880557 - [Bluetooth]Error occurs when Adjust the HBS700 headset volume

Status: RESOLVED WONTFIX

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

Description

[leo.bugzilla.gecko]

step>
1. bluetooth on
2. HBS700 headset pair
3. volume key click
(listen to music or when no action)

log>
Error occurs in javascript files.

06-07 09:59:36.660 V       145      BluetoothHfpManager    ReceiveSocketData:731 [LGBT_gecko] received Socket message = AT+VGS=13
06-07 09:59:36.690 V       145      BluetoothHfpManager    SendLine:1151 [LGBT_gecko] sending message = OK
06-07 09:59:36.740 E       145      GeckoConsole           Content JS LOG at app://system.gaiamobile.org/shared/js/event_helper.js:4 in handle_mozChromeEvent: event_helper :handle_mozChromeEvent: bluetooth-volumeset
06-07 09:59:36.800 I       145      AudioManager           SetStreamVolumeIndex aStream=6, aIndex=13
06-07 09:59:36.800 V       150      AudioPolicyManagerBase setStreamVolumeIndex() stream 6, index 13
06-07 09:59:36.800 V       150      AudioPolicyManager     checkAndSetVolume() cannot set stream 6 volume with force use = 0 for comm
06-07 09:59:36.810 E       145      GeckoConsole           [JavaScript Error: "NS_ERROR_FAILURE: Component returned failure code: 0x80004005 (NS_ERROR_FAILURE) [nsIAudioManager.setStreamVolumeIndex]" {file: "chrome://browser/content/settings.js" line: 96}]
06-07 09:59:36.810 E       145      GeckoConsole           [JavaScript Error: "NS_ERROR_FAILURE: Component returned failure code: 0x80004005 (NS_ERROR_FAILURE) [nsIAudioManager.setStreamVolumeIndex]" {file: "chrome://browser/content/settings.js" line: 96}]

Comment 1

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

It seems AudioManager can only handle volume change during "In Call" state, if Bluetooth gets AT+VGS=13, Bluetooth notify AudioManager to set volume but eventually AudioService cannot perform operation for set volume. Randy, can you help to comment on AudioManager behavior? :)

Comment 2

[Randy Lin [:rlin] - Be Mozillian on 2015/01/26]

Audio manager block this behavior
    // do not change in call volume if bluetooth is connected and vice versa
    if ((stream == AudioSystem::VOICE_CALL && mForceUse[AudioSystem::FOR_COMMUNICATION] == AudioSystem::FORCE_BT_SCO) ||
        (stream == AudioSystem::BLUETOOTH_SCO && mForceUse[AudioSystem::FOR_COMMUNICATION] != AudioSystem::FORCE_BT_SCO)) {
        LOGE("checkAndSetVolume() cannot set stream %d volume with force use = %d for comm",
             stream, mForceUse[AudioSystem::FOR_COMMUNICATION]);
        return INVALID_OPERATION;
    }
Gonk AudioManager got fail result and cause js error.
We need to find out when does bt_sco volume set during the voice call on android phone.

Comment 3

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

Hey Marco,
It seems you're checking volume related problems, would you take this bug? Here Bluetooth is correctly notify AudioManager but I think we need to handle the error inside the audio module (I almost think invalid operation error code return from AudioPolicyManager is platform dependent). Randy and I checked Nexus4 and Unagi, both of them behave differently.

Comment 4

[Marco Chen [:mchen]]

Hi Vasanth,

Could you help to give some suggestion on this case due to it is related to Audio HAL?

Issue:
  BT_SCO volume can't be adjusted during non-voice call state.

Our investigation:
  Please refer to comment 2, it shows that AudioPolicyManager will return error directly when we tried to set BT_SCO volume but mForceUse is not FOR_COMMUNICATION.

Question:
  1. Is this a common behavior of preventing to set BT_SCO on non-voice call state?
  2. What is the expected result on Android version?

Thanks.

Comment 5

[Marco Chen [:mchen]]

Hi Leo,

Could you fire a SR to your chip vendor for clarifying the our investigation here?
Because the part from audio HAL is came from CAF.

Thanks.

Comment 6

[leo.bugzilla.gecko]

Dear Mozilla Team,

I think that this should block in gecko side.
Because it seems to be only send vgs when sco is connected.

"+VGS" command is sent as below,
if (IsConnected()) {
    SendCommand("+VGS: ", mCurrentVgs);
}

So if it is received AT+VGS when sco is not connected, bluetooth-volume-change should not be run as above.
Please check this point.

Thanks.


(In reply to Marco Chen [:mchen] from comment #5)
> Hi Leo,
> 
> Could you fire a SR to your chip vendor for clarifying the our investigation
> here?
> Because the part from audio HAL is came from CAF.
> 
> Thanks.

Comment 7

[Marco Chen [:mchen]]

Hi Leo,

Thanks for your reply.

Hi Shawn,

Could you help to check the comment 6 from Leo? Thanks.

Comment 8

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

Hi leo,
I think maybe i misunderstood bug your description? From the log you mentioned, it looks like phone received AT+VGS=13 and javascript error message has been reported, which is volume change from headset. But what you mentioned is different from the original bug description. What comment 6 depicts that volume key was been clicked from phone instead of headset.
And yes, Comment 6 is right.

Comment 9

[leo.bugzilla.gecko]

Hi Shawn,

Sorry for hassle.
As your comment, +VGS is sent by phone.
The reason I mentioned it is reference for AT+VGS.

+VGS is not generated JS error because IsConnected() method.
It checks the sco connection and decide to try a SendCommand("+VGS: ", mCurrentVgs);

I think that AT+VGS seem to be solved in the same way, so I asked for you to review.

Could you check this?

Comment 10

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

Hi leo,
Well, I double checked the code, IsConnected() actually checked RFCOMM socket connection instead of SCO connection. Do you agree?

Comment 11

[leo.bugzilla.gecko]

Hi Shawn,

You are right. IsConnected() checked socket connection and IsScoConnected() checked the sco connection. I missed it. :)

Thanks.

Comment 12

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

Hi leo,
Just try to summarize here that I want to clarify with you again. Bug Description mentioned "3. volume key click
(listen to music or when no action)", I think here we shall discuss "Bluetooth HFP BT_SCO volume", not media stream volume. The Bug Description might need to be corrected.
So the problem we need to fix/figure out are:
Audio HAL AudioPolicyManager returns error directly when we tried to set BT_SCO volume but mForceUse is not FOR_COMMUNICATION. This is why we can see javascript error.

Comment 13

[leo.bugzilla.gecko]

Hi Shawn,

Yes, we shall discuss only BT_SCO volume.
As your comment, AudioPolicyManager returns error directly when set BT_SCO and not FOR_COMMUNICATION.

So I was thinking about how to fix it as below.
1. Does not notify AT+VGS when non-call state.
2. Makes other log return instead of INVALID_OPERATION.
3. Fix in AudioPolicyManager 

I just thought that no.1 is easy for fix.
If you think it should be fix in Audio part, I'll ask to our Audio engineer.

Please check this.
Thanks.

Comment 14

[leo.bugzilla.gecko]

I change leo? and I'll determine whether to apply in the future.

Comment 15

[Lukas Blakk [:lsblakk] use ?needinfo]

If this isn't important enough to block any more, it's not likely to make it in to v1.1 at this stage.

Comment 16

[BMO Automation]

Firefox OS is not being worked on