MPI_Op_create(3)

πŸ“¦
version 5.0.1 Β· 5 versions
, in nixos 26.05, ubuntu stonking

✨ MPI_Op_create(3) β€” the MPI manual page, bestie edition ✨

okay so picture this: you wanna teach MPI your own little math move, and you want to use it in a reduction. That's literally what this is for. MPI_Op_create creates a user-defined combination function handle. Iconic. πŸ‘‘


🌸 SYNOPSIS 🌸

int MPI_Op_create(MPI_User_function *user_fn, int commute, MPI_Op *op)

int MPI_Op_create_c(MPI_User_function_c *user_fn, int commute, MPI_Op *op)

πŸ’œ INPUT PARAMETERS πŸ’œ

  • user_fn β€” the user defined function you're handing over (a function, obvs)
  • commute β€” true if it's commutative; false otherwise (logical, so like, a yes/no vibe)

πŸŽ€ OUTPUT PARAMETERS πŸŽ€

  • op β€” the operation (a handle, so you can pass it around later)

🧚 NOTES ON THE USER FUNCTION 🧚

k so for MPI_Op_create, the calling list for the user function type is literally this:

typedef void (MPI_User_function) (void * a,
void * b, int * len, MPI_Datatype *);

and for MPI_Op_create_c, the calling list for the user function type is this instead:

typedef void (MPI_User_function_c) (void * a,
void * b, MPI_Count * len, MPI_Datatype *);

and the operation happening there is b[i] = a[i] op b[i] , for i=0,...,len-1 . periodt. πŸ’…

Alsooo, a pointer to the datatype that got handed to the MPI collective computation routine (i.e., MPI_Reduce , MPI_Allreduce , MPI_Scan , or MPI_Reduce_scatter ) gets passed to your user-specified routine too. So it knows what type it's looking at, no cap.


😭 NOTES ON COLLECTIVE OPERATIONS 😭

The reduction functions (MPI_Op) do not return an error value. So if they spot an error, all they can do is either call MPI_Abort or silently skip the problem. Which means… if you swap the error handler from MPI_ERRORS_ARE_FATAL to something else, like MPI_ERRORS_RETURN , then no error may be indicated. Yeah. 😬

The reason for this is the performance problems in ensuring that all collective routines return the same error value.

Like, it's a tradeoff bestie, don't come for them.


🫢 THREAD AND INTERRUPT SAFETY 🫢

This routine is thread-safe. That means it may be safely used by multiple threads without needing any user-provided thread locks. Slay. πŸ’…

However, the routine is not interrupt safe. Typically this is because of memory allocation routines like malloc or other non-MPICH runtime routines that are themselves not interrupt-safe.


🌈 NOTES FOR FORTRAN 🌈

All MPI routines in Fortran (except for MPI_WTIME and MPI_WTICK ) have an additional argument named ierr tacked on the end of the argument list. ierr is an integer and has the same meaning as the return value of the routine in C. In Fortran, MPI routines are subroutines, and are invoked with the call statement.

All MPI objects (e.g., MPI_Datatype , MPI_Comm ) are of type INTEGER in Fortran.


πŸ’€ ERRORS πŸ’€

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. Before that value is returned, the current MPI error handler gets called. By default, this error handler aborts the MPI job. 😳

The error handler may be changed with MPI_Comm_set_errhandler (for communicators), MPI_File_set_errhandler (for files), and MPI_Win_set_errhandler (for RMA windows). The MPI-1 routine MPI_Errhandler_set may be used but its use is deprecated. The predefined error handler MPI_ERRORS_RETURN may be used to cause error values to be returned. Note that MPI does not guarantee that an MPI program can continue past an error; however, MPI implementations will attempt to continue whenever possible.

  • MPI_SUCCESS β€” no error; MPI routine completed successfully. πŸŽ‰
  • MPI_ERR_ARG β€” invalid argument. Some argument is invalid and is not identified by a specific error class (e.g., MPI_ERR_RANK ).
  • MPI_ERR_OTHER β€” other error; use MPI_Error_string to get more information about this error code.

πŸ‘‘ SEE ALSO πŸ‘‘

MPI_Op_free


and that's the whole page, bestie πŸ¦‹ keep being the main character of your own reduction. This was the MPI manual, dated 4/10/2026, page MPI_Op_create(3) β€” mwah, ily, go free that op πŸ’–βœ¨πŸŽ€

πŸ“¦ where this page comes from