|
Hello Paul,
Paul Lantinga wrote:
only one bug? ;)
Jeff Garzik is working very closely with Intel and they just released
some specs and excellent code for the e1000 and the eepro100 card.
There's a light at the end of the tunnel, if you find the way out.
There's a patch for bonding.c and ifenslave.c for the 2.4.18, but I
didn't see the eepro100 patch. I'll go take a look and pray that it's
*not* my version. ;)
What patch do you refer to?
That reminds me of another question: what are people using for the
Intel Pro card driver, the eepro100.o linux community driver or the
e100.o Intel driver?
I strongly suggest you use the eepro100.c driver and report failures if
you can so people working on it can address the problems. As I mentioned
before, there is a WIP between Jeff and Intel to merge eepro100.c and
e100.c. Stay tuned ;)
Included with the e100 driver, Intel includes the source for iANS, which
is their Advanced Networking Services. In short, it's their software
for doing
trunking/bonding/etherchannel/teaming/wacking/schlepping/whatever-you-wa
nt-to-call-it-today-ing.
Which fails horribly!
btw, if anyone out there is using broadcom gig interfaces, the broadcom
driver author was good enough to send me the recent version of their
driver source. The only drivers that you'll find out 'there' are old
and crusty. Lemme know and I'll fork it over.
:)
Cheers,
Roberto Nibali, ratz
| 