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

Re: [mpich-discuss] Assertion in MX netmod
by "Antonio J. Peña" 24 Nov '14

24 Nov '14

Dear Kuleshov, In order to accomodate resources for more recent networking APIs we dropped support for the mx netmod, which in fact has been completely removed in our most recent 3.2 releases. So, unfortunately, we are not able to assist you with this issue. Best regards, Antonio On 11/22/2014 01:52 PM, Kuleshov Aleksey wrote: > And the same problem with different approach: > I downloaded from http://www.mcs.anl.gov/research/projects/mpi/mpi-test/tsuite.html mpi2test.tar.gz, built it and try > to run pingping test: > >> MPITEST_VERBOSE=1 MPICH_NEMESIS_NETMOD=mx mpiexec -f /tmp/m -n 2 /tests/pingping > [stdout] > Get new datatypes: send = MPI_INT, recv = MPI_INT > Get new datatypes: send = MPI_INT, recv = MPI_INT > Sending count = 1 of sendtype MPI_INT of total size 4 bytes > Sending count = 1 of sendtype MPI_INT of total size 4 bytes > Get new datatypes: send = MPI_DOUBLE, recv = MPI_DOUBLE > Get new datatypes: send = MPI_DOUBLE, recv = MPI_DOUBLE > Sending count = 1 of sendtype MPI_DOUBLE of total size 8 bytes > Sending count = 1 of sendtype MPI_DOUBLE of total size 8 bytes > Get new datatypes: send = MPI_FLOAT_INT, recv = MPI_FLOAT_INT > Sending count = 1 of sendtype MPI_FLOAT_INT of total size 8 bytes > Get new datatypes: send = MPI_FLOAT_INT, recv = MPI_FLOAT_INT > Sending count = 1 of sendtype MPI_FLOAT_INT of total size 8 bytes > Get new datatypes: send = dup of MPI_INT, recv = dup of MPI_INT > Get new datatypes: send = dup of MPI_INT, recv = dup of MPI_INT > Sending count = 1 of sendtype dup of MPI_INT of total size 4 bytes > Sending count = 1 of sendtype dup of MPI_INT of total size 4 bytes > Get new datatypes: send = int-vector, recv = MPI_INT > Sending count = 1 of sendtype int-vector of total size 4 bytes > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_send.c at line 435: n_iov > 0 > internal ABORT - process 0 > [/stdout] > > 22.11.2014, 18:39, "Kuleshov Aleksey" <rndfax(a)yandex.ru>: >> Hello! Can you please help me with problem? >> >> I'm working on custom myriexpress library and I'm using MX netmod in MPICH v.3.1.2. >> For testing purposes I built OSU Micro Benchmarks v3.8. >> >> To run it on 7 nodes I execute test osu_alltoall as follows: >>> MPICH_NEMESIS_NETMOD=mx mpiexec -f /tmp/m -n 7 /osu_alltoall >> It passed successfully (I also tried it on 2, 3, 4, 5 and 6 nodes - everything is alright). >> >> But now I want to run it on 8 nodes: >>> MPICH_NEMESIS_NETMOD=mx mpiexec -f /tmp/m -n 8 /osu_alltoall >> [stdout] >> # OSU MPI All-to-All Personalized Exchange Latency Test v3.8 >> # Size Avg Latency(us) >> Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 >> Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 >> Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 >> Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 >> Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 >> Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 >> Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 >> Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 >> internal ABORT - process 4 >> internal ABORT - process 7 >> internal ABORT - process 2 >> internal ABORT - process 6 >> internal ABORT - process 3 >> internal ABORT - process 0 >> internal ABORT - process 5 >> internal ABORT - process 1 >> [/stdout] >> >> So, what does these assertions mean? >> Is it something wrong with MX netmod? >> Or in myriexpress library? >> Or in test osu_alltoall itself? >> >> BTW, osu_alltoall on 8 nodes passed successfully for TCP netmod. >> _______________________________________________ >> 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 -- Antonio J. Peña Postdoctoral Appointee Mathematics and Computer Science Division Argonne National Laboratory 9700 South Cass Avenue, Bldg. 240, Of. 3148 Argonne, IL 60439-4847 apenya(a)mcs.anl.gov www.mcs.anl.gov/~apenya _______________________________________________ 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] Assertion in MX netmod
by Kuleshov Aleksey 22 Nov '14

22 Nov '14

And the same problem with different approach: I downloaded from http://www.mcs.anl.gov/research/projects/mpi/mpi-test/tsuite.html mpi2test.tar.gz, built it and try to run pingping test: > MPITEST_VERBOSE=1 MPICH_NEMESIS_NETMOD=mx mpiexec -f /tmp/m -n 2 /tests/pingping [stdout] Get new datatypes: send = MPI_INT, recv = MPI_INT Get new datatypes: send = MPI_INT, recv = MPI_INT Sending count = 1 of sendtype MPI_INT of total size 4 bytes Sending count = 1 of sendtype MPI_INT of total size 4 bytes Get new datatypes: send = MPI_DOUBLE, recv = MPI_DOUBLE Get new datatypes: send = MPI_DOUBLE, recv = MPI_DOUBLE Sending count = 1 of sendtype MPI_DOUBLE of total size 8 bytes Sending count = 1 of sendtype MPI_DOUBLE of total size 8 bytes Get new datatypes: send = MPI_FLOAT_INT, recv = MPI_FLOAT_INT Sending count = 1 of sendtype MPI_FLOAT_INT of total size 8 bytes Get new datatypes: send = MPI_FLOAT_INT, recv = MPI_FLOAT_INT Sending count = 1 of sendtype MPI_FLOAT_INT of total size 8 bytes Get new datatypes: send = dup of MPI_INT, recv = dup of MPI_INT Get new datatypes: send = dup of MPI_INT, recv = dup of MPI_INT Sending count = 1 of sendtype dup of MPI_INT of total size 4 bytes Sending count = 1 of sendtype dup of MPI_INT of total size 4 bytes Get new datatypes: send = int-vector, recv = MPI_INT Sending count = 1 of sendtype int-vector of total size 4 bytes Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_send.c at line 435: n_iov > 0 internal ABORT - process 0 [/stdout] 22.11.2014, 18:39, "Kuleshov Aleksey" <rndfax(a)yandex.ru>: > Hello! Can you please help me with problem? > > I'm working on custom myriexpress library and I'm using MX netmod in MPICH v.3.1.2. > For testing purposes I built OSU Micro Benchmarks v3.8. > > To run it on 7 nodes I execute test osu_alltoall as follows: >> MPICH_NEMESIS_NETMOD=mx mpiexec -f /tmp/m -n 7 /osu_alltoall > > It passed successfully (I also tried it on 2, 3, 4, 5 and 6 nodes - everything is alright). > > But now I want to run it on 8 nodes: >> MPICH_NEMESIS_NETMOD=mx mpiexec -f /tmp/m -n 8 /osu_alltoall > > [stdout] > # OSU MPI All-to-All Personalized Exchange Latency Test v3.8 > # Size Avg Latency(us) > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 > Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 > internal ABORT - process 4 > internal ABORT - process 7 > internal ABORT - process 2 > internal ABORT - process 6 > internal ABORT - process 3 > internal ABORT - process 0 > internal ABORT - process 5 > internal ABORT - process 1 > [/stdout] > > So, what does these assertions mean? > Is it something wrong with MX netmod? > Or in myriexpress library? > Or in test osu_alltoall itself? > > BTW, osu_alltoall on 8 nodes passed successfully for TCP netmod. > _______________________________________________ > 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

Assertion in MX netmod
by Kuleshov Aleksey 22 Nov '14

22 Nov '14

Hello! Can you please help me with problem? I'm working on custom myriexpress library and I'm using MX netmod in MPICH v.3.1.2. For testing purposes I built OSU Micro Benchmarks v3.8. To run it on 7 nodes I execute test osu_alltoall as follows: > MPICH_NEMESIS_NETMOD=mx mpiexec -f /tmp/m -n 7 /osu_alltoall It passed successfully (I also tried it on 2, 3, 4, 5 and 6 nodes - everything is alright). But now I want to run it on 8 nodes: > MPICH_NEMESIS_NETMOD=mx mpiexec -f /tmp/m -n 8 /osu_alltoall [stdout] # OSU MPI All-to-All Personalized Exchange Latency Test v3.8 # Size Avg Latency(us) Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 Assertion failed in file ../src/mpid/ch3/channels/nemesis/netmod/mx/mx_poll.c at line 784: n_iov > 0 internal ABORT - process 4 internal ABORT - process 7 internal ABORT - process 2 internal ABORT - process 6 internal ABORT - process 3 internal ABORT - process 0 internal ABORT - process 5 internal ABORT - process 1 [/stdout] So, what does these assertions mean? Is it something wrong with MX netmod? Or in myriexpress library? Or in test osu_alltoall itself? BTW, osu_alltoall on 8 nodes passed successfully for TCP netmod. _______________________________________________ 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] MPI_Iprobe bug in MPICH for BGQ?
by Antonio J. Peña 20 Nov '14

20 Nov '14

Hi Florin, I dont think youre hitting any bug, and Id say both behaviors are correct. Note that when you do a sleep the MPI implementation is not making any progress, so if the Isend call didnt push the message immediately to the network, which is likely to be more the case on an HPC network than on a socket-based one, the sleep is unproductive, and so the Iprobe returns nothing arrived. Best, Toni -------------------------------------------------------------- Antonio J. Peña Postdoctoral Appointee Mathematics and Computer Science Division (MCS) Argonne National Laboratory 9700 S. Cass Avenue Argonne, IL 60439 Building 340 - Office 3148 apenya(a)mcs.anl.gov (+1) 630-252-7928 From: Isaila, Florin D. [mailto:[email protected]] Sent: Thursday, November 20, 2014 4:20 PM To: mpich-discuss(a)mcs.anl.gov Subject: [mpich-discuss] MPI_Iprobe bug in MPICH for BGQ? Hi, when I run the program from below on 1 node on BGQ (Vesta), the message is not received (flag is 0). However on a Ubuntu, the message is received (flag is non-zero). If I add another Iprobe (uncomment the Iprobe in the code below) the message is received on both BGQ and Ubuntu. Note that the program sleeps for 1 second after the Isend. Is it a bug? This happens for both MPICH-3.1.3 and MPICH-3.1. #include "mpi.h" #include <stdio.h> #include <unistd.h> int main(int argc, char **argv) { int send_int, recv_int, tag, flag; MPI_Status status; MPI_Request req; MPI_Init(&argc, &argv); tag = 0; send_int = 100; MPI_Isend(&send_int, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &req ); sleep(1); MPI_Iprobe(MPI_ANY_SOURCE , MPI_ANY_TAG, MPI_COMM_WORLD, &flag, &status ); //MPI_Iprobe(MPI_ANY_SOURCE , tag, MPI_COMM_WORLD, &flag, &status ); if (flag) { MPI_Recv( &recv_int, 1, MPI_INT, MPI_ANY_SOURCE, tag, MPI_COMM_WORLD, &status); printf("Received = %d\n", recv_int); } else printf("Message not received yet"); MPI_Waitall(1, &req, MPI_STATUSES_IGNORE); MPI_Finalize(); return 0; } Thanks Florin _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

MPI_Iprobe bug in MPICH for BGQ?
by Isaila, Florin D. 20 Nov '14

20 Nov '14

Hi, when I run the program from below on 1 node on BGQ (Vesta), the message is not received (flag is 0). However on a Ubuntu, the message is received (flag is non-zero). If I add another Iprobe (uncomment the Iprobe in the code below) the message is received on both BGQ and Ubuntu. Note that the program sleeps for 1 second after the Isend. Is it a bug? This happens for both MPICH-3.1.3 and MPICH-3.1. #include "mpi.h" #include <stdio.h> #include <unistd.h> int main(int argc, char **argv) { int send_int, recv_int, tag, flag; MPI_Status status; MPI_Request req; MPI_Init(&argc, &argv); tag = 0; send_int = 100; MPI_Isend(&send_int, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &req ); sleep(1); MPI_Iprobe(MPI_ANY_SOURCE , MPI_ANY_TAG, MPI_COMM_WORLD, &flag, &status ); //MPI_Iprobe(MPI_ANY_SOURCE , tag, MPI_COMM_WORLD, &flag, &status ); if (flag) { MPI_Recv( &recv_int, 1, MPI_INT, MPI_ANY_SOURCE, tag, MPI_COMM_WORLD, &status); printf("Received = %d\n", recv_int); } else printf("Message not received yet"); MPI_Waitall(1, &req, MPI_STATUSES_IGNORE); MPI_Finalize(); return 0; } Thanks Florin

1 0

Re: [mpich-discuss] f77 bindings and profiling
by Isaila, Florin D. 20 Nov '14

20 Nov '14

ldd output: fisaila@howard:f77_program$ gfortran init_finalize.f -I/homes/fisaila/software/mpich/include -I/homes/fisaila/software/mpich/include -L/homes/fisaila/software/mpich/lib -L.. -ltarget -lmpifort -Wl,-rpath -Wl,/homes/fisaila/software/mpich/lib -Wl,--enable-new-dtags -lmpi fisaila@howard:f77_program$ ldd a.out linux-vdso.so.1 => (0x00007fff5c1ff000) libmpifort.so.12 => /homes/fisaila/software/mpich/lib/libmpifort.so.12 (0x00007fd2b4957000) libgfortran.so.3 => /usr/lib/x86_64-linux-gnu/libgfortran.so.3 (0x00007fd2b460f000) libc.so.6 => /lib/x86_64-linux-gnu/libc.so.6 (0x00007fd2b424f000) libmpi.so.12 => /homes/fisaila/software/mpich/lib/libmpi.so.12 (0x00007fd2b394c000) libquadmath.so.0 => /usr/lib/x86_64-linux-gnu/libquadmath.so.0 (0x00007fd2b3716000) libm.so.6 => /lib/x86_64-linux-gnu/libm.so.6 (0x00007fd2b3419000) /lib64/ld-linux-x86-64.so.2 (0x00007fd2b4b91000) librt.so.1 => /lib/x86_64-linux-gnu/librt.so.1 (0x00007fd2b3211000) libpthread.so.0 => /lib/x86_64-linux-gnu/libpthread.so.0 (0x00007fd2b2ff4000) libgcc_s.so.1 => /lib/x86_64-linux-gnu/libgcc_s.so.1 (0x00007fd2b2ddd000) ________________________________________ From: Kenneth Raffenetti [raffenet(a)mcs.anl.gov] Sent: Thursday, November 20, 2014 9:37 AM To: discuss(a)mpich.org Subject: Re: [mpich-discuss] f77 bindings and profiling Can you paste the output of ldd on your binary? Ken On 11/20/2014 09:22 AM, Isaila, Florin D. wrote: > Hi Ken, > > it is not working this way. > > Florin > ________________________________________ > From: Kenneth Raffenetti [raffenet(a)mcs.anl.gov] > Sent: Wednesday, November 19, 2014 10:56 AM > To: discuss(a)mpich.org > Cc: mpich-discuss(a)mcs.anl.gov > Subject: Re: [mpich-discuss] f77 bindings and profiling > > On 11/18/2014 09:47 AM, Isaila, Florin D. wrote: >> fisaila@howard:f77$ mpif77 -show fpi.f >> gfortran fpi.f -I/homes/fisaila/software/mpich/include -I/homes/fisaila/software/mpich/include -L/homes/fisaila/software/mpich/lib -lmpifort -lfoo -Wl,-rpath -Wl,/homes/fisaila/software/mpich/lib -Wl,--enable-new-dtags -lmpi > > This looks to be the problem. PROFILE_PRELIB needs to be added *before* > libmpifort. Some time after 3.1, we re-organized which symbols go into > which libraries and now libmpifort.so contains PMPI_Init and all its > weak aliases (MPI_Init, mpi_init_, etc.) This means the MPI_Init symbol > is resolved before your library is searched. > > Can you manually try a compile line with -lfoo before -lmpifort and > confirm that it works as expected? I.e. > > gfortran fpi.f -I/homes/fisaila/software/mpich/include > -I/homes/fisaila/software/mpich/include > -L/homes/fisaila/software/mpich/lib -lfoo -lmpifort -Wl,-rpath > -Wl,/homes/fisaila/software/mpich/lib -Wl,--enable-new-dtags -lmpi > > Ken > _______________________________________________ > 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 _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Re : Client hangs if server dies in dynamic process management
by Roy, Hirak 20 Nov '14

20 Nov '14

Hi Huiwei, Thanks for acknowledgement. http://trac.mpich.org/projects/mpich/ticket/2205 Could you please let me know what would be the target release for this? Thanks, Hirak _______________________________________________ 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] f77 bindings and profiling
by Kenneth Raffenetti 20 Nov '14

20 Nov '14

Can you paste the output of ldd on your binary? Ken On 11/20/2014 09:22 AM, Isaila, Florin D. wrote: > Hi Ken, > > it is not working this way. > > Florin > ________________________________________ > From: Kenneth Raffenetti [raffenet(a)mcs.anl.gov] > Sent: Wednesday, November 19, 2014 10:56 AM > To: discuss(a)mpich.org > Cc: mpich-discuss(a)mcs.anl.gov > Subject: Re: [mpich-discuss] f77 bindings and profiling > > On 11/18/2014 09:47 AM, Isaila, Florin D. wrote: >> fisaila@howard:f77$ mpif77 -show fpi.f >> gfortran fpi.f -I/homes/fisaila/software/mpich/include -I/homes/fisaila/software/mpich/include -L/homes/fisaila/software/mpich/lib -lmpifort -lfoo -Wl,-rpath -Wl,/homes/fisaila/software/mpich/lib -Wl,--enable-new-dtags -lmpi > > This looks to be the problem. PROFILE_PRELIB needs to be added *before* > libmpifort. Some time after 3.1, we re-organized which symbols go into > which libraries and now libmpifort.so contains PMPI_Init and all its > weak aliases (MPI_Init, mpi_init_, etc.) This means the MPI_Init symbol > is resolved before your library is searched. > > Can you manually try a compile line with -lfoo before -lmpifort and > confirm that it works as expected? I.e. > > gfortran fpi.f -I/homes/fisaila/software/mpich/include > -I/homes/fisaila/software/mpich/include > -L/homes/fisaila/software/mpich/lib -lfoo -lmpifort -Wl,-rpath > -Wl,/homes/fisaila/software/mpich/lib -Wl,--enable-new-dtags -lmpi > > Ken > _______________________________________________ > 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] f77 bindings and profiling
by Isaila, Florin D. 20 Nov '14

20 Nov '14

Hi Ken, it is not working this way. Florin ________________________________________ From: Kenneth Raffenetti [raffenet(a)mcs.anl.gov] Sent: Wednesday, November 19, 2014 10:56 AM To: discuss(a)mpich.org Cc: mpich-discuss(a)mcs.anl.gov Subject: Re: [mpich-discuss] f77 bindings and profiling On 11/18/2014 09:47 AM, Isaila, Florin D. wrote: > fisaila@howard:f77$ mpif77 -show fpi.f > gfortran fpi.f -I/homes/fisaila/software/mpich/include -I/homes/fisaila/software/mpich/include -L/homes/fisaila/software/mpich/lib -lmpifort -lfoo -Wl,-rpath -Wl,/homes/fisaila/software/mpich/lib -Wl,--enable-new-dtags -lmpi This looks to be the problem. PROFILE_PRELIB needs to be added *before* libmpifort. Some time after 3.1, we re-organized which symbols go into which libraries and now libmpifort.so contains PMPI_Init and all its weak aliases (MPI_Init, mpi_init_, etc.) This means the MPI_Init symbol is resolved before your library is searched. Can you manually try a compile line with -lfoo before -lmpifort and confirm that it works as expected? I.e. gfortran fpi.f -I/homes/fisaila/software/mpich/include -I/homes/fisaila/software/mpich/include -L/homes/fisaila/software/mpich/lib -lfoo -lmpifort -Wl,-rpath -Wl,/homes/fisaila/software/mpich/lib -Wl,--enable-new-dtags -lmpi Ken _______________________________________________ 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] Client hangs if server dies in dynamic process management
by Lu, Huiwei 19 Nov '14

19 Nov '14

Hi, Hirak, Yes I can repeat the bug on both MacOS and Ubuntu with socket channel. I have created a ticket for it. You can track the progress here: http://trac.mpich.org/projects/mpich/ticket/2205 Thanks for reporting the bug. — Huiwei > On Nov 18, 2014, at 12:49 PM, Roy, Hirak <Hirak_Roy(a)mentor.com> wrote: > > > Hi Huiwei, > > 1> Did you start your nameserver ? > 2> Did the server program crash? > I see the same hang (incomplete MPI_Finalize in client). > > Here is my command line: > Ø hydra_namserver & > Ø mpiexec –n 1 –nameserver <hostname> ./server > Ø mpiexec –n 1 –nameserver <hostname> ./client > > > > MPICH Version: 3.2a2 > MPICH Release date: Sun Nov 16 11:09:31 CST 2014 > MPICH Device: ch3:sock > MPICH configure: --prefix /home/hroy/local//mpich-3.2a2/linux_x86_64 --disable-f77 --disable-fc --disable-f90modules --disable-cxx --enable-fast=nochkmsg --enable-fast=notiming --enable-fast=ndebug --enable-fast=O3 --with-device=ch3:sock --enable-g=dbg --disable-fortran --without-valgrind CFLAGS=-O3 -fPIC CXXFLAGS=-O3 -fPIC > MPICH CC: /u/prod/gnu/gcc/20121129/gcc-4.5.0-linux_x86_64/bin/gcc -O3 -fPIC -g -O3 > MPICH CXX: no -O3 -fPIC -g > MPICH F77: no -g > MPICH FC: no -g > > > Thanks, > Hirak > > Could you try with the latest mpich-3.2a2? > The client exit successfully on my Macbook with sock channel. > > — > Huiwei > > > On Nov 16, 2014, at 10:44 PM, Hirak Roy <hirak_roy at mentor.com> wrote: > > > > Hi All, > > > > Here is my sample program. I am using channel sock of mpich-3.0.4. > > > > I am running it as > > > mpiexec -n 1 ./server.out > > > mpiexec -n 1 ./client.out > > > > Here my client program (client.c) hangs in MPI_Finalize. > > There is an assert in the server.c where server exits. > > > > There is no way to detect that in client. > > Even if we detect that using some timeout strategy, the client hangs in the finalize step. > > Could you please suggest what is going wrong here or is this a bug in sock channel? > > > > Thanks, > > Hirak > > <client.c><server.c>_______________________________________________ > > discuss mailing list discuss at 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