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

Re: [mpich-discuss] fail to configure mpich 3.1.4 use gcc 5.2.0
by Halim Amer 20 Oct '15

20 Oct '15

Can you first test that your gcc installation works fine (https://gcc.gnu.org/install/test.html) Then if the problem persists, please send your *config.log* to the list instead of the output of configure so that we look at it. --Halim www.mcs.anl.gov/~aamer On 10/19/15 8:32 PM, 广州刘丹 wrote: > Hello, everyone: > After I upgrade gcc from 4.4.7 to 5.2.0 under oracle linux 6.7, the > system can't pass config mpich 3.1.4 when I run: > ./configure -prefix=/usr/local/share/mpich-3.1.4/ |& tee c.txt > > below is error message: > configure: error: Unable to configure with Fortran support because > configure could not determine the size of a Fortran INTEGER. Consider > setting CROSS_F77_SIZEOF_INTEGER to the length in bytes of a Fortran INTEGER > > When I use gcc 4.4.7, there's no error to make mpich 3.1.4. > > and below is the process how to upgrade gcc: > wget ftp://ftp.gnu.org/gnu/gcc/gcc-5.2.0/gcc-5.2.0.tar.bz2 > tar -jxvf gcc-5.2.0.tar.bz2 > cd gcc-5.2.0 > ./contrib/download_prerequisites > cd .. > mkdir /usr/local/share/gcc-5.2.0 > ./gcc-5.2.0/configure --prefix=/usr/local/share/gcc-5.2.0/ > --disable-multilib --enable-languages=c,c++,objc,obj-c++,java,fortran > make -j4 > sudo make install > > > > _______________________________________________ > 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] Semantics of MPI_Send and MPI_Isend
by Jeff Hammond 20 Oct '15

20 Oct '15

I recommend that you read "Semantics of Point-to-Point Communication" in the MPI standard (section 3.5 in MPI 3.1). It has all of the information that you seek. Best, Jeff On Tue, Oct 20, 2015 at 7:58 AM, Dorier, Matthieu <mdorier(a)anl.gov> wrote: > > Hi, > > In my program I have a loop of this form: > > loop { > MPI_Send(small message) > MPI_Isend(big message) > } > do_some_computation() > MPI_Waitall > > The sends and isends all have the same target rank. > > Is there a risk that an MPI_Send forces a previously posted MPI_Isend to complete before it can complete itself (in which case I would loose the benefit of using non-blocking sends)? > > Thanks, > > Matthieu > > _______________________________________________ > 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] fail to configure mpich 3.1.4 use gcc 5.2.0
by Guo, Yanfei 20 Oct '15

20 Oct '15

Are you using fortran-5 as well? If you don't need fortran, you can disable it by passing --disable-fortran to configure. Yanfei Guo Postdoctoral Appointee MCS Division, ANL On 10/19/15, 8:32 PM, "广州刘丹" <liudaniel(a)qq.com> wrote: > > > >Hello, everyone: > After I upgrade gcc from 4.4.7 to 5.2.0 under oracle linux 6.7, the system can't pass config mpich 3.1.4 when I run: > >./configure -prefix=/usr/local/share/mpich-3.1.4/ |& tee c.txt > > >below is error message: >configure: error: Unable to configure with Fortran support because configure could not determine the size of a Fortran INTEGER. Consider setting CROSS_F77_SIZEOF_INTEGER to the length in bytes of a Fortran INTEGER > > > >When I use gcc 4.4.7, there's no error to make mpich 3.1.4. > > >and below is the process how to upgrade gcc: >wget ftp://ftp.gnu.org/gnu/gcc/gcc-5.2.0/gcc-5.2.0.tar.bz2 >tar -jxvf gcc-5.2.0.tar.bz2 >cd gcc-5.2.0 >./contrib/download_prerequisites >cd .. >mkdir /usr/local/share/gcc-5.2.0 >./gcc-5.2.0/configure --prefix=/usr/local/share/gcc-5.2.0/ --disable-multilib --enable-languages=c,c++,objc,obj-c++,java,fortran >make -j4 >sudo make install > > > _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Semantics of MPI_Send and MPI_Isend
by Dorier, Matthieu 20 Oct '15

20 Oct '15

Hi, In my program I have a loop of this form: loop { MPI_Send(small message) MPI_Isend(big message) } do_some_computation() MPI_Waitall The sends and isends all have the same target rank. Is there a risk that an MPI_Send forces a previously posted MPI_Isend to complete before it can complete itself (in which case I would loose the benefit of using non-blocking sends)? Thanks, Matthieu _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

mpich 3.1.4 with PBS Pro 12
by T. H. Chiang 20 Oct '15

20 Oct '15

Hi, I'm having trouble compiling hydra with the "--with-pbs" options, something seem to be wrong with the configure steps which causes files such as pbs_launch.c to be skipped during compilation. Anybody encountered similar problem/needs? -- yours truly, Tsung Han Chiang / 江宗翰 _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

fail to configure mpich 3.1.4 use gcc 5.2.0
by 广州刘丹 19 Oct '15

19 Oct '15

Hello, everyone: After I upgrade gcc from 4.4.7 to 5.2.0 under oracle linux 6.7, the system can't pass config mpich 3.1.4 when I run: ./configure -prefix=/usr/local/share/mpich-3.1.4/ |& tee c.txt below is error message: configure: error: Unable to configure with Fortran support because configure could not determine the size of a Fortran INTEGER. Consider setting CROSS_F77_SIZEOF_INTEGER to the length in bytes of a Fortran INTEGER When I use gcc 4.4.7, there's no error to make mpich 3.1.4. and below is the process how to upgrade gcc: wget ftp://ftp.gnu.org/gnu/gcc/gcc-5.2.0/gcc-5.2.0.tar.bz2 tar -jxvf gcc-5.2.0.tar.bz2 cd gcc-5.2.0 ./contrib/download_prerequisites cd .. mkdir /usr/local/share/gcc-5.2.0 ./gcc-5.2.0/configure --prefix=/usr/local/share/gcc-5.2.0/ --disable-multilib --enable-languages=c,c++,objc,obj-c++,java,fortran make -j4 sudo make install _______________________________________________ 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 symmetric mode execution on Xeon Phi
by mahdiyeh ghazi 19 Oct '15

19 Oct '15

Thank you Dmitry, Here is the output in verbose mode: host: aster01-mic0 host: aster01 ================================================================================================== mpiexec options: ---------------- Base path: /opt/new-mpich-host/mpich-3.1.4/1/bin/./ Launcher: (null) Debug level: 1 Enable X: -1 Global environment: ------------------- MANPATH=/opt/intel/impi/ 4.1.0.024/man:/usr/local/share/man:/usr/share/man XDG_SESSION_ID=75 HOSTNAME=aster01 TERM=xterm SHELL=/bin/bash HISTSIZE=1000 SSH_CLIENT=192.168.2.1 47980 22 OLDPWD=/opt/new-mpich/mpich-3.1.4/1/bin SSH_TTY=/dev/pts/0 USER=root LD_LIBRARY_PATH=/opt/intel/composer_xe_2015.3.187/compiler/lib/intel64:/opt/intel/impi/ 4.1.0.024/intel64/lib:/usr/local/cuda-6.5/lib64/:/usr/local/cuda-6.5/lib64/:/home/farajii/installation-files/fedora20/gdrcopy-master LS_COLORS=rs=0:di=01;34:ln=01;36:mh=00:pi=40;33:so=01;35:do=01;35:bd=40;33;01:cd=40;33;01:or=40;31;01:mi=01;05;37;41:su=37;41:sg=30;43:ca=30;41:tw=30;42:ow=34;42:st=37;44:ex=01;32:*.tar=01;31:*.tgz=01;31:*.arc=01;31:*.arj=01;31:*.taz=01;31:*.lha=01;31:*.lzh=01;31:*.lzma=01;31:*.tlz=01;31:*.txz=01;31:*.tzo=01;31:*.t7z=01;31:*.zip=01;31:*.z=01;31:*.Z=01;31:*.dz=01;31:*.gz=01;31:*.lrz=01;31:*.lz=01;31:*.lzo=01;31:*.xz=01;31:*.bz2=01;31:*.bz=01;31:*.tbz=01;31:*.tbz2=01;31:*.tz=01;31:*.deb=01;31:*.rpm=01;31:*.jar=01;31:*.war=01;31:*.ear=01;31:*.sar=01;31:*.rar=01;31:*.alz=01;31:*.ace=01;31:*.zoo=01;31:*.cpio=01;31:*.7z=01;31:*.rz=01;31:*.cab=01;31:*.jpg=01;35:*.jpeg=01;35:*.gif=01;35:*.bmp=01;35:*.pbm=01;35:*.pgm=01;35:*.ppm=01;35:*.tga=01;35:*.xbm=01;35:*.xpm=01;35:*.tif=01;35:*.tiff=01;35:*.png=01;35:*.svg=01;35:*.svgz=01;35:*.mng=01;35:*.pcx=01;35:*.mov=01;35:*.mpg=01;35:*.mpeg=01;35:*.m2v=01;35:*.mkv=01;35:*.ogm=01;35:*.mp4=01;35:*.m4v=01;35:*.mp4v=01;35:*.vob=01;35:*.qt=01;35:*.nuv=01;35:*.wmv=01;35:*.asf=01;35:*.rm=01;35:*.rmvb=01;35:*.flc=01;35:*.avi=01;35:*.fli=01;35:*.flv=01;35:*.gl=01;35:*.dl=01;35:*.xcf=01;35:*.xwd=01;35:*.yuv=01;35:*.cgm=01;35:*.emf=01;35:*.axv=01;35:*.anx=01;35:*.ogv=01;35:*.ogx=01;35:*.aac=01;36:*.au=01;36:*.flac=01;36:*.mid=01;36:*.midi=01;36:*.mka=01;36:*.mp3=01;36:*.mpc=01;36:*.ogg=01;36:*.ra=01;36:*.wav=01;36:*.axa=01;36:*.oga=01;36:*.spx=01;36:*.xspf=01;36: MAIL=/var/spool/mail/root PATH=/opt/intel/composer_xe_2015.3.187/bin/intel64:/opt/intel/impi/ 4.1.0.024/intel64/bin:/opt/intel/impi/4.1.0.024/intel64/bin:/opt/intel/impi/4.1.0.024/intel64/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin::/home/ghazim/build/aster/intel/composer_xe_2015.3.187/bin/intel64:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin:/home/farajii/builds/11-mv2.2.1/bin/:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin:/home/farajii/builds/11-mv2.2.1/bin/ PWD=/opt/new-mpich-host/mpich-3.1.4/1/bin LANG=en_US.UTF-8 HISTCONTROL=ignoredups SHLVL=1 HOME=/root LOGNAME=root CVS_RSH=ssh SSH_CONNECTION=192.168.2.1 47980 192.168.2.121 22 LESSOPEN=||/usr/bin/lesspipe.sh %s XDG_RUNTIME_DIR=/run/user/0 I_MPI_ROOT=/opt/intel/impi/4.1.0.024 _=./mpiexec.hydra Hydra internal environment: --------------------------- GFORTRAN_UNBUFFERED_PRECONNECTED=y Proxy information: ********************* [1] proxy: aster01-mic0 (1 cores) Exec list: ./hellomic (1 processes); [2] proxy: aster01 (1 cores) Exec list: ./hello-host (1 processes); ================================================================================================== [mpiexec@aster01] Timeout set to -1 (-1 means infinite) [mpiexec@aster01] Got a control port string of aster01:51327 Proxy launch args: /opt/new-mpich-host/mpich-3.1.4/1/bin/./hydra_pmi_proxy --control-port aster01:51327 --debug --rmk user --launcher ssh --demux poll --pgid 0 --retries 10 --usize -2 --proxy-id Arguments being passed to proxy 0: --version 3.1.4 --iface-ip-env-name MPIR_CVAR_CH3_INTERFACE_HOSTNAME --hostname aster01-mic0 --global-core-map 0,1,2 --pmi-id-map 0,0 --global-process-count 2 --auto-cleanup 1 --pmi-kvsname kvs_5294_0 --pmi-process-mapping (vector,(0,2,1)) --ckpoint-num -1 --global-inherited-env 26 'MANPATH=/opt/intel/impi/ 4.1.0.024/man:/usr/local/share/man:/usr/share/man' 'XDG_SESSION_ID=75' 'HOSTNAME=aster01' 'TERM=xterm' 'SHELL=/bin/bash' 'HISTSIZE=1000' 'SSH_CLIENT=192.168.2.1 47980 22' 'OLDPWD=/opt/new-mpich/mpich-3.1.4/1/bin' 'SSH_TTY=/dev/pts/0' 'USER=root' 'LD_LIBRARY_PATH=/opt/intel/composer_xe_2015.3.187/compiler/lib/intel64:/opt/intel/impi/ 4.1.0.024/intel64/lib:/usr/local/cuda-6.5/lib64/:/usr/local/cuda-6.5/lib64/:/home/farajii/installation-files/fedora20/gdrcopy-master' 'LS_COLORS=rs=0:di=01;34:ln=01;36:mh=00:pi=40;33:so=01;35:do=01;35:bd=40;33;01:cd=40;33;01:or=40;31;01:mi=01;05;37;41:su=37;41:sg=30;43:ca=30;41:tw=30;42:ow=34;42:st=37;44:ex=01;32:*.tar=01;31:*.tgz=01;31:*.arc=01;31:*.arj=01;31:*.taz=01;31:*.lha=01;31:*.lzh=01;31:*.lzma=01;31:*.tlz=01;31:*.txz=01;31:*.tzo=01;31:*.t7z=01;31:*.zip=01;31:*.z=01;31:*.Z=01;31:*.dz=01;31:*.gz=01;31:*.lrz=01;31:*.lz=01;31:*.lzo=01;31:*.xz=01;31:*.bz2=01;31:*.bz=01;31:*.tbz=01;31:*.tbz2=01;31:*.tz=01;31:*.deb=01;31:*.rpm=01;31:*.jar=01;31:*.war=01;31:*.ear=01;31:*.sar=01;31:*.rar=01;31:*.alz=01;31:*.ace=01;31:*.zoo=01;31:*.cpio=01;31:*.7z=01;31:*.rz=01;31:*.cab=01;31:*.jpg=01;35:*.jpeg=01;35:*.gif=01;35:*.bmp=01;35:*.pbm=01;35:*.pgm=01;35:*.ppm=01;35:*.tga=01;35:*.xbm=01;35:*.xpm=01;35:*.tif=01;35:*.tiff=01;35:*.png=01;35:*.svg=01;35:*.svgz=01;35:*.mng=01;35:*.pcx=01;35:*.mov=01;35:*.mpg=01;35:*.mpeg=01;35:*.m2v=01;35:*.mkv=01;35:*.ogm=01;35:*.mp4=01;35:*.m4v=01;35:*.mp4v=01;35:*.vob=01;35:*.qt=01;35:*.nuv=01;35:*.wmv=01;35:*.asf=01;35:*.rm=01;35:*.rmvb=01;35:*.flc=01;35:*.avi=01;35:*.fli=01;35:*.flv=01;35:*.gl=01;35:*.dl=01;35:*.xcf=01;35:*.xwd=01;35:*.yuv=01;35:*.cgm=01;35:*.emf=01;35:*.axv=01;35:*.anx=01;35:*.ogv=01;35:*.ogx=01;35:*.aac=01;36:*.au=01;36:*.flac=01;36:*.mid=01;36:*.midi=01;36:*.mka=01;36:*.mp3=01;36:*.mpc=01;36:*.ogg=01;36:*.ra=01;36:*.wav=01;36:*.axa=01;36:*.oga=01;36:*.spx=01;36:*.xspf=01;36:' 'MAIL=/var/spool/mail/root' 'PATH=/opt/intel/composer_xe_2015.3.187/bin/intel64:/opt/intel/impi/ 4.1.0.024/intel64/bin:/opt/intel/impi/4.1.0.024/intel64/bin:/opt/intel/impi/4.1.0.024/intel64/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin::/home/ghazim/build/aster/intel/composer_xe_2015.3.187/bin/intel64:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin:/home/farajii/builds/11-mv2.2.1/bin/:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin:/home/farajii/builds/11-mv2.2.1/bin/' 'PWD=/opt/new-mpich-host/mpich-3.1.4/1/bin' 'LANG=en_US.UTF-8' 'HISTCONTROL=ignoredups' 'SHLVL=1' 'HOME=/root' 'LOGNAME=root' 'CVS_RSH=ssh' 'SSH_CONNECTION=192.168.2.1 47980 192.168.2.121 22' 'LESSOPEN=||/usr/bin/lesspipe.sh %s' 'XDG_RUNTIME_DIR=/run/user/0' 'I_MPI_ROOT=/opt/intel/impi/4.1.0.024' '_=./mpiexec.hydra' --global-user-env 0 --global-system-env 1 'GFORTRAN_UNBUFFERED_PRECONNECTED=y' --proxy-core-count 1 --exec --exec-appnum 0 --exec-proc-count 1 --exec-local-env 0 --exec-wdir /opt/new-mpich-host/mpich-3.1.4/1/bin --exec-args 1 ./hellomic Arguments being passed to proxy 1: --version 3.1.4 --iface-ip-env-name MPIR_CVAR_CH3_INTERFACE_HOSTNAME --hostname aster01 --global-core-map 0,1,2 --pmi-id-map 0,1 --global-process-count 2 --auto-cleanup 1 --pmi-kvsname kvs_5294_0 --pmi-process-mapping (vector,(0,2,1)) --ckpoint-num -1 --global-inherited-env 26 'MANPATH=/opt/intel/impi/ 4.1.0.024/man:/usr/local/share/man:/usr/share/man' 'XDG_SESSION_ID=75' 'HOSTNAME=aster01' 'TERM=xterm' 'SHELL=/bin/bash' 'HISTSIZE=1000' 'SSH_CLIENT=192.168.2.1 47980 22' 'OLDPWD=/opt/new-mpich/mpich-3.1.4/1/bin' 'SSH_TTY=/dev/pts/0' 'USER=root' 'LD_LIBRARY_PATH=/opt/intel/composer_xe_2015.3.187/compiler/lib/intel64:/opt/intel/impi/ 4.1.0.024/intel64/lib:/usr/local/cuda-6.5/lib64/:/usr/local/cuda-6.5/lib64/:/home/farajii/installation-files/fedora20/gdrcopy-master' 'LS_COLORS=rs=0:di=01;34:ln=01;36:mh=00:pi=40;33:so=01;35:do=01;35:bd=40;33;01:cd=40;33;01:or=40;31;01:mi=01;05;37;41:su=37;41:sg=30;43:ca=30;41:tw=30;42:ow=34;42:st=37;44:ex=01;32:*.tar=01;31:*.tgz=01;31:*.arc=01;31:*.arj=01;31:*.taz=01;31:*.lha=01;31:*.lzh=01;31:*.lzma=01;31:*.tlz=01;31:*.txz=01;31:*.tzo=01;31:*.t7z=01;31:*.zip=01;31:*.z=01;31:*.Z=01;31:*.dz=01;31:*.gz=01;31:*.lrz=01;31:*.lz=01;31:*.lzo=01;31:*.xz=01;31:*.bz2=01;31:*.bz=01;31:*.tbz=01;31:*.tbz2=01;31:*.tz=01;31:*.deb=01;31:*.rpm=01;31:*.jar=01;31:*.war=01;31:*.ear=01;31:*.sar=01;31:*.rar=01;31:*.alz=01;31:*.ace=01;31:*.zoo=01;31:*.cpio=01;31:*.7z=01;31:*.rz=01;31:*.cab=01;31:*.jpg=01;35:*.jpeg=01;35:*.gif=01;35:*.bmp=01;35:*.pbm=01;35:*.pgm=01;35:*.ppm=01;35:*.tga=01;35:*.xbm=01;35:*.xpm=01;35:*.tif=01;35:*.tiff=01;35:*.png=01;35:*.svg=01;35:*.svgz=01;35:*.mng=01;35:*.pcx=01;35:*.mov=01;35:*.mpg=01;35:*.mpeg=01;35:*.m2v=01;35:*.mkv=01;35:*.ogm=01;35:*.mp4=01;35:*.m4v=01;35:*.mp4v=01;35:*.vob=01;35:*.qt=01;35:*.nuv=01;35:*.wmv=01;35:*.asf=01;35:*.rm=01;35:*.rmvb=01;35:*.flc=01;35:*.avi=01;35:*.fli=01;35:*.flv=01;35:*.gl=01;35:*.dl=01;35:*.xcf=01;35:*.xwd=01;35:*.yuv=01;35:*.cgm=01;35:*.emf=01;35:*.axv=01;35:*.anx=01;35:*.ogv=01;35:*.ogx=01;35:*.aac=01;36:*.au=01;36:*.flac=01;36:*.mid=01;36:*.midi=01;36:*.mka=01;36:*.mp3=01;36:*.mpc=01;36:*.ogg=01;36:*.ra=01;36:*.wav=01;36:*.axa=01;36:*.oga=01;36:*.spx=01;36:*.xspf=01;36:' 'MAIL=/var/spool/mail/root' 'PATH=/opt/intel/composer_xe_2015.3.187/bin/intel64:/opt/intel/impi/ 4.1.0.024/intel64/bin:/opt/intel/impi/4.1.0.024/intel64/bin:/opt/intel/impi/4.1.0.024/intel64/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin::/home/ghazim/build/aster/intel/composer_xe_2015.3.187/bin/intel64:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin:/home/farajii/builds/11-mv2.2.1/bin/:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin:/usr/local/cuda-6.5/bin:/home/farajii/builds/11-mv2.2.1/bin/' 'PWD=/opt/new-mpich-host/mpich-3.1.4/1/bin' 'LANG=en_US.UTF-8' 'HISTCONTROL=ignoredups' 'SHLVL=1' 'HOME=/root' 'LOGNAME=root' 'CVS_RSH=ssh' 'SSH_CONNECTION=192.168.2.1 47980 192.168.2.121 22' 'LESSOPEN=||/usr/bin/lesspipe.sh %s' 'XDG_RUNTIME_DIR=/run/user/0' 'I_MPI_ROOT=/opt/intel/impi/4.1.0.024' '_=./mpiexec.hydra' --global-user-env 0 --global-system-env 1 'GFORTRAN_UNBUFFERED_PRECONNECTED=y' --proxy-core-count 1 --exec --exec-appnum 1 --exec-proc-count 1 --exec-local-env 0 --exec-wdir /opt/new-mpich-host/mpich-3.1.4/1/bin --exec-args 1 ./hello-host [mpiexec@aster01] Launch arguments: /usr/bin/ssh -x aster01-mic0 "/opt/new-mpich-host/mpich-3.1.4/1/bin/./hydra_pmi_proxy" --control-port aster01:51327 --debug --rmk user --launcher ssh --demux poll --pgid 0 --retries 10 --usize -2 --proxy-id 0 [mpiexec@aster01] Launch arguments: /opt/new-mpich-host/mpich-3.1.4/1/bin/./hydra_pmi_proxy --control-port aster01:51327 --debug --rmk user --launcher ssh --demux poll --pgid 0 --retries 10 --usize -2 --proxy-id 1 equal [proxy:unset@aster01] HYD_pmcd_pmip_get_params (pm/pmiserv/pmip_utils.c:701): server name not available [proxy:unset@aster01] main (pm/pmiserv/pmip.c:169): bad parameters passed to the proxy [mpiexec@aster01] HYD_pmcd_pmiserv_proxy_init_cb (pm/pmiserv/pmiserv_cb.c:445): assert (!closed) failed [mpiexec@aster01] HYDT_dmxu_poll_wait_for_event (tools/demux/demux_poll.c:76): callback returned error status [mpiexec@aster01] HYD_pmci_wait_for_completion (pm/pmiserv/pmiserv_pmci.c:198): error waiting for event [mpiexec@aster01] main (ui/mpich/mpiexec.c:344): process manager error waiting for completion any suggestion? Thanks, Mahdieh _______________________________________________ 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 symmetric mode execution on Xeon Phi
by Durnov, Dmitry 19 Oct '15

19 Oct '15

Hi Mahdieh, Most likely there is an issue with host name resolution. I would recommend to use “-verbose” hydra argument for further troubleshooting. BR, Dmitry From: mahdiyeh ghazi [mailto:[email protected]] Sent: Saturday, October 17, 2015 12:13 AM To: discuss(a)mpich.org Subject: Re: [mpich-discuss] MPICH symmetric mode execution on Xeon Phi Thanks for your reply. So I created a hydra_pmi_proxy shell script as follows: #!/bin/bash ARCH=$(uname -m) if [ $ARCH == 'x86_64' ]; then #host ./hydra_pmi_proxy.host elif [ $ARCH == 'k1om' ]; then #coprocessor /opt/new-mpich-host/mpich-3.1.4/1/bin/hydra_pmi_proxy.mic fi and I run the command: ./mpirun -f ./hosts -n 1 ./hellomic : -n 1 ./hello-host in the hosts, I specified the mic0 and the host. Now I get the following error: [proxy:unset@aster01] HYD_pmcd_pmip_get_params (pm/pmiserv/pmip_utils.c:701): server name not available [proxy:unset@aster01] main (pm/pmiserv/pmip.c:169): bad parameters passed to the proxy [mpiexec@aster01] HYD_pmcd_pmiserv_proxy_init_cb (pm/pmiserv/pmiserv_cb.c:445): assert (!closed) failed [mpiexec@aster01] HYDT_dmxu_poll_wait_for_event (tools/demux/demux_poll.c:76): callback returned error status [mpiexec@aster01] HYD_pmci_wait_for_completion (pm/pmiserv/pmiserv_pmci.c:198): error waiting for event [mpiexec@aster01] main (ui/mpich/mpiexec.c:344): process manager error waiting for completion Please let me know if I am missing something. Thanks, Mahdieh ---------------------------------------------------------------------------------------------------------------- Hello Mahdieh, There was a related discussion recently: link<http://lists.mpich.org/pipermail/discuss/2015-September/004175.html> You may try similar approach if I?ve got your question in the correct way. BR, Dmitry ----------------------------------------------------------------------------------------------------------------- From: mahdiyeh ghazi [mailto:[email protected]<mailto:[email protected]>] Sent: Thursday, October 15, 2015 5:44 PM To: discuss(a)mpich.org<mailto:[email protected]> Subject: [mpich-discuss] MPICH symmetric mode execution on Xeon Phi Hi, I'm trying to use MPICH to run a code in symmetric mode on Xeon Phi. First, I installed MPICH-3.1.4 with the following configuration for Xeon Phi: ./configure --prefix=/opt/new-mpich/mpich-3.1.4/1 --with-device=ch3:nemesis --host=x86_64 --build=x86_64-k1om-linux CC=icc CXX=icpc --disable-f77 --disable-fc CFLAGS=-mmic CXXFLAGS=-mmic --disable-mcast --disable-fortran LDFLAGS='-L/home/ghazim/build/aster/intel/composer_xe_2015.3.187/compiler/lib/mic/ -Wl,-rpath,/home/ghazim/build/aster/intel/composer_xe_2015.3.187/compiler/lib/mic/' With this configuration, I could successfully run a code in native mode on Xeon Phi. In order to run the code in symmetric mode, I installed MPICH once again, this time for the host with following configuration: ./configure --prefix=/opt/new-mpich-host/mpich-3.1.4/1 Then I compiled the code two times, once with -mmic option to make executable for Xeon Phi (from /opt/new-mpich/mpich-3.1.4/1) and once again without -mmic option to make executable for the host (from /opt/new-mpich-host/mpich-3.1.4/1). Now, my question is how I can run the program with these two executable from Xeon Phi and the host. I know that MPICH supports running parallel programs with different executables with some devices like ch_p4, ch_mpl and globus2 using MPMD but I don't know how I can use this feature to run the code in symmetric mode on Xeon Phi and whether it is possible or not. Thanks, Mahdieh -------------------------------------------------------------------- Joint Stock Company Intel A/O Registered legal address: Krylatsky Hills Business Park, 17 Krylatskaya Str., Bldg 4, Moscow 121614, Russian Federation This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies. _______________________________________________ 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 symmetric mode execution on Xeon Phi
by mahdiyeh ghazi 16 Oct '15

16 Oct '15

Thanks for your reply. So I created a hydra_pmi_proxy shell script as follows: #!/bin/bash ARCH=$(uname -m) if [ $ARCH == 'x86_64' ]; then #host ./hydra_pmi_proxy.host elif [ $ARCH == 'k1om' ]; then #coprocessor /opt/new-mpich-host/mpich-3.1.4/1/bin/hydra_pmi_proxy.mic fi and I run the command: ./mpirun -f ./hosts -n 1 ./hellomic : -n 1 ./hello-host in the hosts, I specified the mic0 and the host. Now I get the following error: [proxy:unset@aster01] HYD_pmcd_pmip_get_params (pm/pmiserv/pmip_utils.c:701): server name not available [proxy:unset@aster01] main (pm/pmiserv/pmip.c:169): bad parameters passed to the proxy [mpiexec@aster01] HYD_pmcd_pmiserv_proxy_init_cb (pm/pmiserv/pmiserv_cb.c:445): assert (!closed) failed [mpiexec@aster01] HYDT_dmxu_poll_wait_for_event (tools/demux/demux_poll.c:76): callback returned error status [mpiexec@aster01] HYD_pmci_wait_for_completion (pm/pmiserv/pmiserv_pmci.c:198): error waiting for event [mpiexec@aster01] main (ui/mpich/mpiexec.c:344): process manager error waiting for completion Please let me know if I am missing something. Thanks, Mahdieh ---------------------------------------------------------------------------------------------------------------- Hello Mahdieh, There was a related discussion recently: link< http://lists.mpich.org/pipermail/discuss/2015-September/004175.html> You may try similar approach if I?ve got your question in the correct way. BR, Dmitry ----------------------------------------------------------------------------------------------------------------- From: mahdiyeh ghazi [mailto:[email protected]] Sent: Thursday, October 15, 2015 5:44 PM To: discuss(a)mpich.org Subject: [mpich-discuss] MPICH symmetric mode execution on Xeon Phi Hi, I'm trying to use MPICH to run a code in symmetric mode on Xeon Phi. First, I installed MPICH-3.1.4 with the following configuration for Xeon Phi: ./configure --prefix=/opt/new-mpich/mpich-3.1.4/1 --with-device=ch3:nemesis --host=x86_64 --build=x86_64-k1om-linux CC=icc CXX=icpc --disable-f77 --disable-fc CFLAGS=-mmic CXXFLAGS=-mmic --disable-mcast --disable-fortran LDFLAGS='-L/home/ghazim/build/aster/intel/composer_xe_2015.3.187/compiler/lib/mic/ -Wl,-rpath,/home/ghazim/build/aster/intel/composer_xe_2015. 3.187/compiler/lib/mic/' With this configuration, I could successfully run a code in native mode on Xeon Phi. In order to run the code in symmetric mode, I installed MPICH once again, this time for the host with following configuration: ./configure --prefix=/opt/new-mpich-host/mpich-3.1.4/1 Then I compiled the code two times, once with -mmic option to make executable for Xeon Phi (from /opt/new-mpich/mpich-3.1.4/1) and once again without -mmic option to make executable for the host (from /opt/new-mpich-host/mpich-3.1.4/1). Now, my question is how I can run the program with these two executable from Xeon Phi and the host. I know that MPICH supports running parallel programs with different executables with some devices like ch_p4, ch_mpl and globus2 using MPMD but I don't know how I can use this feature to run the code in symmetric mode on Xeon Phi and whether it is possible or not. Thanks, Mahdieh _______________________________________________ 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 3.1.4 with Slurm 14.11.9
by Guo, Yanfei 15 Oct '15

15 Oct '15

Hi Davide, I have tried your configuration on our machines and I could not reproduce the problem. Can you provide some more information about how you run into this problem? Yanfei Guo Postdoctoral Appointee MCS Division, ANL On 10/15/15, 10:40 AM, "Balaji, Pavan" <balaji(a)anl.gov> wrote: >Yanfei, Ken, > >Notice that the user is having trouble because the SLURM path is not passed to the mpich test suite (not mpich itself). My guess is that we did something weird in passing the paths downstream since setting xFLAGS didn't help the user either. > > -- Pavan > > > > >On 10/15/15, 10:37 AM, "Guo, Yanfei" <yguo(a)anl.gov> wrote: > >>Then, can you check if “/usr/scheduler/slurm/lib” is in ld’s search path. You can get that by running ldconfig -v 2>/dev/null | grep -v ^$'\t’ >> >>If not, you will need to set >>LIBRARY_PATH=“/usr/scheduler/slurm/lib” >>LD_LIBRARY_PATH=“/usr/scheduler/slurm/lib” >> >> >>Yanfei Guo >>Postdoctoral Appointee >>MCS Division, ANL >> >> >> >> >> >> >> >>On 10/15/15, 10:34 AM, "Kenneth Raffenetti" <raffenet(a)mcs.anl.gov> wrote: >> >>>I think this is the right solution, but should actually be >>>"—with-slurm=/usr/scheduler/slurm". See the help text from configure: >>> >>> -with-slurm=PATH specify path where slurm include directory and >>>lib directory can be found >>> >>>Ken >>> >>>On 10/15/2015 10:25 AM, Guo, Yanfei wrote: >>>> Can you try using “—with-slurm=/usr/scheduler/slurm/lib” when you run ./configure? >>>> >>>> Yanfei Guo >>>> Postdoctoral Appointee >>>> MCS Division, ANL >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> On 10/14/15, 3:21 PM, "Davide Vanzo" <vanzod(a)accre.vanderbilt.edu> wrote: >>>> >>>>> Pavan, >>>>> maybe you missed in the crowd of my previous email. Anyway, yes, the library is there: >>>>> >>>>> >>>>> $ ls /usr/scheduler/slurm/lib >>>>> libpmi2.a libpmi.la libslurmdb.la libslurm.so.28 >>>>> libpmi2.la libpmi.so libslurmdb.so libslurm.so.28.0.0 >>>>> libpmi2.so libpmi.so.0 libslurmdb.so.28 slurm >>>>> libpmi2.so.0 libpmi.so.0.0.0 libslurmdb.so.28.0.0 >>>>> libpmi2.so.0.0.0 libslurm.a libslurm.la >>>>> libpmi.a libslurmdb.a libslurm.so >>>>> >>>>> >>>>> Davide >>>>> >>>>> >>>>> >>>>> >>>>> On Wed, 2015-10-14 at 20:17 +0000, Balaji, Pavan wrote: >>>>> >>>>> >>>>> >>>>> Can you make sure you have SLURM's libpmi in the location you specified below (/usr/scheduler/slurm/lib)? >>>>> >>>>> >>>>> -- Pavan >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> From: Davide Vanzo >>>>> Reply-To: "discuss(a)mpich.org" >>>>> Date: Wednesday, October 14, 2015 at 2:29 PM >>>>> To: "discuss(a)mpich.org" >>>>> Subject: [mpich-discuss] MPICH 3.1.4 with Slurm 14.11.9 >>>>> >>>>> >>>>> >>>>> Hello all, >>>>> I'm trying to compile MPICH 3.1.4 with GNU compilers 4.9.0 on a system with Slurm 14.11.9. I want to test the performance fall when using srun vs. mpiexec.hydra. >>>>> I followed the guide and I added the slurm flags to configure as follows: >>>>> >>>>> >>>>> >>>>> >>>>> ./configure --with-device=ch3:nemesis --prefix=$PREFIX --enable-fortran=yes --enable-cxx --enable-shared --with-pmi=slurm --with-pm=no --with-slurm=/usr/scheduler/slurm >>>>> >>>>> >>>>> >>>>> >>>>> The only compiler options I'm using are: >>>>> >>>>> >>>>> MPICHLIB_xFLAGS="-O3 -march=corei7" >>>>> >>>>> >>>>> >>>>> >>>>> MPICH builds correctly, but when I run make testing, I get the following error: >>>>> >>>>> >>>>> >>>>> >>>>> ( cd test && make testing ) >>>>> make[1]: Entering directory `/gpfs22/local/centos6/mpich/3.1.4/nehalem/gcc49/ether/source/test' >>>>> (NOXMLCLOSE=YES && export NOXMLCLOSE && cd mpi && make testing) >>>>> make[2]: Entering directory `/gpfs22/local/centos6/mpich/3.1.4/nehalem/gcc49/ether/source/test/mpi' >>>>> ./runtests -srcdir=. -tests=testlist \ >>>>> -mpiexec=/usr/local/mpich/3.1.4/nehalem/gcc49/ether/bin/mpiexec -xmlfile=summary.xml \ >>>>> -tapfile=summary.tap -junitfile=summary.junit.xml >>>>> Looking in ./testlist >>>>> Processing directory attr >>>>> Looking in ./attr/testlist >>>>> Failed to build attrt; make[3]: Entering directory `/gpfs22/local/centos6/mpich/3.1.4/nehalem/gcc49/ether/source/test/mpi/attr' >>>>> CC attrt.o >>>>> CCLD attrt >>>>> /usr/bin/ld: cannot find -lpmi >>>>> collect2: ld returned 1 exit status >>>>> make[3]: *** [attrt] Error 1 >>>>> make[3]: Leaving directory `/gpfs22/local/centos6/mpich/3.1.4/nehalem/gcc49/ether/source/test/mpi/attr' >>>>> >>>>> >>>>> >>>>> >>>>> although /usr/scheduler/slurm/lib/libpmi.la exists. >>>>> So I tried to explicitly use the lib and include path in the configure options: no luck. >>>>> I also set "-L/usr/scheduler/slurm/lib" in the xFLAGS, but no effect. >>>>> Any suggestion? >>>>> Please let me know if I missed some information and thank you in advance for your help. >>>>> >>>>> >>>>> Best regards, >>>>> Davide >>>>> >>>>> >>>>> -- >>>>> Davide Vanzo, PhD >>>>> Application Developer >>>>> Advanced Computing Center for Research and Education (ACCRE) >>>>> Vanderbilt University - Hill Center 201 >>>>> www.accre.vanderbilt.edu <http://www.accre.vanderbilt.edu/> >>>>> >>>>> >>>>> >>>>> >>>>> _______________________________________________ >>>>> 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 >_______________________________________________ >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