Merge pull request #3087 from ROCm/develop-upstream-sync-250825

Develop upstream sync 250825

Author: mmakevic-amd

.bazelrc

@@ -76,9 +76,9 @@
 # The generated version suffix is used in
 # third_party/tensorflow/core/public/release_version.h and
 # third_party/tensorflow/tools/pip_package/setup.oss.py.tpl
-build --repo_env=ML_WHEEL_TYPE="nightly"
-build --repo_env=ML_WHEEL_BUILD_DATE="20250716"
-build --repo_env=ML_WHEEL_VERSION_SUFFIX=".630"
+build --repo_env=ML_WHEEL_TYPE="snapshot"
+build --repo_env=ML_WHEEL_BUILD_DATE=""
+build --repo_env=ML_WHEEL_VERSION_SUFFIX=""
 
 # 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
@@ -163,9 +163,13 @@ common --incompatible_enforce_config_setting_visibility
 # TODO: Enable Bzlmod
 common --noenable_bzlmod
 
+build --incompatible_enable_cc_toolchain_resolution
+build --repo_env USE_HERMETIC_CC_TOOLCHAIN=1
+
 # TODO: Migrate for https://github.com/bazelbuild/bazel/issues/7260
-common --noincompatible_enable_cc_toolchain_resolution
-common --noincompatible_enable_android_toolchain_resolution
+build:clang_local --noincompatible_enable_cc_toolchain_resolution
+build:clang_local --noincompatible_enable_android_toolchain_resolution
+build:clang_local --repo_env USE_HERMETIC_CC_TOOLCHAIN=0
 
 # Print a stacktrace when a test is killed
 test --test_env="GTEST_INSTALL_FAILURE_SIGNAL_HANDLER=1"
@@ -176,6 +180,7 @@ test --test_env="GTEST_INSTALL_FAILURE_SIGNAL_HANDLER=1"
 # 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 --config=clang_local
 build:android --crosstool_top=//external:android/crosstool
 build:android --host_crosstool_top=@bazel_tools//tools/cpp:toolchain
 build:android_arm --config=android
@@ -214,7 +219,8 @@ build:macos --features=archive_param_file
 build:macos --linkopt=-Wl,-undefined,dynamic_lookup
 build:macos --host_linkopt=-Wl,-undefined,dynamic_lookup
 
-# Use the Apple toolchain for MacOS builds.
+# Use the old Apple toolchain for MacOS builds.
+build:macos --config=clang_local
 build:macos --config=apple-toolchain
 
 # Use cc toolchains from apple_support for Apple builds (ios, macos, etc).
@@ -226,12 +232,14 @@ build:apple-toolchain --host_crosstool_top=@local_config_apple_cc//:toolchain
 # Settings for MacOS on ARM CPUs.
 build:macos_arm64 --cpu=darwin_arm64
 build:macos_arm64 --macos_minimum_os=11.0
+build:macos_arm64 --config=clang_local
 build:macos_arm64 --platforms=@build_bazel_apple_support//configs/platforms:darwin_arm64
 
 # iOS configs for each architecture and the fat binary builds.
 build:ios --apple_platform_type=ios
 build:ios --copt=-fembed-bitcode
 build:ios --copt=-Wno-c++11-narrowing
+build:ios --config=clang_local
 build:ios --config=apple-toolchain
 build:ios_armv7 --config=ios
 build:ios_armv7 --cpu=ios_armv7
@@ -273,16 +281,12 @@ 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).
-build:mkl_aarch64 --define=build_with_mkl_aarch64=true
-build:mkl_aarch64 --define=build_with_openmp=true
-build:mkl_aarch64 --define=build_with_acl=true
-build:mkl_aarch64 -c opt
-
 # Config setting to build oneDNN with Compute Library for the Arm Architecture (ACL).
 # with Eigen threadpool support
 build:mkl_aarch64_threadpool --define=build_with_mkl_aarch64=true
 build:mkl_aarch64_threadpool -c opt
+# This is an alias for the mkl_aarch64_threadpool build.
+build:mkl_aarch64 --config=mkl_aarch64_threadpool
 
 # Default CUDA, CUDNN and NVSHMEM versions.
 build:cuda_version --repo_env=HERMETIC_CUDA_VERSION="12.5.1"
@@ -299,7 +303,6 @@ build:cuda --@local_config_cuda//cuda:include_cuda_libs=true
 
 # This configuration is used for building the wheels.
 build:cuda_wheel --@local_config_cuda//cuda:include_cuda_libs=false
-build:cuda_wheel --@local_config_nvshmem//:include_nvshmem_libs=false
 
 # CUDA: This config refers to building CUDA op kernels with clang.
 build:cuda_clang --config=cuda
@@ -358,6 +361,7 @@ build:tpu --define=framework_shared_object=true
 build:tpu --copt=-DLIBTPU_ON_GCE
 build:tpu --define=enable_mlir_bridge=true
 
+build:rocm --config=clang_local
 build:rocm_base --crosstool_top=@local_config_rocm//crosstool:toolchain
 build:rocm_base --define=using_rocm_hipcc=true
 build:rocm_base --define=tensorflow_mkldnn_contraction_kernel=0
@@ -398,6 +402,7 @@ build:rocm_ci_hermetic --repo_env="OS=ubuntu_22.04"
 build:rocm_ci_hermetic --repo_env="ROCM_VERSION=6.2.0"
 build:rocm_ci_hermetic --@local_config_rocm//rocm:use_rocm_hermetic_rpath=True
 
+build:sycl --config=clang_local
 build:sycl --crosstool_top=@local_config_sycl//crosstool:toolchain
 build:sycl --define=using_sycl=true
 build:sycl --define=tensorflow_mkldnn_contraction_kernel=0
@@ -537,6 +542,9 @@ build:windows --verbose_failures
 # See: https://github.com/bazelbuild/bazel/issues/5163
 build:windows --features=compiler_param_file
 
+# Use old toolchains for Windows builds.
+build:windows --config=clang_local
+
 # Do not risk cache corruption. See:
 # https://github.com/bazelbuild/bazel/issues/3360
 build:linux --experimental_guard_against_concurrent_changes
@@ -550,7 +558,9 @@ build:verbose_logs --output_filter=
 #   avoid having to define linux/win separately.
 build:avx_linux --copt=-mavx
 build:avx_linux --host_copt=-mavx
+build:avx_linux --copt="-mf16c"
 build:avx_win --copt=/arch:AVX
+build:avx_win --copt="-mf16c"
 
 build:win_clang_base --@com_google_protobuf//build_defs:use_dlls=True
 build:win_clang_base --@com_google_absl//absl:use_dlls=True
@@ -567,6 +577,7 @@ test:win_clang_base --host_linkopt=/FORCE:MULTIPLE
 test:win_clang_base --build_tests_only --keep_going --test_output=errors --verbose_failures=true --test_summary=short
 
 build:win_clang --config=win_clang_base
+build:win_clang --config=clang_local
 build:win_clang --extra_toolchains=@local_config_cc//:cc-toolchain-x64_windows-clang-cl
 build:win_clang --extra_execution_platforms=//tensorflow/tools/toolchains/win:x64_windows-clang-cl
 build:win_clang --host_platform=//tensorflow/tools/toolchains/win:x64_windows-clang-cl
@@ -604,7 +615,13 @@ build:resultstore --bes_timeout=600s
 # Flag to enable remote config
 common --experimental_repo_remote_exec
 
+build:use_tar_archive_files --repo_env=USE_CUDA_TAR_ARCHIVE_FILES=1
+build:use_tar_archive_files --repo_env=USE_NVSHMEM_TAR_ARCHIVE_FILES=1
+build:use_tar_archive_files --repo_env=USE_LLVM_TAR_ARCHIVE_FILES=1
+build:use_tar_archive_files --repo_env=USE_MIRRORED_TAR_ARCHIVE_FILES=1
+
 # Make Bazel not try to probe the host system for a C++ toolchain.
+build:rbe_base --config=use_tar_archive_files
 build:rbe_base --config=resultstore
 build:rbe_base --repo_env=BAZEL_DO_NOT_DETECT_CPP_TOOLCHAIN=1
 build:rbe_base --define=EXECUTOR=remote
@@ -629,12 +646,6 @@ build:rbe_linux --linkopt=-lm
 build:rbe_linux --host_linkopt=-lm
 
 build:rbe_linux_cpu --config=rbe_linux
-# Linux cpu and cuda builds share the same toolchain now.
-build:rbe_linux_cpu --host_crosstool_top="@local_config_cuda//crosstool:toolchain"
-build:rbe_linux_cpu --crosstool_top="@local_config_cuda//crosstool:toolchain"
-build:rbe_linux_cpu --extra_toolchains="@local_config_cuda//crosstool:toolchain-linux-x86_64"
-build:rbe_linux_cpu --repo_env=CC="/usr/lib/llvm-18/bin/clang"
-build:rbe_linux_cpu --repo_env=TF_SYSROOT="/dt9"
 build:rbe_linux_cpu --extra_execution_platforms="@ml_build_config_platform//:platform"
 build:rbe_linux_cpu --host_platform="@ml_build_config_platform//:platform"
 build:rbe_linux_cpu --platforms="@ml_build_config_platform//:platform"
@@ -656,6 +667,15 @@ common:rbe_linux_cpu --remote_instance_name=projects/tensorflow-testing/instance
 # build:rbe_linux_cpu --repo_env USE_CUDA_REDISTRIBUTIONS=1
 # build:rbe_linux_cpu --config=cuda_version
 
+# Deprecated RBE config with non-hermetic toolchains.
+build:rbe_linux_cpu_clang_local --config=rbe_linux_cpu
+build:rbe_linux_cpu_clang_local --config=clang_local
+build:rbe_linux_cpu_clang_local --host_crosstool_top="@local_config_cuda//crosstool:toolchain"
+build:rbe_linux_cpu_clang_local --crosstool_top="@local_config_cuda//crosstool:toolchain"
+build:rbe_linux_cpu_clang_local --extra_toolchains="@local_config_cuda//crosstool:toolchain-linux-x86_64"
+build:rbe_linux_cpu_clang_local --repo_env=CC="/usr/lib/llvm-18/bin/clang"
+build:rbe_linux_cpu_clang_local --repo_env=TF_SYSROOT="/dt9"
+
 # TODO(kanglan): Remove it after toolchain update is complete.
 build:rbe_linux_cpu_old --config=rbe_linux
 build:rbe_linux_cpu_old --host_crosstool_top="@ubuntu20.04-gcc9_manylinux2014-cuda11.2-cudnn8.1-tensorrt7.2_config_cuda//crosstool:toolchain"
@@ -669,8 +689,6 @@ common:rbe_linux_cpu_old --remote_instance_name=projects/tensorflow-testing/inst
 
 build:rbe_linux_cuda --config=cuda_clang_official
 build:rbe_linux_cuda --config=rbe_linux_cpu
-build:rbe_linux_cuda --repo_env=USE_CUDA_TAR_ARCHIVE_FILES=1
-build:rbe_linux_cuda --repo_env=USE_NVSHMEM_TAR_ARCHIVE_FILES=1
 # For Remote build execution -- GPU configuration
 build:rbe_linux_cuda --repo_env=REMOTE_GPU_TESTING=1
 
@@ -692,6 +710,7 @@ build:rbe_linux_cuda_nvcc --config=cuda_nvcc
 build:rbe_linux_cuda_nvcc --repo_env TF_NCCL_USE_STUB=1
 
 build:rbe_win_base --config=rbe_base
+build:rbe_win_base --config=clang_local
 build:rbe_win_base --shell_executable=C:\\tools\\msys64\\usr\\bin\\bash.exe
 build:rbe_win_base --remote_instance_name=projects/tensorflow-testing/instances/windows
 # Don't build the python zip archive in the RBE build.
@@ -707,6 +726,7 @@ build:rbe_windows_x86_cpu_2022 --config=rbe_win_base --config=windows_x86_cpu_20
 # END TF REMOTE BUILD EXECUTION OPTIONS
 
 # TFLite build configs for generic embedded Linux
+build:elinux --config=clang_local
 build:elinux --crosstool_top=@local_config_embedded_arm//:toolchain
 build:elinux --host_crosstool_top=@bazel_tools//tools/cpp:toolchain
 build:elinux_aarch64 --config=elinux
@@ -781,8 +801,23 @@ build:release_gpu_linux --config=cuda_clang_official
 # Local test jobs has to be 4 because parallel_gpu_execute is fragile, I think
 test:release_gpu_linux --test_timeout=300,450,1200,3600 --local_test_jobs=4 --run_under=//tensorflow/tools/ci_build/gpu_build:parallel_gpu_execute
 
+# Deprecated release CPU config with non-hermetic toolchains.
+build:release_cpu_linux_clang_local --config=release_cpu_linux
+build:release_cpu_linux_clang_local --config=clang_local
+build:release_cpu_linux_clang_local --repo_env=CC="/usr/lib/llvm-18/bin/clang"
+build:release_cpu_linux_clang_local --repo_env=BAZEL_COMPILER="/usr/lib/llvm-18/bin/clang"
+build:release_cpu_linux_clang_local --crosstool_top="@local_config_cuda//crosstool:toolchain"
+build:release_cpu_linux_clang_local --repo_env=TF_SYSROOT="/dt9"
+
+# Deprecated release GPU config with non-hermetic toolchains.
+build:release_gpu_linux_clang_local --config=release_gpu_linux
+build:release_gpu_linux_clang_local --config=release_cpu_linux_clang_local
+
 build:release_arm64_linux --config=release_linux_base
 build:release_arm64_linux --config=linux_arm64
+build:release_arm64_linux --config=clang_local
+build:release_arm64_linux --repo_env=CC="/usr/lib/llvm-18/bin/clang"
+build:release_arm64_linux --repo_env=BAZEL_COMPILER="/usr/lib/llvm-18/bin/clang"
 build:release_arm64_linux --crosstool_top="@ml2014_clang_aarch64_config_aarch64//crosstool:toolchain"
 build:release_arm64_linux --config=mkl_aarch64_threadpool
 build:release_arm64_linux --copt=-flax-vector-conversions
@@ -791,6 +826,7 @@ test:release_arm64_linux --flaky_test_attempts=3
 build:release_cpu_macos --config=avx_linux
 
 # Base build configs for macOS
+build:release_macos_base --config=clang_local
 build:release_macos_base --action_env  DEVELOPER_DIR=/Applications/Xcode.app/Contents/Developer
 build:release_macos_base --define=no_nccl_support=true --output_filter=^$
 
@@ -956,6 +992,7 @@ test:windows_x86_cpu_2022_pycpp_test --config=windows_x86_cpu_2022_pycpp_test_op
 # flags seem to be actually used to specify the execution platform details. It
 # seems it is this way because these flags are old and predate the distinction
 # between host and execution platform.
+build:cross_compile_base --config=clang_local
 build:cross_compile_base --host_cpu=k8
 build:cross_compile_base --host_crosstool_top=//tensorflow/tools/toolchains/cross_compile/cc:cross_compile_toolchain_suite
 build:cross_compile_base --extra_execution_platforms=//tensorflow/tools/toolchains/cross_compile/config:linux_x86_64

.github/workflows/osv-scanner-scheduled.yml

@@ -28,7 +28,7 @@ permissions:
 jobs:
   scan-scheduled:
     if: github.repository == 'tensorflow/tensorflow'
-    uses: "google/osv-scanner-action/.github/workflows/osv-scanner-reusable.yml@v2.0.3"
+    uses: "google/osv-scanner-action/.github/workflows/osv-scanner-reusable.yml@v2.1.0"
     with:
       scan-args: |-
         --lockfile=requirements.txt:./requirements_lock_3_9.txt

.github/workflows/scorecards-analysis.yml

@@ -64,6 +64,6 @@ jobs:
       # Upload the results to GitHub's code scanning dashboard (optional).
       # Commenting out will disable upload of results to your repo's Code Scanning dashboard
       - name: "Upload to code-scanning"
-        uses: github/codeql-action/upload-sarif@181d5eefc20863364f96762470ba6f862bdef56b # v3.29.2
+        uses: github/codeql-action/upload-sarif@51f77329afa6477de8c49fc9c7046c15b9a4e79d # v3.29.5
         with:
           sarif_file: results.sarif

README.md

@@ -87,7 +87,7 @@ $ pip install tensorflow
 ```
 
 Other devices (DirectX and MacOS-metal) are supported using
-[Device plugins](https://www.tensorflow.org/install/gpu_plugins#available_devices).
+[Device Plugins](https://www.tensorflow.org/install/gpu_plugins#available_devices).
 
 A smaller CPU-only package is also available:
 
@@ -118,16 +118,16 @@ b'Hello, TensorFlow!'
 ```
 
 For more examples, see the
-[TensorFlow tutorials](https://www.tensorflow.org/tutorials/).
+[TensorFlow Tutorials](https://www.tensorflow.org/tutorials/).
 
 ## Contribution guidelines
 
 **If you want to contribute to TensorFlow, be sure to review the
-[contribution guidelines](CONTRIBUTING.md). This project adheres to TensorFlow's
-[code of conduct](CODE_OF_CONDUCT.md). By participating, you are expected to
+[Contribution Guidelines](CONTRIBUTING.md). This project adheres to TensorFlow's
+[Code of Conduct](CODE_OF_CONDUCT.md). By participating, you are expected to
 uphold this code.**
 
-**We use [GitHub issues](https://github.com/tensorflow/tensorflow/issues) for
+**We use [GitHub Issues](https://github.com/tensorflow/tensorflow/issues) for
 tracking requests and bugs, please see
 [TensorFlow Forum](https://discuss.tensorflow.org/) for general questions and
 discussion, and please direct specific questions to
@@ -141,18 +141,18 @@ open-source software development.
 Follow these steps to patch a specific version of TensorFlow, for example, to
 apply fixes to bugs or security vulnerabilities:
 
-*   Clone the TensorFlow repo and switch to the corresponding branch for your
-    desired TensorFlow version, for example, branch `r2.8` for version 2.8.
-*   Apply (that is, cherry-pick) the desired changes and resolve any code
-    conflicts.
+*    Clone the TensorFlow repository and switch to the appropriate branch for
+  your desired version—for example, `r2.8` for version 2.8.
+*   Apply the desired changes (i.e., cherry-pick them) and resolve any code
+  conflicts.
 *   Run TensorFlow tests and ensure they pass.
 *   [Build](https://www.tensorflow.org/install/source) the TensorFlow pip
     package from source.
 
 ## Continuous build status
 
 You can find more community-supported platforms and configurations in the
-[TensorFlow SIG Build community builds table](https://github.com/tensorflow/build#community-supported-tensorflow-builds).
+[TensorFlow SIG Build Community Builds Table](https://github.com/tensorflow/build#community-supported-tensorflow-builds).
 
 ### Official Builds
 
@@ -190,8 +190,8 @@ Build Type                    | Status
 *   [TensorFlow Code Search](https://cs.opensource.google/tensorflow/tensorflow)
 
 Learn more about the
-[TensorFlow community](https://www.tensorflow.org/community) and how to
-[contribute](https://www.tensorflow.org/community/contribute).
+[TensorFlow Community](https://www.tensorflow.org/community) and how to
+[Contribute](https://www.tensorflow.org/community/contribute).
 
 ## Courses