Closed mpichbot closed 7 years ago
Originally by dinan on 2012-12-14 11:24:30 -0600
Attachment added: ftmain.f90
(3.7 KiB)
Test case #1
Originally by dinan on 2012-12-14 11:24:41 -0600
Attachment added: ftmain2.f90
(3.8 KiB)
Test case #2
Originally by dinan on 2012-12-17 14:03:19 -0600
Resolved in [3c720d0887e5aab1a21ada789717d29795ccbd46].
Originally by dinan on 2012-12-14 11:24:17 -0600
Reported by Bob Cernhous @ IBM:
I've had a hang reported on BG/Q after about 2K MPI_Comm_create's.
It hangs on the latest 2 releases (mpich2 v1.5.x and v1.4.x) on BG/Q.
It also hangs on linux: 64bit (MPI over PAMI) MPICH2 library.
On older mpich 1.? (BG/P) it failed with 'too many communicators' and didn't hang, which is what they expected.
It seems like it's stuck in the while (*context_id == 0) loop repeatedly calling allreduce and never settling on a context id in commutil.c. I didn't do a lot of debug but seems like it's in vanilla mpich code, not something we modified.
ftmain.f90 fails if you run it on >2k ranks (creates one comm per rank). This was the original customer testcase.
ftmain2.f90 fails by looping so you can run on fewer ranks.
I just noticed that with --np 1, I get the 'too many communicators' from ftmain2. But --np 2 and up hangs.