Summary: | media-sound/pulseaudio-15.0-r1 fails thread-test test | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Paolo Pedroni <paolo.pedroni> |
Component: | Current packages | Assignee: | Gentoo Linux Gnome Desktop Team <gnome> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | jstein |
Priority: | Normal | Keywords: | TESTFAILURE |
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | pulseaudio-15.0-r1:20211115-164722.log.gz |
Description
Paolo Pedroni
2021-11-15 17:03:00 UTC
If you still can reproduce this, please try again with the following change to thread-test.c - this test is seen to seldom fail on VMs with limited CPU resources: diff --git a/src/tests/thread-test.c b/src/tests/thread-test.c index 0c83e67e0..4fcbfd4d3 100644 --- a/src/tests/thread-test.c +++ b/src/tests/thread-test.c @@ -153,6 +153,7 @@ int main(int argc, char *argv[]) { s = suite_create("Thread"); tc = tcase_create("thread"); tcase_add_test(tc, thread_test); + tcase_set_timeout(tc, 60 * 60); suite_add_tcase(s, tc); sr = srunner_create(s); (In reply to Paolo Pedroni from comment #0) ... > > 44/51 thread-test > FAIL > 4.02s exit status 1 Hi Paolo, could you please check the environment where this test fails, and if this is a slower/resource-constrained VM the patch helps? (In reply to Igor V. Kovalenko from comment #2) > (In reply to Paolo Pedroni from comment #0) > ... > > > > 44/51 thread-test > > FAIL > > 4.02s exit status 1 > Hi Paolo, could you please check the environment where this test fails, and > if this is a slower/resource-constrained VM the patch helps? It is not a VM, just an old, slow PC. The patch actually helps: [...] 40/51 cpu-sconv-test OK 1.80s 41/51 mainloop-test-glib OK 0.41s 42/51 mainloop-test OK 2.16s 43/51 volume-test OK 3.60s 44/51 mult-s16-test OK 5.33s 45/51 cpu-volume-test OK 6.60s 46/51 thread-test OK 6.14s 47/51 cpu-mix-test OK 8.93s 48/51 thread-mainloop-test OK 10.02s 49/51 cpu-remap-test OK 11.19s 50/51 once-test OK 10.61s 51/51 lock-autospawn-test OK 20.02s Ok: 51 Expected Fail: 0 Fail: 0 Unexpected Pass: 0 Skipped: 0 Timeout: 0 Full log written to /var/tmp/portage/media-sound/pulseaudio-15.0-r2/work/pulseaudio-15.0-abi_x86_64.amd64/meson-logs/testlog.txt >>> Completed testing media-sound/pulseaudio-15.0-r2 (In reply to Paolo Pedroni from comment #3) > (In reply to Igor V. Kovalenko from comment #2) > > (In reply to Paolo Pedroni from comment #0) > > ... > > > > > > 44/51 thread-test > > > FAIL > > > 4.02s exit status 1 > > Hi Paolo, could you please check the environment where this test fails, and > > if this is a slower/resource-constrained VM the patch helps? > > It is not a VM, just an old, slow PC. The patch actually helps: > [...] Thanks for confirming this! Do you have any runtime experience with pulseaudio on this old, slow PC? This test is not a performance test of any kind, it checks that pulseaudio use of thread conditional variables results in execution progress. I think we can just bump the timeout upstream. Just in case, upstream PR https://gitlab.freedesktop.org/pulseaudio/pulseaudio/-/merge_requests/696 (In reply to Igor V. Kovalenko from comment #4) > Thanks for confirming this! Do you have any runtime experience with > pulseaudio on this old, slow PC? Not really. It's a media player with Kodi on it. Pulseaudio is installed as a dependency of something, but Kodi is configured to use Pipewire for audio. > This test is not a performance test of any kind, it checks that pulseaudio > use of thread conditional variables results in execution progress. I think > we can just bump the timeout upstream. Thanks. Upstream change is now merged and should appear in upcoming 16.0 release. Can we close this issue now? (In reply to Igor V. Kovalenko from comment #7) > Can we close this issue now? Fine for me. |