[Debian-med-packaging] Bug#1032550: igdiscover: FTBFS in testing: dh_auto_test: error: pybuild --test --test-pytest -i python{version} -p 3.11 returned exit code 13

[Santiago Vila]

El 9/3/23 a las 16:18, Andreas Tille escribió:
> I've rebuild igdiscover 0.11-3 in a testing and unstable
> environment.  Both builds are passing all tests.

Hello. This one FTBFS randomly for me. Therefore,
to "reproduce the randomness" you might have to try several times,
not just once.

My build environment is currently made of VMs with 1 and 2 CPUs.

For this package, I have two successful builds using 2 CPUs,
and nothing more, so it may be random there, or maybe not,
because I've not tested it enough times to know.

When using 1 CPU, this is my build history so far:

Status: successful  igdiscover_0.11-3_amd64-20220501T161126.901Z
Status: failed      igdiscover_0.11-3_amd64-20221108T225319.839Z
Status: successful  igdiscover_0.11-3_amd64-20221115T130535.295Z
Status: successful  igdiscover_0.11-3_amd64-20221115T130920.067Z
Status: failed      igdiscover_0.11-3_amd64-20221115T133708.367Z
Status: successful  igdiscover_0.11-3_amd64-20221115T135147.584Z
Status: failed      igdiscover_0.11-3_amd64-20221115T144835.514Z
Status: successful  igdiscover_0.11-3_amd64-20221115T144843.540Z
Status: successful  igdiscover_0.11-3_amd64-20221115T145126.644Z
Status: successful  igdiscover_0.11-3_amd64-20221115T145404.560Z
Status: successful  igdiscover_0.11-3_amd64-20221115T144945.690Z
Status: successful  igdiscover_0.11-3_amd64-20221115T145128.070Z
Status: successful  igdiscover_0.11-3_amd64-20230111T155218.678Z
Status: successful  igdiscover_0.11-3_amd64-20230111T155235.799Z
Status: successful  igdiscover_0.11-3_amd64-20230125T193926.333Z
Status: successful  igdiscover_0.11-3_amd64-20230125T194042.500Z
Status: successful  igdiscover_0.11-3_amd64-20230125T195323.679Z
Status: failed      igdiscover_0.11-3_amd64-20230125T195500.696Z
Status: successful  igdiscover_0.11-3_amd64-20230125T200024.475Z
Status: successful  igdiscover_0.11-3_amd64-20230125T200000.253Z
Status: successful  igdiscover_0.11-3_amd64-20230125T200433.788Z
Status: successful  igdiscover_0.11-3_amd64-20230125T201954.420Z
Status: successful  igdiscover_0.11-3_amd64-20230125T202301.598Z
Status: successful  igdiscover_0.11-3_amd64-20230125T203341.230Z
Status: failed      igdiscover_0.11-3_amd64-20230125T203408.732Z
Status: failed      igdiscover_0.11-3_amd64-20230125T204109.497Z
Status: successful  igdiscover_0.11-3_amd64-20230125T204244.306Z
Status: successful  igdiscover_0.11-3_amd64-20230125T204258.004Z
Status: successful  igdiscover_0.11-3_amd64-20230125T205136.438Z
Status: successful  igdiscover_0.11-3_amd64-20230125T205808.095Z
Status: failed      igdiscover_0.11-3_amd64-20230125T210534.566Z
Status: successful  igdiscover_0.11-3_amd64-20230125T210502.213Z
Status: successful  igdiscover_0.11-3_amd64-20230125T210815.549Z
Status: successful  igdiscover_0.11-3_amd64-20230125T211006.621Z
Status: successful  igdiscover_0.11-3_amd64-20230126T180837.356Z
Status: successful  igdiscover_0.11-3_amd64-20230126T180948.067Z
Status: failed      igdiscover_0.11-3_amd64-20230126T182131.288Z
Status: failed      igdiscover_0.11-3_amd64-20230126T183203.578Z


My advice, therefore, is to try using a machine with a single CPU.

This is very easy to achieve by setting GRUB_CMDLINE_LINUX="nr_cpus=1"
in /etc/default/grub.

Nilesh Patra wrote:
> I wonder if it is another of those "$specific-CPU-configuration" failures.

Please note that no specific configuration is mandated by Debian policy, so
I understand that packages must build from source regardless of the number
of CPUs. This is what usually happens, and it's also what our users expect.

If you could not reproduce the randomness with the grub trick above, please
contact me privately, as I could offer ssh access to a machine where
such randomness happens.

Thanks.