Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: XAUI and 64b/66b and State Machine i/o definitions

Title: RE: XAUI and 64b/66b and State Machine i/o definitions


As you will well recall -

The 1000BASE-X PCS Transmit state machine was defined in terms of its
GMII inputs (TXEN, etc) and its encoded 8B/10B outputs. Equally, the
1000BASE-X PCS Receive state machine was defined in terms of its 8B/10B
encoded inputs and GMII outputs (RXDV etc). As we know, the hard part of
defining this state machine was dealing with all the possible error
conditions, but it ensured that we had a detailed definition of how the
PCS worked.

Looking at XGXS, I imagine it will be necessary to define the XGXS Transmit
state machine in terms of its XGMII inputs, and encoded XAUI outputs. Equally,
it makes sense to define the XGXS Receive state machine in terms of its XAUI
inputs and XGMII outputs. Once this has been done (and all the gorey details
of error cases defined!), I think it is obvious that you need to instantiate
an XGXS Transmit state machine and an XGXS Receive state machine at each end
of the XAUI link.

As a consequence of this, the PCS Transmit state machine will have XGMII inputs
and the PCS Receive state machine will have XGMII outputs.

I think this is the best way to go for a number of reasons, one being that the
XGXS/XAUI is optional. There will be no extra work involved in doing things this
way, because there will be an identical XGXS Transmit and Receive state machine
defined at the 'RS end of the XAUI link' and at the 'PCS end of the XAUI link'.

Also, in the Layer drawings, I think it would be worth illustrating the option where
the XGXS layer is bypassed, showing that the RS sublayer can talk directly to the PCS.

Una Quinlan

-----Original Message-----
From: Rich Taborek [mailto:rtaborek@xxxxxxxxxxx]
Sent: 21 March 2000 18:32
Subject: Re: XAUI and 64b/66b


Good point. I ran out of room on the the front page to list the many folks from
the same company in support of this proposal :-)

I'll allow the chair to rule on your question since that's his job.

In this case, I had obtained advance permission from each individual referenced
on the front page of the XAUI/XGXS proposal to also list the company name.

Best Regards,

Patrick Gilliland wrote:
> Rich,
> 1.) I believe it is proper to mention the support
> of 27 individuals for the 8B/10b XAUI/XGXS proosal.
> Typically, at IEEE we do not have companies voting
> their support as a block, and it has been the intent
> of the IEEE not to engage companies formally as
> voting members.
> I have seen quite a number of these corporate references
> in the last few days, and I thought the chair might have
> noticed by now and stepped in.
> 2.) It raises an interesting theoretical question about
> individuals who are incorporated as one-man consulting
> and design services.  Should they be barred from voting
> because their vote might also be interpreted as representing
> the interest of their company?
> Before anyone flames me for 2.), please consider the source
> and the purely diversionary motives.
> Best Regards,
> Patrick Gilliland
Richard Taborek Sr.                 Phone: 408-845-6102      
Chief Technology Officer             Cell: 408-832-3957
nSerial Corporation                   Fax: 408-845-6114
2500-5 Augustine Dr.        mailto:rtaborek@xxxxxxxxxxx
Santa Clara, CA 95054