Closed Bug 1071768 Opened 10 years ago Closed 6 years ago

[Rocket bar] In a flash based website zooming in turning the flame to landscape mode and locking phone causes browser to close out.

Categories

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

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(tracking-b2g:backlog, b2g-v2.0 unaffected, b2g-v2.1 affected, b2g-v2.2 affected)

RESOLVED WONTFIX
tracking-b2g backlog
Tracking Status
b2g-v2.0 --- unaffected
b2g-v2.1 --- affected
b2g-v2.2 --- affected

People

(Reporter: rmitchell, Unassigned)

References

()

Details

(Keywords: regression, Whiteboard: [2.1-exploratory-2] [systemsfe])

Attachments

(2 files)

Attached file log cat
Description:
In a flash based website(newgrounds, theprp, metalinjection, ect) with x-heavy on SD card zooming in turning the flame to landscape mode and locking phone causes browser to close out.


Setup: run X-heavy script 

Repro Steps:
1) Update a Flame to 20140923003005
2) Search a website with flash elements on the Rocket bar
3) Zoom into the website as far as flame will allow
4) Turn to landscape mode > tap power button to lock the screen


Actual:
Website is closed out completely when screen is unlocked 


Expected:
Website is still launched and reloads the page 

Environmental Variables:
Device: Flame 2.1
Build ID: 20140923003005
Gaia: 3742913e11f69e789dcb0aa0dedf2e5572da0129
Gecko: df42b05782aa
Version: 34.0a2 (2.1)
Firmware Version: 123
User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0


Repro frequency:100%
See attached: logcat, video clip: https://www.youtube.com/watch?v=hxMS0sv5eDI
This issue DOES occur on Flame 2.2(319mb) Kit Kat, Flame 2.2(319mb) Jelly bean, ,Open C 2.2, , Open C 2.1, Flame 2.1 jelly bean 


In a flash based website zooming in turning the flame to landscape mode and locking phone causes browser to close out.

Flame 2.2 Jelly Bean Base (319mb)

Environmental Variables:
----------------------------------------
Device: Flame 2.2 Master
BuildID: 20140923040203
Gaia: 3c898380b47f298cd3b7a0dacb3a6529e94322d4
Gecko: 790f41c631cc
Version: 35.0a1 (2.2 Master)
Firmware: V123
User Agent: Mozilla/5.0 (Mobile; rv:35.0) Gecko/35.0 Firefox/35.0



2.2 Open C

Environmental Variables:
Device: Open_C Master
Build ID: 20140923040203
Gaia: 3c898380b47f298cd3b7a0dacb3a6529e94322d4
Gecko: 790f41c631cc
Version: 35.0a1 (Master)
Firmware Version: P821A10V1.0.0B06_LOG_DL
User Agent: Mozilla/5.0 (Mobile; rv:35.0) Gecko/35.0 Firefox/35.0

Flame 2.1 Jelly Bean Base (319mb)

Enviromental Variables:
Device: Flame 2.1
BuildID: 20140923000204
Gaia: 3742913e11f69e789dcb0aa0dedf2e5572da0129
Gecko: df42b05782aa
Version: 34.0a2 (2.1)
Firmware: V123
User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0

2.1 Open C

Environmental Variables:
Device: Open_C 2.1
Build ID: 20140923000204
Gaia: 3742913e11f69e789dcb0aa0dedf2e5572da0129
Gecko: df42b05782aa
Version: 34.0a2 
Firmware Version: P821A10V1.0.0B06_LOG_DL
User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0



Issue does not occur in Flame 2.1 Kitkat(512mb) 2.0 KitKat Base (319mb) flame 2.0(319 mb) open C 2.0, 
Web site is still launched and reloads the page 

Flame 2.1 KitKat Base (512mb)

Environmental Variables:
Device: Flame 2.1
BuildID: 20140923003005
Gaia: 3742913e11f69e789dcb0aa0dedf2e5572da0129
Gecko: df42b05782aa
Version: 34.0a2 (2.1)
Firmware: V180
User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0

Flame 2.0 KitKat Base (319mb)

Environmental Variables:
Device: Flame 2.0
Build ID: 20140923063014
Gaia: 6449cc35a8f0704d95acac374ba857bde4b86d6c
Gecko: b930730dba81
Version: 32.0 (2.0)
Firmware Version: 123
User Agent: Mozilla/5.0 (Mobile; rv:32.0) Gecko/32.0 Firefox/32.0

Flame 2.0 Jelly Bean Base (319mb)

Environmental Variables:
Device: Flame 2.0
Build ID: 20140923000205
Gaia: 8d7f2ac85f3154bdb149d67e5c2f9b035f5e4105
Gecko: 6dd19beda1c2
Version: 32.0 (2.0)
Firmware Version: 123
User Agent: Mozilla/5.0 (Mobile; rv:32.0) Gecko/32.0 Firefox/32.0


Open_C 2.0

Environmental Variables:
Device: Open_C 2.0
BuildID: 20140923000205
Gaia: 8d7f2ac85f3154bdb149d67e5c2f9b035f5e4105
Gecko: 6dd19beda1c2
Version: 32.0 (2.0)
Firmware: P821A10v1.0.0B06_LOG_DL
User Agent: Mozilla/5.0 (Mobile; rv:32.0) Gecko/32.0 Firefox/32.0
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(dharris)
[Blocking Requested - why for this release]:

Nominating this as a blocker because the webpage will LMK after a short time, and is a regression
blocking-b2g: --- → 2.1?
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(dharris)
QA Contact: smiko
While working on the regression window, I noticed that this bug repros without hitting the power button in step 4 and without running the X-heavy script.
We are unable to get a regression window due to several blocking issues listed below:

Mozilla central Flame builds:(JB base) (these are prior to the first broken)
20140417000006-
20140611072409- 
20140701171354- 

Actual result: Site functions as expected. 
______________________________________________________________________________________________

Mozilla central Flame builds:(JB base)
20140706091720-
20140711132312

Actual result: An error message is displayed informing the user that the site cannot be loaded.
______________________________________________________________________________________________

Mozilla central Flame builds (JB base):
20140724040112
20140724171005

Actual result: The phone becomes unresponsive. The home and power button do not function. The only way to recover is to remove the battery and reboot.
________________________________________________________________________________________________

Mozilla central Flame builds:
20140806090025- JB base - ("Last Working" area) 

Actual result: The listed bug repros. The user is brought to the home screen.
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(jmitchell)
QA Contact: smiko
To elaborate on the above post - the regression window was blocked by several other issues and we were not going to be able to get a clear-cut "First Broken" build as there were several different forms of breakage prior to our "Last Working", sorry for any inconvenience.
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(jmitchell)
Kevin, Ben, any ideas?
Flags: needinfo?(kgrandon)
Flags: needinfo?(bfrancis)
We can't block on this if we can't identify the regressing bug. Its also possible that many small memory regression are causing this.

Milan, should the gfx team take this on?
blocking-b2g: 2.1? → backlog
Flags: needinfo?(milan)
Let's triage this without the regression range?  Does turning off tiling help, does it happen when the browser is the only app running, how much memory is browser using or trying to use, etc.
Flags: needinfo?(milan)
We're also running into some memory leaks in the system app which could be impacting this. See bug 1074703.
Flags: needinfo?(kgrandon)
I'm not sure what the X-heavy script is but I haven't been able to reproduce this issue without it.

I'm not sure what we could do to fix this in Gaia except find memory leaks where they are identified, as Gregor and Kevin point out.
Flags: needinfo?(bfrancis)
Adding qawanted to answer the questions in Comment 9.
Keywords: qawanted
QA Whiteboard: [QAnalyst-Triage+]
Whiteboard: [2.1-exploratory-2] → [2.1-exploratory-2] [systemsfe]
Attached file dmesg log
(In reply to Milan Sreckovic [:milan] from comment #9)
> Does turning off tiling help,

No. un-checking 'Tiling' in Settings > Developer still reproduces the bug. On one instance I even had a Firefox OS crash (device reboots itself) while trying to repro the bug without tiling.

> does it happen when the browser is the only app running

Yes.

> how much memory is browser using or trying to use, etc.

Attached a dmesg log. Ran dmesg several times during an attempt to reproduce the bug. Also comment 0 attached a logcat.

(In reply to Ben Francis [:benfrancis] from comment #11)
> I'm not sure what the X-heavy script is but I haven't been able to reproduce
> this issue without it.

As comment 3 mentioned, X-heavy script is NOT required in order to reproduce the bug.
I wasn't able to reproduce the bug in 512MB memory Flame, but I was able to repro the bug 7 out of 7 attempts with 319MB memory on latest Flame 2.2.

Tested on:
Device: Flame (shallow flash)
BuildID: 20141006092353
Gaia: 470826d13ae130a5c3d572d1029e595105485fb0
Gecko: 4534f97c4633
Version: 35.0a1 (2.2 Master)
Firmware: V180
User Agent: Mozilla/5.0 (Mobile; rv:35.0) Gecko/35.0 Firefox/35.0
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(jmitchell)
Keywords: qawanted
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(jmitchell)
Moving this bug as it has nothing to do with browser chrome. So what is the conclusion here? Is this a platform issue or a memory leak in the system app, or a combination of both?
Component: Gaia::System::Browser Chrome → Gaia::System
blocking-b2g: backlog → ---
Firefox OS is not being worked on
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: