Regression for btrfs send when an inode only has extended references
associated to it (no regular references present). This used to cause
incorrect access to a b+tree leaf, where an extended reference item
was accessed as if it were a regular reference item, causing unexpected
and unpredictable behaviour such as producing a random/weird path string
or a crash.
This issue is fixed by the following linux kernel btrfs patch:
Btrfs: send, fix incorrect ref access when using extrefs
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Kill any lingering fsstress processes and wait properly should we abort
the test. This prevents the workload from inadvertently affecting
subsequent tests.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
This patch adds a regression test to verify btrfs can not
reuse inode id until we have committed transaction. Which was
addressed by the following kernel patch:
Btrfs: fix inode cache vs tree log
Signed-off-by: Wang Shilong <wangsl.fnst@cn.fujitsu.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Create a stress test for the free inode btree. Allocate a set of inodes
sequentually and run a hard link clone and random replacement algorithm
across the set. Background removal of the oldest directories creates a
sparse set of free inodes over time. Run an fsstress workload
concurrently to exercise the fs.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
The finobt creates a duplicate subset of inode allocation metadata from
the inobt. xfs_repair should detect and repair inconsistencies in the
finobt that could be caused by bugs or corruption. This test uses xfs_db
to cause targeted corruptions in the finobt and verify repair detects
and corrects the filesystem.
In particular, the test corrupts individual finobt records to cause
inconsistency between the inode allocation count fields as well as
causing the finobt to contain a record with no free inodes.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
xfs/030 nukes various on-disk data structures to test for repair. This
can result in extra output when testing finobt enabled filesystems. For
example, xfs_repair detects an invalid free inode btree root block when
the agi is zeroed.
Filter this output directly in xfs/030 such that the test passes for
finobt and non-finobt filesystems.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
If TEST_DEV or SCRATCH_DEV is symlink(mostly a lvm lv), a simple
basename is not enough, symlink should be followed.
This task is common enough, so introduce new helper functions and
replace all readlink calls in
ext4/305
generic/009
generic/019
generic/285
generic/312
Signed-off-by: Eryu Guan <eguan@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Number of helpers for checking xfs_io functionality is slowly
growing. But it's as easy to simply use _require_xfs_io_command()
directly and just specify the command we want to check. It will also
avoid the need to create helper every time we need to check a new
command in xfs_io.
Remove all the helpers and use _require_xfs_io_command() in the
tests.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Having just removed the largeacl test from the shared ACL test,
reintroduce the same test as an generic test so that we can
handle the different limits in supported ACL count appropriately
across different filesystems and different configurations within
filesystem types.
Filesystems have to add support to _acl_get_max to run
this test - the default behaviour right now is to throw a
notrun error like this:
generic/026 14s ... [not run] ext4 does not define maximum ACL count
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Different versions of XFS support different numbers of ACLs on disk.
Remove that subtest from this shared test to prevent it form causing
spurious failures on filesystems that support more than 25 ACLs.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
There is no mainline kernel support for DMAPI in XFS, and so every
time the xfstests auto group is run it throws a large number of
"[not run] Assuming DMAPI modules are not loaded". I've noted that
people are excluding the dmapi group to avoid this, so rather than
inflict pain on everyone, make hose few that need to do dmapi
testing include it specifically.
IOWs, remove the DMAPI tests from the auto group.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
A couple of tests leave behind large files or directory structures
when they complete, which leads to small TEST_DEVs running out of
space during other tests. Make those space hogs clean up after
themselves so that random tests don't fail with ENOSPC errors.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
On small block size filesystems, the reserve pool size is kept
constant at 4MB. filesystems with smaller blocks use comparitively
more blocks for indexing metadata (e.g. in the inode and extent
btrees) and so having a higher indirect block usage. Hence we need
to leave the reserve pool at 1024 block and not scale it for a
constant size.
This makes the test pass on a filesystem made with MKFS_OPTIONS="-b
size=1024 -m crc=1".
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
When running on a ramdisk, the fsstress background workload consumes
a GB of disk space every 5 seconds. This leads to the test failing
with ENOSPC because the test file cannot be created due otthe
background load cosuming it all. Hence don't run this test unless
the scratch device is large enough not to hit ENOSPC conditions.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Generic/009 fails when run on a file system that does not support byte range
zeroing. For example, an EOPNOTSUPP failure occurs when the test is run
on a pre-3.15 extent-mapped file system. The code in the test intended
to prevent this contains an apparent typo that results in a check for
fallocate() rather than zero range support.
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Ext4/001 fails when run on a file system that does not support byte range
zeroing. For example, an EOPNOTSUPP failure occurs when the test is run
on a pre-3.15 extent-mapped file system. The code in the test intended
to prevent this contains an apparent typo that results in a check for
fallocate() rather than zero range support.
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
The Q_XQUOTARM quotactl was not working properly, because
we weren't passing in proper flags. The xfs_fs_set_xstate()
ioctl handler used the same flags for Q_XQUOTAON/OFF as
well as Q_XQUOTARM, but Q_XQUOTAON/OFF look for
XFS_UQUOTA_ACCT, XFS_UQUOTA_ENFD, XFS_GQUOTA_ACCT etc,
i.e. quota type + state, while Q_XQUOTARM looks only for
the type of quota, i.e. XFS_DQ_USER, XFS_DQ_GROUP etc.
Unfortunately these flag spaces overlap a bit, so we
got semi-random results for Q_XQUOTARM; i.e. the value
for XFS_DQ_USER == XFS_UQUOTA_ACCT, etc. yeargh.
Anyway, here's a simple test that demonstrates it,
kernel patch to fix it will follow.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Just like FSSTRESS_AVOID, FSX_AVOID can be used to add
options at the end of the default fsx runs in each test.
i.e. FSX_AVOID="-H -z -C" will disable punch hole, zero range,
and collapse range calls in all tests which run fsx.
This should handle Ted's concerns about buggy ext4 fallocate
code without needing to add tunables to the kernel to reject
these operations during xfstests runs.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Files that consist of an inline extent, have the corresponding
data in the filesystem btree and not on a dedicated extent. For
such extents filefrag (fiemap) will report a physical location
of 0 for that extent and set the 'inline' flag.
The btrfs inspect-internal logical-resolve command will cause a
lookup in the extent tree for the extent address we give it as
an argument, which fails with errno ENOENT if it is 0.
This error didn't happen always, as the test uses fsstress to
generate a random filesystem, which needed to generate at least
one file that could be inlined (content less than 4018 bytes).
Example, taken from results/btrfs/004.full:
# filefrag -v /home/fdmanana/btrfs-tests/scratch_1/snap1/p0/de/d1b/dcb/fb1
Filesystem type is: 9123683e
File size of /home/fdmanana/btrfs-tests/scratch_1/snap1/p0/de/d1b/dcb/fb1 is 3860 (1 block of 4096 bytes)
ext: logical_offset: physical_offset: length: expected: flags:
0: 0.. 4095: 0.. 4095: 4096: not_aligned,inline,eof
1: 280.. 344: 35190.. 35254: 65: 1: eof
/home/fdmanana/btrfs-tests/scratch_1/snap1/p0/de/d1b/dcb/fb1: 2 extents found
after filter: 0#0#0 0#0#0
# stat -c %i /home/fdmanana/btrfs-tests/scratch_1/snap1/p0/de/d1b/dcb/fb1
403
# /home/fdmanana/git/hub/btrfs-progs/btrfs inspect-internal logical-resolve -P 0 /home/fdmanana/btrfs-tests/scratch_1
ioctl ret=-1, error: No such file or directory
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
If the file consists of a single block, then filefrag mentions
'1 block of ...', and the filter expected 'blocks of ...'.
Example:
$ echo qwerty > foobar
$ filefrag -v foobar
Filesystem type is: ef53
File size of foobar is 7 (1 block of 4096 bytes)
ext: logical_offset: physical_offset: length: expected: flags:
0: 0.. 0: 0.. 0: 1: unknown,delalloc,eof
foobar: 1 extent found
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
This test case verifies the btrfs properties feature, a new feature
introduced in the linux kernel version 3.14.
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
This test verifies that after an incremental btrfs send the
replicated file has the same exact hole and data structure as in
the origin filesystem. This didn't use to be the case before the
send stream version 2 - holes were sent as write operations of 0
valued bytes instead of punching holes with the fallocate system
call, and pre-allocated extents were sent as well as write
operations of 0 valued bytes instead of intructions for the
receiver to use the fallocate system call.
It also checks that prealloc extents that lie beyond the file's
size are replicated by an incremental send.
Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Didn't update a patch correctly when renumbering it. This time
on a test that doesn't run on XFS yet, so it avoided smoke tests...
Signed-off-by: Dave Chinner <david@fromorbit.com>
The failure message goes to stderr, so we need to redirect stderr to
stdout before running sed.
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Generic/311 fails when run on a test filesystem that does not
support fallocate(). Its I/O load is produced by fsync-tester,
which uses fallocate() system calls to allocate blocks for some of
its test cases. This causes EOPNOTSUPP failures when the test is
run on indirect block-mapped ext4 filesystems.
Verify that the test filesystem supports fallocate() before
proceeding with the test, checking for block allocation
capabilities. Also, fix a minor error message typo.
Signed-off-by: Eric Whitney <enwlinux@gmail.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Filipe David Borba Manana
Brian Foster
Wang Shilong
Eryu Guan
Lukas Czerner
Dave Chinner
Eric Whitney
Eric Sandeen
Theodore Ts'o