[Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Lars Grunewaldt]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi there,

right now I'm using two BT devices with my Epox dongle, an MX900 mice
and a Sony Ericcson HBH-35 Headset with snd-bt-sco (voice mode 0x0060,
before you ask ;)

both devices work fluently when used without the other, but when I use
headset and mouse at the same time there are various problems during
connect, scan and usage.

The main usage problem was that the sco layer got scrambled audio
(sounded much like random data) on about every third sco connection.
After a while, mouse and headset began to block each other, and the
stack seemed to get more and more, well, muddy. After some more
connection retries even the mouse, before only "hanging" when the
headset tried to connect, stopped working fully.

dmesg only noticed this:

hci_scodata_packet: hci0 SCO packet for unknown connection handle 76

When even connection was impossible, I plugged the BT dongle out and put
it back in. After reloading the hid module and programs the mouse and
headset worked again.

Are you interested in more specific reports? I could enable debugging
for some modules and try to collect some more data than this, if you
want to have something to debug on. I myself don't have the time right
now to drop too deeply into the subject :(

at least, mostly it works :)


I hope I can deliver some actual patches for snd-bt-sco against
alsa-1.0.5a soon, but right now I'm totally busy :/

thanks,
~  Lars
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFBF6kHQWC6DTWkDAoRAur6AKCqRgv5pB5rL9PE1V+rtuLmfNklRwCglOrb
NEzF/QeL2Q0z0vlF5H7Bn7s=
=cR7L
-----END PGP SIGNATURE-----


-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Marcel Holtmann]

Hi Lars,

> right now I'm using two BT devices with my Epox dongle, an MX900 mice
> and a Sony Ericcson HBH-35 Headset with snd-bt-sco (voice mode 0x0060,
> before you ask ;)

what does "hcitool info ..." say about your headset? What is the
firmware version of your EPoX dongle?

> both devices work fluently when used without the other, but when I use
> headset and mouse at the same time there are various problems during
> connect, scan and usage.
> 
> The main usage problem was that the sco layer got scrambled audio
> (sounded much like random data) on about every third sco connection.
> After a while, mouse and headset began to block each other, and the
> stack seemed to get more and more, well, muddy. After some more
> connection retries even the mouse, before only "hanging" when the
> headset tried to connect, stopped working fully.
> 
> dmesg only noticed this:
> 
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 76
> 
> When even connection was impossible, I plugged the BT dongle out and put
> it back in. After reloading the hid module and programs the mouse and
> headset worked again.

I think that is a hardware problem. A protocol analyser would help to
track this down ;)

Regards

Marcel




-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Lars Grunewaldt]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Marcel Holtmann wrote:
| Hi Lars,
|
|
|>right now I'm using two BT devices with my Epox dongle, an MX900 mice
|>and a Sony Ericcson HBH-35 Headset with snd-bt-sco (voice mode 0x0060,
|>before you ask ;)
|
|
| what does "hcitool info ..." say about your headset? What is the
| firmware version of your EPoX dongle?
|
|
Headset:
~        BD Address:  00:0E:07:33:DE:2B
~        Device Name: HBH-35
~        LMP Version: 1.1 (0x1) LMP Subversion: 0x8d40
~        Manufacturer: Ericsson Technology Licensing (0)
~        Features: 0xbc 0x28 0x01 0x00 0x00 0x00 0x00 0x00
~                <encryption> <slot offset> <timing accuracy> <role switch>
~                <sniff mode> <SCO link> <HV3 packets> <CVSD>


Mouse:
~        BD Address:  00:07:61:04:D6:7D
~        Device Name: Logitech MX900 Mouse
~        LMP Version: 1.1 (0x1) LMP Subversion: 0x350
~        Manufacturer: Cambridge Silicon Radio (10)
~        Features: 0xfc 0xff 0x0f 0x00 0x00 0x00 0x00 0x00
~                <encryption> <slot offset> <timing accuracy> <role switch>
~                <hold mode> <sniff mode> <park state> <RSSI> <channel
quality>
~                <SCO link> <HV2 packets> <HV3 packets> <u-law log>
<A-law log>
~                <CVSD> <paging scheme> <power control> <transparent SCO>


How can I find out what firmware my dongle has? It's a BT-DG03...

cu,
~  Lars

PS: great mouse, by the way. Works perfectly with Bluetooth and connects
easily to different systems.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFBF7N2QWC6DTWkDAoRArpEAJ4mMDV59gdZUyPBzsZ3a+iqrZsBtwCfZPPZ
TJWAeT+qMUXQ3nGs7DxbrYA=
=gvqb
-----END PGP SIGNATURE-----


-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Marcel Holtmann]

Hi Lars,

> | what does "hcitool info ..." say about your headset? What is the
> | firmware version of your EPoX dongle?
> |
> |
> Headset:
> ~        BD Address:  00:0E:07:33:DE:2B
> ~        Device Name: HBH-35
> ~        LMP Version: 1.1 (0x1) LMP Subversion: 0x8d40
> ~        Manufacturer: Ericsson Technology Licensing (0)
> ~        Features: 0xbc 0x28 0x01 0x00 0x00 0x00 0x00 0x00
> ~                <encryption> <slot offset> <timing accuracy> <role switch>
> ~                <sniff mode> <SCO link> <HV3 packets> <CVSD>

I don't see the hold mode. This can cause problem if they form a
scatternet. When both are connect what does "hcitool con" say?

> How can I find out what firmware my dongle has? It's a BT-DG03...

Use "hciconfig hci0 revision" as root.

Regards

Marcel




-------------------------------------------------------
This SF.Net email is sponsored by OSTG. Have you noticed the changes on
Linux.com, ITManagersJournal and NewsForge in the past few weeks? Now,
one more big change to announce. We are now OSTG- Open Source Technology
Group. Come see the changes on the new OSTG site. www.ostg.com
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Marcel Holtmann]

Hi Lars,

> | I don't see the hold mode. This can cause problem if they form a
> | scatternet. When both are connect what does "hcitool con" say?
> 
> without SCO connection active:
> Connections:
> ~        < ACL 00:0E:07:33:DE:2B handle 45 state 1 lm MASTER
> ~        > ACL 00:07:61:04:D6:7D handle 41 state 1 lm MASTER
> 
> 
> with active SCO connection:
> Connections:
> ~        < SCO 00:0E:07:33:DE:2B handle 44 state 1 lm SLAVE
> ~        < ACL 00:0E:07:33:DE:2B handle 45 state 1 lm MASTER
> ~        > ACL 00:07:61:04:D6:7D handle 41 state 1 lm MASTER

so both devices are in the same piconet. There should be no problem, but
from my position I can't tell you anything more.

> | Use "hciconfig hci0 revision" as root.
> 
> hci0:   Type: USB
> ~        BD Address: 00:04:61:81:4A:3E ACL MTU: 192:8  SCO MTU: 64:8
> ~        HCI 16.4
> ~        Chip version: BlueCore02
> ~        Max key size: 56 bit
> ~        SCO mapping:  HCI

This is a recent firmware version. So it should be fine.

> BTW, do you know what is the difference between using channel "1" and
> "no" channel for the SCO connection? The btsco client has this option,
> and if it's used as "1" the language for the RFCOMM commands seems to be
> different (i.e. if I press the take/hold button on the headset, I get
> "AT+BVRA=?" when using without "channel", but "AT+CKPD=200" when used
> with channel 1).

I've never used the btsco software.

Regards

Marcel




-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Lars Grunewaldt]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Marcel Holtmann wrote:
| Hi Lars,
|
|
|>| I don't see the hold mode. This can cause problem if they form a
|>| scatternet. When both are connect what does "hcitool con" say?
|>
|>without SCO connection active:
|>Connections:
|>~        < ACL 00:0E:07:33:DE:2B handle 45 state 1 lm MASTER
|>~        > ACL 00:07:61:04:D6:7D handle 41 state 1 lm MASTER
|>
|>
|>with active SCO connection:
|>Connections:
|>~        < SCO 00:0E:07:33:DE:2B handle 44 state 1 lm SLAVE
|>~        < ACL 00:0E:07:33:DE:2B handle 45 state 1 lm MASTER
|>~        > ACL 00:07:61:04:D6:7D handle 41 state 1 lm MASTER
|
|
| so both devices are in the same piconet. There should be no problem, but
| from my position I can't tell you anything more.
|

Hm OK thanks anyway. I'll try to find out what's happening.

|>BTW, do you know what is the difference between using channel "1" and
|>"no" channel for the SCO connection? The btsco client has this option,
|>and if it's used as "1" the language for the RFCOMM commands seems to be
|>different (i.e. if I press the take/hold button on the headset, I get
|>"AT+BVRA=?" when using without "channel", but "AT+CKPD=200" when used
|>with channel 1).
|
| I've never used the btsco software.

I'd not have asked if I thought it would have to do (much) with btsco.

Here's what this channel does (default if not given is 2):

	addr.rc_family = AF_BLUETOOTH;
	bacpy(&addr.rc_bdaddr, dst);
	addr.rc_channel = channel;
	if (connect(s, (struct sockaddr *) &addr, sizeof(addr)) < 0) {
		close(s);
		return -1;
	}

I just thought you might remember something from the BT specs or so. But
if you don't, I don't want to bother you, I can read up myself :)

thanks,
~  Lars
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFBGBC+QWC6DTWkDAoRAsD3AKCiWO+t6jNpm35Sq3EzoXG/+oPUeQCfdmAm
dlxxjGyWd45MV3RvVQvPJFI=
=nZcH
-----END PGP SIGNATURE-----


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Marcel Holtmann]

Hi Lars,

> I'd not have asked if I thought it would have to do (much) with btsco.
> 
> Here's what this channel does (default if not given is 2):
> 
> 	addr.rc_family = AF_BLUETOOTH;
> 	bacpy(&addr.rc_bdaddr, dst);
> 	addr.rc_channel = channel;
> 	if (connect(s, (struct sockaddr *) &addr, sizeof(addr)) < 0) {
> 		close(s);
> 		return -1;
> 	}
> 
> I just thought you might remember something from the BT specs or so. But
> if you don't, I don't want to bother you, I can read up myself :)

the channel is headset specific. You must use SDP to retrieve the
correct value. The first headset on the market (HBH-10) used RFCOMM
channel 6. Some others uses channel 1 etc.

Regards

Marcel




-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Lars Grunewaldt]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Marcel Holtmann wrote:
| Hi Lars,

| so both devices are in the same piconet. There should be no problem, but
| from my position I can't tell you anything more.

I think this message might have something to do with the problem:

hci_scodata_packet: hci0 SCO packet for unknown connection handle 44

My main problem does not even seem to be connected to the mouse, because
it also happens when the mouse is far far away and not connected.

In about 50% of SCO connections the sound coming from the headset is
totally garbage (not the sound send TO the headset!); I think this might
be connected to another report here on the list when someone mentioned
he had to cut the first byte when recording audio from an SCO channel
(because I'm using 16bit right now, too). This never happend with 8bit
transmissions over SCO.

Any ideas?

Kernel is 2.6.7 with mh4 patch.

cu,
~  Lars
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFBGL6+QWC6DTWkDAoRAl2lAJ4ktMZ3xgDH5euwJmISKE/bM6l4XwCfci3X
kXpVLb9uy8Nwu8qXP5rzBVg=
=qY1P
-----END PGP SIGNATURE-----


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] "comlicated" setup - known issues for hid and SCO?

[Marcel Holtmann]

Hi Lars,

> | so both devices are in the same piconet. There should be no problem, but
> | from my position I can't tell you anything more.
> 
> I think this message might have something to do with the problem:
> 
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 44
> 
> My main problem does not even seem to be connected to the mouse, because
> it also happens when the mouse is far far away and not connected.

this can also happen when the connection is already closed and then some
data packets arrive. Maybe it is a chip problem, because as I know CSR
never official supported the SCO over HCI/USB transport. Or the USB host
controller has problems with the ISOC transfers.

> In about 50% of SCO connections the sound coming from the headset is
> totally garbage (not the sound send TO the headset!); I think this might
> be connected to another report here on the list when someone mentioned
> he had to cut the first byte when recording audio from an SCO channel
> (because I'm using 16bit right now, too). This never happend with 8bit
> transmissions over SCO.

This can be that problem. We need an algorithm in the hci_usb driver to
detect this.

Regards

Marcel




-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel