PLASMA
2.8.0
PLASMA - Parallel Linear Algebra for Scalable Multi-core Architectures
|
int PLASMA_dgetmi | ( | int | m, |
int | n, | ||
double * | A, | ||
PLASMA_enum | f_in, | ||
int | imb, | ||
int | inb | ||
) |
PLASMA_dgetmi Implementation of inplace transposition based on the GKK algorithm by Gustavson, Karlsson, Kagstrom. This algorithm shift some cycles to transpose the matrix.
[in] | m | Number of rows of matrix A |
[in] | n | Number of columns of matrix A |
[in,out] | A | Matrix of size L*m*n. |
[in] | f_in | Original layout of the matrix A. Must be part of: PlasmaCM: if A is stored in Column Major, PlasmaRM: if A is stored in Row Major, PlasmaCCRB: if A is stored in Column Column Rectangular Block layout, PlasmaCRRB: if A is stored in Column Row Rectangular Block layout, PlasmaRCRB: if A is stored in Row Column Rectangular Block layout, PlasmaRRRB: if A is stored in Row Row Rectangular Block layout; where the first Column/Row stand for the order in which tiles are sorted, and the secon Column/Row stand for the layout inside each tile. |
[in] | imb | Number of rows of the problem |
[in] | inb | Number of columns in the problem |
PLASMA_SUCCESS | successful exit |
<0 | if -i, the i-th argument had an illegal value |