Closed Bug 1162207 Opened 9 years ago Closed 9 years ago

[Homescreen] Overscroll Animation 'tears' screen when switching between column layout of Three to Four (3 -> 4)

Categories

(Core :: Graphics: Layers, defect, P3)

ARM
Gonk (Firefox OS)
defect

Tracking

()

RESOLVED WORKSFORME
blocking-b2g 2.5?
Tracking Status
b2g-v2.1 --- unaffected
b2g-v2.2 --- affected
b2g-master --- unaffected

People

(Reporter: onelson, Assigned: nical)

References

()

Details

(Keywords: polish, regression, Whiteboard: [3.0-Daily-Testing][platform])

Attachments

(1 file)

Description:
When a user changes their column layout at homescreen from 'Three Columns' to 'Four Columns', they may observe an overscroll animation from the bottom of the screen (scrolling 'into' the bottom) will tear between layers of icons, as the graphic displays the remnants of the 'Three Column' layout that stood before it. Once the user locks their screen and unlocks, the animations appear to reset and the tearing no longer occurs.

Repro Steps:
1) Update a Flame to 20150506010204
2) Long press at homescreen to open System Dialog
3) Access 'Homescreen Settings'
4) Change setting fro 'Three Columns' to 'Four Columns'
5) Return to homescreen
6) Scroll 'into' the bottom (drag scroll up)
7) Observe screen

Actual:
Overscroll artifacts from 'Three Column' layout persist until screen reset

Expected:
Graphic display is as expected, no artifacts


Environmental Variables:
-------------------------------------------

Device: Flame 3.0
Build ID: 20150506010204
Gaia: 3e6fd1e0a478af2c95d09ce95c2c6de2de2fec14
Gecko: ba44099cbd07
Gonk: a9f3f8fb8b0844724de32426b7bcc4e6dc4fa2ed
Version: 40.0a1 (3.0)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:40.0) Gecko/40.0 Firefox/40.0

Device: Flame 2.2
BuildID: 20150506002501
Gaia: 772a9491909abd02dc67278dd453746e2dd358a8
Gecko: 3af6a0a79227
Gonk: ab265fb203390c70b8f2a054f38cf4b2f2dad70a
Version: 37.0 (2.2) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:37.0) Gecko/37.0 Firefox/37.0
*****************************************************

Issue DOES NOT REPRO on 2.1 for flame devices:
Results: Graphic display is as expected, no artifacts

Device: Flame 2.1
BuildID: 20150506001242
Gaia: b4a03b7ee61de5a479b3cf0916f47e91a43b0f50
Gecko: 4493015380ab
Gonk: ab265fb203390c70b8f2a054f38cf4b2f2dad70a
Version: 34.0 (2.1) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0
-------------------------------------------

Repro frequency: 5/5, stops after user resets their screen (display off to unlock)
See attached: 
video
logcat
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(pbylenga)
Keywords: regression
Whiteboard: [3.0-Daily-Testing] → [3.0-Daily-Testing], [systemsfe]
[Blocking Requested - why for this release]:
Visible graphical regression.

Starting with 3.0 nomination, 2.2 is affected as well.  Requesting a window.
blocking-b2g: --- → 3.0?
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(pbylenga)
QA Contact: pcheng
Kats, any idea?
Flags: needinfo?(bugmail.mozilla)
It's really hard to tell from the video what's going on. I can sort of reproduce, I see a 1-pixel-high row of glitchy pixels but I haven't seen anything like this before, so not sure what the root cause is. I'd like to see what the regression window turns up.
Flags: needinfo?(bugmail.mozilla)
mozilla-inbound regression window:

Last Working
Device: Flame
BuildID: 20141010070103
Gaia: cc5da7b055e2b06fdeb46fa94970550392ee571d
Gecko: 0fbe7cd05a11
Version: 35.0a1 (2.2 Master)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:35.0) Gecko/35.0 Firefox/35.0

First Broken
Device: Flame
BuildID: 20141010070603
Gaia: cc5da7b055e2b06fdeb46fa94970550392ee571d
Gecko: c61c91ea40fe
Version: 35.0a1 (2.2 Master)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:35.0) Gecko/35.0 Firefox/35.0

Gaia is the same so it's a Gecko issue.

Gecko pushlog:
http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=0fbe7cd05a11&tochange=c61c91ea40fe

Possibly caused by bug 1071769.
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(ktucker)
Nicolas, can you take a look at this please? Looks like this was caused by the work done on bug 1071769.
Blocks: 1071769
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(ktucker) → needinfo?(nical.bugzilla)
Huh, this is surprising. The relevant patch in this pushlog was backed out shortly after it landed and hasn't made it back since.
This is not in a specific branch that would have forked around that time or with specific prefs?
Flags: needinfo?(nical.bugzilla)
I wouldn't be surprised to see these artifacts if the pref "layers.tiled-drawtarget.enabled" is set to true (it is false by default everywhere except on mac), because edge-padding (the thing that is supposed to fix this issue) for DrawTargetTiled hasn't landed, but edge-padding should work for the code path currently used by default on b2g.
I haven't managed to reproduce this on my phone yet (using gecko master).
If I understand bug 1071769 comment 21 correctly, it only backed out one commit out of the four that were landed.
(In reply to Pi Wei Cheng [:piwei] from comment #8)
> If I understand bug 1071769 comment 21 correctly, it only backed out one
> commit out of the four that were landed.

I meant that in the pushlog from comment 4 there is only a patch that fuzzes a reftest and the one that got backed out.
I am looking at the other patches from the same bug but so far I haven't found anything that would cause this.
Assignee: nobody → nical.bugzilla
blocking-b2g: 3.0? → 3.0+
Component: Gaia::Homescreen → Graphics: Layers
Keywords: polish
Product: Firefox OS → Core
Whiteboard: [3.0-Daily-Testing], [systemsfe] → [3.0-Daily-Testing]
Does this show up in Aries as well, or just on Flame?
Making it a P3 for 2.5
Priority: -- → P3
Whiteboard: [3.0-Daily-Testing] → [3.0-Daily-Testing][platform]
QAnalysts, Can you please check if this issue still exists?

Thanks
Keywords: qawanted
[Blocking Requested - why for this release]:Removing blocker for 2.5 and nominating it for 2.5?
blocking-b2g: 2.5+ → 2.5?
I can no longer reproduce this issue on Aries and Flame 2.5. But the issue is still reproducible on Flame 2.2. On Aries and Flame 2.5 I tried on both legacy and default home screens and couldn't repro the bug.

Bug Repro's on:
Device: Flame 2.2
BuildID: 20151023032506
Gaia: 885647d92208fb67574ced44004ab2f29d23cb45
Gecko: 9e0aa88fea7a
Gonk: bd9cb3af2a0354577a6903917bc826489050b40d
Version: 37.0 (2.2) 
Firmware Version: v18Dv4
User Agent: Mozilla/5.0 (Mobile; rv:37.0) Gecko/37.0 Firefox/37.0

Bug does NOT repro on:
Device: Flame 2.5
BuildID: 20151023030241
Gaia: 410e91ddabc7ba82a9b43b3711a3fdf2cb8de309
Gecko: 0625c68c0abcfe4d10880d15d8fe7d06df3369c9
Gonk: 205ac4204bbbb2098a8046444acba551ba5dc75a
Version: 44.0a1 (2.5) 
Firmware Version: v18Dv4
User Agent: Mozilla/5.0 (Mobile; rv:44.0) Gecko/44.0 Firefox/44.0

Device: Aries 2.5
BuildID: 20151023103340
Gaia: 410e91ddabc7ba82a9b43b3711a3fdf2cb8de309
Gecko: 8ffcf080c975b372cc39a58b8b76d61de1767417
Gonk: 2916e2368074b5383c80bf5a0fba3fc83ba310bd
Version: 44.0a1 (2.5) 
Firmware Version: D5803_23.1.A.1.28_NCB.ftf
User Agent: Mozilla/5.0 (Mobile; rv:44.0) Gecko/44.0 Firefox/44.0
QA Whiteboard: [QAnalyst-Triage+] → [QAnalyst-Triage?]
Flags: needinfo?(jmercado)
Keywords: qawanted
Mahendra this issue no longer occurs on 2.5 but does still occur on 2.2 Flame.
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(jmercado) → needinfo?(mpotharaju)
Thanks Jayme/Pi Wie. 

Milan - Your question is answered now. Please look at this when you get a chance. Please mind that its a P3 for 2.5.
Flags: needinfo?(mpotharaju) → needinfo?(milan)
Good stuff.  If it shows up again, let's reopen the bug, otherwise, since it works in the latest build, we can close it.
Status: NEW → RESOLVED
Closed: 9 years ago
Flags: needinfo?(milan)
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: