# TensorFlow Bazel configuration file. # This file tries to group and simplify build options for TensorFlow # # ----CONFIG OPTIONS---- # Android options: # android: # android_arm: # android_arm64: # android_x86: # android_x86_64: # # iOS options: # ios: # ios_armv7: # ios_arm64: # ios_i386: # ios_x86_64: # ios_fat: # # Macosx options # darwin_arm64: # # Compiler options: # cuda_clang: Use clang when building CUDA code. # c++17: Build with C++17 options (links with libc++) # c++1z: Build with C++17 options (links with libc++) # c++17_gcc: Build with C++17 options (links with stdlibc++) # c++1z_gcc: Build with C++17 options (links with stdlibc++) # avx_linux: Build with avx instruction set on linux. # avx2_linux: Build with avx2 instruction set on linux. # native_arch_linux: Build with instruction sets available to the host machine on linux # avx_win: Build with avx instruction set on windows # avx2_win: Build with avx2 instruction set on windows # # Other build options: # short_logs: Only log errors during build, skip warnings. # verbose_logs: Show all compiler warnings during build. # monolithic: Build all TF C++ code into a single shared object. # dynamic_kernels: Try to link all kernels dynamically (experimental). # libc++: Link against libc++ instead of stdlibc++ # asan: Build with the clang address sanitizer # msan: Build with the clang memory sanitizer # ubsan: Build with the clang undefined behavior sanitizer # dbg: Build with debug info # # # TF version options; # v1: Build TF V1 (without contrib) # v2: Build TF v2 # # Feature and Third party library support options: # xla: Build TF with XLA # tpu: Build TF with TPU support # cuda: Build with full cuda support. # rocm: Build with AMD GPU support (rocm). # mkl: Enable full mkl support. # tensorrt: Enable Tensorrt support. # numa: Enable numa using hwloc. # noaws: Disable AWS S3 storage support # nogcp: Disable GCS support. # nohdfs: Disable hadoop hdfs support. # nonccl: Disable nccl support. # # # Remote build execution options (only configured to work with TF team projects for now.) # rbe: General RBE options shared by all flavors. # rbe_linux: General RBE options used on all linux builds. # rbe_win: General RBE options used on all windows builds. # # rbe_cpu_linux: RBE options to build with only CPU support. # rbe_linux_cuda_nvcc_py*: RBE options to build with GPU support using nvcc. # # rbe_linux_py2: Linux Python 2 RBE config. # rbe_linux_py3: Linux Python 3 RBE config # # rbe_win_py37: Windows Python 3.7 RBE config # rbe_win_py38: Windows Python 3.8 RBE config # # tensorflow_testing_rbe_linux: RBE options to use RBE with tensorflow-testing project on linux # tensorflow_testing_rbe_win: RBE options to use RBE with tensorflow-testing project on windows # # Embedded Linux options (experimental and only tested with TFLite build yet) # elinux: General Embedded Linux options shared by all flavors. # elinux_aarch64: Embedded Linux options for aarch64 (ARM64) CPU support. # elinux_armhf: Embedded Linux options for armhf (ARMv7) CPU support. # # Release build options (for all operating systems) # release_base: Common options for all builds on all operating systems. # release_gpu_base: Common options for GPU builds on Linux and Windows. # release_cpu_linux: Toolchain and CUDA options for Linux CPU builds. # release_cpu_macos: Toolchain and CUDA options for MacOS CPU builds. # release_gpu_linux: Toolchain and CUDA options for Linux GPU builds. # release_cpu_windows: Toolchain and CUDA options for Windows CPU builds. # release_gpu_windows: Toolchain and CUDA options for Windows GPU builds. # Default build options. These are applied first and unconditionally. # For projects which use TensorFlow as part of a Bazel build process, putting # nothing in a bazelrc will default to a monolithic build. The following line # opts in to modular op registration support by default. build --define framework_shared_object=true # For workaround https://github.com/bazelbuild/bazel/issues/8772 with Bazel >= 0.29.1 build --java_toolchain=@tf_toolchains//toolchains/java:tf_java_toolchain build --host_java_toolchain=@tf_toolchains//toolchains/java:tf_java_toolchain build --define=use_fast_cpp_protos=true build --define=allow_oversize_protos=true build --spawn_strategy=standalone build -c opt # Make Bazel print out all options from rc files. build --announce_rc build --define=grpc_no_ares=true # See https://github.com/bazelbuild/bazel/issues/7362 for information on what # --incompatible_remove_legacy_whole_archive flag does. # This flag is set to true in Bazel 1.0 and newer versions. We tried to migrate # Tensorflow to the default, however test coverage wasn't enough to catch the # errors. # There is ongoing work on Bazel team's side to provide support for transitive # shared libraries. As part of migrating to transitive shared libraries, we # hope to provide a better mechanism for control over symbol exporting, and # then tackle this issue again. # # TODO: Remove this line once TF doesn't depend on Bazel wrapping all library # archives in -whole_archive -no_whole_archive. build --noincompatible_remove_legacy_whole_archive build --enable_platform_specific_config # Enable XLA support by default. build --define=with_xla_support=true build --config=short_logs build --config=v2 # Disable AWS/HDFS support by default build --define=no_aws_support=true build --define=no_hdfs_support=true # Default options should come above this line. # Allow builds using libc++ as a linker library # This is mostly for OSSFuzz, so we also pass in the flags from environment to clean build file build:libc++ --action_env=CC build:libc++ --action_env=CXX build:libc++ --action_env=CXXFLAGS=-stdlib=libc++ build:libc++ --action_env=PATH build:libc++ --define force_libcpp=enabled build:libc++ --linkopt -fuse-ld=lld # Android configs. Bazel needs to have --cpu and --fat_apk_cpu both set to the # target CPU to build transient dependencies correctly. See # https://docs.bazel.build/versions/master/user-manual.html#flag--fat_apk_cpu build:android --crosstool_top=//external:android/crosstool build:android --host_crosstool_top=@bazel_tools//tools/cpp:toolchain build:android_arm --config=android build:android_arm --cpu=armeabi-v7a build:android_arm --fat_apk_cpu=armeabi-v7a build:android_arm64 --config=android build:android_arm64 --cpu=arm64-v8a build:android_arm64 --fat_apk_cpu=arm64-v8a build:android_x86 --config=android build:android_x86 --cpu=x86 build:android_x86 --fat_apk_cpu=x86 build:android_x86_64 --config=android build:android_x86_64 --cpu=x86_64 build:android_x86_64 --fat_apk_cpu=x86_64 # Sets the default Apple platform to macOS. build:macos --apple_platform_type=macos # gRPC on MacOS requires this #define build:macos --copt=-DGRPC_BAZEL_BUILD # Settings for MacOS on ARM CPUs. build:macos_arm64 --cpu=darwin_arm64 # iOS configs for each architecture and the fat binary builds. build:ios --apple_platform_type=ios build:ios --apple_bitcode=embedded --copt=-fembed-bitcode build:ios --copt=-Wno-c++11-narrowing build:ios_armv7 --config=ios build:ios_armv7 --cpu=ios_armv7 build:ios_arm64 --config=ios build:ios_arm64 --cpu=ios_arm64 build:ios_i386 --config=ios build:ios_i386 --cpu=ios_i386 build:ios_x86_64 --config=ios build:ios_x86_64 --cpu=ios_x86_64 build:ios_fat --config=ios build:ios_fat --ios_multi_cpus=armv7,arm64,i386,x86_64 # Config to use a mostly-static build and disable modular op registration # support (this will revert to loading TensorFlow with RTLD_GLOBAL in Python). # By default, TensorFlow will build with a dependence on # //tensorflow:libtensorflow_framework.so. build:monolithic --define framework_shared_object=false # Please note that MKL on MacOS or windows is still not supported. # If you would like to use a local MKL instead of downloading, please set the # environment variable "TF_MKL_ROOT" every time before build. build:mkl --define=build_with_mkl=true --define=enable_mkl=true build:mkl --define=tensorflow_mkldnn_contraction_kernel=0 build:mkl --define=build_with_openmp=true build:mkl -c opt # config to build OneDNN backend with a user specified threadpool. build:mkl_threadpool --define=build_with_mkl=true --define=enable_mkl=true build:mkl_threadpool --define=tensorflow_mkldnn_contraction_kernel=0 build:mkl_threadpool --define=build_with_mkl_opensource=true build:mkl_threadpool -c opt # Config setting to build oneDNN with Compute Library for the Arm Architecture (ACL). # This build is for the inference regime only. build:mkl_aarch64 --define=build_with_mkl_aarch64=true --define=enable_mkl=true build:mkl_aarch64 --define=tensorflow_mkldnn_contraction_kernel=0 build:mkl_aarch64 --define=build_with_mkl_opensource=true build:mkl_aarch64 --define=build_with_openmp=true build:mkl_aarch64 -c opt # This config refers to building CUDA op kernels with nvcc. build:cuda --repo_env TF_NEED_CUDA=1 build:cuda --crosstool_top=@local_config_cuda//crosstool:toolchain build:cuda --@local_config_cuda//:enable_cuda # This config refers to building CUDA op kernels with clang. build:cuda_clang --config=cuda build:cuda_clang --repo_env TF_CUDA_CLANG=1 build:cuda_clang --@local_config_cuda//:cuda_compiler=clang # Debug config build:dbg -c dbg # Only include debug info for files under tensorflow/, excluding kernels, to # reduce the size of the debug info in the binary. This is because if the debug # sections in the ELF binary are too large, errors can occur. See # https://github.com/tensorflow/tensorflow/issues/48919. # Users can still include debug info for a specific kernel, e.g. with: # --config=dbg --per_file_copt=+tensorflow/core/kernels/identity_op.*@-g build:dbg --per_file_copt=+.*,-tensorflow.*@-g0 build:dbg --per_file_copt=+tensorflow/core/kernels.*@-g0 # for now, disable arm_neon. see: https://github.com/tensorflow/tensorflow/issues/33360 build:dbg --cxxopt -DTF_LITE_DISABLE_X86_NEON # AWS SDK must be compiled in release mode. see: https://github.com/tensorflow/tensorflow/issues/37498 build:dbg --copt -DDEBUG_BUILD # Config to build TPU backend build:tpu --define=with_tpu_support=true build:tensorrt --repo_env TF_NEED_TENSORRT=1 build:rocm --crosstool_top=@local_config_rocm//crosstool:toolchain build:rocm --define=using_rocm_hipcc=true build:rocm --repo_env TF_NEED_ROCM=1 # Options extracted from configure script build:numa --define=with_numa_support=true # Options to disable default on features build:noaws --define=no_aws_support=true build:nogcp --define=no_gcp_support=true build:nohdfs --define=no_hdfs_support=true build:nonccl --define=no_nccl_support=true build:stackdriver_support --define=stackdriver_support=true # Modular TF build options build:dynamic_kernels --define=dynamic_loaded_kernels=true build:dynamic_kernels --copt=-DAUTOLOAD_DYNAMIC_KERNELS # Build TF with C++ 17 features. build:c++17 --cxxopt=-std=c++1z build:c++17 --cxxopt=-stdlib=libc++ build:c++1z --config=c++17 build:c++17_gcc --cxxopt=-std=c++1z build:c++1z_gcc --config=c++17_gcc # Don't trigger --config= when cross-compiling. build:android --noenable_platform_specific_config build:ios --noenable_platform_specific_config # Suppress C++ compiler warnings, otherwise build logs become 10s of MBs. build:android --copt=-w build:ios --copt=-w build:linux --copt=-w build:linux --host_copt=-w build:macos --copt=-w build:windows --copt=/W0 # Tensorflow uses M_* math constants that only get defined by MSVC headers if # _USE_MATH_DEFINES is defined. build:windows --copt=/D_USE_MATH_DEFINES build:windows --host_copt=/D_USE_MATH_DEFINES # Default paths for TF_SYSTEM_LIBS build:linux --define=PREFIX=/usr build:linux --define=LIBDIR=$(PREFIX)/lib build:linux --define=INCLUDEDIR=$(PREFIX)/include build:linux --define=PROTOBUF_INCLUDE_PATH=$(PREFIX)/include build:macos --define=PREFIX=/usr build:macos --define=LIBDIR=$(PREFIX)/lib build:macos --define=INCLUDEDIR=$(PREFIX)/include build:macos --define=PROTOBUF_INCLUDE_PATH=$(PREFIX)/include # TF_SYSTEM_LIBS do not work on windows. # By default, build TF in C++ 14 mode. build:android --cxxopt=-std=c++14 build:android --host_cxxopt=-std=c++14 build:ios --cxxopt=-std=c++14 build:ios --host_cxxopt=-std=c++14 build:linux --cxxopt=-std=c++14 build:linux --host_cxxopt=-std=c++14 build:macos --cxxopt=-std=c++14 build:macos --host_cxxopt=-std=c++14 build:windows --cxxopt=/std:c++14 build:windows --host_cxxopt=/std:c++14 # On windows, we still link everything into a single DLL. build:windows --config=monolithic # On linux, we dynamically link small amount of kernels build:linux --config=dynamic_kernels # Make sure to include as little of windows.h as possible build:windows --copt=-DWIN32_LEAN_AND_MEAN build:windows --host_copt=-DWIN32_LEAN_AND_MEAN build:windows --copt=-DNOGDI build:windows --host_copt=-DNOGDI # MSVC (Windows): Standards-conformant preprocessor mode # See https://docs.microsoft.com/en-us/cpp/preprocessor/preprocessor-experimental-overview build:windows --copt=/experimental:preprocessor build:windows --host_copt=/experimental:preprocessor # Misc build options we need for windows. build:windows --linkopt=/DEBUG build:windows --host_linkopt=/DEBUG build:windows --linkopt=/OPT:REF build:windows --host_linkopt=/OPT:REF build:windows --linkopt=/OPT:ICF build:windows --host_linkopt=/OPT:ICF # Verbose failure logs when something goes wrong build:windows --verbose_failures # On windows, we never cross compile build:windows --distinct_host_configuration=false # On linux, don't cross compile by default build:linux --distinct_host_configuration=false # Do not risk cache corruption. See: # https://github.com/bazelbuild/bazel/issues/3360 build:linux --experimental_guard_against_concurrent_changes # Configure short or long logs build:short_logs --output_filter=DONT_MATCH_ANYTHING build:verbose_logs --output_filter= # Instruction set optimizations # TODO(gunan): Create a feature in toolchains for avx/avx2 to # avoid having to define linux/win separately. build:avx_linux --copt=-mavx build:avx_linux --host_copt=-mavx build:avx2_linux --copt=-mavx2 build:native_arch_linux --copt=-march=native build:avx_win --copt=/arch=AVX build:avx2_win --copt=/arch=AVX2 # Options to build TensorFlow 1.x or 2.x. build:v1 --define=tf_api_version=1 --action_env=TF2_BEHAVIOR=0 build:v2 --define=tf_api_version=2 --action_env=TF2_BEHAVIOR=1 # Disable XLA on mobile. build:xla --define=with_xla_support=true # TODO: remove, it's on by default. build:android --define=with_xla_support=false build:ios --define=with_xla_support=false # BEGIN TF REMOTE BUILD EXECUTION OPTIONS # Options when using remote execution # WARNING: THESE OPTIONS WONT WORK IF YOU DO NOT HAVE PROPER AUTHENTICATION AND PERMISSIONS # Flag to enable remote config common --experimental_repo_remote_exec build:rbe --repo_env=BAZEL_DO_NOT_DETECT_CPP_TOOLCHAIN=1 build:rbe --google_default_credentials build:rbe --bes_backend=buildeventservice.googleapis.com build:rbe --bes_results_url="https://source.cloud.google.com/results/invocations" build:rbe --bes_timeout=600s build:rbe --define=EXECUTOR=remote build:rbe --distinct_host_configuration=false build:rbe --flaky_test_attempts=3 build:rbe --jobs=200 build:rbe --remote_executor=grpcs://remotebuildexecution.googleapis.com build:rbe --remote_timeout=3600 build:rbe --spawn_strategy=remote,worker,standalone,local test:rbe --test_env=USER=anon # Attempt to minimize the amount of data transfer between bazel and the remote # workers: build:rbe --remote_download_toplevel build:rbe_linux --config=rbe build:rbe_linux --action_env=PATH="/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/local/go/bin" build:rbe_linux --host_javabase=@bazel_toolchains//configs/ubuntu16_04_clang/1.1:jdk8 build:rbe_linux --javabase=@bazel_toolchains//configs/ubuntu16_04_clang/1.1:jdk8 build:rbe_linux --host_java_toolchain=@bazel_tools//tools/jdk:toolchain_hostjdk8 build:rbe_linux --java_toolchain=@bazel_tools//tools/jdk:toolchain_hostjdk8 # Non-rbe settings we should include because we do not run configure build:rbe_linux --config=avx_linux # TODO(gunan): Check why we need this specified in rbe, but not in other builds. build:rbe_linux --linkopt=-lrt build:rbe_linux --host_linkopt=-lrt build:rbe_linux --linkopt=-lm build:rbe_linux --host_linkopt=-lm # Use the GPU toolchain until the CPU one is ready. # https://github.com/bazelbuild/bazel/issues/13623 build:rbe_cpu_linux --config=rbe_linux build:rbe_cpu_linux --host_crosstool_top="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain" build:rbe_cpu_linux --crosstool_top="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain" build:rbe_cpu_linux --extra_toolchains="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain-linux-x86_64" build:rbe_cpu_linux --extra_execution_platforms="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_cpu_linux --host_platform="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_cpu_linux --platforms="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_linux_cuda_base --config=rbe_linux build:rbe_linux_cuda_base --config=cuda build:rbe_linux_cuda_base --config=tensorrt build:rbe_linux_cuda_base --action_env=TF_CUDA_VERSION=11 build:rbe_linux_cuda_base --action_env=TF_CUDNN_VERSION=8 build:rbe_linux_cuda_base --repo_env=REMOTE_GPU_TESTING=1 # TensorRT 7 for CUDA 11.1 is compatible with CUDA 11.2, but requires # libnvrtc.so.11.1. See https://github.com/NVIDIA/TensorRT/issues/1064. # TODO(b/187962120): Remove when upgrading to TensorRT 8. test:rbe_linux_cuda_base --test_env=LD_LIBRARY_PATH="/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64:/usr/local/cuda-11.1/lib64" build:rbe_linux_cuda11.2_nvcc_base --config=rbe_linux_cuda_base build:rbe_linux_cuda11.2_nvcc_base --host_crosstool_top="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain" build:rbe_linux_cuda11.2_nvcc_base --crosstool_top="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain" build:rbe_linux_cuda11.2_nvcc_base --extra_toolchains="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain-linux-x86_64" build:rbe_linux_cuda11.2_nvcc_base --extra_execution_platforms="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_linux_cuda11.2_nvcc_base --host_platform="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_linux_cuda11.2_nvcc_base --platforms="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_linux_cuda11.2_nvcc_base --repo_env=TF_CUDA_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda" build:rbe_linux_cuda11.2_nvcc_base --repo_env=TF_TENSORRT_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_tensorrt" build:rbe_linux_cuda11.2_nvcc_base --repo_env=TF_NCCL_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_nccl" build:rbe_linux_cuda11.2_nvcc_py3.6 --config=rbe_linux_cuda11.2_nvcc_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.6" build:rbe_linux_cuda11.2_nvcc_py3.7 --config=rbe_linux_cuda11.2_nvcc_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.7" build:rbe_linux_cuda11.2_nvcc_py3.8 --config=rbe_linux_cuda11.2_nvcc_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.8" build:rbe_linux_cuda11.2_nvcc_py3.9 --config=rbe_linux_cuda11.2_nvcc_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.9" # Map default to CUDA 11.2. build:rbe_linux_cuda_nvcc_py36 --config=rbe_linux_cuda11.2_nvcc_py3.6 build:rbe_linux_cuda_nvcc_py37 --config=rbe_linux_cuda11.2_nvcc_py3.7 build:rbe_linux_cuda_nvcc_py38 --config=rbe_linux_cuda11.2_nvcc_py3.8 build:rbe_linux_cuda_nvcc_py39 --config=rbe_linux_cuda11.2_nvcc_py3.9 # Deprecated configs that people might still use. build:rbe_linux_cuda_nvcc --config=rbe_linux_cuda_nvcc_py36 build:rbe_gpu_linux --config=rbe_linux_cuda_nvcc build:rbe_linux_cuda_clang_base --config=rbe_linux_cuda_base build:rbe_linux_cuda_clang_base --repo_env TF_CUDA_CLANG=1 build:rbe_linux_cuda_clang_base --@local_config_cuda//:cuda_compiler=clang build:rbe_linux_cuda_clang_base --crosstool_top="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain" build:rbe_linux_cuda_clang_base --extra_toolchains="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain-linux-x86_64" build:rbe_linux_cuda_clang_base --extra_execution_platforms="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_linux_cuda_clang_base --host_platform="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_linux_cuda_clang_base --platforms="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_platform//:platform" build:rbe_linux_cuda_clang_base --repo_env=TF_CUDA_CONFIG_REPO="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda" build:rbe_linux_cuda_clang_base --repo_env=TF_TENSORRT_CONFIG_REPO="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_tensorrt" build:rbe_linux_cuda_clang_base --repo_env=TF_NCCL_CONFIG_REPO="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_nccl" build:rbe_linux_cuda_clang_py27 --config=rbe_linux_cuda_clang_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python2.7" build:rbe_linux_cuda_clang_py35 --config=rbe_linux_cuda_clang_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.5" build:rbe_linux_cuda_clang_py36 --config=rbe_linux_cuda_clang_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.6" build:rbe_linux_cuda_clang_py37 --config=rbe_linux_cuda_clang_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.7" build:rbe_linux_cuda_clang_py38 --config=rbe_linux_cuda_clang_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-clang_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.8" # ROCm build:rbe_linux_rocm_base --config=rocm build:rbe_linux_rocm_base --config=rbe_linux build:rbe_linux_rocm_base --crosstool_top="@ubuntu18.04-gcc7_manylinux2010-rocm_config_rocm//crosstool:toolchain" build:rbe_linux_rocm_base --extra_toolchains="@ubuntu18.04-gcc7_manylinux2010-rocm_config_rocm//crosstool:toolchain-linux-x86_64" build:rbe_linux_rocm_base --extra_execution_platforms="@ubuntu18.04-gcc7_manylinux2010-rocm_config_platform//:platform" build:rbe_linux_rocm_base --host_platform="@ubuntu18.04-gcc7_manylinux2010-rocm_config_platform//:platform" build:rbe_linux_rocm_base --platforms="@ubuntu18.04-gcc7_manylinux2010-rocm_config_platform//:platform" build:rbe_linux_rocm_base --action_env=TF_ROCM_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-rocm_config_rocm" build:rbe_linux_rocm_py2.7 --config=rbe_linux_rocm_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-rocm_config_python2.7" build:rbe_linux_rocm_py3.5 --config=rbe_linux_rocm_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-rocm_config_python3.5" build:rbe_linux_rocm_py3.6 --config=rbe_linux_rocm_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-rocm_config_python3.6" build:rbe_linux_rocm_py3.7 --config=rbe_linux_rocm_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-rocm_config_python3.7" build:rbe_linux_rocm_py3.8 --config=rbe_linux_rocm_base --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-rocm_config_python3.8" # Linux CPU build:rbe_linux_py3 --config=rbe_linux build:rbe_linux_py3 --python_path="/usr/local/bin/python3.9" build:rbe_linux_py3 --repo_env=TF_PYTHON_CONFIG_REPO="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_python3.9" build:rbe_win --config=rbe build:rbe_win --crosstool_top="@tf_toolchains//toolchains/win/tf_win_06242021:toolchain" build:rbe_win --extra_toolchains="@tf_toolchains//toolchains/win/tf_win_06242021:cc-toolchain-x64_windows" build:rbe_win --host_javabase="@tf_toolchains//toolchains/win:windows_jdk8" build:rbe_win --javabase="@tf_toolchains//toolchains/win:windows_jdk8" build:rbe_win --extra_execution_platforms="@tf_toolchains//toolchains/win:rbe_windows_ltsc2019" build:rbe_win --host_platform="@tf_toolchains//toolchains/win:rbe_windows_ltsc2019" build:rbe_win --platforms="@tf_toolchains//toolchains/win:rbe_windows_ltsc2019" build:rbe_win --shell_executable=C:\\tools\\msys64\\usr\\bin\\bash.exe build:rbe_win --experimental_strict_action_env=true # TODO(gunan): Remove once we use MSVC 2019 with latest patches. build:rbe_win --define=override_eigen_strong_inline=true build:rbe_win --jobs=100 # Don't build the python zip archive in the RBE build. build:rbe_win --remote_download_minimal build:rbe_win --enable_runfiles build:rbe_win --nobuild_python_zip build:rbe_win_py37 --config=rbe build:rbe_win_py37 --repo_env=TF_PYTHON_CONFIG_REPO="@windows_py37_config_python" build:rbe_win_py37 --python_path=C:\\Python37\\python.exe build:rbe_win_py38 --config=rbe build:rbe_win_py38 --repo_env=PYTHON_BIN_PATH=C:\\Python38\\python.exe build:rbe_win_py38 --repo_env=PYTHON_LIB_PATH=C:\\Python38\\lib\\site-packages build:rbe_win_py38 --repo_env=TF_PYTHON_CONFIG_REPO=@tf_toolchains//toolchains/win_1803/py38 build:rbe_win_py38 --python_path=C:\\Python38\\python.exe # These you may need to change for your own GCP project. build:tensorflow_testing_rbe --project_id=tensorflow-testing common:tensorflow_testing_rbe_linux --remote_instance_name=projects/tensorflow-testing/instances/default_instance build:tensorflow_testing_rbe_linux --config=tensorflow_testing_rbe common:tensorflow_testing_rbe_win --remote_instance_name=projects/tensorflow-testing/instances/windows build:tensorflow_testing_rbe_win --config=tensorflow_testing_rbe # TFLite build configs for generic embedded Linux build:elinux --crosstool_top=@local_config_embedded_arm//:toolchain build:elinux --host_crosstool_top=@bazel_tools//tools/cpp:toolchain build:elinux_aarch64 --config=elinux build:elinux_aarch64 --cpu=aarch64 build:elinux_aarch64 --distinct_host_configuration=true build:elinux_armhf --config=elinux build:elinux_armhf --cpu=armhf build:elinux_armhf --distinct_host_configuration=true # END TF REMOTE BUILD EXECUTION OPTIONS # Config-specific options should come above this line. # Load rc file written by ./configure. try-import %workspace%/.tf_configure.bazelrc # Load rc file with user-specific options. try-import %workspace%/.bazelrc.user # Here are bazelrc configs for release builds build:release_base --config=v2 build:release_base --distinct_host_configuration=false test:release_base --flaky_test_attempts=3 test:release_base --test_size_filters=small,medium build:release_cpu_linux --config=release_base build:release_cpu_linux --config=avx_linux build:release_cpu_linux --crosstool_top="@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain" test:release_cpu_linux --test_env=LD_LIBRARY_PATH build:release_cpu_macos --config=release_base build:release_cpu_macos --config=avx_linux build:release_gpu_base --config=cuda build:release_gpu_base --action_env=TF_CUDA_VERSION="11" build:release_gpu_base --action_env=TF_CUDNN_VERSION="8" build:release_gpu_base --repo_env=TF_CUDA_COMPUTE_CAPABILITIES="sm_35,sm_50,sm_60,sm_70,sm_75,compute_80" build:release_gpu_linux --config=release_cpu_linux build:release_gpu_linux --config=release_gpu_base build:release_gpu_linux --config=tensorrt build:release_gpu_linux --action_env=CUDA_TOOLKIT_PATH="/usr/local/cuda-11.2" build:release_gpu_linux --action_env=LD_LIBRARY_PATH="/usr/local/cuda:/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64:/usr/local/tensorrt/lib" build:release_gpu_linux --action_env=GCC_HOST_COMPILER_PATH="/dt7/usr/bin/gcc" build:release_gpu_linux --crosstool_top=@ubuntu18.04-gcc7_manylinux2010-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain build:release_cpu_windows --config=release_base build:release_cpu_windows --config=avx_win build:release_cpu_windows --define=no_tensorflow_py_deps=true # First available in VS 16.4. Speeds Windows compile times by a lot. See # https://groups.google.com/a/tensorflow.org/d/topic/build/SsW98Eo7l3o/discussion build:release_cpu_windows --copt=/d2ReducedOptimizeHugeFunctions --host_copt=/d2ReducedOptimizeHugeFunctions build:release_gpu_windows --config=release_cpu_windows build:release_gpu_windows --config=release_gpu_base # Address sanitizer # CC=clang bazel build --config asan build:asan --strip=never build:asan --copt -fsanitize=address build:asan --copt -DADDRESS_SANITIZER build:asan --copt -g build:asan --copt -O3 build:asan --copt -fno-omit-frame-pointer build:asan --linkopt -fsanitize=address # Memory sanitizer # CC=clang bazel build --config msan build:msan --strip=never build:msan --copt -fsanitize=memory build:msan --copt -DMEMORY_SANITIZER build:msan --copt -g build:msan --copt -O3 build:msan --copt -fno-omit-frame-pointer build:msan --linkopt -fsanitize=memory # Undefined Behavior Sanitizer # CC=clang bazel build --config ubsan build:ubsan --strip=never build:ubsan --copt -fsanitize=undefined build:ubsan --copt -DUNDEFINED_BEHAVIOR_SANITIZER build:ubsan --copt -g build:ubsan --copt -O3 build:ubsan --copt -fno-omit-frame-pointer build:ubsan --linkopt -fsanitize=undefined build:ubsan --linkopt -lubsan # Exclude TFRT integration for anything but Linux. build:android --config=no_tfrt build:macos --config=no_tfrt build:windows --config=no_tfrt build:rocm --config=no_tfrt build:no_tfrt --deleted_packages=tensorflow/compiler/mlir/tfrt,tensorflow/compiler/mlir/tfrt/benchmarks,tensorflow/compiler/mlir/tfrt/jit/python_binding,tensorflow/compiler/mlir/tfrt/python_tests,tensorflow/compiler/mlir/tfrt/tests,tensorflow/compiler/mlir/tfrt/tests/saved_model,tensorflow/compiler/mlir/tfrt/transforms/lhlo_gpu_to_tfrt_gpu,tensorflow/core/runtime_fallback,tensorflow/core/runtime_fallback/conversion,tensorflow/core/runtime_fallback/kernel,tensorflow/core/runtime_fallback/opdefs,tensorflow/core/runtime_fallback/runtime,tensorflow/core/runtime_fallback/util,tensorflow/core/tfrt/common,tensorflow/core/tfrt/eager,tensorflow/core/tfrt/eager/backends/cpu,tensorflow/core/tfrt/eager/backends/gpu,tensorflow/core/tfrt/eager/core_runtime,tensorflow/core/tfrt/eager/cpp_tests/core_runtime,tensorflow/core/tfrt/fallback,tensorflow/core/tfrt/gpu,tensorflow/core/tfrt/run_handler_thread_pool,tensorflow/core/tfrt/runtime,tensorflow/core/tfrt/saved_model,tensorflow/core/tfrt/saved_model/tests,tensorflow/core/tfrt/tpu,tensorflow/core/tfrt/utils # Experimental configuration for testing XLA GPU lowering to TFRT BEF thunks. # bazel test --config=experimental_enable_bef_thunk \ # //tensorflow/compiler/xla/service/gpu/tests:mlir_gemm_test build:experimental_enable_bef_thunk --config=cuda build:experimental_enable_bef_thunk --//tensorflow/compiler/xla/service/gpu:enable_bef_thunk build:experimental_enable_bef_thunk --@tf_runtime//:enable_gpu build:experimental_enable_bef_thunk --@rules_cuda//cuda:enable_cuda build:experimental_enable_bef_thunk --nocheck_visibility build:experimental_enable_bef_thunk --incompatible_strict_action_env build:experimental_enable_bef_thunk --config=monolithic