[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: if_dc support for 21140

Hi Chris,

Here are my results. They might not be determinent though.

On 05-Jan-00 Chris Cappuccio wrote:
> Would anyone with 21140 PCI cards mind testing the patch below? (needs
> current...Make sure dc is in your kernel config, and de isn't.)
> This makes my Znyx ZX346 (with 21140A chips and ICS1890 PHYs) work nicely,
> even with mii selections (instead of almost not at all as with the buggy
> if_de driver)
> (The MX98713A, already supported by if_dc, appears to be a very close clone
> of the 21140.)
> ...

My setup consists of some cards, one is a Cogent EM-110TX with a DEC-21140-AB
chip. Some other chips on the card are a Motorola XC68836-FNB, a Seeq NQ80C24
and a National Semi DP83223 (I don't know which one is the MII, if there is
one) .

dmesg before (de)
de0 at pci0 dev 9 function 0 "DEC DECchip 21140 (FasterNet)" rev 0x12: irq 5
de0: Cogent 21140 [10-100Mb/s] pass 1.2 address 00:00:92:96:0e:d2
de0: enabling 10baseT port

ifconfig before
        media: Ethernet 10baseT
        status: active
        inet6 fe80:3::200:92ff:fe96:ed2 prefixlen 64

dmesg after (dc)
dc0 at pci0 dev 9 function 0 "DEC DECchip 21140 (FasterNet)" rev 0x12: irq 5
address 00:00:92:96:0e:d2

ifconfig after
        media: Ethernet none (none)
        inet6 fe80:1::200:92ff:fe96:ed2 prefixlen 64

It's like the configuration of the card is less precise with the dc driver...

While the card didn't even work with the de driver (with or without inet6), it
still doesn't work with the dc driver. Though I remember the card was working
with both WinNT and OpenStep, the card may have a failure now (I might retest
on a WinNT) .

Does it have something to do with the PHY negociation ?

E-Mail: Thierry Deval <TDeval@PrimeOBJ.COM>
Date: 08-Jan-00
Time: 14:00:45

         Prime Objective
   OpenBSD - Linux - Windows NT
 OO Development - Network Systems