April 2016 - discuss - mailman3.cels.anl.gov

Re: [mpich-discuss] gcc 6.0 miscompiles mpich ?
by Jeff Hammond 06 Apr '16

06 Apr '16

On Wed, Apr 6, 2016 at 9:32 AM, Halim Amer <aamer(a)anl.gov> wrote: > This information is insufficient to blame the compiler. > This statement sure seems sufficient to me: "mpich gets miscompiled when building with CFLAGS='-O2 -ftree-vectorize -g' (but not with just -O2 -g or -O3 -g)." If changing compiler flags alone causes MPICH to go from working to nonworking, it's hard not to blame the compiler. Of course, if the app flags changed, then all bets are off. Anyways, I concur that using the latest development tree of a compiler isn't a good idea. I'm sure the GCC folks would readily admit that there are bugs in the trunk. It never hurts to file this bug, but if there is no MCVE, it will be hard for anyone to make progress. Jeff > We don't know if the user (application code), MPICH, or gcc is doing > something wrong. > > First, you need to use a stable compiler version (not trunk) and a recent > MPICH (3.2 not the old 3.1 version). > > Second, you need to make sure the user code is not doing something wrong > (e.g. MPI_Alltoall could have gotten an invalid buffer address or count). > If you send us a toy program that reproduces the problem, we could help > with this. > > --Halim > www.mcs.anl.gov/~aamer > > > On 4/6/16 10:24 AM, VandeVondele Joost wrote: > >> Hi, >> >> when compiling mpich-3.1.2 using gcc trunk (which will be gcc 6.0 in a >> couple of days/weeks), I noticed mpich gets miscompiled when building >> with CFLAGS='-O2 -ftree-vectorize -g' (but not with just -O2 -g or -O3 >> -g). >> >> I believe that the miscompilation happens with the macro >> MPIDI_Request_create_sreq since a typical trace looks like: >> >> Program received signal SIGSEGV: Segmentation fault - invalid memory >> reference. >> >> Backtrace for this error: >> #0 0x33ac83269f in ??? >> #1 0x7f3743953d69 in MPIDI_Isend_self >> at src/mpid/ch3/src/mpidi_isend_self.c:34 >> #2 0x7f374394dfa3 in MPID_Isend >> at src/mpid/ch3/src/mpid_isend.c:55 >> #3 0x7f374390793a in MPIC_Isend >> at src/mpi/coll/helper_fns.c:646 >> #4 0x7f37438628c5 in MPIR_Alltoall_intra >> at src/mpi/coll/alltoall.c:520 >> #5 0x7f374386376e in MPIR_Alltoall >> at src/mpi/coll/alltoall.c:726 >> #6 0x7f3743863827 in MPIR_Alltoall_impl >> at src/mpi/coll/alltoall.c:761 >> #7 0x7f3743863e7a in PMPI_Alltoall >> at src/mpi/coll/alltoall.c:884 >> #8 0x7f37424181bf in pmpi_alltoall_ >> at src/binding/fortran/mpif_h/alltoallf.c:272 >> >> I don't know mpich well enough to extract a testcase, but it would be >> great if somebody could extract that and report to the gcc team (I'm >> happy to do the reporting if needed). >> >> thanks in advance, >> >> Joost >> >> >> >> _______________________________________________ >> 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 > -- 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] gcc 6.0 miscompiles mpich ?
by Halim Amer 06 Apr '16

06 Apr '16

This information is insufficient to blame the compiler. We don't know if the user (application code), MPICH, or gcc is doing something wrong. First, you need to use a stable compiler version (not trunk) and a recent MPICH (3.2 not the old 3.1 version). Second, you need to make sure the user code is not doing something wrong (e.g. MPI_Alltoall could have gotten an invalid buffer address or count). If you send us a toy program that reproduces the problem, we could help with this. --Halim www.mcs.anl.gov/~aamer On 4/6/16 10:24 AM, VandeVondele Joost wrote: > Hi, > > when compiling mpich-3.1.2 using gcc trunk (which will be gcc 6.0 in a > couple of days/weeks), I noticed mpich gets miscompiled when building > with CFLAGS='-O2 -ftree-vectorize -g' (but not with just -O2 -g or -O3 -g). > > I believe that the miscompilation happens with the macro > MPIDI_Request_create_sreq since a typical trace looks like: > > Program received signal SIGSEGV: Segmentation fault - invalid memory > reference. > > Backtrace for this error: > #0 0x33ac83269f in ??? > #1 0x7f3743953d69 in MPIDI_Isend_self > at src/mpid/ch3/src/mpidi_isend_self.c:34 > #2 0x7f374394dfa3 in MPID_Isend > at src/mpid/ch3/src/mpid_isend.c:55 > #3 0x7f374390793a in MPIC_Isend > at src/mpi/coll/helper_fns.c:646 > #4 0x7f37438628c5 in MPIR_Alltoall_intra > at src/mpi/coll/alltoall.c:520 > #5 0x7f374386376e in MPIR_Alltoall > at src/mpi/coll/alltoall.c:726 > #6 0x7f3743863827 in MPIR_Alltoall_impl > at src/mpi/coll/alltoall.c:761 > #7 0x7f3743863e7a in PMPI_Alltoall > at src/mpi/coll/alltoall.c:884 > #8 0x7f37424181bf in pmpi_alltoall_ > at src/binding/fortran/mpif_h/alltoallf.c:272 > > I don't know mpich well enough to extract a testcase, but it would be > great if somebody could extract that and report to the gcc team (I'm > happy to do the reporting if needed). > > thanks in advance, > > Joost > > > > _______________________________________________ > 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] mpiexec -ip\hosts
by Oden, Lena 06 Apr '16

06 Apr '16

you are missing “-host" before the IP-Address. Please use mpiexec —help, the web-documentation (http://www.mpich.org/static/docs/latest/) or stack overflow to answer questions like this in the future. This group is mainly for mpich-relatated questions, to report bugs or the like. Thanks, Lena On Apr 6, 2016, at 10:50 AM, יוסף אלון <yos104104(a)gmail.com<mailto:[email protected]>> wrote: hii i am trying to execute an mpi command to specific node. but i am not succed. the command i am trying is: mpiexec 147.161.3.201 -n 1 hostname i try to excute the command from a c program by system("mpiexec 147.161.3.201 -n 1 hostname") -- בברכה, יוסף אלון 050-4916740 _______________________________________________ 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

mpiexec -ip\hosts
by יוסף אלון 06 Apr '16

06 Apr '16

hii i am trying to execute an mpi command to specific node. but i am not succed. the command i am trying is: *mpiexec 147.161.3.201 -n 1 hostname* i try to excute the command from a c program by system("*mpiexec 147.161.3.201 -n 1 hostname")* -- בברכה, יוסף אלון 050-4916740 _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

gcc 6.0 miscompiles mpich ?
by VandeVondele Joost 06 Apr '16

06 Apr '16

Hi, when compiling mpich-3.1.2 using gcc trunk (which will be gcc 6.0 in a couple of days/weeks), I noticed mpich gets miscompiled when building with CFLAGS='-O2 -ftree-vectorize -g' (but not with just -O2 -g or -O3 -g). I believe that the miscompilation happens with the macro MPIDI_Request_create_sreq since a typical trace looks like: Program received signal SIGSEGV: Segmentation fault - invalid memory reference. Backtrace for this error: #0 0x33ac83269f in ??? #1 0x7f3743953d69 in MPIDI_Isend_self at src/mpid/ch3/src/mpidi_isend_self.c:34 #2 0x7f374394dfa3 in MPID_Isend at src/mpid/ch3/src/mpid_isend.c:55 #3 0x7f374390793a in MPIC_Isend at src/mpi/coll/helper_fns.c:646 #4 0x7f37438628c5 in MPIR_Alltoall_intra at src/mpi/coll/alltoall.c:520 #5 0x7f374386376e in MPIR_Alltoall at src/mpi/coll/alltoall.c:726 #6 0x7f3743863827 in MPIR_Alltoall_impl at src/mpi/coll/alltoall.c:761 #7 0x7f3743863e7a in PMPI_Alltoall at src/mpi/coll/alltoall.c:884 #8 0x7f37424181bf in pmpi_alltoall_ at src/binding/fortran/mpif_h/alltoallf.c:272 I don't know mpich well enough to extract a testcase, but it would be great if somebody could extract that and report to the gcc team (I'm happy to do the reporting if needed). thanks in advance, Joost _______________________________________________ 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 comm attributes not working?
by Jeff Hammond 05 Apr '16

05 Apr '16

Thanks, Ken. When I set *flag=1 in the copyfn, the output becomes: $ ./a.out In ibcast_copy: keyval = -1539309568 In ibcast_delete: keyval = -1539309568 In ibcast_delete: keyval = -1539309568 Jeff On Tue, Apr 5, 2016 at 11:08 AM, Kenneth Raffenetti <raffenet(a)mcs.anl.gov> wrote: > > I think in the example, it's undefined if the attribute will be set on the duped communicator. See pg. 269 in MPI-3.1. If the callback does not set the output flag to true/false, how will the implementation know what to do? > > Ken > > On 04/05/2016 11:05 AM, Jeff Hammond wrote: >> >> I guess I don't know what the right answer is for MPI_COMM_WORLD. I'll >> try that test again with attributes set only on subcomms to see how it >> behaves. >> >> Jeff >> >> On Tue, Apr 5, 2016 at 10:24 AM, Balaji, Pavan <balaji(a)anl.gov >> <mailto:[email protected]>> wrote: >> >> >> Hmm. I would think the mpich output is correct: the attributed >> should only be deleted once. But I could be wrong. >> >> -- Pavan >> >> > On Apr 5, 2016, at 8:10 AM, Jeff Hammond <jeff.science(a)gmail.com >> <mailto:[email protected]>> wrote: >> > >> > My concern is the number of times the string "ibcast_delete" >> appears in the output (2 vs 1)... >> > >> > Jeff >> > >> > On Mon, Apr 4, 2016 at 10:02 PM, Balaji, Pavan <balaji(a)anl.gov >> <mailto:[email protected]>> wrote: >> > >> > You are printing a number that's generated by the MPI >> implementation. It's value will be different for different MPI >> implementations. >> > >> > -- Pavan >> > >> >> On Apr 4, 2016, at 11:56 PM, Jeff Hammond >> <jeff.science(a)gmail.com <mailto:[email protected]>> wrote: >> >> >> >> I am trying to understand how communicator attributes work, so I >> found >> http://stackoverflow.com/questions/25198353/callbacks-provided-in-mpi-comm-… , >> which contains a nice example (the code in >> http://stackoverflow.com/a/25205611/2189128) >> >> >> >> However, the posted output of this code is: >> >> >> >> $ mpirun -np 1 ./comm-attr >> >> In ibcast_copy: keyval = 10 >> >> In ibcast_delete: keyval = 10 >> >> In ibcast_delete: keyval = 10 >> >> >> >> Whereas I observe: >> >> >> >> $ mpirun -np 1 ./a.out >> >> In ibcast_copy: keyval = -1539309568 >> >> In ibcast_delete: keyval = -1539309568 >> >> >> >> Can anyone help me understand what the correct output of this >> program is? If it is not what I am seeing, is this a bug in MPICH >> or am I doing something wrong? >> >> >> >> I am using MPICH git from November 2015, so I will git pull && >> ./autogen.sh && $CONFIGUREMAKEINSTALL and try again just to be sure. >> >> >> >> Thanks, >> >> >> >> Jeff >> >> >> >> $ mpirun --version >> >> HYDRA build details: >> >> Version: 3.2 >> >> Release Date: unreleased >> development copy >> >> CC: icc >> >> CXX: icpc >> >> F77: ifort >> >> F90: ifort >> >> Configure options: >> '--disable-option-checking' >> '--prefix=/opt/mpich/dev/intel/default' 'CC=icc' 'CXX=icpc' >> 'FC=ifort' 'F77=ifort' '--enable-cxx' '--enable-fortran' >> '--enable-threads=runtime' '--enable-g=dbg' '--with-pm=hydra' >> '--enable-wrapper-rpath' '--enable-static' '--enable-shared' >> '--cache-file=/dev/null' '--srcdir=../../../../src/pm/hydra' >> 'CFLAGS= -g -O2' 'LDFLAGS=' 'LIBS=-lpthread ' 'CPPFLAGS= >> -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpl/include >> -I/Users/jrhammon/Work/MPI/MPICH/git/src/mpl/include >> -I/Users/jrhammon/Work/MPI/MPICH/git/src/openpa/src >> -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/openpa/src >> -D_REENTRANT >> -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpi/romio/include' >> >> Process Manager: pmi >> >> Launchers available: ssh rsh fork slurm >> ll lsf sge manual persist >> >> Topology libraries available: hwloc >> >> Resource management kernels available: user slurm ll lsf >> sge pbs cobalt >> >> Checkpointing libraries available: >> >> Demux engines available: poll select >> >> >> >> >> >> -- >> >> Jeff Hammond >> >> jeff.science(a)gmail.com <mailto:[email protected]> >> >> http://jeffhammond.github.io/ >> >> _______________________________________________ >> >> 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 <mailto:[email protected]> >> > To manage subscription options or unsubscribe: >> > https://lists.mpich.org/mailman/listinfo/discuss >> > >> > >> > >> > -- >> > Jeff Hammond >> > jeff.science(a)gmail.com <mailto:[email protected]> >> > http://jeffhammond.github.io/ >> > _______________________________________________ >> > 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 <mailto:[email protected]> >> To manage subscription options or unsubscribe: >> https://lists.mpich.org/mailman/listinfo/discuss >> >> >> >> >> -- >> Jeff Hammond >> jeff.science(a)gmail.com <mailto:[email protected]> >> http://jeffhammond.github.io/ >> >> >> _______________________________________________ >> 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 -- 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] MPI comm attributes not working?
by Kenneth Raffenetti 05 Apr '16

05 Apr '16

I think in the example, it's undefined if the attribute will be set on the duped communicator. See pg. 269 in MPI-3.1. If the callback does not set the output flag to true/false, how will the implementation know what to do? Ken On 04/05/2016 11:05 AM, Jeff Hammond wrote: > I guess I don't know what the right answer is for MPI_COMM_WORLD. I'll > try that test again with attributes set only on subcomms to see how it > behaves. > > Jeff > > On Tue, Apr 5, 2016 at 10:24 AM, Balaji, Pavan <balaji(a)anl.gov > <mailto:[email protected]>> wrote: > > > Hmm. I would think the mpich output is correct: the attributed > should only be deleted once. But I could be wrong. > > -- Pavan > > > On Apr 5, 2016, at 8:10 AM, Jeff Hammond <jeff.science(a)gmail.com > <mailto:[email protected]>> wrote: > > > > My concern is the number of times the string "ibcast_delete" > appears in the output (2 vs 1)... > > > > Jeff > > > > On Mon, Apr 4, 2016 at 10:02 PM, Balaji, Pavan <balaji(a)anl.gov > <mailto:[email protected]>> wrote: > > > > You are printing a number that's generated by the MPI > implementation. It's value will be different for different MPI > implementations. > > > > -- Pavan > > > >> On Apr 4, 2016, at 11:56 PM, Jeff Hammond > <jeff.science(a)gmail.com <mailto:[email protected]>> wrote: > >> > >> I am trying to understand how communicator attributes work, so I > found > http://stackoverflow.com/questions/25198353/callbacks-provided-in-mpi-comm-…, > which contains a nice example (the code in > http://stackoverflow.com/a/25205611/2189128) > >> > >> However, the posted output of this code is: > >> > >> $ mpirun -np 1 ./comm-attr > >> In ibcast_copy: keyval = 10 > >> In ibcast_delete: keyval = 10 > >> In ibcast_delete: keyval = 10 > >> > >> Whereas I observe: > >> > >> $ mpirun -np 1 ./a.out > >> In ibcast_copy: keyval = -1539309568 > >> In ibcast_delete: keyval = -1539309568 > >> > >> Can anyone help me understand what the correct output of this > program is? If it is not what I am seeing, is this a bug in MPICH > or am I doing something wrong? > >> > >> I am using MPICH git from November 2015, so I will git pull && > ./autogen.sh && $CONFIGUREMAKEINSTALL and try again just to be sure. > >> > >> Thanks, > >> > >> Jeff > >> > >> $ mpirun --version > >> HYDRA build details: > >> Version: 3.2 > >> Release Date: unreleased > development copy > >> CC: icc > >> CXX: icpc > >> F77: ifort > >> F90: ifort > >> Configure options: > '--disable-option-checking' > '--prefix=/opt/mpich/dev/intel/default' 'CC=icc' 'CXX=icpc' > 'FC=ifort' 'F77=ifort' '--enable-cxx' '--enable-fortran' > '--enable-threads=runtime' '--enable-g=dbg' '--with-pm=hydra' > '--enable-wrapper-rpath' '--enable-static' '--enable-shared' > '--cache-file=/dev/null' '--srcdir=../../../../src/pm/hydra' > 'CFLAGS= -g -O2' 'LDFLAGS=' 'LIBS=-lpthread ' 'CPPFLAGS= > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpl/include > -I/Users/jrhammon/Work/MPI/MPICH/git/src/mpl/include > -I/Users/jrhammon/Work/MPI/MPICH/git/src/openpa/src > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/openpa/src > -D_REENTRANT > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpi/romio/include' > >> Process Manager: pmi > >> Launchers available: ssh rsh fork slurm > ll lsf sge manual persist > >> Topology libraries available: hwloc > >> Resource management kernels available: user slurm ll lsf > sge pbs cobalt > >> Checkpointing libraries available: > >> Demux engines available: poll select > >> > >> > >> -- > >> Jeff Hammond > >> jeff.science(a)gmail.com <mailto:[email protected]> > >> http://jeffhammond.github.io/ > >> _______________________________________________ > >> 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 <mailto:[email protected]> > > To manage subscription options or unsubscribe: > > https://lists.mpich.org/mailman/listinfo/discuss > > > > > > > > -- > > Jeff Hammond > > jeff.science(a)gmail.com <mailto:[email protected]> > > http://jeffhammond.github.io/ > > _______________________________________________ > > 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 <mailto:[email protected]> > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > > > > > -- > Jeff Hammond > jeff.science(a)gmail.com <mailto:[email protected]> > http://jeffhammond.github.io/ > > > _______________________________________________ > 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] MPI comm attributes not working?
by Jeff Hammond 05 Apr '16

05 Apr '16

I guess I don't know what the right answer is for MPI_COMM_WORLD. I'll try that test again with attributes set only on subcomms to see how it behaves. Jeff On Tue, Apr 5, 2016 at 10:24 AM, Balaji, Pavan <balaji(a)anl.gov> wrote: > > Hmm. I would think the mpich output is correct: the attributed should > only be deleted once. But I could be wrong. > > -- Pavan > > > On Apr 5, 2016, at 8:10 AM, Jeff Hammond <jeff.science(a)gmail.com> wrote: > > > > My concern is the number of times the string "ibcast_delete" appears in > the output (2 vs 1)... > > > > Jeff > > > > On Mon, Apr 4, 2016 at 10:02 PM, Balaji, Pavan <balaji(a)anl.gov> wrote: > > > > You are printing a number that's generated by the MPI implementation. > It's value will be different for different MPI implementations. > > > > -- Pavan > > > >> On Apr 4, 2016, at 11:56 PM, Jeff Hammond <jeff.science(a)gmail.com> > wrote: > >> > >> I am trying to understand how communicator attributes work, so I found > http://stackoverflow.com/questions/25198353/callbacks-provided-in-mpi-comm-…, > which contains a nice example (the code in > http://stackoverflow.com/a/25205611/2189128) > >> > >> However, the posted output of this code is: > >> > >> $ mpirun -np 1 ./comm-attr > >> In ibcast_copy: keyval = 10 > >> In ibcast_delete: keyval = 10 > >> In ibcast_delete: keyval = 10 > >> > >> Whereas I observe: > >> > >> $ mpirun -np 1 ./a.out > >> In ibcast_copy: keyval = -1539309568 > >> In ibcast_delete: keyval = -1539309568 > >> > >> Can anyone help me understand what the correct output of this program > is? If it is not what I am seeing, is this a bug in MPICH or am I doing > something wrong? > >> > >> I am using MPICH git from November 2015, so I will git pull && > ./autogen.sh && $CONFIGUREMAKEINSTALL and try again just to be sure. > >> > >> Thanks, > >> > >> Jeff > >> > >> $ mpirun --version > >> HYDRA build details: > >> Version: 3.2 > >> Release Date: unreleased development copy > >> CC: icc > >> CXX: icpc > >> F77: ifort > >> F90: ifort > >> Configure options: '--disable-option-checking' > '--prefix=/opt/mpich/dev/intel/default' 'CC=icc' 'CXX=icpc' 'FC=ifort' > 'F77=ifort' '--enable-cxx' '--enable-fortran' '--enable-threads=runtime' > '--enable-g=dbg' '--with-pm=hydra' '--enable-wrapper-rpath' > '--enable-static' '--enable-shared' '--cache-file=/dev/null' > '--srcdir=../../../../src/pm/hydra' 'CFLAGS= -g -O2' 'LDFLAGS=' > 'LIBS=-lpthread ' 'CPPFLAGS= > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpl/include > -I/Users/jrhammon/Work/MPI/MPICH/git/src/mpl/include > -I/Users/jrhammon/Work/MPI/MPICH/git/src/openpa/src > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/openpa/src -D_REENTRANT > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpi/romio/include' > >> Process Manager: pmi > >> Launchers available: ssh rsh fork slurm ll lsf > sge manual persist > >> Topology libraries available: hwloc > >> Resource management kernels available: user slurm ll lsf sge pbs > cobalt > >> Checkpointing libraries available: > >> Demux engines available: poll select > >> > >> > >> -- > >> 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 > > > > _______________________________________________ > > 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 > > _______________________________________________ > 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] MPI comm attributes not working?
by Balaji, Pavan 05 Apr '16

05 Apr '16

Hmm. I would think the mpich output is correct: the attributed should only be deleted once. But I could be wrong. -- Pavan > On Apr 5, 2016, at 8:10 AM, Jeff Hammond <jeff.science(a)gmail.com> wrote: > > My concern is the number of times the string "ibcast_delete" appears in the output (2 vs 1)... > > Jeff > > On Mon, Apr 4, 2016 at 10:02 PM, Balaji, Pavan <balaji(a)anl.gov> wrote: > > You are printing a number that's generated by the MPI implementation. It's value will be different for different MPI implementations. > > -- Pavan > >> On Apr 4, 2016, at 11:56 PM, Jeff Hammond <jeff.science(a)gmail.com> wrote: >> >> I am trying to understand how communicator attributes work, so I found http://stackoverflow.com/questions/25198353/callbacks-provided-in-mpi-comm-…, which contains a nice example (the code in http://stackoverflow.com/a/25205611/2189128) >> >> However, the posted output of this code is: >> >> $ mpirun -np 1 ./comm-attr >> In ibcast_copy: keyval = 10 >> In ibcast_delete: keyval = 10 >> In ibcast_delete: keyval = 10 >> >> Whereas I observe: >> >> $ mpirun -np 1 ./a.out >> In ibcast_copy: keyval = -1539309568 >> In ibcast_delete: keyval = -1539309568 >> >> Can anyone help me understand what the correct output of this program is? If it is not what I am seeing, is this a bug in MPICH or am I doing something wrong? >> >> I am using MPICH git from November 2015, so I will git pull && ./autogen.sh && $CONFIGUREMAKEINSTALL and try again just to be sure. >> >> Thanks, >> >> Jeff >> >> $ mpirun --version >> HYDRA build details: >> Version: 3.2 >> Release Date: unreleased development copy >> CC: icc >> CXX: icpc >> F77: ifort >> F90: ifort >> Configure options: '--disable-option-checking' '--prefix=/opt/mpich/dev/intel/default' 'CC=icc' 'CXX=icpc' 'FC=ifort' 'F77=ifort' '--enable-cxx' '--enable-fortran' '--enable-threads=runtime' '--enable-g=dbg' '--with-pm=hydra' '--enable-wrapper-rpath' '--enable-static' '--enable-shared' '--cache-file=/dev/null' '--srcdir=../../../../src/pm/hydra' 'CFLAGS= -g -O2' 'LDFLAGS=' 'LIBS=-lpthread ' 'CPPFLAGS= -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpl/include -I/Users/jrhammon/Work/MPI/MPICH/git/src/mpl/include -I/Users/jrhammon/Work/MPI/MPICH/git/src/openpa/src -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/openpa/src -D_REENTRANT -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpi/romio/include' >> Process Manager: pmi >> Launchers available: ssh rsh fork slurm ll lsf sge manual persist >> Topology libraries available: hwloc >> Resource management kernels available: user slurm ll lsf sge pbs cobalt >> Checkpointing libraries available: >> Demux engines available: poll select >> >> >> -- >> 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 > > _______________________________________________ > 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 _______________________________________________ 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 comm attributes not working?
by Jeff Hammond 05 Apr '16

05 Apr '16

My concern is the number of times the string "ibcast_delete" appears in the output (2 vs 1)... Jeff On Mon, Apr 4, 2016 at 10:02 PM, Balaji, Pavan <balaji(a)anl.gov> wrote: > > You are printing a number that's generated by the MPI implementation. > It's value will be different for different MPI implementations. > > -- Pavan > > > On Apr 4, 2016, at 11:56 PM, Jeff Hammond <jeff.science(a)gmail.com> > wrote: > > > > I am trying to understand how communicator attributes work, so I found > http://stackoverflow.com/questions/25198353/callbacks-provided-in-mpi-comm-…, > which contains a nice example (the code in > http://stackoverflow.com/a/25205611/2189128) > > > > However, the posted output of this code is: > > > > $ mpirun -np 1 ./comm-attr > > In ibcast_copy: keyval = 10 > > In ibcast_delete: keyval = 10 > > In ibcast_delete: keyval = 10 > > > > Whereas I observe: > > > > $ mpirun -np 1 ./a.out > > In ibcast_copy: keyval = -1539309568 > > In ibcast_delete: keyval = -1539309568 > > > > Can anyone help me understand what the correct output of this program > is? If it is not what I am seeing, is this a bug in MPICH or am I doing > something wrong? > > > > I am using MPICH git from November 2015, so I will git pull && > ./autogen.sh && $CONFIGUREMAKEINSTALL and try again just to be sure. > > > > Thanks, > > > > Jeff > > > > $ mpirun --version > > HYDRA build details: > > Version: 3.2 > > Release Date: unreleased development copy > > CC: icc > > CXX: icpc > > F77: ifort > > F90: ifort > > Configure options: '--disable-option-checking' > '--prefix=/opt/mpich/dev/intel/default' 'CC=icc' 'CXX=icpc' 'FC=ifort' > 'F77=ifort' '--enable-cxx' '--enable-fortran' '--enable-threads=runtime' > '--enable-g=dbg' '--with-pm=hydra' '--enable-wrapper-rpath' > '--enable-static' '--enable-shared' '--cache-file=/dev/null' > '--srcdir=../../../../src/pm/hydra' 'CFLAGS= -g -O2' 'LDFLAGS=' > 'LIBS=-lpthread ' 'CPPFLAGS= > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpl/include > -I/Users/jrhammon/Work/MPI/MPICH/git/src/mpl/include > -I/Users/jrhammon/Work/MPI/MPICH/git/src/openpa/src > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/openpa/src -D_REENTRANT > -I/Users/jrhammon/Work/MPI/MPICH/git/build/src/mpi/romio/include' > > Process Manager: pmi > > Launchers available: ssh rsh fork slurm ll lsf > sge manual persist > > Topology libraries available: hwloc > > Resource management kernels available: user slurm ll lsf sge pbs > cobalt > > Checkpointing libraries available: > > Demux engines available: poll select > > > > > > -- > > 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 > > _______________________________________________ > 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