vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

ci: add support for optimized Windows ARM64 builds with MSVC and LLVM

Browse source
This commit is contained in:
Max Krasnyansky 2024-05-14 20:36:57 -07:00
parent 18d20c0727
commit e838a3d459
1 changed files with 33 additions and 28 deletions
Download patch file Download diff file Expand all files Collapse all files

61
.github/workflows/build.yml vendored
View file

@ -693,26 +693,28 @@ jobs:
strategy: strategy:
matrix: matrix:
include: include:
- build: 'rpc' - build: 'rpc-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_RPC=ON -DBUILD_SHARED_LIBS=ON' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_RPC=ON -DBUILD_SHARED_LIBS=ON'
- build: 'noavx' - build: 'noavx-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_AVX=OFF -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF -DBUILD_SHARED_LIBS=ON' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_AVX=OFF -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF -DBUILD_SHARED_LIBS=ON'
- build: 'avx2' - build: 'avx2-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=ON' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=ON'
- build: 'avx' - build: 'avx-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_AVX2=OFF -DBUILD_SHARED_LIBS=ON' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_AVX2=OFF -DBUILD_SHARED_LIBS=ON'
- build: 'avx512' - build: 'avx512-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_AVX512=ON -DBUILD_SHARED_LIBS=ON' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_AVX512=ON -DBUILD_SHARED_LIBS=ON'
- build: 'clblast' - build: 'clblast-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_CLBLAST=ON -DBUILD_SHARED_LIBS=ON -DCMAKE_PREFIX_PATH="$env:RUNNER_TEMP/clblast"' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_CLBLAST=ON -DBUILD_SHARED_LIBS=ON -DCMAKE_PREFIX_PATH="$env:RUNNER_TEMP/clblast"'
- build: 'openblas' - build: 'openblas-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_BLAS=ON -DBUILD_SHARED_LIBS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_BLAS=ON -DBUILD_SHARED_LIBS=ON -DLLAMA_BLAS_VENDOR=OpenBLAS -DBLAS_INCLUDE_DIRS="$env:RUNNER_TEMP/openblas/include" -DBLAS_LIBRARIES="$env:RUNNER_TEMP/openblas/lib/openblas.lib"'
- build: 'kompute' - build: 'kompute-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_KOMPUTE=ON -DKOMPUTE_OPT_DISABLE_VULKAN_VERSION_CHECK=ON -DBUILD_SHARED_LIBS=ON' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_KOMPUTE=ON -DKOMPUTE_OPT_DISABLE_VULKAN_VERSION_CHECK=ON -DBUILD_SHARED_LIBS=ON'
- build: 'vulkan' - build: 'vulkan-x64'
defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_VULKAN=ON -DBUILD_SHARED_LIBS=ON' defines: '-DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DLLAMA_VULKAN=ON -DBUILD_SHARED_LIBS=ON'
- build: 'arm64' - build: 'llvm-arm64'
defines: '-A ARM64 -DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=ON' defines: '-G Ninja -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-llvm.cmake -DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=ON'
- build: 'msvc-arm64'
defines: '-G Ninja -D CMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-msvc.cmake -DLLAMA_NATIVE=OFF -DLLAMA_BUILD_SERVER=ON -DBUILD_SHARED_LIBS=ON'
steps: steps:
- name: Clone - name: Clone
@ -723,13 +725,13 @@ jobs:
- name: Clone Kompute submodule - name: Clone Kompute submodule
id: clone_kompute id: clone_kompute
if: ${{ matrix.build == 'kompute' }} if: ${{ matrix.build == 'kompute-x64' }}
run: | run: |
git submodule update --init kompute git submodule update --init kompute
- name: Download OpenCL SDK - name: Download OpenCL SDK
id: get_opencl id: get_opencl
if: ${{ matrix.build == 'clblast' }} if: ${{ matrix.build == 'clblast-x64' }}
run: | run: |
curl.exe -o $env:RUNNER_TEMP/opencl.zip -L "https://github.com/KhronosGroup/OpenCL-SDK/releases/download/v${env:OPENCL_VERSION}/OpenCL-SDK-v${env:OPENCL_VERSION}-Win-x64.zip" curl.exe -o $env:RUNNER_TEMP/opencl.zip -L "https://github.com/KhronosGroup/OpenCL-SDK/releases/download/v${env:OPENCL_VERSION}/OpenCL-SDK-v${env:OPENCL_VERSION}-Win-x64.zip"
mkdir $env:RUNNER_TEMP/opencl mkdir $env:RUNNER_TEMP/opencl
@ -737,7 +739,7 @@ jobs:
- name: Download CLBlast - name: Download CLBlast
id: get_clblast id: get_clblast
if: ${{ matrix.build == 'clblast' }} if: ${{ matrix.build == 'clblast-x64' }}
run: | run: |
curl.exe -o $env:RUNNER_TEMP/clblast.7z -L "https://github.com/CNugteren/CLBlast/releases/download/${env:CLBLAST_VERSION}/CLBlast-${env:CLBLAST_VERSION}-windows-x64.7z" curl.exe -o $env:RUNNER_TEMP/clblast.7z -L "https://github.com/CNugteren/CLBlast/releases/download/${env:CLBLAST_VERSION}/CLBlast-${env:CLBLAST_VERSION}-windows-x64.7z"
curl.exe -o $env:RUNNER_TEMP/CLBlast.LICENSE.txt -L "https://github.com/CNugteren/CLBlast/raw/${env:CLBLAST_VERSION}/LICENSE" curl.exe -o $env:RUNNER_TEMP/CLBlast.LICENSE.txt -L "https://github.com/CNugteren/CLBlast/raw/${env:CLBLAST_VERSION}/LICENSE"
@ -750,7 +752,7 @@ jobs:
- name: Download OpenBLAS - name: Download OpenBLAS
id: get_openblas id: get_openblas
if: ${{ matrix.build == 'openblas' }} if: ${{ matrix.build == 'openblas-x64' }}
run: | run: |
curl.exe -o $env:RUNNER_TEMP/openblas.zip -L "https://github.com/xianyi/OpenBLAS/releases/download/v${env:OPENBLAS_VERSION}/OpenBLAS-${env:OPENBLAS_VERSION}-x64.zip" curl.exe -o $env:RUNNER_TEMP/openblas.zip -L "https://github.com/xianyi/OpenBLAS/releases/download/v${env:OPENBLAS_VERSION}/OpenBLAS-${env:OPENBLAS_VERSION}-x64.zip"
curl.exe -o $env:RUNNER_TEMP/OpenBLAS.LICENSE.txt -L "https://github.com/xianyi/OpenBLAS/raw/v${env:OPENBLAS_VERSION}/LICENSE" curl.exe -o $env:RUNNER_TEMP/OpenBLAS.LICENSE.txt -L "https://github.com/xianyi/OpenBLAS/raw/v${env:OPENBLAS_VERSION}/LICENSE"
@ -763,38 +765,41 @@ jobs:
- name: Install Vulkan SDK - name: Install Vulkan SDK
id: get_vulkan id: get_vulkan
if: ${{ matrix.build == 'kompute' || matrix.build == 'vulkan' }} if: ${{ matrix.build == 'kompute-x64' || matrix.build == 'vulkan-x64' }}
run: | run: |
curl.exe -o $env:RUNNER_TEMP/VulkanSDK-Installer.exe -L "https://sdk.lunarg.com/sdk/download/${env:VULKAN_VERSION}/windows/VulkanSDK-${env:VULKAN_VERSION}-Installer.exe" curl.exe -o $env:RUNNER_TEMP/VulkanSDK-Installer.exe -L "https://sdk.lunarg.com/sdk/download/${env:VULKAN_VERSION}/windows/VulkanSDK-${env:VULKAN_VERSION}-Installer.exe"
& "$env:RUNNER_TEMP\VulkanSDK-Installer.exe" --accept-licenses --default-answer --confirm-command install & "$env:RUNNER_TEMP\VulkanSDK-Installer.exe" --accept-licenses --default-answer --confirm-command install
Add-Content $env:GITHUB_ENV "VULKAN_SDK=C:\VulkanSDK\${env:VULKAN_VERSION}" Add-Content $env:GITHUB_ENV "VULKAN_SDK=C:\VulkanSDK\${env:VULKAN_VERSION}"
Add-Content $env:GITHUB_PATH "C:\VulkanSDK\${env:VULKAN_VERSION}\bin" Add-Content $env:GITHUB_PATH "C:\VulkanSDK\${env:VULKAN_VERSION}\bin"
- name: Install Ninja
id: install_ninja
run: |
choco install ninja
- name: Build - name: Build
id: cmake_build id: cmake_build
run: | run: |
mkdir build cmake -S . -B build ${{ matrix.defines }}
cd build cmake --build build --config Release -j ${env:NUMBER_OF_PROCESSORS}
cmake .. ${{ matrix.defines }}
cmake --build . --config Release -j ${env:NUMBER_OF_PROCESSORS}
- name: Add clblast.dll - name: Add clblast.dll
id: add_clblast_dll id: add_clblast_dll
if: ${{ matrix.build == 'clblast' }} if: ${{ matrix.build == 'clblast-x64' }}
run: | run: |
cp $env:RUNNER_TEMP/clblast/lib/clblast.dll ./build/bin/Release cp $env:RUNNER_TEMP/clblast/lib/clblast.dll ./build/bin/Release
cp $env:RUNNER_TEMP/CLBlast.LICENSE.txt ./build/bin/Release/CLBlast-${env:CLBLAST_VERSION}.txt cp $env:RUNNER_TEMP/CLBlast.LICENSE.txt ./build/bin/Release/CLBlast-${env:CLBLAST_VERSION}.txt
- name: Add libopenblas.dll - name: Add libopenblas.dll
id: add_libopenblas_dll id: add_libopenblas_dll
if: ${{ matrix.build == 'openblas' }} if: ${{ matrix.build == 'openblas-x64' }}
run: | run: |
cp $env:RUNNER_TEMP/openblas/bin/libopenblas.dll ./build/bin/Release/openblas.dll cp $env:RUNNER_TEMP/openblas/bin/libopenblas.dll ./build/bin/Release/openblas.dll
cp $env:RUNNER_TEMP/OpenBLAS.LICENSE.txt ./build/bin/Release/OpenBLAS-${env:OPENBLAS_VERSION}.txt cp $env:RUNNER_TEMP/OpenBLAS.LICENSE.txt ./build/bin/Release/OpenBLAS-${env:OPENBLAS_VERSION}.txt
- name: Check AVX512F support - name: Check AVX512F support
id: check_avx512f id: check_avx512f
if: ${{ matrix.build == 'avx512' }} if: ${{ matrix.build == 'avx512-x64' }}
continue-on-error: true continue-on-error: true
run: | run: |
cd build cd build
@ -808,14 +813,14 @@ jobs:
- name: Test - name: Test
id: cmake_test id: cmake_test
# not all machines have native AVX-512 # not all machines have native AVX-512
if: ${{ matrix.build != 'arm64' && matrix.build != 'clblast' && matrix.build != 'kompute' && matrix.build != 'vulkan' && (matrix.build != 'avx512' || env.HAS_AVX512F == '1') }} if: ${{ matrix.build != 'msvc-arm64' && matrix.build != 'llvm-arm64' && matrix.build != 'clblast-x64' && matrix.build != 'kompute-x64' && matrix.build != 'vulkan-x64' && (matrix.build != 'avx512-x64' || env.HAS_AVX512F == '1') }}
run: | run: |
cd build cd build
ctest -L main -C Release --verbose --timeout 900 ctest -L main -C Release --verbose --timeout 900
- name: Test (Intel SDE) - name: Test (Intel SDE)
id: cmake_test_sde id: cmake_test_sde
if: ${{ matrix.build == 'avx512' && env.HAS_AVX512F == '0' }} # use Intel SDE for AVX-512 emulation if: ${{ matrix.build == 'avx512-x64' && env.HAS_AVX512F == '0' }} # use Intel SDE for AVX-512 emulation
run: | run: |
curl.exe -o $env:RUNNER_TEMP/sde.tar.xz -L "https://downloadmirror.intel.com/813591/sde-external-${env:SDE_VERSION}-win.tar.xz" curl.exe -o $env:RUNNER_TEMP/sde.tar.xz -L "https://downloadmirror.intel.com/813591/sde-external-${env:SDE_VERSION}-win.tar.xz"
# for some weird reason windows tar doesn't like sde tar.xz # for some weird reason windows tar doesn't like sde tar.xz
@ -843,14 +848,14 @@ jobs:
if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }} if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
run: | run: |
Copy-Item LICENSE .\build\bin\Release\llama.cpp.txt Copy-Item LICENSE .\build\bin\Release\llama.cpp.txt
7z a llama-${{ steps.tag.outputs.name }}-bin-win-${{ matrix.build }}-x64.zip .\build\bin\Release\* 7z a llama-${{ steps.tag.outputs.name }}-bin-win-${{ matrix.build }}.zip .\build\bin\Release\*
- name: Upload artifacts - name: Upload artifacts
if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }} if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
uses: actions/upload-artifact@v4 uses: actions/upload-artifact@v4
with: with:
path: llama-${{ steps.tag.outputs.name }}-bin-win-${{ matrix.build }}-x64.zip path: llama-${{ steps.tag.outputs.name }}-bin-win-${{ matrix.build }}.zip
name: llama-bin-win-${{ matrix.build }}-x64.zip name: llama-bin-win-${{ matrix.build }}.zip
windows-latest-cmake-cuda: windows-latest-cmake-cuda:
runs-on: windows-latest runs-on: windows-latest