Triton python backend build failed (main branch)
See original GitHub issueDescription The triton build failed if I try to build without cuda.
./build.py --cmake-dir=$(pwd)/build --build-dir=/tmp/citritonbuild --enable-logging --enable-stats --enable-tracing --enable-metrics --filesystem=azure_storage --endpoint=http --endpoint=grpc --repo-tag=common:main --repo-tag=core:main --repo-tag=backend:main --repo-tag=thirdparty:main --backend=ensemble --backend=identity:main --backend=repeat:main --backend=tensorflow2:main --backend=python:main --repoagent=checksum:main
/tmp/tritonbuild/python/src/python.cc: In member function ‘TRITONSERVER_Error* triton::backend::python::ModelInstanceState::GetInputTensor(uint32_t, triton::backend::python::Tensor*, TRITONBACKEND_Request*, std::vector<TRITONBACKEND_Response*>&)’:
/tmp/tritonbuild/python/src/python.cc:1429:7: error: ‘cudaSetDevice’ was not declared in this scope
1429 | cudaSetDevice(src_memory_type_id);
| ^~~~~~~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1430:7: error: ‘cudaError_t’ was not declared in this scope
1430 | cudaError_t err = cudaIpcGetMemHandle(
| ^~~~~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1432:11: error: ‘err’ was not declared in this scope; did you mean ‘erf’?
1432 | if (err != cudaSuccess) {
| ^~~
| erf
/tmp/tritonbuild/python/src/python.cc:1432:18: error: ‘cudaSuccess’ was not declared in this scope
1432 | if (err != cudaSuccess) {
| ^~~~~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1435:55: error: ‘cudaGetErrorName’ was not declared in this scope
1435 | std::string(cudaGetErrorName(err)))
| ^~~~~~~~~~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1439:7: error: ‘CUdeviceptr’ was not declared in this scope
1439 | CUdeviceptr start_address;
| ^~~~~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1440:7: error: ‘CUresult’ was not declared in this scope
1440 | CUresult cuda_err = cuPointerGetAttribute(
| ^~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1443:11: error: ‘cuda_err’ was not declared in this scope
1443 | if (cuda_err != CUDA_SUCCESS) {
| ^~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1443:23: error: ‘CUDA_SUCCESS’ was not declared in this scope; did you mean ‘EXIT_SUCCESS’?
1443 | if (cuda_err != CUDA_SUCCESS) {
| ^~~~~~~~~~~~
| EXIT_SUCCESS
/tmp/tritonbuild/python/src/python.cc:1445:9: error: ‘cuGetErrorString’ was not declared in this scope
1445 | cuGetErrorString(cuda_err, &error_string);
| ^~~~~~~~~~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1453:45: error: ‘start_address’ was not declared in this scope
1453 | reinterpret_cast<char*>(start_address);
| ^~~~~~~~~~~~~
/tmp/tritonbuild/python/src/python.cc:1454:7: error: ‘gpu_tensors_map_’ was not declared in this scope
1454 | gpu_tensors_map_.insert(
| ^~~~~~~~~~~~~~~~
make[2]: *** [CMakeFiles/triton-python-backend.dir/build.make:82: CMakeFiles/triton-python-backend.dir/src/python.cc.o] Error 1
make[2]: Leaving directory '/tmp/tritonbuild/python/build'
make[1]: *** [CMakeFiles/Makefile2:239: CMakeFiles/triton-python-backend.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs....
[ 97%] Linking CXX executable triton_python_backend_stub
/usr/bin/cmake -E cmake_link_script CMakeFiles/triton-python-backend-stub.dir/link.txt --verbose=0
make[2]: Leaving directory '/tmp/tritonbuild/python/build'
[ 97%] Built target triton-python-backend-stub
make[1]: Leaving directory '/tmp/tritonbuild/python/build'
make: *** [Makefile:149: all] Error 2
version 2.13.0dev
default repo-tag: main
backend "ensemble" at tag/branch "main"
backend "identity" at tag/branch "main"
backend "repeat" at tag/branch "main"
backend "tensorflow2" at tag/branch "main"
backend "python" at tag/branch "main"
repoagent "checksum" at tag/branch "main"
Building Triton Inference Server
component "common" at tag/branch "main"
component "core" at tag/branch "main"
component "backend" at tag/branch "main"
component "thirdparty" at tag/branch "main"
error: make install failed
error: docker run tritonserver_builder failed
Triton Information What version of Triton are you using? Main branch
Are you using the Triton container or did you build it yourself? Try to build by myself
To Reproduce Steps to reproduce the behavior. As above
Describe the models (framework, inputs, outputs), ideally include the model configuration file (if using an ensemble include the model configuration file for that as well). N/A
Expected behavior A clear and concise description of what you expected to happen. Build succeeded.
Issue Analytics
- State:
- Created 2 years ago
- Comments:7 (4 by maintainers)
Top GitHub Comments
Generally main branch is shippable and tested regularly but will some dev changes may cause unexpected issues in the build.
@NonStatic2014 This has been fixed now. Feel free to re-open if you are still having any issues.