API tools faq
paste
Guest User

Flash attention 7900xtx error

a guest
Apr 27th, 2025
52
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 62.72 KB | None | 0 0
raw download clone embed print report
  1. [50/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_casual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_casual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  2. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_casual_gfx110x_hip.o
  3. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_casual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_casual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  4. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_casual_gfx110x_hip.hip:25:
  5. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  6. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  7. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  8. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  9. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  10. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  11. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  12. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  13. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  14. | ^
  15. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  16. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  17. | ^
  18. 2 errors generated when compiling for gfx1036.
  19. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_casual_gfx110x_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_casual_gfx110x_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  20. [51/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_casual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_casual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  21. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_casual_gfx110x_hip.o
  22. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_casual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_casual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  23. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_casual_gfx110x_hip.hip:25:
  24. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  25. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  26. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  27. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  28. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  29. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  30. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  31. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  32. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  33. | ^
  34. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  35. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  36. | ^
  37. 2 errors generated when compiling for gfx1036.
  38. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_casual_gfx110x_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_casual_gfx110x_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  39. [52/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_noncasual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_noncasual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  40. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_noncasual_gfx110x_hip.o
  41. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_noncasual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_noncasual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  42. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_noncasual_gfx110x_hip.hip:25:
  43. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  44. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  45. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  46. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  47. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  48. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  49. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  50. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  51. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  52. | ^
  53. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  54. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  55. | ^
  56. 2 errors generated when compiling for gfx1036.
  57. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_noncasual_gfx110x_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_bf16_noncasual_gfx110x_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  58. [53/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_noncasual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_noncasual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  59. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_noncasual_gfx110x_hip.o
  60. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_noncasual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_noncasual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  61. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_noncasual_gfx110x_hip.hip:25:
  62. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  63. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  64. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  65. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  66. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  67. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  68. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  69. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  70. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  71. | ^
  72. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  73. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  74. | ^
  75. 2 errors generated when compiling for gfx1036.
  76. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_noncasual_gfx110x_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_gqa_fp16_noncasual_gfx110x_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  77. [54/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_casual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_casual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  78. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_casual_gfx110x_hip.o
  79. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_casual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_casual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  80. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_casual_gfx110x_hip.hip:25:
  81. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  82. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  83. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  84. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  85. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  86. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  87. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  88. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  89. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  90. | ^
  91. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  92. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  93. | ^
  94. 2 errors generated when compiling for gfx1036.
  95. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_casual_gfx110x_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_casual_gfx110x_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  96. [55/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_noncasual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_noncasual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  97. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_noncasual_gfx110x_hip.o
  98. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_noncasual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_noncasual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  99. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_noncasual_gfx110x_hip.hip:25:
  100. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  101. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  102. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  103. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  104. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  105. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  106. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  107. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  108. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  109. | ^
  110. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  111. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  112. | ^
  113. 2 errors generated when compiling for gfx1036.
  114. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_noncasual_gfx110x_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_noncasual_gfx110x_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  115. [56/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_casual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_casual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  116. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_casual_gfx110x_hip.o
  117. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_casual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_casual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  118. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_casual_gfx110x_hip.hip:25:
  119. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  120. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  121. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  122. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  123. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  124. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  125. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  126. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  127. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  128. | ^
  129. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  130. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  131. | ^
  132. 2 errors generated when compiling for gfx1036.
  133. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_casual_gfx110x_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_fp16_casual_gfx110x_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  134. [57/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/flash_api_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/flash_api_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  135. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/flash_api_hip.o
  136. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/flash_api_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/flash_api_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  137. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/flash_api_hip.hip:14:
  138. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  139. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  140. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  141. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  142. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  143. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  144. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  145. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  146. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  147. | ^
  148. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  149. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  150. | ^
  151. 2 errors generated when compiling for gfx1036.
  152. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/flash_api_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/flash_api_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  153. [58/58] /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_noncasual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_noncasual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  154. FAILED: /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_noncasual_gfx110x_hip.o
  155. /opt/rocm-6.3.4/bin/hipcc -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_noncasual_gfx110x_hip.hip -o /tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_noncasual_gfx110x_hip.o -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ --offload-arch=native -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1016"' -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  156. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_noncasual_gfx110x_hip.hip:25:
  157. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_runner_hip.hpp:30:
  158. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_invoker_hip.hpp:27:
  159. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/fwd_device_gemm_template_hip.hpp:27:
  160. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/device_gemm_trait_hip.hpp:45:
  161. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_operation/gpu/device/impl/device_grouped_query_attention_forward_wmma_hip.hpp:17:
  162. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/tensor_description/tensor_descriptor_hip.hpp:7:
  163. In file included from /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/common_header_hip.hpp:37:
  164. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:32:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  165. 32 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  166. | ^
  167. /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include/ck/utility/amd_buffer_addressing.hpp:47:48: error: use of undeclared identifier 'CK_BUFFER_RESOURCE_3RD_DWORD'
  168. 47 | wave_buffer_resource.config(Number<3>{}) = CK_BUFFER_RESOURCE_3RD_DWORD;
  169. | ^
  170. 2 errors generated when compiling for gfx1036.
  171. failed to execute:/opt/rocm-6.3.4/lib/llvm/bin/clang++ --offload-arch=native -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/include -I/tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/composable_kernel/library/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/torch/csrc/api/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/include/THH -I/opt/rocm-6.3.4/include -I/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/include -I/usr/include/python3.12 -c -c -x hip /tmp/pip-req-build-jdvqq1lu/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_noncasual_gfx110x_hip.hip -o "/tmp/pip-req-build-jdvqq1lu/build/temp.linux-x86_64-cpython-312/csrc/flash_attn_rocm/src/flash_fwd_runner_batched_mqa_bf16_noncasual_gfx110x_hip.o" -D__HIP_PLATFORM_AMD__=1 -DUSE_ROCM=1 -DHIPBLAS_V2 -fPIC -DCUDA_HAS_FP16=1 -D__HIP_NO_HALF_OPERATORS__=1 -D__HIP_NO_HALF_CONVERSIONS__=1 -DHIP_ENABLE_WARP_SYNC_BUILTINS=1 -O3 -std=c++17 -DNDEBUG -U__HIP_NO_HALF_OPERATORS__ -U__HIP_NO_HALF_CONVERSIONS__ -D__WMMA__ -DTORCH_API_INCLUDE_EXTENSION_H -DPYBIND11_COMPILER_TYPE=\"_gcc\" -DPYBIND11_STDLIB=\"_libstdcpp\" -DPYBIND11_BUILD_ABI=\"_cxxabi1016\" -DTORCH_EXTENSION_NAME=flash_attn_2_cuda -fno-gpu-rdc
  172. ninja: build stopped: subcommand failed.
  173. Traceback (most recent call last):
  174. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 2560, in _run_ninja_build
  175. subprocess.run(
  176. File "/usr/lib/python3.12/subprocess.py", line 571, in run
  177. raise CalledProcessError(retcode, process.args,
  178. subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.
  179.  
  180. The above exception was the direct cause of the following exception:
  181.  
  182. Traceback (most recent call last):
  183. File "<string>", line 2, in <module>
  184. File "<pip-setuptools-caller>", line 35, in <module>
  185. File "/tmp/pip-req-build-jdvqq1lu/setup.py", line 380, in <module>
  186. setup(
  187. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/__init__.py", line 103, in setup
  188. return distutils.core.setup(**attrs)
  189. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  190. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/core.py", line 184, in setup
  191. return run_commands(dist)
  192. ^^^^^^^^^^^^^^^^^^
  193. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/core.py", line 200, in run_commands
  194. dist.run_commands()
  195. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 970, in run_commands
  196. self.run_command(cmd)
  197. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/dist.py", line 974, in run_command
  198. super().run_command(command)
  199. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 989, in run_command
  200. cmd_obj.run()
  201. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/command/bdist_wheel.py", line 373, in run
  202. self.run_command("build")
  203. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/cmd.py", line 316, in run_command
  204. self.distribution.run_command(command)
  205. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/dist.py", line 974, in run_command
  206. super().run_command(command)
  207. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 989, in run_command
  208. cmd_obj.run()
  209. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/command/build.py", line 135, in run
  210. self.run_command(cmd_name)
  211. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/cmd.py", line 316, in run_command
  212. self.distribution.run_command(command)
  213. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/dist.py", line 974, in run_command
  214. super().run_command(command)
  215. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 989, in run_command
  216. cmd_obj.run()
  217. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/command/build_ext.py", line 93, in run
  218. _build_ext.run(self)
  219. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 359, in run
  220. self.build_extensions()
  221. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 1056, in build_extensions
  222. build_ext.build_extensions(self)
  223. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 479, in build_extensions
  224. self._build_extensions_serial()
  225. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 505, in _build_extensions_serial
  226. self.build_extension(ext)
  227. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/command/build_ext.py", line 254, in build_extension
  228. _build_ext.build_extension(self, ext)
  229. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 560, in build_extension
  230. objects = self.compiler.compile(
  231. ^^^^^^^^^^^^^^^^^^^^^^
  232. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 840, in unix_wrap_ninja_compile
  233. _write_ninja_file_and_compile_objects(
  234. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 2213, in _write_ninja_file_and_compile_objects
  235. _run_ninja_build(
  236. File "/home/andrew/Downloads/SwarmUI/dlbackend/ComfyUI/venv/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 2577, in _run_ninja_build
  237. raise RuntimeError(message) from e
  238. RuntimeError: Error compiling objects for extension
  239. [end of output]
  240.  
  241. note: This error originates from a subprocess, and is likely not a problem with pip.
  242. ERROR: Failed building wheel for flash_attn
  243. Running setup.py clean for flash_attn
  244. Failed to build flash_attn
  245. ERROR: Failed to build installable wheels for some pyproject.toml based projects (flash_attn)
Advertisement
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!