https://github.com/torvalds/linux
Revision 866e4ed77448a0c311e1b055eb72ea05423fd799 authored by Christoph Hellwig on 27 August 2011, 05:57:44 UTC, committed by Alex Elder on 31 August 2011, 22:59:39 UTC
During umount we do not add a dirty inode to the lru and wait for it to
become clean first, but force writeback of data and metadata with
I_WILL_FREE set.  Currently there is no way for XFS to detect that the
inode has been redirtied for metadata operations, as we skip the
mark_inode_dirty call during teardown.  Fix this by setting i_update_core
nanually in that case, so that the inode gets flushed during inode reclaim.

Alternatively we could enable calling mark_inode_dirty for inodes in
I_WILL_FREE state, and let the VFS dirty tracking handle this.  I decided
against this as we will get better I/O patterns from reclaim compared to
the synchronous writeout in write_inode_now, and always marking the inode
dirty in some way from xfs_mark_inode_dirty is a better safetly net in
either case.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Alex Elder <aelder@sgi.com>
(cherry picked from commit da6742a5a4cc844a9982fdd936ddb537c0747856)

Signed-off-by: Alex Elder <aelder@sgi.com>
1 parent 242d621
History
Tip revision: 866e4ed77448a0c311e1b055eb72ea05423fd799 authored by Christoph Hellwig on 27 August 2011, 05:57:44 UTC
xfs: fix xfs_mark_inode_dirty during umount
Tip revision: 866e4ed
File Mode Size
Makefile -rw-r--r-- 394 bytes
compat.c -rw-r--r-- 17.0 KB
compat_mq.c -rw-r--r-- 4.1 KB
ipc_sysctl.c -rw-r--r-- 6.0 KB
ipcns_notifier.c -rw-r--r-- 2.2 KB
mq_sysctl.c -rw-r--r-- 2.7 KB
mqueue.c -rw-r--r-- 30.5 KB
msg.c -rw-r--r-- 20.9 KB
msgutil.c -rw-r--r-- 2.8 KB
namespace.c -rw-r--r-- 4.3 KB
sem.c -rw-r--r-- 40.4 KB
shm.c -rw-r--r-- 29.1 KB
syscall.c -rw-r--r-- 2.3 KB
util.c -rw-r--r-- 22.8 KB
util.h -rw-r--r-- 5.3 KB

back to top