1130035 - [System] Device displays a white screen and becomes unresponsive while trying to load a webpage.

Status: RESOLVED WONTFIX

(Core :: Hardware Abstraction Layer (HAL), defect)

Description

[Yeojin Chung [:YeojinC]]

Description:
The same issue as bug 1122119 occurrs with 100% repro rate when going to a specific webpage. This webpage contains a heavy load of pictures, and the white screen occurs while the page is loading.

Repro Steps:
1) Update a Flame to 20150205010209.
2) Connect to a Wi-Fi network or turn on the data connection.
3) Go to http://goo.gl/sQPu8j on the browser.

Actual:
A white screen appears, and the device becomes unresponsive. When connected to a computer, it tries to open device folders indefinitely.

Expected:
The page loads properly.

Environmental Variables:
Device: Flame 3.0
Build ID: 20150205010209
Gaia: 2b83a6d5d1185a438b5bbd287497ac2743b501db
Gecko: 34a66aaaca81
Gonk: e7c90613521145db090dd24147afd5ceb5703190
Version: 38.0a1 (3.0)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:38.0) Gecko/38.0 Firefox/38.0

Device: Flame 2.2 (319mb, full flash)
Build ID: 20150205002503
Gaia: c2047a46e29696238e9b4c9caaba47736421449a
Gecko: adfba0a07e9b
Gonk: e7c90613521145db090dd24147afd5ceb5703190
Version: 37.0a2 (2.2)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:37.0) Gecko/37.0 Firefox/37.0

Repro frequency: 5/5
See attached: logcat, video clip
http://youtu.be/uWfr4IC9RbY

Comment 1

[Yeojin Chung [:YeojinC]]

This issue does NOT reproduce on Flame 2.1.

Result: While trying to load the page, the browser app crashes, but the white screen does not appears.

Device: Flame 2.1 (319mb, full flash)
Build ID: 20150205001711
Gaia: 17bf14f12e43043654498330d610d469b8b55e64
Gecko: bdebcc47ec7a
Gonk: e7c90613521145db090dd24147afd5ceb5703190
Version: 34.0 (2.1)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0

Comment 2

[Yeojin Chung [:YeojinC]]

Attachment: logcat_whitescreen

Comment 3

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

Nominating 2.2? the user should not encounter a white screen.

Let's get a regression window.

Comment 4

[Gabriele Svelto [:gsvelto]]

(In reply to Yeojin Chung [:YeojinC] from comment #0)
> Environmental Variables:
> Device: Flame 3.0
> Build ID: 20150205010209
> Gaia: 2b83a6d5d1185a438b5bbd287497ac2743b501db
> Gecko: 34a66aaaca81
> Gonk: e7c90613521145db090dd24147afd5ceb5703190
> Version: 38.0a1 (3.0)
> Firmware Version: v18D-1
> User Agent: Mozilla/5.0 (Mobile; rv:38.0) Gecko/38.0 Firefox/38.0

I've just tested this configuration but I can't seem to reproduce the problem on my Flame.

Comment 5

[Lance（leave from Mozilla）]

I have verified this bug on latest Flame 2.1/2.2 and oldest Flame 2.1/2.2 we can find, but the bug 

still exist in these versions.

Steps:
1. Connect to a Wi-Fi network or turn on the data connection.
2. Go to http://goo.gl/sQPu8j on the browser.

Actual:
A white screen appears, and the device becomes unresponsive. When connected to a computer, it tries to 

open device folders indefinitely.

Expected:
The page loads properly.

Fail rate:5/5

Flame 2.2:
Build ID               20140923160207
Gaia Revision          ff6dbb006e4e60edba697639aa2a2a199b07069b
Gaia Date              2014-09-23 18:22:24
Gecko Revision         https://hg.mozilla.org/mozilla-central/rev/30920bcb070a
Gecko Version          35.0a1
Device Name            flame
Firmware(Release)      4.4.2
Firmware(Incremental)  eng.cltbld.20140923.192209
Firmware Date          Tue Sep 23 19:22:19 EDT 2014
Bootloader             L1TC000118D0

Build ID               20150205002503
Gaia Revision          c2047a46e29696238e9b4c9caaba47736421449a
Gaia Date              2015-02-04 20:34:04
Gecko Revision         https://hg.mozilla.org/releases/mozilla-b2g37_v2_2/rev/adfba0a07e9b
Gecko Version          37.0a2
Device Name            flame
Firmware(Release)      4.4.2
Firmware(Incremental)  eng.cltbld.20150205.045043
Firmware Date          Thu Feb  5 04:50:54 EST 2015
Bootloader             L1TC000118D0

Flame 2.1:
Build ID               20141008161201
Gaia Revision          7ef2e1e59637a34ca4489c329b3bdee93df3ac6c
Gaia Date              2014-10-08 12:32:21
Gecko Revision         https://hg.mozilla.org/releases/mozilla-b2g34_v2_1/rev/e3d495eb85c6
Gecko Version          34.0a2
Device Name            flame
Firmware(Release)      4.4.2
Firmware(Incremental)  eng.cltbld.20141008.195811
Firmware Date          Wed Oct  8 19:58:25 EDT 2014
Bootloader             L1TC000118D0

Build ID               20150205001711
Gaia Revision          17bf14f12e43043654498330d610d469b8b55e64
Gaia Date              2015-02-03 05:19:41
Gecko Revision         https://hg.mozilla.org/releases/mozilla-b2g34_v2_1/rev/bdebcc47ec7a
Gecko Version          34.0
Device Name            flame
Firmware(Release)      4.4.2
Firmware(Incremental)  eng.cltbld.20150205.035050
Firmware Date          Thu Feb  5 03:51:01 EST 2015
Bootloader             L1TC000118D0

Comment 6

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

Repro rate for this bug is not 100%, especially on 2.1 and 2.0. On these two branches it could take up to 11 attempts before the issue occurs, but I've seen it happened on first try after flashing as well.

This issue reproduces on 2.1, 2.0, base image v18D-1 only, and base image v188-1 only. Removing regression keyword.

Comment 7

[bhavana bajaj [:bajaj]]

wesly looks like this could be related to base image as it reproduces with 18D-1, could you have a look at this?

Comment 8

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

triage: 
per comment 5, removing the regression keyword.
the experience is too bad so let's block on it. if it's base image problem, please change component to vendcom.

Comment 9

[Wesly Huang (EPM)]

Hi Youlong:

per comment#6 this can be repro. in v18D base image, would you help investigate first? Thanks!

Comment 10

[youlong.jiang]

(In reply to Wesly Huang from comment #9)
> Hi Youlong:
> 
> per comment#6 this can be repro. in v18D base image, would you help
> investigate first? Thanks!

hi wesly -

we'll try to reproduce this issue on our side and feedback.

tks.

Comment 11

[bhavana bajaj [:bajaj]]

wesly, can you please help follow-up on this?

Comment 12

[Wesly Huang (EPM)]

Yes this is(In reply to bhavana bajaj [:bajaj] from comment #11)
> wesly, can you please help follow-up on this?

Yes I still have weekly issue review meeting with T2M that covering this one, the thing is their responsible engineer take some vacation before and after Chinese New Year (a big holiday in China) so we expect a delayed feedback.

@Youlong: as discussed last time please still try to arrange substitute team member to look into this soon, then keep us updated, thank you!

Comment 13

[jingming]

test many times(20+) on both v18D-1 and v188-1. but still can't reproduce now.

Comment 14

[howie [:howie]]

Hi  Yeojin, can you verify Comment 13?

Comment 15

[Yeojin Chung [:YeojinC]]

(In reply to howie [:howie] from comment #14)
> Hi  Yeojin, can you verify Comment 13?

I was able to reproduce this issue on the first try on today's nightly Flame Master. Here's my environmental variables:

Device: Flame Master (KK, 319mb, full flash)
Build ID: 20150305010212
Gaia: eff3321ab4e65da3f906688ebb55ddf1e93d9452
Gecko: 56492f7244a9
Gonk: e7c90613521145db090dd24147afd5ceb5703190
Version: 39.0a1 (3.0)
Firmware Version: v18D-1
User Agent: Mozilla/5.0 (Mobile; rv:39.0) Gecko/39.0 Firefox/39.0

Comment 16

[bhavana bajaj [:bajaj]]

(In reply to Yeojin Chung [:YeojinC] from comment #15)
> (In reply to howie [:howie] from comment #14)
> > Hi  Yeojin, can you verify Comment 13?
> 
> I was able to reproduce this issue on the first try on today's nightly Flame
> Master. Here's my environmental variables:
> 
> Device: Flame Master (KK, 319mb, full flash)
> Build ID: 20150305010212
> Gaia: eff3321ab4e65da3f906688ebb55ddf1e93d9452
> Gecko: 56492f7244a9
> Gonk: e7c90613521145db090dd24147afd5ceb5703190
> Version: 39.0a1 (3.0)
> Firmware Version: v18D-1
> User Agent: Mozilla/5.0 (Mobile; rv:39.0) Gecko/39.0 Firefox/39.0

NI :wesly to make sure T2M is using the exact build listed here and then trying this issue.

Comment 17

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

FW NI to Wesly

Comment 18

[Wesly Huang (EPM)]

> > I was able to reproduce this issue on the first try on today's nightly Flame
> > Master. Here's my environmental variables:
> > 
> > Device: Flame Master (KK, 319mb, full flash)
> > Build ID: 20150305010212
> > Gaia: eff3321ab4e65da3f906688ebb55ddf1e93d9452
> > Gecko: 56492f7244a9
> > Gonk: e7c90613521145db090dd24147afd5ceb5703190
> > Version: 39.0a1 (3.0)
> > Firmware Version: v18D-1
> > User Agent: Mozilla/5.0 (Mobile; rv:39.0) Gecko/39.0 Firefox/39.0
> 
> NI :wesly to make sure T2M is using the exact build listed here and then
> trying this issue.

Bhavana: 

this is not the build T2M have and made, so will be good if we can repro. & get log with base image to eliminate other factors introduced by 2.1/2.2/3.0 Gaia/Gecko.

@Youlong:

As discussed in phone please still try to take a look at the log provided, and let us know any finding you see and reply here, thank you.

@Pi Wei:

I saw you've repro. the issue in T2M's based image per comment#6, would you help me understand the repro. rate you see? Also share a log based on it so let T2M to check further.

Comment 19

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

Adding qawanted to get a logcat and repro rate on v18D base.

Comment 20

[Yeojin Chung [:YeojinC]]

This issue does reproduce on v18D base image only. The white screen appeared when the browser tried to load the website.

Comment 21

[Mike Lien[:mlien]]

I verify with base image v18D only for 10 times but always cannot reproduce it even I adjust memory to 319MB
If any log or video can be referenced here?

Comment 22

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

Yeojin, can you provide a logcat and video for this issue reproducing on the base image only?

Comment 23

[Yeojin Chung [:YeojinC]]

I cannot reproduce the white screen issue anymore on v18D base image any more. Instead of the white screen, the error message shows up on the browser: "Well, this is embarrassing.We tried to display this Web page, but it's not responding."

Then sometimes the device becomes unresponsive. When the user presses the home button, nothing happens. The device has to be restarted to get out of the unresponsive state.

Comment 24

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

Yeojin, can you please try again today?

Comment 25

[Yeojin Chung [:YeojinC]]

Attachment: logcat_20150316_WhiteScreen.txt

I was able to reproduce this bug on v18D base only, but the repro rate is below 10%. Before I reproduced it, I tried to load the page about 12 times.

I was unable to get the video, but here's the logcat.

Comment 26

[Mike Lien[:mlien]]

I'm a little confused what is the problem now. White screen? Or Browser show error message: "Well, this is embarrassing.We tried to display this Web page, but it's not responding."?

If current problem is the later one, I think we should file another bug and close this one since it's indeed cannot reproduce the "white screen" issue.

Comment 27

[Ken Chang[:kenkai]]

Peter, According to comment 25, could you please have someone to check the log?

Comment 28

[Yeojin Chung [:YeojinC]]

(In reply to Mike Lien[:mlien] from comment #26)
> I'm a little confused what is the problem now. White screen? Or Browser show
> error message: "Well, this is embarrassing.We tried to display this Web
> page, but it's not responding."?
> 
> If current problem is the later one, I think we should file another bug and
> close this one since it's indeed cannot reproduce the "white screen" issue.


I DID reproduce the white screen issue, although with a very low repro rate. So I'm not sure if we can close this issue. If you'd like, I can open up a new bug addressing the other issue, but I guessed the cause of those two symptoms is the same. Please let me know. :)

Comment 29

[Peter Chang[:pchang]]

(In reply to Ken Chang[:ken](OOO from 2/18 to 3/1) from comment #27)
> Peter, According to comment 25, could you please have someone to check the
> log?
Found some ipc and memory pressure errors happened and looked like the system got memory pressure.

03-16 13:30:24.659   204   689 I GonkMemoryPressure: Checking to see if memory pressure is over.
03-16 13:30:24.669   204   689 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:24.669  1290  1295 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:24.679  1461  1466 I GonkMemoryPressure: Checking to see if memory pressure is over.
03-16 13:30:24.679  1461  1466 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:29.779   204   689 I GonkMemoryPressure: Checking to see if memory pressure is over.
03-16 13:30:29.779   204   689 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:29.779  1290  1295 I GonkMemoryPressure: Checking to see if memory pressure is over.
03-16 13:30:29.779  1290  1295 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:29.789  1461  1466 I GonkMemoryPressure: Checking to see if memory pressure is over.
03-16 13:30:29.789  1461  1466 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:37.999   204   689 I GonkMemoryPressure: Checking to see if memory pressure is over.
03-16 13:30:37.999  1290  1295 I GonkMemoryPressure: Checking to see if memory pressure is over.
03-16 13:30:38.009   204   689 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:38.009  1290  1295 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:38.009  1461  1466 I GonkMemoryPressure: Checking to see if memory pressure is over.
03-16 13:30:38.019  1461  1466 I GonkMemoryPressure: Memory pressure is over.
03-16 13:30:38.579   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.589   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.589   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.589   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.589   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.589   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.599   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.599   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.599   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv
03-16 13:30:38.599   204   884 I Gecko   : ###!!! [Child][MessageChannel] Error: Channel error: cannot send/recv

Comment 30

[Gabriele Svelto [:gsvelto]]

The memory pressure messages are harmless. They are caused by the memory pressure watcher which is used to detect when we're low on memory and send memory pressure events to all applications. Applications will respond to these by trying to reduce their memory consumption. The following messages about IPC errors can be caused by an app having died due to the lack of memory.

Comment 31

[howie [:howie]]

Triage: based on Comment 28 and 30, remove nom.

Comment 32

[youlong.jiang]

Dears -

we've checked logcat you provided yet, and couldn't find valued info.

from my view, if more stable in base image and hard to reproduce, how about you roll back to check patches in nightly build that may cause this problem.

tks.

Comment 33

[No-Jun Park [:njpark]]

[Tracking Requested - why for this release]:

This still reproduces rather easily on v18D-3 on 319MB config flame, but i think it should be made clear that this page contains MASSIVE amount of image files stitched together.  I can't repro this on aries, which renders the page quite fast compared to my android phone (Oneplus). (it takes about slightly less than a minute to completely load the site)  I'm strongly suspecting that flame device with 319MB can't handle this load reliably, as it even often LMKs on some CNN websites. 

Although there is no way to reset the device other than pulling the battery out once whitescreen happens (which is pretty serious- the sorry message would've been much nicer), I'm moving it to backlog for now until we find the resource to take a look at issue, || similar bug happens in devices with higher memory config.

Comment 34

[Sylvestre Ledru [:Sylvestre]]

Closing bugs with b2g-master=affected as it is likely to be out dated.