[Webkit-unassigned] [Bug 68318] DFG JIT should inline Math.min, Math.max, and Math.sqrt

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Oct 4 14:05:30 PDT 2011


https://bugs.webkit.org/show_bug.cgi?id=68318





--- Comment #6 from Filip Pizlo <fpizlo at apple.com>  2011-10-04 14:05:30 PST ---
(In reply to comment #5)
> It appears that this patch introduced some regression in JSC. I can no longer load internal builds of Google docs. It repeatedly encounters the following error in WebKit2 (regardless of whether I use single process or not):
> 
> ERROR: Failed to send message to real exception port, error 10000003
> /Volumes/Data/webkit/Source/WebKit2/Platform/CoreIPC/mac/ConnectionMac.cpp(407) : void CoreIPC::Connection::exceptionSourceEventHandler()
> 
> I've manually bisected the builds and r95397 worked fine so I'm pretty certain this patch caused it. Unfortunately, I don't know ways to provide access to our internal builds of Google docs at the moment :( so we can only inspect the code change manually.

Here's a simple test you could do to see if and how this changeset affects the problem you're seeing:

Go to Source/JavaScriptCore/dfg/DFGByteCodeParser.cpp

Replace the contents of handleIntrinsic() method with just "return false".

Check if you still see the problem.  If it makes the problem go away, then it's highly likely the fault of this patch.  If the problem persists, then it's highly unlikely that this patch caused it.

Thanks for your help.  I understand this can be complicated when the only repro is in internal code!

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



More information about the webkit-unassigned mailing list