1170825 - [Messages][Draft] When deleting a draft in the messages app, the transition back to the main messages landing page exhibits flickering/stuttering

Status: RESOLVED FIXED

(Core :: Graphics, defect)

Description

[Derek Harris [:DerekH] - Dont need info. Last day 6/5/15]

Attachment: Messages Transition Logcat

Description:
When the user is transitioning back to the messages landing page by deleting a draft there will be a stuttering/flickering black effect on the right side of the screen. This issue is very noticiable.

Prerequisite: Do not have any existing message threads within the Messages app

Repro Steps:
1) Update a Flame to 20150602055237
2) Open Messages app
3) Type a number into the "To:" field
4) Tap the back button> Select "Delete Draft"


Actual:
A stuttering/flickering transition occurs when returning to the messages landing page


Expected:
The screen smoothly transitions back to the messages landing page


Environmental Variables:
Device: Flame 3.0 (319mb)(Kitkat)(Full Flash)
Build ID: 20150602055237
Gaia: 6d477a7884273886605049b20f60af5c1583a150
Gecko: 9eae3880b132
Gonk: 040bb1e9ac8a5b6dd756fdd696aa37a8868b5c67
Version: 41.0a1 (3.0)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:41.0) Gecko/41.0 Firefox/41.0


Repro frequency: 10/10
See attached: Logcat, Video - https://youtu.be/tbVv9uQJLsY

Comment 1

[Derek Harris [:DerekH] - Dont need info. Last day 6/5/15]

This issue does NOT occur on Flame 2.2, or Aries

The screen smoothly transitions back to the messages landing page

Environmental Variables:
Device: Flame 2.2 (319mb)(Kitkat)(Full Flash)
Build ID: 20150602002504
Gaia: a675e5472944a9d7a16e091ba8a59901e37df9e0
Gecko: c163c2a5b921
Gonk: bd9cb3af2a0354577a6903917bc826489050b40d
Version: 37.0 (2.2)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:37.0) Gecko/37.0 Firefox/37.0


Environmental Variables:
Device: Xperia Z3 Compact (B2G) 3.0
Build ID: 20150602115822
Gaia: Unknown
Gecko: 9eae3880b132
Version: 41.0a1 (3.0)
Firmware Version: D5803_23.1.A.1.28_NCB.ftf
User Agent: Mozilla/5.0 (Mobile; rv:41.0) Gecko/41.0 Firefox/41.0

Comment 2

[Oleg Zasypkin [:azasypkin]]

I can't see at the video well, but it looks like the same graphics issue we have in bug 1166211 (see attached video there as well).

Comment 3

[Julien Wajsberg [:julienw]]

yeah, I think so; let's revisit once bug 1166211 is fixed.

Comment 4

[Peter Bylenga [:PBylenga]]

NI reporter to check once bug 1166211 is fixed.

Comment 5

[Derek Harris [:DerekH] - Dont need info. Last day 6/5/15]

Moving NI

Comment 6

[Johan Lorenzo [:jlorenzo]]

Bug 1166211 landed. Verifyme as asked in comment 4.

Comment 7

[Duane Dixon [:ddixon]]

This issue still occurs in Flame 2.5, latest engineering build. 

Environmental Variables:
Device: Flame 2.5 (Shallow Flash)
BuildID: 20150730085341
Gaia: 1d3595836bd55b70478923d771051268a5dabf91
Gecko: f9ccbf328382
Version: 42.0a1 (2.5) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:42.0) Gecko/42.0 Firefox/42.0

Actual Results: 
Graphical issues occur after user selects "Delete Draft" or "Save as Draft" in Messaging app.

Comment 8

[Duane Dixon [:ddixon]]

Also, confirming that this issue DOES occur in the latest nightly Flame 2.5 build. 

Environmental Variables:
Device: Flame 2.5
BuildID: 20150730030209
Gaia: bf8565e0c3ad216ccb3f109c17f8a2eb2c42f6b8
Gecko: 62469b20ec84
Gonk: 41d3e221039d1c4486fc13ff26793a7a39226423
Version: 42.0a1 (2.5) 
Firmware Version: v18D (v4)
User Agent: Mozilla/5.0 (Mobile; rv:42.0) Gecko/42.0 Firefox/42.0

Comment 9

[KTucker [:KTucker][Inactive 3/4/2016]]

Let's find a regression window here.

Comment 10

[Oleg Zasypkin [:azasypkin]]

Hey,

Please make sure that you use the latest base image, it's likely a dupe of bug 1166211.

Thanks!

Comment 11

[Julien Wajsberg [:julienw]]

In comment 7, I see :
Firmware Version: v18D-1

But in comment 8, I see:
Firmware Version: v18D (v4)

Not sure which firmware these versions really relates.
You need the v3/v5 base image.

And I'm not really sure if we need to shallow flash or full flash on top of it (I was told "full flash" is the only correct way, but you'll tell me if this changes something).

Comment 12

[Pi Wei Cheng [:piwei] (inactive)]

v18D v4 is the latest base. v5 is actually a v3.5 which is older than v4 that's what I heard. The bug will reproduce regardless of base and full/shallow flashing.

Removing NI since the request has been fulfilled at comment 7/8.

Comment 13

[Pi Wei Cheng [:piwei] (inactive)]

mozilla-inbound regression window:

Last Working
Device: Flame
BuildID: 20150514061055
Gaia: 338f66e6a96491d2f5854b188c6b141ceb690d97
Gecko: 1189f6ffc65e
Version: 41.0a1 (2.5 Master) 
Firmware Version: v18Dv4
User Agent: Mozilla/5.0 (Mobile; rv:41.0) Gecko/41.0 Firefox/41.0

First Broken
Device: Flame
BuildID: 20150514090744
Gaia: 338f66e6a96491d2f5854b188c6b141ceb690d97
Gecko: fde31eaa2638
Version: 41.0a1 (2.5 Master) 
Firmware Version: v18Dv4
User Agent: Mozilla/5.0 (Mobile; rv:41.0) Gecko/41.0 Firefox/41.0

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

Gecko pushlog:
http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=1189f6ffc65e&tochange=fde31eaa2638

There are 4 graphics related bugs in the pushlog that could have caused the issue: Bug 1155828, Bug 1152461, Bug 1162726, and Bug 1160789.

We're picking bug 1162726 as a possible culprit now, but please do look at the pushlog and help on picking a culprit if you're able to.

Comment 14

[Markus Stange [:mstange]]

(In reply to Pi Wei Cheng [:piwei] from comment #13)
> There are 4 graphics related bugs in the pushlog that could have caused the
> issue: Bug 1155828, Bug 1152461, Bug 1162726, and Bug 1160789.
> 
> We're picking bug 1162726 as a possible culprit now, but please do look at
> the pushlog and help on picking a culprit if you're able to.

Bug 1152461 is a better pick, since that's the bug that made us use HardwareComposer more often, and this bug is likely a HardwareComposer bug.

Sotaro, is it possible that the base image regressed and reintroduced the multi-rect bug?

Comment 15

[Sotaro Ikeda [:sotaro]]

(In reply to Markus Stange [:mstange] from comment #14)
> 
> Bug 1152461 is a better pick, since that's the bug that made us use
> HardwareComposer more often, and this bug is likely a HardwareComposer bug.
> 
> Sotaro, is it possible that the base image regressed and reintroduced the
> multi-rect bug?

It is not clear now. I am going to check it next week.

Comment 16

[Sotaro Ikeda [:sotaro]]

Direct cause of the problem is optimizeLayerRects() works incorrectly. Disabling it fixed the problem. It seems to be caused by incorrect hwc layer blending setting. The setting is set by HwcComposer2D.

Comment 17

[Sotaro Ikeda [:sotaro]]

From the investigation, it becomes clear that this bug is dup of Bug 1085593. The fix is not applied to moz build flame-kk. optimizeLayerRects() does not handle a case that hwc layer has multiple rect.

Comment 18

[Sotaro Ikeda [:sotaro]]

Locally applying Bug 1085593 fixed the problem on master flame-kk.

Comment 19

[Sotaro Ikeda [:sotaro]]

(In reply to Sotaro Ikeda [:sotaro] from comment #18)
> Locally applying Bug 1085593 fixed the problem on master flame-kk.

To update ROM, full flash is necessary.

Comment 20

[Sotaro Ikeda [:sotaro]]

recent gecko uses multiple rect in one gfx layer more often this seems to trigger Bug 1085593.

Comment 21

[Sotaro Ikeda [:sotaro]]

Bug 1085593 Comment 56 applied the fix to moz build master flame-kk. And confirmed the problem is fixed on master flame-kk with full flash.

Comment 22

[Sotaro Ikeda [:sotaro]]

piwei, can you confirm the problem is fixed on master flame-kk with full flash. If the problem happen with shallow flash. It is a base image's problem, it does not have Bug 1085593 fix and POVB problem.

Comment 23

[Pi Wei Cheng [:piwei] (inactive)]

Confirmed that this issue is no longer occurring on latest Flame mozilla-inbound build. Setting bug as fixed.

Device: Flame (319MB, KK, full flashed)
BuildID: 20150805115039
Gaia: 581de383687dc441a878d2c91a0167c6ec688fef
Gecko: 6249e4edb023
Gonk: 41d3e221039d1c4486fc13ff26793a7a39226423
Version: 42.0a1 (2.5) 
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:42.0) Gecko/42.0 Firefox/42.0