- discuss - mailman3.cels.anl.gov

Re: [mpich-discuss] assert (!closed) failed
by Wesley Bland 13 May '13

13 May '13

You might try looking at the HPCC documentation in the README that's distributed with the package, along with the website: http://icl.cs.utk.edu/hpcc/software/index.html If you have trouble with the specific benchmark, I would suggest contacting the HPCC people at the address provided on their website. Wesley On May 13, 2013, at 1:15 PM, Pavan Balaji <balaji(a)mcs.anl.gov> wrote: > > Please keep discuss(a)mpich.org cc'ed. We tend to ignore emails sent to > developers directly. > > On 05/13/2013 01:13 PM US Central Time, Reza Bakhshayeshi wrote: >> Thanks for replying, but I can run hpcc on a single node successfully. >> Isn't it mpich responsibility to run it distributive? >> Can you suggest how I can find the main problem? > > Is it an MPI program? > > From the error message, all I can tell is that the application process > died for some reason (e.g., it might have segfaulted), and mpich cleaned > it up for your convenience. > > -- Pavan > > -- > Pavan Balaji > http://www.mcs.anl.gov/~balaji > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss

1 0

GCC 4.8.0 warnings
by Yauheni Zelenko 13 May '13

13 May '13

Hi! I compiled recent nightly build (3.0.4 as of May 13) with GCC 4.8.0. There are some warnings which will be good idea to fix. Eugene.

1 0

Re: [mpich-discuss] assert (!closed) failed
by Reza Bakhshayeshi 13 May '13

13 May '13

Sorry, I just hit the reply button and didn't notice it's going to your inbox only. Yes, It is a MPI program and running it on a single node is somehow useless. Thanks anyway. On 13 May 2013 22:45, Pavan Balaji <balaji(a)mcs.anl.gov> wrote: > > Please keep discuss(a)mpich.org cc'ed. We tend to ignore emails sent to > developers directly. > > On 05/13/2013 01:13 PM US Central Time, Reza Bakhshayeshi wrote: > > Thanks for replying, but I can run hpcc on a single node successfully. > > Isn't it mpich responsibility to run it distributive? > > Can you suggest how I can find the main problem? > > Is it an MPI program? > > From the error message, all I can tell is that the application process > died for some reason (e.g., it might have segfaulted), and mpich cleaned > it up for your convenience. > > -- Pavan > > -- > Pavan Balaji > http://www.mcs.anl.gov/~balaji >

1 0

Re: [mpich-discuss] assert (!closed) failed
by Pavan Balaji 13 May '13

13 May '13

Please keep discuss(a)mpich.org cc'ed. We tend to ignore emails sent to developers directly. On 05/13/2013 01:13 PM US Central Time, Reza Bakhshayeshi wrote: > Thanks for replying, but I can run hpcc on a single node successfully. > Isn't it mpich responsibility to run it distributive? > Can you suggest how I can find the main problem? Is it an MPI program? >From the error message, all I can tell is that the application process died for some reason (e.g., it might have segfaulted), and mpich cleaned it up for your convenience. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] assert (!closed) failed
by Pavan Balaji 13 May '13

13 May '13

On 05/13/2013 12:35 PM US Central Time, Reza Bakhshayeshi wrote: > and hydra via: > apt-get install hydra You don't need this -- this is probably some other package. > and I want to run hpcc on two nodes. When I enter: > localadmin@ubuntu-benchmark01:~/hpcc-1.4.2$ mpiexec -f hosts2 -n 2 ./hpcc > I receive the following error: > > ===================================================================================== > = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES Here's the key part -- it sounds like your application process died badly. Sounds like a problem with your benchmark. > = EXIT CODE: 139 > = CLEANING UP REMAINING PROCESSES > = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES The rest is just cleanup of the remaining processes. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

assert (!closed) failed
by Reza Bakhshayeshi 13 May '13

13 May '13

hi I have installed mpich2 via: apt-get install mpich2 mpich-bin mpich2-doc and hydra via: apt-get install hydra and I want to run hpcc on two nodes. When I enter: localadmin@ubuntu-benchmark01:~/hpcc-1.4.2$ mpiexec -f hosts2 -n 2 ./hpcc I receive the following error: ===================================================================================== = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES = EXIT CODE: 139 = CLEANING UP REMAINING PROCESSES = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES ===================================================================================== [proxy:0:1@ubuntu-benchmark02] HYD_pmcd_pmip_control_cmd_cb (./pm/pmiserv/pmip_cb.c:928): assert (!closed) failed [proxy:0:1@ubuntu-benchmark02] HYDT_dmxu_poll_wait_for_event (./tools/demux/demux_poll.c:77): callback returned error status [proxy:0:1@ubuntu-benchmark02] main (./pm/pmiserv/pmip.c:226): demux engine error waiting for event [mpiexec@ubuntu-benchmark01] HYDT_bscu_wait_for_completion (./tools/bootstrap/utils/bscu_wait.c:70): one of the processes terminated badly; aborting [mpiexec@ubuntu-benchmark01] HYDT_bsci_wait_for_completion (./tools/bootstrap/src/bsci_wait.c:23): launcher returned error waiting for completion [mpiexec@ubuntu-benchmark01] HYD_pmci_wait_for_completion (./pm/pmiserv/pmiserv_pmci.c:191): launcher returned error waiting for completion [mpiexec@ubuntu-benchmark01] main (./ui/mpich/mpiexec.c:405): process manager error waiting for completion Would you mind please help me solve the problem? Regards, Reza

1 0

Re: [mpich-discuss] mpich-3.0.4: building shared libraries w/cygwin
by Pavan Balaji 10 May '13

10 May '13

On 05/10/2013 07:34 PM US Central Time, Brad Aagaard wrote: > As shown in the config.log that I attached in my original email, all of > the tests for shared libraries give a result of "yes". I am pretty sure > this is related to weak symbols on cygwin and libtool. How libtool > decides to build shared libraries seems pretty opaque to me. That's my guess too, but we don't have a cygwin setup for me to try it out. Did you try setting CC=gcc CXX=g++ to make sure the cc compiler is not pointing to something nasty? > Has building shared libraries for mpich-3.0.4 been tested on cygwin? If > so, what were the configure parameters and what version of cygwin was used? Not really. Windows support has been dropped since the 1.5.x releases. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] mpich-3.0.4: building shared libraries w/cygwin
by Brad Aagaard 10 May '13

10 May '13

On 05/10/2013 05:26 PM, Pavan Balaji wrote: > > On 05/08/2013 07:06 PM US Central Time, Brad Aagaard wrote: >> On 05/06/2013 05:45 PM, Pavan Balaji wrote: >>> >>> On 05/02/2013 04:36 PM US Central Time, Brad Aagaard wrote: >>>> What configure/make arguments should I use to build shared libraries on >>>> cygwin? >>> >>> Just use --enable-shared without any additional --enable-sharedlibs >>> argument. You might also want to give --disable-static if you don't >>> want static libraries. >>> >> >> Dropping --enable-sharedlibs from configure gives me the same result as >> when I use --enable-sharedlibs=libtool or >> --enable-sharedlibs=gcc-cygwin. I end up with static libraries and no >> shared libraries even though I also use --enable-shared --disable-static. > > This is strange. > >> Is there a way to check the Makefile or libtool to see why it doesn't >> want to build shared libraries? > > You could look through the config.log files to see what's going on. As shown in the config.log that I attached in my original email, all of the tests for shared libraries give a result of "yes". I am pretty sure this is related to weak symbols on cygwin and libtool. How libtool decides to build shared libraries seems pretty opaque to me. Has building shared libraries for mpich-3.0.4 been tested on cygwin? If so, what were the configure parameters and what version of cygwin was used? Thanks, Brad

1 0

Re: [mpich-discuss] mpich-3.0.4: building shared libraries w/cygwin
by Pavan Balaji 10 May '13

10 May '13

On 05/08/2013 07:06 PM US Central Time, Brad Aagaard wrote: > On 05/06/2013 05:45 PM, Pavan Balaji wrote: >> >> On 05/02/2013 04:36 PM US Central Time, Brad Aagaard wrote: >>> What configure/make arguments should I use to build shared libraries on >>> cygwin? >> >> Just use --enable-shared without any additional --enable-sharedlibs >> argument. You might also want to give --disable-static if you don't >> want static libraries. >> > > Dropping --enable-sharedlibs from configure gives me the same result as > when I use --enable-sharedlibs=libtool or > --enable-sharedlibs=gcc-cygwin. I end up with static libraries and no > shared libraries even though I also use --enable-shared --disable-static. This is strange. > Is there a way to check the Makefile or libtool to see why it doesn't > want to build shared libraries? You could look through the config.log files to see what's going on. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] Nemesis mode
by Pavan Balaji 09 May '13

09 May '13

Just configure the latest version of mpich. It'll figure out the machine architecture and use nemesis shared memory by default. -- Pavan On 05/09/2013 01:28 AM US Central Time, Bob Ilgner wrote: > Dear all, > > I am using a large shared memory machine ( > 100 cores) and want to use > mpi in nemesis mode. How do I do that ? There is some material in the > faq about using ch3. Where does that come in ? > > Regards, > > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0