72081 - mutation events on anonymous content crash the browser

Status: VERIFIED FIXED

(Core :: XBL, defect, P2)

Description

[Alex Fritze]

see attached test case.
the problem is a null PresContext in nsDOMEvent::GetTarget().

Comment 1

[Alex Fritze]

Attachment: testcase (zipped xul/css/xbl)

Comment 2

[Heikki Toivonen (remove -bugzilla when emailing directly)]

I am pretty sure this was actually fixed by [e-mail] last
week/early this week. What is your build id? Can you retest with a recent build?

Comment 3

[David Hyatt]

Please give me the mutation event bugs.  They're all my fault. :)

Comment 4

[Alex Fritze]

Looks like this is indeed fixed. Testcase works in today's build.

Comment 5

[David Hyatt]

good

Comment 6

[Vladimir Ermakov]

On windows 98 build 2001-03-29-04-Mtrunk opening test.xul still crashes the 
browser... reopening

Comment 7

[David Hyatt]

Fixed.  Note your test case referenced a non-existent stylesheet which caused a 
crash.

Comment 8

[Vladimir Ermakov]

After removing 
<?xml-stylesheet href="test.css" type="text/css"?>

from test.xul there is no crash. What is the bug # for this stylesheet problem?

Verifying this bug on build 2001-04-10-04-trunk

Comment 9

[John Morrison]

What non-existent style sheet. Not in the attached testcase. 'test.css' 
exists, and the binding file 'test-binding.xml' also exists, but loading
test.xul crashes. (Sorry, no stack right now).

Comment 10

[Vladimir Ermakov]

John Morrison, [e-mail],

I tried to investigate further... its actually 
<mywidget/>
that is crashing the browser. I dont know much about xml, but this doesnt look 
like events issue. Can you please open a new bug with this problem? I'm 
resolving the events bug as worksforme

Comment 11

[John Morrison]

.

Comment 12

[John Morrison]

.

Comment 13

[Peter Trudelle]

.?  ->moz0.9.1/P2

Comment 14

[David Hyatt]

The chrome://math stylesheet.  It doesn't exist.  THat's why it crashes.

Comment 15

[John Morrison]

Okay, I'm blind and didn't see that in the binding file. So, it's not about
the mutation event.

But, is there a bug on crashing this way. Should a bug go to nisheeth, et al
for the XML content sink, where this is crashing?

Comment 16

[David Hyatt]

Yeah.  Give nisheeth the present.  YOu could just reopen this bug and change its
summary, since a convenient testcase is attached.

Comment 17

[John Morrison]

I think I wrap the present up in a shiny, brand-new bug. (Half the calories,
tastes great).