LAM/MPI logo

LAM/MPI General User's Mailing List Archives

  |   Home   |   Download   |   Documentation   |   FAQ   |   all just in this list

From: Josh Hursey (jjhursey_at_[hidden])
Date: 2005-04-20 08:22:14


On Apr 20, 2005, at 5:13 AM, Ru-Zhen Li wrote:

> Dear all,
>  
> I am having error messages like this
>  
> 0 - MPI_RECV : Message truncated
> [0]  Aborting program !
> [0] Aborting program !
> p0_29298: p4_error: : 14
> Killed by signal 2.
> .
> .
> Killed by signal 2
> p0_29298: (1873.003170) net_send: could not write to fd=4, errno = 32
>  
>

It appears that you are using MPICH. This list is for support related
to the LAM/MPI implementation of the MPI standard. For support on
MPICH, you should contact the authors at Argonne National Labs:

   http://www-unix.mcs.anl.gov/mpi/mpich/

That being said, truncation errors like this are usually the result of
a mismatch in the send and receive buffer sizes/types. You may want to
check your program to make sure that if you send N doubles, that you
are receiving N doubles in the corresponding process.

Cheers,
Josh

>  
>  
> I have checked on the internet, haven't found any answers yet, does
> any one have any idea how this happens? it is very strange because the
> program quited half way, it started, with no error, and quited before
> finish, with no error message in the output files of the program
> itself, but in the standard output file of mpi, Thank you very much
> indeed in advance!
>  
> Regards,
> Ruzhen_______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/

----
Josh Hursey
jjhursey_at_[hidden]
http://www.lam-mpi.org/