home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.linux
- Path: sparky!uunet!paladin.american.edu!darwin.sura.net!jvnc.net!yale.edu!qt.cs.utexas.edu!cs.utexas.edu!torn!cunews!revcan!quantum!aboyd
- From: aboyd@quantum.on.ca (Andrew Boyd)
- Subject: Re: Ethernet drivers
- Message-ID: <#nmnf6#@quantum.on.ca>
- Date: Fri, 04 Sep 92 12:55:40 GMT
- Organization: Quantum Software Systems
- References: <GNAT.92Sep1191408@kauri.kauri.vuw.ac.nz> <715399538snx@crynwr.com>
- Lines: 35
-
- In article <715399538snx@crynwr.com> nelson@crynwr.com (Russell Nelson)
- writes:
- >In article <GNAT.92Sep1191408@kauri.kauri.vuw.ac.nz> gnat@kauri.vuw.ac.nz
- >writes:
- > In article <#xdngk+@quantum.on.ca> aboyd@quantum.on.ca (Andrew Boyd):
- > > Actually, the 503 is very similar to the wd (now smc) 8003 series of
- > > cards - both use the 8390 NIC. If you folks already have a driver for
- > > the 8003, you won't find it too big a job to support the 503 - just
- > > different locate and init code. The runtime code is identical.
- >
- >Well, that was WRONG. First of all, WD has their own silicon, which
- >has different bugs than the National silicon.
-
- um, it's not *that* different. And until recently, WD cards were shipped
- with nat semi 8390's installed. Check _your_ board - unless it's pretty
- new, it won't have a WD83C690 installed.
-
- >Plus, the 3c503 when I/O mapped has a 16 byte FIFO that you need to
- >ensure is filled.
-
- I didn't say anything about I/O mapped mode - excuse me for forgetting
- to mention memory mapped mode, in which case you can use the same runtime
- code for the 503 & the 8003.
-
- >Plus, the wd8013e has a 16-bit interface that decodes its address
- >early and so needs to be read/written to with interrupts off.
-
- um, I was only talking about the 8003 (8-bit) NOT the 8013 (16-bit) card.
- Reread my posting.
-
- Excuse me for making the mistake of trying to help - I won't do it again.
-
- ----
- #include <std.disclaimer>
-
-