[Webkit-unassigned] [Bug 6083] REGRESSION: 5-10% performance regression on JS iBench from getter/setter change

[bugzilla-daemon at opendarwin.org]

http://bugzilla.opendarwin.org/show_bug.cgi?id=6083


mjs at apple.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|webkit-                     |mjs at apple.com
                   |unassigned at opendarwin.org   |
             Status|NEW                         |ASSIGNED




------- Additional Comments From mjs at apple.com  2005-12-20 04:24 -------
Created an attachment (id=5178)
 --> (http://bugzilla.opendarwin.org/attachment.cgi?id=5178&action=view)
latest patch

Here's the latest patch I had, much of it is irrelevant or wrong but what is
surprising is that it still causes a slowdown, solely due to the
getOwnPropertySlot changes (put changes make no difference any more, nor does
the extra argument to getValue as far as I can tell). The problem is that
JSObject::getOwnPropertSlot no longer gets inlined into derived classes calling
the base class, most notably ActivationImp. Even though I have it marked
always_inline! Not sure why.

Some of the fixes here are for warnings from -Winline.


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