Skip to content
CI

docs: correct per-endpoint queue cost math for MAX_ENDPOINTS 32 -> 64 #44

Sign in to view logs

docs: correct per-endpoint queue cost math for MAX_ENDPOINTS 32 -> 64

docs: correct per-endpoint queue cost math for MAX_ENDPOINTS 32 -> 64 #44

Workflow file for this run

.github/workflows/ci.yml at 126dbe1
# SPDX-License-Identifier: AGPL-3.0-or-later
# Copyright (C) 2024-2026 Jason Ricca
# CambiOS CI — Tier A: the core regression gate.
#
# Mirrors the local pre-commit discipline (CLAUDE.md § Tri-Architecture
# Regression Discipline, ADR-013) on GitHub-hosted Linux runners:
#
# * tri-arch release build — `make check-all` (x86_64 + aarch64 + riscv64)
# * host unit tests — `make test`, retargeted to the Linux host
# * missing_safety_doc = "deny" — `cargo clippy --lib` (Dev Convention 1)
# * don't-grow-the-baseline lints — the pure-Python check-* gates
#
# The pinned nightly and the three bare-metal targets come from
# rust-toolchain.toml; rustup installs them automatically on first use. The
# *-unknown-none targets are tier-2 with prebuilt core/alloc, so every build
# here is a pure Rust cross-compile — no C toolchain, no build-std.
#
# NOT in scope (deliberately):
# * Tier B — Kani proofs (`make verify-all`). Heavier setup (CBMC + a
# bundled toolchain, worth caching); lands as a separate workflow.
# * Tier C — QEMU boot smoke (`make run-quiet`). Needs qemu-system-* plus
# ISO/FAT-image tooling and seed signing; deferred.
# * The commit-msg / pre-commit / pre-push provenance hooks. They enforce
# Claude-authored-commit provenance (CLAUDE_PREFLIGHT_SESSION, edit logs)
# and are inherently local — meaningless in CI. This workflow is additive;
# that discipline stays where it is.
name: CI
on:
push:
branches: [main]
pull_request:
workflow_dispatch:
permissions:
contents: read
# Supersede in-flight runs for the same ref instead of stacking them.
concurrency:
group: ci-${{ github.ref }}
cancel-in-progress: true
jobs:
# ---------------------------------------------------------------------------
# Tri-arch release build — the `make check-all` payload, one job per target
# so the three cross-compiles run in parallel. fail-fast: false so a break on
# one arch still reports the status of the other two.
# ---------------------------------------------------------------------------
build:
name: build (${{ matrix.target }})
runs-on: ubuntu-latest
strategy:
fail-fast: false
matrix:
target:
- x86_64-unknown-none
- aarch64-unknown-none
- riscv64gc-unknown-none-elf
steps:
- uses: actions/checkout@v4
- name: Install pinned toolchain + target (rust-toolchain.toml)
run: |
rustup show
rustup target add ${{ matrix.target }}
- uses: Swatinem/rust-cache@v2
with:
key: ${{ matrix.target }}
- name: cargo build --release
run: cargo build --target ${{ matrix.target }} --release
# ---------------------------------------------------------------------------
# Host unit tests — `make test`, retargeted from x86_64-apple-darwin to the
# runner's native triple. The darwin target in the Makefile exists only to
# override .cargo/config.toml's bare-metal default; the --lib tests are pure
# logic. RUST_MIN_STACK matches the Makefile (buddy-allocator deep recursion).
# ---------------------------------------------------------------------------
test:
name: unit tests (--lib)
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Install pinned toolchain (rust-toolchain.toml)
run: rustup show
- uses: Swatinem/rust-cache@v2
with:
key: host-test
- name: cargo test --lib
env:
RUST_MIN_STACK: "8388608"
run: cargo test --lib --target x86_64-unknown-linux-gnu
# ---------------------------------------------------------------------------
# clippy — enforces the unsafe-discipline deny lints on ALL THREE backends
# (`make check-clippy`): missing_safety_doc (Cargo.toml [lints.clippy]) +
# multiple_unsafe_ops_per_block + undocumented_unsafe_blocks (src/lib.rs
# #![deny], Dev Convention 1). Tri-arch because x86_64 clippy never compiles
# the cfg-gated aarch64/riscv64 paths, so a per-op violation in an arch
# backend only surfaces when clippy runs on that target. No `-D warnings`, so
# the gate is exactly the deny lints, not a flood of unrelated style lints.
# ---------------------------------------------------------------------------
clippy:
name: clippy (unsafe discipline, tri-arch)
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Install pinned toolchain + clippy
run: |
rustup show
rustup component add clippy
rustup target add x86_64-unknown-none aarch64-unknown-none riscv64gc-unknown-none-elf
- uses: Swatinem/rust-cache@v2
with:
key: clippy
- name: make check-clippy (tri-arch)
run: make check-clippy
# ---------------------------------------------------------------------------
# Don't-grow-the-baseline content lints — pure Python (CLAUDE.md Dev
# Conventions 1/8/9, ADR-021). No Rust toolchain needed; ubuntu-latest ships
# python3 + make. One step per gate so a reviewer sees exactly which is green.
#
# Commit-boundary / provenance gates are excluded on purpose: they read the
# staged diff or session state and are no-ops or meaningless in CI —
# check-index-isolation, check-lockfile (advisory), claude-preflight.
# check-status-freshness runs here too (Kani-harness drift, --strict): it is
# compile-free now that STATUS.md no longer pins a unit-test total, so it
# belongs with the other pure-python gates. Its date heuristic stays
# local/advisory and is not run in CI.
#
# Note: check-adrs regenerates docs/adr/INDEX.md as a side effect (exit 0 when
# cross-refs are valid). CI does not commit, so a stale committed INDEX.md is
# not caught here — it fails only on broken/duplicate ADR references.
# ---------------------------------------------------------------------------
lints:
name: baseline lints
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: check-adrs (ADR cross-references)
run: make check-adrs
- name: check-assumptions (Convention 8 numeric bounds)
run: make check-assumptions
- name: check-deferrals (Convention 9 Revisit-when triggers)
run: make check-deferrals
- name: check-boot-panics (ADR-021 boot-init panics)
run: make check-boot-panics
- name: check-banned-paths (deleted-stays-deleted)
run: make check-banned-paths
- name: check-unsafe-coverage (Convention 1 SAFETY comments)
run: make check-unsafe-coverage
- name: check-status-freshness (Kani-harness drift, --strict)
run: python3 tools/check-status-freshness.py --strict