Update copy_strip_binary.sh: use "make install" instead (#21464)

### Description
Before this change, copy_strip_binary.sh manually copies each file from
onnx runtime's build folder to an artifact folder. It can be hard when
dealing with symbolic link for shared libraries.
This PR will change the packaging pipelines to run "make install" first,
before packaging shared libs .


### Motivation and Context

Recently because of feature request #21281 , we changed
libonnxruntime.so's SONAME. Now every package that contains this shared
library must also contains libonnxruntime.so.1. Therefore we need to
change the packaging scripts to include this file. Instead of manually
construct the symlink layout, using `make install` is much easier and
will make things more consistent because it is a standard way of making
packages.

**Breaking change:**
After this change, our **inference** tarballs that are published to our
Github release pages will be not contain ORT **training** headers.

cmake/onnxruntime.cmake

@@ -27,6 +27,7 @@ function(get_c_cxx_api_headers HEADERS_VAR)
     "${REPO_ROOT}/include/onnxruntime/core/session/onnxruntime_float16.h"
     "${REPO_ROOT}/include/onnxruntime/core/session/onnxruntime_run_options_config_keys.h"
     "${REPO_ROOT}/include/onnxruntime/core/session/onnxruntime_session_options_config_keys.h"
+    "${REPO_ROOT}/include/onnxruntime/core/session/onnxruntime_lite_custom_op.h"
   )
 
   if (onnxruntime_ENABLE_TRAINING_APIS)

cmake/onnxruntime_framework.cmake

@@ -123,7 +123,9 @@ if (WIN32)
   target_compile_definitions(onnxruntime_framework PRIVATE _SCL_SECURE_NO_WARNINGS)
 endif()
 
-if (NOT onnxruntime_BUILD_SHARED_LIB)
+if (onnxruntime_BUILD_SHARED_LIB)
+  install(FILES ${PROJECT_SOURCE_DIR}/../include/onnxruntime/core/framework/provider_options.h  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/onnxruntime/)
+else()
   install(DIRECTORY ${PROJECT_SOURCE_DIR}/../include/onnxruntime/core/framework  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/onnxruntime/core)
   install(TARGETS onnxruntime_framework
             ARCHIVE   DESTINATION ${CMAKE_INSTALL_LIBDIR}

onnxruntime/core/optimizer/graph_transformer_utils.cc

@@ -189,7 +189,7 @@ InlinedVector<std::unique_ptr<GraphTransformer>> GenerateTransformers(
     const SessionOptions& session_options,
     const IExecutionProvider& cpu_execution_provider, /*required by constant folding*/
     const InlinedHashSet<std::string>& rules_and_transformers_to_disable,
-    concurrency::ThreadPool* intra_op_thread_pool) {
+    [[maybe_unused]] concurrency::ThreadPool* intra_op_thread_pool) {
   InlinedVector<std::unique_ptr<GraphTransformer>> transformers;
   const bool disable_quant_qdq =
       session_options.config_options.GetConfigOrDefault(kOrtSessionOptionsDisableQuantQDQ, "0") == "1";
@@ -419,7 +419,7 @@ InlinedVector<std::unique_ptr<GraphTransformer>> GenerateTransformersForMinimalB
     const SatApplyContextVariant& apply_context,
     const IExecutionProvider& cpu_execution_provider,
     const InlinedHashSet<std::string>& rules_and_transformers_to_disable,
-    concurrency::ThreadPool* intra_op_thread_pool) {
+    [[maybe_unused]] concurrency::ThreadPool* intra_op_thread_pool) {
   InlinedVector<std::unique_ptr<GraphTransformer>> transformers;
   const bool saving = std::holds_alternative<SatRuntimeOptimizationSaveContext>(apply_context);
 

tools/ci_build/github/azure-pipelines/templates/c-api-linux-cpu.yml

@@ -69,7 +69,7 @@ jobs:
           docker run --rm --volume /data/onnx:/data/onnx:ro --volume $(Build.SourcesDirectory):/onnxruntime_src --volume $(Build.BinariesDirectory):/build \
           --volume $HOME/.onnx:/home/onnxruntimedev/.onnx -e NIGHTLY_BUILD onnxruntimecpubuildcentos8${{parameters.OnnxruntimeArch}} /bin/bash -c "python3.9 \
           /onnxruntime_src/tools/ci_build/build.py --enable_lto --build_java --build_nodejs --build_dir /build --config Release \
-          --skip_submodule_sync --parallel --use_binskim_compliant_compile_flags --build_shared_lib ${{ parameters.AdditionalBuildFlags }} && cd /build/Release && make install DESTDIR=/build/linux-${{parameters.OnnxruntimeArch}}"
+          --skip_submodule_sync --parallel --use_binskim_compliant_compile_flags --build_shared_lib ${{ parameters.AdditionalBuildFlags }} && cd /build/Release && make install DESTDIR=/build/installed"
         workingDirectory: $(Build.SourcesDirectory)
       displayName: 'Build'
 

tools/ci_build/github/azure-pipelines/templates/final-jar-testing.yml

@@ -68,8 +68,9 @@ stages:
         inputs:
           targetType: 'inline'
           script: |
+            set -e -x
             echo "Java Version"
-            java --version
+            java -version
             mkdir test
             pushd test
             jar xf '$(Build.BinariesDirectory)/final-jar/testing.jar'

tools/ci_build/github/azure-pipelines/templates/mac-cpu-packaging-steps.yml

@@ -46,8 +46,11 @@ steps:
     ChangeEveryCommit: true
     BuildStep:
       - script: |
+          set -e -x
           rm -rf $(Build.BinariesDirectory)/Release
           python3 $(Build.SourcesDirectory)/tools/ci_build/build.py --update --build ${{ parameters.AdditionalBuildFlags }} --build_dir $(Build.BinariesDirectory) --skip_submodule_sync --parallel --use_binskim_compliant_compile_flags --build_shared_lib --config Release
+          cd $(Build.BinariesDirectory)/Release
+          make install DESTDIR=$(Build.BinariesDirectory)/installed
         displayName: 'Build ${{ parameters.MacosArch }}'
         env:
           CCACHE_DIR: ${{ parameters.CacheDir }}

tools/ci_build/github/linux/build_cuda_c_api_package.sh

@@ -1,10 +1,5 @@
 #!/bin/bash
 set -e -x
 docker run --rm --volume \
-$BUILD_SOURCESDIRECTORY:/onnxruntime_src --volume $BUILD_BINARIESDIRECTORY:/build \
---volume /data/models:/build/models:ro --volume /data/onnx:/data/onnx:ro -e NIGHTLY_BUILD onnxruntimecuda${CUDA_VERSION_MAJOR}build \
-/usr/bin/python3.9 /onnxruntime_src/tools/ci_build/build.py --enable_lto --build_java --build_nodejs --build_dir /build --config Release \
---skip_submodule_sync  --parallel --use_binskim_compliant_compile_flags --build_shared_lib --use_cuda --cuda_version=$CUDA_VERSION \
---cuda_home=/usr/local/cuda-$CUDA_VERSION --cudnn_home=/usr/local/cuda-$CUDA_VERSION \
---skip_tests \
---cmake_extra_defines 'CMAKE_CUDA_ARCHITECTURES=60;61;70;75;80'
+$BUILD_SOURCESDIRECTORY:/onnxruntime_src --volume $BUILD_BINARIESDIRECTORY:/build -e NIGHTLY_BUILD onnxruntimecuda${CUDA_VERSION_MAJOR}build \
+/bin/bash -c "/usr/bin/python3.9 /onnxruntime_src/tools/ci_build/build.py --enable_lto --build_java --build_nodejs --build_dir /build --config Release --skip_submodule_sync  --parallel --use_binskim_compliant_compile_flags --build_shared_lib --use_cuda --cuda_version=$CUDA_VERSION --cuda_home=/usr/local/cuda-$CUDA_VERSION --cudnn_home=/usr/local/cuda-$CUDA_VERSION --skip_tests --cmake_extra_defines 'CMAKE_CUDA_ARCHITECTURES=60;61;70;75;80' && cd /build/Release && make install DESTDIR=/build/installed"

tools/ci_build/github/linux/build_rocm_c_api_package.sh

@@ -31,14 +31,7 @@ docker run --rm \
   --volume /data/onnx:/data/onnx:ro \
   --workdir /onnxruntime_src \
   $IMAGE \
-  ${PYTHON_BIN:-python} /onnxruntime_src/tools/ci_build/build.py \
-    --config Release \
-    --build_dir /build \
-    --parallel \
-    --use_rocm --rocm_version=$ROCM_VERSION --rocm_home $ROCM_HOME --nccl_home $ROCM_HOME \
-    --build_shared_lib \
-    --skip_submodule_sync \
-    --skip_tests --cmake_extra_defines FETCHCONTENT_TRY_FIND_PACKAGE_MODE=NEVER
+  /bin/bash -c "${PYTHON_BIN:-python} /onnxruntime_src/tools/ci_build/build.py --config Release --build_dir /build --parallel --use_rocm --use_binskim_compliant_compile_flags --rocm_version=$ROCM_VERSION --rocm_home $ROCM_HOME --nccl_home $ROCM_HOME --build_shared_lib --skip_submodule_sync --skip_tests --cmake_extra_defines FETCHCONTENT_TRY_FIND_PACKAGE_MODE=NEVER && cd /build/Release && make install DESTDIR=/build/installed"
 
 
 EXIT_CODE=$?

tools/ci_build/github/linux/build_tensorrt_c_api_package.sh

@@ -3,6 +3,4 @@ set -e -x
 mkdir -p $HOME/.onnx
 docker run --rm --volume /data/onnx:/data/onnx:ro --volume $BUILD_SOURCESDIRECTORY:/onnxruntime_src --volume $BUILD_BINARIESDIRECTORY:/build \
 --volume /data/models:/build/models:ro --volume $HOME/.onnx:/home/onnxruntimedev/.onnx -e NIGHTLY_BUILD onnxruntimecuda${CUDA_VERSION_MAJOR}xtrt86build \
-/usr/bin/python3.9 /onnxruntime_src/tools/ci_build/build.py --build_dir /build --config Release \
---skip_tests \
---skip_submodule_sync --parallel --use_binskim_compliant_compile_flags --build_shared_lib --build_java --build_nodejs --use_tensorrt --cuda_version=$CUDA_VERSION --cuda_home=/usr/local/cuda-$CUDA_VERSION --cudnn_home=/usr --tensorrt_home=/usr --cmake_extra_defines 'CMAKE_CUDA_ARCHITECTURES=60;61;70;75;80'
+/bin/bash -c "/usr/bin/python3.9 /onnxruntime_src/tools/ci_build/build.py --build_dir /build --config Release --skip_tests --skip_submodule_sync --parallel --use_binskim_compliant_compile_flags --build_shared_lib --build_java --build_nodejs --use_tensorrt --cuda_version=$CUDA_VERSION --cuda_home=/usr/local/cuda-$CUDA_VERSION --cudnn_home=/usr --tensorrt_home=/usr --cmake_extra_defines 'CMAKE_CUDA_ARCHITECTURES=60;61;70;75;80' && cd /build/Release && make install DESTDIR=/build/installed"

tools/ci_build/github/linux/copy_strip_binary.sh

@@ -16,46 +16,22 @@ done
 EXIT_CODE=1
 
 uname -a
-mkdir $BINARY_DIR/$ARTIFACT_NAME
-mkdir $BINARY_DIR/$ARTIFACT_NAME/lib
-mkdir $BINARY_DIR/$ARTIFACT_NAME/include
-echo "Directories created"
-cp $BINARY_DIR/$BUILD_CONFIG/$LIB_NAME $BINARY_DIR/$ARTIFACT_NAME/lib
-if [[ -f "$BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_cuda.so" ]]; then
-    cp $BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_shared.so $BINARY_DIR/$ARTIFACT_NAME/lib
-    cp $BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_cuda.so $BINARY_DIR/$ARTIFACT_NAME/lib
-fi
-if [[ -f "$BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_tensorrt.so" ]]; then
-    cp $BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_tensorrt.so $BINARY_DIR/$ARTIFACT_NAME/lib
-fi
-if [[ -f "$BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_rocm.so" ]]; then
-    cp $BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_shared.so $BINARY_DIR/$ARTIFACT_NAME/lib
-    cp $BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_rocm.so $BINARY_DIR/$ARTIFACT_NAME/lib
-fi
+cd "$BINARY_DIR"
+mv installed/usr/local $ARTIFACT_NAME
+mv $ARTIFACT_NAME/include/onnxruntime/* $ARTIFACT_NAME/include
+rmdir $ARTIFACT_NAME/include/onnxruntime
+# Do not ship onnx_test_runner
+rm -rf $ARTIFACT_NAME/bin
 echo "Copy debug symbols in a separate file and strip the original binary."
 if [[ $LIB_NAME == *.dylib ]]
 then
     dsymutil $BINARY_DIR/$ARTIFACT_NAME/lib/$LIB_NAME -o $BINARY_DIR/$ARTIFACT_NAME/lib/$LIB_NAME.dSYM
     strip -S $BINARY_DIR/$ARTIFACT_NAME/lib/$LIB_NAME
-    ln -s $LIB_NAME $BINARY_DIR/$ARTIFACT_NAME/lib/libonnxruntime.dylib
     # copy the CoreML EP header for macOS build (libs with .dylib ext)
     cp $SOURCE_DIR/include/onnxruntime/core/providers/coreml/coreml_provider_factory.h  $BINARY_DIR/$ARTIFACT_NAME/include
-elif [[ $LIB_NAME == *.so.* ]]
-then
-    ln -s $LIB_NAME $BINARY_DIR/$ARTIFACT_NAME/lib/libonnxruntime.so
-fi
-cp $SOURCE_DIR/include/onnxruntime/core/session/onnxruntime_*.h $BINARY_DIR/$ARTIFACT_NAME/include
-cp $SOURCE_DIR/include/onnxruntime/core/framework/provider_options.h  $BINARY_DIR/$ARTIFACT_NAME/include
-cp $SOURCE_DIR/include/onnxruntime/core/providers/cpu/cpu_provider_factory.h  $BINARY_DIR/$ARTIFACT_NAME/include
-cp $SOURCE_DIR/orttraining/orttraining/training_api/include/onnxruntime_training_*.h  $BINARY_DIR/$ARTIFACT_NAME/include
-
-if [[ -f "$BINARY_DIR/$BUILD_CONFIG/libonnxruntime_providers_cuda.so" ]]; then
-# copy headers for context context used in custom ops
-mkdir -p $BINARY_DIR/$ARTIFACT_NAME/include/core/providers/cuda
-cp $SOURCE_DIR/include/onnxruntime/core/providers/custom_op_context.h $BINARY_DIR/$ARTIFACT_NAME/include/core/providers/custom_op_context.h
-cp $SOURCE_DIR/include/onnxruntime/core/providers/resource.h $BINARY_DIR/$ARTIFACT_NAME/include/core/providers/resource.h
-cp $SOURCE_DIR/include/onnxruntime/core/providers/cuda/cuda_context.h $BINARY_DIR/$ARTIFACT_NAME/include/core/providers/cuda/cuda_context.h
-cp $SOURCE_DIR/include/onnxruntime/core/providers/cuda/cuda_resource.h $BINARY_DIR/$ARTIFACT_NAME/include/core/providers/cuda/cuda_resource.h
+else
+   # Linux
+   mv $ARTIFACT_NAME/lib64 $ARTIFACT_NAME/lib
 fi
 
 # copy the README, licence and TPN