Bug 16161 - AB-INT: ltp test_posix_groups times out
Summary: AB-INT: ltp test_posix_groups times out
Status: NEW
Alias: None
Product: Functional (self) Testing
Classification: QA/Testing
Component: oe-selftest (show other bugs)
Version: unspecified
Hardware: x86 Multiple
: Medium normal
Target Milestone: 6.2
Assignee: Mathieu Dubois-Briand
QA Contact:
URL:
Whiteboard: AB-INT
Depends on:
Blocks:
 
Reported: 2026-02-05 17:23 UTC by Mathieu Dubois-Briand
Modified: 2026-11-02 13:12 UTC (History)
2 users (show)

See Also:
OS type for building Yocto: ---
Type of Regression: ---
Verified:
Documentation change: No (bug/feature does not impact docs)


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Mathieu Dubois-Briand 2026-02-05 17:23:03 UTC
Forking this off https://bugzilla.yoctoproject.org/show_bug.cgi?id=15999

We have seen ltp test_posix_groups failing on a regular basis, with the following trace:

  File "/srv/pokybuild/yocto-worker/qemux86-64-ltp/build/layers/openembedded-core/meta/lib/oeqa/runtime/cases/ltp_compliance.py", line 97, in test_posix_groups
    self.runltp(posix_group)
  File "/srv/pokybuild/yocto-worker/qemux86-64-ltp/build/layers/openembedded-core/meta/lib/oeqa/runtime/cases/ltp_compliance.py", line 65, in runltp
    (status, output) = self.target.run(cmd)
                       ^^^^^^^^^^^^^^^^^^^^
...
AssertionError: ssh exited with status '255' for command '['ssh', '-l', 'root', '-o', 'ServerAliveCountMax=2', '-o', 'ServerAliveInterval=30', '-o', 'UserKnownHostsFile=/dev/null', '-o', 'StrictHostKeyChecking=no', '-o', 'LogLevel=ERROR', '192.168.7.2', 'export PATH=/usr/sbin:/sbin:/usr/bin:/bin; /opt/ltp/bin/run-posix-option-group-test.sh SIG 2>@1 | tee /opt/ltp/results/SIG']': this is likely an SSH failure
Executing signals tests

The whole status is in the other ticket, mixed with other kinda unrelated issues.

Quick summary:
- It's unlikely to be a memory issue, even if the OOM killer triggers at some point: it also triggers in succeeding builds.
- It might be a test taking a bit too much time, but I'm not really convinced: I added a patch in my branch extending the timeout.

Looking at the log and trying to match it with the test sequence, I believe open_posix_testsuite/conformance/interfaces/sigwait/6-1.c test is running when the timeout occurs. So I'm also adding a second patch to my branch, adding a bit of debug messages.
Comment 1 Mathieu Dubois-Briand 2026-02-27 11:05:33 UTC
qemux86-64-ltp fedora40-vk-1 master completed at 2026-02-27 10:40:01+00:00
https://autobuilder.yoctoproject.org/valkyrie/#/builders/63/builds/3070/steps/13/logs/stdio
Comment 2 Mathieu Dubois-Briand 2026-03-23 08:55:31 UTC
This one is a bit different, but still in test_posix_groups. LTP was recently
upgraded, maybe the behaviour is a little bit different.

qemux86-64-ltp rocky9-vk-2 master&master-next completed at 2026-03-20 17:52:06+00:00
https://autobuilder.yoctoproject.org/valkyrie/#/builders/63/builds/3239/steps/13/logs/stdio
Comment 3 Mathieu Dubois-Briand 2026-04-07 08:45:35 UTC
This one has some extra logs, I need to analyze this a bit later.

qemux86-64-ltp ubuntu2404-vk-3 mathieu/master-next completed at 2026-04-03 19:18:23+00:00
https://autobuilder.yoctoproject.org/valkyrie/#/builders/63/builds/3333/steps/13/logs/stdio

qemux86-64-ltp debian12-vk-7 master-next completed at 2026-04-05 09:51:18+00:00
https://autobuilder.yoctoproject.org/valkyrie/#/builders/63/builds/3341/steps/13/logs/stdio
Comment 4 Mathieu Dubois-Briand 2026-04-07 14:32:35 UTC
(In reply to Mathieu Dubois-Briand from comment #3)
> This one has some extra logs, I need to analyze this a bit later.
> 

Logs do not have expected lines, I'm modifying the patch in my branch to add more aggressive logs.
Comment 5 Mathieu Dubois-Briand 2026-05-15 08:53:20 UTC
qemux86-64-ltp ubuntu2204-vk-2 mathieu/master-next completed at 2026-05-06 13:43:58+00:00
https://autobuilder.yoctoproject.org/valkyrie/#/builders/63/builds/3527/steps/13/logs/stdio

Last log lines:
/usr/src/debug/ltp/20260130/testcases/open_posix_testsuite/conformance/interfaces/sigtimedwait/6-1.c 102: Test PASSED
/usr/src/debug/ltp/20260130/testcases/open_posix_testsuite/conformance/interfaces/sigwait/1-1.c 88: Test PASSED
/usr/src/debug/ltp/20260130/testcases/open_posix_testsuite/conformance/interfaces/sigwait/2-1.c 128: Test PASSED
/usr/src/debug/ltp/20260130/testcases/open_posix_testsuite/conformance/interfaces/sigwait/3-1.c 98: Test PASSED
/usr/src/debug/ltp/20260130/testcases/open_posix_testsuite/conformance/interfaces/sigwait/4-1.c 80: Test PASSED

Process killed - no output for 300 seconds. Total running time: 375 seconds.

So, we again a lack a bit of logging, but I'm fairly convinced the failing test
is in testcases/open_posix_testsuite/conformance/interfaces/sigwait/, either
6-1.c, 6-2.c or 7-1.c.

Last seen successful test is 4-1.c.
6-1.c, 6-2.c and 7-1.c use a different function to print data.
8-1.c is using printf, as for 4-1.c, so we should see traces if it was running

Adding more traces. Again.
Comment 6 Mathieu Dubois-Briand 2026-06-25 10:23:34 UTC
qemuarm64-ltp ubuntu2504-vk-arm1 master completed at 2026-06-24 03:11:13+00:00
https://autobuilder.yoctoproject.org/valkyrie/#/builders/98/builds/3850/steps/13/logs/stdio
Comment 7 Antonin Godard 2026-09-23 14:55:12 UTC
Issue is really similar although this time for the threads LTP test suite:

qemux86-64-ltp debian13-vk-1 agodard/master-next completed at 2026-09-15 15:57:34+00:00
https://autobuilder.yoctoproject.org/valkyrie/#/builders/63/builds/4434/steps/15/logs/stdio
Comment 8 Mathieu Dubois-Briand 2026-10-27 08:55:37 UTC
qemux86-64-ltp debian11-vk-1 mathieu/master-next-tests&mathieu/master-next completed at 2026-10-23 17:39:22+00:00
https://autobuilder.yoctoproject.org/valkyrie/#/builders/63/builds/4670/steps/16/logs/stdio
Comment 9 Paul Barker 2026-10-29 14:51:18 UTC
Can we try to improve logging?
Comment 10 Mathieu Dubois-Briand 2026-11-02 13:12:51 UTC
I have some patch in my branch that tries do that for the few tests involved in there. I will see if I can come with something generic that we can merge.