Cell Relay Archive

Cell Relay Retreat>List Archive>month:1997-Jun> msg00058



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

Re: ATM using m68360

  • From: "h.j. bae" <hjb@pso.com>
  • Date: Fri, 06 Jun 1997 21:44:32 -0700

Malcolm Herring wrote:
> 
> The ATOM1 ATM Microcode Users Manual can be downloaded from
> www.mot.com/netcomm. This contains a complete description of the functions
> performed. In a nutshell, the ATOM1 microcode is essentially a TC layer
> function with some ATM/AAL features (such as an AAL5 CRC calculator). By
> using a CAM and the scatter/gather capabilities of the QUICC, all the ATM &
> AAL layer functions can be realised.

thanks.  after reading the datasheet, i'm inclined to
believe that ATOM1 microcode requires SAR functions (other
than crc32) to be done in software since transmit buffers
have to be provided as 52 byte cells.  it also seems like
that you have to implement all the signalling and OAM
in software, which is not surprising.  i've worked on
signalling and OAM stuff in software before.  i'm just
surprised to see that software is expected to actually
segment a packet into cells and vice versa.  does that
not incur too much overhead?  can m68360 handle that
and still do other things like task switching, tcp/ip
protocol processing, analog voice -> digital conversion, 
etc.?  it seems that perhaps powerQUICC is what i need.
anyone have real-life experiences with powerQUICC?

thanks again.