782660 - Rendering glitches when resizing a textarea with rows and cols attributes

Status: RESOLVED FIXED

(Core :: Layout, defect)

Description

[(no longer active)]

Try resizing the first textarea here: https://bug82711.bugzilla.mozilla.org/attachment.cgi?id=36296

If I were a mean person, I would blame DLBI patches landed yesterday!

Comment 1

[Matt Woodrow (:mattwoodrow)]

Not obviously something that my patches would have caused, and flipping the disable pref for the biggest change doesn't fix it.

Worth getting a regression window for this.

Comment 2

[(no longer active)]

CCing Alice who knows all of the bisection magic!

Comment 3

[Alice0775 White]

I can reproduce in Aurora16.a2 and Nightly17.0a1 with/without HWA.

#1 Initial-Regression window(m-i)
Good:
http://hg.mozilla.org/integration/mozilla-inbound/rev/daf9c42be518
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/16.0 Firefox/16.0 ID:20120623010444
Bad:
http://hg.mozilla.org/integration/mozilla-inbound/rev/ad006bd70b37
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/16.0 Firefox/16.0 ID:20120623060947
Pushlog:
http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=daf9c42be518&tochange=ad006bd70b37

#2 Progression window(m-i)
Bad:
http://hg.mozilla.org/integration/mozilla-inbound/rev/0d9f7fb55226
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/16.0 Firefox/16.0 ID:20120629192951
Working again
http://hg.mozilla.org/integration/mozilla-inbound/rev/cd6d52bdf2d8
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/16.0 Firefox/16.0 ID:20120629200651
Pushlog:
http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=0d9f7fb55226&tochange=cd6d52bdf2d8

#3 Re-Regression window(m-i)
Good:
http://hg.mozilla.org/integration/mozilla-inbound/rev/de483b66eb94
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/16.0 Firefox/16.0 ID:20120703174600
Broken again
http://hg.mozilla.org/integration/mozilla-inbound/rev/0a009343d59d
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:16.0) Gecko/16.0 Firefox/16.0 ID:20120703175500
Pushlog:
http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=de483b66eb94&tochange=0a009343d59d

Comment 4

[Matt Woodrow (:mattwoodrow)]

Awesome, thanks Alice!

Looks like a regression from bug 716875 which was temporarily fixed by DLBI (until it got backed out again).

CC'ing relevant people.

Comment 5

[Alice0775 White]

>#1 Initial-Regression window(m-i)
Gray vertical border remains when I extend the width of the textarea,

However, before #1 regression,there is rendering bug where white dots are left in the top border when I extend the width of the textarea since Fireox4.

Comment 6

[Charly Molter :lahabana]

Hmm so I don't really know what could caused that. I'm ok to work on that but I think I'll need a hand to really understand.
I guess that's caused by something wrong in the reflow....

Comment 7

[Matt Woodrow (:mattwoodrow)]

We're probably using the visual overflow rect to invalidate, and it's not big enough to cover all the pixel that get painted.

I'd start by checking what's different with the calculation of that.

Comment 8

[Charly Molter :lahabana]

Hmm that's clearer for me I'll have a look too tomorrow

Comment 9

[Charly Molter :lahabana]

Attachment: test case

This is just the test case that was given with a bigger border. It seems that the border is not included in the visual-overflow rect... So I would think it comes from nsControlFrame::reflow maybe we lose the border when we UnionWith in ReflowTextControlChild?
I'm honestly a bit lost on that if somebody could help me.

Comment 10

[Boris Zbarsky [:bzbarsky]]

I just stepped through nsTextControlFrame::Reflow, and at the end the overflow areas sure seem to match the actual frame size....

Comment 11

[Boris Zbarsky [:bzbarsky]]

The problem is that there's no CheckInvalidateSizeChange in nsTextControlFrame::Reflow.

Which is why DLBI fixes this, of course.

Comment 12

[Boris Zbarsky [:bzbarsky]]

Attachment: Resizing a text control frame with borders should invalidate the borders as needed.

Not sure how to write a test for this....

Comment 13

[Boris Zbarsky [:bzbarsky]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/747584155b62

Comment 14

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 655616 [details] [diff] [review]
Resizing a text control frame with borders should invalidate the borders as needed.

[Approval Request Comment]
Bug caused by (feature/regressing bug #): Bug 716875
User impact if declined: Turds when textareas are resized
Testing completed (on m-c, etc.): Passes manual tests.
Risk to taking this patch (and alternatives if risky): Risk is low; the worst
   case would be an overinvalidation leading to performance issues, but that
   should be pretty unlikely: we use CheckInvalidateSizeChange pretty extensively
   already.
String or UUID changes made by this patch: I know better by now.  None.  ;)

Comment 15

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/747584155b62

Comment 16

[Boris Zbarsky [:bzbarsky]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/a84cbf12b48d
https://hg.mozilla.org/releases/mozilla-beta/rev/c428ed8cdb16

Comment 17

[Scoobidiver (away)]

Backout from Beta: http://hg.mozilla.org/releases/mozilla-beta/rev/6f7d29816c57

Comment 18

[Boris Zbarsky [:bzbarsky]]

You meant "unaffected": the bug that caused this one was backed out too.  But maybe it should have stayed "fixed".  Who knows.  In any case, "affected" is wrong.

Comment 19

[Simona Badau, Desktop QA]

Mozilla/5.0 (Macintosh; Intel Mac OS X 10.7; rv:17.0) Gecko/17.0 Firefox/17.0

Verified that there are no rendering glitches on Firefox 17 beta 3 when resizing a textarea with rows and cols attributes (used the test cases attached in the Description and in Comment 9).

Comment 20

[[:Aleksej]]

WFM with ftp://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/18.0b1-candidates/build1/linux-x86_64/ru/ , compared to 2012-08-27-03-05-49-mozilla-central-firefox-17.0a1.ru.linux-x86_64

Comment 21

[Tracy Walker [:tracy]]

mass remove verifyme requests greater than 4 months old