Skip to content

Testing and CI

How the driver is checked: host tests against a model of the pixels written from the datasheets, a fuzzer that controls the configuration, the tick rate and the port's failures, compiled documentation, an install test, ESP32 and Cortex-M builds, static analysis and measured quality gates, on every pull request.

Host tests

cmake -S . -B build -DWS2812_RMT_BUILD_TESTS=ON
cmake --build build
ctest --test-dir build --output-on-failure
File Covers
tests/test_waveform.cpp ResolveWaveform() at known rates, the edges of every window, the period, reset and gap-limit rules, ResolveRmtRate() (the divider arithmetic of the ESP32 port), a sweep of hostile rates
tests/test_timing_models.cpp Every profile equals the intersection of its documents; the nominals lie inside every document; two WS2812B generations share no waveform; default orders
tests/test_model.cpp The model itself: its decoding, its windows, its latch, with hand-made symbol streams
tests/test_encode.cpp, test_colour.cpp Byte packing for every colour order, Scale8 over all 65 536 pairs, gamma, white, wheel
tests/test_frame.cpp Frame time, refresh rate, pixels per frame, the input-level function
tests/test_chain.cpp End to end: every variant, every document, every order, strips of 1 to 1024 pixels, a tick-rate error of 1000 ppm
tests/test_lifecycle.cpp, test_errors.cpp, test_bus_faults.cpp Start-up, every error value, failing and underrunning ports
tests/test_edges.cpp The boundaries a one-edit change would move: the reset wait at a rate that is not whole megahertz, the rounded-up tick time, the longer bit code in the frame time, the exclusive in-frame cap, the recovery after a failed PowerOff()
tests/esp32_stub/ The ESP32 port and both examples over stand-ins of the ESP-IDF headers: each example's Start() with the default pins reaches a shown frame, the channel settings, encoder, waits and recovery, the ESP32-H2 default rate
tests/test_kconfig.cpp The menuconfig values and their defaults
tests/mock_ws2812_rmt.hpp The model: the documents' windows, a pixel chain that latches, and the fake ports
tests/fuzz/ The fuzz target; CTest replays its corpus (ws2812_rmt_fuzz_ws2812_rmt_corpus), see below
tests/compile_fail/ Ports and configurations that must not compile
tests/docs/ Every code block of the site, compiled
tests/check_install.cmake find_package(ws2812_rmt) from an installed package

The tests build with -fno-exceptions -fno-rtti and warnings as errors, on GCC and Clang.

What a test cannot show on the host

Every behaviour fix of the review rounds has a host test that fails without it, except these, which a host cannot exercise. Documentation, label and figure fixes, the typed constants (checked by the compiler) and the reset-wait busy loop are reviewed, not tested:

  • The ESP32 port on the chip (the real clock, the memory size, the interrupt priority, the line level after a timeout) needs ESP-IDF and a chip. On the host, tests/esp32_stub/ builds the port and both examples against stand-ins of the ESP-IDF headers and runs each example's start-up, the channel settings, the encoder, the waits and the recovery; its divider arithmetic is the pure ResolveRmtRate(), which is tested on its own as well.
  • Whether the RMT driver reports an underrun is a property of ESP-IDF: the docs state that this port cannot see one, and the model test of an underrun shows the silent outcome a port that cannot see it produces.
  • The reference circuit figures (docs/hardware.yml) are checked by kokeb-docs hardware --check and kokeb-docs figures --check for validity and routing, not for electrical correctness.

The model stands alone

tests/mock_ws2812_rmt.hpp carries its own table of the windows of each datasheet document (kDocuments), typed in from the PDFs and never from the driver's profile table. A test cross-checks that every named variant is the intersection of its documents. The pixel chain decodes the half-symbols the port was given into bits, judges every duration against the window, latches a frame when a low exceeds the reset, and reports pixels that received too few or too many bits. So a transcription error in either table, a waveform outside a window, or a missing reset fails a test, and green tests cannot come from one author's single reading of a datasheet.

Fuzz tests

tests/fuzz/fuzz_ws2812_rmt.cpp lets libFuzzer choose a configuration (any variant, Custom with arbitrary windows, any order, brightness, gamma, reset), the port (any tick rate including 0 and 2^32 - 1, failing hooks, with or without the optional hooks) and up to 48 driver calls. After every call it checks the driver's promises and aborts, naming the line, when one breaks:

  • a failed call returns an Error this driver returns (never Timeout, Crc or Busy) and T{};
  • before EnsureInitialized() succeeded the port saw no Transmit(), and InvalidArgument, NotInitialized and TimingUnreachable mean nothing was transmitted;
  • every transmitted waveform lies inside the windows of the declared profile and its reset is above the minimum: a hostile tick rate can never produce an illegal waveform;
  • the wire has exactly Length() x channels bytes at every Transmit(), equal to a second encoding written in the target;
  • after a failed Transmit() the pixels are intact, and the port's waits before the next Transmit() add up to strictly more than the reset (the port records every DelayUs() and every Transmit() in order);
  • the supply pin write fails when the input says so: a failed write is Bus, PowerOn() then leaves the strip as it was, and after any PowerOff() (also one whose write failed) Show() and TraceWire() transmit nothing and return NotPowered;
  • TraceWire() sends exactly the bytes it was given, and an empty span or one above four bytes per pixel of capacity is InvalidArgument with nothing sent;
  • every log line is a non-empty plain-text message.
# libFuzzer needs Clang (on macOS, Homebrew's LLVM)
cmake -S . -B build-fuzz -DCMAKE_CXX_COMPILER=clang++ -DWS2812_RMT_BUILD_TESTS=ON -DWS2812_RMT_BUILD_FUZZERS=ON
cmake --build build-fuzz --target fuzz_ws2812_rmt
./build-fuzz/tests/fuzz/fuzz_ws2812_rmt -max_total_time=120 tests/fuzz/corpus/fuzz_ws2812_rmt

The same file also builds without libFuzzer as fuzz_replay_fuzz_ws2812_rmt, and CTest replays every input in tests/fuzz/corpus/fuzz_ws2812_rmt/ on GCC and Clang (ws2812_rmt_fuzz_ws2812_rmt_corpus): the seeds, and every input that ever broke a promise (name it regress-*), so a fixed bug stays fixed. CI fuzzes for 120 s on every pull request, from the seeds plus a corpus it keeps between runs (job Host quality / fuzz).

Quality gates

The Host quality job measures the library code, inc/ and src/ only (tests and examples are left out), and fails when a value is over the limits of this repository's assurance level, high (set in .kokeb/handbook.yml, defined by the handbook's ADR-0020):

Gate Tool and how it is measured Limit at high
Complexity lizard 1.24.1, modified cyclomatic complexity (-m: a switch counts once) per function CCN at most 15
Function length lizard, lines of code per function at most 60
Parameters lizard, per function at most 6
Line coverage GCC 14 --coverage -O0, every CTest, gcovr 8.7 --merge-lines (a template's lines count once across its instances), --exclude-unreachable-branches --exclude-throw-branches at least 85 %
Branch coverage same run at least 75 %
MC/DC GCC -fcondition-coverage, same run measured, not gated at high (100 % at safety)

Measured on the last run: lines 97.2 % (522 of 537), functions 93.9 %, branches 89.1 % (344 of 386), conditions 88.9 % (304 of 342), decisions 84.2 % (133 of 158). | Sanitizers | Clang, -fsanitize=address,undefined -fno-sanitize-recover=all, every CTest | no report | | Fuzzing | libFuzzer with ASan and UBSan, 120 s per target | no broken promise |

The reports are in the run's host-quality-reports artifact and its step summary. To run the gates locally:

# Coverage (GCC 14 or later), then the complexity limits
cmake -S . -B build-cov -DCMAKE_CXX_COMPILER=g++-14 -DWS2812_RMT_BUILD_TESTS=ON \
  -DCMAKE_CXX_FLAGS="--coverage -O0 -fcondition-coverage"
cmake --build build-cov && ctest --test-dir build-cov
gcovr --root . --filter inc/ --filter src/ --merge-lines --exclude-unreachable-branches --txt build-cov
lizard -m -C 15 -L 60 -a 6 inc src

# Sanitizers (Clang)
cmake -S . -B build-san -DCMAKE_CXX_COMPILER=clang++ -DWS2812_RMT_BUILD_TESTS=ON \
  -DCMAKE_CXX_FLAGS="-fsanitize=address,undefined -fno-sanitize-recover=all -O1 -g"
cmake --build build-san && ctest --test-dir build-san --output-on-failure

The fuzz gate runs locally as shown in Fuzz tests.

clang-tidy 20 (_config/.clang-tidy: bugprone, CERT, clang-analyzer, cppcoreguidelines, modernize, performance, portability and readability, each disabled check with its reason) and cppcheck (strict) run in the C++ lint and Repository checks jobs and fail on any finding.

The MISRA job (ci-misra.yml) runs the shared ru-misra workflow of ci-tools at level high: it traces a host build and the Cortex-M build (ws2812_rmt_optimised_build) of every translation unit, checks them against MISRA C++:2023 and the handbook's C++20 profile (the CodeQL coding-standards and profile packs, plus the validated clang-tidy and cppcheck checks), and reports each finding against the handbook's compliance matrix and deviation records. A finding is fixed in the code; a deliberate exception names its deviation on the line before it (// codeql::misra_deviation_next_line(MD-CPP-009) above the .ipp include of inc/ws2812_rmt.hpp). Until the repository is public the job runs with open_source_prerelease: true, the declaration that the code is released under the licence in LICENSE and is to be published.

CI workflows

Everything runs from ci.yml, which calls one workflow file per check and ends in the CI gate job, the single required check (CI gate / gate).

Job What it checks
Host tests GCC, then Clang (one job), build and run every test above, then build and run examples/host on its own
Host quality The gates above: complexity, coverage, sanitizers, fuzzing
Cross-compile arm-none-eabi-g++ for Cortex-M0+, M4F and M7F with the strict flags, warnings as errors (nothing linked)
C++ lint clang-format and clang-tidy over the sources
MISRA MISRA C++:2023 and the C++20 profile on traced host and Cortex-M builds, at level high (see Quality gates)
Docs Strict site build (_config/mkdocs.yml), generated-region drift, Doxygen (_config/Doxyfile), offline link check
Repository checks One job: DCO on pull requests, the org standard, Markdown and YAML lint, the Kokeb Labs handbook rules for .kokeb/handbook.yml (by rule ID), template drift (warns only) and cppcheck over the driver sources
CI gate Passes only if every job above passed: the one required check

The ESP32 examples job builds every app in examples/esp32/app_config.yml with the repository root as the driver's component. The gate needs it too.