MPI_WIN_ALLOCATE_SHARED(3)

📦
version 5.0.10 · 7 versions
, in ubuntu stonking

✨ MPI_Win_allocate_shared(3) ✨

okay bestie, listen up — this is a page from Open MPI, and it's about a one-sided MPI call that allocates shared memory and hands you back a window object for RMA operations. no cap, it's giving "we're all sharing a big memory vibes" and honestly? iconic. 💖

SYNOPSIS 🎀

C Syntax

#include <mpi.h>

int MPI_Win_allocate_shared (MPI_Aint size, int disp_unit, MPI_Info info,
                             MPI_Comm comm, void *baseptr, MPI_Win *win)

Fortran Syntax

USE MPI
! or the older form: INCLUDE 'mpif.h'
MPI_WIN_ALLOCATE_SHARED(SIZE, DISP_UNIT, INFO, COMM, BASEPTR, WIN, IERROR)
     INTEGER(KIND=MPI_ADDRESS_KIND) SIZE, BASEPTR
     INTEGER DISP_UNIT, INFO, COMM, WIN, IERROR

Fortran 2008 Syntax

USE mpi_f08
MPI_Win_allocate_shared(size, disp_unit, info, comm, baseptr, win, ierror)
     USE, INTRINSIC :: ISO_C_BINDING, ONLY : C_PTR
     INTEGER(KIND=MPI_ADDRESS_KIND), INTENT(IN) :: size
     INTEGER, INTENT(IN) :: disp_unit
     TYPE(MPI_Info), INTENT(IN) :: info
     TYPE(MPI_Comm), INTENT(IN) :: comm
     TYPE(C_PTR), INTENT(OUT) :: baseptr
     TYPE(MPI_Win), INTENT(OUT) :: win
     INTEGER, OPTIONAL, INTENT(OUT) :: ierror

INPUT PARAMETERS 💅

  • size — how big the window is, in bytes. it's a nonnegative integer, so like, zero is fine, negative is not the vibe. 🚫
  • disp_unit — the local unit size for displacements, in bytes. positive integer only bestie. ✨
  • info — the info argument (handle).
  • comm — the communicator (handle).

OUTPUT PARAMETERS 🫶

  • baseptr — the initial address of the window.
  • win — the window object the call gives back to you (handle).
  • ierror — Fortran only: the error status (integer).

DESCRIPTION 👑

MPI_Win_allocate_shared is a collective call — meaning every single process in the group of comm has to show up, no ghosting. ✋ On each process, it allocates at least size bytes of memory that's shared among all processes in comm, and returns a pointer to the locally allocated segment in baseptr. That pointer is what you use for load/store accesses on the calling process. and the way that your locally allocated memory can literally be the target of load/store accesses by other processes? main character energy fr. 🔥

The base pointers for the other processes can be queried using the function MPI_Win_shared_query.

The call also returns a window object that all processes in comm can use to do RMA operations. The size argument may be different at each process, and size = 0 is valid. 💀 (yes, zero. slay.)

It is the user's responsibility to make sure the communicator comm represents a group of processes that can create a shared memory segment accessible by everyone in the group. so like, do your homework, bestie.

The discussions of rationales for MPI_Alloc_mem and MPI_Free_mem in MPI-3.1 section 8.2 also apply to MPI_Win_allocate_shared. In particular, peep the rationale in MPI-3.1 section 8.2 for an explanation of the type used for baseptr. 👀

The allocated memory is contiguous across process ranks unless the info key alloc_shared_noncontig is specified. "Contiguous across process ranks" means the first address in the memory segment of process i is consecutive with the last address in the memory segment of process i - 1. This may let the user calculate remote address offsets with local information only. iykyk. 🌈

the info keys, bestie 💜

alloc_shared_noncontig

If not set to true, the allocation strategy is to allocate contiguous memory across process ranks. This may limit the performance on some architectures because it does not allow the implementation to modify the data layout (e.g., padding to reduce access latency). so like, tradesies.

blocking_fence

If set to true, the osc/sm component will use MPI_Barrier for MPI_Win_fence. If set to false, a condition variable and counter will be used instead. The default value is false. This info key is Open MPI specific. 🫢

For additional supported info keys, see MPI_Win_create.

NOTES 🦋

Common choices for disp_unit are 1 (no scaling), and — in C syntax — sizeof(type), for a window that consists of an array of elements of type type. The latter choice will let you use array indices in RMA calls and have those scaled correctly to byte displacements, even in a heterogeneous environment. so yes, the second one is kinda that girl. 💅

Calling MPI_Win_free will deallocate the memory allocated by MPI_Win_allocate_shared. It is thus erroneous to manually free baseptr. do NOT do it. put the free() down. 🙈

C NOTES 🌸

While baseptr is a void * type, this is to allow easy use of any pointer object for this parameter. This argument is really a void ** type. plot twist, but make it C. 😭

ERRORS 💀

Almost all MPI routines return an error value. C routines return it as the return result of the function, and Fortran routines put it 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 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 you can use:

  • MPI_ERRORS_ARE_FATAL — causes the program to abort all connected MPI processes. no survivors. 😭
  • 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, it 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, it aborts only the local process.
  • MPI_ERRORS_RETURN — returns an error code to the application.

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. 😬

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:

  • MPI_Alloc_mem
  • MPI_Free_mem
  • MPI_Win_allocate
  • MPI_Win_create
  • MPI_Win_shared_query
  • MPI_Win_free

2003-2026, The Open MPI Community

okay that's the whole page bestie, you ate it UP. 💖 little sign-off moment: this was the Open MPI MPI_WIN_ALLOCATE_SHARED(3) page, dated Sep 10, 2026, and the copyright belongs to 2003-2026, The Open MPI Community. periodt. 💅✨

📦 where this page comes from