February 2010 - Mpich2-maint - mailman3.cels.anl.gov

Re: [mpich2-maint] mpich configuration cross compile issue
by Rajeev Thakur 23 Feb '10

23 Feb '10

MPICH-1 is old and unsupported. Can you try using MPICH2 instead? Rajeev > -----Original Message----- > From: mpich2-bugs-bounces(a)mcs.anl.gov > [mailto:[email protected]] On Behalf Of John Xu > Sent: Tuesday, February 23, 2010 2:35 AM > To: mpi-bugs(a)mcs.anl.gov > Subject: [mpich2-maint] mpich configuration cross compile issue > > Hi, > > I am trying to build MPICH for alpha linux with a cross > compiler build with > cross tools(gcc-4.0.2-glibc-2.3.5). I am trying to use > ch_shmem without ROMIO. > > Am keep getting the shared memory lock problem with the compiler. > > I tried many different ways but can not get the mpi configure to work. > > I guess that I am not the only one having this problem, but I searched > everything on web > and can not find anything directly related or solving my problem. > > Could somebody give me a pointer of such CROSS compilation guide for > MPICH? Or give > me some hint about what is going on here. Thanks a lot for your help. > > Version of MPICH, mpich-1.2.7p1 > > Host, > Linux ubuntu 2.6.27-9-generic #1 SMP Thu Nov 20 21:57:00 UTC > 2008 i686 GNU/Linux > > The command I am using > setenv CC > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/gcc > setenv GCC > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/gcc > setenv AR > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/ar > setenv RANLIB > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/ranlib > setenv C_LINKER > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/ld > setenv CC_LINKER > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/ld > ./configure --with-device=ch_shmem --with-arch=alpha > --prefix=/host/benchmarks/mpich/mpich --without-romio --without-mpe > --disable-f77 --disable-f90 --disable-cxx | & tee cfg.log > > Configure output, > ubuntu:~/benchmarks/mpich/mpich-1.2.7p1> ./configure > --with-device=ch_shmem --with-arch=alpha > --prefix=/host/benchmarks/mpich/mpich --without-romio --without-mpe > --disable-f77 --disable-f90 --disable-cxx | & tee cfg.log > Configuring with args --with-device=ch_shmem --with-arch=alpha > --prefix=/host/benchmarks/mpich/mpich --without-romio --without-mpe > --disable-f77 --disable-f90 --disable-cxx > Configuring MPICH Version 1.2.7 (release) of : 2005/11/04 11:54:51 > Make will not build MPE routines > Don't build the Fortran interfaces > checking whether filesystem respects case in file names... yes > checking for current directory name... > /home/john/benchmarks/mpich/mpich-1.2.7p1 > checking for install > checking for ar... no > checking for gnumake... yes using --no-print-directory > checking whether make supports include... yes > checking for OSF V3 make... no > checking for virtual path format... VPATH > checking for cc... no > checking whether the compiler > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/gcc > accepts ANSI prototypes... yes > Will not build the Fortran interfaces > checking for ar... found > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/ar > (ar) > checking for whether compiler works... yes > checking whether the compiler > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/gcc > accepts ANSI prototypes... yes > checking whether the compiler > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/gcc > accepts ## for concatenation in cpp... yes > checking whether the compiler > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/gcc > accepts const modifier... yes > checking for extension on executables... none > checking how to run the C preprocessor > checking for AIX > checking for minix/config.h > checking for volatile... yes > Running device-specific setup program > checking whether cross-compiling... yes > checking for mutex_init in -lthread... no > checking how to run the C preprocessor... > /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/a lpha-unknown-linux-gnu/bin/gcc > -E > checking for sched.h... yes > checking for errno.h... yes > checking for library containing sched_yield... none required > checking for usnewlock... no > checking for tslock_init... no > checking for library containing gethostbyname... none required > checking for msem_init... no > checking for mutex_init... no > checking for strerror... yes > checking for shmat... yes > checking for semop... yes > checking for mmap... yes > checking for sem_post... no > checking for atexit... yes > checking that shared, anonymous mmap works... no > checking whether semget works... checking for setsid... yes > checking for isatty... yes > checking for union semun... undefined > checking whether semctl needs union semun... no > checking that usable shared memory locks were found... no > Aborting device setup because device configure failed > > Best regards, > john > _______________________________________________ > mpich2-maint mailing list > mpich2-maint(a)lists.mcs.anl.gov > https://lists.mcs.anl.gov/mailman/listinfo/mpich2-maint >

1 0

mpich configuration cross compile issue
by John Xu 23 Feb '10

23 Feb '10

Hi, I am trying to build MPICH for alpha linux with a cross compiler build with cross tools(gcc-4.0.2-glibc-2.3.5). I am trying to use ch_shmem without ROMIO. Am keep getting the shared memory lock problem with the compiler. I tried many different ways but can not get the mpi configure to work. I guess that I am not the only one having this problem, but I searched everything on web and can not find anything directly related or solving my problem. Could somebody give me a pointer of such CROSS compilation guide for MPICH? Or give me some hint about what is going on here. Thanks a lot for your help. Version of MPICH, mpich-1.2.7p1 Host, Linux ubuntu 2.6.27-9-generic #1 SMP Thu Nov 20 21:57:00 UTC 2008 i686 GNU/Linux The command I am using setenv CC /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/gcc setenv GCC /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/gcc setenv AR /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/ar setenv RANLIB /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/ranlib setenv C_LINKER /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/ld setenv CC_LINKER /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/ld ./configure --with-device=ch_shmem --with-arch=alpha --prefix=/host/benchmarks/mpich/mpich --without-romio --without-mpe --disable-f77 --disable-f90 --disable-cxx | & tee cfg.log Configure output, ubuntu:~/benchmarks/mpich/mpich-1.2.7p1> ./configure --with-device=ch_shmem --with-arch=alpha --prefix=/host/benchmarks/mpich/mpich --without-romio --without-mpe --disable-f77 --disable-f90 --disable-cxx | & tee cfg.log Configuring with args --with-device=ch_shmem --with-arch=alpha --prefix=/host/benchmarks/mpich/mpich --without-romio --without-mpe --disable-f77 --disable-f90 --disable-cxx Configuring MPICH Version 1.2.7 (release) of : 2005/11/04 11:54:51 Make will not build MPE routines Don't build the Fortran interfaces checking whether filesystem respects case in file names... yes checking for current directory name... /home/john/benchmarks/mpich/mpich-1.2.7p1 checking for install checking for ar... no checking for gnumake... yes using --no-print-directory checking whether make supports include... yes checking for OSF V3 make... no checking for virtual path format... VPATH checking for cc... no checking whether the compiler /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/gcc accepts ANSI prototypes... yes Will not build the Fortran interfaces checking for ar... found /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/ar (ar) checking for whether compiler works... yes checking whether the compiler /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/gcc accepts ANSI prototypes... yes checking whether the compiler /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/gcc accepts ## for concatenation in cpp... yes checking whether the compiler /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/gcc accepts const modifier... yes checking for extension on executables... none checking how to run the C preprocessor checking for AIX checking for minix/config.h checking for volatile... yes Running device-specific setup program checking whether cross-compiling... yes checking for mutex_init in -lthread... no checking how to run the C preprocessor... /opt/crosstool/gcc-4.0.2-glibc-2.3.5/alpha-unknown-linux-gnu/alpha-unknown-linux-gnu/bin/gcc -E checking for sched.h... yes checking for errno.h... yes checking for library containing sched_yield... none required checking for usnewlock... no checking for tslock_init... no checking for library containing gethostbyname... none required checking for msem_init... no checking for mutex_init... no checking for strerror... yes checking for shmat... yes checking for semop... yes checking for mmap... yes checking for sem_post... no checking for atexit... yes checking that shared, anonymous mmap works... no checking whether semget works... checking for setsid... yes checking for isatty... yes checking for union semun... undefined checking whether semctl needs union semun... no checking that usable shared memory locks were found... no Aborting device setup because device configure failed Best regards, john

1 0

ICNC'10-FSKD'10 Final Deadline 10 March, Yantai, China [EI/ISTP/IEEE Xplore]
by ICNC'10-FSKD'10 22 Feb '10

22 Feb '10

Dear Colleague, Due to Lunar New Year holidays and numerous requests, the 6th International Conference on Natural Computation (ICNC'10) and the 7th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD'10) is extending the submission deadline to 10 March 2010. We cordially invite you to submit a paper or invited session proposal. Yantai was listed as one of the world's most inhabitable places by the United Nations and was recognized as the "most charming city of China" by China Central Television. Undulating hills rise above the area's many rivers and are framed by beaches and neighboring islands. Famous tourist attractions include the Tashan Mountain, Kongdong Island, and Penglai Pavilion Scenic Area. Seafood and fruits are plentiful in Yantai. Selected best papers will appear in SCI-indexed journal(s). All papers in conference proceedings will be indexed by both EI Compendex and ISTP, as well as the IEEE Xplore. ICNC-FSKD is a premier international forum for scientists and researchers to present the state of the art of data mining and intelligent methods inspired from nature, particularly biological, linguistic, and physical systems, with applications to signal processing, design, and more. This is an exciting and emerging interdisciplinary area in which a wide range of theory and methodologies are being investigated and developed to tackle complex and challenging problems. Previously, the joint conferences in 2005 through 2009 each attracted over 3000 submissions from around the world. ICNC'10-FSKD'10 is technically co-sponsored by the IEEE Circuits and Systems Society. The registration fee of US*D 390 includes proceedings, lunches, dinners, banquet, and all technical sessions. To promote international participation of researchers from outside the country/region where the conference is held (i.e., China's mainland), researchers outside of China's mainland are encouraged to propose invited sessions. The first author of each paper in an invited session must not be affiliated with an organization in China's mainland. All papers in the invited sessions can be marked as "Invited Paper". One organizer for each invited session with at least 6 registered papers will enjoy an honorarium of US*D 400. Invited session organizers will solicit submissions, conduct reviews and recommend accept/reject decisions on the submitted papers. Invited session organizers will be able to set their own submission and review schedules, as long as a list of recommended papers is determined by 30 March 2010. Each invited session proposal should include: (1) the name, bio, and contact information of each organizer of the invited session; (2) the title and a short synopsis of the invited session. Please send your proposal to icnc2010(a)ytu.edu.cn For more information, visit the conference web page: http://icnc-fskd2010.ytu.edu.cn/ If you have any questions after visiting the conference web page, please email the secretariat at icnc2010(a)ytu.edu.cn Join us at this major event in beautiful Yantai !!! Organizing Committee icnc2010(a)ytu.edu.cn P.S.: Kindly forward to your colleagues or students who may be interested. If you wish to unsubscribe, in which case we apologize, please reply with "unsubscribe mpich2-maint(a)mcs.anl.gov " in your email subject. Thanks.

1 0

ICNC'10-FSKD'10 Final Deadline 10 March, Yantai, China [EI/ISTP/IEEE Xplore]
by ICNC'10-FSKD'10 22 Feb '10

22 Feb '10

Dear Colleague, Due to Lunar New Year holidays and numerous requests, the 6th International Conference on Natural Computation (ICNC'10) and the 7th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD'10) is extending the submission deadline to 10 March 2010. We cordially invite you to submit a paper or invited session proposal. Yantai was listed as one of the world's most inhabitable places by the United Nations and was recognized as the "most charming city of China" by China Central Television. Undulating hills rise above the area's many rivers and are framed by beaches and neighboring islands. Famous tourist attractions include the Tashan Mountain, Kongdong Island, and Penglai Pavilion Scenic Area. Seafood and fruits are plentiful in Yantai. Selected best papers will appear in SCI-indexed journal(s). All papers in conference proceedings will be indexed by both EI Compendex and ISTP, as well as the IEEE Xplore. ICNC-FSKD is a premier international forum for scientists and researchers to present the state of the art of data mining and intelligent methods inspired from nature, particularly biological, linguistic, and physical systems, with applications to signal processing, design, and more. This is an exciting and emerging interdisciplinary area in which a wide range of theory and methodologies are being investigated and developed to tackle complex and challenging problems. Previously, the joint conferences in 2005 through 2009 each attracted over 3000 submissions from around the world. ICNC'10-FSKD'10 is technically co-sponsored by the IEEE Circuits and Systems Society. The registration fee of US*D 390 includes proceedings, lunches, dinners, banquet, and all technical sessions. To promote international participation of researchers from outside the country/region where the conference is held (i.e., China's mainland), researchers outside of China's mainland are encouraged to propose invited sessions. The first author of each paper in an invited session must not be affiliated with an organization in China's mainland. All papers in the invited sessions can be marked as "Invited Paper". One organizer for each invited session with at least 6 registered papers will enjoy an honorarium of US*D 400. Invited session organizers will solicit submissions, conduct reviews and recommend accept/reject decisions on the submitted papers. Invited session organizers will be able to set their own submission and review schedules, as long as a list of recommended papers is determined by 30 March 2010. Each invited session proposal should include: (1) the name, bio, and contact information of each organizer of the invited session; (2) the title and a short synopsis of the invited session. Please send your proposal to icnc2010(a)ytu.edu.cn For more information, visit the conference web page: http://icnc-fskd2010.ytu.edu.cn/ If you have any questions after visiting the conference web page, please email the secretariat at icnc2010(a)ytu.edu.cn Join us at this major event in beautiful Yantai !!! Organizing Committee icnc2010(a)ytu.edu.cn P.S.: Kindly forward to your colleagues or students who may be interested. If you wish to unsubscribe, in which case we apologize, please reply with "unsubscribe mpi-maint(a)mcs.anl.gov " in your email subject. Thanks.

1 0

Re: [mpich2-maint] #1010: mpiexec/Hydra dumps core on NetBSD/i386
by Aleksej Saushev 21 Feb '10

21 Feb '10

"mpich2" <mpich2-maint(a)mcs.anl.gov> writes: > #1010: mpiexec/Hydra dumps core on NetBSD/i386 > > * owner: => balaji > * priority: major => blocker > * status: new => assigned > * milestone: => mpich2-1.3 > > Comment: > > Thanks for reporting this. I looked through the code, and I do see a > problem. But I'm surprised that our tests didn't catch this. Would you be > willing to try out a patch once I have one ready? Sure.

1 0

This is pretty interesting...
by Shikhar 20 Feb '10

20 Feb '10

1 0

Re: [mpich2-maint] configure and build problem
by Rajeev Thakur 17 Feb '10

17 Feb '10

Looks like you were installing MPICH-1, not MPICH2. (The make.log file refers to MPICH 1.2.7). Please download the latest version of MPICH2 from the web site and try again. Rajeev _____ From: mpich2-bugs-bounces(a)mcs.anl.gov [mailto:[email protected]] On Behalf Of Jankins Sent: Wednesday, February 17, 2010 8:47 PM To: mpich2-maint(a)mcs.anl.gov Subject: [mpich2-maint] configure and build problem Hi, I followed the instructions of "MPICH2Installer'sGuide" and failed to configure and install MPICH2. There is no error message at the "configuration", "make" and "make install" steps. But there is no 'mpd', 'mpiexec',etc. programs generated. The attachment is the "config.log", "make.log" and "install.log". Thanks so much for your help. Yours sincerely, Jankins

1 0

configure and build problem
by Jankins 17 Feb '10

17 Feb '10

Hi, I followed the instructions of "MPICH2Installer’sGuide" and failed to configure and install MPICH2. There is no error message at the "configuration", "make" and "make install" steps. But there is no 'mpd', 'mpiexec',etc. programs generated. The attachment is the "config.log", "make.log" and "install.log". Thanks so much for your help. Yours sincerely, Jankins

1 0

NetUltimate is looking for Tele Marketers to work from Home
by kirtana 16 Feb '10

16 Feb '10

NetULtimate.com is Currenlty looking for Tele Marketers who want to work from home between 9.30 AM to 5.30 PM EST. Training will be provided. Good Spoken English manditory. If you are interested contact us at ASAP in Yahoo Messanger 'dilipan(a)yahoo.com' or by email dilipan(a)netultimate.com or (703) 849 1269 (USA) / (416) 238 0270 (Canada) for further discussions. We also have work from home data entry spare time position with diltana.com This position is not for companies and only for individuals who want to work from home. Best Regards, Unsubscribe

1 0

Problem in MPI executing program using MPICH
by Priyank Raj 06 Feb '10

06 Feb '10

Hi I follow the all instruction given in http://202.117.122.28/old/newindex/huohongwei/MPI.pdf till page 9/13. I am getting error "No smpd passphrase specified through the registry or .smpd file, exiting" while executing program in MPIEXEC. I am not able to find its solution, Kindly guide me in appropriate direction Regards Priyank

1 0