From 9dffd12f9070a75aee1a9eb9ac137b48dba2468e Mon Sep 17 00:00:00 2001 From: Jithun Nair Date: Wed, 4 Dec 2024 07:00:25 +0000 Subject: [PATCH] Upgrade ROCm wheels to manylinux2_28 - 2 of 2 (binaries) (#141423) Depends on https://github.com/pytorch/pytorch/pull/140681 and https://github.com/pytorch/pytorch/pull/141609 Highlights: * Upgrade binaries to ROCm6.2.4 to use latest docker images * Remove pre-cxx11 builds for libtorch on ROCm * Use manylinux2_28 docker images for ROCm * Set `DESIRED_DEVTOOLSET=cxx-abi` (and hence `_GLIBCXX_USE_CXX11_ABI=1`) for ROCm manylinux2_28 wheels (ROCm RHEL8 packages also have GCC_ABI=1, so it keeps it consistent) Pull Request resolved: https://github.com/pytorch/pytorch/pull/141423 Approved by: https://github.com/jeffdaily Co-authored-by: Jeff Daily Co-authored-by: Pruthvi Madugundu --- .ci/manywheel/build_rocm.sh | 28 ++- .github/scripts/amd/package_triton_wheel.sh | 7 +- .../scripts/generate_binary_build_matrix.py | 18 +- .github/workflows/build-triton-wheel.yml | 8 +- ...inux-binary-libtorch-cxx11-abi-nightly.yml | 36 +-- ...inux-binary-libtorch-pre-cxx11-nightly.yml | 220 ------------------ ...nerated-linux-binary-manywheel-nightly.yml | 200 +++++++++------- 7 files changed, 168 insertions(+), 349 deletions(-) diff --git a/.ci/manywheel/build_rocm.sh b/.ci/manywheel/build_rocm.sh index 1e14c9d81d24..b9bb67631927 100755 --- a/.ci/manywheel/build_rocm.sh +++ b/.ci/manywheel/build_rocm.sh @@ -107,17 +107,29 @@ if [[ $ROCM_INT -ge 60200 ]]; then fi OS_NAME=`awk -F= '/^NAME/{print $2}' /etc/os-release` -if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then +if [[ "$OS_NAME" == *"CentOS Linux"* || "$OS_NAME" == *"AlmaLinux"* ]]; then LIBGOMP_PATH="/usr/lib64/libgomp.so.1" LIBNUMA_PATH="/usr/lib64/libnuma.so.1" LIBELF_PATH="/usr/lib64/libelf.so.1" - LIBTINFO_PATH="/usr/lib64/libtinfo.so.5" + if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then + LIBTINFO_PATH="/usr/lib64/libtinfo.so.5" + else + LIBTINFO_PATH="/usr/lib64/libtinfo.so.6" + fi LIBDRM_PATH="/opt/amdgpu/lib64/libdrm.so.2" LIBDRM_AMDGPU_PATH="/opt/amdgpu/lib64/libdrm_amdgpu.so.1" if [[ $ROCM_INT -ge 60100 ]]; then # Below libs are direct dependencies of libhipsolver LIBSUITESPARSE_CONFIG_PATH="/lib64/libsuitesparseconfig.so.4" - LIBCHOLMOD_PATH="/lib64/libcholmod.so.2" + if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then + LIBCHOLMOD_PATH="/lib64/libcholmod.so.2" + # Below libs are direct dependencies of libsatlas + LIBGFORTRAN_PATH="/lib64/libgfortran.so.3" + else + LIBCHOLMOD_PATH="/lib64/libcholmod.so.3" + # Below libs are direct dependencies of libsatlas + LIBGFORTRAN_PATH="/lib64/libgfortran.so.5" + fi # Below libs are direct dependencies of libcholmod LIBAMD_PATH="/lib64/libamd.so.2" LIBCAMD_PATH="/lib64/libcamd.so.2" @@ -125,7 +137,6 @@ if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then LIBCOLAMD_PATH="/lib64/libcolamd.so.2" LIBSATLAS_PATH="/lib64/atlas/libsatlas.so.3" # Below libs are direct dependencies of libsatlas - LIBGFORTRAN_PATH="/lib64/libgfortran.so.3" LIBQUADMATH_PATH="/lib64/libquadmath.so.0" fi MAYBE_LIB64=lib64 @@ -175,9 +186,12 @@ do OS_SO_FILES[${#OS_SO_FILES[@]}]=$file_name # Append lib to array done -# PyTorch-version specific -# AOTriton dependency only for PyTorch >= 2.4 -if (( $(echo "${PYTORCH_VERSION} 2.4" | awk '{print ($1 >= $2)}') )); then +# FIXME: Temporary until https://github.com/pytorch/pytorch/pull/137443 lands +# Install AOTriton +if [ -e ${PYTORCH_ROOT}/.ci/docker/aotriton_version.txt ]; then + cp -a ${PYTORCH_ROOT}/.ci/docker/aotriton_version.txt aotriton_version.txt + bash ${PYTORCH_ROOT}/.ci/docker/common/install_aotriton.sh ${ROCM_HOME} && rm aotriton_version.txt + export AOTRITON_INSTALLED_PREFIX=${ROCM_HOME}/aotriton ROCM_SO_FILES+=("libaotriton_v2.so") fi diff --git a/.github/scripts/amd/package_triton_wheel.sh b/.github/scripts/amd/package_triton_wheel.sh index e08e58dbfd44..d10986c4338a 100755 --- a/.github/scripts/amd/package_triton_wheel.sh +++ b/.github/scripts/amd/package_triton_wheel.sh @@ -30,9 +30,14 @@ fi # Remove packaged libs and headers rm -rf $TRITON_ROCM_DIR/include/* -LIBTINFO_PATH="/usr/lib64/libtinfo.so.5" LIBNUMA_PATH="/usr/lib64/libnuma.so.1" LIBELF_PATH="/usr/lib64/libelf.so.1" +OS_NAME=`awk -F= '/^NAME/{print $2}' /etc/os-release` +if [[ "$OS_NAME" == *"CentOS Linux"* ]]; then + LIBTINFO_PATH="/usr/lib64/libtinfo.so.5" +else + LIBTINFO_PATH="/usr/lib64/libtinfo.so.6" +fi OS_SO_PATHS=( $LIBELF_PATH diff --git a/.github/scripts/generate_binary_build_matrix.py b/.github/scripts/generate_binary_build_matrix.py index 6b7dd41f84fb..892daee224a2 100644 --- a/.github/scripts/generate_binary_build_matrix.py +++ b/.github/scripts/generate_binary_build_matrix.py @@ -20,7 +20,7 @@ CUDA_ARCHES = ["11.8", "12.4", "12.6"] CUDA_ARCHES_FULL_VERSION = {"11.8": "11.8.0", "12.4": "12.4.1", "12.6": "12.6.3"} CUDA_ARCHES_CUDNN_VERSION = {"11.8": "9", "12.4": "9", "12.6": "9"} -ROCM_ARCHES = ["6.1", "6.2"] +ROCM_ARCHES = ["6.1", "6.2.4"] XPU_ARCHES = ["xpu"] @@ -162,7 +162,7 @@ WHEEL_CONTAINER_IMAGES = { "12.4": f"pytorch/manylinux-builder:cuda12.4-{DEFAULT_TAG}", "12.6": f"pytorch/manylinux2_28-builder:cuda12.6-{DEFAULT_TAG}", **{ - gpu_arch: f"pytorch/manylinux-builder:rocm{gpu_arch}-{DEFAULT_TAG}" + gpu_arch: f"pytorch/manylinux2_28-builder:rocm{gpu_arch}-{DEFAULT_TAG}" for gpu_arch in ROCM_ARCHES }, "xpu": f"pytorch/manylinux2_28-builder:xpu-{DEFAULT_TAG}", @@ -194,13 +194,6 @@ LIBTORCH_CONTAINER_IMAGES: Dict[Tuple[str, str], str] = { ): f"pytorch/libtorch-cxx11-builder:cuda{gpu_arch}-{DEFAULT_TAG}" for gpu_arch in CUDA_ARCHES }, - **{ - ( - gpu_arch, - PRE_CXX11_ABI, - ): f"pytorch/manylinux-builder:rocm{gpu_arch}-{DEFAULT_TAG}" - for gpu_arch in ROCM_ARCHES - }, **{ ( gpu_arch, @@ -262,7 +255,9 @@ def generate_libtorch_matrix( gpu_arch_type = arch_type(arch_version) gpu_arch_version = "" if arch_version == "cpu" else arch_version # ROCm builds without-deps failed even in ROCm runners; skip for now - if gpu_arch_type == "rocm" and "without-deps" in libtorch_variant: + if gpu_arch_type == "rocm" and ( + "without-deps" in libtorch_variant or "pre-cxx11" in abi_version + ): continue ret.append( { @@ -428,7 +423,8 @@ def generate_wheels_matrix( "use_split_build": "True" if use_split_build else "False", "devtoolset": ( "cxx11-abi" - if arch_version in ["cpu-cxx11-abi", "cpu-aarch64"] + if (arch_version in ["cpu-cxx11-abi", "cpu-aarch64"]) + or gpu_arch_type == "rocm" else "" ), "container_image": WHEEL_CONTAINER_IMAGES[arch_version], diff --git a/.github/workflows/build-triton-wheel.yml b/.github/workflows/build-triton-wheel.yml index f780690ac4e0..6caf064d4ed0 100644 --- a/.github/workflows/build-triton-wheel.yml +++ b/.github/workflows/build-triton-wheel.yml @@ -49,17 +49,17 @@ jobs: docker-image: ["pytorch/manylinux-builder:cpu", "pytorch/manylinux2_28-builder:cpu"] exclude: - device: "rocm" - docker-image: "pytorch/manylinux2_28-builder:cpu" + docker-image: "pytorch/manylinux-builder:cpu" - device: "xpu" docker-image: "pytorch/manylinux2_28-builder:cpu" include: - device: "rocm" - rocm_version: "6.2" + rocm_version: "6.2.4" - device: "cuda" rocm_version: "" timeout-minutes: 40 env: - DOCKER_IMAGE: ${{ matrix.device == 'rocm' && format('pytorch/manylinux-builder:rocm{0}', matrix.rocm_version) || matrix.docker-image }} + DOCKER_IMAGE: ${{ matrix.device == 'rocm' && format('pytorch/manylinux2_28-builder:rocm{0}', matrix.rocm_version) || matrix.docker-image }} PY_VERS: ${{ matrix.py_vers }} BUILD_DEVICE: ${{ matrix.device }} PLATFORM: ${{ contains(matrix.docker-image, '2_28') && 'manylinux_2_28_x86_64' || 'manylinux2014_x86_64' }} @@ -128,7 +128,7 @@ jobs: docker exec -t "${container_name}" yum install -y zlib-devel zip docker exec -t "${container_name}" "${PYTHON_EXECUTABLE}" -m pip install -U setuptools==67.4.0 pybind11==2.13.1 auditwheel - if [[ "${{ matrix.device }}" == "cuda" && "${PLATFORM}" == "manylinux_2_28_x86_64" ]]; then + if [[ ("${{ matrix.device }}" == "cuda" || "${{ matrix.device }}" == "rocm") && "${PLATFORM}" == "manylinux_2_28_x86_64" ]]; then # With this install, it gets clang 16.0.6. docker exec -t "${container_name}" dnf install clang lld -y WITH_CLANG_LDD="--with-clang-ldd" diff --git a/.github/workflows/generated-linux-binary-libtorch-cxx11-abi-nightly.yml b/.github/workflows/generated-linux-binary-libtorch-cxx11-abi-nightly.yml index 96e9ef651285..5f513f2a16ac 100644 --- a/.github/workflows/generated-linux-binary-libtorch-cxx11-abi-nightly.yml +++ b/.github/workflows/generated-linux-binary-libtorch-cxx11-abi-nightly.yml @@ -433,7 +433,7 @@ jobs: conda-pytorchbot-token-test: ${{ secrets.CONDA_PYTORCHBOT_TOKEN_TEST }} uses: ./.github/workflows/_binary-upload.yml - libtorch-rocm6_2-shared-with-deps-cxx11-abi-build: + libtorch-rocm6_2_4-shared-with-deps-cxx11-abi-build: if: ${{ github.repository_owner == 'pytorch' }} uses: ./.github/workflows/_binary-build-linux.yml needs: get-label-type @@ -443,21 +443,21 @@ jobs: PACKAGE_TYPE: libtorch # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/libtorch-cxx11-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/libtorch-cxx11-builder:rocm6.2.4-main LIBTORCH_VARIANT: shared-with-deps DESIRED_DEVTOOLSET: cxx11-abi runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" - build_name: libtorch-rocm6_2-shared-with-deps-cxx11-abi + build_name: libtorch-rocm6_2_4-shared-with-deps-cxx11-abi build_environment: linux-binary-libtorch-cxx11-abi secrets: github-token: ${{ secrets.GITHUB_TOKEN }} - libtorch-rocm6_2-shared-with-deps-cxx11-abi-test: # Testing + libtorch-rocm6_2_4-shared-with-deps-cxx11-abi-test: # Testing if: ${{ github.repository_owner == 'pytorch' }} needs: - - libtorch-rocm6_2-shared-with-deps-cxx11-abi-build + - libtorch-rocm6_2_4-shared-with-deps-cxx11-abi-build - get-label-type runs-on: linux.rocm.gpu timeout-minutes: 240 @@ -467,11 +467,11 @@ jobs: PACKAGE_TYPE: libtorch # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/libtorch-cxx11-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/libtorch-cxx11-builder:rocm6.2.4-main LIBTORCH_VARIANT: shared-with-deps DESIRED_DEVTOOLSET: cxx11-abi steps: @@ -480,7 +480,7 @@ jobs: - uses: actions/download-artifact@v4.1.7 name: Download Build Artifacts with: - name: libtorch-rocm6_2-shared-with-deps-cxx11-abi + name: libtorch-rocm6_2_4-shared-with-deps-cxx11-abi path: "${{ runner.temp }}/artifacts/" - name: Checkout PyTorch uses: malfet/checkout@silent-checkout @@ -513,30 +513,30 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/libtorch-cxx11-builder:rocm6.2-main + docker-image: pytorch/libtorch-cxx11-builder:rocm6.2.4-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm uses: ./.github/actions/teardown-rocm - libtorch-rocm6_2-shared-with-deps-cxx11-abi-upload: # Uploading + libtorch-rocm6_2_4-shared-with-deps-cxx11-abi-upload: # Uploading if: ${{ github.repository_owner == 'pytorch' }} permissions: id-token: write contents: read - needs: libtorch-rocm6_2-shared-with-deps-cxx11-abi-test + needs: libtorch-rocm6_2_4-shared-with-deps-cxx11-abi-test with: PYTORCH_ROOT: /pytorch BUILDER_ROOT: /builder PACKAGE_TYPE: libtorch # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/libtorch-cxx11-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/libtorch-cxx11-builder:rocm6.2.4-main LIBTORCH_VARIANT: shared-with-deps DESIRED_DEVTOOLSET: cxx11-abi - build_name: libtorch-rocm6_2-shared-with-deps-cxx11-abi + build_name: libtorch-rocm6_2_4-shared-with-deps-cxx11-abi secrets: github-token: ${{ secrets.GITHUB_TOKEN }} conda-pytorchbot-token: ${{ secrets.CONDA_PYTORCHBOT_TOKEN }} diff --git a/.github/workflows/generated-linux-binary-libtorch-pre-cxx11-nightly.yml b/.github/workflows/generated-linux-binary-libtorch-pre-cxx11-nightly.yml index f4a8964e7865..d20a6f36506c 100644 --- a/.github/workflows/generated-linux-binary-libtorch-pre-cxx11-nightly.yml +++ b/.github/workflows/generated-linux-binary-libtorch-pre-cxx11-nightly.yml @@ -322,223 +322,3 @@ jobs: conda-pytorchbot-token: ${{ secrets.CONDA_PYTORCHBOT_TOKEN }} conda-pytorchbot-token-test: ${{ secrets.CONDA_PYTORCHBOT_TOKEN_TEST }} uses: ./.github/workflows/_binary-upload.yml - - libtorch-rocm6_1-shared-with-deps-pre-cxx11-build: - if: ${{ github.repository_owner == 'pytorch' }} - uses: ./.github/workflows/_binary-build-linux.yml - needs: get-label-type - with: - PYTORCH_ROOT: /pytorch - BUILDER_ROOT: /builder - PACKAGE_TYPE: libtorch - # TODO: This is a legacy variable that we eventually want to get rid of in - # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.1 - GPU_ARCH_VERSION: 6.1 - GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main - LIBTORCH_VARIANT: shared-with-deps - DESIRED_DEVTOOLSET: pre-cxx11 - runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" - build_name: libtorch-rocm6_1-shared-with-deps-pre-cxx11 - build_environment: linux-binary-libtorch-pre-cxx11 - secrets: - github-token: ${{ secrets.GITHUB_TOKEN }} - libtorch-rocm6_1-shared-with-deps-pre-cxx11-test: # Testing - if: ${{ github.repository_owner == 'pytorch' }} - needs: - - libtorch-rocm6_1-shared-with-deps-pre-cxx11-build - - get-label-type - runs-on: linux.rocm.gpu - timeout-minutes: 240 - env: - PYTORCH_ROOT: /pytorch - BUILDER_ROOT: /builder - PACKAGE_TYPE: libtorch - # TODO: This is a legacy variable that we eventually want to get rid of in - # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.1 - GPU_ARCH_VERSION: 6.1 - GPU_ARCH_TYPE: rocm - SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main - LIBTORCH_VARIANT: shared-with-deps - DESIRED_DEVTOOLSET: pre-cxx11 - steps: - - name: Setup ROCm - uses: ./.github/actions/setup-rocm - - uses: actions/download-artifact@v4.1.7 - name: Download Build Artifacts - with: - name: libtorch-rocm6_1-shared-with-deps-pre-cxx11 - path: "${{ runner.temp }}/artifacts/" - - name: Checkout PyTorch - uses: malfet/checkout@silent-checkout - with: - ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }} - submodules: recursive - path: pytorch - quiet-checkout: true - - name: Clean PyTorch checkout - run: | - # Remove any artifacts from the previous checkouts - git clean -fxd - working-directory: pytorch - - name: Checkout pytorch/builder - uses: malfet/checkout@silent-checkout - with: - ref: main - submodules: recursive - repository: pytorch/builder - path: builder - quiet-checkout: true - - name: Clean pytorch/builder checkout - run: | - # Remove any artifacts from the previous checkouts - git clean -fxd - working-directory: builder - - name: ROCm set GPU_FLAG - run: | - echo "GPU_FLAG=--device=/dev/mem --device=/dev/kfd --device=/dev/dri --group-add video --group-add daemon" >> "${GITHUB_ENV}" - - name: Pull Docker image - uses: pytorch/test-infra/.github/actions/pull-docker-image@main - with: - docker-image: pytorch/manylinux-builder:rocm6.1-main - - name: Test Pytorch binary - uses: ./pytorch/.github/actions/test-pytorch-binary - - name: Teardown ROCm - uses: ./.github/actions/teardown-rocm - libtorch-rocm6_1-shared-with-deps-pre-cxx11-upload: # Uploading - if: ${{ github.repository_owner == 'pytorch' }} - permissions: - id-token: write - contents: read - needs: libtorch-rocm6_1-shared-with-deps-pre-cxx11-test - with: - PYTORCH_ROOT: /pytorch - BUILDER_ROOT: /builder - PACKAGE_TYPE: libtorch - # TODO: This is a legacy variable that we eventually want to get rid of in - # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.1 - GPU_ARCH_VERSION: 6.1 - GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main - LIBTORCH_VARIANT: shared-with-deps - DESIRED_DEVTOOLSET: pre-cxx11 - build_name: libtorch-rocm6_1-shared-with-deps-pre-cxx11 - secrets: - github-token: ${{ secrets.GITHUB_TOKEN }} - conda-pytorchbot-token: ${{ secrets.CONDA_PYTORCHBOT_TOKEN }} - conda-pytorchbot-token-test: ${{ secrets.CONDA_PYTORCHBOT_TOKEN_TEST }} - uses: ./.github/workflows/_binary-upload.yml - - libtorch-rocm6_2-shared-with-deps-pre-cxx11-build: - if: ${{ github.repository_owner == 'pytorch' }} - uses: ./.github/workflows/_binary-build-linux.yml - needs: get-label-type - with: - PYTORCH_ROOT: /pytorch - BUILDER_ROOT: /builder - PACKAGE_TYPE: libtorch - # TODO: This is a legacy variable that we eventually want to get rid of in - # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 - GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main - LIBTORCH_VARIANT: shared-with-deps - DESIRED_DEVTOOLSET: pre-cxx11 - runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" - build_name: libtorch-rocm6_2-shared-with-deps-pre-cxx11 - build_environment: linux-binary-libtorch-pre-cxx11 - secrets: - github-token: ${{ secrets.GITHUB_TOKEN }} - libtorch-rocm6_2-shared-with-deps-pre-cxx11-test: # Testing - if: ${{ github.repository_owner == 'pytorch' }} - needs: - - libtorch-rocm6_2-shared-with-deps-pre-cxx11-build - - get-label-type - runs-on: linux.rocm.gpu - timeout-minutes: 240 - env: - PYTORCH_ROOT: /pytorch - BUILDER_ROOT: /builder - PACKAGE_TYPE: libtorch - # TODO: This is a legacy variable that we eventually want to get rid of in - # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 - GPU_ARCH_TYPE: rocm - SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main - LIBTORCH_VARIANT: shared-with-deps - DESIRED_DEVTOOLSET: pre-cxx11 - steps: - - name: Setup ROCm - uses: ./.github/actions/setup-rocm - - uses: actions/download-artifact@v4.1.7 - name: Download Build Artifacts - with: - name: libtorch-rocm6_2-shared-with-deps-pre-cxx11 - path: "${{ runner.temp }}/artifacts/" - - name: Checkout PyTorch - uses: malfet/checkout@silent-checkout - with: - ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }} - submodules: recursive - path: pytorch - quiet-checkout: true - - name: Clean PyTorch checkout - run: | - # Remove any artifacts from the previous checkouts - git clean -fxd - working-directory: pytorch - - name: Checkout pytorch/builder - uses: malfet/checkout@silent-checkout - with: - ref: main - submodules: recursive - repository: pytorch/builder - path: builder - quiet-checkout: true - - name: Clean pytorch/builder checkout - run: | - # Remove any artifacts from the previous checkouts - git clean -fxd - working-directory: builder - - name: ROCm set GPU_FLAG - run: | - echo "GPU_FLAG=--device=/dev/mem --device=/dev/kfd --device=/dev/dri --group-add video --group-add daemon" >> "${GITHUB_ENV}" - - name: Pull Docker image - uses: pytorch/test-infra/.github/actions/pull-docker-image@main - with: - docker-image: pytorch/manylinux-builder:rocm6.2-main - - name: Test Pytorch binary - uses: ./pytorch/.github/actions/test-pytorch-binary - - name: Teardown ROCm - uses: ./.github/actions/teardown-rocm - libtorch-rocm6_2-shared-with-deps-pre-cxx11-upload: # Uploading - if: ${{ github.repository_owner == 'pytorch' }} - permissions: - id-token: write - contents: read - needs: libtorch-rocm6_2-shared-with-deps-pre-cxx11-test - with: - PYTORCH_ROOT: /pytorch - BUILDER_ROOT: /builder - PACKAGE_TYPE: libtorch - # TODO: This is a legacy variable that we eventually want to get rid of in - # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 - GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main - LIBTORCH_VARIANT: shared-with-deps - DESIRED_DEVTOOLSET: pre-cxx11 - build_name: libtorch-rocm6_2-shared-with-deps-pre-cxx11 - secrets: - github-token: ${{ secrets.GITHUB_TOKEN }} - conda-pytorchbot-token: ${{ secrets.CONDA_PYTORCHBOT_TOKEN }} - conda-pytorchbot-token-test: ${{ secrets.CONDA_PYTORCHBOT_TOKEN_TEST }} - uses: ./.github/workflows/_binary-upload.yml diff --git a/.github/workflows/generated-linux-binary-manywheel-nightly.yml b/.github/workflows/generated-linux-binary-manywheel-nightly.yml index 75efa0b8f59f..e70b37ff491d 100644 --- a/.github/workflows/generated-linux-binary-manywheel-nightly.yml +++ b/.github/workflows/generated-linux-binary-manywheel-nightly.yml @@ -409,7 +409,8 @@ jobs: DESIRED_CUDA: rocm6.1 GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.9" runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" @@ -434,7 +435,8 @@ jobs: GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.9" steps: @@ -476,7 +478,7 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/manylinux-builder:rocm6.1-main + docker-image: pytorch/manylinux2_28-builder:rocm6.1-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm @@ -496,7 +498,8 @@ jobs: DESIRED_CUDA: rocm6.1 GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.9" build_name: manywheel-py3_9-rocm6_1 @@ -506,7 +509,7 @@ jobs: conda-pytorchbot-token-test: ${{ secrets.CONDA_PYTORCHBOT_TOKEN_TEST }} uses: ./.github/workflows/_binary-upload.yml - manywheel-py3_9-rocm6_2-build: + manywheel-py3_9-rocm6_2_4-build: if: ${{ github.repository_owner == 'pytorch' }} uses: ./.github/workflows/_binary-build-linux.yml needs: get-label-type @@ -516,21 +519,22 @@ jobs: PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.9" runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" - build_name: manywheel-py3_9-rocm6_2 + build_name: manywheel-py3_9-rocm6_2_4 build_environment: linux-binary-manywheel secrets: github-token: ${{ secrets.GITHUB_TOKEN }} - manywheel-py3_9-rocm6_2-test: # Testing + manywheel-py3_9-rocm6_2_4-test: # Testing if: ${{ github.repository_owner == 'pytorch' }} needs: - - manywheel-py3_9-rocm6_2-build + - manywheel-py3_9-rocm6_2_4-build - get-label-type runs-on: linux.rocm.gpu timeout-minutes: 240 @@ -540,11 +544,12 @@ jobs: PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.9" steps: @@ -553,7 +558,7 @@ jobs: - uses: actions/download-artifact@v4.1.7 name: Download Build Artifacts with: - name: manywheel-py3_9-rocm6_2 + name: manywheel-py3_9-rocm6_2_4 path: "${{ runner.temp }}/artifacts/" - name: Checkout PyTorch uses: malfet/checkout@silent-checkout @@ -586,30 +591,31 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/manylinux-builder:rocm6.2-main + docker-image: pytorch/manylinux2_28-builder:rocm6.2.4-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm uses: ./.github/actions/teardown-rocm - manywheel-py3_9-rocm6_2-upload: # Uploading + manywheel-py3_9-rocm6_2_4-upload: # Uploading if: ${{ github.repository_owner == 'pytorch' }} permissions: id-token: write contents: read - needs: manywheel-py3_9-rocm6_2-test + needs: manywheel-py3_9-rocm6_2_4-test with: PYTORCH_ROOT: /pytorch BUILDER_ROOT: /builder PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.9" - build_name: manywheel-py3_9-rocm6_2 + build_name: manywheel-py3_9-rocm6_2_4 secrets: github-token: ${{ secrets.GITHUB_TOKEN }} conda-pytorchbot-token: ${{ secrets.CONDA_PYTORCHBOT_TOKEN }} @@ -1096,7 +1102,8 @@ jobs: DESIRED_CUDA: rocm6.1 GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.10" runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" @@ -1121,7 +1128,8 @@ jobs: GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.10" steps: @@ -1163,7 +1171,7 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/manylinux-builder:rocm6.1-main + docker-image: pytorch/manylinux2_28-builder:rocm6.1-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm @@ -1183,7 +1191,8 @@ jobs: DESIRED_CUDA: rocm6.1 GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.10" build_name: manywheel-py3_10-rocm6_1 @@ -1193,7 +1202,7 @@ jobs: conda-pytorchbot-token-test: ${{ secrets.CONDA_PYTORCHBOT_TOKEN_TEST }} uses: ./.github/workflows/_binary-upload.yml - manywheel-py3_10-rocm6_2-build: + manywheel-py3_10-rocm6_2_4-build: if: ${{ github.repository_owner == 'pytorch' }} uses: ./.github/workflows/_binary-build-linux.yml needs: get-label-type @@ -1203,21 +1212,22 @@ jobs: PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.10" runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" - build_name: manywheel-py3_10-rocm6_2 + build_name: manywheel-py3_10-rocm6_2_4 build_environment: linux-binary-manywheel secrets: github-token: ${{ secrets.GITHUB_TOKEN }} - manywheel-py3_10-rocm6_2-test: # Testing + manywheel-py3_10-rocm6_2_4-test: # Testing if: ${{ github.repository_owner == 'pytorch' }} needs: - - manywheel-py3_10-rocm6_2-build + - manywheel-py3_10-rocm6_2_4-build - get-label-type runs-on: linux.rocm.gpu timeout-minutes: 240 @@ -1227,11 +1237,12 @@ jobs: PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.10" steps: @@ -1240,7 +1251,7 @@ jobs: - uses: actions/download-artifact@v4.1.7 name: Download Build Artifacts with: - name: manywheel-py3_10-rocm6_2 + name: manywheel-py3_10-rocm6_2_4 path: "${{ runner.temp }}/artifacts/" - name: Checkout PyTorch uses: malfet/checkout@silent-checkout @@ -1273,30 +1284,31 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/manylinux-builder:rocm6.2-main + docker-image: pytorch/manylinux2_28-builder:rocm6.2.4-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm uses: ./.github/actions/teardown-rocm - manywheel-py3_10-rocm6_2-upload: # Uploading + manywheel-py3_10-rocm6_2_4-upload: # Uploading if: ${{ github.repository_owner == 'pytorch' }} permissions: id-token: write contents: read - needs: manywheel-py3_10-rocm6_2-test + needs: manywheel-py3_10-rocm6_2_4-test with: PYTORCH_ROOT: /pytorch BUILDER_ROOT: /builder PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.10" - build_name: manywheel-py3_10-rocm6_2 + build_name: manywheel-py3_10-rocm6_2_4 secrets: github-token: ${{ secrets.GITHUB_TOKEN }} conda-pytorchbot-token: ${{ secrets.CONDA_PYTORCHBOT_TOKEN }} @@ -1853,7 +1865,8 @@ jobs: DESIRED_CUDA: rocm6.1 GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.11" runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" @@ -1878,7 +1891,8 @@ jobs: GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.11" steps: @@ -1920,7 +1934,7 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/manylinux-builder:rocm6.1-main + docker-image: pytorch/manylinux2_28-builder:rocm6.1-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm @@ -1940,7 +1954,8 @@ jobs: DESIRED_CUDA: rocm6.1 GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.11" build_name: manywheel-py3_11-rocm6_1 @@ -1950,7 +1965,7 @@ jobs: conda-pytorchbot-token-test: ${{ secrets.CONDA_PYTORCHBOT_TOKEN_TEST }} uses: ./.github/workflows/_binary-upload.yml - manywheel-py3_11-rocm6_2-build: + manywheel-py3_11-rocm6_2_4-build: if: ${{ github.repository_owner == 'pytorch' }} uses: ./.github/workflows/_binary-build-linux.yml needs: get-label-type @@ -1960,21 +1975,22 @@ jobs: PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.11" runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" - build_name: manywheel-py3_11-rocm6_2 + build_name: manywheel-py3_11-rocm6_2_4 build_environment: linux-binary-manywheel secrets: github-token: ${{ secrets.GITHUB_TOKEN }} - manywheel-py3_11-rocm6_2-test: # Testing + manywheel-py3_11-rocm6_2_4-test: # Testing if: ${{ github.repository_owner == 'pytorch' }} needs: - - manywheel-py3_11-rocm6_2-build + - manywheel-py3_11-rocm6_2_4-build - get-label-type runs-on: linux.rocm.gpu timeout-minutes: 240 @@ -1984,11 +2000,12 @@ jobs: PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.11" steps: @@ -1997,7 +2014,7 @@ jobs: - uses: actions/download-artifact@v4.1.7 name: Download Build Artifacts with: - name: manywheel-py3_11-rocm6_2 + name: manywheel-py3_11-rocm6_2_4 path: "${{ runner.temp }}/artifacts/" - name: Checkout PyTorch uses: malfet/checkout@silent-checkout @@ -2030,30 +2047,31 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/manylinux-builder:rocm6.2-main + docker-image: pytorch/manylinux2_28-builder:rocm6.2.4-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm uses: ./.github/actions/teardown-rocm - manywheel-py3_11-rocm6_2-upload: # Uploading + manywheel-py3_11-rocm6_2_4-upload: # Uploading if: ${{ github.repository_owner == 'pytorch' }} permissions: id-token: write contents: read - needs: manywheel-py3_11-rocm6_2-test + needs: manywheel-py3_11-rocm6_2_4-test with: PYTORCH_ROOT: /pytorch BUILDER_ROOT: /builder PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.11" - build_name: manywheel-py3_11-rocm6_2 + build_name: manywheel-py3_11-rocm6_2_4 secrets: github-token: ${{ secrets.GITHUB_TOKEN }} conda-pytorchbot-token: ${{ secrets.CONDA_PYTORCHBOT_TOKEN }} @@ -2540,7 +2558,8 @@ jobs: DESIRED_CUDA: rocm6.1 GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.12" runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" @@ -2565,7 +2584,8 @@ jobs: GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.12" steps: @@ -2607,7 +2627,7 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/manylinux-builder:rocm6.1-main + docker-image: pytorch/manylinux2_28-builder:rocm6.1-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm @@ -2627,7 +2647,8 @@ jobs: DESIRED_CUDA: rocm6.1 GPU_ARCH_VERSION: 6.1 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.1-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.1-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.12" build_name: manywheel-py3_12-rocm6_1 @@ -2637,7 +2658,7 @@ jobs: conda-pytorchbot-token-test: ${{ secrets.CONDA_PYTORCHBOT_TOKEN_TEST }} uses: ./.github/workflows/_binary-upload.yml - manywheel-py3_12-rocm6_2-build: + manywheel-py3_12-rocm6_2_4-build: if: ${{ github.repository_owner == 'pytorch' }} uses: ./.github/workflows/_binary-build-linux.yml needs: get-label-type @@ -2647,21 +2668,22 @@ jobs: PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.12" runner_prefix: "${{ needs.get-label-type.outputs.label-type }}" - build_name: manywheel-py3_12-rocm6_2 + build_name: manywheel-py3_12-rocm6_2_4 build_environment: linux-binary-manywheel secrets: github-token: ${{ secrets.GITHUB_TOKEN }} - manywheel-py3_12-rocm6_2-test: # Testing + manywheel-py3_12-rocm6_2_4-test: # Testing if: ${{ github.repository_owner == 'pytorch' }} needs: - - manywheel-py3_12-rocm6_2-build + - manywheel-py3_12-rocm6_2_4-build - get-label-type runs-on: linux.rocm.gpu timeout-minutes: 240 @@ -2671,11 +2693,12 @@ jobs: PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm SKIP_ALL_TESTS: 1 - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.12" steps: @@ -2684,7 +2707,7 @@ jobs: - uses: actions/download-artifact@v4.1.7 name: Download Build Artifacts with: - name: manywheel-py3_12-rocm6_2 + name: manywheel-py3_12-rocm6_2_4 path: "${{ runner.temp }}/artifacts/" - name: Checkout PyTorch uses: malfet/checkout@silent-checkout @@ -2717,30 +2740,31 @@ jobs: - name: Pull Docker image uses: pytorch/test-infra/.github/actions/pull-docker-image@main with: - docker-image: pytorch/manylinux-builder:rocm6.2-main + docker-image: pytorch/manylinux2_28-builder:rocm6.2.4-main - name: Test Pytorch binary uses: ./pytorch/.github/actions/test-pytorch-binary - name: Teardown ROCm uses: ./.github/actions/teardown-rocm - manywheel-py3_12-rocm6_2-upload: # Uploading + manywheel-py3_12-rocm6_2_4-upload: # Uploading if: ${{ github.repository_owner == 'pytorch' }} permissions: id-token: write contents: read - needs: manywheel-py3_12-rocm6_2-test + needs: manywheel-py3_12-rocm6_2_4-test with: PYTORCH_ROOT: /pytorch BUILDER_ROOT: /builder PACKAGE_TYPE: manywheel # TODO: This is a legacy variable that we eventually want to get rid of in # favor of GPU_ARCH_VERSION - DESIRED_CUDA: rocm6.2 - GPU_ARCH_VERSION: 6.2 + DESIRED_CUDA: rocm6.2.4 + GPU_ARCH_VERSION: 6.2.4 GPU_ARCH_TYPE: rocm - DOCKER_IMAGE: pytorch/manylinux-builder:rocm6.2-main + DOCKER_IMAGE: pytorch/manylinux2_28-builder:rocm6.2.4-main + DESIRED_DEVTOOLSET: cxx11-abi use_split_build: False DESIRED_PYTHON: "3.12" - build_name: manywheel-py3_12-rocm6_2 + build_name: manywheel-py3_12-rocm6_2_4 secrets: github-token: ${{ secrets.GITHUB_TOKEN }} conda-pytorchbot-token: ${{ secrets.CONDA_PYTORCHBOT_TOKEN }}