Cherry-picks to the release branch (microsoft#16017)

### Description
Cherry-picks 26 commits to the release branch. 
Most cherry-picks are clean merges. Except:

1. When I got conflicts in cgmanifest.json and download-deps.yml, I
choose to ignore the conflicts and regenerate the two files
2. There were some conflicts in cmake/deps.txt, onnxruntime_c_api.cc


PR list:

[js/webgpu] fix Transpose with non-float tensor (microsoft#15819)
[js/web] fix terser reserved symbols for worker (microsoft#15864)
[JSEP] fix constructor for OrtDevice (microsoft#15805)
Bump engine.io from 6.4.1 to 6.4.2 in /js/web (microsoft#15799)
Bump engine.io from 6.4.0 to 6.4.2 in /onnxruntime/test/wasm (microsoft#15798)
[wasm] revert emsdk to v3.1.19 (microsoft#15793)
[wasm/JSEP] add threaded build to artifacts (microsoft#15777)
[js/web] add target ort.webgpu.min.js (microsoft#15780)
update ort extensions to 94142d8391c9791ec71c38336436319a2d4ac7a0 (microsoft#15688)
fix: setting builder optimization level to TRT 8.6 default (microsoft#15897)
Adust GetVersionString() GetBuildInfoString() signatures and move them to OrtApi (microsoft#15921)
Fix segfault for multiple GPU run (regression) (microsoft#15823)
android package fix (microsoft#15999)
[CoreML EP] Minor changes to allow CoreML EP to handle more nodes and models. (microsoft#15993)
Adding support for conv fp16 fusion on Resnet50v1 (microsoft#15474)
update onnx release 1.14 for docker files (microsoft#15680)
Avoid generating training documentation during packaging (microsoft#15795)
Update Conv-Add-Relu Fusion Transformation (microsoft#15834)
Fix symbolic shape infer empty value_info (microsoft#15842)
NhwcFusedConv: Add before Activation (microsoft#15837)
use __hmul2 instead of __hmul2_rn (microsoft#15852)
change the EP device to default OrtDevice() for memoryType equals CPU Input (microsoft#15903)
Fixing NhwcFusedConv fp16 (microsoft#15950)
fix topo sort in quantization tool (microsoft#16003)
[doc] add LeakyRelu to coreml supported ops (microsoft#15944)
[DML EP] Add frequent upload heap flushing (microsoft#15960)

Co-authored-by: Yulong Wang 
Co-authored-by: dependabot[bot] 
Co-authored-by: Guenther Schmuelling 
Co-authored-by: Shalva Mist 
Co-authored-by: Maximilian Müller 
Co-authored-by: Dmitri Smirnov 
Co-authored-by: pengwa 
Co-authored-by: Ashwini Khade 
Co-authored-by: Edward Chen 
Co-authored-by: Jian Chen 
Co-authored-by: liqun Fu 
Co-authored-by: Baiju Meswani 
Co-authored-by: Tianlei Wu 
Co-authored-by: Chen Fu 
Co-authored-by: Ye Wang 
Co-authored-by: cao lei 
Co-authored-by: Yufeng Li 
Co-authored-by: Rachel Guo 
Co-authored-by: Patrice Vignola

cgmanifests/generated/cgmanifest.json

@@ -112,7 +112,7 @@
       "component": {
         "type": "git",
         "git": {
-          "commitHash": "9b7bca2a723ff94edcd007d93b5d0cf1838591dc",
+          "commitHash": "a0d77f18516d2da7468a96b0de3b737266f23176",
           "repositoryUrl": "https://github.com/onnx/onnx.git"
         },
         "comments": "git submodule at cmake/external/onnx"
@@ -288,16 +288,6 @@
         "comments": "mp11"
       }
     },
-    {
-      "component": {
-        "type": "git",
-        "git": {
-          "commitHash": "3b58938e025c41d2fcd89fa22028eefaa81a18ad",
-          "repositoryUrl": "https://github.com/onnx/onnx.git"
-        },
-        "comments": "onnx"
-      }
-    },
     {
       "component": {
         "type": "git",
@@ -448,6 +438,16 @@
         "comments": "triton"
       }
     },
+    {
+      "component": {
+        "type": "git",
+        "git": {
+          "commitHash": "94142d8391c9791ec71c38336436319a2d4ac7a0",
+          "repositoryUrl": "https://github.com/microsoft/onnxruntime-extensions.git"
+        },
+        "comments": "extensions"
+      }
+    },
     {
       "component": {
         "type": "git",

cmake/deps.txt

@@ -23,7 +23,7 @@ microsoft_gsl;https://github.com/microsoft/GSL/archive/refs/tags/v4.0.0.zip;cf36
 microsoft_wil;https://github.com/microsoft/wil/archive/5f4caba4e7a9017816e47becdd918fcc872039ba.zip;fd119887d0d17c37adf1fc227b054befa28158ad
 mimalloc;https://github.com/microsoft/mimalloc/archive/refs/tags/v2.1.1.zip;d5ee7d34223d0567892db5179849939c8769dc41
 mp11;https://github.com/boostorg/mp11/archive/refs/tags/boost-1.79.0.zip;c8f04e378535ededbe5af52c8f969d2dedbe73d5
-onnx;https://github.com/onnx/onnx/archive/3b58938e025c41d2fcd89fa22028eefaa81a18ad.zip;e0e5dda9eea5cd5ecae3bd8be86e477016b6be02
+onnx;https://github.com/onnx/onnx/archive/refs/tags/v1.14.0.zip;3c6e43a36f94addc15afe939860127a1d74a9488
 #use the last commit of 8.6-EA branch (https://github.com/onnx/onnx-tensorrt/commit/ba6a4fb34fdeaa3613bf981610c657e7b663a699)
 onnx_tensorrt;https://github.com/onnx/onnx-tensorrt/archive/ba6a4fb34fdeaa3613bf981610c657e7b663a699.zip;5a474ed86e2c4ee4085d3daeff8222866e933dc0
 protobuf;https://github.com/protocolbuffers/protobuf/archive/refs/tags/v21.12.zip;7cf2733949036c7d52fda017badcab093fe73bfa
@@ -48,5 +48,5 @@ b64;https://github.com/libb64/libb64/archive/refs/tags/v2.0.0.1.zip;815b6d31d50d
 pthread;https://sourceforge.net/projects/pthreads4w/files/pthreads4w-code-v3.0.0.zip;3b9e417e4474c34542b76ad40529e396ac109fb4
 triton;https://github.com/triton-inference-server/server/archive/refs/tags/v2.28.0.zip;4b305570aa1e889946e20e36050b6770e4108fee
 # above are deps introduced by triton client, might remove after 1.14 release
-extensions;https://github.com/microsoft/onnxruntime-extensions/archive/81e7799c69044c745239202085eb0a98f102937b.zip;d53487035174a046628359289ad27aa0ac0380c9
-curl;https://github.com/curl/curl/archive/refs/tags/curl-8_0_1.zip;b16d1fa8ee567b52c09a0f89940b07d8491b881d
+extensions;https://github.com/microsoft/onnxruntime-extensions/archive/94142d8391c9791ec71c38336436319a2d4ac7a0.zip;4365ac5140338b4cb75a39944a4be276e3829b3c
+curl;https://github.com/curl/curl/archive/refs/tags/curl-8_0_1.zip;b16d1fa8ee567b52c09a0f89940b07d8491b881d

cmake/onnxruntime.cmake

@@ -28,9 +28,9 @@ macro(get_mobile_api_headers _HEADERS)
   )
 
   if (onnxruntime_ENABLE_TRAINING_APIS)
-    list(APPEND ${_HEADERS} "${REPO_ROOT/orttraining/orttraining/training_api/include/onnxruntime_training_c_api.h}")
-    list(APPEND ${_HEADERS} "${REPO_ROOT/orttraining/orttraining/training_api/include/onnxruntime_training_cxx_api.h}")
-    list(APPEND ${_HEADERS} "${REPO_ROOT/orttraining/orttraining/training_api/include/onnxruntime_training_cxx_inline_api.h}")
+    list(APPEND ${_HEADERS} "${REPO_ROOT}/orttraining/orttraining/training_api/include/onnxruntime_training_c_api.h")
+    list(APPEND ${_HEADERS} "${REPO_ROOT}/orttraining/orttraining/training_api/include/onnxruntime_training_cxx_api.h")
+    list(APPEND ${_HEADERS} "${REPO_ROOT}/orttraining/orttraining/training_api/include/onnxruntime_training_cxx_inline.h")
   endif()
 
   # need to add header files for enabled EPs

cmake/onnxruntime_config.h.in

@@ -21,5 +21,5 @@
 #cmakedefine HAS_FORMAT_TRUNCATION
 #cmakedefine HAS_BITWISE_INSTEAD_OF_LOGICAL
 #cmakedefine HAS_REALLOCARRAY
-#cmakedefine ORT_VERSION "@ORT_VERSION@"
-#cmakedefine ORT_BUILD_INFO "@ORT_BUILD_INFO@"
+#cmakedefine ORT_VERSION u8"@ORT_VERSION@"
+#cmakedefine ORT_BUILD_INFO u8"@ORT_BUILD_INFO@"

csharp/src/Microsoft.ML.OnnxRuntime/NativeMethods.shared.cs

@@ -286,6 +286,7 @@ public struct OrtApi
         public IntPtr CastTypeInfoToOptionalTypeInfo;
         public IntPtr GetOptionalContainedTypeInfo;
         public IntPtr GetResizedStringTensorElementBuffer;
+        public IntPtr KernelContext_GetAllocator;
     }
 
     internal static class NativeMethods
@@ -1654,7 +1655,6 @@ IntPtr[] outputValues /* An array of output value pointers. Array must be alloca
 
         public static DOrtModelMetadataLookupCustomMetadataMap OrtModelMetadataLookupCustomMetadataMap;
 
-
         /// <summary>
         /// Frees ModelMetadata instance
         /// </summary>

csharp/src/Microsoft.ML.OnnxRuntime/OrtEnv.shared.cs

@@ -276,7 +276,7 @@ public static OrtEnv Instance()
         /// <param name="options"></param>
         /// <returns></returns>
         /// <exception cref="OnnxRuntimeException">if the singleton has already been created</exception>
-        public static OrtEnv CreateInstanceWithOptions(EnvironmentCreationOptions options)
+        public static OrtEnv CreateInstanceWithOptions(ref EnvironmentCreationOptions options)
         {
             // Non-thread safe, best effort hopefully helpful check.
             // Environment is usually created once per process, so this should be fine.

csharp/test/Microsoft.ML.OnnxRuntime.Tests.Common/OrtEnvTests.cs

@@ -80,7 +80,7 @@ public void TestUpdatingEnvWithCustomLogLevel()
                 logLevel = OrtLoggingLevel.ORT_LOGGING_LEVEL_FATAL
             };
 
-            ortEnvInstance = OrtEnv.CreateInstanceWithOptions(envOptions);
+            ortEnvInstance = OrtEnv.CreateInstanceWithOptions(ref envOptions);
             Assert.True(OrtEnv.IsCreated);
             Assert.Equal(OrtLoggingLevel.ORT_LOGGING_LEVEL_FATAL, ortEnvInstance.EnvLogLevel);
 
@@ -92,7 +92,7 @@ public void TestUpdatingEnvWithCustomLogLevel()
                 logId = "CSharpOnnxRuntimeTestLogid"
             };
 
-            ortEnvInstance = OrtEnv.CreateInstanceWithOptions(envOptions);
+            ortEnvInstance = OrtEnv.CreateInstanceWithOptions(ref envOptions);
             Assert.Equal(OrtLoggingLevel.ORT_LOGGING_LEVEL_WARNING, ortEnvInstance.EnvLogLevel);
 
             // Change and see if this takes effect
@@ -118,7 +118,7 @@ public void TestUpdatingEnvWithThreadingOptions()
                 };
 
                 // Make sure we start anew
-                var env = OrtEnv.CreateInstanceWithOptions(envOptions);
+                var env = OrtEnv.CreateInstanceWithOptions(ref envOptions);
                 Assert.True(OrtEnv.IsCreated);
             }
         }
@@ -164,7 +164,7 @@ public void TesEnvWithCustomLogger()
 
             LoggingInvokes = 0;
 
-            var env = OrtEnv.CreateInstanceWithOptions(envOptions);
+            var env = OrtEnv.CreateInstanceWithOptions(ref envOptions);
             Assert.True(OrtEnv.IsCreated);
 
             var model = TestDataLoader.LoadModelFromEmbeddedResource("squeezenet.onnx");
@@ -199,7 +199,7 @@ public void TestEnvWithCustomLoggerAndThredingOptions()
 
                 LoggingInvokes = 0;
 
-                var env = OrtEnv.CreateInstanceWithOptions(envOptions);
+                var env = OrtEnv.CreateInstanceWithOptions(ref envOptions);
                 Assert.True(OrtEnv.IsCreated);
 
                 var model = TestDataLoader.LoadModelFromEmbeddedResource("squeezenet.onnx");

include/onnxruntime/core/providers/tensorrt/tensorrt_provider_options.h

@@ -36,7 +36,7 @@ struct OrtTensorRTProviderOptionsV2 {
   int trt_detailed_build_log;                   // Enable detailed build step logging on TensorRT EP with timing for each engine build. Default 0 = false, nonzero = true
   int trt_build_heuristics_enable;              // Build engine using heuristics to reduce build time. Default 0 = false, nonzero = true
   int trt_sparsity_enable;                      // Control if sparsity can be used by TRT. Default 0 = false, 1 = true
-  int trt_builder_optimization_level;           // Set the builder optimization level. WARNING: levels below 2 do not guarantee good engine performance, but greatly improve build time.  Default 2, valid range [0-4]
+  int trt_builder_optimization_level;           // Set the builder optimization level. WARNING: levels below 3 do not guarantee good engine performance, but greatly improve build time.  Default 3, valid range [0-5]
   int trt_auxiliary_streams;                    // Set maximum number of auxiliary streams per inference stream. Setting this value to 0 will lead to optimal memory usage. Default -1 = heuristics
   const char* trt_tactic_sources;               // pecify the tactics to be used by adding (+) or removing (-) tactics from the default
                                                 // tactic sources (default = all available tactics) e.g. "-CUDNN,+CUBLAS" available keys: "CUBLAS"|"CUBLAS_LT"|"CUDNN"|"EDGE_MASK_CONVOLUTIONS"

include/onnxruntime/core/session/onnxruntime_c_api.h

@@ -95,6 +95,8 @@ extern "C" {
 #define ORTCHAR_T char
 #endif
 
+/// ORTCHAR_T, ORT_TSTR are reserved specifically for path handling.
+/// All other strings are UTF-8 encoded, use char and std::string
 #ifndef ORT_TSTR
 #ifdef _WIN32
 #define ORT_TSTR(X) L##X
@@ -627,11 +629,19 @@ struct OrtApiBase {
    * \param[in] version Must be ::ORT_API_VERSION
    * \return The ::OrtApi for the version requested, nullptr will be returned if this version is unsupported, for example when using a runtime
    *   older than the version created with this header file.
+   *
+   * One can call GetVersionString() to get the version of the Onnxruntime library for logging
+   * and error reporting purposes.
    */
   const OrtApi*(ORT_API_CALL* GetApi)(uint32_t version)NO_EXCEPTION;
-  const ORTCHAR_T*(ORT_API_CALL* GetVersionString)(void)NO_EXCEPTION;    ///< Returns a null terminated string of the version of the Onnxruntime library (eg: "1.8.1")
-  const ORTCHAR_T*(ORT_API_CALL* GetBuildInfoString)(void)NO_EXCEPTION;  ///< Returns a null terminated string of the build info including git info and cxx flags
+
+  /** \brief Returns a null terminated string of the version of the Onnxruntime library (eg: "1.8.1")
+   *
+   *  \return UTF-8 encoded version string. Do not deallocate the returned buffer.
+   */
+  const char*(ORT_API_CALL* GetVersionString)(void)NO_EXCEPTION;
 };
+
 typedef struct OrtApiBase OrtApiBase;
 
 /** \brief The Onnxruntime library's entry point to access the C API
@@ -4192,6 +4202,14 @@ struct OrtApi {
    * \since Version 1.15.
    */
   ORT_API2_STATUS(KernelContext_GetAllocator, _In_ const OrtKernelContext* context, _In_ const OrtMemoryInfo* mem_info, _Outptr_ OrtAllocator** out);
+
+  /** \brief Returns a null terminated string of the build info including git info and cxx flags
+   *
+   * \return UTF-8 encoded version string. Do not deallocate the returned buffer.
+   *
+   * \since Version 1.15.
+   */
+  const char*(ORT_API_CALL* GetBuildInfoString)(void);
 };
 
 /*

include/onnxruntime/core/session/onnxruntime_cxx_api.h

@@ -125,14 +125,14 @@ inline const OrtApi& GetApi() noexcept { return *Global<void>::api_; }
 /// This function returns the onnxruntime version string
 /// </summary>
 /// <returns>version string major.minor.rev</returns>
-std::basic_string<ORTCHAR_T> GetVersionString();
+std::string GetVersionString();
 
 /// <summary>
 /// This function returns the onnxruntime build information: including git branch,
 /// git commit id, build type(Debug/Release/RelWithDebInfo) and cmake cpp flags.
 /// </summary>
 /// <returns>string</returns>
-std::basic_string<ORTCHAR_T> GetBuildInfoString();
+std::string GetBuildInfoString();
 
 /// <summary>
 /// This is a C++ wrapper for OrtApi::GetAvailableProviders() and

include/onnxruntime/core/session/onnxruntime_cxx_inline.h

@@ -1987,14 +1987,12 @@ inline void CustomOpApi::ReleaseKernelInfo(_Frees_ptr_opt_ OrtKernelInfo* info_c
   api_.ReleaseKernelInfo(info_copy);
 }
 
-inline std::basic_string<ORTCHAR_T> GetVersionString() {
-  std::basic_string<ORTCHAR_T> result = OrtGetApiBase()->GetVersionString();
-  return result;
+inline std::string GetVersionString() {
+  return OrtGetApiBase()->GetVersionString();
 }
 
-inline std::basic_string<ORTCHAR_T> GetBuildInfoString() {
-  std::basic_string<ORTCHAR_T> result = OrtGetApiBase()->GetBuildInfoString();
-  return result;
+inline std::string GetBuildInfoString() {
+  return GetApi().GetBuildInfoString();
 }
 
 inline std::vector<std::string> GetAvailableProviders() {

java/src/main/native/ai_onnxruntime_OnnxRuntime.c

@@ -76,13 +76,7 @@ JNIEXPORT jobjectArray JNICALL Java_ai_onnxruntime_OnnxRuntime_getAvailableProvi
 JNIEXPORT jstring JNICALL Java_ai_onnxruntime_OnnxRuntime_initialiseVersion
   (JNIEnv * jniEnv, jclass clazz) {
   (void)clazz;  // required JNI parameter not needed by functions which don't access their host class.
-  const ORTCHAR_T* version = OrtGetApiBase()->GetVersionString();
+  const char* version = ORT_VERSION;
   assert(version != NULL);
-#ifdef _WIN32
-  jsize len = (jsize)(wcslen(version));
-  jstring versionStr = (*jniEnv)->NewString(jniEnv, (const jchar*)version, len);
-#else
-  jstring versionStr = (*jniEnv)->NewStringUTF(jniEnv, version);
-#endif
-  return versionStr;
+  return (*jniEnv)->NewStringUTF(jniEnv, version);
 }