March 2013 - discuss - mailman3.cels.anl.gov

Possible bug: --enable-g=all configure option does not include all other --enable-g options
by Jeremiah Willcock 11 Mar '13

11 Mar '13

In MPICH 3.0.2, the help message for the --enable-g=all configure option states that it enables "All of the above choices", but in fact (according to configure.ac) it does not enable the "handle" or "memarena" debugging flags. Is that the intended behavior? Thank you for your help. -- Jeremiah Willcock

1 0

Hydra fills in the MPIR_proctable[] incorrectly with multiple processes per node
by John DelSignore 11 Mar '13

11 Mar '13

Hi, I'm pretty sure this is a MPICH Hydra bug, but I wanted to ask this group before I go through the trouble of figuring out how to file a MPICH bug report, which I think requires me to create an MPICH Trac account, which I don't know how to do. As far as I can tell, Hydra fills in the MPIR_proctable[] incorrectly with multiple processes per node. The index into the MPIR_proctable[] is supposed to be the MPI process's rank in MPI_COMM_WORLD. To demonstrate this problem. I created a simple MPI "hello world" program where each MPI process prints out its rank and pid; I attached it to this email. This is the version of MPICH I am using: shell% /home/mware/argonne/mpich2/1.4.1p1/x86_64-linux/bin/mpirun --version|head -3 HYDRA build details: Version: 1.4.1p1 Release Date: Thu Sep 1 13:53:02 CDT 2011 shell% I ran the code under TotalView using 2 nodes and 6 processes (3 per node). I enabled logging so that TotalView would output the contents of the MPIR_proctable[] as it extracted it from the mpirun process. Here is the output of the run: shell% tv8cli \ -verbosity errors \ -x15 \ -parallel_stop no \ -debug_file debug.log \ -args \ /home/mware/argonne/mpich2/1.4.1p1/x86_64-linux/bin/mpirun \ -hosts 127.0.0.2,127.0.0.3 \ -n 6 \ ./mpichbug d1.<> dcont Hello from rank 0 of 6, getpid()==2691 Hello from rank 1 of 6, getpid()==2729 Hello from rank 2 of 6, getpid()==2693 Hello from rank 3 of 6, getpid()==2730 Hello from rank 4 of 6, getpid()==2694 Hello from rank 5 of 6, getpid()==2734 d1.<> quit -force shell% Grepping for "proctable" in the debugger's log file shows the contents of the MPIR_proctable[]: shell% grep proctable debug.log mpir_proctable_t::create: extracting hostname/execname/pids for 6 processes mpir_proctable_t::create: MPIR_proctable[0]: host_name(0x0056be20)="127.0.0.2", executable_name(0x0056be80)="./mpichbug", pid=2691 mpir_proctable_t::create: MPIR_proctable[1]: host_name(0x0056be00)="127.0.0.2", executable_name(0x0056bde0)="./mpichbug", pid=2693 mpir_proctable_t::create: MPIR_proctable[2]: host_name(0x005859a0)="127.0.0.2", executable_name(0x0056c080)="./mpichbug", pid=2694 mpir_proctable_t::create: MPIR_proctable[3]: host_name(0x005856b0)="127.0.0.3", executable_name(0x005856d0)="./mpichbug", pid=2729 mpir_proctable_t::create: MPIR_proctable[4]: host_name(0x005856f0)="127.0.0.3", executable_name(0x00585710)="./mpichbug", pid=2730 mpir_proctable_t::create: MPIR_proctable[5]: host_name(0x00585730)="127.0.0.3", executable_name(0x00585750)="./mpichbug", pid=2734 shell% Matching up the pid values shows that the MPIR_proctable[rank] does not match the rank returned to the program by MPI_Comm_rank() for some of the MPI processes. Here's the MPI rank to MPIR_proctable rank mapping: 0 => 0 1 => 3 2 => 1 3 => 4 4 => 2 5 => 5 Do you agree that this is an MPICH Hydra bug? Any advice on how to create an MPICH Trac account so that I can report the bug? Thanks, John D.

1 0

Re: [mpich-discuss] Help with eclipse juno and mpich2
by Jayesh Krishna 11 Mar '13

11 Mar '13

Hi, Are you able to build/compile your MPI program? You can run your MPI job on a single pc (mpiexec -n 2 mympipgm.exe). Regards, Jayesh ----- Original Message ----- From: "George Papageorgakis" <ngeopapas(a)gmail.com> To: "Jayesh Krishna" <jayesh(a)mcs.anl.gov> Cc: discuss(a)mpich.org Sent: Monday, March 11, 2013 11:40:13 AM Subject: Re: [mpich-discuss] Help with eclipse juno and mpich2 Well. I think i have done that correctly. Is it even possible to run a debugger without connecting to a cluster? Is there a simulation for mpi or something that can be done only on one pc? Maybe its not possible to work in x-platform?

1 0

Re: [mpich-discuss] Help with eclipse juno and mpich2
by George Papageorgakis 11 Mar '13

11 Mar '13

Well. I think i have done that correctly. Is it even possible to run a debugger without connecting to a cluster? Is there a simulation for mpi or something that can be done only on one pc? Maybe its not possible to work in x-platform?

1 0

Re: [mpich-discuss] Help with eclipse juno and mpich2
by Jayesh Krishna 11 Mar '13

11 Mar '13

Hi, You need to find out how to specify the include and lib directories for a project on Eclipse, and add "C:\Program Files\MPICH2\include" as your include directory and "C:\Program Files\MPICH2\lib" are your lib directory. After that you need to specify mpi.lib as the library to link your program. (PS: Look into eclipse help - http://help.eclipse.org/galileo/index.jsp?topic=/org.eclipse.cdt.doc.user/r… OR just "google" for the information.) Regards, Jayesh ----- Original Message ----- From: "George Papageorgakis" <ngeopapas(a)gmail.com> To: discuss(a)mpich.org Sent: Wednesday, March 6, 2013 7:44:23 AM Subject: [mpich-discuss] Help with eclipse juno and mpich2 Hello people! I am new in mpi. I have been trying to setup eclipse to play with mpich2 in windows 7. I have installed mpich2 and managed to make the eclipse editor identify mpi libs, functions etc but it wont compile correctly. I am using the example of mpi_heat2D and i would like to run it on a ubuntu cluster of my university. I have done so with a kinda slow and weary methodology using WinSCP to copy the .c file from my pc to a cluster pc, compile it and run it via console not being able to see any graphs etc, only the output file with the results. I would like to know what can i do to run and see everything from eclipse, including graphs for optimization etc. I didnt find much help on the internet and i am kinda frustrated. Is what i ask possible? Thanks in advance! _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] Running MPICH installer in "hidden mode"
by Jayesh Krishna 11 Mar '13

11 Mar '13

Hi, Did you check out the different options in msiexec ? Regards, Jayesh ----- Original Message ----- From: "Luiz Carlos da Costa Junior" <lcjunior(a)ufrj.br> To: "MPICH Discuss" <mpich-discuss(a)mcs.anl.gov> Sent: Monday, March 11, 2013 9:50:29 AM Subject: [mpich-discuss] Running MPICH installer in "hidden mode" Hi all, Is there any way to run MPICH Windows installer in "hidden mode", i.e., using default options (maybe choosing another passphrase)? I deliver MPICH together with my application and I would like to make the installation process easier. Any hint on this? Regards, Luiz _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Running MPICH installer in "hidden mode"
by Luiz Carlos da Costa Junior 11 Mar '13

11 Mar '13

Hi all, Is there any way to run MPICH Windows installer in "hidden mode", i.e., using default options (maybe choosing another passphrase)? I deliver MPICH together with my application and I would like to make the installation process easier. Any hint on this? Regards, Luiz

1 0

Re: [mpich-discuss] Semantics of asynchronous I/O operations in MPI3
by Matthieu Dorier 08 Mar '13

08 Mar '13

Thank you very much! Matthieu ----- Mail original ----- > De: "Rajeev Thakur" <thakur(a)mcs.anl.gov> > À: discuss(a)mpich.org > Envoyé: Vendredi 8 Mars 2013 20:19:38 > Objet: Re: [mpich-discuss] Semantics of asynchronous I/O operations in MPI3 > > The file pointers are updated in the order in which the functions are > called. The functions don't have to complete in the order issued. > For example, if you had called MPI_Waitany instead of Waitall, you > can't say which iwrite will be indicated as completed when the > waitany returns. > > Rajeev > > On Mar 8, 2013, at 8:04 AM, Matthieu Dorier wrote: > > > Hi, > > > > I can't manage to find a clear answer to this question in the MPI3 > > standard book; > > Assume we chain several asynchronous I/O operation on the same > > file, such as > > > > MPI_File_iwrite(...) > > MPI_File_set_view(...) > > MPI_File_iwrite(...) > > MPI_File_iwrite(...) > > MPI_Wait_all(all requests from previous calls) > > > > (equivalently with any other collective or independent asynchronous > > file accesses) > > Is an implementation of the MPI3 standard supposed to perform these > > operations in the same order than the functions are called, or can > > they be re-ordered if they are not in conflict (regions written > > are distinct)? > > > > Thanks! > > > > Matthieu Dorier > > PhD student at ENS Cachan Brittany and IRISA > > http://people.irisa.fr/Matthieu.Dorier > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss >

1 0

Re: [mpich-discuss] Semantics of asynchronous I/O operations in MPI3
by Rajeev Thakur 08 Mar '13

08 Mar '13

The file pointers are updated in the order in which the functions are called. The functions don't have to complete in the order issued. For example, if you had called MPI_Waitany instead of Waitall, you can't say which iwrite will be indicated as completed when the waitany returns. Rajeev On Mar 8, 2013, at 8:04 AM, Matthieu Dorier wrote: > Hi, > > I can't manage to find a clear answer to this question in the MPI3 standard book; > Assume we chain several asynchronous I/O operation on the same file, such as > > MPI_File_iwrite(...) > MPI_File_set_view(...) > MPI_File_iwrite(...) > MPI_File_iwrite(...) > MPI_Wait_all(all requests from previous calls) > > (equivalently with any other collective or independent asynchronous file accesses) > Is an implementation of the MPI3 standard supposed to perform these operations in the same order than the functions are called, or can they be re-ordered if they are not in conflict (regions written are distinct)? > > Thanks! > > Matthieu Dorier > PhD student at ENS Cachan Brittany and IRISA > http://people.irisa.fr/Matthieu.Dorier > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re: [mpich-discuss] MPICH2-1.5 installation problems on Linux
by Dave Goodell 08 Mar '13

08 Mar '13

This is an old ia64 machine with a very old (gcc) compiler. Generally speaking, modern MPICH simply won't work on this platform because we do not have one available for testing. I can see at least two or three issues that you will hit, and it's unfortunately not worth the time investment to fix them. -Dave On Mar 8, 2013, at 12:32 AM CST, psj58285@gmail <psj58285(a)gmail.com> wrote: > Dear Sir, > > I am trying to install MPICH2-1.5 on a Linux System. It succeeded at the configure step, but when I was building MPICH2, it failed. > > My information: > > baoys@altix-1:~> uname -a > Linux altix-1 2.6.5-7.244-sn2 #1 SMP Mon Dec 12 18:32:25 UTC 2005 ia64 ia64 ia64 GNU/Linux > > baoys@altix-1:~> icc -v > Version 11.1 > > baoys@altix-1:~> gcc -v > Reading specs from /usr/lib/gcc-lib/ia64-suse-linux/3.3.3/specs > Configured with: ../configure --enable-threads=posix --prefix=/usr --with-local-prefix=/usr/local --infodir=/usr/share/info --mandir=/usr/share/man --enable-languages=c,c++,f77,objc,java,ada --disable-checking --libdir=/usr/lib --enable-libgcj --with-gxx-include-dir=/usr/include/g++ --with-slibdir=/lib --with-system-zlib --enable-shared --enable-__cxa_atexit ia64-suse-linux > Thread model: posix > gcc version 3.3.3 (SuSE Linux) > > Here are the attachments, including the c.txt and the m.txt(generate by 'make VERBOSE=1 2>&1 | tee m.txt '). > > Looking forward to your help! > > Thanks, > Sijie Pan > > psj58285@gmail > <c.txt><m.txt>_______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss

1 0