877716 - tbpl.mozilla.org is cut off due to improper multicolumn rendering

Status: NEW

(Core :: Layout, defect)

Description

[Scott Johnson (:jwir3)]

(See also bug 857324, comment 49 and bug 857324, comment 50).

Tbpl.mozilla.org is cut off after loading a talos test (e.g. 'd') and then restoring the screen from a maximized state.

This still happens as of the 29-May-2013 Nightly, but happens as early as:

parent: 123959:99ec1a721466 
 Bug 841289 - Remove prefable option from WebIDL codegen. r=bz

Since bug 857324 is specifically about a regression from bug 810726, which landed in revision 123960, this is a separate issue.

Comment 1

[Scott Johnson (:jwir3)]

Using mozregression, this apparently dates back as far as the nightly for 2011-03-17. Prior to that, mozregression was giving me builds that simply crashed, but I know for certain that the nightly for 2010-02-07 seemed to work fine.

Comment 2

[Paul Silaghi, QA [:pauly]]

I've investigated this a little deeper and from the beginning until now, I found 3 ranges:
1)
Last good nightly: 2011-09-29
First bad nightly: 2011-09-30
http://img14.imageshack.us/img14/1838/scroll1.png
Pushlog:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=cb4b93331e4f&tochange=b5b082d183d0

2)
Last good nightly: 2012-07-31
First bad nightly: 2012-08-01
http://img89.imageshack.us/img89/1570/scroll2s.png
Pushlog:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=9d2a7a8ca1c7&tochange=582d4c67b3a7
Considering this I would say bug 857324 isn't a regression from bug 810726 which landed on 2013-03-07.

3)
Last good nightly: 2013-04-24
First bad nightly: 2013-04-25
http://img10.imageshack.us/img10/4294/scroll3.png
Pushlog:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=fef5f202b2dc&tochange=690b5e0f6562

Comment 3

[Scott Johnson (:jwir3)]

Ok, this makes a bit more sense. It looks like this was a much more complicated regression range that I had originally anticipated.

Comment 4

[Scott Johnson (:jwir3)]

The second regression range in comment 2 is being tracked in bug 898452. We should re-evaluate whether this bug has been fixed after that one lands.