scnc
/
quantum-espresso
mirror of https://gitlab.com/QEF/q-e.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Remove unnecessary mydger_c and add mydger2

This commit is contained in:
Giacomo Rossi 2023-04-11 05:16:27 -07:00
parent 8fc1ea73a8
commit 8e6c4d45c4
1 changed files with 55 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

80
UtilXlib/device_helper.f90
View File

@ -15,6 +15,12 @@ SUBROUTINE MYDGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
#if defined(__CUDA) #if defined(__CUDA)
use cudafor use cudafor
use cublas use cublas
#elif defined(__OPENMP_GPU)
#if defined(__ONEMKL)
use onemkl_blas_gpu
#elif defined(__ROCBLAS)
use rocblas_utils
#endif
#endif #endif
! .. Scalar Arguments .. ! .. Scalar Arguments ..
DOUBLE PRECISION :: ALPHA DOUBLE PRECISION :: ALPHA
@ -23,37 +29,20 @@ SUBROUTINE MYDGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
DOUBLE PRECISION :: A( LDA, * ), X( * ), Y( * ) DOUBLE PRECISION :: A( LDA, * ), X( * ), Y( * )
#if defined(__CUDA) #if defined(__CUDA)
attributes(device) :: A, X, Y attributes(device) :: A, X, Y
#endif
CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA ) CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
END SUBROUTINE MYDGER
SUBROUTINE MYDGER_C ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
#if defined(__CUDA)
use cudafor
use cublas
#elif defined(__OPENMP_GPU) #elif defined(__OPENMP_GPU)
#if defined(__ONEMKL)
use onemkl_blas_no_array_check_gpu
#endif
#endif
! .. Scalar Arguments ..
DOUBLE PRECISION :: ALPHA
INTEGER :: INCX, INCY, LDA, M, N
! .. Array Arguments ..
COMPLEX*16 :: A( LDA, * ), X( * ), Y( * )
#if defined(__CUDA)
attributes(device) :: A, X, Y
#endif
#if defined(__ONEMKL) #if defined(__ONEMKL)
!$omp target variant dispatch use_device_ptr(A, X, Y) !$omp target variant dispatch use_device_ptr(A, X, Y)
#endif
CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA ) CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
#if defined(__ONEMKL)
!$omp end target variant dispatch !$omp end target variant dispatch
#elif defined(__ROCBLAS)
CALL rocblas_dger( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
#endif
#else
CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
#endif #endif
END SUBROUTINE MYDGER_C END SUBROUTINE MYDGER
!=----------------------------------------------------------------------------=! !=----------------------------------------------------------------------------=!
@ -65,8 +54,7 @@ SUBROUTINE MYDGEMM( TRANSA, TRANSB, M, N, K, ALPHA, A, LDA, B, LDB, BETA, C, LDC
#elif defined(__OPENMP_GPU) #elif defined(__OPENMP_GPU)
#if defined(__ONEMKL) #if defined(__ONEMKL)
use onemkl_blas_gpu use onemkl_blas_gpu
#endif #elif defined(__ROCBLAS)
#if defined(__ROCBLAS)
use rocblas_utils use rocblas_utils
#endif #endif
#endif #endif
@ -131,6 +119,48 @@ END SUBROUTINE MYZGEMM
! MYZGEMM, but with an additional variable (OMP_OFFLOAD) to decide wether to perform a cpu ! MYZGEMM, but with an additional variable (OMP_OFFLOAD) to decide wether to perform a cpu
! _gemm or call a rocblas _gemm which takes gpu_only arguments. ! _gemm or call a rocblas _gemm which takes gpu_only arguments.
! !
SUBROUTINE MYDGER2 ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA, OMP_OFFLOAD )
#if defined(__CUDA)
use cudafor
use cublas
#elif defined(__OPENMP_GPU)
#if defined(__ONEMKL)
use onemkl_blas_gpu
#elif defined(__ROCBLAS)
use rocblas_utils
#endif
#endif
! .. Scalar Arguments ..
DOUBLE PRECISION :: ALPHA
INTEGER :: INCX, INCY, LDA, M, N
! .. Array Arguments ..
DOUBLE PRECISION :: A( LDA, * ), X( * ), Y( * )
LOGICAL, INTENT(IN) :: OMP_OFFLOAD
#if defined(__CUDA)
attributes(device) :: A, X, Y
CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
#elif defined(__OPENMP_GPU)
#if defined(__ONEMKL)
IF (OMP_OFFLOAD) THEN
!$omp target variant dispatch use_device_ptr(A, X, Y)
CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
!$omp end target variant dispatch
ELSE
CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
ENDIF
#elif defined(__ROCBLAS)
IF (OMP_OFFLOAD) THEN
CALL rocblas_dger( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
ELSE
CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
ENDIF
#endif
#else
CALL DGER ( M, N, ALPHA, X, INCX, Y, INCY, A, LDA )
#endif
END SUBROUTINE MYDGER2
SUBROUTINE MYDGEMM2( TRANSA, TRANSB, M, N, K, ALPHA, A, LDA, B, LDB, BETA, C, LDC, OMP_OFFLOAD ) SUBROUTINE MYDGEMM2( TRANSA, TRANSB, M, N, K, ALPHA, A, LDA, B, LDB, BETA, C, LDC, OMP_OFFLOAD )
#if defined(__CUDA) #if defined(__CUDA)
use cudafor use cudafor