Bug 9463 – REGRESSION (r14879): Assertion failure in CSSParser::sinkFloatingSelector() (m_floatingSelectors.contains(selector)) in some CSS tests

• Component: CSS
• Status: RESOLVED FIXED

The debug build of TOT fails on css1/conformance/forward_compatible_parsing.html and fast/css/css3-modsel-22.html with this error:

ASSERTION FAILED: m_floatingSelectors.contains(selector) (WebCore/css/cssparser.cpp:2864 WebCore::CSSSelector* WebCore::CSSParser::sinkFloatingSelector(WebCore::CSSSelector*))

[David Kilzer (:ddkilzer)]

Probably related to the fix for Bug 3233 (r14879).

(In reply to comment #1)
> Probably related to the fix for Bug 3233 (r14879).
> 

Yes, this is a regression from that fix (says so in the bug summary ;-) ).

[David Kilzer (:ddkilzer)]

Created attachment 8866 [details]
Patch v1

This patch fixes the two regressions in this bug.  The full test suite is still running on my PB G4, though.

[David Kilzer (:ddkilzer)]

(In reply to comment #2)
> (In reply to comment #1)
> > Probably related to the fix for Bug 3233 (r14879).
> Yes, this is a regression from that fix (says so in the bug summary ;-) ).

It was 5:30 AM local time when I wrote that.  I was not fully mentally engaged yet.  :)

[David Kilzer (:ddkilzer)]

(In reply to comment #3)
> Created an attachment (id=8866) [edit]
> Patch v1
> 
> This patch fixes the two regressions in this bug.  The full test suite is still
> running on my PB G4, though.

All of the tests passed.  This looks like the fix.

[Nicholas Shanks]

Thanks. I had no clue what the cause was, nor how to go about fixing it. :-/

[Darin Adler]

Comment on attachment 8866 [details]
Patch v1

r=me

[David Kilzer (:ddkilzer)]

Committed revision 14895.

[David Kilzer (:ddkilzer)]

(In reply to comment #6)
> Thanks. I had no clue what the cause was, nor how to go about fixing it. :-/

This fix was brought to you by Sesame Street:  "One of these things is not like the others.  One of these things just doesn't belong."

http://members.tripod.com/tiny_dancer/one.html