Subject: New Tulip driver committed
To: None <current-users@netbsd.org>
From: Jason Thorpe <thorpej@nas.nasa.gov>
List: current-users
Date: 08/31/1999 17:41:08
Hi folks...

I've committed my new Tulip Ethernet chip driver to -current, which
currently has support for the Winbond 89C840F clone, and (possibly
broken) support for the Lite-On 82C168/82C169 "PNIC", like on the
new NetGear and Linksys boards (my NetGear board may be busted).

Support for the Macronix and ASIX clones should be coming soon.

Anyway, I'm running this with:
.
.
.
tlp0 at pci0 dev 8 function 0: Winbond 89C840F Ethernet, pass 0.0
tlp0: interrupting at kn20aa irq 16
tlp0: Ethernet address 00:80:48:d7:3a:75
ukphy0 at tlp0 phy 1: Generic IEEE 802.3u media interface
ukphy0: OUI 0x00606e, model 0x0000, rev. 2
ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
.
.
.
tlp1 at pci0 dev 12 function 0: Lite-On 82C169 Ethernet, pass 2.1
tlp1: interrupting at kn20aa irq 4
tlp1: Ethernet address a0:00:3c:cc:3b:43
lxtphy0 at tlp1 phy 1: LXT970 10/100 media interface, rev. 0
lxtphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
.
.
.
The Windbond board works apparently flawlessly, where as the Lite-On board
has a constant barrage of DMA underruns and visibly smashed packets that
appear on the wire.  Receive filter setup on the Lite-On seems busted,
too... but since it uses the DMA engine to perform that function... well,
it's all a little suspect to me.  I'd like to hear from others what their
experiences are with this board...

Have fun!

        -- Jason R. Thorpe <thorpej@nas.nasa.gov>