[mpich-discuss] beginner for remote mem access

Jeff Hammond jeff.science at gmail.com
Wed Jul 3 14:50:44 CDT 2013


1) When you reply to the digest emails, change the subject to that of
the thread you care to respond to.  That will make it much easier for
others to track the thread.
2) MPI_Win_fence is not similar to MPI_Barrier in some cases.  The
MPICH implementation can turn some calls into no-ops and others into
MPI_Reduce_scatter.
3) Regarding "use MPI window as thread level", I really don't
understand your thinking at all.  MPI RMA is not shared memory nor is
MPI a threading model.

Jeff

On Wed, Jul 3, 2013 at 2:46 PM, Sufeng Niu <sniu at hawk.iit.edu> wrote:
>
> Thanks a lot for your guys reply. I should initial b array, I figure out the
> MPI_Win_fence should be called by all processes, which will solve the
> assertion failed issue.
>
> it seems that MPI_Win_fence is very similar to MPI_Barrier, Can I say
> MPI_Win_fence is the RMA version of MPI_Barrier?
>
> sorry about so many questions. Can I use MPI window as thread level? let's
> say I have 4 processes, each process has 8 threads. if only one thread is
> used to create mem access window, other process copy it and internal 8
> threads use shared memory to perform some operations. Is that ok?
>
> Thanks a lot!
>
>
>
> On Wed, Jul 3, 2013 at 1:28 PM, <discuss-request at mpich.org> wrote:
>>
>> Send discuss mailing list submissions to
>>         discuss at 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 at mpich.org
>>
>> You can reach the person managing the list at
>>         discuss-owner at mpich.org
>>
>> When replying, please edit your Subject line so it is more specific
>> than "Re: Contents of discuss digest..."
>>
>>
>> Today's Topics:
>>
>>    1.  beginner for remote mem access (Sufeng Niu)
>>    2. Re:  beginner for remote mem access (Yi Gu)
>>    3. Re:  beginner for remote mem access (Yi Gu)
>>    4. Re:  beginner for remote mem access (Jeff Hammond)
>>
>>
>> ----------------------------------------------------------------------
>>
>> Message: 1
>> Date: Wed, 3 Jul 2013 13:12:46 -0500
>> From: Sufeng Niu <sniu at hawk.iit.edu>
>> To: discuss at mpich.org
>> Subject: [mpich-discuss] beginner for remote mem access
>> Message-ID:
>>
>> <CAFNNHkwxqXUB_+b8_tZ=L5K7VcDVhsz4ZMDDMGF55v-FAhgeqg at mail.gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Hi,
>>
>> I am a beginner and just try to use remote memory access, and I wrote a
>> simple program to test it:
>>
>> #include "mpi.h"
>> #include <stdio.h>
>> #define SIZE 8
>>
>> int main(int argc, char *argv[])
>> {
>>         int numtasks, rank, source=0, dest, tag=1, i;
>>         float a[64] =
>>         {
>>                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> 12.0, 13.0, 14.0, 15.0, 16.0,
>>                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> 12.0, 13.0, 14.0, 15.0, 16.0,
>>                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> 12.0, 13.0, 14.0, 15.0, 16.0,
>>                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> 12.0, 13.0, 14.0, 15.0, 16.0,
>>         };
>>         float b[SIZE];
>>
>>         MPI_Status stat;
>>
>>         MPI_Init(&argc,&argv);
>>         MPI_Comm_rank(MPI_COMM_WORLD, &rank);
>>         MPI_Comm_size(MPI_COMM_WORLD, &numtasks);
>>
>>         MPI_Win win;
>>
>>         // check processor rank
>>         char processor_name[MPI_MAX_PROCESSOR_NAME];
>>         int name_len;
>>         MPI_Get_processor_name(processor_name, &name_len);
>>         printf("-- processor %s, rank %d out of %d processors\n",
>> processor_name, rank, numtasks);
>>
>>         MPI_Barrier(MPI_COMM_WORLD);
>>
>>         if (numtasks == 4) {
>>                 if (rank == 0) {
>>                         printf("create window \n");
>>                         MPI_Win_create(a, 8*sizeof(float), sizeof(float),
>> MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>>
>>                 }
>>                 else {
>>                         MPI_Win_create(MPI_BOTTOM, 0, sizeof(float),
>> MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>>                 }
>>
>>                 MPI_Win_fence(0, win);
>>
>>                 if (rank == 1){
>>                         MPI_Get(b, SIZE, MPI_FLOAT, 0, 8, SIZE, MPI_FLOAT,
>> win);
>>
>>                         MPI_Win_fence(0, win);
>>                 }
>>
>>                 printf("rank= %d  b= %3.1f %3.1f %3.1f %3.1f %3.1f %3.1f
>> %3.1f %3.1f\n", rank,b[0],b[1],b[2],b[3],b[4],b[5],b[6],b[7]);
>>         }
>>         else
>>                 printf("Must specify %d processors. Terminating.\n",SIZE);
>>
>>         MPI_Win_free(&win);
>>         MPI_Finalize();
>> }
>>
>> However the terminal gives some odd results:
>> rank= 0  b= 0.0 0.0 0.0 0.0 0.0 0.0 -71847793475452928.0 0.0
>> rank= 2  b= 0.0 0.0 0.0 0.0 0.0 0.0 222086852849451401216.0 0.0
>> rank= 3  b= 0.0 0.0 0.0 0.0 0.0 0.0 -74882.4 0.0
>> rank= 1  b= 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0
>> the rank 1 is correct results, but others should be all zero.
>> terminal also give some comments: "Assertion failed in file
>> src/mpid/ch3/src/ch3u_rma_sync.c at line 5061: win_ptr->my_counter >= 0
>> internal ABORT - process 0"
>>
>> another question is if I use remote memory access. all process which does
>> not create window for share must add additional line:
>> MPI_Win_create(MPI_BOTTOM, 0, data_type, MPI_INFO_NULL, MPI_COMM_WORLD,
>> &win); correct?
>>
>> Thanks a lot!
>>
>> --
>> Best Regards,
>> Sufeng Niu
>> ECASP lab, ECE department, Illinois Institute of Technology
>> Tel: 312-731-7219
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:
>> <http://lists.mpich.org/pipermail/discuss/attachments/20130703/91c4ec09/attachment-0001.html>
>>
>> ------------------------------
>>
>> Message: 2
>> Date: Wed, 3 Jul 2013 13:18:26 -0500
>> From: Yi Gu <gyi at mtu.edu>
>> To: discuss at mpich.org
>> Subject: Re: [mpich-discuss] beginner for remote mem access
>> Message-ID:
>>
>> <CAE5iO_SNJKyC9sG3XbTu4vbS=boDrOEV-Y69jUXZR6MQDanUZA at mail.gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Hi, sufeng:
>>
>> I think you may first initialize array first, since you print out b
>> without
>> initialization,
>> it could print out anything.
>>
>> Yi
>>
>>
>> On Wed, Jul 3, 2013 at 1:12 PM, Sufeng Niu <sniu at hawk.iit.edu> wrote:
>>
>> > Hi,
>> >
>> > I am a beginner and just try to use remote memory access, and I wrote a
>> > simple program to test it:
>> >
>> > #include "mpi.h"
>> > #include <stdio.h>
>> > #define SIZE 8
>> >
>> > int main(int argc, char *argv[])
>> > {
>> >         int numtasks, rank, source=0, dest, tag=1, i;
>> >         float a[64] =
>> >         {
>> >                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> > 12.0, 13.0, 14.0, 15.0, 16.0,
>> >                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> > 12.0, 13.0, 14.0, 15.0, 16.0,
>> >                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> > 12.0, 13.0, 14.0, 15.0, 16.0,
>> >                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> > 12.0, 13.0, 14.0, 15.0, 16.0,
>> >         };
>> >         float b[SIZE];
>> >
>> >         MPI_Status stat;
>> >
>> >         MPI_Init(&argc,&argv);
>> >         MPI_Comm_rank(MPI_COMM_WORLD, &rank);
>> >         MPI_Comm_size(MPI_COMM_WORLD, &numtasks);
>> >
>> >         MPI_Win win;
>> >
>> >         // check processor rank
>> >         char processor_name[MPI_MAX_PROCESSOR_NAME];
>> >         int name_len;
>> >         MPI_Get_processor_name(processor_name, &name_len);
>> >         printf("-- processor %s, rank %d out of %d processors\n",
>> > processor_name, rank, numtasks);
>> >
>> >         MPI_Barrier(MPI_COMM_WORLD);
>> >
>> >         if (numtasks == 4) {
>> >                 if (rank == 0) {
>> >                         printf("create window \n");
>> >                         MPI_Win_create(a, 8*sizeof(float),
>> > sizeof(float),
>> > MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>> >
>> >                 }
>> >                 else {
>> >                         MPI_Win_create(MPI_BOTTOM, 0, sizeof(float),
>> > MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>> >                 }
>> >
>> >                 MPI_Win_fence(0, win);
>> >
>> >                 if (rank == 1){
>> >                         MPI_Get(b, SIZE, MPI_FLOAT, 0, 8, SIZE,
>> > MPI_FLOAT,
>> > win);
>> >
>> >                         MPI_Win_fence(0, win);
>> >                 }
>> >
>> >                 printf("rank= %d  b= %3.1f %3.1f %3.1f %3.1f %3.1f %3.1f
>> > %3.1f %3.1f\n", rank,b[0],b[1],b[2],b[3],b[4],b[5],b[6],b[7]);
>> >         }
>> >         else
>> >                 printf("Must specify %d processors.
>> > Terminating.\n",SIZE);
>> >
>> >         MPI_Win_free(&win);
>> >         MPI_Finalize();
>> > }
>> >
>> > However the terminal gives some odd results:
>> > rank= 0  b= 0.0 0.0 0.0 0.0 0.0 0.0 -71847793475452928.0 0.0
>> > rank= 2  b= 0.0 0.0 0.0 0.0 0.0 0.0 222086852849451401216.0 0.0
>> > rank= 3  b= 0.0 0.0 0.0 0.0 0.0 0.0 -74882.4 0.0
>> > rank= 1  b= 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0
>> > the rank 1 is correct results, but others should be all zero.
>> > terminal also give some comments: "Assertion failed in file
>> > src/mpid/ch3/src/ch3u_rma_sync.c at line 5061: win_ptr->my_counter >= 0
>> > internal ABORT - process 0"
>> >
>> > another question is if I use remote memory access. all process which
>> > does
>> > not create window for share must add additional line:
>> > MPI_Win_create(MPI_BOTTOM, 0, data_type, MPI_INFO_NULL, MPI_COMM_WORLD,
>> > &win); correct?
>> >
>> > Thanks a lot!
>> >
>> > --
>> > Best Regards,
>> > Sufeng Niu
>> > ECASP lab, ECE department, Illinois Institute of Technology
>> > Tel: 312-731-7219
>> >
>> > _______________________________________________
>> > discuss mailing list     discuss at mpich.org
>> > To manage subscription options or unsubscribe:
>> > https://lists.mpich.org/mailman/listinfo/discuss
>> >
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:
>> <http://lists.mpich.org/pipermail/discuss/attachments/20130703/4414c5cd/attachment-0001.html>
>>
>> ------------------------------
>>
>> Message: 3
>> Date: Wed, 3 Jul 2013 13:26:29 -0500
>> From: Yi Gu <gyi at mtu.edu>
>> To: discuss at mpich.org
>> Subject: Re: [mpich-discuss] beginner for remote mem access
>> Message-ID:
>>
>> <CAE5iO_Rj7uTaVsEmJ1LBHe+OpTCJjLqpMmvBmjZRd=6WLWC=EQ at mail.gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Hi, sufeng:
>>
>> I am not sure the second problem is, but I think
>> MPI_Win_create(MPI_BOTTOM, 0, sizeof(float), MPI_INFO_NULL,
>> MPI_COMM_WORLD,
>> &win);
>> should be
>> MPI_Win_create(MPI_BOTTOM, 0, 1, MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>>
>> For the third question, if they need to access, yes.
>>
>> Yi
>>
>>
>> On Wed, Jul 3, 2013 at 1:18 PM, Yi Gu <gyi at mtu.edu> wrote:
>>
>> > Hi, sufeng:
>> >
>> > I think you may first initialize array first, since you print out b
>> > without initialization,
>> > it could print out anything.
>> >
>> > Yi
>> >
>> >
>> > On Wed, Jul 3, 2013 at 1:12 PM, Sufeng Niu <sniu at hawk.iit.edu> wrote:
>> >
>> >> Hi,
>> >>
>> >> I am a beginner and just try to use remote memory access, and I wrote a
>> >> simple program to test it:
>> >>
>> >> #include "mpi.h"
>> >> #include <stdio.h>
>> >> #define SIZE 8
>> >>
>> >> int main(int argc, char *argv[])
>> >> {
>> >>         int numtasks, rank, source=0, dest, tag=1, i;
>> >>         float a[64] =
>> >>         {
>> >>                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0,
>> >> 11.0,
>> >> 12.0, 13.0, 14.0, 15.0, 16.0,
>> >>                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0,
>> >> 11.0,
>> >> 12.0, 13.0, 14.0, 15.0, 16.0,
>> >>                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0,
>> >> 11.0,
>> >> 12.0, 13.0, 14.0, 15.0, 16.0,
>> >>                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0,
>> >> 11.0,
>> >> 12.0, 13.0, 14.0, 15.0, 16.0,
>> >>         };
>> >>         float b[SIZE];
>> >>
>> >>         MPI_Status stat;
>> >>
>> >>         MPI_Init(&argc,&argv);
>> >>         MPI_Comm_rank(MPI_COMM_WORLD, &rank);
>> >>         MPI_Comm_size(MPI_COMM_WORLD, &numtasks);
>> >>
>> >>         MPI_Win win;
>> >>
>> >>         // check processor rank
>> >>         char processor_name[MPI_MAX_PROCESSOR_NAME];
>> >>         int name_len;
>> >>         MPI_Get_processor_name(processor_name, &name_len);
>> >>         printf("-- processor %s, rank %d out of %d processors\n",
>> >> processor_name, rank, numtasks);
>> >>
>> >>         MPI_Barrier(MPI_COMM_WORLD);
>> >>
>> >>         if (numtasks == 4) {
>> >>                 if (rank == 0) {
>> >>                         printf("create window \n");
>> >>                         MPI_Win_create(a, 8*sizeof(float),
>> >> sizeof(float),
>> >> MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>> >>
>> >>                 }
>> >>                 else {
>> >>                         MPI_Win_create(MPI_BOTTOM, 0, sizeof(float),
>> >> MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>> >>                 }
>> >>
>> >>                 MPI_Win_fence(0, win);
>> >>
>> >>                 if (rank == 1){
>> >>                         MPI_Get(b, SIZE, MPI_FLOAT, 0, 8, SIZE,
>> >> MPI_FLOAT, win);
>> >>
>> >>                         MPI_Win_fence(0, win);
>> >>                 }
>> >>
>> >>                 printf("rank= %d  b= %3.1f %3.1f %3.1f %3.1f %3.1f
>> >> %3.1f
>> >> %3.1f %3.1f\n", rank,b[0],b[1],b[2],b[3],b[4],b[5],b[6],b[7]);
>> >>         }
>> >>         else
>> >>                 printf("Must specify %d processors.
>> >> Terminating.\n",SIZE);
>> >>
>> >>         MPI_Win_free(&win);
>> >>         MPI_Finalize();
>> >> }
>> >>
>> >> However the terminal gives some odd results:
>> >> rank= 0  b= 0.0 0.0 0.0 0.0 0.0 0.0 -71847793475452928.0 0.0
>> >> rank= 2  b= 0.0 0.0 0.0 0.0 0.0 0.0 222086852849451401216.0 0.0
>> >> rank= 3  b= 0.0 0.0 0.0 0.0 0.0 0.0 -74882.4 0.0
>> >> rank= 1  b= 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0
>> >> the rank 1 is correct results, but others should be all zero.
>> >> terminal also give some comments: "Assertion failed in file
>> >> src/mpid/ch3/src/ch3u_rma_sync.c at line 5061: win_ptr->my_counter >= 0
>> >> internal ABORT - process 0"
>> >>
>> >> another question is if I use remote memory access. all process which
>> >> does
>> >> not create window for share must add additional line:
>> >> MPI_Win_create(MPI_BOTTOM, 0, data_type, MPI_INFO_NULL, MPI_COMM_WORLD,
>> >> &win); correct?
>> >>
>> >> Thanks a lot!
>> >>
>> >> --
>> >> Best Regards,
>> >> Sufeng Niu
>> >> ECASP lab, ECE department, Illinois Institute of Technology
>> >> Tel: 312-731-7219
>> >>
>> >> _______________________________________________
>> >> discuss mailing list     discuss at mpich.org
>> >> To manage subscription options or unsubscribe:
>> >> https://lists.mpich.org/mailman/listinfo/discuss
>> >>
>> >
>> >
>> -------------- next part --------------
>> An HTML attachment was scrubbed...
>> URL:
>> <http://lists.mpich.org/pipermail/discuss/attachments/20130703/362a0cda/attachment-0001.html>
>>
>> ------------------------------
>>
>> Message: 4
>> Date: Wed, 3 Jul 2013 13:28:31 -0500
>> From: Jeff Hammond <jeff.science at gmail.com>
>> To: discuss at mpich.org
>> Subject: Re: [mpich-discuss] beginner for remote mem access
>> Message-ID:
>>
>> <CAGKz=uLwki+cCY2oL_2vFwkwCi=A1Qd9VV1qXHno20gHxqY6iw at mail.gmail.com>
>> Content-Type: text/plain; charset=ISO-8859-1
>>
>> Look at the MPICH test suite for numerous examples of correct RMA
>> programs.
>>
>> Jeff
>>
>> On Wed, Jul 3, 2013 at 1:12 PM, Sufeng Niu <sniu at hawk.iit.edu> wrote:
>> > Hi,
>> >
>> > I am a beginner and just try to use remote memory access, and I wrote a
>> > simple program to test it:
>> >
>> > #include "mpi.h"
>> > #include <stdio.h>
>> > #define SIZE 8
>> >
>> > int main(int argc, char *argv[])
>> > {
>> >         int numtasks, rank, source=0, dest, tag=1, i;
>> >         float a[64] =
>> >         {
>> >                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> > 12.0, 13.0, 14.0, 15.0, 16.0,
>> >                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> > 12.0, 13.0, 14.0, 15.0, 16.0,
>> >                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> > 12.0, 13.0, 14.0, 15.0, 16.0,
>> >                 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0,
>> > 12.0, 13.0, 14.0, 15.0, 16.0,
>> >         };
>> >         float b[SIZE];
>> >
>> >         MPI_Status stat;
>> >
>> >         MPI_Init(&argc,&argv);
>> >         MPI_Comm_rank(MPI_COMM_WORLD, &rank);
>> >         MPI_Comm_size(MPI_COMM_WORLD, &numtasks);
>> >
>> >         MPI_Win win;
>> >
>> >         // check processor rank
>> >         char processor_name[MPI_MAX_PROCESSOR_NAME];
>> >         int name_len;
>> >         MPI_Get_processor_name(processor_name, &name_len);
>> >         printf("-- processor %s, rank %d out of %d processors\n",
>> > processor_name, rank, numtasks);
>> >
>> >         MPI_Barrier(MPI_COMM_WORLD);
>> >
>> >         if (numtasks == 4) {
>> >                 if (rank == 0) {
>> >                         printf("create window \n");
>> >                         MPI_Win_create(a, 8*sizeof(float),
>> > sizeof(float),
>> > MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>> >
>> >                 }
>> >                 else {
>> >                         MPI_Win_create(MPI_BOTTOM, 0, sizeof(float),
>> > MPI_INFO_NULL, MPI_COMM_WORLD, &win);
>> >                 }
>> >
>> >                 MPI_Win_fence(0, win);
>> >
>> >                 if (rank == 1){
>> >                         MPI_Get(b, SIZE, MPI_FLOAT, 0, 8, SIZE,
>> > MPI_FLOAT,
>> > win);
>> >
>> >                         MPI_Win_fence(0, win);
>> >                 }
>> >
>> >                 printf("rank= %d  b= %3.1f %3.1f %3.1f %3.1f %3.1f %3.1f
>> > %3.1f %3.1f\n", rank,b[0],b[1],b[2],b[3],b[4],b[5],b[6],b[7]);
>> >         }
>> >         else
>> >                 printf("Must specify %d processors.
>> > Terminating.\n",SIZE);
>> >
>> >         MPI_Win_free(&win);
>> >         MPI_Finalize();
>> > }
>> >
>> > However the terminal gives some odd results:
>> > rank= 0  b= 0.0 0.0 0.0 0.0 0.0 0.0 -71847793475452928.0 0.0
>> > rank= 2  b= 0.0 0.0 0.0 0.0 0.0 0.0 222086852849451401216.0 0.0
>> > rank= 3  b= 0.0 0.0 0.0 0.0 0.0 0.0 -74882.4 0.0
>> > rank= 1  b= 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0
>> > the rank 1 is correct results, but others should be all zero.
>> > terminal also give some comments: "Assertion failed in file
>> > src/mpid/ch3/src/ch3u_rma_sync.c at line 5061: win_ptr->my_counter >= 0
>> > internal ABORT - process 0"
>> >
>> > another question is if I use remote memory access. all process which
>> > does
>> > not create window for share must add additional line:
>> > MPI_Win_create(MPI_BOTTOM, 0, data_type, MPI_INFO_NULL, MPI_COMM_WORLD,
>> > &win); correct?
>> >
>> > Thanks a lot!
>> >
>> > --
>> > Best Regards,
>> > Sufeng Niu
>> > ECASP lab, ECE department, Illinois Institute of Technology
>> > Tel: 312-731-7219
>> >
>> > _______________________________________________
>> > discuss mailing list     discuss at mpich.org
>> > To manage subscription options or unsubscribe:
>> > https://lists.mpich.org/mailman/listinfo/discuss
>>
>>
>>
>> --
>> Jeff Hammond
>> jeff.science at gmail.com
>>
>>
>> ------------------------------
>>
>> _______________________________________________
>> discuss mailing list
>> discuss at mpich.org
>> https://lists.mpich.org/mailman/listinfo/discuss
>>
>> End of discuss Digest, Vol 9, Issue 5
>> *************************************
>
>
>
>
> --
> Best Regards,
> Sufeng Niu
> ECASP lab, ECE department, Illinois Institute of Technology
> Tel: 312-731-7219
>
> _______________________________________________
> discuss mailing list     discuss at mpich.org
> To manage subscription options or unsubscribe:
> https://lists.mpich.org/mailman/listinfo/discuss



-- 
Jeff Hammond
jeff.science at gmail.com



More information about the discuss mailing list