Work at SourceForge, help us to make it a better place! We have an immediate need for a Support Technician in our San Francisco or Denver office.

Close

Git Log


Commit Date  
[62a7ab] (20130503) by Shubham Shubham

LTP 20130503

2013-05-03 06:42:13 Tree
[64ae94] by Cyril Hrubis Cyril Hrubis

openposix/...pthread_mutexattr_gettype/speculative/3-1

The testcase tests for optional behavior. On linux the function just
returns member of the passed attr structure and as a such cannot fail.

This commit changes the test to return UNTESTED when the call succeeded.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-05-02 16:16:57 Tree
[64f445] by Cyril Hrubis Cyril Hrubis

openposix/.../pthread_key_create/speculative/5-1.c

* The pthread_key_create was called twice in the loop
no wonder it failed with EAGAIN before the loop has
reached the limit.

* Cleaned up the code a litte.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-05-02 13:54:56 Tree
[f2ec5c] by Cyril Hrubis Cyril Hrubis

mem/thp03: Return TCONF if KSM is not supported

The madvise() with MADV_MERGEABLE will return EINVAL when kernel is not
compiled with KSM. Make the test return TCONF rather than TBROK in this
case.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-30 14:32:13 Tree
[d27f4b] by Cyril Hrubis Cyril Hrubis

mem/ksm05: Add check for KSM sysfs existence.

Add check (the same as in ksm01) for the existence of KSM sysfs files.
Without it the testcase prints 665 lines of TBROK and then Segfaults on
systems without KSM support.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-30 14:03:30 Tree
[25f426] by Cyril Hrubis Cyril Hrubis

syscalls/accept4: Fix test when accept4 returns ENOSYS

* Make the test actually exit with TCONF when accept4 returned ENOSYS

* Remove ifdefs around the main as due tu linux_syscall_numbers.h
both __NR_accept4 and __NR_socketcall are always defined.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-30 11:00:02 Tree
[be0fe2] by Cyril Hrubis Cyril Hrubis

openposix/.../aio_fsync/5-1.c: Fixes.

* Don't use ret variable to store return value from aio_error() as this
overwrites the value that should be returned (coincidentally both
PTS_PASS and return from aio_error() when I/O has finised are zero).

* Print "Test PASSED" when test has passed.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-29 16:45:12 Tree
[db2511] by Cyril Hrubis Cyril Hrubis

openposix/.../aio_fsync/2-1,3-1: Add "Test PASSED"

Make the tests to print "Test PASSED" when the test has passed.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-29 16:15:41 Tree
[a25f89] by Mike Frysinger Mike Frysinger

adjust copyright/license on a few files to match reality

From the author:
I looked at the four files in question. I did write getitimer03.c and I
believe using the standard IBM copyright is correct.

I ported the getresgid01.c code to run under Linux. As I recall it was
originally a Sequent test case for the Dynix/ptx OS. Again I believe
using the standard IBM copyright is correct.

I did not work with the last two files, but they follow the same pattern
as the getresgid01.c code. There were originally part of the Dynix/ptx
test cases.

Also note: the Makefile in question was rewritten by Cisco as part of
the build system overhaul, so use the standard copyright for that.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

2013-04-27 06:35:01 Tree
[f1b9dd] by Mike Frysinger Mike Frysinger

kernel: include: regen linux_syscall_numbers.h

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

2013-04-27 06:27:46 Tree
[a0359f] by Mike Frysinger Mike Frysinger

kernel: include: fix dependency on regen script and input files

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

2013-04-27 06:15:44 Tree
[bce0f5] by Mike Frysinger Mike Frysinger

kernel: include: generate headers in parallel

Signed-off-by: Mike Frysinger <vapier@gentoo.org>

2013-04-27 06:12:42 Tree
[b3e9e5] by Wanlong Gao Wanlong Gao

access05: code cleanups

Cleanup the code.

Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>

2013-04-26 07:40:47 Tree
[330f4e] by Wanlong Gao Wanlong Gao

access04: code cleanups

Cleanup the code.

Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>

2013-04-26 07:40:47 Tree
[41dae8] by Wanlong Gao Wanlong Gao

access03: code cleanups

Cleanup the code.

Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>

2013-04-26 07:40:46 Tree
[332fe2] by Wanlong Gao Wanlong Gao

access02: code cleanups

Cleanup the code.

Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>

2013-04-26 07:40:46 Tree
[f85fea] by Wanlong Gao Wanlong Gao

access01: code cleanups

Cleanup the code.

Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>

2013-04-26 07:40:46 Tree
[b66220] by Cyril Hrubis Cyril Hrubis

Merge ../ltp-dev

2013-04-25 10:42:04 Tree
[cd5812] by Cyril Hrubis Cyril Hrubis

openposix/stress/mqueues: Rename args to match.

This has been broken by 3a48c063457d2bc61605fdcfc9f30a228b2b19c0

As the binary names were renamed, the arg files no longer match. Fix
this by renaming the args files to match again.

Reported-by: Jan Stancek <jstancek@redhat.com>
Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-25 10:32:09 Tree
[865afa] by DAN LI DAN LI , pushed by Wanlong Gao Wanlong Gao

syscalls/getitimer01.c: Test for getting ITIMER_VIRTUAL and ITIMER_PROF

Additional tests of getting ITIMER_VIRTUAL and ITIMER_PROF.

Signed-off-by: DAN LI <li.dan@cn.fujitsu.com>
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>

2013-04-25 05:44:57 Tree
[dc8c2f] by DAN LI DAN LI , pushed by Wanlong Gao Wanlong Gao

syscalls/getitimer: cleanup

1. Remove redundant comments

2. Revise code to follow ltp-code-style

Signed-off-by: DAN LI <li.dan@cn.fujitsu.com>
Signed-off-by: Wanlong Gao <gaowanlong@cn.fujitsu.com>

2013-04-25 05:44:56 Tree
[db7f43] by Cyril Hrubis Cyril Hrubis

fs/proc01.c: Add known issue.

Since 786d7e1612f0b0adb6046f19b906609e4fe8b1ba upstream kernel commit
the proc operations on regular files are done via proc_reg_*() proxies
so the default errno for unimplemented read on proc file changed to EIO
returned from the proxy rather than EINVAL that was previously returned
from vfs_read().

The only affected known issue I've found is the '/proc/xen/privcmd'
file which makes the testcase fail on certain xen installations.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-24 15:05:54 Tree
[3dfde4] by Cyril Hrubis Cyril Hrubis

openposix/.../mmap/14-1: Fix cleanup.

The test wasn't unlinking the temporary file on exit.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-24 12:46:03 Tree
[8e9b99] by Cyril Hrubis Cyril Hrubis

syscalls/mincore*: Fixes.

* Make use of tst_tmpdir() to create temporary files

* Add option parsing and looping to mincore02

* Changed mmap() first parameter from 0 to NULL

* etc.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-24 12:24:25 Tree
[8e3afe] by Cyril Hrubis Cyril Hrubis

syscalls/micore*: Fix.

Remove the UCLINUX ifdefs, add mincore directory
into UCLINUX FILTER_OUT_DIRS in syscalls Makefile.

Signed-off-by: Cyril Hrubis <chrubis@suse.cz>

2013-04-24 12:24:20 Tree
Older >