feat: per-source RX demux — source on Unicast + per-source E2E state#136
Closed
JustinKovacich wants to merge 3 commits into
Closed
feat: per-source RX demux — source on Unicast + per-source E2E state#136JustinKovacich wants to merge 3 commits into
JustinKovacich wants to merge 3 commits into
Conversation
The SOME/IP unicast header carries no instance id, so on a shared subnet the sender's source address is the only way to attribute an event to a device. The source was already received (ReceivedMessage.source) but dropped before forwarding; carry it on ClientUpdate::Unicast instead.
On a shared subnet, several devices send the same (service, method) under one fixed instance id; a single per-key receive counter collides their interleaved sequences into spurious WrongSequence. Split the registry into endpoint-agnostic profile config, per-(source,key) receive state, and per-key transmit state. check() now takes the source address; protect() is unchanged. Reset a source's receive state on its reboot.
Contributor
Author
|
Folded into |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Two changes that let a single client attribute SOME/IP telemetry to a specific device on a shared subnet, correct under E2E:
ClientUpdate::Unicastnow carriessource: SocketAddr. The source was already received (ReceivedMessage.source) but dropped via..before forwarding. The SOME/IP header has no instance id, so on a shared subnet the source address is the only demux key.E2ERegistrynow holds endpoint-agnostic profile config + per-(source, key)receive state + per-keytransmit state.check()takes the source;protect()is unchanged (fan-out keeps one counter; per-recipient TX counters live a layer up). A rebooted source's receive state is reset.Downstream this unblocks per-sensor telemetry in
iris_someip_client/ EnVision / the ROS FFI (the consumer isfeature/env_multisensor_rx_demuxin dft).Tests
unicast_update_carries_sourcedistinct_sources_have_independent_e2e_state(interleaved counters from two sources don't tripWrongSequence)reset_source_clears_only_that_source--all-features; clippy-D warnings -D clippy::pedanticclean; fmt clean.⚠ Versioning decision needed (why this is a draft)
Adding a field to the
Unicaststruct variant is semver-breaking (ClientUpdateis not#[non_exhaustive]). So:cargo-semver-checkswill require a 0.8.0 minor bump (this is the 0.7.2 line; a 0.7.3 patch would fail the SemVer gate).feat:commits will make release-plz propose 0.8.0.Intent was to ship as 0.7.3 off
release/0.7.2, then forward-port into the 0.8.0 spine. Opening as draft to read the CI verdict before choosing: force a 0.7.3 patch (override the SemVer gate, internal-crate consumers) vs. ship 0.8.0 and reconcile with the existing 0.8.0 spine.🤖 Generated with Claude Code