If an error occurs in an MPI function, the current MPI error handler is called to handle it. By default,
this error handler aborts the MPI job. The error handler may be changed with MPI_Errhandler_set ; the
predefined error handler MPI_ERRORS_RETURN may be used to cause error values to be returned (in C and
Fortran; this error handler is less useful in with the C++ MPI bindings. The predefined error handler
MPI::ERRORS_THROW_EXCEPTIONS should be used in C++ if the error value needs to be recovered). Note that
MPI does not guarantee that an MPI program can continue past an error.
All MPI routines (except MPI_Wtime and MPI_Wtick ) return an error value; C routines as the value of the
function and Fortran routines in the last argument. The C++ bindings for MPI do not return error values;
instead, error values are communicated by throwing exceptions of type MPI::Exception (but not by
default). Exceptions are only thrown if the error value is not MPI::SUCCESS .
Note that if the MPI::ERRORS_RETURN handler is set in C++, while MPI functions will return upon an error,
there will be no way to recover what the actual error value was.
MPI_SUCCESS
- No error; MPI routine completed successfully.
MPI_ERR_COUNT
- Invalid count argument. Count arguments must be non-negative; a count of zero is often valid.
MPI_ERR_ARG
- Invalid argument. Some argument is invalid and is not identified by a specific error class.
This is typically a NULL pointer or other such error.
MPI_ERR_IN_STATUS
- The actual error value is in the MPI_Status argument. Note that if this error occurs and
MPI_STATUS_IGNORE or MPI_STATUSES_IGNORE was used as the status argument, the actual error will be
lost. This error class is returned only from the multiple-completion routines ( MPI_Testall ,
MPI_Testany , MPI_Testsome , MPI_Waitall , MPI_Waitany , and MPI_Waitsome ). The field MPI_ERROR
in the status argument contains the error value or MPI_SUCCESS (no error and complete) or
MPI_ERR_PENDING to indicate that the request has not completed. The MPI Standard does not specify
what the result of the multiple completion routines is when an error occurs. For example, in an
MPI_WAITALL , does the routine wait for all requests to either fail or complete, or does it return
immediately (with the MPI definition of immediately, which means independent of actions of other
MPI processes)? LAM/MPI has chosen to make the return immediate (alternately, local in MPI
terms), and to use the error class MPI_ERR_PENDING (introduced in MPI 1.1) to indicate which
requests have not completed. In most cases, only one request with an error will be detected in
each call to an MPI routine that tests multiple requests. The requests that have not been
processed (because an error occured in one of the requests) will have their MPI_ERROR field marked
with MPI_ERR_PENDING .