It is often convenient to define explicitly the lower bound and upper bound of a type map, and override the definition given by Equation 1 on page 1 . This allows one to define a datatype that has ``holes'' at its beginning or its end, or a datatype with entries that extend above the upper bound or below the lower bound. Examples of such usage are provided in Sec. Examples . To achieve this, we add two additional ``pseudo-datatypes,'' MPI_LB and MPI_UB, that can be used, respectively, to mark the lower bound or the upper bound of a datatype. These pseudo-datatypes occupy no space ( ). They do not affect the size or count of a datatype, and do not affect the the content of a message created with this datatype. However, they do affect the definition of the extent of a datatype and, therefore, affect the outcome of a replication of this datatype by a datatype constructor.
Let D = (-3, 0, 6); T = (MPI_LB, MPI_INT, MPI_UB), and B = (1, 1, 1). Then a call to MPI_TYPE_STRUCT(3, B, D, T, type1) creates a new datatype that has an extent of 9 (from -3 to 5, 5 included), and contains an integer at displacement 0. This is the datatype defined by the sequence {(lb, -3), (int, 0), (ub, 6)} . If this type is replicated twice by a call to MPI_TYPE_CONTIGUOUS(2, type1, type2) then the newly created type can be described by the sequence {(lb, -3), (int, 0), (int,9), (ub, 15)} . (Entries of type lb or ub can be deleted if they are not at the end-points of the datatype.)In general, if
then the lower bound ofTypemap is defined to be
Similarly, the upper bound ofTypemap is defined to be
Then
If requires alignment to a byte address that is a multiple of , then is the least nonnegative increment needed to round extent(Typemap) to the next multiple of .
The formal definitions given for the various datatype constructors apply now, with the amended definition of extent.
The two functions below can be used for finding the lower bound and the upper bound of a datatype.
MPI_TYPE_LB( datatype, displacement)
[ IN datatype] datatype (handle)
[ OUT displacement] displacement of lower bound from origin,
in bytes (integer)
int MPI_Type_lb(MPI_Datatype datatype, int* displacement)
MPI_TYPE_LB( DATATYPE, DISPLACEMENT, IERROR)
INTEGER DATATYPE, DISPLACEMENT, IERROR
MPI_TYPE_UB( datatype, displacement)
[ IN datatype] datatype (handle)
[ OUT displacement] displacement of upper bound from origin,
in bytes (integer)
int MPI_Type_ub(MPI_Datatype datatype, int* displacement)
MPI_TYPE_UB( DATATYPE, DISPLACEMENT, IERROR)
INTEGER DATATYPE, DISPLACEMENT, IERROR
[] Rationale.
Note that the
rules given in Sec. Correct use of addresses
imply that
it is erroneous to call MPI_TYPE_EXTENT,
MPI_TYPE_LB, and MPI_TYPE_UB
with a datatype argument that contains absolute addresses,
unless all these addreses are within the same sequential storage.
For this reason, the displacement for the C binding in
MPI_TYPE_UB is an int and not of type MPI_Aint.
( End of rationale.)