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

Re: [mpich-discuss] Problems with mpich-3.1.4 and slurm-14.11
by Kenneth Raffenetti 27 May '15

27 May '15

I wonder if your MPICH build was able to detect and build in Slurm support. Can you send the src/pm/hydra/config.log file from your MPICH build directory? Ken On 05/23/2015 08:41 PM, Bill Broadley wrote: > I have an ubuntu 14.04 cluster where slurm-14.11 and OpenMPI 1.8.5 work > well. I'm trying to get mpich-3.1.4 working. An OpenMPI + slurm example: > $ mpicc.openmpi relay.c -o relay > $ mpicc.openmpi hello.c -o relay > > A run on a single node: > $ salloc -N 1 -n 2 mpirun ./relay 1 > salloc: Granted job allocation 682 > c7-13 c7-13 > size= 1, 16384 hops, 2 nodes in 0.01 sec ( 0.37 us/hop) 10468 > KB/sec > > A run on 2 nodes: > $ salloc -N 2 -n 2 mpirun ./relay 1 > salloc: Granted job allocation 683 > c7-13 c7-14 > size= 1, 16384 hops, 2 nodes in 0.10 sec ( 5.95 us/hop) 657 > KB/sec > > I compiled mpich-3.14, ./configure setting only the --prefix. I just > used the ubuntu default gcc-4.8.2. From config.log: > $ ./configure --prefix=/share/apps/mpich-3.1.4/gcc > > Now to test: > $ export PATH=/share/apps/mpich-3.1.4/gcc/bin:$PATH > $ export LD_LIBRARY_PATH=/share/apps/mpich-3.1.4/gcc/lib:$LD_LIBRARY_PATH > $ which mpicc > /share/apps/mpich-3.1.4/gcc/bin/mpicc > $ which mpiexec > /share/apps/mpich-3.1.4/gcc/bin/mpiexec > $ mpicc relay.c -o relay > $ mpicc hello.c -o hello > $ salloc -N 1 -n 2 mpiexec ./hello > salloc: Granted job allocation 688 > Hello world from process 0 of 2 > Hello world from process 1 of 2 > $ salloc -N 2 -n 2 mpiexec ./hello > salloc: Granted job allocation 689 > Hello world from process 0 of 2 > Hello world from process 1 of 2 > > Great, the basics are working. But no actual communications are happening. > > Communications within a single node seems to work: > $ salloc -N 1 -n 2 mpiexec ./relay 1 > salloc: Granted job allocation 690 > c7-13 c7-13 > size= 1, 16384 hops, 2 nodes in 0.01 sec ( 0.55 us/hop) 7074 > KB/sec > > But not between more than one node: > bill@hpc1:~/src/relay$ salloc -N 2 -n 2 mpiexec ./relay 1 > salloc: Granted job allocation 691 > c7-13 Fatal error in MPI_Recv: Unknown error class, error stack: > MPI_Recv(187)...................: MPI_Recv(buf=0x7ffdc56aa630, > count=129, MPI_CHAR, src=1, tag=1, MPI_COMM_WORLD, > status=0x7ffdc56aa610) failed > MPIDI_CH3U_Recvq_FDU_or_AEP(636): Communication error with rank 1 > > =================================================================================== > = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES > = PID 9734 RUNNING AT c7-13 > = EXIT CODE: 1 > = CLEANING UP REMAINING PROCESSES > = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES > =================================================================================== > [proxy:0:1@c7-14] HYD_pmcd_pmip_control_cmd_cb > (pm/pmiserv/pmip_cb.c:885): assert (!closed) failed > [proxy:0:1@c7-14] HYDT_dmxu_poll_wait_for_event > (tools/demux/demux_poll.c:76): callback returned error status > [proxy:0:1@c7-14] main (pm/pmiserv/pmip.c:206): demux engine error > waiting for event > srun: error: c7-14: task 1: Exited with exit code 7 > [mpiexec@hpc1] HYDT_bscu_wait_for_completion > (tools/bootstrap/utils/bscu_wait.c:76): one of the processes terminated > badly; aborting > [mpiexec@hpc1] HYDT_bsci_wait_for_completion > (tools/bootstrap/src/bsci_wait.c:23): launcher returned error waiting > for completion > [mpiexec@hpc1] HYD_pmci_wait_for_completion > (pm/pmiserv/pmiserv_pmci.c:218): launcher returned error waiting for > completion > [mpiexec@hpc1] main (ui/mpich/mpiexec.c:344): process manager error > waiting for completion > salloc: Relinquishing job allocation 691 > > DNS, and ssh seem to be fine: > $ ssh c7-13 "hostname; ssh c7-14 'hostname'" > c7-13 > c7-14 > > Any ideas? > _______________________________________________ > 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] Running HPL on RPi cluster, seems like MPI is somehow not configured properly since it work with 1 node but not more
by Kenneth Raffenetti 27 May '15

27 May '15

Please try a simple test to ensure your mpiexec is working correctly. Something like: mpiexec -machinefile ~/machinefile -n 4 hostname That should output the hostnames of the machines you are attempting to run on. Second, you should ensure that the mpiexec and MPI library your program is linked against come from the same distribution. A common error we see is mixing MPICH's mpiexec with Open MPI's library, or vice versa. Ken On 05/26/2015 03:32 PM, Heerdt, Lanze M. wrote: > I realize this may be a bit off topic, but since what I am doing seems > to be a pretty commonly done thing I am hoping to find someone who has > done it before/can help since I’ve been at my wits end for so long they > are calling me Mr. Whittaker. > > I am trying to run HPL on a Raspberry Pi cluster. I used the following > guides to get to where I am now: > > http://www.tinkernut.com/2014/04/make-cluster-computer/ > > http://www.tinkernut.com/2014/05/make-cluster-computer-part-2/ > > https://www.howtoforge.com/tutorial/hpl-high-performance-linpack-benchmark-… > > and a bit of: > https://www.raspberrypi.org/forums/viewtopic.php?p=301458#p301458 when > the above guide wasn’t working > > basically when I run: “mpiexec -machinefile ~/machinefile -n 1 xhpl” it > works just fine > > but when I run “mpiexec -machinefile ~/machinefile -n 4 xhpl” it errors > with the attached image. (if I use “mpirun…” I get the exact same behavior) > > [Note: I HAVE changed the HPL.dat to have “2 Ps” and “2 Qs” from 1 > and 1 for when I try to run it with 4 processes] > > This is for a project of mine which I need done by the end of the week > so if you see this after 5/29 thank you but don’t bother responding > > I have hpl-2.1, mpi4py-1.3.1, mpich-3.1, and openmpi-1.8.5 at my disposal > > In the machinefile are the 4 IP addresses of my 4 RPi nodes > > 10.15.106.107 > > 10.15.101.29 > > 10.15.106.108 > > 10.15.101.30 > > Any other information you need I can easily get to you so please do not > hesitate to ask. I have nothing else to do but try and get this to work :P > > > > _______________________________________________ > 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

Running HPL on RPi cluster, seems like MPI is somehow not configured properly since it work with 1 node but not more
by Heerdt, Lanze M. 26 May '15

26 May '15

I realize this may be a bit off topic, but since what I am doing seems to be a pretty commonly done thing I am hoping to find someone who has done it before/can help since I've been at my wits end for so long they are calling me Mr. Whittaker. I am trying to run HPL on a Raspberry Pi cluster. I used the following guides to get to where I am now: http://www.tinkernut.com/2014/04/make-cluster-computer/ http://www.tinkernut.com/2014/05/make-cluster-computer-part-2/ https://www.howtoforge.com/tutorial/hpl-high-performance-linpack-benchmark-… and a bit of: https://www.raspberrypi.org/forums/viewtopic.php?p=301458#p301458 when the above guide wasn't working basically when I run: "mpiexec -machinefile ~/machinefile -n 1 xhpl" it works just fine but when I run "mpiexec -machinefile ~/machinefile -n 4 xhpl" it errors with the attached image. (if I use "mpirun..." I get the exact same behavior) [Note: I HAVE changed the HPL.dat to have "2 Ps" and "2 Qs" from 1 and 1 for when I try to run it with 4 processes] This is for a project of mine which I need done by the end of the week so if you see this after 5/29 thank you but don't bother responding I have hpl-2.1, mpi4py-1.3.1, mpich-3.1, and openmpi-1.8.5 at my disposal In the machinefile are the 4 IP addresses of my 4 RPi nodes 10.15.106.107 10.15.101.29 10.15.106.108 10.15.101.30 Any other information you need I can easily get to you so please do not hesitate to ask. I have nothing else to do but try and get this to work :P _______________________________________________ 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] Installation problem
by Eong Sheng Goh 24 May '15

24 May '15

I have found an even more serious problem regarding libmpi.so.12: Unexpected output in exscanf: forrtl: severe (174): SIGSEGV, segmentation fault occurredUnexpected output in exscanf: Image PC Routine Line Source Unexpected output in exscanf: libmpi.so.12 00002ADB339C6F2C Unknown Unknown UnknownUnexpected output in exscanf: libmpi.so.12 00002ADB3394DA84 Unknown Unknown UnknownUnexpected output in exscanf: libmpi.so.12 00002ADB33925B33 Unknown Unknown UnknownUnexpected output in exscanf: libmpi.so.12 00002ADB338E9819 Unknown Unknown UnknownUnexpected output in exscanf: libmpifort.so.12 00002ADB3361D830 Unknown Unknown UnknownUnexpected output in exscanf: exscanf 00000000004087F6 Unknown Unknown UnknownUnexpected output in exscanf: exscanf 0000000000407824 Unknown Unknown UnknownUnexpected output in exscanf: exscanf 00000000004077AC Unknown Unknown UnknownUnexpected output in exscanf: libc.so.6 0000003050E1ED1D Unknown Unknown UnknownUnexpected output in exscanf: exscanf 00000000004076A9 Unknown Unknown UnknownUnexpected output in exscanf: Unexpected output in exscanf: ===================================================================================Unexpected output in exscanf: = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSESUnexpected output in exscanf: = PID 31438 RUNNING AT comsics.usm.myUnexpected output in exscanf: = EXIT CODE: 174Unexpected output in exscanf: = CLEANING UP REMAINING PROCESSESUnexpected output in exscanf: = YOU CAN IGNORE THE BELOW CLEANUP MESSAGESUnexpected output in exscanf: =================================================================================== All the required library path should be included already.I will appreciate any help.thank you. _______________________________________________ discuss mailing list discuss(a)mpich.org To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/discuss

1 0

Problems with mpich-3.1.4 and slurm-14.11
by Bill Broadley 23 May '15

23 May '15

I have an ubuntu 14.04 cluster where slurm-14.11 and OpenMPI 1.8.5 work well. I'm trying to get mpich-3.1.4 working. An OpenMPI + slurm example: $ mpicc.openmpi relay.c -o relay $ mpicc.openmpi hello.c -o relay A run on a single node: $ salloc -N 1 -n 2 mpirun ./relay 1 salloc: Granted job allocation 682 c7-13 c7-13 size= 1, 16384 hops, 2 nodes in 0.01 sec ( 0.37 us/hop) 10468 KB/sec A run on 2 nodes: $ salloc -N 2 -n 2 mpirun ./relay 1 salloc: Granted job allocation 683 c7-13 c7-14 size= 1, 16384 hops, 2 nodes in 0.10 sec ( 5.95 us/hop) 657 KB/sec I compiled mpich-3.14, ./configure setting only the --prefix. I just used the ubuntu default gcc-4.8.2. From config.log: $ ./configure --prefix=/share/apps/mpich-3.1.4/gcc Now to test: $ export PATH=/share/apps/mpich-3.1.4/gcc/bin:$PATH $ export LD_LIBRARY_PATH=/share/apps/mpich-3.1.4/gcc/lib:$LD_LIBRARY_PATH $ which mpicc /share/apps/mpich-3.1.4/gcc/bin/mpicc $ which mpiexec /share/apps/mpich-3.1.4/gcc/bin/mpiexec $ mpicc relay.c -o relay $ mpicc hello.c -o hello $ salloc -N 1 -n 2 mpiexec ./hello salloc: Granted job allocation 688 Hello world from process 0 of 2 Hello world from process 1 of 2 $ salloc -N 2 -n 2 mpiexec ./hello salloc: Granted job allocation 689 Hello world from process 0 of 2 Hello world from process 1 of 2 Great, the basics are working. But no actual communications are happening. Communications within a single node seems to work: $ salloc -N 1 -n 2 mpiexec ./relay 1 salloc: Granted job allocation 690 c7-13 c7-13 size= 1, 16384 hops, 2 nodes in 0.01 sec ( 0.55 us/hop) 7074 KB/sec But not between more than one node: bill@hpc1:~/src/relay$ salloc -N 2 -n 2 mpiexec ./relay 1 salloc: Granted job allocation 691 c7-13 Fatal error in MPI_Recv: Unknown error class, error stack: MPI_Recv(187)...................: MPI_Recv(buf=0x7ffdc56aa630, count=129, MPI_CHAR, src=1, tag=1, MPI_COMM_WORLD, status=0x7ffdc56aa610) failed MPIDI_CH3U_Recvq_FDU_or_AEP(636): Communication error with rank 1 =================================================================================== = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES = PID 9734 RUNNING AT c7-13 = EXIT CODE: 1 = CLEANING UP REMAINING PROCESSES = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES =================================================================================== [proxy:0:1@c7-14] HYD_pmcd_pmip_control_cmd_cb (pm/pmiserv/pmip_cb.c:885): assert (!closed) failed [proxy:0:1@c7-14] HYDT_dmxu_poll_wait_for_event (tools/demux/demux_poll.c:76): callback returned error status [proxy:0:1@c7-14] main (pm/pmiserv/pmip.c:206): demux engine error waiting for event srun: error: c7-14: task 1: Exited with exit code 7 [mpiexec@hpc1] HYDT_bscu_wait_for_completion (tools/bootstrap/utils/bscu_wait.c:76): one of the processes terminated badly; aborting [mpiexec@hpc1] HYDT_bsci_wait_for_completion (tools/bootstrap/src/bsci_wait.c:23): launcher returned error waiting for completion [mpiexec@hpc1] HYD_pmci_wait_for_completion (pm/pmiserv/pmiserv_pmci.c:218): launcher returned error waiting for completion [mpiexec@hpc1] main (ui/mpich/mpiexec.c:344): process manager error waiting for completion salloc: Relinquishing job allocation 691 DNS, and ssh seem to be fine: $ ssh c7-13 "hostname; ssh c7-14 'hostname'" c7-13 c7-14 Any ideas? _______________________________________________ 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] Installation problem
by Eong Sheng Goh 22 May '15

22 May '15

Thanks for the "--with-pbs=/opt/torque" configure option, it is much easier.But I remember that I don't find this option in configure --help, that's why I resort to my previous method. I did run the compilervars.sh script, and /share/apps/intel/Compiler/11.1/072/lib/intel64 is in my LD_LIBRARY_PATH.The same error is still there. > Date: Fri, 22 May 2015 13:14:47 -0500 > From: raffenet(a)mcs.anl.gov > To: discuss(a)mpich.org > Subject: Re: [mpich-discuss] Installation problem > > Hi, > > In response to question 1, I would suggest using the > "--with-pbs=/opt/torque" configure option to point to your installation > and ensure all flags are set appropriately. > > For question 2, have you run the compilervars.sh script from your Intel > compiler installation? Should be in > /share/apps/intel/Compiler/11.1/072/bin. That script should add any > needed library paths to your environment, and avoid the error you are > seeing. > > Ken > > On 05/22/2015 10:50 AM, Eong Sheng Goh wrote: > > Hi, > > > > I am installing mpich 3.1.4 and the following the configure script I use: > > > > CC=/share/apps/intel/Compiler/11.0/081/bin/intel64/icc > > CXX=/share/apps/intel/Compiler/11.0/081/bin/intel64/icpc > > F77=/share/apps/intel/Compiler/11.1/072/bin/intel64/ifort > > FC=/share/apps/intel/Compiler/11.1/072/bin/intel64/ifort > > CPPFLAGS=-I/opt/torque/include LDFLAGS="-L/opt/torque/lib > > -L/opt/torque/lib64" ../mpich-3.1.4/configure > > --prefix=/share/apps/estools/mpich2/3.1.4/intel --enable-cxx > > --enable-fortran=all --enable-threads=multiple > > --with-thread-package=posix --enable-romio --with-pm=hydra > > --enable-shared --enable-static > > > > However I have two inquiries: > > > > 1. I am installing it on a cluster with Torque PBS system and wish bind > > mpich against Torque. The configure output states that: > > checking tm.h usability... yes > > checking tm.h presence... yes > > checking for tm.h... yes > > checking for tm_init in -ltorque... yes > > But I do this by adding "CPPFLAGS=-I/opt/torque/include ", but do I > > have to repeat it for other flags (CFLAGS, CXXFLAGS, FFLAGS, > > FCFLAGS ...) ? > > > > 2. After I perform make and make install, I proceed to make testing, > > but with a lot of this error: > > > > /share/apps/intel/Compiler/11.1/072/lib/intel64/libifport.so.5: > > undefined reference to `__intel_sse2_strlen' > > By some googling I found some hints that this can be resolved by > > adding the library libifcoremt.so, but when I add it to the LIBS > > parameter the configure script cannot even start. So what is the > > correct way to do this? Or is there any other solution? > > > > > > Thank you. > > > > > > _______________________________________________ > > 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

Re: [mpich-discuss] Installation problem
by Kenneth Raffenetti 22 May '15

22 May '15

Hi, In response to question 1, I would suggest using the "--with-pbs=/opt/torque" configure option to point to your installation and ensure all flags are set appropriately. For question 2, have you run the compilervars.sh script from your Intel compiler installation? Should be in /share/apps/intel/Compiler/11.1/072/bin. That script should add any needed library paths to your environment, and avoid the error you are seeing. Ken On 05/22/2015 10:50 AM, Eong Sheng Goh wrote: > Hi, > > I am installing mpich 3.1.4 and the following the configure script I use: > > CC=/share/apps/intel/Compiler/11.0/081/bin/intel64/icc > CXX=/share/apps/intel/Compiler/11.0/081/bin/intel64/icpc > F77=/share/apps/intel/Compiler/11.1/072/bin/intel64/ifort > FC=/share/apps/intel/Compiler/11.1/072/bin/intel64/ifort > CPPFLAGS=-I/opt/torque/include LDFLAGS="-L/opt/torque/lib > -L/opt/torque/lib64" ../mpich-3.1.4/configure > --prefix=/share/apps/estools/mpich2/3.1.4/intel --enable-cxx > --enable-fortran=all --enable-threads=multiple > --with-thread-package=posix --enable-romio --with-pm=hydra > --enable-shared --enable-static > > However I have two inquiries: > > 1. I am installing it on a cluster with Torque PBS system and wish bind > mpich against Torque. The configure output states that: > checking tm.h usability... yes > checking tm.h presence... yes > checking for tm.h... yes > checking for tm_init in -ltorque... yes > But I do this by adding "CPPFLAGS=-I/opt/torque/include ", but do I > have to repeat it for other flags (CFLAGS, CXXFLAGS, FFLAGS, > FCFLAGS ...) ? > > 2. After I perform make and make install, I proceed to make testing, > but with a lot of this error: > > /share/apps/intel/Compiler/11.1/072/lib/intel64/libifport.so.5: > undefined reference to `__intel_sse2_strlen' > By some googling I found some hints that this can be resolved by > adding the library libifcoremt.so, but when I add it to the LIBS > parameter the configure script cannot even start. So what is the > correct way to do this? Or is there any other solution? > > > Thank you. > > > _______________________________________________ > 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

Installation problem
by Eong Sheng Goh 22 May '15

22 May '15

Hi, I am installing mpich 3.1.4 and the following the configure script I use: CC=/share/apps/intel/Compiler/11.0/081/bin/intel64/icc CXX=/share/apps/intel/Compiler/11.0/081/bin/intel64/icpc F77=/share/apps/intel/Compiler/11.1/072/bin/intel64/ifort FC=/share/apps/intel/Compiler/11.1/072/bin/intel64/ifort CPPFLAGS=-I/opt/torque/include LDFLAGS="-L/opt/torque/lib -L/opt/torque/lib64" ../mpich-3.1.4/configure --prefix=/share/apps/estools/mpich2/3.1.4/intel --enable-cxx --enable-fortran=all --enable-threads=multiple --with-thread-package=posix --enable-romio --with-pm=hydra --enable-shared --enable-static However I have two inquiries: I am installing it on a cluster with Torque PBS system and wish bind mpich against Torque. The configure output states that: checking tm.h usability... yes checking tm.h presence... yes checking for tm.h... yes checking for tm_init in -ltorque... yes But I do this by adding "CPPFLAGS=-I/opt/torque/include ", but do I have to repeat it for other flags (CFLAGS, CXXFLAGS, FFLAGS, FCFLAGS ...) ? After I perform make and make install, I proceed to make testing, but with a lot of this error: /share/apps/intel/Compiler/11.1/072/lib/intel64/libifport.so.5: undefined reference to `__intel_sse2_strlen' By some googling I found some hints that this can be resolved by adding the library libifcoremt.so, but when I add it to the LIBS parameter the configure script cannot even start. So what is the correct way to do this? Or is there any other solution? Thank you. _______________________________________________ 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] Error while using MPI platform with CPP program
by Umesh Shisode 21 May '15

21 May '15

Hello Sir, Sorry for your inconvenience. I will ask this on forum. Thanks for all your kind help. Regards, Umesh Sharad Shisode, M.Tech in Modeling and Simulation(2013-15), Centre for Modeling and Simulation, University of Pune. On Thu, May 21, 2015 at 6:55 PM, Kenneth Raffenetti <raffenet(a)mcs.anl.gov> wrote: > Hi, > > Please do not use this list for general programming questions. Either take > this discussion offline, or move it to a more appropriate forum. This list > is for questions about MPICH. > > Thanks, > Ken > > On 05/21/2015 06:54 AM, Umesh Shisode wrote: > >> Hello Antoine Rougier, >> >> I did not understand, [You have to change your code if you want to >> keep using blocking communication.] you ask me about. Also now the >> code is working fine, but it is only for same process and same numbers >> of rows. For other cases it is giving me error (segmentation fault). >> Please check it, I am attaching here the tar file of compile code. But >> it is working for only following commands >> >> >> mpic++ main.cpp -o result >> >> mpirun -np 4 result >> >> But for this case input matrix is (4 X 4) only. >> >> If you find any mistake please let me know. >> >> >> >> Regards, >> >> Umesh Sharad Shisode, >> M.Tech in Modeling and Simulation(2013-15), >> Centre for Modeling and Simulation, >> University of Pune. >> >> On Wed, May 20, 2015 at 11:29 PM, Antoine Rougier >> <rougier(a)par.tuwien.ac.at <mailto:[email protected]>> wrote: >> >> First, as far as I understand, the broadcast should not be between >> blocking send on a process and blocking recv on the others. You have >> to >> change your code if you want to keep using blocking communication. >> >> Then, by running (with the previous modifications), I have got: >> Fatal error in MPI_Recv: Invalid buffer pointer, error stack: >> MPI_Recv(187): MPI_Recv(buf=(nil), count=8, MPI_DOUBLE, src=0, tag=3, >> MPI_COMM_WORLD, status=0x7fffbc4e02d0) failed >> MPI_Recv(126): Null buffer pointer >> >> It is probably an issue on your side, such as bad index. >> >> Antoine Rougier >> >> On 05/20/2015 05:17 PM, Umesh Shisode wrote: >> > Hello Antoine Rougier, >> > >> > Thanks for your reply. Now I have code which can broadcast vector >> as well >> > as Matrix object (Attached here). But it is giving error when I >> tried to >> > send and receive part of matrix object it is giving me error. >> Thanks again >> > for your time. Please help me out. As it is not giving me correct >> answer. I >> > am waiting for your reply. >> > Sorry for your inconvenience. >> > >> > I have attached here new code. It compiles when I want to >> broadcast. But >> > giving error for send and receive. I also attached error in png >> file. >> > >> > Regards, >> > >> > Umesh Sharad Shisode, >> > M.Tech in Modeling and Simulation(2013-15), >> > Centre for Modeling and Simulation, >> > University of Pune. >> > >> > On Wed, May 20, 2015 at 4:07 PM, Antoine Rougier >> <rougier(a)par.tuwien.ac.at <mailto:[email protected]>> >> >> > wrote: >> > >> >> Hello, >> >> >> >> I think your issue comes from b.vect=NULL (according to the used >> >> constructor) which is not a checked value in T >> Vector<T>::get(int i). By >> >> the way Ncol is no initialized before using it in MPI_Bcast. >> >> >> >> best regards, >> >> >> >> Antoine Rougier >> >> >> >> >> >> On 05/20/2015 11:55 AM, Umesh Shisode wrote: >> >>> Hello Junchao, >> >>> >> >>> I initialize vector b, but even then it is not working for >> broadcasting. >> >> It >> >>> gives same error. As the vector class is user defined and also >> I use >> >>> dynamic memory allocation. This may be the reason for failing. >> If you >> >> know >> >>> about it then please tell. >> >>> I am here attaching the zip file which contains only the vector >> class and >> >>> main file as well as error I am facing. Please have a look at >> it. Thanks >> >>> for your time. I am waiting for your reply. >> >>> >> >>> >> >>> Regards, >> >>> >> >>> Umesh Sharad Shisode, >> >>> M.Tech in Modeling and Simulation(2013-15), >> >>> Centre for Modeling and Simulation, >> >>> University of Pune. >> >>> >> >>> On Tue, May 19, 2015 at 10:57 PM, Junchao Zhang >> <jczhang(a)mcs.anl.gov <mailto:[email protected]>> >> >> wrote: >> >>> >> >>>> Your code "MPI_Bcast(&b(0), b.get_row(), MPI_DOUBLE, 1, >> MPI_COMM_WORLD)" >> >>>> is OK. You can find manual of MPI_Bcast at >> >>>> http://mpi.deino.net/mpi_functions/MPI_Bcast.html >> >>>> Don't forget to initialize the vector b, in other words, >> allocate memory >> >>>> for it, on ALL ranks. >> >>>> >> >>>> >> >>>> --Junchao Zhang >> >>>> >> >>>> On Tue, May 19, 2015 at 9:27 AM, Umesh Shisode >> <ushisode1(a)gmail.com <mailto:[email protected]>> >> >>>> wrote: >> >>>> >> >>>>> Hi Junchao, >> >>>>> >> >>>>> I sort out the problem for MPI_Send() and MPI_Recv(), as I >> requested in >> >>>>> last mail. Now I have one more difficulty which is how can I >> broadcast >> >>>>> class object using MPI_Bcast() function. In my case, class is >> Vector >> >> (you >> >>>>> can check it at vect_fun.h header file). I want to broadcast >> it to all >> >>>>> processes. I have used dynamic memory allocation in the code. >> Please >> >> help >> >>>>> me to solved the issue. I am waiting for your reply. Thanks. >> >>>>> >> >>>>> >> >>>>> Regards, >> >>>>> >> >>>>> Umesh Sharad Shisode, >> >>>>> M.Tech in Modeling and Simulation(2013-15), >> >>>>> Centre for Modeling and Simulation, >> >>>>> University of Pune. >> >>>>> >> >>>>> On Tue, May 19, 2015 at 7:51 AM, Umesh Shisode >> <ushisode1(a)gmail.com <mailto:[email protected]>> >> >>>>> wrote: >> >>>>> >> >>>>>> Sorry Junchao, >> >>>>>> >> >>>>>> Thanks for your reply. But my problem is with MPICH only. My >> question >> >> is >> >>>>>> how to send part of class object (in my case matrix object) >> from one >> >>>>>> process to another process using MPI_Send and receive using >> MPI_Recv >> >>>>>> functions. I tried a lot of way but couldn't do it. I hope >> you will >> >> help >> >>>>>> me. Thanks in advance. I have write a function to get the >> pointer >> >> address >> >>>>>> for first parameter of MPI_Send() function. Waiting for your >> reply. >> >>>>>> >> >>>>>> Regards, >> >>>>>> >> >>>>>> Umesh Sharad Shisode, >> >>>>>> M.Tech in Modeling and Simulation(2013-15), >> >>>>>> Centre for Modeling and Simulation, >> >>>>>> University of Pune. >> >>>>>> >> >>>>>> On Mon, May 18, 2015 at 7:31 PM, Junchao Zhang >> <jczhang(a)mcs.anl.gov <mailto:[email protected]>> >> >>>>>> wrote: >> >>>>>> >> >>>>>>> Could not reproduce it with the latest MPICH. If the >> problem is not >> >>>>>>> specific to MPICH, you'd better move your question to other >> forums >> >> like >> >>>>>>> stackoverflow. >> >>>>>>> >> >>>>>>> $mpirun -n 2 ./test >> >>>>>>> Matrix A : >> >>>>>>> >> >>>>>>> 1 2 3 4 >> >>>>>>> 1 2 3 4 >> >>>>>>> 1 2 3 4 >> >>>>>>> 1 2 3 4 >> >>>>>>> >> >>>>>>> Vector b before broadcast: >> >>>>>>> 1.000000, 2.000000, 3.000000, 4.000000 >> >>>>>>> Matrix A column : 4 >> >>>>>>> after sending: lLim = 2 uLim = 4 for rank : 1 >> >>>>>>> count = 8 >> >>>>>>> >> >>>>>>> Matrix A column : 0for rank = 1 >> >>>>>>> after receive: lLim = 2 uLim = 4 for rank : 1 >> >>>>>>> Ncol after broadcasting = 4 >> >>>>>>> >> >>>>>>> >> >>>>>>> --Junchao Zhang >> >>>>>>> >> >>>>>>> On Mon, May 18, 2015 at 6:33 AM, Umesh Shisode >> <ushisode1(a)gmail.com <mailto:[email protected]>> >> >> >>>>>>> wrote: >> >>>>>>> >> >>>>>>>> Hello Sir / Madam, >> >>>>>>>> >> >>>>>>>> This is Umesh Sharad Shisode, pursuing Masters in Modeling >> and >> >>>>>>>> Simulation, at Centre for Modeling and Simulation, Pune >> University. >> >> I am in >> >>>>>>>> the final stage of my course. My Masters Thesis title is : >> >>>>>>>> >> >>>>>>>> Development of Conjugate Gradient solver in OpenFOAM / >> CPP using >> >>>>>>>> MPI >> >>>>>>>> >> >>>>>>>> I am now doing MPI part of my project and facing some >> problems with >> >>>>>>>> MPI functions syntax. I am trying to use CPP class object >> with MPI >> >>>>>>>> functions but it is giving some error. Which I don't >> understand. >> >>>>>>>> >> >>>>>>>> Operating system : LINUX >> >>>>>>>> Programming Language : CPP >> >>>>>>>> >> >>>>>>>> Commands I am using are : >> >>>>>>>> >> >>>>>>>> > mpic++ main.cpp -o result >> >>>>>>>> > mpirun -np 2 result >> >>>>>>>> >> >>>>>>>> Please help me to sort out this issue . I am attaching >> herewith zip >> >>>>>>>> file containing the code which I want to parallelize with >> the png >> >> file >> >>>>>>>> containing error. Waiting for your reply. Thanks in advance. >> >>>>>>>> >> >>>>>>>> Regards, >> >>>>>>>> >> >>>>>>>> Umesh Sharad Shisode, >> >>>>>>>> M.Tech in Modeling and Simulation(2013-15), >> >>>>>>>> Centre for Modeling and Simulation, >> >>>>>>>> University of Pune. >> >>>>>>>> >> >>>>>>>> >> >>>>>>>> _______________________________________________ >> >>>>>>>> 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 >> >>>>>>> >> >>>>>> >> >>>>>> >> >>>>> >> >>>>> _______________________________________________ >> >>>>> discuss mailing list discuss(a)mpich.org <mailto: >> discuss(a)mpich.org> >> >>>>> To manage subscription options or unsubscribe: >> >>>>> https://lists.mpich.org/mailman/listinfo/discuss >> >>>>> >> >>>> >> >>>> >> >>>> _______________________________________________ >> >>>> discuss mailing list discuss(a)mpich.org <mailto: >> discuss(a)mpich.org> >> >>>> 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 >> >>> >> >> >> >> >> >> _______________________________________________ >> >> 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 >> > >> >> >> _______________________________________________ >> 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 >> >> _______________________________________________ > 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] Error while using MPI platform with CPP program
by Kenneth Raffenetti 21 May '15

21 May '15

Hi, Please do not use this list for general programming questions. Either take this discussion offline, or move it to a more appropriate forum. This list is for questions about MPICH. Thanks, Ken On 05/21/2015 06:54 AM, Umesh Shisode wrote: > Hello Antoine Rougier, > > I did not understand, [You have to change your code if you want to > keep using blocking communication.] you ask me about. Also now the > code is working fine, but it is only for same process and same numbers > of rows. For other cases it is giving me error (segmentation fault). > Please check it, I am attaching here the tar file of compile code. But > it is working for only following commands > > >> mpic++ main.cpp -o result > >> mpirun -np 4 result > > But for this case input matrix is (4 X 4) only. > > If you find any mistake please let me know. > > > > Regards, > > Umesh Sharad Shisode, > M.Tech in Modeling and Simulation(2013-15), > Centre for Modeling and Simulation, > University of Pune. > > On Wed, May 20, 2015 at 11:29 PM, Antoine Rougier > <rougier(a)par.tuwien.ac.at <mailto:[email protected]>> wrote: > > First, as far as I understand, the broadcast should not be between > blocking send on a process and blocking recv on the others. You have to > change your code if you want to keep using blocking communication. > > Then, by running (with the previous modifications), I have got: > Fatal error in MPI_Recv: Invalid buffer pointer, error stack: > MPI_Recv(187): MPI_Recv(buf=(nil), count=8, MPI_DOUBLE, src=0, tag=3, > MPI_COMM_WORLD, status=0x7fffbc4e02d0) failed > MPI_Recv(126): Null buffer pointer > > It is probably an issue on your side, such as bad index. > > Antoine Rougier > > On 05/20/2015 05:17 PM, Umesh Shisode wrote: > > Hello Antoine Rougier, > > > > Thanks for your reply. Now I have code which can broadcast vector > as well > > as Matrix object (Attached here). But it is giving error when I > tried to > > send and receive part of matrix object it is giving me error. > Thanks again > > for your time. Please help me out. As it is not giving me correct > answer. I > > am waiting for your reply. > > Sorry for your inconvenience. > > > > I have attached here new code. It compiles when I want to > broadcast. But > > giving error for send and receive. I also attached error in png file. > > > > Regards, > > > > Umesh Sharad Shisode, > > M.Tech in Modeling and Simulation(2013-15), > > Centre for Modeling and Simulation, > > University of Pune. > > > > On Wed, May 20, 2015 at 4:07 PM, Antoine Rougier > <rougier(a)par.tuwien.ac.at <mailto:[email protected]>> > > wrote: > > > >> Hello, > >> > >> I think your issue comes from b.vect=NULL (according to the used > >> constructor) which is not a checked value in T > Vector<T>::get(int i). By > >> the way Ncol is no initialized before using it in MPI_Bcast. > >> > >> best regards, > >> > >> Antoine Rougier > >> > >> > >> On 05/20/2015 11:55 AM, Umesh Shisode wrote: > >>> Hello Junchao, > >>> > >>> I initialize vector b, but even then it is not working for > broadcasting. > >> It > >>> gives same error. As the vector class is user defined and also > I use > >>> dynamic memory allocation. This may be the reason for failing. > If you > >> know > >>> about it then please tell. > >>> I am here attaching the zip file which contains only the vector > class and > >>> main file as well as error I am facing. Please have a look at > it. Thanks > >>> for your time. I am waiting for your reply. > >>> > >>> > >>> Regards, > >>> > >>> Umesh Sharad Shisode, > >>> M.Tech in Modeling and Simulation(2013-15), > >>> Centre for Modeling and Simulation, > >>> University of Pune. > >>> > >>> On Tue, May 19, 2015 at 10:57 PM, Junchao Zhang > <jczhang(a)mcs.anl.gov <mailto:[email protected]>> > >> wrote: > >>> > >>>> Your code "MPI_Bcast(&b(0), b.get_row(), MPI_DOUBLE, 1, > MPI_COMM_WORLD)" > >>>> is OK. You can find manual of MPI_Bcast at > >>>> http://mpi.deino.net/mpi_functions/MPI_Bcast.html > >>>> Don't forget to initialize the vector b, in other words, > allocate memory > >>>> for it, on ALL ranks. > >>>> > >>>> > >>>> --Junchao Zhang > >>>> > >>>> On Tue, May 19, 2015 at 9:27 AM, Umesh Shisode > <ushisode1(a)gmail.com <mailto:[email protected]>> > >>>> wrote: > >>>> > >>>>> Hi Junchao, > >>>>> > >>>>> I sort out the problem for MPI_Send() and MPI_Recv(), as I > requested in > >>>>> last mail. Now I have one more difficulty which is how can I > broadcast > >>>>> class object using MPI_Bcast() function. In my case, class is > Vector > >> (you > >>>>> can check it at vect_fun.h header file). I want to broadcast > it to all > >>>>> processes. I have used dynamic memory allocation in the code. > Please > >> help > >>>>> me to solved the issue. I am waiting for your reply. Thanks. > >>>>> > >>>>> > >>>>> Regards, > >>>>> > >>>>> Umesh Sharad Shisode, > >>>>> M.Tech in Modeling and Simulation(2013-15), > >>>>> Centre for Modeling and Simulation, > >>>>> University of Pune. > >>>>> > >>>>> On Tue, May 19, 2015 at 7:51 AM, Umesh Shisode > <ushisode1(a)gmail.com <mailto:[email protected]>> > >>>>> wrote: > >>>>> > >>>>>> Sorry Junchao, > >>>>>> > >>>>>> Thanks for your reply. But my problem is with MPICH only. My > question > >> is > >>>>>> how to send part of class object (in my case matrix object) > from one > >>>>>> process to another process using MPI_Send and receive using > MPI_Recv > >>>>>> functions. I tried a lot of way but couldn't do it. I hope > you will > >> help > >>>>>> me. Thanks in advance. I have write a function to get the > pointer > >> address > >>>>>> for first parameter of MPI_Send() function. Waiting for your > reply. > >>>>>> > >>>>>> Regards, > >>>>>> > >>>>>> Umesh Sharad Shisode, > >>>>>> M.Tech in Modeling and Simulation(2013-15), > >>>>>> Centre for Modeling and Simulation, > >>>>>> University of Pune. > >>>>>> > >>>>>> On Mon, May 18, 2015 at 7:31 PM, Junchao Zhang > <jczhang(a)mcs.anl.gov <mailto:[email protected]>> > >>>>>> wrote: > >>>>>> > >>>>>>> Could not reproduce it with the latest MPICH. If the > problem is not > >>>>>>> specific to MPICH, you'd better move your question to other > forums > >> like > >>>>>>> stackoverflow. > >>>>>>> > >>>>>>> $mpirun -n 2 ./test > >>>>>>> Matrix A : > >>>>>>> > >>>>>>> 1 2 3 4 > >>>>>>> 1 2 3 4 > >>>>>>> 1 2 3 4 > >>>>>>> 1 2 3 4 > >>>>>>> > >>>>>>> Vector b before broadcast: > >>>>>>> 1.000000, 2.000000, 3.000000, 4.000000 > >>>>>>> Matrix A column : 4 > >>>>>>> after sending: lLim = 2 uLim = 4 for rank : 1 > >>>>>>> count = 8 > >>>>>>> > >>>>>>> Matrix A column : 0for rank = 1 > >>>>>>> after receive: lLim = 2 uLim = 4 for rank : 1 > >>>>>>> Ncol after broadcasting = 4 > >>>>>>> > >>>>>>> > >>>>>>> --Junchao Zhang > >>>>>>> > >>>>>>> On Mon, May 18, 2015 at 6:33 AM, Umesh Shisode > <ushisode1(a)gmail.com <mailto:[email protected]>> > >>>>>>> wrote: > >>>>>>> > >>>>>>>> Hello Sir / Madam, > >>>>>>>> > >>>>>>>> This is Umesh Sharad Shisode, pursuing Masters in Modeling and > >>>>>>>> Simulation, at Centre for Modeling and Simulation, Pune > University. > >> I am in > >>>>>>>> the final stage of my course. My Masters Thesis title is : > >>>>>>>> > >>>>>>>> Development of Conjugate Gradient solver in OpenFOAM / > CPP using > >>>>>>>> MPI > >>>>>>>> > >>>>>>>> I am now doing MPI part of my project and facing some > problems with > >>>>>>>> MPI functions syntax. I am trying to use CPP class object > with MPI > >>>>>>>> functions but it is giving some error. Which I don't > understand. > >>>>>>>> > >>>>>>>> Operating system : LINUX > >>>>>>>> Programming Language : CPP > >>>>>>>> > >>>>>>>> Commands I am using are : > >>>>>>>> > >>>>>>>> > mpic++ main.cpp -o result > >>>>>>>> > mpirun -np 2 result > >>>>>>>> > >>>>>>>> Please help me to sort out this issue . I am attaching > herewith zip > >>>>>>>> file containing the code which I want to parallelize with > the png > >> file > >>>>>>>> containing error. Waiting for your reply. Thanks in advance. > >>>>>>>> > >>>>>>>> Regards, > >>>>>>>> > >>>>>>>> Umesh Sharad Shisode, > >>>>>>>> M.Tech in Modeling and Simulation(2013-15), > >>>>>>>> Centre for Modeling and Simulation, > >>>>>>>> University of Pune. > >>>>>>>> > >>>>>>>> > >>>>>>>> _______________________________________________ > >>>>>>>> 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 > >>>>>>> > >>>>>> > >>>>>> > >>>>> > >>>>> _______________________________________________ > >>>>> 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 > >>>> > >>> > >>> > >>> > >>> _______________________________________________ > >>> 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 > >> > > > > > > > > _______________________________________________ > > 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 > > > > > _______________________________________________ > 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