[Debichem-devel] Bug#1094090: Bug#1094090: dbcsr: FTBFS: dh_auto_test: error: cd obj-x86_64-linux-gnu && make -j1 test ARGS\+=--verbose ARGS\+=-j1 returned exit code 2
Michael Banck
mbanck at debian.org
Wed Jan 29 20:56:52 GMT 2025
Hi,
On Fri, Jan 24, 2025 at 09:20:23PM +0100, Lucas Nussbaum wrote:
> During a rebuild of all packages in sid, your package failed to build
> on amd64.
>
> This looks similar to #1087439 but this one is reproducible on amd64.
Hrm, this one seems to be a runtime OpenMPI failure, #1087439 appears to
have been a build time issue with cmake:
> > test 11
> > Start 11: dbcsr_unittest1
> >
> > 11: Test command: /usr/bin/mpiexec "-n" "4" "/build/reproducible-path/dbcsr-2.6.0/obj-x86_64-linux-gnu/tests/dbcsr_unittest1"
> > 11: Working Directory: /build/reproducible-path/dbcsr-2.6.0/obj-x86_64-linux-gnu/tests
> > 11: Environment variables:
> > 11: OMP_NUM_THREADS=2
> > 11: Test timeout computed to be: 1500
> > 11: [sbuild:02582] PMIX ERROR: PMIX_ERR_EXISTS in file ../../../../../../3rd-party/prrte/src/util/session_dir.c at line 95
> > 11: [sbuild:02582] PRTE ERROR: Exists in file ../../../../../../3rd-party/prrte/src/util/session_dir.c at line 172
> > 11: --------------------------------------------------------------------------
> > 11: It looks like "prte_init()" failed for some reason. There are many
> > 11: reasons that can cause PRRTE to fail during "prte_init()", some of
> > 11: which are due to configuration or environment problems. This failure
> > 11: appears to be an internal failure — here's some additional information
> > 11: (which may only be relevant to a PRRTE developer):
> > 11:
> > 11: prte_session_dir failed
> > 11: --> Returned value Exists (-14) instead of PRTE_SUCCESS
> > 11: --------------------------------------------------------------------------
> > 11/19 Test #11: dbcsr_unittest1 .......................................***Failed 0.02 sec
Not sure what exactly to make of it, I'll retest here locally. I have
the 2.8.0-1 packaging for dbcsr ready locally, and that one ran the
above unittest #11 fine as well.
Michael
More information about the Debichem-devel
mailing list