minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

[MUSA] Fix sglang-kernel build (#31634)

Browse Source 
Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>
This commit is contained in:
R0CKSTAR
2026-07-17 20:51:09 -07:00
committed by GitHub
parent 359009fa00
commit 87dc211b87
7 changed files with 4 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/nightly-test-musa.yml
-1
View File
@@ -74,7 +74,6 @@ jobs:
if: steps.gate.outputs.run_job == 'true' if: steps.gate.outputs.run_job == 'true'
timeout-minutes: 30 timeout-minutes: 30
run: | run: |
pytest sgl-kernel/tests/test_dsv3_router_gemm.py
pytest sgl-kernel/tests/test_per_token_quant_fp8.py pytest sgl-kernel/tests/test_per_token_quant_fp8.py
pytest sgl-kernel/tests/speculative/test_eagle_utils.py pytest sgl-kernel/tests/speculative/test_eagle_utils.py
pytest sgl-kernel/tests/speculative/test_ngram_utils.py pytest sgl-kernel/tests/speculative/test_ngram_utils.py
.github/workflows/pr-test-musa.yml
-1
View File
@@ -217,7 +217,6 @@ jobs:
- name: Run sgl-kernel test - name: Run sgl-kernel test
timeout-minutes: 20 timeout-minutes: 20
run: | run: |
pytest sgl-kernel/tests/test_dsv3_router_gemm.py
pytest sgl-kernel/tests/test_per_token_quant_fp8.py pytest sgl-kernel/tests/test_per_token_quant_fp8.py
pytest sgl-kernel/tests/speculative/test_eagle_utils.py pytest sgl-kernel/tests/speculative/test_eagle_utils.py
pytest sgl-kernel/tests/speculative/test_ngram_utils.py pytest sgl-kernel/tests/speculative/test_ngram_utils.py
3rdparty/amd/wheel/sglang/pyproject.toml
Vendored
+1 -1
View File
@@ -124,7 +124,7 @@ srt_musa = [
"sglang[runtime_common]", "sglang[runtime_common]",
"torch", "torch",
"torch_musa", "torch_musa",
"torchada>=0.1.68", "torchada>=0.1.74",
"mthreads-ml-py", "mthreads-ml-py",
"mate>=0.2.0", "mate>=0.2.0",
"deep-gemm>=0.1.3", "deep-gemm>=0.1.3",
python/pyproject_other.toml
+1 -1
View File
@@ -126,7 +126,7 @@ srt_musa = [
"sglang[runtime_common]", "sglang[runtime_common]",
"torch", "torch",
"torch_musa", "torch_musa",
"torchada>=0.1.68", "torchada>=0.1.74",
] ]
diffusion_musa = [ diffusion_musa = [
sgl-kernel/csrc/common_extension_musa.cc
-3
View File
@@ -100,9 +100,6 @@ TORCH_LIBRARY_EXPAND(sgl_kernel, m) {
m.def("dsv3_fused_a_gemm(Tensor! output, Tensor mat_a, Tensor mat_b) -> ()"); m.def("dsv3_fused_a_gemm(Tensor! output, Tensor mat_a, Tensor mat_b) -> ()");
m.impl("dsv3_fused_a_gemm", torch::kMUSA, &dsv3_fused_a_gemm); m.impl("dsv3_fused_a_gemm", torch::kMUSA, &dsv3_fused_a_gemm);
m.def("dsv3_router_gemm(Tensor! output, Tensor mat_a, Tensor mat_b) -> ()");
m.impl("dsv3_router_gemm", torch::kMUSA, &dsv3_router_gemm);
/* /*
* From csrc/moe * From csrc/moe
*/ */
sgl-kernel/csrc/musa/top_k_top_p_sampling.mu
+1 -1
View File
@@ -21,7 +21,7 @@
#include <torch/all.h> #include <torch/all.h>
#include "torch_musa/csrc/aten/musa/UnpackRaw.muh" #include "torch_musa/csrc/aten/musa/UnpackRaw.muh"
#include <flashinfer/sampling.muh> #include <flashinfer/sampling.cuh>
#include <mutex> #include <mutex>
#include "musa.h" #include "musa.h"
sgl-kernel/pyproject_musa.toml
+1 -1
View File
@@ -3,7 +3,7 @@ requires = [
"setuptools>=75.0", "setuptools>=75.0",
"scikit-build-core>=0.10", "scikit-build-core>=0.10",
"torch", "torch",
"torchada>=0.1.68", "torchada>=0.1.74",
"wheel", "wheel",
] ]
build-backend = "setuptools.build_meta" build-backend = "setuptools.build_meta"