When inodes are marked stale in a transaction, they are treated
specially when the inode log item is being inserted into the AIL.
It tries to avoid moving the log item forward in the AIL due to a
race condition with the writing the underlying buffer back to disk.
The was "fixed" in commit de25c18 ("xfs: avoid moving stale inodes
in the AIL").

To avoid moving the item forward, we return a LSN smaller than the
commit_lsn of the completing transaction, thereby trying to trick
the commit code into not moving the inode forward at all. I'm not
sure this ever worked as intended - it assumes the inode is already
in the AIL, but I don't think the returned LSN would have been small
enough to prevent moving the inode. It appears that the reason it
worked is that the lower LSN of the inodes meant they were inserted
into the AIL and flushed before the inode buffer (which was moved to
the commit_lsn of the transaction).

The big problem is that with delayed logging, the returning of the
different LSN means insertion takes the slow, non-bulk path.  Worse
yet is that insertion is to a position -before- the commit_lsn so it
is doing a AIL traversal on every insertion, and has to walk over
all the items that have already been inserted into the AIL. It's
expensive.

To compound the matter further, with delayed logging inodes are
likely to go from clean to stale in a single checkpoint, which means
they aren't even in the AIL at all when we come across them at AIL
insertion time. Hence these were all getting inserted into the AIL
when they simply do not need to be as inodes marked XFS_ISTALE are
never written back.

Transactional/recovery integrity is maintained in this case by the
other items in the unlink transaction that were modified (e.g. the
AGI btree blocks) and committed in the same checkpoint.

So to fix this, simply unpin the stale inodes directly in
xfs_inode_item_committed() and return -1 to indicate that the AIL
insertion code does not need to do any further processing of these
inodes.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>

If the attribute fork on an inode is in btree format and has
multiple levels (i.e node format rather than leaf format), then a
lookup failure will trigger an assert failure in xfs_da_path_shift
if the flag XFS_DA_OP_OKNOENT is not set. This flag is used to
indicate to the directory btree code that not finding an entry is
not a fatal error. In the case of doing a lookup for a directory
name removal, this is valid as a user cannot insert an arbitrary
name to remove from the directory btree.

However, in the case of the attribute tree, a user has direct
control over the attribute name and can ask for any random name to
be removed without any validation. In this case, fsstress is asking
for a non-existent user.selinux attribute to be removed, and that is
causing xfs_da_path_shift() to fall off the bottom of the tree where
it asserts that a lookup failure is allowed. Because the flag is not
set, we die a horrible death on a debug enable kernel.

Prevent this assert from firing on attribute removes by adding the
op_flag XFS_DA_OP_OKNOENT to atribute removal operations.

Discovered when testing on a SELinux enabled system by fsstress in
test 070 by trying to remove a non-existent user.selinux attribute.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>

When an inode is truncated down, speculative preallocation is
removed from the inode. This should also reset the state bits for
controlling whether preallocation is subsequently removed when the
file is next closed. The flag is not being cleared, so repeated
operations on a file that first involve a truncate (e.g. multiple
repeated dd invocations on a file) give different file layouts for
the second and subsequent invocations.

Fix this by clearing the XFS_IDIRTY_RELEASE state bit when the
XFS_ITRUNCATED bit is detected in xfs_release() and hence ensure
that speculative delalloc is removed on files that have been
truncated down.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>

XFS inodes has several per-lifetime state fields that determine the
behaviour of the inode. These state fields are not all reset when an
inode is reused from the reclaimable state.

This can lead to unexpected behaviour of the new inode such as
speculative preallocation not being truncated away in the expected
manner for local files until the inode is subsequently truncated,
freed or cycles out of the cache. It can also lead to an inode being
considered to be a filestream inode or having been truncated when
that is not the case.

Rework the reinitialisation of the inode when it is recycled to
ensure that it is pristine before it is reused. While there, also
fix the resetting of state flags in the recycling error paths so the
inode does not become unreclaimable.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Alex Elder <aelder@sgi.com>

There's no reason not to support cache flushing on external log devices.
The only thing this really requires is flushing the data device first
both in fsync and log commits.  A side effect is that we also have to
remove the barrier write test during mount, which has been superflous
since the new FLUSH+FUA code anyway.  Also use the chance to flush the
RT subvolume write cache before the fsync commit, which is required
for correct semantics.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>

->mknod() should return negative on errors and PTR_ERR() gives
already negative value...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>

Tell the filesystem if we just updated timestamp (I_DIRTY_SYNC) or
anything else, so that the filesystem can track internally if it
needs to push out a transaction for fdatasync or not.

This is just the prototype change with no user for it yet.  I plan
to push large XFS changes for the next merge window, and getting
this trivial infrastructure in this window would help a lot to avoid
tree interdependencies.

Also remove incorrect comments that ->dirty_inode can't block.  That
has been changed a long time ago, and many implementations rely on it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

* 'for-linus' of git://oss.sgi.com/xfs/xfs:
  xfs: correctly decrement the extent buffer index in xfs_bmap_del_extent
  xfs: check for valid indices in xfs_iext_get_ext and xfs_iext_idx_to_irec
  xfs: fix up asserts in xfs_iflush_fork
  xfs: do not do pointer arithmetic on extent records
  xfs: do not use unchecked extent indices in xfs_bunmapi
  xfs: do not use unchecked extent indices in xfs_bmapi
  xfs: do not use unchecked extent indices in xfs_bmap_add_extent_*
  xfs: remove if_lastex
  xfs: remove the unused XFS_BMAPI_RSVBLOCKS flag
  xfs: do not discard alloc btree blocks
  xfs: add online discard support

The code in xfs_bmap_del_extent does not correctly decrement the
extent buffer index when deleting a whole extent.  Most of the time
this gets caught by checks in xfs_bmapi that work around it and
decrement it manually and thus wasn't noticed so far.

Based on an earlier patch from Lachlan McIlroy.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Lachlan McIlroy <lmcilroy@redhat.com>
Signed-off-by: Alex Elder <aelder@sgi.com>

Based on an earlier patch from Lachlan McIlroy.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Lachlan McIlroy <lmcilroy@redhat.com>
Signed-off-by: Alex Elder <aelder@sgi.com>