2bfeeed885
. all invocations were S or D, so can safely be dropped to prepare for the segmentless world . still assign D to the SCP_SEG field in the message to make previous kernels usable |
||
---|---|---|
.. | ||
3c501.c | ||
3c501.h | ||
3c503.c | ||
3c503.h | ||
3c509.c | ||
3c509.h | ||
8390.c | ||
8390.h | ||
devio.c | ||
dp.c | ||
dp.h | ||
Makefile | ||
ne.c | ||
ne.h | ||
netbuff.c | ||
README | ||
wd.c | ||
wd.h |
This is my implementation of a new network task for the Minix kernel. I did it initially to handle a 3c501 board (Etherlink), but those board are so unstable that it is not worth using them except for learning how to implement a driver. When I got a 3c509b board (Etherlink III) it was easier to write the code to handle them. The Minix code in 'dp8390.c' is too specific for the National chip set, so what I did was to remove as much as I needed of the code dependant from the chip and produce a generic task that, I hope, will be able to handle many more cards. $Log$ Revision 1.1 2005/06/29 10:16:46 beng Import of dpeth 3c501/3c509b/.. ethernet driver by Giovanni Falzoni <fgalzoni@inwind.it>. Revision 1.3 2004/04/14 12:49:07 lsodgf0 Changes for porting to Minix 2.0.4 run on BOCHS Revision 1.2 2002/03/25 14:16:09 lsodgf0 The driver for the NEx000 has been rewritten to be operational with the ACCTON 18xx (an NE1000 clone) The I/O routines for 16 bit cards are still untested.. Revision 1.1 2002/02/09 09:35:09 lsodgf0 Initial revision The package is not fully tested, i.e. I had only 3Com boards (3c501, 3c503, 3c503/16 and 3c509b) and WD8003. I got also a NE1000 clone but it was not fully operational and I could not appreciate the results. For this reason the changes done to the interface to I/O for 8 and 16 bits are not tested.