Skip to content
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.

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

[Build] SegFault while testing build with Nuphar EP #12894

Closed
dmt-obd opened this issue Sep 8, 2022 · 1 comment
Closed

[Build] SegFault while testing build with Nuphar EP #12894

dmt-obd opened this issue Sep 8, 2022 · 1 comment
Labels
build build issues; typically submitted using template

Comments

@dmt-obd
Copy link

dmt-obd commented Sep 8, 2022

Describe the issue

I'm trying to build package with Nuphar EP. Speedup figures from demo notebook are impressive and I can not find information that with code is deprecated now.

Urgency

Not urgent.

Target platform

5.14.0-1049-oem #56-Ubuntu SMP Fri Aug 12 10:23:08 UTC 2022 x86_64 GNU/Linux

Build script

I have to run different command rather than recommended because some arguments are deprecated with current version.

$ ./build.sh --use_dnnl --use_nuphar --llvm_path=/usr/lib/llvm-6.0/lib/cmake/llvm/ --config Release  --build_shared_lib --parallel --skip_tests --build_wheel
...
-- Link with dynamic LLVM library
-- Found LLVM_INCLUDE_DIRS=/usr/lib/llvm-6.0/include
-- Found LLVM_DEFINITIONS=-D_GNU_SOURCE -D__STDC_CONSTANT_MACROS -D__STDC_FORMAT_MACROS -D__STDC_LIMIT_MACROS
-- Found TVM_LLVM_VERSION=60
-- Build with LLVM 6.0.1
-- Set TVM_LLVM_VERSION=60
-- Build with contrib.hybriddump
-- TVM AFTER USE_LLVM=ON USE_OPENMP=OFF CMAKE_BUILD_TYPE=Release USE_CUDA=OFF USE_GTEST= CMAKE_BUILD_TYPE=Release
Generated: /home/do-op/Projects/onnxruntime/build/Linux/Release/external/onnx/onnx/onnx-ml.proto
Generated: /home/do-op/Projects/onnxruntime/build/Linux/Release/external/onnx/onnx/onnx-operators-ml.proto
Generated: /home/do-op/Projects/onnxruntime/build/Linux/Release/external/onnx/onnx/onnx-data.proto
-- 
-- ******** Summary ********
--   CMake version             : 3.22.1
--   CMake command             : /usr/local/bin/cmake
--   System                    : Linux
--   C++ compiler              : /usr/bin/c++
--   C++ compiler version      : 9.4.0
--   CXX flags                 :  -ffunction-sections -fdata-sections -DCPUINFO_SUPPORTED -Wnon-virtual-dtor
--   Build type                : Release
--   Compile definitions       : EIGEN_MPL2_ONLY;PLATFORM_POSIX;USE_NUPHAR_TVM_WITH_LLVM;USE_NUPHAR_TVM;__STDC_FORMAT_MACROS
--   CMAKE_PREFIX_PATH         : 
--   CMAKE_INSTALL_PREFIX      : /usr/local
--   CMAKE_MODULE_PATH         : /home/do-op/Projects/onnxruntime/cmake/external
-- 
--   ONNX version              : 1.12.0
--   ONNX NAMESPACE            : onnx
--   ONNX_USE_LITE_PROTO       : ON
--   USE_PROTOBUF_SHARED_LIBS  : OFF
--   Protobuf_USE_STATIC_LIBS  : ON
--   ONNX_DISABLE_EXCEPTIONS   : OFF
--   ONNX_WERROR               : OFF
--   ONNX_BUILD_TESTS          : OFF
--   ONNX_BUILD_BENCHMARKS     : OFF
--   ONNXIFI_DUMMY_BACKEND     : OFF
--   ONNXIFI_ENABLE_EXT        : OFF
-- 
--   Protobuf compiler         : 
--   Protobuf includes         : 
--   Protobuf libraries        : 
--   BUILD_ONNX_PYTHON         : OFF
...
2022-09-07 14:40:44,299 util.run [DEBUG] - Subprocess completed. Return code: 0
2022-09-07 14:40:44,299 build [INFO] - Build complete
$ ls build/Linux/Release/dist/
onnxruntime_dnnl-1.12.1-cp38-cp38-linux_x86_64.whl  onnxruntime_nuphar-1.12.1-cp38-cp38-linux_x86_64.whl

Error / output

$ ./onnxruntime_test_all > test_all.log 2>&1
Segmentation fault (core dumped)
$ cat test_all.log | grep '[       OK ]' | wc -l
3032
$ tail test_all.log 
[ RUN      ] MathOpTest.DNNL_Pow_Broadcast_Scalar1
[       OK ] MathOpTest.DNNL_Pow_Broadcast_Scalar1 (7 ms)
[ RUN      ] MathOpTest.DNNL_Pow_Broadcast_Scalar1_12
[       OK ] MathOpTest.DNNL_Pow_Broadcast_Scalar1_12 (7 ms)
[ RUN      ] MathOpTest.DNNL_Pow_Broadcast_Scalar1_float_int32_12
[       OK ] MathOpTest.DNNL_Pow_Broadcast_Scalar1_float_int32_12 (7 ms)
[ RUN      ] MathOpTest.DNNL_Pow_Broadcast_Scalar1_float_int8_12
[       OK ] MathOpTest.DNNL_Pow_Broadcast_Scalar1_float_int8_12 (7 ms)
[ RUN      ] MathOpTest.DNNL_Pow_Broadcast_Scalar1_float_uint8_12
LLVM ERROR: Program used external function 'pow_f32' which could not be resolved!

Visual Studio Version

No response

GCC / Compiler Version

c++ (Ubuntu 9.4.0-1ubuntu1~20.04.1) 9.4.0

@dmt-obd dmt-obd added the build build issues; typically submitted using template label Sep 8, 2022
@snnn
Copy link
Member

snnn commented Sep 8, 2022

Nuphar is removed from our codebase. Please consider using the new TVM EP.

@snnn snnn closed this as completed Sep 8, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
build build issues; typically submitted using template
Projects
None yet
Development

No branches or pull requests

2 participants