Bug#997029: pysph: FTBFS in testing and unstable

Antonio Valentino antonio.valentino at tiscali.it
Fri Dec 31 12:22:32 GMT 2021


Dear Graham,
sorry for the late reply.


On Fri, 22 Oct 2021 18:45:01 +0200 Graham Inggs <ginggs at debian.org> wrote:
> Source: pysph
> Version: 1.0~b0~20191115.gite3d5e10-5
> Severity: serious
> Tags: ftbfs
> 
> Hi Maintainer
> 
> As can be seen in reproducible builds [1], pysph sometimes FTBS in
> both testing and unstable.  I've copied what I hope is the relevant
> part of the log below.
> 
> Regards
> Graham
> 
> 
> [1] https://tests.reproducible-builds.org/debian/rb-pkg/pysph.html
> 
> 
> =================================== FAILURES ===================================
> _________________ MPIReduceArrayTestCase.test_parallel_reduce __________________
> 
> self = <pysph.parallel.tests.test_parallel.MPIReduceArrayTestCase
> testMethod=test_parallel_reduce>
> 
>     @mark.parallel
>     def test_parallel_reduce(self):
>         args = ['--directory=%s' % self.root]
> >       run_parallel_script.run(
>             filename='simple_reduction.py', args=args, nprocs=nprocs, path=path
>         )
> 
> pysph/parallel/tests/test_parallel.py:101:
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> 
> filename = 'simple_reduction.py', args = ['--directory=/tmp/tmpr8h_5ft1']
> nprocs = 2, timeout = 30.0
> path = '/build/1st/pysph-1.0~b0~20191115.gite3d5e10/.pybuild/cpython3_3.9/build/pysph/parallel/tests'
> 
>     def run(filename, args=None, nprocs=2, timeout=30.0, path=None):
>         """Run a python script with MPI or in serial (if nprocs=1).
> Kill process
>         if it takes longer than the specified timeout.
> 
>         Parameters:
>         -----------
>         filename - filename of python script to run under mpi.
>         args - List of arguments to pass to script.
>         nprocs - number of processes to run (1 => serial non-mpi run).
>         timeout - time in seconds to wait for the script to finish running,
>             else raise a RuntimeError exception.
>         path - the path under which the script is located
>             Defaults to the location of this file (__file__), not curdir.
> 
>         """
>         if args is None:
>             args = []
>         file_path = abspath(join(path, filename))
>         cmd = [sys.executable, file_path] + args
>         if nprocs > 1:
>             cmd = ['mpiexec', '-n', str(nprocs)] + cmd


The package seems to build correctly now on all platforms that provide 
the required dependencies.
Also the status in debci [1] seems to be fine (please not that the 
package is currently not buildable on some platform).

In the past there was some issue related to some weird behavior of mpi 
libraries but now the problem seems to be solved.

Do you have an updated pointer to a build failure?
... or can we consider to close or at least reduce the severity of this 
issue?

[1] https://ci.debian.net/packages/p/pysph/


kind regards

-- 
Antonio Valentino



More information about the debian-science-maintainers mailing list