unification of single/double precission for real elpa1 and elpa2 compute templates

added printing of test-suite.log to gitlab ci

elpa 1 real/complex unification

first part of elpa2 single/double

next phase

part 3

till line 2010

till line 2164

till line 2346

till line 2425

till line 2280

till line 2656, two fixes

rest of elpa2_real_template

shortened line

Conflicts:
	.gitlab-ci.yml
	src/elpa1_compute_real_template.X90
	src/elpa2_compute_real_template.X90

Makefile.am

@@ -54,7 +54,8 @@ EXTRA_libelpa@SUFFIX@_private_la_DEPENDENCIES = \
 	src/elpa2_kernels/elpa2_kernels_real_template.X90 \
 	src/elpa2_kernels/elpa2_kernels_complex_template.X90 \
 	src/elpa2_kernels/elpa2_kernels_simple_template.X90 \
-        src/redist_band.X90
+        src/redist_band.X90 \
+        src/precision_macros.h
 
 lib_LTLIBRARIES = libelpa@SUFFIX@.la
 libelpa@SUFFIX@_la_LINK = $(FCLINK) $(AM_LDFLAGS) -version-info $(ELPA_SO_VERSION)
@@ -846,6 +847,7 @@ EXTRA_DIST = \
   src/elpa1_compute_complex_template.X90 \
   src/elpa2_compute_real_template.X90 \
   src/elpa2_compute_complex_template.X90 \
+  src/precision_macros.h \
   src/elpa2_kernels/elpa2_kernels_real_template.X90 \
   src/elpa2_kernels/elpa2_kernels_complex_template.X90 \
   src/elpa2_kernels/elpa2_kernels_simple_template.X90 \

src/elpa1_compute_private.F90

@@ -215,6 +215,7 @@ module ELPA1_COMPUTE
 #define DOUBLE_PRECISION_REAL 1
 #define REAL_DATATYPE rk8
 
+#include "precision_macros.h"
 #include "elpa1_compute_real_template.X90"
 
 #undef DOUBLE_PRECISION_REAL
@@ -226,6 +227,7 @@ module ELPA1_COMPUTE
 #undef DOUBLE_PRECISION_REAL
 #define REAL_DATATYPE rk4
 
+#include "precision_macros.h"
 #include "elpa1_compute_real_template.X90"
 
 #undef DOUBLE_PRECISION_REAL

src/elpa2_compute.F90

@@ -167,6 +167,7 @@ module ELPA2_compute
 #define DOUBLE_PRECISION_REAL 1
 #define REAL_DATATYPE rk8
 
+#include "precision_macros.h"
 #include "elpa2_compute_real_template.X90"
 
 #undef DOUBLE_PRECISION_REAL
@@ -178,6 +179,7 @@ module ELPA2_compute
 #undef DOUBLE_PRECISION_REAL
 #define REAL_DATATYPE rk4
 
+#include "precision_macros.h"
 #include "elpa2_compute_real_template.X90"
 
 #undef DOUBLE_PRECISION_REAL
@@ -189,6 +191,7 @@ module ELPA2_compute
 #define DOUBLE_PRECISION_COMPLEX 1
 #define REAL_DATATYPE rk8
 #define COMPLEX_DATATYPE ck8
+
 #include "elpa2_compute_complex_template.X90"
 
 #undef DOUBLE_PRECISION_COMPLEX

src/precision_macros.h

+#ifdef DOUBLE_PRECISION_REAL
+#define M_elpa_transpose_vectors_real_PRECISSION elpa_transpose_vectors_real_double
+#define M_elpa_reduce_add_vectors_real_PRECISSION elpa_reduce_add_vectors_real_double
+
+#define M_bandred_real_PRECISSION bandred_real_double
+#define M_trans_ev_band_to_full_real_PRECISSION trans_ev_band_to_full_real_double
+#define M_tridiag_band_real_PRECISSION tridiag_band_real_double
+#define M_trans_ev_tridi_to_band_real_PRECISSION trans_ev_tridi_to_band_real_double
+#define M_band_band_real_PRECISSION band_band_real_double
+#define M_tridiag_real_PRECISSION tridiag_real_double
+#define M_trans_ev_real_PRECISSION trans_ev_real_double
+#define M_solve_tridi_PRECISSION solve_tridi_double
+#define M_solve_tridi_col_PRECISSION solve_tridi_col_double
+#define M_solve_tridi_single_problem_PRECISSION solve_tridi_single_problem_double
+
+#define M_qr_pdgeqrf_2dcomm_PRECISSION qr_pdgeqrf_2dcomm_double
+#define M_hh_transform_real_PRECISSION hh_transform_real_double
+#define M_symm_matrix_allreduce_PRECISSION symm_matrix_allreduce_double
+#define M_redist_band_real_PRECISSION redist_band_real_double
+#define M_unpack_row_real_cpu_PRECISSION unpack_row_real_cpu_double
+#define M_unpack_row_real_cpu_openmp_PRECISSION unpack_row_real_cpu_openmp_double
+#define M_unpack_and_prepare_row_group_real_gpu_PRECISSION unpack_and_prepare_row_group_real_gpu_double
+#define M_extract_hh_tau_real_gpu_PRECISSION extract_hh_tau_real_gpu_double
+#define M_compute_hh_dot_products_real_gpu_PRECISSION compute_hh_dot_products_real_gpu_double
+#define M_compute_hh_trafo_real_cpu_openmp_PRECISSION compute_hh_trafo_real_cpu_openmp_double
+#define M_compute_hh_trafo_real_cpu_PRECISSION compute_hh_trafo_real_cpu_double
+#define M_pack_row_group_real_gpu_PRECISSION pack_row_group_real_gpu_double
+#define M_pack_row_real_cpu_openmp_PRECISSION pack_row_real_cpu_openmp_double
+#define M_pack_row_real_cpu_PRECISSION pack_row_real_cpu_double
+#define M_wy_gen_PRECISSION wy_gen_double
+#define M_wy_right_PRECISSION wy_right_double
+#define M_wy_left_PRECISSION wy_left_double
+#define M_wy_symm_PRECISSION wy_symm_double
+#define M_merge_recursive_PRECISSION merge_recursive_double
+#define M_merge_systems_PRECISSION merge_systems_double
+#define M_distribute_global_column_PRECISSION distribute_global_column_double
+#define M_check_monotony_PRECISSION check_monotony_double
+#define M_global_gather_PRECISSION global_gather_double
+#define M_resort_ev_PRECISSION resort_ev_double
+#define M_transform_columns_PRECISSION transform_columns_double
+#define M_solve_secular_equation_PRECISSION solve_secular_equation_double
+#define M_global_product_PRECISSION global_product_double
+#define M_add_tmp_PRECISSION add_tmp_double
+#define M_v_add_s_PRECISSION v_add_s_double
+
+#define M_PRECISSION_SYRK DSYRK
+#define M_PRECISSION_TRMV DTRMV
+#define M_PRECISSION_GEMM DGEMM
+#define M_PRECISSION_GEMV DGEMV 
+#define M_PRECISSION_TRMM DTRMM
+#define M_PRECISSION_SYMV DSYMV
+#define M_PRECISSION_SYMM DSYMM
+#define M_PRECISSION_SYR2 DSYR2
+#define M_PRECISSION_SYR2K DSYR2K
+#define M_PRECISSION_GEQRF dgeqrf
+#define M_PRECISSION_STEDC dstedc
+#define M_PRECISSION_STEQR dsteqr
+#define M_PRECISSION_LAMRG DLAMRG
+#define M_PRECISSION_LAMCH DLAMCH
+#define M_PRECISSION_LAPY2 DLAPY2
+#define M_PRECISSION_LAED4 DLAED4
+#define M_PRECISSION_LAED5 DLAED5
+
+#define M_cublas_PRECISSION_gemm cublas_dgemm
+#define M_cublas_PRECISSION_trmm cublas_dtrmm
+
+#define M_PRECISSION_SUFFIX "_double"
+#define M_CONST_0_0 0.0_rk8
+#define M_CONST_0_5 0.5_rk8
+#define M_CONST_1_0 1.0_rk8
+#define M_CONST_2_0 2.0_rk8
+#define M_CONST_8_0 8.0_rk8
+#define M_size_of_PRECISSION_real_datatape size_of_double_real_datatype
+#define M_MPI_REAL_PRECISSION MPI_REAL8
+
+#else
+
+#undef M_elpa_transpose_vectors_real_PRECISSION
+#undef M_elpa_reduce_add_vectors_real_PRECISSION
+
+#undef M_bandred_real_PRECISSION
+#undef M_trans_ev_band_to_full_real_PRECISSION
+#undef M_tridiag_band_real_PRECISSION 
+#undef M_trans_ev_tridi_to_band_real_PRECISSION
+#undef M_band_band_real_PRECISSION
+#undef M_tridiag_real_PRECISSION 
+#undef M_trans_ev_real_PRECISSION
+#undef M_solve_tridi_PRECISSION
+#undef M_solve_tridi_col_PRECISSION
+#undef M_solve_tridi_single_problem_PRECISSION 
+
+#undef M_qr_pdgeqrf_2dcomm_PRECISSION
+#undef M_hh_transform_real_PRECISSION
+#undef M_symm_matrix_allreduce_PRECISSION
+#undef M_redist_band_real_PRECISSION
+#undef M_unpack_row_real_cpu_PRECISSION 
+#undef M_unpack_row_real_cpu_openmp_PRECISSION 
+#undef M_unpack_and_prepare_row_group_real_gpu_PRECISSION 
+#undef M_extract_hh_tau_real_gpu_PRECISSION 
+#undef M_compute_hh_dot_products_real_gpu_PRECISSION 
+#undef M_compute_hh_trafo_real_cpu_openmp_PRECISSION 
+#undef M_compute_hh_trafo_real_cpu_PRECISSION 
+#undef M_pack_row_group_real_gpu_PRECISSION
+#undef M_pack_row_real_cpu_openmp_PRECISSION 
+#undef M_pack_row_real_cpu_PRECISSION
+#undef M_wy_gen_PRECISSION 
+#undef M_wy_right_PRECISSION
+#undef M_wy_left_PRECISSION
+#undef M_wy_symm_PRECISSION
+#undef M_merge_recursive_PRECISSION
+#undef M_merge_systems_PRECISSION
+#undef M_distribute_global_column_PRECISSION 
+#undef M_check_monotony_PRECISSION
+#undef M_global_gather_PRECISSION 
+#undef M_resort_ev_PRECISSION
+#undef M_transform_columns_PRECISSION
+#undef M_solve_secular_equation_PRECISSION 
+#undef M_global_product_PRECISSION 
+#undef M_add_tmp_PRECISSION
+#undef M_v_add_s_PRECISSION
+
+#undef M_PRECISSION_SYRK 
+#undef M_PRECISSION_TRMV 
+#undef M_PRECISSION_GEMM 
+#undef M_PRECISSION_GEMV
+#undef M_PRECISSION_TRMM 
+#undef M_PRECISSION_SYMV 
+#undef M_PRECISSION_SYMM 
+#undef M_PRECISSION_SYR2
+#undef M_PRECISSION_SYR2K
+#undef M_PRECISSION_GEQRF
+#undef M_PRECISSION_STEDC 
+#undef M_PRECISSION_STEQR 
+#undef M_PRECISSION_LAMRG
+#undef M_PRECISSION_LAMCH
+#undef M_PRECISSION_LAPY2
+#undef M_PRECISSION_LAED4
+#undef M_PRECISSION_LAED5
+
+#undef M_cublas_PRECISSION_gemm
+#undef M_cublas_PRECISSION_trmm 
+
+#undef M_PRECISSION_SUFFIX
+#undef M_CONST_0_0 
+#undef M_CONST_0_5 
+#undef M_CONST_1_0 
+#undef M_CONST_2_0
+#undef M_CONST_8_0
+#undef M_size_of_PRECISSION_real_datatape
+#undef M_MPI_REAL_PRECISSION
+
+#define M_elpa_transpose_vectors_real_PRECISSION elpa_transpose_vectors_real_single
+#define M_elpa_reduce_add_vectors_real_PRECISSION elpa_reduce_add_vectors_real_single
+
+#define M_bandred_real_PRECISSION bandred_real_single
+#define M_trans_ev_band_to_full_real_PRECISSION trans_ev_band_to_full_real_single
+#define M_tridiag_band_real_PRECISSION tridiag_band_real_single
+#define M_trans_ev_tridi_to_band_real_PRECISSION trans_ev_tridi_to_band_real_single
+#define M_band_band_real_PRECISSION band_band_real_single
+#define M_tridiag_real_PRECISSION tridiag_real_single
+#define M_trans_ev_real_PRECISSION trans_ev_real_single
+#define M_solve_tridi_PRECISSION solve_tridi_single
+#define M_solve_tridi_col_PRECISSION solve_tridi_col_single
+#define M_solve_tridi_single_problem_PRECISSION solve_tridi_single_problem_single
+
+#define M_qr_pdgeqrf_2dcomm_PRECISSION qr_pdgeqrf_2dcomm_single
+#define M_hh_transform_real_PRECISSION hh_transform_real_single
+#define M_symm_matrix_allreduce_PRECISSION symm_matrix_allreduce_single
+#define M_redist_band_real_PRECISSION redist_band_real_single
+#define M_unpack_row_real_cpu_PRECISSION unpack_row_real_cpu_single
+#define M_unpack_row_real_cpu_openmp_PRECISSION unpack_row_real_cpu_openmp_single
+#define M_unpack_and_prepare_row_group_real_gpu_PRECISSION unpack_and_prepare_row_group_real_gpu_single
+#define M_extract_hh_tau_real_gpu_PRECISSION extract_hh_tau_real_gpu_single
+#define M_compute_hh_dot_products_real_gpu_PRECISSION compute_hh_dot_products_real_gpu_single
+#define M_compute_hh_trafo_real_cpu_openmp_PRECISSION compute_hh_trafo_real_cpu_openmp_single
+#define M_compute_hh_trafo_real_cpu_PRECISSION compute_hh_trafo_real_cpu_single
+#define M_pack_row_group_real_gpu_PRECISSION pack_row_group_real_gpu_single
+#define M_pack_row_real_cpu_openmp_PRECISSION pack_row_real_cpu_openmp_single
+#define M_pack_row_real_cpu_PRECISSION pack_row_real_cpu_single
+#define M_wy_gen_PRECISSION wy_gen_single
+#define M_wy_right_PRECISSION wy_right_single
+#define M_wy_left_PRECISSION wy_left_single
+#define M_wy_symm_PRECISSION wy_symm_single
+#define M_merge_recursive_PRECISSION merge_recursive_single
+#define M_merge_systems_PRECISSION merge_systems_single
+#define M_distribute_global_column_PRECISSION distribute_global_column_single
+#define M_check_monotony_PRECISSION check_monotony_single
+#define M_global_gather_PRECISSION global_gather_single
+#define M_resort_ev_PRECISSION resort_ev_single
+#define M_transform_columns_PRECISSION transform_columns_single
+#define M_solve_secular_equation_PRECISSION solve_secular_equation_single
+#define M_global_product_PRECISSION global_product_single
+#define M_add_tmp_PRECISSION add_tmp_single
+#define M_v_add_s_PRECISSION v_add_s_single
+
+#define M_PRECISSION_SYRK SSYRK
+#define M_PRECISSION_TRMV STRMV
+#define M_PRECISSION_GEMM SGEMM
+#define M_PRECISSION_GEMV SGEMV
+#define M_PRECISSION_TRMM STRMM
+#define M_PRECISSION_SYMV SSYMV
+#define M_PRECISSION_SYMM SSYMM
+#define M_PRECISSION_SYR2 SSYR2
+#define M_PRECISSION_SYR2K SSYR2K
+#define M_PRECISSION_GEQRF sgeqrf
+#define M_PRECISSION_STEDC sstedc
+#define M_PRECISSION_STEQR ssteqr
+#define M_PRECISSION_LAMRG SLAMRG
+#define M_PRECISSION_LAMCH SLAMCH
+#define M_PRECISSION_LAPY2 SLAPY2
+#define M_PRECISSION_LAED4 SLAED4
+#define M_PRECISSION_LAED5 SLAED5
+
+#define M_cublas_PRECISSION_gemm cublas_sgemm
+#define M_cublas_PRECISSION_trmm cublas_strmm
+
+#define M_PRECISSION_SUFFIX "_single"
+#define M_CONST_0_0 0.0_rk4
+#define M_CONST_0_5 0.5_rk4
+#define M_CONST_1_0 1.0_rk4
+#define M_CONST_2_0 2.0_rk4
+#define M_CONST_8_0 8.0_rk4
+#define M_size_of_PRECISSION_real_datatape size_of_single_real_datatype
+#define M_MPI_REAL_PRECISSION MPI_REAL4
+#endif
+