When I test this case on xfs, it may fail as below:
--------------------------------------------
=== link SCRATCH_MNT/A/foo SCRATCH_MNT/bar with fsync SCRATCH_MNT/A ===
+umount: /mnt/xfstests/scratch: target is busy.
+ (In some cases useful info about processes that use
+ the device is found by lsof(8) or fuser(1))
---------------------------------------------
It fails because somethings is still using the fs when we call sync and then
try to unmount it. We can simply remove sync as the unmount is supposed to
persist the file/directory removals.
Signed-off-by: Yang Xu <xuyang2018.jy@cn.fujitsu.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
fsx has a closeopen option where it will close and then re-open the
file. This is handy, but what is really more useful is to drop the file
from cache completely, so add a drop_caches into this operation so that
the file is read back completely from disk to be really evil.
[Eryu: fix drop cache failure number to stay within 190]
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
I was running down a i_size problem and was missing the failure until
the next iteration of fsx operations because we do the file size check
_before_ the closeopen operation. Move it after the closeopen operation
so we can catch problems where the file gets messed up on disk.
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Tests need the exact operations to reproduce some issues by
--replay-ops so skip tests rather than one operation if a required
operation/flag in tests is not supported.
Signed-off-by: Xiao Yang <yangx.jy@cn.fujitsu.com>
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Also test that d_ino of readdir entries and i_ino from /proc/locks are
consistent with st_ino and that inode numbers persist after rename to
new parent, drop caches and mount cycle.
With nested xino configuration, directory st_ino is not persistent and
its st_ino/d_ino/i_ino values are not consistent, so test only non-dir
in this test.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Also test that d_ino of readdir entries and i_ino from /proc/locks are
consistent with st_ino and that inode numbers persist after rename to
new parent, drop caches and mount cycle.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
This is a variant of overlay file handles test for an overlayfs that
is nested over another lower overlayfs on non-samefs.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
This is a variant of overlay file handles test for an overlayfs that
is nested over another lower overlayfs on the same fs.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
The test is merged under index 202 but the golden output contains
201. This makes the test always fail. Correct the number.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Check that files from middle layer on same fs as upper layer
are not allowed to export the real inode st_dev;st_ino.
This is a regression test for kernel commit:
9c6d8f13e9da ("ovl: fix corner case of non-unique st_dev;st_ino")
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Compilation failed on Fedora 20 because stdbool.h is not included in
xfs/platform_defs-x86_64.h or xfs/linux.h on Fedora 20.
Also, yang xiao fixed similar problem(commit 234f51ebbd) for fsx.c in
2016.2, but after that, fsstress.c started to use bool variable without
including stdbool.h file. It may fail on old linux distributions, so
fix it.
Signed-off-by: Yang Xu <xuyang2018.jy@cn.fujitsu.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Test removal of a subvolume via rmdir after it has been renamed into a
snapshot of the volume that originally contained the subvolume
reference.
This currently fails on btrfs but is fixed by the patch with the title
"btrfs: fix invalid removal of root ref"
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
On 32-bit systems, the offsets are 'unsigned long' (32-bit) which means
that we must cast the explicitly to unsigned long long before feeding
them to llabs. Without the type conversion we fail to sign-extend the
llabs parameter, try to make a copy/clone/dedupe call with overlapping
ranges, and fsx aborts and the test fails.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
IMHO, if kernel doesn't supprt realtime, we should skip test.
So add it. Also, when we use _scratch_mkfs on xfs, we will get
the following error:
mkfs failed with extra mkfs options added to "-bsize=4096" by test 590 **
This failure occurs because we have used "export XFS_MKFS_OPTIONS=
${XFS_MKFS_OPTIONS:=-bsize=4096}" in common/config, we don't need to
set it again in extra_options, so remove it.
Signed-off-by: Yang Xu <xuyang2018.jy@cn.fujitsu.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
The ability to use a mounted device node as the primary argument
to xfs_growfs was added back in with:
7e8275f8 xfs_growfs: allow mounted device node as argument
because it was an undocumented behavior that some userspace depended on.
This test exercises that functionality.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Zorro Lang <zlang@redhat.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
For historical reasons having to do with Solaris ACL behavior, the Linux
client treats an ACL like the one used as an example here as equivalent
to a mode, causing listxattr to report that no ACL is set on the file.
(See the comment at the top of fs/nfs_common/nfsacl.c in the kernel
source for details, and the "bogus ACL_MASK entry" comment in the same
source file.) This causes a spurious generic/529 failure on NFS.
As far as I can tell any ACL should trigger the original XFS problem.
So, modify it so as not to hit this odd NFS corner case.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
In my testing on 1GB zram devices btrfs/139 usually fails with
ENOSPC.
Add a requirement for 2GB scratch devices (empirically measured).
Signed-off-by: Johannes Thumshirn <jth@kernel.org>
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
[BUG]
With btrfs-progs v5.4, btrfs/140 and btrfs/141 will fail.
[CAUSE]
Both tests are testing re-silvering of RAID1, thus they need to corrupt
on-disk data.
This requires to do manual logical -> physical bytes mapping in the test
case.
However the test case itself uses too many hard coded helper to grab
physical offset, which will change with mkfs.btrfs.
[FIX]
Use more flex helper, to get both devid and physical for such
corruption.
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
When we're reloading the error-target dm table, always resume the device
even if the reload fails because (a) we shouldn't leave dm state for the
callers to clean up and (b) the caller don't clean up the state which
just leads to the scratch filesystem hanging on the suspended dm error
device.
Resume the dm device when we're cleaning up so that cleaning up the
scratch filesystem won't hang.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Verify ciphertext for v2 encryption policies that use the IV_INO_LBLK_64
flag and use AES-256-XTS to encrypt file contents and AES-256-CTS-CBC to
encrypt file names.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>
Update _verify_ciphertext_for_encryption_policy() to support encryption
policies with the IV_INO_LBLK_64 flag set.
This flag modifies the encryption to include the inode number in the IVs
and to use a key derived from the tuple [master_key, fs_uuid, mode_num].
Since the file nonce is *not* included in this key derivation, multiple
files can use the same key.
This flag is supported by v2 encryption policies only -- not by v1.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Yang Xu
Josef Bacik
Xiao Yang
Amir Goldstein
Nikolay Borisov
Qu Wenruo
Darrick J. Wong
Eric Sandeen
J. Bruce Fields
Johannes Thumshirn
Eric Biggers