December 2012 - discuss - mailman3.cels.anl.gov

Re: [mpich-discuss] Why does MPI_Init() get NULL argc and argv when running Fortran programs?
by Rajeev Thakur 03 Dec '12

03 Dec '12

The Fortran binding for MPI_Init is MPI_INIT(IERROR) There are no command line arguments. Rajeev On Dec 3, 2012, at 6:22 AM, clara wrote: > Hi, > > I added a print statement at the beginning of MPI_Init() in mpich2-1.4 in order to check the arguments of MPI_Init(). When running a Fortran program, e.g. ep.S.1, the arguments of MPI_Init() are NULL; however, when running a C program, e.g. is.S.1, *argc of MPI_Init() is 1 and (*argv)[0] points to the name of the executed file is.S.1. > Can anyone tell us why there is difference in dealing with arguments of MPI_Init() between C and Fortran programs? How could arguments be transferred to MPI_Init() for Fortran programs in mpich2? Thanks. > > Best, > Clara > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss

1 0

Why does MPI_Init() get NULL argc and argv when running Fortran programs?
by clara 03 Dec '12

03 Dec '12

Hi, I added a print statement at the beginning of MPI_Init() in mpich2-1.4 in order to check the arguments of MPI_Init(). When running a Fortran program, e.g. ep.S.1, the arguments of MPI_Init() are NULL; however, when running a C program, e.g. is.S.1, *argc of MPI_Init() is 1 and (*argv)[0] points to the name of the executed file is.S.1. Can anyone tell us why there is difference in dealing with arguments of MPI_Init() between C and Fortran programs? How could arguments be transferred to MPI_Init() for Fortran programs in mpich2? Thanks. Best, Clara

1 0

Why does MPI_Init() get NULL argc and argv when running Fortran programs?
by clara 03 Dec '12

03 Dec '12

Hi, I added a printf statement at the beginning of MPI_Init() in mpich2-1.4 in order to check the arguments of MPI_Init(). When running a Fortran program, e.g. ep.S.1, the arguments of MPI_Init() are NULL; however, when running a C program, e.g. is.S.1, *argc of MPI_Init() is 1 and (*argv)[0] points to the name of the executed file is.S.1. Can anyone tell us why there is difference in dealing with arguments of MPI_Init() between C and Fortran programs? How could arguments be transferred to MPI_Init() for Fortran programs in mpich2? Thanks. Best, Clara

1 0

Re: [mpich-discuss] send and recieve not working over ethernet
by D Griffiths 01 Dec '12

01 Dec '12

Okay, I'll try reinstalling a 64-bit OS. Thanks for your time.

1 0

Re: [mpich-discuss] Support for MIC in mpich2-1.5
by Pavan Balaji 01 Dec '12

01 Dec '12

Sorry for the delay in responding, John. I'm catching up on my email. Looks like some bug in the scif code. I tried to reproduce it on my machine, but I can't. Can you give a little more information on the setup? -- Pavan On 11/30/2012 04:36 PM US Central Time, John Fettig wrote: > Any thoughts about this? > > Regards, > John > > > On Tue, Nov 13, 2012 at 5:07 PM, John Fettig <john.fettig(a)gmail.com > <mailto:[email protected]>> wrote: > > On Mon, Nov 5, 2012 at 9:37 PM, Pavan Balaji <balaji(a)mcs.anl.gov > <mailto:[email protected]>> wrote: > > > On 11/05/12 13:12, John Fettig wrote: > > I believe I have a working build, I'll append my cross file > to the end > of this email if anybody else wants to try it. > > > Thanks! > > > I have a followup question: is there any support for > launching jobs > that use both the MIC and the host CPU? > > > Yes. Once you have setup MPICH on both the host and MIC, you > can launch jobs across them. > > If you didn't pass any configure option, it'll use TCP/IP, which > is very slow. If you configure with > --with-device=ch3:nemesis:scif, it'll use the SCIF protocol, > which is much faster. > > > I compiled examples/hellow.c for both the MIC and the host CPU, and > copied it to the card. This seems to work: > > $ mpiexec -hosts 172.31.1.1:1 <http://172.31.1.1:1>,172.31.1.254:1 > <http://172.31.1.254:1> -n 1 ./hellow.mic : -n 1 ./hellow > Hello world from process 1 of 2 > Hello world from process 0 of 2 > > However, if I try to run more processes it crashes: > > $ mpiexec -hosts 172.31.1.1:3 <http://172.31.1.1:3>,172.31.1.254:3 > <http://172.31.1.254:3> -n 3 ./hellow.mic : -n 3 ./hellow > Hello world from process 4 of 6 > Hello world from process 0 of 6 > Hello world from process 3 of 6 > Hello world from process 1 of 6 > 0: 3: 00000033: 00000042: readv err 0 > Fatal error in MPI_Finalize: Other MPI error, error stack: > MPI_Finalize(293).................: MPI_Finalize failed > MPI_Finalize(213).................: > MPID_Finalize(117)................: > MPIDI_CH3U_VC_WaitForClose(385)...: an error occurred while the > device was waiting for all open connections to close > MPIDI_CH3I_Progress(367)..........: > MPID_nem_mpich2_blocking_recv(904): > state_commrdy_handler(175)........: > state_commrdy_handler(138)........: > MPID_nem_scif_recv_handler(115)...: Communication error with rank 3 > MPID_nem_scif_recv_handler(35)....: scif_scif_read failed > (scif_scif_read failed with error 'Success') > 1: 3: 00000033: 00000042: readv err 0 > Fatal error in MPI_Finalize: Other MPI error, error stack: > MPI_Finalize(293).................: MPI_Finalize failed > MPI_Finalize(213).................: > MPID_Finalize(117)................: > MPIDI_CH3U_VC_WaitForClose(385)...: an error occurred while the > device was waiting for all open connections to close > MPIDI_CH3I_Progress(367)..........: > MPID_nem_mpich2_blocking_recv(904): > state_commrdy_handler(175)........: > state_commrdy_handler(138)........: > MPID_nem_scif_recv_handler(115)...: Communication error with rank 3 > MPID_nem_scif_recv_handler(35)....: scif_scif_read failed > (scif_scif_read failed with error 'Success') > Hello world from process 5 of 6 > Fatal error in MPI_Finalize: Other MPI error, error stack: > MPI_Finalize(293).................: MPI_Finalize failed > MPI_Finalize(213).................: > MPID_Finalize(117)................: > MPIDI_CH3U_VC_WaitForClose(385)...: an error occurred while the > device was waiting for all open connections to close > MPIDI_CH3I_Progress(367)..........: > MPID_nem_mpich2_blocking_recv(904): > state_commrdy_handler(184)........: poll of socket fds failed > Fatal error in MPI_Finalize: Other MPI error, error stack: > MPI_Finalize(293).................: MPI_Finalize failed > MPI_Finalize(213).................: > MPID_Finalize(117)................: > MPIDI_CH3U_VC_WaitForClose(385)...: an error occurred while the > device was waiting for all open connections to close > MPIDI_CH3I_Progress(367)..........: > MPID_nem_mpich2_blocking_recv(904): > state_commrdy_handler(184)........: poll of socket fds failed > > =================================================================================== > = BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES > = EXIT CODE: 1 > = CLEANING UP REMAINING PROCESSES > = YOU CAN IGNORE THE BELOW CLEANUP MESSAGES > =================================================================================== > [proxy:0:[email protected]] HYD_pmcd_pmip_control_cmd_cb > (./pm/pmiserv/pmip_cb.c:883): assert (!closed) failed > [proxy:0:[email protected]] HYDT_dmxu_poll_wait_for_event > (./tools/demux/demux_poll.c:77): callback returned error status > [proxy:0:[email protected]] main (./pm/pmiserv/pmip.c:210): demux engine > error waiting for event > [mpiexec@host] HYDT_bscu_wait_for_completion > (./tools/bootstrap/utils/bscu_wait.c:76): one of the processes > terminated badly; aborting > [mpiexec@host] HYDT_bsci_wait_for_completion > (./tools/bootstrap/src/bsci_wait.c:23): launcher returned error > waiting for completion > [mpiexec@host] HYD_pmci_wait_for_completion > (./pm/pmiserv/pmiserv_pmci.c:216): launcher returned error waiting > for completion > [mpiexec@host] main (./ui/mpich/mpiexec.c:325): process manager > error waiting for completion > > Any ideas? > > John > > -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] send and recieve not working over ethernet
by Pavan Balaji 01 Dec '12

01 Dec '12

Please keep the discuss@mpich mailing list cc'ed. Yes, this is a problem. Is one of your machines actually a 32-bit machine or did you just install a 32-bit OS on a 64-bit machine? If it's the latter, you could reinstall a 64-bit OS. If you can't do that, another option is to build MPICH in 32-bit mode on both machines by passing CFLAGS=-m32 to configure on the 64-bit machine. This will require that you have the right 32-bit libraries installed on your 64-bit machine (such as the 32-bit libc library). This option is typically for expert users, and is not a recommended option for most users. -- Pavan On 12/01/2012 09:34 AM US Central Time, D Griffiths wrote: > x86_64 for the one, i686 for the other. I take it this is the problem? > > Thanks for getting back to me so quickly. -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

Re: [mpich-discuss] send and recieve not working over ethernet
by Pavan Balaji 01 Dec '12

01 Dec '12

Are the architectures of the two machines identical? Both x86 64-bit? Did you install a 64-bit OS on both machines ("uname -m" == "x86_64")? -- Pavan On 12/01/2012 08:34 AM US Central Time, D Griffiths wrote: > I am completely new to MPI so I hope Iï¿½m not asking anything too silly. > I have two computers, one running Ubuntu and the other running > Scientific Linux. They recognise each other as hosts, I can ssh from > each without a password and have disabled any firewalls. > > Although I can run a simple Hello World application, the trouble starts > when I try to use MPI_SEND and MPI_RECV. I am trying a simple program to > send an int variable from one node to the other (see source code below). > > If I run the code with two processes on the same computer I get the > correct output: > > dan@siriusa:~/mpich2/tutorial2$ mpiexec -n 2 ./result2 > > Process 0 sending number 4 > Process 1 received number 4 from process 0 > > However, if I try to run the code using both computers, I get: > > dan@siriusa:~/mpich2/tutorial2$ mpiexec -n 2 -f hosts ./result2 > > Process 0 sending number 4 > Process 1 received number -1078477356 from process 0 > > My host file: > > dan@siriusa:~/mpich2/tutorial2$ cat hosts > > 192.168.88.1 > 192.168.88.2 > > So there is some sort of problemwith using the send/recieve functions > over the network. As I said I am a beginner at this, so any help will be > greatly appreciated. Thanks in advance. > > Code: > > #include <mpi.h> > > #include <stdio.h> > #include <stdlib.h> > > int main(int argc, char** argv) { > > // Initialize the MPI environment > MPI_Init(NULL, NULL); > // Find out rank, size > int world_rank; > MPI_Comm_rank(MPI_COMM_WORLD, &world_rank); > int world_size; > MPI_Comm_size(MPI_COMM_WORLD, &world_size); > // Get the name of the processor > char processor_name[MPI_MAX_PROCESSOR_NAME]; > int name_len; > MPI_Get_processor_name(processor_name, &name_len); > > // We are assuming at least 2 processes for this task > if (world_size < 2) { > fprintf(stderr, "World size must be greater than 1 for %s\n", > argv[0]); > MPI_Abort(MPI_COMM_WORLD, 1); > } > > int number; > if (world_rank == 0) { > // If we are rank 0, set the number to -1 and send it to process 1 > number = 4; > printf("Process 0 sending number %d\n",number); > MPI_Send(&number, 1, MPI_INT, 1, 0, MPI_COMM_WORLD); > > } else if (world_rank == 1) { > MPI_Recv(&number, 1, MPI_INT, 0, 0, MPI_COMM_WORLD, > MPI_STATUS_IGNORE); > printf("Process 1 received number %d from process 0\n", number); > } > > MPI_Finalize(); > } > > > _______________________________________________ > discuss mailing list discuss(a)mpich.org > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/discuss > -- Pavan Balaji http://www.mcs.anl.gov/~balaji

1 0

send and recieve not working over ethernet
by D Griffiths 01 Dec '12

01 Dec '12

I am completely new to MPI so I hope I’m not asking anything too silly. I have two computers, one running Ubuntu and the other running Scientific Linux. They recognise each other as hosts, I can ssh from each without a password and have disabled any firewalls. Although I can run a simple Hello World application, the trouble starts when I try to use MPI_SEND and MPI_RECV. I am trying a simple program to send an int variable from one node to the other (see source code below). If I run the code with two processes on the same computer I get the correct output: dan@siriusa:~/mpich2/tutorial2$ mpiexec -n 2 ./result2 Process 0 sending number 4 Process 1 received number 4 from process 0 However, if I try to run the code using both computers, I get: dan@siriusa:~/mpich2/tutorial2$ mpiexec -n 2 -f hosts ./result2 Process 0 sending number 4 Process 1 received number -1078477356 from process 0 My host file: dan@siriusa:~/mpich2/tutorial2$ cat hosts 192.168.88.1 192.168.88.2 So there is some sort of problem with using the send/recieve functions over the network. As I said I am a beginner at this, so any help will be greatly appreciated. Thanks in advance. Code: #include <mpi.h> #include <stdio.h> #include <stdlib.h> int main(int argc, char** argv) { // Initialize the MPI environment MPI_Init(NULL, NULL); // Find out rank, size int world_rank; MPI_Comm_rank(MPI_COMM_WORLD, &world_rank); int world_size; MPI_Comm_size(MPI_COMM_WORLD, &world_size); // Get the name of the processor char processor_name[MPI_MAX_PROCESSOR_NAME]; int name_len; MPI_Get_processor_name(processor_name, &name_len); // We are assuming at least 2 processes for this task if (world_size < 2) { fprintf(stderr, "World size must be greater than 1 for %s\n", argv[0]); MPI_Abort(MPI_COMM_WORLD, 1); } int number; if (world_rank == 0) { // If we are rank 0, set the number to -1 and send it to process 1 number = 4; printf("Process 0 sending number %d\n",number); MPI_Send(&number, 1, MPI_INT, 1, 0, MPI_COMM_WORLD); } else if (world_rank == 1) { MPI_Recv(&number, 1, MPI_INT, 0, 0, MPI_COMM_WORLD, MPI_STATUS_IGNORE); printf("Process 1 received number %d from process 0\n", number); } MPI_Finalize(); }

1 0