18.2.134. MPI_File_open

MPI_File_open — Opens a file (collective).

18.2.134.1. SYNTAX

18.2.134.1.1. C Syntax

int MPI_File_open(MPI_Comm comm, const char* filename, int amode,
    MPI_Info info, MPI_File* fh)

18.2.134.1.2. Fortran Syntax

USE MPI
! or the older form: INCLUDE 'mpif.h'
MPI_FILE_OPEN(COMM, FILENAME, AMODE, INFO, FH, IERROR)
    INTEGER COMM, AMODE, INFO, FH, IERROR
    CHARACTER*(*) FILENAME

18.2.134.1.3. Fortran 2008 Syntax

USE mpi_f08
MPI_File_open(comm, filename, amode, info, fh, ierror)
    TYPE(MPI_Comm), INTENT(IN) :: comm
    CHARACTER(LEN=*), INTENT(IN) :: filename
    INTEGER, INTENT(IN) :: amode
    TYPE(MPI_Info), INTENT(IN) :: info
    TYPE(MPI_File), INTENT(OUT) :: fh
    INTEGER, OPTIONAL, INTENT(OUT) :: ierror

18.2.134.2. INPUT PARAMETERS

• comm: Communicator (handle).

• filename: Name of file to open (string).

• amode: File access mode (integer).

• info: Info object (handle).

18.2.134.3. OUTPUT PARAMETERS

• fh: New file handle (handle).

• ierror: Fortran only: Error status (integer).

18.2.134.4. DESCRIPTION

MPI_File_open opens the file identified by the filename filename on all processes in the comm communicator group. MPI_File_open is a collective routine; all processes must provide the same value for amode, and all processes must provide filenames that reference the same file which are textually identical (note: Open MPI I/O plugins may have restrictions on characters that can be used in filenames. A process can open a file independently of other processes by using the MPI_COMM_SELF communicator. The file handle returned, fh, can be subsequently used to access the file until the file is closed using MPI_File_close. Before calling MPI_Finalize, the user is required to close (via MPI_File_close) all files that were opened with MPI_File_open. Note that the communicator comm is unaffected by MPI_File_open and continues to be usable in all MPI routines. Furthermore, use of comm will not interfere with I/O behavior.

Initially, all processes view the file as a linear byte stream; that is, the etype and filetype are both MPI_BYTE. The file view can be changed via the MPI_File_set_view routine.

The following access modes are supported (specified in amode, in a bit-vector OR in one of the following integer constants):

• MPI_MODE_APPEND

• MPI_MODE_CREATE: Create the file if it does not exist.

• MPI_MODE_DELETE_ON_CLOSE

• MPI_MODE_EXCL: Error creating a file that already exists.

• MPI_MODE_RDONLY: Read only.

• MPI_MODE_RDWR: Reading and writing.

• MPI_MODE_SEQUENTIAL

• MPI__MODE_WRONLY: Write only.

• MPI_MODE_UNIQUE_OPEN

The modes MPI_MODE_RDONLY, MPI_MODE_RDWR, MPI_MODE_WRONLY, and MPI_MODE_CREATE have identical semantics to their POSIX counterparts. It is erroneous to specify MPI_MODE_CREATE in conjunction with MPI_MODE_RDONLY. Errors related to the access mode are raised in the class MPI_ERR_AMODE.

On single-node clusters, files are opened by default using nonatomic mode file consistency semantics. The more stringent atomic-mode consistency semantics, required for atomicity of overlapping accesses, are the default when processors in a communicator group reside on more than one node. This setting can be changed using MPI_File_set_atomicity.

The MPI_File_open interface allows the user to pass information via the info argument. It can be set to MPI_INFO_NULL. See the HINTS section for a list of hints that can be set.

18.2.134.5. HINTS

MPI allows applications to pass MPI_Info hints to MPI_File_open, MPI_File_set_view, and MPI_File_set_info. Hints are advisory: OMPIO may ignore hints that are not supported in the selected component stack. The MPI_File_get_info routine returns a new MPI_Info object containing the public hints that OMPIO currently associates with the file. The caller is responsible for freeing that returned object with MPI_Info_free.

OMPIO preserves the spelling of a supported hint that was accepted from the user’s MPI_Info object. When OMPIO reports an internal default that did not come from a user-supplied hint, it uses the canonical public spelling shown below.

18.2.134.5.1. OMPIO common hints

Hint	Accepted by	Default / MCA parameter	Notes
cb_buffer_size	MPI_File_open, MPI_File_set_info, MPI_File_set_view	io_ompio_bytes_per_agg	Size, in bytes, of the temporary collective I/O buffer on each aggregator.
cb_nodes	MPI_File_open, MPI_File_set_info, MPI_File_set_view	io_ompio_num_aggregators when that MCA parameter is set to a non-negative value.	Number of collective I/O aggregators. When OMPIO is using its automatic aggregator selection, this hint is not returned as -1.
collective_buffering	MPI_File_open, MPI_File_set_info, MPI_File_set_view	true	Controls whether collective buffering is enabled for collective I/O decisions.

18.2.134.5.2. sharedfp component hints

Component	Hint	Accepted by	Default / MCA parameter	Notes
sharedfp/individual	OMPIO_SHAREDFP_RELAXED_ORDERING	MPI_File_open	No default; no corresponding MCA parameter.	If present on a write-capable file open, this hint raises the priority of the individual shared file pointer component. OMPIO reports it only when that component is selected for the file.

18.2.134.5.3. Lustre fs component hints

The Lustre layout hints affect file layout creation. Lustre applies these settings only when a file is created in a mode where layout can be set safely; changing them after a file has been opened does not change the already-created file layout.

Canonical hint	Accepted alias	Accepted by	Default / MCA parameter	Notes
striping_unit	stripe_size	MPI_File_open	fs_lustre_stripe_size when greater than zero.	Stripe size in bytes.
striping_factor	stripe_width	MPI_File_open	fs_lustre_stripe_width when greater than zero.	Number of Lustre object storage targets to stripe across.

If an application supplies an alias, MPI_File_get_info returns the alias spelling. If both a canonical hint and its alias are supplied, OMPIO prefers the canonical spelling. If only an MCA parameter supplies the value, OMPIO reports the canonical spelling.

18.2.134.5.4. GPFS fs component hints

The GPFS component accepts the hints listed below when the GPFS component is built and selected. These hints are consumed when the file is opened. Later MPI_File_set_info or MPI_File_set_view calls do not reapply them to GPFS, and MPI_File_get_info continues to report the open-time accepted values.

Hint	Accepted by	Default / MCA parameter	Notes
useSIOXLib	MPI_File_open	No default; no corresponding MCA parameter.	Enables optional SIOX-assisted GPFS I/O selection when SIOX support is compiled in.
gpfsAccessRange	MPI_File_open	No default; no corresponding MCA parameter.	Passed to GPFS as an access range hint.
gpfsFreeRange	MPI_File_open	No default; no corresponding MCA parameter.	Passed to GPFS as a free range hint.
gpfsClearFileCache	MPI_File_open	No default; no corresponding MCA parameter.	Requests GPFS file-cache clearing.
gpfsCancelHints	MPI_File_open	No default; no corresponding MCA parameter.	Requests cancellation of GPFS hints.
gpfsSetReplication	MPI_File_open	No default; no corresponding MCA parameter.	Passed to GPFS as a replication hint.
gpfsByteRange	MPI_File_open	No default; no corresponding MCA parameter.	Passed to GPFS as a byte range hint.
gpfsRestripeData	MPI_File_open	No default; no corresponding MCA parameter.	Requests GPFS data restriping.

When SIOX support is compiled in, the GPFS component also accepts: sioxAccessRange, sioxFreeRange, sioxClearFileCache, sioxCancelHints, sioxDataShipStart, sioxDataShipStop, sioxSetReplication, sioxByteRange, and sioxRestripeData.

18.2.134.6. ERRORS

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 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.

• 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.

• 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-4.1 standard for more information.