1231228 - Stretch Homescreen then swipe to the bottom it will not return to the initial state

Status: VERIFIED FIXED

(Core :: Panning and Zooming, defect)

Description

[Max Ivanov[:MaxIvanov]]

Description:
After stretching Homescreen and swiping to the bottom it will stuck and not return to the initial state  

Repro Steps:
1) Update a Aries to 20151208121557
2) Scroll Homescreen to the right and make swipe to the bottom
3) Observe Homescreen state

Actual:
Homescreen stuck and not returning to the initial state

Expected:
We expect that Homescreen should returns to initial state

Environmental Variables:
Device: Aries 2.6
Build ID: 20151208121557
Gaia: 6b430ea7274af4c352de16b75e6bb85d7621ca83
Gecko: 2bdd9ec79799eff3ceec0a318f5a0632d918a527
Gonk: a19052e4389c3ae2d8fc3e7a74a475401baacc56
Version: 45.0a1 (2.6)
Firmware Version: D5803_23.1.A.1.28_NCB.ftf
User Agent: Mozilla/5.0 (Mobile; rv:45.0) Gecko/45.0 Firefox/45.0

Repro frequency: (100 %)
See attached: (video clip, logcat)
https://youtu.be/J5kM-sD5_ZM

Comment 1

[Max Ivanov[:MaxIvanov]]

This does occur on  Flame 2.6 and Flame 2.5
Homescreen stuck and not returns to the initial state

Device: Flame 2.6 (KK)(512mb)
Build ID: 20151208030225
Gaia: 6b430ea7274af4c352de16b75e6bb85d7621ca83
Gecko: 2bdd9ec79799eff3ceec0a318f5a0632d918a527
Gonk: 205ac4204bbbb2098a8046444acba551ba5dc75a
Version: 45.0a1 (2.6)
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:45.0) Gecko/45.0 Firefox/45.0

Device: Flame 2.5 (KK)(512mb)
Build ID: 20151208120554
Gaia: 2d54c29f429bed790b5d8284633812dc2b782518
Gecko: ff31a251b2f6149edf4fc0a199133ef2e190ceac
Gonk: 205ac4204bbbb2098a8046444acba551ba5dc75a
Version: 44.0a2 (2.5)
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:44.0) Gecko/44.0 Firefox/44.0
----------------------------------------------

This does NOT occur on  Flame 2.2
There is no stretch effect when users scroll Homescreen to the right

Device: Flame 2.2 (KK)(512mb)
Build ID: 20151207032502
Gaia: 885647d92208fb67574ced44004ab2f29d23cb45
Gecko: 4381c4b69b9c
Gonk: bd9cb3af2a0354577a6903917bc826489050b40d
Version: 37.0 (2.2)
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:37.0) Gecko/37.0 Firefox/37.0

Comment 2

[Max Ivanov[:MaxIvanov]]

Attachment: logs.txt

Comment 3

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

Let's get a window here.  Not nominating to block since scrolling the screen in any direction will fix the issue.

Comment 4

[Sxean Lee-David [:sxean]]

Regression Window: Mozilla-Inbound

Last Working: 
Device: Flame 2.5
BuildID: 20151016035729
Gaia: 8ea9029190af2ffeb04dcd97b323738125e31a0e
Gecko: 33b06ee164424bfec45d78b2bc33a741c87cd4d1

First Broken:  
Device: Flame 2.5
BuildID: 20151016063734
Gaia: 8999f0ba6326d815c8366e3c1155b7e4e9763b40
Gecko: 4fda345579d439eabcc83745accc2eeac9fbb56f

Last Working Gaia First Broken Gecko: Issue does NOT reproduce 
Device: Flame 2.5
BuildID: 20151016035729
Gaia: 8ea9029190af2ffeb04dcd97b323738125e31a0e
Gecko: 4fda345579d439eabcc83745accc2eeac9fbb56f

First Broken Gaia Last Working Gecko: Issue DOES reproduce  
Device: Flame 2.5
BuildID: 20151016063734
Gaia: 8999f0ba6326d815c8366e3c1155b7e4e9763b40
Gecko: 33b06ee164424bfec45d78b2bc33a741c87cd4d1

Mozilla Inbound Pushlog: 
http://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=33b06ee164424bfec45d78b2bc33a741c87cd4d1&tochange=4fda345579d439eabcc83745accc2eeac9fbb56f

May have been caused due to changes made by Bug 1205559

Comment 5

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

That seems *very* unlikely, can we retest the window? None of the changes in the regression window seem like a likely culprit.

Comment 6

[Sxean Lee-David [:sxean]]

(In reply to Kartikaya Gupta (email:[e-mail]) from comment #5)
> That seems *very* unlikely, can we retest the window? None of the changes in
> the regression window seem like a likely culprit.

I will double check the Regression Window, again.

Comment 7

[Sxean Lee-David [:sxean]]

(In reply to Kartikaya Gupta (email:[e-mail]) from comment #5)
> That seems *very* unlikely, can we retest the window? None of the changes in
> the regression window seem like a likely culprit.

The RW is "good" and repeatable based on the builds given. IMO, the code that enabled the horizontal stretch is where the culprit resides. However, the RW is not pointing to those particular changes. 
The point where the new Homescreen was implemented with the horizontal stretching feature is most likely.

Comment 8

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

Let's have someone else look at the window just in case.

Comment 9

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

1. comment 4 window is for when 'new homescreen' becomes the default homescreen.

2. the below window is found using the 'new home screen' option that hides within settings before 'new home screen' becomes default.

3. the bug occurs as long as home screen is horizontally scrollable. so the root cause could be long before this change, it's just that this change made the bug visible.

Last Working
Device: Flame 2.5
BuildID: 20151015113536
Gaia: 8ea9029190af2ffeb04dcd97b323738125e31a0e
Gecko: d374d16cbb251c9dac5af69f8e186e821ce82fe2
Version: 44.0a1 (2.5) 
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:44.0) Gecko/44.0 Firefox/44.0

First Broken
Device: Flame 2.5
BuildID: 20151016041835
Gaia: 8999f0ba6326d815c8366e3c1155b7e4e9763b40
Gecko: ccf288f658211b6cfab33c458aaf033baed2375b
Version: 44.0a1 (2.5) 
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:44.0) Gecko/44.0 Firefox/44.0

Last Working Gaia First Broken Gecko - 'new home screen' is NOT horizontally scrollable
Gaia: 8ea9029190af2ffeb04dcd97b323738125e31a0e
Gecko: ccf288f658211b6cfab33c458aaf033baed2375b

Last Working Gecko First Broken Gaia - 'new home screen' is horizontally scrollable, and bug repros
Gaia: 8999f0ba6326d815c8366e3c1155b7e4e9763b40
Gecko: d374d16cbb251c9dac5af69f8e186e821ce82fe2

Gaia pushlog:
https://github.com/mozilla-b2g/gaia/compare/8ea9029190af2ffeb04dcd97b323738125e31a0e...8999f0ba6326d815c8366e3c1155b7e4e9763b40

This issue is possibly caused by changes made in Bug 1214607.

Comment 10

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

Chris it seems the earliest we can see this issue is your change for bug 1214607, but it doesn't seem likely that's the true cause.  Hopefully you can see something we can't.

Comment 11

[Chris Lord [:cwiiis]]

(In reply to Jayme Mercado [:JMercado] from comment #10)
> Chris it seems the earliest we can see this issue is your change for bug
> 1214607, but it doesn't seem likely that's the true cause.  Hopefully you
> can see something we can't.

Right, I expect this is just a bug in the handling of nested scrolling layers in apz. Probably unrelated, but I'll mention bug 1221395 in case. Passing needinfo to botond to have a look when there's time.

If you want to test beyond bug 1214607, pinning a page will let you test further, but I expect you'll just hit the commit the switched to the new homescreen by default (and if you switch the homescreen manually, I expect you'll hit the commit that added pinned page support). Might be worth checking all the same.

Comment 12

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

Thanks, I agree that it's probably not a regression but a pre-exiating bug that was exposed. We can look into this at some point since it's in 2.5 but again probably not very high priority right now.

Comment 13

[Botond Ballo [:botond]]

Looks like the the "filter out a second touch if the second finger goes down while the first finger has panned into overscroll" code added in bug 1022719 is not kicking in.

Comment 14

[Botond Ballo [:botond]]

We're testing for "is any APZC along the handoff chain panned into overscroll" with:

  BuildOverscrollHandoffChain(mApzcForInputBlock)->HasApzcPannedIntoOverscroll()

where that checks for each APZC:

  IsOverscrolled() && IsInPanningState()

The problem with this is that the APZC that's in the panning state will always be the event-target APZC, but the overscrolled APZC can be any APZC along the chain; if it's not the event-target APZC, then no APZC will simultaneously be in the panning state and overscrolled.

Comment 15

[Botond Ballo [:botond]]

Attachment: MozReview Request: Bug 1231228 - Correctly check for whether an APZC along the handoff chain is panned into overscroll. r=kats

Bug 1231228 - Correctly check for whether an APZC along the handoff chain is panned into overscroll. r=kats

Comment 16

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

Comment on attachment 8698283 [details]
MozReview Request: Bug 1231228 - Correctly check for whether an APZC along the handoff chain is panned into overscroll. r=kats

https://reviewboard.mozilla.org/r/27911/#review25059

r+ assuming this doesn't regress bug 1193930.

Comment 17

[Botond Ballo [:botond]]

Verified that the patch does not regress bug 1193930 (because we still have the IsInPanningState() check).

Comment 18

[Botond Ballo [:botond]]

Attachment: MozReview Request: Bug 1231228 - Gtest. r=kats

Review commit: https://reviewboard.mozilla.org/r/28175/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/28175/

Comment 19

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

Comment on attachment 8699013 [details]
MozReview Request: Bug 1231228 - Gtest. r=kats

https://reviewboard.mozilla.org/r/28175/#review25205

Thanks!

Comment 20

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/ca4192a91755
https://hg.mozilla.org/integration/mozilla-inbound/rev/d65e209839b4

Comment 21

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/ca4192a91755
https://hg.mozilla.org/mozilla-central/rev/d65e209839b4

Comment 23

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

I am still able to reproduce this issue ONLY when the steps are performed immediately upon reaching the homescreen for the first time since the phone has been flashed or restarted.  Kartikaya would you like a separate bug filed for this instance?

Environmental Variables:
Device: Aries 2.6
BuildID: 20160112105459
Gaia: 3c97d6a8ac5a69662e1e2c22a84ea59bf50c305e
Gecko: e790bba372f14241addda469a4bdb7ab00786ab3
Gonk: a19052e4389c3ae2d8fc3e7a74a475401baacc56
Version: 46.0a1 (2.6) 
Firmware Version: D5803_23.1.A.1.28_NCB.ftf
User Agent: Mozilla/5.0 (Mobile; rv:46.0) Gecko/46.0 Firefox/46.0

Environmental Variables:
Device: Flame 2.6
BuildID: 20160112030203
Gaia: 3c97d6a8ac5a69662e1e2c22a84ea59bf50c305e
Gecko: e790bba372f14241addda469a4bdb7ab00786ab3
Gonk: 205ac4204bbbb2098a8046444acba551ba5dc75a
Version: 46.0a1 (2.6) 
Firmware Version: v18D
User Agent: Mozilla/5.0 (Mobile; rv:46.0) Gecko/46.0 Firefox/46.0

Comment 24

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

If you have reliable STR, then yes, please file a new bug for this issue. Thanks!

Comment 25

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

Comment 23 has been written up as bug 1241605.

Comment 26

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

Marking this as verified based on comment 23 since bug 1240202 has been written.