rocsparse_sgebsr2gebsr Interface Reference

• hipfort_rocsparse
• rocsparse_sgebsr2gebsr

Public Member Functions | List of all members

hipfort_rocsparse::rocsparse_sgebsr2gebsr Interface Reference

This function converts the general BSR sparse matrix \(A\) to another general BSR sparse matrix \(C\). More...

Public Member Functions
integer(kind(rocsparse_status_success)) function	rocsparse_sgebsr2gebsr_ (handle, dir, mb, nb, nnzb, descr_a, bsr_val_a, bsr_row_ptr_a, bsr_col_ind_a, row_block_dim_a, col_block_dim_a, descr_c, bsr_val_c, bsr_row_ptr_c, bsr_col_ind_c, row_block_dim_c, col_block_dim_c, temp_buffer)

Detailed Description

This function converts the general BSR sparse matrix \(A\) to another general BSR sparse matrix \(C\).

rocsparse_gebsr2gebsr converts a general BSR matrix \(A\) into a general BSR matrix \(C\). The input and output matrices can have different row and column block dimensions. The input matrix \(A\) is assumed to be allocated such that array bsr_row_ptr_A has length mb+1, bsr_col_ind_A has length nnzb, and bsr_val_A has length nnzb*row_block_dim_A*col_block_dim_A. The output matrix \(C\) is assumed to be allocated such that array bsr_row_ptr_C has length mb_C+1, bsr_col_ind_C has length nnzb_C, and bsr_val_C has length nnzb_C*row_block_dim_C*col_block_dim_C where:

\[ m = mb * row\_block\_dim\_A \\% n = nb * col\_block\_dim\_A \]

and

\[ mb\_C = (m + row\_block\_dim\_C - 1) / row\_block\_dim\_C \\% nb\_C = (n + col\_block\_dim\_C - 1) / col\_block\_dim\_C \]

The number of non-zero blocks in the output sparse \(C\) matrix (i.e. nnzb_C) is computed using rocsparse_gebsr2gebsr_nnz() which also fills in the bsr_row_ptr_C array.

Converting from a sparse general BSR matrix to a sparse general BSR matrix requires three steps. First, call rocsparse_Xgebsr2gebsr_buffer_size() to determine the size of the required temporary storage buffer. After this has been determined, allocate this buffer. Also allocate the bsr_row_ptr_C array to have length mb_C+1 and pass it to the function rocsparse_gebsr2gebsr_nnz. This will fill the bsr_row_ptr_C array and also compute the total number of non-zero blocks in the general BSR output \(C\) matrix. Now that the total number of non-zero blocks is known, allocate the bsr_col_ind_C and bsr_val_C arrays. Finally, call rocsparse_gebsr2gebsr to complete the conversion. After the conversion is complete, the temporary storage buffer can be deallocated. See the example below.

Note

This function is blocking with respect to the host.

This routine does not support execution in a hipGraph context.

Parameters

[in]	handle	- handle to the rocSPARSE library context queue.
[in]	dir	- the storage format of the blocks, rocsparse_direction_row or rocsparse_direction_column.
[in]	mb	- number of block rows of the general BSR sparse matrix \(A\).
[in]	nb	- number of block columns of the general BSR sparse matrix \(A\).
[in]	nnzb	- number of blocks in the general BSR sparse matrix \(A\).
[in]	descr_A	- the descriptor of the general BSR sparse matrix \(A\). The supported matrix type is rocsparse_matrix_type_general and also any valid value of the rocsparse_index_base.
[in]	bsr_val_A	- array of nnzb*row_block_dim_A*col_block_dim_A containing the values of the sparse general BSR matrix \(A\).
[in]	bsr_row_ptr_A	- array of mb+1 elements that point to the start of every block row of the sparse general BSR matrix \(A\).
[in]	bsr_col_ind_A	- array of nnzb elements containing the block column indices of the sparse general BSR matrix \(A\).
[in]	row_block_dim_A	- row size of the blocks in the sparse general BSR matrix \(A\).
[in]	col_block_dim_A	- column size of the blocks in the sparse general BSR matrix \(A\).
[in]	descr_C	- the descriptor of the general BSR sparse matrix \(C\). The supported matrix type is rocsparse_matrix_type_general and also any valid value of the rocsparse_index_base.
[in]	bsr_val_C	- array of nnzb_C*row_block_dim_C*col_block_dim_C containing the values of the sparse general BSR matrix \(C\).
[in]	bsr_row_ptr_C	- array of mb_C+1 elements that point to the start of every block row of the sparse general BSR matrix \(C\).
[in]	bsr_col_ind_C	- array of nnzb_C elements containing the block column indices of the sparse general BSR matrix \(C\).
[in]	row_block_dim_C	- row size of the blocks in the sparse general BSR matrix \(C\).
[in]	col_block_dim_C	- column size of the blocks in the sparse general BSR matrix \(C\).
[out]	temp_buffer	- buffer allocated by the user. Its size is determined by calling rocsparse_Xgebsr2gebsr_buffer_size().

Return values

rocsparse_status_success	the operation completed successfully.
rocsparse_status_invalid_handle	the library context was not initialized.
rocsparse_status_invalid_size	mb, nb, nnzb, row_block_dim_A, col_block_dim_A, row_block_dim_C, or col_block_dim_C is invalid.
rocsparse_status_invalid_pointer	bsr_row_ptr_A, bsr_col_ind_A, bsr_val_A, bsr_row_ptr_C, bsr_col_ind_C, bsr_val_C, descr_A, descr_C, or temp_buffer pointer is invalid.

Example

This example converts a general BSR matrix into an general BSR matrix.

Member Function/Subroutine Documentation

◆ rocsparse_sgebsr2gebsr_()

integer(kind(rocsparse_status_success)) function hipfort_rocsparse::rocsparse_sgebsr2gebsr::rocsparse_sgebsr2gebsr_	(	type(c_ptr), value	handle,
		integer(kind(rocsparse_direction_row)), value	dir,
		integer(c_int), value	mb,
		integer(c_int), value	nb,
		integer(c_int), value	nnzb,
		type(c_ptr), value	descr_a,
		type(c_ptr), value	bsr_val_a,
		type(c_ptr), value	bsr_row_ptr_a,
		type(c_ptr), value	bsr_col_ind_a,
		integer(c_int), value	row_block_dim_a,
		integer(c_int), value	col_block_dim_a,
		type(c_ptr), value	descr_c,
		type(c_ptr), value	bsr_val_c,
		type(c_ptr), value	bsr_row_ptr_c,
		type(c_ptr), value	bsr_col_ind_c,
		integer(c_int), value	row_block_dim_c,
		integer(c_int), value	col_block_dim_c,
		type(c_ptr), value	temp_buffer
	)

---

The documentation for this interface was generated from the following file:

• hipfort_rocsparse.F90