May 2014 - discuss - mailman3.cels.anl.gov

Re: [mpich-discuss] unable to bind socket to port
by Jed Brown 16 May '14

16 May '14

张国熙 <altriaex86(a)gmail.com> writes: > Hi, all > > I'm using SLEPc as an eigensolver, which implemented MPICH for multiprocess > usage. > > I compile SLEPc code into a .so library B. > > My executable program A calls .so library B, and .so libraryB calls > SLEPclibrary. > I used to put MPI_Init() only in A. But it will got stuck. Attach a debugger when it "gets stuck". > Then I try to put it in A,B( and C,SLEPc init will call this > function). SlepcInitialize checks (via PetscInitialize) whether MPI has been initialized. You can call MPI_Init before all of them, but you still have to call SlepcInitialize collectively.

1 0

Re: [mpich-discuss] unable to bind socket to port
by Balaji, Pavan 16 May '14

16 May '14

This doesn’t look like an error with multiple initializations. If you do that MPICH will throw an error saying "Cannot call MPI_INIT or MPI_INIT_THREAD more than once”. It looks like you might not be using SLEPc correctly. Perhaps ask for help on their mailing list? — Pavan On May 16, 2014, at 6:56 PM, 张国熙 <altriaex86(a)gmail.com> wrote: > Hi, all > > I'm using SLEPc as an eigensolver, which implemented MPICH for multiprocess usage. > > I compile SLEPc code into a .so library B. > > My executable program A calls .so library B, and .so libraryB calls SLEPclibrary. > I used to put MPI_Init() only in A. But it will got stuck. Then I try to put it in A,B( and C,SLEPc init will call this function). > > INTERNAL ERROR: Invalid error class (59) encountered while returning from > MPI_Init. Please file a bug report. > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > INTERNAL ERROR: Invalid error class (59) encountered while returning from > MPI_Init. Please file a bug report. > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > [cli_1]: aborting job: > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > INTERNAL ERROR: Invalid error class (59) encountered while returning from > MPI_Init. Please file a bug report. > INTERNAL ERROR: Invalid error class (59) encountered while returning from > MPI_Init. Please file a bug report. > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > [cli_3]: aborting job: > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > [cli_0]: aborting job: > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > [cli_2]: aborting job: > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > > =================================================================================== > = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES > = EXIT CODE: 1 > = CLEANING UP REMAINING PROCESSES > = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES > =================================================================================== > root@altria-Aspire-5830TG:/home/altria/software/rememode-1.1alpha/Examples/c++# /home/altria/software/petsc-3.4.4/arch-linux2-c-debug/bin/mpiexec -np 4 ./fd_guide > INTERNAL ERROR: Invalid error class (59) encountered while returning from > MPI_Init. Please file a bug report. > INTERNAL ERROR: Invalid error class (59) encountered while returning from > MPI_Init. Please file a bug report. > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > [cli_1]: aborting job: > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > INTERNAL ERROR: Invalid error class (59) encountered while returning from > MPI_Init. Please file a bug report. > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > [cli_2]: aborting job: > Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: > (unknown)(): unable to bind socket to port > > =================================================================================== > = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES > = EXIT CODE: 1 > = CLEANING UP REMAINING PROCESSES > = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES > =================================================================================== > > I think it is because I called MPI_Init for more than one times. Because if I only call it in A, and C(SLEPc call), code in A,B will be executed, but stuck at C. If I call it in A,B,C. Code in A will be executed, and stop at B. > > I think even if I call MPI_Init for more than one times, it should return to me something like "MPI_Init has been called". Do anyone know how to fix it?Thanks. > > Your sincerely, > Guoxi > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss

1 0

unable to bind socket to port
by 张国熙 16 May '14

16 May '14

Hi, all I'm using SLEPc as an eigensolver, which implemented MPICH for multiprocess usage. I compile SLEPc code into a .so library B. My executable program A calls .so library B, and .so libraryB calls SLEPclibrary. I used to put MPI_Init() only in A. But it will got stuck. Then I try to put it in A,B( and C,SLEPc init will call this function). INTERNAL ERROR: Invalid error class (59) encountered while returning from MPI_Init. Please file a bug report. Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port INTERNAL ERROR: Invalid error class (59) encountered while returning from MPI_Init. Please file a bug report. Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port [cli_1]: aborting job: Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port INTERNAL ERROR: Invalid error class (59) encountered while returning from MPI_Init. Please file a bug report. INTERNAL ERROR: Invalid error class (59) encountered while returning from MPI_Init. Please file a bug report. Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port [cli_3]: aborting job: Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port [cli_0]: aborting job: Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port [cli_2]: aborting job: Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port =================================================================================== = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES = EXIT CODE: 1 = CLEANING UP REMAINING PROCESSES = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES =================================================================================== root@altria-Aspire-5830TG:/home/altria/software/rememode-1.1alpha/Examples/c++# /home/altria/software/petsc-3.4.4/arch-linux2-c-debug/bin/mpiexec -np 4 ./fd_guide INTERNAL ERROR: Invalid error class (59) encountered while returning from MPI_Init. Please file a bug report. INTERNAL ERROR: Invalid error class (59) encountered while returning from MPI_Init. Please file a bug report. Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port [cli_1]: aborting job: Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port INTERNAL ERROR: Invalid error class (59) encountered while returning from MPI_Init. Please file a bug report. Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port [cli_2]: aborting job: Fatal error in MPI_Init: Unknown error. Please file a bug report., error stack: (unknown)(): unable to bind socket to port =================================================================================== = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES = EXIT CODE: 1 = CLEANING UP REMAINING PROCESSES = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES =================================================================================== I think it is because I called MPI_Init for more than one times. Because if I only call it in A, and C(SLEPc call), code in A,B will be executed, but stuck at C. If I call it in A,B,C. Code in A will be executed, and stop at B. I think even if I call MPI_Init for more than one times, it should return to me something like "MPI_Init has been called". Do anyone know how to fix it?Thanks. Your sincerely, Guoxi

1 0

Re: [mpich-discuss] bug in MPI_File_write_all?
by Rajeev Thakur 16 May '14

16 May '14

Can you send us a small test program that can be compiled and run? Rajeev On May 16, 2014, at 11:42 AM, CANELA-XANDRI Oriol <Oriol.CAnela-Xandri(a)roslin.ed.ac.uk> wrote: > I also tried with the 3.0.4 (for which I have a distribution package) version with same results. If you think the result could change with 3.1, I will try with it. > > Oriol > > > -- > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > > -----Original Message----- > From: discuss-bounces(a)mpich.org [mailto:[email protected]] On Behalf Of Rajeev Thakur > Sent: 16 May 2014 17:14 > To: discuss(a)mpich.org > Subject: Re: [mpich-discuss] bug in MPI_File_write_all? > > 1.4.1 is a very old version of MPICH. Can you try with the latest release, 3.1? > > Rajeev > > On May 16, 2014, at 8:50 AM, CANELA-XANDRI Oriol <Oriol.CAnela-Xandri(a)roslin.ed.ac.uk> wrote: > >> Hello, >> >> I am writing a function for storing a block cyclic distributed array using MPI_Type_create_darray, MPI_File_set_view, and MPI_File_write_all functions. After some testing I found that for some particular cases, sometimes, some of the values of the array have been not stored properly on the file. The error seems random, but always affects the same number in the same position (where instead of a number, there is a nan, it seems that the position is not written). I tried with other implementations (e.g. openmpi) and seems everything works well (I am using MPICH2 Version 1.4.1). >> >> For instance, I saw the error for a matrix with 5 rows and 4 columns. Block size for rows = 1 and for columns = 3. There are 9 MPI processes distributed in a 3x3 array. My function is (where nBlockRows/nBlockCols define the size of the blocks, nGlobRows/nGlobCols define the global size of the matrix, nProcRows/nProcCols define the dimensions of the process grid, and fname is the name of the file.): >> >> void Matrix::writeMatrixMPI(std::string fname) { int dims[] = >> {this->nGlobRows, this->nGlobCols}; int dargs[] = {this->nBlockRows, >> this->nBlockCols}; int distribs[] = {MPI_DISTRIBUTE_CYCLIC, >> MPI_DISTRIBUTE_CYCLIC}; int dim[] = {communicator->nProcRows, >> communicator->nProcCols}; char nat[] = "native"; int rc; >> MPI_Datatype dcarray; MPI_File cFile; MPI_Status status; >> >> MPI_Type_create_darray(communicator->mpiNumTasks, >> communicator->mpiRank, 2, dims, distribs, dargs, dim, >> MPI_ORDER_FORTRAN, MPI_DOUBLE, &dcarray); MPI_Type_commit(&dcarray); >> >> std::vector<char> fn(fname.begin(), fname.end()); >> fn.push_back('\0'); rc = MPI_File_open(MPI_COMM_WORLD, &fn[0], >> MPI_MODE_EXCL | MPI_MODE_CREATE | MPI_MODE_WRONLY, MPI_INFO_NULL, >> &cFile); if(rc){ >> ss << "Error: Failed to open file: " << rc; >> misc.error(ss.str(), 0); >> } >> else >> { >> MPI_File_set_view(cFile, 0, MPI_DOUBLE, dcarray, nat, MPI_INFO_NULL); >> MPI_File_write_all(cFile, this->m, this->nRows*this->nCols, MPI_DOUBLE, &status); >> } >> MPI_File_close(&cFile); >> MPI_Type_free(&dcarray); >> } >> >> If I call the function a repeated number of times without deleting the file between calls, the file seems to be always correct. If I delete the file between calls or create a file with a different name each time, then I see the error in some of the files. The error is a nan always in the same position. For me it is very strange, but as I said, with other implementations, I do not have the problem. >> >> -- >> The University of Edinburgh is a charitable body, registered in >> Scotland, with registration number SC005336. >> >> _______________________________________________ >> 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 > _______________________________________________ > 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] bug in MPI_File_write_all?
by CANELA-XANDRI Oriol 16 May '14

16 May '14

I also tried with the 3.0.4 (for which I have a distribution package) version with same results. If you think the result could change with 3.1, I will try with it. Oriol -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. -----Original Message----- From: discuss-bounces(a)mpich.org [mailto:[email protected]] On Behalf Of Rajeev Thakur Sent: 16 May 2014 17:14 To: discuss(a)mpich.org Subject: Re: [mpich-discuss] bug in MPI_File_write_all? 1.4.1 is a very old version of MPICH. Can you try with the latest release, 3.1? Rajeev On May 16, 2014, at 8:50 AM, CANELA-XANDRI Oriol <Oriol.CAnela-Xandri(a)roslin.ed.ac.uk> wrote: > Hello, > > I am writing a function for storing a block cyclic distributed array using MPI_Type_create_darray, MPI_File_set_view, and MPI_File_write_all functions. After some testing I found that for some particular cases, sometimes, some of the values of the array have been not stored properly on the file. The error seems random, but always affects the same number in the same position (where instead of a number, there is a nan, it seems that the position is not written). I tried with other implementations (e.g. openmpi) and seems everything works well (I am using MPICH2 Version 1.4.1). > > For instance, I saw the error for a matrix with 5 rows and 4 columns. Block size for rows = 1 and for columns = 3. There are 9 MPI processes distributed in a 3x3 array. My function is (where nBlockRows/nBlockCols define the size of the blocks, nGlobRows/nGlobCols define the global size of the matrix, nProcRows/nProcCols define the dimensions of the process grid, and fname is the name of the file.): > > void Matrix::writeMatrixMPI(std::string fname) { int dims[] = > {this->nGlobRows, this->nGlobCols}; int dargs[] = {this->nBlockRows, > this->nBlockCols}; int distribs[] = {MPI_DISTRIBUTE_CYCLIC, > MPI_DISTRIBUTE_CYCLIC}; int dim[] = {communicator->nProcRows, > communicator->nProcCols}; char nat[] = "native"; int rc; > MPI_Datatype dcarray; MPI_File cFile; MPI_Status status; > > MPI_Type_create_darray(communicator->mpiNumTasks, > communicator->mpiRank, 2, dims, distribs, dargs, dim, > MPI_ORDER_FORTRAN, MPI_DOUBLE, &dcarray); MPI_Type_commit(&dcarray); > > std::vector<char> fn(fname.begin(), fname.end()); > fn.push_back('\0'); rc = MPI_File_open(MPI_COMM_WORLD, &fn[0], > MPI_MODE_EXCL | MPI_MODE_CREATE | MPI_MODE_WRONLY, MPI_INFO_NULL, > &cFile); if(rc){ > ss << "Error: Failed to open file: " << rc; > misc.error(ss.str(), 0); > } > else > { > MPI_File_set_view(cFile, 0, MPI_DOUBLE, dcarray, nat, MPI_INFO_NULL); > MPI_File_write_all(cFile, this->m, this->nRows*this->nCols, MPI_DOUBLE, &status); > } > MPI_File_close(&cFile); > MPI_Type_free(&dcarray); > } > > If I call the function a repeated number of times without deleting the file between calls, the file seems to be always correct. If I delete the file between calls or create a file with a different name each time, then I see the error in some of the files. The error is a nan always in the same position. For me it is very strange, but as I said, with other implementations, I do not have the problem. > > -- > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > _______________________________________________ > 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] bug in MPI_File_write_all?
by Rajeev Thakur 16 May '14

16 May '14

1.4.1 is a very old version of MPICH. Can you try with the latest release, 3.1? Rajeev On May 16, 2014, at 8:50 AM, CANELA-XANDRI Oriol <Oriol.CAnela-Xandri(a)roslin.ed.ac.uk> wrote: > Hello, > > I am writing a function for storing a block cyclic distributed array using MPI_Type_create_darray, MPI_File_set_view, and MPI_File_write_all functions. After some testing I found that for some particular cases, sometimes, some of the values of the array have been not stored properly on the file. The error seems random, but always affects the same number in the same position (where instead of a number, there is a nan, it seems that the position is not written). I tried with other implementations (e.g. openmpi) and seems everything works well (I am using MPICH2 Version 1.4.1). > > For instance, I saw the error for a matrix with 5 rows and 4 columns. Block size for rows = 1 and for columns = 3. There are 9 MPI processes distributed in a 3x3 array. My function is (where nBlockRows/nBlockCols define the size of the blocks, nGlobRows/nGlobCols define the global size of the matrix, nProcRows/nProcCols define the dimensions of the process grid, and fname is the name of the file.): > > void Matrix::writeMatrixMPI(std::string fname) > { > int dims[] = {this->nGlobRows, this->nGlobCols}; > int dargs[] = {this->nBlockRows, this->nBlockCols}; > int distribs[] = {MPI_DISTRIBUTE_CYCLIC, MPI_DISTRIBUTE_CYCLIC}; > int dim[] = {communicator->nProcRows, communicator->nProcCols}; > char nat[] = "native"; > int rc; > MPI_Datatype dcarray; > MPI_File cFile; > MPI_Status status; > > MPI_Type_create_darray(communicator->mpiNumTasks, communicator->mpiRank, 2, dims, distribs, dargs, dim, MPI_ORDER_FORTRAN, MPI_DOUBLE, &dcarray); > MPI_Type_commit(&dcarray); > > std::vector<char> fn(fname.begin(), fname.end()); > fn.push_back('\0'); > rc = MPI_File_open(MPI_COMM_WORLD, &fn[0], MPI_MODE_EXCL | MPI_MODE_CREATE | MPI_MODE_WRONLY, MPI_INFO_NULL, &cFile); > if(rc){ > ss << "Error: Failed to open file: " << rc; > misc.error(ss.str(), 0); > } > else > { > MPI_File_set_view(cFile, 0, MPI_DOUBLE, dcarray, nat, MPI_INFO_NULL); > MPI_File_write_all(cFile, this->m, this->nRows*this->nCols, MPI_DOUBLE, &status); > } > MPI_File_close(&cFile); > MPI_Type_free(&dcarray); > } > > If I call the function a repeated number of times without deleting the file between calls, the file seems to be always correct. If I delete the file between calls or create a file with a different name each time, then I see the error in some of the files. The error is a nan always in the same position. For me it is very strange, but as I said, with other implementations, I do not have the problem. > > -- > The University of Edinburgh is a charitable body, registered in > Scotland, with registration number SC005336. > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss

1 0

bug in MPI_File_write_all?
by CANELA-XANDRI Oriol 16 May '14

16 May '14

Hello, I am writing a function for storing a block cyclic distributed array using MPI_Type_create_darray, MPI_File_set_view, and MPI_File_write_all functions. After some testing I found that for some particular cases, sometimes, some of the values of the array have been not stored properly on the file. The error seems random, but always affects the same number in the same position (where instead of a number, there is a nan, it seems that the position is not written). I tried with other implementations (e.g. openmpi) and seems everything works well (I am using MPICH2 Version 1.4.1). For instance, I saw the error for a matrix with 5 rows and 4 columns. Block size for rows = 1 and for columns = 3. There are 9 MPI processes distributed in a 3x3 array. My function is (where nBlockRows/nBlockCols define the size of the blocks, nGlobRows/nGlobCols define the global size of the matrix, nProcRows/nProcCols define the dimensions of the process grid, and fname is the name of the file.): void Matrix::writeMatrixMPI(std::string fname) { int dims[] = {this->nGlobRows, this->nGlobCols}; int dargs[] = {this->nBlockRows, this->nBlockCols}; int distribs[] = {MPI_DISTRIBUTE_CYCLIC, MPI_DISTRIBUTE_CYCLIC}; int dim[] = {communicator->nProcRows, communicator->nProcCols}; char nat[] = "native"; int rc; MPI_Datatype dcarray; MPI_File cFile; MPI_Status status; MPI_Type_create_darray(communicator->mpiNumTasks, communicator->mpiRank, 2, dims, distribs, dargs, dim, MPI_ORDER_FORTRAN, MPI_DOUBLE, &dcarray); MPI_Type_commit(&dcarray); std::vector<char> fn(fname.begin(), fname.end()); fn.push_back('\0'); rc = MPI_File_open(MPI_COMM_WORLD, &fn[0], MPI_MODE_EXCL | MPI_MODE_CREATE | MPI_MODE_WRONLY, MPI_INFO_NULL, &cFile); if(rc){ ss << "Error: Failed to open file: " << rc; misc.error(ss.str(), 0); } else { MPI_File_set_view(cFile, 0, MPI_DOUBLE, dcarray, nat, MPI_INFO_NULL); MPI_File_write_all(cFile, this->m, this->nRows*this->nCols, MPI_DOUBLE, &status); } MPI_File_close(&cFile); MPI_Type_free(&dcarray); } If I call the function a repeated number of times without deleting the file between calls, the file seems to be always correct. If I delete the file between calls or create a file with a different name each time, then I see the error in some of the files. The error is a nan always in the same position. For me it is very strange, but as I said, with other implementations, I do not have the problem. -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

1 0

Question about launch speed of program compiled with different version of MPICH
by Luo Han 12 May '14

12 May '14

Dear MPI users and developers, Hello! My name is Han Luo and I'm a new hand of MPI. I have some questions to ask you about the speed to launch MPI program. I compiled and ran the "helloworld" program with MPICH.nt. 1.2.5 (download link: http://www.mcs.anl.gov/research/projects/mpi/mpich1-old/mpich-nt/) and MPICH2-1.4.1p-ia32 successfully under Windows 8.1 32bit (2 core cpu, the program was written in Fortran language). But it took a long time, more than 5 seconds for the second one to be launched and stopped. I guess the time might cost in the process of MPI initialization and finalization. I also find that for MPICH2-1.3.2p1-ia32 and MPICH2-1.2.1p-ia32, this problem always exists. Is the problem related to my development environment or configuration? Or it is a common problem? My development environment is VS2012 update4+Intel Cluster Studio 2013 SP1 ( I didn't install Intel MPI) . I do hope to use new MPICH but this problem is really confusing. I have searched for related content by Google but there is nothing help. Thanks for your kind help! I'm looking forward to your reply. Best regards, Luo

1 0

Re: [mpich-discuss] MPE2 and OpenMPI 1.8.1
by Lu, Huiwei 12 May '14

12 May '14

Thank you for the patch, Steven. We will look into it. You can track the progress here: https://trac.mpich.org/projects/mpich/ticket/2090 — Huiwei Lu Postdoc Appointee Mathematics and Computer Science Division, Argonne National Laboratory http://www.mcs.anl.gov/~huiweilu/ On May 6, 2014, at 9:19 AM, Steven Wright <steven.wright(a)warwick.ac.uk> wrote: > Hi, > > The API for OpenMPI has changed at version 1.8.1 (in that many parameters have been made const). This breaks compatibility with MPE when using GCC (though Intel reports these as warnings, rather than errors). > > Attached is a patch to make MPE2 compliant with the current MPI3.0 specification (and therefore build correctly with OpenMPI 1.8.1). > > Apply in mpe2-1.3.0 directory: > ~/mpe2-1.3.0$ patch -p1 < ompi-1.8.1.patch > > Thanks, > Steven > > -- > Steven Wright > CS202 High Performance and Scientific Computing > Performance Computing and Visualisation > Department of Computer Science > University of Warwick > Tel: +44 (0)24 7657 3775 > <ompi-1.8.1.patch>_______________________________________________ > 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] problem when configure mpich
by Junchao Zhang 09 May '14

09 May '14

You may try a clean configure and build with gfortran to see if it works. --Junchao Zhang On Wed, May 7, 2014 at 7:14 PM, Yan Bai <yanbai06(a)gmail.com> wrote: > Many thanks to your quick response. It still doesn’t work. > > I used this configure line > ./configure --prefix=/users/ybai5/mpich2-install F77=$ABSOFT/bin/af90\ > FC=$ABSOFT/bin/af90 --with-pm=mpd > and got this error: > configure: error: The selected Fortran 90 compiler > /Applications/Absoft11.5/bin/af90 does not work with the selected Fortran > 77 compiler /Applications/Absoft11.5/bin/af90. Use the environment > variables FC and F77 respectively to select compatible Fortran compilers. > The check here tests to see if a main program compiled with the Fortran 90 > compiler can link with a subroutine compiled with the Fortran 77 compiler. > > It’s so weird. If you need, I can send the log files. > > Thanks a million! > > Best > > Yan > > > On May 7, 2014, at 7:54 PM, Rajeev Thakur <thakur(a)mcs.anl.gov> wrote: > > Try setting both FC and F77 to af90. > > Rajeev > > On May 7, 2014, at 6:33 PM, Yan Bai <yanbai06(a)gmail.com> wrote: > > Hello > > I was trying to install mpich-3.1 but stopped at the first step: > configuration. The Fortran compiler I have is absoft. I tried both the > configuration command given by Readme file of mpich-3.1 and the > configuration line given by Absoft. Both have the following error message: > > configure: error: The selected Fortran 90 compiler > /Applications/Absoft11.5/bin/af90 does not work with the selected Fortran > 77 compiler /Applications/Absoft11.5/bin/af77. Use the environment > variables FC and F77 respectively to select compatible Fortran compilers. > The check here tests to see if a main program compiled with the Fortran 90 > compiler can link with a subroutine compiled with the Fortran 77 compiler. > > Attached please find the log files. > > I really appreciate your help! > > Best regards > > Yan > ------------------ > Yan Bai > 216 Harkness Hall, RC 270156 > Department of Economics > University of Rochester > 585.275.4196 (office) > 585.256.2309 (fax) > yanbai06(a)gmail.com > https://sites.google.com/site/yanbai06/ > > <sentMPI.zip> > > > _______________________________________________ > 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