[Bug 482386] New: openmpi mpirun crashes with Floating point exception
https://bugzilla.novell.com/show_bug.cgi?id=482386 Summary: openmpi mpirun crashes with Floating point exception Classification: openSUSE Product: openSUSE 11.1 Version: Final Platform: All OS/Version: openSUSE 11.1 Status: NEW Severity: Critical Priority: P5 - None Component: Development AssignedTo: pth@novell.com ReportedBy: felix.richter2@uni-rostock.de QAContact: qa@suse.de Found By: --- User-Agent: Mozilla/5.0 (X11; U; Linux i686; de; rv:1.9.0.6) Gecko/2009012700 SUSE/3.0.6-0.1.2 Firefox/3.0.6 mpirun crashes immediately, independent of the program to be run or the number of nodes or... same commandline works in a 11.0 installation tested on three different machines, i686 and x86_64. I don't know what more information could be needed (except the error message below, of course), please ask. Reproducible: Always Steps to Reproduce: $ mpirun time $ mpirun -np 2 time $ mpirun python Actual Results: Error message: [myhost:03001] *** Process received signal *** [myhost:03001] Signal: Floating point exception (8) [myhost:03001] Signal code: Integer divide-by-zero (1) [myhost:03001] Failing at address: 0xb7e55226 [myhost:03001] [ 0] [0xffffe410] [myhost:03001] [ 1] /lib/libc.so.6 [0xb7e55226] [myhost:03001] [ 2] /lib/libc.so.6(cfree+0x9c) [0xb7e56f3c] [myhost:03001] [ 3] /usr/lib/libcrypto.so.0.9.8(CRYPTO_free+0x2d) [0xb7af5d8d] [myhost:03001] [ 4] /usr/lib/libssl.so.0.9.8 [0xb7c540eb] [myhost:03001] [ 5] /usr/lib/libssl.so.0.9.8(ssl_create_cipher_list+0x4c1) [0xb7c54781] [myhost:03001] [ 6] /usr/lib/libssl.so.0.9.8(SSL_CTX_new+0x1d0) [0xb7c4df20] [myhost:03001] [ 7] /usr/lib/libldap-2.4.so.2 [0xb7daa645] [myhost:03001] [ 8] /usr/lib/libldap-2.4.so.2 [0xb7daacaa] [myhost:03001] [ 9] /usr/lib/libldap-2.4.so.2 [0xb7daaeb2] [myhost:03001] [10] /usr/lib/libldap-2.4.so.2(ldap_int_tls_start+0x47) [0xb7dab047] [myhost:03001] [11] /usr/lib/libldap-2.4.so.2(ldap_install_tls+0x45) [0xb7dab225] [myhost:03001] [12] /lib/libnss_ldap.so.2 [0xb7dc1b19] [myhost:03001] [13] /lib/libnss_ldap.so.2 [0xb7dc2249] [myhost:03001] [14] /lib/libnss_ldap.so.2 [0xb7dc2b6e] [myhost:03001] [15] /lib/libnss_ldap.so.2 [0xb7dc333a] [myhost:03001] [16] /lib/libnss_ldap.so.2(_nss_ldap_getpwuid_r+0x70) [0xb7dc39b0] [myhost:03001] [17] /lib/libnss_compat.so.2 [0xb7de079c] [myhost:03001] [18] /lib/libnss_compat.so.2 [0xb7de0986] [myhost:03001] [19] /lib/libnss_compat.so.2(_nss_compat_getpwuid_r+0xc6) [0xb7de0d06] [myhost:03001] [20] /lib/libc.so.6(getpwuid_r+0xd2) [0xb7e81e72] [myhost:03001] [21] /lib/libc.so.6(getpwuid+0x77) [0xb7e817a7] [myhost:03001] [22] /usr/lib/mpi/gcc/openmpi/lib/libopen-rte.so.0(orte_sys_info+0xc2) [0xb803fa52] [myhost:03001] [23] /usr/lib/mpi/gcc/openmpi/lib/libopen-rte.so.0(orte_init_stage1+0xc5) [0xb8039845] [myhost:03001] [24] /usr/lib/mpi/gcc/openmpi/lib/libopen-rte.so.0(orte_system_init+0x23) [0xb803ccd3] [myhost:03001] [25] /usr/lib/mpi/gcc/openmpi/lib/libopen-rte.so.0(orte_init+0x58) [0xb80394d8] [myhost:03001] [26] mpirun(orterun+0x15e) [0x804a2de] [myhost:03001] [27] mpirun(main+0x22) [0x804a176] [myhost:03001] [28] /lib/libc.so.6(__libc_start_main+0xe5) [0xb7dff705] [myhost:03001] [29] mpirun [0x804a0c1] [myhost:03001] *** End of error message *** Gleitkomma-Ausnahme -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=482386 User felix.richter2@uni-rostock.de added comment https://bugzilla.novell.com/show_bug.cgi?id=482386#c1 --- Comment #1 from Felix Richter <felix.richter2@uni-rostock.de> 2009-03-06 01:33:42 MST --- I packaged openmpi version 1.2.9 in my OBS home project using the opensuse 11.1 spec file -- same result, crashes immediately -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=482386 User felix.richter2@uni-rostock.de added comment https://bugzilla.novell.com/show_bug.cgi?id=482386#c2 --- Comment #2 from Felix Richter <felix.richter2@uni-rostock.de> 2009-03-06 03:51:13 MST --- Now I packaged openmpi version 1.3 -- this one works (specfile has to be changed to include some more files)! -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=482386 User pth@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=482386#c3 Philipp Thomas <pth@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |pth@novell.com AssignedTo|pth@novell.com |jjolly@novell.com --- Comment #3 from Philipp Thomas <pth@novell.com> 2009-03-06 06:01:33 MST --- Assigning to openmpi maintainer. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=482386 User jjolly@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=482386#c4 John Jolly <jjolly@novell.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P5 - None |P2 - High Status|NEW |ASSIGNED --- Comment #4 from John Jolly <jjolly@novell.com> 2009-03-09 06:19:04 MST --- I will investigate upgrading the openMPI. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=482386 User jjolly@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=482386#c5 --- Comment #5 from John Jolly <jjolly@novell.com> 2009-03-16 08:40:31 MST --- I have verified that the bug is not an issue with OpenMPI v1.3, but it is because they have removed the use of getpwuid. Since this is a problem with a shipping product, I won't be able to upgrade from 1.2.8 to 1.3. I can make this upgrade for OpenSUSE 11.2, though. The problem with getpwuid seems to be with the OpenMPI build process. I am still investigating. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply@novell.com