Test complex AVX Block2 kernel

Makefile.am

@@ -794,7 +794,6 @@ EXTRA_DIST = \
   src/elpa2/kernels/complex_avx512_1hv_template.c \
   src/elpa2/kernels/complex_avx512_2hv_template.c \
   src/elpa2/kernels/complex_128bit_256bit_512bit_BLOCK_template.c \
-  src/elpa2/kernels/complex_avx-avx2_2hv_template.c \
   src/elpa2/kernels/complex_template.F90 \
   src/elpa2/kernels/real_vsx_4hv_template.c \
   src/elpa2/kernels/real_vsx_6hv_template.c \

ci_test_scripts/run_ci_tests.sh

@@ -190,7 +190,10 @@ then
 
   #if [ $exitCode -ne 0 ]
   #then
+  if [ -f ./test-suite.log ]
+  then
     cat ./test-suite.log
+  fi
   #fi
 
   exit $exitCode

ci_test_scripts/run_distcheck_tests.sh

@@ -130,6 +130,10 @@ then
     #then
       cat ./ELPA_CI.err.*
     #fi
+    if [ -f ./test-suite.log ]
+    then
+      cat ./test-suite.log
+    fi
 
   fi
 

ci_test_scripts/run_project_tests.sh

@@ -188,6 +188,10 @@ then
     #then
       cat ./ELPA_CI.err.*
     #fi
+    if [ -f ./test-suite.log ]
+    then
+      cat ./test-suite.log
+    fi
 
   fi
 

src/elpa2/kernels/complex_128bit_256bit_512bit_BLOCK_template.c

@@ -528,6 +528,36 @@ static __forceinline void CONCAT_8ARGS(hh_trafo_complex_kernel_,ROW_LENGTH,_,SIM
 !f>#endif
 */
 
+/*
+!f>#if defined(HAVE_AVX) || defined(HAVE_AVX2)
+!f> interface
+!f>   subroutine double_hh_trafo_complex_AVX_AVX2_2hv_double(q, hh, pnb, pnq, pldq, pldh) &
+!f>                                bind(C, name="double_hh_trafo_complex_AVX_AVX2_2hv_double")
+!f>        use, intrinsic :: iso_c_binding
+!f>        integer(kind=c_int)        :: pnb, pnq, pldq, pldh
+!f>        ! complex(kind=c_double_complex)     :: q(*)
+!f>        type(c_ptr), value                     :: q
+!f>        complex(kind=c_double_complex)           :: hh(pnb,2)
+!f>   end subroutine
+!f> end interface
+!f>#endif
+*/
+
+/*
+!f>#if defined(HAVE_AVX) || defined(HAVE_AVX2)
+!f> interface
+!f>   subroutine double_hh_trafo_complex_AVX_AVX2_2hv_single(q, hh, pnb, pnq, pldq, pldh) &
+!f>                                bind(C, name="double_hh_trafo_complex_AVX_AVX2_2hv_single")
+!f>        use, intrinsic :: iso_c_binding
+!f>        integer(kind=c_int)        :: pnb, pnq, pldq, pldh
+!f>        ! complex(kind=c_float_complex)   :: q(*)
+!f>        type(c_ptr), value                  :: q
+!f>        complex(kind=c_float_complex)        :: hh(pnb,2)
+!f>   end subroutine
+!f> end interface
+!f>#endif
+*/
+
 void CONCAT_7ARGS(PREFIX,_hh_trafo_complex_,SIMD_SET,_,BLOCK,hv_,WORD_LENGTH) (DATA_TYPE_PTR q, DATA_TYPE_PTR hh, int* pnb, int* pnq, int* pldq
 #ifdef BLOCK1
 		  )

src/elpa2/kernels/complex_avx-avx2_2hv_double_precision.c

@@ -51,7 +51,7 @@
 #define VEC_SET AVX_256
 #define BLOCK2 1
 #include "../../general/precision_macros.h"
-#include "complex_avx-avx2_2hv_template.c"
+#include "complex_128bit_256bit_512bit_BLOCK_template.c"
 #undef VEC_SET
 #undef BLOCK2
 #undef DOUBLE_PRECISION

src/elpa2/kernels/complex_avx-avx2_2hv_single_precision.c

@@ -51,7 +51,7 @@
 #define VEC_SET AVX_256
 #define BLOCK2 1
 #include "../../general/precision_macros.h"
-#include "complex_avx-avx2_2hv_template.c"
+#include "complex_128bit_256bit_512bit_BLOCK_template.c"
 #undef VEC_SET
 #undef BLOCK2
 #undef SINGLE_PRECISION