I think that we've fixed everything with respect to f90 except the "large" interface.  Let us know if we either missed something or you find something new.


 -----Original Message-----
From:   Michael Kluskens [mailto:mkluskens@ieee.org]
Sent:   Mon Jun 12 09:47:52 2006
To:     Open MPI Users
Subject:        Re: [OMPI users] F90 interfaces again

On Jun 9, 2006, at 12:33 PM, Brian W. Barrett wrote:

> On Thu, 8 Jun 2006, Michael Kluskens wrote:
>>             call MPI_WAITALL(3,sp_request,MPI_STATUSES_IGNORE,ier)
>>                                                                   1
>> Error: Generic subroutine 'mpi_waitall' at (1) is not consistent with
>> a specific subroutine interface
>> Issue, 3rd argument of MPI_WAITALL expects an integer array normally,
>> but this constant is permitted by the standard.
>> This is with OpenMPI 1.2a1r10186,  I can check the details of the
>> scripts and generated files next week for whatever is the latest
>> version.  But odds are this has not been spotted.
> Michael -
> Which compiler are you using?  I'm trying to replicate, and 
> gfortran isn't getting mad at me (which is weird, because I would 
> have expected it to get very angry at me).

I'm using g95 and I configure with:

./configure F77=g95 FC=g95 LDFLAGS=-lSystemStubs --with-mpi-f90-
size=large --enable-static --with-f90-max-array-dim=3

Downloaded "openmpi-1.2a1r10297" today.

Looking at the scripts I believe this problem was fixed between 10186 
and 10297.

I can't test until I check for and if needed implement my fixes to 
the other interfaces I mentioned previously.


users mailing list