Hou Tao [Thu, 11 Jan 2018 06:49:31 +0000 (14:49 +0800)]
dmflakey: support error_writes feature for dm-flakey
Add _require_flakey_with_error_writes() to check the availability of
dm-flakey target and its error_writes feature, and support for
enabling FLAKEY_TABLE_ERROR table by passing FLAKEY_ERROR_WRITES to
_load_flakey_table().
Hou Tao [Thu, 11 Jan 2018 06:49:30 +0000 (14:49 +0800)]
dmflakey: support multiple dm targets for a dm-flakey device
dm-flakey can be used to emulate IO write error, however, when we
also need to prevent the IO error for a specific range of the block
device (eg., the log region of a XFS), we need to specify multiple
dm targets for the dm device.
Option --table can not accommodate the multiple dm targets case, so
let dmsetup get the possible-multiple-targets table from standard
input.
Eric Whitney [Wed, 10 Jan 2018 21:19:05 +0000 (16:19 -0500)]
common/rc: ext4 doesn't support O_DIRECT with data journaling
Modify _require_odirect() to skip O_DIRECT tests if the ext4 mount
option data=journal has been applied. Because ext4 disables its
O_DIRECT support when in data journaling mode, no significant value
is obtained by running these tests and they consume a significant
amount of run time. Additionally, this change eliminates the need
to triage false negative O_DIRECT test results caused by test
problems, as has occurred from time to time.
Chengguang Xu [Sat, 6 Jan 2018 08:23:58 +0000 (16:23 +0800)]
common/rc: perform _require_metadata_journaling check on overlay base fs
Currently generic shutdown tests are enabled on overlayfs by
extending _require_scratch_shutdown() to check shutdown support for
overlayfs, some of those tests also call
_require_metadata_journaling() to inspect whether test filesystem
support metadata journaling or not, so need to extend
_require_metadata_journaling() as well to check metadata journaling
support for overlayfs in case shutdown tests run on uncommon
overlayfs setup and fail. For example, base filesystem of overlayfs
is ext2 or ext4 with noload mode.
Metadata journaling check is based on base filesystem configurations
and because -overlay option saves those configurations to
OVL_BASE_*, adding restore/override the configurations before/after
the check.
Tested ext2, xfs, ext4(load/no-load mode) as base filesystem of
overlayfs in single and multi section configurations.
Ross Zwisler [Thu, 4 Jan 2018 08:11:21 +0000 (01:11 -0700)]
ext4: test for inline data + DAX corruption
Add a regression test for the following kernel commit:
ext4: prevent data corruption with inline data + DAX
The test passes either if we don't encounter corruption, or if
mounting with DAX + inline data fails. The latter is the way that
we prevent this issue in the kernel.
Ross Zwisler [Thu, 4 Jan 2018 08:11:20 +0000 (01:11 -0700)]
ext4: test for DAX + journaling corruption
Add a regression test for the following kernel commit:
ext4: prevent data corruption with journaling + DAX
The test passes if either we successfully compare the data between
the mmap with journaling turned on and the one with journaling
turned off, or if we fail the chattr command to turn on or off
journaling. The latter is how we prevent this issue in the kernel.
Darrick J. Wong [Wed, 13 Dec 2017 23:28:05 +0000 (15:28 -0800)]
xfs/068: fix clonerange problems in file/dir count output
In this test we use a fixed sequence of operations in fsstress to
create some number of files and dirs and then exercise
xfsdump/xfsrestore on them. Since clonerange/deduperange are not
supported on all xfs configurations, detect if they're in fsstress
and disable them so that we always execute exactly the same sequence
of operations no matter how the filesystem is configured.
[eguan: move it from xfs/068 to common/dump as Dave suggested]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Suggested-by: Dave Chinner <david@fromorbit.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Luis Henriques [Tue, 2 Jan 2018 17:17:05 +0000 (17:17 +0000)]
generic/050: fix _require_local_device $SCRATCH_DEV check order
Commit 2b4eae7fd825 ("common/rc: add scratch shutdown support for
overlayfs") added a _require_local_device check to generic tests 042
and 050. However, for test 050, this check was added _before_
actually verifying that a SCRATCH_DEV actually exists. This patch
simply re-orders the _require_local_device to the right place.
Darrick J. Wong [Wed, 13 Dec 2017 06:03:48 +0000 (22:03 -0800)]
generic: test error shutdown while stressing filesystem
Test log recovery with repeated (simulated) disk failures. We kick
off fsstress on the scratch fs, then switch out the underlying
device with dm-error to see what happens when the disk goes down.
Having taken down the fs in this manner, remount it and repeat.
This test is a Good Enough (tm) simulation of our internal multipath
failure testing efforts.
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>
Theodore Ts'o [Fri, 22 Dec 2017 18:26:07 +0000 (13:26 -0500)]
ext4: use a slightly bigger file system in ext4/021
An 8 MB file system may not be big enough for certain file system
configurations --- in particular, if the inode size is 2048 bytes.
Make the test file system 10MB instead.
Ari Sundholm [Thu, 21 Dec 2017 22:10:05 +0000 (00:10 +0200)]
common/encrypt: Create an encrypted equivalent of _scratch_mkfs_sized
Test case generic/399 hardcodes "-O encrypt" in MKFS_OPTIONS when
calling _scratch_mkfs_sized, which only works with the mkfs of
certain filesystems. Create a new helper,
_scratch_mkfs_sized_encrypted, for handling the differences between
the mkfs tools of different filesystems. It also allows those
filesystems whose mkfs doesn't accept "-O encrypt" to skip the test
gracefully until proper support is added for them in the helper.
ubifs is not supported in the new helper despite supporting
encryption, as _scratch_mkfs_sized has no ubifs support and adding
that should be done in a separate patch.
Getting xfs/040 to "pass" takes a bit of effort, however the effort
to require updating xfsprogs is purely an xfsprogs maintainer task
only. There no functional gain by users of xfs or a QA team to get
this test to pass. This is not trivial from the current description
so document this.
Signed-off-by: Luis R. Rodriguez <mcgrof@kernel.org> Reviewed-by: Bill O'Donnell <billodo@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Chengguang Xu [Fri, 15 Dec 2017 07:47:33 +0000 (15:47 +0800)]
common/rc: add scratch shutdown support for overlayfs
Enable shutdown tests on overlayfs. This requires configuring
overlayfs with TEST|SCRATCH_DEV pointing to block device, which act
as base fs of overlay. The legacy config (pointing TEST|SCRATCH_DEV
to existing directories) won't support shutdown.
Extend _require_scratch_shutdown() to deal with overlay by checking
shutdown support on $OVL_BASE_SCRATCH_MNT instead of $SCRATCH_MNT,
so we check shutdown support status against the underlying upper fs
of overlay.
Introduce new _scratch_shutdown() helper to do the actual shutdown
work and shutdown upper fs in the case of overlay. Also converting
existing bare 'src/godown' calls to use this helper to avoid false
failures when testing overlayfs.
generic/042 and generic/050 assume $SCRATCH_DEV to be a local
device, so add _require_local_device $SCRATCH_DEV rule.
Ari Sundholm [Wed, 20 Dec 2017 17:00:11 +0000 (19:00 +0200)]
fsstress: getcwd() return value fixes
Add a check for the return value of getcwd(). Fix another check
which mistakenly checks if the return value is less than zero
instead of checking whether the return value is NULL.
Ari Sundholm [Wed, 20 Dec 2017 17:00:10 +0000 (19:00 +0200)]
fsstress: Eliminate valgrind noise
This patch plugs what we think are the remaining sources of valgrind
noise we found when running fsstress under valgrind. The noise is
caused by memory being left unfreed before process termination.
Darrick J. Wong [Wed, 13 Dec 2017 22:45:01 +0000 (14:45 -0800)]
xfs: fix tests to handle removal of no-alloc create nonfeature
We're removing from XFS the ability to perform no-allocation file
creation. This was added years ago because some customer of SGI
demanded that we still be able to create (empty?) files with zero
free blocks remaining so long as there were free inodes and space in
existing directory blocks. This came at an unacceptable risk of
ENOSPC'ing midway through a transaction and shutting down the fs, so
we're removing it for the create case having changed our minds 20
years later.
However, some tests fail as a result, so fix them to be more
flexible about not failing when a dir/file creation fails due to
ENOSPC.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
build: update AC_PACKAGE_WANT_GDBM() and src/dbtest.c to build
Modern gdbm-devel packages bundle together gdbm.h and ndbm.h. The
old m4 macro had detection support for some old gdbm libraries but
not for new ones.
We fix compilation of src/dbtest.c by making the autoconf helper
check for this new arrangement:
If both gdbm.h and ndbm.h are found define set both gdbm_ndbm_=true,
and have_db=true, and define HAVE_GDBM_H. The src/dbtest.c already
had a HAVE_GDBM_H but there was never a respective autoconf settter
for it. We can just re-use this and fix it for new arrangement.
Signed-off-by: Luis R. Rodriguez <mcgrof@kernel.org> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Nikolay Borisov [Thu, 30 Nov 2017 16:05:27 +0000 (18:05 +0200)]
generic: initial fiemap range query test
Fiemap gained support for passing in optional offset len which
denote the range requested, so this patch adds testcases for this
functionality. Aditionally, a special "ranged" argument is added to
the require_xfs_io_command which checks for the presence of fiemap
range support.
Getting an error midway through a direct write would return an error
and the error-code is returned in the write() call. However, partial
data is over-written during the call.
This tests simulates the ENOSPC error to check for partial direct
write consistency.
common/rc: Check pwrite parameters in _require_xfs_io_command
There are some parameters added with xfs_io. Check if the pwrite
parameters are available. For some cases, xfs_io now returns
"command -%c not supported", so added "not supported" to count as
error.
Theodore Ts'o [Tue, 12 Dec 2017 05:19:37 +0000 (00:19 -0500)]
common/populate: fix S_IFDIR.FMT_BLOCK and ATTR.FMT_LOCAL for ext4 encryption
When ext4 encryption is enabled, the directory entries are encrypted
so we need to create fewer directory entries to guarantee that they
will all fit in a single block.
Also, the encryption metadata takes up extended attribute room so we
can only add a single xattr to guarantee that the xattrs will fit in
the inode.
Theodore Ts'o [Mon, 11 Dec 2017 22:49:19 +0000 (17:49 -0500)]
common: rework _require_ext4_mkfs_feature
In all of the places where we need check to see if mkfs.ext4 can
support a set of file system features, we also should be checking to
see if the kernel can support those file system features. So rename
_require_ext4_mkfs_feature to _require_scratch_ext4_feature, and
actually format the file system in $SCRATCH_DEV.
Also allow ext4/306 to run on systems where mke2fs doesn't support
the "64bit" option.
Ross Zwisler [Thu, 7 Dec 2017 21:35:03 +0000 (14:35 -0700)]
build: fix Wlog_Error_String overflow issues
The 'Wlog_Error_String' string is defined to be 256 bytes in length,
but in two places we write into it with a format that contains a
string (wfile->w_file) that has length 1024. This can overflow
Wlog_Error_String, as we see in the new compiler warnings from gcc
7.2.1:
write_log.c:124:37: warning: ‘%s’ directive writing up to 1023 bytes into
a region of size 224 [-Wformat-overflow=]
"Could not open write_log - open(%s, %#o, %#o) failed: %s\n",
^~
Fix this by increasing the length of Wlog_Error_String to 1280
characters (1024 for wfile->w_file plus 256 for the rest of the
format string), and by using snprintf() instead of sprintf() so we
are sure we don't overflow.
Ross Zwisler [Thu, 7 Dec 2017 21:35:02 +0000 (14:35 -0700)]
build: fix TagName overflow issues
The 'TagName' string is defined to be 40 characters in length, but
in three places we write into it with a format of "(%.39s)". This
can result in a string of up to 42 characters, the 39 character user
string plus "()\0". This overflows TagName, as we see in the new
complier warnings from gcc 7.2.1:
iogen.c:1277:6: note: 'sprintf' output between 3 and 42 bytes into a
destination of size 40
sprintf( TagName, "(%.39s)", optarg );
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Fix this by limiting the user string to 37 characters.
Ross Zwisler [Thu, 7 Dec 2017 21:35:01 +0000 (14:35 -0700)]
build: fix _BSD_SOURCE complier warning
When compiling xfstests with either gcc 6.4.1 or 7.2.1 I see the
following warning:
In file included from /usr/include/stdio.h:27:0,
from fssum.c:25:
/usr/include/features.h:148:3: warning: #warning "_BSD_SOURCE and _SVID_SOURCE
are deprecated, use _DEFAULT_SOURCE" [-Wcpp]
# warning "_BSD_SOURCE and _SVID_SOURCE are deprecated, use _DEFAULT_SOURCE"
^~~~~~~
The feature_test_macros(7) man page says this about _BSD_SOURCE:
Since glibc 2.20, this macro is deprecated. It now has the same
effect as defining _DEFAULT_SOURCE, but generates a compile-time
warning (unless _DEFAULT_SOURCE is also defined). Use
_DEFAULT_SOURCE instead. To allow code that requires _BSD_SOURCE
in glibc 2.19 and earlier and _DEFAULT_SOURCE in glibc 2.20 and
later to compile without warnings, define both _BSD_SOURCE and
_DEFAULT_SOURCE.
Keep backwards compatibility with older code by defining both
_BSD_SOURCE and _DEFAULT_SOURCE.
Ross Zwisler [Thu, 7 Dec 2017 23:19:50 +0000 (16:19 -0700)]
generic: add test for DAX MAP_SYNC support
This test creates a file and writes to it via an mmap(), but never
syncs via fsync/msync. This process is tracked via dm-log-writes,
then replayed.
If MAP_SYNC is working the dm-log-writes replay will show the test
file with 1 MiB of on-media block allocations. This is because each
allocating page fault included an implicit metadata sync. If
MAP_SYNC isn't working (which you can test by removing the "-S" flag
to xfs_io mmap) the file will be smaller or missing entirely.
Note that dm-log-writes doesn't track the data that we write via the
mmap(), so we can't do any data integrity checking. We can only
verify that the metadata writes for the page faults happened.
[eguan: add comments on _require_log_writes_dax and fix its cleanup]
Ross Zwisler [Wed, 6 Dec 2017 00:37:43 +0000 (17:37 -0700)]
dm-log-writes: only replay log to marks that exist
The 'replay-log' executable will replay the dm-log-writes log until
the given mark, or until the end of the log if the mark isn't found.
This means that if the mark you're looking for was never inserted in
the log or if you give garbage to _log_writes_replay_log() the
entire log will be replayed. This can cause unexpected test
results.
Fix this by making sure that the mark we're given actually exists in
the log before we allow the replay.
Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com> Reviewed-by: Amir Goldstein <amir73il@gmail.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
yang xu [Fri, 1 Dec 2017 10:05:36 +0000 (18:05 +0800)]
generic/459: Fix check for ro-remount in extN
Currently ,freeze failure caused by the lack of space can not
guarantee to remount extN filesystem in read-only mode, and test
failed due to "ro" mount option not found. We can add a touch to
trigger the action which aborts journal and ro-remounts the fs.
Qu Wenruo [Wed, 29 Nov 2017 06:14:19 +0000 (14:14 +0800)]
btrfs: make sure btrfs can handle full fs trim correctly
Ancient commit f4c697e6406d ("btrfs: return EINVAL if start >
total_bytes in fitrim ioctl") introduced a regression where btrfs
may fail to trim any free space in existing block groups.
It's caused by confusion with btrfs_super_block->total_bytes and
btrfs logical address space.
Unlike physical address, any aligned bytenr in range [0, U64_MAX) is
valid in btrfs logical address space, and it's chunk mapping
mechanism of btrfs to handle the logical<->physical mapping.
The test case will craft a btrfs with the following features:
0) Single data/meta profile
Make trimmed bytes reporting and chunk allocation more predictable.
1) All chunks start beyond super_block->total_bytes (1G)
By relocating these blocks several times.
2) Unallocated space is less than 50% of the whole fs
3) Fragmented data chunks
Data chunks will be full of fragments, 50% of data chunks will be
free space.
So in theory fstrim should be able to trim over 50% space of the fs.
(after fix, 64% of the fs can be trimmed)
While the regression makes btrfs only able to trim unallocated
space, which is less than 50% of the total space.
(without fix, it's only 31%)
Fixed by patch named "btrfs: Ensure btrfs_trim_fs can trim the whole
fs".
Amir Goldstein [Tue, 28 Nov 2017 16:42:02 +0000 (18:42 +0200)]
overlay/042: remove wrong check for empty index
The check for empty index in this test was wrongly copied from test
overlay/034. In test overlay/034 lower file starts as a hardlink, so
nlink accounting is done from the first copy up and index can be cleaned
on last upper hardlink unlink. In this test, lower starts as
non-hardlink, so first copy up does not perform nlink accounting and
therefore, union nlink count does not drop to 0 at the end of the test
and the index is not expected to be cleaned.
Amir Goldstein [Tue, 28 Nov 2017 16:42:00 +0000 (18:42 +0200)]
overlay: regression test for hardlink breakage after unlink and mount cycle
- file A and B are hardlinked in lower
- modify A to trigger copy up and index lower
- unlink A and mount cycle
- check that B still contains the modified data
Filipe Manana [Fri, 17 Nov 2017 14:49:05 +0000 (14:49 +0000)]
btrfs: test send for files with multiple hard links renamed
Test that an incremental send operation works if a file that has
multiple hard links has some of its hard links renamed in the send
snapshot, with one of them getting the same path that some other
inode had in the send snapshot.
At the moment this test fails on btrfs and a fix is provived by a
linux kernel patch titled:
"Btrfs: incremental send, fix wrong unlink path after renaming file"
Xiao Yang [Mon, 27 Nov 2017 02:34:27 +0000 (10:34 +0800)]
generic/465: just check the actual read data under dio read/write
I got the following message when running generic/465 in ext4
data=journal mode
---------------------------------------------------------------
QA output created by 465
non-aio dio test
encounter an error: block 0 offset 4096, content 62
encounter an error: block 0 offset 122880, content 62
encounter an error: block 0 offset 274432, content 62
encounter an error: block 0 offset 86016, content 62
aio-dio test
encounter an error: block 0 offset 28672, content 62
encounter an error: block 0 offset 12288, content 62
encounter an error: block 2 offset 16384, content 62
encounter an error: block 1 offset 565248, content 62
---------------------------------------------------------------
In ext4 data=journal mode, direct read will fall back to buffer
read, and buffer read doesn't take inode lock, so it doesn't need to
wait for the writer to finish first and sees the intermediate inode
size and returns data less than 1M.
We can just check the actual read data instead of the whole read
buffer.
Eryu Guan [Tue, 14 Nov 2017 10:26:45 +0000 (18:26 +0800)]
fstests: filter readonly mount error messages
util-linux commit 6dede2f2f7c5 ("libmount: support MS_RDONLY on
write-protected devices") changed the error message on read-only
block device, and in the failure case printed one line message
instead of two (for details please see comments in common/filter),
and this change broke generic/050 and overlay/035.
Fix it by adding more filter rules to _filter_ro_mount and updating
associated .out files to unify the output from both old and new
util-linux versions.
Reviewed-by: Amir Goldstein <amir73il@gmail.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Eryu Guan [Tue, 14 Nov 2017 07:49:36 +0000 (15:49 +0800)]
fstests: filter mount error message for EUCLEAN and ESTALE
util-linux commit ea848180dd34 ("libmount: add
mnt_context_get_excode()") since v2.30 changed the error message on
EUCLEAN and ESTALE again (and maybe other errno too):
- mount: <device> on <mountpoint> failed: Structure needs cleaning
+ mount: <mountpoint>: mount(2) system call failed: Structure needs cleaning.
and it causes xfs/005, overlay/037 to fail (and probably xfs/333 too,
but it's always _notrun for now).
And what's more, the mentioned tests would also fail when testing
with util-linux prior to v2.21, no one complained just because the
tests are usually _notrun on such old distributions that ship
util-linux < v2.21.
So let's filter out the changing parts and keep the error message
simple.
Reviewed-by: Amir Goldstein <amir73il@gmail.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Chandan Rajendra [Mon, 20 Nov 2017 14:34:10 +0000 (20:04 +0530)]
overlay/038: test consistent values of st_ino/d_ino for subdirs
This commit implements similar changes made to overlay/038 test i.e. in
addition to checking consistent values of st_ino/d_ino for "." and ".."
entries, check also the values for subdir inside each tested directory
setup.
Chandan Rajendra [Mon, 20 Nov 2017 14:34:09 +0000 (20:04 +0530)]
overlay/038: code cleanup
This commit implements similar changes made to overlay/041 test i.e.
- Remove duplicate merged dir tests
- Remove duplicate pure upper dir inside merged dir tests
- Sort out comments and error messages
Josef Bacik [Tue, 7 Nov 2017 21:53:32 +0000 (16:53 -0500)]
fstests: add fio perf results support
This patch does the nuts and bolts of grabbing fio results and
storing them in a database in order to check against for future
runs. This works by storing the results in resuts/fio-results.db as
a sqlite database. The src/perf directory has all the supporting
python code for parsing the fio json results, storing it in the
database, and loading previous results from the database to compare
with the current results.
This also adds a PERF_CONFIGNAME option that must be set for this to
work. Since we all have various ways we run fstests it doesn't make
sense to compare different configurations with each other (unless
specifically desired). The PERF_CONFIGNAME will allow us to
separate out results for different test run configurations to make
sure we're comparing results correctly.
Currently we only check against the last perf result. In the future
I will flesh this out to compare against the average of N number of
runs to be a little more complete, and hopefully that will allow us
to also watch latencies as well.
Amir Goldstein [Sun, 19 Nov 2017 11:56:29 +0000 (13:56 +0200)]
overlay/041: use user defined overlay mount option
Mount non-samefs dirs with OVERLAY_MOUNT_OPTIONS, so user can opt-in
for constant inode numbers on non-samefs setup (i.e. -o xino) when
it is implemented.
Amir Goldstein [Sun, 19 Nov 2017 11:56:28 +0000 (13:56 +0200)]
overlay/041: test consistent values of st_ino/d_ino for subdirs
In addition to checking consistent values of st_ino/d_ino for "."
and ".." entries, check also the values for subdir inside each
tested directory setup.
Chandan Rajendra [Sat, 18 Nov 2017 05:57:00 +0000 (11:27 +0530)]
overlay: Test consistent st_ino numbers for non-samefs scenario
This commit adds a test to verify consistent st_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 overlayfs returns consistent
st_ino for the following scenarios,
- Copy-up of lowerdir files
- Rename files and drop dentry/inode cache
- Remount the overlayfs instance
Chandan Rajendra [Sat, 18 Nov 2017 05:58:02 +0000 (11:28 +0530)]
overlay/017: Remove constant st_ino/d_ino test for hardlinks
Constant st_ino/d_ino for hardlinks requires "index" config
feature/mount option to be enabled. overlay/018 already tests
constant st_ino feature for hardlinks. A future commit will add
constant d_ino test to overlay/018 test.
Eryu Guan [Thu, 9 Nov 2017 08:32:42 +0000 (16:32 +0800)]
generic: test zero page cache beyond new EOF on truncate down
From mmap(2) manpage, "a file is mapped in multiples of the page
size. For a file that is not a multiple of the page size, the
remaining memory is zeroed when mapped", this test is to test this
behavior on truncate down.
This is inspired by an XFS bug that truncate down fails to zero page
cache beyond new EOF and causes stale data written to disk
unexpectedly and a subsequent mmap sees non-zeros post EOF.
Patch "xfs: truncate pagecache before writeback in
xfs_setattr_size()" fixed the bug on XFS.
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Liu Bo [Wed, 15 Nov 2017 23:47:59 +0000 (16:47 -0700)]
btrfs/143: make test case more reliable
This changes to use '_scratch_cycle_mount' to drop all caches btrfs
could have in order to avoid an issue that drop_caches somehow
doesn't work on Nikolay's box.
Also use bash -c to run 'read' only when %pid is odd so that we can
read the faulty disk.
Reported-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: Liu Bo <bo.li.liu@oracle.com> Tested-by: Nikolay Borisov <nborisov@suse.com> Reviewed-by: Eryu Guan <eguan@redhat.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Darrick J. Wong [Thu, 16 Nov 2017 01:22:37 +0000 (17:22 -0800)]
aio-dio-append-write-read-race: abort if we encounter syscall errors
If any of the library calls return error codes, just print out a message
and abort the test. Whoever wrote the write test did not check for
write failures, which means that if we ENOSPC without writing anything
then the reader thread will loop forever trying to read.
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>
Chao Yu [Wed, 15 Nov 2017 08:58:33 +0000 (16:58 +0800)]
generic: add testcase to test fallocate & f{data,}sync
f2fs can skip isize updating in fsync(), since during mount, f2fs
tries to recovery isize according to valid block address or
preallocated flag in last fsynced dnode block.
However, fallocate() breaks our rule with setting
FALLOC_FL_KEEP_SIZE flag, since it can preallocated block cross EOF,
once the file is fsynced, in POR, we will recover isize incorrectly
based on these fallocated blocks.
This patch adds a new testcase to test fallocate, in order to verify
whether filesystem will do correct recovery on isize.
Chandan Rajendra [Mon, 13 Nov 2017 14:45:26 +0000 (20:15 +0530)]
src/t_dir_type.c: Use strtoul() instead of atoll()
An overlayfs filesystem instance with one lowerdir filesystem and
with "xino" mount option enabled can have the layer index encoded in
the 63rd bit of the inode number. A signed 64 bit integer won't
suffice to store this inode number. Hence this commit uses strtoul()
to convert the inode number in string form to unsigned integer form.
Darrick J. Wong [Tue, 7 Nov 2017 01:54:11 +0000 (17:54 -0800)]
generic/204: use available blocks to determine the number of files to create
Use the available block count to compute the number of files we think
we can create, rather than hardcoding a particular size. This fixes
the ENOSPC failures for xfs filesystems with rmap/reflink support.
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>
Darrick J. Wong [Fri, 10 Nov 2017 00:44:34 +0000 (16:44 -0800)]
common/fuzzy: ignore padding fields on xfs
Padding fields are never checked and can have arbitrary values (if we
ever put them to use there'll be a feature flag) so there's no point
in fuzz-testing them.
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, 10 Nov 2017 00:44:28 +0000 (16:44 -0800)]
xfs/31[2-7]: hide pwrite fsync errors that are supposed to fail
xfs_io 4.14 will gain the ability to print error messages when
pwrite+fsync fail. Certain tests use the error injector to cause
failures, so the errors are expected. Since we test for a shut down
filesystem after the error injection, we can push the error messages to
the log.
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>
Eryu Guan [Sun, 5 Nov 2017 06:14:11 +0000 (14:14 +0800)]
fstests: do not call _scratch_mkfs_sized in a pipe
_scratch_mkfs_sized will create a filesystem of the given size, and
call _notrun and exit if current $FSTYP doesn't support sized mkfs.
But when it's called in a pipe, the exit in _notrun only exits from
the subshell created by the pipe not the test itself, and test
continues to run unnecessarily, though the test is still reported as
[notrun] due to existence of $seqres.notrun file.
Fix it by not calling _scratch_mkfs_sized in a pipe, but dumping the
output to a tmp file, which will be fed to _filter_mkfs later.
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Eryu Guan [Wed, 25 Oct 2017 04:06:06 +0000 (12:06 +0800)]
ltp/fsx: allow comments when reading operations from logs
We use '--replay-ops' option to replay operations in the specified
operation log file, but we're not allowed to add comments for the
operations in the log, which might be useful when writing regression
tests that replay a given sequence of operations.
Now treat lines starting with '#' as comments and skip them when
reading operations.
Reviewed-by: Amir Goldstein <amir73il@gmail.com> Signed-off-by: Eryu Guan <eguan@redhat.com>
Amir Goldstein [Thu, 2 Nov 2017 10:15:39 +0000 (12:15 +0200)]
generic: add test with more open by file handle use cases
This test is a variant of test generic/426 that tests with less
files and more use cases:
- Create test dir with non empty files with known content and verify
their content after opening file by handle.
- Check open by handle of directory.
- Check open by handle of files that have been unlinked, but still open.
- Check open by handle of files that have been renamed in same dir,
moved to new dir and whose parent dir has been renamed.
Amir Goldstein [Thu, 2 Nov 2017 10:15:38 +0000 (12:15 +0200)]
generic/426: factor out helper functions
Helper test_file_handles() outputs the sub-test command to output,
so if errors are detected in one of the sub-tests, it is easier
to know which sub-tests have failed.
Amir Goldstein [Thu, 2 Nov 2017 10:15:37 +0000 (12:15 +0200)]
open_by_handle: test file handles of open files
usage: open_by_handle -dk <test_dir> [N]
Get file handles for existing test set, keep open file handles for all
test files, unlink all test files, drop caches and try to open all files
by handle.
With -p flag, create/delete also the test_dir itself and try to open by
handle also test_dir itself after droping caches and use dir fd to test
faccessat() of a file inside dir.
mount_fd argument to open_by_handle_at() is open fd of test_dir's parent.
Amir Goldstein [Thu, 2 Nov 2017 10:15:35 +0000 (12:15 +0200)]
open_by_handle: test content of open file handle
usage: open_by_handle -rwa <test_dir> [N]
Get file handles for existing test set, write data to files,
drop caches, open all files by handle, read data and verify old data,
write new data to files.
This is needed for testing that overlay decoded file handles are not
pointing the lower inodes after new data is already written to upper
inodes after copy up.
open_by_handle -a is needed for testing copy up of disconnected overlay
decoded file handles (to index dir).
Darrick J. Wong [Fri, 3 Nov 2017 04:27:17 +0000 (21:27 -0700)]
generic: test IO at maximum file offset
Make sure we can write to and read from the highest possible offset
that Linux will allow. Format the filesystem with a variety of
possible blocksizes to stress the filesystem.
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 21:47:48 +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>
projects / xfstests-dev.git / log