Skip to content

Remove old cutlass mla #23961

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 4 commits into from
Sep 17, 2025
Merged

Remove old cutlass mla #23961

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
6be6496
Remove old cutlass MLA kernel
MatthewBonanni Aug 29, 2025
fb84464
Remove FORCE_OLD_CUTLASS_MLA option
MatthewBonanni Aug 29, 2025
d091fd2
Address pre-commit
MatthewBonanni Sep 16, 2025
0e2d591
Merge branch 'main' into remove_old_cutlass_mla
MatthewBonanni Sep 17, 2025
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
2 changes: 0 additions & 2 deletions CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -298,7 +298,6 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
"csrc/quantization/fp4/nvfp4_blockwise_moe_kernel.cu"
"csrc/sparse/cutlass/sparse_scaled_mm_entry.cu"
"csrc/cutlass_extensions/common.cpp"
"csrc/attention/mla/cutlass_mla_entry.cu"
"csrc/quantization/fp8/per_token_group_quant.cu")

set_gencode_flags_for_srcs(
Expand Down Expand Up @@ -585,7 +584,6 @@ if(VLLM_GPU_LANG STREQUAL "CUDA")
cuda_archs_loose_intersection(MLA_ARCHS "10.0a" "${CUDA_ARCHS}")
if(${CMAKE_CUDA_COMPILER_VERSION} VERSION_GREATER_EQUAL 12.8 AND MLA_ARCHS)
set(SRCS
"csrc/attention/mla/cutlass_mla_kernels.cu"
"csrc/attention/mla/sm100_cutlass_mla_kernel.cu")
set_gencode_flags_for_srcs(
SRCS "${SRCS}"
Expand Down
38 changes: 0 additions & 38 deletions csrc/attention/mla/cutlass_mla_entry.cu
View file
Open in desktop

This file was deleted.

225 changes: 0 additions & 225 deletions csrc/attention/mla/cutlass_mla_kernels.cu
View file
Open in desktop

This file was deleted.

7 changes: 0 additions & 7 deletions csrc/torch_bindings.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -510,13 +510,6 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
ops.def("cutlass_sparse_compress(Tensor a) -> Tensor[]");
ops.impl("cutlass_sparse_compress", &cutlass_sparse_compress);

// CUTLASS MLA decode
ops.def(
"cutlass_mla_decode(Tensor! out, Tensor q_nope, Tensor q_pe,"
" Tensor kv_c_and_k_pe_cache, Tensor seq_lens,"
" Tensor page_table, float scale) -> ()");
ops.impl("cutlass_mla_decode", torch::kCUDA, &cutlass_mla_decode);

// SM100 CUTLASS MLA decode
ops.def(
"sm100_cutlass_mla_decode(Tensor! out, Tensor! lse, Tensor q_nope,"
Expand Down
9 changes: 0 additions & 9 deletions vllm/_custom_ops.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1823,15 +1823,6 @@ def flash_mla_with_kvcache(
return out, softmax_lse


def cutlass_mla_decode(out: torch.Tensor, q_nope: torch.Tensor,
q_pe: torch.Tensor, kv_c_and_k_pe_cache: torch.Tensor,
seq_lens: torch.Tensor, page_table: torch.Tensor,
scale: float) -> torch.Tensor:
torch.ops._C.cutlass_mla_decode(out, q_nope, q_pe, kv_c_and_k_pe_cache,
seq_lens, page_table, scale)
return out


def sm100_cutlass_mla_decode(out: torch.Tensor, lse: torch.Tensor,
q_nope: torch.Tensor, q_pe: torch.Tensor,
kv_c_and_k_pe_cache: torch.Tensor,
Expand Down
Loading