linux-toradex.git/fs/fat, branch v2.6.27.12

vfat: fix 'sync' mount deadlock due to BKL->lock_super conversion

2008-08-20T15:31:19+00:00

There was another FAT BKL conversion deadlock reported by Bart
Trojanowski due to the BKL being used as a recursive lock by FAT, which
was missed because it only triggers with 'sync' (or 'dirsync') mounts.

The recursion worked for the BKL, but after the conversion to lock_super
(which uses a mutex), it just deadlocks.

Thanks to Bart for debugging this and testing the fix.  The lock
debugging information from the original report:

  =============================================
  [ INFO: possible recursive locking detected ]
  2.6.27-rc3-bisect-00448-ga7f5aaf #16
  ---------------------------------------------
  mv/4020 is trying to acquire lock:
   (&type->s_lock_key#9){--..}, at: [] lock_super+0x1e/0x20

  but task is already holding lock:
   (&type->s_lock_key#9){--..}, at: [] lock_super+0x1e/0x20

  other info that might help us debug this:
  3 locks held by mv/4020:
   #0:  (&sb->s_type->i_mutex_key#9/1){--..}, at: [] do_unlinkat+0x66/0x140
   #1:  (&sb->s_type->i_mutex_key#9){--..}, at: [] vfs_unlink+0x84/0x110
   #2:  (&type->s_lock_key#9){--..}, at: [] lock_super+0x1e/0x20

  stack backtrace:
  Pid: 4020, comm: mv Not tainted 2.6.27-rc3-bisect-00448-ga7f5aaf #16
   [] validate_chain+0x984/0xea0
   [] ? native_sched_clock+0x0/0xf0
   [] __lock_acquire+0x2ec/0x9b0
   [] lock_acquire+0x6f/0x90
   [] ? lock_super+0x1e/0x20
   [] mutex_lock_nested+0xad/0x300
   [] ? lock_super+0x1e/0x20
   [] ? lock_super+0x1e/0x20
   [] lock_super+0x1e/0x20
   [] fat_write_inode+0x60/0x2b0 [fat]
   [] ? _spin_unlock_irqrestore+0x48/0x80
   [] ? fat_sync_inode+0x3/0x20 [fat]
   [] fat_sync_inode+0x12/0x20 [fat]
   [] fat_remove_entries+0xbe/0x120 [fat]
   [] vfat_unlink+0x5f/0x90 [vfat]
   [] ? vfat_unlink+0x0/0x90 [vfat]
   [] vfs_unlink+0x98/0x110
   [] do_unlinkat+0x130/0x140
   [] ? audit_syscall_entry+0x105/0x150
   [] sys_unlinkat+0x3b/0x40
   [] sysenter_do_call+0x12/0x3f
   =======================

where the deadlock is due to the nesting of lock_super from vfat_unlink
to fat_write_inode:

 - do_unlinkat
   - vfs_unlink
     - vfat_unlink
       * lock_super
       - fat_remove_entries
         - fat_sync_inode
           - fat_write_inode
             * lock_super

and the fix is to simply remove the use of lock_super() in fat_write_inode.

The lock_super() there had been just an automatic conversion of the
kernel lock to the superblock lock, but no locking was actually needed
there, since the code in fat_write_inode already protected all relevant
accesses with a spinlock (sbi->inode_hash_lock to be exact).  The only
code inside the BKL (and thus the superblock lock) was accesses tp local
variables or calls to functions that have long been SMP-safe (i.e.
sb_bread, mark_buffe_dirty and brlese).

Bart reports:
 "Looks good.  I ran 10 parallel processes creating 1M files truncating
  them, writing to them again and then deleting them.  This patch fixes
  the issue I ran into.

  Signed-off-by: Bart Trojanowski "

Reported-and-tested-by: Bart Trojanowski 
Signed-off-by: Linus Torvalds

fat: Fix allow_utime option

2008-08-02T16:13:55+00:00

FAT has to handle the newly introduced ATTR_TIMES_SET for allow_utime
option.

Signed-off-by: OGAWA Hirofumi 
Signed-off-by: Linus Torvalds

[patch 3/4] fat: dont call notify_change

2008-07-27T00:53:27+00:00

The FAT_IOCTL_SET_ATTRIBUTES ioctl() calls notify_change() to change
the file mode before changing the inode attributes.  Replace with
explicit calls to security_inode_setattr(), fat_setattr() and
fsnotify_change().

This is equivalent to the original.  The reason it is needed, is that
later in the series we move the immutable check into notify_change().
That would break the FAT_IOCTL_SET_ATTRIBUTES ioctl, as it needs to
perform the mode change regardless of the immutability of the file.

[Fix error if fat is built as a module.  Thanks to OGAWA Hirofumi for
noticing.]

Signed-off-by: Miklos Szeredi 
Acked-by: OGAWA Hirofumi 
Signed-off-by: Al Viro

SL*B: drop kmem cache argument from constructor

2008-07-26T19:00:07+00:00

Kmem cache passed to constructor is only needed for constructors that are
themselves multiplexeres.  Nobody uses this "feature", nor does anybody uses
passed kmem cache in non-trivial way, so pass only pointer to object.

Non-trivial places are:
	arch/powerpc/mm/init_64.c
	arch/powerpc/mm/hugetlbpage.c

This is flag day, yes.

Signed-off-by: Alexey Dobriyan 
Acked-by: Pekka Enberg 
Acked-by: Christoph Lameter 
Cc: Jon Tollefson 
Cc: Nick Piggin 
Cc: Matt Mackall 
[akpm@linux-foundation.org: fix arch/powerpc/mm/hugetlbpage.c]
[akpm@linux-foundation.org: fix mm/slab.c]
[akpm@linux-foundation.org: fix ubifs]
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fatfs: add UTC timestamp option

2008-07-25T17:53:34+00:00

Provide a new mount option ("tz=UTC") for DOS (vfat/msdos) filesystems,
allowing timestamps to be in coordinated universal time (UTC) rather than
local time in applications where doing this is advantageous.

In particular, portable devices that use fat/vfat (such as digital
cameras) can benefit from using UTC in their internal clocks, thus
avoiding daylight saving time errors and general time ambiguity issues.
The user of the device does not have to worry about changing the time when
moving from place or when daylight saving changes.

The new mount option, when set, disables the counter-adjustment that Linux
currently makes to FAT timestamp info in anticipation of the normal
userspace time zone correction.  When used in this new mode, all daylight
saving time and time zone handling is done in userspace as is normal for
many other filesystems (like ext3).  The default mode, which remains
unchanged, is still appropriate when mounting volumes written in Windows
(because of its use of local time).

I originally based this patch on one submitted last year by Paul Collins,
but I updated it to work with current source and changed variable/option
naming.  Ogawa Hirofumi (who maintains these filesystems) and I discussed
this patch at length on lkml, and he suggested using the option name in
the attached version of the patch.  Barry Bouwsma pointed out a good
addition to the patch as well.

Signed-off-by: Joe Peterson 
Signed-off-by: Paul Collins 
Acked-by: OGAWA Hirofumi 
Cc: Barry Bouwsma 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: small optimization to __fat_readdir()

2008-07-25T17:53:34+00:00

This removes unnecessary parsing for directory entries.

If short_only, we don't need to parse longname.  And if !both and it found
the longname, we don't need shortname.

Signed-off-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: use same logic in fat_search_long() and __fat_readdir()

2008-07-25T17:53:34+00:00

This uses uses stack for shortname, and uses __getname() for longname in
fat_search_long() and __fat_readdir().  By this, it removes unneeded
__getname() for shortname.

Signed-off-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: cleanup fs/fat/dir.c

2008-07-25T17:53:34+00:00

This is no logic changes, just cleans fs/fat/dir.c up.

Signed-off-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat/dir.c: switch to struct __fat_dirent

2008-07-25T17:53:34+00:00

struct __fat_dirent is what was formerly the kernel struct dirent (that
was different from the userspace struct dirent).

Converting all fat users to struct __fat_dirent will allow us to get rid
of the conflicting struct dirent definition.

Signed-off-by: Adrian Bunk 
Signed-off-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fat: fix parse_options()

2008-07-25T17:53:34+00:00

Current parse_options() exits too early.  We need to run the code of
bottom in this function even if users doesn't specify options.

Signed-off-by: OGAWA Hirofumi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds