Hi folks...

So, as some of you may know, I am writing a new Tulip driver from
scratch, partially to address maintenance problems with the old
driver, partially to add support for the Tulip clone chips that
are becoming available (sadly, in lieu of geniune Tulips).

In any case, this is a teaser, to let you know that I might be
soliciting testers soon, _ESPECIALLY_ for geniune Tulip boards:

tlp0 at pci0 dev 12 function 0: Lite-On 82C169 Ethernet, pass 2.1
tlp0: interrupting at kn20aa irq 4
tlp0: Ethernet address a0:00:3c:cc:3b:43
ukphy0 at tlp0 phy 1: Generic IEEE 802.3u media interface
ukphy0: OUI 0x1e0400, model 0x0000, rev. 0
ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto

That's my NetGear FA-310TX in an AlphaStation 500.

I haven't done much more testing that a boot, so far... I can't stay
at work late tonight to do that :-)  I'll make sure it's at least
minimally functional and commit it tomorrow.

Note, this will only handle PNICs w/ MII for now!  My next goal will be
to get the Macronix chips working (should be ~trivial in this new driver).
For that, I'll definitely need testers, as I only have one of the several
Macronix chips, one without MII (only internal NWAY).

After the Macronix chips, I suspect I'll tackle the 21142/21143.  It's
known to be flaky on Digital Personal Workstations w/ the "de" driver,
and it needs to be in front-end/back-end style for the forthcoming
CardBus code.

Oh yes... if anyone out there has a PNIC-based card that does NOT
have PHY on the board, PLEASE CONTACT ME!  I have no way of testing
the PNIC's internal NWAY transciever, as the only boards I can find
(NetGear and Linksys) all have PHYs on them.

Note, the while support for more chips migrates to the new driver from
the old, it will be possible to have BOTH in your kernel, if you have
multiple types of boards.  The new "tlp" driver matches supported chips
with a higher priority, whereas the "de" driver will then get chips not
yet supported by "tlp".

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