Created attachment 1063074 [details] Disable mdrun test on ppc64 and s390 Description of problem: mdrun test case times out on ppc64 and s390[x]. Version-Release number of selected component (if applicable): 5.0.6-4 How reproducible: koji-build gromacs on ppc64 and s390[x]. Actual results: Start 15: MdrunTests 15/15 Test #15: MdrunTests .......................***Timeout 1500.02 sec [...] WARNING: Using the slow plain C kernels. This should not happen during routine usage on supported platforms. [...] 93% tests passed, 1 tests failed out of 15 Label Time Summary: GTest = 3.38 sec IntegrationTest = 1500.85 sec UnitTest = 3.38 sec Total Test time (real) = 1504.25 sec The following tests FAILED: 15 - MdrunTests (Timeout) CMakeFiles/run-ctest.dir/build.make:52: recipe for target 'CMakeFiles/run-ctest' failed make[3]: Leaving directory '/builddir/build/BUILD/gromacs-5.0.6/openmpi' CMakeFiles/Makefile2:963: recipe for target 'CMakeFiles/run-ctest.dir/all' failed make[2]: Leaving directory '/builddir/build/BUILD/gromacs-5.0.6/openmpi' CMakeFiles/Makefile2:940: recipe for target 'CMakeFiles/check.dir/rule' failed make[1]: Leaving directory '/builddir/build/BUILD/gromacs-5.0.6/openmpi' Makefile:547: recipe for target 'check' failed RPM build errors: Errors while running CTest make[3]: *** [CMakeFiles/run-ctest] Error 8 make[2]: *** [CMakeFiles/run-ctest.dir/all] Error 2 make[1]: *** [CMakeFiles/check.dir/rule] Error 2 make: *** [check] Error 2 error: Bad exit status from /var/tmp/rpm-tmp.rnrPs8 (%check) Additional info: this looks like a hardware problem. Looking for the warning found in the build log, I found this thread [1] explaining that devs assume one will run gromacs on GPU accelerated platforms, or at least with SIMD-capable CPU, and the P7 is still too slow. The attached patch disables such a test on the affected platforms. [1] http://permalink.gmane.org/gmane.science.biology.gromacs.user/77673
Thanks for the report. I'm in the process of upgrading gromacs to 5.1 for f24 (and will probably backport to f23 once bugs are shaken out). It doesn't time out. There are a couple of other issues, though. One is an out-of-memory error when running the testsuite on s390 (but s390x runs fine). Another is a compilation error on ppc64/ppc64le (reported upstream). I'll push the updated version to git soon.
Fixed in rawhide. I tested on ppc.koji and s390.koji, but feel free to test yourself and reopen if there are still some issues.
It is indeed fixed for f>=23. Could you apply the fixes for f22 as well? Gromacs fails to build on f22/ppc64 [1]. [1] http://ppc.koji.fedoraproject.org/koji/taskinfo?taskID=2938260 Thank you.
Bumping down to F22.
gromacs-5.0.7-2.fc22 has been submitted as an update to Fedora 22. https://bodhi.fedoraproject.org/updates/FEDORA-2016-39b236aba6
gromacs-5.0.7-2.fc22 has been pushed to the Fedora 22 testing repository. If problems still persist, please make note of it in this bug report. See https://fedoraproject.org/wiki/QA:Updates_Testing for instructions on how to install test updates. You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2016-39b236aba6
gromacs-5.0.7-2.fc22 has been pushed to the Fedora 22 stable repository. If problems still persist, please make note of it in this bug report.