[mpich-discuss] Destination value is being overwritten during MPI_Isend

[coolhead.pranay at gmail.com]

the value returned by the function which generates the destination value is
less than nprocs but it magically changes to 99 once it is passed to Isend.

assert(dest<nprocs) is successful in the function but fails in the caller
which contains Isend.

On Fri, Feb 17, 2012 at 9:32 AM, Rajeev Thakur <thakur at mcs.anl.gov> wrote:

> Can you print the value of dest just before calling MPI_Isend?
>
>
> On Feb 17, 2012, at 12:33 AM, coolhead.pranay at gmail.com wrote:
>
> > Hi,
> >
> > I encountered a strange problem when trying to send a message using
> MPI_ISend. The program exits with the following error message
> >
> > Fatal error in MPI_Isend: Invalid rank, error stack:
> > MPI_Isend(146): MPI_Isend(buf=0x6bbfc0, count=129, MPI_DOUBLE, dest=99,
> tag=99, MPI_COMM_WORLD, request=0x6bd000) failed
> > MPI_Isend(96).: Invalid rank has value 99 but must be nonnegative and
> less than 4
> > rank 1 in job 129  kratos_43452   caused collective abort of all ranks
> >   exit status of rank 1: return code 1
> >
> > I am pretty sure that the value of destination is < nprocs. Another
> strange error is that when I change the tag to 88 the error is almost the
> same except dest=88, "Invalid rank has a value 88".
> >
> > Can tell me possible reasons for this strange behavior ?
> >
> >
> >
> >
> > _______________________________________________
> > mpich-discuss mailing list     mpich-discuss at mcs.anl.gov
> > To manage subscription options or unsubscribe:
> > https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>
> _______________________________________________
> mpich-discuss mailing list     mpich-discuss at mcs.anl.gov
> To manage subscription options or unsubscribe:
> https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss
>



-- 
Pranay Anchuri
Computer Science Graduate Student
Rensselaer Polytechnic Institute
Troy,NY
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/mpich-discuss/attachments/20120217/6fa43faa/attachment.htm>