1038165 - Music App paused/blocked when wired headset is inserted during audio playback through BT headset

Status: RESOLVED FIXED

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

Description

[Pooja]

Steps to Reproduce:
1.	Bootup the target
2.	Pair and connect a Bluetooth headset
3.	Start playing a music clip
4.	Insert a wired headset

Issue:	A message "Music Paused during the call" pops up and Music app gets hanged. Even when there is no call on device

Comment 1

[bhavana bajaj [:bajaj]]

Adding qawanted to see if we can reproduce the issue and help find the cause here.

Comment 2

[bhavana bajaj [:bajaj]]

NI Alison to get urgent testing.

Comment 3

[Cody Roesch [:croesch]]

This bug does NOT repro on: Flame 2.1, Flame 2.0, Flame 1.4, Buri 2.1

Actual Result: Playing music over bluetooth attached to a Airband bluetooth headset then inserting a separate wired headset into the phone only produces a loudness warning message.

Environmental Variables:
Device: Flame Master
Build ID: 20140714061512
Gaia: 88e0a972280bb35847c010b8c3f1481fa80f3847
Gecko: 340b19c14d3d
Version: 33.0a1 (Master)
Firmware Version: v122
-----------------------------------------------
Environmental Variables:
Device: Flame 2.0
Build ID: 20140714071406
Gaia: 726ca25ef1dbb83d8ed8a902b46fb340a3f95927
Gecko: 002aee8237a3
Version: 32.0a2 (2.0)
Firmware Version: v122
-----------------------------------------------
Environmental Variables:
Device: Flame 1.4
Build ID: 20140711225112
Gaia: b7d36622c7df92c976c37520ccab25199c7ada91
Gecko: dbebcdab47aa
Version: 30.0 (1.4)
Firmware Version: v122
-----------------------------------------------
Environmental Variables:
Device: Buri Master
Build ID: 20140714061512
Gaia: 88e0a972280bb35847c010b8c3f1481fa80f3847
Gecko: 340b19c14d3d
Version: 33.0a1 (Master)
Firmware Version: v1.2device.cfg

Comment 4

[Hema Koka [:hema]]

Pooja,

Please provide more information on the DUT. Comment #3 says we cannot reproduce on Flame/Buri

Thanks
Hema

Comment 5

[Alison Shiue]

Attachment: music_issue.log

Test on 
Gaia      2c6c413ed729d465c52d6c2d5d458e2eee79e956
Gecko     https://hg.mozilla.org/releases/mozilla-aurora/rev/d32649a24965
BuildID   20140714160201
Version   32.0a2
Firmware Version: v122

This issue does not happen when I first tried this case, and the actual result just as what Cody mentioned; But after I reboot the device and tested again, this issue happened.

Comment 6

[Hema Koka [:hema]]

(In reply to ashiue from comment #5)
> Created attachment 8455860 [details]
> music_issue.log
> 
> Test on 
> Gaia      2c6c413ed729d465c52d6c2d5d458e2eee79e956
> Gecko     https://hg.mozilla.org/releases/mozilla-aurora/rev/d32649a24965
> BuildID   20140714160201
> Version   32.0a2
> Firmware Version: v122
> 
> This issue does not happen when I first tried this case, and the actual
> result just as what Cody mentioned; But after I reboot the device and tested
> again, this issue happened.

Did you try this with 273MB mem config on Flame? 

Jim, can you help investigate this?

Comment 7

[Jim Porter (:squib)]

(In reply to Hema Koka [:hema] from comment #6)
> Jim, can you help investigate this?

Not at the moment, no. My bluetooth headset is a couple of thousand miles away.

Comment 8

[Alison Shiue]

(In reply to Hema Koka [:hema] from comment #6)
> Did you try this with 273MB mem config on Flame? 

No, I did not try with 273MB.  
And this issue could not be reproduced when I tune memory to 273MB.

Comment 9

[Pooja]

(In reply to Hema Koka [:hema] from comment #4)
> Pooja,
> 
> Please provide more information on the DUT. Comment #3 says we cannot
> reproduce on Flame/Buri
> 

Target: 8x10 QRD
Gaia: "35a9b715e7348ec738ff6c8a59f50190390a06f2"
Gecko "2fb60c777d3f82d580cba249e5e01a167a01de39"
Version 32.0a2
Build ID 20140710185233

And as Ashiue mentioned this is not every time reproducible. reproducibility is 3/5.

Comment 10

[bhavana bajaj [:bajaj]]

I'll set teh status falgs here, Given the reproducibility rate, this is worth investigating.

Allision, can give this a couple of tries to see if this is reproducible in 1.4 or 2.1 as well ?

Comment 11

[Alison Shiue]

(In reply to bhavana bajaj [:bajaj] [NOT reading Bugmail, needInfo please] from comment #10)
> I'll set teh status falgs here, Given the reproducibility rate, this is
> worth investigating.
> 
> Allision, can give this a couple of tries to see if this is reproducible in
> 1.4 or 2.1 as well ?

Yes, I can reproduce this issue on both 1.4 and 2.1 version. (default memory)

[2.1]
Gaia             0910be495385d90acdeaddbeaf1fba315aff57b0
Gecko            https://hg.mozilla.org/mozilla-central/rev/7883d8e9f210
BuildID          20140715160202
Version          33.0a1
Firmware Version v122

[1.4]
Gaia             393d72937727ad20e82b2ff7b13e3d7ff077a9f0
Gecko            https://hg.mozilla.org/releases/mozilla-b2g30_v1_4/rev/932c37978d37
BuildID          20140715160201
Version          30.0
Firmware Version v122

Comment 12

[Hema Koka [:hema]]

Jim, 

Dominic has a bluetooth headset, please borrow it from him and investigate this out tomorrow. 

Thanks
Hema

Comment 13

[bhavana bajaj [:bajaj]]

(In reply to Cody Roesch [:croesch] from comment #3)
> This bug does NOT repro on: Flame 2.1, Flame 2.0, Flame 1.4, Buri 2.1
> 
> Actual Result: Playing music over bluetooth attached to a Airband bluetooth
> headset then inserting a separate wired headset into the phone only produces
> a loudness warning message.
> 
> Environmental Variables:
> Device: Flame Master
> Build ID: 20140714061512
> Gaia: 88e0a972280bb35847c010b8c3f1481fa80f3847
> Gecko: 340b19c14d3d
> Version: 33.0a1 (Master)
> Firmware Version: v122
> -----------------------------------------------
> Environmental Variables:
> Device: Flame 2.0
> Build ID: 20140714071406
> Gaia: 726ca25ef1dbb83d8ed8a902b46fb340a3f95927
> Gecko: 002aee8237a3
> Version: 32.0a2 (2.0)
> Firmware Version: v122
> -----------------------------------------------
> Environmental Variables:
> Device: Flame 1.4
> Build ID: 20140711225112
> Gaia: b7d36622c7df92c976c37520ccab25199c7ada91
> Gecko: dbebcdab47aa
> Version: 30.0 (1.4)
> Firmware Version: v122
> -----------------------------------------------
> Environmental Variables:
> Device: Buri Master
> Build ID: 20140714061512
> Gaia: 88e0a972280bb35847c010b8c3f1481fa80f3847
> Gecko: 340b19c14d3d
> Version: 33.0a1 (Master)
> Firmware Version: v1.2device.cfg

Cody, can you confirm you are trying the issue a couple of times here as I see allison be able to reproduce this. Wanted to identify that rate of reproducibility is the key issue here as I see some discrepancies.

Comment 14

[Cody Roesch [:croesch]]

After testing this a little more, I'm able to get this issue across all devices. Not sure why i wasn't able to get this yesterday.

This bug repro's on: Flame 2.1 Master, Flame 2.0, Flame 1.4 and Buri 2.1

Actual Results: When connected to a BT headset, and listening to music, if a wired headset is plugged into the DuT, the music will pause and a message "Music paused during the call" will appear.

Environmental Variables:
Device: Flame Master
BuildID: 20140716065602
Gaia: d29773d2a011825fd77d1c0915a96eb0911417b6
Gecko: f6e46d1fc903
Version: 33.0a1 (Master) 
Firmware Version: v122
-----------------------------------------------
Environmental Variables:
Device: Flame 2.0
Build ID: 20140716084350
Gaia: 5fa5611fc747d397dbaa2c5ccdc776f1dfd2b6e8
Gecko: 3045ff641a0b
Version: 32.0a2 (2.0)
Firmware Version: v122
-----------------------------------------------
Environmental Variables:
Device: Flame 1.4
Build ID: 20140716073943
Gaia: edec55eb65b4b8be3e44bedab2949f295784dff1
Gecko: 90162ab32776
Version: 30.0 (1.4)
Firmware Version: v122
-----------------------------------------------
Environmental Variables:
Device: Buri Master
Build ID: 20140716065602
Gaia: d29773d2a011825fd77d1c0915a96eb0911417b6
Gecko: f6e46d1fc903
Version: 33.0a1 (Master)
Firmware Version: v1.2device.cfg

Comment 15

[Cody Roesch [:croesch]]

3 of 4 builds tested today, the issue occurred on the first try. Yesterdays builds were not as generous for some reason.

Comment 16

[Marcia Knous [:marcia]]

I tested this with Flame using:

Gaia   5f8b1b8a2da9e3b531eee817a669f57fa4d9b9c6
SourceStamp 913827496f65
BuildID 20140716000201
Version 32.0a2
Base: V122 + fonts
512 Memory

Wired headphones were Bose
Sony bluetooth headset was: Sony MDR 10RBT

I also tried rebooting the device and wasn't able to reproduce it with this scenario.

When I plugged in the wired headset while the BT was playing, all that happened was that the music player stopped, but the app did not hang.

I then repeated the same STR with 273 and wasn't able to reproduce it either.

Comment 17

[Jim Porter (:squib)]

So, we get an onscostatuschanged in shared/js/media/remote_controls.js, which fires self._commandHandler(AVRCP.PAUSE_PRESS); a bit later. From there, we wind up in apps/music/js/Player.js, and since we're 1) pausing, and 2) SCO is enabled, we show the little overlay.

However, I have no idea why all this is happening. It would probably be best for someone familiar with Bluetooth to look into this (or at least point me at some documentation).

Comment 18

[Hema Koka [:hema]]

shuang/eric, can we get some help on the bluetooth side please.

Comment 19

[Shawn Huang [:shawnjohnjr][:shuang] (as a happy gecko contributor)]

It looks like |connectSco| called via callscreen call_handler.js, this is why we can see SCO connection status changed event suddenly was been dispatched due to onheadphoneschange invoked.

7-18 11:25:41.419   292   292 I GeckoDump: !!!!!!!!!!!!!!!!switchToDefaultOut connectSco!!!!!!!!!!!!!!!
07-18 11:25:41.419   292   292 I GeckoDump: -----------connectSco-------------
07-18 11:25:41.419   292   292 I GeckoDump: connectSco dumpstack:BluetoothHelper/<.connectSco@app://callscreen.gaiamobile.org/shared/js/bluetooth_helper.js:98:15
07-18 11:25:41.419   292   292 I GeckoDump:  switchToDefaultOut@app://callscreen.gaiamobile.org/js/calls_handler.js:664:7
07-18 11:25:41.419   292   292 I GeckoDump:  cs_switchToDefaultOut@app://callscreen.gaiamobile.org/js/call_screen.js:408:5
07-18 11:25:41.419   292   292 I GeckoDump:  onheadphoneschange@app://callscreen.gaiamobile.org/js/calls_handler.js:62:11


After that connectSco was been triggered again by callscreen while onscostatuschaged (because previously sco connected).

07-18 11:25:41.479   292   292 I GeckoDump: connectSco dumpstack:BluetoothHelper/<.connectSco@app://callscreen.gaiamobile.org/shared/js/bluetooth_helper.js:98:15
07-18 11:25:41.479   292   292 I GeckoDump:  switchToDefaultOut@app://callscreen.gaiamobile.org/js/calls_handler.js:664:7
07-18 11:25:41.479   292   292 I GeckoDump:  cs_switchToDefaultOut@app://callscreen.gaiamobile.org/js/call_screen.js:408:5
07-18 11:25:41.479   292   292 I GeckoDump:  onscostatuschanged@app://callscreen.gaiamobile.org/js/calls_handler.js:69:9

Comment 20

[Shawn Huang [:shawnjohnjr][:shuang] (as a happy gecko contributor)]

We are not in the middle of incoming/outgoing call session in this case, but callscreen still handle onheadphoneschange and trigger BT SCO connection during a2dp streaming, I think that doesn't make sense here. ni? Etienne.

Comment 21

[Shawn Huang [:shawnjohnjr][:shuang] (as a happy gecko contributor)]

SCO/A2DP should be mutual exclusive, so I think this bug is not related to Music app but call screen.

Comment 22

[Dominic Kuo [:dkuo]]

Thanks Shawn, I think I got the issue and will take it over from Jim because the fix might need some knowledge of the bluetooth module in the system app, which I should be familiar with.

Comment 23

[Hema Koka [:hema]]

Dominic has identified the issue (not a music problem), but this probably won't make it by FC date. He will update with more information.

Comment 24

[Dominic Kuo [:dkuo]]

Found bug 1026475 caused this regression.

Ben, can you please describe why you made that change in bug 1026475? to connect the SCO when the telephony is not connected seems confused, and caused the other apps(like music) listen to SCO status behave wrong.

I will probably fix this issue by checking the telephony status then decide to connect SCO or not, it should keep this original behavior and fix the issue in music.

Comment 25

[Dominic Kuo [:dkuo]]

Attachment: patch

Etienne, would you please review this? thanks.

Comment 26

[Dominic Kuo [:dkuo]]

Changing the title to fit the reality.

Comment 27

[Ben Tian (inactive)]

(In reply to Dominic Kuo [:dkuo] (Media Apps Work Week, July 14-18) from comment #24)
> Found bug 1026475 caused this regression.
> 
> Ben, can you please describe why you made that change in bug 1026475? to
> connect the SCO when the telephony is not connected seems confused, and
> caused the other apps(like music) listen to SCO status behave wrong.

Bug 1026475 alters in-call audio path to match pre-2.0 UI. When both headphone and BT headset are connected, the callscreen audio path switch UI didn't show "headphone" but 3 options only: speaker, receiver, and BT headset. Therefore I made the behavior favor BT headset. The change intends to affect in-call audio path only.
 
Also refer to bug 989233. Original behavior was to connect SCO when headphone is plugged in. In bug 989233 I thought headphone should be favored so removed it. And then in bug 1026475 I found the UI inconsistency so restored the original behavior.

> I will probably fix this issue by checking the telephony status then decide
> to connect SCO or not, it should keep this original behavior and fix the
> issue in music.

Agree. This is the correct fix to make other apps immune from callscreen audio path changes.

Comment 28

[Dominic Kuo [:dkuo]]

Thanks Ben, then I think my patch should be right to fix this issue since I didn't change the behaviour on selecting the destination which the audio will be routed. Let's wait for Etienne's review :)

Comment 29

[Etienne Segonzac (:etienne)]

Comment on attachment 8458853 [details] [review]
patch

Redirecting to Rik, but you're probably missing a new unit-test :)

Comment 30

[Anthony Ricaud (:rik)]

Comment on attachment 8458853 [details] [review]
patch

Instead of telephony.active, we should test 'displayed', a "global" to this module. This will make it more robust. telephony.active could be false if we only have one call and it is on hold. 'displayed' is true when the callscreen is open.

And yup, we'll need one more test here. By using 'displayed', a test breaks so we need to update its title. And the new test should open the callscreen before trying to connect to SCO.

Let me know if you want me to help with this test.

Comment 31

[Hema Koka [:hema]]

Thanks Dominic for addressing this issue. This should probably be in Bluetooth component

Comment 32

[Dominic Kuo [:dkuo]]

Thanks guys, I didn't aware I should add an unit test for this, I will address it then request review again.

(In reply to Hema Koka [:hema] from comment #31)
> Thanks Dominic for addressing this issue. This should probably be in
> Bluetooth component

Hema, this should be a callscreen issue because it connects SCO in a wrong timing then causes music to have wrong behaviour, I will change the component to dialer first since there is no callscreen component.

Comment 33

[Dominic Kuo [:dkuo]]

Comment on attachment 8458853 [details] [review]
patch

Rik,

I have addressed the issue with the new tests, would you please review it again? thanks.

Comment 34

[Anthony Ricaud (:rik)]

Comment on attachment 8458853 [details] [review]
patch

r+ with the tests modified for readability.

Thanks!

Comment 35

[Dominic Kuo [:dkuo]]

Thanks Rik!

master: cabe1a8891961e8654ef4a60491e9907c7217465

Comment 36

[Dominic Kuo [:dkuo]]

v2.0: 29266e18c35f4e72e35f1bba0e34f2fb6b995cc3

Comment 38

[Shawn Huang [:shawnjohnjr][:shuang] (as a happy gecko contributor)]

status-b2g-v1.4 already marked as affected. Please triage for v1.4.

Comment 39

[James Zhang (Spreadtrum)]

Please land it to 1.4.

Comment 40

[Rachelle Yang [:ryang][[e-mail]]]

Hi Dominic,could you please kindly help us to rebase 1.4 patch? Thank you so much.

Comment 41

[Kai-Chih Hu [:khu]]

I think we should mark approval‑gaia‑v1.4, right?

Comment 42

[Dominic Kuo [:dkuo]]

v1.4: eb3b185325901d4c04e2d43eb58d90835213bea9

I already uplifted to v1.4...please needinfo me again if we don't want this on v1.4.

Comment 43

[[:SalvadorR]]

Needs to be covered by new test case to be covered.

Comment 44

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

Test case added in moztrap:

https://moztrap.mozilla.org/manage/case/14331/

Comment 45

[Paladin]

Attachment: Verify_video.3gp

This issue has been verified successfully on Flame 2.0,2.1

See attachment: Verify_video.3gp
Reproducing rate: 0/5
Flame2.0  build:
Gaia-Rev        8d1e868864c8a8f1e037685f0656d1da70d08c06
Gecko-Rev       https://hg.mozilla.org/releases/mozilla-b2g32_v2_0/rev/c756bd8bf3c3
Build-ID        20141127000203
Version         32.0
Flame2.1 build:
Gaia-Rev        5372b675e018b6aac97d95ff5db8d4bd16addb9b
Gecko-Rev       https://hg.mozilla.org/releases/mozilla-b2g34_v2_1/rev/f34377ae402b
Build-ID        20141127001201
Version         34.0