Closed Bug 1038938 Opened 11 years ago Closed 8 years ago

[B2G] [Camera][273MB Flame] Long-pressing the Home button while self timer is counting down will cause various errors

Categories

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

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

(b2g-v1.4 unaffected, b2g-v2.0 affected, b2g-v2.1 affected)

RESOLVED WONTFIX
Tracking Status
b2g-v1.4 --- unaffected
b2g-v2.0 --- affected
b2g-v2.1 --- affected

People

(Reporter: ckreinbring, Unassigned)

References

()

Details

(Keywords: regression, Whiteboard: [273MB-Flame-Support] [2.0-exploratory])

Attachments

(1 file)

Description: Various issues occur when the user long taps the Home button while the self-timer is running in the Camera app. Repro Steps: 1) Update a Flame to 20140715000201 2) Launch the Camera app. 3) Tap the drawer icon, select Self-Timer, then set it to something other than Off. 4) Tap the drawer icon again then tap the camera icon to start the timer. 5) When the timer starts counting down, long-press the Home button. 6) Observe the device as the Camera app is taken into card view. Move the device around as well. Actual: There is usually a small delay in entering the card view. After a few seconds, beeping can be heard that usually indicates the timer is about to expire, and sometimes a shutter sound plays. In rare instances, moving the camera around will cause the image in the card to move around. Expected: The Camera app enters card view with no errors, there is no countdown beeping, and the image in the Camera card remains static. Environmental Variables: Device: Flame 2.0 (273 MB) Build ID: 20140715000201 Gaia: 2c6c413ed729d465c52d6c2d5d458e2eee79e956 Gecko: d32649a24965 Platform Version: 32.0a2 Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:32.0) Gecko/32.0 Firefox/32.0 Repro frequency: 40% See attached video clip and logcat logs
Bug repros on Flame 2.1 (273 MB) Build ID: 20140715040206 Gaia: 46cd188fdda2397d2b8f3303a184dcd52952e2b2 Gecko: e032c429908b Platform Version: 33.0a1 Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:33.0) Gecko/33.0 Firefox/33.0 Actual result: After a few seconds, beeping can be heard that usually indicates the timer is about to expire, and sometimes a shutter sound plays. In rare instances, moving the camera around will cause the image in the card to move around. -------------------------------------------------------------------------------------------------------- The bug does not repro on Flame 2.0 (512 MB), Flame 1.4, Buri 2.1, Buri 2.0 or Buri 1.4 Flame 2.0 (512 MB) Build ID: 20140715000201 Gaia: 2c6c413ed729d465c52d6c2d5d458e2eee79e956 Gecko: d32649a24965 Platform Version: 32.0a2 Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:32.0) Gecko/32.0 Firefox/32.0 Flame 1.4 (273 MB) Build ID: 20140715000202 Gaia: b7d36622c7df92c976c37520ccab25199c7ada91 Gecko: de7ecfb00955 Platform Version: 30.0 Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:30.0) Gecko/30.0 Firefox/30.0 Buri 2.1 (273 MB) Build ID: 20140715040206 Gaia: 46cd188fdda2397d2b8f3303a184dcd52952e2b2 Gecko: e032c429908b Platform Version: 33.0a1 Firmware Version: v1.2device.cfg User Agent: Mozilla/5.0 (Mobile; rv:33.0) Gecko/33.0 Firefox/33.0 Buri 2.0 (273 MB) Build ID: 20140715000201 Gaia: 2c6c413ed729d465c52d6c2d5d458e2eee79e956 Gecko: d32649a24965 Platform Version: 32.0a2 Firmware Version: v1.2device.cfg User Agent: Mozilla/5.0 (Mobile; rv:32.0) Gecko/32.0 Firefox/32.0 Buri 1.4 (273 MB) Build ID: 20140715000202 Gaia: b7d36622c7df92c976c37520ccab25199c7ada91 Gecko: de7ecfb00955 Platform Version: 30.0 Firmware Version: v1.2device.cfg User Agent: Mozilla/5.0 (Mobile; rv:30.0) Gecko/30.0 Firefox/30.0 Actual result: The countdown sound does not play unless the self timer is at the 4 or 5 second mark, and the card view always shows a static image.
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(ktucker)
Another thing that has been noticed is that tapping the Camera app card will sometimes have no response. This happens about 70% of the time.
I checked the issue in following build and the issue is not reproducible Hardware : Mako Platform Version : 33.0a1 Build Identifier: 20140715124800
This is a regression on the Flame but i feel like this issue is minor and not a blocker.
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(ktucker)
Keywords: regression
Please disregard Comment 4. After reproducing this issues a few times, the user completely lost the view finder on the camera. Nominating this 2.0? since this is a regression and can lead to more serious issues in camera.
blocking-b2g: --- → 2.0?
QA Whiteboard: [QAnalyst-Triage+]
blocking-b2g: 2.0? → 2.0+
Summary: [B2G] [Camera] Long-pressing the Home button while self timer is counting down will cause various errors → [B2G] [Camera][273MB Flame] Long-pressing the Home button while self timer is counting down will cause various errors
David, Please investigate. Thanks!
Assignee: nobody → dflanagan
QA Contact: ddixon
Does this reproduce on QRD?
Flags: needinfo?(ddixon)
We are not sure what you mean by "QRD". This bug DOES reproduce on our Flame device with a 2.1 build installed (memory: 273 MB). Comment 1 has our complete Branch Check.
Flags: needinfo?(ddixon)
Issue DOES occur on the earliest Flame 2.1 build we have access to. Repro rate: about 25% Environmental Variables: Device: Flame Master Build ID: 20140417000006 Gaia: 7591e9dc782ac2e97d63a96f9deb71c7b3588328 Gecko: e71ed4135461 Version: 31.0a1 (Master) Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:31.0) Gecko/31.0 Firefox/31.0
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(jmitchell)
Regression window unavailable - occurs on the oldest build we have access to
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(jmitchell)
This is not a camera bug, or at least not a bug that we can do anything about in the camera app. As best as I can tell, what is going on is that the System, Homescreen and Cameras apps are all running at the same time and are using more memory than the device has, so zmem swapping is required to keep them all alive. When I try to take a picture, parts of the camera app need to be swapped in, so taking a picture is slow. When I quick press on the home button, parts of the homescreen need to be swapped in, so returning to the homescreen is slow. When I long press on the home button, parts of the system app need to be swapped in, so displaying the task switcher is slow (really slow in this case). The task switcher shrinks the running window down to a small size and keeps it running while it takes a screenshot of it. Since the camera is displaying a live video feed we can see the image continue to move while the system app processes the screenshot. Because running a live video feed is memory-intensive, we probably have a situation where memory is thrashing between the camera and the system app. There are all sorts of bug here even without a countdown timer, but with the timer running it appears that the camera app does not receive the event telling it that it has been hidden for more than 10 seconds. So the countdown timer keeps running and we hear the beeps and even the shutter sound. In summary: 1) there is not a camera bug here. 2) In the 273mb configuration the camera is barely useable in any case, even ignoring this particular bug. So I'm not sure it makes sense to block on this one. Alive: could you take a look at this? In particular, note that on a 273mb Flame, it can take more than 10 seconds to bring up the task switcher from the camera app. The screenshot seems to be generated quickly, but it doesn't display for much longer, and meanwhile the camera viewfinder continues to display a live video feed. Is there some way you can modify the system app so that it sends the app to the background more quickly? I think this bug would be minimized if the camera knew to stop its video stream sooner. Or is there some other signal that the camera app could listen for to know when the task manager is being displayed? If so, then it could stop the preview stream in advance.
Alive: in addition to the comment above, I found one more thing to mention to you. While investigating this, I found a few times that I could get into a state where the home button would not work any more. I think what triggered it was long pressing on the home button to bring up the task switcher, and then before the task switcher appeared, I did the same thing again. Then somehow after that I got back to a working camera app with a broken home button. Do the hardware buttons still work via a state machine? Could we be getting into an unexpected state with multiple long presses and somehow breaking the home button? I tried to reproduce this and file a bug, but could not reproduce.
Flags: needinfo?(alive)
Component: Gaia::Camera → Gaia::System::Window Mgmt
If this is the wrong component, feel free to reassign. But I don't think there is anything we can do in the camera app to resolve these issues.
This is the right comp if the above assessment is correct. Unassign David accordingly.
Assignee: dflanagan → nobody
(In reply to David Flanagan [:djf] from comment #12) > Alive: in addition to the comment above, I found one more thing to mention > to you. While investigating this, I found a few times that I could get into > a state where the home button would not work any more. I think what > triggered it was long pressing on the home button to bring up the task > switcher, and then before the task switcher appeared, I did the same thing > again. Then somehow after that I got back to a working camera app with a > broken home button. Do the hardware buttons still work via a state machine? > Could we be getting into an unexpected state with multiple long presses and > somehow breaking the home button? I tried to reproduce this and file a bug, > but could not reproduce. Hardware button doesn't change for a long time. The hardware key event in browser API is still on going. If your issue is reproduce-able we could have someone take a look. I suspect it is related to bug 1037041.
Flags: needinfo?(alive)
Assignee: nobody → alive
I cannot reproduce it with master + flame. The camera app goes to background within 1 second when long press home button. "[Dump: AppWindow][Camera][AppWindow_2][233.726]close with to-cardview" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.726]opened,closing,::,close" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.728]timer to ensure closed does occur." app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.730] publishing internal event: closing" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.732]handling _closing" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.734] publishing external event: closingundefined" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.738] publishing internal event: willclose" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.741] publishing external event: willcloseundefined" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.943]closeAppTowardsCardView has been ENDED!" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.944]closing,closed,::,complete" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.946] publishing internal event: closed" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.947]set visibility -> ,false" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.949]screenshot state -> ,screenshot" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.950] publishing external event: closedundefined" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.951] publishing internal event: close" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.951] publishing external event: closeundefined" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][233.968] has been ENDED!" app_window.js:981 "[AppUsage]" "app://camera.gaiamobile.org/manifest.webapp" "ran for" 10 app_usage_metrics.js:92 "[Dump: AppWindow][Camera][AppWindow_2][234.321]getScreenshot succeed!" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][234.322]setVisible on browser element:false" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][234.333] Handling mozbrowservisibilitychange event..." app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][234.334] publishing internal event: background" app_window.js:981 "[Dump: AppWindow][Camera][AppWindow_2][234.334] publishing external event: backgroundundefined" And there's no 2.0 repro-able info in this bug by comment 1. Why is this blocking 2.0 release?
blocking-b2g: 2.0+ → 2.0?
(In reply to Alive Kuo [:alive][NEEDINFO!] from comment #16) > I cannot reproduce it with master + flame. The camera app goes to background > within 1 second when long press home button. > "[Dump: AppWindow][Camera][AppWindow_2][233.726]close with to-cardview" > app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.726]opened,closing,::,close" > app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.728]timer to ensure closed does > occur." app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.730] publishing internal event: > closing" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.732]handling _closing" > app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.734] publishing external event: > closingundefined" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.738] publishing internal event: > willclose" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.741] publishing external event: > willcloseundefined" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.943]closeAppTowardsCardView has > been ENDED!" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.944]closing,closed,::,complete" > app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.946] publishing internal event: > closed" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.947]set visibility -> ,false" > app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.949]screenshot state -> > ,screenshot" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.950] publishing external event: > closedundefined" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.951] publishing internal event: > close" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.951] publishing external event: > closeundefined" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][233.968] has been ENDED!" > app_window.js:981 > "[AppUsage]" "app://camera.gaiamobile.org/manifest.webapp" "ran for" 10 > app_usage_metrics.js:92 > "[Dump: AppWindow][Camera][AppWindow_2][234.321]getScreenshot succeed!" > app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][234.322]setVisible on browser > element:false" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][234.333] Handling > mozbrowservisibilitychange event..." app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][234.334] publishing internal event: > background" app_window.js:981 > "[Dump: AppWindow][Camera][AppWindow_2][234.334] publishing external event: > backgroundundefined" > > > And there's no 2.0 repro-able info in this bug by comment 1. Why is this > blocking 2.0 release? comment 0 indicates this reproduces on 2.0 40% of the time, but only with 273 MB Flame.
qawanted, see if the problem still exists after several camera 273mb fix.
Keywords: qawanted
Assignee: alive → nobody
Qa-wanted: Check repro in the latest 2.0 and 2.1
QA Whiteboard: [QAnalyst-Triage+]
Issue DOES reproduce on latest Flame 2.0 and 2.1 builds. (273 MB) Actual Results: Long pressing home button as Self-Timer hits 5 seconds causes various problems such as: countdown chime continues after app is put in tile view, camera may take photo when in tile mode, user at times cannot re-enter the app once in tile mode, and phone becomes very sluggish and unresponsive. 2.0 Device: Flame 2.0 Build ID: 20140721082721 Gaia: b9d19011123487009c80d1200937652d58c434a0 Gecko: d69cd84b6824 Version: 32.0a2 (2.0) Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:32.0) Gecko/32.0 Firefox/32.0 2.1 Device: Flame Master Build ID: 20140723040045 Gaia: 01d86c8cc615658694b114ca5711763efc4ef205 Gecko: d0f6259e8446 Version: 34.0a1 (Master) Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0
QA Whiteboard: [QAnalyst-Triage?]
Flags: needinfo?(jmitchell)
Keywords: qawanted
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(jmitchell)
QA Wanted - we need to retest this on 319 MB Flame.
Keywords: qawanted
Issue DOES occur but is less severe and doesn't freeze the phone in 2.0 and 2.1 Flame builds with 319 MB memory. Actual Results: Long pressing home button as Self-Timer hits 5 seconds causes the countdown chime to continue after camera app is put in tile view. Environmental Variables: Device: Flame 2.0 Build ID: 20140721082721 Gaia: b9d19011123487009c80d1200937652d58c434a0 Gecko: d69cd84b6824 Version: 32.0a2 (2.0) Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:32.0) Gecko/32.0 Firefox/32.0 Environmental Variables: Device: Flame Master Build ID: 20140724063605 Gaia: c72257b2d27135bfcd68e89dd584182797784016 Gecko: 616e6924cb0b Version: 34.0a1 (Master) Firmware Version: v122 User Agent: Mozilla/5.0 (Mobile; rv:34.0) Gecko/34.0 Firefox/34.0
QA Whiteboard: [QAnalyst-Triage+] → [QAnalyst-Triage?]
Flags: needinfo?(jmitchell)
Keywords: qawanted
QA Whiteboard: [QAnalyst-Triage?] → [QAnalyst-Triage+]
Flags: needinfo?(jmitchell)
QA Whiteboard: [QAnalyst-Triage+] → [QAnalyst-Triage+][lead-review+]
Triage: this is corner case and does not freeze in 319 MB configuration.
blocking-b2g: 2.0? → ---
Firefox OS is not being worked on
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: