965415 - crash in application.zip@0x1a1ffe

Status: RESOLVED DUPLICATE of bug 948326

(Core :: Audio/Video, defect)

Description

[pfield]

This bug was filed from the Socorro interface and is 
report bp-23795fb0-6fdd-4ee5-b8c8-b5a922140129.
=============================================================

Description: 
A crash can occur if the user attempts to skip through a very long song on SoundCloud.com.

Repro Steps:
1) Updated Buri to Build ID: 20140127004002
2) tap to open Browser app.
3) Use the search bar to navigate to Soundcloud.com
4) Search for a long song that has a run time of over 1 hour.
5) Play the song and attempt to skip around to various points in the song.

Actual:
A crash may occur.

Expected:
User is able to skip through a song without a crash occurring.

Environmental Variables
Device: Buri v 1.3.0 Mozilla
Build ID: 20140127004002
Gecko: http://hg.mozilla.org/releases/mozilla-aurora/rev/c40099a42c1f
Gaia: 25a45a836a4a21a30f63fa7b544b42e8b781180a
Platform Version: 28.0a2
Firmware Version: V1.2-device.cfg

Comment 1

[pfield]

Attachment: crash-log.txt

Adding logcat

Comment 2

[Jason Smith [:jsmith]]

Can you give an example of a song that runs over an hour?

Comment 3

[pfield]

Yeah certainly Jsmith! A song that I used to get this issue is "Bearly Legal" by -The White Panda.-

Comment 4

[pfield]

I should clarify, I used this song in a playlist that connected with other songs and made for a very long "set of songs" but you can shift between all of them in the same time bar. But, you can also use a song that is over 1 hour and that will also trigger the issue, such as "https://soundcloud.com/jimanoli1/power-hour-classics" which is one long song and not many in a stream.

Comment 5

[Jason Smith [:jsmith]]

QA Wanted - Can we check what happens on 1.1 & 1.2?

Comment 6

[Peter Bylenga [:PBylenga]]

This issue does not reproduce on v1.2 or v1.1, user is able to seek around a long song without encountering a crash.  Tested the 'power-hour-classics' link in Comment 4.

v1.2 Environmental Variables:
Device: Buri v1.2 MOZ
BuildID: 20140130004002
Gaia: 539a25e1887b902b8b25038c547048e691bd97f6
Gecko: f9f469d5d1e1
Version: 26.0
Firmware Version: v1.2-device.cfg

v1.1 Environmental Variables:
Device: Buri v1.1 MOZ
BuildID: 20140130041201
Gaia: c434fe9a0e823029796805e141cfa983cda2d246
Gecko: 1b24a4178317
Version: 18.0
Firmware Version: v1.2-device.cfg

Comment 7

[Peter Bylenga [:PBylenga]]

Upon further investigation this issue is not 100% reproducible.  It's closer to 10% but once it reproduces it'll reproduce often and can happen quickly.  Other times I've repeated the STRs for 10 minutes until I saw the crash. Restarting the device doesn't permanently fix it either, eventually it'll reproduce again.  I was able to reproduce it as early as 1-20-2014.  At this point in time it's unrealistic to grab a regression window that I'm confident upon.

Comment 8

[Jason Smith [:jsmith]]

(In reply to Peter Bylenga from comment #7)
> Upon further investigation this issue is not 100% reproducible.  It's closer
> to 10% but once it reproduces it'll reproduce often and can happen quickly. 
> Other times I've repeated the STRs for 10 minutes until I saw the crash.
> Restarting the device doesn't permanently fix it either, eventually it'll
> reproduce again.  I was able to reproduce it as early as 1-20-2014.  At this
> point in time it's unrealistic to grab a regression window that I'm
> confident upon.

Does this reproduce with shorter songs on soundcloud? Like 3 - 5 minutes in length.

Comment 9

[Peter Bylenga [:PBylenga]]

I've tried multiple songs within that length on the Build ID: 20140127004002 stated in Comment 0.  Have not encountered the crash, also discussed with the reporter offline and he had went through a lot of regular songs on that day without encountering the issue as well.

Comment 10

[Jason Smith [:jsmith]]

Given the low reproduction rate & low likelihood of hitting this crash, I'm going to pull the nom. However, I'd like to get technical input on whether this can happen elsewhere & how likely, as this could be possible to happen elsewhere.

Sotaro - Can you weigh in what's not working here & indicate how likely this could happen in the web?

Comment 11

[Sotaro Ikeda [:sotaro]]

From the debugging, I confirmed that the crash happens when "mSamplesPerSecond == 0". It is a dup of Bug 918135. The bug is committed only to master, not to v1.3.

Comment 12

[Sotaro Ikeda [:sotaro]]

(In reply to Sotaro Ikeda [:sotaro] from comment #11)
> From the debugging, I confirmed that the crash happens when
> "mSamplesPerSecond == 0". It is a dup of Bug 918135. The bug is committed
> only to master, not to v1.3.

Sorry, the above was incorrect. The problem was already fixed by Bug 922334. But it is degraded by Bug 918135.

Comment 13

[Sotaro Ikeda [:sotaro]]

(In reply to Sotaro Ikeda [:sotaro] from comment #12)
> (In reply to Sotaro Ikeda [:sotaro] from comment #11)
> > From the debugging, I confirmed that the crash happens when
> > "mSamplesPerSecond == 0". It is a dup of Bug 918135. The bug is committed
> > only to master, not to v1.3.
> 
> Sorry, the above was incorrect. The problem was already fixed by Bug 922334.
> But it is degraded by Bug 918135.

After that the regression is fixed by Bug 948326. But it is only on master, not v1.3.

Comment 14

[Jason Smith [:jsmith]]

Just to clarify - so bug 948326 is what we need here, right?

Comment 15

[Sotaro Ikeda [:sotaro]]

(In reply to Jason Smith [:jsmith] from comment #14)
> Just to clarify - so bug 948326 is what we need here, right?

Yes. Need to uplift to v1.3.