Forbid further execution in jsc shell if execution is terminated.

[Mark Lam]

Also re-implement WorkerOrWorkletScriptController::forbidExecution() and isExecutionForbidden() using the VM's notion of the flag.

rdar://77548608

[Mark Lam]

Created attachment 427797 [details]
propose patch.

[Michael Saboff]

Comment on attachment 427797 [details]
propose patch.

View in context: https://bugs.webkit.org/attachment.cgi?id=427797&action=review

r=me with a suggested change.

> Source/JavaScriptCore/runtime/VM.h:345
> +    void setExecutionForbidden(bool value) { m_executionForbidden = value; }

Seem like we only want to go from "execution allowed", e.g. m_executionForbidden == false, to "execution forbidden", e.g. m_executionForbidden == true.  I suggest we eliminate the argument to setExecutionForbidden() and have it always set m_executionForbidden to true.

[Mark Lam]

(In reply to Michael Saboff from comment #2)
> Comment on attachment 427797 [details]
> propose patch.
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=427797&action=review
> 
> r=me with a suggested change.
> 
> > Source/JavaScriptCore/runtime/VM.h:345
> > +    void setExecutionForbidden(bool value) { m_executionForbidden = value; }
> 
> Seem like we only want to go from "execution allowed", e.g.
> m_executionForbidden == false, to "execution forbidden", e.g.
> m_executionForbidden == true.  I suggest we eliminate the argument to
> setExecutionForbidden() and have it always set m_executionForbidden to true.

I agree.  That's probably the better approach until we find evidence to the contrary.

[Mark Lam]

Thanks for the review.  I've made the suggested change.

Landed in r277094: <http://trac.webkit.org/r277094>.