Skip to content

[AUTOGENERATED] [release/2.9] [ROCm] Adjust grid size for non-unit stride backwards indexing #2724

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
Oct 17, 2025
Merged

[AUTOGENERATED] [release/2.9] [ROCm] Adjust grid size for non-unit stride backwards indexing #2724

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
5 changes: 4 additions & 1 deletion aten/src/ATen/native/cuda/Indexing.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -710,6 +710,9 @@ void index_put_with_sort_kernel(Tensor & self, const c10::List<std::optional<Ten
dim3 block(warp_size, indices_per_block);

#ifdef USE_ROCM
dim3 new_grid_many_indices(ceil_div(num_indices, (int64_t) (indices_per_block * warp_size)),
grid.y == 1 ? std::min<int>(at::cuda::getCurrentDeviceProperties()->maxGridSize[1], ceil_div(sliceSize, (int64_t) (warp_size))) : grid.y,
grid.z);
dim3 new_grid(ceil_div(num_indices, (int64_t) (indices_per_block * warp_size)), grid.y, grid.z);
size_t smem_dups_size = indices_per_block * warp_size * sizeof(int64_t);
#define KERNEL_GRID new_grid
Expand Down Expand Up @@ -788,7 +791,7 @@ void index_put_with_sort_kernel(Tensor & self, const c10::List<std::optional<Ten
expandedValue.scalar_type(),
"indexing_backward_many_indices",
AT_WRAP([&] {
indexing_backward_kernel_many_indices<scalar_t, UNROLL><<<new_grid, block, smem_dups_size, stream>>>(
indexing_backward_kernel_many_indices<scalar_t, UNROLL><<<new_grid_many_indices, block, smem_dups_size, stream>>>(
sorted_indices.const_data_ptr<int64_t>(),
orig_indices.const_data_ptr<int64_t>(),
expandedValue.const_data_ptr<scalar_t>(),
Expand Down