linux-stable.git/fs, branch v3.4.111

get rid of s_files and files_lock

2016-03-21T01:17:55+00:00

commit eee5cc2702929fd41cce28058dc6d6717f723f87 upstream.

The only thing we need it for is alt-sysrq-r (emergency remount r/o)
and these days we can do just as well without going through the
list of files.

Signed-off-by: Al Viro 
[lizf: Backported to 3.4: adjust context]
Signed-off-by: Zefan Li

ocfs2: fix BUG in ocfs2_downconvert_thread_do_work()

2016-03-21T01:17:51+00:00

commit 209f7512d007980fd111a74a064d70a3656079cf upstream.

The "BUG_ON(list_empty(&osb->blocked_lock_list))" in
ocfs2_downconvert_thread_do_work can be triggered in the following case:

ocfs2dc has firstly saved osb->blocked_lock_count to local varibale
processed, and then processes the dentry lockres.  During the dentry
put, it calls iput and then deletes rw, inode and open lockres from
blocked list in ocfs2_mark_lockres_freeing.  And this causes the
variable `processed' to not reflect the number of blocked lockres to be
processed, which triggers the BUG.

Signed-off-by: Joseph Qi 
Cc: Mark Fasheh 
Cc: Joel Becker 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds 
Signed-off-by: Zefan Li

9p: don't leave a half-initialized inode sitting around

2016-03-21T01:17:41+00:00

commit 0a73d0a204a4a04a1e110539c5a524ae51f91d6d upstream.

Signed-off-by: Al Viro 
Signed-off-by: Zefan Li

hpfs: hpfs_error: Remove static buffer, use vsprintf extension %pV instead

2016-03-21T01:17:41+00:00

commit a28e4b2b18ccb90df402da3f21e1a83c9d4f8ec1 upstream.

Removing unnecessary static buffers is good.
Use the vsprintf %pV extension instead.

Signed-off-by: Joe Perches 
Signed-off-by: Mikulas Patocka 
Signed-off-by: Linus Torvalds 
[Mikulas:
 - The bug corrected by the patch is - if hpfs_error is called concurrently
   on multiple filesystems, it could corrupt the string because the text
   buffer is shared. That's why I marked the patch for stable.]
[lizf: Backported to 3.4: adjust context]
Signed-off-by: Zefan Li

hpfs: kstrdup() out of memory handling

2016-03-21T01:17:40+00:00

commit ce657611baf902f14ae559ce4e0787ead6712067 upstream.

There is a possibility of nothing being allocated to the new_opts in
case of memory pressure, therefore return ENOMEM for such case.

Signed-off-by: Sanidhya Kashyap 
Signed-off-by: Mikulas Patocka 
Signed-off-by: Linus Torvalds 
[lizf: Backported to 3.4: adjust context]
Signed-off-by: Zefan Li

ext4: replace open coded nofail allocation in ext4_free_blocks()

2016-03-21T01:17:40+00:00

commit 7444a072c387a93ebee7066e8aee776954ab0e41 upstream.

ext4_free_blocks is looping around the allocation request and mimics
__GFP_NOFAIL behavior without any allocation fallback strategy. Let's
remove the open coded loop and replace it with __GFP_NOFAIL. Without the
flag the allocator has no way to find out never-fail requirement and
cannot help in any way.

Signed-off-by: Michal Hocko 
Signed-off-by: Theodore Ts'o 
Signed-off-by: Zefan Li

ext4: avoid deadlocks in the writeback path by using sb_getblk_gfp

2016-03-21T01:17:40+00:00

commit c45653c341f5c8a0ce19c8f0ad4678640849cb86 upstream.

Switch ext4 to using sb_getblk_gfp with GFP_NOFS added to fix possible
deadlocks in the page writeback path.

Signed-off-by: Nikolay Borisov 
Signed-off-by: Theodore Ts'o 
[lizf: Backported to 3.4: adjust context]
Signed-off-by: Zefan Li

fs/buffer.c: support buffer cache allocations with gfp modifiers

2016-03-21T01:17:39+00:00

commit 3b5e6454aaf6b4439b19400d8365e2ec2d24e411 upstream.

A buffer cache is allocated from movable area because it is referred
for a while and released soon.  But some filesystems are taking buffer
cache for a long time and it can disturb page migration.

New APIs are introduced to allocate buffer cache with user specific
flag.  *_gfp APIs are for user want to set page allocation flag for
page cache allocation.  And *_unmovable APIs are for the user wants to
allocate page cache from non-movable area.

Signed-off-by: Gioh Kim 
Signed-off-by: Theodore Ts'o 
Reviewed-by: Jan Kara 
[lizf: Backported to 3.4: adjust context]
Signed-off-by: Zefan Li

Btrfs: use kmem_cache_free when freeing entry in inode cache

2016-03-21T01:17:39+00:00

commit c3f4a1685bb87e59c886ee68f7967eae07d4dffa upstream.

The free space entries are allocated using kmem_cache_zalloc(),
through __btrfs_add_free_space(), therefore we should use
kmem_cache_free() and not kfree() to avoid any confusion and
any potential problem. Looking at the kfree() definition at
mm/slab.c it has the following comment:

  /*
   * (...)
   *
   * Don't free memory not originally allocated by kmalloc()
   * or you will run into trouble.
   */

So better be safe and use kmem_cache_free().

Signed-off-by: Filipe Manana 
Reviewed-by: David Sterba 
Signed-off-by: Chris Mason 
Signed-off-by: Zefan Li

Failing to send a CLOSE if file is opened WRONLY and server reboots on a 4.x mount

2015-10-22T01:20:08+00:00

commit a41cbe86df3afbc82311a1640e20858c0cd7e065 upstream.

A test case is as the description says:
open(foobar, O_WRONLY);
sleep()  --> reboot the server
close(foobar)

The bug is because in nfs4state.c in nfs4_reclaim_open_state() a few
line before going to restart, there is
clear_bit(NFS4CLNT_RECLAIM_NOGRACE, &state->flags).

NFS4CLNT_RECLAIM_NOGRACE is a flag for the client states not open
owner states. Value of NFS4CLNT_RECLAIM_NOGRACE is 4 which is the
value of NFS_O_WRONLY_STATE in nfs4_state->flags. So clearing it wipes
out state and when we go to close it, “call_close” doesn’t get set as
state flag is not set and CLOSE doesn’t go on the wire.

Signed-off-by: Olga Kornievskaia 
Signed-off-by: Trond Myklebust 
Signed-off-by: Zefan Li