Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
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
8 changes: 8 additions & 0 deletions common/defines.h
Original file line number Diff line number Diff line change
Expand Up @@ -166,4 +166,12 @@ enum {C=0,N=1,O=2,H=3,XX=4,P=5,S=6}; // see "bond_index" in the "AD4.1_bound.da
#define SYCL_ATOMICS_MEM_SCOPE sycl::memory_scope::device
#endif

#define SYCL_INTEL_KERNEL_ARGS_RESTRICT

#ifdef SYCL_INTEL_KERNEL_ARGS_RESTRICT
#define SYCL_KERNEL_ARGS_RESTRICT [[intel::kernel_args_restrict]]
#else
#define SYCL_KERNEL_ARGS_RESTRICT
#endif

#endif /* DEFINES_H_ */
3 changes: 2 additions & 1 deletion dpcpp/kernel1.dp.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -88,7 +88,8 @@ void gpu_calc_initpop(
sycl::nd_range<3>(sycl::range<3>(1, 1, blocks) *
sycl::range<3>(1, 1, threadsPerBlock),
sycl::range<3>(1, 1, threadsPerBlock)),
[=](sycl::nd_item<3> item_ct1) {
[=](sycl::nd_item<3> item_ct1)
SYCL_KERNEL_ARGS_RESTRICT {
gpu_calc_initpop_kernel(
pConformations_current, pEnergies_current,
item_ct1, *cData_ptr_ct1,
Expand Down
3 changes: 2 additions & 1 deletion dpcpp/kernel2.dp.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -79,7 +79,8 @@ void gpu_sum_evals(uint32_t blocks, uint32_t threadsPerBlock)
sycl::nd_range<3>(sycl::range<3>(1, 1, blocks) *
sycl::range<3>(1, 1, threadsPerBlock),
sycl::range<3>(1, 1, threadsPerBlock)),
[=](sycl::nd_item<3> item_ct1) {
[=](sycl::nd_item<3> item_ct1)
SYCL_KERNEL_ARGS_RESTRICT {
gpu_sum_evals_kernel(
item_ct1, *cData_ptr_ct1,
sSum_evals_acc_ct1.get_pointer());
Expand Down
3 changes: 2 additions & 1 deletion dpcpp/kernel3.dp.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -418,7 +418,8 @@ void gpu_perform_LS(
sycl::nd_range<3>(sycl::range<3>(1, 1, blocks) *
sycl::range<3>(1, 1, threads),
sycl::range<3>(1, 1, threads)),
[=](sycl::nd_item<3> item_ct1) {
[=](sycl::nd_item<3> item_ct1)
SYCL_KERNEL_ARGS_RESTRICT {
gpu_perform_LS_kernel(
pMem_conformations_next, pMem_energies_next,
item_ct1, dpct_local_acc_ct1.get_pointer(),
Expand Down
5 changes: 3 additions & 2 deletions dpcpp/kernel4.dp.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -465,8 +465,9 @@ void gpu_gen_and_eval_newpops(
sycl::nd_range<3>(sycl::range<3>(1, 1, blocks) *
sycl::range<3>(1, 1, threadsPerBlock),
sycl::range<3>(1, 1, threadsPerBlock)),
[=](sycl::nd_item<3> item_ct1)
[[intel::reqd_sub_group_size(32)]] {
[=](sycl::nd_item<3> item_ct1)
[[intel::reqd_sub_group_size(32)]]
SYCL_KERNEL_ARGS_RESTRICT {
gpu_gen_and_eval_newpops_kernel(
pMem_conformations_current,
pMem_energies_current, pMem_conformations_next,
Expand Down
3 changes: 2 additions & 1 deletion dpcpp/kernel_ad.dp.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -492,7 +492,8 @@ void gpu_gradient_minAD(
sycl::nd_range<3>(sycl::range<3>(1, 1, blocks) *
sycl::range<3>(1, 1, threads),
sycl::range<3>(1, 1, threads)),
[=](sycl::nd_item<3> item_ct1) {
[=](sycl::nd_item<3> item_ct1)
SYCL_KERNEL_ARGS_RESTRICT {
gpu_gradient_minAD_kernel(
pMem_conformations_next, pMem_energies_next,
item_ct1, dpct_local_acc_ct1.get_pointer(),
Expand Down
3 changes: 2 additions & 1 deletion dpcpp/kernel_adam.dp.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -486,7 +486,8 @@ void gpu_gradient_minAdam(
sycl::nd_range<3>(sycl::range<3>(1, 1, blocks) *
sycl::range<3>(1, 1, threads),
sycl::range<3>(1, 1, threads)),
[=](sycl::nd_item<3> item_ct1) {
[=](sycl::nd_item<3> item_ct1)
SYCL_KERNEL_ARGS_RESTRICT {
gpu_gradient_minAdam_kernel(
pMem_conformations_next, pMem_energies_next,
item_ct1, dpct_local_acc_ct1.get_pointer(),
Expand Down