Skip to content

Update strided copy kernel #72662

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

Merged
merged 1 commit into from
May 13, 2025
Merged

Update strided copy kernel #72662

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
12 changes: 6 additions & 6 deletions paddle/phi/kernels/gpu/strided_copy_kernel.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -221,7 +221,7 @@ bool LaunchStridedCopyCaseOneKernel(
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& output_stride,
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& dims,
int rank,
int numel) {
int64_t numel) {
dim3 grid(1, 1, 1), block(1, 1, 1);
phi::Array<int64_t, 6> cur_dims;
block.x = 512;
Expand Down Expand Up @@ -398,7 +398,7 @@ void LaunchStridedCopyDefaultKernel(
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& output_stride,
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& dims,
int rank,
int numel) {
int64_t numel) {
int64_t block = 512;
int64_t grid = (numel + block - 1) / block;

Expand Down Expand Up @@ -648,7 +648,7 @@ bool LaunchStrided2ContiguousCaseOneKernel(
T* output_data,
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& dims,
int rank,
int numel) {
int64_t numel) {
dim3 grid(1, 1, 1), block(1, 1, 1);
phi::Array<int64_t, 6> cur_dims;
block.x = 512;
Expand Down Expand Up @@ -803,7 +803,7 @@ void LaunchStrided2ContiguousDefaultKernel(
T* output_data,
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& dims,
int rank,
int numel) {
int64_t numel) {
int64_t block = 512;
int64_t grid = (numel + block - 1) / block;

Expand Down Expand Up @@ -1054,7 +1054,7 @@ bool LaunchContiguous2StridedCaseOneKernel(
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& output_stride,
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& dims,
int rank,
int numel) {
int64_t numel) {
dim3 grid(1, 1, 1), block(1, 1, 1);
phi::Array<int64_t, 6> cur_dims;
block.x = 512;
Expand Down Expand Up @@ -1209,7 +1209,7 @@ void LaunchContiguous2StridedDefaultKernel(
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& output_stride,
const phi::Array<int64_t, phi::DDim::kMaxRank + 1>& dims,
int rank,
int numel) {
int64_t numel) {
int64_t block = 512;
int64_t grid = (numel + block - 1) / block;

Expand Down
Loading