When trying to run test 70, we encountered several problems.
Sergey has strace'd some of the problems to attempts to
set xattr on the wrong test dir (/mnt/test/), which appears to be
hardcoded in common.attr.
Reported-by: Sergey Ivanov <sergey57@gmail.com>
Signed-off-by: Amir Goldstein <amir73il@users.sf.net>
Signed-off-by: Christoph Hellwig <hch@lst.de>
This test just runs through all of the basic btrfs commands that manipulate our
subvolume stuff. It creates a snapshot, a subvolume, sets the subvolume as a
default, lists the volumes and deletes the snapshot. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
ltp/fsx.c tests whether the filesystem it is run on supports fallocate.
If it is not supported the fsx will print warning to stderr. This leads
to fails of tests 075, 112, 127 for the filesystems that do not support
fallocate. The tests use ltp/fsx but do not filter out stderr. Since
ltp/fsx.c can work without fallocate support I propose to move this
message to stdout unless quiet output is not requested. Previous patch
printed the message even if -q flag was used. This patch honours the flag.
This simple patch fixes the issue for me, tested on all the mentioned tests:
Signed-off-by: Boris Ranto <branto@redhat.com>
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Because of udevs complaining device mapper now creates /dev/dm-N as the real
device nodes, and just symlinks the /dev/mapper/ names to it. This would be
easy if everything used the /dev/mapper clear names, but most system utilities
translate them back to the /dev/mapper/ names and thus confuse various test
cases. Add support to _is_block_dev to read symlinks, and add documentation
on how to run xfstests on device mapper volumes.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
When I added fallocate support to fsx I inadvertently added
a duplicate fallocate test.
Consolidate them both into one test (the link test, not the
compile test) and make all tests use "true" rather than "yes"
to be more consistent with other tests.
Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
Reviewed-by: Alex Elder <aelder@sgi.com>
I don't know why, but discard tests is 251 in xfs, but 251.out has
number of 248 in it, So it fails. Change it to 251 now.
Cc: Lukas Czerner <lczerner@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Alex Elder <aelder@sgi.com>
Signed-off-by: Tao Ma <boyu.mt@taobao.com>
Signed-off-by: Alex Elder <aelder@sgi.com>
This patch creates a test that exercises xfs_metadump, with a focus
on its obfuscation of names. It was created to verify fixes that
avoided a hang condition when running "xfs_metadump" on a directory
containing files having particular bit patterns in their name.
Arkadiusz MiÅkiewicz first reported seeing this while attempting
to create a metadump for a filesystem containing a file named
"R\323\257NE".
For now this script checks the following (using only filenames, not
attributes):
- that short names (4 characters or less) aren't obfuscated
- that long names get obfuscated
- that (long) directory names get obfuscated
- that names that are known to produce bit patterns that lead
to invalid path components still generate obfuscated names
(this could previously lead to a hang)
- that many names of the same length can still generate new
obfuscated names (this could previously lead to a hang)
- that neither "lost+found" nor orphaned files stored in it ge
obfuscated
Right now there are two sets of "ls" commands executed (one before
and one after obfuscation). This produces repeatable results for
me on one filesystem, but on a different filesystem I expect the
inode numbers to change (and random number generation might change
the output too). I'm interested in suggestions on how to filter
the output so the results can be verified. If nothing else, the
test serves its purpose if I simply comment out those commands,
and will do that if there's not a better suggstion.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
The specific set of extent sizes allocated to a file is not always
deterministic. In particular, sometimes a range of unwritten blocks
is covered by a single extent, while in other cases it might be
represented by multiple consecutive unwritten extents. This can
result in spurious errors being reported in tests that check file
extent maps.
Add a filter that finds adjacent extents in what gets produced for
fiemap and bmap output and coalesces them as if all consective
extents of the same time were really just one extent. (Note that
as implemented here this applies to all extent types, not just
unwritten extents.)
Update the golden output for test 242 to reflect the change.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
By convention all of the tests in the xfstests suite have mode 0755.
A few have recently committed without the execute bit set.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Bill Kendall's latest update to xfsdump dropped a line of output
that served no real purpose. This change updates the golden output
for the two tests that included that line of output.
Updated to filter out that line from dump output also, so old
versions of the code will still produce the same output (suggested
by Dave Chinner).
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Eliminate build warnings reported on files located under
dmapi/src/suite2/src/.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Eliminate build warnings reported on files located under
dmapi/src/suite1/cmd/ (third of three).
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Eliminate build warnings reported on files located under
dmapi/src/suite1/cmd/ (second of three).
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Eliminate build warnings reported on files located under
dmapi/src/suite1/cmd/ (first of three).
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Eliminate build warnings reported on files located under these
subdirectories:
dmapi/src/common/cmd/
dmapi/src/common/lib/
dmapi/src/sample_hsm/
dmapi/src/simple/
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Move the major test meat of 242 into common.punch, and reuse it for
a new testcase the exercises the new fallocate implementation
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Once we get one short write, we know the test has failed. We don't
ened to keep writing and getting more short writes and spewing
thousands of errors to the console.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Recent kernels allow more than 40 nested symlinks, so up the limit
to still reproduce a failure.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
219 and 235 were failing on ext3 with selinux, because the extra
xattrs upped the quota usage.
Do the same trick we've used in the past to mount with an selinux
global context when doing quota tests.
Note the gross hack for remount,ro, though.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
Test 249 was appearing to pass on ext4, but it wasn't really
exercising the test due to lack of "-F" in the xfs_io arguments.
Without -F the files were created (oddly enough); neither pwrite
nor sendfile were executed, and the diff of the two (empty)
files passed, resulting in a passed test without testing anything.
So add the -F, capture the output, and test the result of each
xfs_io invocation.
Also, when it fails, the diff output is huge. Make diff silent,
but describe the diff failure and exit.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
The "dmapi" subtree was developed separate from and sort of wedged
into the rest of the "xfstests" code. As a result, it has a lot of
build infrastructure that's just different from the unified way used
for everything else.
This patch changes all that, making the "dmapi" subtree be a more
normal component of "xfstests" with respect to its build process.
This involves removing all the cruft needed and used by the dmapi
"configure" script, and replacing each "Makefile.am" file with a
proper "Makefile" that includes a simple set of rules that are
compatible with the broader "xfstests" build.
The result is a much cleaner, consistent build. It also deletes
a considerable amount of code.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Make it so "make depend" is a generic target, like "make clean".
Each Makefile has a "depend" target that indicates whether making
dependencies means creating ".dep" or creating ".ltdep" (or, I
suppose, both, though none do that right now). Both files get
created even if there are no CFILES to scan (to ensure the target
up-to-date). The "default" target now depends on "depend" (there is
no "ltdepend" any more).
Remove the "depend" and "ltdepend" definitions from the "buildrules"
file; only the actual generated files (".dep" and ".ltdep") remain
as generic targets. The "depend' target is still defined as phony.
Do a shell trick when expanding the value of CFILES, to avoid a
problem that occurs if it is created by "make" by concatentating two
empty strings. The problem was that in that case CFILES will
contain a space, and that wasn't getting treated as empty as
desired.
Make the rule for tool/lib dependencies more generic, to reflect the
general desire that "lib" subdirectories need to be built before
things in the "tool" subdirectories.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Change the top-level Makefile, to make it clearer just what makes
what, and what depends on what:
- Separate the rules for "configure" and "include/builddefs" into
two parts, each of which generate one of the files
- Get rid of the rule for include/config.h, and group it with the
one for include/builddefs (the same command creates both files)
Having done this, we find that having both "include/builddefs" and
"include/config.h" as dependencies for the default target results in
a parallel invocation of "make" spawning two concurrent attempts to
do the configure step--and that doesn't work.
Creating one of those two will result in the other getting created,
so just list one of them as a dependency for the default rule.
A couple of other small fixes:
- Get rid of the "new", "remake" and "check" dependencies for the
default rule, which serv no purpose
- Use the $(Q) convention in a few missed spots
- Stop a DMAPI-only comment from getting echoed on default build
- Delete the "
This updated version pulls in the content of a patch previously
posted separately to fix the problem with parallel builds.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Eric Sandeen <sandeen@sandeen.net>
This patch fixes a few build warnings. I have built the code using
i386, x86_64, and ia64 architectures and each has ends up with
complaints of one sort or anther. This gets rid of all of them
*except* those reported by files under the "ltp" (Linux Test
Project) sub-tree.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Eric Sandeen <sandeen@sandeen.net>
Tests 127 and 134 leave temp files around when they complete.
Fix (or enable) their cleanup functions to remedy this.
Signed-off-by: Alex Elder <aelder@sgi.com>
Reviewed-by: Eric Sandeen <sandeen@sandeen.net>
Amir Goldstein
Josef Bacik
Boris Ranto
Christoph Hellwig
Eric Sandeen
Tao Ma
Alex Elder
Dave Chinner