perf(shred): skip primary FEC reparse — reduce packet-worker overhead

[Lythaeon]

Description

Reduce packet-worker FEC overhead without changing SOF API contracts or feature scope.
This PR removes a redundant primary-packet shred reparse in the FEC ingest path by reusing the already-validated parsed header that packet workers already carry.

Changes

Detailed list of what changed:

• crates/sof-observer/src/app/runtime/runloop/packet_workers.rs: pass the existing ParsedShredHeader into the FEC recoverer instead of forcing the recoverer to parse the primary packet bytes again.
• crates/sof-observer/src/shred/fec/core.rs: change primary-packet ingest to consume ParsedShredHeader and use it for slot/FEC-set/variant routing, avoiding a second full parse on the hot path.
• Any architecture/runtime/infra implications: none; this is an internal hot-path optimization inside the existing packet-worker/FEC pipeline.

For slice-related changes, include:

• Affected slices: app/runtime/runloop, shred/fec.
• Cross-slice communication changes (if any) and why: packet workers now hand the FEC layer the parsed header they already own, removing duplicated work across the runloop and shred slices.
• Migration requirements (if any): none.

Motivation

Business motivation:

• Improve throughput on the current feature set before expanding scope.

Technical motivation:

• Packet workers already parse and retain ParsedShredHeader for accepted packets.
• Repeating full shred parsing inside FecRecoverer::ingest_packet wastes CPU on the primary-path hot loop.
• Reusing the existing header reduces duplicated parsing without changing behavior.

Alternative approaches considered:

• Leaving the duplicate parse in place.
• Bundling this with a larger FEC/storage refactor instead of landing the smallest measurable improvement first.

Scope and impact

• Affected slices: packet-worker/FEC hot path.
• Data/API changes: none.
• Backward compatibility: fully backward compatible.
• Performance impact: positive in the primary FEC profile fixture; local A/B dropped from 110 ms to 94 ms over 50,000 iterations.
• Security impact: none.

Testing

• Unit tests
• Integration tests
• Manual verification
• Performance checks (if applicable)
• Security checks (if applicable)

Commands/results:

cargo fmt --all
cargo test -p sof --lib --no-run
cargo test -p sof --lib packet_worker_ -- --nocapture
cargo test -p sof --lib purge_older_than_removes_old_sets -- --nocapture
SOF_PACKET_WORKER_FEC_PROFILE_ITERS=50000 CARGO_TARGET_DIR=/home/ac/.cache/sof-fec-current/target cargo test -p sof --lib packet_worker_primary_fec_profile_fixture -- --ignored --nocapture
cd /tmp/sof-baseline-import-hygiene && SOF_PACKET_WORKER_FEC_PROFILE_ITERS=50000 CARGO_TARGET_DIR=/home/ac/.cache/sof-fec-baseline/target cargo test -p sof --lib packet_worker_primary_fec_profile_fixture -- --ignored --nocapture

Results: passed.

Related issues and documentation

• Fixes:
• Related: packet/FEC perf roadmap follow-up in docs/operations/tx-reconstruction-and-packet-processing-todo-2026-03-24.md
• Architecture docs: docs/architecture/README.md
• Relevant ARD/ADR:
• Operations/runbook updates: none.

Reviewer checklist

• Code follows project standards and architecture constraints
• Slice boundaries are respected (docs/architecture/ard/0003-slice-dependency-contracts.md)
• Tests added/updated and passing
• Documentation updated (README/docs/operations as needed)
• No undocumented breaking change
• Performance trade-offs documented where relevant
• Security considerations addressed where relevant

Additional notes

This PR is independent of the completed-dataset perf work and can be reviewed directly on top of refactor/import-hygiene.