Bug 21267 - Crash in KJS::Heap::isBusy() running Google Top Draw
Summary: Crash in KJS::Heap::isBusy() running Google Top Draw
Status: RESOLVED INVALID
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: 528+ (Nightly build)
Hardware: Mac OS X 10.5
: P1 Normal
Assignee: Nobody
URL:
Keywords: NeedsReduction
Depends on:
Blocks:
 
Reported: 2008-09-30 21:38 PDT by Matt Lilek
Modified: 2008-10-01 16:16 PDT (History)
2 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Matt Lilek 2008-09-30 21:38:06 PDT
Crashes while rendering at the 2 second mark on a 2GHz Core 2 Duo with 4GB RAM with JSCore from Safari 4 DPUpdate.

Source is available at <http://code.google.com/p/topdraw/>.

Thread 0 Crashed:
0   com.apple.JavaScriptCore      	0x93b65876 KJS::Heap::isBusy() + 6
1   com.apple.JavaScriptCore      	0x93bee3fc JSGarbageCollect + 28
2   TopDrawRenderer               	0x0001393c -[Runtime dealloc] + 46
3   TopDrawRenderer               	0x00004c69 -[Compositor evaluateWithSeed:] + 1116
4   TopDrawRenderer               	0x000118ef Process + 590
5   TopDrawRenderer               	0x000121e4 main + 1326
6   TopDrawRenderer               	0x0000271e start + 54
Comment 1 Mark Rowe (bdash) 2008-09-30 23:40:04 PDT
Does this reproduce with a TOT build?  It'd be great if you could provide steps to reproduce.  It's not at all clear *what* you are trying to render when it crashes.
Comment 2 Matt Lilek 2008-10-01 16:16:15 PDT
(In reply to comment #1)
> It's not at all clear *what* you are trying to render when it crashes.

I was just launching it and it was crashing, I should have been more clear.

> Does this reproduce with a TOT build?  It'd be great if you could provide steps
> to reproduce.

No, it doesn't crash in TOT and I'm closing it as INVALID.  I had grabbed the source and linked it against my local WebKit build, but for some reason it was still using the system version.  This should have been obvious from the "KJS" not "JSC" namespace in the backtrace.  Apologies for the misfire.