Skip to content

Sage Attention supports minimal modifications when compiling MS Visual Studio on Windows. #323

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

Open
mengqin wants to merge 8 commits into
base: main
Choose a base branch
from
+326 −60
Open

Sage Attention supports minimal modifications when compiling MS Visual Studio on Windows. #323

Changes from 1 commit
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
2e0b948
Modify setup.py to make it suitable for Windows builds and
mengqin Dec 7, 2025
2799f20
Modify sageattn3's setup.py and header files to make it compatible with
mengqin Dec 7, 2025
444e8b0
Update README.md
mengqin Dec 8, 2025
8bb81e4
Fixed a runtime crash in Sage Attention 3, added the `/Zc:__cplus_plus`
mengqin Dec 14, 2025
7220213
Merge remote-tracking branch 'upstream/main'
mengqin Dec 14, 2025
8f57f60
Fixed an error caused by a comment.
mengqin Dec 14, 2025
821da7b
Merge remote-tracking branch 'thu-ml/SageAttention'
mengqin Dec 30, 2025
142c02a
Merge branch 'thu-ml:main' into main
mengqin Jan 19, 2026
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
Next Next commit
Modify setup.py to make it suitable for Windows builds and 
multi-architecture builds.
  • Loading branch information
@mengqin
mengqin committed Dec 7, 2025
commit 2e0b94810e4613bab15dab94cbaaa60c4c0619bf
64 changes: 49 additions & 15 deletions setup.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,17 +48,33 @@
SUPPORTED_ARCHS = {"8.0", "8.6", "8.9", "9.0", "10.0", "12.0", "12.1"}

# Compiler flags.
CXX_FLAGS = ["-g", "-O3", "-fopenmp", "-lgomp", "-std=c++17", "-DENABLE_BF16"]
NVCC_FLAGS = [
"-O3",
"-std=c++17",
"-U__CUDA_NO_HALF_OPERATORS__",
"-U__CUDA_NO_HALF_CONVERSIONS__",
"--use_fast_math",
"--threads=8",
"-Xptxas=-v",
"-diag-suppress=174",
]
if os.name == "nt":
CXX_FLAGS = ["/Zi", "/O2", "/openmp", "/std:c++17", "/DENABLE_BF16", "/MD", "/permissive-"]
NVCC_FLAGS = [
"-O3",
"-std=c++17",
"-U__CUDA_NO_HALF_OPERATORS__",
"-U__CUDA_NO_HALF_CONVERSIONS__",
"--use_fast_math",
"--threads=8",
#"-Xptxas=-v",
"-diag-suppress=174",
"-diag-suppress=177",
"-D_WIN32=1",
"-DUSE_CUDA=1",
]
else:
CXX_FLAGS = ["-g", "-O3", "-fopenmp", "-lgomp", "-std=c++17", "-DENABLE_BF16"]
NVCC_FLAGS = [
"-O3",
"-std=c++17",
"-U__CUDA_NO_HALF_OPERATORS__",
"-U__CUDA_NO_HALF_CONVERSIONS__",
"--use_fast_math",
"--threads=8",
"-Xptxas=-v",
"-diag-suppress=174",
]

# Append flags from env if provided
cxx_append = os.getenv("CXX_APPEND_FLAGS", "").strip()
Expand Down Expand Up @@ -140,19 +156,31 @@ def get_nvcc_cuda_version(cuda_dir: str) -> Version:
"CUDA 12.8 or higher is required for compute capability 12.0.")

# Add target compute capabilities to NVCC flags.
NVCC_FLAGS_SM80 = list(NVCC_FLAGS)
NVCC_FLAGS_SM89 = list(NVCC_FLAGS)
NVCC_FLAGS_SM90 = list(NVCC_FLAGS)
for capability in compute_capabilities:
if capability.startswith("8.0"):
HAS_SM80 = True
num = "80"
NVCC_FLAGS_SM80 += ["-gencode", f"arch=compute_{num},code=sm_{num}"]
if capability.endswith("+PTX"):
NVCC_FLAGS_SM80 += ["-gencode", f"arch=compute_{num},code=compute_{num}"]
elif capability.startswith("8.6"):
HAS_SM86 = True
num = "86"
elif capability.startswith("8.9"):
HAS_SM89 = True
num = "89"
NVCC_FLAGS_SM89 += ["-gencode", f"arch=compute_{num},code=sm_{num}"]
if capability.endswith("+PTX"):
NVCC_FLAGS_SM89 += ["-gencode", f"arch=compute_{num},code=compute_{num}"]
elif capability.startswith("9.0"):
HAS_SM90 = True
num = "90a"
NVCC_FLAGS_SM90 += ["-gencode", f"arch=compute_{num},code=sm_{num}"]
if capability.endswith("+PTX"):
NVCC_FLAGS_SM90 += ["-gencode", f"arch=compute_{num},code=compute_{num}"]
elif capability.startswith("10.0"):
HAS_SM100 = True
num = "100a"
Expand All @@ -179,7 +207,7 @@ def get_nvcc_cuda_version(cuda_dir: str) -> Version:
"csrc/qattn/pybind_sm80.cpp",
"csrc/qattn/qk_int_sv_f16_cuda_sm80.cu",
],
extra_compile_args={"cxx": CXX_FLAGS, "nvcc": NVCC_FLAGS},
extra_compile_args={"cxx": CXX_FLAGS, "nvcc": NVCC_FLAGS_SM80},
)
)

Expand All @@ -197,10 +225,16 @@ def get_nvcc_cuda_version(cuda_dir: str) -> Version:
"csrc/qattn/sm89_qk_int8_sv_f8_accum_f32_fuse_v_scale_attn_inst_buf.cu",
"csrc/qattn/sm89_qk_int8_sv_f8_accum_f16_fuse_v_scale_attn_inst_buf.cu",
],
extra_compile_args={"cxx": CXX_FLAGS, "nvcc": NVCC_FLAGS},
extra_compile_args={"cxx": CXX_FLAGS, "nvcc": NVCC_FLAGS_SM89},
)
)

cuda_lib = []
if os.name == "nt":
cuda_lib = ["cuda.lib"]
else:
cuda_lib = ["-lcuda"]

if HAS_SM90:
ext_modules.append(
CUDAExtension(
Expand All @@ -209,8 +243,8 @@ def get_nvcc_cuda_version(cuda_dir: str) -> Version:
"csrc/qattn/pybind_sm90.cpp",
"csrc/qattn/qk_int_sv_f8_cuda_sm90.cu",
],
extra_compile_args={"cxx": CXX_FLAGS, "nvcc": NVCC_FLAGS},
extra_link_args=['-lcuda'],
extra_compile_args={"cxx": CXX_FLAGS, "nvcc": NVCC_FLAGS_SM90},
extra_link_args=cuda_lib,
)
)

Expand Down