Bug#965352: libopenmpi3: breaks tests in client programs

From	Drew Parsons <dparsons@debian.org>
Newsgroups	linux.debian.bugs.rc
Subject	Bug#965352: libopenmpi3: breaks tests in client programs
Date	2020-07-20 07:50 +0200
Message-ID	<AuwQx-7Aw-3@gated-at.bofh.it> (permalink)
Organization	linux.* mail to news gateway

Show all headers | View raw

Package: libopenmpi3
Version: 4.0.4-2
Severity: serious
Justification: breaks dependent packages

A number of packages have lately started failing debci tests,
in particular mpi4py and dolfin. dolfin is also failing some build-time
tests.

I've run mpi4py tests on debci for both testing and unstable. They're
still passing in testing, where openmpi is 4.0.4-1, but failing in
unstable with openmpi 4.0.4-2. I'm not absolutely certain, but it
seems like one of the changes in openmpi 4.0.4-2 might be making
problems.

Debci indicates that mpi4py 3.0.3-5 isn't passing
test/test_rma.py::TestRMASelf::testGetAccumulate
Test log in https://ci.debian.net/data/autopkgtest/unstable/amd64/m/mpi4py/6342921/log.gz
Apparently there's a segfault in an mpi test (single process test pass).

The problem can be reproduced at the command line in an mpi4py source
dir with python3-mpi4py installed (manual testing also shows
TestRMASelf::testStartComplete is failing, separate from
testGetAccumulate):

$ mpirun -n 2 python3 -m pytest test/test_rma.py -v
============================= test session starts ==============================
platform linux -- Python 3.8.4, pytest-4.6.11, py-1.8.1, pluggy-0.13.0 -- /usr/bin/python3
============================= test session starts ==============================
platform linux -- Python 3.8.4, pytest-4.6.11, py-1.8.1, pluggy-0.13.0 -- /usr/bin/python3
cachedir: .pytest_cache
hypothesis profile 'default' -> database=DirectoryBasedExampleDatabase('/home/drew/projects/python/build/mpi4py/.hypothesis/examples')
rootdir: /home/drew/projects/python/build/mpi4py, inifile: setup.cfg
plugins: xdist-1.32.0, doctestplus-0.7.0, astropy-header-0.1.2, filter-subpackage-0.1.1, cov-2.8.1, remotedata-0.3.2, mpi-0+unknown, forked-1.2.0, arraydiff-0.3, openfiles-0.5.0, hypothesis-5.16.0
collecting ... cachedir: .pytest_cache
hypothesis profile 'default' -> database=DirectoryBasedExampleDatabase('/home/drew/projects/python/build/mpi4py/.hypothesis/examples')
rootdir: /home/drew/projects/python/build/mpi4py, inifile: setup.cfg
plugins: xdist-1.32.0, doctestplus-0.7.0, astropy-header-0.1.2, filter-subpackage-0.1.1, cov-2.8.1, remotedata-0.3.2, mpi-0+unknown, forked-1.2.0, arraydiff-0.3, openfiles-0.5.0, hypothesis-5.16.0
collected 36 items                                                             
collected 36 items                                                             

test/test_rma.py::TestRMASelf::testAccumulate 
test/test_rma.py::TestRMASelf::testAccumulate PASSED                     [  2%]
test/test_rma.py::TestRMASelf::testAccumulateProcNullReplace PASSED                     [  2%]PASSED      [  5%]
test/test_rma.py::TestRMASelf::testAccumulateProcNullSum 
test/test_rma.py::TestRMASelf::testAccumulateProcNullReplace PASSED          [  8%]PASSED      [  5%]
test/test_rma.py::TestRMASelf::testCompareAndSwap 
test/test_rma.py::TestRMASelf::testAccumulateProcNullSum PASSED          [  8%]PASSED
                 [ 11%]test/test_rma.py::TestRMASelf::testCompareAndSwap 
test/test_rma.py::TestRMASelf::testFence PASSED                          [ 13%]
test/test_rma.py::TestRMASelf::testFenceAll PASSED                       [ 16%]
test/test_rma.py::TestRMASelf::testFetchAndOp PASSED                     [ 19%]
test/test_rma.py::TestRMASelf::testFlush PASSED                          [ 22%]
test/test_rma.py::TestRMASelf::testGetAccumulate PASSED                  [ 25%]
test/test_rma.py::TestRMASelf::testGetAccumulateProcNull PASSED          [ 27%]
test/test_rma.py::TestRMASelf::testGetProcNull PASSED                    [ 30%]
test/test_rma.py::TestRMASelf::testPostWait PASSED                       [ 33%]
test/test_rma.py::TestRMASelf::testPutGet PASSED                         [ 36%]
test/test_rma.py::TestRMASelf::testPutProcNull PASSED                    [ 38%]
test/test_rma.py::TestRMASelf::testStartComplete FAILED                  [ 41%]
test/test_rma.py::TestRMASelf::testStartCompletePostTest PASSED          [ 44%]
test/test_rma.py::TestRMASelf::testStartCompletePostWait PASSED          [ 47%]
test/test_rma.py::TestRMASelf::testSync PASSED                           [ 50%]
test/test_rma.py::TestRMAWorld::testAccumulate --------------------------------------------------------------------------
Primary job  terminated normally, but 1 process returned
a non-zero exit code. Per user-direction, the job has been aborted.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
mpirun noticed that process rank 1 with PID 0 on node monte exited on signal 11 (Segmentation fault).
--------------------------------------------------------------------------


The tests pass with mpirun -n 1.




-- System Information:
Debian Release: bullseye/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 5.7.0-1-amd64 (SMP w/8 CPU threads)
Locale: LANG=en_AU.UTF-8, LC_CTYPE=en_AU.UTF-8 (charmap=UTF-8), LANGUAGE=en_AU:en
Shell: /bin/sh linked to /usr/bin/dash
Init: systemd (via /run/systemd/system)
LSM: AppArmor: enabled

Versions of packages libopenmpi3 depends on:
ii  libc6                    2.31-1
ii  libevent-core-2.1-7      2.1.12-stable-1
ii  libevent-pthreads-2.1-7  2.1.12-stable-1
ii  libfabric1               1.6.2-3+b1
ii  libgcc-s1                10.1.0-6
ii  libhwloc-plugins         2.2.0+dfsg-3
ii  libhwloc15               2.2.0+dfsg-3
ii  libibverbs1              29.0-1
ii  libnl-3-200              3.4.0-1+b1
ii  libpmix2                 3.1.5-1
ii  libpsm-infinipath1       3.3+20.604758e7-6+b1
ii  libpsm2-2                11.2.86-1
ii  libstdc++6               10.1.0-6
ii  libucx0                  1.8.1-1
ii  zlib1g                   1:1.2.11.dfsg-2

libopenmpi3 recommends no packages.

libopenmpi3 suggests no packages.

-- no debconf information

Back to linux.debian.bugs.rc | Previous | Next — Next in thread | Find similar | Unroll thread

Thread

Bug#965352: libopenmpi3: breaks tests in client programs Drew Parsons <dparsons@debian.org> - 2020-07-20 07:50 +0200
  Bug#965352: libopenmpi3: breaks tests in client programs Drew Parsons <dparsons@debian.org> - 2020-07-20 12:10 +0200
  Processed: Re: libopenmpi3: breaks tests in client programs "Debian Bug Tracking System" <owner@bugs.debian.org> - 2020-07-23 07:40 +0200
  Bug#965352: libopenmpi3: breaks tests in client programs Drew Parsons <dparsons@debian.org> - 2020-07-24 04:20 +0200
  Processed: Re: libopenmpi3: breaks tests in client programs "Debian Bug Tracking System" <owner@bugs.debian.org> - 2020-07-24 04:20 +0200
  Processed: Re: libopenmpi3: breaks tests in client programs "Debian Bug Tracking System" <owner@bugs.debian.org> - 2020-08-07 15:20 +0200
  Bug#965352: marked as done (libopenmpi3: breaks tests in client  programs) "Debian Bug Tracking System" <owner@bugs.debian.org> - 2020-09-12 15:00 +0200

csiph-web