Merge branch 'main' into main

Author: psiddh

.ci/scripts/bloaty-measure.sh

@@ -0,0 +1,46 @@
+#!/usr/bin/env bash
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+# Usage: bash .ci/scripts/bloaty-measure.sh <job_name> <head_elf> <strip_tool>
+#
+# Runs bloaty against the head ELF, writes metadata.json + full.txt +
+# head_only.txt to artifacts-to-be-uploaded/, and appends a markdown table
+# to $GITHUB_STEP_SUMMARY.
+#
+# Best-effort: never exits non-zero — the size jobs that source this should
+# not fail because of a bloaty hiccup.
+
+set -uo pipefail
+
+job_name=$1
+head_elf=$2
+strip_tool=$3
+head_sha=${GITHUB_HEAD_SHA:-${GITHUB_SHA:-unknown}}
+
+(
+  # conda-forge bloaty depends on a newer libstdc++ than the ubuntu-22.04
+  # docker images ship, so pull libstdcxx-ng into the same env and invoke
+  # via `conda run` so library paths are set correctly.
+  bloaty_env=/tmp/bloaty-conda-env
+  if [[ ! -x "${bloaty_env}/bin/bloaty" ]]; then
+    conda create -y -p "${bloaty_env}" -c conda-forge bloaty libstdcxx-ng || exit 1
+  fi
+  bloaty_cmd=("conda" "run" "--no-capture-output" "-p" "${bloaty_env}" "bloaty")
+  "${bloaty_cmd[@]}" --version || exit 1
+
+  tmp_out=/tmp/bloaty-out
+  rm -rf "${tmp_out}" && mkdir -p "${tmp_out}"
+  BLOATY="${bloaty_cmd[*]}" python3 .github/scripts/bloaty_diff.py measure \
+    --head "${head_elf}" \
+    --job "${job_name}" \
+    --binary-name size_test \
+    --head-sha "${head_sha}" \
+    --strip-tool "${strip_tool}" \
+    --out "${tmp_out}" || exit 1
+  mkdir -p artifacts-to-be-uploaded
+  mv "${tmp_out}"/* artifacts-to-be-uploaded/
+) || echo "bloaty report failed; continuing"

.ci/scripts/setup-webgpu-linux-deps.sh

@@ -0,0 +1,30 @@
+#!/bin/bash
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+set -ex
+
+# SwiftShader: software Vulkan adapter for GPU-less CI (LunarG SDK not needed).
+install_swiftshader() {
+  _https_amazon_aws=https://ossci-android.s3.amazonaws.com
+  _swiftshader_archive=swiftshader-abe07b943-prebuilt.tar.gz
+  _swiftshader_dir=/tmp/swiftshader
+  mkdir -p $_swiftshader_dir
+
+  _tmp_archive="/tmp/${_swiftshader_archive}"
+
+  curl --silent --show-error --location --fail --retry 3 --retry-all-errors \
+    --output "${_tmp_archive}" "$_https_amazon_aws/${_swiftshader_archive}"
+
+  tar -C "${_swiftshader_dir}" -xzf "${_tmp_archive}"
+
+  export VK_ICD_FILENAMES="${_swiftshader_dir}/swiftshader/build/Linux/vk_swiftshader_icd.json"
+  export LD_LIBRARY_PATH="${_swiftshader_dir}/swiftshader/build/Linux/:${LD_LIBRARY_PATH}"
+  export ETVK_USING_SWIFTSHADER=1
+}
+
+install_swiftshader
+bash backends/webgpu/scripts/setup-wgpu-native.sh

.ci/scripts/test_backend.sh

@@ -57,6 +57,14 @@ if [[ "$FLOW" == *vulkan* ]]; then
     EXTRA_BUILD_ARGS+=" -DEXECUTORCH_BUILD_VULKAN=ON"
 fi
 
+if [[ "$FLOW" == *webgpu* ]]; then
+    # Setup swiftshader (software Vulkan adapter for GPU-less runners) and wgpu-native,
+    # which are required to build and run the WebGPU delegate.
+    source .ci/scripts/setup-webgpu-linux-deps.sh
+
+    EXTRA_BUILD_ARGS+=" -DEXECUTORCH_BUILD_WEBGPU=ON"
+fi
+
 if [[ "$FLOW" == *arm* ]]; then
     if [[ "$SUITE" == "operators" ]]; then
         PYTEST_RETRY_ARGS=(--reruns 2 --reruns-delay 1)

.claude/skills/qualcomm/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: qualcomm
-description: Build, test, or develop the QNN (Qualcomm AI Engine Direct) backend. Use when working on backends/qualcomm/, building QNN (use backends/qualcomm/scripts/build.sh), adding new ops or passes, running QNN delegate tests, or exporting models for Qualcomm HTP/GPU targets. Also exposes a Buck-vs-CMake parity workflow — invoke as `/qualcomm buck-fix`, `/qualcomm buck-cmake fix`, `/qualcomm buck-parity`, or any user request to fix `test-qnn-buck-build-linux` CI failures or check buck/cmake drift in backends/qualcomm/.
+description: Build, test, or develop the QNN (Qualcomm AI Engine Direct) backend. Use when working on backends/qualcomm/, building QNN (use backends/qualcomm/scripts/build.sh), adding new ops or passes, running QNN delegate tests, or exporting models for Qualcomm HTP/GPU targets. Also exposes a Buck-vs-CMake parity workflow — invoke as `/qualcomm buck-fix`, `/qualcomm buck-cmake fix`, `/qualcomm buck-parity`, or any user request to fix `test-qnn-buck-build-linux` CI failures or check buck/cmake drift in backends/qualcomm/. Also covers QNN intermediate-output / per-layer accuracy debugging — trigger on phrases like "QNN accuracy issue", "QNN output doesn't match CPU", "debug per-layer for QNN", "find which QNN layer is wrong".
 ---
 
 # QNN (Qualcomm AI Engine Direct) Backend
@@ -25,6 +25,7 @@ When the user's request falls into one of these areas, read the corresponding fi
 | Model enablement | `model_enablement.md` | User asks to enable a new model end-to-end |
 | Buck vs CMake parity (pre-PR or fix red CI) | `buck_parity.md` | User changed BUCK / TARGETS / `targets.bzl` or `CMakeLists.txt` under `backends/qualcomm/`, added new `.cpp` / `.h` / `#include` there, is preparing to push a PR that touches QNN, **or** the `test-qnn-buck-build-linux` CI check on their PR is red and they want to fix it locally. Direct trigger: `/qualcomm buck-fix`. |
 | Profiling & debugging | `profiling.md` | User asks about profiling, optrace, QHAS, QAIRT Visualizer *(file TBD)* |
+| QNN intermediate-output / per-layer accuracy debugging | `qnn_intermediate_debugger.md` | User reports QNN-vs-CPU accuracy divergence, asks to debug per-layer / intermediate output for QNN, mentions `QNNIntermediateDebugger` / `QcomNumericalComparator`, or wants to find which layer causes a QNN accuracy drop. Workflow generates a new debug script from the user's existing example script. |
 
 ## Building