1166211 - [Messages] There are black artifacts visible when user navigates to another panel right after dialog dismissal

Status: RESOLVED WORKSFORME

(Core :: Graphics, defect)

Description

[Oleg Zasypkin [:azasypkin]]

Attachment: Issue

STR:
* Open Messages app and enter new message panel;
* Add some text and recipients;
* Tap on Back button and choose "Save as a draft";
* Observe Message UI;

Expected behavior: "Save/Delete draft" dialog should be dismissed and user navigated to Inbox smoothly;

Actual result: user sees black artifacts and transition is jugged (see attached video).

I see this issue on the latest PVT flame-kk eng build. v2.2 looks fine. Likely Gecko regression.

Comment 1

[Brogan Zumwalt [Inactive]]

Last working Mozilla-Inbound build:
Device: Flame 3.0
Build ID: 20150514061055
Gaia: 338f66e6a96491d2f5854b188c6b141ceb690d97
Gecko: 1189f6ffc65e
Version: 41.0a1 (3.0)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:41.0) Gecko/41.0 Firefox/41.0

First broken Mozilla-Inbound build:
Device: Flame 3.0
Build ID: 20150514090744
Gaia: 338f66e6a96491d2f5854b188c6b141ceb690d97
Gecko: fde31eaa2638
Version: 41.0a1 (3.0)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:41.0) Gecko/41.0 Firefox/41.0


Working Gaia with Broken Gecko issue DOES reproduce:
Gaia: 338f66e6a96491d2f5854b188c6b141ceb690d97
Gecko: fde31eaa2638

Working Gecko with Broken Gaia issue does NOT reproduce:
Gaia: 338f66e6a96491d2f5854b188c6b141ceb690d97
Gecko: 1189f6ffc65e


Mozilla-Inbound Pushlog:
http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=1189f6ffc65e&tochange=fde31eaa2638


Issue appears to occur due to changes made in either bug 1152461 or bug 1162726

Comment 2

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

Markus, can you take a look at this please? This might have been caused by either the landing for bug 1152461 or bug 1162726.

Comment 3

[Markus Strangl]

This seems to have been misdirected - I'm no Mozilla developer, just a user/bugreporter.

Comment 4

[Oleg Zasypkin [:azasypkin]]

Hey Markus!

Any  news on this issue?

It see this quite frequently when using Messages app on Flame with the latest PVT eng build (not on spark/aries though).

Thanks!

Comment 5

[Markus Stange [:mstange]]

I haven't been able to reproduce this yet on my flame. I'm going to make a new build from master and test again.

Comment 6

[Oleg Zasypkin [:azasypkin]]

Okay, seems something wrong with our latest Flame's v18D base image. If I use v18D base image and shallow flash latest PVT Gecko and Gaia via b2g-flash-tool - I always see this issue, if I flash full image via b2g-flash-tool - I don't see this issue.

Not sure who we should raise this concern to.

Comment 7

[Markus Stange [:mstange]]

Interesting! I recently did a full B2G local build and flashed everything, so I might have picked up a fix from some upstream project.

Comment 8

[Julien Wajsberg [:julienw] (PTO -> Aug 14)]

Hey Naoki, does that ring some bell to you ? (re: the image respin)

Comment 9

[Sotaro Ikeda [:sotaro]]

> https://youtu.be/D4MX8Ydq6Io

From the video, hwc hal seems not correctly render a layer when the layer has multiple rectangles.

Comment 10

[Sotaro Ikeda [:sotaro]]

Since bug 1152461 fix, a layer often have multiple rectangles. Then some hwc hal's implementation's problem in comment 9 seems to cause a problem.

Comment 11

[Sotaro Ikeda [:sotaro]]

In the past Bug 1100255 and Bug 1085593 caused this symptom.

Recently Bug 1171201 caused this symptom on aries.

Comment 12

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

I think Sotaro-san answered the question and way better than I probably could have.  There are some fixes and most of the major fixes are listed in the intranet page.  I'm not sure if it's ok to open up the fixes to the public, I need to get Wesley's word on that.  Wesley, is it ok to open the fix list to the public?

Passing it back to you Julien.

Comment 13

[Sotaro Ikeda [:sotaro]]

(In reply to Oleg Zasypkin [:azasypkin] from comment #6)
> Okay, seems something wrong with our latest Flame's v18D base image. If I
> use v18D base image and shallow flash latest PVT Gecko and Gaia via
> b2g-flash-tool - I always see this issue, if I flash full image via
> b2g-flash-tool - I don't see this issue.
> 
> Not sure who we should raise this concern to.

This is POVB problem. And we do not have a source code of the ROM.

Comment 14

[Julien Wajsberg [:julienw] (PTO -> Aug 14)]

Naoki, Sotaro, do you know if the builds we have in pvtbuilds have different blobs than the v18D base build ?

I mean, how is it possible we get different results here? This is what I don't understand...

Comment 15

[Sotaro Ikeda [:sotaro]]

(In reply to Julien Wajsberg [:julienw] from comment #14)
> Naoki, Sotaro, do you know if the builds we have in pvtbuilds have different
> blobs than the v18D base build ?

We do not have source code of v18D base. But I think the code is different, because the flame-kk moz build refers very old caf code.

Comment 16

[Wesly Huang (EPM)]

(In reply to Naoki Hirata :nhirata (please use needinfo instead of cc) from comment #12)
> There are some fixes and most of the major fixes are listed in
> the intranet page.  I'm not sure if it's ok to open up the fixes to the
> public, I need to get Wesley's word on that.  Wesley, is it ok to open the
> fix list to the public?

Hi Naoki,

Sorry for late. I read all comments but not sure which list you mentioned here, would you give me more context?

Comment 17

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Wesley, the fix list for T2M builds, and the different fixes in between builds.  Would it be ok to open that to the public?  The ones that we know about that is...

Referring to the release change listed here:
https://intranet.mozilla.org/QA/B2G_Tips_and_Tricks#lastest_OEM_KitKat_Base_Image_:_v18D

Comment 18

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Hrm, well it's probably fixed with the contributor build or a fix in 18D.  Should we close this off as works for me?

Comment 19

[Oleg Zasypkin [:azasypkin]]

Yep, I don't see this issue anymore in v18D_v3 (or v5 how it's named when unpacked).

Thanks Naoki for noticing this!

Comment 20

[Wesly Huang (EPM)]

(In reply to Naoki Hirata :nhirata (please use needinfo instead of cc) from comment #17)
> Wesley, the fix list for T2M builds, and the different fixes in between
> builds.  Would it be ok to open that to the public?  The ones that we know
> about that is...
> 
> Referring to the release change listed here:
> https://intranet.mozilla.org/QA/
> B2G_Tips_and_Tricks#lastest_OEM_KitKat_Base_Image_:_v18D

Hi Naoki:

I think the "change list" itself should be fine to public since related issues are all opened in bugzilla already per my memory. But Asa as the lead of Flame would be a properer man to say it. (by the way is it a must-have to public? Or instead a good-to-have?)

@Asa: would you share you view here? Thanks.