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

[bugzilla-daemon at webkit.org]

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


eric at webkit.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mjs at apple.com,
                   |                            |darin at apple.com




------- Comment #7 from eric at webkit.org  2007-10-24 04:33 PDT -------
So the memory check is gone after the patch on bug 15658 lands.  This macro is,
of course, still super hot.  I think it can safely be removed from hot loops
with a little work.

I think if the semantic on ExecState were to change such the setting an
exception when another was already set was an error, then we could just remove
these checks.  Assume that any sort of throwError function will set the
exception (including proper line numbers, etc) ONLY if there isn't already an
exception on the ExecState.  Then we just need to make sure that before leaving
any node we should call handleExeception once to make sure the line numbers,
etc. are updated if necessary.

We just need to be careful to educate the node tree to make sure continuing in
an "exception was thrown, but I'm not yet ready to check it" state is OK.  So
long as we record the proper line numbers, etc in the Error object, and make
sure to update them before leaving any node... I think it should work.  Maybe
I'll look at that tomorrow (when I'm more awake) if no one else has.


-- 
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.