1147194 - [Cost Control] Settings page displays a black screen

Status: VERIFIED FIXED

(Core :: Graphics: Layers, defect)

Description

[Yeojin Chung [:YeojinC]]

Attachment: logcat_20150324_1605.txt

Description:
Settings page displays a black screen after the mobile data usage is changed (ex. going on a website).

Pre-requisite: Data connection is enabled.

Repro Steps:
1) Update a Flame to 20150324010202.
2) Open Usage app, and go through the FTU steps.
3) Long-press the home button to bring out the task manager.
4) Close the Usage app.
5) Open Browser, and go to any website.
6) Press the home button to return to the homescreen, and open Usage app again.
7) Select the settings icon.

Actual:
A black screen appears below the status bar.

Expected:
The settings page appears properly.

Environmental Variables:
Device: Flame 3.0 (KK, 319mb, full flash)
Build ID: 20150324010202
Gaia: efebbafd12fc42ddcd378948b683a51106517660
Gecko: 840cfd5bc971
Gonk: b83fc73de7b64594cd74b33e498bf08332b5d87b
Version: 39.0a1 (3.0)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:39.0) Gecko/39.0 Firefox/39.0

Repro frequency: 5/5
See attached: video, logcat
https://youtu.be/TbclvTM5xsk

Comment 1

[Yeojin Chung [:YeojinC]]

This issue does NOT reproduce on Flame 2.2.

Result: The settings page appears properly.

Device: Flame 2.2 (KK, 319mb, full flash)
Build ID: 20150324002504
Gaia: 014d38f7ad3912b8b33cb08ce7535a5dc5aced59
Gecko: 7a9f2a248e57
Gonk: ebad7da532429a6f5efadc00bf6ad8a41288a429
Version: 37.0 (2.2)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:37.0) Gecko/37.0 Firefox/37.0

Comment 2

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

[Blocking Requested - why for this release]:

Nominating 3.0? since this a regression from 2.2 and leads to a broken settings page.

Let's get a regression window.

Comment 3

[Yeojin Chung [:YeojinC]]

Mozilla-inbound Regression Window:

Last Working Environmental Variables:
Device: Flame 3.0
BuildID: 20150316231229
Gaia: 4868c56c0a3b7a1e51d55b24457e44a7709ea1ae
Gecko: a80676087e0a
Version: 39.0a1 (3.0) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:39.0) Gecko/39.0 Firefox/39.0

First Broken Environmental Variables:
Device: Flame 3.0
BuildID: 20150316231828
Gaia: 4868c56c0a3b7a1e51d55b24457e44a7709ea1ae
Gecko: 18e46b2b3315
Version: 39.0a1 (3.0) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:39.0) Gecko/39.0 Firefox/39.0

Last Working Gaia First Broken Gecko: Issue DOES reproduce 
Gaia: 4868c56c0a3b7a1e51d55b24457e44a7709ea1ae
Gecko: 18e46b2b3315

First Broken Gaia Last Working Gecko: Issue does NOT reproduce
Gaia: 4868c56c0a3b7a1e51d55b24457e44a7709ea1ae
Gecko: a80676087e0a

http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=a80676087e0a&tochange=18e46b2b3315

possibly caused by bug 1135907

Comment 4

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

David, could you take a look at this please? This looks to be caused by the landing for bug 1135907.

Comment 5

[Wesley Huang [:wesley_huang] (EPM) (NI me)]

triage: major regression. but we should revisit if the 3.0 scope has changed

Comment 6

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

This reproduce on B2G 2.2 20150409002503 on Flame (v18D).

Comment 7

[Wesley Huang [:wesley_huang] (EPM) (NI me)]

triage: blocking.

Comment 8

[Johan Lorenzo [:jlorenzo]]

Based on regression window.

Comment 9

[Ethan Lin[:ethlin]]

There is a simple reproduce way:
1. Let the Mobile usage more than 0KB.
2. Open the Usage app and the open the setting page.

Wifi usage will not cause this problem!

I think the regression window is wrong because I tried "Gecko: a80676087e0a" also has this problem. That is because the reproduce way must be Mobile. I will look at this bug.

Comment 10

[Ethan Lin[:ethlin]]

Yeojin, I found the regression changeset should be Gecko:9d08f61b03f9. Can you help me confirm the changeset?

Issue DOES reproduce: 
Gaia: 4868c56c0a3b7a1e51d55b24457e44a7709ea1ae
Gecko: 9d08f61b03f9

Issue does NOT reproduce:
Gaia: 4868c56c0a3b7a1e51d55b24457e44a7709ea1ae
Gecko: c7dd154a7e83

Comment 11

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

That new regression range seems more likely to be correct based on the symptoms described.

Comment 12

[Yeojin Chung [:YeojinC]]

Could you provide the build ID from the pvt server? I cannot find those changeset just by that. And can you specify what branch those are?

Can you also provide me the pushlog please?

Comment 13

[Ethan Lin[:ethlin]]

Yeojin, the push log range should be from Gecko:24653749f9e0 to Gecko:9d08f61b03f9. 

http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=24653749f9e0&tochange=9d08f61b03f9

The branch is mozilla-center, and the build ID 20150317073344 is the first version which can reproduce. The build ID 20150317010203 should be the last version does NOT reproduce.

it's possibly caused by bug 1141595. Please check again, Thanks.

Comment 14

[Yeojin Chung [:YeojinC]]

Dear Ethan,

I was unable to reproduce on the central build that you pointed as first broken build:

Environmental Variables:
Device: Flame 3.0
BuildID: 20150317073344
Gaia: 738987bd80b0ddb4ccf853855388c2627e19dcc1
Gecko: 008b3f65a7e0
Version: 39.0a1 (3.0) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:39.0) Gecko/39.0 Firefox/39.0

I did not see the black screen, and the settings page appears fine. And I am still unable to find the correct builds with the gecko changeset. I'm sorry but I don't think I am able to help you confirming the issue anymore.

Comment 15

[Ethan Lin[:ethlin]]

I backed out the patches in bug 1141595 and the black screen problem is solved. With the patch, there is an additional container layer which contains a black paintedlayer. That should be the hoisted ScrollInfoLayer. Markus, can you take a look at this please?

Comment 16

[Markus Stange [:mstange]]

Ethan, is it possible that there are two black screen bugs here, with different regression ranges? I'm assuming that you tested your steps from comment 9 in order to get the regression range in comment 10, and Yeojin used the steps from comment 0 in order to get the regression range in comment 3.

Comment 17

[Milan Sreckovic [:milan] (needinfo for best results)]

Markus, please unassign this from you if it turns out you don't have an action on it.

Comment 18

[Ethan Lin[:ethlin]]

(In reply to Markus Stange [:mstange] from comment #16)
> Ethan, is it possible that there are two black screen bugs here, with
> different regression ranges? I'm assuming that you tested your steps from
> comment 9 in order to get the regression range in comment 10, and Yeojin
> used the steps from comment 0 in order to get the regression range in
> comment 3.

Markus, the two kinds of test steps is same. The steps in comment 9 is just a simple way to get the same result. I also tried to back out the patches in bug 1135907 which is found in comments 3, and I used exactly the same steps in comment1. But the black screen still appeared. So I think there should be only one black screen bug.

Comment 19

[Markus Stange [:mstange]]

Attachment: asserting testcase

Comment 20

[Markus Stange [:mstange]]

Attachment: patch

Comment 21

[Markus Stange [:mstange]]

The wrapping transform can turn inactive because at the time that we determine its layer state, the nsDisplayScrollInfoLayer's layer state is still LAYER_NONE, because its mHoisted is false. We only call MarkHoisted on it once we process the item in ContainerState::ProcessDisplayItems, and by then it's already too late to make the wrapping transform active.

Comment 22

[Markus Stange [:mstange]]

Attachment: rewrite hoisting a third time

As discussed on IRC. This looks much better.

Comment 23

[Markus Stange [:mstange]]

Comment on attachment 8599990 [details] [diff] [review]
rewrite hoisting a third time

Review of attachment 8599990 [details] [diff] [review]:
-----------------------------------------------------------------

Actually, let me try to come up with some better method names first.

Comment 24

[Markus Stange [:mstange]]

Attachment: rewrite hoisting a third time

Comment 25

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/ade50933c777

Comment 26

[Phil Ringnalda (:philor)]

https://hg.mozilla.org/mozilla-central/rev/ade50933c777

Comment 27

[Ryan VanderMeulen [:RyanVM]]

Please request b2g37 approval on this patch when you get a chance.

Comment 28

[Markus Stange [:mstange]]

Comment on attachment 8600005 [details] [diff] [review]
rewrite hoisting a third time

NOTE: Please see https://wiki.mozilla.org/Release_Management/B2G_Landing to better understand the B2G approval process and landings.

[Approval Request Comment]
Bug caused by (feature/regressing bug #): bug 1141595
User impact if declined: intermittent brief black screens
Testing completed: a little manual testing
Risk to taking this patch (and alternatives if risky): not particularly risky, has baked for weeks now
String or UUID changes made by this patch: none

Comment 29

[Milan Sreckovic [:milan] (needinfo for best results)]

No-Jun, any manual testing you want to do?

Comment 30

[Josh Cheng [:josh]]

Comment on attachment 8600005 [details] [diff] [review]
rewrite hoisting a third time

Approving and requesting QA verify after patch landed on 2.2

Comment 31

[No-Jun Park [:njpark]]

sure, Will check it out on master today

Comment 32

[Ryan VanderMeulen [:RyanVM]]

This needs rebasing for b2g37 uplift.

Comment 33

[Jayme Mercado [:JMercado] (Inactive 3/4/2016)]

This issue is verified fixed on the latest 3.0 Nightly Flame build.  Leaving the verifyme tag for when this fix gets uplifted to 2.2.

Actual Results: The settings screen looks as expected.

Environmental Variables:
Device: Flame 3.0
BuildID: 20150528010203
Gaia: 05380df3158fa39e1dde1687c0bf11a71f8c6868
Gecko: baa9c64fea6f
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

Comment 34

[Markus Stange [:mstange]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=c859217d2819

Comment 35

[Markus Stange [:mstange]]

Attachment: merged to b2g37

Giving this to tn for a quick look to make sure I didn't make any mistakes during the merge.

Comment 36

[Markus Stange [:mstange]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=d32ff8da77f9

Comment 37

[Markus Stange [:mstange]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=04dacb7a5f85

Comment 38

[Markus Stange [:mstange]]

Attachment: merged to b2g37

The previous patch didn't build. Hopefully this one will.

https://treeherder.mozilla.org/#/jobs?repo=try&revision=f03e44088204

Comment 39

[Markus Stange [:mstange]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=fa80e4e21e83

Comment 40

[Markus Stange [:mstange]]

Attachment: merged to b2g37

another try

https://treeherder.mozilla.org/#/jobs?repo=try&revision=681ff619ba4d

Comment 41

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-b2g37_v2_2/rev/6a53213b652f

Comment 42

[Shally Li（leave from Mozilla）]

This bug has been verified as pass on latest Nightly build of Flame v2.2 and Nexus 5 v2.2 by the STR in Comment 0.

Actual results: Settings page displays normally in Usage app after user views website in browser.
See attachment: verified_v2.2.mp4
Reproduce rate: 0/15


Device: Flame v2.2 build(Pass)
Build ID               20150607002503
Gaia Revision          8fc797527a3eca7665bc1d1828848f2fb77ca99f
Gaia Date              2015-06-04 07:46:11
Gecko Revision         https://hg.mozilla.org/releases/mozilla-b2g37_v2_2/rev/d213237e11e9
Gecko Version          37.0
Device Name            flame
Firmware(Release)      4.4.2
Firmware(Incremental)  eng.cltbld.20150607.035848
Firmware Date          Sun Jun  7 03:58:59 EDT 2015
Bootloader             L1TC000118D0

Device: Nexus 5 v2.2 build(Pass)
Build ID               20150607002503
Gaia Revision          8fc797527a3eca7665bc1d1828848f2fb77ca99f
Gaia Date              2015-06-04 07:46:11
Gecko Revision         https://hg.mozilla.org/releases/mozilla-b2g37_v2_2/rev/d213237e11e9
Gecko Version          37.0
Device Name            hammerhead
Firmware(Release)      5.1
Firmware(Incremental)  eng.cltbld.20150607.035410
Firmware Date          Sun Jun  7 03:54:26 EDT 2015
Bootloader             HHZ12f

Comment 43

[Shally Li（leave from Mozilla）]

Attachment: verified_v2.2.mp4

Leaving "verifyme" for "firefox40" verification