linux-stable/fs/ocfs2
Ryan Ding e63890f38a ocfs2: fix ip_unaligned_aio deadlock with dio work queue
In the current implementation of unaligned aio+dio, lock order behave as
follow:

in user process context:
  -> call io_submit()
    -> get i_mutex
		<== window1
      -> get ip_unaligned_aio
        -> submit direct io to block device
    -> release i_mutex
  -> io_submit() return

in dio work queue context(the work queue is created in __blockdev_direct_IO):
  -> release ip_unaligned_aio
		<== window2
    -> get i_mutex
      -> clear unwritten flag & change i_size
    -> release i_mutex

There is a limitation to the thread number of dio work queue.  256 at
default.  If all 256 thread are in the above 'window2' stage, and there
is a user process in the 'window1' stage, the system will became
deadlock.  Since the user process hold i_mutex to wait ip_unaligned_aio
lock, while there is a direct bio hold ip_unaligned_aio mutex who is
waiting for a dio work queue thread to be schedule.  But all the dio
work queue thread is waiting for i_mutex lock in 'window2'.

This case only happened in a test which send a large number(more than
256) of aio at one io_submit() call.

My design is to remove ip_unaligned_aio lock.  Change it to a sync io
instead.  Just like ip_unaligned_aio lock, serialize the unaligned aio
dio.

[akpm@linux-foundation.org: remove OCFS2_IOCB_UNALIGNED_IO, per Junxiao Bi]
Signed-off-by: Ryan Ding <ryan.ding@oracle.com>
Reviewed-by: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Joseph Qi <joseph.qi@huawei.com>
Cc: Mark Fasheh <mfasheh@suse.de>
Cc: Joel Becker <jlbec@evilplan.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-03-25 16:37:42 -07:00
..
cluster ocfs2: o2hb: fix double free bug 2016-03-25 16:37:42 -07:00
dlm
dlmfs
acl.c
acl.h
alloc.c
alloc.h
aops.c ocfs2: fix ip_unaligned_aio deadlock with dio work queue 2016-03-25 16:37:42 -07:00
aops.h ocfs2: fix ip_unaligned_aio deadlock with dio work queue 2016-03-25 16:37:42 -07:00
blockcheck.c
blockcheck.h
buffer_head_io.c
buffer_head_io.h
dcache.c
dcache.h
dir.c
dir.h
dlmglue.c
dlmglue.h
export.c
export.h
extent_map.c
extent_map.h
file.c ocfs2: fix ip_unaligned_aio deadlock with dio work queue 2016-03-25 16:37:42 -07:00
file.h
filecheck.c ocfs2: sysfile interfaces for online file check 2016-03-22 15:36:02 -07:00
filecheck.h ocfs2: sysfile interfaces for online file check 2016-03-22 15:36:02 -07:00
heartbeat.c
heartbeat.h
inode.c ocfs2: record UNWRITTEN extents when populate write desc 2016-03-25 16:37:42 -07:00
inode.h ocfs2: fix ip_unaligned_aio deadlock with dio work queue 2016-03-25 16:37:42 -07:00
ioctl.c
ioctl.h
journal.c
journal.h
Kconfig
localalloc.c
localalloc.h
locks.c
locks.h
Makefile ocfs2: sysfile interfaces for online file check 2016-03-22 15:36:02 -07:00
mmap.c ocfs2: add ocfs2_write_type_t type to identify the caller of write 2016-03-25 16:37:42 -07:00
mmap.h
move_extents.c
move_extents.h
namei.c
namei.h
ocfs1_fs_compat.h
ocfs2.h
ocfs2_fs.h
ocfs2_ioctl.h
ocfs2_lockid.h
ocfs2_lockingver.h
ocfs2_trace.h ocfs2: code clean up for direct io 2016-03-25 16:37:42 -07:00
quota.h
quota_global.c
quota_local.c
refcounttree.c
refcounttree.h
reservations.c
reservations.h
resize.c
resize.h
slot_map.c
slot_map.h
stack_o2cb.c
stack_user.c
stackglue.c ocfs2: export ocfs2_kset for online file check 2016-03-22 15:36:02 -07:00
stackglue.h ocfs2: export ocfs2_kset for online file check 2016-03-22 15:36:02 -07:00
suballoc.c
suballoc.h
super.c ocfs2: fix ip_unaligned_aio deadlock with dio work queue 2016-03-25 16:37:42 -07:00
super.h
symlink.c
symlink.h
sysfile.c
sysfile.h
uptodate.c
uptodate.h
xattr.c
xattr.h