diff --git a/Makefile.am b/Makefile.am index 7c93ae230b882075500a3110bbb2acf37e89c98a..a346618957d4f6c12f5a26cfe0aea22f5079ab4a 100644 --- a/Makefile.am +++ b/Makefile.am @@ -26,8 +26,8 @@ libelpa@SUFFIX@_private_la_SOURCES = \ src/helpers/mod_precision.F90 \ src/helpers/mod_mpi.F90 \ src/helpers/mod_mpi_stubs.F90 \ - src/mod_redist_band.F90 \ - src/mod_pack_unpack_cpu.F90 \ + src/elpa2/mod_redist_band.F90 \ + src/elpa2/mod_pack_unpack_cpu.F90 \ src/elpa2/mod_compute_hh_trafo.F90 \ src/helpers/aligned_mem.F90 \ src/elpa1/elpa1_compute_private.F90 \ @@ -38,8 +38,8 @@ libelpa@SUFFIX@_private_la_SOURCES = \ src/elpa_c_interface.F90 \ src/GPU/check_for_gpu.F90 \ src/GPU/mod_cuda.F90 \ - src/interface_c_kernel.F90 \ - src/mod_pack_unpack_gpu.F90 \ + src/elpa2/GPU/interface_c_kernel.F90 \ + src/elpa2/mod_pack_unpack_gpu.F90 \ src/elpa2/qr/qr_utils.F90 \ src/elpa2/qr/elpa_qrkernels.F90 \ src/elpa2/qr/elpa_pdlarfb.F90 \ @@ -70,10 +70,10 @@ 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/pack_unpack_cpu.X90 \ - src/pack_unpack_gpu.X90 \ + src/elpa2/pack_unpack_cpu.X90 \ + src/elpa2/pack_unpack_gpu.X90 \ src/elpa2/compute_hh_trafo.X90 \ - src/redist_band.X90 \ + src/elpa2/redist_band.X90 \ src/sanity.X90 \ src/elpa1/elpa_cholesky_template.X90 \ src/elpa1/elpa_invert_trm.X90 \ @@ -104,8 +104,8 @@ else endif if WITH_GPU_VERSION - libelpa@SUFFIX@_private_la_SOURCES += src/GPU/cudaFunctions.cu src/GPU/cuUtils.cu src/ev_tridi_band_gpu_c_v2.cu - EXTRA_libelpa@SUFFIX@_private_la_DEPENDENCIES += src/ev_tridi_band_gpu_c_v2_complex_template.Xcu src/ev_tridi_band_gpu_c_v2_real_template.Xcu + libelpa@SUFFIX@_private_la_SOURCES += src/GPU/cudaFunctions.cu src/GPU/cuUtils.cu src/elpa2/GPU/ev_tridi_band_gpu_c_v2.cu + EXTRA_libelpa@SUFFIX@_private_la_DEPENDENCIES += src/elpa2/GPU/ev_tridi_band_gpu_c_v2_complex_template.Xcu src/elpa2/GPU/ev_tridi_band_gpu_c_v2_real_template.Xcu endif if !WITH_MPI @@ -381,7 +381,7 @@ dist_files_DATA = \ test/Fortran/test_cholesky_complex.F90 \ test/Fortran/test_invert_trm_complex.F90 \ test/Fortran/elpa_tests.F90 \ - src/elpa2_print_kernels.F90 + src/elpa2/elpa2_print_kernels.F90 #end needed @@ -634,7 +634,7 @@ elpa_driver_complex@SUFFIX@_LDADD = $(build_lib) elpa_driver_complex@SUFFIX@_FCFLAGS = $(AM_FCFLAGS) @FC_MODOUT@private_modules @FC_MODINC@private_modules EXTRA_elpa_driver_complex@SUFFIX@_DEPENDENCIES = test/Fortran/elpa_print_headers.X90 -elpa2_print_kernels@SUFFIX@_SOURCES = src/elpa2_print_kernels.F90 +elpa2_print_kernels@SUFFIX@_SOURCES = src/elpa2/elpa2_print_kernels.F90 elpa2_print_kernels@SUFFIX@_LDADD = $(build_lib) elpa2_print_kernels@SUFFIX@_FCFLAGS = $(AM_FCFLAGS) @FC_MODOUT@private_modules @FC_MODINC@private_modules @@ -996,9 +996,9 @@ EXTRA_DIST = \ 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/pack_unpack_cpu.X90 \ - src/pack_unpack_gpu.X90 \ + src/elpa2/redist_band.X90 \ + src/elpa2/pack_unpack_cpu.X90 \ + src/elpa2/pack_unpack_gpu.X90 \ src/elpa2/compute_hh_trafo.X90 \ src/sanity.X90 \ src/elpa1/elpa_cholesky_template.X90 \ @@ -1006,8 +1006,8 @@ EXTRA_DIST = \ src/elpa1/elpa_multiply_a_b.X90 \ src/elpa1/elpa_solve_tridi.X90 \ src/elpa2/qr/elpa_qrkernels.X90 \ - src/ev_tridi_band_gpu_c_v2_complex_template.Xcu \ - src/ev_tridi_band_gpu_c_v2_real_template.Xcu \ + src/elpa2/GPU/ev_tridi_band_gpu_c_v2_complex_template.Xcu \ + src/elpa2/GPU/ev_tridi_band_gpu_c_v2_real_template.Xcu \ src/GPU/cuUtils_template.Xcu \ nvcc_wrap \ test_project/Makefile.am \ diff --git a/src/interface_cuda.F90 b/src/GPU/interface_cuda.F90 similarity index 100% rename from src/interface_cuda.F90 rename to src/GPU/interface_cuda.F90 diff --git a/src/ev_tridi_band_gpu_c_v2.cu b/src/elpa2/GPU/ev_tridi_band_gpu_c_v2.cu similarity index 100% rename from src/ev_tridi_band_gpu_c_v2.cu rename to src/elpa2/GPU/ev_tridi_band_gpu_c_v2.cu diff --git a/src/ev_tridi_band_gpu_c_v2_complex_template.Xcu b/src/elpa2/GPU/ev_tridi_band_gpu_c_v2_complex_template.Xcu similarity index 100% rename from src/ev_tridi_band_gpu_c_v2_complex_template.Xcu rename to src/elpa2/GPU/ev_tridi_band_gpu_c_v2_complex_template.Xcu diff --git a/src/ev_tridi_band_gpu_c_v2_real_template.Xcu b/src/elpa2/GPU/ev_tridi_band_gpu_c_v2_real_template.Xcu similarity index 100% rename from src/ev_tridi_band_gpu_c_v2_real_template.Xcu rename to src/elpa2/GPU/ev_tridi_band_gpu_c_v2_real_template.Xcu diff --git a/src/interface_c_kernel.F90 b/src/elpa2/GPU/interface_c_kernel.F90 similarity index 100% rename from src/interface_c_kernel.F90 rename to src/elpa2/GPU/interface_c_kernel.F90 diff --git a/src/elpa2_print_kernels.F90 b/src/elpa2/elpa2_print_kernels.F90 similarity index 100% rename from src/elpa2_print_kernels.F90 rename to src/elpa2/elpa2_print_kernels.F90 diff --git a/src/mod_pack_unpack_cpu.F90 b/src/elpa2/mod_pack_unpack_cpu.F90 similarity index 96% rename from src/mod_pack_unpack_cpu.F90 rename to src/elpa2/mod_pack_unpack_cpu.F90 index 139d706beae2b64b20316c0e88cf88bf204c4d62..acdc9815e92bd0f9cd4db2a829a111097af2d4d3 100644 --- a/src/mod_pack_unpack_cpu.F90 +++ b/src/elpa2/mod_pack_unpack_cpu.F90 @@ -72,7 +72,7 @@ module pack_unpack_cpu !real double precision #define REALCASE 1 #define DOUBLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "pack_unpack_cpu.X90" #undef REALCASE #undef DOUBLE_PRECISION @@ -81,7 +81,7 @@ module pack_unpack_cpu #if defined(WANT_SINGLE_PRECISION_REAL) #define REALCASE 1 #define SINGLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "pack_unpack_cpu.X90" #undef REALCASE #undef SINGLE_PRECISION @@ -90,7 +90,7 @@ module pack_unpack_cpu !complex double precision #define COMPLEXCASE 1 #define DOUBLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "pack_unpack_cpu.X90" #undef COMPLEXCASE #undef DOUBLE_PRECISION @@ -99,7 +99,7 @@ module pack_unpack_cpu #if defined(WANT_SINGLE_PRECISION_COMPLEX) #define COMPLEXCASE 1 #define SINGLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "pack_unpack_cpu.X90" #undef COMPLEXCASE #undef SINGLE_PRECISION diff --git a/src/mod_pack_unpack_gpu.F90 b/src/elpa2/mod_pack_unpack_gpu.F90 similarity index 96% rename from src/mod_pack_unpack_gpu.F90 rename to src/elpa2/mod_pack_unpack_gpu.F90 index a02114797452e7ac7630ea55079ecb97eb91066a..a49b4a90e226810316534b07934a04ecffccf0d8 100644 --- a/src/mod_pack_unpack_gpu.F90 +++ b/src/elpa2/mod_pack_unpack_gpu.F90 @@ -72,7 +72,7 @@ module pack_unpack_gpu !real double precision #define REALCASE 1 #define DOUBLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "pack_unpack_gpu.X90" #undef REALCASE #undef DOUBLE_PRECISION @@ -81,7 +81,7 @@ module pack_unpack_gpu #if defined(WANT_SINGLE_PRECISION_REAL) #define REALCASE 1 #define SINGLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "pack_unpack_gpu.X90" #undef REALCASE #undef SINGLE_PRECISION @@ -90,7 +90,7 @@ module pack_unpack_gpu !complex double precision #define COMPLEXCASE 1 #define DOUBLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "pack_unpack_gpu.X90" #undef COMPLEXCASE #undef DOUBLE_PRECISION @@ -99,7 +99,7 @@ module pack_unpack_gpu #if defined(WANT_SINGLE_PRECISION_COMPLEX) #define COMPLEXCASE 1 #define SINGLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "pack_unpack_gpu.X90" #undef COMPLEXCASE #undef SINGLE_PRECISION diff --git a/src/mod_redist_band.F90 b/src/elpa2/mod_redist_band.F90 similarity index 96% rename from src/mod_redist_band.F90 rename to src/elpa2/mod_redist_band.F90 index bfd6d52ddd5ebea650670b3694d42e2071508e8d..0409612e8146df4d1336eb8e99a04b17680ea51b 100644 --- a/src/mod_redist_band.F90 +++ b/src/elpa2/mod_redist_band.F90 @@ -68,7 +68,7 @@ module redist #define REALCASE 1 #define DOUBLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "redist_band.X90" #undef REALCASE @@ -79,7 +79,7 @@ module redist #define REALCASE 1 #define SINGLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "redist_band.X90" #undef REALCASE @@ -90,7 +90,7 @@ module redist ! double precision #define COMPLEXCASE 1 #define DOUBLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "redist_band.X90" #undef COMPLEXCASE @@ -101,7 +101,7 @@ module redist #define COMPLEXCASE 1 #define SINGLE_PRECISION 1 -#include "precision_macros.h" +#include "../precision_macros.h" #include "redist_band.X90" #undef COMPLEXCASE diff --git a/src/pack_unpack_cpu.X90 b/src/elpa2/pack_unpack_cpu.X90 similarity index 100% rename from src/pack_unpack_cpu.X90 rename to src/elpa2/pack_unpack_cpu.X90 diff --git a/src/pack_unpack_gpu.X90 b/src/elpa2/pack_unpack_gpu.X90 similarity index 100% rename from src/pack_unpack_gpu.X90 rename to src/elpa2/pack_unpack_gpu.X90 diff --git a/src/redist_band.X90 b/src/elpa2/redist_band.X90 similarity index 100% rename from src/redist_band.X90 rename to src/elpa2/redist_band.X90