Bug 1170639 - perl-Image-Sane test gets stuck on 1-core VM
perl-Image-Sane test gets stuck on 1-core VM
Status: NEW
Classification: openSUSE
Product: openSUSE Tumbleweed
Classification: openSUSE
Component: Other
Current
Other Other
: P5 - None : Normal (vote)
: ---
Assigned To: Malcolm Lewis
E-mail List
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2020-04-28 04:37 UTC by Bernhard Wiedemann
Modified: 2020-07-05 04:22 UTC (History)
1 user (show)

See Also:
Found By: ---
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Bernhard Wiedemann 2020-04-28 04:37:00 UTC
Steps To Reproduce:
osc build --noservice --vm-type=kvm -j1

Actual Result:
test output gets stuck after
t/enums.t .............. ok

in t/err.t
in $test->cancel


In successful parallel runs, the loop gets repeated 11 times
so maybe there is some background-asynchronous work happening
that gets stuck?
Not sure if this is a bug in sane-backends or in how it is used.
Comment 1 Bernhard Wiedemann 2020-04-28 04:44:25 UTC
It can also get stuck in t/81_scanimage-perl.t

It is also possible to trigger the problem on machines with more CPU cores using this patch:

 %check
-make test
+taskset 1 make test
Comment 2 Malcolm Lewis 2020-05-27 14:41:12 UTC
(In reply to Bernhard Wiedemann from comment #1)
> It can also get stuck in t/81_scanimage-perl.t
> 
> It is also possible to trigger the problem on machines with more CPU cores
> using this patch:
> 
>  %check
> -make test
> +taskset 1 make test
Hi
I normally use just a build-root, no kvm, so the first one (-j1) works, the above test fails in a build root and can duplicate the kvm build hang.

For me it gets stuck in t/81_scanimage-perl.t only. I'm assuming this is tied in with reproducible builds?
Comment 3 Bernhard Wiedemann 2020-07-05 04:22:19 UTC
> I'm assuming this is tied in with reproducible builds?

yes, I am building with -j1 and -j4 to find variations from parallelism.
Stuck builds are not yet handled in my reproducibleopensuse tools,
so they are pretty annoying.

Apart from that, there could be an actual bug that we want to squash.