Hi,
I have checked Henryś patch for rt_sigqueueinfo on my system
(Linux-2.6.29.1), It seems to be working fine.
I'm not sure about the parent- child syncronization that Garrett is
talking about, haven't checked it thoroughly.
On Fri, Aug 7, 2009 at 6:11 PM, Subrata Modak wrote:
> On Thu, 2009-08
On Thu, Aug 13, 2009 at 10:54 AM, Paul Larson wrote:
> You were not providing the full path. Here's a new version that should
> let you either provide absolute or relative path (including no patch
> assuming the file is in LTPROOT). See if this works better for you.
>
>
> Add a -S option to runlt
n Thu, Aug 13, 2009 at 10:18 AM, Subrata
Modak wrote:
> Hey,
>
> Thanks for sending that update once again. But still issues remain to be
> addressed:
>
> 1) # ls -l testcases/kernel/controllers/cpuacct/
> total 72
> -rwxr-xr-x 1 root root 4903 Aug 13 23:08 cpuacct_task
> -rw-r--r-- 1 root root
You were not providing the full path. Here's a new version that should
let you either provide absolute or relative path (including no patch
assuming the file is in LTPROOT). See if this works better for you.
Add a -S option to runltp that allows the user to specify a SKIPFILE of
tests to skip.
Hey,
Thanks for sending that update once again. But still issues remain to be
addressed:
1) # ls -l testcases/kernel/controllers/cpuacct/
total 72
-rwxr-xr-x 1 root root 4903 Aug 13 23:08 cpuacct_task
-rw-r--r-- 1 root root 121 Aug 13 23:05 cpuacct_task.c
-rw-r--r-- 1 root root 1665 Aug 13 23
Hi Paul,
On Wed, 2009-08-12 at 16:50 -0500, Paul Larson wrote:
> One problem that LTP has long suffered is that certain tests are known
> to be broken, sometimes for long periods of time. In an ideal world, I
> think it best that these should be skipped, or at least removed from the
> default li
Mike Frysinger wrote:
> On Friday 31 July 2009 17:29:55 Darren Hart wrote:
>> Mike Frysinger wrote:
>>> are there any futex syscall tests already in ltp that i cant seem to
>>> locate ? if not, i can adapt the few that already exist in the libfutex
>>> code ...
>> I recently done a lot of cleanup a
On Thursday 30 July 2009 05:57:54 Jiri Palecek wrote:
> the commit "extend the test result to a bit field so we can extend the
> output further" from July 20th unfortunately changes the code, that was
> previously a bitfield, to a non-bitfield. This causes FAILED tests to
> return erroneously 0.
>
On Thursday 13 August 2009 05:04:13 Garrett Cooper wrote:
> On Thu, Aug 13, 2009 at 12:46 AM, Mike Frysinger wrote:
> > On Tuesday 04 August 2009 04:48:25 Michal Simek wrote:
> >> I am thinking that will be good to extend return flag for function not
> >> implemented (errno 38) because
> >> some of
On Friday 31 July 2009 17:29:55 Darren Hart wrote:
> Mike Frysinger wrote:
> > are there any futex syscall tests already in ltp that i cant seem to
> > locate ? if not, i can adapt the few that already exist in the libfutex
> > code ...
>
> I recently done a lot of cleanup and a couple feature addi
On Thu, Aug 13, 2009 at 12:46 AM, Mike Frysinger wrote:
> On Tuesday 04 August 2009 04:48:25 Michal Simek wrote:
>> I am thinking that will be good to extend return flag for function not
>> implemented (errno 38) because
>> some of tests return it when you don't have enable that option is kernel
>>
On Wed, Aug 12, 2009 at 10:36 PM, Mike Frysinger wrote:
> On Sunday 19 July 2009 00:38:39 Garrett Cooper wrote:
>> int errnochoose(void){ //choose the relative errno
>>
>> switch (TEST_ERRNO){
>> - case 0: tst_exit();
>> - case ECHILD: strerror((int)"ECHILD");
>> -
Hi Wang Yong,
I have observed the following thing in case02 of the test case
utimes01 in latest ltp release.
Expected error value is EACCES (In the code below)
But the actual error value is EPERM.I have checked this on 2.6.27 and
2.6.29 kernel.
static struct test_case tcase[] = {
{ //
Hi Subrata, I have done the changes to the patch as you have
suggested, please review it and let me know if it needs any
changes thanku.
I am psoting the results also here.
Signed-off-by: Duddu Rajasekhar
diff -rupN
ltp-full-20090731//testcases/kernel/controllers/cpuacct/cpuacct_task.c
ltp-ful
On Monday 03 August 2009 11:14:49 Serge E. Hallyn wrote:
> Quoting Michal Simek (michal.si...@petalogix.com):
> > can you please to run clone01 syscall test on any x86 machine? I am
> > getting fault there when I run it 10 times for example.
> > The same problem I have on Microblaze.
> >
> > ./clo
On Tuesday 04 August 2009 04:48:25 Michal Simek wrote:
> I am thinking that will be good to extend return flag for function not
> implemented (errno 38) because
> some of tests return it when you don't have enable that option is kernel
> config.
> What do you think?
what "return flag" are you talk
On Friday 31 July 2009 06:03:41 Garrett Cooper wrote:
> After installing the time utility (I was a bit amazed it wasn't
> installed by default on Gentoo), and fixing up a few minor Makefile
> items, I've gotten to the point where only the networking and selinux
> items are noticeably missing...
On Monday 03 August 2009 02:32:59 Garrett Cooper wrote:
> export SHELL_DEBUG=${SHELL_DEBUG:=0}
does this really need to be exported ? otherwise, the simpler way to write a
default value is:
: ${SHELL_DEBUG:=0}
there are a bunch of places this can be fixed in this file
> exists()
> {
> for
On Friday 31 July 2009 10:35:44 Serge E. Hallyn wrote:
> Quoting Garrett Cooper (yaneg...@gmail.com):
> > clone(2) is externally defined in sched.h, and as such
> > testcases/kernel/security/digsig/writeexec/libwritetest.c fails to
> > compile with -Wall due to a shadowed declaration.
> >
> > Signe
On Thursday 06 August 2009 20:56:58 Henry Yei wrote:
> Michal,
please do not top post
> This is a known issue. If you look at the manpages it states the following
> under the BUGS section:
>
> Linux is not careful to distinguish between the EACCES and EPERM
> error returns. On the other hand
On Friday 07 August 2009 04:49:45 Geert Uytterhoeven wrote:
> Commit "move leading __NR_ to script to make all the .in files simpler"
> accidentally removed all plus signs surrounded by spaces from the syscall
> definition files for hppa/powerpc{,64}/s390{,x}/sh/sparc{,64}, breaking the
> build lik
On Monday 10 August 2009 11:44:57 Garrett Cooper wrote:
> On Mon, Aug 10, 2009 at 1:29 AM, srikanth krishnakar wrote:
> > On Mon, Aug 10, 2009 at 1:50 PM, Subrata Modak wrote:
> >> On Fri, 2009-08-07 at 20:56 +0530, srikanth krishnakar wrote:
> >> > I see these two test cases are failing on PowerPC
On Tuesday 11 August 2009 13:31:01 Subrata Modak wrote:
> + if [ $INJECT_KERNEL_FAULT -eq 1 ] ; then
> + #See if Debugfs is mounted, and
> + #Fault Injection Framework available through Debugfs
> + if [ -d "/debug/fail_io_timeout" -o \
> +
On Wednesday 12 August 2009 07:21:00 Subrata Modak wrote:
> On Wed, 2009-08-12 at 01:42 -0700, Garrett Cooper wrote:
> > On Tue, Aug 11, 2009 at 10:30 AM, Subrata Modak wrote:
> > > At the heart of this infrastructure is this Script, which will
> > > actually: 1) Change the temporary command file g
On Wednesday 12 August 2009 17:13:11 Paul Larson wrote:
> If you run file_test.sh on a system without rpm installed, it should
> return tconf instead of tfail.
this has the same problems as i mentioned with your mail_tests patch
-mike
signature.asc
Description: This is a digitally signed message
On Wednesday 12 August 2009 17:14:09 Paul Larson wrote:
> +`which mail`
> +if [ $? != 0 ]; then
this results in pretty ugly output, and the return value/output of `which` is
not portable. use the type builtin instead.
> +MAIL_NOT_INSTALLED=1
you should make sure this is always set/unset at
26 matches
Mail list logo
