1161621 - [Flame] Factory reset via Settings doesn't enter FTE automatically, instead the screen turns off

Status: VERIFIED FIXED

(Firefox OS Graveyard :: Gaia::System, defect)

Description

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

Attachment: logcat of issue

Description:
During the process of factory reseting via Settings, the device will not automatically enter FTE, instead it turns screen off infinitely after showing the Firefox logo with blue background. If user presses power button during this screen off time, there is a chance they will interrupt with reset process and wake the device up to lockscreen.

STR:
1) Go to Settings > Device Information > Reset Phone > Reset
2) Observe the device go into reset process. Approximately 5 seconds after it shows the Firefox logo with blue background and turns the screen off, press the power button.

Expected: Device can be reset and automatically takes user to FTE when it's done without showing prolonged screen off time.

Actual: Device does NOT automatically takes user to FTE, instead the screen goes off after showing Firefox logo. If user presses power button while screen is off, there is a chance they will interrupt with reset process and reset will fail.

See video:
https://www.youtube.com/watch?v=DmLoM33cuYk

Repro rate:
4 out of 10 attempts. It seems to be related to timing, but I can't figure out the exact timing when will reliably occur.

Also attaching a logcat.

Device: Flame 3.0 Master (full flashed, 319MB, KK)
BuildID: 20150505010204
Gaia: 70077825aab2c7a79611befb40a5fe7e610d5443
Gecko: 102d0e9aa9e1
Gonk: a9f3f8fb8b0844724de32426b7bcc4e6dc4fa2ed
Version: 40.0a1 (3.0 Master)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:40.0) Gecko/40.0 Firefox/40.0

Comment 1

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

This issue does NOT occur on Flame 2.2. Immediately after it shows the Firefox logo blue background, it shows the Firefox logo with white background, and after that it displays the FTE without showing any kind of prolonged black screen. Bug repro rate: 0 out of 10 attempts.

Device: Flame 2.2 (full flashed, 319MB, KK)
BuildID: 20150505002501
Gaia: 772a9491909abd02dc67278dd453746e2dd358a8
Gecko: 2df83538ae20
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

Comment 2

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

[Blocking Requested - why for this release]:

Let's try to get a window here if possible. This seems like a recent regression.

Comment 3

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

I can't seem to reproduce the power-button-interrupt-reset bug using shallow flash. However the screen going off without auto-entering FTE bug is consistently reproducible (this applies to full flash as well) so I think I'll be focusing on that aspect of the bug. User is kind of forced to press the power button to see what's going on because their screen is turned off for an unreasonable amount of time.

Comment 4

[Gregor Wagner [:gwagner]]

(In reply to Pi Wei Cheng [:piwei] from comment #3)
> I can't seem to reproduce the power-button-interrupt-reset bug using shallow
> flash. However the screen going off without auto-entering FTE bug is
> consistently reproducible (this applies to full flash as well) so I think
> I'll be focusing on that aspect of the bug. User is kind of forced to press
> the power button to see what's going on because their screen is turned off
> for an unreasonable amount of time.

Please do and update the bug title accordingly.

Comment 5

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

b2g-inbound regression window:

Last Working
Device: Flame
BuildID: 20150430035620
Gaia: c85b9cb965fd1e718e9e86578e2c783d11124439
Gecko: 694eac9d36bb
Version: 40.0a1 (3.0 Master) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:40.0) Gecko/40.0 Firefox/40.0

First Broken
Device: Flame
BuildID: 20150430043019
Gaia: efb8e902e467bf7aa640cfe70d6ad470d0a4ceea
Gecko: 5e958d738bbb
Version: 40.0a1 (3.0 Master) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:40.0) Gecko/40.0 Firefox/40.0

Last Working Gaia First Broken Gecko - issue does NOT repro
Gaia: c85b9cb965fd1e718e9e86578e2c783d11124439
Gecko: 5e958d738bbb

Last Working Gecko First Broken Gaia - issue DOES repro
Gaia: efb8e902e467bf7aa640cfe70d6ad470d0a4ceea
Gecko: 694eac9d36bb

Gaia pushlog:
https://github.com/mozilla-b2g/gaia/compare/c85b9cb965fd1e718e9e86578e2c783d11124439...efb8e902e467bf7aa640cfe70d6ad470d0a4ceea

Caused by Bug 1115652.

Comment 6

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

Evan, can you take a look at this please? This might have been caused by the landing for bug 1115652.

Comment 7

[Evan Tseng [:evanxd]]

Hi Alive,

Looks like the root cause of this bug is that the `lockscreen-appopened` event is just triggered too early, right?

You said you have a patch to fix the `lockscreen-appopened` issue before. What is the bug number? We could set that this bug depends on it.

Thanks.

Comment 8

[Evan Tseng [:evanxd]]

Or if we want to have other way to fix this bug, I would like to do it.

Comment 9

[(inactive after 6/18) Alive Kuo [:alive]]

My patch is too big. Let's try to Query Service.query('LockscreenWindowManager.ready') on ScreenManager start to decide what to do.

If LockScreenWindowManager.ready => check Service.locked to reset the timeout to 10sec or the settings by the user (screen.timeout)

If LockScreenWindowManager.ready is false or undefined => listen to lockscreen-appopened to reset the timeout to 10secs.

Comment 10

[Henry Chang [:hchang]]

(In reply to Evan Tseng [:evanxd][:愛聞插低] from comment #8)
> Or if we want to have other way to fix this bug, I would like to do it.

I am also willing to verify if the patch fixes the emulator issue once 
there comes one.

Comment 11

[GitHub Autolander Bot]

Attachment: [gaia] evanxd:bug-1161621 > mozilla-b2g:master

Comment 12

[Evan Tseng [:evanxd]]

The bug is fixed by the patch.
Continue to add tests.

Comment 13

[Evan Tseng [:evanxd]]

Added tests and updated patch[1].

[1]: https://github.com/evanxd/gaia/commit/db608722ed141a7cde106a15addb5373a7ffe8fa

Comment 14

[Evan Tseng [:evanxd]]

Comment on attachment 8604044 [details] [review]
[gaia] evanxd:bug-1161621 > mozilla-b2g:master

Hi Alive,

Could you help to review the patch?

Thanks.

Comment 15

[(inactive after 6/18) Alive Kuo [:alive]]

Comment on attachment 8604044 [details] [review]
[gaia] evanxd:bug-1161621 > mozilla-b2g:master

Looks sane

Comment 16

[Evan Tseng [:evanxd]]

Thanks for the review, Alive.

Now waiting for CI[1].

[1]: https://treeherder.mozilla.org/#/jobs?repo=gaia&revision=4fea8b588438a4e77dcfa5858c92ea0cefd058a2

Comment 17

[Evan Tseng [:evanxd]]

master: https://github.com/mozilla-b2g/gaia/commit/702e614de7941c49ead2594cbbeface93fc9155e

Comment 18

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

This issue is verified fixed. Factory resetting the phone or just plain flashing to it now enters FTE automatically, leaving no chance for user to encounter the original issue that I reported about power button could interrupt the process. Bug repro rate: 0 out of 7 attempts.

Device: Flame (KK, full flashed, 319MB)
BuildID: 20150514010203
Gaia: 338f66e6a96491d2f5854b188c6b141ceb690d97
Gecko: 1fab94ad196c
Gonk: 040bb1e9ac8a5b6dd756fdd696aa37a8868b5c67
Version: 41.0a1 (3.0 Master) 
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:41.0) Gecko/41.0 Firefox/41.0

Comment 19

[GitHub Autolander Bot]

Attachment: [gaia] evanxd:v2.2 > mozilla-b2g:master