Phase A: cellular modem control stack + bench runbooks (pre-v0.1.0)#2
Merged
Conversation
A stalled auth handshake (a bus that returns after a crash but is slow to accept, or a stale socket left by a SIGKILLed daemon) left the reconnect loop's await on establish() hanging forever, so the reconnected event never fired. This surfaced as an intermittent full-suite-load failure of the transport reconnect reliability test on slower CI runners. Bound each connect attempt with a 2s timeout and close+swallow the abandoned connection so the loop always makes progress and retries with backoff.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
The Phase-A integration branch for
modem-stack— the complete cellular modem control stacklanded as wave-ordered coherent commits (bootstrap → A2 domain/transport/ports → A3 backend
→ A4 NM/USB-mode/usage → A5 bookworm packaging → A6 bench CLI/certify/runbooks). This single
PR brings all of it onto
mainahead of the first release,v0.1.0.Final commit adds
docs/BENCH.md: agent-executable, non-interactive, machine-checkablerunbooks for every hardware-gated item (system-bus probe, slot-UID stability across replug,
MM-1.24-from-artifacts install, PIN/PUK on a real SIM, a certified USB-mode transition, and a
numeric usage-meter accuracy check), plus the Phase-A iteration loop.
Why
The plan's commit strategy cuts Phase A as ONE integration branch/PR (R2), merged just before
the first
0.xrelease.maincurrently holds only the repo bootstrap; this PR lands theengine, packaging, CLI, and bench docs together so a single
v0.1.0tag can release both thenpm package (
@ceralive/modem-control) and the ModemManager-stack.debset from the samecommit.
How to verify
bun install --frozen-lockfile && bun run lint && bun run typecheckdbus-run-session -- bun test— full suite (355 pass at A6.2)bash packaging/ci/contract.sh— packaging contract lane (bookworm containers)cli/smoke/run.sh— compiled cross-arch D-Bus probe smoke (amd64 native + arm64 QEMU)docs/BENCH.md: every hardware-gated runbook is[PARTIAL]with an exactnon-interactive command, a greppable expected line, and a repo-local evidence path.
Risks
[PARTIAL]by design — their CI proxies are green (compiled probesmoke both arches, packaging contract + daemon smoke, full unit suite), but on-hardware
evidence is captured later on real bench devices. No product consumes this repo in Phase A.
autonomous switching; recovery ladder disabled by default.