MPI_GROUP_F2C(3)
π MPI_GROUP_F2C(3) π
okay bestie, so this is the manual page for a whole little family of Open MPI functions π and it's giving "translator" energy, because that's literally what they do. Folding the manual name in right here so you know what you're reading: MPI_GROUP_F2C(3) β iconic, we stan a page that does the most with sixteen functions and one job.
Here's the whole girlgang of them β¨:
MPI_Comm_f2c<#mpi-comm-f2c>MPI_Comm_c2f<#mpi-comm-c2f>MPI_File_f2c<#mpi-file-f2c>MPI_File_c2f<#mpi-file-c2f>MPI_Info_f2c<#mpi-info-f2c>MPI_Info_c2f<#mpi-info-c2f>MPI_Message_f2c<#mpi-message-f2c>MPI_Message_c2f<#mpi-message-c2f>MPI_Op_f2c<#mpi-op-f2c>MPI_Op_c2f<#mpi-op-c2f>MPI_Request_f2c<#mpi-request-f2c>MPI_Request_c2f<#mpi-request-c2f>MPI_Type_f2c<#mpi-type-f2c>MPI_Type_c2f<#mpi-type-c2f>MPI_Win_f2c<#mpi-win-f2c>MPI_Win_c2f<#mpi-win-c2f>
β¦and what do they do? They translate a C handle into a Fortran handle, or the other way around. That's it, that's the whole vibe. π
β¨ SYNTAX β¨
π«Ά C Syntax π«Ά
#include <mpi.h>
MPI_Comm MPI_Comm_f2c(MPI_Fint comm)
MPI_Fint MPI_Comm_c2f(MPI_Comm comm)
MPI_File MPI_File_f2c(MPI_Fint file)
MPI_Fint MPI_File_c2f(MPI_File file)
MPI_Group MPI_Group_f2c(MPI Fint group)
MPI_Fint MPI_Group_c2f(MPI Group group)
MPI_Info MPI_Info_f2c(MPI_Fint info)
MPI_Fint MPI_Info_c2f(MPI_Info info)
MPI_Message MPI_Message_f2c(MPI_Fint message)
MPI_Fint MPI_Message_c2f(MPI_Message message)
MPI_Op MPI_Op_f2c(MPI_Fint op)
MPI_Fint MPI_Op_c2f(MPI_Op op)
MPI_Request MPI_Request_f2c(MPI_Fint request)
MPI_Fint MPI_Request_c2f(MPI_Request request)
MPI_Datatype MPI_Type_f2c(MPI_Fint datatype)
MPI_Fint MPI_Type_c2f(MPI_Datatype datatype)
MPI_Win MPI_Win_f2c(MPI_Fint win)
MPI_Fint MPI_Win_c2f(MPI_Win win)
π DESCRIPTION π
okay so here's the tea β: handles get passed between Fortran and C by using an explicit C wrapper to convert Fortran handles into C handles. Periodt. There is no direct access to C handles in Fortran. None. Don't even try it.
The type definition MPI_Fint is provided in C for an integer of the size that matches a Fortran INTEGER; usually, MPI_Fint will be equivalent to int. (yeah, "usually" β it's giving mostly, and that's the note we're working with π« )
And the handle translation functions? They're provided in C to convert from a Fortran handle (which is literally just an integer, bestie) to a C handle, and vice versa. π
for example β and I love a good example, it's the main character of this section:
- if
commis a valid Fortran handle to a communicator, thenMPI_Comm_f2c<#mpi-comm-f2c> returns a valid C handle to that same communicator β - if
comm = MPI_COMM_NULL(that's the Fortran value, ok), thenMPI_Comm_f2c<#mpi-comm-f2c> returns a null C handle π«₯ - if
commis an invalid Fortran handle, thenMPI_Comm_f2c<#mpi-comm-f2c> returns an invalid C handle π
π NOTE π
okay this part is important, no cap:
This function does not return an error value. Consequently, the result of calling it beforeMPI_Init<#mpi-init> or afterMPI_Finalize<#mpi-finalize> is undefined.
so likeβ¦ don't. seriously. undefined behaviour is not the vibe check you want to fail. π
π Copyright π
2003-2026, The Open MPI Community
and that's the whole page bestie, you ate it up π β¨ it's from the Open MPI manual, page MPI_GROUP_F2C(3), dated Sep 10, 2026 β go forth and translate those handles, you've got this π«Άπ¦
π¦ where this page comes from
- ubuntu stonking: openmpi-doc 5.0.10-6 Β· other
- π girlified by deepseek-ai/DeepSeek-V4.1-Flash on : a language model's rewrite of the original, under its terms