Hi Community,
I kept debugging this issue for the past a couple of weeks, didn't
make much advances yet but found one interesting fact.
The default Front_Center.wav file's sample rate is 48000, when using
gdb to observe the calltrace for bluetooth_hsp_write(), I found that
alsa-lib has introduced a nested plugins structure, in particular, a
"rate" plugin to master another slave "ioplug" plugin. It seems that
the master "rate" plugin is used to convert the sample rate from 48000
to 8000.
It is when this rate plugin is used that the low level
bluetooth_hsp_write() is not correctly invoked by aplay's pcm_write -
no more than two mtu amount of frames are transmitted before pcm_write
thought all data has been sent successfully.
So on my laptop I used arecord to generate a wav file with 8000 sample
rate, then I found it can be properly played by "aplay -D btheadset"
successfully ! :-)
I tried to register the alsa-lib mailing list for further opinions but
my subscription not approved in one week.
Moreover, although aplay can play a 8000 sample rate wav file through
bluetooth headset, after playing it a couple of times, there is no
sound output anymore, although aplay can process the wav file as
normal. Also the debug info of bluetoothd doesn't look fishy:
$ hcitool con
Connections:
< ACL 1C:48:F9:AD:19:EE handle 70 state 1 lm MASTER AUTH ENCRYPT
$ aplay -D btheadset my_record_8k.wav
Playing WAVE 'my_record_8k.wav' : Signed 16 bit Little Endian, Rate
8000 Hz, Mono
$
bluetoothd[4102]: audio/unix.c:server_cb() Accepted new client
connection on unix socket (fd=33)
bluetoothd[4102]: audio/unix.c:client_cb() Audio API: BT_REQUEST <-
BT_GET_CAPABILITIES
bluetoothd[4102]: audio/unix.c:unix_ipc_sendmsg() Audio API:
BT_RESPONSE -> BT_GET_CAPABILITIES
bluetoothd[4102]: audio/unix.c:client_cb() Audio API: BT_REQUEST <- BT_OPEN
bluetoothd[4102]: audio/unix.c:handle_sco_open() open sco - object=ANY
source=ANY destination=1C:48:F9:AD:19:EE lock=write
bluetoothd[4102]: audio/unix.c:unix_ipc_sendmsg() Audio API:
BT_RESPONSE -> BT_OPEN
bluetoothd[4102]: audio/unix.c:client_cb() Audio API: BT_REQUEST <-
BT_SET_CONFIGURATION
bluetoothd[4102]: audio/unix.c:unix_ipc_sendmsg() Audio API:
BT_RESPONSE -> BT_SET_CONFIGURATION
bluetoothd[4102]: audio/unix.c:client_cb() Audio API: BT_REQUEST <-
BT_START_STREAM
bluetoothd[4102]: audio/headset.c:headset_set_state() State changed
/org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE: HEADSET_STATE_CONNECTED ->
HEADSET_STATE_PLAY_IN_PROGRESS
bluetoothd[4102]: audio/media.c:headset_state_changed()
bluetoothd[4102]: audio/media.c:headset_state_changed()
bluetoothd[4102]: audio/headset.c:sco_connect_cb() SCO socket opened
for headset /org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE
bluetoothd[4102]: audio/headset.c:sco_connect_cb() SCO fd=35
bluetoothd[4102]: audio/unix.c:unix_ipc_sendmsg() Audio API:
BT_RESPONSE -> BT_START_STREAM
bluetoothd[4102]: audio/unix.c:unix_ipc_sendmsg() Audio API:
BT_INDICATION -> BT_NEW_STREAM
bluetoothd[4102]: audio/headset.c:headset_set_state() State changed
/org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE:
HEADSET_STATE_PLAY_IN_PROGRESS -> HEADSET_STATE_PLAYING
bluetoothd[4102]: audio/media.c:headset_state_changed()
bluetoothd[4102]: audio/media.c:headset_state_changed()
bluetoothd[4102]: audio/transport.c:media_transport_acquire()
Transport /org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE/fd5: read lock
acquired
bluetoothd[4102]: audio/transport.c:media_transport_acquire()
Transport /org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE/fd5: write lock
acquired
bluetoothd[4102]: audio/transport.c:media_owner_create() Owner
created: sender=:1.76 accesstype=rw
bluetoothd[4102]: audio/transport.c:media_transport_release()
Transport /org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE/fd5: read lock
released
bluetoothd[4102]: audio/transport.c:media_transport_release()
Transport /org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE/fd5: write lock
released
bluetoothd[4102]: audio/transport.c:media_owner_free() Owner :1.76
bluetoothd[4102]: audio/unix.c:client_cb() Unix client disconnected (fd=33)
bluetoothd[4102]: audio/headset.c:headset_set_state() State changed
/org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE: HEADSET_STATE_PLAYING ->
HEADSET_STATE_CONNECTED
bluetoothd[4102]: audio/media.c:headset_state_changed()
bluetoothd[4102]: audio/media.c:headset_state_changed()
bluetoothd[4102]: audio/unix.c:client_free() client_free(0x7f9e16071e00)
I tried to restart the bluetoothd and re-pair/connect the bluetooth
headset, but got the same result. Any idea?
Moreover, how can I use the VolumeUp/Down method on org.bluez.Control
interface to control the volume of the bluetooth headset properly?
$ echo $DEVICE
/org/bluez/4102/hci1/dev_1C_48_F9_AD_19_EE
$ dbus-send --system --print-reply --dest=org.bluez $DEVICE
org.bluez.Control.VolumeUp
Error org.freedesktop.DBus.Error.UnknownMethod: Method "VolumeUp" with
signature "" on interface "org.bluez.Control" doesn't exist
Cheers,
Harry
On Fri, Mar 18, 2016 at 7:57 PM, Qingtao Cao <qingtao.cao.au@xxxxxxxxx> wrote:
> Thanks Lukasz for your kind reply!
>
> Frankly speaking I have been working on bluez 4.101 for several months
> and validated quite some aspects of it, so it could be too costly for
> me to upgrade to bluez 5 at the end of the game.
>
> Cheers,
> Harry
>
> On Fri, Mar 18, 2016 at 5:01 PM, Łukasz Rymanowski
> <lukasz.rymanowski@xxxxxxxxxxx> wrote:
>> Hi Harry,
>>
>> On 18 March 2016 at 06:57, Qingtao Cao <qingtao.cao.au@xxxxxxxxx> wrote:
>>> Hi Community,
>>>
>>> I kept debugging this issue today and found out that aplay failed to
>>> send out the .wav file content to bluetooth headset.
>>>
>>> From the "strace aplay -D btheadset .wav" I can see that the wav file
>>> is read in 12000B chunk, but sent to bluetooth socket in 48B chunk of
>>> data. Compared with the strace result for aplay through speaker, the
>>> wav file can be read correctly, however, when using bluetooth headset
>>> aplay always seems to send 48B zero again and again.
>>>
>>> This explains why there is no sound output at all.
>>>
>>> fd 6 is the opened wav file.
>>> fd 7 is the connection with /org/bluez/audio.
>>>
>>> ...
>>> read(6, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 12000) = 12000
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> read(6, "w\37\210 \21!\6!\320 q
>>> \244\37u\36b\35\222\34a\33\314\31\267\30C\30\265\27\234\26"..., 12000)
>>> = 12000
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> read(6, "\t\23\205\23\27\24\206\24\330\24G\25\301\25!\26x\26\335\26B\27\211\27\301\27\24\30`\30\211\30"...,
>>> 12000) = 12000
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> read(6, "\334\377\343\377\347\377\337\377\332\377\340\377\344\377\342\377\337\377\327\377\324\377\345\377\350\377\331\377\343\377\362\377"...,
>>> 12000) = 12000
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> clock_gettime(CLOCK_REALTIME, {1457576840, 978074823}) = 0
>>> read(6, "\374\377\361\377\345\377\363\377\366\377\361\377\364\377\364\377\367\377\370\377\352\377\352\377\371\377\362\377\360\377\4\0"...,
>>> 12000) = 12000
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> send(7, "\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"...,
>>> 48, 0) = 48
>>> ... (endless sending 48b of zeros) ...
>>>
>>> Any comments?
>>>
>>> Thanks,
>>> Harry
>>>
>>> On Thu, Mar 17, 2016 at 2:47 PM, Qingtao Cao <qingtao.cao.au@xxxxxxxxx> wrote:
>>>> Hi Community,
>>>>
>>>> I run into a bluetooth headset issue with bluez 4.101. After a
>>>> successful pairing and connection, I failed to play a .wav file
>>>> through the bluetooth headset by aplay. I have been searching around
>>>> for a couple days and looked at bluez 4.101 source code to ensure the
>>>> audio.conf settings make sense but still clueless.
>>>>
>>>> My laptop is Ubuntu 15.04 with bluez 4.101-ubuntu25 (kernel:
>>>> 3.19.0-25-generic), the target embedded uses yocto "dizzy" release
>>>> with bluez 4.101 as well (kernel: 3.4).
>>
>> I recommend to use ubuntu 15.10 which have new BlueZ 5.x on it.
>>
>> On both laptop and target
>>>> machine the libasound-module-bluez has been installed properly.
>>>>
>>>> My settings are:
>>>>
>>>> 1. /etc/bluetooth/main.conf
>>>> The default one, in particular, Class=0x00100
>>>>
>>>> 2. /etc/bluetooth/audio.conf
>>>>
>>>> [General]
>>>> Enable=Socket
>>>>
>>>> [Headset]
>>>> HFP=true
>>>> MaxConnected=1
>>>> FastConnectable=false
>>>>
>>>> [A2DP]
>>>> SBCSources=1
>>>> SBCSinks=1
>>>> MPEG12Sources=1
>>>> MPEG12Sinks=1
>>>>
>>>> 3. ~/.asoundrc
>>>> pcm.btheadset {
>>>> type plug
>>>> slave {
>>>> pcm {
>>>> type bluetooth
>>>> device "1C:48:F9:AD:19:EE"
>>>> profile "auto"
>>>> }
>>>> }
>>>>
>>>> hint {
>>>> show on
>>>> description "BT Headset"
>>>> }
>>>> }
>>>>
>>>> ctl.btheadset {
>>>> type bluetooth
>>>> }
>>>>
>>>> where I have hardcoded the bdaddr of my bluetooth headset.
>>>>
>>>> 4. My bluetooth headset has been paired and connected successfully:
>>>>
>>>> $ hcitool con
>>>> Connections:
>>>> < eSCO 1C:48:F9:AD:19:EE handle 70 state 1 lm MASTER
>>>> < ACL 1C:48:F9:AD:19:EE handle 69 state 1 lm MASTER AUTH ENCRYPT
>>>>
>>>>
>>>> 5. On my laptop, I can play through bluetooth headset with mplayer successfully:
>>>>
>>>> $ mplayer -ao alsa:device=btheadset /usr/share/sounds/alsa/Front_Center.wav
>>>>
>>>> MPlayer2 2.0-728-g2c378c7-4 (C) 2000-2012 MPlayer Team
>>>> Cannot open file '/home/harry/.mplayer/input.conf': No such file
>>>> or directory
>>>> Failed to open /home/harry/.mplayer/input.conf.
>>>> Cannot open file '/etc/mplayer/input.conf': No such file or directory
>>>> Failed to open /etc/mplayer/input.conf.
>>>>
>>>> Playing /usr/share/sounds/alsa/Front_Center.wav.
>>>> Detected file format: WAV / WAVE (Waveform Audio) (libavformat)
>>>> [lavf] stream 0: audio (pcm_s16le), -aid 0
>>>> Load subtitles in /usr/share/sounds/alsa/
>>>> Selected audio codec: Uncompressed PCM [pcm]
>>>> AUDIO: 48000 Hz, 1 ch, s16le, 768.0 kbit/100.00% (ratio: 96000->96000)
>>>> AO: [alsa] 8000Hz 1ch s16le (2 bytes per sample)
>>>> Video: no video
>>>> Starting playback...
>>>> A: 1.4 (01.4) of 1.4 (01.4) 0.1%
>>>>
>>>> Exiting... (End of file)
>>>>
>>>>
>>>> 6. However, on my laptop, play with aplay failed:
>>>>
>>>> $ aplay -D btheadset /usr/share/sounds/alsa/Front_Center.wav
>>>>
>>>> Playing WAVE '/usr/share/sounds/alsa/Front_Center.wav' : Signed 16
>>>> bit Little Endian, Rate 48000 Hz, Mono
>>>> ^CAborted by signal Interrupt...
>>>>
>>>> All I got is constant beeeeep noises and I have to press Ctrl^C to terminate it.
>>>>
>>>> 7. On my target machine, the rootfs is provided by "dizzy" yocto
>>>> release also with bluez 4.101. I have modified main.conf to define
>>>> "Class=0x5A020C" (or simply 0x000200) and audio.conf to enable
>>>> "SCORouting=PCM".
>>>>
>>>> 8. On my target machine, the same aplay command failed as well. But it
>>>> doesn't output any sound, no constant beep noise. Only when I
>>>> terminated it could I hear a short beep sound.
>>>>
>>>>
>>>> Given that mplayer works on bluetooth headset, I thought the bridging
>>>> from alsa to bluez works. However, why aplay failed? On my target
>>>> there is only aplay now and it's difficult to add mplay, so I would
>>>> have to make aplay work with bluetooth audio.
>>>>
>>>> Any comments are much appreciated!
>>>>
>>>> Cheers,
>>>> Harry
>>>>
>>>>
>>>> P.S.
>>>>
>>>> The following is the "bluetoothd -d -n" result during "aplay -D btheadset ...":
>>>>
>>>> bluetoothd[9490]: audio/unix.c:server_cb() Accepted new client
>>>> connection on unix socket (fd=33)
>>>> bluetoothd[9490]: audio/unix.c:client_cb() Audio API: BT_REQUEST <-
>>>> BT_GET_CAPABILITIES
>>>> bluetoothd[9490]: audio/unix.c:unix_ipc_sendmsg() Audio API:
>>>> BT_RESPONSE -> BT_GET_CAPABILITIES
>>>> bluetoothd[9490]: audio/unix.c:client_cb() Audio API: BT_REQUEST <- BT_OPEN
>>>> bluetoothd[9490]: audio/unix.c:handle_sco_open() open sco - object=ANY
>>>> source=ANY destination=1C:48:F9:AD:19:EE lock=write
>>>> bluetoothd[9490]: audio/unix.c:unix_ipc_sendmsg() Audio API:
>>>> BT_RESPONSE -> BT_OPEN
>>>> bluetoothd[9490]: audio/unix.c:client_cb() Audio API: BT_REQUEST <-
>>>> BT_SET_CONFIGURATION
>>>> bluetoothd[9490]: audio/unix.c:unix_ipc_sendmsg() Audio API:
>>>> BT_RESPONSE -> BT_SET_CONFIGURATION
>>>> bluetoothd[9490]: audio/unix.c:client_cb() Audio API: BT_REQUEST <-
>>>> BT_START_STREAM
>>>> bluetoothd[9490]: audio/headset.c:headset_set_state() State changed
>>>> /org/bluez/9490/hci1/dev_1C_48_F9_AD_19_EE: HEADSET_STATE_CONNECTED ->
>>>> HEADSET_STATE_PLAY_IN_PROGRESS
>>>> bluetoothd[9490]: audio/media.c:headset_state_changed()
>>>> bluetoothd[9490]: audio/media.c:headset_state_changed()
>>>> bluetoothd[9490]: audio/headset.c:sco_connect_cb() SCO socket opened
>>>> for headset /org/bluez/9490/hci1/dev_1C_48_F9_AD_19_EE
>>>> bluetoothd[9490]: audio/headset.c:sco_connect_cb() SCO fd=35
>>>> bluetoothd[9490]: audio/unix.c:unix_ipc_sendmsg() Audio API:
>>>> BT_RESPONSE -> BT_START_STREAM
>>>> bluetoothd[9490]: audio/unix.c:unix_ipc_sendmsg() Audio API:
>>>> BT_INDICATION -> BT_NEW_STREAM
>>>> bluetoothd[9490]: audio/headset.c:headset_set_state() State changed
>>>> /org/bluez/9490/hci1/dev_1C_48_F9_AD_19_EE:
>>>> HEADSET_STATE_PLAY_IN_PROGRESS -> HEADSET_STATE_PLAYING
>>>> bluetoothd[9490]: audio/media.c:headset_state_changed()
>>>> bluetoothd[9490]: audio/media.c:headset_state_changed()
>>>> bluetoothd[9490]: audio/transport.c:media_transport_acquire()
>>>> Transport /org/bluez/9490/hci1/dev_1C_48_F9_AD_19_EE/fd0: read lock
>>>> acquired
>>>> bluetoothd[9490]: audio/transport.c:media_transport_acquire()
>>>> Transport /org/bluez/9490/hci1/dev_1C_48_F9_AD_19_EE/fd0: write lock
>>>> acquired
>>>> bluetoothd[9490]: audio/transport.c:media_owner_create() Owner
>>>> created: sender=:1.309 accesstype=rw
>>>> bluetoothd[9490]: audio/transport.c:media_transport_release()
>>>> Transport /org/bluez/9490/hci1/dev_1C_48_F9_AD_19_EE/fd0: read lock
>>>> released
>>>> bluetoothd[9490]: audio/transport.c:media_transport_release()
>>>> Transport /org/bluez/9490/hci1/dev_1C_48_F9_AD_19_EE/fd0: write lock
>>>> released
>>>> bluetoothd[9490]: audio/transport.c:media_owner_free() Owner :1.309
>>>> bluetoothd[9490]: audio/unix.c:client_cb() Unix client disconnected (fd=33)
>>>> bluetoothd[9490]: audio/headset.c:headset_set_state() State changed
>>>> /org/bluez/9490/hci1/dev_1C_48_F9_AD_19_EE: HEADSET_STATE_PLAYING ->
>>>> HEADSET_STATE_CONNECTED
>>>> bluetoothd[9490]: audio/media.c:headset_state_changed()
>>>> bluetoothd[9490]: audio/media.c:headset_state_changed()
>>>> bluetoothd[9490]: audio/unix.c:client_free() client_free(0x7f59b52e54b0)
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in
>>> the body of a message to majordomo@xxxxxxxxxxxxxxx
>>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>>
>>
>>
>> --
>> BR / Pozdrawiam
>> Łukasz
--
To unsubscribe from this list: send the line "unsubscribe linux-bluetooth" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
