register HID with SDP error

[Iain Hibbert]

save the binary with 'hcidump -w file' and let's see it

On 1 April 2015 16:24:12 BST, Waitman Gobble <gobble.wa at gmail.com> wrote:
>On Apr 1, 2015 8:15 AM, "Maksim Yevmenkin" <maksim.yevmenkin at gmail.com>
>wrote:
>>
>> On Wed, Apr 1, 2015 at 8:09 AM, Waitman Gobble <gobble.wa at gmail.com>
>wrote:
>> > On Wed, Apr 1, 2015 at 8:05 AM, Maksim Yevmenkin
>> > <maksim.yevmenkin at gmail.com> wrote:
>> >> Hello,
>> >>
>> >> [...]
>> >>
>> >>> Can this 'syntax error' problem be related to the requested MTU?
>> >>>
>> >>> with essentially the 'same' connection request,
>> >>>
>> >>> here's the transaction log from an MS Windows client. No syntax
>> >>> errors. The requested MTU is 1024
>> >>>
>> >>> https://gist.github.com/waitman/2d5a72e9f9544848e832
>> >>>
>> >>> L2CAP(d): cid 0x40 len 346 [psm 1]
>> >>>
>> >>>
>> >>> and a log from android client. Requested MTU is 256.
>> >>>
>> >>> https://gist.github.com/waitman/e365debcd4a96d849153
>> >>>
>> >>> L2CAP(d): cid 0x43 len 247 [psm 1]
>> >>>
>> >>> note it's bombing out with something about aid 0x0200, which
>isn't
>> >>> actually an aid specified in my 0x1124 record.
>> >>
>> >> just a wild guess. please make sure that you maintain continuation
>> >> state properly. when sdp response is bigger than l2cap mtu,
>response
>> >> is chunked.
>> >
>> > Thank you, I was tinkering with things for awhile to try to find
>out
>> > why I was seeing 'syntax errors', but I'm thinking now there isn't
>a
>> > syntax error with the record, it's the hcidump program reading the
>> > broken up response.
>>
>> hcidump tries do be smart and re-assemble spd responses. if your
>> continuation state is wrong, hcidump might not be able to re-assemble
>> and parse response correctly
>>
>> thanks,
>> max
>
>Interesting. I will check into this. Thanks for the info.
>
>Waitman

iain