summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--.mailmap2
-rw-r--r--Documentation/CodingStyle16
-rw-r--r--Documentation/filesystems/Locking3
-rw-r--r--Documentation/filesystems/proc.txt23
-rw-r--r--Documentation/filesystems/vfs.txt4
-rw-r--r--Documentation/sysctl/fs.txt7
-rw-r--r--Documentation/vm/pagemap.txt2
-rw-r--r--arch/alpha/include/asm/posix_types.h3
-rw-r--r--arch/alpha/kernel/signal.c20
-rw-r--r--arch/arm/include/asm/posix_types.h3
-rw-r--r--arch/arm/kernel/signal.c49
-rw-r--r--arch/arm/kernel/smp.c8
-rw-r--r--arch/avr32/include/asm/posix_types.h3
-rw-r--r--arch/avr32/kernel/entry-avr32b.S4
-rw-r--r--arch/avr32/kernel/signal.c43
-rw-r--r--arch/blackfin/include/asm/posix_types.h3
-rw-r--r--arch/blackfin/include/asm/thread_info.h2
-rw-r--r--arch/blackfin/kernel/signal.c69
-rw-r--r--arch/blackfin/kernel/trace.c32
-rw-r--r--arch/blackfin/mach-common/entry.S2
-rw-r--r--arch/c6x/kernel/signal.c45
-rw-r--r--arch/cris/arch-v10/kernel/signal.c34
-rw-r--r--arch/cris/arch-v32/kernel/signal.c36
-rw-r--r--arch/cris/include/asm/posix_types.h3
-rw-r--r--arch/cris/kernel/ptrace.c2
-rw-r--r--arch/frv/include/asm/posix_types.h3
-rw-r--r--arch/frv/include/asm/thread_info.h16
-rw-r--r--arch/frv/kernel/entry.S29
-rw-r--r--arch/frv/kernel/signal.c59
-rw-r--r--arch/h8300/include/asm/posix_types.h3
-rw-r--r--arch/h8300/kernel/signal.c30
-rw-r--r--arch/hexagon/kernel/signal.c50
-rw-r--r--arch/ia64/include/asm/posix_types.h3
-rw-r--r--arch/ia64/include/asm/thread_info.h18
-rw-r--r--arch/ia64/kernel/perfmon.c10
-rw-r--r--arch/ia64/kernel/process.c2
-rw-r--r--arch/ia64/kernel/signal.c34
-rw-r--r--arch/ia64/kernel/sys_ia64.c19
-rw-r--r--arch/m32r/include/asm/posix_types.h3
-rw-r--r--arch/m32r/kernel/signal.c34
-rw-r--r--arch/m68k/include/asm/posix_types.h3
-rw-r--r--arch/m68k/kernel/signal.c29
-rw-r--r--arch/microblaze/include/asm/thread_info.h18
-rw-r--r--arch/microblaze/kernel/signal.c41
-rw-r--r--arch/mips/include/asm/posix_types.h5
-rw-r--r--arch/mips/include/asm/stat.h6
-rw-r--r--arch/mips/kernel/signal-common.h2
-rw-r--r--arch/mips/kernel/signal.c40
-rw-r--r--arch/mips/kernel/signal32.c2
-rw-r--r--arch/mips/kernel/signal_n32.c1
-rw-r--r--arch/mn10300/include/asm/posix_types.h3
-rw-r--r--arch/mn10300/kernel/signal.c41
-rw-r--r--arch/openrisc/kernel/signal.c42
-rw-r--r--arch/parisc/include/asm/posix_types.h3
-rw-r--r--arch/parisc/include/asm/stat.h4
-rw-r--r--arch/parisc/include/asm/thread_info.h2
-rw-r--r--arch/parisc/kernel/entry.S4
-rw-r--r--arch/parisc/kernel/signal.c49
-rw-r--r--arch/parisc/kernel/signal32.c2
-rw-r--r--arch/powerpc/include/asm/posix_types.h3
-rw-r--r--arch/powerpc/include/asm/stat.h4
-rw-r--r--arch/powerpc/include/asm/thread_info.h18
-rw-r--r--arch/powerpc/kernel/signal.c38
-rw-r--r--arch/powerpc/kernel/signal.h3
-rw-r--r--arch/powerpc/kernel/signal_32.c4
-rw-r--r--arch/powerpc/kernel/signal_64.c4
-rw-r--r--arch/powerpc/mm/mmu_context_nohash.c11
-rw-r--r--arch/s390/include/asm/posix_types.h3
-rw-r--r--arch/s390/kernel/compat_signal.c12
-rw-r--r--arch/s390/kernel/entry.h2
-rw-r--r--arch/s390/kernel/signal.c49
-rw-r--r--arch/score/kernel/signal.c42
-rw-r--r--arch/sh/include/asm/posix_types_32.h2
-rw-r--r--arch/sh/include/asm/posix_types_64.h2
-rw-r--r--arch/sh/include/asm/thread_info.h19
-rw-r--r--arch/sh/kernel/signal_32.c45
-rw-r--r--arch/sh/kernel/signal_64.c49
-rw-r--r--arch/sh/kernel/smp.c7
-rw-r--r--arch/sparc/include/asm/posix_types.h5
-rw-r--r--arch/sparc/include/asm/thread_info_32.h3
-rw-r--r--arch/sparc/include/asm/thread_info_64.h18
-rw-r--r--arch/sparc/kernel/signal32.c27
-rw-r--r--arch/sparc/kernel/signal_32.c41
-rw-r--r--arch/sparc/kernel/signal_64.c36
-rw-r--r--arch/sparc/kernel/sys_sparc_64.c11
-rw-r--r--arch/tile/include/asm/compat.h1
-rw-r--r--arch/tile/include/asm/thread_info.h18
-rw-r--r--arch/tile/kernel/compat_signal.c3
-rw-r--r--arch/tile/kernel/process.c2
-rw-r--r--arch/tile/kernel/signal.c42
-rw-r--r--arch/um/include/shared/frame_kern.h3
-rw-r--r--arch/um/kernel/process.c5
-rw-r--r--arch/um/kernel/reboot.c13
-rw-r--r--arch/um/kernel/signal.c37
-rw-r--r--arch/um/kernel/trap.c24
-rw-r--r--arch/unicore32/kernel/signal.c49
-rw-r--r--arch/x86/ia32/ia32_signal.c2
-rw-r--r--arch/x86/include/asm/posix_types_32.h3
-rw-r--r--arch/x86/include/asm/sighandling.h2
-rw-r--r--arch/x86/include/asm/thread_info.h18
-rw-r--r--arch/x86/kernel/entry_32.S13
-rw-r--r--arch/x86/kernel/signal.c62
-rw-r--r--arch/x86/syscalls/syscall_32.tbl1
-rw-r--r--arch/x86/syscalls/syscall_64.tbl2
-rw-r--r--arch/x86/um/signal.c2
-rw-r--r--arch/xtensa/kernel/signal.c26
-rw-r--r--drivers/base/soc.c2
-rw-r--r--drivers/gpu/drm/cirrus/cirrus_drv.c19
-rw-r--r--drivers/gpu/drm/cirrus/cirrus_drv.h2
-rw-r--r--drivers/gpu/drm/cirrus/cirrus_ttm.c5
-rw-r--r--drivers/gpu/drm/drm_edid.c11
-rw-r--r--drivers/gpu/drm/i810/i810_dma.c4
-rw-r--r--drivers/gpu/drm/i915/i915_drv.h3
-rw-r--r--drivers/gpu/drm/i915/i915_gem_dmabuf.c61
-rw-r--r--drivers/gpu/drm/mgag200/mgag200_drv.c19
-rw-r--r--drivers/gpu/drm/nouveau/nouveau_drv.h3
-rw-r--r--drivers/gpu/drm/nouveau/nouveau_prime.c45
-rw-r--r--drivers/gpu/drm/radeon/evergreen.c382
-rw-r--r--drivers/gpu/drm/radeon/evergreend.h11
-rw-r--r--drivers/gpu/drm/radeon/ni.c360
-rw-r--r--drivers/gpu/drm/radeon/nid.h11
-rw-r--r--drivers/gpu/drm/radeon/r600.c199
-rw-r--r--drivers/gpu/drm/radeon/r600d.h2
-rw-r--r--drivers/gpu/drm/radeon/radeon.h8
-rw-r--r--drivers/gpu/drm/radeon/radeon_cs.c31
-rw-r--r--drivers/gpu/drm/radeon/radeon_prime.c44
-rw-r--r--drivers/gpu/drm/radeon/rv770.c274
-rw-r--r--drivers/gpu/drm/radeon/rv770d.h4
-rw-r--r--drivers/gpu/drm/ttm/ttm_bo.c1
-rw-r--r--drivers/gpu/drm/udl/udl_fb.c13
-rw-r--r--drivers/gpu/drm/udl/udl_gem.c25
-rw-r--r--drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c2
-rw-r--r--drivers/message/fusion/mptbase.c12
-rw-r--r--drivers/message/fusion/mptctl.c4
-rw-r--r--drivers/rapidio/Kconfig14
-rw-r--r--drivers/rapidio/devices/Makefile3
-rw-r--r--drivers/rapidio/devices/tsi721.c211
-rw-r--r--drivers/rapidio/devices/tsi721.h105
-rw-r--r--drivers/rapidio/devices/tsi721_dma.c823
-rw-r--r--drivers/rapidio/rio.c81
-rw-r--r--fs/9p/vfs_inode_dotl.c24
-rw-r--r--fs/affs/affs.h8
-rw-r--r--fs/aio.c10
-rw-r--r--fs/attr.c5
-rw-r--r--fs/binfmt_elf.c8
-rw-r--r--fs/binfmt_flat.c8
-rw-r--r--fs/btrfs/acl.c4
-rw-r--r--fs/btrfs/backref.c495
-rw-r--r--fs/btrfs/backref.h3
-rw-r--r--fs/btrfs/btrfs_inode.h50
-rw-r--r--fs/btrfs/check-integrity.c584
-rw-r--r--fs/btrfs/ctree.c861
-rw-r--r--fs/btrfs/ctree.h78
-rw-r--r--fs/btrfs/delayed-inode.c8
-rw-r--r--fs/btrfs/delayed-ref.c10
-rw-r--r--fs/btrfs/delayed-ref.h24
-rw-r--r--fs/btrfs/disk-io.c57
-rw-r--r--fs/btrfs/disk-io.h1
-rw-r--r--fs/btrfs/export.c15
-rw-r--r--fs/btrfs/extent-tree.c23
-rw-r--r--fs/btrfs/extent_io.c168
-rw-r--r--fs/btrfs/extent_io.h8
-rw-r--r--fs/btrfs/file.c78
-rw-r--r--fs/btrfs/free-space-cache.c52
-rw-r--r--fs/btrfs/inode.c317
-rw-r--r--fs/btrfs/ioctl.c50
-rw-r--r--fs/btrfs/ioctl.h33
-rw-r--r--fs/btrfs/ordered-data.c165
-rw-r--r--fs/btrfs/ordered-data.h13
-rw-r--r--fs/btrfs/print-tree.c3
-rw-r--r--fs/btrfs/reada.c5
-rw-r--r--fs/btrfs/scrub.c65
-rw-r--r--fs/btrfs/super.c117
-rw-r--r--fs/btrfs/transaction.c59
-rw-r--r--fs/btrfs/tree-log.c35
-rw-r--r--fs/btrfs/ulist.c38
-rw-r--r--fs/btrfs/ulist.h15
-rw-r--r--fs/btrfs/volumes.c306
-rw-r--r--fs/btrfs/volumes.h52
-rw-r--r--fs/btrfs/xattr.c1
-rw-r--r--fs/buffer.c2
-rw-r--r--fs/ceph/export.c32
-rw-r--r--fs/ceph/snap.c2
-rw-r--r--fs/compat.c43
-rw-r--r--fs/dcache.c20
-rw-r--r--fs/ecryptfs/inode.c48
-rw-r--r--fs/eventfd.c12
-rw-r--r--fs/eventpoll.c4
-rw-r--r--fs/exec.c4
-rw-r--r--fs/exportfs/expfs.c33
-rw-r--r--fs/ext4/Kconfig2
-rw-r--r--fs/ext4/balloc.c41
-rw-r--r--fs/ext4/bitmap.c83
-rw-r--r--fs/ext4/dir.c12
-rw-r--r--fs/ext4/ext4.h130
-rw-r--r--fs/ext4/ext4_extents.h24
-rw-r--r--fs/ext4/ext4_jbd2.c9
-rw-r--r--fs/ext4/ext4_jbd2.h7
-rw-r--r--fs/ext4/extents.c91
-rw-r--r--fs/ext4/file.c2
-rw-r--r--fs/ext4/ialloc.c81
-rw-r--r--fs/ext4/inode.c119
-rw-r--r--fs/ext4/ioctl.c19
-rw-r--r--fs/ext4/mballoc.c30
-rw-r--r--fs/ext4/mmp.c44
-rw-r--r--fs/ext4/namei.c445
-rw-r--r--fs/ext4/resize.c71
-rw-r--r--fs/ext4/super.c253
-rw-r--r--fs/ext4/xattr.c92
-rw-r--r--fs/ext4/xattr.h4
-rw-r--r--fs/fat/dir.c4
-rw-r--r--fs/fat/fat.h6
-rw-r--r--fs/fat/fatent.c21
-rw-r--r--fs/fat/inode.c63
-rw-r--r--fs/fcntl.c42
-rw-r--r--fs/file_table.c17
-rw-r--r--fs/fuse/file.c4
-rw-r--r--fs/fuse/inode.c17
-rw-r--r--fs/gfs2/export.c17
-rw-r--r--fs/hpfs/alloc.c14
-rw-r--r--fs/hpfs/anode.c43
-rw-r--r--fs/hpfs/buffer.c1
-rw-r--r--fs/hpfs/dir.c2
-rw-r--r--fs/hpfs/dnode.c10
-rw-r--r--fs/hpfs/ea.c60
-rw-r--r--fs/hpfs/hpfs.h289
-rw-r--r--fs/hpfs/hpfs_fn.h23
-rw-r--r--fs/hpfs/inode.c2
-rw-r--r--fs/hpfs/map.c20
-rw-r--r--fs/hpfs/namei.c2
-rw-r--r--fs/hpfs/super.c4
-rw-r--r--fs/inode.c124
-rw-r--r--fs/internal.h3
-rw-r--r--fs/isofs/export.c13
-rw-r--r--fs/jbd2/Kconfig2
-rw-r--r--fs/jbd2/commit.c70
-rw-r--r--fs/jbd2/journal.c132
-rw-r--r--fs/jbd2/recovery.c126
-rw-r--r--fs/jbd2/revoke.c27
-rw-r--r--fs/jbd2/transaction.c4
-rw-r--r--fs/jffs2/jffs2_fs_sb.h4
-rw-r--r--fs/jffs2/os-linux.h7
-rw-r--r--fs/jffs2/super.c21
-rw-r--r--fs/jffs2/wbuf.c55
-rw-r--r--fs/lockd/clntlock.c13
-rw-r--r--fs/lockd/svc.c148
-rw-r--r--fs/locks.c5
-rw-r--r--fs/namei.c177
-rw-r--r--fs/namespace.c142
-rw-r--r--fs/ncpfs/file.c6
-rw-r--r--fs/ncpfs/ncp_fs_sb.h10
-rw-r--r--fs/nfs/callback.c13
-rw-r--r--fs/nfs/dir.c56
-rw-r--r--fs/nfs/file.c77
-rw-r--r--fs/nfsd/auth.c2
-rw-r--r--fs/nfsd/export.c181
-rw-r--r--fs/nfsd/fault_inject.c1
-rw-r--r--fs/nfsd/idmap.h8
-rw-r--r--fs/nfsd/netns.h6
-rw-r--r--fs/nfsd/nfs4callback.c5
-rw-r--r--fs/nfsd/nfs4idmap.c113
-rw-r--r--fs/nfsd/nfs4recover.c4
-rw-r--r--fs/nfsd/nfs4state.c538
-rw-r--r--fs/nfsd/nfs4xdr.c62
-rw-r--r--fs/nfsd/nfsctl.c67
-rw-r--r--fs/nfsd/nfsfh.c2
-rw-r--r--fs/nfsd/nfssvc.c31
-rw-r--r--fs/nfsd/state.h1
-rw-r--r--fs/nfsd/vfs.c2
-rw-r--r--fs/nfsd/xdr4.h6
-rw-r--r--fs/nilfs2/file.c24
-rw-r--r--fs/nilfs2/ioctl.c8
-rw-r--r--fs/nilfs2/namei.c22
-rw-r--r--fs/nls/Kconfig157
-rw-r--r--fs/nls/Makefile23
-rw-r--r--fs/nls/nls_macceltic.c602
-rw-r--r--fs/nls/nls_maccenteuro.c532
-rw-r--r--fs/nls/nls_maccroatian.c602
-rw-r--r--fs/nls/nls_maccyrillic.c497
-rw-r--r--fs/nls/nls_macgaelic.c567
-rw-r--r--fs/nls/nls_macgreek.c497
-rw-r--r--fs/nls/nls_maciceland.c602
-rw-r--r--fs/nls/nls_macinuit.c532
-rw-r--r--fs/nls/nls_macroman.c637
-rw-r--r--fs/nls/nls_macromanian.c602
-rw-r--r--fs/nls/nls_macturkish.c602
-rw-r--r--fs/notify/fsnotify.c12
-rw-r--r--fs/ntfs/file.c4
-rw-r--r--fs/ocfs2/blockcheck.c42
-rw-r--r--fs/ocfs2/dlm/dlmast.c2
-rw-r--r--fs/ocfs2/dlm/dlmcommon.h6
-rw-r--r--fs/ocfs2/dlm/dlmdomain.c2
-rw-r--r--fs/ocfs2/export.c19
-rw-r--r--fs/ocfs2/inode.c13
-rw-r--r--fs/ocfs2/ioctl.c31
-rw-r--r--fs/ocfs2/move_extents.c6
-rw-r--r--fs/ocfs2/namei.c5
-rw-r--r--fs/ocfs2/symlink.c115
-rw-r--r--fs/ocfs2/symlink.h2
-rw-r--r--fs/open.c76
-rw-r--r--fs/pipe.c9
-rw-r--r--fs/pnode.c4
-rw-r--r--fs/proc/array.c147
-rw-r--r--fs/proc/base.c81
-rw-r--r--fs/proc/internal.h3
-rw-r--r--fs/proc/task_mmu.c82
-rw-r--r--fs/proc/task_nommu.c2
-rw-r--r--fs/proc_namespace.c4
-rw-r--r--fs/read_write.c7
-rw-r--r--fs/readdir.c33
-rw-r--r--fs/reiserfs/inode.c30
-rw-r--r--fs/reiserfs/journal.c15
-rw-r--r--fs/reiserfs/reiserfs.h12
-rw-r--r--fs/reiserfs/resize.c1
-rw-r--r--fs/reiserfs/super.c74
-rw-r--r--fs/select.c4
-rw-r--r--fs/signalfd.c7
-rw-r--r--fs/splice.c6
-rw-r--r--fs/statfs.c5
-rw-r--r--fs/sync.c5
-rw-r--r--fs/ubifs/dir.c11
-rw-r--r--fs/udf/namei.c14
-rw-r--r--fs/utimes.c5
-rw-r--r--fs/xattr.c20
-rw-r--r--fs/xfs/kmem.c10
-rw-r--r--fs/xfs/kmem.h21
-rw-r--r--fs/xfs/xfs_export.c23
-rw-r--r--fs/xfs/xfs_file.c7
-rw-r--r--fs/xfs/xfs_log.c2
-rw-r--r--fs/xfs/xfs_log_priv.h2
-rw-r--r--fs/xfs/xfs_trans.c2
-rw-r--r--fs/xfs/xfs_trans.h2
-rw-r--r--include/asm-generic/bitsperlong.h4
-rw-r--r--include/asm-generic/posix_types.h4
-rw-r--r--include/drm/drm_mem_util.h4
-rw-r--r--include/linux/Kbuild1
-rw-r--r--include/linux/compat.h3
-rw-r--r--include/linux/cpu.h1
-rw-r--r--include/linux/cred.h10
-rw-r--r--include/linux/dmaengine.h12
-rw-r--r--include/linux/errno.h1
-rw-r--r--include/linux/eventfd.h2
-rw-r--r--include/linux/exportfs.h4
-rw-r--r--include/linux/fs.h22
-rw-r--r--include/linux/fsnotify_backend.h2
-rw-r--r--include/linux/interrupt.h4
-rw-r--r--include/linux/ipc_namespace.h42
-rw-r--r--include/linux/jbd2.h59
-rw-r--r--include/linux/jbd_common.h2
-rw-r--r--include/linux/kcmp.h17
-rw-r--r--include/linux/kernel.h1
-rw-r--r--include/linux/kexec.h75
-rw-r--r--include/linux/key.h4
-rw-r--r--include/linux/kmod.h34
-rw-r--r--include/linux/lglock.h179
-rw-r--r--include/linux/lockd/bind.h4
-rw-r--r--include/linux/mm.h2
-rw-r--r--include/linux/msdos_fs.h3
-rw-r--r--include/linux/nfsd/export.h13
-rw-r--r--include/linux/prctl.h6
-rw-r--r--include/linux/rio.h47
-rw-r--r--include/linux/rio_drv.h9
-rw-r--r--include/linux/sched.h28
-rw-r--r--include/linux/security.h40
-rw-r--r--include/linux/signal.h5
-rw-r--r--include/linux/slab.h2
-rw-r--r--include/linux/sunrpc/svc.h2
-rw-r--r--include/linux/sunrpc/svcauth.h13
-rw-r--r--include/linux/sunrpc/svcauth_gss.h1
-rw-r--r--include/linux/syscalls.h2
-rw-r--r--include/linux/task_work.h33
-rw-r--r--include/linux/thread_info.h19
-rw-r--r--include/linux/tracehook.h13
-rw-r--r--include/linux/types.h2
-rw-r--r--init/Kconfig11
-rw-r--r--init/do_mounts.c14
-rw-r--r--init/do_mounts_initrd.c10
-rw-r--r--init/do_mounts_md.c12
-rw-r--r--init/do_mounts_rd.c13
-rw-r--r--init/initramfs.c16
-rw-r--r--ipc/mq_sysctl.c49
-rw-r--r--ipc/mqueue.c292
-rw-r--r--ipc/shm.c7
-rw-r--r--kernel/Makefile7
-rw-r--r--kernel/cpu.c44
-rw-r--r--kernel/cpu_pm.c16
-rw-r--r--kernel/cred.c9
-rw-r--r--kernel/exit.c13
-rw-r--r--kernel/fork.c11
-rw-r--r--kernel/irq/manage.c80
-rw-r--r--kernel/kcmp.c196
-rw-r--r--kernel/kmod.c30
-rw-r--r--kernel/lglock.c89
-rw-r--r--kernel/pid_namespace.c13
-rw-r--r--kernel/resource.c4
-rw-r--r--kernel/signal.c59
-rw-r--r--kernel/sys.c213
-rw-r--r--kernel/sys_ni.c3
-rw-r--r--kernel/task_work.c84
-rw-r--r--lib/vsprintf.c289
-rw-r--r--mm/cleancache.c6
-rw-r--r--mm/filemap.c69
-rw-r--r--mm/filemap_xip.c4
-rw-r--r--mm/internal.h4
-rw-r--r--mm/mmap.c54
-rw-r--r--mm/mremap.c26
-rw-r--r--mm/nommu.c35
-rw-r--r--mm/process_vm_access.c16
-rw-r--r--mm/shmem.c6
-rw-r--r--mm/util.c30
-rw-r--r--net/sched/sch_atm.c2
-rw-r--r--net/sunrpc/auth_gss/gss_krb5_wrap.c61
-rw-r--r--net/sunrpc/auth_gss/svcauth_gss.c127
-rw-r--r--net/sunrpc/rpcb_clnt.c12
-rw-r--r--net/sunrpc/svc.c23
-rw-r--r--net/sunrpc/svc_xprt.c4
-rw-r--r--net/sunrpc/svcauth_unix.c19
-rwxr-xr-xscripts/checkpatch.pl20
-rw-r--r--security/apparmor/lsm.c15
-rw-r--r--security/capability.c3
-rw-r--r--security/commoncap.c17
-rw-r--r--security/keys/compat.c2
-rw-r--r--security/keys/internal.h2
-rw-r--r--security/keys/keyctl.c77
-rw-r--r--security/keys/process_keys.c20
-rw-r--r--security/keys/request_key.c13
-rw-r--r--security/security.c51
-rw-r--r--security/selinux/hooks.c15
-rw-r--r--security/selinux/selinuxfs.c36
-rw-r--r--security/smack/smack_lsm.c15
-rw-r--r--sound/pci/rme9652/hdspm.c7
-rw-r--r--sound/soc/fsl/imx-ssi.c6
-rw-r--r--sound/soc/sh/fsi.c25
-rw-r--r--sound/usb/pcm.c3
-rw-r--r--tools/testing/selftests/Makefile2
-rw-r--r--tools/testing/selftests/kcmp/Makefile29
-rw-r--r--tools/testing/selftests/kcmp/kcmp_test.c94
-rw-r--r--tools/testing/selftests/mqueue/.gitignore2
-rw-r--r--tools/testing/selftests/mqueue/Makefile10
-rw-r--r--tools/testing/selftests/mqueue/mq_open_tests.c492
-rw-r--r--tools/testing/selftests/mqueue/mq_perf_tests.c741
-rw-r--r--usr/Kconfig10
442 files changed, 18830 insertions, 6329 deletions
diff --git a/.mailmap b/.mailmap
index 9b0d0267a3c..2909c33bc54 100644
--- a/.mailmap
+++ b/.mailmap
@@ -113,3 +113,5 @@ Uwe Kleine-K├Ânig <Uwe.Kleine-Koenig@digi.com>
Valdis Kletnieks <Valdis.Kletnieks@vt.edu>
Takashi YOSHII <takashi.yoshii.zj@renesas.com>
Yusuke Goda <goda.yusuke@renesas.com>
+Gustavo Padovan <gustavo@las.ic.unicamp.br>
+Gustavo Padovan <padovan@profusion.mobi>
diff --git a/Documentation/CodingStyle b/Documentation/CodingStyle
index c58b236bbe0..cb9258b8fd3 100644
--- a/Documentation/CodingStyle
+++ b/Documentation/CodingStyle
@@ -671,8 +671,9 @@ ones already enabled by DEBUG.
Chapter 14: Allocating memory
The kernel provides the following general purpose memory allocators:
-kmalloc(), kzalloc(), kcalloc(), vmalloc(), and vzalloc(). Please refer to
-the API documentation for further information about them.
+kmalloc(), kzalloc(), kmalloc_array(), kcalloc(), vmalloc(), and
+vzalloc(). Please refer to the API documentation for further information
+about them.
The preferred form for passing a size of a struct is the following:
@@ -686,6 +687,17 @@ Casting the return value which is a void pointer is redundant. The conversion
from void pointer to any other pointer type is guaranteed by the C programming
language.
+The preferred form for allocating an array is the following:
+
+ p = kmalloc_array(n, sizeof(...), ...);
+
+The preferred form for allocating a zeroed array is the following:
+
+ p = kcalloc(n, sizeof(...), ...);
+
+Both forms check for overflow on the allocation size n * sizeof(...),
+and return NULL if that occurred.
+
Chapter 15: The inline disease
diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking
index d449e632e6a..8e2da1e06e3 100644
--- a/Documentation/filesystems/Locking
+++ b/Documentation/filesystems/Locking
@@ -61,6 +61,7 @@ ata *);
ssize_t (*listxattr) (struct dentry *, char *, size_t);
int (*removexattr) (struct dentry *, const char *);
int (*fiemap)(struct inode *, struct fiemap_extent_info *, u64 start, u64 len);
+ void (*update_time)(struct inode *, struct timespec *, int);
locking rules:
all may block
@@ -87,6 +88,8 @@ getxattr: no
listxattr: no
removexattr: yes
fiemap: no
+update_time: no
+
Additionally, ->rmdir(), ->unlink() and ->rename() have ->i_mutex on
victim.
cross-directory ->rename() has (per-superblock) ->s_vfs_rename_sem.
diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt
index 912af6ce562..fb0a6aeb936 100644
--- a/Documentation/filesystems/proc.txt
+++ b/Documentation/filesystems/proc.txt
@@ -40,6 +40,7 @@ Table of Contents
3.4 /proc/<pid>/coredump_filter - Core dump filtering settings
3.5 /proc/<pid>/mountinfo - Information about mounts
3.6 /proc/<pid>/comm & /proc/<pid>/task/<tid>/comm
+ 3.7 /proc/<pid>/task/<tid>/children - Information about task children
4 Configuring procfs
4.1 Mount options
@@ -310,6 +311,11 @@ Table 1-4: Contents of the stat files (as of 2.6.30-rc7)
start_data address above which program data+bss is placed
end_data address below which program data+bss is placed
start_brk address above which program heap can be expanded with brk()
+ arg_start address above which program command line is placed
+ arg_end address below which program command line is placed
+ env_start address above which program environment is placed
+ env_end address below which program environment is placed
+ exit_code the thread's exit_code in the form reported by the waitpid system call
..............................................................................
The /proc/PID/maps file containing the currently mapped memory regions and
@@ -1578,6 +1584,23 @@ then the kernel's TASK_COMM_LEN (currently 16 chars) will result in a truncated
comm value.
+3.7 /proc/<pid>/task/<tid>/children - Information about task children
+-------------------------------------------------------------------------
+This file provides a fast way to retrieve first level children pids
+of a task pointed by <pid>/<tid> pair. The format is a space separated
+stream of pids.
+
+Note the "first level" here -- if a child has own children they will
+not be listed here, one needs to read /proc/<children-pid>/task/<tid>/children
+to obtain the descendants.
+
+Since this interface is intended to be fast and cheap it doesn't
+guarantee to provide precise results and some children might be
+skipped, especially if they've exited right after we printed their
+pids, so one need to either stop or freeze processes being inspected
+if precise results are needed.
+
+
------------------------------------------------------------------------------
Configuring procfs
------------------------------------------------------------------------------
diff --git a/Documentation/filesystems/vfs.txt b/Documentation/filesystems/vfs.txt
index ef19f91a0f1..efd23f48170 100644
--- a/Documentation/filesystems/vfs.txt
+++ b/Documentation/filesystems/vfs.txt
@@ -363,6 +363,7 @@ struct inode_operations {
ssize_t (*getxattr) (struct dentry *, const char *, void *, size_t);
ssize_t (*listxattr) (struct dentry *, char *, size_t);
int (*removexattr) (struct dentry *, const char *);
+ void (*update_time)(struct inode *, struct timespec *, int);
};
Again, all methods are called without any locks being held, unless
@@ -471,6 +472,9 @@ otherwise noted.
removexattr: called by the VFS to remove an extended attribute from
a file. This method is called by removexattr(2) system call.
+ update_time: called by the VFS to update a specific time or the i_version of
+ an inode. If this is not defined the VFS will update the inode itself
+ and call mark_inode_dirty_sync.
The Address Space Object
========================
diff --git a/Documentation/sysctl/fs.txt b/Documentation/sysctl/fs.txt
index 88fd7f5c8dc..13d6166d7a2 100644
--- a/Documentation/sysctl/fs.txt
+++ b/Documentation/sysctl/fs.txt
@@ -225,6 +225,13 @@ a queue must be less or equal then msg_max.
maximum message size value (it is every message queue's attribute set during
its creation).
+/proc/sys/fs/mqueue/msg_default is a read/write file for setting/getting the
+default number of messages in a queue value if attr parameter of mq_open(2) is
+NULL. If it exceed msg_max, the default value is initialized msg_max.
+
+/proc/sys/fs/mqueue/msgsize_default is a read/write file for setting/getting
+the default message size value if attr parameter of mq_open(2) is NULL. If it
+exceed msgsize_max, the default value is initialized msgsize_max.
4. /proc/sys/fs/epoll - Configuration options for the epoll interface
--------------------------------------------------------
diff --git a/Documentation/vm/pagemap.txt b/Documentation/vm/pagemap.txt
index 4600cbe3d6b..7587493c67f 100644
--- a/Documentation/vm/pagemap.txt
+++ b/Documentation/vm/pagemap.txt
@@ -16,7 +16,7 @@ There are three components to pagemap:
* Bits 0-4 swap type if swapped
* Bits 5-54 swap offset if swapped
* Bits 55-60 page shift (page size = 1<<page shift)
- * Bit 61 reserved for future use
+ * Bit 61 page is file-page or shared-anon
* Bit 62 page swapped
* Bit 63 page present
diff --git a/arch/alpha/include/asm/posix_types.h b/arch/alpha/include/asm/posix_types.h
index 24779fc9599..5a8a48320ef 100644
--- a/arch/alpha/include/asm/posix_types.h
+++ b/arch/alpha/include/asm/posix_types.h
@@ -10,9 +10,6 @@
typedef unsigned int __kernel_ino_t;
#define __kernel_ino_t __kernel_ino_t
-typedef unsigned int __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned long __kernel_sigset_t; /* at least 32 bits */
#include <asm-generic/posix_types.h>
diff --git a/arch/alpha/kernel/signal.c b/arch/alpha/kernel/signal.c
index 10ab2d74ecb..a8c97d42ec8 100644
--- a/arch/alpha/kernel/signal.c
+++ b/arch/alpha/kernel/signal.c
@@ -226,7 +226,6 @@ do_sigreturn(struct sigcontext __user *sc, struct pt_regs *regs,
if (__get_user(set.sig[0], &sc->sc_mask))
goto give_sigsegv;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(sc, regs, sw))
@@ -261,7 +260,6 @@ do_rt_sigreturn(struct rt_sigframe __user *frame, struct pt_regs *regs,
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto give_sigsegv;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(&frame->uc.uc_mcontext, regs, sw))
@@ -468,12 +466,9 @@ static inline void
handle_signal(int sig, struct k_sigaction *ka, siginfo_t *info,
struct pt_regs * regs, struct switch_stack *sw)
{
- sigset_t *oldset = &current->blocked;
+ sigset_t *oldset = sigmask_to_save();
int ret;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
-
if (ka->sa.sa_flags & SA_SIGINFO)
ret = setup_rt_frame(sig, ka, info, oldset, regs, sw);
else
@@ -483,12 +478,7 @@ handle_signal(int sig, struct k_sigaction *ka, siginfo_t *info,
force_sigsegv(sig, current);
return;
}
- block_sigmask(ka, sig);
- /* A signal was successfully delivered, and the
- saved sigmask was stored on the signal frame,
- and will be restored by sigreturn. So we can
- simply clear the restore sigmask flag. */
- clear_thread_flag(TIF_RESTORE_SIGMASK);
+ signal_delivered(sig, info, ka, regs, 0);
}
static inline void
@@ -572,9 +562,7 @@ do_signal(struct pt_regs * regs, struct switch_stack * sw,
}
/* If there's no signal to deliver, we just restore the saved mask. */
- if (test_and_clear_thread_flag(TIF_RESTORE_SIGMASK))
- set_current_blocked(&current->saved_sigmask);
-
+ restore_saved_sigmask();
if (single_stepping)
ptrace_set_bpt(current); /* re-set breakpoint */
}
@@ -590,7 +578,5 @@ do_notify_resume(struct pt_regs *regs, struct switch_stack *sw,
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/arm/include/asm/posix_types.h b/arch/arm/include/asm/posix_types.h
index efdf99045d8..d2de9cbbcd9 100644
--- a/arch/arm/include/asm/posix_types.h
+++ b/arch/arm/include/asm/posix_types.h
@@ -22,9 +22,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c
index 17fc36c41cf..fd2392a17ac 100644
--- a/arch/arm/kernel/signal.c
+++ b/arch/arm/kernel/signal.c
@@ -22,8 +22,6 @@
#include "signal.h"
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/*
* For ARM syscalls, we encode the syscall number into the instruction.
*/
@@ -210,10 +208,8 @@ static int restore_sigframe(struct pt_regs *regs, struct sigframe __user *sf)
int err;
err = __copy_from_user(&set, &sf->uc.uc_sigmask, sizeof(set));
- if (err == 0) {
- sigdelsetmask(&set, ~_BLOCKABLE);
+ if (err == 0)
set_current_blocked(&set);
- }
__get_user_error(regs->ARM_r0, &sf->uc.uc_mcontext.arm_r0, err);
__get_user_error(regs->ARM_r1, &sf->uc.uc_mcontext.arm_r1, err);
@@ -528,13 +524,13 @@ setup_rt_frame(int usig, struct k_sigaction *ka, siginfo_t *info,
/*
* OK, we're invoking a handler
*/
-static int
+static void
handle_signal(unsigned long sig, struct k_sigaction *ka,
- siginfo_t *info, sigset_t *oldset,
- struct pt_regs * regs)
+ siginfo_t *info, struct pt_regs *regs)
{
struct thread_info *thread = current_thread_info();
struct task_struct *tsk = current;
+ sigset_t *oldset = sigmask_to_save();
int usig = sig;
int ret;
@@ -559,17 +555,9 @@ handle_signal(unsigned long sig, struct k_sigaction *ka,
if (ret != 0) {
force_sigsegv(sig, tsk);
- return ret;
+ return;
}
-
- /*
- * Block the signal if we were successful.
- */
- block_sigmask(ka, sig);
-
- tracehook_signal_handler(sig, info, ka, regs, 0);
-
- return 0;
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -617,8 +605,6 @@ static void do_signal(struct pt_regs *regs, int syscall)
*/
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
- sigset_t *oldset;
-
/*
* Depending on the signal settings we may need to revert the
* decision to restart the system call. But skip this if a
@@ -635,20 +621,7 @@ static void do_signal(struct pt_regs *regs, int syscall)
clear_thread_flag(TIF_SYSCALL_RESTARTSYS);
}
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
- if (handle_signal(signr, &ka, &info, oldset, regs) == 0) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag.
- */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
+ handle_signal(signr, &ka, &info, regs);
return;
}
@@ -663,11 +636,7 @@ static void do_signal(struct pt_regs *regs, int syscall)
set_thread_flag(TIF_SYSCALL_RESTARTSYS);
}
- /* If there's no signal to deliver, we just put the saved sigmask
- * back.
- */
- if (test_and_clear_thread_flag(TIF_RESTORE_SIGMASK))
- set_current_blocked(&current->saved_sigmask);
+ restore_saved_sigmask();
}
asmlinkage void
@@ -679,7 +648,5 @@ do_notify_resume(struct pt_regs *regs, unsigned int thread_flags, int syscall)
if (thread_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c
index b735521a4a5..2c7217d971d 100644
--- a/arch/arm/kernel/smp.c
+++ b/arch/arm/kernel/smp.c
@@ -109,7 +109,6 @@ static void percpu_timer_stop(void);
int __cpu_disable(void)
{
unsigned int cpu = smp_processor_id();
- struct task_struct *p;
int ret;
ret = platform_cpu_disable(cpu);
@@ -139,12 +138,7 @@ int __cpu_disable(void)
flush_cache_all();
local_flush_tlb_all();
- read_lock(&tasklist_lock);
- for_each_process(p) {
- if (p->mm)
- cpumask_clear_cpu(cpu, mm_cpumask(p->mm));
- }
- read_unlock(&tasklist_lock);
+ clear_tasks_mm_cpumask(cpu);
return 0;
}
diff --git a/arch/avr32/include/asm/posix_types.h b/arch/avr32/include/asm/posix_types.h
index 74667bfc88c..9ba9e749b3f 100644
--- a/arch/avr32/include/asm/posix_types.h
+++ b/arch/avr32/include/asm/posix_types.h
@@ -17,9 +17,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/avr32/kernel/entry-avr32b.S b/arch/avr32/kernel/entry-avr32b.S
index 169268c40ae..df288418131 100644
--- a/arch/avr32/kernel/entry-avr32b.S
+++ b/arch/avr32/kernel/entry-avr32b.S
@@ -281,7 +281,7 @@ syscall_exit_work:
ld.w r1, r0[TI_flags]
rjmp 1b
-2: mov r2, _TIF_SIGPENDING | _TIF_RESTORE_SIGMASK | _TIF_NOTIFY_RESUME
+2: mov r2, _TIF_SIGPENDING | _TIF_NOTIFY_RESUME
tst r1, r2
breq 3f
unmask_interrupts
@@ -587,7 +587,7 @@ fault_exit_work:
ld.w r1, r0[TI_flags]
rjmp fault_exit_work
-1: mov r2, _TIF_SIGPENDING | _TIF_RESTORE_SIGMASK
+1: mov r2, _TIF_SIGPENDING | _TIF_NOTIFY_RESUME
tst r1, r2
breq 2f
unmask_interrupts
diff --git a/arch/avr32/kernel/signal.c b/arch/avr32/kernel/signal.c
index ae386c304be..c140f9b41dc 100644
--- a/arch/avr32/kernel/signal.c
+++ b/arch/avr32/kernel/signal.c
@@ -22,8 +22,6 @@
#include <asm/ucontext.h>
#include <asm/syscalls.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
asmlinkage int sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss,
struct pt_regs *regs)
{
@@ -89,7 +87,6 @@ asmlinkage int sys_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext))
@@ -224,30 +221,27 @@ static inline void setup_syscall_restart(struct pt_regs *regs)
static inline void
handle_signal(unsigned long sig, struct k_sigaction *ka, siginfo_t *info,
- sigset_t *oldset, struct pt_regs *regs, int syscall)
+ struct pt_regs *regs, int syscall)
{
int ret;
/*
* Set up the stack frame
*/
- ret = setup_rt_frame(sig, ka, info, oldset, regs);
+ ret = setup_rt_frame(sig, ka, info, sigmask_to_save(), regs);
/*
* Check that the resulting registers are sane
*/
ret |= !valid_user_regs(regs);
- if (ret != 0) {
- force_sigsegv(sig, current);
- return;
- }
-
/*
* Block the signal if we were successful.
*/
- block_sigmask(ka, sig);
- clear_thread_flag(TIF_RESTORE_SIGMASK);
+ if (ret != 0)
+ force_sigsegv(sig, current);
+ else
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -255,7 +249,7 @@ handle_signal(unsigned long sig, struct k_sigaction *ka, siginfo_t *info,
* doesn't want to handle. Thus you cannot kill init even with a
* SIGKILL even by mistake.
*/
-int do_signal(struct pt_regs *regs, sigset_t *oldset, int syscall)
+static void do_signal(struct pt_regs *regs, int syscall)
{
siginfo_t info;
int signr;
@@ -267,12 +261,7 @@ int do_signal(struct pt_regs *regs, sigset_t *oldset, int syscall)
* without doing anything if so.
*/
if (!user_mode(regs))
- return 0;
-
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else if (!oldset)
- oldset = &current->blocked;
+ return;
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (syscall) {
@@ -297,15 +286,11 @@ int do_signal(struct pt_regs *regs, sigset_t *oldset, int syscall)
if (signr == 0) {
/* No signal to deliver -- put the saved sigmask back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
- return 0;
+ restore_saved_sigmask();
+ return;
}
- handle_signal(signr, &ka, &info, oldset, regs, syscall);
- return 1;
+ handle_signal(signr, &ka, &info, regs, syscall);
}
asmlinkage void do_notify_resume(struct pt_regs *regs, struct thread_info *ti)
@@ -315,13 +300,11 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, struct thread_info *ti)
if ((sysreg_read(SR) & MODE_MASK) == MODE_SUPERVISOR)
syscall = 1;
- if (ti->flags & (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK))
- do_signal(regs, &current->blocked, syscall);
+ if (ti->flags & _TIF_SIGPENDING))
+ do_signal(regs, syscall);
if (ti->flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/blackfin/include/asm/posix_types.h b/arch/blackfin/include/asm/posix_types.h
index 41bc1875c4d..1bd3436db6a 100644
--- a/arch/blackfin/include/asm/posix_types.h
+++ b/arch/blackfin/include/asm/posix_types.h
@@ -10,9 +10,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned int __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/blackfin/include/asm/thread_info.h b/arch/blackfin/include/asm/thread_info.h
index 02560fd8a12..53ad10005ae 100644
--- a/arch/blackfin/include/asm/thread_info.h
+++ b/arch/blackfin/include/asm/thread_info.h
@@ -100,7 +100,6 @@ static inline struct thread_info *current_thread_info(void)
TIF_NEED_RESCHED */
#define TIF_MEMDIE 4 /* is terminating due to OOM killer */
#define TIF_RESTORE_SIGMASK 5 /* restore signal mask in do_signal() */
-#define TIF_FREEZE 6 /* is freezing for suspend */
#define TIF_IRQ_SYNC 7 /* sync pipeline stage */
#define TIF_NOTIFY_RESUME 8 /* callback before returning to user */
#define TIF_SINGLESTEP 9
@@ -111,7 +110,6 @@ static inline struct thread_info *current_thread_info(void)
#define _TIF_NEED_RESCHED (1<<TIF_NEED_RESCHED)
#define _TIF_POLLING_NRFLAG (1<<TIF_POLLING_NRFLAG)
#define _TIF_RESTORE_SIGMASK (1<<TIF_RESTORE_SIGMASK)
-#define _TIF_FREEZE (1<<TIF_FREEZE)
#define _TIF_IRQ_SYNC (1<<TIF_IRQ_SYNC)
#define _TIF_NOTIFY_RESUME (1<<TIF_NOTIFY_RESUME)
#define _TIF_SINGLESTEP (1<<TIF_SINGLESTEP)
diff --git a/arch/blackfin/kernel/signal.c b/arch/blackfin/kernel/signal.c
index e5bbc1a5edc..6682b73a852 100644
--- a/arch/blackfin/kernel/signal.c
+++ b/arch/blackfin/kernel/signal.c
@@ -19,8 +19,6 @@
#include <asm/fixed_code.h>
#include <asm/syscall.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/* Location of the trace bit in SYSCFG. */
#define TRACE_BITS 0x0001
@@ -98,7 +96,6 @@ asmlinkage int do_rt_sigreturn(unsigned long __unused)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (rt_restore_sigcontext(regs, &frame->uc.uc_mcontext, &r0))
@@ -190,17 +187,22 @@ setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t * info,
err |= copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
if (err)
- goto give_sigsegv;
+ return -EFAULT;
/* Set up registers for signal handler */
- wrusp((unsigned long)frame);
if (current->personality & FDPIC_FUNCPTRS) {
struct fdpic_func_descriptor __user *funcptr =
(struct fdpic_func_descriptor *) ka->sa.sa_handler;
- __get_user(regs->pc, &funcptr->text);
- __get_user(regs->p3, &funcptr->GOT);
+ u32 pc, p3;
+ err |= __get_user(pc, &funcptr->text);
+ err |= __get_user(p3, &funcptr->GOT);
+ if (err)
+ return -EFAULT;
+ regs->pc = pc;
+ regs->p3 = p3;
} else
regs->pc = (unsigned long)ka->sa.sa_handler;
+ wrusp((unsigned long)frame);
regs->rets = SIGRETURN_STUB;
regs->r0 = frame->sig;
@@ -208,10 +210,6 @@ setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t * info,
regs->r2 = (unsigned long)(&frame->uc);
return 0;
-
- give_sigsegv:
- force_sigsegv(sig, current);
- return -EFAULT;
}
static inline void
@@ -247,24 +245,21 @@ handle_restart(struct pt_regs *regs, struct k_sigaction *ka, int has_handler)
/*
* OK, we're invoking a handler
*/
-static int
+static void
handle_signal(int sig, siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs *regs)
+ struct pt_regs *regs)
{
- int ret;
-
/* are we from a system call? to see pt_regs->orig_p0 */
if (regs->orig_p0 >= 0)
/* If so, check system call restarting.. */
handle_restart(regs, ka, 1);
/* set up the stack frame */
- ret = setup_rt_frame(sig, ka, info, oldset, regs);
-
- if (ret == 0)
- block_sigmask(ka, sig);
-
- return ret;
+ if (setup_rt_frame(sig, ka, info, sigmask_to_save(), regs) < 0)
+ force_sigsegv(sig, current);
+ else
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLESTEP));
}
/*
@@ -281,37 +276,16 @@ asmlinkage void do_signal(struct pt_regs *regs)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
current->thread.esp0 = (unsigned long)regs;
- if (try_to_freeze())
- goto no_signal;
-
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
- if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
- /* a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
-
- tracehook_signal_handler(signr, &info, &ka, regs,
- test_thread_flag(TIF_SINGLESTEP));
- }
-
+ handle_signal(signr, &info, &ka, regs);
return;
}
- no_signal:
/* Did we come from a system call? */
if (regs->orig_p0 >= 0)
/* Restart the system call - no handlers present */
@@ -319,10 +293,7 @@ asmlinkage void do_signal(struct pt_regs *regs)
/* if there's no signal to deliver, we just put the saved sigmask
* back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
/*
@@ -330,14 +301,12 @@ asmlinkage void do_signal(struct pt_regs *regs)
*/
asmlinkage void do_notify_resume(struct pt_regs *regs)
{
- if (test_thread_flag(TIF_SIGPENDING) || test_thread_flag(TIF_RESTORE_SIGMASK))
+ if (test_thread_flag(TIF_SIGPENDING))
do_signal(regs);
if (test_thread_flag(TIF_NOTIFY_RESUME)) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/blackfin/kernel/trace.c b/arch/blackfin/kernel/trace.c
index 44bbf2f564c..f7f7a18abca 100644
--- a/arch/blackfin/kernel/trace.c
+++ b/arch/blackfin/kernel/trace.c
@@ -10,6 +10,8 @@
#include <linux/hardirq.h>
#include <linux/thread_info.h>
#include <linux/mm.h>
+#include <linux/oom.h>
+#include <linux/sched.h>
#include <linux/uaccess.h>
#include <linux/module.h>
#include <linux/kallsyms.h>
@@ -27,8 +29,7 @@ void decode_address(char *buf, unsigned long address)
{
struct task_struct *p;
struct mm_struct *mm;
- unsigned long flags, offset;
- unsigned char in_atomic = (bfin_read_IPEND() & 0x10) || in_atomic();
+ unsigned long offset;
struct rb_node *n;
#ifdef CONFIG_KALLSYMS
@@ -112,17 +113,17 @@ void decode_address(char *buf, unsigned long address)
* mappings of all our processes and see if we can't be a whee
* bit more specific
*/
- write_lock_irqsave(&tasklist_lock, flags);
+ read_lock(&tasklist_lock);
for_each_process(p) {
- mm = (in_atomic ? p->mm : get_task_mm(p));
- if (!mm)
- continue;
+ struct task_struct *t;
- if (!down_read_trylock(&mm->mmap_sem)) {
- if (!in_atomic)
- mmput(mm);
+ t = find_lock_task_mm(p);
+ if (!t)
continue;
- }
+
+ mm = t->mm;
+ if (!down_read_trylock(&mm->mmap_sem))
+ goto __continue;
for (n = rb_first(&mm->mm_rb); n; n = rb_next(n)) {
struct vm_area_struct *vma;
@@ -131,7 +132,7 @@ void decode_address(char *buf, unsigned long address)
if (address >= vma->vm_start && address < vma->vm_end) {
char _tmpbuf[256];
- char *name = p->comm;
+ char *name = t->comm;
struct file *file = vma->vm_file;
if (file) {
@@ -164,8 +165,7 @@ void decode_address(char *buf, unsigned long address)
name, vma->vm_start, vma->vm_end);
up_read(&mm->mmap_sem);
- if (!in_atomic)
- mmput(mm);
+ task_unlock(t);
if (buf[0] == '\0')
sprintf(buf, "[ %s ] dynamic memory", name);
@@ -175,8 +175,8 @@ void decode_address(char *buf, unsigned long address)
}
up_read(&mm->mmap_sem);
- if (!in_atomic)
- mmput(mm);
+__continue:
+ task_unlock(t);
}
/*
@@ -186,7 +186,7 @@ void decode_address(char *buf, unsigned long address)
sprintf(buf, "/* kernel dynamic memory */");
done:
- write_unlock_irqrestore(&tasklist_lock, flags);
+ read_unlock(&tasklist_lock);
}
#define EXPAND_LEN ((1 << CONFIG_DEBUG_BFIN_HWTRACE_EXPAND_LEN) * 256 - 1)
diff --git a/arch/blackfin/mach-common/entry.S b/arch/blackfin/mach-common/entry.S
index 80aa2535e2c..04c2fbe41a7 100644
--- a/arch/blackfin/mach-common/entry.S
+++ b/arch/blackfin/mach-common/entry.S
@@ -711,8 +711,6 @@ ENTRY(_system_call)
jump .Lresume_userspace_1;
.Lsyscall_sigpending:
- cc = BITTST(r7, TIF_RESTORE_SIGMASK);
- if cc jump .Lsyscall_do_signals;
cc = BITTST(r7, TIF_SIGPENDING);
if cc jump .Lsyscall_do_signals;
cc = BITTST(r7, TIF_NOTIFY_RESUME);
diff --git a/arch/c6x/kernel/signal.c b/arch/c6x/kernel/signal.c
index cf37478c116..3d8f3c22a94 100644
--- a/arch/c6x/kernel/signal.c
+++ b/arch/c6x/kernel/signal.c
@@ -20,8 +20,6 @@
#include <asm/cacheflush.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/*
* Do a signal return, undo the signal stack.
*/
@@ -87,7 +85,6 @@ asmlinkage int do_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext))
@@ -248,10 +245,9 @@ do_restart:
/*
* handle the actual delivery of a signal to userspace
*/
-static int handle_signal(int sig,
+static void handle_signal(int sig,
siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs *regs,
- int syscall)
+ struct pt_regs *regs, int syscall)
{
int ret;
@@ -278,11 +274,9 @@ static int handle_signal(int sig,
}
/* Set up the stack frame */
- ret = setup_rt_frame(sig, ka, info, oldset, regs);
- if (ret == 0)
- block_sigmask(ka, sig);
-
- return ret;
+ if (setup_rt_frame(sig, ka, info, sigmask_to_save(), regs) < 0)
+ return;
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -292,7 +286,6 @@ static void do_signal(struct pt_regs *regs, int syscall)
{
struct k_sigaction ka;
siginfo_t info;
- sigset_t *oldset;
int signr;
/* we want the common case to go fast, which is why we may in certain
@@ -300,25 +293,9 @@ static void do_signal(struct pt_regs *regs, int syscall)
if (!user_mode(regs))
return;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
- if (handle_signal(signr, &info, &ka, oldset,
- regs, syscall) == 0) {
- /* a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
-
- tracehook_signal_handler(signr, &info, &ka, regs, 0);
- }
-
+ handle_signal(signr, &info, &ka, regs, syscall);
return;
}
@@ -343,10 +320,7 @@ static void do_signal(struct pt_regs *regs, int syscall)
/* if there's no signal to deliver, we just put the saved sigmask
* back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
/*
@@ -357,14 +331,11 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, u32 thread_info_flags,
int syscall)
{
/* deal with pending signal delivery */
- if (thread_info_flags & ((1 << TIF_SIGPENDING) |
- (1 << TIF_RESTORE_SIGMASK)))
+ if (thread_info_flags & (1 << TIF_SIGPENDING))
do_signal(regs, syscall);
if (thread_info_flags & (1 << TIF_NOTIFY_RESUME)) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/cris/arch-v10/kernel/signal.c b/arch/cris/arch-v10/kernel/signal.c
index e16f8f297f6..0bb477c13a4 100644
--- a/arch/cris/arch-v10/kernel/signal.c
+++ b/arch/cris/arch-v10/kernel/signal.c
@@ -31,8 +31,6 @@
#define DEBUG_SIG 0
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/* a syscall in Linux/CRIS is a break 13 instruction which is 2 bytes */
/* manipulate regs so that upon return, it will be re-executed */
@@ -176,7 +174,6 @@ asmlinkage int sys_sigreturn(long r10, long r11, long r12, long r13, long mof,
sizeof(frame->extramask))))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->sc))
@@ -212,7 +209,6 @@ asmlinkage int sys_rt_sigreturn(long r10, long r11, long r12, long r13,
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext))
@@ -415,10 +411,11 @@ give_sigsegv:
* OK, we're invoking a handler
*/
-static inline int handle_signal(int canrestart, unsigned long sig,
+static inline void handle_signal(int canrestart, unsigned long sig,
siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs *regs)
+ struct pt_regs *regs)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* Are we from a system call? */
@@ -456,9 +453,7 @@ static inline int handle_signal(int canrestart, unsigned long sig,
ret = setup_frame(sig, ka, oldset, regs);
if (ret == 0)
- block_sigmask(ka, sig);
-
- return ret;
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -478,7 +473,6 @@ void do_signal(int canrestart, struct pt_regs *regs)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
/*
* We want the common case to go fast, which
@@ -489,23 +483,10 @@ void do_signal(int canrestart, struct pt_regs *regs)
if (!user_mode(regs))
return;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
- if (handle_signal(canrestart, signr, &info, &ka,
- oldset, regs)) {
- /* a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
+ handle_signal(canrestart, signr, &info, &ka, regs);
return;
}
@@ -525,8 +506,5 @@ void do_signal(int canrestart, struct pt_regs *regs)
/* if there's no signal to deliver, we just put the saved sigmask
* back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
diff --git a/arch/cris/arch-v32/kernel/signal.c b/arch/cris/arch-v32/kernel/signal.c
index b338d8fc0c1..b60d1b65a42 100644
--- a/arch/cris/arch-v32/kernel/signal.c
+++ b/arch/cris/arch-v32/kernel/signal.c
@@ -24,9 +24,6 @@
extern unsigned long cris_signal_return_page;
-/* Flag to check if a signal is blockable. */
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/*
* A syscall in CRIS is really a "break 13" instruction, which is 2
* bytes. The registers is manipulated so upon return the instruction
@@ -167,7 +164,6 @@ sys_sigreturn(long r10, long r11, long r12, long r13, long mof, long srp,
sizeof(frame->extramask))))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->sc))
@@ -208,7 +204,6 @@ sys_rt_sigreturn(long r10, long r11, long r12, long r13, long mof, long srp,
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext))
@@ -434,11 +429,12 @@ give_sigsegv:
}
/* Invoke a signal handler to, well, handle the signal. */
-static inline int
+static inline void
handle_signal(int canrestart, unsigned long sig,
siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs * regs)
+ struct pt_regs * regs)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* Check if this got called from a system call. */
@@ -489,9 +485,7 @@ handle_signal(int canrestart, unsigned long sig,
ret = setup_frame(sig, ka, oldset, regs);
if (ret == 0)
- block_sigmask(ka, sig);
-
- return ret;
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -511,7 +505,6 @@ do_signal(int canrestart, struct pt_regs *regs)
int signr;
siginfo_t info;
struct k_sigaction ka;
- sigset_t *oldset;
/*
* The common case should go fast, which is why this point is
@@ -521,25 +514,11 @@ do_signal(int canrestart, struct pt_regs *regs)
if (!user_mode(regs))
return;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
- if (handle_signal(canrestart, signr, &info, &ka,
- oldset, regs)) {
- /* a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
-
+ handle_signal(canrestart, signr, &info, &ka, regs);
return;
}
@@ -560,10 +539,7 @@ do_signal(int canrestart, struct pt_regs *regs)
/* if there's no signal to deliver, we just put the saved sigmask
* back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
asmlinkage void
diff --git a/arch/cris/include/asm/posix_types.h b/arch/cris/include/asm/posix_types.h
index 234891c74e2..ce4e5179315 100644
--- a/arch/cris/include/asm/posix_types.h
+++ b/arch/cris/include/asm/posix_types.h
@@ -15,9 +15,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/cris/kernel/ptrace.c b/arch/cris/kernel/ptrace.c
index d114ad3da9b..58d44ee1a71 100644
--- a/arch/cris/kernel/ptrace.c
+++ b/arch/cris/kernel/ptrace.c
@@ -40,7 +40,5 @@ void do_notify_resume(int canrestart, struct pt_regs *regs,
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/frv/include/asm/posix_types.h b/arch/frv/include/asm/posix_types.h
index 3f34cb45fbb..fe512af74a5 100644
--- a/arch/frv/include/asm/posix_types.h
+++ b/arch/frv/include/asm/posix_types.h
@@ -10,9 +10,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/frv/include/asm/thread_info.h b/arch/frv/include/asm/thread_info.h
index 54ab13a0de4..0ff03a33c81 100644
--- a/arch/frv/include/asm/thread_info.h
+++ b/arch/frv/include/asm/thread_info.h
@@ -94,8 +94,8 @@ register struct thread_info *__current_thread_info asm("gr15");
#define TIF_NEED_RESCHED 3 /* rescheduling necessary */
#define TIF_SINGLESTEP 4 /* restore singlestep on return to user mode */
#define TIF_RESTORE_SIGMASK 5 /* restore signal mask in do_signal() */
-#define TIF_POLLING_NRFLAG 16 /* true if poll_idle() is polling TIF_NEED_RESCHED */
-#define TIF_MEMDIE 17 /* is terminating due to OOM killer */
+#define TIF_POLLING_NRFLAG 6 /* true if poll_idle() is polling TIF_NEED_RESCHED */
+#define TIF_MEMDIE 7 /* is terminating due to OOM killer */
#define _TIF_SYSCALL_TRACE (1 << TIF_SYSCALL_TRACE)
#define _TIF_NOTIFY_RESUME (1 << TIF_NOTIFY_RESUME)
@@ -105,8 +105,16 @@ register struct thread_info *__current_thread_info asm("gr15");
#define _TIF_RESTORE_SIGMASK (1 << TIF_RESTORE_SIGMASK)
#define _TIF_POLLING_NRFLAG (1 << TIF_POLLING_NRFLAG)
-#define _TIF_WORK_MASK 0x0000FFFE /* work to do on interrupt/exception return */
-#define _TIF_ALLWORK_MASK 0x0000FFFF /* work to do on any return to u-space */
+/* work to do on interrupt/exception return */
+#define _TIF_WORK_MASK \
+ (_TIF_NOTIFY_RESUME | _TIF_SIGPENDING | _TIF_NEED_RESCHED | _TIF_SINGLESTEP)
+
+/* work to do on any return to u-space */
+#define _TIF_ALLWORK_MASK (_TIF_WORK_MASK | _TIF_SYSCALL_TRACE)
+
+#if _TIF_ALLWORK_MASK >= 0x2000
+#error "_TIF_ALLWORK_MASK won't fit in an ANDI now (see entry.S)"
+#endif
/*
* Thread-synchronous status.
diff --git a/arch/frv/kernel/entry.S b/arch/frv/kernel/entry.S
index 5ba23f715ea..7d5e000fd32 100644
--- a/arch/frv/kernel/entry.S
+++ b/arch/frv/kernel/entry.S
@@ -905,18 +905,19 @@ __syscall_call:
__syscall_exit:
LEDS 0x6300
- sti gr8,@(gr28,#REG_GR(8)) ; save return value
+ # keep current PSR in GR23
+ movsg psr,gr23
- # rebuild saved psr - execve will change it for init/main.c
ldi @(gr28,#REG_PSR),gr22
+
+ sti.p gr8,@(gr28,#REG_GR(8)) ; save return value
+
+ # rebuild saved psr - execve will change it for init/main.c
srli gr22,#1,gr5
andi.p gr22,#~PSR_PS,gr22
andi gr5,#PSR_PS,gr5
or gr5,gr22,gr22
- ori gr22,#PSR_S,gr22
-
- # keep current PSR in GR23
- movsg psr,gr23
+ ori.p gr22,#PSR_S,gr22
# make sure we don't miss an interrupt setting need_resched or sigpending between
# sampling and the RETT
@@ -924,9 +925,7 @@ __syscall_exit:
movgs gr23,psr
ldi @(gr15,#TI_FLAGS),gr4
- sethi.p %hi(_TIF_ALLWORK_MASK),gr5
- setlo %lo(_TIF_ALLWORK_MASK),gr5
- andcc gr4,gr5,gr0,icc0
+ andicc gr4,#_TIF_ALLWORK_MASK,gr0,icc0
bne icc0,#0,__syscall_exit_work
# restore all registers and return
@@ -1111,9 +1110,7 @@ __entry_resume_userspace:
__entry_return_from_user_interrupt:
LEDS 0x6402
ldi @(gr15,#TI_FLAGS),gr4
- sethi.p %hi(_TIF_WORK_MASK),gr5
- setlo %lo(_TIF_WORK_MASK),gr5
- andcc gr4,gr5,gr0,icc0
+ andicc gr4,#_TIF_WORK_MASK,gr0,icc0
beq icc0,#1,__entry_return_direct
__entry_work_pending:
@@ -1133,9 +1130,7 @@ __entry_work_resched:
LEDS 0x6401
ldi @(gr15,#TI_FLAGS),gr4
- sethi.p %hi(_TIF_WORK_MASK),gr5
- setlo %lo(_TIF_WORK_MASK),gr5
- andcc gr4,gr5,gr0,icc0
+ andicc gr4,#_TIF_WORK_MASK,gr0,icc0
beq icc0,#1,__entry_return_direct
andicc gr4,#_TIF_NEED_RESCHED,gr0,icc0
bne icc0,#1,__entry_work_resched
@@ -1163,7 +1158,9 @@ __syscall_trace_entry:
# perform syscall exit tracing
__syscall_exit_work:
LEDS 0x6340
- andicc gr4,#_TIF_SYSCALL_TRACE,gr0,icc0
+ andicc gr22,#PSR_PS,gr0,icc1 ; don't handle on return to kernel mode
+ andicc.p gr4,#_TIF_SYSCALL_TRACE,gr0,icc0
+ bne icc1,#0,__entry_return_direct
beq icc0,#1,__entry_work_pending
movsg psr,gr23
diff --git a/arch/frv/kernel/signal.c b/arch/frv/kernel/signal.c
index 8cf5dca0175..f3b9064c548 100644
--- a/arch/frv/kernel/signal.c
+++ b/arch/frv/kernel/signal.c
@@ -28,8 +28,6 @@
#define DEBUG_SIG 0
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
struct fdpic_func_descriptor {
unsigned long text;
unsigned long GOT;
@@ -149,7 +147,6 @@ asmlinkage int sys_sigreturn(void)
__copy_from_user(&set.sig[1], &frame->extramask, sizeof(frame->extramask)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(&frame->sc, &gr8))
@@ -172,7 +169,6 @@ asmlinkage int sys_rt_sigreturn(void)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(&frame->uc.uc_mcontext, &gr8))
@@ -426,9 +422,10 @@ give_sigsegv:
/*
* OK, we're invoking a handler
*/
-static int handle_signal(unsigned long sig, siginfo_t *info,
- struct k_sigaction *ka, sigset_t *oldset)
+static void handle_signal(unsigned long sig, siginfo_t *info,
+ struct k_sigaction *ka)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* Are we from a system call? */
@@ -460,11 +457,11 @@ static int handle_signal(unsigned long sig, siginfo_t *info,
else
ret = setup_frame(sig, ka, oldset);
- if (ret == 0)
- block_sigmask(ka, sig);
-
- return ret;
+ if (ret)
+ return;
+ signal_delivered(sig, info, ka, __frame,
+ test_thread_flag(TIF_SINGLESTEP));
} /* end handle_signal() */
/*****************************************************************************/
@@ -477,44 +474,14 @@ static void do_signal(void)
{
struct k_sigaction ka;
siginfo_t info;
- sigset_t *oldset;
int signr;
- /*
- * We want the common case to go fast, which
- * is why we may in certain cases get here from
- * kernel mode. Just return without doing anything
- * if so.
- */
- if (!user_mode(__frame))
- return;
-
- if (try_to_freeze())
- goto no_signal;
-
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, __frame, NULL);
if (signr > 0) {
- if (handle_signal(signr, &info, &ka, oldset) == 0) {
- /* a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
-
- tracehook_signal_handler(signr, &info, &ka, __frame,
- test_thread_flag(TIF_SINGLESTEP));
- }
-
+ handle_signal(signr, &info, &ka);
return;
}
-no_signal:
/* Did we come from a system call? */
if (__frame->syscallno != -1) {
/* Restart the system call - no handlers present */
@@ -536,11 +503,7 @@ no_signal:
/* if there's no signal to deliver, we just put the saved sigmask
* back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
-
+ restore_saved_sigmask();
} /* end do_signal() */
/*****************************************************************************/
@@ -555,15 +518,13 @@ asmlinkage void do_notify_resume(__u32 thread_info_flags)
clear_thread_flag(TIF_SINGLESTEP);
/* deal with pending signal delivery */
- if (thread_info_flags & (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK))
+ if (thread_info_flags & _TIF_SIGPENDING))
do_signal();
/* deal with notification on about to resume userspace execution */
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(__frame);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
} /* end do_notify_resume() */
diff --git a/arch/h8300/include/asm/posix_types.h b/arch/h8300/include/asm/posix_types.h
index bc4c34efb1a..91e62ba4c7b 100644
--- a/arch/h8300/include/asm/posix_types.h
+++ b/arch/h8300/include/asm/posix_types.h
@@ -10,9 +10,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/h8300/kernel/signal.c b/arch/h8300/kernel/signal.c
index d4b0555d290..fca10378701 100644
--- a/arch/h8300/kernel/signal.c
+++ b/arch/h8300/kernel/signal.c
@@ -47,8 +47,6 @@
#include <asm/traps.h>
#include <asm/ucontext.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/*
* Atomically swap in the new signal mask, and wait for a signal.
*/
@@ -186,7 +184,6 @@ asmlinkage int do_sigreturn(unsigned long __unused,...)
sizeof(frame->extramask))))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->sc, &er0))
@@ -211,7 +208,6 @@ asmlinkage int do_rt_sigreturn(unsigned long __unused,...)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &er0))
@@ -412,8 +408,9 @@ give_sigsegv:
*/
static void
handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs * regs)
+ struct pt_regs * regs)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* are we from a system call? */
if (regs->orig_er0 >= 0) {
@@ -441,10 +438,8 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
else
ret = setup_frame(sig, ka, oldset, regs);
- if (!ret) {
- block_sigmask(ka, sig);
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
+ if (!ret)
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -457,7 +452,6 @@ statis void do_signal(struct pt_regs *regs)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
/*
* We want the common case to go fast, which
@@ -468,23 +462,14 @@ statis void do_signal(struct pt_regs *regs)
if ((regs->ccr & 0x10))
return;
- if (try_to_freeze())
- goto no_signal;
-
current->thread.esp0 = (unsigned long) regs;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
- handle_signal(signr, &info, &ka, oldset, regs);
+ handle_signal(signr, &info, &ka, regs);
return;
}
- no_signal:
/* Did we come from a system call? */
if (regs->orig_er0 >= 0) {
/* Restart the system call - no handlers present */
@@ -501,8 +486,7 @@ statis void do_signal(struct pt_regs *regs)
}
/* If there's no signal to deliver, we just restore the saved mask. */
- if (test_and_clear_thread_flag(TIF_RESTORE_SIGMASK))
- set_current_blocked(&current->saved_sigmask);
+ restore_saved_sigmask();
}
asmlinkage void do_notify_resume(struct pt_regs *regs, u32 thread_info_flags)
@@ -513,7 +497,5 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, u32 thread_info_flags)
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/hexagon/kernel/signal.c b/arch/hexagon/kernel/signal.c
index 434866eb0f1..304b0808d07 100644
--- a/arch/hexagon/kernel/signal.c
+++ b/arch/hexagon/kernel/signal.c
@@ -31,8 +31,6 @@
#include <asm/signal.h>
#include <asm/vdso.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
struct rt_sigframe {
unsigned long tramp[2];
struct siginfo info;
@@ -149,11 +147,9 @@ sigsegv:
/*
* Setup invocation of signal handler
*/
-static int handle_signal(int sig, siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs *regs)
+static void handle_signal(int sig, siginfo_t *info, struct k_sigaction *ka,
+ struct pt_regs *regs)
{
- int rc;
-
/*
* If we're handling a signal that aborted a system call,
* set up the error return value before adding the signal
@@ -186,15 +182,12 @@ static int handle_signal(int sig, siginfo_t *info, struct k_sigaction *ka,
* Set up the stack frame; not doing the SA_SIGINFO thing. We
* only set up the rt_frame flavor.
*/
- rc = setup_rt_frame(sig, ka, info, oldset, regs);
-
/* If there was an error on setup, no signal was delivered. */
- if (rc)
- return rc;
-
- block_sigmask(ka, sig);
+ if (setup_rt_frame(sig, ka, info, sigmask_to_save(), regs) < 0)
+ return;
- return 0;
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLESTEP));
}
/*
@@ -209,34 +202,13 @@ static void do_signal(struct pt_regs *regs)
if (!user_mode(regs))
return;
- if (try_to_freeze())
- goto no_signal;
-
signo = get_signal_to_deliver(&info, &sigact, regs, NULL);
if (signo > 0) {
- sigset_t *oldset;
-
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
- if (handle_signal(signo, &info, &sigact, oldset, regs) == 0) {
- /*
- * Successful delivery case. The saved sigmask is
- * stored in the signal frame, and will be restored
- * by sigreturn. We can clear the TIF flag.
- */
- clear_thread_flag(TIF_RESTORE_SIGMASK);
-
- tracehook_signal_handler(signo, &info, &sigact, regs,
- test_thread_flag(TIF_SINGLESTEP));
- }
+ handle_signal(signo, &info, &sigact, regs);
return;
}
-no_signal:
/*
* If we came from a system call, handle the restart.
*/
@@ -259,10 +231,7 @@ no_signal:
no_restart:
/* If there's no signal to deliver, put the saved sigmask back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
void do_notify_resume(struct pt_regs *regs, unsigned long thread_info_flags)
@@ -273,8 +242,6 @@ void do_notify_resume(struct pt_regs *regs, unsigned long thread_info_flags)
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
@@ -303,7 +270,6 @@ asmlinkage int sys_rt_sigreturn(void)
if (__copy_from_user(&blocked, &frame->uc.uc_sigmask, sizeof(blocked)))
goto badframe;
- sigdelsetmask(&blocked, ~_BLOCKABLE);
set_current_blocked(&blocked);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext))
diff --git a/arch/ia64/include/asm/posix_types.h b/arch/ia64/include/asm/posix_types.h
index 7323ab9467e..99ee1d6510c 100644
--- a/arch/ia64/include/asm/posix_types.h
+++ b/arch/ia64/include/asm/posix_types.h
@@ -1,9 +1,6 @@
#ifndef _ASM_IA64_POSIX_TYPES_H
#define _ASM_IA64_POSIX_TYPES_H
-typedef unsigned int __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned long __kernel_sigset_t; /* at least 32 bits */
#include <asm-generic/posix_types.h>
diff --git a/arch/ia64/include/asm/thread_info.h b/arch/ia64/include/asm/thread_info.h
index 310d9734f02..f7ee8537831 100644
--- a/arch/ia64/include/asm/thread_info.h
+++ b/arch/ia64/include/asm/thread_info.h
@@ -141,7 +141,23 @@ static inline void set_restore_sigmask(void)
{
struct thread_info *ti = current_thread_info();
ti->status |= TS_RESTORE_SIGMASK;
- set_bit(TIF_SIGPENDING, &ti->flags);
+ WARN_ON(!test_bit(TIF_SIGPENDING, &ti->flags));
+}
+static inline void clear_restore_sigmask(void)
+{
+ current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
+}
+static inline bool test_restore_sigmask(void)
+{
+ return current_thread_info()->status & TS_RESTORE_SIGMASK;
+}
+static inline bool test_and_clear_restore_sigmask(void)
+{
+ struct thread_info *ti = current_thread_info();
+ if (!(ti->status & TS_RESTORE_SIGMASK))
+ return false;
+ ti->status &= ~TS_RESTORE_SIGMASK;
+ return true;
}
#endif /* !__ASSEMBLY__ */
diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c
index f00ba025375..d7f558c1e71 100644
--- a/arch/ia64/kernel/perfmon.c
+++ b/arch/ia64/kernel/perfmon.c
@@ -604,12 +604,6 @@ pfm_unprotect_ctx_ctxsw(pfm_context_t *x, unsigned long f)
spin_unlock(&(x)->ctx_lock);
}
-static inline unsigned long
-pfm_get_unmapped_area(struct file *file, unsigned long addr, unsigned long len, unsigned long pgoff, unsigned long flags, unsigned long exec)
-{
- return get_unmapped_area(file, addr, len, pgoff, flags);
-}
-
/* forward declaration */
static const struct dentry_operations pfmfs_dentry_operations;
@@ -2333,8 +2327,8 @@ pfm_smpl_buffer_alloc(struct task_struct *task, struct file *filp, pfm_context_t
down_write(&task->mm->mmap_sem);
/* find some free area in address space, must have mmap sem held */
- vma->vm_start = pfm_get_unmapped_area(NULL, 0, size, 0, MAP_PRIVATE|MAP_ANONYMOUS, 0);
- if (vma->vm_start == 0UL) {
+ vma->vm_start = get_unmapped_area(NULL, 0, size, 0, MAP_PRIVATE|MAP_ANONYMOUS);
+ if (IS_ERR_VALUE(vma->vm_start)) {
DPRINT(("Cannot find unmapped area for size %ld\n", size));
up_write(&task->mm->mmap_sem);
goto error;
diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c
index 5e0e86ddb12..dd6fc144974 100644
--- a/arch/ia64/kernel/process.c
+++ b/arch/ia64/kernel/process.c
@@ -199,8 +199,6 @@ do_notify_resume_user(sigset_t *unused, struct sigscratch *scr, long in_syscall)
if (test_thread_flag(TIF_NOTIFY_RESUME)) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(&scr->pt);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
/* copy user rbs to kernel rbs */
diff --git a/arch/ia64/kernel/signal.c b/arch/ia64/kernel/signal.c
index 7523501d3bc..a199be1fe61 100644
--- a/arch/ia64/kernel/signal.c
+++ b/arch/ia64/kernel/signal.c
@@ -30,7 +30,6 @@
#define DEBUG_SIG 0
#define STACK_ALIGN 16 /* minimal alignment for stack pointer */
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
#if _NSIG_WORDS > 1
# define PUT_SIGSET(k,u) __copy_to_user((u)->sig, (k)->sig, sizeof(sigset_t))
@@ -200,7 +199,6 @@ ia64_rt_sigreturn (struct sigscratch *scr)
if (GET_SIGSET(&set, &sc->sc_mask))
goto give_sigsegv;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(sc, scr))
@@ -415,18 +413,13 @@ setup_frame (int sig, struct k_sigaction *ka, siginfo_t *info, sigset_t *set,
}
static long
-handle_signal (unsigned long sig, struct k_sigaction *ka, siginfo_t *info, sigset_t *oldset,
+handle_signal (unsigned long sig, struct k_sigaction *ka, siginfo_t *info,
struct sigscratch *scr)
{
- if (!setup_frame(sig, ka, info, oldset, scr))
+ if (!setup_frame(sig, ka, info, sigmask_to_save(), scr))
return 0;
- block_sigmask(ka, sig);
-
- /*
- * Let tracing know that we've done the handler setup.
- */
- tracehook_signal_handler(sig, info, ka, &scr->pt,
+ signal_delivered(sig, info, ka, &scr->pt,
test_thread_flag(TIF_SINGLESTEP));
return 1;
@@ -440,7 +433,6 @@ void
ia64_do_signal (struct sigscratch *scr, long in_syscall)
{
struct k_sigaction ka;
- sigset_t *oldset;
siginfo_t info;
long restart = in_syscall;
long errno = scr->pt.r8;
@@ -453,11 +445,6 @@ ia64_do_signal (struct sigscratch *scr, long in_syscall)
if (!user_mode(&scr->pt))
return;
- if (current_thread_info()->status & TS_RESTORE_SIGMASK)
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
/*
* This only loops in the rare cases of handle_signal() failing, in which case we
* need to push through a forced SIGSEGV.
@@ -507,16 +494,8 @@ ia64_do_signal (struct sigscratch *scr, long in_syscall)
* Whee! Actually deliver the signal. If the delivery failed, we need to
* continue to iterate in this loop so we can deliver the SIGSEGV...
*/
- if (handle_signal(signr, &ka, &info, oldset, scr)) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TS_RESTORE_SIGMASK flag.
- */
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
+ if (handle_signal(signr, &ka, &info, scr))
return;
- }
}
/* Did we come from a system call? */
@@ -538,8 +517,5 @@ ia64_do_signal (struct sigscratch *scr, long in_syscall)
/* if there's no signal to deliver, we just put the saved sigmask
* back */
- if (current_thread_info()->status & TS_RESTORE_SIGMASK) {
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
diff --git a/arch/ia64/kernel/sys_ia64.c b/arch/ia64/kernel/sys_ia64.c
index 609d50056a6..d9439ef2f66 100644
--- a/arch/ia64/kernel/sys_ia64.c
+++ b/arch/ia64/kernel/sys_ia64.c
@@ -171,22 +171,9 @@ asmlinkage unsigned long
ia64_mremap (unsigned long addr, unsigned long old_len, unsigned long new_len, unsigned long flags,
unsigned long new_addr)
{
- extern unsigned long do_mremap (unsigned long addr,
- unsigned long old_len,
- unsigned long new_len,
- unsigned long flags,
- unsigned long new_addr);
-
- down_write(&current->mm->mmap_sem);
- {
- addr = do_mremap(addr, old_len, new_len, flags, new_addr);
- }
- up_write(&current->mm->mmap_sem);
-
- if (IS_ERR((void *) addr))
- return addr;
-
- force_successful_syscall_return();
+ addr = sys_mremap(addr, old_len, new_len, flags, new_addr);
+ if (!IS_ERR((void *) addr))
+ force_successful_syscall_return();
return addr;
}
diff --git a/arch/m32r/include/asm/posix_types.h b/arch/m32r/include/asm/posix_types.h
index 0195850e1f8..236de26a409 100644
--- a/arch/m32r/include/asm/posix_types.h
+++ b/arch/m32r/include/asm/posix_types.h
@@ -10,9 +10,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/m32r/kernel/signal.c b/arch/m32r/kernel/signal.c
index f54d96993ea..f3fb2c029cf 100644
--- a/arch/m32r/kernel/signal.c
+++ b/arch/m32r/kernel/signal.c
@@ -28,8 +28,6 @@
#define DEBUG_SIG 0
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
asmlinkage int
sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss,
unsigned long r2, unsigned long r3, unsigned long r4,
@@ -111,7 +109,6 @@ sys_rt_sigreturn(unsigned long r0, unsigned long r1,
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &result))
@@ -267,9 +264,9 @@ static int prev_insn(struct pt_regs *regs)
* OK, we're invoking a handler
*/
-static int
+static void
handle_signal(unsigned long sig, struct k_sigaction *ka, siginfo_t *info,
- sigset_t *oldset, struct pt_regs *regs)
+ struct pt_regs *regs)
{
/* Are we from a system call? */
if (regs->syscall_nr >= 0) {
@@ -294,11 +291,10 @@ handle_signal(unsigned long sig, struct k_sigaction *ka, siginfo_t *info,
}
/* Set up the stack frame */
- if (setup_rt_frame(sig, ka, info, oldset, regs))
- return -EFAULT;
+ if (setup_rt_frame(sig, ka, info, sigmask_to_save(), regs))
+ return;
- block_sigmask(ka, sig);
- return 0;
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -311,7 +307,6 @@ static void do_signal(struct pt_regs *regs)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
/*
* We want the common case to go fast, which
@@ -322,14 +317,6 @@ static void do_signal(struct pt_regs *regs)
if (!user_mode(regs))
return;
- if (try_to_freeze())
- goto no_signal;
-
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Re-enable any watchpoints before delivering the
@@ -339,13 +326,11 @@ static void do_signal(struct pt_regs *regs)
*/
/* Whee! Actually deliver the signal. */
- if (handle_signal(signr, &ka, &info, oldset, regs) == 0)
- clear_thread_flag(TIF_RESTORE_SIGMASK);
+ handle_signal(signr, &ka, &info, regs);
return;
}
- no_signal:
/* Did we come from a system call? */
if (regs->syscall_nr >= 0) {
/* Restart the system call - no handlers present */
@@ -360,10 +345,7 @@ static void do_signal(struct pt_regs *regs)
prev_insn(regs);
}
}
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
/*
@@ -383,8 +365,6 @@ void do_notify_resume(struct pt_regs *regs, __u32 thread_info_flags)
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
clear_thread_flag(TIF_IRET);
diff --git a/arch/m68k/include/asm/posix_types.h b/arch/m68k/include/asm/posix_types.h
index 6373093be72..cf4dbf70fdc 100644
--- a/arch/m68k/include/asm/posix_types.h
+++ b/arch/m68k/include/asm/posix_types.h
@@ -10,9 +10,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/m68k/kernel/signal.c b/arch/m68k/kernel/signal.c
index d9f3d1900ee..710a528b928 100644
--- a/arch/m68k/kernel/signal.c
+++ b/arch/m68k/kernel/signal.c
@@ -51,8 +51,6 @@
#include <asm/traps.h>
#include <asm/ucontext.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
#ifdef CONFIG_MMU
/*
@@ -795,7 +793,6 @@ asmlinkage int do_sigreturn(unsigned long __unused)
sizeof(frame->extramask))))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->sc, frame + 1))
@@ -820,7 +817,6 @@ asmlinkage int do_rt_sigreturn(unsigned long __unused)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (rt_restore_ucontext(regs, sw, &frame->uc))
@@ -1123,8 +1119,9 @@ handle_restart(struct pt_regs *regs, struct k_sigaction *ka, int has_handler)
*/
static void
handle_signal(int sig, struct k_sigaction *ka, siginfo_t *info,
- sigset_t *oldset, struct pt_regs *regs)
+ struct pt_regs *regs)
{
+ sigset_t *oldset = sigmask_to_save();
int err;
/* are we from a system call? */
if (regs->orig_d0 >= 0)
@@ -1140,14 +1137,12 @@ handle_signal(int sig, struct k_sigaction *ka, siginfo_t *info,
if (err)
return;
- block_sigmask(ka, sig);
+ signal_delivered(sig, info, ka, regs, 0);
if (test_thread_flag(TIF_DELAYED_TRACE)) {
regs->sr &= ~0x8000;
send_sig(SIGTRAP, current, 1);
}
-
- clear_thread_flag(TIF_RESTORE_SIGMASK);
}
/*
@@ -1160,19 +1155,13 @@ static void do_signal(struct pt_regs *regs)
siginfo_t info;
struct k_sigaction ka;
int signr;
- sigset_t *oldset;
current->thread.esp0 = (unsigned long) regs;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
- handle_signal(signr, &ka, &info, oldset, regs);
+ handle_signal(signr, &ka, &info, regs);
return;
}
@@ -1182,10 +1171,7 @@ static void do_signal(struct pt_regs *regs)
handle_restart(regs, NULL, 0);
/* If there's no signal to deliver, we just restore the saved mask. */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
void do_notify_resume(struct pt_regs *regs)
@@ -1193,9 +1179,6 @@ void do_notify_resume(struct pt_regs *regs)
if (test_thread_flag(TIF_SIGPENDING))
do_signal(regs);
- if (test_and_clear_thread_flag(TIF_NOTIFY_RESUME)) {
+ if (test_and_clear_thread_flag(TIF_NOTIFY_RESUME))
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
- }
}
diff --git a/arch/microblaze/include/asm/thread_info.h b/arch/microblaze/include/asm/thread_info.h
index 1a8ab6a5c03..6c610234ffa 100644
--- a/arch/microblaze/include/asm/thread_info.h
+++ b/arch/microblaze/include/asm/thread_info.h
@@ -166,7 +166,23 @@ static inline void set_restore_sigmask(void)
{
struct thread_info *ti = current_thread_info();
ti->status |= TS_RESTORE_SIGMASK;
- set_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags);
+ WARN_ON(!test_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags));
+}
+static inline void clear_restore_sigmask(void)
+{
+ current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
+}
+static inline bool test_restore_sigmask(void)
+{
+ return current_thread_info()->status & TS_RESTORE_SIGMASK;
+}
+static inline bool test_and_clear_restore_sigmask(void)
+{
+ struct thread_info *ti = current_thread_info();
+ if (!(ti->status & TS_RESTORE_SIGMASK))
+ return false;
+ ti->status &= ~TS_RESTORE_SIGMASK;
+ return true;
}
#endif
diff --git a/arch/microblaze/kernel/signal.c b/arch/microblaze/kernel/signal.c
index 7f4c7bef164..76b9722557d 100644
--- a/arch/microblaze/kernel/signal.c
+++ b/arch/microblaze/kernel/signal.c
@@ -41,8 +41,6 @@
#include <asm/cacheflush.h>
#include <asm/syscalls.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
asmlinkage long
sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss,
struct pt_regs *regs)
@@ -106,7 +104,6 @@ asmlinkage long sys_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &rval))
@@ -310,10 +307,11 @@ do_restart:
* OK, we're invoking a handler
*/
-static int
+static void
handle_signal(unsigned long sig, struct k_sigaction *ka,
- siginfo_t *info, sigset_t *oldset, struct pt_regs *regs)
+ siginfo_t *info, struct pt_regs *regs)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* Set up the stack frame */
@@ -323,11 +321,9 @@ handle_signal(unsigned long sig, struct k_sigaction *ka,
ret = setup_rt_frame(sig, ka, NULL, oldset, regs);
if (ret)
- return ret;
-
- block_sigmask(ka, sig);
+ return;
- return 0;
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -344,33 +340,18 @@ static void do_signal(struct pt_regs *regs, int in_syscall)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
#ifdef DEBUG_SIG
printk(KERN_INFO "do signal: %p %d\n", regs, in_syscall);
printk(KERN_INFO "do signal2: %lx %lx %ld [%lx]\n", regs->pc, regs->r1,
regs->r12, current_thread_info()->flags);
#endif
- if (current_thread_info()->status & TS_RESTORE_SIGMASK)
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
if (in_syscall)
handle_restart(regs, &ka, 1);
- if (!handle_signal(signr, &ka, &info, oldset, regs)) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TS_RESTORE_SIGMASK flag.
- */
- current_thread_info()->status &=
- ~TS_RESTORE_SIGMASK;
- }
+ handle_signal(signr, &ka, &info, regs);
return;
}
@@ -381,10 +362,7 @@ static void do_signal(struct pt_regs *regs, int in_syscall)
* If there's no signal to deliver, we just put the saved sigmask
* back.
*/
- if (current_thread_info()->status & TS_RESTORE_SIGMASK) {
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
void do_notify_resume(struct pt_regs *regs, int in_syscall)
@@ -401,9 +379,6 @@ void do_notify_resume(struct pt_regs *regs, int in_syscall)
if (test_thread_flag(TIF_SIGPENDING))
do_signal(regs, in_syscall);
- if (test_and_clear_thread_flag(TIF_NOTIFY_RESUME)) {
+ if (test_and_clear_thread_flag(TIF_NOTIFY_RESUME))
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
- }
}
diff --git a/arch/mips/include/asm/posix_types.h b/arch/mips/include/asm/posix_types.h
index e0308dcca13..fa03ec3fbf8 100644
--- a/arch/mips/include/asm/posix_types.h
+++ b/arch/mips/include/asm/posix_types.h
@@ -17,11 +17,6 @@
* assume GCC is being used.
*/
-#if (_MIPS_SZLONG == 64)
-typedef unsigned int __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-#endif
-
typedef long __kernel_daddr_t;
#define __kernel_daddr_t __kernel_daddr_t
diff --git a/arch/mips/include/asm/stat.h b/arch/mips/include/asm/stat.h
index 6e00f751ab6..fe9a4c3ec5a 100644
--- a/arch/mips/include/asm/stat.h
+++ b/arch/mips/include/asm/stat.h
@@ -20,7 +20,7 @@ struct stat {
long st_pad1[3]; /* Reserved for network id */
ino_t st_ino;
mode_t st_mode;
- nlink_t st_nlink;
+ __u32 st_nlink;
uid_t st_uid;
gid_t st_gid;
unsigned st_rdev;
@@ -55,7 +55,7 @@ struct stat64 {
unsigned long long st_ino;
mode_t st_mode;
- nlink_t st_nlink;
+ __u32 st_nlink;
uid_t st_uid;
gid_t st_gid;
@@ -96,7 +96,7 @@ struct stat {
unsigned long st_ino;
mode_t st_mode;
- nlink_t st_nlink;
+ __u32 st_nlink;
uid_t st_uid;
gid_t st_gid;
diff --git a/arch/mips/kernel/signal-common.h b/arch/mips/kernel/signal-common.h
index 10263b40598..9c60d09e62a 100644
--- a/arch/mips/kernel/signal-common.h
+++ b/arch/mips/kernel/signal-common.h
@@ -19,8 +19,6 @@
# define DEBUGP(fmt, args...)
#endif
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/*
* Determine which stack to use..
*/
diff --git a/arch/mips/kernel/signal.c b/arch/mips/kernel/signal.c
index 17f6ee30ad0..f2c09cfc60a 100644
--- a/arch/mips/kernel/signal.c
+++ b/arch/mips/kernel/signal.c
@@ -339,7 +339,6 @@ asmlinkage void sys_sigreturn(nabi_no_regargs struct pt_regs regs)
if (__copy_from_user(&blocked, &frame->sf_mask, sizeof(blocked)))
goto badframe;
- sigdelsetmask(&blocked, ~_BLOCKABLE);
set_current_blocked(&blocked);
sig = restore_sigcontext(&regs, &frame->sf_sc);
@@ -375,7 +374,6 @@ asmlinkage void sys_rt_sigreturn(nabi_no_regargs struct pt_regs regs)
if (__copy_from_user(&set, &frame->rs_uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
sig = restore_sigcontext(&regs, &frame->rs_uc.uc_mcontext);
@@ -514,9 +512,10 @@ struct mips_abi mips_abi = {
.restart = __NR_restart_syscall
};
-static int handle_signal(unsigned long sig, siginfo_t *info,
- struct k_sigaction *ka, sigset_t *oldset, struct pt_regs *regs)
+static void handle_signal(unsigned long sig, siginfo_t *info,
+ struct k_sigaction *ka, struct pt_regs *regs)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
struct mips_abi *abi = current->thread.abi;
void *vdso = current->mm->context.vdso;
@@ -550,17 +549,14 @@ static int handle_signal(unsigned long sig, siginfo_t *info,
ka, regs, sig, oldset);
if (ret)
- return ret;
-
- block_sigmask(ka, sig);
+ return;
- return ret;
+ signal_delivered(sig, info, ka, regs, 0);
}
static void do_signal(struct pt_regs *regs)
{
struct k_sigaction ka;
- sigset_t *oldset;
siginfo_t info;
int signr;
@@ -572,25 +568,10 @@ static void do_signal(struct pt_regs *regs)
if (!user_mode(regs))
return;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
- if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag.
- */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
-
+ handle_signal(signr, &info, &ka, regs);
return;
}
@@ -614,10 +595,7 @@ static void do_signal(struct pt_regs *regs)
* If there's no signal to deliver, we just put the saved sigmask
* back
*/
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
/*
@@ -630,14 +608,12 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, void *unused,
local_irq_enable();
/* deal with pending signal delivery */
- if (thread_info_flags & (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK))
+ if (thread_info_flags & _TIF_SIGPENDING)
do_signal(regs);
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/mips/kernel/signal32.c b/arch/mips/kernel/signal32.c
index b4fe2eacbd5..da1b56a39ac 100644
--- a/arch/mips/kernel/signal32.c
+++ b/arch/mips/kernel/signal32.c
@@ -465,7 +465,6 @@ asmlinkage void sys32_sigreturn(nabi_no_regargs struct pt_regs regs)
if (__copy_conv_sigset_from_user(&blocked, &frame->sf_mask))
goto badframe;
- sigdelsetmask(&blocked, ~_BLOCKABLE);
set_current_blocked(&blocked);
sig = restore_sigcontext32(&regs, &frame->sf_sc);
@@ -503,7 +502,6 @@ asmlinkage void sys32_rt_sigreturn(nabi_no_regargs struct pt_regs regs)
if (__copy_conv_sigset_from_user(&set, &frame->rs_uc.uc_sigmask))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
sig = restore_sigcontext32(&regs, &frame->rs_uc.uc_mcontext);
diff --git a/arch/mips/kernel/signal_n32.c b/arch/mips/kernel/signal_n32.c
index 63ffac9af7c..3574c145511 100644
--- a/arch/mips/kernel/signal_n32.c
+++ b/arch/mips/kernel/signal_n32.c
@@ -109,7 +109,6 @@ asmlinkage void sysn32_rt_sigreturn(nabi_no_regargs struct pt_regs regs)
if (__copy_conv_sigset_from_user(&set, &frame->rs_uc.uc_sigmask))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
sig = restore_sigcontext(&regs, &frame->rs_uc.uc_mcontext);
diff --git a/arch/mn10300/include/asm/posix_types.h b/arch/mn10300/include/asm/posix_types.h
index ab506181ec3..d31eeea480c 100644
--- a/arch/mn10300/include/asm/posix_types.h
+++ b/arch/mn10300/include/asm/posix_types.h
@@ -20,9 +20,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/mn10300/kernel/signal.c b/arch/mn10300/kernel/signal.c
index 890cf91767c..6ab0bee2a54 100644
--- a/arch/mn10300/kernel/signal.c
+++ b/arch/mn10300/kernel/signal.c
@@ -31,8 +31,6 @@
#define DEBUG_SIG 0
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/*
* atomically swap in the new signal mask, and wait for a signal.
*/
@@ -163,7 +161,6 @@ asmlinkage long sys_sigreturn(void)
sizeof(frame->extramask)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(current_frame(), &frame->sc, &d0))
@@ -191,7 +188,6 @@ asmlinkage long sys_rt_sigreturn(void)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(current_frame(), &frame->uc.uc_mcontext, &d0))
@@ -430,8 +426,9 @@ static inline void stepback(struct pt_regs *regs)
*/
static int handle_signal(int sig,
siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs *regs)
+ struct pt_regs *regs)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* Are we from a system call? */
@@ -461,11 +458,11 @@ static int handle_signal(int sig,
ret = setup_rt_frame(sig, ka, info, oldset, regs);
else
ret = setup_frame(sig, ka, oldset, regs);
+ if (ret)
+ return;
- if (ret == 0)
- block_sigmask(ka, sig);
-
- return ret;
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLESTEP));
}
/*
@@ -475,7 +472,6 @@ static void do_signal(struct pt_regs *regs)
{
struct k_sigaction ka;
siginfo_t info;
- sigset_t *oldset;
int signr;
/* we want the common case to go fast, which is why we may in certain
@@ -483,23 +479,9 @@ static void do_signal(struct pt_regs *regs)
if (!user_mode(regs))
return;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
- if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
- /* a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
-
- tracehook_signal_handler(signr, &info, &ka, regs,
- test_thread_flag(TIF_SINGLESTEP));
+ if (handle_signal(signr, &info, &ka, regs) == 0) {
}
return;
@@ -525,10 +507,7 @@ static void do_signal(struct pt_regs *regs)
/* if there's no signal to deliver, we just put the saved sigmask
* back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
/*
@@ -548,13 +527,11 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, u32 thread_info_flags)
}
/* deal with pending signal delivery */
- if (thread_info_flags & (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK))
+ if (thread_info_flags & _TIF_SIGPENDING)
do_signal(regs);
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(current_frame());
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/openrisc/kernel/signal.c b/arch/openrisc/kernel/signal.c
index e970743251a..30110297f4f 100644
--- a/arch/openrisc/kernel/signal.c
+++ b/arch/openrisc/kernel/signal.c
@@ -33,8 +33,6 @@
#define DEBUG_SIG 0
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
asmlinkage long
_sys_sigaltstack(const stack_t *uss, stack_t *uoss, struct pt_regs *regs)
{
@@ -101,7 +99,6 @@ asmlinkage long _sys_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext))
@@ -251,20 +248,19 @@ give_sigsegv:
return -EFAULT;
}
-static inline int
+static inline void
handle_signal(unsigned long sig,
siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs *regs)
+ struct pt_regs *regs)
{
int ret;
- ret = setup_rt_frame(sig, ka, info, oldset, regs);
+ ret = setup_rt_frame(sig, ka, info, sigmask_to_save(), regs);
if (ret)
- return ret;
-
- block_sigmask(ka, sig);
+ return;
- return 0;
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLESTEP));
}
/*
@@ -339,30 +335,10 @@ void do_signal(struct pt_regs *regs)
if (signr <= 0) {
/* no signal to deliver so we just put the saved sigmask
* back */
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
-
+ restore_saved_sigmask();
} else { /* signr > 0 */
- sigset_t *oldset;
-
- if (current_thread_info()->flags & _TIF_RESTORE_SIGMASK)
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
/* Whee! Actually deliver the signal. */
- if (!handle_signal(signr, &info, &ka, oldset, regs)) {
- /* a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag */
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
-
- tracehook_signal_handler(signr, &info, &ka, regs,
- test_thread_flag(TIF_SINGLESTEP));
+ handle_signal(signr, &info, &ka, regs);
}
return;
@@ -376,7 +352,5 @@ asmlinkage void do_notify_resume(struct pt_regs *regs)
if (current_thread_info()->flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/parisc/include/asm/posix_types.h b/arch/parisc/include/asm/posix_types.h
index 5212b0357da..b9344256f76 100644
--- a/arch/parisc/include/asm/posix_types.h
+++ b/arch/parisc/include/asm/posix_types.h
@@ -10,9 +10,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/parisc/include/asm/stat.h b/arch/parisc/include/asm/stat.h
index 9d5fbbc5c31..d76fbda5d62 100644
--- a/arch/parisc/include/asm/stat.h
+++ b/arch/parisc/include/asm/stat.h
@@ -7,7 +7,7 @@ struct stat {
unsigned int st_dev; /* dev_t is 32 bits on parisc */
ino_t st_ino; /* 32 bits */
mode_t st_mode; /* 16 bits */
- nlink_t st_nlink; /* 16 bits */
+ unsigned short st_nlink; /* 16 bits */
unsigned short st_reserved1; /* old st_uid */
unsigned short st_reserved2; /* old st_gid */
unsigned int st_rdev;
@@ -42,7 +42,7 @@ struct hpux_stat64 {
unsigned int st_dev; /* dev_t is 32 bits on parisc */
ino_t st_ino; /* 32 bits */
mode_t st_mode; /* 16 bits */
- nlink_t st_nlink; /* 16 bits */
+ unsigned short st_nlink; /* 16 bits */
unsigned short st_reserved1; /* old st_uid */
unsigned short st_reserved2; /* old st_gid */
unsigned int st_rdev;
diff --git a/arch/parisc/include/asm/thread_info.h b/arch/parisc/include/asm/thread_info.h
index 83ae7dd4d99..22b4726dee4 100644
--- a/arch/parisc/include/asm/thread_info.h
+++ b/arch/parisc/include/asm/thread_info.h
@@ -74,7 +74,7 @@ struct thread_info {
#define _TIF_BLOCKSTEP (1 << TIF_BLOCKSTEP)
#define _TIF_USER_WORK_MASK (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | \
- _TIF_NEED_RESCHED | _TIF_RESTORE_SIGMASK)
+ _TIF_NEED_RESCHED)
#endif /* __KERNEL__ */
diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S
index 07ef351edd5..18670a07884 100644
--- a/arch/parisc/kernel/entry.S
+++ b/arch/parisc/kernel/entry.S
@@ -924,7 +924,7 @@ intr_check_sig:
/* As above */
mfctl %cr30,%r1
LDREG TI_FLAGS(%r1),%r19
- ldi (_TIF_SIGPENDING|_TIF_RESTORE_SIGMASK|_TIF_NOTIFY_RESUME), %r20
+ ldi (_TIF_SIGPENDING|_TIF_NOTIFY_RESUME), %r20
and,COND(<>) %r19, %r20, %r0
b,n intr_restore /* skip past if we've nothing to do */
@@ -2032,7 +2032,7 @@ syscall_check_resched:
.import do_signal,code
syscall_check_sig:
LDREG TI_FLAGS-THREAD_SZ_ALGN-FRAME_SIZE(%r30),%r19
- ldi (_TIF_SIGPENDING|_TIF_RESTORE_SIGMASK), %r26
+ ldi (_TIF_SIGPENDING|_TIF_NOTIFY_RESUME), %r26
and,COND(<>) %r19, %r26, %r0
b,n syscall_restore /* skip past if we've nothing to do */
diff --git a/arch/parisc/kernel/signal.c b/arch/parisc/kernel/signal.c
index 4b9cb0d546d..594459bde14 100644
--- a/arch/parisc/kernel/signal.c
+++ b/arch/parisc/kernel/signal.c
@@ -48,9 +48,6 @@
#define DBG(LEVEL, ...)
#endif
-
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/* gcc will complain if a pointer is cast to an integer of different
* size. If you really need to do this (and we do for an ELF32 user
* application in an ELF64 kernel) then you have to do a cast to an
@@ -131,7 +128,6 @@ sys_rt_sigreturn(struct pt_regs *regs, int in_syscall)
goto give_sigsegv;
}
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
/* Good thing we saved the old gr[30], eh? */
@@ -443,8 +439,9 @@ give_sigsegv:
static long
handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs *regs, int in_syscall)
+ struct pt_regs *regs, int in_syscall)
{
+ sigset_t *oldset = sigmask_to_save();
DBG(1,"handle_signal: sig=%ld, ka=%p, info=%p, oldset=%p, regs=%p\n",
sig, ka, info, oldset, regs);
@@ -452,12 +449,13 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
if (!setup_rt_frame(sig, ka, info, oldset, regs, in_syscall))
return 0;
- block_sigmask(ka, sig);
-
- tracehook_signal_handler(sig, info, ka, regs,
+ signal_delivered(sig, info, ka, regs,
test_thread_flag(TIF_SINGLESTEP) ||
test_thread_flag(TIF_BLOCKSTEP));
+ DBG(1,KERN_DEBUG "do_signal: Exit (success), regs->gr[28] = %ld\n",
+ regs->gr[28]);
+
return 1;
}
@@ -568,28 +566,17 @@ do_signal(struct pt_regs *regs, long in_syscall)
siginfo_t info;
struct k_sigaction ka;
int signr;
- sigset_t *oldset;
- DBG(1,"\ndo_signal: oldset=0x%p, regs=0x%p, sr7 %#lx, in_syscall=%d\n",
- oldset, regs, regs->sr[7], in_syscall);
+ DBG(1,"\ndo_signal: regs=0x%p, sr7 %#lx, in_syscall=%d\n",
+ regs, regs->sr[7], in_syscall);
/* Everyone else checks to see if they are in kernel mode at
this point and exits if that's the case. I'm not sure why
we would be called in that case, but for some reason we
are. */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
- DBG(1,"do_signal: oldset %08lx / %08lx\n",
- oldset->sig[0], oldset->sig[1]);
-
-
/* May need to force signal if handle_signal failed to deliver */
while (1) {
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
DBG(3,"do_signal: signr = %d, regs->gr[28] = %ld\n", signr, regs->gr[28]);
@@ -603,14 +590,8 @@ do_signal(struct pt_regs *regs, long in_syscall)
/* Whee! Actually deliver the signal. If the
delivery failed, we need to continue to iterate in
this loop so we can deliver the SIGSEGV... */
- if (handle_signal(signr, &info, &ka, oldset,
- regs, in_syscall)) {
- DBG(1,KERN_DEBUG "do_signal: Exit (success), regs->gr[28] = %ld\n",
- regs->gr[28]);
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
+ if (handle_signal(signr, &info, &ka, regs, in_syscall))
return;
- }
}
/* end of while(1) looping forever if we can't force a signal */
@@ -621,24 +602,16 @@ do_signal(struct pt_regs *regs, long in_syscall)
DBG(1,"do_signal: Exit (not delivered), regs->gr[28] = %ld\n",
regs->gr[28]);
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
-
- return;
+ restore_saved_sigmask();
}
void do_notify_resume(struct pt_regs *regs, long in_syscall)
{
- if (test_thread_flag(TIF_SIGPENDING) ||
- test_thread_flag(TIF_RESTORE_SIGMASK))
+ if (test_thread_flag(TIF_SIGPENDING))
do_signal(regs, in_syscall);
if (test_thread_flag(TIF_NOTIFY_RESUME)) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/parisc/kernel/signal32.c b/arch/parisc/kernel/signal32.c
index e1413243076..fd49aeda9eb 100644
--- a/arch/parisc/kernel/signal32.c
+++ b/arch/parisc/kernel/signal32.c
@@ -47,8 +47,6 @@
#define DBG(LEVEL, ...)
#endif
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
inline void
sigset_32to64(sigset_t *s64, compat_sigset_t *s32)
{
diff --git a/arch/powerpc/include/asm/posix_types.h b/arch/powerpc/include/asm/posix_types.h
index f1393252bbd..2958c5b97b2 100644
--- a/arch/powerpc/include/asm/posix_types.h
+++ b/arch/powerpc/include/asm/posix_types.h
@@ -16,9 +16,6 @@ typedef int __kernel_ssize_t;
typedef long __kernel_ptrdiff_t;
#define __kernel_size_t __kernel_size_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
#endif
diff --git a/arch/powerpc/include/asm/stat.h b/arch/powerpc/include/asm/stat.h
index e4edc510b53..10cfb558e0f 100644
--- a/arch/powerpc/include/asm/stat.h
+++ b/arch/powerpc/include/asm/stat.h
@@ -30,11 +30,11 @@ struct stat {
unsigned long st_dev;
ino_t st_ino;
#ifdef __powerpc64__
- nlink_t st_nlink;
+ unsigned short st_nlink;
mode_t st_mode;
#else
mode_t st_mode;
- nlink_t st_nlink;
+ unsigned short st_nlink;
#endif
uid_t st_uid;
gid_t st_gid;
diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h
index a556ccc16b5..68831e9cf82 100644
--- a/arch/powerpc/include/asm/thread_info.h
+++ b/arch/powerpc/include/asm/thread_info.h
@@ -140,7 +140,23 @@ static inline void set_restore_sigmask(void)
{
struct thread_info *ti = current_thread_info();
ti->local_flags |= _TLF_RESTORE_SIGMASK;
- set_bit(TIF_SIGPENDING, &ti->flags);
+ WARN_ON(!test_bit(TIF_SIGPENDING, &ti->flags));
+}
+static inline void clear_restore_sigmask(void)
+{
+ current_thread_info()->local_flags &= ~_TLF_RESTORE_SIGMASK;
+}
+static inline bool test_restore_sigmask(void)
+{
+ return current_thread_info()->local_flags & _TLF_RESTORE_SIGMASK;
+}
+static inline bool test_and_clear_restore_sigmask(void)
+{
+ struct thread_info *ti = current_thread_info();
+ if (!(ti->local_flags & _TLF_RESTORE_SIGMASK))
+ return false;
+ ti->local_flags &= ~_TLF_RESTORE_SIGMASK;
+ return true;
}
static inline bool test_thread_local_flags(unsigned int flags)
diff --git a/arch/powerpc/kernel/signal.c b/arch/powerpc/kernel/signal.c
index 651c5963662..5c023c9cf16 100644
--- a/arch/powerpc/kernel/signal.c
+++ b/arch/powerpc/kernel/signal.c
@@ -51,16 +51,6 @@ void __user * get_sigframe(struct k_sigaction *ka, struct pt_regs *regs,
return (void __user *)newsp;
}
-
-/*
- * Restore the user process's signal mask
- */
-void restore_sigmask(sigset_t *set)
-{
- sigdelsetmask(set, ~_BLOCKABLE);
- set_current_blocked(set);
-}
-
static void check_syscall_restart(struct pt_regs *regs, struct k_sigaction *ka,
int has_handler)
{
@@ -114,30 +104,21 @@ static void check_syscall_restart(struct pt_regs *regs, struct k_sigaction *ka,
static int do_signal(struct pt_regs *regs)
{
- sigset_t *oldset;
+ sigset_t *oldset = sigmask_to_save();
siginfo_t info;
int signr;
struct k_sigaction ka;
int ret;
int is32 = is_32bit_task();
- if (current_thread_info()->local_flags & _TLF_RESTORE_SIGMASK)
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
/* Is there any syscall restart business here ? */
check_syscall_restart(regs, &ka, signr > 0);
if (signr <= 0) {
- struct thread_info *ti = current_thread_info();
/* No signal to deliver -- put the saved sigmask back */
- if (ti->local_flags & _TLF_RESTORE_SIGMASK) {
- ti->local_flags &= ~_TLF_RESTORE_SIGMASK;
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
regs->trap = 0;
return 0; /* no signals delivered */
}
@@ -167,18 +148,7 @@ static int do_signal(struct pt_regs *regs)
regs->trap = 0;
if (ret) {
- block_sigmask(&ka, signr);
-
- /*
- * A signal was successfully delivered; the saved sigmask is in
- * its frame, and we can clear the TLF_RESTORE_SIGMASK flag.
- */
- current_thread_info()->local_flags &= ~_TLF_RESTORE_SIGMASK;
-
- /*
- * Let tracing know that we've done the handler setup.
- */
- tracehook_signal_handler(signr, &info, &ka, regs,
+ signal_delivered(signr, &info, &ka, regs,
test_thread_flag(TIF_SINGLESTEP));
}
@@ -193,8 +163,6 @@ void do_notify_resume(struct pt_regs *regs, unsigned long thread_info_flags)
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/powerpc/kernel/signal.h b/arch/powerpc/kernel/signal.h
index 8dde973aaaf..e00acb41393 100644
--- a/arch/powerpc/kernel/signal.h
+++ b/arch/powerpc/kernel/signal.h
@@ -10,13 +10,10 @@
#ifndef _POWERPC_ARCH_SIGNAL_H
#define _POWERPC_ARCH_SIGNAL_H
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
extern void do_notify_resume(struct pt_regs *regs, unsigned long thread_info_flags);
extern void __user * get_sigframe(struct k_sigaction *ka, struct pt_regs *regs,
size_t frame_size, int is_32);
-extern void restore_sigmask(sigset_t *set);
extern int handle_signal32(unsigned long sig, struct k_sigaction *ka,
siginfo_t *info, sigset_t *oldset,
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 61f6aff25ed..8b4c049aee2 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -919,7 +919,7 @@ static int do_setcontext(struct ucontext __user *ucp, struct pt_regs *regs, int
if (!access_ok(VERIFY_READ, mcp, sizeof(*mcp)))
return -EFAULT;
#endif
- restore_sigmask(&set);
+ set_current_blocked(&set);
if (restore_user_regs(regs, mcp, sig))
return -EFAULT;
@@ -1273,7 +1273,7 @@ long sys_sigreturn(int r3, int r4, int r5, int r6, int r7, int r8,
set.sig[0] = sigctx.oldmask;
set.sig[1] = sigctx._unused[3];
#endif
- restore_sigmask(&set);
+ set_current_blocked(&set);
sr = (struct mcontext __user *)from_user_ptr(sigctx.regs);
addr = sr;
diff --git a/arch/powerpc/kernel/signal_64.c b/arch/powerpc/kernel/signal_64.c
index 2692efdb154..d183f8719a5 100644
--- a/arch/powerpc/kernel/signal_64.c
+++ b/arch/powerpc/kernel/signal_64.c
@@ -335,7 +335,7 @@ int sys_swapcontext(struct ucontext __user *old_ctx,
if (__copy_from_user(&set, &new_ctx->uc_sigmask, sizeof(set)))
do_exit(SIGSEGV);
- restore_sigmask(&set);
+ set_current_blocked(&set);
if (restore_sigcontext(regs, NULL, 0, &new_ctx->uc_mcontext))
do_exit(SIGSEGV);
@@ -364,7 +364,7 @@ int sys_rt_sigreturn(unsigned long r3, unsigned long r4, unsigned long r5,
if (__copy_from_user(&set, &uc->uc_sigmask, sizeof(set)))
goto badframe;
- restore_sigmask(&set);
+ set_current_blocked(&set);
if (restore_sigcontext(regs, NULL, 1, &uc->uc_mcontext))
goto badframe;
diff --git a/arch/powerpc/mm/mmu_context_nohash.c b/arch/powerpc/mm/mmu_context_nohash.c
index 5b63bd3da4a..e779642c25e 100644
--- a/arch/powerpc/mm/mmu_context_nohash.c
+++ b/arch/powerpc/mm/mmu_context_nohash.c
@@ -333,9 +333,7 @@ static int __cpuinit mmu_context_cpu_notify(struct notifier_block *self,
unsigned long action, void *hcpu)
{
unsigned int cpu = (unsigned int)(long)hcpu;
-#ifdef CONFIG_HOTPLUG_CPU
- struct task_struct *p;
-#endif
+
/* We don't touch CPU 0 map, it's allocated at aboot and kept
* around forever
*/
@@ -358,12 +356,7 @@ static int __cpuinit mmu_context_cpu_notify(struct notifier_block *self,
stale_map[cpu] = NULL;
/* We also clear the cpu_vm_mask bits of CPUs going away */
- read_lock(&tasklist_lock);
- for_each_process(p) {
- if (p->mm)
- cpumask_clear_cpu(cpu, mm_cpumask(p->mm));
- }
- read_unlock(&tasklist_lock);
+ clear_tasks_mm_cpumask(cpu);
break;
#endif /* CONFIG_HOTPLUG_CPU */
}
diff --git a/arch/s390/include/asm/posix_types.h b/arch/s390/include/asm/posix_types.h
index edf8527ff08..7be104c0f19 100644
--- a/arch/s390/include/asm/posix_types.h
+++ b/arch/s390/include/asm/posix_types.h
@@ -24,7 +24,6 @@ typedef unsigned short __kernel_old_dev_t;
typedef unsigned long __kernel_ino_t;
typedef unsigned short __kernel_mode_t;
-typedef unsigned short __kernel_nlink_t;
typedef unsigned short __kernel_ipc_pid_t;
typedef unsigned short __kernel_uid_t;
typedef unsigned short __kernel_gid_t;
@@ -35,7 +34,6 @@ typedef int __kernel_ptrdiff_t;
typedef unsigned int __kernel_ino_t;
typedef unsigned int __kernel_mode_t;
-typedef unsigned int __kernel_nlink_t;
typedef int __kernel_ipc_pid_t;
typedef unsigned int __kernel_uid_t;
typedef unsigned int __kernel_gid_t;
@@ -47,7 +45,6 @@ typedef unsigned long __kernel_sigset_t; /* at least 32 bits */
#define __kernel_ino_t __kernel_ino_t
#define __kernel_mode_t __kernel_mode_t
-#define __kernel_nlink_t __kernel_nlink_t
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
#define __kernel_uid_t __kernel_uid_t
#define __kernel_gid_t __kernel_gid_t
diff --git a/arch/s390/kernel/compat_signal.c b/arch/s390/kernel/compat_signal.c
index 377c096ca4a..3c0c19830c3 100644
--- a/arch/s390/kernel/compat_signal.c
+++ b/arch/s390/kernel/compat_signal.c
@@ -32,8 +32,6 @@
#include "compat_ptrace.h"
#include "entry.h"
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
typedef struct
{
__u8 callee_used_stack[__SIGNAL_FRAMESIZE32];
@@ -364,7 +362,6 @@ asmlinkage long sys32_sigreturn(void)
goto badframe;
if (__copy_from_user(&set.sig, &frame->sc.oldmask, _SIGMASK_COPY_SIZE32))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigregs32(regs, &frame->sregs))
goto badframe;
@@ -390,7 +387,6 @@ asmlinkage long sys32_rt_sigreturn(void)
goto badframe;
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigregs32(regs, &frame->uc.uc_mcontext))
goto badframe;
@@ -572,7 +568,7 @@ give_sigsegv:
* OK, we're invoking a handler
*/
-int handle_signal32(unsigned long sig, struct k_sigaction *ka,
+void handle_signal32(unsigned long sig, struct k_sigaction *ka,
siginfo_t *info, sigset_t *oldset, struct pt_regs *regs)
{
int ret;
@@ -583,8 +579,8 @@ int handle_signal32(unsigned long sig, struct k_sigaction *ka,
else
ret = setup_frame32(sig, ka, oldset, regs);
if (ret)
- return ret;
- block_sigmask(ka, sig);
- return 0;
+ return;
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLE_STEP));
}
diff --git a/arch/s390/kernel/entry.h b/arch/s390/kernel/entry.h
index 6cdddac93a2..f66a229ab0b 100644
--- a/arch/s390/kernel/entry.h
+++ b/arch/s390/kernel/entry.h
@@ -31,7 +31,7 @@ void do_per_trap(struct pt_regs *regs);
void syscall_trace(struct pt_regs *regs, int entryexit);
void kernel_stack_overflow(struct pt_regs * regs);
void do_signal(struct pt_regs *regs);
-int handle_signal32(unsigned long sig, struct k_sigaction *ka,
+void handle_signal32(unsigned long sig, struct k_sigaction *ka,
siginfo_t *info, sigset_t *oldset, struct pt_regs *regs);
void do_notify_resume(struct pt_regs *regs);
diff --git a/arch/s390/kernel/signal.c b/arch/s390/kernel/signal.c
index f626232e216..ac565b44aab 100644
--- a/arch/s390/kernel/signal.c
+++ b/arch/s390/kernel/signal.c
@@ -33,9 +33,6 @@
#include <asm/switch_to.h>
#include "entry.h"
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
-
typedef struct
{
__u8 callee_used_stack[__SIGNAL_FRAMESIZE];
@@ -169,7 +166,6 @@ SYSCALL_DEFINE0(sigreturn)
goto badframe;
if (__copy_from_user(&set.sig, &frame->sc.oldmask, _SIGMASK_COPY_SIZE))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigregs(regs, &frame->sregs))
goto badframe;
@@ -189,7 +185,6 @@ SYSCALL_DEFINE0(rt_sigreturn)
goto badframe;
if (__copy_from_user(&set.sig, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigregs(regs, &frame->uc.uc_mcontext))
goto badframe;
@@ -367,7 +362,7 @@ give_sigsegv:
return -EFAULT;
}
-static int handle_signal(unsigned long sig, struct k_sigaction *ka,
+static void handle_signal(unsigned long sig, struct k_sigaction *ka,
siginfo_t *info, sigset_t *oldset,
struct pt_regs *regs)
{
@@ -379,9 +374,9 @@ static int handle_signal(unsigned long sig, struct k_sigaction *ka,
else
ret = setup_frame(sig, ka, oldset, regs);
if (ret)
- return ret;
- block_sigmask(ka, sig);
- return 0;
+ return;
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLE_STEP));
}
/*
@@ -398,12 +393,7 @@ void do_signal(struct pt_regs *regs)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
-
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
+ sigset_t *oldset = sigmask_to_save();
/*
* Get signal to deliver. When running under ptrace, at this point
@@ -441,24 +431,10 @@ void do_signal(struct pt_regs *regs)
/* No longer in a system call */
clear_thread_flag(TIF_SYSCALL);
- if ((is_compat_task() ?
- handle_signal32(signr, &ka, &info, oldset, regs) :
- handle_signal(signr, &ka, &info, oldset, regs)) == 0) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag.
- */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
-
- /*
- * Let tracing know that we've done the handler setup.
- */
- tracehook_signal_handler(signr, &info, &ka, regs,
- test_thread_flag(TIF_SINGLE_STEP));
- }
+ if (is_compat_task())
+ handle_signal32(signr, &ka, &info, oldset, regs);
+ else
+ handle_signal(signr, &ka, &info, oldset, regs);
return;
}
@@ -484,16 +460,11 @@ void do_signal(struct pt_regs *regs)
/*
* If there's no signal to deliver, we just put the saved sigmask back.
*/
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
void do_notify_resume(struct pt_regs *regs)
{
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
diff --git a/arch/score/kernel/signal.c b/arch/score/kernel/signal.c
index d4a49011c48..e382c52ca0d 100644
--- a/arch/score/kernel/signal.c
+++ b/arch/score/kernel/signal.c
@@ -34,8 +34,6 @@
#include <asm/syscalls.h>
#include <asm/ucontext.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
struct rt_sigframe {
u32 rs_ass[4]; /* argument save space */
u32 rs_code[2]; /* signal trampoline */
@@ -162,7 +160,6 @@ score_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->rs_uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
sig = restore_sigcontext(regs, &frame->rs_uc.uc_mcontext);
@@ -241,11 +238,9 @@ give_sigsegv:
return -EFAULT;
}
-static int handle_signal(unsigned long sig, siginfo_t *info,
- struct k_sigaction *ka, sigset_t *oldset, struct pt_regs *regs)
+static void handle_signal(unsigned long sig, siginfo_t *info,
+ struct k_sigaction *ka, struct pt_regs *regs)
{
- int ret;
-
if (regs->is_syscall) {
switch (regs->regs[4]) {
case ERESTART_RESTARTBLOCK:
@@ -269,18 +264,15 @@ static int handle_signal(unsigned long sig, siginfo_t *info,
/*
* Set up the stack frame
*/
- ret = setup_rt_frame(ka, regs, sig, oldset, info);
-
- if (ret == 0)
- block_sigmask(ka, sig);
+ if (setup_rt_frame(ka, regs, sig, sigmask_to_save(), info) < 0)
+ return;
- return ret;
+ signal_delivered(sig, info, ka, regs, 0);
}
static void do_signal(struct pt_regs *regs)
{
struct k_sigaction ka;
- sigset_t *oldset;
siginfo_t info;
int signr;
@@ -292,25 +284,10 @@ static void do_signal(struct pt_regs *regs)
if (!user_mode(regs))
return;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Actually deliver the signal. */
- if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag.
- */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
-
+ handle_signal(signr, &info, &ka, regs);
return;
}
@@ -337,10 +314,7 @@ static void do_signal(struct pt_regs *regs)
* If there's no signal to deliver, we just put the saved sigmask
* back
*/
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
/*
@@ -356,7 +330,5 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, void *unused,
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/sh/include/asm/posix_types_32.h b/arch/sh/include/asm/posix_types_32.h
index abda58467ec..ba0bdc423b0 100644
--- a/arch/sh/include/asm/posix_types_32.h
+++ b/arch/sh/include/asm/posix_types_32.h
@@ -3,8 +3,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
typedef unsigned short __kernel_uid_t;
diff --git a/arch/sh/include/asm/posix_types_64.h b/arch/sh/include/asm/posix_types_64.h
index fcda07b4a61..244f7e950e1 100644
--- a/arch/sh/include/asm/posix_types_64.h
+++ b/arch/sh/include/asm/posix_types_64.h
@@ -3,8 +3,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
typedef unsigned short __kernel_uid_t;
diff --git a/arch/sh/include/asm/thread_info.h b/arch/sh/include/asm/thread_info.h
index 0c04ffc4f12..bc13b57cdc8 100644
--- a/arch/sh/include/asm/thread_info.h
+++ b/arch/sh/include/asm/thread_info.h
@@ -169,7 +169,7 @@ static inline void set_restore_sigmask(void)
{
struct thread_info *ti = current_thread_info();
ti->status |= TS_RESTORE_SIGMASK;
- set_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags);
+ WARN_ON(!test_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags));
}
#define TI_FLAG_FAULT_CODE_SHIFT 24
@@ -189,6 +189,23 @@ static inline unsigned int get_thread_fault_code(void)
struct thread_info *ti = current_thread_info();
return ti->flags >> TI_FLAG_FAULT_CODE_SHIFT;
}
+
+static inline void clear_restore_sigmask(void)
+{
+ current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
+}
+static inline bool test_restore_sigmask(void)
+{
+ return current_thread_info()->status & TS_RESTORE_SIGMASK;
+}
+static inline bool test_and_clear_restore_sigmask(void)
+{
+ struct thread_info *ti = current_thread_info();
+ if (!(ti->status & TS_RESTORE_SIGMASK))
+ return false;
+ ti->status &= ~TS_RESTORE_SIGMASK;
+ return true;
+}
#endif /* !__ASSEMBLY__ */
#endif /* __KERNEL__ */
diff --git a/arch/sh/kernel/signal_32.c b/arch/sh/kernel/signal_32.c
index cb4172c8af7..d6b7b6154f8 100644
--- a/arch/sh/kernel/signal_32.c
+++ b/arch/sh/kernel/signal_32.c
@@ -32,8 +32,6 @@
#include <asm/syscalls.h>
#include <asm/fpu.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
struct fdpic_func_descriptor {
unsigned long text;
unsigned long GOT;
@@ -226,7 +224,6 @@ asmlinkage int sys_sigreturn(unsigned long r4, unsigned long r5,
sizeof(frame->extramask))))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->sc, &r0))
@@ -256,7 +253,6 @@ asmlinkage int sys_rt_sigreturn(unsigned long r4, unsigned long r5,
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &r0))
@@ -522,10 +518,11 @@ handle_syscall_restart(unsigned long save_r0, struct pt_regs *regs,
/*
* OK, we're invoking a handler
*/
-static int
+static void
handle_signal(unsigned long sig, struct k_sigaction *ka, siginfo_t *info,
- sigset_t *oldset, struct pt_regs *regs, unsigned int save_r0)
+ struct pt_regs *regs, unsigned int save_r0)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* Set up the stack frame */
@@ -534,10 +531,10 @@ handle_signal(unsigned long sig, struct k_sigaction *ka, siginfo_t *info,
else
ret = setup_frame(sig, ka, oldset, regs);
- if (ret == 0)
- block_sigmask(ka, sig);
-
- return ret;
+ if (ret)
+ return;
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLESTEP));
}
/*
@@ -554,7 +551,6 @@ static void do_signal(struct pt_regs *regs, unsigned int save_r0)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
/*
* We want the common case to go fast, which
@@ -565,30 +561,12 @@ static void do_signal(struct pt_regs *regs, unsigned int save_r0)
if (!user_mode(regs))
return;
- if (current_thread_info()->status & TS_RESTORE_SIGMASK)
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
handle_syscall_restart(save_r0, regs, &ka.sa);
/* Whee! Actually deliver the signal. */
- if (handle_signal(signr, &ka, &info, oldset,
- regs, save_r0) == 0) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TS_RESTORE_SIGMASK flag
- */
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
-
- tracehook_signal_handler(signr, &info, &ka, regs,
- test_thread_flag(TIF_SINGLESTEP));
- }
-
+ handle_signal(signr, &ka, &info, regs, save_r0);
return;
}
@@ -610,10 +588,7 @@ static void do_signal(struct pt_regs *regs, unsigned int save_r0)
* If there's no signal to deliver, we just put the saved sigmask
* back.
*/
- if (current_thread_info()->status & TS_RESTORE_SIGMASK) {
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
}
asmlinkage void do_notify_resume(struct pt_regs *regs, unsigned int save_r0,
@@ -626,7 +601,5 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, unsigned int save_r0,
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/sh/kernel/signal_64.c b/arch/sh/kernel/signal_64.c
index b589a354c06..6b5b3dfe886 100644
--- a/arch/sh/kernel/signal_64.c
+++ b/arch/sh/kernel/signal_64.c
@@ -41,11 +41,9 @@
#define DEBUG_SIG 0
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
-static int
+static void
handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs * regs);
+ struct pt_regs * regs);
static inline void
handle_syscall_restart(struct pt_regs *regs, struct sigaction *sa)
@@ -88,7 +86,6 @@ static void do_signal(struct pt_regs *regs)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
/*
* We want the common case to go fast, which
@@ -99,28 +96,13 @@ static void do_signal(struct pt_regs *regs)
if (!user_mode(regs))
return;
- if (current_thread_info()->status & TS_RESTORE_SIGMASK)
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, 0);
if (signr > 0) {
handle_syscall_restart(regs, &ka.sa);
/* Whee! Actually deliver the signal. */
- if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
- /*
- * If a signal was successfully delivered, the
- * saved sigmask is in its frame, and we can
- * clear the TS_RESTORE_SIGMASK flag.
- */
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
-
- tracehook_signal_handler(signr, &info, &ka, regs,
- test_thread_flag(TIF_SINGLESTEP));
- return;
- }
+ handle_signal(signr, &info, &ka, regs);
+ return;
}
/* Did we come from a system call? */
@@ -143,12 +125,7 @@ static void do_signal(struct pt_regs *regs)
}
/* No signal to deliver -- put the saved sigmask back */
- if (current_thread_info()->status & TS_RESTORE_SIGMASK) {
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
-
- return;
+ restore_saved_sigmask();
}
/*
@@ -351,7 +328,6 @@ asmlinkage int sys_sigreturn(unsigned long r2, unsigned long r3,
sizeof(frame->extramask))))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->sc, &ret))
@@ -384,7 +360,6 @@ asmlinkage int sys_rt_sigreturn(unsigned long r2, unsigned long r3,
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ret))
@@ -659,10 +634,11 @@ give_sigsegv:
/*
* OK, we're invoking a handler
*/
-static int
+static void
handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
- sigset_t *oldset, struct pt_regs * regs)
+ struct pt_regs * regs)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* Set up the stack frame */
@@ -671,10 +647,11 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
else
ret = setup_frame(sig, ka, oldset, regs);
- if (ret == 0)
- block_sigmask(ka, sig);
+ if (ret)
+ return;
- return ret;
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLESTEP));
}
asmlinkage void do_notify_resume(struct pt_regs *regs, unsigned long thread_info_flags)
@@ -685,7 +662,5 @@ asmlinkage void do_notify_resume(struct pt_regs *regs, unsigned long thread_info
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/sh/kernel/smp.c b/arch/sh/kernel/smp.c
index b86e9ca7945..2062aa88af4 100644
--- a/arch/sh/kernel/smp.c
+++ b/arch/sh/kernel/smp.c
@@ -123,7 +123,6 @@ void native_play_dead(void)
int __cpu_disable(void)
{
unsigned int cpu = smp_processor_id();
- struct task_struct *p;
int ret;
ret = mp_ops->cpu_disable(cpu);
@@ -153,11 +152,7 @@ int __cpu_disable(void)
flush_cache_all();
local_flush_tlb_all();
- read_lock(&tasklist_lock);
- for_each_process(p)
- if (p->mm)
- cpumask_clear_cpu(cpu, mm_cpumask(p->mm));
- read_unlock(&tasklist_lock);
+ clear_tasks_mm_cpumask(cpu);
return 0;
}
diff --git a/arch/sparc/include/asm/posix_types.h b/arch/sparc/include/asm/posix_types.h
index 3070f25ae90..156220ed99e 100644
--- a/arch/sparc/include/asm/posix_types.h
+++ b/arch/sparc/include/asm/posix_types.h
@@ -9,8 +9,6 @@
#if defined(__sparc__) && defined(__arch64__)
/* sparc 64 bit */
-typedef unsigned int __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
typedef unsigned short __kernel_old_uid_t;
typedef unsigned short __kernel_old_gid_t;
@@ -38,9 +36,6 @@ typedef unsigned short __kernel_gid_t;
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef long __kernel_daddr_t;
#define __kernel_daddr_t __kernel_daddr_t
diff --git a/arch/sparc/include/asm/thread_info_32.h b/arch/sparc/include/asm/thread_info_32.h
index 5af66493245..e6cd224506a 100644
--- a/arch/sparc/include/asm/thread_info_32.h
+++ b/arch/sparc/include/asm/thread_info_32.h
@@ -131,8 +131,7 @@ register struct thread_info *current_thread_info_reg asm("g6");
#define _TIF_POLLING_NRFLAG (1<<TIF_POLLING_NRFLAG)
#define _TIF_DO_NOTIFY_RESUME_MASK (_TIF_NOTIFY_RESUME | \
- _TIF_SIGPENDING | \
- _TIF_RESTORE_SIGMASK)
+ _TIF_SIGPENDING)
#endif /* __KERNEL__ */
diff --git a/arch/sparc/include/asm/thread_info_64.h b/arch/sparc/include/asm/thread_info_64.h
index 7f0981b0945..cfa8c38fb9c 100644
--- a/arch/sparc/include/asm/thread_info_64.h
+++ b/arch/sparc/include/asm/thread_info_64.h
@@ -238,7 +238,23 @@ static inline void set_restore_sigmask(void)
{
struct thread_info *ti = current_thread_info();
ti->status |= TS_RESTORE_SIGMASK;
- set_bit(TIF_SIGPENDING, &ti->flags);
+ WARN_ON(!test_bit(TIF_SIGPENDING, &ti->flags));
+}
+static inline void clear_restore_sigmask(void)
+{
+ current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
+}
+static inline bool test_restore_sigmask(void)
+{
+ return current_thread_info()->status & TS_RESTORE_SIGMASK;
+}
+static inline bool test_and_clear_restore_sigmask(void)
+{
+ struct thread_info *ti = current_thread_info();
+ if (!(ti->status & TS_RESTORE_SIGMASK))
+ return false;
+ ti->status &= ~TS_RESTORE_SIGMASK;
+ return true;
}
#endif /* !__ASSEMBLY__ */
diff --git a/arch/sparc/kernel/signal32.c b/arch/sparc/kernel/signal32.c
index bb1513e45f1..a53e0a5fd3a 100644
--- a/arch/sparc/kernel/signal32.c
+++ b/arch/sparc/kernel/signal32.c
@@ -32,8 +32,6 @@
#include "sigutil.h"
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/* This magic should be in g_upper[0] for all upper parts
* to be valid.
*/
@@ -274,7 +272,6 @@ void do_sigreturn32(struct pt_regs *regs)
case 2: set.sig[1] = seta[2] + (((long)seta[3]) << 32);
case 1: set.sig[0] = seta[0] + (((long)seta[1]) << 32);
}
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
return;
@@ -376,7 +373,6 @@ asmlinkage void do_rt_sigreturn32(struct pt_regs *regs)
case 2: set.sig[1] = seta.sig[2] + (((long)seta.sig[3]) << 32);
case 1: set.sig[0] = seta.sig[0] + (((long)seta.sig[1]) << 32);
}
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
return;
segv:
@@ -775,7 +771,7 @@ sigsegv:
return -EFAULT;
}
-static inline int handle_signal32(unsigned long signr, struct k_sigaction *ka,
+static inline void handle_signal32(unsigned long signr, struct k_sigaction *ka,
siginfo_t *info,
sigset_t *oldset, struct pt_regs *regs)
{
@@ -787,12 +783,9 @@ static inline int handle_signal32(unsigned long signr, struct k_sigaction *ka,
err = setup_frame32(ka, regs, signr, oldset);
if (err)
- return err;
-
- block_sigmask(ka, signr);
- tracehook_signal_handler(signr, info, ka, regs, 0);
+ return;
- return 0;
+ signal_delivered(signr, info, ka, regs, 0);
}
static inline void syscall_restart32(unsigned long orig_i0, struct pt_regs *regs,
@@ -841,14 +834,7 @@ void do_signal32(sigset_t *oldset, struct pt_regs * regs)
if (signr > 0) {
if (restart_syscall)
syscall_restart32(orig_i0, regs, &ka.sa);
- if (handle_signal32(signr, &ka, &info, oldset, regs) == 0) {
- /* A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TS_RESTORE_SIGMASK flag.
- */
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- }
+ handle_signal32(signr, &ka, &info, oldset, regs);
return;
}
if (restart_syscall &&
@@ -872,10 +858,7 @@ void do_signal32(sigset_t *oldset, struct pt_regs * regs)
/* If there's no signal to deliver, we just put the saved sigmask
* back
*/
- if (current_thread_info()->status & TS_RESTORE_SIGMASK) {
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- set_current_blocked(&current->saved_sigmask);
- }
+ restore_saved_sigmask();
}
struct sigstack32 {
diff --git a/arch/sparc/kernel/signal_32.c b/arch/sparc/kernel/signal_32.c
index 2b7e849f7c6..68f9c8650af 100644
--- a/arch/sparc/kernel/signal_32.c
+++ b/arch/sparc/kernel/signal_32.c
@@ -29,8 +29,6 @@
#include "sigutil.h"
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
extern void fpsave(unsigned long *fpregs, unsigned long *fsr,
void *fpqueue, unsigned long *fpqdepth);
extern void fpload(unsigned long *fpregs, unsigned long *fsr);
@@ -130,7 +128,6 @@ asmlinkage void do_sigreturn(struct pt_regs *regs)
if (err)
goto segv_and_exit;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
return;
@@ -197,7 +194,6 @@ asmlinkage void do_rt_sigreturn(struct pt_regs *regs)
goto segv;
}
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
return;
segv:
@@ -449,10 +445,11 @@ sigsegv:
return -EFAULT;
}
-static inline int
+static inline void
handle_signal(unsigned long signr, struct k_sigaction *ka,
- siginfo_t *info, sigset_t *oldset, struct pt_regs *regs)
+ siginfo_t *info, struct pt_regs *regs)
{
+ sigset_t *oldset = sigmask_to_save();
int err;
if (ka->sa.sa_flags & SA_SIGINFO)
@@ -461,12 +458,9 @@ handle_signal(unsigned long signr, struct k_sigaction *ka,
err = setup_frame(ka, regs, signr, oldset);
if (err)
- return err;
-
- block_sigmask(ka, signr);
- tracehook_signal_handler(signr, info, ka, regs, 0);
+ return;
- return 0;
+ signal_delivered(signr, info, ka, regs, 0);
}
static inline void syscall_restart(unsigned long orig_i0, struct pt_regs *regs,
@@ -498,7 +492,6 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0)
{
struct k_sigaction ka;
int restart_syscall;
- sigset_t *oldset;
siginfo_t info;
int signr;
@@ -523,11 +516,6 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0)
if (pt_regs_is_syscall(regs) && (regs->psr & PSR_C))
regs->u_regs[UREG_G6] = orig_i0;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
/* If the debugger messes with the program counter, it clears
@@ -544,15 +532,7 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0)
if (signr > 0) {
if (restart_syscall)
syscall_restart(orig_i0, regs, &ka.sa);
- if (handle_signal(signr, &ka, &info, oldset, regs) == 0) {
- /* a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag.
- */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
+ handle_signal(signr, &ka, &info, regs);
return;
}
if (restart_syscall &&
@@ -576,22 +556,17 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0)
/* if there's no signal to deliver, we just put the saved sigmask
* back
*/
- if (test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- set_current_blocked(&current->saved_sigmask);
- }
+ restore_saved_sigmask();
}
void do_notify_resume(struct pt_regs *regs, unsigned long orig_i0,
unsigned long thread_info_flags)
{
- if (thread_info_flags & (_TIF_SIGPENDING | _TIF_RESTORE_SIGMASK))
+ if (thread_info_flags & _TIF_SIGPENDING)
do_signal(regs, orig_i0);
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/sparc/kernel/signal_64.c b/arch/sparc/kernel/signal_64.c
index eafaab486b2..867de2f8189 100644
--- a/arch/sparc/kernel/signal_64.c
+++ b/arch/sparc/kernel/signal_64.c
@@ -38,8 +38,6 @@
#include "systbls.h"
#include "sigutil.h"
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/* {set, get}context() needed for 64-bit SparcLinux userland. */
asmlinkage void sparc64_set_context(struct pt_regs *regs)
{
@@ -71,7 +69,6 @@ asmlinkage void sparc64_set_context(struct pt_regs *regs)
if (__copy_from_user(&set, &ucp->uc_sigmask, sizeof(sigset_t)))
goto do_sigsegv;
}
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
}
if (test_thread_flag(TIF_32BIT)) {
@@ -315,7 +312,6 @@ void do_rt_sigreturn(struct pt_regs *regs)
/* Prevent syscall restart. */
pt_regs_clear_syscall(regs);
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
return;
segv:
@@ -466,7 +462,7 @@ sigsegv:
return -EFAULT;
}
-static inline int handle_signal(unsigned long signr, struct k_sigaction *ka,
+static inline void handle_signal(unsigned long signr, struct k_sigaction *ka,
siginfo_t *info,
sigset_t *oldset, struct pt_regs *regs)
{
@@ -475,12 +471,9 @@ static inline int handle_signal(unsigned long signr, struct k_sigaction *ka,
err = setup_rt_frame(ka, regs, signr, oldset,
(ka->sa.sa_flags & SA_SIGINFO) ? info : NULL);
if (err)
- return err;
-
- block_sigmask(ka, signr);
- tracehook_signal_handler(signr, info, ka, regs, 0);
+ return;
- return 0;
+ signal_delivered(signr, info, ka, regs, 0);
}
static inline void syscall_restart(unsigned long orig_i0, struct pt_regs *regs,
@@ -512,7 +505,7 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0)
{
struct k_sigaction ka;
int restart_syscall;
- sigset_t *oldset;
+ sigset_t *oldset = sigmask_to_save();
siginfo_t info;
int signr;
@@ -538,11 +531,6 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0)
(regs->tstate & (TSTATE_XCARRY | TSTATE_ICARRY)))
regs->u_regs[UREG_G6] = orig_i0;
- if (current_thread_info()->status & TS_RESTORE_SIGMASK)
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
#ifdef CONFIG_COMPAT
if (test_thread_flag(TIF_32BIT)) {
extern void do_signal32(sigset_t *, struct pt_regs *);
@@ -563,14 +551,7 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0)
if (signr > 0) {
if (restart_syscall)
syscall_restart(orig_i0, regs, &ka.sa);
- if (handle_signal(signr, &ka, &info, oldset, regs) == 0) {
- /* A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TS_RESTORE_SIGMASK flag.
- */
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- }
+ handle_signal(signr, &ka, &info, oldset, regs);
return;
}
if (restart_syscall &&
@@ -594,10 +575,7 @@ static void do_signal(struct pt_regs *regs, unsigned long orig_i0)
/* If there's no signal to deliver, we just put the saved sigmask
* back
*/
- if (current_thread_info()->status & TS_RESTORE_SIGMASK) {
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- set_current_blocked(&current->saved_sigmask);
- }
+ restore_saved_sigmask();
}
void do_notify_resume(struct pt_regs *regs, unsigned long orig_i0, unsigned long thread_info_flags)
@@ -607,8 +585,6 @@ void do_notify_resume(struct pt_regs *regs, unsigned long orig_i0, unsigned long
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/sparc/kernel/sys_sparc_64.c b/arch/sparc/kernel/sys_sparc_64.c
index 3ee51f189a5..275f74fd6f6 100644
--- a/arch/sparc/kernel/sys_sparc_64.c
+++ b/arch/sparc/kernel/sys_sparc_64.c
@@ -580,16 +580,9 @@ SYSCALL_DEFINE5(64_mremap, unsigned long, addr, unsigned long, old_len,
unsigned long, new_len, unsigned long, flags,
unsigned long, new_addr)
{
- unsigned long ret = -EINVAL;
-
if (test_thread_flag(TIF_32BIT))
- goto out;
-
- down_write(&current->mm->mmap_sem);
- ret = do_mremap(addr, old_len, new_len, flags, new_addr);
- up_write(&current->mm->mmap_sem);
-out:
- return ret;
+ return -EINVAL;
+ return sys_mremap(addr, old_len, new_len, flags, new_addr);
}
/* we come to here via sys_nis_syscall so it can setup the regs argument */
diff --git a/arch/tile/include/asm/compat.h b/arch/tile/include/asm/compat.h
index 69adc08d36a..6e74450ff0a 100644
--- a/arch/tile/include/asm/compat.h
+++ b/arch/tile/include/asm/compat.h
@@ -44,7 +44,6 @@ typedef __kernel_uid32_t __compat_gid32_t;
typedef __kernel_mode_t compat_mode_t;
typedef __kernel_dev_t compat_dev_t;
typedef __kernel_loff_t compat_loff_t;
-typedef __kernel_nlink_t compat_nlink_t;
typedef __kernel_ipc_pid_t compat_ipc_pid_t;
typedef __kernel_daddr_t compat_daddr_t;
typedef __kernel_fsid_t compat_fsid_t;
diff --git a/arch/tile/include/asm/thread_info.h b/arch/tile/include/asm/thread_info.h
index 656c486e64f..7e1fef36bde 100644
--- a/arch/tile/include/asm/thread_info.h
+++ b/arch/tile/include/asm/thread_info.h
@@ -166,7 +166,23 @@ static inline void set_restore_sigmask(void)
{
struct thread_info *ti = current_thread_info();
ti->status |= TS_RESTORE_SIGMASK;
- set_bit(TIF_SIGPENDING, &ti->flags);
+ WARN_ON(!test_bit(TIF_SIGPENDING, &ti->flags));
+}
+static inline void clear_restore_sigmask(void)
+{
+ current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
+}
+static inline bool test_restore_sigmask(void)
+{
+ return current_thread_info()->status & TS_RESTORE_SIGMASK;
+}
+static inline bool test_and_clear_restore_sigmask(void)
+{
+ struct thread_info *ti = current_thread_info();
+ if (!(ti->status & TS_RESTORE_SIGMASK))
+ return false;
+ ti->status &= ~TS_RESTORE_SIGMASK;
+ return true;
}
#endif /* !__ASSEMBLY__ */
diff --git a/arch/tile/kernel/compat_signal.c b/arch/tile/kernel/compat_signal.c
index cdef6e5ec02..474571b8408 100644
--- a/arch/tile/kernel/compat_signal.c
+++ b/arch/tile/kernel/compat_signal.c
@@ -118,8 +118,6 @@ struct compat_rt_sigframe {
struct compat_ucontext uc;
};
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
long compat_sys_rt_sigaction(int sig, struct compat_sigaction __user *act,
struct compat_sigaction __user *oact,
size_t sigsetsize)
@@ -302,7 +300,6 @@ long compat_sys_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext))
diff --git a/arch/tile/kernel/process.c b/arch/tile/kernel/process.c
index ba1023d8a02..6be79915050 100644
--- a/arch/tile/kernel/process.c
+++ b/arch/tile/kernel/process.c
@@ -565,8 +565,6 @@ int do_work_pending(struct pt_regs *regs, u32 thread_info_flags)
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
return 1;
}
if (thread_info_flags & _TIF_SINGLESTEP) {
diff --git a/arch/tile/kernel/signal.c b/arch/tile/kernel/signal.c
index f79d4b88c74..e29b0553211 100644
--- a/arch/tile/kernel/signal.c
+++ b/arch/tile/kernel/signal.c
@@ -37,8 +37,6 @@
#define DEBUG_SIG 0
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
SYSCALL_DEFINE3(sigaltstack, const stack_t __user *, uss,
stack_t __user *, uoss, struct pt_regs *, regs)
{
@@ -96,7 +94,6 @@ SYSCALL_DEFINE1(rt_sigreturn, struct pt_regs *, regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext))
@@ -242,10 +239,11 @@ give_sigsegv:
* OK, we're invoking a handler
*/
-static int handle_signal(unsigned long sig, siginfo_t *info,
- struct k_sigaction *ka, sigset_t *oldset,
+static void handle_signal(unsigned long sig, siginfo_t *info,
+ struct k_sigaction *ka,
struct pt_regs *regs)
{
+ sigset_t *oldset = sigmask_to_save();
int ret;
/* Are we from a system call? */
@@ -278,15 +276,9 @@ static int handle_signal(unsigned long sig, siginfo_t *info,
else
#endif
ret = setup_rt_frame(sig, ka, info, oldset, regs);
- if (ret == 0) {
- /* This code is only called from system calls or from
- * the work_pending path in the return-to-user code, and
- * either way we can re-enable interrupts unconditionally.
- */
- block_sigmask(ka, sig);
- }
-
- return ret;
+ if (ret)
+ return;
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -299,7 +291,6 @@ void do_signal(struct pt_regs *regs)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t *oldset;
/*
* i386 will check if we're coming from kernel mode and bail out
@@ -308,24 +299,10 @@ void do_signal(struct pt_regs *regs)
* helpful, we can reinstate the check on "!user_mode(regs)".
*/
- if (current_thread_info()->status & TS_RESTORE_SIGMASK)
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
- if (handle_signal(signr, &info, &ka, oldset, regs) == 0) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TS_RESTORE_SIGMASK flag.
- */
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- }
-
+ handle_signal(signr, &info, &ka, regs);
goto done;
}
@@ -350,10 +327,7 @@ void do_signal(struct pt_regs *regs)
}
/* If there's no signal to deliver, just put the saved sigmask back. */
- if (current_thread_info()->status & TS_RESTORE_SIGMASK) {
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ restore_saved_sigmask();
done:
/* Avoid double syscall restart if there are nested signals. */
diff --git a/arch/um/include/shared/frame_kern.h b/arch/um/include/shared/frame_kern.h
index 76078490c25..e584e40ee83 100644
--- a/arch/um/include/shared/frame_kern.h
+++ b/arch/um/include/shared/frame_kern.h
@@ -6,9 +6,6 @@
#ifndef __FRAME_KERN_H_
#define __FRAME_KERN_H_
-#define _S(nr) (1<<((nr)-1))
-#define _BLOCKABLE (~(_S(SIGKILL) | _S(SIGSTOP)))
-
extern int setup_signal_stack_sc(unsigned long stack_top, int sig,
struct k_sigaction *ka,
struct pt_regs *regs,
diff --git a/arch/um/kernel/process.c b/arch/um/kernel/process.c
index 3a2235e0abc..ccb9a9d283f 100644
--- a/arch/um/kernel/process.c
+++ b/arch/um/kernel/process.c
@@ -117,11 +117,8 @@ void interrupt_end(void)
schedule();
if (test_thread_flag(TIF_SIGPENDING))
do_signal();
- if (test_and_clear_thread_flag(TIF_NOTIFY_RESUME)) {
+ if (test_and_clear_thread_flag(TIF_NOTIFY_RESUME))
tracehook_notify_resume(&current->thread.regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
- }
}
void exit_thread(void)
diff --git a/arch/um/kernel/reboot.c b/arch/um/kernel/reboot.c
index 4d93dff6b37..3d15243ce69 100644
--- a/arch/um/kernel/reboot.c
+++ b/arch/um/kernel/reboot.c
@@ -4,7 +4,9 @@
*/
#include "linux/sched.h"
+#include "linux/spinlock.h"
#include "linux/slab.h"
+#include "linux/oom.h"
#include "kern_util.h"
#include "os.h"
#include "skas.h"
@@ -22,13 +24,18 @@ static void kill_off_processes(void)
struct task_struct *p;
int pid;
+ read_lock(&tasklist_lock);
for_each_process(p) {
- if (p->mm == NULL)
- continue;
+ struct task_struct *t;
- pid = p->mm->context.id.u.pid;
+ t = find_lock_task_mm(p);
+ if (!t)
+ continue;
+ pid = t->mm->context.id.u.pid;
+ task_unlock(t);
os_kill_ptraced_process(pid, 1);
}
+ read_unlock(&tasklist_lock);
}
}
diff --git a/arch/um/kernel/signal.c b/arch/um/kernel/signal.c
index 292e706016c..7362d58efc2 100644
--- a/arch/um/kernel/signal.c
+++ b/arch/um/kernel/signal.c
@@ -15,17 +15,13 @@
EXPORT_SYMBOL(block_signals);
EXPORT_SYMBOL(unblock_signals);
-#define _S(nr) (1<<((nr)-1))
-
-#define _BLOCKABLE (~(_S(SIGKILL) | _S(SIGSTOP)))
-
/*
* OK, we're invoking a handler
*/
-static int handle_signal(struct pt_regs *regs, unsigned long signr,
- struct k_sigaction *ka, siginfo_t *info,
- sigset_t *oldset)
+static void handle_signal(struct pt_regs *regs, unsigned long signr,
+ struct k_sigaction *ka, siginfo_t *info)
{
+ sigset_t *oldset = sigmask_to_save();
unsigned long sp;
int err;
@@ -65,9 +61,7 @@ static int handle_signal(struct pt_regs *regs, unsigned long signr,
if (err)
force_sigsegv(signr, current);
else
- block_sigmask(ka, signr);
-
- return err;
+ signal_delivered(signr, info, ka, regs, 0);
}
static int kern_do_signal(struct pt_regs *regs)
@@ -77,24 +71,9 @@ static int kern_do_signal(struct pt_regs *regs)
int sig, handled_sig = 0;
while ((sig = get_signal_to_deliver(&info, &ka_copy, regs, NULL)) > 0) {
- sigset_t *oldset;
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
handled_sig = 1;
/* Whee! Actually deliver the signal. */
- if (!handle_signal(regs, sig, &ka_copy, &info, oldset)) {
- /*
- * a signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag
- */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- break;
- }
+ handle_signal(regs, sig, &ka_copy, &info);
}
/* Did we come from a system call? */
@@ -130,10 +109,8 @@ static int kern_do_signal(struct pt_regs *regs)
* if there's no signal to deliver, we just put the saved sigmask
* back
*/
- if (!handled_sig && test_thread_flag(TIF_RESTORE_SIGMASK)) {
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- sigprocmask(SIG_SETMASK, &current->saved_sigmask, NULL);
- }
+ if (!handled_sig)
+ restore_saved_sigmask();
return handled_sig;
}
diff --git a/arch/um/kernel/trap.c b/arch/um/kernel/trap.c
index dafc9471595..3be60765c0e 100644
--- a/arch/um/kernel/trap.c
+++ b/arch/um/kernel/trap.c
@@ -30,6 +30,8 @@ int handle_page_fault(unsigned long address, unsigned long ip,
pmd_t *pmd;
pte_t *pte;
int err = -EFAULT;
+ unsigned int flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE |
+ (is_write ? FAULT_FLAG_WRITE : 0);
*code_out = SEGV_MAPERR;
@@ -40,6 +42,7 @@ int handle_page_fault(unsigned long address, unsigned long ip,
if (in_atomic())
goto out_nosemaphore;
+retry:
down_read(&mm->mmap_sem);
vma = find_vma(mm, address);
if (!vma)
@@ -65,7 +68,11 @@ good_area:
do {
int fault;
- fault = handle_mm_fault(mm, vma, address, is_write ? FAULT_FLAG_WRITE : 0);
+ fault = handle_mm_fault(mm, vma, address, flags);
+
+ if ((fault & VM_FAULT_RETRY) && fatal_signal_pending(current))
+ goto out_nosemaphore;
+
if (unlikely(fault & VM_FAULT_ERROR)) {
if (fault & VM_FAULT_OOM) {
goto out_of_memory;
@@ -75,10 +82,17 @@ good_area:
}
BUG();
}
- if (fault & VM_FAULT_MAJOR)
- current->maj_flt++;
- else
- current->min_flt++;
+ if (flags & FAULT_FLAG_ALLOW_RETRY) {
+ if (fault & VM_FAULT_MAJOR)
+ current->maj_flt++;
+ else
+ current->min_flt++;
+ if (fault & VM_FAULT_RETRY) {
+ flags &= ~FAULT_FLAG_ALLOW_RETRY;
+
+ goto retry;
+ }
+ }
pgd = pgd_offset(mm, address);
pud = pud_offset(pgd, address);
diff --git a/arch/unicore32/kernel/signal.c b/arch/unicore32/kernel/signal.c
index 7754df6ef7d..8adedb37720 100644
--- a/arch/unicore32/kernel/signal.c
+++ b/arch/unicore32/kernel/signal.c
@@ -21,8 +21,6 @@
#include <asm/cacheflush.h>
#include <asm/ucontext.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
/*
* For UniCore syscalls, we encode the syscall number into the instruction.
*/
@@ -61,10 +59,8 @@ static int restore_sigframe(struct pt_regs *regs, struct sigframe __user *sf)
int err;
err = __copy_from_user(&set, &sf->uc.uc_sigmask, sizeof(set));
- if (err == 0) {
- sigdelsetmask(&set, ~_BLOCKABLE);
+ if (err == 0)
set_current_blocked(&set);
- }
err |= __get_user(regs->UCreg_00, &sf->uc.uc_mcontext.regs.UCreg_00);
err |= __get_user(regs->UCreg_01, &sf->uc.uc_mcontext.regs.UCreg_01);
@@ -312,13 +308,12 @@ static inline void setup_syscall_restart(struct pt_regs *regs)
/*
* OK, we're invoking a handler
*/
-static int handle_signal(unsigned long sig, struct k_sigaction *ka,
- siginfo_t *info, sigset_t *oldset,
- struct pt_regs *regs, int syscall)
+static void handle_signal(unsigned long sig, struct k_sigaction *ka,
+ siginfo_t *info, struct pt_regs *regs, int syscall)
{
struct thread_info *thread = current_thread_info();
struct task_struct *tsk = current;
- sigset_t blocked;
+ sigset_t *oldset = sigmask_to_save();
int usig = sig;
int ret;
@@ -364,15 +359,10 @@ static int handle_signal(unsigned long sig, struct k_sigaction *ka,
if (ret != 0) {
force_sigsegv(sig, tsk);
- return ret;
+ return;
}
- /*
- * Block the signal if we were successful.
- */
- block_sigmask(ka, sig);
-
- return 0;
+ signal_delivered(sig, info, ka, regs, 0);
}
/*
@@ -399,32 +389,12 @@ static void do_signal(struct pt_regs *regs, int syscall)
if (!user_mode(regs))
return;
- if (try_to_freeze())
- goto no_signal;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
- sigset_t *oldset;
-
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
- if (handle_signal(signr, &ka, &info, oldset, regs, syscall)
- == 0) {
- /*
- * A signal was successfully delivered; the saved
- * sigmask will have been stored in the signal frame,
- * and will be restored by sigreturn, so we can simply
- * clear the TIF_RESTORE_SIGMASK flag.
- */
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- }
+ handle_signal(signr, &ka, &info, regs, syscall);
return;
}
- no_signal:
/*
* No signal to deliver to the process - restart the syscall.
*/
@@ -451,8 +421,7 @@ static void do_signal(struct pt_regs *regs, int syscall)
/* If there's no signal to deliver, we just put the saved
* sigmask back.
*/
- if (test_and_clear_thread_flag(TIF_RESTORE_SIGMASK))
- set_current_blocked(&current->saved_sigmask);
+ restore_saved_sigmask();
}
asmlinkage void do_notify_resume(struct pt_regs *regs,
@@ -464,8 +433,6 @@ asmlinkage void do_notify_resume(struct pt_regs *regs,
if (thread_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
}
diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c
index 98bd70faccc..daeca56211e 100644
--- a/arch/x86/ia32/ia32_signal.c
+++ b/arch/x86/ia32/ia32_signal.c
@@ -273,7 +273,6 @@ asmlinkage long sys32_sigreturn(struct pt_regs *regs)
sizeof(frame->extramask))))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (ia32_restore_sigcontext(regs, &frame->sc, &ax))
@@ -299,7 +298,6 @@ asmlinkage long sys32_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (ia32_restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax))
diff --git a/arch/x86/include/asm/posix_types_32.h b/arch/x86/include/asm/posix_types_32.h
index 99f262e04b9..8e525059e7d 100644
--- a/arch/x86/include/asm/posix_types_32.h
+++ b/arch/x86/include/asm/posix_types_32.h
@@ -10,9 +10,6 @@
typedef unsigned short __kernel_mode_t;
#define __kernel_mode_t __kernel_mode_t
-typedef unsigned short __kernel_nlink_t;
-#define __kernel_nlink_t __kernel_nlink_t
-
typedef unsigned short __kernel_ipc_pid_t;
#define __kernel_ipc_pid_t __kernel_ipc_pid_t
diff --git a/arch/x86/include/asm/sighandling.h b/arch/x86/include/asm/sighandling.h
index ada93b3b8c6..beff97f7df3 100644
--- a/arch/x86/include/asm/sighandling.h
+++ b/arch/x86/include/asm/sighandling.h
@@ -7,8 +7,6 @@
#include <asm/processor-flags.h>
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
#define __FIX_EFLAGS (X86_EFLAGS_AC | X86_EFLAGS_OF | \
X86_EFLAGS_DF | X86_EFLAGS_TF | X86_EFLAGS_SF | \
X86_EFLAGS_ZF | X86_EFLAGS_AF | X86_EFLAGS_PF | \
diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h
index 5c25de07cba..89f794f007e 100644
--- a/arch/x86/include/asm/thread_info.h
+++ b/arch/x86/include/asm/thread_info.h
@@ -248,7 +248,23 @@ static inline void set_restore_sigmask(void)
{
struct thread_info *ti = current_thread_info();
ti->status |= TS_RESTORE_SIGMASK;
- set_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags);
+ WARN_ON(!test_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags));
+}
+static inline void clear_restore_sigmask(void)
+{
+ current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
+}
+static inline bool test_restore_sigmask(void)
+{
+ return current_thread_info()->status & TS_RESTORE_SIGMASK;
+}
+static inline bool test_and_clear_restore_sigmask(void)
+{
+ struct thread_info *ti = current_thread_info();
+ if (!(ti->status & TS_RESTORE_SIGMASK))
+ return false;
+ ti->status &= ~TS_RESTORE_SIGMASK;
+ return true;
}
static inline bool is_ia32_task(void)
diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
index 01ccf9b7147..623f2883747 100644
--- a/arch/x86/kernel/entry_32.S
+++ b/arch/x86/kernel/entry_32.S
@@ -316,7 +316,6 @@ ret_from_exception:
preempt_stop(CLBR_ANY)
ret_from_intr:
GET_THREAD_INFO(%ebp)
-resume_userspace_sig:
#ifdef CONFIG_VM86
movl PT_EFLAGS(%esp), %eax # mix EFLAGS and CS
movb PT_CS(%esp), %al
@@ -615,9 +614,13 @@ work_notifysig: # deal with pending signals and
# vm86-space
TRACE_IRQS_ON
ENABLE_INTERRUPTS(CLBR_NONE)
+ movb PT_CS(%esp), %bl
+ andb $SEGMENT_RPL_MASK, %bl
+ cmpb $USER_RPL, %bl
+ jb resume_kernel
xorl %edx, %edx
call do_notify_resume
- jmp resume_userspace_sig
+ jmp resume_userspace
ALIGN
work_notifysig_v86:
@@ -630,9 +633,13 @@ work_notifysig_v86:
#endif
TRACE_IRQS_ON
ENABLE_INTERRUPTS(CLBR_NONE)
+ movb PT_CS(%esp), %bl
+ andb $SEGMENT_RPL_MASK, %bl
+ cmpb $USER_RPL, %bl
+ jb resume_kernel
xorl %edx, %edx
call do_notify_resume
- jmp resume_userspace_sig
+ jmp resume_userspace
END(work_pending)
# perform syscall exit tracing
diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c
index 965dfda0fd5..21af737053a 100644
--- a/arch/x86/kernel/signal.c
+++ b/arch/x86/kernel/signal.c
@@ -555,7 +555,6 @@ unsigned long sys_sigreturn(struct pt_regs *regs)
sizeof(frame->extramask))))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->sc, &ax))
@@ -581,7 +580,6 @@ long sys_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax))
@@ -647,42 +645,28 @@ setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
struct pt_regs *regs)
{
int usig = signr_convert(sig);
- sigset_t *set = &current->blocked;
- int ret;
-
- if (current_thread_info()->status & TS_RESTORE_SIGMASK)
- set = &current->saved_sigmask;
+ sigset_t *set = sigmask_to_save();
/* Set up the stack frame */
if (is_ia32) {
if (ka->sa.sa_flags & SA_SIGINFO)
- ret = ia32_setup_rt_frame(usig, ka, info, set, regs);
+ return ia32_setup_rt_frame(usig, ka, info, set, regs);
else
- ret = ia32_setup_frame(usig, ka, set, regs);
+ return ia32_setup_frame(usig, ka, set, regs);
#ifdef CONFIG_X86_X32_ABI
} else if (is_x32) {
- ret = x32_setup_rt_frame(usig, ka, info,
+ return x32_setup_rt_frame(usig, ka, info,
(compat_sigset_t *)set, regs);
#endif
} else {
- ret = __setup_rt_frame(sig, ka, info, set, regs);
- }
-
- if (ret) {
- force_sigsegv(sig, current);
- return -EFAULT;
+ return __setup_rt_frame(sig, ka, info, set, regs);
}
-
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- return ret;
}
-static int
+static void
handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
struct pt_regs *regs)
{
- int ret;
-
/* Are we from a system call? */
if (syscall_get_nr(current, regs) >= 0) {
/* If so, check system call restarting.. */
@@ -713,10 +697,10 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
likely(test_and_clear_thread_flag(TIF_FORCED_TF)))
regs->flags &= ~X86_EFLAGS_TF;
- ret = setup_rt_frame(sig, ka, info, regs);
-
- if (ret)
- return ret;
+ if (setup_rt_frame(sig, ka, info, regs) < 0) {
+ force_sigsegv(sig, current);
+ return;
+ }
/*
* Clear the direction flag as per the ABI for function entry.
@@ -731,12 +715,8 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka,
*/
regs->flags &= ~X86_EFLAGS_TF;
- block_sigmask(ka, sig);
-
- tracehook_signal_handler(sig, info, ka, regs,
- test_thread_flag(TIF_SINGLESTEP));
-
- return 0;
+ signal_delivered(sig, info, ka, regs,
+ test_thread_flag(TIF_SINGLESTEP));
}
#ifdef CONFIG_X86_32
@@ -757,16 +737,6 @@ static void do_signal(struct pt_regs *regs)
siginfo_t info;
int signr;
- /*
- * We want the common case to go fast, which is why we may in certain
- * cases get here from kernel mode. Just return without doing anything
- * if so.
- * X86_32: vm86 regs switched out by assembly code before reaching
- * here, so testing against kernel CS suffices.
- */
- if (!user_mode(regs))
- return;
-
signr = get_signal_to_deliver(&info, &ka, regs, NULL);
if (signr > 0) {
/* Whee! Actually deliver the signal. */
@@ -796,10 +766,7 @@ static void do_signal(struct pt_regs *regs)
* If there's no signal to deliver, we just put the saved sigmask
* back.
*/
- if (current_thread_info()->status & TS_RESTORE_SIGMASK) {
- current_thread_info()->status &= ~TS_RESTORE_SIGMASK;
- set_current_blocked(&current->saved_sigmask);
- }
+ restore_saved_sigmask();
}
/*
@@ -827,8 +794,6 @@ do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags)
if (thread_info_flags & _TIF_NOTIFY_RESUME) {
clear_thread_flag(TIF_NOTIFY_RESUME);
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
}
if (thread_info_flags & _TIF_USER_RETURN_NOTIFY)
fire_user_return_notifiers();
@@ -936,7 +901,6 @@ asmlinkage long sys32_x32_rt_sigreturn(struct pt_regs *regs)
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax))
diff --git a/arch/x86/syscalls/syscall_32.tbl b/arch/x86/syscalls/syscall_32.tbl
index 29f9f0554f7..7a35a6e71d4 100644
--- a/arch/x86/syscalls/syscall_32.tbl
+++ b/arch/x86/syscalls/syscall_32.tbl
@@ -355,3 +355,4 @@
346 i386 setns sys_setns
347 i386 process_vm_readv sys_process_vm_readv compat_sys_process_vm_readv
348 i386 process_vm_writev sys_process_vm_writev compat_sys_process_vm_writev
+349 i386 kcmp sys_kcmp
diff --git a/arch/x86/syscalls/syscall_64.tbl b/arch/x86/syscalls/syscall_64.tbl
index dd29a9ea27c..51171aeff0d 100644
--- a/arch/x86/syscalls/syscall_64.tbl
+++ b/arch/x86/syscalls/syscall_64.tbl
@@ -318,6 +318,8 @@
309 common getcpu sys_getcpu
310 64 process_vm_readv sys_process_vm_readv
311 64 process_vm_writev sys_process_vm_writev
+312 64 kcmp sys_kcmp
+
#
# x32-specific system call numbers start at 512 to avoid cache impact
# for native 64-bit operation.
diff --git a/arch/x86/um/signal.c b/arch/x86/um/signal.c
index bb0fb03b9f8..a508cea1350 100644
--- a/arch/x86/um/signal.c
+++ b/arch/x86/um/signal.c
@@ -486,7 +486,6 @@ long sys_sigreturn(struct pt_regs *regs)
copy_from_user(&set.sig[1], extramask, sig_size))
goto segfault;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (copy_sc_from_user(&current->thread.regs, sc))
@@ -600,7 +599,6 @@ long sys_rt_sigreturn(struct pt_regs *regs)
if (copy_from_user(&set, &uc->uc_sigmask, sizeof(set)))
goto segfault;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (copy_sc_from_user(&current->thread.regs, &uc->uc_mcontext))
diff --git a/arch/xtensa/kernel/signal.c b/arch/xtensa/kernel/signal.c
index c5e4ec0598d..b9f8e5850d3 100644
--- a/arch/xtensa/kernel/signal.c
+++ b/arch/xtensa/kernel/signal.c
@@ -30,8 +30,6 @@
#define DEBUG_SIG 0
-#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP)))
-
extern struct task_struct *coproc_owners[];
struct rt_sigframe
@@ -261,7 +259,6 @@ asmlinkage long xtensa_rt_sigreturn(long a0, long a1, long a2, long a3,
if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
goto badframe;
- sigdelsetmask(&set, ~_BLOCKABLE);
set_current_blocked(&set);
if (restore_sigcontext(regs, frame))
@@ -452,15 +449,6 @@ static void do_signal(struct pt_regs *regs)
siginfo_t info;
int signr;
struct k_sigaction ka;
- sigset_t oldset;
-
- if (try_to_freeze())
- goto no_signal;
-
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = &current->saved_sigmask;
- else
- oldset = &current->blocked;
task_pt_regs(current)->icountlevel = 0;
@@ -501,19 +489,17 @@ static void do_signal(struct pt_regs *regs)
/* Whee! Actually deliver the signal. */
/* Set up the stack frame */
- ret = setup_frame(signr, &ka, &info, oldset, regs);
+ ret = setup_frame(signr, &ka, &info, sigmask_to_save(), regs);
if (ret)
return;
- clear_thread_flag(TIF_RESTORE_SIGMASK);
- block_sigmask(&ka, signr);
+ signal_delivered(signr, info, ka, regs, 0);
if (current->ptrace & PT_SINGLESTEP)
task_pt_regs(current)->icountlevel = 1;
return;
}
-no_signal:
/* Did we come from a system call? */
if ((signed) regs->syscall >= 0) {
/* Restart the system call - no handlers present */
@@ -532,8 +518,7 @@ no_signal:
}
/* If there's no signal to deliver, we just restore the saved mask. */
- if (test_and_clear_thread_flag(TIF_RESTORE_SIGMASK))
- set_current_blocked(&current->saved_sigmask);
+ restore_saved_sigmask();
if (current->ptrace & PT_SINGLESTEP)
task_pt_regs(current)->icountlevel = 1;
@@ -548,9 +533,6 @@ void do_notify_resume(struct pt_regs *regs)
if (test_thread_flag(TIF_SIGPENDING))
do_signal(regs);
- if (test_and_clear_thread_flag(TIF_NOTIFY_RESUME)) {
+ if (test_and_clear_thread_flag(TIF_NOTIFY_RESUME))
tracehook_notify_resume(regs);
- if (current->replacement_session_keyring)
- key_replace_session_keyring();
- }
}
diff --git a/drivers/base/soc.c b/drivers/base/soc.c
index ba29b2e73d4..72b5e7280d1 100644
--- a/drivers/base/soc.c
+++ b/drivers/base/soc.c
@@ -42,7 +42,7 @@ struct device *soc_device_to_device(struct soc_device *soc_dev)
return &soc_dev->dev;
}
-static mode_t soc_attribute_mode(struct kobject *kobj,
+static umode_t soc_attribute_mode(struct kobject *kobj,
struct attribute *attr,
int index)
{
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.c b/drivers/gpu/drm/cirrus/cirrus_drv.c
index d7038230b71..7053140c659 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.c
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.c
@@ -35,9 +35,28 @@ static DEFINE_PCI_DEVICE_TABLE(pciidlist) = {
{0,}
};
+
+static void cirrus_kick_out_firmware_fb(struct pci_dev *pdev)
+{
+ struct apertures_struct *ap;
+ bool primary = false;
+
+ ap = alloc_apertures(1);
+ ap->ranges[0].base = pci_resource_start(pdev, 0);
+ ap->ranges[0].size = pci_resource_len(pdev, 0);
+
+#ifdef CONFIG_X86
+ primary = pdev->resource[PCI_ROM_RESOURCE].flags & IORESOURCE_ROM_SHADOW;
+#endif
+ remove_conflicting_framebuffers(ap, "cirrusdrmfb", primary);
+ kfree(ap);
+}
+
static int __devinit
cirrus_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
{
+ cirrus_kick_out_firmware_fb(pdev);
+
return drm_get_pci_dev(pdev, ent, &driver);
}
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h
index 21bdfa8836f..64ea597cb6d 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.h
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.h
@@ -145,7 +145,7 @@ struct cirrus_device {
struct ttm_bo_device bdev;
atomic_t validate_sequence;
} ttm;
-
+ bool mm_inited;
};
diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c
index 2ebcd11a502..50e170f879d 100644
--- a/drivers/gpu/drm/cirrus/cirrus_ttm.c
+++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c
@@ -275,12 +275,17 @@ int cirrus_mm_init(struct cirrus_device *cirrus)
pci_resource_len(dev->pdev, 0),
DRM_MTRR_WC);
+ cirrus->mm_inited = true;
return 0;
}
void cirrus_mm_fini(struct cirrus_device *cirrus)
{
struct drm_device *dev = cirrus->dev;
+
+ if (!cirrus->mm_inited)
+ return;
+
ttm_bo_device_release(&cirrus->ttm.bdev);
cirrus_ttm_global_release(cirrus);
diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index c3b5139eba7..eb92fe257a3 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -30,7 +30,7 @@
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/i2c.h>
-#include <linux/export.h>
+#include <linux/module.h>
#include "drmP.h"
#include "drm_edid.h"
#include "drm_edid_modes.h"
@@ -149,6 +149,10 @@ int drm_edid_header_is_valid(const u8 *raw_edid)
}
EXPORT_SYMBOL(drm_edid_header_is_valid);
+static int edid_fixup __read_mostly = 6;
+module_param_named(edid_fixup, edid_fixup, int, 0400);
+MODULE_PARM_DESC(edid_fixup,
+ "Minimum number of valid EDID header bytes (0-8, default 6)");
/*
* Sanity check the EDID block (base or extension). Return 0 if the block
@@ -160,10 +164,13 @@ bool drm_edid_block_valid(u8 *raw_edid, int block)
u8 csum = 0;
struct edid *edid = (struct edid *)raw_edid;
+ if (edid_fixup > 8 || edid_fixup < 0)
+ edid_fixup = 6;
+
if (block == 0) {
int score = drm_edid_header_is_valid(raw_edid);
if (score == 8) ;
- else if (score >= 6) {
+ else if (score >= edid_fixup) {
DRM_DEBUG("Fixing EDID header, your hardware may be failing\n");
memcpy(raw_edid, edid_header, sizeof(edid_header));
} else {
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index f920fb5e42b..fa9439159eb 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -130,11 +130,10 @@ static int i810_map_buffer(struct drm_buf *buf, struct drm_file *file_priv)
return -EINVAL;
/* This is all entirely broken */
- down_write(&current->mm->mmap_sem);
old_fops = file_priv->filp->f_op;
file_priv->filp->f_op = &i810_buffer_fops;
dev_priv->mmap_buffer = buf;
- buf_priv->virtual = (void *)do_mmap(file_priv->filp, 0, buf->total,
+ buf_priv->virtual = (void *)vm_mmap(file_priv->filp, 0, buf->total,
PROT_READ | PROT_WRITE,
MAP_SHARED, buf->bus_address);
dev_priv->mmap_buffer = NULL;
@@ -145,7 +144,6 @@ static int i810_map_buffer(struct drm_buf *buf, struct drm_file *file_priv)
retcode = PTR_ERR(buf_priv->virtual);
buf_priv->virtual = NULL;
}
- up_write(&current->mm->mmap_sem);
return retcode;
}
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 377c21f531e..c9cfc67c2cf 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -942,6 +942,9 @@ struct drm_i915_gem_object {
/* prime dma-buf support */
struct sg_table *sg_table;
+ void *dma_buf_vmapping;
+ int vmapping_count;
+
/**
* Used for performing relocations during execbuffer insertion.
*/
diff --git a/drivers/gpu/drm/i915/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
index 8e269178d6a..aa308e1337d 100644
--- a/drivers/gpu/drm/i915/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
@@ -74,6 +74,59 @@ static void i915_gem_dmabuf_release(struct dma_buf *dma_buf)
}
}
+static void *i915_gem_dmabuf_vmap(struct dma_buf *dma_buf)
+{
+ struct drm_i915_gem_object *obj = dma_buf->priv;
+ struct drm_device *dev = obj->base.dev;
+ int ret;
+
+ ret = i915_mutex_lock_interruptible(dev);
+ if (ret)
+ return ERR_PTR(ret);
+
+ if (obj->dma_buf_vmapping) {
+ obj->vmapping_count++;
+ goto out_unlock;
+ }
+
+ if (!obj->pages) {
+ ret = i915_gem_object_get_pages_gtt(obj, __GFP_NORETRY | __GFP_NOWARN);
+ if (ret) {
+ mutex_unlock(&dev->struct_mutex);
+ return ERR_PTR(ret);
+ }
+ }
+
+ obj->dma_buf_vmapping = vmap(obj->pages, obj->base.size / PAGE_SIZE, 0, PAGE_KERNEL);
+ if (!obj->dma_buf_vmapping) {
+ DRM_ERROR("failed to vmap object\n");
+ goto out_unlock;
+ }
+
+ obj->vmapping_count = 1;
+out_unlock:
+ mutex_unlock(&dev->struct_mutex);
+ return obj->dma_buf_vmapping;
+}
+
+static void i915_gem_dmabuf_vunmap(struct dma_buf *dma_buf, void *vaddr)
+{
+ struct drm_i915_gem_object *obj = dma_buf->priv;
+ struct drm_device *dev = obj->base.dev;
+ int ret;
+
+ ret = i915_mutex_lock_interruptible(dev);
+ if (ret)
+ return;
+
+ --obj->vmapping_count;
+ if (obj->vmapping_count == 0) {
+ vunmap(obj->dma_buf_vmapping);
+ obj->dma_buf_vmapping = NULL;
+ }
+ mutex_unlock(&dev->struct_mutex);
+}
+
static void *i915_gem_dmabuf_kmap_atomic(struct dma_buf *dma_buf, unsigned long page_num)
{
return NULL;
@@ -93,6 +146,11 @@ static void i915_gem_dmabuf_kunmap(struct dma_buf *dma_buf, unsigned long page_n
}
+static int i915_gem_dmabuf_mmap(struct dma_buf *dma_buf, struct vm_area_struct *vma)
+{
+ return -EINVAL;
+}
+
static const struct dma_buf_ops i915_dmabuf_ops = {
.map_dma_buf = i915_gem_map_dma_buf,
.unmap_dma_buf = i915_gem_unmap_dma_buf,
@@ -101,6 +159,9 @@ static const struct dma_buf_ops i915_dmabuf_ops = {
.kmap_atomic = i915_gem_dmabuf_kmap_atomic,
.kunmap = i915_gem_dmabuf_kunmap,
.kunmap_atomic = i915_gem_dmabuf_kunmap_atomic,
+ .mmap = i915_gem_dmabuf_mmap,
+ .vmap = i915_gem_dmabuf_vmap,
+ .vunmap = i915_gem_dmabuf_vunmap,
};
struct dma_buf *i915_gem_prime_export(struct drm_device *dev,
diff --git a/drivers/gpu/drm/mgag200/mgag200_drv.c b/drivers/gpu/drm/mgag200/mgag200_drv.c
index 3c8e04f5471..93e832d6c32 100644
--- a/drivers/gpu/drm/mgag200/mgag200_drv.c
+++ b/drivers/gpu/drm/mgag200/mgag200_drv.c
@@ -41,9 +41,28 @@ static DEFINE_PCI_DEVICE_TABLE(pciidlist) = {
MODULE_DEVICE_TABLE(pci, pciidlist);
+static void mgag200_kick_out_firmware_fb(struct pci_dev *pdev)
+{
+ struct apertures_struct *ap;
+ bool primary = false;
+
+ ap = alloc_apertures(1);
+ ap->ranges[0].base = pci_resource_start(pdev, 0);
+ ap->ranges[0].size = pci_resource_len(pdev, 0);
+
+#ifdef CONFIG_X86
+ primary = pdev->resource[PCI_ROM_RESOURCE].flags & IORESOURCE_ROM_SHADOW;
+#endif
+ remove_conflicting_framebuffers(ap, "mgag200drmfb", primary);
+ kfree(ap);
+}
+
+
static int __devinit
mga_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
{
+ mgag200_kick_out_firmware_fb(pdev);
+
return drm_get_pci_dev(pdev, ent, &driver);
}
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h
index 634d222c93d..8613cb23808 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
@@ -123,6 +123,9 @@ struct nouveau_bo {
struct drm_gem_object *gem;
int pin_refcnt;
+
+ struct ttm_bo_kmap_obj dma_buf_vmap;
+ int vmapping_count;
};
#define nouveau_bo_tile_layout(nvbo) \
diff --git a/drivers/gpu/drm/nouveau/nouveau_prime.c b/drivers/gpu/drm/nouveau/nouveau_prime.c
index c58aab7370c..a89240e5fb2 100644
--- a/drivers/gpu/drm/nouveau/nouveau_prime.c
+++ b/drivers/gpu/drm/nouveau/nouveau_prime.c
@@ -61,6 +61,48 @@ static void nouveau_gem_kunmap(struct dma_buf *dma_buf, unsigned long page_num,
}
+static int nouveau_gem_prime_mmap(struct dma_buf *dma_buf, struct vm_area_struct *vma)
+{
+ return -EINVAL;
+}
+
+static void *nouveau_gem_prime_vmap(struct dma_buf *dma_buf)
+{
+ struct nouveau_bo *nvbo = dma_buf->priv;
+ struct drm_device *dev = nvbo->gem->dev;
+ int ret;
+
+ mutex_lock(&dev->struct_mutex);
+ if (nvbo->vmapping_count) {
+ nvbo->vmapping_count++;
+ goto out_unlock;
+ }
+
+ ret = ttm_bo_kmap(&nvbo->bo, 0, nvbo->bo.num_pages,
+ &nvbo->dma_buf_vmap);
+ if (ret) {
+ mutex_unlock(&dev->struct_mutex);
+ return ERR_PTR(ret);
+ }
+ nvbo->vmapping_count = 1;
+out_unlock:
+ mutex_unlock(&dev->struct_mutex);
+ return nvbo->dma_buf_vmap.virtual;
+}
+
+static void nouveau_gem_prime_vunmap(struct dma_buf *dma_buf, void *vaddr)
+{
+ struct nouveau_bo *nvbo = dma_buf->priv;
+ struct drm_device *dev = nvbo->gem->dev;
+
+ mutex_lock(&dev->struct_mutex);
+ nvbo->vmapping_count--;
+ if (nvbo->vmapping_count == 0) {
+ ttm_bo_kunmap(&nvbo->dma_buf_vmap);
+ }
+ mutex_unlock(&dev->struct_mutex);
+}
+
static const struct dma_buf_ops nouveau_dmabuf_ops = {
.map_dma_buf = nouveau_gem_map_dma_buf,
.unmap_dma_buf = nouveau_gem_unmap_dma_buf,
@@ -69,6 +111,9 @@ static const struct dma_buf_ops nouveau_dmabuf_ops = {
.kmap_atomic = nouveau_gem_kmap_atomic,
.kunmap = nouveau_gem_kunmap,
.kunmap_atomic = nouveau_gem_kunmap_atomic,
+ .mmap = nouveau_gem_prime_mmap,
+ .vmap = nouveau_gem_prime_vmap,
+ .vunmap = nouveau_gem_prime_vunmap,
};
static int
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index 58991af9050..01550d05e27 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -1029,6 +1029,11 @@ int evergreen_pcie_gart_enable(struct radeon_device *rdev)
WREG32(MC_VM_MD_L1_TLB0_CNTL, tmp);
WREG32(MC_VM_MD_L1_TLB1_CNTL, tmp);
WREG32(MC_VM_MD_L1_TLB2_CNTL, tmp);
+ if ((rdev->family == CHIP_JUNIPER) ||
+ (rdev->family == CHIP_CYPRESS) ||
+ (rdev->family == CHIP_HEMLOCK) ||
+ (rdev->family == CHIP_BARTS))
+ WREG32(MC_VM_MD_L1_TLB3_CNTL, tmp);
}
WREG32(MC_VM_MB_L1_TLB0_CNTL, tmp);
WREG32(MC_VM_MB_L1_TLB1_CNTL, tmp);
@@ -1553,163 +1558,10 @@ int evergreen_cp_resume(struct radeon_device *rdev)
/*
* Core functions
*/
-static u32 evergreen_get_tile_pipe_to_backend_map(struct radeon_device *rdev,
- u32 num_tile_pipes,
- u32 num_backends,
- u32 backend_disable_mask)
-{
- u32 backend_map = 0;
- u32 enabled_backends_mask = 0;
- u32 enabled_backends_count = 0;
- u32 cur_pipe;
- u32 swizzle_pipe[EVERGREEN_MAX_PIPES];
- u32 cur_backend = 0;
- u32 i;
- bool force_no_swizzle;
-
- if (num_tile_pipes > EVERGREEN_MAX_PIPES)
- num_tile_pipes = EVERGREEN_MAX_PIPES;
- if (num_tile_pipes < 1)
- num_tile_pipes = 1;
- if (num_backends > EVERGREEN_MAX_BACKENDS)
- num_backends = EVERGREEN_MAX_BACKENDS;
- if (num_backends < 1)
- num_backends = 1;
-
- for (i = 0; i < EVERGREEN_MAX_BACKENDS; ++i) {
- if (((backend_disable_mask >> i) & 1) == 0) {
- enabled_backends_mask |= (1 << i);
- ++enabled_backends_count;
- }
- if (enabled_backends_count == num_backends)
- break;
- }
-
- if (enabled_backends_count == 0) {
- enabled_backends_mask = 1;
- enabled_backends_count = 1;
- }
-
- if (enabled_backends_count != num_backends)
- num_backends = enabled_backends_count;
-
- memset((uint8_t *)&swizzle_pipe[0], 0, sizeof(u32) * EVERGREEN_MAX_PIPES);
- switch (rdev->family) {
- case CHIP_CEDAR:
- case CHIP_REDWOOD:
- case CHIP_PALM:
- case CHIP_SUMO:
- case CHIP_SUMO2:
- case CHIP_TURKS:
- case CHIP_CAICOS:
- force_no_swizzle = false;
- break;
- case CHIP_CYPRESS:
- case CHIP_HEMLOCK:
- case CHIP_JUNIPER:
- case CHIP_BARTS:
- default:
- force_no_swizzle = true;
- break;
- }
- if (force_no_swizzle) {
- bool last_backend_enabled = false;
-
- force_no_swizzle = false;
- for (i = 0; i < EVERGREEN_MAX_BACKENDS; ++i) {
- if (((enabled_backends_mask >> i) & 1) == 1) {
- if (last_backend_enabled)
- force_no_swizzle = true;
- last_backend_enabled = true;
- } else
- last_backend_enabled = false;
- }
- }
-
- switch (num_tile_pipes) {
- case 1:
- case 3:
- case 5:
- case 7:
- DRM_ERROR("odd number of pipes!\n");
- break;
- case 2:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- break;
- case 4:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 1;
- swizzle_pipe[3] = 3;
- }
- break;
- case 6:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- swizzle_pipe[5] = 5;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 1;
- swizzle_pipe[4] = 3;
- swizzle_pipe[5] = 5;
- }
- break;
- case 8:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- swizzle_pipe[5] = 5;
- swizzle_pipe[6] = 6;
- swizzle_pipe[7] = 7;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 6;
- swizzle_pipe[4] = 1;
- swizzle_pipe[5] = 3;
- swizzle_pipe[6] = 5;
- swizzle_pipe[7] = 7;
- }
- break;
- }
-
- for (cur_pipe = 0; cur_pipe < num_tile_pipes; ++cur_pipe) {
- while (((1 << cur_backend) & enabled_backends_mask) == 0)
- cur_backend = (cur_backend + 1) % EVERGREEN_MAX_BACKENDS;
-
- backend_map |= (((cur_backend & 0xf) << (swizzle_pipe[cur_pipe] * 4)));
-
- cur_backend = (cur_backend + 1) % EVERGREEN_MAX_BACKENDS;
- }
-
- return backend_map;
-}
-
static void evergreen_gpu_init(struct radeon_device *rdev)
{
- u32 cc_rb_backend_disable = 0;
- u32 cc_gc_shader_pipe_config;
- u32 gb_addr_config = 0;
+ u32 gb_addr_config;
u32 mc_shared_chmap, mc_arb_ramcfg;
- u32 gb_backend_map;
- u32 grbm_gfx_index;
u32 sx_debug_1;
u32 smx_dc_ctl0;
u32 sq_config;
@@ -1724,6 +1576,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
u32 sq_stack_resource_mgmt_3;
u32 vgt_cache_invalidation;
u32 hdp_host_path_cntl, tmp;
+ u32 disabled_rb_mask;
int i, j, num_shader_engines, ps_thread_count;
switch (rdev->family) {
@@ -1748,6 +1601,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x100;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = CYPRESS_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_JUNIPER:
rdev->config.evergreen.num_ses = 1;
@@ -1769,6 +1623,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x100;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = JUNIPER_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_REDWOOD:
rdev->config.evergreen.num_ses = 1;
@@ -1790,6 +1645,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x100;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = REDWOOD_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_CEDAR:
default:
@@ -1812,6 +1668,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x40;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = CEDAR_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_PALM:
rdev->config.evergreen.num_ses = 1;
@@ -1833,6 +1690,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x40;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = CEDAR_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_SUMO:
rdev->config.evergreen.num_ses = 1;
@@ -1860,6 +1718,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x40;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = REDWOOD_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_SUMO2:
rdev->config.evergreen.num_ses = 1;
@@ -1881,6 +1740,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x40;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = REDWOOD_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_BARTS:
rdev->config.evergreen.num_ses = 2;
@@ -1902,6 +1762,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x100;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = BARTS_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_TURKS:
rdev->config.evergreen.num_ses = 1;
@@ -1923,6 +1784,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x100;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = TURKS_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_CAICOS:
rdev->config.evergreen.num_ses = 1;
@@ -1944,6 +1806,7 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
rdev->config.evergreen.sc_prim_fifo_size = 0x40;
rdev->config.evergreen.sc_hiz_tile_fifo_size = 0x30;
rdev->config.evergreen.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = CAICOS_GB_ADDR_CONFIG_GOLDEN;
break;
}
@@ -1960,20 +1823,6 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
evergreen_fix_pci_max_read_req_size(rdev);
- cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG) & ~2;
-
- cc_gc_shader_pipe_config |=
- INACTIVE_QD_PIPES((EVERGREEN_MAX_PIPES_MASK << rdev->config.evergreen.max_pipes)
- & EVERGREEN_MAX_PIPES_MASK);
- cc_gc_shader_pipe_config |=
- INACTIVE_SIMDS((EVERGREEN_MAX_SIMDS_MASK << rdev->config.evergreen.max_simds)
- & EVERGREEN_MAX_SIMDS_MASK);
-
- cc_rb_backend_disable =
- BACKEND_DISABLE((EVERGREEN_MAX_BACKENDS_MASK << rdev->config.evergreen.max_backends)
- & EVERGREEN_MAX_BACKENDS_MASK);
-
-
mc_shared_chmap = RREG32(MC_SHARED_CHMAP);
if ((rdev->family == CHIP_PALM) ||
(rdev->family == CHIP_SUMO) ||
@@ -1982,134 +1831,6 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
else
mc_arb_ramcfg = RREG32(MC_ARB_RAMCFG);
- switch (rdev->config.evergreen.max_tile_pipes) {
- case 1:
- default:
- gb_addr_config |= NUM_PIPES(0);
- break;
- case 2:
- gb_addr_config |= NUM_PIPES(1);
- break;
- case 4:
- gb_addr_config |= NUM_PIPES(2);
- break;
- case 8:
- gb_addr_config |= NUM_PIPES(3);
- break;
- }
-
- gb_addr_config |= PIPE_INTERLEAVE_SIZE((mc_arb_ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT);
- gb_addr_config |= BANK_INTERLEAVE_SIZE(0);
- gb_addr_config |= NUM_SHADER_ENGINES(rdev->config.evergreen.num_ses - 1);
- gb_addr_config |= SHADER_ENGINE_TILE_SIZE(1);
- gb_addr_config |= NUM_GPUS(0); /* Hemlock? */
- gb_addr_config |= MULTI_GPU_TILE_SIZE(2);
-
- if (((mc_arb_ramcfg & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT) > 2)
- gb_addr_config |= ROW_SIZE(2);
- else
- gb_addr_config |= ROW_SIZE((mc_arb_ramcfg & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT);
-
- if (rdev->ddev->pdev->device == 0x689e) {
- u32 efuse_straps_4;
- u32 efuse_straps_3;
- u8 efuse_box_bit_131_124;
-
- WREG32(RCU_IND_INDEX, 0x204);
- efuse_straps_4 = RREG32(RCU_IND_DATA);
- WREG32(RCU_IND_INDEX, 0x203);
- efuse_straps_3 = RREG32(RCU_IND_DATA);
- efuse_box_bit_131_124 = (u8)(((efuse_straps_4 & 0xf) << 4) | ((efuse_straps_3 & 0xf0000000) >> 28));
-
- switch(efuse_box_bit_131_124) {
- case 0x00:
- gb_backend_map = 0x76543210;
- break;
- case 0x55:
- gb_backend_map = 0x77553311;
- break;
- case 0x56:
- gb_backend_map = 0x77553300;
- break;
- case 0x59:
- gb_backend_map = 0x77552211;
- break;
- case 0x66:
- gb_backend_map = 0x77443300;
- break;
- case 0x99:
- gb_backend_map = 0x66552211;
- break;
- case 0x5a:
- gb_backend_map = 0x77552200;
- break;
- case 0xaa:
- gb_backend_map = 0x66442200;
- break;
- case 0x95:
- gb_backend_map = 0x66553311;
- break;
- default:
- DRM_ERROR("bad backend map, using default\n");
- gb_backend_map =
- evergreen_get_tile_pipe_to_backend_map(rdev,
- rdev->config.evergreen.max_tile_pipes,
- rdev->config.evergreen.max_backends,
- ((EVERGREEN_MAX_BACKENDS_MASK <<
- rdev->config.evergreen.max_backends) &
- EVERGREEN_MAX_BACKENDS_MASK));
- break;
- }
- } else if (rdev->ddev->pdev->device == 0x68b9) {
- u32 efuse_straps_3;
- u8 efuse_box_bit_127_124;
-
- WREG32(RCU_IND_INDEX, 0x203);
- efuse_straps_3 = RREG32(RCU_IND_DATA);
- efuse_box_bit_127_124 = (u8)((efuse_straps_3 & 0xF0000000) >> 28);
-
- switch(efuse_box_bit_127_124) {
- case 0x0:
- gb_backend_map = 0x00003210;
- break;
- case 0x5:
- case 0x6:
- case 0x9:
- case 0xa:
- gb_backend_map = 0x00003311;
- break;
- default:
- DRM_ERROR("bad backend map, using default\n");
- gb_backend_map =
- evergreen_get_tile_pipe_to_backend_map(rdev,
- rdev->config.evergreen.max_tile_pipes,
- rdev->config.evergreen.max_backends,
- ((EVERGREEN_MAX_BACKENDS_MASK <<
- rdev->config.evergreen.max_backends) &
- EVERGREEN_MAX_BACKENDS_MASK));
- break;
- }
- } else {
- switch (rdev->family) {
- case CHIP_CYPRESS:
- case CHIP_HEMLOCK:
- case CHIP_BARTS:
- gb_backend_map = 0x66442200;
- break;
- case CHIP_JUNIPER:
- gb_backend_map = 0x00002200;
- break;
- default:
- gb_backend_map =
- evergreen_get_tile_pipe_to_backend_map(rdev,
- rdev->config.evergreen.max_tile_pipes,
- rdev->config.evergreen.max_backends,
- ((EVERGREEN_MAX_BACKENDS_MASK <<
- rdev->config.evergreen.max_backends) &
- EVERGREEN_MAX_BACKENDS_MASK));
- }
- }
-
/* setup tiling info dword. gb_addr_config is not adequate since it does
* not have bank info, so create a custom tiling dword.
* bits 3:0 num_pipes
@@ -2136,45 +1857,54 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
/* num banks is 8 on all fusion asics. 0 = 4, 1 = 8, 2 = 16 */
if (rdev->flags & RADEON_IS_IGP)
rdev->config.evergreen.tile_config |= 1 << 4;
- else
- rdev->config.evergreen.tile_config |=
- ((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT) << 4;
- rdev->config.evergreen.tile_config |=
- ((mc_arb_ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT) << 8;
+ else {
+ if ((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT)
+ rdev->config.evergreen.tile_config |= 1 << 4;
+ else
+ rdev->config.evergreen.tile_config |= 0 << 4;
+ }
+ rdev->config.evergreen.tile_config |= 0 << 8;
rdev->config.evergreen.tile_config |=
((gb_addr_config & 0x30000000) >> 28) << 12;
- rdev->config.evergreen.backend_map = gb_backend_map;
- WREG32(GB_BACKEND_MAP, gb_backend_map);
- WREG32(GB_ADDR_CONFIG, gb_addr_config);
- WREG32(DMIF_ADDR_CONFIG, gb_addr_config);
- WREG32(HDP_ADDR_CONFIG, gb_addr_config);
-
- num_shader_engines = ((RREG32(GB_ADDR_CONFIG) & NUM_SHADER_ENGINES(3)) >> 12) + 1;
- grbm_gfx_index = INSTANCE_BROADCAST_WRITES;
+ num_shader_engines = (gb_addr_config & NUM_SHADER_ENGINES(3) >> 12) + 1;
- for (i = 0; i < rdev->config.evergreen.num_ses; i++) {
- u32 rb = cc_rb_backend_disable | (0xf0 << 16);
- u32 sp = cc_gc_shader_pipe_config;
- u32 gfx = grbm_gfx_index | SE_INDEX(i);
+ if ((rdev->family >= CHIP_CEDAR) && (rdev->family <= CHIP_HEMLOCK)) {
+ u32 efuse_straps_4;
+ u32 efuse_straps_3;
- if (i == num_shader_engines) {
- rb |= BACKEND_DISABLE(EVERGREEN_MAX_BACKENDS_MASK);
- sp |= INACTIVE_SIMDS(EVERGREEN_MAX_SIMDS_MASK);
+ WREG32(RCU_IND_INDEX, 0x204);
+ efuse_straps_4 = RREG32(RCU_IND_DATA);
+ WREG32(RCU_IND_INDEX, 0x203);
+ efuse_straps_3 = RREG32(RCU_IND_DATA);
+ tmp = (((efuse_straps_4 & 0xf) << 4) |
+ ((efuse_straps_3 & 0xf0000000) >> 28));
+ } else {
+ tmp = 0;
+ for (i = (rdev->config.evergreen.num_ses - 1); i >= 0; i--) {
+ u32 rb_disable_bitmap;
+
+ WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i));
+ WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i));
+ rb_disable_bitmap = (RREG32(CC_RB_BACKEND_DISABLE) & 0x00ff0000) >> 16;
+ tmp <<= 4;
+ tmp |= rb_disable_bitmap;
}
+ }
+ /* enabled rb are just the one not disabled :) */
+ disabled_rb_mask = tmp;
- WREG32(GRBM_GFX_INDEX, gfx);
- WREG32(RLC_GFX_INDEX, gfx);
+ WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES);
+ WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES);
- WREG32(CC_RB_BACKEND_DISABLE, rb);
- WREG32(CC_SYS_RB_BACKEND_DISABLE, rb);
- WREG32(GC_USER_RB_BACKEND_DISABLE, rb);
- WREG32(CC_GC_SHADER_PIPE_CONFIG, sp);
- }
+ WREG32(GB_ADDR_CONFIG, gb_addr_config);
+ WREG32(DMIF_ADDR_CONFIG, gb_addr_config);
+ WREG32(HDP_ADDR_CONFIG, gb_addr_config);
- grbm_gfx_index |= SE_BROADCAST_WRITES;
- WREG32(GRBM_GFX_INDEX, grbm_gfx_index);
- WREG32(RLC_GFX_INDEX, grbm_gfx_index);
+ tmp = gb_addr_config & NUM_PIPES_MASK;
+ tmp = r6xx_remap_render_backend(rdev, tmp, rdev->config.evergreen.max_backends,
+ EVERGREEN_MAX_BACKENDS, disabled_rb_mask);
+ WREG32(GB_BACKEND_MAP, tmp);
WREG32(CGTS_SYS_TCC_DISABLE, 0);
WREG32(CGTS_TCC_DISABLE, 0);
diff --git a/drivers/gpu/drm/radeon/evergreend.h b/drivers/gpu/drm/radeon/evergreend.h
index 79130bfd1d6..2773039b490 100644
--- a/drivers/gpu/drm/radeon/evergreend.h
+++ b/drivers/gpu/drm/radeon/evergreend.h
@@ -37,6 +37,15 @@
#define EVERGREEN_MAX_PIPES_MASK 0xFF
#define EVERGREEN_MAX_LDS_NUM 0xFFFF
+#define CYPRESS_GB_ADDR_CONFIG_GOLDEN 0x02011003
+#define BARTS_GB_ADDR_CONFIG_GOLDEN 0x02011003
+#define CAYMAN_GB_ADDR_CONFIG_GOLDEN 0x02011003
+#define JUNIPER_GB_ADDR_CONFIG_GOLDEN 0x02010002
+#define REDWOOD_GB_ADDR_CONFIG_GOLDEN 0x02010002
+#define TURKS_GB_ADDR_CONFIG_GOLDEN 0x02010002
+#define CEDAR_GB_ADDR_CONFIG_GOLDEN 0x02010001
+#define CAICOS_GB_ADDR_CONFIG_GOLDEN 0x02010001
+
/* Registers */
#define RCU_IND_INDEX 0x100
@@ -54,6 +63,7 @@
#define BACKEND_DISABLE(x) ((x) << 16)
#define GB_ADDR_CONFIG 0x98F8
#define NUM_PIPES(x) ((x) << 0)
+#define NUM_PIPES_MASK 0x0000000f
#define PIPE_INTERLEAVE_SIZE(x) ((x) << 4)
#define BANK_INTERLEAVE_SIZE(x) ((x) << 8)
#define NUM_SHADER_ENGINES(x) ((x) << 12)
@@ -452,6 +462,7 @@
#define MC_VM_MD_L1_TLB0_CNTL 0x2654
#define MC_VM_MD_L1_TLB1_CNTL 0x2658
#define MC_VM_MD_L1_TLB2_CNTL 0x265C
+#define MC_VM_MD_L1_TLB3_CNTL 0x2698
#define FUS_MC_VM_MD_L1_TLB0_CNTL 0x265C
#define FUS_MC_VM_MD_L1_TLB1_CNTL 0x2660
diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c
index ce4e7cc6c90..3df4efa1194 100644
--- a/drivers/gpu/drm/radeon/ni.c
+++ b/drivers/gpu/drm/radeon/ni.c
@@ -417,215 +417,17 @@ out:
/*
* Core functions
*/
-static u32 cayman_get_tile_pipe_to_backend_map(struct radeon_device *rdev,
- u32 num_tile_pipes,
- u32 num_backends_per_asic,
- u32 *backend_disable_mask_per_asic,
- u32 num_shader_engines)
-{
- u32 backend_map = 0;
- u32 enabled_backends_mask = 0;
- u32 enabled_backends_count = 0;
- u32 num_backends_per_se;
- u32 cur_pipe;
- u32 swizzle_pipe[CAYMAN_MAX_PIPES];
- u32 cur_backend = 0;
- u32 i;
- bool force_no_swizzle;
-
- /* force legal values */
- if (num_tile_pipes < 1)
- num_tile_pipes = 1;
- if (num_tile_pipes > rdev->config.cayman.max_tile_pipes)
- num_tile_pipes = rdev->config.cayman.max_tile_pipes;
- if (num_shader_engines < 1)
- num_shader_engines = 1;
- if (num_shader_engines > rdev->config.cayman.max_shader_engines)
- num_shader_engines = rdev->config.cayman.max_shader_engines;
- if (num_backends_per_asic < num_shader_engines)
- num_backends_per_asic = num_shader_engines;
- if (num_backends_per_asic > (rdev->config.cayman.max_backends_per_se * num_shader_engines))
- num_backends_per_asic = rdev->config.cayman.max_backends_per_se * num_shader_engines;
-
- /* make sure we have the same number of backends per se */
- num_backends_per_asic = ALIGN(num_backends_per_asic, num_shader_engines);
- /* set up the number of backends per se */
- num_backends_per_se = num_backends_per_asic / num_shader_engines;
- if (num_backends_per_se > rdev->config.cayman.max_backends_per_se) {
- num_backends_per_se = rdev->config.cayman.max_backends_per_se;
- num_backends_per_asic = num_backends_per_se * num_shader_engines;
- }
-
- /* create enable mask and count for enabled backends */
- for (i = 0; i < CAYMAN_MAX_BACKENDS; ++i) {
- if (((*backend_disable_mask_per_asic >> i) & 1) == 0) {
- enabled_backends_mask |= (1 << i);
- ++enabled_backends_count;
- }
- if (enabled_backends_count == num_backends_per_asic)
- break;
- }
-
- /* force the backends mask to match the current number of backends */
- if (enabled_backends_count != num_backends_per_asic) {
- u32 this_backend_enabled;
- u32 shader_engine;
- u32 backend_per_se;
-
- enabled_backends_mask = 0;
- enabled_backends_count = 0;
- *backend_disable_mask_per_asic = CAYMAN_MAX_BACKENDS_MASK;
- for (i = 0; i < CAYMAN_MAX_BACKENDS; ++i) {
- /* calc the current se */
- shader_engine = i / rdev->config.cayman.max_backends_per_se;
- /* calc the backend per se */
- backend_per_se = i % rdev->config.cayman.max_backends_per_se;
- /* default to not enabled */
- this_backend_enabled = 0;
- if ((shader_engine < num_shader_engines) &&
- (backend_per_se < num_backends_per_se))
- this_backend_enabled = 1;
- if (this_backend_enabled) {
- enabled_backends_mask |= (1 << i);
- *backend_disable_mask_per_asic &= ~(1 << i);
- ++enabled_backends_count;
- }
- }
- }
-
-
- memset((uint8_t *)&swizzle_pipe[0], 0, sizeof(u32) * CAYMAN_MAX_PIPES);
- switch (rdev->family) {
- case CHIP_CAYMAN:
- case CHIP_ARUBA:
- force_no_swizzle = true;
- break;
- default:
- force_no_swizzle = false;
- break;
- }
- if (force_no_swizzle) {
- bool last_backend_enabled = false;
-
- force_no_swizzle = false;
- for (i = 0; i < CAYMAN_MAX_BACKENDS; ++i) {
- if (((enabled_backends_mask >> i) & 1) == 1) {
- if (last_backend_enabled)
- force_no_swizzle = true;
- last_backend_enabled = true;
- } else
- last_backend_enabled = false;
- }
- }
-
- switch (num_tile_pipes) {
- case 1:
- case 3:
- case 5:
- case 7:
- DRM_ERROR("odd number of pipes!\n");
- break;
- case 2:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- break;
- case 4:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 1;
- swizzle_pipe[3] = 3;
- }
- break;
- case 6:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- swizzle_pipe[5] = 5;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 1;
- swizzle_pipe[4] = 3;
- swizzle_pipe[5] = 5;
- }
- break;
- case 8:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- swizzle_pipe[5] = 5;
- swizzle_pipe[6] = 6;
- swizzle_pipe[7] = 7;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 6;
- swizzle_pipe[4] = 1;
- swizzle_pipe[5] = 3;
- swizzle_pipe[6] = 5;
- swizzle_pipe[7] = 7;
- }
- break;
- }
-
- for (cur_pipe = 0; cur_pipe < num_tile_pipes; ++cur_pipe) {
- while (((1 << cur_backend) & enabled_backends_mask) == 0)
- cur_backend = (cur_backend + 1) % CAYMAN_MAX_BACKENDS;
-
- backend_map |= (((cur_backend & 0xf) << (swizzle_pipe[cur_pipe] * 4)));
-
- cur_backend = (cur_backend + 1) % CAYMAN_MAX_BACKENDS;
- }
-
- return backend_map;
-}
-
-static u32 cayman_get_disable_mask_per_asic(struct radeon_device *rdev,
- u32 disable_mask_per_se,
- u32 max_disable_mask_per_se,
- u32 num_shader_engines)
-{
- u32 disable_field_width_per_se = r600_count_pipe_bits(disable_mask_per_se);
- u32 disable_mask_per_asic = disable_mask_per_se & max_disable_mask_per_se;
-
- if (num_shader_engines == 1)
- return disable_mask_per_asic;
- else if (num_shader_engines == 2)
- return disable_mask_per_asic | (disable_mask_per_asic << disable_field_width_per_se);
- else
- return 0xffffffff;
-}
-
static void cayman_gpu_init(struct radeon_device *rdev)
{
- u32 cc_rb_backend_disable = 0;
- u32 cc_gc_shader_pipe_config;
u32 gb_addr_config = 0;
u32 mc_shared_chmap, mc_arb_ramcfg;
- u32 gb_backend_map;
u32 cgts_tcc_disable;
u32 sx_debug_1;
u32 smx_dc_ctl0;
- u32 gc_user_shader_pipe_config;
- u32 gc_user_rb_backend_disable;
- u32 cgts_user_tcc_disable;
u32 cgts_sm_ctrl_reg;
u32 hdp_host_path_cntl;
u32 tmp;
+ u32 disabled_rb_mask;
int i, j;
switch (rdev->family) {
@@ -650,6 +452,7 @@ static void cayman_gpu_init(struct radeon_device *rdev)
rdev->config.cayman.sc_prim_fifo_size = 0x100;
rdev->config.cayman.sc_hiz_tile_fifo_size = 0x30;
rdev->config.cayman.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = CAYMAN_GB_ADDR_CONFIG_GOLDEN;
break;
case CHIP_ARUBA:
default:
@@ -687,6 +490,7 @@ static void cayman_gpu_init(struct radeon_device *rdev)
rdev->config.cayman.sc_prim_fifo_size = 0x40;
rdev->config.cayman.sc_hiz_tile_fifo_size = 0x30;
rdev->config.cayman.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = ARUBA_GB_ADDR_CONFIG_GOLDEN;
break;
}
@@ -706,39 +510,6 @@ static void cayman_gpu_init(struct radeon_device *rdev)
mc_shared_chmap = RREG32(MC_SHARED_CHMAP);
mc_arb_ramcfg = RREG32(MC_ARB_RAMCFG);
- cc_rb_backend_disable = RREG32(CC_RB_BACKEND_DISABLE);
- cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG);
- cgts_tcc_disable = 0xffff0000;
- for (i = 0; i < rdev->config.cayman.max_texture_channel_caches; i++)
- cgts_tcc_disable &= ~(1 << (16 + i));
- gc_user_rb_backend_disable = RREG32(GC_USER_RB_BACKEND_DISABLE);
- gc_user_shader_pipe_config = RREG32(GC_USER_SHADER_PIPE_CONFIG);
- cgts_user_tcc_disable = RREG32(CGTS_USER_TCC_DISABLE);
-
- rdev->config.cayman.num_shader_engines = rdev->config.cayman.max_shader_engines;
- tmp = ((~gc_user_shader_pipe_config) & INACTIVE_QD_PIPES_MASK) >> INACTIVE_QD_PIPES_SHIFT;
- rdev->config.cayman.num_shader_pipes_per_simd = r600_count_pipe_bits(tmp);
- rdev->config.cayman.num_tile_pipes = rdev->config.cayman.max_tile_pipes;
- tmp = ((~gc_user_shader_pipe_config) & INACTIVE_SIMDS_MASK) >> INACTIVE_SIMDS_SHIFT;
- rdev->config.cayman.num_simds_per_se = r600_count_pipe_bits(tmp);
- tmp = ((~gc_user_rb_backend_disable) & BACKEND_DISABLE_MASK) >> BACKEND_DISABLE_SHIFT;
- rdev->config.cayman.num_backends_per_se = r600_count_pipe_bits(tmp);
- tmp = (gc_user_rb_backend_disable & BACKEND_DISABLE_MASK) >> BACKEND_DISABLE_SHIFT;
- rdev->config.cayman.backend_disable_mask_per_asic =
- cayman_get_disable_mask_per_asic(rdev, tmp, CAYMAN_MAX_BACKENDS_PER_SE_MASK,
- rdev->config.cayman.num_shader_engines);
- rdev->config.cayman.backend_map =
- cayman_get_tile_pipe_to_backend_map(rdev, rdev->config.cayman.num_tile_pipes,
- rdev->config.cayman.num_backends_per_se *
- rdev->config.cayman.num_shader_engines,
- &rdev->config.cayman.backend_disable_mask_per_asic,
- rdev->config.cayman.num_shader_engines);
- tmp = ((~cgts_user_tcc_disable) & TCC_DISABLE_MASK) >> TCC_DISABLE_SHIFT;
- rdev->config.cayman.num_texture_channel_caches = r600_count_pipe_bits(tmp);
- tmp = (mc_arb_ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT;
- rdev->config.cayman.mem_max_burst_length_bytes = (tmp + 1) * 256;
- if (rdev->config.cayman.mem_max_burst_length_bytes > 512)
- rdev->config.cayman.mem_max_burst_length_bytes = 512;
tmp = (mc_arb_ramcfg & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT;
rdev->config.cayman.mem_row_size_in_kb = (4 * (1 << (8 + tmp))) / 1024;
if (rdev->config.cayman.mem_row_size_in_kb > 4)
@@ -748,73 +519,6 @@ static void cayman_gpu_init(struct radeon_device *rdev)
rdev->config.cayman.num_gpus = 1;
rdev->config.cayman.multi_gpu_tile_size = 64;
- //gb_addr_config = 0x02011003
-#if 0
- gb_addr_config = RREG32(GB_ADDR_CONFIG);
-#else
- gb_addr_config = 0;
- switch (rdev->config.cayman.num_tile_pipes) {
- case 1:
- default:
- gb_addr_config |= NUM_PIPES(0);
- break;
- case 2:
- gb_addr_config |= NUM_PIPES(1);
- break;
- case 4:
- gb_addr_config |= NUM_PIPES(2);
- break;
- case 8:
- gb_addr_config |= NUM_PIPES(3);
- break;
- }
-
- tmp = (rdev->config.cayman.mem_max_burst_length_bytes / 256) - 1;
- gb_addr_config |= PIPE_INTERLEAVE_SIZE(tmp);
- gb_addr_config |= NUM_SHADER_ENGINES(rdev->config.cayman.num_shader_engines - 1);
- tmp = (rdev->config.cayman.shader_engine_tile_size / 16) - 1;
- gb_addr_config |= SHADER_ENGINE_TILE_SIZE(tmp);
- switch (rdev->config.cayman.num_gpus) {
- case 1:
- default:
- gb_addr_config |= NUM_GPUS(0);
- break;
- case 2:
- gb_addr_config |= NUM_GPUS(1);
- break;
- case 4:
- gb_addr_config |= NUM_GPUS(2);
- break;
- }
- switch (rdev->config.cayman.multi_gpu_tile_size) {
- case 16:
- gb_addr_config |= MULTI_GPU_TILE_SIZE(0);
- break;
- case 32:
- default:
- gb_addr_config |= MULTI_GPU_TILE_SIZE(1);
- break;
- case 64:
- gb_addr_config |= MULTI_GPU_TILE_SIZE(2);
- break;
- case 128:
- gb_addr_config |= MULTI_GPU_TILE_SIZE(3);
- break;
- }
- switch (rdev->config.cayman.mem_row_size_in_kb) {
- case 1:
- default:
- gb_addr_config |= ROW_SIZE(0);
- break;
- case 2:
- gb_addr_config |= ROW_SIZE(1);
- break;
- case 4:
- gb_addr_config |= ROW_SIZE(2);
- break;
- }
-#endif
-
tmp = (gb_addr_config & NUM_PIPES_MASK) >> NUM_PIPES_SHIFT;
rdev->config.cayman.num_tile_pipes = (1 << tmp);
tmp = (gb_addr_config & PIPE_INTERLEAVE_SIZE_MASK) >> PIPE_INTERLEAVE_SIZE_SHIFT;
@@ -828,17 +532,7 @@ static void cayman_gpu_init(struct radeon_device *rdev)
tmp = (gb_addr_config & ROW_SIZE_MASK) >> ROW_SIZE_SHIFT;
rdev->config.cayman.mem_row_size_in_kb = 1 << tmp;
- //gb_backend_map = 0x76541032;
-#if 0
- gb_backend_map = RREG32(GB_BACKEND_MAP);
-#else
- gb_backend_map =
- cayman_get_tile_pipe_to_backend_map(rdev, rdev->config.cayman.num_tile_pipes,
- rdev->config.cayman.num_backends_per_se *
- rdev->config.cayman.num_shader_engines,
- &rdev->config.cayman.backend_disable_mask_per_asic,
- rdev->config.cayman.num_shader_engines);
-#endif
+
/* setup tiling info dword. gb_addr_config is not adequate since it does
* not have bank info, so create a custom tiling dword.
* bits 3:0 num_pipes
@@ -866,33 +560,49 @@ static void cayman_gpu_init(struct radeon_device *rdev)
/* num banks is 8 on all fusion asics. 0 = 4, 1 = 8, 2 = 16 */
if (rdev->flags & RADEON_IS_IGP)
rdev->config.cayman.tile_config |= 1 << 4;
- else
- rdev->config.cayman.tile_config |=
- ((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT) << 4;
+ else {
+ if ((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT)
+ rdev->config.cayman.tile_config |= 1 << 4;
+ else
+ rdev->config.cayman.tile_config |= 0 << 4;
+ }
rdev->config.cayman.tile_config |=
((gb_addr_config & PIPE_INTERLEAVE_SIZE_MASK) >> PIPE_INTERLEAVE_SIZE_SHIFT) << 8;
rdev->config.cayman.tile_config |=
((gb_addr_config & ROW_SIZE_MASK) >> ROW_SIZE_SHIFT) << 12;
- rdev->config.cayman.backend_map = gb_backend_map;
- WREG32(GB_BACKEND_MAP, gb_backend_map);
+ tmp = 0;
+ for (i = (rdev->config.cayman.max_shader_engines - 1); i >= 0; i--) {
+ u32 rb_disable_bitmap;
+
+ WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i));
+ WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_INDEX(i));
+ rb_disable_bitmap = (RREG32(CC_RB_BACKEND_DISABLE) & 0x00ff0000) >> 16;
+ tmp <<= 4;
+ tmp |= rb_disable_bitmap;
+ }
+ /* enabled rb are just the one not disabled :) */
+ disabled_rb_mask = tmp;
+
+ WREG32(GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES);
+ WREG32(RLC_GFX_INDEX, INSTANCE_BROADCAST_WRITES | SE_BROADCAST_WRITES);
+
WREG32(GB_ADDR_CONFIG, gb_addr_config);
WREG32(DMIF_ADDR_CONFIG, gb_addr_config);
WREG32(HDP_ADDR_CONFIG, gb_addr_config);
- /* primary versions */
- WREG32(CC_RB_BACKEND_DISABLE, cc_rb_backend_disable);
- WREG32(CC_SYS_RB_BACKEND_DISABLE, cc_rb_backend_disable);
- WREG32(CC_GC_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config);
+ tmp = gb_addr_config & NUM_PIPES_MASK;
+ tmp = r6xx_remap_render_backend(rdev, tmp,
+ rdev->config.cayman.max_backends_per_se *
+ rdev->config.cayman.max_shader_engines,
+ CAYMAN_MAX_BACKENDS, disabled_rb_mask);
+ WREG32(GB_BACKEND_MAP, tmp);
+ cgts_tcc_disable = 0xffff0000;
+ for (i = 0; i < rdev->config.cayman.max_texture_channel_caches; i++)
+ cgts_tcc_disable &= ~(1 << (16 + i));
WREG32(CGTS_TCC_DISABLE, cgts_tcc_disable);
WREG32(CGTS_SYS_TCC_DISABLE, cgts_tcc_disable);
-
- /* user versions */
- WREG32(GC_USER_RB_BACKEND_DISABLE, cc_rb_backend_disable);
- WREG32(GC_USER_SYS_RB_BACKEND_DISABLE, cc_rb_backend_disable);
- WREG32(GC_USER_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config);
-
WREG32(CGTS_USER_SYS_TCC_DISABLE, cgts_tcc_disable);
WREG32(CGTS_USER_TCC_DISABLE, cgts_tcc_disable);
diff --git a/drivers/gpu/drm/radeon/nid.h b/drivers/gpu/drm/radeon/nid.h
index 2aa7046ada5..a0b98066e20 100644
--- a/drivers/gpu/drm/radeon/nid.h
+++ b/drivers/gpu/drm/radeon/nid.h
@@ -41,6 +41,9 @@
#define CAYMAN_MAX_TCC 16
#define CAYMAN_MAX_TCC_MASK 0xFF
+#define CAYMAN_GB_ADDR_CONFIG_GOLDEN 0x02011003
+#define ARUBA_GB_ADDR_CONFIG_GOLDEN 0x12010001
+
#define DMIF_ADDR_CONFIG 0xBD4
#define SRBM_GFX_CNTL 0x0E44
#define RINGID(x) (((x) & 0x3) << 0)
@@ -148,6 +151,8 @@
#define CGTS_SYS_TCC_DISABLE 0x3F90
#define CGTS_USER_SYS_TCC_DISABLE 0x3F94
+#define RLC_GFX_INDEX 0x3FC4
+
#define CONFIG_MEMSIZE 0x5428
#define HDP_MEM_COHERENCY_FLUSH_CNTL 0x5480
@@ -212,6 +217,12 @@
#define SOFT_RESET_VGT (1 << 14)
#define SOFT_RESET_IA (1 << 15)
+#define GRBM_GFX_INDEX 0x802C
+#define INSTANCE_INDEX(x) ((x) << 0)
+#define SE_INDEX(x) ((x) << 16)
+#define INSTANCE_BROADCAST_WRITES (1 << 30)
+#define SE_BROADCAST_WRITES (1 << 31)
+
#define SCRATCH_REG0 0x8500
#define SCRATCH_REG1 0x8504
#define SCRATCH_REG2 0x8508
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index f388a1d73b6..45cfcea6350 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -1376,113 +1376,51 @@ int r600_asic_reset(struct radeon_device *rdev)
return r600_gpu_soft_reset(rdev);
}
-static u32 r600_get_tile_pipe_to_backend_map(u32 num_tile_pipes,
- u32 num_backends,
- u32 backend_disable_mask)
-{
- u32 backend_map = 0;
- u32 enabled_backends_mask;
- u32 enabled_backends_count;
- u32 cur_pipe;
- u32 swizzle_pipe[R6XX_MAX_PIPES];
- u32 cur_backend;
- u32 i;
-
- if (num_tile_pipes > R6XX_MAX_PIPES)
- num_tile_pipes = R6XX_MAX_PIPES;
- if (num_tile_pipes < 1)
- num_tile_pipes = 1;
- if (num_backends > R6XX_MAX_BACKENDS)
- num_backends = R6XX_MAX_BACKENDS;
- if (num_backends < 1)
- num_backends = 1;
-
- enabled_backends_mask = 0;
- enabled_backends_count = 0;
- for (i = 0; i < R6XX_MAX_BACKENDS; ++i) {
- if (((backend_disable_mask >> i) & 1) == 0) {
- enabled_backends_mask |= (1 << i);
- ++enabled_backends_count;
- }
- if (enabled_backends_count == num_backends)
- break;
- }
-
- if (enabled_backends_count == 0) {
- enabled_backends_mask = 1;
- enabled_backends_count = 1;
- }
-
- if (enabled_backends_count != num_backends)
- num_backends = enabled_backends_count;
-
- memset((uint8_t *)&swizzle_pipe[0], 0, sizeof(u32) * R6XX_MAX_PIPES);
- switch (num_tile_pipes) {
- case 1:
- swizzle_pipe[0] = 0;
- break;
- case 2:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- break;
- case 3:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- break;
- case 4:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- break;
- case 5:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- break;
- case 6:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 5;
- swizzle_pipe[4] = 1;
- swizzle_pipe[5] = 3;
- break;
- case 7:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 6;
- swizzle_pipe[4] = 1;
- swizzle_pipe[5] = 3;
- swizzle_pipe[6] = 5;
- break;
- case 8:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 6;
- swizzle_pipe[4] = 1;
- swizzle_pipe[5] = 3;
- swizzle_pipe[6] = 5;
- swizzle_pipe[7] = 7;
- break;
+u32 r6xx_remap_render_backend(struct radeon_device *rdev,
+ u32 tiling_pipe_num,
+ u32 max_rb_num,
+ u32 total_max_rb_num,
+ u32 disabled_rb_mask)
+{
+ u32 rendering_pipe_num, rb_num_width, req_rb_num;
+ u32 pipe_rb_ratio, pipe_rb_remain;
+ u32 data = 0, mask = 1 << (max_rb_num - 1);
+ unsigned i, j;
+
+ /* mask out the RBs that don't exist on that asic */
+ disabled_rb_mask |= (0xff << max_rb_num) & 0xff;
+
+ rendering_pipe_num = 1 << tiling_pipe_num;
+ req_rb_num = total_max_rb_num - r600_count_pipe_bits(disabled_rb_mask);
+ BUG_ON(rendering_pipe_num < req_rb_num);
+
+ pipe_rb_ratio = rendering_pipe_num / req_rb_num;
+ pipe_rb_remain = rendering_pipe_num - pipe_rb_ratio * req_rb_num;
+
+ if (rdev->family <= CHIP_RV740) {
+ /* r6xx/r7xx */
+ rb_num_width = 2;
+ } else {
+ /* eg+ */
+ rb_num_width = 4;
}
- cur_backend = 0;
- for (cur_pipe = 0; cur_pipe < num_tile_pipes; ++cur_pipe) {
- while (((1 << cur_backend) & enabled_backends_mask) == 0)
- cur_backend = (cur_backend + 1) % R6XX_MAX_BACKENDS;
-
- backend_map |= (u32)(((cur_backend & 3) << (swizzle_pipe[cur_pipe] * 2)));
-
- cur_backend = (cur_backend + 1) % R6XX_MAX_BACKENDS;
+ for (i = 0; i < max_rb_num; i++) {
+ if (!(mask & disabled_rb_mask)) {
+ for (j = 0; j < pipe_rb_ratio; j++) {
+ data <<= rb_num_width;
+ data |= max_rb_num - i - 1;
+ }
+ if (pipe_rb_remain) {
+ data <<= rb_num_width;
+ data |= max_rb_num - i - 1;
+ pipe_rb_remain--;
+ }
+ }
+ mask >>= 1;
}
- return backend_map;
+ return data;
}
int r600_count_pipe_bits(uint32_t val)
@@ -1500,7 +1438,6 @@ void r600_gpu_init(struct radeon_device *rdev)
{
u32 tiling_config;
u32 ramcfg;
- u32 backend_map;
u32 cc_rb_backend_disable;
u32 cc_gc_shader_pipe_config;
u32 tmp;
@@ -1511,8 +1448,9 @@ void r600_gpu_init(struct radeon_device *rdev)
u32 sq_thread_resource_mgmt = 0;
u32 sq_stack_resource_mgmt_1 = 0;
u32 sq_stack_resource_mgmt_2 = 0;
+ u32 disabled_rb_mask;
- /* FIXME: implement */
+ rdev->config.r600.tiling_group_size = 256;
switch (rdev->family) {
case CHIP_R600:
rdev->config.r600.max_pipes = 4;
@@ -1616,10 +1554,7 @@ void r600_gpu_init(struct radeon_device *rdev)
rdev->config.r600.tiling_nbanks = 4 << ((ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT);
tiling_config |= BANK_TILING((ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT);
tiling_config |= GROUP_SIZE((ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT);
- if ((ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT)
- rdev->config.r600.tiling_group_size = 512;
- else
- rdev->config.r600.tiling_group_size = 256;
+
tmp = (ramcfg & NOOFROWS_MASK) >> NOOFROWS_SHIFT;
if (tmp > 3) {
tiling_config |= ROW_TILING(3);
@@ -1631,32 +1566,36 @@ void r600_gpu_init(struct radeon_device *rdev)
tiling_config |= BANK_SWAPS(1);
cc_rb_backend_disable = RREG32(CC_RB_BACKEND_DISABLE) & 0x00ff0000;
- cc_rb_backend_disable |=
- BACKEND_DISABLE((R6XX_MAX_BACKENDS_MASK << rdev->config.r600.max_backends) & R6XX_MAX_BACKENDS_MASK);
-
- cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG) & 0xffffff00;
- cc_gc_shader_pipe_config |=
- INACTIVE_QD_PIPES((R6XX_MAX_PIPES_MASK << rdev->config.r600.max_pipes) & R6XX_MAX_PIPES_MASK);
- cc_gc_shader_pipe_config |=
- INACTIVE_SIMDS((R6XX_MAX_SIMDS_MASK << rdev->config.r600.max_simds) & R6XX_MAX_SIMDS_MASK);
-
- backend_map = r600_get_tile_pipe_to_backend_map(rdev->config.r600.max_tile_pipes,
- (R6XX_MAX_BACKENDS -
- r600_count_pipe_bits((cc_rb_backend_disable &
- R6XX_MAX_BACKENDS_MASK) >> 16)),
- (cc_rb_backend_disable >> 16));
+ tmp = R6XX_MAX_BACKENDS -
+ r600_count_pipe_bits((cc_rb_backend_disable >> 16) & R6XX_MAX_BACKENDS_MASK);
+ if (tmp < rdev->config.r600.max_backends) {
+ rdev->config.r600.max_backends = tmp;
+ }
+
+ cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG) & 0x00ffff00;
+ tmp = R6XX_MAX_PIPES -
+ r600_count_pipe_bits((cc_gc_shader_pipe_config >> 8) & R6XX_MAX_PIPES_MASK);
+ if (tmp < rdev->config.r600.max_pipes) {
+ rdev->config.r600.max_pipes = tmp;
+ }
+ tmp = R6XX_MAX_SIMDS -
+ r600_count_pipe_bits((cc_gc_shader_pipe_config >> 16) & R6XX_MAX_SIMDS_MASK);
+ if (tmp < rdev->config.r600.max_simds) {
+ rdev->config.r600.max_simds = tmp;
+ }
+
+ disabled_rb_mask = (RREG32(CC_RB_BACKEND_DISABLE) >> 16) & R6XX_MAX_BACKENDS_MASK;
+ tmp = (tiling_config & PIPE_TILING__MASK) >> PIPE_TILING__SHIFT;
+ tmp = r6xx_remap_render_backend(rdev, tmp, rdev->config.r600.max_backends,
+ R6XX_MAX_BACKENDS, disabled_rb_mask);
+ tiling_config |= tmp << 16;
+ rdev->config.r600.backend_map = tmp;
+
rdev->config.r600.tile_config = tiling_config;
- rdev->config.r600.backend_map = backend_map;
- tiling_config |= BACKEND_MAP(backend_map);
WREG32(GB_TILING_CONFIG, tiling_config);
WREG32(DCP_TILING_CONFIG, tiling_config & 0xffff);
WREG32(HDP_TILING_CONFIG, tiling_config & 0xffff);
- /* Setup pipes */
- WREG32(CC_RB_BACKEND_DISABLE, cc_rb_backend_disable);
- WREG32(CC_GC_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config);
- WREG32(GC_USER_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config);
-
tmp = R6XX_MAX_PIPES - r600_count_pipe_bits((cc_gc_shader_pipe_config & INACTIVE_QD_PIPES_MASK) >> 8);
WREG32(VGT_OUT_DEALLOC_CNTL, (tmp * 4) & DEALLOC_DIST_MASK);
WREG32(VGT_VERTEX_REUSE_BLOCK_CNTL, ((tmp * 4) - 2) & VTX_REUSE_DEPTH_MASK);
diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h
index 15bd3b21624..a0dbf1fe6a4 100644
--- a/drivers/gpu/drm/radeon/r600d.h
+++ b/drivers/gpu/drm/radeon/r600d.h
@@ -219,6 +219,8 @@
#define BACKEND_MAP(x) ((x) << 16)
#define GB_TILING_CONFIG 0x98F0
+#define PIPE_TILING__SHIFT 1
+#define PIPE_TILING__MASK 0x0000000e
#define GC_USER_SHADER_PIPE_CONFIG 0x8954
#define INACTIVE_QD_PIPES(x) ((x) << 8)
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 492654f8ee7..85dac33e3cc 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -346,6 +346,9 @@ struct radeon_bo {
/* Constant after initialization */
struct radeon_device *rdev;
struct drm_gem_object gem_base;
+
+ struct ttm_bo_kmap_obj dma_buf_vmap;
+ int vmapping_count;
};
#define gem_to_radeon_bo(gobj) container_of((gobj), struct radeon_bo, gem_base)
@@ -1845,6 +1848,11 @@ extern struct radeon_hdmi_acr r600_hdmi_acr(uint32_t clock);
extern void r600_hdmi_enable(struct drm_encoder *encoder);
extern void r600_hdmi_disable(struct drm_encoder *encoder);
extern void r600_hdmi_setmode(struct drm_encoder *encoder, struct drm_display_mode *mode);
+extern u32 r6xx_remap_render_backend(struct radeon_device *rdev,
+ u32 tiling_pipe_num,
+ u32 max_rb_num,
+ u32 total_max_rb_num,
+ u32 enabled_rb_mask);
/*
* evergreen functions used by radeon_encoder.c
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index 0137689ed46..142f89462aa 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -147,6 +147,7 @@ static int radeon_cs_sync_rings(struct radeon_cs_parser *p)
sync_to_ring, p->ring);
}
+/* XXX: note that this is called from the legacy UMS CS ioctl as well */
int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
{
struct drm_radeon_cs *cs = data;
@@ -245,22 +246,24 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
}
}
- if ((p->cs_flags & RADEON_CS_USE_VM) &&
- !p->rdev->vm_manager.enabled) {
- DRM_ERROR("VM not active on asic!\n");
- return -EINVAL;
- }
-
- /* we only support VM on SI+ */
- if ((p->rdev->family >= CHIP_TAHITI) &&
- ((p->cs_flags & RADEON_CS_USE_VM) == 0)) {
- DRM_ERROR("VM required on SI+!\n");
- return -EINVAL;
- }
+ /* these are KMS only */
+ if (p->rdev) {
+ if ((p->cs_flags & RADEON_CS_USE_VM) &&
+ !p->rdev->vm_manager.enabled) {
+ DRM_ERROR("VM not active on asic!\n");
+ return -EINVAL;
+ }
- if (radeon_cs_get_ring(p, ring, priority))
- return -EINVAL;
+ /* we only support VM on SI+ */
+ if ((p->rdev->family >= CHIP_TAHITI) &&
+ ((p->cs_flags & RADEON_CS_USE_VM) == 0)) {
+ DRM_ERROR("VM required on SI+!\n");
+ return -EINVAL;
+ }
+ if (radeon_cs_get_ring(p, ring, priority))
+ return -EINVAL;
+ }
/* deal with non-vm */
if ((p->chunk_ib_idx != -1) &&
diff --git a/drivers/gpu/drm/radeon/radeon_prime.c b/drivers/gpu/drm/radeon/radeon_prime.c
index b8f835d8ecb..8ddab4c7671 100644
--- a/drivers/gpu/drm/radeon/radeon_prime.c
+++ b/drivers/gpu/drm/radeon/radeon_prime.c
@@ -85,6 +85,47 @@ static void radeon_gem_kunmap(struct dma_buf *dma_buf, unsigned long page_num, v
}
+static int radeon_gem_prime_mmap(struct dma_buf *dma_buf, struct vm_area_struct *vma)
+{
+ return -EINVAL;
+}
+
+static void *radeon_gem_prime_vmap(struct dma_buf *dma_buf)
+{
+ struct radeon_bo *bo = dma_buf->priv;
+ struct drm_device *dev = bo->rdev->ddev;
+ int ret;
+
+ mutex_lock(&dev->struct_mutex);
+ if (bo->vmapping_count) {
+ bo->vmapping_count++;
+ goto out_unlock;
+ }
+
+ ret = ttm_bo_kmap(&bo->tbo, 0, bo->tbo.num_pages,
+ &bo->dma_buf_vmap);
+ if (ret) {
+ mutex_unlock(&dev->struct_mutex);
+ return ERR_PTR(ret);
+ }
+ bo->vmapping_count = 1;
+out_unlock:
+ mutex_unlock(&dev->struct_mutex);
+ return bo->dma_buf_vmap.virtual;
+}
+
+static void radeon_gem_prime_vunmap(struct dma_buf *dma_buf, void *vaddr)
+{
+ struct radeon_bo *bo = dma_buf->priv;
+ struct drm_device *dev = bo->rdev->ddev;
+
+ mutex_lock(&dev->struct_mutex);
+ bo->vmapping_count--;
+ if (bo->vmapping_count == 0) {
+ ttm_bo_kunmap(&bo->dma_buf_vmap);
+ }
+ mutex_unlock(&dev->struct_mutex);
+}
const static struct dma_buf_ops radeon_dmabuf_ops = {
.map_dma_buf = radeon_gem_map_dma_buf,
.unmap_dma_buf = radeon_gem_unmap_dma_buf,
@@ -93,6 +134,9 @@ const static struct dma_buf_ops radeon_dmabuf_ops = {
.kmap_atomic = radeon_gem_kmap_atomic,
.kunmap = radeon_gem_kunmap,
.kunmap_atomic = radeon_gem_kunmap_atomic,
+ .mmap = radeon_gem_prime_mmap,
+ .vmap = radeon_gem_prime_vmap,
+ .vunmap = radeon_gem_prime_vunmap,
};
static int radeon_prime_create(struct drm_device *dev,
diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c
index c2f473bc13b..04ddc365a90 100644
--- a/drivers/gpu/drm/radeon/rv770.c
+++ b/drivers/gpu/drm/radeon/rv770.c
@@ -151,6 +151,8 @@ int rv770_pcie_gart_enable(struct radeon_device *rdev)
WREG32(MC_VM_MD_L1_TLB0_CNTL, tmp);
WREG32(MC_VM_MD_L1_TLB1_CNTL, tmp);
WREG32(MC_VM_MD_L1_TLB2_CNTL, tmp);
+ if (rdev->family == CHIP_RV740)
+ WREG32(MC_VM_MD_L1_TLB3_CNTL, tmp);
WREG32(MC_VM_MB_L1_TLB0_CNTL, tmp);
WREG32(MC_VM_MB_L1_TLB1_CNTL, tmp);
WREG32(MC_VM_MB_L1_TLB2_CNTL, tmp);
@@ -363,180 +365,6 @@ void r700_cp_fini(struct radeon_device *rdev)
/*
* Core functions
*/
-static u32 r700_get_tile_pipe_to_backend_map(struct radeon_device *rdev,
- u32 num_tile_pipes,
- u32 num_backends,
- u32 backend_disable_mask)
-{
- u32 backend_map = 0;
- u32 enabled_backends_mask;
- u32 enabled_backends_count;
- u32 cur_pipe;
- u32 swizzle_pipe[R7XX_MAX_PIPES];
- u32 cur_backend;
- u32 i;
- bool force_no_swizzle;
-
- if (num_tile_pipes > R7XX_MAX_PIPES)
- num_tile_pipes = R7XX_MAX_PIPES;
- if (num_tile_pipes < 1)
- num_tile_pipes = 1;
- if (num_backends > R7XX_MAX_BACKENDS)
- num_backends = R7XX_MAX_BACKENDS;
- if (num_backends < 1)
- num_backends = 1;
-
- enabled_backends_mask = 0;
- enabled_backends_count = 0;
- for (i = 0; i < R7XX_MAX_BACKENDS; ++i) {
- if (((backend_disable_mask >> i) & 1) == 0) {
- enabled_backends_mask |= (1 << i);
- ++enabled_backends_count;
- }
- if (enabled_backends_count == num_backends)
- break;
- }
-
- if (enabled_backends_count == 0) {
- enabled_backends_mask = 1;
- enabled_backends_count = 1;
- }
-
- if (enabled_backends_count != num_backends)
- num_backends = enabled_backends_count;
-
- switch (rdev->family) {
- case CHIP_RV770:
- case CHIP_RV730:
- force_no_swizzle = false;
- break;
- case CHIP_RV710:
- case CHIP_RV740:
- default:
- force_no_swizzle = true;
- break;
- }
-
- memset((uint8_t *)&swizzle_pipe[0], 0, sizeof(u32) * R7XX_MAX_PIPES);
- switch (num_tile_pipes) {
- case 1:
- swizzle_pipe[0] = 0;
- break;
- case 2:
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- break;
- case 3:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 1;
- }
- break;
- case 4:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 3;
- swizzle_pipe[3] = 1;
- }
- break;
- case 5:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 1;
- swizzle_pipe[4] = 3;
- }
- break;
- case 6:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- swizzle_pipe[5] = 5;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 5;
- swizzle_pipe[4] = 3;
- swizzle_pipe[5] = 1;
- }
- break;
- case 7:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- swizzle_pipe[5] = 5;
- swizzle_pipe[6] = 6;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 6;
- swizzle_pipe[4] = 3;
- swizzle_pipe[5] = 1;
- swizzle_pipe[6] = 5;
- }
- break;
- case 8:
- if (force_no_swizzle) {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 1;
- swizzle_pipe[2] = 2;
- swizzle_pipe[3] = 3;
- swizzle_pipe[4] = 4;
- swizzle_pipe[5] = 5;
- swizzle_pipe[6] = 6;
- swizzle_pipe[7] = 7;
- } else {
- swizzle_pipe[0] = 0;
- swizzle_pipe[1] = 2;
- swizzle_pipe[2] = 4;
- swizzle_pipe[3] = 6;
- swizzle_pipe[4] = 3;
- swizzle_pipe[5] = 1;
- swizzle_pipe[6] = 7;
- swizzle_pipe[7] = 5;
- }
- break;
- }
-
- cur_backend = 0;
- for (cur_pipe = 0; cur_pipe < num_tile_pipes; ++cur_pipe) {
- while (((1 << cur_backend) & enabled_backends_mask) == 0)
- cur_backend = (cur_backend + 1) % R7XX_MAX_BACKENDS;
-
- backend_map |= (u32)(((cur_backend & 3) << (swizzle_pipe[cur_pipe] * 2)));
-
- cur_backend = (cur_backend + 1) % R7XX_MAX_BACKENDS;
- }
-
- return backend_map;
-}
-
static void rv770_gpu_init(struct radeon_device *rdev)
{
int i, j, num_qd_pipes;
@@ -552,14 +380,17 @@ static void rv770_gpu_init(struct radeon_device *rdev)
u32 sq_thread_resource_mgmt;
u32 hdp_host_path_cntl;
u32 sq_dyn_gpr_size_simd_ab_0;
- u32 backend_map;
u32 gb_tiling_config = 0;
u32 cc_rb_backend_disable = 0;
u32 cc_gc_shader_pipe_config = 0;
u32 mc_arb_ramcfg;
- u32 db_debug4;
+ u32 db_debug4, tmp;
+ u32 inactive_pipes, shader_pipe_config;
+ u32 disabled_rb_mask;
+ unsigned active_number;
/* setup chip specs */
+ rdev->config.rv770.tiling_group_size = 256;
switch (rdev->family) {
case CHIP_RV770:
rdev->config.rv770.max_pipes = 4;
@@ -670,33 +501,70 @@ static void rv770_gpu_init(struct radeon_device *rdev)
/* setup tiling, simd, pipe config */
mc_arb_ramcfg = RREG32(MC_ARB_RAMCFG);
+ shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG);
+ inactive_pipes = (shader_pipe_config & INACTIVE_QD_PIPES_MASK) >> INACTIVE_QD_PIPES_SHIFT;
+ for (i = 0, tmp = 1, active_number = 0; i < R7XX_MAX_PIPES; i++) {
+ if (!(inactive_pipes & tmp)) {
+ active_number++;
+ }
+ tmp <<= 1;
+ }
+ if (active_number == 1) {
+ WREG32(SPI_CONFIG_CNTL, DISABLE_INTERP_1);
+ } else {
+ WREG32(SPI_CONFIG_CNTL, 0);
+ }
+
+ cc_rb_backend_disable = RREG32(CC_RB_BACKEND_DISABLE) & 0x00ff0000;
+ tmp = R7XX_MAX_BACKENDS - r600_count_pipe_bits(cc_rb_backend_disable >> 16);
+ if (tmp < rdev->config.rv770.max_backends) {
+ rdev->config.rv770.max_backends = tmp;
+ }
+
+ cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG) & 0xffffff00;
+ tmp = R7XX_MAX_PIPES - r600_count_pipe_bits((cc_gc_shader_pipe_config >> 8) & R7XX_MAX_PIPES_MASK);
+ if (tmp < rdev->config.rv770.max_pipes) {
+ rdev->config.rv770.max_pipes = tmp;
+ }
+ tmp = R7XX_MAX_SIMDS - r600_count_pipe_bits((cc_gc_shader_pipe_config >> 16) & R7XX_MAX_SIMDS_MASK);
+ if (tmp < rdev->config.rv770.max_simds) {
+ rdev->config.rv770.max_simds = tmp;
+ }
+
switch (rdev->config.rv770.max_tile_pipes) {
case 1:
default:
- gb_tiling_config |= PIPE_TILING(0);
+ gb_tiling_config = PIPE_TILING(0);
break;
case 2:
- gb_tiling_config |= PIPE_TILING(1);
+ gb_tiling_config = PIPE_TILING(1);
break;
case 4:
- gb_tiling_config |= PIPE_TILING(2);
+ gb_tiling_config = PIPE_TILING(2);
break;
case 8:
- gb_tiling_config |= PIPE_TILING(3);
+ gb_tiling_config = PIPE_TILING(3);
break;
}
rdev->config.rv770.tiling_npipes = rdev->config.rv770.max_tile_pipes;
+ disabled_rb_mask = (RREG32(CC_RB_BACKEND_DISABLE) >> 16) & R7XX_MAX_BACKENDS_MASK;
+ tmp = (gb_tiling_config & PIPE_TILING__MASK) >> PIPE_TILING__SHIFT;
+ tmp = r6xx_remap_render_backend(rdev, tmp, rdev->config.rv770.max_backends,
+ R7XX_MAX_BACKENDS, disabled_rb_mask);
+ gb_tiling_config |= tmp << 16;
+ rdev->config.rv770.backend_map = tmp;
+
if (rdev->family == CHIP_RV770)
gb_tiling_config |= BANK_TILING(1);
- else
- gb_tiling_config |= BANK_TILING((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT);
+ else {
+ if ((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT)
+ gb_tiling_config |= BANK_TILING(1);
+ else
+ gb_tiling_config |= BANK_TILING(0);
+ }
rdev->config.rv770.tiling_nbanks = 4 << ((gb_tiling_config >> 4) & 0x3);
gb_tiling_config |= GROUP_SIZE((mc_arb_ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT);
- if ((mc_arb_ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT)
- rdev->config.rv770.tiling_group_size = 512;
- else
- rdev->config.rv770.tiling_group_size = 256;
if (((mc_arb_ramcfg & NOOFROWS_MASK) >> NOOFROWS_SHIFT) > 3) {
gb_tiling_config |= ROW_TILING(3);
gb_tiling_config |= SAMPLE_SPLIT(3);
@@ -708,47 +576,19 @@ static void rv770_gpu_init(struct radeon_device *rdev)
}
gb_tiling_config |= BANK_SWAPS(1);
-
- cc_rb_backend_disable = RREG32(CC_RB_BACKEND_DISABLE) & 0x00ff0000;
- cc_rb_backend_disable |=
- BACKEND_DISABLE((R7XX_MAX_BACKENDS_MASK << rdev->config.rv770.max_backends) & R7XX_MAX_BACKENDS_MASK);
-
- cc_gc_shader_pipe_config = RREG32(CC_GC_SHADER_PIPE_CONFIG) & 0xffffff00;
- cc_gc_shader_pipe_config |=
- INACTIVE_QD_PIPES((R7XX_MAX_PIPES_MASK << rdev->config.rv770.max_pipes) & R7XX_MAX_PIPES_MASK);
- cc_gc_shader_pipe_config |=
- INACTIVE_SIMDS((R7XX_MAX_SIMDS_MASK << rdev->config.rv770.max_simds) & R7XX_MAX_SIMDS_MASK);
-
- if (rdev->family == CHIP_RV740)
- backend_map = 0x28;
- else
- backend_map = r700_get_tile_pipe_to_backend_map(rdev,
- rdev->config.rv770.max_tile_pipes,
- (R7XX_MAX_BACKENDS -
- r600_count_pipe_bits((cc_rb_backend_disable &
- R7XX_MAX_BACKENDS_MASK) >> 16)),
- (cc_rb_backend_disable >> 16));
-
rdev->config.rv770.tile_config = gb_tiling_config;
- rdev->config.rv770.backend_map = backend_map;
- gb_tiling_config |= BACKEND_MAP(backend_map);
WREG32(GB_TILING_CONFIG, gb_tiling_config);
WREG32(DCP_TILING_CONFIG, (gb_tiling_config & 0xffff));
WREG32(HDP_TILING_CONFIG, (gb_tiling_config & 0xffff));
- WREG32(CC_RB_BACKEND_DISABLE, cc_rb_backend_disable);
- WREG32(CC_GC_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config);
- WREG32(GC_USER_SHADER_PIPE_CONFIG, cc_gc_shader_pipe_config);
- WREG32(CC_SYS_RB_BACKEND_DISABLE, cc_rb_backend_disable);
-
WREG32(CGTS_SYS_TCC_DISABLE, 0);
WREG32(CGTS_TCC_DISABLE, 0);
WREG32(CGTS_USER_SYS_TCC_DISABLE, 0);
WREG32(CGTS_USER_TCC_DISABLE, 0);
- num_qd_pipes =
- R7XX_MAX_PIPES - r600_count_pipe_bits((cc_gc_shader_pipe_config & INACTIVE_QD_PIPES_MASK) >> 8);
+
+ num_qd_pipes = R7XX_MAX_PIPES - r600_count_pipe_bits((cc_gc_shader_pipe_config & INACTIVE_QD_PIPES_MASK) >> 8);
WREG32(VGT_OUT_DEALLOC_CNTL, (num_qd_pipes * 4) & DEALLOC_DIST_MASK);
WREG32(VGT_VERTEX_REUSE_BLOCK_CNTL, ((num_qd_pipes * 4) - 2) & VTX_REUSE_DEPTH_MASK);
@@ -809,8 +649,6 @@ static void rv770_gpu_init(struct radeon_device *rdev)
WREG32(VGT_NUM_INSTANCES, 1);
- WREG32(SPI_CONFIG_CNTL, GPR_WRITE_PRIORITY(0));
-
WREG32(SPI_CONFIG_CNTL_1, VTX_DONE_DELAY(4));
WREG32(CP_PERFMON_CNTL, 0);
diff --git a/drivers/gpu/drm/radeon/rv770d.h b/drivers/gpu/drm/radeon/rv770d.h
index 9c549f702f2..fdc08989601 100644
--- a/drivers/gpu/drm/radeon/rv770d.h
+++ b/drivers/gpu/drm/radeon/rv770d.h
@@ -106,10 +106,13 @@
#define BACKEND_MAP(x) ((x) << 16)
#define GB_TILING_CONFIG 0x98F0
+#define PIPE_TILING__SHIFT 1
+#define PIPE_TILING__MASK 0x0000000e
#define GC_USER_SHADER_PIPE_CONFIG 0x8954
#define INACTIVE_QD_PIPES(x) ((x) << 8)
#define INACTIVE_QD_PIPES_MASK 0x0000FF00
+#define INACTIVE_QD_PIPES_SHIFT 8
#define INACTIVE_SIMDS(x) ((x) << 16)
#define INACTIVE_SIMDS_MASK 0x00FF0000
@@ -174,6 +177,7 @@
#define MC_VM_MD_L1_TLB0_CNTL 0x2654
#define MC_VM_MD_L1_TLB1_CNTL 0x2658
#define MC_VM_MD_L1_TLB2_CNTL 0x265C
+#define MC_VM_MD_L1_TLB3_CNTL 0x2698
#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR 0x203C
#define MC_VM_SYSTEM_APERTURE_HIGH_ADDR 0x2038
#define MC_VM_SYSTEM_APERTURE_LOW_ADDR 0x2034
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 36792bd4da7..b67cfcaa661 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -1834,6 +1834,7 @@ static int ttm_bo_swapout(struct ttm_mem_shrink *shrink)
spin_unlock(&glob->lru_lock);
(void) ttm_bo_cleanup_refs(bo, false, false, false);
kref_put(&bo->list_kref, ttm_bo_release_list);
+ spin_lock(&glob->lru_lock);
continue;
}
diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c
index a029ee39b0c..ce9a6117992 100644
--- a/drivers/gpu/drm/udl/udl_fb.c
+++ b/drivers/gpu/drm/udl/udl_fb.c
@@ -156,8 +156,17 @@ int udl_handle_damage(struct udl_framebuffer *fb, int x, int y,
if (!fb->active_16)
return 0;
- if (!fb->obj->vmapping)
- udl_gem_vmap(fb->obj);
+ if (!fb->obj->vmapping) {
+ ret = udl_gem_vmap(fb->obj);
+ if (ret == -ENOMEM) {
+ DRM_ERROR("failed to vmap fb\n");
+ return 0;
+ }
+ if (!fb->obj->vmapping) {
+ DRM_ERROR("failed to vmapping\n");
+ return 0;
+ }
+ }
start_cycles = get_cycles();
diff --git a/drivers/gpu/drm/udl/udl_gem.c b/drivers/gpu/drm/udl/udl_gem.c
index 97acc9c6c95..7bd65bdd15a 100644
--- a/drivers/gpu/drm/udl/udl_gem.c
+++ b/drivers/gpu/drm/udl/udl_gem.c
@@ -180,6 +180,18 @@ int udl_gem_vmap(struct udl_gem_object *obj)
int page_count = obj->base.size / PAGE_SIZE;
int ret;
+ if (obj->base.import_attach) {
+ ret = dma_buf_begin_cpu_access(obj->base.import_attach->dmabuf,
+ 0, obj->base.size, DMA_BIDIRECTIONAL);
+ if (ret)
+ return -EINVAL;
+
+ obj->vmapping = dma_buf_vmap(obj->base.import_attach->dmabuf);
+ if (!obj->vmapping)
+ return -ENOMEM;
+ return 0;
+ }
+
ret = udl_gem_get_pages(obj, GFP_KERNEL);
if (ret)
return ret;
@@ -192,6 +204,13 @@ int udl_gem_vmap(struct udl_gem_object *obj)
void udl_gem_vunmap(struct udl_gem_object *obj)
{
+ if (obj->base.import_attach) {
+ dma_buf_vunmap(obj->base.import_attach->dmabuf, obj->vmapping);
+ dma_buf_end_cpu_access(obj->base.import_attach->dmabuf, 0,
+ obj->base.size, DMA_BIDIRECTIONAL);
+ return;
+ }
+
if (obj->vmapping)
vunmap(obj->vmapping);
@@ -202,12 +221,12 @@ void udl_gem_free_object(struct drm_gem_object *gem_obj)
{
struct udl_gem_object *obj = to_udl_bo(gem_obj);
- if (gem_obj->import_attach)
- drm_prime_gem_destroy(gem_obj, obj->sg);
-
if (obj->vmapping)
udl_gem_vunmap(obj);
+ if (gem_obj->import_attach)
+ drm_prime_gem_destroy(gem_obj, obj->sg);
+
if (obj->pages)
udl_gem_put_pages(obj);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
index 51c9ba5cd2f..21ee7822656 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
@@ -66,7 +66,7 @@ static int vmw_gmr2_bind(struct vmw_private *dev_priv,
cmd += sizeof(remap_cmd) / sizeof(uint32);
for (i = 0; i < num_pages; ++i) {
- if (VMW_PPN_SIZE > 4)
+ if (VMW_PPN_SIZE <= 4)
*cmd = page_to_pfn(*pages++);
else
*((uint64_t *)cmd) = page_to_pfn(*pages++);
diff --git a/drivers/message/fusion/mptbase.c b/drivers/message/fusion/mptbase.c
index 444143e5f28..d99db5623ac 100644
--- a/drivers/message/fusion/mptbase.c
+++ b/drivers/message/fusion/mptbase.c
@@ -1653,7 +1653,6 @@ mpt_mapresources(MPT_ADAPTER *ioc)
unsigned long port;
u32 msize;
u32 psize;
- u8 revision;
int r = -ENODEV;
struct pci_dev *pdev;
@@ -1670,8 +1669,6 @@ mpt_mapresources(MPT_ADAPTER *ioc)
return r;
}
- pci_read_config_byte(pdev, PCI_CLASS_REVISION, &revision);
-
if (sizeof(dma_addr_t) > 4) {
const uint64_t required_mask = dma_get_required_mask
(&pdev->dev);
@@ -1779,7 +1776,6 @@ mpt_attach(struct pci_dev *pdev, const struct pci_device_id *id)
MPT_ADAPTER *ioc;
u8 cb_idx;
int r = -ENODEV;
- u8 revision;
u8 pcixcmd;
static int mpt_ids = 0;
#ifdef CONFIG_PROC_FS
@@ -1887,8 +1883,8 @@ mpt_attach(struct pci_dev *pdev, const struct pci_device_id *id)
dinitprintk(ioc, printk(MYIOC_s_INFO_FMT "facts @ %p, pfacts[0] @ %p\n",
ioc->name, &ioc->facts, &ioc->pfacts[0]));
- pci_read_config_byte(pdev, PCI_CLASS_REVISION, &revision);
- mpt_get_product_name(pdev->vendor, pdev->device, revision, ioc->prod_name);
+ mpt_get_product_name(pdev->vendor, pdev->device, pdev->revision,
+ ioc->prod_name);
switch (pdev->device)
{
@@ -1903,7 +1899,7 @@ mpt_attach(struct pci_dev *pdev, const struct pci_device_id *id)
break;
case MPI_MANUFACTPAGE_DEVICEID_FC929X:
- if (revision < XL_929) {
+ if (pdev->revision < XL_929) {
/* 929X Chip Fix. Set Split transactions level
* for PCIX. Set MOST bits to zero.
*/
@@ -1934,7 +1930,7 @@ mpt_attach(struct pci_dev *pdev, const struct pci_device_id *id)
/* 1030 Chip Fix. Disable Split transactions
* for PCIX. Set MOST bits to zero if Rev < C0( = 8).
*/
- if (revision < C0_1030) {
+ if (pdev->revision < C0_1030) {
pci_read_config_byte(pdev, 0x6a, &pcixcmd);
pcixcmd &= 0x8F;
pci_write_config_byte(pdev, 0x6a, pcixcmd);
diff --git a/drivers/message/fusion/mptctl.c b/drivers/message/fusion/mptctl.c
index 6e6e16aab9d..b383b6961e5 100644
--- a/drivers/message/fusion/mptctl.c
+++ b/drivers/message/fusion/mptctl.c
@@ -1250,7 +1250,6 @@ mptctl_getiocinfo (unsigned long arg, unsigned int data_size)
int iocnum;
unsigned int port;
int cim_rev;
- u8 revision;
struct scsi_device *sdev;
VirtDevice *vdevice;
@@ -1324,8 +1323,7 @@ mptctl_getiocinfo (unsigned long arg, unsigned int data_size)
pdev = (struct pci_dev *) ioc->pcidev;
karg->pciId = pdev->device;
- pci_read_config_byte(pdev, PCI_CLASS_REVISION, &revision);
- karg->hwRev = revision;
+ karg->hwRev = pdev->revision;
karg->subSystemDevice = pdev->subsystem_device;
karg->subSystemVendor = pdev->subsystem_vendor;
diff --git a/drivers/rapidio/Kconfig b/drivers/rapidio/Kconfig
index bc871923879..6194d35ebb9 100644
--- a/drivers/rapidio/Kconfig
+++ b/drivers/rapidio/Kconfig
@@ -22,6 +22,20 @@ config RAPIDIO_ENABLE_RX_TX_PORTS
ports for Input/Output direction to allow other traffic
than Maintenance transfers.
+config RAPIDIO_DMA_ENGINE
+ bool "DMA Engine support for RapidIO"
+ depends on RAPIDIO
+ select DMADEVICES
+ select DMA_ENGINE
+ help
+ Say Y here if you want to use DMA Engine frameork for RapidIO data
+ transfers to/from target RIO devices. RapidIO uses NREAD and
+ NWRITE (NWRITE_R, SWRITE) requests to transfer data between local
+ memory and memory on remote target device. You need a DMA controller
+ capable to perform data transfers to/from RapidIO.
+
+ If you are unsure about this, say Y here.
+
config RAPIDIO_DEBUG
bool "RapidIO subsystem debug messages"
depends on RAPIDIO
diff --git a/drivers/rapidio/devices/Makefile b/drivers/rapidio/devices/Makefile
index 3b7b4e2dff7..7b62860f34f 100644
--- a/drivers/rapidio/devices/Makefile
+++ b/drivers/rapidio/devices/Makefile
@@ -3,3 +3,6 @@
#
obj-$(CONFIG_RAPIDIO_TSI721) += tsi721.o
+ifeq ($(CONFIG_RAPIDIO_DMA_ENGINE),y)
+obj-$(CONFIG_RAPIDIO_TSI721) += tsi721_dma.o
+endif
diff --git a/drivers/rapidio/devices/tsi721.c b/drivers/rapidio/devices/tsi721.c
index 30d2072f480..722246cf20a 100644
--- a/drivers/rapidio/devices/tsi721.c
+++ b/drivers/rapidio/devices/tsi721.c
@@ -108,6 +108,7 @@ static int tsi721_maint_dma(struct tsi721_device *priv, u32 sys_size,
u16 destid, u8 hopcount, u32 offset, int len,
u32 *data, int do_wr)
{
+ void __iomem *regs = priv->regs + TSI721_DMAC_BASE(priv->mdma.ch_id);
struct tsi721_dma_desc *bd_ptr;
u32 rd_count, swr_ptr, ch_stat;
int i, err = 0;
@@ -116,10 +117,9 @@ static int tsi721_maint_dma(struct tsi721_device *priv, u32 sys_size,
if (offset > (RIO_MAINT_SPACE_SZ - len) || (len != sizeof(u32)))
return -EINVAL;
- bd_ptr = priv->bdma[TSI721_DMACH_MAINT].bd_base;
+ bd_ptr = priv->mdma.bd_base;
- rd_count = ioread32(
- priv->regs + TSI721_DMAC_DRDCNT(TSI721_DMACH_MAINT));
+ rd_count = ioread32(regs + TSI721_DMAC_DRDCNT);
/* Initialize DMA descriptor */
bd_ptr[0].type_id = cpu_to_le32((DTYPE2 << 29) | (op << 19) | destid);
@@ -134,19 +134,18 @@ static int tsi721_maint_dma(struct tsi721_device *priv, u32 sys_size,
mb();
/* Start DMA operation */
- iowrite32(rd_count + 2,
- priv->regs + TSI721_DMAC_DWRCNT(TSI721_DMACH_MAINT));
- ioread32(priv->regs + TSI721_DMAC_DWRCNT(TSI721_DMACH_MAINT));
+ iowrite32(rd_count + 2, regs + TSI721_DMAC_DWRCNT);
+ ioread32(regs + TSI721_DMAC_DWRCNT);
i = 0;
/* Wait until DMA transfer is finished */
- while ((ch_stat = ioread32(priv->regs +
- TSI721_DMAC_STS(TSI721_DMACH_MAINT))) & TSI721_DMAC_STS_RUN) {
+ while ((ch_stat = ioread32(regs + TSI721_DMAC_STS))
+ & TSI721_DMAC_STS_RUN) {
udelay(1);
if (++i >= 5000000) {
dev_dbg(&priv->pdev->dev,
"%s : DMA[%d] read timeout ch_status=%x\n",
- __func__, TSI721_DMACH_MAINT, ch_stat);
+ __func__, priv->mdma.ch_id, ch_stat);
if (!do_wr)
*data = 0xffffffff;
err = -EIO;
@@ -162,13 +161,10 @@ static int tsi721_maint_dma(struct tsi721_device *priv, u32 sys_size,
__func__, ch_stat);
dev_dbg(&priv->pdev->dev, "OP=%d : destid=%x hc=%x off=%x\n",
do_wr ? MAINT_WR : MAINT_RD, destid, hopcount, offset);
- iowrite32(TSI721_DMAC_INT_ALL,
- priv->regs + TSI721_DMAC_INT(TSI721_DMACH_MAINT));
- iowrite32(TSI721_DMAC_CTL_INIT,
- priv->regs + TSI721_DMAC_CTL(TSI721_DMACH_MAINT));
+ iowrite32(TSI721_DMAC_INT_ALL, regs + TSI721_DMAC_INT);
+ iowrite32(TSI721_DMAC_CTL_INIT, regs + TSI721_DMAC_CTL);
udelay(10);
- iowrite32(0, priv->regs +
- TSI721_DMAC_DWRCNT(TSI721_DMACH_MAINT));
+ iowrite32(0, regs + TSI721_DMAC_DWRCNT);
udelay(1);
if (!do_wr)
*data = 0xffffffff;
@@ -184,8 +180,8 @@ static int tsi721_maint_dma(struct tsi721_device *priv, u32 sys_size,
* NOTE: Skipping check and clear FIFO entries because we are waiting
* for transfer to be completed.
*/
- swr_ptr = ioread32(priv->regs + TSI721_DMAC_DSWP(TSI721_DMACH_MAINT));
- iowrite32(swr_ptr, priv->regs + TSI721_DMAC_DSRP(TSI721_DMACH_MAINT));
+ swr_ptr = ioread32(regs + TSI721_DMAC_DSWP);
+ iowrite32(swr_ptr, regs + TSI721_DMAC_DSRP);
err_out:
return err;
@@ -541,6 +537,22 @@ static irqreturn_t tsi721_irqhandler(int irq, void *ptr)
tsi721_pw_handler(mport);
}
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+ if (dev_int & TSI721_DEV_INT_BDMA_CH) {
+ int ch;
+
+ if (dev_ch_int & TSI721_INT_BDMA_CHAN_M) {
+ dev_dbg(&priv->pdev->dev,
+ "IRQ from DMA channel 0x%08x\n", dev_ch_int);
+
+ for (ch = 0; ch < TSI721_DMA_MAXCH; ch++) {
+ if (!(dev_ch_int & TSI721_INT_BDMA_CHAN(ch)))
+ continue;
+ tsi721_bdma_handler(&priv->bdma[ch]);
+ }
+ }
+ }
+#endif
return IRQ_HANDLED;
}
@@ -553,18 +565,26 @@ static void tsi721_interrupts_init(struct tsi721_device *priv)
priv->regs + TSI721_SR_CHINT(IDB_QUEUE));
iowrite32(TSI721_SR_CHINT_IDBQRCV,
priv->regs + TSI721_SR_CHINTE(IDB_QUEUE));
- iowrite32(TSI721_INT_SR2PC_CHAN(IDB_QUEUE),
- priv->regs + TSI721_DEV_CHAN_INTE);
/* Enable SRIO MAC interrupts */
iowrite32(TSI721_RIO_EM_DEV_INT_EN_INT,
priv->regs + TSI721_RIO_EM_DEV_INT_EN);
+ /* Enable interrupts from channels in use */
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+ intr = TSI721_INT_SR2PC_CHAN(IDB_QUEUE) |
+ (TSI721_INT_BDMA_CHAN_M &
+ ~TSI721_INT_BDMA_CHAN(TSI721_DMACH_MAINT));
+#else
+ intr = TSI721_INT_SR2PC_CHAN(IDB_QUEUE);
+#endif
+ iowrite32(intr, priv->regs + TSI721_DEV_CHAN_INTE);
+
if (priv->flags & TSI721_USING_MSIX)
intr = TSI721_DEV_INT_SRIO;
else
intr = TSI721_DEV_INT_SR2PC_CH | TSI721_DEV_INT_SRIO |
- TSI721_DEV_INT_SMSG_CH;
+ TSI721_DEV_INT_SMSG_CH | TSI721_DEV_INT_BDMA_CH;
iowrite32(intr, priv->regs + TSI721_DEV_INTE);
ioread32(priv->regs + TSI721_DEV_INTE);
@@ -715,12 +735,29 @@ static int tsi721_enable_msix(struct tsi721_device *priv)
TSI721_MSIX_OMSG_INT(i);
}
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+ /*
+ * Initialize MSI-X entries for Block DMA Engine:
+ * this driver supports XXX DMA channels
+ * (one is reserved for SRIO maintenance transactions)
+ */
+ for (i = 0; i < TSI721_DMA_CHNUM; i++) {
+ entries[TSI721_VECT_DMA0_DONE + i].entry =
+ TSI721_MSIX_DMACH_DONE(i);
+ entries[TSI721_VECT_DMA0_INT + i].entry =
+ TSI721_MSIX_DMACH_INT(i);
+ }
+#endif /* CONFIG_RAPIDIO_DMA_ENGINE */
+
err = pci_enable_msix(priv->pdev, entries, ARRAY_SIZE(entries));
if (err) {
if (err > 0)
dev_info(&priv->pdev->dev,
"Only %d MSI-X vectors available, "
"not using MSI-X\n", err);
+ else
+ dev_err(&priv->pdev->dev,
+ "Failed to enable MSI-X (err=%d)\n", err);
return err;
}
@@ -760,6 +797,22 @@ static int tsi721_enable_msix(struct tsi721_device *priv)
i, pci_name(priv->pdev));
}
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+ for (i = 0; i < TSI721_DMA_CHNUM; i++) {
+ priv->msix[TSI721_VECT_DMA0_DONE + i].vector =
+ entries[TSI721_VECT_DMA0_DONE + i].vector;
+ snprintf(priv->msix[TSI721_VECT_DMA0_DONE + i].irq_name,
+ IRQ_DEVICE_NAME_MAX, DRV_NAME "-dmad%d@pci:%s",
+ i, pci_name(priv->pdev));
+
+ priv->msix[TSI721_VECT_DMA0_INT + i].vector =
+ entries[TSI721_VECT_DMA0_INT + i].vector;
+ snprintf(priv->msix[TSI721_VECT_DMA0_INT + i].irq_name,
+ IRQ_DEVICE_NAME_MAX, DRV_NAME "-dmai%d@pci:%s",
+ i, pci_name(priv->pdev));
+ }
+#endif /* CONFIG_RAPIDIO_DMA_ENGINE */
+
return 0;
}
#endif /* CONFIG_PCI_MSI */
@@ -888,20 +941,34 @@ static void tsi721_doorbell_free(struct tsi721_device *priv)
priv->idb_base = NULL;
}
-static int tsi721_bdma_ch_init(struct tsi721_device *priv, int chnum)
+/**
+ * tsi721_bdma_maint_init - Initialize maintenance request BDMA channel.
+ * @priv: pointer to tsi721 private data
+ *
+ * Initialize BDMA channel allocated for RapidIO maintenance read/write
+ * request generation
+ * Returns %0 on success or %-ENOMEM on failure.
+ */
+static int tsi721_bdma_maint_init(struct tsi721_device *priv)
{
struct tsi721_dma_desc *bd_ptr;
u64 *sts_ptr;
dma_addr_t bd_phys, sts_phys;
int sts_size;
- int bd_num = priv->bdma[chnum].bd_num;
+ int bd_num = 2;
+ void __iomem *regs;
- dev_dbg(&priv->pdev->dev, "Init Block DMA Engine, CH%d\n", chnum);
+ dev_dbg(&priv->pdev->dev,
+ "Init Block DMA Engine for Maintenance requests, CH%d\n",
+ TSI721_DMACH_MAINT);
/*
* Initialize DMA channel for maintenance requests
*/
+ priv->mdma.ch_id = TSI721_DMACH_MAINT;
+ regs = priv->regs + TSI721_DMAC_BASE(TSI721_DMACH_MAINT);
+
/* Allocate space for DMA descriptors */
bd_ptr = dma_zalloc_coherent(&priv->pdev->dev,
bd_num * sizeof(struct tsi721_dma_desc),
@@ -909,8 +976,9 @@ static int tsi721_bdma_ch_init(struct tsi721_device *priv, int chnum)
if (!bd_ptr)
return -ENOMEM;
- priv->bdma[chnum].bd_phys = bd_phys;
- priv->bdma[chnum].bd_base = bd_ptr;
+ priv->mdma.bd_num = bd_num;
+ priv->mdma.bd_phys = bd_phys;
+ priv->mdma.bd_base = bd_ptr;
dev_dbg(&priv->pdev->dev, "DMA descriptors @ %p (phys = %llx)\n",
bd_ptr, (unsigned long long)bd_phys);
@@ -927,13 +995,13 @@ static int tsi721_bdma_ch_init(struct tsi721_device *priv, int chnum)
dma_free_coherent(&priv->pdev->dev,
bd_num * sizeof(struct tsi721_dma_desc),
bd_ptr, bd_phys);
- priv->bdma[chnum].bd_base = NULL;
+ priv->mdma.bd_base = NULL;
return -ENOMEM;
}
- priv->bdma[chnum].sts_phys = sts_phys;
- priv->bdma[chnum].sts_base = sts_ptr;
- priv->bdma[chnum].sts_size = sts_size;
+ priv->mdma.sts_phys = sts_phys;
+ priv->mdma.sts_base = sts_ptr;
+ priv->mdma.sts_size = sts_size;
dev_dbg(&priv->pdev->dev,
"desc status FIFO @ %p (phys = %llx) size=0x%x\n",
@@ -946,83 +1014,61 @@ static int tsi721_bdma_ch_init(struct tsi721_device *priv, int chnum)
bd_ptr[bd_num - 1].next_hi = cpu_to_le32((u64)bd_phys >> 32);
/* Setup DMA descriptor pointers */
- iowrite32(((u64)bd_phys >> 32),
- priv->regs + TSI721_DMAC_DPTRH(chnum));
+ iowrite32(((u64)bd_phys >> 32), regs + TSI721_DMAC_DPTRH);
iowrite32(((u64)bd_phys & TSI721_DMAC_DPTRL_MASK),
- priv->regs + TSI721_DMAC_DPTRL(chnum));
+ regs + TSI721_DMAC_DPTRL);
/* Setup descriptor status FIFO */
- iowrite32(((u64)sts_phys >> 32),
- priv->regs + TSI721_DMAC_DSBH(chnum));
+ iowrite32(((u64)sts_phys >> 32), regs + TSI721_DMAC_DSBH);
iowrite32(((u64)sts_phys & TSI721_DMAC_DSBL_MASK),
- priv->regs + TSI721_DMAC_DSBL(chnum));
+ regs + TSI721_DMAC_DSBL);
iowrite32(TSI721_DMAC_DSSZ_SIZE(sts_size),
- priv->regs + TSI721_DMAC_DSSZ(chnum));
+ regs + TSI721_DMAC_DSSZ);
/* Clear interrupt bits */
- iowrite32(TSI721_DMAC_INT_ALL,
- priv->regs + TSI721_DMAC_INT(chnum));
+ iowrite32(TSI721_DMAC_INT_ALL, regs + TSI721_DMAC_INT);
- ioread32(priv->regs + TSI721_DMAC_INT(chnum));
+ ioread32(regs + TSI721_DMAC_INT);
/* Toggle DMA channel initialization */
- iowrite32(TSI721_DMAC_CTL_INIT, priv->regs + TSI721_DMAC_CTL(chnum));
- ioread32(priv->regs + TSI721_DMAC_CTL(chnum));
+ iowrite32(TSI721_DMAC_CTL_INIT, regs + TSI721_DMAC_CTL);
+ ioread32(regs + TSI721_DMAC_CTL);
udelay(10);
return 0;
}
-static int tsi721_bdma_ch_free(struct tsi721_device *priv, int chnum)
+static int tsi721_bdma_maint_free(struct tsi721_device *priv)
{
u32 ch_stat;
+ struct tsi721_bdma_maint *mdma = &priv->mdma;
+ void __iomem *regs = priv->regs + TSI721_DMAC_BASE(mdma->ch_id);
- if (priv->bdma[chnum].bd_base == NULL)
+ if (mdma->bd_base == NULL)
return 0;
/* Check if DMA channel still running */
- ch_stat = ioread32(priv->regs + TSI721_DMAC_STS(chnum));
+ ch_stat = ioread32(regs + TSI721_DMAC_STS);
if (ch_stat & TSI721_DMAC_STS_RUN)
return -EFAULT;
/* Put DMA channel into init state */
- iowrite32(TSI721_DMAC_CTL_INIT,
- priv->regs + TSI721_DMAC_CTL(chnum));
+ iowrite32(TSI721_DMAC_CTL_INIT, regs + TSI721_DMAC_CTL);
/* Free space allocated for DMA descriptors */
dma_free_coherent(&priv->pdev->dev,
- priv->bdma[chnum].bd_num * sizeof(struct tsi721_dma_desc),
- priv->bdma[chnum].bd_base, priv->bdma[chnum].bd_phys);
- priv->bdma[chnum].bd_base = NULL;
+ mdma->bd_num * sizeof(struct tsi721_dma_desc),
+ mdma->bd_base, mdma->bd_phys);
+ mdma->bd_base = NULL;
/* Free space allocated for status FIFO */
dma_free_coherent(&priv->pdev->dev,
- priv->bdma[chnum].sts_size * sizeof(struct tsi721_dma_sts),
- priv->bdma[chnum].sts_base, priv->bdma[chnum].sts_phys);
- priv->bdma[chnum].sts_base = NULL;
- return 0;
-}
-
-static int tsi721_bdma_init(struct tsi721_device *priv)
-{
- /* Initialize BDMA channel allocated for RapidIO maintenance read/write
- * request generation
- */
- priv->bdma[TSI721_DMACH_MAINT].bd_num = 2;
- if (tsi721_bdma_ch_init(priv, TSI721_DMACH_MAINT)) {
- dev_err(&priv->pdev->dev, "Unable to initialize maintenance DMA"
- " channel %d, aborting\n", TSI721_DMACH_MAINT);
- return -ENOMEM;
- }
-
+ mdma->sts_size * sizeof(struct tsi721_dma_sts),
+ mdma->sts_base, mdma->sts_phys);
+ mdma->sts_base = NULL;
return 0;
}
-static void tsi721_bdma_free(struct tsi721_device *priv)
-{
- tsi721_bdma_ch_free(priv, TSI721_DMACH_MAINT);
-}
-
/* Enable Inbound Messaging Interrupts */
static void
tsi721_imsg_interrupt_enable(struct tsi721_device *priv, int ch,
@@ -2035,7 +2081,8 @@ static void tsi721_disable_ints(struct tsi721_device *priv)
/* Disable all BDMA Channel interrupts */
for (ch = 0; ch < TSI721_DMA_MAXCH; ch++)
- iowrite32(0, priv->regs + TSI721_DMAC_INTE(ch));
+ iowrite32(0,
+ priv->regs + TSI721_DMAC_BASE(ch) + TSI721_DMAC_INTE);
/* Disable all general BDMA interrupts */
iowrite32(0, priv->regs + TSI721_BDMA_INTE);
@@ -2104,6 +2151,7 @@ static int __devinit tsi721_setup_mport(struct tsi721_device *priv)
mport->phy_type = RIO_PHY_SERIAL;
mport->priv = (void *)priv;
mport->phys_efptr = 0x100;
+ priv->mport = mport;
INIT_LIST_HEAD(&mport->dbells);
@@ -2129,17 +2177,21 @@ static int __devinit tsi721_setup_mport(struct tsi721_device *priv)
if (!err) {
tsi721_interrupts_init(priv);
ops->pwenable = tsi721_pw_enable;
- } else
+ } else {
dev_err(&pdev->dev, "Unable to get assigned PCI IRQ "
"vector %02X err=0x%x\n", pdev->irq, err);
+ goto err_exit;
+ }
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+ tsi721_register_dma(priv);
+#endif
/* Enable SRIO link */
iowrite32(ioread32(priv->regs + TSI721_DEVCTL) |
TSI721_DEVCTL_SRBOOT_CMPL,
priv->regs + TSI721_DEVCTL);
rio_register_mport(mport);
- priv->mport = mport;
if (mport->host_deviceid >= 0)
iowrite32(RIO_PORT_GEN_HOST | RIO_PORT_GEN_MASTER |
@@ -2149,6 +2201,11 @@ static int __devinit tsi721_setup_mport(struct tsi721_device *priv)
iowrite32(0, priv->regs + (0x100 + RIO_PORT_GEN_CTL_CSR));
return 0;
+
+err_exit:
+ kfree(mport);
+ kfree(ops);
+ return err;
}
static int __devinit tsi721_probe(struct pci_dev *pdev,
@@ -2294,7 +2351,7 @@ static int __devinit tsi721_probe(struct pci_dev *pdev,
tsi721_init_pc2sr_mapping(priv);
tsi721_init_sr2pc_mapping(priv);
- if (tsi721_bdma_init(priv)) {
+ if (tsi721_bdma_maint_init(priv)) {
dev_err(&pdev->dev, "BDMA initialization failed, aborting\n");
err = -ENOMEM;
goto err_unmap_bars;
@@ -2319,7 +2376,7 @@ static int __devinit tsi721_probe(struct pci_dev *pdev,
err_free_consistent:
tsi721_doorbell_free(priv);
err_free_bdma:
- tsi721_bdma_free(priv);
+ tsi721_bdma_maint_free(priv);
err_unmap_bars:
if (priv->regs)
iounmap(priv->regs);
diff --git a/drivers/rapidio/devices/tsi721.h b/drivers/rapidio/devices/tsi721.h
index 1c226b31af1..59de9d7be34 100644
--- a/drivers/rapidio/devices/tsi721.h
+++ b/drivers/rapidio/devices/tsi721.h
@@ -167,6 +167,8 @@
#define TSI721_DEV_INTE 0x29840
#define TSI721_DEV_INT 0x29844
#define TSI721_DEV_INTSET 0x29848
+#define TSI721_DEV_INT_BDMA_CH 0x00002000
+#define TSI721_DEV_INT_BDMA_NCH 0x00001000
#define TSI721_DEV_INT_SMSG_CH 0x00000800
#define TSI721_DEV_INT_SMSG_NCH 0x00000400
#define TSI721_DEV_INT_SR2PC_CH 0x00000200
@@ -181,6 +183,8 @@
#define TSI721_INT_IMSG_CHAN(x) (1 << (16 + (x)))
#define TSI721_INT_OMSG_CHAN_M 0x0000ff00
#define TSI721_INT_OMSG_CHAN(x) (1 << (8 + (x)))
+#define TSI721_INT_BDMA_CHAN_M 0x000000ff
+#define TSI721_INT_BDMA_CHAN(x) (1 << (x))
/*
* PC2SR block registers
@@ -235,14 +239,16 @@
* x = 0..7
*/
-#define TSI721_DMAC_DWRCNT(x) (0x51000 + (x) * 0x1000)
-#define TSI721_DMAC_DRDCNT(x) (0x51004 + (x) * 0x1000)
+#define TSI721_DMAC_BASE(x) (0x51000 + (x) * 0x1000)
-#define TSI721_DMAC_CTL(x) (0x51008 + (x) * 0x1000)
+#define TSI721_DMAC_DWRCNT 0x000
+#define TSI721_DMAC_DRDCNT 0x004
+
+#define TSI721_DMAC_CTL 0x008
#define TSI721_DMAC_CTL_SUSP 0x00000002
#define TSI721_DMAC_CTL_INIT 0x00000001
-#define TSI721_DMAC_INT(x) (0x5100c + (x) * 0x1000)
+#define TSI721_DMAC_INT 0x00c
#define TSI721_DMAC_INT_STFULL 0x00000010
#define TSI721_DMAC_INT_DONE 0x00000008
#define TSI721_DMAC_INT_SUSP 0x00000004
@@ -250,34 +256,33 @@
#define TSI721_DMAC_INT_IOFDONE 0x00000001
#define TSI721_DMAC_INT_ALL 0x0000001f
-#define TSI721_DMAC_INTSET(x) (0x51010 + (x) * 0x1000)
+#define TSI721_DMAC_INTSET 0x010
-#define TSI721_DMAC_STS(x) (0x51014 + (x) * 0x1000)
+#define TSI721_DMAC_STS 0x014
#define TSI721_DMAC_STS_ABORT 0x00400000
#define TSI721_DMAC_STS_RUN 0x00200000
#define TSI721_DMAC_STS_CS 0x001f0000
-#define TSI721_DMAC_INTE(x) (0x51018 + (x) * 0x1000)
+#define TSI721_DMAC_INTE 0x018
-#define TSI721_DMAC_DPTRL(x) (0x51024 + (x) * 0x1000)
+#define TSI721_DMAC_DPTRL 0x024
#define TSI721_DMAC_DPTRL_MASK 0xffffffe0
-#define TSI721_DMAC_DPTRH(x) (0x51028 + (x) * 0x1000)
+#define TSI721_DMAC_DPTRH 0x028
-#define TSI721_DMAC_DSBL(x) (0x5102c + (x) * 0x1000)
+#define TSI721_DMAC_DSBL 0x02c
#define TSI721_DMAC_DSBL_MASK 0xffffffc0
-#define TSI721_DMAC_DSBH(x) (0x51030 + (x) * 0x1000)
+#define TSI721_DMAC_DSBH 0x030
-#define TSI721_DMAC_DSSZ(x) (0x51034 + (x) * 0x1000)
+#define TSI721_DMAC_DSSZ 0x034
#define TSI721_DMAC_DSSZ_SIZE_M 0x0000000f
#define TSI721_DMAC_DSSZ_SIZE(size) (__fls(size) - 4)
-
-#define TSI721_DMAC_DSRP(x) (0x51038 + (x) * 0x1000)
+#define TSI721_DMAC_DSRP 0x038
#define TSI721_DMAC_DSRP_MASK 0x0007ffff
-#define TSI721_DMAC_DSWP(x) (0x5103c + (x) * 0x1000)
+#define TSI721_DMAC_DSWP 0x03c
#define TSI721_DMAC_DSWP_MASK 0x0007ffff
#define TSI721_BDMA_INTE 0x5f000
@@ -612,6 +617,8 @@ enum dma_rtype {
#define TSI721_DMACH_MAINT 0 /* DMA channel for maint requests */
#define TSI721_DMACH_MAINT_NBD 32 /* Number of BDs for maint requests */
+#define TSI721_DMACH_DMA 1 /* DMA channel for data transfers */
+
#define MSG_DMA_ENTRY_INX_TO_SIZE(x) ((0x10 << (x)) & 0xFFFF0)
enum tsi721_smsg_int_flag {
@@ -626,7 +633,48 @@ enum tsi721_smsg_int_flag {
/* Structures */
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+
+struct tsi721_tx_desc {
+ struct dma_async_tx_descriptor txd;
+ struct tsi721_dma_desc *hw_desc;
+ u16 destid;
+ /* low 64-bits of 66-bit RIO address */
+ u64 rio_addr;
+ /* upper 2-bits of 66-bit RIO address */
+ u8 rio_addr_u;
+ bool interrupt;
+ struct list_head desc_node;
+ struct list_head tx_list;
+};
+
struct tsi721_bdma_chan {
+ int id;
+ void __iomem *regs;
+ int bd_num; /* number of buffer descriptors */
+ void *bd_base; /* start of DMA descriptors */
+ dma_addr_t bd_phys;
+ void *sts_base; /* start of DMA BD status FIFO */
+ dma_addr_t sts_phys;
+ int sts_size;
+ u32 sts_rdptr;
+ u32 wr_count;
+ u32 wr_count_next;
+
+ struct dma_chan dchan;
+ struct tsi721_tx_desc *tx_desc;
+ spinlock_t lock;
+ struct list_head active_list;
+ struct list_head queue;
+ struct list_head free_list;
+ dma_cookie_t completed_cookie;
+ struct tasklet_struct tasklet;
+};
+
+#endif /* CONFIG_RAPIDIO_DMA_ENGINE */
+
+struct tsi721_bdma_maint {
+ int ch_id; /* BDMA channel number */
int bd_num; /* number of buffer descriptors */
void *bd_base; /* start of DMA descriptors */
dma_addr_t bd_phys;
@@ -721,6 +769,24 @@ enum tsi721_msix_vect {
TSI721_VECT_IMB1_INT,
TSI721_VECT_IMB2_INT,
TSI721_VECT_IMB3_INT,
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+ TSI721_VECT_DMA0_DONE,
+ TSI721_VECT_DMA1_DONE,
+ TSI721_VECT_DMA2_DONE,
+ TSI721_VECT_DMA3_DONE,
+ TSI721_VECT_DMA4_DONE,
+ TSI721_VECT_DMA5_DONE,
+ TSI721_VECT_DMA6_DONE,
+ TSI721_VECT_DMA7_DONE,
+ TSI721_VECT_DMA0_INT,
+ TSI721_VECT_DMA1_INT,
+ TSI721_VECT_DMA2_INT,
+ TSI721_VECT_DMA3_INT,
+ TSI721_VECT_DMA4_INT,
+ TSI721_VECT_DMA5_INT,
+ TSI721_VECT_DMA6_INT,
+ TSI721_VECT_DMA7_INT,
+#endif /* CONFIG_RAPIDIO_DMA_ENGINE */
TSI721_VECT_MAX
};
@@ -754,7 +820,11 @@ struct tsi721_device {
u32 pw_discard_count;
/* BDMA Engine */
+ struct tsi721_bdma_maint mdma; /* Maintenance rd/wr request channel */
+
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
struct tsi721_bdma_chan bdma[TSI721_DMA_CHNUM];
+#endif
/* Inbound Messaging */
int imsg_init[TSI721_IMSG_CHNUM];
@@ -765,4 +835,9 @@ struct tsi721_device {
struct tsi721_omsg_ring omsg_ring[TSI721_OMSG_CHNUM];
};
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+extern void tsi721_bdma_handler(struct tsi721_bdma_chan *bdma_chan);
+extern int __devinit tsi721_register_dma(struct tsi721_device *priv);
+#endif
+
#endif
diff --git a/drivers/rapidio/devices/tsi721_dma.c b/drivers/rapidio/devices/tsi721_dma.c
new file mode 100644
index 00000000000..92e06a5c62e
--- /dev/null
+++ b/drivers/rapidio/devices/tsi721_dma.c
@@ -0,0 +1,823 @@
+/*
+ * DMA Engine support for Tsi721 PCIExpress-to-SRIO bridge
+ *
+ * Copyright 2011 Integrated Device Technology, Inc.
+ * Alexandre Bounine <alexandre.bounine@idt.com>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program; if not, write to the Free Software Foundation, Inc., 59
+ * Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ */
+
+#include <linux/io.h>
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/ioport.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/pci.h>
+#include <linux/rio.h>
+#include <linux/rio_drv.h>
+#include <linux/dma-mapping.h>
+#include <linux/interrupt.h>
+#include <linux/kfifo.h>
+#include <linux/delay.h>
+
+#include "tsi721.h"
+
+static inline struct tsi721_bdma_chan *to_tsi721_chan(struct dma_chan *chan)
+{
+ return container_of(chan, struct tsi721_bdma_chan, dchan);
+}
+
+static inline struct tsi721_device *to_tsi721(struct dma_device *ddev)
+{
+ return container_of(ddev, struct rio_mport, dma)->priv;
+}
+
+static inline
+struct tsi721_tx_desc *to_tsi721_desc(struct dma_async_tx_descriptor *txd)
+{
+ return container_of(txd, struct tsi721_tx_desc, txd);
+}
+
+static inline
+struct tsi721_tx_desc *tsi721_dma_first_active(
+ struct tsi721_bdma_chan *bdma_chan)
+{
+ return list_first_entry(&bdma_chan->active_list,
+ struct tsi721_tx_desc, desc_node);
+}
+
+static int tsi721_bdma_ch_init(struct tsi721_bdma_chan *bdma_chan)
+{
+ struct tsi721_dma_desc *bd_ptr;
+ struct device *dev = bdma_chan->dchan.device->dev;
+ u64 *sts_ptr;
+ dma_addr_t bd_phys;
+ dma_addr_t sts_phys;
+ int sts_size;
+ int bd_num = bdma_chan->bd_num;
+
+ dev_dbg(dev, "Init Block DMA Engine, CH%d\n", bdma_chan->id);
+
+ /* Allocate space for DMA descriptors */
+ bd_ptr = dma_zalloc_coherent(dev,
+ bd_num * sizeof(struct tsi721_dma_desc),
+ &bd_phys, GFP_KERNEL);
+ if (!bd_ptr)
+ return -ENOMEM;
+
+ bdma_chan->bd_phys = bd_phys;
+ bdma_chan->bd_base = bd_ptr;
+
+ dev_dbg(dev, "DMA descriptors @ %p (phys = %llx)\n",
+ bd_ptr, (unsigned long long)bd_phys);
+
+ /* Allocate space for descriptor status FIFO */
+ sts_size = (bd_num >= TSI721_DMA_MINSTSSZ) ?
+ bd_num : TSI721_DMA_MINSTSSZ;
+ sts_size = roundup_pow_of_two(sts_size);
+ sts_ptr = dma_zalloc_coherent(dev,
+ sts_size * sizeof(struct tsi721_dma_sts),
+ &sts_phys, GFP_KERNEL);
+ if (!sts_ptr) {
+ /* Free space allocated for DMA descriptors */
+ dma_free_coherent(dev,
+ bd_num * sizeof(struct tsi721_dma_desc),
+ bd_ptr, bd_phys);
+ bdma_chan->bd_base = NULL;
+ return -ENOMEM;
+ }
+
+ bdma_chan->sts_phys = sts_phys;
+ bdma_chan->sts_base = sts_ptr;
+ bdma_chan->sts_size = sts_size;
+
+ dev_dbg(dev,
+ "desc status FIFO @ %p (phys = %llx) size=0x%x\n",
+ sts_ptr, (unsigned long long)sts_phys, sts_size);
+
+ /* Initialize DMA descriptors ring */
+ bd_ptr[bd_num - 1].type_id = cpu_to_le32(DTYPE3 << 29);
+ bd_ptr[bd_num - 1].next_lo = cpu_to_le32((u64)bd_phys &
+ TSI721_DMAC_DPTRL_MASK);
+ bd_ptr[bd_num - 1].next_hi = cpu_to_le32((u64)bd_phys >> 32);
+
+ /* Setup DMA descriptor pointers */
+ iowrite32(((u64)bd_phys >> 32),
+ bdma_chan->regs + TSI721_DMAC_DPTRH);
+ iowrite32(((u64)bd_phys & TSI721_DMAC_DPTRL_MASK),
+ bdma_chan->regs + TSI721_DMAC_DPTRL);
+
+ /* Setup descriptor status FIFO */
+ iowrite32(((u64)sts_phys >> 32),
+ bdma_chan->regs + TSI721_DMAC_DSBH);
+ iowrite32(((u64)sts_phys & TSI721_DMAC_DSBL_MASK),
+ bdma_chan->regs + TSI721_DMAC_DSBL);
+ iowrite32(TSI721_DMAC_DSSZ_SIZE(sts_size),
+ bdma_chan->regs + TSI721_DMAC_DSSZ);
+
+ /* Clear interrupt bits */
+ iowrite32(TSI721_DMAC_INT_ALL,
+ bdma_chan->regs + TSI721_DMAC_INT);
+
+ ioread32(bdma_chan->regs + TSI721_DMAC_INT);
+
+ /* Toggle DMA channel initialization */
+ iowrite32(TSI721_DMAC_CTL_INIT, bdma_chan->regs + TSI721_DMAC_CTL);
+ ioread32(bdma_chan->regs + TSI721_DMAC_CTL);
+ bdma_chan->wr_count = bdma_chan->wr_count_next = 0;
+ bdma_chan->sts_rdptr = 0;
+ udelay(10);
+
+ return 0;
+}
+
+static int tsi721_bdma_ch_free(struct tsi721_bdma_chan *bdma_chan)
+{
+ u32 ch_stat;
+
+ if (bdma_chan->bd_base == NULL)
+ return 0;
+
+ /* Check if DMA channel still running */
+ ch_stat = ioread32(bdma_chan->regs + TSI721_DMAC_STS);
+ if (ch_stat & TSI721_DMAC_STS_RUN)
+ return -EFAULT;
+
+ /* Put DMA channel into init state */
+ iowrite32(TSI721_DMAC_CTL_INIT, bdma_chan->regs + TSI721_DMAC_CTL);
+
+ /* Free space allocated for DMA descriptors */
+ dma_free_coherent(bdma_chan->dchan.device->dev,
+ bdma_chan->bd_num * sizeof(struct tsi721_dma_desc),
+ bdma_chan->bd_base, bdma_chan->bd_phys);
+ bdma_chan->bd_base = NULL;
+
+ /* Free space allocated for status FIFO */
+ dma_free_coherent(bdma_chan->dchan.device->dev,
+ bdma_chan->sts_size * sizeof(struct tsi721_dma_sts),
+ bdma_chan->sts_base, bdma_chan->sts_phys);
+ bdma_chan->sts_base = NULL;
+ return 0;
+}
+
+static void
+tsi721_bdma_interrupt_enable(struct tsi721_bdma_chan *bdma_chan, int enable)
+{
+ if (enable) {
+ /* Clear pending BDMA channel interrupts */
+ iowrite32(TSI721_DMAC_INT_ALL,
+ bdma_chan->regs + TSI721_DMAC_INT);
+ ioread32(bdma_chan->regs + TSI721_DMAC_INT);
+ /* Enable BDMA channel interrupts */
+ iowrite32(TSI721_DMAC_INT_ALL,
+ bdma_chan->regs + TSI721_DMAC_INTE);
+ } else {
+ /* Disable BDMA channel interrupts */
+ iowrite32(0, bdma_chan->regs + TSI721_DMAC_INTE);
+ /* Clear pending BDMA channel interrupts */
+ iowrite32(TSI721_DMAC_INT_ALL,
+ bdma_chan->regs + TSI721_DMAC_INT);
+ }
+
+}
+
+static bool tsi721_dma_is_idle(struct tsi721_bdma_chan *bdma_chan)
+{
+ u32 sts;
+
+ sts = ioread32(bdma_chan->regs + TSI721_DMAC_STS);
+ return ((sts & TSI721_DMAC_STS_RUN) == 0);
+}
+
+void tsi721_bdma_handler(struct tsi721_bdma_chan *bdma_chan)
+{
+ /* Disable BDMA channel interrupts */
+ iowrite32(0, bdma_chan->regs + TSI721_DMAC_INTE);
+
+ tasklet_schedule(&bdma_chan->tasklet);
+}
+
+#ifdef CONFIG_PCI_MSI
+/**
+ * tsi721_omsg_msix - MSI-X interrupt handler for BDMA channels
+ * @irq: Linux interrupt number
+ * @ptr: Pointer to interrupt-specific data (BDMA channel structure)
+ *
+ * Handles BDMA channel interrupts signaled using MSI-X.
+ */
+static irqreturn_t tsi721_bdma_msix(int irq, void *ptr)
+{
+ struct tsi721_bdma_chan *bdma_chan = ptr;
+
+ tsi721_bdma_handler(bdma_chan);
+ return IRQ_HANDLED;
+}
+#endif /* CONFIG_PCI_MSI */
+
+/* Must be called with the spinlock held */
+static void tsi721_start_dma(struct tsi721_bdma_chan *bdma_chan)
+{
+ if (!tsi721_dma_is_idle(bdma_chan)) {
+ dev_err(bdma_chan->dchan.device->dev,
+ "BUG: Attempt to start non-idle channel\n");
+ return;
+ }
+
+ if (bdma_chan->wr_count == bdma_chan->wr_count_next) {
+ dev_err(bdma_chan->dchan.device->dev,
+ "BUG: Attempt to start DMA with no BDs ready\n");
+ return;
+ }
+
+ dev_dbg(bdma_chan->dchan.device->dev,
+ "tx_chan: %p, chan: %d, regs: %p\n",
+ bdma_chan, bdma_chan->dchan.chan_id, bdma_chan->regs);
+
+ iowrite32(bdma_chan->wr_count_next,
+ bdma_chan->regs + TSI721_DMAC_DWRCNT);
+ ioread32(bdma_chan->regs + TSI721_DMAC_DWRCNT);
+
+ bdma_chan->wr_count = bdma_chan->wr_count_next;
+}
+
+static void tsi721_desc_put(struct tsi721_bdma_chan *bdma_chan,
+ struct tsi721_tx_desc *desc)
+{
+ dev_dbg(bdma_chan->dchan.device->dev,
+ "Put desc: %p into free list\n", desc);
+
+ if (desc) {
+ spin_lock_bh(&bdma_chan->lock);
+ list_splice_init(&desc->tx_list, &bdma_chan->free_list);
+ list_add(&desc->desc_node, &bdma_chan->free_list);
+ bdma_chan->wr_count_next = bdma_chan->wr_count;
+ spin_unlock_bh(&bdma_chan->lock);
+ }
+}
+
+static
+struct tsi721_tx_desc *tsi721_desc_get(struct tsi721_bdma_chan *bdma_chan)
+{
+ struct tsi721_tx_desc *tx_desc, *_tx_desc;
+ struct tsi721_tx_desc *ret = NULL;
+ int i;
+
+ spin_lock_bh(&bdma_chan->lock);
+ list_for_each_entry_safe(tx_desc, _tx_desc,
+ &bdma_chan->free_list, desc_node) {
+ if (async_tx_test_ack(&tx_desc->txd)) {
+ list_del(&tx_desc->desc_node);
+ ret = tx_desc;
+ break;
+ }
+ dev_dbg(bdma_chan->dchan.device->dev,
+ "desc %p not ACKed\n", tx_desc);
+ }
+
+ i = bdma_chan->wr_count_next % bdma_chan->bd_num;
+ if (i == bdma_chan->bd_num - 1) {
+ i = 0;
+ bdma_chan->wr_count_next++; /* skip link descriptor */
+ }
+
+ bdma_chan->wr_count_next++;
+ tx_desc->txd.phys = bdma_chan->bd_phys +
+ i * sizeof(struct tsi721_dma_desc);
+ tx_desc->hw_desc = &((struct tsi721_dma_desc *)bdma_chan->bd_base)[i];
+
+ spin_unlock_bh(&bdma_chan->lock);
+
+ return ret;
+}
+
+static int
+tsi721_fill_desc(struct tsi721_bdma_chan *bdma_chan,
+ struct tsi721_tx_desc *desc, struct scatterlist *sg,
+ enum dma_rtype rtype, u32 sys_size)
+{
+ struct tsi721_dma_desc *bd_ptr = desc->hw_desc;
+ u64 rio_addr;
+
+ if (sg_dma_len(sg) > TSI721_DMAD_BCOUNT1 + 1) {
+ dev_err(bdma_chan->dchan.device->dev,
+ "SG element is too large\n");
+ return -EINVAL;
+ }
+
+ dev_dbg(bdma_chan->dchan.device->dev,
+ "desc: 0x%llx, addr: 0x%llx len: 0x%x\n",
+ (u64)desc->txd.phys, (unsigned long long)sg_dma_address(sg),
+ sg_dma_len(sg));
+
+ dev_dbg(bdma_chan->dchan.device->dev,
+ "bd_ptr = %p did=%d raddr=0x%llx\n",
+ bd_ptr, desc->destid, desc->rio_addr);
+
+ /* Initialize DMA descriptor */
+ bd_ptr->type_id = cpu_to_le32((DTYPE1 << 29) |
+ (rtype << 19) | desc->destid);
+ if (desc->interrupt)
+ bd_ptr->type_id |= cpu_to_le32(TSI721_DMAD_IOF);
+ bd_ptr->bcount = cpu_to_le32(((desc->rio_addr & 0x3) << 30) |
+ (sys_size << 26) | sg_dma_len(sg));
+ rio_addr = (desc->rio_addr >> 2) |
+ ((u64)(desc->rio_addr_u & 0x3) << 62);
+ bd_ptr->raddr_lo = cpu_to_le32(rio_addr & 0xffffffff);
+ bd_ptr->raddr_hi = cpu_to_le32(rio_addr >> 32);
+ bd_ptr->t1.bufptr_lo = cpu_to_le32(
+ (u64)sg_dma_address(sg) & 0xffffffff);
+ bd_ptr->t1.bufptr_hi = cpu_to_le32((u64)sg_dma_address(sg) >> 32);
+ bd_ptr->t1.s_dist = 0;
+ bd_ptr->t1.s_size = 0;
+
+ return 0;
+}
+
+static void tsi721_dma_chain_complete(struct tsi721_bdma_chan *bdma_chan,
+ struct tsi721_tx_desc *desc)
+{
+ struct dma_async_tx_descriptor *txd = &desc->txd;
+ dma_async_tx_callback callback = txd->callback;
+ void *param = txd->callback_param;
+
+ list_splice_init(&desc->tx_list, &bdma_chan->free_list);
+ list_move(&desc->desc_node, &bdma_chan->free_list);
+ bdma_chan->completed_cookie = txd->cookie;
+
+ if (callback)
+ callback(param);
+}
+
+static void tsi721_dma_complete_all(struct tsi721_bdma_chan *bdma_chan)
+{
+ struct tsi721_tx_desc *desc, *_d;
+ LIST_HEAD(list);
+
+ BUG_ON(!tsi721_dma_is_idle(bdma_chan));
+
+ if (!list_empty(&bdma_chan->queue))
+ tsi721_start_dma(bdma_chan);
+
+ list_splice_init(&bdma_chan->active_list, &list);
+ list_splice_init(&bdma_chan->queue, &bdma_chan->active_list);
+
+ list_for_each_entry_safe(desc, _d, &list, desc_node)
+ tsi721_dma_chain_complete(bdma_chan, desc);
+}
+
+static void tsi721_clr_stat(struct tsi721_bdma_chan *bdma_chan)
+{
+ u32 srd_ptr;
+ u64 *sts_ptr;
+ int i, j;
+
+ /* Check and clear descriptor status FIFO entries */
+ srd_ptr = bdma_chan->sts_rdptr;
+ sts_ptr = bdma_chan->sts_base;
+ j = srd_ptr * 8;
+ while (sts_ptr[j]) {
+ for (i = 0; i < 8 && sts_ptr[j]; i++, j++)
+ sts_ptr[j] = 0;
+
+ ++srd_ptr;
+ srd_ptr %= bdma_chan->sts_size;
+ j = srd_ptr * 8;
+ }
+
+ iowrite32(srd_ptr, bdma_chan->regs + TSI721_DMAC_DSRP);
+ bdma_chan->sts_rdptr = srd_ptr;
+}
+
+static void tsi721_advance_work(struct tsi721_bdma_chan *bdma_chan)
+{
+ if (list_empty(&bdma_chan->active_list) ||
+ list_is_singular(&bdma_chan->active_list)) {
+ dev_dbg(bdma_chan->dchan.device->dev,
+ "%s: Active_list empty\n", __func__);
+ tsi721_dma_complete_all(bdma_chan);
+ } else {
+ dev_dbg(bdma_chan->dchan.device->dev,
+ "%s: Active_list NOT empty\n", __func__);
+ tsi721_dma_chain_complete(bdma_chan,
+ tsi721_dma_first_active(bdma_chan));
+ tsi721_start_dma(bdma_chan);
+ }
+}
+
+static void tsi721_dma_tasklet(unsigned long data)
+{
+ struct tsi721_bdma_chan *bdma_chan = (struct tsi721_bdma_chan *)data;
+ u32 dmac_int, dmac_sts;
+
+ dmac_int = ioread32(bdma_chan->regs + TSI721_DMAC_INT);
+ dev_dbg(bdma_chan->dchan.device->dev, "%s: DMAC%d_INT = 0x%x\n",
+ __func__, bdma_chan->id, dmac_int);
+ /* Clear channel interrupts */
+ iowrite32(dmac_int, bdma_chan->regs + TSI721_DMAC_INT);
+
+ if (dmac_int & TSI721_DMAC_INT_ERR) {
+ dmac_sts = ioread32(bdma_chan->regs + TSI721_DMAC_STS);
+ dev_err(bdma_chan->dchan.device->dev,
+ "%s: DMA ERROR - DMAC%d_STS = 0x%x\n",
+ __func__, bdma_chan->id, dmac_sts);
+ }
+
+ if (dmac_int & TSI721_DMAC_INT_STFULL) {
+ dev_err(bdma_chan->dchan.device->dev,
+ "%s: DMAC%d descriptor status FIFO is full\n",
+ __func__, bdma_chan->id);
+ }
+
+ if (dmac_int & (TSI721_DMAC_INT_DONE | TSI721_DMAC_INT_IOFDONE)) {
+ tsi721_clr_stat(bdma_chan);
+ spin_lock(&bdma_chan->lock);
+ tsi721_advance_work(bdma_chan);
+ spin_unlock(&bdma_chan->lock);
+ }
+
+ /* Re-Enable BDMA channel interrupts */
+ iowrite32(TSI721_DMAC_INT_ALL, bdma_chan->regs + TSI721_DMAC_INTE);
+}
+
+static dma_cookie_t tsi721_tx_submit(struct dma_async_tx_descriptor *txd)
+{
+ struct tsi721_tx_desc *desc = to_tsi721_desc(txd);
+ struct tsi721_bdma_chan *bdma_chan = to_tsi721_chan(txd->chan);
+ dma_cookie_t cookie;
+
+ spin_lock_bh(&bdma_chan->lock);
+
+ cookie = txd->chan->cookie;
+ if (++cookie < 0)
+ cookie = 1;
+ txd->chan->cookie = cookie;
+ txd->cookie = cookie;
+
+ if (list_empty(&bdma_chan->active_list)) {
+ list_add_tail(&desc->desc_node, &bdma_chan->active_list);
+ tsi721_start_dma(bdma_chan);
+ } else {
+ list_add_tail(&desc->desc_node, &bdma_chan->queue);
+ }
+
+ spin_unlock_bh(&bdma_chan->lock);
+ return cookie;
+}
+
+static int tsi721_alloc_chan_resources(struct dma_chan *dchan)
+{
+ struct tsi721_bdma_chan *bdma_chan = to_tsi721_chan(dchan);
+#ifdef CONFIG_PCI_MSI
+ struct tsi721_device *priv = to_tsi721(dchan->device);
+#endif
+ struct tsi721_tx_desc *desc = NULL;
+ LIST_HEAD(tmp_list);
+ int i;
+ int rc;
+
+ if (bdma_chan->bd_base)
+ return bdma_chan->bd_num - 1;
+
+ /* Initialize BDMA channel */
+ if (tsi721_bdma_ch_init(bdma_chan)) {
+ dev_err(dchan->device->dev, "Unable to initialize data DMA"
+ " channel %d, aborting\n", bdma_chan->id);
+ return -ENOMEM;
+ }
+
+ /* Alocate matching number of logical descriptors */
+ desc = kcalloc((bdma_chan->bd_num - 1), sizeof(struct tsi721_tx_desc),
+ GFP_KERNEL);
+ if (!desc) {
+ dev_err(dchan->device->dev,
+ "Failed to allocate logical descriptors\n");
+ rc = -ENOMEM;
+ goto err_out;
+ }
+
+ bdma_chan->tx_desc = desc;
+
+ for (i = 0; i < bdma_chan->bd_num - 1; i++) {
+ dma_async_tx_descriptor_init(&desc[i].txd, dchan);
+ desc[i].txd.tx_submit = tsi721_tx_submit;
+ desc[i].txd.flags = DMA_CTRL_ACK;
+ INIT_LIST_HEAD(&desc[i].tx_list);
+ list_add_tail(&desc[i].desc_node, &tmp_list);
+ }
+
+ spin_lock_bh(&bdma_chan->lock);
+ list_splice(&tmp_list, &bdma_chan->free_list);
+ bdma_chan->completed_cookie = dchan->cookie = 1;
+ spin_unlock_bh(&bdma_chan->lock);
+
+#ifdef CONFIG_PCI_MSI
+ if (priv->flags & TSI721_USING_MSIX) {
+ /* Request interrupt service if we are in MSI-X mode */
+ rc = request_irq(
+ priv->msix[TSI721_VECT_DMA0_DONE +
+ bdma_chan->id].vector,
+ tsi721_bdma_msix, 0,
+ priv->msix[TSI721_VECT_DMA0_DONE +
+ bdma_chan->id].irq_name,
+ (void *)bdma_chan);
+
+ if (rc) {
+ dev_dbg(dchan->device->dev,
+ "Unable to allocate MSI-X interrupt for "
+ "BDMA%d-DONE\n", bdma_chan->id);
+ goto err_out;
+ }
+
+ rc = request_irq(priv->msix[TSI721_VECT_DMA0_INT +
+ bdma_chan->id].vector,
+ tsi721_bdma_msix, 0,
+ priv->msix[TSI721_VECT_DMA0_INT +
+ bdma_chan->id].irq_name,
+ (void *)bdma_chan);
+
+ if (rc) {
+ dev_dbg(dchan->device->dev,
+ "Unable to allocate MSI-X interrupt for "
+ "BDMA%d-INT\n", bdma_chan->id);
+ free_irq(
+ priv->msix[TSI721_VECT_DMA0_DONE +
+ bdma_chan->id].vector,
+ (void *)bdma_chan);
+ rc = -EIO;
+ goto err_out;
+ }
+ }
+#endif /* CONFIG_PCI_MSI */
+
+ tasklet_enable(&bdma_chan->tasklet);
+ tsi721_bdma_interrupt_enable(bdma_chan, 1);
+
+ return bdma_chan->bd_num - 1;
+
+err_out:
+ kfree(desc);
+ tsi721_bdma_ch_free(bdma_chan);
+ return rc;
+}
+
+static void tsi721_free_chan_resources(struct dma_chan *dchan)
+{
+ struct tsi721_bdma_chan *bdma_chan = to_tsi721_chan(dchan);
+#ifdef CONFIG_PCI_MSI
+ struct tsi721_device *priv = to_tsi721(dchan->device);
+#endif
+ LIST_HEAD(list);
+
+ dev_dbg(dchan->device->dev, "%s: Entry\n", __func__);
+
+ if (bdma_chan->bd_base == NULL)
+ return;
+
+ BUG_ON(!list_empty(&bdma_chan->active_list));
+ BUG_ON(!list_empty(&bdma_chan->queue));
+
+ tasklet_disable(&bdma_chan->tasklet);
+
+ spin_lock_bh(&bdma_chan->lock);
+ list_splice_init(&bdma_chan->free_list, &list);
+ spin_unlock_bh(&bdma_chan->lock);
+
+ tsi721_bdma_interrupt_enable(bdma_chan, 0);
+
+#ifdef CONFIG_PCI_MSI
+ if (priv->flags & TSI721_USING_MSIX) {
+ free_irq(priv->msix[TSI721_VECT_DMA0_DONE +
+ bdma_chan->id].vector, (void *)bdma_chan);
+ free_irq(priv->msix[TSI721_VECT_DMA0_INT +
+ bdma_chan->id].vector, (void *)bdma_chan);
+ }
+#endif /* CONFIG_PCI_MSI */
+
+ tsi721_bdma_ch_free(bdma_chan);
+ kfree(bdma_chan->tx_desc);
+}
+
+static
+enum dma_status tsi721_tx_status(struct dma_chan *dchan, dma_cookie_t cookie,
+ struct dma_tx_state *txstate)
+{
+ struct tsi721_bdma_chan *bdma_chan = to_tsi721_chan(dchan);
+ dma_cookie_t last_used;
+ dma_cookie_t last_completed;
+ int ret;
+
+ spin_lock_bh(&bdma_chan->lock);
+ last_completed = bdma_chan->completed_cookie;
+ last_used = dchan->cookie;
+ spin_unlock_bh(&bdma_chan->lock);
+
+ ret = dma_async_is_complete(cookie, last_completed, last_used);
+
+ dma_set_tx_state(txstate, last_completed, last_used, 0);
+
+ dev_dbg(dchan->device->dev,
+ "%s: exit, ret: %d, last_completed: %d, last_used: %d\n",
+ __func__, ret, last_completed, last_used);
+
+ return ret;
+}
+
+static void tsi721_issue_pending(struct dma_chan *dchan)
+{
+ struct tsi721_bdma_chan *bdma_chan = to_tsi721_chan(dchan);
+
+ dev_dbg(dchan->device->dev, "%s: Entry\n", __func__);
+
+ if (tsi721_dma_is_idle(bdma_chan)) {
+ spin_lock_bh(&bdma_chan->lock);
+ tsi721_advance_work(bdma_chan);
+ spin_unlock_bh(&bdma_chan->lock);
+ } else
+ dev_dbg(dchan->device->dev,
+ "%s: DMA channel still busy\n", __func__);
+}
+
+static
+struct dma_async_tx_descriptor *tsi721_prep_rio_sg(struct dma_chan *dchan,
+ struct scatterlist *sgl, unsigned int sg_len,
+ enum dma_transfer_direction dir, unsigned long flags,
+ void *tinfo)
+{
+ struct tsi721_bdma_chan *bdma_chan = to_tsi721_chan(dchan);
+ struct tsi721_tx_desc *desc = NULL;
+ struct tsi721_tx_desc *first = NULL;
+ struct scatterlist *sg;
+ struct rio_dma_ext *rext = tinfo;
+ u64 rio_addr = rext->rio_addr; /* limited to 64-bit rio_addr for now */
+ unsigned int i;
+ u32 sys_size = dma_to_mport(dchan->device)->sys_size;
+ enum dma_rtype rtype;
+
+ if (!sgl || !sg_len) {
+ dev_err(dchan->device->dev, "%s: No SG list\n", __func__);
+ return NULL;
+ }
+
+ if (dir == DMA_DEV_TO_MEM)
+ rtype = NREAD;
+ else if (dir == DMA_MEM_TO_DEV) {
+ switch (rext->wr_type) {
+ case RDW_ALL_NWRITE:
+ rtype = ALL_NWRITE;
+ break;
+ case RDW_ALL_NWRITE_R:
+ rtype = ALL_NWRITE_R;
+ break;
+ case RDW_LAST_NWRITE_R:
+ default:
+ rtype = LAST_NWRITE_R;
+ break;
+ }
+ } else {
+ dev_err(dchan->device->dev,
+ "%s: Unsupported DMA direction option\n", __func__);
+ return NULL;
+ }
+
+ for_each_sg(sgl, sg, sg_len, i) {
+ int err;
+
+ dev_dbg(dchan->device->dev, "%s: sg #%d\n", __func__, i);
+ desc = tsi721_desc_get(bdma_chan);
+ if (!desc) {
+ dev_err(dchan->device->dev,
+ "Not enough descriptors available\n");
+ goto err_desc_get;
+ }
+
+ if (sg_is_last(sg))
+ desc->interrupt = (flags & DMA_PREP_INTERRUPT) != 0;
+ else
+ desc->interrupt = false;
+
+ desc->destid = rext->destid;
+ desc->rio_addr = rio_addr;
+ desc->rio_addr_u = 0;
+
+ err = tsi721_fill_desc(bdma_chan, desc, sg, rtype, sys_size);
+ if (err) {
+ dev_err(dchan->device->dev,
+ "Failed to build desc: %d\n", err);
+ goto err_desc_get;
+ }
+
+ rio_addr += sg_dma_len(sg);
+
+ if (!first)
+ first = desc;
+ else
+ list_add_tail(&desc->desc_node, &first->tx_list);
+ }
+
+ first->txd.cookie = -EBUSY;
+ desc->txd.flags = flags;
+
+ return &first->txd;
+
+err_desc_get:
+ tsi721_desc_put(bdma_chan, first);
+ return NULL;
+}
+
+static int tsi721_device_control(struct dma_chan *dchan, enum dma_ctrl_cmd cmd,
+ unsigned long arg)
+{
+ struct tsi721_bdma_chan *bdma_chan = to_tsi721_chan(dchan);
+ struct tsi721_tx_desc *desc, *_d;
+ LIST_HEAD(list);
+
+ dev_dbg(dchan->device->dev, "%s: Entry\n", __func__);
+
+ if (cmd != DMA_TERMINATE_ALL)
+ return -ENXIO;
+
+ spin_lock_bh(&bdma_chan->lock);
+
+ /* make sure to stop the transfer */
+ iowrite32(TSI721_DMAC_CTL_SUSP, bdma_chan->regs + TSI721_DMAC_CTL);
+
+ list_splice_init(&bdma_chan->active_list, &list);
+ list_splice_init(&bdma_chan->queue, &list);
+
+ list_for_each_entry_safe(desc, _d, &list, desc_node)
+ tsi721_dma_chain_complete(bdma_chan, desc);
+
+ spin_unlock_bh(&bdma_chan->lock);
+
+ return 0;
+}
+
+int __devinit tsi721_register_dma(struct tsi721_device *priv)
+{
+ int i;
+ int nr_channels = TSI721_DMA_MAXCH;
+ int err;
+ struct rio_mport *mport = priv->mport;
+
+ mport->dma.dev = &priv->pdev->dev;
+ mport->dma.chancnt = nr_channels;
+
+ INIT_LIST_HEAD(&mport->dma.channels);
+
+ for (i = 0; i < nr_channels; i++) {
+ struct tsi721_bdma_chan *bdma_chan = &priv->bdma[i];
+
+ if (i == TSI721_DMACH_MAINT)
+ continue;
+
+ bdma_chan->bd_num = 64;
+ bdma_chan->regs = priv->regs + TSI721_DMAC_BASE(i);
+
+ bdma_chan->dchan.device = &mport->dma;
+ bdma_chan->dchan.cookie = 1;
+ bdma_chan->dchan.chan_id = i;
+ bdma_chan->id = i;
+
+ spin_lock_init(&bdma_chan->lock);
+
+ INIT_LIST_HEAD(&bdma_chan->active_list);
+ INIT_LIST_HEAD(&bdma_chan->queue);
+ INIT_LIST_HEAD(&bdma_chan->free_list);
+
+ tasklet_init(&bdma_chan->tasklet, tsi721_dma_tasklet,
+ (unsigned long)bdma_chan);
+ tasklet_disable(&bdma_chan->tasklet);
+ list_add_tail(&bdma_chan->dchan.device_node,
+ &mport->dma.channels);
+ }
+
+ dma_cap_zero(mport->dma.cap_mask);
+ dma_cap_set(DMA_PRIVATE, mport->dma.cap_mask);
+ dma_cap_set(DMA_SLAVE, mport->dma.cap_mask);
+
+ mport->dma.device_alloc_chan_resources = tsi721_alloc_chan_resources;
+ mport->dma.device_free_chan_resources = tsi721_free_chan_resources;
+ mport->dma.device_tx_status = tsi721_tx_status;
+ mport->dma.device_issue_pending = tsi721_issue_pending;
+ mport->dma.device_prep_slave_sg = tsi721_prep_rio_sg;
+ mport->dma.device_control = tsi721_device_control;
+
+ err = dma_async_device_register(&mport->dma);
+ if (err)
+ dev_err(&priv->pdev->dev, "Failed to register DMA device\n");
+
+ return err;
+}
diff --git a/drivers/rapidio/rio.c b/drivers/rapidio/rio.c
index 86c9a091a2f..c40665a4fa3 100644
--- a/drivers/rapidio/rio.c
+++ b/drivers/rapidio/rio.c
@@ -1121,6 +1121,87 @@ int rio_std_route_clr_table(struct rio_mport *mport, u16 destid, u8 hopcount,
return 0;
}
+#ifdef CONFIG_RAPIDIO_DMA_ENGINE
+
+static bool rio_chan_filter(struct dma_chan *chan, void *arg)
+{
+ struct rio_dev *rdev = arg;
+
+ /* Check that DMA device belongs to the right MPORT */
+ return (rdev->net->hport ==
+ container_of(chan->device, struct rio_mport, dma));
+}
+
+/**
+ * rio_request_dma - request RapidIO capable DMA channel that supports
+ * specified target RapidIO device.
+ * @rdev: RIO device control structure
+ *
+ * Returns pointer to allocated DMA channel or NULL if failed.
+ */
+struct dma_chan *rio_request_dma(struct rio_dev *rdev)
+{
+ dma_cap_mask_t mask;
+ struct dma_chan *dchan;
+
+ dma_cap_zero(mask);
+ dma_cap_set(DMA_SLAVE, mask);
+ dchan = dma_request_channel(mask, rio_chan_filter, rdev);
+
+ return dchan;
+}
+EXPORT_SYMBOL_GPL(rio_request_dma);
+
+/**
+ * rio_release_dma - release specified DMA channel
+ * @dchan: DMA channel to release
+ */
+void rio_release_dma(struct dma_chan *dchan)
+{
+ dma_release_channel(dchan);
+}
+EXPORT_SYMBOL_GPL(rio_release_dma);
+
+/**
+ * rio_dma_prep_slave_sg - RapidIO specific wrapper
+ * for device_prep_slave_sg callback defined by DMAENGINE.
+ * @rdev: RIO device control structure
+ * @dchan: DMA channel to configure
+ * @data: RIO specific data descriptor
+ * @direction: DMA data transfer direction (TO or FROM the device)
+ * @flags: dmaengine defined flags
+ *
+ * Initializes RapidIO capable DMA channel for the specified data transfer.
+ * Uses DMA channel private extension to pass information related to remote
+ * target RIO device.
+ * Returns pointer to DMA transaction descriptor or NULL if failed.
+ */
+struct dma_async_tx_descriptor *rio_dma_prep_slave_sg(struct rio_dev *rdev,
+ struct dma_chan *dchan, struct rio_dma_data *data,
+ enum dma_transfer_direction direction, unsigned long flags)
+{
+ struct dma_async_tx_descriptor *txd = NULL;
+ struct rio_dma_ext rio_ext;
+
+ if (dchan->device->device_prep_slave_sg == NULL) {
+ pr_err("%s: prep_rio_sg == NULL\n", __func__);
+ return NULL;
+ }
+
+ rio_ext.destid = rdev->destid;
+ rio_ext.rio_addr_u = data->rio_addr_u;
+ rio_ext.rio_addr = data->rio_addr;
+ rio_ext.wr_type = data->wr_type;
+
+ txd = dmaengine_prep_rio_sg(dchan, data->sg, data->sg_len,
+ direction, flags, &rio_ext);
+
+ return txd;
+}
+EXPORT_SYMBOL_GPL(rio_dma_prep_slave_sg);
+
+#endif /* CONFIG_RAPIDIO_DMA_ENGINE */
+
static void rio_fixup_device(struct rio_dev *dev)
{
}
diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c
index a1e6c990cd4..e3dd2a1e2bf 100644
--- a/fs/9p/vfs_inode_dotl.c
+++ b/fs/9p/vfs_inode_dotl.c
@@ -68,24 +68,6 @@ static gid_t v9fs_get_fsgid_for_create(struct inode *dir_inode)
return current_fsgid();
}
-/**
- * v9fs_dentry_from_dir_inode - helper function to get the dentry from
- * dir inode.
- *
- */
-
-static struct dentry *v9fs_dentry_from_dir_inode(struct inode *inode)
-{
- struct dentry *dentry;
-
- spin_lock(&inode->i_lock);
- /* Directory should have only one entry. */
- BUG_ON(S_ISDIR(inode->i_mode) && !list_is_singular(&inode->i_dentry));
- dentry = list_entry(inode->i_dentry.next, struct dentry, d_alias);
- spin_unlock(&inode->i_lock);
- return dentry;
-}
-
static int v9fs_test_inode_dotl(struct inode *inode, void *data)
{
struct v9fs_inode *v9inode = V9FS_I(inode);
@@ -415,7 +397,7 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir,
if (dir->i_mode & S_ISGID)
omode |= S_ISGID;
- dir_dentry = v9fs_dentry_from_dir_inode(dir);
+ dir_dentry = dentry->d_parent;
dfid = v9fs_fid_lookup(dir_dentry);
if (IS_ERR(dfid)) {
err = PTR_ERR(dfid);
@@ -793,7 +775,7 @@ v9fs_vfs_link_dotl(struct dentry *old_dentry, struct inode *dir,
dir->i_ino, old_dentry->d_name.name, dentry->d_name.name);
v9ses = v9fs_inode2v9ses(dir);
- dir_dentry = v9fs_dentry_from_dir_inode(dir);
+ dir_dentry = dentry->d_parent;
dfid = v9fs_fid_lookup(dir_dentry);
if (IS_ERR(dfid))
return PTR_ERR(dfid);
@@ -858,7 +840,7 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, umode_t omode,
return -EINVAL;
v9ses = v9fs_inode2v9ses(dir);
- dir_dentry = v9fs_dentry_from_dir_inode(dir);
+ dir_dentry = dentry->d_parent;
dfid = v9fs_fid_lookup(dir_dentry);
if (IS_ERR(dfid)) {
err = PTR_ERR(dfid);
diff --git a/fs/affs/affs.h b/fs/affs/affs.h
index 45a0ce45d7b..1fceb320d2f 100644
--- a/fs/affs/affs.h
+++ b/fs/affs/affs.h
@@ -18,14 +18,6 @@
#define AFFS_GET_HASHENTRY(data,hashkey) be32_to_cpu(((struct dir_front *)data)->hashtable[hashkey])
#define AFFS_BLOCK(sb, bh, blk) (AFFS_HEAD(bh)->table[AFFS_SB(sb)->s_hashsize-1-(blk)])
-#ifdef __LITTLE_ENDIAN
-#define BO_EXBITS 0x18UL
-#elif defined(__BIG_ENDIAN)
-#define BO_EXBITS 0x00UL
-#else
-#error Endianness must be known for affs to work.
-#endif
-
#define AFFS_HEAD(bh) ((struct affs_head *)(bh)->b_data)
#define AFFS_TAIL(sb, bh) ((struct affs_tail *)((bh)->b_data+(sb)->s_blocksize-sizeof(struct affs_tail)))
#define AFFS_ROOT_HEAD(bh) ((struct affs_root_head *)(bh)->b_data)
diff --git a/fs/aio.c b/fs/aio.c
index e7f2fad7b4c..55c4c765605 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -134,9 +134,9 @@ static int aio_setup_ring(struct kioctx *ctx)
info->mmap_size = nr_pages * PAGE_SIZE;
dprintk("attempting mmap of %lu bytes\n", info->mmap_size);
down_write(&ctx->mm->mmap_sem);
- info->mmap_base = do_mmap(NULL, 0, info->mmap_size,
- PROT_READ|PROT_WRITE, MAP_ANONYMOUS|MAP_PRIVATE,
- 0);
+ info->mmap_base = do_mmap_pgoff(NULL, 0, info->mmap_size,
+ PROT_READ|PROT_WRITE,
+ MAP_ANONYMOUS|MAP_PRIVATE, 0);
if (IS_ERR((void *)info->mmap_base)) {
up_write(&ctx->mm->mmap_sem);
info->mmap_size = 0;
@@ -1446,13 +1446,13 @@ static ssize_t aio_setup_vectored_rw(int type, struct kiocb *kiocb, bool compat)
ret = compat_rw_copy_check_uvector(type,
(struct compat_iovec __user *)kiocb->ki_buf,
kiocb->ki_nbytes, 1, &kiocb->ki_inline_vec,
- &kiocb->ki_iovec, 1);
+ &kiocb->ki_iovec);
else
#endif
ret = rw_copy_check_uvector(type,
(struct iovec __user *)kiocb->ki_buf,
kiocb->ki_nbytes, 1, &kiocb->ki_inline_vec,
- &kiocb->ki_iovec, 1);
+ &kiocb->ki_iovec);
if (ret < 0)
goto out;
diff --git a/fs/attr.c b/fs/attr.c
index 584620e5dee..0da90951d27 100644
--- a/fs/attr.c
+++ b/fs/attr.c
@@ -176,6 +176,11 @@ int notify_change(struct dentry * dentry, struct iattr * attr)
return -EPERM;
}
+ if ((ia_valid & ATTR_SIZE) && IS_I_VERSION(inode)) {
+ if (attr->ia_size != inode->i_size)
+ inode_inc_iversion(inode);
+ }
+
if ((ia_valid & ATTR_MODE)) {
umode_t amode = attr->ia_mode;
/* Flag setting protected by i_mutex */
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index e658dd134b9..1b52956afe3 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -329,7 +329,6 @@ static unsigned long elf_map(struct file *filep, unsigned long addr,
if (!size)
return addr;
- down_write(&current->mm->mmap_sem);
/*
* total_size is the size of the ELF (interpreter) image.
* The _first_ mmap needs to know the full size, otherwise
@@ -340,13 +339,12 @@ static unsigned long elf_map(struct file *filep, unsigned long addr,
*/
if (total_size) {
total_size = ELF_PAGEALIGN(total_size);
- map_addr = do_mmap(filep, addr, total_size, prot, type, off);
+ map_addr = vm_mmap(filep, addr, total_size, prot, type, off);
if (!BAD_ADDR(map_addr))
- do_munmap(current->mm, map_addr+size, total_size-size);
+ vm_munmap(map_addr+size, total_size-size);
} else
- map_addr = do_mmap(filep, addr, size, prot, type, off);
+ map_addr = vm_mmap(filep, addr, size, prot, type, off);
- up_write(&current->mm->mmap_sem);
return(map_addr);
}
diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c
index 6b2daf99fab..178cb70acc2 100644
--- a/fs/binfmt_flat.c
+++ b/fs/binfmt_flat.c
@@ -562,7 +562,7 @@ static int load_flat_file(struct linux_binprm * bprm,
realdatastart = (unsigned long) -ENOMEM;
printk("Unable to allocate RAM for process data, errno %d\n",
(int)-realdatastart);
- do_munmap(current->mm, textpos, text_len);
+ vm_munmap(textpos, text_len);
ret = realdatastart;
goto err;
}
@@ -586,8 +586,8 @@ static int load_flat_file(struct linux_binprm * bprm,
}
if (IS_ERR_VALUE(result)) {
printk("Unable to read data+bss, errno %d\n", (int)-result);
- do_munmap(current->mm, textpos, text_len);
- do_munmap(current->mm, realdatastart, len);
+ vm_munmap(textpos, text_len);
+ vm_munmap(realdatastart, len);
ret = result;
goto err;
}
@@ -654,7 +654,7 @@ static int load_flat_file(struct linux_binprm * bprm,
}
if (IS_ERR_VALUE(result)) {
printk("Unable to read code+data+bss, errno %d\n",(int)-result);
- do_munmap(current->mm, textpos, text_len + data_len + extra +
+ vm_munmap(textpos, text_len + data_len + extra +
MAX_SHARED_LIBS * sizeof(unsigned long));
ret = result;
goto err;
diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c
index 89b156d85d6..761e2cd8fed 100644
--- a/fs/btrfs/acl.c
+++ b/fs/btrfs/acl.c
@@ -227,7 +227,11 @@ int btrfs_init_acl(struct btrfs_trans_handle *trans,
if (ret > 0) {
/* we need an acl */
ret = btrfs_set_acl(trans, inode, acl, ACL_TYPE_ACCESS);
+ } else {
+ cache_no_acl(inode);
}
+ } else {
+ cache_no_acl(inode);
}
failed:
posix_acl_release(acl);
diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c
index bcec0675023..3f75895c919 100644
--- a/fs/btrfs/backref.c
+++ b/fs/btrfs/backref.c
@@ -24,22 +24,135 @@
#include "delayed-ref.h"
#include "locking.h"
+struct extent_inode_elem {
+ u64 inum;
+ u64 offset;
+ struct extent_inode_elem *next;
+};
+
+static int check_extent_in_eb(struct btrfs_key *key, struct extent_buffer *eb,
+ struct btrfs_file_extent_item *fi,
+ u64 extent_item_pos,
+ struct extent_inode_elem **eie)
+{
+ u64 data_offset;
+ u64 data_len;
+ struct extent_inode_elem *e;
+
+ data_offset = btrfs_file_extent_offset(eb, fi);
+ data_len = btrfs_file_extent_num_bytes(eb, fi);
+
+ if (extent_item_pos < data_offset ||
+ extent_item_pos >= data_offset + data_len)
+ return 1;
+
+ e = kmalloc(sizeof(*e), GFP_NOFS);
+ if (!e)
+ return -ENOMEM;
+
+ e->next = *eie;
+ e->inum = key->objectid;
+ e->offset = key->offset + (extent_item_pos - data_offset);
+ *eie = e;
+
+ return 0;
+}
+
+static int find_extent_in_eb(struct extent_buffer *eb, u64 wanted_disk_byte,
+ u64 extent_item_pos,
+ struct extent_inode_elem **eie)
+{
+ u64 disk_byte;
+ struct btrfs_key key;
+ struct btrfs_file_extent_item *fi;
+ int slot;
+ int nritems;
+ int extent_type;
+ int ret;
+
+ /*
+ * from the shared data ref, we only have the leaf but we need
+ * the key. thus, we must look into all items and see that we
+ * find one (some) with a reference to our extent item.
+ */
+ nritems = btrfs_header_nritems(eb);
+ for (slot = 0; slot < nritems; ++slot) {
+ btrfs_item_key_to_cpu(eb, &key, slot);
+ if (key.type != BTRFS_EXTENT_DATA_KEY)
+ continue;
+ fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
+ extent_type = btrfs_file_extent_type(eb, fi);
+ if (extent_type == BTRFS_FILE_EXTENT_INLINE)
+ continue;
+ /* don't skip BTRFS_FILE_EXTENT_PREALLOC, we can handle that */
+ disk_byte = btrfs_file_extent_disk_bytenr(eb, fi);
+ if (disk_byte != wanted_disk_byte)
+ continue;
+
+ ret = check_extent_in_eb(&key, eb, fi, extent_item_pos, eie);
+ if (ret < 0)
+ return ret;
+ }
+
+ return 0;
+}
+
/*
* this structure records all encountered refs on the way up to the root
*/
struct __prelim_ref {
struct list_head list;
u64 root_id;
- struct btrfs_key key;
+ struct btrfs_key key_for_search;
int level;
int count;
+ struct extent_inode_elem *inode_list;
u64 parent;
u64 wanted_disk_byte;
};
+/*
+ * the rules for all callers of this function are:
+ * - obtaining the parent is the goal
+ * - if you add a key, you must know that it is a correct key
+ * - if you cannot add the parent or a correct key, then we will look into the
+ * block later to set a correct key
+ *
+ * delayed refs
+ * ============
+ * backref type | shared | indirect | shared | indirect
+ * information | tree | tree | data | data
+ * --------------------+--------+----------+--------+----------
+ * parent logical | y | - | - | -
+ * key to resolve | - | y | y | y
+ * tree block logical | - | - | - | -
+ * root for resolving | y | y | y | y
+ *
+ * - column 1: we've the parent -> done
+ * - column 2, 3, 4: we use the key to find the parent
+ *
+ * on disk refs (inline or keyed)
+ * ==============================
+ * backref type | shared | indirect | shared | indirect
+ * information | tree | tree | data | data
+ * --------------------+--------+----------+--------+----------
+ * parent logical | y | - | y | -
+ * key to resolve | - | - | - | y
+ * tree block logical | y | y | y | y
+ * root for resolving | - | y | y | y
+ *
+ * - column 1, 3: we've the parent -> done
+ * - column 2: we take the first key from the block to find the parent
+ * (see __add_missing_keys)
+ * - column 4: we use the key to find the parent
+ *
+ * additional information that's available but not required to find the parent
+ * block might help in merging entries to gain some speed.
+ */
+
static int __add_prelim_ref(struct list_head *head, u64 root_id,
- struct btrfs_key *key, int level, u64 parent,
- u64 wanted_disk_byte, int count)
+ struct btrfs_key *key, int level,
+ u64 parent, u64 wanted_disk_byte, int count)
{
struct __prelim_ref *ref;
@@ -50,10 +163,11 @@ static int __add_prelim_ref(struct list_head *head, u64 root_id,
ref->root_id = root_id;
if (key)
- ref->key = *key;
+ ref->key_for_search = *key;
else
- memset(&ref->key, 0, sizeof(ref->key));
+ memset(&ref->key_for_search, 0, sizeof(ref->key_for_search));
+ ref->inode_list = NULL;
ref->level = level;
ref->count = count;
ref->parent = parent;
@@ -64,18 +178,26 @@ static int __add_prelim_ref(struct list_head *head, u64 root_id,
}
static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path,
- struct ulist *parents,
- struct extent_buffer *eb, int level,
- u64 wanted_objectid, u64 wanted_disk_byte)
+ struct ulist *parents, int level,
+ struct btrfs_key *key, u64 wanted_disk_byte,
+ const u64 *extent_item_pos)
{
int ret;
- int slot;
+ int slot = path->slots[level];
+ struct extent_buffer *eb = path->nodes[level];
struct btrfs_file_extent_item *fi;
- struct btrfs_key key;
+ struct extent_inode_elem *eie = NULL;
u64 disk_byte;
+ u64 wanted_objectid = key->objectid;
add_parent:
- ret = ulist_add(parents, eb->start, 0, GFP_NOFS);
+ if (level == 0 && extent_item_pos) {
+ fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
+ ret = check_extent_in_eb(key, eb, fi, *extent_item_pos, &eie);
+ if (ret < 0)
+ return ret;
+ }
+ ret = ulist_add(parents, eb->start, (unsigned long)eie, GFP_NOFS);
if (ret < 0)
return ret;
@@ -89,6 +211,7 @@ add_parent:
* repeat this until we don't find any additional EXTENT_DATA items.
*/
while (1) {
+ eie = NULL;
ret = btrfs_next_leaf(root, path);
if (ret < 0)
return ret;
@@ -97,9 +220,9 @@ add_parent:
eb = path->nodes[0];
for (slot = 0; slot < btrfs_header_nritems(eb); ++slot) {
- btrfs_item_key_to_cpu(eb, &key, slot);
- if (key.objectid != wanted_objectid ||
- key.type != BTRFS_EXTENT_DATA_KEY)
+ btrfs_item_key_to_cpu(eb, key, slot);
+ if (key->objectid != wanted_objectid ||
+ key->type != BTRFS_EXTENT_DATA_KEY)
return 0;
fi = btrfs_item_ptr(eb, slot,
struct btrfs_file_extent_item);
@@ -118,8 +241,10 @@ add_parent:
*/
static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
int search_commit_root,
+ u64 time_seq,
struct __prelim_ref *ref,
- struct ulist *parents)
+ struct ulist *parents,
+ const u64 *extent_item_pos)
{
struct btrfs_path *path;
struct btrfs_root *root;
@@ -152,12 +277,13 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
goto out;
path->lowest_level = level;
- ret = btrfs_search_slot(NULL, root, &ref->key, path, 0, 0);
+ ret = btrfs_search_old_slot(root, &ref->key_for_search, path, time_seq);
pr_debug("search slot in root %llu (level %d, ref count %d) returned "
"%d for key (%llu %u %llu)\n",
(unsigned long long)ref->root_id, level, ref->count, ret,
- (unsigned long long)ref->key.objectid, ref->key.type,
- (unsigned long long)ref->key.offset);
+ (unsigned long long)ref->key_for_search.objectid,
+ ref->key_for_search.type,
+ (unsigned long long)ref->key_for_search.offset);
if (ret < 0)
goto out;
@@ -179,9 +305,8 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info,
btrfs_item_key_to_cpu(eb, &key, path->slots[0]);
}
- /* the last two parameters will only be used for level == 0 */
- ret = add_all_parents(root, path, parents, eb, level, key.objectid,
- ref->wanted_disk_byte);
+ ret = add_all_parents(root, path, parents, level, &key,
+ ref->wanted_disk_byte, extent_item_pos);
out:
btrfs_free_path(path);
return ret;
@@ -191,8 +316,9 @@ out:
* resolve all indirect backrefs from the list
*/
static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
- int search_commit_root,
- struct list_head *head)
+ int search_commit_root, u64 time_seq,
+ struct list_head *head,
+ const u64 *extent_item_pos)
{
int err;
int ret = 0;
@@ -201,6 +327,7 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
struct __prelim_ref *new_ref;
struct ulist *parents;
struct ulist_node *node;
+ struct ulist_iterator uiter;
parents = ulist_alloc(GFP_NOFS);
if (!parents)
@@ -217,7 +344,8 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
if (ref->count == 0)
continue;
err = __resolve_indirect_ref(fs_info, search_commit_root,
- ref, parents);
+ time_seq, ref, parents,
+ extent_item_pos);
if (err) {
if (ret == 0)
ret = err;
@@ -225,11 +353,14 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
}
/* we put the first parent into the ref at hand */
- node = ulist_next(parents, NULL);
+ ULIST_ITER_INIT(&uiter);
+ node = ulist_next(parents, &uiter);
ref->parent = node ? node->val : 0;
+ ref->inode_list =
+ node ? (struct extent_inode_elem *)node->aux : 0;
/* additional parents require new refs being added here */
- while ((node = ulist_next(parents, node))) {
+ while ((node = ulist_next(parents, &uiter))) {
new_ref = kmalloc(sizeof(*new_ref), GFP_NOFS);
if (!new_ref) {
ret = -ENOMEM;
@@ -237,6 +368,8 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
}
memcpy(new_ref, ref, sizeof(*ref));
new_ref->parent = node->val;
+ new_ref->inode_list =
+ (struct extent_inode_elem *)node->aux;
list_add(&new_ref->list, &ref->list);
}
ulist_reinit(parents);
@@ -246,10 +379,65 @@ static int __resolve_indirect_refs(struct btrfs_fs_info *fs_info,
return ret;
}
+static inline int ref_for_same_block(struct __prelim_ref *ref1,
+ struct __prelim_ref *ref2)
+{
+ if (ref1->level != ref2->level)
+ return 0;
+ if (ref1->root_id != ref2->root_id)
+ return 0;
+ if (ref1->key_for_search.type != ref2->key_for_search.type)
+ return 0;
+ if (ref1->key_for_search.objectid != ref2->key_for_search.objectid)
+ return 0;
+ if (ref1->key_for_search.offset != ref2->key_for_search.offset)
+ return 0;
+ if (ref1->parent != ref2->parent)
+ return 0;
+
+ return 1;
+}
+
+/*
+ * read tree blocks and add keys where required.
+ */
+static int __add_missing_keys(struct btrfs_fs_info *fs_info,
+ struct list_head *head)
+{
+ struct list_head *pos;
+ struct extent_buffer *eb;
+
+ list_for_each(pos, head) {
+ struct __prelim_ref *ref;
+ ref = list_entry(pos, struct __prelim_ref, list);
+
+ if (ref->parent)
+ continue;
+ if (ref->key_for_search.type)
+ continue;
+ BUG_ON(!ref->wanted_disk_byte);
+ eb = read_tree_block(fs_info->tree_root, ref->wanted_disk_byte,
+ fs_info->tree_root->leafsize, 0);
+ BUG_ON(!eb);
+ btrfs_tree_read_lock(eb);
+ if (btrfs_header_level(eb) == 0)
+ btrfs_item_key_to_cpu(eb, &ref->key_for_search, 0);
+ else
+ btrfs_node_key_to_cpu(eb, &ref->key_for_search, 0);
+ btrfs_tree_read_unlock(eb);
+ free_extent_buffer(eb);
+ }
+ return 0;
+}
+
/*
* merge two lists of backrefs and adjust counts accordingly
*
* mode = 1: merge identical keys, if key is set
+ * FIXME: if we add more keys in __add_prelim_ref, we can merge more here.
+ * additionally, we could even add a key range for the blocks we
+ * looked into to merge even more (-> replace unresolved refs by those
+ * having a parent).
* mode = 2: merge identical parents
*/
static int __merge_refs(struct list_head *head, int mode)
@@ -263,20 +451,21 @@ static int __merge_refs(struct list_head *head, int mode)
ref1 = list_entry(pos1, struct __prelim_ref, list);
- if (mode == 1 && ref1->key.type == 0)
- continue;
for (pos2 = pos1->next, n2 = pos2->next; pos2 != head;
pos2 = n2, n2 = pos2->next) {
struct __prelim_ref *ref2;
+ struct __prelim_ref *xchg;
ref2 = list_entry(pos2, struct __prelim_ref, list);
if (mode == 1) {
- if (memcmp(&ref1->key, &ref2->key,
- sizeof(ref1->key)) ||
- ref1->level != ref2->level ||
- ref1->root_id != ref2->root_id)
+ if (!ref_for_same_block(ref1, ref2))
continue;
+ if (!ref1->parent && ref2->parent) {
+ xchg = ref1;
+ ref1 = ref2;
+ ref2 = xchg;
+ }
ref1->count += ref2->count;
} else {
if (ref1->parent != ref2->parent)
@@ -296,16 +485,17 @@ static int __merge_refs(struct list_head *head, int mode)
* smaller or equal that seq to the list
*/
static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
- struct btrfs_key *info_key,
struct list_head *prefs)
{
struct btrfs_delayed_extent_op *extent_op = head->extent_op;
struct rb_node *n = &head->node.rb_node;
+ struct btrfs_key key;
+ struct btrfs_key op_key = {0};
int sgn;
int ret = 0;
if (extent_op && extent_op->update_key)
- btrfs_disk_key_to_cpu(info_key, &extent_op->key);
+ btrfs_disk_key_to_cpu(&op_key, &extent_op->key);
while ((n = rb_prev(n))) {
struct btrfs_delayed_ref_node *node;
@@ -337,7 +527,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
struct btrfs_delayed_tree_ref *ref;
ref = btrfs_delayed_node_to_tree_ref(node);
- ret = __add_prelim_ref(prefs, ref->root, info_key,
+ ret = __add_prelim_ref(prefs, ref->root, &op_key,
ref->level + 1, 0, node->bytenr,
node->ref_mod * sgn);
break;
@@ -346,7 +536,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
struct btrfs_delayed_tree_ref *ref;
ref = btrfs_delayed_node_to_tree_ref(node);
- ret = __add_prelim_ref(prefs, ref->root, info_key,
+ ret = __add_prelim_ref(prefs, ref->root, NULL,
ref->level + 1, ref->parent,
node->bytenr,
node->ref_mod * sgn);
@@ -354,8 +544,6 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
}
case BTRFS_EXTENT_DATA_REF_KEY: {
struct btrfs_delayed_data_ref *ref;
- struct btrfs_key key;
-
ref = btrfs_delayed_node_to_data_ref(node);
key.objectid = ref->objectid;
@@ -368,7 +556,6 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
}
case BTRFS_SHARED_DATA_REF_KEY: {
struct btrfs_delayed_data_ref *ref;
- struct btrfs_key key;
ref = btrfs_delayed_node_to_data_ref(node);
@@ -394,8 +581,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq,
*/
static int __add_inline_refs(struct btrfs_fs_info *fs_info,
struct btrfs_path *path, u64 bytenr,
- struct btrfs_key *info_key, int *info_level,
- struct list_head *prefs)
+ int *info_level, struct list_head *prefs)
{
int ret = 0;
int slot;
@@ -411,7 +597,7 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
* enumerate all inline refs
*/
leaf = path->nodes[0];
- slot = path->slots[0] - 1;
+ slot = path->slots[0];
item_size = btrfs_item_size_nr(leaf, slot);
BUG_ON(item_size < sizeof(*ei));
@@ -424,12 +610,9 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) {
struct btrfs_tree_block_info *info;
- struct btrfs_disk_key disk_key;
info = (struct btrfs_tree_block_info *)ptr;
*info_level = btrfs_tree_block_level(leaf, info);
- btrfs_tree_block_key(leaf, info, &disk_key);
- btrfs_disk_key_to_cpu(info_key, &disk_key);
ptr += sizeof(struct btrfs_tree_block_info);
BUG_ON(ptr > end);
} else {
@@ -447,7 +630,7 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
switch (type) {
case BTRFS_SHARED_BLOCK_REF_KEY:
- ret = __add_prelim_ref(prefs, 0, info_key,
+ ret = __add_prelim_ref(prefs, 0, NULL,
*info_level + 1, offset,
bytenr, 1);
break;
@@ -462,8 +645,9 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
break;
}
case BTRFS_TREE_BLOCK_REF_KEY:
- ret = __add_prelim_ref(prefs, offset, info_key,
- *info_level + 1, 0, bytenr, 1);
+ ret = __add_prelim_ref(prefs, offset, NULL,
+ *info_level + 1, 0,
+ bytenr, 1);
break;
case BTRFS_EXTENT_DATA_REF_KEY: {
struct btrfs_extent_data_ref *dref;
@@ -477,8 +661,8 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
key.type = BTRFS_EXTENT_DATA_KEY;
key.offset = btrfs_extent_data_ref_offset(leaf, dref);
root = btrfs_extent_data_ref_root(leaf, dref);
- ret = __add_prelim_ref(prefs, root, &key, 0, 0, bytenr,
- count);
+ ret = __add_prelim_ref(prefs, root, &key, 0, 0,
+ bytenr, count);
break;
}
default:
@@ -496,8 +680,7 @@ static int __add_inline_refs(struct btrfs_fs_info *fs_info,
*/
static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
struct btrfs_path *path, u64 bytenr,
- struct btrfs_key *info_key, int info_level,
- struct list_head *prefs)
+ int info_level, struct list_head *prefs)
{
struct btrfs_root *extent_root = fs_info->extent_root;
int ret;
@@ -527,7 +710,7 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
switch (key.type) {
case BTRFS_SHARED_BLOCK_REF_KEY:
- ret = __add_prelim_ref(prefs, 0, info_key,
+ ret = __add_prelim_ref(prefs, 0, NULL,
info_level + 1, key.offset,
bytenr, 1);
break;
@@ -543,8 +726,9 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
break;
}
case BTRFS_TREE_BLOCK_REF_KEY:
- ret = __add_prelim_ref(prefs, key.offset, info_key,
- info_level + 1, 0, bytenr, 1);
+ ret = __add_prelim_ref(prefs, key.offset, NULL,
+ info_level + 1, 0,
+ bytenr, 1);
break;
case BTRFS_EXTENT_DATA_REF_KEY: {
struct btrfs_extent_data_ref *dref;
@@ -560,7 +744,7 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
key.offset = btrfs_extent_data_ref_offset(leaf, dref);
root = btrfs_extent_data_ref_root(leaf, dref);
ret = __add_prelim_ref(prefs, root, &key, 0, 0,
- bytenr, count);
+ bytenr, count);
break;
}
default:
@@ -582,11 +766,12 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info,
*/
static int find_parent_nodes(struct btrfs_trans_handle *trans,
struct btrfs_fs_info *fs_info, u64 bytenr,
- u64 seq, struct ulist *refs, struct ulist *roots)
+ u64 delayed_ref_seq, u64 time_seq,
+ struct ulist *refs, struct ulist *roots,
+ const u64 *extent_item_pos)
{
struct btrfs_key key;
struct btrfs_path *path;
- struct btrfs_key info_key = { 0 };
struct btrfs_delayed_ref_root *delayed_refs = NULL;
struct btrfs_delayed_ref_head *head;
int info_level = 0;
@@ -645,7 +830,7 @@ again:
btrfs_put_delayed_ref(&head->node);
goto again;
}
- ret = __add_delayed_refs(head, seq, &info_key,
+ ret = __add_delayed_refs(head, delayed_ref_seq,
&prefs_delayed);
if (ret) {
spin_unlock(&delayed_refs->lock);
@@ -659,16 +844,17 @@ again:
struct extent_buffer *leaf;
int slot;
+ path->slots[0]--;
leaf = path->nodes[0];
- slot = path->slots[0] - 1;
+ slot = path->slots[0];
btrfs_item_key_to_cpu(leaf, &key, slot);
if (key.objectid == bytenr &&
key.type == BTRFS_EXTENT_ITEM_KEY) {
ret = __add_inline_refs(fs_info, path, bytenr,
- &info_key, &info_level, &prefs);
+ &info_level, &prefs);
if (ret)
goto out;
- ret = __add_keyed_refs(fs_info, path, bytenr, &info_key,
+ ret = __add_keyed_refs(fs_info, path, bytenr,
info_level, &prefs);
if (ret)
goto out;
@@ -676,21 +862,18 @@ again:
}
btrfs_release_path(path);
- /*
- * when adding the delayed refs above, the info_key might not have
- * been known yet. Go over the list and replace the missing keys
- */
- list_for_each_entry(ref, &prefs_delayed, list) {
- if ((ref->key.offset | ref->key.type | ref->key.objectid) == 0)
- memcpy(&ref->key, &info_key, sizeof(ref->key));
- }
list_splice_init(&prefs_delayed, &prefs);
+ ret = __add_missing_keys(fs_info, &prefs);
+ if (ret)
+ goto out;
+
ret = __merge_refs(&prefs, 1);
if (ret)
goto out;
- ret = __resolve_indirect_refs(fs_info, search_commit_root, &prefs);
+ ret = __resolve_indirect_refs(fs_info, search_commit_root, time_seq,
+ &prefs, extent_item_pos);
if (ret)
goto out;
@@ -709,7 +892,33 @@ again:
BUG_ON(ret < 0);
}
if (ref->count && ref->parent) {
- ret = ulist_add(refs, ref->parent, 0, GFP_NOFS);
+ struct extent_inode_elem *eie = NULL;
+ if (extent_item_pos && !ref->inode_list) {
+ u32 bsz;
+ struct extent_buffer *eb;
+ bsz = btrfs_level_size(fs_info->extent_root,
+ info_level);
+ eb = read_tree_block(fs_info->extent_root,
+ ref->parent, bsz, 0);
+ BUG_ON(!eb);
+ ret = find_extent_in_eb(eb, bytenr,
+ *extent_item_pos, &eie);
+ ref->inode_list = eie;
+ free_extent_buffer(eb);
+ }
+ ret = ulist_add_merge(refs, ref->parent,
+ (unsigned long)ref->inode_list,
+ (unsigned long *)&eie, GFP_NOFS);
+ if (!ret && extent_item_pos) {
+ /*
+ * we've recorded that parent, so we must extend
+ * its inode list here
+ */
+ BUG_ON(!eie);
+ while (eie->next)
+ eie = eie->next;
+ eie->next = ref->inode_list;
+ }
BUG_ON(ret < 0);
}
kfree(ref);
@@ -734,6 +943,28 @@ out:
return ret;
}
+static void free_leaf_list(struct ulist *blocks)
+{
+ struct ulist_node *node = NULL;
+ struct extent_inode_elem *eie;
+ struct extent_inode_elem *eie_next;
+ struct ulist_iterator uiter;
+
+ ULIST_ITER_INIT(&uiter);
+ while ((node = ulist_next(blocks, &uiter))) {
+ if (!node->aux)
+ continue;
+ eie = (struct extent_inode_elem *)node->aux;
+ for (; eie; eie = eie_next) {
+ eie_next = eie->next;
+ kfree(eie);
+ }
+ node->aux = 0;
+ }
+
+ ulist_free(blocks);
+}
+
/*
* Finds all leafs with a reference to the specified combination of bytenr and
* offset. key_list_head will point to a list of corresponding keys (caller must
@@ -744,7 +975,9 @@ out:
*/
static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans,
struct btrfs_fs_info *fs_info, u64 bytenr,
- u64 num_bytes, u64 seq, struct ulist **leafs)
+ u64 delayed_ref_seq, u64 time_seq,
+ struct ulist **leafs,
+ const u64 *extent_item_pos)
{
struct ulist *tmp;
int ret;
@@ -758,11 +991,12 @@ static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans,
return -ENOMEM;
}
- ret = find_parent_nodes(trans, fs_info, bytenr, seq, *leafs, tmp);
+ ret = find_parent_nodes(trans, fs_info, bytenr, delayed_ref_seq,
+ time_seq, *leafs, tmp, extent_item_pos);
ulist_free(tmp);
if (ret < 0 && ret != -ENOENT) {
- ulist_free(*leafs);
+ free_leaf_list(*leafs);
return ret;
}
@@ -784,10 +1018,12 @@ static int btrfs_find_all_leafs(struct btrfs_trans_handle *trans,
*/
int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
struct btrfs_fs_info *fs_info, u64 bytenr,
- u64 num_bytes, u64 seq, struct ulist **roots)
+ u64 delayed_ref_seq, u64 time_seq,
+ struct ulist **roots)
{
struct ulist *tmp;
struct ulist_node *node = NULL;
+ struct ulist_iterator uiter;
int ret;
tmp = ulist_alloc(GFP_NOFS);
@@ -799,15 +1035,16 @@ int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
return -ENOMEM;
}
+ ULIST_ITER_INIT(&uiter);
while (1) {
- ret = find_parent_nodes(trans, fs_info, bytenr, seq,
- tmp, *roots);
+ ret = find_parent_nodes(trans, fs_info, bytenr, delayed_ref_seq,
+ time_seq, tmp, *roots, NULL);
if (ret < 0 && ret != -ENOENT) {
ulist_free(tmp);
ulist_free(*roots);
return ret;
}
- node = ulist_next(tmp, node);
+ node = ulist_next(tmp, &uiter);
if (!node)
break;
bytenr = node->val;
@@ -1093,67 +1330,25 @@ int tree_backref_for_extent(unsigned long *ptr, struct extent_buffer *eb,
return 0;
}
-static int iterate_leaf_refs(struct btrfs_fs_info *fs_info, u64 logical,
- u64 orig_extent_item_objectid,
- u64 extent_item_pos, u64 root,
+static int iterate_leaf_refs(struct extent_inode_elem *inode_list,
+ u64 root, u64 extent_item_objectid,
iterate_extent_inodes_t *iterate, void *ctx)
{
- u64 disk_byte;
- struct btrfs_key key;
- struct btrfs_file_extent_item *fi;
- struct extent_buffer *eb;
- int slot;
- int nritems;
+ struct extent_inode_elem *eie;
int ret = 0;
- int extent_type;
- u64 data_offset;
- u64 data_len;
-
- eb = read_tree_block(fs_info->tree_root, logical,
- fs_info->tree_root->leafsize, 0);
- if (!eb)
- return -EIO;
-
- /*
- * from the shared data ref, we only have the leaf but we need
- * the key. thus, we must look into all items and see that we
- * find one (some) with a reference to our extent item.
- */
- nritems = btrfs_header_nritems(eb);
- for (slot = 0; slot < nritems; ++slot) {
- btrfs_item_key_to_cpu(eb, &key, slot);
- if (key.type != BTRFS_EXTENT_DATA_KEY)
- continue;
- fi = btrfs_item_ptr(eb, slot, struct btrfs_file_extent_item);
- extent_type = btrfs_file_extent_type(eb, fi);
- if (extent_type == BTRFS_FILE_EXTENT_INLINE)
- continue;
- /* don't skip BTRFS_FILE_EXTENT_PREALLOC, we can handle that */
- disk_byte = btrfs_file_extent_disk_bytenr(eb, fi);
- if (disk_byte != orig_extent_item_objectid)
- continue;
-
- data_offset = btrfs_file_extent_offset(eb, fi);
- data_len = btrfs_file_extent_num_bytes(eb, fi);
-
- if (extent_item_pos < data_offset ||
- extent_item_pos >= data_offset + data_len)
- continue;
+ for (eie = inode_list; eie; eie = eie->next) {
pr_debug("ref for %llu resolved, key (%llu EXTEND_DATA %llu), "
- "root %llu\n", orig_extent_item_objectid,
- key.objectid, key.offset, root);
- ret = iterate(key.objectid,
- key.offset + (extent_item_pos - data_offset),
- root, ctx);
+ "root %llu\n", extent_item_objectid,
+ eie->inum, eie->offset, root);
+ ret = iterate(eie->inum, eie->offset, root, ctx);
if (ret) {
- pr_debug("stopping iteration because ret=%d\n", ret);
+ pr_debug("stopping iteration for %llu due to ret=%d\n",
+ extent_item_objectid, ret);
break;
}
}
- free_extent_buffer(eb);
-
return ret;
}
@@ -1175,7 +1370,10 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info,
struct ulist *roots = NULL;
struct ulist_node *ref_node = NULL;
struct ulist_node *root_node = NULL;
- struct seq_list seq_elem;
+ struct seq_list seq_elem = {};
+ struct seq_list tree_mod_seq_elem = {};
+ struct ulist_iterator ref_uiter;
+ struct ulist_iterator root_uiter;
struct btrfs_delayed_ref_root *delayed_refs = NULL;
pr_debug("resolving all inodes for extent %llu\n",
@@ -1192,34 +1390,41 @@ int iterate_extent_inodes(struct btrfs_fs_info *fs_info,
spin_lock(&delayed_refs->lock);
btrfs_get_delayed_seq(delayed_refs, &seq_elem);
spin_unlock(&delayed_refs->lock);
+ btrfs_get_tree_mod_seq(fs_info, &tree_mod_seq_elem);
}
ret = btrfs_find_all_leafs(trans, fs_info, extent_item_objectid,
- extent_item_pos, seq_elem.seq,
- &refs);
-
+ seq_elem.seq, tree_mod_seq_elem.seq, &refs,
+ &extent_item_pos);
if (ret)
goto out;
- while (!ret && (ref_node = ulist_next(refs, ref_node))) {
- ret = btrfs_find_all_roots(trans, fs_info, ref_node->val, -1,
- seq_elem.seq, &roots);
+ ULIST_ITER_INIT(&ref_uiter);
+ while (!ret && (ref_node = ulist_next(refs, &ref_uiter))) {
+ ret = btrfs_find_all_roots(trans, fs_info, ref_node->val,
+ seq_elem.seq,
+ tree_mod_seq_elem.seq, &roots);
if (ret)
break;
- while (!ret && (root_node = ulist_next(roots, root_node))) {
- pr_debug("root %llu references leaf %llu\n",
- root_node->val, ref_node->val);
- ret = iterate_leaf_refs(fs_info, ref_node->val,
- extent_item_objectid,
- extent_item_pos, root_node->val,
- iterate, ctx);
+ ULIST_ITER_INIT(&root_uiter);
+ while (!ret && (root_node = ulist_next(roots, &root_uiter))) {
+ pr_debug("root %llu references leaf %llu, data list "
+ "%#lx\n", root_node->val, ref_node->val,
+ ref_node->aux);
+ ret = iterate_leaf_refs(
+ (struct extent_inode_elem *)ref_node->aux,
+ root_node->val, extent_item_objectid,
+ iterate, ctx);
}
+ ulist_free(roots);
+ roots = NULL;
}
- ulist_free(refs);
+ free_leaf_list(refs);
ulist_free(roots);
out:
if (!search_commit_root) {
+ btrfs_put_tree_mod_seq(fs_info, &tree_mod_seq_elem);
btrfs_put_delayed_seq(delayed_refs, &seq_elem);
btrfs_end_transaction(trans, fs_info->extent_root);
}
diff --git a/fs/btrfs/backref.h b/fs/btrfs/backref.h
index 57ea2e959e4..c18d8ac7b79 100644
--- a/fs/btrfs/backref.h
+++ b/fs/btrfs/backref.h
@@ -58,7 +58,8 @@ int paths_from_inode(u64 inum, struct inode_fs_paths *ipath);
int btrfs_find_all_roots(struct btrfs_trans_handle *trans,
struct btrfs_fs_info *fs_info, u64 bytenr,
- u64 num_bytes, u64 seq, struct ulist **roots);
+ u64 delayed_ref_seq, u64 time_seq,
+ struct ulist **roots);
struct btrfs_data_container *init_data_container(u32 total_bytes);
struct inode_fs_paths *init_ipath(s32 total_bytes, struct btrfs_root *fs_root,
diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index 9b9b15fd520..e616f8872e6 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -24,6 +24,20 @@
#include "ordered-data.h"
#include "delayed-inode.h"
+/*
+ * ordered_data_close is set by truncate when a file that used
+ * to have good data has been truncated to zero. When it is set
+ * the btrfs file release call will add this inode to the
+ * ordered operations list so that we make sure to flush out any
+ * new data the application may have written before commit.
+ */
+#define BTRFS_INODE_ORDERED_DATA_CLOSE 0
+#define BTRFS_INODE_ORPHAN_META_RESERVED 1
+#define BTRFS_INODE_DUMMY 2
+#define BTRFS_INODE_IN_DEFRAG 3
+#define BTRFS_INODE_DELALLOC_META_RESERVED 4
+#define BTRFS_INODE_HAS_ORPHAN_ITEM 5
+
/* in memory btrfs inode */
struct btrfs_inode {
/* which subvolume this inode belongs to */
@@ -57,9 +71,6 @@ struct btrfs_inode {
/* used to order data wrt metadata */
struct btrfs_ordered_inode_tree ordered_tree;
- /* for keeping track of orphaned inodes */
- struct list_head i_orphan;
-
/* list of all the delalloc inodes in the FS. There are times we need
* to write all the delalloc pages to disk, and this list is used
* to walk them all.
@@ -78,14 +89,13 @@ struct btrfs_inode {
/* the space_info for where this inode's data allocations are done */
struct btrfs_space_info *space_info;
+ unsigned long runtime_flags;
+
/* full 64 bit generation number, struct vfs_inode doesn't have a big
* enough field for this.
*/
u64 generation;
- /* sequence number for NFS changes */
- u64 sequence;
-
/*
* transid of the trans_handle that last modified this inode
*/
@@ -145,22 +155,9 @@ struct btrfs_inode {
unsigned reserved_extents;
/*
- * ordered_data_close is set by truncate when a file that used
- * to have good data has been truncated to zero. When it is set
- * the btrfs file release call will add this inode to the
- * ordered operations list so that we make sure to flush out any
- * new data the application may have written before commit.
- */
- unsigned ordered_data_close:1;
- unsigned orphan_meta_reserved:1;
- unsigned dummy_inode:1;
- unsigned in_defrag:1;
- unsigned delalloc_meta_reserved:1;
-
- /*
* always compress this one file
*/
- unsigned force_compress:4;
+ unsigned force_compress;
struct btrfs_delayed_node *delayed_node;
@@ -202,4 +199,17 @@ static inline bool btrfs_is_free_space_inode(struct btrfs_root *root,
return false;
}
+static inline int btrfs_inode_in_log(struct inode *inode, u64 generation)
+{
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+ int ret = 0;
+
+ mutex_lock(&root->log_mutex);
+ if (BTRFS_I(inode)->logged_trans == generation &&
+ BTRFS_I(inode)->last_sub_trans <= root->last_log_commit)
+ ret = 1;
+ mutex_unlock(&root->log_mutex);
+ return ret;
+}
+
#endif
diff --git a/fs/btrfs/check-integrity.c b/fs/btrfs/check-integrity.c
index c053e90f200..9cebb1fd6a3 100644
--- a/fs/btrfs/check-integrity.c
+++ b/fs/btrfs/check-integrity.c
@@ -103,8 +103,6 @@
#define BTRFSIC_BLOCK_STACK_FRAME_MAGIC_NUMBER 20111300
#define BTRFSIC_TREE_DUMP_MAX_INDENT_LEVEL (200 - 6) /* in characters,
* excluding " [...]" */
-#define BTRFSIC_BLOCK_SIZE PAGE_SIZE
-
#define BTRFSIC_GENERATION_UNKNOWN ((u64)-1)
/*
@@ -210,8 +208,9 @@ struct btrfsic_block_data_ctx {
u64 dev_bytenr; /* physical bytenr on device */
u32 len;
struct btrfsic_dev_state *dev;
- char *data;
- struct buffer_head *bh; /* do not use if set to NULL */
+ char **datav;
+ struct page **pagev;
+ void *mem_to_free;
};
/* This structure is used to implement recursion without occupying
@@ -243,6 +242,8 @@ struct btrfsic_state {
struct btrfs_root *root;
u64 max_superblock_generation;
struct btrfsic_block *latest_superblock;
+ u32 metablock_size;
+ u32 datablock_size;
};
static void btrfsic_block_init(struct btrfsic_block *b);
@@ -290,8 +291,10 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
static int btrfsic_process_metablock(struct btrfsic_state *state,
struct btrfsic_block *block,
struct btrfsic_block_data_ctx *block_ctx,
- struct btrfs_header *hdr,
int limit_nesting, int force_iodone_flag);
+static void btrfsic_read_from_block_data(
+ struct btrfsic_block_data_ctx *block_ctx,
+ void *dst, u32 offset, size_t len);
static int btrfsic_create_link_to_next_block(
struct btrfsic_state *state,
struct btrfsic_block *block,
@@ -318,12 +321,13 @@ static void btrfsic_release_block_ctx(struct btrfsic_block_data_ctx *block_ctx);
static int btrfsic_read_block(struct btrfsic_state *state,
struct btrfsic_block_data_ctx *block_ctx);
static void btrfsic_dump_database(struct btrfsic_state *state);
+static void btrfsic_complete_bio_end_io(struct bio *bio, int err);
static int btrfsic_test_for_metadata(struct btrfsic_state *state,
- const u8 *data, unsigned int size);
+ char **datav, unsigned int num_pages);
static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
- u64 dev_bytenr, u8 *mapped_data,
- unsigned int len, struct bio *bio,
- int *bio_is_patched,
+ u64 dev_bytenr, char **mapped_datav,
+ unsigned int num_pages,
+ struct bio *bio, int *bio_is_patched,
struct buffer_head *bh,
int submit_bio_bh_rw);
static int btrfsic_process_written_superblock(
@@ -375,7 +379,7 @@ static struct btrfsic_dev_state *btrfsic_dev_state_lookup(
static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
u64 bytenr,
struct btrfsic_dev_state *dev_state,
- u64 dev_bytenr, char *data);
+ u64 dev_bytenr);
static struct mutex btrfsic_mutex;
static int btrfsic_is_initialized;
@@ -651,7 +655,7 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
int pass;
BUG_ON(NULL == state);
- selected_super = kmalloc(sizeof(*selected_super), GFP_NOFS);
+ selected_super = kzalloc(sizeof(*selected_super), GFP_NOFS);
if (NULL == selected_super) {
printk(KERN_INFO "btrfsic: error, kmalloc failed!\n");
return -1;
@@ -718,7 +722,7 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
num_copies =
btrfs_num_copies(&state->root->fs_info->mapping_tree,
- next_bytenr, PAGE_SIZE);
+ next_bytenr, state->metablock_size);
if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
(unsigned long long)next_bytenr, num_copies);
@@ -727,9 +731,9 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
struct btrfsic_block *next_block;
struct btrfsic_block_data_ctx tmp_next_block_ctx;
struct btrfsic_block_link *l;
- struct btrfs_header *hdr;
- ret = btrfsic_map_block(state, next_bytenr, PAGE_SIZE,
+ ret = btrfsic_map_block(state, next_bytenr,
+ state->metablock_size,
&tmp_next_block_ctx,
mirror_num);
if (ret) {
@@ -758,7 +762,7 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
BUG_ON(NULL == l);
ret = btrfsic_read_block(state, &tmp_next_block_ctx);
- if (ret < (int)BTRFSIC_BLOCK_SIZE) {
+ if (ret < (int)PAGE_CACHE_SIZE) {
printk(KERN_INFO
"btrfsic: read @logical %llu failed!\n",
(unsigned long long)
@@ -768,11 +772,9 @@ static int btrfsic_process_superblock(struct btrfsic_state *state,
return -1;
}
- hdr = (struct btrfs_header *)tmp_next_block_ctx.data;
ret = btrfsic_process_metablock(state,
next_block,
&tmp_next_block_ctx,
- hdr,
BTRFS_MAX_LEVEL + 3, 1);
btrfsic_release_block_ctx(&tmp_next_block_ctx);
}
@@ -799,7 +801,10 @@ static int btrfsic_process_superblock_dev_mirror(
/* super block bytenr is always the unmapped device bytenr */
dev_bytenr = btrfs_sb_offset(superblock_mirror_num);
- bh = __bread(superblock_bdev, dev_bytenr / 4096, 4096);
+ if (dev_bytenr + BTRFS_SUPER_INFO_SIZE > device->total_bytes)
+ return -1;
+ bh = __bread(superblock_bdev, dev_bytenr / 4096,
+ BTRFS_SUPER_INFO_SIZE);
if (NULL == bh)
return -1;
super_tmp = (struct btrfs_super_block *)
@@ -808,7 +813,10 @@ static int btrfsic_process_superblock_dev_mirror(
if (btrfs_super_bytenr(super_tmp) != dev_bytenr ||
strncmp((char *)(&(super_tmp->magic)), BTRFS_MAGIC,
sizeof(super_tmp->magic)) ||
- memcmp(device->uuid, super_tmp->dev_item.uuid, BTRFS_UUID_SIZE)) {
+ memcmp(device->uuid, super_tmp->dev_item.uuid, BTRFS_UUID_SIZE) ||
+ btrfs_super_nodesize(super_tmp) != state->metablock_size ||
+ btrfs_super_leafsize(super_tmp) != state->metablock_size ||
+ btrfs_super_sectorsize(super_tmp) != state->datablock_size) {
brelse(bh);
return 0;
}
@@ -893,7 +901,7 @@ static int btrfsic_process_superblock_dev_mirror(
num_copies =
btrfs_num_copies(&state->root->fs_info->mapping_tree,
- next_bytenr, PAGE_SIZE);
+ next_bytenr, state->metablock_size);
if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
(unsigned long long)next_bytenr, num_copies);
@@ -902,7 +910,8 @@ static int btrfsic_process_superblock_dev_mirror(
struct btrfsic_block_data_ctx tmp_next_block_ctx;
struct btrfsic_block_link *l;
- if (btrfsic_map_block(state, next_bytenr, PAGE_SIZE,
+ if (btrfsic_map_block(state, next_bytenr,
+ state->metablock_size,
&tmp_next_block_ctx,
mirror_num)) {
printk(KERN_INFO "btrfsic: btrfsic_map_block("
@@ -966,13 +975,15 @@ static int btrfsic_process_metablock(
struct btrfsic_state *state,
struct btrfsic_block *const first_block,
struct btrfsic_block_data_ctx *const first_block_ctx,
- struct btrfs_header *const first_hdr,
int first_limit_nesting, int force_iodone_flag)
{
struct btrfsic_stack_frame initial_stack_frame = { 0 };
struct btrfsic_stack_frame *sf;
struct btrfsic_stack_frame *next_stack;
+ struct btrfs_header *const first_hdr =
+ (struct btrfs_header *)first_block_ctx->datav[0];
+ BUG_ON(!first_hdr);
sf = &initial_stack_frame;
sf->error = 0;
sf->i = -1;
@@ -1012,21 +1023,47 @@ continue_with_current_leaf_stack_frame:
}
if (sf->i < sf->nr) {
- struct btrfs_item *disk_item = leafhdr->items + sf->i;
- struct btrfs_disk_key *disk_key = &disk_item->key;
+ struct btrfs_item disk_item;
+ u32 disk_item_offset =
+ (uintptr_t)(leafhdr->items + sf->i) -
+ (uintptr_t)leafhdr;
+ struct btrfs_disk_key *disk_key;
u8 type;
- const u32 item_offset = le32_to_cpu(disk_item->offset);
+ u32 item_offset;
+ if (disk_item_offset + sizeof(struct btrfs_item) >
+ sf->block_ctx->len) {
+leaf_item_out_of_bounce_error:
+ printk(KERN_INFO
+ "btrfsic: leaf item out of bounce at logical %llu, dev %s\n",
+ sf->block_ctx->start,
+ sf->block_ctx->dev->name);
+ goto one_stack_frame_backwards;
+ }
+ btrfsic_read_from_block_data(sf->block_ctx,
+ &disk_item,
+ disk_item_offset,
+ sizeof(struct btrfs_item));
+ item_offset = le32_to_cpu(disk_item.offset);
+ disk_key = &disk_item.key;
type = disk_key->type;
if (BTRFS_ROOT_ITEM_KEY == type) {
- const struct btrfs_root_item *const root_item =
- (struct btrfs_root_item *)
- (sf->block_ctx->data +
- offsetof(struct btrfs_leaf, items) +
- item_offset);
- const u64 next_bytenr =
- le64_to_cpu(root_item->bytenr);
+ struct btrfs_root_item root_item;
+ u32 root_item_offset;
+ u64 next_bytenr;
+
+ root_item_offset = item_offset +
+ offsetof(struct btrfs_leaf, items);
+ if (root_item_offset +
+ sizeof(struct btrfs_root_item) >
+ sf->block_ctx->len)
+ goto leaf_item_out_of_bounce_error;
+ btrfsic_read_from_block_data(
+ sf->block_ctx, &root_item,
+ root_item_offset,
+ sizeof(struct btrfs_root_item));
+ next_bytenr = le64_to_cpu(root_item.bytenr);
sf->error =
btrfsic_create_link_to_next_block(
@@ -1041,7 +1078,7 @@ continue_with_current_leaf_stack_frame:
&sf->num_copies,
&sf->mirror_num,
disk_key,
- le64_to_cpu(root_item->
+ le64_to_cpu(root_item.
generation));
if (sf->error)
goto one_stack_frame_backwards;
@@ -1049,7 +1086,7 @@ continue_with_current_leaf_stack_frame:
if (NULL != sf->next_block) {
struct btrfs_header *const next_hdr =
(struct btrfs_header *)
- sf->next_block_ctx.data;
+ sf->next_block_ctx.datav[0];
next_stack =
btrfsic_stack_frame_alloc();
@@ -1111,10 +1148,24 @@ continue_with_current_node_stack_frame:
}
if (sf->i < sf->nr) {
- struct btrfs_key_ptr *disk_key_ptr =
- nodehdr->ptrs + sf->i;
- const u64 next_bytenr =
- le64_to_cpu(disk_key_ptr->blockptr);
+ struct btrfs_key_ptr key_ptr;
+ u32 key_ptr_offset;
+ u64 next_bytenr;
+
+ key_ptr_offset = (uintptr_t)(nodehdr->ptrs + sf->i) -
+ (uintptr_t)nodehdr;
+ if (key_ptr_offset + sizeof(struct btrfs_key_ptr) >
+ sf->block_ctx->len) {
+ printk(KERN_INFO
+ "btrfsic: node item out of bounce at logical %llu, dev %s\n",
+ sf->block_ctx->start,
+ sf->block_ctx->dev->name);
+ goto one_stack_frame_backwards;
+ }
+ btrfsic_read_from_block_data(
+ sf->block_ctx, &key_ptr, key_ptr_offset,
+ sizeof(struct btrfs_key_ptr));
+ next_bytenr = le64_to_cpu(key_ptr.blockptr);
sf->error = btrfsic_create_link_to_next_block(
state,
@@ -1127,15 +1178,15 @@ continue_with_current_node_stack_frame:
force_iodone_flag,
&sf->num_copies,
&sf->mirror_num,
- &disk_key_ptr->key,
- le64_to_cpu(disk_key_ptr->generation));
+ &key_ptr.key,
+ le64_to_cpu(key_ptr.generation));
if (sf->error)
goto one_stack_frame_backwards;
if (NULL != sf->next_block) {
struct btrfs_header *const next_hdr =
(struct btrfs_header *)
- sf->next_block_ctx.data;
+ sf->next_block_ctx.datav[0];
next_stack = btrfsic_stack_frame_alloc();
if (NULL == next_stack)
@@ -1181,6 +1232,35 @@ one_stack_frame_backwards:
return sf->error;
}
+static void btrfsic_read_from_block_data(
+ struct btrfsic_block_data_ctx *block_ctx,
+ void *dstv, u32 offset, size_t len)
+{
+ size_t cur;
+ size_t offset_in_page;
+ char *kaddr;
+ char *dst = (char *)dstv;
+ size_t start_offset = block_ctx->start & ((u64)PAGE_CACHE_SIZE - 1);
+ unsigned long i = (start_offset + offset) >> PAGE_CACHE_SHIFT;
+
+ WARN_ON(offset + len > block_ctx->len);
+ offset_in_page = (start_offset + offset) &
+ ((unsigned long)PAGE_CACHE_SIZE - 1);
+
+ while (len > 0) {
+ cur = min(len, ((size_t)PAGE_CACHE_SIZE - offset_in_page));
+ BUG_ON(i >= (block_ctx->len + PAGE_CACHE_SIZE - 1) >>
+ PAGE_CACHE_SHIFT);
+ kaddr = block_ctx->datav[i];
+ memcpy(dst, kaddr + offset_in_page, cur);
+
+ dst += cur;
+ len -= cur;
+ offset_in_page = 0;
+ i++;
+ }
+}
+
static int btrfsic_create_link_to_next_block(
struct btrfsic_state *state,
struct btrfsic_block *block,
@@ -1204,7 +1284,7 @@ static int btrfsic_create_link_to_next_block(
if (0 == *num_copiesp) {
*num_copiesp =
btrfs_num_copies(&state->root->fs_info->mapping_tree,
- next_bytenr, PAGE_SIZE);
+ next_bytenr, state->metablock_size);
if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
(unsigned long long)next_bytenr, *num_copiesp);
@@ -1219,7 +1299,7 @@ static int btrfsic_create_link_to_next_block(
"btrfsic_create_link_to_next_block(mirror_num=%d)\n",
*mirror_nump);
ret = btrfsic_map_block(state, next_bytenr,
- BTRFSIC_BLOCK_SIZE,
+ state->metablock_size,
next_block_ctx, *mirror_nump);
if (ret) {
printk(KERN_INFO
@@ -1314,7 +1394,7 @@ static int btrfsic_create_link_to_next_block(
if (limit_nesting > 0 && did_alloc_block_link) {
ret = btrfsic_read_block(state, next_block_ctx);
- if (ret < (int)BTRFSIC_BLOCK_SIZE) {
+ if (ret < (int)next_block_ctx->len) {
printk(KERN_INFO
"btrfsic: read block @logical %llu failed!\n",
(unsigned long long)next_bytenr);
@@ -1339,43 +1419,74 @@ static int btrfsic_handle_extent_data(
u32 item_offset, int force_iodone_flag)
{
int ret;
- struct btrfs_file_extent_item *file_extent_item =
- (struct btrfs_file_extent_item *)(block_ctx->data +
- offsetof(struct btrfs_leaf,
- items) + item_offset);
- u64 next_bytenr =
- le64_to_cpu(file_extent_item->disk_bytenr) +
- le64_to_cpu(file_extent_item->offset);
- u64 num_bytes = le64_to_cpu(file_extent_item->num_bytes);
- u64 generation = le64_to_cpu(file_extent_item->generation);
+ struct btrfs_file_extent_item file_extent_item;
+ u64 file_extent_item_offset;
+ u64 next_bytenr;
+ u64 num_bytes;
+ u64 generation;
struct btrfsic_block_link *l;
+ file_extent_item_offset = offsetof(struct btrfs_leaf, items) +
+ item_offset;
+ if (file_extent_item_offset +
+ offsetof(struct btrfs_file_extent_item, disk_num_bytes) >
+ block_ctx->len) {
+ printk(KERN_INFO
+ "btrfsic: file item out of bounce at logical %llu, dev %s\n",
+ block_ctx->start, block_ctx->dev->name);
+ return -1;
+ }
+
+ btrfsic_read_from_block_data(block_ctx, &file_extent_item,
+ file_extent_item_offset,
+ offsetof(struct btrfs_file_extent_item, disk_num_bytes));
+ if (BTRFS_FILE_EXTENT_REG != file_extent_item.type ||
+ ((u64)0) == le64_to_cpu(file_extent_item.disk_bytenr)) {
+ if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)
+ printk(KERN_INFO "extent_data: type %u, disk_bytenr = %llu\n",
+ file_extent_item.type,
+ (unsigned long long)
+ le64_to_cpu(file_extent_item.disk_bytenr));
+ return 0;
+ }
+
+ if (file_extent_item_offset + sizeof(struct btrfs_file_extent_item) >
+ block_ctx->len) {
+ printk(KERN_INFO
+ "btrfsic: file item out of bounce at logical %llu, dev %s\n",
+ block_ctx->start, block_ctx->dev->name);
+ return -1;
+ }
+ btrfsic_read_from_block_data(block_ctx, &file_extent_item,
+ file_extent_item_offset,
+ sizeof(struct btrfs_file_extent_item));
+ next_bytenr = le64_to_cpu(file_extent_item.disk_bytenr) +
+ le64_to_cpu(file_extent_item.offset);
+ generation = le64_to_cpu(file_extent_item.generation);
+ num_bytes = le64_to_cpu(file_extent_item.num_bytes);
+ generation = le64_to_cpu(file_extent_item.generation);
+
if (state->print_mask & BTRFSIC_PRINT_MASK_VERY_VERBOSE)
printk(KERN_INFO "extent_data: type %u, disk_bytenr = %llu,"
" offset = %llu, num_bytes = %llu\n",
- file_extent_item->type,
+ file_extent_item.type,
(unsigned long long)
- le64_to_cpu(file_extent_item->disk_bytenr),
- (unsigned long long)
- le64_to_cpu(file_extent_item->offset),
- (unsigned long long)
- le64_to_cpu(file_extent_item->num_bytes));
- if (BTRFS_FILE_EXTENT_REG != file_extent_item->type ||
- ((u64)0) == le64_to_cpu(file_extent_item->disk_bytenr))
- return 0;
+ le64_to_cpu(file_extent_item.disk_bytenr),
+ (unsigned long long)le64_to_cpu(file_extent_item.offset),
+ (unsigned long long)num_bytes);
while (num_bytes > 0) {
u32 chunk_len;
int num_copies;
int mirror_num;
- if (num_bytes > BTRFSIC_BLOCK_SIZE)
- chunk_len = BTRFSIC_BLOCK_SIZE;
+ if (num_bytes > state->datablock_size)
+ chunk_len = state->datablock_size;
else
chunk_len = num_bytes;
num_copies =
btrfs_num_copies(&state->root->fs_info->mapping_tree,
- next_bytenr, PAGE_SIZE);
+ next_bytenr, state->datablock_size);
if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
(unsigned long long)next_bytenr, num_copies);
@@ -1475,8 +1586,9 @@ static int btrfsic_map_block(struct btrfsic_state *state, u64 bytenr, u32 len,
block_ctx_out->dev_bytenr = multi->stripes[0].physical;
block_ctx_out->start = bytenr;
block_ctx_out->len = len;
- block_ctx_out->data = NULL;
- block_ctx_out->bh = NULL;
+ block_ctx_out->datav = NULL;
+ block_ctx_out->pagev = NULL;
+ block_ctx_out->mem_to_free = NULL;
if (0 == ret)
kfree(multi);
@@ -1496,8 +1608,9 @@ static int btrfsic_map_superblock(struct btrfsic_state *state, u64 bytenr,
block_ctx_out->dev_bytenr = bytenr;
block_ctx_out->start = bytenr;
block_ctx_out->len = len;
- block_ctx_out->data = NULL;
- block_ctx_out->bh = NULL;
+ block_ctx_out->datav = NULL;
+ block_ctx_out->pagev = NULL;
+ block_ctx_out->mem_to_free = NULL;
if (NULL != block_ctx_out->dev) {
return 0;
} else {
@@ -1508,38 +1621,127 @@ static int btrfsic_map_superblock(struct btrfsic_state *state, u64 bytenr,
static void btrfsic_release_block_ctx(struct btrfsic_block_data_ctx *block_ctx)
{
- if (NULL != block_ctx->bh) {
- brelse(block_ctx->bh);
- block_ctx->bh = NULL;
+ if (block_ctx->mem_to_free) {
+ unsigned int num_pages;
+
+ BUG_ON(!block_ctx->datav);
+ BUG_ON(!block_ctx->pagev);
+ num_pages = (block_ctx->len + (u64)PAGE_CACHE_SIZE - 1) >>
+ PAGE_CACHE_SHIFT;
+ while (num_pages > 0) {
+ num_pages--;
+ if (block_ctx->datav[num_pages]) {
+ kunmap(block_ctx->pagev[num_pages]);
+ block_ctx->datav[num_pages] = NULL;
+ }
+ if (block_ctx->pagev[num_pages]) {
+ __free_page(block_ctx->pagev[num_pages]);
+ block_ctx->pagev[num_pages] = NULL;
+ }
+ }
+
+ kfree(block_ctx->mem_to_free);
+ block_ctx->mem_to_free = NULL;
+ block_ctx->pagev = NULL;
+ block_ctx->datav = NULL;
}
}
static int btrfsic_read_block(struct btrfsic_state *state,
struct btrfsic_block_data_ctx *block_ctx)
{
- block_ctx->bh = NULL;
- if (block_ctx->dev_bytenr & 4095) {
+ unsigned int num_pages;
+ unsigned int i;
+ u64 dev_bytenr;
+ int ret;
+
+ BUG_ON(block_ctx->datav);
+ BUG_ON(block_ctx->pagev);
+ BUG_ON(block_ctx->mem_to_free);
+ if (block_ctx->dev_bytenr & ((u64)PAGE_CACHE_SIZE - 1)) {
printk(KERN_INFO
"btrfsic: read_block() with unaligned bytenr %llu\n",
(unsigned long long)block_ctx->dev_bytenr);
return -1;
}
- if (block_ctx->len > 4096) {
- printk(KERN_INFO
- "btrfsic: read_block() with too huge size %d\n",
- block_ctx->len);
+
+ num_pages = (block_ctx->len + (u64)PAGE_CACHE_SIZE - 1) >>
+ PAGE_CACHE_SHIFT;
+ block_ctx->mem_to_free = kzalloc((sizeof(*block_ctx->datav) +
+ sizeof(*block_ctx->pagev)) *
+ num_pages, GFP_NOFS);
+ if (!block_ctx->mem_to_free)
return -1;
+ block_ctx->datav = block_ctx->mem_to_free;
+ block_ctx->pagev = (struct page **)(block_ctx->datav + num_pages);
+ for (i = 0; i < num_pages; i++) {
+ block_ctx->pagev[i] = alloc_page(GFP_NOFS);
+ if (!block_ctx->pagev[i])
+ return -1;
}
- block_ctx->bh = __bread(block_ctx->dev->bdev,
- block_ctx->dev_bytenr >> 12, 4096);
- if (NULL == block_ctx->bh)
- return -1;
- block_ctx->data = block_ctx->bh->b_data;
+ dev_bytenr = block_ctx->dev_bytenr;
+ for (i = 0; i < num_pages;) {
+ struct bio *bio;
+ unsigned int j;
+ DECLARE_COMPLETION_ONSTACK(complete);
+
+ bio = bio_alloc(GFP_NOFS, num_pages - i);
+ if (!bio) {
+ printk(KERN_INFO
+ "btrfsic: bio_alloc() for %u pages failed!\n",
+ num_pages - i);
+ return -1;
+ }
+ bio->bi_bdev = block_ctx->dev->bdev;
+ bio->bi_sector = dev_bytenr >> 9;
+ bio->bi_end_io = btrfsic_complete_bio_end_io;
+ bio->bi_private = &complete;
+
+ for (j = i; j < num_pages; j++) {
+ ret = bio_add_page(bio, block_ctx->pagev[j],
+ PAGE_CACHE_SIZE, 0);
+ if (PAGE_CACHE_SIZE != ret)
+ break;
+ }
+ if (j == i) {
+ printk(KERN_INFO
+ "btrfsic: error, failed to add a single page!\n");
+ return -1;
+ }
+ submit_bio(READ, bio);
+
+ /* this will also unplug the queue */
+ wait_for_completion(&complete);
+
+ if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) {
+ printk(KERN_INFO
+ "btrfsic: read error at logical %llu dev %s!\n",
+ block_ctx->start, block_ctx->dev->name);
+ bio_put(bio);
+ return -1;
+ }
+ bio_put(bio);
+ dev_bytenr += (j - i) * PAGE_CACHE_SIZE;
+ i = j;
+ }
+ for (i = 0; i < num_pages; i++) {
+ block_ctx->datav[i] = kmap(block_ctx->pagev[i]);
+ if (!block_ctx->datav[i]) {
+ printk(KERN_INFO "btrfsic: kmap() failed (dev %s)!\n",
+ block_ctx->dev->name);
+ return -1;
+ }
+ }
return block_ctx->len;
}
+static void btrfsic_complete_bio_end_io(struct bio *bio, int err)
+{
+ complete((struct completion *)bio->bi_private);
+}
+
static void btrfsic_dump_database(struct btrfsic_state *state)
{
struct list_head *elem_all;
@@ -1617,32 +1819,39 @@ static void btrfsic_dump_database(struct btrfsic_state *state)
* (note that this test fails for the super block)
*/
static int btrfsic_test_for_metadata(struct btrfsic_state *state,
- const u8 *data, unsigned int size)
+ char **datav, unsigned int num_pages)
{
struct btrfs_header *h;
u8 csum[BTRFS_CSUM_SIZE];
u32 crc = ~(u32)0;
- int fail = 0;
- int crc_fail = 0;
+ unsigned int i;
- h = (struct btrfs_header *)data;
+ if (num_pages * PAGE_CACHE_SIZE < state->metablock_size)
+ return 1; /* not metadata */
+ num_pages = state->metablock_size >> PAGE_CACHE_SHIFT;
+ h = (struct btrfs_header *)datav[0];
if (memcmp(h->fsid, state->root->fs_info->fsid, BTRFS_UUID_SIZE))
- fail++;
+ return 1;
+
+ for (i = 0; i < num_pages; i++) {
+ u8 *data = i ? datav[i] : (datav[i] + BTRFS_CSUM_SIZE);
+ size_t sublen = i ? PAGE_CACHE_SIZE :
+ (PAGE_CACHE_SIZE - BTRFS_CSUM_SIZE);
- crc = crc32c(crc, data + BTRFS_CSUM_SIZE, PAGE_SIZE - BTRFS_CSUM_SIZE);
+ crc = crc32c(crc, data, sublen);
+ }
btrfs_csum_final(crc, csum);
if (memcmp(csum, h->csum, state->csum_size))
- crc_fail++;
+ return 1;
- return fail || crc_fail;
+ return 0; /* is metadata */
}
static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
- u64 dev_bytenr,
- u8 *mapped_data, unsigned int len,
- struct bio *bio,
- int *bio_is_patched,
+ u64 dev_bytenr, char **mapped_datav,
+ unsigned int num_pages,
+ struct bio *bio, int *bio_is_patched,
struct buffer_head *bh,
int submit_bio_bh_rw)
{
@@ -1652,12 +1861,19 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
int ret;
struct btrfsic_state *state = dev_state->state;
struct block_device *bdev = dev_state->bdev;
+ unsigned int processed_len;
- WARN_ON(len > PAGE_SIZE);
- is_metadata = (0 == btrfsic_test_for_metadata(state, mapped_data, len));
if (NULL != bio_is_patched)
*bio_is_patched = 0;
+again:
+ if (num_pages == 0)
+ return;
+
+ processed_len = 0;
+ is_metadata = (0 == btrfsic_test_for_metadata(state, mapped_datav,
+ num_pages));
+
block = btrfsic_block_hashtable_lookup(bdev, dev_bytenr,
&state->block_hashtable);
if (NULL != block) {
@@ -1667,8 +1883,16 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
if (block->is_superblock) {
bytenr = le64_to_cpu(((struct btrfs_super_block *)
- mapped_data)->bytenr);
+ mapped_datav[0])->bytenr);
+ if (num_pages * PAGE_CACHE_SIZE <
+ BTRFS_SUPER_INFO_SIZE) {
+ printk(KERN_INFO
+ "btrfsic: cannot work with too short bios!\n");
+ return;
+ }
is_metadata = 1;
+ BUG_ON(BTRFS_SUPER_INFO_SIZE & (PAGE_CACHE_SIZE - 1));
+ processed_len = BTRFS_SUPER_INFO_SIZE;
if (state->print_mask &
BTRFSIC_PRINT_MASK_TREE_BEFORE_SB_WRITE) {
printk(KERN_INFO
@@ -1678,12 +1902,18 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
}
if (is_metadata) {
if (!block->is_superblock) {
+ if (num_pages * PAGE_CACHE_SIZE <
+ state->metablock_size) {
+ printk(KERN_INFO
+ "btrfsic: cannot work with too short bios!\n");
+ return;
+ }
+ processed_len = state->metablock_size;
bytenr = le64_to_cpu(((struct btrfs_header *)
- mapped_data)->bytenr);
+ mapped_datav[0])->bytenr);
btrfsic_cmp_log_and_dev_bytenr(state, bytenr,
dev_state,
- dev_bytenr,
- mapped_data);
+ dev_bytenr);
}
if (block->logical_bytenr != bytenr) {
printk(KERN_INFO
@@ -1710,6 +1940,13 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
block->mirror_num,
btrfsic_get_block_type(state, block));
} else {
+ if (num_pages * PAGE_CACHE_SIZE <
+ state->datablock_size) {
+ printk(KERN_INFO
+ "btrfsic: cannot work with too short bios!\n");
+ return;
+ }
+ processed_len = state->datablock_size;
bytenr = block->logical_bytenr;
if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
printk(KERN_INFO
@@ -1747,7 +1984,7 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
le64_to_cpu(block->disk_key.offset),
(unsigned long long)
le64_to_cpu(((struct btrfs_header *)
- mapped_data)->generation),
+ mapped_datav[0])->generation),
(unsigned long long)
state->max_superblock_generation);
btrfsic_dump_tree(state);
@@ -1765,10 +2002,10 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
(unsigned long long)block->generation,
(unsigned long long)
le64_to_cpu(((struct btrfs_header *)
- mapped_data)->generation));
+ mapped_datav[0])->generation));
/* it would not be safe to go on */
btrfsic_dump_tree(state);
- return;
+ goto continue_loop;
}
/*
@@ -1796,18 +2033,19 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
}
if (block->is_superblock)
- ret = btrfsic_map_superblock(state, bytenr, len,
+ ret = btrfsic_map_superblock(state, bytenr,
+ processed_len,
bdev, &block_ctx);
else
- ret = btrfsic_map_block(state, bytenr, len,
+ ret = btrfsic_map_block(state, bytenr, processed_len,
&block_ctx, 0);
if (ret) {
printk(KERN_INFO
"btrfsic: btrfsic_map_block(root @%llu)"
" failed!\n", (unsigned long long)bytenr);
- return;
+ goto continue_loop;
}
- block_ctx.data = mapped_data;
+ block_ctx.datav = mapped_datav;
/* the following is required in case of writes to mirrors,
* use the same that was used for the lookup */
block_ctx.dev = dev_state;
@@ -1863,11 +2101,13 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
block->logical_bytenr = bytenr;
block->is_metadata = 1;
if (block->is_superblock) {
+ BUG_ON(PAGE_CACHE_SIZE !=
+ BTRFS_SUPER_INFO_SIZE);
ret = btrfsic_process_written_superblock(
state,
block,
(struct btrfs_super_block *)
- mapped_data);
+ mapped_datav[0]);
if (state->print_mask &
BTRFSIC_PRINT_MASK_TREE_AFTER_SB_WRITE) {
printk(KERN_INFO
@@ -1880,8 +2120,6 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
state,
block,
&block_ctx,
- (struct btrfs_header *)
- block_ctx.data,
0, 0);
}
if (ret)
@@ -1912,26 +2150,30 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
u64 bytenr;
if (!is_metadata) {
+ processed_len = state->datablock_size;
if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
printk(KERN_INFO "Written block (%s/%llu/?)"
" !found in hash table, D.\n",
dev_state->name,
(unsigned long long)dev_bytenr);
- if (!state->include_extent_data)
- return; /* ignore that written D block */
+ if (!state->include_extent_data) {
+ /* ignore that written D block */
+ goto continue_loop;
+ }
/* this is getting ugly for the
* include_extent_data case... */
bytenr = 0; /* unknown */
block_ctx.start = bytenr;
- block_ctx.len = len;
- block_ctx.bh = NULL;
+ block_ctx.len = processed_len;
+ block_ctx.mem_to_free = NULL;
+ block_ctx.pagev = NULL;
} else {
+ processed_len = state->metablock_size;
bytenr = le64_to_cpu(((struct btrfs_header *)
- mapped_data)->bytenr);
+ mapped_datav[0])->bytenr);
btrfsic_cmp_log_and_dev_bytenr(state, bytenr, dev_state,
- dev_bytenr,
- mapped_data);
+ dev_bytenr);
if (state->print_mask & BTRFSIC_PRINT_MASK_VERBOSE)
printk(KERN_INFO
"Written block @%llu (%s/%llu/?)"
@@ -1940,17 +2182,17 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
dev_state->name,
(unsigned long long)dev_bytenr);
- ret = btrfsic_map_block(state, bytenr, len, &block_ctx,
- 0);
+ ret = btrfsic_map_block(state, bytenr, processed_len,
+ &block_ctx, 0);
if (ret) {
printk(KERN_INFO
"btrfsic: btrfsic_map_block(root @%llu)"
" failed!\n",
(unsigned long long)dev_bytenr);
- return;
+ goto continue_loop;
}
}
- block_ctx.data = mapped_data;
+ block_ctx.datav = mapped_datav;
/* the following is required in case of writes to mirrors,
* use the same that was used for the lookup */
block_ctx.dev = dev_state;
@@ -1960,7 +2202,7 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
if (NULL == block) {
printk(KERN_INFO "btrfsic: error, kmalloc failed!\n");
btrfsic_release_block_ctx(&block_ctx);
- return;
+ goto continue_loop;
}
block->dev_state = dev_state;
block->dev_bytenr = dev_bytenr;
@@ -2020,9 +2262,7 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
if (is_metadata) {
ret = btrfsic_process_metablock(state, block,
- &block_ctx,
- (struct btrfs_header *)
- block_ctx.data, 0, 0);
+ &block_ctx, 0, 0);
if (ret)
printk(KERN_INFO
"btrfsic: process_metablock(root @%llu)"
@@ -2031,6 +2271,13 @@ static void btrfsic_process_written_block(struct btrfsic_dev_state *dev_state,
}
btrfsic_release_block_ctx(&block_ctx);
}
+
+continue_loop:
+ BUG_ON(!processed_len);
+ dev_bytenr += processed_len;
+ mapped_datav += processed_len >> PAGE_CACHE_SHIFT;
+ num_pages -= processed_len >> PAGE_CACHE_SHIFT;
+ goto again;
}
static void btrfsic_bio_end_io(struct bio *bp, int bio_error_status)
@@ -2213,7 +2460,7 @@ static int btrfsic_process_written_superblock(
num_copies =
btrfs_num_copies(&state->root->fs_info->mapping_tree,
- next_bytenr, PAGE_SIZE);
+ next_bytenr, BTRFS_SUPER_INFO_SIZE);
if (state->print_mask & BTRFSIC_PRINT_MASK_NUM_COPIES)
printk(KERN_INFO "num_copies(log_bytenr=%llu) = %d\n",
(unsigned long long)next_bytenr, num_copies);
@@ -2224,7 +2471,8 @@ static int btrfsic_process_written_superblock(
printk(KERN_INFO
"btrfsic_process_written_superblock("
"mirror_num=%d)\n", mirror_num);
- ret = btrfsic_map_block(state, next_bytenr, PAGE_SIZE,
+ ret = btrfsic_map_block(state, next_bytenr,
+ BTRFS_SUPER_INFO_SIZE,
&tmp_next_block_ctx,
mirror_num);
if (ret) {
@@ -2689,7 +2937,7 @@ static struct btrfsic_block *btrfsic_block_lookup_or_add(
static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
u64 bytenr,
struct btrfsic_dev_state *dev_state,
- u64 dev_bytenr, char *data)
+ u64 dev_bytenr)
{
int num_copies;
int mirror_num;
@@ -2698,10 +2946,10 @@ static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
int match = 0;
num_copies = btrfs_num_copies(&state->root->fs_info->mapping_tree,
- bytenr, PAGE_SIZE);
+ bytenr, state->metablock_size);
for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
- ret = btrfsic_map_block(state, bytenr, PAGE_SIZE,
+ ret = btrfsic_map_block(state, bytenr, state->metablock_size,
&block_ctx, mirror_num);
if (ret) {
printk(KERN_INFO "btrfsic:"
@@ -2727,7 +2975,8 @@ static void btrfsic_cmp_log_and_dev_bytenr(struct btrfsic_state *state,
(unsigned long long)bytenr, dev_state->name,
(unsigned long long)dev_bytenr);
for (mirror_num = 1; mirror_num <= num_copies; mirror_num++) {
- ret = btrfsic_map_block(state, bytenr, PAGE_SIZE,
+ ret = btrfsic_map_block(state, bytenr,
+ state->metablock_size,
&block_ctx, mirror_num);
if (ret)
continue;
@@ -2781,13 +3030,13 @@ int btrfsic_submit_bh(int rw, struct buffer_head *bh)
(unsigned long)bh->b_size, bh->b_data,
bh->b_bdev);
btrfsic_process_written_block(dev_state, dev_bytenr,
- bh->b_data, bh->b_size, NULL,
+ &bh->b_data, 1, NULL,
NULL, bh, rw);
} else if (NULL != dev_state && (rw & REQ_FLUSH)) {
if (dev_state->state->print_mask &
BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
printk(KERN_INFO
- "submit_bh(rw=0x%x) FLUSH, bdev=%p)\n",
+ "submit_bh(rw=0x%x FLUSH, bdev=%p)\n",
rw, bh->b_bdev);
if (!dev_state->dummy_block_for_bio_bh_flush.is_iodone) {
if ((dev_state->state->print_mask &
@@ -2836,6 +3085,7 @@ void btrfsic_submit_bio(int rw, struct bio *bio)
unsigned int i;
u64 dev_bytenr;
int bio_is_patched;
+ char **mapped_datav;
dev_bytenr = 512 * bio->bi_sector;
bio_is_patched = 0;
@@ -2848,35 +3098,46 @@ void btrfsic_submit_bio(int rw, struct bio *bio)
(unsigned long long)dev_bytenr,
bio->bi_bdev);
+ mapped_datav = kmalloc(sizeof(*mapped_datav) * bio->bi_vcnt,
+ GFP_NOFS);
+ if (!mapped_datav)
+ goto leave;
for (i = 0; i < bio->bi_vcnt; i++) {
- u8 *mapped_data;
-
- mapped_data = kmap(bio->bi_io_vec[i].bv_page);
+ BUG_ON(bio->bi_io_vec[i].bv_len != PAGE_CACHE_SIZE);
+ mapped_datav[i] = kmap(bio->bi_io_vec[i].bv_page);
+ if (!mapped_datav[i]) {
+ while (i > 0) {
+ i--;
+ kunmap(bio->bi_io_vec[i].bv_page);
+ }
+ kfree(mapped_datav);
+ goto leave;
+ }
if ((BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH |
BTRFSIC_PRINT_MASK_VERBOSE) ==
(dev_state->state->print_mask &
(BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH |
BTRFSIC_PRINT_MASK_VERBOSE)))
printk(KERN_INFO
- "#%u: page=%p, mapped=%p, len=%u,"
- " offset=%u\n",
+ "#%u: page=%p, len=%u, offset=%u\n",
i, bio->bi_io_vec[i].bv_page,
- mapped_data,
bio->bi_io_vec[i].bv_len,
bio->bi_io_vec[i].bv_offset);
- btrfsic_process_written_block(dev_state, dev_bytenr,
- mapped_data,
- bio->bi_io_vec[i].bv_len,
- bio, &bio_is_patched,
- NULL, rw);
+ }
+ btrfsic_process_written_block(dev_state, dev_bytenr,
+ mapped_datav, bio->bi_vcnt,
+ bio, &bio_is_patched,
+ NULL, rw);
+ while (i > 0) {
+ i--;
kunmap(bio->bi_io_vec[i].bv_page);
- dev_bytenr += bio->bi_io_vec[i].bv_len;
}
+ kfree(mapped_datav);
} else if (NULL != dev_state && (rw & REQ_FLUSH)) {
if (dev_state->state->print_mask &
BTRFSIC_PRINT_MASK_SUBMIT_BIO_BH)
printk(KERN_INFO
- "submit_bio(rw=0x%x) FLUSH, bdev=%p)\n",
+ "submit_bio(rw=0x%x FLUSH, bdev=%p)\n",
rw, bio->bi_bdev);
if (!dev_state->dummy_block_for_bio_bh_flush.is_iodone) {
if ((dev_state->state->print_mask &
@@ -2903,6 +3164,7 @@ void btrfsic_submit_bio(int rw, struct bio *bio)
bio->bi_end_io = btrfsic_bio_end_io;
}
}
+leave:
mutex_unlock(&btrfsic_mutex);
submit_bio(rw, bio);
@@ -2917,6 +3179,30 @@ int btrfsic_mount(struct btrfs_root *root,
struct list_head *dev_head = &fs_devices->devices;
struct btrfs_device *device;
+ if (root->nodesize != root->leafsize) {
+ printk(KERN_INFO
+ "btrfsic: cannot handle nodesize %d != leafsize %d!\n",
+ root->nodesize, root->leafsize);
+ return -1;
+ }
+ if (root->nodesize & ((u64)PAGE_CACHE_SIZE - 1)) {
+ printk(KERN_INFO
+ "btrfsic: cannot handle nodesize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
+ root->nodesize, (unsigned long)PAGE_CACHE_SIZE);
+ return -1;
+ }
+ if (root->leafsize & ((u64)PAGE_CACHE_SIZE - 1)) {
+ printk(KERN_INFO
+ "btrfsic: cannot handle leafsize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
+ root->leafsize, (unsigned long)PAGE_CACHE_SIZE);
+ return -1;
+ }
+ if (root->sectorsize & ((u64)PAGE_CACHE_SIZE - 1)) {
+ printk(KERN_INFO
+ "btrfsic: cannot handle sectorsize %d not being a multiple of PAGE_CACHE_SIZE %ld!\n",
+ root->sectorsize, (unsigned long)PAGE_CACHE_SIZE);
+ return -1;
+ }
state = kzalloc(sizeof(*state), GFP_NOFS);
if (NULL == state) {
printk(KERN_INFO "btrfs check-integrity: kmalloc() failed!\n");
@@ -2933,6 +3219,8 @@ int btrfsic_mount(struct btrfs_root *root,
state->print_mask = print_mask;
state->include_extent_data = including_extent_data;
state->csum_size = 0;
+ state->metablock_size = root->nodesize;
+ state->datablock_size = root->sectorsize;
INIT_LIST_HEAD(&state->all_blocks_list);
btrfsic_block_hashtable_init(&state->block_hashtable);
btrfsic_block_link_hashtable_init(&state->block_link_hashtable);
@@ -3049,7 +3337,7 @@ void btrfsic_unmount(struct btrfs_root *root,
btrfsic_block_link_free(l);
}
- if (b_all->is_iodone)
+ if (b_all->is_iodone || b_all->never_written)
btrfsic_block_free(b_all);
else
printk(KERN_INFO "btrfs: attempt to free %c-block"
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index 4106264fbc6..d7a96cfdc50 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -18,6 +18,7 @@
#include <linux/sched.h>
#include <linux/slab.h>
+#include <linux/rbtree.h>
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
@@ -37,7 +38,16 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
struct extent_buffer *dst_buf,
struct extent_buffer *src_buf);
static void del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
- struct btrfs_path *path, int level, int slot);
+ struct btrfs_path *path, int level, int slot,
+ int tree_mod_log);
+static void tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *eb);
+struct extent_buffer *read_old_tree_block(struct btrfs_root *root, u64 bytenr,
+ u32 blocksize, u64 parent_transid,
+ u64 time_seq);
+struct extent_buffer *btrfs_find_old_tree_block(struct btrfs_root *root,
+ u64 bytenr, u32 blocksize,
+ u64 time_seq);
struct btrfs_path *btrfs_alloc_path(void)
{
@@ -255,7 +265,7 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans,
cow = btrfs_alloc_free_block(trans, root, buf->len, 0,
new_root_objectid, &disk_key, level,
- buf->start, 0, 1);
+ buf->start, 0);
if (IS_ERR(cow))
return PTR_ERR(cow);
@@ -288,6 +298,434 @@ int btrfs_copy_root(struct btrfs_trans_handle *trans,
return 0;
}
+enum mod_log_op {
+ MOD_LOG_KEY_REPLACE,
+ MOD_LOG_KEY_ADD,
+ MOD_LOG_KEY_REMOVE,
+ MOD_LOG_KEY_REMOVE_WHILE_FREEING,
+ MOD_LOG_KEY_REMOVE_WHILE_MOVING,
+ MOD_LOG_MOVE_KEYS,
+ MOD_LOG_ROOT_REPLACE,
+};
+
+struct tree_mod_move {
+ int dst_slot;
+ int nr_items;
+};
+
+struct tree_mod_root {
+ u64 logical;
+ u8 level;
+};
+
+struct tree_mod_elem {
+ struct rb_node node;
+ u64 index; /* shifted logical */
+ struct seq_list elem;
+ enum mod_log_op op;
+
+ /* this is used for MOD_LOG_KEY_* and MOD_LOG_MOVE_KEYS operations */
+ int slot;
+
+ /* this is used for MOD_LOG_KEY* and MOD_LOG_ROOT_REPLACE */
+ u64 generation;
+
+ /* those are used for op == MOD_LOG_KEY_{REPLACE,REMOVE} */
+ struct btrfs_disk_key key;
+ u64 blockptr;
+
+ /* this is used for op == MOD_LOG_MOVE_KEYS */
+ struct tree_mod_move move;
+
+ /* this is used for op == MOD_LOG_ROOT_REPLACE */
+ struct tree_mod_root old_root;
+};
+
+static inline void
+__get_tree_mod_seq(struct btrfs_fs_info *fs_info, struct seq_list *elem)
+{
+ elem->seq = atomic_inc_return(&fs_info->tree_mod_seq);
+ list_add_tail(&elem->list, &fs_info->tree_mod_seq_list);
+}
+
+void btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info,
+ struct seq_list *elem)
+{
+ elem->flags = 1;
+ spin_lock(&fs_info->tree_mod_seq_lock);
+ __get_tree_mod_seq(fs_info, elem);
+ spin_unlock(&fs_info->tree_mod_seq_lock);
+}
+
+void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,
+ struct seq_list *elem)
+{
+ struct rb_root *tm_root;
+ struct rb_node *node;
+ struct rb_node *next;
+ struct seq_list *cur_elem;
+ struct tree_mod_elem *tm;
+ u64 min_seq = (u64)-1;
+ u64 seq_putting = elem->seq;
+
+ if (!seq_putting)
+ return;
+
+ BUG_ON(!(elem->flags & 1));
+ spin_lock(&fs_info->tree_mod_seq_lock);
+ list_del(&elem->list);
+
+ list_for_each_entry(cur_elem, &fs_info->tree_mod_seq_list, list) {
+ if ((cur_elem->flags & 1) && cur_elem->seq < min_seq) {
+ if (seq_putting > cur_elem->seq) {
+ /*
+ * blocker with lower sequence number exists, we
+ * cannot remove anything from the log
+ */
+ goto out;
+ }
+ min_seq = cur_elem->seq;
+ }
+ }
+
+ /*
+ * anything that's lower than the lowest existing (read: blocked)
+ * sequence number can be removed from the tree.
+ */
+ write_lock(&fs_info->tree_mod_log_lock);
+ tm_root = &fs_info->tree_mod_log;
+ for (node = rb_first(tm_root); node; node = next) {
+ next = rb_next(node);
+ tm = container_of(node, struct tree_mod_elem, node);
+ if (tm->elem.seq > min_seq)
+ continue;
+ rb_erase(node, tm_root);
+ list_del(&tm->elem.list);
+ kfree(tm);
+ }
+ write_unlock(&fs_info->tree_mod_log_lock);
+out:
+ spin_unlock(&fs_info->tree_mod_seq_lock);
+}
+
+/*
+ * key order of the log:
+ * index -> sequence
+ *
+ * the index is the shifted logical of the *new* root node for root replace
+ * operations, or the shifted logical of the affected block for all other
+ * operations.
+ */
+static noinline int
+__tree_mod_log_insert(struct btrfs_fs_info *fs_info, struct tree_mod_elem *tm)
+{
+ struct rb_root *tm_root;
+ struct rb_node **new;
+ struct rb_node *parent = NULL;
+ struct tree_mod_elem *cur;
+ int ret = 0;
+
+ BUG_ON(!tm || !tm->elem.seq);
+
+ write_lock(&fs_info->tree_mod_log_lock);
+ tm_root = &fs_info->tree_mod_log;
+ new = &tm_root->rb_node;
+ while (*new) {
+ cur = container_of(*new, struct tree_mod_elem, node);
+ parent = *new;
+ if (cur->index < tm->index)
+ new = &((*new)->rb_left);
+ else if (cur->index > tm->index)
+ new = &((*new)->rb_right);
+ else if (cur->elem.seq < tm->elem.seq)
+ new = &((*new)->rb_left);
+ else if (cur->elem.seq > tm->elem.seq)
+ new = &((*new)->rb_right);
+ else {
+ kfree(tm);
+ ret = -EEXIST;
+ goto unlock;
+ }
+ }
+
+ rb_link_node(&tm->node, parent, new);
+ rb_insert_color(&tm->node, tm_root);
+unlock:
+ write_unlock(&fs_info->tree_mod_log_lock);
+ return ret;
+}
+
+static inline int tree_mod_dont_log(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *eb) {
+ smp_mb();
+ if (list_empty(&(fs_info)->tree_mod_seq_list))
+ return 1;
+ if (!eb)
+ return 0;
+ if (btrfs_header_level(eb) == 0)
+ return 1;
+ return 0;
+}
+
+static inline int tree_mod_alloc(struct btrfs_fs_info *fs_info, gfp_t flags,
+ struct tree_mod_elem **tm_ret)
+{
+ struct tree_mod_elem *tm;
+ int seq;
+
+ if (tree_mod_dont_log(fs_info, NULL))
+ return 0;
+
+ tm = *tm_ret = kzalloc(sizeof(*tm), flags);
+ if (!tm)
+ return -ENOMEM;
+
+ tm->elem.flags = 0;
+ spin_lock(&fs_info->tree_mod_seq_lock);
+ if (list_empty(&fs_info->tree_mod_seq_list)) {
+ /*
+ * someone emptied the list while we were waiting for the lock.
+ * we must not add to the list, because no blocker exists. items
+ * are removed from the list only when the existing blocker is
+ * removed from the list.
+ */
+ kfree(tm);
+ seq = 0;
+ } else {
+ __get_tree_mod_seq(fs_info, &tm->elem);
+ seq = tm->elem.seq;
+ }
+ spin_unlock(&fs_info->tree_mod_seq_lock);
+
+ return seq;
+}
+
+static noinline int
+tree_mod_log_insert_key_mask(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *eb, int slot,
+ enum mod_log_op op, gfp_t flags)
+{
+ struct tree_mod_elem *tm;
+ int ret;
+
+ ret = tree_mod_alloc(fs_info, flags, &tm);
+ if (ret <= 0)
+ return ret;
+
+ tm->index = eb->start >> PAGE_CACHE_SHIFT;
+ if (op != MOD_LOG_KEY_ADD) {
+ btrfs_node_key(eb, &tm->key, slot);
+ tm->blockptr = btrfs_node_blockptr(eb, slot);
+ }
+ tm->op = op;
+ tm->slot = slot;
+ tm->generation = btrfs_node_ptr_generation(eb, slot);
+
+ return __tree_mod_log_insert(fs_info, tm);
+}
+
+static noinline int
+tree_mod_log_insert_key(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
+ int slot, enum mod_log_op op)
+{
+ return tree_mod_log_insert_key_mask(fs_info, eb, slot, op, GFP_NOFS);
+}
+
+static noinline int
+tree_mod_log_insert_move(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *eb, int dst_slot, int src_slot,
+ int nr_items, gfp_t flags)
+{
+ struct tree_mod_elem *tm;
+ int ret;
+ int i;
+
+ if (tree_mod_dont_log(fs_info, eb))
+ return 0;
+
+ for (i = 0; i + dst_slot < src_slot && i < nr_items; i++) {
+ ret = tree_mod_log_insert_key(fs_info, eb, i + dst_slot,
+ MOD_LOG_KEY_REMOVE_WHILE_MOVING);
+ BUG_ON(ret < 0);
+ }
+
+ ret = tree_mod_alloc(fs_info, flags, &tm);
+ if (ret <= 0)
+ return ret;
+
+ tm->index = eb->start >> PAGE_CACHE_SHIFT;
+ tm->slot = src_slot;
+ tm->move.dst_slot = dst_slot;
+ tm->move.nr_items = nr_items;
+ tm->op = MOD_LOG_MOVE_KEYS;
+
+ return __tree_mod_log_insert(fs_info, tm);
+}
+
+static noinline int
+tree_mod_log_insert_root(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *old_root,
+ struct extent_buffer *new_root, gfp_t flags)
+{
+ struct tree_mod_elem *tm;
+ int ret;
+
+ ret = tree_mod_alloc(fs_info, flags, &tm);
+ if (ret <= 0)
+ return ret;
+
+ tm->index = new_root->start >> PAGE_CACHE_SHIFT;
+ tm->old_root.logical = old_root->start;
+ tm->old_root.level = btrfs_header_level(old_root);
+ tm->generation = btrfs_header_generation(old_root);
+ tm->op = MOD_LOG_ROOT_REPLACE;
+
+ return __tree_mod_log_insert(fs_info, tm);
+}
+
+static struct tree_mod_elem *
+__tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq,
+ int smallest)
+{
+ struct rb_root *tm_root;
+ struct rb_node *node;
+ struct tree_mod_elem *cur = NULL;
+ struct tree_mod_elem *found = NULL;
+ u64 index = start >> PAGE_CACHE_SHIFT;
+
+ read_lock(&fs_info->tree_mod_log_lock);
+ tm_root = &fs_info->tree_mod_log;
+ node = tm_root->rb_node;
+ while (node) {
+ cur = container_of(node, struct tree_mod_elem, node);
+ if (cur->index < index) {
+ node = node->rb_left;
+ } else if (cur->index > index) {
+ node = node->rb_right;
+ } else if (cur->elem.seq < min_seq) {
+ node = node->rb_left;
+ } else if (!smallest) {
+ /* we want the node with the highest seq */
+ if (found)
+ BUG_ON(found->elem.seq > cur->elem.seq);
+ found = cur;
+ node = node->rb_left;
+ } else if (cur->elem.seq > min_seq) {
+ /* we want the node with the smallest seq */
+ if (found)
+ BUG_ON(found->elem.seq < cur->elem.seq);
+ found = cur;
+ node = node->rb_right;
+ } else {
+ found = cur;
+ break;
+ }
+ }
+ read_unlock(&fs_info->tree_mod_log_lock);
+
+ return found;
+}
+
+/*
+ * this returns the element from the log with the smallest time sequence
+ * value that's in the log (the oldest log item). any element with a time
+ * sequence lower than min_seq will be ignored.
+ */
+static struct tree_mod_elem *
+tree_mod_log_search_oldest(struct btrfs_fs_info *fs_info, u64 start,
+ u64 min_seq)
+{
+ return __tree_mod_log_search(fs_info, start, min_seq, 1);
+}
+
+/*
+ * this returns the element from the log with the largest time sequence
+ * value that's in the log (the most recent log item). any element with
+ * a time sequence lower than min_seq will be ignored.
+ */
+static struct tree_mod_elem *
+tree_mod_log_search(struct btrfs_fs_info *fs_info, u64 start, u64 min_seq)
+{
+ return __tree_mod_log_search(fs_info, start, min_seq, 0);
+}
+
+static inline void
+tree_mod_log_eb_copy(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
+ struct extent_buffer *src, unsigned long dst_offset,
+ unsigned long src_offset, int nr_items)
+{
+ int ret;
+ int i;
+
+ if (tree_mod_dont_log(fs_info, NULL))
+ return;
+
+ if (btrfs_header_level(dst) == 0 && btrfs_header_level(src) == 0)
+ return;
+
+ /* speed this up by single seq for all operations? */
+ for (i = 0; i < nr_items; i++) {
+ ret = tree_mod_log_insert_key(fs_info, src, i + src_offset,
+ MOD_LOG_KEY_REMOVE);
+ BUG_ON(ret < 0);
+ ret = tree_mod_log_insert_key(fs_info, dst, i + dst_offset,
+ MOD_LOG_KEY_ADD);
+ BUG_ON(ret < 0);
+ }
+}
+
+static inline void
+tree_mod_log_eb_move(struct btrfs_fs_info *fs_info, struct extent_buffer *dst,
+ int dst_offset, int src_offset, int nr_items)
+{
+ int ret;
+ ret = tree_mod_log_insert_move(fs_info, dst, dst_offset, src_offset,
+ nr_items, GFP_NOFS);
+ BUG_ON(ret < 0);
+}
+
+static inline void
+tree_mod_log_set_node_key(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *eb,
+ struct btrfs_disk_key *disk_key, int slot, int atomic)
+{
+ int ret;
+
+ ret = tree_mod_log_insert_key_mask(fs_info, eb, slot,
+ MOD_LOG_KEY_REPLACE,
+ atomic ? GFP_ATOMIC : GFP_NOFS);
+ BUG_ON(ret < 0);
+}
+
+static void tree_mod_log_free_eb(struct btrfs_fs_info *fs_info,
+ struct extent_buffer *eb)
+{
+ int i;
+ int ret;
+ u32 nritems;
+
+ if (tree_mod_dont_log(fs_info, eb))
+ return;
+
+ nritems = btrfs_header_nritems(eb);
+ for (i = nritems - 1; i >= 0; i--) {
+ ret = tree_mod_log_insert_key(fs_info, eb, i,
+ MOD_LOG_KEY_REMOVE_WHILE_FREEING);
+ BUG_ON(ret < 0);
+ }
+}
+
+static inline void
+tree_mod_log_set_root_pointer(struct btrfs_root *root,
+ struct extent_buffer *new_root_node)
+{
+ int ret;
+ tree_mod_log_free_eb(root->fs_info, root->node);
+ ret = tree_mod_log_insert_root(root->fs_info, root->node,
+ new_root_node, GFP_NOFS);
+ BUG_ON(ret < 0);
+}
+
/*
* check if the tree block can be shared by multiple trees
*/
@@ -409,6 +847,12 @@ static noinline int update_ref_for_cow(struct btrfs_trans_handle *trans,
ret = btrfs_dec_ref(trans, root, buf, 1, 1);
BUG_ON(ret); /* -ENOMEM */
}
+ /*
+ * don't log freeing in case we're freeing the root node, this
+ * is done by tree_mod_log_set_root_pointer later
+ */
+ if (buf != root->node && btrfs_header_level(buf) != 0)
+ tree_mod_log_free_eb(root->fs_info, buf);
clean_tree_block(trans, root, buf);
*last_ref = 1;
}
@@ -467,7 +911,7 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
cow = btrfs_alloc_free_block(trans, root, buf->len, parent_start,
root->root_key.objectid, &disk_key,
- level, search_start, empty_size, 1);
+ level, search_start, empty_size);
if (IS_ERR(cow))
return PTR_ERR(cow);
@@ -506,10 +950,11 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
parent_start = 0;
extent_buffer_get(cow);
+ tree_mod_log_set_root_pointer(root, cow);
rcu_assign_pointer(root->node, cow);
btrfs_free_tree_block(trans, root, buf, parent_start,
- last_ref, 1);
+ last_ref);
free_extent_buffer(buf);
add_root_to_dirty_list(root);
} else {
@@ -519,13 +964,15 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
parent_start = 0;
WARN_ON(trans->transid != btrfs_header_generation(parent));
+ tree_mod_log_insert_key(root->fs_info, parent, parent_slot,
+ MOD_LOG_KEY_REPLACE);
btrfs_set_node_blockptr(parent, parent_slot,
cow->start);
btrfs_set_node_ptr_generation(parent, parent_slot,
trans->transid);
btrfs_mark_buffer_dirty(parent);
btrfs_free_tree_block(trans, root, buf, parent_start,
- last_ref, 1);
+ last_ref);
}
if (unlock_orig)
btrfs_tree_unlock(buf);
@@ -535,6 +982,210 @@ static noinline int __btrfs_cow_block(struct btrfs_trans_handle *trans,
return 0;
}
+/*
+ * returns the logical address of the oldest predecessor of the given root.
+ * entries older than time_seq are ignored.
+ */
+static struct tree_mod_elem *
+__tree_mod_log_oldest_root(struct btrfs_fs_info *fs_info,
+ struct btrfs_root *root, u64 time_seq)
+{
+ struct tree_mod_elem *tm;
+ struct tree_mod_elem *found = NULL;
+ u64 root_logical = root->node->start;
+ int looped = 0;
+
+ if (!time_seq)
+ return 0;
+
+ /*
+ * the very last operation that's logged for a root is the replacement
+ * operation (if it is replaced at all). this has the index of the *new*
+ * root, making it the very first operation that's logged for this root.
+ */
+ while (1) {
+ tm = tree_mod_log_search_oldest(fs_info, root_logical,
+ time_seq);
+ if (!looped && !tm)
+ return 0;
+ /*
+ * we must have key remove operations in the log before the
+ * replace operation.
+ */
+ BUG_ON(!tm);
+
+ if (tm->op != MOD_LOG_ROOT_REPLACE)
+ break;
+
+ found = tm;
+ root_logical = tm->old_root.logical;
+ BUG_ON(root_logical == root->node->start);
+ looped = 1;
+ }
+
+ return found;
+}
+
+/*
+ * tm is a pointer to the first operation to rewind within eb. then, all
+ * previous operations will be rewinded (until we reach something older than
+ * time_seq).
+ */
+static void
+__tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq,
+ struct tree_mod_elem *first_tm)
+{
+ u32 n;
+ struct rb_node *next;
+ struct tree_mod_elem *tm = first_tm;
+ unsigned long o_dst;
+ unsigned long o_src;
+ unsigned long p_size = sizeof(struct btrfs_key_ptr);
+
+ n = btrfs_header_nritems(eb);
+ while (tm && tm->elem.seq >= time_seq) {
+ /*
+ * all the operations are recorded with the operator used for
+ * the modification. as we're going backwards, we do the
+ * opposite of each operation here.
+ */
+ switch (tm->op) {
+ case MOD_LOG_KEY_REMOVE_WHILE_FREEING:
+ BUG_ON(tm->slot < n);
+ case MOD_LOG_KEY_REMOVE_WHILE_MOVING:
+ case MOD_LOG_KEY_REMOVE:
+ btrfs_set_node_key(eb, &tm->key, tm->slot);
+ btrfs_set_node_blockptr(eb, tm->slot, tm->blockptr);
+ btrfs_set_node_ptr_generation(eb, tm->slot,
+ tm->generation);
+ n++;
+ break;
+ case MOD_LOG_KEY_REPLACE:
+ BUG_ON(tm->slot >= n);
+ btrfs_set_node_key(eb, &tm->key, tm->slot);
+ btrfs_set_node_blockptr(eb, tm->slot, tm->blockptr);
+ btrfs_set_node_ptr_generation(eb, tm->slot,
+ tm->generation);
+ break;
+ case MOD_LOG_KEY_ADD:
+ if (tm->slot != n - 1) {
+ o_dst = btrfs_node_key_ptr_offset(tm->slot);
+ o_src = btrfs_node_key_ptr_offset(tm->slot + 1);
+ memmove_extent_buffer(eb, o_dst, o_src, p_size);
+ }
+ n--;
+ break;
+ case MOD_LOG_MOVE_KEYS:
+ o_dst = btrfs_node_key_ptr_offset(tm->slot);
+ o_src = btrfs_node_key_ptr_offset(tm->move.dst_slot);
+ memmove_extent_buffer(eb, o_dst, o_src,
+ tm->move.nr_items * p_size);
+ break;
+ case MOD_LOG_ROOT_REPLACE:
+ /*
+ * this operation is special. for roots, this must be
+ * handled explicitly before rewinding.
+ * for non-roots, this operation may exist if the node
+ * was a root: root A -> child B; then A gets empty and
+ * B is promoted to the new root. in the mod log, we'll
+ * have a root-replace operation for B, a tree block
+ * that is no root. we simply ignore that operation.
+ */
+ break;
+ }
+ next = rb_next(&tm->node);
+ if (!next)
+ break;
+ tm = container_of(next, struct tree_mod_elem, node);
+ if (tm->index != first_tm->index)
+ break;
+ }
+ btrfs_set_header_nritems(eb, n);
+}
+
+static struct extent_buffer *
+tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
+ u64 time_seq)
+{
+ struct extent_buffer *eb_rewin;
+ struct tree_mod_elem *tm;
+
+ if (!time_seq)
+ return eb;
+
+ if (btrfs_header_level(eb) == 0)
+ return eb;
+
+ tm = tree_mod_log_search(fs_info, eb->start, time_seq);
+ if (!tm)
+ return eb;
+
+ if (tm->op == MOD_LOG_KEY_REMOVE_WHILE_FREEING) {
+ BUG_ON(tm->slot != 0);
+ eb_rewin = alloc_dummy_extent_buffer(eb->start,
+ fs_info->tree_root->nodesize);
+ BUG_ON(!eb_rewin);
+ btrfs_set_header_bytenr(eb_rewin, eb->start);
+ btrfs_set_header_backref_rev(eb_rewin,
+ btrfs_header_backref_rev(eb));
+ btrfs_set_header_owner(eb_rewin, btrfs_header_owner(eb));
+ btrfs_set_header_level(eb_rewin, btrfs_header_level(eb));
+ } else {
+ eb_rewin = btrfs_clone_extent_buffer(eb);
+ BUG_ON(!eb_rewin);
+ }
+
+ extent_buffer_get(eb_rewin);
+ free_extent_buffer(eb);
+
+ __tree_mod_log_rewind(eb_rewin, time_seq, tm);
+
+ return eb_rewin;
+}
+
+static inline struct extent_buffer *
+get_old_root(struct btrfs_root *root, u64 time_seq)
+{
+ struct tree_mod_elem *tm;
+ struct extent_buffer *eb;
+ struct tree_mod_root *old_root;
+ u64 old_generation;
+
+ tm = __tree_mod_log_oldest_root(root->fs_info, root, time_seq);
+ if (!tm)
+ return root->node;
+
+ old_root = &tm->old_root;
+ old_generation = tm->generation;
+
+ tm = tree_mod_log_search(root->fs_info, old_root->logical, time_seq);
+ /*
+ * there was an item in the log when __tree_mod_log_oldest_root
+ * returned. this one must not go away, because the time_seq passed to
+ * us must be blocking its removal.
+ */
+ BUG_ON(!tm);
+
+ if (old_root->logical == root->node->start) {
+ /* there are logged operations for the current root */
+ eb = btrfs_clone_extent_buffer(root->node);
+ } else {
+ /* there's a root replace operation for the current root */
+ eb = alloc_dummy_extent_buffer(tm->index << PAGE_CACHE_SHIFT,
+ root->nodesize);
+ btrfs_set_header_bytenr(eb, eb->start);
+ btrfs_set_header_backref_rev(eb, BTRFS_MIXED_BACKREF_REV);
+ btrfs_set_header_owner(eb, root->root_key.objectid);
+ }
+ if (!eb)
+ return NULL;
+ btrfs_set_header_level(eb, old_root->level);
+ btrfs_set_header_generation(eb, old_generation);
+ __tree_mod_log_rewind(eb, time_seq, tm);
+
+ return eb;
+}
+
static inline int should_cow_block(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct extent_buffer *buf)
@@ -739,7 +1390,11 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans,
if (!cur)
return -EIO;
} else if (!uptodate) {
- btrfs_read_buffer(cur, gen);
+ err = btrfs_read_buffer(cur, gen);
+ if (err) {
+ free_extent_buffer(cur);
+ return err;
+ }
}
}
if (search_start == 0)
@@ -854,20 +1509,18 @@ static noinline int generic_bin_search(struct extent_buffer *eb,
static int bin_search(struct extent_buffer *eb, struct btrfs_key *key,
int level, int *slot)
{
- if (level == 0) {
+ if (level == 0)
return generic_bin_search(eb,
offsetof(struct btrfs_leaf, items),
sizeof(struct btrfs_item),
key, btrfs_header_nritems(eb),
slot);
- } else {
+ else
return generic_bin_search(eb,
offsetof(struct btrfs_node, ptrs),
sizeof(struct btrfs_key_ptr),
key, btrfs_header_nritems(eb),
slot);
- }
- return -1;
}
int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key,
@@ -974,6 +1627,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
goto enospc;
}
+ tree_mod_log_set_root_pointer(root, child);
rcu_assign_pointer(root->node, child);
add_root_to_dirty_list(root);
@@ -987,7 +1641,7 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
free_extent_buffer(mid);
root_sub_used(root, mid->len);
- btrfs_free_tree_block(trans, root, mid, 0, 1, 0);
+ btrfs_free_tree_block(trans, root, mid, 0, 1);
/* once for the root ptr */
free_extent_buffer_stale(mid);
return 0;
@@ -1040,14 +1694,16 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
if (btrfs_header_nritems(right) == 0) {
clean_tree_block(trans, root, right);
btrfs_tree_unlock(right);
- del_ptr(trans, root, path, level + 1, pslot + 1);
+ del_ptr(trans, root, path, level + 1, pslot + 1, 1);
root_sub_used(root, right->len);
- btrfs_free_tree_block(trans, root, right, 0, 1, 0);
+ btrfs_free_tree_block(trans, root, right, 0, 1);
free_extent_buffer_stale(right);
right = NULL;
} else {
struct btrfs_disk_key right_key;
btrfs_node_key(right, &right_key, 0);
+ tree_mod_log_set_node_key(root->fs_info, parent,
+ &right_key, pslot + 1, 0);
btrfs_set_node_key(parent, &right_key, pslot + 1);
btrfs_mark_buffer_dirty(parent);
}
@@ -1082,15 +1738,17 @@ static noinline int balance_level(struct btrfs_trans_handle *trans,
if (btrfs_header_nritems(mid) == 0) {
clean_tree_block(trans, root, mid);
btrfs_tree_unlock(mid);
- del_ptr(trans, root, path, level + 1, pslot);
+ del_ptr(trans, root, path, level + 1, pslot, 1);
root_sub_used(root, mid->len);
- btrfs_free_tree_block(trans, root, mid, 0, 1, 0);
+ btrfs_free_tree_block(trans, root, mid, 0, 1);
free_extent_buffer_stale(mid);
mid = NULL;
} else {
/* update the parent key to reflect our changes */
struct btrfs_disk_key mid_key;
btrfs_node_key(mid, &mid_key, 0);
+ tree_mod_log_set_node_key(root->fs_info, parent, &mid_key,
+ pslot, 0);
btrfs_set_node_key(parent, &mid_key, pslot);
btrfs_mark_buffer_dirty(parent);
}
@@ -1188,6 +1846,8 @@ static noinline int push_nodes_for_insert(struct btrfs_trans_handle *trans,
struct btrfs_disk_key disk_key;
orig_slot += left_nr;
btrfs_node_key(mid, &disk_key, 0);
+ tree_mod_log_set_node_key(root->fs_info, parent,
+ &disk_key, pslot, 0);
btrfs_set_node_key(parent, &disk_key, pslot);
btrfs_mark_buffer_dirty(parent);
if (btrfs_header_nritems(left) > orig_slot) {
@@ -1239,6 +1899,8 @@ static noinline int push_nodes_for_insert(struct btrfs_trans_handle *trans,
struct btrfs_disk_key disk_key;
btrfs_node_key(right, &disk_key, 0);
+ tree_mod_log_set_node_key(root->fs_info, parent,
+ &disk_key, pslot + 1, 0);
btrfs_set_node_key(parent, &disk_key, pslot + 1);
btrfs_mark_buffer_dirty(parent);
@@ -1496,7 +2158,7 @@ static int
read_block_for_search(struct btrfs_trans_handle *trans,
struct btrfs_root *root, struct btrfs_path *p,
struct extent_buffer **eb_ret, int level, int slot,
- struct btrfs_key *key)
+ struct btrfs_key *key, u64 time_seq)
{
u64 blocknr;
u64 gen;
@@ -1850,7 +2512,7 @@ cow_done:
}
err = read_block_for_search(trans, root, p,
- &b, level, slot, key);
+ &b, level, slot, key, 0);
if (err == -EAGAIN)
goto again;
if (err) {
@@ -1922,6 +2584,115 @@ done:
}
/*
+ * Like btrfs_search_slot, this looks for a key in the given tree. It uses the
+ * current state of the tree together with the operations recorded in the tree
+ * modification log to search for the key in a previous version of this tree, as
+ * denoted by the time_seq parameter.
+ *
+ * Naturally, there is no support for insert, delete or cow operations.
+ *
+ * The resulting path and return value will be set up as if we called
+ * btrfs_search_slot at that point in time with ins_len and cow both set to 0.
+ */
+int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
+ struct btrfs_path *p, u64 time_seq)
+{
+ struct extent_buffer *b;
+ int slot;
+ int ret;
+ int err;
+ int level;
+ int lowest_unlock = 1;
+ u8 lowest_level = 0;
+
+ lowest_level = p->lowest_level;
+ WARN_ON(p->nodes[0] != NULL);
+
+ if (p->search_commit_root) {
+ BUG_ON(time_seq);
+ return btrfs_search_slot(NULL, root, key, p, 0, 0);
+ }
+
+again:
+ b = get_old_root(root, time_seq);
+ extent_buffer_get(b);
+ level = btrfs_header_level(b);
+ btrfs_tree_read_lock(b);
+ p->locks[level] = BTRFS_READ_LOCK;
+
+ while (b) {
+ level = btrfs_header_level(b);
+ p->nodes[level] = b;
+ btrfs_clear_path_blocking(p, NULL, 0);
+
+ /*
+ * we have a lock on b and as long as we aren't changing
+ * the tree, there is no way to for the items in b to change.
+ * It is safe to drop the lock on our parent before we
+ * go through the expensive btree search on b.
+ */
+ btrfs_unlock_up_safe(p, level + 1);
+
+ ret = bin_search(b, key, level, &slot);
+
+ if (level != 0) {
+ int dec = 0;
+ if (ret && slot > 0) {
+ dec = 1;
+ slot -= 1;
+ }
+ p->slots[level] = slot;
+ unlock_up(p, level, lowest_unlock, 0, NULL);
+
+ if (level == lowest_level) {
+ if (dec)
+ p->slots[level]++;
+ goto done;
+ }
+
+ err = read_block_for_search(NULL, root, p, &b, level,
+ slot, key, time_seq);
+ if (err == -EAGAIN)
+ goto again;
+ if (err) {
+ ret = err;
+ goto done;
+ }
+
+ level = btrfs_header_level(b);
+ err = btrfs_try_tree_read_lock(b);
+ if (!err) {
+ btrfs_set_path_blocking(p);
+ btrfs_tree_read_lock(b);
+ btrfs_clear_path_blocking(p, b,
+ BTRFS_READ_LOCK);
+ }
+ p->locks[level] = BTRFS_READ_LOCK;
+ p->nodes[level] = b;
+ b = tree_mod_log_rewind(root->fs_info, b, time_seq);
+ if (b != p->nodes[level]) {
+ btrfs_tree_unlock_rw(p->nodes[level],
+ p->locks[level]);
+ p->locks[level] = 0;
+ p->nodes[level] = b;
+ }
+ } else {
+ p->slots[level] = slot;
+ unlock_up(p, level, lowest_unlock, 0, NULL);
+ goto done;
+ }
+ }
+ ret = 1;
+done:
+ if (!p->leave_spinning)
+ btrfs_set_path_blocking(p);
+ if (ret < 0)
+ btrfs_release_path(p);
+
+ return ret;
+}
+
+/*
* adjust the pointers going up the tree, starting at level
* making sure the right key of each node is points to 'key'.
* This is used after shifting pointers to the left, so it stops
@@ -1941,6 +2712,7 @@ static void fixup_low_keys(struct btrfs_trans_handle *trans,
if (!path->nodes[i])
break;
t = path->nodes[i];
+ tree_mod_log_set_node_key(root->fs_info, t, key, tslot, 1);
btrfs_set_node_key(t, key, tslot);
btrfs_mark_buffer_dirty(path->nodes[i]);
if (tslot != 0)
@@ -2023,12 +2795,16 @@ static int push_node_left(struct btrfs_trans_handle *trans,
} else
push_items = min(src_nritems - 8, push_items);
+ tree_mod_log_eb_copy(root->fs_info, dst, src, dst_nritems, 0,
+ push_items);
copy_extent_buffer(dst, src,
btrfs_node_key_ptr_offset(dst_nritems),
btrfs_node_key_ptr_offset(0),
push_items * sizeof(struct btrfs_key_ptr));
if (push_items < src_nritems) {
+ tree_mod_log_eb_move(root->fs_info, src, 0, push_items,
+ src_nritems - push_items);
memmove_extent_buffer(src, btrfs_node_key_ptr_offset(0),
btrfs_node_key_ptr_offset(push_items),
(src_nritems - push_items) *
@@ -2082,11 +2858,14 @@ static int balance_node_right(struct btrfs_trans_handle *trans,
if (max_push < push_items)
push_items = max_push;
+ tree_mod_log_eb_move(root->fs_info, dst, push_items, 0, dst_nritems);
memmove_extent_buffer(dst, btrfs_node_key_ptr_offset(push_items),
btrfs_node_key_ptr_offset(0),
(dst_nritems) *
sizeof(struct btrfs_key_ptr));
+ tree_mod_log_eb_copy(root->fs_info, dst, src, 0,
+ src_nritems - push_items, push_items);
copy_extent_buffer(dst, src,
btrfs_node_key_ptr_offset(0),
btrfs_node_key_ptr_offset(src_nritems - push_items),
@@ -2129,7 +2908,7 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
c = btrfs_alloc_free_block(trans, root, root->nodesize, 0,
root->root_key.objectid, &lower_key,
- level, root->node->start, 0, 0);
+ level, root->node->start, 0);
if (IS_ERR(c))
return PTR_ERR(c);
@@ -2161,6 +2940,7 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
btrfs_mark_buffer_dirty(c);
old = root->node;
+ tree_mod_log_set_root_pointer(root, c);
rcu_assign_pointer(root->node, c);
/* the super has an extra ref to root->node */
@@ -2184,10 +2964,11 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans,
static void insert_ptr(struct btrfs_trans_handle *trans,
struct btrfs_root *root, struct btrfs_path *path,
struct btrfs_disk_key *key, u64 bytenr,
- int slot, int level)
+ int slot, int level, int tree_mod_log)
{
struct extent_buffer *lower;
int nritems;
+ int ret;
BUG_ON(!path->nodes[level]);
btrfs_assert_tree_locked(path->nodes[level]);
@@ -2196,11 +2977,19 @@ static void insert_ptr(struct btrfs_trans_handle *trans,
BUG_ON(slot > nritems);
BUG_ON(nritems == BTRFS_NODEPTRS_PER_BLOCK(root));
if (slot != nritems) {
+ if (tree_mod_log && level)
+ tree_mod_log_eb_move(root->fs_info, lower, slot + 1,
+ slot, nritems - slot);
memmove_extent_buffer(lower,
btrfs_node_key_ptr_offset(slot + 1),
btrfs_node_key_ptr_offset(slot),
(nritems - slot) * sizeof(struct btrfs_key_ptr));
}
+ if (tree_mod_log && level) {
+ ret = tree_mod_log_insert_key(root->fs_info, lower, slot,
+ MOD_LOG_KEY_ADD);
+ BUG_ON(ret < 0);
+ }
btrfs_set_node_key(lower, key, slot);
btrfs_set_node_blockptr(lower, slot, bytenr);
WARN_ON(trans->transid == 0);
@@ -2252,7 +3041,7 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
split = btrfs_alloc_free_block(trans, root, root->nodesize, 0,
root->root_key.objectid,
- &disk_key, level, c->start, 0, 0);
+ &disk_key, level, c->start, 0);
if (IS_ERR(split))
return PTR_ERR(split);
@@ -2271,7 +3060,7 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
(unsigned long)btrfs_header_chunk_tree_uuid(split),
BTRFS_UUID_SIZE);
-
+ tree_mod_log_eb_copy(root->fs_info, split, c, 0, mid, c_nritems - mid);
copy_extent_buffer(split, c,
btrfs_node_key_ptr_offset(0),
btrfs_node_key_ptr_offset(mid),
@@ -2284,7 +3073,7 @@ static noinline int split_node(struct btrfs_trans_handle *trans,
btrfs_mark_buffer_dirty(split);
insert_ptr(trans, root, path, &disk_key, split->start,
- path->slots[level + 1] + 1, level + 1);
+ path->slots[level + 1] + 1, level + 1, 1);
if (path->slots[level] >= mid) {
path->slots[level] -= mid;
@@ -2821,7 +3610,7 @@ static noinline void copy_for_split(struct btrfs_trans_handle *trans,
btrfs_set_header_nritems(l, mid);
btrfs_item_key(right, &disk_key, 0);
insert_ptr(trans, root, path, &disk_key, right->start,
- path->slots[1] + 1, 1);
+ path->slots[1] + 1, 1, 0);
btrfs_mark_buffer_dirty(right);
btrfs_mark_buffer_dirty(l);
@@ -3004,7 +3793,7 @@ again:
right = btrfs_alloc_free_block(trans, root, root->leafsize, 0,
root->root_key.objectid,
- &disk_key, 0, l->start, 0, 0);
+ &disk_key, 0, l->start, 0);
if (IS_ERR(right))
return PTR_ERR(right);
@@ -3028,7 +3817,7 @@ again:
if (mid <= slot) {
btrfs_set_header_nritems(right, 0);
insert_ptr(trans, root, path, &disk_key, right->start,
- path->slots[1] + 1, 1);
+ path->slots[1] + 1, 1, 0);
btrfs_tree_unlock(path->nodes[0]);
free_extent_buffer(path->nodes[0]);
path->nodes[0] = right;
@@ -3037,7 +3826,7 @@ again:
} else {
btrfs_set_header_nritems(right, 0);
insert_ptr(trans, root, path, &disk_key, right->start,
- path->slots[1], 1);
+ path->slots[1], 1, 0);
btrfs_tree_unlock(path->nodes[0]);
free_extent_buffer(path->nodes[0]);
path->nodes[0] = right;
@@ -3749,19 +4538,29 @@ int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root
* empty a node.
*/
static void del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root,
- struct btrfs_path *path, int level, int slot)
+ struct btrfs_path *path, int level, int slot,
+ int tree_mod_log)
{
struct extent_buffer *parent = path->nodes[level];
u32 nritems;
+ int ret;
nritems = btrfs_header_nritems(parent);
if (slot != nritems - 1) {
+ if (tree_mod_log && level)
+ tree_mod_log_eb_move(root->fs_info, parent, slot,
+ slot + 1, nritems - slot - 1);
memmove_extent_buffer(parent,
btrfs_node_key_ptr_offset(slot),
btrfs_node_key_ptr_offset(slot + 1),
sizeof(struct btrfs_key_ptr) *
(nritems - slot - 1));
+ } else if (tree_mod_log && level) {
+ ret = tree_mod_log_insert_key(root->fs_info, parent, slot,
+ MOD_LOG_KEY_REMOVE);
+ BUG_ON(ret < 0);
}
+
nritems--;
btrfs_set_header_nritems(parent, nritems);
if (nritems == 0 && parent == root->node) {
@@ -3793,7 +4592,7 @@ static noinline void btrfs_del_leaf(struct btrfs_trans_handle *trans,
struct extent_buffer *leaf)
{
WARN_ON(btrfs_header_generation(leaf) != trans->transid);
- del_ptr(trans, root, path, 1, path->slots[1]);
+ del_ptr(trans, root, path, 1, path->slots[1], 1);
/*
* btrfs_free_extent is expensive, we want to make sure we
@@ -3804,7 +4603,7 @@ static noinline void btrfs_del_leaf(struct btrfs_trans_handle *trans,
root_sub_used(root, leaf->len);
extent_buffer_get(leaf);
- btrfs_free_tree_block(trans, root, leaf, 0, 1, 0);
+ btrfs_free_tree_block(trans, root, leaf, 0, 1);
free_extent_buffer_stale(leaf);
}
/*
@@ -4271,7 +5070,7 @@ again:
next = c;
next_rw_lock = path->locks[level];
ret = read_block_for_search(NULL, root, path, &next, level,
- slot, &key);
+ slot, &key, 0);
if (ret == -EAGAIN)
goto again;
@@ -4308,7 +5107,7 @@ again:
break;
ret = read_block_for_search(NULL, root, path, &next, level,
- 0, &key);
+ 0, &key, 0);
if (ret == -EAGAIN)
goto again;
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 8fd72331d60..0236d03c673 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -173,6 +173,9 @@ static int btrfs_csum_sizes[] = { 4, 0 };
#define BTRFS_FT_XATTR 8
#define BTRFS_FT_MAX 9
+/* ioprio of readahead is set to idle */
+#define BTRFS_IOPRIO_READA (IOPRIO_PRIO_VALUE(IOPRIO_CLASS_IDLE, 0))
+
/*
* The key defines the order in the tree, and so it also defines (optimal)
* block layout.
@@ -823,6 +826,14 @@ struct btrfs_csum_item {
u8 csum;
} __attribute__ ((__packed__));
+struct btrfs_dev_stats_item {
+ /*
+ * grow this item struct at the end for future enhancements and keep
+ * the existing values unchanged
+ */
+ __le64 values[BTRFS_DEV_STAT_VALUES_MAX];
+} __attribute__ ((__packed__));
+
/* different types of block groups (and chunks) */
#define BTRFS_BLOCK_GROUP_DATA (1ULL << 0)
#define BTRFS_BLOCK_GROUP_SYSTEM (1ULL << 1)
@@ -1129,6 +1140,15 @@ struct btrfs_fs_info {
spinlock_t delayed_iput_lock;
struct list_head delayed_iputs;
+ /* this protects tree_mod_seq_list */
+ spinlock_t tree_mod_seq_lock;
+ atomic_t tree_mod_seq;
+ struct list_head tree_mod_seq_list;
+
+ /* this protects tree_mod_log */
+ rwlock_t tree_mod_log_lock;
+ struct rb_root tree_mod_log;
+
atomic_t nr_async_submits;
atomic_t async_submit_draining;
atomic_t nr_async_bios;
@@ -1375,7 +1395,7 @@ struct btrfs_root {
struct list_head root_list;
spinlock_t orphan_lock;
- struct list_head orphan_list;
+ atomic_t orphan_inodes;
struct btrfs_block_rsv *orphan_block_rsv;
int orphan_item_inserted;
int orphan_cleanup_state;
@@ -1508,6 +1528,12 @@ struct btrfs_ioctl_defrag_range_args {
#define BTRFS_BALANCE_ITEM_KEY 248
/*
+ * Persistantly stores the io stats in the device tree.
+ * One key for all stats, (0, BTRFS_DEV_STATS_KEY, devid).
+ */
+#define BTRFS_DEV_STATS_KEY 249
+
+/*
* string items are for debugging. They just store a short string of
* data in the FS
*/
@@ -2415,6 +2441,30 @@ static inline u32 btrfs_file_extent_inline_item_len(struct extent_buffer *eb,
return btrfs_item_size(eb, e) - offset;
}
+/* btrfs_dev_stats_item */
+static inline u64 btrfs_dev_stats_value(struct extent_buffer *eb,
+ struct btrfs_dev_stats_item *ptr,
+ int index)
+{
+ u64 val;
+
+ read_extent_buffer(eb, &val,
+ offsetof(struct btrfs_dev_stats_item, values) +
+ ((unsigned long)ptr) + (index * sizeof(u64)),
+ sizeof(val));
+ return val;
+}
+
+static inline void btrfs_set_dev_stats_value(struct extent_buffer *eb,
+ struct btrfs_dev_stats_item *ptr,
+ int index, u64 val)
+{
+ write_extent_buffer(eb, &val,
+ offsetof(struct btrfs_dev_stats_item, values) +
+ ((unsigned long)ptr) + (index * sizeof(u64)),
+ sizeof(val));
+}
+
static inline struct btrfs_fs_info *btrfs_sb(struct super_block *sb)
{
return sb->s_fs_info;
@@ -2496,11 +2546,11 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
struct btrfs_root *root, u32 blocksize,
u64 parent, u64 root_objectid,
struct btrfs_disk_key *key, int level,
- u64 hint, u64 empty_size, int for_cow);
+ u64 hint, u64 empty_size);
void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct extent_buffer *buf,
- u64 parent, int last_ref, int for_cow);
+ u64 parent, int last_ref);
struct extent_buffer *btrfs_init_new_buffer(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
u64 bytenr, u32 blocksize,
@@ -2659,6 +2709,8 @@ int btrfs_duplicate_item(struct btrfs_trans_handle *trans,
int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root
*root, struct btrfs_key *key, struct btrfs_path *p, int
ins_len, int cow);
+int btrfs_search_old_slot(struct btrfs_root *root, struct btrfs_key *key,
+ struct btrfs_path *p, u64 time_seq);
int btrfs_realloc_node(struct btrfs_trans_handle *trans,
struct btrfs_root *root, struct extent_buffer *parent,
int start_slot, int cache_only, u64 *last_ret,
@@ -2922,7 +2974,6 @@ int btrfs_readpage(struct file *file, struct page *page);
void btrfs_evict_inode(struct inode *inode);
int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc);
int btrfs_dirty_inode(struct inode *inode);
-int btrfs_update_time(struct file *file);
struct inode *btrfs_alloc_inode(struct super_block *sb);
void btrfs_destroy_inode(struct inode *inode);
int btrfs_drop_inode(struct inode *inode);
@@ -3098,4 +3149,23 @@ void btrfs_reada_detach(void *handle);
int btree_readahead_hook(struct btrfs_root *root, struct extent_buffer *eb,
u64 start, int err);
+/* delayed seq elem */
+struct seq_list {
+ struct list_head list;
+ u64 seq;
+ u32 flags;
+};
+
+void btrfs_get_tree_mod_seq(struct btrfs_fs_info *fs_info,
+ struct seq_list *elem);
+void btrfs_put_tree_mod_seq(struct btrfs_fs_info *fs_info,
+ struct seq_list *elem);
+
+static inline int is_fstree(u64 rootid)
+{
+ if (rootid == BTRFS_FS_TREE_OBJECTID ||
+ (s64)rootid >= (s64)BTRFS_FIRST_FREE_OBJECTID)
+ return 1;
+ return 0;
+}
#endif
diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c
index 03e3748d84d..c18d0442ae6 100644
--- a/fs/btrfs/delayed-inode.c
+++ b/fs/btrfs/delayed-inode.c
@@ -669,8 +669,8 @@ static int btrfs_delayed_inode_reserve_metadata(
return ret;
} else if (src_rsv == &root->fs_info->delalloc_block_rsv) {
spin_lock(&BTRFS_I(inode)->lock);
- if (BTRFS_I(inode)->delalloc_meta_reserved) {
- BTRFS_I(inode)->delalloc_meta_reserved = 0;
+ if (test_and_clear_bit(BTRFS_INODE_DELALLOC_META_RESERVED,
+ &BTRFS_I(inode)->runtime_flags)) {
spin_unlock(&BTRFS_I(inode)->lock);
release = true;
goto migrate;
@@ -1706,7 +1706,7 @@ static void fill_stack_inode_item(struct btrfs_trans_handle *trans,
btrfs_set_stack_inode_nbytes(inode_item, inode_get_bytes(inode));
btrfs_set_stack_inode_generation(inode_item,
BTRFS_I(inode)->generation);
- btrfs_set_stack_inode_sequence(inode_item, BTRFS_I(inode)->sequence);
+ btrfs_set_stack_inode_sequence(inode_item, inode->i_version);
btrfs_set_stack_inode_transid(inode_item, trans->transid);
btrfs_set_stack_inode_rdev(inode_item, inode->i_rdev);
btrfs_set_stack_inode_flags(inode_item, BTRFS_I(inode)->flags);
@@ -1754,7 +1754,7 @@ int btrfs_fill_inode(struct inode *inode, u32 *rdev)
set_nlink(inode, btrfs_stack_inode_nlink(inode_item));
inode_set_bytes(inode, btrfs_stack_inode_nbytes(inode_item));
BTRFS_I(inode)->generation = btrfs_stack_inode_generation(inode_item);
- BTRFS_I(inode)->sequence = btrfs_stack_inode_sequence(inode_item);
+ inode->i_version = btrfs_stack_inode_sequence(inode_item);
inode->i_rdev = 0;
*rdev = btrfs_stack_inode_rdev(inode_item);
BTRFS_I(inode)->flags = btrfs_stack_inode_flags(inode_item);
diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c
index 69f22e3ab3b..13ae7b04790 100644
--- a/fs/btrfs/delayed-ref.c
+++ b/fs/btrfs/delayed-ref.c
@@ -525,7 +525,7 @@ static noinline void add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
ref->is_head = 0;
ref->in_tree = 1;
- if (need_ref_seq(for_cow, ref_root))
+ if (is_fstree(ref_root))
seq = inc_delayed_seq(delayed_refs);
ref->seq = seq;
@@ -584,7 +584,7 @@ static noinline void add_delayed_data_ref(struct btrfs_fs_info *fs_info,
ref->is_head = 0;
ref->in_tree = 1;
- if (need_ref_seq(for_cow, ref_root))
+ if (is_fstree(ref_root))
seq = inc_delayed_seq(delayed_refs);
ref->seq = seq;
@@ -658,10 +658,11 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info,
add_delayed_tree_ref(fs_info, trans, &ref->node, bytenr,
num_bytes, parent, ref_root, level, action,
for_cow);
- if (!need_ref_seq(for_cow, ref_root) &&
+ if (!is_fstree(ref_root) &&
waitqueue_active(&delayed_refs->seq_wait))
wake_up(&delayed_refs->seq_wait);
spin_unlock(&delayed_refs->lock);
+
return 0;
}
@@ -706,10 +707,11 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info,
add_delayed_data_ref(fs_info, trans, &ref->node, bytenr,
num_bytes, parent, ref_root, owner, offset,
action, for_cow);
- if (!need_ref_seq(for_cow, ref_root) &&
+ if (!is_fstree(ref_root) &&
waitqueue_active(&delayed_refs->seq_wait))
wake_up(&delayed_refs->seq_wait);
spin_unlock(&delayed_refs->lock);
+
return 0;
}
diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h
index d8f244d9492..413927fb995 100644
--- a/fs/btrfs/delayed-ref.h
+++ b/fs/btrfs/delayed-ref.h
@@ -195,11 +195,6 @@ int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans,
int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans,
struct list_head *cluster, u64 search_start);
-struct seq_list {
- struct list_head list;
- u64 seq;
-};
-
static inline u64 inc_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs)
{
assert_spin_locked(&delayed_refs->lock);
@@ -230,25 +225,6 @@ int btrfs_check_delayed_seq(struct btrfs_delayed_ref_root *delayed_refs,
u64 seq);
/*
- * delayed refs with a ref_seq > 0 must be held back during backref walking.
- * this only applies to items in one of the fs-trees. for_cow items never need
- * to be held back, so they won't get a ref_seq number.
- */
-static inline int need_ref_seq(int for_cow, u64 rootid)
-{
- if (for_cow)
- return 0;
-
- if (rootid == BTRFS_FS_TREE_OBJECTID)
- return 1;
-
- if ((s64)rootid >= (s64)BTRFS_FIRST_FREE_OBJECTID)
- return 1;
-
- return 0;
-}
-
-/*
* a node might live in a head or a regular ref, this lets you
* test for the proper type to use.
*/
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index e1fe74a2ce1..7ae51decf6d 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1153,7 +1153,6 @@ static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
root->orphan_block_rsv = NULL;
INIT_LIST_HEAD(&root->dirty_list);
- INIT_LIST_HEAD(&root->orphan_list);
INIT_LIST_HEAD(&root->root_list);
spin_lock_init(&root->orphan_lock);
spin_lock_init(&root->inode_lock);
@@ -1166,6 +1165,7 @@ static void __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
atomic_set(&root->log_commit[0], 0);
atomic_set(&root->log_commit[1], 0);
atomic_set(&root->log_writers, 0);
+ atomic_set(&root->orphan_inodes, 0);
root->log_batch = 0;
root->log_transid = 0;
root->last_log_commit = 0;
@@ -1252,7 +1252,7 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
leaf = btrfs_alloc_free_block(trans, root, root->leafsize, 0,
BTRFS_TREE_LOG_OBJECTID, NULL,
- 0, 0, 0, 0);
+ 0, 0, 0);
if (IS_ERR(leaf)) {
kfree(root);
return ERR_CAST(leaf);
@@ -1914,11 +1914,14 @@ int open_ctree(struct super_block *sb,
spin_lock_init(&fs_info->delayed_iput_lock);
spin_lock_init(&fs_info->defrag_inodes_lock);
spin_lock_init(&fs_info->free_chunk_lock);
+ spin_lock_init(&fs_info->tree_mod_seq_lock);
+ rwlock_init(&fs_info->tree_mod_log_lock);
mutex_init(&fs_info->reloc_mutex);
init_completion(&fs_info->kobj_unregister);
INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots);
INIT_LIST_HEAD(&fs_info->space_info);
+ INIT_LIST_HEAD(&fs_info->tree_mod_seq_list);
btrfs_mapping_init(&fs_info->mapping_tree);
btrfs_init_block_rsv(&fs_info->global_block_rsv);
btrfs_init_block_rsv(&fs_info->delalloc_block_rsv);
@@ -1931,12 +1934,14 @@ int open_ctree(struct super_block *sb,
atomic_set(&fs_info->async_submit_draining, 0);
atomic_set(&fs_info->nr_async_bios, 0);
atomic_set(&fs_info->defrag_running, 0);
+ atomic_set(&fs_info->tree_mod_seq, 0);
fs_info->sb = sb;
fs_info->max_inline = 8192 * 1024;
fs_info->metadata_ratio = 0;
fs_info->defrag_inodes = RB_ROOT;
fs_info->trans_no_join = 0;
fs_info->free_chunk_space = 0;
+ fs_info->tree_mod_log = RB_ROOT;
/* readahead state */
INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_WAIT);
@@ -2001,7 +2006,8 @@ int open_ctree(struct super_block *sb,
BTRFS_I(fs_info->btree_inode)->root = tree_root;
memset(&BTRFS_I(fs_info->btree_inode)->location, 0,
sizeof(struct btrfs_key));
- BTRFS_I(fs_info->btree_inode)->dummy_inode = 1;
+ set_bit(BTRFS_INODE_DUMMY,
+ &BTRFS_I(fs_info->btree_inode)->runtime_flags);
insert_inode_hash(fs_info->btree_inode);
spin_lock_init(&fs_info->block_group_cache_lock);
@@ -2353,6 +2359,13 @@ retry_root_backup:
fs_info->generation = generation;
fs_info->last_trans_committed = generation;
+ ret = btrfs_init_dev_stats(fs_info);
+ if (ret) {
+ printk(KERN_ERR "btrfs: failed to init dev_stats: %d\n",
+ ret);
+ goto fail_block_groups;
+ }
+
ret = btrfs_init_space_info(fs_info);
if (ret) {
printk(KERN_ERR "Failed to initial space info: %d\n", ret);
@@ -2556,18 +2569,19 @@ recovery_tree_root:
static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate)
{
- char b[BDEVNAME_SIZE];
-
if (uptodate) {
set_buffer_uptodate(bh);
} else {
+ struct btrfs_device *device = (struct btrfs_device *)
+ bh->b_private;
+
printk_ratelimited(KERN_WARNING "lost page write due to "
- "I/O error on %s\n",
- bdevname(bh->b_bdev, b));
+ "I/O error on %s\n", device->name);
/* note, we dont' set_buffer_write_io_error because we have
* our own ways of dealing with the IO errors
*/
clear_buffer_uptodate(bh);
+ btrfs_dev_stat_inc_and_print(device, BTRFS_DEV_STAT_WRITE_ERRS);
}
unlock_buffer(bh);
put_bh(bh);
@@ -2682,6 +2696,7 @@ static int write_dev_supers(struct btrfs_device *device,
set_buffer_uptodate(bh);
lock_buffer(bh);
bh->b_end_io = btrfs_end_buffer_write_sync;
+ bh->b_private = device;
}
/*
@@ -2740,6 +2755,9 @@ static int write_dev_flush(struct btrfs_device *device, int wait)
}
if (!bio_flagged(bio, BIO_UPTODATE)) {
ret = -EIO;
+ if (!bio_flagged(bio, BIO_EOPNOTSUPP))
+ btrfs_dev_stat_inc_and_print(device,
+ BTRFS_DEV_STAT_FLUSH_ERRS);
}
/* drop the reference from the wait == 0 run */
@@ -2902,19 +2920,6 @@ int write_ctree_super(struct btrfs_trans_handle *trans,
return ret;
}
-/* Kill all outstanding I/O */
-void btrfs_abort_devices(struct btrfs_root *root)
-{
- struct list_head *head;
- struct btrfs_device *dev;
- mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
- head = &root->fs_info->fs_devices->devices;
- list_for_each_entry_rcu(dev, head, dev_list) {
- blk_abort_queue(dev->bdev->bd_disk->queue);
- }
- mutex_unlock(&root->fs_info->fs_devices->device_list_mutex);
-}
-
void btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
{
spin_lock(&fs_info->fs_roots_radix_lock);
@@ -3671,17 +3676,6 @@ int btrfs_cleanup_transaction(struct btrfs_root *root)
return 0;
}
-static int btree_writepage_io_failed_hook(struct bio *bio, struct page *page,
- u64 start, u64 end,
- struct extent_state *state)
-{
- struct super_block *sb = page->mapping->host->i_sb;
- struct btrfs_fs_info *fs_info = btrfs_sb(sb);
- btrfs_error(fs_info, -EIO,
- "Error occured while writing out btree at %llu", start);
- return -EIO;
-}
-
static struct extent_io_ops btree_extent_io_ops = {
.write_cache_pages_lock_hook = btree_lock_page_hook,
.readpage_end_io_hook = btree_readpage_end_io_hook,
@@ -3689,5 +3683,4 @@ static struct extent_io_ops btree_extent_io_ops = {
.submit_bio_hook = btree_submit_bio_hook,
/* note we're sharing with inode.c for the merge bio hook */
.merge_bio_hook = btrfs_merge_bio_hook,
- .writepage_io_failed_hook = btree_writepage_io_failed_hook,
};
diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h
index ab1830aaf0e..05b3fab39f7 100644
--- a/fs/btrfs/disk-io.h
+++ b/fs/btrfs/disk-io.h
@@ -89,7 +89,6 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
int btrfs_cleanup_transaction(struct btrfs_root *root);
void btrfs_cleanup_one_transaction(struct btrfs_transaction *trans,
struct btrfs_root *root);
-void btrfs_abort_devices(struct btrfs_root *root);
#ifdef CONFIG_DEBUG_LOCK_ALLOC
void btrfs_init_lockdep(void);
diff --git a/fs/btrfs/export.c b/fs/btrfs/export.c
index e887ee62b6d..614f34a899c 100644
--- a/fs/btrfs/export.c
+++ b/fs/btrfs/export.c
@@ -13,15 +13,14 @@
parent_root_objectid) / 4)
#define BTRFS_FID_SIZE_CONNECTABLE_ROOT (sizeof(struct btrfs_fid) / 4)
-static int btrfs_encode_fh(struct dentry *dentry, u32 *fh, int *max_len,
- int connectable)
+static int btrfs_encode_fh(struct inode *inode, u32 *fh, int *max_len,
+ struct inode *parent)
{
struct btrfs_fid *fid = (struct btrfs_fid *)fh;
- struct inode *inode = dentry->d_inode;
int len = *max_len;
int type;
- if (connectable && (len < BTRFS_FID_SIZE_CONNECTABLE)) {
+ if (parent && (len < BTRFS_FID_SIZE_CONNECTABLE)) {
*max_len = BTRFS_FID_SIZE_CONNECTABLE;
return 255;
} else if (len < BTRFS_FID_SIZE_NON_CONNECTABLE) {
@@ -36,19 +35,13 @@ static int btrfs_encode_fh(struct dentry *dentry, u32 *fh, int *max_len,
fid->root_objectid = BTRFS_I(inode)->root->objectid;
fid->gen = inode->i_generation;
- if (connectable && !S_ISDIR(inode->i_mode)) {
- struct inode *parent;
+ if (parent) {
u64 parent_root_id;
- spin_lock(&dentry->d_lock);
-
- parent = dentry->d_parent->d_inode;
fid->parent_objectid = BTRFS_I(parent)->location.objectid;
fid->parent_gen = parent->i_generation;
parent_root_id = BTRFS_I(parent)->root->objectid;
- spin_unlock(&dentry->d_lock);
-
if (parent_root_id != fid->root_objectid) {
fid->parent_root_objectid = parent_root_id;
len = BTRFS_FID_SIZE_CONNECTABLE_ROOT;
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 49fd7b66d57..4b5a1e1bdef 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3578,7 +3578,7 @@ again:
space_info->chunk_alloc = 0;
spin_unlock(&space_info->lock);
out:
- mutex_unlock(&extent_root->fs_info->chunk_mutex);
+ mutex_unlock(&fs_info->chunk_mutex);
return ret;
}
@@ -4355,10 +4355,9 @@ static unsigned drop_outstanding_extent(struct inode *inode)
BTRFS_I(inode)->outstanding_extents--;
if (BTRFS_I(inode)->outstanding_extents == 0 &&
- BTRFS_I(inode)->delalloc_meta_reserved) {
+ test_and_clear_bit(BTRFS_INODE_DELALLOC_META_RESERVED,
+ &BTRFS_I(inode)->runtime_flags))
drop_inode_space = 1;
- BTRFS_I(inode)->delalloc_meta_reserved = 0;
- }
/*
* If we have more or the same amount of outsanding extents than we have
@@ -4465,7 +4464,8 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
* Add an item to reserve for updating the inode when we complete the
* delalloc io.
*/
- if (!BTRFS_I(inode)->delalloc_meta_reserved) {
+ if (!test_bit(BTRFS_INODE_DELALLOC_META_RESERVED,
+ &BTRFS_I(inode)->runtime_flags)) {
nr_extents++;
extra_reserve = 1;
}
@@ -4511,7 +4511,8 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes)
spin_lock(&BTRFS_I(inode)->lock);
if (extra_reserve) {
- BTRFS_I(inode)->delalloc_meta_reserved = 1;
+ set_bit(BTRFS_INODE_DELALLOC_META_RESERVED,
+ &BTRFS_I(inode)->runtime_flags);
nr_extents--;
}
BTRFS_I(inode)->reserved_extents += nr_extents;
@@ -5217,7 +5218,7 @@ out:
void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct extent_buffer *buf,
- u64 parent, int last_ref, int for_cow)
+ u64 parent, int last_ref)
{
struct btrfs_block_group_cache *cache = NULL;
int ret;
@@ -5227,7 +5228,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
buf->start, buf->len,
parent, root->root_key.objectid,
btrfs_header_level(buf),
- BTRFS_DROP_DELAYED_REF, NULL, for_cow);
+ BTRFS_DROP_DELAYED_REF, NULL, 0);
BUG_ON(ret); /* -ENOMEM */
}
@@ -6249,7 +6250,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
struct btrfs_root *root, u32 blocksize,
u64 parent, u64 root_objectid,
struct btrfs_disk_key *key, int level,
- u64 hint, u64 empty_size, int for_cow)
+ u64 hint, u64 empty_size)
{
struct btrfs_key ins;
struct btrfs_block_rsv *block_rsv;
@@ -6297,7 +6298,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans,
ins.objectid,
ins.offset, parent, root_objectid,
level, BTRFS_ADD_DELAYED_EXTENT,
- extent_op, for_cow);
+ extent_op, 0);
BUG_ON(ret); /* -ENOMEM */
}
return buf;
@@ -6715,7 +6716,7 @@ static noinline int walk_up_proc(struct btrfs_trans_handle *trans,
btrfs_header_owner(path->nodes[level + 1]));
}
- btrfs_free_tree_block(trans, root, eb, parent, wc->refs[level] == 1, 0);
+ btrfs_free_tree_block(trans, root, eb, parent, wc->refs[level] == 1);
out:
wc->refs[level] = 0;
wc->flags[level] = 0;
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index c9018a05036..2c8f7b20461 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -186,7 +186,6 @@ static struct rb_node *tree_insert(struct rb_root *root, u64 offset,
return parent;
}
- entry = rb_entry(node, struct tree_entry, rb_node);
rb_link_node(node, parent, p);
rb_insert_color(node, root);
return NULL;
@@ -413,7 +412,7 @@ static struct extent_state *next_state(struct extent_state *state)
/*
* utility function to clear some bits in an extent state struct.
- * it will optionally wake up any one waiting on this state (wake == 1)
+ * it will optionally wake up any one waiting on this state (wake == 1).
*
* If no bits are set on the state struct after clearing things, the
* struct is freed and removed from the tree
@@ -570,10 +569,8 @@ hit_next:
if (err)
goto out;
if (state->end <= end) {
- clear_state_bit(tree, state, &bits, wake);
- if (last_end == (u64)-1)
- goto out;
- start = last_end + 1;
+ state = clear_state_bit(tree, state, &bits, wake);
+ goto next;
}
goto search_again;
}
@@ -781,7 +778,6 @@ hit_next:
* Just lock what we found and keep going
*/
if (state->start == start && state->end <= end) {
- struct rb_node *next_node;
if (state->state & exclusive_bits) {
*failed_start = state->start;
err = -EEXIST;
@@ -789,20 +785,15 @@ hit_next:
}
set_state_bits(tree, state, &bits);
-
cache_state(state, cached_state);
merge_state(tree, state);
if (last_end == (u64)-1)
goto out;
-
start = last_end + 1;
- next_node = rb_next(&state->rb_node);
- if (next_node && start < end && prealloc && !need_resched()) {
- state = rb_entry(next_node, struct extent_state,
- rb_node);
- if (state->start == start)
- goto hit_next;
- }
+ state = next_state(state);
+ if (start < end && state && state->start == start &&
+ !need_resched())
+ goto hit_next;
goto search_again;
}
@@ -845,6 +836,10 @@ hit_next:
if (last_end == (u64)-1)
goto out;
start = last_end + 1;
+ state = next_state(state);
+ if (start < end && state && state->start == start &&
+ !need_resched())
+ goto hit_next;
}
goto search_again;
}
@@ -994,21 +989,14 @@ hit_next:
* Just lock what we found and keep going
*/
if (state->start == start && state->end <= end) {
- struct rb_node *next_node;
-
set_state_bits(tree, state, &bits);
- clear_state_bit(tree, state, &clear_bits, 0);
+ state = clear_state_bit(tree, state, &clear_bits, 0);
if (last_end == (u64)-1)
goto out;
-
start = last_end + 1;
- next_node = rb_next(&state->rb_node);
- if (next_node && start < end && prealloc && !need_resched()) {
- state = rb_entry(next_node, struct extent_state,
- rb_node);
- if (state->start == start)
- goto hit_next;
- }
+ if (start < end && state && state->start == start &&
+ !need_resched())
+ goto hit_next;
goto search_again;
}
@@ -1042,10 +1030,13 @@ hit_next:
goto out;
if (state->end <= end) {
set_state_bits(tree, state, &bits);
- clear_state_bit(tree, state, &clear_bits, 0);
+ state = clear_state_bit(tree, state, &clear_bits, 0);
if (last_end == (u64)-1)
goto out;
start = last_end + 1;
+ if (start < end && state && state->start == start &&
+ !need_resched())
+ goto hit_next;
}
goto search_again;
}
@@ -1173,9 +1164,8 @@ int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
cached_state, mask);
}
-static int clear_extent_uptodate(struct extent_io_tree *tree, u64 start,
- u64 end, struct extent_state **cached_state,
- gfp_t mask)
+int clear_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
+ struct extent_state **cached_state, gfp_t mask)
{
return clear_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, 0,
cached_state, mask);
@@ -1293,7 +1283,7 @@ out:
* returned if we find something, and *start_ret and *end_ret are
* set to reflect the state struct that was found.
*
- * If nothing was found, 1 is returned, < 0 on error
+ * If nothing was found, 1 is returned. If found something, return 0.
*/
int find_first_extent_bit(struct extent_io_tree *tree, u64 start,
u64 *start_ret, u64 *end_ret, int bits)
@@ -1923,6 +1913,7 @@ int repair_io_failure(struct btrfs_mapping_tree *map_tree, u64 start,
if (!test_bit(BIO_UPTODATE, &bio->bi_flags)) {
/* try to remap that extent elsewhere? */
bio_put(bio);
+ btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
return -EIO;
}
@@ -2222,17 +2213,7 @@ int end_extent_writepage(struct page *page, int err, u64 start, u64 end)
uptodate = 0;
}
- if (!uptodate && tree->ops &&
- tree->ops->writepage_io_failed_hook) {
- ret = tree->ops->writepage_io_failed_hook(NULL, page,
- start, end, NULL);
- /* Writeback already completed */
- if (ret == 0)
- return 1;
- }
-
if (!uptodate) {
- clear_extent_uptodate(tree, start, end, NULL, GFP_NOFS);
ClearPageUptodate(page);
SetPageError(page);
}
@@ -2347,10 +2328,23 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
if (uptodate && tree->ops && tree->ops->readpage_end_io_hook) {
ret = tree->ops->readpage_end_io_hook(page, start, end,
state, mirror);
- if (ret)
+ if (ret) {
+ /* no IO indicated but software detected errors
+ * in the block, either checksum errors or
+ * issues with the contents */
+ struct btrfs_root *root =
+ BTRFS_I(page->mapping->host)->root;
+ struct btrfs_device *device;
+
uptodate = 0;
- else
+ device = btrfs_find_device_for_logical(
+ root, start, mirror);
+ if (device)
+ btrfs_dev_stat_inc_and_print(device,
+ BTRFS_DEV_STAT_CORRUPTION_ERRS);
+ } else {
clean_io_failure(start, page);
+ }
}
if (!uptodate && tree->ops && tree->ops->readpage_io_failed_hook) {
@@ -3164,7 +3158,7 @@ static int write_one_eb(struct extent_buffer *eb,
u64 offset = eb->start;
unsigned long i, num_pages;
int rw = (epd->sync_io ? WRITE_SYNC : WRITE);
- int ret;
+ int ret = 0;
clear_bit(EXTENT_BUFFER_IOERR, &eb->bflags);
num_pages = num_extent_pages(eb->start, eb->len);
@@ -3930,6 +3924,7 @@ static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree,
eb->start = start;
eb->len = len;
eb->tree = tree;
+ eb->bflags = 0;
rwlock_init(&eb->lock);
atomic_set(&eb->write_locks, 0);
atomic_set(&eb->read_locks, 0);
@@ -3967,6 +3962,60 @@ static struct extent_buffer *__alloc_extent_buffer(struct extent_io_tree *tree,
return eb;
}
+struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src)
+{
+ unsigned long i;
+ struct page *p;
+ struct extent_buffer *new;
+ unsigned long num_pages = num_extent_pages(src->start, src->len);
+
+ new = __alloc_extent_buffer(NULL, src->start, src->len, GFP_ATOMIC);
+ if (new == NULL)
+ return NULL;
+
+ for (i = 0; i < num_pages; i++) {
+ p = alloc_page(GFP_ATOMIC);
+ BUG_ON(!p);
+ attach_extent_buffer_page(new, p);
+ WARN_ON(PageDirty(p));
+ SetPageUptodate(p);
+ new->pages[i] = p;
+ }
+
+ copy_extent_buffer(new, src, 0, 0, src->len);
+ set_bit(EXTENT_BUFFER_UPTODATE, &new->bflags);
+ set_bit(EXTENT_BUFFER_DUMMY, &new->bflags);
+
+ return new;
+}
+
+struct extent_buffer *alloc_dummy_extent_buffer(u64 start, unsigned long len)
+{
+ struct extent_buffer *eb;
+ unsigned long num_pages = num_extent_pages(0, len);
+ unsigned long i;
+
+ eb = __alloc_extent_buffer(NULL, start, len, GFP_ATOMIC);
+ if (!eb)
+ return NULL;
+
+ for (i = 0; i < num_pages; i++) {
+ eb->pages[i] = alloc_page(GFP_ATOMIC);
+ if (!eb->pages[i])
+ goto err;
+ }
+ set_extent_buffer_uptodate(eb);
+ btrfs_set_header_nritems(eb, 0);
+ set_bit(EXTENT_BUFFER_DUMMY, &eb->bflags);
+
+ return eb;
+err:
+ for (i--; i > 0; i--)
+ __free_page(eb->pages[i]);
+ __free_extent_buffer(eb);
+ return NULL;
+}
+
static int extent_buffer_under_io(struct extent_buffer *eb)
{
return (atomic_read(&eb->io_pages) ||
@@ -3981,18 +4030,21 @@ static void btrfs_release_extent_buffer_page(struct extent_buffer *eb,
unsigned long start_idx)
{
unsigned long index;
+ unsigned long num_pages;
struct page *page;
+ int mapped = !test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags);
BUG_ON(extent_buffer_under_io(eb));
- index = num_extent_pages(eb->start, eb->len);
+ num_pages = num_extent_pages(eb->start, eb->len);
+ index = start_idx + num_pages;
if (start_idx >= index)
return;
do {
index--;
page = extent_buffer_page(eb, index);
- if (page) {
+ if (page && mapped) {
spin_lock(&page->mapping->private_lock);
/*
* We do this since we'll remove the pages after we've
@@ -4017,6 +4069,8 @@ static void btrfs_release_extent_buffer_page(struct extent_buffer *eb,
}
spin_unlock(&page->mapping->private_lock);
+ }
+ if (page) {
/* One for when we alloced the page */
page_cache_release(page);
}
@@ -4235,14 +4289,18 @@ static void release_extent_buffer(struct extent_buffer *eb, gfp_t mask)
{
WARN_ON(atomic_read(&eb->refs) == 0);
if (atomic_dec_and_test(&eb->refs)) {
- struct extent_io_tree *tree = eb->tree;
+ if (test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags)) {
+ spin_unlock(&eb->refs_lock);
+ } else {
+ struct extent_io_tree *tree = eb->tree;
- spin_unlock(&eb->refs_lock);
+ spin_unlock(&eb->refs_lock);
- spin_lock(&tree->buffer_lock);
- radix_tree_delete(&tree->buffer,
- eb->start >> PAGE_CACHE_SHIFT);
- spin_unlock(&tree->buffer_lock);
+ spin_lock(&tree->buffer_lock);
+ radix_tree_delete(&tree->buffer,
+ eb->start >> PAGE_CACHE_SHIFT);
+ spin_unlock(&tree->buffer_lock);
+ }
/* Should be safe to release our pages at this point */
btrfs_release_extent_buffer_page(eb, 0);
@@ -4260,6 +4318,10 @@ void free_extent_buffer(struct extent_buffer *eb)
spin_lock(&eb->refs_lock);
if (atomic_read(&eb->refs) == 2 &&
+ test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags))
+ atomic_dec(&eb->refs);
+
+ if (atomic_read(&eb->refs) == 2 &&
test_bit(EXTENT_BUFFER_STALE, &eb->bflags) &&
!extent_buffer_under_io(eb) &&
test_and_clear_bit(EXTENT_BUFFER_TREE_REF, &eb->bflags))
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index b516c3b8dec..25900af5b15 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -39,6 +39,7 @@
#define EXTENT_BUFFER_STALE 6
#define EXTENT_BUFFER_WRITEBACK 7
#define EXTENT_BUFFER_IOERR 8
+#define EXTENT_BUFFER_DUMMY 9
/* these are flags for extent_clear_unlock_delalloc */
#define EXTENT_CLEAR_UNLOCK_PAGE 0x1
@@ -75,9 +76,6 @@ struct extent_io_ops {
unsigned long bio_flags);
int (*readpage_io_hook)(struct page *page, u64 start, u64 end);
int (*readpage_io_failed_hook)(struct page *page, int failed_mirror);
- int (*writepage_io_failed_hook)(struct bio *bio, struct page *page,
- u64 start, u64 end,
- struct extent_state *state);
int (*readpage_end_io_hook)(struct page *page, u64 start, u64 end,
struct extent_state *state, int mirror);
int (*writepage_end_io_hook)(struct page *page, u64 start, u64 end,
@@ -225,6 +223,8 @@ int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
struct extent_state **cached_state, gfp_t mask);
int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
struct extent_state **cached_state, gfp_t mask);
+int clear_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end,
+ struct extent_state **cached_state, gfp_t mask);
int set_extent_new(struct extent_io_tree *tree, u64 start, u64 end,
gfp_t mask);
int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end,
@@ -265,6 +265,8 @@ void set_page_extent_mapped(struct page *page);
struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree,
u64 start, unsigned long len);
+struct extent_buffer *alloc_dummy_extent_buffer(u64 start, unsigned long len);
+struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src);
struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree,
u64 start, unsigned long len);
void free_extent_buffer(struct extent_buffer *eb);
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 53bf2d764bb..70dc8ca73e2 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -65,6 +65,21 @@ struct inode_defrag {
int cycled;
};
+static int __compare_inode_defrag(struct inode_defrag *defrag1,
+ struct inode_defrag *defrag2)
+{
+ if (defrag1->root > defrag2->root)
+ return 1;
+ else if (defrag1->root < defrag2->root)
+ return -1;
+ else if (defrag1->ino > defrag2->ino)
+ return 1;
+ else if (defrag1->ino < defrag2->ino)
+ return -1;
+ else
+ return 0;
+}
+
/* pop a record for an inode into the defrag tree. The lock
* must be held already
*
@@ -81,15 +96,17 @@ static void __btrfs_add_inode_defrag(struct inode *inode,
struct inode_defrag *entry;
struct rb_node **p;
struct rb_node *parent = NULL;
+ int ret;
p = &root->fs_info->defrag_inodes.rb_node;
while (*p) {
parent = *p;
entry = rb_entry(parent, struct inode_defrag, rb_node);
- if (defrag->ino < entry->ino)
+ ret = __compare_inode_defrag(defrag, entry);
+ if (ret < 0)
p = &parent->rb_left;
- else if (defrag->ino > entry->ino)
+ else if (ret > 0)
p = &parent->rb_right;
else {
/* if we're reinserting an entry for
@@ -103,7 +120,7 @@ static void __btrfs_add_inode_defrag(struct inode *inode,
goto exists;
}
}
- BTRFS_I(inode)->in_defrag = 1;
+ set_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags);
rb_link_node(&defrag->rb_node, parent, p);
rb_insert_color(&defrag->rb_node, &root->fs_info->defrag_inodes);
return;
@@ -131,7 +148,7 @@ int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
if (btrfs_fs_closing(root->fs_info))
return 0;
- if (BTRFS_I(inode)->in_defrag)
+ if (test_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags))
return 0;
if (trans)
@@ -148,7 +165,7 @@ int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
defrag->root = root->root_key.objectid;
spin_lock(&root->fs_info->defrag_inodes_lock);
- if (!BTRFS_I(inode)->in_defrag)
+ if (!test_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags))
__btrfs_add_inode_defrag(inode, defrag);
else
kfree(defrag);
@@ -159,28 +176,35 @@ int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans,
/*
* must be called with the defrag_inodes lock held
*/
-struct inode_defrag *btrfs_find_defrag_inode(struct btrfs_fs_info *info, u64 ino,
+struct inode_defrag *btrfs_find_defrag_inode(struct btrfs_fs_info *info,
+ u64 root, u64 ino,
struct rb_node **next)
{
struct inode_defrag *entry = NULL;
+ struct inode_defrag tmp;
struct rb_node *p;
struct rb_node *parent = NULL;
+ int ret;
+
+ tmp.ino = ino;
+ tmp.root = root;
p = info->defrag_inodes.rb_node;
while (p) {
parent = p;
entry = rb_entry(parent, struct inode_defrag, rb_node);
- if (ino < entry->ino)
+ ret = __compare_inode_defrag(&tmp, entry);
+ if (ret < 0)
p = parent->rb_left;
- else if (ino > entry->ino)
+ else if (ret > 0)
p = parent->rb_right;
else
return entry;
}
if (next) {
- while (parent && ino > entry->ino) {
+ while (parent && __compare_inode_defrag(&tmp, entry) > 0) {
parent = rb_next(parent);
entry = rb_entry(parent, struct inode_defrag, rb_node);
}
@@ -202,6 +226,7 @@ int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info)
struct btrfs_key key;
struct btrfs_ioctl_defrag_range_args range;
u64 first_ino = 0;
+ u64 root_objectid = 0;
int num_defrag;
int defrag_batch = 1024;
@@ -214,11 +239,14 @@ int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info)
n = NULL;
/* find an inode to defrag */
- defrag = btrfs_find_defrag_inode(fs_info, first_ino, &n);
+ defrag = btrfs_find_defrag_inode(fs_info, root_objectid,
+ first_ino, &n);
if (!defrag) {
- if (n)
- defrag = rb_entry(n, struct inode_defrag, rb_node);
- else if (first_ino) {
+ if (n) {
+ defrag = rb_entry(n, struct inode_defrag,
+ rb_node);
+ } else if (root_objectid || first_ino) {
+ root_objectid = 0;
first_ino = 0;
continue;
} else {
@@ -228,6 +256,7 @@ int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info)
/* remove it from the rbtree */
first_ino = defrag->ino + 1;
+ root_objectid = defrag->root;
rb_erase(&defrag->rb_node, &fs_info->defrag_inodes);
if (btrfs_fs_closing(fs_info))
@@ -252,7 +281,7 @@ int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info)
goto next;
/* do a chunk of defrag */
- BTRFS_I(inode)->in_defrag = 0;
+ clear_bit(BTRFS_INODE_IN_DEFRAG, &BTRFS_I(inode)->runtime_flags);
range.start = defrag->last_offset;
num_defrag = btrfs_defrag_file(inode, NULL, &range, defrag->transid,
defrag_batch);
@@ -1404,12 +1433,11 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb,
goto out;
}
- err = btrfs_update_time(file);
+ err = file_update_time(file);
if (err) {
mutex_unlock(&inode->i_mutex);
goto out;
}
- BTRFS_I(inode)->sequence++;
start_pos = round_down(pos, root->sectorsize);
if (start_pos > i_size_read(inode)) {
@@ -1466,8 +1494,8 @@ int btrfs_release_file(struct inode *inode, struct file *filp)
* flush down new bytes that may have been written if the
* application were using truncate to replace a file in place.
*/
- if (BTRFS_I(inode)->ordered_data_close) {
- BTRFS_I(inode)->ordered_data_close = 0;
+ if (test_and_clear_bit(BTRFS_INODE_ORDERED_DATA_CLOSE,
+ &BTRFS_I(inode)->runtime_flags)) {
btrfs_add_ordered_operation(NULL, BTRFS_I(inode)->root, inode);
if (inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT)
filemap_flush(inode->i_mapping);
@@ -1498,14 +1526,15 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
trace_btrfs_sync_file(file, datasync);
- ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
- if (ret)
- return ret;
mutex_lock(&inode->i_mutex);
- /* we wait first, since the writeback may change the inode */
+ /*
+ * we wait first, since the writeback may change the inode, also wait
+ * ordered range does a filemape_write_and_wait_range which is why we
+ * don't do it above like other file systems.
+ */
root->log_batch++;
- btrfs_wait_ordered_range(inode, 0, (u64)-1);
+ btrfs_wait_ordered_range(inode, start, end);
root->log_batch++;
/*
@@ -1523,7 +1552,8 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
* syncing
*/
smp_mb();
- if (BTRFS_I(inode)->last_trans <=
+ if (btrfs_inode_in_log(inode, root->fs_info->generation) ||
+ BTRFS_I(inode)->last_trans <=
root->fs_info->last_trans_committed) {
BTRFS_I(inode)->last_trans = 0;
mutex_unlock(&inode->i_mutex);
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 202008ec367..81296c57405 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -33,6 +33,8 @@
static int link_free_space(struct btrfs_free_space_ctl *ctl,
struct btrfs_free_space *info);
+static void unlink_free_space(struct btrfs_free_space_ctl *ctl,
+ struct btrfs_free_space *info);
static struct inode *__lookup_free_space_inode(struct btrfs_root *root,
struct btrfs_path *path,
@@ -75,7 +77,8 @@ static struct inode *__lookup_free_space_inode(struct btrfs_root *root,
return ERR_PTR(-ENOENT);
}
- inode->i_mapping->flags &= ~__GFP_FS;
+ mapping_set_gfp_mask(inode->i_mapping,
+ mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS);
return inode;
}
@@ -365,7 +368,7 @@ static int io_ctl_prepare_pages(struct io_ctl *io_ctl, struct inode *inode,
static void io_ctl_set_generation(struct io_ctl *io_ctl, u64 generation)
{
- u64 *val;
+ __le64 *val;
io_ctl_map_page(io_ctl, 1);
@@ -388,7 +391,7 @@ static void io_ctl_set_generation(struct io_ctl *io_ctl, u64 generation)
static int io_ctl_check_generation(struct io_ctl *io_ctl, u64 generation)
{
- u64 *gen;
+ __le64 *gen;
/*
* Skip the crc area. If we don't check crcs then we just have a 64bit
@@ -584,6 +587,44 @@ static int io_ctl_read_bitmap(struct io_ctl *io_ctl,
return 0;
}
+/*
+ * Since we attach pinned extents after the fact we can have contiguous sections
+ * of free space that are split up in entries. This poses a problem with the
+ * tree logging stuff since it could have allocated across what appears to be 2
+ * entries since we would have merged the entries when adding the pinned extents
+ * back to the free space cache. So run through the space cache that we just
+ * loaded and merge contiguous entries. This will make the log replay stuff not
+ * blow up and it will make for nicer allocator behavior.
+ */
+static void merge_space_tree(struct btrfs_free_space_ctl *ctl)
+{
+ struct btrfs_free_space *e, *prev = NULL;
+ struct rb_node *n;
+
+again:
+ spin_lock(&ctl->tree_lock);
+ for (n = rb_first(&ctl->free_space_offset); n; n = rb_next(n)) {
+ e = rb_entry(n, struct btrfs_free_space, offset_index);
+ if (!prev)
+ goto next;
+ if (e->bitmap || prev->bitmap)
+ goto next;
+ if (prev->offset + prev->bytes == e->offset) {
+ unlink_free_space(ctl, prev);
+ unlink_free_space(ctl, e);
+ prev->bytes += e->bytes;
+ kmem_cache_free(btrfs_free_space_cachep, e);
+ link_free_space(ctl, prev);
+ prev = NULL;
+ spin_unlock(&ctl->tree_lock);
+ goto again;
+ }
+next:
+ prev = e;
+ }
+ spin_unlock(&ctl->tree_lock);
+}
+
int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
struct btrfs_free_space_ctl *ctl,
struct btrfs_path *path, u64 offset)
@@ -726,6 +767,7 @@ int __load_free_space_cache(struct btrfs_root *root, struct inode *inode,
}
io_ctl_drop_pages(&io_ctl);
+ merge_space_tree(ctl);
ret = 1;
out:
io_ctl_free(&io_ctl);
@@ -972,9 +1014,7 @@ int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode,
goto out;
- ret = filemap_write_and_wait(inode->i_mapping);
- if (ret)
- goto out;
+ btrfs_wait_ordered_range(inode, 0, (u64)-1);
key.objectid = BTRFS_FREE_SPACE_OBJECTID;
key.offset = offset;
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index ceb7b9c9edc..f6ab6f5e635 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -89,7 +89,7 @@ static unsigned char btrfs_type_by_mode[S_IFMT >> S_SHIFT] = {
static int btrfs_setsize(struct inode *inode, loff_t newsize);
static int btrfs_truncate(struct inode *inode);
-static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end);
+static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent);
static noinline int cow_file_range(struct inode *inode,
struct page *locked_page,
u64 start, u64 end, int *page_started,
@@ -257,10 +257,13 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
ret = insert_inline_extent(trans, root, inode, start,
inline_len, compressed_size,
compress_type, compressed_pages);
- if (ret) {
+ if (ret && ret != -ENOSPC) {
btrfs_abort_transaction(trans, root, ret);
return ret;
+ } else if (ret == -ENOSPC) {
+ return 1;
}
+
btrfs_delalloc_release_metadata(inode, end + 1 - start);
btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0);
return 0;
@@ -1572,11 +1575,11 @@ static int btrfs_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
if (btrfs_is_free_space_inode(root, inode))
metadata = 2;
- ret = btrfs_bio_wq_end_io(root->fs_info, bio, metadata);
- if (ret)
- return ret;
-
if (!(rw & REQ_WRITE)) {
+ ret = btrfs_bio_wq_end_io(root->fs_info, bio, metadata);
+ if (ret)
+ return ret;
+
if (bio_flags & EXTENT_BIO_COMPRESSED) {
return btrfs_submit_compressed_read(inode, bio,
mirror_num, bio_flags);
@@ -1815,25 +1818,24 @@ out:
* an ordered extent if the range of bytes in the file it covers are
* fully written.
*/
-static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
+static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
{
+ struct inode *inode = ordered_extent->inode;
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_trans_handle *trans = NULL;
- struct btrfs_ordered_extent *ordered_extent = NULL;
struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
struct extent_state *cached_state = NULL;
int compress_type = 0;
int ret;
bool nolock;
- ret = btrfs_dec_test_ordered_pending(inode, &ordered_extent, start,
- end - start + 1);
- if (!ret)
- return 0;
- BUG_ON(!ordered_extent); /* Logic error */
-
nolock = btrfs_is_free_space_inode(root, inode);
+ if (test_bit(BTRFS_ORDERED_IOERR, &ordered_extent->flags)) {
+ ret = -EIO;
+ goto out;
+ }
+
if (test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags)) {
BUG_ON(!list_empty(&ordered_extent->list)); /* Logic error */
ret = btrfs_ordered_update_i_size(inode, 0, ordered_extent);
@@ -1889,12 +1891,10 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
ordered_extent->file_offset,
ordered_extent->len);
}
- unlock_extent_cached(io_tree, ordered_extent->file_offset,
- ordered_extent->file_offset +
- ordered_extent->len - 1, &cached_state, GFP_NOFS);
+
if (ret < 0) {
btrfs_abort_transaction(trans, root, ret);
- goto out;
+ goto out_unlock;
}
add_pending_csums(trans, inode, ordered_extent->file_offset,
@@ -1905,10 +1905,14 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
ret = btrfs_update_inode_fallback(trans, root, inode);
if (ret) { /* -ENOMEM or corruption */
btrfs_abort_transaction(trans, root, ret);
- goto out;
+ goto out_unlock;
}
}
ret = 0;
+out_unlock:
+ unlock_extent_cached(io_tree, ordered_extent->file_offset,
+ ordered_extent->file_offset +
+ ordered_extent->len - 1, &cached_state, GFP_NOFS);
out:
if (root != root->fs_info->tree_root)
btrfs_delalloc_release_metadata(inode, ordered_extent->len);
@@ -1919,26 +1923,57 @@ out:
btrfs_end_transaction(trans, root);
}
+ if (ret)
+ clear_extent_uptodate(io_tree, ordered_extent->file_offset,
+ ordered_extent->file_offset +
+ ordered_extent->len - 1, NULL, GFP_NOFS);
+
+ /*
+ * This needs to be dont to make sure anybody waiting knows we are done
+ * upating everything for this ordered extent.
+ */
+ btrfs_remove_ordered_extent(inode, ordered_extent);
+
/* once for us */
btrfs_put_ordered_extent(ordered_extent);
/* once for the tree */
btrfs_put_ordered_extent(ordered_extent);
- return 0;
-out_unlock:
- unlock_extent_cached(io_tree, ordered_extent->file_offset,
- ordered_extent->file_offset +
- ordered_extent->len - 1, &cached_state, GFP_NOFS);
- goto out;
+ return ret;
+}
+
+static void finish_ordered_fn(struct btrfs_work *work)
+{
+ struct btrfs_ordered_extent *ordered_extent;
+ ordered_extent = container_of(work, struct btrfs_ordered_extent, work);
+ btrfs_finish_ordered_io(ordered_extent);
}
static int btrfs_writepage_end_io_hook(struct page *page, u64 start, u64 end,
struct extent_state *state, int uptodate)
{
+ struct inode *inode = page->mapping->host;
+ struct btrfs_root *root = BTRFS_I(inode)->root;
+ struct btrfs_ordered_extent *ordered_extent = NULL;
+ struct btrfs_workers *workers;
+
trace_btrfs_writepage_end_io_hook(page, start, end, uptodate);
ClearPagePrivate2(page);
- return btrfs_finish_ordered_io(page->mapping->host, start, end);
+ if (!btrfs_dec_test_ordered_pending(inode, &ordered_extent, start,
+ end - start + 1, uptodate))
+ return 0;
+
+ ordered_extent->work.func = finish_ordered_fn;
+ ordered_extent->work.flags = 0;
+
+ if (btrfs_is_free_space_inode(root, inode))
+ workers = &root->fs_info->endio_freespace_worker;
+ else
+ workers = &root->fs_info->endio_write_workers;
+ btrfs_queue_worker(workers, &ordered_extent->work);
+
+ return 0;
}
/*
@@ -2072,12 +2107,12 @@ void btrfs_orphan_commit_root(struct btrfs_trans_handle *trans,
struct btrfs_block_rsv *block_rsv;
int ret;
- if (!list_empty(&root->orphan_list) ||
+ if (atomic_read(&root->orphan_inodes) ||
root->orphan_cleanup_state != ORPHAN_CLEANUP_DONE)
return;
spin_lock(&root->orphan_lock);
- if (!list_empty(&root->orphan_list)) {
+ if (atomic_read(&root->orphan_inodes)) {
spin_unlock(&root->orphan_lock);
return;
}
@@ -2134,8 +2169,8 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
block_rsv = NULL;
}
- if (list_empty(&BTRFS_I(inode)->i_orphan)) {
- list_add(&BTRFS_I(inode)->i_orphan, &root->orphan_list);
+ if (!test_and_set_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
+ &BTRFS_I(inode)->runtime_flags)) {
#if 0
/*
* For proper ENOSPC handling, we should do orphan
@@ -2148,12 +2183,12 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
insert = 1;
#endif
insert = 1;
+ atomic_dec(&root->orphan_inodes);
}
- if (!BTRFS_I(inode)->orphan_meta_reserved) {
- BTRFS_I(inode)->orphan_meta_reserved = 1;
+ if (!test_and_set_bit(BTRFS_INODE_ORPHAN_META_RESERVED,
+ &BTRFS_I(inode)->runtime_flags))
reserve = 1;
- }
spin_unlock(&root->orphan_lock);
/* grab metadata reservation from transaction handle */
@@ -2166,6 +2201,8 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
if (insert >= 1) {
ret = btrfs_insert_orphan_item(trans, root, btrfs_ino(inode));
if (ret && ret != -EEXIST) {
+ clear_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
+ &BTRFS_I(inode)->runtime_flags);
btrfs_abort_transaction(trans, root, ret);
return ret;
}
@@ -2196,15 +2233,13 @@ int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode)
int ret = 0;
spin_lock(&root->orphan_lock);
- if (!list_empty(&BTRFS_I(inode)->i_orphan)) {
- list_del_init(&BTRFS_I(inode)->i_orphan);
+ if (test_and_clear_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
+ &BTRFS_I(inode)->runtime_flags))
delete_item = 1;
- }
- if (BTRFS_I(inode)->orphan_meta_reserved) {
- BTRFS_I(inode)->orphan_meta_reserved = 0;
+ if (test_and_clear_bit(BTRFS_INODE_ORPHAN_META_RESERVED,
+ &BTRFS_I(inode)->runtime_flags))
release_rsv = 1;
- }
spin_unlock(&root->orphan_lock);
if (trans && delete_item) {
@@ -2212,8 +2247,10 @@ int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode)
BUG_ON(ret); /* -ENOMEM or corruption (JDM: Recheck) */
}
- if (release_rsv)
+ if (release_rsv) {
btrfs_orphan_release_metadata(inode);
+ atomic_dec(&root->orphan_inodes);
+ }
return 0;
}
@@ -2341,6 +2378,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
ret = PTR_ERR(trans);
goto out;
}
+ printk(KERN_ERR "auto deleting %Lu\n",
+ found_key.objectid);
ret = btrfs_del_orphan_item(trans, root,
found_key.objectid);
BUG_ON(ret); /* -ENOMEM or corruption (JDM: Recheck) */
@@ -2352,9 +2391,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root)
* add this inode to the orphan list so btrfs_orphan_del does
* the proper thing when we hit it
*/
- spin_lock(&root->orphan_lock);
- list_add(&BTRFS_I(inode)->i_orphan, &root->orphan_list);
- spin_unlock(&root->orphan_lock);
+ set_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
+ &BTRFS_I(inode)->runtime_flags);
/* if we have links, this was a truncate, lets do that */
if (inode->i_nlink) {
@@ -2510,7 +2548,7 @@ static void btrfs_read_locked_inode(struct inode *inode)
inode_set_bytes(inode, btrfs_inode_nbytes(leaf, inode_item));
BTRFS_I(inode)->generation = btrfs_inode_generation(leaf, inode_item);
- BTRFS_I(inode)->sequence = btrfs_inode_sequence(leaf, inode_item);
+ inode->i_version = btrfs_inode_sequence(leaf, inode_item);
inode->i_generation = BTRFS_I(inode)->generation;
inode->i_rdev = 0;
rdev = btrfs_inode_rdev(leaf, inode_item);
@@ -2594,7 +2632,7 @@ static void fill_inode_item(struct btrfs_trans_handle *trans,
btrfs_set_inode_nbytes(leaf, item, inode_get_bytes(inode));
btrfs_set_inode_generation(leaf, item, BTRFS_I(inode)->generation);
- btrfs_set_inode_sequence(leaf, item, BTRFS_I(inode)->sequence);
+ btrfs_set_inode_sequence(leaf, item, inode->i_version);
btrfs_set_inode_transid(leaf, item, trans->transid);
btrfs_set_inode_rdev(leaf, item, inode->i_rdev);
btrfs_set_inode_flags(leaf, item, BTRFS_I(inode)->flags);
@@ -2752,6 +2790,8 @@ err:
goto out;
btrfs_i_size_write(dir, dir->i_size - name_len * 2);
+ inode_inc_iversion(inode);
+ inode_inc_iversion(dir);
inode->i_ctime = dir->i_mtime = dir->i_ctime = CURRENT_TIME;
btrfs_update_inode(trans, root, dir);
out:
@@ -3089,6 +3129,7 @@ int btrfs_unlink_subvol(struct btrfs_trans_handle *trans,
}
btrfs_i_size_write(dir, dir->i_size - name_len * 2);
+ inode_inc_iversion(dir);
dir->i_mtime = dir->i_ctime = CURRENT_TIME;
ret = btrfs_update_inode(trans, root, dir);
if (ret)
@@ -3607,7 +3648,8 @@ static int btrfs_setsize(struct inode *inode, loff_t newsize)
* any new writes get down to disk quickly.
*/
if (newsize == 0)
- BTRFS_I(inode)->ordered_data_close = 1;
+ set_bit(BTRFS_INODE_ORDERED_DATA_CLOSE,
+ &BTRFS_I(inode)->runtime_flags);
/* we don't support swapfiles, so vmtruncate shouldn't fail */
truncate_setsize(inode, newsize);
@@ -3638,6 +3680,7 @@ static int btrfs_setattr(struct dentry *dentry, struct iattr *attr)
if (attr->ia_valid) {
setattr_copy(inode, attr);
+ inode_inc_iversion(inode);
err = btrfs_dirty_inode(inode);
if (!err && attr->ia_valid & ATTR_MODE)
@@ -3671,7 +3714,8 @@ void btrfs_evict_inode(struct inode *inode)
btrfs_wait_ordered_range(inode, 0, (u64)-1);
if (root->fs_info->log_root_recovering) {
- BUG_ON(!list_empty(&BTRFS_I(inode)->i_orphan));
+ BUG_ON(!test_bit(BTRFS_INODE_HAS_ORPHAN_ITEM,
+ &BTRFS_I(inode)->runtime_flags));
goto no_delete;
}
@@ -4066,7 +4110,7 @@ static struct inode *new_simple_dir(struct super_block *s,
BTRFS_I(inode)->root = root;
memcpy(&BTRFS_I(inode)->location, key, sizeof(*key));
- BTRFS_I(inode)->dummy_inode = 1;
+ set_bit(BTRFS_INODE_DUMMY, &BTRFS_I(inode)->runtime_flags);
inode->i_ino = BTRFS_EMPTY_SUBVOL_DIR_OBJECTID;
inode->i_op = &btrfs_dir_ro_inode_operations;
@@ -4370,7 +4414,7 @@ int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc)
int ret = 0;
bool nolock = false;
- if (BTRFS_I(inode)->dummy_inode)
+ if (test_bit(BTRFS_INODE_DUMMY, &BTRFS_I(inode)->runtime_flags))
return 0;
if (btrfs_fs_closing(root->fs_info) && btrfs_is_free_space_inode(root, inode))
@@ -4403,7 +4447,7 @@ int btrfs_dirty_inode(struct inode *inode)
struct btrfs_trans_handle *trans;
int ret;
- if (BTRFS_I(inode)->dummy_inode)
+ if (test_bit(BTRFS_INODE_DUMMY, &BTRFS_I(inode)->runtime_flags))
return 0;
trans = btrfs_join_transaction(root);
@@ -4431,46 +4475,18 @@ int btrfs_dirty_inode(struct inode *inode)
* This is a copy of file_update_time. We need this so we can return error on
* ENOSPC for updating the inode in the case of file write and mmap writes.
*/
-int btrfs_update_time(struct file *file)
+static int btrfs_update_time(struct inode *inode, struct timespec *now,
+ int flags)
{
- struct inode *inode = file->f_path.dentry->d_inode;
- struct timespec now;
- int ret;
- enum { S_MTIME = 1, S_CTIME = 2, S_VERSION = 4 } sync_it = 0;
-
- /* First try to exhaust all avenues to not sync */
- if (IS_NOCMTIME(inode))
- return 0;
-
- now = current_fs_time(inode->i_sb);
- if (!timespec_equal(&inode->i_mtime, &now))
- sync_it = S_MTIME;
-
- if (!timespec_equal(&inode->i_ctime, &now))
- sync_it |= S_CTIME;
-
- if (IS_I_VERSION(inode))
- sync_it |= S_VERSION;
-
- if (!sync_it)
- return 0;
-
- /* Finally allowed to write? Takes lock. */
- if (mnt_want_write_file(file))
- return 0;
-
- /* Only change inode inside the lock region */
- if (sync_it & S_VERSION)
+ if (flags & S_VERSION)
inode_inc_iversion(inode);
- if (sync_it & S_CTIME)
- inode->i_ctime = now;
- if (sync_it & S_MTIME)
- inode->i_mtime = now;
- ret = btrfs_dirty_inode(inode);
- if (!ret)
- mark_inode_dirty_sync(inode);
- mnt_drop_write(file->f_path.mnt);
- return ret;
+ if (flags & S_CTIME)
+ inode->i_ctime = *now;
+ if (flags & S_MTIME)
+ inode->i_mtime = *now;
+ if (flags & S_ATIME)
+ inode->i_atime = *now;
+ return btrfs_dirty_inode(inode);
}
/*
@@ -4730,6 +4746,7 @@ int btrfs_add_link(struct btrfs_trans_handle *trans,
btrfs_i_size_write(parent_inode, parent_inode->i_size +
name_len * 2);
+ inode_inc_iversion(parent_inode);
parent_inode->i_mtime = parent_inode->i_ctime = CURRENT_TIME;
ret = btrfs_update_inode(trans, root, parent_inode);
if (ret)
@@ -4937,6 +4954,7 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
}
btrfs_inc_nlink(inode);
+ inode_inc_iversion(inode);
inode->i_ctime = CURRENT_TIME;
ihold(inode);
@@ -5903,9 +5921,7 @@ static void btrfs_endio_direct_write(struct bio *bio, int err)
struct btrfs_dio_private *dip = bio->bi_private;
struct inode *inode = dip->inode;
struct btrfs_root *root = BTRFS_I(inode)->root;
- struct btrfs_trans_handle *trans;
struct btrfs_ordered_extent *ordered = NULL;
- struct extent_state *cached_state = NULL;
u64 ordered_offset = dip->logical_offset;
u64 ordered_bytes = dip->bytes;
int ret;
@@ -5915,73 +5931,14 @@ static void btrfs_endio_direct_write(struct bio *bio, int err)
again:
ret = btrfs_dec_test_first_ordered_pending(inode, &ordered,
&ordered_offset,
- ordered_bytes);
+ ordered_bytes, !err);
if (!ret)
goto out_test;
- BUG_ON(!ordered);
-
- trans = btrfs_join_transaction(root);
- if (IS_ERR(trans)) {
- err = -ENOMEM;
- goto out;
- }
- trans->block_rsv = &root->fs_info->delalloc_block_rsv;
-
- if (test_bit(BTRFS_ORDERED_NOCOW, &ordered->flags)) {
- ret = btrfs_ordered_update_i_size(inode, 0, ordered);
- if (!ret)
- err = btrfs_update_inode_fallback(trans, root, inode);
- goto out;
- }
-
- lock_extent_bits(&BTRFS_I(inode)->io_tree, ordered->file_offset,
- ordered->file_offset + ordered->len - 1, 0,
- &cached_state);
-
- if (test_bit(BTRFS_ORDERED_PREALLOC, &ordered->flags)) {
- ret = btrfs_mark_extent_written(trans, inode,
- ordered->file_offset,
- ordered->file_offset +
- ordered->len);
- if (ret) {
- err = ret;
- goto out_unlock;
- }
- } else {
- ret = insert_reserved_file_extent(trans, inode,
- ordered->file_offset,
- ordered->start,
- ordered->disk_len,
- ordered->len,
- ordered->len,
- 0, 0, 0,
- BTRFS_FILE_EXTENT_REG);
- unpin_extent_cache(&BTRFS_I(inode)->extent_tree,
- ordered->file_offset, ordered->len);
- if (ret) {
- err = ret;
- WARN_ON(1);
- goto out_unlock;
- }
- }
-
- add_pending_csums(trans, inode, ordered->file_offset, &ordered->list);
- ret = btrfs_ordered_update_i_size(inode, 0, ordered);
- if (!ret || !test_bit(BTRFS_ORDERED_PREALLOC, &ordered->flags))
- btrfs_update_inode_fallback(trans, root, inode);
- ret = 0;
-out_unlock:
- unlock_extent_cached(&BTRFS_I(inode)->io_tree, ordered->file_offset,
- ordered->file_offset + ordered->len - 1,
- &cached_state, GFP_NOFS);
-out:
- btrfs_delalloc_release_metadata(inode, ordered->len);
- btrfs_end_transaction(trans, root);
- ordered_offset = ordered->file_offset + ordered->len;
- btrfs_put_ordered_extent(ordered);
- btrfs_put_ordered_extent(ordered);
-
+ ordered->work.func = finish_ordered_fn;
+ ordered->work.flags = 0;
+ btrfs_queue_worker(&root->fs_info->endio_write_workers,
+ &ordered->work);
out_test:
/*
* our bio might span multiple ordered extents. If we haven't
@@ -5990,12 +5947,12 @@ out_test:
if (ordered_offset < dip->logical_offset + dip->bytes) {
ordered_bytes = dip->logical_offset + dip->bytes -
ordered_offset;
+ ordered = NULL;
goto again;
}
out_done:
bio->bi_private = dip->private;
- kfree(dip->csums);
kfree(dip);
/* If we had an error make sure to clear the uptodate flag */
@@ -6063,9 +6020,12 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode,
int ret;
bio_get(bio);
- ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
- if (ret)
- goto err;
+
+ if (!write) {
+ ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
+ if (ret)
+ goto err;
+ }
if (skip_sum)
goto map;
@@ -6485,13 +6445,13 @@ static int btrfs_releasepage(struct page *page, gfp_t gfp_flags)
static void btrfs_invalidatepage(struct page *page, unsigned long offset)
{
+ struct inode *inode = page->mapping->host;
struct extent_io_tree *tree;
struct btrfs_ordered_extent *ordered;
struct extent_state *cached_state = NULL;
u64 page_start = page_offset(page);
u64 page_end = page_start + PAGE_CACHE_SIZE - 1;
-
/*
* we have the page locked, so new writeback can't start,
* and the dirty bit won't be cleared while we are here.
@@ -6501,13 +6461,13 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
*/
wait_on_page_writeback(page);
- tree = &BTRFS_I(page->mapping->host)->io_tree;
+ tree = &BTRFS_I(inode)->io_tree;
if (offset) {
btrfs_releasepage(page, GFP_NOFS);
return;
}
lock_extent_bits(tree, page_start, page_end, 0, &cached_state);
- ordered = btrfs_lookup_ordered_extent(page->mapping->host,
+ ordered = btrfs_lookup_ordered_extent(inode,
page_offset(page));
if (ordered) {
/*
@@ -6522,9 +6482,10 @@ static void btrfs_