[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [lwip-users] uIP: inconsistent docs and implementation quirks
From: |
John Taylor |
Subject: |
RE: [lwip-users] uIP: inconsistent docs and implementation quirks |
Date: |
Mon, 26 Apr 2004 09:50:17 -0400 |
Here is a link for an example port of uIP to an SMSC91C96 LAN Controller
(it's SMSC 10Mbit only part). The example is part of the XMK RTOS.
Once you have downloaded the source code (either the 'full' or 'compact'
release is OK). Look in the following directories:
src/aplmt/hardware/smsc/lan91c96/uip
src/aplmt/networking/uip
XMK: http://sourceforge.net/projects/xmk
-----Original Message-----
From: address@hidden
[mailto:address@hidden On Behalf
Of Henning Riedel
Sent: Monday, April 26, 2004 9:03 AM
To: address@hidden
Subject: [lwip-users] uIP: inconsistent docs and implementation quirks
Well, I'm trying to port uIP to 68332 as interim platform, until 68HC12
hardware is here. I have an SMC91C111 eth module hooked.
I get a little pissed by the infos provided, which seem somehow
inconsistent.
You write in your samples for device drivers, to take the header infos
from uip_buf and the paylod from uip_appdata, but somehow it doesn't
work quite as it should.
You define IP and TCP headers to be 20 bytes each, but when I sniff,
some packets have e.g. 24 bytes TCP header. Though, there come up
questions as, when does a driver needs to take data from uip_buf and
when from uip_appdata? If the header is 24 bytes, will the options then
be in uip_buf or uip_appdata, as you say, application can put
uip_appdata somewhere else.
_______________________________________________
lwip-users mailing list
address@hidden
http://mail.gnu.org/mailman/listinfo/lwip-users