August 2015 - discuss - mailman3.cels.anl.gov

Re: [mpich-discuss] From file reading to memory sharing
by Wei-keng Liao 12 Aug '15

12 Aug '15

Hi, Matthieu If you have "a series of subdomains (blocks) to load from the file", one way to get a better I/O performance is to concatenate the filetypes into a single one and then use only one MPI_File_set_view and one MPI_File_read_all to read the data. FYI. This request aggregation strategy is also used in PnetCDF. Wei-keng On Aug 12, 2015, at 11:27 AM, Dorier, Matthieu wrote: > Hi, > > I'm trying to refactor an MPI code using MPI one-sided communications. > > The initial version of the code reads its data from a file containing a 3D array of floats. Each process has a series of subdomains (blocks) to load from the file, so they all open the file and then issue a series of MPI_File_set_view and MPI_File_read. The type passed to MPI_File_set_view is constructed using MPI_Type_create_subarray to match the block that needs to be loaded. > > This code performs very poorly even at small scale: the file is 7GB but the blocks are a few hundreds of bytes, and each process has many blocks to load. > > Instead, I would like to have process rank 0 load the entire file, then expose it over RMA. I'm not familiar at all with MPI one-sided operations, since I never used them before, but I guess there should be a simple way to reuse the subarray datatype of my MPI_File_set_view and use it in the context of an MPI_Get. I'm just not sure what the arguments of this MPI_Get would be. My guess is: origin_count would be the number of floats in a single block, origin_datatype would be MPI_FLOAT, target_rank = 0, target_disp = 0, target_count = 1, target_datatype = my subarray datatype. Would that be correct? > > Thanks, > > Matthieu > _______________________________________________ > 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] discuss Digest, Vol 34, Issue 4
by Jeff Hammond 12 Aug '15

12 Aug '15

1) Please change the subject to something appropriate when responding to email digests. I have violated this recommendation here to preserve thread continuity. 2) Do not post code to the list that does not compile. Using GCC with CFLAGS="-std=c99 -Wall -Werror" is recommended. Jeff On Wed, Aug 12, 2015 at 8:06 AM, Kenneth Raffenetti <raffenet(a)mcs.anl.gov> wrote: > There were a number of bugs in your program. I've attached a new version > that ran fine on my machine. Can you confirm? > > Ken > > > On 08/12/2015 01:25 AM, Riddhi Modi wrote: > >> My program: >> >> #include<stdio.h> >> #include"mpi.h" >> >> int main(int argc,char **argv) >> { >> int total_procs,rank,namelen,i; >> int n=10,m=11; >> char processor_name[MPI_MAX_PROCESSOR_NAME]; >> char filename[256]; >> file *fp1,*fp2; >> MPI_Status status; >> MPI_Init(&argc,&argv); >> MPI_Comm_size(MPI_COMM_WORLD,&total_procs); >> MPI_Comm_rank(MPI_COMM_WORLD,&rank); >> MPI_Get_Processor_name(Processor name,&namelen); >> fprintf(stdout,"Process %d of %d is on %s >> \n",rank,total_procs,processor_name); >> fflush(stdout); >> if(rank==0) >> { >> MPI_Send(&n,1,MPI_INT,1,0,MPI_COMM_WORLD); >> fp1=fopen("n.xx","w"); >> fprintf(fp1,"%d",n); >> } >> >> >> if(rank==1) >> { >> MPI_Recv(&n,1,MPI_INT,0.0,MPI_COMM_WORLD,&status); >> n=11; >> fp2=fopen("n.yy","w"); >> fprintf(fp2,"%d",n); >> fclose(fp2); >> } >> MPI_Finalize(); >> return 0; >> } >> >> >> to execute the program i wrote : mpiexec -n 2 -hosts master,node1 ./sim >> >> but it gives me error like assertion failed in file >> src/mpid/ch3/channels/nemesis/netmod/tcp/socksm.c at line 597 : >> hdr.datalen!=0 >> >> internal ABORT - process 1 . >> >> >> how to solve this error ?? >> >> >> >> On Tue, Aug 11, 2015 at 9:54 PM, <discuss-request(a)mpich.org >> <mailto:[email protected]>> wrote: >> >> Send discuss mailing list submissions to >> discuss(a)mpich.org <mailto:[email protected]> >> >> To subscribe or unsubscribe via the World Wide Web, visit >> https://lists.mpich.org/mailman/listinfo/discuss >> or, via email, send a message with subject or body 'help' to >> discuss-request(a)mpich.org <mailto:[email protected]> >> >> You can reach the person managing the list at >> discuss-owner(a)mpich.org <mailto:[email protected]> >> >> When replying, please edit your Subject line so it is more specific >> than "Re: Contents of discuss digest..." >> >> >> Today's Topics: >> >> 1. Re: [MPICH compilation with NAG Fortran compiler] >> (Thakur, Rajeev) >> 2. Re: Problems with running make (Balaji, Pavan) >> 3. static building of mpich (Doha Ehab) >> 4. Re: static building of mpich (Balaji, Pavan) >> 5. Error : Assertion Failed (Riddhi Modi) >> 6. Re: Error : Assertion Failed (Balaji, Pavan) >> 7. Re: static building of mpich (Doha Ehab) >> >> >> ---------------------------------------------------------------------- >> >> Message: 1 >> Date: Thu, 6 Aug 2015 15:15:32 +0000 >> From: "Thakur, Rajeev" <thakur(a)mcs.anl.gov <mailto:[email protected] >> >> >> To: "discuss(a)mpich.org <mailto:[email protected]>" >> <discuss(a)mpich.org <mailto:[email protected]>> >> Subject: Re: [mpich-discuss] [MPICH compilation with NAG Fortran >> compiler] >> Message-ID: <93932738-717A-466B-96ED-F9A398DD92EE(a)mcs.anl.gov >> <mailto:[email protected]>> >> Content-Type: text/plain; charset="utf-8" >> >> We run nightly tests with the NAG compiler, and they build fine (see >> the 5th column from the right at >> http://www.mpich.org/static/cron/tests/) Can you try configuring >> with just the ?prefix option to see if one of the options is the >> cause. Shared libraries are built by default. >> >> Rajeev >> >> >> > On Aug 6, 2015, at 6:25 AM, Maxim Abalenkov >> <maxim.abalenkov(a)gmail.com <mailto:[email protected]>> wrote: >> > >> > Dear Madam or Sir, >> > >> > I?m facing a problem trying to compile MPICH with NAG Fortran >> compiler. The compilation breaks with an unrecognised option >> ??dynamic lib? error. Please see the make log file called ?m.txt?. >> > >> > To configure MPICH version 3.2b4 I?m using the attached configure >> script. Please see the script for the MPICH configuration options. >> Configuration and compilation log files called ?c.txt? and ?m.txt? >> are also attached. >> > >> > I would greatly appreciate your help in compiling MPICH with NAG >> Fortran. Thank you. Have a good day. >> > >> > Product name and version: NAG Fortran Compiler Release 6.0 >> (Hibiya) Build 1021 >> > OS: Mac OS X Yosemite 10.10.4 (64-bit) >> > >> > ? >> > Best wishes, >> > Maxim Abalenkov >> > >> > >> > >> > _______________________________________________ >> > discuss mailing list discuss(a)mpich.org <mailto:[email protected]> >> > To manage subscription options or unsubscribe: >> > https://lists.mpich.org/mailman/listinfo/discuss >> > <mpich-configure-nag.sh><c.txt><m.txt> >> >> >> ------------------------------ >> >> Message: 2 >> Date: Thu, 6 Aug 2015 15:48:45 +0000 >> From: "Balaji, Pavan" <balaji(a)anl.gov <mailto:[email protected]>> >> To: "discuss(a)mpich.org <mailto:[email protected]>" >> <discuss(a)mpich.org <mailto:[email protected]>> >> Subject: Re: [mpich-discuss] Problems with running make >> Message-ID: <D355EC24-A2D3-4BDA-85A0-01F76B89D250(a)anl.gov >> <mailto:[email protected]>> >> Content-Type: text/plain; charset="utf-8" >> >> >> Sorry, windows is not supported. >> >> -- Pavan >> >> From: hirak chatterjee >> Reply-To: "discuss(a)mpich.org >> <mailto:[email protected]><mailto:[email protected] >> <mailto:[email protected]>>" >> Date: Thursday, August 6, 2015 at 10:04 AM >> To: "discuss(a)mpich.org >> <mailto:[email protected]><mailto:[email protected] >> >> <mailto:[email protected]>>" >> Subject: [mpich-discuss] Problems with running make >> >> >> Sir, >> I am working with windows 8 (64 bit). After configuring >> successfully,following the export step >> >> ./configure --prefix=/home/hirak/mpich-install >> >> --enable-timer-type=gettimeofday CC="gcc-4" CPP="cpp-4" CXX="g++-4" >> >> CXXCPP="cpp-4" 2>&1 | tee c.txt >> >> I am getting the error repeatedly >> >> function 'MPIX_Grequest_class_allocate' part of alias cycle. >> The text file is attached herewith. >> Plz help. >> Hirak >> >> -- >> Hirak Chatterjee<https://www.facebook.com/hirakc1> >> Research Scholar >> Department of Chemistry >> Assam University, Silchar >> Assam-788011 >> INDIA >> Phone No: +918486882543 (ASSAM) >> +919475265107 (WEST BENGAL) >> >> _______________________________________________ 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] From file reading to memory sharing
by Jeff Hammond 12 Aug '15

12 Aug '15

On Wed, Aug 12, 2015 at 9:27 AM, Dorier, Matthieu <mdorier(a)anl.gov> wrote: > Hi, > > I'm trying to refactor an MPI code using MPI one-sided communications. > > The initial version of the code reads its data from a file containing a 3D > array of floats. Each process has a series of subdomains (blocks) to load > from the file, so they all open the file and then issue a series of > MPI_File_set_view and MPI_File_read. The type passed to MPI_File_set_view > is constructed using MPI_Type_create_subarray to match the block that needs > to be loaded. > > This code performs very poorly even at small scale: the file is 7GB but > the blocks are a few hundreds of bytes, and each process has many blocks to > load. > > What is the root cause for the poor performance here? > Instead, I would like to have process rank 0 load the entire file, then > expose it over RMA. I'm not > I don't understand why you want to serialize your I/O through rank 0, limit your code to loading files that fit into the memory of one process, and then force every single process to obtain data via a point-to-point operation with one process. This seems triply unscalable. Perhaps you can address your performance issues without compromising scalability. Do you have a MCVE (http://stackoverflow.com/help/mcve) for this? Jeff > familiar at all with MPI one-sided operations, since I never used them > before, but I guess there should be a simple way to reuse the subarray > datatype of my MPI_File_set_view and use it in the context of an MPI_Get. > I'm just not sure what the arguments of this MPI_Get would be. My guess is: > origin_count would be the number of floats in a single block, > origin_datatype would be MPI_FLOAT, target_rank = 0, target_disp = 0, > target_count = 1, target_datatype = my subarray datatype. Would that be > correct? > > Thanks, > > Matthieu > > _______________________________________________ > 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

Re: [mpich-discuss] From file reading to memory sharing
by Balaji, Pavan 12 Aug '15

12 Aug '15

If you are reading into a subarray in memory, you can simply expose the base address of the buffer in the window and then the GET would use the same subarray datatype to fetch the actual data content. Remember that the GET is using the datatype describing the layout in memory and not that on file. -- Pavan From: "Dorier, Matthieu" Reply-To: "discuss(a)mpich.org<mailto:[email protected]>" Date: Wednesday, August 12, 2015 at 11:27 AM To: "discuss(a)mpich.org<mailto:[email protected]>" Subject: [mpich-discuss] From file reading to memory sharing Hi, I'm trying to refactor an MPI code using MPI one-sided communications. The initial version of the code reads its data from a file containing a 3D array of floats. Each process has a series of subdomains (blocks) to load from the file, so they all open the file and then issue a series of MPI_File_set_view and MPI_File_read. The type passed to MPI_File_set_view is constructed using MPI_Type_create_subarray to match the block that needs to be loaded. This code performs very poorly even at small scale: the file is 7GB but the blocks are a few hundreds of bytes, and each process has many blocks to load. Instead, I would like to have process rank 0 load the entire file, then expose it over RMA. I'm not familiar at all with MPI one-sided operations, since I never used them before, but I guess there should be a simple way to reuse the subarray datatype of my MPI_File_set_view and use it in the context of an MPI_Get. I'm just not sure what the arguments of this MPI_Get would be. My guess is: origin_count would be the number of floats in a single block, origin_datatype would be MPI_FLOAT, target_rank = 0, target_disp = 0, target_count = 1, target_datatype = my subarray datatype. Would that be correct? Thanks, Matthieu _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

From file reading to memory sharing
by Dorier, Matthieu 12 Aug '15

12 Aug '15

Hi, I'm trying to refactor an MPI code using MPI one-sided communications. The initial version of the code reads its data from a file containing a 3D array of floats. Each process has a series of subdomains (blocks) to load from the file, so they all open the file and then issue a series of MPI_File_set_view and MPI_File_read. The type passed to MPI_File_set_view is constructed using MPI_Type_create_subarray to match the block that needs to be loaded. This code performs very poorly even at small scale: the file is 7GB but the blocks are a few hundreds of bytes, and each process has many blocks to load. Instead, I would like to have process rank 0 load the entire file, then expose it over RMA. I'm not familiar at all with MPI one-sided operations, since I never used them before, but I guess there should be a simple way to reuse the subarray datatype of my MPI_File_set_view and use it in the context of an MPI_Get. I'm just not sure what the arguments of this MPI_Get would be. My guess is: origin_count would be the number of floats in a single block, origin_datatype would be MPI_FLOAT, target_rank = 0, target_disp = 0, target_count = 1, target_datatype = my subarray datatype. Would that be correct? Thanks, Matthieu _______________________________________________ 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] discuss Digest, Vol 34, Issue 4
by Kenneth Raffenetti 12 Aug '15

12 Aug '15

There were a number of bugs in your program. I've attached a new version that ran fine on my machine. Can you confirm? Ken On 08/12/2015 01:25 AM, Riddhi Modi wrote: > My program: > > #include<stdio.h> > #include"mpi.h" > > int main(int argc,char **argv) > { > int total_procs,rank,namelen,i; > int n=10,m=11; > char processor_name[MPI_MAX_PROCESSOR_NAME]; > char filename[256]; > file *fp1,*fp2; > MPI_Status status; > MPI_Init(&argc,&argv); > MPI_Comm_size(MPI_COMM_WORLD,&total_procs); > MPI_Comm_rank(MPI_COMM_WORLD,&rank); > MPI_Get_Processor_name(Processor name,&namelen); > fprintf(stdout,"Process %d of %d is on %s > \n",rank,total_procs,processor_name); > fflush(stdout); > if(rank==0) > { > MPI_Send(&n,1,MPI_INT,1,0,MPI_COMM_WORLD); > fp1=fopen("n.xx","w"); > fprintf(fp1,"%d",n); > } > > > if(rank==1) > { > MPI_Recv(&n,1,MPI_INT,0.0,MPI_COMM_WORLD,&status); > n=11; > fp2=fopen("n.yy","w"); > fprintf(fp2,"%d",n); > fclose(fp2); > } > MPI_Finalize(); > return 0; > } > > > to execute the program i wrote : mpiexec -n 2 -hosts master,node1 ./sim > > but it gives me error like assertion failed in file > src/mpid/ch3/channels/nemesis/netmod/tcp/socksm.c at line 597 : > hdr.datalen!=0 > > internal ABORT - process 1 . > > > how to solve this error ?? > > > > On Tue, Aug 11, 2015 at 9:54 PM, <discuss-request(a)mpich.org > <mailto:[email protected]>> wrote: > > Send discuss mailing list submissions to > discuss(a)mpich.org <mailto:[email protected]> > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.mpich.org/mailman/listinfo/discuss > or, via email, send a message with subject or body 'help' to > discuss-request(a)mpich.org <mailto:[email protected]> > > You can reach the person managing the list at > discuss-owner(a)mpich.org <mailto:[email protected]> > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of discuss digest..." > > > Today's Topics: > > 1. Re: [MPICH compilation with NAG Fortran compiler] > (Thakur, Rajeev) > 2. Re: Problems with running make (Balaji, Pavan) > 3. static building of mpich (Doha Ehab) > 4. Re: static building of mpich (Balaji, Pavan) > 5. Error : Assertion Failed (Riddhi Modi) > 6. Re: Error : Assertion Failed (Balaji, Pavan) > 7. Re: static building of mpich (Doha Ehab) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Thu, 6 Aug 2015 15:15:32 +0000 > From: "Thakur, Rajeev" <thakur(a)mcs.anl.gov <mailto:[email protected]>> > To: "discuss(a)mpich.org <mailto:[email protected]>" > <discuss(a)mpich.org <mailto:[email protected]>> > Subject: Re: [mpich-discuss] [MPICH compilation with NAG Fortran > compiler] > Message-ID: <93932738-717A-466B-96ED-F9A398DD92EE(a)mcs.anl.gov > <mailto:[email protected]>> > Content-Type: text/plain; charset="utf-8" > > We run nightly tests with the NAG compiler, and they build fine (see > the 5th column from the right at > http://www.mpich.org/static/cron/tests/) Can you try configuring > with just the ?prefix option to see if one of the options is the > cause. Shared libraries are built by default. > > Rajeev > > > > On Aug 6, 2015, at 6:25 AM, Maxim Abalenkov > <maxim.abalenkov(a)gmail.com <mailto:[email protected]>> wrote: > > > > Dear Madam or Sir, > > > > I?m facing a problem trying to compile MPICH with NAG Fortran > compiler. The compilation breaks with an unrecognised option > ??dynamic lib? error. Please see the make log file called ?m.txt?. > > > > To configure MPICH version 3.2b4 I?m using the attached configure > script. Please see the script for the MPICH configuration options. > Configuration and compilation log files called ?c.txt? and ?m.txt? > are also attached. > > > > I would greatly appreciate your help in compiling MPICH with NAG > Fortran. Thank you. Have a good day. > > > > Product name and version: NAG Fortran Compiler Release 6.0 > (Hibiya) Build 1021 > > OS: Mac OS X Yosemite 10.10.4 (64-bit) > > > > ? > > Best wishes, > > Maxim Abalenkov > > > > > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org <mailto:[email protected]> > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > <mpich-configure-nag.sh><c.txt><m.txt> > > > ------------------------------ > > Message: 2 > Date: Thu, 6 Aug 2015 15:48:45 +0000 > From: "Balaji, Pavan" <balaji(a)anl.gov <mailto:[email protected]>> > To: "discuss(a)mpich.org <mailto:[email protected]>" > <discuss(a)mpich.org <mailto:[email protected]>> > Subject: Re: [mpich-discuss] Problems with running make > Message-ID: <D355EC24-A2D3-4BDA-85A0-01F76B89D250(a)anl.gov > <mailto:[email protected]>> > Content-Type: text/plain; charset="utf-8" > > > Sorry, windows is not supported. > > -- Pavan > > From: hirak chatterjee > Reply-To: "discuss(a)mpich.org > <mailto:[email protected]><mailto:[email protected] > <mailto:[email protected]>>" > Date: Thursday, August 6, 2015 at 10:04 AM > To: "discuss(a)mpich.org > <mailto:[email protected]><mailto:[email protected] > <mailto:[email protected]>>" > Subject: [mpich-discuss] Problems with running make > > > Sir, > I am working with windows 8 (64 bit). After configuring > successfully,following the export step > > ./configure --prefix=/home/hirak/mpich-install > > --enable-timer-type=gettimeofday CC="gcc-4" CPP="cpp-4" CXX="g++-4" > > CXXCPP="cpp-4" 2>&1 | tee c.txt > > I am getting the error repeatedly > > function 'MPIX_Grequest_class_allocate' part of alias cycle. > The text file is attached herewith. > Plz help. > Hirak > > -- > Hirak Chatterjee<https://www.facebook.com/hirakc1> > Research Scholar > Department of Chemistry > Assam University, Silchar > Assam-788011 > INDIA > Phone No: +918486882543 (ASSAM) > +919475265107 (WEST BENGAL) > > _______________________________________________ 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] discuss Digest, Vol 34, Issue 4
by Riddhi Modi 12 Aug '15

12 Aug '15

My program: #include<stdio.h> #include"mpi.h" int main(int argc,char **argv) { int total_procs,rank,namelen,i; int n=10,m=11; char processor_name[MPI_MAX_PROCESSOR_NAME]; char filename[256]; file *fp1,*fp2; MPI_Status status; MPI_Init(&argc,&argv); MPI_Comm_size(MPI_COMM_WORLD,&total_procs); MPI_Comm_rank(MPI_COMM_WORLD,&rank); MPI_Get_Processor_name(Processor name,&namelen); fprintf(stdout,"Process %d of %d is on %s \n",rank,total_procs,processor_name); fflush(stdout); if(rank==0) { MPI_Send(&n,1,MPI_INT,1,0,MPI_COMM_WORLD); fp1=fopen("n.xx","w"); fprintf(fp1,"%d",n); } if(rank==1) { MPI_Recv(&n,1,MPI_INT,0.0,MPI_COMM_WORLD,&status); n=11; fp2=fopen("n.yy","w"); fprintf(fp2,"%d",n); fclose(fp2); } MPI_Finalize(); return 0; } to execute the program i wrote : mpiexec -n 2 -hosts master,node1 ./sim but it gives me error like assertion failed in file src/mpid/ch3/channels/nemesis/netmod/tcp/socksm.c at line 597 : hdr.datalen!=0 internal ABORT - process 1 . how to solve this error ?? On Tue, Aug 11, 2015 at 9:54 PM, <discuss-request(a)mpich.org> wrote: > Send discuss mailing list submissions to > discuss(a)mpich.org > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.mpich.org/mailman/listinfo/discuss > or, via email, send a message with subject or body 'help' to > discuss-request(a)mpich.org > > You can reach the person managing the list at > discuss-owner(a)mpich.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of discuss digest..." > > > Today's Topics: > > 1. Re: [MPICH compilation with NAG Fortran compiler] > (Thakur, Rajeev) > 2. Re: Problems with running make (Balaji, Pavan) > 3. static building of mpich (Doha Ehab) > 4. Re: static building of mpich (Balaji, Pavan) > 5. Error : Assertion Failed (Riddhi Modi) > 6. Re: Error : Assertion Failed (Balaji, Pavan) > 7. Re: static building of mpich (Doha Ehab) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Thu, 6 Aug 2015 15:15:32 +0000 > From: "Thakur, Rajeev" <thakur(a)mcs.anl.gov> > To: "discuss(a)mpich.org" <discuss(a)mpich.org> > Subject: Re: [mpich-discuss] [MPICH compilation with NAG Fortran > compiler] > Message-ID: <93932738-717A-466B-96ED-F9A398DD92EE(a)mcs.anl.gov> > Content-Type: text/plain; charset="utf-8" > > We run nightly tests with the NAG compiler, and they build fine (see the > 5th column from the right at http://www.mpich.org/static/cron/tests/) > Can you try configuring with just the ?prefix option to see if one of the > options is the cause. Shared libraries are built by default. > > Rajeev > > > > On Aug 6, 2015, at 6:25 AM, Maxim Abalenkov <maxim.abalenkov(a)gmail.com> > wrote: > > > > Dear Madam or Sir, > > > > I?m facing a problem trying to compile MPICH with NAG Fortran compiler. > The compilation breaks with an unrecognised option ??dynamic lib? error. > Please see the make log file called ?m.txt?. > > > > To configure MPICH version 3.2b4 I?m using the attached configure > script. Please see the script for the MPICH configuration options. > Configuration and compilation log files called ?c.txt? and ?m.txt? are also > attached. > > > > I would greatly appreciate your help in compiling MPICH with NAG > Fortran. Thank you. Have a good day. > > > > Product name and version: NAG Fortran Compiler Release 6.0 (Hibiya) > Build 1021 > > OS: Mac OS X Yosemite 10.10.4 (64-bit) > > > > ? > > Best wishes, > > Maxim Abalenkov > > > > > > > > _______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > <mpich-configure-nag.sh><c.txt><m.txt> > > > ------------------------------ > > Message: 2 > Date: Thu, 6 Aug 2015 15:48:45 +0000 > From: "Balaji, Pavan" <balaji(a)anl.gov> > To: "discuss(a)mpich.org" <discuss(a)mpich.org> > Subject: Re: [mpich-discuss] Problems with running make > Message-ID: <D355EC24-A2D3-4BDA-85A0-01F76B89D250(a)anl.gov> > Content-Type: text/plain; charset="utf-8" > > > Sorry, windows is not supported. > > -- Pavan > > From: hirak chatterjee > Reply-To: "discuss(a)mpich.org<mailto:[email protected]>" > Date: Thursday, August 6, 2015 at 10:04 AM > To: "discuss(a)mpich.org<mailto:[email protected]>" > Subject: [mpich-discuss] Problems with running make > > > Sir, > I am working with windows 8 (64 bit). After configuring > successfully,following the export step > > ./configure --prefix=/home/hirak/mpich-install > > --enable-timer-type=gettimeofday CC="gcc-4" CPP="cpp-4" CXX="g++-4" > > CXXCPP="cpp-4" 2>&1 | tee c.txt > > I am getting the error repeatedly > > function 'MPIX_Grequest_class_allocate' part of alias cycle. > The text file is attached herewith. > Plz help. > Hirak > > -- > Hirak Chatterjee<https://www.facebook.com/hirakc1> > Research Scholar > Department of Chemistry > Assam University, Silchar > Assam-788011 > INDIA > Phone No: +918486882543 (ASSAM) > +919475265107 (WEST BENGAL) > > _______________________________________________ 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] static building of mpich
by Balaji, Pavan 11 Aug '15

11 Aug '15

The static flag should go into LDFLAGS or LIBS, not CFLAGS. Also, for libtool to interpret it correctly, it'll need to be "--static", not "-static" (double dash instead of single dash). -- Pavan From: Doha Ehab Reply-To: "discuss(a)mpich.org<mailto:[email protected]>" Date: Tuesday, August 11, 2015 at 11:24 AM To: "discuss(a)mpich.org<mailto:[email protected]>" Subject: Re: [mpich-discuss] static building of mpich No , but after I did make and make install I checked whether it was statically built using the following command : file bin/mpiexec mpiexec: symbolic link to `mpiexec.hydra' Then I typed file bin/mpiexec.hydrampiexec.hydra: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.24, BuildID[sha1]=0xb4510fb70ac8047d9028dbafe2d43848140fe894, not stripped I don't understand why it is dynamically linked and not statically linked On Mon, Aug 10, 2015 at 6:13 AM, Balaji, Pavan <balaji(a)anl.gov<mailto:[email protected]>> wrote: Did you do make and make install? -- Pavan From: Doha Ehab Reply-To: "discuss(a)mpich.org<mailto:[email protected]>" Date: Sunday, August 9, 2015 at 7:24 PM To: "discuss(a)mpich.org<mailto:[email protected]>" Subject: [mpich-discuss] static building of mpich Hello, i'm trying to cross compile mpich , I have extracted mpich-3.1.3 the I created a folder name build inside it . In order to statically build mpich I used the following commands: export CC="/home/doha/buildroot-2015.02/output/host/usr/bin/arm-linux-gcc" export AR="/home/doha/buildroot-2015.02/output/host/usr/bin/arm-linux-ar" export CFLAGS="-march=armv5te -static -I/home/doha/buildroot-2015.02/output/host/usr/include/" export LDFLAGS="-L/home/doha/buildroot-2015.02/output/host/usr/lib/" cc=/home/doha/buildroot-2015.02/output/host/usr/bin/arm-linux-gcc ./configure --prefix=/home/doha/mpich-3.1.3/build/ --host=arm-linux --disable-shared --with-pm=hydra --with-device=ch3:nemesis --disable-fortran but nothing appears in the build folder while it shows no errors on the terminal _______________________________________________ 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] static building of mpich
by Doha Ehab 11 Aug '15

11 Aug '15

No , but after I did make and make install I checked whether it was statically built using the following command : file bin/mpiexec mpiexec: symbolic link to `mpiexec.hydra' Then I typed file bin/mpiexec.hydrampiexec.hydra: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.24, BuildID[sha1]=0xb4510fb70ac8047d9028dbafe2d43848140fe894, not stripped I don't understand why it is dynamically linked and not statically linked On Mon, Aug 10, 2015 at 6:13 AM, Balaji, Pavan <balaji(a)anl.gov> wrote: > > Did you do make and make install? > > -- Pavan > > From: Doha Ehab > Reply-To: "discuss(a)mpich.org" > Date: Sunday, August 9, 2015 at 7:24 PM > To: "discuss(a)mpich.org" > Subject: [mpich-discuss] static building of mpich > > Hello, > i'm trying to cross compile mpich , I have extracted mpich-3.1.3 the I > created a folder name build inside it . In order to statically build mpich > I used the following commands: > > export CC="/home/doha/buildroot-2015.02/output/host/usr/bin/arm-linux-gcc" > export AR="/home/doha/buildroot-2015.02/output/host/usr/bin/arm-linux-ar" > export CFLAGS="-march=armv5te -static > -I/home/doha/buildroot-2015.02/output/host/usr/include/" > export LDFLAGS="-L/home/doha/buildroot-2015.02/output/host/usr/lib/" > > > cc=/home/doha/buildroot-2015.02/output/host/usr/bin/arm-linux-gcc > ./configure --prefix=/home/doha/mpich-3.1.3/build/ --host=arm-linux > --disable-shared --with-pm=hydra --with-device=ch3:nemesis > --disable-fortran > > > but nothing appears in the build folder while it shows no errors on the > terminal > > _______________________________________________ > 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] Error : Assertion Failed
by Balaji, Pavan 11 Aug '15

11 Aug '15

I don't think we have seen this error before. Can you send us a small program that reproduces this error? -- Pavan From: Riddhi Modi Reply-To: "discuss(a)mpich.org<mailto:[email protected]>" Date: Tuesday, August 11, 2015 at 1:49 AM To: "discuss(a)mpich.org<mailto:[email protected]>" Subject: [mpich-discuss] Error : Assertion Failed I have one major problem.I established mpich-3.1.4 cluster on 2 machines and run cpi.c program with 2 process which run on 2 different machines.But there is an error like Assertion Failed in file src/mpid/ch3/channels/nemesis/netmod/tcp/socksm.c at line 597: hdr.datalen!=0 Internal ABORT -process 1 how can i solve this kind of error.??/ any idea?? _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0