I've just tested OMPI 1.3.1 w/ GM 2.0.19 and the "ring" test from LAM's
examples directory.
I too see an early death. The backtrace is shown below.
-Paul
$ mpirun -mca btl gm,self -H pcp-i-1,pcp-i-2 ./ring
Rank 0 starting message around the ring -- 1st of 5
[pcp-i-1:27587] *** Process received signal ***
[pcp-i-1:27587] Signal: Segmentation fault (11)
[pcp-i-1:27587] Signal code: Address not mapped (1)
[pcp-i-1:27587] Failing at address: 0x5
[pcp-i-1:27587] [ 0] /lib/tls/libpthread.so.0 [0x401a38f0]
[pcp-i-1:27587] [ 1]
/usr/local/pkg/gm-2.0.19-2.4.20-8smp//lib/libgm.so.0(gm_handle_sent_tokens+0x67)
[0x404c6fbb]
[pcp-i-1:27587] [ 2]
/usr/local/pkg/gm-2.0.19-2.4.20-8smp//lib/libgm.so.0(_gm_unknown+0x42f)
[0x404cbf53]
[pcp-i-1:27587] [ 3]
/usr/local/pkg/gm-2.0.19-2.4.20-8smp//lib/libgm.so.0(gm_unknown+0x20)
[0x404cc068]
[pcp-i-1:27587] [ 4]
/opt/pcp-i/usr/local/pkg/openmpi-1.3.1/lib/openmpi/mca_btl_gm.so(mca_btl_gm_component_progress+0xc2)
[0x404b9c92]
[pcp-i-1:27587] [ 5]
/opt/pcp-i/usr/local/pkg/openmpi-1.3.1/lib/libopen-pal.so.0(opal_progress+0x79)
[0x40100c79]
[pcp-i-1:27587] [ 6]
/opt/pcp-i/usr/local/pkg/openmpi-1.3.1/lib/openmpi/mca_pml_ob1.so(mca_pml_ob1_recv+0x225)
[0x4049d9b5]
[pcp-i-1:27587] [ 7]
/opt/pcp-i/usr/local/pkg/openmpi-1.3.1/lib/libmpi.so.0(PMPI_Recv+0x159)
[0x40074ab9]
[pcp-i-1:27587] [ 8] ./ring(main+0xda) [0x8048876]
[pcp-i-1:27587] [ 9] /lib/tls/libc.so.6(__libc_start_main+0xe4) [0x42015704]
[pcp-i-1:27587] [10] ./ring(printf+0x31) [0x804870d]
[pcp-i-1:27587] *** End of error message ***
--------------------------------------------------------------------------
mpirun noticed that process rank 0 with PID 27587 on node pcp-i-1 exited
on signal 11 (Segmentation fault).
--------------------------------------------------------------------------
Patrick Geoffray wrote:
Hi Christian,
Christian Siebert wrote:
I just gave the new release 1.3.1 a go. While Ethernet and InfiniBand
seem to work properly, I noticed that Myrinet/GM compiles fine but
gives a segmentation violation in the first attempt to communicate
(MPI_Send in a simple "hello world" application). Is GM not supported
anymore or is it just too old so that nobody tested it?
GM itself is supported and maintenance releases are tested (no more
development releases), but Open-MPI/GM is not tested at the moment. GM
does not run on Myri-10G NICs, so we have to use a smaller pool of
machines with Myrinet 2000 NICs in them. Human usage and MTT runs for
Open-MPI/MX have priority and MTT for Open-MPI/GM has not run for a
while :-(
We will try to resume MTT testing with Open-MPI/GM when we have the
resources. In the meantime, we'll look into the segfault.
Patrick
_______________________________________________
devel mailing list
de...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/devel
--
Paul H. Hargrove phhargr...@lbl.gov
Future Technologies Group Tel: +1-510-495-2352
HPC Research Department Fax: +1-510-486-6900
Lawrence Berkeley National Laboratory
