Merge branch 'main' of https://github.com/microsoft/onnxruntime into …

…zhanyi/activationparam

.github/workflows/lint.yml

@@ -45,7 +45,7 @@ jobs:
       - name: Setup Python
         uses: actions/setup-python@v5
         with:
-          # Version range or exact version of Python to use, using SemVer's version range syntax. Reads from .python-version if unset.
+          # Use the version configured in target-version of [tool.black] section in pyproject.toml.
           python-version: "3.10"
       - name: Setup Rust
         uses: actions-rs/toolchain@v1
@@ -55,12 +55,10 @@ jobs:
       - name: Update PATH
         run: |
           echo "$HOME/.local/bin" >> "$GITHUB_PATH"
-
       - name: Install dependencies
         run: |
           set -e -x
           python -m pip install --user -r requirements-dev.txt
-          python -m pip install --user lintrunner lintrunner-adapters
           lintrunner init
       - name: Run lintrunner on all files
         run: |

.lintrunner.toml

@@ -2,31 +2,23 @@
 # You can install the dependencies and initialize with
 #
 # ```sh
-# pip install lintrunner lintrunner-adapters
+# pip install -r requirements-lintrunner.txt
 # lintrunner init
 # ```
 #
 # This will install lintrunner on your system and download all the necessary
 # dependencies to run linters locally.
-# If you want to see what lintrunner init will install, run
-# `lintrunner init --dry-run`.
 #
-# To lint local changes:
+# To format local changes:
 #
 # ```bash
-# lintrunner
+# lintrunner -a
 # ```
 #
-# To lint all files:
+# To format all files:
 #
 # ```bash
-# lintrunner --all-files
-# ```
-#
-# To format files:
-#
-# ```bash
-# lintrunner f --all-files
+# lintrunner -a --all-files
 # ```
 #
 # To read more about lintrunner, see [wiki](https://github.com/pytorch/pytorch/wiki/lintrunner).

CODEOWNERS

@@ -9,10 +9,6 @@
 /onnxruntime/core/graph/contrib_ops/quantization_defs.* @microsoft/onnxruntime-mlas
 /onnxruntime/core/mlas/** @microsoft/onnxruntime-mlas
 
-# build pipelines and workflows
-/tools/ci_build/github/azure-pipelines @microsoft/onnxruntime-es
-/.github/workflows @microsoft/onnxruntime-es
-
 # Dependencies
 requirements-dev.txt @microsoft/onnxruntime-admin
 requirements-doc.txt @microsoft/onnxruntime-admin

cgmanifests/generated/cgmanifest.json

@@ -196,7 +196,7 @@
       "component": {
         "type": "git",
         "git": {
-          "commitHash": "bc0d2e35909b8456abe32f3b30a49bb0c125e8b7",
+          "commitHash": "9c69a24bc2e20c8a511a4e6b06fd49639ec5300a",
           "repositoryUrl": "https://github.com/onnx/onnx-tensorrt.git"
         },
         "comments": "onnx_tensorrt"

cmake/CMakeLists.txt

@@ -130,8 +130,7 @@ option(onnxruntime_DEBUG_NODE_INPUTS_OUTPUTS "Dump debug information about node
 cmake_dependent_option(onnxruntime_DEBUG_NODE_INPUTS_OUTPUTS_ENABLE_DUMP_TO_SQLDB "Build dump debug information about node inputs and outputs with support for sql database." OFF "onnxruntime_DEBUG_NODE_INPUTS_OUTPUTS" OFF)
 
 # When loading a delay loaded DLL, Windows searches the main EXE's folder first.
-# In a Python process, it searches where python.exe lives, but it doesn't search the python package's installation folder. Therefore we cannot enable this flag when Python is enabled.
-cmake_dependent_option(onnxruntime_ENABLE_DELAY_LOADING_WIN_DLLS "Delay load some of the dependent DLls that are part of the OS" ON "WIN32;NOT GDK_PLATFORM;NOT onnxruntime_ENABLE_PYTHON" OFF)
+cmake_dependent_option(onnxruntime_ENABLE_DELAY_LOADING_WIN_DLLS "Delay load some of the dependent DLls that are part of the OS" ON "WIN32;NOT GDK_PLATFORM" OFF)
 option(onnxruntime_USE_DML "Build with DirectML support" OFF)
 option(onnxruntime_USE_MIGRAPHX "Build with AMDMIGraphX support" OFF)
 option(onnxruntime_USE_WINML "Build with WinML support" OFF)

cmake/deps.txt

@@ -36,8 +36,8 @@ microsoft_wil;https://github.com/microsoft/wil/archive/refs/tags/v1.0.230629.1.z
 mimalloc;https://github.com/microsoft/mimalloc/archive/refs/tags/v2.1.1.zip;d5ee7d34223d0567892db5179849939c8769dc41
 mp11;https://github.com/boostorg/mp11/archive/refs/tags/boost-1.82.0.zip;9bc9e01dffb64d9e0773b2e44d2f22c51aace063
 onnx;https://github.com/onnx/onnx/archive/refs/tags/v1.16.1.zip;2eb9198bb352757d5ff13977cbe0634898e0837c
-# Use the latest commit of 10.6-GA-ORT-DDS
-onnx_tensorrt;https://github.com/onnx/onnx-tensorrt/archive/bc0d2e35909b8456abe32f3b30a49bb0c125e8b7.zip;f233ae871ad82c023da62e5dd620639f00bc2d15
+# Use the latest commit of 10.7-GA
+onnx_tensorrt;https://github.com/onnx/onnx-tensorrt/archive/9c69a24bc2e20c8a511a4e6b06fd49639ec5300a.zip;ff1fe9af78eb129b4a4cdcb7450b7390b4436dd3
 protobuf;https://github.com/protocolbuffers/protobuf/archive/refs/tags/v21.12.zip;7cf2733949036c7d52fda017badcab093fe73bfa
 protoc_win64;https://github.com/protocolbuffers/protobuf/releases/download/v21.12/protoc-21.12-win64.zip;b4521f7ada5b260380f94c4bd7f1b7684c76969a
 protoc_win32;https://github.com/protocolbuffers/protobuf/releases/download/v21.12/protoc-21.12-win32.zip;3688010318192c46ce73213cdfb6b3e5656da874

cmake/onnxruntime.cmake

@@ -77,6 +77,7 @@ if(WIN32)
   onnxruntime_add_shared_library(onnxruntime
     ${SYMBOL_FILE}
     "${ONNXRUNTIME_ROOT}/core/dll/dllmain.cc"
+    "${ONNXRUNTIME_ROOT}/core/dll/delay_load_hook.cc"
     "${ONNXRUNTIME_ROOT}/core/dll/onnxruntime.rc"
   )
 elseif(onnxruntime_BUILD_APPLE_FRAMEWORK)

cmake/onnxruntime_nodejs.cmake

@@ -60,15 +60,26 @@ else()
     endif()
 endif()
 
+# a list of DLLs that the Node.js binding depends on
+set(NODEJS_DLL_DEPS)
+
 # setup providers
 if (onnxruntime_USE_CUDA)
     set(NODEJS_BINDING_USE_CUDA "--use_cuda")
 endif()
 if (onnxruntime_USE_DML)
     set(NODEJS_BINDING_USE_DML "--use_dml")
+    list(APPEND NODEJS_DLL_DEPS "$<TARGET_FILE_DIR:onnxruntime>/DirectML.dll")
 endif()
 if (onnxruntime_USE_WEBGPU)
     set(NODEJS_BINDING_USE_WEBGPU "--use_webgpu")
+    if (WIN32 AND onnxruntime_ENABLE_DAWN_BACKEND_D3D12)
+        list(APPEND NODEJS_DLL_DEPS "$<TARGET_FILE_DIR:dxcompiler>/dxil.dll")
+        list(APPEND NODEJS_DLL_DEPS "$<TARGET_FILE_DIR:dxcompiler>/dxcompiler.dll")
+    endif()
+    if (onnxruntime_BUILD_DAWN_MONOLITHIC_LIBRARY)
+        list(APPEND NODEJS_DLL_DEPS "$<TARGET_FILE:dawn::webgpu_dawn>")
+    endif()
 endif()
 if (onnxruntime_USE_TENSORRT)
     set(NODEJS_BINDING_USE_TENSORRT "--use_tensorrt")
@@ -94,9 +105,12 @@ add_custom_target(js_common_npm_ci ALL
 
 add_custom_target(nodejs_binding_wrapper ALL
     COMMAND ${NPM_CLI} ci
-    COMMAND ${NPM_CLI} run build -- --onnxruntime-build-dir=${CMAKE_CURRENT_BINARY_DIR} --config=${CMAKE_BUILD_TYPE} --onnxruntime-generator=${CMAKE_GENERATOR}
-        --arch=${NODEJS_BINDING_ARCH} ${NODEJS_BINDING_USE_CUDA} ${NODEJS_BINDING_USE_DML} ${NODEJS_BINDING_USE_WEBGPU} ${NODEJS_BINDING_USE_TENSORRT}
-        ${NODEJS_BINDING_USE_COREML} ${NODEJS_BINDING_USE_QNN}
+    COMMAND ${NPM_CLI} run build -- "--onnxruntime-build-dir=${CMAKE_CURRENT_BINARY_DIR}"
+        --config=${CMAKE_BUILD_TYPE}
+        "--onnxruntime-generator=${CMAKE_GENERATOR}"
+        "--dll_deps=${NODEJS_DLL_DEPS}"
+        --arch=${NODEJS_BINDING_ARCH} ${NODEJS_BINDING_USE_CUDA} ${NODEJS_BINDING_USE_DML} ${NODEJS_BINDING_USE_WEBGPU}
+        ${NODEJS_BINDING_USE_TENSORRT} ${NODEJS_BINDING_USE_COREML} ${NODEJS_BINDING_USE_QNN}
     WORKING_DIRECTORY ${JS_NODE_ROOT}
     COMMENT "Using cmake-js to build OnnxRuntime Node.js binding")
 

cmake/onnxruntime_providers_webgpu.cmake

@@ -23,24 +23,42 @@
   onnxruntime_add_include_to_target(onnxruntime_providers_webgpu
     onnxruntime_common dawn::dawncpp_headers dawn::dawn_headers onnx onnx_proto flatbuffers::flatbuffers Boost::mp11 safeint_interface)
 
+  set(onnxruntime_providers_webgpu_dll_deps)
+
   if (onnxruntime_BUILD_DAWN_MONOLITHIC_LIBRARY)
     target_link_libraries(onnxruntime_providers_webgpu dawn::webgpu_dawn)
 
-    if (onnxruntime_ENABLE_DELAY_LOADING_WIN_DLLS)
-      list(APPEND onnxruntime_DELAYLOAD_FLAGS "/DELAYLOAD:webgpu_dawn.dll")
-    endif()
+    if (WIN32)
+      if (onnxruntime_ENABLE_DELAY_LOADING_WIN_DLLS)
+        list(APPEND onnxruntime_DELAYLOAD_FLAGS "/DELAYLOAD:webgpu_dawn.dll")
+      endif()
 
-    # Copy webgpu_dawn.dll to the output directory
-    add_custom_command(
-      TARGET onnxruntime_providers_webgpu
-      POST_BUILD
-      COMMAND ${CMAKE_COMMAND} -E copy_if_different "$<TARGET_FILE:dawn::webgpu_dawn>" "$<TARGET_FILE_DIR:onnxruntime_providers_webgpu>"
-      VERBATIM )
+      list(APPEND onnxruntime_providers_webgpu_dll_deps "$<TARGET_FILE:dawn::webgpu_dawn>")
+    endif()
   else()
     if (NOT onnxruntime_USE_EXTERNAL_DAWN)
       target_link_libraries(onnxruntime_providers_webgpu dawn::dawn_native)
     endif()
     target_link_libraries(onnxruntime_providers_webgpu dawn::dawn_proc)
   endif()
 
+  if (WIN32 AND onnxruntime_ENABLE_DAWN_BACKEND_D3D12)
+    # Ensure dxil.dll and dxcompiler.dll exist in the output directory $<TARGET_FILE_DIR:dxcompiler>
+    add_dependencies(onnxruntime_providers_webgpu copy_dxil_dll)
+    add_dependencies(onnxruntime_providers_webgpu dxcompiler)
+
+    list(APPEND onnxruntime_providers_webgpu_dll_deps "$<TARGET_FILE_DIR:dxcompiler>/dxil.dll")
+    list(APPEND onnxruntime_providers_webgpu_dll_deps "$<TARGET_FILE_DIR:dxcompiler>/dxcompiler.dll")
+  endif()
+
+  if (onnxruntime_providers_webgpu_dll_deps)
+    # Copy dependency DLLs to the output directory
+    add_custom_command(
+      TARGET onnxruntime_providers_webgpu
+      POST_BUILD
+      COMMAND ${CMAKE_COMMAND} -E copy_if_different "${onnxruntime_providers_webgpu_dll_deps}" "$<TARGET_FILE_DIR:onnxruntime_providers_webgpu>"
+      COMMAND_EXPAND_LISTS
+      VERBATIM )
+  endif()
+
   set_target_properties(onnxruntime_providers_webgpu PROPERTIES FOLDER "ONNXRuntime")

cmake/onnxruntime_unittests.cmake

@@ -525,6 +525,9 @@ set (onnxruntime_global_thread_pools_test_SRC
 set (onnxruntime_webgpu_external_dawn_test_SRC
           ${TEST_SRC_DIR}/webgpu/external_dawn/main.cc)
 
+set (onnxruntime_webgpu_delay_load_test_SRC
+          ${TEST_SRC_DIR}/webgpu/delay_load/main.cc)
+
 # tests from lowest level library up.
 # the order of libraries should be maintained, with higher libraries being added first in the list
 
@@ -1864,4 +1867,13 @@ if (onnxruntime_USE_WEBGPU AND onnxruntime_USE_EXTERNAL_DAWN)
   onnxruntime_add_include_to_target(onnxruntime_webgpu_external_dawn_test dawn::dawncpp_headers dawn::dawn_headers)
 endif()
 
+if (onnxruntime_USE_WEBGPU AND WIN32 AND onnxruntime_BUILD_SHARED_LIB AND NOT CMAKE_SYSTEM_NAME STREQUAL "Emscripten" AND NOT onnxruntime_MINIMAL_BUILD)
+  AddTest(DYN
+          TARGET onnxruntime_webgpu_delay_load_test
+          SOURCES ${onnxruntime_webgpu_delay_load_test_SRC}
+          LIBS ${SYS_PATH_LIB}
+          DEPENDS ${all_dependencies}
+  )
+endif()
+
 include(onnxruntime_fuzz_test.cmake)

cmake/winml.cmake

@@ -782,7 +782,7 @@ add_dependencies(winml_dll winml_api_native)
 add_dependencies(winml_dll winml_api_native_internal)
 
 # Link libraries
-target_link_libraries(winml_dll PRIVATE re2)
+target_link_libraries(winml_dll PRIVATE re2::re2)
 target_link_libraries(winml_dll PRIVATE ${WIL_TARGET})
 target_link_libraries(winml_dll PRIVATE winml_lib_api)
 if (NOT winml_is_inbox)

docs/Coding_Conventions_and_Standards.md

@@ -164,22 +164,16 @@ dependencies to run linters locally.
 If you want to see what lintrunner init will install, run
 `lintrunner init --dry-run`.
 
-To lint local changes:
-
-```bash
-lintrunner
-```
-
-To format files and apply suggestions:
+To format local changes:
 
 ```bash
 lintrunner -a
 ```
 
-To lint all files:
+To format all files:
 
 ```bash
-lintrunner --all-files
+lintrunner -a --all-files
 ```
 
 To show help text:

docs/ContribOperators.md

@@ -1625,7 +1625,7 @@ This version of the operator has been available since version 1 of the 'com.micr
 #### Type Constraints
 
 <dl>
-<dt><tt>T</tt> : tensor(int8), tensor(int16), tensor(int32), tensor(int64), tensor(uint8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(float16), tensor(float), tensor(double)</dt>
+<dt><tt>T</tt> : tensor(int8), tensor(int16), tensor(int32), tensor(int64), tensor(uint8), tensor(uint16), tensor(uint32), tensor(uint64), tensor(float16), tensor(float), tensor(double), tensor(bfloat16)</dt>
 <dd>Constrain input and output types.</dd>
 </dl>
 

include/onnxruntime/core/framework/op_kernel.h

@@ -7,6 +7,7 @@
 
 // It is safe to include the below header even if SHARED_PROVIDER macro is enabled
 // as it doesn't include any pb headers.
+#include "core/framework/buffer_deleter.h"
 #include "core/framework/prepacked_weights_container.h"
 
 #ifndef SHARED_PROVIDER