Bug#769801: mpi4py: FTBFS under pbuilder: Failure in test_spawn.TestSpawnSelf
Just a datapoint: It builds fine under sbuild: sbuild -A -s -d unstable mpi4py_1.3.1+hg20131106-1.dsc The test run OK there, but maybe sbuild does less environment sanitising? As it's sbuild that gets used on the buildds, does that mean that this is not actually a serious bug? Wookey -- Principal hats: Linaro, Debian, Wookware, ARM http://wookware.org/ -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#769801: mpi4py: FTBFS under pbuilder: Failure in test_spawn.TestSpawnSelf
Source: mpi4py Version: 1.3.1+hg20131106-1 Severity: serious From my pbuilder build log (on amd64): ... set -e; for v in 2.7 3.4; do \ echo I: testing using python$v; \ PYTHONPATH=`/bin/ls -d /tmp/buildd/mpi4py-1.3.1+hg20131106/build/lib.*-$v` \ /usr/bin/python$v /usr/bin/nosetests -v --exclude='testPackUnpackExternal'; \ done I: testing using python2.7 testAttr (test_attributes.TestCommAttrSelf) ... ok testAttrCopyDelete (test_attributes.TestCommAttrSelf) ... ok testAttrCopyFalse (test_attributes.TestCommAttrSelf) ... ok ... testPutGet (test_rma.TestRMAWorld) ... ok testPutProcNull (test_rma.TestRMAWorld) ... ok testArgsOnlyAtRoot (test_spawn.TestSpawnSelf) ... -- At least one pair of MPI processes are unable to reach each other for MPI communications. This means that no Open MPI device has indicated that it can be used to communicate between these processes. This is an error; Open MPI requires that all MPI processes be able to reach each other. This error can sometimes be the result of forgetting to specify the self BTL. Process 1 ([[22838,1],0]) is on host: frobozz Process 2 ([[22838,2],0]) is on host: frobozz BTLs attempted: self sm Your MPI job is now going to abort; sorry. -- ERROR testArgsOnlyAtRootMultiple (test_spawn.TestSpawnSelf) ... [frobozz:11719] *** An error occurred in MPI_Barrier [frobozz:11719] *** on communicator MPI_COMM_PARENT [frobozz:11719] *** MPI_ERR_INTERN: internal error [frobozz:11719] *** MPI_ERRORS_ARE_FATAL: your MPI job will now abort [frobozz:11717] 1 more process has sent help message help-mca-bml-r2.txt / unreachable proc [frobozz:11717] Set MCA parameter orte_base_help_aggregate to 0 to see all help / error messages debian/rules:83: recipe for target 'override_dh_auto_test' failed make[1]: *** [override_dh_auto_test] Error 1 make[1]: Leaving directory '/tmp/buildd/mpi4py-1.3.1+hg20131106' debian/rules:16: recipe for target 'build' failed make: *** [build] Error 2 dpkg-buildpackage: error: debian/rules build gave error exit status 2 This seems like it might be related to the fact that pbuilder now completely disables network access except for a new instance of lo (as opposed to what it did before, just blanking out /etc/resolv.conf). -- Daniel Schepler -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
