kernel_optimize_test/fs/ocfs2
Xue jiufei b246d3d11e ocfs2: fix a deadlock while o2net_wq doing direct memory reclaim
Fix a deadlock problem caused by direct memory reclaim in o2net_wq.  The
situation is as follows:

1) Receive a connect message from another node, node queues a
   work_struct o2net_listen_work.

2) o2net_wq processes this work and call the following functions:

o2net_wq
-> o2net_accept_one
  -> sock_create_lite
    -> sock_alloc()
      -> kmem_cache_alloc with GFP_KERNEL
        -> ____cache_alloc_node
          ->__alloc_pages_nodemask
            -> do_try_to_free_pages
              -> shrink_slab
                -> evict
                  -> ocfs2_evict_inode
                    -> ocfs2_drop_lock
                      -> dlmunlock
                        -> o2net_send_message_vec

   then o2net_wq wait for the unlock reply from master.

3) tcp layer received the reply, call o2net_data_ready() and queue
   sc_rx_work, waiting o2net_wq to process this work.

4) o2net_wq is a single thread workqueue, it process the work one by
   one.  Right now it is still doing o2net_listen_work and cannot handle
   sc_rx_work.  so we deadlock.

Junxiao Bi's patch "mm: clear __GFP_FS when PF_MEMALLOC_NOIO is set"
(http://ozlabs.org/~akpm/mmots/broken-out/mm-clear-__gfp_fs-when-pf_memalloc_noio-is-set.patch)
clears __GFP_FS in memalloc_noio_flags() besides __GFP_IO.  We use
memalloc_noio_save() to set process flag PF_MEMALLOC_NOIO so that all
allocations done by this process are done as if GFP_NOIO was specified.
We are not reentering filesystem while doing memory reclaim.

Signed-off-by: joyce.xue <xuejiufei@huawei.com>
Cc: Junxiao Bi <junxiao.bi@oracle.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-10-09 22:25:58 -04:00
..
cluster ocfs2: fix a deadlock while o2net_wq doing direct memory reclaim 2014-10-09 22:25:58 -04:00
dlm ocfs2: fix deadlock between o2hb thread and o2net_wq 2014-10-09 22:25:47 -04:00
dlmfs
acl.c
acl.h
alloc.c
alloc.h
aops.c ocfs2: fix deadlock due to wrong locking order 2014-10-09 22:25:48 -04:00
aops.h
blockcheck.c
blockcheck.h
buffer_head_io.c
buffer_head_io.h
dcache.c
dcache.h
dir.c
dir.h
dlmglue.c fs/ocfs2/dlmglue.c: use __seq_open_private() not seq_open() 2014-10-09 22:25:47 -04:00
dlmglue.h
export.c
export.h
extent_map.c
extent_map.h
file.c ocfs2: fix deadlock due to wrong locking order 2014-10-09 22:25:48 -04:00
file.h
heartbeat.c
heartbeat.h
inode.c
inode.h ocfs2: fix shift left operations overflow 2014-10-09 22:25:47 -04:00
ioctl.c
ioctl.h
journal.c
journal.h
Kconfig
localalloc.c
localalloc.h
locks.c
locks.h
Makefile
mmap.c
mmap.h
move_extents.c ocfs2: fix shift left operations overflow 2014-10-09 22:25:47 -04:00
move_extents.h
namei.c
namei.h
ocfs1_fs_compat.h
ocfs2_fs.h
ocfs2_ioctl.h
ocfs2_lockid.h
ocfs2_lockingver.h
ocfs2_trace.h
ocfs2.h
quota_global.c
quota_local.c
quota.h
refcounttree.c
refcounttree.h
reservations.c
reservations.h
resize.c
resize.h
slot_map.c
slot_map.h
stack_o2cb.c
stack_user.c fs/ocfs2/stack_user.c: fix typo in ocfs2_control_release() 2014-10-09 22:25:46 -04:00
stackglue.c
stackglue.h
suballoc.c
suballoc.h
super.c ocfs2: free vol_label in ocfs2_delete_osb() 2014-09-26 08:10:34 -07:00
super.h
symlink.c
symlink.h
sysfile.c
sysfile.h
uptodate.c
uptodate.h
xattr.c
xattr.h