GeezerVison

A worker whose entry script sets a throwing onclose and then calls close() crashes the app:

onclose = () => { throw new Error("boom"); };
close();

CloseWorkerCallback reports the error through CallOnErrorHandlers, which read workerIsolate_. That field is only published once the worker's startup function returns, so during the entry it is still null and Caches::Get(nullptr) faults. CallOnErrorHandlers and ReportEntryEvaluationRejection now take the isolate from their caller, which is running on it. The new spec crashes the TestRunner without the fix.

Two more lifetime problems in the same file, found while reading how that field is used. Neither has a spec: one needs a preempted thread, the other needs the runtime shut down mid-suite.

• Terminate() read workerIsolate_ on the parent thread and then called RequestTermination() and TerminateExecution() on it with nothing holding the isolate alive. If the worker has just published the isolate and sees isTerminating_ before entering its run loop, it goes straight to deleting its runtime while the parent may still be between the read and the use. The worker now publishes and withdraws the pointer under a mutex and Terminate() holds it across the use. Only the winner of the isTerminating_ exchange takes it, and the worker holds it just for the two assignments, so the heap-limit callback calling Terminate() from inside a GC on the worker thread cannot deadlock on it. Once the worker is inside CFRunLoopRun the old code was already safe, since the loop only stops through queue_.Terminate(), which runs after the isolate use.
• EndWrapperLifetime() dispatched nsworkerended and then called this->UnrootWorkerObject(). A listener that calls shutdownRuntime (supported from inside a JS callback, see DisposeIsolateWhenPossible) runs ~Runtime on the same thread, and its disposal walk deletes this wrapper, because the worker has already published isDisposed_ by the time the notification is posted. It now reads what the dispatch needs up front and checks the selfRef_ token again before touching this.

Summary by CodeRabbit

• Bug Fixes

  • Fixed worker errors thrown by an onclose handler during entry-script evaluation so they are correctly reported through the parent worker’s onerror event.
  • Improved worker shutdown handling to prevent race conditions and ensure errors and lifecycle events are delivered safely.

• Tests

  • Added regression coverage for errors raised during worker shutdown while the entry script is still running.