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

MPICH install error (UNCLASSIFIED)
by Carrillo, Justin T ERDC-RDE-GSL-MS 15 Jan '15

15 Jan '15

Classification: UNCLASSIFIED Caveats: NONE I am attempting to install MPICH using Cygwin for Windows 7. The process seems to be successful for the majority of the installation. Thanks, Justin T. Carrillo Research Mechanical Engineer U.S. Army Engineer Research and Development Center (USACE - ERDC) Geotechnical and Structures Laboratory (GSL) Mobility Sysytems Branch (MSB) Vicksburg, MS 39180 601-634-3582 769-798-4033 Justin.T.Carrillo(a)erdc.dren.mil<mailto:[email protected]> Justin.T.Carrillo(a)usace.army.mil<mailto:[email protected]> Classification: UNCLASSIFIED Caveats: NONE

1 0

Re: [mpich-discuss] Runing mpich over InfiniBand Open Fabrics
by "Antonio J. Peña" 15 Jan '15

15 Jan '15

Hi Ramiro, Our folks that contributed that netmod are looking into this issue. In the meantime, I'd suggest trying the MXM netmod. MXM is a netmod for IB networks using the MXM API instead of Verbs to interact with the HCA. Best, Antonio On 01/15/2015 05:52 AM, Ramiro Alba wrote: > Hi all, > > I've compiled mpich-3.1.3 on centos 6.5 with the following options: > > --enable-fortran=yes \ > --with-device=ch3:nemesis:ib \ > --with-pm=hydra:gforker \ > > and the package 'libibverbs-devel' installed. > > When I try to run a test hello program using two IB DDR nodes, using the > command: > > mpiexec.hydra -np 16 -bind-to core -launcher rsh -iface ib0 -hosts > jff201,jff202 mpi_hello > > I've got the errors bellow, even running using with root user. > > If I compile with: > > --with-device=ch3:nemesis > > it works with no errors. > > I am also using both openmpi and mvapich2 on Infiniband and they work > fine > > Am I doing something wrong when compiling and/or running? > Any sugestion is welcomed? > > Thanks in advance > Regards > > ########################################################################## > > MPICH OVER IB: RUN ERRORS > ########################################################################## > > Fatal error in MPI_Init: Other MPI error, error stack: > MPIR_Init_thread(498): > MPID_Init(177).......: channel initialization failed > MPIDI_CH3_Init(89)...: > MPID_nem_init(320)...: > MPID_nem_ib_init(264): MPID_nem_ib_com_open failed > Fatal error in MPI_Init: Other MPI error, error stack: > MPIR_Init_thread(498): > MPID_Init(177).......: channel initialization failed > MPIDI_CH3_Init(89)...: > MPID_nem_init(320)...: > MPID_nem_ib_init(264): MPID_nem_ib_com_open failed > [root@jff201 mpich]# mpirun -np 2 -iface eth0 mpi_hello-mpich > IB device not foundFatal error in MPI_Init: Other MPI error, error stack: > MPIR_Init_thread(498): > MPID_Init(177).......: channel initialization failed > MPIDI_CH3_Init(89)...: > MPID_nem_init(320)...: > MPID_nem_ib_init(264): MPID_nem_ib_com_open failed > IB device not foundFatal error in MPI_Init: Other MPI error, error stack: > MPIR_Init_thread(498): > MPID_Init(177).......: channel initialization failed > MPIDI_CH3_Init(89)...: > MPID_nem_init(320)...: > MPID_nem_ib_init(264): MPID_nem_ib_com_open failed > ########################################################################## > > > > _______________________________________________ > 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] In function `MPIR_Waitall_impl': waitall.c:(.text+0x644): undefined reference to `MPI_Testall'
by Wesley Bland 15 Jan '15

15 Jan '15

I'll take a look. It might have come from one of my recent commits. On Wed, Jan 14, 2015 at 10:09 PM, Jeff Hammond <jeff.science(a)gmail.com> wrote: > I see the following issue with the latest git master on CentOS 6.6 > with GCC 4.9. Surely I will be accused of overzealous option > disabling in the name of premature optimization, but I still think > that this should work > > Thanks, > > Jeff > > GEN lib/libpmpi.la > GEN lib/libmpi.la > CCLD src/env/mpichversion > GEN lib/libmpifort.la > CCLD src/env/mpivars > > /home/jrhammon/Work/MPI/MPICH/git/build/lib/.libs/libpmpi.a(lib_libpmpi_la-waitall.o): > In function `MPIR_Waitall_impl': > waitall.c:(.text+0x644): undefined reference to `MPI_Testall' > collect2: error: ld returned 1 exit status > make[2]: *** [src/env/mpivars] Error 1 > make[2]: *** Waiting for unfinished jobs.... > make[2]: Leaving directory `/home/jrhammon/Work/MPI/MPICH/mpich/build' > make[1]: *** [install-recursive] Error 1 > make[1]: Leaving directory `/home/jrhammon/Work/MPI/MPICH/mpich/build' > make: *** [install] Error 2 > > [jrhammon@esgmonster build]$ ../configure CC=gcc CXX=g++ FC=gfortran > --disable-cxx --enable-fortran --with-pm=hydra > --prefix=/opt/mpich/dev/gcc/fast --enable-static > --enable-fast=O3,nochkmsg,notiming,ndebug,nompit > --disable-weak-symbols --enable-threads=funneled && make -j36 install > > [jrhammon@esgmonster build]$ uname -a > Linux esgmonster 2.6.32-504.3.3.el6.centos.plus.x86_64 #1 SMP Wed Dec > 17 01:21:03 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux > > [jrhammon@esgmonster build]$ gcc -v > Using built-in specs. > COLLECT_GCC=gcc > > COLLECT_LTO_WRAPPER=/opt/gcc/4.9.2/libexec/gcc/x86_64-unknown-linux-gnu/4.9.2/lto-wrapper > Target: x86_64-unknown-linux-gnu > Configured with: /home/jrhammon/Work/GCC/gcc-4.9.2/configure > --disable-multilib --enable-threads=posix --enable-checking=release > --enable-__cxa_atexit --enable-languages=c,c++,fortran > --with-tune=native --enable-bootstrap --enable-lto > --prefix=/opt/gcc/4.9.2 --with-isl --with-cloog > Thread model: posix > gcc version 4.9.2 (GCC) > > -- > 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

Runing mpich over InfiniBand Open Fabrics
by Ramiro Alba 15 Jan '15

15 Jan '15

Hi all, I've compiled mpich-3.1.3 on centos 6.5 with the following options: --enable-fortran=yes \ --with-device=ch3:nemesis:ib \ --with-pm=hydra:gforker \ and the package 'libibverbs-devel' installed. When I try to run a test hello program using two IB DDR nodes, using the command: mpiexec.hydra -np 16 -bind-to core -launcher rsh -iface ib0 -hosts jff201,jff202 mpi_hello I've got the errors bellow, even running using with root user. If I compile with: --with-device=ch3:nemesis it works with no errors. I am also using both openmpi and mvapich2 on Infiniband and they work fine Am I doing something wrong when compiling and/or running? Any sugestion is welcomed? Thanks in advance Regards ########################################################################## MPICH OVER IB: RUN ERRORS ########################################################################## Fatal error in MPI_Init: Other MPI error, error stack: MPIR_Init_thread(498): MPID_Init(177).......: channel initialization failed MPIDI_CH3_Init(89)...: MPID_nem_init(320)...: MPID_nem_ib_init(264): MPID_nem_ib_com_open failed Fatal error in MPI_Init: Other MPI error, error stack: MPIR_Init_thread(498): MPID_Init(177).......: channel initialization failed MPIDI_CH3_Init(89)...: MPID_nem_init(320)...: MPID_nem_ib_init(264): MPID_nem_ib_com_open failed [root@jff201 mpich]# mpirun -np 2 -iface eth0 mpi_hello-mpich IB device not foundFatal error in MPI_Init: Other MPI error, error stack: MPIR_Init_thread(498): MPID_Init(177).......: channel initialization failed MPIDI_CH3_Init(89)...: MPID_nem_init(320)...: MPID_nem_ib_init(264): MPID_nem_ib_com_open failed IB device not foundFatal error in MPI_Init: Other MPI error, error stack: MPIR_Init_thread(498): MPID_Init(177).......: channel initialization failed MPIDI_CH3_Init(89)...: MPID_nem_init(320)...: MPID_nem_ib_init(264): MPID_nem_ib_com_open failed ########################################################################## -- Ramiro Alba Centre Tecnològic de Tranferència de Calor http://www.cttc.upc.edu Escola Tècnica Superior d'Enginyeries Industrial i Aeronàutica de Terrassa Colom 11, E-08222, Terrassa, Barcelona, Spain Tel: (+34) 93 739 8928 -- Aquest missatge ha estat analitzat per MailScanner a la cerca de virus i d'altres continguts perillosos, i es considera que est� net. _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

In function `MPIR_Waitall_impl': waitall.c:(.text+0x644): undefined reference to `MPI_Testall'
by Jeff Hammond 14 Jan '15

14 Jan '15

I see the following issue with the latest git master on CentOS 6.6 with GCC 4.9. Surely I will be accused of overzealous option disabling in the name of premature optimization, but I still think that this should work Thanks, Jeff GEN lib/libpmpi.la GEN lib/libmpi.la CCLD src/env/mpichversion GEN lib/libmpifort.la CCLD src/env/mpivars /home/jrhammon/Work/MPI/MPICH/git/build/lib/.libs/libpmpi.a(lib_libpmpi_la-waitall.o): In function `MPIR_Waitall_impl': waitall.c:(.text+0x644): undefined reference to `MPI_Testall' collect2: error: ld returned 1 exit status make[2]: *** [src/env/mpivars] Error 1 make[2]: *** Waiting for unfinished jobs.... make[2]: Leaving directory `/home/jrhammon/Work/MPI/MPICH/mpich/build' make[1]: *** [install-recursive] Error 1 make[1]: Leaving directory `/home/jrhammon/Work/MPI/MPICH/mpich/build' make: *** [install] Error 2 [jrhammon@esgmonster build]$ ../configure CC=gcc CXX=g++ FC=gfortran --disable-cxx --enable-fortran --with-pm=hydra --prefix=/opt/mpich/dev/gcc/fast --enable-static --enable-fast=O3,nochkmsg,notiming,ndebug,nompit --disable-weak-symbols --enable-threads=funneled && make -j36 install [jrhammon@esgmonster build]$ uname -a Linux esgmonster 2.6.32-504.3.3.el6.centos.plus.x86_64 #1 SMP Wed Dec 17 01:21:03 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux [jrhammon@esgmonster build]$ gcc -v Using built-in specs. COLLECT_GCC=gcc COLLECT_LTO_WRAPPER=/opt/gcc/4.9.2/libexec/gcc/x86_64-unknown-linux-gnu/4.9.2/lto-wrapper Target: x86_64-unknown-linux-gnu Configured with: /home/jrhammon/Work/GCC/gcc-4.9.2/configure --disable-multilib --enable-threads=posix --enable-checking=release --enable-__cxa_atexit --enable-languages=c,c++,fortran --with-tune=native --enable-bootstrap --enable-lto --prefix=/opt/gcc/4.9.2 --with-isl --with-cloog Thread model: posix gcc version 4.9.2 (GCC) -- 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] [OMPI users] OpenMPI 1.8.4rc3, 1.6.5 and 1.6.3: segmentation violation in mca_io_romio_dist_MPI_File_close
by Rob Latham 14 Jan '15

14 Jan '15

On 12/17/2014 07:04 PM, Eric Chamberland wrote: > Hi! > > Here is a "poor man's fix" that works for me (the idea is not from me, > thanks to Thomas H.): > > #1- char* lCwd = getcwd(0,0); > #2- chdir(lPathToFile); > #3- MPI_File_open(...,lFileNameWithoutTooLongPath,...); > #4- chdir(lCwd); > #5- ... > > I think there are some limitations but it works very well for our > uses... and until a "real" fix is proposed... Thanks for the bug report and test cases. I just pushed two fixes for master that fix the problem you were seeing: http://git.mpich.org/mpich.git/commit/ed39c901 http://git.mpich.org/mpich.git/commit/a30a4721a2 ==rob -- Rob Latham Mathematics and Computer Science Division Argonne National Lab, IL USA _______________________________________________ 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] bug report: incorrect lower bound from MPI_Type_create_struct
by Rob Latham 14 Jan '15

14 Jan '15

Hi Wei-keng. I've recorded this as http://trac.mpich.org/projects/mpich/ticket/2219 ==rob On 01/14/2015 02:47 PM, Wei-keng Liao wrote: > Attached program is an implementation of example 4.6 from MPI-3 standard. > It prints the lower bound, upper bound, extent, size, etc. of a datatype > created > by MPI_Type_create_struct(). The value of lower bound is incorrect. > > Here is the description of example 4.6. > > Let type1 have type map, {(double, 0), (char, 8)}, with extent 16. > Let B = (2, 1, 3), D = (0, 16, 26), and T = (MPI_FLOAT, type1, MPI_CHAR). > Then a call to MPI_TYPE_CREATE_STRUCT(3, B, D, T, newtype) returns a > datatype with type map, > {(float, 0), (float, 4), (double, 16), (char, 24), (char, 26), (char, > 27), (char, 28)}. > > According to the type map of "newtype", its LB should be 0, but the > result from > MPICH gives 16. Rajeev believes this is because of the LB of a resized > type was mistakenly > carried over to the successive call to MPI_Type_create_struct(). -- Rob Latham Mathematics and Computer Science Division Argonne National Lab, IL USA _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

bug report: incorrect lower bound from MPI_Type_create_struct
by Wei-keng Liao 14 Jan '15

14 Jan '15

1 0

Re: [mpich-discuss] mpich2 1.5: failing tests src/mpe2
by Rob Latham 13 Jan '15

13 Jan '15

On 01/13/2015 01:13 PM, Giordano Lipari wrote: > I have installed mpchi2-1.5 in my Linux desktop (Ubuntu 12.04 LTS > 64-bit) with the standard command sequence > > sudo ./configure --prefix=$MOUNTDIR > sudo make > sudo make install > sudo make check > > The following 9 tests of the check in src/mpe2 fail: Do you need MPE, jumpshot, collcheck, or any of their cousins? You can safely ignore this error if so. If you take Wes's advice and upgrade to 3.1.2, you won't even have MPE to worry about, as we made it a standalone "community supported" project a couple years back: run from git, as only the git version is aware of the new const-ness of the MPI functions: http://git.mpich.org/mpe.git/ ==rob -- Rob Latham Mathematics and Computer Science Division Argonne National Lab, IL USA _______________________________________________ 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] mpich2 1.5: failing tests src/mpe2
by Wesley Bland 13 Jan '15

13 Jan '15

MPICH2 is the same project as regular old MPICH. The name was changed when moving to version 3 to reduce future confusion. I'm not sure what feature your application is using, but it should still be there in the latest versions. Regardless, what I'm trying to say is that while we can help you with problems in the latest few versions, we probably won't be able to help with such old versions. On Tue, Jan 13, 2015 at 1:28 PM, Giordano Lipari < giordano_lipari(a)yahoo.co.uk> wrote: > I do have mpich 3.1.2 installed but I specifically installed mpich2 on > request of another application that uses it. The version of mpich2 > available from the Ubuntu repository is 1.4.1. I actually downloaded my > mpchi2 1-5 from http://www.mpich.org/static/downloads/ in the > understanding that this is the most recent available (last modified April > 2013). > > How does this additional information change the picture? > > ------------------------------ > *From:* Wesley Bland <wbland(a)anl.gov> > *To:* Giordano Lipari <giordano_lipari(a)yahoo.co.uk>; "discuss(a)mpich.org" < > discuss(a)mpich.org> > *Sent:* Tuesday, 13 January 2015, 20:15 > *Subject:* Re: [mpich-discuss] mpich2 1.5: failing tests src/mpe2 > > Have you tried updating to the latest version of MPICH? Our current > version is 3.1.3. The version you are using is years old so we'll have a > hard time helping you fix any issues. > > Thanks, > Wesley > > On Tue, Jan 13, 2015 at 1:13 PM, Giordano Lipari < > giordano_lipari(a)yahoo.co.uk> wrote: > > I have installed mpchi2-1.5 in my Linux desktop (Ubuntu 12.04 LTS > 64-bit) with the standard command sequence > > sudo ./configure --prefix=$MOUNTDIR > sudo make > sudo make install > sudo make check > > The following 9 tests of the check in src/mpe2 fail: > *** Test C program with the MPI tracing library .......................... > No. > *** Test C program with the MPI logging library .......................... > No. > *** Test C program with the MPI and manual logging libraries ............. > No. > *** Test F77 program with the MPI and manual logging libraries ........... > No. > *** Test C program with the MPI animation library ........................ > No. > *** Test C program with the X11 graphics library ......................... > No. > *** Test F77 program with the X11 graphics library ....................... > No. > *** Test C program with the MPI collective/datatype checking library ..... > No. > *** Test F77 program with the MPI collective/datatype checking library ... > No. > > The entire stdout and stderr outputs are enclosed as attachment. > My question is whether this is the regular behaviour and, if not, how I > fix this. > Thanks for helping me out! > > > > > _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0