coro: integrate LLVM 22 runtime fast paths - #160
Merged
Conversation
Publish exact static outcome entries through ProgramIR and library metadata, prove borrowed transaction storage, and make closed-program worker capability explicit.\n\nFuse single-event park and owner-local channel/manual completion paths while preserving the common select, cancellation, and cross-route state machines. Add native/WASM correctness gates and reproducible core performance fixtures.
Keep immutable scheduler, placement, frame-context, and direct-channel proofs at their stable ownership boundaries. Fuse the compiler/runtime channel transaction, retain exact fallback gates, and add race, architecture, LLVM 22, and performance validation.
Emit static itabs and immutable backing for constant interface conversions, and preserve LLVM call-site attributes through ABI rewrites.\n\nMove native worker queue transport to scalar reservations and arguments, add a bounded direct handoff plus per-worker fault setup, and use an advisory executor completion window before retained sleep. Keep source reduction, cancellation, and doorbell fallback unchanged.\n\nNarrow borrowed-allocation proof authority, update exact architecture/source gates, and record controlled file/TCP measurements.
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Integrates the current LLVM 22-only correctness and performance series into
llvm-coro.Highlights:
Validation:
go test -count=1 ./...-O3 -lto=fullbuild plus file 10,000 and TCP 5,000 stress runsSame-host 21-run medians versus the PR parent artifact (500 operations):
The isolated worker-park change improves file median by about 9% and leaves TCP neutral; the stripped workload artifact is 7,666,544 bytes.