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

MPICH for Windows
by Luiz Carlos da Costa Junior 06 Jan '14

06 Jan '14

Hi all, It has been quite a long time since the last tie I updated the MPI's version of the Windows version of my software (I am using still using MPICH 1.4.1). I am having many problems regarding folder permissions and credentials using Windows 8. I went to MPICH's download page and I understood that MPICH's group no longer supports Windows distribution because it now points to Microsoft MPI homepage. Is this correct? Regards, Luiz

1 0

Re: [mpich-discuss] MPI fail on 20 processes, but not on 10.
by Wesley Bland 06 Jan '14

06 Jan '14

Without going through all of your code, I’d suggest making sure that you’re correctly sending and posting receives for all of the messages that you intend. I’d suggest printing out some sort of debugging messages whenever you send a message and make sure they all match up. Wesley On Jan 5, 2014, at 10:38 PM, Anatoly G <anatolyrishon(a)gmail.com> wrote: > Hi. > I have created an application. This application fails on MPI error. > Assertion failed in file src/mpid/ch3/channels/nemesis/src/ch3_progress.c at line 640: pkt->type >= 0 && pkt->type < MPIDI_NEM_PKT_END > internal ABORT - process 0 > > Scenario: > Master receives messages from slaves. > Each slave sends data using MPI_Send. > Master receives using MPI_Irecv and MPI_Recv. > > There are another errors in out*.log files. > Application doesn't fail with 10 processes, but fails with 20. > > execute command: > mpiexec.hydra -genvall -f MpiConfigMachines1.txt -launcher=rsh -n 20 /home/anatol-g/Grape/release_under_constr_MPI_tests_quantum/bin/linux64/rhe6/g++4.4.6/debug/mpi_rcv_any_multithread 100000 1000000 -1 -1 1 out > > Please help, > > Regards, > Anatoly. > > > <MpiConfigMachines1.txt><out_r19.log><mpi_rcv_any_multithread.cpp><out_r0.log><out_r1.log><out_r2.log><out_r3.log><out_r4.log><out_r5.log><out_r6.log><out_r7.log><out_r8.log><out_r9.log><out_r10.log><out_r11.log><out_r12.log><out_r13.log><out_r14.log><out_r15.log><out_r16.log><out_r17.log><out_r18.log>_______________________________________________ > 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] MPICH-3.0.4 on Ubuntu
by Wesley Bland 06 Jan '14

06 Jan '14

You need to change '/home/you/mpich-install/bin' to the path of your mpich installation that you used in earlier steps. The easier way to install on MPICH is to use the built in package manager or one of the pre-built packages available on the MPICH Downloads page. Thanks, Wesley On Jan 5, 2014, at 11:51 AM, Ярослав Иванов <ivanov.jaroslaw(a)gmail.com> wrote: > Hello, Pavan. But I see error on 6 step. > 7 step don't return any error, but I have trouble with 8 step. > setenv PATH /home/you/mpich-install/bin:$PATH > export PATH=/home/you/mpich-install/bin:$PATH > don't work. > > and > > which mpicc > which mpiexec > > don't work > > maybe is there guide for Ubuntu installing? > > > 2014/1/6 Pavan Balaji <balaji(a)mcs.anl.gov> > Hello, > > Your m.txt file does not seem to indicate any such error. > > — Pavan > > On Jan 5, 2014, at 11:13 AM, Ярослав Иванов <ivanov.jaroslaw(a)gmail.com> wrote: > > > Hello! > > > > I want to install MPICH-3.0.4 on my virtual ubuntu machine, but I got some errors. > > > > Install Guide: step 6 > > http://www.mpich.org/static/downloads/3.0.4/mpich-3.0.4-installguide.pdf > > > > When I try: > > "make |& tee m.txt" > > It return error in param_vals.c no such file > > > > So, I remake it with option VERBOSE=1. There are c and m txt files. > > <c.txt><m.txt>_______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > -- > Pavan Balaji > http://www.mcs.anl.gov/~balaji > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > > > > -- > С уважением, Иванов Ярослав > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss

1 0

MPI fail on 20 processes, but not on 10.
by Anatoly G 05 Jan '14

05 Jan '14

Hi. I have created an application. This application fails on MPI error. Assertion failed in file src/mpid/ch3/channels/nemesis/src/ch3_progress.c at line 640: pkt->type >= 0 && pkt->type < MPIDI_NEM_PKT_END internal ABORT - process 0 Scenario: Master receives messages from slaves. Each slave sends data using MPI_Send. Master receives using MPI_Irecv and MPI_Recv. There are another errors in out*.log files. Application doesn't *fail* with 10 processes, but fails with 20. execute command: mpiexec.hydra -genvall -f MpiConfigMachines1.txt -launcher=rsh -n 20 /home/anatol-g/Grape/release_under_constr_MPI_tests_quantum/bin/linux64/rhe6/g++4.4.6/debug/mpi_rcv_any_multithread 100000 1000000 -1 -1 1 out Please help, Regards, Anatoly.

1 0

Re: [mpich-discuss] build broken on 'master': 'mpitypedefs.h' file not found
by Jed Brown 05 Jan '14

05 Jan '14

Robert Latham <robl(a)mcs.anl.gov> writes: > I'll just pull the two lines I need from mpitypedefs.h instead of including the entire file. Until I do that, delete the, #include. It's only needed on bluegene Where is it used (git grep is not helping)? $ git grep MPI_AINT_CAST_TO_VOID src/mpi src/mpi/romio/adio/include/adioi.h: MPI_AINT_CAST_TO_VOID_PTR from configure (mpi.h) */ src/mpi/romio/adio/include/adioi.h: #include "mpitypedefs.h" /* for MPI_AINT_CAST_TO_VOID_PTR */ Why can't that code use ADIOI_AINT_CAST_TO_VOID_PTR, which is equivalent to MPI_AINT_CAST_TO_VOID_PTR?

1 0

Re: [mpich-discuss] MPICH-3.0.4 on Ubuntu
by Ярослав Иванов 05 Jan '14

05 Jan '14

Hello, Pavan. But I see error on 6 step. 7 step don't return any error, but I have trouble with 8 step. setenv PATH /home/you/mpich-install/bin:$PATH export PATH=/home/you/mpich-install/bin:$PATH don't work. and which mpicc which mpiexec don't work maybe is there guide for Ubuntu installing? 2014/1/6 Pavan Balaji <balaji(a)mcs.anl.gov> > Hello, > > Your m.txt file does not seem to indicate any such error. > > -- Pavan > > On Jan 5, 2014, at 11:13 AM, Ярослав Иванов <ivanov.jaroslaw(a)gmail.com> > wrote: > > > Hello! > > > > I want to install MPICH-3.0.4 on my virtual ubuntu machine, but I got > some errors. > > > > Install Guide: step 6 > > http://www.mpich.org/static/downloads/3.0.4/mpich-3.0.4-installguide.pdf > > > > When I try: > > "make |& tee m.txt" > > It return error in param_vals.c no such file > > > > So, I remake it with option VERBOSE=1. There are c and m txt files. > > <c.txt><m.txt>_______________________________________________ > > discuss mailing list discuss(a)mpich.org > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > -- > Pavan Balaji > http://www.mcs.anl.gov/~balaji > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss -- С уважением, Иванов Ярослав

1 0

Re: [mpich-discuss] MPICH-3.0.4 on Ubuntu
by Pavan Balaji 05 Jan '14

05 Jan '14

Hello, Your m.txt file does not seem to indicate any such error. — Pavan On Jan 5, 2014, at 11:13 AM, Ярослав Иванов <ivanov.jaroslaw(a)gmail.com> wrote: > Hello! > > I want to install MPICH-3.0.4 on my virtual ubuntu machine, but I got some errors. > > Install Guide: step 6 > http://www.mpich.org/static/downloads/3.0.4/mpich-3.0.4-installguide.pdf > > When I try: > "make |& tee m.txt" > It return error in param_vals.c no such file > > So, I remake it with option VERBOSE=1. There are c and m txt files. > <c.txt><m.txt>_______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

MPICH-3.0.4 on Ubuntu
by Ярослав Иванов 05 Jan '14

05 Jan '14

Hello! I want to install MPICH-3.0.4 on my virtual ubuntu machine, but I got some errors. Install Guide: step 6 http://www.mpich.org/static/downloads/3.0.4/mpich-3.0.4-installguide.pdf When I try: "make |& tee m.txt" It return error in param_vals.c no such file So, I remake it with option VERBOSE=1. There are c and m txt files.

1 0

Re: [mpich-discuss] build broken on 'master': 'mpitypedefs.h' file not found
by Robert Latham 05 Jan '14

05 Jan '14

I'll just pull the two lines I need from mpitypedefs.h instead of including the entire file. Until I do that, delete the, #include. It's only needed on bluegene -- sent from a phone. Jed Brown <jedbrown(a)mcs.anl.gov> wrote: Rob, this commit does not build, at least not if you configure out of source. commit 3146e104017114b0e9a7e04824cfeb7383155d12 Author: Rob Latham <robl(a)mcs.anl.gov> Date: Fri Jan 3 12:35:21 2014 -0600 bring back include file needed for MPI_AINT_CAST_TO_VOID_PTR In romio-standalone fixes, I zapped the inclusion of mpitypedefs.h, but the "ensure this large value fits into pointer without overflowing" check relied on some MPICH-internal macros/routines. Only Blue Gene driver uses those routines, so nightly tests didn't flag it. No reviewer. diff --git a/src/mpi/romio/adio/include/adioi.h b/src/mpi/romio/adio/include/adioi.h index ea07491..0553dfb 100644 --- a/src/mpi/romio/adio/include/adioi.h +++ b/src/mpi/romio/adio/include/adioi.h @@ -824,6 +824,7 @@ int ADIOI_MPE_iwrite_b; /* Use MPID_Ensure_Aint_fits_in_pointer from mpiutil.h and MPI_AINT_CAST_TO_VOID_PTR from configure (mpi.h) */ #include "glue_romio.h" + #include "mpitypedefs.h" /* for MPI_AINT_CAST_TO_VOID_PTR */ #define ADIOI_AINT_CAST_TO_VOID_PTR (void*)(MPIR_Pint) /* The next two casts are only used when you don't want sign extension /bin/sh ./libtool --tag=CC --mode=compile clang -DHAVE_CONFIG_H -I. -I../../../../src/mpi/romio -I./adio/include -I/home/jed/src/mpich/build-clang/src/include -I./include -I../../../../src/mpi/romio/include -I./mpi-io -I../../../../src/mpi/romio/mpi-io -I./adio/include -I../../../../src/mpi/romio/adio/include -I/home/jed/src/mpich/build-clang/src/mpl/include -I/home/jed/src/mpich/src/mpl/include -I/home/jed/src/mpich/src/openpa/src -I/home/jed/src/mpich/build-clang/src/openpa/src -I/home/jed/src/mpich/build-clang/src/mpi/romio/include -O2 -DFORTRANUNDERSCORE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -DHAVE_ROMIOCONF_H -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -MT mpi-io/close.lo -MD -MP -MF $depbase.Tpo -c -o mpi-io/close.lo ../../../../src/mpi/romio/mpi-io/close.c &&\ mv -f $depbase.Tpo $depbase.Plo libtool: compile: clang -DHAVE_CONFIG_H -I. -I../../../../src/mpi/romio -I./adio/include -I/home/jed/src/mpich/build-clang/src/include -I./include -I../../../../src/mpi/romio/include -I./mpi-io -I../../../../src/mpi/romio/mpi-io -I./adio/include -I../../../../src/mpi/romio/adio/include -I/home/jed/src/mpich/build-clang/src/mpl/include -I/home/jed/src/mpich/src/mpl/include -I/home/jed/src/mpich/src/openpa/src -I/home/jed/src/mpich/build-clang/src/openpa/src -I/home/jed/src/mpich/build-clang/src/mpi/romio/include -O2 -DFORTRANUNDERSCORE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -DHAVE_ROMIOCONF_H -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -MT mpi-io/close.lo -MD -MP -MF mpi-io/.deps/close.Tpo -c ../../../../src/mpi/romio/mpi-io/close.c -fPIC -DPIC -o mpi-io/.libs/close.o In file included from ../../../../src/mpi/romio/mpi-io/close.c:8: In file included from ../../../../src/mpi/romio/mpi-io/mpioimpl.h:15: In file included from ../../../../src/mpi/romio/adio/include/adio.h:440: ../../../../src/mpi/romio/adio/include/adioi.h:827:12: fatal error: 'mpitypedefs.h' file not found #include "mpitypedefs.h" /* for MPI_AINT_CAST_TO_VOID_PTR */ ^ 1 error generated. Now my understanding is that ROMIO is still supposed to be buildable outside the MPI source tree, in which case src/include/mpitypedefs.h will not be available. I don't know how you want to fix this.

1 0

Re: [mpich-discuss] Use of Signed-off-by in MPICH
by Jed Brown 04 Jan '14

04 Jan '14

Pavan Balaji <balaji(a)mcs.anl.gov> writes: > Please use devel(a)mpich.org for developer oriented discussion. Okay, moving it. > We use signoff as a way to get a second confirmation on the > correctness of the patch. All the projects following Linux kernel conventions (where "Signed-off-by" was invented) use Reviewed-by (and Acked-by or Tested-by) for this. I would recommend changing because it creates confusion and I don't see a tangible benefit to using the terms differently from conventional usage. Regardless, conventions like these are good to have in the developer documentation. > In some cases, multiple confirmations. There’s also a "silent > signoff” done by the jenkins testing before things are merged into > mpich/master. I thought the tests were on jenkins-ci.mcs.anl.gov, but I don't see it there any more. In any case, 'master' is currently not buildable (see my other mail).

1 0