Skip to content

The compile_bundle.sh script fails at ipex build step #441

Open
@fcharras

Description

@fcharras

Describe the bug

The command:

LD_LIBRARY_PRELOAD="" bash ./compile_bundle.sh $HOME/intel/oneapi/compiler/latest/ $HOME/intel/oneapi/mkl/latest/ "pvc,ats-m150,tgllp,Gen12LP"

fails with the following error:

[ 97%] Building CXX object csrc/gpu/CMakeFiles/intel-ext-pt-gpu.dir/aten/operators/SoftShrink.cpp.o
[ 97%] Building CXX object csrc/gpu/CMakeFiles/intel-ext-pt-gpu.dir/aten/operators/SparseCsrTensorMath.cpp.o
/home/u103857/torch_compiler/intel-extension-for-pytorch/csrc/gpu/aten/operators/QConv_prepack.cpp:129:3: warning: use of function template name with no prior declaration in function call with explicit template arguments is a C++20 extension [-Wc++20-extensions]
  register_prepack<2>(
  ^
/home/u103857/torch_compiler/intel-extension-for-pytorch/csrc/gpu/aten/operators/QConv_prepack.cpp:130:20: error: no member named 'QXPU' in 'c10::QEngine'
      at::QEngine::QXPU,
      ~~~~~~~~~~~~~^
/home/u103857/torch_compiler/intel-extension-for-pytorch/csrc/gpu/aten/operators/QConv_prepack.cpp:129:3: warning: use of function template name with no prior declaration in function call with explicit template arguments is a C++20 extension [-Wc++20-extensions]
  register_prepack<2>(
  ^
/home/u103857/torch_compiler/intel-extension-for-pytorch/csrc/gpu/aten/operators/QConv_prepack.cpp:132:3: warning: use of function template name with no prior declaration in function call with explicit template arguments is a C++20 extension [-Wc++20-extensions]
  register_prepack<3>(
  ^
/home/u103857/torch_compiler/intel-extension-for-pytorch/csrc/gpu/aten/operators/QConv_prepack.cpp:133:20: error: no member named 'QXPU' in 'c10::QEngine'
      at::QEngine::QXPU,
      ~~~~~~~~~~~~~^
/home/u103857/torch_compiler/intel-extension-for-pytorch/csrc/gpu/aten/operators/QConv_prepack.cpp:132:3: warning: use of function template name with no prior declaration in function call with explicit template arguments is a C++20 extension [-Wc++20-extensions]
  register_prepack<3>(
  ^
/home/u103857/torch_compiler/intel-extension-for-pytorch/csrc/gpu/aten/operators/QConv_prepack.cpp:136:20: error: no member named 'QXPU' in 'c10::QEngine'
      at::QEngine::QXPU,
      ~~~~~~~~~~~~~^
4 warnings and 3 errors generated.
make[2]: *** [csrc/gpu/CMakeFiles/intel-ext-pt-gpu.dir/build.make:2721: csrc/gpu/CMakeFiles/intel-ext-pt-gpu.dir/aten/operators/QConv_prepack.cpp.o] Error 1
make[2]: *** Waiting for unfinished jobs....
make[1]: *** [CMakeFiles/Makefile2:594: csrc/gpu/CMakeFiles/intel-ext-pt-gpu.dir/all] Error 2
make: *** [Makefile:136: all] Error 2

Versions

with oneapi basekit 2023.2.0.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions