Build Error #6
Description
For some reason I was stuck with DLL load failed while importing flash_attn_wmma even after installing python310 then I'm trying to build it by myself, but got this error
Successfully preprocessed all matching files.
Traceback (most recent call last):
File "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\utils\cpp_extension.py", line 2506, in _run_ninja_build
subprocess.run(
File "C:\Python310\lib\subprocess.py", line 526, in run
raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\flash-attention-v2-RDNA3-minimal\bench_with_sdpa.py", line 60, in <module>
from rocwmma_fattn.FlashAttn import FlashAttentionFunction
File "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\flash-attention-v2-RDNA3-minimal\rocwmma_fattn\FlashAttn.py", line 23, in <module>
flash_attn_wmma = torch.utils.cpp_extension.load(
File "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\utils\cpp_extension.py", line 1623, in load
return _jit_compile(
File "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\utils\cpp_extension.py", line 2076, in _jit_compile
_write_ninja_file_and_build_library(
File "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\utils\cpp_extension.py", line 2222, in _write_ninja_file_and_build_library
_run_ninja_build(
File "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\utils\cpp_extension.py", line 2522, in _run_ninja_build
raise RuntimeError(message) from e
RuntimeError: Error building extension 'flash_attn_wmma': [1/3] C:\ROCm\5.7\bin\hipcc.bin.exe -DWITH_HIP -Wno-ignored-attributes -DTORCH_EXTENSION_NAME=flash_attn_wmma -DTORCH_API_INCLUDE_EXTENSION_H -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\torch\csrc\api\include -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\THH -IC:\ROCm\5.7\include -IC:\Python310\Include -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++17 -fPIC -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -Ofast -save-temps -DROCWMMA_ARCH_GFX1100=1 -DROCWMMA_ARCH_GFX1101=1 -DROCWMMA_ARCH_GFX1102=1 -DROCWMMA_ARCH_GFX1103=1 -DROCWMMA_ARCH_GFX11=1 -DROCWMMA_WAVE32_MODE=1 -DROCWMMA_BLOCK_DIM_16_SUPPORTED=1 -mcumode -ffast-math -fgpu-flush-denormals-to-zero --offload-arch=gfx1100 -fno-gpu-rdc -c C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\flash-attention-v2-RDNA3-minimal\rocwmma_fattn\kernel_fp16.hip -o kernel_fp16.cuda.o
FAILED: kernel_fp16.cuda.o
C:\ROCm\5.7\bin\hipcc.bin.exe -DWITH_HIP -Wno-ignored-attributes -DTORCH_EXTENSION_NAME=flash_attn_wmma -DTORCH_API_INCLUDE_EXTENSION_H -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\torch\csrc\api\include -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\THH -IC:\ROCm\5.7\include -IC:\Python310\Include -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++17 -fPIC -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -Ofast -save-temps -DROCWMMA_ARCH_GFX1100=1 -DROCWMMA_ARCH_GFX1101=1 -DROCWMMA_ARCH_GFX1102=1 -DROCWMMA_ARCH_GFX1103=1 -DROCWMMA_ARCH_GFX11=1 -DROCWMMA_WAVE32_MODE=1 -DROCWMMA_BLOCK_DIM_16_SUPPORTED=1 -mcumode -ffast-math -fgpu-flush-denormals-to-zero --offload-arch=gfx1100 -fno-gpu-rdc -c C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\flash-attention-v2-RDNA3-minimal\rocwmma_fattn\kernel_fp16.hip -o kernel_fp16.cuda.o
clang++: error: unsupported option '-fPIC' for target 'x86_64-pc-windows-msvc'
clang++: error: unsupported option '-fPIC' for target 'x86_64-pc-windows-msvc'
clang++: error: unsupported option '-fPIC' for target 'x86_64-pc-windows-msvc'
clang++: error: unsupported option '-fPIC' for target 'x86_64-pc-windows-msvc'
failed to execute:C:\ROCm\5.7\bin\clang++ -isystem "C:\ROCm\5.7\include" --offload-arch=gfx1100 -mllvm -amdgpu-early-inline-all=true -mllvm -amdgpu-function-calls=false "-DWITH_HIP" "-Wno-ignored-attributes" "-DTORCH_EXTENSION_NAME=flash_attn_wmma" "-DTORCH_API_INCLUDE_EXTENSION_H" "-IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include" "-IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\torch\csrc\api\include" "-IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\THH" "-IC:\ROCm\5.7\include" "-IC:\Python310\Include" "-D_GLIBCXX_USE_CXX11_ABI=0" "-std=c++17" "-fPIC" "-D__HIP_PLATFORM_AMD__=1" "-DUSE_ROCM=1" "-DHIPBLAS_V2" "-DCUDA_HAS_FP16=1" "-D__HIP_NO_HALF_OPERATORS__=1" "-D__HIP_NO_HALF_CONVERSIONS__=1" "-Ofast" "-save-temps" "-DROCWMMA_ARCH_GFX1100=1" "-DROCWMMA_ARCH_GFX1101=1" "-DROCWMMA_ARCH_GFX1102=1" "-DROCWMMA_ARCH_GFX1103=1" "-DROCWMMA_ARCH_GFX11=1" "-DROCWMMA_WAVE32_MODE=1" "-DROCWMMA_BLOCK_DIM_16_SUPPORTED=1" "-mcumode" "-ffast-math" "-fgpu-flush-denormals-to-zero" "-fno-gpu-rdc" "-c" -x hip "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\flash-attention-v2-RDNA3-minimal\rocwmma_fattn\kernel_fp16.hip" "-o" "kernel_fp16.cuda.o"
[2/3] C:\ROCm\5.7\bin\hipcc.bin.exe -DWITH_HIP -Wno-ignored-attributes -DTORCH_EXTENSION_NAME=flash_attn_wmma -DTORCH_API_INCLUDE_EXTENSION_H -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\torch\csrc\api\include -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\THH -IC:\ROCm\5.7\include -IC:\Python310\Include -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++17 -fPIC -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -Ofast -save-temps -DROCWMMA_ARCH_GFX1100=1 -DROCWMMA_ARCH_GFX1101=1 -DROCWMMA_ARCH_GFX1102=1 -DROCWMMA_ARCH_GFX1103=1 -DROCWMMA_ARCH_GFX11=1 -DROCWMMA_WAVE32_MODE=1 -DROCWMMA_BLOCK_DIM_16_SUPPORTED=1 -mcumode -ffast-math -fgpu-flush-denormals-to-zero --offload-arch=gfx1100 -fno-gpu-rdc -c C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\flash-attention-v2-RDNA3-minimal\rocwmma_fattn\kernel_bf16.hip -o kernel_bf16.cuda.o
FAILED: kernel_bf16.cuda.o
C:\ROCm\5.7\bin\hipcc.bin.exe -DWITH_HIP -Wno-ignored-attributes -DTORCH_EXTENSION_NAME=flash_attn_wmma -DTORCH_API_INCLUDE_EXTENSION_H -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\torch\csrc\api\include -IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\THH -IC:\ROCm\5.7\include -IC:\Python310\Include -D_GLIBCXX_USE_CXX11_ABI=0 -std=c++17 -fPIC -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -Ofast -save-temps -DROCWMMA_ARCH_GFX1100=1 -DROCWMMA_ARCH_GFX1101=1 -DROCWMMA_ARCH_GFX1102=1 -DROCWMMA_ARCH_GFX1103=1 -DROCWMMA_ARCH_GFX11=1 -DROCWMMA_WAVE32_MODE=1 -DROCWMMA_BLOCK_DIM_16_SUPPORTED=1 -mcumode -ffast-math -fgpu-flush-denormals-to-zero --offload-arch=gfx1100 -fno-gpu-rdc -c C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\flash-attention-v2-RDNA3-minimal\rocwmma_fattn\kernel_bf16.hip -o kernel_bf16.cuda.o
clang++: error: unsupported option '-fPIC' for target 'x86_64-pc-windows-msvc'
clang++: error: unsupported option '-fPIC' for target 'x86_64-pc-windows-msvc'
clang++: error: unsupported option '-fPIC' for target 'x86_64-pc-windows-msvc'
clang++: error: unsupported option '-fPIC' for target 'x86_64-pc-windows-msvc'
failed to execute:C:\ROCm\5.7\bin\clang++ -isystem "C:\ROCm\5.7\include" --offload-arch=gfx1100 -mllvm -amdgpu-early-inline-all=true -mllvm -amdgpu-function-calls=false "-DWITH_HIP" "-Wno-ignored-attributes" "-DTORCH_EXTENSION_NAME=flash_attn_wmma" "-DTORCH_API_INCLUDE_EXTENSION_H" "-IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include" "-IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\torch\csrc\api\include" "-IC:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\venv\lib\site-packages\torch\include\THH" "-IC:\ROCm\5.7\include" "-IC:\Python310\Include" "-D_GLIBCXX_USE_CXX11_ABI=0" "-std=c++17" "-fPIC" "-D__HIP_PLATFORM_AMD__=1" "-DUSE_ROCM=1" "-DHIPBLAS_V2" "-DCUDA_HAS_FP16=1" "-D__HIP_NO_HALF_OPERATORS__=1" "-D__HIP_NO_HALF_CONVERSIONS__=1" "-Ofast" "-save-temps" "-DROCWMMA_ARCH_GFX1100=1" "-DROCWMMA_ARCH_GFX1101=1" "-DROCWMMA_ARCH_GFX1102=1" "-DROCWMMA_ARCH_GFX1103=1" "-DROCWMMA_ARCH_GFX11=1" "-DROCWMMA_WAVE32_MODE=1" "-DROCWMMA_BLOCK_DIM_16_SUPPORTED=1" "-mcumode" "-ffast-math" "-fgpu-flush-denormals-to-zero" "-fno-gpu-rdc" "-c" -x hip "C:\Users\Marcus\Documents\Ferramentas\ComfyUI-Zluda\flash-attention-v2-RDNA3-minimal\rocwmma_fattn\kernel_bf16.hip" "-o" "kernel_bf16.cuda.o"
ninja: build stopped: subcommand failed.