Functionality

Functionality


Up: Caching Next: Attributes Example Previous: Caching

Attributes are attached to communicators. Attributes are local to the process and specific to the communicator to which they are attached. Attributes are not propagated by MPI from one communicator to another except when the communicator is duplicated using MPI_COMM_DUP (and even then the application must give specific permission through callback functions for the attribute to be copied).


[] Advice to users.

Attributes in C are of type void *. Typically, such an attribute will be a pointer to a structure that contains further information, or a handle to an MPI object. In Fortran, attributes are of type INTEGER. Such attribute can be a handle to an MPI object, or just an integer-valued attribute. ( End of advice to users.)


[] Advice to implementors.

Attributes are scalar values, equal in size to, or larger than a C-language pointer. Attributes can always hold an MPI handle. ( End of advice to implementors.)
The caching interface defined here represents that attributes be stored by MPI opaquely within a communicator. Accessor functions include the following:

  • obtain a key value (used to identify an attribute); the user specifies ``callback'' functions by which MPI informs the application when the communicator is destroyed or copied.
  • store and retrieve the value of an attribute;


[] Advice to implementors.

Caching and callback functions are only called synchronously, in response to explicit application requests. This avoid problems that result from repeated crossings between user and system space. (This synchronous calling rule is a general property of MPI.)

The choice of key values is under control of MPI. This allows MPI to optimize its implementation of attribute sets. It also avoids conflict between independent modules caching information on the same communicators.

A much smaller interface, consisting of just a callback facility, would allow the entire caching facility to be implemented by portable code. However, with the minimal callback interface, some form of table searching is implied by the need to handle arbitrary communicators. In contrast, the more complete interface defined here permits rapid access to attributes through the use of pointers in communicators (to find the attribute table) and cleverly chosen key values (to retrieve individual attributes). In light of the efficiency ``hit'' inherent in the minimal interface, the more complete interface defined here is seen to be superior. ( End of advice to implementors.)
MPI provides the following services related to caching. They are all process local.

MPI_KEYVAL_CREATE(copy_fn, delete_fn, keyval, extra_state)
[ IN copy_fn] Copy callback function for keyval
[ IN delete_fn] Delete callback function for keyval
[ OUT keyval] key value for future access (integer)
[ IN extra_state] Extra state for callback functions

int MPI_Keyval_create(MPI_Copy_function *copy_fn, MPI_Delete_function *delete_fn, int *keyval, void* extra_state)

MPI_KEYVAL_CREATE(COPY_FN, DELETE_FN, KEYVAL, EXTRA_STATE, IERROR)
EXTERNAL COPY_FN, DELETE_FN
INTEGER KEYVAL, EXTRA_STATE, IERROR

Generates a new attribute key. Keys are locally unique in a process, and opaque to user, though they are explicitly stored in integers. Once allocated, the key value can be used to associate attributes and access them on any locally defined communicator.

The copy_fn function is invoked when a communicator is duplicated by MPI_COMM_DUP. copy_fn should be of type MPI_Copy_function, which is defined as follows:

typedef int MPI_Copy_function(MPI_Comm oldcomm, int keyval, 
                              void *extra_state, void *attribute_val_in, 
                              void *attribute_val_out, int *flag) 

A Fortran declaration for such a function is as follows:

SUBROUTINE COPY_FUNCTION(OLDCOMM, KEYVAL, EXTRA_STATE, ATTRIBUTE_VAL_IN, ATTRIBUTE_VAL_OUT, FLAG, IERR)
INTEGER OLDCOMM, KEYVAL, EXTRA_STATE, ATTRIBUTE_VAL_IN, ATTRIBUTE_VAL_OUT, IERR
LOGICAL FLAG
The copy callback function is invoked for each key value in oldcomm in arbitrary order. Each call to the copy callback is made with a key value and its corresponding attribute. If it returns flag = 0, then the attribute is deleted in the duplicated communicator. Otherwise ( flag = 1), the new attribute value is set to the value returned in attribute_val_out. The function returns MPI_SUCCESS on success and an error code on failure (in which case MPI_COMM_DUP will fail).

copy_fn may be specified as MPI_NULL_COPY_FN or MPI_DUP_FN from either C or FORTRAN; MPI_NULL_COPY_FN is a function that does nothing other than returning flag = 0 and MPI_SUCCESS. MPI_DUP_FN is a simple-minded copy function that sets flag = 1, returns the value of attribute_val_in in attribute_val_out, and returns MPI_SUCCESS.


[] Advice to users.

Even though both formal arguments attribute_val_in and attribute_val_out are of type void *, their usage differs. The C copy function is passed by MPI in attribute_val_in the value of the attribute, and in attribute_val_out the address of the attribute, so as to allow the function to return the (new) attribute value. The use of type void * for both is to avoid messy type casts.

A valid copy function is one that completely duplicates the information by making a full duplicate copy of the data structures implied by an attribute; another might just make another reference to that data structure, while using a reference-count mechanism. Other types of attributes might not copy at all (they might be specific to oldcomm only). ( End of advice to users.)

[] Advice to implementors.

A C interface should be assumed for copy and delete functions associated with key values created in C; a Fortran calling interface should be assumed for key values created in Fortran. ( End of advice to implementors.)

Analogous to copy_fn is a callback deletion function, defined as follows. The delete_fn function is invoked when a communicator is deleted by MPI_COMM_FREE or when a call is made explicitly to MPI_ATTR_DELETE. delete_fn should be of type MPI_Delete_function, which is defined as follows:


typedef int MPI_Delete_function(MPI_Comm comm, int keyval, 
    void *attribute_val, void *extra_state); 
A Fortran declaration for such a function is as follows:

SUBROUTINE DELETE_FUNCTION(COMM, KEYVAL, ATTRIBUTE_VAL, EXTRA_STATE, IERR)
INTEGER COMM, KEYVAL, ATTRIBUTE_VAL, EXTRA_STATE, IERR

This function is called by MPI_COMM_FREE, MPI_ATTR_DELETE, and MPI_ATTR_PUT to do whatever is needed to remove an attribute. The function returns MPI_SUCCESS on success and an error code on failure (in which case MPI_COMM_FREE will fail).

delete_fn may be specified as MPI_NULL_DELETE_FN from either C or FORTRAN; MPI_NULL_DELETE_FN is a function that does nothing, other than returning MPI_SUCCESS.

The special key value MPI_KEYVAL_INVALID is never returned by MPI_KEYVAL_CREATE. Therefore, it can be used for static initialization of key values.

MPI_KEYVAL_FREE(keyval)
[ INOUT keyval] Frees the integer key value (integer)

int MPI_Keyval_free(int *keyval)

MPI_KEYVAL_FREE(KEYVAL, IERROR)
INTEGER KEYVAL, IERROR

Frees an extant attribute key. This function sets the value of keyval to
MPI_KEYVAL_INVALID. Note that it is not erroneous to free an attribute key that is in use, because the actual free does not transpire until after all references (in other communicators on the process) to the key have been freed. These references need to be explictly freed by the program, either via calls to MPI_ATTR_DELETE that free one attribute instance, or by calls to MPI_COMM_FREE that free all attribute instances associated with the freed communicator.

MPI_ATTR_PUT(comm, keyval, attribute_val)
[ IN comm] communicator to which attribute will be attached (handle)
[ IN keyval] key value, as returned by
MPI_KEYVAL_CREATE (integer)
[ IN attribute_val] attribute value

int MPI_Attr_put(MPI_Comm comm, int keyval, void* attribute_val)

MPI_ATTR_PUT(COMM, KEYVAL, ATTRIBUTE_VAL, IERROR)
INTEGER COMM, KEYVAL, ATTRIBUTE_VAL, IERROR

This function stores the stipulated attribute value attribute_val for subsequent retrieval by MPI_ATTR_GET. If the value is already present, then the outcome is as if MPI_ATTR_DELETE was first called to delete the previous value (and the callback function delete_fn was executed), and a new value was next stored. The call is erroneous if there is no key with value keyval; in particular MPI_KEYVAL_INVALID is an erroneous key value. The call will fail if the delete_fn function returned an error code other than MPI_SUCCESS.

MPI_ATTR_GET(comm, keyval, attribute_val, flag)
[ IN comm] communicator to which attribute is attached (handle)
[ IN keyval] key value (integer)
[ OUT attribute_val] attribute value, unless flag = false
[ OUT flag] true if an attribute value was extracted; false if no attribute is associated with the key

int MPI_Attr_get(MPI_Comm comm, int keyval, void *attribute_val, int *flag)

MPI_ATTR_GET(COMM, KEYVAL, ATTRIBUTE_VAL, FLAG, IERROR)
INTEGER COMM, KEYVAL, ATTRIBUTE_VAL, IERROR
LOGICAL FLAG

Retrieves attribute value by key. The call is erroneous if there is no key with value keyval. On the other hand, the call is correct if the key value exists, but no attribute is attached on comm for that key; in such case, the call returns flag = false. In particular MPI_KEYVAL_INVALID is an erroneous key value.


[] Advice to users.

The call to MPI_Attr_put passes in attribute_val the value of the attribute; the call to MPI_Attr_get passes in attribute_val the address of the the location where the attribute value is to be returned. Thus, if the attribute value itself is a pointer of type void*, the the actual attribute_val parameter to MPI_Attr_put will be of type void* and the actual attribute_val parameter to MPI_Attr_put will be of type void**. ( End of advice to users.)

[] Rationale.

The use of a formal parameter attribute_val or type void* (rather than void**) avoids the messy type casting that would be needed if the attribute value is declared with a type other than void*. ( End of rationale.)

MPI_ATTR_DELETE(comm, keyval)
[ IN comm] communicator to which attribute is attached (handle)
[ IN keyval] The key value of the deleted attribute (integer)

int MPI_Attr_delete(MPI_Comm comm, int keyval)

MPI_ATTR_DELETE(COMM, KEYVAL, IERROR)
INTEGER COMM, KEYVAL, IERROR

Delete attribute from cache by key. This function invokes the attribute delete function delete_fn specified when the keyval was created. The call will fail if the delete_fn function returns an error code other than MPI_SUCCESS.

Whenever a communicator is replicated using the function MPI_COMM_DUP, all call-back copy functions for attributes that are currently set are invoked (in arbitrary order). Whenever a communicator is deleted using the function MPI_COMM_FREE all callback delete functions for attributes that are currently set are invoked.



Up: Caching Next: Attributes Example Previous: Caching


Return to MPI Standard Index
Return to MPI home page