You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Test: make shared GC attribution witnesses stable under full unit sharding #839
Parent: #476. Related: #673 and #838 (observed while validating that compiler fix; no affected loop/suspension test failed in these runs).
Two consecutive zig build test-parallel runs on current main plus the #838 candidate each accounted for all 2,233 tests with 2,231 pass / 1 skip / 1 failure, but the single failure moved between unrelated GC-attribution witnesses:
failed only because ctx.gc_cooperative_attempts == 0
all JS work completed; no crash, race report, leak, or wrong result
immediately passed 3/3 when run with the exact test filter
enable_gc: requested GC runs between microtasks after joined threads
failed because the final collected witness was not true
immediately passed 3/3 when run with the exact test filter
The earlier full run before the final storage-policy correction passed both of these witnesses and instead caught the now-fixed loop/suspension regressions, so neither predicate is deterministically broken by #838. The failures appear only under the full 10-shard load and move between collection-attribution boundaries.
Diagnose why the requested/cooperative collection can complete the workload without recording the expected witness under sharded load. Preserve the production tranche, collection requirements, and exact assertions: no sleeps, larger timeouts, forced extra drains/collections, relaxed predicates, disabled GC, or shard serialization. The test should observe the engine's real request/attempt publication deterministically, and the engine should make progress if the missing event is genuine.
synchronous Thread.join now remains parked until the worker has left its active interpreter and native stack; result publication alone no longer lets join return while Context GC must still treat the thread as live
the cooperative nursery witness now has either lane independently cross the unchanged 512 KiB tranche while its peer remains registered, making the existing attempt/collection/peer-park assertions scheduler-independent
No sleeps, larger timeouts, forced collections, relaxed predicates, disabled GC, or shard serialization were added.
Validation:
cooperative nursery witness: 10/10
requested microtask GC after join: 10/10
all 11 join-focused unit tests: pass
the same 11 tests under unsuppressed TSan: pass, no sanitizer report
full test-parallel: 2,232 passed, 1 skipped, 0 failed, 0 leaked across 10 shards
Thread is a non-standard host API, so test262 is not applicable to this change.
Activity
fix(compiler): isolate iterator state per activation (#838)
chrisbbreuer commentedon Sep 1, 2026
Landed on main in 0b6f4d96.
Two independent causes were corrected:
No sleeps, larger timeouts, forced collections, relaxed predicates, disabled GC, or shard serialization were added.
Validation:
Thread is a non-standard host API, so test262 is not applicable to this change.