diff --git a/compute/pzlange.c b/compute/pzlange.c
index b5f58e47d25fc1c231faefa374e85b9b8e250b79..f909fa25fa34ff17bc6d0c72b811916a187d284e 100644
--- a/compute/pzlange.c
+++ b/compute/pzlange.c
@@ -329,13 +329,14 @@ chameleon_pzlange_frb( cham_uplo_t uplo, cham_diag_t diag, CHAM_desc_t *A, CHAM_
else {
INSERT_TASK_zgessq(
options,
+ ChamEltwise,
tempmm, tempnn,
A(m, n), ldam, W( Welt, m, n) );
}
if ( n >= Q ) {
INSERT_TASK_dplssq(
- options, W( Welt, m, n), W( Welt, m, n%Q) );
+ options, ChamEltwise, 1, 1, W( Welt, m, n), W( Welt, m, n%Q) );
}
}
@@ -345,7 +346,7 @@ chameleon_pzlange_frb( cham_uplo_t uplo, cham_diag_t diag, CHAM_desc_t *A, CHAM_
*/
for(n = 1; n < Q; n++) {
INSERT_TASK_dplssq(
- options, W( Welt, m, n), W( Welt, m, 0) );
+ options, ChamEltwise, 1, 1, W( Welt, m, n), W( Welt, m, 0) );
}
}
@@ -355,7 +356,7 @@ chameleon_pzlange_frb( cham_uplo_t uplo, cham_diag_t diag, CHAM_desc_t *A, CHAM_
*/
for(m = P; m < MT; m++) {
INSERT_TASK_dplssq(
- options, W( Welt, m, 0), W( Welt, m%P, 0) );
+ options, ChamEltwise, 1, 1, W( Welt, m, 0), W( Welt, m%P, 0) );
}
/**
@@ -364,11 +365,11 @@ chameleon_pzlange_frb( cham_uplo_t uplo, cham_diag_t diag, CHAM_desc_t *A, CHAM_
*/
for(m = 1; m < P; m++) {
INSERT_TASK_dplssq(
- options, W( Welt, m, 0), W( Welt, 0, 0) );
+ options, ChamEltwise, 1, 1, W( Welt, m, 0), W( Welt, 0, 0) );
}
INSERT_TASK_dplssq2(
- options, W( Welt, 0, 0) );
+ options, 1, W( Welt, 0, 0) );
}
/**
diff --git a/compute/pzlansy.c b/compute/pzlansy.c
index 71189ef578ffa2fd5f5b3d3feac28508d7bb86c5..e799f57285430e09eccd348abdca54d0c31d6cdc 100644
--- a/compute/pzlansy.c
+++ b/compute/pzlansy.c
@@ -241,21 +241,21 @@ chameleon_pzlansy_frb( cham_trans_t trans, cham_uplo_t uplo,
if ( n == m ) {
if ( trans == ChamConjTrans) {
INSERT_TASK_zhessq(
- options, uplo, tempmm,
+ options, ChamEltwise, uplo, tempmm,
A(m, n), ldam, W( Welt, m, n) );
}
else {
INSERT_TASK_zsyssq(
- options, uplo, tempmm,
+ options, ChamEltwise, uplo, tempmm,
A(m, n), ldam, W( Welt, m, n) );
}
}
else {
INSERT_TASK_zgessq(
- options, tempmm, tempnn,
+ options, ChamEltwise, tempmm, tempnn,
A(m, n), ldam, W( Welt, m, n) );
INSERT_TASK_zgessq(
- options, tempmm, tempnn,
+ options, ChamEltwise, tempmm, tempnn,
A(m, n), ldam, W( Welt, n, m) );
}
}
@@ -264,7 +264,7 @@ chameleon_pzlansy_frb( cham_trans_t trans, cham_uplo_t uplo,
for(m = 0; m < MT; m++) {
for(n = Q; n < NT; n++) {
INSERT_TASK_dplssq(
- options, W( Welt, m, n), W( Welt, m, n%Q) );
+ options, ChamEltwise, 1, 1, W( Welt, m, n), W( Welt, m, n%Q) );
}
/**
@@ -273,7 +273,7 @@ chameleon_pzlansy_frb( cham_trans_t trans, cham_uplo_t uplo,
*/
for(n = 1; n < Q; n++) {
INSERT_TASK_dplssq(
- options, W( Welt, m, n), W( Welt, m, 0) );
+ options, ChamEltwise, 1, 1, W( Welt, m, n), W( Welt, m, 0) );
}
}
@@ -283,7 +283,7 @@ chameleon_pzlansy_frb( cham_trans_t trans, cham_uplo_t uplo,
*/
for(m = P; m < MT; m++) {
INSERT_TASK_dplssq(
- options, W( Welt, m, 0), W( Welt, m%P, 0) );
+ options, ChamEltwise, 1, 1, W( Welt, m, 0), W( Welt, m%P, 0) );
}
/**
@@ -292,11 +292,11 @@ chameleon_pzlansy_frb( cham_trans_t trans, cham_uplo_t uplo,
*/
for(m = 1; m < P; m++) {
INSERT_TASK_dplssq(
- options, W( Welt, m, 0), W( Welt, 0, 0) );
+ options, ChamEltwise, 1, 1, W( Welt, m, 0), W( Welt, 0, 0) );
}
INSERT_TASK_dplssq2(
- options, W( Welt, 0, 0) );
+ options, 1, W( Welt, 0, 0) );
}
/**
diff --git a/coreblas/compute/CMakeLists.txt b/coreblas/compute/CMakeLists.txt
index 009cd37dbb5112325e7622b5eef71715d661fde1..d44a83fd7a2480d0c5197f4b4d4477aebfa4be16 100644
--- a/coreblas/compute/CMakeLists.txt
+++ b/coreblas/compute/CMakeLists.txt
@@ -66,6 +66,7 @@ set(ZSRC
core_zplghe.c
core_zplgsy.c
core_zplrnt.c
+ core_zplssq.c
core_zpotrf.c
core_zssssm.c
core_zsymm.c
diff --git a/coreblas/compute/core_zgessq.c b/coreblas/compute/core_zgessq.c
index aba33e5f5102baf1632332be7cbbc2d97d2259c5..71aa407bdebde121a438604a3d08a9145b802e2f 100644
--- a/coreblas/compute/core_zgessq.c
+++ b/coreblas/compute/core_zgessq.c
@@ -15,23 +15,98 @@
* @comment This file has been automatically generated
* from Plasma 2.6.0 for CHAMELEON 0.9.2
* @author Mathieu Faverge
+ * @author Florent Pruvost
* @date 2014-11-16
* @precisions normal z -> c d s
*
*/
#include <math.h>
#include "coreblas/lapacke.h"
+#include "coreblas/sumsq_update.h"
#include "coreblas.h"
-#define UPDATE( __nb, __value ) \
- if (__value != 0. ){ \
- if ( *scale < __value ) { \
- *sumsq = __nb + (*sumsq) * ( *scale / __value ) * ( *scale / __value ); \
- *scale = __value; \
- } else { \
- *sumsq = *sumsq + __nb * ( __value / *scale ) * ( __value / *scale ); \
- } \
+/**
+ * @brief Subcase storev == ChamColumnwise of CORE_zgessq()
+ */
+static inline int
+CORE_zgessq_col( int M, int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
+{
+ int i, j;
+ double tmp;
+ double *ptr, *tmpScale, *tmpSumsq;
+
+ for(j=0; j<N; j++) {
+ ptr = (double*) ( A + j * LDA );
+ tmpScale = sclssq+2*j;
+ tmpSumsq = sclssq+2*j+1;
+ for(i=0; i<M; i++, ptr++) {
+ tmp = fabs(*ptr);
+ sumsq_update( 1., tmpScale, tmpSumsq, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., tmpScale, tmpSumsq, &tmp );
+#endif
+ }
}
+ return CHAMELEON_SUCCESS;
+}
+/**
+ * @brief Subcase storev == ChamRowwise of CORE_zgessq()
+ */
+static inline int
+CORE_zgessq_row( int M, int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
+{
+ int i, j;
+ double tmp;
+ double *ptr, *tmpScale, *tmpSumsq;
+
+ for(j=0; j<N; j++) {
+ ptr = (double*) ( A + j * LDA );
+ tmpScale = sclssq;
+ tmpSumsq = sclssq+1;
+ for(i=0; i<M; i++, ptr++, tmpScale+=2, tmpSumsq+=2) {
+ tmp = fabs(*ptr);
+ sumsq_update( 1., tmpScale, tmpSumsq, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., tmpScale, tmpSumsq, &tmp );
+#endif
+ }
+ }
+ return CHAMELEON_SUCCESS;
+}
+/**
+ * @brief Subcase storev == ChamEltwise of CORE_zgessq()
+ */
+static inline int
+CORE_zgessq_elt( int M, int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
+{
+ int i, j;
+ double tmp;
+ double *ptr;
+
+ for(j=0; j<N; j++) {
+ ptr = (double*) ( A + j * LDA );
+ for(i=0; i<M; i++, ptr++) {
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq, sclssq+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq, sclssq+1, &tmp );
+#endif
+ }
+ }
+ return CHAMELEON_SUCCESS;
+}
/**
*
@@ -60,6 +135,12 @@
*
*******************************************************************************
*
+ * @param[in] storev
+ * Specifies whether the sums are made per column or row.
+ * = ChamColumnwise: Computes the sum of squares on each column
+ * = ChamRowwise: Computes the sum of squares on each row
+ * = ChamEltwise: Computes the sum of squares on all the matrix
+ *
* @param[in] M
* The number of rows in the tile A.
*
@@ -72,13 +153,16 @@
* @param[in] LDA
* The leading dimension of the tile A. LDA >= max(1,M).
*
- * @param[in,out] scale
- * On entry, the value scale in the equation above.
- * On exit, scale is overwritten with the value scl.
- *
- * @param[in,out] sumsq
- * On entry, the value sumsq in the equation above.
- * On exit, SUMSQ is overwritten with the value ssq.
+ * @param[in,out] sclssq
+ * Dimension: (2,K)
+ * if storev == ChamColumnwise, K = N
+ * if storev == ChamRowwise, K = M
+ * if storev == ChamEltwise, K = 1
+ * On entry, sclssq contains K couples (sclssq[2*i], sclssq[2*i+1])
+ * which corresponds to (scale, sumsq) in the equation below
+ * ( scl**2 )*ssq = sum( A( i, j )**2 ) + ( scale**2 )*sumsq,
+ * respectively for the columns, the rows and the full matrix
+ * On exit, each couple is overwritten with the final result (scl, ssq).
*
*******************************************************************************
*
@@ -86,26 +170,18 @@
* @retval -k, the k-th argument had an illegal value
*
*/
-int CORE_zgessq(int M, int N,
- const CHAMELEON_Complex64_t *A, int LDA,
- double *scale, double *sumsq)
+int CORE_zgessq( cham_store_t storev, int M, int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
{
- int i, j;
- double tmp;
- double *ptr;
-
- for(j=0; j<N; j++) {
- ptr = (double*) ( A + j * LDA );
- for(i=0; i<M; i++, ptr++) {
- tmp = fabs(*ptr);
- UPDATE( 1., tmp );
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- ptr++;
- tmp = fabs(*ptr);
- UPDATE( 1., tmp );
-#endif
- }
+ if (storev == ChamColumnwise) {
+ CORE_zgessq_col( M, N, A, LDA, sclssq );
+ }
+ else if (storev == ChamRowwise) {
+ CORE_zgessq_row( M, N, A, LDA, sclssq );
+ }
+ else {
+ CORE_zgessq_elt( M, N, A, LDA, sclssq );
}
return CHAMELEON_SUCCESS;
}
diff --git a/coreblas/compute/core_zhessq.c b/coreblas/compute/core_zhessq.c
index 482cc6fb42d14acf3f945801f645e734001ce317..255e7e7f716f6b2cd3475c11aefac373814a2ae8 100644
--- a/coreblas/compute/core_zhessq.c
+++ b/coreblas/compute/core_zhessq.c
@@ -19,20 +19,8 @@
* @precisions normal z -> c
*
*/
-#include <math.h>
-#include "coreblas/lapacke.h"
#include "coreblas.h"
-#define UPDATE( __nb, __value ) \
- if (__value != 0. ){ \
- if ( *scale < __value ) { \
- *sumsq = __nb + (*sumsq) * ( *scale / __value ) * ( *scale / __value ); \
- *scale = __value; \
- } else { \
- *sumsq = *sumsq + __nb * ( __value / *scale ) * ( __value / *scale ); \
- } \
- }
-
/**
*
* @ingroup CORE_CHAMELEON_Complex64_t
@@ -56,18 +44,24 @@
* SCALE and SUMSQ are overwritten by scl and ssq respectively.
*
* The routine makes only one pass through the tile triangular part of the
- * hermitian tile A defined by uplo.
+ * symmetric tile A defined by uplo.
* See also LAPACK zlassq.f
*
*******************************************************************************
*
+ * @param[in] storev
+ * Specifies whether the sums are made per column or row.
+ * = ChamColumnwise: Computes the sum of squares on each column
+ * = ChamRowwise: Computes the sum of squares on each row
+ * = ChamEltwise: Computes the sum of squares on all the matrix
+ *
* @param[in] uplo
* Specifies whether the upper or lower triangular part of
- * the hermitian matrix A is to be referenced as follows:
+ * the symmetric matrix A is to be referenced as follows:
* = ChamLower: Only the lower triangular part of the
- * hermitian matrix A is to be referenced.
+ * symmetric matrix A is to be referenced.
* = ChamUpper: Only the upper triangular part of the
- * hermitian matrix A is to be referenced.
+ * symmetric matrix A is to be referenced.
*
* @param[in] N
* The number of columns and rows in the tile A.
@@ -78,13 +72,16 @@
* @param[in] LDA
* The leading dimension of the tile A. LDA >= max(1,N).
*
- * @param[in,out] scale
- * On entry, the value scale in the equation above.
- * On exit, scale is overwritten with the value scl.
- *
- * @param[in,out] sumsq
- * On entry, the value sumsq in the equation above.
- * On exit, SUMSQ is overwritten with the value ssq.
+ * @param[in,out] sclssq
+ * Dimension: (2,K)
+ * if storev == ChamColumnwise, K = N
+ * if storev == ChamRowwise, K = N
+ * if storev == ChamEltwise, K = 1
+ * On entry, sclssq contains K couples (sclssq[2*i], sclssq[2*i+1])
+ * which corresponds to (scale, sumsq) in the equation below
+ * ( scl**2 )*ssq = sum( A( i, j )**2 ) + ( scale**2 )*sumsq,
+ * respectively for the columns, the rows and the full matrix
+ * On exit, each couple is overwritten with the final result (scl, ssq).
*
*******************************************************************************
*
@@ -92,65 +89,9 @@
* @retval -k, the k-th argument had an illegal value
*
*/
-
-int CORE_zhessq(cham_uplo_t uplo, int N,
- const CHAMELEON_Complex64_t *A, int LDA,
- double *scale, double *sumsq)
+int CORE_zhessq( cham_store_t storev, cham_uplo_t uplo, int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
{
- int i, j;
- double tmp;
- double *ptr;
-
- if ( uplo == ChamUpper ) {
- for(j=0; j<N; j++) {
- ptr = (double*) ( A + j * LDA );
-
- for(i=0; i<j; i++, ptr++) {
-
- tmp = fabs(*ptr);
- UPDATE( 2., tmp );
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- ptr++;
- tmp = fabs(*ptr);
- UPDATE( 2., tmp );
-#endif
- }
-
- /* Diagonal */
- tmp = fabs(*ptr);
- UPDATE( 1., tmp );
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- ptr++;
-#endif
- }
- } else {
-
- for(j=0; j<N; j++) {
- ptr = (double*) ( A + j * LDA + j);
-
- /* Diagonal */
- tmp = fabs(*ptr);
- UPDATE( 1., tmp );
- ptr++;
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- ptr++;
-#endif
-
- for(i=j+1; i<N; i++, ptr++) {
-
- tmp = fabs(*ptr);
- UPDATE( 2., tmp );
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- ptr++;
- tmp = fabs(*ptr);
- UPDATE( 2., tmp );
-#endif
- }
- }
- }
- return CHAMELEON_SUCCESS;
+ return CORE_zsyssq( storev, uplo, N, A, LDA, sclssq );
}
diff --git a/coreblas/compute/core_zplssq.c b/coreblas/compute/core_zplssq.c
new file mode 100644
index 0000000000000000000000000000000000000000..9549453dcd50df72fc2780eadc80d81ceeac9a83
--- /dev/null
+++ b/coreblas/compute/core_zplssq.c
@@ -0,0 +1,193 @@
+/**
+ *
+ * @file core_zplssq.c
+ *
+ * @copyright 2009-2014 The University of Tennessee and The University of
+ * Tennessee Research Foundation. All rights reserved.
+ * @copyright 2012-2019 Bordeaux INP, CNRS (LaBRI UMR 5800), Inria,
+ * Univ. Bordeaux. All rights reserved.
+ *
+ ***
+ *
+ * @brief Chameleon core_zplssq CPU kernel
+ *
+ * @version 0.9.2
+ * @author Mathieu Faverge
+ * @author Florent Pruvost
+ * @date 2019-04-01
+ * @precisions normal z -> c d s
+ *
+ */
+#include <math.h>
+#include "coreblas/lapacke.h"
+#include "coreblas/sumsq_update.h"
+#include "coreblas.h"
+
+/**
+ * @brief Subcase storev == ChamColumnwise of CORE_zplssq()
+ */
+static inline int
+CORE_zplssq_col( int M, int N,
+ double *sclssqin,
+ double *sclssqout )
+{
+ int i, j;
+ double *tmpScaleIn, *tmpSumsqIn, *tmpScaleOut, *tmpSumsqOut;
+
+ tmpScaleIn = sclssqin;
+ tmpSumsqIn = sclssqin+1;
+ tmpScaleOut = sclssqout;
+ tmpSumsqOut = sclssqout+1;
+
+ for(j=0; j<N; j++) {
+ for(i=0; i<M; i++) {
+ sumsq_update_2( 1., tmpScaleIn, tmpSumsqIn, tmpScaleOut, tmpSumsqOut );
+ tmpScaleIn+=2;
+ tmpSumsqIn+=2;
+ }
+ tmpScaleOut+=2;
+ tmpSumsqOut+=2;
+ }
+ return CHAMELEON_SUCCESS;
+}
+/**
+ * @brief Subcase storev == ChamRowwise of CORE_zplssq()
+ */
+static inline int
+CORE_zplssq_row( int M, int N,
+ double *sclssqin,
+ double *sclssqout )
+{
+ int i, j;
+ double *tmpScaleIn, *tmpSumsqIn, *tmpScaleOut, *tmpSumsqOut;
+
+ tmpScaleIn = sclssqin;
+ tmpSumsqIn = sclssqin+1;
+
+ for(j=0; j<N; j++) {
+ tmpScaleOut = sclssqout;
+ tmpSumsqOut = sclssqout+1;
+ for(i=0; i<M; i++) {
+ sumsq_update_2( 1., tmpScaleIn, tmpSumsqIn, tmpScaleOut, tmpSumsqOut );
+ tmpScaleIn+=2;
+ tmpSumsqIn+=2;
+ tmpScaleOut+=2;
+ tmpSumsqOut+=2;
+ }
+ }
+ return CHAMELEON_SUCCESS;
+}
+/**
+ * @brief Subcase storev == ChamEltwise of CORE_zplssq()
+ */
+static inline int
+CORE_zplssq_elt( int M, int N,
+ double *sclssqin,
+ double *sclssqout )
+{
+ int i, j;
+ double *tmpScaleIn, *tmpSumsqIn, *tmpScaleOut, *tmpSumsqOut;
+
+ tmpScaleIn = sclssqin;
+ tmpSumsqIn = sclssqin+1;
+ tmpScaleOut = sclssqout;
+ tmpSumsqOut = sclssqout+1;
+
+ for(j=0; j<N; j++) {
+ for(i=0; i<M; i++) {
+ sumsq_update_2( 1., tmpScaleIn, tmpSumsqIn, tmpScaleOut, tmpSumsqOut );
+ tmpScaleIn+=2;
+ tmpSumsqIn+=2;
+ }
+ }
+ return CHAMELEON_SUCCESS;
+}
+
+/**
+ *
+ * @ingroup CORE_CHAMELEON_Complex64_t
+ *
+ * CORE_zplssq adds the values ssq ensuring scl >= 0
+ *
+ *******************************************************************************
+ *
+ * @param[in] storev
+ * Specifies whether the sums are made per column or row.
+ * = ChamColumnwise: Computes the sum of squares on each column
+ * = ChamRowwise: Computes the sum of squares on each row
+ * = ChamEltwise: Computes the sum of squares on all the matrix
+ *
+ * @param[in] M
+ * The number of rows in the tile sclssqin.
+ *
+ * @param[in] N
+ * The number of columns in the tile sclssqin.
+ *
+ * @param[in] sclssqin
+ * The 2*M-by-2*N matrix on which to compute the sums.
+ *
+ * @param[in,out] sclssqout
+ * Dimension: (2,K)
+ * if storev == ChamColumnwise, K = N
+ * if storev == ChamRowwise, K = M
+ * if storev == ChamEltwise, K = 1
+ * On entry, sclssqout contains M-by-N couples (sclssqout[2*i], sclssqout[2*i+1])
+ * which corresponds to (scale, sumsq) in the equation below
+ * ( scl**2 )*ssq = sum( A( i, j )**2 ) + ( scale**2 )*sumsq,
+ * respectively for the columns, the rows and the full matrix
+ * On exit, each couple is overwritten with the final result (scl, ssq).
+ *
+ *******************************************************************************
+ *
+ * @retval CHAMELEON_SUCCESS successful exit
+ * @retval -k, the k-th argument had an illegal value
+ *
+ */
+int CORE_zplssq( cham_store_t storev, int M, int N,
+ double *sclssqin,
+ double *sclssqout )
+{
+ if (storev == ChamColumnwise) {
+ CORE_zplssq_col( M, N, sclssqin, sclssqout );
+ }
+ else if (storev == ChamRowwise) {
+ CORE_zplssq_row( M, N, sclssqin, sclssqout );
+ }
+ else {
+ CORE_zplssq_elt( M, N, sclssqin, sclssqout );
+ }
+ return CHAMELEON_SUCCESS;
+}
+
+/**
+ *
+ * @ingroup CORE_CHAMELEON_Complex64_t
+ *
+ * CORE_zplssq2 computes scl*sqrt(ssq) for each couple in a vector
+ *
+ *******************************************************************************
+ *
+ * @param[in] N
+ * The number of columns in the tile sclssq.
+ *
+ * @param[in,out] sclssq
+ * The 2*N matrix on which to compute scl*sqrt(ssq)
+ * On entry contains all couple (scl,ssq) in (sclssq[i],sclssq[i+1])
+ * On exist returns sclssq[i]=sclssq[i]*sqrt(sclssq[i+1])
+ * so that the result is stored in the first line.
+ *
+ *******************************************************************************
+ *
+ * @retval CHAMELEON_SUCCESS successful exit
+ * @retval -k, the k-th argument had an illegal value
+ *
+ */
+int CORE_zplssq2( int N,
+ double *sclssq )
+{
+ int i;
+ for (i=0; i<N; i+=2) {
+ sclssq[i] *= sqrt ( sclssq[i+1] );
+ }
+ return CHAMELEON_SUCCESS;
+}
diff --git a/coreblas/compute/core_zsyssq.c b/coreblas/compute/core_zsyssq.c
index c7abf5a71d25a37c353d0bcb57409c694d89c63a..dfad10e3e6d90b61e737c6a283452c791c37bea6 100644
--- a/coreblas/compute/core_zsyssq.c
+++ b/coreblas/compute/core_zsyssq.c
@@ -21,17 +21,155 @@
*/
#include <math.h>
#include "coreblas/lapacke.h"
+#include "coreblas/sumsq_update.h"
#include "coreblas.h"
-#define UPDATE( __nb, __value ) \
- if (__value != 0. ){ \
- if ( *scale < __value ) { \
- *sumsq = __nb + (*sumsq) * ( *scale / __value ) * ( *scale / __value ); \
- *scale = __value; \
- } else { \
- *sumsq = *sumsq + __nb * ( __value / *scale ) * ( __value / *scale ); \
- } \
+/**
+ * @brief Subcase uplo == ChamUpper, storev == ChamColumnwise || ChamRowwise of CORE_zsyssq()
+ */
+static inline int
+CORE_zsyssq_up_col( int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
+{
+ int i, j;
+ double tmp;
+ double *ptr;
+
+ for(j=0; j<N; j++) {
+ ptr = (double*) ( A + j * LDA );
+ for(i=0; i<j; i++, ptr++) {
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq+2*i, sclssq+2*i+1, &tmp );
+ sumsq_update( 1., sclssq+2*j, sclssq+2*j+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq+2*i, sclssq+2*i+1, &tmp );
+ sumsq_update( 1., sclssq+2*j, sclssq+2*j+1, &tmp );
+#endif
+ }
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq+2*j, sclssq+2*j+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq+2*j, sclssq+2*j+1, &tmp );
+#endif
+ }
+ return CHAMELEON_SUCCESS;
+}
+
+/**
+ * @brief Subcase uplo == ChamLower, storev == ChamColumnwise || ChamRowwise of CORE_zsyssq()
+ */
+static inline int
+CORE_zsyssq_lo_col( int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
+{
+ int i, j;
+ double tmp;
+ double *ptr;
+
+ for(j=0; j<N; j++) {
+ ptr = (double*) ( A + j * (LDA + 1) );
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq+2*j, sclssq+2*j+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq+2*j, sclssq+2*j+1, &tmp );
+#endif
+ ptr++;
+ for(i=j+1; i<N; i++, ptr++) {
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq+2*i, sclssq+2*i+1, &tmp );
+ sumsq_update( 1., sclssq+2*j, sclssq+2*j+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq+2*i, sclssq+2*i+1, &tmp );
+ sumsq_update( 1., sclssq+2*j, sclssq+2*j+1, &tmp );
+#endif
+ }
}
+ return CHAMELEON_SUCCESS;
+}
+
+/**
+ * @brief Subcase uplo == ChamUpper, storev == ChamEltwise of CORE_zsyssq()
+ */
+static inline int
+CORE_zsyssq_up_elt( int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
+{
+ int i, j;
+ double tmp;
+ double *ptr;
+
+ for(j=0; j<N; j++) {
+ ptr = (double*) ( A + j * LDA );
+
+ for(i=0; i<j; i++, ptr++) {
+ tmp = fabs(*ptr);
+ sumsq_update( 2., sclssq, sclssq+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 2., sclssq, sclssq+1, &tmp );
+#endif
+ }
+
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq, sclssq+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq, sclssq+1, &tmp );
+#endif
+ }
+ return CHAMELEON_SUCCESS;
+}
+
+/**
+ * @brief Subcase uplo == ChamLower, storev == ChamEltwise of CORE_zsyssq()
+ */
+static inline int
+CORE_zsyssq_lo_elt( int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
+{
+ int i, j;
+ double tmp;
+ double *ptr;
+
+ for(j=0; j<N; j++) {
+ ptr = (double*) ( A + j * (LDA + 1) );
+
+
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq, sclssq+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 1., sclssq, sclssq+1, &tmp );
+#endif
+ ptr++;
+
+ for(i=j+1; i<N; i++, ptr++) {
+ tmp = fabs(*ptr);
+ sumsq_update( 2., sclssq, sclssq+1, &tmp );
+#if defined(PRECISION_z) || defined(PRECISION_c)
+ ptr++;
+ tmp = fabs(*ptr);
+ sumsq_update( 2., sclssq, sclssq+1, &tmp );
+#endif
+ }
+ }
+ return CHAMELEON_SUCCESS;
+}
/**
*
@@ -61,6 +199,12 @@
*
*******************************************************************************
*
+ * @param[in] storev
+ * Specifies whether the sums are made per column or row.
+ * = ChamColumnwise: Computes the sum of squares on each column
+ * = ChamRowwise: Computes the sum of squares on each row
+ * = ChamEltwise: Computes the sum of squares on all the matrix
+ *
* @param[in] uplo
* Specifies whether the upper or lower triangular part of
* the symmetric matrix A is to be referenced as follows:
@@ -78,13 +222,16 @@
* @param[in] LDA
* The leading dimension of the tile A. LDA >= max(1,N).
*
- * @param[in,out] scale
- * On entry, the value scale in the equation above.
- * On exit, scale is overwritten with the value scl.
- *
- * @param[in,out] sumsq
- * On entry, the value sumsq in the equation above.
- * On exit, SUMSQ is overwritten with the value ssq.
+ * @param[in,out] sclssq
+ * Dimension: (2,K)
+ * if storev == ChamColumnwise, K = N
+ * if storev == ChamRowwise, K = N
+ * if storev == ChamEltwise, K = 1
+ * On entry, sclssq contains K couples (sclssq[2*i], sclssq[2*i+1])
+ * which corresponds to (scale, sumsq) in the equation below
+ * ( scl**2 )*ssq = sum( A( i, j )**2 ) + ( scale**2 )*sumsq,
+ * respectively for the columns, the rows and the full matrix
+ * On exit, each couple is overwritten with the final result (scl, ssq).
*
*******************************************************************************
*
@@ -93,67 +240,21 @@
*
*/
-int CORE_zsyssq(cham_uplo_t uplo, int N,
- const CHAMELEON_Complex64_t *A, int LDA,
- double *scale, double *sumsq)
+int CORE_zsyssq( cham_store_t storev, cham_uplo_t uplo, int N,
+ const CHAMELEON_Complex64_t *A, int LDA,
+ double *sclssq )
{
- int i, j;
- double tmp;
- double *ptr;
-
if ( uplo == ChamUpper ) {
- for(j=0; j<N; j++) {
- ptr = (double*) ( A + j * LDA );
-
- for(i=0; i<j; i++, ptr++) {
-
- tmp = fabs(*ptr);
- UPDATE( 2., tmp );
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- ptr++;
- tmp = fabs(*ptr);
- UPDATE( 2., tmp );
-#endif
- }
-
- /* Diagonal */
- tmp = fabs(*ptr);
- UPDATE( 1., tmp );
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- ptr++;
- tmp = fabs(*ptr);
- UPDATE( 1., tmp );
-#endif
+ if ( storev == ChamEltwise ) {
+ CORE_zsyssq_up_elt( N, A, LDA, sclssq );
+ } else {
+ CORE_zsyssq_up_col( N, A, LDA, sclssq );
}
} else {
-
- for(j=0; j<N; j++) {
- ptr = (double*) ( A + j * LDA + j);
-
- /* Diagonal */
- tmp = fabs(*ptr);
- UPDATE( 1., tmp );
- ptr++;
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- tmp = fabs(*ptr);
- UPDATE( 1., tmp );
- ptr++;
-#endif
-
- for(i=j+1; i<N; i++, ptr++) {
-
- tmp = fabs(*ptr);
- UPDATE( 2., tmp );
-
-#if defined(PRECISION_z) || defined(PRECISION_c)
- ptr++;
- tmp = fabs(*ptr);
- UPDATE( 2., tmp );
-#endif
- }
+ if ( storev == ChamEltwise ) {
+ CORE_zsyssq_lo_elt( N, A, LDA, sclssq );
+ } else {
+ CORE_zsyssq_lo_col( N, A, LDA, sclssq );
}
}
return CHAMELEON_SUCCESS;
diff --git a/coreblas/eztrace_module/coreblas_eztrace_module b/coreblas/eztrace_module/coreblas_eztrace_module
index 3bca3cbf12c4763eec665bf20a02a23d38ee72ae..389858563abc4cfe8cca9e5d40109f6c2a461e3d 100644
--- a/coreblas/eztrace_module/coreblas_eztrace_module
+++ b/coreblas/eztrace_module/coreblas_eztrace_module
@@ -32,9 +32,9 @@ int CORE_cgessm(int M, int N, int K, int IB,
const int *IPIV,
void *L, int LDL,
void *A, int LDA);
-int CORE_cgessq(int M, int N,
+int CORE_cgessq(cham_store_t storev, int M, int N,
void *A, int LDA,
- float *scale, float *sumsq);
+ float *sclssq);
int CORE_cgetf2_nopiv(int M, int N,
void *A, int LDA);
int CORE_cgetrf(int M, int N,
@@ -70,9 +70,9 @@ void CORE_cher2k(int uplo, int trans,
void *alpha, void *A, int LDA,
void *B, int LDB,
float beta, void *C, int LDC);
-int CORE_chessq(int uplo, int N,
+int CORE_chessq(cham_store_t storev, int uplo, int N,
void *A, int LDA,
- float *scale, float *sumsq);
+ float *sclssq);
int CORE_cherfb(int uplo, int N, int K, int IB, int NB,
void *A, int LDA,
@@ -190,9 +190,9 @@ void CORE_csyr2k(int uplo, int trans,
void *alpha, void *A, int LDA,
void *B, int LDB,
void *beta, void *C, int LDC);
-int CORE_csyssq(int uplo, int N,
+int CORE_csyssq(cham_store_t storev, int uplo, int N,
void *A, int LDA,
- float *scale, float *sumsq);
+ float *sclssq);
int CORE_csytf2_nopiv(int uplo, int n, void *A, int lda);
void CORE_ctrmm(int side, int uplo,
int transA, int diag,
@@ -334,9 +334,9 @@ int CORE_dgessm(int M, int N, int K, int IB,
const int *IPIV,
const double *L, int LDL,
double *A, int LDA);
-int CORE_dgessq(int M, int N,
+int CORE_dgessq(cham_store_t storev, int M, int N,
const double *A, int LDA,
- double *scale, double *sumsq);
+ double *sclssq);
int CORE_dgetf2_nopiv(int M, int N,
double *A, int LDA);
int CORE_dgetrf(int M, int N,
@@ -372,9 +372,9 @@ void CORE_dsyr2k(int uplo, int trans,
double alpha, const double *A, int LDA,
const double *B, int LDB,
double beta, double *C, int LDC);
-int CORE_dsyssq(int uplo, int N,
+int CORE_dsyssq(cham_store_t storev, int uplo, int N,
const double *A, int LDA,
- double *scale, double *sumsq);
+ double *sclssq);
int CORE_dsyrfb(int uplo, int N, int K, int IB, int NB,
const double *A, int LDA,
@@ -621,9 +621,9 @@ int CORE_sgessm(int M, int N, int K, int IB,
const int *IPIV,
const float *L, int LDL,
float *A, int LDA);
-int CORE_sgessq(int M, int N,
+int CORE_sgessq(cham_store_t storev, int M, int N,
const float *A, int LDA,
- float *scale, float *sumsq);
+ float *sclssq);
int CORE_sgetf2_nopiv(int M, int N,
float *A, int LDA);
int CORE_sgetrf(int M, int N,
@@ -659,9 +659,9 @@ void CORE_ssyr2k(int uplo, int trans,
float alpha, const float *A, int LDA,
const float *B, int LDB,
float beta, float *C, int LDC);
-int CORE_ssyssq(int uplo, int N,
+int CORE_ssyssq(cham_store_t storev, int uplo, int N,
const float *A, int LDA,
- float *scale, float *sumsq);
+ float *sclssq);
int CORE_ssyrfb(int uplo, int N, int K, int IB, int NB,
const float *A, int LDA,
@@ -901,9 +901,9 @@ int CORE_zgessm(int M, int N, int K, int IB,
const int *IPIV,
void *L, int LDL,
void *A, int LDA);
-int CORE_zgessq(int M, int N,
+int CORE_zgessq(cham_store_t storev, int M, int N,
void *A, int LDA,
- double *scale, double *sumsq);
+ double *sclssq);
int CORE_zgetf2_nopiv(int M, int N,
void *A, int LDA);
int CORE_zgetrf(int M, int N,
@@ -939,9 +939,9 @@ void CORE_zher2k(int uplo, int trans,
void *alpha, void *A, int LDA,
void *B, int LDB,
double beta, void *C, int LDC);
-int CORE_zhessq(int uplo, int N,
+int CORE_zhessq(cham_store_t storev, int uplo, int N,
void *A, int LDA,
- double *scale, double *sumsq);
+ double *sclssq);
int CORE_zherfb(int uplo, int N, int K, int IB, int NB,
void *A, int LDA,
@@ -1028,6 +1028,9 @@ void CORE_zplgsy(void *bump, int m, int n, void *A, int lda,
int bigM, int m0, int n0, unsigned long long int seed );
void CORE_zplrnt(int m, int n, void *A, int lda,
int bigM, int m0, int n0, unsigned long long int seed );
+int CORE_zplssq( cham_store_t storev, int M, int N,
+ double *sclssqin, double *sclssqout );
+int CORE_zplssq2( int N, double *sclssq );
void CORE_zpotrf(int uplo, int N, void *A, int LDA, int *INFO);
void CORE_zshift(int s, int m, int n, int L,
void *A);
@@ -1053,9 +1056,9 @@ void CORE_zsyr2k(int uplo, int trans,
void *alpha, void *A, int LDA,
void *B, int LDB,
void *beta, void *C, int LDC);
-int CORE_zsyssq(int uplo, int N,
+int CORE_zsyssq(cham_store_t storev, int uplo, int N,
void *A, int LDA,
- double *scale, double *sumsq);
+ double *sclssq);
int CORE_zsytf2_nopiv(int uplo, int n, void *A, int lda);
void CORE_ztrmm(int side, int uplo,
int transA, int diag,
diff --git a/coreblas/include/coreblas/coreblas_z.h b/coreblas/include/coreblas/coreblas_z.h
index ee434de43d2da6c4219b00f8b8e4236a419adb86..2bd8c01067594c0c85ee2d43aeb6432d5b9ba031 100644
--- a/coreblas/include/coreblas/coreblas_z.h
+++ b/coreblas/include/coreblas/coreblas_z.h
@@ -64,9 +64,9 @@ int CORE_zgessm(int M, int N, int K, int IB,
const int *IPIV,
const CHAMELEON_Complex64_t *L, int LDL,
CHAMELEON_Complex64_t *A, int LDA);
-int CORE_zgessq(int M, int N,
+int CORE_zgessq(cham_store_t storev, int M, int N,
const CHAMELEON_Complex64_t *A, int LDA,
- double *scale, double *sumsq);
+ double *sclssq);
int CORE_zgetf2_nopiv(int M, int N,
CHAMELEON_Complex64_t *A, int LDA);
int CORE_zgetrf(int M, int N,
@@ -104,9 +104,9 @@ void CORE_zher2k(cham_uplo_t uplo, cham_trans_t trans,
CHAMELEON_Complex64_t alpha, const CHAMELEON_Complex64_t *A, int LDA,
const CHAMELEON_Complex64_t *B, int LDB,
double beta, CHAMELEON_Complex64_t *C, int LDC);
-int CORE_zhessq(cham_uplo_t uplo, int N,
+int CORE_zhessq(cham_store_t storev, cham_uplo_t uplo, int N,
const CHAMELEON_Complex64_t *A, int LDA,
- double *scale, double *sumsq);
+ double *sclssq);
int CORE_zherfb(cham_uplo_t uplo, int N, int K, int IB, int NB,
const CHAMELEON_Complex64_t *A, int LDA,
const CHAMELEON_Complex64_t *T, int LDT,
@@ -196,6 +196,9 @@ void CORE_zplgsy(CHAMELEON_Complex64_t bump, int m, int n, CHAMELEON_Complex64_t
int bigM, int m0, int n0, unsigned long long int seed );
void CORE_zplrnt(int m, int n, CHAMELEON_Complex64_t *A, int lda,
int bigM, int m0, int n0, unsigned long long int seed );
+int CORE_zplssq( cham_store_t storev, int M, int N,
+ double *sclssqin, double *sclssqout );
+int CORE_zplssq2( int N, double *sclssq );
void CORE_zpotrf(cham_uplo_t uplo, int N, CHAMELEON_Complex64_t *A, int LDA, int *INFO);
void CORE_zshift(int s, int m, int n, int L,
CHAMELEON_Complex64_t *A);
@@ -221,9 +224,9 @@ void CORE_zsyr2k(cham_uplo_t uplo, cham_trans_t trans,
CHAMELEON_Complex64_t alpha, const CHAMELEON_Complex64_t *A, int LDA,
const CHAMELEON_Complex64_t *B, int LDB,
CHAMELEON_Complex64_t beta, CHAMELEON_Complex64_t *C, int LDC);
-int CORE_zsyssq(cham_uplo_t uplo, int N,
+int CORE_zsyssq(cham_store_t storev, cham_uplo_t uplo, int N,
const CHAMELEON_Complex64_t *A, int LDA,
- double *scale, double *sumsq);
+ double *sclssq);
int CORE_zsytf2_nopiv(cham_uplo_t uplo, int n, CHAMELEON_Complex64_t *A, int lda);
int CORE_ztradd(cham_uplo_t uplo, cham_trans_t trans, int M, int N,
CHAMELEON_Complex64_t alpha,
diff --git a/coreblas/include/coreblas/sumsq_update.h b/coreblas/include/coreblas/sumsq_update.h
new file mode 100644
index 0000000000000000000000000000000000000000..1c94e85ba13d540c5b72ad3277937829b9375c5b
--- /dev/null
+++ b/coreblas/include/coreblas/sumsq_update.h
@@ -0,0 +1,132 @@
+/**
+ *
+ * @file sumsq_update.h
+ *
+ * @copyright 2012-2019 Bordeaux INP, CNRS (LaBRI UMR 5800), Inria,
+ * Univ. Bordeaux. All rights reserved.
+ *
+ ***
+ *
+ * @brief Chameleon CPU auxiliary sumsq_update routine
+ *
+ * @version 0.9.2
+ * @author Mathieu Faverge
+ * @date 2019-03-27
+ *
+ */
+#ifndef _sumsq_update_h_
+#define _sumsq_update_h_
+
+/**
+ *******************************************************************************
+ *
+ * @brief Update the couple (scale, sumsq) with one element when
+ * computing the Froebnius norm.
+ *
+ * The frobenius norm is equal to scale * sqrt( sumsq ), this method allows to
+ * avoid overflow in the sum square computation.
+ *
+ *******************************************************************************
+ *
+ * @param[inout] scale
+ * On entry, the former scale
+ * On exit, the update scale to take into account the value
+ *
+ * @param[inout] sumsq
+ * On entry, the former sumsq
+ * On exit, the update sumsq to take into account the value
+ *
+ * @param[in] value
+ * The value to integrate into the couple (scale, sumsq)
+ *
+ *******************************************************************************/
+static inline void
+#if defined(PRECISION_d) || defined(PRECISION_z)
+sumsq_update( int nb, double *scale, double *sumsq, const double *value )
+{
+ double absval = fabs(*value);
+ double ratio;
+ if ( absval != 0. ){
+ if ( (*scale) < absval ) {
+ ratio = (*scale) / absval;
+ *sumsq = (double)nb + (*sumsq) * ratio * ratio;
+ *scale = absval;
+ } else {
+ ratio = absval / (*scale);
+ *sumsq = (*sumsq) + (double)nb * ratio * ratio;
+ }
+ }
+}
+#elif defined(PRECISION_s) || defined(PRECISION_c)
+sumsq_update( int nb, float *scale, float *sumsq, const float *value )
+{
+ float absval = fabs(*value);
+ float ratio;
+ if ( absval != 0. ){
+ if ( (*scale) < absval ) {
+ ratio = (*scale) / absval;
+ *sumsq = (float)nb + (*sumsq) * ratio * ratio;
+ *scale = absval;
+ } else {
+ ratio = absval / (*scale);
+ *sumsq = (*sumsq) + (float)nb * ratio * ratio;
+ }
+ }
+}
+#endif
+
+/**
+ *******************************************************************************
+ *
+ * @brief Update the couple (scale, sumsq) by adding another couple when
+ * computing the Froebnius norm.
+ *
+ * The frobenius norm is equal to scale * sqrt( sumsq ), this method allows to
+ * avoid overflow in the sum square computation.
+ *
+ *******************************************************************************
+ *
+ * @param[inout] scale
+ * On entry, the former scale
+ * On exit, the update scale to take into account the value
+ *
+ * @param[inout] sumsq
+ * On entry, the former sumsq
+ * On exit, the update sumsq to take into account the value
+ *
+ * @param[in] value
+ * The value to integrate into the couple (scale, sumsq)
+ *
+ *******************************************************************************/
+static inline void
+#if defined(PRECISION_d) || defined(PRECISION_z)
+sumsq_update_2( int nb, const double *scalein, const double *sumsqin, double *scaleout, double *sumsqout )
+{
+ if (*scaleout >= 0.) {
+ if ( (*scaleout) < (*scalein) ) {
+ *sumsqout = *sumsqin + (*sumsqout * (( *scaleout / *scalein ) * ( *scaleout / *scalein )));
+ *scaleout = *scalein;
+ } else {
+ if ( (*scaleout) > 0. ){
+ *sumsqout = *sumsqout + (*sumsqin * (( *scalein / *scaleout ) * ( *scalein / *scaleout )));
+ }
+ }
+ }
+}
+#elif defined(PRECISION_s) || defined(PRECISION_c)
+sumsq_update_2( int nb, const float *scalein, const float *sumsqin, float *scaleout, float *sumsqout )
+{
+ if (*scaleout >= 0.) {
+ if ( (*scaleout) < (*scalein) ) {
+ *sumsqout = *sumsqin + (*sumsqout * (( *scaleout / *scalein ) * ( *scaleout / *scalein )));
+ *scaleout = *scalein;
+ } else {
+ if ( (*scaleout) > 0. ){
+ *sumsqout = *sumsqout + (*sumsqin * (( *scalein / *scaleout ) * ( *scalein / *scaleout )));
+ }
+ }
+ }
+}
+#endif
+
+#endif /* _sumsq_update_h_ */
diff --git a/include/chameleon/constants.h b/include/chameleon/constants.h
index 7e623aebf155ab457a8fd15f85e55539553a6967..ec7ea257144d0f1fc0730feb2f5817b98cd956f0 100644
--- a/include/chameleon/constants.h
+++ b/include/chameleon/constants.h
@@ -151,6 +151,7 @@ typedef enum chameleon_dir_e {
typedef enum chameleon_store_e {
ChamColumnwise = 401, /**< Column wise storage */
ChamRowwise = 402, /**< Row wise storage */
+ ChamEltwise = 403, /**< Element by element storage */
} cham_store_t;
diff --git a/include/chameleon/tasks_z.h b/include/chameleon/tasks_z.h
index c82829c9e78db4c216de1d41eaa0d8793384f37f..2d8ec988208595f60f34b9355fa6f192ba89db6b 100644
--- a/include/chameleon/tasks_z.h
+++ b/include/chameleon/tasks_z.h
@@ -66,7 +66,7 @@ void INSERT_TASK_zgessm( const RUNTIME_option_t *options,
const CHAM_desc_t *D, int Dm, int Dn, int ldd,
const CHAM_desc_t *A, int Am, int An, int lda );
void INSERT_TASK_zgessq( const RUNTIME_option_t *options,
- int m, int n,
+ cham_store_t storev, int m, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn );
void INSERT_TASK_zgetrf( const RUNTIME_option_t *options,
@@ -112,7 +112,7 @@ void INSERT_TASK_zherk( const RUNTIME_option_t *options,
double alpha, const CHAM_desc_t *A, int Am, int An, int lda,
double beta, const CHAM_desc_t *C, int Cm, int Cn, int ldc );
void INSERT_TASK_zhessq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
+ cham_store_t storev, cham_uplo_t uplo, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn );
void INSERT_TASK_zlacpy( const RUNTIME_option_t *options,
@@ -172,9 +172,10 @@ void INSERT_TASK_zplrnt( const RUNTIME_option_t *options,
int m, int n, const CHAM_desc_t *A, int Am, int An, int lda,
int bigM, int m0, int n0, unsigned long long int seed );
void INSERT_TASK_zplssq( const RUNTIME_option_t *options,
+ cham_store_t storev, int M, int N,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn,
const CHAM_desc_t *SCLSSQ, int SCLSSQm, int SCLSSQn );
-void INSERT_TASK_zplssq2( const RUNTIME_option_t *options,
+void INSERT_TASK_zplssq2( const RUNTIME_option_t *options, int N,
const CHAM_desc_t *RESULT, int RESULTm, int RESULTn );
void INSERT_TASK_zpotrf( const RUNTIME_option_t *options,
cham_uplo_t uplo, int n, int nb,
@@ -205,7 +206,7 @@ void INSERT_TASK_zsyrk( const RUNTIME_option_t *options,
CHAMELEON_Complex64_t alpha, const CHAM_desc_t *A, int Am, int An, int lda,
CHAMELEON_Complex64_t beta, const CHAM_desc_t *C, int Cm, int Cn, int ldc );
void INSERT_TASK_zsyssq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
+ cham_store_t storev, cham_uplo_t uplo, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn );
void INSERT_TASK_zsytrf_nopiv( const RUNTIME_option_t *options,
diff --git a/runtime/openmp/codelets/codelet_zgessq.c b/runtime/openmp/codelets/codelet_zgessq.c
index ae15172d441dcd6bbf8475b9c3d841a5acfc2be3..42453eaacfde2ddc798a70059df8d417c7bd1edb 100644
--- a/runtime/openmp/codelets/codelet_zgessq.c
+++ b/runtime/openmp/codelets/codelet_zgessq.c
@@ -25,12 +25,12 @@
#include "coreblas/coreblas_z.h"
void INSERT_TASK_zgessq( const RUNTIME_option_t *options,
- int m, int n,
+ cham_store_t storev, int m, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
CHAMELEON_Complex64_t *ptrA = RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An);
double *ptrScaleSum = RTBLKADDR(SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn);
-#pragma omp task firstprivate(m, n, ptrA, lda, ptrScaleSum) depend(in:ptrA[0]) depend(inout:ptrScaleSum[0])
- CORE_zgessq( m, n, ptrA, lda, &ptrScaleSum[0], &ptrScaleSum[1] );
+#pragma omp task firstprivate(storev, m, n, ptrA, lda, ptrScaleSum) depend(in:ptrA[0]) depend(inout:ptrScaleSum[0])
+ CORE_zgessq( storev, m, n, ptrA, lda, ptrScaleSum );
}
diff --git a/runtime/openmp/codelets/codelet_zhessq.c b/runtime/openmp/codelets/codelet_zhessq.c
index 702095aed7388e453b0b9696b0291fc090392bd8..85232d001a1b29286796d1a2e17e77eb095f310c 100644
--- a/runtime/openmp/codelets/codelet_zhessq.c
+++ b/runtime/openmp/codelets/codelet_zhessq.c
@@ -22,15 +22,13 @@
*/
#include "chameleon_openmp.h"
#include "chameleon/tasks_z.h"
-#include "coreblas/coreblas_z.h"
void INSERT_TASK_zhessq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
- const CHAM_desc_t *A, int Am, int An, int lda,
- const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
+ cham_store_t storev, cham_uplo_t uplo, int n,
+ const CHAM_desc_t *A, int Am, int An, int lda,
+ const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
- CHAMELEON_Complex64_t *ptrA = RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An);
- double *ptrScaleSum = RTBLKADDR(SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn);
-#pragma omp task firstprivate(uplo, n, ptrA, lda, ptrScaleSum) depend(in:ptrScaleSum[0]) depend(inout:ptrA[0])
- CORE_zhessq( uplo, n, ptrA, lda, &ptrScaleSum[0], &ptrScaleSum[1] );
+ INSERT_TASK_zsyssq( options, storev, uplo, n,
+ A, Am, An, lda,
+ SCALESUMSQ, SCALESUMSQm, SCALESUMSQn );
}
diff --git a/runtime/openmp/codelets/codelet_zplssq.c b/runtime/openmp/codelets/codelet_zplssq.c
index e41484c09702a4f9deddbf2645f904b03ad0a6e4..61196102c2d1ba7646925bb4ff27c2e858e08dd2 100644
--- a/runtime/openmp/codelets/codelet_zplssq.c
+++ b/runtime/openmp/codelets/codelet_zplssq.c
@@ -25,61 +25,22 @@
#include "chameleon/tasks_z.h"
#include "coreblas/coreblas_z.h"
-/**
- *
- * @ingroup CORE_CHAMELEON_Complex64_t
- *
- * @brief Compute sum( a_ij ^ 2 ) = scl * sqrt(ssq)
- *
- * with scl and ssq such that
- *
- * ( scl**2 )*ssq = sum( A( 2*i )**2 * A( 2*i+1 ) )
- * i
- *
- * The values of A(2*i+1) are assumed to be at least unity.
- * The values of A(2*i) are assumed to be non-negative and scl is
- *
- * scl = max( A( 2*i ) ),
- * i
- *
- * The routine makes only one pass through the matrix A.
- *
- *******************************************************************************
- *
- * @param[in] M
- * The number of couple (scale, sumsq) in the matrix A.
- *
- * @param[in] A
- * The 2-by-M matrix.
- *
- * @param[out] result
- * On exit, result contains scl * sqrt( ssq )
- *
- */
void INSERT_TASK_zplssq( const RUNTIME_option_t *options,
+ cham_store_t storev, int M, int N,
const CHAM_desc_t *IN, int INm, int INn,
const CHAM_desc_t *OUT, int OUTm, int OUTn )
{
double *sclssq_in = RTBLKADDR(IN, double, INm, INn );
double *sclssq_out = RTBLKADDR(OUT, double, OUTm, OUTn);
-#pragma omp task depend(in: sclssq_in[0]) depend(inout: sclssq_out[0])
- {
- if( sclssq_out[0] < sclssq_in[0] ) {
- sclssq_out[1] = sclssq_in[1] + (sclssq_out[1] * (( sclssq_out[0] / sclssq_in[0] ) * ( sclssq_out[0] / sclssq_in[0] )));
- sclssq_out[0] = sclssq_in[0];
- } else {
- sclssq_out[1] = sclssq_out[1] + (sclssq_in[1] * (( sclssq_in[0] / sclssq_out[0] ) * ( sclssq_in[0] / sclssq_out[0] )));
- }
- }
+#pragma omp task firstprivate(storev, M, N) depend(in: sclssq_in[0]) depend(inout: sclssq_out[0])
+ CORE_zplssq(storev, M, N, sclssq_in, sclssq_out);
}
-void INSERT_TASK_zplssq2( const RUNTIME_option_t *options,
+void INSERT_TASK_zplssq2( const RUNTIME_option_t *options, int N,
const CHAM_desc_t *RESULT, int RESULTm, int RESULTn )
{
- CHAMELEON_Complex64_t *res = RTBLKADDR(RESULT, CHAMELEON_Complex64_t, RESULTm, RESULTn);
+ CHAMELEON_Complex64_t *res = RTBLKADDR(RESULT, double, RESULTm, RESULTn);
-#pragma omp task depend(inout: res[0])
- {
- res[0] = res[0] * sqrt( res[1] );
- }
+#pragma omp task firstprivate(N) depend(inout: res[0])
+ CORE_zplssq2(N, res);
}
diff --git a/runtime/openmp/codelets/codelet_zsyssq.c b/runtime/openmp/codelets/codelet_zsyssq.c
index bc616c81beccd9a2bc9e3711991fe1d2d7d587b6..32a7dc9d52775570d9575755b8d5c05d53945b70 100644
--- a/runtime/openmp/codelets/codelet_zsyssq.c
+++ b/runtime/openmp/codelets/codelet_zsyssq.c
@@ -23,12 +23,12 @@
#include "chameleon/tasks_z.h"
void INSERT_TASK_zsyssq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
+ cham_store_t storev, cham_uplo_t uplo, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
CHAMELEON_Complex64_t *ptrA = RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An);
double *ptrSCALESUMSQ = RTBLKADDR(SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn);
-#pragma omp task firstprivate(uplo, n, ptrA, lda, ptrSCALESUMSQ) depend(in:ptrA[0]) depend(inout:ptrSCALESUMSQ[0])
- CORE_zsyssq( uplo, n, ptrA, lda, &ptrSCALESUMSQ[0], &ptrSCALESUMSQ[1] );
+#pragma omp task firstprivate(storev, uplo, n, ptrA, lda, ptrSCALESUMSQ) depend(in:ptrA[0]) depend(inout:ptrSCALESUMSQ[0])
+ CORE_zsyssq( storev, uplo, n, ptrA, lda, ptrSCALESUMSQ );
}
diff --git a/runtime/parsec/codelets/codelet_zgessq.c b/runtime/parsec/codelets/codelet_zgessq.c
index eb1ffd6f66d86816206c819274f48265be2f6e88..93b332cbf9117375286077075141c3f4a7f14560 100644
--- a/runtime/parsec/codelets/codelet_zgessq.c
+++ b/runtime/parsec/codelets/codelet_zgessq.c
@@ -25,6 +25,7 @@ static inline int
CORE_zgessq_parsec( parsec_execution_stream_t *context,
parsec_task_t *this_task )
{
+ cham_store_t storev;
int m;
int n;
CHAMELEON_Complex64_t *A;
@@ -32,16 +33,16 @@ CORE_zgessq_parsec( parsec_execution_stream_t *context,
double *SCALESUMSQ;
parsec_dtd_unpack_args(
- this_task, &m, &n, &A, &lda, &SCALESUMSQ );
+ this_task, &storev, &m, &n, &A, &lda, &SCALESUMSQ );
- CORE_zgessq( m, n, A, lda, SCALESUMSQ, SCALESUMSQ+1 );
+ CORE_zgessq( storev, m, n, A, lda, SCALESUMSQ );
(void)context;
return PARSEC_HOOK_RETURN_DONE;
}
void INSERT_TASK_zgessq( const RUNTIME_option_t *options,
- int m, int n,
+ cham_store_t storev, int m, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
@@ -49,10 +50,11 @@ void INSERT_TASK_zgessq( const RUNTIME_option_t *options,
parsec_dtd_taskpool_insert_task(
PARSEC_dtd_taskpool, CORE_zgessq_parsec, options->priority, "gessq",
- sizeof(int), &m, VALUE,
- sizeof(int), &n, VALUE,
+ sizeof(cham_store_t), &storev, VALUE,
+ sizeof(int), &m, VALUE,
+ sizeof(int), &n, VALUE,
PASSED_BY_REF, RTBLKADDR( A, CHAMELEON_Complex64_t, Am, An ), chameleon_parsec_get_arena_index( A ) | INPUT,
- sizeof(int), &lda, VALUE,
+ sizeof(int), &lda, VALUE,
PASSED_BY_REF, RTBLKADDR( SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn ), INOUT | AFFINITY,
PARSEC_DTD_ARG_END );
}
diff --git a/runtime/parsec/codelets/codelet_zhessq.c b/runtime/parsec/codelets/codelet_zhessq.c
index 238e56bcbc80cb0a74b6d55b0049bb71d32a3072..2229c62eb91c03e52cbc848f482b2702a40a4c0c 100644
--- a/runtime/parsec/codelets/codelet_zhessq.c
+++ b/runtime/parsec/codelets/codelet_zhessq.c
@@ -19,40 +19,13 @@
*/
#include "chameleon_parsec.h"
#include "chameleon/tasks_z.h"
-#include "coreblas/coreblas_z.h"
-
-static inline int
-CORE_zhessq_parsec( parsec_execution_stream_t *context,
- parsec_task_t *this_task )
-{
- cham_uplo_t uplo;
- int n;
- CHAMELEON_Complex64_t *A;
- int lda;
- double *SCALESUMSQ;
-
- parsec_dtd_unpack_args(
- this_task, &uplo, &n, &A, &lda, &SCALESUMSQ );
-
- CORE_zhessq( uplo, n, A, lda, &SCALESUMSQ[0], &SCALESUMSQ[1]);
-
- (void)context;
- return PARSEC_HOOK_RETURN_DONE;
-}
void INSERT_TASK_zhessq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
- const CHAM_desc_t *A, int Am, int An, int lda,
- const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
+ cham_store_t storev, cham_uplo_t uplo, int n,
+ const CHAM_desc_t *A, int Am, int An, int lda,
+ const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
- parsec_taskpool_t* PARSEC_dtd_taskpool = (parsec_taskpool_t *)(options->sequence->schedopt);
-
- parsec_dtd_taskpool_insert_task(
- PARSEC_dtd_taskpool, CORE_zhessq_parsec, options->priority, "hessq",
- sizeof(int), &uplo, VALUE,
- sizeof(int), &n, VALUE,
- PASSED_BY_REF, RTBLKADDR( A, CHAMELEON_Complex64_t, Am, An ), chameleon_parsec_get_arena_index( A ) | INPUT,
- sizeof(int), &lda, VALUE,
- PASSED_BY_REF, RTBLKADDR( SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn ), INOUT | AFFINITY,
- PARSEC_DTD_ARG_END );
+ INSERT_TASK_zsyssq( options, storev, uplo, n,
+ A, Am, An, lda,
+ SCALESUMSQ, SCALESUMSQm, SCALESUMSQn );
}
diff --git a/runtime/parsec/codelets/codelet_zplssq.c b/runtime/parsec/codelets/codelet_zplssq.c
index 6c797829395817c83819a5cca0ab53601d25e295..16f19a21fad3581ec4a170d7fae3a90b657d5350 100644
--- a/runtime/parsec/codelets/codelet_zplssq.c
+++ b/runtime/parsec/codelets/codelet_zplssq.c
@@ -26,58 +26,23 @@ static inline int
CORE_zplssq_parsec( parsec_execution_stream_t *context,
parsec_task_t *this_task )
{
+ cham_store_t storev;
+ int M;
+ int N;
double *SCLSSQ_IN;
double *SCLSSQ_OUT;
parsec_dtd_unpack_args(
- this_task, &SCLSSQ_IN, &SCLSSQ_OUT );
+ this_task, &storev, &M, &N, &SCLSSQ_IN, &SCLSSQ_OUT );
- assert( SCLSSQ_OUT[0] >= 0. );
- if( SCLSSQ_OUT[0] < SCLSSQ_IN[0] ) {
- SCLSSQ_OUT[1] = SCLSSQ_IN[1] + (SCLSSQ_OUT[1] * (( SCLSSQ_OUT[0] / SCLSSQ_IN[0] ) * ( SCLSSQ_OUT[0] / SCLSSQ_IN[0] )));
- SCLSSQ_OUT[0] = SCLSSQ_IN[0];
- } else {
- if ( SCLSSQ_OUT[0] > 0 ) {
- SCLSSQ_OUT[1] = SCLSSQ_OUT[1] + (SCLSSQ_IN[1] * (( SCLSSQ_IN[0] / SCLSSQ_OUT[0] ) * ( SCLSSQ_IN[0] / SCLSSQ_OUT[0] )));
- }
- }
+ CORE_zplssq(storev, M, N, SCLSSQ_IN, SCLSSQ_OUT);
(void)context;
return PARSEC_HOOK_RETURN_DONE;
}
-/**
- *
- * @ingroup INSERT_TASK_Complex64_t
- *
- * @brief Compute sum( a_ij ^ 2 ) = scl * sqrt(ssq)
- *
- * with scl and ssq such that
- *
- * ( scl**2 )*ssq = sum( A( 2*i )**2 * A( 2*i+1 ) )
- * i
- *
- * The values of A(2*i+1) are assumed to be at least unity.
- * The values of A(2*i) are assumed to be non-negative and scl is
- *
- * scl = max( A( 2*i ) ),
- * i
- *
- * The routine makes only one pass through the matrix A.
- *
- *******************************************************************************
- *
- * @param[in] M
- * The number of couple (scale, sumsq) in the matrix A.
- *
- * @param[in] A
- * The 2-by-M matrix.
- *
- * @param[out] result
- * On exit, result contains scl * sqrt( ssq )
- *
- */
void INSERT_TASK_zplssq( const RUNTIME_option_t *options,
+ cham_store_t storev, int M, int N,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn,
const CHAM_desc_t *SCLSSQ, int SCLSSQm, int SCLSSQn )
{
@@ -85,6 +50,9 @@ void INSERT_TASK_zplssq( const RUNTIME_option_t *options,
parsec_dtd_taskpool_insert_task(
PARSEC_dtd_taskpool, CORE_zplssq_parsec, options->priority, "plssq",
+ sizeof(int), &storev, VALUE,
+ sizeof(int), &M, VALUE,
+ sizeof(int), &N, VALUE,
PASSED_BY_REF, RTBLKADDR( SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn ), INPUT,
PASSED_BY_REF, RTBLKADDR( SCLSSQ, double, SCLSSQm, SCLSSQn ), INOUT | AFFINITY,
PARSEC_DTD_ARG_END );
@@ -94,24 +62,26 @@ static inline int
CORE_zplssq2_parsec( parsec_execution_stream_t *context,
parsec_task_t *this_task )
{
+ int N;
double *RESULT;
parsec_dtd_unpack_args(
- this_task, &RESULT );
+ this_task, &N, &RESULT );
- RESULT[0] = RESULT[0] * sqrt( RESULT[1] );
+ CORE_zplssq2(N, RESULT);
(void)context;
return PARSEC_HOOK_RETURN_DONE;
}
-void INSERT_TASK_zplssq2( const RUNTIME_option_t *options,
+void INSERT_TASK_zplssq2( const RUNTIME_option_t *options, int N,
const CHAM_desc_t *RESULT, int RESULTm, int RESULTn )
{
parsec_taskpool_t* PARSEC_dtd_taskpool = (parsec_taskpool_t *)(options->sequence->schedopt);
parsec_dtd_taskpool_insert_task(
PARSEC_dtd_taskpool, CORE_zplssq2_parsec, options->priority, "plssq2",
+ sizeof(int), &N, VALUE,
PASSED_BY_REF, RTBLKADDR( RESULT, double, RESULTm, RESULTn ), INOUT | AFFINITY,
PARSEC_DTD_ARG_END );
}
diff --git a/runtime/parsec/codelets/codelet_zsyssq.c b/runtime/parsec/codelets/codelet_zsyssq.c
index 9f06c6d4c35769048c1817d92fafe694d5543f09..8e86035ccbaafd400985c1ffea92b9e27643e588 100644
--- a/runtime/parsec/codelets/codelet_zsyssq.c
+++ b/runtime/parsec/codelets/codelet_zsyssq.c
@@ -25,6 +25,7 @@ static inline int
CORE_zsyssq_parsec( parsec_execution_stream_t *context,
parsec_task_t *this_task )
{
+ cham_store_t storev;
cham_uplo_t uplo;
int n;
CHAMELEON_Complex64_t *A;
@@ -32,16 +33,16 @@ CORE_zsyssq_parsec( parsec_execution_stream_t *context,
double *SCALESUMSQ;
parsec_dtd_unpack_args(
- this_task, &uplo, &n, &A, &lda, &SCALESUMSQ );
+ this_task, &storev, &uplo, &n, &A, &lda, &SCALESUMSQ );
- CORE_zsyssq( uplo, n, A, lda, &SCALESUMSQ[0], &SCALESUMSQ[1]);
+ CORE_zsyssq( storev, uplo, n, A, lda, SCALESUMSQ );
(void)context;
return PARSEC_HOOK_RETURN_DONE;
}
void INSERT_TASK_zsyssq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
+ cham_store_t storev, cham_uplo_t uplo, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
@@ -49,7 +50,8 @@ void INSERT_TASK_zsyssq( const RUNTIME_option_t *options,
parsec_dtd_taskpool_insert_task(
PARSEC_dtd_taskpool, CORE_zsyssq_parsec, options->priority, "syssq",
- sizeof(int), &uplo, VALUE,
+ sizeof(cham_store_t), &storev, VALUE,
+ sizeof(int), &uplo, VALUE,
sizeof(int), &n, VALUE,
PASSED_BY_REF, RTBLKADDR( A, CHAMELEON_Complex64_t, Am, An ), chameleon_parsec_get_arena_index( A ) | INPUT,
sizeof(int), &lda, VALUE,
diff --git a/runtime/quark/codelets/codelet_zgessq.c b/runtime/quark/codelets/codelet_zgessq.c
index f63b6d99a6fc597a6b97549d24860a839f6364eb..bf5396e4206dbd5f9389377ac11a16c45b4e24d7 100644
--- a/runtime/quark/codelets/codelet_zgessq.c
+++ b/runtime/quark/codelets/codelet_zgessq.c
@@ -25,27 +25,29 @@
void CORE_zgessq_quark(Quark *quark)
{
+ cham_store_t storev;
int m;
int n;
CHAMELEON_Complex64_t *A;
int lda;
double *SCALESUMSQ;
- quark_unpack_args_5( quark, m, n, A, lda, SCALESUMSQ );
- CORE_zgessq( m, n, A, lda, &SCALESUMSQ[0], &SCALESUMSQ[1]);
+ quark_unpack_args_6( quark, storev, m, n, A, lda, SCALESUMSQ );
+ CORE_zgessq( storev, m, n, A, lda, SCALESUMSQ );
}
void INSERT_TASK_zgessq( const RUNTIME_option_t *options,
- int m, int n,
+ cham_store_t storev, int m, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
quark_option_t *opt = (quark_option_t*)(options->schedopt);
QUARK_Insert_Task(opt->quark, CORE_zgessq_quark, (Quark_Task_Flags*)opt,
- sizeof(int), &m, VALUE,
- sizeof(int), &n, VALUE,
+ sizeof(cham_store_t), &storev, VALUE,
+ sizeof(int), &m, VALUE,
+ sizeof(int), &n, VALUE,
sizeof(CHAMELEON_Complex64_t)*lda*n, RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An), INPUT,
- sizeof(int), &lda, VALUE,
+ sizeof(int), &lda, VALUE,
sizeof(double)*2, RTBLKADDR(SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn), INOUT,
0);
}
diff --git a/runtime/quark/codelets/codelet_zhessq.c b/runtime/quark/codelets/codelet_zhessq.c
index 5d797c984297b70b6bbbe570825ba76e069859f9..b66dfd60a9dcda3747993cacec4b4d33b22e6370 100644
--- a/runtime/quark/codelets/codelet_zhessq.c
+++ b/runtime/quark/codelets/codelet_zhessq.c
@@ -21,31 +21,13 @@
*/
#include "chameleon_quark.h"
#include "chameleon/tasks_z.h"
-#include "coreblas/coreblas_z.h"
-
-void CORE_zhessq_quark(Quark *quark)
-{
- cham_uplo_t uplo;
- int n;
- CHAMELEON_Complex64_t *A;
- int lda;
- double *SCALESUMSQ;
-
- quark_unpack_args_5( quark, uplo, n, A, lda, SCALESUMSQ );
- CORE_zhessq( uplo, n, A, lda, &SCALESUMSQ[0], &SCALESUMSQ[1]);
-}
void INSERT_TASK_zhessq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
- const CHAM_desc_t *A, int Am, int An, int lda,
- const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
+ cham_store_t storev, cham_uplo_t uplo, int n,
+ const CHAM_desc_t *A, int Am, int An, int lda,
+ const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
- quark_option_t *opt = (quark_option_t*)(options->schedopt);
- QUARK_Insert_Task(opt->quark, CORE_zhessq_quark, (Quark_Task_Flags*)opt,
- sizeof(int), &uplo, VALUE,
- sizeof(int), &n, VALUE,
- sizeof(CHAMELEON_Complex64_t)*lda*n, RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An), INPUT,
- sizeof(int), &lda, VALUE,
- sizeof(double)*2, RTBLKADDR(SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn), INOUT,
- 0);
+ INSERT_TASK_zsyssq( options, storev, uplo, n,
+ A, Am, An, lda,
+ SCALESUMSQ, SCALESUMSQm, SCALESUMSQn );
}
diff --git a/runtime/quark/codelets/codelet_zplssq.c b/runtime/quark/codelets/codelet_zplssq.c
index 8744b7b3ec319ce347abd8f40d101cb1a92b3e07..70810682c9a89cfba122782dbdff65149a17a09e 100644
--- a/runtime/quark/codelets/codelet_zplssq.c
+++ b/runtime/quark/codelets/codelet_zplssq.c
@@ -26,20 +26,15 @@
void CORE_zplssq_quark(Quark *quark)
{
+ cham_store_t storev;
+ int M;
+ int N;
double *SCLSSQ_IN;
double *SCLSSQ_OUT;
- quark_unpack_args_2( quark, SCLSSQ_IN, SCLSSQ_OUT );
+ quark_unpack_args_5( quark, storev, M, N, SCLSSQ_IN, SCLSSQ_OUT );
- assert( SCLSSQ_OUT[0] >= 0. );
- if( SCLSSQ_OUT[0] < SCLSSQ_IN[0] ) {
- SCLSSQ_OUT[1] = SCLSSQ_IN[1] + (SCLSSQ_OUT[1] * (( SCLSSQ_OUT[0] / SCLSSQ_IN[0] ) * ( SCLSSQ_OUT[0] / SCLSSQ_IN[0] )));
- SCLSSQ_OUT[0] = SCLSSQ_IN[0];
- } else {
- if ( SCLSSQ_OUT[0] > 0 ) {
- SCLSSQ_OUT[1] = SCLSSQ_OUT[1] + (SCLSSQ_IN[1] * (( SCLSSQ_IN[0] / SCLSSQ_OUT[0] ) * ( SCLSSQ_IN[0] / SCLSSQ_OUT[0] )));
- }
- }
+ CORE_zplssq(storev, M, N, SCLSSQ_IN, SCLSSQ_OUT);
}
/**
@@ -74,30 +69,48 @@ void CORE_zplssq_quark(Quark *quark)
*
*/
void INSERT_TASK_zplssq( const RUNTIME_option_t *options,
+ cham_store_t storev, int M, int N,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn,
const CHAM_desc_t *SCLSSQ, int SCLSSQm, int SCLSSQn )
{
+ int sizein = M*N;
+ int sizeout;
+
+ if ( storev == ChamColumnwise ) {
+ sizeout = 2*N;
+ } else if ( storev == ChamRowwise ) {
+ sizeout = 2*M;
+ } else {
+ sizeout = 2;
+ }
+
quark_option_t *opt = (quark_option_t*)(options->schedopt);
QUARK_Insert_Task(opt->quark, CORE_zplssq_quark, (Quark_Task_Flags*)opt,
- sizeof(double)*2, RTBLKADDR(SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn), INPUT,
- sizeof(double)*2, RTBLKADDR(SCLSSQ, double, SCLSSQm, SCLSSQn), INOUT,
+ sizeof(int), &storev, VALUE,
+ sizeof(int), &M, VALUE,
+ sizeof(int), &N, VALUE,
+ sizeof(double)*sizein, RTBLKADDR(SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn), INPUT,
+ sizeof(double)*sizeout, RTBLKADDR(SCLSSQ, double, SCLSSQm, SCLSSQn), INOUT,
0);
}
void CORE_zplssq2_quark(Quark *quark)
{
+ int N;
double *RESULT;
- quark_unpack_args_1( quark, RESULT );
- RESULT[0] = RESULT[0] * sqrt( RESULT[1] );
+ quark_unpack_args_2( quark, N, RESULT );
+
+ CORE_zplssq2(N, RESULT);
}
-void INSERT_TASK_zplssq2( const RUNTIME_option_t *options,
+void INSERT_TASK_zplssq2( const RUNTIME_option_t *options, int N,
const CHAM_desc_t *RESULT, int RESULTm, int RESULTn )
{
quark_option_t *opt = (quark_option_t*)(options->schedopt);
QUARK_Insert_Task(opt->quark, CORE_zplssq2_quark, (Quark_Task_Flags*)opt,
- sizeof(double)*2, RTBLKADDR(RESULT, double, RESULTm, RESULTn), INOUT,
+ sizeof(int), &N, VALUE,
+ sizeof(double)*2*N, RTBLKADDR(RESULT, double, RESULTm, RESULTn), INOUT,
0);
}
diff --git a/runtime/quark/codelets/codelet_zsyssq.c b/runtime/quark/codelets/codelet_zsyssq.c
index 9611e2082f9fd7c971f196d3318c66edb4353a31..cf1f39442c015139d869ae1fc08a5e4a79bd83a9 100644
--- a/runtime/quark/codelets/codelet_zsyssq.c
+++ b/runtime/quark/codelets/codelet_zsyssq.c
@@ -25,24 +25,26 @@
void CORE_zsyssq_quark(Quark *quark)
{
+ cham_store_t storev;
cham_uplo_t uplo;
int n;
CHAMELEON_Complex64_t *A;
int lda;
double *SCALESUMSQ;
- quark_unpack_args_5( quark, uplo, n, A, lda, SCALESUMSQ );
- CORE_zsyssq( uplo, n, A, lda, &SCALESUMSQ[0], &SCALESUMSQ[1]);
+ quark_unpack_args_6( quark, storev, uplo, n, A, lda, SCALESUMSQ );
+ CORE_zsyssq( storev, uplo, n, A, lda, SCALESUMSQ );
}
void INSERT_TASK_zsyssq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
+ cham_store_t storev, cham_uplo_t uplo, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
quark_option_t *opt = (quark_option_t*)(options->schedopt);
QUARK_Insert_Task(opt->quark, CORE_zsyssq_quark, (Quark_Task_Flags*)opt,
- sizeof(int), &uplo, VALUE,
+ sizeof(cham_store_t), &storev, VALUE,
+ sizeof(int), &uplo, VALUE,
sizeof(int), &n, VALUE,
sizeof(CHAMELEON_Complex64_t)*lda*n, RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An), INPUT,
sizeof(int), &lda, VALUE,
diff --git a/runtime/starpu/codelets/codelet_zcallback.c b/runtime/starpu/codelets/codelet_zcallback.c
index 35aea3122a2aa896ef9044c3a37c13191235296f..dddcb6994da86b889f7d484828d86a9d71cc9814 100644
--- a/runtime/starpu/codelets/codelet_zcallback.c
+++ b/runtime/starpu/codelets/codelet_zcallback.c
@@ -54,8 +54,8 @@ CHAMELEON_CL_CB(zsytrf_nopiv, starpu_matrix_get_nx(task->handles[0]), 0, 0,
CHAMELEON_CL_CB(zplgsy, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, M*N)
CHAMELEON_CL_CB(zplrnt, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, M*N)
CHAMELEON_CL_CB(zbuild, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, M*N)
-CHAMELEON_CL_CB(zplssq, 1, 1, 0, 4)
-CHAMELEON_CL_CB(zplssq2, 1, 1, 0, 1)
+CHAMELEON_CL_CB(zplssq, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, M*N)
+CHAMELEON_CL_CB(zplssq2, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, 2*N)
CHAMELEON_CL_CB(zpotrf, starpu_matrix_get_nx(task->handles[0]), 0, 0, (1./3.)*M* M*M)
CHAMELEON_CL_CB(zssssm, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_nx(task->handles[0]), M*M*(2.*M+starpu_matrix_get_nx(task->handles[2])))
CHAMELEON_CL_CB(zsymm, starpu_matrix_get_nx(task->handles[2]), starpu_matrix_get_ny(task->handles[2]), 0, 2.*M*M *N)
diff --git a/runtime/starpu/codelets/codelet_zgessq.c b/runtime/starpu/codelets/codelet_zgessq.c
index 3c37cb1c6858ab4af43475fd19e17f199d2cea65..c9141aefbd8aaa67be87429b7e0f29a34e2344eb 100644
--- a/runtime/starpu/codelets/codelet_zgessq.c
+++ b/runtime/starpu/codelets/codelet_zgessq.c
@@ -25,6 +25,7 @@
#if !defined(CHAMELEON_SIMULATION)
static void cl_zgessq_cpu_func(void *descr[], void *cl_arg)
{
+ cham_store_t storev;
int m;
int n;
CHAMELEON_Complex64_t *A;
@@ -33,8 +34,8 @@ static void cl_zgessq_cpu_func(void *descr[], void *cl_arg)
A = (CHAMELEON_Complex64_t *)STARPU_MATRIX_GET_PTR(descr[0]);
SCALESUMSQ = (double *)STARPU_MATRIX_GET_PTR(descr[1]);
- starpu_codelet_unpack_args(cl_arg, &m, &n, &lda);
- CORE_zgessq( m, n, A, lda, &SCALESUMSQ[0], &SCALESUMSQ[1] );
+ starpu_codelet_unpack_args(cl_arg, &storev, &m, &n, &lda);
+ CORE_zgessq( storev, m, n, A, lda, SCALESUMSQ );
}
#endif /* !defined(CHAMELEON_SIMULATION) */
@@ -44,7 +45,7 @@ static void cl_zgessq_cpu_func(void *descr[], void *cl_arg)
CODELETS_CPU(zgessq, 2, cl_zgessq_cpu_func)
void INSERT_TASK_zgessq( const RUNTIME_option_t *options,
- int m, int n,
+ cham_store_t storev, int m, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
@@ -58,6 +59,7 @@ void INSERT_TASK_zgessq( const RUNTIME_option_t *options,
starpu_insert_task(
starpu_mpi_codelet(codelet),
+ STARPU_VALUE, &storev, sizeof(cham_store_t),
STARPU_VALUE, &m, sizeof(int),
STARPU_VALUE, &n, sizeof(int),
STARPU_R, RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An),
diff --git a/runtime/starpu/codelets/codelet_zhessq.c b/runtime/starpu/codelets/codelet_zhessq.c
index 724f5e6fba78f76a62541b8e1a0e40f207716fe8..c1b18c8bc388df6890ff44a1e74058036987451c 100644
--- a/runtime/starpu/codelets/codelet_zhessq.c
+++ b/runtime/starpu/codelets/codelet_zhessq.c
@@ -22,51 +22,12 @@
#include "chameleon_starpu.h"
#include "runtime_codelet_z.h"
-#if !defined(CHAMELEON_SIMULATION)
-static void cl_zhessq_cpu_func(void *descr[], void *cl_arg)
-{
- cham_uplo_t uplo;
- int n;
- CHAMELEON_Complex64_t *A;
- int lda;
- double *SCALESUMSQ;
-
- A = (CHAMELEON_Complex64_t *)STARPU_MATRIX_GET_PTR(descr[0]);
- SCALESUMSQ = (double *)STARPU_MATRIX_GET_PTR(descr[1]);
- starpu_codelet_unpack_args(cl_arg, &uplo, &n, &lda);
- CORE_zhessq( uplo, n, A, lda, &SCALESUMSQ[0], &SCALESUMSQ[1] );
-}
-#endif /* !defined(CHAMELEON_SIMULATION) */
-
-/*
- * Codelet definition
- */
-CODELETS_CPU(zhessq, 2, cl_zhessq_cpu_func)
-
void INSERT_TASK_zhessq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
- const CHAM_desc_t *A, int Am, int An, int lda,
- const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
+ cham_store_t storev, cham_uplo_t uplo, int n,
+ const CHAM_desc_t *A, int Am, int An, int lda,
+ const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
- struct starpu_codelet *codelet = &cl_zhessq;
- void (*callback)(void*) = options->profiling ? cl_zgessq_callback : NULL;
-
- CHAMELEON_BEGIN_ACCESS_DECLARATION;
- CHAMELEON_ACCESS_R(A, Am, An);
- CHAMELEON_ACCESS_RW(SCALESUMSQ, SCALESUMSQm, SCALESUMSQn);
- CHAMELEON_END_ACCESS_DECLARATION;
-
- starpu_insert_task(
- starpu_mpi_codelet(codelet),
- STARPU_VALUE, &uplo, sizeof(int),
- STARPU_VALUE, &n, sizeof(int),
- STARPU_R, RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An),
- STARPU_VALUE, &lda, sizeof(int),
- STARPU_RW, RTBLKADDR(SCALESUMSQ, double, SCALESUMSQm, SCALESUMSQn),
- STARPU_PRIORITY, options->priority,
- STARPU_CALLBACK, callback,
-#if defined(CHAMELEON_CODELETS_HAVE_NAME)
- STARPU_NAME, "zhessq",
-#endif
- 0);
+ INSERT_TASK_zsyssq( options, storev, uplo, n,
+ A, Am, An, lda,
+ SCALESUMSQ, SCALESUMSQm, SCALESUMSQn );
}
diff --git a/runtime/starpu/codelets/codelet_zplssq.c b/runtime/starpu/codelets/codelet_zplssq.c
index 683aedf2f070bccb037af1139abdd9d3c24ab15c..96a41eb75bf9e30d9ad939acd8c29f89c71cc7c0 100644
--- a/runtime/starpu/codelets/codelet_zplssq.c
+++ b/runtime/starpu/codelets/codelet_zplssq.c
@@ -26,21 +26,17 @@
#if !defined(CHAMELEON_SIMULATION)
static void cl_zplssq_cpu_func(void *descr[], void *cl_arg)
{
+ cham_store_t storev;
+ int M;
+ int N;
double *SCLSSQ_IN;
double *SCLSSQ_OUT;
+ starpu_codelet_unpack_args(cl_arg, &storev, &M, &N);
SCLSSQ_IN = (double *)STARPU_MATRIX_GET_PTR(descr[0]);
SCLSSQ_OUT = (double *)STARPU_MATRIX_GET_PTR(descr[1]);
- assert( SCLSSQ_OUT[0] >= 0. );
- if( SCLSSQ_OUT[0] < SCLSSQ_IN[0] ) {
- SCLSSQ_OUT[1] = SCLSSQ_IN[1] + (SCLSSQ_OUT[1] * (( SCLSSQ_OUT[0] / SCLSSQ_IN[0] ) * ( SCLSSQ_OUT[0] / SCLSSQ_IN[0] )));
- SCLSSQ_OUT[0] = SCLSSQ_IN[0];
- } else {
- if ( SCLSSQ_OUT[0] > 0 ) {
- SCLSSQ_OUT[1] = SCLSSQ_OUT[1] + (SCLSSQ_IN[1] * (( SCLSSQ_IN[0] / SCLSSQ_OUT[0] ) * ( SCLSSQ_IN[0] / SCLSSQ_OUT[0] )));
- }
- }
+ CORE_zplssq(storev, M, N, SCLSSQ_IN, SCLSSQ_OUT);
(void)cl_arg;
}
@@ -51,40 +47,10 @@ static void cl_zplssq_cpu_func(void *descr[], void *cl_arg)
*/
CODELETS_CPU(zplssq, 2, cl_zplssq_cpu_func)
-/**
- *
- * @ingroup INSERT_TASK_Complex64_t
- *
- * @brief Compute sum( a_ij ^ 2 ) = scl * sqrt(ssq)
- *
- * with scl and ssq such that
- *
- * ( scl**2 )*ssq = sum( A( 2*i )**2 * A( 2*i+1 ) )
- * i
- *
- * The values of A(2*i+1) are assumed to be at least unity.
- * The values of A(2*i) are assumed to be non-negative and scl is
- *
- * scl = max( A( 2*i ) ),
- * i
- *
- * The routine makes only one pass through the matrix A.
- *
- *******************************************************************************
- *
- * @param[in] M
- * The number of couple (scale, sumsq) in the matrix A.
- *
- * @param[in] A
- * The 2-by-M matrix.
- *
- * @param[out] result
- * On exit, result contains scl * sqrt( ssq )
- *
- */
void INSERT_TASK_zplssq( const RUNTIME_option_t *options,
- const CHAM_desc_t *SCLSSQ_IN, int SCLSSQ_INm, int SCLSSQ_INn,
- const CHAM_desc_t *SCLSSQ_OUT, int SCLSSQ_OUTm, int SCLSSQ_OUTn )
+ cham_store_t storev, int M, int N,
+ const CHAM_desc_t *SCLSSQ_IN, int SCLSSQ_INm, int SCLSSQ_INn,
+ const CHAM_desc_t *SCLSSQ_OUT, int SCLSSQ_OUTm, int SCLSSQ_OUTn )
{
struct starpu_codelet *codelet = &cl_zplssq;
void (*callback)(void*) = options->profiling ? cl_zplssq_callback : NULL;
@@ -96,6 +62,9 @@ void INSERT_TASK_zplssq( const RUNTIME_option_t *options,
starpu_insert_task(
starpu_mpi_codelet(codelet),
+ STARPU_VALUE, &storev, sizeof(int),
+ STARPU_VALUE, &M, sizeof(int),
+ STARPU_VALUE, &N, sizeof(int),
STARPU_R, RTBLKADDR( SCLSSQ_IN, double, SCLSSQ_INm, SCLSSQ_INn ),
STARPU_RW, RTBLKADDR( SCLSSQ_OUT, double, SCLSSQ_OUTm, SCLSSQ_OUTn ),
STARPU_PRIORITY, options->priority,
@@ -109,11 +78,13 @@ void INSERT_TASK_zplssq( const RUNTIME_option_t *options,
#if !defined(CHAMELEON_SIMULATION)
static void cl_zplssq2_cpu_func(void *descr[], void *cl_arg)
{
+ int N;
double *RESULT;
+ starpu_codelet_unpack_args(cl_arg, &N);
RESULT = (double *)STARPU_MATRIX_GET_PTR(descr[0]);
- RESULT[0] = RESULT[0] * sqrt( RESULT[1] );
+ CORE_zplssq2(N, RESULT);
(void)cl_arg;
}
@@ -124,7 +95,7 @@ static void cl_zplssq2_cpu_func(void *descr[], void *cl_arg)
*/
CODELETS_CPU(zplssq2, 1, cl_zplssq2_cpu_func)
-void INSERT_TASK_zplssq2( const RUNTIME_option_t *options,
+void INSERT_TASK_zplssq2( const RUNTIME_option_t *options, int N,
const CHAM_desc_t *RESULT, int RESULTm, int RESULTn )
{
struct starpu_codelet *codelet = &cl_zplssq2;
@@ -136,6 +107,7 @@ void INSERT_TASK_zplssq2( const RUNTIME_option_t *options,
starpu_insert_task(
starpu_mpi_codelet(codelet),
+ STARPU_VALUE, &N, sizeof(int),
STARPU_RW, RTBLKADDR(RESULT, double, RESULTm, RESULTn),
STARPU_PRIORITY, options->priority,
STARPU_CALLBACK, callback,
diff --git a/runtime/starpu/codelets/codelet_zsyssq.c b/runtime/starpu/codelets/codelet_zsyssq.c
index 543b6e1b6b5ef298c47f9dee64c1481596df35a5..30980d669565b05567026c20649459be00ac8549 100644
--- a/runtime/starpu/codelets/codelet_zsyssq.c
+++ b/runtime/starpu/codelets/codelet_zsyssq.c
@@ -25,6 +25,7 @@
#if !defined(CHAMELEON_SIMULATION)
static void cl_zsyssq_cpu_func(void *descr[], void *cl_arg)
{
+ cham_store_t storev;
cham_uplo_t uplo;
int n;
CHAMELEON_Complex64_t *A;
@@ -33,8 +34,8 @@ static void cl_zsyssq_cpu_func(void *descr[], void *cl_arg)
A = (CHAMELEON_Complex64_t *)STARPU_MATRIX_GET_PTR(descr[0]);
SCALESUMSQ = (double *)STARPU_MATRIX_GET_PTR(descr[1]);
- starpu_codelet_unpack_args(cl_arg, &uplo, &n, &lda);
- CORE_zsyssq( uplo, n, A, lda, &SCALESUMSQ[0], &SCALESUMSQ[1] );
+ starpu_codelet_unpack_args(cl_arg, &storev, &uplo, &n, &lda);
+ CORE_zsyssq( storev, uplo, n, A, lda, SCALESUMSQ );
}
#endif /* !defined(CHAMELEON_SIMULATION) */
@@ -44,7 +45,7 @@ static void cl_zsyssq_cpu_func(void *descr[], void *cl_arg)
CODELETS_CPU(zsyssq, 2, cl_zsyssq_cpu_func)
void INSERT_TASK_zsyssq( const RUNTIME_option_t *options,
- cham_uplo_t uplo, int n,
+ cham_store_t storev, cham_uplo_t uplo, int n,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *SCALESUMSQ, int SCALESUMSQm, int SCALESUMSQn )
{
@@ -58,6 +59,7 @@ void INSERT_TASK_zsyssq( const RUNTIME_option_t *options,
starpu_insert_task(
starpu_mpi_codelet(codelet),
+ STARPU_VALUE, &storev, sizeof(cham_store_t),
STARPU_VALUE, &uplo, sizeof(int),
STARPU_VALUE, &n, sizeof(int),
STARPU_R, RTBLKADDR(A, CHAMELEON_Complex64_t, Am, An),