commit 5fcd57505c002efc5823a7355e21f48dd02d5a51 upstream.
The only use of I_DIRTY_TIME_EXPIRE is to detect in
__writeback_single_inode() that inode got there because flush worker
decided it's time to writeback the dirty inode time stamps (either
because we are syncing or because of age). However we can detect this
directly in __writeback_single_inode() and there's no need for the
strange propagation with I_DIRTY_TIME_EXPIRE flag.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jan Kara <jack@suse.cz>
Cc: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit eb8409071a1d47e3593cfe077107ac46853182ab ]
This reverts commit 6ff646b2ceb0eec916101877f38da0b73e3a5b7f.
Your maintainer committed a major braino in the rmap code by adding the
attr fork, bmbt, and unwritten extent usage bits into rmap record key
comparisons. While XFS uses the usage bits *in the rmap records* for
cross-referencing metadata in xfs_scrub and xfs_repair, it only needs
the owner and offset information to distinguish between reverse mappings
of the same physical extent into the data fork of a file at multiple
offsets. The other bits are not important for key comparisons for index
lookups, and never have been.
Eric Sandeen reports that this causes regressions in generic/299, so
undo this patch before it does more damage.
Reported-by: Eric Sandeen <sandeen@sandeen.net>
Fixes: 6ff646b2ceb0 ("xfs: fix rmap key and record comparison functions")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6ff646b2ceb0eec916101877f38da0b73e3a5b7f ]
Keys for extent interval records in the reverse mapping btree are
supposed to be computed as follows:
(physical block, owner, fork, is_btree, is_unwritten, offset)
This provides users the ability to look up a reverse mapping from a bmbt
record -- start with the physical block; then if there are multiple
records for the same block, move on to the owner; then the inode fork
type; and so on to the file offset.
However, the key comparison functions incorrectly remove the
fork/btree/unwritten information that's encoded in the on-disk offset.
This means that lookup comparisons are only done with:
(physical block, owner, offset)
This means that queries can return incorrect results. On consistent
filesystems this hasn't been an issue because blocks are never shared
between forks or with bmbt blocks; and are never unwritten. However,
this bug means that online repair cannot always detect corruption in the
key information in internal rmapbt nodes.
Found by fuzzing keys[1].attrfork = ones on xfs/371.
Fixes: 4b8ed67794 ("xfs: add rmap btree operations")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ea8439899c0b15a176664df62aff928010fad276 ]
Pass the same oldext argument (which contains the existing rmapping's
unwritten state) to xfs_rmap_lookup_le_range at the start of
xfs_rmap_convert_shared. At this point in the code, flags is zero,
which means that we perform lookups using the wrong key.
Fixes: 3f165b334e ("xfs: convert unwritten status of reverse mappings for shared files")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2c334e12f957cd8c6bb66b4aa3f79848b7c33cab ]
Make sure that we actually initialize xefi_discard when we're scheduling
a deferred free of an AGFL block. This was (eventually) found by the
UBSAN while I was banging on realtime rmap problems, but it exists in
the upstream codebase. While we're at it, rearrange the structure to
reduce the struct size from 64 to 56 bytes.
Fixes: fcb762f5de ("xfs: add bmapi nodiscard flag")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8df0fa39bdd86ca81a8d706a6ed9d33cc65ca625 ]
When callers pass XFS_BMAPI_REMAP into xfs_bunmapi, they want the extent
to be unmapped from the given file fork without the extent being freed.
We do this for non-rt files, but we forgot to do this for realtime
files. So far this isn't a big deal since nobody makes a bunmapi call
to a rt file with the REMAP flag set, but don't leave a logic bomb.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d88850bd5516a77c6f727e8b6cefb64e0cc929c7 ]
Fix some off-by-one errors in xfs_rtalloc_query_range. The highest key
in the realtime bitmap is always one less than the number of rt extents,
which means that the key clamp at the start of the function is wrong.
The 4th argument to xfs_rtfind_forw is the highest rt extent that we
want to probe, which means that passing 1 less than the high key is
wrong. Finally, drop the rem variable that controls the loop because we
can compare the iteration point (rtstart) against the high key directly.
The sordid history of this function is that the original commit (fb3c3)
incorrectly passed (high_rec->ar_startblock - 1) as the 'limit' parameter
to xfs_rtfind_forw. This was wrong because the "high key" is supposed
to be the largest key for which the caller wants result rows, not the
key for the first row that could possibly be outside the range that the
caller wants to see.
A subsequent attempt (8ad56) to strengthen the parameter checking added
incorrect clamping of the parameters to the number of rt blocks in the
system (despite the bitmap functions all taking units of rt extents) to
avoid querying ranges past the end of rt bitmap file but failed to fix
the incorrect _rtfind_forw parameter. The original _rtfind_forw
parameter error then survived the conversion of the startblock and
blockcount fields to rt extents (a0e5c), and the most recent off-by-one
fix (a3a37) thought it was patching a problem when the end of the rt
volume is not in use, but none of these fixes actually solved the
original problem that the author was confused about the "limit" argument
to xfs_rtfind_forw.
Sadly, all four of these patches were written by this author and even
his own usage of this function and rt testing were inadequate to get
this fixed quickly.
Original-problem: fb3c3de2f6 ("xfs: add a couple of queries to iterate free extents in the rtbitmap")
Not-fixed-by: 8ad560d256 ("xfs: strengthen rtalloc query range checks")
Not-fixed-by: a0e5c435ba ("xfs: fix xfs_rtalloc_rec units")
Fixes: a3a374bf18 ("xfs: fix off-by-one error in xfs_rtalloc_query_range")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1cb5deb5bc095c070c09a4540c45f9c9ba24be43 ]
If we decide that a directory free block is corrupt, we must take care
not to leak a buffer pointer to the caller. After xfs_trans_brelse
returns, the buffer can be freed or reused, which means that we have to
set *bpp back to NULL.
Callers are supposed to notice the nonzero return value and not use the
buffer pointer, but we should code more defensively, even if all current
callers handle this situation correctly.
Fixes: de14c5f541 ("xfs: verify free block header fields")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b1de6fc7520fe12949c070af0e8c0e4044cd3420 ]
Omar Sandoval reported that a 4G fallocate on the realtime device causes
filesystem shutdowns due to a log reservation overflow that happens when
we log the rtbitmap updates. Factor rtbitmap/rtsummary updates into the
the tr_write and tr_itruncate log reservation calculation.
"The following reproducer results in a transaction log overrun warning
for me:
mkfs.xfs -f -r rtdev=/dev/vdc -d rtinherit=1 -m reflink=0 /dev/vdb
mount -o rtdev=/dev/vdc /dev/vdb /mnt
fallocate -l 4G /mnt/foo
Reported-by: Omar Sandoval <osandov@osandov.com>
Tested-by: Omar Sandoval <osandov@osandov.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0c4da70c83d41a8461fdf50a3f7b292ecb04e378 ]
Realtime files in XFS allocate extents in rextsize units. However, the
written/unwritten state of those extents is still tracked in blocksize
units. Therefore, a realtime file can be split up into written and
unwritten extents that are not necessarily aligned to the realtime
extent size. __xfs_bunmapi() has some logic to handle these various
corner cases. Consider how it handles the following case:
1. The last extent is unwritten.
2. The last extent is smaller than the realtime extent size.
3. startblock of the last extent is not aligned to the realtime extent
size, but startblock + blockcount is.
In this case, __xfs_bunmapi() calls xfs_bmap_add_extent_unwritten_real()
to set the second-to-last extent to unwritten. This should merge the
last and second-to-last extents, so __xfs_bunmapi() moves on to the
second-to-last extent.
However, if the size of the last and second-to-last extents combined is
greater than MAXEXTLEN, xfs_bmap_add_extent_unwritten_real() does not
merge the two extents. When that happens, __xfs_bunmapi() skips past the
last extent without unmapping it, thus leaking the space.
Fix it by only unwriting the minimum amount needed to align the last
extent to the realtime extent size, which is guaranteed to merge with
the last extent.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2a2b5932db67586bacc560cc065d62faece5b996 ]
The leaf format xattr addition helper xfs_attr3_leaf_add_work()
adjusts the block freemap in a couple places. The first update drops
the size of the freemap that the caller had already selected to
place the xattr name/value data. Before the function returns, it
also checks whether the entries array has encroached on a freemap
range by virtue of the new entry addition. This is necessary because
the entries array grows from the start of the block (but end of the
block header) towards the end of the block while the name/value data
grows from the end of the block in the opposite direction. If the
associated freemap is already empty, however, size is zero and the
subtraction underflows the field and causes corruption.
This is reproduced rarely by generic/070. The observed behavior is
that a smaller sized freemap is aligned to the end of the entries
list, several subsequent xattr additions land in larger freemaps and
the entries list expands into the smaller freemap until it is fully
consumed and then underflows. Note that it is not otherwise a
corruption for the entries array to consume an empty freemap because
the nameval list (i.e. the firstused pointer in the xattr header)
starts beyond the end of the corrupted freemap.
Update the freemap size modification to account for the fact that
the freemap entry can be empty and thus stale.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3f8a4f1d876d3e3e49e50b0396eaffcc4ba71b08 ]
[commit message is verbose for discussion purposes - will trim it
down later. Some questions about implementation details at the end.]
Zorro Lang recently ran a new test to stress single inode extent
counts now that they are no longer limited by memory allocation.
The test was simply:
# xfs_io -f -c "falloc 0 40t" /mnt/scratch/big-file
# ~/src/xfstests-dev/punch-alternating /mnt/scratch/big-file
This test uncovered a problem where the hole punching operation
appeared to finish with no error, but apparently only created 268M
extents instead of the 10 billion it was supposed to.
Further, trying to punch out extents that should have been present
resulted in success, but no change in the extent count. It looked
like a silent failure.
While running the test and observing the behaviour in real time,
I observed the extent coutn growing at ~2M extents/minute, and saw
this after about an hour:
# xfs_io -f -c "stat" /mnt/scratch/big-file |grep next ; \
> sleep 60 ; \
> xfs_io -f -c "stat" /mnt/scratch/big-file |grep next
fsxattr.nextents = 127657993
fsxattr.nextents = 129683339
#
And a few minutes later this:
# xfs_io -f -c "stat" /mnt/scratch/big-file |grep next
fsxattr.nextents = 4177861124
#
Ah, what? Where did that 4 billion extra extents suddenly come from?
Stop the workload, unmount, mount:
# xfs_io -f -c "stat" /mnt/scratch/big-file |grep next
fsxattr.nextents = 166044375
#
And it's back at the expected number. i.e. the extent count is
correct on disk, but it's screwed up in memory. I loaded up the
extent list, and immediately:
# xfs_io -f -c "stat" /mnt/scratch/big-file |grep next
fsxattr.nextents = 4192576215
#
It's bad again. So, where does that number come from?
xfs_fill_fsxattr():
if (ip->i_df.if_flags & XFS_IFEXTENTS)
fa->fsx_nextents = xfs_iext_count(&ip->i_df);
else
fa->fsx_nextents = ip->i_d.di_nextents;
And that's the behaviour I just saw in a nutshell. The on disk count
is correct, but once the tree is loaded into memory, it goes whacky.
Clearly there's something wrong with xfs_iext_count():
inline xfs_extnum_t xfs_iext_count(struct xfs_ifork *ifp)
{
return ifp->if_bytes / sizeof(struct xfs_iext_rec);
}
Simple enough, but 134M extents is 2**27, and that's right about
where things went wrong. A struct xfs_iext_rec is 16 bytes in size,
which means 2**27 * 2**4 = 2**31 and we're right on target for an
integer overflow. And, sure enough:
struct xfs_ifork {
int if_bytes; /* bytes in if_u1 */
....
Once we get 2**27 extents in a file, we overflow if_bytes and the
in-core extent count goes wrong. And when we reach 2**28 extents,
if_bytes wraps back to zero and things really start to go wrong
there. This is where the silent failure comes from - only the first
2**28 extents can be looked up directly due to the overflow, all the
extents above this index wrap back to somewhere in the first 2**28
extents. Hence with a regular pattern, trying to punch a hole in the
range that didn't have holes mapped to a hole in the first 2**28
extents and so "succeeded" without changing anything. Hence "silent
failure"...
Fix this by converting if_bytes to a int64_t and converting all the
index variables and size calculations to use int64_t types to avoid
overflows in future. Signed integers are still used to enable easy
detection of extent count underflows. This enables scalability of
extent counts to the limits of the on-disk format - MAXEXTNUM
(2**31) extents.
Current testing is at over 500M extents and still going:
fsxattr.nextents = 517310478
Reported-by: Zorro Lang <zlang@redhat.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 657f101930bc6c5b41bd7d6c22565c4302a80d33 ]
The inode chunk allocation transaction reserves inobt_maxlevels-1
blocks to accommodate a full split of the inode btree. A full split
requires an allocation for every existing level and a new root
block, which means inobt_maxlevels is the worst case block
requirement for a transaction that inserts to the inobt. This can
lead to a transaction block reservation overrun when tmpfile
creation allocates an inode chunk and expands the inobt to its
maximum depth. This problem has been observed in conjunction with
overlayfs, which makes frequent use of tmpfiles internally.
The existing reservation code goes back as far as the Linux git repo
history (v2.6.12). It was likely never observed as a problem because
the traditional file/directory creation transactions also include
worst case block reservation for directory modifications, which most
likely is able to make up for a single block deficiency in the inode
allocation portion of the calculation. tmpfile support is relatively
more recent (v3.15), less heavily used, and only includes the inode
allocation block reservation as tmpfiles aren't linked into the
directory tree on creation.
Fix up the inode alloc block reservation macro and a couple of the
block allocator minleft parameters that enforce an allocation to
leave enough free blocks in the AG for a full inobt split.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d0c20d38af135b2b4b90aa59df7878ef0c8fbef4 ]
The realtime flag only applies to the data fork, so don't use the
realtime block number checks on the attr fork of a realtime file.
Fixes: 30b0984d91 ("xfs: refactor bmap record validation")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f4020438fab05364018c91f7e02ebdd192085933 ]
The boundary test for the fixed-offset parts of xfs_attr_sf_entry in
xfs_attr_shortform_verify is off by one, because the variable array
at the end is defined as nameval[1] not nameval[].
Hence we need to subtract 1 from the calculation.
This can be shown by:
# touch file
# setfattr -n root.a file
and verifications will fail when it's written to disk.
This only matters for a last attribute which has a single-byte name
and no value, otherwise the combination of namelen & valuelen will
push endp further out and this test won't fail.
Fixes: 1e1bbd8e7e ("xfs: create structure verifier function for shortform xattrs")
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 96355d5a1f0ee6dcc182c37db4894ec0c29f1692 ]
In tracking down a problem in this patchset, I discovered we are
reclaiming dirty stale inodes. This wasn't discovered until inodes
were always attached to the cluster buffer and then the rcu callback
that freed inodes was assert failing because the inode still had an
active pointer to the cluster buffer after it had been reclaimed.
Debugging the issue indicated that this was a pre-existing issue
resulting from the way the inodes are handled in xfs_inactive_ifree.
When we free a cluster buffer from xfs_ifree_cluster, all the inodes
in cache are marked XFS_ISTALE. Those that are clean have nothing
else done to them and so eventually get cleaned up by background
reclaim. i.e. it is assumed we'll never dirty/relog an inode marked
XFS_ISTALE.
On journal commit dirty stale inodes as are handled by both
buffer and inode log items to run though xfs_istale_done() and
removed from the AIL (buffer log item commit) or the log item will
simply unpin it because the buffer log item will clean it. What happens
to any specific inode is entirely dependent on which log item wins
the commit race, but the result is the same - stale inodes are
clean, not attached to the cluster buffer, and not in the AIL. Hence
inode reclaim can just free these inodes without further care.
However, if the stale inode is relogged, it gets dirtied again and
relogged into the CIL. Most of the time this isn't an issue, because
relogging simply changes the inode's location in the current
checkpoint. Problems arise, however, when the CIL checkpoints
between two transactions in the xfs_inactive_ifree() deferops
processing. This results in the XFS_ISTALE inode being redirtied
and inserted into the CIL without any of the other stale cluster
buffer infrastructure being in place.
Hence on journal commit, it simply gets unpinned, so it remains
dirty in memory. Everything in inode writeback avoids XFS_ISTALE
inodes so it can't be written back, and it is not tracked in the AIL
so there's not even a trigger to attempt to clean the inode. Hence
the inode just sits dirty in memory until inode reclaim comes along,
sees that it is XFS_ISTALE, and goes to reclaim it. This reclaiming
of a dirty inode caused use after free, list corruptions and other
nasty issues later in this patchset.
Hence this patch addresses a violation of the "never log XFS_ISTALE
inodes" caused by the deferops processing rolling a transaction
and relogging a stale inode in xfs_inactive_free. It also adds a
bunch of asserts to catch this problem in debug kernels so that
we don't reintroduce this problem in future.
Reproducer for this issue was generic/558 on a v4 filesystem.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b2a8864728683443f34a9fd33a2b78b860934cc1 ]
The block reservation calculation for inode allocation is supposed
to consist of the blocks required for the inode chunk plus
(maxlevels-1) of the inode btree multiplied by the number of inode
btrees in the fs (2 when finobt is enabled, 1 otherwise).
Instead, the macro returns (ialloc_blocks + 2) due to a precedence
error in the calculation logic. This leads to block reservation
overruns via generic/531 on small block filesystems with finobt
enabled. Add braces to fix the calculation and reserve the
appropriate number of blocks.
Fixes: 9d43b180af ("xfs: update inode allocation/free transaction reservations for finobt")
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d0c7feaf87678371c2c09b3709400be416b2dc62 ]
We recently used fuzz(hydra) to test XFS and automatically generate
tmp.img(XFS v5 format, but some metadata is wrong)
xfs_repair information(just one AG):
agf_freeblks 0, counted 3224 in ag 0
agf_longest 536874136, counted 3224 in ag 0
sb_fdblocks 613, counted 3228
Test as follows:
mount tmp.img tmpdir
cp file1M tmpdir
sync
In 4.19-stable, sync will stuck, the reason is:
xfs_mountfs
xfs_check_summary_counts
if ((!xfs_sb_version_haslazysbcount(&mp->m_sb) ||
XFS_LAST_UNMOUNT_WAS_CLEAN(mp)) &&
!xfs_fs_has_sickness(mp, XFS_SICK_FS_COUNTERS))
return 0; -->just return, incore sb_fdblocks still be 613
xfs_initialize_perag_data
cp file1M tmpdir -->ok(write file to pagecache)
sync -->stuck(write pagecache to disk)
xfs_map_blocks
xfs_iomap_write_allocate
while (count_fsb != 0) {
nimaps = 0;
while (nimaps == 0) { --> endless loop
nimaps = 1;
xfs_bmapi_write(..., &nimaps) --> nimaps becomes 0 again
xfs_bmapi_write
xfs_bmap_alloc
xfs_bmap_btalloc
xfs_alloc_vextent
xfs_alloc_fix_freelist
xfs_alloc_space_available -->fail(agf_freeblks is 0)
In linux-next, sync not stuck, cause commit c2b3164320 ("xfs:
use the latest extent at writeback delalloc conversion time") remove
the above while, dmesg is as follows:
[ 55.250114] XFS (loop0): page discard on page ffffea0008bc7380, inode 0x1b0c, offset 0.
Users do not know why this page is discard, the better soultion is:
1. Like xfs_repair, make sure sb_fdblocks is equal to counted
(xfs_initialize_perag_data did this, who is not called at this mount)
2. Add agf verify, if fail, will tell users to repair
This patch use the second soultion.
Signed-off-by: Zheng Bin <zhengbin13@huawei.com>
Signed-off-by: Ren Xudong <renxudong1@huawei.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 953aa9d136f53e226448dbd801a905c28f8071bf upstream.
Don't allow passing arbitrary flags as they change behavior including
memory allocation that the call stack is not prepared for.
Fixes: ddbca70cc4 ("xfs: allocate xattr buffer on demand")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 69ffe5960df16938bccfe1b65382af0b3de51265 upstream.
Commit 5b094d6dac ("xfs: fix multi-AG deadlock in xfs_bunmapi") added
a check in __xfs_bunmapi() to stop early if we would touch multiple AGs
in the wrong order. However, this check isn't applicable for realtime
files. In most cases, it just makes us do unnecessary commits. However,
without the fix from the previous commit ("xfs: fix realtime file data
space leak"), if the last and second-to-last extents also happen to have
different "AG numbers", then the break actually causes __xfs_bunmapi()
to return without making any progress, which sends
xfs_itruncate_extents_flags() into an infinite loop.
Fixes: 5b094d6dac ("xfs: fix multi-AG deadlock in xfs_bunmapi")
Signed-off-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
64-bit time is a signed quantity in the kernel, so the bulkstat
structure should reflect that. Note that the structure size stays
the same and that we have not yet published userspace headers for this
new ioctl so there are no users to break.
Fixes: 7035f9724f ("xfs: introduce new v5 bulkstat structure")
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
The callers of xfs_bmap_local_to_extents_empty() log the inode
external to the function, yet this function is where the on-disk
format value is updated. Push the inode logging down into the
function itself to help prevent future mistakes.
Note that internal bmap callers track the inode logging flags
independently and thus may log the inode core twice due to this
change. This is harmless, so leave this code around for consistency
with the other attr fork conversion functions.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
xfs_attr_shortform_to_leaf() attempts to put the shortform fork back
together after a failed attempt to convert from shortform to leaf
format. While this code reallocates and copies back the shortform
attr fork data, it never resets the inode format field back to local
format. Further, now that the inode is properly logged after the
initial switch from local format, any error that triggers the
recovery code will eventually abort the transaction and shutdown the
fs. Therefore, remove the broken and unnecessary error handling
code.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
When a directory changes from shortform (sf) to block format, the sf
format is copied to a temporary buffer, the inode format is modified
and the updated format filled with the dentries from the temporary
buffer. If the inode format is modified and attempt to grow the
inode fails (due to I/O error, for example), it is possible to
return an error while leaving the directory in an inconsistent state
and with an otherwise clean transaction. This results in corruption
of the associated directory and leads to xfs_dabuf_map() errors as
subsequent lookups cannot accurately determine the format of the
directory. This problem is reproduced occasionally by generic/475.
The fundamental problem is that xfs_dir2_sf_to_block() changes the
on-disk inode format without logging the inode. The inode is
eventually logged by the bmapi layer in the common case, but error
checking introduces the possibility of failing the high level
request before this happens.
Update both of the dir2 and attr callers of
xfs_bmap_local_to_extents_empty() to log the inode core as
consistent with the bmap local to extent format change codepath.
This ensures that any subsequent errors after the format has changed
cause the transaction to abort.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The flags arg is always passed as zero, so remove it.
(xfs_buf_get_uncached takes flags to support XBF_NO_IOACCT for
the sb, but that should never be relevant for xfs_get_aghdr_buf)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
xfs_trans_log_buf takes first byte, last byte as args. In this
case, it should be from 0 to sizeof() - 1.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The collapse range operation can merge extents if two newly adjacent
extents are physically contiguous. If the extent count is reduced on
a btree format inode, a change to extent format might be necessary.
This format change currently occurs as a side effect of the file
size update after extents have been shifted for the collapse. This
codepath ultimately calls xfs_bunmapi(), which happens to check for
and execute the format conversion even if there were no blocks
removed from the mapping.
While this ultimately puts the inode into the correct state, the
fact the format conversion occurs in a separate transaction from the
change that called for it is a problem. If an extent shift
transaction commits and the filesystem happens to crash before the
format conversion, the inode fork is left in a corrupted state after
log recovery. The inode fork verifier fails and xfs_repair
ultimately nukes the inode. This problem was originally reproduced
by generic/388.
Similar to how the insert range extent split code handles extent to
btree conversion, update the collapse range extent merge code to
handle btree to extent format conversion in the same transaction
that merges the extents. This ensures that the inode fork format
remains consistent if the filesystem happens to crash in the middle
of a collapse range operation that changes the inode fork format.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Define a flags field for the AG geometry ioctl structure.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Add a helper that validates the startblock is valid. This checks for a
non-zero block on the main device, but skips that check for blocks on
the realtime device.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
When doing file lookups and checking for permissions, we end up in
xfs_get_acl() to see if there are any ACLs on the inode. This
requires and xattr lookup, and to do that we have to supply a buffer
large enough to hold an maximum sized xattr.
On workloads were we are accessing a wide range of cache cold files
under memory pressure (e.g. NFS fileservers) we end up spending a
lot of time allocating the buffer. The buffer is 64k in length, so
is a contiguous multi-page allocation, and if that then fails we
fall back to vmalloc(). Hence the allocation here is /expensive/
when we are looking up hundreds of thousands of files a second.
Initial numbers from a bpf trace show average time in xfs_get_acl()
is ~32us, with ~19us of that in the memory allocation. Note these
are average times, so there are going to be affected by the worst
case allocations more than the common fast case...
To avoid this, we could just do a "null" lookup to see if the ACL
xattr exists and then only do the allocation if it exists. This,
however, optimises the path for the "no ACL present" case at the
expense of the "acl present" case. i.e. we can halve the time in
xfs_get_acl() for the no acl case (i.e down to ~10-15us), but that
then increases the ACL case by 30% (i.e. up to 40-45us).
To solve this and speed up both cases, drive the xattr buffer
allocation into the attribute code once we know what the actual
xattr length is. For the no-xattr case, we avoid the allocation
completely, speeding up that case. For the common ACL case, we'll
end up with a fast heap allocation (because it'll be smaller than a
page), and only for the rarer "we have a remote xattr" will we have
a multi-page allocation occur. Hence the common ACL case will be
much faster, too.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The same code is used to copy do the attribute copying in three
different places. Consolidate them into a single function in
preparation from on-demand buffer allocation.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Because we repeat exactly the same code to get the remote attribute
value after both calls to xfs_attr3_leaf_getvalue() if it's a remote
attr. Just do it in xfs_attr3_leaf_getvalue() so the callers don't
have to care about it.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Shortform, leaf and remote value attr value retrieval return
different values for success. This makes it more complex to handle
actual errors xfs_attr_get() as some errors mean success and some
mean failure. Make the return values consistent for success and
failure consistent for all attribute formats.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
When a directory is growing rapidly, new blocks tend to get added at
the end of the directory. These end up at the end of the freespace
index, and when the directory gets large finding these new
freespaces gets expensive. The code does a linear search across the
frespace index from the first block in the directory to the last,
hence meaning the newly added space is the last index searched.
Instead, do a reverse order index search, starting from the last
block and index in the freespace index. This makes most lookups for
free space on rapidly growing directories O(1) instead of O(N), but
should not have any impact on random insert workloads because the
average search length is the same regardless of which end of the
array we start at.
The result is a major improvement in large directory grow rates:
create time(sec) / rate (files/s)
File count vanilla Prev commit Patched
10k 0.41 / 24.3k 0.42 / 23.8k 0.41 / 24.3k
20k 0.74 / 27.0k 0.76 / 26.3k 0.75 / 26.7k
100k 3.81 / 26.4k 3.47 / 28.8k 3.27 / 30.6k
200k 8.58 / 23.3k 7.19 / 27.8k 6.71 / 29.8k
1M 85.69 / 11.7k 48.53 / 20.6k 37.67 / 26.5k
2M 280.31 / 7.1k 130.14 / 15.3k 79.55 / 25.2k
10M 3913.26 / 2.5k 552.89 / 18.1k
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
When running a "create millions inodes in a directory" test
recently, I noticed we were spending a huge amount of time
converting freespace block headers from disk format to in-memory
format:
31.47% [kernel] [k] xfs_dir2_node_addname
17.86% [kernel] [k] xfs_dir3_free_hdr_from_disk
3.55% [kernel] [k] xfs_dir3_free_bests_p
We shouldn't be hitting the best free block scanning code so hard
when doing sequential directory creates, and it turns out there's
a highly suboptimal loop searching the the best free array in
the freespace block - it decodes the block header before checking
each entry inside a loop, instead of decoding the header once before
running the entry search loop.
This makes a massive difference to create rates. Profile now looks
like this:
13.15% [kernel] [k] xfs_dir2_node_addname
3.52% [kernel] [k] xfs_dir3_leaf_check_int
3.11% [kernel] [k] xfs_log_commit_cil
And the wall time/average file create rate differences are
just as stark:
create time(sec) / rate (files/s)
File count vanilla patched
10k 0.41 / 24.3k 0.42 / 23.8k
20k 0.74 / 27.0k 0.76 / 26.3k
100k 3.81 / 26.4k 3.47 / 28.8k
200k 8.58 / 23.3k 7.19 / 27.8k
1M 85.69 / 11.7k 48.53 / 20.6k
2M 280.31 / 7.1k 130.14 / 15.3k
The larger the directory, the bigger the performance improvement.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Simplify the logic in xfs_dir2_node_addname_int() by factoring out
the free block index lookup code that finds a block with enough free
space for the entry to be added. The code that is moved gets a major
cleanup at the same time, but there is no algorithm change here.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Factor out the code that adds a data block to a directory from
xfs_dir2_node_addname_int(). This makes the code flow cleaner and
more obvious and provides clear isolation of upcoming optimsations.
Signed-off-By: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
This gets rid of the need for a forward declaration of the static
function xfs_dir2_addname_int() and readies the code for factoring
of xfs_dir2_addname_int().
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Iterator functions already use 0 to signal "continue iterating", so get
rid of the #defines and just do it directly.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Use -ECANCELED to signal "stop iterating" instead of these magical
*_ITER_ABORT values, since it's duplicative.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
In xfs_rmap_irec_offset_unpack, we should always clear the contents of
rm_flags before we begin unpacking the encoded (ondisk) offset into the
incore rm_offset and incore rm_flags fields. Remove the open-coded
field zeroing as this encourages api misuse.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Remove the return value from the functions that schedule deferred bmap
operations since they never fail and do not return status.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Remove the return value from the functions that schedule deferred
refcount operations since they never fail and do not return status.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Remove the return value from the functions that schedule deferred rmap
operations since they never fail and do not return status.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
This function doesn't use the @state parameter, so get rid of it.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
In xfs_bmbt_diff_two_keys, we perform a signed int64_t subtraction with
two unsigned 64-bit quantities. If the second quantity is actually the
"maximum" key (all ones) as used in _query_all, the subtraction
effectively becomes addition of two positive numbers and the function
returns incorrect results. Fix this with explicit comparisons of the
unsigned values. Nobody needs this now, but the online repair patches
will need this to work properly.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
