how to compile on last gcc?

Open
#2,081 1 comment 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

Nobody has claimed this yet.

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
35/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Stale
Tech stack
cmake, python
Domain
build-system

Research direction

The error occurs during CMake's CUDA compiler identification when building llama-cpp-python with CUDA 11.4 and GCC 12.2.0. Look at the CMake configuration in the project's build scripts, particularly flags passed to nvcc. The issue is compatibility between CUDA 11.4 and newer GCC versions; check if the project documents supported compiler versions or if there's a workaround flag needed. Testing involves modifying CMake arguments or environment variables to see if the build succeeds.

Written by the indexing model from the issue text.

Description

export PATH=/usr/local/cuda-11.4/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-11.4/lib64:$LD_LIBRARY_PATH
NVCC_PREPEND_FLAGS="-allow-unsupported-compiler" CMAKE_ARGS="-DGGML_CUDA=on -DLLAVA_BUILD=off -DCMAKE_CUDA_FLAGS=-allow-unsupported-compiler" proxychains ../../bin/pip install -U llama-cpp-python
ProxyChains-3.1 (http://proxychains.sf.net)
|DNS-request| pypi.org 
|S-chain|-<>-127.0.0.1:4445-<><>-4.2.2.2:53-<><>-OK
|DNS-response| pypi.org is 151.101.0.223
|S-chain|-<>-127.0.0.1:4445-<><>-151.101.0.223:443-<><>-OK
Collecting llama-cpp-python
  Using cached llama_cpp_python-0.3.16.tar.gz (50.7 MB)
  Installing build dependencies ... done
  Getting requirements to build wheel ... done
  Installing backend dependencies ... done
  Preparing metadata (pyproject.toml) ... done
Requirement already satisfied: typing-extensions>=4.5.0 in /mnt/SOLR/lib/python3.11/site-packages (from llama-cpp-python) (4.15.0)
Requirement already satisfied: numpy>=1.20.0 in /mnt/SOLR/lib/python3.11/site-packages (from llama-cpp-python) (1.26.4)
Collecting diskcache>=5.6.1
  Using cached diskcache-5.6.3-py3-none-any.whl (45 kB)
Requirement already satisfied: jinja2>=2.11.3 in /mnt/SOLR/lib/python3.11/site-packages (from llama-cpp-python) (3.1.6)
Requirement already satisfied: MarkupSafe>=2.0 in /mnt/SOLR/lib/python3.11/site-packages (from jinja2>=2.11.3->llama-cpp-python) (3.0.3)
Building wheels for collected packages: llama-cpp-python
  Building wheel for llama-cpp-python (pyproject.toml) ... error
  error: subprocess-exited-with-error
  
  × Building wheel for llama-cpp-python (pyproject.toml) did not run successfully.
  │ exit code: 1
  ╰─> [379 lines of output]
      *** scikit-build-core 0.11.6 using CMake 3.25.1 (wheel)
      *** Configuring CMake...
      loading initial cache file /tmp/tmpsrjhdqci/build/CMakeInit.txt
      -- The C compiler identification is GNU 12.2.0
      -- The CXX compiler identification is GNU 12.2.0
      -- Detecting C compiler ABI info
      -- Detecting C compiler ABI info - done
      -- Check for working C compiler: /usr/bin/x86_64-linux-gnu-gcc - skipped
      -- Detecting C compile features
      -- Detecting C compile features - done
      -- Detecting CXX compiler ABI info
      -- Detecting CXX compiler ABI info - done
      -- Check for working CXX compiler: /usr/bin/x86_64-linux-gnu-g++ - skipped
      -- Detecting CXX compile features
      -- Detecting CXX compile features - done
      CMAKE_BUILD_TYPE=Release
      -- Found Git: /usr/bin/git (found version "2.39.5")
      -- Performing Test CMAKE_HAVE_LIBC_PTHREAD
      -- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
      -- Found Threads: TRUE
      -- Warning: ccache not found - consider installing it for faster compilation or disable this warning with GGML_CCACHE=OFF
      -- CMAKE_SYSTEM_PROCESSOR: x86_64
      -- GGML_SYSTEM_ARCH: x86
      -- Including CPU backend
      -- Found OpenMP_C: -fopenmp (found version "4.5")
      -- Found OpenMP_CXX: -fopenmp (found version "4.5")
      -- Found OpenMP: TRUE (found version "4.5")
      -- x86 detected
      -- Adding CPU backend variant ggml-cpu: -march=native
      -- Found CUDAToolkit: /usr/local/cuda-11.4/include (found version "11.4.48")
      -- CUDA Toolkit found
      -- Using CUDA architectures: 50-virtual;61-virtual;70-virtual;75-virtual;80-virtual;86-real
      CMake Error at /usr/share/cmake-3.25/Modules/CMakeDetermineCompilerId.cmake:739 (message):
        Compiling the CUDA compiler identification source file
        "CMakeCUDACompilerId.cu" failed.
      
        Compiler: /usr/local/cuda-11.4/bin/nvcc
      
        Build flags: -allow-unsupported-compiler
      
        Id flags: --keep;--keep-dir;tmp -v
      
      
      
        The output was:
      
        1
      
        #$ NVCC_PREPEND_FLAGS="-allow-unsupported-compiler"
      
        #$ _NVVM_BRANCH_=nvvm
      
        #$ _SPACE_=
      
        #$ _CUDART_=cudart
      
        #$ _HERE_=/usr/local/cuda-11.4/bin
      
        #$ _THERE_=/usr/local/cuda-11.4/bin
      
        #$ _TARGET_SIZE_=
      
        #$ _TARGET_DIR_=
      
        #$ _TARGET_DIR_=targets/x86_64-linux
      
        #$ TOP=/usr/local/cuda-11.4/bin/..
      
        #$ NVVMIR_LIBRARY_DIR=/usr/local/cuda-11.4/bin/../nvvm/libdevice
      
        #$
        LD_LIBRARY_PATH=/usr/local/cuda-11.4/bin/../lib:/usr/local/cuda-11.4/lib64:
      
      
        #$
        PATH=/usr/local/cuda-11.4/bin/../nvvm/bin:/usr/local/cuda-11.4/bin:/tmp/pip-build-env-zklleoev/overlay/bin:/tmp/pip-build-env-zklleoev/normal/bin:/usr/local/cuda-11.4/bin:/home/user/.cargo/bin:/usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games
      
      
        #$ INCLUDES="-I/usr/local/cuda-11.4/bin/../targets/x86_64-linux/include"
      
        #$ LIBRARIES=
        "-L/usr/local/cuda-11.4/bin/../targets/x86_64-linux/lib/stubs"
        "-L/usr/local/cuda-11.4/bin/../targets/x86_64-linux/lib"
      
        #$ CUDAFE_FLAGS=
      
        #$ PTXAS_FLAGS=
      
        #$ rm tmp/a_dlink.reg.c
      
        #$ gcc -D__NV_NO_HOST_COMPILER_CHECK=1 -D__CUDA_ARCH__=520 -E -x c++
        -DCUDA_DOUBLE_MATH_FUNCTIONS -D__CUDACC__ -D__NVCC__
        "-I/usr/local/cuda-11.4/bin/../targets/x86_64-linux/include"
        -D__CUDACC_VER_MAJOR__=11 -D__CUDACC_VER_MINOR__=4
        -D__CUDACC_VER_BUILD__=48 -D__CUDA_API_VER_MAJOR__=11
        -D__CUDA_API_VER_MINOR__=4 -include "cuda_runtime.h" -m64
        "CMakeCUDACompilerId.cu" -o "tmp/CMakeCUDACompilerId.cpp1.ii"
      
        #$ cicc --c++14 --gnu_version=120200 --orig_src_file_name
        "CMakeCUDACompilerId.cu" --allow_managed -arch compute_52 -m64
        --no-version-ident -ftz=0 -prec_div=1 -prec_sqrt=1 -fmad=1
        --include_file_name "CMakeCUDACompilerId.fatbin.c" -tused
        --gen_module_id_file --module_id_file_name
        "tmp/CMakeCUDACompilerId.module_id" --gen_c_file_name
        "tmp/CMakeCUDACompilerId.cudafe1.c" --stub_file_name
        "tmp/CMakeCUDACompilerId.cudafe1.stub.c" --gen_device_file_name
        "tmp/CMakeCUDACompilerId.cudafe1.gpu" "tmp/CMakeCUDACompilerId.cpp1.ii" -o
        "tmp/CMakeCUDACompilerId.ptx"
      
        /usr/include/stdio.h(189): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(201): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(223): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(260): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(285): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(294): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(303): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(309): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(315): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(830): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdlib.h(579): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdlib.h(583): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdlib.h(812): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/c++/12/bits/stl_pair.h(634): error: argument list for class
        template "std::pair" is missing
      
      
      
        /usr/include/c++/12/bits/stl_pair.h(634): error: expected a ")"
      
      
      
        /usr/include/c++/12/bits/stl_pair.h(634): error: template parameter "_T1"
        may not be redeclared in this scope
      
      
      
        /usr/include/c++/12/bits/stl_pair.h(634): error: expected a ";"
      
      
      
        17 errors detected in the compilation of "CMakeCUDACompilerId.cu".
      
        # --error 0x1 --
      
      
      
      
      
        Compiling the CUDA compiler identification source file
        "CMakeCUDACompilerId.cu" failed.
      
        Compiler: /usr/local/cuda-11.4/bin/nvcc
      
        Build flags:
      
        Id flags: --keep;--keep-dir;tmp -v
      
      
      
        The output was:
      
        1
      
        #$ NVCC_PREPEND_FLAGS="-allow-unsupported-compiler"
      
        #$ _NVVM_BRANCH_=nvvm
      
        #$ _SPACE_=
      
        #$ _CUDART_=cudart
      
        #$ _HERE_=/usr/local/cuda-11.4/bin
      
        #$ _THERE_=/usr/local/cuda-11.4/bin
      
        #$ _TARGET_SIZE_=
      
        #$ _TARGET_DIR_=
      
        #$ _TARGET_DIR_=targets/x86_64-linux
      
        #$ TOP=/usr/local/cuda-11.4/bin/..
      
        #$ NVVMIR_LIBRARY_DIR=/usr/local/cuda-11.4/bin/../nvvm/libdevice
      
        #$
        LD_LIBRARY_PATH=/usr/local/cuda-11.4/bin/../lib:/usr/local/cuda-11.4/lib64:
      
      
        #$
        PATH=/usr/local/cuda-11.4/bin/../nvvm/bin:/usr/local/cuda-11.4/bin:/tmp/pip-build-env-zklleoev/overlay/bin:/tmp/pip-build-env-zklleoev/normal/bin:/usr/local/cuda-11.4/bin:/home/user/.cargo/bin:/usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games
      
      
        #$ INCLUDES="-I/usr/local/cuda-11.4/bin/../targets/x86_64-linux/include"
      
        #$ LIBRARIES=
        "-L/usr/local/cuda-11.4/bin/../targets/x86_64-linux/lib/stubs"
        "-L/usr/local/cuda-11.4/bin/../targets/x86_64-linux/lib"
      
        #$ CUDAFE_FLAGS=
      
        #$ PTXAS_FLAGS=
      
        #$ rm tmp/a_dlink.reg.c
      
        #$ gcc -D__NV_NO_HOST_COMPILER_CHECK=1 -D__CUDA_ARCH__=520 -E -x c++
        -DCUDA_DOUBLE_MATH_FUNCTIONS -D__CUDACC__ -D__NVCC__
        "-I/usr/local/cuda-11.4/bin/../targets/x86_64-linux/include"
        -D__CUDACC_VER_MAJOR__=11 -D__CUDACC_VER_MINOR__=4
        -D__CUDACC_VER_BUILD__=48 -D__CUDA_API_VER_MAJOR__=11
        -D__CUDA_API_VER_MINOR__=4 -include "cuda_runtime.h" -m64
        "CMakeCUDACompilerId.cu" -o "tmp/CMakeCUDACompilerId.cpp1.ii"
      
        #$ cicc --c++14 --gnu_version=120200 --orig_src_file_name
        "CMakeCUDACompilerId.cu" --allow_managed -arch compute_52 -m64
        --no-version-ident -ftz=0 -prec_div=1 -prec_sqrt=1 -fmad=1
        --include_file_name "CMakeCUDACompilerId.fatbin.c" -tused
        --gen_module_id_file --module_id_file_name
        "tmp/CMakeCUDACompilerId.module_id" --gen_c_file_name
        "tmp/CMakeCUDACompilerId.cudafe1.c" --stub_file_name
        "tmp/CMakeCUDACompilerId.cudafe1.stub.c" --gen_device_file_name
        "tmp/CMakeCUDACompilerId.cudafe1.gpu" "tmp/CMakeCUDACompilerId.cpp1.ii" -o
        "tmp/CMakeCUDACompilerId.ptx"
      
        /usr/include/stdio.h(189): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(201): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(223): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(260): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(285): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(294): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(303): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(309): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(315): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdio.h(830): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdlib.h(579): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdlib.h(583): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/stdlib.h(812): error: attribute "__malloc__" does not take
        arguments
      
      
      
        /usr/include/c++/12/bits/stl_pair.h(634): error: argument list for class
        template "std::pair" is missing
      
      
      
        /usr/include/c++/12/bits/stl_pair.h(634): error: expected a ")"
      
      
      
        /usr/include/c++/12/bits/stl_pair.h(634): error: template parameter "_T1"
        may not be redeclared in this scope
      
      
      
        /usr/include/c++/12/bits/stl_pair.h(634): error: expected a ";"
      
      
      
        17 errors detected in the compilation of "CMakeCUDACompilerId.cu".
      
        # --error 0x1 --
      
      
      
      
      
      Call Stack (most recent call first):
        /usr/share/cmake-3.25/Modules/CMakeDetermineCompilerId.cmake:6 (CMAKE_DETERMINE_COMPILER_ID_BUILD)
        /usr/share/cmake-3.25/Modules/CMakeDetermineCompilerId.cmake:48 (__determine_compiler_id_test)
        /usr/share/cmake-3.25/Modules/CMakeDetermineCUDACompiler.cmake:307 (CMAKE_DETERMINE_COMPILER_ID)
        vendor/llama.cpp/ggml/src/ggml-cuda/CMakeLists.txt:43 (enable_language)
      
      
      -- Configuring incomplete, errors occurred!
      See also "/tmp/tmpsrjhdqci/build/CMakeFiles/CMakeOutput.log".
      See also "/tmp/tmpsrjhdqci/build/CMakeFiles/CMakeError.log".
      
      *** CMake configuration failed
      [end of output]
  
  note: This error originates from a subprocess, and is likely not a problem with pip.
  ERROR: Failed building wheel for llama-cpp-python
Failed to build llama-cpp-python
ERROR: Could not build wheels for llama-cpp-python, which is required to install pyproject.toml-based projects

It need gcc10, I should compile gcc10?
llama-cpp-python have exists version for cuda 11?

Dominant language
Python
Stars
10.7k
Forks
1.5k
Avg merge
1h 40m
Merged PRs (30d)
5

Getting set up

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from abetlen/llama-cpp-python

All issues in abetlen/llama-cpp-python

Similar issues

More Python issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.