Bug 6884 - GC after destroying the very last frame to avoid node leak false positives
Summary: GC after destroying the very last frame to avoid node leak false positives
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Frames (show other bugs)
Version: 420+
Hardware: Mac OS X 10.4
: P2 Normal
Assignee: Maciej Stachowiak
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2006-01-27 17:43 PST by Maciej Stachowiak
Modified: 2006-01-27 21:24 PST (History)
0 users

See Also:

Attachments
GC after destroying the very last Frame, from higher up on the stack (1.70 KB, patch)
2006-01-27 17:45 PST, Maciej Stachowiak 		ggaren: review+
Details | Formatted Diff | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Maciej Stachowiak 2006-01-27 17:43:35 PST 
The NodeImpl leak counter can give false positives sometimes if the GC after destroying the last ScriptInterpreter doesn't collect everything due to stack references. Thus, in debug mode, it would be good to do a GC from Page's destructor after destroying the last frame, so we're at a point on the stack where there wont be JS object references.
Comment 1 Maciej Stachowiak 2006-01-27 17:45:19 PST 
Created attachment 6033 [details]
GC after destroying the very last Frame, from higher up on the stack
Comment 2 Geoffrey Garen 2006-01-27 17:58:25 PST 
Comment on attachment 6033 [details]
GC after destroying the very last Frame, from higher up on the stack

r=me

I'm a little worried that we'll be finessing the garbage collector to avoid false leak reports in a lot more places. RenderObject comes to mind.