feat: cross-validation harness and performance regression tracking #55

Workflow file for this run

	name: CI

	on:
	push:
	branches: [main]
	pull_request:
	branches: [main]

	jobs:
	python:
	name: Python Tests & Lint
	runs-on: ubuntu-latest
	strategy:
	matrix:
	python-version: ["3.10", "3.11", "3.12"]
	steps:
	- uses: actions/checkout@v4

	- name: Set up Python ${{ matrix.python-version }}
	uses: actions/setup-python@v5
	with:
	python-version: ${{ matrix.python-version }}

	- name: Install dependencies
	run: \|
	python -m pip install --upgrade pip
	pip install -e ".[dev]"

	- name: Lint with ruff
	run: ruff check python/

	- name: Type check with mypy
	run: mypy python/arbiter/

	- name: Run tests
	run: pytest tests/python -v

	- name: Validate sample model
	run: arbiterc validate samples/battery_policy/models/battery.arb.yaml --strict

	- name: Profile validation — compile all samples on standard profile
	run: \|
	for model in samples//models/.arb.yaml; do
	echo "--- $model ---"
	arbiterc compile "$model" --profile standard --out-c /dev/null --out-h /dev/null
	done

	clang-tidy:
	name: clang-tidy Static Analysis
	runs-on: ubuntu-latest
	steps:
	- uses: actions/checkout@v4

	- name: Install clang-tidy
	run: sudo apt-get install -y --no-install-recommends clang-tidy

	- name: Run clang-tidy on engine sources
	run: \|
	clang-tidy \
	lib/arbiter_engine.c \
	lib/arbiter_eval.c \
	lib/arbiter_fact_store.c \
	lib/arbiter_trace.c \
	lib/arbiter_blob.c \
	lib/arbiter_action.c \
	-- \
	-I include \
	-isystem tools/clang-tidy-stubs \
	-std=c11 \
	-DCONFIG_ARBITER_LOG_LEVEL=3 \
	-DCONFIG_ARBITER_MAX_FACTS=64 \
	-DCONFIG_ARBITER_MAX_ACTIONS_PER_EVAL=16 \
	-DCONFIG_ARBITER_MAX_TRACE_ENTRIES=64 \
	-DCONFIG_ARBITER_MAX_TRACE_INPUTS=8

	spdx:
	name: SPDX Header Check
	runs-on: ubuntu-latest
	steps:
	- uses: actions/checkout@v4

	- name: Check SPDX headers
	run: \|
	missing=0
	for f in $(find lib/ include/ subsys/ python/ -name '.c' -o -name '.h' -o -name '*.py' \| grep -v __pycache__); do
	if ! head -5 "$f" \| grep -q 'SPDX-License-Identifier'; then
	echo "Missing SPDX header: $f"
	missing=1
	fi
	done
	exit $missing

	misra:
	name: MISRA-C Analysis
	runs-on: ubuntu-latest
	steps:
	- uses: actions/checkout@v4

	- name: Install cppcheck
	run: sudo apt-get update && sudo apt-get install -y --no-install-recommends cppcheck

	- name: Run cppcheck with MISRA C 2012 rules
	run: \|
	# cppcheck's MISRA addon may not be available on ubuntu-latest;
	# fall back to built-in checks if the addon fails.
	if cppcheck --addon=misra --version 2>/dev/null; then
	cppcheck --addon=misra \
	--suppress=misra-c2012-1.1 \
	--suppress=misra-c2012-20.9 \
	--suppress=misra-c2012-21.6 \
	-I include/ \
	-isystem tools/clang-tidy-stubs \
	--error-exitcode=1 \
	--inline-suppr \
	lib/*.c
	else
	echo "MISRA addon not available — falling back to built-in checks"
	cppcheck \
	--enable=all \
	--suppress=missingIncludeSystem \
	-I include/ \
	-isystem tools/clang-tidy-stubs \
	--error-exitcode=1 \
	--inline-suppr \
	lib/*.c
	fi

	zephyr:
	name: Zephyr Twister Tests
	runs-on: ubuntu-latest
	env:
	# native_sim uses host GCC; skip cross-compiler toolchain lookup.
	ZEPHYR_TOOLCHAIN_VARIANT: host
	# Tell Zephyr's module system where arbiter lives. west skips
	# cloning the manifest project to its declared module path, so
	# ZEPHYR_EXTRA_MODULES is the clean solution. Now that the
	# Kconfig symbol (ARBITER) and CMake guard (CONFIG_ARBITER) match,
	# and sources use CMAKE_CURRENT_LIST_DIR, this resolves fully.
	ZEPHYR_EXTRA_MODULES: ${{ github.workspace }}/app
	steps:
	- uses: actions/checkout@v4
	with:
	# west.yml declares self.path: app — match it so west module
	# resolution works correctly alongside modules/lib/arbiter/
	path: app

	- uses: actions/setup-python@v5
	with:
	python-version: "3.12"

	- name: Install west
	run: pip install west

	# Cache the Zephyr tree and west state keyed on west.yml.
	# All three paths are saved/restored together so a partial
	# hit cannot leave the workspace in an inconsistent state.
	- name: Cache west workspace
	id: cache-west
	uses: actions/cache@v4
	with:
	path: \|
	zephyr
	modules
	.west
	key: west-${{ hashFiles('app/west.yml') }}
	restore-keys: west-

	# Only run init+update on a cache miss; .west/config is
	# restored from cache and is sufficient for subsequent steps.
	- name: West init
	if: steps.cache-west.outputs.cache-hit != 'true'
	run: west init -l app

	# --narrow: only fetch projects in our top-level manifest
	# (zephyr + arbiter), skipping Zephyr's full HAL set.
	# -o=--depth=1: shallow clone to keep the runner fast.
	# native_sim needs only the Zephyr core tree, no HALs.
	- name: West update
	if: steps.cache-west.outputs.cache-hit != 'true'
	run: west update --narrow -o=--depth=1

	# native_sim/native targets 32-bit by default; gcc-multilib provides
	# the 32-bit glibc headers (bits/libc-header-start.h) on ubuntu-latest.
	- name: Install native_sim build dependencies
	run: sudo apt-get install -y --no-install-recommends gcc-multilib g++-multilib

	- name: Install Zephyr Python requirements
	run: pip install -r zephyr/scripts/requirements.txt

	# Zephyr CMake requires find_package(Zephyr-sdk) even for native_sim
	# (host tools: DTC, cmake config files). The minimal bundle is ~10 MB
	# and contains no cross-compiler — only the cmake registration files.
	# setup.sh -c writes ~/.cmake/packages/Zephyr-sdk/ so CMake finds it.
	# SDK 1.0.1 matches pinned Zephyr v4.4.0 (Zephyr v4.4.0 requires >= 1.0.0).
	# sdk-version file does not exist in v4.4.0; version is hardcoded here.
	- name: Install Zephyr SDK (minimal, cmake files only)
	env:
	SDK_VER: "1.0.1"
	run: \|
	wget -q \
	"https://github.com/zephyrproject-rtos/sdk-ng/releases/download/v${SDK_VER}/zephyr-sdk-${SDK_VER}_linux-x86_64_minimal.tar.xz" \
	-O /tmp/zephyr-sdk-minimal.tar.xz
	mkdir -p ~/zephyr-sdk
	tar -xf /tmp/zephyr-sdk-minimal.tar.xz --strip-components=1 -C ~/zephyr-sdk
	~/zephyr-sdk/setup.sh -c

	# Unit tests run as native executables on native_sim.
	# -fstack-usage emits .su files for stack analysis.
	- name: Twister — unit tests
	run: \|
	west twister \
	-T app/tests/unit \
	-p native_sim \
	--inline-logs -v \
	-O twister-out/unit \
	-- -DEXTRA_CFLAGS=-fstack-usage

	- name: Check stack usage (max 512 bytes per function)
	run: \|
	echo "--- Stack usage analysis ---"
	err=0
	while IFS= read -r sufile; do
	while IFS=$'\t' read -r func bytes type; do
	if [ "${bytes:-0}" -gt 512 ] 2>/dev/null; then
	echo "STACK EXCEEDED: $sufile: $func $bytes $type"
	err=1
	fi
	done < "$sufile"
	done < <(find twister-out/unit -name '*.su' 2>/dev/null)
	if [ $err -ne 0 ]; then
	echo "::error::One or more functions exceed 512-byte stack limit"
	exit 1
	fi
	echo "All functions within 512-byte stack limit"

	# Benchmarks execute on native_sim; Twister captures timing output
	# as an artifact. pass/fail is gated on the final log line regex.
	- name: Twister — benchmarks
	run: \|
	west twister \
	-T app/tests/benchmarks \
	-p native_sim \
	--inline-logs -v \
	-O twister-out/benchmarks

	# Parse benchmark timing and print summary table.
	# No fail threshold yet — we need baseline data first.
	- name: Benchmark timing summary
	if: always()
	run: python app/tools/parse_benchmark.py twister-out/benchmarks/

	# All 17 samples are build_only; CI proves they compile clean.
	- name: Twister — samples
	run: \|
	west twister \
	-T app/samples \
	-p native_sim \
	--inline-logs -v \
	-O twister-out/samples

	- name: Upload Twister results
	if: always()
	uses: actions/upload-artifact@v4
	with:
	name: twister-results
	path: twister-out/

	- name: Upload stack usage reports
	if: always()
	uses: actions/upload-artifact@v4
	with:
	name: stack-usage-reports
	path: twister-out/*/.su
	if-no-files-found: ignore

Navigation Menu

Uh oh!

feat: cross-validation harness and performance regression tracking #55

Workflow file

feat: cross-validation harness and performance regression tracking #55

Uh oh!

Workflow file for this run