Harold --
Along those lines, would you mind downloading a nightly trunk tarball
and giving it a whirl to see if the IOF patch we have soaking does fix
the problem in your environment? Part of the difficulty we've had in
testing the patch is that the problem was only intermittently
repeatable on some machines in some environments.
http://www.open-mpi.org/nightly/trunk/
On Dec 12, 2008, at 11:45 AM, Ralph Castain wrote:
Hi Harald
There is a patch for the IOF in 1.3 "soaking" in the trunk right
now. I'll check to ensure it fixes this issue too. Hopefully, it
will come over to the 1.3 branch early next week.
Ralph
On Dec 12, 2008, at 8:21 AM, Harald Anlauf wrote:
Hello,
I am having problems with OMPI-1.3beta with an interactive job
where rank 0
reads stdin from a terminal. The problem does not show up when stdin
is redirected from a file. The problem also does not exist with
OMPI 1.2.[5-9].
Has there been any change in OMPI between 1.2 and 1.3 that I should
take
care of?
Please find attached a famous sample program that was modified to
aid debugging.
The program reads the number of intervals used to calculate pi. 0
means exit.
I first enter 1000, then 0.
Interactive run, without mpirun:
% ./a.out
Process 0 of 1 is alive
Process 0 before read
Enter the number of intervals: (0 quits)
1000
Process 0 read: n = 1000
Process 0 before MPI_BCAST
Process 0 after MPI_BCAST
pi is approximately: 3.1415927369231227 Error is: 0.0000000833333296
Process 0 before read
Enter the number of intervals: (0 quits)
0
Process 0 read: n = 0
Process 0 before MPI_BCAST
Process 0 after MPI_BCAST
Process 0 Normal exit
With mpirun:
% mpirun -np 1 ./a.out
Process 0 of 1 is alive
Process 0 before read
Enter the number of intervals: (0 quits)
1000
--------------------------------------------------------------------------
mpirun has exited due to process rank 0 with PID 10909 on
node oflws105 exiting without calling "finalize". This may
have caused other processes in the application to be
terminated by signals sent by mpirun (as reported here).
--------------------------------------------------------------------------
Now with stdin redirected from a here-document:
% mpirun -np 1 ./a.out <<EOF
1000
0
EOF
Process 0 of 1 is alive
Process 0 before read
Enter the number of intervals: (0 quits)
Process 0 read: n = 1000
Process 0 before MPI_BCAST
Process 0 after MPI_BCAST
pi is approximately: 3.1415927369231227 Error is: 0.0000000833333296
Process 0 before read
Enter the number of intervals: (0 quits)
Process 0 read: n = 0
Process 0 before MPI_BCAST
Process 0 after MPI_BCAST
Similarly for np > 1, with minor variations, but the same error
message.
Can anybody reproduce this behavior?
% ompi_info |grep SVN
Open MPI SVN revision: r20119
Open RTE SVN revision: r20119
OPAL SVN revision: r20119
Regards,
Harald
--
Sensationsangebot verlängert: GMX FreeDSL - Telefonanschluss + DSL
für nur 16,37 Euro/mtl.!* http://dsl.gmx.de/?
ac=OM.AD.PD003K1308T4569a
<pi3f90-test.f90>_______________________________________________
users mailing list
us...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/users
_______________________________________________
users mailing list
us...@open-mpi.org
http://www.open-mpi.org/mailman/listinfo.cgi/users
--
Jeff Squyres
Cisco Systems
