Re: [OMPI devel] mpirun --prefix question

2007-03-22 Thread Jeff Squyres
FWIW, I believe that we had intended --prefix to handle simple cases which is why this probably doesn't work for you. But as long as the different prefixes are specified for different nodes, it could probably be made to work. Which launcher are you using this with? On Mar 21, 2007, at 1

Re: [OMPI devel] mpirun --prefix question

2007-03-22 Thread David Daniel
This is a development system for roadrunner using ssh. David On Mar 22, 2007, at 5:19 AM, Jeff Squyres wrote: FWIW, I believe that we had intended --prefix to handle simple cases which is why this probably doesn't work for you. But as long as the different prefixes are specified for different

Re: [OMPI devel] mpirun --prefix question

2007-03-22 Thread Ralph Castain
We had a nice chat about this on the OpenRTE telecon this morning. The question of what to do with multiple prefix's has been a long-running issue, most recently captured in bug trac report #497. The problem is that prefix is intended to tell us where to find the ORTE/OMPI executables, and therefor

[OMPI devel] RH Enterprise Linux issue

2007-03-22 Thread Greg Watson
Is this a known problem? Building ompi 1.2 on RHEL4: ./configure --with-devel-headers --without-threads (actually tried without '--without-threads' too, but no change) $ mpirun -np 2 test [beth:06029] *** Process received signal *** [beth:06029] Signal: Segmentation fault (11) [beth:06029] Sign

Re: [OMPI devel] RH Enterprise Linux issue

2007-03-22 Thread Greg Watson
Oh, and this is a single x86 machine. Just trying to launch locally. $uname -a Linux 2.6.9-42.0.2.ELsmp #1 SMP Thu Aug 17 18:00:32 EDT 2006 i686 i686 i386 GNU/Linux Greg On Mar 22, 2007, at 12:17 PM, Greg Watson wrote: Is this a known problem? Building ompi 1.2 on RHEL4: ./configure --w

Re: [OMPI devel] RH Enterprise Linux issue

2007-03-22 Thread Jeff Squyres
No, not a known problem -- my cluster is RHEL4U4 -- I use it for many thousands of runs of the OMPI v1.2 branch every day... Can you see where it's dying in orte_init_stage1? On Mar 22, 2007, at 2:17 PM, Greg Watson wrote: Is this a known problem? Building ompi 1.2 on RHEL4: ./configure --

Re: [OMPI devel] RH Enterprise Linux issue

2007-03-22 Thread Greg Watson
gdb says this: #0 0x2e342e33 in ?? () #1 0xb7fe1d31 in orte_pls_base_select () from /usr/local/lib/libopen- rte.so.0 #2 0xb7fc50cb in orte_init_stage1 () from /usr/local/lib/libopen- rte.so.0 #3 0xb7fc84be in orte_system_init () from /usr/local/lib/libopen- rte.so.0 #4 0xb7fc4cee in orte

Re: [OMPI devel] RH Enterprise Linux issue

2007-03-22 Thread Jeff Squyres
Yes, if you could recompile with debugging, that would be great. What launcher are you trying to use? On Mar 22, 2007, at 2:35 PM, Greg Watson wrote: gdb says this: #0 0x2e342e33 in ?? () #1 0xb7fe1d31 in orte_pls_base_select () from /usr/local/lib/libopen- rte.so.0 #2 0xb7fc50cb in orte_

Re: [OMPI devel] RH Enterprise Linux issue

2007-03-22 Thread Greg Watson
Scratch that. The problem was an installation over an old copy of ompi. Obviously picking up some old stuff. Sorry for the disturbance. Back to the bat cave... Greg On Mar 22, 2007, at 12:46 PM, Jeff Squyres wrote: Yes, if you could recompile with debugging, that would be great. What launc

Re: [OMPI devel] RH Enterprise Linux issue

2007-03-22 Thread Jeff Squyres
Whew! You had me worried there for a minute... :-) On Mar 22, 2007, at 3:15 PM, Greg Watson wrote: Scratch that. The problem was an installation over an old copy of ompi. Obviously picking up some old stuff. Sorry for the disturbance. Back to the bat cave... Greg On Mar 22, 2007, at 12:46

Re: [OMPI devel] mpirun --prefix question

2007-03-22 Thread David Daniel
OK. This sounds sensible. Thanks, David On Mar 22, 2007, at 10:38 AM, Ralph Castain wrote: We had a nice chat about this on the OpenRTE telecon this morning. The question of what to do with multiple prefix's has been a long- running issue, most recently captured in bug trac report #497. The p