diff options
Diffstat (limited to 'arch/x86/kernel')
74 files changed, 1772 insertions, 1293 deletions
diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index b67efd1cf59..bf04201b657 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -24,6 +24,10 @@ CFLAGS_vsyscall_64.o := $(PROFILING) -g0 $(nostackp) CFLAGS_hpet.o := $(nostackp) CFLAGS_tsc.o := $(nostackp) CFLAGS_paravirt.o := $(nostackp) +GCOV_PROFILE_vsyscall_64.o := n +GCOV_PROFILE_hpet.o := n +GCOV_PROFILE_tsc.o := n +GCOV_PROFILE_paravirt.o := n obj-y := process_$(BITS).o signal.o entry_$(BITS).o obj-y += traps.o irq.o irq_$(BITS).o dumpstack_$(BITS).o diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 631086159c5..6b8ca3a0285 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -44,11 +44,7 @@ static int __initdata acpi_force = 0; u32 acpi_rsdt_forced; -#ifdef CONFIG_ACPI -int acpi_disabled = 0; -#else -int acpi_disabled = 1; -#endif +int acpi_disabled; EXPORT_SYMBOL(acpi_disabled); #ifdef CONFIG_X86_64 @@ -122,72 +118,6 @@ void __init __acpi_unmap_table(char *map, unsigned long size) early_iounmap(map, size); } -#ifdef CONFIG_PCI_MMCONFIG - -static int acpi_mcfg_64bit_base_addr __initdata = FALSE; - -/* The physical address of the MMCONFIG aperture. Set from ACPI tables. */ -struct acpi_mcfg_allocation *pci_mmcfg_config; -int pci_mmcfg_config_num; - -static int __init acpi_mcfg_oem_check(struct acpi_table_mcfg *mcfg) -{ - if (!strcmp(mcfg->header.oem_id, "SGI")) - acpi_mcfg_64bit_base_addr = TRUE; - - return 0; -} - -int __init acpi_parse_mcfg(struct acpi_table_header *header) -{ - struct acpi_table_mcfg *mcfg; - unsigned long i; - int config_size; - - if (!header) - return -EINVAL; - - mcfg = (struct acpi_table_mcfg *)header; - - /* how many config structures do we have */ - pci_mmcfg_config_num = 0; - i = header->length - sizeof(struct acpi_table_mcfg); - while (i >= sizeof(struct acpi_mcfg_allocation)) { - ++pci_mmcfg_config_num; - i -= sizeof(struct acpi_mcfg_allocation); - }; - if (pci_mmcfg_config_num == 0) { - printk(KERN_ERR PREFIX "MMCONFIG has no entries\n"); - return -ENODEV; - } - - config_size = pci_mmcfg_config_num * sizeof(*pci_mmcfg_config); - pci_mmcfg_config = kmalloc(config_size, GFP_KERNEL); - if (!pci_mmcfg_config) { - printk(KERN_WARNING PREFIX - "No memory for MCFG config tables\n"); - return -ENOMEM; - } - - memcpy(pci_mmcfg_config, &mcfg[1], config_size); - - acpi_mcfg_oem_check(mcfg); - - for (i = 0; i < pci_mmcfg_config_num; ++i) { - if ((pci_mmcfg_config[i].address > 0xFFFFFFFF) && - !acpi_mcfg_64bit_base_addr) { - printk(KERN_ERR PREFIX - "MMCONFIG not in low 4GB of memory\n"); - kfree(pci_mmcfg_config); - pci_mmcfg_config_num = 0; - return -ENODEV; - } - } - - return 0; -} -#endif /* CONFIG_PCI_MMCONFIG */ - #ifdef CONFIG_X86_LOCAL_APIC static int __init acpi_parse_madt(struct acpi_table_header *table) { @@ -1519,14 +1449,6 @@ static struct dmi_system_id __initdata acpi_dmi_table[] = { }, { .callback = force_acpi_ht, - .ident = "ASUS P4B266", - .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), - DMI_MATCH(DMI_BOARD_NAME, "P4B266"), - }, - }, - { - .callback = force_acpi_ht, .ident = "ASUS P2B-DS", .matches = { DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), diff --git a/arch/x86/kernel/acpi/cstate.c b/arch/x86/kernel/acpi/cstate.c index bbbe4bbb6f3..8c44c232efc 100644 --- a/arch/x86/kernel/acpi/cstate.c +++ b/arch/x86/kernel/acpi/cstate.c @@ -34,12 +34,22 @@ void acpi_processor_power_init_bm_check(struct acpi_processor_flags *flags, flags->bm_check = 1; else if (c->x86_vendor == X86_VENDOR_INTEL) { /* - * Today all CPUs that support C3 share cache. - * TBD: This needs to look at cache shared map, once - * multi-core detection patch makes to the base. + * Today all MP CPUs that support C3 share cache. + * And caches should not be flushed by software while + * entering C3 type state. */ flags->bm_check = 1; } + + /* + * On all recent Intel platforms, ARB_DISABLE is a nop. + * So, set bm_control to zero to indicate that ARB_DISABLE + * is not required while entering C3 type state on + * P4, Core and beyond CPUs + */ + if (c->x86_vendor == X86_VENDOR_INTEL && + (c->x86 > 0x6 || (c->x86 == 6 && c->x86_model >= 14))) + flags->bm_control = 0; } EXPORT_SYMBOL(acpi_processor_power_init_bm_check); diff --git a/arch/x86/kernel/acpi/processor.c b/arch/x86/kernel/acpi/processor.c index 7c074eec39f..d296f4a195c 100644 --- a/arch/x86/kernel/acpi/processor.c +++ b/arch/x86/kernel/acpi/processor.c @@ -72,6 +72,7 @@ static void init_intel_pdc(struct acpi_processor *pr, struct cpuinfo_x86 *c) return; } + /* Initialize _PDC data based on the CPU vendor */ void arch_acpi_processor_init_pdc(struct acpi_processor *pr) { @@ -85,3 +86,15 @@ void arch_acpi_processor_init_pdc(struct acpi_processor *pr) } EXPORT_SYMBOL(arch_acpi_processor_init_pdc); + +void arch_acpi_processor_cleanup_pdc(struct acpi_processor *pr) +{ + if (pr->pdc) { + kfree(pr->pdc->pointer->buffer.pointer); + kfree(pr->pdc->pointer); + kfree(pr->pdc); + pr->pdc = NULL; + } +} + +EXPORT_SYMBOL(arch_acpi_processor_cleanup_pdc); diff --git a/arch/x86/kernel/acpi/realmode/Makefile b/arch/x86/kernel/acpi/realmode/Makefile index 167bc16ce0e..6a564ac67ef 100644 --- a/arch/x86/kernel/acpi/realmode/Makefile +++ b/arch/x86/kernel/acpi/realmode/Makefile @@ -42,6 +42,7 @@ KBUILD_CFLAGS := $(LINUXINCLUDE) -g -Os -D_SETUP -D_WAKEUP -D__KERNEL__ \ $(call cc-option, -mpreferred-stack-boundary=2) KBUILD_CFLAGS += $(call cc-option, -m32) KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ +GCOV_PROFILE := n WAKEUP_OBJS = $(addprefix $(obj)/,$(wakeup-y)) diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 1c60554537c..6c99f503780 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -434,6 +434,16 @@ static void iommu_flush_tlb(struct amd_iommu *iommu, u16 domid) iommu_queue_inv_iommu_pages(iommu, address, domid, 0, 1); } +/* Flush the whole IO/TLB for a given protection domain - including PDE */ +static void iommu_flush_tlb_pde(struct amd_iommu *iommu, u16 domid) +{ + u64 address = CMD_INV_IOMMU_ALL_PAGES_ADDRESS; + + INC_STATS_COUNTER(domain_flush_single); + + iommu_queue_inv_iommu_pages(iommu, address, domid, 1, 1); +} + /* * This function is used to flush the IO/TLB for a given protection domain * on every IOMMU in the system @@ -1078,7 +1088,13 @@ static void attach_device(struct amd_iommu *iommu, amd_iommu_pd_table[devid] = domain; write_unlock_irqrestore(&amd_iommu_devtable_lock, flags); + /* + * We might boot into a crash-kernel here. The crashed kernel + * left the caches in the IOMMU dirty. So we have to flush + * here to evict all dirty stuff. + */ iommu_queue_inv_dev_entry(iommu, devid); + iommu_flush_tlb_pde(iommu, domain->id); } /* @@ -1176,7 +1192,7 @@ out: return 0; } -struct notifier_block device_nb = { +static struct notifier_block device_nb = { .notifier_call = device_change_notifier, }; @@ -1747,7 +1763,7 @@ static void *alloc_coherent(struct device *dev, size_t size, flag |= __GFP_ZERO; virt_addr = (void *)__get_free_pages(flag, get_order(size)); if (!virt_addr) - return 0; + return NULL; paddr = virt_to_phys(virt_addr); diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c index 238989ec077..c1b17e97252 100644 --- a/arch/x86/kernel/amd_iommu_init.c +++ b/arch/x86/kernel/amd_iommu_init.c @@ -260,6 +260,14 @@ static void iommu_enable(struct amd_iommu *iommu) static void iommu_disable(struct amd_iommu *iommu) { + /* Disable command buffer */ + iommu_feature_disable(iommu, CONTROL_CMDBUF_EN); + + /* Disable event logging and event interrupts */ + iommu_feature_disable(iommu, CONTROL_EVT_INT_EN); + iommu_feature_disable(iommu, CONTROL_EVT_LOG_EN); + + /* Disable IOMMU hardware itself */ iommu_feature_disable(iommu, CONTROL_IOMMU_EN); } @@ -464,6 +472,8 @@ static u8 * __init alloc_event_buffer(struct amd_iommu *iommu) if (iommu->evt_buf == NULL) return NULL; + iommu->evt_buf_size = EVT_BUFFER_SIZE; + return iommu->evt_buf; } @@ -478,6 +488,10 @@ static void iommu_enable_event_buffer(struct amd_iommu *iommu) memcpy_toio(iommu->mmio_base + MMIO_EVT_BUF_OFFSET, &entry, sizeof(entry)); + /* set head and tail to zero manually */ + writel(0x00, iommu->mmio_base + MMIO_EVT_HEAD_OFFSET); + writel(0x00, iommu->mmio_base + MMIO_EVT_TAIL_OFFSET); + iommu_feature_enable(iommu, CONTROL_EVT_LOG_EN); } @@ -679,6 +693,7 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu, 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; @@ -737,11 +752,13 @@ static void __init init_iommu_from_acpi(struct amd_iommu *iommu, devid = e->devid; for (dev_i = devid_start; dev_i <= devid; ++dev_i) { - if (alias) + if (alias) { amd_iommu_alias_table[dev_i] = devid_to; - set_dev_entry_from_acpi(iommu, - amd_iommu_alias_table[dev_i], - flags, ext_flags); + set_dev_entry_from_acpi(iommu, + devid_to, flags, ext_flags); + } + set_dev_entry_from_acpi(iommu, dev_i, + flags, ext_flags); } break; default: @@ -1042,6 +1059,7 @@ static void enable_iommus(void) struct amd_iommu *iommu; for_each_iommu(iommu) { + iommu_disable(iommu); iommu_set_device_table(iommu); iommu_enable_command_buffer(iommu); iommu_enable_event_buffer(iommu); @@ -1066,12 +1084,6 @@ static void disable_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(); @@ -1079,8 +1091,8 @@ static int amd_iommu_resume(struct sys_device *dev) * 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(); + amd_iommu_flush_all_domains(); return 0; } @@ -1273,6 +1285,11 @@ free: goto out; } +void amd_iommu_shutdown(void) +{ + disable_iommus(); +} + /**************************************************************************** * * Early detect code. This code runs at IOMMU detection time in the DMA diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 8c7c042ecad..0a1c2830ec6 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -140,7 +140,6 @@ int x2apic_mode; #ifdef CONFIG_X86_X2APIC /* x2apic enabled before OS handover */ static int x2apic_preenabled; -static int disable_x2apic; static __init int setup_nox2apic(char *str) { if (x2apic_enabled()) { @@ -149,7 +148,6 @@ static __init int setup_nox2apic(char *str) return 0; } - disable_x2apic = 1; setup_clear_cpu_cap(X86_FEATURE_X2APIC); return 0; } diff --git a/arch/x86/kernel/apic/es7000_32.c b/arch/x86/kernel/apic/es7000_32.c index 69328ac8de9..8952a589028 100644 --- a/arch/x86/kernel/apic/es7000_32.c +++ b/arch/x86/kernel/apic/es7000_32.c @@ -652,7 +652,8 @@ static int es7000_mps_oem_check_cluster(struct mpc_table *mpc, char *oem, return ret && es7000_apic_is_cluster(); } -struct apic apic_es7000_cluster = { +/* We've been warned by a false positive warning.Use __refdata to keep calm. */ +struct apic __refdata apic_es7000_cluster = { .name = "es7000", .probe = probe_es7000, diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index ef8d9290c7e..d2ed6c5ddc8 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -462,7 +462,8 @@ static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin) static void __ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) { - union entry_union eu; + union entry_union eu = {{0, 0}}; + eu.entry = e; io_apic_write(apic, 0x11 + 2*pin, eu.w2); io_apic_write(apic, 0x10 + 2*pin, eu.w1); @@ -1413,6 +1414,9 @@ int setup_ioapic_entry(int apic_id, int irq, irte.vector = vector; irte.dest_id = IRTE_DEST(destination); + /* Set source-id of interrupt request */ + set_ioapic_sid(&irte, apic_id); + modify_irte(irq, &irte); ir_entry->index2 = (index >> 15) & 0x1; @@ -1712,25 +1716,19 @@ __apicdebuginit(void) print_IO_APIC(void) return; } -__apicdebuginit(void) print_APIC_bitfield(int base) +__apicdebuginit(void) print_APIC_field(int base) { - unsigned int v; - int i, j; + int i; if (apic_verbosity == APIC_QUIET) return; - printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG); - for (i = 0; i < 8; i++) { - v = apic_read(base + i*0x10); - for (j = 0; j < 32; j++) { - if (v & (1<<j)) - printk("1"); - else - printk("0"); - } - printk("\n"); - } + printk(KERN_DEBUG); + + for (i = 0; i < 8; i++) + printk(KERN_CONT "%08x", apic_read(base + i*0x10)); + + printk(KERN_CONT "\n"); } __apicdebuginit(void) print_local_APIC(void *dummy) @@ -1741,7 +1739,7 @@ __apicdebuginit(void) print_local_APIC(void *dummy) if (apic_verbosity == APIC_QUIET) return; - printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n", + printk(KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n", smp_processor_id(), hard_smp_processor_id()); v = apic_read(APIC_ID); printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, read_apic_id()); @@ -1782,11 +1780,11 @@ __apicdebuginit(void) print_local_APIC(void *dummy) printk(KERN_DEBUG "... APIC SPIV: %08x\n", v); printk(KERN_DEBUG "... APIC ISR field:\n"); - print_APIC_bitfield(APIC_ISR); + print_APIC_field(APIC_ISR); printk(KERN_DEBUG "... APIC TMR field:\n"); - print_APIC_bitfield(APIC_TMR); + print_APIC_field(APIC_TMR); printk(KERN_DEBUG "... APIC IRR field:\n"); - print_APIC_bitfield(APIC_IRR); + print_APIC_field(APIC_IRR); if (APIC_INTEGRATED(ver)) { /* !82489DX */ if (maxlvt > 3) /* Due to the Pentium erratum 3AP. */ @@ -2003,7 +2001,9 @@ void disable_IO_APIC(void) /* * Use virtual wire A mode when interrupt remapping is enabled. */ - disconnect_bsp_APIC(!intr_remapping_enabled && ioapic_i8259.pin != -1); + if (cpu_has_apic) + disconnect_bsp_APIC(!intr_remapping_enabled && + ioapic_i8259.pin != -1); } #ifdef CONFIG_X86_32 @@ -3287,6 +3287,9 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms irte.vector = cfg->vector; irte.dest_id = IRTE_DEST(dest); + /* Set source-id of interrupt request */ + set_msi_sid(&irte, pdev); + modify_irte(irq, &irte); msg->address_hi = MSI_ADDR_BASE_HI; @@ -3567,7 +3570,7 @@ static int dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask) #endif /* CONFIG_SMP */ -struct irq_chip dmar_msi_type = { +static struct irq_chip dmar_msi_type = { .name = "DMAR_MSI", .unmask = dmar_msi_unmask, .mask = dmar_msi_mask, @@ -3790,6 +3793,9 @@ int arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade, mmr_pnode = uv_blade_to_pnode(mmr_blade); uv_write_global_mmr64(mmr_pnode, mmr_offset, mmr_value); + if (cfg->move_in_progress) + send_cleanup_vector(cfg); + return irq; } @@ -4178,28 +4184,20 @@ fake_ioapic_page: } } -static int __init ioapic_insert_resources(void) +void __init ioapic_insert_resources(void) { int i; struct resource *r = ioapic_resources; if (!r) { - if (nr_ioapics > 0) { + if (nr_ioapics > 0) printk(KERN_ERR "IO APIC resources couldn't be allocated.\n"); - return -1; - } - return 0; + return; } for (i = 0; i < nr_ioapics; i++) { insert_resource(&iomem_resource, r); r++; } - - return 0; } - -/* Insert the IO APIC resources after PCI initialization has occured to handle - * IO APICS that are mapped in on a BAR in PCI space. */ -late_initcall(ioapic_insert_resources); diff --git a/arch/x86/kernel/apic/ipi.c b/arch/x86/kernel/apic/ipi.c index dbf5445727a..6ef00ba4c88 100644 --- a/arch/x86/kernel/apic/ipi.c +++ b/arch/x86/kernel/apic/ipi.c @@ -106,6 +106,9 @@ void default_send_IPI_mask_logical(const struct cpumask *cpumask, int vector) unsigned long mask = cpumask_bits(cpumask)[0]; unsigned long flags; + if (WARN_ONCE(!mask, "empty IPI mask")) + return; + local_irq_save(flags); WARN_ON(mask & ~cpumask_bits(cpu_online_mask)[0]); __default_send_IPI_dest_field(mask, vector, apic->dest_logical); diff --git a/arch/x86/kernel/apic/numaq_32.c b/arch/x86/kernel/apic/numaq_32.c index 533e59c6fc8..ca96e68f0d2 100644 --- a/arch/x86/kernel/apic/numaq_32.c +++ b/arch/x86/kernel/apic/numaq_32.c @@ -493,7 +493,8 @@ static void numaq_setup_portio_remap(void) (u_long) xquad_portio, (u_long) num_quads*XQUAD_PORTIO_QUAD); } -struct apic apic_numaq = { +/* Use __refdata to keep false positive warning calm. */ +struct apic __refdata apic_numaq = { .name = "NUMAQ", .probe = probe_numaq, diff --git a/arch/x86/kernel/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c index 440a8bccd91..0c0182cc947 100644 --- a/arch/x86/kernel/apic/probe_32.c +++ b/arch/x86/kernel/apic/probe_32.c @@ -20,23 +20,12 @@ #include <asm/apic.h> #include <asm/setup.h> -#include <linux/threads.h> -#include <linux/cpumask.h> -#include <asm/mpspec.h> -#include <asm/fixmap.h> -#include <asm/apicdef.h> -#include <linux/kernel.h> -#include <linux/string.h> #include <linux/smp.h> -#include <linux/init.h> #include <asm/ipi.h> -#include <linux/smp.h> -#include <linux/init.h> #include <linux/interrupt.h> #include <asm/acpi.h> #include <asm/e820.h> -#include <asm/setup.h> #ifdef CONFIG_HOTPLUG_CPU #define DEFAULT_SEND_IPI (1) diff --git a/arch/x86/kernel/apic/probe_64.c b/arch/x86/kernel/apic/probe_64.c index bc3e880f9b8..fcec2f1d34a 100644 --- a/arch/x86/kernel/apic/probe_64.c +++ b/arch/x86/kernel/apic/probe_64.c @@ -44,6 +44,11 @@ static struct apic *apic_probe[] __initdata = { NULL, }; +static int apicid_phys_pkg_id(int initial_apic_id, int index_msb) +{ + return hard_smp_processor_id() >> index_msb; +} + /* * Check the APIC IDs in bios_cpu_apicid and choose the APIC mode. */ @@ -69,6 +74,11 @@ void __init default_setup_apic_routing(void) printk(KERN_INFO "Setting APIC routing to %s\n", apic->name); } + if (is_vsmp_box()) { + /* need to update phys_pkg_id */ + apic->phys_pkg_id = apicid_phys_pkg_id; + } + /* * Now that apic routing model is selected, configure the * fault handling for intr remapping. diff --git a/arch/x86/kernel/apic/summit_32.c b/arch/x86/kernel/apic/summit_32.c index 344eee4ac0a..eafdfbd1ea9 100644 --- a/arch/x86/kernel/apic/summit_32.c +++ b/arch/x86/kernel/apic/summit_32.c @@ -44,7 +44,6 @@ #include <asm/ipi.h> #include <linux/kernel.h> #include <linux/string.h> -#include <linux/init.h> #include <linux/gfp.h> #include <linux/smp.h> diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c index 8e4cbb255c3..a5371ec3677 100644 --- a/arch/x86/kernel/apic/x2apic_cluster.c +++ b/arch/x86/kernel/apic/x2apic_cluster.c @@ -17,11 +17,13 @@ static int x2apic_acpi_madt_oem_check(char *oem_id, char *oem_table_id) return x2apic_enabled(); } -/* Start with all IRQs pointing to boot CPU. IRQ balancing will shift them. */ - +/* + * need to use more than cpu 0, because we need more vectors when + * MSI-X are used. + */ static const struct cpumask *x2apic_target_cpus(void) { - return cpumask_of(0); + return cpu_online_mask; } /* @@ -170,7 +172,7 @@ static unsigned long set_apic_id(unsigned int id) static int x2apic_cluster_phys_pkg_id(int initial_apicid, int index_msb) { - return current_cpu_data.initial_apicid >> index_msb; + return initial_apicid >> index_msb; } static void x2apic_send_IPI_self(int vector) diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c index a284359627e..a8989aadc99 100644 --- a/arch/x86/kernel/apic/x2apic_phys.c +++ b/arch/x86/kernel/apic/x2apic_phys.c @@ -27,11 +27,13 @@ static int x2apic_acpi_madt_oem_check(char *oem_id, char *oem_table_id) return 0; } -/* Start with all IRQs pointing to boot CPU. IRQ balancing will shift them. */ - +/* + * need to use more than cpu 0, because we need more vectors when + * MSI-X are used. + */ static const struct cpumask *x2apic_target_cpus(void) { - return cpumask_of(0); + return cpu_online_mask; } static void x2apic_vector_allocation_domain(int cpu, struct cpumask *retmask) @@ -162,7 +164,7 @@ static unsigned long set_apic_id(unsigned int id) static int x2apic_phys_pkg_id(int initial_apicid, int index_msb) { - return current_cpu_data.initial_apicid >> index_msb; + return initial_apicid >> index_msb; } static void x2apic_send_IPI_self(int vector) diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 096d19aea2f..601159374e8 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -46,7 +46,7 @@ static int early_get_nodeid(void) return node_id.s.node_id; } -static int uv_acpi_madt_oem_check(char *oem_id, char *oem_table_id) +static int __init uv_acpi_madt_oem_check(char *oem_id, char *oem_table_id) { if (!strcmp(oem_id, "SGI")) { if (!strcmp(oem_table_id, "UVL")) @@ -253,7 +253,7 @@ static void uv_send_IPI_self(int vector) apic_write(APIC_SELF_IPI, vector); } -struct apic apic_x2apic_uv_x = { +struct apic __refdata apic_x2apic_uv_x = { .name = "UV large system", .probe = NULL, @@ -261,7 +261,7 @@ struct apic apic_x2apic_uv_x = { .apic_id_registered = uv_apic_id_registered, .irq_delivery_mode = dest_Fixed, - .irq_dest_mode = 1, /* logical */ + .irq_dest_mode = 0, /* physical */ .target_cpus = uv_target_cpus, .disable_esr = 0, @@ -362,12 +362,6 @@ static __init void get_lowmem_redirect(unsigned long *base, unsigned long *size) BUG(); } -static __init void map_low_mmrs(void) -{ - init_extra_mapping_uc(UV_GLOBAL_MMR32_BASE, UV_GLOBAL_MMR32_SIZE); - init_extra_mapping_uc(UV_LOCAL_MMR_BASE, UV_LOCAL_MMR_SIZE); -} - enum map_type {map_wb, map_uc}; static __init void map_high(char *id, unsigned long base, int shift, @@ -395,26 +389,6 @@ static __init void map_gru_high(int max_pnode) map_high("GRU", gru.s.base, shift, max_pnode, map_wb); } -static __init void map_config_high(int max_pnode) -{ - union uvh_rh_gam_cfg_overlay_config_mmr_u cfg; - int shift = UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR_BASE_SHFT; - - cfg.v = uv_read_local_mmr(UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR); - if (cfg.s.enable) - map_high("CONFIG", cfg.s.base, shift, max_pnode, map_uc); -} - -static __init void map_mmr_high(int max_pnode) -{ - union uvh_rh_gam_mmr_overlay_config_mmr_u mmr; - int shift = UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_BASE_SHFT; - - mmr.v = uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR); - if (mmr.s.enable) - map_high("MMR", mmr.s.base, shift, max_pnode, map_uc); -} - static __init void map_mmioh_high(int max_pnode) { union uvh_rh_gam_mmioh_overlay_config_mmr_u mmioh; @@ -566,8 +540,6 @@ void __init uv_system_init(void) unsigned long mmr_base, present, paddr; unsigned short pnode_mask; - map_low_mmrs(); - m_n_config.v = uv_read_local_mmr(UVH_SI_ADDR_MAP_CONFIG); m_val = m_n_config.s.m_skt; n_val = m_n_config.s.n_skt; @@ -591,6 +563,8 @@ 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); + for (blade = 0; blade < uv_num_possible_blades(); blade++) + uv_blade_info[blade].memory_nid = -1; get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size); @@ -629,6 +603,9 @@ void __init uv_system_init(void) lcpu = uv_blade_info[blade].nr_possible_cpus; uv_blade_info[blade].nr_possible_cpus++; + /* Any node on the blade, else will contain -1. */ + uv_blade_info[blade].memory_nid = nid; + uv_cpu_hub_info(cpu)->lowmem_remap_base = lowmem_redir_base; uv_cpu_hub_info(cpu)->lowmem_remap_top = lowmem_redir_size; uv_cpu_hub_info(cpu)->m_val = m_val; @@ -662,11 +639,10 @@ void __init uv_system_init(void) pnode = (paddr >> m_val) & pnode_mask; blade = boot_pnode_to_blade(pnode); uv_node_to_blade[nid] = blade; + max_pnode = max(pnode, max_pnode); } map_gru_high(max_pnode); - map_mmr_high(max_pnode); - map_config_high(max_pnode); map_mmioh_high(max_pnode); uv_cpu_init(); diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index 79302e9a33a..442b5508893 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -811,7 +811,7 @@ static int apm_do_idle(void) u8 ret = 0; int idled = 0; int polling; - int err; + int err = 0; polling = !!(current_thread_info()->status & TS_POLLING); if (polling) { diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c index 898ecc47e12..4a6aeedcd96 100644 --- a/arch/x86/kernel/asm-offsets_64.c +++ b/arch/x86/kernel/asm-offsets_64.c @@ -3,6 +3,7 @@ * This code generates raw asm output which is post-processed to extract * and format the required data. */ +#define COMPILE_OFFSETS #include <linux/crypto.h> #include <linux/sched.h> diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index 3efcb2b96a1..c1f253dac15 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -7,6 +7,10 @@ ifdef CONFIG_FUNCTION_TRACER CFLAGS_REMOVE_common.o = -pg endif +# Make sure load_percpu_segment has no stackprotector +nostackp := $(call cc-option, -fno-stack-protector) +CFLAGS_common.o := $(nostackp) + obj-y := intel_cacheinfo.o addon_cpuid_features.o obj-y += proc.o capflags.o powerflags.o common.o obj-y += vmware.o hypervisor.o diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index e5b27d8f1b4..63fddcd082c 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -258,13 +258,15 @@ static void __cpuinit amd_detect_cmp(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_HT unsigned bits; + int cpu = smp_processor_id(); bits = c->x86_coreid_bits; - /* Low order bits define the core id (index of core in socket) */ c->cpu_core_id = c->initial_apicid & ((1 << bits)-1); /* Convert the initial APIC ID into the socket ID */ c->phys_proc_id = c->initial_apicid >> bits; + /* use socket ID also for last level cache */ + per_cpu(cpu_llc_id, cpu) = c->phys_proc_id; #endif } @@ -354,7 +356,7 @@ static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) #endif #if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_PCI) /* check CPU config space for extended APIC ID */ - if (c->x86 >= 0xf) { + if (cpu_has_apic && c->x86 >= 0xf) { unsigned int val; val = read_pci_config(0, 24, 0, 0x68); if ((val & ((1 << 17) | (1 << 18))) == ((1 << 17) | (1 << 18))) @@ -398,6 +400,13 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) level = cpuid_eax(1); if((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58) set_cpu_cap(c, X86_FEATURE_REP_GOOD); + + /* + * Some BIOSes incorrectly force this feature, but only K8 + * revision D (model = 0x14) and later actually support it. + */ + if (c->x86_model < 0x14) + clear_cpu_cap(c, X86_FEATURE_LAHF_LM); } if (c->x86 == 0x10 || c->x86 == 0x11) set_cpu_cap(c, X86_FEATURE_REP_GOOD); diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 9fa33886c0d..5ce60a88027 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -59,7 +59,30 @@ void __init setup_cpu_local_masks(void) alloc_bootmem_cpumask_var(&cpu_sibling_setup_mask); } -static const struct cpu_dev *this_cpu __cpuinitdata; +static void __cpuinit default_init(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_X86_64 + display_cacheinfo(c); +#else + /* Not much we can do here... */ + /* Check if at least it has cpuid */ + if (c->cpuid_level == -1) { + /* No cpuid. It must be an ancient CPU */ + if (c->x86 == 4) + strcpy(c->x86_model_id, "486"); + else if (c->x86 == 3) + strcpy(c->x86_model_id, "386"); + } +#endif +} + +static const struct cpu_dev __cpuinitconst default_cpu = { + .c_init = default_init, + .c_vendor = "Unknown", + .c_x86_vendor = X86_VENDOR_UNKNOWN, +}; + +static const struct cpu_dev *this_cpu __cpuinitdata = &default_cpu; DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = { #ifdef CONFIG_X86_64 @@ -108,7 +131,7 @@ DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = { /* data */ [GDT_ENTRY_APMBIOS_BASE+2] = { { { 0x0000ffff, 0x00409200 } } }, - [GDT_ENTRY_ESPFIX_SS] = { { { 0x00000000, 0x00c09200 } } }, + [GDT_ENTRY_ESPFIX_SS] = { { { 0x0000ffff, 0x00cf9200 } } }, [GDT_ENTRY_PERCPU] = { { { 0x0000ffff, 0x00cf9200 } } }, GDT_STACK_CANARY_INIT #endif @@ -332,29 +355,6 @@ void switch_to_new_gdt(int cpu) static const struct cpu_dev *__cpuinitdata cpu_devs[X86_VENDOR_NUM] = {}; -static void __cpuinit default_init(struct cpuinfo_x86 *c) -{ -#ifdef CONFIG_X86_64 - display_cacheinfo(c); -#else - /* Not much we can do here... */ - /* Check if at least it has cpuid */ - if (c->cpuid_level == -1) { - /* No cpuid. It must be an ancient CPU */ - if (c->x86 == 4) - strcpy(c->x86_model_id, "486"); - else if (c->x86 == 3) - strcpy(c->x86_model_id, "386"); - } -#endif -} - -static const struct cpu_dev __cpuinitconst default_cpu = { - .c_init = default_init, - .c_vendor = "Unknown", - .c_x86_vendor = X86_VENDOR_UNKNOWN, -}; - static void __cpuinit get_model_name(struct cpuinfo_x86 *c) { unsigned int *v; diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c index cf52215d9eb..2a50ef89100 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c @@ -1,3 +1,4 @@ + /* * (c) 2003-2006 Advanced Micro Devices, Inc. * Your use of this code is subject to the terms and conditions of the @@ -117,20 +118,17 @@ static int query_current_values_with_pending_wait(struct powernow_k8_data *data) u32 i = 0; if (cpu_family == CPU_HW_PSTATE) { - if (data->currpstate == HW_PSTATE_INVALID) { - /* read (initial) hw pstate if not yet set */ - rdmsr(MSR_PSTATE_STATUS, lo, hi); - i = lo & HW_PSTATE_MASK; - - /* - * a workaround for family 11h erratum 311 might cause - * an "out-of-range Pstate if the core is in Pstate-0 - */ - if (i >= data->numps) - data->currpstate = HW_PSTATE_0; - else - data->currpstate = i; - } + rdmsr(MSR_PSTATE_STATUS, lo, hi); + i = lo & HW_PSTATE_MASK; + data->currpstate = i; + + /* + * a workaround for family 11h erratum 311 might cause + * an "out-of-range Pstate if the core is in Pstate-0 + */ + if ((boot_cpu_data.x86 == 0x11) && (i >= data->numps)) + data->currpstate = HW_PSTATE_0; + return 0; } do { @@ -301,7 +299,7 @@ static int transition_pstate(struct powernow_k8_data *data, u32 pstate) static int transition_fid_vid(struct powernow_k8_data *data, u32 reqfid, u32 reqvid) { - if (core_voltage_pre_transition(data, reqvid)) + if (core_voltage_pre_transition(data, reqvid, reqfid)) return 1; if (core_frequency_transition(data, reqfid)) @@ -329,17 +327,20 @@ static int transition_fid_vid(struct powernow_k8_data *data, /* Phase 1 - core voltage transition ... setup voltage */ static int core_voltage_pre_transition(struct powernow_k8_data *data, - u32 reqvid) + u32 reqvid, u32 reqfid) { u32 rvosteps = data->rvo; u32 savefid = data->currfid; - u32 maxvid, lo; + u32 maxvid, lo, rvomult = 1; dprintk("ph1 (cpu%d): start, currfid 0x%x, currvid 0x%x, " "reqvid 0x%x, rvo 0x%x\n", smp_processor_id(), data->currfid, data->currvid, reqvid, data->rvo); + if ((savefid < LO_FID_TABLE_TOP) && (reqfid < LO_FID_TABLE_TOP)) + rvomult = 2; + rvosteps *= rvomult; rdmsr(MSR_FIDVID_STATUS, lo, maxvid); maxvid = 0x1f & (maxvid >> 16); dprintk("ph1 maxvid=0x%x\n", maxvid); @@ -353,7 +354,8 @@ static int core_voltage_pre_transition(struct powernow_k8_data *data, return 1; } - while ((rvosteps > 0) && ((data->rvo + data->currvid) > reqvid)) { + while ((rvosteps > 0) && + ((rvomult * data->rvo + data->currvid) > reqvid)) { if (data->currvid == maxvid) { rvosteps = 0; } else { @@ -386,13 +388,6 @@ static int core_frequency_transition(struct powernow_k8_data *data, u32 reqfid) u32 vcoreqfid, vcocurrfid, vcofiddiff; u32 fid_interval, savevid = data->currvid; - if ((reqfid < HI_FID_TABLE_BOTTOM) && - (data->currfid < HI_FID_TABLE_BOTTOM)) { - printk(KERN_ERR PFX "ph2: illegal lo-lo transition " - "0x%x 0x%x\n", reqfid, data->currfid); - return 1; - } - if (data->currfid == reqfid) { printk(KERN_ERR PFX "ph2 null fid transition 0x%x\n", data->currfid); @@ -409,6 +404,9 @@ static int core_frequency_transition(struct powernow_k8_data *data, u32 reqfid) vcofiddiff = vcocurrfid > vcoreqfid ? vcocurrfid - vcoreqfid : vcoreqfid - vcocurrfid; + if ((reqfid <= LO_FID_TABLE_TOP) && (data->currfid <= LO_FID_TABLE_TOP)) + vcofiddiff = 0; + while (vcofiddiff > 2) { (data->currfid & 1) ? (fid_interval = 1) : (fid_interval = 2); @@ -510,41 +508,34 @@ static int core_voltage_post_transition(struct powernow_k8_data *data, return 0; } -static int check_supported_cpu(unsigned int cpu) +static void check_supported_cpu(void *_rc) { - cpumask_t oldmask; u32 eax, ebx, ecx, edx; - unsigned int rc = 0; + int *rc = _rc; - oldmask = current->cpus_allowed; - set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); - - if (smp_processor_id() != cpu) { - printk(KERN_ERR PFX "limiting to cpu %u failed\n", cpu); - goto out; - } + *rc = -ENODEV; if (current_cpu_data.x86_vendor != X86_VENDOR_AMD) - goto out; + return; eax = cpuid_eax(CPUID_PROCESSOR_SIGNATURE); if (((eax & CPUID_XFAM) != CPUID_XFAM_K8) && ((eax & CPUID_XFAM) < CPUID_XFAM_10H)) - goto out; + return; if ((eax & CPUID_XFAM) == CPUID_XFAM_K8) { if (((eax & CPUID_USE_XFAM_XMOD) != CPUID_USE_XFAM_XMOD) || ((eax & CPUID_XMOD) > CPUID_XMOD_REV_MASK)) { printk(KERN_INFO PFX "Processor cpuid %x not supported\n", eax); - goto out; + return; } eax = cpuid_eax(CPUID_GET_MAX_CAPABILITIES); if (eax < CPUID_FREQ_VOLT_CAPABILITIES) { printk(KERN_INFO PFX "No frequency change capabilities detected\n"); - goto out; + return; } cpuid(CPUID_FREQ_VOLT_CAPABILITIES, &eax, &ebx, &ecx, &edx); @@ -552,21 +543,17 @@ static int check_supported_cpu(unsigned int cpu) != P_STATE_TRANSITION_CAPABLE) { printk(KERN_INFO PFX "Power state transitions not supported\n"); - goto out; + return; } } else { /* must be a HW Pstate capable processor */ cpuid(CPUID_FREQ_VOLT_CAPABILITIES, &eax, &ebx, &ecx, &edx); if ((edx & USE_HW_PSTATE) == USE_HW_PSTATE) cpu_family = CPU_HW_PSTATE; else - goto out; + return; } - rc = 1; - -out: - set_cpus_allowed_ptr(current, &oldmask); - return rc; + *rc = 0; } static int check_pst_table(struct powernow_k8_data *data, struct pst_s *pst, @@ -823,13 +810,14 @@ static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, if (!data->acpi_data.state_count || (cpu_family == CPU_HW_PSTATE)) return; - control = data->acpi_data.states[index].control; data->irt = (control - >> IRT_SHIFT) & IRT_MASK; data->rvo = (control >> - RVO_SHIFT) & RVO_MASK; data->exttype = (control - >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK; - data->plllock = (control >> PLL_L_SHIFT) & PLL_L_MASK; data->vidmvs = 1 - << ((control >> MVS_SHIFT) & MVS_MASK); data->vstable = - (control >> VST_SHIFT) & VST_MASK; } + control = data->acpi_data.states[index].control; + data->irt = (control >> IRT_SHIFT) & IRT_MASK; + data->rvo = (control >> RVO_SHIFT) & RVO_MASK; + data->exttype = (control >> EXT_TYPE_SHIFT) & EXT_TYPE_MASK; + data->plllock = (control >> PLL_L_SHIFT) & PLL_L_MASK; + data->vidmvs = 1 << ((control >> MVS_SHIFT) & MVS_MASK); + data->vstable = (control >> VST_SHIFT) & VST_MASK; +} static int powernow_k8_cpu_init_acpi(struct powernow_k8_data *data) { @@ -1046,6 +1034,19 @@ static int get_transition_latency(struct powernow_k8_data *data) if (cur_latency > max_latency) max_latency = cur_latency; } + if (max_latency == 0) { + /* + * Fam 11h always returns 0 as transition latency. + * This is intended and means "very fast". While cpufreq core + * and governors currently can handle that gracefully, better + * set it to 1 to avoid problems in the future. + * For all others it's a BIOS bug. + */ + if (!boot_cpu_data.x86 == 0x11) + printk(KERN_ERR FW_WARN PFX "Invalid zero transition " + "latency\n"); + max_latency = 1; + } /* value in usecs, needs to be in nanoseconds */ return 1000 * max_latency; } @@ -1080,20 +1081,12 @@ static int transition_frequency_fidvid(struct powernow_k8_data *data, return 0; } - if ((fid < HI_FID_TABLE_BOTTOM) && - (data->currfid < HI_FID_TABLE_BOTTOM)) { - printk(KERN_ERR PFX - "ignoring illegal change in lo freq table-%x to 0x%x\n", - data->currfid, fid); - return 1; - } - dprintk("cpu %d, changing to fid 0x%x, vid 0x%x\n", smp_processor_id(), fid, vid); freqs.old = find_khz_freq_from_fid(data->currfid); freqs.new = find_khz_freq_from_fid(fid); - for_each_cpu_mask_nr(i, *(data->available_cores)) { + for_each_cpu(i, data->available_cores) { freqs.cpu = i; cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); } @@ -1101,7 +1094,7 @@ static int transition_frequency_fidvid(struct powernow_k8_data *data, res = transition_fid_vid(data, fid, vid); freqs.new = find_khz_freq_from_fid(data->currfid); - for_each_cpu_mask_nr(i, *(data->available_cores)) { + for_each_cpu(i, data->available_cores) { freqs.cpu = i; cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); } @@ -1126,7 +1119,7 @@ static int transition_frequency_pstate(struct powernow_k8_data *data, data->currpstate); freqs.new = find_khz_freq_from_pstate(data->powernow_table, pstate); - for_each_cpu_mask_nr(i, *(data->available_cores)) { + for_each_cpu(i, data->available_cores) { freqs.cpu = i; cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); } @@ -1134,7 +1127,7 @@ static int transition_frequency_pstate(struct powernow_k8_data *data, res = transition_pstate(data, pstate); freqs.new = find_khz_freq_from_pstate(data->powernow_table, pstate); - for_each_cpu_mask_nr(i, *(data->available_cores)) { + for_each_cpu(i, data->available_cores) { freqs.cpu = i; cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); } @@ -1235,21 +1228,47 @@ static int powernowk8_verify(struct cpufreq_policy *pol) return cpufreq_frequency_table_verify(pol, data->powernow_table); } -static const char ACPI_PSS_BIOS_BUG_MSG[] = - KERN_ERR FW_BUG PFX "No compatible ACPI _PSS objects found.\n" - KERN_ERR FW_BUG PFX "Try again with latest BIOS.\n"; +struct init_on_cpu { + struct powernow_k8_data *data; + int rc; +}; + +static void __cpuinit powernowk8_cpu_init_on_cpu(void *_init_on_cpu) +{ + struct init_on_cpu *init_on_cpu = _init_on_cpu; + + if (pending_bit_stuck()) { + printk(KERN_ERR PFX "failing init, change pending bit set\n"); + init_on_cpu->rc = -ENODEV; + return; + } + + if (query_current_values_with_pending_wait(init_on_cpu->data)) { + init_on_cpu->rc = -ENODEV; + return; + } + + if (cpu_family == CPU_OPTERON) + fidvid_msr_init(); + + init_on_cpu->rc = 0; +} /* per CPU init entry point to the driver */ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) { + static const char ACPI_PSS_BIOS_BUG_MSG[] = + KERN_ERR FW_BUG PFX "No compatible ACPI _PSS objects found.\n" + FW_BUG PFX "Try again with latest BIOS.\n"; struct powernow_k8_data *data; - cpumask_t oldmask; + struct init_on_cpu init_on_cpu; int rc; if (!cpu_online(pol->cpu)) return -ENODEV; - if (!check_supported_cpu(pol->cpu)) + smp_call_function_single(pol->cpu, check_supported_cpu, &rc, 1); + if (rc) return -ENODEV; data = kzalloc(sizeof(struct powernow_k8_data), GFP_KERNEL); @@ -1289,27 +1308,12 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) pol->cpuinfo.transition_latency = get_transition_latency(data); /* only run on specific CPU from here on */ - oldmask = current->cpus_allowed; - set_cpus_allowed_ptr(current, &cpumask_of_cpu(pol->cpu)); - - if (smp_processor_id() != pol->cpu) { - printk(KERN_ERR PFX "limiting to cpu %u failed\n", pol->cpu); - goto err_out_unmask; - } - - if (pending_bit_stuck()) { - printk(KERN_ERR PFX "failing init, change pending bit set\n"); - goto err_out_unmask; - } - - if (query_current_values_with_pending_wait(data)) - goto err_out_unmask; - - if (cpu_family == CPU_OPTERON) - fidvid_msr_init(); - - /* run on any CPU again */ - set_cpus_allowed_ptr(current, &oldmask); + init_on_cpu.data = data; + smp_call_function_single(data->cpu, powernowk8_cpu_init_on_cpu, + &init_on_cpu, 1); + rc = init_on_cpu.rc; + if (rc != 0) + goto err_out_exit_acpi; if (cpu_family == CPU_HW_PSTATE) cpumask_copy(pol->cpus, cpumask_of(pol->cpu)); @@ -1346,8 +1350,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) return 0; -err_out_unmask: - set_cpus_allowed_ptr(current, &oldmask); +err_out_exit_acpi: powernow_k8_cpu_exit_acpi(data); err_out: @@ -1372,28 +1375,25 @@ static int __devexit powernowk8_cpu_exit(struct cpufreq_policy *pol) return 0; } +static void query_values_on_cpu(void *_err) +{ + int *err = _err; + struct powernow_k8_data *data = __get_cpu_var(powernow_data); + + *err = query_current_values_with_pending_wait(data); +} + static unsigned int powernowk8_get(unsigned int cpu) { - struct powernow_k8_data *data; - cpumask_t oldmask = current->cpus_allowed; + struct powernow_k8_data *data = per_cpu(powernow_data, cpu); unsigned int khz = 0; - unsigned int first; - - first = cpumask_first(cpu_core_mask(cpu)); - data = per_cpu(powernow_data, first); + int err; if (!data) return -EINVAL; - set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); - if (smp_processor_id() != cpu) { - printk(KERN_ERR PFX - "limiting to CPU %d failed in powernowk8_get\n", cpu); - set_cpus_allowed_ptr(current, &oldmask); - return 0; - } - - if (query_current_values_with_pending_wait(data)) + smp_call_function_single(cpu, query_values_on_cpu, &err, true); + if (err) goto out; if (cpu_family == CPU_HW_PSTATE) @@ -1404,7 +1404,6 @@ static unsigned int powernowk8_get(unsigned int cpu) out: - set_cpus_allowed_ptr(current, &oldmask); return khz; } @@ -1430,7 +1429,9 @@ static int __cpuinit powernowk8_init(void) unsigned int i, supported_cpus = 0; for_each_online_cpu(i) { - if (check_supported_cpu(i)) + int rc; + smp_call_function_single(i, check_supported_cpu, &rc, 1); + if (rc == 0) supported_cpus++; } diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.h b/arch/x86/kernel/cpu/cpufreq/powernow-k8.h index 6c6698feade..02ce824073c 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.h +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.h @@ -215,7 +215,8 @@ struct pst_s { #define dprintk(msg...) cpufreq_debug_printk(CPUFREQ_DEBUG_DRIVER, "powernow-k8", msg) -static int core_voltage_pre_transition(struct powernow_k8_data *data, u32 reqvid); +static int core_voltage_pre_transition(struct powernow_k8_data *data, + u32 reqvid, u32 regfid); static int core_voltage_post_transition(struct powernow_k8_data *data, u32 reqvid); static int core_frequency_transition(struct powernow_k8_data *data, u32 reqfid); @@ -223,14 +224,3 @@ static void powernow_k8_acpi_pst_values(struct powernow_k8_data *data, unsigned static int fill_powernow_table_pstate(struct powernow_k8_data *data, struct cpufreq_frequency_table *powernow_table); static int fill_powernow_table_fidvid(struct powernow_k8_data *data, struct cpufreq_frequency_table *powernow_table); - -#ifdef CONFIG_SMP -static inline void define_siblings(int cpu, cpumask_t cpu_sharedcore_mask[]) -{ -} -#else -static inline void define_siblings(int cpu, cpumask_t cpu_sharedcore_mask[]) -{ - cpu_set(0, cpu_sharedcore_mask[0]); -} -#endif diff --git a/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c index 55c831ed71c..8d672ef162c 100644 --- a/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c +++ b/arch/x86/kernel/cpu/cpufreq/speedstep-centrino.c @@ -323,14 +323,8 @@ static unsigned int get_cur_freq(unsigned int cpu) { unsigned l, h; unsigned clock_freq; - cpumask_t saved_mask; - saved_mask = current->cpus_allowed; - set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu)); - if (smp_processor_id() != cpu) - return 0; - - rdmsr(MSR_IA32_PERF_STATUS, l, h); + rdmsr_on_cpu(cpu, MSR_IA32_PERF_STATUS, &l, &h); clock_freq = extract_clock(l, cpu, 0); if (unlikely(clock_freq == 0)) { @@ -340,11 +334,9 @@ static unsigned int get_cur_freq(unsigned int cpu) * P-state transition (like TM2). Get the last freq set * in PERF_CTL. */ - rdmsr(MSR_IA32_PERF_CTL, l, h); + rdmsr_on_cpu(cpu, MSR_IA32_PERF_CTL, &l, &h); clock_freq = extract_clock(l, cpu, 1); } - - set_cpus_allowed_ptr(current, &saved_mask); return clock_freq; } @@ -467,15 +459,10 @@ static int centrino_target (struct cpufreq_policy *policy, struct cpufreq_freqs freqs; int retval = 0; unsigned int j, k, first_cpu, tmp; - cpumask_var_t saved_mask, covered_cpus; + cpumask_var_t covered_cpus; - if (unlikely(!alloc_cpumask_var(&saved_mask, GFP_KERNEL))) - return -ENOMEM; - if (unlikely(!zalloc_cpumask_var(&covered_cpus, GFP_KERNEL))) { - free_cpumask_var(saved_mask); + if (unlikely(!zalloc_cpumask_var(&covered_cpus, GFP_KERNEL))) return -ENOMEM; - } - cpumask_copy(saved_mask, ¤t->cpus_allowed); if (unlikely(per_cpu(centrino_model, cpu) == NULL)) { retval = -ENODEV; @@ -493,7 +480,7 @@ static int centrino_target (struct cpufreq_policy *policy, first_cpu = 1; for_each_cpu(j, policy->cpus) { - const struct cpumask *mask; + int good_cpu; /* cpufreq holds the hotplug lock, so we are safe here */ if (!cpu_online(j)) @@ -504,32 +491,30 @@ static int centrino_target (struct cpufreq_policy *policy, * Make sure we are running on CPU that wants to change freq */ if (policy->shared_type == CPUFREQ_SHARED_TYPE_ANY) - mask = policy->cpus; + good_cpu = cpumask_any_and(policy->cpus, + cpu_online_mask); else - mask = cpumask_of(j); + good_cpu = j; - set_cpus_allowed_ptr(current, mask); - preempt_disable(); - if (unlikely(!cpu_isset(smp_processor_id(), *mask))) { + if (good_cpu >= nr_cpu_ids) { dprintk("couldn't limit to CPUs in this domain\n"); retval = -EAGAIN; if (first_cpu) { /* We haven't started the transition yet. */ - goto migrate_end; + goto out; } - preempt_enable(); break; } msr = per_cpu(centrino_model, cpu)->op_points[newstate].index; if (first_cpu) { - rdmsr(MSR_IA32_PERF_CTL, oldmsr, h); + rdmsr_on_cpu(good_cpu, MSR_IA32_PERF_CTL, &oldmsr, &h); if (msr == (oldmsr & 0xffff)) { dprintk("no change needed - msr was and needs " "to be %x\n", oldmsr); retval = 0; - goto migrate_end; + goto out; } freqs.old = extract_clock(oldmsr, cpu, 0); @@ -553,14 +538,11 @@ static int centrino_target (struct cpufreq_policy *policy, oldmsr |= msr; } - wrmsr(MSR_IA32_PERF_CTL, oldmsr, h); - if (policy->shared_type == CPUFREQ_SHARED_TYPE_ANY) { - preempt_enable(); + wrmsr_on_cpu(good_cpu, MSR_IA32_PERF_CTL, oldmsr, h); + if (policy->shared_type == CPUFREQ_SHARED_TYPE_ANY) break; - } - cpu_set(j, *covered_cpus); - preempt_enable(); + cpumask_set_cpu(j, covered_cpus); } for_each_cpu(k, policy->cpus) { @@ -578,10 +560,8 @@ static int centrino_target (struct cpufreq_policy *policy, * Best effort undo.. */ - for_each_cpu_mask_nr(j, *covered_cpus) { - set_cpus_allowed_ptr(current, &cpumask_of_cpu(j)); - wrmsr(MSR_IA32_PERF_CTL, oldmsr, h); - } + for_each_cpu(j, covered_cpus) + wrmsr_on_cpu(j, MSR_IA32_PERF_CTL, oldmsr, h); tmp = freqs.new; freqs.new = freqs.old; @@ -593,15 +573,9 @@ static int centrino_target (struct cpufreq_policy *policy, cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); } } - set_cpus_allowed_ptr(current, saved_mask); retval = 0; - goto out; -migrate_end: - preempt_enable(); - set_cpus_allowed_ptr(current, saved_mask); out: - free_cpumask_var(saved_mask); free_cpumask_var(covered_cpus); return retval; } diff --git a/arch/x86/kernel/cpu/cpufreq/speedstep-ich.c b/arch/x86/kernel/cpu/cpufreq/speedstep-ich.c index 016c1a4fa3f..6911e91fb4f 100644 --- a/arch/x86/kernel/cpu/cpufreq/speedstep-ich.c +++ b/arch/x86/kernel/cpu/cpufreq/speedstep-ich.c @@ -89,7 +89,8 @@ static int speedstep_find_register(void) * speedstep_set_state - set the SpeedStep state * @state: new processor frequency state (SPEEDSTEP_LOW or SPEEDSTEP_HIGH) * - * Tries to change the SpeedStep state. + * Tries to change the SpeedStep state. Can be called from + * smp_call_function_single. */ static void speedstep_set_state(unsigned int state) { @@ -143,6 +144,11 @@ static void speedstep_set_state(unsigned int state) return; } +/* Wrapper for smp_call_function_single. */ +static void _speedstep_set_state(void *_state) +{ + speedstep_set_state(*(unsigned int *)_state); +} /** * speedstep_activate - activate SpeedStep control in the chipset @@ -226,22 +232,28 @@ static unsigned int speedstep_detect_chipset(void) return 0; } -static unsigned int _speedstep_get(const struct cpumask *cpus) -{ +struct get_freq_data { unsigned int speed; - cpumask_t cpus_allowed; - - cpus_allowed = current->cpus_allowed; - set_cpus_allowed_ptr(current, cpus); - speed = speedstep_get_frequency(speedstep_processor); - set_cpus_allowed_ptr(current, &cpus_allowed); - dprintk("detected %u kHz as current frequency\n", speed); - return speed; + unsigned int processor; +}; + +static void get_freq_data(void *_data) +{ + struct get_freq_data *data = _data; + + data->speed = speedstep_get_frequency(data->processor); } static unsigned int speedstep_get(unsigned int cpu) { - return _speedstep_get(cpumask_of(cpu)); + struct get_freq_data data = { .processor = cpu }; + + /* You're supposed to ensure CPU is online. */ + if (smp_call_function_single(cpu, get_freq_data, &data, 1) != 0) + BUG(); + + dprintk("detected %u kHz as current frequency\n", data.speed); + return data.speed; } /** @@ -257,16 +269,16 @@ static int speedstep_target(struct cpufreq_policy *policy, unsigned int target_freq, unsigned int relation) { - unsigned int newstate = 0; + unsigned int newstate = 0, policy_cpu; struct cpufreq_freqs freqs; - cpumask_t cpus_allowed; int i; if (cpufreq_frequency_table_target(policy, &speedstep_freqs[0], target_freq, relation, &newstate)) return -EINVAL; - freqs.old = _speedstep_get(policy->cpus); + policy_cpu = cpumask_any_and(policy->cpus, cpu_online_mask); + freqs.old = speedstep_get(policy_cpu); freqs.new = speedstep_freqs[newstate].frequency; freqs.cpu = policy->cpu; @@ -276,20 +288,13 @@ static int speedstep_target(struct cpufreq_policy *policy, if (freqs.old == freqs.new) return 0; - cpus_allowed = current->cpus_allowed; - for_each_cpu(i, policy->cpus) { freqs.cpu = i; cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); } - /* switch to physical CPU where state is to be changed */ - set_cpus_allowed_ptr(current, policy->cpus); - - speedstep_set_state(newstate); - - /* allow to be run on all CPUs */ - set_cpus_allowed_ptr(current, &cpus_allowed); + smp_call_function_single(policy_cpu, _speedstep_set_state, &newstate, + true); for_each_cpu(i, policy->cpus) { freqs.cpu = i; @@ -312,33 +317,43 @@ static int speedstep_verify(struct cpufreq_policy *policy) return cpufreq_frequency_table_verify(policy, &speedstep_freqs[0]); } +struct get_freqs { + struct cpufreq_policy *policy; + int ret; +}; + +static void get_freqs_on_cpu(void *_get_freqs) +{ + struct get_freqs *get_freqs = _get_freqs; + + get_freqs->ret = + speedstep_get_freqs(speedstep_processor, + &speedstep_freqs[SPEEDSTEP_LOW].frequency, + &speedstep_freqs[SPEEDSTEP_HIGH].frequency, + &get_freqs->policy->cpuinfo.transition_latency, + &speedstep_set_state); +} static int speedstep_cpu_init(struct cpufreq_policy *policy) { - int result = 0; - unsigned int speed; - cpumask_t cpus_allowed; + int result; + unsigned int policy_cpu, speed; + struct get_freqs gf; /* only run on CPU to be set, or on its sibling */ #ifdef CONFIG_SMP cpumask_copy(policy->cpus, cpu_sibling_mask(policy->cpu)); #endif - - cpus_allowed = current->cpus_allowed; - set_cpus_allowed_ptr(current, policy->cpus); + policy_cpu = cpumask_any_and(policy->cpus, cpu_online_mask); /* detect low and high frequency and transition latency */ - result = speedstep_get_freqs(speedstep_processor, - &speedstep_freqs[SPEEDSTEP_LOW].frequency, - &speedstep_freqs[SPEEDSTEP_HIGH].frequency, - &policy->cpuinfo.transition_latency, - &speedstep_set_state); - set_cpus_allowed_ptr(current, &cpus_allowed); - if (result) - return result; + gf.policy = policy; + smp_call_function_single(policy_cpu, get_freqs_on_cpu, &gf, 1); + if (gf.ret) + return gf.ret; /* get current speed setting */ - speed = _speedstep_get(policy->cpus); + speed = speedstep_get(policy_cpu); if (!speed) return -EIO; diff --git a/arch/x86/kernel/cpu/cpufreq/speedstep-lib.c b/arch/x86/kernel/cpu/cpufreq/speedstep-lib.c index 2e3c6862657..f4c290b8482 100644 --- a/arch/x86/kernel/cpu/cpufreq/speedstep-lib.c +++ b/arch/x86/kernel/cpu/cpufreq/speedstep-lib.c @@ -226,6 +226,7 @@ static unsigned int pentium4_get_frequency(void) } +/* Warning: may get called from smp_call_function_single. */ unsigned int speedstep_get_frequency(unsigned int processor) { switch (processor) { diff --git a/arch/x86/kernel/cpu/mcheck/Makefile b/arch/x86/kernel/cpu/mcheck/Makefile index 45004faf67e..188a1ca5ad2 100644 --- a/arch/x86/kernel/cpu/mcheck/Makefile +++ b/arch/x86/kernel/cpu/mcheck/Makefile @@ -1,11 +1,12 @@ -obj-y = mce.o therm_throt.o +obj-y = mce.o obj-$(CONFIG_X86_NEW_MCE) += mce-severity.o obj-$(CONFIG_X86_OLD_MCE) += k7.o p4.o p6.o obj-$(CONFIG_X86_ANCIENT_MCE) += winchip.o p5.o -obj-$(CONFIG_X86_MCE_P4THERMAL) += mce_intel.o -obj-$(CONFIG_X86_MCE_INTEL) += mce_intel_64.o mce_intel.o -obj-$(CONFIG_X86_MCE_AMD) += mce_amd_64.o +obj-$(CONFIG_X86_MCE_INTEL) += mce_intel.o +obj-$(CONFIG_X86_MCE_AMD) += mce_amd.o obj-$(CONFIG_X86_MCE_NONFATAL) += non-fatal.o obj-$(CONFIG_X86_MCE_THRESHOLD) += threshold.o obj-$(CONFIG_X86_MCE_INJECT) += mce-inject.o + +obj-$(CONFIG_X86_THERMAL_VECTOR) += therm_throt.o diff --git a/arch/x86/kernel/cpu/mcheck/k7.c b/arch/x86/kernel/cpu/mcheck/k7.c index 89e51042415..b945d5dbc60 100644 --- a/arch/x86/kernel/cpu/mcheck/k7.c +++ b/arch/x86/kernel/cpu/mcheck/k7.c @@ -10,10 +10,9 @@ #include <asm/processor.h> #include <asm/system.h> +#include <asm/mce.h> #include <asm/msr.h> -#include "mce.h" - /* Machine Check Handler For AMD Athlon/Duron: */ static void k7_machine_check(struct pt_regs *regs, long error_code) { diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c index fabba15e455..01213048f62 100644 --- a/arch/x86/kernel/cpu/mcheck/mce.c +++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -44,7 +44,6 @@ #include <asm/msr.h> #include "mce-internal.h" -#include "mce.h" /* Handle unconfigured int18 (should never happen) */ static void unexpected_machine_check(struct pt_regs *regs, long error_code) @@ -57,7 +56,7 @@ static void unexpected_machine_check(struct pt_regs *regs, long error_code) void (*machine_check_vector)(struct pt_regs *, long error_code) = unexpected_machine_check; -int mce_disabled; +int mce_disabled __read_mostly; #ifdef CONFIG_X86_NEW_MCE @@ -76,21 +75,22 @@ DEFINE_PER_CPU(unsigned, mce_exception_count); * 2: SIGBUS or log uncorrected errors (if possible), log corrected errors * 3: never panic or SIGBUS, log all errors (for testing only) */ -static int tolerant = 1; -static int banks; -static u64 *bank; -static unsigned long notify_user; -static int rip_msr; -static int mce_bootlog = -1; -static int monarch_timeout = -1; -static int mce_panic_timeout; -static int mce_dont_log_ce; -int mce_cmci_disabled; -int mce_ignore_ce; -int mce_ser; - -static char trigger[128]; -static char *trigger_argv[2] = { trigger, NULL }; +static int tolerant __read_mostly = 1; +static int banks __read_mostly; +static u64 *bank __read_mostly; +static int rip_msr __read_mostly; +static int mce_bootlog __read_mostly = -1; +static int monarch_timeout __read_mostly = -1; +static int mce_panic_timeout __read_mostly; +static int mce_dont_log_ce __read_mostly; +int mce_cmci_disabled __read_mostly; +int mce_ignore_ce __read_mostly; +int mce_ser __read_mostly; + +/* User mode helper program triggered by machine check event */ +static unsigned long mce_need_notify; +static char mce_helper[128]; +static char *mce_helper_argv[2] = { mce_helper, NULL }; static unsigned long dont_init_banks; @@ -180,7 +180,7 @@ void mce_log(struct mce *mce) wmb(); mce->finished = 1; - set_bit(0, ¬ify_user); + set_bit(0, &mce_need_notify); } static void print_mce(struct mce *m) @@ -194,14 +194,14 @@ static void print_mce(struct mce *m) m->cs, m->ip); if (m->cs == __KERNEL_CS) print_symbol("{%s}", m->ip); - printk("\n"); + printk(KERN_CONT "\n"); } printk(KERN_EMERG "TSC %llx ", m->tsc); if (m->addr) - printk("ADDR %llx ", m->addr); + printk(KERN_CONT "ADDR %llx ", m->addr); if (m->misc) - printk("MISC %llx ", m->misc); - printk("\n"); + printk(KERN_CONT "MISC %llx ", m->misc); + printk(KERN_CONT "\n"); printk(KERN_EMERG "PROCESSOR %u:%x TIME %llu SOCKET %u APIC %x\n", m->cpuvendor, m->cpuid, m->time, m->socketid, m->apicid); @@ -209,13 +209,13 @@ static void print_mce(struct mce *m) static void print_mce_head(void) { - printk(KERN_EMERG "\n" KERN_EMERG "HARDWARE ERROR\n"); + printk(KERN_EMERG "\nHARDWARE ERROR\n"); } static void print_mce_tail(void) { printk(KERN_EMERG "This is not a software problem!\n" - KERN_EMERG "Run through mcelog --ascii to decode and contact your hardware vendor\n"); + "Run through mcelog --ascii to decode and contact your hardware vendor\n"); } #define PANIC_TIMEOUT 5 /* 5 seconds */ @@ -691,18 +691,21 @@ static atomic_t global_nwo; * in the entry order. * TBD double check parallel CPU hotunplug */ -static int mce_start(int no_way_out, int *order) +static int mce_start(int *no_way_out) { - int nwo; + int order; int cpus = num_online_cpus(); u64 timeout = (u64)monarch_timeout * NSEC_PER_USEC; - if (!timeout) { - *order = -1; - return no_way_out; - } + if (!timeout) + return -1; - atomic_add(no_way_out, &global_nwo); + atomic_add(*no_way_out, &global_nwo); + /* + * global_nwo should be updated before mce_callin + */ + smp_wmb(); + order = atomic_add_return(1, &mce_callin); /* * Wait for everyone. @@ -710,40 +713,43 @@ static int mce_start(int no_way_out, int *order) while (atomic_read(&mce_callin) != cpus) { if (mce_timed_out(&timeout)) { atomic_set(&global_nwo, 0); - *order = -1; - return no_way_out; + return -1; } ndelay(SPINUNIT); } /* - * Cache the global no_way_out state. + * mce_callin should be read before global_nwo */ - nwo = atomic_read(&global_nwo); + smp_rmb(); - /* - * Monarch starts executing now, the others wait. - */ - if (*order == 1) { + if (order == 1) { + /* + * Monarch: Starts executing now, the others wait. + */ atomic_set(&mce_executing, 1); - return nwo; + } else { + /* + * Subject: Now start the scanning loop one by one in + * the original callin order. + * This way when there are any shared banks it will be + * only seen by one CPU before cleared, avoiding duplicates. + */ + while (atomic_read(&mce_executing) < order) { + if (mce_timed_out(&timeout)) { + atomic_set(&global_nwo, 0); + return -1; + } + ndelay(SPINUNIT); + } } /* - * Now start the scanning loop one by one - * in the original callin order. - * This way when there are any shared banks it will - * be only seen by one CPU before cleared, avoiding duplicates. + * Cache the global no_way_out state. */ - while (atomic_read(&mce_executing) < *order) { - if (mce_timed_out(&timeout)) { - atomic_set(&global_nwo, 0); - *order = -1; - return no_way_out; - } - ndelay(SPINUNIT); - } - return nwo; + *no_way_out = atomic_read(&global_nwo); + + return order; } /* @@ -863,7 +869,6 @@ void do_machine_check(struct pt_regs *regs, long error_code) * check handler. */ int order; - /* * If no_way_out gets set, there is no safe way to recover from this * MCE. If tolerant is cranked up, we'll try anyway. @@ -887,7 +892,6 @@ void do_machine_check(struct pt_regs *regs, long error_code) if (!banks) goto out; - order = atomic_add_return(1, &mce_callin); mce_setup(&m); m.mcgstatus = mce_rdmsrl(MSR_IA32_MCG_STATUS); @@ -909,7 +913,7 @@ void do_machine_check(struct pt_regs *regs, long error_code) * This way we don't report duplicated events on shared banks * because the first one to see it will clear it. */ - no_way_out = mce_start(no_way_out, &order); + order = mce_start(&no_way_out); for (i = 0; i < banks; i++) { __clear_bit(i, toclear); if (!bank[i]) @@ -1113,12 +1117,12 @@ static void mcheck_timer(unsigned long data) *n = min(*n*2, (int)round_jiffies_relative(check_interval*HZ)); t->expires = jiffies + *n; - add_timer(t); + add_timer_on(t, smp_processor_id()); } static void mce_do_trigger(struct work_struct *work) { - call_usermodehelper(trigger, trigger_argv, NULL, UMH_NO_WAIT); + call_usermodehelper(mce_helper, mce_helper_argv, NULL, UMH_NO_WAIT); } static DECLARE_WORK(mce_trigger_work, mce_do_trigger); @@ -1135,7 +1139,7 @@ int mce_notify_irq(void) clear_thread_flag(TIF_MCE_NOTIFY); - if (test_and_clear_bit(0, ¬ify_user)) { + if (test_and_clear_bit(0, &mce_need_notify)) { wake_up_interruptible(&mce_wait); /* @@ -1143,7 +1147,7 @@ int mce_notify_irq(void) * work_pending is always cleared before the function is * executed. */ - if (trigger[0] && !work_pending(&mce_trigger_work)) + if (mce_helper[0] && !work_pending(&mce_trigger_work)) schedule_work(&mce_trigger_work); if (__ratelimit(&ratelimit)) @@ -1222,8 +1226,13 @@ static void mce_init(void) } /* Add per CPU specific workarounds here */ -static void mce_cpu_quirks(struct cpuinfo_x86 *c) +static int mce_cpu_quirks(struct cpuinfo_x86 *c) { + if (c->x86_vendor == X86_VENDOR_UNKNOWN) { + pr_info("MCE: unknown CPU type - not enabling MCE support.\n"); + return -EOPNOTSUPP; + } + /* This should be disabled by the BIOS, but isn't always */ if (c->x86_vendor == X86_VENDOR_AMD) { if (c->x86 == 15 && banks > 4) { @@ -1245,7 +1254,7 @@ static void mce_cpu_quirks(struct cpuinfo_x86 *c) * Various K7s with broken bank 0 around. Always disable * by default. */ - if (c->x86 == 6) + if (c->x86 == 6 && banks > 0) bank[0] = 0; } @@ -1269,11 +1278,20 @@ static void mce_cpu_quirks(struct cpuinfo_x86 *c) if ((c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xe)) && monarch_timeout < 0) monarch_timeout = USEC_PER_SEC; + + /* + * There are also broken BIOSes on some Pentium M and + * earlier systems: + */ + if (c->x86 == 6 && c->x86_model <= 13 && mce_bootlog < 0) + mce_bootlog = 0; } if (monarch_timeout < 0) monarch_timeout = 0; if (mce_bootlog != 0) mce_panic_timeout = 30; + + return 0; } static void __cpuinit mce_ancient_init(struct cpuinfo_x86 *c) @@ -1282,8 +1300,7 @@ static void __cpuinit mce_ancient_init(struct cpuinfo_x86 *c) return; switch (c->x86_vendor) { case X86_VENDOR_INTEL: - if (mce_p5_enabled()) - intel_p5_mcheck_init(c); + intel_p5_mcheck_init(c); break; case X86_VENDOR_CENTAUR: winchip_mcheck_init(c); @@ -1318,7 +1335,7 @@ static void mce_init_timer(void) return; setup_timer(t, mcheck_timer, smp_processor_id()); t->expires = round_jiffies(jiffies + *n); - add_timer(t); + add_timer_on(t, smp_processor_id()); } /* @@ -1335,11 +1352,10 @@ void __cpuinit mcheck_init(struct cpuinfo_x86 *c) if (!mce_available(c)) return; - if (mce_cap_init() < 0) { + if (mce_cap_init() < 0 || mce_cpu_quirks(c) < 0) { mce_disabled = 1; return; } - mce_cpu_quirks(c); machine_check_vector = do_machine_check; @@ -1609,8 +1625,9 @@ static int mce_resume(struct sys_device *dev) static void mce_cpu_restart(void *data) { del_timer_sync(&__get_cpu_var(mce_timer)); - if (mce_available(¤t_cpu_data)) - mce_init(); + if (!mce_available(¤t_cpu_data)) + return; + mce_init(); mce_init_timer(); } @@ -1620,6 +1637,26 @@ static void mce_restart(void) on_each_cpu(mce_cpu_restart, NULL, 1); } +/* Toggle features for corrected errors */ +static void mce_disable_ce(void *all) +{ + if (!mce_available(¤t_cpu_data)) + return; + if (all) + del_timer_sync(&__get_cpu_var(mce_timer)); + cmci_clear(); +} + +static void mce_enable_ce(void *all) +{ + if (!mce_available(¤t_cpu_data)) + return; + cmci_reenable(); + cmci_recheck(); + if (all) + mce_init_timer(); +} + static struct sysdev_class mce_sysclass = { .suspend = mce_suspend, .shutdown = mce_shutdown, @@ -1659,26 +1696,70 @@ static ssize_t set_bank(struct sys_device *s, struct sysdev_attribute *attr, static ssize_t show_trigger(struct sys_device *s, struct sysdev_attribute *attr, char *buf) { - strcpy(buf, trigger); + strcpy(buf, mce_helper); strcat(buf, "\n"); - return strlen(trigger) + 1; + return strlen(mce_helper) + 1; } static ssize_t set_trigger(struct sys_device *s, struct sysdev_attribute *attr, const char *buf, size_t siz) { char *p; - int len; - strncpy(trigger, buf, sizeof(trigger)); - trigger[sizeof(trigger)-1] = 0; - len = strlen(trigger); - p = strchr(trigger, '\n'); + strncpy(mce_helper, buf, sizeof(mce_helper)); + mce_helper[sizeof(mce_helper)-1] = 0; + p = strchr(mce_helper, '\n'); - if (*p) + if (p) *p = 0; - return len; + return strlen(mce_helper) + !!p; +} + +static ssize_t set_ignore_ce(struct sys_device *s, + struct sysdev_attribute *attr, + const char *buf, size_t size) +{ + u64 new; + + if (strict_strtoull(buf, 0, &new) < 0) + return -EINVAL; + + if (mce_ignore_ce ^ !!new) { + if (new) { + /* disable ce features */ + on_each_cpu(mce_disable_ce, (void *)1, 1); + mce_ignore_ce = 1; + } else { + /* enable ce features */ + mce_ignore_ce = 0; + on_each_cpu(mce_enable_ce, (void *)1, 1); + } + } + return size; +} + +static ssize_t set_cmci_disabled(struct sys_device *s, + struct sysdev_attribute *attr, + const char *buf, size_t size) +{ + u64 new; + + if (strict_strtoull(buf, 0, &new) < 0) + return -EINVAL; + + if (mce_cmci_disabled ^ !!new) { + if (new) { + /* disable cmci */ + on_each_cpu(mce_disable_ce, NULL, 1); + mce_cmci_disabled = 1; + } else { + /* enable cmci */ + mce_cmci_disabled = 0; + on_each_cpu(mce_enable_ce, NULL, 1); + } + } + return size; } static ssize_t store_int_with_restart(struct sys_device *s, @@ -1693,6 +1774,7 @@ static ssize_t store_int_with_restart(struct sys_device *s, static SYSDEV_ATTR(trigger, 0644, show_trigger, set_trigger); static SYSDEV_INT_ATTR(tolerant, 0644, tolerant); static SYSDEV_INT_ATTR(monarch_timeout, 0644, monarch_timeout); +static SYSDEV_INT_ATTR(dont_log_ce, 0644, mce_dont_log_ce); static struct sysdev_ext_attribute attr_check_interval = { _SYSDEV_ATTR(check_interval, 0644, sysdev_show_int, @@ -1700,9 +1782,24 @@ static struct sysdev_ext_attribute attr_check_interval = { &check_interval }; +static struct sysdev_ext_attribute attr_ignore_ce = { + _SYSDEV_ATTR(ignore_ce, 0644, sysdev_show_int, set_ignore_ce), + &mce_ignore_ce +}; + +static struct sysdev_ext_attribute attr_cmci_disabled = { + _SYSDEV_ATTR(cmci_disabled, 0644, sysdev_show_int, set_cmci_disabled), + &mce_cmci_disabled +}; + static struct sysdev_attribute *mce_attrs[] = { - &attr_tolerant.attr, &attr_check_interval.attr, &attr_trigger, + &attr_tolerant.attr, + &attr_check_interval.attr, + &attr_trigger, &attr_monarch_timeout.attr, + &attr_dont_log_ce.attr, + &attr_ignore_ce.attr, + &attr_cmci_disabled.attr, NULL }; @@ -1712,7 +1809,7 @@ static cpumask_var_t mce_dev_initialized; static __cpuinit int mce_create_device(unsigned int cpu) { int err; - int i; + int i, j; if (!mce_available(&boot_cpu_data)) return -EIO; @@ -1730,9 +1827,9 @@ static __cpuinit int mce_create_device(unsigned int cpu) if (err) goto error; } - for (i = 0; i < banks; i++) { + for (j = 0; j < banks; j++) { err = sysdev_create_file(&per_cpu(mce_dev, cpu), - &bank_attrs[i]); + &bank_attrs[j]); if (err) goto error2; } @@ -1740,8 +1837,8 @@ static __cpuinit int mce_create_device(unsigned int cpu) return 0; error2: - while (--i >= 0) - sysdev_remove_file(&per_cpu(mce_dev, cpu), &bank_attrs[i]); + while (--j >= 0) + sysdev_remove_file(&per_cpu(mce_dev, cpu), &bank_attrs[j]); error: while (--i >= 0) sysdev_remove_file(&per_cpu(mce_dev, cpu), mce_attrs[i]); @@ -1883,7 +1980,7 @@ static __init int mce_init_device(void) if (!mce_available(&boot_cpu_data)) return -EIO; - alloc_cpumask_var(&mce_dev_initialized, GFP_KERNEL); + zalloc_cpumask_var(&mce_dev_initialized, GFP_KERNEL); err = mce_init_banks(); if (err) @@ -1915,7 +2012,7 @@ EXPORT_SYMBOL_GPL(nr_mce_banks); /* non-fatal.o */ /* This has to be run for each processor */ void mcheck_init(struct cpuinfo_x86 *c) { - if (mce_disabled == 1) + if (mce_disabled) return; switch (c->x86_vendor) { @@ -1945,10 +2042,9 @@ void mcheck_init(struct cpuinfo_x86 *c) static int __init mcheck_enable(char *str) { - mce_disabled = -1; + mce_p5_enabled = 1; return 1; } - __setup("mce", mcheck_enable); #endif /* CONFIG_X86_OLD_MCE */ diff --git a/arch/x86/kernel/cpu/mcheck/mce.h b/arch/x86/kernel/cpu/mcheck/mce.h deleted file mode 100644 index 84a552b458c..00000000000 --- a/arch/x86/kernel/cpu/mcheck/mce.h +++ /dev/null @@ -1,38 +0,0 @@ -#include <linux/init.h> -#include <asm/mce.h> - -#ifdef CONFIG_X86_OLD_MCE -void amd_mcheck_init(struct cpuinfo_x86 *c); -void intel_p4_mcheck_init(struct cpuinfo_x86 *c); -void intel_p6_mcheck_init(struct cpuinfo_x86 *c); -#endif - -#ifdef CONFIG_X86_ANCIENT_MCE -void intel_p5_mcheck_init(struct cpuinfo_x86 *c); -void winchip_mcheck_init(struct cpuinfo_x86 *c); -extern int mce_p5_enable; -static inline int mce_p5_enabled(void) { return mce_p5_enable; } -static inline void enable_p5_mce(void) { mce_p5_enable = 1; } -#else -static inline void intel_p5_mcheck_init(struct cpuinfo_x86 *c) {} -static inline void winchip_mcheck_init(struct cpuinfo_x86 *c) {} -static inline int mce_p5_enabled(void) { return 0; } -static inline void enable_p5_mce(void) { } -#endif - -/* Call the installed machine check handler for this CPU setup. */ -extern void (*machine_check_vector)(struct pt_regs *, long error_code); - -#ifdef CONFIG_X86_OLD_MCE - -extern int nr_mce_banks; - -void intel_set_thermal_handler(void); - -#else - -static inline void intel_set_thermal_handler(void) { } - -#endif - -void intel_init_thermal(struct cpuinfo_x86 *c); diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c index ddae21620bd..ddae21620bd 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_amd_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel.c b/arch/x86/kernel/cpu/mcheck/mce_intel.c index 2b011d2d857..e1acec0f7a3 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel.c @@ -1,74 +1,226 @@ /* - * Common code for Intel machine checks + * Intel specific MCE features. + * Copyright 2004 Zwane Mwaikambo <zwane@linuxpower.ca> + * Copyright (C) 2008, 2009 Intel Corporation + * Author: Andi Kleen */ -#include <linux/interrupt.h> -#include <linux/kernel.h> -#include <linux/types.h> -#include <linux/init.h> -#include <linux/smp.h> -#include <asm/therm_throt.h> -#include <asm/processor.h> -#include <asm/system.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/percpu.h> #include <asm/apic.h> +#include <asm/processor.h> #include <asm/msr.h> +#include <asm/mce.h> + +/* + * Support for Intel Correct Machine Check Interrupts. This allows + * the CPU to raise an interrupt when a corrected machine check happened. + * Normally we pick those up using a regular polling timer. + * Also supports reliable discovery of shared banks. + */ -#include "mce.h" +static DEFINE_PER_CPU(mce_banks_t, mce_banks_owned); -void intel_init_thermal(struct cpuinfo_x86 *c) +/* + * cmci_discover_lock protects against parallel discovery attempts + * which could race against each other. + */ +static DEFINE_SPINLOCK(cmci_discover_lock); + +#define CMCI_THRESHOLD 1 + +static int cmci_supported(int *banks) { - unsigned int cpu = smp_processor_id(); - int tm2 = 0; - u32 l, h; + u64 cap; - /* Thermal monitoring depends on ACPI and clock modulation*/ - if (!cpu_has(c, X86_FEATURE_ACPI) || !cpu_has(c, X86_FEATURE_ACC)) - return; + if (mce_cmci_disabled || mce_ignore_ce) + return 0; /* - * First check if its enabled already, in which case there might - * be some SMM goo which handles it, so we can't even put a handler - * since it might be delivered via SMI already: + * Vendor check is not strictly needed, but the initial + * initialization is vendor keyed and this + * makes sure none of the backdoors are entered otherwise. */ - rdmsr(MSR_IA32_MISC_ENABLE, l, h); - h = apic_read(APIC_LVTTHMR); - if ((l & MSR_IA32_MISC_ENABLE_TM1) && (h & APIC_DM_SMI)) { - printk(KERN_DEBUG - "CPU%d: Thermal monitoring handled by SMI\n", cpu); - return; + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) + return 0; + if (!cpu_has_apic || lapic_get_maxlvt() < 6) + return 0; + rdmsrl(MSR_IA32_MCG_CAP, cap); + *banks = min_t(unsigned, MAX_NR_BANKS, cap & 0xff); + return !!(cap & MCG_CMCI_P); +} + +/* + * The interrupt handler. This is called on every event. + * Just call the poller directly to log any events. + * This could in theory increase the threshold under high load, + * but doesn't for now. + */ +static void intel_threshold_interrupt(void) +{ + machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned)); + mce_notify_irq(); +} + +static void print_update(char *type, int *hdr, int num) +{ + if (*hdr == 0) + printk(KERN_INFO "CPU %d MCA banks", smp_processor_id()); + *hdr = 1; + printk(KERN_CONT " %s:%d", type, num); +} + +/* + * Enable CMCI (Corrected Machine Check Interrupt) for available MCE banks + * on this CPU. Use the algorithm recommended in the SDM to discover shared + * banks. + */ +static void cmci_discover(int banks, int boot) +{ + unsigned long *owned = (void *)&__get_cpu_var(mce_banks_owned); + unsigned long flags; + int hdr = 0; + int i; + + spin_lock_irqsave(&cmci_discover_lock, flags); + for (i = 0; i < banks; i++) { + u64 val; + + if (test_bit(i, owned)) + continue; + + rdmsrl(MSR_IA32_MC0_CTL2 + i, val); + + /* Already owned by someone else? */ + if (val & CMCI_EN) { + if (test_and_clear_bit(i, owned) || boot) + print_update("SHD", &hdr, i); + __clear_bit(i, __get_cpu_var(mce_poll_banks)); + continue; + } + + val |= CMCI_EN | CMCI_THRESHOLD; + wrmsrl(MSR_IA32_MC0_CTL2 + i, val); + rdmsrl(MSR_IA32_MC0_CTL2 + i, val); + + /* Did the enable bit stick? -- the bank supports CMCI */ + if (val & CMCI_EN) { + if (!test_and_set_bit(i, owned) || boot) + print_update("CMCI", &hdr, i); + __clear_bit(i, __get_cpu_var(mce_poll_banks)); + } else { + WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks))); + } } + spin_unlock_irqrestore(&cmci_discover_lock, flags); + if (hdr) + printk(KERN_CONT "\n"); +} + +/* + * Just in case we missed an event during initialization check + * all the CMCI owned banks. + */ +void cmci_recheck(void) +{ + unsigned long flags; + int banks; + + if (!mce_available(¤t_cpu_data) || !cmci_supported(&banks)) + return; + local_irq_save(flags); + machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned)); + local_irq_restore(flags); +} - if (cpu_has(c, X86_FEATURE_TM2) && (l & MSR_IA32_MISC_ENABLE_TM2)) - tm2 = 1; +/* + * Disable CMCI on this CPU for all banks it owns when it goes down. + * This allows other CPUs to claim the banks on rediscovery. + */ +void cmci_clear(void) +{ + unsigned long flags; + int i; + int banks; + u64 val; - /* Check whether a vector already exists */ - if (h & APIC_VECTOR_MASK) { - printk(KERN_DEBUG - "CPU%d: Thermal LVT vector (%#x) already installed\n", - cpu, (h & APIC_VECTOR_MASK)); + if (!cmci_supported(&banks)) return; + spin_lock_irqsave(&cmci_discover_lock, flags); + for (i = 0; i < banks; i++) { + if (!test_bit(i, __get_cpu_var(mce_banks_owned))) + continue; + /* Disable CMCI */ + rdmsrl(MSR_IA32_MC0_CTL2 + i, val); + val &= ~(CMCI_EN|CMCI_THRESHOLD_MASK); + wrmsrl(MSR_IA32_MC0_CTL2 + i, val); + __clear_bit(i, __get_cpu_var(mce_banks_owned)); } + spin_unlock_irqrestore(&cmci_discover_lock, flags); +} + +/* + * After a CPU went down cycle through all the others and rediscover + * Must run in process context. + */ +void cmci_rediscover(int dying) +{ + int banks; + int cpu; + cpumask_var_t old; + + if (!cmci_supported(&banks)) + return; + if (!alloc_cpumask_var(&old, GFP_KERNEL)) + return; + cpumask_copy(old, ¤t->cpus_allowed); - /* We'll mask the thermal vector in the lapic till we're ready: */ - h = THERMAL_APIC_VECTOR | APIC_DM_FIXED | APIC_LVT_MASKED; - apic_write(APIC_LVTTHMR, h); + for_each_online_cpu(cpu) { + if (cpu == dying) + continue; + if (set_cpus_allowed_ptr(current, cpumask_of(cpu))) + continue; + /* Recheck banks in case CPUs don't all have the same */ + if (cmci_supported(&banks)) + cmci_discover(banks, 0); + } - rdmsr(MSR_IA32_THERM_INTERRUPT, l, h); - wrmsr(MSR_IA32_THERM_INTERRUPT, - l | (THERM_INT_LOW_ENABLE | THERM_INT_HIGH_ENABLE), h); + set_cpus_allowed_ptr(current, old); + free_cpumask_var(old); +} - intel_set_thermal_handler(); +/* + * Reenable CMCI on this CPU in case a CPU down failed. + */ +void cmci_reenable(void) +{ + int banks; + if (cmci_supported(&banks)) + cmci_discover(banks, 0); +} - rdmsr(MSR_IA32_MISC_ENABLE, l, h); - wrmsr(MSR_IA32_MISC_ENABLE, l | MSR_IA32_MISC_ENABLE_TM1, h); +static void intel_init_cmci(void) +{ + int banks; - /* Unmask the thermal vector: */ - l = apic_read(APIC_LVTTHMR); - apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED); + if (!cmci_supported(&banks)) + return; - printk(KERN_INFO "CPU%d: Thermal monitoring enabled (%s)\n", - cpu, tm2 ? "TM2" : "TM1"); + mce_threshold_vector = intel_threshold_interrupt; + cmci_discover(banks, 1); + /* + * For CPU #0 this runs with still disabled APIC, but that's + * ok because only the vector is set up. We still do another + * check for the banks later for CPU #0 just to make sure + * to not miss any events. + */ + apic_write(APIC_LVTCMCI, THRESHOLD_APIC_VECTOR|APIC_DM_FIXED); + cmci_recheck(); +} - /* enable thermal throttle processing */ - atomic_set(&therm_throt_en, 1); +void mce_intel_feature_init(struct cpuinfo_x86 *c) +{ + intel_init_thermal(c); + intel_init_cmci(); } diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c b/arch/x86/kernel/cpu/mcheck/mce_intel_64.c deleted file mode 100644 index f2ef6952c40..00000000000 --- a/arch/x86/kernel/cpu/mcheck/mce_intel_64.c +++ /dev/null @@ -1,248 +0,0 @@ -/* - * Intel specific MCE features. - * Copyright 2004 Zwane Mwaikambo <zwane@linuxpower.ca> - * Copyright (C) 2008, 2009 Intel Corporation - * Author: Andi Kleen - */ - -#include <linux/init.h> -#include <linux/interrupt.h> -#include <linux/percpu.h> -#include <asm/processor.h> -#include <asm/apic.h> -#include <asm/msr.h> -#include <asm/mce.h> -#include <asm/hw_irq.h> -#include <asm/idle.h> -#include <asm/therm_throt.h> - -#include "mce.h" - -asmlinkage void smp_thermal_interrupt(void) -{ - __u64 msr_val; - - ack_APIC_irq(); - - exit_idle(); - irq_enter(); - - rdmsrl(MSR_IA32_THERM_STATUS, msr_val); - if (therm_throt_process(msr_val & THERM_STATUS_PROCHOT)) - mce_log_therm_throt_event(msr_val); - - inc_irq_stat(irq_thermal_count); - irq_exit(); -} - -/* - * Support for Intel Correct Machine Check Interrupts. This allows - * the CPU to raise an interrupt when a corrected machine check happened. - * Normally we pick those up using a regular polling timer. - * Also supports reliable discovery of shared banks. - */ - -static DEFINE_PER_CPU(mce_banks_t, mce_banks_owned); - -/* - * cmci_discover_lock protects against parallel discovery attempts - * which could race against each other. - */ -static DEFINE_SPINLOCK(cmci_discover_lock); - -#define CMCI_THRESHOLD 1 - -static int cmci_supported(int *banks) -{ - u64 cap; - - if (mce_cmci_disabled || mce_ignore_ce) - return 0; - - /* - * Vendor check is not strictly needed, but the initial - * initialization is vendor keyed and this - * makes sure none of the backdoors are entered otherwise. - */ - if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) - return 0; - if (!cpu_has_apic || lapic_get_maxlvt() < 6) - return 0; - rdmsrl(MSR_IA32_MCG_CAP, cap); - *banks = min_t(unsigned, MAX_NR_BANKS, cap & 0xff); - return !!(cap & MCG_CMCI_P); -} - -/* - * The interrupt handler. This is called on every event. - * Just call the poller directly to log any events. - * This could in theory increase the threshold under high load, - * but doesn't for now. - */ -static void intel_threshold_interrupt(void) -{ - machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned)); - mce_notify_irq(); -} - -static void print_update(char *type, int *hdr, int num) -{ - if (*hdr == 0) - printk(KERN_INFO "CPU %d MCA banks", smp_processor_id()); - *hdr = 1; - printk(KERN_CONT " %s:%d", type, num); -} - -/* - * Enable CMCI (Corrected Machine Check Interrupt) for available MCE banks - * on this CPU. Use the algorithm recommended in the SDM to discover shared - * banks. - */ -static void cmci_discover(int banks, int boot) -{ - unsigned long *owned = (void *)&__get_cpu_var(mce_banks_owned); - unsigned long flags; - int hdr = 0; - int i; - - spin_lock_irqsave(&cmci_discover_lock, flags); - for (i = 0; i < banks; i++) { - u64 val; - - if (test_bit(i, owned)) - continue; - - rdmsrl(MSR_IA32_MC0_CTL2 + i, val); - - /* Already owned by someone else? */ - if (val & CMCI_EN) { - if (test_and_clear_bit(i, owned) || boot) - print_update("SHD", &hdr, i); - __clear_bit(i, __get_cpu_var(mce_poll_banks)); - continue; - } - - val |= CMCI_EN | CMCI_THRESHOLD; - wrmsrl(MSR_IA32_MC0_CTL2 + i, val); - rdmsrl(MSR_IA32_MC0_CTL2 + i, val); - - /* Did the enable bit stick? -- the bank supports CMCI */ - if (val & CMCI_EN) { - if (!test_and_set_bit(i, owned) || boot) - print_update("CMCI", &hdr, i); - __clear_bit(i, __get_cpu_var(mce_poll_banks)); - } else { - WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks))); - } - } - spin_unlock_irqrestore(&cmci_discover_lock, flags); - if (hdr) - printk(KERN_CONT "\n"); -} - -/* - * Just in case we missed an event during initialization check - * all the CMCI owned banks. - */ -void cmci_recheck(void) -{ - unsigned long flags; - int banks; - - if (!mce_available(¤t_cpu_data) || !cmci_supported(&banks)) - return; - local_irq_save(flags); - machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned)); - local_irq_restore(flags); -} - -/* - * Disable CMCI on this CPU for all banks it owns when it goes down. - * This allows other CPUs to claim the banks on rediscovery. - */ -void cmci_clear(void) -{ - unsigned long flags; - int i; - int banks; - u64 val; - - if (!cmci_supported(&banks)) - return; - spin_lock_irqsave(&cmci_discover_lock, flags); - for (i = 0; i < banks; i++) { - if (!test_bit(i, __get_cpu_var(mce_banks_owned))) - continue; - /* Disable CMCI */ - rdmsrl(MSR_IA32_MC0_CTL2 + i, val); - val &= ~(CMCI_EN|CMCI_THRESHOLD_MASK); - wrmsrl(MSR_IA32_MC0_CTL2 + i, val); - __clear_bit(i, __get_cpu_var(mce_banks_owned)); - } - spin_unlock_irqrestore(&cmci_discover_lock, flags); -} - -/* - * After a CPU went down cycle through all the others and rediscover - * Must run in process context. - */ -void cmci_rediscover(int dying) -{ - int banks; - int cpu; - cpumask_var_t old; - - if (!cmci_supported(&banks)) - return; - if (!alloc_cpumask_var(&old, GFP_KERNEL)) - return; - cpumask_copy(old, ¤t->cpus_allowed); - - for_each_online_cpu(cpu) { - if (cpu == dying) - continue; - if (set_cpus_allowed_ptr(current, cpumask_of(cpu))) - continue; - /* Recheck banks in case CPUs don't all have the same */ - if (cmci_supported(&banks)) - cmci_discover(banks, 0); - } - - set_cpus_allowed_ptr(current, old); - free_cpumask_var(old); -} - -/* - * Reenable CMCI on this CPU in case a CPU down failed. - */ -void cmci_reenable(void) -{ - int banks; - if (cmci_supported(&banks)) - cmci_discover(banks, 0); -} - -static void intel_init_cmci(void) -{ - int banks; - - if (!cmci_supported(&banks)) - return; - - mce_threshold_vector = intel_threshold_interrupt; - cmci_discover(banks, 1); - /* - * For CPU #0 this runs with still disabled APIC, but that's - * ok because only the vector is set up. We still do another - * check for the banks later for CPU #0 just to make sure - * to not miss any events. - */ - apic_write(APIC_LVTCMCI, THRESHOLD_APIC_VECTOR|APIC_DM_FIXED); - cmci_recheck(); -} - -void mce_intel_feature_init(struct cpuinfo_x86 *c) -{ - intel_init_thermal(c); - intel_init_cmci(); -} diff --git a/arch/x86/kernel/cpu/mcheck/non-fatal.c b/arch/x86/kernel/cpu/mcheck/non-fatal.c index 70b710420f7..f5f2d6f71fb 100644 --- a/arch/x86/kernel/cpu/mcheck/non-fatal.c +++ b/arch/x86/kernel/cpu/mcheck/non-fatal.c @@ -17,10 +17,9 @@ #include <asm/processor.h> #include <asm/system.h> +#include <asm/mce.h> #include <asm/msr.h> -#include "mce.h" - static int firstbank; #define MCE_RATE (15*HZ) /* timer rate is 15s */ diff --git a/arch/x86/kernel/cpu/mcheck/p4.c b/arch/x86/kernel/cpu/mcheck/p4.c index 82cee108a2d..4482aea9aa2 100644 --- a/arch/x86/kernel/cpu/mcheck/p4.c +++ b/arch/x86/kernel/cpu/mcheck/p4.c @@ -1,21 +1,15 @@ /* * P4 specific Machine Check Exception Reporting */ - -#include <linux/interrupt.h> #include <linux/kernel.h> #include <linux/types.h> #include <linux/init.h> #include <linux/smp.h> -#include <asm/therm_throt.h> #include <asm/processor.h> -#include <asm/system.h> -#include <asm/apic.h> +#include <asm/mce.h> #include <asm/msr.h> -#include "mce.h" - /* as supported by the P4/Xeon family */ struct intel_mce_extended_msrs { u32 eax; @@ -33,46 +27,6 @@ struct intel_mce_extended_msrs { static int mce_num_extended_msrs; - -#ifdef CONFIG_X86_MCE_P4THERMAL - -static void unexpected_thermal_interrupt(struct pt_regs *regs) -{ - printk(KERN_ERR "CPU%d: Unexpected LVT TMR interrupt!\n", - smp_processor_id()); - add_taint(TAINT_MACHINE_CHECK); -} - -/* P4/Xeon Thermal transition interrupt handler: */ -static void intel_thermal_interrupt(struct pt_regs *regs) -{ - __u64 msr_val; - - ack_APIC_irq(); - - rdmsrl(MSR_IA32_THERM_STATUS, msr_val); - therm_throt_process(msr_val & THERM_STATUS_PROCHOT); -} - -/* Thermal interrupt handler for this CPU setup: */ -static void (*vendor_thermal_interrupt)(struct pt_regs *regs) = - unexpected_thermal_interrupt; - -void smp_thermal_interrupt(struct pt_regs *regs) -{ - irq_enter(); - vendor_thermal_interrupt(regs); - __get_cpu_var(irq_stat).irq_thermal_count++; - irq_exit(); -} - -void intel_set_thermal_handler(void) -{ - vendor_thermal_interrupt = intel_thermal_interrupt; -} - -#endif /* CONFIG_X86_MCE_P4THERMAL */ - /* P4/Xeon Extended MCE MSR retrieval, return 0 if unsupported */ static void intel_get_extended_msrs(struct intel_mce_extended_msrs *r) { diff --git a/arch/x86/kernel/cpu/mcheck/p5.c b/arch/x86/kernel/cpu/mcheck/p5.c index 015f481ab1b..5c0e6533d9b 100644 --- a/arch/x86/kernel/cpu/mcheck/p5.c +++ b/arch/x86/kernel/cpu/mcheck/p5.c @@ -10,12 +10,11 @@ #include <asm/processor.h> #include <asm/system.h> +#include <asm/mce.h> #include <asm/msr.h> -#include "mce.h" - /* By default disabled */ -int mce_p5_enable; +int mce_p5_enabled __read_mostly; /* Machine check handler for Pentium class Intel CPUs: */ static void pentium_machine_check(struct pt_regs *regs, long error_code) @@ -43,15 +42,13 @@ void intel_p5_mcheck_init(struct cpuinfo_x86 *c) { u32 l, h; - /* Check for MCE support: */ - if (!cpu_has(c, X86_FEATURE_MCE)) + /* Default P5 to off as its often misconnected: */ + if (!mce_p5_enabled) return; -#ifdef CONFIG_X86_OLD_MCE - /* Default P5 to off as its often misconnected: */ - if (mce_disabled != -1) + /* Check for MCE support: */ + if (!cpu_has(c, X86_FEATURE_MCE)) return; -#endif machine_check_vector = pentium_machine_check; /* Make sure the vector pointer is visible before we enable MCEs: */ diff --git a/arch/x86/kernel/cpu/mcheck/p6.c b/arch/x86/kernel/cpu/mcheck/p6.c index 43c24e66745..01e4f817818 100644 --- a/arch/x86/kernel/cpu/mcheck/p6.c +++ b/arch/x86/kernel/cpu/mcheck/p6.c @@ -10,10 +10,9 @@ #include <asm/processor.h> #include <asm/system.h> +#include <asm/mce.h> #include <asm/msr.h> -#include "mce.h" - /* Machine Check Handler For PII/PIII */ static void intel_machine_check(struct pt_regs *regs, long error_code) { diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c index 7b1ae2e20ba..5957a93e517 100644 --- a/arch/x86/kernel/cpu/mcheck/therm_throt.c +++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c @@ -13,21 +13,32 @@ * Credits: Adapted from Zwane Mwaikambo's original code in mce_intel.c. * Inspired by Ross Biro's and Al Borchers' counter code. */ +#include <linux/interrupt.h> #include <linux/notifier.h> #include <linux/jiffies.h> +#include <linux/kernel.h> #include <linux/percpu.h> #include <linux/sysdev.h> +#include <linux/types.h> +#include <linux/init.h> +#include <linux/smp.h> #include <linux/cpu.h> -#include <asm/therm_throt.h> +#include <asm/processor.h> +#include <asm/system.h> +#include <asm/apic.h> +#include <asm/idle.h> +#include <asm/mce.h> +#include <asm/msr.h> /* How long to wait between reporting thermal events */ #define CHECK_INTERVAL (300 * HZ) static DEFINE_PER_CPU(__u64, next_check) = INITIAL_JIFFIES; static DEFINE_PER_CPU(unsigned long, thermal_throttle_count); +static DEFINE_PER_CPU(bool, thermal_throttle_active); -atomic_t therm_throt_en = ATOMIC_INIT(0); +static atomic_t therm_throt_en = ATOMIC_INIT(0); #ifdef CONFIG_SYSFS #define define_therm_throt_sysdev_one_ro(_name) \ @@ -82,31 +93,37 @@ static struct attribute_group thermal_throttle_attr_group = { * 1 : Event should be logged further, and a message has been * printed to the syslog. */ -int therm_throt_process(int curr) +static int therm_throt_process(int curr) { unsigned int cpu = smp_processor_id(); __u64 tmp_jiffs = get_jiffies_64(); + bool was_throttled = __get_cpu_var(thermal_throttle_active); + bool is_throttled = __get_cpu_var(thermal_throttle_active) = curr; - if (curr) + if (is_throttled) __get_cpu_var(thermal_throttle_count)++; - if (time_before64(tmp_jiffs, __get_cpu_var(next_check))) + if (!(was_throttled ^ is_throttled) && + time_before64(tmp_jiffs, __get_cpu_var(next_check))) return 0; __get_cpu_var(next_check) = tmp_jiffs + CHECK_INTERVAL; /* if we just entered the thermal event */ - if (curr) { + if (is_throttled) { printk(KERN_CRIT "CPU%d: Temperature above threshold, " - "cpu clock throttled (total events = %lu)\n", cpu, - __get_cpu_var(thermal_throttle_count)); + "cpu clock throttled (total events = %lu)\n", + cpu, __get_cpu_var(thermal_throttle_count)); add_taint(TAINT_MACHINE_CHECK); - } else { - printk(KERN_CRIT "CPU%d: Temperature/speed normal\n", cpu); + return 1; + } + if (was_throttled) { + printk(KERN_INFO "CPU%d: Temperature/speed normal\n", cpu); + return 1; } - return 1; + return 0; } #ifdef CONFIG_SYSFS @@ -186,6 +203,94 @@ static __init int thermal_throttle_init_device(void) return 0; } - device_initcall(thermal_throttle_init_device); + #endif /* CONFIG_SYSFS */ + +/* Thermal transition interrupt handler */ +static void intel_thermal_interrupt(void) +{ + __u64 msr_val; + + rdmsrl(MSR_IA32_THERM_STATUS, msr_val); + if (therm_throt_process(msr_val & THERM_STATUS_PROCHOT)) + mce_log_therm_throt_event(msr_val); +} + +static void unexpected_thermal_interrupt(void) +{ + printk(KERN_ERR "CPU%d: Unexpected LVT TMR interrupt!\n", + smp_processor_id()); + add_taint(TAINT_MACHINE_CHECK); +} + +static void (*smp_thermal_vector)(void) = unexpected_thermal_interrupt; + +asmlinkage void smp_thermal_interrupt(struct pt_regs *regs) +{ + exit_idle(); + irq_enter(); + inc_irq_stat(irq_thermal_count); + smp_thermal_vector(); + irq_exit(); + /* Ack only at the end to avoid potential reentry */ + ack_APIC_irq(); +} + +void intel_init_thermal(struct cpuinfo_x86 *c) +{ + unsigned int cpu = smp_processor_id(); + int tm2 = 0; + u32 l, h; + + /* Thermal monitoring depends on ACPI and clock modulation*/ + if (!cpu_has(c, X86_FEATURE_ACPI) || !cpu_has(c, X86_FEATURE_ACC)) + return; + + /* + * First check if its enabled already, in which case there might + * be some SMM goo which handles it, so we can't even put a handler + * since it might be delivered via SMI already: + */ + rdmsr(MSR_IA32_MISC_ENABLE, l, h); + h = apic_read(APIC_LVTTHMR); + if ((l & MSR_IA32_MISC_ENABLE_TM1) && (h & APIC_DM_SMI)) { + printk(KERN_DEBUG + "CPU%d: Thermal monitoring handled by SMI\n", cpu); + return; + } + + if (cpu_has(c, X86_FEATURE_TM2) && (l & MSR_IA32_MISC_ENABLE_TM2)) + tm2 = 1; + + /* Check whether a vector already exists */ + if (h & APIC_VECTOR_MASK) { + printk(KERN_DEBUG + "CPU%d: Thermal LVT vector (%#x) already installed\n", + cpu, (h & APIC_VECTOR_MASK)); + return; + } + + /* We'll mask the thermal vector in the lapic till we're ready: */ + h = THERMAL_APIC_VECTOR | APIC_DM_FIXED | APIC_LVT_MASKED; + apic_write(APIC_LVTTHMR, h); + + rdmsr(MSR_IA32_THERM_INTERRUPT, l, h); + wrmsr(MSR_IA32_THERM_INTERRUPT, + l | (THERM_INT_LOW_ENABLE | THERM_INT_HIGH_ENABLE), h); + + smp_thermal_vector = intel_thermal_interrupt; + + rdmsr(MSR_IA32_MISC_ENABLE, l, h); + wrmsr(MSR_IA32_MISC_ENABLE, l | MSR_IA32_MISC_ENABLE_TM1, h); + + /* Unmask the thermal vector: */ + l = apic_read(APIC_LVTTHMR); + apic_write(APIC_LVTTHMR, l & ~APIC_LVT_MASKED); + + printk(KERN_INFO "CPU%d: Thermal monitoring enabled (%s)\n", + cpu, tm2 ? "TM2" : "TM1"); + + /* enable thermal throttle processing */ + atomic_set(&therm_throt_en, 1); +} diff --git a/arch/x86/kernel/cpu/mcheck/winchip.c b/arch/x86/kernel/cpu/mcheck/winchip.c index 81b02487090..54060f56597 100644 --- a/arch/x86/kernel/cpu/mcheck/winchip.c +++ b/arch/x86/kernel/cpu/mcheck/winchip.c @@ -9,10 +9,9 @@ #include <asm/processor.h> #include <asm/system.h> +#include <asm/mce.h> #include <asm/msr.h> -#include "mce.h" - /* Machine check handler for WinChip C6: */ static void winchip_machine_check(struct pt_regs *regs, long error_code) { diff --git a/arch/x86/kernel/cpu/perf_counter.c b/arch/x86/kernel/cpu/perf_counter.c index 275bc142cd5..900332b800f 100644 --- a/arch/x86/kernel/cpu/perf_counter.c +++ b/arch/x86/kernel/cpu/perf_counter.c @@ -19,6 +19,7 @@ #include <linux/kdebug.h> #include <linux/sched.h> #include <linux/uaccess.h> +#include <linux/highmem.h> #include <asm/apic.h> #include <asm/stacktrace.h> @@ -54,6 +55,7 @@ struct x86_pmu { int num_counters_fixed; int counter_bits; u64 counter_mask; + int apic; u64 max_period; u64 intel_ctrl; }; @@ -65,6 +67,52 @@ static DEFINE_PER_CPU(struct cpu_hw_counters, cpu_hw_counters) = { }; /* + * Not sure about some of these + */ +static const u64 p6_perfmon_event_map[] = +{ + [PERF_COUNT_HW_CPU_CYCLES] = 0x0079, + [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x0f2e, + [PERF_COUNT_HW_CACHE_MISSES] = 0x012e, + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c4, + [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c5, + [PERF_COUNT_HW_BUS_CYCLES] = 0x0062, +}; + +static u64 p6_pmu_event_map(int event) +{ + return p6_perfmon_event_map[event]; +} + +/* + * Counter setting that is specified not to count anything. + * We use this to effectively disable a counter. + * + * L2_RQSTS with 0 MESI unit mask. + */ +#define P6_NOP_COUNTER 0x0000002EULL + +static u64 p6_pmu_raw_event(u64 event) +{ +#define P6_EVNTSEL_EVENT_MASK 0x000000FFULL +#define P6_EVNTSEL_UNIT_MASK 0x0000FF00ULL +#define P6_EVNTSEL_EDGE_MASK 0x00040000ULL +#define P6_EVNTSEL_INV_MASK 0x00800000ULL +#define P6_EVNTSEL_COUNTER_MASK 0xFF000000ULL + +#define P6_EVNTSEL_MASK \ + (P6_EVNTSEL_EVENT_MASK | \ + P6_EVNTSEL_UNIT_MASK | \ + P6_EVNTSEL_EDGE_MASK | \ + P6_EVNTSEL_INV_MASK | \ + P6_EVNTSEL_COUNTER_MASK) + + return event & P6_EVNTSEL_MASK; +} + + +/* * Intel PerfMon v3. Used on Core2 and later. */ static const u64 intel_perfmon_event_map[] = @@ -389,23 +437,23 @@ static u64 intel_pmu_raw_event(u64 event) return event & CORE_EVNTSEL_MASK; } -static const u64 amd_0f_hw_cache_event_ids +static const u64 amd_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(RESULT_ACCESS) ] = 0x0040, /* Data Cache Accesses */ + [ C(RESULT_MISS) ] = 0x0041, /* Data Cache Misses */ }, [ C(OP_WRITE) ] = { - [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_ACCESS) ] = 0x0142, /* Data Cache Refills :system */ [ C(RESULT_MISS) ] = 0, }, [ C(OP_PREFETCH) ] = { - [ C(RESULT_ACCESS) ] = 0, - [ C(RESULT_MISS) ] = 0, + [ C(RESULT_ACCESS) ] = 0x0267, /* Data Prefetcher :attempts */ + [ C(RESULT_MISS) ] = 0x0167, /* Data Prefetcher :cancelled */ }, }, [ C(L1I ) ] = { @@ -418,17 +466,17 @@ static const u64 amd_0f_hw_cache_event_ids [ C(RESULT_MISS) ] = -1, }, [ C(OP_PREFETCH) ] = { - [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_ACCESS) ] = 0x014B, /* Prefetch Instructions :Load */ [ C(RESULT_MISS) ] = 0, }, }, [ C(LL ) ] = { [ C(OP_READ) ] = { - [ C(RESULT_ACCESS) ] = 0, - [ C(RESULT_MISS) ] = 0, + [ C(RESULT_ACCESS) ] = 0x037D, /* Requests to L2 Cache :IC+DC */ + [ C(RESULT_MISS) ] = 0x037E, /* L2 Cache Misses : IC+DC */ }, [ C(OP_WRITE) ] = { - [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_ACCESS) ] = 0x017F, /* L2 Fill/Writeback */ [ C(RESULT_MISS) ] = 0, }, [ C(OP_PREFETCH) ] = { @@ -438,8 +486,8 @@ static const u64 amd_0f_hw_cache_event_ids }, [ C(DTLB) ] = { [ C(OP_READ) ] = { - [ C(RESULT_ACCESS) ] = 0, - [ C(RESULT_MISS) ] = 0, + [ C(RESULT_ACCESS) ] = 0x0040, /* Data Cache Accesses */ + [ C(RESULT_MISS) ] = 0x0046, /* L1 DTLB and L2 DLTB Miss */ }, [ C(OP_WRITE) ] = { [ C(RESULT_ACCESS) ] = 0, @@ -566,6 +614,7 @@ static DEFINE_MUTEX(pmc_reserve_mutex); static bool reserve_pmc_hardware(void) { +#ifdef CONFIG_X86_LOCAL_APIC int i; if (nmi_watchdog == NMI_LOCAL_APIC) @@ -580,9 +629,11 @@ static bool reserve_pmc_hardware(void) if (!reserve_evntsel_nmi(x86_pmu.eventsel + i)) goto eventsel_fail; } +#endif return true; +#ifdef CONFIG_X86_LOCAL_APIC eventsel_fail: for (i--; i >= 0; i--) release_evntsel_nmi(x86_pmu.eventsel + i); @@ -597,10 +648,12 @@ perfctr_fail: enable_lapic_nmi_watchdog(); return false; +#endif } static void release_pmc_hardware(void) { +#ifdef CONFIG_X86_LOCAL_APIC int i; for (i = 0; i < x86_pmu.num_counters; i++) { @@ -610,6 +663,7 @@ static void release_pmc_hardware(void) if (nmi_watchdog == NMI_LOCAL_APIC) enable_lapic_nmi_watchdog(); +#endif } static void hw_perf_counter_destroy(struct perf_counter *counter) @@ -665,6 +719,7 @@ static int __hw_perf_counter_init(struct perf_counter *counter) { struct perf_counter_attr *attr = &counter->attr; struct hw_perf_counter *hwc = &counter->hw; + u64 config; int err; if (!x86_pmu_initialized()) @@ -700,6 +755,15 @@ static int __hw_perf_counter_init(struct perf_counter *counter) hwc->sample_period = x86_pmu.max_period; hwc->last_period = hwc->sample_period; atomic64_set(&hwc->period_left, hwc->sample_period); + } else { + /* + * If we have a PMU initialized but no APIC + * interrupts, we cannot sample hardware + * counters (user-space has to fall back and + * sample via a hrtimer based software counter): + */ + if (!x86_pmu.apic) + return -EOPNOTSUPP; } counter->destroy = hw_perf_counter_destroy; @@ -717,14 +781,40 @@ static int __hw_perf_counter_init(struct perf_counter *counter) if (attr->config >= x86_pmu.max_events) return -EINVAL; + /* * The generic map: */ - hwc->config |= x86_pmu.event_map(attr->config); + config = x86_pmu.event_map(attr->config); + + if (config == 0) + return -ENOENT; + + if (config == -1LL) + return -EINVAL; + + hwc->config |= config; return 0; } +static void p6_pmu_disable_all(void) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + u64 val; + + if (!cpuc->enabled) + return; + + cpuc->enabled = 0; + barrier(); + + /* p6 only has one enable register */ + rdmsrl(MSR_P6_EVNTSEL0, val); + val &= ~ARCH_PERFMON_EVENTSEL0_ENABLE; + wrmsrl(MSR_P6_EVNTSEL0, val); +} + static void intel_pmu_disable_all(void) { wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0); @@ -766,6 +856,23 @@ void hw_perf_disable(void) return x86_pmu.disable_all(); } +static void p6_pmu_enable_all(void) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + unsigned long val; + + if (cpuc->enabled) + return; + + cpuc->enabled = 1; + barrier(); + + /* p6 only has one enable register */ + rdmsrl(MSR_P6_EVNTSEL0, val); + val |= ARCH_PERFMON_EVENTSEL0_ENABLE; + wrmsrl(MSR_P6_EVNTSEL0, val); +} + static void intel_pmu_enable_all(void) { wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, x86_pmu.intel_ctrl); @@ -783,13 +890,13 @@ static void amd_pmu_enable_all(void) barrier(); for (idx = 0; idx < x86_pmu.num_counters; idx++) { + struct perf_counter *counter = cpuc->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 = counter->hw.config; val |= ARCH_PERFMON_EVENTSEL0_ENABLE; wrmsrl(MSR_K7_EVNTSEL0 + idx, val); } @@ -818,16 +925,13 @@ static inline void intel_pmu_ack_status(u64 ack) static inline void x86_pmu_enable_counter(struct hw_perf_counter *hwc, int idx) { - int err; - err = checking_wrmsrl(hwc->config_base + idx, + (void)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); + (void)checking_wrmsrl(hwc->config_base + idx, hwc->config); } static inline void @@ -835,13 +939,24 @@ 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); + (void)checking_wrmsrl(hwc->config_base, ctrl_val); +} + +static inline void +p6_pmu_disable_counter(struct hw_perf_counter *hwc, int idx) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + u64 val = P6_NOP_COUNTER; + + if (cpuc->enabled) + val |= ARCH_PERFMON_EVENTSEL0_ENABLE; + + (void)checking_wrmsrl(hwc->config_base + idx, val); } static inline void @@ -911,6 +1026,8 @@ x86_perf_counter_set_period(struct perf_counter *counter, err = checking_wrmsrl(hwc->counter_base + idx, (u64)(-left) & x86_pmu.counter_mask); + perf_counter_update_userpage(counter); + return ret; } @@ -940,6 +1057,19 @@ intel_pmu_enable_fixed(struct hw_perf_counter *hwc, int __idx) err = checking_wrmsrl(hwc->config_base, ctrl_val); } +static void p6_pmu_enable_counter(struct hw_perf_counter *hwc, int idx) +{ + struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters); + u64 val; + + val = hwc->config; + if (cpuc->enabled) + val |= ARCH_PERFMON_EVENTSEL0_ENABLE; + + (void)checking_wrmsrl(hwc->config_base + idx, 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)) { @@ -956,8 +1086,6 @@ static void amd_pmu_enable_counter(struct hw_perf_counter *hwc, int idx) if (cpuc->enabled) x86_pmu_enable_counter(hwc, idx); - else - x86_pmu_disable_counter(hwc, idx); } static int @@ -968,13 +1096,6 @@ fixed_mode_idx(struct perf_counter *counter, struct hw_perf_counter *hwc) if (!x86_pmu.num_counters_fixed) return -1; - /* - * Quirk, IA32_FIXED_CTRs do not work on current Atom processors: - */ - if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && - boot_cpu_data.x86_model == 28) - return -1; - event = hwc->config & ARCH_PERFMON_EVENT_MASK; if (unlikely(event == x86_pmu.event_map(PERF_COUNT_HW_INSTRUCTIONS))) @@ -1040,6 +1161,8 @@ try_generic: x86_perf_counter_set_period(counter, hwc, idx); x86_pmu.enable(hwc, idx); + perf_counter_update_userpage(counter); + return 0; } @@ -1132,6 +1255,8 @@ static void x86_pmu_disable(struct perf_counter *counter) x86_perf_counter_update(counter, hwc, idx); cpuc->counters[idx] = NULL; clear_bit(idx, cpuc->used_mask); + + perf_counter_update_userpage(counter); } /* @@ -1176,6 +1301,49 @@ static void intel_pmu_reset(void) local_irq_restore(flags); } +static int p6_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 idx, handled = 0; + u64 val; + + data.regs = regs; + data.addr = 0; + + cpuc = &__get_cpu_var(cpu_hw_counters); + + 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)) + p6_pmu_disable_counter(hwc, idx); + } + + if (handled) + inc_irq_stat(apic_perf_irqs); + + return handled; +} /* * This handler is triggered by the local APIC, so the APIC IRQ handling @@ -1185,14 +1353,13 @@ static int intel_pmu_handle_irq(struct pt_regs *regs) { struct perf_sample_data data; struct cpu_hw_counters *cpuc; - int bit, cpu, loops; + int bit, loops; u64 ack, status; data.regs = regs; data.addr = 0; - cpu = smp_processor_id(); - cpuc = &per_cpu(cpu_hw_counters, cpu); + cpuc = &__get_cpu_var(cpu_hw_counters); perf_disable(); status = intel_pmu_get_status(); @@ -1223,6 +1390,8 @@ again: if (!intel_pmu_save_and_restart(counter)) continue; + data.period = counter->hw.last_period; + if (perf_counter_overflow(counter, 1, &data)) intel_pmu_disable_counter(&counter->hw, bit); } @@ -1247,14 +1416,13 @@ static int amd_pmu_handle_irq(struct pt_regs *regs) struct cpu_hw_counters *cpuc; struct perf_counter *counter; struct hw_perf_counter *hwc; - int cpu, idx, handled = 0; + int idx, handled = 0; u64 val; data.regs = regs; data.addr = 0; - cpu = smp_processor_id(); - cpuc = &per_cpu(cpu_hw_counters, cpu); + cpuc = &__get_cpu_var(cpu_hw_counters); for (idx = 0; idx < x86_pmu.num_counters; idx++) { if (!test_bit(idx, cpuc->active_mask)) @@ -1297,18 +1465,22 @@ void smp_perf_pending_interrupt(struct pt_regs *regs) void set_perf_counter_pending(void) { +#ifdef CONFIG_X86_LOCAL_APIC apic->send_IPI_self(LOCAL_PENDING_VECTOR); +#endif } void perf_counters_lapic_init(void) { - if (!x86_pmu_initialized()) +#ifdef CONFIG_X86_LOCAL_APIC + if (!x86_pmu.apic || !x86_pmu_initialized()) return; /* * Always use NMI for PMU */ apic_write(APIC_LVTPC, APIC_DM_NMI); +#endif } static int __kprobes @@ -1332,7 +1504,9 @@ perf_counter_nmi_handler(struct notifier_block *self, regs = args->regs; +#ifdef CONFIG_X86_LOCAL_APIC apic_write(APIC_LVTPC, APIC_DM_NMI); +#endif /* * 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. @@ -1351,6 +1525,33 @@ static __read_mostly struct notifier_block perf_counter_nmi_notifier = { .priority = 1 }; +static struct x86_pmu p6_pmu = { + .name = "p6", + .handle_irq = p6_pmu_handle_irq, + .disable_all = p6_pmu_disable_all, + .enable_all = p6_pmu_enable_all, + .enable = p6_pmu_enable_counter, + .disable = p6_pmu_disable_counter, + .eventsel = MSR_P6_EVNTSEL0, + .perfctr = MSR_P6_PERFCTR0, + .event_map = p6_pmu_event_map, + .raw_event = p6_pmu_raw_event, + .max_events = ARRAY_SIZE(p6_perfmon_event_map), + .apic = 1, + .max_period = (1ULL << 31) - 1, + .version = 0, + .num_counters = 2, + /* + * Counters have 40 bits implemented. However they are designed such + * that bits [32-39] are sign extensions of bit 31. As such the + * effective width of a counter for P6-like PMU is 32 bits only. + * + * See IA-32 Intel Architecture Software developer manual Vol 3B + */ + .counter_bits = 32, + .counter_mask = (1ULL << 32) - 1, +}; + static struct x86_pmu intel_pmu = { .name = "Intel", .handle_irq = intel_pmu_handle_irq, @@ -1363,6 +1564,7 @@ static struct x86_pmu intel_pmu = { .event_map = intel_pmu_event_map, .raw_event = intel_pmu_raw_event, .max_events = ARRAY_SIZE(intel_perfmon_event_map), + .apic = 1, /* * Intel PMCs cannot be accessed sanely above 32 bit width, * so we install an artificial 1<<31 period regardless of @@ -1386,10 +1588,43 @@ static struct x86_pmu amd_pmu = { .num_counters = 4, .counter_bits = 48, .counter_mask = (1ULL << 48) - 1, + .apic = 1, /* use highest bit to detect overflow */ .max_period = (1ULL << 47) - 1, }; +static int p6_pmu_init(void) +{ + switch (boot_cpu_data.x86_model) { + case 1: + case 3: /* Pentium Pro */ + case 5: + case 6: /* Pentium II */ + case 7: + case 8: + case 11: /* Pentium III */ + break; + case 9: + case 13: + /* Pentium M */ + break; + default: + pr_cont("unsupported p6 CPU model %d ", + boot_cpu_data.x86_model); + return -ENODEV; + } + + x86_pmu = p6_pmu; + + if (!cpu_has_apic) { + pr_info("no APIC, boot with the \"lapic\" boot parameter to force-enable it.\n"); + pr_info("no hardware sampling interrupt available.\n"); + x86_pmu.apic = 0; + } + + return 0; +} + static int intel_pmu_init(void) { union cpuid10_edx edx; @@ -1398,8 +1633,14 @@ static int intel_pmu_init(void) unsigned int ebx; int version; - if (!cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) + if (!cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) { + /* check for P6 processor family */ + if (boot_cpu_data.x86 == 6) { + return p6_pmu_init(); + } else { return -ENODEV; + } + } /* * Check whether the Architectural PerfMon supports @@ -1425,8 +1666,6 @@ static int intel_pmu_init(void) */ 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: */ @@ -1459,18 +1698,16 @@ static int intel_pmu_init(void) static int amd_pmu_init(void) { + /* Performance-monitoring supported from K7 and later: */ + if (boot_cpu_data.x86 < 6) + return -ENODEV; + 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)); + /* Events are common for all AMDs */ + memcpy(hw_cache_event_ids, amd_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); - pr_cont("AMD Family 0f/10/11 events, "); - break; - } return 0; } @@ -1498,21 +1735,22 @@ void __init init_hw_perf_counters(void) 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); + 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); + x86_pmu.num_counters_fixed = X86_PMC_MAX_FIXED; } perf_counter_mask |= ((1LL << x86_pmu.num_counters_fixed)-1) << X86_PMC_IDX_FIXED; + x86_pmu.intel_ctrl = perf_counter_mask; perf_counters_lapic_init(); register_die_notifier(&perf_counter_nmi_notifier); @@ -1554,14 +1792,15 @@ const struct pmu *hw_perf_counter_init(struct perf_counter *counter) */ static inline -void callchain_store(struct perf_callchain_entry *entry, unsigned long ip) +void callchain_store(struct perf_callchain_entry *entry, u64 ip) { - if (entry->nr < MAX_STACK_DEPTH) + if (entry->nr < PERF_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 DEFINE_PER_CPU(int, in_nmi_frame); static void @@ -1577,14 +1816,19 @@ static void backtrace_warning(void *data, char *msg) static int backtrace_stack(void *data, char *name) { - /* Don't bother with IRQ stacks for now */ - return -1; + per_cpu(in_nmi_frame, smp_processor_id()) = + x86_is_stack_id(NMI_STACK, name); + + return 0; } static void backtrace_address(void *data, unsigned long addr, int reliable) { struct perf_callchain_entry *entry = data; + if (per_cpu(in_nmi_frame, smp_processor_id())) + return; + if (reliable) callchain_store(entry, addr); } @@ -1596,47 +1840,59 @@ static const struct stacktrace_ops backtrace_ops = { .address = backtrace_address, }; +#include "../dumpstack.h" + 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, PERF_CONTEXT_KERNEL); + callchain_store(entry, regs->ip); - callchain_store(entry, instruction_pointer(regs)); + dump_trace(NULL, regs, NULL, 0, &backtrace_ops, entry); +} - stack = ((char *)regs + sizeof(struct pt_regs)); -#ifdef CONFIG_FRAME_POINTER - bp = frame_pointer(regs); -#else - bp = 0; -#endif +/* + * best effort, GUP based copy_from_user() that assumes IRQ or NMI context + */ +static unsigned long +copy_from_user_nmi(void *to, const void __user *from, unsigned long n) +{ + unsigned long offset, addr = (unsigned long)from; + int type = in_nmi() ? KM_NMI : KM_IRQ0; + unsigned long size, len = 0; + struct page *page; + void *map; + int ret; - dump_trace(NULL, regs, (void *)stack, bp, &backtrace_ops, entry); + do { + ret = __get_user_pages_fast(addr, 1, 0, &page); + if (!ret) + break; - entry->kernel = entry->nr - nr; -} + offset = addr & (PAGE_SIZE - 1); + size = min(PAGE_SIZE - offset, n - len); + map = kmap_atomic(page, type); + memcpy(to, map+offset, size); + kunmap_atomic(map, type); + put_page(page); -struct stack_frame { - const void __user *next_fp; - unsigned long return_address; -}; + len += size; + to += size; + addr += size; + + } while (len < n); + + return len; +} 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; + unsigned long bytes; - ret = 1; - pagefault_disable(); - if (__copy_from_user_inatomic(frame, fp, sizeof(*frame))) - ret = 0; - pagefault_enable(); + bytes = copy_from_user_nmi(frame, fp, sizeof(*frame)); - return ret; + return bytes == sizeof(*frame); } static void @@ -1644,28 +1900,28 @@ 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; + if (!user_mode(regs)) + regs = task_pt_regs(current); + + fp = (void __user *)regs->bp; + callchain_store(entry, PERF_CONTEXT_USER); callchain_store(entry, regs->ip); - while (entry->nr < MAX_STACK_DEPTH) { - frame.next_fp = NULL; + while (entry->nr < PERF_MAX_STACK_DEPTH) { + frame.next_frame = NULL; frame.return_address = 0; if (!copy_stack_frame(fp, &frame)) break; - if ((unsigned long)fp < user_stack_pointer(regs)) + if ((unsigned long)fp < regs->sp) break; callchain_store(entry, frame.return_address); - fp = frame.next_fp; + fp = frame.next_frame; } - - entry->user = entry->nr - nr; } static void @@ -1701,9 +1957,6 @@ struct perf_callchain_entry *perf_callchain(struct pt_regs *regs) entry = &__get_cpu_var(irq_entry); entry->nr = 0; - entry->hv = 0; - entry->kernel = 0; - entry->user = 0; perf_do_callchain(regs, entry); diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c index d6f5b9fbde3..e60ed740d2b 100644 --- a/arch/x86/kernel/cpu/perfctr-watchdog.c +++ b/arch/x86/kernel/cpu/perfctr-watchdog.c @@ -716,11 +716,15 @@ static void probe_nmi_watchdog(void) wd_ops = &k7_wd_ops; break; case X86_VENDOR_INTEL: - /* - * Work around Core Duo (Yonah) errata AE49 where perfctr1 - * doesn't have a working enable bit. + /* Work around where perfctr1 doesn't have a working enable + * bit as described in the following errata: + * AE49 Core Duo and Intel Core Solo 65 nm + * AN49 Intel Pentium Dual-Core + * AF49 Dual-Core Intel Xeon Processor LV */ - if (boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 14) { + if ((boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 14) || + ((boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 15 && + boot_cpu_data.x86_mask == 4))) { intel_arch_wd_ops.perfctr = MSR_ARCH_PERFMON_PERFCTR0; intel_arch_wd_ops.evntsel = MSR_ARCH_PERFMON_EVENTSEL0; } @@ -799,8 +803,3 @@ int __kprobes lapic_wd_event(unsigned nmi_hz) wd_ops->rearm(wd, nmi_hz); return 1; } - -int lapic_watchdog_ok(void) -{ - return wd_ops != NULL; -} diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c index ff958248e61..5e409dc298a 100644 --- a/arch/x86/kernel/crash.c +++ b/arch/x86/kernel/crash.c @@ -27,6 +27,7 @@ #include <asm/cpu.h> #include <asm/reboot.h> #include <asm/virtext.h> +#include <asm/iommu.h> #if defined(CONFIG_SMP) && defined(CONFIG_X86_LOCAL_APIC) @@ -103,5 +104,10 @@ void native_machine_crash_shutdown(struct pt_regs *regs) #ifdef CONFIG_HPET_TIMER hpet_disable(); #endif + +#ifdef CONFIG_X86_64 + pci_iommu_shutdown(); +#endif + crash_save_cpu(regs, safe_smp_processor_id()); } diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index 95ea5fa7d44..c8405718a4c 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -22,6 +22,7 @@ #include "dumpstack.h" int panic_on_unrecovered_nmi; +int panic_on_io_nmi; unsigned int code_bytes = 64; int kstack_depth_to_print = 3 * STACKSLOTS_PER_LINE; static int die_counter; diff --git a/arch/x86/kernel/dumpstack_32.c b/arch/x86/kernel/dumpstack_32.c index d593cd1f58d..bca5fba91c9 100644 --- a/arch/x86/kernel/dumpstack_32.c +++ b/arch/x86/kernel/dumpstack_32.c @@ -19,6 +19,12 @@ #include "dumpstack.h" +/* Just a stub for now */ +int x86_is_stack_id(int id, char *name) +{ + return 0; +} + void dump_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *stack, unsigned long bp, const struct stacktrace_ops *ops, void *data) diff --git a/arch/x86/kernel/dumpstack_64.c b/arch/x86/kernel/dumpstack_64.c index d35db5993fd..54b0a327676 100644 --- a/arch/x86/kernel/dumpstack_64.c +++ b/arch/x86/kernel/dumpstack_64.c @@ -19,10 +19,8 @@ #include "dumpstack.h" -static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack, - unsigned *usedp, char **idp) -{ - static char ids[][8] = { + +static char x86_stack_ids[][8] = { [DEBUG_STACK - 1] = "#DB", [NMI_STACK - 1] = "NMI", [DOUBLEFAULT_STACK - 1] = "#DF", @@ -33,6 +31,15 @@ static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack, N_EXCEPTION_STACKS + DEBUG_STKSZ / EXCEPTION_STKSZ - 2] = "#DB[?]" #endif }; + +int x86_is_stack_id(int id, char *name) +{ + return x86_stack_ids[id - 1] == name; +} + +static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack, + unsigned *usedp, char **idp) +{ unsigned k; /* @@ -61,7 +68,7 @@ static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack, if (*usedp & (1U << k)) break; *usedp |= 1U << k; - *idp = ids[k]; + *idp = x86_stack_ids[k]; return (unsigned long *)end; } /* @@ -81,12 +88,13 @@ static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack, do { ++j; end -= EXCEPTION_STKSZ; - ids[j][4] = '1' + (j - N_EXCEPTION_STACKS); + x86_stack_ids[j][4] = '1' + + (j - N_EXCEPTION_STACKS); } while (stack < end - EXCEPTION_STKSZ); if (*usedp & (1U << j)) break; *usedp |= 1U << j; - *idp = ids[j]; + *idp = x86_stack_ids[j]; return (unsigned long *)end; } #endif diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 7271fa33d79..5cb5725b2ba 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -627,10 +627,9 @@ __init void e820_setup_gap(void) #ifdef CONFIG_X86_64 if (!found) { gapstart = (max_pfn << PAGE_SHIFT) + 1024*1024; - printk(KERN_ERR "PCI: Warning: Cannot find a gap in the 32bit " - "address range\n" - KERN_ERR "PCI: Unassigned devices with 32bit resource " - "registers may break!\n"); + printk(KERN_ERR + "PCI: Warning: Cannot find a gap in the 32bit address range\n" + "PCI: Unassigned devices with 32bit resource registers may break!\n"); } #endif @@ -1383,6 +1382,8 @@ static unsigned long ram_alignment(resource_size_t pos) return 32*1024*1024; } +#define MAX_RESOURCE_SIZE ((resource_size_t)-1) + void __init e820_reserve_resources_late(void) { int i; @@ -1400,17 +1401,19 @@ void __init e820_reserve_resources_late(void) * avoid stolen RAM: */ for (i = 0; i < e820.nr_map; i++) { - struct e820entry *entry = &e820_saved.map[i]; - resource_size_t start, end; + struct e820entry *entry = &e820.map[i]; + u64 start, end; if (entry->type != E820_RAM) continue; start = entry->addr + entry->size; - end = round_up(start, ram_alignment(start)); - if (start == end) + end = round_up(start, ram_alignment(start)) - 1; + if (end > MAX_RESOURCE_SIZE) + end = MAX_RESOURCE_SIZE; + if (start >= end) continue; - reserve_region_with_split(&iomem_resource, start, - end - 1, "RAM buffer"); + reserve_region_with_split(&iomem_resource, start, end, + "RAM buffer"); } } diff --git a/arch/x86/kernel/efi.c b/arch/x86/kernel/efi.c index 1736acc4d7a..fe26ba3e345 100644 --- a/arch/x86/kernel/efi.c +++ b/arch/x86/kernel/efi.c @@ -240,10 +240,35 @@ static void __init do_add_efi_memmap(void) unsigned long long size = md->num_pages << EFI_PAGE_SHIFT; int e820_type; - if (md->attribute & EFI_MEMORY_WB) - e820_type = E820_RAM; - else + switch (md->type) { + case EFI_LOADER_CODE: + case EFI_LOADER_DATA: + case EFI_BOOT_SERVICES_CODE: + case EFI_BOOT_SERVICES_DATA: + case EFI_CONVENTIONAL_MEMORY: + if (md->attribute & EFI_MEMORY_WB) + e820_type = E820_RAM; + else + e820_type = E820_RESERVED; + break; + case EFI_ACPI_RECLAIM_MEMORY: + e820_type = E820_ACPI; + break; + case EFI_ACPI_MEMORY_NVS: + e820_type = E820_NVS; + break; + case EFI_UNUSABLE_MEMORY: + e820_type = E820_UNUSABLE; + break; + default: + /* + * EFI_RESERVED_TYPE EFI_RUNTIME_SERVICES_CODE + * EFI_RUNTIME_SERVICES_DATA EFI_MEMORY_MAPPED_IO + * EFI_MEMORY_MAPPED_IO_PORT_SPACE EFI_PAL_CODE + */ e820_type = E820_RESERVED; + break; + } e820_add_region(start, size, e820_type); } sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); @@ -329,7 +354,7 @@ void __init efi_init(void) */ c16 = tmp = early_ioremap(efi.systab->fw_vendor, 2); if (c16) { - for (i = 0; i < sizeof(vendor) && *c16; ++i) + for (i = 0; i < sizeof(vendor) - 1 && *c16; ++i) vendor[i] = *c16++; vendor[i] = '\0'; } else @@ -487,7 +512,7 @@ void __init efi_enter_virtual_mode(void) && end_pfn <= max_pfn_mapped)) va = __va(md->phys_addr); else - va = efi_ioremap(md->phys_addr, size); + va = efi_ioremap(md->phys_addr, size, md->type); md->virt_addr = (u64) (unsigned long) va; diff --git a/arch/x86/kernel/efi_64.c b/arch/x86/kernel/efi_64.c index 22c3b7828c5..ac0621a7ac3 100644 --- a/arch/x86/kernel/efi_64.c +++ b/arch/x86/kernel/efi_64.c @@ -98,10 +98,14 @@ void __init efi_call_phys_epilog(void) early_runtime_code_mapping_set_exec(0); } -void __iomem *__init efi_ioremap(unsigned long phys_addr, unsigned long size) +void __iomem *__init efi_ioremap(unsigned long phys_addr, unsigned long size, + u32 type) { unsigned long last_map_pfn; + if (type == EFI_MEMORY_MAPPED_IO) + return ioremap(phys_addr, size); + last_map_pfn = init_memory_mapping(phys_addr, phys_addr + size); if ((last_map_pfn << PAGE_SHIFT) < phys_addr + size) return NULL; diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index c929add475c..c097e7d607c 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -48,7 +48,6 @@ #include <asm/segment.h> #include <asm/smp.h> #include <asm/page_types.h> -#include <asm/desc.h> #include <asm/percpu.h> #include <asm/dwarf2.h> #include <asm/processor-flags.h> @@ -84,7 +83,7 @@ #define preempt_stop(clobbers) DISABLE_INTERRUPTS(clobbers); TRACE_IRQS_OFF #else #define preempt_stop(clobbers) -#define resume_kernel restore_nocheck +#define resume_kernel restore_all #endif .macro TRACE_IRQS_IRET @@ -372,7 +371,7 @@ END(ret_from_exception) ENTRY(resume_kernel) DISABLE_INTERRUPTS(CLBR_ANY) cmpl $0,TI_preempt_count(%ebp) # non-zero preempt_count ? - jnz restore_nocheck + jnz restore_all need_resched: movl TI_flags(%ebp), %ecx # need_resched set ? testb $_TIF_NEED_RESCHED, %cl @@ -540,6 +539,8 @@ syscall_exit: jne syscall_exit_work restore_all: + TRACE_IRQS_IRET +restore_all_notrace: movl PT_EFLAGS(%esp), %eax # mix EFLAGS, SS and CS # Warning: PT_OLDSS(%esp) contains the wrong/random values if we # are returning to the kernel. @@ -551,8 +552,6 @@ restore_all: CFI_REMEMBER_STATE je ldt_ss # returning to user-space with LDT SS restore_nocheck: - TRACE_IRQS_IRET -restore_nocheck_notrace: RESTORE_REGS 4 # skip orig_eax/error_code CFI_ADJUST_CFA_OFFSET -4 irq_return: @@ -588,22 +587,34 @@ ldt_ss: jne restore_nocheck #endif - /* If returning to userspace with 16bit stack, - * try to fix the higher word of ESP, as the CPU - * won't restore it. - * This is an "official" bug of all the x86-compatible - * CPUs, which we can try to work around to make - * dosemu and wine happy. */ - movl PT_OLDESP(%esp), %eax - movl %esp, %edx - call patch_espfix_desc +/* + * Setup and switch to ESPFIX stack + * + * We're returning to userspace with a 16 bit stack. The CPU will not + * restore the high word of ESP for us on executing iret... This is an + * "official" bug of all the x86-compatible CPUs, which we can work + * around to make dosemu and wine happy. We do this by preloading the + * high word of ESP with the high word of the userspace ESP while + * compensating for the offset by changing to the ESPFIX segment with + * a base address that matches for the difference. + */ + mov %esp, %edx /* load kernel esp */ + mov PT_OLDESP(%esp), %eax /* load userspace esp */ + mov %dx, %ax /* eax: new kernel esp */ + sub %eax, %edx /* offset (low word is 0) */ + PER_CPU(gdt_page, %ebx) + shr $16, %edx + mov %dl, GDT_ENTRY_ESPFIX_SS * 8 + 4(%ebx) /* bits 16..23 */ + mov %dh, GDT_ENTRY_ESPFIX_SS * 8 + 7(%ebx) /* bits 24..31 */ pushl $__ESPFIX_SS CFI_ADJUST_CFA_OFFSET 4 - pushl %eax + push %eax /* new kernel esp */ CFI_ADJUST_CFA_OFFSET 4 + /* Disable interrupts, but do not irqtrace this section: we + * will soon execute iret and the tracer was already set to + * the irqstate after the iret */ DISABLE_INTERRUPTS(CLBR_EAX) - TRACE_IRQS_OFF - lss (%esp), %esp + lss (%esp), %esp /* switch to espfix segment */ CFI_ADJUST_CFA_OFFSET -8 jmp restore_nocheck CFI_ENDPROC @@ -716,15 +727,24 @@ PTREGSCALL(vm86) PTREGSCALL(vm86old) .macro FIXUP_ESPFIX_STACK - /* since we are on a wrong stack, we cant make it a C code :( */ +/* + * Switch back for ESPFIX stack to the normal zerobased stack + * + * We can't call C functions using the ESPFIX stack. This code reads + * the high word of the segment base from the GDT and swiches to the + * normal stack and adjusts ESP with the matching offset. + */ + /* fixup the stack */ PER_CPU(gdt_page, %ebx) - GET_DESC_BASE(GDT_ENTRY_ESPFIX_SS, %ebx, %eax, %ax, %al, %ah) - addl %esp, %eax + mov GDT_ENTRY_ESPFIX_SS * 8 + 4(%ebx), %al /* bits 16..23 */ + mov GDT_ENTRY_ESPFIX_SS * 8 + 7(%ebx), %ah /* bits 24..31 */ + shl $16, %eax + addl %esp, %eax /* the adjusted stack pointer */ pushl $__KERNEL_DS CFI_ADJUST_CFA_OFFSET 4 pushl %eax CFI_ADJUST_CFA_OFFSET 4 - lss (%esp), %esp + lss (%esp), %esp /* switch to the normal stack segment */ CFI_ADJUST_CFA_OFFSET -8 .endm .macro UNWIND_ESPFIX_STACK @@ -1154,6 +1174,7 @@ ENTRY(ftrace_graph_caller) pushl %edx movl 0xc(%esp), %edx lea 0x4(%ebp), %eax + movl (%ebp), %ecx subl $MCOUNT_INSN_SIZE, %edx call prepare_ftrace_return popl %edx @@ -1168,6 +1189,7 @@ return_to_handler: pushl %eax pushl %ecx pushl %edx + movl %ebp, %eax call ftrace_return_to_handler movl %eax, 0xc(%esp) popl %edx @@ -1329,7 +1351,7 @@ nmi_stack_correct: xorl %edx,%edx # zero error code movl %esp,%eax # pt_regs pointer call do_nmi - jmp restore_nocheck_notrace + jmp restore_all_notrace CFI_ENDPROC nmi_stack_fixup: diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index de74f0a3e0e..c251be74510 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -135,6 +135,7 @@ ENTRY(ftrace_graph_caller) leaq 8(%rbp), %rdi movq 0x38(%rsp), %rsi + movq (%rbp), %rdx subq $MCOUNT_INSN_SIZE, %rsi call prepare_ftrace_return @@ -150,6 +151,7 @@ GLOBAL(return_to_handler) /* Save the return values */ movq %rax, (%rsp) movq %rdx, 8(%rsp) + movq %rbp, %rdi call ftrace_return_to_handler diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index b79c5533c42..9dbb527e165 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -408,7 +408,8 @@ int ftrace_disable_ftrace_graph_caller(void) * Hook the return address and push it in the stack of return addrs * in current thread info. */ -void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) +void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr, + unsigned long frame_pointer) { unsigned long old; int faulted; @@ -416,10 +417,6 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) unsigned long return_hooker = (unsigned long) &return_to_handler; - /* Nmi's are currently unsupported */ - if (unlikely(in_nmi())) - return; - if (unlikely(atomic_read(¤t->tracing_graph_pause))) return; @@ -453,7 +450,8 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) return; } - if (ftrace_push_return_trace(old, self_addr, &trace.depth) == -EBUSY) { + if (ftrace_push_return_trace(old, self_addr, &trace.depth, + frame_pointer) == -EBUSY) { *parent = old; return; } @@ -496,37 +494,56 @@ static struct syscall_metadata *find_syscall_meta(unsigned long *syscall) struct syscall_metadata *syscall_nr_to_meta(int nr) { - if (!syscalls_metadata || nr >= FTRACE_SYSCALL_MAX || nr < 0) + if (!syscalls_metadata || nr >= NR_syscalls || nr < 0) return NULL; return syscalls_metadata[nr]; } -void arch_init_ftrace_syscalls(void) +int syscall_name_to_nr(char *name) +{ + int i; + + if (!syscalls_metadata) + return -1; + + for (i = 0; i < NR_syscalls; i++) { + if (syscalls_metadata[i]) { + if (!strcmp(syscalls_metadata[i]->name, name)) + return i; + } + } + return -1; +} + +void set_syscall_enter_id(int num, int id) +{ + syscalls_metadata[num]->enter_id = id; +} + +void set_syscall_exit_id(int num, int id) +{ + syscalls_metadata[num]->exit_id = id; +} + +static int __init arch_init_ftrace_syscalls(void) { int i; struct syscall_metadata *meta; unsigned long **psys_syscall_table = &sys_call_table; - static atomic_t refs; - - if (atomic_inc_return(&refs) != 1) - goto end; syscalls_metadata = kzalloc(sizeof(*syscalls_metadata) * - FTRACE_SYSCALL_MAX, GFP_KERNEL); + NR_syscalls, GFP_KERNEL); if (!syscalls_metadata) { WARN_ON(1); - return; + return -ENOMEM; } - for (i = 0; i < FTRACE_SYSCALL_MAX; i++) { + for (i = 0; i < NR_syscalls; i++) { meta = find_syscall_meta(psys_syscall_table[i]); syscalls_metadata[i] = meta; } - return; - - /* Paranoid: avoid overflow */ -end: - atomic_dec(&refs); + return 0; } +arch_initcall(arch_init_ftrace_syscalls); #endif diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index dc5ed4bdd88..cc827ac9e8d 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -13,7 +13,6 @@ #include <asm/segment.h> #include <asm/page_types.h> #include <asm/pgtable_types.h> -#include <asm/desc.h> #include <asm/cache.h> #include <asm/thread_info.h> #include <asm/asm-offsets.h> @@ -262,9 +261,7 @@ page_pde_offset = (__PAGE_OFFSET >> 20); * which will be freed later */ -#ifndef CONFIG_HOTPLUG_CPU -.section .init.text,"ax",@progbits -#endif +__CPUINIT #ifdef CONFIG_SMP ENTRY(startup_32_smp) @@ -603,7 +600,7 @@ ignore_int: #endif iret -.section .cpuinit.data,"wa" + __REFDATA .align 4 ENTRY(initial_code) .long i386_start_kernel diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index 54b29bb24e7..fa54f78e2a0 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -12,7 +12,6 @@ #include <linux/linkage.h> #include <linux/threads.h> #include <linux/init.h> -#include <asm/desc.h> #include <asm/segment.h> #include <asm/pgtable.h> #include <asm/page.h> diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index 81408b93f88..dedc2bddf7a 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -510,7 +510,8 @@ static int hpet_setup_irq(struct hpet_dev *dev) { if (request_irq(dev->irq, hpet_interrupt_handler, - IRQF_DISABLED|IRQF_NOBALANCING, dev->name, dev)) + IRQF_TIMER | IRQF_DISABLED | IRQF_NOBALANCING, + dev->name, dev)) return -1; disable_irq(dev->irq); diff --git a/arch/x86/kernel/irqinit.c b/arch/x86/kernel/irqinit.c index 696f0e475c2..92b7703d3d5 100644 --- a/arch/x86/kernel/irqinit.c +++ b/arch/x86/kernel/irqinit.c @@ -187,7 +187,7 @@ static void __init apic_intr_init(void) #ifdef CONFIG_X86_THERMAL_VECTOR alloc_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt); #endif -#ifdef CONFIG_X86_THRESHOLD +#ifdef CONFIG_X86_MCE_THRESHOLD alloc_intr_gate(THRESHOLD_APIC_VECTOR, threshold_interrupt); #endif #if defined(CONFIG_X86_NEW_MCE) && defined(CONFIG_X86_LOCAL_APIC) diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c index a78ecad0c90..c664d515f61 100644 --- a/arch/x86/kernel/kvm.c +++ b/arch/x86/kernel/kvm.c @@ -200,7 +200,7 @@ static void kvm_leave_lazy_mmu(void) state->mode = paravirt_get_lazy_mode(); } -static void paravirt_ops_setup(void) +static void __init paravirt_ops_setup(void) { pv_info.name = "KVM"; pv_info.paravirt_enabled = 1; diff --git a/arch/x86/kernel/mfgpt_32.c b/arch/x86/kernel/mfgpt_32.c index 846510b78a0..2a62d843f01 100644 --- a/arch/x86/kernel/mfgpt_32.c +++ b/arch/x86/kernel/mfgpt_32.c @@ -347,7 +347,7 @@ static irqreturn_t mfgpt_tick(int irq, void *dev_id) static struct irqaction mfgptirq = { .handler = mfgpt_tick, - .flags = IRQF_DISABLED | IRQF_NOBALANCING, + .flags = IRQF_DISABLED | IRQF_NOBALANCING | IRQF_TIMER, .name = "mfgpt-timer" }; diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index 745579bc825..1a041bcf506 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -32,6 +32,8 @@ int no_iommu __read_mostly; /* Set this to 1 if there is a HW IOMMU in the system */ int iommu_detected __read_mostly = 0; +int iommu_pass_through; + dma_addr_t bad_dma_address __read_mostly = 0; EXPORT_SYMBOL(bad_dma_address); @@ -210,6 +212,10 @@ static __init int iommu_setup(char *p) if (!strncmp(p, "soft", 4)) swiotlb = 1; #endif + if (!strncmp(p, "pt", 2)) { + iommu_pass_through = 1; + return 1; + } gart_parse_options(p); @@ -290,6 +296,8 @@ static int __init pci_iommu_init(void) void pci_iommu_shutdown(void) { gart_iommu_shutdown(); + + amd_iommu_shutdown(); } /* Must execute after PCI subsystem */ fs_initcall(pci_iommu_init); diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index cfd9f906389..d2e56b8f48e 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -675,7 +675,7 @@ static __init int init_k8_gatt(struct agp_kern_info *info) nommu: /* Should not happen anymore */ printk(KERN_WARNING "PCI-DMA: More than 4GB of RAM and no IOMMU\n" - KERN_WARNING "falling back to iommu=soft.\n"); + "falling back to iommu=soft.\n"); return -1; } diff --git a/arch/x86/kernel/pci-swiotlb.c b/arch/x86/kernel/pci-swiotlb.c index a1712f2b50f..6af96ee4420 100644 --- a/arch/x86/kernel/pci-swiotlb.c +++ b/arch/x86/kernel/pci-swiotlb.c @@ -71,7 +71,8 @@ void __init pci_swiotlb_init(void) { /* don't initialize swiotlb if iommu=off (no_iommu=1) */ #ifdef CONFIG_X86_64 - if (!iommu_detected && !no_iommu && max_pfn > MAX_DMA32_PFN) + if ((!iommu_detected && !no_iommu && max_pfn > MAX_DMA32_PFN) || + iommu_pass_through) swiotlb = 1; #endif if (swiotlb_force) diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index fc6e4b773fc..1092a1a2fbe 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -509,16 +509,12 @@ static void c1e_idle(void) if (!cpumask_test_cpu(cpu, c1e_mask)) { cpumask_set_cpu(cpu, c1e_mask); /* - * Force broadcast so ACPI can not interfere. Needs - * to run with interrupts enabled as it uses - * smp_function_call. + * Force broadcast so ACPI can not interfere. */ - local_irq_enable(); clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_FORCE, &cpu); printk(KERN_INFO "Switch to broadcast mode on CPU%d\n", cpu); - local_irq_disable(); } clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_ENTER, &cpu); diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index cabdabce3cb..113b8927c82 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -36,10 +36,11 @@ #include <asm/ds.h> #include <asm/hw_breakpoint.h> -#include <trace/syscall.h> - #include "tls.h" +#define CREATE_TRACE_POINTS +#include <trace/events/syscalls.h> + enum x86_regset { REGSET_GENERAL, REGSET_FP, @@ -1548,8 +1549,8 @@ asmregparm long syscall_trace_enter(struct pt_regs *regs) tracehook_report_syscall_entry(regs)) ret = -1L; - if (unlikely(test_thread_flag(TIF_SYSCALL_FTRACE))) - ftrace_syscall_enter(regs); + if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) + trace_sys_enter(regs, regs->orig_ax); if (unlikely(current->audit_context)) { if (IS_IA32) @@ -1574,8 +1575,8 @@ asmregparm void syscall_trace_leave(struct pt_regs *regs) if (unlikely(current->audit_context)) audit_syscall_exit(AUDITSC_RESULT(regs->ax), regs->ax); - if (unlikely(test_thread_flag(TIF_SYSCALL_FTRACE))) - ftrace_syscall_exit(regs); + if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) + trace_sys_exit(regs, regs->ax); if (test_thread_flag(TIF_SYSCALL_TRACE)) tracehook_report_syscall_exit(regs, 0); diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c index 4f9c55f3a7c..03801f2f761 100644 --- a/arch/x86/kernel/pvclock.c +++ b/arch/x86/kernel/pvclock.c @@ -60,7 +60,7 @@ static inline u64 scale_delta(u64 delta, u32 mul_frac, int shift) "adc %5,%%edx ; " : "=A" (product), "=r" (tmp1), "=r" (tmp2) : "a" ((u32)delta), "1" ((u32)(delta >> 32)), "2" (mul_frac) ); -#elif __x86_64__ +#elif defined(__x86_64__) __asm__ ( "mul %%rdx ; shrd $32,%%rdx,%%rax" : "=a" (product) : "0" (delta), "d" ((u64)mul_frac) ); diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index d2d1ce8170f..a06e8d10184 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -3,6 +3,7 @@ #include <linux/init.h> #include <linux/pm.h> #include <linux/efi.h> +#include <linux/dmi.h> #include <acpi/reboot.h> #include <asm/io.h> #include <asm/apic.h> @@ -17,7 +18,6 @@ #include <asm/cpu.h> #ifdef CONFIG_X86_32 -# include <linux/dmi.h> # include <linux/ctype.h> # include <linux/mc146818rtc.h> #else @@ -249,6 +249,14 @@ static struct dmi_system_id __initdata reboot_dmi_table[] = { DMI_MATCH(DMI_PRODUCT_NAME, "VGN-Z540N"), }, }, + { /* Handle problems with rebooting on CompuLab SBC-FITPC2 */ + .callback = set_bios_reboot, + .ident = "CompuLab SBC-FITPC2", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "CompuLab"), + DMI_MATCH(DMI_PRODUCT_NAME, "SBC-FITPC2"), + }, + }, { } }; @@ -396,6 +404,46 @@ EXPORT_SYMBOL(machine_real_restart); #endif /* CONFIG_X86_32 */ +/* + * Some Apple MacBook and MacBookPro's needs reboot=p to be able to reboot + */ +static int __init set_pci_reboot(const struct dmi_system_id *d) +{ + if (reboot_type != BOOT_CF9) { + reboot_type = BOOT_CF9; + printk(KERN_INFO "%s series board detected. " + "Selecting PCI-method for reboots.\n", d->ident); + } + return 0; +} + +static struct dmi_system_id __initdata pci_reboot_dmi_table[] = { + { /* Handle problems with rebooting on Apple MacBook5 */ + .callback = set_pci_reboot, + .ident = "Apple MacBook5", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "MacBook5"), + }, + }, + { /* Handle problems with rebooting on Apple MacBookPro5 */ + .callback = set_pci_reboot, + .ident = "Apple MacBookPro5", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro5"), + }, + }, + { } +}; + +static int __init pci_reboot_init(void) +{ + dmi_check_system(pci_reboot_dmi_table); + return 0; +} +core_initcall(pci_reboot_init); + static inline void kb_wait(void) { int i; diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index be5ae80f897..63f32d220ef 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -289,6 +289,20 @@ void * __init extend_brk(size_t size, size_t align) return ret; } +#ifdef CONFIG_X86_64 +static void __init init_gbpages(void) +{ + if (direct_gbpages && cpu_has_gbpages) + printk(KERN_INFO "Using GB pages for direct mapping\n"); + else + direct_gbpages = 0; +} +#else +static inline void init_gbpages(void) +{ +} +#endif + static void __init reserve_brk(void) { if (_brk_end > _brk_start) @@ -658,6 +672,19 @@ static struct dmi_system_id __initdata bad_bios_dmi_table[] = { DMI_MATCH(DMI_BIOS_VENDOR, "Phoenix Technologies"), }, }, + { + /* + * AMI BIOS with low memory corruption was found on Intel DG45ID board. + * It hase different DMI_BIOS_VENDOR = "Intel Corp.", for now we will + * match only DMI_BOARD_NAME and see if there is more bad products + * with this vendor. + */ + .callback = dmi_low_memory_corruption, + .ident = "AMI BIOS", + .matches = { + DMI_MATCH(DMI_BOARD_NAME, "DG45ID"), + }, + }, #endif {} }; @@ -871,6 +898,8 @@ void __init setup_arch(char **cmdline_p) reserve_brk(); + init_gbpages(); + /* max_pfn_mapped is updated here */ max_low_pfn_mapped = init_memory_mapping(0, max_low_pfn<<PAGE_SHIFT); max_pfn_mapped = max_low_pfn_mapped; diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 9c3f0823e6a..07d81916f21 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -124,7 +124,7 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size, } /* - * Remap allocator + * Large page remap allocator * * This allocator uses PMD page as unit. A PMD page is allocated for * each cpu and each is remapped into vmalloc area using PMD mapping. @@ -137,105 +137,185 @@ static void * __init pcpu_alloc_bootmem(unsigned int cpu, unsigned long size, * better than only using 4k mappings while still being NUMA friendly. */ #ifdef CONFIG_NEED_MULTIPLE_NODES -static size_t pcpur_size __initdata; -static void **pcpur_ptrs __initdata; +struct pcpul_ent { + unsigned int cpu; + void *ptr; +}; + +static size_t pcpul_size; +static struct pcpul_ent *pcpul_map; +static struct vm_struct pcpul_vm; -static struct page * __init pcpur_get_page(unsigned int cpu, int pageno) +static struct page * __init pcpul_get_page(unsigned int cpu, int pageno) { size_t off = (size_t)pageno << PAGE_SHIFT; - if (off >= pcpur_size) + if (off >= pcpul_size) return NULL; - return virt_to_page(pcpur_ptrs[cpu] + off); + return virt_to_page(pcpul_map[cpu].ptr + off); } -static ssize_t __init setup_pcpu_remap(size_t static_size) +static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) { - static struct vm_struct vm; - size_t ptrs_size, dyn_size; + size_t map_size, dyn_size; unsigned int cpu; + int i, j; ssize_t ret; - /* - * If large page isn't supported, there's no benefit in doing - * this. Also, on non-NUMA, embedding is better. - * - * NOTE: disabled for now. - */ - if (true || !cpu_has_pse || !pcpu_need_numa()) + if (!chosen) { + size_t vm_size = VMALLOC_END - VMALLOC_START; + size_t tot_size = nr_cpu_ids * PMD_SIZE; + + /* on non-NUMA, embedding is better */ + if (!pcpu_need_numa()) + return -EINVAL; + + /* don't consume more than 20% of vmalloc area */ + if (tot_size > vm_size / 5) { + pr_info("PERCPU: too large chunk size %zuMB for " + "large page remap\n", tot_size >> 20); + return -EINVAL; + } + } + + /* need PSE */ + if (!cpu_has_pse) { + pr_warning("PERCPU: lpage allocator requires PSE\n"); return -EINVAL; + } /* * Currently supports only single page. Supporting multiple * pages won't be too difficult if it ever becomes necessary. */ - pcpur_size = PFN_ALIGN(static_size + PERCPU_MODULE_RESERVE + + pcpul_size = PFN_ALIGN(static_size + PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE); - if (pcpur_size > PMD_SIZE) { + if (pcpul_size > PMD_SIZE) { pr_warning("PERCPU: static data is larger than large page, " "can't use large page\n"); return -EINVAL; } - dyn_size = pcpur_size - static_size - PERCPU_FIRST_CHUNK_RESERVE; + dyn_size = pcpul_size - static_size - PERCPU_FIRST_CHUNK_RESERVE; /* allocate pointer array and alloc large pages */ - ptrs_size = PFN_ALIGN(num_possible_cpus() * sizeof(pcpur_ptrs[0])); - pcpur_ptrs = alloc_bootmem(ptrs_size); + map_size = PFN_ALIGN(nr_cpu_ids * sizeof(pcpul_map[0])); + pcpul_map = alloc_bootmem(map_size); for_each_possible_cpu(cpu) { - pcpur_ptrs[cpu] = pcpu_alloc_bootmem(cpu, PMD_SIZE, PMD_SIZE); - if (!pcpur_ptrs[cpu]) + pcpul_map[cpu].cpu = cpu; + pcpul_map[cpu].ptr = pcpu_alloc_bootmem(cpu, PMD_SIZE, + PMD_SIZE); + if (!pcpul_map[cpu].ptr) { + pr_warning("PERCPU: failed to allocate large page " + "for cpu%u\n", cpu); goto enomem; + } /* - * Only use pcpur_size bytes and give back the rest. + * Only use pcpul_size bytes and give back the rest. * * Ingo: The 2MB up-rounding bootmem is needed to make * sure the partial 2MB page is still fully RAM - it's * not well-specified to have a PAT-incompatible area * (unmapped RAM, device memory, etc.) in that hole. */ - free_bootmem(__pa(pcpur_ptrs[cpu] + pcpur_size), - PMD_SIZE - pcpur_size); + free_bootmem(__pa(pcpul_map[cpu].ptr + pcpul_size), + PMD_SIZE - pcpul_size); - memcpy(pcpur_ptrs[cpu], __per_cpu_load, static_size); + memcpy(pcpul_map[cpu].ptr, __per_cpu_load, static_size); } /* allocate address and map */ - vm.flags = VM_ALLOC; - vm.size = num_possible_cpus() * PMD_SIZE; - vm_area_register_early(&vm, PMD_SIZE); + pcpul_vm.flags = VM_ALLOC; + pcpul_vm.size = nr_cpu_ids * PMD_SIZE; + vm_area_register_early(&pcpul_vm, PMD_SIZE); for_each_possible_cpu(cpu) { - pmd_t *pmd; + pmd_t *pmd, pmd_v; - pmd = populate_extra_pmd((unsigned long)vm.addr - + cpu * PMD_SIZE); - set_pmd(pmd, pfn_pmd(page_to_pfn(virt_to_page(pcpur_ptrs[cpu])), - PAGE_KERNEL_LARGE)); + pmd = populate_extra_pmd((unsigned long)pcpul_vm.addr + + cpu * PMD_SIZE); + pmd_v = pfn_pmd(page_to_pfn(virt_to_page(pcpul_map[cpu].ptr)), + PAGE_KERNEL_LARGE); + set_pmd(pmd, pmd_v); } /* we're ready, commit */ pr_info("PERCPU: Remapped at %p with large pages, static data " - "%zu bytes\n", vm.addr, static_size); + "%zu bytes\n", pcpul_vm.addr, static_size); - ret = pcpu_setup_first_chunk(pcpur_get_page, static_size, + ret = pcpu_setup_first_chunk(pcpul_get_page, static_size, PERCPU_FIRST_CHUNK_RESERVE, dyn_size, - PMD_SIZE, vm.addr, NULL); - goto out_free_ar; + PMD_SIZE, pcpul_vm.addr, NULL); + + /* sort pcpul_map array for pcpu_lpage_remapped() */ + for (i = 0; i < nr_cpu_ids - 1; i++) + for (j = i + 1; j < nr_cpu_ids; j++) + if (pcpul_map[i].ptr > pcpul_map[j].ptr) { + struct pcpul_ent tmp = pcpul_map[i]; + pcpul_map[i] = pcpul_map[j]; + pcpul_map[j] = tmp; + } + + return ret; enomem: for_each_possible_cpu(cpu) - if (pcpur_ptrs[cpu]) - free_bootmem(__pa(pcpur_ptrs[cpu]), PMD_SIZE); - ret = -ENOMEM; -out_free_ar: - free_bootmem(__pa(pcpur_ptrs), ptrs_size); - return ret; + if (pcpul_map[cpu].ptr) + free_bootmem(__pa(pcpul_map[cpu].ptr), pcpul_size); + free_bootmem(__pa(pcpul_map), map_size); + return -ENOMEM; +} + +/** + * pcpu_lpage_remapped - determine whether a kaddr is in pcpul recycled area + * @kaddr: the kernel address in question + * + * Determine whether @kaddr falls in the pcpul recycled area. This is + * used by pageattr to detect VM aliases and break up the pcpu PMD + * mapping such that the same physical page is not mapped under + * different attributes. + * + * The recycled area is always at the tail of a partially used PMD + * page. + * + * RETURNS: + * Address of corresponding remapped pcpu address if match is found; + * otherwise, NULL. + */ +void *pcpu_lpage_remapped(void *kaddr) +{ + void *pmd_addr = (void *)((unsigned long)kaddr & PMD_MASK); + unsigned long offset = (unsigned long)kaddr & ~PMD_MASK; + int left = 0, right = nr_cpu_ids - 1; + int pos; + + /* pcpul in use at all? */ + if (!pcpul_map) + return NULL; + + /* okay, perform binary search */ + while (left <= right) { + pos = (left + right) / 2; + + if (pcpul_map[pos].ptr < pmd_addr) + left = pos + 1; + else if (pcpul_map[pos].ptr > pmd_addr) + right = pos - 1; + else { + /* it shouldn't be in the area for the first chunk */ + WARN_ON(offset < pcpul_size); + + return pcpul_vm.addr + + pcpul_map[pos].cpu * PMD_SIZE + offset; + } + } + + return NULL; } #else -static ssize_t __init setup_pcpu_remap(size_t static_size) +static ssize_t __init setup_pcpu_lpage(size_t static_size, bool chosen) { return -EINVAL; } @@ -249,7 +329,7 @@ static ssize_t __init setup_pcpu_remap(size_t static_size) * mapping so that it can use PMD mapping without additional TLB * pressure. */ -static ssize_t __init setup_pcpu_embed(size_t static_size) +static ssize_t __init setup_pcpu_embed(size_t static_size, bool chosen) { size_t reserve = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE; @@ -258,7 +338,7 @@ static ssize_t __init setup_pcpu_embed(size_t static_size) * this. Also, embedding allocation doesn't play well with * NUMA. */ - if (!cpu_has_pse || pcpu_need_numa()) + if (!chosen && (!cpu_has_pse || pcpu_need_numa())) return -EINVAL; return pcpu_embed_first_chunk(static_size, PERCPU_FIRST_CHUNK_RESERVE, @@ -297,7 +377,7 @@ static ssize_t __init setup_pcpu_4k(size_t static_size) pcpu4k_nr_static_pages = PFN_UP(static_size); /* unaligned allocations can't be freed, round up to page size */ - pages_size = PFN_ALIGN(pcpu4k_nr_static_pages * num_possible_cpus() + pages_size = PFN_ALIGN(pcpu4k_nr_static_pages * nr_cpu_ids * sizeof(pcpu4k_pages[0])); pcpu4k_pages = alloc_bootmem(pages_size); @@ -308,8 +388,11 @@ static ssize_t __init setup_pcpu_4k(size_t static_size) void *ptr; ptr = pcpu_alloc_bootmem(cpu, PAGE_SIZE, PAGE_SIZE); - if (!ptr) + if (!ptr) { + pr_warning("PERCPU: failed to allocate " + "4k page for cpu%u\n", cpu); goto enomem; + } memcpy(ptr, __per_cpu_load + i * PAGE_SIZE, PAGE_SIZE); pcpu4k_pages[j++] = virt_to_page(ptr); @@ -333,6 +416,16 @@ out_free_ar: return ret; } +/* for explicit first chunk allocator selection */ +static char pcpu_chosen_alloc[16] __initdata; + +static int __init percpu_alloc_setup(char *str) +{ + strncpy(pcpu_chosen_alloc, str, sizeof(pcpu_chosen_alloc) - 1); + return 0; +} +early_param("percpu_alloc", percpu_alloc_setup); + static inline void setup_percpu_segment(int cpu) { #ifdef CONFIG_X86_32 @@ -346,11 +439,6 @@ static inline void setup_percpu_segment(int cpu) #endif } -/* - * Great future plan: - * Declare PDA itself and support (irqstack,tss,pgd) as per cpu data. - * Always point %gs to its beginning - */ void __init setup_per_cpu_areas(void) { size_t static_size = __per_cpu_end - __per_cpu_start; @@ -367,9 +455,26 @@ void __init setup_per_cpu_areas(void) * of large page mappings. Please read comments on top of * each allocator for details. */ - ret = setup_pcpu_remap(static_size); - if (ret < 0) - ret = setup_pcpu_embed(static_size); + ret = -EINVAL; + if (strlen(pcpu_chosen_alloc)) { + if (strcmp(pcpu_chosen_alloc, "4k")) { + if (!strcmp(pcpu_chosen_alloc, "lpage")) + ret = setup_pcpu_lpage(static_size, true); + else if (!strcmp(pcpu_chosen_alloc, "embed")) + ret = setup_pcpu_embed(static_size, true); + else + pr_warning("PERCPU: unknown allocator %s " + "specified\n", pcpu_chosen_alloc); + if (ret < 0) + pr_warning("PERCPU: %s allocator failed (%zd), " + "falling back to 4k\n", + pcpu_chosen_alloc, ret); + } + } else { + ret = setup_pcpu_lpage(static_size, false); + if (ret < 0) + ret = setup_pcpu_embed(static_size, false); + } if (ret < 0) ret = setup_pcpu_4k(static_size); if (ret < 0) diff --git a/arch/x86/kernel/sys_x86_64.c b/arch/x86/kernel/sys_x86_64.c index 6bc211accf0..45e00eb09c3 100644 --- a/arch/x86/kernel/sys_x86_64.c +++ b/arch/x86/kernel/sys_x86_64.c @@ -18,9 +18,9 @@ #include <asm/ia32.h> #include <asm/syscalls.h> -asmlinkage long sys_mmap(unsigned long addr, unsigned long len, - unsigned long prot, unsigned long flags, - unsigned long fd, unsigned long off) +SYSCALL_DEFINE6(mmap, unsigned long, addr, unsigned long, len, + unsigned long, prot, unsigned long, flags, + unsigned long, fd, unsigned long, off) { long error; struct file *file; @@ -226,7 +226,7 @@ bottomup: } -asmlinkage long sys_uname(struct new_utsname __user *name) +SYSCALL_DEFINE1(uname, struct new_utsname __user *, name) { int err; down_read(&uts_sem); diff --git a/arch/x86/kernel/tlb_uv.c b/arch/x86/kernel/tlb_uv.c index 124d40c575d..77b9689f8ed 100644 --- a/arch/x86/kernel/tlb_uv.c +++ b/arch/x86/kernel/tlb_uv.c @@ -711,7 +711,6 @@ uv_activation_descriptor_init(int node, int pnode) unsigned long pa; unsigned long m; unsigned long n; - unsigned long mmr_image; struct bau_desc *adp; struct bau_desc *ad2; @@ -727,12 +726,8 @@ uv_activation_descriptor_init(int node, int pnode) n = pa >> uv_nshift; m = pa & uv_mmask; - mmr_image = uv_read_global_mmr64(pnode, UVH_LB_BAU_SB_DESCRIPTOR_BASE); - if (mmr_image) { - uv_write_global_mmr64(pnode, (unsigned long) - UVH_LB_BAU_SB_DESCRIPTOR_BASE, - (n << UV_DESC_BASE_PNODE_SHIFT | m)); - } + uv_write_global_mmr64(pnode, UVH_LB_BAU_SB_DESCRIPTOR_BASE, + (n << UV_DESC_BASE_PNODE_SHIFT | m)); /* * initializing all 8 (UV_ITEMS_PER_DESCRIPTOR) descriptors for each @@ -749,6 +744,7 @@ uv_activation_descriptor_init(int node, int pnode) * note that base_dest_nodeid is actually a nasid. */ ad2->header.base_dest_nodeid = uv_partition_base_pnode << 1; + ad2->header.dest_subnodeid = 0x10; /* the LB */ ad2->header.command = UV_NET_ENDPOINT_INTD; ad2->header.int_both = 1; /* diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 286d64eba31..ae04589a579 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -54,6 +54,7 @@ #include <asm/traps.h> #include <asm/desc.h> #include <asm/i387.h> +#include <asm/mce.h> #include <asm/mach_traps.h> @@ -65,8 +66,6 @@ #include <asm/setup.h> #include <asm/traps.h> -#include "cpu/mcheck/mce.h" - asmlinkage int system_call(void); /* Do we ignore FPU interrupts ? */ @@ -347,6 +346,9 @@ io_check_error(unsigned char reason, struct pt_regs *regs) printk(KERN_EMERG "NMI: IOCK error (debug interrupt?)\n"); show_registers(regs); + if (panic_on_io_nmi) + panic("NMI IOCK error: Not continuing"); + /* Re-enable the IOCK line, wait for a few seconds */ reason = (reason & 0xf) | 8; outb(reason, 0x61); diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index ae3180c506a..71f4368b357 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -275,15 +275,20 @@ static unsigned long pit_calibrate_tsc(u32 latch, unsigned long ms, int loopmin) * use the TSC value at the transitions to calculate a pretty * good value for the TSC frequencty. */ +static inline int pit_verify_msb(unsigned char val) +{ + /* Ignore LSB */ + inb(0x42); + return inb(0x42) == val; +} + static inline int pit_expect_msb(unsigned char val, u64 *tscp, unsigned long *deltap) { int count; u64 tsc = 0; for (count = 0; count < 50000; count++) { - /* Ignore LSB */ - inb(0x42); - if (inb(0x42) != val) + if (!pit_verify_msb(val)) break; tsc = get_cycles(); } @@ -336,8 +341,7 @@ static unsigned long quick_pit_calibrate(void) * to do that is to just read back the 16-bit counter * once from the PIT. */ - inb(0x42); - inb(0x42); + pit_verify_msb(0); if (pit_expect_msb(0xff, &tsc, &d1)) { for (i = 1; i <= MAX_QUICK_PIT_ITERATIONS; i++) { @@ -348,8 +352,19 @@ static unsigned long quick_pit_calibrate(void) * Iterate until the error is less than 500 ppm */ delta -= tsc; - if (d1+d2 < delta >> 11) - goto success; + if (d1+d2 >= delta >> 11) + continue; + + /* + * Check the PIT one more time to verify that + * all TSC reads were stable wrt the PIT. + * + * This also guarantees serialization of the + * last cycle read ('d2') in pit_expect_msb. + */ + if (!pit_verify_msb(0xfe - i)) + break; + goto success; } } printk("Fast TSC calibration failed\n"); @@ -590,22 +605,26 @@ EXPORT_SYMBOL(recalibrate_cpu_khz); */ DEFINE_PER_CPU(unsigned long, cyc2ns); +DEFINE_PER_CPU(unsigned long long, cyc2ns_offset); static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) { - unsigned long long tsc_now, ns_now; + unsigned long long tsc_now, ns_now, *offset; unsigned long flags, *scale; local_irq_save(flags); sched_clock_idle_sleep_event(); scale = &per_cpu(cyc2ns, cpu); + offset = &per_cpu(cyc2ns_offset, cpu); rdtscll(tsc_now); ns_now = __cycles_2_ns(tsc_now); - if (cpu_khz) + if (cpu_khz) { *scale = (NSEC_PER_MSEC << CYC2NS_SCALE_FACTOR)/cpu_khz; + *offset = ns_now - (tsc_now * *scale >> CYC2NS_SCALE_FACTOR); + } sched_clock_idle_wakeup_event(0); local_irq_restore(flags); @@ -632,17 +651,15 @@ static int time_cpufreq_notifier(struct notifier_block *nb, unsigned long val, void *data) { struct cpufreq_freqs *freq = data; - unsigned long *lpj, dummy; + unsigned long *lpj; if (cpu_has(&cpu_data(freq->cpu), X86_FEATURE_CONSTANT_TSC)) return 0; - lpj = &dummy; - if (!(freq->flags & CPUFREQ_CONST_LOOPS)) + lpj = &boot_cpu_data.loops_per_jiffy; #ifdef CONFIG_SMP + if (!(freq->flags & CPUFREQ_CONST_LOOPS)) lpj = &cpu_data(freq->cpu).loops_per_jiffy; -#else - lpj = &boot_cpu_data.loops_per_jiffy; #endif if (!ref_freq) { diff --git a/arch/x86/kernel/vmi_32.c b/arch/x86/kernel/vmi_32.c index b263423fbe2..95a7289e4b0 100644 --- a/arch/x86/kernel/vmi_32.c +++ b/arch/x86/kernel/vmi_32.c @@ -441,7 +441,7 @@ vmi_startup_ipi_hook(int phys_apicid, unsigned long start_eip, ap.ds = __USER_DS; ap.es = __USER_DS; ap.fs = __KERNEL_PERCPU; - ap.gs = 0; + ap.gs = __KERNEL_STACK_CANARY; ap.eflags = 0; diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S index 367e8788204..9fc178255c0 100644 --- a/arch/x86/kernel/vmlinux.lds.S +++ b/arch/x86/kernel/vmlinux.lds.S @@ -46,11 +46,10 @@ PHDRS { 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 */ + init PT_LOAD FLAGS(7); /* RWE */ #endif note PT_NOTE FLAGS(0); /* ___ */ } @@ -103,72 +102,43 @@ SECTIONS __stop___ex_table = .; } :text = 0x9090 - RODATA + RO_DATA(PAGE_SIZE) /* Data */ - . = ALIGN(PAGE_SIZE); .data : AT(ADDR(.data) - LOAD_OFFSET) { /* Start of data section */ _sdata = .; - DATA_DATA - CONSTRUCTORS -#ifdef CONFIG_X86_64 - /* End of data section */ - _edata = .; -#endif - } :data + /* init_task */ + INIT_TASK_DATA(THREAD_SIZE) #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 = .; - } + /* 32 bit has nosave before _edata */ + NOSAVE_DATA #endif - . = ALIGN(PAGE_SIZE); - .data.page_aligned : AT(ADDR(.data.page_aligned) - LOAD_OFFSET) { - *(.data.page_aligned) + PAGE_ALIGNED_DATA(PAGE_SIZE) *(.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) - } + CACHELINE_ALIGNED_DATA(CONFIG_X86_L1_CACHE_BYTES) - /* 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) + DATA_DATA + CONSTRUCTORS + + /* rarely changed data like cpu maps */ + READ_MOSTLY_DATA(CONFIG_X86_INTERNODE_CACHE_BYTES) -#ifdef CONFIG_X86_32 /* End of data section */ _edata = .; -#endif - } + } :data #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 VSYSCALL_PHYS_ADDR ((LOADADDR(.data) + SIZEOF(.data) + \ + PAGE_SIZE - 1) & ~(PAGE_SIZE - 1)) +#define VSYSCALL_VIRT_ADDR ((ADDR(.data) + SIZEOF(.data) + \ + PAGE_SIZE - 1) & ~(PAGE_SIZE - 1)) #define VLOAD_OFFSET (VSYSCALL_ADDR - VSYSCALL_PHYS_ADDR) #define VLOAD(x) (ADDR(x) - VLOAD_OFFSET) @@ -234,35 +204,29 @@ SECTIONS #endif /* CONFIG_X86_64 */ - /* init_task */ - . = ALIGN(THREAD_SIZE); - .data.init_task : AT(ADDR(.data.init_task) - LOAD_OFFSET) { - *(.data.init_task) + /* Init code and data - will be freed after init */ + . = ALIGN(PAGE_SIZE); + .init.begin : AT(ADDR(.init.begin) - LOAD_OFFSET) { + __init_begin = .; /* paired with __init_end */ } -#ifdef CONFIG_X86_64 - :data.init -#endif +#if defined(CONFIG_X86_64) && defined(CONFIG_SMP) /* - * smp_locks might be freed after init - * start/end must be page aligned + * percpu offsets are zero-based on SMP. PERCPU_VADDR() changes the + * output PHDR, so the next output section - .init.text - should + * start another segment - init. */ - . = ALIGN(PAGE_SIZE); - .smp_locks : AT(ADDR(.smp_locks) - LOAD_OFFSET) { - __smp_locks = .; - *(.smp_locks) - __smp_locks_end = .; - . = ALIGN(PAGE_SIZE); - } + PERCPU_VADDR(0, :percpu) +#endif - /* 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 = .; } +#ifdef CONFIG_X86_64 + :init +#endif .init.data : AT(ADDR(.init.data) - LOAD_OFFSET) { INIT_DATA @@ -333,17 +297,7 @@ SECTIONS } #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 +#if !defined(CONFIG_X86_64) || !defined(CONFIG_SMP) PERCPU(PAGE_SIZE) #endif @@ -354,15 +308,22 @@ SECTIONS __init_end = .; } + /* + * 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); + } + #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 */ + NOSAVE_DATA + } #endif /* BSS */ @@ -400,8 +361,8 @@ SECTIONS #ifdef CONFIG_X86_32 -ASSERT((_end - LOAD_OFFSET <= KERNEL_IMAGE_SIZE), - "kernel image bigger than KERNEL_IMAGE_SIZE") +. = 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 @@ -414,12 +375,12 @@ 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") +. = 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"); +. = ASSERT((per_cpu__irq_stack_union == 0), + "irq_stack_union is not at start of per-cpu area"); #endif #endif /* CONFIG_X86_32 */ @@ -427,7 +388,7 @@ ASSERT((per_cpu__irq_stack_union == 0), #ifdef CONFIG_KEXEC #include <asm/kexec.h> -ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE, - "kexec control code size is too big") +. = ASSERT(kexec_control_code_size <= KEXEC_CONTROL_CODE_MAX_SIZE, + "kexec control code size is too big"); #endif |