#1021057 cp2k autopkgtest times outs on armel

Package:
src:cp2k
Source:
src:cp2k
Submitter:
Paul Gevers
Date:
2026-11-03 16:11:04 UTC
Severity:
normal
#1021057#5
Date:
2022-10-01 09:53:27 UTC
From:
To:
Dear maintainer(s),

Your package has an autopkgtest, great. However, it fails on armel.
What's worse, it fails because it seems to hang and eventually times out
due to autopkgtest. Can you please investigate the situation and fix it?
I copied some of the output at the bottom of this report.

Tests that time out (while normally running in much less time) are bad
for our infrastructure. I have added your package to our reject_list on
armel and will remove it once this bug is closed.

More information about this bug and the reason for filing it can be
found on
https://wiki.debian.org/ContinuousIntegration/RegressionEmailInformation

Paul

[1] https://lists.debian.org/debian-devel-announce/2019/07/msg00002.html

https://ci.debian.net/packages/c/cp2k/

https://ci.debian.net/data/autopkgtest/testing/armel/c/cp2k/26510840/log.gz


Running MC_QS_cluster.inp
 >>>
/tmp/autopkgtest-lxc._uoweqoc/downtmp/build.l8C/src/tools/regtesting/../..//TEST-Linux-armv8l-gfortran-psmp-2022-09-28_01-25-33/tests/MC/regtest
     MC_QS.inp
-51.34794469     OK (  31.92 sec)
     MC_QS_restart.inp
-51.34909000     OK (  38.61 sec)
     GEMC_NpT_box1.inp
0.49826574     OK (   5.68 sec)
     GEMC_NVT_box1.inp
0.49803069     OK (   5.25 sec)
     GEMC_swap_box1.inp
0.55872200     OK (   9.25 sec)
     canonical.inp
0.55333815     OK (   0.94 sec)
     canonical_bias.inp
0.38566238     OK (   3.88 sec)
     virial.inp
-.723498E+03     OK (   3.35 sec)
     hmc.inp
                        KILLED
     MC_QS_cluster.inp
-123.50157228768187               KILLED
<<<
/tmp/autopkgtest-lxc._uoweqoc/downtmp/build.l8C/src/tools/regtesting/../..//TEST-Linux-armv8l-gfortran-psmp-2022-09-28_01-25-33/tests/MC/regtest
(9 of 53) done in 551.00 sec
Starting regression tests in
/tmp/autopkgtest-lxc._uoweqoc/downtmp/build.l8C/src/tools/regtesting/../..//TEST-Linux-armv8l-gfortran-psmp-2022-09-28_01-25-33/tests/QS/regtest-ri-mp2
(10 of 53)
Running RI_MP2_H2O.inp
autopkgtest [04:12:14]: ERROR: timed out on command [...]
#1021057#10
Date:
2022-10-30 09:27:47 UTC
From:
To:
I think both paragraphs are wrong for cp2k...

It is not normally running in much less time, the faster runs are
failures for actual breakage (the last one in unstable that still
has a test log was some openmpi/pmix breakage).

I am not convinced that it actually hangs,
I would guess your timeout is just too low.

Running FPU-heavy code without FPU is slow.

https://buildd.debian.org/status/logs.php?pkg=cp2k&arch=armhf
https://buildd.debian.org/status/logs.php?pkg=cp2k&arch=armel

This is build time (which should be similar on armel and armhf)
plus test time, where an order of magnitude difference between
armhf and armel is expected for FPU-heavy code.

Could you try setting the timeout to 10 hours for cp2k on armel?
I'd be optimistic that this will result in a pass.

cu
Adrian
#1021057#17
Date:
2022-10-30 15:58:27 UTC
From:
To:
Hi Paul,

I guess that was supposed to be 9.1-2, as that is (was) current and the
log you referenced also has that.

I agree with Adrian and downgraded this bug to important for now.


Michael
#1021057#22
Date:
2022-10-30 18:50:33 UTC
From:
To:
Hi,

We don't have that flexibility. If the test can't run in a reasonable
time (below 2:47h per stanza in d/t/control) on armel, please skip the
test on armel until the infrastructure grows that option.

Paul
#1021057#27
Date:
2026-11-03 14:32:15 UTC
From:
To:
Hi,

We don't support armel in Debian anymore in unstable/forky, so let's
consider this bug done.

Paul