common/rc: introduce minimal fs size check

_scratch_mkfs_sized() create a file system with specified size
limit. It can, however, too small for certain kind of devices. For
example, zoned btrfs requires at least 5 zones to make a file system.

This commit introduces MIN_FSSIZE, which specify the minimum size of the
possible file system. We can set this variable e.g. $ZONE_SIZE *
$MIN_ZONE_COUNT.

Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: use udevadm settle after pvremove

As with creation, we also need to use udevadm settle
when removing a pv, otherwise we can trip on races with
module removals for the block devices in use.

This reduces the amount of time in which a block device
module refcnt for test modules such as scsi_debug spends
outside of 0.

Races with the refcnt being greater than 0 means module
removal can fail causing false positives. This helps
ensure that the pv is really long gone. These issues
are tracked for scsi_debug [0] and later found to be a
generic issue regardless of filesystem with pvremove [1].

Using udevadm settle *helps*, it does not address all
possible races with the refcnt as noted in the generic
bug entry [1].

[0] https://bugzilla.kernel.org/show_bug.cgi?id=212337
[1] https://bugzilla.kernel.org/show_bug.cgi?id=214015

Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs/177: Handle the different fi resize output formats

Commit 78aa1d95dd99 ("btrfs-progs: fi resize: make output more
readable") added the device id of the resized fs along with a pretty
printed size. Create a new function to simplify the output message
using size in bytes.

Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

dmerror: export configuration so that subprograms don't have to reinit

Export the dmerror configuration variables so that subprograms don't
have to reinitialize the configuration in their own subprograms before
calling the helpers.  In the next patch (where we allow dmerror for log
and rt devices) it will become important to avoid these
reinitializations.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

dmflakey: support external log and realtime devices

Upgrade the dmflakey code to coordinate making external scratch log and
scratch realtime devices flakey along with the scratch device.  Note
that unlike SCRATCH_DEV, we save the old rt/log devices as separate
variables and replace SCRATCH_{RT,LOG}DEV so that helper functions
continue to work without modification.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

overlay/078: Fix cleanup with unpatched kernel

Darrick wrote: "...
I noticed a massive regression with this week's fstests,
probably because something didn't get cleaned up properly:
...
+rm: cannot remove '/opt/ovl-upper/testfile': Operation not permitted
...

and then the tests after it (e.g. generic/030) fail with:

+mount: /opt/ovl-mnt: mount(2) system call failed: Stale file handle.
..."

Link: https://lore.kernel.org/fstests/20210802230727.GC3601425@magnolia/
Reported-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs: quota state command should report upg grace times separately

Regression test for xfsprogs d8a94546 ("xfs_quota: state command
should report ugp grace times"). When give "-ugp" or "-a" options to
xfs_quota state command, it should report grace times for all three
types separately.

Signed-off-by: Zorro Lang <zlang@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs: verify the behavior of deleting non-existing device

There is a kernel regression for btrfs, that when passing
non-existing devid to "btrfs device remove" command, kernel will
crash due to NULL pointer dereference.

The test case is for such regression, it will:

- Create and mount an empty single-device btrfs
- Try to remove devid 3, which doesn't exist for above fs
- Make sure the command exits properly with expected error message

The kernel fix is titled "btrfs: fix NULL pointer dereference when
deleting device by invalid id".

Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/rc: only force nfs4.2 non-default SEEK_HOLE behaviour

Only NFSv4.2 supports non-defautl SEEK_HOLE behaviour. Thus default
SEEK_HOLE behaviour shall be allowed for NFSv4.0/4.1, or it will fail
generic/285, generic/448, generic/490 on NFSv4.0/4.1, complaining they
should support non-default SEEK_HOLE behaviour.

The *.full log is like:
File system supports the default behavior.
Default behavior is not allowed. Aborting.

Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/attr: Reduce MAX_ATTRS to leave some overhead for 64K blocksize

Test generic/020 fails for ext4 with 64K blocksize.
This adds changes in common/attr for MAX_ATTRS calculations for
ext2|ext3|ext4 along with comments explaining the calculations.

Suggested-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Ritesh Harjani <riteshh@linux.ibm.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/attr: Cleanup end of line whitespaces issues

This patch clears the end of line whitespace issues in this file.
Mostly since many kernel developers also keep this editor config to clear
any end of line whitespaces on file save.

Reviewed-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Ritesh Harjani <riteshh@linux.ibm.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

ext4/022: exclude this test for dax config on 64KB pagesize platform

This test case assumes blocksize to be 4KB and hence it fails
to mount with "-o dax" option on a 64kb pagesize platform (e.g. PPC64).
This leads to test case reported as failed with dax config on PPC64.

This patch exclude this test when pagesize is 64KB and for dax config.

Reviewed-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Ritesh Harjani <riteshh@linux.ibm.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

ext4/306: Add -b blocksize parameter too to avoid failure with DAX config

mkfs.ext4 by default uses 4K blocksize. On DAX config with a 64K
pagesize platform (PPC64), this will fail to mount since DAX requires bs
== ps.
Hence add the -b blocksize paramter in ext4/306.

Reviewed-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Ritesh Harjani <riteshh@linux.ibm.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

ext4/027: Correct the right code of block and inode bitmap

Observed occasional failure of this test sometimes say with 64k config
and small device size. Reason is we were grepping for wrong values for
inode and block bitmap.

Correct those values according to [1] to fix this test.

[1]: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/fs/ext4/fsmap.h#n53

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Ritesh Harjani <riteshh@linux.ibm.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

ext4/003: Fix this test on 64K platform for dax config

mkfs.ext4 by default uses 4K blocksize which doesn't mount when
testing with dax config and the test fails. This patch fixes it.

Signed-off-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/scsi_debug: use udevadm settle instead of sleeping

The variable UDEV_SETTLE_PROG is already defined and used for
lvm to either use `udevadm settle` in case it is available, and
if not, use 'sleep 1' otherwise (ancient distros or CONFIG_NET
is missing).

Use it on scsi_debug to replace the sleep calls sprinkled in
place after module removal. The correct thing to do is to
just use udevadm settle when available, and only fall back to
calling sleep when udevadm is not available or CONFIG_NET
is disabled.

Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/config: disable udevadm settle if CONFIG_NET is disabled

If CONFIG_NET is disabled kobject_uevent_net_broadcast() will be a no-op
and so no uevent are sent and so 'udevadm settle' won't really do
anything for you.

We check for /proc/net to see if CONFIG_NET was enabled.

Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/007: fix regressions on V4 filesystems

Following commit eae40404, I noticed the following regression when
running a V4 fstests run on an 5.13 kernel:

  --- /tmp/fstests/tests/xfs/007.out      2021-05-13 11:47:55.793859995 -0700
  +++ /var/tmp/fstests/xfs/007.out.bad    2021-07-28 09:23:42.856000000 -0700
  @@ -16,4 +16,4 @@
   *** umount
   *** Usage after quotarm ***
   core.nblocks = 0
  -core.nblocks = 0
  +core.nblocks = 1

The underlying cause of this problem is the fact that we now remount the
filesystem with no quota options because that will (soon) become the
only means to turn off quota accounting on XFS.  Because V4 filesystems
don't support simultaneous project and group quotas and play weird
remapping games with the incore superblock field, we actually have to
issue a remove command for the group quota file if we're trying to
truncate the project quota file on a V4 filesystem.

Due to stupid limitations in xfs_quota we actually have to issue a
separate 'remove' command.

Fixes: eae40404 ("xfs/007: unmount after disabling quota")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/530: skip test if user MKFS_OPTIONS screw up formatting

Skip this test if the user's MKFS_OPTIONS are not compatible with the
realtime parameters that we're injecting in order to test growfs bugs.
Because this test is trying to trigger a specific kernel bug, we need
mkfs to format a filesystem with very specific geometry parameters.

The first problem stems from the fact that the test performs a default
mkfs, computes a suitable realtime geometry from that filesystem, and
then formats a second time with an explicit blocksize option to mkfs.
If the original MKFS_OPTS contained a blocksize directive, the mkfs will
fail because the option was respecified.  The two blocksize options will
be the same, so we drop the explicit blocksize option.

However, this exposes a second problem: MKFS_OPTIONS might contain
options that are not compatible with any realtime filesystem.  If that
happens, _scratch_do_mkfs will "helpfully" drop MKFS_OPTIONS and try
again with only the options specified by the test.  This gets us a
filesystem with the given rt geometry, but it could be missing critical
parameters from MKFS_OPTIONS (like blocksize).  The test will then fail
to exercise the growfs bugfix, so the second part of the fix is to check
that the filesystem we're going to test actually has the geometry
parameters that we require.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

check: back off the OOM score adjustment to -500

Dave Chinner complained that fstests really shouldn't be running at
-1000 oom score adjustment because that makes it more "important" than
certain system daemons (e.g. journald, udev).  That's true, so increase
it to -500.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

check: don't leave the scratch filesystem mounted after _notrun

Unmount the scratch filesystem if a test decides to _notrun itself
because _try_wipe_scratch_devs will not be able to wipe the scratch
device prior to the next test run.  We don't want to let scratch state
from one test leak into subsequent tests if we can help it.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/106: fix golden output regression in quota off test

In commit 4c5df338, we reshuffled this test a bit in preparation to
disable quotaoff by rearranging the test to testing disabling of quota
by remounting the filesystem.  Unfortunately, extra blank lines were
added to the golden output, leading to test regressions.

The "extra" blank lines are a result of the "echo ; test_off";
test_off() itself doesn't print anything.  Make it print /something/ so
that we know what the test was trying to do when a particular line of
golden output appears, then fix the blank lines.

Fixes: 4c5df338 ("xfs/106: don't test disabling quota accounting")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

misc: tag all tests that examine crash recovery in a loop

Given all the recent problems that we've been finding with log recovery,
I think it would be useful to create a 'recoveryloop' group so that
developers have a convenient way to run every single test that rolls
around in a fs shutdown loop looking for subtle errors in recovery.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic: test small swapfile without page-aligned contiguous blocks

If a swapfile doesn't contain even a single page-aligned contiguous
range of blocks, it's an invalid swapfile, and might cause kernel
issue. This case covered commit 5808fecc5723 ("iomap: Fix negative
assignment to unsigned sis->pages in iomap_swapfile_activate").

Signed-off-by: Zorro Lang <zlang@redhat.com>
Reviewed-by: Ritesh Harjani <riteshh@linux.ibm.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic: test for file loss after mix of rename, fsync and inode eviction

Test that if we fsync a directory A, evict A's inode, move one file
from directory A to directory B, fsync some other inode that is not
directory A, B or any inode inside these two directories, and then
power fail, the file that was moved is not lost.

This currently fails on btrfs and is fixed by commit ecc64fab7d49
("btrfs: fix lost inode on log replay after mix of fsync, rename and
inode eviction")

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

tools: add a dm-logwrites replay tool

Trying to decipher a dm-logwrites replay failure from generic/482 is
difficult. The test destroys all the dm state and devices when the
test fails, leaving nothing behind for post-mortem analysis. This
script is intended to allow replay of the dm-logwrites device one
FUA at a time similar to generic/482, and is used in conjunction
with a modifed g/482 test that does not tear down the dm volumes
or state when the test fails.

This allows the developer to replay the logwrites up to just prior
to failure to examine just the differences between the last good
state and the first failure state, greatly reducing the complexity
of analysing failures.

Instructions for use are in the tools/dm-logwrite-replay script
itself.

Signed-off-by: Dave Chinner <dchinner@redhat.com>

overlay: add test for copy up of lower file attributes

Overlayfs copies up a subset of lower file attributes since kernel
commits:
173ff5c9ec37 ("ovl: consistent behavior for immutable/append-only inodes")
2e3f6e87c2b0 ("ovl: copy up sync/noatime fileattr flags")

This test verifies this functionality works correctly and that it
survives power failure and/or 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>

overlay/077: check for inconsistent d_ino/st_ino

After unlink of a directory entry, that entry may still apear in getdents
results of an already open directory fd, but it should return a d_ino
value that is consistent with the already observed st_ino of that entry.

Remove redundant break condition from gendents read loop.

For testing of inconsistent d_ino/st_ino we need to unlink an entry
whose st_ino is not that of the upper inode.

In the former merge dir setup we unlink all the files in the lower
dir after copyup, so they all use st_ino of the upper inode.

Let the unlinked file f100 reside in a lower path that is not being
unlinked so it will have the st_ino of the lower inode.

This is a regression test for kernel commit fcb7f373684d
("ovl: skip stale entries in merge dir cache iteration")

Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/152: avoid failure when quotaoff is not supported

Switch the test that removes the quota files to just disable enforcement
and then unmount the file system as disabling quota accounting is about
to go away.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs: use $XFS_QUOTA_PROG instead of hardcoding xfs_quota

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/305: don't turn quota accounting off

The test case tests just as much when just testing turning quota
enforcement off, so switch it to that.  This is in preparation for
removing support to turn quota accounting off.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/304: don't turn quota accounting off

The test case tests just as much when just testing turning quota
enforcement off, so switch it to that.  This is in preparation for
removing support to turn quota accounting off.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/220: avoid failure when disabling quota accounting is not supported

Doing a proper _requires for quotaoff support is rather hard, as we need
to test it on a specific file system.  Instead just use sed to remove
the warning and let the test case pass.  Eventually it should just be
removed entirely.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/106: don't test disabling quota accounting

Switch the test that removes the quota files to just disable
enforcement and then unmount the file system as disabling quota
accounting is about to go away.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/007: unmount after disabling quota

With the pending patches to remove support for disabling quota
accounting on a mounted file system we need to unmount the
file system first before removing the quota files.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/quota: allow removing quota options entirely in _qmount_option

Add support for dropping all quota related options instead of only
overriding them with new ones to _qmount_option.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs: source common/module when requiring module reload

Tests 163, 219, 225 and 242 require the btrfs module to be reloaded
in the middle of the test. If btrfs is compiled builtin, these tests
should be skipped, since we cannot unload it.

Today, this is what happens:

    QA output created by 242
    +./common/btrfs: line 405: _require_loadable_fs_module: command not found

Sourcing common/module in these tests fixes the issue by skipping the
test:
    btrfs/242       [not run] btrfs: must be a module.
    Ran: btrfs/242
    Not run: btrfs/242

Other tests liker btrfs/124 and btrfs/125 already source the same file
for the same reason, so follow the pattern.

Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/encrypt: accept '-' character in no-key names

Add the '-' character to the regex that generic/{419,429} use to match
no-key filenames.  This is needed to prevent these tests from failing
after the kernel is changed to use a more standard variant of Base64
(https://lkml.kernel.org/r/20210718000125.59701-1-ebiggers@kernel.org).

Note that despite breaking these tests, the kernel change is not
expected to break any real users, as the fscrypt no-key name encoding
has always been considered an implementation detail.  So it is
appropriate to just update these tests.

Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/encrypt: add helper function for filtering no-key names

Add a helper function _filter_nokey_filenames() which replaces no-key
filenames with "NOKEY_NAME".  Use it in generic/419 and generic/429.

Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic: update encryption tests to use term "no-key names"

To avoid ambiguity, don't use the terms "ciphertext names" or "encrypted
names" when we really mean "no-key names" (the names that the filesystem
shows when userspace lists an encrypted directory without the
directory's encryption key being present).  This aligns with changes
that have been made to the kernel source code and documentation
(https://lore.kernel.org/r/20200924042624.98439-1-ebiggers@kernel.org).

No change to the actual test logic.

Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/423: test stx_size of symlink

Update generic/423 to test that the reported size of a symlink is equal
to the length of the symlink target, as required by POSIX and as
documented in various man pages.

When the test_dummy_encryption mount option is enabled, this serves as a
regression test for the bug that will be fixed by the patch series
"fscrypt: report correct st_size for encrypted symlinks"
(https://lkml.kernel.org/r/20210702065350.209646-1-ebiggers@kernel.org).
However, this test is applicable to all filesystems and mount options.

Signed-off-by: Eric Biggers <ebiggers@google.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/561: hide assertions when duperemove is killed

Use some bash redirection trickery to capture in $seqres.full all of
bash's warnings about duperemove being killed due to assertions
triggering.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

new: allow users to specify a new test id

Alter the ./new script so that one can set the test id explicitly.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/019: don't dump cores when fio/fsstress hit io errors

Disable coredumps so that fstests won't mark the test failed when the
EIO injector causes an mmap write to abort with SIGBUS.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/371: disable speculative preallocation regressions on XFS

Once in a very long while, the fallocate calls in this test will fail
due to ENOSPC conditions.  While in theory this test is careful only to
allocate at most 160M of space from a 256M filesystem, there's a twist
on XFS: speculative preallocation.

The first loop in this test is an 80M appending write done in units of
4k.  Once the file size hits 64k, XFS will begin speculatively
preallocating blocks past the end of the file; as the file grows larger,
so will the speculative preallocation.

Since the pwrite/rm loop races with the fallocate/rm loop, it's possible
that the fallocate loop will free that file just before the buffered
write extends the speculative preallocation out to 160MB.  With fs and
log overhead, that doesn't leave enough free space to start the 80MB
fallocate request, which tries to avoid disappointing the caller by
freeing all speculative preallocations.  That fails if the pwriter
thread owns the IOLOCK on $testfile1, so fallocate returns ENOSPC and
the test fails.

The simple solution here is to disable speculative preallocation by
setting an extent size hint if the fs is XFS.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/084: fix test program status collection and processing

On a test VM with 1.2GB memory, I noticed that the test will
sometimes fail because resvtest leaks too much memory and gets OOM
killed.  It would be useful to _notrun the test when this happens so
that it doesn't appear as an intermittent regression.

The exit code processing in this test is incorrect, since "$?" will
get us the exit status of _filter_resv, not $here/src/resvtest.  Fix
that as part of learning to detect a SIGKILL and skip the test.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

check: run _check_filesystems in an OOM-happy subshell

While running fstests one night, I observed that fstests stopped
abruptly because ./check ran _check_filesystems to run xfs_repair.
In turn, repair (which inherited oom_score_adj=-1000 from ./check)
consumed so much memory that the OOM killer ran around killing other
daemons, rendering the system nonfunctional.

This is silly -- we set an OOM score adjustment of -1000 on the
./check process so that the test framework itself wouldn't get
OOM-killed, because that aborts the entire run.  Everything else is
fair game for that, including subprocesses started by
_check_filesystems.

Therefore, adapt _check_filesystems (and its children) to run in a
subshell with a much higher oom score adjustment.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

dmthin: erase the metadata device properly before starting

Every now and then I see the following failure when running generic/347:

  --- generic/347.out
  +++ generic/347.out.bad
  @@ -1,2 +1,2 @@
   QA output created by 347
  -=== completed
  +failed to create dm thin pool device

Accompanied by the following dmesg spew:

device-mapper: thin metadata: sb_check failed: blocknr 7016996765293437281: wanted 0
device-mapper: block manager: superblock validator check failed for block 0
device-mapper: thin metadata: couldn't read superblock
device-mapper: table: 253:2: thin-pool: Error creating metadata object
device-mapper: ioctl: error adding target to table

7016996765293437281 is of course the magic number 0x6161616161616161,
which are stale ondisk contents left behind by previous tests that wrote
known tests patterns to files on the scratch device.  This is a bit
surprising, since _dmthin_init supposedly zeroes the first 4k of the
thin pool metadata device before initializing the pool.  Or does it?

dd if=/dev/zero of=$DMTHIN_META_DEV bs=4096 count=1 &>/dev/null

Herein lies the problem: the dd process writes zeroes into the page
cache and exits.  Normally the block layer will flush the page cache
after the last file descriptor is closed, but once in a while the
terminating dd process won't be the only process in the system with an
open file descriptor!

That process is of course udev.  The write() call from dd triggers a
kernel uevent, which starts udev.  If udev is running particularly
slowly, it'll still be running an instant later when dd terminates,
thereby preventing the page cache flush.  If udev is still running a
moment later when we call dmsetup to set up the thin pool, the pool
creation will issue a bio to read the ondisk superblock.  This read
isn't coherent with the page cache, so it sees old disk contents and the
test fails even though we supposedly formatted the metadata device.

Fix this by explicitly flushing the page cache after writing the zeroes.

Fixes: 4b52fffb ("dm-thinp helpers in common/dmthin")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

shared/298: fix random deletion when filenames contain spaces

Correct the deletion loop in this test to work properly when there are
files in $here that have spaces in their name.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/172: disable test when file writes don't use delayed allocation

This test tries to exploit an interaction between delayed allocation and
writeback on full filesystems to see if it can trip up the filestreams
allocator.  The behaviors do not present if the filesystem allocates
space at write time, so disable it under these scenarios.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

src/seek_sanity_test: test seeks from LLONG_MIN

We don't currently test seeking from LLONG_MIN, which apparently can
produce an UBSAN warning, although I've been unable to reproduce
that.  In any case, it's a good corner case to test and
straightforward to add.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/459: catch mount failure

Because if we don't, we freeze the root filesystem instead of
$SCRATCH_MNT and it all goes downhill from there.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/{453,454}: Don't run for FSs restricting names

Running generic/{453,454} tests is probably only useful for filesystems
that allow unrestricted byte streams for names.

Signed-off-by: Pavel Reichl <preichl@redhat.com>
Suggested-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: install btrfs_crc32c_forged_name.py

btrfs/154 uses btrfs_crc32c_forged_name.py but the file is not
installed in the PKG_LIB_DIR. Also consolidate install target files
into EXTRA_EXECS.

Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs: test fsync after increasing file size with truncate

Test that if we explicitly fsync a file that was previously renamed and
its size was increased through a truncate operation, after a power failure
the file has the size set by the truncate operation. Also, in between the
truncation and the fsync, there was a rename of another file in the same
directory and that file was also fsynced before we fsynced the file that
was truncated.

This currently fails on a 5.13 kernel and on Linus' master branch. It is
fixed by a patch with the following subject:

  "btrfs: fix unpersisted i_size on fsync after expanding truncate"

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs: test case to fstrim on a degraded filesystem

Create a degraded btrfs filesystem and run fstrim on it.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

ext4/048: skip test of filename wipe if journal checkpoint is not supported

ext4/048 will fail when running on older kernels that don't support
the filename wipe feature.  The journal checkpoint ioctl is a related
feature, and landed just a little bit after filename wipe feature, so
use support for the journal checkpoint ioctl as a proxy for support
for the filename wipe feature.

Without this change, this test will fail when tesing 5.10, 5.4, and
other LTS kernels.

Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: Leah Rumancik <leah.rumancik@gmail.com>
Tested-by: Yang Xu <xuyang2018.jy@fujitsu.com>
Reviewed-by: Leah Rumancik <leah.rumancik@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

overlay/Makefile: Use the unified name rules

Since xfs/ext4 uses XFS_DIR/EXT4_DIR in Makefile(also for other filesystems),
I think we don't have any reason not to use OVERLAY_DIR in here.

Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

Makefile: fix bug when running make install under tests/* directory

The TESTS_DIR value is defined in TOPDIR makefile, it is empty here.
So running make install on ext4 directory will get the following info:

../../install-sh -o root -g root -m 755 -d /var/lib/xfstests//ext4

We can see they aren't installed under /var/lib/xfstests/tests/ext4 directory.
Fix this by moving TESTS_DIR definition to include/builddefs.in file.

Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

tools: make convert-group ignore converted test files

Teach the tool that rewrites tests files to use _begin_fstest group
tagging to ignore tests that have already been treated.  This will make
it easier for people to rebase their dev branches.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic: ensure that page is properly filled before write

We had a broken optimization in cephfs and netfs lib that could
cause part of a page to be improperly zeroed-out when writing to an
offset that was beyond the EOF but in an existing page.

Add a simple test that would have caught this.

Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs/504: Add scratch_mount before checking for xfs_scrub unicode support

We may not detect the error `Inappropriate ioctl for device`, while running
`$XFS_IO_PROG -c "scrub probe" "$mountpoint"`, if scratch device is not
mounted before invoking _check_xfs_scrub_does_unicode(). So do
_scratch_mount before checking for xfs_scrub support.

Signed-off-by: Anju T Sudhakar <anju@linux.vnet.ibm.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

check: exit with exit code 1 after printing the usage message

If check is passed an invalid command line option, exit with a
non-zero exit code so that a script calling check can detect the
failure.  The check script already performs an "exit 1" if a valid
option has an invalid argument, so this is consistent with existing
practice.

Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/260: f2fs is also special

It fail on f2fs:

 [+] Default length with start set (should succeed)
 [+] Length beyond the end of fs (should succeed)
 [+] Length beyond the end of fs with start set (should succeed)
+After the full fs discard 0 bytes were discarded however the file system is 12882804736 bytes long.
+It seems that fs logic handling len argument overflows

The root cause is f2fs can tag a special flag TRIMMED_FLAG to
indicate the whole filesystem is trimmed, so after mkfs/fstrim(),
following fstrim() won't trim any block.

Suggested-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Sun Ke <sunke32@huawei.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/103: special left calculation for f2fs

It failed on f2fs:
      QA output created by 103
     +fallocate: No space left on device
      Silence is golden.
     ...

f2fs uses index(radix) tree as mapping metadata, its space overhead
is about one thousandth of the data.

Suggested-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Sun Ke <sunke32@huawei.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/042: make a bigger img for f2fs

f2fs-utils 1.9.0 needs at least 38 MB space for f2fs image. However,
f2fs-utils 1.14.0 needs at least 52 MB. Not sure if it will change
again.  So, just set it to 128M.

Suggested-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Sun Ke <sunke32@huawei.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/019: fix test cleanup

This never worked properly and the failure was exposed by the recent
rework of the fstest setup infrastructure.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

misc: update documentation to reflect auto-generated group files

Update the documentation to outline the new requirements for test files
so that we can generate group files during build.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: remove test group management code

Remove all the code that manages group files, since we now generate
them at build time.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: remove group files

From: Darrick J. Wong <djwong@kernel.org>

Now that we autogenerate group files, get rid of them in the source
tree.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

check: use generated group files

Convert the ./check script to use the automatically generated group list
membership files, as the transition is now complete.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: adapt the new test script to our new group tagging scheme

Now that we autogenerate group files, adapt the new test creation script
to use autogenerated group files and to set the group data in the new
test.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: convert nextid to use automatic group generation

Convert the nextid script to use the automatic group file generation to
figure out the next available test id.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: automatically generate group files

Now that we've moved the group membership details into the test case
files themselves, automatically generate the group files during build.
The autogenerated files are named "group.list" instead of "group" to
avoid conflicts between generated and (stale) SCM files as everyone
rebases.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: move test group info to test files

Refactor every test in the entire test suite to use the new boilerplate
functions.  This also migrates all the test group information into the
test files.  This patch has been autogenerated via the command:

./tools/convert-group btrfs ceph cifs ext4 f2fs generic nfs ocfs2 overlay perf shared udf xfs

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: clean up open-coded golden output

Fix the handful of tests that open-coded 'QA output created by XXX'.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: add tool migrate group membership data to test files

Create a tool to migrate the mapping of tests <-> groups out of the
group file and into the individual test file as a _begin_fstest
call.  In the next patches we'll rewrite all the test files and auto
generate the group files from the tests.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: refactor test boilerplate code

Create two new helper functions to deal with boilerplate test code:

A helper function to set the seq and seqnum variables.  We will expand
on this in the next patch so that fstests can autogenerate group files
from now on.

A helper function to register cleanup code that will run if the test
exits or trips over a standard range of signals.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

misc: move exit status into trap handler

Move the "exit $status" clause of the _cleanup function into the
argument to the "trap" command so that we can standardize the
registration of the atexit cleanup code in the next few patches.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: fix group check in new script

In the tests/*/group files, group names are found in the Nth columns of
the file, where N > 1.  The grep expression to warn about unknown groups
is not correct (since it currently checks column 1), so fix this.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Chandan Babu R <chandanrlinux@gmail.com>
Reviewed-by: Allison Henderson <allison.henderson@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs: test incremental send swapping location of a directory with a file

Test that an incremental send operation succeeds, and produces the
correct results, after renaming and moving around directories and files
with multiple hardlinks, in such a way that one of the files gets the old
name and location of a directory and another name (hardlink) with the old
name and location of another file that was located in that same directory.

This currently fails on btrfs but is fixed by a kernel patch with the
following subject:

  "btrfs: send: fix invalid path for unlink operations after parent orphanization"

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

ext4: test journal checkpoint ioctl

Test for commit "ext4: add ioctl EXT4_IOC_CHECKPOINT". Tests journal
checkpointing and journal erasing via EXT4_IOC_CHECKPOINT with flag
EXT4_IOC_CHECKPOINT_FLAG_ZEROOUT set.

Signed-off-by: Leah Rumancik <leah.rumancik@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

ext4/048: add check for od endian flag

od --endian=little is used to read bytes in little endian when on a
big endian machine. Update test ext4/048 to require od endian flag
support on big endian machines, otherwise, skip test.

Signed-off-by: Leah Rumancik <leah.rumancik@gmail.com>
Reviewed-by: Yang Xu <xuyang2018.jy@fujitsu.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/rc: fix _scratch_mkfs_blocksized indentation

No function change. Fix indentation.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs: support other sectorsizes in _scratch_mkfs_blocksized

When btrfs supports sectorsize != pagesize it can run these test cases
now,
generic/205 generic/206 generic/216 generic/217 generic/218 generic/220
generic/222 generic/227 generic/229 generic/238

This change is backward compatible for kernels without non pagesize
sectorsize support.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

overlay/075: add test coverage for clearing immutable/append-only flags

overlay/075 is a variant of check -overlay generic/079.

check -overlay generic/079 does the following operations on overlay fs:
1. Create files
2. Set immutable/append-only flags
3. Verify files/dirs behaving as immutable/append-only
4. Clear immutable/append-only flags
5. Remove files

overlay/075 performs steps 1,2,4,5 on upper and lower layers
and only step 3 is performed on overlay fs (before and after copy up
and mount cycle).

Add also steps 4,5 to be performed on overlay fs to increase the
test coverage of the "merged" inode xflags feature.

Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/507: support more filesystems

The commit message introducing the test says:
"We only check below attribute modification which most filesystem
 supports:
    - no atime updates (A)
    - secure deletion (s)
    - synchronous updates (S)
    - undeletable (u)
"
But in fact, very few filesystems support the (s) and (u) flags.
xfs and btrfs do not support them for example.

The test doesn't need to check those specific flags, so replace those
flags with immutable (i) and append-only (a), which most filesystems
really do support.

Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/config: remove default 4k blocksize from XFS_MKFS_OPTIONS

xfstests set "-bsize=4k" to XFS_MKFS_OPTIONS by default, then give
it to MKFS_OPTIOPNS. So MKFS_OPTIOPNS always contains "-bsize=4k"
except we set XFS_MKFS_OPTIONS manually.

It's useless to set XFS_MKFS_OPTIONS to "-bsize=4096" by default,
especially that will cause all cases with _scratch_mkfs_blocksized()
always fail as "-b size option respecified", when test on XFS. For
exmaple: generic/222

Signed-off-by: Zorro Lang <zlang@redhat.com>
Acked-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

overlay/075: fix wrong invocation of t_immutable

t_immutable cannot be run twice on the same test directoty, because
append-only directory tests create files in append-only.d and those
file already exist from the first run.

Use separate test directories for the first and second t_immutable runs.

Signed-off-by: Amir Goldstein <amir73il@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs/029: fix the test compatible with older cp(1)

cp(1) versions 8.30 and 8.32 are compared with its --reflink=always option
and they have different semantic if the target-file (with zero sizes) must be
created when the cp --reflink=alaways fails with a cross-device link
error. As shown below.

$ cp --version | head -1
cp (GNU coreutils) 8.30

$ cp --reflink=always /mnt/scratch/original /mnt/test/test-029/copy
cp: failed to clone '/mnt/test/test-029/copy' from '/mnt/scratch/original': Invalid cross-device link

$ ls -l /mnt/test/test-029/copy
ls: cannot access '/mnt/test/test-029/copy': No such file or directory

$ cp --version | head -1
cp (GNU coreutils) 8.32

$ cp --reflink=always /mnt/scratch/original /mnt/test/test-029/copy;
cp: failed to clone '/mnt/test/test-029/copy' from '/mnt/scratch/original': Invalid cross-device link

$ ls -l /mnt/test/test-029/copy
-rw------- 1 root root 0 Jun  4 13:29 /mnt/test/test-029/copy

Reported-by: Wang Yugui <wangyugui@e16-tech.com>
Signed-off-by: Anand Jain <anand.jain@oracle.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

ext4: prevent initializing reserved inodes

Check the block group zero and prevent initializing reserved inodes.
But in some special cases, the reserved inode may not all belong to
the group zero, it may exist into the second group if we format 
filesystem below.

  mkfs.ext4 -b 4096 -g 8192 -N 1024 -I 4096 /dev/sda

So, it will end up triggering a false positive report of a corrupted
file system

It's a regression test for kernel commit a149d2a5cabb ("ext4: fix
check to prevent false positive report of incorrect used inodes")

Signed-off-by: Chen Long <chenlongcl.chen@huawei.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: initial bcachefs support

To get started, you'll need to get bcachefs from the kernel repository
  https://evilpiepirate.org/git/bcachefs.git/

and tools from
  https://evilpiepirate.org/git/bcache-tools.git/

Build kernel as normal, enabling CONFIG_BCACHEFS_FS (and probably
CONFIG_BCACHEFS_DEBUG), and build and install tools. Then running
fstests is exactly the same as other local filesystems - just set
FSTYP=bcachefs

Also see https://evilpiepirate.org/git/ktest.git/ for a tool for
conveniently building kernel and running fstests inside a qemu VM.

This patch also updates generic/441 to run the more thorough test on
bcachefs, and generic/425 to not run on bcachefs (since bcachefs does
not store xattrs in blocks)

Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

fstests: improved .gitignore

Ignore dotfiles, tags, and verifier state.

Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

generic/042: set size to 38M for f2fs

25M is too small for f2fs:

Info: Disable heap-based policy
Info: Debug level = 0
Info: Label =
Info: Trim is enabled
Info: Segments per section = 1
Info: Sections per zone = 1
Info: sector size = 512
Info: total sectors = 51200 (25 MB)
Info: zone aligned segment0 blkaddr: 512
    Error: Device size is not sufficient for F2FS volume
    Error: Failed to prepare a super block!!!
    Error: Could not format the device!!!

So, set size to 38M for f2fs.

Signed-off-by: Sun Ke <sunke32@huawei.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

btrfs: add test for multiple fsync with adjacent preallocated extents

Test a scenario where we do several partial writes into multiple
preallocated extents across two transactions and with several fsyncs
in between. The goal is to check that the fsyncs succeed.

Currently the last fsync fails with an -EIO error, and it aborts the
current transaction. This issue is fixed by a patch with the following
subject:

  "btrfs: fix fsync failure and transaction abort after writes to prealloc extents"

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

common/xfs: Fix _require_scratch_xfs_shrink bug

Since local declaration can't pass function return value, this
helper never detects whether kernel or xfsprogs supports xfs shrink
feature successfully. Fit it by separating declaration and
assignment of local variables.

Signed-off-by: Yang Xu <xuyang2018.jy@fujitsu.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

xfs: fix check of whether the mkfs fail

The return value ("$?") is always the status of the last command in
the pipe.

Signed-off-by: Sun Ke <sunke32@huawei.com>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

idmapped-mounts: split out run_test() function

to make it easier to run subsets of tests.

Cc: fstests@vger.kernel.org
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

idmapped-mounts: add missing newline to print_r()

The function missed to print a newline making the output difficult to
read when running with DEBUG_TRACE.

Cc: fstests@vger.kernel.org
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>

idmapped-mounts: remove unused set_cloexec() helper

This function has never been used so remove it.

Cc: fstests@vger.kernel.org
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Reviewed-by: Eryu Guan <guaneryu@gmail.com>
Signed-off-by: Eryu Guan <guaneryu@gmail.com>