Merge branch 'main' into gasoonjia/drop-cuda-12.6

Author: Gasoonjia

.ci/scripts/test-native.sh

@@ -20,14 +20,18 @@ case "${component}" in
     test_regex='^native_runtime_test$'
     ;;
   cpu)
+    # The CPU plan test covers ETDump tracing, including operator shapes.
     cmake_args+=(
       -DEXECUTORCH_BUILD_NATIVE_CPU=ON
       -DEXECUTORCH_BUILD_XNNPACK=ON
       -DEXECUTORCH_BUILD_KERNELS_OPTIMIZED=ON
       -DEXECUTORCH_BUILD_NATIVE_VISION_RUNNER=ON
+      -DEXECUTORCH_BUILD_DEVTOOLS=ON
+      -DEXECUTORCH_ENABLE_EVENT_TRACER=ON
+      -DEXECUTORCH_CPU_XNNPACK_OPERATOR_SHAPES=ON
     )
-    targets+=(native_cpu_engine_test run_classification_cpu)
-    test_regex='^native_cpu_engine_test$'
+    targets+=(native_cpu_engine_test cpu_plan_test run_classification_cpu)
+    test_regex='^(native_cpu_engine|cpu_plan)_test$'
     ;;
   ggml)
     cmake_args+=(
@@ -68,7 +72,7 @@ cmake -S . -B "${build_dir}" -G Ninja \
 cmake --build "${build_dir}" \
   --parallel "${build_jobs}" \
   --target "${targets[@]}"
-ctest --test-dir "${build_dir}/backends/native/test" \
+ctest --test-dir "${build_dir}" \
   -R "${test_regex}" --output-on-failure --no-tests=error --timeout 120
 
 if [[ "${component}" == "ggml" ]]; then

.claude/faq.md

@@ -7,7 +7,7 @@ Error codes defined in `runtime/core/error.h`.
 |------|------|--------------|
 | 0x10 | InvalidArgument | Input shape mismatch - inputs don't match export shapes. Use dynamic shapes if needed. |
 | 0x14 | OperatorMissing | Selective build missing operator. Regenerate `et_operator_library` from current model. |
-| 0x20 | NotFound | Missing backend. Link with `--whole-archive`: `-Wl,--whole-archive libxnnpack_backend.a -Wl,--no-whole-archive` |
+| 0x20 | NotFound | Missing backend. Link with `--whole-archive`: `-Wl,--whole-archive libxnnpack_backend.a -Wl,--no-whole-archive libxnnpack_profiler.a` |
 
 ## Export Issues
 

.claude/skills/building/SKILL.md

@@ -202,7 +202,7 @@ Most commonly needed flags (full list: `CMakeLists.txt`):
 | Artifact | Location |
 |----------|----------|
 | Core runtime | `cmake-out/lib/libexecutorch.a` |
-| XNNPACK backend | `cmake-out/lib/libxnnpack_backend.a` |
+| XNNPACK backend | `cmake-out/lib/libxnnpack_backend.a` and `cmake-out/lib/libxnnpack_profiler.a` |
 | executor_runner | `cmake-out/executor_runner` (Ninja/Make) or `cmake-out/Release/executor_runner` (Xcode) |
 | Model runners | `cmake-out/examples/models/<model>/<runner>` |
 

.github/workflows/_test_cortex_m_ops.yml

@@ -59,6 +59,6 @@ jobs:
         # op set, implementation tests check FVP numerics. Both are parametrized
         # over --cortex-m-target, so a future target-dependent lowering change is
         # caught here. (cortex-m55 runs on pull via the full-suite job.)
-        pytest --config-file=backends/arm/test/pytest.ini \
+        pytest -c backends/arm/test/pytest.ini \
           backends/cortex_m/test/ops \
           --cortex-m-target=${{ matrix.target }}

.github/workflows/cuda.yml

@@ -48,6 +48,7 @@ on:
       - exir/passes/insert_write_back_for_buffers_pass.py
       - exir/passes/propagate_device_config.py
       - exir/passes/propagate_device_pass.py
+      - exir/passes/replace_complex_with_real_pass.py
   workflow_dispatch:
     inputs:
       models:
@@ -476,9 +477,9 @@ jobs:
   unittest-cuda:
     name: unittest-cuda (${{ matrix.shard }})
     needs: [changed-files, run-decision]
-    # This job runs the pybindings device test and the device placement test,
-    # so it also has to fire on the pybindings and device pass paths the
-    # workflow filter lists.
+    # This job runs the pybindings device test, the device placement test and
+    # the complex number tests, so it also has to fire on the pybindings and
+    # exir pass paths the workflow filter lists.
     if: |
       contains(needs.changed-files.outputs.changed-files, 'backends/cuda') ||
       contains(needs.changed-files.outputs.changed-files, 'backends/aoti') ||
@@ -493,6 +494,7 @@ jobs:
       contains(needs.changed-files.outputs.changed-files, 'exir/passes/insert_write_back_for_buffers_pass.py') ||
       contains(needs.changed-files.outputs.changed-files, 'exir/passes/propagate_device_config.py') ||
       contains(needs.changed-files.outputs.changed-files, 'exir/passes/propagate_device_pass.py') ||
+      contains(needs.changed-files.outputs.changed-files, 'exir/passes/replace_complex_with_real_pass.py') ||
       needs.run-decision.outputs.is-full-run == 'true'
     # kernel: Triton kernels and their autotuning. system: lowering passes, op
     # dispatch, export/runtime, and model tests. Together they do not fit the

.github/workflows/pull.yml

@@ -859,7 +859,7 @@ jobs:
         backends/cortex_m/test/build_test_runner.sh
 
         # To run cortex_m tests
-        pytest --config-file=backends/arm/test/pytest.ini backends/cortex_m/test
+        pytest -c backends/arm/test/pytest.ini backends/cortex_m/test
 
   android:
     uses: ./.github/workflows/_android.yml
@@ -2084,7 +2084,8 @@ jobs:
         # Custom operator tests
         PYTHON_EXECUTABLE=python bash backends/vulkan/test/ops/benchmark_v2/build.sh
         for op in binary/add gemm/linear/qw/q8csw convolution/qw/q8csw gemm/linear/qw/q4gsw \
-                  quantization/choose_qparams_per_row quantize/q8ta clone/q8ta binary/q8ta; do
+                  quantization/choose_qparams_per_row quantize/q8ta clone/q8ta binary/q8ta \
+                  warp; do
           ./cmake-out/backends/vulkan/test/ops/benchmark_v2/etvk_ops_bench --validation_only \
             --config=backends/vulkan/test/ops/benchmark_v2/operators/${op}/cases/correctness.json
         done

.github/workflows/test-native.yml

@@ -33,13 +33,15 @@ on:
       - backends/vulkan_shared/runtime/**
       - backends/xnnpack/CMakeLists.txt
       - backends/xnnpack/cmake/**
+      - backends/xnnpack/runtime/profiling/**
       - backends/xnnpack/third-party/FXdiv
       - backends/xnnpack/third-party/XNNPACK
       - backends/xnnpack/third-party/cpuinfo
       - backends/xnnpack/third-party/pthreadpool
       - examples/models/llama/tokenizer/**
       - extension/llm/tokenizers
       - extension/llm/export/**
+      - devtools/etdump/**
       - exir/**
       - configurations/**
       - extension/memory_allocator/**

.github/workflows/trunk.yml

@@ -1143,6 +1143,9 @@ jobs:
         trap 'bash .ci/scripts/cleanup-samsung-linux-deps.sh' EXIT TERM INT
         source .ci/scripts/setup-samsung-linux-deps.sh
 
+        # Graph-level pass unit tests (no device needed)
+        python -m unittest discover -s backends/samsung/test/passes -p "test_*.py"
+
         # Check if device was reserved
         if [[ "${DEVICE_ACQUIRED:-0}" != "1" ]]; then
           echo "::warning::Skipping tests - no Samsung device available"

.github/workflows/vulkan.yml

@@ -123,7 +123,8 @@ jobs:
         # selected above.
         PYTHON_EXECUTABLE=python bash backends/vulkan/test/ops/benchmark_v2/build.sh
         for op in binary/add gemm/linear/qw/q8csw convolution/qw/q8csw gemm/linear/qw/q4gsw \
-                  quantization/choose_qparams_per_row quantize/q8ta clone/q8ta binary/q8ta; do
+                  quantization/choose_qparams_per_row quantize/q8ta clone/q8ta binary/q8ta \
+                  warp; do
           ./cmake-out/backends/vulkan/test/ops/benchmark_v2/etvk_ops_bench --validation_only \
             --config=backends/vulkan/test/ops/benchmark_v2/operators/${op}/cases/correctness.json
         done

.lintrunner.toml

@@ -397,6 +397,8 @@ exclude_patterns = [
     # muse_glimmer is a symlink to this directory; the linter resolves it and
     # then fails on a directory.
     'examples/models/muse-glimmer',
+    # src/executorch/frontends is a symlink to this directory.
+    'frontends',
     'backends/cadence/utils/FACTO',
     'examples/cuda',
     'examples/qualcomm',