From e46b6449fb20f81c9446dd4c019a896946e986fc Mon Sep 17 00:00:00 2001
From: Mathieu Faverge <mathieu.faverge@inria.fr>
Date: Thu, 11 Apr 2019 23:43:00 +0200
Subject: [PATCH] Rules precision
---
CMakeLists.txt | 4 +-
cmake_modules/local_subs.py | 91 +++++++++
cmake_modules/morse_cmake | 2 +-
compute/zgels.c | 4 +-
compute/zgels_param.c | 4 +-
compute/zgesvd.c | 34 ++--
compute/zgetrs_incpiv.c | 4 +-
compute/zgetrs_nopiv.c | 4 +-
compute/zhetrd.c | 4 +-
compute/zposv.c | 4 +-
compute/zpotrf.c | 4 +-
compute/zpotri.c | 12 +-
compute/zpotrimm.c | 12 +-
compute/zpotrs.c | 6 +-
compute/zsysv.c | 4 +-
compute/zsytrf.c | 4 +-
compute/zsytrs.c | 6 +-
compute/zunmlq.c | 18 +-
compute/zunmlq_param.c | 18 +-
compute/zunmqr.c | 18 +-
compute/zunmqr_param.c | 18 +-
control/compute_z.h | 4 -
coreblas/compute/core_zherfb.c | 4 +-
coreblas/compute/core_zpamm.c | 2 +-
coreblas/compute/core_zparfb.c | 6 +-
coreblas/compute/core_zpemv.c | 6 +-
coreblas/compute/core_ztpmlqt.c | 14 +-
coreblas/compute/core_ztpmqrt.c | 14 +-
coreblas/compute/core_ztsmlq.c | 10 +-
coreblas/compute/core_ztsmlq_hetra1.c | 6 +-
coreblas/compute/core_ztsmqr.c | 10 +-
coreblas/compute/core_ztsmqr_hetra1.c | 6 +-
coreblas/compute/core_zttmlq.c | 10 +-
coreblas/compute/core_zttmqr.c | 10 +-
coreblas/compute/core_zunmlq.c | 10 +-
coreblas/compute/core_zunmqr.c | 10 +-
.../eztrace_module/coreblas_eztrace_module | 172 ------------------
coreblas/include/coreblas/coreblas_z.h | 43 -----
cudablas/compute/cuda_zparfb.c | 6 +-
cudablas/compute/cuda_ztpmlqt.c | 14 +-
cudablas/compute/cuda_ztpmqrt.c | 14 +-
runtime/CMakeLists.txt | 2 +-
.../{codelet_zasum.c => codelet_dzasum.c} | 4 +-
runtime/openmp/codelets/codelet_zunmlq.c | 10 +-
runtime/openmp/codelets/codelet_zunmqr.c | 10 +-
.../{codelet_zasum.c => codelet_dzasum.c} | 6 +-
.../{codelet_zasum.c => codelet_dzasum.c} | 4 +-
runtime/quark/codelets/codelet_zunmlq.c | 10 +-
runtime/quark/codelets/codelet_zunmqr.c | 10 +-
runtime/quark/include/core_blas_dag.h | 2 +
.../{codelet_zasum.c => codelet_dzasum.c} | 12 +-
runtime/starpu/codelets/codelet_zcallback.c | 2 +-
runtime/starpu/codelets/codelet_zunmlq.c | 10 +-
runtime/starpu/codelets/codelet_zunmqr.c | 10 +-
runtime/starpu/include/runtime_codelet_z.h | 114 ++++++------
runtime/starpu/include/runtime_codelets.h | 5 -
testing/lin/clagsy.f | 2 +-
testing/lin/clarhs.f | 6 +-
testing/lin/clatrs.f | 24 +--
testing/lin/cpocon.f | 4 +-
testing/lin/cporfs.f | 2 +-
testing/lin/cposvx.f | 12 +-
testing/lin/cpotri.f | 4 +-
testing/lin/dpocon.f | 4 +-
testing/lin/dporfs.f | 2 +-
testing/lin/dposvx.f | 12 +-
testing/lin/dpotri.f | 4 +-
testing/lin/spocon.f | 4 +-
testing/lin/sporfs.f | 2 +-
testing/lin/sposvx.f | 12 +-
testing/lin/spotri.f | 4 +-
testing/lin/zlagsy.f | 2 +-
testing/lin/zlarhs.f | 6 +-
testing/lin/zlatrs.f | 24 +--
testing/lin/zpocon.f | 4 +-
testing/lin/zporfs.f | 2 +-
testing/lin/zposvx.f | 12 +-
testing/lin/zpotri.f | 4 +-
timing/timing_zauxiliary.c | 4 +-
timing/timing_zauxiliary.h | 4 +-
80 files changed, 443 insertions(+), 574 deletions(-)
create mode 100644 cmake_modules/local_subs.py
rename runtime/openmp/codelets/{codelet_zasum.c => codelet_dzasum.c} (94%)
rename runtime/parsec/codelets/{codelet_zasum.c => codelet_dzasum.c} (95%)
rename runtime/quark/codelets/{codelet_zasum.c => codelet_dzasum.c} (96%)
rename runtime/starpu/codelets/{codelet_zasum.c => codelet_dzasum.c} (88%)
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 4d3309fee..4e2c41e4a 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -97,7 +97,7 @@ option(BUILD_SHARED_LIBS "Build shared libraries" OFF)
# Define precision supported by CHAMELEON
# -----------------------------------------
-set( RP_CHAMELEON_DICTIONNARY ${CHAMELEON_CMAKE_MODULE_PATH}/precision_generator/subs.py )
+set( RP_CHAMELEON_DICTIONNARY ${CMAKE_SOURCE_DIR}/cmake_modules/local_subs.py )
set( RP_CHAMELEON_PRECISIONS "s;d;c;z" )
include(RulesPrecisions)
@@ -586,7 +586,7 @@ endif(NOT CHAMELEON_SIMULATION)
# -------------------------------
if( CHAMELEON_SCHED_STARPU )
- set(CHAMELEON_STARPU_VERSION "1.1" CACHE STRING "oldest STARPU version desired")
+ set(CHAMELEON_STARPU_VERSION "1.3" CACHE STRING "oldest STARPU version desired")
# create list of components in order to make a single call to find_package(starpu...)
if(NOT CHAMELEON_SIMULATION)
diff --git a/cmake_modules/local_subs.py b/cmake_modules/local_subs.py
new file mode 100644
index 000000000..4608e59b8
--- /dev/null
+++ b/cmake_modules/local_subs.py
@@ -0,0 +1,91 @@
+_extra_blas = [
+ # ----- Additional BLAS
+ ('', 'dsgesv', 'dsgesv', 'zcgesv', 'zcgesv' ),
+ ('', 'sgesplit', 'dgesplit', 'cgesplit', 'zgesplit' ),
+ ('', 'slascal', 'dlascal', 'clascal', 'zlascal' ),
+ ('', 'slapack', 'dlapack', 'clapack', 'zlapack' ),
+ ('', 'stile', 'dtile', 'ctile', 'ztile' ),
+ ('', 'sgecon', 'dgecon', 'cgecon', 'zgecon' ),
+ ('', 'spocon', 'dpocon', 'cpocon', 'zpocon' ),
+ ('', 'strasm', 'dtrasm', 'ctrasm', 'ztrasm' ),
+ ('', 'sgecfi', 'dgecfi', 'cgecfi', 'zgecfi' ),
+ ('', 'splssq', 'dplssq', 'cplssq', 'zplssq' ),
+ ('', 'sy2sb', 'sy2sb' , 'he2hb', 'he2hb' ),
+ ('', 'she2ge', 'dhe2ge', 'che2ge', 'zhe2ge' ),
+ ('', 'slatro', 'dlatro', 'clatro', 'zlatro' ), #=> Replace by getmo/gecmo as in essl
+ ('', 'sbuild', 'dbuild', 'cbuild', 'zbuild' ), #=> Replace by map function
+]
+
+_extra_BLAS = [ [ x.upper() for x in row ] for row in _extra_blas ]
+
+subs = {
+ # ------------------------------------------------------------
+ # replacements applied to mixed precision files.
+ 'mixed' : [
+ # double/single, double/single-complex
+ #'12345678901234567890', '12345678901234567890')
+ (r'\bdouble', r'\bCHAMELEON_Complex64_t'),
+ (r'\bChamRealDouble', r'\bChamComplexDouble' ),
+ (r'\bfloat', r'\bCHAMELEON_Complex32_t'),
+ (r'\bChamRealFloat', r'\bChamComplexFloat' ),
+ (r'\breal\b', r'\bcomplex\b' ),
+
+ ('dsgels', 'zcgels' ),
+ ('dsorgesv', 'zcungesv' ),
+ ],
+ # ------------------------------------------------------------
+ # replacements applied to mixed precision files.
+ 'normal': [
+ # pattern single double single-complex double-complex
+ #'12345678901234567890', '12345678901234567890', '12345678901234567890', '12345678901234567890', '12345678901234567890')
+ ('int', 'float', 'double', 'CHAMELEON_Complex32_t', r'\bCHAMELEON_Complex64_t'),
+ ('ChamPattern', 'ChamRealFloat', 'ChamRealDouble', 'ChamComplexFloat', r'\bChamComplexDouble' ),
+ ('ChamPattern', 'ChamRealFloat', 'ChamRealDouble', 'ChamRealFloat', r'\bChamRealDouble' ),
+
+ # ----- Additional BLAS
+ ('', 'sTile', 'dTile', 'cTile', 'zTile' ),
+ ('', 'sLapack', 'dLapack', 'cLapack', 'zLapack' ),
+ ('', 'ORMQR', 'ORMQR', 'UNMQR', 'UNMQR' ),
+ ('', 'ORMLQ', 'ORMLQ', 'UNMLQ', 'UNMLQ' ),
+ ('', 'SYEV', 'SYEV', 'HEEV', 'HEEV' ),
+ ('', 'SYG', 'SYG', 'HEG', 'HEG' ),
+ ]
+ + _extra_blas
+ + _extra_BLAS
+ + [
+
+ # ----- For norms: compute result in Real or Double
+ ('', 'slange', 'dlange', 'slange', 'dlange' ),
+ ('', 'slaset', 'dlaset', 'slaset', 'dlaset' ),
+ ('', 'splssq', 'dplssq', 'splssq', 'dplssq' ),
+ ('', 'slacpy', 'dlacpy', 'slacpy', 'dlacpy' ),
+ ('', 'saxpy', 'daxpy', 'saxpy', 'daxpy' ),
+
+ (r'\b', r'szero\b', r'dzero\b', r'czero\b', r'zzero\b' ),
+# (r'\b', r'sone\b', r'done\b', r'cone\b', r'zone\b' ),
+
+ # ----- Chameleon Prefixes
+ ('CHAMELEON_P', 'CHAMELEON_S', 'CHAMELEON_D', 'CHAMELEON_C', 'CHAMELEON_Z' ),
+ ('RUNTIME_P', 'RUNTIME_s', 'RUNTIME_d', 'RUNTIME_c', 'RUNTIME_z' ),
+ ('chameleon_p', 'chameleon_s', 'chameleon_d', 'chameleon_c', 'chameleon_z' ),
+ ('codelet_p', 'codelet_s', 'codelet_d', 'codelet_c', 'codelet_z' ),
+ ('runtime_p', 'runtime_s', 'runtime_d', 'runtime_c', 'runtime_z' ),
+ ('testing_p', 'testing_s', 'testing_d', 'testing_c', 'testing_z' ),
+ ('timing_p', 'timing_s', 'timing_d', 'timing_c', 'timing_z' ),
+ ('workspace_p', 'workspace_s', 'workspace_d', 'workspace_c', 'workspace_z' ),
+# ('CORE_P', 'CORE_S', 'CORE_D', 'CORE_C', 'CORE_Z' ),
+# ('vec_p', 'vec_s', 'vec_d', 'vec_c', 'vec_z' ),
+
+ # ('', 'starpu_s', 'starpu_d', 'starpu_c', 'starpu_z' ),
+ # ('', 'STARPU_S', 'STARPU_D', 'STARPU_C', 'STARPU_Z' ),
+ # ('', 's_', 'd_', 'c_', 'z_' ),
+ # ('', 'S_', 'D_', 'C_', 'Z_' ),
+ # ('', 'FLT_EPSILON', 'DBL_EPSILON', 'FLT_EPSILON', 'DBL_EPSILON' ),
+ # ('', 's_RAFF_FLOAT', 'd_RAFF_FLOAT', 'c_RAFF_FLOAT', 'z_RAFF_FLOAT' ),
+ # # ----- unused?
+ # ('', 's_check', 'd_check', 'c_check', 'z_check' ),
+ # ('', 'stesting', 'dtesting', 'ctesting', 'ztesting' ),
+ # ('', 'SAUXILIARY', 'DAUXILIARY', 'CAUXILIARY', 'ZAUXILIARY' ),
+ # ('', 'sbuild', 'dbuild', 'cbuild', 'zbuild' ),
+ ]
+}
diff --git a/cmake_modules/morse_cmake b/cmake_modules/morse_cmake
index 33a182878..ade499661 160000
--- a/cmake_modules/morse_cmake
+++ b/cmake_modules/morse_cmake
@@ -1 +1 @@
-Subproject commit 33a182878f9049c47af1fce3e86e72b9a10e7f7a
+Subproject commit ade499661b58c71fe0586c2bbb98ea9725a88c52
diff --git a/compute/zgels.c b/compute/zgels.c
index 444021e55..2aed98e7d 100644
--- a/compute/zgels.c
+++ b/compute/zgels.c
@@ -48,7 +48,7 @@
* @param[in] trans
* Intended usage:
* = ChamNoTrans: the linear system involves A;
- * = ChamConjTrans: the linear system involves A**H.
+ * = ChamConjTrans: the linear system involves A^H.
* Currently only ChamNoTrans is supported.
*
* @param[in] M
@@ -218,7 +218,7 @@ int CHAMELEON_zgels( cham_trans_t trans, int M, int N, int NRHS,
* @param[in] trans
* Intended usage:
* = ChamNoTrans: the linear system involves A;
- * = ChamConjTrans: the linear system involves A**H.
+ * = ChamConjTrans: the linear system involves A^H.
* Currently only ChamNoTrans is supported.
*
* @param[in,out] A
diff --git a/compute/zgels_param.c b/compute/zgels_param.c
index fd5a99ddb..697e1b4ff 100644
--- a/compute/zgels_param.c
+++ b/compute/zgels_param.c
@@ -48,7 +48,7 @@
* @param[in] trans
* Intended usage:
* = ChamNoTrans: the linear system involves A;
- * = ChamConjTrans: the linear system involves A**H.
+ * = ChamConjTrans: the linear system involves A^H.
* Currently only ChamNoTrans is supported.
*
* @param[in] M
@@ -221,7 +221,7 @@ int CHAMELEON_zgels_param( const libhqr_tree_t *qrtree, cham_trans_t trans, int
* @param[in] trans
* Intended usage:
* = ChamNoTrans: the linear system involves A;
- * = ChamConjTrans: the linear system involves A**H.
+ * = ChamConjTrans: the linear system involves A^H.
* Currently only ChamNoTrans is supported.
*
* @param[in,out] A
diff --git a/compute/zgesvd.c b/compute/zgesvd.c
index c52fa59b3..7d12a10ba 100644
--- a/compute/zgesvd.c
+++ b/compute/zgesvd.c
@@ -44,7 +44,7 @@
* are returned in descending order. The first min(m,n) columns of
* U and V are the left and right singular vectors of A.
*
- * Note that the routine returns V**T, not V.
+ * Note that the routine returns V^T, not V.
*******************************************************************************
*
* @param[in] jobu
@@ -62,16 +62,16 @@
* NOT SUPPORTTED YET
*
* @param[in] jobvt
- * Specifies options for computing all or part of the matrix V**H.
+ * Specifies options for computing all or part of the matrix V^H.
* Intended usage:
- * = ChamVec = 'A'(lapack): all N rows of V**H are returned
+ * = ChamVec = 'A'(lapack): all N rows of V^H are returned
* in the array VT;
- * = ChamNoVec = 'N': no rows of V**H (no right singular vectors)
+ * = ChamNoVec = 'N': no rows of V^H (no right singular vectors)
* are computed.
- * = ChamSVec = 'S': the first min(m,n) rows of V**H (the right
+ * = ChamSVec = 'S': the first min(m,n) rows of V^H (the right
* singular vectors) are returned in the array VT;
* NOT SUPPORTTED YET
- * = ChamOVec = 'O': the first min(m,n) rows of V**H (the right
+ * = ChamOVec = 'O': the first min(m,n) rows of V^H (the right
* singular vectors) are overwritten on the array A;
* NOT SUPPORTTED YET
*
@@ -90,7 +90,7 @@
* columns of U (the left singular vectors,
* stored columnwise);
* if JOBVT = 'O', A is overwritten with the first min(m,n)
- * rows of V**H (the right singular vectors,
+ * rows of V^H (the right singular vectors,
* stored rowwise);
* if JOBU .ne. 'O' and JOBVT .ne. 'O', the contents of A
* are destroyed.
@@ -118,9 +118,9 @@
*
* @param[out] VT
* If JOBVT = 'A', VT contains the N-by-N unitary matrix
- * V**H;
+ * V^H;
* if JOBVT = 'S', VT contains the first min(m,n) rows of
- * V**H (the right singular vectors, stored rowwise);
+ * V^H (the right singular vectors, stored rowwise);
* if JOBVT = 'N' or 'O', VT is not referenced.
*
* @param[in] LDVT
@@ -259,16 +259,16 @@ int CHAMELEON_zgesvd( cham_job_t jobu, cham_job_t jobvt,
* NOT SUPPORTTED YET
*
* @param[in] jobvt
- * Specifies options for computing all or part of the matrix V**H.
+ * Specifies options for computing all or part of the matrix V^H.
* Intended usage:
- * = ChamVec = 'A'(lapack): all N rows of V**H are returned
+ * = ChamVec = 'A'(lapack): all N rows of V^H are returned
* in the array VT;
- * = ChamNoVec = 'N': no rows of V**H (no right singular vectors)
+ * = ChamNoVec = 'N': no rows of V^H (no right singular vectors)
* are computed.
- * = ChamSVec = 'S': the first min(m,n) rows of V**H (the right
+ * = ChamSVec = 'S': the first min(m,n) rows of V^H (the right
* singular vectors) are returned in the array VT;
* NOT SUPPORTTED YET
- * = ChamOVec = 'O': the first min(m,n) rows of V**H (the right
+ * = ChamOVec = 'O': the first min(m,n) rows of V^H (the right
* singular vectors) are overwritten on the array A;
* NOT SUPPORTTED YET
*
@@ -281,7 +281,7 @@ int CHAMELEON_zgesvd( cham_job_t jobu, cham_job_t jobvt,
* columns of U (the left singular vectors,
* stored columnwise);
* if JOBVT = 'O', A is overwritten with the first min(m,n)
- * rows of V**H (the right singular vectors,
+ * rows of V^H (the right singular vectors,
* stored rowwise);
* if JOBU .ne. 'O' and JOBVT .ne. 'O', the contents of A
* are destroyed.
@@ -306,9 +306,9 @@ int CHAMELEON_zgesvd( cham_job_t jobu, cham_job_t jobvt,
*
* @param[out] VT
* If JOBVT = 'A', VT contains the N-by-N unitary matrix
- * V**H;
+ * V^H;
* if JOBVT = 'S', VT contains the first min(m,n) rows of
- * V**H (the right singular vectors, stored rowwise);
+ * V^H (the right singular vectors, stored rowwise);
* if JOBVT = 'N' or 'O', VT is not referenced.
*
* @param[in] LDVT
diff --git a/compute/zgetrs_incpiv.c b/compute/zgetrs_incpiv.c
index d3e6324f7..225cb125f 100644
--- a/compute/zgetrs_incpiv.c
+++ b/compute/zgetrs_incpiv.c
@@ -37,8 +37,8 @@
* @param[in] trans
* Intended to specify the the form of the system of equations:
* = ChamNoTrans: A * X = B (No transpose)
- * = ChamTrans: A**T * X = B (Transpose)
- * = ChamConjTrans: A**H * X = B (Conjugate transpose)
+ * = ChamTrans: A^T * X = B (Transpose)
+ * = ChamConjTrans: A^H * X = B (Conjugate transpose)
* Currently only ChamNoTrans is supported.
*
* @param[in] N
diff --git a/compute/zgetrs_nopiv.c b/compute/zgetrs_nopiv.c
index bb0de9875..3a3dfe360 100644
--- a/compute/zgetrs_nopiv.c
+++ b/compute/zgetrs_nopiv.c
@@ -38,8 +38,8 @@
* @param[in] trans
* Intended to specify the the form of the system of equations:
* = ChamNoTrans: A * X = B (No transpose)
- * = ChamTrans: A**T * X = B (Transpose)
- * = ChamConjTrans: A**H * X = B (Conjugate transpose)
+ * = ChamTrans: A^T * X = B (Transpose)
+ * = ChamConjTrans: A^H * X = B (Conjugate transpose)
* Currently only ChamNoTrans is supported.
*
* @param[in] N
diff --git a/compute/zhetrd.c b/compute/zhetrd.c
index 0c7e3c66b..43c4fc59d 100644
--- a/compute/zhetrd.c
+++ b/compute/zhetrd.c
@@ -34,7 +34,7 @@
* tridiagonal form S using a two-stage approach
* First stage: reduction to band tridiagonal form (unitary Q1);
* Second stage: reduction from band to tridiagonal form (unitary
- * Q2). Let Q = Q1 * Q2 be the global unitary transformation; Q**H *
+ * Q2). Let Q = Q1 * Q2 be the global unitary transformation; Q^H *
* A * Q = S.
*
*******************************************************************************
@@ -190,7 +190,7 @@ int CHAMELEON_zhetrd( cham_job_t jobz, cham_uplo_t uplo, int N,
* First stage: reduction to band tridiagonal form (unitary Q1);
* Second stage: reduction from band to tridiagonal form (unitary Q2).
* Let Q = Q1 * Q2 be the global unitary transformation;
- * Q**H * A * Q = S.
+ * Q^H * A * Q = S.
* Tile equivalent of CHAMELEON_zhetrd().
* Operates on matrices stored by tiles.
* All matrices are passed through descriptors.
diff --git a/compute/zposv.c b/compute/zposv.c
index 27ab796eb..317f21f14 100644
--- a/compute/zposv.c
+++ b/compute/zposv.c
@@ -61,7 +61,7 @@
* triangular part of the matrix A, and the strictly upper triangular part of A is not
* referenced.
* On exit, if return value = 0, the factor U or L from the Cholesky factorization
- * A = U**H*U or A = L*L**H.
+ * A = U^H*U or A = L*L^H.
*
* @param[in] LDA
* The leading dimension of the array A. LDA >= max(1,N).
@@ -197,7 +197,7 @@ int CHAMELEON_zposv( cham_uplo_t uplo, int N, int NRHS,
* triangular part of the matrix A, and the strictly upper triangular part of A is not
* referenced.
* On exit, if return value = 0, the factor U or L from the Cholesky factorization
- * A = U**H*U or A = L*L**H.
+ * A = U^H*U or A = L*L^H.
*
* @param[in,out] B
* On entry, the N-by-NRHS right hand side matrix B.
diff --git a/compute/zpotrf.c b/compute/zpotrf.c
index c898ca145..41093b0ca 100644
--- a/compute/zpotrf.c
+++ b/compute/zpotrf.c
@@ -55,7 +55,7 @@
* triangular part of the matrix A, and the strictly upper triangular part of A is not
* referenced.
* On exit, if return value = 0, the factor U or L from the Cholesky factorization
- * A = U**H*U or A = L*L**H.
+ * A = U^H*U or A = L*L^H.
*
* @param[in] LDA
* The leading dimension of the array A. LDA >= max(1,N).
@@ -170,7 +170,7 @@ int CHAMELEON_zpotrf( cham_uplo_t uplo, int N,
* triangular part of the matrix A, and the strictly upper triangular part of A is not
* referenced.
* On exit, if return value = 0, the factor U or L from the Cholesky factorization
- * A = U**H*U or A = L*L**H.
+ * A = U^H*U or A = L*L^H.
*
*******************************************************************************
*
diff --git a/compute/zpotri.c b/compute/zpotri.c
index eb1e95d61..899e13b89 100644
--- a/compute/zpotri.c
+++ b/compute/zpotri.c
@@ -29,7 +29,7 @@
* @ingroup CHAMELEON_Complex64_t
*
* CHAMELEON_zpotri - Computes the inverse of a complex Hermitian positive definite
- * matrix A using the Cholesky factorization A = U**H*U or A = L*L**H
+ * matrix A using the Cholesky factorization A = U^H*U or A = L*L^H
* computed by CHAMELEON_zpotrf.
*
*******************************************************************************
@@ -43,7 +43,7 @@
*
* @param[in,out] A
* On entry, the triangular factor U or L from the Cholesky
- * factorization A = U**H*U or A = L*L**H, as computed by
+ * factorization A = U^H*U or A = L*L^H, as computed by
* CHAMELEON_zpotrf.
* On exit, the upper or lower triangle of the (Hermitian)
* inverse of A, overwriting the input factor U or L.
@@ -140,7 +140,7 @@ int CHAMELEON_zpotri( cham_uplo_t uplo, int N,
*
* CHAMELEON_zpotri_Tile - Computes the inverse of a complex Hermitian
* positive definite matrix A using the Cholesky factorization
- * A = U**H*U or A = L*L**H computed by CHAMELEON_zpotrf.
+ * A = U^H*U or A = L*L^H computed by CHAMELEON_zpotrf.
* Tile equivalent of CHAMELEON_zpotri().
* Operates on matrices stored by tiles.
* All matrices are passed through descriptors.
@@ -154,7 +154,7 @@ int CHAMELEON_zpotri( cham_uplo_t uplo, int N,
*
* @param[in] A
* On entry, the triangular factor U or L from the Cholesky
- * factorization A = U**H*U or A = L*L**H, as computed by
+ * factorization A = U^H*U or A = L*L^H, as computed by
* CHAMELEON_zpotrf.
* On exit, the upper or lower triangle of the (Hermitian)
* inverse of A, overwriting the input factor U or L.
@@ -206,8 +206,8 @@ int CHAMELEON_zpotri_Tile( cham_uplo_t uplo, CHAM_desc_t *A )
* @ingroup CHAMELEON_Complex64_t_Tile_Async
*
* CHAMELEON_zpotri_Tile_Async - Computes the inverse of a complex Hermitian
- * positive definite matrix A using the Cholesky factorization A = U**H*U
- * or A = L*L**H computed by CHAMELEON_zpotrf.
+ * positive definite matrix A using the Cholesky factorization A = U^H*U
+ * or A = L*L^H computed by CHAMELEON_zpotrf.
* Non-blocking equivalent of CHAMELEON_zpotri_Tile().
* May return before the computation is finished.
* Allows for pipelining of operations at runtime.
diff --git a/compute/zpotrimm.c b/compute/zpotrimm.c
index f903d52e9..7d5cda488 100644
--- a/compute/zpotrimm.c
+++ b/compute/zpotrimm.c
@@ -29,7 +29,7 @@
* @ingroup CHAMELEON_Complex64_t
*
* CHAMELEON_zpotrimm - Computes the inverse of a complex Hermitian positive definite
- * matrix A using the Cholesky factorization A = U**H*U or A = L*L**H
+ * matrix A using the Cholesky factorization A = U^H*U or A = L*L^H
* computed by CHAMELEON_zpotrf.
*
*******************************************************************************
@@ -43,7 +43,7 @@
*
* @param[in,out] A
* On entry, the triangular factor U or L from the Cholesky
- * factorization A = U**H*U or A = L*L**H, as computed by
+ * factorization A = U^H*U or A = L*L^H, as computed by
* CHAMELEON_zpotrf.
* On exit, the upper or lower triangle of the (Hermitian)
* inverse of A, overwriting the input factor U or L.
@@ -162,7 +162,7 @@ int CHAMELEON_zpotrimm( cham_uplo_t uplo, int N,
*
* CHAMELEON_zpotrimm_Tile - Computes the inverse of a complex Hermitian
* positive definite matrix A using the Cholesky factorization
- * A = U**H*U or A = L*L**H computed by CHAMELEON_zpotrf.
+ * A = U^H*U or A = L*L^H computed by CHAMELEON_zpotrf.
* Tile equivalent of CHAMELEON_zpotrimm().
* Operates on matrices stored by tiles.
* All matrices are passed through descriptors.
@@ -176,7 +176,7 @@ int CHAMELEON_zpotrimm( cham_uplo_t uplo, int N,
*
* @param[in] A
* On entry, the triangular factor U or L from the Cholesky
- * factorization A = U**H*U or A = L*L**H, as computed by
+ * factorization A = U^H*U or A = L*L^H, as computed by
* CHAMELEON_zpotrf.
* On exit, the upper or lower triangle of the (Hermitian)
* inverse of A, overwriting the input factor U or L.
@@ -230,8 +230,8 @@ int CHAMELEON_zpotrimm_Tile( cham_uplo_t uplo, CHAM_desc_t *A, CHAM_desc_t *B, C
* @ingroup CHAMELEON_Complex64_t_Tile_Async
*
* CHAMELEON_zpotrimm_Tile_Async - Computes the inverse of a complex Hermitian
- * positive definite matrix A using the Cholesky factorization A = U**H*U
- * or A = L*L**H computed by CHAMELEON_zpotrf.
+ * positive definite matrix A using the Cholesky factorization A = U^H*U
+ * or A = L*L^H computed by CHAMELEON_zpotrf.
* Non-blocking equivalent of CHAMELEON_zpotrimm_Tile().
* May return before the computation is finished.
* Allows for pipelining of operations at runtime.
diff --git a/compute/zpotrs.c b/compute/zpotrs.c
index 72bd62a7f..1d290f64f 100644
--- a/compute/zpotrs.c
+++ b/compute/zpotrs.c
@@ -31,7 +31,7 @@
*
* CHAMELEON_zpotrs - Solves a system of linear equations A * X = B with a symmetric positive
* definite (or Hermitian positive definite in the complex case) matrix A using the Cholesky
- * factorization A = U**H*U or A = L*L**H computed by CHAMELEON_zpotrf.
+ * factorization A = U^H*U or A = L*L^H computed by CHAMELEON_zpotrf.
*
*******************************************************************************
*
@@ -46,7 +46,7 @@
* The number of right hand sides, i.e., the number of columns of the matrix B. NRHS >= 0.
*
* @param[in] A
- * The triangular factor U or L from the Cholesky factorization A = U**H*U or A = L*L**H,
+ * The triangular factor U or L from the Cholesky factorization A = U^H*U or A = L*L^H,
* computed by CHAMELEON_zpotrf.
*
* @param[in] LDA
@@ -173,7 +173,7 @@ int CHAMELEON_zpotrs( cham_uplo_t uplo, int N, int NRHS,
* = ChamLower: Lower triangle of A is stored.
*
* @param[in] A
- * The triangular factor U or L from the Cholesky factorization A = U**H*U or A = L*L**H,
+ * The triangular factor U or L from the Cholesky factorization A = U^H*U or A = L*L^H,
* computed by CHAMELEON_zpotrf.
*
* @param[in,out] B
diff --git a/compute/zsysv.c b/compute/zsysv.c
index 256e27de9..ebee2c6f1 100644
--- a/compute/zsysv.c
+++ b/compute/zsysv.c
@@ -62,7 +62,7 @@
* triangular part of the matrix A, and the strictly upper triangular part of A is not
* referenced.
* On exit, if return value = 0, the factor U or L from the Cholesky factorization
- * A = U**T*U or A = L*L**T.
+ * A = U^T*U or A = L*L^T.
*
* @param[in] LDA
* The leading dimension of the array A. LDA >= max(1,N).
@@ -194,7 +194,7 @@ int CHAMELEON_zsysv( cham_uplo_t uplo, int N, int NRHS,
* triangular part of the matrix A, and the strictly upper triangular part of A is not
* referenced.
* On exit, if return value = 0, the factor U or L from the Cholesky factorization
- * A = U**T*U or A = L*L**T.
+ * A = U^T*U or A = L*L^T.
*
* @param[in,out] B
* On entry, the N-by-NRHS right hand side matrix B.
diff --git a/compute/zsytrf.c b/compute/zsytrf.c
index e32b1f8f4..b603ddde5 100644
--- a/compute/zsytrf.c
+++ b/compute/zsytrf.c
@@ -50,7 +50,7 @@
* triangular part of the matrix A, and the strictly upper triangular part of A is not
* referenced.
* On exit, if return value = 0, the factor U or L from the Cholesky factorization
- * A = U**H*U or A = L*L**H.
+ * A = U^H*U or A = L*L^H.
*
* @param[in] LDA
* The leading dimension of the array A. LDA >= max(1,N).
@@ -164,7 +164,7 @@ int CHAMELEON_zsytrf( cham_uplo_t uplo, int N,
* triangular part of the matrix A, and the strictly upper triangular part of A is not
* referenced.
* On exit, if return value = 0, the factor U or L from the Cholesky factorization
- * A = U**T*U or A = L*L**T.
+ * A = U^T*U or A = L*L^T.
*
*******************************************************************************
*
diff --git a/compute/zsytrs.c b/compute/zsytrs.c
index b50289edd..84a2c778f 100644
--- a/compute/zsytrs.c
+++ b/compute/zsytrs.c
@@ -33,7 +33,7 @@
*
* CHAMELEON_zsytrs - Solves a system of linear equations A * X = B with a complex
* symmetric matrix A using the Cholesky factorization
- * A = U**H*U or A = L*L**H computed by CHAMELEON_zsytrf.
+ * A = U^H*U or A = L*L^H computed by CHAMELEON_zsytrf.
*
*******************************************************************************
*
@@ -48,7 +48,7 @@
* The number of right hand sides, i.e., the number of columns of the matrix B. NRHS >= 0.
*
* @param[in] A
- * The triangular factor U or L from the Cholesky factorization A = U**T*U or A = L*L**T,
+ * The triangular factor U or L from the Cholesky factorization A = U^T*U or A = L*L^T,
* computed by CHAMELEON_zsytrf.
*
* @param[in] LDA
@@ -172,7 +172,7 @@ int CHAMELEON_zsytrs( cham_uplo_t uplo, int N, int NRHS,
* = ChamLower: Lower triangle of A is stored.
*
* @param[in] A
- * The triangular factor U or L from the Cholesky factorization A = U**T*U or A = L*L**T,
+ * The triangular factor U or L from the Cholesky factorization A = U^T*U or A = L*L^T,
* computed by CHAMELEON_zsytrf.
*
* @param[in,out] B
diff --git a/compute/zunmlq.c b/compute/zunmlq.c
index 51506cdaa..bbc4a6a77 100644
--- a/compute/zunmlq.c
+++ b/compute/zunmlq.c
@@ -35,7 +35,7 @@
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -49,13 +49,13 @@
*
* @param[in] side
* Intended usage:
- * = ChamLeft: apply Q or Q**H from the left;
- * = ChamRight: apply Q or Q**H from the right.
+ * = ChamLeft: apply Q or Q^H from the left;
+ * = ChamRight: apply Q or Q^H from the right.
*
* @param[in] trans
* Intended usage:
* = ChamNoTrans: no transpose, apply Q;
- * = ChamConjTrans: conjugate transpose, apply Q**H.
+ * = ChamConjTrans: conjugate transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the matrix C. M >= 0.
@@ -79,7 +79,7 @@
*
* @param[in,out] C
* On entry, the M-by-N matrix C.
- * On exit, C is overwritten by Q*C or Q**H*C.
+ * On exit, C is overwritten by Q*C or Q^H*C.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
@@ -211,14 +211,14 @@ int CHAMELEON_zunmlq( cham_side_t side, cham_trans_t trans, int M, int N, int K,
*
* @param[in] side
* Intended usage:
- * = ChamLeft: apply Q or Q**H from the left;
- * = ChamRight: apply Q or Q**H from the right.
+ * = ChamLeft: apply Q or Q^H from the left;
+ * = ChamRight: apply Q or Q^H from the right.
* Currently only ChamLeft is supported.
*
* @param[in] trans
* Intended usage:
* = ChamNoTrans: no transpose, apply Q;
- * = ChamConjTrans: conjugate transpose, apply Q**H.
+ * = ChamConjTrans: conjugate transpose, apply Q^H.
* Currently only ChamConjTrans is supported.
*
* @param[in] A
@@ -229,7 +229,7 @@ int CHAMELEON_zunmlq( cham_side_t side, cham_trans_t trans, int M, int N, int K,
*
* @param[in,out] C
* On entry, the M-by-N matrix C.
- * On exit, C is overwritten by Q*C or Q**H*C.
+ * On exit, C is overwritten by Q*C or Q^H*C.
*
*******************************************************************************
*
diff --git a/compute/zunmlq_param.c b/compute/zunmlq_param.c
index 7199aed91..df0fba8ba 100644
--- a/compute/zunmlq_param.c
+++ b/compute/zunmlq_param.c
@@ -29,7 +29,7 @@
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -46,13 +46,13 @@
*
* @param[in] side
* Intended usage:
- * = ChamLeft: apply Q or Q**H from the left;
- * = ChamRight: apply Q or Q**H from the right.
+ * = ChamLeft: apply Q or Q^H from the left;
+ * = ChamRight: apply Q or Q^H from the right.
*
* @param[in] trans
* Intended usage:
* = ChamNoTrans: no transpose, apply Q;
- * = ChamConjTrans: conjugate transpose, apply Q**H.
+ * = ChamConjTrans: conjugate transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the matrix C. M >= 0.
@@ -79,7 +79,7 @@
*
* @param[in,out] C
* On entry, the M-by-N matrix C.
- * On exit, C is overwritten by Q*C or Q**H*C.
+ * On exit, C is overwritten by Q*C or Q^H*C.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
@@ -210,14 +210,14 @@ int CHAMELEON_zunmlq_param( const libhqr_tree_t *qrtree, cham_side_t side, cham_
*
* @param[in] side
* Intended usage:
- * = ChamLeft: apply Q or Q**H from the left;
- * = ChamRight: apply Q or Q**H from the right.
+ * = ChamLeft: apply Q or Q^H from the left;
+ * = ChamRight: apply Q or Q^H from the right.
* Currently only ChamLeft is supported.
*
* @param[in] trans
* Intended usage:
* = ChamNoTrans: no transpose, apply Q;
- * = ChamConjTrans: conjugate transpose, apply Q**H.
+ * = ChamConjTrans: conjugate transpose, apply Q^H.
* Currently only ChamConjTrans is supported.
*
* @param[in] A
@@ -228,7 +228,7 @@ int CHAMELEON_zunmlq_param( const libhqr_tree_t *qrtree, cham_side_t side, cham_
*
* @param[in,out] C
* On entry, the M-by-N matrix C.
- * On exit, C is overwritten by Q*C or Q**H*C.
+ * On exit, C is overwritten by Q*C or Q^H*C.
*
*******************************************************************************
*
diff --git a/compute/zunmqr.c b/compute/zunmqr.c
index c7adba166..4139f579a 100644
--- a/compute/zunmqr.c
+++ b/compute/zunmqr.c
@@ -34,7 +34,7 @@
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -48,13 +48,13 @@
*
* @param[in] side
* Intended usage:
- * = ChamLeft: apply Q or Q**H from the left;
- * = ChamRight: apply Q or Q**H from the right.
+ * = ChamLeft: apply Q or Q^H from the left;
+ * = ChamRight: apply Q or Q^H from the right.
*
* @param[in] trans
* Intended usage:
* = ChamNoTrans: no transpose, apply Q;
- * = ChamConjTrans: conjugate transpose, apply Q**H.
+ * = ChamConjTrans: conjugate transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the matrix C. M >= 0.
@@ -81,7 +81,7 @@
*
* @param[in,out] C
* On entry, the M-by-N matrix C.
- * On exit, C is overwritten by Q*C or Q**H*C.
+ * On exit, C is overwritten by Q*C or Q^H*C.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
@@ -212,14 +212,14 @@ int CHAMELEON_zunmqr( cham_side_t side, cham_trans_t trans, int M, int N, int K,
*
* @param[in] side
* Intended usage:
- * = ChamLeft: apply Q or Q**H from the left;
- * = ChamRight: apply Q or Q**H from the right.
+ * = ChamLeft: apply Q or Q^H from the left;
+ * = ChamRight: apply Q or Q^H from the right.
* Currently only ChamLeft is supported.
*
* @param[in] trans
* Intended usage:
* = ChamNoTrans: no transpose, apply Q;
- * = ChamConjTrans: conjugate transpose, apply Q**H.
+ * = ChamConjTrans: conjugate transpose, apply Q^H.
* Currently only ChamConjTrans is supported.
*
* @param[in] A
@@ -231,7 +231,7 @@ int CHAMELEON_zunmqr( cham_side_t side, cham_trans_t trans, int M, int N, int K,
*
* @param[in,out] C
* On entry, the M-by-N matrix C.
- * On exit, C is overwritten by Q*C or Q**H*C.
+ * On exit, C is overwritten by Q*C or Q^H*C.
*
*******************************************************************************
*
diff --git a/compute/zunmqr_param.c b/compute/zunmqr_param.c
index 152257e77..e7b31e742 100644
--- a/compute/zunmqr_param.c
+++ b/compute/zunmqr_param.c
@@ -29,7 +29,7 @@
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -46,13 +46,13 @@
*
* @param[in] side
* Intended usage:
- * = ChamLeft: apply Q or Q**H from the left;
- * = ChamRight: apply Q or Q**H from the right.
+ * = ChamLeft: apply Q or Q^H from the left;
+ * = ChamRight: apply Q or Q^H from the right.
*
* @param[in] trans
* Intended usage:
* = ChamNoTrans: no transpose, apply Q;
- * = ChamConjTrans: conjugate transpose, apply Q**H.
+ * = ChamConjTrans: conjugate transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the matrix C. M >= 0.
@@ -82,7 +82,7 @@
*
* @param[in,out] C
* On entry, the M-by-N matrix C.
- * On exit, C is overwritten by Q*C or Q**H*C.
+ * On exit, C is overwritten by Q*C or Q^H*C.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
@@ -215,14 +215,14 @@ int CHAMELEON_zunmqr_param( const libhqr_tree_t *qrtree,
*
* @param[in] side
* Intended usage:
- * = ChamLeft: apply Q or Q**H from the left;
- * = ChamRight: apply Q or Q**H from the right.
+ * = ChamLeft: apply Q or Q^H from the left;
+ * = ChamRight: apply Q or Q^H from the right.
* Currently only ChamLeft is supported.
*
* @param[in] trans
* Intended usage:
* = ChamNoTrans: no transpose, apply Q;
- * = ChamConjTrans: conjugate transpose, apply Q**H.
+ * = ChamConjTrans: conjugate transpose, apply Q^H.
* Currently only ChamConjTrans is supported.
*
* @param[in] A
@@ -234,7 +234,7 @@ int CHAMELEON_zunmqr_param( const libhqr_tree_t *qrtree,
*
* @param[in,out] C
* On entry, the M-by-N matrix C.
- * On exit, C is overwritten by Q*C or Q**H*C.
+ * On exit, C is overwritten by Q*C or Q^H*C.
*
*******************************************************************************
*
diff --git a/control/compute_z.h b/control/compute_z.h
index 6a78fe620..e34ec4dfc 100644
--- a/control/compute_z.h
+++ b/control/compute_z.h
@@ -35,10 +35,6 @@ int chameleon_zshift(CHAM_context_t *chamctxt, int m, int n, CHAMELEON_Complex64
/**
* Declarations of parallel functions (dynamic scheduling) - alphabetical order
*/
-void chameleon_pzbarrier_pnl2tl(CHAM_desc_t *A, RUNTIME_sequence_t *sequence, RUNTIME_request_t *request);
-void chameleon_pzbarrier_row2tl(CHAM_desc_t *A, RUNTIME_sequence_t *sequence, RUNTIME_request_t *request);
-void chameleon_pzbarrier_tl2pnl(CHAM_desc_t *A, RUNTIME_sequence_t *sequence, RUNTIME_request_t *request);
-void chameleon_pzbarrier_tl2row(CHAM_desc_t *A, RUNTIME_sequence_t *sequence, RUNTIME_request_t *request);
void chameleon_pzgebrd_gb2bd(cham_uplo_t uplo, CHAM_desc_t *A, double *D, double *E, CHAM_desc_t *T, RUNTIME_sequence_t *sequence, RUNTIME_request_t *request);
void chameleon_pzgebrd_ge2gb( int genD, CHAM_desc_t *A, CHAM_desc_t *T, CHAM_desc_t *D, RUNTIME_sequence_t *sequence, RUNTIME_request_t *request);
void chameleon_pzgelqf( int genD, CHAM_desc_t *A, CHAM_desc_t *T, CHAM_desc_t *D, RUNTIME_sequence_t *sequence, RUNTIME_request_t *request);
diff --git a/coreblas/compute/core_zherfb.c b/coreblas/compute/core_zherfb.c
index d71c18028..a54de4a5a 100644
--- a/coreblas/compute/core_zherfb.c
+++ b/coreblas/compute/core_zherfb.c
@@ -27,7 +27,7 @@
*
* CORE_zherfb overwrites the symmetric complex N-by-N tile C with
*
- * Q**T*C*Q
+ * Q^T*C*Q
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -72,7 +72,7 @@
*
* @param[in,out] C
* On entry, the symmetric N-by-N tile C.
- * On exit, C is overwritten by Q**T*C*Q.
+ * On exit, C is overwritten by Q^T*C*Q.
*
* @param[in] ldc
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/coreblas/compute/core_zpamm.c b/coreblas/compute/core_zpamm.c
index eb995720c..01a25ea55 100644
--- a/coreblas/compute/core_zpamm.c
+++ b/coreblas/compute/core_zpamm.c
@@ -52,7 +52,7 @@ static inline int CORE_zpamm_w(cham_side_t side, cham_trans_t trans, cham_uplo_t
*
* where op( V ) is one of
*
- * op( V ) = V or op( V ) = V**T or op( V ) = V**H,
+ * op( V ) = V or op( V ) = V^T or op( V ) = V^H,
*
* A1, A2 and W are general matrices, and V is:
*
diff --git a/coreblas/compute/core_zparfb.c b/coreblas/compute/core_zparfb.c
index 4b2246994..9bd1e809a 100644
--- a/coreblas/compute/core_zparfb.c
+++ b/coreblas/compute/core_zparfb.c
@@ -52,12 +52,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] direct
* Indicates how H is formed from a product of elementary
diff --git a/coreblas/compute/core_zpemv.c b/coreblas/compute/core_zpemv.c
index f96ef1e14..0144020a7 100644
--- a/coreblas/compute/core_zpemv.c
+++ b/coreblas/compute/core_zpemv.c
@@ -36,7 +36,7 @@
*
* where op( A ) is one of
*
- * op( A ) = A or op( A ) = A**T or op( A ) = A**H,
+ * op( A ) = A or op( A ) = A^T or op( A ) = A^H,
*
* alpha and beta are scalars, x and y are vectors and A is a
* pentagonal matrix (see further details).
@@ -52,8 +52,8 @@
* @param[in] trans
*
* @arg ChamNoTrans : y := alpha*A*x + beta*y.
- * @arg ChamTrans : y := alpha*A**T*x + beta*y.
- * @arg ChamConjTrans : y := alpha*A**H*x + beta*y.
+ * @arg ChamTrans : y := alpha*A^T*x + beta*y.
+ * @arg ChamConjTrans : y := alpha*A^H*x + beta*y.
*
* @param[in] M
* Number of rows of the matrix A.
diff --git a/coreblas/compute/core_ztpmlqt.c b/coreblas/compute/core_ztpmlqt.c
index 7a0ecf7ae..a15c7db3d 100644
--- a/coreblas/compute/core_ztpmlqt.c
+++ b/coreblas/compute/core_ztpmlqt.c
@@ -33,12 +33,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile B. M >= 0.
@@ -78,7 +78,7 @@
* or (LDA,K) if SIDE = ChamRight
* On entry, the K-by-N or M-by-K matrix A.
* On exit, A is overwritten by the corresponding block of
- * Q*C or Q**H*C or C*Q or C*Q**H. See Further Details.
+ * Q*C or Q^H*C or C*Q or C*Q^H. See Further Details.
*
* @param[in] LDA
* The leading dimension of the array A. LDA >= max(1,M).
@@ -88,7 +88,7 @@
* @param[in,out] B
* On entry, the M-by-N tile B.
* On exit, B is overwritten by the corresponding block of
- * Q*C or Q**H*C or C*Q or C*Q**H. See Further Details.
+ * Q*C or Q^H*C or C*Q or C*Q^H. See Further Details.
*
* @param[in] LDB
* The leading dimension of the tile B. LDB >= max(1,M).
@@ -122,11 +122,11 @@
*
* If trans='N' and side='L', C is on exit replaced with Q * C.
*
- * If trans='C' and side='L', C is on exit replaced with Q**H * C.
+ * If trans='C' and side='L', C is on exit replaced with Q^H * C.
*
* If trans='N' and side='R', C is on exit replaced with C * Q.
*
- * If trans='C' and side='R', C is on exit replaced with C * Q**H.
+ * If trans='C' and side='R', C is on exit replaced with C * Q^H.
*
*******************************************************************************
*
diff --git a/coreblas/compute/core_ztpmqrt.c b/coreblas/compute/core_ztpmqrt.c
index f308f6389..68dfb9744 100644
--- a/coreblas/compute/core_ztpmqrt.c
+++ b/coreblas/compute/core_ztpmqrt.c
@@ -31,12 +31,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile B. M >= 0.
@@ -76,7 +76,7 @@
* or (LDA,K) if SIDE = ChamRight
* On entry, the K-by-N or M-by-K matrix A.
* On exit, A is overwritten by the corresponding block of
- * Q*C or Q**H*C or C*Q or C*Q**H. See Further Details.
+ * Q*C or Q^H*C or C*Q or C*Q^H. See Further Details.
*
* @param[in] LDA
* The leading dimension of the array A. LDA >= max(1,M).
@@ -86,7 +86,7 @@
* @param[in,out] B
* On entry, the M-by-N tile B.
* On exit, B is overwritten by the corresponding block of
- * Q*C or Q**H*C or C*Q or C*Q**H. See Further Details.
+ * Q*C or Q^H*C or C*Q or C*Q^H. See Further Details.
*
* @param[in] LDB
* The leading dimension of the tile B. LDB >= max(1,M).
@@ -121,11 +121,11 @@
*
* If trans='N' and side='L', C is on exit replaced with Q * C.
*
- * If trans='C' and side='L', C is on exit replaced with Q**H * C.
+ * If trans='C' and side='L', C is on exit replaced with Q^H * C.
*
* If trans='N' and side='R', C is on exit replaced with C * Q.
*
- * If trans='C' and side='R', C is on exit replaced with C * Q**H.
+ * If trans='C' and side='R', C is on exit replaced with C * Q^H.
*
*******************************************************************************
*
diff --git a/coreblas/compute/core_ztsmlq.c b/coreblas/compute/core_ztsmlq.c
index 34a3b2016..d0f55f225 100644
--- a/coreblas/compute/core_ztsmlq.c
+++ b/coreblas/compute/core_ztsmlq.c
@@ -38,7 +38,7 @@
* TRANS = 'N': Q * | A1 | | A1 A2 | * Q
* | A2 |
*
- * TRANS = 'C': Q**H * | A1 | | A1 A2 | * Q**H
+ * TRANS = 'C': Q^H * | A1 | | A1 A2 | * Q^H
* | A2 |
*
* where Q is a complex unitary matrix defined as the product of k
@@ -51,12 +51,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] M1
* The number of rows of the tile A1. M1 >= 0.
@@ -248,7 +248,7 @@ int CORE_ztsmlq(cham_side_t side, cham_trans_t trans,
jc = i;
}
/*
- * Apply H or H' (NOTE: CORE_zparfb used to be CORE_ztsrfb)
+ * Apply H or H'
*/
CORE_zparfb(
side, trans, ChamDirForward, ChamRowwise,
diff --git a/coreblas/compute/core_ztsmlq_hetra1.c b/coreblas/compute/core_ztsmlq_hetra1.c
index bd3d8dc40..c4e947942 100644
--- a/coreblas/compute/core_ztsmlq_hetra1.c
+++ b/coreblas/compute/core_ztsmlq_hetra1.c
@@ -38,12 +38,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] m1
* The number of rows of the tile A1. m1 >= 0.
diff --git a/coreblas/compute/core_ztsmqr.c b/coreblas/compute/core_ztsmqr.c
index 0e48144dc..98b5b58d0 100644
--- a/coreblas/compute/core_ztsmqr.c
+++ b/coreblas/compute/core_ztsmqr.c
@@ -38,7 +38,7 @@
* TRANS = 'N': Q * | A1 | | A1 A2 | * Q
* | A2 |
*
- * TRANS = 'C': Q**H * | A1 | | A1 A2 | * Q**H
+ * TRANS = 'C': Q^H * | A1 | | A1 A2 | * Q^H
* | A2 |
*
* where Q is a complex unitary matrix defined as the product of k
@@ -51,12 +51,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] M1
* The number of rows of the tile A1. M1 >= 0.
@@ -243,7 +243,7 @@ int CORE_ztsmqr(cham_side_t side, cham_trans_t trans,
jc = i;
}
/*
- * Apply H or H' (NOTE: CORE_zparfb used to be CORE_ztsrfb)
+ * Apply H or H'
*/
CORE_zparfb(
side, trans, ChamDirForward, ChamColumnwise,
diff --git a/coreblas/compute/core_ztsmqr_hetra1.c b/coreblas/compute/core_ztsmqr_hetra1.c
index 825fd30c7..2b9f8f661 100644
--- a/coreblas/compute/core_ztsmqr_hetra1.c
+++ b/coreblas/compute/core_ztsmqr_hetra1.c
@@ -40,12 +40,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] m1
* The number of rows of the tile A1. M1 >= 0.
diff --git a/coreblas/compute/core_zttmlq.c b/coreblas/compute/core_zttmlq.c
index d28b8057a..0226f47dd 100644
--- a/coreblas/compute/core_zttmlq.c
+++ b/coreblas/compute/core_zttmlq.c
@@ -36,7 +36,7 @@
* TRANS = 'N': Q * | A1 | | A1 | * Q
* | A2 | | A2 |
*
- * TRANS = 'C': Q**H * | A1 | | A1 | * Q**H
+ * TRANS = 'C': Q^H * | A1 | | A1 | * Q^H
* | A2 | | A2 |
*
* where Q is a complex unitary matrix defined as the product of k
@@ -49,12 +49,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] M1
* The number of rows of the tile A1. M1 >= 0.
@@ -244,7 +244,7 @@ int CORE_zttmlq(cham_side_t side, cham_trans_t trans,
}
/*
- * Apply H or H' (NOTE: CORE_zparfb used to be CORE_zttrfb)
+ * Apply H or H'
*/
CORE_zparfb(
side, trans, ChamDirForward, ChamRowwise,
diff --git a/coreblas/compute/core_zttmqr.c b/coreblas/compute/core_zttmqr.c
index 896b0c898..e04a22ac5 100644
--- a/coreblas/compute/core_zttmqr.c
+++ b/coreblas/compute/core_zttmqr.c
@@ -35,7 +35,7 @@
* TRANS = 'N': Q * | A1 | | A1 | * Q
* | A2 | | A2 |
*
- * TRANS = 'C': Q**H * | A1 | | A1 | * Q**H
+ * TRANS = 'C': Q^H * | A1 | | A1 | * Q^H
* | A2 | | A2 |
*
* where Q is a complex unitary matrix defined as the product of k
@@ -48,12 +48,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] M1
* The number of rows of the tile A1. M1 >= 0.
@@ -235,7 +235,7 @@ int CORE_zttmqr(cham_side_t side, cham_trans_t trans,
}
/*
- * Apply H or H' (NOTE: CORE_zparfb used to be CORE_zttrfb)
+ * Apply H or H'
*/
CORE_zparfb(
side, trans, ChamDirForward, ChamColumnwise,
diff --git a/coreblas/compute/core_zunmlq.c b/coreblas/compute/core_zunmlq.c
index 3b5d17c06..6a310b859 100644
--- a/coreblas/compute/core_zunmlq.c
+++ b/coreblas/compute/core_zunmlq.c
@@ -35,7 +35,7 @@
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -48,12 +48,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : Transpose, apply Q**H.
+ * @arg ChamConjTrans : Transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile C. M >= 0.
@@ -90,7 +90,7 @@
*
* @param[in,out] C
* On entry, the M-by-N tile C.
- * On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
+ * On exit, C is overwritten by Q*C or Q^T*C or C*Q^T or C*Q.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/coreblas/compute/core_zunmqr.c b/coreblas/compute/core_zunmqr.c
index 48e75b281..712da7e6c 100644
--- a/coreblas/compute/core_zunmqr.c
+++ b/coreblas/compute/core_zunmqr.c
@@ -35,7 +35,7 @@
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -48,12 +48,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : Transpose, apply Q**H.
+ * @arg ChamConjTrans : Transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile C. M >= 0.
@@ -91,7 +91,7 @@
*
* @param[in,out] C
* On entry, the M-by-N tile C.
- * On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
+ * On exit, C is overwritten by Q*C or Q^T*C or C*Q^T or C*Q.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/coreblas/eztrace_module/coreblas_eztrace_module b/coreblas/eztrace_module/coreblas_eztrace_module
index 4a8688192..3bca3cbf1 100644
--- a/coreblas/eztrace_module/coreblas_eztrace_module
+++ b/coreblas/eztrace_module/coreblas_eztrace_module
@@ -7,18 +7,6 @@ ID 7770
void CORE_scasum(int storev, int uplo, int M, int N,
void *A, int lda, float *work);
-void CORE_cbrdalg(int uplo, int N, int NB,
- void *pA, void *C, void *S,
- int i, int j, int m, int grsiz);
-int CORE_cgbelr(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-int CORE_cgbrce(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-int CORE_cgblrx(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
int CORE_cgeadd(int M, int N, void *alpha,
void *A, int LDA,
void *B, int LDB);
@@ -64,30 +52,6 @@ int CORE_cgetrf_reclap(int M, int N,
int CORE_cgetrf_rectil(void *A, int *IPIV, int *info);
void CORE_cgetrip(int m, int n, void *A,
void *work);
-int CORE_chbelr(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-int CORE_chblrx(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-int CORE_chbrce(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-void CORE_chbtype1cb(int N, int NB,
- void *A, int LDA,
- void *V, void *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- void *WORK);
-void CORE_chbtype2cb(int N, int NB,
- void *A, int LDA,
- void *V, void *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- void *WORK);
-void CORE_chbtype3cb(int N, int NB,
- void *A, int LDA,
- void *V, void *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- void *WORK);
void CORE_chegst(int itype, int uplo, int N,
void *A, int LDA,
void *B, int LDB, int *INFO);
@@ -230,13 +194,6 @@ int CORE_csyssq(int uplo, int N,
void *A, int LDA,
float *scale, float *sumsq);
int CORE_csytf2_nopiv(int uplo, int n, void *A, int lda);
-void CORE_cswpab(int i, int n1, int n2,
- void *A, void *work);
-int CORE_cswptr_ontile(void *descA, int i1, int i2, const int *ipiv, int inc,
- void *Akk, int ldak);
-void CORE_ctrdalg(int uplo, int N, int NB,
- void *pA, void *V, void *TAU,
- int i, int j, int m, int grsiz);
void CORE_ctrmm(int side, int uplo,
int transA, int diag,
int M, int N,
@@ -352,18 +309,6 @@ int CORE_cunmqr(int side, int trans,
void CORE_dasum(int storev, int uplo, int M, int N,
const double *A, int lda, double *work);
-void CORE_dbrdalg(int uplo, int N, int NB,
- void *pA, double *C, double *S,
- int i, int j, int m, int grsiz);
-int CORE_dgbelr(int uplo, int N,
- void **A, double *V, double *TAU,
- int st, int ed, int eltsize);
-int CORE_dgbrce(int uplo, int N,
- void **A, double *V, double *TAU,
- int st, int ed, int eltsize);
-int CORE_dgblrx(int uplo, int N,
- void **A, double *V, double *TAU,
- int st, int ed, int eltsize);
int CORE_dgeadd(int M, int N, double alpha,
const double *A, int LDA,
double *B, int LDB);
@@ -409,30 +354,6 @@ int CORE_dgetrf_reclap(int M, int N,
int CORE_dgetrf_rectil(void *A, int *IPIV, int *info);
void CORE_dgetrip(int m, int n, double *A,
double *work);
-int CORE_dhbelr(int uplo, int N,
- void **A, double *V, double *TAU,
- int st, int ed, int eltsize);
-int CORE_dhblrx(int uplo, int N,
- void **A, double *V, double *TAU,
- int st, int ed, int eltsize);
-int CORE_dhbrce(int uplo, int N,
- void **A, double *V, double *TAU,
- int st, int ed, int eltsize);
-void CORE_dhbtype1cb(int N, int NB,
- double *A, int LDA,
- double *V, double *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- double *WORK);
-void CORE_dhbtype2cb(int N, int NB,
- double *A, int LDA,
- double *V, double *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- double *WORK);
-void CORE_dhbtype3cb(int N, int NB,
- double *A, int LDA,
- const double *V, const double *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- double *WORK);
void CORE_dsygst(int itype, int uplo, int N,
double *A, int LDA,
double *B, int LDB, int *INFO);
@@ -553,13 +474,6 @@ int CORE_dssssm(int M1, int N1, int M2, int N2, int K, int IB,
const double *L2, int LDL2,
const int *IPIV);
int CORE_dsytf2_nopiv(int uplo, int n, double *A, int lda);
-void CORE_dswpab(int i, int n1, int n2,
- double *A, double *work);
-int CORE_dswptr_ontile(void *descA, int i1, int i2, const int *ipiv, int inc,
- const double *Akk, int ldak);
-void CORE_dtrdalg(int uplo, int N, int NB,
- void *pA, double *V, double *TAU,
- int i, int j, int m, int grsiz);
void CORE_dtrmm(int side, int uplo,
int transA, int diag,
int M, int N,
@@ -682,18 +596,6 @@ void CORE_dlag2s(int m, int n,
void CORE_sasum(int storev, int uplo, int M, int N,
const float *A, int lda, float *work);
-void CORE_sbrdalg(int uplo, int N, int NB,
- void *pA, float *C, float *S,
- int i, int j, int m, int grsiz);
-int CORE_sgbelr(int uplo, int N,
- void **A, float *V, float *TAU,
- int st, int ed, int eltsize);
-int CORE_sgbrce(int uplo, int N,
- void **A, float *V, float *TAU,
- int st, int ed, int eltsize);
-int CORE_sgblrx(int uplo, int N,
- void **A, float *V, float *TAU,
- int st, int ed, int eltsize);
int CORE_sgeadd(int M, int N, float alpha,
const float *A, int LDA,
float *B, int LDB);
@@ -739,30 +641,6 @@ int CORE_sgetrf_reclap(int M, int N,
int CORE_sgetrf_rectil(void *A, int *IPIV, int *info);
void CORE_sgetrip(int m, int n, float *A,
float *work);
-int CORE_shbelr(int uplo, int N,
- void **A, float *V, float *TAU,
- int st, int ed, int eltsize);
-int CORE_shblrx(int uplo, int N,
- void **A, float *V, float *TAU,
- int st, int ed, int eltsize);
-int CORE_shbrce(int uplo, int N,
- void **A, float *V, float *TAU,
- int st, int ed, int eltsize);
-void CORE_shbtype1cb(int N, int NB,
- float *A, int LDA,
- float *V, float *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- float *WORK);
-void CORE_shbtype2cb(int N, int NB,
- float *A, int LDA,
- float *V, float *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- float *WORK);
-void CORE_shbtype3cb(int N, int NB,
- float *A, int LDA,
- const float *V, const float *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- float *WORK);
void CORE_ssygst(int itype, int uplo, int N,
float *A, int LDA,
float *B, int LDB, int *INFO);
@@ -883,13 +761,6 @@ int CORE_sssssm(int M1, int N1, int M2, int N2, int K, int IB,
const float *L2, int LDL2,
const int *IPIV);
int CORE_ssytf2_nopiv(int uplo, int n, float *A, int lda);
-void CORE_sswpab(int i, int n1, int n2,
- float *A, float *work);
-int CORE_sswptr_ontile(void *descA, int i1, int i2, const int *ipiv, int inc,
- const float *Akk, int ldak);
-void CORE_strdalg(int uplo, int N, int NB,
- void *pA, float *V, float *TAU,
- int i, int j, int m, int grsiz);
void CORE_strmm(int side, int uplo,
int transA, int diag,
int M, int N,
@@ -1005,18 +876,6 @@ int CORE_sormqr(int side, int trans,
void CORE_dzasum(int storev, int uplo, int M, int N,
void *A, int lda, double *work);
-void CORE_zbrdalg(int uplo, int N, int NB,
- void *pA, void *C, void *S,
- int i, int j, int m, int grsiz);
-int CORE_zgbelr(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-int CORE_zgbrce(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-int CORE_zgblrx(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
int CORE_zgeadd(int M, int N, void *alpha,
void *A, int LDA,
void *B, int LDB);
@@ -1062,30 +921,6 @@ int CORE_zgetrf_reclap(int M, int N,
int CORE_zgetrf_rectil(void *A, int *IPIV, int *info);
void CORE_zgetrip(int m, int n, void *A,
void *work);
-int CORE_zhbelr(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-int CORE_zhblrx(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-int CORE_zhbrce(int uplo, int N,
- void **A, void *V, void *TAU,
- int st, int ed, int eltsize);
-void CORE_zhbtype1cb(int N, int NB,
- void *A, int LDA,
- void *V, void *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- void *WORK);
-void CORE_zhbtype2cb(int N, int NB,
- void *A, int LDA,
- void *V, void *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- void *WORK);
-void CORE_zhbtype3cb(int N, int NB,
- void *A, int LDA,
- void *V, void *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- void *WORK);
void CORE_zhegst(int itype, int uplo, int N,
void *A, int LDA,
void *B, int LDB, int *INFO);
@@ -1222,13 +1057,6 @@ int CORE_zsyssq(int uplo, int N,
void *A, int LDA,
double *scale, double *sumsq);
int CORE_zsytf2_nopiv(int uplo, int n, void *A, int lda);
-void CORE_zswpab(int i, int n1, int n2,
- void *A, void *work);
-int CORE_zswptr_ontile(void *descA, int i1, int i2, const int *ipiv, int inc,
- void *Akk, int ldak);
-void CORE_ztrdalg(int uplo, int N, int NB,
- void *pA, void *V, void *TAU,
- int i, int j, int m, int grsiz);
void CORE_ztrmm(int side, int uplo,
int transA, int diag,
int M, int N,
diff --git a/coreblas/include/coreblas/coreblas_z.h b/coreblas/include/coreblas/coreblas_z.h
index 8a755db76..ee434de43 100644
--- a/coreblas/include/coreblas/coreblas_z.h
+++ b/coreblas/include/coreblas/coreblas_z.h
@@ -32,18 +32,6 @@
*/
void CORE_dzasum(cham_store_t storev, cham_uplo_t uplo, int M, int N,
const CHAMELEON_Complex64_t *A, int lda, double *work);
-void CORE_zbrdalg(cham_uplo_t uplo, int N, int NB,
- const CHAM_desc_t *pA, CHAMELEON_Complex64_t *C, CHAMELEON_Complex64_t *S,
- int i, int j, int m, int grsiz);
-int CORE_zgbelr(cham_uplo_t uplo, int N,
- CHAM_desc_t *A, CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int st, int ed, int eltsize);
-int CORE_zgbrce(cham_uplo_t uplo, int N,
- CHAM_desc_t *A, CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int st, int ed, int eltsize);
-int CORE_zgblrx(cham_uplo_t uplo, int N,
- CHAM_desc_t *A, CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int st, int ed, int eltsize);
int CORE_zaxpy(int M, CHAMELEON_Complex64_t alpha,
const CHAMELEON_Complex64_t *A, int incA,
CHAMELEON_Complex64_t *B, int incB);
@@ -99,30 +87,6 @@ void CORE_zgetrip(int m, int n, CHAMELEON_Complex64_t *A,
void CORE_zhe2ge(cham_uplo_t uplo, int M, int N,
const CHAMELEON_Complex64_t *A, int LDA,
CHAMELEON_Complex64_t *B, int LDB);
-int CORE_zhbelr(cham_uplo_t uplo, int N,
- CHAM_desc_t *A, CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int st, int ed, int eltsize);
-int CORE_zhblrx(cham_uplo_t uplo, int N,
- CHAM_desc_t *A, CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int st, int ed, int eltsize);
-int CORE_zhbrce(cham_uplo_t uplo, int N,
- CHAM_desc_t *A, CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int st, int ed, int eltsize);
-void CORE_zhbtype1cb(int N, int NB,
- CHAMELEON_Complex64_t *A, int LDA,
- CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- CHAMELEON_Complex64_t *WORK);
-void CORE_zhbtype2cb(int N, int NB,
- CHAMELEON_Complex64_t *A, int LDA,
- CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- CHAMELEON_Complex64_t *WORK);
-void CORE_zhbtype3cb(int N, int NB,
- CHAMELEON_Complex64_t *A, int LDA,
- const CHAMELEON_Complex64_t *V, const CHAMELEON_Complex64_t *TAU,
- int st, int ed, int sweep, int Vblksiz, int WANTZ,
- CHAMELEON_Complex64_t *WORK);
void CORE_zhegst(int itype, cham_uplo_t uplo, int N,
CHAMELEON_Complex64_t *A, int LDA,
CHAMELEON_Complex64_t *B, int LDB, int *INFO);
@@ -261,10 +225,6 @@ int CORE_zsyssq(cham_uplo_t uplo, int N,
const CHAMELEON_Complex64_t *A, int LDA,
double *scale, double *sumsq);
int CORE_zsytf2_nopiv(cham_uplo_t uplo, int n, CHAMELEON_Complex64_t *A, int lda);
-void CORE_zswpab(int i, int n1, int n2,
- CHAMELEON_Complex64_t *A, CHAMELEON_Complex64_t *work);
-int CORE_zswptr_ontile(CHAM_desc_t descA, int i1, int i2, const int *ipiv, int inc,
- const CHAMELEON_Complex64_t *Akk, int ldak);
int CORE_ztradd(cham_uplo_t uplo, cham_trans_t trans, int M, int N,
CHAMELEON_Complex64_t alpha,
const CHAMELEON_Complex64_t *A, int LDA,
@@ -273,9 +233,6 @@ int CORE_ztradd(cham_uplo_t uplo, cham_trans_t trans, int M, int N,
void CORE_ztrasm(cham_store_t storev, cham_uplo_t uplo, cham_diag_t diag,
int M, int N,
const CHAMELEON_Complex64_t *A, int lda, double *work);
-void CORE_ztrdalg(cham_uplo_t uplo, int N, int NB,
- const CHAM_desc_t *pA, CHAMELEON_Complex64_t *V, CHAMELEON_Complex64_t *TAU,
- int i, int j, int m, int grsiz);
void CORE_ztrmm(cham_side_t side, cham_uplo_t uplo,
cham_trans_t transA, cham_diag_t diag,
int M, int N,
diff --git a/cudablas/compute/cuda_zparfb.c b/cudablas/compute/cuda_zparfb.c
index 95bbe2115..b4e6f6c19 100644
--- a/cudablas/compute/cuda_zparfb.c
+++ b/cudablas/compute/cuda_zparfb.c
@@ -46,12 +46,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] direct
* Indicates how H is formed from a product of elementary
diff --git a/cudablas/compute/cuda_ztpmlqt.c b/cudablas/compute/cuda_ztpmlqt.c
index a8ff6ce5a..aaf70c231 100644
--- a/cudablas/compute/cuda_ztpmlqt.c
+++ b/cudablas/compute/cuda_ztpmlqt.c
@@ -33,12 +33,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile B. M >= 0.
@@ -78,7 +78,7 @@
* or (LDA,K) if SIDE = ChamRight
* On entry, the K-by-N or M-by-K matrix A.
* On exit, A is overwritten by the corresponding block of
- * Q*C or Q**H*C or C*Q or C*Q**H. See Further Details.
+ * Q*C or Q^H*C or C*Q or C*Q^H. See Further Details.
*
* @param[in] LDA
* The leading dimension of the array A. LDA >= max(1,M).
@@ -88,7 +88,7 @@
* @param[in,out] B
* On entry, the M-by-N tile B.
* On exit, B is overwritten by the corresponding block of
- * Q*C or Q**H*C or C*Q or C*Q**H. See Further Details.
+ * Q*C or Q^H*C or C*Q or C*Q^H. See Further Details.
*
* @param[in] LDB
* The leading dimension of the tile B. LDB >= max(1,M).
@@ -122,11 +122,11 @@
*
* If trans='N' and side='L', C is on exit replaced with Q * C.
*
- * If trans='C' and side='L', C is on exit replaced with Q**H * C.
+ * If trans='C' and side='L', C is on exit replaced with Q^H * C.
*
* If trans='N' and side='R', C is on exit replaced with C * Q.
*
- * If trans='C' and side='R', C is on exit replaced with C * Q**H.
+ * If trans='C' and side='R', C is on exit replaced with C * Q^H.
*
*******************************************************************************
*
diff --git a/cudablas/compute/cuda_ztpmqrt.c b/cudablas/compute/cuda_ztpmqrt.c
index 22319e42d..c9a1fea2a 100644
--- a/cudablas/compute/cuda_ztpmqrt.c
+++ b/cudablas/compute/cuda_ztpmqrt.c
@@ -33,12 +33,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : ConjTranspose, apply Q**H.
+ * @arg ChamConjTrans : ConjTranspose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile B. M >= 0.
@@ -78,7 +78,7 @@
* or (LDA,K) if SIDE = ChamRight
* On entry, the K-by-N or M-by-K matrix A.
* On exit, A is overwritten by the corresponding block of
- * Q*C or Q**H*C or C*Q or C*Q**H. See Further Details.
+ * Q*C or Q^H*C or C*Q or C*Q^H. See Further Details.
*
* @param[in] LDA
* The leading dimension of the array A. LDA >= max(1,M).
@@ -88,7 +88,7 @@
* @param[in,out] B
* On entry, the M-by-N tile B.
* On exit, B is overwritten by the corresponding block of
- * Q*C or Q**H*C or C*Q or C*Q**H. See Further Details.
+ * Q*C or Q^H*C or C*Q or C*Q^H. See Further Details.
*
* @param[in] LDB
* The leading dimension of the tile B. LDB >= max(1,M).
@@ -123,11 +123,11 @@
*
* If trans='N' and side='L', C is on exit replaced with Q * C.
*
- * If trans='C' and side='L', C is on exit replaced with Q**H * C.
+ * If trans='C' and side='L', C is on exit replaced with Q^H * C.
*
* If trans='N' and side='R', C is on exit replaced with C * Q.
*
- * If trans='C' and side='R', C is on exit replaced with C * Q**H.
+ * If trans='C' and side='R', C is on exit replaced with C * Q^H.
*
*******************************************************************************
*
diff --git a/runtime/CMakeLists.txt b/runtime/CMakeLists.txt
index f756adc45..e184ca536 100644
--- a/runtime/CMakeLists.txt
+++ b/runtime/CMakeLists.txt
@@ -28,7 +28,7 @@
# List of codelets required by all runtimes
# -----------------------------------------
set(CODELETS_ZSRC
- codelets/codelet_zasum.c
+ codelets/codelet_dzasum.c
##################
# BLAS 1
##################
diff --git a/runtime/openmp/codelets/codelet_zasum.c b/runtime/openmp/codelets/codelet_dzasum.c
similarity index 94%
rename from runtime/openmp/codelets/codelet_zasum.c
rename to runtime/openmp/codelets/codelet_dzasum.c
index 0ec02b39c..1ce65879b 100644
--- a/runtime/openmp/codelets/codelet_zasum.c
+++ b/runtime/openmp/codelets/codelet_dzasum.c
@@ -1,6 +1,6 @@
/**
*
- * @file openmp/codelet_zasum.c
+ * @file openmp/codelet_dzasum.c
*
* @copyright 2009-2014 The University of Tennessee and The University of
* Tennessee Research Foundation. All rights reserved.
@@ -9,7 +9,7 @@
*
***
*
- * @brief Chameleon zasum OpenMP codelet
+ * @brief Chameleon dzasum OpenMP codelet
*
* @version 0.9.2
* @comment This file has been automatically generated
diff --git a/runtime/openmp/codelets/codelet_zunmlq.c b/runtime/openmp/codelets/codelet_zunmlq.c
index 6de8202b2..92d6e71f8 100644
--- a/runtime/openmp/codelets/codelet_zunmlq.c
+++ b/runtime/openmp/codelets/codelet_zunmlq.c
@@ -35,7 +35,7 @@
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -48,12 +48,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : Transpose, apply Q**H.
+ * @arg ChamConjTrans : Transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile C. M >= 0.
@@ -90,7 +90,7 @@
*
* @param[in,out] C
* On entry, the M-by-N tile C.
- * On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
+ * On exit, C is overwritten by Q*C or Q^T*C or C*Q^T or C*Q.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/runtime/openmp/codelets/codelet_zunmqr.c b/runtime/openmp/codelets/codelet_zunmqr.c
index 93190251e..66aa62b5d 100644
--- a/runtime/openmp/codelets/codelet_zunmqr.c
+++ b/runtime/openmp/codelets/codelet_zunmqr.c
@@ -34,7 +34,7 @@
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -47,12 +47,12 @@
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : Transpose, apply Q**H.
+ * @arg ChamConjTrans : Transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile C. M >= 0.
@@ -90,7 +90,7 @@
*
* @param[in,out] C
* On entry, the M-by-N tile C.
- * On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
+ * On exit, C is overwritten by Q*C or Q^T*C or C*Q^T or C*Q.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/runtime/parsec/codelets/codelet_zasum.c b/runtime/parsec/codelets/codelet_dzasum.c
similarity index 95%
rename from runtime/parsec/codelets/codelet_zasum.c
rename to runtime/parsec/codelets/codelet_dzasum.c
index ccea81743..e0faa8dd2 100644
--- a/runtime/parsec/codelets/codelet_zasum.c
+++ b/runtime/parsec/codelets/codelet_dzasum.c
@@ -1,6 +1,6 @@
/**
*
- * @file parsec/codelet_zasum.c
+ * @file parsec/codelet_dzasum.c
*
* @copyright 2009-2015 The University of Tennessee and The University of
* Tennessee Research Foundation. All rights reserved.
@@ -9,7 +9,7 @@
*
***
*
- * @brief Chameleon zasum PaRSEC codelet
+ * @brief Chameleon dzasum PaRSEC codelet
*
* @version 0.9.2
* @author Reazul Hoque
@@ -50,7 +50,7 @@ void INSERT_TASK_dzasum(const RUNTIME_option_t *options,
parsec_taskpool_t* PARSEC_dtd_taskpool = (parsec_taskpool_t *)(options->sequence->schedopt);
parsec_dtd_taskpool_insert_task(
- PARSEC_dtd_taskpool, CORE_dzasum_parsec, options->priority, "zasum",
+ PARSEC_dtd_taskpool, CORE_dzasum_parsec, options->priority, "dzasum",
sizeof(int), &storev, VALUE,
sizeof(int), &uplo, VALUE,
sizeof(int), &M, VALUE,
diff --git a/runtime/quark/codelets/codelet_zasum.c b/runtime/quark/codelets/codelet_dzasum.c
similarity index 96%
rename from runtime/quark/codelets/codelet_zasum.c
rename to runtime/quark/codelets/codelet_dzasum.c
index 6dd5cce03..8c49a6f76 100644
--- a/runtime/quark/codelets/codelet_zasum.c
+++ b/runtime/quark/codelets/codelet_dzasum.c
@@ -1,6 +1,6 @@
/**
*
- * @file quark/codelet_zasum.c
+ * @file quark/codelet_dzasum.c
*
* @copyright 2009-2014 The University of Tennessee and The University of
* Tennessee Research Foundation. All rights reserved.
@@ -9,7 +9,7 @@
*
***
*
- * @brief Chameleon zasum Quark codelet
+ * @brief Chameleon dzasum Quark codelet
*
* @version 0.9.2
* @comment This file has been automatically generated
diff --git a/runtime/quark/codelets/codelet_zunmlq.c b/runtime/quark/codelets/codelet_zunmlq.c
index 76141454b..5b8687571 100644
--- a/runtime/quark/codelets/codelet_zunmlq.c
+++ b/runtime/quark/codelets/codelet_zunmlq.c
@@ -59,7 +59,7 @@ void CORE_zunmlq_quark(Quark *quark)
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -72,12 +72,12 @@ void CORE_zunmlq_quark(Quark *quark)
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : Transpose, apply Q**H.
+ * @arg ChamConjTrans : Transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile C. M >= 0.
@@ -114,7 +114,7 @@ void CORE_zunmlq_quark(Quark *quark)
*
* @param[in,out] C
* On entry, the M-by-N tile C.
- * On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
+ * On exit, C is overwritten by Q*C or Q^T*C or C*Q^T or C*Q.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/runtime/quark/codelets/codelet_zunmqr.c b/runtime/quark/codelets/codelet_zunmqr.c
index 75968addd..f03746016 100644
--- a/runtime/quark/codelets/codelet_zunmqr.c
+++ b/runtime/quark/codelets/codelet_zunmqr.c
@@ -58,7 +58,7 @@ void CORE_zunmqr_quark(Quark *quark)
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -71,12 +71,12 @@ void CORE_zunmqr_quark(Quark *quark)
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : Transpose, apply Q**H.
+ * @arg ChamConjTrans : Transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile C. M >= 0.
@@ -114,7 +114,7 @@ void CORE_zunmqr_quark(Quark *quark)
*
* @param[in,out] C
* On entry, the M-by-N tile C.
- * On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
+ * On exit, C is overwritten by Q*C or Q^T*C or C*Q^T or C*Q.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/runtime/quark/include/core_blas_dag.h b/runtime/quark/include/core_blas_dag.h
index f5ba7073d..83bdbd531 100644
--- a/runtime/quark/include/core_blas_dag.h
+++ b/runtime/quark/include/core_blas_dag.h
@@ -78,6 +78,8 @@
#define DAG_CORE_TSTRF DAG_SET_PROPERTIES( "TSTRF" , "red" )
#define DAG_CORE_UNMLQ DAG_SET_PROPERTIES( "UNMLQ" , "cyan" )
#define DAG_CORE_UNMQR DAG_SET_PROPERTIES( "UNMQR" , "cyan" )
+#define DAG_CORE_ORMLQ DAG_SET_PROPERTIES( "ORMLQ" , "cyan" )
+#define DAG_CORE_ORMQR DAG_SET_PROPERTIES( "ORMQR" , "cyan" )
#define DAG_CORE_TSLQT DAG_CORE_TPLQT
#define DAG_CORE_TSMLQ DAG_CORE_TPMLQT
diff --git a/runtime/starpu/codelets/codelet_zasum.c b/runtime/starpu/codelets/codelet_dzasum.c
similarity index 88%
rename from runtime/starpu/codelets/codelet_zasum.c
rename to runtime/starpu/codelets/codelet_dzasum.c
index 1cbd3e6b2..0e94fc672 100644
--- a/runtime/starpu/codelets/codelet_zasum.c
+++ b/runtime/starpu/codelets/codelet_dzasum.c
@@ -1,6 +1,6 @@
/**
*
- * @file starpu/codelet_zasum.c
+ * @file starpu/codelet_dzasum.c
*
* @copyright 2009-2014 The University of Tennessee and The University of
* Tennessee Research Foundation. All rights reserved.
@@ -9,7 +9,7 @@
*
***
*
- * @brief Chameleon zasum StarPU codelet
+ * @brief Chameleon dzasum StarPU codelet
*
* @version 0.9.2
* @comment This file has been automatically generated
@@ -43,15 +43,15 @@ static void cl_dzasum_cpu_func(void *descr[], void *cl_arg)
/*
* Codelet definition
*/
-CODELETS_CPU(zasum, 2, cl_dzasum_cpu_func)
+CODELETS_CPU(dzasum, 2, cl_dzasum_cpu_func)
void INSERT_TASK_dzasum( const RUNTIME_option_t *options,
cham_store_t storev, cham_uplo_t uplo, int M, int N,
const CHAM_desc_t *A, int Am, int An, int lda,
const CHAM_desc_t *B, int Bm, int Bn )
{
- struct starpu_codelet *codelet = &cl_zasum;
- void (*callback)(void*) = options->profiling ? cl_zasum_callback : NULL;
+ struct starpu_codelet *codelet = &cl_dzasum;
+ void (*callback)(void*) = options->profiling ? cl_dzasum_callback : NULL;
CHAMELEON_BEGIN_ACCESS_DECLARATION;
CHAMELEON_ACCESS_R(A, Am, An);
@@ -70,7 +70,7 @@ void INSERT_TASK_dzasum( const RUNTIME_option_t *options,
STARPU_PRIORITY, options->priority,
STARPU_CALLBACK, callback,
#if defined(CHAMELEON_CODELETS_HAVE_NAME)
- STARPU_NAME, "zasum",
+ STARPU_NAME, "dzasum",
#endif
0);
}
diff --git a/runtime/starpu/codelets/codelet_zcallback.c b/runtime/starpu/codelets/codelet_zcallback.c
index 0c31aa653..35aea3122 100644
--- a/runtime/starpu/codelets/codelet_zcallback.c
+++ b/runtime/starpu/codelets/codelet_zcallback.c
@@ -22,7 +22,7 @@
#include "chameleon_starpu.h"
#include "runtime_codelet_z.h"
-CHAMELEON_CL_CB(zasum, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, M*N)
+CHAMELEON_CL_CB(dzasum, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, M*N)
CHAMELEON_CL_CB(zaxpy, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_nx(task->handles[1]), 0, M)
CHAMELEON_CL_CB(zgeadd, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, M*N)
CHAMELEON_CL_CB(zlascal, starpu_matrix_get_nx(task->handles[0]), starpu_matrix_get_ny(task->handles[0]), 0, M*N)
diff --git a/runtime/starpu/codelets/codelet_zunmlq.c b/runtime/starpu/codelets/codelet_zunmlq.c
index bab65bfcc..be36f957d 100644
--- a/runtime/starpu/codelets/codelet_zunmlq.c
+++ b/runtime/starpu/codelets/codelet_zunmlq.c
@@ -104,7 +104,7 @@ CODELETS(zunmlq, 4, cl_zunmlq_cpu_func, cl_zunmlq_cuda_func, STARPU_CUDA_ASYNC)
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -117,12 +117,12 @@ CODELETS(zunmlq, 4, cl_zunmlq_cpu_func, cl_zunmlq_cuda_func, STARPU_CUDA_ASYNC)
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : Transpose, apply Q**H.
+ * @arg ChamConjTrans : Transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile C. M >= 0.
@@ -159,7 +159,7 @@ CODELETS(zunmlq, 4, cl_zunmlq_cpu_func, cl_zunmlq_cuda_func, STARPU_CUDA_ASYNC)
*
* @param[in,out] C
* On entry, the M-by-N tile C.
- * On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
+ * On exit, C is overwritten by Q*C or Q^T*C or C*Q^T or C*Q.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/runtime/starpu/codelets/codelet_zunmqr.c b/runtime/starpu/codelets/codelet_zunmqr.c
index ed8cbccd5..8ff98bc79 100644
--- a/runtime/starpu/codelets/codelet_zunmqr.c
+++ b/runtime/starpu/codelets/codelet_zunmqr.c
@@ -103,7 +103,7 @@ CODELETS(zunmqr, 4, cl_zunmqr_cpu_func, cl_zunmqr_cuda_func, STARPU_CUDA_ASYNC)
*
* SIDE = 'L' SIDE = 'R'
* TRANS = 'N': Q * C C * Q
- * TRANS = 'C': Q**H * C C * Q**H
+ * TRANS = 'C': Q^H * C C * Q^H
*
* where Q is a complex unitary matrix defined as the product of k
* elementary reflectors
@@ -116,12 +116,12 @@ CODELETS(zunmqr, 4, cl_zunmqr_cpu_func, cl_zunmqr_cuda_func, STARPU_CUDA_ASYNC)
*******************************************************************************
*
* @param[in] side
- * @arg ChamLeft : apply Q or Q**H from the Left;
- * @arg ChamRight : apply Q or Q**H from the Right.
+ * @arg ChamLeft : apply Q or Q^H from the Left;
+ * @arg ChamRight : apply Q or Q^H from the Right.
*
* @param[in] trans
* @arg ChamNoTrans : No transpose, apply Q;
- * @arg ChamConjTrans : Transpose, apply Q**H.
+ * @arg ChamConjTrans : Transpose, apply Q^H.
*
* @param[in] M
* The number of rows of the tile C. M >= 0.
@@ -159,7 +159,7 @@ CODELETS(zunmqr, 4, cl_zunmqr_cpu_func, cl_zunmqr_cuda_func, STARPU_CUDA_ASYNC)
*
* @param[in,out] C
* On entry, the M-by-N tile C.
- * On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
+ * On exit, C is overwritten by Q*C or Q^T*C or C*Q^T or C*Q.
*
* @param[in] LDC
* The leading dimension of the array C. LDC >= max(1,M).
diff --git a/runtime/starpu/include/runtime_codelet_z.h b/runtime/starpu/include/runtime_codelet_z.h
index d9c6e915e..4936d329c 100644
--- a/runtime/starpu/include/runtime_codelet_z.h
+++ b/runtime/starpu/include/runtime_codelet_z.h
@@ -36,89 +36,89 @@
/*
* BLAS 1 functions
*/
-ZCODELETS_HEADER(axpy)
+CODELETS_HEADER(zaxpy)
/*
* BLAS 3 functions
*/
-ZCODELETS_HEADER(gemm)
-ZCODELETS_HEADER(hemm)
-ZCODELETS_HEADER(her2k)
-ZCODELETS_HEADER(herk)
-ZCODELETS_HEADER(symm)
-ZCODELETS_HEADER(syr2k)
-ZCODELETS_HEADER(syrk)
-ZCODELETS_HEADER(trmm)
-ZCODELETS_HEADER(trsm)
+CODELETS_HEADER(zgemm)
+CODELETS_HEADER(zhemm)
+CODELETS_HEADER(zher2k)
+CODELETS_HEADER(zherk)
+CODELETS_HEADER(zsymm)
+CODELETS_HEADER(zsyr2k)
+CODELETS_HEADER(zsyrk)
+CODELETS_HEADER(ztrmm)
+CODELETS_HEADER(ztrsm)
/*
* LAPACK functions
*/
-ZCODELETS_HEADER(gelqt)
-ZCODELETS_HEADER(geqrt)
-ZCODELETS_HEADER(gessm)
-ZCODELETS_HEADER(gessq)
-ZCODELETS_HEADER(getrf)
-ZCODELETS_HEADER(getrf_incpiv)
-ZCODELETS_HEADER(getrf_nopiv)
-ZCODELETS_HEADER(herfb)
-ZCODELETS_HEADER(lauum)
-ZCODELETS_HEADER(potrf)
-ZCODELETS_HEADER(ssssm)
-ZCODELETS_HEADER(syssq)
-ZCODELETS_HEADER(trasm)
-ZCODELETS_HEADER(trssq)
-ZCODELETS_HEADER(trtri)
-ZCODELETS_HEADER(tplqt)
-ZCODELETS_HEADER(tpqrt)
-ZCODELETS_HEADER(tpmlqt)
-ZCODELETS_HEADER(tpmqrt)
-ZCODELETS_HEADER(tsmlq_hetra1)
-ZCODELETS_HEADER(tsmqr_hetra1)
-ZCODELETS_HEADER(tstrf)
-ZCODELETS_HEADER(unmlq)
-ZCODELETS_HEADER(unmqr)
+CODELETS_HEADER(zgelqt)
+CODELETS_HEADER(zgeqrt)
+CODELETS_HEADER(zgessm)
+CODELETS_HEADER(zgessq)
+CODELETS_HEADER(zgetrf)
+CODELETS_HEADER(zgetrf_incpiv)
+CODELETS_HEADER(zgetrf_nopiv)
+CODELETS_HEADER(zherfb)
+CODELETS_HEADER(zlauum)
+CODELETS_HEADER(zpotrf)
+CODELETS_HEADER(zssssm)
+CODELETS_HEADER(zsyssq)
+CODELETS_HEADER(ztrasm)
+CODELETS_HEADER(ztrssq)
+CODELETS_HEADER(ztrtri)
+CODELETS_HEADER(ztplqt)
+CODELETS_HEADER(ztpqrt)
+CODELETS_HEADER(ztpmlqt)
+CODELETS_HEADER(ztpmqrt)
+CODELETS_HEADER(ztsmlq_hetra1)
+CODELETS_HEADER(ztsmqr_hetra1)
+CODELETS_HEADER(ztstrf)
+CODELETS_HEADER(zunmlq)
+CODELETS_HEADER(zunmqr)
/*
* Auxiliary functions
*/
-ZCODELETS_HEADER(geadd)
-ZCODELETS_HEADER(he2ge)
-ZCODELETS_HEADER(lascal)
-ZCODELETS_HEADER(tradd)
-ZCODELETS_HEADER(lacpy)
-ZCODELETS_HEADER(lange)
-ZCODELETS_HEADER(lange_max)
-ZCODELETS_HEADER(lansy)
-ZCODELETS_HEADER(lantr)
-ZCODELETS_HEADER(laset)
-ZCODELETS_HEADER(laset2)
-ZCODELETS_HEADER(latro)
-ZCODELETS_HEADER(plssq)
-ZCODELETS_HEADER(plssq2)
+CODELETS_HEADER(zgeadd)
+CODELETS_HEADER(zhe2ge)
+CODELETS_HEADER(zlascal)
+CODELETS_HEADER(ztradd)
+CODELETS_HEADER(zlacpy)
+CODELETS_HEADER(zlange)
+CODELETS_HEADER(zlange_max)
+CODELETS_HEADER(zlansy)
+CODELETS_HEADER(zlantr)
+CODELETS_HEADER(zlaset)
+CODELETS_HEADER(zlaset2)
+CODELETS_HEADER(zlatro)
+CODELETS_HEADER(zplssq)
+CODELETS_HEADER(zplssq2)
/*
* MIXED PRECISION functions
*/
-ZCODELETS_HEADER(lag2c)
+CODELETS_HEADER(zlag2c)
/*
* DZ functions
*/
-ZCODELETS_HEADER(asum)
+CODELETS_HEADER(dzasum)
/*
* CPU only functions
*/
-ZCODELETS_HEADER(plrnt)
-ZCODELETS_HEADER(build)
+CODELETS_HEADER(zplrnt)
+CODELETS_HEADER(zbuild)
#if defined(PRECISION_z) || defined(PRECISION_c)
-ZCODELETS_HEADER(hessq)
-ZCODELETS_HEADER(lanhe)
-ZCODELETS_HEADER(plghe)
-ZCODELETS_HEADER(sytrf_nopiv)
+CODELETS_HEADER(zhessq)
+CODELETS_HEADER(zlanhe)
+CODELETS_HEADER(zplghe)
+CODELETS_HEADER(zsytrf_nopiv)
#endif
-ZCODELETS_HEADER(plgsy)
+CODELETS_HEADER(zplgsy)
#endif /* _runtime_codelet_z_h_ */
diff --git a/runtime/starpu/include/runtime_codelets.h b/runtime/starpu/include/runtime_codelets.h
index a68724449..4fb8b9cf8 100644
--- a/runtime/starpu/include/runtime_codelets.h
+++ b/runtime/starpu/include/runtime_codelets.h
@@ -112,11 +112,6 @@
#define CODELETS_HEADER(name) CODELETS_ALL_HEADER(name)
#endif
-#define SCODELETS_HEADER(name) CODELETS_HEADER(s##name)
-#define DCODELETS_HEADER(name) CODELETS_HEADER(d##name)
-#define CCODELETS_HEADER(name) CODELETS_HEADER(c##name)
-#define ZCODELETS_HEADER(name) CODELETS_HEADER(z##name)
-
CODELETS_HEADER(map);
#endif /* _runtime_codelets_h_ */
diff --git a/testing/lin/clagsy.f b/testing/lin/clagsy.f
index c5fea1b7d..0522d0600 100644
--- a/testing/lin/clagsy.f
+++ b/testing/lin/clagsy.f
@@ -55,7 +55,7 @@
*
* CLAGSY generates a complex symmetric matrix A, by pre- and post-
* multiplying a real diagonal matrix D with a random unitary matrix:
-* A = U*D*U**T. The semi-bandwidth may then be reduced to k by
+* A = U*D*U^T. The semi-bandwidth may then be reduced to k by
* additional unitary transformations.
*
* Arguments
diff --git a/testing/lin/clarhs.f b/testing/lin/clarhs.f
index 46f4d68c2..22165f3a3 100644
--- a/testing/lin/clarhs.f
+++ b/testing/lin/clarhs.f
@@ -58,7 +58,7 @@
* CLARHS chooses a set of NRHS random solution vectors and sets
* up the right hand sides for the linear system
* op( A ) * X = B,
-* where op( A ) may be A, A**T (transpose of A), or A**H (conjugate
+* where op( A ) may be A, A^T (transpose of A), or A^H (conjugate
* transpose of A).
*
* Arguments
@@ -102,8 +102,8 @@
* Used only if A is nonsymmetric; specifies the operation
* applied to the matrix A.
* = 'N': B := A * X
-* = 'T': B := A**T * X
-* = 'C': B := A**H * X
+* = 'T': B := A^T * X
+* = 'C': B := A^H * X
*
* M (input) INTEGER
* The number of rows of the matrix A. M >= 0.
diff --git a/testing/lin/clatrs.f b/testing/lin/clatrs.f
index 87da087b8..38cc5a2a2 100644
--- a/testing/lin/clatrs.f
+++ b/testing/lin/clatrs.f
@@ -57,10 +57,10 @@
*
* CLATRS solves one of the triangular systems
*
-* A * x = s*b, A**T * x = s*b, or A**H * x = s*b,
+* A * x = s*b, A^T * x = s*b, or A^H * x = s*b,
*
* with scaling to prevent overflow. Here A is an upper or lower
-* triangular matrix, A**T denotes the transpose of A, A**H denotes the
+* triangular matrix, A^T denotes the transpose of A, A^H denotes the
* conjugate transpose of A, x and b are n-element vectors, and s is a
* scaling factor, usually less than or equal to 1, chosen so that the
* components of x will be less than the overflow threshold. If the
@@ -79,8 +79,8 @@
* TRANS (input) CHARACTER*1
* Specifies the operation applied to A.
* = 'N': Solve A * x = s*b (No transpose)
-* = 'T': Solve A**T * x = s*b (Transpose)
-* = 'C': Solve A**H * x = s*b (Conjugate transpose)
+* = 'T': Solve A^T * x = s*b (Transpose)
+* = 'C': Solve A^H * x = s*b (Conjugate transpose)
*
* DIAG (input) CHARACTER*1
* Specifies whether or not the matrix A is unit triangular.
@@ -115,7 +115,7 @@
*
* SCALE (output) REAL
* The scaling factor s for the triangular system
-* A * x = s*b, A**T * x = s*b, or A**H * x = s*b.
+* A * x = s*b, A^T * x = s*b, or A^H * x = s*b.
* If SCALE = 0, the matrix A is singular or badly scaled, and
* the vector x is an exact or approximate solution to A*x = 0.
*
@@ -181,8 +181,8 @@
* prevent overflow, but if the bound overflows, x is set to 0, x(j) to
* 1, and scale to 0, and a non-trivial solution to A*x = 0 is found.
*
-* Similarly, a row-wise scheme is used to solve A**T *x = b or
-* A**H *x = b. The basic algorithm for A upper triangular is
+* Similarly, a row-wise scheme is used to solve A^T *x = b or
+* A^H *x = b. The basic algorithm for A upper triangular is
*
* for j = 1, ..., n
* x(j) := ( b(j) - A[1:j-1,j]' * x[1:j-1] ) / A(j,j)
@@ -412,7 +412,7 @@
*
ELSE
*
-* Compute the growth in A**T * x = b or A**H * x = b.
+* Compute the growth in A^T * x = b or A^H * x = b.
*
IF( UPPER ) THEN
JFIRST = 1
@@ -632,7 +632,7 @@
*
ELSE IF( LSAME( TRANS, 'T' ) ) THEN
*
-* Solve A**T * x = b
+* Solve A^T * x = b
*
DO 150 J = JFIRST, JLAST, JINC
*
@@ -744,7 +744,7 @@
ELSE
*
* A(j,j) = 0: Set x(1:n) = 0, x(j) = 1, and
-* scale = 0 and compute a solution to A**T *x = 0.
+* scale = 0 and compute a solution to A^T *x = 0.
*
DO 140 I = 1, N
X( I ) = ZERO
@@ -766,7 +766,7 @@
*
ELSE
*
-* Solve A**H * x = b
+* Solve A^H * x = b
*
DO 190 J = JFIRST, JLAST, JINC
*
@@ -880,7 +880,7 @@
ELSE
*
* A(j,j) = 0: Set x(1:n) = 0, x(j) = 1, and
-* scale = 0 and compute a solution to A**H *x = 0.
+* scale = 0 and compute a solution to A^H *x = 0.
*
DO 180 I = 1, N
X( I ) = ZERO
diff --git a/testing/lin/cpocon.f b/testing/lin/cpocon.f
index a5469bb8e..3fd5f26c4 100644
--- a/testing/lin/cpocon.f
+++ b/testing/lin/cpocon.f
@@ -59,7 +59,7 @@
*
* CPOCON estimates the reciprocal of the condition number (in the
* 1-norm) of a complex Hermitian positive definite matrix using the
-* Cholesky factorization A = U**H*U or A = L*L**H computed by CPOTRF.
+* Cholesky factorization A = U^H*U or A = L*L^H computed by CPOTRF.
*
* An estimate is obtained for norm(inv(A)), and the reciprocal of the
* condition number is computed as RCOND = 1 / (ANORM * norm(inv(A))).
@@ -76,7 +76,7 @@
*
* A (input) COMPLEX array, dimension (LDA,N)
* The triangular factor U or L from the Cholesky factorization
-* A = U**H*U or A = L*L**H, as computed by CPOTRF.
+* A = U^H*U or A = L*L^H, as computed by CPOTRF.
*
* LDA (input) INTEGER
* The leading dimension of the array A. LDA >= max(1,N).
diff --git a/testing/lin/cporfs.f b/testing/lin/cporfs.f
index 14708b5d3..616a71ce7 100644
--- a/testing/lin/cporfs.f
+++ b/testing/lin/cporfs.f
@@ -92,7 +92,7 @@
*
* AF (input) COMPLEX array, dimension (LDAF,N)
* The triangular factor U or L from the Cholesky factorization
-* A = U**H*U or A = L*L**H, as computed by CPOTRF.
+* A = U^H*U or A = L*L^H, as computed by CPOTRF.
*
* LDAF (input) INTEGER
* The leading dimension of the array AF. LDAF >= max(1,N).
diff --git a/testing/lin/cposvx.f b/testing/lin/cposvx.f
index 950a1b514..d502bd56f 100644
--- a/testing/lin/cposvx.f
+++ b/testing/lin/cposvx.f
@@ -59,7 +59,7 @@
* Purpose
* =======
*
-* CPOSVX uses the Cholesky factorization A = U**H*U or A = L*L**H to
+* CPOSVX uses the Cholesky factorization A = U^H*U or A = L*L^H to
* compute the solution to a complex system of linear equations
* A * X = B,
* where A is an N-by-N Hermitian positive definite matrix and X and B
@@ -82,8 +82,8 @@
*
* 2. If FACT = 'N' or 'E', the Cholesky decomposition is used to
* factor the matrix A (after equilibration if FACT = 'E') as
-* A = U**H* U, if UPLO = 'U', or
-* A = L * L**H, if UPLO = 'L',
+* A = U^H* U, if UPLO = 'U', or
+* A = L * L^H, if UPLO = 'L',
* where U is an upper triangular matrix and L is a lower triangular
* matrix.
*
@@ -154,18 +154,18 @@
* AF (input or output) COMPLEX array, dimension (LDAF,N)
* If FACT = 'F', then AF is an input argument and on entry
* contains the triangular factor U or L from the Cholesky
-* factorization A = U**H*U or A = L*L**H, in the same storage
+* factorization A = U^H*U or A = L*L^H, in the same storage
* format as A. If EQUED .ne. 'N', then AF is the factored form
* of the equilibrated matrix diag(S)*A*diag(S).
*
* If FACT = 'N', then AF is an output argument and on exit
* returns the triangular factor U or L from the Cholesky
-* factorization A = U**H*U or A = L*L**H of the original
+* factorization A = U^H*U or A = L*L^H of the original
* matrix A.
*
* If FACT = 'E', then AF is an output argument and on exit
* returns the triangular factor U or L from the Cholesky
-* factorization A = U**H*U or A = L*L**H of the equilibrated
+* factorization A = U^H*U or A = L*L^H of the equilibrated
* matrix A (see the description of A for the form of the
* equilibrated matrix).
*
diff --git a/testing/lin/cpotri.f b/testing/lin/cpotri.f
index 839559633..dea6b9d01 100644
--- a/testing/lin/cpotri.f
+++ b/testing/lin/cpotri.f
@@ -53,7 +53,7 @@
* =======
*
* CPOTRI computes the inverse of a complex Hermitian positive definite
-* matrix A using the Cholesky factorization A = U**H*U or A = L*L**H
+* matrix A using the Cholesky factorization A = U^H*U or A = L*L^H
* computed by CPOTRF.
*
* Arguments
@@ -68,7 +68,7 @@
*
* A (input/output) COMPLEX array, dimension (LDA,N)
* On entry, the triangular factor U or L from the Cholesky
-* factorization A = U**H*U or A = L*L**H, as computed by
+* factorization A = U^H*U or A = L*L^H, as computed by
* CPOTRF.
* On exit, the upper or lower triangle of the (Hermitian)
* inverse of A, overwriting the input factor U or L.
diff --git a/testing/lin/dpocon.f b/testing/lin/dpocon.f
index 43c957d38..1a4c1b67a 100644
--- a/testing/lin/dpocon.f
+++ b/testing/lin/dpocon.f
@@ -59,7 +59,7 @@
*
* DPOCON estimates the reciprocal of the condition number (in the
* 1-norm) of a real symmetric positive definite matrix using the
-* Cholesky factorization A = U**T*U or A = L*L**T computed by DPOTRF.
+* Cholesky factorization A = U^T*U or A = L*L^T computed by DPOTRF.
*
* An estimate is obtained for norm(inv(A)), and the reciprocal of the
* condition number is computed as RCOND = 1 / (ANORM * norm(inv(A))).
@@ -76,7 +76,7 @@
*
* A (input) DOUBLE PRECISION array, dimension (LDA,N)
* The triangular factor U or L from the Cholesky factorization
-* A = U**T*U or A = L*L**T, as computed by DPOTRF.
+* A = U^T*U or A = L*L^T, as computed by DPOTRF.
*
* LDA (input) INTEGER
* The leading dimension of the array A. LDA >= max(1,N).
diff --git a/testing/lin/dporfs.f b/testing/lin/dporfs.f
index 3a1496638..c93d5793a 100644
--- a/testing/lin/dporfs.f
+++ b/testing/lin/dporfs.f
@@ -92,7 +92,7 @@
*
* AF (input) DOUBLE PRECISION array, dimension (LDAF,N)
* The triangular factor U or L from the Cholesky factorization
-* A = U**T*U or A = L*L**T, as computed by DPOTRF.
+* A = U^T*U or A = L*L^T, as computed by DPOTRF.
*
* LDAF (input) INTEGER
* The leading dimension of the array AF. LDAF >= max(1,N).
diff --git a/testing/lin/dposvx.f b/testing/lin/dposvx.f
index aeca6aee0..79d723a27 100644
--- a/testing/lin/dposvx.f
+++ b/testing/lin/dposvx.f
@@ -61,7 +61,7 @@
* Purpose
* =======
*
-* DPOSVX uses the Cholesky factorization A = U**T*U or A = L*L**T to
+* DPOSVX uses the Cholesky factorization A = U^T*U or A = L*L^T to
* compute the solution to a real system of linear equations
* A * X = B,
* where A is an N-by-N symmetric positive definite matrix and X and B
@@ -84,8 +84,8 @@
*
* 2. If FACT = 'N' or 'E', the Cholesky decomposition is used to
* factor the matrix A (after equilibration if FACT = 'E') as
-* A = U**T* U, if UPLO = 'U', or
-* A = L * L**T, if UPLO = 'L',
+* A = U^T* U, if UPLO = 'U', or
+* A = L * L^T, if UPLO = 'L',
* where U is an upper triangular matrix and L is a lower triangular
* matrix.
*
@@ -156,18 +156,18 @@
* AF (input or output) DOUBLE PRECISION array, dimension (LDAF,N)
* If FACT = 'F', then AF is an input argument and on entry
* contains the triangular factor U or L from the Cholesky
-* factorization A = U**T*U or A = L*L**T, in the same storage
+* factorization A = U^T*U or A = L*L^T, in the same storage
* format as A. If EQUED .ne. 'N', then AF is the factored form
* of the equilibrated matrix diag(S)*A*diag(S).
*
* If FACT = 'N', then AF is an output argument and on exit
* returns the triangular factor U or L from the Cholesky
-* factorization A = U**T*U or A = L*L**T of the original
+* factorization A = U^T*U or A = L*L^T of the original
* matrix A.
*
* If FACT = 'E', then AF is an output argument and on exit
* returns the triangular factor U or L from the Cholesky
-* factorization A = U**T*U or A = L*L**T of the equilibrated
+* factorization A = U^T*U or A = L*L^T of the equilibrated
* matrix A (see the description of A for the form of the
* equilibrated matrix).
*
diff --git a/testing/lin/dpotri.f b/testing/lin/dpotri.f
index f8585b348..2a5f4c2dd 100644
--- a/testing/lin/dpotri.f
+++ b/testing/lin/dpotri.f
@@ -53,7 +53,7 @@
* =======
*
* DPOTRI computes the inverse of a real symmetric positive definite
-* matrix A using the Cholesky factorization A = U**T*U or A = L*L**T
+* matrix A using the Cholesky factorization A = U^T*U or A = L*L^T
* computed by DPOTRF.
*
* Arguments
@@ -68,7 +68,7 @@
*
* A (input/output) DOUBLE PRECISION array, dimension (LDA,N)
* On entry, the triangular factor U or L from the Cholesky
-* factorization A = U**T*U or A = L*L**T, as computed by
+* factorization A = U^T*U or A = L*L^T, as computed by
* DPOTRF.
* On exit, the upper or lower triangle of the (symmetric)
* inverse of A, overwriting the input factor U or L.
diff --git a/testing/lin/spocon.f b/testing/lin/spocon.f
index 02392607f..380896480 100644
--- a/testing/lin/spocon.f
+++ b/testing/lin/spocon.f
@@ -59,7 +59,7 @@
*
* SPOCON estimates the reciprocal of the condition number (in the
* 1-norm) of a real symmetric positive definite matrix using the
-* Cholesky factorization A = U**T*U or A = L*L**T computed by SPOTRF.
+* Cholesky factorization A = U^T*U or A = L*L^T computed by SPOTRF.
*
* An estimate is obtained for norm(inv(A)), and the reciprocal of the
* condition number is computed as RCOND = 1 / (ANORM * norm(inv(A))).
@@ -76,7 +76,7 @@
*
* A (input) REAL array, dimension (LDA,N)
* The triangular factor U or L from the Cholesky factorization
-* A = U**T*U or A = L*L**T, as computed by SPOTRF.
+* A = U^T*U or A = L*L^T, as computed by SPOTRF.
*
* LDA (input) INTEGER
* The leading dimension of the array A. LDA >= max(1,N).
diff --git a/testing/lin/sporfs.f b/testing/lin/sporfs.f
index 8dcdea760..e633b0978 100644
--- a/testing/lin/sporfs.f
+++ b/testing/lin/sporfs.f
@@ -92,7 +92,7 @@
*
* AF (input) REAL array, dimension (LDAF,N)
* The triangular factor U or L from the Cholesky factorization
-* A = U**T*U or A = L*L**T, as computed by SPOTRF.
+* A = U^T*U or A = L*L^T, as computed by SPOTRF.
*
* LDAF (input) INTEGER
* The leading dimension of the array AF. LDAF >= max(1,N).
diff --git a/testing/lin/sposvx.f b/testing/lin/sposvx.f
index b8a94475a..8a8f53564 100644
--- a/testing/lin/sposvx.f
+++ b/testing/lin/sposvx.f
@@ -61,7 +61,7 @@
* Purpose
* =======
*
-* SPOSVX uses the Cholesky factorization A = U**T*U or A = L*L**T to
+* SPOSVX uses the Cholesky factorization A = U^T*U or A = L*L^T to
* compute the solution to a real system of linear equations
* A * X = B,
* where A is an N-by-N symmetric positive definite matrix and X and B
@@ -84,8 +84,8 @@
*
* 2. If FACT = 'N' or 'E', the Cholesky decomposition is used to
* factor the matrix A (after equilibration if FACT = 'E') as
-* A = U**T* U, if UPLO = 'U', or
-* A = L * L**T, if UPLO = 'L',
+* A = U^T* U, if UPLO = 'U', or
+* A = L * L^T, if UPLO = 'L',
* where U is an upper triangular matrix and L is a lower triangular
* matrix.
*
@@ -156,18 +156,18 @@
* AF (input or output) REAL array, dimension (LDAF,N)
* If FACT = 'F', then AF is an input argument and on entry
* contains the triangular factor U or L from the Cholesky
-* factorization A = U**T*U or A = L*L**T, in the same storage
+* factorization A = U^T*U or A = L*L^T, in the same storage
* format as A. If EQUED .ne. 'N', then AF is the factored form
* of the equilibrated matrix diag(S)*A*diag(S).
*
* If FACT = 'N', then AF is an output argument and on exit
* returns the triangular factor U or L from the Cholesky
-* factorization A = U**T*U or A = L*L**T of the original
+* factorization A = U^T*U or A = L*L^T of the original
* matrix A.
*
* If FACT = 'E', then AF is an output argument and on exit
* returns the triangular factor U or L from the Cholesky
-* factorization A = U**T*U or A = L*L**T of the equilibrated
+* factorization A = U^T*U or A = L*L^T of the equilibrated
* matrix A (see the description of A for the form of the
* equilibrated matrix).
*
diff --git a/testing/lin/spotri.f b/testing/lin/spotri.f
index 13885e2fd..d52f05699 100644
--- a/testing/lin/spotri.f
+++ b/testing/lin/spotri.f
@@ -53,7 +53,7 @@
* =======
*
* SPOTRI computes the inverse of a real symmetric positive definite
-* matrix A using the Cholesky factorization A = U**T*U or A = L*L**T
+* matrix A using the Cholesky factorization A = U^T*U or A = L*L^T
* computed by SPOTRF.
*
* Arguments
@@ -68,7 +68,7 @@
*
* A (input/output) REAL array, dimension (LDA,N)
* On entry, the triangular factor U or L from the Cholesky
-* factorization A = U**T*U or A = L*L**T, as computed by
+* factorization A = U^T*U or A = L*L^T, as computed by
* SPOTRF.
* On exit, the upper or lower triangle of the (symmetric)
* inverse of A, overwriting the input factor U or L.
diff --git a/testing/lin/zlagsy.f b/testing/lin/zlagsy.f
index d2a05500d..a9366c90c 100644
--- a/testing/lin/zlagsy.f
+++ b/testing/lin/zlagsy.f
@@ -55,7 +55,7 @@
*
* ZLAGSY generates a complex symmetric matrix A, by pre- and post-
* multiplying a real diagonal matrix D with a random unitary matrix:
-* A = U*D*U**T. The semi-bandwidth may then be reduced to k by
+* A = U*D*U^T. The semi-bandwidth may then be reduced to k by
* additional unitary transformations.
*
* Arguments
diff --git a/testing/lin/zlarhs.f b/testing/lin/zlarhs.f
index 1da073177..333feeb71 100644
--- a/testing/lin/zlarhs.f
+++ b/testing/lin/zlarhs.f
@@ -58,7 +58,7 @@
* ZLARHS chooses a set of NRHS random solution vectors and sets
* up the right hand sides for the linear system
* op( A ) * X = B,
-* where op( A ) may be A, A**T (transpose of A), or A**H (conjugate
+* where op( A ) may be A, A^T (transpose of A), or A^H (conjugate
* transpose of A).
*
* Arguments
@@ -102,8 +102,8 @@
* Used only if A is nonsymmetric; specifies the operation
* applied to the matrix A.
* = 'N': B := A * X
-* = 'T': B := A**T * X
-* = 'C': B := A**H * X
+* = 'T': B := A^T * X
+* = 'C': B := A^H * X
*
* M (input) INTEGER
* The number of rows of the matrix A. M >= 0.
diff --git a/testing/lin/zlatrs.f b/testing/lin/zlatrs.f
index c4271a045..ba7f497ef 100644
--- a/testing/lin/zlatrs.f
+++ b/testing/lin/zlatrs.f
@@ -57,10 +57,10 @@
*
* ZLATRS solves one of the triangular systems
*
-* A * x = s*b, A**T * x = s*b, or A**H * x = s*b,
+* A * x = s*b, A^T * x = s*b, or A^H * x = s*b,
*
* with scaling to prevent overflow. Here A is an upper or lower
-* triangular matrix, A**T denotes the transpose of A, A**H denotes the
+* triangular matrix, A^T denotes the transpose of A, A^H denotes the
* conjugate transpose of A, x and b are n-element vectors, and s is a
* scaling factor, usually less than or equal to 1, chosen so that the
* components of x will be less than the overflow threshold. If the
@@ -79,8 +79,8 @@
* TRANS (input) CHARACTER*1
* Specifies the operation applied to A.
* = 'N': Solve A * x = s*b (No transpose)
-* = 'T': Solve A**T * x = s*b (Transpose)
-* = 'C': Solve A**H * x = s*b (Conjugate transpose)
+* = 'T': Solve A^T * x = s*b (Transpose)
+* = 'C': Solve A^H * x = s*b (Conjugate transpose)
*
* DIAG (input) CHARACTER*1
* Specifies whether or not the matrix A is unit triangular.
@@ -115,7 +115,7 @@
*
* SCALE (output) DOUBLE PRECISION
* The scaling factor s for the triangular system
-* A * x = s*b, A**T * x = s*b, or A**H * x = s*b.
+* A * x = s*b, A^T * x = s*b, or A^H * x = s*b.
* If SCALE = 0, the matrix A is singular or badly scaled, and
* the vector x is an exact or approximate solution to A*x = 0.
*
@@ -181,8 +181,8 @@
* prevent overflow, but if the bound overflows, x is set to 0, x(j) to
* 1, and scale to 0, and a non-trivial solution to A*x = 0 is found.
*
-* Similarly, a row-wise scheme is used to solve A**T *x = b or
-* A**H *x = b. The basic algorithm for A upper triangular is
+* Similarly, a row-wise scheme is used to solve A^T *x = b or
+* A^H *x = b. The basic algorithm for A upper triangular is
*
* for j = 1, ..., n
* x(j) := ( b(j) - A[1:j-1,j]' * x[1:j-1] ) / A(j,j)
@@ -412,7 +412,7 @@
*
ELSE
*
-* Compute the growth in A**T * x = b or A**H * x = b.
+* Compute the growth in A^T * x = b or A^H * x = b.
*
IF( UPPER ) THEN
JFIRST = 1
@@ -632,7 +632,7 @@
*
ELSE IF( LSAME( TRANS, 'T' ) ) THEN
*
-* Solve A**T * x = b
+* Solve A^T * x = b
*
DO 170 J = JFIRST, JLAST, JINC
*
@@ -744,7 +744,7 @@
ELSE
*
* A(j,j) = 0: Set x(1:n) = 0, x(j) = 1, and
-* scale = 0 and compute a solution to A**T *x = 0.
+* scale = 0 and compute a solution to A^T *x = 0.
*
DO 150 I = 1, N
X( I ) = ZERO
@@ -766,7 +766,7 @@
*
ELSE
*
-* Solve A**H * x = b
+* Solve A^H * x = b
*
DO 220 J = JFIRST, JLAST, JINC
*
@@ -880,7 +880,7 @@
ELSE
*
* A(j,j) = 0: Set x(1:n) = 0, x(j) = 1, and
-* scale = 0 and compute a solution to A**H *x = 0.
+* scale = 0 and compute a solution to A^H *x = 0.
*
DO 200 I = 1, N
X( I ) = ZERO
diff --git a/testing/lin/zpocon.f b/testing/lin/zpocon.f
index 4ead889ed..b3e91f057 100644
--- a/testing/lin/zpocon.f
+++ b/testing/lin/zpocon.f
@@ -59,7 +59,7 @@
*
* ZPOCON estimates the reciprocal of the condition number (in the
* 1-norm) of a complex Hermitian positive definite matrix using the
-* Cholesky factorization A = U**H*U or A = L*L**H computed by ZPOTRF.
+* Cholesky factorization A = U^H*U or A = L*L^H computed by ZPOTRF.
*
* An estimate is obtained for norm(inv(A)), and the reciprocal of the
* condition number is computed as RCOND = 1 / (ANORM * norm(inv(A))).
@@ -76,7 +76,7 @@
*
* A (input) COMPLEX*16 array, dimension (LDA,N)
* The triangular factor U or L from the Cholesky factorization
-* A = U**H*U or A = L*L**H, as computed by ZPOTRF.
+* A = U^H*U or A = L*L^H, as computed by ZPOTRF.
*
* LDA (input) INTEGER
* The leading dimension of the array A. LDA >= max(1,N).
diff --git a/testing/lin/zporfs.f b/testing/lin/zporfs.f
index 4503ef94a..696739a6c 100644
--- a/testing/lin/zporfs.f
+++ b/testing/lin/zporfs.f
@@ -92,7 +92,7 @@
*
* AF (input) COMPLEX*16 array, dimension (LDAF,N)
* The triangular factor U or L from the Cholesky factorization
-* A = U**H*U or A = L*L**H, as computed by ZPOTRF.
+* A = U^H*U or A = L*L^H, as computed by ZPOTRF.
*
* LDAF (input) INTEGER
* The leading dimension of the array AF. LDAF >= max(1,N).
diff --git a/testing/lin/zposvx.f b/testing/lin/zposvx.f
index 5fe686b41..4d8d322fe 100644
--- a/testing/lin/zposvx.f
+++ b/testing/lin/zposvx.f
@@ -59,7 +59,7 @@
* Purpose
* =======
*
-* ZPOSVX uses the Cholesky factorization A = U**H*U or A = L*L**H to
+* ZPOSVX uses the Cholesky factorization A = U^H*U or A = L*L^H to
* compute the solution to a complex system of linear equations
* A * X = B,
* where A is an N-by-N Hermitian positive definite matrix and X and B
@@ -82,8 +82,8 @@
*
* 2. If FACT = 'N' or 'E', the Cholesky decomposition is used to
* factor the matrix A (after equilibration if FACT = 'E') as
-* A = U**H* U, if UPLO = 'U', or
-* A = L * L**H, if UPLO = 'L',
+* A = U^H* U, if UPLO = 'U', or
+* A = L * L^H, if UPLO = 'L',
* where U is an upper triangular matrix and L is a lower triangular
* matrix.
*
@@ -154,18 +154,18 @@
* AF (input or output) COMPLEX*16 array, dimension (LDAF,N)
* If FACT = 'F', then AF is an input argument and on entry
* contains the triangular factor U or L from the Cholesky
-* factorization A = U**H*U or A = L*L**H, in the same storage
+* factorization A = U^H*U or A = L*L^H, in the same storage
* format as A. If EQUED .ne. 'N', then AF is the factored form
* of the equilibrated matrix diag(S)*A*diag(S).
*
* If FACT = 'N', then AF is an output argument and on exit
* returns the triangular factor U or L from the Cholesky
-* factorization A = U**H*U or A = L*L**H of the original
+* factorization A = U^H*U or A = L*L^H of the original
* matrix A.
*
* If FACT = 'E', then AF is an output argument and on exit
* returns the triangular factor U or L from the Cholesky
-* factorization A = U**H*U or A = L*L**H of the equilibrated
+* factorization A = U^H*U or A = L*L^H of the equilibrated
* matrix A (see the description of A for the form of the
* equilibrated matrix).
*
diff --git a/testing/lin/zpotri.f b/testing/lin/zpotri.f
index bf86de0ee..ed9fd8dd6 100644
--- a/testing/lin/zpotri.f
+++ b/testing/lin/zpotri.f
@@ -53,7 +53,7 @@
* =======
*
* ZPOTRI computes the inverse of a complex Hermitian positive definite
-* matrix A using the Cholesky factorization A = U**H*U or A = L*L**H
+* matrix A using the Cholesky factorization A = U^H*U or A = L*L^H
* computed by ZPOTRF.
*
* Arguments
@@ -68,7 +68,7 @@
*
* A (input/output) COMPLEX*16 array, dimension (LDA,N)
* On entry, the triangular factor U or L from the Cholesky
-* factorization A = U**H*U or A = L*L**H, as computed by
+* factorization A = U^H*U or A = L*L^H, as computed by
* ZPOTRF.
* On exit, the upper or lower triangle of the (Hermitian)
* inverse of A, overwriting the input factor U or L.
diff --git a/timing/timing_zauxiliary.c b/timing/timing_zauxiliary.c
index f804164b6..007d57c85 100644
--- a/timing/timing_zauxiliary.c
+++ b/timing/timing_zauxiliary.c
@@ -314,8 +314,8 @@ double z_check_solution(int M, int N, int NRHS, CHAMELEON_Complex64_t *A, int LD
* * Check the accuracy of the computed inverse
* */
-int zcheck_inverse(int N, CHAMELEON_Complex64_t *A1, CHAMELEON_Complex64_t *A2, int LDA,
- cham_uplo_t uplo, double *rnorm, double *anorm, double *ainvnorm )
+int z_check_inverse( int N, CHAMELEON_Complex64_t *A1, CHAMELEON_Complex64_t *A2, int LDA,
+ cham_uplo_t uplo, double *rnorm, double *anorm, double *ainvnorm )
{
int info_inverse;
int i, j;
diff --git a/timing/timing_zauxiliary.h b/timing/timing_zauxiliary.h
index 2c9957f52..6fbb0ff15 100644
--- a/timing/timing_zauxiliary.h
+++ b/timing/timing_zauxiliary.h
@@ -39,8 +39,8 @@ double z_check_solution(int M, int N, int NRHS,
CHAMELEON_Complex64_t *B1, CHAMELEON_Complex64_t *B2, int LDB,
double *anorm, double *bnorm, double *xnorm);
-int zcheck_inverse(int N, CHAMELEON_Complex64_t *A1, CHAMELEON_Complex64_t *A2,
- int LDA, cham_uplo_t uplo, double *rnorm, double *anorm, double *ainvnorm);
+int z_check_inverse( int N, CHAMELEON_Complex64_t *A1, CHAMELEON_Complex64_t *A2,
+ int LDA, cham_uplo_t uplo, double *rnorm, double *anorm, double *ainvnorm );
#endif /* _timing_zauxiliary_h_ */
--
GitLab