Discussion:
[OMPI users] OpenMPI 3.0.1 debug crashes eclipse due to mpirun -display-map bug
Érico
2018-04-24 01:54:14 UTC
Permalink
Hello,

Because of the error below, Eclipse is not able to run PTP debugger with OpenMPI 3.0.1.

Can someone help me??

I use CentOS 7.

Thanks!

Erico

---------------------

[***@centos64 ContainerServiceDebug]$ mpirun -mca orte_show_resolved_nodenames 1 -display-map -np 1 pwd
Data for JOB [23564,1] offset 0 Total slots allocated 4
[centos64:98315] *** Process received signal ***
[centos64:98315] Signal: Segmentation fault (11)
[centos64:98315] Signal code: Address not mapped (1)
[centos64:98315] Failing at address: (nil)
[centos64:98315] [ 0] /usr/lib64/libpthread.so.0(+0xf100)[0x7f74537c9100]
[centos64:98315] [ 1] /usr/local/lib/libopen-rte.so.40(orte_dt_print_node+0x451)[0x7f7454a6a35f]
[centos64:98315] [ 2] /usr/local/lib/libopen-pal.so.40(opal_dss_print+0x68)[0x7f745474d1d5]
[centos64:98315] [ 3] /usr/local/lib/libopen-rte.so.40(orte_dt_print_map+0x517)[0x7f7454a6b834]
[centos64:98315] [ 4] /usr/local/lib/libopen-pal.so.40(opal_dss_print+0x68)[0x7f745474d1d5]
[centos64:98315] [ 5] /usr/local/lib/libopen-rte.so.40(orte_rmaps_base_display_map+0x53b)[0x7f7454aefd0c]
[centos64:98315] [ 6] /usr/local/lib/libopen-rte.so.40(orte_odls_base_default_construct_child_list+0x13f7)[0x7f7454acf090]
[centos64:98315] [ 7] /usr/local/lib/openmpi/mca_odls_default.so(+0x2c7c)[0x7f744d5f3c7c]
[centos64:98315] [ 8] /usr/local/lib/libopen-rte.so.40(orte_daemon_recv+0x6d7)[0x7f7454a9bdb5]
[centos64:98315] [ 9] /usr/local/lib/libopen-rte.so.40(orte_rml_base_process_msg+0x2e5)[0x7f7454afbde8]
[centos64:98315] [10] /usr/local/lib/libopen-pal.so.40(opal_libevent2022_event_base_loop+0x8fc)[0x7f74547a246c]
[centos64:98315] [11] mpirun[0x4016f7]
[centos64:98315] [12] mpirun[0x4010e0]
[centos64:98315] [13] /usr/lib64/libc.so.6(__libc_start_main+0xf5)[0x7f7453419b15]
[centos64:98315] [14] mpirun[0x400ff9]
[centos64:98315] *** End of error message ***
Segmentation fault (core dumped)
[***@centos64 ContainerServiceDebug]$
g***@rist.or.jp
2018-04-24 06:47:25 UTC
Permalink
Thanks for the report,

This is indeed an issue i fixed at https://github.com/open-mpi/ompi/pull/5088



Cheers,

Gilles

----- Original Message -----
Post by Érico
Hello,
Because of the error below, Eclipse is not able to run PTP debugger with OpenMPI 3.0.1.
Can someone help me??
I use CentOS 7.
Thanks!
Erico
---------------------
_nodenames 1 -display-map -np 1 pwd
Post by Érico
Data for JOB [23564,1] offset 0 Total slots allocated 4
[centos64:98315] *** Process received signal ***
[centos64:98315] Signal: Segmentation fault (11)
[centos64:98315] Signal code: Address not mapped (1)
[centos64:98315] Failing at address: (nil)
[centos64:98315] [ 0] /usr/lib64/libpthread.so.0(+0xf100)[
0x7f74537c9100]
Post by Érico
[centos64:98315] [ 1] /usr/local/lib/libopen-rte.so.40(orte_dt_print_
node+0x451)[0x7f7454a6a35f]
Post by Érico
[centos64:98315] [ 2] /usr/local/lib/libopen-pal.so.40(opal_dss_print+
0x68)[0x7f745474d1d5]
Post by Érico
[centos64:98315] [ 3] /usr/local/lib/libopen-rte.so.40(orte_dt_print_
map+0x517)[0x7f7454a6b834]
Post by Érico
[centos64:98315] [ 4] /usr/local/lib/libopen-pal.so.40(opal_dss_print+
0x68)[0x7f745474d1d5]
Post by Érico
[centos64:98315] [ 5] /usr/local/lib/libopen-rte.so.40(orte_rmaps_base
_display_map+0x53b)[0x7f7454aefd0c]
Post by Érico
[centos64:98315] [ 6] /usr/local/lib/libopen-rte.so.40(orte_odls_base_
default_construct_child_list+0x13f7)[0x7f7454acf090]
Post by Érico
[centos64:98315] [ 7] /usr/local/lib/openmpi/mca_odls_default.so(+
0x2c7c)[0x7f744d5f3c7c]
Post by Érico
[centos64:98315] [ 8] /usr/local/lib/libopen-rte.so.40(orte_daemon_
recv+0x6d7)[0x7f7454a9bdb5]
Post by Érico
[centos64:98315] [ 9] /usr/local/lib/libopen-rte.so.40(orte_rml_base_
process_msg+0x2e5)[0x7f7454afbde8]
Post by Érico
[centos64:98315] [10] /usr/local/lib/libopen-pal.so.40(opal_
libevent2022_event_base_loop+0x8fc)[0x7f74547a246c]
Post by Érico
[centos64:98315] [11] mpirun[0x4016f7]
[centos64:98315] [12] mpirun[0x4010e0]
[centos64:98315] [13] /usr/lib64/libc.so.6(__libc_start_main+0xf5)[
0x7f7453419b15]
Post by Érico
[centos64:98315] [14] mpirun[0x400ff9]
[centos64:98315] *** End of error message ***
Segmentation fault (core dumped)
_______________________________________________
users mailing list
https://lists.open-mpi.org/mailman/listinfo/users
Loading...