xfs: Check new inode size is OK before preallocating
authorDave Chinner <dchinner@redhat.com>
Fri, 21 May 2010 05:47:59 +0000 (05:47 +0000)
committerAlex Elder <aelder@sgi.com>
Fri, 28 May 2010 20:19:12 +0000 (15:19 -0500)
commit07f1a4f5e89cd4e6c79d67d41e8a18c451214ae2
treea4efa89cd8536b6ff6b75713fd5cc342f4da4f62
parentfdc07f44c891d3fdee7722a03e3881614a293b3c
xfs: Check new inode size is OK before preallocating

The new xfsqa test 228 tries to preallocate more space than the
filesystem contains. it should fail, but instead triggers an assert
about lock flags.  The failure is due to the size extension failing
in vmtruncate() due to rlimit being set. Check this before we start
the preallocation to avoid allocating space that will never be used.

Also the path through xfs_vn_allocate already holds the IO lock, so
it should not be present in the lock flags when the setattr fails.
Hence the assert needs to take this into account. This will prevent
other such callers from hitting this incorrect ASSERT.

(Fixed a reference to "newsize" to read "new_size". -Alex)

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>
fs/xfs/linux-2.6/xfs_iops.c
fs/xfs/xfs_vnodeops.c