Git Log


Commit Date  
[6ccf55] by Garrett Cooper Garrett Cooper

Merge branch 'master' of ssh://ltp.git.sourceforge.net/gitroot/ltp/ltp

2011-10-20 17:03:14 Tree
[cbf234] by Garrett Cooper Garrett Cooper

Fix a typo and update some of the instructions.

Signed-off-by: Garrett Cooper <yanegomi@gmail.com>

2011-10-20 17:02:25 Tree
[ab837b] by Jan Stancek Jan Stancek , pushed by Cyril Hrubis Cyril Hrubis

cgroup/cpuctl: clean cpuctl_task_* in case of OOM

If one of cpuctl_task_* gets killed by OOM, it won't
signal run script with SIGUSR1, and trap action
"killall cpuctl_task_*;" never runs.

Try to kill all left over cpuctl_task_* in cleanup() to be sure,
none are left behind.

Signed-off-by: Jan Stancek <jstancek@redhat.com>

2011-09-19 08:57:28 Tree
[fc0986] by Peng Haitao Peng Haitao , pushed by Cyril Hrubis Cyril Hrubis

cron_deny01: fix error of minute over 60

Change to fix hour error when the addition of 2 minutes over 60.

Signed-off-by: Peng Haitao <penght@cn.fujitsu.com>
Signed-off-by: Yang Dongqing <yangdongqing@cn.fujitsu.com>

2011-10-14 02:43:52 Tree
[775fbe] by Jan Stancek Jan Stancek , pushed by Cyril Hrubis Cyril Hrubis

cgroup/memcg_regression_test: treat failed mount as error

If test fails to mount memory subsytem, it should exit
with non-zero code.

Signed-off-by: Jan Stancek <jstancek@redhat.com>

2011-09-19 07:49:53 Tree
[21fcf8] by Jan Stancek Jan Stancek , pushed by Cyril Hrubis Cyril Hrubis

cgroup/memcg_regression_test: fix test4 cleanup

If test_4.sh gets killed, it won't clean cgroup it created,
so clean it here.

Signed-off-by: Jan Stancek <jstancek@redhat.com>

2011-09-19 07:49:47 Tree
[81b12e] by Jan Stancek Jan Stancek , pushed by Cyril Hrubis Cyril Hrubis

cgroup/memcg_regression_test: clear dmesg before test

dmesg can rotate and number of found bugs can actually go down,
so clear the buffer before test to avoid this.

Signed-off-by: Jan Stancek <jstancek@redhat.com>

2011-09-19 07:49:40 Tree
[489fc0] by Jan Stancek Jan Stancek , pushed by Cyril Hrubis Cyril Hrubis

cpuset_base_ops_testset.sh: drop sched_relax_domain_level test

At the moment we don't have reliable way to enumerate
all possible values, which can be used. They are dependent
on both kernel version and kernel config.

Based on previous discussion here:
http://article.gmane.org/gmane.linux.ltp/14146/match=sched_relax_domain_level
it seems better to get rid of this testcase, rather than testing
internal knobs, which can change at anytime.

Signed-off-by: Jan Stancek <jstancek@redhat.com>

2011-09-19 07:08:30 Tree
[c9dd96] by Andrew Chen Andrew Chen , pushed by Cyril Hrubis Cyril Hrubis

runtest/fs: change parameters of growfiles

gf16/gf17 create 10 files and keep them growing and shrinking.
But when the number of created file (-S) can be divided exactly
by the rate shrinkfile() is involked (-T), the shrinking will
always happen on the same file. This may cause the tests run out
of memory and unexpected OOM killer be launched.

The amount of consumed memory depends on the speed of CPU/Bus.
In our machine, CPU 400 MHz/Bus 100 MHz, about 240 MB free memory
was consumed before the test finished.

To prevent this, we adjusts the shrinkfile() rate (-T) slightly
to ensure it cannot be divided by the number of created file (-S).

Signed-off-by: Andrew Chen <ypchen@faraday-tech.com>

2011-09-19 05:54:05 Tree
[e9d874] by Caspar Zhang Caspar Zhang , pushed by Cyril Hrubis Cyril Hrubis

mm/ksm: more fixes when ksm config not enabled

This patch follows up Nageswara R Sastry's fix, which checks whether ksm
configuration is enabled in ksm NUMA cases (ksm02,ksm04).

Signed-off-by: Caspar Zhang <czhang@redhat.com>

2011-09-15 08:17:41 Tree
[e53335] by Caspar Zhang Caspar Zhang , pushed by Cyril Hrubis Cyril Hrubis

mm/oom: tiny fixes during code review

This patch fixes some issues during code review. One is that there
would be a potential failure if path /sys/devices/system/node does not
exist, which might cause oom02/04 fail. The other one is moving
tst_exit() from cleanup() to main(), this should make us get rid of a
compile warning.

Signed-off-by: Caspar Zhang <czhang@redhat.com>

2011-09-15 08:17:40 Tree
[fb0051] by Caspar Zhang Caspar Zhang , pushed by Cyril Hrubis Cyril Hrubis

mm/vma01: remove useless mmap

addr v has been proved unless in test case, so remove it in this patch.
Also a tst_resm/printf would call mmap() which might cause noise in this
test, remove it as well.

Signed-off-by: Caspar Zhang <caspar@casparzhang.com>

2011-09-27 09:27:18 Tree
[13fd84] by Cyril Hrubis Cyril Hrubis

Fix kdump/runkdump.sh

The shell was looking for runkdump.conf in $PATH rather
than in current directory where the file actually is.

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

2011-10-13 12:33:09 Tree
[811fd8] by Peng Haitao Peng Haitao , pushed by Cyril Hrubis Cyril Hrubis

ldd: fix error with -m32 in x86_64

When making with -m32 in x86_64 kernel, the following error messages may be outputted:
/usr/bin/ld: i386 architecture of input file `/root/ltp-dev/testcases/commands/ade/ldd/lddfile1.obj' is incompatible with i386:x86-64 output
collect2: ld returned 1 exit status

The following patch can fix this.

Signed-off-by: Peng Haitao <penght@cn.fujitsu.com>

2011-09-29 02:32:35 Tree
[2fbbc9] by Nageswara R Sastry Nageswara R Sastry , pushed by Cyril Hrubis Cyril Hrubis

KSM test case fixes - mmap10

Disable mmap10 test if KSM is not configured

P.S. patch checked with 'checkpatch.pl'

# ./mmap10 -s
mmap10 1 TCONF : KSM configuration is not enabled
mmap10 2 TCONF : Remaining cases not appropriate for configuration

Signed-off-by: Nageswara R Sastry <rnsastry@linux.vnet.ibm.com>

--
Regards
R.Nageswara Sastry

2011-09-12 11:55:19 Tree
[786e3a] by Dmitry Artamonow Dmitry Artamonow , pushed by Cyril Hrubis Cyril Hrubis

fs/proc01: namespaces fix for kernel >= 3.0

Add exclusion for /proc/self/.../ns/{net,uts,ipc} files introduced
in kernel 3.0. These files aren't meant to be read and merely
used for getting filedescriptor as a handle to corresponding namespace.
Any file operation on them including read() returns -EINVAL.

Signed-off-by: Dmitry Artamonow <mad_soft@inbox.ru>

2011-09-30 12:32:50 Tree
[cfb676] by Michal Simek Michal Simek , pushed by Cyril Hrubis Cyril Hrubis

creat08: Fix test to be able to run on NFS

NFS creates temporary .nfsXXX files for every open/created file.
Based on this rmdir syscall failed because directory is not empty.
The solution is to close all open/created files. They are also unlink
in cleanup.

Error log:
creat08 5 TBROK : rmdir testdir.A.167 failed: errno=???(39): Directory not empty
creat08 6 TBROK : Remaining cases broken

Example of created files(listed by readdir):
testdir.A.155
.nfs00000000070201f60000007d
..
.nfs00000000070201f70000007c
.

testdir.B.155
.nfs00000000000dc13c0000007f
.nfs00000000000dc1050000007e
..
.
.nfs00000000000dc0f600000080

Signed-off-by: Michal Simek <monstr@monstr.eu>

2011-09-15 11:09:02 Tree
[0cc751] by Nageswara R Sastry Nageswara R Sastry , pushed by Cyril Hrubis Cyril Hrubis

readlink03 test case fix

In readlink03 testcase,
When 'Symlink Pathname is empty' the return value was 'ENOENT' but with
newer kernels the return value is 'EINVAL'. The following patch will fix
this.

readlink03:
Before patch:
...
readlink03 9 TFAIL : readlink() failed, Symlink Pathname is empty,
errno=22, expected errno=2
...

After patch:
...
readlink03 9 TPASS : readlink(), Symlink Pathname is empty,
returned errno 22
readlink03 10 TPASS : readlink(), Symlink Pathname is empty,
returned errno 22
...

Signed-off-by: Nageswara R Sastry <rnsastry@linux.vnet.ibm.com>
Tested-by: Michal Simek <michal.simek@petalogix.com>

2011-09-12 11:47:27 Tree
[53053d] by Nageswara R Sastry Nageswara R Sastry , pushed by Cyril Hrubis Cyril Hrubis

sockioctl01 test case fix

In sockioctl01 testcase,
Current 'inappropriate ioctl operation returning EINVAL'
With the git commit '41c31f318a5209922d051e293c61e4724daad11c' more
details at http://permalink.gmane.org/gmane.linux.kernel.commits.head/296060
'inappropriate ioctl operation returning ENOTTY'

The following patch will fix this.

sockioctl01:
Before patch:
...
sockioctl01 4 TFAIL : ATMARK on UDP ; returned -1 (expected -1),
errno 25 (expected 22)
...

After patch:
...
sockioctl01 4 TPASS : ATMARK on UDP successful
...

Signed-off-by: Nageswara R Sastry <rnsastry@linux.vnet.ibm.com>
Tested-by: Michal Simek <michal.simek@petalogix.com>

2011-09-12 11:45:40 Tree
[313837] (20110915) by Cyril Hrubis Cyril Hrubis

LTP 20110915

2011-09-27 11:51:50 Tree
[2c321c] by Nageswara R Sastry Nageswara R Sastry , pushed by Cyril Hrubis Cyril Hrubis

KSM test case fixes - oom01 and oom03

Disable oom01, oom03 tests if KSM is not configured

Signed-off-by: Nageswara R Sastry <rnsastry@linux.vnet.ibm.com>

2011-09-08 09:53:45 Tree
[bca9a8] by Nageswara R Sastry Nageswara R Sastry , pushed by Cyril Hrubis Cyril Hrubis

KSM test case fixes - ksm03

Disable ksm03 test if KSM is not configured

Signed-off-by: Nageswara R Sastry <rnsastry@linux.vnet.ibm.com>

2011-09-08 09:51:03 Tree
[c0299b] by Nageswara R Sastry Nageswara R Sastry , pushed by Cyril Hrubis Cyril Hrubis

KSM test case fixes - ksm01

Disable ksm01 test if KSM is not configured

Signed-off-by: Nageswara R Sastry <rnsastry@linux.vnet.ibm.com>

2011-09-08 09:50:32 Tree
[89f480] by Peng Haitao Peng Haitao , pushed by Cyril Hrubis Cyril Hrubis

pthread_create/10-1: correct return value if test fails

This patch fixes return value always being 0 even if test
fails. Using _exit(value) instead of pthread_exit();.

Signed-off-by: Peng Haitao <penght@cn.fujitsu.com>

2011-09-06 08:55:47 Tree
[0bf1a7] by Peter W Morreale Peter W Morreale , pushed by Cyril Hrubis Cyril Hrubis

Apply coding style to timer_gettime 1-3

Fix coding style issues.

Signed-off-by: Peter W. Morreale <pmorreale@novell.com>

2011-08-30 13:17:56 Tree
Older >

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:





No, thanks