Re: [Bluez-devel] Re: Problems with anycom usb dongle

[Marcel Holtmann <marcel@holtmann.org>]

Hi Sebastian,

>  > What does "hcitool con" show when both (ACL + SCO) links were open?
> 
> (btsco's last line is RFCOMM channel 1 connected)
> # hcitool con
> Connections:
>          < ACL 00:08:1B:00:FF:33 handle 1 state 1 lm MASTER
> 
> (pressing the button at my headset starts the SCO stream)
> # hcitool con
> Connections:
>          < SCO 00:08:1B:00:FF:33 handle 1025 state 1 lm SLAVE
>          < ACL 00:08:1B:00:FF:33 handle 1 state 1 lm MASTER

using handle 1025 is weird. I've never seen this before.

> I've tested the anycom dongle with both headsets once again.
> Here is a (heavy striped) output of hcidump while testing one of those:
> 
> < HCI Command: Create Connection (0x01|0x0005) plen 13
>    0000: 33 ff 00 1b 08 00 18 cc  02 00 00 00 01
> ... (normal connect)

There must be "SCO add" in between.

>  > HCI Event: Connect Complete (0x03) plen 11
>    0000: 00 01 04 33 ff 00 1b 08  00 00 01

This is handle 1025 after the "SCO add" and this is fine.

>  > SCO data: handle 0x0000 dlen 0
> ... (a lot of those packets)
>  > SCO data: handle 0x0000 dlen 0

And this is broken hardware. You can't send something for handle 0 with
zero data if no connection handle is available.

> (I guess, here was my first button press)
>  > SCO data: handle 0x0401 dlen 48
>      0000: 58 00 60 ff f0 ff a0 ff  7c ff 98 ff d8 00 38 00
>      0010: 68 00 48 ff d0 ff 6c ff  5c ff a8 ff f0 00 50 00
>      0020: 70 00 38 ff c8 ff 7c ff  88 ff c0 00 20 00 70 00
> ... (a lot of those packets)
>  > SCO data: handle 0x0401 dlen 48
>      0000: f0 ff f0 ff f8 ff f8 ff  f0 ff f0 ff f0 ff f8 ff
>      0010: f8 ff f8 ff f0 ff f0 ff  f8 ff f8 ff f8 ff f0 ff
>      0020: f0 ff f8 ff f8 ff f8 ff  f0 ff f0 ff f8 ff f8 ff
> (the second button press)
>  > ACL data: handle 0x0001 flags 0x01 dlen 4
>      L2CAP(d): cid 0x0041 len 17 [psm 3]
>        RFCOMM(d): UIH: cr 0 dlci 2 pf 1 ilen 12 fcs 0x5c credits 1
>        0000: 41 54 2b 43 4b 50 44 3d  32 30 30 0d
> **
> < ACL data: handle 0x0001 flags 0x02 dlen 14
>      L2CAP(d): cid 0x01c3 len 10 [psm 3]
>        RFCOMM(d): UIH: cr 1 dlci 2 pf 0 ilen 6 fcs 0x9a
>        0000: 0d 0a 4f 4b 0d 0a
> **
>  > HCI Event: Number of Completed Packets (0x13) plen 5
>    0000: 01 01 00 01 00
> < HCI Command: Disconnect (0x01|0x0006) plen 3
>    0000: 01 04 13
> **
>  > HCI Event: Command Status (0x0f) plen 4
>    0000: 00 01 06 04
> **
>  > HCI Event: Disconn Complete (0x05) plen 4
>    0000: 00 01 04 13
>  > HCI Event: Max Slots Change (0x1b) plen 3
>    0000: 01 00 05
>  > SCO data: handle 0x0000 dlen 0
>  > SCO data: handle 0x0000 dlen 0
> ... flood ;)

The SCO part of this dongle seem totally broken. Should I mark it as
broken in hci_usb driver? Or should I simply drop SCO packets with no
data on handle 0 in the HCI core? Is handle 0 a valid handle?

Regards

Marcel




-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
http://productguide.itmanagersjournal.com/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel