868806 - [Buri][Video]The video can not played when the SD card have a HD-Video.

Status: RESOLVED FIXED

(Firefox OS Graveyard :: General, defect, P1)

Description

[sync-1]

AU_LINUX_GECKO_ICS_STRAWBERRY_V1.01.00.01.019.092
 Firefox os  v1.0.1
 Mozilla build ID:20130430070201
 
 +++ This bug was initially created as a clone of Bug #449489 +++
 
 Created an attachment (id=402945)
 Special video
 
 DEFECT DESCRIPTION:
 The video can not played when the SD card have a Special HD-Video.
 
  REPRODUCING PROCEDURES:
 1.SD card have a Special HD-Videos;
 2.Launch video ,can read the video list;
 3.Choose a video to play ,screen black and can not play the video ,if you deleted the special HD-video, others videos can be played normally.->KO
 
  EXPECTED BEHAVIOUR:
 Can be play video normally whatever have the special video.
 
  ASSOCIATE SPECIFICATION:
 
  TEST PLAN REFERENCE:
 
  TOOLS AND PLATFORMS USED:
 
  USER IMPACT:
 
  REPRODUCING RATE:
 5/5
 
  For FT PR, Please list reference mobile's behavior:
 
 ++++++++++ end of initial bug #449489 description ++++++++++

Comment 1

[sync-1]

Clone from brother

Comment 2

[sync-1]

Attachment: Special video

Clone from brother

Comment 3

[Sotaro Ikeda [:sotaro]]

attachment 745627 [details] seems like an abnormal content. Actual playable content is shorter than content duration in mp4 container.

Comment 4

[Sotaro Ikeda [:sotaro]]

During thumbnail creation, nsBuiltinDecoderReader::DecodeToTarget() falled into infinite loop. During thumbnail creation by video app, video app try to create a thumbnail of 1/4 duration time. But the video does not have data to decode.

https://github.com/sotaroikeda/gaia/blob/master/apps/video/js/video.js#L310

Comment 5

[Sotaro Ikeda [:sotaro]]

read() to audio codec return "-1004", but it is not handled in OmxDecoder::ReadAudio() and the function return true even in the error code.

Comment 6

[Sotaro Ikeda [:sotaro]]

Attachment: patch - handle OMXCodec's error

By applying the patch, the video is handled as normal. No infinite loop.

Comment 7

[Sotaro Ikeda [:sotaro]]

Without the patch, video an audio app could come into infinite loop by some abnormal contents and freeze untill the apps are killed.

Comment 8

[Sotaro Ikeda [:sotaro]]

Attachment: patch v2 - handle OMXCodec's error

committable patch. Carry "chris.double: review+".

Comment 9

[Sotaro Ikeda [:sotaro]]

https://tbpl.mozilla.org/?tree=Try&rev=8e4b158acc25

Comment 10

[Sotaro Ikeda [:sotaro]]

I am going to confirm if attachment 746988 [details] [diff] [review] regress youtube video playback. I am going to keep how to handle codec timeout error as before for not to regress youtube video playback.

Comment 11

[Sotaro Ikeda [:sotaro]]

Attachment: patch v3 - handle OMXCodec's error

change from v2: handle -ETIMEDOUT (timeout error) same as current gecko for not to regress youtube video playback.

Comment 12

[Sotaro Ikeda [:sotaro]]

Comment on attachment 749616 [details] [diff] [review]
patch v3 - handle OMXCodec's error

doublec, can you review the patch again?

Comment 13

[Sotaro Ikeda [:sotaro]]

https://tbpl.mozilla.org/?tree=Try&rev=8dfc19154435

Comment 14

[Sotaro Ikeda [:sotaro]]

Attachment: patch v4 - handle OMXCodec's error

Comittable patch. Carry "chris.double: review+".

Comment 15

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/projects/birch/rev/ed2c321c9bf6

Comment 16

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/ed2c321c9bf6

Comment 17

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-b2g18/rev/75ebaadf2078
https://hg.mozilla.org/releases/mozilla-b2g18_v1_0_1/rev/2e3deabbcb38

Comment 18

[Sarah Parsons]

Issue reproduces on Unagi 1.1 Build ID: 20130531070205

Environmental  Variables:
Kernel Date: Dec 5
Gecko: http://hg.mozilla.org/releases/mozilla-b2g18/rev/4f318822e72c
Gaia: e1c59baed29e4665d1da9392f86239272073f07a

The issue did not reproduce on Unagi 1.0.1 Build ID: 20130531070208

The bug was checked on the Inari Build ID: 20130528231041 the video appeared but being able to actually play it was inconsistent. Sometimes it would start 5+ seconds into the video run for 3 seconds then close. Sometimes playing the video would only appear as a gray screen. 

Issue could not be checked on the Leo due to bug#874615

Comment 19

[Sotaro Ikeda [:sotaro]]

(In reply to Sarah Parsons from comment #18)
> Issue reproduces on Unagi 1.1 Build ID: 20130531070205

I checked it on v1.1 unagi. It seems different bug.

Comment 20

[Sotaro Ikeda [:sotaro]]

OMXCodec destruct stopped in OMXCodec::stop() at following code.

    while (isIntermediateState(mState)) {
        mAsyncCompletion.wait(mLock);
    }

Comment 21

[Sotaro Ikeda [:sotaro]]

Attachment: stack trace of the infinite loop

stack trace of the problem.

Comment 22

[Sotaro Ikeda [:sotaro]]

I am going to create another bug for it.

Comment 23

[Sotaro Ikeda [:sotaro]]

(In reply to Sarah Parsons from comment #18)
> Issue could not be checked on the Leo due to bug#874615

On leo device, it works as normal.

Comment 24

[Sotaro Ikeda [:sotaro]]

(In reply to Sotaro Ikeda [:sotaro] from comment #22)
> I am going to create another bug for it.

Created Bug 878981.

Comment 25

[Sotaro Ikeda [:sotaro]]

(In reply to Sarah Parsons from comment #18)
> Issue reproduces on Unagi 1.1 Build ID: 20130531070205
> 
> The bug was checked on the Inari Build ID: 20130528231041 the video appeared
> but being able to actually play it was inconsistent. Sometimes it would
> start 5+ seconds into the video run for 3 seconds then close. Sometimes
> playing the video would only appear as a gray screen. 

The problem is in CAF's OMXCodec. Unagi and inari uses ics_chocolate as base platform. It is older qcom's code. New one is ics_strawberry, it is used by hamachi(buri) and leo. As in Bug 878981 comment #7 , ics_chocolate is not supported in v1.1. Need to confirm on hamachi(buri) or leo.