Re: [OMPI users] Conflict between OpenMPI 1.7.3 and libevent

2013-11-05 Thread Jeff Squyres (jsquyres)
Excellent -- thank you! The real fix will be in 1.7.4. On Nov 5, 2013, at 5:42 PM, Dominique Orban wrote: > Many thanks for your patience. Patching Makefile.in did the job. I submitted > an updated formula to Homebrew. > > On 5 November, 2013 at 2:22:36 PM, Jeff Squyres (jsquyres) > (jsquy

Re: [OMPI users] Conflict between OpenMPI 1.7.3 and libevent

2013-11-05 Thread Dominique Orban
Many thanks for your patience. Patching Makefile.in did the job. I submitted an updated formula to Homebrew.  On 5 November, 2013 at 2:22:36 PM, Jeff Squyres (jsquyres) (jsquy...@cisco.com) wrote: > > On Nov 5, 2013, at 11:00 AM, Dominique Orban > wrote: > > > I'm installing in a fresh

Re: [OMPI users] proper use of MPI_Abort

2013-11-05 Thread Andrus, Brian Contractor
Jeff, We are using the latest version: 1.6.5 Brian Andrus ITACS/Research Computing Naval Postgraduate School Monterey, California voice: 831-656-6238 > -Original Message- > From: users [mailto:users-boun...@open-mpi.org] On Behalf Of Jeff > Squyres (jsquyres) > Sent: Tuesday, November

Re: [OMPI users] No spaces between concatenated string literals causes warning on GCC 4.9 prerelease in C++11 mode

2013-11-05 Thread Jeremiah Willcock
On Tue, 5 Nov 2013, Jeff Squyres (jsquyres) wrote: Thanks for reporting the issue.  Can you try this patch, and see if it fixes the issue on your trunk checkout? I don't have a trunk checkout; I just looked at the file in SVN and saw the same issue as in 1.6.1. Applying the patch to my inst

Re: [OMPI users] No spaces between concatenated string literals causes warning on GCC 4.9 prerelease in C++11 mode

2013-11-05 Thread Jeff Squyres (jsquyres)
Thanks for reporting the issue. Can you try this patch, and see if it fixes the issue on your trunk checkout? On Nov 5, 2013, at 11:12 AM, Jeremiah Willcock wrote: > In version 1.6.1 (and the same issue appears to be in SVN trunk), there are > some warnings when using from code compiled i

Re: [OMPI users] Conflict between OpenMPI 1.7.3 and libevent

2013-11-05 Thread Jeff Squyres (jsquyres)
On Nov 5, 2013, at 11:00 AM, Dominique Orban wrote: > I'm installing in a fresh location and include/event2 is still there. I don't > speak configure so I may be doing something wrong. It's enough to change the > Makefile.am, isn't it? No, it may well not be sufficient. If you don't have the

[OMPI users] No spaces between concatenated string literals causes warning on GCC 4.9 prerelease in C++11 mode

2013-11-05 Thread Jeremiah Willcock
In version 1.6.1 (and the same issue appears to be in SVN trunk), there are some warnings when using from code compiled in C++11 mode with more recent snapshots of GCC 4.9: .../include/mpi_portable_platform.h:102:38: warning: invalid suffix on literal; C++11 requires a space between literal a

Re: [OMPI users] Conflict between OpenMPI 1.7.3 and libevent

2013-11-05 Thread Dominique Orban
I'm installing in a fresh location and include/event2 is still there. I don't speak configure so I may be doing something wrong. It's enough to change the Makefile.am, isn't it? Here are my commands: $ cd /tmp/openmpi-1.7.3  # fresh download $ cp ../ompi-trunk/opal/mca/event/libevent2021/libeve

Re: [OMPI users] Conflict between OpenMPI 1.7.3 and libevent

2013-11-05 Thread Jeff Squyres (jsquyres)
On Nov 5, 2013, at 7:44 AM, Dominique Orban wrote: > Thanks for patch. Unless I'm completely confused, it is still not resolving > the issue. I downloaded Open-MPI 1.7.3 and replaced the Makefile.am with the > one obtained from a fresh svn checkout (I checked that your changes are in > there).

Re: [OMPI users] Conflict between OpenMPI 1.7.3 and libevent

2013-11-05 Thread Dominique Orban
Thanks for patch. Unless I'm completely confused, it is still not resolving the issue. I downloaded Open-MPI 1.7.3 and replaced the Makefile.am with the one obtained from a fresh svn checkout (I checked that your changes are in there). But `make install` still leaves me with an `event2` subdirec

Re: [OMPI users] proper use of MPI_Abort

2013-11-05 Thread Jeff Squyres (jsquyres)
You're correct -- you don't need to call MPI_Finalize after MPI_Abort. Can you cite what version of Open MPI you are using? On Nov 4, 2013, at 9:01 AM, "Andrus, Brian Contractor" wrote: > All, > > I have some sample code that has a syntax message and then an MPI_Abort call > if the program

[OMPI users] Fwd: experiencing some problems in MPI_reduce usage

2013-11-05 Thread Osman Khalid
Kindly discard below email. I have corrected the mistakes after running the program in parallel debug mode. Thanks. -- Forwarded message -- From: Osman Khalid List-Post: users@lists.open-mpi.org Date: Mon, Nov 4, 2013 at 11:02 PM Subject: experiencing some problems in MPI_reduce u

[OMPI users] experiencing some problems in MPI_reduce usage

2013-11-05 Thread Osman Khalid
Hi I am very new to Open MPI. I have made a small program that computes the sum of an array, by splitting array into pieces equal to the number of processes. The problem in my program is that each process is computing right sum of its share of the array, but the individually computed sums are not