1224209 - Header Bar on various website aren't fixed positioned properly and tend to float away or disappear during scrolling

Status: VERIFIED FIXED

(Core :: Layout, defect)

Description

[bull500]

Hello

On several website that tend to use a fixed header bar, the header bar seems to float away from their actual fixed positions upon scrolling the page

For example on http://www.theatlantic.com/world/ the jumping header bar issue occurs.  

And other situation is on http://foxtrotalpha.jalopnik.com/  
Here was happens is that the header bar disappears once you scroll down but when you start scrolling up slow/fast you can see the header bar to become visible in a shredded pattern along with the jump.  

Another site with this problem is http://www.popsci.com/

This was all tested with the FF45(2015-11-12) and Full screen mode enabled

Comment 1

[Kartikaya Gupta (email:[e-mail])]

Investigating these made me realize I screwed up the patch in bug 1214267 so I'll put a follow-up there. However I don't think my screwup is causing this bug, because it happens even with that fixed. It looks like the fixed-pos items aren't generating fixed layers but are getting flattened into the main content layer for some reason, and so we can't move them asynchronously. I'll try to track down a regression range for this.

Comment 2

[Kartikaya Gupta (email:[e-mail])]

I bisected fennec nightlies using popsci.com as the test URL. The regression occurred between the nov03 and nov04 nightlies, m-c pushlog range is http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=bb4d614a0b09&tochange=6077f51254c6

It seems to be a layout-side bug because the nov03 layers dump has the fixed-pos layer but the nov04 layers dump doesn't. I'll bisect inbound as well.

Comment 3

[Kartikaya Gupta (email:[e-mail])]

Regressed by bug 1156238; I checked the inbound builds with [1] and without [2] those patches. tn, can you take a look?

[1] http://archive.mozilla.org/pub/mobile/tinderbox-builds/mozilla-inbound-android-api-11/1446537955/
[2] http://archive.mozilla.org/pub/mobile/tinderbox-builds/mozilla-inbound-android-api-11/1446537726/

Comment 4

[Timothy Nikkel (:tnikkel)]

Looking into this.

Interestingly aurora doesn't have this bug, so I was right to go with a more conservative patch for uplift in bug 1199131.

Comment 6

[Kartikaya Gupta (email:[e-mail])]

[Tracking Requested - why for this release]: Pretty bad regression. The bug I just duped here was tracking 45+

Comment 7

[Kartikaya Gupta (email:[e-mail])]

Whoops, wrong flag.

Comment 8

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/7529d41e6f1b

Comment 9

[Kartikaya Gupta (email:[e-mail])]

*sigh* wrong bug number in the above push. I'll back it out and reland it with the right bug number.

Comment 10

[Timothy Nikkel (:tnikkel)]

Attachment: patch

https://treeherder.mozilla.org/#/jobs?repo=try&revision=12fcd48201b7

Comment 11

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/894ebb2becda

Comment 12

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/894ebb2becda

Comment 13

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/144f18fa10e9

Comment 14

[Timothy Nikkel (:tnikkel)]

^-- a reftest

Comment 15

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/144f18fa10e9

Comment 16

[Teodora Vermesan (:TeoVermesan)]

Verified as fixed using:
Device: Moto X (Android 4.4)
Build: Firefox for Android 45.0a1 (2015-11-22)

Comment 17

[Timothy Nikkel (:tnikkel)]

(In reply to Kartikaya Gupta (email:[e-mail]) from comment #9)
> *sigh* wrong bug number in the above push. I'll back it out and reland it
> with the right bug number.

That bug is bug 1214267 for future reference.