Last Comment Bug 709384 - "ASSERTION: Don't schedule the same document multiple times"
: "ASSERTION: Don't schedule the same document multiple times"
Status: RESOLVED FIXED
: assertion, testcase
Product: Core
Classification: Components
Component: DOM (show other bugs)
: Trunk
: x86_64 Mac OS X
: P1 normal (vote)
: mozilla11
Assigned To: Boris Zbarsky [:bz] (TPAC)
:
Mentors:
Depends on:
Blocks: 594645 647518
  Show dependency tree / graph
 
Reported: 2011-12-09 15:47 PST by Jesse Ruderman
Modified: 2011-12-15 10:35 PST (History)
4 users (show)
bzbarsky: in‑testsuite+
See Also:
Crash Signature:
(edit)
QA Whiteboard:
Iteration: ---
Points: ---
Has Regression Range: ---
Has STR: ---

Attachments
testcase (193 bytes, text/html)
2011-12-09 15:47 PST, Jesse Ruderman 		no flags Details
Manage our scheduled state better when animation frame requests are canceled. (2.05 KB, patch)
2011-12-14 22:57 PST, Boris Zbarsky [:bz] (TPAC) 		roc: review+
Details | Diff | Splinter Review
View All Add an attachment (proposed patch, testcase, etc.)

Description Jesse Ruderman 2011-12-09 15:47:46 PST 
Created attachment 580578 [details]
testcase

###!!! ASSERTION: Don't schedule the same document multiple times: 'mFrameRequestCallbackDocs.IndexOf(aDocument) == mFrameRequestCallbackDocs.NoIndex', file layout/base/nsRefreshDriver.cpp, line 533
Comment 1 Boris Zbarsky [:bz] (TPAC) 2011-12-09 20:37:43 PST 
Worksforme with an m-c debug build pulled today (on Mac).
Comment 2 Jesse Ruderman 2011-12-10 11:17:58 PST 
I can reproduce with the same rev (built from http://hg.mozilla.org/mozilla-central/rev/63bff373cb94). Hmm. I'm on Mac OS X 10.6, if that matters.
Comment 3 Boris Zbarsky [:bz] (TPAC) 2011-12-14 14:23:51 PST 
I am too...  I guess I'll try building from your rev and reproducing.  Using a clean profile?
Comment 4 Jesse Ruderman 2011-12-14 14:57:39 PST 
I can reproduce with a clean profile. (Tested with rev 7e74a12601a4 this time.)
Comment 5 Boris Zbarsky [:bz] (TPAC) 2011-12-14 15:57:55 PST 
OK, now I can reproduce; I needed a narrower browser window to trigger tab scrolling.  This is totally my fault: canceling a callback doesn't unregister us (on purpose, sorta), but other code depends on the invariant that registered == have callbacks.
Comment 6 Boris Zbarsky [:bz] (TPAC) 2011-12-14 22:57:36 PST 
Created attachment 581883 [details] [diff] [review]
Manage our scheduled state better when animation frame requests are canceled.
Comment 7 Boris Zbarsky [:bz] (TPAC) 2011-12-15 05:54:41 PST 
https://hg.mozilla.org/integration/mozilla-inbound/rev/3eb0d7445611
Comment 8 Matt Brubeck (:mbrubeck) 2011-12-15 10:35:47 PST 
https://hg.mozilla.org/mozilla-central/rev/3eb0d7445611
Note You need to log in before you can comment on or make changes to this bug.