October 2016 - discuss - mailman3.cels.anl.gov

Re: [mpich-discuss] length of errcodes array in MPI_Comm_spawn_multiple
by Dorier, Matthieu 10 Oct '16

10 Oct '16

Thanks! Matthieu ________________________________ From: Thakur, Rajeev [thakur(a)mcs.anl.gov] Sent: Monday, October 10, 2016 2:23 PM To: discuss(a)mpich.org Subject: Re: [mpich-discuss] length of errcodes array in MPI_Comm_spawn_multiple It’s at the top of pg 382 of MPI 3.1 [cid:[email protected]] On Oct 10, 2016, at 2:19 PM, Dorier, Matthieu <mdorier(a)anl.gov<mailto:[email protected]>> wrote: Hi, I can't find this information in online documentations: In MPI_Comm_spawn_multiple, what is the size of the errcodes output array? the sum of the elements in the maxprocs input array? Thanks! Matthieu _______________________________________________ discuss mailing list discuss(a)mpich.org<mailto:[email protected]> 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] length of errcodes array in MPI_Comm_spawn_multiple
by Thakur, Rajeev 10 Oct '16

10 Oct '16

It’s at the top of pg 382 of MPI 3.1 [cid:[email protected]] On Oct 10, 2016, at 2:19 PM, Dorier, Matthieu <mdorier(a)anl.gov<mailto:[email protected]>> wrote: Hi, I can't find this information in online documentations: In MPI_Comm_spawn_multiple, what is the size of the errcodes output array? the sum of the elements in the maxprocs input array? Thanks! Matthieu _______________________________________________ discuss mailing list discuss(a)mpich.org<mailto:[email protected]> 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

length of errcodes array in MPI_Comm_spawn_multiple
by Dorier, Matthieu 10 Oct '16

10 Oct '16

Hi, I can't find this information in online documentations: In MPI_Comm_spawn_multiple, what is the size of the errcodes output array? the sum of the elements in the maxprocs input array? Thanks! Matthieu _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Errors building MPICH on Windows 10 Ubuntu subsystem
by Damian Rouson 10 Oct '16

10 Oct '16

All, In attempting to build MPICH 3.1.4 or 3.2 in the Windows 10 Subsystem for Linux, which provides a full Ubuntu bash shell, I receive the following error message after configuring: mpich-3.2/confdb/missing: line 81: aclocal-1.15: command not found WARNING: 'aclocal-1.15' is missing on your system. You should only need it if you modified 'acinclude.m4' or 'configure.ac <http://configure.ac/>' or m4 files included by 'configure.ac <http://configure.ac/>'. The 'aclocal' program is part of the GNU Automake package: <http://www.gnu.org/software/automake <http://www.gnu.org/software/automake>> It also requires GNU Autoconf, GNU m4 and Perl in order to run: <http://www.gnu.org/software/autoconf <http://www.gnu.org/software/autoconf>> <http://www.gnu.org/software/m4/ <http://www.gnu.org/software/m4/>> <http://www.perl.org/ <http://www.perl.org/>> make: *** [/home/rouson/Desktop-shortcut/opencoarrays/prerequisites/downloads/mpich-3.2/aclocal.m4] Error 127 Prior to configuring, I installed aclocal installed via “sudo apt-get install automake,” which provided aclocal-1.14. I attempted unsuccessfully to build aclocal-1.15, but the discussion at the following URL suggests that the problem might relate to time stamps on file MPICH tar ball: http://stackoverflow.com/questions/18769770/user-of-autotools-generated-tar… <http://stackoverflow.com/questions/18769770/user-of-autotools-generated-tar…> I tried several of the steps suggested at the above link, but each leads to the following error message during the make step: $ make ( cd ./src/binding/fortran/mpif_h && ./buildiface ) Replacing mpif.h.in <http://mpif.h.in/> ( cd ./src/binding/fortran/use_mpi && ./buildiface ) Replacing mpi_base.f90.in <http://mpi_base.f90.in/> Replacing mpi_constants.f90.in <http://mpi_constants.f90.in/> Replacing mpi_t1.f90 ( cd . && rm -f .err unusederr.txt ; rm -rf .tmp ) ( cd . && \ ./maint/extracterrmsgs -careful=unusederr.txt \ -skip=src/util/multichannel/mpi.c \ `cat maint/errmsgdirs` > .tmp 2>.err ) ( cd . && if test -s .err ; then rm -f .tmp ; cat .err ; exit 1 ; fi ) Warning: attempt to redefine **fail. Duplicate ignored. make: *** [src/mpi/errhan/defmsg.h] Error 1 Can anyone advise me about how to resolve this problem? Information on the Windows 10 Ubuntu distribution is here: http://www.howtogeek.com/249966/how-to-install-and-use-the-linux-bash-shell… My goa is to support users of OpenCoarrays (https://github.com/sourceryinstitute/opencoarrays) which uses MPICH to build and execute coarray Fortran codes with the GNU Fortran compiler (gfortran). Best Regards, ________________________________ Damian Rouson, Ph.D., P.E. President, Sourcery Institute http://www.sourceryinstitute.org +1-510-600-2992 (mobile) _______________________________________________ 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] 'nonblocking' test needs revision
by Kenneth Raffenetti 05 Oct '16

05 Oct '16

Thanks, patch applied. On 10/05/2016 08:18 AM, Jeff Hammond wrote: > Attached, pushed to mpich-review/fix-mpich-test-dt-types and pushed > to https://github.com/jeffhammond/mpich/tree/fix-mpich-test-dt-types. > > Jeff > > > > On Wed, Oct 5, 2016 at 3:25 AM, Chris Ward <TJCW(a)uk.ibm.com > <mailto:[email protected]>> wrote: > > I'm using the MPICH BVT tests to test a build of OpenMPI. While > doing that, I find that 'nonblocking.c' needs to be revised as follows > // int *types = NULL; > MPI_Datatype *types = NULL; > > // types = malloc(size * sizeof(int)); > types = malloc(size * sizeof(MPI_Datatype)); > > Presumably it works as-is for mpich because MPI_Datatype is an > 'int'; but this isn't the case for other MPI implementations. Would > one of the mpich developers care to make these changes to the source > code ? > *T J (Chris) Ward, IBM Research. > Scalable Data-Centric Computing - Active Storage Fabrics - IBM > System BlueGene > IBM United Kingdom Ltd., Hursley Park, Winchester, Hants, SO21 2JN > 011-44-1962-818679 > LinkedIn **_https://www.linkedin.com/profile/view?id=60628729_* > <https://www.linkedin.com/profile/view?id=60628729>* > ResearchGate **_https://www.researchgate.net/profile/T_Ward2_* > <https://www.researchgate.net/profile/T_Ward2>* > **_ > _**_IBM System BlueGene Research_* > <http://www.research.ibm.com/bluegene/>***_ > _**_IBM System BlueGene Marketing_* > <http://www-03.ibm.com/systems/deepcomputing/bluegene/>* > **_ > _**_IBM Resources for Global Servants_* > <http://hurgsa.ibm.com/~tjcw/Compete/>***_ > _**_IBM Branded Products_* <http://www.ibm.com/shop>***_IBM Branded > Swag_* <http://ibmlogogear.com/default.aspx>** > <http://www.ibm.com/developerworks/linux/library/l-bluegene/index.html> > > > UNIX in the Cloud - Find A Place Where There's Room To Grow, with > the original Open Standard. _Free Trial Here Today_ > <http://sdf.lonestar.org/index.cgi?telnet> > New Lamps For Old - _Diskless Remote Boot Linux_ > <http://drbl.sourceforge.net/>from _National Center for > High-Performance Computing, Taiwan_ <http://www.nchc.org.tw/en/> > > > > Unless stated otherwise above: > IBM United Kingdom Limited - Registered in England and Wales with > number 741598. > Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire > PO6 3AU > > _______________________________________________ > discuss mailing list discuss(a)mpich.org <mailto:[email protected]> > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > <https://lists.mpich.org/mailman/listinfo/discuss> > > > > > -- > Jeff Hammond > jeff.science(a)gmail.com <mailto:[email protected]> > http://jeffhammond.github.io/ > > > _______________________________________________ > 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] 'nonblocking' test needs revision
by Jeff Hammond 05 Oct '16

05 Oct '16

Attached, pushed to mpich-review/fix-mpich-test-dt-types and pushed to https://github.com/jeffhammond/mpich/tree/fix-mpich-test-dt-types. Jeff On Wed, Oct 5, 2016 at 3:25 AM, Chris Ward <TJCW(a)uk.ibm.com> wrote: > I'm using the MPICH BVT tests to test a build of OpenMPI. While doing > that, I find that 'nonblocking.c' needs to be revised as follows > // int *types = NULL; > MPI_Datatype *types = NULL; > > // types = malloc(size * sizeof(int)); > types = malloc(size * sizeof(MPI_Datatype)); > > Presumably it works as-is for mpich because MPI_Datatype is an 'int'; but > this isn't the case for other MPI implementations. Would one of the mpich > developers care to make these changes to the source code ? > > > > > *T J (Chris) Ward, IBM Research. Scalable Data-Centric Computing - Active > Storage Fabrics - IBM System BlueGene IBM United Kingdom Ltd., Hursley > Park, Winchester, Hants, SO21 2JN 011-44-1962-818679 LinkedIn * > *https://www.linkedin.com/profile/view?id=60628729* > <https://www.linkedin.com/profile/view?id=60628729> > * ResearchGate **https://www.researchgate.net/profile/T_Ward2* > <https://www.researchgate.net/profile/T_Ward2> > > *IBM System BlueGene Research* <http://www.research.ibm.com/bluegene/> > *IBM System BlueGene Marketing* > <http://www-03.ibm.com/systems/deepcomputing/bluegene/> > > *IBM Resources for Global Servants* <http://hurgsa.ibm.com/~tjcw/Compete/> > *IBM Branded Products* <http://www.ibm.com/shop> *IBM Branded Swag* > <http://ibmlogogear.com/default.aspx> > <http://www.ibm.com/developerworks/linux/library/l-bluegene/index.html> > > UNIX in the Cloud - Find A Place Where There's Room To Grow, with the > original Open Standard. *Free Trial Here Today* > <http://sdf.lonestar.org/index.cgi?telnet> > New Lamps For Old - *Diskless Remote Boot Linux* > <http://drbl.sourceforge.net/> from *National Center for High-Performance > Computing, Taiwan* <http://www.nchc.org.tw/en/> > > Unless stated otherwise above: > IBM United Kingdom Limited - Registered in England and Wales with number > 741598. > Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > -- Jeff Hammond jeff.science(a)gmail.com http://jeffhammond.github.io/ _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

'nonblocking' test needs revision
by Chris Ward 05 Oct '16

05 Oct '16

I'm using the MPICH BVT tests to test a build of OpenMPI. While doing that, I find that 'nonblocking.c' needs to be revised as follows // int *types = NULL; MPI_Datatype *types = NULL; // types = malloc(size * sizeof(int)); types = malloc(size * sizeof(MPI_Datatype)); Presumably it works as-is for mpich because MPI_Datatype is an 'int'; but this isn't the case for other MPI implementations. Would one of the mpich developers care to make these changes to the source code ? T J (Chris) Ward, IBM Research. Scalable Data-Centric Computing - Active Storage Fabrics - IBM System BlueGene IBM United Kingdom Ltd., Hursley Park, Winchester, Hants, SO21 2JN 011-44-1962-818679 LinkedIn https://www.linkedin.com/profile/view?id=60628729 ResearchGate https://www.researchgate.net/profile/T_Ward2 IBM System BlueGene Research IBM System BlueGene Marketing IBM Resources for Global Servants IBM Branded Products IBM Branded Swag UNIX in the Cloud - Find A Place Where There's Room To Grow, with the original Open Standard. Free Trial Here Today New Lamps For Old - Diskless Remote Boot Linux from National Center for High-Performance Computing, Taiwan Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU _______________________________________________ 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] Why does MPICH perform busy wait?
by Tim Jammer 02 Oct '16

02 Oct '16

hi Thank you for your insights. I am still wondering, why the busy wait is in the MPI_Barrier. As the purpose of the Barrier is to just wait for all involved processes. I think the latentency introduced with one or another way of saving energy is not such important in the Barrier. For all other communication it might be best to perform a short busy wait in order to guarantee a maximum message rate. Was it considered before to introduce an "energy aware wait" for the MPI_Barrier function? So that the user can still enjoy the high performance solution for all other communication. Regards Tim Jammer Am 29.09.2016 um 17:08 schrieb Rob Latham: > Your proposal introduces latency. Data centers want power efficiency, > but several large classes of applications care about message rate. _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0