Skip to content

Align CUDA sources with C++17 #223

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
JensWehner wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Align CUDA sources with C++17 #223

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 5 additions & 5 deletions attic/src/integrator/cuda/cuda_eval_denvars.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@ __global__ void eval_uvars_lda_kernel( size_t ntasks,
const int tid_x = blockIdx.x * blockDim.x + threadIdx.x;
const int tid_y = blockIdx.y * blockDim.y + threadIdx.y;

register double den_reg = 0.;
double den_reg = 0.;

if( tid_x < nbf and tid_y < npts ) {

Expand Down Expand Up @@ -115,10 +115,10 @@ __global__ void eval_uvars_gga_kernel( size_t ntasks,

for (int sm_y = threadIdx.y; sm_y < GGA_KERNEL_SM_BLOCK_Y; sm_y += blockDim.y) {
const int tid_y = bid_y + sm_y;
register double den_reg = den_shared[0][sm_y][threadIdx.x];
register double dx_reg = den_shared[1][sm_y][threadIdx.x];
register double dy_reg = den_shared[2][sm_y][threadIdx.x];
register double dz_reg = den_shared[3][sm_y][threadIdx.x];
double den_reg = den_shared[0][sm_y][threadIdx.x];
double dx_reg = den_shared[1][sm_y][threadIdx.x];
double dy_reg = den_shared[2][sm_y][threadIdx.x];
double dz_reg = den_shared[3][sm_y][threadIdx.x];

// Warp blocks are stored col major
den_reg = 2 * warpReduceSum( den_reg );
Expand Down
2 changes: 1 addition & 1 deletion attic/src/integrator/cuda/cuda_weights.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,7 +30,7 @@ __global__ void compute_point_center_dist(
) {

__shared__ double3 point_buffer[warp_size];
register double3 coord_reg;
double3 coord_reg;

const int natoms_block = (natoms + warp_size-1) / warp_size;
const int coords_block = (npts + warp_size-1) / warp_size;
Expand Down
2 changes: 1 addition & 1 deletion attic/src/integrator/cuda/gauxc-cuda_integrator.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,7 +32,7 @@ target_sources( gauxc PRIVATE cuda/collocation_device.cu
cuda/cuda_device_properties.cxx
)

target_compile_features( gauxc PRIVATE cuda_std_14 )
target_compile_features( gauxc PRIVATE cuda_std_17 )
#target_include_directories( gauxc
# PRIVATE
# $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/src/integrator/cuda>
Expand Down
10 changes: 5 additions & 5 deletions attic/src/new_integrator/device/cuda/cuda_eval_denvars.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@ __global__ void eval_uvars_lda_kernel( size_t ntasks,
const int tid_x = blockIdx.x * blockDim.x + threadIdx.x;
const int tid_y = blockIdx.y * blockDim.y + threadIdx.y;

register double den_reg = 0.;
double den_reg = 0.;

if( tid_x < nbf and tid_y < npts ) {

Expand Down Expand Up @@ -115,10 +115,10 @@ __global__ void eval_uvars_gga_kernel( size_t ntasks,

for (int sm_y = threadIdx.y; sm_y < GGA_KERNEL_SM_BLOCK_Y; sm_y += blockDim.y) {
const int tid_y = bid_y + sm_y;
register double den_reg = den_shared[0][sm_y][threadIdx.x];
register double dx_reg = den_shared[1][sm_y][threadIdx.x];
register double dy_reg = den_shared[2][sm_y][threadIdx.x];
register double dz_reg = den_shared[3][sm_y][threadIdx.x];
double den_reg = den_shared[0][sm_y][threadIdx.x];
double dx_reg = den_shared[1][sm_y][threadIdx.x];
double dy_reg = den_shared[2][sm_y][threadIdx.x];
double dz_reg = den_shared[3][sm_y][threadIdx.x];

// Warp blocks are stored col major
den_reg = 2 * warpReduceSum( den_reg );
Expand Down
2 changes: 1 addition & 1 deletion attic/src/new_integrator/device/cuda/cuda_weights.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,7 +30,7 @@ __global__ void compute_point_center_dist(
) {

__shared__ double3 point_buffer[warp_size];
register double3 coord_reg;
double3 coord_reg;

const int natoms_block = (natoms + warp_size-1) / warp_size;
const int coords_block = (npts + warp_size-1) / warp_size;
Expand Down
2 changes: 1 addition & 1 deletion attic/src/new_integrator/device/cuda/gauxc-cuda.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,7 +22,7 @@ target_sources( gauxc PRIVATE

)

target_compile_features( gauxc PRIVATE cuda_std_14 )
target_compile_features( gauxc PRIVATE cuda_std_17 )
target_compile_options( gauxc
PRIVATE
$<$<COMPILE_LANGUAGE:CUDA>: -Xcudafe --diag_suppress=partial_override -Xptxas -v >
Expand Down
2 changes: 1 addition & 1 deletion src/runtime_environment/device/cuda/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,7 @@ if( NOT TARGET CUDA::cublas )
find_package( CUDAToolkit REQUIRED )
endif()

target_compile_features( gauxc PRIVATE cuda_std_14 )
target_compile_features( gauxc PRIVATE cuda_std_17 )
target_sources( gauxc PRIVATE cuda_backend.cxx )

if(NOT GAUXC_LINK_CUDA_STATIC)
Expand Down
2 changes: 1 addition & 1 deletion src/xc_integrator/local_work_driver/device/cuda/kernels/grid_to_center.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -28,7 +28,7 @@ __global__ void compute_grid_to_center_dist(
) {

__shared__ double3 point_buffer[cuda::warp_size];
register double3 coord_reg;
double3 coord_reg;

const int natoms_block = (natoms + cuda::warp_size-1) / cuda::warp_size;
const int coords_block = (npts + cuda::warp_size-1) / cuda::warp_size;
Expand Down
8 changes: 4 additions & 4 deletions src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_gga.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -131,10 +131,10 @@ __global__ void eval_vvar_gga_kern( size_t ntasks,

for (int sm_y = threadIdx.y; sm_y < VVAR_KERNEL_SM_BLOCK; sm_y += blockDim.y) {
const int tid_y = bid_y + sm_y;
register double den_reg = den_shared[0][sm_y][threadIdx.x];
register double dx_reg = den_shared[1][sm_y][threadIdx.x];
register double dy_reg = den_shared[2][sm_y][threadIdx.x];
register double dz_reg = den_shared[3][sm_y][threadIdx.x];
double den_reg = den_shared[0][sm_y][threadIdx.x];
double dx_reg = den_shared[1][sm_y][threadIdx.x];
double dy_reg = den_shared[2][sm_y][threadIdx.x];
double dz_reg = den_shared[3][sm_y][threadIdx.x];

// Warp blocks are stored col major
den_reg = cuda::warp_reduce_sum<warp_size>( den_reg );
Expand Down
2 changes: 1 addition & 1 deletion src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_lda.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,7 @@ __global__ void eval_vvar_lda_kern( size_t ntasks,
const int tid_x = blockIdx.x * blockDim.x + threadIdx.x;
const int tid_y = blockIdx.y * blockDim.y + threadIdx.y;

register double den_reg = 0.;
double den_reg = 0.;

if( tid_x < nbf and tid_y < npts ) {

Expand Down
10 changes: 5 additions & 5 deletions src/xc_integrator/local_work_driver/device/cuda/kernels/uvvars_mgga.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -128,16 +128,16 @@ __global__ void eval_vvar_mgga_kern( size_t ntasks,
for (int sm_y = threadIdx.y; sm_y < MGGA_KERNEL_SM_BLOCK; sm_y += blockDim.y) {
const int tid_y = bid_y + sm_y;

register double tx_reg = den_shared[0][sm_y][threadIdx.x];
register double ty_reg = den_shared[1][sm_y][threadIdx.x];
register double tz_reg = den_shared[2][sm_y][threadIdx.x];
double tx_reg = den_shared[0][sm_y][threadIdx.x];
double ty_reg = den_shared[1][sm_y][threadIdx.x];
double tz_reg = den_shared[2][sm_y][threadIdx.x];
// Warp blocks are stored col major
register double tau_reg = 0.0;
double tau_reg = 0.0;
tau_reg = 0.5 * cuda::warp_reduce_sum<warp_size>( tx_reg );
tau_reg += 0.5 * cuda::warp_reduce_sum<warp_size>( ty_reg );
tau_reg += 0.5 * cuda::warp_reduce_sum<warp_size>( tz_reg );

register double lapl_reg = 0.0;
double lapl_reg = 0.0;
if constexpr (need_lapl) {
lapl_reg = den_shared[3][sm_y][threadIdx.x];
lapl_reg = cuda::warp_reduce_sum<warp_size>(lapl_reg);
Expand Down
Loading