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

smpd version mismatch
by Fernando Barreto 08 May '13

08 May '13

HI, my name is Fernando. I'm trying to execute a model in my pc. Mpich2 is already installed and running (I executed the cpi example with success). However when I try to execute my program I get the following error: Aborting: unable to connect to Boussines, smpd version mismatch

1 0

Re: [mpich-discuss] BLCR kernel module not present
by Pavan Balaji 06 May '13

06 May '13

On 05/01/2013 01:08 PM US Central Time, basma a.azeem wrote: > it works but i need to run these commands every time i shut down > i think that there is a method to have them installed permanently > > i am new to Linux and parallel programming also > > i need to find long running c/mpi applications to test blcr , the > applications i will use > to test blcr should have an enough long execution time , i found many > applications but with a short execution time . Please only use this mailing list for mpich related questions. -- Pavan -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] Issue regarding MPI_Comm_spawn_multiple
by Pavan Balaji 06 May '13

06 May '13

It's impossible to tell the problem from your description below. Can you please send us a small test program that reproduces your problem? -- Pavan On 05/06/2013 06:45 AM US Central Time, Mahesh Doijade wrote: > Hi, > I am using MPI_Comm_spawn_multiple for dynamic process creation > and passing different arguments to each process through this function, > but the code runs perfectly fine for some inputs and throws SEGFAULT for > some other inputs the code sample is > > > /*Arguments generation starts*/ > char ***procs_argv = procs_argv = (char ***) malloc(sizeof(char **) > * no_processes_to_launch); > > for(i=0; i < no_processes_to_launch; i++) > { > procs_argv[i] = ( char **)malloc(sizeof(char *) * 2); > procs_argv[i][0] = ( char *) malloc(sizeof(char)*4); > procs_argv[i][1] = ( char *) malloc(sizeof(char)*10); > sprintf(procs_argv[i][0], "%d", ppstat_ptr[i].no_of_pivots); > //Total no. of pivot variables > sprintf(procs_argv[i][1], "%d", > all_sizes_elems_fep[i+1]);//Total bytes of all sizes data. > } > > /*Arguments generation finish*/ > and process is launched as -- > MPI_Comm_spawn_multiple(no_processes_to_launch, cmds, procs_argv, > nprocs, infos, 0, MPI_COMM_WORLD, &worker_comm, errcodes); > > Error given when debugged through gdb : > > in _IO_vfprintf_internal (s=<value optimized out>, format=<value > optimized out>, ap=<value optimized out>) at vfprintf.c : 1603 , > process_string_arg(((struct printf_spec *)NULL)); > > > > > > -- > > Regards, > -- Mahesh Doijade > > > > > > _______________________________________________ > 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

Re: [mpich-discuss] mpich-3.0.4: building shared libraries w/cygwin
by Pavan Balaji 06 May '13

06 May '13

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. -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] where is two-phase I/O implemented
by Matthieu Dorier 06 May '13

06 May '13

Thanks Rob. Good to know that it's the messiest part, I managed to find everything I need so I'm confident with the rest of ROMIO :) Matthieu ----- Mail original ----- > De: "Rob Latham" <robl(a)mcs.anl.gov> > À: discuss(a)mpich.org > Envoyé: Lundi 6 Mai 2013 17:38:11 > Objet: Re: [mpich-discuss] where is two-phase I/O implemented > > On Wed, May 01, 2013 at 11:41:52AM +0200, Matthieu Dorier wrote: > > Btw how does Mpich knows which ADIO backend to use, in general? > > (and > > more specifically how does ADIOI_GEN_ReadStridedColl knows that it > > has to use ADIOI_BGL_ReadContig?) Does it have something to do with > > the fact that for instance with PVFS2 when the kernel module is not > > used, the filename should be prefixed by pvfs2:// ? If I want to > > develop a new ADIO backend, how can I make it such that filenames > > prefixed with myfilesystem:// will use this backend? (Maybe there's > > a documentation for that somewhere?) > > To add a bit upon what Dries answered > > Yes, if you prefix the file name, ROMIO will parse that and select > the > right set of function pointers. > > ROMIO will also try to guess the file system type (usually with a > 'statfs' call.) See the functions in adio/common/ad_fstype.c. > > It's a bit of a mess because on some platforms one calls statfs, > others one calls fstatfs, etc. If you squint around the preprocessor > directives I hope it eventually makes sense. Don't hesitate to ask > me > questions -- it's probably the messiest part of ROMIO ... > > ==rob > > > Thanks, > > > > Matthieu > > > > ----- Mail original ----- > > > De: "Pavan Balaji" <balaji(a)mcs.anl.gov> > > > À: discuss(a)mpich.org > > > Cc: "Matthieu Dorier" <matthieu.dorier(a)irisa.fr> > > > Envoyé: Mardi 30 Avril 2013 18:01:11 > > > Objet: Re: [mpich-discuss] where is two-phase I/O implemented > > > > > > > > > Please don't call it mpich3; it's just "mpich" :-). > > > > > > -- Pavan > > > > > > On 04/30/2013 10:32 AM US Central Time, Matthieu Dorier wrote: > > > > Hi, > > > > > > > > A question more for the mpich3 developers: > > > > > > > > If I were to implement an ADIO interface to a file system in > > > > mpich3, > > > > would I have to implement optimizations like two-phase I/O > > > > myself > > > > or are > > > > these optimizations implemented outside ADIO, independently of > > > > the > > > > file > > > > system? > > > > > > > > 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 > > > > > > > > > > -- > > > 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 > > -- > Rob Latham > Mathematics and Computer Science Division > Argonne National Lab, IL USA > _______________________________________________ > 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] where is two-phase I/O implemented
by Rob Latham 06 May '13

06 May '13

On Wed, May 01, 2013 at 11:41:52AM +0200, Matthieu Dorier wrote: > Btw how does Mpich knows which ADIO backend to use, in general? (and > more specifically how does ADIOI_GEN_ReadStridedColl knows that it > has to use ADIOI_BGL_ReadContig?) Does it have something to do with > the fact that for instance with PVFS2 when the kernel module is not > used, the filename should be prefixed by pvfs2:// ? If I want to > develop a new ADIO backend, how can I make it such that filenames > prefixed with myfilesystem:// will use this backend? (Maybe there's > a documentation for that somewhere?) To add a bit upon what Dries answered Yes, if you prefix the file name, ROMIO will parse that and select the right set of function pointers. ROMIO will also try to guess the file system type (usually with a 'statfs' call.) See the functions in adio/common/ad_fstype.c. It's a bit of a mess because on some platforms one calls statfs, others one calls fstatfs, etc. If you squint around the preprocessor directives I hope it eventually makes sense. Don't hesitate to ask me questions -- it's probably the messiest part of ROMIO ... ==rob > Thanks, > > Matthieu > > ----- Mail original ----- > > De: "Pavan Balaji" <balaji(a)mcs.anl.gov> > > ï¿½: discuss(a)mpich.org > > Cc: "Matthieu Dorier" <matthieu.dorier(a)irisa.fr> > > Envoyï¿½: Mardi 30 Avril 2013 18:01:11 > > Objet: Re: [mpich-discuss] where is two-phase I/O implemented > > > > > > Please don't call it mpich3; it's just "mpich" :-). > > > > -- Pavan > > > > On 04/30/2013 10:32 AM US Central Time, Matthieu Dorier wrote: > > > Hi, > > > > > > A question more for the mpich3 developers: > > > > > > If I were to implement an ADIO interface to a file system in > > > mpich3, > > > would I have to implement optimizations like two-phase I/O myself > > > or are > > > these optimizations implemented outside ADIO, independently of the > > > file > > > system? > > > > > > 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 > > > > > > > -- > > 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 -- Rob Latham Mathematics and Computer Science Division Argonne National Lab, IL USA

1 0

Issue regarding MPI_Comm_spawn_multiple
by Mahesh Doijade 06 May '13

06 May '13

Hi, I am using MPI_Comm_spawn_multiple for dynamic process creation and passing different arguments to each process through this function, but the code runs perfectly fine for some inputs and throws SEGFAULT for some other inputs the code sample is /*Arguments generation starts*/ char ***procs_argv = procs_argv = (char ***) malloc(sizeof(char **) * no_processes_to_launch); for(i=0; i < no_processes_to_launch; i++) { procs_argv[i] = ( char **)malloc(sizeof(char *) * 2); procs_argv[i][0] = ( char *) malloc(sizeof(char)*4); procs_argv[i][1] = ( char *) malloc(sizeof(char)*10); sprintf(procs_argv[i][0], "%d", ppstat_ptr[i].no_of_pivots); //Total no. of pivot variables sprintf(procs_argv[i][1], "%d", all_sizes_elems_fep[i+1]);//Total bytes of all sizes data. } /*Arguments generation finish*/ and process is launched as -- MPI_Comm_spawn_multiple(no_processes_to_launch, cmds, procs_argv, nprocs, infos, 0, MPI_COMM_WORLD, &worker_comm, errcodes); Error given when debugged through gdb : in _IO_vfprintf_internal (s=<value optimized out>, format=<value optimized out>, ap=<value optimized out>) at vfprintf.c : 1603 , process_string_arg(((struct printf_spec *)NULL)); -- Regards, -- Mahesh Doijade

1 0

Re: [mpich-discuss] mpich 3.0.4 build problem
by Pavan Balaji 03 May '13

03 May '13

Likewise for F77 and FC. Both should be set to gfortran-mp-4.7 if you want consistency. -- Pavan On 05/03/2013 01:24 PM US Central Time, Rajeev Thakur wrote: > In addition to FCFLAGS if you had set F77FLAGS to -m64, it probably would have worked in the third case. > > On May 3, 2013, at 1:17 PM, Thomas Evans wrote: > >> Operating System OSX 10.8.3 >> Xcode v4.6.1 >> MPICH 3.0.4 >> >> macports >> % port list automake libtool autoconf gcc47 >> automake @1.13.1 devel/automake >> libtool @2.4.2 devel/libtool >> autoconf @2.69 devel/autoconf >> gcc47 @4.7.3 lang/gcc47 >> >> >> ========================== My first attempt ============================= >> setenv CC "gcc-mp-4.7 -m64" >> setenv FC "gfortran-mp-4.7 -m64" >> setenv F77 "gfortran-mp-4.7 -m64" >> setenv F90 "gfortran-mp-4.7 -m64" >> ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis >> ... >> configure: ===== done with src/openpa configure ===== >> ... >> configure: error: F90 and F90FLAGS are replaced by FC and FCFLAGS respectively in this configure, please unset F90/F90FLAGS and set FC/FCFLAGS instead and rerun configure again >> >> Remarks: While odd, this is at least says what it wants. >> >> ========================== My second attempt ============================= >> unsetenv F77 >> unsetenv F90 >> ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis >> ... >> configure: ===== done with src/openpa configure ===== >> ... >> configure: error: **** Incompatible Fortran and C Object File Types! **** >> F77 Object File Type produced by "gfortran -O2" is : : Mach-O object i386. >> C Object File Type produced by "gcc-mp-4.7 -m64 -O2" is : : Mach-O 64-bit object x86_64. >> >> Remark: sometimes I see "gfortran", sometimes I see "gfortran-mp-4.7 -m64" >> >> ========================== My third attempt ============================= >> >> setenv FC "gfortran-mp-4.7" >> setenv FCFLAGS "-m64" >> ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis >> ... >> configure: ===== done with src/openpa configure ===== >> ... >> configure: error: **** Incompatible Fortran and C Object File Types! **** >> F77 Object File Type produced by "gfortran -O2" is : : Mach-O object i386. >> C Object File Type produced by "gcc-mp-4.7 -m64 -O2" is : : Mach-O 64-bit object x86_64. >> Remark: Well, that didn't help. >> >> ========================== My workaround ============================= >> In the end I created a csh script called gfortran at the start of my path and simply made it execute >> gfortran-mp-4.7 -m64 $argv >> This built okay and runs just fine. >> >> >> >> _______________________________________________ >> 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 > -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] mpich 3.0.4 build problem
by Rajeev Thakur 03 May '13

03 May '13

In addition to FCFLAGS if you had set F77FLAGS to -m64, it probably would have worked in the third case. On May 3, 2013, at 1:17 PM, Thomas Evans wrote: > Operating System OSX 10.8.3 > Xcode v4.6.1 > MPICH 3.0.4 > > macports > % port list automake libtool autoconf gcc47 > automake @1.13.1 devel/automake > libtool @2.4.2 devel/libtool > autoconf @2.69 devel/autoconf > gcc47 @4.7.3 lang/gcc47 > > > ========================== My first attempt ============================= > setenv CC "gcc-mp-4.7 -m64" > setenv FC "gfortran-mp-4.7 -m64" > setenv F77 "gfortran-mp-4.7 -m64" > setenv F90 "gfortran-mp-4.7 -m64" > ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis > ... > configure: ===== done with src/openpa configure ===== > ... > configure: error: F90 and F90FLAGS are replaced by FC and FCFLAGS respectively in this configure, please unset F90/F90FLAGS and set FC/FCFLAGS instead and rerun configure again > > Remarks: While odd, this is at least says what it wants. > > ========================== My second attempt ============================= > unsetenv F77 > unsetenv F90 > ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis > ... > configure: ===== done with src/openpa configure ===== > ... > configure: error: **** Incompatible Fortran and C Object File Types! **** > F77 Object File Type produced by "gfortran -O2" is : : Mach-O object i386. > C Object File Type produced by "gcc-mp-4.7 -m64 -O2" is : : Mach-O 64-bit object x86_64. > > Remark: sometimes I see "gfortran", sometimes I see "gfortran-mp-4.7 -m64" > > ========================== My third attempt ============================= > > setenv FC "gfortran-mp-4.7" > setenv FCFLAGS "-m64" > ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis > ... > configure: ===== done with src/openpa configure ===== > ... > configure: error: **** Incompatible Fortran and C Object File Types! **** > F77 Object File Type produced by "gfortran -O2" is : : Mach-O object i386. > C Object File Type produced by "gcc-mp-4.7 -m64 -O2" is : : Mach-O 64-bit object x86_64. > Remark: Well, that didn't help. > > ========================== My workaround ============================= > In the end I created a csh script called gfortran at the start of my path and simply made it execute > gfortran-mp-4.7 -m64 $argv > This built okay and runs just fine. > > > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss

1 0

mpich 3.0.4 build problem
by Thomas Evans 03 May '13

03 May '13

Operating System OSX 10.8.3 Xcode v4.6.1 MPICH 3.0.4 macports % port list automake libtool autoconf gcc47 automake @1.13.1 devel/automake libtool @2.4.2 devel/libtool autoconf @2.69 devel/autoconf gcc47 @4.7.3 lang/gcc47 ========================== My first attempt ============================= setenv CC "gcc-mp-4.7 -m64" setenv FC "gfortran-mp-4.7 -m64" setenv F77 "gfortran-mp-4.7 -m64" setenv F90 "gfortran-mp-4.7 -m64" ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis ... configure: ===== done with src/openpa configure ===== ... configure: error: F90 and F90FLAGS are replaced by FC and FCFLAGS respectively in this configure, please unset F90/F90FLAGS and set FC/FCFLAGS instead and rerun configure again Remarks: While odd, this is at least says what it wants. ========================== My second attempt ============================= unsetenv F77 unsetenv F90 ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis ... configure: ===== done with src/openpa configure ===== ... configure: error: **** Incompatible Fortran and C Object File Types! **** F77 Object File Type produced by "gfortran -O2" is : : Mach-O object i386. C Object File Type produced by "gcc-mp-4.7 -m64 -O2" is : : Mach-O 64-bit object x86_64. Remark: sometimes I see "gfortran", sometimes I see "gfortran-mp-4.7 -m64" ========================== My third attempt ============================= setenv FC "gfortran-mp-4.7" setenv FCFLAGS "-m64" ./configure --prefix=$HOME/mpi3gnu473 --enable-fc --disable-cxx --with-device=ch3:nemesis ... configure: ===== done with src/openpa configure ===== ... configure: error: **** Incompatible Fortran and C Object File Types! **** F77 Object File Type produced by "gfortran -O2" is : : Mach-O object i386. C Object File Type produced by "gcc-mp-4.7 -m64 -O2" is : : Mach-O 64-bit object x86_64. Remark: Well, that didn't help. ========================== My workaround ============================= In the end I created a csh script called gfortran at the start of my path and simply made it execute gfortran-mp-4.7 -m64 $argv This built okay and runs just fine.

1 0