Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Development mailing list

Subject: [OMPI devel] r26565 (orte progress threads and libevent thread support by default) causing segfaults
From: Eugene Loh (eugene.loh_at_[hidden])
Date: 2012-06-09 15:06:06


With r26565:
     Enable orte progress threads and libevent thread support by default
Oracle MTT testing started showing new spawn_multiple failures. I've
only seen this in 64-bit. Here are two segfaults, both from Linux/x86
systems running over TCP:

This one with GNU compilers:
     [...]
     parent: MPI_Comm_spawn #300 return : 0
     [burl-ct-v20z-26:28518] *** Process received signal ***
     [burl-ct-v20z-26:28518] Signal: Segmentation fault (11)
     [burl-ct-v20z-26:28518] Signal code: Address not mapped (1)
     [burl-ct-v20z-26:28518] Failing at address: (nil)
     [burl-ct-v20z-26:28518] [ 0] /lib64/libpthread.so.0 [0x3a21c0e7c0]
     [burl-ct-v20z-26:28518] [ 1] /lib64/libc.so.6(memcpy+0x35)
[0x3a2107bde5]
     [burl-ct-v20z-26:28518] [ 2]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-v20z-26/ompi-tarball-testing/installs/smMv/install/lib/lib64/openmpi/mca_oob_tcp.so(mca_oob_tcp_msg_copy+0x58)
     [burl-ct-v20z-26:28518] [ 3]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-v20z-26/ompi-tarball-testing/installs/smMv/install/lib/lib64/openmpi/mca_oob_tcp.so
     [burl-ct-v20z-26:28518] [ 4]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-v20z-26/ompi-tarball-testing/installs/smMv/install/lib/lib64/openmpi/mca_oob_tcp.so(mca_oob_tcp_recv_nb+0x314)
     [burl-ct-v20z-26:28518] [ 5]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-v20z-26/ompi-tarball-testing/installs/smMv/install/lib/lib64/openmpi/mca_rml_oob.so(orte_rml_oob_recv_buffer_nb+0xff)
     [burl-ct-v20z-26:28518] [ 6]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-v20z-26/ompi-tarball-testing/installs/smMv/install/lib/lib64/openmpi/mca_dpm_orte.so
     [burl-ct-v20z-26:28518] [ 7]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-v20z-26/ompi-tarball-testing/installs/smMv/install/lib/lib64/libmpi.so.0(PMPI_Comm_spawn+0x2ee)
     [burl-ct-v20z-26:28518] [ 8] dynamic/loop_spawn [0x40120b]
     [burl-ct-v20z-26:28518] [ 9]
/lib64/libc.so.6(__libc_start_main+0xf4) [0x3a2101d994]
     [burl-ct-v20z-26:28518] [10] dynamic/loop_spawn [0x400dd9]
     [burl-ct-v20z-26:28518] *** End of error message ***

This one with Oracle Studio compilers:
     parent: MPI_Comm_spawn #0 return : 0
     parent: MPI_Comm_spawn #20 return : 0
     [burl-ct-x2200-12:02348] *** Process received signal ***
     [burl-ct-x2200-12:02348] Signal: Segmentation fault (11)
     [burl-ct-x2200-12:02348] Signal code: Address not mapped (1)
     [burl-ct-x2200-12:02348] Failing at address: 0x10
     [burl-ct-x2200-12:02348] [ 0] /lib64/libpthread.so.0 [0x318ac0de80]
     [burl-ct-x2200-12:02348] [ 1]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-x2200-12/ompi-tarball-testing/installs/Q7wT/install/lib/lib64/openmpi/mca_oob_tcp.so(mca_oob_tcp_msg_recv_handler+0xe3)
     [burl-ct-x2200-12:02348] [ 2]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-x2200-12/ompi-tarball-testing/installs/Q7wT/install/lib/lib64/openmpi/mca_oob_tcp.so
     [burl-ct-x2200-12:02348] [ 3]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-x2200-12/ompi-tarball-testing/installs/Q7wT/install/lib/lib64/libmpi.so.0
     [burl-ct-x2200-12:02348] [ 4]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-x2200-12/ompi-tarball-testing/installs/Q7wT/install/lib/lib64/libmpi.so.0(opal_libevent2019_event_base_loop+0x7c7)
     [burl-ct-x2200-12:02348] [ 5]
/workspace/tdontje/hpc/mtt-scratch/burl-ct-x2200-12/ompi-tarball-testing/installs/Q7wT/install/lib/lib64/libmpi.so.0
     [burl-ct-x2200-12:02348] [ 6] /lib64/libpthread.so.0 [0x318ac06307]
     [burl-ct-x2200-12:02348] [ 7] /lib64/libc.so.6(clone+0x6d)
[0x318a0d1ded]
     [burl-ct-x2200-12:02348] *** End of error message ***

Sometimes, I see a hang rather than a segfault.