[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [bluetooth] LMP interface



> I suggest to have a register for in the baseband
> controller for the LMP opcode and one for the operand.
> The controller will decode these information and form
> the packet and transmit it.
>
[Ling] I didn't quite catch your idea, can you give a clear example, like
what is opcode and what is operand, what is the usage cases?

> Other suggestion could be a single register (buffer)
> that holds all LMP packet and the base band controller
> will not look at its contents.

[Ling] LMP packet usually will be short and need to be send in higher
priority, it should have a seperate queue with SCO and ACL packets. And, of
course, baseband controller won't look at its contents, since it doesn't
bother him. :) I think this queue should be handle in LMP software, so the
packet can be in RAM, and be DMA or write to Link Controller FIFO whenever
this is slot available to send it.

> I have a question, does the LMp control the baseband
> controller it is connected to or just it is
> responsible for transmitting packets to the other end?

[Ling] In networking, the layer concept had better to be kept as clean as
possible. LMP won't control baseband controller, it just handle packet
communication with other peers(of course, another LMP on  the peers). It has
no idea on underlying layer, which is just a reliable channel can
intelligently send/recieve packets, this it, no more and no less. Hope this
is helpful..

Regards,
-Ling


--
To unsubscribe from bluetooth mailing list please visit http://www.opencores.org/mailinglists.shtml