[Webkit-unassigned] [Bug 15633] SunSpider shows JSC spending time in KJSCHECKEXCEPTIONVALUE

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Oct 24 16:41:30 PDT 2007


http://bugs.webkit.org/show_bug.cgi?id=15633





------- Comment #9 from darin at apple.com  2007-10-24 16:41 PDT -------
(In reply to comment #8)
> Ok, so I've further revised my theory for how we can remove all these if
> (exec-hadException()) checks.  Like before, we make exec->setException() ASSERT
> when already had an exception.  We remove *all* exception checks, except before
> any user-visible change.  User visible changes come in 3 flavors:
> 1.  Assignment nodes (we have to pretend execution stopped immediately after
> the exception, make sure it stops before we change any variable)
> 2.  Loops (hanging due to assignments never happening, thus loops never
> reaching their termination condition is a bad thing, loops need to check
> exceptions)
> 3.  non-source function calls (functions provided by the interpreter must not
> be called after an exception).

I think this sounds OK. But I believe "user-visible" is a pretty large
category; get/set on an arbitrary object, toString on an arbitrary object, etc.
I'm not sure how many of the checks you could remove with this trick.


-- 
Configure bugmail: http://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.



More information about the webkit-unassigned mailing list