linux.git/fs/ext4, branch v2.6.25

vfs: fix possible deadlock in ext2, ext3, ext4 when using xattrs

2008-04-16T02:35:41+00:00

mb_cache_entry_alloc() was allocating cache entries with GFP_KERNEL.  But
filesystems are calling this function while holding xattr_sem so possible
recursion into the fs violates locking ordering of xattr_sem and transaction
start / i_mutex for ext2-4.  Change mb_cache_entry_alloc() so that filesystems
can specify desired gfp mask and use GFP_NOFS from all of them.

Signed-off-by: Jan Kara 
Reported-by: Dave Jones 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

ext4: add missing ext4_journal_stop()

2008-02-25T20:37:42+00:00

Add missing ext4_journal_stop() in error handling.

Signed-off-by: Akinobu Mita 
Signed-off-by: "Theodore Ts'o" 
Cc: Stephen Tweedie 
Cc: adilger@clusterfs.com
Cc: Andrew Morton 
Cc: Mingming Cao

ext4: ext4_find_next_zero_bit needs an aligned address on some arch

2008-02-23T06:38:34+00:00

ext4_find_next_zero_bit and ext4_find_next_bit needs a long aligned
address on x8_64. Add mb_find_next_zero_bit and mb_find_next_bit
and use them in the mballoc.

Fix: https://bugzilla.redhat.com/show_bug.cgi?id=433286

Eric Sandeen debugged the problem and suggested the fix.

Signed-off-by: Aneesh Kumar K.V 
Acked-by:      Eric Sandeen 
Signed-off-by: Mingming Cao 
Signed-off-by: "Theodore Ts'o"

ext4: set EXT4_EXTENTS_FL only for directory and regular files

2008-02-25T21:38:03+00:00

In addition, don't inherit EXT4_EXTENTS_FL from parent directory.
If we have a directory with extent flag set and later mount the file
system with -o noextents, the files created in that directory will also
have extent flag set but we would not have called ext4_ext_tree_init for
them. This will cause error later when we are verifying the extent header

Signed-off-by: Aneesh Kumar K.V 
Acked-off-by: Eric Sandeen 
Signed-off-by: Mingming Cao 
Signed-off-by: "Theodore Ts'o"

ext4: Don't mark filesystem error if fallocate fails

2008-02-25T20:41:35+00:00

If we fail to allocate blocks don't call ext4_error. Also don't hide
errors from ext4_get_blocks_wrap

Signed-off-by: Aneesh Kumar K.V 
Signed-off-by: "Theodore Ts'o"

ext4: Fix BUG when writing to an unitialized extent

2008-02-25T20:29:55+00:00

This patch fixes a bug when writing to preallocated but uninitialized
blocks, which resulted in a BUG in fs/buffer.c saying that the buffer
is not mapped.

When writing to a file, ext4_get_block_wrap() is called with create=1 in
order to request that blocks be allocated if necessary.  It currently
calls ext4_get_blocks() with create=0 in order to do a lookup first.  If
the inode contains an unitialized data block, the buffer head is left
unampped, which ext4_get_blocks_wrap() returns, causing the BUG.

We fix this by checking to see if the buffer head is unmapped, and if
so, we make sure the the buffer head is mapped by calling
ext4_ext_get_blocks with create=1.

Signed-off-by: Mingming Cao 
Signed-off-by: "Theodore Ts'o"

ext4: Don't use ext4_dec_count() if not needed

2008-02-15T20:00:38+00:00

The ext4_dec_count() function is only needed when dropping the i_nlink
count on inodes which are (or which could be) directories.  If we
*know* that the inode in question can't possibly be a directory, use
drop_nlink or clear_nlink() if we know i_nlink is 1.

Signed-off-by: "Theodore Ts'o"

ext4: modify block allocation algorithm for the last group

2008-02-15T18:43:07+00:00

When a directory inode is allocated in the last group and the last group
contains less than s_blocks_per_group blocks, the initial block allocated
for the directory is not always allocated in the same group as the
directory inode, but in one of the first groups of the filesystem (group 1
for example).
Depending on the current process's pid, ext4_find_near() and 
ext4_ext_find_goal() can return a block number greater than the maximum
blocks count in the filesystem and in that case the block will be not
allocated in the same group as the inode.

The following patch fixes the problem.

Should the modification also be done in ext2/3 code?

Signed-off-by: Valerie Clement 
Signed-off-by: Mingming Cao 
Signed-off-by: "Theodore Ts'o"

ext4: Don't claim block from group which has corrupt bitmap

2008-02-15T18:48:21+00:00

In ext4_mb_complex_scan_group, if the extent length of the newly
found extentet is greater than than the total free blocks counted
in group info, break without claiming the block.

Document different ext4_error usage, explaining the state with which we
continue if we mount with errors=continue

Signed-off-by: Aneesh Kumar K.V 
Signed-off-by: Mingming Cao 
Signed-off-by: "Theodore Ts'o"

ext4: Get journal write access before modifying the extent tree

2008-02-22T11:17:31+00:00

When the user was writing into an unitialized extent,
ext4_ext_convert_to_initialize() was not requesting journal write access
before it started to modify the extent tree.   Fix this oversight.

Signed-off-by: Aneesh Kumar K.V 
Signed-off-by: Mingming Cao 
Signed-off-by: "Theodore Ts'o"