(mod->kp, mod->num_kp); - /* release any pointers to mcount in this module */ - ftrace_release(mod->module_core, mod->core_size); - /* This may be NULL, but that's OK */ module_free(mod, mod->module_init); kfree(mod->args); @@ -1892,11 +1890,9 @@ static noinline struct module *load_module(void __user *umod, unsigned int symindex = 0; unsigned int strindex = 0; unsigned int modindex, versindex, infoindex, pcpuindex; - unsigned int num_mcount; struct module *mod; long err = 0; void *percpu = NULL, *ptr = NULL; /* Stops spurious gcc warning */ - unsigned long *mseg; mm_segment_t old_fs; DEBUGP("load_module: umod=%p, len=%lu, uargs=%p\n", @@ -2172,7 +2168,19 @@ static noinline struct module *load_module(void __user *umod, sizeof(*mod->tracepoints), &mod->num_tracepoints); #endif - +#ifdef CONFIG_EVENT_TRACING + mod->trace_events = section_objs(hdr, sechdrs, secstrings, + "_ftrace_events", + sizeof(*mod->trace_events), + &mod->num_trace_events); +#endif +#ifdef CONFIG_FTRACE_MCOUNT_RECORD + /* sechdrs[0].sh_size is always zero */ + mod->ftrace_callsites = section_objs(hdr, sechdrs, secstrings, + "__mcount_loc", + sizeof(*mod->ftrace_callsites), + &mod->num_ftrace_callsites); +#endif #ifdef CONFIG_MODVERSIONS if ((mod->num_syms && !mod->crcs) || (mod->num_gpl_syms && !mod->gpl_crcs) @@ -2237,11 +2245,6 @@ static noinline struct module *load_module(void __user *umod, dynamic_debug_setup(debug, num_debug); } - /* sechdrs[0].sh_size is always zero */ - mseg = section_objs(hdr, sechdrs, secstrings, "__mcount_loc", - sizeof(*mseg), &num_mcount); - ftrace_init_module(mod, mseg, mseg + num_mcount); - err = module_finalize(hdr, sechdrs, mod); if (err < 0) goto cleanup; @@ -2302,7 +2305,6 @@ static noinline struct module *load_module(void __user *umod, cleanup: kobject_del(&mod->mkobj.kobj); kobject_put(&mod->mkobj.kobj); - ftrace_release(mod->module_core, mod->core_size); free_unload: module_unload_free(mod); #if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP) diff --git a/kernel/mutex.c b/kernel/mutex.c index 507cf2b5e9f..947b3ad551f 100644 --- a/kernel/mutex.c +++ b/kernel/mutex.c @@ -89,7 +89,7 @@ __mutex_lock_slowpath(atomic_t *lock_count); * * This function is similar to (but not equivalent to) down(). */ -void inline __sched mutex_lock(struct mutex *lock) +void __sched mutex_lock(struct mutex *lock) { might_sleep(); /* @@ -249,7 +249,9 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass, /* didnt get the lock, go to sleep: */ spin_unlock_mutex(&lock->wait_lock, flags); - __schedule(); + preempt_enable_no_resched(); + schedule(); + preempt_disable(); spin_lock_mutex(&lock->wait_lock, flags); } @@ -471,5 +473,28 @@ int __sched mutex_trylock(struct mutex *lock) return ret; } - EXPORT_SYMBOL(mutex_trylock); + +/** + * atomic_dec_and_mutex_lock - return holding mutex if we dec to 0 + * @cnt: the atomic which we are to dec + * @lock: the mutex to return holding if we dec to 0 + * + * return true and hold lock if we dec to 0, return false otherwise + */ +int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock) +{ + /* dec if we can't possibly hit 0 */ + if (atomic_add_unless(cnt, -1, 1)) + return 0; + /* we might hit 0, so take the lock */ + mutex_lock(lock); + if (!atomic_dec_and_test(cnt)) { + /* when we actually did the dec, we didn't hit 0 */ + mutex_unlock(lock); + return 0; + } + /* we hit 0, and we hold the lock */ + return 1; +} +EXPORT_SYMBOL(atomic_dec_and_mutex_lock); diff --git a/kernel/perf_counter.c b/kernel/perf_counter.c new file mode 100644 index 00000000000..ef5d8a5b245 --- /dev/null +++ b/kernel/perf_counter.c @@ -0,0 +1,4260 @@ +/* + * Performance counter core code + * + * Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de> + * Copyright (C) 2008-2009 Red Hat, Inc., Ingo Molnar + * Copyright (C) 2008-2009 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com> + * Copyright © 2009 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com> + * + * For licensing details see kernel-base/COPYING + */ + +#include <linux/fs.h> +#include <linux/mm.h> +#include <linux/cpu.h> +#include <linux/smp.h> +#include <linux/file.h> +#include <linux/poll.h> +#include <linux/sysfs.h> +#include <linux/dcache.h> +#include <linux/percpu.h> +#include <linux/ptrace.h> +#include <linux/vmstat.h> +#include <linux/hardirq.h> +#include <linux/rculist.h> +#include <linux/uaccess.h> +#include <linux/syscalls.h> +#include <linux/anon_inodes.h> +#include <linux/kernel_stat.h> +#include <linux/perf_counter.h> + +#include <asm/irq_regs.h> + +/* + * Each CPU has a list of per CPU counters: + */ +DEFINE_PER_CPU(struct perf_cpu_context, perf_cpu_context); + +int perf_max_counters __read_mostly = 1; +static int perf_reserved_percpu __read_mostly; +static int perf_overcommit __read_mostly = 1; + +static atomic_t nr_counters __read_mostly; +static atomic_t nr_mmap_counters __read_mostly; +static atomic_t nr_comm_counters __read_mostly; + +/* + * perf counter paranoia level: + * 0 - not paranoid + * 1 - disallow cpu counters to unpriv + * 2 - disallow kernel profiling to unpriv + */ +int sysctl_perf_counter_paranoid __read_mostly; + +static inline bool perf_paranoid_cpu(void) +{ + return sysctl_perf_counter_paranoid > 0; +} + +static inline bool perf_paranoid_kernel(void) +{ + return sysctl_perf_counter_paranoid > 1; +} + +int sysctl_perf_counter_mlock __read_mostly = 512; /* 'free' kb per user */ + +/* + * max perf counter sample rate + */ +int sysctl_perf_counter_sample_rate __read_mostly = 100000; + +static atomic64_t perf_counter_id; + +/* + * Lock for (sysadmin-configurable) counter reservations: + */ +static DEFINE_SPINLOCK(perf_resource_lock); + +/* + * Architecture provided APIs - weak aliases: + */ +extern __weak const struct pmu *hw_perf_counter_init(struct perf_counter *counter) +{ + return NULL; +} + +void __weak hw_perf_disable(void) { barrier(); } +void __weak hw_perf_enable(void) { barrier(); } + +void __weak hw_perf_counter_setup(int cpu) { barrier(); } + +int __weak +hw_perf_group_sched_in(struct perf_counter *group_leader, + struct perf_cpu_context *cpuctx, + struct perf_counter_context *ctx, int cpu) +{ + return 0; +} + +void __weak perf_counter_print_debug(void) { } + +static DEFINE_PER_CPU(int, disable_count); + +void __perf_disable(void) +{ + __get_cpu_var(disable_count)++; +} + +bool __perf_enable(void) +{ + return !--__get_cpu_var(disable_count); +} + +void perf_disable(void) +{ + __perf_disable(); + hw_perf_disable(); +} + +void perf_enable(void) +{ + if (__perf_enable()) + hw_perf_enable(); +} + +static void get_ctx(struct perf_counter_context *ctx) +{ + atomic_inc(&ctx->refcount); +} + +static void free_ctx(struct rcu_head *head) +{ + struct perf_counter_context *ctx; + + ctx = container_of(head, struct perf_counter_context, rcu_head); + kfree(ctx); +} + +static void put_ctx(struct perf_counter_context *ctx) +{ + if (atomic_dec_and_test(&ctx->refcount)) { + if (ctx->parent_ctx) + put_ctx(ctx->parent_ctx); + if (ctx->task) + put_task_struct(ctx->task); + call_rcu(&ctx->rcu_head, free_ctx); + } +} + +/* + * Get the perf_counter_context for a task and lock it. + * This has to cope with with the fact that until it is locked, + * the context could get moved to another task. + */ +static struct perf_counter_context * +perf_lock_task_context(struct task_struct *task, unsigned long *flags) +{ + struct perf_counter_context *ctx; + + rcu_read_lock(); + retry: + ctx = rcu_dereference(task->perf_counter_ctxp); + if (ctx) { + /* + * If this context is a clone of another, it might + * get swapped for another underneath us by + * perf_counter_task_sched_out, though the + * rcu_read_lock() protects us from any context + * getting freed. Lock the context and check if it + * got swapped before we could get the lock, and retry + * if so. If we locked the right context, then it + * can't get swapped on us any more. + */ + spin_lock_irqsave(&ctx->lock, *flags); + if (ctx != rcu_dereference(task->perf_counter_ctxp)) { + spin_unlock_irqrestore(&ctx->lock, *flags); + goto retry; + } + } + rcu_read_unlock(); + return ctx; +} + +/* + * Get the context for a task and increment its pin_count so it + * can't get swapped to another task. This also increments its + * reference count so that the context can't get freed. + */ +static struct perf_counter_context *perf_pin_task_context(struct task_struct *task) +{ + struct perf_counter_context *ctx; + unsigned long flags; + + ctx = perf_lock_task_context(task, &flags); + if (ctx) { + ++ctx->pin_count; + get_ctx(ctx); + spin_unlock_irqrestore(&ctx->lock, flags); + } + return ctx; +} + +static void perf_unpin_context(struct perf_counter_context *ctx) +{ + unsigned long flags; + + spin_lock_irqsave(&ctx->lock, flags); + --ctx->pin_count; + spin_unlock_irqrestore(&ctx->lock, flags); + put_ctx(ctx); +} + +/* + * Add a counter from the lists for its context. + * Must be called with ctx->mutex and ctx->lock held. + */ +static void +list_add_counter(struct perf_counter *counter, struct perf_counter_context *ctx) +{ + struct perf_counter *group_leader = counter->group_leader; + + /* + * Depending on whether it is a standalone or sibling counter, + * add it straight to the context's counter list, or to the group + * leader's sibling list: + */ + if (group_leader == counter) + list_add_tail(&counter->list_entry, &ctx->counter_list); + else { + list_add_tail(&counter->list_entry, &group_leader->sibling_list); + group_leader->nr_siblings++; + } + + list_add_rcu(&counter->event_entry, &ctx->event_list); + ctx->nr_counters++; +} + +/* + * Remove a counter from the lists for its context. + * Must be called with ctx->mutex and ctx->lock held. + */ +static void +list_del_counter(struct perf_counter *counter, struct perf_counter_context *ctx) +{ + struct perf_counter *sibling, *tmp; + + if (list_empty(&counter->list_entry)) + return; + ctx->nr_counters--; + + list_del_init(&counter->list_entry); + list_del_rcu(&counter->event_entry); + + if (counter->group_leader != counter) + counter->group_leader->nr_siblings--; + + /* + * If this was a group counter with sibling counters then + * upgrade the siblings to singleton counters by adding them + * to the context list directly: + */ + list_for_each_entry_safe(sibling, tmp, + &counter->sibling_list, list_entry) { + + list_move_tail(&sibling->list_entry, &ctx->counter_list); + sibling->group_leader = sibling; + } +} + +static void +counter_sched_out(struct perf_counter *counter, + struct perf_cpu_context *cpuctx, + struct perf_counter_context *ctx) +{ + if (counter->state != PERF_COUNTER_STATE_ACTIVE) + return; + + counter->state = PERF_COUNTER_STATE_INACTIVE; + counter->tstamp_stopped = ctx->time; + counter->pmu->disable(counter); + counter->oncpu = -1; + + if (!is_software_counter(counter)) + cpuctx->active_oncpu--; + ctx->nr_active--; + if (counter->attr.exclusive || !cpuctx->active_oncpu) + cpuctx->exclusive = 0; +} + +static void +group_sched_out(struct perf_counter *group_counter, + struct perf_cpu_context *cpuctx, + struct perf_counter_context *ctx) +{ + struct perf_counter *counter; + + if (group_counter->state != PERF_COUNTER_STATE_ACTIVE) + return; + + counter_sched_out(group_counter, cpuctx, ctx); + + /* + * Schedule out siblings (if any): + */ + list_for_each_entry(counter, &group_counter->sibling_list, list_entry) + counter_sched_out(counter, cpuctx, ctx); + + if (group_counter->attr.exclusive) + cpuctx->exclusive = 0; +} + +/* + * Cross CPU call to remove a performance counter + * + * We disable the counter on the hardware level first. After that we + * remove it from the context list. + */ +static void __perf_counter_remove_from_context(void *info) +{ + struct perf_cpu_context *cpuctx = &__get_cpu_var(perf_cpu_context); + struct perf_counter *counter = info; + struct perf_counter_context *ctx = counter->ctx; + + /* + * If this is a task context, we need to check whether it is + * the current task context of this cpu. If not it has been + * scheduled out before the smp call arrived. + */ + if (ctx->task && cpuctx->task_ctx != ctx) + return; + + spin_lock(&ctx->lock); + /* + * Protect the list operation against NMI by disabling the + * counters on a global level. + */ + perf_disable(); + + counter_sched_out(counter, cpuctx, ctx); + + list_del_counter(counter, ctx); + + if (!ctx->task) { + /* + * Allow more per task counters with respect to the + * reservation: + */ + cpuctx->max_pertask = + min(perf_max_counters - ctx->nr_counters, + perf_max_counters - perf_reserved_percpu); + } + + perf_enable(); + spin_unlock(&ctx->lock); +} + + +/* + * Remove the counter from a task's (or a CPU's) list of counters. + * + * Must be called with ctx->mutex held. + * + * CPU counters are removed with a smp call. For task counters we only + * call when the task is on a CPU. + * + * If counter->ctx is a cloned context, callers must make sure that + * every task struct that counter->ctx->task could possibly point to + * remains valid. This is OK when called from perf_release since + * that only calls us on the top-level context, which can't be a clone. + * When called from perf_counter_exit_task, it's OK because the + * context has been detached from its task. + */ +static void perf_counter_remove_from_context(struct perf_counter *counter) +{ + struct perf_counter_context *ctx = counter->ctx; + struct task_struct *task = ctx->task; + + if (!task) { + /* + * Per cpu counters are removed via an smp call and + * the removal is always sucessful. + */ + smp_call_function_single(counter->cpu, + __perf_counter_remove_from_context, + counter, 1); + return; + } + +retry: + task_oncpu_function_call(task, __perf_counter_remove_from_context, + counter); + + spin_lock_irq(&ctx->lock); + /* + * If the context is active we need to retry the smp call. + */ + if (ctx->nr_active && !list_empty(&counter->list_entry)) { + spin_unlock_irq(&ctx->lock); + goto retry; + } + + /* + * The lock prevents that this context is scheduled in so we + * can remove the counter safely, if the call above did not + * succeed. + */ + if (!list_empty(&counter->list_entry)) { + list_del_counter(counter, ctx); + } + spin_unlock_irq(&ctx->lock); +} + +static inline u64 perf_clock(void) +{ + return cpu_clock(smp_processor_id()); +} + +/* + * Update the record of the current time in a context. + */ +static void update_context_time(struct perf_counter_context *ctx) +{ + u64 now = perf_clock(); + + ctx->time += now - ctx->timestamp; + ctx->timestamp = now; +} + +/* + * Update the total_time_enabled and total_time_running fields for a counter. + */ +static void update_counter_times(struct perf_counter *counter) +{ + struct perf_counter_context *ctx = counter->ctx; + u64 run_end; + + if (counter->state < PERF_COUNTER_STATE_INACTIVE) + return; + + counter->total_time_enabled = ctx->time - counter->tstamp_enabled; + + if (counter->state == PERF_COUNTER_STATE_INACTIVE) + run_end = counter->tstamp_stopped; + else + run_end = ctx->time; + + counter->total_time_running = run_end - counter->tstamp_running; +} + +/* + * Update total_time_enabled and total_time_running for all counters in a group. + */ +static void update_group_times(struct perf_counter *leader) +{ + struct perf_counter *counter; + + update_counter_times(leader); + list_for_each_entry(counter, &leader->sibling_list, list_entry) + update_counter_times(counter); +} + +/* + * Cross CPU call to disable a performance counter + */ +static void __perf_counter_disable(void *info) +{ + struct perf_counter *counter = info; + struct perf_cpu_context *cpuctx = &__get_cpu_var(perf_cpu_context); + struct perf_counter_context *ctx = counter->ctx; + + /* + * If this is a per-task counter, need to check whether this + * counter's task is the current task on this cpu. + */ + if (ctx->task && cpuctx->task_ctx != ctx) + return; + + spin_lock(&ctx->lock); + + /* + * If the counter is on, turn it off. + * If it is in error state, leave it in error state. + */ + if (counter->state >= PERF_COUNTER_STATE_INACTIVE) { + update_context_time(ctx); + update_counter_times(counter); + if (counter == counter->group_leader) + group_sched_out(counter, cpuctx, ctx); + else + counter_sched_out(counter, cpuctx, ctx); + counter->state = PERF_COUNTER_STATE_OFF; + } + + spin_unlock(&ctx->lock); +} + +/* + * Disable a counter. + * + * If counter->ctx is a cloned context, callers must make sure that + * every task struct that counter->ctx->task could possibly point to + * remains valid. This condition is satisifed when called through + * perf_counter_for_each_child or perf_counter_for_each because they + * hold the top-level counter's child_mutex, so any descendant that + * goes to exit will block in sync_child_counter. + * When called from perf_pending_counter it's OK because counter->ctx + * is the current context on this CPU and preemption is disabled, + * hence we can't get into perf_counter_task_sched_out for this context. + */ +static void perf_counter_disable(struct perf_counter *counter) +{ + struct perf_counter_context *ctx = counter->ctx; + struct task_struct *task = ctx->task; + + if (!task) { + /* + * Disable the counter on the cpu that it's on + */ + smp_call_function_single(counter->cpu, __perf_counter_disable, + counter, 1); + return; + } + + retry: + task_oncpu_function_call(task, __perf_counter_disable, counter); + + spin_lock_irq(&ctx->lock); + /* + * If the counter is still active, we need to retry the cross-call. + */ + if (counter->state == PERF_COUNTER_STATE_ACTIVE) { + spin_unlock_irq(&ctx->lock); + goto retry; + } + + /* + * Since we have the lock this context can't be scheduled + * in, so we can change the state safely. + */ + if (counter->state == PERF_COUNTER_STATE_INACTIVE) { + update_counter_times(counter); + counter->state = PERF_COUNTER_STATE_OFF; + } + + spin_unlock_irq(&ctx->lock); +} + +static int +counter_sched_in(struct perf_counter *counter, + struct perf_cpu_context *cpuctx, + struct perf_counter_context *ctx, + int cpu) +{ + if (counter->state <= PERF_COUNTER_STATE_OFF) + return 0; + + counter->state = PERF_COUNTER_STATE_ACTIVE; + counter->oncpu = cpu; /* TODO: put 'cpu' into cpuctx->cpu */ + /* + * The new state must be visible before we turn it on in the hardware: + */ + smp_wmb(); + + if (counter->pmu->enable(counter)) { + counter->state = PERF_COUNTER_STATE_INACTIVE; + counter->oncpu = -1; + return -EAGAIN; + } + + counter->tstamp_running += ctx->time - counter->tstamp_stopped; + + if (!is_software_counter(counter)) + cpuctx->active_oncpu++; + ctx->nr_active++; + + if (counter->attr.exclusive) + cpuctx->exclusive = 1; + + return 0; +} + +static int +group_sched_in(struct perf_counter *group_counter, + struct perf_cpu_context *cpuctx, + struct perf_counter_context *ctx, + int cpu) +{ + struct perf_counter *counter, *partial_group; + int ret; + + if (group_counter->state == PERF_COUNTER_STATE_OFF) + return 0; + + ret = hw_perf_group_sched_in(group_counter, cpuctx, ctx, cpu); + if (ret) + return ret < 0 ? ret : 0; + + if (counter_sched_in(group_counter, cpuctx, ctx, cpu)) + return -EAGAIN; + + /* + * Schedule in siblings as one group (if any): + */ + list_for_each_entry(counter, &group_counter->sibling_list, list_entry) { + if (counter_sched_in(counter, cpuctx, ctx, cpu)) { + partial_group = counter; + goto group_error; + } + } + + return 0; + +group_error: + /* + * Groups can be scheduled in as one unit only, so undo any + * partial group before returning: + */ + list_for_each_entry(counter, &group_counter->sibling_list, list_entry) { + if (counter == partial_group) + break; + counter_sched_out(counter, cpuctx, ctx); + } + counter_sched_out(group_counter, cpuctx, ctx); + + return -EAGAIN; +} + +/* + * Return 1 for a group consisting entirely of software counters, + * 0 if the group contains any hardware counters. + */ +static int is_software_only_group(struct perf_counter *leader) +{ + struct perf_counter *counter; + + if (!is_software_counter(leader)) + return 0; + + list_for_each_entry(counter, &leader->sibling_list, list_entry) + if (!is_software_counter(counter)) + return 0; + + return 1; +} + +/* + * Work out whether we can put this counter group on the CPU now. + */ +static int group_can_go_on(struct perf_counter *counter, + struct perf_cpu_context *cpuctx, + int can_add_hw) +{ + /* + * Groups consisting entirely of software counters can always go on. + */ + if (is_software_only_group(counter)) + return 1; + /* + * If an exclusive group is already on, no other hardware + * counters can go on. + */ + if (cpuctx->exclusive) + return 0; + /* + * If this group is exclusive and there are already + * counters on the CPU, it can't go on. + */ + if (counter->attr.exclusive && cpuctx->active_oncpu) + return 0; + /* + * Otherwise, try to add it if all previous groups were able + * to go on. + */ + return can_add_hw; +} + +static void add_counter_to_ctx(struct perf_counter *counter, + struct perf_counter_context *ctx) +{ + list_add_counter(counter, ctx); + counter->tstamp_enabled = ctx->time; + counter->tstamp_running = ctx->time; + counter->tstamp_stopped = ctx->time; +} + +/* + * Cross CPU call to install and enable a performance counter + * + * Must be called with ctx->mutex held + */ +static void __perf_install_in_context(void *info) +{ + struct perf_cpu_context *cpuctx = &__get_cpu_var(perf_cpu_context); + struct perf_counter *counter = info; + struct perf_counter_context *ctx = counter->ctx; + struct perf_counter *leader = counter->group_leader; + int cpu = smp_processor_id(); + int err; + + /* + * If this is a task context, we need to check whether it is + * the current task context of this cpu. If not it has been + * scheduled out before the smp call arrived. + * Or possibly this is the right context but it isn't + * on this cpu because it had no counters. + */ + if (ctx->task && cpuctx->task_ctx != ctx) { + if (cpuctx->task_ctx || ctx->task != current) + return; + cpuctx->task_ctx = ctx; + } + + spin_lock(&ctx->lock); + ctx->is_active = 1; + update_context_time(ctx); + + /* + * Protect the list operation against NMI by disabling the + * counters on a global level. NOP for non NMI based counters. + */ + perf_disable(); + + add_counter_to_ctx(counter, ctx); + + /* + * Don't put the counter on if it is disabled or if + * it is in a group and the group isn't on. + */ + if (counter->state != PERF_COUNTER_STATE_INACTIVE || + (leader != counter && leader->state != PERF_COUNTER_STATE_ACTIVE)) + goto unlock; + + /* + * An exclusive counter can't go on if there are already active + * hardware counters, and no hardware counter can go on if there + * is already an exclusive counter on. + */ + if (!group_can_go_on(counter, cpuctx, 1)) + err = -EEXIST; + else + err = counter_sched_in(counter, cpuctx, ctx, cpu); + + if (err) { + /* + * This counter couldn't go on. If it is in a group + * then we have to pull the whole group off. + * If the counter group is pinned then put it in error state. + */ + if (leader != counter) + group_sched_out(leader, cpuctx, ctx); + if (leader->attr.pinned) { + update_group_times(leader); + leader->state = PERF_COUNTER_STATE_ERROR; + } + } + + if (!err && !ctx->task && cpuctx->max_pertask) + cpuctx->max_pertask--; + + unlock: + perf_enable(); + + spin_unlock(&ctx->lock); +} + +/* + * Attach a performance counter to a context + * + * First we add the counter to the list with the hardware enable bit + * in counter->hw_config cleared. + * + * If the counter is attached to a task which is on a CPU we use a smp + * call to enable it in the task context. The task might have been + * scheduled away, but we check this in the smp call again. + * + * Must be called with ctx->mutex held. + */ +static void +perf_install_in_context(struct perf_counter_context *ctx, + struct perf_counter *counter, + int cpu) +{ + struct task_struct *task = ctx->task; + + if (!task) { + /* + * Per cpu counters are installed via an smp call and + * the install is always sucessful. + */ + smp_call_function_single(cpu, __perf_install_in_context, + counter, 1); + return; + } + +retry: + task_oncpu_function_call(task, __perf_install_in_context, + counter); + + spin_lock_irq(&ctx->lock); + /* + * we need to retry the smp call. + */ + if (ctx->is_active && list_empty(&counter->list_entry)) { + spin_unlock_irq(&ctx->lock); + goto retry; + } + + /* + * The lock prevents that this context is scheduled in so we + * can add the counter safely, if it the call above did not + * succeed. + */ + if (list_empty(&counter->list_entry)) + add_counter_to_ctx(counter, ctx); + spin_unlock_irq(&ctx->lock); +} + +/* + * Cross CPU call to enable a performance counter + */ +static void __perf_counter_enable(void *info) +{ + struct perf_counter *counter = info; + struct perf_cpu_context *cpuctx = &__get_cpu_var(perf_cpu_context); + struct perf_counter_context *ctx = counter->ctx; + struct perf_counter *leader = counter->group_leader; + int err; + + /* + * If this is a per-task counter, need to check whether this + * counter's task is the current task on this cpu. + */ + if (ctx->task && cpuctx->task_ctx != ctx) { + if (cpuctx->task_ctx || ctx->task != current) + return; + cpuctx->task_ctx = ctx; + } + + spin_lock(&ctx->lock); + ctx->is_active = 1; + update_context_time(ctx); + + if (counter->state >= PERF_COUNTER_STATE_INACTIVE) + goto unlock; + counter->state = PERF_COUNTER_STATE_INACTIVE; + counter->tstamp_enabled = ctx->time - counter->total_time_enabled; + + /* + * If the counter is in a group and isn't the group leader, + * then don't put it on unless the group is on. + */ + if (leader != counter && leader->state != PERF_COUNTER_STATE_ACTIVE) + goto unlock; + + if (!group_can_go_on(counter, cpuctx, 1)) { + err = -EEXIST; + } else { + perf_disable(); + if (counter == leader) + err = group_sched_in(counter, cpuctx, ctx, + smp_processor_id()); + else + err = counter_sched_in(counter, cpuctx, ctx, + smp_processor_id()); + perf_enable(); + } + + if (err) { + /* + * If this counter can't go on and it's part of a + * group, then the whole group has to come off. + */ + if (leader != counter) + group_sched_out(leader, cpuctx, ctx); + if (leader->attr.pinned) { + update_group_times(leader); + leader->state = PERF_COUNTER_STATE_ERROR; + } + } + + unlock: + spin_unlock(&ctx->lock); +} + +/* + * Enable a counter. + * + * If counter->ctx is a cloned context, callers must make sure that + * every task struct that counter->ctx->task could possibly point to + * remains valid. This condition is satisfied when called through + * perf_counter_for_each_child or perf_counter_for_each as described + * for perf_counter_disable. + */ +static void perf_counter_enable(struct perf_counter *counter) +{ + struct perf_counter_context *ctx = counter->ctx; + struct task_struct *task = ctx->task; + + if (!task) { + /* + * Enable the counter on the cpu that it's on + */ + smp_call_function_single(counter->cpu, __perf_counter_enable, + counter, 1); + return; + } + + spin_lock_irq(&ctx->lock); + if (counter->state >= PERF_COUNTER_STATE_INACTIVE) + goto out; + + /* + * If the counter is in error state, clear that first. + * That way, if we see the counter in error state below, we + * know that it has gone back into error state, as distinct + * from the task having been scheduled away before the + * cross-call arrived. + */ + if (counter->state == PERF_COUNTER_STATE_ERROR) + counter->state = PERF_COUNTER_STATE_OFF; + + retry: + spin_unlock_irq(&ctx->lock); + task_oncpu_function_call(task, __perf_counter_enable, counter); + + spin_lock_irq(&ctx->lock); + + /* + * If the context is active and the counter is still off, + * we need to retry the cross-call. + */ + if (ctx->is_active && counter->state == PERF_COUNTER_STATE_OFF) + goto retry; + + /* + * Since we have the lock this context can't be scheduled + * in, so we can change the state safely. + */ + if (counter->state == PERF_COUNTER_STATE_OFF) { + counter->state = PERF_COUNTER_STATE_INACTIVE; + counter->tstamp_enabled = + ctx->time - counter->total_time_enabled; + } + out: + spin_unlock_irq(&ctx->lock); +} + +static int perf_counter_refresh(struct perf_counter *counter, int refresh) +{ + /* + * not supported on inherited counters + */ + if (counter->attr.inherit) + return -EINVAL; + + atomic_add(refresh, &counter->event_limit); + perf_counter_enable(counter); + + return 0; +} + +void __perf_counter_sched_out(struct perf_counter_context *ctx, + struct perf_cpu_context *cpuctx) +{ + struct perf_counter *counter; + + spin_lock(&ctx->lock); + ctx->is_active = 0; + if (likely(!ctx->nr_counters)) + goto out; + update_context_time(ctx); + + perf_disable(); + if (ctx->nr_active) { + list_for_each_entry(counter, &ctx->counter_list, list_entry) { + if (counter != counter->group_leader) + counter_sched_out(counter, cpuctx, ctx); + else + group_sched_out(counter, cpuctx, ctx); + } + } + perf_enable(); + out: + spin_unlock(&ctx->lock); +} + +/* + * Test whether two contexts are equivalent, i.e. whether they + * have both been cloned from the same version of the same context + * and they both have the same number of enabled counters. + * If the number of enabled counters is the same, then the set + * of enabled counters should be the same, because these are both + * inherited contexts, therefore we can't access individual counters + * in them directly with an fd; we can only enable/disable all + * counters via prctl, or enable/disable all counters in a family + * via ioctl, which will have the same effect on both contexts. + */ +static int context_equiv(struct perf_counter_context *ctx1, + struct perf_counter_context *ctx2) +{ + return ctx1->parent_ctx && ctx1->parent_ctx == ctx2->parent_ctx + && ctx1->parent_gen == ctx2->parent_gen + && !ctx1->pin_count && !ctx2->pin_count; +} + +/* + * Called from scheduler to remove the counters of the current task, + * with interrupts disabled. + * + * We stop each counter and update the counter value in counter->count. + * + * This does not protect us against NMI, but disable() + * sets the disabled bit in the control field of counter _before_ + * accessing the counter control register. If a NMI hits, then it will + * not restart the counter. + */ +void perf_counter_task_sched_out(struct task_struct *task, + struct task_struct *next, int cpu) +{ + struct perf_cpu_context *cpuctx = &per_cpu(perf_cpu_context, cpu); + struct perf_counter_context *ctx = task->perf_counter_ctxp; + struct perf_counter_context *next_ctx; + struct perf_counter_context *parent; + struct pt_regs *regs; + int do_switch = 1; + + regs = task_pt_regs(task); + perf_swcounter_event(PERF_COUNT_SW_CONTEXT_SWITCHES, 1, 1, regs, 0); + + if (likely(!ctx || !cpuctx->task_ctx)) + return; + + update_context_time(ctx); + + rcu_read_lock(); + parent = rcu_dereference(ctx->parent_ctx); + next_ctx = next->perf_counter_ctxp; + if (parent && next_ctx && + rcu_dereference(next_ctx->parent_ctx) == parent) { + /* + * Looks like the two contexts are clones, so we might be + * able to optimize the context switch. We lock both + * contexts and check that they are clones under the + * lock (including re-checking that neither has been + * uncloned in the meantime). It doesn't matter which + * order we take the locks because no other cpu could + * be trying to lock both of these tasks. + */ + spin_lock(&ctx->lock); + spin_lock_nested(&next_ctx->lock, SINGLE_DEPTH_NESTING); + if (context_equiv(ctx, next_ctx)) { + /* + * XXX do we need a memory barrier of sorts + * wrt to rcu_dereference() of perf_counter_ctxp + */ + task->perf_counter_ctxp = next_ctx; + next->perf_counter_ctxp = ctx; + ctx->task = next; + next_ctx->task = task; + do_switch = 0; + } + spin_unlock(&next_ctx->lock); + spin_unlock(&ctx->lock); + } + rcu_read_unlock(); + + if (do_switch) { + __perf_counter_sched_out(ctx, cpuctx); + cpuctx->task_ctx = NULL; + } +} + +/* + * Called with IRQs disabled + */ +static void __perf_counter_task_sched_out(struct perf_counter_context *ctx) +{ + struct perf_cpu_context *cpuctx = &__get_cpu_var(perf_cpu_context); + + if (!cpuctx->task_ctx) + return; + + if (WARN_ON_ONCE(ctx != cpuctx->task_ctx)) + return; + + __perf_counter_sched_out(ctx, cpuctx); + cpuctx->task_ctx = NULL; +} + +/* + * Called with IRQs disabled + */ +static void perf_counter_cpu_sched_out(struct perf_cpu_context *cpuctx) +{ + __perf_counter_sched_out(&cpuctx->ctx, cpuctx); +} + +static void +__perf_counter_sched_in(struct perf_counter_context *ctx, + struct perf_cpu_context *cpuctx, int cpu) +{ + struct perf_counter *counter; + int can_add_hw = 1; + + spin_lock(&ctx->lock); + ctx->is_active = 1; + if (likely(!ctx->nr_counters)) + goto out; + + ctx->timestamp = perf_clock(); + + perf_disable(); + + /* + * First go through the list and put on any pinned groups + * in order to give them the best chance of going on. + */ + list_for_each_entry(counter, &ctx->counter_list, list_entry) { + if (counter->state <= PERF_COUNTER_STATE_OFF || + !counter->attr.pinned) + continue; + if (counter->cpu != -1 && counter->cpu != cpu) + continue; + + if (counter != counter->group_leader) + counter_sched_in(counter, cpuctx, ctx, cpu); + else { + if (group_can_go_on(counter, cpuctx, 1)) + group_sched_in(counter, cpuctx, ctx, cpu); + } + + /* + * If this pinned group hasn't been scheduled, + * put it in error state. + */ + if (counter->state == PERF_COUNTER_STATE_INACTIVE) { + update_group_times(counter); + counter->state = PERF_COUNTER_STATE_ERROR; + } + } + + list_for_each_entry(counter, &ctx->counter_list, list_entry) { + /* + * Ignore counters in OFF or ERROR state, and + * ignore pinned counters since we did them already. + */ + if (counter->state <= PERF_COUNTER_STATE_OFF || + counter->attr.pinned) + continue; + + /* + * Listen to the 'cpu' scheduling filter constraint + * of counters: + */ + if (counter->cpu != -1 && counter->cpu != cpu) + continue; + + if (counter != counter->group_leader) { + if (counter_sched_in(counter, cpuctx, ctx, cpu)) + can_add_hw = 0; + } else { + if (group_can_go_on(counter, cpuctx, can_add_hw)) { + if (group_sched_in(counter, cpuctx, ctx, cpu)) + can_add_hw = 0; + } + } + } + perf_enable(); + out: + spin_unlock(&ctx->lock); +} + +/* + * Called from scheduler to add the counters of the current task + * with interrupts disabled. + * + * We restore the counter value and then enable it. + * + * This does not protect us against NMI, but enable() + * sets the enabled bit in the control field of counter _before_ + * accessing the counter control register. If a NMI hits, then it will + * keep the counter running. + */ +void perf_counter_task_sched_in(struct task_struct *task, int cpu) +{ + struct perf_cpu_context *cpuctx = &per_cpu(perf_cpu_context, cpu); + struct perf_counter_context *ctx = task->perf_counter_ctxp; + + if (likely(!ctx)) + return; + if (cpuctx->task_ctx == ctx) + return; + __perf_counter_sched_in(ctx, cpuctx, cpu); + cpuctx->task_ctx = ctx; +} + +static void perf_counter_cpu_sched_in(struct perf_cpu_context *cpuctx, int cpu) +{ + struct perf_counter_context *ctx = &cpuctx->ctx; + + __perf_counter_sched_in(ctx, cpuctx, cpu); +} + +#define MAX_INTERRUPTS (~0ULL) + +static void perf_log_throttle(struct perf_counter *counter, int enable); +static void perf_log_period(struct perf_counter *counter, u64 period); + +static void perf_adjust_period(struct perf_counter *counter, u64 events) +{ + struct hw_perf_counter *hwc = &counter->hw; + u64 period, sample_period; + s64 delta; + + events *= hwc->sample_period; + period = div64_u64(events, counter->attr.sample_freq); + + delta = (s64)(period - hwc->sample_period); + delta = (delta + 7) / 8; /* low pass filter */ + + sample_period = hwc->sample_period + delta; + + if (!sample_period) + sample_period = 1; + + perf_log_period(counter, sample_period); + + hwc->sample_period = sample_period; +} + +static void perf_ctx_adjust_freq(struct perf_counter_context *ctx) +{ + struct perf_counter *counter; + struct hw_perf_counter *hwc; + u64 interrupts, freq; + + spin_lock(&ctx->lock); + list_for_each_entry(counter, &ctx->counter_list, list_entry) { + if (counter->state != PERF_COUNTER_STATE_ACTIVE) + continue; + + hwc = &counter->hw; + + interrupts = hwc->interrupts; + hwc->interrupts = 0; + + /* + * unthrottle counters on the tick + */ + if (interrupts == MAX_INTERRUPTS) { + perf_log_throttle(counter, 1); + counter->pmu->unthrottle(counter); + interrupts = 2*sysctl_perf_counter_sample_rate/HZ; + } + + if (!counter->attr.freq || !counter->attr.sample_freq) + continue; + + /* + * if the specified freq < HZ then we need to skip ticks + */ + if (counter->attr.sample_freq < HZ) { + freq = counter->attr.sample_freq; + + hwc->freq_count += freq; + hwc->freq_interrupts += interrupts; + + if (hwc->freq_count < HZ) + continue; + + interrupts = hwc->freq_interrupts; + hwc->freq_interrupts = 0; + hwc->freq_count -= HZ; + } else + freq = HZ; + + perf_adjust_period(counter, freq * interrupts); + + /* + * In order to avoid being stalled by an (accidental) huge + * sample period, force reset the sample period if we didn't + * get any events in this freq period. + */ + if (!interrupts) { + perf_disable(); + counter->pmu->disable(counter); + atomic_set(&hwc->period_left, 0); + counter->pmu->enable(counter); + perf_enable(); + } + } + spin_unlock(&ctx->lock); +} + +/* + * Round-robin a context's counters: + */ +static void rotate_ctx(struct perf_counter_context *ctx) +{ + struct perf_counter *counter; + + if (!ctx->nr_counters) + return; + + spin_lock(&ctx->lock); + /* + * Rotate the first entry last (works just fine for group counters too): + */ + perf_disable(); + list_for_each_entry(counter, &ctx->counter_list, list_entry) { + list_move_tail(&counter->list_entry, &ctx->counter_list); + break; + } + perf_enable(); + + spin_unlock(&ctx->lock); +} + +void perf_counter_task_tick(struct task_struct *curr, int cpu) +{ + struct perf_cpu_context *cpuctx; + struct perf_counter_context *ctx; + + if (!atomic_read(&nr_counters)) + return; + + cpuctx = &per_cpu(perf_cpu_context, cpu); + ctx = curr->perf_counter_ctxp; + + perf_ctx_adjust_freq(&cpuctx->ctx); + if (ctx) + perf_ctx_adjust_freq(ctx); + + perf_counter_cpu_sched_out(cpuctx); + if (ctx) + __perf_counter_task_sched_out(ctx); + + rotate_ctx(&cpuctx->ctx); + if (ctx) + rotate_ctx(ctx); + + perf_counter_cpu_sched_in(cpuctx, cpu); + if (ctx) + perf_counter_task_sched_in(curr, cpu); +} + +/* + * Cross CPU call to read the hardware counter + */ +static void __read(void *info) +{ + struct perf_counter *counter = info; + struct perf_counter_context *ctx = counter->ctx; + unsigned long flags; + + local_irq_save(flags); + if (ctx->is_active) + update_context_time(ctx); + counter->pmu->read(counter); + update_counter_times(counter); + local_irq_restore(flags); +} + +static u64 perf_counter_read(struct perf_counter *counter) +{ + /* + * If counter is enabled and currently active on a CPU, update the + * value in the counter structure: + */ + if (counter->state == PERF_COUNTER_STATE_ACTIVE) { + smp_call_function_single(counter->oncpu, + __read, counter, 1); + } else if (counter->state == PERF_COUNTER_STATE_INACTIVE) { + update_counter_times(counter); + } + + return atomic64_read(&counter->count); +} + +/* + * Initialize the perf_counter context in a task_struct: + */ +static void +__perf_counter_init_context(struct perf_counter_context *ctx, + struct task_struct *task) +{ + memset(ctx, 0, sizeof(*ctx)); + spin_lock_init(&ctx->lock); + mutex_init(&ctx->mutex); + INIT_LIST_HEAD(&ctx->counter_list); + INIT_LIST_HEAD(&ctx->event_list); + atomic_set(&ctx->refcount, 1); + ctx->task = task; +} + +static struct perf_counter_context *find_get_context(pid_t pid, int cpu) +{ + struct perf_counter_context *parent_ctx; + struct perf_counter_context *ctx; + struct perf_cpu_context *cpuctx; + struct task_struct *task; + unsigned long flags; + int err; + + /* + * If cpu is not a wildcard then this is a percpu counter: + */ + if (cpu != -1) { + /* Must be root to operate on a CPU counter: */ + if (perf_paranoid_cpu() && !capable(CAP_SYS_ADMIN)) + return ERR_PTR(-EACCES); + + if (cpu < 0 || cpu > num_possible_cpus()) + return ERR_PTR(-EINVAL); + + /* + * We could be clever and allow to attach a counter to an + * offline CPU and activate it when the CPU comes up, but + * that's for later. + */ + if (!cpu_isset(cpu, cpu_online_map)) + return ERR_PTR(-ENODEV); + + cpuctx = &per_cpu(perf_cpu_context, cpu); + ctx = &cpuctx->ctx; + get_ctx(ctx); + + return ctx; + } + + rcu_read_lock(); + if (!pid) + task = current; + else + task = find_task_by_vpid(pid); + if (task) + get_task_struct(task); + rcu_read_unlock(); + + if (!task) + return ERR_PTR(-ESRCH); + + /* + * Can't attach counters to a dying task. + */ + err = -ESRCH; + if (task->flags & PF_EXITING) + goto errout; + + /* Reuse ptrace permission checks for now. */ + err = -EACCES; + if (!ptrace_may_access(task, PTRACE_MODE_READ)) + goto errout; + + retry: + ctx = perf_lock_task_context(task, &flags); + if (ctx) { + parent_ctx = ctx->parent_ctx; + if (parent_ctx) { + put_ctx(parent_ctx); + ctx->parent_ctx = NULL; /* no longer a clone */ + } + /* + * Get an extra reference before dropping the lock so that + * this context won't get freed if the task exits. + */ + get_ctx(ctx); + spin_unlock_irqrestore(&ctx->lock, flags); + } + + if (!ctx) { + ctx = kmalloc(sizeof(struct perf_counter_context), GFP_KERNEL); + err = -ENOMEM; + if (!ctx) + goto errout; + __perf_counter_init_context(ctx, task); + get_ctx(ctx); + if (cmpxchg(&task->perf_counter_ctxp, NULL, ctx)) { + /* + * We raced with some other task; use + * the context they set. + */ + kfree(ctx); + goto retry; + } + get_task_struct(task); + } + + put_task_struct(task); + return ctx; + + errout: + put_task_struct(task); + return ERR_PTR(err); +} + +static void free_counter_rcu(struct rcu_head *head) +{ + struct perf_counter *counter; + + counter = container_of(head, struct perf_counter, rcu_head); + if (counter->ns) + put_pid_ns(counter->ns); + kfree(counter); +} + +static void perf_pending_sync(struct perf_counter *counter); + +static void free_counter(struct perf_counter *counter) +{ + perf_pending_sync(counter); + + atomic_dec(&nr_counters); + if (counter->attr.mmap) + atomic_dec(&nr_mmap_counters); + if (counter->attr.comm) + atomic_dec(&nr_comm_counters); + + if (counter->destroy) + counter->destroy(counter); + + put_ctx(counter->ctx); + call_rcu(&counter->rcu_head, free_counter_rcu); +} + +/* + * Called when the last reference to the file is gone. + */ +static int perf_release(struct inode *inode, struct file *file) +{ + struct perf_counter *counter = file->private_data; + struct perf_counter_context *ctx = counter->ctx; + + file->private_data = NULL; + + WARN_ON_ONCE(ctx->parent_ctx); + mutex_lock(&ctx->mutex); + perf_counter_remove_from_context(counter); + mutex_unlock(&ctx->mutex); + + mutex_lock(&counter->owner->perf_counter_mutex); + list_del_init(&counter->owner_entry); + mutex_unlock(&counter->owner->perf_counter_mutex); + put_task_struct(counter->owner); + + free_counter(counter); + + return 0; +} + +/* + * Read the performance counter - simple non blocking version for now + */ +static ssize_t +perf_read_hw(struct perf_counter *counter, char __user *buf, size_t count) +{ + u64 values[3]; + int n; + + /* + * Return end-of-file for a read on a counter that is in + * error state (i.e. because it was pinned but it couldn't be + * scheduled on to the CPU at some point). + */ + if (counter->state == PERF_COUNTER_STATE_ERROR) + return 0; + + WARN_ON_ONCE(counter->ctx->parent_ctx); + mutex_lock(&counter->child_mutex); + values[0] = perf_counter_read(counter); + n = 1; + if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + values[n++] = counter->total_time_enabled + + atomic64_read(&counter->child_total_time_enabled); + if (counter->attr.read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + values[n++] = counter->total_time_running + + atomic64_read(&counter->child_total_time_running); + if (counter->attr.read_format & PERF_FORMAT_ID) + values[n++] = counter->id; + mutex_unlock(&counter->child_mutex); + + if (count < n * sizeof(u64)) + return -EINVAL; + count = n * sizeof(u64); + + if (copy_to_user(buf, values, count)) + return -EFAULT; + + return count; +} + +static ssize_t +perf_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) +{ + struct perf_counter *counter = file->private_data; + + return perf_read_hw(counter, buf, count); +} + +static unsigned int perf_poll(struct file *file, poll_table *wait) +{ + struct perf_counter *counter = file->private_data; + struct perf_mmap_data *data; + unsigned int events = POLL_HUP; + + rcu_read_lock(); + data = rcu_dereference(counter->data); + if (data) + events = atomic_xchg(&data->poll, 0); + rcu_read_unlock(); + + poll_wait(file, &counter->waitq, wait); + + return events; +} + +static void perf_counter_reset(struct perf_counter *counter) +{ + (void)perf_counter_read(counter); + atomic64_set(&counter->count, 0); + perf_counter_update_userpage(counter); +} + +static void perf_counter_for_each_sibling(struct perf_counter *counter, + void (*func)(struct perf_counter *)) +{ + struct perf_counter_context *ctx = counter->ctx; + struct perf_counter *sibling; + + WARN_ON_ONCE(ctx->parent_ctx); + mutex_lock(&ctx->mutex); + counter = counter->group_leader; + + func(counter); + list_for_each_entry(sibling, &counter->sibling_list, list_entry) + func(sibling); + mutex_unlock(&ctx->mutex); +} + +/* + * Holding the top-level counter's child_mutex means that any + * descendant process that has inherited this counter will block + * in sync_child_counter if it goes to exit, thus satisfying the + * task existence requirements of perf_counter_enable/disable. + */ +static void perf_counter_for_each_child(struct perf_counter *counter, + void (*func)(struct perf_counter *)) +{ + struct perf_counter *child; + + WARN_ON_ONCE(counter->ctx->parent_ctx); + mutex_lock(&counter->child_mutex); + func(counter); + list_for_each_entry(child, &counter->child_list, child_list) + func(child); + mutex_unlock(&counter->child_mutex); +} + +static void perf_counter_for_each(struct perf_counter *counter, + void (*func)(struct perf_counter *)) +{ + struct perf_counter *child; + + WARN_ON_ONCE(counter->ctx->parent_ctx); + mutex_lock(&counter->child_mutex); + perf_counter_for_each_sibling(counter, func); + list_for_each_entry(child, &counter->child_list, child_list) + perf_counter_for_each_sibling(child, func); + mutex_unlock(&counter->child_mutex); +} + +static int perf_counter_period(struct perf_counter *counter, u64 __user *arg) +{ + struct perf_counter_context *ctx = counter->ctx; + unsigned long size; + int ret = 0; + u64 value; + + if (!counter->attr.sample_period) + return -EINVAL; + + size = copy_from_user(&value, arg, sizeof(value)); + if (size != sizeof(value)) + return -EFAULT; + + if (!value) + return -EINVAL; + + spin_lock_irq(&ctx->lock); + if (counter->attr.freq) { + if (value > sysctl_perf_counter_sample_rate) { + ret = -EINVAL; + goto unlock; + } + + counter->attr.sample_freq = value; + } else { + perf_log_period(counter, value); + + counter->attr.sample_period = value; + counter->hw.sample_period = value; + } +unlock: + spin_unlock_irq(&ctx->lock); + + return ret; +} + +static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + struct perf_counter *counter = file->private_data; + void (*func)(struct perf_counter *); + u32 flags = arg; + + switch (cmd) { + case PERF_COUNTER_IOC_ENABLE: + func = perf_counter_enable; + break; + case PERF_COUNTER_IOC_DISABLE: + func = perf_counter_disable; + break; + case PERF_COUNTER_IOC_RESET: + func = perf_counter_reset; + break; + + case PERF_COUNTER_IOC_REFRESH: + return perf_counter_refresh(counter, arg); + + case PERF_COUNTER_IOC_PERIOD: + return perf_counter_period(counter, (u64 __user *)arg); + + default: + return -ENOTTY; + } + + if (flags & PERF_IOC_FLAG_GROUP) + perf_counter_for_each(counter, func); + else + perf_counter_for_each_child(counter, func); + + return 0; +} + +int perf_counter_task_enable(void) +{ + struct perf_counter *counter; + + mutex_lock(¤t->perf_counter_mutex); + list_for_each_entry(counter, ¤t->perf_counter_list, owner_entry) + perf_counter_for_each_child(counter, perf_counter_enable); + mutex_unlock(¤t->perf_counter_mutex); + + return 0; +} + +int perf_counter_task_disable(void) +{ + struct perf_counter *counter; + + mutex_lock(¤t->perf_counter_mutex); + list_for_each_entry(counter, ¤t->perf_counter_list, owner_entry) + perf_counter_for_each_child(counter, perf_counter_disable); + mutex_unlock(¤t->perf_counter_mutex); + + return 0; +} + +/* + * Callers need to ensure there can be no nesting of this function, otherwise + * the seqlock logic goes bad. We can not serialize this because the arch + * code calls this from NMI context. + */ +void perf_counter_update_userpage(struct perf_counter *counter) +{ + struct perf_counter_mmap_page *userpg; + struct perf_mmap_data *data; + + rcu_read_lock(); + data = rcu_dereference(counter->data); + if (!data) + goto unlock; + + userpg = data->user_page; + + /* + * Disable preemption so as to not let the corresponding user-space + * spin too long if we get preempted. + */ + preempt_disable(); + ++userpg->lock; + barrier(); + userpg->index = counter->hw.idx; + userpg->offset = atomic64_read(&counter->count); + if (counter->state == PERF_COUNTER_STATE_ACTIVE) + userpg->offset -= atomic64_read(&counter->hw.prev_count); + + barrier(); + ++userpg->lock; + preempt_enable(); +unlock: + rcu_read_unlock(); +} + +static int perf_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf) +{ + struct perf_counter *counter = vma->vm_file->private_data; + struct perf_mmap_data *data; + int ret = VM_FAULT_SIGBUS; + + rcu_read_lock(); + data = rcu_dereference(counter->data); + if (!data) + goto unlock; + + if (vmf->pgoff == 0) { + vmf->page = virt_to_page(data->user_page); + } else { + int nr = vmf->pgoff - 1; + + if ((unsigned)nr > data->nr_pages) + goto unlock; + + vmf->page = virt_to_page(data->data_pages[nr]); + } + get_page(vmf->page); + ret = 0; +unlock: + rcu_read_unlock(); + + return ret; +} + +static int perf_mmap_data_alloc(struct perf_counter *counter, int nr_pages) +{ + struct perf_mmap_data *data; + unsigned long size; + int i; + + WARN_ON(atomic_read(&counter->mmap_count)); + + size = sizeof(struct perf_mmap_data); + size += nr_pages * sizeof(void *); + + data = kzalloc(size, GFP_KERNEL); + if (!data) + goto fail; + + data->user_page = (void *)get_zeroed_page(GFP_KERNEL); + if (!data->user_page) + goto fail_user_page; + + for (i = 0; i < nr_pages; i++) { + data->data_pages[i] = (void *)get_zeroed_page(GFP_KERNEL); + if (!data->data_pages[i]) + goto fail_data_pages; + } + + data->nr_pages = nr_pages; + atomic_set(&data->lock, -1); + + rcu_assign_pointer(counter->data, data); + + return 0; + +fail_data_pages: + for (i--; i >= 0; i--) + free_page((unsigned long)data->data_pages[i]); + + free_page((unsigned long)data->user_page); + +fail_user_page: + kfree(data); + +fail: + return -ENOMEM; +} + +static void __perf_mmap_data_free(struct rcu_head *rcu_head) +{ + struct perf_mmap_data *data; + int i; + + data = container_of(rcu_head, struct perf_mmap_data, rcu_head); + + free_page((unsigned long)data->user_page); + for (i = 0; i < data->nr_pages; i++) + free_page((unsigned long)data->data_pages[i]); + kfree(data); +} + +static void perf_mmap_data_free(struct perf_counter *counter) +{ + struct perf_mmap_data *data = counter->data; + + WARN_ON(atomic_read(&counter->mmap_count)); + + rcu_assign_pointer(counter->data, NULL); + call_rcu(&data->rcu_head, __perf_mmap_data_free); +} + +static void perf_mmap_open(struct vm_area_struct *vma) +{ + struct perf_counter *counter = vma->vm_file->private_data; + + atomic_inc(&counter->mmap_count); +} + +static void perf_mmap_close(struct vm_area_struct *vma) +{ + struct perf_counter *counter = vma->vm_file->private_data; + + WARN_ON_ONCE(counter->ctx->parent_ctx); + if (atomic_dec_and_mutex_lock(&counter->mmap_count, &counter->mmap_mutex)) { + struct user_struct *user = current_user(); + + atomic_long_sub(counter->data->nr_pages + 1, &user->locked_vm); + vma->vm_mm->locked_vm -= counter->data->nr_locked; + perf_mmap_data_free(counter); + mutex_unlock(&counter->mmap_mutex); + } +} + +static struct vm_operations_struct perf_mmap_vmops = { + .open = perf_mmap_open, + .close = perf_mmap_close, + .fault = perf_mmap_fault, +}; + +static int perf_mmap(struct file *file, struct vm_area_struct *vma) +{ + struct perf_counter *counter = file->private_data; + unsigned long user_locked, user_lock_limit; + struct user_struct *user = current_user(); + unsigned long locked, lock_limit; + unsigned long vma_size; + unsigned long nr_pages; + long user_extra, extra; + int ret = 0; + + if (!(vma->vm_flags & VM_SHARED) || (vma->vm_flags & VM_WRITE)) + return -EINVAL; + + vma_size = vma->vm_end - vma->vm_start; + nr_pages = (vma_size / PAGE_SIZE) - 1; + + /* + * If we have data pages ensure they're a power-of-two number, so we + * can do bitmasks instead of modulo. + */ + if (nr_pages != 0 && !is_power_of_2(nr_pages)) + return -EINVAL; + + if (vma_size != PAGE_SIZE * (1 + nr_pages)) + return -EINVAL; + + if (vma->vm_pgoff != 0) + return -EINVAL; + + WARN_ON_ONCE(counter->ctx->parent_ctx); + mutex_lock(&counter->mmap_mutex); + if (atomic_inc_not_zero(&counter->mmap_count)) { + if (nr_pages != counter->data->nr_pages) + ret = -EINVAL; + goto unlock; + } + + user_extra = nr_pages + 1; + user_lock_limit = sysctl_perf_counter_mlock >> (PAGE_SHIFT - 10); + + /* + * Increase the limit linearly with more CPUs: + */ + user_lock_limit *= num_online_cpus(); + + user_locked = atomic_long_read(&user->locked_vm) + user_extra; + + extra = 0; + if (user_locked > user_lock_limit) + extra = user_locked - user_lock_limit; + + lock_limit = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur; + lock_limit >>= PAGE_SHIFT; + locked = vma->vm_mm->locked_vm + extra; + + if ((locked > lock_limit) && !capable(CAP_IPC_LOCK)) { + ret = -EPERM; + goto unlock; + } + + WARN_ON(counter->data); + ret = perf_mmap_data_alloc(counter, nr_pages); + if (ret) + goto unlock; + + atomic_set(&counter->mmap_count, 1); + atomic_long_add(user_extra, &user->locked_vm); + vma->vm_mm->locked_vm += extra; + counter->data->nr_locked = extra; +unlock: + mutex_unlock(&counter->mmap_mutex); + + vma->vm_flags &= ~VM_MAYWRITE; + vma->vm_flags |= VM_RESERVED; + vma->vm_ops = &perf_mmap_vmops; + + return ret; +} + +static int perf_fasync(int fd, struct file *filp, int on) +{ + struct inode *inode = filp->f_path.dentry->d_inode; + struct perf_counter *counter = filp->private_data; + int retval; + + mutex_lock(&inode->i_mutex); + retval = fasync_helper(fd, filp, on, &counter->fasync); + mutex_unlock(&inode->i_mutex); + + if (retval < 0) + return retval; + + return 0; +} + +static const struct file_operations perf_fops = { + .release = perf_release, + .read = perf_read, + .poll = perf_poll, + .unlocked_ioctl = perf_ioctl, + .compat_ioctl = perf_ioctl, + .mmap = perf_mmap, + .fasync = perf_fasync, +}; + +/* + * Perf counter wakeup + * + * If there's data, ensure we set the poll() state and publish everything + * to user-space before waking everybody up. + */ + +void perf_counter_wakeup(struct perf_counter *counter) +{ + wake_up_all(&counter->waitq); + + if (counter->pending_kill) { + kill_fasync(&counter->fasync, SIGIO, counter->pending_kill); + counter->pending_kill = 0; + } +} + +/* + * Pending wakeups + * + * Handle the case where we need to wakeup up from NMI (or rq->lock) context. + * + * The NMI bit means we cannot possibly take locks. Therefore, maintain a + * single linked list and use cmpxchg() to add entries lockless. + */ + +static void perf_pending_counter(struct perf_pending_entry *entry) +{ + struct perf_counter *counter = container_of(entry, + struct perf_counter, pending); + + if (counter->pending_disable) { + counter->pending_disable = 0; + perf_counter_disable(counter); + } + + if (counter->pending_wakeup) { + counter->pending_wakeup = 0; + perf_counter_wakeup(counter); + } +} + +#define PENDING_TAIL ((struct perf_pending_entry *)-1UL) + +static DEFINE_PER_CPU(struct perf_pending_entry *, perf_pending_head) = { + PENDING_TAIL, +}; + +static void perf_pending_queue(struct perf_pending_entry *entry, + void (*func)(struct perf_pending_entry *)) +{ + struct perf_pending_entry **head; + + if (cmpxchg(&entry->next, NULL, PENDING_TAIL) != NULL) + return; + + entry->func = func; + + head = &get_cpu_var(perf_pending_head); + + do { + entry->next = *head; + } while (cmpxchg(head, entry->next, entry) != entry->next); + + set_perf_counter_pending(); + + put_cpu_var(perf_pending_head); +} + +static int __perf_pending_run(void) +{ + struct perf_pending_entry *list; + int nr = 0; + + list = xchg(&__get_cpu_var(perf_pending_head), PENDING_TAIL); + while (list != PENDING_TAIL) { + void (*func)(struct perf_pending_entry *); + struct perf_pending_entry *entry = list; + + list = list->next; + + func = entry->func; + entry->next = NULL; + /* + * Ensure we observe the unqueue before we issue the wakeup, + * so that we won't be waiting forever. + * -- see perf_not_pending(). + */ + smp_wmb(); + + func(entry); + nr++; + } + + return nr; +} + +static inline int perf_not_pending(struct perf_counter *counter) +{ + /* + * If we flush on whatever cpu we run, there is a chance we don't + * need to wait. + */ + get_cpu(); + __perf_pending_run(); + put_cpu(); + + /* + * Ensure we see the proper queue state before going to sleep + * so that we do not miss the wakeup. -- see perf_pending_handle() + */ + smp_rmb(); + return counter->pending.next == NULL; +} + +static void perf_pending_sync(struct perf_counter *counter) +{ + wait_event(counter->waitq, perf_not_pending(counter)); +} + +void perf_counter_do_pending(void) +{ + __perf_pending_run(); +} + +/* + * Callchain support -- arch specific + */ + +__weak struct perf_callchain_entry *perf_callchain(struct pt_regs *regs) +{ + return NULL; +} + +/* + * Output + */ + +struct perf_output_handle { + struct perf_counter *counter; + struct perf_mmap_data *data; + unsigned long head; + unsigned long offset; + int nmi; + int overflow; + int locked; + unsigned long flags; +}; + +static void perf_output_wakeup(struct perf_output_handle *handle) +{ + atomic_set(&handle->data->poll, POLL_IN); + + if (handle->nmi) { + handle->counter->pending_wakeup = 1; + perf_pending_queue(&handle->counter->pending, + perf_pending_counter); + } else + perf_counter_wakeup(handle->counter); +} + +/* + * Curious locking construct. + * + * We need to ensure a later event doesn't publish a head when a former + * event isn't done writing. However since we need to deal with NMIs we + * cannot fully serialize things. + * + * What we do is serialize between CPUs so we only have to deal with NMI + * nesting on a single CPU. + * + * We only publish the head (and generate a wakeup) when the outer-most + * event completes. + */ +static void perf_output_lock(struct perf_output_handle *handle) +{ + struct perf_mmap_data *data = handle->data; + int cpu; + + handle->locked = 0; + + local_irq_save(handle->flags); + cpu = smp_processor_id(); + + if (in_nmi() && atomic_read(&data->lock) == cpu) + return; + + while (atomic_cmpxchg(&data->lock, -1, cpu) != -1) + cpu_relax(); + + handle->locked = 1; +} + +static void perf_output_unlock(struct perf_output_handle *handle) +{ + struct perf_mmap_data *data = handle->data; + unsigned long head; + int cpu; + + data->done_head = data->head; + + if (!handle->locked) + goto out; + +again: + /* + * The xchg implies a full barrier that ensures all writes are done + * before we publish the new head, matched by a rmb() in userspace when + * reading this position. + */ + while ((head = atomic_long_xchg(&data->done_head, 0))) + data->user_page->data_head = head; + + /* + * NMI can happen here, which means we can miss a done_head update. + */ + + cpu = atomic_xchg(&data->lock, -1); + WARN_ON_ONCE(cpu != smp_processor_id()); + + /* + * Therefore we have to validate we did not indeed do so. + */ + if (unlikely(atomic_long_read(&data->done_head))) { + /* + * Since we had it locked, we can lock it again. + */ + while (atomic_cmpxchg(&data->lock, -1, cpu) != -1) + cpu_relax(); + + goto again; + } + + if (atomic_xchg(&data->wakeup, 0)) + perf_output_wakeup(handle); +out: + local_irq_restore(handle->flags); +} + +static int perf_output_begin(struct perf_output_handle *handle, + struct perf_counter *counter, unsigned int size, + int nmi, int overflow) +{ + struct perf_mmap_data *data; + unsigned int offset, head; + + /* + * For inherited counters we send all the output towards the parent. + */ + if (counter->parent) + counter = counter->parent; + + rcu_read_lock(); + data = rcu_dereference(counter->data); + if (!data) + goto out; + + handle->data = data; + handle->counter = counter; + handle->nmi = nmi; + handle->overflow = overflow; + + if (!data->nr_pages) + goto fail; + + perf_output_lock(handle); + + do { + offset = head = atomic_long_read(&data->head); + head += size; + } while (atomic_long_cmpxchg(&data->head, offset, head) != offset); + + handle->offset = offset; + handle->head = head; + + if ((offset >> PAGE_SHIFT) != (head >> PAGE_SHIFT)) + atomic_set(&data->wakeup, 1); + + return 0; + +fail: + perf_output_wakeup(handle); +out: + rcu_read_unlock(); + + return -ENOSPC; +} + +static void perf_output_copy(struct perf_output_handle *handle, + const void *buf, unsigned int len) +{ + unsigned int pages_mask; + unsigned int offset; + unsigned int size; + void **pages; + + offset = handle->offset; + pages_mask = handle->data->nr_pages - 1; + pages = handle->data->data_pages; + + do { + unsigned int page_offset; + int nr; + + nr = (offset >> PAGE_SHIFT) & pages_mask; + page_offset = offset & (PAGE_SIZE - 1); + size = min_t(unsigned int, PAGE_SIZE - page_offset, len); + + memcpy(pages[nr] + page_offset, buf, size); + + len -= size; + buf += size; + offset += size; + } while (len); + + handle->offset = offset; + + /* + * Check we didn't copy past our reservation window, taking the + * possible unsigned int wrap into account. + */ + WARN_ON_ONCE(((long)(handle->head - handle->offset)) < 0); +} + +#define perf_output_put(handle, x) \ + perf_output_copy((handle), &(x), sizeof(x)) + +static void perf_output_end(struct perf_output_handle *handle) +{ + struct perf_counter *counter = handle->counter; + struct perf_mmap_data *data = handle->data; + + int wakeup_events = counter->attr.wakeup_events; + + if (handle->overflow && wakeup_events) { + int events = atomic_inc_return(&data->events); + if (events >= wakeup_events) { + atomic_sub(wakeup_events, &data->events); + atomic_set(&data->wakeup, 1); + } + } + + perf_output_unlock(handle); + rcu_read_unlock(); +} + +static u32 perf_counter_pid(struct perf_counter *counter, struct task_struct *p) +{ + /* + * only top level counters have the pid namespace they were created in + */ + if (counter->parent) + counter = counter->parent; + + return task_tgid_nr_ns(p, counter->ns); +} + +static u32 perf_counter_tid(struct perf_counter *counter, struct task_struct *p) +{ + /* + * only top level counters have the pid namespace they were created in + */ + if (counter->parent) + counter = counter->parent; + + return task_pid_nr_ns(p, counter->ns); +} + +static void perf_counter_output(struct perf_counter *counter, int nmi, + struct perf_sample_data *data) +{ + int ret; + u64 sample_type = counter->attr.sample_type; + struct perf_output_handle handle; + struct perf_event_header header; + u64 ip; + struct { + u32 pid, tid; + } tid_entry; + struct { + u64 id; + u64 counter; + } group_entry; + struct perf_callchain_entry *callchain = NULL; + int callchain_size = 0; + u64 time; + struct { + u32 cpu, reserved; + } cpu_entry; + + header.type = 0; + header.size = sizeof(header); + + header.misc = PERF_EVENT_MISC_OVERFLOW; + header.misc |= perf_misc_flags(data->regs); + + if (sample_type & PERF_SAMPLE_IP) { + ip = perf_instruction_pointer(data->regs); + header.type |= PERF_SAMPLE_IP; + header.size += sizeof(ip); + } + + if (sample_type & PERF_SAMPLE_TID) { + /* namespace issues */ + tid_entry.pid = perf_counter_pid(counter, current); + tid_entry.tid = perf_counter_tid(counter, current); + + header.type |= PERF_SAMPLE_TID; + header.size += sizeof(tid_entry); + } + + if (sample_type & PERF_SAMPLE_TIME) { + /* + * Maybe do better on x86 and provide cpu_clock_nmi() + */ + time = sched_clock(); + + header.type |= PERF_SAMPLE_TIME; + header.size += sizeof(u64); + } + + if (sample_type & PERF_SAMPLE_ADDR) { + header.type |= PERF_SAMPLE_ADDR; + header.size += sizeof(u64); + } + + if (sample_type & PERF_SAMPLE_ID) { + header.type |= PERF_SAMPLE_ID; + header.size += sizeof(u64); + } + + if (sample_type & PERF_SAMPLE_CPU) { + header.type |= PERF_SAMPLE_CPU; + header.size += sizeof(cpu_entry); + + cpu_entry.cpu = raw_smp_processor_id(); + } + + if (sample_type & PERF_SAMPLE_PERIOD) { + header.type |= PERF_SAMPLE_PERIOD; + header.size += sizeof(u64); + } + + if (sample_type & PERF_SAMPLE_GROUP) { + header.type |= PERF_SAMPLE_GROUP; + header.size += sizeof(u64) + + counter->nr_siblings * sizeof(group_entry); + } + + if (sample_type & PERF_SAMPLE_CALLCHAIN) { + callchain = perf_callchain(data->regs); + + if (callchain) { + callchain_size = (1 + callchain->nr) * sizeof(u64); + + header.type |= PERF_SAMPLE_CALLCHAIN; + header.size += callchain_size; + } + } + + ret = perf_output_begin(&handle, counter, header.size, nmi, 1); + if (ret) + return; + + perf_output_put(&handle, header); + + if (sample_type & PERF_SAMPLE_IP) + perf_output_put(&handle, ip); + + if (sample_type & PERF_SAMPLE_TID) + perf_output_put(&handle, tid_entry); + + if (sample_type & PERF_SAMPLE_TIME) + perf_output_put(&handle, time); + + if (sample_type & PERF_SAMPLE_ADDR) + perf_output_put(&handle, data->addr); + + if (sample_type & PERF_SAMPLE_ID) + perf_output_put(&handle, counter->id); + + if (sample_type & PERF_SAMPLE_CPU) + perf_output_put(&handle, cpu_entry); + + if (sample_type & PERF_SAMPLE_PERIOD) + perf_output_put(&handle, data->period); + + /* + * XXX PERF_SAMPLE_GROUP vs inherited counters seems difficult. + */ + if (sample_type & PERF_SAMPLE_GROUP) { + struct perf_counter *leader, *sub; + u64 nr = counter->nr_siblings; + + perf_output_put(&handle, nr); + + leader = counter->group_leader; + list_for_each_entry(sub, &leader->sibling_list, list_entry) { + if (sub != counter) + sub->pmu->read(sub); + + group_entry.id = sub->id; + group_entry.counter = atomic64_read(&sub->count); + + perf_output_put(&handle, group_entry); + } + } + + if (callchain) + perf_output_copy(&handle, callchain, callchain_size); + + perf_output_end(&handle); +} + +/* + * fork tracking + */ + +struct perf_fork_event { + struct task_struct *task; + + struct { + struct perf_event_header header; + + u32 pid; + u32 ppid; + } event; +}; + +static void perf_counter_fork_output(struct perf_counter *counter, + struct perf_fork_event *fork_event) +{ + struct perf_output_handle handle; + int size = fork_event->event.header.size; + struct task_struct *task = fork_event->task; + int ret = perf_output_begin(&handle, counter, size, 0, 0); + + if (ret) + return; + + fork_event->event.pid = perf_counter_pid(counter, task); + fork_event->event.ppid = perf_counter_pid(counter, task->real_parent); + + perf_output_put(&handle, fork_event->event); + perf_output_end(&handle); +} + +static int perf_counter_fork_match(struct perf_counter *counter) +{ + if (counter->attr.comm || counter->attr.mmap) + return 1; + + return 0; +} + +static void perf_counter_fork_ctx(struct perf_counter_context *ctx, + struct perf_fork_event *fork_event) +{ + struct perf_counter *counter; + + if (system_state != SYSTEM_RUNNING || list_empty(&ctx->event_list)) + return; + + rcu_read_lock(); + list_for_each_entry_rcu(counter, &ctx->event_list, event_entry) { + if (perf_counter_fork_match(counter)) + perf_counter_fork_output(counter, fork_event); + } + rcu_read_unlock(); +} + +static void perf_counter_fork_event(struct perf_fork_event *fork_event) +{ + struct perf_cpu_context *cpuctx; + struct perf_counter_context *ctx; + + cpuctx = &get_cpu_var(perf_cpu_context); + perf_counter_fork_ctx(&cpuctx->ctx, fork_event); + put_cpu_var(perf_cpu_context); + + rcu_read_lock(); + /* + * doesn't really matter which of the child contexts the + * events ends up in. + */ + ctx = rcu_dereference(current->perf_counter_ctxp); + if (ctx) + perf_counter_fork_ctx(ctx, fork_event); + rcu_read_unlock(); +} + +void perf_counter_fork(struct task_struct *task) +{ + struct perf_fork_event fork_event; + + if (!atomic_read(&nr_comm_counters) && + !atomic_read(&nr_mmap_counters)) + return; + + fork_event = (struct perf_fork_event){ + .task = task, + .event = { + .header = { + .type = PERF_EVENT_FORK, + .size = sizeof(fork_event.event), + }, + }, + }; + + perf_counter_fork_event(&fork_event); +} + +/* + * comm tracking + */ + +struct perf_comm_event { + struct task_struct *task; + char *comm; + int comm_size; + + struct { + struct perf_event_header header; + + u32 pid; + u32 tid; + } event; +}; + +static void perf_counter_comm_output(struct perf_counter *counter, + struct perf_comm_event *comm_event) +{ + struct perf_output_handle handle; + int size = comm_event->event.header.size; + int ret = perf_output_begin(&handle, counter, size, 0, 0); + + if (ret) + return; + + comm_event->event.pid = perf_counter_pid(counter, comm_event->task); + comm_event->event.tid = perf_counter_tid(counter, comm_event->task); + + perf_output_put(&handle, comm_event->event); + perf_output_copy(&handle, comm_event->comm, + comm_event->comm_size); + perf_output_end(&handle); +} + +static int perf_counter_comm_match(struct perf_counter *counter) +{ + if (counter->attr.comm) + return 1; + + return 0; +} + +static void perf_counter_comm_ctx(struct perf_counter_context *ctx, + struct perf_comm_event *comm_event) +{ + struct perf_counter *counter; + + if (system_state != SYSTEM_RUNNING || list_empty(&ctx->event_list)) + return; + + rcu_read_lock(); + list_for_each_entry_rcu(counter, &ctx->event_list, event_entry) { + if (perf_counter_comm_match(counter)) + perf_counter_comm_output(counter, comm_event); + } + rcu_read_unlock(); +} + +static void perf_counter_comm_event(struct perf_comm_event *comm_event) +{ + struct perf_cpu_context *cpuctx; + struct perf_counter_context *ctx; + unsigned int size; + char *comm = comm_event->task->comm; + + size = ALIGN(strlen(comm)+1, sizeof(u64)); + + comm_event->comm = comm; + comm_event->comm_size = size; + + comm_event->event.header.size = sizeof(comm_event->event) + size; + + cpuctx = &get_cpu_var(perf_cpu_context); + perf_counter_comm_ctx(&cpuctx->ctx, comm_event); + put_cpu_var(perf_cpu_context); + + rcu_read_lock(); + /* + * doesn't really matter which of the child contexts the + * events ends up in. + */ + ctx = rcu_dereference(current->perf_counter_ctxp); + if (ctx) + perf_counter_comm_ctx(ctx, comm_event); + rcu_read_unlock(); +} + +void perf_counter_comm(struct task_struct *task) +{ + struct perf_comm_event comm_event; + + if (!atomic_read(&nr_comm_counters)) + return; + + comm_event = (struct perf_comm_event){ + .task = task, + .event = { + .header = { .type = PERF_EVENT_COMM, }, + }, + }; + + perf_counter_comm_event(&comm_event); +} + +/* + * mmap tracking + */ + +struct perf_mmap_event { + struct vm_area_struct *vma; + + const char *file_name; + int file_size; + + struct { + struct perf_event_header header; + + u32 pid; + u32 tid; + u64 start; + u64 len; + u64 pgoff; + } event; +}; + +static void perf_counter_mmap_output(struct perf_counter *counter, + struct perf_mmap_event *mmap_event) +{ + struct perf_output_handle handle; + int size = mmap_event->event.header.size; + int ret = perf_output_begin(&handle, counter, size, 0, 0); + + if (ret) + return; + + mmap_event->event.pid = perf_counter_pid(counter, current); + mmap_event->event.tid = perf_counter_tid(counter, current); + + perf_output_put(&handle, mmap_event->event); + perf_output_copy(&handle, mmap_event->file_name, + mmap_event->file_size); + perf_output_end(&handle); +} + +static int perf_counter_mmap_match(struct perf_counter *counter, + struct perf_mmap_event *mmap_event) +{ + if (counter->attr.mmap) + return 1; + + return 0; +} + +static void perf_counter_mmap_ctx(struct perf_counter_context *ctx, + struct perf_mmap_event *mmap_event) +{ + struct perf_counter *counter; + + if (system_state != SYSTEM_RUNNING || list_empty(&ctx->event_list)) + return; + + rcu_read_lock(); + list_for_each_entry_rcu(counter, &ctx->event_list, event_entry) { + if (perf_counter_mmap_match(counter, mmap_event)) + perf_counter_mmap_output(counter, mmap_event); + } + rcu_read_unlock(); +} + +static void perf_counter_mmap_event(struct perf_mmap_event *mmap_event) +{ + struct perf_cpu_context *cpuctx; + struct perf_counter_context *ctx; + struct vm_area_struct *vma = mmap_event->vma; + struct file *file = vma->vm_file; + unsigned int size; + char tmp[16]; + char *buf = NULL; + const char *name; + + if (file) { + buf = kzalloc(PATH_MAX, GFP_KERNEL); + if (!buf) { + name = strncpy(tmp, "//enomem", sizeof(tmp)); + goto got_name; + } + name = d_path(&file->f_path, buf, PATH_MAX); + if (IS_ERR(name)) { + name = strncpy(tmp, "//toolong", sizeof(tmp)); + goto got_name; + } + } else { + name = arch_vma_name(mmap_event->vma); + if (name) + goto got_name; + + if (!vma->vm_mm) { + name = strncpy(tmp, "[vdso]", sizeof(tmp)); + goto got_name; + } + + name = strncpy(tmp, "//anon", sizeof(tmp)); + goto got_name; + } + +got_name: + size = ALIGN(strlen(name)+1, sizeof(u64)); + + mmap_event->file_name = name; + mmap_event->file_size = size; + + mmap_event->event.header.size = sizeof(mmap_event->event) + size; + + cpuctx = &get_cpu_var(perf_cpu_context); + perf_counter_mmap_ctx(&cpuctx->ctx, mmap_event); + put_cpu_var(perf_cpu_context); + + rcu_read_lock(); + /* + * doesn't really matter which of the child contexts the + * events ends up in. + */ + ctx = rcu_dereference(current->perf_counter_ctxp); + if (ctx) + perf_counter_mmap_ctx(ctx, mmap_event); + rcu_read_unlock(); + + kfree(buf); +} + +void __perf_counter_mmap(struct vm_area_struct *vma) +{ + struct perf_mmap_event mmap_event; + + if (!atomic_read(&nr_mmap_counters)) + return; + + mmap_event = (struct perf_mmap_event){ + .vma = vma, + .event = { + .header = { .type = PERF_EVENT_MMAP, }, + .start = vma->vm_start, + .len = vma->vm_end - vma->vm_start, + .pgoff = vma->vm_pgoff, + }, + }; + + perf_counter_mmap_event(&mmap_event); +} + +/* + * Log sample_period changes so that analyzing tools can re-normalize the + * event flow. + */ + +struct freq_event { + struct perf_event_header header; + u64 time; + u64 id; + u64 period; +}; + +static void perf_log_period(struct perf_counter *counter, u64 period) +{ + struct perf_output_handle handle; + struct freq_event event; + int ret; + + if (counter->hw.sample_period == period) + return; + + if (counter->attr.sample_type & PERF_SAMPLE_PERIOD) + return; + + event = (struct freq_event) { + .header = { + .type = PERF_EVENT_PERIOD, + .misc = 0, + .size = sizeof(event), + }, + .time = sched_clock(), + .id = counter->id, + .period = period, + }; + + ret = perf_output_begin(&handle, counter, sizeof(event), 1, 0); + if (ret) + return; + + perf_output_put(&handle, event); + perf_output_end(&handle); +} + +/* + * IRQ throttle logging + */ + +static void perf_log_throttle(struct perf_counter *counter, int enable) +{ + struct perf_output_handle handle; + int ret; + + struct { + struct perf_event_header header; + u64 time; + u64 id; + } throttle_event = { + .header = { + .type = PERF_EVENT_THROTTLE + 1, + .misc = 0, + .size = sizeof(throttle_event), + }, + .time = sched_clock(), + .id = counter->id, + }; + + ret = perf_output_begin(&handle, counter, sizeof(throttle_event), 1, 0); + if (ret) + return; + + perf_output_put(&handle, throttle_event); + perf_output_end(&handle); +} + +/* + * Generic counter overflow handling. + */ + +int perf_counter_overflow(struct perf_counter *counter, int nmi, + struct perf_sample_data *data) +{ + int events = atomic_read(&counter->event_limit); + int throttle = counter->pmu->unthrottle != NULL; + struct hw_perf_counter *hwc = &counter->hw; + int ret = 0; + + if (!throttle) { + hwc->interrupts++; + } else { + if (hwc->interrupts != MAX_INTERRUPTS) { + hwc->interrupts++; + if (HZ * hwc->interrupts > + (u64)sysctl_perf_counter_sample_rate) { + hwc->interrupts = MAX_INTERRUPTS; + perf_log_throttle(counter, 0); + ret = 1; + } + } else { + /* + * Keep re-disabling counters even though on the previous + * pass we disabled it - just in case we raced with a + * sched-in and the counter got enabled again: + */ + ret = 1; + } + } + + if (counter->attr.freq) { + u64 now = sched_clock(); + s64 delta = now - hwc->freq_stamp; + + hwc->freq_stamp = now; + + if (delta > 0 && delta < TICK_NSEC) + perf_adjust_period(counter, NSEC_PER_SEC / (int)delta); + } + + /* + * XXX event_limit might not quite work as expected on inherited + * counters + */ + + counter->pending_kill = POLL_IN; + if (events && atomic_dec_and_test(&counter->event_limit)) { + ret = 1; + counter->pending_kill = POLL_HUP; + if (nmi) { + counter->pending_disable = 1; + perf_pending_queue(&counter->pending, + perf_pending_counter); + } else + perf_counter_disable(counter); + } + + perf_counter_output(counter, nmi, data); + return ret; +} + +/* + * Generic software counter infrastructure + */ + +static void perf_swcounter_update(struct perf_counter *counter) +{ + struct hw_perf_counter *hwc = &counter->hw; + u64 prev, now; + s64 delta; + +again: + prev = atomic64_read(&hwc->prev_count); + now = atomic64_read(&hwc->count); + if (atomic64_cmpxchg(&hwc->prev_count, prev, now) != prev) + goto again; + + delta = now - prev; + + atomic64_add(delta, &counter->count); + atomic64_sub(delta, &hwc->period_left); +} + +static void perf_swcounter_set_period(struct perf_counter *counter) +{ + struct hw_perf_counter *hwc = &counter->hw; + s64 left = atomic64_read(&hwc->period_left); + s64 period = hwc->sample_period; + + if (unlikely(left <= -period)) { + left = period; + atomic64_set(&hwc->period_left, left); + hwc->last_period = period; + } + + if (unlikely(left <= 0)) { + left += period; + atomic64_add(period, &hwc->period_left); + hwc->last_period = period; + } + + atomic64_set(&hwc->prev_count, -left); + atomic64_set(&hwc->count, -left); +} + +static enum hrtimer_restart perf_swcounter_hrtimer(struct hrtimer *hrtimer) +{ + enum hrtimer_restart ret = HRTIMER_RESTART; + struct perf_sample_data data; + struct perf_counter *counter; + u64 period; + + counter = container_of(hrtimer, struct perf_counter, hw.hrtimer); + counter->pmu->read(counter); + + data.addr = 0; + data.regs = get_irq_regs(); + /* + * In case we exclude kernel IPs or are somehow not in interrupt + * context, provide the next best thing, the user IP. + */ + if ((counter->attr.exclude_kernel || !data.regs) && + !counter->attr.exclude_user) + data.regs = task_pt_regs(current); + + if (data.regs) { + if (perf_counter_overflow(counter, 0, &data)) + ret = HRTIMER_NORESTART; + } + + period = max_t(u64, 10000, counter->hw.sample_period); + hrtimer_forward_now(hrtimer, ns_to_ktime(period)); + + return ret; +} + +static void perf_swcounter_overflow(struct perf_counter *counter, + int nmi, struct pt_regs *regs, u64 addr) +{ + struct perf_sample_data data = { + .regs = regs, + .addr = addr, + .period = counter->hw.last_period, + }; + + perf_swcounter_update(counter); + perf_swcounter_set_period(counter); + if (perf_counter_overflow(counter, nmi, &data)) + /* soft-disable the counter */ + ; + +} + +static int perf_swcounter_is_counting(struct perf_counter *counter) +{ + struct perf_counter_context *ctx; + unsigned long flags; + int count; + + if (counter->state == PERF_COUNTER_STATE_ACTIVE) + return 1; + + if (counter->state != PERF_COUNTER_STATE_INACTIVE) + return 0; + + /* + * If the counter is inactive, it could be just because + * its task is scheduled out, or because it's in a group + * which could not go on the PMU. We want to count in + * the first case but not the second. If the context is + * currently active then an inactive software counter must + * be the second case. If it's not currently active then + * we need to know whether the counter was active when the + * context was last active, which we can determine by + * comparing counter->tstamp_stopped with ctx->time. + * + * We are within an RCU read-side critical section, + * which protects the existence of *ctx. + */ + ctx = counter->ctx; + spin_lock_irqsave(&ctx->lock, flags); + count = 1; + /* Re-check state now we have the lock */ + if (counter->state < PERF_COUNTER_STATE_INACTIVE || + counter->ctx->is_active || + counter->tstamp_stopped < ctx->time) + count = 0; + spin_unlock_irqrestore(&ctx->lock, flags); + return count; +} + +static int perf_swcounter_match(struct perf_counter *counter, + enum perf_type_id type, + u32 event, struct pt_regs *regs) +{ + if (!perf_swcounter_is_counting(counter)) + return 0; + + if (counter->attr.type != type) + return 0; + if (counter->attr.config != event) + return 0; + + if (regs) { + if (counter->attr.exclude_user && user_mode(regs)) + return 0; + + if (counter->attr.exclude_kernel && !user_mode(regs)) + return 0; + } + + return 1; +} + +static void perf_swcounter_add(struct perf_counter *counter, u64 nr, + int nmi, struct pt_regs *regs, u64 addr) +{ + int neg = atomic64_add_negative(nr, &counter->hw.count); + + if (counter->hw.sample_period && !neg && regs) + perf_swcounter_overflow(counter, nmi, regs, addr); +} + +static void perf_swcounter_ctx_event(struct perf_counter_context *ctx, + enum perf_type_id type, u32 event, + u64 nr, int nmi, struct pt_regs *regs, + u64 addr) +{ + struct perf_counter *counter; + + if (system_state != SYSTEM_RUNNING || list_empty(&ctx->event_list)) + return; + + rcu_read_lock(); + list_for_each_entry_rcu(counter, &ctx->event_list, event_entry) { + if (perf_swcounter_match(counter, type, event, regs)) + perf_swcounter_add(counter, nr, nmi, regs, addr); + } + rcu_read_unlock(); +} + +static int *perf_swcounter_recursion_context(struct perf_cpu_context *cpuctx) +{ + if (in_nmi()) + return &cpuctx->recursion[3]; + + if (in_irq()) + return &cpuctx->recursion[2]; + + if (in_softirq()) + return &cpuctx->recursion[1]; + + return &cpuctx->recursion[0]; +} + +static void __perf_swcounter_event(enum perf_type_id type, u32 event, + u64 nr, int nmi, struct pt_regs *regs, + u64 addr) +{ + struct perf_cpu_context *cpuctx = &get_cpu_var(perf_cpu_context); + int *recursion = perf_swcounter_recursion_context(cpuctx); + struct perf_counter_context *ctx; + + if (*recursion) + goto out; + + (*recursion)++; + barrier(); + + perf_swcounter_ctx_event(&cpuctx->ctx, type, event, + nr, nmi, regs, addr); + rcu_read_lock(); + /* + * doesn't really matter which of the child contexts the + * events ends up in. + */ + ctx = rcu_dereference(current->perf_counter_ctxp); + if (ctx) + perf_swcounter_ctx_event(ctx, type, event, nr, nmi, regs, addr); + rcu_read_unlock(); + + barrier(); + (*recursion)--; + +out: + put_cpu_var(perf_cpu_context); +} + +void +perf_swcounter_event(u32 event, u64 nr, int nmi, struct pt_regs *regs, u64 addr) +{ + __perf_swcounter_event(PERF_TYPE_SOFTWARE, event, nr, nmi, regs, addr); +} + +static void perf_swcounter_read(struct perf_counter *counter) +{ + perf_swcounter_update(counter); +} + +static int perf_swcounter_enable(struct perf_counter *counter) +{ + perf_swcounter_set_period(counter); + return 0; +} + +static void perf_swcounter_disable(struct perf_counter *counter) +{ + perf_swcounter_update(counter); +} + +static const struct pmu perf_ops_generic = { + .enable = perf_swcounter_enable, + .disable = perf_swcounter_disable, + .read = perf_swcounter_read, +}; + +/* + * Software counter: cpu wall time clock + */ + +static void cpu_clock_perf_counter_update(struct perf_counter *counter) +{ + int cpu = raw_smp_processor_id(); + s64 prev; + u64 now; + + now = cpu_clock(cpu); + prev = atomic64_read(&counter->hw.prev_count); + atomic64_set(&counter->hw.prev_count, now); + atomic64_add(now - prev, &counter->count); +} + +static int cpu_clock_perf_counter_enable(struct perf_counter *counter) +{ + struct hw_perf_counter *hwc = &counter->hw; + int cpu = raw_smp_processor_id(); + + atomic64_set(&hwc->prev_count, cpu_clock(cpu)); + hrtimer_init(&hwc->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hwc->hrtimer.function = perf_swcounter_hrtimer; + if (hwc->sample_period) { + u64 period = max_t(u64, 10000, hwc->sample_period); + __hrtimer_start_range_ns(&hwc->hrtimer, + ns_to_ktime(period), 0, + HRTIMER_MODE_REL, 0); + } + + return 0; +} + +static void cpu_clock_perf_counter_disable(struct perf_counter *counter) +{ + if (counter->hw.sample_period) + hrtimer_cancel(&counter->hw.hrtimer); + cpu_clock_perf_counter_update(counter); +} + +static void cpu_clock_perf_counter_read(struct perf_counter *counter) +{ + cpu_clock_perf_counter_update(counter); +} + +static const struct pmu perf_ops_cpu_clock = { + .enable = cpu_clock_perf_counter_enable, + .disable = cpu_clock_perf_counter_disable, + .read = cpu_clock_perf_counter_read, +}; + +/* + * Software counter: task time clock + */ + +static void task_clock_perf_counter_update(struct perf_counter *counter, u64 now) +{ + u64 prev; + s64 delta; + + prev = atomic64_xchg(&counter->hw.prev_count, now); + delta = now - prev; + atomic64_add(delta, &counter->count); +} + +static int task_clock_perf_counter_enable(struct perf_counter *counter) +{ + struct hw_perf_counter *hwc = &counter->hw; + u64 now; + + now = counter->ctx->time; + + atomic64_set(&hwc->prev_count, now); + hrtimer_init(&hwc->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hwc->hrtimer.function = perf_swcounter_hrtimer; + if (hwc->sample_period) { + u64 period = max_t(u64, 10000, hwc->sample_period); + __hrtimer_start_range_ns(&hwc->hrtimer, + ns_to_ktime(period), 0, + HRTIMER_MODE_REL, 0); + } + + return 0; +} + +static void task_clock_perf_counter_disable(struct perf_counter *counter) +{ + if (counter->hw.sample_period) + hrtimer_cancel(&counter->hw.hrtimer); + task_clock_perf_counter_update(counter, counter->ctx->time); + +} + +static void task_clock_perf_counter_read(struct perf_counter *counter) +{ + u64 time; + + if (!in_nmi()) { + update_context_time(counter->ctx); + time = counter->ctx->time; + } else { + u64 now = perf_clock(); + u64 delta = now - counter->ctx->timestamp; + time = counter->ctx->time + delta; + } + + task_clock_perf_counter_update(counter, time); +} + +static const struct pmu perf_ops_task_clock = { + .enable = task_clock_perf_counter_enable, + .disable = task_clock_perf_counter_disable, + .read = task_clock_perf_counter_read, +}; + +/* + * Software counter: cpu migrations + */ +void perf_counter_task_migration(struct task_struct *task, int cpu) +{ + struct perf_cpu_context *cpuctx = &per_cpu(perf_cpu_context, cpu); + struct perf_counter_context *ctx; + + perf_swcounter_ctx_event(&cpuctx->ctx, PERF_TYPE_SOFTWARE, + PERF_COUNT_SW_CPU_MIGRATIONS, + 1, 1, NULL, 0); + + ctx = perf_pin_task_context(task); + if (ctx) { + perf_swcounter_ctx_event(ctx, PERF_TYPE_SOFTWARE, + PERF_COUNT_SW_CPU_MIGRATIONS, + 1, 1, NULL, 0); + perf_unpin_context(ctx); + } +} + +#ifdef CONFIG_EVENT_PROFILE +void perf_tpcounter_event(int event_id) +{ + struct pt_regs *regs = get_irq_regs(); + + if (!regs) + regs = task_pt_regs(current); + + __perf_swcounter_event(PERF_TYPE_TRACEPOINT, event_id, 1, 1, regs, 0); +} +EXPORT_SYMBOL_GPL(perf_tpcounter_event); + +extern int ftrace_profile_enable(int); +extern void ftrace_profile_disable(int); + +static void tp_perf_counter_destroy(struct perf_counter *counter) +{ + ftrace_profile_disable(perf_event_id(&counter->attr)); +} + +static const struct pmu *tp_perf_counter_init(struct perf_counter *counter) +{ + int event_id = perf_event_id(&counter->attr); + int ret; + + ret = ftrace_profile_enable(event_id); + if (ret) + return NULL; + + counter->destroy = tp_perf_counter_destroy; + + return &perf_ops_generic; +} +#else +static const struct pmu *tp_perf_counter_init(struct perf_counter *counter) +{ + return NULL; +} +#endif + +static const struct pmu *sw_perf_counter_init(struct perf_counter *counter) +{ + const struct pmu *pmu = NULL; + + /* + * Software counters (currently) can't in general distinguish + * between user, kernel and hypervisor events. + * However, context switches and cpu migrations are considered + * to be kernel events, and page faults are never hypervisor + * events. + */ + switch (counter->attr.config) { + case PERF_COUNT_SW_CPU_CLOCK: + pmu = &perf_ops_cpu_clock; + + break; + case PERF_COUNT_SW_TASK_CLOCK: + /* + * If the user instantiates this as a per-cpu counter, + * use the cpu_clock counter instead. + */ + if (counter->ctx->task) + pmu = &perf_ops_task_clock; + else + pmu = &perf_ops_cpu_clock; + + break; + case PERF_COUNT_SW_PAGE_FAULTS: + case PERF_COUNT_SW_PAGE_FAULTS_MIN: + case PERF_COUNT_SW_PAGE_FAULTS_MAJ: + case PERF_COUNT_SW_CONTEXT_SWITCHES: + case PERF_COUNT_SW_CPU_MIGRATIONS: + pmu = &perf_ops_generic; + break; + } + + return pmu; +} + +/* + * Allocate and initialize a counter structure + */ +static struct perf_counter * +perf_counter_alloc(struct perf_counter_attr *attr, + int cpu, + struct perf_counter_context *ctx, + struct perf_counter *group_leader, + gfp_t gfpflags) +{ + const struct pmu *pmu; + struct perf_counter *counter; + struct hw_perf_counter *hwc; + long err; + + counter = kzalloc(sizeof(*counter), gfpflags); + if (!counter) + return ERR_PTR(-ENOMEM); + + /* + * Single counters are their own group leaders, with an + * empty sibling list: + */ + if (!group_leader) + group_leader = counter; + + mutex_init(&counter->child_mutex); + INIT_LIST_HEAD(&counter->child_list); + + INIT_LIST_HEAD(&counter->list_entry); + INIT_LIST_HEAD(&counter->event_entry); + INIT_LIST_HEAD(&counter->sibling_list); + init_waitqueue_head(&counter->waitq); + + mutex_init(&counter->mmap_mutex); + + counter->cpu = cpu; + counter->attr = *attr; + counter->group_leader = group_leader; + counter->pmu = NULL; + counter->ctx = ctx; + counter->oncpu = -1; + + counter->ns = get_pid_ns(current->nsproxy->pid_ns); + counter->id = atomic64_inc_return(&perf_counter_id); + + counter->state = PERF_COUNTER_STATE_INACTIVE; + + if (attr->disabled) + counter->state = PERF_COUNTER_STATE_OFF; + + pmu = NULL; + + hwc = &counter->hw; + hwc->sample_period = attr->sample_period; + if (attr->freq && attr->sample_freq) + hwc->sample_period = 1; + + atomic64_set(&hwc->period_left, hwc->sample_period); + + /* + * we currently do not support PERF_SAMPLE_GROUP on inherited counters + */ + if (attr->inherit && (attr->sample_type & PERF_SAMPLE_GROUP)) + goto done; + + if (attr->type == PERF_TYPE_RAW) { + pmu = hw_perf_counter_init(counter); + goto done; + } + + switch (attr->type) { + case PERF_TYPE_HARDWARE: + case PERF_TYPE_HW_CACHE: + pmu = hw_perf_counter_init(counter); + break; + + case PERF_TYPE_SOFTWARE: + pmu = sw_perf_counter_init(counter); + break; + + case PERF_TYPE_TRACEPOINT: + pmu = tp_perf_counter_init(counter); + break; + } +done: + err = 0; + if (!pmu) + err = -EINVAL; + else if (IS_ERR(pmu)) + err = PTR_ERR(pmu); + + if (err) { + if (counter->ns) + put_pid_ns(counter->ns); + kfree(counter); + return ERR_PTR(err); + } + + counter->pmu = pmu; + + atomic_inc(&nr_counters); + if (counter->attr.mmap) + atomic_inc(&nr_mmap_counters); + if (counter->attr.comm) + atomic_inc(&nr_comm_counters); + + return counter; +} + +/** + * sys_perf_counter_open - open a performance counter, associate it to a task/cpu + * + * @attr_uptr: event type attributes for monitoring/sampling + * @pid: target pid + * @cpu: target cpu + * @group_fd: group leader counter fd + */ +SYSCALL_DEFINE5(perf_counter_open, + const struct perf_counter_attr __user *, attr_uptr, + pid_t, pid, int, cpu, int, group_fd, unsigned long, flags) +{ + struct perf_counter *counter, *group_leader; + struct perf_counter_attr attr; + struct perf_counter_context *ctx; + struct file *counter_file = NULL; + struct file *group_file = NULL; + int fput_needed = 0; + int fput_needed2 = 0; + int ret; + + /* for future expandability... */ + if (flags) + return -EINVAL; + + if (copy_from_user(&attr, attr_uptr, sizeof(attr)) != 0) + return -EFAULT; + + if (!attr.exclude_kernel) { + if (perf_paranoid_kernel() && !capable(CAP_SYS_ADMIN)) + return -EACCES; + } + + if (attr.freq) { + if (attr.sample_freq > sysctl_perf_counter_sample_rate) + return -EINVAL; + } + + /* + * Get the target context (task or percpu): + */ + ctx = find_get_context(pid, cpu); + if (IS_ERR(ctx)) + return PTR_ERR(ctx); + + /* + * Look up the group leader (we will attach this counter to it): + */ + group_leader = NULL; + if (group_fd != -1) { + ret = -EINVAL; + group_file = fget_light(group_fd, &fput_needed); + if (!group_file) + goto err_put_context; + if (group_file->f_op != &perf_fops) + goto err_put_context; + + group_leader = group_file->private_data; + /* + * Do not allow a recursive hierarchy (this new sibling + * becoming part of another group-sibling): + */ + if (group_leader->group_leader != group_leader) + goto err_put_context; + /* + * Do not allow to attach to a group in a different + * task or CPU context: + */ + if (group_leader->ctx != ctx) + goto err_put_context; + /* + * Only a group leader can be exclusive or pinned + */ + if (attr.exclusive || attr.pinned) + goto err_put_context; + } + + counter = perf_counter_alloc(&attr, cpu, ctx, group_leader, + GFP_KERNEL); + ret = PTR_ERR(counter); + if (IS_ERR(counter)) + goto err_put_context; + + ret = anon_inode_getfd("[perf_counter]", &perf_fops, counter, 0); + if (ret < 0) + goto err_free_put_context; + + counter_file = fget_light(ret, &fput_needed2); + if (!counter_file) + goto err_free_put_context; + + counter->filp = counter_file; + WARN_ON_ONCE(ctx->parent_ctx); + mutex_lock(&ctx->mutex); + perf_install_in_context(ctx, counter, cpu); + ++ctx->generation; + mutex_unlock(&ctx->mutex); + + counter->owner = current; + get_task_struct(current); + mutex_lock(¤t->perf_counter_mutex); + list_add_tail(&counter->owner_entry, ¤t->perf_counter_list); + mutex_unlock(¤t->perf_counter_mutex); + + fput_light(counter_file, fput_needed2); + +out_fput: + fput_light(group_file, fput_needed); + + return ret; + +err_free_put_context: + kfree(counter); + +err_put_context: + put_ctx(ctx); + + goto out_fput; +} + +/* + * inherit a counter from parent task to child task: + */ +static struct perf_counter * +inherit_counter(struct perf_counter *parent_counter, + struct task_struct *parent, + struct perf_counter_context *parent_ctx, + struct task_struct *child, + struct perf_counter *group_leader, + struct perf_counter_context *child_ctx) +{ + struct perf_counter *child_counter; + + /* + * Instead of creating recursive hierarchies of counters, + * we link inherited counters back to the original parent, + * which has a filp for sure, which we use as the reference + * count: + */ + if (parent_counter->parent) + parent_counter = parent_counter->parent; + + child_counter = perf_counter_alloc(&parent_counter->attr, + parent_counter->cpu, child_ctx, + group_leader, GFP_KERNEL); + if (IS_ERR(child_counter)) + return child_counter; + get_ctx(child_ctx); + + /* + * Make the child state follow the state of the parent counter, + * not its attr.disabled bit. We hold the parent's mutex, + * so we won't race with perf_counter_{en, dis}able_family. + */ + if (parent_counter->state >= PERF_COUNTER_STATE_INACTIVE) + child_counter->state = PERF_COUNTER_STATE_INACTIVE; + else + child_counter->state = PERF_COUNTER_STATE_OFF; + + if (parent_counter->attr.freq) + child_counter->hw.sample_period = parent_counter->hw.sample_period; + + /* + * Link it up in the child's context: + */ + add_counter_to_ctx(child_counter, child_ctx); + + child_counter->parent = parent_counter; + /* + * inherit into child's child as well: + */ + child_counter->attr.inherit = 1; + + /* + * Get a reference to the parent filp - we will fput it + * when the child counter exits. This is safe to do because + * we are in the parent and we know that the filp still + * exists and has a nonzero count: + */ + atomic_long_inc(&parent_counter->filp->f_count); + + /* + * Link this into the parent counter's child list + */ + WARN_ON_ONCE(parent_counter->ctx->parent_ctx); + mutex_lock(&parent_counter->child_mutex); + list_add_tail(&child_counter->child_list, &parent_counter->child_list); + mutex_unlock(&parent_counter->child_mutex); + + return child_counter; +} + +static int inherit_group(struct perf_counter *parent_counter, + struct task_struct *parent, + struct perf_counter_context *parent_ctx, + struct task_struct *child, + struct perf_counter_context *child_ctx) +{ + struct perf_counter *leader; + struct perf_counter *sub; + struct perf_counter *child_ctr; + + leader = inherit_counter(parent_counter, parent, parent_ctx, + child, NULL, child_ctx); + if (IS_ERR(leader)) + return PTR_ERR(leader); + list_for_each_entry(sub, &parent_counter->sibling_list, list_entry) { + child_ctr = inherit_counter(sub, parent, parent_ctx, + child, leader, child_ctx); + if (IS_ERR(child_ctr)) + return PTR_ERR(child_ctr); + } + return 0; +} + +static void sync_child_counter(struct perf_counter *child_counter, + struct perf_counter *parent_counter) +{ + u64 child_val; + + child_val = atomic64_read(&child_counter->count); + + /* + * Add back the child's count to the parent's count: + */ + atomic64_add(child_val, &parent_counter->count); + atomic64_add(child_counter->total_time_enabled, + &parent_counter->child_total_time_enabled); + atomic64_add(child_counter->total_time_running, + &parent_counter->child_total_time_running); + + /* + * Remove this counter from the parent's list + */ + WARN_ON_ONCE(parent_counter->ctx->parent_ctx); + mutex_lock(&parent_counter->child_mutex); + list_del_init(&child_counter->child_list); + mutex_unlock(&parent_counter->child_mutex); + + /* + * Release the parent counter, if this was the last + * reference to it. + */ + fput(parent_counter->filp); +} + +static void +__perf_counter_exit_task(struct perf_counter *child_counter, + struct perf_counter_context *child_ctx) +{ + struct perf_counter *parent_counter; + + update_counter_times(child_counter); + perf_counter_remove_from_context(child_counter); + + parent_counter = child_counter->parent; + /* + * It can happen that parent exits first, and has counters + * that are still around due to the child reference. These + * counters need to be zapped - but otherwise linger. + */ + if (parent_counter) { + sync_child_counter(child_counter, parent_counter); + free_counter(child_counter); + } +} + +/* + * When a child task exits, feed back counter values to parent counters. + */ +void perf_counter_exit_task(struct task_struct *child) +{ + struct perf_counter *child_counter, *tmp; + struct perf_counter_context *child_ctx; + unsigned long flags; + + if (likely(!child->perf_counter_ctxp)) + return; + + local_irq_save(flags); + /* + * We can't reschedule here because interrupts are disabled, + * and either child is current or it is a task that can't be + * scheduled, so we are now safe from rescheduling changing + * our context. + */ + child_ctx = child->perf_counter_ctxp; + __perf_counter_task_sched_out(child_ctx); + + /* + * Take the context lock here so that if find_get_context is + * reading child->perf_counter_ctxp, we wait until it has + * incremented the context's refcount before we do put_ctx below. + */ + spin_lock(&child_ctx->lock); + child->perf_counter_ctxp = NULL; + if (child_ctx->parent_ctx) { + /* + * This context is a clone; unclone it so it can't get + * swapped to another process while we're removing all + * the counters from it. + */ + put_ctx(child_ctx->parent_ctx); + child_ctx->parent_ctx = NULL; + } + spin_unlock(&child_ctx->lock); + local_irq_restore(flags); + + /* + * We can recurse on the same lock type through: + * + * __perf_counter_exit_task() + * sync_child_counter() + * fput(parent_counter->filp) + * perf_release() + * mutex_lock(&ctx->mutex) + * + * But since its the parent context it won't be the same instance. + */ + mutex_lock_nested(&child_ctx->mutex, SINGLE_DEPTH_NESTING); + +again: + list_for_each_entry_safe(child_counter, tmp, &child_ctx->counter_list, + list_entry) + __perf_counter_exit_task(child_counter, child_ctx); + + /* + * If the last counter was a group counter, it will have appended all + * its siblings to the list, but we obtained 'tmp' before that which + * will still point to the list head terminating the iteration. + */ + if (!list_empty(&child_ctx->counter_list)) + goto again; + + mutex_unlock(&child_ctx->mutex); + + put_ctx(child_ctx); +} + +/* + * free an unexposed, unused context as created by inheritance by + * init_task below, used by fork() in case of fail. + */ +void perf_counter_free_task(struct task_struct *task) +{ + struct perf_counter_context *ctx = task->perf_counter_ctxp; + struct perf_counter *counter, *tmp; + + if (!ctx) + return; + + mutex_lock(&ctx->mutex); +again: + list_for_each_entry_safe(counter, tmp, &ctx->counter_list, list_entry) { + struct perf_counter *parent = counter->parent; + + if (WARN_ON_ONCE(!parent)) + continue; + + mutex_lock(&parent->child_mutex); + list_del_init(&counter->child_list); + mutex_unlock(&parent->child_mutex); + + fput(parent->filp); + + list_del_counter(counter, ctx); + free_counter(counter); + } + + if (!list_empty(&ctx->counter_list)) + goto again; + + mutex_unlock(&ctx->mutex); + + put_ctx(ctx); +} + +/* + * Initialize the perf_counter context in task_struct + */ +int perf_counter_init_task(struct task_struct *child) +{ + struct perf_counter_context *child_ctx, *parent_ctx; + struct perf_counter_context *cloned_ctx; + struct perf_counter *counter; + struct task_struct *parent = current; + int inherited_all = 1; + int ret = 0; + + child->perf_counter_ctxp = NULL; + + mutex_init(&child->perf_counter_mutex); + INIT_LIST_HEAD(&child->perf_counter_list); + + if (likely(!parent->perf_counter_ctxp)) + return 0; + + /* + * This is executed from the parent task context, so inherit + * counters that have been marked for cloning. + * First allocate and initialize a context for the child. + */ + + child_ctx = kmalloc(sizeof(struct perf_counter_context), GFP_KERNEL); + if (!child_ctx) + return -ENOMEM; + + __perf_counter_init_context(child_ctx, child); + child->perf_counter_ctxp = child_ctx; + get_task_struct(child); + + /* + * If the parent's context is a clone, pin it so it won't get + * swapped under us. + */ + parent_ctx = perf_pin_task_context(parent); + + /* + * No need to check if parent_ctx != NULL here; since we saw + * it non-NULL earlier, the only reason for it to become NULL + * is if we exit, and since we're currently in the middle of + * a fork we can't be exiting at the same time. + */ + + /* + * Lock the parent list. No need to lock the child - not PID + * hashed yet and not running, so nobody can access it. + */ + mutex_lock(&parent_ctx->mutex); + + /* + * We dont have to disable NMIs - we are only looking at + * the list, not manipulating it: + */ + list_for_each_entry_rcu(counter, &parent_ctx->event_list, event_entry) { + if (counter != counter->group_leader) + continue; + + if (!counter->attr.inherit) { + inherited_all = 0; + continue; + } + + ret = inherit_group(counter, parent, parent_ctx, + child, child_ctx); + if (ret) { + inherited_all = 0; + break; + } + } + + if (inherited_all) { + /* + * Mark the child context as a clone of the parent + * context, or of whatever the parent is a clone of. + * Note that if the parent is a clone, it could get + * uncloned at any point, but that doesn't matter + * because the list of counters and the generation + * count can't have changed since we took the mutex. + */ + cloned_ctx = rcu_dereference(parent_ctx->parent_ctx); + if (cloned_ctx) { + child_ctx->parent_ctx = cloned_ctx; + child_ctx->parent_gen = parent_ctx->parent_gen; + } else { + child_ctx->parent_ctx = parent_ctx; + child_ctx->parent_gen = parent_ctx->generation; + } + get_ctx(child_ctx->parent_ctx); + } + + mutex_unlock(&parent_ctx->mutex); + + perf_unpin_context(parent_ctx); + + return ret; +} + +static void __cpuinit perf_counter_init_cpu(int cpu) +{ + struct perf_cpu_context *cpuctx; + + cpuctx = &per_cpu(perf_cpu_context, cpu); + __perf_counter_init_context(&cpuctx->ctx, NULL); + + spin_lock(&perf_resource_lock); + cpuctx->max_pertask = perf_max_counters - perf_reserved_percpu; + spin_unlock(&perf_resource_lock); + + hw_perf_counter_setup(cpu); +} + +#ifdef CONFIG_HOTPLUG_CPU +static void __perf_counter_exit_cpu(void *info) +{ + struct perf_cpu_context *cpuctx = &__get_cpu_var(perf_cpu_context); + struct perf_counter_context *ctx = &cpuctx->ctx; + struct perf_counter *counter, *tmp; + + list_for_each_entry_safe(counter, tmp, &ctx->counter_list, list_entry) + __perf_counter_remove_from_context(counter); +} +static void perf_counter_exit_cpu(int cpu) +{ + struct perf_cpu_context *cpuctx = &per_cpu(perf_cpu_context, cpu); + struct perf_counter_context *ctx = &cpuctx->ctx; + + mutex_lock(&ctx->mutex); + smp_call_function_single(cpu, __perf_counter_exit_cpu, NULL, 1); + mutex_unlock(&ctx->mutex); +} +#else +static inline void perf_counter_exit_cpu(int cpu) { } +#endif + +static int __cpuinit +perf_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) +{ + unsigned int cpu = (long)hcpu; + + switch (action) { + + case CPU_UP_PREPARE: + case CPU_UP_PREPARE_FROZEN: + perf_counter_init_cpu(cpu); + break; + + case CPU_DOWN_PREPARE: + case CPU_DOWN_PREPARE_FROZEN: + perf_counter_exit_cpu(cpu); + break; + + default: + break; + } + + return NOTIFY_OK; +} + +/* + * This has to have a higher priority than migration_notifier in sched.c. + */ +static struct notifier_block __cpuinitdata perf_cpu_nb = { + .notifier_call = perf_cpu_notify, + .priority = 20, +}; + +void __init perf_counter_init(void) +{ + perf_cpu_notify(&perf_cpu_nb, (unsigned long)CPU_UP_PREPARE, + (void *)(long)smp_processor_id()); + register_cpu_notifier(&perf_cpu_nb); +} + +static ssize_t perf_show_reserve_percpu(struct sysdev_class *class, char *buf) +{ + return sprintf(buf, "%d\n", perf_reserved_percpu); +} + +static ssize_t +perf_set_reserve_percpu(struct sysdev_class *class, + const char *buf, + size_t count) +{ + struct perf_cpu_context *cpuctx; + unsigned long val; + int err, cpu, mpt; + + err = strict_strtoul(buf, 10, &val); + if (err) + return err; + if (val > perf_max_counters) + return -EINVAL; + + spin_lock(&perf_resource_lock); + perf_reserved_percpu = val; + for_each_online_cpu(cpu) { + cpuctx = &per_cpu(perf_cpu_context, cpu); + spin_lock_irq(&cpuctx->ctx.lock); + mpt = min(perf_max_counters - cpuctx->ctx.nr_counters, + perf_max_counters - perf_reserved_percpu); + cpuctx->max_pertask = mpt; + spin_unlock_irq(&cpuctx->ctx.lock); + } + spin_unlock(&perf_resource_lock); + + return count; +} + +static ssize_t perf_show_overcommit(struct sysdev_class *class, char *buf) +{ + return sprintf(buf, "%d\n", perf_overcommit); +} + +static ssize_t +perf_set_overcommit(struct sysdev_class *class, const char *buf, size_t count) +{ + unsigned long val; + int err; + + err = strict_strtoul(buf, 10, &val); + if (err) + return err; + if (val > 1) + return -EINVAL; + + spin_lock(&perf_resource_lock); + perf_overcommit = val; + spin_unlock(&perf_resource_lock); + + return count; +} + +static SYSDEV_CLASS_ATTR( + reserve_percpu, + 0644, + perf_show_reserve_percpu, + perf_set_reserve_percpu + ); + +static SYSDEV_CLASS_ATTR( + overcommit, + 0644, + perf_show_overcommit, + perf_set_overcommit + ); + +static struct attribute *perfclass_attrs[] = { + &attr_reserve_percpu.attr, + &attr_overcommit.attr, + NULL +}; + +static struct attribute_group perfclass_attr_group = { + .attrs = perfclass_attrs, + .name = "perf_counters", +}; + +static int __init perf_counter_sysfs_init(void) +{ + return sysfs_create_group(&cpu_sysdev_class.kset.kobj, + &perfclass_attr_group); +} +device_initcall(perf_counter_sysfs_init); diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 42c317874cf..2442d140bd9 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -25,16 +25,6 @@ /* - * Initialize a new task whose father had been ptraced. - * - * Called from copy_process(). - */ -void ptrace_fork(struct task_struct *child, unsigned long clone_flags) -{ - arch_ptrace_fork(child, clone_flags); -} - -/* * ptrace a task: make the debugger its new parent and * move it to the ptrace list. * diff --git a/kernel/rcupreempt.c b/kernel/rcupreempt.c index ce97a4df64d..beb0e659adc 100644 --- a/kernel/rcupreempt.c +++ b/kernel/rcupreempt.c @@ -1356,17 +1356,11 @@ static int rcu_sched_grace_period(void *arg) rcu_ctrlblk.sched_sleep = rcu_sched_sleeping; spin_unlock_irqrestore(&rcu_ctrlblk.schedlock, flags); - ret = 0; + ret = 0; /* unused */ __wait_event_interruptible(rcu_ctrlblk.sched_wq, rcu_ctrlblk.sched_sleep != rcu_sched_sleeping, ret); - /* - * Signals would prevent us from sleeping, and we cannot - * do much with them in any case. So flush them. - */ - if (ret) - flush_signals(current); couldsleepnext = 0; } while (!kthread_should_stop()); diff --git a/kernel/rcutree.c b/kernel/rcutree.c index d2a372fb0b9..0dccfbba6d2 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -1259,31 +1259,44 @@ static int __rcu_pending(struct rcu_state *rsp, struct rcu_data *rdp) check_cpu_stall(rsp, rdp); /* Is the RCU core waiting for a quiescent state from this CPU? */ - if (rdp->qs_pending) + if (rdp->qs_pending) { + rdp->n_rp_qs_pending++; return 1; + } /* Does this CPU have callbacks ready to invoke? */ - if (cpu_has_callbacks_ready_to_invoke(rdp)) + if (cpu_has_callbacks_ready_to_invoke(rdp)) { + rdp->n_rp_cb_ready++; return 1; + } /* Has RCU gone idle with this CPU needing another grace period? */ - if (cpu_needs_another_gp(rsp, rdp)) + if (cpu_needs_another_gp(rsp, rdp)) { + rdp->n_rp_cpu_needs_gp++; return 1; + } /* Has another RCU grace period completed? */ - if (ACCESS_ONCE(rsp->completed) != rdp->completed) /* outside of lock */ + if (ACCESS_ONCE(rsp->completed) != rdp->completed) { /* outside lock */ + rdp->n_rp_gp_completed++; return 1; + } /* Has a new RCU grace period started? */ - if (ACCESS_ONCE(rsp->gpnum) != rdp->gpnum) /* outside of lock */ + if (ACCESS_ONCE(rsp->gpnum) != rdp->gpnum) { /* outside lock */ + rdp->n_rp_gp_started++; return 1; + } /* Has an RCU GP gone long enough to send resched IPIs &c? */ if (ACCESS_ONCE(rsp->completed) != ACCESS_ONCE(rsp->gpnum) && - ((long)(ACCESS_ONCE(rsp->jiffies_force_qs) - jiffies) < 0)) + ((long)(ACCESS_ONCE(rsp->jiffies_force_qs) - jiffies) < 0)) { + rdp->n_rp_need_fqs++; return 1; + } /* nothing to do */ + rdp->n_rp_need_nothing++; return 0; } diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c index 4b1875ba940..fe1dcdbf1ca 100644 --- a/kernel/rcutree_trace.c +++ b/kernel/rcutree_trace.c @@ -213,7 +213,63 @@ static struct file_operations rcugp_fops = { .release = single_release, }; -static struct dentry *rcudir, *datadir, *datadir_csv, *hierdir, *gpdir; +static void print_one_rcu_pending(struct seq_file *m, struct rcu_data *rdp) +{ + seq_printf(m, "%3d%cnp=%ld " + "qsp=%ld cbr=%ld cng=%ld gpc=%ld gps=%ld nf=%ld nn=%ld\n", + rdp->cpu, + cpu_is_offline(rdp->cpu) ? '!' : ' ', + rdp->n_rcu_pending, + rdp->n_rp_qs_pending, + rdp->n_rp_cb_ready, + rdp->n_rp_cpu_needs_gp, + rdp->n_rp_gp_completed, + rdp->n_rp_gp_started, + rdp->n_rp_need_fqs, + rdp->n_rp_need_nothing); +} + +static void print_rcu_pendings(struct seq_file *m, struct rcu_state *rsp) +{ + int cpu; + struct rcu_data *rdp; + + for_each_possible_cpu(cpu) { + rdp = rsp->rda[cpu]; + if (rdp->beenonline) + print_one_rcu_pending(m, rdp); + } +} + +static int show_rcu_pending(struct seq_file *m, void *unused) +{ + seq_puts(m, "rcu:\n"); + print_rcu_pendings(m, &rcu_state); + seq_puts(m, "rcu_bh:\n"); + print_rcu_pendings(m, &rcu_bh_state); + return 0; +} + +static int rcu_pending_open(struct inode *inode, struct file *file) +{ + return single_open(file, show_rcu_pending, NULL); +} + +static struct file_operations rcu_pending_fops = { + .owner = THIS_MODULE, + .open = rcu_pending_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static struct dentry *rcudir; +static struct dentry *datadir; +static struct dentry *datadir_csv; +static struct dentry *gpdir; +static struct dentry *hierdir; +static struct dentry *rcu_pendingdir; + static int __init rcuclassic_trace_init(void) { rcudir = debugfs_create_dir("rcu", NULL); @@ -238,6 +294,11 @@ static int __init rcuclassic_trace_init(void) NULL, &rcuhier_fops); if (!hierdir) goto free_out; + + rcu_pendingdir = debugfs_create_file("rcu_pending", 0444, rcudir, + NULL, &rcu_pending_fops); + if (!rcu_pendingdir) + goto free_out; return 0; free_out: if (datadir) @@ -257,6 +318,7 @@ static void __exit rcuclassic_trace_cleanup(void) debugfs_remove(datadir_csv); debugfs_remove(gpdir); debugfs_remove(hierdir); + debugfs_remove(rcu_pendingdir); debugfs_remove(rcudir); } diff --git a/kernel/rtmutex.c b/kernel/rtmutex.c index 69d9cb921ff..820c5af44f3 100644 --- a/kernel/rtmutex.c +++ b/kernel/rtmutex.c @@ -300,7 +300,8 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task, * assigned pending owner [which might not have taken the * lock yet]: */ -static inline int try_to_steal_lock(struct rt_mutex *lock) +static inline int try_to_steal_lock(struct rt_mutex *lock, + struct task_struct *task) { struct task_struct *pendowner = rt_mutex_owner(lock); struct rt_mutex_waiter *next; @@ -309,11 +310,11 @@ static inline int try_to_steal_lock(struct rt_mutex *lock) if (!rt_mutex_owner_pending(lock)) return 0; - if (pendowner == current) + if (pendowner == task) return 1; spin_lock_irqsave(&pendowner->pi_lock, flags); - if (current->prio >= pendowner->prio) { + if (task->prio >= pendowner->prio) { spin_unlock_irqrestore(&pendowner->pi_lock, flags); return 0; } @@ -338,21 +339,21 @@ static inline int try_to_steal_lock(struct rt_mutex *lock) * We are going to steal the lock and a waiter was * enqueued on the pending owners pi_waiters queue. So * we have to enqueue this waiter into - * current->pi_waiters list. This covers the case, - * where current is boosted because it holds another + * task->pi_waiters list. This covers the case, + * where task is boosted because it holds another * lock and gets unboosted because the booster is * interrupted, so we would delay a waiter with higher - * priority as current->normal_prio. + * priority as task->normal_prio. * * Note: in the rare case of a SCHED_OTHER task changing * its priority and thus stealing the lock, next->task - * might be current: + * might be task: */ - if (likely(next->task != current)) { - spin_lock_irqsave(¤t->pi_lock, flags); - plist_add(&next->pi_list_entry, ¤t->pi_waiters); - __rt_mutex_adjust_prio(current); - spin_unlock_irqrestore(¤t->pi_lock, flags); + if (likely(next->task != task)) { + spin_lock_irqsave(&task->pi_lock, flags); + plist_add(&next->pi_list_entry, &task->pi_waiters); + __rt_mutex_adjust_prio(task); + spin_unlock_irqrestore(&task->pi_lock, flags); } return 1; } @@ -389,7 +390,7 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock) */ mark_rt_mutex_waiters(lock); - if (rt_mutex_owner(lock) && !try_to_steal_lock(lock)) + if (rt_mutex_owner(lock) && !try_to_steal_lock(lock, current)) return 0; /* We got the lock. */ @@ -411,6 +412,7 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock) */ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, struct rt_mutex_waiter *waiter, + struct task_struct *task, int detect_deadlock) { struct task_struct *owner = rt_mutex_owner(lock); @@ -418,21 +420,21 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, unsigned long flags; int chain_walk = 0, res; - spin_lock_irqsave(¤t->pi_lock, flags); - __rt_mutex_adjust_prio(current); - waiter->task = current; + spin_lock_irqsave(&task->pi_lock, flags); + __rt_mutex_adjust_prio(task); + waiter->task = task; waiter->lock = lock; - plist_node_init(&waiter->list_entry, current->prio); - plist_node_init(&waiter->pi_list_entry, current->prio); + plist_node_init(&waiter->list_entry, task->prio); + plist_node_init(&waiter->pi_list_entry, task->prio); /* Get the top priority waiter on the lock */ if (rt_mutex_has_waiters(lock)) top_waiter = rt_mutex_top_waiter(lock); plist_add(&waiter->list_entry, &lock->wait_list); - current->pi_blocked_on = waiter; + task->pi_blocked_on = waiter; - spin_unlock_irqrestore(¤t->pi_lock, flags); + spin_unlock_irqrestore(&task->pi_lock, flags); if (waiter == rt_mutex_top_waiter(lock)) { spin_lock_irqsave(&owner->pi_lock, flags); @@ -460,7 +462,7 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock, spin_unlock(&lock->wait_lock); res = rt_mutex_adjust_prio_chain(owner, detect_deadlock, lock, waiter, - current); + task); spin_lock(&lock->wait_lock); @@ -605,37 +607,25 @@ void rt_mutex_adjust_pi(struct task_struct *task) rt_mutex_adjust_prio_chain(task, 0, NULL, NULL, task); } -/* - * Slow path lock function: +/** + * __rt_mutex_slowlock() - Perform the wait-wake-try-to-take loop + * @lock: the rt_mutex to take + * @state: the state the task should block in (TASK_INTERRUPTIBLE + * or TASK_UNINTERRUPTIBLE) + * @timeout: the pre-initialized and started timer, or NULL for none + * @waiter: the pre-initialized rt_mutex_waiter + * @detect_deadlock: passed to task_blocks_on_rt_mutex + * + * lock->wait_lock must be held by the caller. */ static int __sched -rt_mutex_slowlock(struct rt_mutex *lock, int state, - struct hrtimer_sleeper *timeout, - int detect_deadlock) +__rt_mutex_slowlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, + struct rt_mutex_waiter *waiter, + int detect_deadlock) { - struct rt_mutex_waiter waiter; int ret = 0; - debug_rt_mutex_init_waiter(&waiter); - waiter.task = NULL; - - spin_lock(&lock->wait_lock); - - /* Try to acquire the lock again: */ - if (try_to_take_rt_mutex(lock)) { - spin_unlock(&lock->wait_lock); - return 0; - } - - set_current_state(state); - - /* Setup the timer, when timeout != NULL */ - if (unlikely(timeout)) { - hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); - if (!hrtimer_active(&timeout->timer)) - timeout->task = NULL; - } - for (;;) { /* Try to acquire the lock: */ if (try_to_take_rt_mutex(lock)) @@ -656,19 +646,19 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state, } /* - * waiter.task is NULL the first time we come here and + * waiter->task is NULL the first time we come here and * when we have been woken up by the previous owner * but the lock got stolen by a higher prio task. */ - if (!waiter.task) { - ret = task_blocks_on_rt_mutex(lock, &waiter, + if (!waiter->task) { + ret = task_blocks_on_rt_mutex(lock, waiter, current, detect_deadlock); /* * If we got woken up by the owner then start loop * all over without going into schedule to try * to get the lock now: */ - if (unlikely(!waiter.task)) { + if (unlikely(!waiter->task)) { /* * Reset the return value. We might * have returned with -EDEADLK and the @@ -684,15 +674,52 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state, spin_unlock(&lock->wait_lock); - debug_rt_mutex_print_deadlock(&waiter); + debug_rt_mutex_print_deadlock(waiter); - if (waiter.task) + if (waiter->task) schedule_rt_mutex(lock); spin_lock(&lock->wait_lock); set_current_state(state); } + return ret; +} + +/* + * Slow path lock function: + */ +static int __sched +rt_mutex_slowlock(struct rt_mutex *lock, int state, + struct hrtimer_sleeper *timeout, + int detect_deadlock) +{ + struct rt_mutex_waiter waiter; + int ret = 0; + + debug_rt_mutex_init_waiter(&waiter); + waiter.task = NULL; + + spin_lock(&lock->wait_lock); + + /* Try to acquire the lock again: */ + if (try_to_take_rt_mutex(lock)) { + spin_unlock(&lock->wait_lock); + return 0; + } + + set_current_state(state); + + /* Setup the timer, when timeout != NULL */ + if (unlikely(timeout)) { + hrtimer_start_expires(&timeout->timer, HRTIMER_MODE_ABS); + if (!hrtimer_active(&timeout->timer)) + timeout->task = NULL; + } + + ret = __rt_mutex_slowlock(lock, state, timeout, &waiter, + detect_deadlock); + set_current_state(TASK_RUNNING); if (unlikely(waiter.task)) @@ -864,9 +891,9 @@ int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock, EXPORT_SYMBOL_GPL(rt_mutex_lock_interruptible); /** - * rt_mutex_lock_interruptible_ktime - lock a rt_mutex interruptible - * the timeout structure is provided - * by the caller + * rt_mutex_timed_lock - lock a rt_mutex interruptible + * the timeout structure is provided + * by the caller * * @lock: the rt_mutex to be locked * @timeout: timeout structure or NULL (no timeout) @@ -913,7 +940,7 @@ void __sched rt_mutex_unlock(struct rt_mutex *lock) } EXPORT_SYMBOL_GPL(rt_mutex_unlock); -/*** +/** * rt_mutex_destroy - mark a mutex unusable * @lock: the mutex to be destroyed * @@ -986,6 +1013,59 @@ void rt_mutex_proxy_unlock(struct rt_mutex *lock, } /** + * rt_mutex_start_proxy_lock() - Start lock acquisition for another task + * @lock: the rt_mutex to take + * @waiter: the pre-initialized rt_mutex_waiter + * @task: the task to prepare + * @detect_deadlock: perform deadlock detection (1) or not (0) + * + * Returns: + * 0 - task blocked on lock + * 1 - acquired the lock for task, caller should wake it up + * <0 - error + * + * Special API call for FUTEX_REQUEUE_PI support. + */ +int rt_mutex_start_proxy_lock(struct rt_mutex *lock, + struct rt_mutex_waiter *waiter, + struct task_struct *task, int detect_deadlock) +{ + int ret; + + spin_lock(&lock->wait_lock); + + mark_rt_mutex_waiters(lock); + + if (!rt_mutex_owner(lock) || try_to_steal_lock(lock, task)) { + /* We got the lock for task. */ + debug_rt_mutex_lock(lock); + + rt_mutex_set_owner(lock, task, 0); + + rt_mutex_deadlock_account_lock(lock, task); + return 1; + } + + ret = task_blocks_on_rt_mutex(lock, waiter, task, detect_deadlock); + + + if (ret && !waiter->task) { + /* + * Reset the return value. We might have + * returned with -EDEADLK and the owner + * released the lock while we were walking the + * pi chain. Let the waiter sort it out. + */ + ret = 0; + } + spin_unlock(&lock->wait_lock); + + debug_rt_mutex_print_deadlock(waiter); + + return ret; +} + +/** * rt_mutex_next_owner - return the next owner of the lock * * @lock: the rt lock query @@ -1004,3 +1084,57 @@ struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock) return rt_mutex_top_waiter(lock)->task; } + +/** + * rt_mutex_finish_proxy_lock() - Complete lock acquisition + * @lock: the rt_mutex we were woken on + * @to: the timeout, null if none. hrtimer should already have + * been started. + * @waiter: the pre-initialized rt_mutex_waiter + * @detect_deadlock: perform deadlock detection (1) or not (0) + * + * Complete the lock acquisition started our behalf by another thread. + * + * Returns: + * 0 - success + * <0 - error, one of -EINTR, -ETIMEDOUT, or -EDEADLK + * + * Special API call for PI-futex requeue support + */ +int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, + struct hrtimer_sleeper *to, + struct rt_mutex_waiter *waiter, + int detect_deadlock) +{ + int ret; + + spin_lock(&lock->wait_lock); + + set_current_state(TASK_INTERRUPTIBLE); + + ret = __rt_mutex_slowlock(lock, TASK_INTERRUPTIBLE, to, waiter, + detect_deadlock); + + set_current_state(TASK_RUNNING); + + if (unlikely(waiter->task)) + remove_waiter(lock, waiter); + + /* + * try_to_take_rt_mutex() sets the waiter bit unconditionally. We might + * have to fix that up. + */ + fixup_rt_mutex_waiters(lock); + + spin_unlock(&lock->wait_lock); + + /* + * Readjust priority, when we did not get the lock. We might have been + * the pending owner and boosted. Since we did not take the lock, the + * PI boost has to go. + */ + if (unlikely(ret)) + rt_mutex_adjust_prio(current); + + return ret; +} diff --git a/kernel/rtmutex_common.h b/kernel/rtmutex_common.h index e124bf5800e..97a2f81866a 100644 --- a/kernel/rtmutex_common.h +++ b/kernel/rtmutex_common.h @@ -120,6 +120,14 @@ extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock, struct task_struct *proxy_owner); extern void rt_mutex_proxy_unlock(struct rt_mutex *lock, struct task_struct *proxy_owner); +extern int rt_mutex_start_proxy_lock(struct rt_mutex *lock, + struct rt_mutex_waiter *waiter, + struct task_struct *task, + int detect_deadlock); +extern int rt_mutex_finish_proxy_lock(struct rt_mutex *lock, + struct hrtimer_sleeper *to, + struct rt_mutex_waiter *waiter, + int detect_deadlock); #ifdef CONFIG_DEBUG_RT_MUTEXES # include "rtmutex-debug.h" diff --git a/kernel/sched.c b/kernel/sched.c index 26efa475bdc..5b3f6ec1b0b 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -39,6 +39,7 @@ #include <linux/completion.h> #include <linux/kernel_stat.h> #include <linux/debug_locks.h> +#include <linux/perf_counter.h> #include <linux/security.h> #include <linux/notifier.h> #include <linux/profile.h> @@ -72,13 +73,15 @@ #include <linux/debugfs.h> #include <linux/ctype.h> #include <linux/ftrace.h> -#include <trace/sched.h> #include <asm/tlb.h> #include <asm/irq_regs.h> #include "sched_cpupri.h" +#define CREATE_TRACE_POINTS +#include <trace/events/sched.h> + /* * Convert user-nice values [ -20 ... 0 ... 19 ] * to static priority [ MAX_RT_PRIO..MAX_PRIO-1 ], @@ -118,12 +121,6 @@ */ #define RUNTIME_INF ((u64)~0ULL) -DEFINE_TRACE(sched_wait_task); -DEFINE_TRACE(sched_wakeup); -DEFINE_TRACE(sched_wakeup_new); -DEFINE_TRACE(sched_switch); -DEFINE_TRACE(sched_migrate_task); - #ifdef CONFIG_SMP static void double_rq_lock(struct rq *rq1, struct rq *rq2); @@ -584,6 +581,7 @@ struct rq { struct load_weight load; unsigned long nr_load_updates; u64 nr_switches; + u64 nr_migrations_in; struct cfs_rq cfs; struct rt_rq rt; @@ -630,6 +628,10 @@ struct rq { struct list_head migration_queue; #endif + /* calc_load related fields */ + unsigned long calc_load_update; + long calc_load_active; + #ifdef CONFIG_SCHED_HRTICK #ifdef CONFIG_SMP int hrtick_csd_pending; @@ -692,7 +694,7 @@ static inline int cpu_of(struct rq *rq) #define task_rq(p) cpu_rq(task_cpu(p)) #define cpu_curr(cpu) (cpu_rq(cpu)->curr) -static inline void update_rq_clock(struct rq *rq) +inline void update_rq_clock(struct rq *rq) { rq->clock = sched_clock_cpu(cpu_of(rq)); } @@ -1728,6 +1730,8 @@ static void cfs_rq_set_shares(struct cfs_rq *cfs_rq, unsigned long shares) } #endif +static void calc_load_account_active(struct rq *this_rq); + #include "sched_stats.h" #include "sched_idletask.c" #include "sched_fair.c" @@ -1958,7 +1962,7 @@ void set_task_cpu(struct task_struct *p, unsigned int new_cpu) clock_offset = old_rq->clock - new_rq->clock; - trace_sched_migrate_task(p, task_cpu(p), new_cpu); + trace_sched_migrate_task(p, new_cpu); #ifdef CONFIG_SCHEDSTATS if (p->se.wait_start) @@ -1967,12 +1971,16 @@ void set_task_cpu(struct task_struct *p, unsigned int new_cpu) p->se.sleep_start -= clock_offset; if (p->se.block_start) p->se.block_start -= clock_offset; +#endif if (old_cpu != new_cpu) { - schedstat_inc(p, se.nr_migrations); + p->se.nr_migrations++; + new_rq->nr_migrations_in++; +#ifdef CONFIG_SCHEDSTATS if (task_hot(p, old_rq->clock, NULL)) schedstat_inc(p, se.nr_forced2_migrations); - } #endif + perf_counter_task_migration(p, new_cpu); + } p->se.vruntime -= old_cfsrq->min_vruntime - new_cfsrq->min_vruntime; @@ -2015,6 +2023,49 @@ migrate_task(struct task_struct *p, int dest_cpu, struct migration_req *req) } /* + * wait_task_context_switch - wait for a thread to complete at least one + * context switch. + * + * @p must not be current. + */ +void wait_task_context_switch(struct task_struct *p) +{ + unsigned long nvcsw, nivcsw, flags; + int running; + struct rq *rq; + + nvcsw = p->nvcsw; + nivcsw = p->nivcsw; + for (;;) { + /* + * The runqueue is assigned before the actual context + * switch. We need to take the runqueue lock. + * + * We could check initially without the lock but it is + * very likely that we need to take the lock in every + * iteration. + */ + rq = task_rq_lock(p, &flags); + running = task_running(rq, p); + task_rq_unlock(rq, &flags); + + if (likely(!running)) + break; + /* + * The switch count is incremented before the actual + * context switch. We thus wait for two switches to be + * sure at least one completed. + */ + if ((p->nvcsw - nvcsw) > 1) + break; + if ((p->nivcsw - nivcsw) > 1) + break; + + cpu_relax(); + } +} + +/* * wait_task_inactive - wait for a thread to unschedule. * * If @match_state is nonzero, it's the @p->state value just checked and @@ -2324,6 +2375,27 @@ static int sched_balance_self(int cpu, int flag) #endif /* CONFIG_SMP */ +/** + * task_oncpu_function_call - call a function on the cpu on which a task runs + * @p: the task to evaluate + * @func: the function to be called + * @info: the function call argument + * + * Calls the function @func when the task is currently running. This might + * be on the current CPU, which just calls the function directly + */ +void task_oncpu_function_call(struct task_struct *p, + void (*func) (void *info), void *info) +{ + int cpu; + + preempt_disable(); + cpu = task_cpu(p); + if (task_curr(p)) + smp_call_function_single(cpu, func, info, 1); + preempt_enable(); +} + /*** * try_to_wake_up - wake up a thread * @p: the to-be-woken-up thread @@ -2458,6 +2530,17 @@ out: return success; } +/** + * wake_up_process - Wake up a specific process + * @p: The process to be woken up. + * + * Attempt to wake up the nominated process and move it to the set of runnable + * processes. Returns 1 if the process was woken up, 0 if it was already + * running. + * + * It may be assumed that this function implies a write memory barrier before + * changing the task state if and only if any tasks are woken up. + */ int wake_up_process(struct task_struct *p) { return try_to_wake_up(p, TASK_ALL, 0); @@ -2480,6 +2563,7 @@ static void __sched_fork(struct task_struct *p) p->se.exec_start = 0; p->se.sum_exec_runtime = 0; p->se.prev_sum_exec_runtime = 0; + p->se.nr_migrations = 0; p->se.last_wakeup = 0; p->se.avg_overlap = 0; p->se.start_runtime = 0; @@ -2710,6 +2794,7 @@ static void finish_task_switch(struct rq *rq, struct task_struct *prev) */ prev_state = prev->state; finish_arch_switch(prev); + perf_counter_task_sched_in(current, cpu_of(rq)); finish_lock_switch(rq, prev); #ifdef CONFIG_SMP if (post_schedule) @@ -2766,7 +2851,7 @@ context_switch(struct rq *rq, struct task_struct *prev, * combine the page table reload and the switch backend into * one hypercall. */ - arch_enter_lazy_cpu_mode(); + arch_start_context_switch(prev); if (unlikely(!mm)) { next->active_mm = oldmm; @@ -2856,19 +2941,81 @@ unsigned long nr_iowait(void) return sum; } -unsigned long nr_active(void) +/* Variables and functions for calc_load */ +static atomic_long_t calc_load_tasks; +static unsigned long calc_load_update; +unsigned long avenrun[3]; +EXPORT_SYMBOL(avenrun); + +/** + * get_avenrun - get the load average array + * @loads: pointer to dest load array + * @offset: offset to add + * @shift: shift count to shift the result left + * + * These values are estimates at best, so no need for locking. + */ +void get_avenrun(unsigned long *loads, unsigned long offset, int shift) { - unsigned long i, running = 0, uninterruptible = 0; + loads[0] = (avenrun[0] + offset) << shift; + loads[1] = (avenrun[1] + offset) << shift; + loads[2] = (avenrun[2] + offset) << shift; +} - for_each_online_cpu(i) { - running += cpu_rq(i)->nr_running; - uninterruptible += cpu_rq(i)->nr_uninterruptible; - } +static unsigned long +calc_load(unsigned long load, unsigned long exp, unsigned long active) +{ + load *= exp; + load += active * (FIXED_1 - exp); + return load >> FSHIFT; +} - if (unlikely((long)uninterruptible < 0)) - uninterruptible = 0; +/* + * calc_load - update the avenrun load estimates 10 ticks after the + * CPUs have updated calc_load_tasks. + */ +void calc_global_load(void) +{ + unsigned long upd = calc_load_update + 10; + long active; - return running + uninterruptible; + if (time_before(jiffies, upd)) + return; + + active = atomic_long_read(&calc_load_tasks); + active = active > 0 ? active * FIXED_1 : 0; + + avenrun[0] = calc_load(avenrun[0], EXP_1, active); + avenrun[1] = calc_load(avenrun[1], EXP_5, active); + avenrun[2] = calc_load(avenrun[2], EXP_15, active); + + calc_load_update += LOAD_FREQ; +} + +/* + * Either called from update_cpu_load() or from a cpu going idle + */ +static void calc_load_account_active(struct rq *this_rq) +{ + long nr_active, delta; + + nr_active = this_rq->nr_running; + nr_active += (long) this_rq->nr_uninterruptible; + + if (nr_active != this_rq->calc_load_active) { + delta = nr_active - this_rq->calc_load_active; + this_rq->calc_load_active = nr_active; + atomic_long_add(delta, &calc_load_tasks); + } +} + +/* + * Externally visible per-cpu scheduler statistics: + * cpu_nr_migrations(cpu) - number of migrations into that cpu + */ +u64 cpu_nr_migrations(int cpu) +{ + return cpu_rq(cpu)->nr_migrations_in; } /* @@ -2899,6 +3046,11 @@ static void update_cpu_load(struct rq *this_rq) new_load += scale-1; this_rq->cpu_load[i] = (old_load*(scale-1) + new_load) >> i; } + + if (time_after_eq(jiffies, this_rq->calc_load_update)) { + this_rq->calc_load_update += LOAD_FREQ; + calc_load_account_active(this_rq); + } } #ifdef CONFIG_SMP @@ -4240,10 +4392,126 @@ static void active_load_balance(struct rq *busiest_rq, int busiest_cpu) static struct { atomic_t load_balancer; cpumask_var_t cpu_mask; + cpumask_var_t ilb_grp_nohz_mask; } nohz ____cacheline_aligned = { .load_balancer = ATOMIC_INIT(-1), }; +#if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) +/** + * lowest_flag_domain - Return lowest sched_domain containing flag. + * @cpu: The cpu whose lowest level of sched domain is to + * be returned. + * @flag: The flag to check for the lowest sched_domain + * for the given cpu. + * + * Returns the lowest sched_domain of a cpu which contains the given flag. + */ +static inline struct sched_domain *lowest_flag_domain(int cpu, int flag) +{ + struct sched_domain *sd; + + for_each_domain(cpu, sd) + if (sd && (sd->flags & flag)) + break; + + return sd; +} + +/** + * for_each_flag_domain - Iterates over sched_domains containing the flag. + * @cpu: The cpu whose domains we're iterating over. + * @sd: variable holding the value of the power_savings_sd + * for cpu. + * @flag: The flag to filter the sched_domains to be iterated. + * + * Iterates over all the scheduler domains for a given cpu that has the 'flag' + * set, starting from the lowest sched_domain to the highest. + */ +#define for_each_flag_domain(cpu, sd, flag) \ + for (sd = lowest_flag_domain(cpu, flag); \ + (sd && (sd->flags & flag)); sd = sd->parent) + +/** + * is_semi_idle_group - Checks if the given sched_group is semi-idle. + * @ilb_group: group to be checked for semi-idleness + * + * Returns: 1 if the group is semi-idle. 0 otherwise. + * + * We define a sched_group to be semi idle if it has atleast one idle-CPU + * and atleast one non-idle CPU. This helper function checks if the given + * sched_group is semi-idle or not. + */ +static inline int is_semi_idle_group(struct sched_group *ilb_group) +{ + cpumask_and(nohz.ilb_grp_nohz_mask, nohz.cpu_mask, + sched_group_cpus(ilb_group)); + + /* + * A sched_group is semi-idle when it has atleast one busy cpu + * and atleast one idle cpu. + */ + if (cpumask_empty(nohz.ilb_grp_nohz_mask)) + return 0; + + if (cpumask_equal(nohz.ilb_grp_nohz_mask, sched_group_cpus(ilb_group))) + return 0; + + return 1; +} +/** + * find_new_ilb - Finds the optimum idle load balancer for nomination. + * @cpu: The cpu which is nominating a new idle_load_balancer. + * + * Returns: Returns the id of the idle load balancer if it exists, + * Else, returns >= nr_cpu_ids. + * + * This algorithm picks the idle load balancer such that it belongs to a + * semi-idle powersavings sched_domain. The idea is to try and avoid + * completely idle packages/cores just for the purpose of idle load balancing + * when there are other idle cpu's which are better suited for that job. + */ +static int find_new_ilb(int cpu) +{ + struct sched_domain *sd; + struct sched_group *ilb_group; + + /* + * Have idle load balancer selection from semi-idle packages only + * when power-aware load balancing is enabled + */ + if (!(sched_smt_power_savings || sched_mc_power_savings)) + goto out_done; + + /* + * Optimize for the case when we have no idle CPUs or only one + * idle CPU. Don't walk the sched_domain hierarchy in such cases + */ + if (cpumask_weight(nohz.cpu_mask) < 2) + goto out_done; + + for_each_flag_domain(cpu, sd, SD_POWERSAVINGS_BALANCE) { + ilb_group = sd->groups; + + do { + if (is_semi_idle_group(ilb_group)) + return cpumask_first(nohz.ilb_grp_nohz_mask); + + ilb_group = ilb_group->next; + + } while (ilb_group != sd->groups); + } + +out_done: + return cpumask_first(nohz.cpu_mask); +} +#else /* (CONFIG_SCHED_MC || CONFIG_SCHED_SMT) */ +static inline int find_new_ilb(int call_cpu) +{ + return cpumask_first(nohz.cpu_mask); +} +#endif + /* * This routine will try to nominate the ilb (idle load balancing) * owner among the cpus whose ticks are stopped. ilb owner will do the idle @@ -4298,8 +4566,24 @@ int select_nohz_load_balancer(int stop_tick) /* make me the ilb owner */ if (atomic_cmpxchg(&nohz.load_balancer, -1, cpu) == -1) return 1; - } else if (atomic_read(&nohz.load_balancer) == cpu) + } else if (atomic_read(&nohz.load_balancer) == cpu) { + int new_ilb; + + if (!(sched_smt_power_savings || + sched_mc_power_savings)) + return 1; + /* + * Check to see if there is a more power-efficient + * ilb. + */ + new_ilb = find_new_ilb(cpu); + if (new_ilb < nr_cpu_ids && new_ilb != cpu) { + atomic_set(&nohz.load_balancer, -1); + resched_cpu(new_ilb); + return 0; + } return 1; + } } else { if (!cpumask_test_cpu(cpu, nohz.cpu_mask)) return 0; @@ -4468,15 +4752,7 @@ static inline void trigger_load_balance(struct rq *rq, int cpu) } if (atomic_read(&nohz.load_balancer) == -1) { - /* - * simple selection for now: Nominate the - * first cpu in the nohz list to be the next - * ilb owner. - * - * TBD: Traverse the sched domains and nominate - * the nearest cpu in the nohz.cpu_mask. - */ - int ilb = cpumask_first(nohz.cpu_mask); + int ilb = find_new_ilb(cpu); if (ilb < nr_cpu_ids) resched_cpu(ilb); @@ -4840,6 +5116,8 @@ void scheduler_tick(void) curr->sched_class->task_tick(rq, curr, 0); spin_unlock(&rq->lock); + perf_counter_task_tick(curr, cpu); + #ifdef CONFIG_SMP rq->idle_at_tick = idle_cpu(cpu); trigger_load_balance(rq, cpu); @@ -5007,13 +5285,15 @@ pick_next_task(struct rq *rq) /* * schedule() is the main scheduler function. */ -asmlinkage void __sched __schedule(void) +asmlinkage void __sched schedule(void) { struct task_struct *prev, *next; unsigned long *switch_count; struct rq *rq; int cpu; +need_resched: + preempt_disable(); cpu = smp_processor_id(); rq = cpu_rq(cpu); rcu_qsctr_inc(cpu); @@ -5053,6 +5333,7 @@ need_resched_nonpreemptible: if (likely(prev != next)) { sched_info_switch(prev, next); + perf_counter_task_sched_out(prev, next, cpu); rq->nr_switches++; rq->curr = next; @@ -5070,15 +5351,9 @@ need_resched_nonpreemptible: if (unlikely(reacquire_kernel_lock(current) < 0)) goto need_resched_nonpreemptible; -} -asmlinkage void __sched schedule(void) -{ -need_resched: - preempt_disable(); - __schedule(); preempt_enable_no_resched(); - if (unlikely(test_thread_flag(TIF_NEED_RESCHED))) + if (need_resched()) goto need_resched; } EXPORT_SYMBOL(schedule); @@ -5221,7 +5496,7 @@ EXPORT_SYMBOL(default_wake_function); * started to run but is not in state TASK_RUNNING. try_to_wake_up() returns * zero in this (rare) case, and we handle it by continuing to scan the queue. */ -void __wake_up_common(wait_queue_head_t *q, unsigned int mode, +static void __wake_up_common(wait_queue_head_t *q, unsigned int mode, int nr_exclusive, int sync, void *key) { wait_queue_t *curr, *next; @@ -5241,6 +5516,9 @@ void __wake_up_common(wait_queue_head_t *q, unsigned int mode, * @mode: which threads * @nr_exclusive: how many wake-one or wake-many threads to wake up * @key: is directly passed to the wakeup function + * + * It may be assumed that this function implies a write memory barrier before + * changing the task state if and only if any tasks are woken up. */ void __wake_up(wait_queue_head_t *q, unsigned int mode, int nr_exclusive, void *key) @@ -5279,6 +5557,9 @@ void __wake_up_locked_key(wait_queue_head_t *q, unsigned int mode, void *key) * with each other. This can prevent needless bouncing between CPUs. * * On UP it can prevent extra preemption. + * + * It may be assumed that this function implies a write memory barrier before + * changing the task state if and only if any tasks are woken up. */ void __wake_up_sync_key(wait_queue_head_t *q, unsigned int mode, int nr_exclusive, void *key) @@ -5315,6 +5596,9 @@ EXPORT_SYMBOL_GPL(__wake_up_sync); /* For internal use only */ * awakened in the same order in which they were queued. * * See also complete_all(), wait_for_completion() and related routines. + * + * It may be assumed that this function implies a write memory barrier before + * changing the task state if and only if any tasks are woken up. */ void complete(struct completion *x) { @@ -5332,6 +5616,9 @@ EXPORT_SYMBOL(complete); * @x: holds the state of this particular completion * * This will wake up all threads waiting on this particular completion event. + * + * It may be assumed that this function implies a write memory barrier before + * changing the task state if and only if any tasks are woken up. */ void complete_all(struct completion *x) { @@ -6490,8 +6777,9 @@ void sched_show_task(struct task_struct *p) #ifdef CONFIG_DEBUG_STACK_USAGE free = stack_not_used(p); #endif - printk(KERN_CONT "%5lu %5d %6d\n", free, - task_pid_nr(p), task_pid_nr(p->real_parent)); + printk(KERN_CONT "%5lu %5d %6d 0x%08lx\n", free, + task_pid_nr(p), task_pid_nr(p->real_parent), + (unsigned long)task_thread_info(p)->flags); show_stack(p, NULL); } @@ -6970,6 +7258,14 @@ static void migrate_dead_tasks(unsigned int dead_cpu) } } + +/* + * remove the tasks which were accounted by rq from calc_load_tasks. + */ +static void calc_global_load_remove(struct rq *rq) +{ + atomic_long_sub(rq->calc_load_active, &calc_load_tasks); +} #endif /* CONFIG_HOTPLUG_CPU */ #if defined(CONFIG_SCHED_DEBUG) && defined(CONFIG_SYSCTL) @@ -7204,6 +7500,8 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) /* Update our root-domain */ rq = cpu_rq(cpu); spin_lock_irqsave(&rq->lock, flags); + rq->calc_load_update = calc_load_update; + rq->calc_load_active = 0; if (rq->rd) { BUG_ON(!cpumask_test_cpu(cpu, rq->rd->span)); @@ -7243,7 +7541,7 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) cpuset_unlock(); migrate_nr_uninterruptible(rq); BUG_ON(rq->nr_running != 0); - + calc_global_load_remove(rq); /* * No need to migrate the tasks: it was best-effort if * they didn't take sched_hotcpu_mutex. Just wake up @@ -7279,8 +7577,10 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) return NOTIFY_OK; } -/* Register at highest priority so that task migration (migrate_all_tasks) - * happens before everything else. +/* + * Register at high priority so that task migration (migrate_all_tasks) + * happens before everything else. This has to be lower priority than + * the notifier in the perf_counter subsystem, though. */ static struct notifier_block __cpuinitdata migration_notifier = { .notifier_call = migration_call, @@ -7753,8 +8053,9 @@ int sched_smt_power_savings = 0, sched_mc_power_savings = 0; /* * The cpus mask in sched_group and sched_domain hangs off the end. - * FIXME: use cpumask_var_t or dynamic percpu alloc to avoid wasting space - * for nr_cpu_ids < CONFIG_NR_CPUS. + * + * ( See the the comments in include/linux/sched.h:struct sched_group + * and struct sched_domain. ) */ struct static_sched_group { struct sched_group sg; @@ -7875,7 +8176,7 @@ static void init_numa_sched_groups_power(struct sched_group *group_head) struct sched_domain *sd; sd = &per_cpu(phys_domains, j).sd; - if (j != cpumask_first(sched_group_cpus(sd->groups))) { + if (j != group_first_cpu(sd->groups)) { /* * Only add "power" once for each * physical package. @@ -7953,7 +8254,7 @@ static void init_sched_groups_power(int cpu, struct sched_domain *sd) WARN_ON(!sd || !sd->groups); - if (cpu != cpumask_first(sched_group_cpus(sd->groups))) + if (cpu != group_first_cpu(sd->groups)) return; child = sd->child; @@ -8938,6 +9239,8 @@ void __init sched_init(void) rq = cpu_rq(i); spin_lock_init(&rq->lock); rq->nr_running = 0; + rq->calc_load_active = 0; + rq->calc_load_update = jiffies + LOAD_FREQ; init_cfs_rq(&rq->cfs, rq); init_rt_rq(&rq->rt, rq); #ifdef CONFIG_FAIR_GROUP_SCHED @@ -8958,7 +9261,7 @@ void __init sched_init(void) * 1024) and two child groups A0 and A1 (of weight 1024 each), * then A0's share of the cpu resource is: * - * A0's bandwidth = 1024 / (10*1024 + 1024 + 1024) = 8.33% + * A0's bandwidth = 1024 / (10*1024 + 1024 + 1024) = 8.33% * * We achieve this by letting init_task_group's tasks sit * directly in rq->cfs (i.e init_task_group->se[] = NULL). @@ -9045,6 +9348,9 @@ void __init sched_init(void) * when this runqueue becomes "idle". */ init_idle(current, smp_processor_id()); + + calc_load_update = jiffies + LOAD_FREQ; + /* * During early bootup we pretend to be a normal task: */ @@ -9055,10 +9361,13 @@ void __init sched_init(void) #ifdef CONFIG_SMP #ifdef CONFIG_NO_HZ alloc_bootmem_cpumask_var(&nohz.cpu_mask); + alloc_bootmem_cpumask_var(&nohz.ilb_grp_nohz_mask); #endif alloc_bootmem_cpumask_var(&cpu_isolated_map); #endif /* SMP */ + perf_counter_init(); + scheduler_running = 1; } @@ -9800,6 +10109,13 @@ static int sched_rt_global_constraints(void) if (sysctl_sched_rt_period <= 0) return -EINVAL; + /* + * There's always some RT tasks in the root group + * -- migration, kstopmachine etc.. + */ + if (sysctl_sched_rt_runtime == 0) + return -EBUSY; + spin_lock_irqsave(&def_rt_bandwidth.rt_runtime_lock, flags); for_each_possible_cpu(i) { struct rt_rq *rt_rq = &cpu_rq(i)->rt; diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 3816f217f11..5f9650e8fe7 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1487,17 +1487,10 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int sync) find_matching_se(&se, &pse); - while (se) { - BUG_ON(!pse); + BUG_ON(!pse); - if (wakeup_preempt_entity(se, pse) == 1) { - resched_task(curr); - break; - } - - se = parent_entity(se); - pse = parent_entity(pse); - } + if (wakeup_preempt_entity(se, pse) == 1) + resched_task(curr); } static struct task_struct *pick_next_task_fair(struct rq *rq) diff --git a/kernel/sched_idletask.c b/kernel/sched_idletask.c index 8a21a2e28c1..499672c10cb 100644 --- a/kernel/sched_idletask.c +++ b/kernel/sched_idletask.c @@ -22,7 +22,8 @@ static void check_preempt_curr_idle(struct rq *rq, struct task_struct *p, int sy static struct task_struct *pick_next_task_idle(struct rq *rq) { schedstat_inc(rq, sched_goidle); - + /* adjust the active tasks as we might go into a long sleep */ + calc_load_account_active(rq); return rq->idle; } diff --git a/kernel/signal.c b/kernel/signal.c index d8034737db4..dba6ae99978 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -27,7 +27,7 @@ #include <linux/freezer.h> #include <linux/pid_namespace.h> #include <linux/nsproxy.h> -#include <trace/sched.h> +#include <trace/events/sched.h> #include <asm/param.h> #include <asm/uaccess.h> @@ -41,8 +41,6 @@ static struct kmem_cache *sigqueue_cachep; -DEFINE_TRACE(sched_signal_send); - static void __user *sig_handler(struct task_struct *t, int sig) { return t->sighand->action[sig - 1].sa.sa_handler; @@ -2278,24 +2276,17 @@ SYSCALL_DEFINE2(kill, pid_t, pid, int, sig) return kill_something_info(sig, &info, pid); } -static int do_tkill(pid_t tgid, pid_t pid, int sig) +static int +do_send_specific(pid_t tgid, pid_t pid, int sig, struct siginfo *info) { - int error; - struct siginfo info; struct task_struct *p; unsigned long flags; - - error = -ESRCH; - info.si_signo = sig; - info.si_errno = 0; - info.si_code = SI_TKILL; - info.si_pid = task_tgid_vnr(current); - info.si_uid = current_uid(); + int error = -ESRCH; rcu_read_lock(); p = find_task_by_vpid(pid); if (p && (tgid <= 0 || task_tgid_vnr(p) == tgid)) { - error = check_kill_permission(sig, &info, p); + error = check_kill_permission(sig, info, p); /* * The null signal is a permissions and process existence * probe. No signal is actually delivered. @@ -2305,7 +2296,7 @@ static int do_tkill(pid_t tgid, pid_t pid, int sig) * signal is private anyway. */ if (!error && sig && lock_task_sighand(p, &flags)) { - error = specific_send_sig_info(sig, &info, p); + error = specific_send_sig_info(sig, info, p); unlock_task_sighand(p, &flags); } } @@ -2314,6 +2305,19 @@ static int do_tkill(pid_t tgid, pid_t pid, int sig) return error; } +static int do_tkill(pid_t tgid, pid_t pid, int sig) +{ + struct siginfo info; + + info.si_signo = sig; + info.si_errno = 0; + info.si_code = SI_TKILL; + info.si_pid = task_tgid_vnr(current); + info.si_uid = current_uid(); + + return do_send_specific(tgid, pid, sig, &info); +} + /** * sys_tgkill - send signal to one specific thread * @tgid: the thread group ID of the thread @@ -2363,6 +2367,32 @@ SYSCALL_DEFINE3(rt_sigqueueinfo, pid_t, pid, int, sig, return kill_proc_info(sig, &info, pid); } +long do_rt_tgsigqueueinfo(pid_t tgid, pid_t pid, int sig, siginfo_t *info) +{ + /* This is only valid for single tasks */ + if (pid <= 0 || tgid <= 0) + return -EINVAL; + + /* Not even root can pretend to send signals from the kernel. + Nor can they impersonate a kill(), which adds source info. */ + if (info->si_code >= 0) + return -EPERM; + info->si_signo = sig; + + return do_send_specific(tgid, pid, sig, info); +} + +SYSCALL_DEFINE4(rt_tgsigqueueinfo, pid_t, tgid, pid_t, pid, int, sig, + siginfo_t __user *, uinfo) +{ + siginfo_t info; + + if (copy_from_user(&info, uinfo, sizeof(siginfo_t))) + return -EFAULT; + + return do_rt_tgsigqueueinfo(tgid, pid, sig, &info); +} + int do_sigaction(int sig, struct k_sigaction *act, struct k_sigaction *oact) { struct task_struct *t = current; diff --git a/kernel/softirq.c b/kernel/softirq.c index b525dd34851..258885a543d 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -24,7 +24,9 @@ #include <linux/ftrace.h> #include <linux/smp.h> #include <linux/tick.h> -#include <trace/irq.h> + +#define CREATE_TRACE_POINTS +#include <trace/events/irq.h> #include <asm/irq.h> /* @@ -186,9 +188,6 @@ EXPORT_SYMBOL(local_bh_enable_ip); */ #define MAX_SOFTIRQ_RESTART 10 -DEFINE_TRACE(softirq_entry); -DEFINE_TRACE(softirq_exit); - asmlinkage void __do_softirq(void) { struct softirq_action *h; @@ -828,7 +827,7 @@ int __init __weak arch_early_irq_init(void) return 0; } -int __weak arch_init_chip_data(struct irq_desc *desc, int cpu) +int __weak arch_init_chip_data(struct irq_desc *desc, int node) { return 0; } diff --git a/kernel/sys.c b/kernel/sys.c index e7998cf3149..438d99a38c8 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -14,6 +14,7 @@ #include <linux/prctl.h> #include <linux/highuid.h> #include <linux/fs.h> +#include <linux/perf_counter.h> #include <linux/resource.h> #include <linux/kernel.h> #include <linux/kexec.h> @@ -1793,6 +1794,12 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3, case PR_SET_TSC: error = SET_TSC_CTL(arg2); break; + case PR_TASK_PERF_COUNTERS_DISABLE: + error = perf_counter_task_disable(); + break; + case PR_TASK_PERF_COUNTERS_ENABLE: + error = perf_counter_task_enable(); + break; case PR_GET_TIMERSLACK: error = current->timer_slack_ns; break; diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c index 27dad296738..68320f6b07b 100644 --- a/kernel/sys_ni.c +++ b/kernel/sys_ni.c @@ -175,3 +175,6 @@ cond_syscall(compat_sys_timerfd_settime); cond_syscall(compat_sys_timerfd_gettime); cond_syscall(sys_eventfd); cond_syscall(sys_eventfd2); + +/* performance counters: */ +cond_syscall(sys_perf_counter_open); diff --git a/kernel/sysctl.c b/kernel/sysctl.c index b2970d56fb7..a7e4eb0525b 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -49,6 +49,7 @@ #include <linux/reboot.h> #include <linux/ftrace.h> #include <linux/slow-work.h> +#include <linux/perf_counter.h> #include <asm/uaccess.h> #include <asm/processor.h> @@ -731,6 +732,14 @@ static struct ctl_table kern_table[] = { }, { .ctl_name = CTL_UNNUMBERED, + .procname = "bootloader_version", + .data = &bootloader_version, + .maxlen = sizeof (int), + .mode = 0444, + .proc_handler = &proc_dointvec, + }, + { + .ctl_name = CTL_UNNUMBERED, .procname = "kstack_depth_to_print", .data = &kstack_depth_to_print, .maxlen = sizeof(int), @@ -912,6 +921,32 @@ static struct ctl_table kern_table[] = { .child = slow_work_sysctls, }, #endif +#ifdef CONFIG_PERF_COUNTERS + { + .ctl_name = CTL_UNNUMBERED, + .procname = "perf_counter_paranoid", + .data = &sysctl_perf_counter_paranoid, + .maxlen = sizeof(sysctl_perf_counter_paranoid), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, + { + .ctl_name = CTL_UNNUMBERED, + .procname = "perf_counter_mlock_kb", + .data = &sysctl_perf_counter_mlock, + .maxlen = sizeof(sysctl_perf_counter_mlock), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, + { + .ctl_name = CTL_UNNUMBERED, + .procname = "perf_counter_max_sample_rate", + .data = &sysctl_perf_counter_sample_rate, + .maxlen = sizeof(sysctl_perf_counter_sample_rate), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, +#endif /* * NOTE: do not add new entries to this table unless you have read * Documentation/sysctl/ctl_unnumbered.txt diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 687dff49f6e..52a8bf8931f 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -22,7 +22,7 @@ /* * This read-write spinlock protects us from races in SMP while - * playing with xtime and avenrun. + * playing with xtime. */ __cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock); diff --git a/kernel/timer.c b/kernel/timer.c index cffffad01c3..c01e568935e 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -37,6 +37,7 @@ #include <linux/delay.h> #include <linux/tick.h> #include <linux/kallsyms.h> +#include <linux/perf_counter.h> #include <asm/uaccess.h> #include <asm/unistd.h> @@ -1123,53 +1124,14 @@ void update_process_times(int user_tick) } /* - * Nr of active tasks - counted in fixed-point numbers - */ -static unsigned long count_active_tasks(void) -{ - return nr_active() * FIXED_1; -} - -/* - * Hmm.. Changed this, as the GNU make sources (load.c) seems to - * imply that avenrun[] is the standard name for this kind of thing. - * Nothing else seems to be standardized: the fractional size etc - * all seem to differ on different machines. - * - * Requires xtime_lock to access. - */ -unsigned long avenrun[3]; - -EXPORT_SYMBOL(avenrun); - -/* - * calc_load - given tick count, update the avenrun load estimates. - * This is called while holding a write_lock on xtime_lock. - */ -static inline void calc_load(unsigned long ticks) -{ - unsigned long active_tasks; /* fixed-point */ - static int count = LOAD_FREQ; - - count -= ticks; - if (unlikely(count < 0)) { - active_tasks = count_active_tasks(); - do { - CALC_LOAD(avenrun[0], EXP_1, active_tasks); - CALC_LOAD(avenrun[1], EXP_5, active_tasks); - CALC_LOAD(avenrun[2], EXP_15, active_tasks); - count += LOAD_FREQ; - } while (count < 0); - } -} - -/* * This function runs timers and the timer-tq in bottom half context. */ static void run_timer_softirq(struct softirq_action *h) { struct tvec_base *base = __get_cpu_var(tvec_bases); + perf_counter_do_pending(); + hrtimer_run_pending(); if (time_after_eq(jiffies, base->timer_jiffies)) @@ -1187,16 +1149,6 @@ void run_local_timers(void) } /* - * Called by the timer interrupt. xtime_lock must already be taken - * by the timer IRQ! - */ -static inline void update_times(unsigned long ticks) -{ - update_wall_time(); - calc_load(ticks); -} - -/* * The 64-bit jiffies value is not atomic - you MUST NOT read it * without sampling the sequence number in xtime_lock. * jiffies is defined in the linker script... @@ -1205,7 +1157,8 @@ static inline void update_times(unsigned long ticks) void do_timer(unsigned long ticks) { jiffies_64 += ticks; - update_times(ticks); + update_wall_time(); + calc_global_load(); } #ifdef __ARCH_WANT_SYS_ALARM @@ -1406,37 +1359,17 @@ int do_sysinfo(struct sysinfo *info) { unsigned long mem_total, sav_total; unsigned int mem_unit, bitcount; - unsigned long seq; + struct timespec tp; memset(info, 0, sizeof(struct sysinfo)); - do { - struct timespec tp; - seq = read_seqbegin(&xtime_lock); - - /* - * This is annoying. The below is the same thing - * posix_get_clock_monotonic() does, but it wants to - * take the lock which we want to cover the loads stuff - * too. - */ - - getnstimeofday(&tp); - tp.tv_sec += wall_to_monotonic.tv_sec; - tp.tv_nsec += wall_to_monotonic.tv_nsec; - monotonic_to_bootbased(&tp); - if (tp.tv_nsec - NSEC_PER_SEC >= 0) { - tp.tv_nsec = tp.tv_nsec - NSEC_PER_SEC; - tp.tv_sec++; - } - info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0); + ktime_get_ts(&tp); + monotonic_to_bootbased(&tp); + info->uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0); - info->loads[0] = avenrun[0] << (SI_LOAD_SHIFT - FSHIFT); - info->loads[1] = avenrun[1] << (SI_LOAD_SHIFT - FSHIFT); - info->loads[2] = avenrun[2] << (SI_LOAD_SHIFT - FSHIFT); + get_avenrun(info->loads, 0, SI_LOAD_SHIFT - FSHIFT); - info->procs = nr_threads; - } while (read_seqretry(&xtime_lock, seq)); + info->procs = nr_threads; si_meminfo(info); si_swapinfo(info); diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 417d1985e29..4a13e5a01ce 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -48,6 +48,21 @@ config FTRACE_NMI_ENTER depends on HAVE_FTRACE_NMI_ENTER default y +config EVENT_TRACING + select CONTEXT_SWITCH_TRACER + bool + +config CONTEXT_SWITCH_TRACER + select MARKERS + bool + +# All tracer options should select GENERIC_TRACER. For those options that are +# enabled by all tracers (context switch and event tracer) they select TRACING. +# This allows those options to appear when no other tracer is selected. But the +# options do not appear when something else selects it. We need the two options +# GENERIC_TRACER and TRACING to avoid circular dependencies to accomplish the +# hidding of the automatic options options. + config TRACING bool select DEBUG_FS @@ -56,6 +71,11 @@ config TRACING select TRACEPOINTS select NOP_TRACER select BINARY_PRINTF + select EVENT_TRACING + +config GENERIC_TRACER + bool + select TRACING # # Minimum requirements an architecture has to meet for us to @@ -73,14 +93,20 @@ config TRACING_SUPPORT if TRACING_SUPPORT -menu "Tracers" +menuconfig FTRACE + bool "Tracers" + default y if DEBUG_KERNEL + help + Enable the kernel tracing infrastructure. + +if FTRACE config FUNCTION_TRACER bool "Kernel Function Tracer" depends on HAVE_FUNCTION_TRACER select FRAME_POINTER select KALLSYMS - select TRACING + select GENERIC_TRACER select CONTEXT_SWITCH_TRACER help Enable the kernel to trace every kernel function. This is done @@ -104,13 +130,14 @@ config FUNCTION_GRAPH_TRACER the return value. This is done by setting the current return address on the current task structure into a stack of calls. + config IRQSOFF_TRACER bool "Interrupts-off Latency Tracer" default n depends on TRACE_IRQFLAGS_SUPPORT depends on GENERIC_TIME select TRACE_IRQFLAGS - select TRACING + select GENERIC_TRACER select TRACER_MAX_TRACE help This option measures the time spent in irqs-off critical @@ -131,7 +158,7 @@ config PREEMPT_TRACER default n depends on GENERIC_TIME depends on PREEMPT - select TRACING + select GENERIC_TRACER select TRACER_MAX_TRACE help This option measures the time spent in preemption off critical @@ -150,7 +177,7 @@ config PREEMPT_TRACER config SYSPROF_TRACER bool "Sysprof Tracer" depends on X86 - select TRACING + select GENERIC_TRACER select CONTEXT_SWITCH_TRACER help This tracer provides the trace needed by the 'Sysprof' userspace @@ -158,40 +185,33 @@ config SYSPROF_TRACER config SCHED_TRACER bool "Scheduling Latency Tracer" - select TRACING + select GENERIC_TRACER select CONTEXT_SWITCH_TRACER select TRACER_MAX_TRACE help This tracer tracks the latency of the highest priority task to be scheduled in, starting from the point it has woken up. -config CONTEXT_SWITCH_TRACER - bool "Trace process context switches" - select TRACING - select MARKERS - help - This tracer gets called from the context switch and records - all switching of tasks. - -config EVENT_TRACER - bool "Trace various events in the kernel" +config ENABLE_DEFAULT_TRACERS + bool "Trace process context switches and events" + depends on !GENERIC_TRACER select TRACING help This tracer hooks to various trace points in the kernel allowing the user to pick and choose which trace point they - want to trace. + want to trace. It also includes the sched_switch tracer plugin. config FTRACE_SYSCALLS bool "Trace syscalls" depends on HAVE_FTRACE_SYSCALLS - select TRACING + select GENERIC_TRACER select KALLSYMS help Basic tracer to catch the syscall entry and exit events. config BOOT_TRACER bool "Trace boot initcalls" - select TRACING + select GENERIC_TRACER select CONTEXT_SWITCH_TRACER help This tracer helps developers to optimize boot times: it records @@ -207,8 +227,36 @@ config BOOT_TRACER to enable this on bootup. config TRACE_BRANCH_PROFILING + bool + select GENERIC_TRACER + +choice + prompt "Branch Profiling" + default BRANCH_PROFILE_NONE + help + The branch profiling is a software profiler. It will add hooks + into the C conditionals to test which path a branch takes. + + The likely/unlikely profiler only looks at the conditions that + are annotated with a likely or unlikely macro. + + The "all branch" profiler will profile every if statement in the + kernel. This profiler will also enable the likely/unlikely + profiler as well. + + Either of the above profilers add a bit of overhead to the system. + If unsure choose "No branch profiling". + +config BRANCH_PROFILE_NONE + bool "No branch profiling" + help + No branch profiling. Branch profiling adds a bit of overhead. + Only enable it if you want to analyse the branching behavior. + Otherwise keep it disabled. + +config PROFILE_ANNOTATED_BRANCHES bool "Trace likely/unlikely profiler" - select TRACING + select TRACE_BRANCH_PROFILING help This tracer profiles all the the likely and unlikely macros in the kernel. It will display the results in: @@ -218,11 +266,9 @@ config TRACE_BRANCH_PROFILING Note: this will add a significant overhead, only turn this on if you need to profile the system's use of these macros. - Say N if unsure. - config PROFILE_ALL_BRANCHES bool "Profile all if conditionals" - depends on TRACE_BRANCH_PROFILING + select TRACE_BRANCH_PROFILING help This tracer profiles all branch conditions. Every if () taken in the kernel is recorded whether it hit or miss. @@ -230,11 +276,12 @@ config PROFILE_ALL_BRANCHES /debugfs/tracing/profile_branch + This option also enables the likely/unlikely profiler. + This configuration, when enabled, will impose a great overhead on the system. This should only be enabled when the system is to be analyzed - - Say N if unsure. +endchoice config TRACING_BRANCHES bool @@ -261,7 +308,7 @@ config BRANCH_TRACER config POWER_TRACER bool "Trace power consumption behavior" depends on X86 - select TRACING + select GENERIC_TRACER help This tracer helps developers to analyze and optimize the kernels power management decisions, specifically the C-state and P-state @@ -295,14 +342,14 @@ config STACK_TRACER config HW_BRANCH_TRACER depends on HAVE_HW_BRANCH_TRACER bool "Trace hw branches" - select TRACING + select GENERIC_TRACER help This tracer records all branches on the system in a circular buffer giving access to the last N branches for each cpu. config KMEMTRACE bool "Trace SLAB allocations" - select TRACING + select GENERIC_TRACER help kmemtrace provides tracing for slab allocator functions, such as kmalloc, kfree, kmem_cache_alloc, kmem_cache_free etc.. Collected @@ -322,7 +369,7 @@ config KMEMTRACE config WORKQUEUE_TRACER bool "Trace workqueues" - select TRACING + select GENERIC_TRACER help The workqueue tracer provides some statistical informations about each cpu workqueue thread such as the number of the @@ -338,7 +385,7 @@ config BLK_DEV_IO_TRACE select RELAY select DEBUG_FS select TRACEPOINTS - select TRACING + select GENERIC_TRACER select STACKTRACE help Say Y here if you want to be able to trace the block layer actions @@ -375,6 +422,20 @@ config DYNAMIC_FTRACE were made. If so, it runs stop_machine (stops all CPUS) and modifies the code to jump over the call to ftrace. +config FUNCTION_PROFILER + bool "Kernel function profiler" + depends on FUNCTION_TRACER + default n + help + This option enables the kernel function profiler. A file is created + in debugfs called function_profile_enabled which defaults to zero. + When a 1 is echoed into this file profiling begins, and when a + zero is entered, profiling stops. A file in the trace_stats + directory called functions, that show the list of functions that + have been hit and their counters. + + If in doubt, say N + config FTRACE_MCOUNT_RECORD def_bool y depends on DYNAMIC_FTRACE @@ -385,7 +446,7 @@ config FTRACE_SELFTEST config FTRACE_STARTUP_TEST bool "Perform a startup test on ftrace" - depends on TRACING + depends on GENERIC_TRACER select FTRACE_SELFTEST help This option performs a series of startup tests on ftrace. On bootup @@ -396,7 +457,7 @@ config FTRACE_STARTUP_TEST config MMIOTRACE bool "Memory mapped IO tracing" depends on HAVE_MMIOTRACE_SUPPORT && PCI - select TRACING + select GENERIC_TRACER help Mmiotrace traces Memory Mapped I/O access and is meant for debugging and reverse engineering. It is called from the ioremap @@ -416,7 +477,23 @@ config MMIOTRACE_TEST Say N, unless you absolutely know what you are doing. -endmenu +config RING_BUFFER_BENCHMARK + tristate "Ring buffer benchmark stress tester" + depends on RING_BUFFER + help + This option creates a test to stress the ring buffer and bench mark it. + It creates its own ring buffer such that it will not interfer with + any other users of the ring buffer (such as ftrace). It then creates + a producer and consumer that will run for 10 seconds and sleep for + 10 seconds. Each interval it will print out the number of events + it recorded and give a rough estimate of how long each iteration took. + + It does not disable interrupts or raise its priority, so it may be + affected by processes that are running. + + If unsure, say N + +endif # FTRACE endif # TRACING_SUPPORT diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index 2630f5121ec..844164dca90 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -15,11 +15,17 @@ ifdef CONFIG_TRACING_BRANCHES KBUILD_CFLAGS += -DDISABLE_BRANCH_PROFILING endif +# +# Make the trace clocks available generally: it's infrastructure +# relied on by ptrace for example: +# +obj-y += trace_clock.o + obj-$(CONFIG_FUNCTION_TRACER) += libftrace.o obj-$(CONFIG_RING_BUFFER) += ring_buffer.o +obj-$(CONFIG_RING_BUFFER_BENCHMARK) += ring_buffer_benchmark.o obj-$(CONFIG_TRACING) += trace.o -obj-$(CONFIG_TRACING) += trace_clock.o obj-$(CONFIG_TRACING) += trace_output.o obj-$(CONFIG_TRACING) += trace_stat.o obj-$(CONFIG_TRACING) += trace_printk.o @@ -39,12 +45,14 @@ obj-$(CONFIG_HW_BRANCH_TRACER) += trace_hw_branches.o obj-$(CONFIG_POWER_TRACER) += trace_power.o obj-$(CONFIG_KMEMTRACE) += kmemtrace.o obj-$(CONFIG_WORKQUEUE_TRACER) += trace_workqueue.o -obj-$(CONFIG_BLK_DEV_IO_TRACE) += blktrace.o -obj-$(CONFIG_EVENT_TRACER) += trace_events.o -obj-$(CONFIG_EVENT_TRACER) += events.o -obj-$(CONFIG_EVENT_TRACER) += trace_export.o +obj-$(CONFIG_BLK_DEV_IO_TRACE) += blktrace.o +ifeq ($(CONFIG_BLOCK),y) +obj-$(CONFIG_EVENT_TRACING) += blktrace.o +endif +obj-$(CONFIG_EVENT_TRACING) += trace_events.o +obj-$(CONFIG_EVENT_TRACING) += trace_export.o obj-$(CONFIG_FTRACE_SYSCALLS) += trace_syscalls.o obj-$(CONFIG_EVENT_PROFILE) += trace_event_profile.o -obj-$(CONFIG_EVENT_TRACER) += trace_events_filter.o +obj-$(CONFIG_EVENT_TRACING) += trace_events_filter.o libftrace-y := ftrace.o diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 921ef5d1f0b..7bd6a9893c2 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -23,10 +23,14 @@ #include <linux/mutex.h> #include <linux/debugfs.h> #include <linux/time.h> -#include <trace/block.h> #include <linux/uaccess.h> + +#include <trace/events/block.h> + #include "trace_output.h" +#ifdef CONFIG_BLK_DEV_IO_TRACE + static unsigned int blktrace_seq __read_mostly = 1; static struct trace_array *blk_tr; @@ -147,7 +151,7 @@ static int act_log_check(struct blk_trace *bt, u32 what, sector_t sector, { if (((bt->act_mask << BLK_TC_SHIFT) & what) == 0) return 1; - if (sector < bt->start_lba || sector > bt->end_lba) + if (sector && (sector < bt->start_lba || sector > bt->end_lba)) return 1; if (bt->pid && pid != bt->pid) return 1; @@ -192,7 +196,7 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes, what |= MASK_TC_BIT(rw, DISCARD); pid = tsk->pid; - if (unlikely(act_log_check(bt, what, sector, pid))) + if (act_log_check(bt, what, sector, pid)) return; cpu = raw_smp_processor_id(); @@ -262,6 +266,7 @@ static void blk_trace_free(struct blk_trace *bt) { debugfs_remove(bt->msg_file); debugfs_remove(bt->dropped_file); + debugfs_remove(bt->dir); relay_close(bt->rchan); free_percpu(bt->sequence); free_percpu(bt->msg_data); @@ -403,11 +408,29 @@ static struct rchan_callbacks blk_relay_callbacks = { .remove_buf_file = blk_remove_buf_file_callback, }; +static void blk_trace_setup_lba(struct blk_trace *bt, + struct block_device *bdev) +{ + struct hd_struct *part = NULL; + + if (bdev) + part = bdev->bd_part; + + if (part) { + bt->start_lba = part->start_sect; + bt->end_lba = part->start_sect + part->nr_sects; + } else { + bt->start_lba = 0; + bt->end_lba = -1ULL; + } +} + /* * Setup everything required to start tracing */ int do_blk_trace_setup(struct request_queue *q, char *name, dev_t dev, - struct blk_user_trace_setup *buts) + struct block_device *bdev, + struct blk_user_trace_setup *buts) { struct blk_trace *old_bt, *bt = NULL; struct dentry *dir = NULL; @@ -480,10 +503,13 @@ int do_blk_trace_setup(struct request_queue *q, char *name, dev_t dev, if (!bt->act_mask) bt->act_mask = (u16) -1; - bt->start_lba = buts->start_lba; - bt->end_lba = buts->end_lba; - if (!bt->end_lba) - bt->end_lba = -1ULL; + blk_trace_setup_lba(bt, bdev); + + /* overwrite with user settings */ + if (buts->start_lba) + bt->start_lba = buts->start_lba; + if (buts->end_lba) + bt->end_lba = buts->end_lba; bt->pid = buts->pid; bt->trace_state = Blktrace_setup; @@ -505,6 +531,7 @@ err: } int blk_trace_setup(struct request_queue *q, char *name, dev_t dev, + struct block_device *bdev, char __user *arg) { struct blk_user_trace_setup buts; @@ -514,7 +541,7 @@ int blk_trace_setup(struct request_queue *q, char *name, dev_t dev, if (ret) return -EFAULT; - ret = do_blk_trace_setup(q, name, dev, &buts); + ret = do_blk_trace_setup(q, name, dev, bdev, &buts); if (ret) return ret; @@ -582,7 +609,7 @@ int blk_trace_ioctl(struct block_device *bdev, unsigned cmd, char __user *arg) switch (cmd) { case BLKTRACESETUP: bdevname(bdev, b); - ret = blk_trace_setup(q, b, bdev->bd_dev, arg); + ret = blk_trace_setup(q, b, bdev->bd_dev, bdev, arg); break; case BLKTRACESTART: start = 1; @@ -809,7 +836,6 @@ static void blk_add_trace_split(struct request_queue *q, struct bio *bio, * @bio: the source bio * @dev: target device * @from: source sector - * @to: target sector * * Description: * Device mapper or raid target sometimes need to split a bio because @@ -817,7 +843,7 @@ static void blk_add_trace_split(struct request_queue *q, struct bio *bio, * **/ static void blk_add_trace_remap(struct request_queue *q, struct bio *bio, - dev_t dev, sector_t from, sector_t to) + dev_t dev, sector_t from) { struct blk_trace *bt = q->blk_trace; struct blk_io_trace_remap r; @@ -825,12 +851,13 @@ static void blk_add_trace_remap(struct request_queue *q, struct bio *bio, if (likely(!bt)) return; - r.device = cpu_to_be32(dev); - r.device_from = cpu_to_be32(bio->bi_bdev->bd_dev); - r.sector = cpu_to_be64(to); + r.device_from = cpu_to_be32(dev); + r.device_to = cpu_to_be32(bio->bi_bdev->bd_dev); + r.sector_from = cpu_to_be64(from); - __blk_add_trace(bt, from, bio->bi_size, bio->bi_rw, BLK_TA_REMAP, - !bio_flagged(bio, BIO_UPTODATE), sizeof(r), &r); + __blk_add_trace(bt, bio->bi_sector, bio->bi_size, bio->bi_rw, + BLK_TA_REMAP, !bio_flagged(bio, BIO_UPTODATE), + sizeof(r), &r); } /** @@ -971,6 +998,16 @@ static inline const void *pdu_start(const struct trace_entry *ent) return te_blk_io_trace(ent) + 1; } +static inline u32 t_action(const struct trace_entry *ent) +{ + return te_blk_io_trace(ent)->action; +} + +static inline u32 t_bytes(const struct trace_entry *ent) +{ + return te_blk_io_trace(ent)->bytes; +} + static inline u32 t_sec(const struct trace_entry *ent) { return te_blk_io_trace(ent)->bytes >> 9; @@ -996,11 +1033,11 @@ static void get_pdu_remap(const struct trace_entry *ent, struct blk_io_trace_remap *r) { const struct blk_io_trace_remap *__r = pdu_start(ent); - __u64 sector = __r->sector; + __u64 sector_from = __r->sector_from; - r->device = be32_to_cpu(__r->device); r->device_from = be32_to_cpu(__r->device_from); - r->sector = be64_to_cpu(sector); + r->device_to = be32_to_cpu(__r->device_to); + r->sector_from = be64_to_cpu(sector_from); } typedef int (blk_log_action_t) (struct trace_iterator *iter, const char *act); @@ -1031,36 +1068,98 @@ static int blk_log_action(struct trace_iterator *iter, const char *act) MAJOR(t->device), MINOR(t->device), act, rwbs); } +static int blk_log_dump_pdu(struct trace_seq *s, const struct trace_entry *ent) +{ + const unsigned char *pdu_buf; + int pdu_len; + int i, end, ret; + + pdu_buf = pdu_start(ent); + pdu_len = te_blk_io_trace(ent)->pdu_len; + + if (!pdu_len) + return 1; + + /* find the last zero that needs to be printed */ + for (end = pdu_len - 1; end >= 0; end--) + if (pdu_buf[end]) + break; + end++; + + if (!trace_seq_putc(s, '(')) + return 0; + + for (i = 0; i < pdu_len; i++) { + + ret = trace_seq_printf(s, "%s%02x", + i == 0 ? "" : " ", pdu_buf[i]); + if (!ret) + return ret; + + /* + * stop when the rest is just zeroes and indicate so + * with a ".." appended + */ + if (i == end && end != pdu_len - 1) + return trace_seq_puts(s, " ..) "); + } + + return trace_seq_puts(s, ") "); +} + static int blk_log_generic(struct trace_seq *s, const struct trace_entry *ent) { char cmd[TASK_COMM_LEN]; trace_find_cmdline(ent->pid, cmd); - if (t_sec(ent)) - return trace_seq_printf(s, "%llu + %u [%s]\n", - t_sector(ent), t_sec(ent), cmd); - return trace_seq_printf(s, "[%s]\n", cmd); + if (t_action(ent) & BLK_TC_ACT(BLK_TC_PC)) { + int ret; + + ret = trace_seq_printf(s, "%u ", t_bytes(ent)); + if (!ret) + return 0; + ret = blk_log_dump_pdu(s, ent); + if (!ret) + return 0; + return trace_seq_printf(s, "[%s]\n", cmd); + } else { + if (t_sec(ent)) + return trace_seq_printf(s, "%llu + %u [%s]\n", + t_sector(ent), t_sec(ent), cmd); + return trace_seq_printf(s, "[%s]\n", cmd); + } } static int blk_log_with_error(struct trace_seq *s, const struct trace_entry *ent) { - if (t_sec(ent)) - return trace_seq_printf(s, "%llu + %u [%d]\n", t_sector(ent), - t_sec(ent), t_error(ent)); - return trace_seq_printf(s, "%llu [%d]\n", t_sector(ent), t_error(ent)); + if (t_action(ent) & BLK_TC_ACT(BLK_TC_PC)) { + int ret; + + ret = blk_log_dump_pdu(s, ent); + if (ret) + return trace_seq_printf(s, "[%d]\n", t_error(ent)); + return 0; + } else { + if (t_sec(ent)) + return trace_seq_printf(s, "%llu + %u [%d]\n", + t_sector(ent), + t_sec(ent), t_error(ent)); + return trace_seq_printf(s, "%llu [%d]\n", + t_sector(ent), t_error(ent)); + } } static int blk_log_remap(struct trace_seq *s, const struct trace_entry *ent) { - struct blk_io_trace_remap r = { .device = 0, }; + struct blk_io_trace_remap r = { .device_from = 0, }; get_pdu_remap(ent, &r); return trace_seq_printf(s, "%llu + %u <- (%d,%d) %llu\n", - t_sector(ent), - t_sec(ent), MAJOR(r.device), MINOR(r.device), - (unsigned long long)r.sector); + t_sector(ent), t_sec(ent), + MAJOR(r.device_from), MINOR(r.device_from), + (unsigned long long)r.sector_from); } static int blk_log_plug(struct trace_seq *s, const struct trace_entry *ent) @@ -1117,7 +1216,6 @@ static void blk_tracer_print_header(struct seq_file *m) static void blk_tracer_start(struct trace_array *tr) { blk_tracer_enabled = true; - trace_flags &= ~TRACE_ITER_CONTEXT_INFO; } static int blk_tracer_init(struct trace_array *tr) @@ -1130,7 +1228,6 @@ static int blk_tracer_init(struct trace_array *tr) static void blk_tracer_stop(struct trace_array *tr) { blk_tracer_enabled = false; - trace_flags |= TRACE_ITER_CONTEXT_INFO; } static void blk_tracer_reset(struct trace_array *tr) @@ -1182,7 +1279,7 @@ static enum print_line_t print_one_line(struct trace_iterator *iter, } if (unlikely(what == 0 || what >= ARRAY_SIZE(what2act))) - ret = trace_seq_printf(s, "Bad pc action %x\n", what); + ret = trace_seq_printf(s, "Unknown action %x\n", what); else { ret = log_action(iter, what2act[what].act[long_act]); if (ret) @@ -1195,9 +1292,6 @@ out: static enum print_line_t blk_trace_event_print(struct trace_iterator *iter, int flags) { - if (!trace_print_context(iter)) - return TRACE_TYPE_PARTIAL_LINE; - return print_one_line(iter, false); } @@ -1232,6 +1326,18 @@ static enum print_line_t blk_tracer_print_line(struct trace_iterator *iter) return print_one_line(iter, true); } +static int blk_tracer_set_flag(u32 old_flags, u32 bit, int set) +{ + /* don't output context-info for blk_classic output */ + if (bit == TRACE_BLK_OPT_CLASSIC) { + if (set) + trace_flags &= ~TRACE_ITER_CONTEXT_INFO; + else + trace_flags |= TRACE_ITER_CONTEXT_INFO; + } + return 0; +} + static struct tracer blk_tracer __read_mostly = { .name = "blk", .init = blk_tracer_init, @@ -1241,6 +1347,7 @@ static struct tracer blk_tracer __read_mostly = { .print_header = blk_tracer_print_header, .print_line = blk_tracer_print_line, .flags = &blk_tracer_flags, + .set_flag = blk_tracer_set_flag, }; static struct trace_event trace_blk_event = { @@ -1285,7 +1392,8 @@ static int blk_trace_remove_queue(struct request_queue *q) /* * Setup everything required to start tracing */ -static int blk_trace_setup_queue(struct request_queue *q, dev_t dev) +static int blk_trace_setup_queue(struct request_queue *q, + struct block_device *bdev) { struct blk_trace *old_bt, *bt = NULL; int ret = -ENOMEM; @@ -1298,9 +1406,10 @@ static int blk_trace_setup_queue(struct request_queue *q, dev_t dev) if (!bt->msg_data) goto free_bt; - bt->dev = dev; + bt->dev = bdev->bd_dev; bt->act_mask = (u16)-1; - bt->end_lba = -1ULL; + + blk_trace_setup_lba(bt, bdev); old_bt = xchg(&q->blk_trace, bt); if (old_bt != NULL) { @@ -1517,7 +1626,7 @@ static ssize_t sysfs_blk_trace_attr_store(struct device *dev, if (attr == &dev_attr_enable) { if (value) - ret = blk_trace_setup_queue(q, bdev->bd_dev); + ret = blk_trace_setup_queue(q, bdev); else ret = blk_trace_remove_queue(q); goto out_unlock_bdev; @@ -1525,7 +1634,7 @@ static ssize_t sysfs_blk_trace_attr_store(struct device *dev, ret = 0; if (q->blk_trace == NULL) - ret = blk_trace_setup_queue(q, bdev->bd_dev); + ret = blk_trace_setup_queue(q, bdev); if (ret == 0) { if (attr == &dev_attr_act_mask) @@ -1548,3 +1657,80 @@ out: return ret ? ret : count; } +int blk_trace_init_sysfs(struct device *dev) +{ + return sysfs_create_group(&dev->kobj, &blk_trace_attr_group); +} + +#endif /* CONFIG_BLK_DEV_IO_TRACE */ + +#ifdef CONFIG_EVENT_TRACING + +void blk_dump_cmd(char *buf, struct request *rq) +{ + int i, end; + int len = rq->cmd_len; + unsigned char *cmd = rq->cmd; + + if (!blk_pc_request(rq)) { + buf[0] = '\0'; + return; + } + + for (end = len - 1; end >= 0; end--) + if (cmd[end]) + break; + end++; + + for (i = 0; i < len; i++) { + buf += sprintf(buf, "%s%02x", i == 0 ? "" : " ", cmd[i]); + if (i == end && end != len - 1) { + sprintf(buf, " .."); + break; + } + } +} + +void blk_fill_rwbs(char *rwbs, u32 rw, int bytes) +{ + int i = 0; + + if (rw & WRITE) + rwbs[i++] = 'W'; + else if (rw & 1 << BIO_RW_DISCARD) + rwbs[i++] = 'D'; + else if (bytes) + rwbs[i++] = 'R'; + else + rwbs[i++] = 'N'; + + if (rw & 1 << BIO_RW_AHEAD) + rwbs[i++] = 'A'; + if (rw & 1 << BIO_RW_BARRIER) + rwbs[i++] = 'B'; + if (rw & 1 << BIO_RW_SYNCIO) + rwbs[i++] = 'S'; + if (rw & 1 << BIO_RW_META) + rwbs[i++] = 'M'; + + rwbs[i] = '\0'; +} + +void blk_fill_rwbs_rq(char *rwbs, struct request *rq) +{ + int rw = rq->cmd_flags & 0x03; + int bytes; + + if (blk_discard_rq(rq)) + rw |= (1 << BIO_RW_DISCARD); + + if (blk_pc_request(rq)) + bytes = rq->data_len; + else + bytes = rq->hard_nr_sectors << 9; + + blk_fill_rwbs(rwbs, rw, bytes); +} + +#endif /* CONFIG_EVENT_TRACING */ + diff --git a/kernel/trace/events.c b/kernel/trace/events.c deleted file mode 100644 index 246f2aa6dc4..00000000000 --- a/kernel/trace/events.c +++ /dev/null @@ -1,14 +0,0 @@ -/* - * This is the place to register all trace points as events. - */ - -#include <linux/stringify.h> - -#include <trace/trace_events.h> - -#include "trace_output.h" - -#include "trace_events_stage_1.h" -#include "trace_events_stage_2.h" -#include "trace_events_stage_3.h" - diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index f1ed080406c..bb60732ade0 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -29,11 +29,13 @@ #include <linux/list.h> #include <linux/hash.h> -#include <trace/sched.h> +#include <trace/events/sched.h> #include <asm/ftrace.h> +#include <asm/setup.h> -#include "trace.h" +#include "trace_output.h" +#include "trace_stat.h" #define FTRACE_WARN_ON(cond) \ do { \ @@ -68,7 +70,7 @@ static DEFINE_MUTEX(ftrace_lock); static struct ftrace_ops ftrace_list_end __read_mostly = { - .func = ftrace_stub, + .func = ftrace_stub, }; static struct ftrace_ops *ftrace_list __read_mostly = &ftrace_list_end; @@ -240,6 +242,580 @@ static void ftrace_update_pid_func(void) #endif } +#ifdef CONFIG_FUNCTION_PROFILER +struct ftrace_profile { + struct hlist_node node; + unsigned long ip; + unsigned long counter; +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + unsigned long long time; +#endif +}; + +struct ftrace_profile_page { + struct ftrace_profile_page *next; + unsigned long index; + struct ftrace_profile records[]; +}; + +struct ftrace_profile_stat { + atomic_t disabled; + struct hlist_head *hash; + struct ftrace_profile_page *pages; + struct ftrace_profile_page *start; + struct tracer_stat stat; +}; + +#define PROFILE_RECORDS_SIZE \ + (PAGE_SIZE - offsetof(struct ftrace_profile_page, records)) + +#define PROFILES_PER_PAGE \ + (PROFILE_RECORDS_SIZE / sizeof(struct ftrace_profile)) + +static int ftrace_profile_bits __read_mostly; +static int ftrace_profile_enabled __read_mostly; + +/* ftrace_profile_lock - synchronize the enable and disable of the profiler */ +static DEFINE_MUTEX(ftrace_profile_lock); + +static DEFINE_PER_CPU(struct ftrace_profile_stat, ftrace_profile_stats); + +#define FTRACE_PROFILE_HASH_SIZE 1024 /* must be power of 2 */ + +static void * +function_stat_next(void *v, int idx) +{ + struct ftrace_profile *rec = v; + struct ftrace_profile_page *pg; + + pg = (struct ftrace_profile_page *)((unsigned long)rec & PAGE_MASK); + + again: + rec++; + if ((void *)rec >= (void *)&pg->records[pg->index]) { + pg = pg->next; + if (!pg) + return NULL; + rec = &pg->records[0]; + if (!rec->counter) + goto again; + } + + return rec; +} + +static void *function_stat_start(struct tracer_stat *trace) +{ + struct ftrace_profile_stat *stat = + container_of(trace, struct ftrace_profile_stat, stat); + + if (!stat || !stat->start) + return NULL; + + return function_stat_next(&stat->start->records[0], 0); +} + +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +/* function graph compares on total time */ +static int function_stat_cmp(void *p1, void *p2) +{ + struct ftrace_profile *a = p1; + struct ftrace_profile *b = p2; + + if (a->time < b->time) + return -1; + if (a->time > b->time) + return 1; + else + return 0; +} +#else +/* not function graph compares against hits */ +static int function_stat_cmp(void *p1, void *p2) +{ + struct ftrace_profile *a = p1; + struct ftrace_profile *b = p2; + + if (a->counter < b->counter) + return -1; + if (a->counter > b->counter) + return 1; + else + return 0; +} +#endif + +static int function_stat_headers(struct seq_file *m) +{ +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + seq_printf(m, " Function " + "Hit Time Avg\n" + " -------- " + "--- ---- ---\n"); +#else + seq_printf(m, " Function Hit\n" + " -------- ---\n"); +#endif + return 0; +} + +static int function_stat_show(struct seq_file *m, void *v) +{ + struct ftrace_profile *rec = v; + char str[KSYM_SYMBOL_LEN]; +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + static DEFINE_MUTEX(mutex); + static struct trace_seq s; + unsigned long long avg; +#endif + + kallsyms_lookup(rec->ip, NULL, NULL, NULL, str); + seq_printf(m, " %-30.30s %10lu", str, rec->counter); + +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + seq_printf(m, " "); + avg = rec->time; + do_div(avg, rec->counter); + + mutex_lock(&mutex); + trace_seq_init(&s); + trace_print_graph_duration(rec->time, &s); + trace_seq_puts(&s, " "); + trace_print_graph_duration(avg, &s); + trace_print_seq(m, &s); + mutex_unlock(&mutex); +#endif + seq_putc(m, '\n'); + + return 0; +} + +static void ftrace_profile_reset(struct ftrace_profile_stat *stat) +{ + struct ftrace_profile_page *pg; + + pg = stat->pages = stat->start; + + while (pg) { + memset(pg->records, 0, PROFILE_RECORDS_SIZE); + pg->index = 0; + pg = pg->next; + } + + memset(stat->hash, 0, + FTRACE_PROFILE_HASH_SIZE * sizeof(struct hlist_head)); +} + +int ftrace_profile_pages_init(struct ftrace_profile_stat *stat) +{ + struct ftrace_profile_page *pg; + int functions; + int pages; + int i; + + /* If we already allocated, do nothing */ + if (stat->pages) + return 0; + + stat->pages = (void *)get_zeroed_page(GFP_KERNEL); + if (!stat->pages) + return -ENOMEM; + +#ifdef CONFIG_DYNAMIC_FTRACE + functions = ftrace_update_tot_cnt; +#else + /* + * We do not know the number of functions that exist because + * dynamic tracing is what counts them. With past experience + * we have around 20K functions. That should be more than enough. + * It is highly unlikely we will execute every function in + * the kernel. + */ + functions = 20000; +#endif + + pg = stat->start = stat->pages; + + pages = DIV_ROUND_UP(functions, PROFILES_PER_PAGE); + + for (i = 0; i < pages; i++) { + pg->next = (void *)get_zeroed_page(GFP_KERNEL); + if (!pg->next) + goto out_free; + pg = pg->next; + } + + return 0; + + out_free: + pg = stat->start; + while (pg) { + unsigned long tmp = (unsigned long)pg; + + pg = pg->next; + free_page(tmp); + } + + free_page((unsigned long)stat->pages); + stat->pages = NULL; + stat->start = NULL; + + return -ENOMEM; +} + +static int ftrace_profile_init_cpu(int cpu) +{ + struct ftrace_profile_stat *stat; + int size; + + stat = &per_cpu(ftrace_profile_stats, cpu); + + if (stat->hash) { + /* If the profile is already created, simply reset it */ + ftrace_profile_reset(stat); + return 0; + } + + /* + * We are profiling all functions, but usually only a few thousand + * functions are hit. We'll make a hash of 1024 items. + */ + size = FTRACE_PROFILE_HASH_SIZE; + + stat->hash = kzalloc(sizeof(struct hlist_head) * size, GFP_KERNEL); + + if (!stat->hash) + return -ENOMEM; + + if (!ftrace_profile_bits) { + size--; + + for (; size; size >>= 1) + ftrace_profile_bits++; + } + + /* Preallocate the function profiling pages */ + if (ftrace_profile_pages_init(stat) < 0) { + kfree(stat->hash); + stat->hash = NULL; + return -ENOMEM; + } + + return 0; +} + +static int ftrace_profile_init(void) +{ + int cpu; + int ret = 0; + + for_each_online_cpu(cpu) { + ret = ftrace_profile_init_cpu(cpu); + if (ret) + break; + } + + return ret; +} + +/* interrupts must be disabled */ +static struct ftrace_profile * +ftrace_find_profiled_func(struct ftrace_profile_stat *stat, unsigned long ip) +{ + struct ftrace_profile *rec; + struct hlist_head *hhd; + struct hlist_node *n; + unsigned long key; + + key = hash_long(ip, ftrace_profile_bits); + hhd = &stat->hash[key]; + + if (hlist_empty(hhd)) + return NULL; + + hlist_for_each_entry_rcu(rec, n, hhd, node) { + if (rec->ip == ip) + return rec; + } + + return NULL; +} + +static void ftrace_add_profile(struct ftrace_profile_stat *stat, + struct ftrace_profile *rec) +{ + unsigned long key; + + key = hash_long(rec->ip, ftrace_profile_bits); + hlist_add_head_rcu(&rec->node, &stat->hash[key]); +} + +/* + * The memory is already allocated, this simply finds a new record to use. + */ +static struct ftrace_profile * +ftrace_profile_alloc(struct ftrace_profile_stat *stat, unsigned long ip) +{ + struct ftrace_profile *rec = NULL; + + /* prevent recursion (from NMIs) */ + if (atomic_inc_return(&stat->disabled) != 1) + goto out; + + /* + * Try to find the function again since an NMI + * could have added it + */ + rec = ftrace_find_profiled_func(stat, ip); + if (rec) + goto out; + + if (stat->pages->index == PROFILES_PER_PAGE) { + if (!stat->pages->next) + goto out; + stat->pages = stat->pages->next; + } + + rec = &stat->pages->records[stat->pages->index++]; + rec->ip = ip; + ftrace_add_profile(stat, rec); + + out: + atomic_dec(&stat->disabled); + + return rec; +} + +static void +function_profile_call(unsigned long ip, unsigned long parent_ip) +{ + struct ftrace_profile_stat *stat; + struct ftrace_profile *rec; + unsigned long flags; + + if (!ftrace_profile_enabled) + return; + + local_irq_save(flags); + + stat = &__get_cpu_var(ftrace_profile_stats); + if (!stat->hash || !ftrace_profile_enabled) + goto out; + + rec = ftrace_find_profiled_func(stat, ip); + if (!rec) { + rec = ftrace_profile_alloc(stat, ip); + if (!rec) + goto out; + } + + rec->counter++; + out: + local_irq_restore(flags); +} + +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +static int profile_graph_entry(struct ftrace_graph_ent *trace) +{ + function_profile_call(trace->func, 0); + return 1; +} + +static void profile_graph_return(struct ftrace_graph_ret *trace) +{ + struct ftrace_profile_stat *stat; + unsigned long long calltime; + struct ftrace_profile *rec; + unsigned long flags; + + local_irq_save(flags); + stat = &__get_cpu_var(ftrace_profile_stats); + if (!stat->hash || !ftrace_profile_enabled) + goto out; + + calltime = trace->rettime - trace->calltime; + + if (!(trace_flags & TRACE_ITER_GRAPH_TIME)) { + int index; + + index = trace->depth; + + /* Append this call time to the parent time to subtract */ + if (index) + current->ret_stack[index - 1].subtime += calltime; + + if (current->ret_stack[index].subtime < calltime) + calltime -= current->ret_stack[index].subtime; + else + calltime = 0; + } + + rec = ftrace_find_profiled_func(stat, trace->func); + if (rec) + rec->time += calltime; + + out: + local_irq_restore(flags); +} + +static int register_ftrace_profiler(void) +{ + return register_ftrace_graph(&profile_graph_return, + &profile_graph_entry); +} + +static void unregister_ftrace_profiler(void) +{ + unregister_ftrace_graph(); +} +#else +static struct ftrace_ops ftrace_profile_ops __read_mostly = +{ + .func = function_profile_call, +}; + +static int register_ftrace_profiler(void) +{ + return register_ftrace_function(&ftrace_profile_ops); +} + +static void unregister_ftrace_profiler(void) +{ + unregister_ftrace_function(&ftrace_profile_ops); +} +#endif /* CONFIG_FUNCTION_GRAPH_TRACER */ + +static ssize_t +ftrace_profile_write(struct file *filp, const char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + unsigned long val; + char buf[64]; /* big enough to hold a number */ + int ret; + + if (cnt >= sizeof(buf)) + return -EINVAL; + + if (copy_from_user(&buf, ubuf, cnt)) + return -EFAULT; + + buf[cnt] = 0; + + ret = strict_strtoul(buf, 10, &val); + if (ret < 0) + return ret; + + val = !!val; + + mutex_lock(&ftrace_profile_lock); + if (ftrace_profile_enabled ^ val) { + if (val) { + ret = ftrace_profile_init(); + if (ret < 0) { + cnt = ret; + goto out; + } + + ret = register_ftrace_profiler(); + if (ret < 0) { + cnt = ret; + goto out; + } + ftrace_profile_enabled = 1; + } else { + ftrace_profile_enabled = 0; + /* + * unregister_ftrace_profiler calls stop_machine + * so this acts like an synchronize_sched. + */ + unregister_ftrace_profiler(); + } + } + out: + mutex_unlock(&ftrace_profile_lock); + + filp->f_pos += cnt; + + return cnt; +} + +static ssize_t +ftrace_profile_read(struct file *filp, char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + char buf[64]; /* big enough to hold a number */ + int r; + + r = sprintf(buf, "%u\n", ftrace_profile_enabled); + return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); +} + +static const struct file_operations ftrace_profile_fops = { + .open = tracing_open_generic, + .read = ftrace_profile_read, + .write = ftrace_profile_write, +}; + +/* used to initialize the real stat files */ +static struct tracer_stat function_stats __initdata = { + .name = "functions", + .stat_start = function_stat_start, + .stat_next = function_stat_next, + .stat_cmp = function_stat_cmp, + .stat_headers = function_stat_headers, + .stat_show = function_stat_show +}; + +static void ftrace_profile_debugfs(struct dentry *d_tracer) +{ + struct ftrace_profile_stat *stat; + struct dentry *entry; + char *name; + int ret; + int cpu; + + for_each_possible_cpu(cpu) { + stat = &per_cpu(ftrace_profile_stats, cpu); + + /* allocate enough for function name + cpu number */ + name = kmalloc(32, GFP_KERNEL); + if (!name) { + /* + * The files created are permanent, if something happens + * we still do not free memory. + */ + kfree(stat); + WARN(1, + "Could not allocate stat file for cpu %d\n", + cpu); + return; + } + stat->stat = function_stats; + snprintf(name, 32, "function%d", cpu); + stat->stat.name = name; + ret = register_stat_tracer(&stat->stat); + if (ret) { + WARN(1, + "Could not register function stat for cpu %d\n", + cpu); + kfree(name); + return; + } + } + + entry = debugfs_create_file("function_profile_enabled", 0644, + d_tracer, NULL, &ftrace_profile_fops); + if (!entry) + pr_warning("Could not create debugfs " + "'function_profile_enabled' entry\n"); +} + +#else /* CONFIG_FUNCTION_PROFILER */ +static void ftrace_profile_debugfs(struct dentry *d_tracer) +{ +} +#endif /* CONFIG_FUNCTION_PROFILER */ + /* set when tracing only a pid */ struct pid *ftrace_pid_trace; static struct pid * const ftrace_swapper_pid = &init_struct_pid; @@ -261,7 +837,6 @@ struct ftrace_func_probe { struct rcu_head rcu; }; - enum { FTRACE_ENABLE_CALLS = (1 << 0), FTRACE_DISABLE_CALLS = (1 << 1), @@ -346,30 +921,6 @@ static void ftrace_free_rec(struct dyn_ftrace *rec) rec->flags |= FTRACE_FL_FREE; } -void ftrace_release(void *start, unsigned long size) -{ - struct dyn_ftrace *rec; - struct ftrace_page *pg; - unsigned long s = (unsigned long)start; - unsigned long e = s + size; - - if (ftrace_disabled || !start) - return; - - mutex_lock(&ftrace_lock); - do_for_each_ftrace_rec(pg, rec) { - if ((rec->ip >= s) && (rec->ip < e)) { - /* - * rec->ip is changed in ftrace_free_rec() - * It should not between s and e if record was freed. - */ - FTRACE_WARN_ON(rec->flags & FTRACE_FL_FREE); - ftrace_free_rec(rec); - } - } while_for_each_ftrace_rec(); - mutex_unlock(&ftrace_lock); -} - static struct dyn_ftrace *ftrace_alloc_dyn_node(unsigned long ip) { struct dyn_ftrace *rec; @@ -1408,7 +1959,7 @@ function_trace_probe_call(unsigned long ip, unsigned long parent_ip) static struct ftrace_ops trace_probe_ops __read_mostly = { - .func = function_trace_probe_call, + .func = function_trace_probe_call, }; static int ftrace_probe_registered; @@ -1823,6 +2374,45 @@ void ftrace_set_notrace(unsigned char *buf, int len, int reset) ftrace_set_regex(buf, len, reset, 0); } +/* + * command line interface to allow users to set filters on boot up. + */ +#define FTRACE_FILTER_SIZE COMMAND_LINE_SIZE +static char ftrace_notrace_buf[FTRACE_FILTER_SIZE] __initdata; +static char ftrace_filter_buf[FTRACE_FILTER_SIZE] __initdata; + +static int __init set_ftrace_notrace(char *str) +{ + strncpy(ftrace_notrace_buf, str, FTRACE_FILTER_SIZE); + return 1; +} +__setup("ftrace_notrace=", set_ftrace_notrace); + +static int __init set_ftrace_filter(char *str) +{ + strncpy(ftrace_filter_buf, str, FTRACE_FILTER_SIZE); + return 1; +} +__setup("ftrace_filter=", set_ftrace_filter); + +static void __init set_ftrace_early_filter(char *buf, int enable) +{ + char *func; + + while (buf) { + func = strsep(&buf, ","); + ftrace_set_regex(func, strlen(func), 0, enable); + } +} + +static void __init set_ftrace_early_filters(void) +{ + if (ftrace_filter_buf[0]) + set_ftrace_early_filter(ftrace_filter_buf, 1); + if (ftrace_notrace_buf[0]) + set_ftrace_early_filter(ftrace_notrace_buf, 0); +} + static int ftrace_regex_release(struct inode *inode, struct file *file, int enable) { @@ -2128,38 +2718,23 @@ static const struct file_operations ftrace_graph_fops = { static __init int ftrace_init_dyn_debugfs(struct dentry *d_tracer) { - struct dentry *entry; - entry = debugfs_create_file("available_filter_functions", 0444, - d_tracer, NULL, &ftrace_avail_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'available_filter_functions' entry\n"); + trace_create_file("available_filter_functions", 0444, + d_tracer, NULL, &ftrace_avail_fops); - entry = debugfs_create_file("failures", 0444, - d_tracer, NULL, &ftrace_failures_fops); - if (!entry) - pr_warning("Could not create debugfs 'failures' entry\n"); + trace_create_file("failures", 0444, + d_tracer, NULL, &ftrace_failures_fops); - entry = debugfs_create_file("set_ftrace_filter", 0644, d_tracer, - NULL, &ftrace_filter_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'set_ftrace_filter' entry\n"); + trace_create_file("set_ftrace_filter", 0644, d_tracer, + NULL, &ftrace_filter_fops); - entry = debugfs_create_file("set_ftrace_notrace", 0644, d_tracer, + trace_create_file("set_ftrace_notrace", 0644, d_tracer, NULL, &ftrace_notrace_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'set_ftrace_notrace' entry\n"); #ifdef CONFIG_FUNCTION_GRAPH_TRACER - entry = debugfs_create_file("set_graph_function", 0444, d_tracer, + trace_create_file("set_graph_function", 0444, d_tracer, NULL, &ftrace_graph_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'set_graph_function' entry\n"); #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ return 0; @@ -2197,14 +2772,72 @@ static int ftrace_convert_nops(struct module *mod, return 0; } -void ftrace_init_module(struct module *mod, - unsigned long *start, unsigned long *end) +#ifdef CONFIG_MODULES +void ftrace_release(void *start, void *end) +{ + struct dyn_ftrace *rec; + struct ftrace_page *pg; + unsigned long s = (unsigned long)start; + unsigned long e = (unsigned long)end; + + if (ftrace_disabled || !start || start == end) + return; + + mutex_lock(&ftrace_lock); + do_for_each_ftrace_rec(pg, rec) { + if ((rec->ip >= s) && (rec->ip < e)) { + /* + * rec->ip is changed in ftrace_free_rec() + * It should not between s and e if record was freed. + */ + FTRACE_WARN_ON(rec->flags & FTRACE_FL_FREE); + ftrace_free_rec(rec); + } + } while_for_each_ftrace_rec(); + mutex_unlock(&ftrace_lock); +} + +static void ftrace_init_module(struct module *mod, + unsigned long *start, unsigned long *end) { if (ftrace_disabled || start == end) return; ftrace_convert_nops(mod, start, end); } +static int ftrace_module_notify(struct notifier_block *self, + unsigned long val, void *data) +{ + struct module *mod = data; + + switch (val) { + case MODULE_STATE_COMING: + ftrace_init_module(mod, mod->ftrace_callsites, + mod->ftrace_callsites + + mod->num_ftrace_callsites); + break; + case MODULE_STATE_GOING: + ftrace_release(mod->ftrace_callsites, + mod->ftrace_callsites + + mod->num_ftrace_callsites); + break; + } + + return 0; +} +#else +static int ftrace_module_notify(struct notifier_block *self, + unsigned long val, void *data) +{ + return 0; +} +#endif /* CONFIG_MODULES */ + +struct notifier_block ftrace_module_nb = { + .notifier_call = ftrace_module_notify, + .priority = 0, +}; + extern unsigned long __start_mcount_loc[]; extern unsigned long __stop_mcount_loc[]; @@ -2236,6 +2869,12 @@ void __init ftrace_init(void) __start_mcount_loc, __stop_mcount_loc); + ret = register_module_notifier(&ftrace_module_nb); + if (ret) + pr_warning("Failed to register trace ftrace module notifier\n"); + + set_ftrace_early_filters(); + return; failed: ftrace_disabled = 1; @@ -2417,7 +3056,6 @@ static const struct file_operations ftrace_pid_fops = { static __init int ftrace_init_debugfs(void) { struct dentry *d_tracer; - struct dentry *entry; d_tracer = tracing_init_dentry(); if (!d_tracer) @@ -2425,11 +3063,11 @@ static __init int ftrace_init_debugfs(void) ftrace_init_dyn_debugfs(d_tracer); - entry = debugfs_create_file("set_ftrace_pid", 0644, d_tracer, - NULL, &ftrace_pid_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'set_ftrace_pid' entry\n"); + trace_create_file("set_ftrace_pid", 0644, d_tracer, + NULL, &ftrace_pid_fops); + + ftrace_profile_debugfs(d_tracer); + return 0; } fs_initcall(ftrace_init_debugfs); @@ -2538,7 +3176,7 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, #ifdef CONFIG_FUNCTION_GRAPH_TRACER -static atomic_t ftrace_graph_active; +static int ftrace_graph_active; static struct notifier_block ftrace_suspend_notifier; int ftrace_graph_entry_stub(struct ftrace_graph_ent *trace) @@ -2580,12 +3218,12 @@ static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list) } if (t->ret_stack == NULL) { - t->curr_ret_stack = -1; - /* Make sure IRQs see the -1 first: */ - barrier(); - t->ret_stack = ret_stack_list[start++]; atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); + t->curr_ret_stack = -1; + /* Make sure the tasks see the -1 first: */ + smp_wmb(); + t->ret_stack = ret_stack_list[start++]; } } while_each_thread(g, t); @@ -2643,8 +3281,10 @@ static int start_graph_tracing(void) return -ENOMEM; /* The cpu_boot init_task->ret_stack will never be freed */ - for_each_online_cpu(cpu) - ftrace_graph_init_task(idle_task(cpu)); + for_each_online_cpu(cpu) { + if (!idle_task(cpu)->ret_stack) + ftrace_graph_init_task(idle_task(cpu)); + } do { ret = alloc_retstack_tasklist(ret_stack_list); @@ -2690,7 +3330,7 @@ int register_ftrace_graph(trace_func_graph_ret_t retfunc, mutex_lock(&ftrace_lock); /* we currently allow only one tracer registered at a time */ - if (atomic_read(&ftrace_graph_active)) { + if (ftrace_graph_active) { ret = -EBUSY; goto out; } @@ -2698,10 +3338,10 @@ int register_ftrace_graph(trace_func_graph_ret_t retfunc, ftrace_suspend_notifier.notifier_call = ftrace_suspend_notifier_call; register_pm_notifier(&ftrace_suspend_notifier); - atomic_inc(&ftrace_graph_active); + ftrace_graph_active++; ret = start_graph_tracing(); if (ret) { - atomic_dec(&ftrace_graph_active); + ftrace_graph_active--; goto out; } @@ -2719,10 +3359,10 @@ void unregister_ftrace_graph(void) { mutex_lock(&ftrace_lock); - if (!unlikely(atomic_read(&ftrace_graph_active))) + if (unlikely(!ftrace_graph_active)) goto out; - atomic_dec(&ftrace_graph_active); + ftrace_graph_active--; unregister_trace_sched_switch(ftrace_graph_probe_sched_switch); ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub; ftrace_graph_entry = ftrace_graph_entry_stub; @@ -2736,18 +3376,25 @@ void unregister_ftrace_graph(void) /* Allocate a return stack for newly created task */ void ftrace_graph_init_task(struct task_struct *t) { - if (atomic_read(&ftrace_graph_active)) { - t->ret_stack = kmalloc(FTRACE_RETFUNC_DEPTH + /* Make sure we do not use the parent ret_stack */ + t->ret_stack = NULL; + + if (ftrace_graph_active) { + struct ftrace_ret_stack *ret_stack; + + ret_stack = kmalloc(FTRACE_RETFUNC_DEPTH * sizeof(struct ftrace_ret_stack), GFP_KERNEL); - if (!t->ret_stack) + if (!ret_stack) return; t->curr_ret_stack = -1; atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); t->ftrace_timestamp = 0; - } else - t->ret_stack = NULL; + /* make curr_ret_stack visable before we add the ret_stack */ + smp_wmb(); + t->ret_stack = ret_stack; + } } void ftrace_graph_exit_task(struct task_struct *t) diff --git a/kernel/trace/kmemtrace.c b/kernel/trace/kmemtrace.c index 5011f4d91e3..86cdf671d7e 100644 --- a/kernel/trace/kmemtrace.c +++ b/kernel/trace/kmemtrace.c @@ -12,7 +12,7 @@ #include <linux/dcache.h> #include <linux/fs.h> -#include <trace/kmemtrace.h> +#include <linux/kmemtrace.h> #include "trace_output.h" #include "trace.h" @@ -42,6 +42,7 @@ static inline void kmemtrace_alloc(enum kmemtrace_type_id type_id, gfp_t gfp_flags, int node) { + struct ftrace_event_call *call = &event_kmem_alloc; struct trace_array *tr = kmemtrace_array; struct kmemtrace_alloc_entry *entry; struct ring_buffer_event *event; @@ -62,7 +63,8 @@ static inline void kmemtrace_alloc(enum kmemtrace_type_id type_id, entry->gfp_flags = gfp_flags; entry->node = node; - ring_buffer_unlock_commit(tr->buffer, event); + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); trace_wake_up(); } @@ -71,6 +73,7 @@ static inline void kmemtrace_free(enum kmemtrace_type_id type_id, unsigned long call_site, const void *ptr) { + struct ftrace_event_call *call = &event_kmem_free; struct trace_array *tr = kmemtrace_array; struct kmemtrace_free_entry *entry; struct ring_buffer_event *event; @@ -86,7 +89,8 @@ static inline void kmemtrace_free(enum kmemtrace_type_id type_id, entry->call_site = call_site; entry->ptr = ptr; - ring_buffer_unlock_commit(tr->buffer, event); + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); trace_wake_up(); } diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 960cbf44c84..2e642b2b725 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -22,6 +22,28 @@ #include "trace.h" /* + * The ring buffer header is special. We must manually up keep it. + */ +int ring_buffer_print_entry_header(struct trace_seq *s) +{ + int ret; + + ret = trace_seq_printf(s, "# compressed entry header\n"); + ret = trace_seq_printf(s, "\ttype_len : 5 bits\n"); + ret = trace_seq_printf(s, "\ttime_delta : 27 bits\n"); + ret = trace_seq_printf(s, "\tarray : 32 bits\n"); + ret = trace_seq_printf(s, "\n"); + ret = trace_seq_printf(s, "\tpadding : type == %d\n", + RINGBUF_TYPE_PADDING); + ret = trace_seq_printf(s, "\ttime_extend : type == %d\n", + RINGBUF_TYPE_TIME_EXTEND); + ret = trace_seq_printf(s, "\tdata max type_len == %d\n", + RINGBUF_TYPE_DATA_TYPE_LEN_MAX); + + return ret; +} + +/* * The ring buffer is made up of a list of pages. A separate list of pages is * allocated for each CPU. A writer may only write to a buffer that is * associated with the CPU it is currently executing on. A reader may read @@ -182,7 +204,10 @@ EXPORT_SYMBOL_GPL(tracing_is_on); #define RB_EVNT_HDR_SIZE (offsetof(struct ring_buffer_event, array)) #define RB_ALIGNMENT 4U -#define RB_MAX_SMALL_DATA 28 +#define RB_MAX_SMALL_DATA (RB_ALIGNMENT * RINGBUF_TYPE_DATA_TYPE_LEN_MAX) + +/* define RINGBUF_TYPE_DATA for 'case RINGBUF_TYPE_DATA:' */ +#define RINGBUF_TYPE_DATA 0 ... RINGBUF_TYPE_DATA_TYPE_LEN_MAX enum { RB_LEN_TIME_EXTEND = 8, @@ -191,48 +216,28 @@ enum { static inline int rb_null_event(struct ring_buffer_event *event) { - return event->type == RINGBUF_TYPE_PADDING && event->time_delta == 0; + return event->type_len == RINGBUF_TYPE_PADDING + && event->time_delta == 0; } static inline int rb_discarded_event(struct ring_buffer_event *event) { - return event->type == RINGBUF_TYPE_PADDING && event->time_delta; + return event->type_len == RINGBUF_TYPE_PADDING && event->time_delta; } static void rb_event_set_padding(struct ring_buffer_event *event) { - event->type = RINGBUF_TYPE_PADDING; + event->type_len = RINGBUF_TYPE_PADDING; event->time_delta = 0; } -/** - * ring_buffer_event_discard - discard an event in the ring buffer - * @buffer: the ring buffer - * @event: the event to discard - * - * Sometimes a event that is in the ring buffer needs to be ignored. - * This function lets the user discard an event in the ring buffer - * and then that event will not be read later. - * - * Note, it is up to the user to be careful with this, and protect - * against races. If the user discards an event that has been consumed - * it is possible that it could corrupt the ring buffer. - */ -void ring_buffer_event_discard(struct ring_buffer_event *event) -{ - event->type = RINGBUF_TYPE_PADDING; - /* time delta must be non zero */ - if (!event->time_delta) - event->time_delta = 1; -} - static unsigned rb_event_data_length(struct ring_buffer_event *event) { unsigned length; - if (event->len) - length = event->len * RB_ALIGNMENT; + if (event->type_len) + length = event->type_len * RB_ALIGNMENT; else length = event->array[0]; return length + RB_EVNT_HDR_SIZE; @@ -242,12 +247,12 @@ rb_event_data_length(struct ring_buffer_event *event) static unsigned rb_event_length(struct ring_buffer_event *event) { - switch (event->type) { + switch (event->type_len) { case RINGBUF_TYPE_PADDING: if (rb_null_event(event)) /* undefined */ return -1; - return rb_event_data_length(event); + return event->array[0] + RB_EVNT_HDR_SIZE; case RINGBUF_TYPE_TIME_EXTEND: return RB_LEN_TIME_EXTEND; @@ -271,7 +276,7 @@ rb_event_length(struct ring_buffer_event *event) unsigned ring_buffer_event_length(struct ring_buffer_event *event) { unsigned length = rb_event_length(event); - if (event->type != RINGBUF_TYPE_DATA) + if (event->type_len > RINGBUF_TYPE_DATA_TYPE_LEN_MAX) return length; length -= RB_EVNT_HDR_SIZE; if (length > RB_MAX_SMALL_DATA + sizeof(event->array[0])) @@ -284,9 +289,9 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_length); static void * rb_event_data(struct ring_buffer_event *event) { - BUG_ON(event->type != RINGBUF_TYPE_DATA); + BUG_ON(event->type_len > RINGBUF_TYPE_DATA_TYPE_LEN_MAX); /* If length is in len field, then array[0] has the data */ - if (event->len) + if (event->type_len) return (void *)&event->array[0]; /* Otherwise length is in array[0] and array[1] has the data */ return (void *)&event->array[1]; @@ -316,9 +321,10 @@ struct buffer_data_page { }; struct buffer_page { + struct list_head list; /* list of buffer pages */ local_t write; /* index for next write */ unsigned read; /* index for next read */ - struct list_head list; /* list of free pages */ + local_t entries; /* entries on this page */ struct buffer_data_page *page; /* Actual data page */ }; @@ -361,6 +367,34 @@ static inline int test_time_stamp(u64 delta) #define BUF_PAGE_SIZE (PAGE_SIZE - BUF_PAGE_HDR_SIZE) +/* Max payload is BUF_PAGE_SIZE - header (8bytes) */ +#define BUF_MAX_DATA_SIZE (BUF_PAGE_SIZE - (sizeof(u32) * 2)) + +/* Max number of timestamps that can fit on a page */ +#define RB_TIMESTAMPS_PER_PAGE (BUF_PAGE_SIZE / RB_LEN_TIME_STAMP) + +int ring_buffer_print_page_header(struct trace_seq *s) +{ + struct buffer_data_page field; + int ret; + + ret = trace_seq_printf(s, "\tfield: u64 timestamp;\t" + "offset:0;\tsize:%u;\n", + (unsigned int)sizeof(field.time_stamp)); + + ret = trace_seq_printf(s, "\tfield: local_t commit;\t" + "offset:%u;\tsize:%u;\n", + (unsigned int)offsetof(typeof(field), commit), + (unsigned int)sizeof(field.commit)); + + ret = trace_seq_printf(s, "\tfield: char data;\t" + "offset:%u;\tsize:%u;\n", + (unsigned int)offsetof(typeof(field), data), + (unsigned int)BUF_PAGE_SIZE); + + return ret; +} + /* * head_page == tail_page && head == tail then buffer is empty. */ @@ -375,8 +409,11 @@ struct ring_buffer_per_cpu { struct buffer_page *tail_page; /* write to tail */ struct buffer_page *commit_page; /* committed pages */ struct buffer_page *reader_page; + unsigned long nmi_dropped; + unsigned long commit_overrun; unsigned long overrun; - unsigned long entries; + unsigned long read; + local_t entries; u64 write_stamp; u64 read_stamp; atomic_t record_disabled; @@ -389,6 +426,8 @@ struct ring_buffer { atomic_t record_disabled; cpumask_var_t cpumask; + struct lock_class_key *reader_lock_key; + struct mutex mutex; struct ring_buffer_per_cpu **buffers; @@ -420,13 +459,18 @@ struct ring_buffer_iter { /* Up this if you want to test the TIME_EXTENTS and normalization */ #define DEBUG_SHIFT 0 +static inline u64 rb_time_stamp(struct ring_buffer *buffer, int cpu) +{ + /* shift to debug/test normalization and TIME_EXTENTS */ + return buffer->clock() << DEBUG_SHIFT; +} + u64 ring_buffer_time_stamp(struct ring_buffer *buffer, int cpu) { u64 time; preempt_disable_notrace(); - /* shift to debug/test normalization and TIME_EXTENTS */ - time = buffer->clock() << DEBUG_SHIFT; + time = rb_time_stamp(buffer, cpu); preempt_enable_no_resched_notrace(); return time; @@ -523,6 +567,7 @@ rb_allocate_cpu_buffer(struct ring_buffer *buffer, int cpu) cpu_buffer->cpu = cpu; cpu_buffer->buffer = buffer; spin_lock_init(&cpu_buffer->reader_lock); + lockdep_set_class(&cpu_buffer->reader_lock, buffer->reader_lock_key); cpu_buffer->lock = (raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED; INIT_LIST_HEAD(&cpu_buffer->pages); @@ -593,7 +638,8 @@ static int rb_cpu_notify(struct notifier_block *self, * when the buffer wraps. If this flag is not set, the buffer will * drop data when the tail hits the head. */ -struct ring_buffer *ring_buffer_alloc(unsigned long size, unsigned flags) +struct ring_buffer *__ring_buffer_alloc(unsigned long size, unsigned flags, + struct lock_class_key *key) { struct ring_buffer *buffer; int bsize; @@ -616,6 +662,7 @@ struct ring_buffer *ring_buffer_alloc(unsigned long size, unsigned flags) buffer->pages = DIV_ROUND_UP(size, BUF_PAGE_SIZE); buffer->flags = flags; buffer->clock = trace_clock_local; + buffer->reader_lock_key = key; /* need at least two pages */ if (buffer->pages == 1) @@ -673,7 +720,7 @@ struct ring_buffer *ring_buffer_alloc(unsigned long size, unsigned flags) kfree(buffer); return NULL; } -EXPORT_SYMBOL_GPL(ring_buffer_alloc); +EXPORT_SYMBOL_GPL(__ring_buffer_alloc); /** * ring_buffer_free - free a ring buffer. @@ -947,31 +994,6 @@ static inline unsigned rb_head_size(struct ring_buffer_per_cpu *cpu_buffer) return rb_page_commit(cpu_buffer->head_page); } -/* - * When the tail hits the head and the buffer is in overwrite mode, - * the head jumps to the next page and all content on the previous - * page is discarded. But before doing so, we update the overrun - * variable of the buffer. - */ -static void rb_update_overflow(struct ring_buffer_per_cpu *cpu_buffer) -{ - struct ring_buffer_event *event; - unsigned long head; - - for (head = 0; head < rb_head_size(cpu_buffer); - head += rb_event_length(event)) { - - event = __rb_page_index(cpu_buffer->head_page, head); - if (RB_WARN_ON(cpu_buffer, rb_null_event(event))) - return; - /* Only count data entries */ - if (event->type != RINGBUF_TYPE_DATA) - continue; - cpu_buffer->overrun++; - cpu_buffer->entries--; - } -} - static inline void rb_inc_page(struct ring_buffer_per_cpu *cpu_buffer, struct buffer_page **bpage) { @@ -991,7 +1013,7 @@ rb_event_index(struct ring_buffer_event *event) return (addr & ~PAGE_MASK) - (PAGE_SIZE - BUF_PAGE_SIZE); } -static int +static inline int rb_is_commit(struct ring_buffer_per_cpu *cpu_buffer, struct ring_buffer_event *event) { @@ -1110,28 +1132,21 @@ static void rb_update_event(struct ring_buffer_event *event, unsigned type, unsigned length) { - event->type = type; + event->type_len = type; switch (type) { case RINGBUF_TYPE_PADDING: - break; - case RINGBUF_TYPE_TIME_EXTEND: - event->len = DIV_ROUND_UP(RB_LEN_TIME_EXTEND, RB_ALIGNMENT); - break; - case RINGBUF_TYPE_TIME_STAMP: - event->len = DIV_ROUND_UP(RB_LEN_TIME_STAMP, RB_ALIGNMENT); break; - case RINGBUF_TYPE_DATA: + case 0: length -= RB_EVNT_HDR_SIZE; - if (length > RB_MAX_SMALL_DATA) { - event->len = 0; + if (length > RB_MAX_SMALL_DATA) event->array[0] = length; - } else - event->len = DIV_ROUND_UP(length, RB_ALIGNMENT); + else + event->type_len = DIV_ROUND_UP(length, RB_ALIGNMENT); break; default: BUG(); @@ -1155,131 +1170,156 @@ static unsigned rb_calculate_event_length(unsigned length) return length; } + static struct ring_buffer_event * -__rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer, - unsigned type, unsigned long length, u64 *ts) +rb_move_tail(struct ring_buffer_per_cpu *cpu_buffer, + unsigned long length, unsigned long tail, + struct buffer_page *commit_page, + struct buffer_page *tail_page, u64 *ts) { - struct buffer_page *tail_page, *head_page, *reader_page, *commit_page; - unsigned long tail, write; + struct buffer_page *next_page, *head_page, *reader_page; struct ring_buffer *buffer = cpu_buffer->buffer; struct ring_buffer_event *event; - unsigned long flags; bool lock_taken = false; + unsigned long flags; - commit_page = cpu_buffer->commit_page; - /* we just need to protect against interrupts */ - barrier(); - tail_page = cpu_buffer->tail_page; - write = local_add_return(length, &tail_page->write); - tail = write - length; + next_page = tail_page; - /* See if we shot pass the end of this buffer page */ - if (write > BUF_PAGE_SIZE) { - struct buffer_page *next_page = tail_page; + local_irq_save(flags); + /* + * Since the write to the buffer is still not + * fully lockless, we must be careful with NMIs. + * The locks in the writers are taken when a write + * crosses to a new page. The locks protect against + * races with the readers (this will soon be fixed + * with a lockless solution). + * + * Because we can not protect against NMIs, and we + * want to keep traces reentrant, we need to manage + * what happens when we are in an NMI. + * + * NMIs can happen after we take the lock. + * If we are in an NMI, only take the lock + * if it is not already taken. Otherwise + * simply fail. + */ + if (unlikely(in_nmi())) { + if (!__raw_spin_trylock(&cpu_buffer->lock)) { + cpu_buffer->nmi_dropped++; + goto out_reset; + } + } else + __raw_spin_lock(&cpu_buffer->lock); - local_irq_save(flags); - /* - * Since the write to the buffer is still not - * fully lockless, we must be careful with NMIs. - * The locks in the writers are taken when a write - * crosses to a new page. The locks protect against - * races with the readers (this will soon be fixed - * with a lockless solution). - * - * Because we can not protect against NMIs, and we - * want to keep traces reentrant, we need to manage - * what happens when we are in an NMI. - * - * NMIs can happen after we take the lock. - * If we are in an NMI, only take the lock - * if it is not already taken. Otherwise - * simply fail. - */ - if (unlikely(in_nmi())) { - if (!__raw_spin_trylock(&cpu_buffer->lock)) - goto out_reset; - } else - __raw_spin_lock(&cpu_buffer->lock); + lock_taken = true; - lock_taken = true; + rb_inc_page(cpu_buffer, &next_page); - rb_inc_page(cpu_buffer, &next_page); + head_page = cpu_buffer->head_page; + reader_page = cpu_buffer->reader_page; - head_page = cpu_buffer->head_page; - reader_page = cpu_buffer->reader_page; + /* we grabbed the lock before incrementing */ + if (RB_WARN_ON(cpu_buffer, next_page == reader_page)) + goto out_reset; - /* we grabbed the lock before incrementing */ - if (RB_WARN_ON(cpu_buffer, next_page == reader_page)) - goto out_reset; + /* + * If for some reason, we had an interrupt storm that made + * it all the way around the buffer, bail, and warn + * about it. + */ + if (unlikely(next_page == commit_page)) { + cpu_buffer->commit_overrun++; + goto out_reset; + } - /* - * If for some reason, we had an interrupt storm that made - * it all the way around the buffer, bail, and warn - * about it. - */ - if (unlikely(next_page == commit_page)) { - WARN_ON_ONCE(1); + if (next_page == head_page) { + if (!(buffer->flags & RB_FL_OVERWRITE)) goto out_reset; - } - if (next_page == head_page) { - if (!(buffer->flags & RB_FL_OVERWRITE)) - goto out_reset; - - /* tail_page has not moved yet? */ - if (tail_page == cpu_buffer->tail_page) { - /* count overflows */ - rb_update_overflow(cpu_buffer); + /* tail_page has not moved yet? */ + if (tail_page == cpu_buffer->tail_page) { + /* count overflows */ + cpu_buffer->overrun += + local_read(&head_page->entries); - rb_inc_page(cpu_buffer, &head_page); - cpu_buffer->head_page = head_page; - cpu_buffer->head_page->read = 0; - } + rb_inc_page(cpu_buffer, &head_page); + cpu_buffer->head_page = head_page; + cpu_buffer->head_page->read = 0; } + } - /* - * If the tail page is still the same as what we think - * it is, then it is up to us to update the tail - * pointer. - */ - if (tail_page == cpu_buffer->tail_page) { - local_set(&next_page->write, 0); - local_set(&next_page->page->commit, 0); - cpu_buffer->tail_page = next_page; + /* + * If the tail page is still the same as what we think + * it is, then it is up to us to update the tail + * pointer. + */ + if (tail_page == cpu_buffer->tail_page) { + local_set(&next_page->write, 0); + local_set(&next_page->entries, 0); + local_set(&next_page->page->commit, 0); + cpu_buffer->tail_page = next_page; + + /* reread the time stamp */ + *ts = rb_time_stamp(buffer, cpu_buffer->cpu); + cpu_buffer->tail_page->page->time_stamp = *ts; + } - /* reread the time stamp */ - *ts = ring_buffer_time_stamp(buffer, cpu_buffer->cpu); - cpu_buffer->tail_page->page->time_stamp = *ts; - } + /* + * The actual tail page has moved forward. + */ + if (tail < BUF_PAGE_SIZE) { + /* Mark the rest of the page with padding */ + event = __rb_page_index(tail_page, tail); + rb_event_set_padding(event); + } - /* - * The actual tail page has moved forward. - */ - if (tail < BUF_PAGE_SIZE) { - /* Mark the rest of the page with padding */ - event = __rb_page_index(tail_page, tail); - rb_event_set_padding(event); - } + /* Set the write back to the previous setting */ + local_sub(length, &tail_page->write); - if (tail <= BUF_PAGE_SIZE) - /* Set the write back to the previous setting */ - local_set(&tail_page->write, tail); + /* + * If this was a commit entry that failed, + * increment that too + */ + if (tail_page == cpu_buffer->commit_page && + tail == rb_commit_index(cpu_buffer)) { + rb_set_commit_to_write(cpu_buffer); + } - /* - * If this was a commit entry that failed, - * increment that too - */ - if (tail_page == cpu_buffer->commit_page && - tail == rb_commit_index(cpu_buffer)) { - rb_set_commit_to_write(cpu_buffer); - } + __raw_spin_unlock(&cpu_buffer->lock); + local_irq_restore(flags); + + /* fail and let the caller try again */ + return ERR_PTR(-EAGAIN); + + out_reset: + /* reset write */ + local_sub(length, &tail_page->write); + if (likely(lock_taken)) __raw_spin_unlock(&cpu_buffer->lock); - local_irq_restore(flags); + local_irq_restore(flags); + return NULL; +} - /* fail and let the caller try again */ - return ERR_PTR(-EAGAIN); - } +static struct ring_buffer_event * +__rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer, + unsigned type, unsigned long length, u64 *ts) +{ + struct buffer_page *tail_page, *commit_page; + struct ring_buffer_event *event; + unsigned long tail, write; + + commit_page = cpu_buffer->commit_page; + /* we just need to protect against interrupts */ + barrier(); + tail_page = cpu_buffer->tail_page; + write = local_add_return(length, &tail_page->write); + tail = write - length; + + /* See if we shot pass the end of this buffer page */ + if (write > BUF_PAGE_SIZE) + return rb_move_tail(cpu_buffer, length, tail, + commit_page, tail_page, ts); /* We reserved something on the buffer */ @@ -1289,6 +1329,10 @@ __rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer, event = __rb_page_index(tail_page, tail); rb_update_event(event, type, length); + /* The passed in type is zero for DATA */ + if (likely(!type)) + local_inc(&tail_page->entries); + /* * If this is a commit and the tail is zero, then update * this page's time stamp. @@ -1297,16 +1341,38 @@ __rb_reserve_next(struct ring_buffer_per_cpu *cpu_buffer, cpu_buffer->commit_page->page->time_stamp = *ts; return event; +} - out_reset: - /* reset write */ - if (tail <= BUF_PAGE_SIZE) - local_set(&tail_page->write, tail); +static inline int +rb_try_to_discard(struct ring_buffer_per_cpu *cpu_buffer, + struct ring_buffer_event *event) +{ + unsigned long new_index, old_index; + struct buffer_page *bpage; + unsigned long index; + unsigned long addr; - if (likely(lock_taken)) - __raw_spin_unlock(&cpu_buffer->lock); - local_irq_restore(flags); - return NULL; + new_index = rb_event_index(event); + old_index = new_index + rb_event_length(event); + addr = (unsigned long)event; + addr &= PAGE_MASK; + + bpage = cpu_buffer->tail_page; + + if (bpage->page == (void *)addr && rb_page_write(bpage) == old_index) { + /* + * This is on the tail page. It is possible that + * a write could come in and move the tail page + * and write to the next page. That is fine + * because we just shorten what is on this page. + */ + index = local_cmpxchg(&bpage->write, old_index, new_index); + if (index == old_index) + return 1; + } + + /* could not discard */ + return 0; } static int @@ -1351,16 +1417,23 @@ rb_add_time_stamp(struct ring_buffer_per_cpu *cpu_buffer, event->array[0] = *delta >> TS_SHIFT; } else { cpu_buffer->commit_page->page->time_stamp = *ts; - event->time_delta = 0; - event->array[0] = 0; + /* try to discard, since we do not need this */ + if (!rb_try_to_discard(cpu_buffer, event)) { + /* nope, just zero it */ + event->time_delta = 0; + event->array[0] = 0; + } } cpu_buffer->write_stamp = *ts; /* let the caller know this was the commit */ ret = 1; } else { - /* Darn, this is just wasted space */ - event->time_delta = 0; - event->array[0] = 0; + /* Try to discard the event */ + if (!rb_try_to_discard(cpu_buffer, event)) { + /* Darn, this is just wasted space */ + event->time_delta = 0; + event->array[0] = 0; + } ret = 0; } @@ -1371,13 +1444,14 @@ rb_add_time_stamp(struct ring_buffer_per_cpu *cpu_buffer, static struct ring_buffer_event * rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer, - unsigned type, unsigned long length) + unsigned long length) { struct ring_buffer_event *event; - u64 ts, delta; + u64 ts, delta = 0; int commit = 0; int nr_loops = 0; + length = rb_calculate_event_length(length); again: /* * We allow for interrupts to reenter here and do a trace. @@ -1391,7 +1465,7 @@ rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer, if (RB_WARN_ON(cpu_buffer, ++nr_loops > 1000)) return NULL; - ts = ring_buffer_time_stamp(cpu_buffer->buffer, cpu_buffer->cpu); + ts = rb_time_stamp(cpu_buffer->buffer, cpu_buffer->cpu); /* * Only the first commit can update the timestamp. @@ -1401,23 +1475,24 @@ rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer, * also be made. But only the entry that did the actual * commit will be something other than zero. */ - if (cpu_buffer->tail_page == cpu_buffer->commit_page && - rb_page_write(cpu_buffer->tail_page) == - rb_commit_index(cpu_buffer)) { + if (likely(cpu_buffer->tail_page == cpu_buffer->commit_page && + rb_page_write(cpu_buffer->tail_page) == + rb_commit_index(cpu_buffer))) { + u64 diff; - delta = ts - cpu_buffer->write_stamp; + diff = ts - cpu_buffer->write_stamp; - /* make sure this delta is calculated here */ + /* make sure this diff is calculated here */ barrier(); /* Did the write stamp get updated already? */ if (unlikely(ts < cpu_buffer->write_stamp)) - delta = 0; + goto get_event; - if (test_time_stamp(delta)) { + delta = diff; + if (unlikely(test_time_stamp(delta))) { commit = rb_add_time_stamp(cpu_buffer, &ts, &delta); - if (commit == -EBUSY) return NULL; @@ -1426,12 +1501,11 @@ rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer, RB_WARN_ON(cpu_buffer, commit < 0); } - } else - /* Non commits have zero deltas */ - delta = 0; + } - event = __rb_reserve_next(cpu_buffer, type, length, &ts); - if (PTR_ERR(event) == -EAGAIN) + get_event: + event = __rb_reserve_next(cpu_buffer, 0, length, &ts); + if (unlikely(PTR_ERR(event) == -EAGAIN)) goto again; if (!event) { @@ -1448,7 +1522,7 @@ rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer, * If the timestamp was commited, make the commit our entry * now so that we will update it when needed. */ - if (commit) + if (unlikely(commit)) rb_set_commit_event(cpu_buffer, event); else if (!rb_is_commit(cpu_buffer, event)) delta = 0; @@ -1458,6 +1532,36 @@ rb_reserve_next_event(struct ring_buffer_per_cpu *cpu_buffer, return event; } +#define TRACE_RECURSIVE_DEPTH 16 + +static int trace_recursive_lock(void) +{ + current->trace_recursion++; + + if (likely(current->trace_recursion < TRACE_RECURSIVE_DEPTH)) + return 0; + + /* Disable all tracing before we do anything else */ + tracing_off_permanent(); + + printk_once(KERN_WARNING "Tracing recursion: depth[%ld]:" + "HC[%lu]:SC[%lu]:NMI[%lu]\n", + current->trace_recursion, + hardirq_count() >> HARDIRQ_SHIFT, + softirq_count() >> SOFTIRQ_SHIFT, + in_nmi()); + + WARN_ON_ONCE(1); + return -1; +} + +static void trace_recursive_unlock(void) +{ + WARN_ON_ONCE(!current->trace_recursion); + + current->trace_recursion--; +} + static DEFINE_PER_CPU(int, rb_need_resched); /** @@ -1491,6 +1595,9 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length) /* If we are tracing schedule, we don't want to recurse */ resched = ftrace_preempt_disable(); + if (trace_recursive_lock()) + goto out_nocheck; + cpu = raw_smp_processor_id(); if (!cpumask_test_cpu(cpu, buffer->cpumask)) @@ -1501,11 +1608,10 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length) if (atomic_read(&cpu_buffer->record_disabled)) goto out; - length = rb_calculate_event_length(length); - if (length > BUF_PAGE_SIZE) + if (length > BUF_MAX_DATA_SIZE) goto out; - event = rb_reserve_next_event(cpu_buffer, RINGBUF_TYPE_DATA, length); + event = rb_reserve_next_event(cpu_buffer, length); if (!event) goto out; @@ -1520,6 +1626,9 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length) return event; out: + trace_recursive_unlock(); + + out_nocheck: ftrace_preempt_enable(resched); return NULL; } @@ -1528,7 +1637,7 @@ EXPORT_SYMBOL_GPL(ring_buffer_lock_reserve); static void rb_commit(struct ring_buffer_per_cpu *cpu_buffer, struct ring_buffer_event *event) { - cpu_buffer->entries++; + local_inc(&cpu_buffer->entries); /* Only process further if we own the commit */ if (!rb_is_commit(cpu_buffer, event)) @@ -1558,6 +1667,8 @@ int ring_buffer_unlock_commit(struct ring_buffer *buffer, rb_commit(cpu_buffer, event); + trace_recursive_unlock(); + /* * Only the last preempt count needs to restore preemption. */ @@ -1570,6 +1681,99 @@ int ring_buffer_unlock_commit(struct ring_buffer *buffer, } EXPORT_SYMBOL_GPL(ring_buffer_unlock_commit); +static inline void rb_event_discard(struct ring_buffer_event *event) +{ + /* array[0] holds the actual length for the discarded event */ + event->array[0] = rb_event_data_length(event) - RB_EVNT_HDR_SIZE; + event->type_len = RINGBUF_TYPE_PADDING; + /* time delta must be non zero */ + if (!event->time_delta) + event->time_delta = 1; +} + +/** + * ring_buffer_event_discard - discard any event in the ring buffer + * @event: the event to discard + * + * Sometimes a event that is in the ring buffer needs to be ignored. + * This function lets the user discard an event in the ring buffer + * and then that event will not be read later. + * + * Note, it is up to the user to be careful with this, and protect + * against races. If the user discards an event that has been consumed + * it is possible that it could corrupt the ring buffer. + */ +void ring_buffer_event_discard(struct ring_buffer_event *event) +{ + rb_event_discard(event); +} +EXPORT_SYMBOL_GPL(ring_buffer_event_discard); + +/** + * ring_buffer_commit_discard - discard an event that has not been committed + * @buffer: the ring buffer + * @event: non committed event to discard + * + * This is similar to ring_buffer_event_discard but must only be + * performed on an event that has not been committed yet. The difference + * is that this will also try to free the event from the ring buffer + * if another event has not been added behind it. + * + * If another event has been added behind it, it will set the event + * up as discarded, and perform the commit. + * + * If this function is called, do not call ring_buffer_unlock_commit on + * the event. + */ +void ring_buffer_discard_commit(struct ring_buffer *buffer, + struct ring_buffer_event *event) +{ + struct ring_buffer_per_cpu *cpu_buffer; + int cpu; + + /* The event is discarded regardless */ + rb_event_discard(event); + + /* + * This must only be called if the event has not been + * committed yet. Thus we can assume that preemption + * is still disabled. + */ + RB_WARN_ON(buffer, preemptible()); + + cpu = smp_processor_id(); + cpu_buffer = buffer->buffers[cpu]; + + if (!rb_try_to_discard(cpu_buffer, event)) + goto out; + + /* + * The commit is still visible by the reader, so we + * must increment entries. + */ + local_inc(&cpu_buffer->entries); + out: + /* + * If a write came in and pushed the tail page + * we still need to update the commit pointer + * if we were the commit. + */ + if (rb_is_commit(cpu_buffer, event)) + rb_set_commit_to_write(cpu_buffer); + + trace_recursive_unlock(); + + /* + * Only the last preempt count needs to restore preemption. + */ + if (preempt_count() == 1) + ftrace_preempt_enable(per_cpu(rb_need_resched, cpu)); + else + preempt_enable_no_resched_notrace(); + +} +EXPORT_SYMBOL_GPL(ring_buffer_discard_commit); + /** * ring_buffer_write - write data to the buffer without reserving * @buffer: The ring buffer to write to. @@ -1589,7 +1793,6 @@ int ring_buffer_write(struct ring_buffer *buffer, { struct ring_buffer_per_cpu *cpu_buffer; struct ring_buffer_event *event; - unsigned long event_length; void *body; int ret = -EBUSY; int cpu, resched; @@ -1612,9 +1815,10 @@ int ring_buffer_write(struct ring_buffer *buffer, if (atomic_read(&cpu_buffer->record_disabled)) goto out; - event_length = rb_calculate_event_length(length); - event = rb_reserve_next_event(cpu_buffer, - RINGBUF_TYPE_DATA, event_length); + if (length > BUF_MAX_DATA_SIZE) + goto out; + + event = rb_reserve_next_event(cpu_buffer, length); if (!event) goto out; @@ -1728,7 +1932,8 @@ unsigned long ring_buffer_entries_cpu(struct ring_buffer *buffer, int cpu) return 0; cpu_buffer = buffer->buffers[cpu]; - ret = cpu_buffer->entries; + ret = (local_read(&cpu_buffer->entries) - cpu_buffer->overrun) + - cpu_buffer->read; return ret; } @@ -1755,6 +1960,47 @@ unsigned long ring_buffer_overrun_cpu(struct ring_buffer *buffer, int cpu) EXPORT_SYMBOL_GPL(ring_buffer_overrun_cpu); /** + * ring_buffer_nmi_dropped_cpu - get the number of nmis that were dropped + * @buffer: The ring buffer + * @cpu: The per CPU buffer to get the number of overruns from + */ +unsigned long ring_buffer_nmi_dropped_cpu(struct ring_buffer *buffer, int cpu) +{ + struct ring_buffer_per_cpu *cpu_buffer; + unsigned long ret; + + if (!cpumask_test_cpu(cpu, buffer->cpumask)) + return 0; + + cpu_buffer = buffer->buffers[cpu]; + ret = cpu_buffer->nmi_dropped; + + return ret; +} +EXPORT_SYMBOL_GPL(ring_buffer_nmi_dropped_cpu); + +/** + * ring_buffer_commit_overrun_cpu - get the number of overruns caused by commits + * @buffer: The ring buffer + * @cpu: The per CPU buffer to get the number of overruns from + */ +unsigned long +ring_buffer_commit_overrun_cpu(struct ring_buffer *buffer, int cpu) +{ + struct ring_buffer_per_cpu *cpu_buffer; + unsigned long ret; + + if (!cpumask_test_cpu(cpu, buffer->cpumask)) + return 0; + + cpu_buffer = buffer->buffers[cpu]; + ret = cpu_buffer->commit_overrun; + + return ret; +} +EXPORT_SYMBOL_GPL(ring_buffer_commit_overrun_cpu); + +/** * ring_buffer_entries - get the number of entries in a buffer * @buffer: The ring buffer * @@ -1770,7 +2016,8 @@ unsigned long ring_buffer_entries(struct ring_buffer *buffer) /* if you care about this being correct, lock the buffer */ for_each_buffer_cpu(buffer, cpu) { cpu_buffer = buffer->buffers[cpu]; - entries += cpu_buffer->entries; + entries += (local_read(&cpu_buffer->entries) - + cpu_buffer->overrun) - cpu_buffer->read; } return entries; @@ -1862,7 +2109,7 @@ rb_update_read_stamp(struct ring_buffer_per_cpu *cpu_buffer, { u64 delta; - switch (event->type) { + switch (event->type_len) { case RINGBUF_TYPE_PADDING: return; @@ -1893,7 +2140,7 @@ rb_update_iter_read_stamp(struct ring_buffer_iter *iter, { u64 delta; - switch (event->type) { + switch (event->type_len) { case RINGBUF_TYPE_PADDING: return; @@ -1966,6 +2213,7 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) cpu_buffer->reader_page->list.prev = reader->list.prev; local_set(&cpu_buffer->reader_page->write, 0); + local_set(&cpu_buffer->reader_page->entries, 0); local_set(&cpu_buffer->reader_page->page->commit, 0); /* Make the reader page now replace the head */ @@ -2008,8 +2256,9 @@ static void rb_advance_reader(struct ring_buffer_per_cpu *cpu_buffer) event = rb_reader_event(cpu_buffer); - if (event->type == RINGBUF_TYPE_DATA || rb_discarded_event(event)) - cpu_buffer->entries--; + if (event->type_len <= RINGBUF_TYPE_DATA_TYPE_LEN_MAX + || rb_discarded_event(event)) + cpu_buffer->read++; rb_update_read_stamp(cpu_buffer, event); @@ -2031,8 +2280,8 @@ static void rb_advance_iter(struct ring_buffer_iter *iter) * Check if we are at the end of the buffer. */ if (iter->head >= rb_page_size(iter->head_page)) { - if (RB_WARN_ON(buffer, - iter->head_page == cpu_buffer->commit_page)) + /* discarded commits can make the page empty */ + if (iter->head_page == cpu_buffer->commit_page) return; rb_inc_iter(iter); return; @@ -2075,12 +2324,10 @@ rb_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) /* * We repeat when a timestamp is encountered. It is possible * to get multiple timestamps from an interrupt entering just - * as one timestamp is about to be written. The max times - * that this can happen is the number of nested interrupts we - * can have. Nesting 10 deep of interrupts is clearly - * an anomaly. + * as one timestamp is about to be written, or from discarded + * commits. The most that we can have is the number on a single page. */ - if (RB_WARN_ON(cpu_buffer, ++nr_loops > 10)) + if (RB_WARN_ON(cpu_buffer, ++nr_loops > RB_TIMESTAMPS_PER_PAGE)) return NULL; reader = rb_get_reader_page(cpu_buffer); @@ -2089,7 +2336,7 @@ rb_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) event = rb_reader_event(cpu_buffer); - switch (event->type) { + switch (event->type_len) { case RINGBUF_TYPE_PADDING: if (rb_null_event(event)) RB_WARN_ON(cpu_buffer, 1); @@ -2146,14 +2393,14 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts) again: /* - * We repeat when a timestamp is encountered. It is possible - * to get multiple timestamps from an interrupt entering just - * as one timestamp is about to be written. The max times - * that this can happen is the number of nested interrupts we - * can have. Nesting 10 deep of interrupts is clearly - * an anomaly. + * We repeat when a timestamp is encountered. + * We can get multiple timestamps by nested interrupts or also + * if filtering is on (discarding commits). Since discarding + * commits can be frequent we can get a lot of timestamps. + * But we limit them by not adding timestamps if they begin + * at the start of a page. */ - if (RB_WARN_ON(cpu_buffer, ++nr_loops > 10)) + if (RB_WARN_ON(cpu_buffer, ++nr_loops > RB_TIMESTAMPS_PER_PAGE)) return NULL; if (rb_per_cpu_empty(cpu_buffer)) @@ -2161,7 +2408,7 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts) event = rb_iter_head_event(iter); - switch (event->type) { + switch (event->type_len) { case RINGBUF_TYPE_PADDING: if (rb_null_event(event)) { rb_inc_iter(iter); @@ -2220,7 +2467,7 @@ ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) event = rb_buffer_peek(buffer, cpu, ts); spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); - if (event && event->type == RINGBUF_TYPE_PADDING) { + if (event && event->type_len == RINGBUF_TYPE_PADDING) { cpu_relax(); goto again; } @@ -2248,7 +2495,7 @@ ring_buffer_iter_peek(struct ring_buffer_iter *iter, u64 *ts) event = rb_iter_peek(iter, ts); spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); - if (event && event->type == RINGBUF_TYPE_PADDING) { + if (event && event->type_len == RINGBUF_TYPE_PADDING) { cpu_relax(); goto again; } @@ -2293,7 +2540,7 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts) out: preempt_enable(); - if (event && event->type == RINGBUF_TYPE_PADDING) { + if (event && event->type_len == RINGBUF_TYPE_PADDING) { cpu_relax(); goto again; } @@ -2386,7 +2633,7 @@ ring_buffer_read(struct ring_buffer_iter *iter, u64 *ts) out: spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); - if (event && event->type == RINGBUF_TYPE_PADDING) { + if (event && event->type_len == RINGBUF_TYPE_PADDING) { cpu_relax(); goto again; } @@ -2411,6 +2658,7 @@ rb_reset_cpu(struct ring_buffer_per_cpu *cpu_buffer) cpu_buffer->head_page = list_entry(cpu_buffer->pages.next, struct buffer_page, list); local_set(&cpu_buffer->head_page->write, 0); + local_set(&cpu_buffer->head_page->entries, 0); local_set(&cpu_buffer->head_page->page->commit, 0); cpu_buffer->head_page->read = 0; @@ -2420,11 +2668,15 @@ rb_reset_cpu(struct ring_buffer_per_cpu *cpu_buffer) INIT_LIST_HEAD(&cpu_buffer->reader_page->list); local_set(&cpu_buffer->reader_page->write, 0); + local_set(&cpu_buffer->reader_page->entries, 0); local_set(&cpu_buffer->reader_page->page->commit, 0); cpu_buffer->reader_page->read = 0; + cpu_buffer->nmi_dropped = 0; + cpu_buffer->commit_overrun = 0; cpu_buffer->overrun = 0; - cpu_buffer->entries = 0; + cpu_buffer->read = 0; + local_set(&cpu_buffer->entries, 0); cpu_buffer->write_stamp = 0; cpu_buffer->read_stamp = 0; @@ -2443,6 +2695,8 @@ void ring_buffer_reset_cpu(struct ring_buffer *buffer, int cpu) if (!cpumask_test_cpu(cpu, buffer->cpumask)) return; + atomic_inc(&cpu_buffer->record_disabled); + spin_lock_irqsave(&cpu_buffer->reader_lock, flags); __raw_spin_lock(&cpu_buffer->lock); @@ -2452,6 +2706,8 @@ void ring_buffer_reset_cpu(struct ring_buffer *buffer, int cpu) __raw_spin_unlock(&cpu_buffer->lock); spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); + + atomic_dec(&cpu_buffer->record_disabled); } EXPORT_SYMBOL_GPL(ring_buffer_reset_cpu); @@ -2578,28 +2834,6 @@ out: } EXPORT_SYMBOL_GPL(ring_buffer_swap_cpu); -static void rb_remove_entries(struct ring_buffer_per_cpu *cpu_buffer, - struct buffer_data_page *bpage, - unsigned int offset) -{ - struct ring_buffer_event *event; - unsigned long head; - - __raw_spin_lock(&cpu_buffer->lock); - for (head = offset; head < local_read(&bpage->commit); - head += rb_event_length(event)) { - - event = __rb_data_page_index(bpage, head); - if (RB_WARN_ON(cpu_buffer, rb_null_event(event))) - return; - /* Only count data entries */ - if (event->type != RINGBUF_TYPE_DATA) - continue; - cpu_buffer->entries--; - } - __raw_spin_unlock(&cpu_buffer->lock); -} - /** * ring_buffer_alloc_read_page - allocate a page to read from buffer * @buffer: the buffer to allocate for. @@ -2630,6 +2864,7 @@ void *ring_buffer_alloc_read_page(struct ring_buffer *buffer) return bpage; } +EXPORT_SYMBOL_GPL(ring_buffer_alloc_read_page); /** * ring_buffer_free_read_page - free an allocated read page @@ -2642,6 +2877,7 @@ void ring_buffer_free_read_page(struct ring_buffer *buffer, void *data) { free_page((unsigned long)data); } +EXPORT_SYMBOL_GPL(ring_buffer_free_read_page); /** * ring_buffer_read_page - extract a page from the ring buffer @@ -2768,16 +3004,17 @@ int ring_buffer_read_page(struct ring_buffer *buffer, /* we copied everything to the beginning */ read = 0; } else { + /* update the entry counter */ + cpu_buffer->read += local_read(&reader->entries); + /* swap the pages */ rb_init_page(bpage); bpage = reader->page; reader->page = *data_page; local_set(&reader->write, 0); + local_set(&reader->entries, 0); reader->read = 0; *data_page = bpage; - - /* update the entry counter */ - rb_remove_entries(cpu_buffer, bpage, read); } ret = read; @@ -2787,6 +3024,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer, out: return ret; } +EXPORT_SYMBOL_GPL(ring_buffer_read_page); static ssize_t rb_simple_read(struct file *filp, char __user *ubuf, @@ -2845,14 +3083,11 @@ static const struct file_operations rb_simple_fops = { static __init int rb_init_debugfs(void) { struct dentry *d_tracer; - struct dentry *entry; d_tracer = tracing_init_dentry(); - entry = debugfs_create_file("tracing_on", 0644, d_tracer, - &ring_buffer_flags, &rb_simple_fops); - if (!entry) - pr_warning("Could not create debugfs 'tracing_on' entry\n"); + trace_create_file("tracing_on", 0644, d_tracer, + &ring_buffer_flags, &rb_simple_fops); return 0; } diff --git a/kernel/trace/ring_buffer_benchmark.c b/kernel/trace/ring_buffer_benchmark.c new file mode 100644 index 00000000000..8d68e149a8b --- /dev/null +++ b/kernel/trace/ring_buffer_benchmark.c @@ -0,0 +1,416 @@ +/* + * ring buffer tester and benchmark + * + * Copyright (C) 2009 Steven Rostedt <srostedt@redhat.com> + */ +#include <linux/ring_buffer.h> +#include <linux/completion.h> +#include <linux/kthread.h> +#include <linux/module.h> +#include <linux/time.h> + +struct rb_page { + u64 ts; + local_t commit; + char data[4080]; +}; + +/* run time and sleep time in seconds */ +#define RUN_TIME 10 +#define SLEEP_TIME 10 + +/* number of events for writer to wake up the reader */ +static int wakeup_interval = 100; + +static int reader_finish; +static struct completion read_start; +static struct completion read_done; + +static struct ring_buffer *buffer; +static struct task_struct *producer; +static struct task_struct *consumer; +static unsigned long read; + +static int disable_reader; +module_param(disable_reader, uint, 0644); +MODULE_PARM_DESC(disable_reader, "only run producer"); + +static int read_events; + +static int kill_test; + +#define KILL_TEST() \ + do { \ + if (!kill_test) { \ + kill_test = 1; \ + WARN_ON(1); \ + } \ + } while (0) + +enum event_status { + EVENT_FOUND, + EVENT_DROPPED, +}; + +static enum event_status read_event(int cpu) +{ + struct ring_buffer_event *event; + int *entry; + u64 ts; + + event = ring_buffer_consume(buffer, cpu, &ts); + if (!event) + return EVENT_DROPPED; + + entry = ring_buffer_event_data(event); + if (*entry != cpu) { + KILL_TEST(); + return EVENT_DROPPED; + } + + read++; + return EVENT_FOUND; +} + +static enum event_status read_page(int cpu) +{ + struct ring_buffer_event *event; + struct rb_page *rpage; + unsigned long commit; + void *bpage; + int *entry; + int ret; + int inc; + int i; + + bpage = ring_buffer_alloc_read_page(buffer); + if (!bpage) + return EVENT_DROPPED; + + ret = ring_buffer_read_page(buffer, &bpage, PAGE_SIZE, cpu, 1); + if (ret >= 0) { + rpage = bpage; + commit = local_read(&rpage->commit); + for (i = 0; i < commit && !kill_test; i += inc) { + + if (i >= (PAGE_SIZE - offsetof(struct rb_page, data))) { + KILL_TEST(); + break; + } + + inc = -1; + event = (void *)&rpage->data[i]; + switch (event->type_len) { + case RINGBUF_TYPE_PADDING: + /* We don't expect any padding */ + KILL_TEST(); + break; + case RINGBUF_TYPE_TIME_EXTEND: + inc = 8; + break; + case 0: + entry = ring_buffer_event_data(event); + if (*entry != cpu) { + KILL_TEST(); + break; + } + read++; + if (!event->array[0]) { + KILL_TEST(); + break; + } + inc = event->array[0]; + break; + default: + entry = ring_buffer_event_data(event); + if (*entry != cpu) { + KILL_TEST(); + break; + } + read++; + inc = ((event->type_len + 1) * 4); + } + if (kill_test) + break; + + if (inc <= 0) { + KILL_TEST(); + break; + } + } + } + ring_buffer_free_read_page(buffer, bpage); + + if (ret < 0) + return EVENT_DROPPED; + return EVENT_FOUND; +} + +static void ring_buffer_consumer(void) +{ + /* toggle between reading pages and events */ + read_events ^= 1; + + read = 0; + while (!reader_finish && !kill_test) { + int found; + + do { + int cpu; + + found = 0; + for_each_online_cpu(cpu) { + enum event_status stat; + + if (read_events) + stat = read_event(cpu); + else + stat = read_page(cpu); + + if (kill_test) + break; + if (stat == EVENT_FOUND) + found = 1; + } + } while (found && !kill_test); + + set_current_state(TASK_INTERRUPTIBLE); + if (reader_finish) + break; + + schedule(); + __set_current_state(TASK_RUNNING); + } + reader_finish = 0; + complete(&read_done); +} + +static void ring_buffer_producer(void) +{ + struct timeval start_tv; + struct timeval end_tv; + unsigned long long time; + unsigned long long entries; + unsigned long long overruns; + unsigned long missed = 0; + unsigned long hit = 0; + unsigned long avg; + int cnt = 0; + + /* + * Hammer the buffer for 10 secs (this may + * make the system stall) + */ + pr_info("Starting ring buffer hammer\n"); + do_gettimeofday(&start_tv); + do { + struct ring_buffer_event *event; + int *entry; + + event = ring_buffer_lock_reserve(buffer, 10); + if (!event) { + missed++; + } else { + hit++; + entry = ring_buffer_event_data(event); + *entry = smp_processor_id(); + ring_buffer_unlock_commit(buffer, event); + } + do_gettimeofday(&end_tv); + + cnt++; + if (consumer && !(cnt % wakeup_interval)) + wake_up_process(consumer); + +#ifndef CONFIG_PREEMPT + /* + * If we are a non preempt kernel, the 10 second run will + * stop everything while it runs. Instead, we will call + * cond_resched and also add any time that was lost by a + * rescedule. + * + * Do a cond resched at the same frequency we would wake up + * the reader. + */ + if (cnt % wakeup_interval) + cond_resched(); +#endif + + } while (end_tv.tv_sec < (start_tv.tv_sec + RUN_TIME) && !kill_test); + pr_info("End ring buffer hammer\n"); + + if (consumer) { + /* Init both completions here to avoid races */ + init_completion(&read_start); + init_completion(&read_done); + /* the completions must be visible before the finish var */ + smp_wmb(); + reader_finish = 1; + /* finish var visible before waking up the consumer */ + smp_wmb(); + wake_up_process(consumer); + wait_for_completion(&read_done); + } + + time = end_tv.tv_sec - start_tv.tv_sec; + time *= USEC_PER_SEC; + time += (long long)((long)end_tv.tv_usec - (long)start_tv.tv_usec); + + entries = ring_buffer_entries(buffer); + overruns = ring_buffer_overruns(buffer); + + if (kill_test) + pr_info("ERROR!\n"); + pr_info("Time: %lld (usecs)\n", time); + pr_info("Overruns: %lld\n", overruns); + if (disable_reader) + pr_info("Read: (reader disabled)\n"); + else + pr_info("Read: %ld (by %s)\n", read, + read_events ? "events" : "pages"); + pr_info("Entries: %lld\n", entries); + pr_info("Total: %lld\n", entries + overruns + read); + pr_info("Missed: %ld\n", missed); + pr_info("Hit: %ld\n", hit); + + /* Convert time from usecs to millisecs */ + do_div(time, USEC_PER_MSEC); + if (time) + hit /= (long)time; + else + pr_info("TIME IS ZERO??\n"); + + pr_info("Entries per millisec: %ld\n", hit); + + if (hit) { + /* Calculate the average time in nanosecs */ + avg = NSEC_PER_MSEC / hit; + pr_info("%ld ns per entry\n", avg); + } + + if (missed) { + if (time) + missed /= (long)time; + + pr_info("Total iterations per millisec: %ld\n", hit + missed); + + /* it is possible that hit + missed will overflow and be zero */ + if (!(hit + missed)) { + pr_info("hit + missed overflowed and totalled zero!\n"); + hit--; /* make it non zero */ + } + + /* Caculate the average time in nanosecs */ + avg = NSEC_PER_MSEC / (hit + missed); + pr_info("%ld ns per entry\n", avg); + } +} + +static void wait_to_die(void) +{ + set_current_state(TASK_INTERRUPTIBLE); + while (!kthread_should_stop()) { + schedule(); + set_current_state(TASK_INTERRUPTIBLE); + } + __set_current_state(TASK_RUNNING); +} + +static int ring_buffer_consumer_thread(void *arg) +{ + while (!kthread_should_stop() && !kill_test) { + complete(&read_start); + + ring_buffer_consumer(); + + set_current_state(TASK_INTERRUPTIBLE); + if (kthread_should_stop() || kill_test) + break; + + schedule(); + __set_current_state(TASK_RUNNING); + } + __set_current_state(TASK_RUNNING); + + if (kill_test) + wait_to_die(); + + return 0; +} + +static int ring_buffer_producer_thread(void *arg) +{ + init_completion(&read_start); + + while (!kthread_should_stop() && !kill_test) { + ring_buffer_reset(buffer); + + if (consumer) { + smp_wmb(); + wake_up_process(consumer); + wait_for_completion(&read_start); + } + + ring_buffer_producer(); + + pr_info("Sleeping for 10 secs\n"); + set_current_state(TASK_INTERRUPTIBLE); + schedule_timeout(HZ * SLEEP_TIME); + __set_current_state(TASK_RUNNING); + } + + if (kill_test) + wait_to_die(); + + return 0; +} + +static int __init ring_buffer_benchmark_init(void) +{ + int ret; + + /* make a one meg buffer in overwite mode */ + buffer = ring_buffer_alloc(1000000, RB_FL_OVERWRITE); + if (!buffer) + return -ENOMEM; + + if (!disable_reader) { + consumer = kthread_create(ring_buffer_consumer_thread, + NULL, "rb_consumer"); + ret = PTR_ERR(consumer); + if (IS_ERR(consumer)) + goto out_fail; + } + + producer = kthread_run(ring_buffer_producer_thread, + NULL, "rb_producer"); + ret = PTR_ERR(producer); + + if (IS_ERR(producer)) + goto out_kill; + + return 0; + + out_kill: + if (consumer) + kthread_stop(consumer); + + out_fail: + ring_buffer_free(buffer); + return ret; +} + +static void __exit ring_buffer_benchmark_exit(void) +{ + kthread_stop(producer); + if (consumer) + kthread_stop(consumer); + ring_buffer_free(buffer); +} + +module_init(ring_buffer_benchmark_init); +module_exit(ring_buffer_benchmark_exit); + +MODULE_AUTHOR("Steven Rostedt"); +MODULE_DESCRIPTION("ring_buffer_benchmark"); +MODULE_LICENSE("GPL"); diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index cda81ec58d9..8acd9b81a5d 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -171,6 +171,13 @@ static struct trace_array global_trace; static DEFINE_PER_CPU(struct trace_array_cpu, global_trace_cpu); +int filter_current_check_discard(struct ftrace_event_call *call, void *rec, + struct ring_buffer_event *event) +{ + return filter_check_discard(call, rec, global_trace.buffer, event); +} +EXPORT_SYMBOL_GPL(filter_current_check_discard); + cycle_t ftrace_now(int cpu) { u64 ts; @@ -255,7 +262,8 @@ static DECLARE_WAIT_QUEUE_HEAD(trace_wait); /* trace_flags holds trace_options default values */ unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK | - TRACE_ITER_ANNOTATE | TRACE_ITER_CONTEXT_INFO | TRACE_ITER_SLEEP_TIME; + TRACE_ITER_ANNOTATE | TRACE_ITER_CONTEXT_INFO | TRACE_ITER_SLEEP_TIME | + TRACE_ITER_GRAPH_TIME; /** * trace_wake_up - wake up tasks waiting for trace input @@ -317,6 +325,7 @@ static const char *trace_options[] = { "latency-format", "global-clock", "sleep-time", + "graph-time", NULL }; @@ -402,17 +411,6 @@ static ssize_t trace_seq_to_buffer(struct trace_seq *s, void *buf, size_t cnt) return cnt; } -static void -trace_print_seq(struct seq_file *m, struct trace_seq *s) -{ - int len = s->len >= PAGE_SIZE ? PAGE_SIZE - 1 : s->len; - - s->buffer[len] = 0; - seq_puts(m, s->buffer); - - trace_seq_init(s); -} - /** * update_max_tr - snapshot all trace buffers from global_trace to max_tr * @tr: tracer @@ -641,6 +639,16 @@ void tracing_reset_online_cpus(struct trace_array *tr) tracing_reset(tr, cpu); } +void tracing_reset_current(int cpu) +{ + tracing_reset(&global_trace, cpu); +} + +void tracing_reset_current_online_cpus(void) +{ + tracing_reset_online_cpus(&global_trace); +} + #define SAVED_CMDLINES 128 #define NO_CMDLINE_MAP UINT_MAX static unsigned map_pid_to_cmdline[PID_MAX_DEFAULT+1]; @@ -800,6 +808,7 @@ void trace_find_cmdline(int pid, char comm[]) return; } + preempt_disable(); __raw_spin_lock(&trace_cmdline_lock); map = map_pid_to_cmdline[pid]; if (map != NO_CMDLINE_MAP) @@ -808,6 +817,7 @@ void trace_find_cmdline(int pid, char comm[]) strcpy(comm, "<...>"); __raw_spin_unlock(&trace_cmdline_lock); + preempt_enable(); } void tracing_record_cmdline(struct task_struct *tsk) @@ -840,7 +850,7 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags, } struct ring_buffer_event *trace_buffer_lock_reserve(struct trace_array *tr, - unsigned char type, + int type, unsigned long len, unsigned long flags, int pc) { @@ -883,30 +893,40 @@ void trace_buffer_unlock_commit(struct trace_array *tr, } struct ring_buffer_event * -trace_current_buffer_lock_reserve(unsigned char type, unsigned long len, +trace_current_buffer_lock_reserve(int type, unsigned long len, unsigned long flags, int pc) { return trace_buffer_lock_reserve(&global_trace, type, len, flags, pc); } +EXPORT_SYMBOL_GPL(trace_current_buffer_lock_reserve); void trace_current_buffer_unlock_commit(struct ring_buffer_event *event, unsigned long flags, int pc) { - return __trace_buffer_unlock_commit(&global_trace, event, flags, pc, 1); + __trace_buffer_unlock_commit(&global_trace, event, flags, pc, 1); } +EXPORT_SYMBOL_GPL(trace_current_buffer_unlock_commit); void trace_nowake_buffer_unlock_commit(struct ring_buffer_event *event, unsigned long flags, int pc) { - return __trace_buffer_unlock_commit(&global_trace, event, flags, pc, 0); + __trace_buffer_unlock_commit(&global_trace, event, flags, pc, 0); +} +EXPORT_SYMBOL_GPL(trace_nowake_buffer_unlock_commit); + +void trace_current_buffer_discard_commit(struct ring_buffer_event *event) +{ + ring_buffer_discard_commit(global_trace.buffer, event); } +EXPORT_SYMBOL_GPL(trace_current_buffer_discard_commit); void trace_function(struct trace_array *tr, unsigned long ip, unsigned long parent_ip, unsigned long flags, int pc) { + struct ftrace_event_call *call = &event_function; struct ring_buffer_event *event; struct ftrace_entry *entry; @@ -921,7 +941,9 @@ trace_function(struct trace_array *tr, entry = ring_buffer_event_data(event); entry->ip = ip; entry->parent_ip = parent_ip; - ring_buffer_unlock_commit(tr->buffer, event); + + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); } #ifdef CONFIG_FUNCTION_GRAPH_TRACER @@ -930,6 +952,7 @@ static int __trace_graph_entry(struct trace_array *tr, unsigned long flags, int pc) { + struct ftrace_event_call *call = &event_funcgraph_entry; struct ring_buffer_event *event; struct ftrace_graph_ent_entry *entry; @@ -942,7 +965,8 @@ static int __trace_graph_entry(struct trace_array *tr, return 0; entry = ring_buffer_event_data(event); entry->graph_ent = *trace; - ring_buffer_unlock_commit(global_trace.buffer, event); + if (!filter_current_check_discard(call, entry, event)) + ring_buffer_unlock_commit(global_trace.buffer, event); return 1; } @@ -952,6 +976,7 @@ static void __trace_graph_return(struct trace_array *tr, unsigned long flags, int pc) { + struct ftrace_event_call *call = &event_funcgraph_exit; struct ring_buffer_event *event; struct ftrace_graph_ret_entry *entry; @@ -964,7 +989,8 @@ static void __trace_graph_return(struct trace_array *tr, return; entry = ring_buffer_event_data(event); entry->ret = *trace; - ring_buffer_unlock_commit(global_trace.buffer, event); + if (!filter_current_check_discard(call, entry, event)) + ring_buffer_unlock_commit(global_trace.buffer, event); } #endif @@ -982,6 +1008,7 @@ static void __ftrace_trace_stack(struct trace_array *tr, int skip, int pc) { #ifdef CONFIG_STACKTRACE + struct ftrace_event_call *call = &event_kernel_stack; struct ring_buffer_event *event; struct stack_entry *entry; struct stack_trace trace; @@ -999,7 +1026,8 @@ static void __ftrace_trace_stack(struct trace_array *tr, trace.entries = entry->caller; save_stack_trace(&trace); - ring_buffer_unlock_commit(tr->buffer, event); + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); #endif } @@ -1024,6 +1052,7 @@ static void ftrace_trace_userstack(struct trace_array *tr, unsigned long flags, int pc) { #ifdef CONFIG_STACKTRACE + struct ftrace_event_call *call = &event_user_stack; struct ring_buffer_event *event; struct userstack_entry *entry; struct stack_trace trace; @@ -1045,7 +1074,8 @@ static void ftrace_trace_userstack(struct trace_array *tr, trace.entries = entry->caller; save_stack_trace_user(&trace); - ring_buffer_unlock_commit(tr->buffer, event); + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); #endif } @@ -1089,6 +1119,7 @@ tracing_sched_switch_trace(struct trace_array *tr, struct task_struct *next, unsigned long flags, int pc) { + struct ftrace_event_call *call = &event_context_switch; struct ring_buffer_event *event; struct ctx_switch_entry *entry; @@ -1104,7 +1135,9 @@ tracing_sched_switch_trace(struct trace_array *tr, entry->next_prio = next->prio; entry->next_state = next->state; entry->next_cpu = task_cpu(next); - trace_buffer_unlock_commit(tr, event, flags, pc); + + if (!filter_check_discard(call, entry, tr->buffer, event)) + trace_buffer_unlock_commit(tr, event, flags, pc); } void @@ -1113,6 +1146,7 @@ tracing_sched_wakeup_trace(struct trace_array *tr, struct task_struct *curr, unsigned long flags, int pc) { + struct ftrace_event_call *call = &event_wakeup; struct ring_buffer_event *event; struct ctx_switch_entry *entry; @@ -1129,7 +1163,8 @@ tracing_sched_wakeup_trace(struct trace_array *tr, entry->next_state = wakee->state; entry->next_cpu = task_cpu(wakee); - ring_buffer_unlock_commit(tr->buffer, event); + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); ftrace_trace_stack(tr, flags, 6, pc); ftrace_trace_userstack(tr, flags, pc); } @@ -1230,11 +1265,13 @@ int trace_vbprintk(unsigned long ip, const char *fmt, va_list args) (raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED; static u32 trace_buf[TRACE_BUF_SIZE]; + struct ftrace_event_call *call = &event_bprint; struct ring_buffer_event *event; struct trace_array *tr = &global_trace; struct trace_array_cpu *data; struct bprint_entry *entry; unsigned long flags; + int disable; int resched; int cpu, len = 0, size, pc; @@ -1249,7 +1286,8 @@ int trace_vbprintk(unsigned long ip, const char *fmt, va_list args) cpu = raw_smp_processor_id(); data = tr->data[cpu]; - if (unlikely(atomic_read(&data->disabled))) + disable = atomic_inc_return(&data->disabled); + if (unlikely(disable != 1)) goto out; /* Lockdep uses trace_printk for lock tracing */ @@ -1269,13 +1307,15 @@ int trace_vbprintk(unsigned long ip, const char *fmt, va_list args) entry->fmt = fmt; memcpy(entry->buf, trace_buf, sizeof(u32) * len); - ring_buffer_unlock_commit(tr->buffer, event); + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); out_unlock: __raw_spin_unlock(&trace_buf_lock); local_irq_restore(flags); out: + atomic_dec_return(&data->disabled); ftrace_preempt_enable(resched); unpause_graph_tracing(); @@ -1288,12 +1328,14 @@ int trace_vprintk(unsigned long ip, const char *fmt, va_list args) static raw_spinlock_t trace_buf_lock = __RAW_SPIN_LOCK_UNLOCKED; static char trace_buf[TRACE_BUF_SIZE]; + struct ftrace_event_call *call = &event_print; struct ring_buffer_event *event; struct trace_array *tr = &global_trace; struct trace_array_cpu *data; int cpu, len = 0, size, pc; struct print_entry *entry; unsigned long irq_flags; + int disable; if (tracing_disabled || tracing_selftest_running) return 0; @@ -1303,7 +1345,8 @@ int trace_vprintk(unsigned long ip, const char *fmt, va_list args) cpu = raw_smp_processor_id(); data = tr->data[cpu]; - if (unlikely(atomic_read(&data->disabled))) + disable = atomic_inc_return(&data->disabled); + if (unlikely(disable != 1)) goto out; pause_graph_tracing(); @@ -1323,13 +1366,15 @@ int trace_vprintk(unsigned long ip, const char *fmt, va_list args) memcpy(&entry->buf, trace_buf, len); entry->buf[len] = 0; - ring_buffer_unlock_commit(tr->buffer, event); + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); out_unlock: __raw_spin_unlock(&trace_buf_lock); raw_local_irq_restore(irq_flags); unpause_graph_tracing(); out: + atomic_dec_return(&data->disabled); preempt_enable_notrace(); return len; @@ -1526,12 +1571,14 @@ static void *s_start(struct seq_file *m, loff_t *pos) p = s_next(m, p, &l); } + trace_event_read_lock(); return p; } static void s_stop(struct seq_file *m, void *p) { atomic_dec(&trace_record_cmdline_disabled); + trace_event_read_unlock(); } static void print_lat_help_header(struct seq_file *m) @@ -1774,6 +1821,7 @@ static int trace_empty(struct trace_iterator *iter) return 1; } +/* Called with trace_event_read_lock() held. */ static enum print_line_t print_trace_line(struct trace_iterator *iter) { enum print_line_t ret; @@ -2397,6 +2445,56 @@ static const struct file_operations tracing_readme_fops = { }; static ssize_t +tracing_saved_cmdlines_read(struct file *file, char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + char *buf_comm; + char *file_buf; + char *buf; + int len = 0; + int pid; + int i; + + file_buf = kmalloc(SAVED_CMDLINES*(16+TASK_COMM_LEN), GFP_KERNEL); + if (!file_buf) + return -ENOMEM; + + buf_comm = kmalloc(TASK_COMM_LEN, GFP_KERNEL); + if (!buf_comm) { + kfree(file_buf); + return -ENOMEM; + } + + buf = file_buf; + + for (i = 0; i < SAVED_CMDLINES; i++) { + int r; + + pid = map_cmdline_to_pid[i]; + if (pid == -1 || pid == NO_CMDLINE_MAP) + continue; + + trace_find_cmdline(pid, buf_comm); + r = sprintf(buf, "%d %s\n", pid, buf_comm); + buf += r; + len += r; + } + + len = simple_read_from_buffer(ubuf, cnt, ppos, + file_buf, len); + + kfree(file_buf); + kfree(buf_comm); + + return len; +} + +static const struct file_operations tracing_saved_cmdlines_fops = { + .open = tracing_open_generic, + .read = tracing_saved_cmdlines_read, +}; + +static ssize_t tracing_ctrl_read(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) { @@ -2728,6 +2826,9 @@ static int tracing_open_pipe(struct inode *inode, struct file *filp) /* trace pipe does not show start of buffer */ cpumask_setall(iter->started); + if (trace_flags & TRACE_ITER_LATENCY_FMT) + iter->iter_flags |= TRACE_FILE_LAT_FMT; + iter->cpu_file = cpu_file; iter->tr = &global_trace; mutex_init(&iter->mutex); @@ -2915,6 +3016,7 @@ waitagain: offsetof(struct trace_iterator, seq)); iter->pos = -1; + trace_event_read_lock(); while (find_next_entry_inc(iter) != NULL) { enum print_line_t ret; int len = iter->seq.len; @@ -2931,6 +3033,7 @@ waitagain: if (iter->seq.len >= cnt) break; } + trace_event_read_unlock(); /* Now copy what we have to the user */ sret = trace_seq_to_user(&iter->seq, ubuf, cnt); @@ -3053,6 +3156,8 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, goto out_err; } + trace_event_read_lock(); + /* Fill as many pages as possible. */ for (i = 0, rem = len; i < PIPE_BUFFERS && rem; i++) { pages[i] = alloc_page(GFP_KERNEL); @@ -3075,6 +3180,7 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, trace_seq_init(&iter->seq); } + trace_event_read_unlock(); mutex_unlock(&iter->mutex); spd.nr_pages = i; @@ -3425,7 +3531,7 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, .spd_release = buffer_spd_release, }; struct buffer_ref *ref; - int size, i; + int entries, size, i; size_t ret; if (*ppos & (PAGE_SIZE - 1)) { @@ -3440,7 +3546,9 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, len &= PAGE_MASK; } - for (i = 0; i < PIPE_BUFFERS && len; i++, len -= PAGE_SIZE) { + entries = ring_buffer_entries_cpu(info->tr->buffer, info->cpu); + + for (i = 0; i < PIPE_BUFFERS && len && entries; i++, len -= PAGE_SIZE) { struct page *page; int r; @@ -3457,7 +3565,7 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, } r = ring_buffer_read_page(ref->buffer, &ref->page, - len, info->cpu, 0); + len, info->cpu, 1); if (r < 0) { ring_buffer_free_read_page(ref->buffer, ref->page); @@ -3481,6 +3589,8 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, spd.partial[i].private = (unsigned long)ref; spd.nr_pages++; *ppos += PAGE_SIZE; + + entries = ring_buffer_entries_cpu(info->tr->buffer, info->cpu); } spd.nr_pages = i; @@ -3508,6 +3618,45 @@ static const struct file_operations tracing_buffers_fops = { .llseek = no_llseek, }; +static ssize_t +tracing_stats_read(struct file *filp, char __user *ubuf, + size_t count, loff_t *ppos) +{ + unsigned long cpu = (unsigned long)filp->private_data; + struct trace_array *tr = &global_trace; + struct trace_seq *s; + unsigned long cnt; + + s = kmalloc(sizeof(*s), GFP_ATOMIC); + if (!s) + return ENOMEM; + + trace_seq_init(s); + + cnt = ring_buffer_entries_cpu(tr->buffer, cpu); + trace_seq_printf(s, "entries: %ld\n", cnt); + + cnt = ring_buffer_overrun_cpu(tr->buffer, cpu); + trace_seq_printf(s, "overrun: %ld\n", cnt); + + cnt = ring_buffer_commit_overrun_cpu(tr->buffer, cpu); + trace_seq_printf(s, "commit overrun: %ld\n", cnt); + + cnt = ring_buffer_nmi_dropped_cpu(tr->buffer, cpu); + trace_seq_printf(s, "nmi dropped: %ld\n", cnt); + + count = simple_read_from_buffer(ubuf, count, ppos, s->buffer, s->len); + + kfree(s); + + return count; +} + +static const struct file_operations tracing_stats_fops = { + .open = tracing_open_generic, + .read = tracing_stats_read, +}; + #ifdef CONFIG_DYNAMIC_FTRACE int __weak ftrace_arch_read_dyn_info(char *buf, int size) @@ -3597,7 +3746,7 @@ struct dentry *tracing_dentry_percpu(void) static void tracing_init_debugfs_percpu(long cpu) { struct dentry *d_percpu = tracing_dentry_percpu(); - struct dentry *entry, *d_cpu; + struct dentry *d_cpu; /* strlen(cpu) + MAX(log10(cpu)) + '\0' */ char cpu_dir[7]; @@ -3612,21 +3761,18 @@ static void tracing_init_debugfs_percpu(long cpu) } /* per cpu trace_pipe */ - entry = debugfs_create_file("trace_pipe", 0444, d_cpu, - (void *) cpu, &tracing_pipe_fops); - if (!entry) - pr_warning("Could not create debugfs 'trace_pipe' entry\n"); + trace_create_file("trace_pipe", 0444, d_cpu, + (void *) cpu, &tracing_pipe_fops); /* per cpu trace */ - entry = debugfs_create_file("trace", 0644, d_cpu, - (void *) cpu, &tracing_fops); - if (!entry) - pr_warning("Could not create debugfs 'trace' entry\n"); + trace_create_file("trace", 0644, d_cpu, + (void *) cpu, &tracing_fops); + + trace_create_file("trace_pipe_raw", 0444, d_cpu, + (void *) cpu, &tracing_buffers_fops); - entry = debugfs_create_file("trace_pipe_raw", 0444, d_cpu, - (void *) cpu, &tracing_buffers_fops); - if (!entry) - pr_warning("Could not create debugfs 'trace_pipe_raw' entry\n"); + trace_create_file("stats", 0444, d_cpu, + (void *) cpu, &tracing_stats_fops); } #ifdef CONFIG_FTRACE_SELFTEST @@ -3782,6 +3928,22 @@ static const struct file_operations trace_options_core_fops = { .write = trace_options_core_write, }; +struct dentry *trace_create_file(const char *name, + mode_t mode, + struct dentry *parent, + void *data, + const struct file_operations *fops) +{ + struct dentry *ret; + + ret = debugfs_create_file(name, mode, parent, data, fops); + if (!ret) + pr_warning("Could not create debugfs '%s' entry\n", name); + + return ret; +} + + static struct dentry *trace_options_init_dentry(void) { struct dentry *d_tracer; @@ -3809,7 +3971,6 @@ create_trace_option_file(struct trace_option_dentry *topt, struct tracer_opt *opt) { struct dentry *t_options; - struct dentry *entry; t_options = trace_options_init_dentry(); if (!t_options) @@ -3818,11 +3979,9 @@ create_trace_option_file(struct trace_option_dentry *topt, topt->flags = flags; topt->opt = opt; - entry = debugfs_create_file(opt->name, 0644, t_options, topt, + topt->entry = trace_create_file(opt->name, 0644, t_options, topt, &trace_options_fops); - topt->entry = entry; - } static struct trace_option_dentry * @@ -3877,123 +4036,84 @@ static struct dentry * create_trace_option_core_file(const char *option, long index) { struct dentry *t_options; - struct dentry *entry; t_options = trace_options_init_dentry(); if (!t_options) return NULL; - entry = debugfs_create_file(option, 0644, t_options, (void *)index, + return trace_create_file(option, 0644, t_options, (void *)index, &trace_options_core_fops); - - return entry; } static __init void create_trace_options_dir(void) { struct dentry *t_options; - struct dentry *entry; int i; t_options = trace_options_init_dentry(); if (!t_options) return; - for (i = 0; trace_options[i]; i++) { - entry = create_trace_option_core_file(trace_options[i], i); - if (!entry) - pr_warning("Could not create debugfs %s entry\n", - trace_options[i]); - } + for (i = 0; trace_options[i]; i++) + create_trace_option_core_file(trace_options[i], i); } static __init int tracer_init_debugfs(void) { struct dentry *d_tracer; - struct dentry *entry; int cpu; d_tracer = tracing_init_dentry(); - entry = debugfs_create_file("tracing_enabled", 0644, d_tracer, - &global_trace, &tracing_ctrl_fops); - if (!entry) - pr_warning("Could not create debugfs 'tracing_enabled' entry\n"); + trace_create_file("tracing_enabled", 0644, d_tracer, + &global_trace, &tracing_ctrl_fops); - entry = debugfs_create_file("trace_options", 0644, d_tracer, - NULL, &tracing_iter_fops); - if (!entry) - pr_warning("Could not create debugfs 'trace_options' entry\n"); + trace_create_file("trace_options", 0644, d_tracer, + NULL, &tracing_iter_fops); - create_trace_options_dir(); + trace_create_file("tracing_cpumask", 0644, d_tracer, + NULL, &tracing_cpumask_fops); + + trace_create_file("trace", 0644, d_tracer, + (void *) TRACE_PIPE_ALL_CPU, &tracing_fops); - entry = debugfs_create_file("tracing_cpumask", 0644, d_tracer, - NULL, &tracing_cpumask_fops); - if (!entry) - pr_warning("Could not create debugfs 'tracing_cpumask' entry\n"); - - entry = debugfs_create_file("trace", 0644, d_tracer, - (void *) TRACE_PIPE_ALL_CPU, &tracing_fops); - if (!entry) - pr_warning("Could not create debugfs 'trace' entry\n"); - - entry = debugfs_create_file("available_tracers", 0444, d_tracer, - &global_trace, &show_traces_fops); - if (!entry) - pr_warning("Could not create debugfs 'available_tracers' entry\n"); - - entry = debugfs_create_file("current_tracer", 0444, d_tracer, - &global_trace, &set_tracer_fops); - if (!entry) - pr_warning("Could not create debugfs 'current_tracer' entry\n"); - - entry = debugfs_create_file("tracing_max_latency", 0644, d_tracer, - &tracing_max_latency, - &tracing_max_lat_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'tracing_max_latency' entry\n"); - - entry = debugfs_create_file("tracing_thresh", 0644, d_tracer, - &tracing_thresh, &tracing_max_lat_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'tracing_thresh' entry\n"); - entry = debugfs_create_file("README", 0644, d_tracer, - NULL, &tracing_readme_fops); - if (!entry) - pr_warning("Could not create debugfs 'README' entry\n"); - - entry = debugfs_create_file("trace_pipe", 0444, d_tracer, + trace_create_file("available_tracers", 0444, d_tracer, + &global_trace, &show_traces_fops); + + trace_create_file("current_tracer", 0644, d_tracer, + &global_trace, &set_tracer_fops); + + trace_create_file("tracing_max_latency", 0644, d_tracer, + &tracing_max_latency, &tracing_max_lat_fops); + + trace_create_file("tracing_thresh", 0644, d_tracer, + &tracing_thresh, &tracing_max_lat_fops); + + trace_create_file("README", 0444, d_tracer, + NULL, &tracing_readme_fops); + + trace_create_file("trace_pipe", 0444, d_tracer, (void *) TRACE_PIPE_ALL_CPU, &tracing_pipe_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'trace_pipe' entry\n"); - - entry = debugfs_create_file("buffer_size_kb", 0644, d_tracer, - &global_trace, &tracing_entries_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'buffer_size_kb' entry\n"); - - entry = debugfs_create_file("trace_marker", 0220, d_tracer, - NULL, &tracing_mark_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'trace_marker' entry\n"); + + trace_create_file("buffer_size_kb", 0644, d_tracer, + &global_trace, &tracing_entries_fops); + + trace_create_file("trace_marker", 0220, d_tracer, + NULL, &tracing_mark_fops); + + trace_create_file("saved_cmdlines", 0444, d_tracer, + NULL, &tracing_saved_cmdlines_fops); #ifdef CONFIG_DYNAMIC_FTRACE - entry = debugfs_create_file("dyn_ftrace_total_info", 0444, d_tracer, - &ftrace_update_tot_cnt, - &tracing_dyn_info_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'dyn_ftrace_total_info' entry\n"); + trace_create_file("dyn_ftrace_total_info", 0444, d_tracer, + &ftrace_update_tot_cnt, &tracing_dyn_info_fops); #endif #ifdef CONFIG_SYSPROF_TRACER init_tracer_sysprof_debugfs(d_tracer); #endif + create_trace_options_dir(); + for_each_tracing_cpu(cpu) tracing_init_debugfs_percpu(cpu); @@ -4064,7 +4184,8 @@ trace_printk_seq(struct trace_seq *s) static void __ftrace_dump(bool disable_tracing) { - static DEFINE_SPINLOCK(ftrace_dump_lock); + static raw_spinlock_t ftrace_dump_lock = + (raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED; /* use static because iter can be a bit big for the stack */ static struct trace_iterator iter; unsigned int old_userobj; @@ -4073,7 +4194,8 @@ static void __ftrace_dump(bool disable_tracing) int cnt = 0, cpu; /* only one dump */ - spin_lock_irqsave(&ftrace_dump_lock, flags); + local_irq_save(flags); + __raw_spin_lock(&ftrace_dump_lock); if (dump_ran) goto out; @@ -4145,7 +4267,8 @@ static void __ftrace_dump(bool disable_tracing) } out: - spin_unlock_irqrestore(&ftrace_dump_lock, flags); + __raw_spin_unlock(&ftrace_dump_lock); + local_irq_restore(flags); } /* By default: disable tracing after the dump */ diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index e685ac2b2ba..6e735d4771f 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -9,9 +9,12 @@ #include <linux/mmiotrace.h> #include <linux/ftrace.h> #include <trace/boot.h> -#include <trace/kmemtrace.h> +#include <linux/kmemtrace.h> #include <trace/power.h> +#include <linux/trace_seq.h> +#include <linux/ftrace_event.h> + enum trace_type { __TRACE_FIRST_TYPE = 0, @@ -42,20 +45,6 @@ enum trace_type { }; /* - * The trace entry - the most basic unit of tracing. This is what - * is printed in the end as a single line in the trace output, such as: - * - * bash-15816 [01] 235.197585: idle_cpu <- irq_enter - */ -struct trace_entry { - unsigned char type; - unsigned char flags; - unsigned char preempt_count; - int pid; - int tgid; -}; - -/* * Function trace entry - function address and parent function addres: */ struct ftrace_entry { @@ -263,8 +252,6 @@ struct trace_array_cpu { char comm[TASK_COMM_LEN]; }; -struct trace_iterator; - /* * The trace array - an array of per-CPU trace arrays. This is the * highest level data structure that individual tracers deal with. @@ -339,15 +326,6 @@ extern void __ftrace_bad_type(void); __ftrace_bad_type(); \ } while (0) -/* Return values for print_line callback */ -enum print_line_t { - TRACE_TYPE_PARTIAL_LINE = 0, /* Retry after flushing the seq */ - TRACE_TYPE_HANDLED = 1, - TRACE_TYPE_UNHANDLED = 2, /* Relay to other output functions */ - TRACE_TYPE_NO_CONSUME = 3 /* Handled but ask to not consume */ -}; - - /* * An option specific to a tracer. This is a boolean value. * The bit is the bit index that sets its value on the @@ -423,60 +401,30 @@ struct tracer { struct tracer_stat *stats; }; -struct trace_seq { - unsigned char buffer[PAGE_SIZE]; - unsigned int len; - unsigned int readpos; -}; - -static inline void -trace_seq_init(struct trace_seq *s) -{ - s->len = 0; - s->readpos = 0; -} - #define TRACE_PIPE_ALL_CPU -1 -/* - * Trace iterator - used by printout routines who present trace - * results to users and which routines might sleep, etc: - */ -struct trace_iterator { - struct trace_array *tr; - struct tracer *trace; - void *private; - int cpu_file; - struct mutex mutex; - struct ring_buffer_iter *buffer_iter[NR_CPUS]; - - /* The below is zeroed out in pipe_read */ - struct trace_seq seq; - struct trace_entry *ent; - int cpu; - u64 ts; - - unsigned long iter_flags; - loff_t pos; - long idx; - - cpumask_var_t started; -}; - int tracer_init(struct tracer *t, struct trace_array *tr); int tracing_is_enabled(void); void trace_wake_up(void); void tracing_reset(struct trace_array *tr, int cpu); void tracing_reset_online_cpus(struct trace_array *tr); +void tracing_reset_current(int cpu); +void tracing_reset_current_online_cpus(void); int tracing_open_generic(struct inode *inode, struct file *filp); +struct dentry *trace_create_file(const char *name, + mode_t mode, + struct dentry *parent, + void *data, + const struct file_operations *fops); + struct dentry *tracing_init_dentry(void); void init_tracer_sysprof_debugfs(struct dentry *d_tracer); struct ring_buffer_event; struct ring_buffer_event *trace_buffer_lock_reserve(struct trace_array *tr, - unsigned char type, + int type, unsigned long len, unsigned long flags, int pc); @@ -484,14 +432,6 @@ void trace_buffer_unlock_commit(struct trace_array *tr, struct ring_buffer_event *event, unsigned long flags, int pc); -struct ring_buffer_event * -trace_current_buffer_lock_reserve(unsigned char type, unsigned long len, - unsigned long flags, int pc); -void trace_current_buffer_unlock_commit(struct ring_buffer_event *event, - unsigned long flags, int pc); -void trace_nowake_buffer_unlock_commit(struct ring_buffer_event *event, - unsigned long flags, int pc); - struct trace_entry *tracing_get_trace_entry(struct trace_array *tr, struct trace_array_cpu *data); @@ -514,7 +454,6 @@ void tracing_sched_switch_trace(struct trace_array *tr, struct task_struct *prev, struct task_struct *next, unsigned long flags, int pc); -void tracing_record_cmdline(struct task_struct *tsk); void tracing_sched_wakeup_trace(struct trace_array *tr, struct task_struct *wakee, @@ -599,6 +538,8 @@ extern int trace_selftest_startup_sysprof(struct tracer *trace, struct trace_array *tr); extern int trace_selftest_startup_branch(struct tracer *trace, struct trace_array *tr); +extern int trace_selftest_startup_hw_branches(struct tracer *trace, + struct trace_array *tr); #endif /* CONFIG_FTRACE_STARTUP_TEST */ extern void *head_page(struct trace_array_cpu *data); @@ -613,6 +554,8 @@ extern unsigned long trace_flags; /* Standard output formatting function used for function return traces */ #ifdef CONFIG_FUNCTION_GRAPH_TRACER extern enum print_line_t print_graph_function(struct trace_iterator *iter); +extern enum print_line_t +trace_print_graph_duration(unsigned long long duration, struct trace_seq *s); #ifdef CONFIG_DYNAMIC_FTRACE /* TODO: make this variable */ @@ -644,7 +587,6 @@ static inline int ftrace_graph_addr(unsigned long addr) return 1; } #endif /* CONFIG_DYNAMIC_FTRACE */ - #else /* CONFIG_FUNCTION_GRAPH_TRACER */ static inline enum print_line_t print_graph_function(struct trace_iterator *iter) @@ -692,6 +634,7 @@ enum trace_iterator_flags { TRACE_ITER_LATENCY_FMT = 0x40000, TRACE_ITER_GLOBAL_CLK = 0x80000, TRACE_ITER_SLEEP_TIME = 0x100000, + TRACE_ITER_GRAPH_TIME = 0x200000, }; /* @@ -790,103 +733,113 @@ struct ftrace_event_field { char *type; int offset; int size; + int is_signed; }; -struct ftrace_event_call { - char *name; - char *system; - struct dentry *dir; - int enabled; - int (*regfunc)(void); - void (*unregfunc)(void); - int id; - int (*raw_init)(void); - int (*show_format)(struct trace_seq *s); - int (*define_fields)(void); - struct list_head fields; +struct event_filter { + int n_preds; struct filter_pred **preds; - -#ifdef CONFIG_EVENT_PROFILE - atomic_t profile_count; - int (*profile_enable)(struct ftrace_event_call *); - void (*profile_disable)(struct ftrace_event_call *); -#endif + char *filter_string; }; struct event_subsystem { struct list_head list; const char *name; struct dentry *entry; - struct filter_pred **preds; + void *filter; }; -#define events_for_each(event) \ - for (event = __start_ftrace_events; \ - (unsigned long)event < (unsigned long)__stop_ftrace_events; \ - event++) - -#define MAX_FILTER_PRED 8 - struct filter_pred; -typedef int (*filter_pred_fn_t) (struct filter_pred *pred, void *event); +typedef int (*filter_pred_fn_t) (struct filter_pred *pred, void *event, + int val1, int val2); struct filter_pred { filter_pred_fn_t fn; u64 val; - char *str_val; + char str_val[MAX_FILTER_STR_VAL]; int str_len; char *field_name; int offset; int not; - int or; - int compound; - int clear; + int op; + int pop_n; }; -int trace_define_field(struct ftrace_event_call *call, char *type, - char *name, int offset, int size); -extern void filter_free_pred(struct filter_pred *pred); -extern void filter_print_preds(struct filter_pred **preds, +extern void print_event_filter(struct ftrace_event_call *call, struct trace_seq *s); -extern int filter_parse(char **pbuf, struct filter_pred *pred); -extern int filter_add_pred(struct ftrace_event_call *call, - struct filter_pred *pred); -extern void filter_free_preds(struct ftrace_event_call *call); -extern int filter_match_preds(struct ftrace_event_call *call, void *rec); -extern void filter_free_subsystem_preds(struct event_subsystem *system); -extern int filter_add_subsystem_pred(struct event_subsystem *system, - struct filter_pred *pred); - -void event_trace_printk(unsigned long ip, const char *fmt, ...); -extern struct ftrace_event_call __start_ftrace_events[]; -extern struct ftrace_event_call __stop_ftrace_events[]; - -#define for_each_event(event) \ - for (event = __start_ftrace_events; \ - (unsigned long)event < (unsigned long)__stop_ftrace_events; \ - event++) +extern int apply_event_filter(struct ftrace_event_call *call, + char *filter_string); +extern int apply_subsystem_event_filter(struct event_subsystem *system, + char *filter_string); +extern void print_subsystem_event_filter(struct event_subsystem *system, + struct trace_seq *s); + +static inline int +filter_check_discard(struct ftrace_event_call *call, void *rec, + struct ring_buffer *buffer, + struct ring_buffer_event *event) +{ + if (unlikely(call->filter_active) && !filter_match_preds(call, rec)) { + ring_buffer_discard_commit(buffer, event); + return 1; + } + + return 0; +} + +#define DEFINE_COMPARISON_PRED(type) \ +static int filter_pred_##type(struct filter_pred *pred, void *event, \ + int val1, int val2) \ +{ \ + type *addr = (type *)(event + pred->offset); \ + type val = (type)pred->val; \ + int match = 0; \ + \ + switch (pred->op) { \ + case OP_LT: \ + match = (*addr < val); \ + break; \ + case OP_LE: \ + match = (*addr <= val); \ + break; \ + case OP_GT: \ + match = (*addr > val); \ + break; \ + case OP_GE: \ + match = (*addr >= val); \ + break; \ + default: \ + break; \ + } \ + \ + return match; \ +} + +#define DEFINE_EQUALITY_PRED(size) \ +static int filter_pred_##size(struct filter_pred *pred, void *event, \ + int val1, int val2) \ +{ \ + u##size *addr = (u##size *)(event + pred->offset); \ + u##size val = (u##size)pred->val; \ + int match; \ + \ + match = (val == *addr) ^ pred->not; \ + \ + return match; \ +} + +extern struct mutex event_mutex; +extern struct list_head ftrace_events; extern const char *__start___trace_bprintk_fmt[]; extern const char *__stop___trace_bprintk_fmt[]; -/* - * The double __builtin_constant_p is because gcc will give us an error - * if we try to allocate the static variable to fmt if it is not a - * constant. Even with the outer if statement optimizing out. - */ -#define event_trace_printk(ip, fmt, args...) \ -do { \ - __trace_printk_check_format(fmt, ##args); \ - tracing_record_cmdline(current); \ - if (__builtin_constant_p(fmt)) { \ - static const char *trace_printk_fmt \ - __attribute__((section("__trace_printk_fmt"))) = \ - __builtin_constant_p(fmt) ? fmt : NULL; \ - \ - __trace_bprintk(ip, trace_printk_fmt, ##args); \ - } else \ - __trace_printk(ip, fmt, ##args); \ -} while (0) +#undef TRACE_EVENT_FORMAT +#define TRACE_EVENT_FORMAT(call, proto, args, fmt, tstruct, tpfmt) \ + extern struct ftrace_event_call event_##call; +#undef TRACE_EVENT_FORMAT_NOFILTER +#define TRACE_EVENT_FORMAT_NOFILTER(call, proto, args, fmt, tstruct, tpfmt) +#include "trace_event_types.h" #endif /* _LINUX_KERNEL_TRACE_H */ diff --git a/kernel/trace/trace_boot.c b/kernel/trace/trace_boot.c index 7a30fc4c364..a29ef23ffb4 100644 --- a/kernel/trace/trace_boot.c +++ b/kernel/trace/trace_boot.c @@ -9,6 +9,7 @@ #include <linux/debugfs.h> #include <linux/ftrace.h> #include <linux/kallsyms.h> +#include <linux/time.h> #include "trace.h" #include "trace_output.h" @@ -67,7 +68,7 @@ initcall_call_print_line(struct trace_iterator *iter) trace_assign_type(field, entry); call = &field->boot_call; ts = iter->ts; - nsec_rem = do_div(ts, 1000000000); + nsec_rem = do_div(ts, NSEC_PER_SEC); ret = trace_seq_printf(s, "[%5ld.%09ld] calling %s @ %i\n", (unsigned long)ts, nsec_rem, call->func, call->caller); @@ -92,7 +93,7 @@ initcall_ret_print_line(struct trace_iterator *iter) trace_assign_type(field, entry); init_ret = &field->boot_ret; ts = iter->ts; - nsec_rem = do_div(ts, 1000000000); + nsec_rem = do_div(ts, NSEC_PER_SEC); ret = trace_seq_printf(s, "[%5ld.%09ld] initcall %s " "returned %d after %llu msecs\n", diff --git a/kernel/trace/trace_branch.c b/kernel/trace/trace_branch.c index 8333715e406..7a7a9fd249a 100644 --- a/kernel/trace/trace_branch.c +++ b/kernel/trace/trace_branch.c @@ -30,6 +30,7 @@ static struct trace_array *branch_tracer; static void probe_likely_condition(struct ftrace_branch_data *f, int val, int expect) { + struct ftrace_event_call *call = &event_branch; struct trace_array *tr = branch_tracer; struct ring_buffer_event *event; struct trace_branch *entry; @@ -73,7 +74,8 @@ probe_likely_condition(struct ftrace_branch_data *f, int val, int expect) entry->line = f->line; entry->correct = val == expect; - ring_buffer_unlock_commit(tr->buffer, event); + if (!filter_check_discard(call, entry, tr->buffer, event)) + ring_buffer_unlock_commit(tr->buffer, event); out: atomic_dec(&tr->data[cpu]->disabled); @@ -271,7 +273,7 @@ static int branch_stat_show(struct seq_file *m, void *v) return 0; } -static void *annotated_branch_stat_start(void) +static void *annotated_branch_stat_start(struct tracer_stat *trace) { return __start_annotated_branch_profile; } @@ -346,7 +348,7 @@ static int all_branch_stat_headers(struct seq_file *m) return 0; } -static void *all_branch_stat_start(void) +static void *all_branch_stat_start(struct tracer_stat *trace) { return __start_branch_profile; } diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c index 22cba997077..5b5895afecf 100644 --- a/kernel/trace/trace_event_profile.c +++ b/kernel/trace/trace_event_profile.c @@ -10,22 +10,30 @@ int ftrace_profile_enable(int event_id) { struct ftrace_event_call *event; + int ret = -EINVAL; - for_each_event(event) { - if (event->id == event_id) - return event->profile_enable(event); + mutex_lock(&event_mutex); + list_for_each_entry(event, &ftrace_events, list) { + if (event->id == event_id) { + ret = event->profile_enable(event); + break; + } } + mutex_unlock(&event_mutex); - return -EINVAL; + return ret; } void ftrace_profile_disable(int event_id) { struct ftrace_event_call *event; - for_each_event(event) { - if (event->id == event_id) - return event->profile_disable(event); + mutex_lock(&event_mutex); + list_for_each_entry(event, &ftrace_events, list) { + if (event->id == event_id) { + event->profile_disable(event); + break; + } } + mutex_unlock(&event_mutex); } - diff --git a/kernel/trace/trace_event_types.h b/kernel/trace/trace_event_types.h index fd78bee71dd..5e32e375134 100644 --- a/kernel/trace/trace_event_types.h +++ b/kernel/trace/trace_event_types.h @@ -57,7 +57,7 @@ TRACE_EVENT_FORMAT(context_switch, TRACE_CTX, ctx_switch_entry, ignore, TP_RAW_FMT("%u:%u:%u ==+ %u:%u:%u [%03u]") ); -TRACE_EVENT_FORMAT(special, TRACE_SPECIAL, special_entry, ignore, +TRACE_EVENT_FORMAT_NOFILTER(special, TRACE_SPECIAL, special_entry, ignore, TRACE_STRUCT( TRACE_FIELD(unsigned long, arg1, arg1) TRACE_FIELD(unsigned long, arg2, arg2) @@ -122,8 +122,10 @@ TRACE_EVENT_FORMAT(print, TRACE_PRINT, print_entry, ignore, TRACE_EVENT_FORMAT(branch, TRACE_BRANCH, trace_branch, ignore, TRACE_STRUCT( TRACE_FIELD(unsigned int, line, line) - TRACE_FIELD_SPECIAL(char func[TRACE_FUNC_SIZE+1], func, func) - TRACE_FIELD_SPECIAL(char file[TRACE_FUNC_SIZE+1], file, file) + TRACE_FIELD_SPECIAL(char func[TRACE_FUNC_SIZE+1], func, + TRACE_FUNC_SIZE+1, func) + TRACE_FIELD_SPECIAL(char file[TRACE_FUNC_SIZE+1], file, + TRACE_FUNC_SIZE+1, file) TRACE_FIELD(char, correct, correct) ), TP_RAW_FMT("%u:%s:%s (%u)") @@ -139,8 +141,8 @@ TRACE_EVENT_FORMAT(hw_branch, TRACE_HW_BRANCHES, hw_branch_entry, ignore, TRACE_EVENT_FORMAT(power, TRACE_POWER, trace_power, ignore, TRACE_STRUCT( - TRACE_FIELD(ktime_t, state_data.stamp, stamp) - TRACE_FIELD(ktime_t, state_data.end, end) + TRACE_FIELD_SIGN(ktime_t, state_data.stamp, stamp, 1) + TRACE_FIELD_SIGN(ktime_t, state_data.end, end, 1) TRACE_FIELD(int, state_data.type, type) TRACE_FIELD(int, state_data.state, state) ), diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 576f4fa2af0..aa08be69a1b 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -8,19 +8,25 @@ * */ +#include <linux/workqueue.h> +#include <linux/spinlock.h> +#include <linux/kthread.h> #include <linux/debugfs.h> #include <linux/uaccess.h> #include <linux/module.h> #include <linux/ctype.h> +#include <linux/delay.h> #include "trace_output.h" #define TRACE_SYSTEM "TRACE_SYSTEM" -static DEFINE_MUTEX(event_mutex); +DEFINE_MUTEX(event_mutex); + +LIST_HEAD(ftrace_events); int trace_define_field(struct ftrace_event_call *call, char *type, - char *name, int offset, int size) + char *name, int offset, int size, int is_signed) { struct ftrace_event_field *field; @@ -38,6 +44,7 @@ int trace_define_field(struct ftrace_event_call *call, char *type, field->offset = offset; field->size = size; + field->is_signed = is_signed; list_add(&field->link, &call->fields); return 0; @@ -51,47 +58,94 @@ err: return -ENOMEM; } +EXPORT_SYMBOL_GPL(trace_define_field); -static void ftrace_clear_events(void) -{ - struct ftrace_event_call *call = (void *)__start_ftrace_events; - +#ifdef CONFIG_MODULES - while ((unsigned long)call < (unsigned long)__stop_ftrace_events) { +static void trace_destroy_fields(struct ftrace_event_call *call) +{ + struct ftrace_event_field *field, *next; - if (call->enabled) { - call->enabled = 0; - call->unregfunc(); - } - call++; + list_for_each_entry_safe(field, next, &call->fields, link) { + list_del(&field->link); + kfree(field->type); + kfree(field->name); + kfree(field); } } +#endif /* CONFIG_MODULES */ + static void ftrace_event_enable_disable(struct ftrace_event_call *call, int enable) { - switch (enable) { case 0: if (call->enabled) { call->enabled = 0; + tracing_stop_cmdline_record(); call->unregfunc(); } break; case 1: if (!call->enabled) { call->enabled = 1; + tracing_start_cmdline_record(); call->regfunc(); } break; } } +static void ftrace_clear_events(void) +{ + struct ftrace_event_call *call; + + mutex_lock(&event_mutex); + list_for_each_entry(call, &ftrace_events, list) { + ftrace_event_enable_disable(call, 0); + } + mutex_unlock(&event_mutex); +} + +/* + * __ftrace_set_clr_event(NULL, NULL, NULL, set) will set/unset all events. + */ +static int __ftrace_set_clr_event(const char *match, const char *sub, + const char *event, int set) +{ + struct ftrace_event_call *call; + int ret = -EINVAL; + + mutex_lock(&event_mutex); + list_for_each_entry(call, &ftrace_events, list) { + + if (!call->name || !call->regfunc) + continue; + + if (match && + strcmp(match, call->name) != 0 && + strcmp(match, call->system) != 0) + continue; + + if (sub && strcmp(sub, call->system) != 0) + continue; + + if (event && strcmp(event, call->name) != 0) + continue; + + ftrace_event_enable_disable(call, set); + + ret = 0; + } + mutex_unlock(&event_mutex); + + return ret; +} + static int ftrace_set_clr_event(char *buf, int set) { - struct ftrace_event_call *call = __start_ftrace_events; char *event = NULL, *sub = NULL, *match; - int ret = -EINVAL; /* * The buf format can be <subsystem>:<event-name> @@ -117,30 +171,24 @@ static int ftrace_set_clr_event(char *buf, int set) event = NULL; } - mutex_lock(&event_mutex); - for_each_event(call) { - - if (!call->name || !call->regfunc) - continue; - - if (match && - strcmp(match, call->name) != 0 && - strcmp(match, call->system) != 0) - continue; - - if (sub && strcmp(sub, call->system) != 0) - continue; - - if (event && strcmp(event, call->name) != 0) - continue; - - ftrace_event_enable_disable(call, set); - - ret = 0; - } - mutex_unlock(&event_mutex); + return __ftrace_set_clr_event(match, sub, event, set); +} - return ret; +/** + * trace_set_clr_event - enable or disable an event + * @system: system name to match (NULL for any system) + * @event: event name to match (NULL for all events, within system) + * @set: 1 to enable, 0 to disable + * + * This is a way for other parts of the kernel to enable or disable + * event recording. + * + * Returns 0 on success, -EINVAL if the parameters do not match any + * registered events. + */ +int trace_set_clr_event(const char *system, const char *event, int set) +{ + return __ftrace_set_clr_event(NULL, system, event, set); } /* 128 should be much more than enough */ @@ -224,15 +272,17 @@ ftrace_event_write(struct file *file, const char __user *ubuf, static void * t_next(struct seq_file *m, void *v, loff_t *pos) { - struct ftrace_event_call *call = m->private; - struct ftrace_event_call *next = call; + struct list_head *list = m->private; + struct ftrace_event_call *call; (*pos)++; for (;;) { - if ((unsigned long)call >= (unsigned long)__stop_ftrace_events) + if (list == &ftrace_events) return NULL; + call = list_entry(list, struct ftrace_event_call, list); + /* * The ftrace subsystem is for showing formats only. * They can not be enabled or disabled via the event files. @@ -240,45 +290,51 @@ t_next(struct seq_file *m, void *v, loff_t *pos) if (call->regfunc) break; - call++; - next = call; + list = list->next; } - m->private = ++next; + m->private = list->next; return call; } static void *t_start(struct seq_file *m, loff_t *pos) { + mutex_lock(&event_mutex); + if (*pos == 0) + m->private = ftrace_events.next; return t_next(m, NULL, pos); } static void * s_next(struct seq_file *m, void *v, loff_t *pos) { - struct ftrace_event_call *call = m->private; - struct ftrace_event_call *next; + struct list_head *list = m->private; + struct ftrace_event_call *call; (*pos)++; retry: - if ((unsigned long)call >= (unsigned long)__stop_ftrace_events) + if (list == &ftrace_events) return NULL; + call = list_entry(list, struct ftrace_event_call, list); + if (!call->enabled) { - call++; + list = list->next; goto retry; } - next = call; - m->private = ++next; + m->private = list->next; return call; } static void *s_start(struct seq_file *m, loff_t *pos) { + mutex_lock(&event_mutex); + if (*pos == 0) + m->private = ftrace_events.next; return s_next(m, NULL, pos); } @@ -295,12 +351,12 @@ static int t_show(struct seq_file *m, void *v) static void t_stop(struct seq_file *m, void *p) { + mutex_unlock(&event_mutex); } static int ftrace_event_seq_open(struct inode *inode, struct file *file) { - int ret; const struct seq_operations *seq_ops; if ((file->f_mode & FMODE_WRITE) && @@ -308,13 +364,7 @@ ftrace_event_seq_open(struct inode *inode, struct file *file) ftrace_clear_events(); seq_ops = inode->i_private; - ret = seq_open(file, seq_ops); - if (!ret) { - struct seq_file *m = file->private_data; - - m->private = __start_ftrace_events; - } - return ret; + return seq_open(file, seq_ops); } static ssize_t @@ -374,8 +424,93 @@ event_enable_write(struct file *filp, const char __user *ubuf, size_t cnt, return cnt; } +static ssize_t +system_enable_read(struct file *filp, char __user *ubuf, size_t cnt, + loff_t *ppos) +{ + const char set_to_char[4] = { '?', '0', '1', 'X' }; + const char *system = filp->private_data; + struct ftrace_event_call *call; + char buf[2]; + int set = 0; + int ret; + + mutex_lock(&event_mutex); + list_for_each_entry(call, &ftrace_events, list) { + if (!call->name || !call->regfunc) + continue; + + if (system && strcmp(call->system, system) != 0) + continue; + + /* + * We need to find out if all the events are set + * or if all events or cleared, or if we have + * a mixture. + */ + set |= (1 << !!call->enabled); + + /* + * If we have a mixture, no need to look further. + */ + if (set == 3) + break; + } + mutex_unlock(&event_mutex); + + buf[0] = set_to_char[set]; + buf[1] = '\n'; + + ret = simple_read_from_buffer(ubuf, cnt, ppos, buf, 2); + + return ret; +} + +static ssize_t +system_enable_write(struct file *filp, const char __user *ubuf, size_t cnt, + loff_t *ppos) +{ + const char *system = filp->private_data; + unsigned long val; + char buf[64]; + ssize_t ret; + + if (cnt >= sizeof(buf)) + return -EINVAL; + + if (copy_from_user(&buf, ubuf, cnt)) + return -EFAULT; + + buf[cnt] = 0; + + ret = strict_strtoul(buf, 10, &val); + if (ret < 0) + return ret; + + ret = tracing_update_buffers(); + if (ret < 0) + return ret; + + if (val != 0 && val != 1) + return -EINVAL; + + ret = __ftrace_set_clr_event(NULL, system, NULL, val); + if (ret) + goto out; + + ret = cnt; + +out: + *ppos += cnt; + + return ret; +} + +extern char *__bad_type_size(void); + #undef FIELD #define FIELD(type, name) \ + sizeof(type) != sizeof(field.name) ? __bad_type_size() : \ #type, "common_" #name, offsetof(typeof(field), name), \ sizeof(field.name) @@ -391,7 +526,7 @@ static int trace_write_header(struct trace_seq *s) "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n" "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\n" "\n", - FIELD(unsigned char, type), + FIELD(unsigned short, type), FIELD(unsigned char, flags), FIELD(unsigned char, preempt_count), FIELD(int, pid), @@ -481,7 +616,7 @@ event_filter_read(struct file *filp, char __user *ubuf, size_t cnt, trace_seq_init(s); - filter_print_preds(call->preds, s); + print_event_filter(call, s); r = simple_read_from_buffer(ubuf, cnt, ppos, s->buffer, s->len); kfree(s); @@ -494,38 +629,26 @@ event_filter_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { struct ftrace_event_call *call = filp->private_data; - char buf[64], *pbuf = buf; - struct filter_pred *pred; + char *buf; int err; - if (cnt >= sizeof(buf)) + if (cnt >= PAGE_SIZE) return -EINVAL; - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - buf[cnt] = '\0'; - - pred = kzalloc(sizeof(*pred), GFP_KERNEL); - if (!pred) + buf = (char *)__get_free_page(GFP_TEMPORARY); + if (!buf) return -ENOMEM; - err = filter_parse(&pbuf, pred); - if (err < 0) { - filter_free_pred(pred); - return err; - } - - if (pred->clear) { - filter_free_preds(call); - filter_free_pred(pred); - return cnt; + if (copy_from_user(buf, ubuf, cnt)) { + free_page((unsigned long) buf); + return -EFAULT; } + buf[cnt] = '\0'; - err = filter_add_pred(call, pred); - if (err < 0) { - filter_free_pred(pred); + err = apply_event_filter(call, buf); + free_page((unsigned long) buf); + if (err < 0) return err; - } *ppos += cnt; @@ -549,7 +672,7 @@ subsystem_filter_read(struct file *filp, char __user *ubuf, size_t cnt, trace_seq_init(s); - filter_print_preds(system->preds, s); + print_subsystem_event_filter(system, s); r = simple_read_from_buffer(ubuf, cnt, ppos, s->buffer, s->len); kfree(s); @@ -562,45 +685,56 @@ subsystem_filter_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { struct event_subsystem *system = filp->private_data; - char buf[64], *pbuf = buf; - struct filter_pred *pred; + char *buf; int err; - if (cnt >= sizeof(buf)) + if (cnt >= PAGE_SIZE) return -EINVAL; - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - buf[cnt] = '\0'; - - pred = kzalloc(sizeof(*pred), GFP_KERNEL); - if (!pred) + buf = (char *)__get_free_page(GFP_TEMPORARY); + if (!buf) return -ENOMEM; - err = filter_parse(&pbuf, pred); - if (err < 0) { - filter_free_pred(pred); - return err; - } - - if (pred->clear) { - filter_free_subsystem_preds(system); - filter_free_pred(pred); - return cnt; + if (copy_from_user(buf, ubuf, cnt)) { + free_page((unsigned long) buf); + return -EFAULT; } + buf[cnt] = '\0'; - err = filter_add_subsystem_pred(system, pred); - if (err < 0) { - filter_free_subsystem_preds(system); - filter_free_pred(pred); + err = apply_subsystem_event_filter(system, buf); + free_page((unsigned long) buf); + if (err < 0) return err; - } *ppos += cnt; return cnt; } +static ssize_t +show_header(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) +{ + int (*func)(struct trace_seq *s) = filp->private_data; + struct trace_seq *s; + int r; + + if (*ppos) + return 0; + + s = kmalloc(sizeof(*s), GFP_KERNEL); + if (!s) + return -ENOMEM; + + trace_seq_init(s); + + func(s); + r = simple_read_from_buffer(ubuf, cnt, ppos, s->buffer, s->len); + + kfree(s); + + return r; +} + static const struct seq_operations show_event_seq_ops = { .start = t_start, .next = t_next, @@ -658,6 +792,17 @@ static const struct file_operations ftrace_subsystem_filter_fops = { .write = subsystem_filter_write, }; +static const struct file_operations ftrace_system_enable_fops = { + .open = tracing_open_generic, + .read = system_enable_read, + .write = system_enable_write, +}; + +static const struct file_operations ftrace_show_header_fops = { + .open = tracing_open_generic, + .read = show_header, +}; + static struct dentry *event_trace_events_dir(void) { static struct dentry *d_tracer; @@ -684,6 +829,7 @@ static struct dentry * event_subsystem_dir(const char *name, struct dentry *d_events) { struct event_subsystem *system; + struct dentry *entry; /* First see if we did not already create this dir */ list_for_each_entry(system, &event_subsystems, list) { @@ -707,16 +853,46 @@ event_subsystem_dir(const char *name, struct dentry *d_events) return d_events; } - system->name = name; + system->name = kstrdup(name, GFP_KERNEL); + if (!system->name) { + debugfs_remove(system->entry); + kfree(system); + return d_events; + } + list_add(&system->list, &event_subsystems); - system->preds = NULL; + system->filter = NULL; + + system->filter = kzalloc(sizeof(struct event_filter), GFP_KERNEL); + if (!system->filter) { + pr_warning("Could not allocate filter for subsystem " + "'%s'\n", name); + return system->entry; + } + + entry = debugfs_create_file("filter", 0644, system->entry, system, + &ftrace_subsystem_filter_fops); + if (!entry) { + kfree(system->filter); + system->filter = NULL; + pr_warning("Could not create debugfs " + "'%s/filter' entry\n", name); + } + + entry = trace_create_file("enable", 0644, system->entry, + (void *)system->name, + &ftrace_system_enable_fops); return system->entry; } static int -event_create_dir(struct ftrace_event_call *call, struct dentry *d_events) +event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, + const struct file_operations *id, + const struct file_operations *enable, + const struct file_operations *filter, + const struct file_operations *format) { struct dentry *entry; int ret; @@ -725,7 +901,7 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events) * If the trace point header did not define TRACE_SYSTEM * then the system would be called "TRACE_SYSTEM". */ - if (strcmp(call->system, "TRACE_SYSTEM") != 0) + if (strcmp(call->system, TRACE_SYSTEM) != 0) d_events = event_subsystem_dir(call->system, d_events); if (call->raw_init) { @@ -744,21 +920,13 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events) return -1; } - if (call->regfunc) { - entry = debugfs_create_file("enable", 0644, call->dir, call, - &ftrace_enable_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'%s/enable' entry\n", call->name); - } + if (call->regfunc) + entry = trace_create_file("enable", 0644, call->dir, call, + enable); - if (call->id) { - entry = debugfs_create_file("id", 0444, call->dir, call, - &ftrace_event_id_fops); - if (!entry) - pr_warning("Could not create debugfs '%s/id' entry\n", - call->name); - } + if (call->id) + entry = trace_create_file("id", 0444, call->dir, call, + id); if (call->define_fields) { ret = call->define_fields(); @@ -767,32 +935,195 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events) " events/%s\n", call->name); return ret; } - entry = debugfs_create_file("filter", 0644, call->dir, call, - &ftrace_event_filter_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'%s/filter' entry\n", call->name); + entry = trace_create_file("filter", 0644, call->dir, call, + filter); } /* A trace may not want to export its format */ if (!call->show_format) return 0; - entry = debugfs_create_file("format", 0444, call->dir, call, - &ftrace_event_format_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'%s/format' entry\n", call->name); + entry = trace_create_file("format", 0444, call->dir, call, + format); + + return 0; +} + +#define for_each_event(event, start, end) \ + for (event = start; \ + (unsigned long)event < (unsigned long)end; \ + event++) + +#ifdef CONFIG_MODULES + +static LIST_HEAD(ftrace_module_file_list); + +/* + * Modules must own their file_operations to keep up with + * reference counting. + */ +struct ftrace_module_file_ops { + struct list_head list; + struct module *mod; + struct file_operations id; + struct file_operations enable; + struct file_operations format; + struct file_operations filter; +}; + +static struct ftrace_module_file_ops * +trace_create_file_ops(struct module *mod) +{ + struct ftrace_module_file_ops *file_ops; + + /* + * This is a bit of a PITA. To allow for correct reference + * counting, modules must "own" their file_operations. + * To do this, we allocate the file operations that will be + * used in the event directory. + */ + + file_ops = kmalloc(sizeof(*file_ops), GFP_KERNEL); + if (!file_ops) + return NULL; + + file_ops->mod = mod; + + file_ops->id = ftrace_event_id_fops; + file_ops->id.owner = mod; + + file_ops->enable = ftrace_enable_fops; + file_ops->enable.owner = mod; + + file_ops->filter = ftrace_event_filter_fops; + file_ops->filter.owner = mod; + + file_ops->format = ftrace_event_format_fops; + file_ops->format.owner = mod; + + list_add(&file_ops->list, &ftrace_module_file_list); + + return file_ops; +} + +static void trace_module_add_events(struct module *mod) +{ + struct ftrace_module_file_ops *file_ops = NULL; + struct ftrace_event_call *call, *start, *end; + struct dentry *d_events; + + start = mod->trace_events; + end = mod->trace_events + mod->num_trace_events; + + if (start == end) + return; + + d_events = event_trace_events_dir(); + if (!d_events) + return; + + for_each_event(call, start, end) { + /* The linker may leave blanks */ + if (!call->name) + continue; + + /* + * This module has events, create file ops for this module + * if not already done. + */ + if (!file_ops) { + file_ops = trace_create_file_ops(mod); + if (!file_ops) + return; + } + call->mod = mod; + list_add(&call->list, &ftrace_events); + event_create_dir(call, d_events, + &file_ops->id, &file_ops->enable, + &file_ops->filter, &file_ops->format); + } +} + +static void trace_module_remove_events(struct module *mod) +{ + struct ftrace_module_file_ops *file_ops; + struct ftrace_event_call *call, *p; + bool found = false; + + down_write(&trace_event_mutex); + list_for_each_entry_safe(call, p, &ftrace_events, list) { + if (call->mod == mod) { + found = true; + ftrace_event_enable_disable(call, 0); + if (call->event) + __unregister_ftrace_event(call->event); + debugfs_remove_recursive(call->dir); + list_del(&call->list); + trace_destroy_fields(call); + destroy_preds(call); + } + } + + /* Now free the file_operations */ + list_for_each_entry(file_ops, &ftrace_module_file_list, list) { + if (file_ops->mod == mod) + break; + } + if (&file_ops->list != &ftrace_module_file_list) { + list_del(&file_ops->list); + kfree(file_ops); + } + + /* + * It is safest to reset the ring buffer if the module being unloaded + * registered any events. + */ + if (found) + tracing_reset_current_online_cpus(); + up_write(&trace_event_mutex); +} + +static int trace_module_notify(struct notifier_block *self, + unsigned long val, void *data) +{ + struct module *mod = data; + + mutex_lock(&event_mutex); + switch (val) { + case MODULE_STATE_COMING: + trace_module_add_events(mod); + break; + case MODULE_STATE_GOING: + trace_module_remove_events(mod); + break; + } + mutex_unlock(&event_mutex); return 0; } +#else +static int trace_module_notify(struct notifier_block *self, + unsigned long val, void *data) +{ + return 0; +} +#endif /* CONFIG_MODULES */ + +struct notifier_block trace_module_nb = { + .notifier_call = trace_module_notify, + .priority = 0, +}; + +extern struct ftrace_event_call __start_ftrace_events[]; +extern struct ftrace_event_call __stop_ftrace_events[]; static __init int event_trace_init(void) { - struct ftrace_event_call *call = __start_ftrace_events; + struct ftrace_event_call *call; struct dentry *d_tracer; struct dentry *entry; struct dentry *d_events; + int ret; d_tracer = tracing_init_dentry(); if (!d_tracer) @@ -816,13 +1147,243 @@ static __init int event_trace_init(void) if (!d_events) return 0; - for_each_event(call) { + /* ring buffer internal formats */ + trace_create_file("header_page", 0444, d_events, + ring_buffer_print_page_header, + &ftrace_show_header_fops); + + trace_create_file("header_event", 0444, d_events, + ring_buffer_print_entry_header, + &ftrace_show_header_fops); + + trace_create_file("enable", 0644, d_events, + NULL, &ftrace_system_enable_fops); + + for_each_event(call, __start_ftrace_events, __stop_ftrace_events) { /* The linker may leave blanks */ if (!call->name) continue; - event_create_dir(call, d_events); + list_add(&call->list, &ftrace_events); + event_create_dir(call, d_events, &ftrace_event_id_fops, + &ftrace_enable_fops, &ftrace_event_filter_fops, + &ftrace_event_format_fops); } + ret = register_module_notifier(&trace_module_nb); + if (ret) + pr_warning("Failed to register trace events module notifier\n"); + return 0; } fs_initcall(event_trace_init); + +#ifdef CONFIG_FTRACE_STARTUP_TEST + +static DEFINE_SPINLOCK(test_spinlock); +static DEFINE_SPINLOCK(test_spinlock_irq); +static DEFINE_MUTEX(test_mutex); + +static __init void test_work(struct work_struct *dummy) +{ + spin_lock(&test_spinlock); + spin_lock_irq(&test_spinlock_irq); + udelay(1); + spin_unlock_irq(&test_spinlock_irq); + spin_unlock(&test_spinlock); + + mutex_lock(&test_mutex); + msleep(1); + mutex_unlock(&test_mutex); +} + +static __init int event_test_thread(void *unused) +{ + void *test_malloc; + + test_malloc = kmalloc(1234, GFP_KERNEL); + if (!test_malloc) + pr_info("failed to kmalloc\n"); + + schedule_on_each_cpu(test_work); + + kfree(test_malloc); + + set_current_state(TASK_INTERRUPTIBLE); + while (!kthread_should_stop()) + schedule(); + + return 0; +} + +/* + * Do various things that may trigger events. + */ +static __init void event_test_stuff(void) +{ + struct task_struct *test_thread; + + test_thread = kthread_run(event_test_thread, NULL, "test-events"); + msleep(1); + kthread_stop(test_thread); +} + +/* + * For every trace event defined, we will test each trace point separately, + * and then by groups, and finally all trace points. + */ +static __init void event_trace_self_tests(void) +{ + struct ftrace_event_call *call; + struct event_subsystem *system; + int ret; + + pr_info("Running tests on trace events:\n"); + + list_for_each_entry(call, &ftrace_events, list) { + + /* Only test those that have a regfunc */ + if (!call->regfunc) + continue; + + pr_info("Testing event %s: ", call->name); + + /* + * If an event is already enabled, someone is using + * it and the self test should not be on. + */ + if (call->enabled) { + pr_warning("Enabled event during self test!\n"); + WARN_ON_ONCE(1); + continue; + } + + ftrace_event_enable_disable(call, 1); + event_test_stuff(); + ftrace_event_enable_disable(call, 0); + + pr_cont("OK\n"); + } + + /* Now test at the sub system level */ + + pr_info("Running tests on trace event systems:\n"); + + list_for_each_entry(system, &event_subsystems, list) { + + /* the ftrace system is special, skip it */ + if (strcmp(system->name, "ftrace") == 0) + continue; + + pr_info("Testing event system %s: ", system->name); + + ret = __ftrace_set_clr_event(NULL, system->name, NULL, 1); + if (WARN_ON_ONCE(ret)) { + pr_warning("error enabling system %s\n", + system->name); + continue; + } + + event_test_stuff(); + + ret = __ftrace_set_clr_event(NULL, system->name, NULL, 0); + if (WARN_ON_ONCE(ret)) + pr_warning("error disabling system %s\n", + system->name); + + pr_cont("OK\n"); + } + + /* Test with all events enabled */ + + pr_info("Running tests on all trace events:\n"); + pr_info("Testing all events: "); + + ret = __ftrace_set_clr_event(NULL, NULL, NULL, 1); + if (WARN_ON_ONCE(ret)) { + pr_warning("error enabling all events\n"); + return; + } + + event_test_stuff(); + + /* reset sysname */ + ret = __ftrace_set_clr_event(NULL, NULL, NULL, 0); + if (WARN_ON_ONCE(ret)) { + pr_warning("error disabling all events\n"); + return; + } + + pr_cont("OK\n"); +} + +#ifdef CONFIG_FUNCTION_TRACER + +static DEFINE_PER_CPU(atomic_t, test_event_disable); + +static void +function_test_events_call(unsigned long ip, unsigned long parent_ip) +{ + struct ring_buffer_event *event; + struct ftrace_entry *entry; + unsigned long flags; + long disabled; + int resched; + int cpu; + int pc; + + pc = preempt_count(); + resched = ftrace_preempt_disable(); + cpu = raw_smp_processor_id(); + disabled = atomic_inc_return(&per_cpu(test_event_disable, cpu)); + + if (disabled != 1) + goto out; + + local_save_flags(flags); + + event = trace_current_buffer_lock_reserve(TRACE_FN, sizeof(*entry), + flags, pc); + if (!event) + goto out; + entry = ring_buffer_event_data(event); + entry->ip = ip; + entry->parent_ip = parent_ip; + + trace_nowake_buffer_unlock_commit(event, flags, pc); + + out: + atomic_dec(&per_cpu(test_event_disable, cpu)); + ftrace_preempt_enable(resched); +} + +static struct ftrace_ops trace_ops __initdata = +{ + .func = function_test_events_call, +}; + +static __init void event_trace_self_test_with_function(void) +{ + register_ftrace_function(&trace_ops); + pr_info("Running tests again, along with the function tracer\n"); + event_trace_self_tests(); + unregister_ftrace_function(&trace_ops); +} +#else +static __init void event_trace_self_test_with_function(void) +{ +} +#endif + +static __init int event_trace_self_tests_init(void) +{ + + event_trace_self_tests(); + + event_trace_self_test_with_function(); + + return 0; +} + +late_initcall(event_trace_self_tests_init); + +#endif diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index e03cbf1e38f..db6e54bdb59 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -22,119 +22,297 @@ #include <linux/uaccess.h> #include <linux/module.h> #include <linux/ctype.h> +#include <linux/mutex.h> #include "trace.h" #include "trace_output.h" -static int filter_pred_64(struct filter_pred *pred, void *event) +static DEFINE_MUTEX(filter_mutex); + +enum filter_op_ids +{ + OP_OR, + OP_AND, + OP_NE, + OP_EQ, + OP_LT, + OP_LE, + OP_GT, + OP_GE, + OP_NONE, + OP_OPEN_PAREN, +}; + +struct filter_op { + int id; + char *string; + int precedence; +}; + +static struct filter_op filter_ops[] = { + { OP_OR, "||", 1 }, + { OP_AND, "&&", 2 }, + { OP_NE, "!=", 4 }, + { OP_EQ, "==", 4 }, + { OP_LT, "<", 5 }, + { OP_LE, "<=", 5 }, + { OP_GT, ">", 5 }, + { OP_GE, ">=", 5 }, + { OP_NONE, "OP_NONE", 0 }, + { OP_OPEN_PAREN, "(", 0 }, +}; + +enum { + FILT_ERR_NONE, + FILT_ERR_INVALID_OP, + FILT_ERR_UNBALANCED_PAREN, + FILT_ERR_TOO_MANY_OPERANDS, + FILT_ERR_OPERAND_TOO_LONG, + FILT_ERR_FIELD_NOT_FOUND, + FILT_ERR_ILLEGAL_FIELD_OP, + FILT_ERR_ILLEGAL_INTVAL, + FILT_ERR_BAD_SUBSYS_FILTER, + FILT_ERR_TOO_MANY_PREDS, + FILT_ERR_MISSING_FIELD, + FILT_ERR_INVALID_FILTER, +}; + +static char *err_text[] = { + "No error", + "Invalid operator", + "Unbalanced parens", + "Too many operands", + "Operand too long", + "Field not found", + "Illegal operation for field type", + "Illegal integer value", + "Couldn't find or set field in one of a subsystem's events", + "Too many terms in predicate expression", + "Missing field name and/or value", + "Meaningless filter expression", +}; + +struct opstack_op { + int op; + struct list_head list; +}; + +struct postfix_elt { + int op; + char *operand; + struct list_head list; +}; + +struct filter_parse_state { + struct filter_op *ops; + struct list_head opstack; + struct list_head postfix; + int lasterr; + int lasterr_pos; + + struct { + char *string; + unsigned int cnt; + unsigned int tail; + } infix; + + struct { + char string[MAX_FILTER_STR_VAL]; + int pos; + unsigned int tail; + } operand; +}; + +DEFINE_COMPARISON_PRED(s64); +DEFINE_COMPARISON_PRED(u64); +DEFINE_COMPARISON_PRED(s32); +DEFINE_COMPARISON_PRED(u32); +DEFINE_COMPARISON_PRED(s16); +DEFINE_COMPARISON_PRED(u16); +DEFINE_COMPARISON_PRED(s8); +DEFINE_COMPARISON_PRED(u8); + +DEFINE_EQUALITY_PRED(64); +DEFINE_EQUALITY_PRED(32); +DEFINE_EQUALITY_PRED(16); +DEFINE_EQUALITY_PRED(8); + +static int filter_pred_and(struct filter_pred *pred __attribute((unused)), + void *event __attribute((unused)), + int val1, int val2) +{ + return val1 && val2; +} + +static int filter_pred_or(struct filter_pred *pred __attribute((unused)), + void *event __attribute((unused)), + int val1, int val2) +{ + return val1 || val2; +} + +/* Filter predicate for fixed sized arrays of characters */ +static int filter_pred_string(struct filter_pred *pred, void *event, + int val1, int val2) { - u64 *addr = (u64 *)(event + pred->offset); - u64 val = (u64)pred->val; - int match; + char *addr = (char *)(event + pred->offset); + int cmp, match; + + cmp = strncmp(addr, pred->str_val, pred->str_len); - match = (val == *addr) ^ pred->not; + match = (!cmp) ^ pred->not; return match; } -static int filter_pred_32(struct filter_pred *pred, void *event) +/* + * Filter predicate for dynamic sized arrays of characters. + * These are implemented through a list of strings at the end + * of the entry. + * Also each of these strings have a field in the entry which + * contains its offset from the beginning of the entry. + * We have then first to get this field, dereference it + * and add it to the address of the entry, and at last we have + * the address of the string. + */ +static int filter_pred_strloc(struct filter_pred *pred, void *event, + int val1, int val2) { - u32 *addr = (u32 *)(event + pred->offset); - u32 val = (u32)pred->val; - int match; + int str_loc = *(int *)(event + pred->offset); + char *addr = (char *)(event + str_loc); + int cmp, match; + + cmp = strncmp(addr, pred->str_val, pred->str_len); - match = (val == *addr) ^ pred->not; + match = (!cmp) ^ pred->not; return match; } -static int filter_pred_16(struct filter_pred *pred, void *event) +static int filter_pred_none(struct filter_pred *pred, void *event, + int val1, int val2) +{ + return 0; +} + +/* return 1 if event matches, 0 otherwise (discard) */ +int filter_match_preds(struct ftrace_event_call *call, void *rec) { - u16 *addr = (u16 *)(event + pred->offset); - u16 val = (u16)pred->val; - int match; + struct event_filter *filter = call->filter; + int match, top = 0, val1 = 0, val2 = 0; + int stack[MAX_FILTER_PRED]; + struct filter_pred *pred; + int i; + + for (i = 0; i < filter->n_preds; i++) { + pred = filter->preds[i]; + if (!pred->pop_n) { + match = pred->fn(pred, rec, val1, val2); + stack[top++] = match; + continue; + } + if (pred->pop_n > top) { + WARN_ON_ONCE(1); + return 0; + } + val1 = stack[--top]; + val2 = stack[--top]; + match = pred->fn(pred, rec, val1, val2); + stack[top++] = match; + } - match = (val == *addr) ^ pred->not; + return stack[--top]; +} +EXPORT_SYMBOL_GPL(filter_match_preds); - return match; +static void parse_error(struct filter_parse_state *ps, int err, int pos) +{ + ps->lasterr = err; + ps->lasterr_pos = pos; } -static int filter_pred_8(struct filter_pred *pred, void *event) +static void remove_filter_string(struct event_filter *filter) { - u8 *addr = (u8 *)(event + pred->offset); - u8 val = (u8)pred->val; - int match; + kfree(filter->filter_string); + filter->filter_string = NULL; +} - match = (val == *addr) ^ pred->not; +static int replace_filter_string(struct event_filter *filter, + char *filter_string) +{ + kfree(filter->filter_string); + filter->filter_string = kstrdup(filter_string, GFP_KERNEL); + if (!filter->filter_string) + return -ENOMEM; - return match; + return 0; } -static int filter_pred_string(struct filter_pred *pred, void *event) +static int append_filter_string(struct event_filter *filter, + char *string) { - char *addr = (char *)(event + pred->offset); - int cmp, match; + int newlen; + char *new_filter_string; - cmp = strncmp(addr, pred->str_val, pred->str_len); + BUG_ON(!filter->filter_string); + newlen = strlen(filter->filter_string) + strlen(string) + 1; + new_filter_string = kmalloc(newlen, GFP_KERNEL); + if (!new_filter_string) + return -ENOMEM; - match = (!cmp) ^ pred->not; + strcpy(new_filter_string, filter->filter_string); + strcat(new_filter_string, string); + kfree(filter->filter_string); + filter->filter_string = new_filter_string; - return match; + return 0; } -/* return 1 if event matches, 0 otherwise (discard) */ -int filter_match_preds(struct ftrace_event_call *call, void *rec) +static void append_filter_err(struct filter_parse_state *ps, + struct event_filter *filter) { - int i, matched, and_failed = 0; - struct filter_pred *pred; + int pos = ps->lasterr_pos; + char *buf, *pbuf; - for (i = 0; i < MAX_FILTER_PRED; i++) { - if (call->preds[i]) { - pred = call->preds[i]; - if (and_failed && !pred->or) - continue; - matched = pred->fn(pred, rec); - if (!matched && !pred->or) { - and_failed = 1; - continue; - } else if (matched && pred->or) - return 1; - } else - break; - } + buf = (char *)__get_free_page(GFP_TEMPORARY); + if (!buf) + return; - if (and_failed) - return 0; + append_filter_string(filter, "\n"); + memset(buf, ' ', PAGE_SIZE); + if (pos > PAGE_SIZE - 128) + pos = 0; + buf[pos] = '^'; + pbuf = &buf[pos] + 1; - return 1; + sprintf(pbuf, "\nparse_error: %s\n", err_text[ps->lasterr]); + append_filter_string(filter, buf); + free_page((unsigned long) buf); } -void filter_print_preds(struct filter_pred **preds, struct trace_seq *s) +void print_event_filter(struct ftrace_event_call *call, struct trace_seq *s) { - char *field_name; - struct filter_pred *pred; - int i; + struct event_filter *filter = call->filter; - if (!preds) { + mutex_lock(&filter_mutex); + if (filter->filter_string) + trace_seq_printf(s, "%s\n", filter->filter_string); + else trace_seq_printf(s, "none\n"); - return; - } + mutex_unlock(&filter_mutex); +} - for (i = 0; i < MAX_FILTER_PRED; i++) { - if (preds[i]) { - pred = preds[i]; - field_name = pred->field_name; - if (i) - trace_seq_printf(s, pred->or ? "|| " : "&& "); - trace_seq_printf(s, "%s ", field_name); - trace_seq_printf(s, pred->not ? "!= " : "== "); - if (pred->str_val) - trace_seq_printf(s, "%s\n", pred->str_val); - else - trace_seq_printf(s, "%llu\n", pred->val); - } else - break; - } +void print_subsystem_event_filter(struct event_subsystem *system, + struct trace_seq *s) +{ + struct event_filter *filter = system->filter; + + mutex_lock(&filter_mutex); + if (filter->filter_string) + trace_seq_printf(s, "%s\n", filter->filter_string); + else + trace_seq_printf(s, "none\n"); + mutex_unlock(&filter_mutex); } static struct ftrace_event_field * @@ -150,284 +328,828 @@ find_event_field(struct ftrace_event_call *call, char *name) return NULL; } -void filter_free_pred(struct filter_pred *pred) +static void filter_free_pred(struct filter_pred *pred) { if (!pred) return; kfree(pred->field_name); - kfree(pred->str_val); kfree(pred); } -void filter_free_preds(struct ftrace_event_call *call) +static void filter_clear_pred(struct filter_pred *pred) { - int i; + kfree(pred->field_name); + pred->field_name = NULL; + pred->str_len = 0; +} - if (call->preds) { - for (i = 0; i < MAX_FILTER_PRED; i++) - filter_free_pred(call->preds[i]); - kfree(call->preds); - call->preds = NULL; +static int filter_set_pred(struct filter_pred *dest, + struct filter_pred *src, + filter_pred_fn_t fn) +{ + *dest = *src; + if (src->field_name) { + dest->field_name = kstrdup(src->field_name, GFP_KERNEL); + if (!dest->field_name) + return -ENOMEM; } + dest->fn = fn; + + return 0; } -void filter_free_subsystem_preds(struct event_subsystem *system) +static void filter_disable_preds(struct ftrace_event_call *call) { - struct ftrace_event_call *call = __start_ftrace_events; + struct event_filter *filter = call->filter; int i; - if (system->preds) { - for (i = 0; i < MAX_FILTER_PRED; i++) - filter_free_pred(system->preds[i]); - kfree(system->preds); - system->preds = NULL; - } + call->filter_active = 0; + filter->n_preds = 0; - events_for_each(call) { - if (!call->name || !call->regfunc) - continue; + for (i = 0; i < MAX_FILTER_PRED; i++) + filter->preds[i]->fn = filter_pred_none; +} + +void destroy_preds(struct ftrace_event_call *call) +{ + struct event_filter *filter = call->filter; + int i; - if (!strcmp(call->system, system->name)) - filter_free_preds(call); + for (i = 0; i < MAX_FILTER_PRED; i++) { + if (filter->preds[i]) + filter_free_pred(filter->preds[i]); } + kfree(filter->preds); + kfree(filter); + call->filter = NULL; } -static int __filter_add_pred(struct ftrace_event_call *call, - struct filter_pred *pred) +int init_preds(struct ftrace_event_call *call) { + struct event_filter *filter; + struct filter_pred *pred; int i; - if (call->preds && !pred->compound) - filter_free_preds(call); + filter = call->filter = kzalloc(sizeof(*filter), GFP_KERNEL); + if (!call->filter) + return -ENOMEM; - if (!call->preds) { - call->preds = kzalloc(MAX_FILTER_PRED * sizeof(pred), - GFP_KERNEL); - if (!call->preds) - return -ENOMEM; - } + call->filter_active = 0; + filter->n_preds = 0; + + filter->preds = kzalloc(MAX_FILTER_PRED * sizeof(pred), GFP_KERNEL); + if (!filter->preds) + goto oom; for (i = 0; i < MAX_FILTER_PRED; i++) { - if (!call->preds[i]) { - call->preds[i] = pred; - return 0; + pred = kzalloc(sizeof(*pred), GFP_KERNEL); + if (!pred) + goto oom; + pred->fn = filter_pred_none; + filter->preds[i] = pred; + } + + return 0; + +oom: + destroy_preds(call); + + return -ENOMEM; +} +EXPORT_SYMBOL_GPL(init_preds); + +static void filter_free_subsystem_preds(struct event_subsystem *system) +{ + struct event_filter *filter = system->filter; + struct ftrace_event_call *call; + int i; + + if (filter->n_preds) { + for (i = 0; i < filter->n_preds; i++) + filter_free_pred(filter->preds[i]); + kfree(filter->preds); + filter->preds = NULL; + filter->n_preds = 0; + } + + mutex_lock(&event_mutex); + list_for_each_entry(call, &ftrace_events, list) { + if (!call->define_fields) + continue; + + if (!strcmp(call->system, system->name)) { + filter_disable_preds(call); + remove_filter_string(call->filter); } } + mutex_unlock(&event_mutex); +} + +static int filter_add_pred_fn(struct filter_parse_state *ps, + struct ftrace_event_call *call, + struct filter_pred *pred, + filter_pred_fn_t fn) +{ + struct event_filter *filter = call->filter; + int idx, err; + + if (filter->n_preds == MAX_FILTER_PRED) { + parse_error(ps, FILT_ERR_TOO_MANY_PREDS, 0); + return -ENOSPC; + } + + idx = filter->n_preds; + filter_clear_pred(filter->preds[idx]); + err = filter_set_pred(filter->preds[idx], pred, fn); + if (err) + return err; - return -ENOSPC; + filter->n_preds++; + call->filter_active = 1; + + return 0; } +enum { + FILTER_STATIC_STRING = 1, + FILTER_DYN_STRING +}; + static int is_string_field(const char *type) { + if (strstr(type, "__data_loc") && strstr(type, "char")) + return FILTER_DYN_STRING; + if (strchr(type, '[') && strstr(type, "char")) - return 1; + return FILTER_STATIC_STRING; return 0; } -int filter_add_pred(struct ftrace_event_call *call, struct filter_pred *pred) +static int is_legal_op(struct ftrace_event_field *field, int op) { - struct ftrace_event_field *field; - - field = find_event_field(call, pred->field_name); - if (!field) - return -EINVAL; + if (is_string_field(field->type) && (op != OP_EQ && op != OP_NE)) + return 0; - pred->offset = field->offset; + return 1; +} - if (is_string_field(field->type)) { - if (!pred->str_val) - return -EINVAL; - pred->fn = filter_pred_string; - pred->str_len = field->size; - return __filter_add_pred(call, pred); - } else { - if (pred->str_val) - return -EINVAL; - } +static filter_pred_fn_t select_comparison_fn(int op, int field_size, + int field_is_signed) +{ + filter_pred_fn_t fn = NULL; - switch (field->size) { + switch (field_size) { case 8: - pred->fn = filter_pred_64; + if (op == OP_EQ || op == OP_NE) + fn = filter_pred_64; + else if (field_is_signed) + fn = filter_pred_s64; + else + fn = filter_pred_u64; break; case 4: - pred->fn = filter_pred_32; + if (op == OP_EQ || op == OP_NE) + fn = filter_pred_32; + else if (field_is_signed) + fn = filter_pred_s32; + else + fn = filter_pred_u32; break; case 2: - pred->fn = filter_pred_16; + if (op == OP_EQ || op == OP_NE) + fn = filter_pred_16; + else if (field_is_signed) + fn = filter_pred_s16; + else + fn = filter_pred_u16; break; case 1: - pred->fn = filter_pred_8; + if (op == OP_EQ || op == OP_NE) + fn = filter_pred_8; + else if (field_is_signed) + fn = filter_pred_s8; + else + fn = filter_pred_u8; break; - default: - return -EINVAL; } - return __filter_add_pred(call, pred); + return fn; } -static struct filter_pred *copy_pred(struct filter_pred *pred) +static int filter_add_pred(struct filter_parse_state *ps, + struct ftrace_event_call *call, + struct filter_pred *pred) { - struct filter_pred *new_pred = kmalloc(sizeof(*pred), GFP_KERNEL); - if (!new_pred) - return NULL; + struct ftrace_event_field *field; + filter_pred_fn_t fn; + unsigned long long val; + int string_type; + + pred->fn = filter_pred_none; + + if (pred->op == OP_AND) { + pred->pop_n = 2; + return filter_add_pred_fn(ps, call, pred, filter_pred_and); + } else if (pred->op == OP_OR) { + pred->pop_n = 2; + return filter_add_pred_fn(ps, call, pred, filter_pred_or); + } + + field = find_event_field(call, pred->field_name); + if (!field) { + parse_error(ps, FILT_ERR_FIELD_NOT_FOUND, 0); + return -EINVAL; + } - memcpy(new_pred, pred, sizeof(*pred)); + pred->offset = field->offset; - if (pred->field_name) { - new_pred->field_name = kstrdup(pred->field_name, GFP_KERNEL); - if (!new_pred->field_name) { - kfree(new_pred); - return NULL; - } + if (!is_legal_op(field, pred->op)) { + parse_error(ps, FILT_ERR_ILLEGAL_FIELD_OP, 0); + return -EINVAL; } - if (pred->str_val) { - new_pred->str_val = kstrdup(pred->str_val, GFP_KERNEL); - if (!new_pred->str_val) { - filter_free_pred(new_pred); - return NULL; + string_type = is_string_field(field->type); + if (string_type) { + if (string_type == FILTER_STATIC_STRING) + fn = filter_pred_string; + else + fn = filter_pred_strloc; + pred->str_len = field->size; + if (pred->op == OP_NE) + pred->not = 1; + return filter_add_pred_fn(ps, call, pred, fn); + } else { + if (strict_strtoull(pred->str_val, 0, &val)) { + parse_error(ps, FILT_ERR_ILLEGAL_INTVAL, 0); + return -EINVAL; } + pred->val = val; + } + + fn = select_comparison_fn(pred->op, field->size, field->is_signed); + if (!fn) { + parse_error(ps, FILT_ERR_INVALID_OP, 0); + return -EINVAL; } - return new_pred; + if (pred->op == OP_NE) + pred->not = 1; + + return filter_add_pred_fn(ps, call, pred, fn); } -int filter_add_subsystem_pred(struct event_subsystem *system, - struct filter_pred *pred) +static int filter_add_subsystem_pred(struct filter_parse_state *ps, + struct event_subsystem *system, + struct filter_pred *pred, + char *filter_string) { - struct ftrace_event_call *call = __start_ftrace_events; - struct filter_pred *event_pred; - int i; - - if (system->preds && !pred->compound) - filter_free_subsystem_preds(system); + struct event_filter *filter = system->filter; + struct ftrace_event_call *call; + int err = 0; - if (!system->preds) { - system->preds = kzalloc(MAX_FILTER_PRED * sizeof(pred), + if (!filter->preds) { + filter->preds = kzalloc(MAX_FILTER_PRED * sizeof(pred), GFP_KERNEL); - if (!system->preds) + + if (!filter->preds) return -ENOMEM; } - for (i = 0; i < MAX_FILTER_PRED; i++) { - if (!system->preds[i]) { - system->preds[i] = pred; - break; - } + if (filter->n_preds == MAX_FILTER_PRED) { + parse_error(ps, FILT_ERR_TOO_MANY_PREDS, 0); + return -ENOSPC; } - if (i == MAX_FILTER_PRED) - return -ENOSPC; + filter->preds[filter->n_preds] = pred; + filter->n_preds++; - events_for_each(call) { - int err; + mutex_lock(&event_mutex); + list_for_each_entry(call, &ftrace_events, list) { - if (!call->name || !call->regfunc) + if (!call->define_fields) continue; if (strcmp(call->system, system->name)) continue; - if (!find_event_field(call, pred->field_name)) - continue; + err = filter_add_pred(ps, call, pred); + if (err) { + mutex_unlock(&event_mutex); + filter_free_subsystem_preds(system); + parse_error(ps, FILT_ERR_BAD_SUBSYS_FILTER, 0); + goto out; + } + replace_filter_string(call->filter, filter_string); + } + mutex_unlock(&event_mutex); +out: + return err; +} - event_pred = copy_pred(pred); - if (!event_pred) - goto oom; +static void parse_init(struct filter_parse_state *ps, + struct filter_op *ops, + char *infix_string) +{ + memset(ps, '\0', sizeof(*ps)); - err = filter_add_pred(call, event_pred); - if (err) - filter_free_pred(event_pred); - if (err == -ENOMEM) - goto oom; + ps->infix.string = infix_string; + ps->infix.cnt = strlen(infix_string); + ps->ops = ops; + + INIT_LIST_HEAD(&ps->opstack); + INIT_LIST_HEAD(&ps->postfix); +} + +static char infix_next(struct filter_parse_state *ps) +{ + ps->infix.cnt--; + + return ps->infix.string[ps->infix.tail++]; +} + +static char infix_peek(struct filter_parse_state *ps) +{ + if (ps->infix.tail == strlen(ps->infix.string)) + return 0; + + return ps->infix.string[ps->infix.tail]; +} + +static void infix_advance(struct filter_parse_state *ps) +{ + ps->infix.cnt--; + ps->infix.tail++; +} + +static inline int is_precedence_lower(struct filter_parse_state *ps, + int a, int b) +{ + return ps->ops[a].precedence < ps->ops[b].precedence; +} + +static inline int is_op_char(struct filter_parse_state *ps, char c) +{ + int i; + + for (i = 0; strcmp(ps->ops[i].string, "OP_NONE"); i++) { + if (ps->ops[i].string[0] == c) + return 1; } return 0; +} -oom: - system->preds[i] = NULL; - return -ENOMEM; +static int infix_get_op(struct filter_parse_state *ps, char firstc) +{ + char nextc = infix_peek(ps); + char opstr[3]; + int i; + + opstr[0] = firstc; + opstr[1] = nextc; + opstr[2] = '\0'; + + for (i = 0; strcmp(ps->ops[i].string, "OP_NONE"); i++) { + if (!strcmp(opstr, ps->ops[i].string)) { + infix_advance(ps); + return ps->ops[i].id; + } + } + + opstr[1] = '\0'; + + for (i = 0; strcmp(ps->ops[i].string, "OP_NONE"); i++) { + if (!strcmp(opstr, ps->ops[i].string)) + return ps->ops[i].id; + } + + return OP_NONE; } -int filter_parse(char **pbuf, struct filter_pred *pred) +static inline void clear_operand_string(struct filter_parse_state *ps) { - char *tmp, *tok, *val_str = NULL; - int tok_n = 0; + memset(ps->operand.string, '\0', MAX_FILTER_STR_VAL); + ps->operand.tail = 0; +} - /* field ==/!= number, or/and field ==/!= number, number */ - while ((tok = strsep(pbuf, " \n"))) { - if (tok_n == 0) { - if (!strcmp(tok, "0")) { - pred->clear = 1; - return 0; - } else if (!strcmp(tok, "&&")) { - pred->or = 0; - pred->compound = 1; - } else if (!strcmp(tok, "||")) { - pred->or = 1; - pred->compound = 1; - } else - pred->field_name = tok; - tok_n = 1; +static inline int append_operand_char(struct filter_parse_state *ps, char c) +{ + if (ps->operand.tail == MAX_FILTER_STR_VAL - 1) + return -EINVAL; + + ps->operand.string[ps->operand.tail++] = c; + + return 0; +} + +static int filter_opstack_push(struct filter_parse_state *ps, int op) +{ + struct opstack_op *opstack_op; + + opstack_op = kmalloc(sizeof(*opstack_op), GFP_KERNEL); + if (!opstack_op) + return -ENOMEM; + + opstack_op->op = op; + list_add(&opstack_op->list, &ps->opstack); + + return 0; +} + +static int filter_opstack_empty(struct filter_parse_state *ps) +{ + return list_empty(&ps->opstack); +} + +static int filter_opstack_top(struct filter_parse_state *ps) +{ + struct opstack_op *opstack_op; + + if (filter_opstack_empty(ps)) + return OP_NONE; + + opstack_op = list_first_entry(&ps->opstack, struct opstack_op, list); + + return opstack_op->op; +} + +static int filter_opstack_pop(struct filter_parse_state *ps) +{ + struct opstack_op *opstack_op; + int op; + + if (filter_opstack_empty(ps)) + return OP_NONE; + + opstack_op = list_first_entry(&ps->opstack, struct opstack_op, list); + op = opstack_op->op; + list_del(&opstack_op->list); + + kfree(opstack_op); + + return op; +} + +static void filter_opstack_clear(struct filter_parse_state *ps) +{ + while (!filter_opstack_empty(ps)) + filter_opstack_pop(ps); +} + +static char *curr_operand(struct filter_parse_state *ps) +{ + return ps->operand.string; +} + +static int postfix_append_operand(struct filter_parse_state *ps, char *operand) +{ + struct postfix_elt *elt; + + elt = kmalloc(sizeof(*elt), GFP_KERNEL); + if (!elt) + return -ENOMEM; + + elt->op = OP_NONE; + elt->operand = kstrdup(operand, GFP_KERNEL); + if (!elt->operand) { + kfree(elt); + return -ENOMEM; + } + + list_add_tail(&elt->list, &ps->postfix); + + return 0; +} + +static int postfix_append_op(struct filter_parse_state *ps, int op) +{ + struct postfix_elt *elt; + + elt = kmalloc(sizeof(*elt), GFP_KERNEL); + if (!elt) + return -ENOMEM; + + elt->op = op; + elt->operand = NULL; + + list_add_tail(&elt->list, &ps->postfix); + + return 0; +} + +static void postfix_clear(struct filter_parse_state *ps) +{ + struct postfix_elt *elt; + + while (!list_empty(&ps->postfix)) { + elt = list_first_entry(&ps->postfix, struct postfix_elt, list); + kfree(elt->operand); + list_del(&elt->list); + } +} + +static int filter_parse(struct filter_parse_state *ps) +{ + int in_string = 0; + int op, top_op; + char ch; + + while ((ch = infix_next(ps))) { + if (ch == '"') { + in_string ^= 1; continue; } - if (tok_n == 1) { - if (!pred->field_name) - pred->field_name = tok; - else if (!strcmp(tok, "!=")) - pred->not = 1; - else if (!strcmp(tok, "==")) - pred->not = 0; - else { - pred->field_name = NULL; + + if (in_string) + goto parse_operand; + + if (isspace(ch)) + continue; + + if (is_op_char(ps, ch)) { + op = infix_get_op(ps, ch); + if (op == OP_NONE) { + parse_error(ps, FILT_ERR_INVALID_OP, 0); return -EINVAL; } - tok_n = 2; + + if (strlen(curr_operand(ps))) { + postfix_append_operand(ps, curr_operand(ps)); + clear_operand_string(ps); + } + + while (!filter_opstack_empty(ps)) { + top_op = filter_opstack_top(ps); + if (!is_precedence_lower(ps, top_op, op)) { + top_op = filter_opstack_pop(ps); + postfix_append_op(ps, top_op); + continue; + } + break; + } + + filter_opstack_push(ps, op); continue; } - if (tok_n == 2) { - if (pred->compound) { - if (!strcmp(tok, "!=")) - pred->not = 1; - else if (!strcmp(tok, "==")) - pred->not = 0; - else { - pred->field_name = NULL; - return -EINVAL; - } - } else { - val_str = tok; - break; /* done */ + + if (ch == '(') { + filter_opstack_push(ps, OP_OPEN_PAREN); + continue; + } + + if (ch == ')') { + if (strlen(curr_operand(ps))) { + postfix_append_operand(ps, curr_operand(ps)); + clear_operand_string(ps); + } + + top_op = filter_opstack_pop(ps); + while (top_op != OP_NONE) { + if (top_op == OP_OPEN_PAREN) + break; + postfix_append_op(ps, top_op); + top_op = filter_opstack_pop(ps); + } + if (top_op == OP_NONE) { + parse_error(ps, FILT_ERR_UNBALANCED_PAREN, 0); + return -EINVAL; } - tok_n = 3; continue; } - if (tok_n == 3) { - val_str = tok; - break; /* done */ +parse_operand: + if (append_operand_char(ps, ch)) { + parse_error(ps, FILT_ERR_OPERAND_TOO_LONG, 0); + return -EINVAL; } } - if (!val_str) { - pred->field_name = NULL; - return -EINVAL; + if (strlen(curr_operand(ps))) + postfix_append_operand(ps, curr_operand(ps)); + + while (!filter_opstack_empty(ps)) { + top_op = filter_opstack_pop(ps); + if (top_op == OP_NONE) + break; + if (top_op == OP_OPEN_PAREN) { + parse_error(ps, FILT_ERR_UNBALANCED_PAREN, 0); + return -EINVAL; + } + postfix_append_op(ps, top_op); } - pred->field_name = kstrdup(pred->field_name, GFP_KERNEL); - if (!pred->field_name) - return -ENOMEM; + return 0; +} - pred->val = simple_strtoull(val_str, &tmp, 0); - if (tmp == val_str) { - pred->str_val = kstrdup(val_str, GFP_KERNEL); - if (!pred->str_val) - return -ENOMEM; - } else if (*tmp != '\0') +static struct filter_pred *create_pred(int op, char *operand1, char *operand2) +{ + struct filter_pred *pred; + + pred = kzalloc(sizeof(*pred), GFP_KERNEL); + if (!pred) + return NULL; + + pred->field_name = kstrdup(operand1, GFP_KERNEL); + if (!pred->field_name) { + kfree(pred); + return NULL; + } + + strcpy(pred->str_val, operand2); + pred->str_len = strlen(operand2); + + pred->op = op; + + return pred; +} + +static struct filter_pred *create_logical_pred(int op) +{ + struct filter_pred *pred; + + pred = kzalloc(sizeof(*pred), GFP_KERNEL); + if (!pred) + return NULL; + + pred->op = op; + + return pred; +} + +static int check_preds(struct filter_parse_state *ps) +{ + int n_normal_preds = 0, n_logical_preds = 0; + struct postfix_elt *elt; + + list_for_each_entry(elt, &ps->postfix, list) { + if (elt->op == OP_NONE) + continue; + + if (elt->op == OP_AND || elt->op == OP_OR) { + n_logical_preds++; + continue; + } + n_normal_preds++; + } + + if (!n_normal_preds || n_logical_preds >= n_normal_preds) { + parse_error(ps, FILT_ERR_INVALID_FILTER, 0); return -EINVAL; + } return 0; } +static int replace_preds(struct event_subsystem *system, + struct ftrace_event_call *call, + struct filter_parse_state *ps, + char *filter_string) +{ + char *operand1 = NULL, *operand2 = NULL; + struct filter_pred *pred; + struct postfix_elt *elt; + int err; + + err = check_preds(ps); + if (err) + return err; + + list_for_each_entry(elt, &ps->postfix, list) { + if (elt->op == OP_NONE) { + if (!operand1) + operand1 = elt->operand; + else if (!operand2) + operand2 = elt->operand; + else { + parse_error(ps, FILT_ERR_TOO_MANY_OPERANDS, 0); + return -EINVAL; + } + continue; + } + + if (elt->op == OP_AND || elt->op == OP_OR) { + pred = create_logical_pred(elt->op); + if (call) { + err = filter_add_pred(ps, call, pred); + filter_free_pred(pred); + } else + err = filter_add_subsystem_pred(ps, system, + pred, filter_string); + if (err) + return err; + + operand1 = operand2 = NULL; + continue; + } + + if (!operand1 || !operand2) { + parse_error(ps, FILT_ERR_MISSING_FIELD, 0); + return -EINVAL; + } + + pred = create_pred(elt->op, operand1, operand2); + if (call) { + err = filter_add_pred(ps, call, pred); + filter_free_pred(pred); + } else + err = filter_add_subsystem_pred(ps, system, pred, + filter_string); + if (err) + return err; + + operand1 = operand2 = NULL; + } + + return 0; +} + +int apply_event_filter(struct ftrace_event_call *call, char *filter_string) +{ + int err; + + struct filter_parse_state *ps; + + mutex_lock(&filter_mutex); + + if (!strcmp(strstrip(filter_string), "0")) { + filter_disable_preds(call); + remove_filter_string(call->filter); + mutex_unlock(&filter_mutex); + return 0; + } + + err = -ENOMEM; + ps = kzalloc(sizeof(*ps), GFP_KERNEL); + if (!ps) + goto out_unlock; + + filter_disable_preds(call); + replace_filter_string(call->filter, filter_string); + + parse_init(ps, filter_ops, filter_string); + err = filter_parse(ps); + if (err) { + append_filter_err(ps, call->filter); + goto out; + } + + err = replace_preds(NULL, call, ps, filter_string); + if (err) + append_filter_err(ps, call->filter); + +out: + filter_opstack_clear(ps); + postfix_clear(ps); + kfree(ps); +out_unlock: + mutex_unlock(&filter_mutex); + + return err; +} + +int apply_subsystem_event_filter(struct event_subsystem *system, + char *filter_string) +{ + int err; + + struct filter_parse_state *ps; + + mutex_lock(&filter_mutex); + + if (!strcmp(strstrip(filter_string), "0")) { + filter_free_subsystem_preds(system); + remove_filter_string(system->filter); + mutex_unlock(&filter_mutex); + return 0; + } + + err = -ENOMEM; + ps = kzalloc(sizeof(*ps), GFP_KERNEL); + if (!ps) + goto out_unlock; + + filter_free_subsystem_preds(system); + replace_filter_string(system->filter, filter_string); + + parse_init(ps, filter_ops, filter_string); + err = filter_parse(ps); + if (err) { + append_filter_err(ps, system->filter); + goto out; + } + + err = replace_preds(system, NULL, ps, filter_string); + if (err) + append_filter_err(ps, system->filter); + +out: + filter_opstack_clear(ps); + postfix_clear(ps); + kfree(ps); +out_unlock: + mutex_unlock(&filter_mutex); + + return err; +} diff --git a/kernel/trace/trace_events_stage_1.h b/kernel/trace/trace_events_stage_1.h deleted file mode 100644 index 38985f9b379..00000000000 --- a/kernel/trace/trace_events_stage_1.h +++ /dev/null @@ -1,39 +0,0 @@ -/* - * Stage 1 of the trace events. - * - * Override the macros in <trace/trace_event_types.h> to include the following: - * - * struct ftrace_raw_<call> { - * struct trace_entry ent; - * <type> <item>; - * <type2> <item2>[<len>]; - * [...] - * }; - * - * The <type> <item> is created by the __field(type, item) macro or - * the __array(type2, item2, len) macro. - * We simply do "type item;", and that will create the fields - * in the structure. - */ - -#undef TRACE_FORMAT -#define TRACE_FORMAT(call, proto, args, fmt) - -#undef __array -#define __array(type, item, len) type item[len]; - -#undef __field -#define __field(type, item) type item; - -#undef TP_STRUCT__entry -#define TP_STRUCT__entry(args...) args - -#undef TRACE_EVENT -#define TRACE_EVENT(name, proto, args, tstruct, assign, print) \ - struct ftrace_raw_##name { \ - struct trace_entry ent; \ - tstruct \ - }; \ - static struct ftrace_event_call event_##name - -#include <trace/trace_event_types.h> diff --git a/kernel/trace/trace_events_stage_2.h b/kernel/trace/trace_events_stage_2.h deleted file mode 100644 index d363c6672c6..00000000000 --- a/kernel/trace/trace_events_stage_2.h +++ /dev/null @@ -1,176 +0,0 @@ -/* - * Stage 2 of the trace events. - * - * Override the macros in <trace/trace_event_types.h> to include the following: - * - * enum print_line_t - * ftrace_raw_output_<call>(struct trace_iterator *iter, int flags) - * { - * struct trace_seq *s = &iter->seq; - * struct ftrace_raw_<call> *field; <-- defined in stage 1 - * struct trace_entry *entry; - * int ret; - * - * entry = iter->ent; - * - * if (entry->type != event_<call>.id) { - * WARN_ON_ONCE(1); - * return TRACE_TYPE_UNHANDLED; - * } - * - * field = (typeof(field))entry; - * - * ret = trace_seq_printf(s, <TP_printk> "\n"); - * if (!ret) - * return TRACE_TYPE_PARTIAL_LINE; - * - * return TRACE_TYPE_HANDLED; - * } - * - * This is the method used to print the raw event to the trace - * output format. Note, this is not needed if the data is read - * in binary. - */ - -#undef __entry -#define __entry field - -#undef TP_printk -#define TP_printk(fmt, args...) fmt "\n", args - -#undef TRACE_EVENT -#define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ -enum print_line_t \ -ftrace_raw_output_##call(struct trace_iterator *iter, int flags) \ -{ \ - struct trace_seq *s = &iter->seq; \ - struct ftrace_raw_##call *field; \ - struct trace_entry *entry; \ - int ret; \ - \ - entry = iter->ent; \ - \ - if (entry->type != event_##call.id) { \ - WARN_ON_ONCE(1); \ - return TRACE_TYPE_UNHANDLED; \ - } \ - \ - field = (typeof(field))entry; \ - \ - ret = trace_seq_printf(s, #call ": " print); \ - if (!ret) \ - return TRACE_TYPE_PARTIAL_LINE; \ - \ - return TRACE_TYPE_HANDLED; \ -} - -#include <trace/trace_event_types.h> - -/* - * Setup the showing format of trace point. - * - * int - * ftrace_format_##call(struct trace_seq *s) - * { - * struct ftrace_raw_##call field; - * int ret; - * - * ret = trace_seq_printf(s, #type " " #item ";" - * " offset:%u; size:%u;\n", - * offsetof(struct ftrace_raw_##call, item), - * sizeof(field.type)); - * - * } - */ - -#undef TP_STRUCT__entry -#define TP_STRUCT__entry(args...) args - -#undef __field -#define __field(type, item) \ - ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t" \ - "offset:%u;\tsize:%u;\n", \ - (unsigned int)offsetof(typeof(field), item), \ - (unsigned int)sizeof(field.item)); \ - if (!ret) \ - return 0; - -#undef __array -#define __array(type, item, len) \ - ret = trace_seq_printf(s, "\tfield:" #type " " #item "[" #len "];\t" \ - "offset:%u;\tsize:%u;\n", \ - (unsigned int)offsetof(typeof(field), item), \ - (unsigned int)sizeof(field.item)); \ - if (!ret) \ - return 0; - -#undef __entry -#define __entry REC - -#undef TP_printk -#define TP_printk(fmt, args...) "%s, %s\n", #fmt, __stringify(args) - -#undef TP_fast_assign -#define TP_fast_assign(args...) args - -#undef TRACE_EVENT -#define TRACE_EVENT(call, proto, args, tstruct, func, print) \ -static int \ -ftrace_format_##call(struct trace_seq *s) \ -{ \ - struct ftrace_raw_##call field; \ - int ret; \ - \ - tstruct; \ - \ - trace_seq_printf(s, "\nprint fmt: " print); \ - \ - return ret; \ -} - -#include <trace/trace_event_types.h> - -#undef __field -#define __field(type, item) \ - ret = trace_define_field(event_call, #type, #item, \ - offsetof(typeof(field), item), \ - sizeof(field.item)); \ - if (ret) \ - return ret; - -#undef __array -#define __array(type, item, len) \ - ret = trace_define_field(event_call, #type "[" #len "]", #item, \ - offsetof(typeof(field), item), \ - sizeof(field.item)); \ - if (ret) \ - return ret; - -#define __common_field(type, item) \ - ret = trace_define_field(event_call, #type, "common_" #item, \ - offsetof(typeof(field.ent), item), \ - sizeof(field.ent.item)); \ - if (ret) \ - return ret; - -#undef TRACE_EVENT -#define TRACE_EVENT(call, proto, args, tstruct, func, print) \ -int \ -ftrace_define_fields_##call(void) \ -{ \ - struct ftrace_raw_##call field; \ - struct ftrace_event_call *event_call = &event_##call; \ - int ret; \ - \ - __common_field(unsigned char, type); \ - __common_field(unsigned char, flags); \ - __common_field(unsigned char, preempt_count); \ - __common_field(int, pid); \ - __common_field(int, tgid); \ - \ - tstruct; \ - \ - return ret; \ -} - -#include <trace/trace_event_types.h> diff --git a/kernel/trace/trace_events_stage_3.h b/kernel/trace/trace_events_stage_3.h deleted file mode 100644 index 9d2fa78cecc..00000000000 --- a/kernel/trace/trace_events_stage_3.h +++ /dev/null @@ -1,281 +0,0 @@ -/* - * Stage 3 of the trace events. - * - * Override the macros in <trace/trace_event_types.h> to include the following: - * - * static void ftrace_event_<call>(proto) - * { - * event_trace_printk(_RET_IP_, "<call>: " <fmt>); - * } - * - * static int ftrace_reg_event_<call>(void) - * { - * int ret; - * - * ret = register_trace_<call>(ftrace_event_<call>); - * if (!ret) - * pr_info("event trace: Could not activate trace point " - * "probe to <call>"); - * return ret; - * } - * - * static void ftrace_unreg_event_<call>(void) - * { - * unregister_trace_<call>(ftrace_event_<call>); - * } - * - * For those macros defined with TRACE_FORMAT: - * - * static struct ftrace_event_call __used - * __attribute__((__aligned__(4))) - * __attribute__((section("_ftrace_events"))) event_<call> = { - * .name = "<call>", - * .regfunc = ftrace_reg_event_<call>, - * .unregfunc = ftrace_unreg_event_<call>, - * } - * - * - * For those macros defined with TRACE_EVENT: - * - * static struct ftrace_event_call event_<call>; - * - * static void ftrace_raw_event_<call>(proto) - * { - * struct ring_buffer_event *event; - * struct ftrace_raw_<call> *entry; <-- defined in stage 1 - * unsigned long irq_flags; - * int pc; - * - * local_save_flags(irq_flags); - * pc = preempt_count(); - * - * event = trace_current_buffer_lock_reserve(event_<call>.id, - * sizeof(struct ftrace_raw_<call>), - * irq_flags, pc); - * if (!event) - * return; - * entry = ring_buffer_event_data(event); - * - * <assign>; <-- Here we assign the entries by the __field and - * __array macros. - * - * trace_current_buffer_unlock_commit(event, irq_flags, pc); - * } - * - * static int ftrace_raw_reg_event_<call>(void) - * { - * int ret; - * - * ret = register_trace_<call>(ftrace_raw_event_<call>); - * if (!ret) - * pr_info("event trace: Could not activate trace point " - * "probe to <call>"); - * return ret; - * } - * - * static void ftrace_unreg_event_<call>(void) - * { - * unregister_trace_<call>(ftrace_raw_event_<call>); - * } - * - * static struct trace_event ftrace_event_type_<call> = { - * .trace = ftrace_raw_output_<call>, <-- stage 2 - * }; - * - * static int ftrace_raw_init_event_<call>(void) - * { - * int id; - * - * id = register_ftrace_event(&ftrace_event_type_<call>); - * if (!id) - * return -ENODEV; - * event_<call>.id = id; - * return 0; - * } - * - * static struct ftrace_event_call __used - * __attribute__((__aligned__(4))) - * __attribute__((section("_ftrace_events"))) event_<call> = { - * .name = "<call>", - * .system = "<system>", - * .raw_init = ftrace_raw_init_event_<call>, - * .regfunc = ftrace_reg_event_<call>, - * .unregfunc = ftrace_unreg_event_<call>, - * .show_format = ftrace_format_<call>, - * } - * - */ - -#undef TP_FMT -#define TP_FMT(fmt, args...) fmt "\n", ##args - -#ifdef CONFIG_EVENT_PROFILE -#define _TRACE_PROFILE(call, proto, args) \ -static void ftrace_profile_##call(proto) \ -{ \ - extern void perf_tpcounter_event(int); \ - perf_tpcounter_event(event_##call.id); \ -} \ - \ -static int ftrace_profile_enable_##call(struct ftrace_event_call *call) \ -{ \ - int ret = 0; \ - \ - if (!atomic_inc_return(&call->profile_count)) \ - ret = register_trace_##call(ftrace_profile_##call); \ - \ - return ret; \ -} \ - \ -static void ftrace_profile_disable_##call(struct ftrace_event_call *call) \ -{ \ - if (atomic_add_negative(-1, &call->profile_count)) \ - unregister_trace_##call(ftrace_profile_##call); \ -} - -#define _TRACE_PROFILE_INIT(call) \ - .profile_count = ATOMIC_INIT(-1), \ - .profile_enable = ftrace_profile_enable_##call, \ - .profile_disable = ftrace_profile_disable_##call, - -#else -#define _TRACE_PROFILE(call, proto, args) -#define _TRACE_PROFILE_INIT(call) -#endif - -#define _TRACE_FORMAT(call, proto, args, fmt) \ -static void ftrace_event_##call(proto) \ -{ \ - event_trace_printk(_RET_IP_, #call ": " fmt); \ -} \ - \ -static int ftrace_reg_event_##call(void) \ -{ \ - int ret; \ - \ - ret = register_trace_##call(ftrace_event_##call); \ - if (ret) \ - pr_info("event trace: Could not activate trace point " \ - "probe to " #call "\n"); \ - return ret; \ -} \ - \ -static void ftrace_unreg_event_##call(void) \ -{ \ - unregister_trace_##call(ftrace_event_##call); \ -} \ - \ -static struct ftrace_event_call event_##call; \ - \ -static int ftrace_init_event_##call(void) \ -{ \ - int id; \ - \ - id = register_ftrace_event(NULL); \ - if (!id) \ - return -ENODEV; \ - event_##call.id = id; \ - return 0; \ -} - -#undef TRACE_FORMAT -#define TRACE_FORMAT(call, proto, args, fmt) \ -_TRACE_FORMAT(call, PARAMS(proto), PARAMS(args), PARAMS(fmt)) \ -_TRACE_PROFILE(call, PARAMS(proto), PARAMS(args)) \ -static struct ftrace_event_call __used \ -__attribute__((__aligned__(4))) \ -__attribute__((section("_ftrace_events"))) event_##call = { \ - .name = #call, \ - .system = __stringify(TRACE_SYSTEM), \ - .raw_init = ftrace_init_event_##call, \ - .regfunc = ftrace_reg_event_##call, \ - .unregfunc = ftrace_unreg_event_##call, \ - _TRACE_PROFILE_INIT(call) \ -} - -#undef __entry -#define __entry entry - -#undef TRACE_EVENT -#define TRACE_EVENT(call, proto, args, tstruct, assign, print) \ -_TRACE_PROFILE(call, PARAMS(proto), PARAMS(args)) \ - \ -static struct ftrace_event_call event_##call; \ - \ -static void ftrace_raw_event_##call(proto) \ -{ \ - struct ftrace_event_call *call = &event_##call; \ - struct ring_buffer_event *event; \ - struct ftrace_raw_##call *entry; \ - unsigned long irq_flags; \ - int pc; \ - \ - local_save_flags(irq_flags); \ - pc = preempt_count(); \ - \ - event = trace_current_buffer_lock_reserve(event_##call.id, \ - sizeof(struct ftrace_raw_##call), \ - irq_flags, pc); \ - if (!event) \ - return; \ - entry = ring_buffer_event_data(event); \ - \ - assign; \ - \ - if (call->preds && !filter_match_preds(call, entry)) \ - ring_buffer_event_discard(event); \ - \ - trace_nowake_buffer_unlock_commit(event, irq_flags, pc); \ - \ -} \ - \ -static int ftrace_raw_reg_event_##call(void) \ -{ \ - int ret; \ - \ - ret = register_trace_##call(ftrace_raw_event_##call); \ - if (ret) \ - pr_info("event trace: Could not activate trace point " \ - "probe to " #call "\n"); \ - return ret; \ -} \ - \ -static void ftrace_raw_unreg_event_##call(void) \ -{ \ - unregister_trace_##call(ftrace_raw_event_##call); \ -} \ - \ -static struct trace_event ftrace_event_type_##call = { \ - .trace = ftrace_raw_output_##call, \ -}; \ - \ -static int ftrace_raw_init_event_##call(void) \ -{ \ - int id; \ - \ - id = register_ftrace_event(&ftrace_event_type_##call); \ - if (!id) \ - return -ENODEV; \ - event_##call.id = id; \ - INIT_LIST_HEAD(&event_##call.fields); \ - return 0; \ -} \ - \ -static struct ftrace_event_call __used \ -__attribute__((__aligned__(4))) \ -__attribute__((section("_ftrace_events"))) event_##call = { \ - .name = #call, \ - .system = __stringify(TRACE_SYSTEM), \ - .raw_init = ftrace_raw_init_event_##call, \ - .regfunc = ftrace_raw_reg_event_##call, \ - .unregfunc = ftrace_raw_unreg_event_##call, \ - .show_format = ftrace_format_##call, \ - .define_fields = ftrace_define_fields_##call, \ - _TRACE_PROFILE_INIT(call) \ -} - -#include <trace/trace_event_types.h> - -#undef _TRACE_PROFILE -#undef _TRACE_PROFILE_INIT - diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 07a22c33ebf..d06cf898dc8 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -19,8 +19,12 @@ #undef TRACE_STRUCT #define TRACE_STRUCT(args...) args +extern void __bad_type_size(void); + #undef TRACE_FIELD #define TRACE_FIELD(type, item, assign) \ + if (sizeof(type) != sizeof(field.item)) \ + __bad_type_size(); \ ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t" \ "offset:%u;\tsize:%u;\n", \ (unsigned int)offsetof(typeof(field), item), \ @@ -30,7 +34,7 @@ #undef TRACE_FIELD_SPECIAL -#define TRACE_FIELD_SPECIAL(type_item, item, cmd) \ +#define TRACE_FIELD_SPECIAL(type_item, item, len, cmd) \ ret = trace_seq_printf(s, "\tfield special:" #type_item ";\t" \ "offset:%u;\tsize:%u;\n", \ (unsigned int)offsetof(typeof(field), item), \ @@ -46,6 +50,9 @@ if (!ret) \ return 0; +#undef TRACE_FIELD_SIGN +#define TRACE_FIELD_SIGN(type, item, assign, is_signed) \ + TRACE_FIELD(type, item, assign) #undef TP_RAW_FMT #define TP_RAW_FMT(args...) args @@ -65,6 +72,22 @@ ftrace_format_##call(struct trace_seq *s) \ return ret; \ } +#undef TRACE_EVENT_FORMAT_NOFILTER +#define TRACE_EVENT_FORMAT_NOFILTER(call, proto, args, fmt, tstruct, \ + tpfmt) \ +static int \ +ftrace_format_##call(struct trace_seq *s) \ +{ \ + struct args field; \ + int ret; \ + \ + tstruct; \ + \ + trace_seq_printf(s, "\nprint fmt: \"%s\"\n", tpfmt); \ + \ + return ret; \ +} + #include "trace_event_types.h" #undef TRACE_ZERO_CHAR @@ -78,6 +101,10 @@ ftrace_format_##call(struct trace_seq *s) \ #define TRACE_FIELD(type, item, assign)\ entry->item = assign; +#undef TRACE_FIELD_SIGN +#define TRACE_FIELD_SIGN(type, item, assign, is_signed) \ + TRACE_FIELD(type, item, assign) + #undef TP_CMD #define TP_CMD(cmd...) cmd @@ -85,18 +112,95 @@ ftrace_format_##call(struct trace_seq *s) \ #define TRACE_ENTRY entry #undef TRACE_FIELD_SPECIAL -#define TRACE_FIELD_SPECIAL(type_item, item, cmd) \ +#define TRACE_FIELD_SPECIAL(type_item, item, len, cmd) \ cmd; #undef TRACE_EVENT_FORMAT #define TRACE_EVENT_FORMAT(call, proto, args, fmt, tstruct, tpfmt) \ +int ftrace_define_fields_##call(void); \ +static int ftrace_raw_init_event_##call(void); \ + \ +struct ftrace_event_call __used \ +__attribute__((__aligned__(4))) \ +__attribute__((section("_ftrace_events"))) event_##call = { \ + .name = #call, \ + .id = proto, \ + .system = __stringify(TRACE_SYSTEM), \ + .raw_init = ftrace_raw_init_event_##call, \ + .show_format = ftrace_format_##call, \ + .define_fields = ftrace_define_fields_##call, \ +}; \ +static int ftrace_raw_init_event_##call(void) \ +{ \ + INIT_LIST_HEAD(&event_##call.fields); \ + init_preds(&event_##call); \ + return 0; \ +} \ + +#undef TRACE_EVENT_FORMAT_NOFILTER +#define TRACE_EVENT_FORMAT_NOFILTER(call, proto, args, fmt, tstruct, \ + tpfmt) \ \ -static struct ftrace_event_call __used \ +struct ftrace_event_call __used \ __attribute__((__aligned__(4))) \ __attribute__((section("_ftrace_events"))) event_##call = { \ .name = #call, \ .id = proto, \ .system = __stringify(TRACE_SYSTEM), \ .show_format = ftrace_format_##call, \ +}; + +#include "trace_event_types.h" + +#undef TRACE_FIELD +#define TRACE_FIELD(type, item, assign) \ + ret = trace_define_field(event_call, #type, #item, \ + offsetof(typeof(field), item), \ + sizeof(field.item), is_signed_type(type)); \ + if (ret) \ + return ret; + +#undef TRACE_FIELD_SPECIAL +#define TRACE_FIELD_SPECIAL(type, item, len, cmd) \ + ret = trace_define_field(event_call, #type "[" #len "]", #item, \ + offsetof(typeof(field), item), \ + sizeof(field.item), 0); \ + if (ret) \ + return ret; + +#undef TRACE_FIELD_SIGN +#define TRACE_FIELD_SIGN(type, item, assign, is_signed) \ + ret = trace_define_field(event_call, #type, #item, \ + offsetof(typeof(field), item), \ + sizeof(field.item), is_signed); \ + if (ret) \ + return ret; + +#undef TRACE_FIELD_ZERO_CHAR +#define TRACE_FIELD_ZERO_CHAR(item) + +#undef TRACE_EVENT_FORMAT +#define TRACE_EVENT_FORMAT(call, proto, args, fmt, tstruct, tpfmt) \ +int \ +ftrace_define_fields_##call(void) \ +{ \ + struct ftrace_event_call *event_call = &event_##call; \ + struct args field; \ + int ret; \ + \ + __common_field(unsigned char, type, 0); \ + __common_field(unsigned char, flags, 0); \ + __common_field(unsigned char, preempt_count, 0); \ + __common_field(int, pid, 1); \ + __common_field(int, tgid, 1); \ + \ + tstruct; \ + \ + return ret; \ } + +#undef TRACE_EVENT_FORMAT_NOFILTER +#define TRACE_EVENT_FORMAT_NOFILTER(call, proto, args, fmt, tstruct, \ + tpfmt) + #include "trace_event_types.h" diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index d28687e7b3a..8b592418d8b 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -65,6 +65,12 @@ ftrace_push_return_trace(unsigned long ret, unsigned long func, int *depth) if (!current->ret_stack) return -EBUSY; + /* + * We must make sure the ret_stack is tested before we read + * anything else. + */ + smp_rmb(); + /* The return trace stack is full */ if (current->curr_ret_stack == FTRACE_RETFUNC_DEPTH - 1) { atomic_inc(¤t->trace_overrun); @@ -78,13 +84,14 @@ ftrace_push_return_trace(unsigned long ret, unsigned long func, int *depth) current->ret_stack[index].ret = ret; current->ret_stack[index].func = func; current->ret_stack[index].calltime = calltime; + current->ret_stack[index].subtime = 0; *depth = index; return 0; } /* Retrieve a function return address to the trace stack on thread info.*/ -void +static void ftrace_pop_return_trace(struct ftrace_graph_ret *trace, unsigned long *ret) { int index; @@ -104,9 +111,6 @@ ftrace_pop_return_trace(struct ftrace_graph_ret *trace, unsigned long *ret) trace->calltime = current->ret_stack[index].calltime; trace->overrun = atomic_read(¤t->trace_overrun); trace->depth = index; - barrier(); - current->curr_ret_stack--; - } /* @@ -121,6 +125,8 @@ unsigned long ftrace_return_to_handler(void) ftrace_pop_return_trace(&trace, &ret); trace.rettime = trace_clock_local(); ftrace_graph_return(&trace); + barrier(); + current->curr_ret_stack--; if (unlikely(!ret)) { ftrace_graph_stop(); @@ -426,8 +432,8 @@ print_graph_irq(struct trace_iterator *iter, unsigned long addr, return TRACE_TYPE_HANDLED; } -static enum print_line_t -print_graph_duration(unsigned long long duration, struct trace_seq *s) +enum print_line_t +trace_print_graph_duration(unsigned long long duration, struct trace_seq *s) { unsigned long nsecs_rem = do_div(duration, 1000); /* log10(ULONG_MAX) + '\0' */ @@ -464,12 +470,23 @@ print_graph_duration(unsigned long long duration, struct trace_seq *s) if (!ret) return TRACE_TYPE_PARTIAL_LINE; } + return TRACE_TYPE_HANDLED; +} + +static enum print_line_t +print_graph_duration(unsigned long long duration, struct trace_seq *s) +{ + int ret; + + ret = trace_print_graph_duration(duration, s); + if (ret != TRACE_TYPE_HANDLED) + return ret; ret = trace_seq_printf(s, "| "); if (!ret) return TRACE_TYPE_PARTIAL_LINE; - return TRACE_TYPE_HANDLED; + return TRACE_TYPE_HANDLED; } /* Case of a leaf function on its call entry */ diff --git a/kernel/trace/trace_hw_branches.c b/kernel/trace/trace_hw_branches.c index 7bfdf4c2347..ca7d7c4d0c2 100644 --- a/kernel/trace/trace_hw_branches.c +++ b/kernel/trace/trace_hw_branches.c @@ -1,10 +1,9 @@ /* - * h/w branch tracer for x86 based on bts + * h/w branch tracer for x86 based on BTS * * Copyright (C) 2008-2009 Intel Corporation. * Markus Metzger <markus.t.metzger@gmail.com>, 2008-2009 */ -#include <linux/spinlock.h> #include <linux/kallsyms.h> #include <linux/debugfs.h> #include <linux/ftrace.h> @@ -15,110 +14,119 @@ #include <asm/ds.h> -#include "trace.h" #include "trace_output.h" +#include "trace.h" -#define SIZEOF_BTS (1 << 13) +#define BTS_BUFFER_SIZE (1 << 13) -/* - * The tracer lock protects the below per-cpu tracer array. - * It needs to be held to: - * - start tracing on all cpus - * - stop tracing on all cpus - * - start tracing on a single hotplug cpu - * - stop tracing on a single hotplug cpu - * - read the trace from all cpus - * - read the trace from a single cpu - */ -static DEFINE_SPINLOCK(bts_tracer_lock); static DEFINE_PER_CPU(struct bts_tracer *, tracer); -static DEFINE_PER_CPU(unsigned char[SIZEOF_BTS], buffer); +static DEFINE_PER_CPU(unsigned char[BTS_BUFFER_SIZE], buffer); #define this_tracer per_cpu(tracer, smp_processor_id()) -#define this_buffer per_cpu(buffer, smp_processor_id()) -static int __read_mostly trace_hw_branches_enabled; +static int trace_hw_branches_enabled __read_mostly; +static int trace_hw_branches_suspended __read_mostly; static struct trace_array *hw_branch_trace __read_mostly; -/* - * Start tracing on the current cpu. - * The argument is ignored. - * - * pre: bts_tracer_lock must be locked. - */ -static void bts_trace_start_cpu(void *arg) +static void bts_trace_init_cpu(int cpu) { - if (this_tracer) - ds_release_bts(this_tracer); - - this_tracer = - ds_request_bts(/* task = */ NULL, this_buffer, SIZEOF_BTS, - /* ovfl = */ NULL, /* th = */ (size_t)-1, - BTS_KERNEL); - if (IS_ERR(this_tracer)) { - this_tracer = NULL; - return; - } + per_cpu(tracer, cpu) = + ds_request_bts_cpu(cpu, per_cpu(buffer, cpu), BTS_BUFFER_SIZE, + NULL, (size_t)-1, BTS_KERNEL); + + if (IS_ERR(per_cpu(tracer, cpu))) + per_cpu(tracer, cpu) = NULL; } -static void bts_trace_start(struct trace_array *tr) +static int bts_trace_init(struct trace_array *tr) { - spin_lock(&bts_tracer_lock); + int cpu; + + hw_branch_trace = tr; + trace_hw_branches_enabled = 0; - on_each_cpu(bts_trace_start_cpu, NULL, 1); - trace_hw_branches_enabled = 1; + get_online_cpus(); + for_each_online_cpu(cpu) { + bts_trace_init_cpu(cpu); - spin_unlock(&bts_tracer_lock); + if (likely(per_cpu(tracer, cpu))) + trace_hw_branches_enabled = 1; + } + trace_hw_branches_suspended = 0; + put_online_cpus(); + + /* If we could not enable tracing on a single cpu, we fail. */ + return trace_hw_branches_enabled ? 0 : -EOPNOTSUPP; } -/* - * Stop tracing on the current cpu. - * The argument is ignored. - * - * pre: bts_tracer_lock must be locked. - */ -static void bts_trace_stop_cpu(void *arg) +static void bts_trace_reset(struct trace_array *tr) { - if (this_tracer) { - ds_release_bts(this_tracer); - this_tracer = NULL; + int cpu; + + get_online_cpus(); + for_each_online_cpu(cpu) { + if (likely(per_cpu(tracer, cpu))) { + ds_release_bts(per_cpu(tracer, cpu)); + per_cpu(tracer, cpu) = NULL; + } } + trace_hw_branches_enabled = 0; + trace_hw_branches_suspended = 0; + put_online_cpus(); } -static void bts_trace_stop(struct trace_array *tr) +static void bts_trace_start(struct trace_array *tr) { - spin_lock(&bts_tracer_lock); + int cpu; - trace_hw_branches_enabled = 0; - on_each_cpu(bts_trace_stop_cpu, NULL, 1); + get_online_cpus(); + for_each_online_cpu(cpu) + if (likely(per_cpu(tracer, cpu))) + ds_resume_bts(per_cpu(tracer, cpu)); + trace_hw_branches_suspended = 0; + put_online_cpus(); +} - spin_unlock(&bts_tracer_lock); +static void bts_trace_stop(struct trace_array *tr) +{ + int cpu; + + get_online_cpus(); + for_each_online_cpu(cpu) + if (likely(per_cpu(tracer, cpu))) + ds_suspend_bts(per_cpu(tracer, cpu)); + trace_hw_branches_suspended = 1; + put_online_cpus(); } static int __cpuinit bts_hotcpu_handler(struct notifier_block *nfb, unsigned long action, void *hcpu) { - unsigned int cpu = (unsigned long)hcpu; - - spin_lock(&bts_tracer_lock); - - if (!trace_hw_branches_enabled) - goto out; + int cpu = (long)hcpu; switch (action) { case CPU_ONLINE: case CPU_DOWN_FAILED: - smp_call_function_single(cpu, bts_trace_start_cpu, NULL, 1); + /* The notification is sent with interrupts enabled. */ + if (trace_hw_branches_enabled) { + bts_trace_init_cpu(cpu); + + if (trace_hw_branches_suspended && + likely(per_cpu(tracer, cpu))) + ds_suspend_bts(per_cpu(tracer, cpu)); + } break; + case CPU_DOWN_PREPARE: - smp_call_function_single(cpu, bts_trace_stop_cpu, NULL, 1); - break; + /* The notification is sent with interrupts enabled. */ + if (likely(per_cpu(tracer, cpu))) { + ds_release_bts(per_cpu(tracer, cpu)); + per_cpu(tracer, cpu) = NULL; + } } - out: - spin_unlock(&bts_tracer_lock); return NOTIFY_DONE; } @@ -126,20 +134,6 @@ static struct notifier_block bts_hotcpu_notifier __cpuinitdata = { .notifier_call = bts_hotcpu_handler }; -static int bts_trace_init(struct trace_array *tr) -{ - hw_branch_trace = tr; - - bts_trace_start(tr); - - return 0; -} - -static void bts_trace_reset(struct trace_array *tr) -{ - bts_trace_stop(tr); -} - static void bts_trace_print_header(struct seq_file *m) { seq_puts(m, "# CPU# TO <- FROM\n"); @@ -147,10 +141,10 @@ static void bts_trace_print_header(struct seq_file *m) static enum print_line_t bts_trace_print_line(struct trace_iterator *iter) { + unsigned long symflags = TRACE_ITER_SYM_OFFSET; struct trace_entry *entry = iter->ent; struct trace_seq *seq = &iter->seq; struct hw_branch_entry *it; - unsigned long symflags = TRACE_ITER_SYM_OFFSET; trace_assign_type(it, entry); @@ -168,6 +162,7 @@ static enum print_line_t bts_trace_print_line(struct trace_iterator *iter) void trace_hw_branch(u64 from, u64 to) { + struct ftrace_event_call *call = &event_hw_branch; struct trace_array *tr = hw_branch_trace; struct ring_buffer_event *event; struct hw_branch_entry *entry; @@ -194,7 +189,8 @@ void trace_hw_branch(u64 from, u64 to) entry->ent.type = TRACE_HW_BRANCHES; entry->from = from; entry->to = to; - trace_buffer_unlock_commit(tr, event, 0, 0); + if (!filter_check_discard(call, entry, tr->buffer, event)) + trace_buffer_unlock_commit(tr, event, 0, 0); out: atomic_dec(&tr->data[cpu]->disabled); @@ -224,11 +220,11 @@ static void trace_bts_at(const struct bts_trace *trace, void *at) /* * Collect the trace on the current cpu and write it into the ftrace buffer. * - * pre: bts_tracer_lock must be locked + * pre: tracing must be suspended on the current cpu */ static void trace_bts_cpu(void *arg) { - struct trace_array *tr = (struct trace_array *) arg; + struct trace_array *tr = (struct trace_array *)arg; const struct bts_trace *trace; unsigned char *at; @@ -241,10 +237,9 @@ static void trace_bts_cpu(void *arg) if (unlikely(!this_tracer)) return; - ds_suspend_bts(this_tracer); trace = ds_read_bts(this_tracer); if (!trace) - goto out; + return; for (at = trace->ds.top; (void *)at < trace->ds.end; at += trace->ds.size) @@ -253,18 +248,27 @@ static void trace_bts_cpu(void *arg) for (at = trace->ds.begin; (void *)at < trace->ds.top; at += trace->ds.size) trace_bts_at(trace, at); - -out: - ds_resume_bts(this_tracer); } static void trace_bts_prepare(struct trace_iterator *iter) { - spin_lock(&bts_tracer_lock); + int cpu; + get_online_cpus(); + for_each_online_cpu(cpu) + if (likely(per_cpu(tracer, cpu))) + ds_suspend_bts(per_cpu(tracer, cpu)); + /* + * We need to collect the trace on the respective cpu since ftrace + * implicitly adds the record for the current cpu. + * Once that is more flexible, we could collect the data from any cpu. + */ on_each_cpu(trace_bts_cpu, iter->tr, 1); - spin_unlock(&bts_tracer_lock); + for_each_online_cpu(cpu) + if (likely(per_cpu(tracer, cpu))) + ds_resume_bts(per_cpu(tracer, cpu)); + put_online_cpus(); } static void trace_bts_close(struct trace_iterator *iter) @@ -274,11 +278,11 @@ static void trace_bts_close(struct trace_iterator *iter) void trace_hw_branch_oops(void) { - spin_lock(&bts_tracer_lock); - - trace_bts_cpu(hw_branch_trace); - - spin_unlock(&bts_tracer_lock); + if (this_tracer) { + ds_suspend_bts_noirq(this_tracer); + trace_bts_cpu(hw_branch_trace); + ds_resume_bts_noirq(this_tracer); + } } struct tracer bts_tracer __read_mostly = @@ -291,7 +295,10 @@ struct tracer bts_tracer __read_mostly = .start = bts_trace_start, .stop = bts_trace_stop, .open = trace_bts_prepare, - .close = trace_bts_close + .close = trace_bts_close, +#ifdef CONFIG_FTRACE_SELFTEST + .selftest = trace_selftest_startup_hw_branches, +#endif /* CONFIG_FTRACE_SELFTEST */ }; __init static int init_bts_trace(void) diff --git a/kernel/trace/trace_mmiotrace.c b/kernel/trace/trace_mmiotrace.c index 8e37fcddd8b..d53b45ed080 100644 --- a/kernel/trace/trace_mmiotrace.c +++ b/kernel/trace/trace_mmiotrace.c @@ -9,6 +9,8 @@ #include <linux/kernel.h> #include <linux/mmiotrace.h> #include <linux/pci.h> +#include <linux/time.h> + #include <asm/atomic.h> #include "trace.h" @@ -174,7 +176,7 @@ static enum print_line_t mmio_print_rw(struct trace_iterator *iter) struct mmiotrace_rw *rw; struct trace_seq *s = &iter->seq; unsigned long long t = ns2usecs(iter->ts); - unsigned long usec_rem = do_div(t, 1000000ULL); + unsigned long usec_rem = do_div(t, USEC_PER_SEC); unsigned secs = (unsigned long)t; int ret = 1; @@ -221,7 +223,7 @@ static enum print_line_t mmio_print_map(struct trace_iterator *iter) struct mmiotrace_map *m; struct trace_seq *s = &iter->seq; unsigned long long t = ns2usecs(iter->ts); - unsigned long usec_rem = do_div(t, 1000000ULL); + unsigned long usec_rem = do_div(t, USEC_PER_SEC); unsigned secs = (unsigned long)t; int ret; diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index 64b54a59c55..7938f3ae93e 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -14,11 +14,25 @@ /* must be a power of 2 */ #define EVENT_HASHSIZE 128 -static DEFINE_MUTEX(trace_event_mutex); +DECLARE_RWSEM(trace_event_mutex); + +DEFINE_PER_CPU(struct trace_seq, ftrace_event_seq); +EXPORT_PER_CPU_SYMBOL(ftrace_event_seq); + static struct hlist_head event_hash[EVENT_HASHSIZE] __read_mostly; static int next_event_type = __TRACE_LAST_TYPE + 1; +void trace_print_seq(struct seq_file *m, struct trace_seq *s) +{ + int len = s->len >= PAGE_SIZE ? PAGE_SIZE - 1 : s->len; + + s->buffer[len] = 0; + seq_puts(m, s->buffer); + + trace_seq_init(s); +} + enum print_line_t trace_print_bprintk_msg_only(struct trace_iterator *iter) { struct trace_seq *s = &iter->seq; @@ -84,6 +98,39 @@ trace_seq_printf(struct trace_seq *s, const char *fmt, ...) return len; } +EXPORT_SYMBOL_GPL(trace_seq_printf); + +/** + * trace_seq_vprintf - sequence printing of trace information + * @s: trace sequence descriptor + * @fmt: printf format string + * + * The tracer may use either sequence operations or its own + * copy to user routines. To simplify formating of a trace + * trace_seq_printf is used to store strings into a special + * buffer (@s). Then the output may be either used by + * the sequencer or pulled into another buffer. + */ +int +trace_seq_vprintf(struct trace_seq *s, const char *fmt, va_list args) +{ + int len = (PAGE_SIZE - 1) - s->len; + int ret; + + if (!len) + return 0; + + ret = vsnprintf(s->buffer + s->len, len, fmt, args); + + /* If we can't write it all, don't bother writing anything */ + if (ret >= len) + return 0; + + s->len += ret; + + return len; +} +EXPORT_SYMBOL_GPL(trace_seq_vprintf); int trace_seq_bprintf(struct trace_seq *s, const char *fmt, const u32 *binary) { @@ -201,6 +248,67 @@ int trace_seq_path(struct trace_seq *s, struct path *path) return 0; } +const char * +ftrace_print_flags_seq(struct trace_seq *p, const char *delim, + unsigned long flags, + const struct trace_print_flags *flag_array) +{ + unsigned long mask; + const char *str; + const char *ret = p->buffer + p->len; + int i; + + for (i = 0; flag_array[i].name && flags; i++) { + + mask = flag_array[i].mask; + if ((flags & mask) != mask) + continue; + + str = flag_array[i].name; + flags &= ~mask; + if (p->len && delim) + trace_seq_puts(p, delim); + trace_seq_puts(p, str); + } + + /* check for left over flags */ + if (flags) { + if (p->len && delim) + trace_seq_puts(p, delim); + trace_seq_printf(p, "0x%lx", flags); + } + + trace_seq_putc(p, 0); + + return ret; +} +EXPORT_SYMBOL(ftrace_print_flags_seq); + +const char * +ftrace_print_symbols_seq(struct trace_seq *p, unsigned long val, + const struct trace_print_flags *symbol_array) +{ + int i; + const char *ret = p->buffer + p->len; + + for (i = 0; symbol_array[i].name; i++) { + + if (val != symbol_array[i].mask) + continue; + + trace_seq_puts(p, symbol_array[i].name); + break; + } + + if (!p->len) + trace_seq_printf(p, "0x%lx", val); + + trace_seq_putc(p, 0); + + return ret; +} +EXPORT_SYMBOL(ftrace_print_symbols_seq); + #ifdef CONFIG_KRETPROBES static inline const char *kretprobed(const char *name) { @@ -311,17 +419,20 @@ seq_print_userip_objs(const struct userstack_entry *entry, struct trace_seq *s, if (ip == ULONG_MAX || !ret) break; - if (i && ret) - ret = trace_seq_puts(s, " <- "); + if (ret) + ret = trace_seq_puts(s, " => "); if (!ip) { if (ret) ret = trace_seq_puts(s, "??"); + if (ret) + ret = trace_seq_puts(s, "\n"); continue; } if (!ret) break; if (ret) ret = seq_print_user_ip(s, mm, ip, sym_flags); + ret = trace_seq_puts(s, "\n"); } if (mm) @@ -455,6 +566,7 @@ static int task_state_char(unsigned long state) * @type: the type of event to look for * * Returns an event of type @type otherwise NULL + * Called with trace_event_read_lock() held. */ struct trace_event *ftrace_find_event(int type) { @@ -464,7 +576,7 @@ struct trace_event *ftrace_find_event(int type) key = type & (EVENT_HASHSIZE - 1); - hlist_for_each_entry_rcu(event, n, &event_hash[key], node) { + hlist_for_each_entry(event, n, &event_hash[key], node) { if (event->type == type) return event; } @@ -472,6 +584,46 @@ struct trace_event *ftrace_find_event(int type) return NULL; } +static LIST_HEAD(ftrace_event_list); + +static int trace_search_list(struct list_head **list) +{ + struct trace_event *e; + int last = __TRACE_LAST_TYPE; + + if (list_empty(&ftrace_event_list)) { + *list = &ftrace_event_list; + return last + 1; + } + + /* + * We used up all possible max events, + * lets see if somebody freed one. + */ + list_for_each_entry(e, &ftrace_event_list, list) { + if (e->type != last + 1) + break; + last++; + } + + /* Did we used up all 65 thousand events??? */ + if ((last + 1) > FTRACE_MAX_EVENT) + return 0; + + *list = &e->list; + return last + 1; +} + +void trace_event_read_lock(void) +{ + down_read(&trace_event_mutex); +} + +void trace_event_read_unlock(void) +{ + up_read(&trace_event_mutex); +} + /** * register_ftrace_event - register output for an event type * @event: the event type to register @@ -492,22 +644,42 @@ int register_ftrace_event(struct trace_event *event) unsigned key; int ret = 0; - mutex_lock(&trace_event_mutex); + down_write(&trace_event_mutex); - if (!event) { - ret = next_event_type++; + if (WARN_ON(!event)) goto out; - } - if (!event->type) - event->type = next_event_type++; - else if (event->type > __TRACE_LAST_TYPE) { + INIT_LIST_HEAD(&event->list); + + if (!event->type) { + struct list_head *list = NULL; + + if (next_event_type > FTRACE_MAX_EVENT) { + + event->type = trace_search_list(&list); + if (!event->type) + goto out; + + } else { + + event->type = next_event_type++; + list = &ftrace_event_list; + } + + if (WARN_ON(ftrace_find_event(event->type))) + goto out; + + list_add_tail(&event->list, list); + + } else if (event->type > __TRACE_LAST_TYPE) { printk(KERN_WARNING "Need to add type to trace.h\n"); WARN_ON(1); - } - - if (ftrace_find_event(event->type)) goto out; + } else { + /* Is this event already used */ + if (ftrace_find_event(event->type)) + goto out; + } if (event->trace == NULL) event->trace = trace_nop_print; @@ -520,14 +692,25 @@ int register_ftrace_event(struct trace_event *event) key = event->type & (EVENT_HASHSIZE - 1); - hlist_add_head_rcu(&event->node, &event_hash[key]); + hlist_add_head(&event->node, &event_hash[key]); ret = event->type; out: - mutex_unlock(&trace_event_mutex); + up_write(&trace_event_mutex); return ret; } +EXPORT_SYMBOL_GPL(register_ftrace_event); + +/* + * Used by module code with the trace_event_mutex held for write. + */ +int __unregister_ftrace_event(struct trace_event *event) +{ + hlist_del(&event->node); + list_del(&event->list); + return 0; +} /** * unregister_ftrace_event - remove a no longer used event @@ -535,12 +718,13 @@ int register_ftrace_event(struct trace_event *event) */ int unregister_ftrace_event(struct trace_event *event) { - mutex_lock(&trace_event_mutex); - hlist_del(&event->node); - mutex_unlock(&trace_event_mutex); + down_write(&trace_event_mutex); + __unregister_ftrace_event(event); + up_write(&trace_event_mutex); return 0; } +EXPORT_SYMBOL_GPL(unregister_ftrace_event); /* * Standard events @@ -833,14 +1017,16 @@ static enum print_line_t trace_stack_print(struct trace_iterator *iter, trace_assign_type(field, iter->ent); + if (!trace_seq_puts(s, "<stack trace>\n")) + goto partial; for (i = 0; i < FTRACE_STACK_ENTRIES; i++) { - if (i) { - if (!trace_seq_puts(s, " <= ")) - goto partial; + if (!field->caller[i] || (field->caller[i] == ULONG_MAX)) + break; + if (!trace_seq_puts(s, " => ")) + goto partial; - if (!seq_print_ip_sym(s, field->caller[i], flags)) - goto partial; - } + if (!seq_print_ip_sym(s, field->caller[i], flags)) + goto partial; if (!trace_seq_puts(s, "\n")) goto partial; } @@ -868,10 +1054,10 @@ static enum print_line_t trace_user_stack_print(struct trace_iterator *iter, trace_assign_type(field, iter->ent); - if (!seq_print_userip_objs(field, s, flags)) + if (!trace_seq_puts(s, "<user stack trace>\n")) goto partial; - if (!trace_seq_putc(s, '\n')) + if (!seq_print_userip_objs(field, s, flags)) goto partial; return TRACE_TYPE_HANDLED; diff --git a/kernel/trace/trace_output.h b/kernel/trace/trace_output.h index e0bde39c2dd..d38bec4a9c3 100644 --- a/kernel/trace/trace_output.h +++ b/kernel/trace/trace_output.h @@ -1,41 +1,17 @@ #ifndef __TRACE_EVENTS_H #define __TRACE_EVENTS_H +#include <linux/trace_seq.h> #include "trace.h" -typedef enum print_line_t (*trace_print_func)(struct trace_iterator *iter, - int flags); - -struct trace_event { - struct hlist_node node; - int type; - trace_print_func trace; - trace_print_func raw; - trace_print_func hex; - trace_print_func binary; -}; - extern enum print_line_t trace_print_bprintk_msg_only(struct trace_iterator *iter); extern enum print_line_t trace_print_printk_msg_only(struct trace_iterator *iter); -extern int trace_seq_printf(struct trace_seq *s, const char *fmt, ...) - __attribute__ ((format (printf, 2, 3))); -extern int -trace_seq_bprintf(struct trace_seq *s, const char *fmt, const u32 *binary); extern int seq_print_ip_sym(struct trace_seq *s, unsigned long ip, unsigned long sym_flags); -extern ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, - size_t cnt); -extern int trace_seq_puts(struct trace_seq *s, const char *str); -extern int trace_seq_putc(struct trace_seq *s, unsigned char c); -extern int trace_seq_putmem(struct trace_seq *s, const void *mem, size_t len); -extern int trace_seq_putmem_hex(struct trace_seq *s, const void *mem, - size_t len); -extern void *trace_seq_reserve(struct trace_seq *s, size_t len); -extern int trace_seq_path(struct trace_seq *s, struct path *path); extern int seq_print_userip_objs(const struct userstack_entry *entry, struct trace_seq *s, unsigned long sym_flags); extern int seq_print_user_ip(struct trace_seq *s, struct mm_struct *mm, @@ -44,13 +20,17 @@ extern int seq_print_user_ip(struct trace_seq *s, struct mm_struct *mm, extern int trace_print_context(struct trace_iterator *iter); extern int trace_print_lat_context(struct trace_iterator *iter); +extern void trace_event_read_lock(void); +extern void trace_event_read_unlock(void); extern struct trace_event *ftrace_find_event(int type); -extern int register_ftrace_event(struct trace_event *event); -extern int unregister_ftrace_event(struct trace_event *event); extern enum print_line_t trace_nop_print(struct trace_iterator *iter, int flags); +/* used by module unregistering */ +extern int __unregister_ftrace_event(struct trace_event *event); +extern struct rw_semaphore trace_event_mutex; + #define MAX_MEMHEX_BYTES 8 #define HEX_CHARS (MAX_MEMHEX_BYTES*2 + 1) diff --git a/kernel/trace/trace_power.c b/kernel/trace/trace_power.c index 118439709fb..8a30d9874cd 100644 --- a/kernel/trace/trace_power.c +++ b/kernel/trace/trace_power.c @@ -36,6 +36,7 @@ static void probe_power_start(struct power_trace *it, unsigned int type, static void probe_power_end(struct power_trace *it) { + struct ftrace_event_call *call = &event_power; struct ring_buffer_event *event; struct trace_power *entry; struct trace_array_cpu *data; @@ -54,7 +55,8 @@ static void probe_power_end(struct power_trace *it) goto out; entry = ring_buffer_event_data(event); entry->state_data = *it; - trace_buffer_unlock_commit(tr, event, 0, 0); + if (!filter_check_discard(call, entry, tr->buffer, event)) + trace_buffer_unlock_commit(tr, event, 0, 0); out: preempt_enable(); } @@ -62,6 +64,7 @@ static void probe_power_end(struct power_trace *it) static void probe_power_mark(struct power_trace *it, unsigned int type, unsigned int level) { + struct ftrace_event_call *call = &event_power; struct ring_buffer_event *event; struct trace_power *entry; struct trace_array_cpu *data; @@ -84,7 +87,8 @@ static void probe_power_mark(struct power_trace *it, unsigned int type, goto out; entry = ring_buffer_event_data(event); entry->state_data = *it; - trace_buffer_unlock_commit(tr, event, 0, 0); + if (!filter_check_discard(call, entry, tr->buffer, event)) + trace_buffer_unlock_commit(tr, event, 0, 0); out: preempt_enable(); } diff --git a/kernel/trace/trace_printk.c b/kernel/trace/trace_printk.c index eb81556107f..9bece9687b6 100644 --- a/kernel/trace/trace_printk.c +++ b/kernel/trace/trace_printk.c @@ -245,17 +245,13 @@ static const struct file_operations ftrace_formats_fops = { static __init int init_trace_printk_function_export(void) { struct dentry *d_tracer; - struct dentry *entry; d_tracer = tracing_init_dentry(); if (!d_tracer) return 0; - entry = debugfs_create_file("printk_formats", 0444, d_tracer, + trace_create_file("printk_formats", 0444, d_tracer, NULL, &ftrace_formats_fops); - if (!entry) - pr_warning("Could not create debugfs " - "'printk_formats' entry\n"); return 0; } diff --git a/kernel/trace/trace_sched_switch.c b/kernel/trace/trace_sched_switch.c index 9117cea6f1a..a98106dd979 100644 --- a/kernel/trace/trace_sched_switch.c +++ b/kernel/trace/trace_sched_switch.c @@ -10,7 +10,7 @@ #include <linux/kallsyms.h> #include <linux/uaccess.h> #include <linux/ftrace.h> -#include <trace/sched.h> +#include <trace/events/sched.h> #include "trace.h" @@ -29,13 +29,13 @@ probe_sched_switch(struct rq *__rq, struct task_struct *prev, int cpu; int pc; - if (!sched_ref || sched_stopped) + if (unlikely(!sched_ref)) return; tracing_record_cmdline(prev); tracing_record_cmdline(next); - if (!tracer_enabled) + if (!tracer_enabled || sched_stopped) return; pc = preempt_count(); @@ -56,15 +56,15 @@ probe_sched_wakeup(struct rq *__rq, struct task_struct *wakee, int success) unsigned long flags; int cpu, pc; - if (!likely(tracer_enabled)) + if (unlikely(!sched_ref)) return; - pc = preempt_count(); tracing_record_cmdline(current); - if (sched_stopped) + if (!tracer_enabled || sched_stopped) return; + pc = preempt_count(); local_irq_save(flags); cpu = raw_smp_processor_id(); data = ctx_trace->data[cpu]; diff --git a/kernel/trace/trace_sched_wakeup.c b/kernel/trace/trace_sched_wakeup.c index 5bc00e8f153..eacb2722517 100644 --- a/kernel/trace/trace_sched_wakeup.c +++ b/kernel/trace/trace_sched_wakeup.c @@ -15,7 +15,7 @@ #include <linux/kallsyms.h> #include <linux/uaccess.h> #include <linux/ftrace.h> -#include <trace/sched.h> +#include <trace/events/sched.h> #include "trace.h" @@ -138,9 +138,6 @@ probe_wakeup_sched_switch(struct rq *rq, struct task_struct *prev, pc = preempt_count(); - /* The task we are waiting for is waking up */ - data = wakeup_trace->data[wakeup_cpu]; - /* disable local data, not wakeup_cpu data */ cpu = raw_smp_processor_id(); disabled = atomic_inc_return(&wakeup_trace->data[cpu]->disabled); @@ -154,6 +151,9 @@ probe_wakeup_sched_switch(struct rq *rq, struct task_struct *prev, if (unlikely(!tracer_enabled || next != wakeup_task)) goto out_unlock; + /* The task we are waiting for is waking up */ + data = wakeup_trace->data[wakeup_cpu]; + trace_function(wakeup_trace, CALLER_ADDR0, CALLER_ADDR1, flags, pc); tracing_sched_switch_trace(wakeup_trace, prev, next, flags, pc); diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index 08f4eb2763d..00dd6485bdd 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -16,6 +16,7 @@ static inline int trace_valid_entry(struct trace_entry *entry) case TRACE_BRANCH: case TRACE_GRAPH_ENT: case TRACE_GRAPH_RET: + case TRACE_HW_BRANCHES: return 1; } return 0; @@ -188,6 +189,7 @@ int trace_selftest_startup_dynamic_tracing(struct tracer *trace, #else # define trace_selftest_startup_dynamic_tracing(trace, tr, func) ({ 0; }) #endif /* CONFIG_DYNAMIC_FTRACE */ + /* * Simple verification test of ftrace function tracer. * Enable ftrace, sleep 1/10 second, and then read the trace @@ -749,3 +751,59 @@ trace_selftest_startup_branch(struct tracer *trace, struct trace_array *tr) return ret; } #endif /* CONFIG_BRANCH_TRACER */ + +#ifdef CONFIG_HW_BRANCH_TRACER +int +trace_selftest_startup_hw_branches(struct tracer *trace, + struct trace_array *tr) +{ + struct trace_iterator *iter; + struct tracer tracer; + unsigned long count; + int ret; + + if (!trace->open) { + printk(KERN_CONT "missing open function..."); + return -1; + } + + ret = tracer_init(trace, tr); + if (ret) { + warn_failed_init_tracer(trace, ret); + return ret; + } + + /* + * The hw-branch tracer needs to collect the trace from the various + * cpu trace buffers - before tracing is stopped. + */ + iter = kzalloc(sizeof(*iter), GFP_KERNEL); + if (!iter) + return -ENOMEM; + + memcpy(&tracer, trace, sizeof(tracer)); + + iter->trace = &tracer; + iter->tr = tr; + iter->pos = -1; + mutex_init(&iter->mutex); + + trace->open(iter); + + mutex_destroy(&iter->mutex); + kfree(iter); + + tracing_stop(); + + ret = trace_test_buffer(tr, &count); + trace->reset(tr); + tracing_start(); + + if (!ret && !count) { + printk(KERN_CONT "no entries found.."); + ret = -1; + } + + return ret; +} +#endif /* CONFIG_HW_BRANCH_TRACER */ diff --git a/kernel/trace/trace_stack.c b/kernel/trace/trace_stack.c index c750f65f966..2d7aebd71db 100644 --- a/kernel/trace/trace_stack.c +++ b/kernel/trace/trace_stack.c @@ -265,7 +265,7 @@ static int t_show(struct seq_file *m, void *v) seq_printf(m, " Depth Size Location" " (%d entries)\n" " ----- ---- --------\n", - max_stack_trace.nr_entries); + max_stack_trace.nr_entries - 1); if (!stack_tracer_enabled && !max_stack_size) print_disabled(m); @@ -352,19 +352,14 @@ __setup("stacktrace", enable_stacktrace); static __init int stack_trace_init(void) { struct dentry *d_tracer; - struct dentry *entry; d_tracer = tracing_init_dentry(); - entry = debugfs_create_file("stack_max_size", 0644, d_tracer, - &max_stack_size, &stack_max_size_fops); - if (!entry) - pr_warning("Could not create debugfs 'stack_max_size' entry\n"); + trace_create_file("stack_max_size", 0644, d_tracer, + &max_stack_size, &stack_max_size_fops); - entry = debugfs_create_file("stack_trace", 0444, d_tracer, - NULL, &stack_trace_fops); - if (!entry) - pr_warning("Could not create debugfs 'stack_trace' entry\n"); + trace_create_file("stack_trace", 0444, d_tracer, + NULL, &stack_trace_fops); if (stack_tracer_enabled) register_ftrace_function(&trace_ops); diff --git a/kernel/trace/trace_stat.c b/kernel/trace/trace_stat.c index acdebd771a9..c00643733f4 100644 --- a/kernel/trace/trace_stat.c +++ b/kernel/trace/trace_stat.c @@ -1,7 +1,7 @@ /* * Infrastructure for statistic tracing (histogram output). * - * Copyright (C) 2008 Frederic Weisbecker <fweisbec@gmail.com> + * Copyright (C) 2008-2009 Frederic Weisbecker <fweisbec@gmail.com> * * Based on the code from trace_branch.c which is * Copyright (C) 2008 Steven Rostedt <srostedt@redhat.com> @@ -10,22 +10,27 @@ #include <linux/list.h> +#include <linux/rbtree.h> #include <linux/debugfs.h> #include "trace_stat.h" #include "trace.h" -/* List of stat entries from a tracer */ -struct trace_stat_list { - struct list_head list; +/* + * List of stat red-black nodes from a tracer + * We use a such tree to sort quickly the stat + * entries from the tracer. + */ +struct stat_node { + struct rb_node node; void *stat; }; /* A stat session is the stats output in one file */ -struct tracer_stat_session { +struct stat_session { struct list_head session_list; struct tracer_stat *ts; - struct list_head stat_list; + struct rb_root stat_root; struct mutex stat_mutex; struct dentry *file; }; @@ -37,18 +42,48 @@ static DEFINE_MUTEX(all_stat_sessions_mutex); /* The root directory for all stat files */ static struct dentry *stat_dir; +/* + * Iterate through the rbtree using a post order traversal path + * to release the next node. + * It won't necessary release one at each iteration + * but it will at least advance closer to the next one + * to be released. + */ +static struct rb_node *release_next(struct rb_node *node) +{ + struct stat_node *snode; + struct rb_node *parent = rb_parent(node); + + if (node->rb_left) + return node->rb_left; + else if (node->rb_right) + return node->rb_right; + else { + if (!parent) + ; + else if (parent->rb_left == node) + parent->rb_left = NULL; + else + parent->rb_right = NULL; + + snode = container_of(node, struct stat_node, node); + kfree(snode); + + return parent; + } +} -static void reset_stat_session(struct tracer_stat_session *session) +static void reset_stat_session(struct stat_session *session) { - struct trace_stat_list *node, *next; + struct rb_node *node = session->stat_root.rb_node; - list_for_each_entry_safe(node, next, &session->stat_list, list) - kfree(node); + while (node) + node = release_next(node); - INIT_LIST_HEAD(&session->stat_list); + session->stat_root = RB_ROOT; } -static void destroy_session(struct tracer_stat_session *session) +static void destroy_session(struct stat_session *session) { debugfs_remove(session->file); reset_stat_session(session); @@ -56,25 +91,60 @@ static void destroy_session(struct tracer_stat_session *session) kfree(session); } +typedef int (*cmp_stat_t)(void *, void *); + +static int insert_stat(struct rb_root *root, void *stat, cmp_stat_t cmp) +{ + struct rb_node **new = &(root->rb_node), *parent = NULL; + struct stat_node *data; + + data = kzalloc(sizeof(*data), GFP_KERNEL); + if (!data) + return -ENOMEM; + data->stat = stat; + + /* + * Figure out where to put new node + * This is a descendent sorting + */ + while (*new) { + struct stat_node *this; + int result; + + this = container_of(*new, struct stat_node, node); + result = cmp(data->stat, this->stat); + + parent = *new; + if (result >= 0) + new = &((*new)->rb_left); + else + new = &((*new)->rb_right); + } + + rb_link_node(&data->node, parent, new); + rb_insert_color(&data->node, root); + return 0; +} + /* * For tracers that don't provide a stat_cmp callback. - * This one will force an immediate insertion on tail of - * the list. + * This one will force an insertion as right-most node + * in the rbtree. */ static int dummy_cmp(void *p1, void *p2) { - return 1; + return -1; } /* - * Initialize the stat list at each trace_stat file opening. + * Initialize the stat rbtree at each trace_stat file opening. * All of these copies and sorting are required on all opening * since the stats could have changed between two file sessions. */ -static int stat_seq_init(struct tracer_stat_session *session) +static int stat_seq_init(struct stat_session *session) { - struct trace_stat_list *iter_entry, *new_entry; struct tracer_stat *ts = session->ts; + struct rb_root *root = &session->stat_root; void *stat; int ret = 0; int i; @@ -85,29 +155,16 @@ static int stat_seq_init(struct tracer_stat_session *session) if (!ts->stat_cmp) ts->stat_cmp = dummy_cmp; - stat = ts->stat_start(); + stat = ts->stat_start(ts); if (!stat) goto exit; - /* - * The first entry. Actually this is the second, but the first - * one (the stat_list head) is pointless. - */ - new_entry = kmalloc(sizeof(struct trace_stat_list), GFP_KERNEL); - if (!new_entry) { - ret = -ENOMEM; + ret = insert_stat(root, stat, ts->stat_cmp); + if (ret) goto exit; - } - - INIT_LIST_HEAD(&new_entry->list); - - list_add(&new_entry->list, &session->stat_list); - - new_entry->stat = stat; /* - * Iterate over the tracer stat entries and store them in a sorted - * list. + * Iterate over the tracer stat entries and store them in an rbtree. */ for (i = 1; ; i++) { stat = ts->stat_next(stat, i); @@ -116,36 +173,16 @@ static int stat_seq_init(struct tracer_stat_session *session) if (!stat) break; - new_entry = kmalloc(sizeof(struct trace_stat_list), GFP_KERNEL); - if (!new_entry) { - ret = -ENOMEM; - goto exit_free_list; - } - - INIT_LIST_HEAD(&new_entry->list); - new_entry->stat = stat; - - list_for_each_entry_reverse(iter_entry, &session->stat_list, - list) { - - /* Insertion with a descendent sorting */ - if (ts->stat_cmp(iter_entry->stat, - new_entry->stat) >= 0) { - - list_add(&new_entry->list, &iter_entry->list); - break; - } - } - - /* The current larger value */ - if (list_empty(&new_entry->list)) - list_add(&new_entry->list, &session->stat_list); + ret = insert_stat(root, stat, ts->stat_cmp); + if (ret) + goto exit_free_rbtree; } + exit: mutex_unlock(&session->stat_mutex); return ret; -exit_free_list: +exit_free_rbtree: reset_stat_session(session); mutex_unlock(&session->stat_mutex); return ret; @@ -154,38 +191,51 @@ exit_free_list: static void *stat_seq_start(struct seq_file *s, loff_t *pos) { - struct tracer_stat_session *session = s->private; + struct stat_session *session = s->private; + struct rb_node *node; + int i; - /* Prevent from tracer switch or stat_list modification */ + /* Prevent from tracer switch or rbtree modification */ mutex_lock(&session->stat_mutex); /* If we are in the beginning of the file, print the headers */ - if (!*pos && session->ts->stat_headers) + if (!*pos && session->ts->stat_headers) { + (*pos)++; return SEQ_START_TOKEN; + } - return seq_list_start(&session->stat_list, *pos); + node = rb_first(&session->stat_root); + for (i = 0; node && i < *pos; i++) + node = rb_next(node); + + (*pos)++; + + return node; } static void *stat_seq_next(struct seq_file *s, void *p, loff_t *pos) { - struct tracer_stat_session *session = s->private; + struct stat_session *session = s->private; + struct rb_node *node = p; + + (*pos)++; if (p == SEQ_START_TOKEN) - return seq_list_start(&session->stat_list, *pos); + return rb_first(&session->stat_root); - return seq_list_next(p, &session->stat_list, pos); + return rb_next(node); } static void stat_seq_stop(struct seq_file *s, void *p) { - struct tracer_stat_session *session = s->private; + struct stat_session *session = s->private; mutex_unlock(&session->stat_mutex); } static int stat_seq_show(struct seq_file *s, void *v) { - struct tracer_stat_session *session = s->private; - struct trace_stat_list *l = list_entry(v, struct trace_stat_list, list); + struct stat_session *session = s->private; + struct stat_node *l = container_of(v, struct stat_node, node); if (v == SEQ_START_TOKEN) return session->ts->stat_headers(s); @@ -205,7 +255,7 @@ static int tracing_stat_open(struct inode *inode, struct file *file) { int ret; - struct tracer_stat_session *session = inode->i_private; + struct stat_session *session = inode->i_private; ret = seq_open(file, &trace_stat_seq_ops); if (!ret) { @@ -218,11 +268,11 @@ static int tracing_stat_open(struct inode *inode, struct file *file) } /* - * Avoid consuming memory with our now useless list. + * Avoid consuming memory with our now useless rbtree. */ static int tracing_stat_release(struct inode *i, struct file *f) { - struct tracer_stat_session *session = i->i_private; + struct stat_session *session = i->i_private; mutex_lock(&session->stat_mutex); reset_stat_session(session); @@ -251,7 +301,7 @@ static int tracing_stat_init(void) return 0; } -static int init_stat_file(struct tracer_stat_session *session) +static int init_stat_file(struct stat_session *session) { if (!stat_dir && tracing_stat_init()) return -ENODEV; @@ -266,7 +316,7 @@ static int init_stat_file(struct tracer_stat_session *session) int register_stat_tracer(struct tracer_stat *trace) { - struct tracer_stat_session *session, *node, *tmp; + struct stat_session *session, *node; int ret; if (!trace) @@ -277,7 +327,7 @@ int register_stat_tracer(struct tracer_stat *trace) /* Already registered? */ mutex_lock(&all_stat_sessions_mutex); - list_for_each_entry_safe(node, tmp, &all_stat_sessions, session_list) { + list_for_each_entry(node, &all_stat_sessions, session_list) { if (node->ts == trace) { mutex_unlock(&all_stat_sessions_mutex); return -EINVAL; @@ -286,15 +336,13 @@ int register_stat_tracer(struct tracer_stat *trace) mutex_unlock(&all_stat_sessions_mutex); /* Init the session */ - session = kmalloc(sizeof(struct tracer_stat_session), GFP_KERNEL); + session = kzalloc(sizeof(*session), GFP_KERNEL); if (!session) return -ENOMEM; session->ts = trace; INIT_LIST_HEAD(&session->session_list); - INIT_LIST_HEAD(&session->stat_list); mutex_init(&session->stat_mutex); - session->file = NULL; ret = init_stat_file(session); if (ret) { @@ -312,7 +360,7 @@ int register_stat_tracer(struct tracer_stat *trace) void unregister_stat_tracer(struct tracer_stat *trace) { - struct tracer_stat_session *node, *tmp; + struct stat_session *node, *tmp; mutex_lock(&all_stat_sessions_mutex); list_for_each_entry_safe(node, tmp, &all_stat_sessions, session_list) { diff --git a/kernel/trace/trace_stat.h b/kernel/trace/trace_stat.h index 202274cf7f3..f3546a2cd82 100644 --- a/kernel/trace/trace_stat.h +++ b/kernel/trace/trace_stat.h @@ -12,7 +12,7 @@ struct tracer_stat { /* The name of your stat file */ const char *name; /* Iteration over statistic entries */ - void *(*stat_start)(void); + void *(*stat_start)(struct tracer_stat *trace); void *(*stat_next)(void *prev, int idx); /* Compare two entries for stats sorting */ int (*stat_cmp)(void *p1, void *p2); diff --git a/kernel/trace/trace_sysprof.c b/kernel/trace/trace_sysprof.c index 91fd19c2149..e04b76cc238 100644 --- a/kernel/trace/trace_sysprof.c +++ b/kernel/trace/trace_sysprof.c @@ -321,11 +321,7 @@ static const struct file_operations sysprof_sample_fops = { void init_tracer_sysprof_debugfs(struct dentry *d_tracer) { - struct dentry *entry; - entry = debugfs_create_file("sysprof_sample_period", 0644, + trace_create_file("sysprof_sample_period", 0644, d_tracer, NULL, &sysprof_sample_fops); - if (entry) - return; - pr_warning("Could not create debugfs 'sysprof_sample_period' entry\n"); } diff --git a/kernel/trace/trace_workqueue.c b/kernel/trace/trace_workqueue.c index 797201e4a13..97fcea4acce 100644 --- a/kernel/trace/trace_workqueue.c +++ b/kernel/trace/trace_workqueue.c @@ -6,7 +6,7 @@ */ -#include <trace/workqueue.h> +#include <trace/events/workqueue.h> #include <linux/list.h> #include <linux/percpu.h> #include "trace_stat.h" @@ -16,8 +16,6 @@ /* A cpu workqueue thread */ struct cpu_workqueue_stats { struct list_head list; -/* Useful to know if we print the cpu headers */ - bool first_entry; int cpu; pid_t pid; /* Can be inserted from interrupt or user context, need to be atomic */ @@ -47,12 +45,11 @@ probe_workqueue_insertion(struct task_struct *wq_thread, struct work_struct *work) { int cpu = cpumask_first(&wq_thread->cpus_allowed); - struct cpu_workqueue_stats *node, *next; + struct cpu_workqueue_stats *node; unsigned long flags; spin_lock_irqsave(&workqueue_cpu_stat(cpu)->lock, flags); - list_for_each_entry_safe(node, next, &workqueue_cpu_stat(cpu)->list, - list) { + list_for_each_entry(node, &workqueue_cpu_stat(cpu)->list, list) { if (node->pid == wq_thread->pid) { atomic_inc(&node->inserted); goto found; @@ -69,12 +66,11 @@ probe_workqueue_execution(struct task_struct *wq_thread, struct work_struct *work) { int cpu = cpumask_first(&wq_thread->cpus_allowed); - struct cpu_workqueue_stats *node, *next; + struct cpu_workqueue_stats *node; unsigned long flags; spin_lock_irqsave(&workqueue_cpu_stat(cpu)->lock, flags); - list_for_each_entry_safe(node, next, &workqueue_cpu_stat(cpu)->list, - list) { + list_for_each_entry(node, &workqueue_cpu_stat(cpu)->list, list) { if (node->pid == wq_thread->pid) { node->executed++; goto found; @@ -105,8 +101,6 @@ static void probe_workqueue_creation(struct task_struct *wq_thread, int cpu) cws->pid = wq_thread->pid; spin_lock_irqsave(&workqueue_cpu_stat(cpu)->lock, flags); - if (list_empty(&workqueue_cpu_stat(cpu)->list)) - cws->first_entry = true; list_add_tail(&cws->list, &workqueue_cpu_stat(cpu)->list); spin_unlock_irqrestore(&workqueue_cpu_stat(cpu)->lock, flags); } @@ -152,7 +146,7 @@ static struct cpu_workqueue_stats *workqueue_stat_start_cpu(int cpu) return ret; } -static void *workqueue_stat_start(void) +static void *workqueue_stat_start(struct tracer_stat *trace) { int cpu; void *ret = NULL; @@ -191,16 +185,9 @@ static void *workqueue_stat_next(void *prev, int idx) static int workqueue_stat_show(struct seq_file *s, void *p) { struct cpu_workqueue_stats *cws = p; - unsigned long flags; - int cpu = cws->cpu; struct pid *pid; struct task_struct *tsk; - spin_lock_irqsave(&workqueue_cpu_stat(cpu)->lock, flags); - if (&cws->list == workqueue_cpu_stat(cpu)->list.next) - seq_printf(s, "\n"); - spin_unlock_irqrestore(&workqueue_cpu_stat(cpu)->lock, flags); - pid = find_get_pid(cws->pid); if (pid) { tsk = get_pid_task(pid, PIDTYPE_PID); diff --git a/kernel/wait.c b/kernel/wait.c index 42a2dbc181c..ea7c3b4275c 100644 --- a/kernel/wait.c +++ b/kernel/wait.c @@ -154,7 +154,7 @@ void abort_exclusive_wait(wait_queue_head_t *q, wait_queue_t *wait, if (!list_empty(&wait->task_list)) list_del_init(&wait->task_list); else if (waitqueue_active(q)) - __wake_up_common(q, mode, 1, 0, key); + __wake_up_locked_key(q, mode, key); spin_unlock_irqrestore(&q->lock, flags); } EXPORT_SYMBOL(abort_exclusive_wait); diff --git a/kernel/workqueue.c b/kernel/workqueue.c index f71fb2a0895..0668795d881 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -33,7 +33,8 @@ #include <linux/kallsyms.h> #include <linux/debug_locks.h> #include <linux/lockdep.h> -#include <trace/workqueue.h> +#define CREATE_TRACE_POINTS +#include <trace/events/workqueue.h> /* * The per-CPU workqueue (if single thread, we always use the first @@ -124,8 +125,6 @@ struct cpu_workqueue_struct *get_wq_data(struct work_struct *work) return (void *) (atomic_long_read(&work->data) & WORK_STRUCT_WQ_DATA_MASK); } -DEFINE_TRACE(workqueue_insertion); - static void insert_work(struct cpu_workqueue_struct *cwq, struct work_struct *work, struct list_head *head) { @@ -262,8 +261,6 @@ int queue_delayed_work_on(int cpu, struct workqueue_struct *wq, } EXPORT_SYMBOL_GPL(queue_delayed_work_on); -DEFINE_TRACE(workqueue_execution); - static void run_workqueue(struct cpu_workqueue_struct *cwq) { spin_lock_irq(&cwq->lock); @@ -753,8 +750,6 @@ init_cpu_workqueue(struct workqueue_struct *wq, int cpu) return cwq; } -DEFINE_TRACE(workqueue_creation); - static int create_workqueue_thread(struct cpu_workqueue_struct *cwq, int cpu) { struct sched_param param = { .sched_priority = MAX_RT_PRIO-1 }; @@ -860,8 +855,6 @@ struct workqueue_struct *__create_workqueue_key(const char *name, } EXPORT_SYMBOL_GPL(__create_workqueue_key); -DEFINE_TRACE(workqueue_destruction); - static void cleanup_workqueue_thread(struct cpu_workqueue_struct *cwq) { /* diff --git a/lib/dma-debug.c b/lib/dma-debug.c index 69da09a085a..ad65fc0317d 100644 --- a/lib/dma-debug.c +++ b/lib/dma-debug.c @@ -23,9 +23,11 @@ #include <linux/dma-debug.h> #include <linux/spinlock.h> #include <linux/debugfs.h> +#include <linux/uaccess.h> #include <linux/device.h> #include <linux/types.h> #include <linux/sched.h> +#include <linux/ctype.h> #include <linux/list.h> #include <linux/slab.h> @@ -85,6 +87,7 @@ static u32 show_num_errors = 1; static u32 num_free_entries; static u32 min_free_entries; +static u32 nr_total_entries; /* number of preallocated entries requested by kernel cmdline */ static u32 req_entries; @@ -97,6 +100,16 @@ static struct dentry *show_all_errors_dent __read_mostly; static struct dentry *show_num_errors_dent __read_mostly; static struct dentry *num_free_entries_dent __read_mostly; static struct dentry *min_free_entries_dent __read_mostly; +static struct dentry *filter_dent __read_mostly; + +/* per-driver filter related state */ + +#define NAME_MAX_LEN 64 + +static char current_driver_name[NAME_MAX_LEN] __read_mostly; +static struct device_driver *current_driver __read_mostly; + +static DEFINE_RWLOCK(driver_name_lock); static const char *type2name[4] = { "single", "page", "scather-gather", "coherent" }; @@ -104,6 +117,11 @@ static const char *type2name[4] = { "single", "page", static const char *dir2name[4] = { "DMA_BIDIRECTIONAL", "DMA_TO_DEVICE", "DMA_FROM_DEVICE", "DMA_NONE" }; +/* little merge helper - remove it after the merge window */ +#ifndef BUS_NOTIFY_UNBOUND_DRIVER +#define BUS_NOTIFY_UNBOUND_DRIVER 0x0005 +#endif + /* * The access to some variables in this macro is racy. We can't use atomic_t * here because all these variables are exported to debugfs. Some of them even @@ -121,15 +139,54 @@ static inline void dump_entry_trace(struct dma_debug_entry *entry) { #ifdef CONFIG_STACKTRACE if (entry) { - printk(KERN_WARNING "Mapped at:\n"); + pr_warning("Mapped at:\n"); print_stack_trace(&entry->stacktrace, 0); } #endif } +static bool driver_filter(struct device *dev) +{ + struct device_driver *drv; + unsigned long flags; + bool ret; + + /* driver filter off */ + if (likely(!current_driver_name[0])) + return true; + + /* driver filter on and initialized */ + if (current_driver && dev->driver == current_driver) + return true; + + if (current_driver || !current_driver_name[0]) + return false; + + /* driver filter on but not yet initialized */ + drv = get_driver(dev->driver); + if (!drv) + return false; + + /* lock to protect against change of current_driver_name */ + read_lock_irqsave(&driver_name_lock, flags); + + ret = false; + if (drv->name && + strncmp(current_driver_name, drv->name, NAME_MAX_LEN - 1) == 0) { + current_driver = drv; + ret = true; + } + + read_unlock_irqrestore(&driver_name_lock, flags); + put_driver(drv); + + return ret; +} + #define err_printk(dev, entry, format, arg...) do { \ error_count += 1; \ - if (show_all_errors || show_num_errors > 0) { \ + if (driver_filter(dev) && \ + (show_all_errors || show_num_errors > 0)) { \ WARN(1, "%s %s: " format, \ dev_driver_string(dev), \ dev_name(dev) , ## arg); \ @@ -185,15 +242,50 @@ static void put_hash_bucket(struct hash_bucket *bucket, static struct dma_debug_entry *hash_bucket_find(struct hash_bucket *bucket, struct dma_debug_entry *ref) { - struct dma_debug_entry *entry; + struct dma_debug_entry *entry, *ret = NULL; + int matches = 0, match_lvl, last_lvl = 0; list_for_each_entry(entry, &bucket->list, list) { - if ((entry->dev_addr == ref->dev_addr) && - (entry->dev == ref->dev)) + if ((entry->dev_addr != ref->dev_addr) || + (entry->dev != ref->dev)) + continue; + + /* + * Some drivers map the same physical address multiple + * times. Without a hardware IOMMU this results in the + * same device addresses being put into the dma-debug + * hash multiple times too. This can result in false + * positives being reported. Therfore we implement a + * best-fit algorithm here which returns the entry from + * the hash which fits best to the reference value + * instead of the first-fit. + */ + matches += 1; + match_lvl = 0; + entry->size == ref->size ? ++match_lvl : match_lvl; + entry->type == ref->type ? ++match_lvl : match_lvl; + entry->direction == ref->direction ? ++match_lvl : match_lvl; + + if (match_lvl == 3) { + /* perfect-fit - return the result */ return entry; + } else if (match_lvl > last_lvl) { + /* + * We found an entry that fits better then the + * previous one + */ + last_lvl = match_lvl; + ret = entry; + } } - return NULL; + /* + * If we have multiple matches but no perfect-fit, just return + * NULL. + */ + ret = (matches == 1) ? ret : NULL; + + return ret; } /* @@ -257,6 +349,21 @@ static void add_dma_entry(struct dma_debug_entry *entry) put_hash_bucket(bucket, &flags); } +static struct dma_debug_entry *__dma_entry_alloc(void) +{ + struct dma_debug_entry *entry; + + entry = list_entry(free_entries.next, struct dma_debug_entry, list); + list_del(&entry->list); + memset(entry, 0, sizeof(*entry)); + + num_free_entries -= 1; + if (num_free_entries < min_free_entries) + min_free_entries = num_free_entries; + + return entry; +} + /* struct dma_entry allocator * * The next two functions implement the allocator for @@ -270,15 +377,12 @@ static struct dma_debug_entry *dma_entry_alloc(void) spin_lock_irqsave(&free_entries_lock, flags); if (list_empty(&free_entries)) { - printk(KERN_ERR "DMA-API: debugging out of memory " - "- disabling\n"); + pr_err("DMA-API: debugging out of memory - disabling\n"); global_disable = true; goto out; } - entry = list_entry(free_entries.next, struct dma_debug_entry, list); - list_del(&entry->list); - memset(entry, 0, sizeof(*entry)); + entry = __dma_entry_alloc(); #ifdef CONFIG_STACKTRACE entry->stacktrace.max_entries = DMA_DEBUG_STACKTRACE_ENTRIES; @@ -286,9 +390,6 @@ static struct dma_debug_entry *dma_entry_alloc(void) entry->stacktrace.skip = 2; save_stack_trace(&entry->stacktrace); #endif - num_free_entries -= 1; - if (num_free_entries < min_free_entries) - min_free_entries = num_free_entries; out: spin_unlock_irqrestore(&free_entries_lock, flags); @@ -310,6 +411,53 @@ static void dma_entry_free(struct dma_debug_entry *entry) spin_unlock_irqrestore(&free_entries_lock, flags); } +int dma_debug_resize_entries(u32 num_entries) +{ + int i, delta, ret = 0; + unsigned long flags; + struct dma_debug_entry *entry; + LIST_HEAD(tmp); + + spin_lock_irqsave(&free_entries_lock, flags); + + if (nr_total_entries < num_entries) { + delta = num_entries - nr_total_entries; + + spin_unlock_irqrestore(&free_entries_lock, flags); + + for (i = 0; i < delta; i++) { + entry = kzalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) + break; + + list_add_tail(&entry->list, &tmp); + } + + spin_lock_irqsave(&free_entries_lock, flags); + + list_splice(&tmp, &free_entries); + nr_total_entries += i; + num_free_entries += i; + } else { + delta = nr_total_entries - num_entries; + + for (i = 0; i < delta && !list_empty(&free_entries); i++) { + entry = __dma_entry_alloc(); + kfree(entry); + } + + nr_total_entries -= i; + } + + if (nr_total_entries != num_entries) + ret = 1; + + spin_unlock_irqrestore(&free_entries_lock, flags); + + return ret; +} +EXPORT_SYMBOL(dma_debug_resize_entries); + /* * DMA-API debugging init code * @@ -334,8 +482,7 @@ static int prealloc_memory(u32 num_entries) num_free_entries = num_entries; min_free_entries = num_entries; - printk(KERN_INFO "DMA-API: preallocated %d debug entries\n", - num_entries); + pr_info("DMA-API: preallocated %d debug entries\n", num_entries); return 0; @@ -349,11 +496,102 @@ out_err: return -ENOMEM; } +static ssize_t filter_read(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) +{ + char buf[NAME_MAX_LEN + 1]; + unsigned long flags; + int len; + + if (!current_driver_name[0]) + return 0; + + /* + * We can't copy to userspace directly because current_driver_name can + * only be read under the driver_name_lock with irqs disabled. So + * create a temporary copy first. + */ + read_lock_irqsave(&driver_name_lock, flags); + len = scnprintf(buf, NAME_MAX_LEN + 1, "%s\n", current_driver_name); + read_unlock_irqrestore(&driver_name_lock, flags); + + return simple_read_from_buffer(user_buf, count, ppos, buf, len); +} + +static ssize_t filter_write(struct file *file, const char __user *userbuf, + size_t count, loff_t *ppos) +{ + char buf[NAME_MAX_LEN]; + unsigned long flags; + size_t len; + int i; + + /* + * We can't copy from userspace directly. Access to + * current_driver_name is protected with a write_lock with irqs + * disabled. Since copy_from_user can fault and may sleep we + * need to copy to temporary buffer first + */ + len = min(count, (size_t)(NAME_MAX_LEN - 1)); + if (copy_from_user(buf, userbuf, len)) + return -EFAULT; + + buf[len] = 0; + + write_lock_irqsave(&driver_name_lock, flags); + + /* + * Now handle the string we got from userspace very carefully. + * The rules are: + * - only use the first token we got + * - token delimiter is everything looking like a space + * character (' ', '\n', '\t' ...) + * + */ + if (!isalnum(buf[0])) { + /* + * If the first character userspace gave us is not + * alphanumerical then assume the filter should be + * switched off. + */ + if (current_driver_name[0]) + pr_info("DMA-API: switching off dma-debug driver filter\n"); + current_driver_name[0] = 0; + current_driver = NULL; + goto out_unlock; + } + + /* + * Now parse out the first token and use it as the name for the + * driver to filter for. + */ + for (i = 0; i < NAME_MAX_LEN; ++i) { + current_driver_name[i] = buf[i]; + if (isspace(buf[i]) || buf[i] == ' ' || buf[i] == 0) + break; + } + current_driver_name[i] = 0; + current_driver = NULL; + + pr_info("DMA-API: enable driver filter for driver [%s]\n", + current_driver_name); + +out_unlock: + write_unlock_irqrestore(&driver_name_lock, flags); + + return count; +} + +const struct file_operations filter_fops = { + .read = filter_read, + .write = filter_write, +}; + static int dma_debug_fs_init(void) { dma_debug_dent = debugfs_create_dir("dma-api", NULL); if (!dma_debug_dent) { - printk(KERN_ERR "DMA-API: can not create debugfs directory\n"); + pr_err("DMA-API: can not create debugfs directory\n"); return -ENOMEM; } @@ -392,6 +630,11 @@ static int dma_debug_fs_init(void) if (!min_free_entries_dent) goto out_err; + filter_dent = debugfs_create_file("driver_filter", 0644, + dma_debug_dent, NULL, &filter_fops); + if (!filter_dent) + goto out_err; + return 0; out_err: @@ -400,9 +643,64 @@ out_err: return -ENOMEM; } +static int device_dma_allocations(struct device *dev) +{ + struct dma_debug_entry *entry; + unsigned long flags; + int count = 0, i; + + local_irq_save(flags); + + for (i = 0; i < HASH_SIZE; ++i) { + spin_lock(&dma_entry_hash[i].lock); + list_for_each_entry(entry, &dma_entry_hash[i].list, list) { + if (entry->dev == dev) + count += 1; + } + spin_unlock(&dma_entry_hash[i].lock); + } + + local_irq_restore(flags); + + return count; +} + +static int dma_debug_device_change(struct notifier_block *nb, + unsigned long action, void *data) +{ + struct device *dev = data; + int count; + + + switch (action) { + case BUS_NOTIFY_UNBOUND_DRIVER: + count = device_dma_allocations(dev); + if (count == 0) + break; + err_printk(dev, NULL, "DMA-API: device driver has pending " + "DMA allocations while released from device " + "[count=%d]\n", count); + break; + default: + break; + } + + return 0; +} + void dma_debug_add_bus(struct bus_type *bus) { - /* FIXME: register notifier */ + struct notifier_block *nb; + + nb = kzalloc(sizeof(struct notifier_block), GFP_KERNEL); + if (nb == NULL) { + pr_err("dma_debug_add_bus: out of memory\n"); + return; + } + + nb->notifier_call = dma_debug_device_change; + + bus_register_notifier(bus, nb); } /* @@ -421,8 +719,7 @@ void dma_debug_init(u32 num_entries) } if (dma_debug_fs_init() != 0) { - printk(KERN_ERR "DMA-API: error creating debugfs entries " - "- disabling\n"); + pr_err("DMA-API: error creating debugfs entries - disabling\n"); global_disable = true; return; @@ -432,14 +729,15 @@ void dma_debug_init(u32 num_entries) num_entries = req_entries; if (prealloc_memory(num_entries) != 0) { - printk(KERN_ERR "DMA-API: debugging out of memory error " - "- disabled\n"); + pr_err("DMA-API: debugging out of memory error - disabled\n"); global_disable = true; return; } - printk(KERN_INFO "DMA-API: debugging enabled by kernel config\n"); + nr_total_entries = num_free_entries; + + pr_info("DMA-API: debugging enabled by kernel config\n"); } static __init int dma_debug_cmdline(char *str) @@ -448,8 +746,7 @@ static __init int dma_debug_cmdline(char *str) return -EINVAL; if (strncmp(str, "off", 3) == 0) { - printk(KERN_INFO "DMA-API: debugging disabled on kernel " - "command line\n"); + pr_info("DMA-API: debugging disabled on kernel command line\n"); global_disable = true; } @@ -723,15 +1020,15 @@ void debug_dma_map_sg(struct device *dev, struct scatterlist *sg, entry->type = dma_debug_sg; entry->dev = dev; entry->paddr = sg_phys(s); - entry->size = s->length; - entry->dev_addr = s->dma_address; + entry->size = sg_dma_len(s); + entry->dev_addr = sg_dma_address(s); entry->direction = direction; entry->sg_call_ents = nents; entry->sg_mapped_ents = mapped_ents; if (!PageHighMem(sg_page(s))) { check_for_stack(dev, sg_virt(s)); - check_for_illegal_area(dev, sg_virt(s), s->length); + check_for_illegal_area(dev, sg_virt(s), sg_dma_len(s)); } add_dma_entry(entry); @@ -739,13 +1036,33 @@ void debug_dma_map_sg(struct device *dev, struct scatterlist *sg, } EXPORT_SYMBOL(debug_dma_map_sg); +static int get_nr_mapped_entries(struct device *dev, struct scatterlist *s) +{ + struct dma_debug_entry *entry, ref; + struct hash_bucket *bucket; + unsigned long flags; + int mapped_ents; + + ref.dev = dev; + ref.dev_addr = sg_dma_address(s); + ref.size = sg_dma_len(s), + + bucket = get_hash_bucket(&ref, &flags); + entry = hash_bucket_find(bucket, &ref); + mapped_ents = 0; + + if (entry) + mapped_ents = entry->sg_mapped_ents; + put_hash_bucket(bucket, &flags); + + return mapped_ents; +} + void debug_dma_unmap_sg(struct device *dev, struct scatterlist *sglist, int nelems, int dir) { - struct dma_debug_entry *entry; struct scatterlist *s; int mapped_ents = 0, i; - unsigned long flags; if (unlikely(global_disable)) return; @@ -756,8 +1073,8 @@ void debug_dma_unmap_sg(struct device *dev, struct scatterlist *sglist, .type = dma_debug_sg, .dev = dev, .paddr = sg_phys(s), - .dev_addr = s->dma_address, - .size = s->length, + .dev_addr = sg_dma_address(s), + .size = sg_dma_len(s), .direction = dir, .sg_call_ents = 0, }; @@ -765,14 +1082,9 @@ void debug_dma_unmap_sg(struct device *dev, struct scatterlist *sglist, if (mapped_ents && i >= mapped_ents) break; - if (mapped_ents == 0) { - struct hash_bucket *bucket; + if (!i) { ref.sg_call_ents = nelems; - bucket = get_hash_bucket(&ref, &flags); - entry = hash_bucket_find(bucket, &ref); - if (entry) - mapped_ents = entry->sg_mapped_ents; - put_hash_bucket(bucket, &flags); + mapped_ents = get_nr_mapped_entries(dev, s); } check_unmap(&ref); @@ -874,14 +1186,20 @@ void debug_dma_sync_sg_for_cpu(struct device *dev, struct scatterlist *sg, int nelems, int direction) { struct scatterlist *s; - int i; + int mapped_ents = 0, i; if (unlikely(global_disable)) return; for_each_sg(sg, s, nelems, i) { - check_sync(dev, s->dma_address, s->dma_length, 0, - direction, true); + if (!i) + mapped_ents = get_nr_mapped_entries(dev, s); + + if (i >= mapped_ents) + break; + + check_sync(dev, sg_dma_address(s), sg_dma_len(s), 0, + direction, true); } } EXPORT_SYMBOL(debug_dma_sync_sg_for_cpu); @@ -890,15 +1208,39 @@ void debug_dma_sync_sg_for_device(struct device *dev, struct scatterlist *sg, int nelems, int direction) { struct scatterlist *s; - int i; + int mapped_ents = 0, i; if (unlikely(global_disable)) return; for_each_sg(sg, s, nelems, i) { - check_sync(dev, s->dma_address, s->dma_length, 0, - direction, false); + if (!i) + mapped_ents = get_nr_mapped_entries(dev, s); + + if (i >= mapped_ents) + break; + + check_sync(dev, sg_dma_address(s), sg_dma_len(s), 0, + direction, false); } } EXPORT_SYMBOL(debug_dma_sync_sg_for_device); +static int __init dma_debug_driver_setup(char *str) +{ + int i; + + for (i = 0; i < NAME_MAX_LEN - 1; ++i, ++str) { + current_driver_name[i] = *str; + if (*str == 0) + break; + } + + if (current_driver_name[0]) + pr_info("DMA-API: enable driver filter for driver [%s]\n", + current_driver_name); + + + return 1; +} +__setup("dma_debug_driver=", dma_debug_driver_setup); diff --git a/lib/swiotlb.c b/lib/swiotlb.c index 2b0b5a7d2ce..bffe6d7ef9d 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -60,8 +60,8 @@ enum dma_sync_target { int swiotlb_force; /* - * Used to do a quick range check in swiotlb_unmap_single and - * swiotlb_sync_single_*, to see if the memory was in fact allocated by this + * Used to do a quick range check in unmap_single and + * sync_single_*, to see if the memory was in fact allocated by this * API. */ static char *io_tlb_start, *io_tlb_end; @@ -129,7 +129,7 @@ dma_addr_t __weak swiotlb_phys_to_bus(struct device *hwdev, phys_addr_t paddr) return paddr; } -phys_addr_t __weak swiotlb_bus_to_phys(dma_addr_t baddr) +phys_addr_t __weak swiotlb_bus_to_phys(struct device *hwdev, dma_addr_t baddr) { return baddr; } @@ -140,9 +140,15 @@ static dma_addr_t swiotlb_virt_to_bus(struct device *hwdev, return swiotlb_phys_to_bus(hwdev, virt_to_phys(address)); } -static void *swiotlb_bus_to_virt(dma_addr_t address) +void * __weak swiotlb_bus_to_virt(struct device *hwdev, dma_addr_t address) { - return phys_to_virt(swiotlb_bus_to_phys(address)); + return phys_to_virt(swiotlb_bus_to_phys(hwdev, address)); +} + +int __weak swiotlb_arch_address_needs_mapping(struct device *hwdev, + dma_addr_t addr, size_t size) +{ + return !is_buffer_dma_capable(dma_get_mask(hwdev), addr, size); } int __weak swiotlb_arch_range_needs_mapping(phys_addr_t paddr, size_t size) @@ -309,10 +315,10 @@ cleanup1: return -ENOMEM; } -static int +static inline int address_needs_mapping(struct device *hwdev, dma_addr_t addr, size_t size) { - return !is_buffer_dma_capable(dma_get_mask(hwdev), addr, size); + return swiotlb_arch_address_needs_mapping(hwdev, addr, size); } static inline int range_needs_mapping(phys_addr_t paddr, size_t size) @@ -341,7 +347,7 @@ static void swiotlb_bounce(phys_addr_t phys, char *dma_addr, size_t size, unsigned long flags; while (size) { - sz = min(PAGE_SIZE - offset, size); + sz = min_t(size_t, PAGE_SIZE - offset, size); local_irq_save(flags); buffer = kmap_atomic(pfn_to_page(pfn), @@ -476,7 +482,7 @@ found: * dma_addr is the kernel virtual address of the bounce buffer to unmap. */ static void -unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir) +do_unmap_single(struct device *hwdev, char *dma_addr, size_t size, int dir) { unsigned long flags; int i, count, nslots = ALIGN(size, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT; @@ -560,7 +566,6 @@ swiotlb_alloc_coherent(struct device *hwdev, size_t size, size)) { /* * The allocated memory isn't reachable by the device. - * Fall back on swiotlb_map_single(). */ free_pages((unsigned long) ret, order); ret = NULL; @@ -568,9 +573,8 @@ swiotlb_alloc_coherent(struct device *hwdev, size_t size, if (!ret) { /* * We are either out of memory or the device can't DMA - * to GFP_DMA memory; fall back on - * swiotlb_map_single(), which will grab memory from - * the lowest available address range. + * to GFP_DMA memory; fall back on map_single(), which + * will grab memory from the lowest available address range. */ ret = map_single(hwdev, 0, size, DMA_FROM_DEVICE); if (!ret) @@ -587,7 +591,7 @@ swiotlb_alloc_coherent(struct device *hwdev, size_t size, (unsigned long long)dev_addr); /* DMA_TO_DEVICE to avoid memcpy in unmap_single */ - unmap_single(hwdev, ret, size, DMA_TO_DEVICE); + do_unmap_single(hwdev, ret, size, DMA_TO_DEVICE); return NULL; } *dma_handle = dev_addr; @@ -604,7 +608,7 @@ swiotlb_free_coherent(struct device *hwdev, size_t size, void *vaddr, free_pages((unsigned long) vaddr, get_order(size)); else /* DMA_TO_DEVICE to avoid memcpy in unmap_single */ - unmap_single(hwdev, vaddr, size, DMA_TO_DEVICE); + do_unmap_single(hwdev, vaddr, size, DMA_TO_DEVICE); } EXPORT_SYMBOL(swiotlb_free_coherent); @@ -634,7 +638,7 @@ swiotlb_full(struct device *dev, size_t size, int dir, int do_panic) * physical address to use is returned. * * Once the device is given the dma address, the device owns this memory until - * either swiotlb_unmap_single or swiotlb_dma_sync_single is performed. + * either swiotlb_unmap_page or swiotlb_dma_sync_single is performed. */ dma_addr_t swiotlb_map_page(struct device *dev, struct page *page, unsigned long offset, size_t size, @@ -642,18 +646,17 @@ dma_addr_t swiotlb_map_page(struct device *dev, struct page *page, struct dma_attrs *attrs) { phys_addr_t phys = page_to_phys(page) + offset; - void *ptr = page_address(page) + offset; dma_addr_t dev_addr = swiotlb_phys_to_bus(dev, phys); void *map; BUG_ON(dir == DMA_NONE); /* - * If the pointer passed in happens to be in the device's DMA window, + * If the address happens to be in the device's DMA window, * we can safely return the device addr and not worry about bounce * buffering it. */ if (!address_needs_mapping(dev, dev_addr, size) && - !range_needs_mapping(virt_to_phys(ptr), size)) + !range_needs_mapping(phys, size)) return dev_addr; /* @@ -679,23 +682,35 @@ EXPORT_SYMBOL_GPL(swiotlb_map_page); /* * Unmap a single streaming mode DMA translation. The dma_addr and size must - * match what was provided for in a previous swiotlb_map_single call. All + * match what was provided for in a previous swiotlb_map_page call. All * other usages are undefined. * * After this call, reads by the cpu to the buffer are guaranteed to see * whatever the device wrote there. */ +static void unmap_single(struct device *hwdev, dma_addr_t dev_addr, + size_t size, int dir) +{ + char *dma_addr = swiotlb_bus_to_virt(hwdev, dev_addr); + + BUG_ON(dir == DMA_NONE); + + if (is_swiotlb_buffer(dma_addr)) { + do_unmap_single(hwdev, dma_addr, size, dir); + return; + } + + if (dir != DMA_FROM_DEVICE) + return; + + dma_mark_clean(dma_addr, size); +} + void swiotlb_unmap_page(struct device *hwdev, dma_addr_t dev_addr, size_t size, enum dma_data_direction dir, struct dma_attrs *attrs) { - char *dma_addr = swiotlb_bus_to_virt(dev_addr); - - BUG_ON(dir == DMA_NONE); - if (is_swiotlb_buffer(dma_addr)) - unmap_single(hwdev, dma_addr, size, dir); - else if (dir == DMA_FROM_DEVICE) - dma_mark_clean(dma_addr, size); + unmap_single(hwdev, dev_addr, size, dir); } EXPORT_SYMBOL_GPL(swiotlb_unmap_page); @@ -703,7 +718,7 @@ EXPORT_SYMBOL_GPL(swiotlb_unmap_page); * Make physical memory consistent for a single streaming mode DMA translation * after a transfer. * - * If you perform a swiotlb_map_single() but wish to interrogate the buffer + * If you perform a swiotlb_map_page() but wish to interrogate the buffer * using the cpu, yet do not wish to teardown the dma mapping, you must * call this function before doing so. At the next point you give the dma * address back to the card, you must first perform a @@ -713,13 +728,19 @@ static void swiotlb_sync_single(struct device *hwdev, dma_addr_t dev_addr, size_t size, int dir, int target) { - char *dma_addr = swiotlb_bus_to_virt(dev_addr); + char *dma_addr = swiotlb_bus_to_virt(hwdev, dev_addr); BUG_ON(dir == DMA_NONE); - if (is_swiotlb_buffer(dma_addr)) + + if (is_swiotlb_buffer(dma_addr)) { sync_single(hwdev, dma_addr, size, dir, target); - else if (dir == DMA_FROM_DEVICE) - dma_mark_clean(dma_addr, size); + return; + } + + if (dir != DMA_FROM_DEVICE) + return; + + dma_mark_clean(dma_addr, size); } void @@ -746,13 +767,7 @@ swiotlb_sync_single_range(struct device *hwdev, dma_addr_t dev_addr, unsigned long offset, size_t size, int dir, int target) { - char *dma_addr = swiotlb_bus_to_virt(dev_addr) + offset; - - BUG_ON(dir == DMA_NONE); - if (is_swiotlb_buffer(dma_addr)) - sync_single(hwdev, dma_addr, size, dir, target); - else if (dir == DMA_FROM_DEVICE) - dma_mark_clean(dma_addr, size); + swiotlb_sync_single(hwdev, dev_addr + offset, size, dir, target); } void @@ -777,7 +792,7 @@ EXPORT_SYMBOL_GPL(swiotlb_sync_single_range_for_device); /* * Map a set of buffers described by scatterlist in streaming mode for DMA. - * This is the scatter-gather version of the above swiotlb_map_single + * This is the scatter-gather version of the above swiotlb_map_page * interface. Here the scatter gather list elements are each tagged with the * appropriate dma address and length. They are obtained via * sg_dma_{address,length}(SG). @@ -788,7 +803,7 @@ EXPORT_SYMBOL_GPL(swiotlb_sync_single_range_for_device); * The routine returns the number of addr/length pairs actually * used, at most nents. * - * Device ownership issues as mentioned above for swiotlb_map_single are the + * Device ownership issues as mentioned above for swiotlb_map_page are the * same here. */ int @@ -836,7 +851,7 @@ EXPORT_SYMBOL(swiotlb_map_sg); /* * Unmap a set of streaming mode DMA translations. Again, cpu read rules - * concerning calls here are the same as for swiotlb_unmap_single() above. + * concerning calls here are the same as for swiotlb_unmap_page() above. */ void swiotlb_unmap_sg_attrs(struct device *hwdev, struct scatterlist *sgl, @@ -847,13 +862,9 @@ swiotlb_unmap_sg_attrs(struct device *hwdev, struct scatterlist *sgl, BUG_ON(dir == DMA_NONE); - for_each_sg(sgl, sg, nelems, i) { - if (sg->dma_address != swiotlb_phys_to_bus(hwdev, sg_phys(sg))) - unmap_single(hwdev, swiotlb_bus_to_virt(sg->dma_address), - sg->dma_length, dir); - else if (dir == DMA_FROM_DEVICE) - dma_mark_clean(swiotlb_bus_to_virt(sg->dma_address), sg->dma_length); - } + for_each_sg(sgl, sg, nelems, i) + unmap_single(hwdev, sg->dma_address, sg->dma_length, dir); + } EXPORT_SYMBOL(swiotlb_unmap_sg_attrs); @@ -879,15 +890,9 @@ swiotlb_sync_sg(struct device *hwdev, struct scatterlist *sgl, struct scatterlist *sg; int i; - BUG_ON(dir == DMA_NONE); - - for_each_sg(sgl, sg, nelems, i) { - if (sg->dma_address != swiotlb_phys_to_bus(hwdev, sg_phys(sg))) - sync_single(hwdev, swiotlb_bus_to_virt(sg->dma_address), + for_each_sg(sgl, sg, nelems, i) + swiotlb_sync_single(hwdev, sg->dma_address, sg->dma_length, dir, target); - else if (dir == DMA_FROM_DEVICE) - dma_mark_clean(swiotlb_bus_to_virt(sg->dma_address), sg->dma_length); - } } void diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 7536acea135..756ccafa9ce 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -408,6 +408,8 @@ enum format_type { FORMAT_TYPE_LONG_LONG, FORMAT_TYPE_ULONG, FORMAT_TYPE_LONG, + FORMAT_TYPE_UBYTE, + FORMAT_TYPE_BYTE, FORMAT_TYPE_USHORT, FORMAT_TYPE_SHORT, FORMAT_TYPE_UINT, @@ -573,12 +575,15 @@ static char *string(char *buf, char *end, char *s, struct printf_spec spec) } static char *symbol_string(char *buf, char *end, void *ptr, - struct printf_spec spec) + struct printf_spec spec, char ext) { unsigned long value = (unsigned long) ptr; #ifdef CONFIG_KALLSYMS char sym[KSYM_SYMBOL_LEN]; - sprint_symbol(sym, value); + if (ext != 'f') + sprint_symbol(sym, value); + else + kallsyms_lookup(value, NULL, NULL, NULL, sym); return string(buf, end, sym, spec); #else spec.field_width = 2*sizeof(void *); @@ -690,7 +695,8 @@ static char *ip4_addr_string(char *buf, char *end, u8 *addr, * * Right now we handle: * - * - 'F' For symbolic function descriptor pointers + * - 'F' For symbolic function descriptor pointers with offset + * - 'f' For simple symbolic function names without offset * - 'S' For symbolic direct pointers * - 'R' For a struct resource pointer, it prints the range of * addresses (not the name nor the flags) @@ -713,10 +719,11 @@ static char *pointer(const char *fmt, char *buf, char *end, void *ptr, switch (*fmt) { case 'F': + case 'f': ptr = dereference_function_descriptor(ptr); /* Fallthrough */ case 'S': - return symbol_string(buf, end, ptr, spec); + return symbol_string(buf, end, ptr, spec, *fmt); case 'R': return resource_string(buf, end, ptr, spec); case 'm': @@ -853,11 +860,15 @@ qualifier: spec->qualifier = -1; if (*fmt == 'h' || *fmt == 'l' || *fmt == 'L' || *fmt == 'Z' || *fmt == 'z' || *fmt == 't') { - spec->qualifier = *fmt; - ++fmt; - if (spec->qualifier == 'l' && *fmt == 'l') { - spec->qualifier = 'L'; - ++fmt; + spec->qualifier = *fmt++; + if (unlikely(spec->qualifier == *fmt)) { + if (spec->qualifier == 'l') { + spec->qualifier = 'L'; + ++fmt; + } else if (spec->qualifier == 'h') { + spec->qualifier = 'H'; + ++fmt; + } } } @@ -919,6 +930,11 @@ qualifier: spec->type = FORMAT_TYPE_SIZE_T; } else if (spec->qualifier == 't') { spec->type = FORMAT_TYPE_PTRDIFF; + } else if (spec->qualifier == 'H') { + if (spec->flags & SIGN) + spec->type = FORMAT_TYPE_BYTE; + else + spec->type = FORMAT_TYPE_UBYTE; } else if (spec->qualifier == 'h') { if (spec->flags & SIGN) spec->type = FORMAT_TYPE_SHORT; @@ -943,7 +959,8 @@ qualifier: * * This function follows C99 vsnprintf, but has some extensions: * %pS output the name of a text symbol - * %pF output the name of a function pointer + * %pF output the name of a function pointer with its offset + * %pf output the name of a function pointer without its offset * %pR output the address range in a struct resource * * The return value is the number of characters which would @@ -1087,6 +1104,12 @@ int vsnprintf(char *buf, size_t size, const char *fmt, va_list args) case FORMAT_TYPE_PTRDIFF: num = va_arg(args, ptrdiff_t); break; + case FORMAT_TYPE_UBYTE: + num = (unsigned char) va_arg(args, int); + break; + case FORMAT_TYPE_BYTE: + num = (signed char) va_arg(args, int); + break; case FORMAT_TYPE_USHORT: num = (unsigned short) va_arg(args, int); break; @@ -1363,6 +1386,10 @@ do { \ case FORMAT_TYPE_PTRDIFF: save_arg(ptrdiff_t); break; + case FORMAT_TYPE_UBYTE: + case FORMAT_TYPE_BYTE: + save_arg(char); + break; case FORMAT_TYPE_USHORT: case FORMAT_TYPE_SHORT: save_arg(short); @@ -1391,7 +1418,8 @@ EXPORT_SYMBOL_GPL(vbin_printf); * * The format follows C99 vsnprintf, but has some extensions: * %pS output the name of a text symbol - * %pF output the name of a function pointer + * %pF output the name of a function pointer with its offset + * %pf output the name of a function pointer without its offset * %pR output the address range in a struct resource * %n is ignored * @@ -1538,6 +1566,12 @@ int bstr_printf(char *buf, size_t size, const char *fmt, const u32 *bin_buf) case FORMAT_TYPE_PTRDIFF: num = get_arg(ptrdiff_t); break; + case FORMAT_TYPE_UBYTE: + num = get_arg(unsigned char); + break; + case FORMAT_TYPE_BYTE: + num = get_arg(signed char); + break; case FORMAT_TYPE_USHORT: num = get_arg(unsigned short); break; diff --git a/mm/bounce.c b/mm/bounce.c index e590272fe7a..65f5e17e411 100644 --- a/mm/bounce.c +++ b/mm/bounce.c @@ -14,16 +14,15 @@ #include <linux/hash.h> #include <linux/highmem.h> #include <linux/blktrace_api.h> -#include <trace/block.h> #include <asm/tlbflush.h> +#include <trace/events/block.h> + #define POOL_SIZE 64 #define ISA_POOL_SIZE 16 static mempool_t *page_pool, *isa_page_pool; -DEFINE_TRACE(block_bio_bounce); - #ifdef CONFIG_HIGHMEM static __init int init_emergency_pool(void) { diff --git a/mm/mlock.c b/mm/mlock.c index cbe9e0581b7..ac130433c7d 100644 --- a/mm/mlock.c +++ b/mm/mlock.c @@ -629,52 +629,43 @@ void user_shm_unlock(size_t size, struct user_struct *user) free_uid(user); } -void *alloc_locked_buffer(size_t size) +int account_locked_memory(struct mm_struct *mm, struct rlimit *rlim, + size_t size) { - unsigned long rlim, vm, pgsz; - void *buffer = NULL; + unsigned long lim, vm, pgsz; + int error = -ENOMEM; pgsz = PAGE_ALIGN(size) >> PAGE_SHIFT; - down_write(¤t->mm->mmap_sem); - - rlim = current->signal->rlim[RLIMIT_AS].rlim_cur >> PAGE_SHIFT; - vm = current->mm->total_vm + pgsz; - if (rlim < vm) - goto out; + down_write(&mm->mmap_sem); - rlim = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur >> PAGE_SHIFT; - vm = current->mm->locked_vm + pgsz; - if (rlim < vm) + lim = rlim[RLIMIT_AS].rlim_cur >> PAGE_SHIFT; + vm = mm->total_vm + pgsz; + if (lim < vm) goto out; - buffer = kzalloc(size, GFP_KERNEL); - if (!buffer) + lim = rlim[RLIMIT_MEMLOCK].rlim_cur >> PAGE_SHIFT; + vm = mm->locked_vm + pgsz; + if (lim < vm) goto out; - current->mm->total_vm += pgsz; - current->mm->locked_vm += pgsz; + mm->total_vm += pgsz; + mm->locked_vm += pgsz; + error = 0; out: - up_write(¤t->mm->mmap_sem); - return buffer; + up_write(&mm->mmap_sem); + return error; } -void release_locked_buffer(void *buffer, size_t size) +void refund_locked_memory(struct mm_struct *mm, size_t size) { unsigned long pgsz = PAGE_ALIGN(size) >> PAGE_SHIFT; - down_write(¤t->mm->mmap_sem); - - current->mm->total_vm -= pgsz; - current->mm->locked_vm -= pgsz; - - up_write(¤t->mm->mmap_sem); -} + down_write(&mm->mmap_sem); -void free_locked_buffer(void *buffer, size_t size) -{ - release_locked_buffer(buffer, size); + mm->total_vm -= pgsz; + mm->locked_vm -= pgsz; - kfree(buffer); + up_write(&mm->mmap_sem); } diff --git a/mm/mmap.c b/mm/mmap.c index 6b7b1a95944..8101de490c7 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -28,6 +28,7 @@ #include <linux/mempolicy.h> #include <linux/rmap.h> #include <linux/mmu_notifier.h> +#include <linux/perf_counter.h> #include <asm/uaccess.h> #include <asm/cacheflush.h> @@ -1219,6 +1220,8 @@ munmap_back: if (correct_wcount) atomic_inc(&inode->i_writecount); out: + perf_counter_mmap(vma); + mm->total_vm += len >> PAGE_SHIFT; vm_stat_account(mm, vm_flags, file, len >> PAGE_SHIFT); if (vm_flags & VM_LOCKED) { @@ -2305,6 +2308,8 @@ int install_special_mapping(struct mm_struct *mm, mm->total_vm += len >> PAGE_SHIFT; + perf_counter_mmap(vma); + return 0; } diff --git a/mm/mprotect.c b/mm/mprotect.c index 258197b76fb..d80311baeb2 100644 --- a/mm/mprotect.c +++ b/mm/mprotect.c @@ -23,6 +23,7 @@ #include <linux/swapops.h> #include <linux/mmu_notifier.h> #include <linux/migrate.h> +#include <linux/perf_counter.h> #include <asm/uaccess.h> #include <asm/pgtable.h> #include <asm/cacheflush.h> @@ -299,6 +300,7 @@ SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len, error = mprotect_fixup(vma, &prev, nstart, tmp, newflags); if (error) goto out; + perf_counter_mmap(vma); nstart = tmp; if (nstart < prev->vm_end) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index fe753ecf2aa..474c7e9dd51 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -149,10 +149,6 @@ static unsigned long __meminitdata dma_reserve; static int __meminitdata nr_nodemap_entries; static unsigned long __meminitdata arch_zone_lowest_possible_pfn[MAX_NR_ZONES]; static unsigned long __meminitdata arch_zone_highest_possible_pfn[MAX_NR_ZONES]; -#ifdef CONFIG_MEMORY_HOTPLUG_RESERVE - static unsigned long __meminitdata node_boundary_start_pfn[MAX_NUMNODES]; - static unsigned long __meminitdata node_boundary_end_pfn[MAX_NUMNODES]; -#endif /* CONFIG_MEMORY_HOTPLUG_RESERVE */ static unsigned long __initdata required_kernelcore; static unsigned long __initdata required_movablecore; static unsigned long __meminitdata zone_movable_pfn[MAX_NUMNODES]; @@ -3103,64 +3099,6 @@ void __init sparse_memory_present_with_active_regions(int nid) } /** - * push_node_boundaries - Push node boundaries to at least the requested boundary - * @nid: The nid of the node to push the boundary for - * @start_pfn: The start pfn of the node - * @end_pfn: The end pfn of the node - * - * In reserve-based hot-add, mem_map is allocated that is unused until hotadd - * time. Specifically, on x86_64, SRAT will report ranges that can potentially - * be hotplugged even though no physical memory exists. This function allows - * an arch to push out the node boundaries so mem_map is allocated that can - * be used later. - */ -#ifdef CONFIG_MEMORY_HOTPLUG_RESERVE -void __init push_node_boundaries(unsigned int nid, - unsigned long start_pfn, unsigned long end_pfn) -{ - mminit_dprintk(MMINIT_TRACE, "zoneboundary", - "Entering push_node_boundaries(%u, %lu, %lu)\n", - nid, start_pfn, end_pfn); - - /* Initialise the boundary for this node if necessary */ - if (node_boundary_end_pfn[nid] == 0) - node_boundary_start_pfn[nid] = -1UL; - - /* Update the boundaries */ - if (node_boundary_start_pfn[nid] > start_pfn) - node_boundary_start_pfn[nid] = start_pfn; - if (node_boundary_end_pfn[nid] < end_pfn) - node_boundary_end_pfn[nid] = end_pfn; -} - -/* If necessary, push the node boundary out for reserve hotadd */ -static void __meminit account_node_boundary(unsigned int nid, - unsigned long *start_pfn, unsigned long *end_pfn) -{ - mminit_dprintk(MMINIT_TRACE, "zoneboundary", - "Entering account_node_boundary(%u, %lu, %lu)\n", - nid, *start_pfn, *end_pfn); - - /* Return if boundary information has not been provided */ - if (node_boundary_end_pfn[nid] == 0) - return; - - /* Check the boundaries and update if necessary */ - if (node_boundary_start_pfn[nid] < *start_pfn) - *start_pfn = node_boundary_start_pfn[nid]; - if (node_boundary_end_pfn[nid] > *end_pfn) - *end_pfn = node_boundary_end_pfn[nid]; -} -#else -void __init push_node_boundaries(unsigned int nid, - unsigned long start_pfn, unsigned long end_pfn) {} - -static void __meminit account_node_boundary(unsigned int nid, - unsigned long *start_pfn, unsigned long *end_pfn) {} -#endif - - -/** * get_pfn_range_for_nid - Return the start and end page frames for a node * @nid: The nid to return the range for. If MAX_NUMNODES, the min and max PFN are returned. * @start_pfn: Passed by reference. On return, it will have the node start_pfn. @@ -3185,9 +3123,6 @@ void __meminit get_pfn_range_for_nid(unsigned int nid, if (*start_pfn == -1UL) *start_pfn = 0; - - /* Push the node boundaries out if requested */ - account_node_boundary(nid, start_pfn, end_pfn); } /* @@ -3793,10 +3728,6 @@ void __init remove_all_active_ranges(void) { memset(early_node_map, 0, sizeof(early_node_map)); nr_nodemap_entries = 0; -#ifdef CONFIG_MEMORY_HOTPLUG_RESERVE - memset(node_boundary_start_pfn, 0, sizeof(node_boundary_start_pfn)); - memset(node_boundary_end_pfn, 0, sizeof(node_boundary_end_pfn)); -#endif /* CONFIG_MEMORY_HOTPLUG_RESERVE */ } /* Compare two active node_active_regions */ diff --git a/mm/percpu.c b/mm/percpu.c index 1aa5d8fbca1..c0b2c1a76e8 100644 --- a/mm/percpu.c +++ b/mm/percpu.c @@ -23,7 +23,7 @@ * Allocation is done in offset-size areas of single unit space. Ie, * an area of 512 bytes at 6k in c1 occupies 512 bytes at 6k of c1:u0, * c1:u1, c1:u2 and c1:u3. Percpu access can be done by configuring - * percpu base registers UNIT_SIZE apart. + * percpu base registers pcpu_unit_size apart. * * There are usually many small percpu allocations many of them as * small as 4 bytes. The allocator organizes chunks into lists @@ -38,8 +38,8 @@ * region and negative allocated. Allocation inside a chunk is done * by scanning this map sequentially and serving the first matching * entry. This is mostly copied from the percpu_modalloc() allocator. - * Chunks are also linked into a rb tree to ease address to chunk - * mapping during free. + * Chunks can be determined from the address using the index field + * in the page struct. The index field contains a pointer to the chunk. * * To use this allocator, arch code should do the followings. * @@ -61,7 +61,6 @@ #include <linux/mutex.h> #include <linux/percpu.h> #include <linux/pfn.h> -#include <linux/rbtree.h> #include <linux/slab.h> #include <linux/spinlock.h> #include <linux/vmalloc.h> @@ -88,7 +87,6 @@ struct pcpu_chunk { struct list_head list; /* linked to pcpu_slot lists */ - struct rb_node rb_node; /* key is chunk->vm->addr */ int free_size; /* free bytes in the chunk */ int contig_hint; /* max contiguous size hint */ struct vm_struct *vm; /* mapped vmalloc region */ @@ -110,9 +108,21 @@ static size_t pcpu_chunk_struct_size __read_mostly; void *pcpu_base_addr __read_mostly; EXPORT_SYMBOL_GPL(pcpu_base_addr); -/* optional reserved chunk, only accessible for reserved allocations */ +/* + * The first chunk which always exists. Note that unlike other + * chunks, this one can be allocated and mapped in several different + * ways and thus often doesn't live in the vmalloc area. + */ +static struct pcpu_chunk *pcpu_first_chunk; + +/* + * Optional reserved chunk. This chunk reserves part of the first + * chunk and serves it for reserved allocations. The amount of + * reserved offset is in pcpu_reserved_chunk_limit. When reserved + * area doesn't exist, the following variables contain NULL and 0 + * respectively. + */ static struct pcpu_chunk *pcpu_reserved_chunk; -/* offset limit of the reserved chunk */ static int pcpu_reserved_chunk_limit; /* @@ -121,7 +131,7 @@ static int pcpu_reserved_chunk_limit; * There are two locks - pcpu_alloc_mutex and pcpu_lock. The former * protects allocation/reclaim paths, chunks and chunk->page arrays. * The latter is a spinlock and protects the index data structures - - * chunk slots, rbtree, chunks and area maps in chunks. + * chunk slots, chunks and area maps in chunks. * * During allocation, pcpu_alloc_mutex is kept locked all the time and * pcpu_lock is grabbed and released as necessary. All actual memory @@ -140,7 +150,6 @@ static DEFINE_MUTEX(pcpu_alloc_mutex); /* protects whole alloc and reclaim */ static DEFINE_SPINLOCK(pcpu_lock); /* protects index data structures */ static struct list_head *pcpu_slot __read_mostly; /* chunk list slots */ -static struct rb_root pcpu_addr_root = RB_ROOT; /* chunks by address */ /* reclaim work to release fully free chunks, scheduled from free path */ static void pcpu_reclaim(struct work_struct *work); @@ -191,6 +200,18 @@ static bool pcpu_chunk_page_occupied(struct pcpu_chunk *chunk, return *pcpu_chunk_pagep(chunk, 0, page_idx) != NULL; } +/* set the pointer to a chunk in a page struct */ +static void pcpu_set_page_chunk(struct page *page, struct pcpu_chunk *pcpu) +{ + page->index = (unsigned long)pcpu; +} + +/* obtain pointer to a chunk from a page struct */ +static struct pcpu_chunk *pcpu_get_page_chunk(struct page *page) +{ + return (struct pcpu_chunk *)page->index; +} + /** * pcpu_mem_alloc - allocate memory * @size: bytes to allocate @@ -257,93 +278,26 @@ static void pcpu_chunk_relocate(struct pcpu_chunk *chunk, int oslot) } } -static struct rb_node **pcpu_chunk_rb_search(void *addr, - struct rb_node **parentp) -{ - struct rb_node **p = &pcpu_addr_root.rb_node; - struct rb_node *parent = NULL; - struct pcpu_chunk *chunk; - - while (*p) { - parent = *p; - chunk = rb_entry(parent, struct pcpu_chunk, rb_node); - - if (addr < chunk->vm->addr) - p = &(*p)->rb_left; - else if (addr > chunk->vm->addr) - p = &(*p)->rb_right; - else - break; - } - - if (parentp) - *parentp = parent; - return p; -} - /** - * pcpu_chunk_addr_search - search for chunk containing specified address - * @addr: address to search for - * - * Look for chunk which might contain @addr. More specifically, it - * searchs for the chunk with the highest start address which isn't - * beyond @addr. - * - * CONTEXT: - * pcpu_lock. + * pcpu_chunk_addr_search - determine chunk containing specified address + * @addr: address for which the chunk needs to be determined. * * RETURNS: * The address of the found chunk. */ static struct pcpu_chunk *pcpu_chunk_addr_search(void *addr) { - struct rb_node *n, *parent; - struct pcpu_chunk *chunk; + void *first_start = pcpu_first_chunk->vm->addr; - /* is it in the reserved chunk? */ - if (pcpu_reserved_chunk) { - void *start = pcpu_reserved_chunk->vm->addr; - - if (addr >= start && addr < start + pcpu_reserved_chunk_limit) + /* is it in the first chunk? */ + if (addr >= first_start && addr < first_start + pcpu_chunk_size) { + /* is it in the reserved area? */ + if (addr < first_start + pcpu_reserved_chunk_limit) return pcpu_reserved_chunk; + return pcpu_first_chunk; } - /* nah... search the regular ones */ - n = *pcpu_chunk_rb_search(addr, &parent); - if (!n) { - /* no exactly matching chunk, the parent is the closest */ - n = parent; - BUG_ON(!n); - } - chunk = rb_entry(n, struct pcpu_chunk, rb_node); - - if (addr < chunk->vm->addr) { - /* the parent was the next one, look for the previous one */ - n = rb_prev(n); - BUG_ON(!n); - chunk = rb_entry(n, struct pcpu_chunk, rb_node); - } - - return chunk; -} - -/** - * pcpu_chunk_addr_insert - insert chunk into address rb tree - * @new: chunk to insert - * - * Insert @new into address rb tree. - * - * CONTEXT: - * pcpu_lock. - */ -static void pcpu_chunk_addr_insert(struct pcpu_chunk *new) -{ - struct rb_node **p, *parent; - - p = pcpu_chunk_rb_search(new->vm->addr, &parent); - BUG_ON(*p); - rb_link_node(&new->rb_node, parent, p); - rb_insert_color(&new->rb_node, &pcpu_addr_root); + return pcpu_get_page_chunk(vmalloc_to_page(addr)); } /** @@ -755,6 +709,7 @@ static int pcpu_populate_chunk(struct pcpu_chunk *chunk, int off, int size) alloc_mask, 0); if (!*pagep) goto err; + pcpu_set_page_chunk(*pagep, chunk); } } @@ -879,7 +834,6 @@ restart: spin_lock_irq(&pcpu_lock); pcpu_chunk_relocate(chunk, -1); - pcpu_chunk_addr_insert(chunk); goto restart; area_found: @@ -968,7 +922,6 @@ static void pcpu_reclaim(struct work_struct *work) if (chunk == list_first_entry(head, struct pcpu_chunk, list)) continue; - rb_erase(&chunk->rb_node, &pcpu_addr_root); list_move(&chunk->list, &todo); } @@ -1147,7 +1100,8 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, if (reserved_size) { schunk->free_size = reserved_size; - pcpu_reserved_chunk = schunk; /* not for dynamic alloc */ + pcpu_reserved_chunk = schunk; + pcpu_reserved_chunk_limit = static_size + reserved_size; } else { schunk->free_size = dyn_size; dyn_size = 0; /* dynamic area covered */ @@ -1158,8 +1112,6 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, if (schunk->free_size) schunk->map[schunk->map_used++] = schunk->free_size; - pcpu_reserved_chunk_limit = static_size + schunk->free_size; - /* init dynamic chunk if necessary */ if (dyn_size) { dchunk = alloc_bootmem(sizeof(struct pcpu_chunk)); @@ -1226,13 +1178,8 @@ size_t __init pcpu_setup_first_chunk(pcpu_get_page_fn_t get_page_fn, } /* link the first chunk in */ - if (!dchunk) { - pcpu_chunk_relocate(schunk, -1); - pcpu_chunk_addr_insert(schunk); - } else { - pcpu_chunk_relocate(dchunk, -1); - pcpu_chunk_addr_insert(dchunk); - } + pcpu_first_chunk = dchunk ?: schunk; + pcpu_chunk_relocate(pcpu_first_chunk, -1); /* we're done */ pcpu_base_addr = (void *)pcpu_chunk_addr(schunk, 0, 0); diff --git a/mm/slab.c b/mm/slab.c index 9a90b00d2f9..f85831da908 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -102,7 +102,7 @@ #include <linux/cpu.h> #include <linux/sysctl.h> #include <linux/module.h> -#include <trace/kmemtrace.h> +#include <linux/kmemtrace.h> #include <linux/rcupdate.h> #include <linux/string.h> #include <linux/uaccess.h> diff --git a/mm/slob.c b/mm/slob.c index f92e66d558b..9b1737b0787 100644 --- a/mm/slob.c +++ b/mm/slob.c @@ -66,7 +66,7 @@ #include <linux/module.h> #include <linux/rcupdate.h> #include <linux/list.h> -#include <trace/kmemtrace.h> +#include <linux/kmemtrace.h> #include <asm/atomic.h> /* diff --git a/mm/slub.c b/mm/slub.c index 65ffda5934b..5e805a6fe36 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -17,7 +17,7 @@ #include <linux/slab.h> #include <linux/proc_fs.h> #include <linux/seq_file.h> -#include <trace/kmemtrace.h> +#include <linux/kmemtrace.h> #include <linux/cpu.h> #include <linux/cpuset.h> #include <linux/mempolicy.h> diff --git a/mm/util.c b/mm/util.c index 55bef160b9f..abc65aa7cdf 100644 --- a/mm/util.c +++ b/mm/util.c @@ -4,9 +4,11 @@ #include <linux/module.h> #include <linux/err.h> #include <linux/sched.h> -#include <linux/tracepoint.h> #include <asm/uaccess.h> +#define CREATE_TRACE_POINTS +#include <trace/events/kmem.h> + /** * kstrdup - allocate space for and copy an existing string * @s: the string to duplicate @@ -255,13 +257,6 @@ int __attribute__((weak)) get_user_pages_fast(unsigned long start, EXPORT_SYMBOL_GPL(get_user_pages_fast); /* Tracepoints definitions. */ -DEFINE_TRACE(kmalloc); -DEFINE_TRACE(kmem_cache_alloc); -DEFINE_TRACE(kmalloc_node); -DEFINE_TRACE(kmem_cache_alloc_node); -DEFINE_TRACE(kfree); -DEFINE_TRACE(kmem_cache_free); - EXPORT_TRACEPOINT_SYMBOL(kmalloc); EXPORT_TRACEPOINT_SYMBOL(kmem_cache_alloc); EXPORT_TRACEPOINT_SYMBOL(kmalloc_node); diff --git a/net/core/drop_monitor.c b/net/core/drop_monitor.c index 9fd0dc3cca9..b75b6cea49d 100644 --- a/net/core/drop_monitor.c +++ b/net/core/drop_monitor.c @@ -23,7 +23,7 @@ #include <linux/bitops.h> #include <net/genetlink.h> -#include <trace/skb.h> +#include <trace/events/skb.h> #include <asm/unaligned.h> diff --git a/net/core/net-traces.c b/net/core/net-traces.c index c8fb45665e4..499a67eaf3a 100644 --- a/net/core/net-traces.c +++ b/net/core/net-traces.c @@ -19,11 +19,11 @@ #include <linux/workqueue.h> #include <linux/netlink.h> #include <linux/net_dropmon.h> -#include <trace/skb.h> #include <asm/unaligned.h> #include <asm/bitops.h> +#define CREATE_TRACE_POINTS +#include <trace/events/skb.h> -DEFINE_TRACE(kfree_skb); EXPORT_TRACEPOINT_SYMBOL_GPL(kfree_skb); diff --git a/net/core/skbuff.c b/net/core/skbuff.c index e505b5392e1..c2e4fb8f354 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -65,7 +65,7 @@ #include <asm/uaccess.h> #include <asm/system.h> -#include <trace/skb.h> +#include <trace/events/skb.h> #include "kmap_skb.h" diff --git a/samples/Kconfig b/samples/Kconfig index 4b02f5a0e65..b75d28cba3f 100644 --- a/samples/Kconfig +++ b/samples/Kconfig @@ -19,6 +19,12 @@ config SAMPLE_TRACEPOINTS help This build tracepoints example modules. +config SAMPLE_TRACE_EVENTS + tristate "Build trace_events examples -- loadable modules only" + depends on EVENT_TRACING && m + help + This build trace event example modules. + config SAMPLE_KOBJECT tristate "Build kobject examples" help diff --git a/samples/Makefile b/samples/Makefile index 10eaca89fe1..13e4b470b53 100644 --- a/samples/Makefile +++ b/samples/Makefile @@ -1,3 +1,3 @@ # Makefile for Linux samples code -obj-$(CONFIG_SAMPLES) += markers/ kobject/ kprobes/ tracepoints/ +obj-$(CONFIG_SAMPLES) += markers/ kobject/ kprobes/ tracepoints/ trace_events/ diff --git a/samples/trace_events/Makefile b/samples/trace_events/Makefile new file mode 100644 index 00000000000..0d428dc6728 --- /dev/null +++ b/samples/trace_events/Makefile @@ -0,0 +1,6 @@ +# builds the trace events example kernel modules; +# then to use one (as root): insmod <module_name.ko> + +CFLAGS_trace-events-sample.o := -I$(src) + +obj-$(CONFIG_SAMPLE_TRACE_EVENTS) += trace-events-sample.o diff --git a/samples/trace_events/trace-events-sample.c b/samples/trace_events/trace-events-sample.c new file mode 100644 index 00000000000..aabc4e97091 --- /dev/null +++ b/samples/trace_events/trace-events-sample.c @@ -0,0 +1,52 @@ +#include <linux/module.h> +#include <linux/kthread.h> + +/* + * Any file that uses trace points, must include the header. + * But only one file, must include the header by defining + * CREATE_TRACE_POINTS first. This will make the C code that + * creates the handles for the trace points. + */ +#define CREATE_TRACE_POINTS +#include "trace-events-sample.h" + + +static void simple_thread_func(int cnt) +{ + set_current_state(TASK_INTERRUPTIBLE); + schedule_timeout(HZ); + trace_foo_bar("hello", cnt); +} + +static int simple_thread(void *arg) +{ + int cnt = 0; + + while (!kthread_should_stop()) + simple_thread_func(cnt++); + + return 0; +} + +static struct task_struct *simple_tsk; + +static int __init trace_event_init(void) +{ + simple_tsk = kthread_run(simple_thread, NULL, "event-sample"); + if (IS_ERR(simple_tsk)) + return -1; + + return 0; +} + +static void __exit trace_event_exit(void) +{ + kthread_stop(simple_tsk); +} + +module_init(trace_event_init); +module_exit(trace_event_exit); + +MODULE_AUTHOR("Steven Rostedt"); +MODULE_DESCRIPTION("trace-events-sample"); +MODULE_LICENSE("GPL"); diff --git a/samples/trace_events/trace-events-sample.h b/samples/trace_events/trace-events-sample.h new file mode 100644 index 00000000000..128a897687c --- /dev/null +++ b/samples/trace_events/trace-events-sample.h @@ -0,0 +1,129 @@ +/* + * Notice that this file is not protected like a normal header. + * We also must allow for rereading of this file. The + * + * || defined(TRACE_HEADER_MULTI_READ) + * + * serves this purpose. + */ +#if !defined(_TRACE_EVENT_SAMPLE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_EVENT_SAMPLE_H + +/* + * All trace headers should include tracepoint.h, until we finally + * make it into a standard header. + */ +#include <linux/tracepoint.h> + +/* + * If TRACE_SYSTEM is defined, that will be the directory created + * in the ftrace directory under /debugfs/tracing/events/<system> + * + * The define_trace.h belowe will also look for a file name of + * TRACE_SYSTEM.h where TRACE_SYSTEM is what is defined here. + * + * If you want a different system than file name, you can override + * the header name by defining TRACE_INCLUDE_FILE + * + * If this file was called, goofy.h, then we would define: + * + * #define TRACE_INCLUDE_FILE goofy + * + */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM sample + +/* + * The TRACE_EVENT macro is broken up into 5 parts. + * + * name: name of the trace point. This is also how to enable the tracepoint. + * A function called trace_foo_bar() will be created. + * + * proto: the prototype of the function trace_foo_bar() + * Here it is trace_foo_bar(char *foo, int bar). + * + * args: must match the arguments in the prototype. + * Here it is simply "foo, bar". + * + * struct: This defines the way the data will be stored in the ring buffer. + * There are currently two types of elements. __field and __array. + * a __field is broken up into (type, name). Where type can be any + * type but an array. + * For an array. there are three fields. (type, name, size). The + * type of elements in the array, the name of the field and the size + * of the array. + * + * __array( char, foo, 10) is the same as saying char foo[10]. + * + * fast_assign: This is a C like function that is used to store the items + * into the ring buffer. + * + * printk: This is a way to print out the data in pretty print. This is + * useful if the system crashes and you are logging via a serial line, + * the data can be printed to the console using this "printk" method. + * + * Note, that for both the assign and the printk, __entry is the handler + * to the data structure in the ring buffer, and is defined by the + * TP_STRUCT__entry. + */ +TRACE_EVENT(foo_bar, + + TP_PROTO(char *foo, int bar), + + TP_ARGS(foo, bar), + + TP_STRUCT__entry( + __array( char, foo, 10 ) + __field( int, bar ) + ), + + TP_fast_assign( + strncpy(__entry->foo, foo, 10); + __entry->bar = bar; + ), + + TP_printk("foo %s %d", __entry->foo, __entry->bar) +); +#endif + +/***** NOTICE! The #if protection ends here. *****/ + + +/* + * There are several ways I could have done this. If I left out the + * TRACE_INCLUDE_PATH, then it would default to the kernel source + * include/trace/events directory. + * + * I could specify a path from the define_trace.h file back to this + * file. + * + * #define TRACE_INCLUDE_PATH ../../samples/trace_events + * + * But I chose to simply make it use the current directory and then in + * the Makefile I added: + * + * CFLAGS_trace-events-sample.o := -I$(PWD)/samples/trace_events/ + * + * This will make sure the current path is part of the include + * structure for our file so that we can find it. + * + * I could have made only the top level directory the include: + * + * CFLAGS_trace-events-sample.o := -I$(PWD) + * + * And then let the path to this directory be the TRACE_INCLUDE_PATH: + * + * #define TRACE_INCLUDE_PATH samples/trace_events + * + * But then if something defines "samples" or "trace_events" then we + * could risk that being converted too, and give us an unexpected + * result. + */ +#undef TRACE_INCLUDE_PATH +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_PATH . +/* + * TRACE_INCLUDE_FILE is not needed if the filename and TRACE_SYSTEM are equal + */ +#define TRACE_INCLUDE_FILE trace-events-sample +#include <trace/define_trace.h> diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index cba61ca403c..2b706617c89 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -188,20 +188,34 @@ cmd_objcopy = $(OBJCOPY) $(OBJCOPYFLAGS) $(OBJCOPYFLAGS_$(@F)) $< $@ # --------------------------------------------------------------------------- quiet_cmd_gzip = GZIP $@ -cmd_gzip = gzip -f -9 < $< > $@ +cmd_gzip = (cat $(filter-out FORCE,$^) | gzip -f -9 > $@) || \ + (rm -f $@ ; false) # Bzip2 # --------------------------------------------------------------------------- -# Bzip2 does not include size in file... so we have to fake that -size_append=$(CONFIG_SHELL) $(srctree)/scripts/bin_size - -quiet_cmd_bzip2 = BZIP2 $@ -cmd_bzip2 = (bzip2 -9 < $< && $(size_append) $<) > $@ || (rm -f $@ ; false) +# Bzip2 and LZMA do not include size in file... so we have to fake that; +# append the size as a 32-bit littleendian number as gzip does. +size_append = echo -ne $(shell \ +dec_size=0; \ +for F in $1; do \ + fsize=$$(stat -c "%s" $$F); \ + dec_size=$$(expr $$dec_size + $$fsize); \ +done; \ +printf "%08x" $$dec_size | \ + sed 's/\(..\)\(..\)\(..\)\(..\)/\\\\x\4\\\\x\3\\\\x\2\\\\x\1/g' \ +) + +quiet_cmd_bzip2 = BZIP2 $@ +cmd_bzip2 = (cat $(filter-out FORCE,$^) | \ + bzip2 -9 && $(call size_append, $(filter-out FORCE,$^))) > $@ || \ + (rm -f $@ ; false) # Lzma # --------------------------------------------------------------------------- quiet_cmd_lzma = LZMA $@ -cmd_lzma = (lzma -9 -c $< && $(size_append) $<) >$@ || (rm -f $@ ; false) +cmd_lzma = (cat $(filter-out FORCE,$^) | \ + lzma -9 && $(call size_append, $(filter-out FORCE,$^))) > $@ || \ + (rm -f $@ ; false) diff --git a/scripts/bin_size b/scripts/bin_size deleted file mode 100644 index 43e1b360cee..00000000000 --- a/scripts/bin_size +++ /dev/null @@ -1,10 +0,0 @@ -#!/bin/sh - -if [ $# = 0 ] ; then - echo Usage: $0 file -fi - -size_dec=`stat -c "%s" $1` -size_hex_echo_string=`printf "%08x" $size_dec | - sed 's/\(..\)\(..\)\(..\)\(..\)/\\\\x\4\\\\x\3\\\\x\2\\\\x\1/g'` -/bin/echo -ne $size_hex_echo_string diff --git a/scripts/kernel-doc b/scripts/kernel-doc index 3208a3a7e7f..acd8c4a8e3e 100755 --- a/scripts/kernel-doc +++ b/scripts/kernel-doc @@ -1828,6 +1828,25 @@ sub reset_state { $state = 0; } +sub tracepoint_munge($) { + my $file = shift; + my $tracepointname = 0; + my $tracepointargs = 0; + + if($prototype =~ m/TRACE_EVENT\((.*?),/) { + $tracepointname = $1; + } + if($prototype =~ m/TP_PROTO\((.*?)\)/) { + $tracepointargs = $1; + } + if (($tracepointname eq 0) || ($tracepointargs eq 0)) { + print STDERR "Warning(${file}:$.): Unrecognized tracepoint format: \n". + "$prototype\n"; + } else { + $prototype = "static inline void trace_$tracepointname($tracepointargs)"; + } +} + sub syscall_munge() { my $void = 0; @@ -1882,6 +1901,9 @@ sub process_state3_function($$) { if ($prototype =~ /SYSCALL_DEFINE/) { syscall_munge(); } + if ($prototype =~ /TRACE_EVENT/) { + tracepoint_munge($file); + } dump_function($prototype, $file); reset_state(); } diff --git a/scripts/recordmcount.pl b/scripts/recordmcount.pl index 409596eca12..0fae7da0529 100755 --- a/scripts/recordmcount.pl +++ b/scripts/recordmcount.pl @@ -26,7 +26,7 @@ # which will also be the location of that section after final link. # e.g. # -# .section ".text.sched" +# .section ".sched.text", "ax" # .globl my_func # my_func: # [...] @@ -39,7 +39,7 @@ # [...] # # Both relocation offsets for the mcounts in the above example will be -# offset from .text.sched. If we make another file called tmp.s with: +# offset from .sched.text. If we make another file called tmp.s with: # # .section __mcount_loc # .quad my_func + 0x5 @@ -51,7 +51,7 @@ # But this gets hard if my_func is not globl (a static function). # In such a case we have: # -# .section ".text.sched" +# .section ".sched.text", "ax" # my_func: # [...] # call mcount (offset: 0x5) diff --git a/security/integrity/ima/ima_fs.c b/security/integrity/ima/ima_fs.c index ffbe259700b..510186f0b72 100644 --- a/security/integrity/ima/ima_fs.c +++ b/security/integrity/ima/ima_fs.c @@ -84,8 +84,8 @@ static void *ima_measurements_next(struct seq_file *m, void *v, loff_t *pos) * against concurrent list-extension */ rcu_read_lock(); - qe = list_entry(rcu_dereference(qe->later.next), - struct ima_queue_entry, later); + qe = list_entry_rcu(qe->later.next, + struct ima_queue_entry, later); rcu_read_unlock(); (*pos)++; diff --git a/security/smack/smackfs.c b/security/smack/smackfs.c index e03a7e19c73..11d2cb19d7a 100644 --- a/security/smack/smackfs.c +++ b/security/smack/smackfs.c @@ -734,8 +734,8 @@ static void smk_netlbladdr_insert(struct smk_netlbladdr *new) return; } - m = list_entry(rcu_dereference(smk_netlbladdr_list.next), - struct smk_netlbladdr, list); + m = list_entry_rcu(smk_netlbladdr_list.next, + struct smk_netlbladdr, list); /* the comparison '>' is a bit hacky, but works */ if (new->smk_mask.s_addr > m->smk_mask.s_addr) { @@ -748,8 +748,8 @@ static void smk_netlbladdr_insert(struct smk_netlbladdr *new) list_add_rcu(&new->list, &m->list); return; } - m_next = list_entry(rcu_dereference(m->list.next), - struct smk_netlbladdr, list); + m_next = list_entry_rcu(m->list.next, + struct smk_netlbladdr, list); if (new->smk_mask.s_addr > m_next->smk_mask.s_addr) { list_add_rcu(&new->list, &m->list); return; diff --git a/tools/perf/.gitignore b/tools/perf/.gitignore new file mode 100644 index 00000000000..d69a759a104 --- /dev/null +++ b/tools/perf/.gitignore @@ -0,0 +1,16 @@ +PERF-BUILD-OPTIONS +PERF-CFLAGS +PERF-GUI-VARS +PERF-VERSION-FILE +perf +perf-help +perf-record +perf-report +perf-stat +perf-top +perf*.1 +perf*.xml +common-cmds.h +tags +TAGS +cscope* diff --git a/tools/perf/Documentation/Makefile b/tools/perf/Documentation/Makefile new file mode 100644 index 00000000000..5457192e1b4 --- /dev/null +++ b/tools/perf/Documentation/Makefile @@ -0,0 +1,300 @@ +MAN1_TXT= \ + $(filter-out $(addsuffix .txt, $(ARTICLES) $(SP_ARTICLES)), \ + $(wildcard perf-*.txt)) \ + perf.txt +MAN5_TXT= +MAN7_TXT= + +MAN_TXT = $(MAN1_TXT) $(MAN5_TXT) $(MAN7_TXT) +MAN_XML=$(patsubst %.txt,%.xml,$(MAN_TXT)) +MAN_HTML=$(patsubst %.txt,%.html,$(MAN_TXT)) + +DOC_HTML=$(MAN_HTML) + +ARTICLES = +# with their own formatting rules. +SP_ARTICLES = +API_DOCS = $(patsubst %.txt,%,$(filter-out technical/api-index-skel.txt technical/api-index.txt, $(wildcard technical/api-*.txt))) +SP_ARTICLES += $(API_DOCS) +SP_ARTICLES += technical/api-index + +DOC_HTML += $(patsubst %,%.html,$(ARTICLES) $(SP_ARTICLES)) + +DOC_MAN1=$(patsubst %.txt,%.1,$(MAN1_TXT)) +DOC_MAN5=$(patsubst %.txt,%.5,$(MAN5_TXT)) +DOC_MAN7=$(patsubst %.txt,%.7,$(MAN7_TXT)) + +prefix?=$(HOME) +bindir?=$(prefix)/bin +htmldir?=$(prefix)/share/doc/perf-doc +pdfdir?=$(prefix)/share/doc/perf-doc +mandir?=$(prefix)/share/man +man1dir=$(mandir)/man1 +man5dir=$(mandir)/man5 +man7dir=$(mandir)/man7 +# DESTDIR= + +ASCIIDOC=asciidoc +ASCIIDOC_EXTRA = +MANPAGE_XSL = manpage-normal.xsl +XMLTO_EXTRA = +INSTALL?=install +RM ?= rm -f +DOC_REF = origin/man +HTML_REF = origin/html + +infodir?=$(prefix)/share/info +MAKEINFO=makeinfo +INSTALL_INFO=install-info +DOCBOOK2X_TEXI=docbook2x-texi +DBLATEX=dblatex +ifndef PERL_PATH + PERL_PATH = /usr/bin/perl +endif + +-include ../config.mak.autogen +-include ../config.mak + +# +# For asciidoc ... +# -7.1.2, no extra settings are needed. +# 8.0-, set ASCIIDOC8. +# + +# +# For docbook-xsl ... +# -1.68.1, set ASCIIDOC_NO_ROFF? (based on changelog from 1.73.0) +# 1.69.0, no extra settings are needed? +# 1.69.1-1.71.0, set DOCBOOK_SUPPRESS_SP? +# 1.71.1, no extra settings are needed? +# 1.72.0, set DOCBOOK_XSL_172. +# 1.73.0-, set ASCIIDOC_NO_ROFF +# + +# +# If you had been using DOCBOOK_XSL_172 in an attempt to get rid +# of 'the ".ft C" problem' in your generated manpages, and you +# instead ended up with weird characters around callouts, try +# using ASCIIDOC_NO_ROFF instead (it works fine with ASCIIDOC8). +# + +ifdef ASCIIDOC8 +ASCIIDOC_EXTRA += -a asciidoc7compatible +endif +ifdef DOCBOOK_XSL_172 +ASCIIDOC_EXTRA += -a perf-asciidoc-no-roff +MANPAGE_XSL = manpage-1.72.xsl +else + ifdef ASCIIDOC_NO_ROFF + # docbook-xsl after 1.72 needs the regular XSL, but will not + # pass-thru raw roff codes from asciidoc.conf, so turn them off. + ASCIIDOC_EXTRA += -a perf-asciidoc-no-roff + endif +endif +ifdef MAN_BOLD_LITERAL +XMLTO_EXTRA += -m manpage-bold-literal.xsl +endif +ifdef DOCBOOK_SUPPRESS_SP +XMLTO_EXTRA += -m manpage-suppress-sp.xsl +endif + +SHELL_PATH ?= $(SHELL) +# Shell quote; +SHELL_PATH_SQ = $(subst ','\'',$(SHELL_PATH)) + +# +# Please note that there is a minor bug in asciidoc. +# The version after 6.0.3 _will_ include the patch found here: +# http://marc.theaimsgroup.com/?l=perf&m=111558757202243&w=2 +# +# Until that version is released you may have to apply the patch +# yourself - yes, all 6 characters of it! +# + +QUIET_SUBDIR0 = +$(MAKE) -C # space to separate -C and subdir +QUIET_SUBDIR1 = + +ifneq ($(findstring $(MAKEFLAGS),w),w) +PRINT_DIR = --no-print-directory +else # "make -w" +NO_SUBDIR = : +endif + +ifneq ($(findstring $(MAKEFLAGS),s),s) +ifndef V + QUIET_ASCIIDOC = @echo ' ' ASCIIDOC $@; + QUIET_XMLTO = @echo ' ' XMLTO $@; + QUIET_DB2TEXI = @echo ' ' DB2TEXI $@; + QUIET_MAKEINFO = @echo ' ' MAKEINFO $@; + QUIET_DBLATEX = @echo ' ' DBLATEX $@; + QUIET_XSLTPROC = @echo ' ' XSLTPROC $@; + QUIET_GEN = @echo ' ' GEN $@; + QUIET_STDERR = 2> /dev/null + QUIET_SUBDIR0 = +@subdir= + QUIET_SUBDIR1 = ;$(NO_SUBDIR) echo ' ' SUBDIR $$subdir; \ + $(MAKE) $(PRINT_DIR) -C $$subdir + export V +endif +endif + +all: html man + +html: $(DOC_HTML) + +$(DOC_HTML) $(DOC_MAN1) $(DOC_MAN5) $(DOC_MAN7): asciidoc.conf + +man: man1 man5 man7 +man1: $(DOC_MAN1) +man5: $(DOC_MAN5) +man7: $(DOC_MAN7) + +info: perf.info perfman.info + +pdf: user-manual.pdf + +install: install-man + +install-man: man + $(INSTALL) -d -m 755 $(DESTDIR)$(man1dir) +# $(INSTALL) -d -m 755 $(DESTDIR)$(man5dir) +# $(INSTALL) -d -m 755 $(DESTDIR)$(man7dir) + $(INSTALL) -m 644 $(DOC_MAN1) $(DESTDIR)$(man1dir) +# $(INSTALL) -m 644 $(DOC_MAN5) $(DESTDIR)$(man5dir) +# $(INSTALL) -m 644 $(DOC_MAN7) $(DESTDIR)$(man7dir) + +install-info: info + $(INSTALL) -d -m 755 $(DESTDIR)$(infodir) + $(INSTALL) -m 644 perf.info perfman.info $(DESTDIR)$(infodir) + if test -r $(DESTDIR)$(infodir)/dir; then \ + $(INSTALL_INFO) --info-dir=$(DESTDIR)$(infodir) perf.info ;\ + $(INSTALL_INFO) --info-dir=$(DESTDIR)$(infodir) perfman.info ;\ + else \ + echo "No directory found in $(DESTDIR)$(infodir)" >&2 ; \ + fi + +install-pdf: pdf + $(INSTALL) -d -m 755 $(DESTDIR)$(pdfdir) + $(INSTALL) -m 644 user-manual.pdf $(DESTDIR)$(pdfdir) + +install-html: html + '$(SHELL_PATH_SQ)' ./install-webdoc.sh $(DESTDIR)$(htmldir) + +../PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE + $(QUIET_SUBDIR0)../ $(QUIET_SUBDIR1) PERF-VERSION-FILE + +-include ../PERF-VERSION-FILE + +# +# Determine "include::" file references in asciidoc files. +# +doc.dep : $(wildcard *.txt) build-docdep.perl + $(QUIET_GEN)$(RM) $@+ $@ && \ + $(PERL_PATH) ./build-docdep.perl >$@+ $(QUIET_STDERR) && \ + mv $@+ $@ + +-include doc.dep + +cmds_txt = cmds-ancillaryinterrogators.txt \ + cmds-ancillarymanipulators.txt \ + cmds-mainporcelain.txt \ + cmds-plumbinginterrogators.txt \ + cmds-plumbingmanipulators.txt \ + cmds-synchingrepositories.txt \ + cmds-synchelpers.txt \ + cmds-purehelpers.txt \ + cmds-foreignscminterface.txt + +$(cmds_txt): cmd-list.made + +cmd-list.made: cmd-list.perl ../command-list.txt $(MAN1_TXT) + $(QUIET_GEN)$(RM) $@ && \ + $(PERL_PATH) ./cmd-list.perl ../command-list.txt $(QUIET_STDERR) && \ + date >$@ + +clean: + $(RM) *.xml *.xml+ *.html *.html+ *.1 *.5 *.7 + $(RM) *.texi *.texi+ *.texi++ perf.info perfman.info + $(RM) howto-index.txt howto/*.html doc.dep + $(RM) technical/api-*.html technical/api-index.txt + $(RM) $(cmds_txt) *.made + +$(MAN_HTML): %.html : %.txt + $(QUIET_ASCIIDOC)$(RM) $@+ $@ && \ + $(ASCIIDOC) -b xhtml11 -d manpage -f asciidoc.conf \ + $(ASCIIDOC_EXTRA) -aperf_version=$(PERF_VERSION) -o $@+ $< && \ + mv $@+ $@ + +%.1 %.5 %.7 : %.xml + $(QUIET_XMLTO)$(RM) $@ && \ + xmlto -m $(MANPAGE_XSL) $(XMLTO_EXTRA) man $< + +%.xml : %.txt + $(QUIET_ASCIIDOC)$(RM) $@+ $@ && \ + $(ASCIIDOC) -b docbook -d manpage -f asciidoc.conf \ + $(ASCIIDOC_EXTRA) -aperf_version=$(PERF_VERSION) -o $@+ $< && \ + mv $@+ $@ + +XSLT = docbook.xsl +XSLTOPTS = --xinclude --stringparam html.stylesheet docbook-xsl.css + +user-manual.html: user-manual.xml + $(QUIET_XSLTPROC)xsltproc $(XSLTOPTS) -o $@ $(XSLT) $< + +perf.info: user-manual.texi + $(QUIET_MAKEINFO)$(MAKEINFO) --no-split -o $@ user-manual.texi + +user-manual.texi: user-manual.xml + $(QUIET_DB2TEXI)$(RM) $@+ $@ && \ + $(DOCBOOK2X_TEXI) user-manual.xml --encoding=UTF-8 --to-stdout >$@++ && \ + $(PERL_PATH) fix-texi.perl <$@++ >$@+ && \ + rm $@++ && \ + mv $@+ $@ + +user-manual.pdf: user-manual.xml + $(QUIET_DBLATEX)$(RM) $@+ $@ && \ + $(DBLATEX) -o $@+ -p /etc/asciidoc/dblatex/asciidoc-dblatex.xsl -s /etc/asciidoc/dblatex/asciidoc-dblatex.sty $< && \ + mv $@+ $@ + +perfman.texi: $(MAN_XML) cat-texi.perl + $(QUIET_DB2TEXI)$(RM) $@+ $@ && \ + ($(foreach xml,$(MAN_XML),$(DOCBOOK2X_TEXI) --encoding=UTF-8 \ + --to-stdout $(xml) &&) true) > $@++ && \ + $(PERL_PATH) cat-texi.perl $@ <$@++ >$@+ && \ + rm $@++ && \ + mv $@+ $@ + +perfman.info: perfman.texi + $(QUIET_MAKEINFO)$(MAKEINFO) --no-split --no-validate $*.texi + +$(patsubst %.txt,%.texi,$(MAN_TXT)): %.texi : %.xml + $(QUIET_DB2TEXI)$(RM) $@+ $@ && \ + $(DOCBOOK2X_TEXI) --to-stdout $*.xml >$@+ && \ + mv $@+ $@ + +howto-index.txt: howto-index.sh $(wildcard howto/*.txt) + $(QUIET_GEN)$(RM) $@+ $@ && \ + '$(SHELL_PATH_SQ)' ./howto-index.sh $(wildcard howto/*.txt) >$@+ && \ + mv $@+ $@ + +$(patsubst %,%.html,$(ARTICLES)) : %.html : %.txt + $(QUIET_ASCIIDOC)$(ASCIIDOC) -b xhtml11 $*.txt + +WEBDOC_DEST = /pub/software/tools/perf/docs + +$(patsubst %.txt,%.html,$(wildcard howto/*.txt)): %.html : %.txt + $(QUIET_ASCIIDOC)$(RM) $@+ $@ && \ + sed -e '1,/^$$/d' $< | $(ASCIIDOC) -b xhtml11 - >$@+ && \ + mv $@+ $@ + +install-webdoc : html + '$(SHELL_PATH_SQ)' ./install-webdoc.sh $(WEBDOC_DEST) + +quick-install: quick-install-man + +quick-install-man: + '$(SHELL_PATH_SQ)' ./install-doc-quick.sh $(DOC_REF) $(DESTDIR)$(mandir) + +quick-install-html: + '$(SHELL_PATH_SQ)' ./install-doc-quick.sh $(HTML_REF) $(DESTDIR)$(htmldir) + +.PHONY: .FORCE-PERF-VERSION-FILE diff --git a/tools/perf/Documentation/asciidoc.conf b/tools/perf/Documentation/asciidoc.conf new file mode 100644 index 00000000000..356b23a4033 --- /dev/null +++ b/tools/perf/Documentation/asciidoc.conf @@ -0,0 +1,91 @@ +## linkperf: macro +# +# Usage: linkperf:command[manpage-section] +# +# Note, {0} is the manpage section, while {target} is the command. +# +# Show PERF link as: <command>(<section>); if section is defined, else just show +# the command. + +[macros] +(?su)[\\]?(?P<name>linkperf):(?P<target>\S*?)\[(?P<attrlist>.*?)\]= + +[attributes] +asterisk=* +plus=+ +caret=^ +startsb=[ +endsb=] +tilde=~ + +ifdef::backend-docbook[] +[linkperf-inlinemacro] +{0%{target}} +{0#<citerefentry>} +{0#<refentrytitle>{target}</refentrytitle><manvolnum>{0}</manvolnum>} +{0#</citerefentry>} +endif::backend-docbook[] + +ifdef::backend-docbook[] +ifndef::perf-asciidoc-no-roff[] +# "unbreak" docbook-xsl v1.68 for manpages. v1.69 works with or without this. +# v1.72 breaks with this because it replaces dots not in roff requests. +[listingblock] +<example><title>{title}</title> +<literallayout> +ifdef::doctype-manpage[] + .ft C +endif::doctype-manpage[] +| +ifdef::doctype-manpage[] + .ft +endif::doctype-manpage[] +</literallayout> +{title#}</example> +endif::perf-asciidoc-no-roff[] + +ifdef::perf-asciidoc-no-roff[] +ifdef::doctype-manpage[] +# The following two small workarounds insert a simple paragraph after screen +[listingblock] +<example><title>{title}</title> +<literallayout> +| +</literallayout><simpara></simpara> +{title#}</example> + +[verseblock] +<formalpara{id? id="{id}"}><title>{title}</title><para> +{title%}<literallayout{id? id="{id}"}> +{title#}<literallayout> +| +</literallayout> +{title#}</para></formalpara> +{title%}<simpara></simpara> +endif::doctype-manpage[] +endif::perf-asciidoc-no-roff[] +endif::backend-docbook[] + +ifdef::doctype-manpage[] +ifdef::backend-docbook[] +[header] +template::[header-declarations] +<refentry> +<refmeta> +<refentrytitle>{mantitle}</refentrytitle> +<manvolnum>{manvolnum}</manvolnum> +<refmiscinfo class="source">perf</refmiscinfo> +<refmiscinfo class="version">{perf_version}</refmiscinfo> +<refmiscinfo class="manual">perf Manual</refmiscinfo> +</refmeta> +<refnamediv> + <refname>{manname}</refname> + <refpurpose>{manpurpose}</refpurpose> +</refnamediv> +endif::backend-docbook[] +endif::doctype-manpage[] + +ifdef::backend-xhtml11[] +[linkperf-inlinemacro] +<a href="{target}.html">{target}{0?({0})}</a> +endif::backend-xhtml11[] diff --git a/tools/perf/Documentation/manpage-1.72.xsl b/tools/perf/Documentation/manpage-1.72.xsl new file mode 100644 index 00000000000..b4d315cb8c4 --- /dev/null +++ b/tools/perf/Documentation/manpage-1.72.xsl @@ -0,0 +1,14 @@ +<!-- manpage-1.72.xsl: + special settings for manpages rendered from asciidoc+docbook + handles peculiarities in docbook-xsl 1.72.0 --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<xsl:import href="manpage-base.xsl"/> + +<!-- these are the special values for the roff control characters + needed for docbook-xsl 1.72.0 --> +<xsl:param name="git.docbook.backslash">▓</xsl:param> +<xsl:param name="git.docbook.dot" >⌂</xsl:param> + +</xsl:stylesheet> diff --git a/tools/perf/Documentation/manpage-base.xsl b/tools/perf/Documentation/manpage-base.xsl new file mode 100644 index 00000000000..a264fa61609 --- /dev/null +++ b/tools/perf/Documentation/manpage-base.xsl @@ -0,0 +1,35 @@ +<!-- manpage-base.xsl: + special formatting for manpages rendered from asciidoc+docbook --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- these params silence some output from xmlto --> +<xsl:param name="man.output.quietly" select="1"/> +<xsl:param name="refentry.meta.get.quietly" select="1"/> + +<!-- convert asciidoc callouts to man page format; + git.docbook.backslash and git.docbook.dot params + must be supplied by another XSL file or other means --> +<xsl:template match="co"> + <xsl:value-of select="concat( + $git.docbook.backslash,'fB(', + substring-after(@id,'-'),')', + $git.docbook.backslash,'fR')"/> +</xsl:template> +<xsl:template match="calloutlist"> + <xsl:value-of select="$git.docbook.dot"/> + <xsl:text>sp </xsl:text> + <xsl:apply-templates/> + <xsl:text> </xsl:text> +</xsl:template> +<xsl:template match="callout"> + <xsl:value-of select="concat( + $git.docbook.backslash,'fB', + substring-after(@arearefs,'-'), + '. ',$git.docbook.backslash,'fR')"/> + <xsl:apply-templates/> + <xsl:value-of select="$git.docbook.dot"/> + <xsl:text>br </xsl:text> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/perf/Documentation/manpage-bold-literal.xsl b/tools/perf/Documentation/manpage-bold-literal.xsl new file mode 100644 index 00000000000..608eb5df628 --- /dev/null +++ b/tools/perf/Documentation/manpage-bold-literal.xsl @@ -0,0 +1,17 @@ +<!-- manpage-bold-literal.xsl: + special formatting for manpages rendered from asciidoc+docbook --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- render literal text as bold (instead of plain or monospace); + this makes literal text easier to distinguish in manpages + viewed on a tty --> +<xsl:template match="literal"> + <xsl:value-of select="$git.docbook.backslash"/> + <xsl:text>fB</xsl:text> + <xsl:apply-templates/> + <xsl:value-of select="$git.docbook.backslash"/> + <xsl:text>fR</xsl:text> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/perf/Documentation/manpage-normal.xsl b/tools/perf/Documentation/manpage-normal.xsl new file mode 100644 index 00000000000..a48f5b11f3d --- /dev/null +++ b/tools/perf/Documentation/manpage-normal.xsl @@ -0,0 +1,13 @@ +<!-- manpage-normal.xsl: + special settings for manpages rendered from asciidoc+docbook + handles anything we want to keep away from docbook-xsl 1.72.0 --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<xsl:import href="manpage-base.xsl"/> + +<!-- these are the normal values for the roff control characters --> +<xsl:param name="git.docbook.backslash">\</xsl:param> +<xsl:param name="git.docbook.dot" >.</xsl:param> + +</xsl:stylesheet> diff --git a/tools/perf/Documentation/manpage-suppress-sp.xsl b/tools/perf/Documentation/manpage-suppress-sp.xsl new file mode 100644 index 00000000000..a63c7632a87 --- /dev/null +++ b/tools/perf/Documentation/manpage-suppress-sp.xsl @@ -0,0 +1,21 @@ +<!-- manpage-suppress-sp.xsl: + special settings for manpages rendered from asciidoc+docbook + handles erroneous, inline .sp in manpage output of some + versions of docbook-xsl --> +<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" + version="1.0"> + +<!-- attempt to work around spurious .sp at the tail of the line + that some versions of docbook stylesheets seem to add --> +<xsl:template match="simpara"> + <xsl:variable name="content"> + <xsl:apply-templates/> + </xsl:variable> + <xsl:value-of select="normalize-space($content)"/> + <xsl:if test="not(ancestor::authorblurb) and + not(ancestor::personblurb)"> + <xsl:text> </xsl:text> + </xsl:if> +</xsl:template> + +</xsl:stylesheet> diff --git a/tools/perf/Documentation/perf-annotate.txt b/tools/perf/Documentation/perf-annotate.txt new file mode 100644 index 00000000000..c9dcade0683 --- /dev/null +++ b/tools/perf/Documentation/perf-annotate.txt @@ -0,0 +1,29 @@ +perf-annotate(1) +============== + +NAME +---- +perf-annotate - Read perf.data (created by perf record) and display annotated code + +SYNOPSIS +-------- +[verse] +'perf annotate' [-i <file> | --input=file] symbol_name + +DESCRIPTION +----------- +This command reads the input file and displays an annotated version of the +code. If the object file has debug symbols then the source code will be +displayed alongside assembly code. + +If there is no debug info in the object, then annotated assembly is displayed. + +OPTIONS +------- +-i:: +--input=:: + Input file name. (default: perf.data) + +SEE ALSO +-------- +linkperf:perf-record[1] diff --git a/tools/perf/Documentation/perf-help.txt b/tools/perf/Documentation/perf-help.txt new file mode 100644 index 00000000000..514391818d1 --- /dev/null +++ b/tools/perf/Documentation/perf-help.txt @@ -0,0 +1,38 @@ +perf-help(1) +============ + +NAME +---- +perf-help - display help information about perf + +SYNOPSIS +-------- +'perf help' [-a|--all] [COMMAND] + +DESCRIPTION +----------- + +With no options and no COMMAND given, the synopsis of the 'perf' +command and a list of the most commonly used perf commands are printed +on the standard output. + +If the option '--all' or '-a' is given, then all available commands are +printed on the standard output. + +If a perf command is named, a manual page for that command is brought +up. The 'man' program is used by default for this purpose, but this +can be overridden by other options or configuration variables. + +Note that `perf --help ...` is identical to `perf help ...` because the +former is internally converted into the latter. + +OPTIONS +------- +-a:: +--all:: + Prints all the available commands on the standard output. This + option supersedes any other option. + +PERF +---- +Part of the linkperf:perf[1] suite diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt new file mode 100644 index 00000000000..8290b942266 --- /dev/null +++ b/tools/perf/Documentation/perf-list.txt @@ -0,0 +1,25 @@ +perf-list(1) +============ + +NAME +---- +perf-list - List all symbolic event types + +SYNOPSIS +-------- +[verse] +'perf list' + +DESCRIPTION +----------- +This command displays the symbolic event types which can be selected in the +various perf commands with the -e option. + +OPTIONS +------- +None + +SEE ALSO +-------- +linkperf:perf-stat[1], linkperf:perf-top[1], +linkperf:perf-record[1] diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt new file mode 100644 index 00000000000..1dbc1eeb4c0 --- /dev/null +++ b/tools/perf/Documentation/perf-record.txt @@ -0,0 +1,42 @@ +perf-record(1) +============== + +NAME +---- +perf-record - Run a command and record its profile into perf.data + +SYNOPSIS +-------- +[verse] +'perf record' [-e <EVENT> | --event=EVENT] [-l] [-a] <command> +'perf record' [-e <EVENT> | --event=EVENT] [-l] [-a] -- <command> [<options>] + +DESCRIPTION +----------- +This command runs a command and gathers a performance counter profile +from it, into perf.data - without displaying anything. + +This file can then be inspected later on, using 'perf report'. + + +OPTIONS +------- +<command>...:: + Any command you can specify in a shell. + +-e:: +--event=:: + Select the PMU event. Selection can be a symbolic event name + (use 'perf list' to list all events) or a raw PMU + event (eventsel+umask) in the form of rNNN where NNN is a + hexadecimal event descriptor. + +-a:: + system-wide collection + +-l:: + scale counter values + +SEE ALSO +-------- +linkperf:perf-stat[1], linkperf:perf-list[1] diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt new file mode 100644 index 00000000000..52d3fc6846a --- /dev/null +++ b/tools/perf/Documentation/perf-report.txt @@ -0,0 +1,26 @@ +perf-report(1) +============== + +NAME +---- +perf-report - Read perf.data (created by perf record) and display the profile + +SYNOPSIS +-------- +[verse] +'perf report' [-i <file> | --input=file] + +DESCRIPTION +----------- +This command displays the performance counter profile information recorded +via perf report. + +OPTIONS +------- +-i:: +--input=:: + Input file name. (default: perf.data) + +SEE ALSO +-------- +linkperf:perf-stat[1] diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt new file mode 100644 index 00000000000..c368a72721d --- /dev/null +++ b/tools/perf/Documentation/perf-stat.txt @@ -0,0 +1,66 @@ +perf-stat(1) +============ + +NAME +---- +perf-stat - Run a command and gather performance counter statistics + +SYNOPSIS +-------- +[verse] +'perf stat' [-e <EVENT> | --event=EVENT] [-l] [-a] <command> +'perf stat' [-e <EVENT> | --event=EVENT] [-l] [-a] -- <command> [<options>] + +DESCRIPTION +----------- +This command runs a command and gathers performance counter statistics +from it. + + +OPTIONS +------- +<command>...:: + Any command you can specify in a shell. + + +-e:: +--event=:: + Select the PMU event. Selection can be a symbolic event name + (use 'perf list' to list all events) or a raw PMU + event (eventsel+umask) in the form of rNNN where NNN is a + hexadecimal event descriptor. + +-i:: +--inherit:: + child tasks inherit counters +-p:: +--pid=<pid>:: + stat events on existing pid + +-a:: + system-wide collection + +-l:: + scale counter values + +EXAMPLES +-------- + +$ perf stat -- make -j + + Performance counter stats for 'make -j': + + 8117.370256 task clock ticks # 11.281 CPU utilization factor + 678 context switches # 0.000 M/sec + 133 CPU migrations # 0.000 M/sec + 235724 pagefaults # 0.029 M/sec + 24821162526 CPU cycles # 3057.784 M/sec + 18687303457 instructions # 2302.138 M/sec + 172158895 cache references # 21.209 M/sec + 27075259 cache misses # 3.335 M/sec + + Wall-clock time elapsed: 719.554352 msecs + +SEE ALSO +-------- +linkperf:perf-top[1], linkperf:perf-list[1] diff --git a/tools/perf/Documentation/perf-top.txt b/tools/perf/Documentation/perf-top.txt new file mode 100644 index 00000000000..539d0128972 --- /dev/null +++ b/tools/perf/Documentation/perf-top.txt @@ -0,0 +1,39 @@ +perf-top(1) +=========== + +NAME +---- +perf-top - Run a command and profile it + +SYNOPSIS +-------- +[verse] +'perf top' [-e <EVENT> | --event=EVENT] [-l] [-a] <command> + +DESCRIPTION +----------- +This command runs a command and gathers a performance counter profile +from it. + + +OPTIONS +------- +<command>...:: + Any command you can specify in a shell. + +-e:: +--event=:: + Select the PMU event. Selection can be a symbolic event name + (use 'perf list' to list all events) or a raw PMU + event (eventsel+umask) in the form of rNNN where NNN is a + hexadecimal event descriptor. + +-a:: + system-wide collection + +-l:: + scale counter values + +SEE ALSO +-------- +linkperf:perf-stat[1], linkperf:perf-list[1] diff --git a/tools/perf/Documentation/perf.txt b/tools/perf/Documentation/perf.txt new file mode 100644 index 00000000000..69c83255719 --- /dev/null +++ b/tools/perf/Documentation/perf.txt @@ -0,0 +1,24 @@ +perf(1) +======= + +NAME +---- +perf - Performance analysis tools for Linux + +SYNOPSIS +-------- +[verse] +'perf' [--version] [--help] COMMAND [ARGS] + +DESCRIPTION +----------- +Performance counters for Linux are are a new kernel-based subsystem +that provide a framework for all things performance analysis. It +covers hardware level (CPU/PMU, Performance Monitoring Unit) features +and software features (software counters, tracepoints) as well. + +SEE ALSO +-------- +linkperf:perf-stat[1], linkperf:perf-top[1], +linkperf:perf-record[1], linkperf:perf-report[1], +linkperf:perf-list[1] diff --git a/tools/perf/Makefile b/tools/perf/Makefile new file mode 100644 index 00000000000..0cbd5d6874e --- /dev/null +++ b/tools/perf/Makefile @@ -0,0 +1,929 @@ +# The default target of this Makefile is... +all:: + +# Define V=1 to have a more verbose compile. +# +# Define SNPRINTF_RETURNS_BOGUS if your are on a system which snprintf() +# or vsnprintf() return -1 instead of number of characters which would +# have been written to the final string if enough space had been available. +# +# Define FREAD_READS_DIRECTORIES if your are on a system which succeeds +# when attempting to read from an fopen'ed directory. +# +# Define NO_OPENSSL environment variable if you do not have OpenSSL. +# This also implies MOZILLA_SHA1. +# +# Define CURLDIR=/foo/bar if your curl header and library files are in +# /foo/bar/include and /foo/bar/lib directories. +# +# Define EXPATDIR=/foo/bar if your expat header and library files are in +# /foo/bar/include and /foo/bar/lib directories. +# +# Define NO_D_INO_IN_DIRENT if you don't have d_ino in your struct dirent. +# +# Define NO_D_TYPE_IN_DIRENT if your platform defines DT_UNKNOWN but lacks +# d_type in struct dirent (latest Cygwin -- will be fixed soonish). +# +# Define NO_C99_FORMAT if your formatted IO functions (printf/scanf et.al.) +# do not support the 'size specifiers' introduced by C99, namely ll, hh, +# j, z, t. (representing long long int, char, intmax_t, size_t, ptrdiff_t). +# some C compilers supported these specifiers prior to C99 as an extension. +# +# Define NO_STRCASESTR if you don't have strcasestr. +# +# Define NO_MEMMEM if you don't have memmem. +# +# Define NO_STRTOUMAX if you don't have strtoumax in the C library. +# If your compiler also does not support long long or does not have +# strtoull, define NO_STRTOULL. +# +# Define NO_SETENV if you don't have setenv in the C library. +# +# Define NO_UNSETENV if you don't have unsetenv in the C library. +# +# Define NO_MKDTEMP if you don't have mkdtemp in the C library. +# +# Define NO_SYS_SELECT_H if you don't have sys/select.h. +# +# Define NO_SYMLINK_HEAD if you never want .perf/HEAD to be a symbolic link. +# Enable it on Windows. By default, symrefs are still used. +# +# Define NO_SVN_TESTS if you want to skip time-consuming SVN interoperability +# tests. These tests take up a significant amount of the total test time +# but are not needed unless you plan to talk to SVN repos. +# +# Define NO_FINK if you are building on Darwin/Mac OS X, have Fink +# installed in /sw, but don't want PERF to link against any libraries +# installed there. If defined you may specify your own (or Fink's) +# include directories and library directories by defining CFLAGS +# and LDFLAGS appropriately. +# +# Define NO_DARWIN_PORTS if you are building on Darwin/Mac OS X, +# have DarwinPorts installed in /opt/local, but don't want PERF to +# link against any libraries installed there. If defined you may +# specify your own (or DarwinPort's) include directories and +# library directories by defining CFLAGS and LDFLAGS appropriately. +# +# Define PPC_SHA1 environment variable when running make to make use of +# a bundled SHA1 routine optimized for PowerPC. +# +# Define ARM_SHA1 environment variable when running make to make use of +# a bundled SHA1 routine optimized for ARM. +# +# Define MOZILLA_SHA1 environment variable when running make to make use of +# a bundled SHA1 routine coming from Mozilla. It is GPL'd and should be fast +# on non-x86 architectures (e.g. PowerPC), while the OpenSSL version (default +# choice) has very fast version optimized for i586. +# +# Define NEEDS_SSL_WITH_CRYPTO if you need -lcrypto with -lssl (Darwin). +# +# Define NEEDS_LIBICONV if linking with libc is not enough (Darwin). +# +# Define NEEDS_SOCKET if linking with libc is not enough (SunOS, +# Patrick Mauritz). +# +# Define NO_MMAP if you want to avoid mmap. +# +# Define NO_PTHREADS if you do not have or do not want to use Pthreads. +# +# Define NO_PREAD if you have a problem with pread() system call (e.g. +# cygwin.dll before v1.5.22). +# +# Define NO_FAST_WORKING_DIRECTORY if accessing objects in pack files is +# generally faster on your platform than accessing the working directory. +# +# Define NO_TRUSTABLE_FILEMODE if your filesystem may claim to support +# the executable mode bit, but doesn't really do so. +# +# Define NO_IPV6 if you lack IPv6 support and getaddrinfo(). +# +# Define NO_SOCKADDR_STORAGE if your platform does not have struct +# sockaddr_storage. +# +# Define NO_ICONV if your libc does not properly support iconv. +# +# Define OLD_ICONV if your library has an old iconv(), where the second +# (input buffer pointer) parameter is declared with type (const char **). +# +# Define NO_DEFLATE_BOUND if your zlib does not have deflateBound. +# +# Define NO_R_TO_GCC_LINKER if your gcc does not like "-R/path/lib" +# that tells runtime paths to dynamic libraries; +# "-Wl,-rpath=/path/lib" is used instead. +# +# Define USE_NSEC below if you want perf to care about sub-second file mtimes +# and ctimes. Note that you need recent glibc (at least 2.2.4) for this, and +# it will BREAK YOUR LOCAL DIFFS! show-diff and anything using it will likely +# randomly break unless your underlying filesystem supports those sub-second +# times (my ext3 doesn't). +# +# Define USE_ST_TIMESPEC if your "struct stat" uses "st_ctimespec" instead of +# "st_ctim" +# +# Define NO_NSEC if your "struct stat" does not have "st_ctim.tv_nsec" +# available. This automatically turns USE_NSEC off. +# +# Define USE_STDEV below if you want perf to care about the underlying device +# change being considered an inode change from the update-index perspective. +# +# Define NO_ST_BLOCKS_IN_STRUCT_STAT if your platform does not have st_blocks +# field that counts the on-disk footprint in 512-byte blocks. +# +# Define ASCIIDOC8 if you want to format documentation with AsciiDoc 8 +# +# Define DOCBOOK_XSL_172 if you want to format man pages with DocBook XSL v1.72. +# +# Define NO_PERL_MAKEMAKER if you cannot use Makefiles generated by perl's +# MakeMaker (e.g. using ActiveState under Cygwin). +# +# Define NO_PERL if you do not want Perl scripts or libraries at all. +# +# Define INTERNAL_QSORT to use Git's implementation of qsort(), which +# is a simplified version of the merge sort used in glibc. This is +# recommended if Git triggers O(n^2) behavior in your platform's qsort(). +# +# Define NO_EXTERNAL_GREP if you don't want "perf grep" to ever call +# your external grep (e.g., if your system lacks grep, if its grep is +# broken, or spawning external process is slower than built-in grep perf has). + +PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE + @$(SHELL_PATH) util/PERF-VERSION-GEN +-include PERF-VERSION-FILE + +uname_S := $(shell sh -c 'uname -s 2>/dev/null || echo not') +uname_M := $(shell sh -c 'uname -m 2>/dev/null || echo not') +uname_O := $(shell sh -c 'uname -o 2>/dev/null || echo not') +uname_R := $(shell sh -c 'uname -r 2>/dev/null || echo not') +uname_P := $(shell sh -c 'uname -p 2>/dev/null || echo not') +uname_V := $(shell sh -c 'uname -v 2>/dev/null || echo not') + +# CFLAGS and LDFLAGS are for the users to override from the command line. + +CFLAGS = -ggdb3 -Wall -Werror -Wstrict-prototypes -Wmissing-declarations -Wmissing-prototypes -std=gnu99 -Wdeclaration-after-statement -O6 +LDFLAGS = -lpthread -lrt -lelf +ALL_CFLAGS = $(CFLAGS) +ALL_LDFLAGS = $(LDFLAGS) +STRIP ?= strip + +# Among the variables below, these: +# perfexecdir +# template_dir +# mandir +# infodir +# htmldir +# ETC_PERFCONFIG (but not sysconfdir) +# can be specified as a relative path some/where/else; +# this is interpreted as relative to $(prefix) and "perf" at +# runtime figures out where they are based on the path to the executable. +# This can help installing the suite in a relocatable way. + +prefix = $(HOME) +bindir_relative = bin +bindir = $(prefix)/$(bindir_relative) +mandir = share/man +infodir = share/info +perfexecdir = libexec/perf-core +sharedir = $(prefix)/share +template_dir = share/perf-core/templates +htmldir = share/doc/perf-doc +ifeq ($(prefix),/usr) +sysconfdir = /etc +ETC_PERFCONFIG = $(sysconfdir)/perfconfig +else +sysconfdir = $(prefix)/etc +ETC_PERFCONFIG = etc/perfconfig +endif +lib = lib +# DESTDIR= + +export prefix bindir sharedir sysconfdir + +CC = gcc +AR = ar +RM = rm -f +TAR = tar +FIND = find +INSTALL = install +RPMBUILD = rpmbuild +PTHREAD_LIBS = -lpthread + +# sparse is architecture-neutral, which means that we need to tell it +# explicitly what architecture to check for. Fix this up for yours.. +SPARSE_FLAGS = -D__BIG_ENDIAN__ -D__powerpc__ + + + +### --- END CONFIGURATION SECTION --- + +# Those must not be GNU-specific; they are shared with perl/ which may +# be built by a different compiler. (Note that this is an artifact now +# but it still might be nice to keep that distinction.) +BASIC_CFLAGS = +BASIC_LDFLAGS = + +# Guard against environment variables +BUILTIN_OBJS = +BUILT_INS = +COMPAT_CFLAGS = +COMPAT_OBJS = +LIB_H = +LIB_OBJS = +SCRIPT_PERL = +SCRIPT_SH = +TEST_PROGRAMS = + +# +# No scripts right now: +# + +# SCRIPT_SH += perf-am.sh + +# +# No Perl scripts right now: +# + +# SCRIPT_PERL += perf-add--interactive.perl + +SCRIPTS = $(patsubst %.sh,%,$(SCRIPT_SH)) \ + $(patsubst %.perl,%,$(SCRIPT_PERL)) + +# Empty... +EXTRA_PROGRAMS = + +# ... and all the rest that could be moved out of bindir to perfexecdir +PROGRAMS += $(EXTRA_PROGRAMS) + +# +# Single 'perf' binary right now: +# +PROGRAMS += perf + +# List built-in command $C whose implementation cmd_$C() is not in +# builtin-$C.o but is linked in as part of some other command. +# +# None right now: +# +# BUILT_INS += perf-init $X + +# what 'all' will build and 'install' will install, in perfexecdir +ALL_PROGRAMS = $(PROGRAMS) $(SCRIPTS) + +# what 'all' will build but not install in perfexecdir +OTHER_PROGRAMS = perf$X + +# Set paths to tools early so that they can be used for version tests. +ifndef SHELL_PATH + SHELL_PATH = /bin/sh +endif +ifndef PERL_PATH + PERL_PATH = /usr/bin/perl +endif + +export PERL_PATH + +LIB_FILE=libperf.a + +LIB_H += ../../include/linux/perf_counter.h +LIB_H += perf.h +LIB_H += util/list.h +LIB_H += util/rbtree.h +LIB_H += util/levenshtein.h +LIB_H += util/parse-options.h +LIB_H += util/parse-events.h +LIB_H += util/quote.h +LIB_H += util/util.h +LIB_H += util/help.h +LIB_H += util/strbuf.h +LIB_H += util/string.h +LIB_H += util/run-command.h +LIB_H += util/sigchain.h +LIB_H += util/symbol.h +LIB_H += util/color.h + +LIB_OBJS += util/abspath.o +LIB_OBJS += util/alias.o +LIB_OBJS += util/config.o +LIB_OBJS += util/ctype.o +LIB_OBJS += util/environment.o +LIB_OBJS += util/exec_cmd.o +LIB_OBJS += util/help.o +LIB_OBJS += util/levenshtein.o +LIB_OBJS += util/parse-options.o +LIB_OBJS += util/parse-events.o +LIB_OBJS += util/path.o +LIB_OBJS += util/rbtree.o +LIB_OBJS += util/run-command.o +LIB_OBJS += util/quote.o +LIB_OBJS += util/strbuf.o +LIB_OBJS += util/string.o +LIB_OBJS += util/usage.o +LIB_OBJS += util/wrapper.o +LIB_OBJS += util/sigchain.o +LIB_OBJS += util/symbol.o +LIB_OBJS += util/color.o +LIB_OBJS += util/pager.o + +BUILTIN_OBJS += builtin-annotate.o +BUILTIN_OBJS += builtin-help.o +BUILTIN_OBJS += builtin-list.o +BUILTIN_OBJS += builtin-record.o +BUILTIN_OBJS += builtin-report.o +BUILTIN_OBJS += builtin-stat.o +BUILTIN_OBJS += builtin-top.o + +PERFLIBS = $(LIB_FILE) +EXTLIBS = + +# +# Platform specific tweaks +# + +# We choose to avoid "if .. else if .. else .. endif endif" +# because maintaining the nesting to match is a pain. If +# we had "elif" things would have been much nicer... + +-include config.mak.autogen +-include config.mak + +ifeq ($(uname_S),Darwin) + ifndef NO_FINK + ifeq ($(shell test -d /sw/lib && echo y),y) + BASIC_CFLAGS += -I/sw/include + BASIC_LDFLAGS += -L/sw/lib + endif + endif + ifndef NO_DARWIN_PORTS + ifeq ($(shell test -d /opt/local/lib && echo y),y) + BASIC_CFLAGS += -I/opt/local/include + BASIC_LDFLAGS += -L/opt/local/lib + endif + endif + PTHREAD_LIBS = +endif + +ifndef CC_LD_DYNPATH + ifdef NO_R_TO_GCC_LINKER + # Some gcc does not accept and pass -R to the linker to specify + # the runtime dynamic library path. + CC_LD_DYNPATH = -Wl,-rpath, + else + CC_LD_DYNPATH = -R + endif +endif + +ifdef ZLIB_PATH + BASIC_CFLAGS += -I$(ZLIB_PATH)/include + EXTLIBS += -L$(ZLIB_PATH)/$(lib) $(CC_LD_DYNPATH)$(ZLIB_PATH)/$(lib) +endif +EXTLIBS += -lz + +ifdef NEEDS_SOCKET + EXTLIBS += -lsocket +endif +ifdef NEEDS_NSL + EXTLIBS += -lnsl +endif +ifdef NO_D_TYPE_IN_DIRENT + BASIC_CFLAGS += -DNO_D_TYPE_IN_DIRENT +endif +ifdef NO_D_INO_IN_DIRENT + BASIC_CFLAGS += -DNO_D_INO_IN_DIRENT +endif +ifdef NO_ST_BLOCKS_IN_STRUCT_STAT + BASIC_CFLAGS += -DNO_ST_BLOCKS_IN_STRUCT_STAT +endif +ifdef USE_NSEC + BASIC_CFLAGS += -DUSE_NSEC +endif +ifdef USE_ST_TIMESPEC + BASIC_CFLAGS += -DUSE_ST_TIMESPEC +endif +ifdef NO_NSEC + BASIC_CFLAGS += -DNO_NSEC +endif +ifdef NO_C99_FORMAT + BASIC_CFLAGS += -DNO_C99_FORMAT +endif +ifdef SNPRINTF_RETURNS_BOGUS + COMPAT_CFLAGS += -DSNPRINTF_RETURNS_BOGUS + COMPAT_OBJS += compat/snprintf.o +endif +ifdef FREAD_READS_DIRECTORIES + COMPAT_CFLAGS += -DFREAD_READS_DIRECTORIES + COMPAT_OBJS += compat/fopen.o +endif +ifdef NO_SYMLINK_HEAD + BASIC_CFLAGS += -DNO_SYMLINK_HEAD +endif +ifdef NO_STRCASESTR + COMPAT_CFLAGS += -DNO_STRCASESTR + COMPAT_OBJS += compat/strcasestr.o +endif +ifdef NO_STRTOUMAX + COMPAT_CFLAGS += -DNO_STRTOUMAX + COMPAT_OBJS += compat/strtoumax.o +endif +ifdef NO_STRTOULL + COMPAT_CFLAGS += -DNO_STRTOULL +endif +ifdef NO_SETENV + COMPAT_CFLAGS += -DNO_SETENV + COMPAT_OBJS += compat/setenv.o +endif +ifdef NO_MKDTEMP + COMPAT_CFLAGS += -DNO_MKDTEMP + COMPAT_OBJS += compat/mkdtemp.o +endif +ifdef NO_UNSETENV + COMPAT_CFLAGS += -DNO_UNSETENV + COMPAT_OBJS += compat/unsetenv.o +endif +ifdef NO_SYS_SELECT_H + BASIC_CFLAGS += -DNO_SYS_SELECT_H +endif +ifdef NO_MMAP + COMPAT_CFLAGS += -DNO_MMAP + COMPAT_OBJS += compat/mmap.o +else + ifdef USE_WIN32_MMAP + COMPAT_CFLAGS += -DUSE_WIN32_MMAP + COMPAT_OBJS += compat/win32mmap.o + endif +endif +ifdef NO_PREAD + COMPAT_CFLAGS += -DNO_PREAD + COMPAT_OBJS += compat/pread.o +endif +ifdef NO_FAST_WORKING_DIRECTORY + BASIC_CFLAGS += -DNO_FAST_WORKING_DIRECTORY +endif +ifdef NO_TRUSTABLE_FILEMODE + BASIC_CFLAGS += -DNO_TRUSTABLE_FILEMODE +endif +ifdef NO_IPV6 + BASIC_CFLAGS += -DNO_IPV6 +endif +ifdef NO_UINTMAX_T + BASIC_CFLAGS += -Duintmax_t=uint32_t +endif +ifdef NO_SOCKADDR_STORAGE +ifdef NO_IPV6 + BASIC_CFLAGS += -Dsockaddr_storage=sockaddr_in +else + BASIC_CFLAGS += -Dsockaddr_storage=sockaddr_in6 +endif +endif +ifdef NO_INET_NTOP + LIB_OBJS += compat/inet_ntop.o +endif +ifdef NO_INET_PTON + LIB_OBJS += compat/inet_pton.o +endif + +ifdef NO_ICONV + BASIC_CFLAGS += -DNO_ICONV +endif + +ifdef OLD_ICONV + BASIC_CFLAGS += -DOLD_ICONV +endif + +ifdef NO_DEFLATE_BOUND + BASIC_CFLAGS += -DNO_DEFLATE_BOUND +endif + +ifdef PPC_SHA1 + SHA1_HEADER = "ppc/sha1.h" + LIB_OBJS += ppc/sha1.o ppc/sha1ppc.o +else +ifdef ARM_SHA1 + SHA1_HEADER = "arm/sha1.h" + LIB_OBJS += arm/sha1.o arm/sha1_arm.o +else +ifdef MOZILLA_SHA1 + SHA1_HEADER = "mozilla-sha1/sha1.h" + LIB_OBJS += mozilla-sha1/sha1.o +else + SHA1_HEADER = <openssl/sha.h> + EXTLIBS += $(LIB_4_CRYPTO) +endif +endif +endif +ifdef NO_PERL_MAKEMAKER + export NO_PERL_MAKEMAKER +endif +ifdef NO_HSTRERROR + COMPAT_CFLAGS += -DNO_HSTRERROR + COMPAT_OBJS += compat/hstrerror.o +endif +ifdef NO_MEMMEM + COMPAT_CFLAGS += -DNO_MEMMEM + COMPAT_OBJS += compat/memmem.o +endif +ifdef INTERNAL_QSORT + COMPAT_CFLAGS += -DINTERNAL_QSORT + COMPAT_OBJS += compat/qsort.o +endif +ifdef RUNTIME_PREFIX + COMPAT_CFLAGS += -DRUNTIME_PREFIX +endif + +ifdef DIR_HAS_BSD_GROUP_SEMANTICS + COMPAT_CFLAGS += -DDIR_HAS_BSD_GROUP_SEMANTICS +endif +ifdef NO_EXTERNAL_GREP + BASIC_CFLAGS += -DNO_EXTERNAL_GREP +endif + +ifeq ($(PERL_PATH),) +NO_PERL=NoThanks +endif + +QUIET_SUBDIR0 = +$(MAKE) -C # space to separate -C and subdir +QUIET_SUBDIR1 = + +ifneq ($(findstring $(MAKEFLAGS),w),w) +PRINT_DIR = --no-print-directory +else # "make -w" +NO_SUBDIR = : +endif + +ifneq ($(findstring $(MAKEFLAGS),s),s) +ifndef V + QUIET_CC = @echo ' ' CC $@; + QUIET_AR = @echo ' ' AR $@; + QUIET_LINK = @echo ' ' LINK $@; + QUIET_BUILT_IN = @echo ' ' BUILTIN $@; + QUIET_GEN = @echo ' ' GEN $@; + QUIET_SUBDIR0 = +@subdir= + QUIET_SUBDIR1 = ;$(NO_SUBDIR) echo ' ' SUBDIR $$subdir; \ + $(MAKE) $(PRINT_DIR) -C $$subdir + export V + export QUIET_GEN + export QUIET_BUILT_IN +endif +endif + +ifdef ASCIIDOC8 + export ASCIIDOC8 +endif + +# Shell quote (do not use $(call) to accommodate ancient setups); + +SHA1_HEADER_SQ = $(subst ','\'',$(SHA1_HEADER)) +ETC_PERFCONFIG_SQ = $(subst ','\'',$(ETC_PERFCONFIG)) + +DESTDIR_SQ = $(subst ','\'',$(DESTDIR)) +bindir_SQ = $(subst ','\'',$(bindir)) +bindir_relative_SQ = $(subst ','\'',$(bindir_relative)) +mandir_SQ = $(subst ','\'',$(mandir)) +infodir_SQ = $(subst ','\'',$(infodir)) +perfexecdir_SQ = $(subst ','\'',$(perfexecdir)) +template_dir_SQ = $(subst ','\'',$(template_dir)) +htmldir_SQ = $(subst ','\'',$(htmldir)) +prefix_SQ = $(subst ','\'',$(prefix)) + +SHELL_PATH_SQ = $(subst ','\'',$(SHELL_PATH)) +PERL_PATH_SQ = $(subst ','\'',$(PERL_PATH)) + +LIBS = $(PERFLIBS) $(EXTLIBS) + +BASIC_CFLAGS += -DSHA1_HEADER='$(SHA1_HEADER_SQ)' \ + $(COMPAT_CFLAGS) +LIB_OBJS += $(COMPAT_OBJS) + +ALL_CFLAGS += $(BASIC_CFLAGS) +ALL_LDFLAGS += $(BASIC_LDFLAGS) + +export TAR INSTALL DESTDIR SHELL_PATH + + +### Build rules + +SHELL = $(SHELL_PATH) + +all:: shell_compatibility_test $(ALL_PROGRAMS) $(BUILT_INS) $(OTHER_PROGRAMS) PERF-BUILD-OPTIONS +ifneq (,$X) + $(foreach p,$(patsubst %$X,%,$(filter %$X,$(ALL_PROGRAMS) $(BUILT_INS) perf$X)), test '$p' -ef '$p$X' || $(RM) '$p';) +endif + +all:: + +please_set_SHELL_PATH_to_a_more_modern_shell: + @$$(:) + +shell_compatibility_test: please_set_SHELL_PATH_to_a_more_modern_shell + +strip: $(PROGRAMS) perf$X + $(STRIP) $(STRIP_OPTS) $(PROGRAMS) perf$X + +perf.o: perf.c common-cmds.h PERF-CFLAGS + $(QUIET_CC)$(CC) -DPERF_VERSION='"$(PERF_VERSION)"' \ + '-DPERF_HTML_PATH="$(htmldir_SQ)"' \ + $(ALL_CFLAGS) -c $(filter %.c,$^) + +perf$X: perf.o $(BUILTIN_OBJS) $(PERFLIBS) + $(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ perf.o \ + $(BUILTIN_OBJS) $(ALL_LDFLAGS) $(LIBS) + +builtin-help.o: builtin-help.c common-cmds.h PERF-CFLAGS + $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) \ + '-DPERF_HTML_PATH="$(htmldir_SQ)"' \ + '-DPERF_MAN_PATH="$(mandir_SQ)"' \ + '-DPERF_INFO_PATH="$(infodir_SQ)"' $< + +$(BUILT_INS): perf$X + $(QUIET_BUILT_IN)$(RM) $@ && \ + ln perf$X $@ 2>/dev/null || \ + ln -s perf$X $@ 2>/dev/null || \ + cp perf$X $@ + +common-cmds.h: util/generate-cmdlist.sh command-list.txt + +common-cmds.h: $(wildcard Documentation/perf-*.txt) + $(QUIET_GEN)util/generate-cmdlist.sh > $@+ && mv $@+ $@ + +$(patsubst %.sh,%,$(SCRIPT_SH)) : % : %.sh + $(QUIET_GEN)$(RM) $@ $@+ && \ + sed -e '1s|#!.*/sh|#!$(SHELL_PATH_SQ)|' \ + -e 's|@SHELL_PATH@|$(SHELL_PATH_SQ)|' \ + -e 's|@@PERL@@|$(PERL_PATH_SQ)|g' \ + -e 's/@@PERF_VERSION@@/$(PERF_VERSION)/g' \ + -e 's/@@NO_CURL@@/$(NO_CURL)/g' \ + $@.sh >$@+ && \ + chmod +x $@+ && \ + mv $@+ $@ + +configure: configure.ac + $(QUIET_GEN)$(RM) $@ $<+ && \ + sed -e 's/@@PERF_VERSION@@/$(PERF_VERSION)/g' \ + $< > $<+ && \ + autoconf -o $@ $<+ && \ + $(RM) $<+ + +# These can record PERF_VERSION +perf.o perf.spec \ + $(patsubst %.sh,%,$(SCRIPT_SH)) \ + $(patsubst %.perl,%,$(SCRIPT_PERL)) \ + : PERF-VERSION-FILE + +%.o: %.c PERF-CFLAGS + $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) $< +%.s: %.c PERF-CFLAGS + $(QUIET_CC)$(CC) -S $(ALL_CFLAGS) $< +%.o: %.S + $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) $< + +util/exec_cmd.o: util/exec_cmd.c PERF-CFLAGS + $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) \ + '-DPERF_EXEC_PATH="$(perfexecdir_SQ)"' \ + '-DBINDIR="$(bindir_relative_SQ)"' \ + '-DPREFIX="$(prefix_SQ)"' \ + $< + +builtin-init-db.o: builtin-init-db.c PERF-CFLAGS + $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) -DDEFAULT_PERF_TEMPLATE_DIR='"$(template_dir_SQ)"' $< + +util/config.o: util/config.c PERF-CFLAGS + $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< + +perf-%$X: %.o $(PERFLIBS) + $(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(ALL_LDFLAGS) $(filter %.o,$^) $(LIBS) + +$(LIB_OBJS) $(BUILTIN_OBJS): $(LIB_H) +$(patsubst perf-%$X,%.o,$(PROGRAMS)): $(LIB_H) $(wildcard */*.h) +builtin-revert.o wt-status.o: wt-status.h + +$(LIB_FILE): $(LIB_OBJS) + $(QUIET_AR)$(RM) $@ && $(AR) rcs $@ $(LIB_OBJS) + +doc: + $(MAKE) -C Documentation all + +man: + $(MAKE) -C Documentation man + +html: + $(MAKE) -C Documentation html + +info: + $(MAKE) -C Documentation info + +pdf: + $(MAKE) -C Documentation pdf + +TAGS: + $(RM) TAGS + $(FIND) . -name '*.[hcS]' -print | xargs etags -a + +tags: + $(RM) tags + $(FIND) . -name '*.[hcS]' -print | xargs ctags -a + +cscope: + $(RM) cscope* + $(FIND) . -name '*.[hcS]' -print | xargs cscope -b + +### Detect prefix changes +TRACK_CFLAGS = $(subst ','\'',$(ALL_CFLAGS)):\ + $(bindir_SQ):$(perfexecdir_SQ):$(template_dir_SQ):$(prefix_SQ) + +PERF-CFLAGS: .FORCE-PERF-CFLAGS + @FLAGS='$(TRACK_CFLAGS)'; \ + if test x"$$FLAGS" != x"`cat PERF-CFLAGS 2>/dev/null`" ; then \ + echo 1>&2 " * new build flags or prefix"; \ + echo "$$FLAGS" >PERF-CFLAGS; \ + fi + +# We need to apply sq twice, once to protect from the shell +# that runs PERF-BUILD-OPTIONS, and then again to protect it +# and the first level quoting from the shell that runs "echo". +PERF-BUILD-OPTIONS: .FORCE-PERF-BUILD-OPTIONS + @echo SHELL_PATH=\''$(subst ','\'',$(SHELL_PATH_SQ))'\' >$@ + @echo TAR=\''$(subst ','\'',$(subst ','\'',$(TAR)))'\' >>$@ + @echo NO_CURL=\''$(subst ','\'',$(subst ','\'',$(NO_CURL)))'\' >>$@ + @echo NO_PERL=\''$(subst ','\'',$(subst ','\'',$(NO_PERL)))'\' >>$@ + +### Testing rules + +# +# None right now: +# +# TEST_PROGRAMS += test-something$X + +all:: $(TEST_PROGRAMS) + +# GNU make supports exporting all variables by "export" without parameters. +# However, the environment gets quite big, and some programs have problems +# with that. + +export NO_SVN_TESTS + +check: common-cmds.h + if sparse; \ + then \ + for i in *.c */*.c; \ + do \ + sparse $(ALL_CFLAGS) $(SPARSE_FLAGS) $$i || exit; \ + done; \ + else \ + echo 2>&1 "Did you mean 'make test'?"; \ + exit 1; \ + fi + +remove-dashes: + ./fixup-builtins $(BUILT_INS) $(PROGRAMS) $(SCRIPTS) + +### Installation rules + +ifneq ($(filter /%,$(firstword $(template_dir))),) +template_instdir = $(template_dir) +else +template_instdir = $(prefix)/$(template_dir) +endif +export template_instdir + +ifneq ($(filter /%,$(firstword $(perfexecdir))),) +perfexec_instdir = $(perfexecdir) +else +perfexec_instdir = $(prefix)/$(perfexecdir) +endif +perfexec_instdir_SQ = $(subst ','\'',$(perfexec_instdir)) +export perfexec_instdir + +install: all + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(bindir_SQ)' + $(INSTALL) perf$X '$(DESTDIR_SQ)$(bindir_SQ)' +ifdef BUILT_INS + $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' + $(INSTALL) $(BUILT_INS) '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' +ifneq (,$X) + $(foreach p,$(patsubst %$X,%,$(filter %$X,$(ALL_PROGRAMS) $(BUILT_INS) perf$X)), $(RM) '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/$p';) +endif +endif + +install-doc: + $(MAKE) -C Documentation install + +install-man: + $(MAKE) -C Documentation install-man + +install-html: + $(MAKE) -C Documentation install-html + +install-info: + $(MAKE) -C Documentation install-info + +install-pdf: + $(MAKE) -C Documentation install-pdf + +quick-install-doc: + $(MAKE) -C Documentation quick-install + +quick-install-man: + $(MAKE) -C Documentation quick-install-man + +quick-install-html: + $(MAKE) -C Documentation quick-install-html + + +### Maintainer's dist rules +# +# None right now +# +# +# perf.spec: perf.spec.in +# sed -e 's/@@VERSION@@/$(PERF_VERSION)/g' < $< > $@+ +# mv $@+ $@ +# +# PERF_TARNAME=perf-$(PERF_VERSION) +# dist: perf.spec perf-archive$(X) configure +# ./perf-archive --format=tar \ +# --prefix=$(PERF_TARNAME)/ HEAD^{tree} > $(PERF_TARNAME).tar +# @mkdir -p $(PERF_TARNAME) +# @cp perf.spec configure $(PERF_TARNAME) +# @echo $(PERF_VERSION) > $(PERF_TARNAME)/version +# $(TAR) rf $(PERF_TARNAME).tar \ +# $(PERF_TARNAME)/perf.spec \ +# $(PERF_TARNAME)/configure \ +# $(PERF_TARNAME)/version +# @$(RM) -r $(PERF_TARNAME) +# gzip -f -9 $(PERF_TARNAME).tar +# +# htmldocs = perf-htmldocs-$(PERF_VERSION) +# manpages = perf-manpages-$(PERF_VERSION) +# dist-doc: +# $(RM) -r .doc-tmp-dir +# mkdir .doc-tmp-dir +# $(MAKE) -C Documentation WEBDOC_DEST=../.doc-tmp-dir install-webdoc +# cd .doc-tmp-dir && $(TAR) cf ../$(htmldocs).tar . +# gzip -n -9 -f $(htmldocs).tar +# : +# $(RM) -r .doc-tmp-dir +# mkdir -p .doc-tmp-dir/man1 .doc-tmp-dir/man5 .doc-tmp-dir/man7 +# $(MAKE) -C Documentation DESTDIR=./ \ +# man1dir=../.doc-tmp-dir/man1 \ +# man5dir=../.doc-tmp-dir/man5 \ +# man7dir=../.doc-tmp-dir/man7 \ +# install +# cd .doc-tmp-dir && $(TAR) cf ../$(manpages).tar . +# gzip -n -9 -f $(manpages).tar +# $(RM) -r .doc-tmp-dir +# +# rpm: dist +# $(RPMBUILD) -ta $(PERF_TARNAME).tar.gz + +### Cleaning rules + +distclean: clean +# $(RM) configure + +clean: + $(RM) *.o */*.o $(LIB_FILE) + $(RM) $(ALL_PROGRAMS) $(BUILT_INS) perf$X + $(RM) $(TEST_PROGRAMS) + $(RM) *.spec *.pyc *.pyo */*.pyc */*.pyo common-cmds.h TAGS tags cscope* + $(RM) -r autom4te.cache + $(RM) config.log config.mak.autogen config.mak.append config.status config.cache + $(RM) -r $(PERF_TARNAME) .doc-tmp-dir + $(RM) $(PERF_TARNAME).tar.gz perf-core_$(PERF_VERSION)-*.tar.gz + $(RM) $(htmldocs).tar.gz $(manpages).tar.gz + $(MAKE) -C Documentation/ clean + $(RM) PERF-VERSION-FILE PERF-CFLAGS PERF-BUILD-OPTIONS + +.PHONY: all install clean strip +.PHONY: shell_compatibility_test please_set_SHELL_PATH_to_a_more_modern_shell +.PHONY: .FORCE-PERF-VERSION-FILE TAGS tags cscope .FORCE-PERF-CFLAGS +.PHONY: .FORCE-PERF-BUILD-OPTIONS + +### Make sure built-ins do not have dups and listed in perf.c +# +check-builtins:: + ./check-builtins.sh + +### Test suite coverage testing +# +# None right now +# +# .PHONY: coverage coverage-clean coverage-build coverage-report +# +# coverage: +# $(MAKE) coverage-build +# $(MAKE) coverage-report +# +# coverage-clean: +# rm -f *.gcda *.gcno +# +# COVERAGE_CFLAGS = $(CFLAGS) -O0 -ftest-coverage -fprofile-arcs +# COVERAGE_LDFLAGS = $(CFLAGS) -O0 -lgcov +# +# coverage-build: coverage-clean +# $(MAKE) CFLAGS="$(COVERAGE_CFLAGS)" LDFLAGS="$(COVERAGE_LDFLAGS)" all +# $(MAKE) CFLAGS="$(COVERAGE_CFLAGS)" LDFLAGS="$(COVERAGE_LDFLAGS)" \ +# -j1 test +# +# coverage-report: +# gcov -b *.c */*.c +# grep '^function.*called 0 ' *.c.gcov */*.c.gcov \ +# | sed -e 's/\([^:]*\)\.gcov: *function \([^ ]*\) called.*/\1: \2/' \ +# | tee coverage-untested-functions diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c new file mode 100644 index 00000000000..b1ed5f766cb --- /dev/null +++ b/tools/perf/builtin-annotate.c @@ -0,0 +1,1356 @@ +/* + * builtin-annotate.c + * + * Builtin annotate command: Analyze the perf.data input file, + * look up and read DSOs and symbol information and display + * a histogram of results, along various sorting keys. + */ +#include "builtin.h" + +#include "util/util.h" + +#include "util/color.h" +#include "util/list.h" +#include "util/cache.h" +#include "util/rbtree.h" +#include "util/symbol.h" +#include "util/string.h" + +#include "perf.h" + +#include "util/parse-options.h" +#include "util/parse-events.h" + +#define SHOW_KERNEL 1 +#define SHOW_USER 2 +#define SHOW_HV 4 + +static char const *input_name = "perf.data"; +static char *vmlinux = "vmlinux"; + +static char default_sort_order[] = "comm,symbol"; +static char *sort_order = default_sort_order; + +static int input; +static int show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV; + +static int dump_trace = 0; +#define dprintf(x...) do { if (dump_trace) printf(x); } while (0) + +static int verbose; + +static unsigned long page_size; +static unsigned long mmap_window = 32; + +struct ip_event { + struct perf_event_header header; + __u64 ip; + __u32 pid, tid; +}; + +struct mmap_event { + struct perf_event_header header; + __u32 pid, tid; + __u64 start; + __u64 len; + __u64 pgoff; + char filename[PATH_MAX]; +}; + +struct comm_event { + struct perf_event_header header; + __u32 pid, tid; + char comm[16]; +}; + +struct fork_event { + struct perf_event_header header; + __u32 pid, ppid; +}; + +struct period_event { + struct perf_event_header header; + __u64 time; + __u64 id; + __u64 sample_period; +}; + +typedef union event_union { + struct perf_event_header header; + struct ip_event ip; + struct mmap_event mmap; + struct comm_event comm; + struct fork_event fork; + struct period_event period; +} event_t; + +static LIST_HEAD(dsos); +static struct dso *kernel_dso; +static struct dso *vdso; + + +static void dsos__add(struct dso *dso) +{ + list_add_tail(&dso->node, &dsos); +} + +static struct dso *dsos__find(const char *name) +{ + struct dso *pos; + + list_for_each_entry(pos, &dsos, node) + if (strcmp(pos->name, name) == 0) + return pos; + return NULL; +} + +static struct dso *dsos__findnew(const char *name) +{ + struct dso *dso = dsos__find(name); + int nr; + + if (dso) + return dso; + + dso = dso__new(name, 0); + if (!dso) + goto out_delete_dso; + + nr = dso__load(dso, NULL, verbose); + if (nr < 0) { + if (verbose) + fprintf(stderr, "Failed to open: %s\n", name); + goto out_delete_dso; + } + if (!nr && verbose) { + fprintf(stderr, + "No symbols found in: %s, maybe install a debug package?\n", + name); + } + + dsos__add(dso); + + return dso; + +out_delete_dso: + dso__delete(dso); + return NULL; +} + +static void dsos__fprintf(FILE *fp) +{ + struct dso *pos; + + list_for_each_entry(pos, &dsos, node) + dso__fprintf(pos, fp); +} + +static struct symbol *vdso__find_symbol(struct dso *dso, __u64 ip) +{ + return dso__find_symbol(kernel_dso, ip); +} + +static int load_kernel(void) +{ + int err; + + kernel_dso = dso__new("[kernel]", 0); + if (!kernel_dso) + return -1; + + err = dso__load_kernel(kernel_dso, vmlinux, NULL, verbose); + if (err) { + dso__delete(kernel_dso); + kernel_dso = NULL; + } else + dsos__add(kernel_dso); + + vdso = dso__new("[vdso]", 0); + if (!vdso) + return -1; + + vdso->find_symbol = vdso__find_symbol; + + dsos__add(vdso); + + return err; +} + +struct map { + struct list_head node; + __u64 start; + __u64 end; + __u64 pgoff; + __u64 (*map_ip)(struct map *, __u64); + struct dso *dso; +}; + +static __u64 map__map_ip(struct map *map, __u64 ip) +{ + return ip - map->start + map->pgoff; +} + +static __u64 vdso__map_ip(struct map *map, __u64 ip) +{ + return ip; +} + +static struct map *map__new(struct mmap_event *event) +{ + struct map *self = malloc(sizeof(*self)); + + if (self != NULL) { + const char *filename = event->filename; + + self->start = event->start; + self->end = event->start + event->len; + self->pgoff = event->pgoff; + + self->dso = dsos__findnew(filename); + if (self->dso == NULL) + goto out_delete; + + if (self->dso == vdso) + self->map_ip = vdso__map_ip; + else + self->map_ip = map__map_ip; + } + return self; +out_delete: + free(self); + return NULL; +} + +static struct map *map__clone(struct map *self) +{ + struct map *map = malloc(sizeof(*self)); + + if (!map) + return NULL; + + memcpy(map, self, sizeof(*self)); + + return map; +} + +static int map__overlap(struct map *l, struct map *r) +{ + if (l->start > r->start) { + struct map *t = l; + l = r; + r = t; + } + + if (l->end > r->start) + return 1; + + return 0; +} + +static size_t map__fprintf(struct map *self, FILE *fp) +{ + return fprintf(fp, " %Lx-%Lx %Lx %s\n", + self->start, self->end, self->pgoff, self->dso->name); +} + + +struct thread { + struct rb_node rb_node; + struct list_head maps; + pid_t pid; + char *comm; +}; + +static struct thread *thread__new(pid_t pid) +{ + struct thread *self = malloc(sizeof(*self)); + + if (self != NULL) { + self->pid = pid; + self->comm = malloc(32); + if (self->comm) + snprintf(self->comm, 32, ":%d", self->pid); + INIT_LIST_HEAD(&self->maps); + } + + return self; +} + +static int thread__set_comm(struct thread *self, const char *comm) +{ + if (self->comm) + free(self->comm); + self->comm = strdup(comm); + return self->comm ? 0 : -ENOMEM; +} + +static size_t thread__fprintf(struct thread *self, FILE *fp) +{ + struct map *pos; + size_t ret = fprintf(fp, "Thread %d %s\n", self->pid, self->comm); + + list_for_each_entry(pos, &self->maps, node) + ret += map__fprintf(pos, fp); + + return ret; +} + + +static struct rb_root threads; +static struct thread *last_match; + +static struct thread *threads__findnew(pid_t pid) +{ + struct rb_node **p = &threads.rb_node; + struct rb_node *parent = NULL; + struct thread *th; + + /* + * Font-end cache - PID lookups come in blocks, + * so most of the time we dont have to look up + * the full rbtree: + */ + if (last_match && last_match->pid == pid) + return last_match; + + while (*p != NULL) { + parent = *p; + th = rb_entry(parent, struct thread, rb_node); + + if (th->pid == pid) { + last_match = th; + return th; + } + + if (pid < th->pid) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + th = thread__new(pid); + if (th != NULL) { + rb_link_node(&th->rb_node, parent, p); + rb_insert_color(&th->rb_node, &threads); + last_match = th; + } + + return th; +} + +static void thread__insert_map(struct thread *self, struct map *map) +{ + struct map *pos, *tmp; + + list_for_each_entry_safe(pos, tmp, &self->maps, node) { + if (map__overlap(pos, map)) { + list_del_init(&pos->node); + /* XXX leaks dsos */ + free(pos); + } + } + + list_add_tail(&map->node, &self->maps); +} + +static int thread__fork(struct thread *self, struct thread *parent) +{ + struct map *map; + + if (self->comm) + free(self->comm); + self->comm = strdup(parent->comm); + if (!self->comm) + return -ENOMEM; + + list_for_each_entry(map, &parent->maps, node) { + struct map *new = map__clone(map); + if (!new) + return -ENOMEM; + thread__insert_map(self, new); + } + + return 0; +} + +static struct map *thread__find_map(struct thread *self, __u64 ip) +{ + struct map *pos; + + if (self == NULL) + return NULL; + + list_for_each_entry(pos, &self->maps, node) + if (ip >= pos->start && ip <= pos->end) + return pos; + + return NULL; +} + +static size_t threads__fprintf(FILE *fp) +{ + size_t ret = 0; + struct rb_node *nd; + + for (nd = rb_first(&threads); nd; nd = rb_next(nd)) { + struct thread *pos = rb_entry(nd, struct thread, rb_node); + + ret += thread__fprintf(pos, fp); + } + + return ret; +} + +/* + * histogram, sorted on item, collects counts + */ + +static struct rb_root hist; + +struct hist_entry { + struct rb_node rb_node; + + struct thread *thread; + struct map *map; + struct dso *dso; + struct symbol *sym; + __u64 ip; + char level; + + uint32_t count; +}; + +/* + * configurable sorting bits + */ + +struct sort_entry { + struct list_head list; + + char *header; + + int64_t (*cmp)(struct hist_entry *, struct hist_entry *); + int64_t (*collapse)(struct hist_entry *, struct hist_entry *); + size_t (*print)(FILE *fp, struct hist_entry *); +}; + +/* --sort pid */ + +static int64_t +sort__thread_cmp(struct hist_entry *left, struct hist_entry *right) +{ + return right->thread->pid - left->thread->pid; +} + +static size_t +sort__thread_print(FILE *fp, struct hist_entry *self) +{ + return fprintf(fp, "%16s:%5d", self->thread->comm ?: "", self->thread->pid); +} + +static struct sort_entry sort_thread = { + .header = " Command: Pid", + .cmp = sort__thread_cmp, + .print = sort__thread_print, +}; + +/* --sort comm */ + +static int64_t +sort__comm_cmp(struct hist_entry *left, struct hist_entry *right) +{ + return right->thread->pid - left->thread->pid; +} + +static int64_t +sort__comm_collapse(struct hist_entry *left, struct hist_entry *right) +{ + char *comm_l = left->thread->comm; + char *comm_r = right->thread->comm; + + if (!comm_l || !comm_r) { + if (!comm_l && !comm_r) + return 0; + else if (!comm_l) + return -1; + else + return 1; + } + + return strcmp(comm_l, comm_r); +} + +static size_t +sort__comm_print(FILE *fp, struct hist_entry *self) +{ + return fprintf(fp, "%16s", self->thread->comm); +} + +static struct sort_entry sort_comm = { + .header = " Command", + .cmp = sort__comm_cmp, + .collapse = sort__comm_collapse, + .print = sort__comm_print, +}; + +/* --sort dso */ + +static int64_t +sort__dso_cmp(struct hist_entry *left, struct hist_entry *right) +{ + struct dso *dso_l = left->dso; + struct dso *dso_r = right->dso; + + if (!dso_l || !dso_r) { + if (!dso_l && !dso_r) + return 0; + else if (!dso_l) + return -1; + else + return 1; + } + + return strcmp(dso_l->name, dso_r->name); +} + +static size_t +sort__dso_print(FILE *fp, struct hist_entry *self) +{ + if (self->dso) + return fprintf(fp, "%-25s", self->dso->name); + + return fprintf(fp, "%016llx ", (__u64)self->ip); +} + +static struct sort_entry sort_dso = { + .header = "Shared Object ", + .cmp = sort__dso_cmp, + .print = sort__dso_print, +}; + +/* --sort symbol */ + +static int64_t +sort__sym_cmp(struct hist_entry *left, struct hist_entry *right) +{ + __u64 ip_l, ip_r; + + if (left->sym == right->sym) + return 0; + + ip_l = left->sym ? left->sym->start : left->ip; + ip_r = right->sym ? right->sym->start : right->ip; + + return (int64_t)(ip_r - ip_l); +} + +static size_t +sort__sym_print(FILE *fp, struct hist_entry *self) +{ + size_t ret = 0; + + if (verbose) + ret += fprintf(fp, "%#018llx ", (__u64)self->ip); + + if (self->sym) { + ret += fprintf(fp, "[%c] %s", + self->dso == kernel_dso ? 'k' : '.', self->sym->name); + } else { + ret += fprintf(fp, "%#016llx", (__u64)self->ip); + } + + return ret; +} + +static struct sort_entry sort_sym = { + .header = "Symbol", + .cmp = sort__sym_cmp, + .print = sort__sym_print, +}; + +static int sort__need_collapse = 0; + +struct sort_dimension { + char *name; + struct sort_entry *entry; + int taken; +}; + +static struct sort_dimension sort_dimensions[] = { + { .name = "pid", .entry = &sort_thread, }, + { .name = "comm", .entry = &sort_comm, }, + { .name = "dso", .entry = &sort_dso, }, + { .name = "symbol", .entry = &sort_sym, }, +}; + +static LIST_HEAD(hist_entry__sort_list); + +static int sort_dimension__add(char *tok) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(sort_dimensions); i++) { + struct sort_dimension *sd = &sort_dimensions[i]; + + if (sd->taken) + continue; + + if (strncasecmp(tok, sd->name, strlen(tok))) + continue; + + if (sd->entry->collapse) + sort__need_collapse = 1; + + list_add_tail(&sd->entry->list, &hist_entry__sort_list); + sd->taken = 1; + + return 0; + } + + return -ESRCH; +} + +static int64_t +hist_entry__cmp(struct hist_entry *left, struct hist_entry *right) +{ + struct sort_entry *se; + int64_t cmp = 0; + + list_for_each_entry(se, &hist_entry__sort_list, list) { + cmp = se->cmp(left, right); + if (cmp) + break; + } + + return cmp; +} + +static int64_t +hist_entry__collapse(struct hist_entry *left, struct hist_entry *right) +{ + struct sort_entry *se; + int64_t cmp = 0; + + list_for_each_entry(se, &hist_entry__sort_list, list) { + int64_t (*f)(struct hist_entry *, struct hist_entry *); + + f = se->collapse ?: se->cmp; + + cmp = f(left, right); + if (cmp) + break; + } + + return cmp; +} + +/* + * collect histogram counts + */ +static void hist_hit(struct hist_entry *he, __u64 ip) +{ + unsigned int sym_size, offset; + struct symbol *sym = he->sym; + + he->count++; + + if (!sym || !sym->hist) + return; + + sym_size = sym->end - sym->start; + offset = ip - sym->start; + + if (offset >= sym_size) + return; + + sym->hist_sum++; + sym->hist[offset]++; + + if (verbose >= 3) + printf("%p %s: count++ [ip: %p, %08Lx] => %Ld\n", + (void *)(unsigned long)he->sym->start, + he->sym->name, + (void *)(unsigned long)ip, ip - he->sym->start, + sym->hist[offset]); +} + +static int +hist_entry__add(struct thread *thread, struct map *map, struct dso *dso, + struct symbol *sym, __u64 ip, char level) +{ + struct rb_node **p = &hist.rb_node; + struct rb_node *parent = NULL; + struct hist_entry *he; + struct hist_entry entry = { + .thread = thread, + .map = map, + .dso = dso, + .sym = sym, + .ip = ip, + .level = level, + .count = 1, + }; + int cmp; + + while (*p != NULL) { + parent = *p; + he = rb_entry(parent, struct hist_entry, rb_node); + + cmp = hist_entry__cmp(&entry, he); + + if (!cmp) { + hist_hit(he, ip); + + return 0; + } + + if (cmp < 0) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + he = malloc(sizeof(*he)); + if (!he) + return -ENOMEM; + *he = entry; + rb_link_node(&he->rb_node, parent, p); + rb_insert_color(&he->rb_node, &hist); + + return 0; +} + +static void hist_entry__free(struct hist_entry *he) +{ + free(he); +} + +/* + * collapse the histogram + */ + +static struct rb_root collapse_hists; + +static void collapse__insert_entry(struct hist_entry *he) +{ + struct rb_node **p = &collapse_hists.rb_node; + struct rb_node *parent = NULL; + struct hist_entry *iter; + int64_t cmp; + + while (*p != NULL) { + parent = *p; + iter = rb_entry(parent, struct hist_entry, rb_node); + + cmp = hist_entry__collapse(iter, he); + + if (!cmp) { + iter->count += he->count; + hist_entry__free(he); + return; + } + + if (cmp < 0) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + rb_link_node(&he->rb_node, parent, p); + rb_insert_color(&he->rb_node, &collapse_hists); +} + +static void collapse__resort(void) +{ + struct rb_node *next; + struct hist_entry *n; + + if (!sort__need_collapse) + return; + + next = rb_first(&hist); + while (next) { + n = rb_entry(next, struct hist_entry, rb_node); + next = rb_next(&n->rb_node); + + rb_erase(&n->rb_node, &hist); + collapse__insert_entry(n); + } +} + +/* + * reverse the map, sort on count. + */ + +static struct rb_root output_hists; + +static void output__insert_entry(struct hist_entry *he) +{ + struct rb_node **p = &output_hists.rb_node; + struct rb_node *parent = NULL; + struct hist_entry *iter; + + while (*p != NULL) { + parent = *p; + iter = rb_entry(parent, struct hist_entry, rb_node); + + if (he->count > iter->count) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + rb_link_node(&he->rb_node, parent, p); + rb_insert_color(&he->rb_node, &output_hists); +} + +static void output__resort(void) +{ + struct rb_node *next; + struct hist_entry *n; + struct rb_root *tree = &hist; + + if (sort__need_collapse) + tree = &collapse_hists; + + next = rb_first(tree); + + while (next) { + n = rb_entry(next, struct hist_entry, rb_node); + next = rb_next(&n->rb_node); + + rb_erase(&n->rb_node, tree); + output__insert_entry(n); + } +} + +static void register_idle_thread(void) +{ + struct thread *thread = threads__findnew(0); + + if (thread == NULL || + thread__set_comm(thread, "[idle]")) { + fprintf(stderr, "problem inserting idle task.\n"); + exit(-1); + } +} + +static unsigned long total = 0, + total_mmap = 0, + total_comm = 0, + total_fork = 0, + total_unknown = 0; + +static int +process_overflow_event(event_t *event, unsigned long offset, unsigned long head) +{ + char level; + int show = 0; + struct dso *dso = NULL; + struct thread *thread = threads__findnew(event->ip.pid); + __u64 ip = event->ip.ip; + struct map *map = NULL; + + dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->header.misc, + event->ip.pid, + (void *)(long)ip); + + dprintf(" ... thread: %s:%d\n", thread->comm, thread->pid); + + if (thread == NULL) { + fprintf(stderr, "problem processing %d event, skipping it.\n", + event->header.type); + return -1; + } + + if (event->header.misc & PERF_EVENT_MISC_KERNEL) { + show = SHOW_KERNEL; + level = 'k'; + + dso = kernel_dso; + + dprintf(" ...... dso: %s\n", dso->name); + + } else if (event->header.misc & PERF_EVENT_MISC_USER) { + + show = SHOW_USER; + level = '.'; + + map = thread__find_map(thread, ip); + if (map != NULL) { + ip = map->map_ip(map, ip); + dso = map->dso; + } else { + /* + * If this is outside of all known maps, + * and is a negative address, try to look it + * up in the kernel dso, as it might be a + * vsyscall (which executes in user-mode): + */ + if ((long long)ip < 0) + dso = kernel_dso; + } + dprintf(" ...... dso: %s\n", dso ? dso->name : "<not found>"); + + } else { + show = SHOW_HV; + level = 'H'; + dprintf(" ...... dso: [hypervisor]\n"); + } + + if (show & show_mask) { + struct symbol *sym = NULL; + + if (dso) + sym = dso->find_symbol(dso, ip); + + if (hist_entry__add(thread, map, dso, sym, ip, level)) { + fprintf(stderr, + "problem incrementing symbol count, skipping event\n"); + return -1; + } + } + total++; + + return 0; +} + +static int +process_mmap_event(event_t *event, unsigned long offset, unsigned long head) +{ + struct thread *thread = threads__findnew(event->mmap.pid); + struct map *map = map__new(&event->mmap); + + dprintf("%p [%p]: PERF_EVENT_MMAP %d: [%p(%p) @ %p]: %s\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->mmap.pid, + (void *)(long)event->mmap.start, + (void *)(long)event->mmap.len, + (void *)(long)event->mmap.pgoff, + event->mmap.filename); + + if (thread == NULL || map == NULL) { + dprintf("problem processing PERF_EVENT_MMAP, skipping event.\n"); + return 0; + } + + thread__insert_map(thread, map); + total_mmap++; + + return 0; +} + +static int +process_comm_event(event_t *event, unsigned long offset, unsigned long head) +{ + struct thread *thread = threads__findnew(event->comm.pid); + + dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->comm.comm, event->comm.pid); + + if (thread == NULL || + thread__set_comm(thread, event->comm.comm)) { + dprintf("problem processing PERF_EVENT_COMM, skipping event.\n"); + return -1; + } + total_comm++; + + return 0; +} + +static int +process_fork_event(event_t *event, unsigned long offset, unsigned long head) +{ + struct thread *thread = threads__findnew(event->fork.pid); + struct thread *parent = threads__findnew(event->fork.ppid); + + dprintf("%p [%p]: PERF_EVENT_FORK: %d:%d\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->fork.pid, event->fork.ppid); + + if (!thread || !parent || thread__fork(thread, parent)) { + dprintf("problem processing PERF_EVENT_FORK, skipping event.\n"); + return -1; + } + total_fork++; + + return 0; +} + +static int +process_period_event(event_t *event, unsigned long offset, unsigned long head) +{ + dprintf("%p [%p]: PERF_EVENT_PERIOD: time:%Ld, id:%Ld: period:%Ld\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->period.time, + event->period.id, + event->period.sample_period); + + return 0; +} + +static int +process_event(event_t *event, unsigned long offset, unsigned long head) +{ + if (event->header.misc & PERF_EVENT_MISC_OVERFLOW) + return process_overflow_event(event, offset, head); + + switch (event->header.type) { + case PERF_EVENT_MMAP: + return process_mmap_event(event, offset, head); + + case PERF_EVENT_COMM: + return process_comm_event(event, offset, head); + + case PERF_EVENT_FORK: + return process_fork_event(event, offset, head); + + case PERF_EVENT_PERIOD: + return process_period_event(event, offset, head); + /* + * We dont process them right now but they are fine: + */ + + case PERF_EVENT_THROTTLE: + case PERF_EVENT_UNTHROTTLE: + return 0; + + default: + return -1; + } + + return 0; +} + +static int +parse_line(FILE *file, struct symbol *sym, __u64 start, __u64 len) +{ + char *line = NULL, *tmp, *tmp2; + unsigned int offset; + size_t line_len; + __u64 line_ip; + int ret; + char *c; + + if (getline(&line, &line_len, file) < 0) + return -1; + if (!line) + return -1; + + c = strchr(line, '\n'); + if (c) + *c = 0; + + line_ip = -1; + offset = 0; + ret = -2; + + /* + * Strip leading spaces: + */ + tmp = line; + while (*tmp) { + if (*tmp != ' ') + break; + tmp++; + } + + if (*tmp) { + /* + * Parse hexa addresses followed by ':' + */ + line_ip = strtoull(tmp, &tmp2, 16); + if (*tmp2 != ':') + line_ip = -1; + } + + if (line_ip != -1) { + unsigned int hits = 0; + double percent = 0.0; + char *color = PERF_COLOR_NORMAL; + + offset = line_ip - start; + if (offset < len) + hits = sym->hist[offset]; + + if (sym->hist_sum) + percent = 100.0 * hits / sym->hist_sum; + + /* + * We color high-overhead entries in red, mid-overhead + * entries in green - and keep the low overhead places + * normal: + */ + if (percent >= 5.0) + color = PERF_COLOR_RED; + else { + if (percent > 0.5) + color = PERF_COLOR_GREEN; + } + + color_fprintf(stdout, color, " %7.2f", percent); + printf(" : "); + color_fprintf(stdout, PERF_COLOR_BLUE, "%s\n", line); + } else { + if (!*line) + printf(" :\n"); + else + printf(" : %s\n", line); + } + + return 0; +} + +static void annotate_sym(struct dso *dso, struct symbol *sym) +{ + char *filename = dso->name; + __u64 start, end, len; + char command[PATH_MAX*2]; + FILE *file; + + if (!filename) + return; + if (dso == kernel_dso) + filename = vmlinux; + + printf("\n------------------------------------------------\n"); + printf(" Percent | Source code & Disassembly of %s\n", filename); + printf("------------------------------------------------\n"); + + if (verbose >= 2) + printf("annotating [%p] %30s : [%p] %30s\n", dso, dso->name, sym, sym->name); + + start = sym->obj_start; + if (!start) + start = sym->start; + + end = start + sym->end - sym->start + 1; + len = sym->end - sym->start; + + sprintf(command, "objdump --start-address=0x%016Lx --stop-address=0x%016Lx -dS %s", (__u64)start, (__u64)end, filename); + + if (verbose >= 3) + printf("doing: %s\n", command); + + file = popen(command, "r"); + if (!file) + return; + + while (!feof(file)) { + if (parse_line(file, sym, start, len) < 0) + break; + } + + pclose(file); +} + +static void find_annotations(void) +{ + struct rb_node *nd; + struct dso *dso; + int count = 0; + + list_for_each_entry(dso, &dsos, node) { + + for (nd = rb_first(&dso->syms); nd; nd = rb_next(nd)) { + struct symbol *sym = rb_entry(nd, struct symbol, rb_node); + + if (sym->hist) { + annotate_sym(dso, sym); + count++; + } + } + } + + if (!count) + printf(" Error: symbol '%s' not present amongst the samples.\n", sym_hist_filter); +} + +static int __cmd_annotate(void) +{ + int ret, rc = EXIT_FAILURE; + unsigned long offset = 0; + unsigned long head = 0; + struct stat stat; + event_t *event; + uint32_t size; + char *buf; + + register_idle_thread(); + + input = open(input_name, O_RDONLY); + if (input < 0) { + perror("failed to open file"); + exit(-1); + } + + ret = fstat(input, &stat); + if (ret < 0) { + perror("failed to stat file"); + exit(-1); + } + + if (!stat.st_size) { + fprintf(stderr, "zero-sized file, nothing to do!\n"); + exit(0); + } + + if (load_kernel() < 0) { + perror("failed to load kernel symbols"); + return EXIT_FAILURE; + } + +remap: + buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ, + MAP_SHARED, input, offset); + if (buf == MAP_FAILED) { + perror("failed to mmap file"); + exit(-1); + } + +more: + event = (event_t *)(buf + head); + + size = event->header.size; + if (!size) + size = 8; + + if (head + event->header.size >= page_size * mmap_window) { + unsigned long shift = page_size * (head / page_size); + int ret; + + ret = munmap(buf, page_size * mmap_window); + assert(ret == 0); + + offset += shift; + head -= shift; + goto remap; + } + + size = event->header.size; + + dprintf("%p [%p]: event: %d\n", + (void *)(offset + head), + (void *)(long)event->header.size, + event->header.type); + + if (!size || process_event(event, offset, head) < 0) { + + dprintf("%p [%p]: skipping unknown header type: %d\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->header.type); + + total_unknown++; + + /* + * assume we lost track of the stream, check alignment, and + * increment a single u64 in the hope to catch on again 'soon'. + */ + + if (unlikely(head & 7)) + head &= ~7ULL; + + size = 8; + } + + head += size; + + if (offset + head < stat.st_size) + goto more; + + rc = EXIT_SUCCESS; + close(input); + + dprintf(" IP events: %10ld\n", total); + dprintf(" mmap events: %10ld\n", total_mmap); + dprintf(" comm events: %10ld\n", total_comm); + dprintf(" fork events: %10ld\n", total_fork); + dprintf(" unknown events: %10ld\n", total_unknown); + + if (dump_trace) + return 0; + + if (verbose >= 3) + threads__fprintf(stdout); + + if (verbose >= 2) + dsos__fprintf(stdout); + + collapse__resort(); + output__resort(); + + find_annotations(); + + return rc; +} + +static const char * const annotate_usage[] = { + "perf annotate [<options>] <command>", + NULL +}; + +static const struct option options[] = { + OPT_STRING('i', "input", &input_name, "file", + "input file name"), + OPT_STRING('s', "symbol", &sym_hist_filter, "symbol", + "symbol to annotate"), + OPT_BOOLEAN('v', "verbose", &verbose, + "be more verbose (show symbol address, etc)"), + OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace, + "dump raw trace in ASCII"), + OPT_STRING('k', "vmlinux", &vmlinux, "file", "vmlinux pathname"), + OPT_END() +}; + +static void setup_sorting(void) +{ + char *tmp, *tok, *str = strdup(sort_order); + + for (tok = strtok_r(str, ", ", &tmp); + tok; tok = strtok_r(NULL, ", ", &tmp)) { + if (sort_dimension__add(tok) < 0) { + error("Unknown --sort key: `%s'", tok); + usage_with_options(annotate_usage, options); + } + } + + free(str); +} + +int cmd_annotate(int argc, const char **argv, const char *prefix) +{ + symbol__init(); + + page_size = getpagesize(); + + argc = parse_options(argc, argv, options, annotate_usage, 0); + + setup_sorting(); + + if (argc) { + /* + * Special case: if there's an argument left then assume tha + * it's a symbol filter: + */ + if (argc > 1) + usage_with_options(annotate_usage, options); + + sym_hist_filter = argv[0]; + } + + if (!sym_hist_filter) + usage_with_options(annotate_usage, options); + + setup_pager(); + + return __cmd_annotate(); +} diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c new file mode 100644 index 00000000000..0f32dc3f3c4 --- /dev/null +++ b/tools/perf/builtin-help.c @@ -0,0 +1,461 @@ +/* + * builtin-help.c + * + * Builtin help command + */ +#include "util/cache.h" +#include "builtin.h" +#include "util/exec_cmd.h" +#include "common-cmds.h" +#include "util/parse-options.h" +#include "util/run-command.h" +#include "util/help.h" + +static struct man_viewer_list { + struct man_viewer_list *next; + char name[FLEX_ARRAY]; +} *man_viewer_list; + +static struct man_viewer_info_list { + struct man_viewer_info_list *next; + const char *info; + char name[FLEX_ARRAY]; +} *man_viewer_info_list; + +enum help_format { + HELP_FORMAT_MAN, + HELP_FORMAT_INFO, + HELP_FORMAT_WEB, +}; + +static int show_all = 0; +static enum help_format help_format = HELP_FORMAT_MAN; +static struct option builtin_help_options[] = { + OPT_BOOLEAN('a', "all", &show_all, "print all available commands"), + OPT_SET_INT('m', "man", &help_format, "show man page", HELP_FORMAT_MAN), + OPT_SET_INT('w', "web", &help_format, "show manual in web browser", + HELP_FORMAT_WEB), + OPT_SET_INT('i', "info", &help_format, "show info page", + HELP_FORMAT_INFO), + OPT_END(), +}; + +static const char * const builtin_help_usage[] = { + "perf help [--all] [--man|--web|--info] [command]", + NULL +}; + +static enum help_format parse_help_format(const char *format) +{ + if (!strcmp(format, "man")) + return HELP_FORMAT_MAN; + if (!strcmp(format, "info")) + return HELP_FORMAT_INFO; + if (!strcmp(format, "web") || !strcmp(format, "html")) + return HELP_FORMAT_WEB; + die("unrecognized help format '%s'", format); +} + +static const char *get_man_viewer_info(const char *name) +{ + struct man_viewer_info_list *viewer; + + for (viewer = man_viewer_info_list; viewer; viewer = viewer->next) + { + if (!strcasecmp(name, viewer->name)) + return viewer->info; + } + return NULL; +} + +static int check_emacsclient_version(void) +{ + struct strbuf buffer = STRBUF_INIT; + struct child_process ec_process; + const char *argv_ec[] = { "emacsclient", "--version", NULL }; + int version; + + /* emacsclient prints its version number on stderr */ + memset(&ec_process, 0, sizeof(ec_process)); + ec_process.argv = argv_ec; + ec_process.err = -1; + ec_process.stdout_to_stderr = 1; + if (start_command(&ec_process)) { + fprintf(stderr, "Failed to start emacsclient.\n"); + return -1; + } + strbuf_read(&buffer, ec_process.err, 20); + close(ec_process.err); + + /* + * Don't bother checking return value, because "emacsclient --version" + * seems to always exits with code 1. + */ + finish_command(&ec_process); + + if (prefixcmp(buffer.buf, "emacsclient")) { + fprintf(stderr, "Failed to parse emacsclient version.\n"); + strbuf_release(&buffer); + return -1; + } + + strbuf_remove(&buffer, 0, strlen("emacsclient")); + version = atoi(buffer.buf); + + if (version < 22) { + fprintf(stderr, + "emacsclient version '%d' too old (< 22).\n", + version); + strbuf_release(&buffer); + return -1; + } + + strbuf_release(&buffer); + return 0; +} + +static void exec_woman_emacs(const char* path, const char *page) +{ + if (!check_emacsclient_version()) { + /* This works only with emacsclient version >= 22. */ + struct strbuf man_page = STRBUF_INIT; + + if (!path) + path = "emacsclient"; + strbuf_addf(&man_page, "(woman \"%s\")", page); + execlp(path, "emacsclient", "-e", man_page.buf, NULL); + warning("failed to exec '%s': %s", path, strerror(errno)); + } +} + +static void exec_man_konqueror(const char* path, const char *page) +{ + const char *display = getenv("DISPLAY"); + if (display && *display) { + struct strbuf man_page = STRBUF_INIT; + const char *filename = "kfmclient"; + + /* It's simpler to launch konqueror using kfmclient. */ + if (path) { + const char *file = strrchr(path, '/'); + if (file && !strcmp(file + 1, "konqueror")) { + char *new = strdup(path); + char *dest = strrchr(new, '/'); + + /* strlen("konqueror") == strlen("kfmclient") */ + strcpy(dest + 1, "kfmclient"); + path = new; + } + if (file) + filename = file; + } else + path = "kfmclient"; + strbuf_addf(&man_page, "man:%s(1)", page); + execlp(path, filename, "newTab", man_page.buf, NULL); + warning("failed to exec '%s': %s", path, strerror(errno)); + } +} + +static void exec_man_man(const char* path, const char *page) +{ + if (!path) + path = "man"; + execlp(path, "man", page, NULL); + warning("failed to exec '%s': %s", path, strerror(errno)); +} + +static void exec_man_cmd(const char *cmd, const char *page) +{ + struct strbuf shell_cmd = STRBUF_INIT; + strbuf_addf(&shell_cmd, "%s %s", cmd, page); + execl("/bin/sh", "sh", "-c", shell_cmd.buf, NULL); + warning("failed to exec '%s': %s", cmd, strerror(errno)); +} + +static void add_man_viewer(const char *name) +{ + struct man_viewer_list **p = &man_viewer_list; + size_t len = strlen(name); + + while (*p) + p = &((*p)->next); + *p = calloc(1, (sizeof(**p) + len + 1)); + strncpy((*p)->name, name, len); +} + +static int supported_man_viewer(const char *name, size_t len) +{ + return (!strncasecmp("man", name, len) || + !strncasecmp("woman", name, len) || + !strncasecmp("konqueror", name, len)); +} + +static void do_add_man_viewer_info(const char *name, + size_t len, + const char *value) +{ + struct man_viewer_info_list *new = calloc(1, sizeof(*new) + len + 1); + + strncpy(new->name, name, len); + new->info = strdup(value); + new->next = man_viewer_info_list; + man_viewer_info_list = new; +} + +static int add_man_viewer_path(const char *name, + size_t len, + const char *value) +{ + if (supported_man_viewer(name, len)) + do_add_man_viewer_info(name, len, value); + else + warning("'%s': path for unsupported man viewer.\n" + "Please consider using 'man.<tool>.cmd' instead.", + name); + + return 0; +} + +static int add_man_viewer_cmd(const char *name, + size_t len, + const char *value) +{ + if (supported_man_viewer(name, len)) + warning("'%s': cmd for supported man viewer.\n" + "Please consider using 'man.<tool>.path' instead.", + name); + else + do_add_man_viewer_info(name, len, value); + + return 0; +} + +static int add_man_viewer_info(const char *var, const char *value) +{ + const char *name = var + 4; + const char *subkey = strrchr(name, '.'); + + if (!subkey) + return error("Config with no key for man viewer: %s", name); + + if (!strcmp(subkey, ".path")) { + if (!value) + return config_error_nonbool(var); + return add_man_viewer_path(name, subkey - name, value); + } + if (!strcmp(subkey, ".cmd")) { + if (!value) + return config_error_nonbool(var); + return add_man_viewer_cmd(name, subkey - name, value); + } + + warning("'%s': unsupported man viewer sub key.", subkey); + return 0; +} + +static int perf_help_config(const char *var, const char *value, void *cb) +{ + if (!strcmp(var, "help.format")) { + if (!value) + return config_error_nonbool(var); + help_format = parse_help_format(value); + return 0; + } + if (!strcmp(var, "man.viewer")) { + if (!value) + return config_error_nonbool(var); + add_man_viewer(value); + return 0; + } + if (!prefixcmp(var, "man.")) + return add_man_viewer_info(var, value); + + return perf_default_config(var, value, cb); +} + +static struct cmdnames main_cmds, other_cmds; + +void list_common_cmds_help(void) +{ + int i, longest = 0; + + for (i = 0; i < ARRAY_SIZE(common_cmds); i++) { + if (longest < strlen(common_cmds[i].name)) + longest = strlen(common_cmds[i].name); + } + + puts(" The most commonly used perf commands are:"); + for (i = 0; i < ARRAY_SIZE(common_cmds); i++) { + printf(" %s ", common_cmds[i].name); + mput_char(' ', longest - strlen(common_cmds[i].name)); + puts(common_cmds[i].help); + } +} + +static int is_perf_command(const char *s) +{ + return is_in_cmdlist(&main_cmds, s) || + is_in_cmdlist(&other_cmds, s); +} + +static const char *prepend(const char *prefix, const char *cmd) +{ + size_t pre_len = strlen(prefix); + size_t cmd_len = strlen(cmd); + char *p = malloc(pre_len + cmd_len + 1); + memcpy(p, prefix, pre_len); + strcpy(p + pre_len, cmd); + return p; +} + +static const char *cmd_to_page(const char *perf_cmd) +{ + if (!perf_cmd) + return "perf"; + else if (!prefixcmp(perf_cmd, "perf")) + return perf_cmd; + else if (is_perf_command(perf_cmd)) + return prepend("perf-", perf_cmd); + else + return prepend("perf-", perf_cmd); +} + +static void setup_man_path(void) +{ + struct strbuf new_path = STRBUF_INIT; + const char *old_path = getenv("MANPATH"); + + /* We should always put ':' after our path. If there is no + * old_path, the ':' at the end will let 'man' to try + * system-wide paths after ours to find the manual page. If + * there is old_path, we need ':' as delimiter. */ + strbuf_addstr(&new_path, system_path(PERF_MAN_PATH)); + strbuf_addch(&new_path, ':'); + if (old_path) + strbuf_addstr(&new_path, old_path); + + setenv("MANPATH", new_path.buf, 1); + + strbuf_release(&new_path); +} + +static void exec_viewer(const char *name, const char *page) +{ + const char *info = get_man_viewer_info(name); + + if (!strcasecmp(name, "man")) + exec_man_man(info, page); + else if (!strcasecmp(name, "woman")) + exec_woman_emacs(info, page); + else if (!strcasecmp(name, "konqueror")) + exec_man_konqueror(info, page); + else if (info) + exec_man_cmd(info, page); + else + warning("'%s': unknown man viewer.", name); +} + +static void show_man_page(const char *perf_cmd) +{ + struct man_viewer_list *viewer; + const char *page = cmd_to_page(perf_cmd); + const char *fallback = getenv("PERF_MAN_VIEWER"); + + setup_man_path(); + for (viewer = man_viewer_list; viewer; viewer = viewer->next) + { + exec_viewer(viewer->name, page); /* will return when unable */ + } + if (fallback) + exec_viewer(fallback, page); + exec_viewer("man", page); + die("no man viewer handled the request"); +} + +static void show_info_page(const char *perf_cmd) +{ + const char *page = cmd_to_page(perf_cmd); + setenv("INFOPATH", system_path(PERF_INFO_PATH), 1); + execlp("info", "info", "perfman", page, NULL); +} + +static void get_html_page_path(struct strbuf *page_path, const char *page) +{ + struct stat st; + const char *html_path = system_path(PERF_HTML_PATH); + + /* Check that we have a perf documentation directory. */ + if (stat(mkpath("%s/perf.html", html_path), &st) + || !S_ISREG(st.st_mode)) + die("'%s': not a documentation directory.", html_path); + + strbuf_init(page_path, 0); + strbuf_addf(page_path, "%s/%s.html", html_path, page); +} + +/* + * If open_html is not defined in a platform-specific way (see for + * example compat/mingw.h), we use the script web--browse to display + * HTML. + */ +#ifndef open_html +static void open_html(const char *path) +{ + execl_perf_cmd("web--browse", "-c", "help.browser", path, NULL); +} +#endif + +static void show_html_page(const char *perf_cmd) +{ + const char *page = cmd_to_page(perf_cmd); + struct strbuf page_path; /* it leaks but we exec bellow */ + + get_html_page_path(&page_path, page); + + open_html(page_path.buf); +} + +int cmd_help(int argc, const char **argv, const char *prefix) +{ + const char *alias; + load_command_list("perf-", &main_cmds, &other_cmds); + + perf_config(perf_help_config, NULL); + + argc = parse_options(argc, argv, builtin_help_options, + builtin_help_usage, 0); + + if (show_all) { + printf("\n usage: %s\n\n", perf_usage_string); + list_commands("perf commands", &main_cmds, &other_cmds); + printf(" %s\n\n", perf_more_info_string); + return 0; + } + + if (!argv[0]) { + printf("\n usage: %s\n\n", perf_usage_string); + list_common_cmds_help(); + printf("\n %s\n\n", perf_more_info_string); + return 0; + } + + alias = alias_lookup(argv[0]); + if (alias && !is_perf_command(argv[0])) { + printf("`perf %s' is aliased to `%s'\n", argv[0], alias); + return 0; + } + + switch (help_format) { + case HELP_FORMAT_MAN: + show_man_page(argv[0]); + break; + case HELP_FORMAT_INFO: + show_info_page(argv[0]); + break; + case HELP_FORMAT_WEB: + show_html_page(argv[0]); + break; + } + + return 0; +} diff --git a/tools/perf/builtin-list.c b/tools/perf/builtin-list.c new file mode 100644 index 00000000000..fe60e37c96e --- /dev/null +++ b/tools/perf/builtin-list.c @@ -0,0 +1,20 @@ +/* + * builtin-list.c + * + * Builtin list command: list all event types + * + * Copyright (C) 2009, Thomas Gleixner <tglx@linutronix.de> + * Copyright (C) 2008-2009, Red Hat Inc, Ingo Molnar <mingo@redhat.com> + */ +#include "builtin.h" + +#include "perf.h" + +#include "util/parse-options.h" +#include "util/parse-events.h" + +int cmd_list(int argc, const char **argv, const char *prefix) +{ + print_events(); + return 0; +} diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c new file mode 100644 index 00000000000..29259e74dcf --- /dev/null +++ b/tools/perf/builtin-record.c @@ -0,0 +1,582 @@ +/* + * builtin-record.c + * + * Builtin record command: Record the profile of a workload + * (or a CPU, or a PID) into the perf.data output file - for + * later analysis via perf report. + */ +#include "builtin.h" + +#include "perf.h" + +#include "util/util.h" +#include "util/parse-options.h" +#include "util/parse-events.h" +#include "util/string.h" + +#include <unistd.h> +#include <sched.h> + +#define ALIGN(x, a) __ALIGN_MASK(x, (typeof(x))(a)-1) +#define __ALIGN_MASK(x, mask) (((x)+(mask))&~(mask)) + +static int fd[MAX_NR_CPUS][MAX_COUNTERS]; + +static long default_interval = 100000; + +static int nr_cpus = 0; +static unsigned int page_size; +static unsigned int mmap_pages = 128; +static int freq = 0; +static int output; +static const char *output_name = "perf.data"; +static int group = 0; +static unsigned int realtime_prio = 0; +static int system_wide = 0; +static pid_t target_pid = -1; +static int inherit = 1; +static int force = 0; +static int append_file = 0; +static int verbose = 0; + +static long samples; +static struct timeval last_read; +static struct timeval this_read; + +static __u64 bytes_written; + +static struct pollfd event_array[MAX_NR_CPUS * MAX_COUNTERS]; + +static int nr_poll; +static int nr_cpu; + +struct mmap_event { + struct perf_event_header header; + __u32 pid; + __u32 tid; + __u64 start; + __u64 len; + __u64 pgoff; + char filename[PATH_MAX]; +}; + +struct comm_event { + struct perf_event_header header; + __u32 pid; + __u32 tid; + char comm[16]; +}; + + +struct mmap_data { + int counter; + void *base; + unsigned int mask; + unsigned int prev; +}; + +static struct mmap_data mmap_array[MAX_NR_CPUS][MAX_COUNTERS]; + +static unsigned int mmap_read_head(struct mmap_data *md) +{ + struct perf_counter_mmap_page *pc = md->base; + int head; + + head = pc->data_head; + rmb(); + + return head; +} + +static void mmap_read(struct mmap_data *md) +{ + unsigned int head = mmap_read_head(md); + unsigned int old = md->prev; + unsigned char *data = md->base + page_size; + unsigned long size; + void *buf; + int diff; + + gettimeofday(&this_read, NULL); + + /* + * If we're further behind than half the buffer, there's a chance + * the writer will bite our tail and mess up the samples under us. + * + * If we somehow ended up ahead of the head, we got messed up. + * + * In either case, truncate and restart at head. + */ + diff = head - old; + if (diff > md->mask / 2 || diff < 0) { + struct timeval iv; + unsigned long msecs; + + timersub(&this_read, &last_read, &iv); + msecs = iv.tv_sec*1000 + iv.tv_usec/1000; + + fprintf(stderr, "WARNING: failed to keep up with mmap data." + " Last read %lu msecs ago.\n", msecs); + + /* + * head points to a known good entry, start there. + */ + old = head; + } + + last_read = this_read; + + if (old != head) + samples++; + + size = head - old; + + if ((old & md->mask) + size != (head & md->mask)) { + buf = &data[old & md->mask]; + size = md->mask + 1 - (old & md->mask); + old += size; + + while (size) { + int ret = write(output, buf, size); + + if (ret < 0) + die("failed to write"); + + size -= ret; + buf += ret; + + bytes_written += ret; + } + } + + buf = &data[old & md->mask]; + size = head - old; + old += size; + + while (size) { + int ret = write(output, buf, size); + + if (ret < 0) + die("failed to write"); + + size -= ret; + buf += ret; + + bytes_written += ret; + } + + md->prev = old; +} + +static volatile int done = 0; +static volatile int signr = -1; + +static void sig_handler(int sig) +{ + done = 1; + signr = sig; +} + +static void sig_atexit(void) +{ + if (signr == -1) + return; + + signal(signr, SIG_DFL); + kill(getpid(), signr); +} + +static void pid_synthesize_comm_event(pid_t pid, int full) +{ + struct comm_event comm_ev; + char filename[PATH_MAX]; + char bf[BUFSIZ]; + int fd, ret; + size_t size; + char *field, *sep; + DIR *tasks; + struct dirent dirent, *next; + + snprintf(filename, sizeof(filename), "/proc/%d/stat", pid); + + fd = open(filename, O_RDONLY); + if (fd < 0) { + fprintf(stderr, "couldn't open %s\n", filename); + exit(EXIT_FAILURE); + } + if (read(fd, bf, sizeof(bf)) < 0) { + fprintf(stderr, "couldn't read %s\n", filename); + exit(EXIT_FAILURE); + } + close(fd); + + /* 9027 (cat) R 6747 9027 6747 34816 9027 ... */ + memset(&comm_ev, 0, sizeof(comm_ev)); + field = strchr(bf, '('); + if (field == NULL) + goto out_failure; + sep = strchr(++field, ')'); + if (sep == NULL) + goto out_failure; + size = sep - field; + memcpy(comm_ev.comm, field, size++); + + comm_ev.pid = pid; + comm_ev.header.type = PERF_EVENT_COMM; + size = ALIGN(size, sizeof(__u64)); + comm_ev.header.size = sizeof(comm_ev) - (sizeof(comm_ev.comm) - size); + + if (!full) { + comm_ev.tid = pid; + + ret = write(output, &comm_ev, comm_ev.header.size); + if (ret < 0) { + perror("failed to write"); + exit(-1); + } + return; + } + + snprintf(filename, sizeof(filename), "/proc/%d/task", pid); + + tasks = opendir(filename); + while (!readdir_r(tasks, &dirent, &next) && next) { + char *end; + pid = strtol(dirent.d_name, &end, 10); + if (*end) + continue; + + comm_ev.tid = pid; + + ret = write(output, &comm_ev, comm_ev.header.size); + if (ret < 0) { + perror("failed to write"); + exit(-1); + } + } + closedir(tasks); + return; + +out_failure: + fprintf(stderr, "couldn't get COMM and pgid, malformed %s\n", + filename); + exit(EXIT_FAILURE); +} + +static void pid_synthesize_mmap_samples(pid_t pid) +{ + char filename[PATH_MAX]; + FILE *fp; + + snprintf(filename, sizeof(filename), "/proc/%d/maps", pid); + + fp = fopen(filename, "r"); + if (fp == NULL) { + fprintf(stderr, "couldn't open %s\n", filename); + exit(EXIT_FAILURE); + } + while (1) { + char bf[BUFSIZ], *pbf = bf; + struct mmap_event mmap_ev = { + .header.type = PERF_EVENT_MMAP, + }; + int n; + size_t size; + if (fgets(bf, sizeof(bf), fp) == NULL) + break; + + /* 00400000-0040c000 r-xp 00000000 fd:01 41038 /bin/cat */ + n = hex2u64(pbf, &mmap_ev.start); + if (n < 0) + continue; + pbf += n + 1; + n = hex2u64(pbf, &mmap_ev.len); + if (n < 0) + continue; + pbf += n + 3; + if (*pbf == 'x') { /* vm_exec */ + char *execname = strrchr(bf, ' '); + + if (execname == NULL || execname[1] != '/') + continue; + + execname += 1; + size = strlen(execname); + execname[size - 1] = '\0'; /* Remove \n */ + memcpy(mmap_ev.filename, execname, size); + size = ALIGN(size, sizeof(__u64)); + mmap_ev.len -= mmap_ev.start; + mmap_ev.header.size = (sizeof(mmap_ev) - + (sizeof(mmap_ev.filename) - size)); + mmap_ev.pid = pid; + mmap_ev.tid = pid; + + if (write(output, &mmap_ev, mmap_ev.header.size) < 0) { + perror("failed to write"); + exit(-1); + } + } + } + + fclose(fp); +} + +static void synthesize_samples(void) +{ + DIR *proc; + struct dirent dirent, *next; + + proc = opendir("/proc"); + + while (!readdir_r(proc, &dirent, &next) && next) { + char *end; + pid_t pid; + + pid = strtol(dirent.d_name, &end, 10); + if (*end) /* only interested in proper numerical dirents */ + continue; + + pid_synthesize_comm_event(pid, 1); + pid_synthesize_mmap_samples(pid); + } + + closedir(proc); +} + +static int group_fd; + +static void create_counter(int counter, int cpu, pid_t pid) +{ + struct perf_counter_attr *attr = attrs + counter; + int track = 1; + + attr->sample_type = PERF_SAMPLE_IP | PERF_SAMPLE_TID; + if (freq) { + attr->sample_type |= PERF_SAMPLE_PERIOD; + attr->freq = 1; + attr->sample_freq = freq; + } + attr->mmap = track; + attr->comm = track; + attr->inherit = (cpu < 0) && inherit; + attr->disabled = 1; + + track = 0; /* only the first counter needs these */ + +try_again: + fd[nr_cpu][counter] = sys_perf_counter_open(attr, pid, cpu, group_fd, 0); + + if (fd[nr_cpu][counter] < 0) { + int err = errno; + + if (err == EPERM) + die("Permission error - are you root?\n"); + + /* + * If it's cycles then fall back to hrtimer + * based cpu-clock-tick sw counter, which + * is always available even if no PMU support: + */ + if (attr->type == PERF_TYPE_HARDWARE + && attr->config == PERF_COUNT_HW_CPU_CYCLES) { + + if (verbose) + warning(" ... trying to fall back to cpu-clock-ticks\n"); + attr->type = PERF_TYPE_SOFTWARE; + attr->config = PERF_COUNT_SW_CPU_CLOCK; + goto try_again; + } + printf("\n"); + error("perfcounter syscall returned with %d (%s)\n", + fd[nr_cpu][counter], strerror(err)); + die("No CONFIG_PERF_COUNTERS=y kernel support configured?\n"); + exit(-1); + } + + assert(fd[nr_cpu][counter] >= 0); + fcntl(fd[nr_cpu][counter], F_SETFL, O_NONBLOCK); + + /* + * First counter acts as the group leader: + */ + if (group && group_fd == -1) + group_fd = fd[nr_cpu][counter]; + + event_array[nr_poll].fd = fd[nr_cpu][counter]; + event_array[nr_poll].events = POLLIN; + nr_poll++; + + mmap_array[nr_cpu][counter].counter = counter; + mmap_array[nr_cpu][counter].prev = 0; + mmap_array[nr_cpu][counter].mask = mmap_pages*page_size - 1; + mmap_array[nr_cpu][counter].base = mmap(NULL, (mmap_pages+1)*page_size, + PROT_READ, MAP_SHARED, fd[nr_cpu][counter], 0); + if (mmap_array[nr_cpu][counter].base == MAP_FAILED) { + error("failed to mmap with %d (%s)\n", errno, strerror(errno)); + exit(-1); + } + + ioctl(fd[nr_cpu][counter], PERF_COUNTER_IOC_ENABLE); +} + +static void open_counters(int cpu, pid_t pid) +{ + int counter; + + if (pid > 0) { + pid_synthesize_comm_event(pid, 0); + pid_synthesize_mmap_samples(pid); + } + + group_fd = -1; + for (counter = 0; counter < nr_counters; counter++) + create_counter(counter, cpu, pid); + + nr_cpu++; +} + +static int __cmd_record(int argc, const char **argv) +{ + int i, counter; + struct stat st; + pid_t pid; + int flags; + int ret; + + page_size = sysconf(_SC_PAGE_SIZE); + nr_cpus = sysconf(_SC_NPROCESSORS_ONLN); + assert(nr_cpus <= MAX_NR_CPUS); + assert(nr_cpus >= 0); + + if (!stat(output_name, &st) && !force && !append_file) { + fprintf(stderr, "Error, output file %s exists, use -A to append or -f to overwrite.\n", + output_name); + exit(-1); + } + + flags = O_CREAT|O_RDWR; + if (append_file) + flags |= O_APPEND; + else + flags |= O_TRUNC; + + output = open(output_name, flags, S_IRUSR|S_IWUSR); + if (output < 0) { + perror("failed to create output file"); + exit(-1); + } + + if (!system_wide) { + open_counters(-1, target_pid != -1 ? target_pid : getpid()); + } else for (i = 0; i < nr_cpus; i++) + open_counters(i, target_pid); + + atexit(sig_atexit); + signal(SIGCHLD, sig_handler); + signal(SIGINT, sig_handler); + + if (target_pid == -1 && argc) { + pid = fork(); + if (pid < 0) + perror("failed to fork"); + + if (!pid) { + if (execvp(argv[0], (char **)argv)) { + perror(argv[0]); + exit(-1); + } + } + } + + if (realtime_prio) { + struct sched_param param; + + param.sched_priority = realtime_prio; + if (sched_setscheduler(0, SCHED_FIFO, ¶m)) { + printf("Could not set realtime priority.\n"); + exit(-1); + } + } + + if (system_wide) + synthesize_samples(); + + while (!done) { + int hits = samples; + + for (i = 0; i < nr_cpu; i++) { + for (counter = 0; counter < nr_counters; counter++) + mmap_read(&mmap_array[i][counter]); + } + + if (hits == samples) + ret = poll(event_array, nr_poll, 100); + } + + /* + * Approximate RIP event size: 24 bytes. + */ + fprintf(stderr, + "[ perf record: Captured and wrote %.3f MB %s (~%lld samples) ]\n", + (double)bytes_written / 1024.0 / 1024.0, + output_name, + bytes_written / 24); + + return 0; +} + +static const char * const record_usage[] = { + "perf record [<options>] [<command>]", + "perf record [<options>] -- <command> [<options>]", + NULL +}; + +static const struct option options[] = { + OPT_CALLBACK('e', "event", NULL, "event", + "event selector. use 'perf list' to list available events", + parse_events), + OPT_INTEGER('p', "pid", &target_pid, + "record events on existing pid"), + OPT_INTEGER('r', "realtime", &realtime_prio, + "collect data with this RT SCHED_FIFO priority"), + OPT_BOOLEAN('a', "all-cpus", &system_wide, + "system-wide collection from all CPUs"), + OPT_BOOLEAN('A', "append", &append_file, + "append to the output file to do incremental profiling"), + OPT_BOOLEAN('f', "force", &force, + "overwrite existing data file"), + OPT_LONG('c', "count", &default_interval, + "event period to sample"), + OPT_STRING('o', "output", &output_name, "file", + "output file name"), + OPT_BOOLEAN('i', "inherit", &inherit, + "child tasks inherit counters"), + OPT_INTEGER('F', "freq", &freq, + "profile at this frequency"), + OPT_INTEGER('m', "mmap-pages", &mmap_pages, + "number of mmap data pages"), + OPT_BOOLEAN('v', "verbose", &verbose, + "be more verbose (show counter open errors, etc)"), + OPT_END() +}; + +int cmd_record(int argc, const char **argv, const char *prefix) +{ + int counter; + + argc = parse_options(argc, argv, options, record_usage, 0); + if (!argc && target_pid == -1 && !system_wide) + usage_with_options(record_usage, options); + + if (!nr_counters) + nr_counters = 1; + + for (counter = 0; counter < nr_counters; counter++) { + if (attrs[counter].sample_period) + continue; + + attrs[counter].sample_period = default_interval; + } + + return __cmd_record(argc, argv); +} diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c new file mode 100644 index 00000000000..82fa93b4db9 --- /dev/null +++ b/tools/perf/builtin-report.c @@ -0,0 +1,1316 @@ +/* + * builtin-report.c + * + * Builtin report command: Analyze the perf.data input file, + * look up and read DSOs and symbol information and display + * a histogram of results, along various sorting keys. + */ +#include "builtin.h" + +#include "util/util.h" + +#include "util/color.h" +#include "util/list.h" +#include "util/cache.h" +#include "util/rbtree.h" +#include "util/symbol.h" +#include "util/string.h" + +#include "perf.h" + +#include "util/parse-options.h" +#include "util/parse-events.h" + +#define SHOW_KERNEL 1 +#define SHOW_USER 2 +#define SHOW_HV 4 + +static char const *input_name = "perf.data"; +static char *vmlinux = NULL; + +static char default_sort_order[] = "comm,dso"; +static char *sort_order = default_sort_order; + +static int input; +static int show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV; + +static int dump_trace = 0; +#define dprintf(x...) do { if (dump_trace) printf(x); } while (0) + +static int verbose; +static int full_paths; + +static unsigned long page_size; +static unsigned long mmap_window = 32; + +struct ip_event { + struct perf_event_header header; + __u64 ip; + __u32 pid, tid; + __u64 period; +}; + +struct mmap_event { + struct perf_event_header header; + __u32 pid, tid; + __u64 start; + __u64 len; + __u64 pgoff; + char filename[PATH_MAX]; +}; + +struct comm_event { + struct perf_event_header header; + __u32 pid, tid; + char comm[16]; +}; + +struct fork_event { + struct perf_event_header header; + __u32 pid, ppid; +}; + +struct period_event { + struct perf_event_header header; + __u64 time; + __u64 id; + __u64 sample_period; +}; + +typedef union event_union { + struct perf_event_header header; + struct ip_event ip; + struct mmap_event mmap; + struct comm_event comm; + struct fork_event fork; + struct period_event period; +} event_t; + +static LIST_HEAD(dsos); +static struct dso *kernel_dso; +static struct dso *vdso; + +static void dsos__add(struct dso *dso) +{ + list_add_tail(&dso->node, &dsos); +} + +static struct dso *dsos__find(const char *name) +{ + struct dso *pos; + + list_for_each_entry(pos, &dsos, node) + if (strcmp(pos->name, name) == 0) + return pos; + return NULL; +} + +static struct dso *dsos__findnew(const char *name) +{ + struct dso *dso = dsos__find(name); + int nr; + + if (dso) + return dso; + + dso = dso__new(name, 0); + if (!dso) + goto out_delete_dso; + + nr = dso__load(dso, NULL, verbose); + if (nr < 0) { + if (verbose) + fprintf(stderr, "Failed to open: %s\n", name); + goto out_delete_dso; + } + if (!nr && verbose) { + fprintf(stderr, + "No symbols found in: %s, maybe install a debug package?\n", + name); + } + + dsos__add(dso); + + return dso; + +out_delete_dso: + dso__delete(dso); + return NULL; +} + +static void dsos__fprintf(FILE *fp) +{ + struct dso *pos; + + list_for_each_entry(pos, &dsos, node) + dso__fprintf(pos, fp); +} + +static struct symbol *vdso__find_symbol(struct dso *dso, __u64 ip) +{ + return dso__find_symbol(kernel_dso, ip); +} + +static int load_kernel(void) +{ + int err; + + kernel_dso = dso__new("[kernel]", 0); + if (!kernel_dso) + return -1; + + err = dso__load_kernel(kernel_dso, vmlinux, NULL, verbose); + if (err) { + dso__delete(kernel_dso); + kernel_dso = NULL; + } else + dsos__add(kernel_dso); + + vdso = dso__new("[vdso]", 0); + if (!vdso) + return -1; + + vdso->find_symbol = vdso__find_symbol; + + dsos__add(vdso); + + return err; +} + +static char __cwd[PATH_MAX]; +static char *cwd = __cwd; +static int cwdlen; + +static int strcommon(const char *pathname) +{ + int n = 0; + + while (pathname[n] == cwd[n] && n < cwdlen) + ++n; + + return n; +} + +struct map { + struct list_head node; + __u64 start; + __u64 end; + __u64 pgoff; + __u64 (*map_ip)(struct map *, __u64); + struct dso *dso; +}; + +static __u64 map__map_ip(struct map *map, __u64 ip) +{ + return ip - map->start + map->pgoff; +} + +static __u64 vdso__map_ip(struct map *map, __u64 ip) +{ + return ip; +} + +static inline int is_anon_memory(const char *filename) +{ + return strcmp(filename, "//anon") == 0; +} + +static struct map *map__new(struct mmap_event *event) +{ + struct map *self = malloc(sizeof(*self)); + + if (self != NULL) { + const char *filename = event->filename; + char newfilename[PATH_MAX]; + int anon; + + if (cwd) { + int n = strcommon(filename); + + if (n == cwdlen) { + snprintf(newfilename, sizeof(newfilename), + ".%s", filename + n); + filename = newfilename; + } + } + + anon = is_anon_memory(filename); + + if (anon) { + snprintf(newfilename, sizeof(newfilename), "/tmp/perf-%d.map", event->pid); + filename = newfilename; + } + + self->start = event->start; + self->end = event->start + event->len; + self->pgoff = event->pgoff; + + self->dso = dsos__findnew(filename); + if (self->dso == NULL) + goto out_delete; + + if (self->dso == vdso || anon) + self->map_ip = vdso__map_ip; + else + self->map_ip = map__map_ip; + } + return self; +out_delete: + free(self); + return NULL; +} + +static struct map *map__clone(struct map *self) +{ + struct map *map = malloc(sizeof(*self)); + + if (!map) + return NULL; + + memcpy(map, self, sizeof(*self)); + + return map; +} + +static int map__overlap(struct map *l, struct map *r) +{ + if (l->start > r->start) { + struct map *t = l; + l = r; + r = t; + } + + if (l->end > r->start) + return 1; + + return 0; +} + +static size_t map__fprintf(struct map *self, FILE *fp) +{ + return fprintf(fp, " %Lx-%Lx %Lx %s\n", + self->start, self->end, self->pgoff, self->dso->name); +} + + +struct thread { + struct rb_node rb_node; + struct list_head maps; + pid_t pid; + char *comm; +}; + +static struct thread *thread__new(pid_t pid) +{ + struct thread *self = malloc(sizeof(*self)); + + if (self != NULL) { + self->pid = pid; + self->comm = malloc(32); + if (self->comm) + snprintf(self->comm, 32, ":%d", self->pid); + INIT_LIST_HEAD(&self->maps); + } + + return self; +} + +static int thread__set_comm(struct thread *self, const char *comm) +{ + if (self->comm) + free(self->comm); + self->comm = strdup(comm); + return self->comm ? 0 : -ENOMEM; +} + +static size_t thread__fprintf(struct thread *self, FILE *fp) +{ + struct map *pos; + size_t ret = fprintf(fp, "Thread %d %s\n", self->pid, self->comm); + + list_for_each_entry(pos, &self->maps, node) + ret += map__fprintf(pos, fp); + + return ret; +} + + +static struct rb_root threads; +static struct thread *last_match; + +static struct thread *threads__findnew(pid_t pid) +{ + struct rb_node **p = &threads.rb_node; + struct rb_node *parent = NULL; + struct thread *th; + + /* + * Font-end cache - PID lookups come in blocks, + * so most of the time we dont have to look up + * the full rbtree: + */ + if (last_match && last_match->pid == pid) + return last_match; + + while (*p != NULL) { + parent = *p; + th = rb_entry(parent, struct thread, rb_node); + + if (th->pid == pid) { + last_match = th; + return th; + } + + if (pid < th->pid) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + th = thread__new(pid); + if (th != NULL) { + rb_link_node(&th->rb_node, parent, p); + rb_insert_color(&th->rb_node, &threads); + last_match = th; + } + + return th; +} + +static void thread__insert_map(struct thread *self, struct map *map) +{ + struct map *pos, *tmp; + + list_for_each_entry_safe(pos, tmp, &self->maps, node) { + if (map__overlap(pos, map)) { + list_del_init(&pos->node); + /* XXX leaks dsos */ + free(pos); + } + } + + list_add_tail(&map->node, &self->maps); +} + +static int thread__fork(struct thread *self, struct thread *parent) +{ + struct map *map; + + if (self->comm) + free(self->comm); + self->comm = strdup(parent->comm); + if (!self->comm) + return -ENOMEM; + + list_for_each_entry(map, &parent->maps, node) { + struct map *new = map__clone(map); + if (!new) + return -ENOMEM; + thread__insert_map(self, new); + } + + return 0; +} + +static struct map *thread__find_map(struct thread *self, __u64 ip) +{ + struct map *pos; + + if (self == NULL) + return NULL; + + list_for_each_entry(pos, &self->maps, node) + if (ip >= pos->start && ip <= pos->end) + return pos; + + return NULL; +} + +static size_t threads__fprintf(FILE *fp) +{ + size_t ret = 0; + struct rb_node *nd; + + for (nd = rb_first(&threads); nd; nd = rb_next(nd)) { + struct thread *pos = rb_entry(nd, struct thread, rb_node); + + ret += thread__fprintf(pos, fp); + } + + return ret; +} + +/* + * histogram, sorted on item, collects counts + */ + +static struct rb_root hist; + +struct hist_entry { + struct rb_node rb_node; + + struct thread *thread; + struct map *map; + struct dso *dso; + struct symbol *sym; + __u64 ip; + char level; + + __u64 count; +}; + +/* + * configurable sorting bits + */ + +struct sort_entry { + struct list_head list; + + char *header; + + int64_t (*cmp)(struct hist_entry *, struct hist_entry *); + int64_t (*collapse)(struct hist_entry *, struct hist_entry *); + size_t (*print)(FILE *fp, struct hist_entry *); +}; + +/* --sort pid */ + +static int64_t +sort__thread_cmp(struct hist_entry *left, struct hist_entry *right) +{ + return right->thread->pid - left->thread->pid; +} + +static size_t +sort__thread_print(FILE *fp, struct hist_entry *self) +{ + return fprintf(fp, "%16s:%5d", self->thread->comm ?: "", self->thread->pid); +} + +static struct sort_entry sort_thread = { + .header = " Command: Pid", + .cmp = sort__thread_cmp, + .print = sort__thread_print, +}; + +/* --sort comm */ + +static int64_t +sort__comm_cmp(struct hist_entry *left, struct hist_entry *right) +{ + return right->thread->pid - left->thread->pid; +} + +static int64_t +sort__comm_collapse(struct hist_entry *left, struct hist_entry *right) +{ + char *comm_l = left->thread->comm; + char *comm_r = right->thread->comm; + + if (!comm_l || !comm_r) { + if (!comm_l && !comm_r) + return 0; + else if (!comm_l) + return -1; + else + return 1; + } + + return strcmp(comm_l, comm_r); +} + +static size_t +sort__comm_print(FILE *fp, struct hist_entry *self) +{ + return fprintf(fp, "%16s", self->thread->comm); +} + +static struct sort_entry sort_comm = { + .header = " Command", + .cmp = sort__comm_cmp, + .collapse = sort__comm_collapse, + .print = sort__comm_print, +}; + +/* --sort dso */ + +static int64_t +sort__dso_cmp(struct hist_entry *left, struct hist_entry *right) +{ + struct dso *dso_l = left->dso; + struct dso *dso_r = right->dso; + + if (!dso_l || !dso_r) { + if (!dso_l && !dso_r) + return 0; + else if (!dso_l) + return -1; + else + return 1; + } + + return strcmp(dso_l->name, dso_r->name); +} + +static size_t +sort__dso_print(FILE *fp, struct hist_entry *self) +{ + if (self->dso) + return fprintf(fp, "%-25s", self->dso->name); + + return fprintf(fp, "%016llx ", (__u64)self->ip); +} + +static struct sort_entry sort_dso = { + .header = "Shared Object ", + .cmp = sort__dso_cmp, + .print = sort__dso_print, +}; + +/* --sort symbol */ + +static int64_t +sort__sym_cmp(struct hist_entry *left, struct hist_entry *right) +{ + __u64 ip_l, ip_r; + + if (left->sym == right->sym) + return 0; + + ip_l = left->sym ? left->sym->start : left->ip; + ip_r = right->sym ? right->sym->start : right->ip; + + return (int64_t)(ip_r - ip_l); +} + +static size_t +sort__sym_print(FILE *fp, struct hist_entry *self) +{ + size_t ret = 0; + + if (verbose) + ret += fprintf(fp, "%#018llx ", (__u64)self->ip); + + if (self->sym) { + ret += fprintf(fp, "[%c] %s", + self->dso == kernel_dso ? 'k' : '.', self->sym->name); + } else { + ret += fprintf(fp, "%#016llx", (__u64)self->ip); + } + + return ret; +} + +static struct sort_entry sort_sym = { + .header = "Symbol", + .cmp = sort__sym_cmp, + .print = sort__sym_print, +}; + +static int sort__need_collapse = 0; + +struct sort_dimension { + char *name; + struct sort_entry *entry; + int taken; +}; + +static struct sort_dimension sort_dimensions[] = { + { .name = "pid", .entry = &sort_thread, }, + { .name = "comm", .entry = &sort_comm, }, + { .name = "dso", .entry = &sort_dso, }, + { .name = "symbol", .entry = &sort_sym, }, +}; + +static LIST_HEAD(hist_entry__sort_list); + +static int sort_dimension__add(char *tok) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(sort_dimensions); i++) { + struct sort_dimension *sd = &sort_dimensions[i]; + + if (sd->taken) + continue; + + if (strncasecmp(tok, sd->name, strlen(tok))) + continue; + + if (sd->entry->collapse) + sort__need_collapse = 1; + + list_add_tail(&sd->entry->list, &hist_entry__sort_list); + sd->taken = 1; + + return 0; + } + + return -ESRCH; +} + +static int64_t +hist_entry__cmp(struct hist_entry *left, struct hist_entry *right) +{ + struct sort_entry *se; + int64_t cmp = 0; + + list_for_each_entry(se, &hist_entry__sort_list, list) { + cmp = se->cmp(left, right); + if (cmp) + break; + } + + return cmp; +} + +static int64_t +hist_entry__collapse(struct hist_entry *left, struct hist_entry *right) +{ + struct sort_entry *se; + int64_t cmp = 0; + + list_for_each_entry(se, &hist_entry__sort_list, list) { + int64_t (*f)(struct hist_entry *, struct hist_entry *); + + f = se->collapse ?: se->cmp; + + cmp = f(left, right); + if (cmp) + break; + } + + return cmp; +} + +static size_t +hist_entry__fprintf(FILE *fp, struct hist_entry *self, __u64 total_samples) +{ + struct sort_entry *se; + size_t ret; + + if (total_samples) { + double percent = self->count * 100.0 / total_samples; + char *color = PERF_COLOR_NORMAL; + + /* + * We color high-overhead entries in red, mid-overhead + * entries in green - and keep the low overhead places + * normal: + */ + if (percent >= 5.0) { + color = PERF_COLOR_RED; + } else { + if (percent >= 0.5) + color = PERF_COLOR_GREEN; + } + + ret = color_fprintf(fp, color, " %6.2f%%", + (self->count * 100.0) / total_samples); + } else + ret = fprintf(fp, "%12Ld ", self->count); + + list_for_each_entry(se, &hist_entry__sort_list, list) { + fprintf(fp, " "); + ret += se->print(fp, self); + } + + ret += fprintf(fp, "\n"); + + return ret; +} + +/* + * collect histogram counts + */ + +static int +hist_entry__add(struct thread *thread, struct map *map, struct dso *dso, + struct symbol *sym, __u64 ip, char level, __u64 count) +{ + struct rb_node **p = &hist.rb_node; + struct rb_node *parent = NULL; + struct hist_entry *he; + struct hist_entry entry = { + .thread = thread, + .map = map, + .dso = dso, + .sym = sym, + .ip = ip, + .level = level, + .count = count, + }; + int cmp; + + while (*p != NULL) { + parent = *p; + he = rb_entry(parent, struct hist_entry, rb_node); + + cmp = hist_entry__cmp(&entry, he); + + if (!cmp) { + he->count += count; + return 0; + } + + if (cmp < 0) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + he = malloc(sizeof(*he)); + if (!he) + return -ENOMEM; + *he = entry; + rb_link_node(&he->rb_node, parent, p); + rb_insert_color(&he->rb_node, &hist); + + return 0; +} + +static void hist_entry__free(struct hist_entry *he) +{ + free(he); +} + +/* + * collapse the histogram + */ + +static struct rb_root collapse_hists; + +static void collapse__insert_entry(struct hist_entry *he) +{ + struct rb_node **p = &collapse_hists.rb_node; + struct rb_node *parent = NULL; + struct hist_entry *iter; + int64_t cmp; + + while (*p != NULL) { + parent = *p; + iter = rb_entry(parent, struct hist_entry, rb_node); + + cmp = hist_entry__collapse(iter, he); + + if (!cmp) { + iter->count += he->count; + hist_entry__free(he); + return; + } + + if (cmp < 0) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + rb_link_node(&he->rb_node, parent, p); + rb_insert_color(&he->rb_node, &collapse_hists); +} + +static void collapse__resort(void) +{ + struct rb_node *next; + struct hist_entry *n; + + if (!sort__need_collapse) + return; + + next = rb_first(&hist); + while (next) { + n = rb_entry(next, struct hist_entry, rb_node); + next = rb_next(&n->rb_node); + + rb_erase(&n->rb_node, &hist); + collapse__insert_entry(n); + } +} + +/* + * reverse the map, sort on count. + */ + +static struct rb_root output_hists; + +static void output__insert_entry(struct hist_entry *he) +{ + struct rb_node **p = &output_hists.rb_node; + struct rb_node *parent = NULL; + struct hist_entry *iter; + + while (*p != NULL) { + parent = *p; + iter = rb_entry(parent, struct hist_entry, rb_node); + + if (he->count > iter->count) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + rb_link_node(&he->rb_node, parent, p); + rb_insert_color(&he->rb_node, &output_hists); +} + +static void output__resort(void) +{ + struct rb_node *next; + struct hist_entry *n; + struct rb_root *tree = &hist; + + if (sort__need_collapse) + tree = &collapse_hists; + + next = rb_first(tree); + + while (next) { + n = rb_entry(next, struct hist_entry, rb_node); + next = rb_next(&n->rb_node); + + rb_erase(&n->rb_node, tree); + output__insert_entry(n); + } +} + +static size_t output__fprintf(FILE *fp, __u64 total_samples) +{ + struct hist_entry *pos; + struct sort_entry *se; + struct rb_node *nd; + size_t ret = 0; + + fprintf(fp, "\n"); + fprintf(fp, "#\n"); + fprintf(fp, "# (%Ld samples)\n", (__u64)total_samples); + fprintf(fp, "#\n"); + + fprintf(fp, "# Overhead"); + list_for_each_entry(se, &hist_entry__sort_list, list) + fprintf(fp, " %s", se->header); + fprintf(fp, "\n"); + + fprintf(fp, "# ........"); + list_for_each_entry(se, &hist_entry__sort_list, list) { + int i; + + fprintf(fp, " "); + for (i = 0; i < strlen(se->header); i++) + fprintf(fp, "."); + } + fprintf(fp, "\n"); + + fprintf(fp, "#\n"); + + for (nd = rb_first(&output_hists); nd; nd = rb_next(nd)) { + pos = rb_entry(nd, struct hist_entry, rb_node); + ret += hist_entry__fprintf(fp, pos, total_samples); + } + + if (!strcmp(sort_order, default_sort_order)) { + fprintf(fp, "#\n"); + fprintf(fp, "# (For more details, try: perf report --sort comm,dso,symbol)\n"); + fprintf(fp, "#\n"); + } + fprintf(fp, "\n"); + + return ret; +} + +static void register_idle_thread(void) +{ + struct thread *thread = threads__findnew(0); + + if (thread == NULL || + thread__set_comm(thread, "[idle]")) { + fprintf(stderr, "problem inserting idle task.\n"); + exit(-1); + } +} + +static unsigned long total = 0, + total_mmap = 0, + total_comm = 0, + total_fork = 0, + total_unknown = 0; + +static int +process_overflow_event(event_t *event, unsigned long offset, unsigned long head) +{ + char level; + int show = 0; + struct dso *dso = NULL; + struct thread *thread = threads__findnew(event->ip.pid); + __u64 ip = event->ip.ip; + __u64 period = 1; + struct map *map = NULL; + + if (event->header.type & PERF_SAMPLE_PERIOD) + period = event->ip.period; + + dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p period: %Ld\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->header.misc, + event->ip.pid, + (void *)(long)ip, + (long long)period); + + dprintf(" ... thread: %s:%d\n", thread->comm, thread->pid); + + if (thread == NULL) { + fprintf(stderr, "problem processing %d event, skipping it.\n", + event->header.type); + return -1; + } + + if (event->header.misc & PERF_EVENT_MISC_KERNEL) { + show = SHOW_KERNEL; + level = 'k'; + + dso = kernel_dso; + + dprintf(" ...... dso: %s\n", dso->name); + + } else if (event->header.misc & PERF_EVENT_MISC_USER) { + + show = SHOW_USER; + level = '.'; + + map = thread__find_map(thread, ip); + if (map != NULL) { + ip = map->map_ip(map, ip); + dso = map->dso; + } else { + /* + * If this is outside of all known maps, + * and is a negative address, try to look it + * up in the kernel dso, as it might be a + * vsyscall (which executes in user-mode): + */ + if ((long long)ip < 0) + dso = kernel_dso; + } + dprintf(" ...... dso: %s\n", dso ? dso->name : "<not found>"); + + } else { + show = SHOW_HV; + level = 'H'; + dprintf(" ...... dso: [hypervisor]\n"); + } + + if (show & show_mask) { + struct symbol *sym = NULL; + + if (dso) + sym = dso->find_symbol(dso, ip); + + if (hist_entry__add(thread, map, dso, sym, ip, level, period)) { + fprintf(stderr, + "problem incrementing symbol count, skipping event\n"); + return -1; + } + } + total += period; + + return 0; +} + +static int +process_mmap_event(event_t *event, unsigned long offset, unsigned long head) +{ + struct thread *thread = threads__findnew(event->mmap.pid); + struct map *map = map__new(&event->mmap); + + dprintf("%p [%p]: PERF_EVENT_MMAP %d: [%p(%p) @ %p]: %s\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->mmap.pid, + (void *)(long)event->mmap.start, + (void *)(long)event->mmap.len, + (void *)(long)event->mmap.pgoff, + event->mmap.filename); + + if (thread == NULL || map == NULL) { + dprintf("problem processing PERF_EVENT_MMAP, skipping event.\n"); + return 0; + } + + thread__insert_map(thread, map); + total_mmap++; + + return 0; +} + +static int +process_comm_event(event_t *event, unsigned long offset, unsigned long head) +{ + struct thread *thread = threads__findnew(event->comm.pid); + + dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->comm.comm, event->comm.pid); + + if (thread == NULL || + thread__set_comm(thread, event->comm.comm)) { + dprintf("problem processing PERF_EVENT_COMM, skipping event.\n"); + return -1; + } + total_comm++; + + return 0; +} + +static int +process_fork_event(event_t *event, unsigned long offset, unsigned long head) +{ + struct thread *thread = threads__findnew(event->fork.pid); + struct thread *parent = threads__findnew(event->fork.ppid); + + dprintf("%p [%p]: PERF_EVENT_FORK: %d:%d\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->fork.pid, event->fork.ppid); + + if (!thread || !parent || thread__fork(thread, parent)) { + dprintf("problem processing PERF_EVENT_FORK, skipping event.\n"); + return -1; + } + total_fork++; + + return 0; +} + +static int +process_period_event(event_t *event, unsigned long offset, unsigned long head) +{ + dprintf("%p [%p]: PERF_EVENT_PERIOD: time:%Ld, id:%Ld: period:%Ld\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->period.time, + event->period.id, + event->period.sample_period); + + return 0; +} + +static int +process_event(event_t *event, unsigned long offset, unsigned long head) +{ + if (event->header.misc & PERF_EVENT_MISC_OVERFLOW) + return process_overflow_event(event, offset, head); + + switch (event->header.type) { + case PERF_EVENT_MMAP: + return process_mmap_event(event, offset, head); + + case PERF_EVENT_COMM: + return process_comm_event(event, offset, head); + + case PERF_EVENT_FORK: + return process_fork_event(event, offset, head); + + case PERF_EVENT_PERIOD: + return process_period_event(event, offset, head); + /* + * We dont process them right now but they are fine: + */ + + case PERF_EVENT_THROTTLE: + case PERF_EVENT_UNTHROTTLE: + return 0; + + default: + return -1; + } + + return 0; +} + +static int __cmd_report(void) +{ + int ret, rc = EXIT_FAILURE; + unsigned long offset = 0; + unsigned long head = 0; + struct stat stat; + event_t *event; + uint32_t size; + char *buf; + + register_idle_thread(); + + input = open(input_name, O_RDONLY); + if (input < 0) { + fprintf(stderr, " failed to open file: %s", input_name); + if (!strcmp(input_name, "perf.data")) + fprintf(stderr, " (try 'perf record' first)"); + fprintf(stderr, "\n"); + exit(-1); + } + + ret = fstat(input, &stat); + if (ret < 0) { + perror("failed to stat file"); + exit(-1); + } + + if (!stat.st_size) { + fprintf(stderr, "zero-sized file, nothing to do!\n"); + exit(0); + } + + if (load_kernel() < 0) { + perror("failed to load kernel symbols"); + return EXIT_FAILURE; + } + + if (!full_paths) { + if (getcwd(__cwd, sizeof(__cwd)) == NULL) { + perror("failed to get the current directory"); + return EXIT_FAILURE; + } + cwdlen = strlen(cwd); + } else { + cwd = NULL; + cwdlen = 0; + } +remap: + buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ, + MAP_SHARED, input, offset); + if (buf == MAP_FAILED) { + perror("failed to mmap file"); + exit(-1); + } + +more: + event = (event_t *)(buf + head); + + size = event->header.size; + if (!size) + size = 8; + + if (head + event->header.size >= page_size * mmap_window) { + unsigned long shift = page_size * (head / page_size); + int ret; + + ret = munmap(buf, page_size * mmap_window); + assert(ret == 0); + + offset += shift; + head -= shift; + goto remap; + } + + size = event->header.size; + + dprintf("%p [%p]: event: %d\n", + (void *)(offset + head), + (void *)(long)event->header.size, + event->header.type); + + if (!size || process_event(event, offset, head) < 0) { + + dprintf("%p [%p]: skipping unknown header type: %d\n", + (void *)(offset + head), + (void *)(long)(event->header.size), + event->header.type); + + total_unknown++; + + /* + * assume we lost track of the stream, check alignment, and + * increment a single u64 in the hope to catch on again 'soon'. + */ + + if (unlikely(head & 7)) + head &= ~7ULL; + + size = 8; + } + + head += size; + + if (offset + head < stat.st_size) + goto more; + + rc = EXIT_SUCCESS; + close(input); + + dprintf(" IP events: %10ld\n", total); + dprintf(" mmap events: %10ld\n", total_mmap); + dprintf(" comm events: %10ld\n", total_comm); + dprintf(" fork events: %10ld\n", total_fork); + dprintf(" unknown events: %10ld\n", total_unknown); + + if (dump_trace) + return 0; + + if (verbose >= 3) + threads__fprintf(stdout); + + if (verbose >= 2) + dsos__fprintf(stdout); + + collapse__resort(); + output__resort(); + output__fprintf(stdout, total); + + return rc; +} + +static const char * const report_usage[] = { + "perf report [<options>] <command>", + NULL +}; + +static const struct option options[] = { + OPT_STRING('i', "input", &input_name, "file", + "input file name"), + OPT_BOOLEAN('v', "verbose", &verbose, + "be more verbose (show symbol address, etc)"), + OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace, + "dump raw trace in ASCII"), + OPT_STRING('k', "vmlinux", &vmlinux, "file", "vmlinux pathname"), + OPT_STRING('s', "sort", &sort_order, "key[,key2...]", + "sort by key(s): pid, comm, dso, symbol. Default: pid,symbol"), + OPT_BOOLEAN('P', "full-paths", &full_paths, + "Don't shorten the pathnames taking into account the cwd"), + OPT_END() +}; + +static void setup_sorting(void) +{ + char *tmp, *tok, *str = strdup(sort_order); + + for (tok = strtok_r(str, ", ", &tmp); + tok; tok = strtok_r(NULL, ", ", &tmp)) { + if (sort_dimension__add(tok) < 0) { + error("Unknown --sort key: `%s'", tok); + usage_with_options(report_usage, options); + } + } + + free(str); +} + +int cmd_report(int argc, const char **argv, const char *prefix) +{ + symbol__init(); + + page_size = getpagesize(); + + argc = parse_options(argc, argv, options, report_usage, 0); + + setup_sorting(); + + /* + * Any (unrecognized) arguments left? + */ + if (argc) + usage_with_options(report_usage, options); + + setup_pager(); + + return __cmd_report(); +} diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c new file mode 100644 index 00000000000..c43e4a97dc4 --- /dev/null +++ b/tools/perf/builtin-stat.c @@ -0,0 +1,367 @@ +/* + * builtin-stat.c + * + * Builtin stat command: Give a precise performance counters summary + * overview about any workload, CPU or specific PID. + * + * Sample output: + + $ perf stat ~/hackbench 10 + Time: 0.104 + + Performance counter stats for '/home/mingo/hackbench': + + 1255.538611 task clock ticks # 10.143 CPU utilization factor + 54011 context switches # 0.043 M/sec + 385 CPU migrations # 0.000 M/sec + 17755 pagefaults # 0.014 M/sec + 3808323185 CPU cycles # 3033.219 M/sec + 1575111190 instructions # 1254.530 M/sec + 17367895 cache references # 13.833 M/sec + 7674421 cache misses # 6.112 M/sec + + Wall-clock time elapsed: 123.786620 msecs + + * + * Copyright (C) 2008, Red Hat Inc, Ingo Molnar <mingo@redhat.com> + * + * Improvements and fixes by: + * + * Arjan van de Ven <arjan@linux.intel.com> + * Yanmin Zhang <yanmin.zhang@intel.com> + * Wu Fengguang <fengguang.wu@intel.com> + * Mike Galbraith <efault@gmx.de> + * Paul Mackerras <paulus@samba.org> + * + * Released under the GPL v2. (and only v2, not any later version) + */ + +#include "perf.h" +#include "builtin.h" +#include "util/util.h" +#include "util/parse-options.h" +#include "util/parse-events.h" + +#include <sys/prctl.h> + +static struct perf_counter_attr default_attrs[MAX_COUNTERS] = { + + { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK }, + { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES}, + { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS }, + { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS }, + + { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES }, + { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS }, + { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_REFERENCES}, + { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_MISSES }, + +}; + +static int system_wide = 0; +static int inherit = 1; +static int verbose = 0; + +static int fd[MAX_NR_CPUS][MAX_COUNTERS]; + +static int target_pid = -1; +static int nr_cpus = 0; +static unsigned int page_size; + +static int scale = 1; + +static const unsigned int default_count[] = { + 1000000, + 1000000, + 10000, + 10000, + 1000000, + 10000, +}; + +static __u64 event_res[MAX_COUNTERS][3]; +static __u64 event_scaled[MAX_COUNTERS]; + +static __u64 runtime_nsecs; +static __u64 walltime_nsecs; +static __u64 runtime_cycles; + +static void create_perf_stat_counter(int counter) +{ + struct perf_counter_attr *attr = attrs + counter; + + if (scale) + attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED | + PERF_FORMAT_TOTAL_TIME_RUNNING; + + if (system_wide) { + int cpu; + for (cpu = 0; cpu < nr_cpus; cpu ++) { + fd[cpu][counter] = sys_perf_counter_open(attr, -1, cpu, -1, 0); + if (fd[cpu][counter] < 0 && verbose) { + printf("Error: counter %d, sys_perf_counter_open() syscall returned with %d (%s)\n", counter, fd[cpu][counter], strerror(errno)); + } + } + } else { + attr->inherit = inherit; + attr->disabled = 1; + + fd[0][counter] = sys_perf_counter_open(attr, 0, -1, -1, 0); + if (fd[0][counter] < 0 && verbose) { + printf("Error: counter %d, sys_perf_counter_open() syscall returned with %d (%s)\n", counter, fd[0][counter], strerror(errno)); + } + } +} + +/* + * Does the counter have nsecs as a unit? + */ +static inline int nsec_counter(int counter) +{ + if (attrs[counter].type != PERF_TYPE_SOFTWARE) + return 0; + + if (attrs[counter].config == PERF_COUNT_SW_CPU_CLOCK) + return 1; + + if (attrs[counter].config == PERF_COUNT_SW_TASK_CLOCK) + return 1; + + return 0; +} + +/* + * Read out the results of a single counter: + */ +static void read_counter(int counter) +{ + __u64 *count, single_count[3]; + ssize_t res; + int cpu, nv; + int scaled; + + count = event_res[counter]; + + count[0] = count[1] = count[2] = 0; + + nv = scale ? 3 : 1; + for (cpu = 0; cpu < nr_cpus; cpu ++) { + if (fd[cpu][counter] < 0) + continue; + + res = read(fd[cpu][counter], single_count, nv * sizeof(__u64)); + assert(res == nv * sizeof(__u64)); + + count[0] += single_count[0]; + if (scale) { + count[1] += single_count[1]; + count[2] += single_count[2]; + } + } + + scaled = 0; + if (scale) { + if (count[2] == 0) { + event_scaled[counter] = -1; + count[0] = 0; + return; + } + + if (count[2] < count[1]) { + event_scaled[counter] = 1; + count[0] = (unsigned long long) + ((double)count[0] * count[1] / count[2] + 0.5); + } + } + /* + * Save the full runtime - to allow normalization during printout: + */ + if (attrs[counter].type == PERF_TYPE_SOFTWARE && + attrs[counter].config == PERF_COUNT_SW_TASK_CLOCK) + runtime_nsecs = count[0]; + if (attrs[counter].type == PERF_TYPE_HARDWARE && + attrs[counter].config == PERF_COUNT_HW_CPU_CYCLES) + runtime_cycles = count[0]; +} + +/* + * Print out the results of a single counter: + */ +static void print_counter(int counter) +{ + __u64 *count; + int scaled; + + count = event_res[counter]; + scaled = event_scaled[counter]; + + if (scaled == -1) { + fprintf(stderr, " %14s %-20s\n", + "<not counted>", event_name(counter)); + return; + } + + if (nsec_counter(counter)) { + double msecs = (double)count[0] / 1000000; + + fprintf(stderr, " %14.6f %-20s", + msecs, event_name(counter)); + if (attrs[counter].type == PERF_TYPE_SOFTWARE && + attrs[counter].config == PERF_COUNT_SW_TASK_CLOCK) { + + if (walltime_nsecs) + fprintf(stderr, " # %11.3f CPU utilization factor", + (double)count[0] / (double)walltime_nsecs); + } + } else { + fprintf(stderr, " %14Ld %-20s", + count[0], event_name(counter)); + if (runtime_nsecs) + fprintf(stderr, " # %11.3f M/sec", + (double)count[0]/runtime_nsecs*1000.0); + if (runtime_cycles && + attrs[counter].type == PERF_TYPE_HARDWARE && + attrs[counter].config == PERF_COUNT_HW_INSTRUCTIONS) { + + fprintf(stderr, " # %1.3f per cycle", + (double)count[0] / (double)runtime_cycles); + } + } + if (scaled) + fprintf(stderr, " (scaled from %.2f%%)", + (double) count[2] / count[1] * 100); + fprintf(stderr, "\n"); +} + +static int do_perf_stat(int argc, const char **argv) +{ + unsigned long long t0, t1; + int counter; + int status; + int pid; + int i; + + if (!system_wide) + nr_cpus = 1; + + for (counter = 0; counter < nr_counters; counter++) + create_perf_stat_counter(counter); + + /* + * Enable counters and exec the command: + */ + t0 = rdclock(); + prctl(PR_TASK_PERF_COUNTERS_ENABLE); + + if ((pid = fork()) < 0) + perror("failed to fork"); + + if (!pid) { + if (execvp(argv[0], (char **)argv)) { + perror(argv[0]); + exit(-1); + } + } + + while (wait(&status) >= 0) + ; + + prctl(PR_TASK_PERF_COUNTERS_DISABLE); + t1 = rdclock(); + + walltime_nsecs = t1 - t0; + + fflush(stdout); + + fprintf(stderr, "\n"); + fprintf(stderr, " Performance counter stats for \'%s", argv[0]); + + for (i = 1; i < argc; i++) + fprintf(stderr, " %s", argv[i]); + + fprintf(stderr, "\':\n"); + fprintf(stderr, "\n"); + + for (counter = 0; counter < nr_counters; counter++) + read_counter(counter); + + for (counter = 0; counter < nr_counters; counter++) + print_counter(counter); + + + fprintf(stderr, "\n"); + fprintf(stderr, " Wall-clock time elapsed: %12.6f msecs\n", + (double)(t1-t0)/1e6); + fprintf(stderr, "\n"); + + return 0; +} + +static volatile int signr = -1; + +static void skip_signal(int signo) +{ + signr = signo; +} + +static void sig_atexit(void) +{ + if (signr == -1) + return; + + signal(signr, SIG_DFL); + kill(getpid(), signr); +} + +static const char * const stat_usage[] = { + "perf stat [<options>] <command>", + NULL +}; + +static const struct option options[] = { + OPT_CALLBACK('e', "event", NULL, "event", + "event selector. use 'perf list' to list available events", + parse_events), + OPT_BOOLEAN('i', "inherit", &inherit, + "child tasks inherit counters"), + OPT_INTEGER('p', "pid", &target_pid, + "stat events on existing pid"), + OPT_BOOLEAN('a', "all-cpus", &system_wide, + "system-wide collection from all CPUs"), + OPT_BOOLEAN('S', "scale", &scale, + "scale/normalize counters"), + OPT_BOOLEAN('v', "verbose", &verbose, + "be more verbose (show counter open errors, etc)"), + OPT_END() +}; + +int cmd_stat(int argc, const char **argv, const char *prefix) +{ + page_size = sysconf(_SC_PAGE_SIZE); + + memcpy(attrs, default_attrs, sizeof(attrs)); + + argc = parse_options(argc, argv, options, stat_usage, 0); + if (!argc) + usage_with_options(stat_usage, options); + + if (!nr_counters) + nr_counters = 8; + + nr_cpus = sysconf(_SC_NPROCESSORS_ONLN); + assert(nr_cpus <= MAX_NR_CPUS); + assert(nr_cpus >= 0); + + /* + * We dont want to block the signals - that would cause + * child tasks to inherit that and Ctrl-C would not work. + * What we want is for Ctrl-C to work in the exec()-ed + * task, but being ignored by perf stat itself: + */ + atexit(sig_atexit); + signal(SIGINT, skip_signal); + signal(SIGALRM, skip_signal); + signal(SIGABRT, skip_signal); + + return do_perf_stat(argc, argv); +} diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c new file mode 100644 index 00000000000..fe338d3c5d7 --- /dev/null +++ b/tools/perf/builtin-top.c @@ -0,0 +1,736 @@ +/* + * builtin-top.c + * + * Builtin top command: Display a continuously updated profile of + * any workload, CPU or specific PID. + * + * Copyright (C) 2008, Red Hat Inc, Ingo Molnar <mingo@redhat.com> + * + * Improvements and fixes by: + * + * Arjan van de Ven <arjan@linux.intel.com> + * Yanmin Zhang <yanmin.zhang@intel.com> + * Wu Fengguang <fengguang.wu@intel.com> + * Mike Galbraith <efault@gmx.de> + * Paul Mackerras <paulus@samba.org> + * + * Released under the GPL v2. (and only v2, not any later version) + */ +#include "builtin.h" + +#include "perf.h" + +#include "util/symbol.h" +#include "util/color.h" +#include "util/util.h" +#include "util/rbtree.h" +#include "util/parse-options.h" +#include "util/parse-events.h" + +#include <assert.h> +#include <fcntl.h> + +#include <stdio.h> + +#include <errno.h> +#include <time.h> +#include <sched.h> +#include <pthread.h> + +#include <sys/syscall.h> +#include <sys/ioctl.h> +#include <sys/poll.h> +#include <sys/prctl.h> +#include <sys/wait.h> +#include <sys/uio.h> +#include <sys/mman.h> + +#include <linux/unistd.h> +#include <linux/types.h> + +static int fd[MAX_NR_CPUS][MAX_COUNTERS]; + +static int system_wide = 0; + +static int default_interval = 100000; + +static __u64 count_filter = 5; +static int print_entries = 15; + +static int target_pid = -1; +static int profile_cpu = -1; +static int nr_cpus = 0; +static unsigned int realtime_prio = 0; +static int group = 0; +static unsigned int page_size; +static unsigned int mmap_pages = 16; +static int freq = 0; +static int verbose = 0; + +static char *sym_filter; +static unsigned long filter_start; +static unsigned long filter_end; + +static int delay_secs = 2; +static int zero; +static int dump_symtab; + +/* + * Symbols + */ + +static __u64 min_ip; +static __u64 max_ip = -1ll; + +struct sym_entry { + struct rb_node rb_node; + struct list_head node; + unsigned long count[MAX_COUNTERS]; + unsigned long snap_count; + double weight; + int skip; +}; + +struct sym_entry *sym_filter_entry; + +struct dso *kernel_dso; + +/* + * Symbols will be added here in record_ip and will get out + * after decayed. + */ +static LIST_HEAD(active_symbols); +static pthread_mutex_t active_symbols_lock = PTHREAD_MUTEX_INITIALIZER; + +/* + * Ordering weight: count-1 * count-2 * ... / count-n + */ +static double sym_weight(const struct sym_entry *sym) +{ + double weight = sym->snap_count; + int counter; + + for (counter = 1; counter < nr_counters-1; counter++) + weight *= sym->count[counter]; + + weight /= (sym->count[counter] + 1); + + return weight; +} + +static long samples; +static long userspace_samples; +static const char CONSOLE_CLEAR[] = "[H[2J"; + +static void __list_insert_active_sym(struct sym_entry *syme) +{ + list_add(&syme->node, &active_symbols); +} + +static void list_remove_active_sym(struct sym_entry *syme) +{ + pthread_mutex_lock(&active_symbols_lock); + list_del_init(&syme->node); + pthread_mutex_unlock(&active_symbols_lock); +} + +static void rb_insert_active_sym(struct rb_root *tree, struct sym_entry *se) +{ + struct rb_node **p = &tree->rb_node; + struct rb_node *parent = NULL; + struct sym_entry *iter; + + while (*p != NULL) { + parent = *p; + iter = rb_entry(parent, struct sym_entry, rb_node); + + if (se->weight > iter->weight) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + rb_link_node(&se->rb_node, parent, p); + rb_insert_color(&se->rb_node, tree); +} + +static void print_sym_table(void) +{ + int printed = 0, j; + int counter; + float samples_per_sec = samples/delay_secs; + float ksamples_per_sec = (samples-userspace_samples)/delay_secs; + float sum_ksamples = 0.0; + struct sym_entry *syme, *n; + struct rb_root tmp = RB_ROOT; + struct rb_node *nd; + + samples = userspace_samples = 0; + + /* Sort the active symbols */ + pthread_mutex_lock(&active_symbols_lock); + syme = list_entry(active_symbols.next, struct sym_entry, node); + pthread_mutex_unlock(&active_symbols_lock); + + list_for_each_entry_safe_from(syme, n, &active_symbols, node) { + syme->snap_count = syme->count[0]; + if (syme->snap_count != 0) { + syme->weight = sym_weight(syme); + rb_insert_active_sym(&tmp, syme); + sum_ksamples += syme->snap_count; + + for (j = 0; j < nr_counters; j++) + syme->count[j] = zero ? 0 : syme->count[j] * 7 / 8; + } else + list_remove_active_sym(syme); + } + + puts(CONSOLE_CLEAR); + + printf( +"------------------------------------------------------------------------------\n"); + printf( " PerfTop:%8.0f irqs/sec kernel:%4.1f%% [", + samples_per_sec, + 100.0 - (100.0*((samples_per_sec-ksamples_per_sec)/samples_per_sec))); + + if (nr_counters == 1) { + printf("%Ld", attrs[0].sample_period); + if (freq) + printf("Hz "); + else + printf(" "); + } + + for (counter = 0; counter < nr_counters; counter++) { + if (counter) + printf("/"); + + printf("%s", event_name(counter)); + } + + printf( "], "); + + if (target_pid != -1) + printf(" (target_pid: %d", target_pid); + else + printf(" (all"); + + if (profile_cpu != -1) + printf(", cpu: %d)\n", profile_cpu); + else { + if (target_pid != -1) + printf(")\n"); + else + printf(", %d CPUs)\n", nr_cpus); + } + + printf("------------------------------------------------------------------------------\n\n"); + + if (nr_counters == 1) + printf(" samples pcnt"); + else + printf(" weight samples pcnt"); + + printf(" RIP kernel function\n" + " ______ _______ _____ ________________ _______________\n\n" + ); + + for (nd = rb_first(&tmp); nd; nd = rb_next(nd)) { + struct sym_entry *syme = rb_entry(nd, struct sym_entry, rb_node); + struct symbol *sym = (struct symbol *)(syme + 1); + char *color = PERF_COLOR_NORMAL; + double pcnt; + + if (++printed > print_entries || syme->snap_count < count_filter) + continue; + + pcnt = 100.0 - (100.0 * ((sum_ksamples - syme->snap_count) / + sum_ksamples)); + + /* + * We color high-overhead entries in red, mid-overhead + * entries in green - and keep the low overhead places + * normal: + */ + if (pcnt >= 5.0) { + color = PERF_COLOR_RED; + } else { + if (pcnt >= 0.5) + color = PERF_COLOR_GREEN; + } + + if (nr_counters == 1) + printf("%20.2f - ", syme->weight); + else + printf("%9.1f %10ld - ", syme->weight, syme->snap_count); + + color_fprintf(stdout, color, "%4.1f%%", pcnt); + printf(" - %016llx : %s\n", sym->start, sym->name); + } +} + +static void *display_thread(void *arg) +{ + struct pollfd stdin_poll = { .fd = 0, .events = POLLIN }; + int delay_msecs = delay_secs * 1000; + + printf("PerfTop refresh period: %d seconds\n", delay_secs); + + do { + print_sym_table(); + } while (!poll(&stdin_poll, 1, delay_msecs) == 1); + + printf("key pressed - exiting.\n"); + exit(0); + + return NULL; +} + +static int symbol_filter(struct dso *self, struct symbol *sym) +{ + static int filter_match; + struct sym_entry *syme; + const char *name = sym->name; + + if (!strcmp(name, "_text") || + !strcmp(name, "_etext") || + !strcmp(name, "_sinittext") || + !strncmp("init_module", name, 11) || + !strncmp("cleanup_module", name, 14) || + strstr(name, "_text_start") || + strstr(name, "_text_end")) + return 1; + + syme = dso__sym_priv(self, sym); + /* Tag samples to be skipped. */ + if (!strcmp("default_idle", name) || + !strcmp("cpu_idle", name) || + !strcmp("enter_idle", name) || + !strcmp("exit_idle", name) || + !strcmp("mwait_idle", name)) + syme->skip = 1; + + if (filter_match == 1) { + filter_end = sym->start; + filter_match = -1; + if (filter_end - filter_start > 10000) { + fprintf(stderr, + "hm, too large filter symbol <%s> - skipping.\n", + sym_filter); + fprintf(stderr, "symbol filter start: %016lx\n", + filter_start); + fprintf(stderr, " end: %016lx\n", + filter_end); + filter_end = filter_start = 0; + sym_filter = NULL; + sleep(1); + } + } + + if (filter_match == 0 && sym_filter && !strcmp(name, sym_filter)) { + filter_match = 1; + filter_start = sym->start; + } + + + return 0; +} + +static int parse_symbols(void) +{ + struct rb_node *node; + struct symbol *sym; + + kernel_dso = dso__new("[kernel]", sizeof(struct sym_entry)); + if (kernel_dso == NULL) + return -1; + + if (dso__load_kernel(kernel_dso, NULL, symbol_filter, 1) != 0) + goto out_delete_dso; + + node = rb_first(&kernel_dso->syms); + sym = rb_entry(node, struct symbol, rb_node); + min_ip = sym->start; + + node = rb_last(&kernel_dso->syms); + sym = rb_entry(node, struct symbol, rb_node); + max_ip = sym->end; + + if (dump_symtab) + dso__fprintf(kernel_dso, stderr); + + return 0; + +out_delete_dso: + dso__delete(kernel_dso); + kernel_dso = NULL; + return -1; +} + +#define TRACE_COUNT 3 + +/* + * Binary search in the histogram table and record the hit: + */ +static void record_ip(__u64 ip, int counter) +{ + struct symbol *sym = dso__find_symbol(kernel_dso, ip); + + if (sym != NULL) { + struct sym_entry *syme = dso__sym_priv(kernel_dso, sym); + + if (!syme->skip) { + syme->count[counter]++; + pthread_mutex_lock(&active_symbols_lock); + if (list_empty(&syme->node) || !syme->node.next) + __list_insert_active_sym(syme); + pthread_mutex_unlock(&active_symbols_lock); + return; + } + } + + samples--; +} + +static void process_event(__u64 ip, int counter) +{ + samples++; + + if (ip < min_ip || ip > max_ip) { + userspace_samples++; + return; + } + + record_ip(ip, counter); +} + +struct mmap_data { + int counter; + void *base; + unsigned int mask; + unsigned int prev; +}; + +static unsigned int mmap_read_head(struct mmap_data *md) +{ + struct perf_counter_mmap_page *pc = md->base; + int head; + + head = pc->data_head; + rmb(); + + return head; +} + +struct timeval last_read, this_read; + +static void mmap_read_counter(struct mmap_data *md) +{ + unsigned int head = mmap_read_head(md); + unsigned int old = md->prev; + unsigned char *data = md->base + page_size; + int diff; + + gettimeofday(&this_read, NULL); + + /* + * If we're further behind than half the buffer, there's a chance + * the writer will bite our tail and mess up the samples under us. + * + * If we somehow ended up ahead of the head, we got messed up. + * + * In either case, truncate and restart at head. + */ + diff = head - old; + if (diff > md->mask / 2 || diff < 0) { + struct timeval iv; + unsigned long msecs; + + timersub(&this_read, &last_read, &iv); + msecs = iv.tv_sec*1000 + iv.tv_usec/1000; + + fprintf(stderr, "WARNING: failed to keep up with mmap data." + " Last read %lu msecs ago.\n", msecs); + + /* + * head points to a known good entry, start there. + */ + old = head; + } + + last_read = this_read; + + for (; old != head;) { + struct ip_event { + struct perf_event_header header; + __u64 ip; + __u32 pid, target_pid; + }; + struct mmap_event { + struct perf_event_header header; + __u32 pid, target_pid; + __u64 start; + __u64 len; + __u64 pgoff; + char filename[PATH_MAX]; + }; + + typedef union event_union { + struct perf_event_header header; + struct ip_event ip; + struct mmap_event mmap; + } event_t; + + event_t *event = (event_t *)&data[old & md->mask]; + + event_t event_copy; + + size_t size = event->header.size; + + /* + * Event straddles the mmap boundary -- header should always + * be inside due to u64 alignment of output. + */ + if ((old & md->mask) + size != ((old + size) & md->mask)) { + unsigned int offset = old; + unsigned int len = min(sizeof(*event), size), cpy; + void *dst = &event_copy; + + do { + cpy = min(md->mask + 1 - (offset & md->mask), len); + memcpy(dst, &data[offset & md->mask], cpy); + offset += cpy; + dst += cpy; + len -= cpy; + } while (len); + + event = &event_copy; + } + + old += size; + + if (event->header.misc & PERF_EVENT_MISC_OVERFLOW) { + if (event->header.type & PERF_SAMPLE_IP) + process_event(event->ip.ip, md->counter); + } + } + + md->prev = old; +} + +static struct pollfd event_array[MAX_NR_CPUS * MAX_COUNTERS]; +static struct mmap_data mmap_array[MAX_NR_CPUS][MAX_COUNTERS]; + +static void mmap_read(void) +{ + int i, counter; + + for (i = 0; i < nr_cpus; i++) { + for (counter = 0; counter < nr_counters; counter++) + mmap_read_counter(&mmap_array[i][counter]); + } +} + +int nr_poll; +int group_fd; + +static void start_counter(int i, int counter) +{ + struct perf_counter_attr *attr; + unsigned int cpu; + + cpu = profile_cpu; + if (target_pid == -1 && profile_cpu == -1) + cpu = i; + + attr = attrs + counter; + + attr->sample_type = PERF_SAMPLE_IP | PERF_SAMPLE_TID; + attr->freq = freq; + +try_again: + fd[i][counter] = sys_perf_counter_open(attr, target_pid, cpu, group_fd, 0); + + if (fd[i][counter] < 0) { + int err = errno; + + if (err == EPERM) + die("No permission - are you root?\n"); + /* + * If it's cycles then fall back to hrtimer + * based cpu-clock-tick sw counter, which + * is always available even if no PMU support: + */ + if (attr->type == PERF_TYPE_HARDWARE + && attr->config == PERF_COUNT_HW_CPU_CYCLES) { + + if (verbose) + warning(" ... trying to fall back to cpu-clock-ticks\n"); + + attr->type = PERF_TYPE_SOFTWARE; + attr->config = PERF_COUNT_SW_CPU_CLOCK; + goto try_again; + } + printf("\n"); + error("perfcounter syscall returned with %d (%s)\n", + fd[i][counter], strerror(err)); + die("No CONFIG_PERF_COUNTERS=y kernel support configured?\n"); + exit(-1); + } + assert(fd[i][counter] >= 0); + fcntl(fd[i][counter], F_SETFL, O_NONBLOCK); + + /* + * First counter acts as the group leader: + */ + if (group && group_fd == -1) + group_fd = fd[i][counter]; + + event_array[nr_poll].fd = fd[i][counter]; + event_array[nr_poll].events = POLLIN; + nr_poll++; + + mmap_array[i][counter].counter = counter; + mmap_array[i][counter].prev = 0; + mmap_array[i][counter].mask = mmap_pages*page_size - 1; + mmap_array[i][counter].base = mmap(NULL, (mmap_pages+1)*page_size, + PROT_READ, MAP_SHARED, fd[i][counter], 0); + if (mmap_array[i][counter].base == MAP_FAILED) + die("failed to mmap with %d (%s)\n", errno, strerror(errno)); +} + +static int __cmd_top(void) +{ + pthread_t thread; + int i, counter; + int ret; + + for (i = 0; i < nr_cpus; i++) { + group_fd = -1; + for (counter = 0; counter < nr_counters; counter++) + start_counter(i, counter); + } + + /* Wait for a minimal set of events before starting the snapshot */ + poll(event_array, nr_poll, 100); + + mmap_read(); + + if (pthread_create(&thread, NULL, display_thread, NULL)) { + printf("Could not create display thread.\n"); + exit(-1); + } + + if (realtime_prio) { + struct sched_param param; + + param.sched_priority = realtime_prio; + if (sched_setscheduler(0, SCHED_FIFO, ¶m)) { + printf("Could not set realtime priority.\n"); + exit(-1); + } + } + + while (1) { + int hits = samples; + + mmap_read(); + + if (hits == samples) + ret = poll(event_array, nr_poll, 100); + } + + return 0; +} + +static const char * const top_usage[] = { + "perf top [<options>]", + NULL +}; + +static const struct option options[] = { + OPT_CALLBACK('e', "event", NULL, "event", + "event selector. use 'perf list' to list available events", + parse_events), + OPT_INTEGER('c', "count", &default_interval, + "event period to sample"), + OPT_INTEGER('p', "pid", &target_pid, + "profile events on existing pid"), + OPT_BOOLEAN('a', "all-cpus", &system_wide, + "system-wide collection from all CPUs"), + OPT_INTEGER('C', "CPU", &profile_cpu, + "CPU to profile on"), + OPT_INTEGER('m', "mmap-pages", &mmap_pages, + "number of mmap data pages"), + OPT_INTEGER('r', "realtime", &realtime_prio, + "collect data with this RT SCHED_FIFO priority"), + OPT_INTEGER('d', "delay", &delay_secs, + "number of seconds to delay between refreshes"), + OPT_BOOLEAN('D', "dump-symtab", &dump_symtab, + "dump the symbol table used for profiling"), + OPT_INTEGER('f', "count-filter", &count_filter, + "only display functions with more events than this"), + OPT_BOOLEAN('g', "group", &group, + "put the counters into a counter group"), + OPT_STRING('s', "sym-filter", &sym_filter, "pattern", + "only display symbols matchig this pattern"), + OPT_BOOLEAN('z', "zero", &group, + "zero history across updates"), + OPT_INTEGER('F', "freq", &freq, + "profile at this frequency"), + OPT_INTEGER('E', "entries", &print_entries, + "display this many functions"), + OPT_BOOLEAN('v', "verbose", &verbose, + "be more verbose (show counter open errors, etc)"), + OPT_END() +}; + +int cmd_top(int argc, const char **argv, const char *prefix) +{ + int counter; + + page_size = sysconf(_SC_PAGE_SIZE); + + argc = parse_options(argc, argv, options, top_usage, 0); + if (argc) + usage_with_options(top_usage, options); + + if (freq) { + default_interval = freq; + freq = 1; + } + + /* CPU and PID are mutually exclusive */ + if (target_pid != -1 && profile_cpu != -1) { + printf("WARNING: PID switch overriding CPU\n"); + sleep(1); + profile_cpu = -1; + } + + if (!nr_counters) + nr_counters = 1; + + if (delay_secs < 1) + delay_secs = 1; + + parse_symbols(); + + /* + * Fill in the ones not specifically initialized via -c: + */ + for (counter = 0; counter < nr_counters; counter++) { + if (attrs[counter].sample_period) + continue; + + attrs[counter].sample_period = default_interval; + } + + nr_cpus = sysconf(_SC_NPROCESSORS_ONLN); + assert(nr_cpus <= MAX_NR_CPUS); + assert(nr_cpus >= 0); + + if (target_pid != -1 || profile_cpu != -1) + nr_cpus = 1; + + return __cmd_top(); +} diff --git a/tools/perf/builtin.h b/tools/perf/builtin.h new file mode 100644 index 00000000000..51d168230ee --- /dev/null +++ b/tools/perf/builtin.h @@ -0,0 +1,26 @@ +#ifndef BUILTIN_H +#define BUILTIN_H + +#include "util/util.h" +#include "util/strbuf.h" + +extern const char perf_version_string[]; +extern const char perf_usage_string[]; +extern const char perf_more_info_string[]; + +extern void list_common_cmds_help(void); +extern const char *help_unknown_cmd(const char *cmd); +extern void prune_packed_objects(int); +extern int read_line_with_nul(char *buf, int size, FILE *file); +extern int check_pager_config(const char *cmd); + +extern int cmd_annotate(int argc, const char **argv, const char *prefix); +extern int cmd_help(int argc, const char **argv, const char *prefix); +extern int cmd_record(int argc, const char **argv, const char *prefix); +extern int cmd_report(int argc, const char **argv, const char *prefix); +extern int cmd_stat(int argc, const char **argv, const char *prefix); +extern int cmd_top(int argc, const char **argv, const char *prefix); +extern int cmd_version(int argc, const char **argv, const char *prefix); +extern int cmd_list(int argc, const char **argv, const char *prefix); + +#endif diff --git a/tools/perf/command-list.txt b/tools/perf/command-list.txt new file mode 100644 index 00000000000..eebce30afbc --- /dev/null +++ b/tools/perf/command-list.txt @@ -0,0 +1,10 @@ +# +# List of known perf commands. +# command name category [deprecated] [common] +# +perf-annotate mainporcelain common +perf-list mainporcelain common +perf-record mainporcelain common +perf-report mainporcelain common +perf-stat mainporcelain common +perf-top mainporcelain common diff --git a/tools/perf/design.txt b/tools/perf/design.txt new file mode 100644 index 00000000000..860e116d979 --- /dev/null +++ b/tools/perf/design.txt @@ -0,0 +1,442 @@ + +Performance Counters for Linux +------------------------------ + +Performance counters are special hardware registers available on most modern +CPUs. These registers count the number of certain types of hw events: such +as instructions executed, cachemisses suffered, or branches mis-predicted - +without slowing down the kernel or applications. These registers can also +trigger interrupts when a threshold number of events have passed - and can +thus be used to profile the code that runs on that CPU. + +The Linux Performance Counter subsystem provides an abstraction of these +hardware capabilities. It provides per task and per CPU counters, counter +groups, and it provides event capabilities on top of those. It +provides "virtual" 64-bit counters, regardless of the width of the +underlying hardware counters. + +Performance counters are accessed via special file descriptors. +There's one file descriptor per virtual counter used. + +The special file descriptor is opened via the perf_counter_open() +system call: + + int sys_perf_counter_open(struct perf_counter_hw_event *hw_event_uptr, + pid_t pid, int cpu, int group_fd, + unsigned long flags); + +The syscall returns the new fd. The fd can be used via the normal +VFS system calls: read() can be used to read the counter, fcntl() +can be used to set the blocking mode, etc. + +Multiple counters can be kept open at a time, and the counters +can be poll()ed. + +When creating a new counter fd, 'perf_counter_hw_event' is: + +struct perf_counter_hw_event { + /* + * The MSB of the config word signifies if the rest contains cpu + * specific (raw) counter configuration data, if unset, the next + * 7 bits are an event type and the rest of the bits are the event + * identifier. + */ + __u64 config; + + __u64 irq_period; + __u32 record_type; + __u32 read_format; + + __u64 disabled : 1, /* off by default */ + inherit : 1, /* children inherit it */ + pinned : 1, /* must always be on PMU */ + exclusive : 1, /* only group on PMU */ + exclude_user : 1, /* don't count user */ + exclude_kernel : 1, /* ditto kernel */ + exclude_hv : 1, /* ditto hypervisor */ + exclude_idle : 1, /* don't count when idle */ + mmap : 1, /* include mmap data */ + munmap : 1, /* include munmap data */ + comm : 1, /* include comm data */ + + __reserved_1 : 52; + + __u32 extra_config_len; + __u32 wakeup_events; /* wakeup every n events */ + + __u64 __reserved_2; + __u64 __reserved_3; +}; + +The 'config' field specifies what the counter should count. It +is divided into 3 bit-fields: + +raw_type: 1 bit (most significant bit) 0x8000_0000_0000_0000 +type: 7 bits (next most significant) 0x7f00_0000_0000_0000 +event_id: 56 bits (least significant) 0x00ff_ffff_ffff_ffff + +If 'raw_type' is 1, then the counter will count a hardware event +specified by the remaining 63 bits of event_config. The encoding is +machine-specific. + +If 'raw_type' is 0, then the 'type' field says what kind of counter +this is, with the following encoding: + +enum perf_event_types { + PERF_TYPE_HARDWARE = 0, + PERF_TYPE_SOFTWARE = 1, + PERF_TYPE_TRACEPOINT = 2, +}; + +A counter of PERF_TYPE_HARDWARE will count the hardware event +specified by 'event_id': + +/* + * Generalized performance counter event types, used by the hw_event.event_id + * parameter of the sys_perf_counter_open() syscall: + */ +enum hw_event_ids { + /* + * Common hardware events, generalized by the kernel: + */ + PERF_COUNT_HW_CPU_CYCLES = 0, + PERF_COUNT_HW_INSTRUCTIONS = 1, + PERF_COUNT_HW_CACHE_REFERENCES = 2, + PERF_COUNT_HW_CACHE_MISSES = 3, + PERF_COUNT_HW_BRANCH_INSTRUCTIONS = 4, + PERF_COUNT_HW_BRANCH_MISSES = 5, + PERF_COUNT_HW_BUS_CYCLES = 6, +}; + +These are standardized types of events that work relatively uniformly +on all CPUs that implement Performance Counters support under Linux, +although there may be variations (e.g., different CPUs might count +cache references and misses at different levels of the cache hierarchy). +If a CPU is not able to count the selected event, then the system call +will return -EINVAL. + +More hw_event_types are supported as well, but they are CPU-specific +and accessed as raw events. For example, to count "External bus +cycles while bus lock signal asserted" events on Intel Core CPUs, pass +in a 0x4064 event_id value and set hw_event.raw_type to 1. + +A counter of type PERF_TYPE_SOFTWARE will count one of the available +software events, selected by 'event_id': + +/* + * Special "software" counters provided by the kernel, even if the hardware + * does not support performance counters. These counters measure various + * physical and sw events of the kernel (and allow the profiling of them as + * well): + */ +enum sw_event_ids { + PERF_COUNT_SW_CPU_CLOCK = 0, + PERF_COUNT_SW_TASK_CLOCK = 1, + PERF_COUNT_SW_PAGE_FAULTS = 2, + PERF_COUNT_SW_CONTEXT_SWITCHES = 3, + PERF_COUNT_SW_CPU_MIGRATIONS = 4, + PERF_COUNT_SW_PAGE_FAULTS_MIN = 5, + PERF_COUNT_SW_PAGE_FAULTS_MAJ = 6, +}; + +Counters of the type PERF_TYPE_TRACEPOINT are available when the ftrace event +tracer is available, and event_id values can be obtained from +/debug/tracing/events/*/*/id + + +Counters come in two flavours: counting counters and sampling +counters. A "counting" counter is one that is used for counting the +number of events that occur, and is characterised by having +irq_period = 0. + + +A read() on a counter returns the current value of the counter and possible +additional values as specified by 'read_format', each value is a u64 (8 bytes) +in size. + +/* + * Bits that can be set in hw_event.read_format to request that + * reads on the counter should return the indicated quantities, + * in increasing order of bit value, after the counter value. + */ +enum perf_counter_read_format { + PERF_FORMAT_TOTAL_TIME_ENABLED = 1, + PERF_FORMAT_TOTAL_TIME_RUNNING = 2, +}; + +Using these additional values one can establish the overcommit ratio for a +particular counter allowing one to take the round-robin scheduling effect +into account. + + +A "sampling" counter is one that is set up to generate an interrupt +every N events, where N is given by 'irq_period'. A sampling counter +has irq_period > 0. The record_type controls what data is recorded on each +interrupt: + +/* + * Bits that can be set in hw_event.record_type to request information + * in the overflow packets. + */ +enum perf_counter_record_format { + PERF_RECORD_IP = 1U << 0, + PERF_RECORD_TID = 1U << 1, + PERF_RECORD_TIME = 1U << 2, + PERF_RECORD_ADDR = 1U << 3, + PERF_RECORD_GROUP = 1U << 4, + PERF_RECORD_CALLCHAIN = 1U << 5, +}; + +Such (and other) events will be recorded in a ring-buffer, which is +available to user-space using mmap() (see below). + +The 'disabled' bit specifies whether the counter starts out disabled +or enabled. If it is initially disabled, it can be enabled by ioctl +or prctl (see below). + +The 'inherit' bit, if set, specifies that this counter should count +events on descendant tasks as well as the task specified. This only +applies to new descendents, not to any existing descendents at the +time the counter is created (nor to any new descendents of existing +descendents). + +The 'pinned' bit, if set, specifies that the counter should always be +on the CPU if at all possible. It only applies to hardware counters +and only to group leaders. If a pinned counter cannot be put onto the +CPU (e.g. because there are not enough hardware counters or because of +a conflict with some other event), then the counter goes into an +'error' state, where reads return end-of-file (i.e. read() returns 0) +until the counter is subsequently enabled or disabled. + +The 'exclusive' bit, if set, specifies that when this counter's group +is on the CPU, it should be the only group using the CPU's counters. +In future, this will allow sophisticated monitoring programs to supply +extra configuration information via 'extra_config_len' to exploit +advanced features of the CPU's Performance Monitor Unit (PMU) that are +not otherwise accessible and that might disrupt other hardware +counters. + +The 'exclude_user', 'exclude_kernel' and 'exclude_hv' bits provide a +way to request that counting of events be restricted to times when the +CPU is in user, kernel and/or hypervisor mode. + +The 'mmap' and 'munmap' bits allow recording of PROT_EXEC mmap/munmap +operations, these can be used to relate userspace IP addresses to actual +code, even after the mapping (or even the whole process) is gone, +these events are recorded in the ring-buffer (see below). + +The 'comm' bit allows tracking of process comm data on process creation. +This too is recorded in the ring-buffer (see below). + +The 'pid' parameter to the perf_counter_open() system call allows the +counter to be specific to a task: + + pid == 0: if the pid parameter is zero, the counter is attached to the + current task. + + pid > 0: the counter is attached to a specific task (if the current task + has sufficient privilege to do so) + + pid < 0: all tasks are counted (per cpu counters) + +The 'cpu' parameter allows a counter to be made specific to a CPU: + + cpu >= 0: the counter is restricted to a specific CPU + cpu == -1: the counter counts on all CPUs + +(Note: the combination of 'pid == -1' and 'cpu == -1' is not valid.) + +A 'pid > 0' and 'cpu == -1' counter is a per task counter that counts +events of that task and 'follows' that task to whatever CPU the task +gets schedule to. Per task counters can be created by any user, for +their own tasks. + +A 'pid == -1' and 'cpu == x' counter is a per CPU counter that counts +all events on CPU-x. Per CPU counters need CAP_SYS_ADMIN privilege. + +The 'flags' parameter is currently unused and must be zero. + +The 'group_fd' parameter allows counter "groups" to be set up. A +counter group has one counter which is the group "leader". The leader +is created first, with group_fd = -1 in the perf_counter_open call +that creates it. The rest of the group members are created +subsequently, with group_fd giving the fd of the group leader. +(A single counter on its own is created with group_fd = -1 and is +considered to be a group with only 1 member.) + +A counter group is scheduled onto the CPU as a unit, that is, it will +only be put onto the CPU if all of the counters in the group can be +put onto the CPU. This means that the values of the member counters +can be meaningfully compared, added, divided (to get ratios), etc., +with each other, since they have counted events for the same set of +executed instructions. + + +Like stated, asynchronous events, like counter overflow or PROT_EXEC mmap +tracking are logged into a ring-buffer. This ring-buffer is created and +accessed through mmap(). + +The mmap size should be 1+2^n pages, where the first page is a meta-data page +(struct perf_counter_mmap_page) that contains various bits of information such +as where the ring-buffer head is. + +/* + * Structure of the page that can be mapped via mmap + */ +struct perf_counter_mmap_page { + __u32 version; /* version number of this structure */ + __u32 compat_version; /* lowest version this is compat with */ + + /* + * Bits needed to read the hw counters in user-space. + * + * u32 seq; + * s64 count; + * + * do { + * seq = pc->lock; + * + * barrier() + * if (pc->index) { + * count = pmc_read(pc->index - 1); + * count += pc->offset; + * } else + * goto regular_read; + * + * barrier(); + * } while (pc->lock != seq); + * + * NOTE: for obvious reason this only works on self-monitoring + * processes. + */ + __u32 lock; /* seqlock for synchronization */ + __u32 index; /* hardware counter identifier */ + __s64 offset; /* add to hardware counter value */ + + /* + * Control data for the mmap() data buffer. + * + * User-space reading this value should issue an rmb(), on SMP capable + * platforms, after reading this value -- see perf_counter_wakeup(). + */ + __u32 data_head; /* head in the data section */ +}; + +NOTE: the hw-counter userspace bits are arch specific and are currently only + implemented on powerpc. + +The following 2^n pages are the ring-buffer which contains events of the form: + +#define PERF_EVENT_MISC_KERNEL (1 << 0) +#define PERF_EVENT_MISC_USER (1 << 1) +#define PERF_EVENT_MISC_OVERFLOW (1 << 2) + +struct perf_event_header { + __u32 type; + __u16 misc; + __u16 size; +}; + +enum perf_event_type { + + /* + * The MMAP events record the PROT_EXEC mappings so that we can + * correlate userspace IPs to code. They have the following structure: + * + * struct { + * struct perf_event_header header; + * + * u32 pid, tid; + * u64 addr; + * u64 len; + * u64 pgoff; + * char filename[]; + * }; + */ + PERF_EVENT_MMAP = 1, + PERF_EVENT_MUNMAP = 2, + + /* + * struct { + * struct perf_event_header header; + * + * u32 pid, tid; + * char comm[]; + * }; + */ + PERF_EVENT_COMM = 3, + + /* + * When header.misc & PERF_EVENT_MISC_OVERFLOW the event_type field + * will be PERF_RECORD_* + * + * struct { + * struct perf_event_header header; + * + * { u64 ip; } && PERF_RECORD_IP + * { u32 pid, tid; } && PERF_RECORD_TID + * { u64 time; } && PERF_RECORD_TIME + * { u64 addr; } && PERF_RECORD_ADDR + * + * { u64 nr; + * { u64 event, val; } cnt[nr]; } && PERF_RECORD_GROUP + * + * { u16 nr, + * hv, + * kernel, + * user; + * u64 ips[nr]; } && PERF_RECORD_CALLCHAIN + * }; + */ +}; + +NOTE: PERF_RECORD_CALLCHAIN is arch specific and currently only implemented + on x86. + +Notification of new events is possible through poll()/select()/epoll() and +fcntl() managing signals. + +Normally a notification is generated for every page filled, however one can +additionally set perf_counter_hw_event.wakeup_events to generate one every +so many counter overflow events. + +Future work will include a splice() interface to the ring-buffer. + + +Counters can be enabled and disabled in two ways: via ioctl and via +prctl. When a counter is disabled, it doesn't count or generate +events but does continue to exist and maintain its count value. + +An individual counter or counter group can be enabled with + + ioctl(fd, PERF_COUNTER_IOC_ENABLE); + +or disabled with + + ioctl(fd, PERF_COUNTER_IOC_DISABLE); + +Enabling or disabling the leader of a group enables or disables the +whole group; that is, while the group leader is disabled, none of the +counters in the group will count. Enabling or disabling a member of a +group other than the leader only affects that counter - disabling an +non-leader stops that counter from counting but doesn't affect any +other counter. + +Additionally, non-inherited overflow counters can use + + ioctl(fd, PERF_COUNTER_IOC_REFRESH, nr); + +to enable a counter for 'nr' events, after which it gets disabled again. + +A process can enable or disable all the counter groups that are +attached to it, using prctl: + + prctl(PR_TASK_PERF_COUNTERS_ENABLE); + + prctl(PR_TASK_PERF_COUNTERS_DISABLE); + +This applies to all counters on the current process, whether created +by this process or by another, and doesn't affect any counters that +this process has created on other processes. It only enables or +disables the group leaders, not any other members in the groups. + diff --git a/tools/perf/perf.c b/tools/perf/perf.c new file mode 100644 index 00000000000..4eb72593370 --- /dev/null +++ b/tools/perf/perf.c @@ -0,0 +1,428 @@ +/* + * perf.c + * + * Performance analysis utility. + * + * This is the main hub from which the sub-commands (perf stat, + * perf top, perf record, perf report, etc.) are started. + */ +#include "builtin.h" + +#include "util/exec_cmd.h" +#include "util/cache.h" +#include "util/quote.h" +#include "util/run-command.h" + +const char perf_usage_string[] = + "perf [--version] [--help] COMMAND [ARGS]"; + +const char perf_more_info_string[] = + "See 'perf help COMMAND' for more information on a specific command."; + +static int use_pager = -1; +struct pager_config { + const char *cmd; + int val; +}; + +static int pager_command_config(const char *var, const char *value, void *data) +{ + struct pager_config *c = data; + if (!prefixcmp(var, "pager.") && !strcmp(var + 6, c->cmd)) + c->val = perf_config_bool(var, value); + return 0; +} + +/* returns 0 for "no pager", 1 for "use pager", and -1 for "not specified" */ +int check_pager_config(const char *cmd) +{ + struct pager_config c; + c.cmd = cmd; + c.val = -1; + perf_config(pager_command_config, &c); + return c.val; +} + +static void commit_pager_choice(void) { + switch (use_pager) { + case 0: + setenv("PERF_PAGER", "cat", 1); + break; + case 1: + /* setup_pager(); */ + break; + default: + break; + } +} + +static int handle_options(const char*** argv, int* argc, int* envchanged) +{ + int handled = 0; + + while (*argc > 0) { + const char *cmd = (*argv)[0]; + if (cmd[0] != '-') + break; + + /* + * For legacy reasons, the "version" and "help" + * commands can be written with "--" prepended + * to make them look like flags. + */ + if (!strcmp(cmd, "--help") || !strcmp(cmd, "--version")) + break; + + /* + * Check remaining flags. + */ + if (!prefixcmp(cmd, "--exec-path")) { + cmd += 11; + if (*cmd == '=') + perf_set_argv_exec_path(cmd + 1); + else { + puts(perf_exec_path()); + exit(0); + } + } else if (!strcmp(cmd, "--html-path")) { + puts(system_path(PERF_HTML_PATH)); + exit(0); + } else if (!strcmp(cmd, "-p") || !strcmp(cmd, "--paginate")) { + use_pager = 1; + } else if (!strcmp(cmd, "--no-pager")) { + use_pager = 0; + if (envchanged) + *envchanged = 1; + } else if (!strcmp(cmd, "--perf-dir")) { + if (*argc < 2) { + fprintf(stderr, "No directory given for --perf-dir.\n" ); + usage(perf_usage_string); + } + setenv(PERF_DIR_ENVIRONMENT, (*argv)[1], 1); + if (envchanged) + *envchanged = 1; + (*argv)++; + (*argc)--; + handled++; + } else if (!prefixcmp(cmd, "--perf-dir=")) { + setenv(PERF_DIR_ENVIRONMENT, cmd + 10, 1); + if (envchanged) + *envchanged = 1; + } else if (!strcmp(cmd, "--work-tree")) { + if (*argc < 2) { + fprintf(stderr, "No directory given for --work-tree.\n" ); + usage(perf_usage_string); + } + setenv(PERF_WORK_TREE_ENVIRONMENT, (*argv)[1], 1); + if (envchanged) + *envchanged = 1; + (*argv)++; + (*argc)--; + } else if (!prefixcmp(cmd, "--work-tree=")) { + setenv(PERF_WORK_TREE_ENVIRONMENT, cmd + 12, 1); + if (envchanged) + *envchanged = 1; + } else { + fprintf(stderr, "Unknown option: %s\n", cmd); + usage(perf_usage_string); + } + + (*argv)++; + (*argc)--; + handled++; + } + return handled; +} + +static int handle_alias(int *argcp, const char ***argv) +{ + int envchanged = 0, ret = 0, saved_errno = errno; + int count, option_count; + const char** new_argv; + const char *alias_command; + char *alias_string; + + alias_command = (*argv)[0]; + alias_string = alias_lookup(alias_command); + if (alias_string) { + if (alias_string[0] == '!') { + if (*argcp > 1) { + struct strbuf buf; + + strbuf_init(&buf, PATH_MAX); + strbuf_addstr(&buf, alias_string); + sq_quote_argv(&buf, (*argv) + 1, PATH_MAX); + free(alias_string); + alias_string = buf.buf; + } + ret = system(alias_string + 1); + if (ret >= 0 && WIFEXITED(ret) && + WEXITSTATUS(ret) != 127) + exit(WEXITSTATUS(ret)); + die("Failed to run '%s' when expanding alias '%s'", + alias_string + 1, alias_command); + } + count = split_cmdline(alias_string, &new_argv); + if (count < 0) + die("Bad alias.%s string", alias_command); + option_count = handle_options(&new_argv, &count, &envchanged); + if (envchanged) + die("alias '%s' changes environment variables\n" + "You can use '!perf' in the alias to do this.", + alias_command); + memmove(new_argv - option_count, new_argv, + count * sizeof(char *)); + new_argv -= option_count; + + if (count < 1) + die("empty alias for %s", alias_command); + + if (!strcmp(alias_command, new_argv[0])) + die("recursive alias: %s", alias_command); + + new_argv = realloc(new_argv, sizeof(char*) * + (count + *argcp + 1)); + /* insert after command name */ + memcpy(new_argv + count, *argv + 1, sizeof(char*) * *argcp); + new_argv[count+*argcp] = NULL; + + *argv = new_argv; + *argcp += count - 1; + + ret = 1; + } + + errno = saved_errno; + + return ret; +} + +const char perf_version_string[] = PERF_VERSION; + +#define RUN_SETUP (1<<0) +#define USE_PAGER (1<<1) +/* + * require working tree to be present -- anything uses this needs + * RUN_SETUP for reading from the configuration file. + */ +#define NEED_WORK_TREE (1<<2) + +struct cmd_struct { + const char *cmd; + int (*fn)(int, const char **, const char *); + int option; +}; + +static int run_builtin(struct cmd_struct *p, int argc, const char **argv) +{ + int status; + struct stat st; + const char *prefix; + + prefix = NULL; + if (p->option & RUN_SETUP) + prefix = NULL; /* setup_perf_directory(); */ + + if (use_pager == -1 && p->option & RUN_SETUP) + use_pager = check_pager_config(p->cmd); + if (use_pager == -1 && p->option & USE_PAGER) + use_pager = 1; + commit_pager_choice(); + + if (p->option & NEED_WORK_TREE) + /* setup_work_tree() */; + + status = p->fn(argc, argv, prefix); + if (status) + return status & 0xff; + + /* Somebody closed stdout? */ + if (fstat(fileno(stdout), &st)) + return 0; + /* Ignore write errors for pipes and sockets.. */ + if (S_ISFIFO(st.st_mode) || S_ISSOCK(st.st_mode)) + return 0; + + /* Check for ENOSPC and EIO errors.. */ + if (fflush(stdout)) + die("write failure on standard output: %s", strerror(errno)); + if (ferror(stdout)) + die("unknown write failure on standard output"); + if (fclose(stdout)) + die("close failed on standard output: %s", strerror(errno)); + return 0; +} + +static void handle_internal_command(int argc, const char **argv) +{ + const char *cmd = argv[0]; + static struct cmd_struct commands[] = { + { "help", cmd_help, 0 }, + { "list", cmd_list, 0 }, + { "record", cmd_record, 0 }, + { "report", cmd_report, 0 }, + { "stat", cmd_stat, 0 }, + { "top", cmd_top, 0 }, + { "annotate", cmd_annotate, 0 }, + { "version", cmd_version, 0 }, + }; + int i; + static const char ext[] = STRIP_EXTENSION; + + if (sizeof(ext) > 1) { + i = strlen(argv[0]) - strlen(ext); + if (i > 0 && !strcmp(argv[0] + i, ext)) { + char *argv0 = strdup(argv[0]); + argv[0] = cmd = argv0; + argv0[i] = '\0'; + } + } + + /* Turn "perf cmd --help" into "perf help cmd" */ + if (argc > 1 && !strcmp(argv[1], "--help")) { + argv[1] = argv[0]; + argv[0] = cmd = "help"; + } + + for (i = 0; i < ARRAY_SIZE(commands); i++) { + struct cmd_struct *p = commands+i; + if (strcmp(p->cmd, cmd)) + continue; + exit(run_builtin(p, argc, argv)); + } +} + +static void execv_dashed_external(const char **argv) +{ + struct strbuf cmd = STRBUF_INIT; + const char *tmp; + int status; + + strbuf_addf(&cmd, "perf-%s", argv[0]); + + /* + * argv[0] must be the perf command, but the argv array + * belongs to the caller, and may be reused in + * subsequent loop iterations. Save argv[0] and + * restore it on error. + */ + tmp = argv[0]; + argv[0] = cmd.buf; + + /* + * if we fail because the command is not found, it is + * OK to return. Otherwise, we just pass along the status code. + */ + status = run_command_v_opt(argv, 0); + if (status != -ERR_RUN_COMMAND_EXEC) { + if (IS_RUN_COMMAND_ERR(status)) + die("unable to run '%s'", argv[0]); + exit(-status); + } + errno = ENOENT; /* as if we called execvp */ + + argv[0] = tmp; + + strbuf_release(&cmd); +} + +static int run_argv(int *argcp, const char ***argv) +{ + int done_alias = 0; + + while (1) { + /* See if it's an internal command */ + handle_internal_command(*argcp, *argv); + + /* .. then try the external ones */ + execv_dashed_external(*argv); + + /* It could be an alias -- this works around the insanity + * of overriding "perf log" with "perf show" by having + * alias.log = show + */ + if (done_alias || !handle_alias(argcp, argv)) + break; + done_alias = 1; + } + + return done_alias; +} + + +int main(int argc, const char **argv) +{ + const char *cmd; + + cmd = perf_extract_argv0_path(argv[0]); + if (!cmd) + cmd = "perf-help"; + + /* + * "perf-xxxx" is the same as "perf xxxx", but we obviously: + * + * - cannot take flags in between the "perf" and the "xxxx". + * - cannot execute it externally (since it would just do + * the same thing over again) + * + * So we just directly call the internal command handler, and + * die if that one cannot handle it. + */ + if (!prefixcmp(cmd, "perf-")) { + cmd += 5; + argv[0] = cmd; + handle_internal_command(argc, argv); + die("cannot handle %s internally", cmd); + } + + /* Look for flags.. */ + argv++; + argc--; + handle_options(&argv, &argc, NULL); + commit_pager_choice(); + if (argc > 0) { + if (!prefixcmp(argv[0], "--")) + argv[0] += 2; + } else { + /* The user didn't specify a command; give them help */ + printf("\n usage: %s\n\n", perf_usage_string); + list_common_cmds_help(); + printf("\n %s\n\n", perf_more_info_string); + exit(1); + } + cmd = argv[0]; + + /* + * We use PATH to find perf commands, but we prepend some higher + * precidence paths: the "--exec-path" option, the PERF_EXEC_PATH + * environment, and the $(perfexecdir) from the Makefile at build + * time. + */ + setup_path(); + + while (1) { + static int done_help = 0; + static int was_alias = 0; + + was_alias = run_argv(&argc, &argv); + if (errno != ENOENT) + break; + + if (was_alias) { + fprintf(stderr, "Expansion of alias '%s' failed; " + "'%s' is not a perf-command\n", + cmd, argv[0]); + exit(1); + } + if (!done_help) { + cmd = argv[0] = help_unknown_cmd(cmd); + done_help = 1; + } else + break; + } + + fprintf(stderr, "Failed to run command '%s': %s\n", + cmd, strerror(errno)); + + return 1; +} diff --git a/tools/perf/perf.h b/tools/perf/perf.h new file mode 100644 index 00000000000..af0a5046d74 --- /dev/null +++ b/tools/perf/perf.h @@ -0,0 +1,67 @@ +#ifndef _PERF_PERF_H +#define _PERF_PERF_H + +#if defined(__x86_64__) || defined(__i386__) +#include "../../arch/x86/include/asm/unistd.h" +#define rmb() asm volatile("lfence" ::: "memory") +#define cpu_relax() asm volatile("rep; nop" ::: "memory"); +#endif + +#ifdef __powerpc__ +#include "../../arch/powerpc/include/asm/unistd.h" +#define rmb() asm volatile ("sync" ::: "memory") +#define cpu_relax() asm volatile ("" ::: "memory"); +#endif + +#include <time.h> +#include <unistd.h> +#include <sys/types.h> +#include <sys/syscall.h> + +#include "../../include/linux/perf_counter.h" + +/* + * prctl(PR_TASK_PERF_COUNTERS_DISABLE) will (cheaply) disable all + * counters in the current task. + */ +#define PR_TASK_PERF_COUNTERS_DISABLE 31 +#define PR_TASK_PERF_COUNTERS_ENABLE 32 + +#ifndef NSEC_PER_SEC +# define NSEC_PER_SEC 1000000000ULL +#endif + +static inline unsigned long long rdclock(void) +{ + struct timespec ts; + + clock_gettime(CLOCK_MONOTONIC, &ts); + return ts.tv_sec * 1000000000ULL + ts.tv_nsec; +} + +/* + * Pick up some kernel type conventions: + */ +#define __user +#define asmlinkage + +#define unlikely(x) __builtin_expect(!!(x), 0) +#define min(x, y) ({ \ + typeof(x) _min1 = (x); \ + typeof(y) _min2 = (y); \ + (void) (&_min1 == &_min2); \ + _min1 < _min2 ? _min1 : _min2; }) + +static inline int +sys_perf_counter_open(struct perf_counter_attr *attr_uptr, + pid_t pid, int cpu, int group_fd, + unsigned long flags) +{ + return syscall(__NR_perf_counter_open, attr_uptr, pid, cpu, + group_fd, flags); +} + +#define MAX_COUNTERS 256 +#define MAX_NR_CPUS 256 + +#endif diff --git a/tools/perf/util/PERF-VERSION-GEN b/tools/perf/util/PERF-VERSION-GEN new file mode 100755 index 00000000000..c561d1538c0 --- /dev/null +++ b/tools/perf/util/PERF-VERSION-GEN @@ -0,0 +1,42 @@ +#!/bin/sh + +GVF=PERF-VERSION-FILE +DEF_VER=v0.0.1.PERF + +LF=' +' + +# First see if there is a version file (included in release tarballs), +# then try git-describe, then default. +if test -f version +then + VN=$(cat version) || VN="$DEF_VER" +elif test -d .git -o -f .git && + VN=$(git describe --abbrev=4 HEAD 2>/dev/null) && + case "$VN" in + *$LF*) (exit 1) ;; + v[0-9]*) + git update-index -q --refresh + test -z "$(git diff-index --name-only HEAD --)" || + VN="$VN-dirty" ;; + esac +then + VN=$(echo "$VN" | sed -e 's/-/./g'); +else + VN="$DEF_VER" +fi + +VN=$(expr "$VN" : v*'\(.*\)') + +if test -r $GVF +then + VC=$(sed -e 's/^PERF_VERSION = //' <$GVF) +else + VC=unset +fi +test "$VN" = "$VC" || { + echo >&2 "PERF_VERSION = $VN" + echo "PERF_VERSION = $VN" >$GVF +} + + diff --git a/tools/perf/util/abspath.c b/tools/perf/util/abspath.c new file mode 100644 index 00000000000..61d33b81fc9 --- /dev/null +++ b/tools/perf/util/abspath.c @@ -0,0 +1,117 @@ +#include "cache.h" + +/* + * Do not use this for inspecting *tracked* content. When path is a + * symlink to a directory, we do not want to say it is a directory when + * dealing with tracked content in the working tree. + */ +static int is_directory(const char *path) +{ + struct stat st; + return (!stat(path, &st) && S_ISDIR(st.st_mode)); +} + +/* We allow "recursive" symbolic links. Only within reason, though. */ +#define MAXDEPTH 5 + +const char *make_absolute_path(const char *path) +{ + static char bufs[2][PATH_MAX + 1], *buf = bufs[0], *next_buf = bufs[1]; + char cwd[1024] = ""; + int buf_index = 1, len; + + int depth = MAXDEPTH; + char *last_elem = NULL; + struct stat st; + + if (strlcpy(buf, path, PATH_MAX) >= PATH_MAX) + die ("Too long path: %.*s", 60, path); + + while (depth--) { + if (!is_directory(buf)) { + char *last_slash = strrchr(buf, '/'); + if (last_slash) { + *last_slash = '\0'; + last_elem = xstrdup(last_slash + 1); + } else { + last_elem = xstrdup(buf); + *buf = '\0'; + } + } + + if (*buf) { + if (!*cwd && !getcwd(cwd, sizeof(cwd))) + die ("Could not get current working directory"); + + if (chdir(buf)) + die ("Could not switch to '%s'", buf); + } + if (!getcwd(buf, PATH_MAX)) + die ("Could not get current working directory"); + + if (last_elem) { + int len = strlen(buf); + if (len + strlen(last_elem) + 2 > PATH_MAX) + die ("Too long path name: '%s/%s'", + buf, last_elem); + buf[len] = '/'; + strcpy(buf + len + 1, last_elem); + free(last_elem); + last_elem = NULL; + } + + if (!lstat(buf, &st) && S_ISLNK(st.st_mode)) { + len = readlink(buf, next_buf, PATH_MAX); + if (len < 0) + die ("Invalid symlink: %s", buf); + if (PATH_MAX <= len) + die("symbolic link too long: %s", buf); + next_buf[len] = '\0'; + buf = next_buf; + buf_index = 1 - buf_index; + next_buf = bufs[buf_index]; + } else + break; + } + + if (*cwd && chdir(cwd)) + die ("Could not change back to '%s'", cwd); + + return buf; +} + +static const char *get_pwd_cwd(void) +{ + static char cwd[PATH_MAX + 1]; + char *pwd; + struct stat cwd_stat, pwd_stat; + if (getcwd(cwd, PATH_MAX) == NULL) + return NULL; + pwd = getenv("PWD"); + if (pwd && strcmp(pwd, cwd)) { + stat(cwd, &cwd_stat); + if (!stat(pwd, &pwd_stat) && + pwd_stat.st_dev == cwd_stat.st_dev && + pwd_stat.st_ino == cwd_stat.st_ino) { + strlcpy(cwd, pwd, PATH_MAX); + } + } + return cwd; +} + +const char *make_nonrelative_path(const char *path) +{ + static char buf[PATH_MAX + 1]; + + if (is_absolute_path(path)) { + if (strlcpy(buf, path, PATH_MAX) >= PATH_MAX) + die("Too long path: %.*s", 60, path); + } else { + const char *cwd = get_pwd_cwd(); + if (!cwd) + die("Cannot determine the current working directory"); + if (snprintf(buf, PATH_MAX, "%s/%s", cwd, path) >= PATH_MAX) + die("Too long path: %.*s", 60, path); + } + return buf; +} diff --git a/tools/perf/util/alias.c b/tools/perf/util/alias.c new file mode 100644 index 00000000000..9b3dd2b428d --- /dev/null +++ b/tools/perf/util/alias.c @@ -0,0 +1,77 @@ +#include "cache.h" + +static const char *alias_key; +static char *alias_val; + +static int alias_lookup_cb(const char *k, const char *v, void *cb) +{ + if (!prefixcmp(k, "alias.") && !strcmp(k+6, alias_key)) { + if (!v) + return config_error_nonbool(k); + alias_val = strdup(v); + return 0; + } + return 0; +} + +char *alias_lookup(const char *alias) +{ + alias_key = alias; + alias_val = NULL; + perf_config(alias_lookup_cb, NULL); + return alias_val; +} + +int split_cmdline(char *cmdline, const char ***argv) +{ + int src, dst, count = 0, size = 16; + char quoted = 0; + + *argv = malloc(sizeof(char*) * size); + + /* split alias_string */ + (*argv)[count++] = cmdline; + for (src = dst = 0; cmdline[src];) { + char c = cmdline[src]; + if (!quoted && isspace(c)) { + cmdline[dst++] = 0; + while (cmdline[++src] + && isspace(cmdline[src])) + ; /* skip */ + if (count >= size) { + size += 16; + *argv = realloc(*argv, sizeof(char*) * size); + } + (*argv)[count++] = cmdline + dst; + } else if (!quoted && (c == '\'' || c == '"')) { + quoted = c; + src++; + } else if (c == quoted) { + quoted = 0; + src++; + } else { + if (c == '\\' && quoted != '\'') { + src++; + c = cmdline[src]; + if (!c) { + free(*argv); + *argv = NULL; + return error("cmdline ends with \\"); + } + } + cmdline[dst++] = c; + src++; + } + } + + cmdline[dst] = 0; + + if (quoted) { + free(*argv); + *argv = NULL; + return error("unclosed quote"); + } + + return count; +} + diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h new file mode 100644 index 00000000000..393d6146d13 --- /dev/null +++ b/tools/perf/util/cache.h @@ -0,0 +1,119 @@ +#ifndef CACHE_H +#define CACHE_H + +#include "util.h" +#include "strbuf.h" + +#define PERF_DIR_ENVIRONMENT "PERF_DIR" +#define PERF_WORK_TREE_ENVIRONMENT "PERF_WORK_TREE" +#define DEFAULT_PERF_DIR_ENVIRONMENT ".perf" +#define DB_ENVIRONMENT "PERF_OBJECT_DIRECTORY" +#define INDEX_ENVIRONMENT "PERF_INDEX_FILE" +#define GRAFT_ENVIRONMENT "PERF_GRAFT_FILE" +#define TEMPLATE_DIR_ENVIRONMENT "PERF_TEMPLATE_DIR" +#define CONFIG_ENVIRONMENT "PERF_CONFIG" +#define EXEC_PATH_ENVIRONMENT "PERF_EXEC_PATH" +#define CEILING_DIRECTORIES_ENVIRONMENT "PERF_CEILING_DIRECTORIES" +#define PERFATTRIBUTES_FILE ".perfattributes" +#define INFOATTRIBUTES_FILE "info/attributes" +#define ATTRIBUTE_MACRO_PREFIX "[attr]" + +typedef int (*config_fn_t)(const char *, const char *, void *); +extern int perf_default_config(const char *, const char *, void *); +extern int perf_config_from_file(config_fn_t fn, const char *, void *); +extern int perf_config(config_fn_t fn, void *); +extern int perf_parse_ulong(const char *, unsigned long *); +extern int perf_config_int(const char *, const char *); +extern unsigned long perf_config_ulong(const char *, const char *); +extern int perf_config_bool_or_int(const char *, const char *, int *); +extern int perf_config_bool(const char *, const char *); +extern int perf_config_string(const char **, const char *, const char *); +extern int perf_config_set(const char *, const char *); +extern int perf_config_set_multivar(const char *, const char *, const char *, int); +extern int perf_config_rename_section(const char *, const char *); +extern const char *perf_etc_perfconfig(void); +extern int check_repository_format_version(const char *var, const char *value, void *cb); +extern int perf_config_system(void); +extern int perf_config_global(void); +extern int config_error_nonbool(const char *); +extern const char *config_exclusive_filename; + +#define MAX_PERFNAME (1000) +extern char perf_default_email[MAX_PERFNAME]; +extern char perf_default_name[MAX_PERFNAME]; +extern int user_ident_explicitly_given; + +extern const char *perf_log_output_encoding; +extern const char *perf_mailmap_file; + +/* IO helper functions */ +extern void maybe_flush_or_die(FILE *, const char *); +extern int copy_fd(int ifd, int ofd); +extern int copy_file(const char *dst, const char *src, int mode); +extern ssize_t read_in_full(int fd, void *buf, size_t count); +extern ssize_t write_in_full(int fd, const void *buf, size_t count); +extern void write_or_die(int fd, const void *buf, size_t count); +extern int write_or_whine(int fd, const void *buf, size_t count, const char *msg); +extern int write_or_whine_pipe(int fd, const void *buf, size_t count, const char *msg); +extern void fsync_or_die(int fd, const char *); + +/* pager.c */ +extern void setup_pager(void); +extern const char *pager_program; +extern int pager_in_use(void); +extern int pager_use_color; + +extern const char *editor_program; +extern const char *excludes_file; + +char *alias_lookup(const char *alias); +int split_cmdline(char *cmdline, const char ***argv); + +#define alloc_nr(x) (((x)+16)*3/2) + +/* + * Realloc the buffer pointed at by variable 'x' so that it can hold + * at least 'nr' entries; the number of entries currently allocated + * is 'alloc', using the standard growing factor alloc_nr() macro. + * + * DO NOT USE any expression with side-effect for 'x' or 'alloc'. + */ +#define ALLOC_GROW(x, nr, alloc) \ + do { \ + if ((nr) > alloc) { \ + if (alloc_nr(alloc) < (nr)) \ + alloc = (nr); \ + else \ + alloc = alloc_nr(alloc); \ + x = xrealloc((x), alloc * sizeof(*(x))); \ + } \ + } while(0) + + +static inline int is_absolute_path(const char *path) +{ + return path[0] == '/'; +} + +const char *make_absolute_path(const char *path); +const char *make_nonrelative_path(const char *path); +const char *make_relative_path(const char *abs, const char *base); +int normalize_path_copy(char *dst, const char *src); +int longest_ancestor_length(const char *path, const char *prefix_list); +char *strip_path_suffix(const char *path, const char *suffix); + +extern char *mkpath(const char *fmt, ...) __attribute__((format (printf, 1, 2))); +extern char *perf_path(const char *fmt, ...) __attribute__((format (printf, 1, 2))); +/* perf_mkstemp() - create tmp file honoring TMPDIR variable */ +extern int perf_mkstemp(char *path, size_t len, const char *template); + +extern char *mksnpath(char *buf, size_t n, const char *fmt, ...) + __attribute__((format (printf, 3, 4))); +extern char *perf_snpath(char *buf, size_t n, const char *fmt, ...) + __attribute__((format (printf, 3, 4))); +extern char *perf_pathdup(const char *fmt, ...) + __attribute__((format (printf, 1, 2))); + +extern size_t strlcpy(char *dest, const char *src, size_t size); + +#endif /* CACHE_H */ diff --git a/tools/perf/util/color.c b/tools/perf/util/color.c new file mode 100644 index 00000000000..9a8c20ccc53 --- /dev/null +++ b/tools/perf/util/color.c @@ -0,0 +1,241 @@ +#include "cache.h" +#include "color.h" + +int perf_use_color_default = -1; + +static int parse_color(const char *name, int len) +{ + static const char * const color_names[] = { + "normal", "black", "red", "green", "yellow", + "blue", "magenta", "cyan", "white" + }; + char *end; + int i; + for (i = 0; i < ARRAY_SIZE(color_names); i++) { + const char *str = color_names[i]; + if (!strncasecmp(name, str, len) && !str[len]) + return i - 1; + } + i = strtol(name, &end, 10); + if (end - name == len && i >= -1 && i <= 255) + return i; + return -2; +} + +static int parse_attr(const char *name, int len) +{ + static const int attr_values[] = { 1, 2, 4, 5, 7 }; + static const char * const attr_names[] = { + "bold", "dim", "ul", "blink", "reverse" + }; + int i; + for (i = 0; i < ARRAY_SIZE(attr_names); i++) { + const char *str = attr_names[i]; + if (!strncasecmp(name, str, len) && !str[len]) + return attr_values[i]; + } + return -1; +} + +void color_parse(const char *value, const char *var, char *dst) +{ + color_parse_mem(value, strlen(value), var, dst); +} + +void color_parse_mem(const char *value, int value_len, const char *var, + char *dst) +{ + const char *ptr = value; + int len = value_len; + int attr = -1; + int fg = -2; + int bg = -2; + + if (!strncasecmp(value, "reset", len)) { + strcpy(dst, PERF_COLOR_RESET); + return; + } + + /* [fg [bg]] [attr] */ + while (len > 0) { + const char *word = ptr; + int val, wordlen = 0; + + while (len > 0 && !isspace(word[wordlen])) { + wordlen++; + len--; + } + + ptr = word + wordlen; + while (len > 0 && isspace(*ptr)) { + ptr++; + len--; + } + + val = parse_color(word, wordlen); + if (val >= -1) { + if (fg == -2) { + fg = val; + continue; + } + if (bg == -2) { + bg = val; + continue; + } + goto bad; + } + val = parse_attr(word, wordlen); + if (val < 0 || attr != -1) + goto bad; + attr = val; + } + + if (attr >= 0 || fg >= 0 || bg >= 0) { + int sep = 0; + + *dst++ = '\033'; + *dst++ = '['; + if (attr >= 0) { + *dst++ = '0' + attr; + sep++; + } + if (fg >= 0) { + if (sep++) + *dst++ = ';'; + if (fg < 8) { + *dst++ = '3'; + *dst++ = '0' + fg; + } else { + dst += sprintf(dst, "38;5;%d", fg); + } + } + if (bg >= 0) { + if (sep++) + *dst++ = ';'; + if (bg < 8) { + *dst++ = '4'; + *dst++ = '0' + bg; + } else { + dst += sprintf(dst, "48;5;%d", bg); + } + } + *dst++ = 'm'; + } + *dst = 0; + return; +bad: + die("bad color value '%.*s' for variable '%s'", value_len, value, var); +} + +int perf_config_colorbool(const char *var, const char *value, int stdout_is_tty) +{ + if (value) { + if (!strcasecmp(value, "never")) + return 0; + if (!strcasecmp(value, "always")) + return 1; + if (!strcasecmp(value, "auto")) + goto auto_color; + } + + /* Missing or explicit false to turn off colorization */ + if (!perf_config_bool(var, value)) + return 0; + + /* any normal truth value defaults to 'auto' */ + auto_color: + if (stdout_is_tty < 0) + stdout_is_tty = isatty(1); + if (stdout_is_tty || (pager_in_use() && pager_use_color)) { + char *term = getenv("TERM"); + if (term && strcmp(term, "dumb")) + return 1; + } + return 0; +} + +int perf_color_default_config(const char *var, const char *value, void *cb) +{ + if (!strcmp(var, "color.ui")) { + perf_use_color_default = perf_config_colorbool(var, value, -1); + return 0; + } + + return perf_default_config(var, value, cb); +} + +static int color_vfprintf(FILE *fp, const char *color, const char *fmt, + va_list args, const char *trail) +{ + int r = 0; + + /* + * Auto-detect: + */ + if (perf_use_color_default < 0) { + if (isatty(1) || pager_in_use()) + perf_use_color_default = 1; + else + perf_use_color_default = 0; + } + + if (perf_use_color_default && *color) + r += fprintf(fp, "%s", color); + r += vfprintf(fp, fmt, args); + if (perf_use_color_default && *color) + r += fprintf(fp, "%s", PERF_COLOR_RESET); + if (trail) + r += fprintf(fp, "%s", trail); + return r; +} + + + +int color_fprintf(FILE *fp, const char *color, const char *fmt, ...) +{ + va_list args; + int r; + + va_start(args, fmt); + r = color_vfprintf(fp, color, fmt, args, NULL); + va_end(args); + return r; +} + +int color_fprintf_ln(FILE *fp, const char *color, const char *fmt, ...) +{ + va_list args; + int r; + va_start(args, fmt); + r = color_vfprintf(fp, color, fmt, args, "\n"); + va_end(args); + return r; +} + +/* + * This function splits the buffer by newlines and colors the lines individually. + * + * Returns 0 on success. + */ +int color_fwrite_lines(FILE *fp, const char *color, + size_t count, const char *buf) +{ + if (!*color) + return fwrite(buf, count, 1, fp) != 1; + while (count) { + char *p = memchr(buf, '\n', count); + if (p != buf && (fputs(color, fp) < 0 || + fwrite(buf, p ? p - buf : count, 1, fp) != 1 || + fputs(PERF_COLOR_RESET, fp) < 0)) + return -1; + if (!p) + return 0; + if (fputc('\n', fp) < 0) + return -1; + count -= p + 1 - buf; + buf = p + 1; + } + return 0; +} + + diff --git a/tools/perf/util/color.h b/tools/perf/util/color.h new file mode 100644 index 00000000000..5abfd379582 --- /dev/null +++ b/tools/perf/util/color.h @@ -0,0 +1,36 @@ +#ifndef COLOR_H +#define COLOR_H + +/* "\033[1;38;5;2xx;48;5;2xxm\0" is 23 bytes */ +#define COLOR_MAXLEN 24 + +#define PERF_COLOR_NORMAL "" +#define PERF_COLOR_RESET "\033[m" +#define PERF_COLOR_BOLD "\033[1m" +#define PERF_COLOR_RED "\033[31m" +#define PERF_COLOR_GREEN "\033[32m" +#define PERF_COLOR_YELLOW "\033[33m" +#define PERF_COLOR_BLUE "\033[34m" +#define PERF_COLOR_MAGENTA "\033[35m" +#define PERF_COLOR_CYAN "\033[36m" +#define PERF_COLOR_BG_RED "\033[41m" + +/* + * This variable stores the value of color.ui + */ +extern int perf_use_color_default; + + +/* + * Use this instead of perf_default_config if you need the value of color.ui. + */ +int perf_color_default_config(const char *var, const char *value, void *cb); + +int perf_config_colorbool(const char *var, const char *value, int stdout_is_tty); +void color_parse(const char *value, const char *var, char *dst); +void color_parse_mem(const char *value, int len, const char *var, char *dst); +int color_fprintf(FILE *fp, const char *color, const char *fmt, ...); +int color_fprintf_ln(FILE *fp, const char *color, const char *fmt, ...); +int color_fwrite_lines(FILE *fp, const char *color, size_t count, const char *buf); + +#endif /* COLOR_H */ diff --git a/tools/perf/util/config.c b/tools/perf/util/config.c new file mode 100644 index 00000000000..3dd13faa6a2 --- /dev/null +++ b/tools/perf/util/config.c @@ -0,0 +1,873 @@ +/* + * GIT - The information manager from hell + * + * Copyright (C) Linus Torvalds, 2005 + * Copyright (C) Johannes Schindelin, 2005 + * + */ +#include "util.h" +#include "cache.h" +#include "exec_cmd.h" + +#define MAXNAME (256) + +static FILE *config_file; +static const char *config_file_name; +static int config_linenr; +static int config_file_eof; + +const char *config_exclusive_filename = NULL; + +static int get_next_char(void) +{ + int c; + FILE *f; + + c = '\n'; + if ((f = config_file) != NULL) { + c = fgetc(f); + if (c == '\r') { + /* DOS like systems */ + c = fgetc(f); + if (c != '\n') { + ungetc(c, f); + c = '\r'; + } + } + if (c == '\n') + config_linenr++; + if (c == EOF) { + config_file_eof = 1; + c = '\n'; + } + } + return c; +} + +static char *parse_value(void) +{ + static char value[1024]; + int quote = 0, comment = 0, len = 0, space = 0; + + for (;;) { + int c = get_next_char(); + if (len >= sizeof(value) - 1) + return NULL; + if (c == '\n') { + if (quote) + return NULL; + value[len] = 0; + return value; + } + if (comment) + continue; + if (isspace(c) && !quote) { + space = 1; + continue; + } + if (!quote) { + if (c == ';' || c == '#') { + comment = 1; + continue; + } + } + if (space) { + if (len) + value[len++] = ' '; + space = 0; + } + if (c == '\\') { + c = get_next_char(); + switch (c) { + case '\n': + continue; + case 't': + c = '\t'; + break; + case 'b': + c = '\b'; + break; + case 'n': + c = '\n'; + break; + /* Some characters escape as themselves */ + case '\\': case '"': + break; + /* Reject unknown escape sequences */ + default: + return NULL; + } + value[len++] = c; + continue; + } + if (c == '"') { + quote = 1-quote; + continue; + } + value[len++] = c; + } +} + +static inline int iskeychar(int c) +{ + return isalnum(c) || c == '-'; +} + +static int get_value(config_fn_t fn, void *data, char *name, unsigned int len) +{ + int c; + char *value; + + /* Get the full name */ + for (;;) { + c = get_next_char(); + if (config_file_eof) + break; + if (!iskeychar(c)) + break; + name[len++] = tolower(c); + if (len >= MAXNAME) + return -1; + } + name[len] = 0; + while (c == ' ' || c == '\t') + c = get_next_char(); + + value = NULL; + if (c != '\n') { + if (c != '=') + return -1; + value = parse_value(); + if (!value) + return -1; + } + return fn(name, value, data); +} + +static int get_extended_base_var(char *name, int baselen, int c) +{ + do { + if (c == '\n') + return -1; + c = get_next_char(); + } while (isspace(c)); + + /* We require the format to be '[base "extension"]' */ + if (c != '"') + return -1; + name[baselen++] = '.'; + + for (;;) { + int c = get_next_char(); + if (c == '\n') + return -1; + if (c == '"') + break; + if (c == '\\') { + c = get_next_char(); + if (c == '\n') + return -1; + } + name[baselen++] = c; + if (baselen > MAXNAME / 2) + return -1; + } + + /* Final ']' */ + if (get_next_char() != ']') + return -1; + return baselen; +} + +static int get_base_var(char *name) +{ + int baselen = 0; + + for (;;) { + int c = get_next_char(); + if (config_file_eof) + return -1; + if (c == ']') + return baselen; + if (isspace(c)) + return get_extended_base_var(name, baselen, c); + if (!iskeychar(c) && c != '.') + return -1; + if (baselen > MAXNAME / 2) + return -1; + name[baselen++] = tolower(c); + } +} + +static int perf_parse_file(config_fn_t fn, void *data) +{ + int comment = 0; + int baselen = 0; + static char var[MAXNAME]; + + /* U+FEFF Byte Order Mark in UTF8 */ + static const unsigned char *utf8_bom = (unsigned char *) "\xef\xbb\xbf"; + const unsigned char *bomptr = utf8_bom; + + for (;;) { + int c = get_next_char(); + if (bomptr && *bomptr) { + /* We are at the file beginning; skip UTF8-encoded BOM + * if present. Sane editors won't put this in on their + * own, but e.g. Windows Notepad will do it happily. */ + if ((unsigned char) c == *bomptr) { + bomptr++; + continue; + } else { + /* Do not tolerate partial BOM. */ + if (bomptr != utf8_bom) + break; + /* No BOM at file beginning. Cool. */ + bomptr = NULL; + } + } + if (c == '\n') { + if (config_file_eof) + return 0; + comment = 0; + continue; + } + if (comment || isspace(c)) + continue; + if (c == '#' || c == ';') { + comment = 1; + continue; + } + if (c == '[') { + baselen = get_base_var(var); + if (baselen <= 0) + break; + var[baselen++] = '.'; + var[baselen] = 0; + continue; + } + if (!isalpha(c)) + break; + var[baselen] = tolower(c); + if (get_value(fn, data, var, baselen+1) < 0) + break; + } + die("bad config file line %d in %s", config_linenr, config_file_name); +} + +static int parse_unit_factor(const char *end, unsigned long *val) +{ + if (!*end) + return 1; + else if (!strcasecmp(end, "k")) { + *val *= 1024; + return 1; + } + else if (!strcasecmp(end, "m")) { + *val *= 1024 * 1024; + return 1; + } + else if (!strcasecmp(end, "g")) { + *val *= 1024 * 1024 * 1024; + return 1; + } + return 0; +} + +static int perf_parse_long(const char *value, long *ret) +{ + if (value && *value) { + char *end; + long val = strtol(value, &end, 0); + unsigned long factor = 1; + if (!parse_unit_factor(end, &factor)) + return 0; + *ret = val * factor; + return 1; + } + return 0; +} + +int perf_parse_ulong(const char *value, unsigned long *ret) +{ + if (value && *value) { + char *end; + unsigned long val = strtoul(value, &end, 0); + if (!parse_unit_factor(end, &val)) + return 0; + *ret = val; + return 1; + } + return 0; +} + +static void die_bad_config(const char *name) +{ + if (config_file_name) + die("bad config value for '%s' in %s", name, config_file_name); + die("bad config value for '%s'", name); +} + +int perf_config_int(const char *name, const char *value) +{ + long ret = 0; + if (!perf_parse_long(value, &ret)) + die_bad_config(name); + return ret; +} + +unsigned long perf_config_ulong(const char *name, const char *value) +{ + unsigned long ret; + if (!perf_parse_ulong(value, &ret)) + die_bad_config(name); + return ret; +} + +int perf_config_bool_or_int(const char *name, const char *value, int *is_bool) +{ + *is_bool = 1; + if (!value) + return 1; + if (!*value) + return 0; + if (!strcasecmp(value, "true") || !strcasecmp(value, "yes") || !strcasecmp(value, "on")) + return 1; + if (!strcasecmp(value, "false") || !strcasecmp(value, "no") || !strcasecmp(value, "off")) + return 0; + *is_bool = 0; + return perf_config_int(name, value); +} + +int perf_config_bool(const char *name, const char *value) +{ + int discard; + return !!perf_config_bool_or_int(name, value, &discard); +} + +int perf_config_string(const char **dest, const char *var, const char *value) +{ + if (!value) + return config_error_nonbool(var); + *dest = strdup(value); + return 0; +} + +static int perf_default_core_config(const char *var, const char *value) +{ + /* Add other config variables here and to Documentation/config.txt. */ + return 0; +} + +int perf_default_config(const char *var, const char *value, void *dummy) +{ + if (!prefixcmp(var, "core.")) + return perf_default_core_config(var, value); + + /* Add other config variables here and to Documentation/config.txt. */ + return 0; +} + +int perf_config_from_file(config_fn_t fn, const char *filename, void *data) +{ + int ret; + FILE *f = fopen(filename, "r"); + + ret = -1; + if (f) { + config_file = f; + config_file_name = filename; + config_linenr = 1; + config_file_eof = 0; + ret = perf_parse_file(fn, data); + fclose(f); + config_file_name = NULL; + } + return ret; +} + +const char *perf_etc_perfconfig(void) +{ + static const char *system_wide; + if (!system_wide) + system_wide = system_path(ETC_PERFCONFIG); + return system_wide; +} + +static int perf_env_bool(const char *k, int def) +{ + const char *v = getenv(k); + return v ? perf_config_bool(k, v) : def; +} + +int perf_config_system(void) +{ + return !perf_env_bool("PERF_CONFIG_NOSYSTEM", 0); +} + +int perf_config_global(void) +{ + return !perf_env_bool("PERF_CONFIG_NOGLOBAL", 0); +} + +int perf_config(config_fn_t fn, void *data) +{ + int ret = 0, found = 0; + char *repo_config = NULL; + const char *home = NULL; + + /* Setting $PERF_CONFIG makes perf read _only_ the given config file. */ + if (config_exclusive_filename) + return perf_config_from_file(fn, config_exclusive_filename, data); + if (perf_config_system() && !access(perf_etc_perfconfig(), R_OK)) { + ret += perf_config_from_file(fn, perf_etc_perfconfig(), + data); + found += 1; + } + + home = getenv("HOME"); + if (perf_config_global() && home) { + char *user_config = strdup(mkpath("%s/.perfconfig", home)); + if (!access(user_config, R_OK)) { + ret += perf_config_from_file(fn, user_config, data); + found += 1; + } + free(user_config); + } + + repo_config = perf_pathdup("config"); + if (!access(repo_config, R_OK)) { + ret += perf_config_from_file(fn, repo_config, data); + found += 1; + } + free(repo_config); + if (found == 0) + return -1; + return ret; +} + +/* + * Find all the stuff for perf_config_set() below. + */ + +#define MAX_MATCHES 512 + +static struct { + int baselen; + char* key; + int do_not_match; + regex_t* value_regex; + int multi_replace; + size_t offset[MAX_MATCHES]; + enum { START, SECTION_SEEN, SECTION_END_SEEN, KEY_SEEN } state; + int seen; +} store; + +static int matches(const char* key, const char* value) +{ + return !strcmp(key, store.key) && + (store.value_regex == NULL || + (store.do_not_match ^ + !regexec(store.value_regex, value, 0, NULL, 0))); +} + +static int store_aux(const char* key, const char* value, void *cb) +{ + const char *ep; + size_t section_len; + + switch (store.state) { + case KEY_SEEN: + if (matches(key, value)) { + if (store.seen == 1 && store.multi_replace == 0) { + warning("%s has multiple values", key); + } else if (store.seen >= MAX_MATCHES) { + error("too many matches for %s", key); + return 1; + } + + store.offset[store.seen] = ftell(config_file); + store.seen++; + } + break; + case SECTION_SEEN: + /* + * What we are looking for is in store.key (both + * section and var), and its section part is baselen + * long. We found key (again, both section and var). + * We would want to know if this key is in the same + * section as what we are looking for. We already + * know we are in the same section as what should + * hold store.key. + */ + ep = strrchr(key, '.'); + section_len = ep - key; + + if ((section_len != store.baselen) || + memcmp(key, store.key, section_len+1)) { + store.state = SECTION_END_SEEN; + break; + } + + /* + * Do not increment matches: this is no match, but we + * just made sure we are in the desired section. + */ + store.offset[store.seen] = ftell(config_file); + /* fallthru */ + case SECTION_END_SEEN: + case START: + if (matches(key, value)) { + store.offset[store.seen] = ftell(config_file); + store.state = KEY_SEEN; + store.seen++; + } else { + if (strrchr(key, '.') - key == store.baselen && + !strncmp(key, store.key, store.baselen)) { + store.state = SECTION_SEEN; + store.offset[store.seen] = ftell(config_file); + } + } + } + return 0; +} + +static int store_write_section(int fd, const char* key) +{ + const char *dot; + int i, success; + struct strbuf sb = STRBUF_INIT; + + dot = memchr(key, '.', store.baselen); + if (dot) { + strbuf_addf(&sb, "[%.*s \"", (int)(dot - key), key); + for (i = dot - key + 1; i < store.baselen; i++) { + if (key[i] == '"' || key[i] == '\\') + strbuf_addch(&sb, '\\'); + strbuf_addch(&sb, key[i]); + } + strbuf_addstr(&sb, "\"]\n"); + } else { + strbuf_addf(&sb, "[%.*s]\n", store.baselen, key); + } + + success = write_in_full(fd, sb.buf, sb.len) == sb.len; + strbuf_release(&sb); + + return success; +} + +static int store_write_pair(int fd, const char* key, const char* value) +{ + int i, success; + int length = strlen(key + store.baselen + 1); + const char *quote = ""; + struct strbuf sb = STRBUF_INIT; + + /* + * Check to see if the value needs to be surrounded with a dq pair. + * Note that problematic characters are always backslash-quoted; this + * check is about not losing leading or trailing SP and strings that + * follow beginning-of-comment characters (i.e. ';' and '#') by the + * configuration parser. + */ + if (value[0] == ' ') + quote = "\""; + for (i = 0; value[i]; i++) + if (value[i] == ';' || value[i] == '#') + quote = "\""; + if (i && value[i - 1] == ' ') + quote = "\""; + + strbuf_addf(&sb, "\t%.*s = %s", + length, key + store.baselen + 1, quote); + + for (i = 0; value[i]; i++) + switch (value[i]) { + case '\n': + strbuf_addstr(&sb, "\\n"); + break; + case '\t': + strbuf_addstr(&sb, "\\t"); + break; + case '"': + case '\\': + strbuf_addch(&sb, '\\'); + default: + strbuf_addch(&sb, value[i]); + break; + } + strbuf_addf(&sb, "%s\n", quote); + + success = write_in_full(fd, sb.buf, sb.len) == sb.len; + strbuf_release(&sb); + + return success; +} + +static ssize_t find_beginning_of_line(const char* contents, size_t size, + size_t offset_, int* found_bracket) +{ + size_t equal_offset = size, bracket_offset = size; + ssize_t offset; + +contline: + for (offset = offset_-2; offset > 0 + && contents[offset] != '\n'; offset--) + switch (contents[offset]) { + case '=': equal_offset = offset; break; + case ']': bracket_offset = offset; break; + } + if (offset > 0 && contents[offset-1] == '\\') { + offset_ = offset; + goto contline; + } + if (bracket_offset < equal_offset) { + *found_bracket = 1; + offset = bracket_offset+1; + } else + offset++; + + return offset; +} + +int perf_config_set(const char* key, const char* value) +{ + return perf_config_set_multivar(key, value, NULL, 0); +} + +/* + * If value==NULL, unset in (remove from) config, + * if value_regex!=NULL, disregard key/value pairs where value does not match. + * if multi_replace==0, nothing, or only one matching key/value is replaced, + * else all matching key/values (regardless how many) are removed, + * before the new pair is written. + * + * Returns 0 on success. + * + * This function does this: + * + * - it locks the config file by creating ".perf/config.lock" + * + * - it then parses the config using store_aux() as validator to find + * the position on the key/value pair to replace. If it is to be unset, + * it must be found exactly once. + * + * - the config file is mmap()ed and the part before the match (if any) is + * written to the lock file, then the changed part and the rest. + * + * - the config file is removed and the lock file rename()d to it. + * + */ +int perf_config_set_multivar(const char* key, const char* value, + const char* value_regex, int multi_replace) +{ + int i, dot; + int fd = -1, in_fd; + int ret = 0; + char* config_filename; + const char* last_dot = strrchr(key, '.'); + + if (config_exclusive_filename) + config_filename = strdup(config_exclusive_filename); + else + config_filename = perf_pathdup("config"); + + /* + * Since "key" actually contains the section name and the real + * key name separated by a dot, we have to know where the dot is. + */ + + if (last_dot == NULL) { + error("key does not contain a section: %s", key); + ret = 2; + goto out_free; + } + store.baselen = last_dot - key; + + store.multi_replace = multi_replace; + + /* + * Validate the key and while at it, lower case it for matching. + */ + store.key = malloc(strlen(key) + 1); + dot = 0; + for (i = 0; key[i]; i++) { + unsigned char c = key[i]; + if (c == '.') + dot = 1; + /* Leave the extended basename untouched.. */ + if (!dot || i > store.baselen) { + if (!iskeychar(c) || (i == store.baselen+1 && !isalpha(c))) { + error("invalid key: %s", key); + free(store.key); + ret = 1; + goto out_free; + } + c = tolower(c); + } else if (c == '\n') { + error("invalid key (newline): %s", key); + free(store.key); + ret = 1; + goto out_free; + } + store.key[i] = c; + } + store.key[i] = 0; + + /* + * If .perf/config does not exist yet, write a minimal version. + */ + in_fd = open(config_filename, O_RDONLY); + if ( in_fd < 0 ) { + free(store.key); + + if ( ENOENT != errno ) { + error("opening %s: %s", config_filename, + strerror(errno)); + ret = 3; /* same as "invalid config file" */ + goto out_free; + } + /* if nothing to unset, error out */ + if (value == NULL) { + ret = 5; + goto out_free; + } + + store.key = (char*)key; + if (!store_write_section(fd, key) || + !store_write_pair(fd, key, value)) + goto write_err_out; + } else { + struct stat st; + char* contents; + size_t contents_sz, copy_begin, copy_end; + int i, new_line = 0; + + if (value_regex == NULL) + store.value_regex = NULL; + else { + if (value_regex[0] == '!') { + store.do_not_match = 1; + value_regex++; + } else + store.do_not_match = 0; + + store.value_regex = (regex_t*)malloc(sizeof(regex_t)); + if (regcomp(store.value_regex, value_regex, + REG_EXTENDED)) { + error("invalid pattern: %s", value_regex); + free(store.value_regex); + ret = 6; + goto out_free; + } + } + + store.offset[0] = 0; + store.state = START; + store.seen = 0; + + /* + * After this, store.offset will contain the *end* offset + * of the last match, or remain at 0 if no match was found. + * As a side effect, we make sure to transform only a valid + * existing config file. + */ + if (perf_config_from_file(store_aux, config_filename, NULL)) { + error("invalid config file %s", config_filename); + free(store.key); + if (store.value_regex != NULL) { + regfree(store.value_regex); + free(store.value_regex); + } + ret = 3; + goto out_free; + } + + free(store.key); + if (store.value_regex != NULL) { + regfree(store.value_regex); + free(store.value_regex); + } + + /* if nothing to unset, or too many matches, error out */ + if ((store.seen == 0 && value == NULL) || + (store.seen > 1 && multi_replace == 0)) { + ret = 5; + goto out_free; + } + + fstat(in_fd, &st); + contents_sz = xsize_t(st.st_size); + contents = mmap(NULL, contents_sz, PROT_READ, + MAP_PRIVATE, in_fd, 0); + close(in_fd); + + if (store.seen == 0) + store.seen = 1; + + for (i = 0, copy_begin = 0; i < store.seen; i++) { + if (store.offset[i] == 0) { + store.offset[i] = copy_end = contents_sz; + } else if (store.state != KEY_SEEN) { + copy_end = store.offset[i]; + } else + copy_end = find_beginning_of_line( + contents, contents_sz, + store.offset[i]-2, &new_line); + + if (copy_end > 0 && contents[copy_end-1] != '\n') + new_line = 1; + + /* write the first part of the config */ + if (copy_end > copy_begin) { + if (write_in_full(fd, contents + copy_begin, + copy_end - copy_begin) < + copy_end - copy_begin) + goto write_err_out; + if (new_line && + write_in_full(fd, "\n", 1) != 1) + goto write_err_out; + } + copy_begin = store.offset[i]; + } + + /* write the pair (value == NULL means unset) */ + if (value != NULL) { + if (store.state == START) { + if (!store_write_section(fd, key)) + goto write_err_out; + } + if (!store_write_pair(fd, key, value)) + goto write_err_out; + } + + /* write the rest of the config */ + if (copy_begin < contents_sz) + if (write_in_full(fd, contents + copy_begin, + contents_sz - copy_begin) < + contents_sz - copy_begin) + goto write_err_out; + + munmap(contents, contents_sz); + } + + ret = 0; + +out_free: + free(config_filename); + return ret; + +write_err_out: + goto out_free; + +} + +/* + * Call this to report error for your variable that should not + * get a boolean value (i.e. "[my] var" means "true"). + */ +int config_error_nonbool(const char *var) +{ + return error("Missing value for '%s'", var); +} diff --git a/tools/perf/util/ctype.c b/tools/perf/util/ctype.c new file mode 100644 index 00000000000..b90ec004f29 --- /dev/null +++ b/tools/perf/util/ctype.c @@ -0,0 +1,26 @@ +/* + * Sane locale-independent, ASCII ctype. + * + * No surprises, and works with signed and unsigned chars. + */ +#include "cache.h" + +enum { + S = GIT_SPACE, + A = GIT_ALPHA, + D = GIT_DIGIT, + G = GIT_GLOB_SPECIAL, /* *, ?, [, \\ */ + R = GIT_REGEX_SPECIAL, /* $, (, ), +, ., ^, {, | * */ +}; + +unsigned char sane_ctype[256] = { + 0, 0, 0, 0, 0, 0, 0, 0, 0, S, S, 0, 0, S, 0, 0, /* 0.. 15 */ + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* 16.. 31 */ + S, 0, 0, 0, R, 0, 0, 0, R, R, G, R, 0, 0, R, 0, /* 32.. 47 */ + D, D, D, D, D, D, D, D, D, D, 0, 0, 0, 0, 0, G, /* 48.. 63 */ + 0, A, A, A, A, A, A, A, A, A, A, A, A, A, A, A, /* 64.. 79 */ + A, A, A, A, A, A, A, A, A, A, A, G, G, 0, R, 0, /* 80.. 95 */ + 0, A, A, A, A, A, A, A, A, A, A, A, A, A, A, A, /* 96..111 */ + A, A, A, A, A, A, A, A, A, A, A, R, R, 0, 0, 0, /* 112..127 */ + /* Nothing in the 128.. range */ +}; diff --git a/tools/perf/util/environment.c b/tools/perf/util/environment.c new file mode 100644 index 00000000000..275b0ee345f --- /dev/null +++ b/tools/perf/util/environment.c @@ -0,0 +1,9 @@ +/* + * We put all the perf config variables in this same object + * file, so that programs can link against the config parser + * without having to link against all the rest of perf. + */ +#include "cache.h" + +const char *pager_program; +int pager_use_color = 1; diff --git a/tools/perf/util/exec_cmd.c b/tools/perf/util/exec_cmd.c new file mode 100644 index 00000000000..d3929226315 --- /dev/null +++ b/tools/perf/util/exec_cmd.c @@ -0,0 +1,165 @@ +#include "cache.h" +#include "exec_cmd.h" +#include "quote.h" +#define MAX_ARGS 32 + +extern char **environ; +static const char *argv_exec_path; +static const char *argv0_path; + +const char *system_path(const char *path) +{ +#ifdef RUNTIME_PREFIX + static const char *prefix; +#else + static const char *prefix = PREFIX; +#endif + struct strbuf d = STRBUF_INIT; + + if (is_absolute_path(path)) + return path; + +#ifdef RUNTIME_PREFIX + assert(argv0_path); + assert(is_absolute_path(argv0_path)); + + if (!prefix && + !(prefix = strip_path_suffix(argv0_path, PERF_EXEC_PATH)) && + !(prefix = strip_path_suffix(argv0_path, BINDIR)) && + !(prefix = strip_path_suffix(argv0_path, "perf"))) { + prefix = PREFIX; + fprintf(stderr, "RUNTIME_PREFIX requested, " + "but prefix computation failed. " + "Using static fallback '%s'.\n", prefix); + } +#endif + + strbuf_addf(&d, "%s/%s", prefix, path); + path = strbuf_detach(&d, NULL); + return path; +} + +const char *perf_extract_argv0_path(const char *argv0) +{ + const char *slash; + + if (!argv0 || !*argv0) + return NULL; + slash = argv0 + strlen(argv0); + + while (argv0 <= slash && !is_dir_sep(*slash)) + slash--; + + if (slash >= argv0) { + argv0_path = strndup(argv0, slash - argv0); + return slash + 1; + } + + return argv0; +} + +void perf_set_argv_exec_path(const char *exec_path) +{ + argv_exec_path = exec_path; + /* + * Propagate this setting to external programs. + */ + setenv(EXEC_PATH_ENVIRONMENT, exec_path, 1); +} + + +/* Returns the highest-priority, location to look for perf programs. */ +const char *perf_exec_path(void) +{ + const char *env; + + if (argv_exec_path) + return argv_exec_path; + + env = getenv(EXEC_PATH_ENVIRONMENT); + if (env && *env) { + return env; + } + + return system_path(PERF_EXEC_PATH); +} + +static void add_path(struct strbuf *out, const char *path) +{ + if (path && *path) { + if (is_absolute_path(path)) + strbuf_addstr(out, path); + else + strbuf_addstr(out, make_nonrelative_path(path)); + + strbuf_addch(out, PATH_SEP); + } +} + +void setup_path(void) +{ + const char *old_path = getenv("PATH"); + struct strbuf new_path = STRBUF_INIT; + + add_path(&new_path, perf_exec_path()); + add_path(&new_path, argv0_path); + + if (old_path) + strbuf_addstr(&new_path, old_path); + else + strbuf_addstr(&new_path, "/usr/local/bin:/usr/bin:/bin"); + + setenv("PATH", new_path.buf, 1); + + strbuf_release(&new_path); +} + +const char **prepare_perf_cmd(const char **argv) +{ + int argc; + const char **nargv; + + for (argc = 0; argv[argc]; argc++) + ; /* just counting */ + nargv = malloc(sizeof(*nargv) * (argc + 2)); + + nargv[0] = "perf"; + for (argc = 0; argv[argc]; argc++) + nargv[argc + 1] = argv[argc]; + nargv[argc + 1] = NULL; + return nargv; +} + +int execv_perf_cmd(const char **argv) { + const char **nargv = prepare_perf_cmd(argv); + + /* execvp() can only ever return if it fails */ + execvp("perf", (char **)nargv); + + free(nargv); + return -1; +} + + +int execl_perf_cmd(const char *cmd,...) +{ + int argc; + const char *argv[MAX_ARGS + 1]; + const char *arg; + va_list param; + + va_start(param, cmd); + argv[0] = cmd; + argc = 1; + while (argc < MAX_ARGS) { + arg = argv[argc++] = va_arg(param, char *); + if (!arg) + break; + } + va_end(param); + if (MAX_ARGS <= argc) + return error("too many args to run %s", cmd); + + argv[argc] = NULL; + return execv_perf_cmd(argv); +} diff --git a/tools/perf/util/exec_cmd.h b/tools/perf/util/exec_cmd.h new file mode 100644 index 00000000000..effe25eb154 --- /dev/null +++ b/tools/perf/util/exec_cmd.h @@ -0,0 +1,13 @@ +#ifndef PERF_EXEC_CMD_H +#define PERF_EXEC_CMD_H + +extern void perf_set_argv_exec_path(const char *exec_path); +extern const char *perf_extract_argv0_path(const char *path); +extern const char *perf_exec_path(void); +extern void setup_path(void); +extern const char **prepare_perf_cmd(const char **argv); +extern int execv_perf_cmd(const char **argv); /* NULL terminated */ +extern int execl_perf_cmd(const char *cmd, ...); +extern const char *system_path(const char *path); + +#endif /* PERF_EXEC_CMD_H */ diff --git a/tools/perf/util/generate-cmdlist.sh b/tools/perf/util/generate-cmdlist.sh new file mode 100755 index 00000000000..f06f6fd148f --- /dev/null +++ b/tools/perf/util/generate-cmdlist.sh @@ -0,0 +1,24 @@ +#!/bin/sh + +echo "/* Automatically generated by $0 */ +struct cmdname_help +{ + char name[16]; + char help[80]; +}; + +static struct cmdname_help common_cmds[] = {" + +sed -n -e 's/^perf-\([^ ]*\)[ ].* common.*/\1/p' command-list.txt | +sort | +while read cmd +do + sed -n ' + /^NAME/,/perf-'"$cmd"'/H + ${ + x + s/.*perf-'"$cmd"' - \(.*\)/ {"'"$cmd"'", "\1"},/ + p + }' "Documentation/perf-$cmd.txt" +done +echo "};" diff --git a/tools/perf/util/help.c b/tools/perf/util/help.c new file mode 100644 index 00000000000..6653f7dd1d7 --- /dev/null +++ b/tools/perf/util/help.c @@ -0,0 +1,367 @@ +#include "cache.h" +#include "../builtin.h" +#include "exec_cmd.h" +#include "levenshtein.h" +#include "help.h" + +/* most GUI terminals set COLUMNS (although some don't export it) */ +static int term_columns(void) +{ + char *col_string = getenv("COLUMNS"); + int n_cols; + + if (col_string && (n_cols = atoi(col_string)) > 0) + return n_cols; + +#ifdef TIOCGWINSZ + { + struct winsize ws; + if (!ioctl(1, TIOCGWINSZ, &ws)) { + if (ws.ws_col) + return ws.ws_col; + } + } +#endif + + return 80; +} + +void add_cmdname(struct cmdnames *cmds, const char *name, int len) +{ + struct cmdname *ent = malloc(sizeof(*ent) + len + 1); + + ent->len = len; + memcpy(ent->name, name, len); + ent->name[len] = 0; + + ALLOC_GROW(cmds->names, cmds->cnt + 1, cmds->alloc); + cmds->names[cmds->cnt++] = ent; +} + +static void clean_cmdnames(struct cmdnames *cmds) +{ + int i; + for (i = 0; i < cmds->cnt; ++i) + free(cmds->names[i]); + free(cmds->names); + cmds->cnt = 0; + cmds->alloc = 0; +} + +static int cmdname_compare(const void *a_, const void *b_) +{ + struct cmdname *a = *(struct cmdname **)a_; + struct cmdname *b = *(struct cmdname **)b_; + return strcmp(a->name, b->name); +} + +static void uniq(struct cmdnames *cmds) +{ + int i, j; + + if (!cmds->cnt) + return; + + for (i = j = 1; i < cmds->cnt; i++) + if (strcmp(cmds->names[i]->name, cmds->names[i-1]->name)) + cmds->names[j++] = cmds->names[i]; + + cmds->cnt = j; +} + +void exclude_cmds(struct cmdnames *cmds, struct cmdnames *excludes) +{ + int ci, cj, ei; + int cmp; + + ci = cj = ei = 0; + while (ci < cmds->cnt && ei < excludes->cnt) { + cmp = strcmp(cmds->names[ci]->name, excludes->names[ei]->name); + if (cmp < 0) + cmds->names[cj++] = cmds->names[ci++]; + else if (cmp == 0) + ci++, ei++; + else if (cmp > 0) + ei++; + } + + while (ci < cmds->cnt) + cmds->names[cj++] = cmds->names[ci++]; + + cmds->cnt = cj; +} + +static void pretty_print_string_list(struct cmdnames *cmds, int longest) +{ + int cols = 1, rows; + int space = longest + 1; /* min 1 SP between words */ + int max_cols = term_columns() - 1; /* don't print *on* the edge */ + int i, j; + + if (space < max_cols) + cols = max_cols / space; + rows = (cmds->cnt + cols - 1) / cols; + + for (i = 0; i < rows; i++) { + printf(" "); + + for (j = 0; j < cols; j++) { + int n = j * rows + i; + int size = space; + if (n >= cmds->cnt) + break; + if (j == cols-1 || n + rows >= cmds->cnt) + size = 1; + printf("%-*s", size, cmds->names[n]->name); + } + putchar('\n'); + } +} + +static int is_executable(const char *name) +{ + struct stat st; + + if (stat(name, &st) || /* stat, not lstat */ + !S_ISREG(st.st_mode)) + return 0; + +#ifdef __MINGW32__ + /* cannot trust the executable bit, peek into the file instead */ + char buf[3] = { 0 }; + int n; + int fd = open(name, O_RDONLY); + st.st_mode &= ~S_IXUSR; + if (fd >= 0) { + n = read(fd, buf, 2); + if (n == 2) + /* DOS executables start with "MZ" */ + if (!strcmp(buf, "#!") || !strcmp(buf, "MZ")) + st.st_mode |= S_IXUSR; + close(fd); + } +#endif + return st.st_mode & S_IXUSR; +} + +static void list_commands_in_dir(struct cmdnames *cmds, + const char *path, + const char *prefix) +{ + int prefix_len; + DIR *dir = opendir(path); + struct dirent *de; + struct strbuf buf = STRBUF_INIT; + int len; + + if (!dir) + return; + if (!prefix) + prefix = "perf-"; + prefix_len = strlen(prefix); + + strbuf_addf(&buf, "%s/", path); + len = buf.len; + + while ((de = readdir(dir)) != NULL) { + int entlen; + + if (prefixcmp(de->d_name, prefix)) + continue; + + strbuf_setlen(&buf, len); + strbuf_addstr(&buf, de->d_name); + if (!is_executable(buf.buf)) + continue; + + entlen = strlen(de->d_name) - prefix_len; + if (has_extension(de->d_name, ".exe")) + entlen -= 4; + + add_cmdname(cmds, de->d_name + prefix_len, entlen); + } + closedir(dir); + strbuf_release(&buf); +} + +void load_command_list(const char *prefix, + struct cmdnames *main_cmds, + struct cmdnames *other_cmds) +{ + const char *env_path = getenv("PATH"); + const char *exec_path = perf_exec_path(); + + if (exec_path) { + list_commands_in_dir(main_cmds, exec_path, prefix); + qsort(main_cmds->names, main_cmds->cnt, + sizeof(*main_cmds->names), cmdname_compare); + uniq(main_cmds); + } + + if (env_path) { + char *paths, *path, *colon; + path = paths = strdup(env_path); + while (1) { + if ((colon = strchr(path, PATH_SEP))) + *colon = 0; + if (!exec_path || strcmp(path, exec_path)) + list_commands_in_dir(other_cmds, path, prefix); + + if (!colon) + break; + path = colon + 1; + } + free(paths); + + qsort(other_cmds->names, other_cmds->cnt, + sizeof(*other_cmds->names), cmdname_compare); + uniq(other_cmds); + } + exclude_cmds(other_cmds, main_cmds); +} + +void list_commands(const char *title, struct cmdnames *main_cmds, + struct cmdnames *other_cmds) +{ + int i, longest = 0; + + for (i = 0; i < main_cmds->cnt; i++) + if (longest < main_cmds->names[i]->len) + longest = main_cmds->names[i]->len; + for (i = 0; i < other_cmds->cnt; i++) + if (longest < other_cmds->names[i]->len) + longest = other_cmds->names[i]->len; + + if (main_cmds->cnt) { + const char *exec_path = perf_exec_path(); + printf("available %s in '%s'\n", title, exec_path); + printf("----------------"); + mput_char('-', strlen(title) + strlen(exec_path)); + putchar('\n'); + pretty_print_string_list(main_cmds, longest); + putchar('\n'); + } + + if (other_cmds->cnt) { + printf("%s available from elsewhere on your $PATH\n", title); + printf("---------------------------------------"); + mput_char('-', strlen(title)); + putchar('\n'); + pretty_print_string_list(other_cmds, longest); + putchar('\n'); + } +} + +int is_in_cmdlist(struct cmdnames *c, const char *s) +{ + int i; + for (i = 0; i < c->cnt; i++) + if (!strcmp(s, c->names[i]->name)) + return 1; + return 0; +} + +static int autocorrect; +static struct cmdnames aliases; + +static int perf_unknown_cmd_config(const char *var, const char *value, void *cb) +{ + if (!strcmp(var, "help.autocorrect")) + autocorrect = perf_config_int(var,value); + /* Also use aliases for command lookup */ + if (!prefixcmp(var, "alias.")) + add_cmdname(&aliases, var + 6, strlen(var + 6)); + + return perf_default_config(var, value, cb); +} + +static int levenshtein_compare(const void *p1, const void *p2) +{ + const struct cmdname *const *c1 = p1, *const *c2 = p2; + const char *s1 = (*c1)->name, *s2 = (*c2)->name; + int l1 = (*c1)->len; + int l2 = (*c2)->len; + return l1 != l2 ? l1 - l2 : strcmp(s1, s2); +} + +static void add_cmd_list(struct cmdnames *cmds, struct cmdnames *old) +{ + int i; + ALLOC_GROW(cmds->names, cmds->cnt + old->cnt, cmds->alloc); + + for (i = 0; i < old->cnt; i++) + cmds->names[cmds->cnt++] = old->names[i]; + free(old->names); + old->cnt = 0; + old->names = NULL; +} + +const char *help_unknown_cmd(const char *cmd) +{ + int i, n = 0, best_similarity = 0; + struct cmdnames main_cmds, other_cmds; + + memset(&main_cmds, 0, sizeof(main_cmds)); + memset(&other_cmds, 0, sizeof(main_cmds)); + memset(&aliases, 0, sizeof(aliases)); + + perf_config(perf_unknown_cmd_config, NULL); + + load_command_list("perf-", &main_cmds, &other_cmds); + + add_cmd_list(&main_cmds, &aliases); + add_cmd_list(&main_cmds, &other_cmds); + qsort(main_cmds.names, main_cmds.cnt, + sizeof(main_cmds.names), cmdname_compare); + uniq(&main_cmds); + + if (main_cmds.cnt) { + /* This reuses cmdname->len for similarity index */ + for (i = 0; i < main_cmds.cnt; ++i) + main_cmds.names[i]->len = + levenshtein(cmd, main_cmds.names[i]->name, 0, 2, 1, 4); + + qsort(main_cmds.names, main_cmds.cnt, + sizeof(*main_cmds.names), levenshtein_compare); + + best_similarity = main_cmds.names[0]->len; + n = 1; + while (n < main_cmds.cnt && best_similarity == main_cmds.names[n]->len) + ++n; + } + + if (autocorrect && n == 1) { + const char *assumed = main_cmds.names[0]->name; + + main_cmds.names[0] = NULL; + clean_cmdnames(&main_cmds); + fprintf(stderr, "WARNING: You called a Git program named '%s', " + "which does not exist.\n" + "Continuing under the assumption that you meant '%s'\n", + cmd, assumed); + if (autocorrect > 0) { + fprintf(stderr, "in %0.1f seconds automatically...\n", + (float)autocorrect/10.0); + poll(NULL, 0, autocorrect * 100); + } + return assumed; + } + + fprintf(stderr, "perf: '%s' is not a perf-command. See 'perf --help'.\n", cmd); + + if (main_cmds.cnt && best_similarity < 6) { + fprintf(stderr, "\nDid you mean %s?\n", + n < 2 ? "this": "one of these"); + + for (i = 0; i < n; i++) + fprintf(stderr, "\t%s\n", main_cmds.names[i]->name); + } + + exit(1); +} + +int cmd_version(int argc, const char **argv, const char *prefix) +{ + printf("perf version %s\n", perf_version_string); + return 0; +} diff --git a/tools/perf/util/help.h b/tools/perf/util/help.h new file mode 100644 index 00000000000..56bc15406ff --- /dev/null +++ b/tools/perf/util/help.h @@ -0,0 +1,29 @@ +#ifndef HELP_H +#define HELP_H + +struct cmdnames { + int alloc; + int cnt; + struct cmdname { + size_t len; /* also used for similarity index in help.c */ + char name[FLEX_ARRAY]; + } **names; +}; + +static inline void mput_char(char c, unsigned int num) +{ + while(num--) + putchar(c); +} + +void load_command_list(const char *prefix, + struct cmdnames *main_cmds, + struct cmdnames *other_cmds); +void add_cmdname(struct cmdnames *cmds, const char *name, int len); +/* Here we require that excludes is a sorted list. */ +void exclude_cmds(struct cmdnames *cmds, struct cmdnames *excludes); +int is_in_cmdlist(struct cmdnames *c, const char *s); +void list_commands(const char *title, struct cmdnames *main_cmds, + struct cmdnames *other_cmds); + +#endif /* HELP_H */ diff --git a/tools/perf/util/levenshtein.c b/tools/perf/util/levenshtein.c new file mode 100644 index 00000000000..e521d1516df --- /dev/null +++ b/tools/perf/util/levenshtein.c @@ -0,0 +1,84 @@ +#include "cache.h" +#include "levenshtein.h" + +/* + * This function implements the Damerau-Levenshtein algorithm to + * calculate a distance between strings. + * + * Basically, it says how many letters need to be swapped, substituted, + * deleted from, or added to string1, at least, to get string2. + * + * The idea is to build a distance matrix for the substrings of both + * strings. To avoid a large space complexity, only the last three rows + * are kept in memory (if swaps had the same or higher cost as one deletion + * plus one insertion, only two rows would be needed). + * + * At any stage, "i + 1" denotes the length of the current substring of + * string1 that the distance is calculated for. + * + * row2 holds the current row, row1 the previous row (i.e. for the substring + * of string1 of length "i"), and row0 the row before that. + * + * In other words, at the start of the big loop, row2[j + 1] contains the + * Damerau-Levenshtein distance between the substring of string1 of length + * "i" and the substring of string2 of length "j + 1". + * + * All the big loop does is determine the partial minimum-cost paths. + * + * It does so by calculating the costs of the path ending in characters + * i (in string1) and j (in string2), respectively, given that the last + * operation is a substition, a swap, a deletion, or an insertion. + * + * This implementation allows the costs to be weighted: + * + * - w (as in "sWap") + * - s (as in "Substitution") + * - a (for insertion, AKA "Add") + * - d (as in "Deletion") + * + * Note that this algorithm calculates a distance _iff_ d == a. + */ +int levenshtein(const char *string1, const char *string2, + int w, int s, int a, int d) +{ + int len1 = strlen(string1), len2 = strlen(string2); + int *row0 = malloc(sizeof(int) * (len2 + 1)); + int *row1 = malloc(sizeof(int) * (len2 + 1)); + int *row2 = malloc(sizeof(int) * (len2 + 1)); + int i, j; + + for (j = 0; j <= len2; j++) + row1[j] = j * a; + for (i = 0; i < len1; i++) { + int *dummy; + + row2[0] = (i + 1) * d; + for (j = 0; j < len2; j++) { + /* substitution */ + row2[j + 1] = row1[j] + s * (string1[i] != string2[j]); + /* swap */ + if (i > 0 && j > 0 && string1[i - 1] == string2[j] && + string1[i] == string2[j - 1] && + row2[j + 1] > row0[j - 1] + w) + row2[j + 1] = row0[j - 1] + w; + /* deletion */ + if (row2[j + 1] > row1[j + 1] + d) + row2[j + 1] = row1[j + 1] + d; + /* insertion */ + if (row2[j + 1] > row2[j] + a) + row2[j + 1] = row2[j] + a; + } + + dummy = row0; + row0 = row1; + row1 = row2; + row2 = dummy; + } + + i = row1[len2]; + free(row0); + free(row1); + free(row2); + + return i; +} diff --git a/tools/perf/util/levenshtein.h b/tools/perf/util/levenshtein.h new file mode 100644 index 00000000000..0173abeef52 --- /dev/null +++ b/tools/perf/util/levenshtein.h @@ -0,0 +1,8 @@ +#ifndef LEVENSHTEIN_H +#define LEVENSHTEIN_H + +int levenshtein(const char *string1, const char *string2, + int swap_penalty, int substition_penalty, + int insertion_penalty, int deletion_penalty); + +#endif diff --git a/tools/perf/util/list.h b/tools/perf/util/list.h new file mode 100644 index 00000000000..e2548e8072c --- /dev/null +++ b/tools/perf/util/list.h @@ -0,0 +1,603 @@ +#ifndef _LINUX_LIST_H +#define _LINUX_LIST_H +/* + Copyright (C) Cast of dozens, comes from the Linux kernel + + This program is free software; you can redistribute it and/or modify it + under the terms of version 2 of the GNU General Public License as + published by the Free Software Foundation. +*/ + +#include <stddef.h> + +/* + * These are non-NULL pointers that will result in page faults + * under normal circumstances, used to verify that nobody uses + * non-initialized list entries. + */ +#define LIST_POISON1 ((void *)0x00100100) +#define LIST_POISON2 ((void *)0x00200200) + +/** + * container_of - cast a member of a structure out to the containing structure + * @ptr: the pointer to the member. + * @type: the type of the container struct this is embedded in. + * @member: the name of the member within the struct. + * + */ +#define container_of(ptr, type, member) ({ \ + const typeof( ((type *)0)->member ) *__mptr = (ptr); \ + (type *)( (char *)__mptr - offsetof(type,member) );}) + +/* + * Simple doubly linked list implementation. + * + * Some of the internal functions ("__xxx") are useful when + * manipulating whole lists rather than single entries, as + * sometimes we already know the next/prev entries and we can + * generate better code by using them directly rather than + * using the generic single-entry routines. + */ + +struct list_head { + struct list_head *next, *prev; +}; + +#define LIST_HEAD_INIT(name) { &(name), &(name) } + +#define LIST_HEAD(name) \ + struct list_head name = LIST_HEAD_INIT(name) + +static inline void INIT_LIST_HEAD(struct list_head *list) +{ + list->next = list; + list->prev = list; +} + +/* + * Insert a new entry between two known consecutive entries. + * + * This is only for internal list manipulation where we know + * the prev/next entries already! + */ +static inline void __list_add(struct list_head *new, + struct list_head *prev, + struct list_head *next) +{ + next->prev = new; + new->next = next; + new->prev = prev; + prev->next = new; +} + +/** + * list_add - add a new entry + * @new: new entry to be added + * @head: list head to add it after + * + * Insert a new entry after the specified head. + * This is good for implementing stacks. + */ +static inline void list_add(struct list_head *new, struct list_head *head) +{ + __list_add(new, head, head->next); +} + +/** + * list_add_tail - add a new entry + * @new: new entry to be added + * @head: list head to add it before + * + * Insert a new entry before the specified head. + * This is useful for implementing queues. + */ +static inline void list_add_tail(struct list_head *new, struct list_head *head) +{ + __list_add(new, head->prev, head); +} + +/* + * Delete a list entry by making the prev/next entries + * point to each other. + * + * This is only for internal list manipulation where we know + * the prev/next entries already! + */ +static inline void __list_del(struct list_head * prev, struct list_head * next) +{ + next->prev = prev; + prev->next = next; +} + +/** + * list_del - deletes entry from list. + * @entry: the element to delete from the list. + * Note: list_empty on entry does not return true after this, the entry is + * in an undefined state. + */ +static inline void list_del(struct list_head *entry) +{ + __list_del(entry->prev, entry->next); + entry->next = LIST_POISON1; + entry->prev = LIST_POISON2; +} + +/** + * list_del_range - deletes range of entries from list. + * @beging: first element in the range to delete from the list. + * @beging: first element in the range to delete from the list. + * Note: list_empty on the range of entries does not return true after this, + * the entries is in an undefined state. + */ +static inline void list_del_range(struct list_head *begin, + struct list_head *end) +{ + begin->prev->next = end->next; + end->next->prev = begin->prev; +} + +/** + * list_replace - replace old entry by new one + * @old : the element to be replaced + * @new : the new element to insert + * Note: if 'old' was empty, it will be overwritten. + */ +static inline void list_replace(struct list_head *old, + struct list_head *new) +{ + new->next = old->next; + new->next->prev = new; + new->prev = old->prev; + new->prev->next = new; +} + +static inline void list_replace_init(struct list_head *old, + struct list_head *new) +{ + list_replace(old, new); + INIT_LIST_HEAD(old); +} + +/** + * list_del_init - deletes entry from list and reinitialize it. + * @entry: the element to delete from the list. + */ +static inline void list_del_init(struct list_head *entry) +{ + __list_del(entry->prev, entry->next); + INIT_LIST_HEAD(entry); +} + +/** + * list_move - delete from one list and add as another's head + * @list: the entry to move + * @head: the head that will precede our entry + */ +static inline void list_move(struct list_head *list, struct list_head *head) +{ + __list_del(list->prev, list->next); + list_add(list, head); +} + +/** + * list_move_tail - delete from one list and add as another's tail + * @list: the entry to move + * @head: the head that will follow our entry + */ +static inline void list_move_tail(struct list_head *list, + struct list_head *head) +{ + __list_del(list->prev, list->next); + list_add_tail(list, head); +} + +/** + * list_is_last - tests whether @list is the last entry in list @head + * @list: the entry to test + * @head: the head of the list + */ +static inline int list_is_last(const struct list_head *list, + const struct list_head *head) +{ + return list->next == head; +} + +/** + * list_empty - tests whether a list is empty + * @head: the list to test. + */ +static inline int list_empty(const struct list_head *head) +{ + return head->next == head; +} + +/** + * list_empty_careful - tests whether a list is empty and not being modified + * @head: the list to test + * + * Description: + * tests whether a list is empty _and_ checks that no other CPU might be + * in the process of modifying either member (next or prev) + * + * NOTE: using list_empty_careful() without synchronization + * can only be safe if the only activity that can happen + * to the list entry is list_del_init(). Eg. it cannot be used + * if another CPU could re-list_add() it. + */ +static inline int list_empty_careful(const struct list_head *head) +{ + struct list_head *next = head->next; + return (next == head) && (next == head->prev); +} + +static inline void __list_splice(struct list_head *list, + struct list_head *head) +{ + struct list_head *first = list->next; + struct list_head *last = list->prev; + struct list_head *at = head->next; + + first->prev = head; + head->next = first; + + last->next = at; + at->prev = last; +} + +/** + * list_splice - join two lists + * @list: the new list to add. + * @head: the place to add it in the first list. + */ +static inline void list_splice(struct list_head *list, struct list_head *head) +{ + if (!list_empty(list)) + __list_splice(list, head); +} + +/** + * list_splice_init - join two lists and reinitialise the emptied list. + * @list: the new list to add. + * @head: the place to add it in the first list. + * + * The list at @list is reinitialised + */ +static inline void list_splice_init(struct list_head *list, + struct list_head *head) +{ + if (!list_empty(list)) { + __list_splice(list, head); + INIT_LIST_HEAD(list); + } +} + +/** + * list_entry - get the struct for this entry + * @ptr: the &struct list_head pointer. + * @type: the type of the struct this is embedded in. + * @member: the name of the list_struct within the struct. + */ +#define list_entry(ptr, type, member) \ + container_of(ptr, type, member) + +/** + * list_first_entry - get the first element from a list + * @ptr: the list head to take the element from. + * @type: the type of the struct this is embedded in. + * @member: the name of the list_struct within the struct. + * + * Note, that list is expected to be not empty. + */ +#define list_first_entry(ptr, type, member) \ + list_entry((ptr)->next, type, member) + +/** + * list_for_each - iterate over a list + * @pos: the &struct list_head to use as a loop cursor. + * @head: the head for your list. + */ +#define list_for_each(pos, head) \ + for (pos = (head)->next; pos != (head); \ + pos = pos->next) + +/** + * __list_for_each - iterate over a list + * @pos: the &struct list_head to use as a loop cursor. + * @head: the head for your list. + * + * This variant differs from list_for_each() in that it's the + * simplest possible list iteration code, no prefetching is done. + * Use this for code that knows the list to be very short (empty + * or 1 entry) most of the time. + */ +#define __list_for_each(pos, head) \ + for (pos = (head)->next; pos != (head); pos = pos->next) + +/** + * list_for_each_prev - iterate over a list backwards + * @pos: the &struct list_head to use as a loop cursor. + * @head: the head for your list. + */ +#define list_for_each_prev(pos, head) \ + for (pos = (head)->prev; pos != (head); \ + pos = pos->prev) + +/** + * list_for_each_safe - iterate over a list safe against removal of list entry + * @pos: the &struct list_head to use as a loop cursor. + * @n: another &struct list_head to use as temporary storage + * @head: the head for your list. + */ +#define list_for_each_safe(pos, n, head) \ + for (pos = (head)->next, n = pos->next; pos != (head); \ + pos = n, n = pos->next) + +/** + * list_for_each_entry - iterate over list of given type + * @pos: the type * to use as a loop cursor. + * @head: the head for your list. + * @member: the name of the list_struct within the struct. + */ +#define list_for_each_entry(pos, head, member) \ + for (pos = list_entry((head)->next, typeof(*pos), member); \ + &pos->member != (head); \ + pos = list_entry(pos->member.next, typeof(*pos), member)) + +/** + * list_for_each_entry_reverse - iterate backwards over list of given type. + * @pos: the type * to use as a loop cursor. + * @head: the head for your list. + * @member: the name of the list_struct within the struct. + */ +#define list_for_each_entry_reverse(pos, head, member) \ + for (pos = list_entry((head)->prev, typeof(*pos), member); \ + &pos->member != (head); \ + pos = list_entry(pos->member.prev, typeof(*pos), member)) + +/** + * list_prepare_entry - prepare a pos entry for use in list_for_each_entry_continue + * @pos: the type * to use as a start point + * @head: the head of the list + * @member: the name of the list_struct within the struct. + * + * Prepares a pos entry for use as a start point in list_for_each_entry_continue. + */ +#define list_prepare_entry(pos, head, member) \ + ((pos) ? : list_entry(head, typeof(*pos), member)) + +/** + * list_for_each_entry_continue - continue iteration over list of given type + * @pos: the type * to use as a loop cursor. + * @head: the head for your list. + * @member: the name of the list_struct within the struct. + * + * Continue to iterate over list of given type, continuing after + * the current position. + */ +#define list_for_each_entry_continue(pos, head, member) \ + for (pos = list_entry(pos->member.next, typeof(*pos), member); \ + &pos->member != (head); \ + pos = list_entry(pos->member.next, typeof(*pos), member)) + +/** + * list_for_each_entry_from - iterate over list of given type from the current point + * @pos: the type * to use as a loop cursor. + * @head: the head for your list. + * @member: the name of the list_struct within the struct. + * + * Iterate over list of given type, continuing from current position. + */ +#define list_for_each_entry_from(pos, head, member) \ + for (; &pos->member != (head); \ + pos = list_entry(pos->member.next, typeof(*pos), member)) + +/** + * list_for_each_entry_safe - iterate over list of given type safe against removal of list entry + * @pos: the type * to use as a loop cursor. + * @n: another type * to use as temporary storage + * @head: the head for your list. + * @member: the name of the list_struct within the struct. + */ +#define list_for_each_entry_safe(pos, n, head, member) \ + for (pos = list_entry((head)->next, typeof(*pos), member), \ + n = list_entry(pos->member.next, typeof(*pos), member); \ + &pos->member != (head); \ + pos = n, n = list_entry(n->member.next, typeof(*n), member)) + +/** + * list_for_each_entry_safe_continue + * @pos: the type * to use as a loop cursor. + * @n: another type * to use as temporary storage + * @head: the head for your list. + * @member: the name of the list_struct within the struct. + * + * Iterate over list of given type, continuing after current point, + * safe against removal of list entry. + */ +#define list_for_each_entry_safe_continue(pos, n, head, member) \ + for (pos = list_entry(pos->member.next, typeof(*pos), member), \ + n = list_entry(pos->member.next, typeof(*pos), member); \ + &pos->member != (head); \ + pos = n, n = list_entry(n->member.next, typeof(*n), member)) + +/** + * list_for_each_entry_safe_from + * @pos: the type * to use as a loop cursor. + * @n: another type * to use as temporary storage + * @head: the head for your list. + * @member: the name of the list_struct within the struct. + * + * Iterate over list of given type from current point, safe against + * removal of list entry. + */ +#define list_for_each_entry_safe_from(pos, n, head, member) \ + for (n = list_entry(pos->member.next, typeof(*pos), member); \ + &pos->member != (head); \ + pos = n, n = list_entry(n->member.next, typeof(*n), member)) + +/** + * list_for_each_entry_safe_reverse + * @pos: the type * to use as a loop cursor. + * @n: another type * to use as temporary storage + * @head: the head for your list. + * @member: the name of the list_struct within the struct. + * + * Iterate backwards over list of given type, safe against removal + * of list entry. + */ +#define list_for_each_entry_safe_reverse(pos, n, head, member) \ + for (pos = list_entry((head)->prev, typeof(*pos), member), \ + n = list_entry(pos->member.prev, typeof(*pos), member); \ + &pos->member != (head); \ + pos = n, n = list_entry(n->member.prev, typeof(*n), member)) + +/* + * Double linked lists with a single pointer list head. + * Mostly useful for hash tables where the two pointer list head is + * too wasteful. + * You lose the ability to access the tail in O(1). + */ + +struct hlist_head { + struct hlist_node *first; +}; + +struct hlist_node { + struct hlist_node *next, **pprev; +}; + +#define HLIST_HEAD_INIT { .first = NULL } +#define HLIST_HEAD(name) struct hlist_head name = { .first = NULL } +#define INIT_HLIST_HEAD(ptr) ((ptr)->first = NULL) +static inline void INIT_HLIST_NODE(struct hlist_node *h) +{ + h->next = NULL; + h->pprev = NULL; +} + +static inline int hlist_unhashed(const struct hlist_node *h) +{ + return !h->pprev; +} + +static inline int hlist_empty(const struct hlist_head *h) +{ + return !h->first; +} + +static inline void __hlist_del(struct hlist_node *n) +{ + struct hlist_node *next = n->next; + struct hlist_node **pprev = n->pprev; + *pprev = next; + if (next) + next->pprev = pprev; +} + +static inline void hlist_del(struct hlist_node *n) +{ + __hlist_del(n); + n->next = LIST_POISON1; + n->pprev = LIST_POISON2; +} + +static inline void hlist_del_init(struct hlist_node *n) +{ + if (!hlist_unhashed(n)) { + __hlist_del(n); + INIT_HLIST_NODE(n); + } +} + +static inline void hlist_add_head(struct hlist_node *n, struct hlist_head *h) +{ + struct hlist_node *first = h->first; + n->next = first; + if (first) + first->pprev = &n->next; + h->first = n; + n->pprev = &h->first; +} + +/* next must be != NULL */ +static inline void hlist_add_before(struct hlist_node *n, + struct hlist_node *next) +{ + n->pprev = next->pprev; + n->next = next; + next->pprev = &n->next; + *(n->pprev) = n; +} + +static inline void hlist_add_after(struct hlist_node *n, + struct hlist_node *next) +{ + next->next = n->next; + n->next = next; + next->pprev = &n->next; + + if(next->next) + next->next->pprev = &next->next; +} + +#define hlist_entry(ptr, type, member) container_of(ptr,type,member) + +#define hlist_for_each(pos, head) \ + for (pos = (head)->first; pos; \ + pos = pos->next) + +#define hlist_for_each_safe(pos, n, head) \ + for (pos = (head)->first; pos && ({ n = pos->next; 1; }); \ + pos = n) + +/** + * hlist_for_each_entry - iterate over list of given type + * @tpos: the type * to use as a loop cursor. + * @pos: the &struct hlist_node to use as a loop cursor. + * @head: the head for your list. + * @member: the name of the hlist_node within the struct. + */ +#define hlist_for_each_entry(tpos, pos, head, member) \ + for (pos = (head)->first; \ + pos && \ + ({ tpos = hlist_entry(pos, typeof(*tpos), member); 1;}); \ + pos = pos->next) + +/** + * hlist_for_each_entry_continue - iterate over a hlist continuing after current point + * @tpos: the type * to use as a loop cursor. + * @pos: the &struct hlist_node to use as a loop cursor. + * @member: the name of the hlist_node within the struct. + */ +#define hlist_for_each_entry_continue(tpos, pos, member) \ + for (pos = (pos)->next; \ + pos && \ + ({ tpos = hlist_entry(pos, typeof(*tpos), member); 1;}); \ + pos = pos->next) + +/** + * hlist_for_each_entry_from - iterate over a hlist continuing from current point + * @tpos: the type * to use as a loop cursor. + * @pos: the &struct hlist_node to use as a loop cursor. + * @member: the name of the hlist_node within the struct. + */ +#define hlist_for_each_entry_from(tpos, pos, member) \ + for (; pos && \ + ({ tpos = hlist_entry(pos, typeof(*tpos), member); 1;}); \ + pos = pos->next) + +/** + * hlist_for_each_entry_safe - iterate over list of given type safe against removal of list entry + * @tpos: the type * to use as a loop cursor. + * @pos: the &struct hlist_node to use as a loop cursor. + * @n: another &struct hlist_node to use as temporary storage + * @head: the head for your list. + * @member: the name of the hlist_node within the struct. + */ +#define hlist_for_each_entry_safe(tpos, pos, n, head, member) \ + for (pos = (head)->first; \ + pos && ({ n = pos->next; 1; }) && \ + ({ tpos = hlist_entry(pos, typeof(*tpos), member); 1;}); \ + pos = n) + +#endif diff --git a/tools/perf/util/pager.c b/tools/perf/util/pager.c new file mode 100644 index 00000000000..a28bccae545 --- /dev/null +++ b/tools/perf/util/pager.c @@ -0,0 +1,99 @@ +#include "cache.h" +#include "run-command.h" +#include "sigchain.h" + +/* + * This is split up from the rest of git so that we can do + * something different on Windows. + */ + +static int spawned_pager; + +#ifndef __MINGW32__ +static void pager_preexec(void) +{ + /* + * Work around bug in "less" by not starting it until we + * have real input + */ + fd_set in; + + FD_ZERO(&in); + FD_SET(0, &in); + select(1, &in, NULL, &in, NULL); + + setenv("LESS", "FRSX", 0); +} +#endif + +static const char *pager_argv[] = { "sh", "-c", NULL, NULL }; +static struct child_process pager_process; + +static void wait_for_pager(void) +{ + fflush(stdout); + fflush(stderr); + /* signal EOF to pager */ + close(1); + close(2); + finish_command(&pager_process); +} + +static void wait_for_pager_signal(int signo) +{ + wait_for_pager(); + sigchain_pop(signo); + raise(signo); +} + +void setup_pager(void) +{ + const char *pager = getenv("PERF_PAGER"); + + if (!isatty(1)) + return; + if (!pager) { + if (!pager_program) + perf_config(perf_default_config, NULL); + pager = pager_program; + } + if (!pager) + pager = getenv("PAGER"); + if (!pager) + pager = "less"; + else if (!*pager || !strcmp(pager, "cat")) + return; + + spawned_pager = 1; /* means we are emitting to terminal */ + + /* spawn the pager */ + pager_argv[2] = pager; + pager_process.argv = pager_argv; + pager_process.in = -1; +#ifndef __MINGW32__ + pager_process.preexec_cb = pager_preexec; +#endif + if (start_command(&pager_process)) + return; + + /* original process continues, but writes to the pipe */ + dup2(pager_process.in, 1); + if (isatty(2)) + dup2(pager_process.in, 2); + close(pager_process.in); + + /* this makes sure that the parent terminates after the pager */ + sigchain_push_common(wait_for_pager_signal); + atexit(wait_for_pager); +} + +int pager_in_use(void) +{ + const char *env; + + if (spawned_pager) + return 1; + + env = getenv("PERF_PAGER_IN_USE"); + return env ? perf_config_bool("PERF_PAGER_IN_USE", env) : 0; +} diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c new file mode 100644 index 00000000000..9d5f1ca50e6 --- /dev/null +++ b/tools/perf/util/parse-events.c @@ -0,0 +1,316 @@ + +#include "../perf.h" +#include "util.h" +#include "parse-options.h" +#include "parse-events.h" +#include "exec_cmd.h" +#include "string.h" + +extern char *strcasestr(const char *haystack, const char *needle); + +int nr_counters; + +struct perf_counter_attr attrs[MAX_COUNTERS]; + +struct event_symbol { + __u8 type; + __u64 config; + char *symbol; +}; + +#define C(x, y) .type = PERF_TYPE_##x, .config = PERF_COUNT_##y +#define CR(x, y) .type = PERF_TYPE_##x, .config = y + +static struct event_symbol event_symbols[] = { + { C(HARDWARE, HW_CPU_CYCLES), "cpu-cycles", }, + { C(HARDWARE, HW_CPU_CYCLES), "cycles", }, + { C(HARDWARE, HW_INSTRUCTIONS), "instructions", }, + { C(HARDWARE, HW_CACHE_REFERENCES), "cache-references", }, + { C(HARDWARE, HW_CACHE_MISSES), "cache-misses", }, + { C(HARDWARE, HW_BRANCH_INSTRUCTIONS),"branch-instructions", }, + { C(HARDWARE, HW_BRANCH_INSTRUCTIONS),"branches", }, + { C(HARDWARE, HW_BRANCH_MISSES), "branch-misses", }, + { C(HARDWARE, HW_BUS_CYCLES), "bus-cycles", }, + + { C(SOFTWARE, SW_CPU_CLOCK), "cpu-clock", }, + { C(SOFTWARE, SW_TASK_CLOCK), "task-clock", }, + { C(SOFTWARE, SW_PAGE_FAULTS), "page-faults", }, + { C(SOFTWARE, SW_PAGE_FAULTS), "faults", }, + { C(SOFTWARE, SW_PAGE_FAULTS_MIN), "minor-faults", }, + { C(SOFTWARE, SW_PAGE_FAULTS_MAJ), "major-faults", }, + { C(SOFTWARE, SW_CONTEXT_SWITCHES), "context-switches", }, + { C(SOFTWARE, SW_CONTEXT_SWITCHES), "cs", }, + { C(SOFTWARE, SW_CPU_MIGRATIONS), "cpu-migrations", }, + { C(SOFTWARE, SW_CPU_MIGRATIONS), "migrations", }, +}; + +#define __PERF_COUNTER_FIELD(config, name) \ + ((config & PERF_COUNTER_##name##_MASK) >> PERF_COUNTER_##name##_SHIFT) + +#define PERF_COUNTER_RAW(config) __PERF_COUNTER_FIELD(config, RAW) +#define PERF_COUNTER_CONFIG(config) __PERF_COUNTER_FIELD(config, CONFIG) +#define PERF_COUNTER_TYPE(config) __PERF_COUNTER_FIELD(config, TYPE) +#define PERF_COUNTER_ID(config) __PERF_COUNTER_FIELD(config, EVENT) + +static char *hw_event_names[] = { + "cycles", + "instructions", + "cache-references", + "cache-misses", + "branches", + "branch-misses", + "bus-cycles", +}; + +static char *sw_event_names[] = { + "cpu-clock-ticks", + "task-clock-ticks", + "page-faults", + "context-switches", + "CPU-migrations", + "minor-faults", + "major-faults", +}; + +#define MAX_ALIASES 8 + +static char *hw_cache [][MAX_ALIASES] = { + { "L1-data" , "l1-d", "l1d", "l1" }, + { "L1-instruction" , "l1-i", "l1i" }, + { "L2" , "l2" }, + { "Data-TLB" , "dtlb", "d-tlb" }, + { "Instruction-TLB" , "itlb", "i-tlb" }, + { "Branch" , "bpu" , "btb", "bpc" }, +}; + +static char *hw_cache_op [][MAX_ALIASES] = { + { "Load" , "read" }, + { "Store" , "write" }, + { "Prefetch" , "speculative-read", "speculative-load" }, +}; + +static char *hw_cache_result [][MAX_ALIASES] = { + { "Reference" , "ops", "access" }, + { "Miss" }, +}; + +char *event_name(int counter) +{ + __u64 config = attrs[counter].config; + int type = attrs[counter].type; + static char buf[32]; + + if (attrs[counter].type == PERF_TYPE_RAW) { + sprintf(buf, "raw 0x%llx", config); + return buf; + } + + switch (type) { + case PERF_TYPE_HARDWARE: + if (config < PERF_COUNT_HW_MAX) + return hw_event_names[config]; + return "unknown-hardware"; + + case PERF_TYPE_HW_CACHE: { + __u8 cache_type, cache_op, cache_result; + static char name[100]; + + cache_type = (config >> 0) & 0xff; + if (cache_type > PERF_COUNT_HW_CACHE_MAX) + return "unknown-ext-hardware-cache-type"; + + cache_op = (config >> 8) & 0xff; + if (cache_op > PERF_COUNT_HW_CACHE_OP_MAX) + return "unknown-ext-hardware-cache-op"; + + cache_result = (config >> 16) & 0xff; + if (cache_result > PERF_COUNT_HW_CACHE_RESULT_MAX) + return "unknown-ext-hardware-cache-result"; + + sprintf(name, "%s-Cache-%s-%ses", + hw_cache[cache_type][0], + hw_cache_op[cache_op][0], + hw_cache_result[cache_result][0]); + + return name; + } + + case PERF_TYPE_SOFTWARE: + if (config < PERF_COUNT_SW_MAX) + return sw_event_names[config]; + return "unknown-software"; + + default: + break; + } + + return "unknown"; +} + +static int parse_aliases(const char *str, char *names[][MAX_ALIASES], int size) +{ + int i, j; + + for (i = 0; i < size; i++) { + for (j = 0; j < MAX_ALIASES; j++) { + if (!names[i][j]) + break; + if (strcasestr(str, names[i][j])) + return i; + } + } + + return -1; +} + +static int parse_generic_hw_symbols(const char *str, struct perf_counter_attr *attr) +{ + int cache_type = -1, cache_op = 0, cache_result = 0; + + cache_type = parse_aliases(str, hw_cache, PERF_COUNT_HW_CACHE_MAX); + /* + * No fallback - if we cannot get a clear cache type + * then bail out: + */ + if (cache_type == -1) + return -EINVAL; + + cache_op = parse_aliases(str, hw_cache_op, PERF_COUNT_HW_CACHE_OP_MAX); + /* + * Fall back to reads: + */ + if (cache_op == -1) + cache_op = PERF_COUNT_HW_CACHE_OP_READ; + + cache_result = parse_aliases(str, hw_cache_result, + PERF_COUNT_HW_CACHE_RESULT_MAX); + /* + * Fall back to accesses: + */ + if (cache_result == -1) + cache_result = PERF_COUNT_HW_CACHE_RESULT_ACCESS; + + attr->config = cache_type | (cache_op << 8) | (cache_result << 16); + attr->type = PERF_TYPE_HW_CACHE; + + return 0; +} + +/* + * Each event can have multiple symbolic names. + * Symbolic names are (almost) exactly matched. + */ +static int parse_event_symbols(const char *str, struct perf_counter_attr *attr) +{ + __u64 config, id; + int type; + unsigned int i; + const char *sep, *pstr; + + if (str[0] == 'r' && hex2u64(str + 1, &config) > 0) { + attr->type = PERF_TYPE_RAW; + attr->config = config; + + return 0; + } + + pstr = str; + sep = strchr(pstr, ':'); + if (sep) { + type = atoi(pstr); + pstr = sep + 1; + id = atoi(pstr); + sep = strchr(pstr, ':'); + if (sep) { + pstr = sep + 1; + if (strchr(pstr, 'k')) + attr->exclude_user = 1; + if (strchr(pstr, 'u')) + attr->exclude_kernel = 1; + } + attr->type = type; + attr->config = id; + + return 0; + } + + for (i = 0; i < ARRAY_SIZE(event_symbols); i++) { + if (!strncmp(str, event_symbols[i].symbol, + strlen(event_symbols[i].symbol))) { + + attr->type = event_symbols[i].type; + attr->config = event_symbols[i].config; + + return 0; + } + } + + return parse_generic_hw_symbols(str, attr); +} + +int parse_events(const struct option *opt, const char *str, int unset) +{ + struct perf_counter_attr attr; + int ret; + + memset(&attr, 0, sizeof(attr)); +again: + if (nr_counters == MAX_COUNTERS) + return -1; + + ret = parse_event_symbols(str, &attr); + if (ret < 0) + return ret; + + attrs[nr_counters] = attr; + nr_counters++; + + str = strstr(str, ","); + if (str) { + str++; + goto again; + } + + return 0; +} + +static const char * const event_type_descriptors[] = { + "", + "Hardware event", + "Software event", + "Tracepoint event", + "Hardware cache event", +}; + +/* + * Print the help text for the event symbols: + */ +void print_events(void) +{ + struct event_symbol *syms = event_symbols; + unsigned int i, type, prev_type = -1; + + fprintf(stderr, "\n"); + fprintf(stderr, "List of pre-defined events (to be used in -e):\n"); + + for (i = 0; i < ARRAY_SIZE(event_symbols); i++, syms++) { + type = syms->type + 1; + if (type > ARRAY_SIZE(event_type_descriptors)) + type = 0; + + if (type != prev_type) + fprintf(stderr, "\n"); + + fprintf(stderr, " %-30s [%s]\n", syms->symbol, + event_type_descriptors[type]); + + prev_type = type; + } + + fprintf(stderr, "\n"); + fprintf(stderr, " %-30s [raw hardware event descriptor]\n", + "rNNN"); + fprintf(stderr, "\n"); + + exit(129); +} diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h new file mode 100644 index 00000000000..e3d552908e6 --- /dev/null +++ b/tools/perf/util/parse-events.h @@ -0,0 +1,17 @@ + +/* + * Parse symbolic events/counts passed in as options: + */ + +extern int nr_counters; + +extern struct perf_counter_attr attrs[MAX_COUNTERS]; + +extern char *event_name(int ctr); + +extern int parse_events(const struct option *opt, const char *str, int unset); + +#define EVENTS_HELP_MAX (128*1024) + +extern void print_events(void); + diff --git a/tools/perf/util/parse-options.c b/tools/perf/util/parse-options.c new file mode 100644 index 00000000000..b3affb1658d --- /dev/null +++ b/tools/perf/util/parse-options.c @@ -0,0 +1,508 @@ +#include "util.h" +#include "parse-options.h" +#include "cache.h" + +#define OPT_SHORT 1 +#define OPT_UNSET 2 + +static int opterror(const struct option *opt, const char *reason, int flags) +{ + if (flags & OPT_SHORT) + return error("switch `%c' %s", opt->short_name, reason); + if (flags & OPT_UNSET) + return error("option `no-%s' %s", opt->long_name, reason); + return error("option `%s' %s", opt->long_name, reason); +} + +static int get_arg(struct parse_opt_ctx_t *p, const struct option *opt, + int flags, const char **arg) +{ + if (p->opt) { + *arg = p->opt; + p->opt = NULL; + } else if (p->argc == 1 && (opt->flags & PARSE_OPT_LASTARG_DEFAULT)) { + *arg = (const char *)opt->defval; + } else if (p->argc > 1) { + p->argc--; + *arg = *++p->argv; + } else + return opterror(opt, "requires a value", flags); + return 0; +} + +static int get_value(struct parse_opt_ctx_t *p, + const struct option *opt, int flags) +{ + const char *s, *arg = NULL; + const int unset = flags & OPT_UNSET; + + if (unset && p->opt) + return opterror(opt, "takes no value", flags); + if (unset && (opt->flags & PARSE_OPT_NONEG)) + return opterror(opt, "isn't available", flags); + + if (!(flags & OPT_SHORT) && p->opt) { + switch (opt->type) { + case OPTION_CALLBACK: + if (!(opt->flags & PARSE_OPT_NOARG)) + break; + /* FALLTHROUGH */ + case OPTION_BOOLEAN: + case OPTION_BIT: + case OPTION_SET_INT: + case OPTION_SET_PTR: + return opterror(opt, "takes no value", flags); + default: + break; + } + } + + switch (opt->type) { + case OPTION_BIT: + if (unset) + *(int *)opt->value &= ~opt->defval; + else + *(int *)opt->value |= opt->defval; + return 0; + + case OPTION_BOOLEAN: + *(int *)opt->value = unset ? 0 : *(int *)opt->value + 1; + return 0; + + case OPTION_SET_INT: + *(int *)opt->value = unset ? 0 : opt->defval; + return 0; + + case OPTION_SET_PTR: + *(void **)opt->value = unset ? NULL : (void *)opt->defval; + return 0; + + case OPTION_STRING: + if (unset) + *(const char **)opt->value = NULL; + else if (opt->flags & PARSE_OPT_OPTARG && !p->opt) + *(const char **)opt->value = (const char *)opt->defval; + else + return get_arg(p, opt, flags, (const char **)opt->value); + return 0; + + case OPTION_CALLBACK: + if (unset) + return (*opt->callback)(opt, NULL, 1) ? (-1) : 0; + if (opt->flags & PARSE_OPT_NOARG) + return (*opt->callback)(opt, NULL, 0) ? (-1) : 0; + if (opt->flags & PARSE_OPT_OPTARG && !p->opt) + return (*opt->callback)(opt, NULL, 0) ? (-1) : 0; + if (get_arg(p, opt, flags, &arg)) + return -1; + return (*opt->callback)(opt, arg, 0) ? (-1) : 0; + + case OPTION_INTEGER: + if (unset) { + *(int *)opt->value = 0; + return 0; + } + if (opt->flags & PARSE_OPT_OPTARG && !p->opt) { + *(int *)opt->value = opt->defval; + return 0; + } + if (get_arg(p, opt, flags, &arg)) + return -1; + *(int *)opt->value = strtol(arg, (char **)&s, 10); + if (*s) + return opterror(opt, "expects a numerical value", flags); + return 0; + + case OPTION_LONG: + if (unset) { + *(long *)opt->value = 0; + return 0; + } + if (opt->flags & PARSE_OPT_OPTARG && !p->opt) { + *(long *)opt->value = opt->defval; + return 0; + } + if (get_arg(p, opt, flags, &arg)) + return -1; + *(long *)opt->value = strtol(arg, (char **)&s, 10); + if (*s) + return opterror(opt, "expects a numerical value", flags); + return 0; + + default: + die("should not happen, someone must be hit on the forehead"); + } +} + +static int parse_short_opt(struct parse_opt_ctx_t *p, const struct option *options) +{ + for (; options->type != OPTION_END; options++) { + if (options->short_name == *p->opt) { + p->opt = p->opt[1] ? p->opt + 1 : NULL; + return get_value(p, options, OPT_SHORT); + } + } + return -2; +} + +static int parse_long_opt(struct parse_opt_ctx_t *p, const char *arg, + const struct option *options) +{ + const char *arg_end = strchr(arg, '='); + const struct option *abbrev_option = NULL, *ambiguous_option = NULL; + int abbrev_flags = 0, ambiguous_flags = 0; + + if (!arg_end) + arg_end = arg + strlen(arg); + + for (; options->type != OPTION_END; options++) { + const char *rest; + int flags = 0; + + if (!options->long_name) + continue; + + rest = skip_prefix(arg, options->long_name); + if (options->type == OPTION_ARGUMENT) { + if (!rest) + continue; + if (*rest == '=') + return opterror(options, "takes no value", flags); + if (*rest) + continue; + p->out[p->cpidx++] = arg - 2; + return 0; + } + if (!rest) { + /* abbreviated? */ + if (!strncmp(options->long_name, arg, arg_end - arg)) { +is_abbreviated: + if (abbrev_option) { + /* + * If this is abbreviated, it is + * ambiguous. So when there is no + * exact match later, we need to + * error out. + */ + ambiguous_option = abbrev_option; + ambiguous_flags = abbrev_flags; + } + if (!(flags & OPT_UNSET) && *arg_end) + p->opt = arg_end + 1; + abbrev_option = options; + abbrev_flags = flags; + continue; + } + /* negated and abbreviated very much? */ + if (!prefixcmp("no-", arg)) { + flags |= OPT_UNSET; + goto is_abbreviated; + } + /* negated? */ + if (strncmp(arg, "no-", 3)) + continue; + flags |= OPT_UNSET; + rest = skip_prefix(arg + 3, options->long_name); + /* abbreviated and negated? */ + if (!rest && !prefixcmp(options->long_name, arg + 3)) + goto is_abbreviated; + if (!rest) + continue; + } + if (*rest) { + if (*rest != '=') + continue; + p->opt = rest + 1; + } + return get_value(p, options, flags); + } + + if (ambiguous_option) + return error("Ambiguous option: %s " + "(could be --%s%s or --%s%s)", + arg, + (ambiguous_flags & OPT_UNSET) ? "no-" : "", + ambiguous_option->long_name, + (abbrev_flags & OPT_UNSET) ? "no-" : "", + abbrev_option->long_name); + if (abbrev_option) + return get_value(p, abbrev_option, abbrev_flags); + return -2; +} + +static void check_typos(const char *arg, const struct option *options) +{ + if (strlen(arg) < 3) + return; + + if (!prefixcmp(arg, "no-")) { + error ("did you mean `--%s` (with two dashes ?)", arg); + exit(129); + } + + for (; options->type != OPTION_END; options++) { + if (!options->long_name) + continue; + if (!prefixcmp(options->long_name, arg)) { + error ("did you mean `--%s` (with two dashes ?)", arg); + exit(129); + } + } +} + +void parse_options_start(struct parse_opt_ctx_t *ctx, + int argc, const char **argv, int flags) +{ + memset(ctx, 0, sizeof(*ctx)); + ctx->argc = argc - 1; + ctx->argv = argv + 1; + ctx->out = argv; + ctx->cpidx = ((flags & PARSE_OPT_KEEP_ARGV0) != 0); + ctx->flags = flags; + if ((flags & PARSE_OPT_KEEP_UNKNOWN) && + (flags & PARSE_OPT_STOP_AT_NON_OPTION)) + die("STOP_AT_NON_OPTION and KEEP_UNKNOWN don't go together"); +} + +static int usage_with_options_internal(const char * const *, + const struct option *, int); + +int parse_options_step(struct parse_opt_ctx_t *ctx, + const struct option *options, + const char * const usagestr[]) +{ + int internal_help = !(ctx->flags & PARSE_OPT_NO_INTERNAL_HELP); + + /* we must reset ->opt, unknown short option leave it dangling */ + ctx->opt = NULL; + + for (; ctx->argc; ctx->argc--, ctx->argv++) { + const char *arg = ctx->argv[0]; + + if (*arg != '-' || !arg[1]) { + if (ctx->flags & PARSE_OPT_STOP_AT_NON_OPTION) + break; + ctx->out[ctx->cpidx++] = ctx->argv[0]; + continue; + } + + if (arg[1] != '-') { + ctx->opt = arg + 1; + if (internal_help && *ctx->opt == 'h') + return parse_options_usage(usagestr, options); + switch (parse_short_opt(ctx, options)) { + case -1: + return parse_options_usage(usagestr, options); + case -2: + goto unknown; + } + if (ctx->opt) + check_typos(arg + 1, options); + while (ctx->opt) { + if (internal_help && *ctx->opt == 'h') + return parse_options_usage(usagestr, options); + switch (parse_short_opt(ctx, options)) { + case -1: + return parse_options_usage(usagestr, options); + case -2: + /* fake a short option thing to hide the fact that we may have + * started to parse aggregated stuff + * + * This is leaky, too bad. + */ + ctx->argv[0] = strdup(ctx->opt - 1); + *(char *)ctx->argv[0] = '-'; + goto unknown; + } + } + continue; + } + + if (!arg[2]) { /* "--" */ + if (!(ctx->flags & PARSE_OPT_KEEP_DASHDASH)) { + ctx->argc--; + ctx->argv++; + } + break; + } + + if (internal_help && !strcmp(arg + 2, "help-all")) + return usage_with_options_internal(usagestr, options, 1); + if (internal_help && !strcmp(arg + 2, "help")) + return parse_options_usage(usagestr, options); + switch (parse_long_opt(ctx, arg + 2, options)) { + case -1: + return parse_options_usage(usagestr, options); + case -2: + goto unknown; + } + continue; +unknown: + if (!(ctx->flags & PARSE_OPT_KEEP_UNKNOWN)) + return PARSE_OPT_UNKNOWN; + ctx->out[ctx->cpidx++] = ctx->argv[0]; + ctx->opt = NULL; + } + return PARSE_OPT_DONE; +} + +int parse_options_end(struct parse_opt_ctx_t *ctx) +{ + memmove(ctx->out + ctx->cpidx, ctx->argv, ctx->argc * sizeof(*ctx->out)); + ctx->out[ctx->cpidx + ctx->argc] = NULL; + return ctx->cpidx + ctx->argc; +} + +int parse_options(int argc, const char **argv, const struct option *options, + const char * const usagestr[], int flags) +{ + struct parse_opt_ctx_t ctx; + + parse_options_start(&ctx, argc, argv, flags); + switch (parse_options_step(&ctx, options, usagestr)) { + case PARSE_OPT_HELP: + exit(129); + case PARSE_OPT_DONE: + break; + default: /* PARSE_OPT_UNKNOWN */ + if (ctx.argv[0][1] == '-') { + error("unknown option `%s'", ctx.argv[0] + 2); + } else { + error("unknown switch `%c'", *ctx.opt); + } + usage_with_options(usagestr, options); + } + + return parse_options_end(&ctx); +} + +#define USAGE_OPTS_WIDTH 24 +#define USAGE_GAP 2 + +int usage_with_options_internal(const char * const *usagestr, + const struct option *opts, int full) +{ + if (!usagestr) + return PARSE_OPT_HELP; + + fprintf(stderr, "\n usage: %s\n", *usagestr++); + while (*usagestr && **usagestr) + fprintf(stderr, " or: %s\n", *usagestr++); + while (*usagestr) { + fprintf(stderr, "%s%s\n", + **usagestr ? " " : "", + *usagestr); + usagestr++; + } + + if (opts->type != OPTION_GROUP) + fputc('\n', stderr); + + for (; opts->type != OPTION_END; opts++) { + size_t pos; + int pad; + + if (opts->type == OPTION_GROUP) { + fputc('\n', stderr); + if (*opts->help) + fprintf(stderr, "%s\n", opts->help); + continue; + } + if (!full && (opts->flags & PARSE_OPT_HIDDEN)) + continue; + + pos = fprintf(stderr, " "); + if (opts->short_name) + pos += fprintf(stderr, "-%c", opts->short_name); + if (opts->long_name && opts->short_name) + pos += fprintf(stderr, ", "); + if (opts->long_name) + pos += fprintf(stderr, "--%s", opts->long_name); + + switch (opts->type) { + case OPTION_ARGUMENT: + break; + case OPTION_INTEGER: + if (opts->flags & PARSE_OPT_OPTARG) + if (opts->long_name) + pos += fprintf(stderr, "[=<n>]"); + else + pos += fprintf(stderr, "[<n>]"); + else + pos += fprintf(stderr, " <n>"); + break; + case OPTION_CALLBACK: + if (opts->flags & PARSE_OPT_NOARG) + break; + /* FALLTHROUGH */ + case OPTION_STRING: + if (opts->argh) { + if (opts->flags & PARSE_OPT_OPTARG) + if (opts->long_name) + pos += fprintf(stderr, "[=<%s>]", opts->argh); + else + pos += fprintf(stderr, "[<%s>]", opts->argh); + else + pos += fprintf(stderr, " <%s>", opts->argh); + } else { + if (opts->flags & PARSE_OPT_OPTARG) + if (opts->long_name) + pos += fprintf(stderr, "[=...]"); + else + pos += fprintf(stderr, "[...]"); + else + pos += fprintf(stderr, " ..."); + } + break; + default: /* OPTION_{BIT,BOOLEAN,SET_INT,SET_PTR} */ + break; + } + + if (pos <= USAGE_OPTS_WIDTH) + pad = USAGE_OPTS_WIDTH - pos; + else { + fputc('\n', stderr); + pad = USAGE_OPTS_WIDTH; + } + fprintf(stderr, "%*s%s\n", pad + USAGE_GAP, "", opts->help); + } + fputc('\n', stderr); + + return PARSE_OPT_HELP; +} + +void usage_with_options(const char * const *usagestr, + const struct option *opts) +{ + usage_with_options_internal(usagestr, opts, 0); + exit(129); +} + +int parse_options_usage(const char * const *usagestr, + const struct option *opts) +{ + return usage_with_options_internal(usagestr, opts, 0); +} + + +int parse_opt_verbosity_cb(const struct option *opt, const char *arg, + int unset) +{ + int *target = opt->value; + + if (unset) + /* --no-quiet, --no-verbose */ + *target = 0; + else if (opt->short_name == 'v') { + if (*target >= 0) + (*target)++; + else + *target = 1; + } else { + if (*target <= 0) + (*target)--; + else + *target = -1; + } + return 0; +} diff --git a/tools/perf/util/parse-options.h b/tools/perf/util/parse-options.h new file mode 100644 index 00000000000..a1039a6ce0e --- /dev/null +++ b/tools/perf/util/parse-options.h @@ -0,0 +1,174 @@ +#ifndef PARSE_OPTIONS_H +#define PARSE_OPTIONS_H + +enum parse_opt_type { + /* special types */ + OPTION_END, + OPTION_ARGUMENT, + OPTION_GROUP, + /* options with no arguments */ + OPTION_BIT, + OPTION_BOOLEAN, /* _INCR would have been a better name */ + OPTION_SET_INT, + OPTION_SET_PTR, + /* options with arguments (usually) */ + OPTION_STRING, + OPTION_INTEGER, + OPTION_LONG, + OPTION_CALLBACK, +}; + +enum parse_opt_flags { + PARSE_OPT_KEEP_DASHDASH = 1, + PARSE_OPT_STOP_AT_NON_OPTION = 2, + PARSE_OPT_KEEP_ARGV0 = 4, + PARSE_OPT_KEEP_UNKNOWN = 8, + PARSE_OPT_NO_INTERNAL_HELP = 16, +}; + +enum parse_opt_option_flags { + PARSE_OPT_OPTARG = 1, + PARSE_OPT_NOARG = 2, + PARSE_OPT_NONEG = 4, + PARSE_OPT_HIDDEN = 8, + PARSE_OPT_LASTARG_DEFAULT = 16, +}; + +struct option; +typedef int parse_opt_cb(const struct option *, const char *arg, int unset); + +/* + * `type`:: + * holds the type of the option, you must have an OPTION_END last in your + * array. + * + * `short_name`:: + * the character to use as a short option name, '\0' if none. + * + * `long_name`:: + * the long option name, without the leading dashes, NULL if none. + * + * `value`:: + * stores pointers to the values to be filled. + * + * `argh`:: + * token to explain the kind of argument this option wants. Keep it + * homogenous across the repository. + * + * `help`:: + * the short help associated to what the option does. + * Must never be NULL (except for OPTION_END). + * OPTION_GROUP uses this pointer to store the group header. + * + * `flags`:: + * mask of parse_opt_option_flags. + * PARSE_OPT_OPTARG: says that the argument is optionnal (not for BOOLEANs) + * PARSE_OPT_NOARG: says that this option takes no argument, for CALLBACKs + * PARSE_OPT_NONEG: says that this option cannot be negated + * PARSE_OPT_HIDDEN this option is skipped in the default usage, showed in + * the long one. + * + * `callback`:: + * pointer to the callback to use for OPTION_CALLBACK. + * + * `defval`:: + * default value to fill (*->value) with for PARSE_OPT_OPTARG. + * OPTION_{BIT,SET_INT,SET_PTR} store the {mask,integer,pointer} to put in + * the value when met. + * CALLBACKS can use it like they want. + */ +struct option { + enum parse_opt_type type; + int short_name; + const char *long_name; + void *value; + const char *argh; + const char *help; + + int flags; + parse_opt_cb *callback; + intptr_t defval; +}; + +#define OPT_END() { OPTION_END } +#define OPT_ARGUMENT(l, h) { OPTION_ARGUMENT, 0, (l), NULL, NULL, (h) } +#define OPT_GROUP(h) { OPTION_GROUP, 0, NULL, NULL, NULL, (h) } +#define OPT_BIT(s, l, v, h, b) { OPTION_BIT, (s), (l), (v), NULL, (h), 0, NULL, (b) } +#define OPT_BOOLEAN(s, l, v, h) { OPTION_BOOLEAN, (s), (l), (v), NULL, (h) } +#define OPT_SET_INT(s, l, v, h, i) { OPTION_SET_INT, (s), (l), (v), NULL, (h), 0, NULL, (i) } +#define OPT_SET_PTR(s, l, v, h, p) { OPTION_SET_PTR, (s), (l), (v), NULL, (h), 0, NULL, (p) } +#define OPT_INTEGER(s, l, v, h) { OPTION_INTEGER, (s), (l), (v), NULL, (h) } +#define OPT_LONG(s, l, v, h) { OPTION_LONG, (s), (l), (v), NULL, (h) } +#define OPT_STRING(s, l, v, a, h) { OPTION_STRING, (s), (l), (v), (a), (h) } +#define OPT_DATE(s, l, v, h) \ + { OPTION_CALLBACK, (s), (l), (v), "time",(h), 0, \ + parse_opt_approxidate_cb } +#define OPT_CALLBACK(s, l, v, a, h, f) \ + { OPTION_CALLBACK, (s), (l), (v), (a), (h), 0, (f) } + +/* parse_options() will filter out the processed options and leave the + * non-option argments in argv[]. + * Returns the number of arguments left in argv[]. + */ +extern int parse_options(int argc, const char **argv, + const struct option *options, + const char * const usagestr[], int flags); + +extern NORETURN void usage_with_options(const char * const *usagestr, + const struct option *options); + +/*----- incremantal advanced APIs -----*/ + +enum { + PARSE_OPT_HELP = -1, + PARSE_OPT_DONE, + PARSE_OPT_UNKNOWN, +}; + +/* + * It's okay for the caller to consume argv/argc in the usual way. + * Other fields of that structure are private to parse-options and should not + * be modified in any way. + */ +struct parse_opt_ctx_t { + const char **argv; + const char **out; + int argc, cpidx; + const char *opt; + int flags; +}; + +extern int parse_options_usage(const char * const *usagestr, + const struct option *opts); + +extern void parse_options_start(struct parse_opt_ctx_t *ctx, + int argc, const char **argv, int flags); + +extern int parse_options_step(struct parse_opt_ctx_t *ctx, + const struct option *options, + const char * const usagestr[]); + +extern int parse_options_end(struct parse_opt_ctx_t *ctx); + + +/*----- some often used options -----*/ +extern int parse_opt_abbrev_cb(const struct option *, const char *, int); +extern int parse_opt_approxidate_cb(const struct option *, const char *, int); +extern int parse_opt_verbosity_cb(const struct option *, const char *, int); + +#define OPT__VERBOSE(var) OPT_BOOLEAN('v', "verbose", (var), "be verbose") +#define OPT__QUIET(var) OPT_BOOLEAN('q', "quiet", (var), "be quiet") +#define OPT__VERBOSITY(var) \ + { OPTION_CALLBACK, 'v', "verbose", (var), NULL, "be more verbose", \ + PARSE_OPT_NOARG, &parse_opt_verbosity_cb, 0 }, \ + { OPTION_CALLBACK, 'q', "quiet", (var), NULL, "be more quiet", \ + PARSE_OPT_NOARG, &parse_opt_verbosity_cb, 0 } +#define OPT__DRY_RUN(var) OPT_BOOLEAN('n', "dry-run", (var), "dry run") +#define OPT__ABBREV(var) \ + { OPTION_CALLBACK, 0, "abbrev", (var), "n", \ + "use <n> digits to display SHA-1s", \ + PARSE_OPT_OPTARG, &parse_opt_abbrev_cb, 0 } + +extern const char *parse_options_fix_filename(const char *prefix, const char *file); + +#endif diff --git a/tools/perf/util/path.c b/tools/perf/util/path.c new file mode 100644 index 00000000000..a501a40dd2c --- /dev/null +++ b/tools/perf/util/path.c @@ -0,0 +1,353 @@ +/* + * I'm tired of doing "vsnprintf()" etc just to open a + * file, so here's a "return static buffer with printf" + * interface for paths. + * + * It's obviously not thread-safe. Sue me. But it's quite + * useful for doing things like + * + * f = open(mkpath("%s/%s.perf", base, name), O_RDONLY); + * + * which is what it's designed for. + */ +#include "cache.h" + +static char bad_path[] = "/bad-path/"; +/* + * Two hacks: + */ + +static char *get_perf_dir(void) +{ + return "."; +} + +size_t strlcpy(char *dest, const char *src, size_t size) +{ + size_t ret = strlen(src); + + if (size) { + size_t len = (ret >= size) ? size - 1 : ret; + memcpy(dest, src, len); + dest[len] = '\0'; + } + return ret; +} + + +static char *get_pathname(void) +{ + static char pathname_array[4][PATH_MAX]; + static int index; + return pathname_array[3 & ++index]; +} + +static char *cleanup_path(char *path) +{ + /* Clean it up */ + if (!memcmp(path, "./", 2)) { + path += 2; + while (*path == '/') + path++; + } + return path; +} + +char *mksnpath(char *buf, size_t n, const char *fmt, ...) +{ + va_list args; + unsigned len; + + va_start(args, fmt); + len = vsnprintf(buf, n, fmt, args); + va_end(args); + if (len >= n) { + strlcpy(buf, bad_path, n); + return buf; + } + return cleanup_path(buf); +} + +static char *perf_vsnpath(char *buf, size_t n, const char *fmt, va_list args) +{ + const char *perf_dir = get_perf_dir(); + size_t len; + + len = strlen(perf_dir); + if (n < len + 1) + goto bad; + memcpy(buf, perf_dir, len); + if (len && !is_dir_sep(perf_dir[len-1])) + buf[len++] = '/'; + len += vsnprintf(buf + len, n - len, fmt, args); + if (len >= n) + goto bad; + return cleanup_path(buf); +bad: + strlcpy(buf, bad_path, n); + return buf; +} + +char *perf_snpath(char *buf, size_t n, const char *fmt, ...) +{ + va_list args; + va_start(args, fmt); + (void)perf_vsnpath(buf, n, fmt, args); + va_end(args); + return buf; +} + +char *perf_pathdup(const char *fmt, ...) +{ + char path[PATH_MAX]; + va_list args; + va_start(args, fmt); + (void)perf_vsnpath(path, sizeof(path), fmt, args); + va_end(args); + return xstrdup(path); +} + +char *mkpath(const char *fmt, ...) +{ + va_list args; + unsigned len; + char *pathname = get_pathname(); + + va_start(args, fmt); + len = vsnprintf(pathname, PATH_MAX, fmt, args); + va_end(args); + if (len >= PATH_MAX) + return bad_path; + return cleanup_path(pathname); +} + +char *perf_path(const char *fmt, ...) +{ + const char *perf_dir = get_perf_dir(); + char *pathname = get_pathname(); + va_list args; + unsigned len; + + len = strlen(perf_dir); + if (len > PATH_MAX-100) + return bad_path; + memcpy(pathname, perf_dir, len); + if (len && perf_dir[len-1] != '/') + pathname[len++] = '/'; + va_start(args, fmt); + len += vsnprintf(pathname + len, PATH_MAX - len, fmt, args); + va_end(args); + if (len >= PATH_MAX) + return bad_path; + return cleanup_path(pathname); +} + + +/* perf_mkstemp() - create tmp file honoring TMPDIR variable */ +int perf_mkstemp(char *path, size_t len, const char *template) +{ + const char *tmp; + size_t n; + + tmp = getenv("TMPDIR"); + if (!tmp) + tmp = "/tmp"; + n = snprintf(path, len, "%s/%s", tmp, template); + if (len <= n) { + errno = ENAMETOOLONG; + return -1; + } + return mkstemp(path); +} + + +const char *make_relative_path(const char *abs, const char *base) +{ + static char buf[PATH_MAX + 1]; + int baselen; + if (!base) + return abs; + baselen = strlen(base); + if (prefixcmp(abs, base)) + return abs; + if (abs[baselen] == '/') + baselen++; + else if (base[baselen - 1] != '/') + return abs; + strcpy(buf, abs + baselen); + return buf; +} + +/* + * It is okay if dst == src, but they should not overlap otherwise. + * + * Performs the following normalizations on src, storing the result in dst: + * - Ensures that components are separated by '/' (Windows only) + * - Squashes sequences of '/'. + * - Removes "." components. + * - Removes ".." components, and the components the precede them. + * Returns failure (non-zero) if a ".." component appears as first path + * component anytime during the normalization. Otherwise, returns success (0). + * + * Note that this function is purely textual. It does not follow symlinks, + * verify the existence of the path, or make any system calls. + */ +int normalize_path_copy(char *dst, const char *src) +{ + char *dst0; + + if (has_dos_drive_prefix(src)) { + *dst++ = *src++; + *dst++ = *src++; + } + dst0 = dst; + + if (is_dir_sep(*src)) { + *dst++ = '/'; + while (is_dir_sep(*src)) + src++; + } + + for (;;) { + char c = *src; + + /* + * A path component that begins with . could be + * special: + * (1) "." and ends -- ignore and terminate. + * (2) "./" -- ignore them, eat slash and continue. + * (3) ".." and ends -- strip one and terminate. + * (4) "../" -- strip one, eat slash and continue. + */ + if (c == '.') { + if (!src[1]) { + /* (1) */ + src++; + } else if (is_dir_sep(src[1])) { + /* (2) */ + src += 2; + while (is_dir_sep(*src)) + src++; + continue; + } else if (src[1] == '.') { + if (!src[2]) { + /* (3) */ + src += 2; + goto up_one; + } else if (is_dir_sep(src[2])) { + /* (4) */ + src += 3; + while (is_dir_sep(*src)) + src++; + goto up_one; + } + } + } + + /* copy up to the next '/', and eat all '/' */ + while ((c = *src++) != '\0' && !is_dir_sep(c)) + *dst++ = c; + if (is_dir_sep(c)) { + *dst++ = '/'; + while (is_dir_sep(c)) + c = *src++; + src--; + } else if (!c) + break; + continue; + + up_one: + /* + * dst0..dst is prefix portion, and dst[-1] is '/'; + * go up one level. + */ + dst--; /* go to trailing '/' */ + if (dst <= dst0) + return -1; + /* Windows: dst[-1] cannot be backslash anymore */ + while (dst0 < dst && dst[-1] != '/') + dst--; + } + *dst = '\0'; + return 0; +} + +/* + * path = Canonical absolute path + * prefix_list = Colon-separated list of absolute paths + * + * Determines, for each path in prefix_list, whether the "prefix" really + * is an ancestor directory of path. Returns the length of the longest + * ancestor directory, excluding any trailing slashes, or -1 if no prefix + * is an ancestor. (Note that this means 0 is returned if prefix_list is + * "/".) "/foo" is not considered an ancestor of "/foobar". Directories + * are not considered to be their own ancestors. path must be in a + * canonical form: empty components, or "." or ".." components are not + * allowed. prefix_list may be null, which is like "". + */ +int longest_ancestor_length(const char *path, const char *prefix_list) +{ + char buf[PATH_MAX+1]; + const char *ceil, *colon; + int len, max_len = -1; + + if (prefix_list == NULL || !strcmp(path, "/")) + return -1; + + for (colon = ceil = prefix_list; *colon; ceil = colon+1) { + for (colon = ceil; *colon && *colon != PATH_SEP; colon++); + len = colon - ceil; + if (len == 0 || len > PATH_MAX || !is_absolute_path(ceil)) + continue; + strlcpy(buf, ceil, len+1); + if (normalize_path_copy(buf, buf) < 0) + continue; + len = strlen(buf); + if (len > 0 && buf[len-1] == '/') + buf[--len] = '\0'; + + if (!strncmp(path, buf, len) && + path[len] == '/' && + len > max_len) { + max_len = len; + } + } + + return max_len; +} + +/* strip arbitrary amount of directory separators at end of path */ +static inline int chomp_trailing_dir_sep(const char *path, int len) +{ + while (len && is_dir_sep(path[len - 1])) + len--; + return len; +} + +/* + * If path ends with suffix (complete path components), returns the + * part before suffix (sans trailing directory separators). + * Otherwise returns NULL. + */ +char *strip_path_suffix(const char *path, const char *suffix) +{ + int path_len = strlen(path), suffix_len = strlen(suffix); + + while (suffix_len) { + if (!path_len) + return NULL; + + if (is_dir_sep(path[path_len - 1])) { + if (!is_dir_sep(suffix[suffix_len - 1])) + return NULL; + path_len = chomp_trailing_dir_sep(path, path_len); + suffix_len = chomp_trailing_dir_sep(suffix, suffix_len); + } + else if (path[--path_len] != suffix[--suffix_len]) + return NULL; + } + + if (path_len && !is_dir_sep(path[path_len - 1])) + return NULL; + return xstrndup(path, chomp_trailing_dir_sep(path, path_len)); +} diff --git a/tools/perf/util/quote.c b/tools/perf/util/quote.c new file mode 100644 index 00000000000..f18c5212bc9 --- /dev/null +++ b/tools/perf/util/quote.c @@ -0,0 +1,481 @@ +#include "cache.h" +#include "quote.h" + +int quote_path_fully = 1; + +/* Help to copy the thing properly quoted for the shell safety. + * any single quote is replaced with '\'', any exclamation point + * is replaced with '\!', and the whole thing is enclosed in a + * + * E.g. + * original sq_quote result + * name ==> name ==> 'name' + * a b ==> a b ==> 'a b' + * a'b ==> a'\''b ==> 'a'\''b' + * a!b ==> a'\!'b ==> 'a'\!'b' + */ +static inline int need_bs_quote(char c) +{ + return (c == '\'' || c == '!'); +} + +void sq_quote_buf(struct strbuf *dst, const char *src) +{ + char *to_free = NULL; + + if (dst->buf == src) + to_free = strbuf_detach(dst, NULL); + + strbuf_addch(dst, '\''); + while (*src) { + size_t len = strcspn(src, "'!"); + strbuf_add(dst, src, len); + src += len; + while (need_bs_quote(*src)) { + strbuf_addstr(dst, "'\\"); + strbuf_addch(dst, *src++); + strbuf_addch(dst, '\''); + } + } + strbuf_addch(dst, '\''); + free(to_free); +} + +void sq_quote_print(FILE *stream, const char *src) +{ + char c; + + fputc('\'', stream); + while ((c = *src++)) { + if (need_bs_quote(c)) { + fputs("'\\", stream); + fputc(c, stream); + fputc('\'', stream); + } else { + fputc(c, stream); + } + } + fputc('\'', stream); +} + +void sq_quote_argv(struct strbuf *dst, const char** argv, size_t maxlen) +{ + int i; + + /* Copy into destination buffer. */ + strbuf_grow(dst, 255); + for (i = 0; argv[i]; ++i) { + strbuf_addch(dst, ' '); + sq_quote_buf(dst, argv[i]); + if (maxlen && dst->len > maxlen) + die("Too many or long arguments"); + } +} + +char *sq_dequote_step(char *arg, char **next) +{ + char *dst = arg; + char *src = arg; + char c; + + if (*src != '\'') + return NULL; + for (;;) { + c = *++src; + if (!c) + return NULL; + if (c != '\'') { + *dst++ = c; + continue; + } + /* We stepped out of sq */ + switch (*++src) { + case '\0': + *dst = 0; + if (next) + *next = NULL; + return arg; + case '\\': + c = *++src; + if (need_bs_quote(c) && *++src == '\'') { + *dst++ = c; + continue; + } + /* Fallthrough */ + default: + if (!next || !isspace(*src)) + return NULL; + do { + c = *++src; + } while (isspace(c)); + *dst = 0; + *next = src; + return arg; + } + } +} + +char *sq_dequote(char *arg) +{ + return sq_dequote_step(arg, NULL); +} + +int sq_dequote_to_argv(char *arg, const char ***argv, int *nr, int *alloc) +{ + char *next = arg; + + if (!*arg) + return 0; + do { + char *dequoted = sq_dequote_step(next, &next); + if (!dequoted) + return -1; + ALLOC_GROW(*argv, *nr + 1, *alloc); + (*argv)[(*nr)++] = dequoted; + } while (next); + + return 0; +} + +/* 1 means: quote as octal + * 0 means: quote as octal if (quote_path_fully) + * -1 means: never quote + * c: quote as "\\c" + */ +#define X8(x) x, x, x, x, x, x, x, x +#define X16(x) X8(x), X8(x) +static signed char const sq_lookup[256] = { + /* 0 1 2 3 4 5 6 7 */ + /* 0x00 */ 1, 1, 1, 1, 1, 1, 1, 'a', + /* 0x08 */ 'b', 't', 'n', 'v', 'f', 'r', 1, 1, + /* 0x10 */ X16(1), + /* 0x20 */ -1, -1, '"', -1, -1, -1, -1, -1, + /* 0x28 */ X16(-1), X16(-1), X16(-1), + /* 0x58 */ -1, -1, -1, -1,'\\', -1, -1, -1, + /* 0x60 */ X16(-1), X8(-1), + /* 0x78 */ -1, -1, -1, -1, -1, -1, -1, 1, + /* 0x80 */ /* set to 0 */ +}; + +static inline int sq_must_quote(char c) +{ + return sq_lookup[(unsigned char)c] + quote_path_fully > 0; +} + +/* returns the longest prefix not needing a quote up to maxlen if positive. + This stops at the first \0 because it's marked as a character needing an + escape */ +static size_t next_quote_pos(const char *s, ssize_t maxlen) +{ + size_t len; + if (maxlen < 0) { + for (len = 0; !sq_must_quote(s[len]); len++); + } else { + for (len = 0; len < maxlen && !sq_must_quote(s[len]); len++); + } + return len; +} + +/* + * C-style name quoting. + * + * (1) if sb and fp are both NULL, inspect the input name and counts the + * number of bytes that are needed to hold c_style quoted version of name, + * counting the double quotes around it but not terminating NUL, and + * returns it. + * However, if name does not need c_style quoting, it returns 0. + * + * (2) if sb or fp are not NULL, it emits the c_style quoted version + * of name, enclosed with double quotes if asked and needed only. + * Return value is the same as in (1). + */ +static size_t quote_c_style_counted(const char *name, ssize_t maxlen, + struct strbuf *sb, FILE *fp, int no_dq) +{ +#undef EMIT +#define EMIT(c) \ + do { \ + if (sb) strbuf_addch(sb, (c)); \ + if (fp) fputc((c), fp); \ + count++; \ + } while (0) +#define EMITBUF(s, l) \ + do { \ + int __ret; \ + if (sb) strbuf_add(sb, (s), (l)); \ + if (fp) __ret = fwrite((s), (l), 1, fp); \ + count += (l); \ + } while (0) + + size_t len, count = 0; + const char *p = name; + + for (;;) { + int ch; + + len = next_quote_pos(p, maxlen); + if (len == maxlen || !p[len]) + break; + + if (!no_dq && p == name) + EMIT('"'); + + EMITBUF(p, len); + EMIT('\\'); + p += len; + ch = (unsigned char)*p++; + if (sq_lookup[ch] >= ' ') { + EMIT(sq_lookup[ch]); + } else { + EMIT(((ch >> 6) & 03) + '0'); + EMIT(((ch >> 3) & 07) + '0'); + EMIT(((ch >> 0) & 07) + '0'); + } + } + + EMITBUF(p, len); + if (p == name) /* no ending quote needed */ + return 0; + + if (!no_dq) + EMIT('"'); + return count; +} + +size_t quote_c_style(const char *name, struct strbuf *sb, FILE *fp, int nodq) +{ + return quote_c_style_counted(name, -1, sb, fp, nodq); +} + +void quote_two_c_style(struct strbuf *sb, const char *prefix, const char *path, int nodq) +{ + if (quote_c_style(prefix, NULL, NULL, 0) || + quote_c_style(path, NULL, NULL, 0)) { + if (!nodq) + strbuf_addch(sb, '"'); + quote_c_style(prefix, sb, NULL, 1); + quote_c_style(path, sb, NULL, 1); + if (!nodq) + strbuf_addch(sb, '"'); + } else { + strbuf_addstr(sb, prefix); + strbuf_addstr(sb, path); + } +} + +void write_name_quoted(const char *name, FILE *fp, int terminator) +{ + if (terminator) { + quote_c_style(name, NULL, fp, 0); + } else { + fputs(name, fp); + } + fputc(terminator, fp); +} + +extern void write_name_quotedpfx(const char *pfx, size_t pfxlen, + const char *name, FILE *fp, int terminator) +{ + int needquote = 0; + + if (terminator) { + needquote = next_quote_pos(pfx, pfxlen) < pfxlen + || name[next_quote_pos(name, -1)]; + } + if (needquote) { + fputc('"', fp); + quote_c_style_counted(pfx, pfxlen, NULL, fp, 1); + quote_c_style(name, NULL, fp, 1); + fputc('"', fp); + } else { + int ret; + + ret = fwrite(pfx, pfxlen, 1, fp); + fputs(name, fp); + } + fputc(terminator, fp); +} + +/* quote path as relative to the given prefix */ +char *quote_path_relative(const char *in, int len, + struct strbuf *out, const char *prefix) +{ + int needquote; + + if (len < 0) + len = strlen(in); + + /* "../" prefix itself does not need quoting, but "in" might. */ + needquote = next_quote_pos(in, len) < len; + strbuf_setlen(out, 0); + strbuf_grow(out, len); + + if (needquote) + strbuf_addch(out, '"'); + if (prefix) { + int off = 0; + while (prefix[off] && off < len && prefix[off] == in[off]) + if (prefix[off] == '/') { + prefix += off + 1; + in += off + 1; + len -= off + 1; + off = 0; + } else + off++; + + for (; *prefix; prefix++) + if (*prefix == '/') + strbuf_addstr(out, "../"); + } + + quote_c_style_counted (in, len, out, NULL, 1); + + if (needquote) + strbuf_addch(out, '"'); + if (!out->len) + strbuf_addstr(out, "./"); + + return out->buf; +} + +/* + * C-style name unquoting. + * + * Quoted should point at the opening double quote. + * + Returns 0 if it was able to unquote the string properly, and appends the + * result in the strbuf `sb'. + * + Returns -1 in case of error, and doesn't touch the strbuf. Though note + * that this function will allocate memory in the strbuf, so calling + * strbuf_release is mandatory whichever result unquote_c_style returns. + * + * Updates endp pointer to point at one past the ending double quote if given. + */ +int unquote_c_style(struct strbuf *sb, const char *quoted, const char **endp) +{ + size_t oldlen = sb->len, len; + int ch, ac; + + if (*quoted++ != '"') + return -1; + + for (;;) { + len = strcspn(quoted, "\"\\"); + strbuf_add(sb, quoted, len); + quoted += len; + + switch (*quoted++) { + case '"': + if (endp) + *endp = quoted; + return 0; + case '\\': + break; + default: + goto error; + } + + switch ((ch = *quoted++)) { + case 'a': ch = '\a'; break; + case 'b': ch = '\b'; break; + case 'f': ch = '\f'; break; + case 'n': ch = '\n'; break; + case 'r': ch = '\r'; break; + case 't': ch = '\t'; break; + case 'v': ch = '\v'; break; + + case '\\': case '"': + break; /* verbatim */ + + /* octal values with first digit over 4 overflow */ + case '0': case '1': case '2': case '3': + ac = ((ch - '0') << 6); + if ((ch = *quoted++) < '0' || '7' < ch) + goto error; + ac |= ((ch - '0') << 3); + if ((ch = *quoted++) < '0' || '7' < ch) + goto error; + ac |= (ch - '0'); + ch = ac; + break; + default: + goto error; + } + strbuf_addch(sb, ch); + } + + error: + strbuf_setlen(sb, oldlen); + return -1; +} + +/* quoting as a string literal for other languages */ + +void perl_quote_print(FILE *stream, const char *src) +{ + const char sq = '\''; + const char bq = '\\'; + char c; + + fputc(sq, stream); + while ((c = *src++)) { + if (c == sq || c == bq) + fputc(bq, stream); + fputc(c, stream); + } + fputc(sq, stream); +} + +void python_quote_print(FILE *stream, const char *src) +{ + const char sq = '\''; + const char bq = '\\'; + const char nl = '\n'; + char c; + + fputc(sq, stream); + while ((c = *src++)) { + if (c == nl) { + fputc(bq, stream); + fputc('n', stream); + continue; + } + if (c == sq || c == bq) + fputc(bq, stream); + fputc(c, stream); + } + fputc(sq, stream); +} + +void tcl_quote_print(FILE *stream, const char *src) +{ + char c; + + fputc('"', stream); + while ((c = *src++)) { + switch (c) { + case '[': case ']': + case '{': case '}': + case '$': case '\\': case '"': + fputc('\\', stream); + default: + fputc(c, stream); + break; + case '\f': + fputs("\\f", stream); + break; + case '\r': + fputs("\\r", stream); + break; + case '\n': + fputs("\\n", stream); + break; + case '\t': + fputs("\\t", stream); + break; + case '\v': + fputs("\\v", stream); + break; + } + } + fputc('"', stream); +} diff --git a/tools/perf/util/quote.h b/tools/perf/util/quote.h new file mode 100644 index 00000000000..5dfad89816d --- /dev/null +++ b/tools/perf/util/quote.h @@ -0,0 +1,68 @@ +#ifndef QUOTE_H +#define QUOTE_H + +#include <stddef.h> +#include <stdio.h> + +/* Help to copy the thing properly quoted for the shell safety. + * any single quote is replaced with '\'', any exclamation point + * is replaced with '\!', and the whole thing is enclosed in a + * single quote pair. + * + * For example, if you are passing the result to system() as an + * argument: + * + * sprintf(cmd, "foobar %s %s", sq_quote(arg0), sq_quote(arg1)) + * + * would be appropriate. If the system() is going to call ssh to + * run the command on the other side: + * + * sprintf(cmd, "git-diff-tree %s %s", sq_quote(arg0), sq_quote(arg1)); + * sprintf(rcmd, "ssh %s %s", sq_util/quote.host), sq_quote(cmd)); + * + * Note that the above examples leak memory! Remember to free result from + * sq_quote() in a real application. + * + * sq_quote_buf() writes to an existing buffer of specified size; it + * will return the number of characters that would have been written + * excluding the final null regardless of the buffer size. + */ + +extern void sq_quote_print(FILE *stream, const char *src); + +extern void sq_quote_buf(struct strbuf *, const char *src); +extern void sq_quote_argv(struct strbuf *, const char **argv, size_t maxlen); + +/* This unwraps what sq_quote() produces in place, but returns + * NULL if the input does not look like what sq_quote would have + * produced. + */ +extern char *sq_dequote(char *); + +/* + * Same as the above, but can be used to unwrap many arguments in the + * same string separated by space. "next" is changed to point to the + * next argument that should be passed as first parameter. When there + * is no more argument to be dequoted, "next" is updated to point to NULL. + */ +extern char *sq_dequote_step(char *arg, char **next); +extern int sq_dequote_to_argv(char *arg, const char ***argv, int *nr, int *alloc); + +extern int unquote_c_style(struct strbuf *, const char *quoted, const char **endp); +extern size_t quote_c_style(const char *name, struct strbuf *, FILE *, int no_dq); +extern void quote_two_c_style(struct strbuf *, const char *, const char *, int); + +extern void write_name_quoted(const char *name, FILE *, int terminator); +extern void write_name_quotedpfx(const char *pfx, size_t pfxlen, + const char *name, FILE *, int terminator); + +/* quote path as relative to the given prefix */ +char *quote_path_relative(const char *in, int len, + struct strbuf *out, const char *prefix); + +/* quoting as a string literal for other languages */ +extern void perl_quote_print(FILE *stream, const char *src); +extern void python_quote_print(FILE *stream, const char *src); +extern void tcl_quote_print(FILE *stream, const char *src); + +#endif diff --git a/tools/perf/util/rbtree.c b/tools/perf/util/rbtree.c new file mode 100644 index 00000000000..b15ba9c7cb3 --- /dev/null +++ b/tools/perf/util/rbtree.c @@ -0,0 +1,383 @@ +/* + Red Black Trees + (C) 1999 Andrea Arcangeli <andrea@suse.de> + (C) 2002 David Woodhouse <dwmw2@infradead.org> + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + + linux/lib/rbtree.c +*/ + +#include "rbtree.h" + +static void __rb_rotate_left(struct rb_node *node, struct rb_root *root) +{ + struct rb_node *right = node->rb_right; + struct rb_node *parent = rb_parent(node); + + if ((node->rb_right = right->rb_left)) + rb_set_parent(right->rb_left, node); + right->rb_left = node; + + rb_set_parent(right, parent); + + if (parent) + { + if (node == parent->rb_left) + parent->rb_left = right; + else + parent->rb_right = right; + } + else + root->rb_node = right; + rb_set_parent(node, right); +} + +static void __rb_rotate_right(struct rb_node *node, struct rb_root *root) +{ + struct rb_node *left = node->rb_left; + struct rb_node *parent = rb_parent(node); + + if ((node->rb_left = left->rb_right)) + rb_set_parent(left->rb_right, node); + left->rb_right = node; + + rb_set_parent(left, parent); + + if (parent) + { + if (node == parent->rb_right) + parent->rb_right = left; + else + parent->rb_left = left; + } + else + root->rb_node = left; + rb_set_parent(node, left); +} + +void rb_insert_color(struct rb_node *node, struct rb_root *root) +{ + struct rb_node *parent, *gparent; + + while ((parent = rb_parent(node)) && rb_is_red(parent)) + { + gparent = rb_parent(parent); + + if (parent == gparent->rb_left) + { + { + register struct rb_node *uncle = gparent->rb_right; + if (uncle && rb_is_red(uncle)) + { + rb_set_black(uncle); + rb_set_black(parent); + rb_set_red(gparent); + node = gparent; + continue; + } + } + + if (parent->rb_right == node) + { + register struct rb_node *tmp; + __rb_rotate_left(parent, root); + tmp = parent; + parent = node; + node = tmp; + } + + rb_set_black(parent); + rb_set_red(gparent); + __rb_rotate_right(gparent, root); + } else { + { + register struct rb_node *uncle = gparent->rb_left; + if (uncle && rb_is_red(uncle)) + { + rb_set_black(uncle); + rb_set_black(parent); + rb_set_red(gparent); + node = gparent; + continue; + } + } + + if (parent->rb_left == node) + { + register struct rb_node *tmp; + __rb_rotate_right(parent, root); + tmp = parent; + parent = node; + node = tmp; + } + + rb_set_black(parent); + rb_set_red(gparent); + __rb_rotate_left(gparent, root); + } + } + + rb_set_black(root->rb_node); +} + +static void __rb_erase_color(struct rb_node *node, struct rb_node *parent, + struct rb_root *root) +{ + struct rb_node *other; + + while ((!node || rb_is_black(node)) && node != root->rb_node) + { + if (parent->rb_left == node) + { + other = parent->rb_right; + if (rb_is_red(other)) + { + rb_set_black(other); + rb_set_red(parent); + __rb_rotate_left(parent, root); + other = parent->rb_right; + } + if ((!other->rb_left || rb_is_black(other->rb_left)) && + (!other->rb_right || rb_is_black(other->rb_right))) + { + rb_set_red(other); + node = parent; + parent = rb_parent(node); + } + else + { + if (!other->rb_right || rb_is_black(other->rb_right)) + { + rb_set_black(other->rb_left); + rb_set_red(other); + __rb_rotate_right(other, root); + other = parent->rb_right; + } + rb_set_color(other, rb_color(parent)); + rb_set_black(parent); + rb_set_black(other->rb_right); + __rb_rotate_left(parent, root); + node = root->rb_node; + break; + } + } + else + { + other = parent->rb_left; + if (rb_is_red(other)) + { + rb_set_black(other); + rb_set_red(parent); + __rb_rotate_right(parent, root); + other = parent->rb_left; + } + if ((!other->rb_left || rb_is_black(other->rb_left)) && + (!other->rb_right || rb_is_black(other->rb_right))) + { + rb_set_red(other); + node = parent; + parent = rb_parent(node); + } + else + { + if (!other->rb_left || rb_is_black(other->rb_left)) + { + rb_set_black(other->rb_right); + rb_set_red(other); + __rb_rotate_left(other, root); + other = parent->rb_left; + } + rb_set_color(other, rb_color(parent)); + rb_set_black(parent); + rb_set_black(other->rb_left); + __rb_rotate_right(parent, root); + node = root->rb_node; + break; + } + } + } + if (node) + rb_set_black(node); +} + +void rb_erase(struct rb_node *node, struct rb_root *root) +{ + struct rb_node *child, *parent; + int color; + + if (!node->rb_left) + child = node->rb_right; + else if (!node->rb_right) + child = node->rb_left; + else + { + struct rb_node *old = node, *left; + + node = node->rb_right; + while ((left = node->rb_left) != NULL) + node = left; + child = node->rb_right; + parent = rb_parent(node); + color = rb_color(node); + + if (child) + rb_set_parent(child, parent); + if (parent == old) { + parent->rb_right = child; + parent = node; + } else + parent->rb_left = child; + + node->rb_parent_color = old->rb_parent_color; + node->rb_right = old->rb_right; + node->rb_left = old->rb_left; + + if (rb_parent(old)) + { + if (rb_parent(old)->rb_left == old) + rb_parent(old)->rb_left = node; + else + rb_parent(old)->rb_right = node; + } else + root->rb_node = node; + + rb_set_parent(old->rb_left, node); + if (old->rb_right) + rb_set_parent(old->rb_right, node); + goto color; + } + + parent = rb_parent(node); + color = rb_color(node); + + if (child) + rb_set_parent(child, parent); + if (parent) + { + if (parent->rb_left == node) + parent->rb_left = child; + else + parent->rb_right = child; + } + else + root->rb_node = child; + + color: + if (color == RB_BLACK) + __rb_erase_color(child, parent, root); +} + +/* + * This function returns the first node (in sort order) of the tree. + */ +struct rb_node *rb_first(const struct rb_root *root) +{ + struct rb_node *n; + + n = root->rb_node; + if (!n) + return NULL; + while (n->rb_left) + n = n->rb_left; + return n; +} + +struct rb_node *rb_last(const struct rb_root *root) +{ + struct rb_node *n; + + n = root->rb_node; + if (!n) + return NULL; + while (n->rb_right) + n = n->rb_right; + return n; +} + +struct rb_node *rb_next(const struct rb_node *node) +{ + struct rb_node *parent; + + if (rb_parent(node) == node) + return NULL; + + /* If we have a right-hand child, go down and then left as far + as we can. */ + if (node->rb_right) { + node = node->rb_right; + while (node->rb_left) + node=node->rb_left; + return (struct rb_node *)node; + } + + /* No right-hand children. Everything down and left is + smaller than us, so any 'next' node must be in the general + direction of our parent. Go up the tree; any time the + ancestor is a right-hand child of its parent, keep going + up. First time it's a left-hand child of its parent, said + parent is our 'next' node. */ + while ((parent = rb_parent(node)) && node == parent->rb_right) + node = parent; + + return parent; +} + +struct rb_node *rb_prev(const struct rb_node *node) +{ + struct rb_node *parent; + + if (rb_parent(node) == node) + return NULL; + + /* If we have a left-hand child, go down and then right as far + as we can. */ + if (node->rb_left) { + node = node->rb_left; + while (node->rb_right) + node=node->rb_right; + return (struct rb_node *)node; + } + + /* No left-hand children. Go up till we find an ancestor which + is a right-hand child of its parent */ + while ((parent = rb_parent(node)) && node == parent->rb_left) + node = parent; + + return parent; +} + +void rb_replace_node(struct rb_node *victim, struct rb_node *new, + struct rb_root *root) +{ + struct rb_node *parent = rb_parent(victim); + + /* Set the surrounding nodes to point to the replacement */ + if (parent) { + if (victim == parent->rb_left) + parent->rb_left = new; + else + parent->rb_right = new; + } else { + root->rb_node = new; + } + if (victim->rb_left) + rb_set_parent(victim->rb_left, new); + if (victim->rb_right) + rb_set_parent(victim->rb_right, new); + + /* Copy the pointers/colour from the victim to the replacement */ + *new = *victim; +} diff --git a/tools/perf/util/rbtree.h b/tools/perf/util/rbtree.h new file mode 100644 index 00000000000..6bdc488a47f --- /dev/null +++ b/tools/perf/util/rbtree.h @@ -0,0 +1,171 @@ +/* + Red Black Trees + (C) 1999 Andrea Arcangeli <andrea@suse.de> + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + + linux/include/linux/rbtree.h + + To use rbtrees you'll have to implement your own insert and search cores. + This will avoid us to use callbacks and to drop drammatically performances. + I know it's not the cleaner way, but in C (not in C++) to get + performances and genericity... + + Some example of insert and search follows here. The search is a plain + normal search over an ordered tree. The insert instead must be implemented + int two steps: as first thing the code must insert the element in + order as a red leaf in the tree, then the support library function + rb_insert_color() must be called. Such function will do the + not trivial work to rebalance the rbtree if necessary. + +----------------------------------------------------------------------- +static inline struct page * rb_search_page_cache(struct inode * inode, + unsigned long offset) +{ + struct rb_node * n = inode->i_rb_page_cache.rb_node; + struct page * page; + + while (n) + { + page = rb_entry(n, struct page, rb_page_cache); + + if (offset < page->offset) + n = n->rb_left; + else if (offset > page->offset) + n = n->rb_right; + else + return page; + } + return NULL; +} + +static inline struct page * __rb_insert_page_cache(struct inode * inode, + unsigned long offset, + struct rb_node * node) +{ + struct rb_node ** p = &inode->i_rb_page_cache.rb_node; + struct rb_node * parent = NULL; + struct page * page; + + while (*p) + { + parent = *p; + page = rb_entry(parent, struct page, rb_page_cache); + + if (offset < page->offset) + p = &(*p)->rb_left; + else if (offset > page->offset) + p = &(*p)->rb_right; + else + return page; + } + + rb_link_node(node, parent, p); + + return NULL; +} + +static inline struct page * rb_insert_page_cache(struct inode * inode, + unsigned long offset, + struct rb_node * node) +{ + struct page * ret; + if ((ret = __rb_insert_page_cache(inode, offset, node))) + goto out; + rb_insert_color(node, &inode->i_rb_page_cache); + out: + return ret; +} +----------------------------------------------------------------------- +*/ + +#ifndef _LINUX_RBTREE_H +#define _LINUX_RBTREE_H + +#include <stddef.h> + +/** + * container_of - cast a member of a structure out to the containing structure + * @ptr: the pointer to the member. + * @type: the type of the container struct this is embedded in. + * @member: the name of the member within the struct. + * + */ +#define container_of(ptr, type, member) ({ \ + const typeof( ((type *)0)->member ) *__mptr = (ptr); \ + (type *)( (char *)__mptr - offsetof(type,member) );}) + +struct rb_node +{ + unsigned long rb_parent_color; +#define RB_RED 0 +#define RB_BLACK 1 + struct rb_node *rb_right; + struct rb_node *rb_left; +} __attribute__((aligned(sizeof(long)))); + /* The alignment might seem pointless, but allegedly CRIS needs it */ + +struct rb_root +{ + struct rb_node *rb_node; +}; + + +#define rb_parent(r) ((struct rb_node *)((r)->rb_parent_color & ~3)) +#define rb_color(r) ((r)->rb_parent_color & 1) +#define rb_is_red(r) (!rb_color(r)) +#define rb_is_black(r) rb_color(r) +#define rb_set_red(r) do { (r)->rb_parent_color &= ~1; } while (0) +#define rb_set_black(r) do { (r)->rb_parent_color |= 1; } while (0) + +static inline void rb_set_parent(struct rb_node *rb, struct rb_node *p) +{ + rb->rb_parent_color = (rb->rb_parent_color & 3) | (unsigned long)p; +} +static inline void rb_set_color(struct rb_node *rb, int color) +{ + rb->rb_parent_color = (rb->rb_parent_color & ~1) | color; +} + +#define RB_ROOT (struct rb_root) { NULL, } +#define rb_entry(ptr, type, member) container_of(ptr, type, member) + +#define RB_EMPTY_ROOT(root) ((root)->rb_node == NULL) +#define RB_EMPTY_NODE(node) (rb_parent(node) == node) +#define RB_CLEAR_NODE(node) (rb_set_parent(node, node)) + +extern void rb_insert_color(struct rb_node *, struct rb_root *); +extern void rb_erase(struct rb_node *, struct rb_root *); + +/* Find logical next and previous nodes in a tree */ +extern struct rb_node *rb_next(const struct rb_node *); +extern struct rb_node *rb_prev(const struct rb_node *); +extern struct rb_node *rb_first(const struct rb_root *); +extern struct rb_node *rb_last(const struct rb_root *); + +/* Fast replacement of a single node without remove/rebalance/add/rebalance */ +extern void rb_replace_node(struct rb_node *victim, struct rb_node *new, + struct rb_root *root); + +static inline void rb_link_node(struct rb_node * node, struct rb_node * parent, + struct rb_node ** rb_link) +{ + node->rb_parent_color = (unsigned long )parent; + node->rb_left = node->rb_right = NULL; + + *rb_link = node; +} + +#endif /* _LINUX_RBTREE_H */ diff --git a/tools/perf/util/run-command.c b/tools/perf/util/run-command.c new file mode 100644 index 00000000000..b2f5e854f40 --- /dev/null +++ b/tools/perf/util/run-command.c @@ -0,0 +1,395 @@ +#include "cache.h" +#include "run-command.h" +#include "exec_cmd.h" + +static inline void close_pair(int fd[2]) +{ + close(fd[0]); + close(fd[1]); +} + +static inline void dup_devnull(int to) +{ + int fd = open("/dev/null", O_RDWR); + dup2(fd, to); + close(fd); +} + +int start_command(struct child_process *cmd) +{ + int need_in, need_out, need_err; + int fdin[2], fdout[2], fderr[2]; + + /* + * In case of errors we must keep the promise to close FDs + * that have been passed in via ->in and ->out. + */ + + need_in = !cmd->no_stdin && cmd->in < 0; + if (need_in) { + if (pipe(fdin) < 0) { + if (cmd->out > 0) + close(cmd->out); + return -ERR_RUN_COMMAND_PIPE; + } + cmd->in = fdin[1]; + } + + need_out = !cmd->no_stdout + && !cmd->stdout_to_stderr + && cmd->out < 0; + if (need_out) { + if (pipe(fdout) < 0) { + if (need_in) + close_pair(fdin); + else if (cmd->in) + close(cmd->in); + return -ERR_RUN_COMMAND_PIPE; + } + cmd->out = fdout[0]; + } + + need_err = !cmd->no_stderr && cmd->err < 0; + if (need_err) { + if (pipe(fderr) < 0) { + if (need_in) + close_pair(fdin); + else if (cmd->in) + close(cmd->in); + if (need_out) + close_pair(fdout); + else if (cmd->out) + close(cmd->out); + return -ERR_RUN_COMMAND_PIPE; + } + cmd->err = fderr[0]; + } + +#ifndef __MINGW32__ + fflush(NULL); + cmd->pid = fork(); + if (!cmd->pid) { + if (cmd->no_stdin) + dup_devnull(0); + else if (need_in) { + dup2(fdin[0], 0); + close_pair(fdin); + } else if (cmd->in) { + dup2(cmd->in, 0); + close(cmd->in); + } + + if (cmd->no_stderr) + dup_devnull(2); + else if (need_err) { + dup2(fderr[1], 2); + close_pair(fderr); + } + + if (cmd->no_stdout) + dup_devnull(1); + else if (cmd->stdout_to_stderr) + dup2(2, 1); + else if (need_out) { + dup2(fdout[1], 1); + close_pair(fdout); + } else if (cmd->out > 1) { + dup2(cmd->out, 1); + close(cmd->out); + } + + if (cmd->dir && chdir(cmd->dir)) + die("exec %s: cd to %s failed (%s)", cmd->argv[0], + cmd->dir, strerror(errno)); + if (cmd->env) { + for (; *cmd->env; cmd->env++) { + if (strchr(*cmd->env, '=')) + putenv((char*)*cmd->env); + else + unsetenv(*cmd->env); + } + } + if (cmd->preexec_cb) + cmd->preexec_cb(); + if (cmd->perf_cmd) { + execv_perf_cmd(cmd->argv); + } else { + execvp(cmd->argv[0], (char *const*) cmd->argv); + } + exit(127); + } +#else + int s0 = -1, s1 = -1, s2 = -1; /* backups of stdin, stdout, stderr */ + const char **sargv = cmd->argv; + char **env = environ; + + if (cmd->no_stdin) { + s0 = dup(0); + dup_devnull(0); + } else if (need_in) { + s0 = dup(0); + dup2(fdin[0], 0); + } else if (cmd->in) { + s0 = dup(0); + dup2(cmd->in, 0); + } + + if (cmd->no_stderr) { + s2 = dup(2); + dup_devnull(2); + } else if (need_err) { + s2 = dup(2); + dup2(fderr[1], 2); + } + + if (cmd->no_stdout) { + s1 = dup(1); + dup_devnull(1); + } else if (cmd->stdout_to_stderr) { + s1 = dup(1); + dup2(2, 1); + } else if (need_out) { + s1 = dup(1); + dup2(fdout[1], 1); + } else if (cmd->out > 1) { + s1 = dup(1); + dup2(cmd->out, 1); + } + + if (cmd->dir) + die("chdir in start_command() not implemented"); + if (cmd->env) { + env = copy_environ(); + for (; *cmd->env; cmd->env++) + env = env_setenv(env, *cmd->env); + } + + if (cmd->perf_cmd) { + cmd->argv = prepare_perf_cmd(cmd->argv); + } + + cmd->pid = mingw_spawnvpe(cmd->argv[0], cmd->argv, env); + + if (cmd->env) + free_environ(env); + if (cmd->perf_cmd) + free(cmd->argv); + + cmd->argv = sargv; + if (s0 >= 0) + dup2(s0, 0), close(s0); + if (s1 >= 0) + dup2(s1, 1), close(s1); + if (s2 >= 0) + dup2(s2, 2), close(s2); +#endif + + if (cmd->pid < 0) { + int err = errno; + if (need_in) + close_pair(fdin); + else if (cmd->in) + close(cmd->in); + if (need_out) + close_pair(fdout); + else if (cmd->out) + close(cmd->out); + if (need_err) + close_pair(fderr); + return err == ENOENT ? + -ERR_RUN_COMMAND_EXEC : + -ERR_RUN_COMMAND_FORK; + } + + if (need_in) + close(fdin[0]); + else if (cmd->in) + close(cmd->in); + + if (need_out) + close(fdout[1]); + else if (cmd->out) + close(cmd->out); + + if (need_err) + close(fderr[1]); + + return 0; +} + +static int wait_or_whine(pid_t pid) +{ + for (;;) { + int status, code; + pid_t waiting = waitpid(pid, &status, 0); + + if (waiting < 0) { + if (errno == EINTR) + continue; + error("waitpid failed (%s)", strerror(errno)); + return -ERR_RUN_COMMAND_WAITPID; + } + if (waiting != pid) + return -ERR_RUN_COMMAND_WAITPID_WRONG_PID; + if (WIFSIGNALED(status)) + return -ERR_RUN_COMMAND_WAITPID_SIGNAL; + + if (!WIFEXITED(status)) + return -ERR_RUN_COMMAND_WAITPID_NOEXIT; + code = WEXITSTATUS(status); + switch (code) { + case 127: + return -ERR_RUN_COMMAND_EXEC; + case 0: + return 0; + default: + return -code; + } + } +} + +int finish_command(struct child_process *cmd) +{ + return wait_or_whine(cmd->pid); +} + +int run_command(struct child_process *cmd) +{ + int code = start_command(cmd); + if (code) + return code; + return finish_command(cmd); +} + +static void prepare_run_command_v_opt(struct child_process *cmd, + const char **argv, + int opt) +{ + memset(cmd, 0, sizeof(*cmd)); + cmd->argv = argv; + cmd->no_stdin = opt & RUN_COMMAND_NO_STDIN ? 1 : 0; + cmd->perf_cmd = opt & RUN_PERF_CMD ? 1 : 0; + cmd->stdout_to_stderr = opt & RUN_COMMAND_STDOUT_TO_STDERR ? 1 : 0; +} + +int run_command_v_opt(const char **argv, int opt) +{ + struct child_process cmd; + prepare_run_command_v_opt(&cmd, argv, opt); + return run_command(&cmd); +} + +int run_command_v_opt_cd_env(const char **argv, int opt, const char *dir, const char *const *env) +{ + struct child_process cmd; + prepare_run_command_v_opt(&cmd, argv, opt); + cmd.dir = dir; + cmd.env = env; + return run_command(&cmd); +} + +#ifdef __MINGW32__ +static __stdcall unsigned run_thread(void *data) +{ + struct async *async = data; + return async->proc(async->fd_for_proc, async->data); +} +#endif + +int start_async(struct async *async) +{ + int pipe_out[2]; + + if (pipe(pipe_out) < 0) + return error("cannot create pipe: %s", strerror(errno)); + async->out = pipe_out[0]; + +#ifndef __MINGW32__ + /* Flush stdio before fork() to avoid cloning buffers */ + fflush(NULL); + + async->pid = fork(); + if (async->pid < 0) { + error("fork (async) failed: %s", strerror(errno)); + close_pair(pipe_out); + return -1; + } + if (!async->pid) { + close(pipe_out[0]); + exit(!!async->proc(pipe_out[1], async->data)); + } + close(pipe_out[1]); +#else + async->fd_for_proc = pipe_out[1]; + async->tid = (HANDLE) _beginthreadex(NULL, 0, run_thread, async, 0, NULL); + if (!async->tid) { + error("cannot create thread: %s", strerror(errno)); + close_pair(pipe_out); + return -1; + } +#endif + return 0; +} + +int finish_async(struct async *async) +{ +#ifndef __MINGW32__ + int ret = 0; + + if (wait_or_whine(async->pid)) + ret = error("waitpid (async) failed"); +#else + DWORD ret = 0; + if (WaitForSingleObject(async->tid, INFINITE) != WAIT_OBJECT_0) + ret = error("waiting for thread failed: %lu", GetLastError()); + else if (!GetExitCodeThread(async->tid, &ret)) + ret = error("cannot get thread exit code: %lu", GetLastError()); + CloseHandle(async->tid); +#endif + return ret; +} + +int run_hook(const char *index_file, const char *name, ...) +{ + struct child_process hook; + const char **argv = NULL, *env[2]; + char index[PATH_MAX]; + va_list args; + int ret; + size_t i = 0, alloc = 0; + + if (access(perf_path("hooks/%s", name), X_OK) < 0) + return 0; + + va_start(args, name); + ALLOC_GROW(argv, i + 1, alloc); + argv[i++] = perf_path("hooks/%s", name); + while (argv[i-1]) { + ALLOC_GROW(argv, i + 1, alloc); + argv[i++] = va_arg(args, const char *); + } + va_end(args); + + memset(&hook, 0, sizeof(hook)); + hook.argv = argv; + hook.no_stdin = 1; + hook.stdout_to_stderr = 1; + if (index_file) { + snprintf(index, sizeof(index), "PERF_INDEX_FILE=%s", index_file); + env[0] = index; + env[1] = NULL; + hook.env = env; + } + + ret = start_command(&hook); + free(argv); + if (ret) { + warning("Could not spawn %s", argv[0]); + return ret; + } + ret = finish_command(&hook); + if (ret == -ERR_RUN_COMMAND_WAITPID_SIGNAL) + warning("%s exited due to uncaught signal", argv[0]); + + return ret; +} diff --git a/tools/perf/util/run-command.h b/tools/perf/util/run-command.h new file mode 100644 index 00000000000..328289f2366 --- /dev/null +++ b/tools/perf/util/run-command.h @@ -0,0 +1,93 @@ +#ifndef RUN_COMMAND_H +#define RUN_COMMAND_H + +enum { + ERR_RUN_COMMAND_FORK = 10000, + ERR_RUN_COMMAND_EXEC, + ERR_RUN_COMMAND_PIPE, + ERR_RUN_COMMAND_WAITPID, + ERR_RUN_COMMAND_WAITPID_WRONG_PID, + ERR_RUN_COMMAND_WAITPID_SIGNAL, + ERR_RUN_COMMAND_WAITPID_NOEXIT, +}; +#define IS_RUN_COMMAND_ERR(x) (-(x) >= ERR_RUN_COMMAND_FORK) + +struct child_process { + const char **argv; + pid_t pid; + /* + * Using .in, .out, .err: + * - Specify 0 for no redirections (child inherits stdin, stdout, + * stderr from parent). + * - Specify -1 to have a pipe allocated as follows: + * .in: returns the writable pipe end; parent writes to it, + * the readable pipe end becomes child's stdin + * .out, .err: returns the readable pipe end; parent reads from + * it, the writable pipe end becomes child's stdout/stderr + * The caller of start_command() must close the returned FDs + * after it has completed reading from/writing to it! + * - Specify > 0 to set a channel to a particular FD as follows: + * .in: a readable FD, becomes child's stdin + * .out: a writable FD, becomes child's stdout/stderr + * .err > 0 not supported + * The specified FD is closed by start_command(), even in case + * of errors! + */ + int in; + int out; + int err; + const char *dir; + const char *const *env; + unsigned no_stdin:1; + unsigned no_stdout:1; + unsigned no_stderr:1; + unsigned perf_cmd:1; /* if this is to be perf sub-command */ + unsigned stdout_to_stderr:1; + void (*preexec_cb)(void); +}; + +int start_command(struct child_process *); +int finish_command(struct child_process *); +int run_command(struct child_process *); + +extern int run_hook(const char *index_file, const char *name, ...); + +#define RUN_COMMAND_NO_STDIN 1 +#define RUN_PERF_CMD 2 /*If this is to be perf sub-command */ +#define RUN_COMMAND_STDOUT_TO_STDERR 4 +int run_command_v_opt(const char **argv, int opt); + +/* + * env (the environment) is to be formatted like environ: "VAR=VALUE". + * To unset an environment variable use just "VAR". + */ +int run_command_v_opt_cd_env(const char **argv, int opt, const char *dir, const char *const *env); + +/* + * The purpose of the following functions is to feed a pipe by running + * a function asynchronously and providing output that the caller reads. + * + * It is expected that no synchronization and mutual exclusion between + * the caller and the feed function is necessary so that the function + * can run in a thread without interfering with the caller. + */ +struct async { + /* + * proc writes to fd and closes it; + * returns 0 on success, non-zero on failure + */ + int (*proc)(int fd, void *data); + void *data; + int out; /* caller reads from here and closes it */ +#ifndef __MINGW32__ + pid_t pid; +#else + HANDLE tid; + int fd_for_proc; +#endif +}; + +int start_async(struct async *async); +int finish_async(struct async *async); + +#endif diff --git a/tools/perf/util/sigchain.c b/tools/perf/util/sigchain.c new file mode 100644 index 00000000000..1118b99e57d --- /dev/null +++ b/tools/perf/util/sigchain.c @@ -0,0 +1,52 @@ +#include "sigchain.h" +#include "cache.h" + +#define SIGCHAIN_MAX_SIGNALS 32 + +struct sigchain_signal { + sigchain_fun *old; + int n; + int alloc; +}; +static struct sigchain_signal signals[SIGCHAIN_MAX_SIGNALS]; + +static void check_signum(int sig) +{ + if (sig < 1 || sig >= SIGCHAIN_MAX_SIGNALS) + die("BUG: signal out of range: %d", sig); +} + +int sigchain_push(int sig, sigchain_fun f) +{ + struct sigchain_signal *s = signals + sig; + check_signum(sig); + + ALLOC_GROW(s->old, s->n + 1, s->alloc); + s->old[s->n] = signal(sig, f); + if (s->old[s->n] == SIG_ERR) + return -1; + s->n++; + return 0; +} + +int sigchain_pop(int sig) +{ + struct sigchain_signal *s = signals + sig; + check_signum(sig); + if (s->n < 1) + return 0; + + if (signal(sig, s->old[s->n - 1]) == SIG_ERR) + return -1; + s->n--; + return 0; +} + +void sigchain_push_common(sigchain_fun f) +{ + sigchain_push(SIGINT, f); + sigchain_push(SIGHUP, f); + sigchain_push(SIGTERM, f); + sigchain_push(SIGQUIT, f); + sigchain_push(SIGPIPE, f); +} diff --git a/tools/perf/util/sigchain.h b/tools/perf/util/sigchain.h new file mode 100644 index 00000000000..618083bce0c --- /dev/null +++ b/tools/perf/util/sigchain.h @@ -0,0 +1,11 @@ +#ifndef SIGCHAIN_H +#define SIGCHAIN_H + +typedef void (*sigchain_fun)(int); + +int sigchain_push(int sig, sigchain_fun f); +int sigchain_pop(int sig); + +void sigchain_push_common(sigchain_fun f); + +#endif /* SIGCHAIN_H */ diff --git a/tools/perf/util/strbuf.c b/tools/perf/util/strbuf.c new file mode 100644 index 00000000000..eaba0930680 --- /dev/null +++ b/tools/perf/util/strbuf.c @@ -0,0 +1,359 @@ +#include "cache.h" + +int prefixcmp(const char *str, const char *prefix) +{ + for (; ; str++, prefix++) + if (!*prefix) + return 0; + else if (*str != *prefix) + return (unsigned char)*prefix - (unsigned char)*str; +} + +/* + * Used as the default ->buf value, so that people can always assume + * buf is non NULL and ->buf is NUL terminated even for a freshly + * initialized strbuf. + */ +char strbuf_slopbuf[1]; + +void strbuf_init(struct strbuf *sb, size_t hint) +{ + sb->alloc = sb->len = 0; + sb->buf = strbuf_slopbuf; + if (hint) + strbuf_grow(sb, hint); +} + +void strbuf_release(struct strbuf *sb) +{ + if (sb->alloc) { + free(sb->buf); + strbuf_init(sb, 0); + } +} + +char *strbuf_detach(struct strbuf *sb, size_t *sz) +{ + char *res = sb->alloc ? sb->buf : NULL; + if (sz) + *sz = sb->len; + strbuf_init(sb, 0); + return res; +} + +void strbuf_attach(struct strbuf *sb, void *buf, size_t len, size_t alloc) +{ + strbuf_release(sb); + sb->buf = buf; + sb->len = len; + sb->alloc = alloc; + strbuf_grow(sb, 0); + sb->buf[sb->len] = '\0'; +} + +void strbuf_grow(struct strbuf *sb, size_t extra) +{ + if (sb->len + extra + 1 <= sb->len) + die("you want to use way too much memory"); + if (!sb->alloc) + sb->buf = NULL; + ALLOC_GROW(sb->buf, sb->len + extra + 1, sb->alloc); +} + +void strbuf_trim(struct strbuf *sb) +{ + char *b = sb->buf; + while (sb->len > 0 && isspace((unsigned char)sb->buf[sb->len - 1])) + sb->len--; + while (sb->len > 0 && isspace(*b)) { + b++; + sb->len--; + } + memmove(sb->buf, b, sb->len); + sb->buf[sb->len] = '\0'; +} +void strbuf_rtrim(struct strbuf *sb) +{ + while (sb->len > 0 && isspace((unsigned char)sb->buf[sb->len - 1])) + sb->len--; + sb->buf[sb->len] = '\0'; +} + +void strbuf_ltrim(struct strbuf *sb) +{ + char *b = sb->buf; + while (sb->len > 0 && isspace(*b)) { + b++; + sb->len--; + } + memmove(sb->buf, b, sb->len); + sb->buf[sb->len] = '\0'; +} + +void strbuf_tolower(struct strbuf *sb) +{ + int i; + for (i = 0; i < sb->len; i++) + sb->buf[i] = tolower(sb->buf[i]); +} + +struct strbuf **strbuf_split(const struct strbuf *sb, int delim) +{ + int alloc = 2, pos = 0; + char *n, *p; + struct strbuf **ret; + struct strbuf *t; + + ret = calloc(alloc, sizeof(struct strbuf *)); + p = n = sb->buf; + while (n < sb->buf + sb->len) { + int len; + n = memchr(n, delim, sb->len - (n - sb->buf)); + if (pos + 1 >= alloc) { + alloc = alloc * 2; + ret = realloc(ret, sizeof(struct strbuf *) * alloc); + } + if (!n) + n = sb->buf + sb->len - 1; + len = n - p + 1; + t = malloc(sizeof(struct strbuf)); + strbuf_init(t, len); + strbuf_add(t, p, len); + ret[pos] = t; + ret[++pos] = NULL; + p = ++n; + } + return ret; +} + +void strbuf_list_free(struct strbuf **sbs) +{ + struct strbuf **s = sbs; + + while (*s) { + strbuf_release(*s); + free(*s++); + } + free(sbs); +} + +int strbuf_cmp(const struct strbuf *a, const struct strbuf *b) +{ + int len = a->len < b->len ? a->len: b->len; + int cmp = memcmp(a->buf, b->buf, len); + if (cmp) + return cmp; + return a->len < b->len ? -1: a->len != b->len; +} + +void strbuf_splice(struct strbuf *sb, size_t pos, size_t len, + const void *data, size_t dlen) +{ + if (pos + len < pos) + die("you want to use way too much memory"); + if (pos > sb->len) + die("`pos' is too far after the end of the buffer"); + if (pos + len > sb->len) + die("`pos + len' is too far after the end of the buffer"); + + if (dlen >= len) + strbuf_grow(sb, dlen - len); + memmove(sb->buf + pos + dlen, + sb->buf + pos + len, + sb->len - pos - len); + memcpy(sb->buf + pos, data, dlen); + strbuf_setlen(sb, sb->len + dlen - len); +} + +void strbuf_insert(struct strbuf *sb, size_t pos, const void *data, size_t len) +{ + strbuf_splice(sb, pos, 0, data, len); +} + +void strbuf_remove(struct strbuf *sb, size_t pos, size_t len) +{ + strbuf_splice(sb, pos, len, NULL, 0); +} + +void strbuf_add(struct strbuf *sb, const void *data, size_t len) +{ + strbuf_grow(sb, len); + memcpy(sb->buf + sb->len, data, len); + strbuf_setlen(sb, sb->len + len); +} + +void strbuf_adddup(struct strbuf *sb, size_t pos, size_t len) +{ + strbuf_grow(sb, len); + memcpy(sb->buf + sb->len, sb->buf + pos, len); + strbuf_setlen(sb, sb->len + len); +} + +void strbuf_addf(struct strbuf *sb, const char *fmt, ...) +{ + int len; + va_list ap; + + if (!strbuf_avail(sb)) + strbuf_grow(sb, 64); + va_start(ap, fmt); + len = vsnprintf(sb->buf + sb->len, sb->alloc - sb->len, fmt, ap); + va_end(ap); + if (len < 0) + die("your vsnprintf is broken"); + if (len > strbuf_avail(sb)) { + strbuf_grow(sb, len); + va_start(ap, fmt); + len = vsnprintf(sb->buf + sb->len, sb->alloc - sb->len, fmt, ap); + va_end(ap); + if (len > strbuf_avail(sb)) { + die("this should not happen, your snprintf is broken"); + } + } + strbuf_setlen(sb, sb->len + len); +} + +void strbuf_expand(struct strbuf *sb, const char *format, expand_fn_t fn, + void *context) +{ + for (;;) { + const char *percent; + size_t consumed; + + percent = strchrnul(format, '%'); + strbuf_add(sb, format, percent - format); + if (!*percent) + break; + format = percent + 1; + + consumed = fn(sb, format, context); + if (consumed) + format += consumed; + else + strbuf_addch(sb, '%'); + } +} + +size_t strbuf_expand_dict_cb(struct strbuf *sb, const char *placeholder, + void *context) +{ + struct strbuf_expand_dict_entry *e = context; + size_t len; + + for (; e->placeholder && (len = strlen(e->placeholder)); e++) { + if (!strncmp(placeholder, e->placeholder, len)) { + if (e->value) + strbuf_addstr(sb, e->value); + return len; + } + } + return 0; +} + +size_t strbuf_fread(struct strbuf *sb, size_t size, FILE *f) +{ + size_t res; + size_t oldalloc = sb->alloc; + + strbuf_grow(sb, size); + res = fread(sb->buf + sb->len, 1, size, f); + if (res > 0) + strbuf_setlen(sb, sb->len + res); + else if (res < 0 && oldalloc == 0) + strbuf_release(sb); + return res; +} + +ssize_t strbuf_read(struct strbuf *sb, int fd, size_t hint) +{ + size_t oldlen = sb->len; + size_t oldalloc = sb->alloc; + + strbuf_grow(sb, hint ? hint : 8192); + for (;;) { + ssize_t cnt; + + cnt = read(fd, sb->buf + sb->len, sb->alloc - sb->len - 1); + if (cnt < 0) { + if (oldalloc == 0) + strbuf_release(sb); + else + strbuf_setlen(sb, oldlen); + return -1; + } + if (!cnt) + break; + sb->len += cnt; + strbuf_grow(sb, 8192); + } + + sb->buf[sb->len] = '\0'; + return sb->len - oldlen; +} + +#define STRBUF_MAXLINK (2*PATH_MAX) + +int strbuf_readlink(struct strbuf *sb, const char *path, size_t hint) +{ + size_t oldalloc = sb->alloc; + + if (hint < 32) + hint = 32; + + while (hint < STRBUF_MAXLINK) { + int len; + + strbuf_grow(sb, hint); + len = readlink(path, sb->buf, hint); + if (len < 0) { + if (errno != ERANGE) + break; + } else if (len < hint) { + strbuf_setlen(sb, len); + return 0; + } + + /* .. the buffer was too small - try again */ + hint *= 2; + } + if (oldalloc == 0) + strbuf_release(sb); + return -1; +} + +int strbuf_getline(struct strbuf *sb, FILE *fp, int term) +{ + int ch; + + strbuf_grow(sb, 0); + if (feof(fp)) + return EOF; + + strbuf_reset(sb); + while ((ch = fgetc(fp)) != EOF) { + if (ch == term) + break; + strbuf_grow(sb, 1); + sb->buf[sb->len++] = ch; + } + if (ch == EOF && sb->len == 0) + return EOF; + + sb->buf[sb->len] = '\0'; + return 0; +} + +int strbuf_read_file(struct strbuf *sb, const char *path, size_t hint) +{ + int fd, len; + + fd = open(path, O_RDONLY); + if (fd < 0) + return -1; + len = strbuf_read(sb, fd, hint); + close(fd); + if (len < 0) + return -1; + + return len; +} diff --git a/tools/perf/util/strbuf.h b/tools/perf/util/strbuf.h new file mode 100644 index 00000000000..9ee908a3ec5 --- /dev/null +++ b/tools/perf/util/strbuf.h @@ -0,0 +1,137 @@ +#ifndef STRBUF_H +#define STRBUF_H + +/* + * Strbuf's can be use in many ways: as a byte array, or to store arbitrary + * long, overflow safe strings. + * + * Strbufs has some invariants that are very important to keep in mind: + * + * 1. the ->buf member is always malloc-ed, hence strbuf's can be used to + * build complex strings/buffers whose final size isn't easily known. + * + * It is NOT legal to copy the ->buf pointer away. + * `strbuf_detach' is the operation that detachs a buffer from its shell + * while keeping the shell valid wrt its invariants. + * + * 2. the ->buf member is a byte array that has at least ->len + 1 bytes + * allocated. The extra byte is used to store a '\0', allowing the ->buf + * member to be a valid C-string. Every strbuf function ensure this + * invariant is preserved. + * + * Note that it is OK to "play" with the buffer directly if you work it + * that way: + * + * strbuf_grow(sb, SOME_SIZE); + * ... Here, the memory array starting at sb->buf, and of length + * ... strbuf_avail(sb) is all yours, and you are sure that + * ... strbuf_avail(sb) is at least SOME_SIZE. + * strbuf_setlen(sb, sb->len + SOME_OTHER_SIZE); + * + * Of course, SOME_OTHER_SIZE must be smaller or equal to strbuf_avail(sb). + * + * Doing so is safe, though if it has to be done in many places, adding the + * missing API to the strbuf module is the way to go. + * + * XXX: do _not_ assume that the area that is yours is of size ->alloc - 1 + * even if it's true in the current implementation. Alloc is somehow a + * "private" member that should not be messed with. + */ + +#include <assert.h> + +extern char strbuf_slopbuf[]; +struct strbuf { + size_t alloc; + size_t len; + char *buf; +}; + +#define STRBUF_INIT { 0, 0, strbuf_slopbuf } + +/*----- strbuf life cycle -----*/ +extern void strbuf_init(struct strbuf *, size_t); +extern void strbuf_release(struct strbuf *); +extern char *strbuf_detach(struct strbuf *, size_t *); +extern void strbuf_attach(struct strbuf *, void *, size_t, size_t); +static inline void strbuf_swap(struct strbuf *a, struct strbuf *b) { + struct strbuf tmp = *a; + *a = *b; + *b = tmp; +} + +/*----- strbuf size related -----*/ +static inline size_t strbuf_avail(const struct strbuf *sb) { + return sb->alloc ? sb->alloc - sb->len - 1 : 0; +} + +extern void strbuf_grow(struct strbuf *, size_t); + +static inline void strbuf_setlen(struct strbuf *sb, size_t len) { + if (!sb->alloc) + strbuf_grow(sb, 0); + assert(len < sb->alloc); + sb->len = len; + sb->buf[len] = '\0'; +} +#define strbuf_reset(sb) strbuf_setlen(sb, 0) + +/*----- content related -----*/ +extern void strbuf_trim(struct strbuf *); +extern void strbuf_rtrim(struct strbuf *); +extern void strbuf_ltrim(struct strbuf *); +extern int strbuf_cmp(const struct strbuf *, const struct strbuf *); +extern void strbuf_tolower(struct strbuf *); + +extern struct strbuf **strbuf_split(const struct strbuf *, int delim); +extern void strbuf_list_free(struct strbuf **); + +/*----- add data in your buffer -----*/ +static inline void strbuf_addch(struct strbuf *sb, int c) { + strbuf_grow(sb, 1); + sb->buf[sb->len++] = c; + sb->buf[sb->len] = '\0'; +} + +extern void strbuf_insert(struct strbuf *, size_t pos, const void *, size_t); +extern void strbuf_remove(struct strbuf *, size_t pos, size_t len); + +/* splice pos..pos+len with given data */ +extern void strbuf_splice(struct strbuf *, size_t pos, size_t len, + const void *, size_t); + +extern void strbuf_add(struct strbuf *, const void *, size_t); +static inline void strbuf_addstr(struct strbuf *sb, const char *s) { + strbuf_add(sb, s, strlen(s)); +} +static inline void strbuf_addbuf(struct strbuf *sb, const struct strbuf *sb2) { + strbuf_add(sb, sb2->buf, sb2->len); +} +extern void strbuf_adddup(struct strbuf *sb, size_t pos, size_t len); + +typedef size_t (*expand_fn_t) (struct strbuf *sb, const char *placeholder, void *context); +extern void strbuf_expand(struct strbuf *sb, const char *format, expand_fn_t fn, void *context); +struct strbuf_expand_dict_entry { + const char *placeholder; + const char *value; +}; +extern size_t strbuf_expand_dict_cb(struct strbuf *sb, const char *placeholder, void *context); + +__attribute__((format(printf,2,3))) +extern void strbuf_addf(struct strbuf *sb, const char *fmt, ...); + +extern size_t strbuf_fread(struct strbuf *, size_t, FILE *); +/* XXX: if read fails, any partial read is undone */ +extern ssize_t strbuf_read(struct strbuf *, int fd, size_t hint); +extern int strbuf_read_file(struct strbuf *sb, const char *path, size_t hint); +extern int strbuf_readlink(struct strbuf *sb, const char *path, size_t hint); + +extern int strbuf_getline(struct strbuf *, FILE *, int); + +extern void stripspace(struct strbuf *buf, int skip_comments); +extern int launch_editor(const char *path, struct strbuf *buffer, const char *const *env); + +extern int strbuf_branchname(struct strbuf *sb, const char *name); +extern int strbuf_check_branch_ref(struct strbuf *sb, const char *name); + +#endif /* STRBUF_H */ diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c new file mode 100644 index 00000000000..ec33c0c7f4e --- /dev/null +++ b/tools/perf/util/string.c @@ -0,0 +1,34 @@ +#include "string.h" + +static int hex(char ch) +{ + if ((ch >= '0') && (ch <= '9')) + return ch - '0'; + if ((ch >= 'a') && (ch <= 'f')) + return ch - 'a' + 10; + if ((ch >= 'A') && (ch <= 'F')) + return ch - 'A' + 10; + return -1; +} + +/* + * While we find nice hex chars, build a long_val. + * Return number of chars processed. + */ +int hex2u64(const char *ptr, __u64 *long_val) +{ + const char *p = ptr; + *long_val = 0; + + while (*p) { + const int hex_val = hex(*p); + + if (hex_val < 0) + break; + + *long_val = (*long_val << 4) | hex_val; + p++; + } + + return p - ptr; +} diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h new file mode 100644 index 00000000000..72812c1c9a7 --- /dev/null +++ b/tools/perf/util/string.h @@ -0,0 +1,8 @@ +#ifndef _PERF_STRING_H_ +#define _PERF_STRING_H_ + +#include <linux/types.h> + +int hex2u64(const char *ptr, __u64 *val); + +#endif diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c new file mode 100644 index 00000000000..49a55f81371 --- /dev/null +++ b/tools/perf/util/symbol.c @@ -0,0 +1,641 @@ +#include "util.h" +#include "../perf.h" +#include "string.h" +#include "symbol.h" + +#include <libelf.h> +#include <gelf.h> +#include <elf.h> + +const char *sym_hist_filter; + +static struct symbol *symbol__new(__u64 start, __u64 len, + const char *name, unsigned int priv_size, + __u64 obj_start, int verbose) +{ + size_t namelen = strlen(name) + 1; + struct symbol *self = calloc(1, priv_size + sizeof(*self) + namelen); + + if (!self) + return NULL; + + if (verbose >= 2) + printf("new symbol: %016Lx [%08lx]: %s, hist: %p, obj_start: %p\n", + (__u64)start, (unsigned long)len, name, self->hist, (void *)(unsigned long)obj_start); + + self->obj_start= obj_start; + self->hist = NULL; + self->hist_sum = 0; + + if (sym_hist_filter && !strcmp(name, sym_hist_filter)) + self->hist = calloc(sizeof(__u64), len); + + if (priv_size) { + memset(self, 0, priv_size); + self = ((void *)self) + priv_size; + } + self->start = start; + self->end = start + len - 1; + memcpy(self->name, name, namelen); + + return self; +} + +static void symbol__delete(struct symbol *self, unsigned int priv_size) +{ + free(((void *)self) - priv_size); +} + +static size_t symbol__fprintf(struct symbol *self, FILE *fp) +{ + return fprintf(fp, " %llx-%llx %s\n", + self->start, self->end, self->name); +} + +struct dso *dso__new(const char *name, unsigned int sym_priv_size) +{ + struct dso *self = malloc(sizeof(*self) + strlen(name) + 1); + + if (self != NULL) { + strcpy(self->name, name); + self->syms = RB_ROOT; + self->sym_priv_size = sym_priv_size; + self->find_symbol = dso__find_symbol; + } + + return self; +} + +static void dso__delete_symbols(struct dso *self) +{ + struct symbol *pos; + struct rb_node *next = rb_first(&self->syms); + + while (next) { + pos = rb_entry(next, struct symbol, rb_node); + next = rb_next(&pos->rb_node); + rb_erase(&pos->rb_node, &self->syms); + symbol__delete(pos, self->sym_priv_size); + } +} + +void dso__delete(struct dso *self) +{ + dso__delete_symbols(self); + free(self); +} + +static void dso__insert_symbol(struct dso *self, struct symbol *sym) +{ + struct rb_node **p = &self->syms.rb_node; + struct rb_node *parent = NULL; + const __u64 ip = sym->start; + struct symbol *s; + + while (*p != NULL) { + parent = *p; + s = rb_entry(parent, struct symbol, rb_node); + if (ip < s->start) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + rb_link_node(&sym->rb_node, parent, p); + rb_insert_color(&sym->rb_node, &self->syms); +} + +struct symbol *dso__find_symbol(struct dso *self, __u64 ip) +{ + struct rb_node *n; + + if (self == NULL) + return NULL; + + n = self->syms.rb_node; + + while (n) { + struct symbol *s = rb_entry(n, struct symbol, rb_node); + + if (ip < s->start) + n = n->rb_left; + else if (ip > s->end) + n = n->rb_right; + else + return s; + } + + return NULL; +} + +size_t dso__fprintf(struct dso *self, FILE *fp) +{ + size_t ret = fprintf(fp, "dso: %s\n", self->name); + + struct rb_node *nd; + for (nd = rb_first(&self->syms); nd; nd = rb_next(nd)) { + struct symbol *pos = rb_entry(nd, struct symbol, rb_node); + ret += symbol__fprintf(pos, fp); + } + + return ret; +} + +static int dso__load_kallsyms(struct dso *self, symbol_filter_t filter, int verbose) +{ + struct rb_node *nd, *prevnd; + char *line = NULL; + size_t n; + FILE *file = fopen("/proc/kallsyms", "r"); + + if (file == NULL) + goto out_failure; + + while (!feof(file)) { + __u64 start; + struct symbol *sym; + int line_len, len; + char symbol_type; + + line_len = getline(&line, &n, file); + if (line_len < 0) + break; + + if (!line) + goto out_failure; + + line[--line_len] = '\0'; /* \n */ + + len = hex2u64(line, &start); + + len++; + if (len + 2 >= line_len) + continue; + + symbol_type = toupper(line[len]); + /* + * We're interested only in code ('T'ext) + */ + if (symbol_type != 'T' && symbol_type != 'W') + continue; + /* + * Well fix up the end later, when we have all sorted. + */ + sym = symbol__new(start, 0xdead, line + len + 2, + self->sym_priv_size, 0, verbose); + + if (sym == NULL) + goto out_delete_line; + + if (filter && filter(self, sym)) + symbol__delete(sym, self->sym_priv_size); + else + dso__insert_symbol(self, sym); + } + + /* + * Now that we have all sorted out, just set the ->end of all + * symbols + */ + prevnd = rb_first(&self->syms); + + if (prevnd == NULL) + goto out_delete_line; + + for (nd = rb_next(prevnd); nd; nd = rb_next(nd)) { + struct symbol *prev = rb_entry(prevnd, struct symbol, rb_node), + *curr = rb_entry(nd, struct symbol, rb_node); + + prev->end = curr->start - 1; + prevnd = nd; + } + + free(line); + fclose(file); + + return 0; + +out_delete_line: + free(line); +out_failure: + return -1; +} + +static int dso__load_perf_map(struct dso *self, symbol_filter_t filter, int verbose) +{ + char *line = NULL; + size_t n; + FILE *file; + int nr_syms = 0; + + file = fopen(self->name, "r"); + if (file == NULL) + goto out_failure; + + while (!feof(file)) { + __u64 start, size; + struct symbol *sym; + int line_len, len; + + line_len = getline(&line, &n, file); + if (line_len < 0) + break; + + if (!line) + goto out_failure; + + line[--line_len] = '\0'; /* \n */ + + len = hex2u64(line, &start); + + len++; + if (len + 2 >= line_len) + continue; + + len += hex2u64(line + len, &size); + + len++; + if (len + 2 >= line_len) + continue; + + sym = symbol__new(start, size, line + len, + self->sym_priv_size, start, verbose); + + if (sym == NULL) + goto out_delete_line; + + if (filter && filter(self, sym)) + symbol__delete(sym, self->sym_priv_size); + else { + dso__insert_symbol(self, sym); + nr_syms++; + } + } + + free(line); + fclose(file); + + return nr_syms; + +out_delete_line: + free(line); +out_failure: + return -1; +} + +/** + * elf_symtab__for_each_symbol - iterate thru all the symbols + * + * @self: struct elf_symtab instance to iterate + * @index: uint32_t index + * @sym: GElf_Sym iterator + */ +#define elf_symtab__for_each_symbol(syms, nr_syms, index, sym) \ + for (index = 0, gelf_getsym(syms, index, &sym);\ + index < nr_syms; \ + index++, gelf_getsym(syms, index, &sym)) + +static inline uint8_t elf_sym__type(const GElf_Sym *sym) +{ + return GELF_ST_TYPE(sym->st_info); +} + +static inline int elf_sym__is_function(const GElf_Sym *sym) +{ + return elf_sym__type(sym) == STT_FUNC && + sym->st_name != 0 && + sym->st_shndx != SHN_UNDEF && + sym->st_size != 0; +} + +static inline const char *elf_sym__name(const GElf_Sym *sym, + const Elf_Data *symstrs) +{ + return symstrs->d_buf + sym->st_name; +} + +static Elf_Scn *elf_section_by_name(Elf *elf, GElf_Ehdr *ep, + GElf_Shdr *shp, const char *name, + size_t *index) +{ + Elf_Scn *sec = NULL; + size_t cnt = 1; + + while ((sec = elf_nextscn(elf, sec)) != NULL) { + char *str; + + gelf_getshdr(sec, shp); + str = elf_strptr(elf, ep->e_shstrndx, shp->sh_name); + if (!strcmp(name, str)) { + if (index) + *index = cnt; + break; + } + ++cnt; + } + + return sec; +} + +#define elf_section__for_each_rel(reldata, pos, pos_mem, idx, nr_entries) \ + for (idx = 0, pos = gelf_getrel(reldata, 0, &pos_mem); \ + idx < nr_entries; \ + ++idx, pos = gelf_getrel(reldata, idx, &pos_mem)) + +#define elf_section__for_each_rela(reldata, pos, pos_mem, idx, nr_entries) \ + for (idx = 0, pos = gelf_getrela(reldata, 0, &pos_mem); \ + idx < nr_entries; \ + ++idx, pos = gelf_getrela(reldata, idx, &pos_mem)) + +static int dso__synthesize_plt_symbols(struct dso *self, Elf *elf, + GElf_Ehdr *ehdr, Elf_Scn *scn_dynsym, + GElf_Shdr *shdr_dynsym, + size_t dynsym_idx, int verbose) +{ + uint32_t nr_rel_entries, idx; + GElf_Sym sym; + __u64 plt_offset; + GElf_Shdr shdr_plt; + struct symbol *f; + GElf_Shdr shdr_rel_plt; + Elf_Data *reldata, *syms, *symstrs; + Elf_Scn *scn_plt_rel, *scn_symstrs; + char sympltname[1024]; + int nr = 0, symidx; + + scn_plt_rel = elf_section_by_name(elf, ehdr, &shdr_rel_plt, + ".rela.plt", NULL); + if (scn_plt_rel == NULL) { + scn_plt_rel = elf_section_by_name(elf, ehdr, &shdr_rel_plt, + ".rel.plt", NULL); + if (scn_plt_rel == NULL) + return 0; + } + + if (shdr_rel_plt.sh_link != dynsym_idx) + return 0; + + if (elf_section_by_name(elf, ehdr, &shdr_plt, ".plt", NULL) == NULL) + return 0; + + /* + * Fetch the relocation section to find the indexes to the GOT + * and the symbols in the .dynsym they refer to. + */ + reldata = elf_getdata(scn_plt_rel, NULL); + if (reldata == NULL) + return -1; + + syms = elf_getdata(scn_dynsym, NULL); + if (syms == NULL) + return -1; + + scn_symstrs = elf_getscn(elf, shdr_dynsym->sh_link); + if (scn_symstrs == NULL) + return -1; + + symstrs = elf_getdata(scn_symstrs, NULL); + if (symstrs == NULL) + return -1; + + nr_rel_entries = shdr_rel_plt.sh_size / shdr_rel_plt.sh_entsize; + plt_offset = shdr_plt.sh_offset; + + if (shdr_rel_plt.sh_type == SHT_RELA) { + GElf_Rela pos_mem, *pos; + + elf_section__for_each_rela(reldata, pos, pos_mem, idx, + nr_rel_entries) { + symidx = GELF_R_SYM(pos->r_info); + plt_offset += shdr_plt.sh_entsize; + gelf_getsym(syms, symidx, &sym); + snprintf(sympltname, sizeof(sympltname), + "%s@plt", elf_sym__name(&sym, symstrs)); + + f = symbol__new(plt_offset, shdr_plt.sh_entsize, + sympltname, self->sym_priv_size, 0, verbose); + if (!f) + return -1; + + dso__insert_symbol(self, f); + ++nr; + } + } else if (shdr_rel_plt.sh_type == SHT_REL) { + GElf_Rel pos_mem, *pos; + elf_section__for_each_rel(reldata, pos, pos_mem, idx, + nr_rel_entries) { + symidx = GELF_R_SYM(pos->r_info); + plt_offset += shdr_plt.sh_entsize; + gelf_getsym(syms, symidx, &sym); + snprintf(sympltname, sizeof(sympltname), + "%s@plt", elf_sym__name(&sym, symstrs)); + + f = symbol__new(plt_offset, shdr_plt.sh_entsize, + sympltname, self->sym_priv_size, 0, verbose); + if (!f) + return -1; + + dso__insert_symbol(self, f); + ++nr; + } + } else { + /* + * TODO: There are still one more shdr_rel_plt.sh_type + * I have to investigate, but probably should be ignored. + */ + } + + return nr; +} + +static int dso__load_sym(struct dso *self, int fd, const char *name, + symbol_filter_t filter, int verbose) +{ + Elf_Data *symstrs; + uint32_t nr_syms; + int err = -1; + uint32_t index; + GElf_Ehdr ehdr; + GElf_Shdr shdr; + Elf_Data *syms; + GElf_Sym sym; + Elf_Scn *sec, *sec_dynsym; + Elf *elf; + size_t dynsym_idx; + int nr = 0; + + elf = elf_begin(fd, ELF_C_READ_MMAP, NULL); + if (elf == NULL) { + if (verbose) + fprintf(stderr, "%s: cannot read %s ELF file.\n", + __func__, name); + goto out_close; + } + + if (gelf_getehdr(elf, &ehdr) == NULL) { + if (verbose) + fprintf(stderr, "%s: cannot get elf header.\n", __func__); + goto out_elf_end; + } + + /* + * We need to check if we have a .dynsym, so that we can handle the + * .plt, synthesizing its symbols, that aren't on the symtabs (be it + * .dynsym or .symtab) + */ + sec_dynsym = elf_section_by_name(elf, &ehdr, &shdr, + ".dynsym", &dynsym_idx); + if (sec_dynsym != NULL) { + nr = dso__synthesize_plt_symbols(self, elf, &ehdr, + sec_dynsym, &shdr, + dynsym_idx, verbose); + if (nr < 0) + goto out_elf_end; + } + + /* + * But if we have a full .symtab (that is a superset of .dynsym) we + * should add the symbols not in the .dynsyn + */ + sec = elf_section_by_name(elf, &ehdr, &shdr, ".symtab", NULL); + if (sec == NULL) { + if (sec_dynsym == NULL) + goto out_elf_end; + + sec = sec_dynsym; + gelf_getshdr(sec, &shdr); + } + + syms = elf_getdata(sec, NULL); + if (syms == NULL) + goto out_elf_end; + + sec = elf_getscn(elf, shdr.sh_link); + if (sec == NULL) + goto out_elf_end; + + symstrs = elf_getdata(sec, NULL); + if (symstrs == NULL) + goto out_elf_end; + + nr_syms = shdr.sh_size / shdr.sh_entsize; + + memset(&sym, 0, sizeof(sym)); + + elf_symtab__for_each_symbol(syms, nr_syms, index, sym) { + struct symbol *f; + __u64 obj_start; + + if (!elf_sym__is_function(&sym)) + continue; + + sec = elf_getscn(elf, sym.st_shndx); + if (!sec) + goto out_elf_end; + + gelf_getshdr(sec, &shdr); + obj_start = sym.st_value; + + sym.st_value -= shdr.sh_addr - shdr.sh_offset; + + f = symbol__new(sym.st_value, sym.st_size, + elf_sym__name(&sym, symstrs), + self->sym_priv_size, obj_start, verbose); + if (!f) + goto out_elf_end; + + if (filter && filter(self, f)) + symbol__delete(f, self->sym_priv_size); + else { + dso__insert_symbol(self, f); + nr++; + } + } + + err = nr; +out_elf_end: + elf_end(elf); +out_close: + return err; +} + +int dso__load(struct dso *self, symbol_filter_t filter, int verbose) +{ + int size = strlen(self->name) + sizeof("/usr/lib/debug%s.debug"); + char *name = malloc(size); + int variant = 0; + int ret = -1; + int fd; + + if (!name) + return -1; + + if (strncmp(self->name, "/tmp/perf-", 10) == 0) + return dso__load_perf_map(self, filter, verbose); + +more: + do { + switch (variant) { + case 0: /* Fedora */ + snprintf(name, size, "/usr/lib/debug%s.debug", self->name); + break; + case 1: /* Ubuntu */ + snprintf(name, size, "/usr/lib/debug%s", self->name); + break; + case 2: /* Sane people */ + snprintf(name, size, "%s", self->name); + break; + + default: + goto out; + } + variant++; + + fd = open(name, O_RDONLY); + } while (fd < 0); + + ret = dso__load_sym(self, fd, name, filter, verbose); + close(fd); + + /* + * Some people seem to have debuginfo files _WITHOUT_ debug info!?!? + */ + if (!ret) + goto more; + +out: + free(name); + return ret; +} + +static int dso__load_vmlinux(struct dso *self, const char *vmlinux, + symbol_filter_t filter, int verbose) +{ + int err, fd = open(vmlinux, O_RDONLY); + + if (fd < 0) + return -1; + + err = dso__load_sym(self, fd, vmlinux, filter, verbose); + close(fd); + + return err; +} + +int dso__load_kernel(struct dso *self, const char *vmlinux, + symbol_filter_t filter, int verbose) +{ + int err = -1; + + if (vmlinux) + err = dso__load_vmlinux(self, vmlinux, filter, verbose); + + if (err) + err = dso__load_kallsyms(self, filter, verbose); + + return err; +} + +void symbol__init(void) +{ + elf_version(EV_CURRENT); +} diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h new file mode 100644 index 00000000000..0d1292bd827 --- /dev/null +++ b/tools/perf/util/symbol.h @@ -0,0 +1,47 @@ +#ifndef _PERF_SYMBOL_ +#define _PERF_SYMBOL_ 1 + +#include <linux/types.h> +#include "list.h" +#include "rbtree.h" + +struct symbol { + struct rb_node rb_node; + __u64 start; + __u64 end; + __u64 obj_start; + __u64 hist_sum; + __u64 *hist; + char name[0]; +}; + +struct dso { + struct list_head node; + struct rb_root syms; + unsigned int sym_priv_size; + struct symbol *(*find_symbol)(struct dso *, __u64 ip); + char name[0]; +}; + +const char *sym_hist_filter; + +typedef int (*symbol_filter_t)(struct dso *self, struct symbol *sym); + +struct dso *dso__new(const char *name, unsigned int sym_priv_size); +void dso__delete(struct dso *self); + +static inline void *dso__sym_priv(struct dso *self, struct symbol *sym) +{ + return ((void *)sym) - self->sym_priv_size; +} + +struct symbol *dso__find_symbol(struct dso *self, __u64 ip); + +int dso__load_kernel(struct dso *self, const char *vmlinux, + symbol_filter_t filter, int verbose); +int dso__load(struct dso *self, symbol_filter_t filter, int verbose); + +size_t dso__fprintf(struct dso *self, FILE *fp); + +void symbol__init(void); +#endif /* _PERF_SYMBOL_ */ diff --git a/tools/perf/util/usage.c b/tools/perf/util/usage.c new file mode 100644 index 00000000000..e16bf9a707e --- /dev/null +++ b/tools/perf/util/usage.c @@ -0,0 +1,80 @@ +/* + * GIT - The information manager from hell + * + * Copyright (C) Linus Torvalds, 2005 + */ +#include "util.h" + +static void report(const char *prefix, const char *err, va_list params) +{ + char msg[1024]; + vsnprintf(msg, sizeof(msg), err, params); + fprintf(stderr, " %s%s\n", prefix, msg); +} + +static NORETURN void usage_builtin(const char *err) +{ + fprintf(stderr, "\n Usage: %s\n", err); + exit(129); +} + +static NORETURN void die_builtin(const char *err, va_list params) +{ + report(" Fatal: ", err, params); + exit(128); +} + +static void error_builtin(const char *err, va_list params) +{ + report(" Error: ", err, params); +} + +static void warn_builtin(const char *warn, va_list params) +{ + report(" Warning: ", warn, params); +} + +/* If we are in a dlopen()ed .so write to a global variable would segfault + * (ugh), so keep things static. */ +static void (*usage_routine)(const char *err) NORETURN = usage_builtin; +static void (*die_routine)(const char *err, va_list params) NORETURN = die_builtin; +static void (*error_routine)(const char *err, va_list params) = error_builtin; +static void (*warn_routine)(const char *err, va_list params) = warn_builtin; + +void set_die_routine(void (*routine)(const char *err, va_list params) NORETURN) +{ + die_routine = routine; +} + +void usage(const char *err) +{ + usage_routine(err); +} + +void die(const char *err, ...) +{ + va_list params; + + va_start(params, err); + die_routine(err, params); + va_end(params); +} + +int error(const char *err, ...) +{ + va_list params; + + va_start(params, err); + error_routine(err, params); + va_end(params); + return -1; +} + +void warning(const char *warn, ...) +{ + va_list params; + + va_start(params, warn); + warn_routine(warn, params); + va_end(params); +} diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h new file mode 100644 index 00000000000..76590a16c27 --- /dev/null +++ b/tools/perf/util/util.h @@ -0,0 +1,410 @@ +#ifndef GIT_COMPAT_UTIL_H +#define GIT_COMPAT_UTIL_H + +#define _FILE_OFFSET_BITS 64 + +#ifndef FLEX_ARRAY +/* + * See if our compiler is known to support flexible array members. + */ +#if defined(__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) +# define FLEX_ARRAY /* empty */ +#elif defined(__GNUC__) +# if (__GNUC__ >= 3) +# define FLEX_ARRAY /* empty */ +# else +# define FLEX_ARRAY 0 /* older GNU extension */ +# endif +#endif + +/* + * Otherwise, default to safer but a bit wasteful traditional style + */ +#ifndef FLEX_ARRAY +# define FLEX_ARRAY 1 +#endif +#endif + +#define ARRAY_SIZE(x) (sizeof(x)/sizeof(x[0])) + +#ifdef __GNUC__ +#define TYPEOF(x) (__typeof__(x)) +#else +#define TYPEOF(x) +#endif + +#define MSB(x, bits) ((x) & TYPEOF(x)(~0ULL << (sizeof(x) * 8 - (bits)))) +#define HAS_MULTI_BITS(i) ((i) & ((i) - 1)) /* checks if an integer has more than 1 bit set */ + +/* Approximation of the length of the decimal representation of this type. */ +#define decimal_length(x) ((int)(sizeof(x) * 2.56 + 0.5) + 1) + +#if !defined(__APPLE__) && !defined(__FreeBSD__) && !defined(__USLC__) && !defined(_M_UNIX) +#define _XOPEN_SOURCE 600 /* glibc2 and AIX 5.3L need 500, OpenBSD needs 600 for S_ISLNK() */ +#define _XOPEN_SOURCE_EXTENDED 1 /* AIX 5.3L needs this */ +#endif +#define _ALL_SOURCE 1 +#define _GNU_SOURCE 1 +#define _BSD_SOURCE 1 + +#include <unistd.h> +#include <stdio.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <stddef.h> +#include <stdlib.h> +#include <stdarg.h> +#include <string.h> +#include <errno.h> +#include <limits.h> +#include <sys/param.h> +#include <sys/types.h> +#include <dirent.h> +#include <sys/time.h> +#include <time.h> +#include <signal.h> +#include <fnmatch.h> +#include <assert.h> +#include <regex.h> +#include <utime.h> +#ifndef __MINGW32__ +#include <sys/wait.h> +#include <sys/poll.h> +#include <sys/socket.h> +#include <sys/ioctl.h> +#ifndef NO_SYS_SELECT_H +#include <sys/select.h> +#endif +#include <netinet/in.h> +#include <netinet/tcp.h> +#include <arpa/inet.h> +#include <netdb.h> +#include <pwd.h> +#include <inttypes.h> +#if defined(__CYGWIN__) +#undef _XOPEN_SOURCE +#include <grp.h> +#define _XOPEN_SOURCE 600 +#include "compat/cygwin.h" +#else +#undef _ALL_SOURCE /* AIX 5.3L defines a struct list with _ALL_SOURCE. */ +#include <grp.h> +#define _ALL_SOURCE 1 +#endif +#else /* __MINGW32__ */ +/* pull in Windows compatibility stuff */ +#include "compat/mingw.h" +#endif /* __MINGW32__ */ + +#ifndef NO_ICONV +#include <iconv.h> +#endif + +#ifndef NO_OPENSSL +#include <openssl/ssl.h> +#include <openssl/err.h> +#endif + +/* On most systems <limits.h> would have given us this, but + * not on some systems (e.g. GNU/Hurd). + */ +#ifndef PATH_MAX +#define PATH_MAX 4096 +#endif + +#ifndef PRIuMAX +#define PRIuMAX "llu" +#endif + +#ifndef PRIu32 +#define PRIu32 "u" +#endif + +#ifndef PRIx32 +#define PRIx32 "x" +#endif + +#ifndef PATH_SEP +#define PATH_SEP ':' +#endif + +#ifndef STRIP_EXTENSION +#define STRIP_EXTENSION "" +#endif + +#ifndef has_dos_drive_prefix +#define has_dos_drive_prefix(path) 0 +#endif + +#ifndef is_dir_sep +#define is_dir_sep(c) ((c) == '/') +#endif + +#ifdef __GNUC__ +#define NORETURN __attribute__((__noreturn__)) +#else +#define NORETURN +#ifndef __attribute__ +#define __attribute__(x) +#endif +#endif + +/* General helper functions */ +extern void usage(const char *err) NORETURN; +extern void die(const char *err, ...) NORETURN __attribute__((format (printf, 1, 2))); +extern int error(const char *err, ...) __attribute__((format (printf, 1, 2))); +extern void warning(const char *err, ...) __attribute__((format (printf, 1, 2))); + +extern void set_die_routine(void (*routine)(const char *err, va_list params) NORETURN); + +extern int prefixcmp(const char *str, const char *prefix); +extern time_t tm_to_time_t(const struct tm *tm); + +static inline const char *skip_prefix(const char *str, const char *prefix) +{ + size_t len = strlen(prefix); + return strncmp(str, prefix, len) ? NULL : str + len; +} + +#if defined(NO_MMAP) || defined(USE_WIN32_MMAP) + +#ifndef PROT_READ +#define PROT_READ 1 +#define PROT_WRITE 2 +#define MAP_PRIVATE 1 +#define MAP_FAILED ((void*)-1) +#endif + +#define mmap git_mmap +#define munmap git_munmap +extern void *git_mmap(void *start, size_t length, int prot, int flags, int fd, off_t offset); +extern int git_munmap(void *start, size_t length); + +#else /* NO_MMAP || USE_WIN32_MMAP */ + +#include <sys/mman.h> + +#endif /* NO_MMAP || USE_WIN32_MMAP */ + +#ifdef NO_MMAP + +/* This value must be multiple of (pagesize * 2) */ +#define DEFAULT_PACKED_GIT_WINDOW_SIZE (1 * 1024 * 1024) + +#else /* NO_MMAP */ + +/* This value must be multiple of (pagesize * 2) */ +#define DEFAULT_PACKED_GIT_WINDOW_SIZE \ + (sizeof(void*) >= 8 \ + ? 1 * 1024 * 1024 * 1024 \ + : 32 * 1024 * 1024) + +#endif /* NO_MMAP */ + +#ifdef NO_ST_BLOCKS_IN_STRUCT_STAT +#define on_disk_bytes(st) ((st).st_size) +#else +#define on_disk_bytes(st) ((st).st_blocks * 512) +#endif + +#define DEFAULT_PACKED_GIT_LIMIT \ + ((1024L * 1024L) * (sizeof(void*) >= 8 ? 8192 : 256)) + +#ifdef NO_PREAD +#define pread git_pread +extern ssize_t git_pread(int fd, void *buf, size_t count, off_t offset); +#endif +/* + * Forward decl that will remind us if its twin in cache.h changes. + * This function is used in compat/pread.c. But we can't include + * cache.h there. + */ +extern ssize_t read_in_full(int fd, void *buf, size_t count); + +#ifdef NO_SETENV +#define setenv gitsetenv +extern int gitsetenv(const char *, const char *, int); +#endif + +#ifdef NO_MKDTEMP +#define mkdtemp gitmkdtemp +extern char *gitmkdtemp(char *); +#endif + +#ifdef NO_UNSETENV +#define unsetenv gitunsetenv +extern void gitunsetenv(const char *); +#endif + +#ifdef NO_STRCASESTR +#define strcasestr gitstrcasestr +extern char *gitstrcasestr(const char *haystack, const char *needle); +#endif + +#ifdef NO_STRLCPY +#define strlcpy gitstrlcpy +extern size_t gitstrlcpy(char *, const char *, size_t); +#endif + +#ifdef NO_STRTOUMAX +#define strtoumax gitstrtoumax +extern uintmax_t gitstrtoumax(const char *, char **, int); +#endif + +#ifdef NO_HSTRERROR +#define hstrerror githstrerror +extern const char *githstrerror(int herror); +#endif + +#ifdef NO_MEMMEM +#define memmem gitmemmem +void *gitmemmem(const void *haystack, size_t haystacklen, + const void *needle, size_t needlelen); +#endif + +#ifdef FREAD_READS_DIRECTORIES +#ifdef fopen +#undef fopen +#endif +#define fopen(a,b) git_fopen(a,b) +extern FILE *git_fopen(const char*, const char*); +#endif + +#ifdef SNPRINTF_RETURNS_BOGUS +#define snprintf git_snprintf +extern int git_snprintf(char *str, size_t maxsize, + const char *format, ...); +#define vsnprintf git_vsnprintf +extern int git_vsnprintf(char *str, size_t maxsize, + const char *format, va_list ap); +#endif + +#ifdef __GLIBC_PREREQ +#if __GLIBC_PREREQ(2, 1) +#define HAVE_STRCHRNUL +#endif +#endif + +#ifndef HAVE_STRCHRNUL +#define strchrnul gitstrchrnul +static inline char *gitstrchrnul(const char *s, int c) +{ + while (*s && *s != c) + s++; + return (char *)s; +} +#endif + +/* + * Wrappers: + */ +extern char *xstrdup(const char *str); +extern void *xmalloc(size_t size); +extern void *xmemdupz(const void *data, size_t len); +extern char *xstrndup(const char *str, size_t len); +extern void *xrealloc(void *ptr, size_t size); +extern void *xcalloc(size_t nmemb, size_t size); +extern void *xmmap(void *start, size_t length, int prot, int flags, int fd, off_t offset); +extern ssize_t xread(int fd, void *buf, size_t len); +extern ssize_t xwrite(int fd, const void *buf, size_t len); +extern int xdup(int fd); +extern FILE *xfdopen(int fd, const char *mode); +extern int xmkstemp(char *template); + +static inline size_t xsize_t(off_t len) +{ + return (size_t)len; +} + +static inline int has_extension(const char *filename, const char *ext) +{ + size_t len = strlen(filename); + size_t extlen = strlen(ext); + return len > extlen && !memcmp(filename + len - extlen, ext, extlen); +} + +/* Sane ctype - no locale, and works with signed chars */ +#undef isascii +#undef isspace +#undef isdigit +#undef isalpha +#undef isalnum +#undef tolower +#undef toupper +extern unsigned char sane_ctype[256]; +#define GIT_SPACE 0x01 +#define GIT_DIGIT 0x02 +#define GIT_ALPHA 0x04 +#define GIT_GLOB_SPECIAL 0x08 +#define GIT_REGEX_SPECIAL 0x10 +#define sane_istest(x,mask) ((sane_ctype[(unsigned char)(x)] & (mask)) != 0) +#define isascii(x) (((x) & ~0x7f) == 0) +#define isspace(x) sane_istest(x,GIT_SPACE) +#define isdigit(x) sane_istest(x,GIT_DIGIT) +#define isalpha(x) sane_istest(x,GIT_ALPHA) +#define isalnum(x) sane_istest(x,GIT_ALPHA | GIT_DIGIT) +#define is_glob_special(x) sane_istest(x,GIT_GLOB_SPECIAL) +#define is_regex_special(x) sane_istest(x,GIT_GLOB_SPECIAL | GIT_REGEX_SPECIAL) +#define tolower(x) sane_case((unsigned char)(x), 0x20) +#define toupper(x) sane_case((unsigned char)(x), 0) + +static inline int sane_case(int x, int high) +{ + if (sane_istest(x, GIT_ALPHA)) + x = (x & ~0x20) | high; + return x; +} + +static inline int strtoul_ui(char const *s, int base, unsigned int *result) +{ + unsigned long ul; + char *p; + + errno = 0; + ul = strtoul(s, &p, base); + if (errno || *p || p == s || (unsigned int) ul != ul) + return -1; + *result = ul; + return 0; +} + +static inline int strtol_i(char const *s, int base, int *result) +{ + long ul; + char *p; + + errno = 0; + ul = strtol(s, &p, base); + if (errno || *p || p == s || (int) ul != ul) + return -1; + *result = ul; + return 0; +} + +#ifdef INTERNAL_QSORT +void git_qsort(void *base, size_t nmemb, size_t size, + int(*compar)(const void *, const void *)); +#define qsort git_qsort +#endif + +#ifndef DIR_HAS_BSD_GROUP_SEMANTICS +# define FORCE_DIR_SET_GID S_ISGID +#else +# define FORCE_DIR_SET_GID 0 +#endif + +#ifdef NO_NSEC +#undef USE_NSEC +#define ST_CTIME_NSEC(st) 0 +#define ST_MTIME_NSEC(st) 0 +#else +#ifdef USE_ST_TIMESPEC +#define ST_CTIME_NSEC(st) ((unsigned int)((st).st_ctimespec.tv_nsec)) +#define ST_MTIME_NSEC(st) ((unsigned int)((st).st_mtimespec.tv_nsec)) +#else +#define ST_CTIME_NSEC(st) ((unsigned int)((st).st_ctim.tv_nsec)) +#define ST_MTIME_NSEC(st) ((unsigned int)((st).st_mtim.tv_nsec)) +#endif +#endif + +#endif diff --git a/tools/perf/util/wrapper.c b/tools/perf/util/wrapper.c new file mode 100644 index 00000000000..6350d65f6d9 --- /dev/null +++ b/tools/perf/util/wrapper.c @@ -0,0 +1,206 @@ +/* + * Various trivial helper wrappers around standard functions + */ +#include "cache.h" + +/* + * There's no pack memory to release - but stay close to the Git + * version so wrap this away: + */ +static inline void release_pack_memory(size_t size, int flag) +{ +} + +char *xstrdup(const char *str) +{ + char *ret = strdup(str); + if (!ret) { + release_pack_memory(strlen(str) + 1, -1); + ret = strdup(str); + if (!ret) + die("Out of memory, strdup failed"); + } + return ret; +} + +void *xmalloc(size_t size) +{ + void *ret = malloc(size); + if (!ret && !size) + ret = malloc(1); + if (!ret) { + release_pack_memory(size, -1); + ret = malloc(size); + if (!ret && !size) + ret = malloc(1); + if (!ret) + die("Out of memory, malloc failed"); + } +#ifdef XMALLOC_POISON + memset(ret, 0xA5, size); +#endif + return ret; +} + +/* + * xmemdupz() allocates (len + 1) bytes of memory, duplicates "len" bytes of + * "data" to the allocated memory, zero terminates the allocated memory, + * and returns a pointer to the allocated memory. If the allocation fails, + * the program dies. + */ +void *xmemdupz(const void *data, size_t len) +{ + char *p = xmalloc(len + 1); + memcpy(p, data, len); + p[len] = '\0'; + return p; +} + +char *xstrndup(const char *str, size_t len) +{ + char *p = memchr(str, '\0', len); + return xmemdupz(str, p ? p - str : len); +} + +void *xrealloc(void *ptr, size_t size) +{ + void *ret = realloc(ptr, size); + if (!ret && !size) + ret = realloc(ptr, 1); + if (!ret) { + release_pack_memory(size, -1); + ret = realloc(ptr, size); + if (!ret && !size) + ret = realloc(ptr, 1); + if (!ret) + die("Out of memory, realloc failed"); + } + return ret; +} + +void *xcalloc(size_t nmemb, size_t size) +{ + void *ret = calloc(nmemb, size); + if (!ret && (!nmemb || !size)) + ret = calloc(1, 1); + if (!ret) { + release_pack_memory(nmemb * size, -1); + ret = calloc(nmemb, size); + if (!ret && (!nmemb || !size)) + ret = calloc(1, 1); + if (!ret) + die("Out of memory, calloc failed"); + } + return ret; +} + +void *xmmap(void *start, size_t length, + int prot, int flags, int fd, off_t offset) +{ + void *ret = mmap(start, length, prot, flags, fd, offset); + if (ret == MAP_FAILED) { + if (!length) + return NULL; + release_pack_memory(length, fd); + ret = mmap(start, length, prot, flags, fd, offset); + if (ret == MAP_FAILED) + die("Out of memory? mmap failed: %s", strerror(errno)); + } + return ret; +} + +/* + * xread() is the same a read(), but it automatically restarts read() + * operations with a recoverable error (EAGAIN and EINTR). xread() + * DOES NOT GUARANTEE that "len" bytes is read even if the data is available. + */ +ssize_t xread(int fd, void *buf, size_t len) +{ + ssize_t nr; + while (1) { + nr = read(fd, buf, len); + if ((nr < 0) && (errno == EAGAIN || errno == EINTR)) + continue; + return nr; + } +} + +/* + * xwrite() is the same a write(), but it automatically restarts write() + * operations with a recoverable error (EAGAIN and EINTR). xwrite() DOES NOT + * GUARANTEE that "len" bytes is written even if the operation is successful. + */ +ssize_t xwrite(int fd, const void *buf, size_t len) +{ + ssize_t nr; + while (1) { + nr = write(fd, buf, len); + if ((nr < 0) && (errno == EAGAIN || errno == EINTR)) + continue; + return nr; + } +} + +ssize_t read_in_full(int fd, void *buf, size_t count) +{ + char *p = buf; + ssize_t total = 0; + + while (count > 0) { + ssize_t loaded = xread(fd, p, count); + if (loaded <= 0) + return total ? total : loaded; + count -= loaded; + p += loaded; + total += loaded; + } + + return total; +} + +ssize_t write_in_full(int fd, const void *buf, size_t count) +{ + const char *p = buf; + ssize_t total = 0; + + while (count > 0) { + ssize_t written = xwrite(fd, p, count); + if (written < 0) + return -1; + if (!written) { + errno = ENOSPC; + return -1; + } + count -= written; + p += written; + total += written; + } + + return total; +} + +int xdup(int fd) +{ + int ret = dup(fd); + if (ret < 0) + die("dup failed: %s", strerror(errno)); + return ret; +} + +FILE *xfdopen(int fd, const char *mode) +{ + FILE *stream = fdopen(fd, mode); + if (stream == NULL) + die("Out of memory? fdopen failed: %s", strerror(errno)); + return stream; +} + +int xmkstemp(char *template) +{ + int fd; + + fd = mkstemp(template); + if (fd < 0) + die("Unable to create temporary file: %s", strerror(errno)); + return fd; +} |