double precision
[GPU kernels for non-symmetric sparse LA]

Functions

magma_int_t magma_dbajac_csr (magma_int_t localiters, magma_d_sparse_matrix D, magma_d_sparse_matrix R, magma_d_vector b, magma_d_vector *x)
 This routine is a block-asynchronous Jacobi iteration performing s local Jacobi-updates within the block.
magma_int_t magma_dbcsrvalcpy (magma_int_t size_b, magma_int_t num_blocks, magma_int_t num_zblocks, double **Aval, double **Bval, double **Bval2)
 For a Block-CSR ILU factorization, this routine copies the filled blocks from the original matrix A and initializes the blocks that will later be filled in the factorization process with zeros.
magma_int_t magma_dbcsrluegemm (magma_int_t size_b, magma_int_t num_brows, magma_int_t kblocks, double **dA, double **dB, double **dC)
 For a Block-CSR ILU factorization, this routine updates all blocks in the trailing matrix.
magma_int_t magma_dbcsrlupivloc (magma_int_t size_b, magma_int_t kblocks, double **dA, magma_int_t *ipiv)
 For a Block-CSR ILU factorization, this routine updates all blocks in the trailing matrix.
magma_int_t magma_dbcsrswp (magma_int_t r_blocks, magma_int_t size_b, magma_int_t *ipiv, double *x)
 For a Block-CSR ILU factorization, this routine swaps rows in the vector *x according to the pivoting in *ipiv.
magma_int_t magma_dbcsrtrsv (magma_uplo_t uplo, magma_int_t r_blocks, magma_int_t c_blocks, magma_int_t size_b, double *A, magma_index_t *blockinfo, double *x)
 For a Block-CSR ILU factorization, this routine performs the triangular solves.
void magma_dcompact (magma_int_t m, magma_int_t n, double *dA, magma_int_t ldda, double *dnorms, double tol, magma_index_t *active, magma_index_t *cBlock)
 ZCOMPACT takes a set of n vectors of size m (in dA) and their norms and compacts them into the cBlock size<=n vectors that have norms > tol.
magma_int_t magma_djacobisetup_vector_gpu (int num_rows, double *b, double *d, double *c, double *x)
 Prepares the Jacobi Iteration according to x^(k+1) = D^(-1) * b - D^(-1) * (L+U) * x^k x^(k+1) = c - M * x^k.
magma_int_t magma_dlobpcg_maxpy (magma_int_t num_rows, magma_int_t num_vecs, double *X, double *Y)
 This routine computes a axpy for a mxn matrix:.
int magma_dbicgmerge1 (int n, double *skp, double *v, double *r, double *p)
 Mergels multiple operations into one kernel:.
int magma_dbicgmerge2 (int n, double *skp, double *r, double *v, double *s)
 Mergels multiple operations into one kernel:.
int magma_dbicgmerge3 (int n, double *skp, double *p, double *s, double *t, double *x, double *r)
 Mergels multiple operations into one kernel:.
int magma_dbicgmerge4 (int type, double *skp)
 Performs some parameter operations for the BiCGSTAB with scalars on GPU.
magma_int_t magma_dbicgmerge_spmv1 (magma_d_sparse_matrix A, double *d1, double *d2, double *d_p, double *d_r, double *d_v, double *skp)
 Merges the first SpmV using CSR with the dot product and the computation of alpha.
magma_int_t magma_dbicgmerge_spmv2 (magma_d_sparse_matrix A, double *d1, double *d2, double *d_s, double *d_t, double *skp)
 Merges the second SpmV using CSR with the dot product and the computation of omega.
magma_int_t magma_dbicgmerge_xrbeta (int n, double *d1, double *d2, double *rr, double *r, double *p, double *s, double *t, double *x, double *skp)
 Merges the second SpmV using CSR with the dot product and the computation of omega.
magma_int_t magma_dcgmerge_spmv1 (magma_d_sparse_matrix A, double *d1, double *d2, double *d_d, double *d_z, double *skp)
 Merges the first SpmV using different formats with the dot product and the computation of rho.

Function Documentation

magma_int_t magma_dbajac_csr ( magma_int_t  localiters,
magma_d_sparse_matrix  D,
magma_d_sparse_matrix  R,
magma_d_vector  b,
magma_d_vector *  x 
)

This routine is a block-asynchronous Jacobi iteration performing s local Jacobi-updates within the block.

Input format is two CSR matrices, one containing the diagonal blocks, one containing the rest.

Parameters:
localiters magma_int_t number of local Jacobi-like updates
D magma_d_sparse_matrix input matrix with diagonal blocks
R magma_d_sparse_matrix input matrix with non-diagonal parts
b magma_d_vector RHS
x magma_d_vector* iterate/solution
magma_int_t magma_dbcsrluegemm ( magma_int_t  size_b,
magma_int_t  num_brows,
magma_int_t  kblocks,
double **  dA,
double **  dB,
double **  dC 
)

For a Block-CSR ILU factorization, this routine updates all blocks in the trailing matrix.

Parameters:
size_b magma_int_t blocksize in BCSR
num_brows magma_int_t number of block rows
kblocks magma_int_t number of blocks in row
dA double** input blocks of matrix A
dB double** input blocks of matrix B
dC double** output blocks of matrix C
magma_int_t magma_dbcsrlupivloc ( magma_int_t  size_b,
magma_int_t  kblocks,
double **  dA,
magma_int_t *  ipiv 
)

For a Block-CSR ILU factorization, this routine updates all blocks in the trailing matrix.

Parameters:
size_b magma_int_t blocksize in BCSR
kblocks magma_int_t number of blocks
dA double** matrix in BCSR
ipiv magma_int_t* array containing pivots
magma_int_t magma_dbcsrswp ( magma_int_t  r_blocks,
magma_int_t  size_b,
magma_int_t *  ipiv,
double *  x 
)

For a Block-CSR ILU factorization, this routine swaps rows in the vector *x according to the pivoting in *ipiv.

Parameters:
r_blocks magma_int_t number of blocks
size_b magma_int_t blocksize in BCSR
ipiv magma_int_t* array containing pivots
x double* input/output vector x
magma_int_t magma_dbcsrtrsv ( magma_uplo_t  uplo,
magma_int_t  r_blocks,
magma_int_t  c_blocks,
magma_int_t  size_b,
double *  A,
magma_index_t *  blockinfo,
double *  x 
)

For a Block-CSR ILU factorization, this routine performs the triangular solves.

Parameters:
uplo magma_uplo_t upper/lower fill structure
r_blocks magma_int_t number of blocks in row
c_blocks magma_int_t number of blocks in column
size_b magma_int_t blocksize in BCSR
A double* upper/lower factor
blockinfo magma_int_t* array containing matrix information
x double* input/output vector x
magma_int_t magma_dbcsrvalcpy ( magma_int_t  size_b,
magma_int_t  num_blocks,
magma_int_t  num_zblocks,
double **  Aval,
double **  Bval,
double **  Bval2 
)

For a Block-CSR ILU factorization, this routine copies the filled blocks from the original matrix A and initializes the blocks that will later be filled in the factorization process with zeros.

Parameters:
size_b magma_int_t blocksize in BCSR
num_blocks magma_int_t number of nonzero blocks
num_zblocks magma_int_t number of zero-blocks (will later be filled)
Aval double** pointers to the nonzero blocks in A
Bval double** pointers to the nonzero blocks in B
Bval2 double** pointers to the zero blocks in B
int magma_dbicgmerge1 ( int  n,
double *  skp,
double *  v,
double *  r,
double *  p 
)

Mergels multiple operations into one kernel:.

p = beta*p p = p-omega*beta*v p = p+r

-> p = r + beta * ( p - omega * v )

Parameters:
n int dimension n
skp double* set of scalar parameters
v double* input v
r double* input r
p double* input/output p
int magma_dbicgmerge2 ( int  n,
double *  skp,
double *  r,
double *  v,
double *  s 
)

Mergels multiple operations into one kernel:.

s=r s=s-alpha*v

-> s = r - alpha * v

Parameters:
n int dimension n
skp double* set of scalar parameters
r double* input r
v double* input v
s double* input/output s
int magma_dbicgmerge3 ( int  n,
double *  skp,
double *  p,
double *  s,
double *  t,
double *  x,
double *  r 
)

Mergels multiple operations into one kernel:.

x=x+alpha*p x=x+omega*s r=s r=r-omega*t

-> x = x + alpha * p + omega * s -> r = s - omega * t

Parameters:
n int dimension n
skp double* set of scalar parameters
p double* input p
s double* input s
t double* input t
x double* input/output x
r double* input/output r
int magma_dbicgmerge4 ( int  type,
double *  skp 
)

Performs some parameter operations for the BiCGSTAB with scalars on GPU.

Parameters:
type int kernel type
skp double* vector with parameters
magma_int_t magma_dbicgmerge_spmv1 ( magma_d_sparse_matrix  A,
double *  d1,
double *  d2,
double *  d_p,
double *  d_r,
double *  d_v,
double *  skp 
)

Merges the first SpmV using CSR with the dot product and the computation of alpha.

Parameters:
A magma_d_sparse_matrix system matrix
d1 double* temporary vector
d2 double* temporary vector
d_p double* input vector p
d_r double* input vector r
d_v double* output vector v
skp double* array for parameters ( skp[0]=alpha )
magma_int_t magma_dbicgmerge_spmv2 ( magma_d_sparse_matrix  A,
double *  d1,
double *  d2,
double *  d_s,
double *  d_t,
double *  skp 
)

Merges the second SpmV using CSR with the dot product and the computation of omega.

Parameters:
A magma_d_sparse_matrix input matrix
d1 double* temporary vector
d2 double* temporary vector
d_s double* input vector s
d_t double* output vector t
skp double* array for parameters
magma_int_t magma_dbicgmerge_xrbeta ( int  n,
double *  d1,
double *  d2,
double *  rr,
double *  r,
double *  p,
double *  s,
double *  t,
double *  x,
double *  skp 
)

Merges the second SpmV using CSR with the dot product and the computation of omega.

Parameters:
n int dimension n
d1 double* temporary vector
d2 double* temporary vector
rr double* input vector rr
r double* input/output vector r
p double* input vector p
s double* input vector s
t double* input vector t
x double* output vector x
skp double* array for parameters
magma_int_t magma_dcgmerge_spmv1 ( magma_d_sparse_matrix  A,
double *  d1,
double *  d2,
double *  d_d,
double *  d_z,
double *  skp 
)

Merges the first SpmV using different formats with the dot product and the computation of rho.

Parameters:
A magma_d_sparse_matrix input matrix
d1 double* temporary vector
d2 double* temporary vector
d_d double* input vector d
d_z double* input vector z
skp double* array for parameters ( skp[3]=rho )
void magma_dcompact ( magma_int_t  m,
magma_int_t  n,
double *  dA,
magma_int_t  ldda,
double *  dnorms,
double  tol,
magma_index_t *  active,
magma_index_t *  cBlock 
)

ZCOMPACT takes a set of n vectors of size m (in dA) and their norms and compacts them into the cBlock size<=n vectors that have norms > tol.

The active mask array has 1 or 0, showing if a vector remained or not in the compacted resulting set of vectors.

Parameters:
[in] m INTEGER The number of rows of the matrix dA. M >= 0.
[in] n INTEGER The number of columns of the matrix dA. N >= 0.
[in,out] dA COMPLEX DOUBLE PRECISION array, dimension (LDDA,N) The m by n matrix dA.
[in] ldda INTEGER The leading dimension of the array dA. LDDA >= max(1,M).
[in] dnorms DOUBLE PRECISION array, dimension N The norms of the N vectors in dA
[in] tol DOUBLE PRECISON The tolerance value used in the criteria to compact or not.
[out] active INTEGER array, dimension N A mask of 1s and 0s showing if a vector remains or has been removed
[out] cBlock magma_index_t* The number of vectors that remain in dA (i.e., with norms > tol).
magma_int_t magma_djacobisetup_vector_gpu ( int  num_rows,
double *  b,
double *  d,
double *  c,
double *  x 
)

Prepares the Jacobi Iteration according to x^(k+1) = D^(-1) * b - D^(-1) * (L+U) * x^k x^(k+1) = c - M * x^k.

Returns the vector c. It calls a GPU kernel

Parameters:
num_rows magma_int_t number of rows
b magma_d_vector RHS b
d magma_d_vector vector with diagonal entries
c magma_d_vector* c = D^(-1) * b
x magma_d_vector* iteration vector
magma_int_t magma_dlobpcg_maxpy ( magma_int_t  num_rows,
magma_int_t  num_vecs,
double *  X,
double *  Y 
)

This routine computes a axpy for a mxn matrix:.

Y = X + Y

It replaces: magma_daxpy(m*n, c_one, Y, 1, X, 1);

/ x1[0] x2[0] x3[0] \ | x1[1] x2[1] x3[1] | X = | x1[2] x2[2] x3[2] | = x1[0] x1[1] x1[2] x1[3] x1[4] x2[0] x2[1] . | x1[3] x2[3] x3[3] | \ x1[4] x2[4] x3[4] /

Parameters:
num_rows magma_int_t number of rows
num_vecs magma_int_t number of vectors
X double* input vector X
Y double* input/output vector Y

Generated on 17 Sep 2014 for MAGMA by  doxygen 1.6.1