Sorry for the delay in replying -- SC and the US holiday conspired
against me to fill up my INBOX and I'm just now getting to the
backlog...
You quite definitely identified a bug in mpiexec. It's a silly
perl-ism that I have now fixed in the repository, and this will be
included in 7.1.2. The nightly tarball tomorrow will include the fix.
Thanks!
On Nov 22, 2005, at 11:03 AM, Michael Lees wrote:
> Hi again,
>
> following my email last week....
>
> It seems when I do
> mpirun -np 1 app arg1 arg2 ... arg7 everything works fine
> argc = 8 and argv is correct
>
> however,
>
> if I perform the precise same call but simply change "run" => "exec"
> then,
> argc = 6 and argv is wrong
>
>
> Jeff kindly sent me a test program which worked fine. I've also tried
> using the absolute path to mpiexec.
> However my application which works fine with mpirun fails with mpiexec
> because an incorrect number of command line arguments are passed. I've
> tried the -v option but that doesn't really tell me anything new.
>
> I think I tracked down the problem it seems if one of the command line
> arguments is '0' the rest of the line is ignored by mpiexec. Using
> hello.c (see below) with the following command line arguments "arg1
> arg2
> arg3 0 arg4 arg5 arg6"
>
> compile hello.c with "mpicc hello.c -o hello"
>
>
> mpirun....
>> mpirun -np 1 hello arg1 arg2 arg3 0 arg4 arg5 arg6
> hello arg1 arg2 arg3 0 arg4 arg5 arg6 Hello, world! I am 0 of 1 (argv:
> arg1)
>
>
> mpiexec....
>> mpiexec -np 1 hello arg1 arg2 arg3 0 arg4 arg5 arg6
> hello arg1 arg2 arg3 Hello, world! I am 0 of 1 (argv: arg1)
>
>
> It shows mpiexec gives a different result to mpirun.
>
>
>> lamboot -v
> LAM 7.1.1/MPI 2 C++/ROMIO - Indiana University
>
>> mpicc -v
> Reading specs from /usr/lib/gcc/i386-redhat-linux/3.4.4/specs
> Configured with: ../configure --prefix=/usr --mandir=/usr/share/man
> --infodir=/usr/share/info --enable-shared --enable-threads=posix
> --disable-checking --with-system-zlib --enable-__cxa_atexit
> --disable-libunwind-exceptions --enable-java-awt=gtk
> --host=i386-redhat-linux
> Thread model: posix
> gcc version 3.4.4 20050721 (Red Hat 3.4.4-2)
>
>
> Is this a bug? If so I'm suprised no one has come across it before?
>
> -- hello.c --
> #include <stdio.h>
> #include <mpi.h>
> #include <stdlib.h>
>
> int main(int argc, char **argv)
> {
> int rank = 0, size = 0;
> char hostname[BUFSIZ];
>
> MPI_Init(&argc, &argv);
> MPI_Comm_rank(MPI_COMM_WORLD, &rank);
> MPI_Comm_size(MPI_COMM_WORLD, &size);
> gethostname(hostname);
> int i =0;
> for(i=0;i<argc;i++){
> printf("%s ",argv[i]);
> }
> printf("Hello, world! I am %d of %d (argv: %s)\n", rank, size,
> argv[1]);
>
> MPI_Finalize();
> return 0;
> }
>
>
> Cheers
>
> --
> Mike
>
> mhl_at_[hidden]
> http://www.cs.nott.ac.uk/~mhl/
>
> This message has been checked for viruses but the contents of an
> attachment
> may still contain software viruses, which could damage your computer
> system:
> you are advised to perform your own checks. Email communications with
> the
> University of Nottingham may be monitored as permitted by UK
> legislation.
>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>
--
{+} Jeff Squyres
{+} The Open MPI Project
{+} http://www.open-mpi.org/
|
