MPI_COMM_JOIN(3)

💫 MPI_COMM_JOIN(3) — the Open MPI glow-up 💫

okay bestie, welcome to MPI_Comm_join(3) from the Open MPI manual, rewritten so your brain actually keeps it. buckle up 🎀

---

🌸 NAME 🌸

MPI_Comm_join — establishes communication between MPI jobs 💖

like, two separate MPI jobs? this is the little socket-powered friendship bracelet that links them. 🫶

---

✨ SYNTAX ✨

🧊 C Syntax

#include <mpi.h>

int MPI_Comm_join(int fd, MPI_Comm *intercomm)

🧚 Fortran Syntax

USE MPI
! or the older form: INCLUDE 'mpif.h'
MPI_COMM_JOIN(FD, INTERCOMM, IERROR)
     INTEGER FD, INTERCOMM, IERROR

🦋 Fortran 2008 Syntax

USE mpi_f08
MPI_Comm_join(fd, intercomm, ierror)
     INTEGER, INTENT(IN) :: fd
     TYPE(MPI_Comm), INTENT(OUT) :: intercomm
     INTEGER, OPTIONAL, INTENT(OUT) :: ierror

---

🍓 INPUT PARAMETER 🍓

• fd — socket file descriptor (socket). yes, an actual socket. not a file, not a vibe, a socket. 🔌

---

💜 OUTPUT PARAMETERS 💜

• intercomm — intercommunicator between processes (handle). this is the whole point, bestie — you get a shiny new handle connecting the two sides. ✨
• ierror — Fortran only: error status (integer).

---

🔥 DESCRIPTION 🔥

okay so MPI_Comm_join creates an intercommunicator from the union of two MPI processes that are connected by a socket. two processes, one socket, one intercommunicator. it's giving arranged marriage but make it MPI. 💅

fd is a file descriptor representing a socket of type SOCK_STREAM (that's a two-way reliable byte-stream connection, iykyk).

couple of hard rules, no cap:

• Nonblocking I/O must not be enabled for the socket.
• Asynchronous notification via SIGIO must not be enabled for the socket.
• The socket must be in a connected state.
• The socket must be quiescent when MPI_Comm_join is called.

MPI_Comm_join must be called by the process at each end of the socket. and it does not return until both processes have called MPI_Comm_join. so if your bestie on the other end is ghosting you, you're just… sitting there. 💀 it's a mutual handshake, periodt.

---

🌈 NOTES 🌈

• ngl, there are no MPI library calls for opening and manipulating a socket. none. zero. the socket fd gets opened using standard socket API calls — that's on you, bestie. 🫡
• MPI uses the socket to bootstrap creation of the intercommunicator, and for nothing else. one job. iconic.
• Upon return, the file descriptor will be open and quiescent. good as new. ✨
• In a multithreaded process, the application must ensure that other threads do not access the socket while one is in the midst of calling MPI_Comm_join. like, boundaries. do not touch. 🙈
• The returned communicator will contain the two processes connected by the socket, and may be used to establish MPI communication with additional processes, through the usual MPI communicator-creation mechanisms. so it's a starter pack — build your whole friend group from there. 👑

---

⭐ ERRORS ⭐

Important 💀 In Open MPI, this function only works properly when the MPI job is launched via mpirun / mpiexec. Using this function in other run-time environments is currently unsupported, and may result in undefined behavior (e.g., the MPI job may hang). so like… actually read this part. please. 🙏

Almost all MPI routines return an error value — C routines as the return result of the function, and Fortran routines in the last argument.

Before the error value is returned, the current MPI error handler associated with the communication object (e.g., communicator, window, file) is called. If no communication object is associated with the MPI call, then the call is considered attached to MPI_COMM_SELF and will call the associated MPI error handler. When MPI_COMM_SELF is not initialized (i.e., before MPI_Init/MPI_Init_thread, after MPI_Finalize, or when using the Sessions Model exclusively) the error raises the initial error handler. 🔔

The initial error handler can be changed by calling MPI_Comm_set_errhandler on MPI_COMM_SELF when using the World model, or the mpi_initial_errhandler CLI argument to mpiexec, or the info key to MPI_Comm_spawn/MPI_Comm_spawn_multiple.

If no other appropriate error handler has been set, then the MPI_ERRORS_RETURN error handler is called for MPI I/O functions and the MPI_ERRORS_ABORT error handler is called for all other MPI functions.

Open MPI includes three predefined error handlers that can be used:

• MPI_ERRORS_ARE_FATAL — causes the program to abort all connected MPI processes. yes, all of them. dramatic but effective. 💀
• MPI_ERRORS_ABORT — an error handler that can be invoked on a communicator, window, file, or session. When called on a communicator, it acts as if MPI_Abort was called on that communicator. If called on a window or file, acts as if MPI_Abort was called on a communicator containing the group of processes in the corresponding window or file. If called on a session, aborts only the local process. situational awareness, love that. ✨
• MPI_ERRORS_RETURN — returns an error code to the application. polite. respectful. hands you the receipt. 🧾

MPI applications can also implement their own error handlers by calling:

• MPI_Comm_create_errhandler then MPI_Comm_set_errhandler
• MPI_File_create_errhandler then MPI_File_set_errhandler
• MPI_Session_create_errhandler then MPI_Session_set_errhandler or at MPI_Session_init
• MPI_Win_create_errhandler then MPI_Win_set_errhandler

Note that MPI does not guarantee that an MPI program can continue past an error. 😭 so don't assume you can just vibe through it.

See the MPI man page for a full list of MPI error codes.

See the Error Handling section of the MPI-3.1 standard for more information.

---

🦋 SEE ALSO 🦋

• socket(3SOCKET)
• MPI_Comm_create
• MPI_Comm_group

---

👑 COPYRIGHT 👑

2003-2026, The Open MPI Community

---

okay that's the whole page, you ate that 💅 stay hydrated, close your sockets, and remember: both sides gotta call it or nobody's going anywhere. mwah! 🎀✨🫶

the Open MPI crew, Sep 11, 2026 — big love, go get 'em 💖