Well, that's a bad choice - it means an unnecessary copy in the Fortran wrapper just in case. And it only has(marginal) value in some of the multiple completion cases. I'll update the wrapper generation to support this mistake (particularly for the single completion routines, where the value should *always* be undefined. Bill Sent from my iPad On Apr 30, 2013, at 4:23 PM, "Rajeev Thakur" <[email protected]> wrote:

For Mprobe, the standard says

"The call matches the same message that would have been received by a call to MPI_RECV(..., source, tag, comm, status) executed at the same point in the program and returns in status the same value that would have been returned by MPI_RECV."

For MPI_Recv, it says

"In general, message-passing calls do not modify the value of the error code field of status variables. This field may be updated only by the functions in Section 3.7.5 which return multiple statuses. The field is updated if and only if such function returns with an error code of MPI_ERR_IN_STATUS."

Rajeev

On Apr 30, 2013, at 2:10 PM, William Gropp wrote:

...

I'm getting failures in the Fortran version of the MPROBE tests because the test assumes that STATUS is an INOUT argument, and it sets the ERROR field and then checks that the value is unchanged. I don't see the requirement that the value of the MPI_ERROR field be unchanged. Is this an error in the test or is there another part of the standard of which I've forgotten?

Bill

William Gropp Director, Parallel Computing Institute Deputy Director for Research Institute for Advanced Computing Applications and Technologies Thomas M. Siebel Chair in Computer Science University of Illinois Urbana-Champaign