99176 - figure out why we can have a canvas frame with no children

Status: NEW

(Core :: Layout, defect)

Description

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

This is a followup bug to bug 97226.  That bug was fixed with a null check that,
in theory, we shouldn't need, since a canvas frame should always (I think) have
children.  (Or is there a state of incremental loading where that might not be
true?)

Anyway, we should figure out whether the null check added for bug 97726 was the
right thing to do.  If it was, we should remove the comment that it's temporary.
 If it wasn't, we should do the right thing.

Comment 1

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Bug 97226 is the right number (and bug 97726 is not).

Comment 2

[Hixie (not reading bugmail)]

Thanks David; I'll look at this once I get a build environment.

In the meantime, do we have any reliable ways of hitting the assertion? Does the
testcase in bug 97226 reliably hit it? From that bug:

------- Additional Comments From Samuel Sieb 2001-08-28 15:37 -------
I've reduced it down to this javascript file:
http://serve.m-ave.net/jscript/MerchandisingAvenue.js

------- Additional Comments From David Baron 2001-09-03 07:40 -------
The URLs reported in talkback report comments are:
http://www.kissasylum.com/
http://tvguide.com/magazine/covers
http://tvguide.com/

Comment 3

[Andrew Schultz]

what is the status here?

The assertion and null check are still in nsCSSRending.cpp, but I was unable to
reproduce the assertion (A canvas with a background image had no child frame...)
using the URLs from bug 97226.

Comment 4

[Ioana (away)]

No assertions reproducible with the links in bug 97226 on the 08/21 debug Nightly.