Darrick J. Wong [Wed, 1 Nov 2017 21:47:48 +0000 (14:47 -0700)]
xfs/013: don't fail because cp ran out of space
Don't fail xfs/013 just because cp -Rl runs out of space to allocate
inodes and sprays the ENOSPC messages into the golden output. We
want to stress the finobt by using cp to push us near ENOSPC
conditions, so it's fine to let cp run out of space.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Wed, 1 Nov 2017 21:47:11 +0000 (14:47 -0700)]
generic/459: explicitly require thin_check
The lvm command can invoke the thin pool utilities as part of
managing a thin volume. It'll fail if the thin provisioning
utilities are not installed, so we need to check for its presence
before running a test.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Justin Maggard [Mon, 30 Oct 2017 22:33:48 +0000 (15:33 -0700)]
btrfs: test for qgroup reservation leaks with prealloc
This test case writes into pre-allocated space, then tries to
fallocate some more within the defined quota limit. Currently
(4.14-rc7) this fails with EDQUOT due to quota reservation leakage
when writing into pre- allocated space.
A possible fix has been sent to the ML as "btrfs: Fix quota
reservation leak on preallocated files"
Justin Maggard [Mon, 30 Oct 2017 22:32:45 +0000 (15:32 -0700)]
btrfs: test if receive with qgroups corrupts metadata
This test case does some concurrent send/receives with qgroups
enabled. Currently (4.14-rc7) this usually results in btrfs check
errors, and often also results in a WARN_ON in
record_root_in_trans().
Bisecting points to 6426c7ad697d (btrfs: qgroup: Fix qgroup
accounting when creating snapshot) as the culprit.
Amir Goldstein [Mon, 30 Oct 2017 12:06:56 +0000 (14:06 +0200)]
overlay: test creating lower hardlinks for copied up files
kernel v4.13 introduced the index=on feature for not breaking
hardlinks on copy-up. With the index feature enabled a regression
was introduced - lower files that are hardlined while overlay is
offline can result in lookup error after overlay in mounted.
The regression was fixed by upstream commit 6eaf011144af ovl: fix EIO from lookup of non-indexed upper
that was merged to v4.14-rc7.
This test verifies that behavior is sane after creating lower
hardlinks for copied up files while overlayfs is offline.
Darrick J. Wong [Sat, 28 Oct 2017 17:08:56 +0000 (10:08 -0700)]
common/rc: force mkfs for xfs in _mkfs_dev
In generic/459, if the DISCARD of the thin device somehow fails with
this message:
device-mapper: thin: Data device (dm-1) discard unsupported: Disabling discard passdown.
Then we can end up with arbitrary gunk in the thin device. This
causes mkfs to fail because it's afraid to format the device. Don't
be afraid, just zap it. FWIW mkfs.xfs thinks that the thinp device
has an xfs external log because sometimes the thinp device just
happen to be backed by the log of the previous test's scratch fs.
Fix this by making the _mkfs_dev helper always format the device,
per Eryu Guan's suggestion.
[eguan: fix subject to reflect the fix is in common/rc]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Fri, 27 Oct 2017 20:24:07 +0000 (13:24 -0700)]
xfs/333: fix errors with new inode pointer verifiers
Fix test failures with new inode pointer verifiers... and also make
sure that the running xfs actually supports realtime rmap. (This
should stop the current crop of weird test failures since nobody has
rtrmap yet anyway...)
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Fri, 27 Oct 2017 20:25:50 +0000 (13:25 -0700)]
common/xfs: remove inode-paths cruft
Remove the inode-paths check from _check_xfs_test_fs because we don't
support inode paths, xfsprogs doesn't have a xfs_{check,repair}_ipaths
tool, and it's broken anyway because we ignore _check_xfs_filesystem
(which tells whether or not the filesystem is even still mounted).
[eguan: also fix the return value of _check_xfs_test_fs]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Fri, 27 Oct 2017 20:23:08 +0000 (13:23 -0700)]
fstests: add module reloading helpers
Add some helper functions to require that we can reload a given
module, and add a helper to actually do that. Refactor the existing
users to use the generics.
We need to hoist completely the behaviors of the old btrfs module
helper because we need to confirm before starting the test that we
actually can remove the module.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Fri, 27 Oct 2017 20:21:46 +0000 (13:21 -0700)]
common/xfs: refactor xfs_scrub presence testing
Move all the requirements checking for xfs_scrub into a helper function.
Make sure the helper properly detects the presence of the scrub ioctl
and situations where we can't run scrub (e.g. norecovery).
Refactor the existing three xfs_scrub call sites to use the helper to
check if it's appropriate to run scrub.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Fri, 27 Oct 2017 00:43:12 +0000 (17:43 -0700)]
common/fuzzy: online re-scrub should not preen
When we're doing the second online scrub (to figure out if the
repair did any good) we shouldn't let that second scrub preen the
filesystem in any way. If scrub finds things it can't/won't preen
that turns into a nonzero return code which gets reported
(incorrectly) as a failure.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Thu, 26 Oct 2017 05:51:58 +0000 (22:51 -0700)]
generic/45[34]: force UTF-8 codeset to enable utf-8 namer checks in xfs_scrub
The upcoming xfs_scrub tool will have the ability to warn about
suspicious UTF-8 normalization collisions. We want generic/45[34]
to be able to test this functionality, but to do that we have to
forcibly set the codeset to UTF-8 via LC_ALL since the rest of
xfstests only uses LC_ALL=C.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Thu, 26 Oct 2017 05:51:39 +0000 (22:51 -0700)]
quota: clear speculative delalloc when checking quota usage
Occasionally speculative preallocation kicks in when writing files
to a filesystem under test. These preallocations consume quota and
/usually/ aren't around after we drop_caches, but there's nothing to
guarantee that they actually have, so the quota reports will be
different before and after the fs remount, causing sporadic test
failures in generic/{23[123],270}.
We now have xfs_spaceman which can instruct XFS to forcibly remove
the speculative preallocations. This fixes the sporadic failures,
at least for XFS.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Robbie Ko [Thu, 26 Oct 2017 06:02:15 +0000 (14:02 +0800)]
btrfs/010: fix offset in test Btrfs delalloc accounting overflow
Found it when test btrfs delalloc accounting overflow, Fix offset
error. We will fill in the gaps between the created extents, then
outstanding extents will all be merged into 1.
Qu Wenruo [Mon, 23 Oct 2017 01:48:54 +0000 (09:48 +0800)]
btrfs/130: Remove from auto group
No agreement on how to fix it in the foreseeable future. So remove
it from auto group to prevent newbie tester from spending days
waiting it to finish.
Reported-by: Po-Hsu Lin <po-hsu.lin@canonical.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Chandan Rajendra [Thu, 19 Oct 2017 13:49:07 +0000 (19:19 +0530)]
overlay/038: Remove usage of _overlay_scratch_unmount
_overlay_scratch_unmount is not supposed to be invoked directly by
tests. Also, since unmounting scratch fs is optional, this commit
removes invocation of _overlay_scratch_unmount.
Brian Foster [Thu, 12 Oct 2017 11:36:27 +0000 (07:36 -0400)]
xfs: test for NULL xattr buffer problem during unlink
XFS had a bug that resulted in an unexpected NULL buffer during
unlink of an inode with a multi-level attr fork tree. This occurred
due to a stale reference to content in a released/reclaimed buffer.
Use the XFS buffer LRU reference count error injection tag to
recreate the conditions for the bug. Create a file with a
multi-level attr fork tree and then unlink it with buffer caching
disabled.
Commit f35c5e10c6ed ("xfs: reinit btree pointer on attr tree
inactivation walk") fixed the bug.
Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Chandan Rajendra [Thu, 12 Oct 2017 13:34:32 +0000 (19:04 +0530)]
overlay: Test consistent d_ino feature for non-samefs setup
This commit adds a test to verify consistent d_ino feature when
the overlayfs instance is composed of two different underlying
filesystem instances.
For example,
$ mount -t xfs /dev/loop0 /mnt/test
$ mount -t xfs /dev/loop1 /mnt/scratch
$ mkdir /mnt/scratch/upper
$ mkdir /mnt/scratch/work
$ mount -t overlay overlay -o lowerdir=/mnt/test \
-o upperdir=/mnt/scratch/upper \
-o workdir=/mnt/scratch/work /mnt/merge
The goal of this test is to verify that the inode numbers returned by
readdir(3) (i.e. dirent->d_ino) are consistent with inode numbers
returned by stat(2) (i.e. stat->st_ino) in all the below listed cases,
- Parent's (i.e. "..") d_ino must always be calculated because a
pure dir can be residing inside a merged dir.
- d_ino for "." must always be calculated because the present
directory can have a copy-up origin.
- Verify d_ino of '.' and '..' before and after dir becomes impure.
While at it also verify if trusted.overlay.impure xattr is
set/reset appropriately and invalidation of readdir cache.
- Verify copied up file's (inside a impure dir) d_ino.
- Verify invalidation of readdir cache.
- Verify d_ino values corresponding to "." and ".." entries of a
pure lower dir.
- Verify d_ino of ".." entry of a merged dir.
- Verify pure lower residing in dir which has another lower layer
Ilya Dryomov [Thu, 12 Oct 2017 14:54:03 +0000 (16:54 +0200)]
common: turn _filter_xfs_dmesg() into _filter_aiodio_dmesg()
With upstream commit 332391a9935d ("fs: Fix page cache inconsistency
when mixing buffered and AIO DIO"), ext4 (and probably other
non-iomap based filesystems) need a _check_dmesg() whitelist entry
for a new WARN in dio_complete() -- generic/095 and 247 trigger a
dio_complete() splat pretty reliably for me.
Add dio_complete() entry to _filter_xfs_dmesg(), rename it and use
it unconditionally in generic/095, 224, 247 and 446.
Ilya Dryomov [Thu, 12 Oct 2017 14:54:02 +0000 (16:54 +0200)]
common: move _filter_xfs_dmesg() to common/filter
Move this XFS-specific _filter_xfs_dmesg() to common/filter so all
tests could use it. It will be renamed & made more generic and used
by more tests in later patches.
Dave Chinner [Wed, 11 Oct 2017 23:15:44 +0000 (10:15 +1100)]
generic/166: speed up on slow disks
generic/166 is takes way too long to run on iscsi disks - over an
*hour* on flash based iscsi targets. In comparison, it takes 18s to
run on a pmem device.
The issue is that it takes 3-4s per file write cycle on slow disks,
and it does a thousand write cycles. The problem is taht reflink is
so much faster than the write cycle that it's doing many more
snapshots on slow disks than fast disks, and this slows it down even
more.
e.g. the pmem system that takes 18s to run does just under 1000
snapshots - roughly one per file write. 20 minutes into the iscsi
based test, it's only done ~300 write cycles but there are almost
10,000 snapshots been taken. IOWs, we're doing 30 snapshots a file
write, not ~1.
Fix this by rate limiting snapshots to at most 1 per whole file
write. This reduces the number of snapshots taken on fast devices by
~50% (runtime on pmem device went from 18s -> 8s) but reduced it to
1000 on slow devices and reduced runtime from 3671s to just 311s.
Signed-Off-By: Dave Chinner <dchinner@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Test generic/422 hardcodes in the output file how much space files it
creates are supposed to take up on disk. This doesn't work when
additional space is allocated for extended attributes for ACLs or
SELinux labels.
Instead, record the actual space used in generic/422.full, and only
check if the writeback changes the space used.
Zheng Liu [Fri, 18 Aug 2017 20:35:02 +0000 (15:35 -0500)]
generic: test race when checking i_size on direct i/o read
In this commit a new test case is added to test that i_size races
don't occur under dio reads/writes. We add a program in /src dir,
which has a writer to issue some append dio writes. Meanwhile it
has a reader in this test to do some dio reads. As we expect,
reader should read nothing or data with 'a'. But it might read some
data with '0'.
The bug can be reproduced by this test case [1].
1. http://patchwork.ozlabs.org/patch/311761/
This ostensibly tests commit: 9fe55eea7 Fix race when checking i_size on direct i/o read
Update by Eric Sandeen:
- update to recent xfstests
- update commit log
Update by Eryu Guan:
- add aio-dio support to the test and add 'aio' group
- add ability to test different alignments
- move test from src/ to src/aio-dio-regress/
- add .gitignore entry
- rebase against latest xfstests with various minor fixes & cleanups
- update commit log
Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Zheng Liu <wenqing.lz@taobao.com> Signed-off-by: Eric Sandeen <sandeen@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Eryu Guan [Tue, 10 Oct 2017 07:42:07 +0000 (15:42 +0800)]
generic: test race between block map change and writeback
Run delalloc writes & append writes & non-data-integrity syncs
concurrently to test the race between block map change vs writeback.
This is to cover an XFS bug that data could be written to wrong
block and delay allocated blocks are leaked because the block map
was changed due to the removal of speculative allocated eofblocks
when writeback is in progress.
And this test partially mimics what lustre-racer[1] test does, using
which this bug was first found.
Amir Goldstein [Wed, 27 Sep 2017 07:04:12 +0000 (10:04 +0300)]
overlay: fix _overlay_config_override of MOUNT_OPTIONS
The config variable OVERLAY_MOUNT_OPTIONS is used to configure
the overlay mount options when running ./check -overlay.
The config variable MOUNT_OPTIONS is used to configure the
mount options for base fs.
If config sets value of OVERLAY_MOUNT_OPTIONS and
does not set MOUNT_OPTIONS, the value of MOUNT_OPTIONS
may be leftover from previous _overlay_config_override, so
don't use that value for base fs mount.
A helper to mount with same options/mnt/dev of scratch mount, but
optionally with different lower/upper/work dirs.
use instead of _overlay_mount_dirs() in all tests where applicable.
Amir Goldstein [Wed, 27 Sep 2017 07:04:10 +0000 (10:04 +0300)]
overlay: use default overlay mount options _overlay_mount_dirs()
Tests that use _overlay_mount_dirs() should also use the
default overlay mount options.
Move mount options from overlay_mount() into _overlay_mount_dirs()
and use helper common_dev_mount_opts() to get options.
OVERLAY_MOUNT_OPTIONS is assigned to MOUNT_OPTIONS, so
there is no need to use OVERLAY_MOUNT_OPTIONS directly.
Theodore Ts'o [Sun, 8 Oct 2017 19:27:46 +0000 (15:27 -0400)]
ext4: skip project quota tests if the kernel does not support them
In _scratch_enable_pquota, use _notrun if the file system with project
quotas enable can't be mounted, since that indicates the kernel
doesn't support that feature.
This commit removes the redundant chown operation which was supposed to
cause the test file to be copied up. Also, _overlay_scratch_unmount() is
used to unmount the overlay filesystem rather than invoking $UMOUNT_PROG.
This can be used to trigger an assert in the current XFS code
because it can't handle the case where there are COW extents on a
file, but none at or below the range converted by the AIO completion
handler.
Note that it doesn't trigger the assert 100% but fairly reliably.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
==11670==WARNING: AddressSanitizer failed to allocate 0xffffffffffffffff bytes
==11670==AddressSanitizer's allocator is terminating the process instead of returning 0
...
#5 0x4bb230 in __interceptor_malloc /home/vak-local/3.9.1/release/final/llvm.src/projects/compiler-rt/lib/asan/asan_malloc_linux.cc:65:10
#6 0x7f97e6491405 in getcwd /build/glibc-6V9RKT/glibc-2.19/io/../sysdeps/unix/sysv/linux/getcwd.c:68
#7 0x454691 in getcwd /home/vak-local/3.9.1/release/final/llvm.src/projects/compiler-rt/lib/asan/../sanitizer_common/sanitizer_common_interceptors.inc:2822:15
#8 0x4f765d in doproc /.../ltp/fsstress.c:933:12
#9 0x4f5f54 in main /.../ltp/fsstress.c:581:5
Theodore Ts'o [Mon, 2 Oct 2017 03:39:32 +0000 (23:39 -0400)]
report: fix summary statistics in xUnit header
The xUnit XML DTD distinguishes between test failures and test errors,
where a test failure indicate that the test has explicitly indicated
that the code under test has behaved in an unexpected fashion, whereas
a test error indicates the test code itself has thrown an error or
there has been some other test implementation error.
Xfstest failures are correctly marked as xUnit failures, but in the
attributes of the testsuite XML element, the number of test failures
was incorrectly reported as the number of errors.
report: encode XML Character Entities in xUnit report
Since the xUnit report is an XML document, special XML characters such
as '<', '>', '&', etc. have to be encoded as "<", ">", etc.
Otherwise programs parsing something like this:
<testcase classname="xfstests.global" name="generic/450" time="0">
<skipped message="Only test on sector size < half of block size" />
</testcase>
Will get choke the unescaped '<' character in the skipped message.
nsexec gives an unaligned child stack address to clone() system
call sometimes. For making sure it's always aligned, use
"__attribute__((aligned))" extension of GCC (Thanks this suggestion
from Eric sandeen).
Signed-off-by: Zorro Lang <zlang@redhat.com> Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Fri, 29 Sep 2017 05:37:39 +0000 (22:37 -0700)]
xfs: test increased overlong directory extent discard threshold
As of 2007, metadump has an interesting "feature" where it discards
directory extents that are longer than 1000 (originally 20) blocks.
This ostensibly was to protect metadump from corrupt bmbt records, but
it also has the effect of omitting from the metadump valid long extents.
The end result is that we create incomplete metadumps, which is
exacerbated by the lack of warning unless -w is passed.
So now that we've fixed the default threshold to MAXEXTLEN, check that
the installed metadump no longer exhibits this behavior.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
common/xfs: add iomap_dio_complete() to the dmesg filter
Kernel commit 332391a9935d ("fs: Fix page cache inconsistency when mixing
buffered and AIO DIO") moved the WARN_ON_ONCE() into iomap_dio_complete(),
along with the page cache invalidation. Let's add iomap_dio_complete() to the
filter whitelist too, so this expected warning when mixing direct I/O with
buffered I/O won't fail tests.
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Amir Goldstein [Thu, 28 Sep 2017 16:40:41 +0000 (19:40 +0300)]
generic/447: not a quick test
It hogged my cpu for a good 300s.
Test was inheritted from generic/176, which is not quick.
Signed-off-by: Amir Goldstein <amir73il@gmail.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Eryu Guan <eguan@redhat.com>
Most shutdown tests only run on filesystems with metadata
journaling, so we lose coverage. Add a shutdown stress test that
doesn't check for consistency, so does not require journaling. This
is a modified version of generic/051, with some extras trimmed and
fs checking removed.
Richard Wareing [Fri, 22 Sep 2017 02:34:46 +0000 (19:34 -0700)]
xfs: Add test for CVE-2017-14340
Verify kernel doesn't panic when user attempts to set realtime flags
on non-realtime FS, using kernel compiled with CONFIG_XFS_RT.
Unpatched kernels will panic during this test. Kernels not compiled
with CONFIG_XFS_RT should pass test.
This bug was fixed via commit b31ff3cdf540 ("xfs:
XFS_IS_REALTIME_INODE() should be false if no rt device present") on
the main kernel tree.
[eguan: don't assume fixed position when grepping 't' and add some
comments about why we do this, also remove testfile after test]
Signed-off-by: Richard Wareing <rwareing@fb.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Amir Goldstein [Tue, 19 Sep 2017 11:39:44 +0000 (14:39 +0300)]
overlay/031: fix test with inodes index enabled
When overlayfs is configured with CONFIG_OVERLAY_FS_INDEX=y,
workdir from previous overlay mount cannot be reused in a new
overlay mount that uses a different upper dir.
Fix the test to use a different workdir when mounting with a
different upper dir.
This change has no effect on older kernels and overlay
configured without CONFIG_OVERLAY_FS_INDEX.
Amir Goldstein [Tue, 19 Sep 2017 11:39:43 +0000 (14:39 +0300)]
overlay/014: correct comment relating to workdir reuse
Clarification: EBUSY is what you get when trying to use the same
upperdir/workdir with two different *concurent* overlayfs mounts.
The EBUSY case is independent of the inodes index feature.
This is not the case in this test, but rather the case of trying to
reuse the same workdir with different upper dirs on *subsequent*
overlayfs mounts.
chattr with [iap..] attributes open file for read-only and invoke
ioctl(). If we chattr a lower file in overlay, it will get the lower
file but not trigger copy-up, so ioctl() lead to modification of
that lower file incorrectly. Add this test for this case.
generic: reserve correct indirect blocks for delalloc write path
Test that XFS reserves reasonable indirect blocks for delalloc and
speculative allocation, and doesn't cause any fdblocks corruption.
This was inspired by an XFS but that too large 'indlen' was returned by
xfs_bmap_worst_indlen() which can't fit in a 17 bits value
(STARTBLOCKVALBITS is defined as 17), then leaked 1 << 17 blocks in
sb_fdblocks.
This was only seen on XFS with rmapbt feature enabled, but nothing
prevents the test from being a generic test.
Reviewed-by: "Darrick J. Wong" <darrick.wong@oracle.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Carlos Maiolino [Fri, 18 Aug 2017 08:51:56 +0000 (10:51 +0200)]
xfs: Test infinite loop while searching for a free inode slot
Tests the search algorithm for a free inode slot in a specific AG,
done in xfs_dialloc_ag_inobt().
When finobt is not used, and agi->freecount is not 0, XFS will scan
the AG inode tree looking for a free inode slot, but if
agi->freecount is corrupted, and there is no free slot at all, it
will end up in an infinite loop.
Carlos Maiolino [Thu, 13 Jul 2017 13:12:08 +0000 (15:12 +0200)]
generic: Test filesystem lockup on full overprovisioned dm-thin
With thin devices, it's possible to have a virtual device larger
than the physical device itself, and such situation can cause
problems to filesystems, once the filesystem 'believe' to have more
space than it actually has.
This can lead the filesystem to several weird behaviors. The one
tested here is filesystem lockup.
In case of XFS, it locks up when trying to writeback AIL metadata
back to the filesystem, but, once there is no physical space
available, XFS locks up and do not gracefuly handle this case.
Other filesystems usually are remounted as read-only, so they
already have this situation covered.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Tests were merged with high seq numbers to avoid conflicts with
other tests. Now renumber them to contiguous numbers, as all other
tests have been merged correctly. This is easier to do than
assigning the final seq numbers at commit time.
Amir Goldstein [Thu, 7 Sep 2017 10:03:13 +0000 (13:03 +0300)]
generic: regression test for xfs leftover CoW extent error
The following error are reported after running this test:
*** xfs_check output ***
leftover CoW extent (0/2147483736) len 1
block 0/2147483736 out of range
blocks 0/2147483736..2147483736 claimed by block 0/6
leftover CoW extent (0/2147483738) len 2
blocks 0/2147483738..2147483739 out of range
blocks 0/2147483738..2147483739 claimed by block 0/6
leftover CoW extent (0/2147483741) len 3
blocks 0/2147483741..2147483743 out of range
blocks 0/2147483741..2147483743 claimed by block 0/6
block 0/88 type unknown not expected
block 0/90 type unknown not expected
block 0/91 type unknown not expected
block 0/93 type unknown not expected
block 0/94 type unknown not expected
block 0/95 type unknown not expected
*** xfs_repair -n output ***
Phase 1 - find and verify superblock...
Phase 2 - using internal log
- zero log...
- scan filesystem freespace and inode maps...
leftover CoW extent (0/88) len 1
leftover CoW extent (0/90) len 2
leftover CoW extent (0/93) len 3
- found root inode chunk
This should be fixed by patch titled:
xfs: evict CoW fork extents when performing finsert/fcollapse
Amir Goldstein [Thu, 7 Sep 2017 10:03:10 +0000 (13:03 +0300)]
generic: crash consistency fsx test using dm-log-writes
Cherry-picked the test from commit 70d41e17164b
in Josef Bacik's fstests tree (https://github.com/josefbacik/fstests).
Quoting from Josef's commit message:
The test just runs some ops and exits, then finds all of the good buffers
in the directory we provided and:
- replays up to the mark given
- mounts the file system and compares the md5sum
- unmounts and fsck's to check for metadata integrity
dm-log-writes will pretend to do discard and the replay-log tool will
replay it properly depending on the underlying device, either by writing
0's or actually calling the discard ioctl, so I've enabled discard in the
test for maximum fun.
[Amir:]
- Removed unneeded _test_falloc_support dynamic FSX_OPTS
- Fold repetitions into for loops
- Added place holders for using constant random seeds
- Add pre umount checkpint
- Add test to new 'replay' group
- Address review comments by Eryu Guan
[eguan: fixed minor code style issues, remove extra newline at eof]
Cc: Josef Bacik <jbacik@fb.com> Signed-off-by: Amir Goldstein <amir73il@gmail.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Amir Goldstein [Thu, 7 Sep 2017 10:03:09 +0000 (13:03 +0300)]
fstests: add support for working with dm-log-writes target
Cherry-picked the relevant common bits from commit 70d41e17164b
in Josef Bacik's fstests tree (https://github.com/josefbacik/fstests).
Quoting from Josef's commit message:
This patch adds the supporting code for using the dm-log-writes
target. The dmlogwrites code is similar to the dmflakey code, it just
gives us functions to build and tear down a dm-log-writes target. We
add a new LOGWRITES_DEV variable to take in the device we will use as
the log and add checks for that.
[Amir:]
- Removed unneeded _test_falloc_support
- Moved _require_log_writes to dmlogwrites
- Document _require_log_writes
- Address review comments by Eryu Guan
Cc: Josef Bacik <jbacik@fb.com> Signed-off-by: Amir Goldstein <amir73il@gmail.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Amir Goldstein [Thu, 7 Sep 2017 10:03:06 +0000 (13:03 +0300)]
log-writes: add replay-log program to replay dm-log-writes target
Imported Josef Bacik's code from:
https://github.com/josefbacik/log-writes.git
Specialized program for replaying a write log that was recorded by
device mapper log-writes target. The tools is used to perform
crash consistency tests, allowing to run an arbitrary check tool
(fsck) at specified checkpoints in the write log.
[Amir:]
- Add project Makefile and SOURCE files
- Document the replay-log auxiliary program
- Address review comments by Eryu Guan
Cc: Josef Bacik <jbacik@fb.com> Signed-off-by: Amir Goldstein <amir73il@gmail.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Amir Goldstein [Thu, 7 Sep 2017 07:26:35 +0000 (10:26 +0300)]
fsx: add optional logid prefix to log messages
When redirecting the intermixed output of several fsx processes
to a single output file, it is usefull to prefix debug log messages
with a log id. Use fsx -j <logid> to define the log messages prefix.
Fix implementation of prt() function to avoid using a temp buffer
and convert some more printf() calls to use ptr() instead.
Amir Goldstein [Thu, 7 Sep 2017 07:26:34 +0000 (10:26 +0300)]
fsx: add support for integrity check with dm-log-writes target
Cherry-picked the relevant fsx bits from commit 70d41e17164b
in Josef Bacik's fstests tree (https://github.com/josefbacik/fstests).
Quoting from Josef's commit message:
I've rigged up fsx to have an integrity check mode. Basically it works
like it normally works, but when it fsync()'s it marks the log with a
unique mark and dumps it's buffer to a file with the mark in the filename.
I did this with a system() call simply because it was the fastest. I can
link the device-mapper libraries and do it programatically if that would
be preferred, but this works pretty well.
Signed-off-by: Josef Bacik <jbacik@fb.com>
[Amir:]
- Fix some exit codes
- Require -P dirpath for -i logdev
This commit adds a test to verify constant d_ino feature. The
following scenarios are checked,
- Parent's (i.e. "..") d_ino must always be calculated because a
pure dir can be residing inside a merged dir.
- d_ino for "." must always be calculated because the present
directory can have a copy-up origin.
- Verify d_ino of '.' and '..' before and after dir becomes impure.
While at it also verify if trusted.overlay.impure xattr is
set/reset appropriately and invalidation of readdir cache.
- Verify copied up file's (inside a impure dir) d_ino.
- Verify d_ino values corresponding to "." and ".." entries of a
pure lower dir.
- Verify d_ino of ".." entry of a merged dir.
- Verify pure lower residing in dir which has another lower layer
Several tests uses both _filter_test_dir and _filter_scratch
concatenated by pipe to filter $TEST_DIR and $SCRATCH_MNT. However,
this would fail if the shorter string is a substring of the other
(like "/mnt" and "/mnt2").
This patch introduces new common filter function to safely call both
_filter_test_dir and _filter_scratch, and update tests and functions
to use this new function.
I checked this with btrfs/029, generic/409,410,411, and
generic/381,383, xfs/106,108 (which calls _filter_quota). Thanks
Eryu for advice.
[eguan: folded 2nd patch into 1st patch and update commit log a bit]
Darrick J. Wong [Wed, 30 Aug 2017 22:52:34 +0000 (15:52 -0700)]
generic: try various unicode normalization games
Linux filesystems generally treat filenames and extended attribute
keys as a bag of bytes, which means that there can be unique
sequences of bytes that render the same on most modern GUIs. So,
let's rig up a test to see if it's really true that we can create
filenames and xattrs that look the same but point to different
files. xfs_scrub will warn about these kinds of situations, though
they're not technically fs "corruption".
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Thu, 31 Aug 2017 14:47:41 +0000 (07:47 -0700)]
xfs: test rmapbt updates are correct with insert/collapse range
Make sure that we update the rmapbt correctly when we collapse-range
a file and the extents on both sides of the hole can be merged. We
can construct this pretty trivially with insert-range and write, so
test that too.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Wed, 30 Aug 2017 22:31:10 +0000 (15:31 -0700)]
generic/173: don't dump core when mwrite fails
In generic/173, we try to force a CoW to a mmap'd region to fail if
there's no space to actually stage the CoW operation. That failure
comes in the form of a SIGBUS to xfs_io. If the tester just happens
to have a nonzero coresize ulimit set, a core dump is generated and
the test is marked as having failed, even though the dump generation
is exactly the correct behavior.
Therefore, set the coresize ulimit to zero while calling
_mwrite_byte.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Zorro Lang [Thu, 31 Aug 2017 05:59:43 +0000 (13:59 +0800)]
fsstress: fallback to block size for min dio size
XFS_IOC_DIOINFO is only used for XFS, but fsstress use it to get
DIO aligned size. If XFS_IOC_DIOINFO returns error, then stop
doing any DIO related test (dread/dwrite/aread/awrite etc). That
means we never do DIO related test on other filesystems by fsstress.
The real minimal dio size is really not so important for DIO test
in fsstress. The multiple of real min dio size is fine too. I think
the stat.st_blksize get from stat() system call can be used to be
a fake minimal dio size, if XFS_IOC_DIOINFO fails (not supported).
Note that the equation about d_maxiosz is copied from kernel
XFS_IOC_DIOINFO ioctl source code:
Ross Zwisler [Thu, 31 Aug 2017 04:09:10 +0000 (22:09 -0600)]
generic: add test for executables on read-only DAX mounts
This adds a regression test for the following kernel patch:
commit 42d4a99b09cb ("ext4: fix fault handling when mounted with -o
dax,ro")
The above patch fixes an issue with ext4 where executables cannot be
run on read-only filesystems mounted with the DAX option.
This issue does not appear to be present in ext2 or XFS, as they
both pass the test. I've also confirmed outside of the test that
they are both indeed able to execute binaries on read-only DAX
mounts.
Thanks to Randy Dodgen for the bug report and reproduction steps.
Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com> Cc: Randy Dodgen <rdodgen@gmail.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Theodore Ts'o <tytso@mit.edu> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Wed, 30 Aug 2017 04:40:56 +0000 (21:40 -0700)]
xfs/013: exit cleaner thread if fsstress dies
In this test, the cleaner thread deletes the directory trees created
by fsstress in order to exercise the free inode btree code.
However, if fsstress dies, the cleaner can end up waiting forever
for a directory that will never be created, which hangs up the test
run. Therefore, abort if fsstress has ended.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Eryu Guan <eguan@redhat.com>
projects / xfstests-dev.git / log