linux-toradex.git/fs/ext3, branch v2.6.18.3

[PATCH] ext3 sequential read regression fix

2006-09-16T19:54:32+00:00

ext3-get-blocks support caused ~20% degrade in Sequential read
performance (tiobench). Problem is with marking the buffer boundary
so IO can be submitted right away. Here is the patch to fix it.

  2.6.18-rc6:
  -----------
  # ./iotest
  1048576+0 records in
  1048576+0 records out
  4294967296 bytes (4.3 GB) copied, 75.2726 seconds, 57.1 MB/s

  real    1m15.285s
  user    0m0.276s
  sys     0m3.884s

  2.6.18-rc6 + fix:
  -----------------
  [root@elm3a241 ~]# ./iotest
  1048576+0 records in
  1048576+0 records out
  4294967296 bytes (4.3 GB) copied, 62.9356 seconds, 68.2 MB/s

The boundary block check in ext3_get_blocks_handle needs to be adjusted
against the count of blocks mapped in this call, now that it can map
more than one block.

Signed-off-by: Suparna Bhattacharya 
Tested-by: Badari Pulavarty 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] knfsd: Make ext3 reject filehandles referring to invalid inode number

2006-09-16T19:54:31+00:00

Inodes earlier than the 'first' inode (e.g.  journal, resize) should be
rejected early - except the root inode.  Also inode numbers that are too
big should be rejected early.

[akpm@osdl.org: cleanup]
Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] ext3_getblk() should handle HOLE correctly

2006-09-08T17:22:50+00:00

It has been reported that ext3_getblk() is not doing the right thing and
triggering following WARN():

BUG: warning at fs/ext3/inode.c:1016/ext3_getblk()
  ext3_getblk+0x98/0x2a6   md_wakeup_thread+0x26/0x2a
  ext3_bread+0x1f/0x88   ext3_quota_read+0x136/0x1ae
  v1_read_dqblk+0x61/0xac   dquot_acquire+0xf6/0x107
  ext3_acquire_dquot+0x46/0x68   dqget+0x155/0x1e7
  dquot_transfer+0x3e0/0x3e9   dput+0x23/0x13e
  ext3_setattr+0xc3/0x240   current_fs_time+0x52/0x6a
  notify_change+0x2bd/0x30d   chown_common+0x9c/0xc5
  strncpy_from_user+0x3b/0x68   do_path_lookup+0xdf/0x266
  __user_walk_fd+0x44/0x5a   sys_chown+0x4a/0x55
  vfs_write+0xe7/0x13c   sys_mkdir+0x1f/0x23
  syscall_call+0x7/0xb

Looking at the code, it looks like it's not handle HOLE correctly.  It ends
up returning -EIO.  Here is the patch to fix it.

If we really want to be paranoid, we can allow return values 0 (HOLE), 1
(we asked for one block) and return -EIO for more than 1 block.  But I
really don't see a reason for doing it - all we need is the block# here.
(doesn't matter how many blocks are mapped).

ext3_get_blocks_handle() returns number of blocks it mapped.  It returns 0
in case of HOLE.  ext3_getblk() should handle HOLE properly (currently its
dumping warning stack and returning -EIO).

Signed-off-by: Badari Pulavarty 
Acked-by: Mingming Cao 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] ext3 filesystem bogus ENOSPC with reservation fix

2006-08-27T18:01:30+00:00

To handle the earlier bogus ENOSPC error caused by filesystem full of block
reservation, current code falls back to non block reservation, starts to
allocate block(s) from the goal allocation block group as if there is no
block reservation.

Current code needs to re-load the corresponding block group descriptor for
the initial goal block group in this case.  The patch fixes this.

Signed-off-by: Mingming Cao 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] ext3 -nobh option causes oops

2006-07-31T20:28:44+00:00

For files other than IFREG, nobh option doesn't make sense.  Modifications
to them are journalled and needs buffer heads to do that.  Without this
patch, we get kernel oops in page_buffers().

Signed-off-by: Badari Pulavarty 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] ext3: avoid triggering ext3_error on bad NFS file handle

2006-07-31T20:28:36+00:00

The inode number out of an NFS file handle gets passed eventually to
ext3_get_inode_block() without any checking.  If ext3_get_inode_block()
allows it to trigger an error, then bad filehandles can have unpleasant
effect - ext3_error() will usually cause a forced read-only remount, or a
panic if `errors=panic' was used.

So remove the call to ext3_error there and put a matching check in
ext3/namei.c where inode numbers are read off storage.

[akpm@osdl.org: fix off-by-one error]
Signed-off-by: Neil Brown 
Signed-off-by: Jan Kara 
Cc: Marcel Holtmann 
Cc: 
Cc: "Stephen C. Tweedie" 
Cc: Eric Sandeen 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] Remove leftover ext3 acl declarations

2006-07-10T20:24:26+00:00

These functions no longer exist; remove their declarations.

Signed-off-by: Andreas Gruenbacher 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] lockdep: annotate the quota code

2006-07-03T22:27:08+00:00

The quota code plays interesting games with the lock ordering; to quote Jan:

| i_mutex of inode containing quota file is acquired after all other
| quota locks. i_mutex of all other inodes is acquired before quota
| locks. Quota code makes sure (by resetting inode operations and
| setting special flag on inode) that noone tries to enter quota code
| while holding i_mutex on a quota file...

The good news is that all of this special case i_mutex grabbing happens in the
(per filesystem) low level quota write function.  For this special case we
need a new I_MUTEX_* nesting level, since this just entirely outside any of
the regular VFS locking rules for i_mutex.  I trust Jan on his blue eyes that
this is not ever going to deadlock; and based on that the patch below is what
it takes to inform lockdep of these very interesting new locking rules.

The new locking rule for the I_MUTEX_QUOTA nesting level is that this is the
deepest possible level of nesting for i_mutex, and that this only should be
used in quota write (and possibly read) function of filesystems.  This makes
the lock ordering of the I_MUTEX_* levels:

I_MUTEX_PARENT -> I_MUTEX_CHILD -> I_MUTEX_NORMAL -> I_MUTEX_QUOTA

Has no effect on non-lockdep kernels.

Signed-off-by: Arjan van de Ven 
Acked-by: Ingo Molnar 
Cc: Jan Kara 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

Remove obsolete #include

2006-06-30T17:25:36+00:00

Signed-off-by: Jörn Engel 
Signed-off-by: Adrian Bunk

[PATCH] mark address_space_operations const

2006-06-28T21:59:04+00:00

Same as with already do with the file operations: keep them in .rodata and
prevents people from doing runtime patching.

Signed-off-by: Christoph Hellwig 
Cc: Steven French 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds