Hi Siegmar, It seems it is still caused by the unaligned memory access issue in MPICH. Unfortunately, the fix patch is still in our reviewing queue. I hope it can be pushed to master as soon as we can. Please check this ticket for tracking the status (it will be closed once we have pushed fix to master). Thanks. https://trac.mpich.org/projects/mpich/ticket/2309 Min On 6/8/16 1:14 AM, Siegmar Gross wrote:

Hi,

I have built mpich-master-v3.2-331-g44fd9c5f39e5 on my machines (Solaris 10 Sparc, Solaris 10 x86_64, and openSUSE Linux 12.1 x86_64) with gcc-5.1.0 and Sun C 5.13. Most of the time I get an error with different error messages spawning processes on a Sparc machine.

tyr spawn 107 mpiexec -np 1 --host tyr,tyr,tyr,tyr,tyr spawn_master

Parent process 0 running on tyr.informatik.hs-fulda.de I create 4 slave processes

Parent process 0: tasks in MPI_COMM_WORLD: 1 tasks in COMM_CHILD_PROCESSES local group: 1 tasks in COMM_CHILD_PROCESSES remote group: 4

Slave process 0 of 4 running on tyr.informatik.hs-fulda.de Slave process 1 of 4 running on tyr.informatik.hs-fulda.de Slave process 2 of 4 running on tyr.informatik.hs-fulda.de Slave process 3 of 4 running on tyr.informatik.hs-fulda.de spawn_slave 0: argv[0]: spawn_slave spawn_slave 1: argv[0]: spawn_slave spawn_slave 2: argv[0]: spawn_slave spawn_slave 3: argv[0]: spawn_slave

tyr spawn 108 mpiexec -np 1 --host tyr,tyr,tyr,tyr,tyr spawn_master

Parent process 0 running on tyr.informatik.hs-fulda.de I create 4 slave processes

Fatal error in MPI_Comm_spawn: Unknown error class, error stack: MPI_Comm_spawn(141)...................: MPI_Comm_spawn(cmd="spawn_slave", argv=0, maxprocs=4, MPI_INFO_NULL, root=0, MPI_COMM_WORLD, intercomm=ffffffff7fffdf58, errors=0) failed MPIDI_Comm_spawn_multiple(274)........: MPID_Comm_accept(153).................: MPIDI_Comm_accept(1039)...............: MPIDU_Complete_posted_with_error(1137): Process failed tyr spawn 109

tyr spawn 111 mpiexec -np 1 --host tyr,tyr,tyr,tyr,tyr spawn_master

Parent process 0 running on tyr.informatik.hs-fulda.de I create 4 slave processes

===================================================================================

= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES = PID 3322 RUNNING AT tyr = EXIT CODE: 10 = CLEANING UP REMAINING PROCESSES = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES ===================================================================================

[proxy:0:[email protected]] HYD_pmcd_pmip_control_cmd_cb (../../../../mpich-master-v3.2-331-g44fd9c5f39e5/src/pm/hydra/pm/pmiserv/pmip_cb.c:883): assert (!closed) failed [proxy:0:[email protected]] HYDT_dmxu_poll_wait_for_event (../../../../mpich-master-v3.2-331-g44fd9c5f39e5/src/pm/hydra/tools/demux/demux_poll.c:77): callback returned error status [proxy:0:[email protected]] main (../../../../mpich-master-v3.2-331-g44fd9c5f39e5/src/pm/hydra/pm/pmiserv/pmip.c:202): demux engine error waiting for event [[email protected]] HYDT_bscu_wait_for_completion (../../../../mpich-master-v3.2-331-g44fd9c5f39e5/src/pm/hydra/tools/bootstrap/utils/bscu_wait.c:76): one of the processes terminated badly; aborting [[email protected]] HYDT_bsci_wait_for_completion (../../../../mpich-master-v3.2-331-g44fd9c5f39e5/src/pm/hydra/tools/bootstrap/src/bsci_wait.c:23): launcher returned error waiting for completion [[email protected]] HYD_pmci_wait_for_completion (../../../../mpich-master-v3.2-331-g44fd9c5f39e5/src/pm/hydra/pm/pmiserv/pmiserv_pmci.c:218): launcher returned error waiting for completion [[email protected]] main (../../../../mpich-master-v3.2-331-g44fd9c5f39e5/src/pm/hydra/ui/mpich/mpiexec.c:340): process manager error waiting for completion tyr spawn 112

I would be grateful if somebody can fix the problem. Please let me know, if you need more information. Thank you very much for any help in advance.

Kind regards

Siegmar _______________________________________________ discuss mailing list [email protected] To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

_______________________________________________ discuss mailing list [email protected] To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss