From: spedpr@thor.cf.ac.uk (Paul Richards) Subject: Ethernet drivers Date: 6 Mar 1992 12:49:23 GMT
I'm looking at implementing an ethernet driver basically because I want
to abandone DOS entirely but I need tcp/ip connections to my sun.
Now I have a few queries. The simplest way to do this would be to set up
a block driver along the lines of what we have and assign a file name to
the ethernet driver. This wouldn't involve too much work I think. The
only hesitation I have is that this is totally different to the way
things are normally set up. On other systems the ethernet driver is
accessed through the sockets interface and nor the file system (I
think). Do people feel that I should proceed with my first proposition
or should we go for the second. If we need the second then I need some
info on how what is required. For the moment I will work on the low
level routines to initialise the card etc that will be needed in either
case.
Another alternative would be to use something based on the clarkson
drivers. I'm using those routines to get the low level code in any case.
Maybe we should consider using their interrupt interface and then we can
build whatever stack we wish on top or several as the case may be. This
may be attractive if we want to develop software to talk to novell.
Dunno really, I'm awaiting your thoughts.
-- Paul Richards at Cardiff university, UK.spedpr@uk.ac.cf.thor Internet: spedpr%thor.cf.ac.uk@nsfnet-relay.ac.uk UUCP: spedpr@thor.cf.UUCP or ...!uunet!mcsun!ukc!cf!thor!spedpr +++