Update base for Update on "[ET-VK][Ops] affine quantization operators registration"

# Context
In order to enable dynamic quantization, especially for the source transform method using `Int8DynActInt4WeightQuantizer` we need to have vulkan versions for `quantize_affine`, `dequantize_affine`, and `choose_qparams_affine`. Currently we do not have a shader that performs block-based quantization as expected from these shaders, so we delegate to the per_tensor variant just to get unblocked. At a later stage, this will likely be developed more on in order to ensure we don't get too much accuracy loss.

# Changes
This creates a schema reference in the TorchAO library for out variants of these respective operators. Then there is a VK_REGISTER_OP done on them to ensure that we can properly register them when lowering the ET model with vulkan. Also the vulkan_quantizer is changed a bit in this to enable a dynamic quantization config so that we aren't purely working with just static quantization anymore.

Furthermore, we have `_annotate_for_static_quantization_config` for parity/legacy reasons, and we simply create an equivalent dynamic quantization config method.

We also changed `Linear.cpp`, particularly to allow a passthrough for weight_data since during dynamic quantization it's possible that it'll be a tensor_data than tensor_ref.

Differential Revision: [D78035354](https://our.internmc.facebook.com/intern/diff/D78035354/)

[ghstack-poisoned]

Author: morelos

.Package.swift/kernels_custom/dummy.swift renamed to .Package.swift/kernels_llm/dummy.swift

@@ -0,0 +1,13 @@
+
+set -ex
+
+install_emscripten() {
+    git clone https://github.com/emscripten-core/emsdk.git
+    pushd emsdk || return
+    ./emsdk install 4.0.10
+    ./emsdk activate 4.0.10
+    source ./emsdk_env.sh
+    popd || return
+}
+
+install_emscripten

.Package.swift/kernels_custom_debug/dummy.swift renamed to .Package.swift/kernels_llm_debug/dummy.swift

@@ -35,15 +35,15 @@ cmake -DPYTHON_EXECUTABLE=python \
     -DEXECUTORCH_BUILD_XNNPACK=OFF \
     -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON \
     -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON \
-    -DEXECUTORCH_BUILD_KERNELS_CUSTOM=ON \
+    -DEXECUTORCH_BUILD_KERNELS_LLM=ON \
     -Bcmake-out .
 cmake --build cmake-out -j16 --target install --config Release
 
 # Install llama runner with torchao
 cmake -DPYTHON_EXECUTABLE=python \
     -DBUILD_TESTING=OFF \
     -DCMAKE_BUILD_TYPE=Release \
-    -DEXECUTORCH_BUILD_KERNELS_CUSTOM=ON \
+    -DEXECUTORCH_BUILD_KERNELS_LLM=ON \
     -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON \
     -DEXECUTORCH_BUILD_XNNPACK=OFF \
     -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON \

.ci/scripts/setup-emscripten.sh

@@ -39,7 +39,7 @@ EXECUTORCH_COMMON_CMAKE_ARGS="                      \
         -DEXECUTORCH_BUILD_EXTENSION_DATA_LOADER=ON \
         -DEXECUTORCH_BUILD_EXTENSION_FLAT_TENSOR=ON \
         -DEXECUTORCH_BUILD_EXTENSION_TENSOR=ON      \
-        -DEXECUTORCH_BUILD_KERNELS_CUSTOM=ON        \
+        -DEXECUTORCH_BUILD_KERNELS_LLM=ON        \
         -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON     \
         -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON     \
         -DEXECUTORCH_BUILD_XNNPACK=ON               \
@@ -69,7 +69,7 @@ LLAVA_COMMON_CMAKE_ARGS="                        \
         -DPYTHON_EXECUTABLE="$PYTHON_EXECUTABLE" \
         -DCMAKE_INSTALL_PREFIX=${BUILD_DIR}      \
         -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE}   \
-        -DEXECUTORCH_BUILD_KERNELS_CUSTOM=ON     \
+        -DEXECUTORCH_BUILD_KERNELS_LLM=ON     \
         -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON  \
         -DEXECUTORCH_BUILD_XNNPACK=ON"
 

.ci/scripts/test_llama_torchao_lowbit.sh

@@ -33,7 +33,7 @@ cmake_install_executorch_libraries() {
       -DEXECUTORCH_BUILD_XNNPACK=ON \
       -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON \
       -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON \
-      -DEXECUTORCH_BUILD_KERNELS_CUSTOM=ON \
+      -DEXECUTORCH_BUILD_KERNELS_LLM=ON \
       -B${BUILD_DIR} .
 
   cmake --build ${BUILD_DIR} -j${NPROC} --target install --config ${BUILD_TYPE}
@@ -43,7 +43,7 @@ cmake_build_phi_3_mini() {
   cmake -DPYTHON_EXECUTABLE=$PYTHON_EXECUTABLE \
       -DCMAKE_INSTALL_PREFIX=${BUILD_DIR} \
       -DCMAKE_BUILD_TYPE=${BUILD_TYPE} \
-      -DEXECUTORCH_BUILD_KERNELS_CUSTOM=ON \
+      -DEXECUTORCH_BUILD_KERNELS_LLM=ON \
       -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON \
       -DEXECUTORCH_BUILD_XNNPACK=ON \
       -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON \

.ci/scripts/test_llava.sh

@@ -152,7 +152,7 @@ jobs:
           "backend_coreml"
           "backend_mps"
           "backend_xnnpack"
-          "kernels_custom"
+          "kernels_llm"
           "kernels_optimized"
           "kernels_quantized"
           "threadpool"

.ci/scripts/test_phi_3_mini.sh

@@ -734,3 +734,31 @@ jobs:
 
         PYTHON_EXECUTABLE=python bash .ci/scripts/setup-openvino.sh
         PYTHON_EXECUTABLE=python bash .ci/scripts/test_openvino.sh
+
+  test-build-wasm-linux:
+    name: test-build-wasm-linux
+    uses: pytorch/test-infra/.github/workflows/linux_job_v2.yml@main
+    permissions:
+      id-token: write
+      contents: read
+    strategy:
+      fail-fast: false
+    with:
+      runner: linux.2xlarge
+      docker-image: executorch-ubuntu-22.04-clang12
+      submodules: 'recursive'
+      ref: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
+      timeout: 90
+      script: |
+        # The generic Linux job chooses to use base env, not the one setup by the image
+        CONDA_ENV=$(conda env list --json | jq -r ".envs | .[-1]")
+        conda activate "${CONDA_ENV}"
+
+        BUILD_TOOL="cmake"
+        PYTHON_EXECUTABLE=python bash .ci/scripts/setup-linux.sh --build-tool "${BUILD_TOOL}"
+
+        # Install Node.js and Emscripten
+        source .ci/scripts/setup-emscripten.sh
+
+        # Test selective build
+        PYTHON_EXECUTABLE=python bash examples/wasm/test_build_wasm.sh

.github/workflows/apple.yml

@@ -645,7 +645,7 @@ jobs:
           -DEXECUTORCH_BUILD_XNNPACK=ON \
           -DEXECUTORCH_BUILD_KERNELS_QUANTIZED=ON \
           -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON \
-          -DEXECUTORCH_BUILD_KERNELS_CUSTOM=ON \
+          -DEXECUTORCH_BUILD_KERNELS_LLM=ON \
           -DEXECUTORCH_BUILD_DEVTOOLS=ON \
           -DEXECUTORCH_ENABLE_EVENT_TRACER=ON \
           -Bcmake-out .

.github/workflows/pull.yml

@@ -27,7 +27,7 @@
 	url = https://github.com/pytorch/cpuinfo.git
 [submodule "backends/xnnpack/third-party/pthreadpool"]
 	path = backends/xnnpack/third-party/pthreadpool
-	url = https://github.com/Maratyszcza/pthreadpool.git
+	url = https://github.com/google/pthreadpool.git
 [submodule "extension/llm/tokenizers"]
 	path = extension/llm/tokenizers
 	url = https://github.com/pytorch-labs/tokenizers.git