On Sat, Apr 15, 2006 at 09:52:13PM +0200, Jurjen Oskam wrote: > On Sat, Apr 15, 2006 at 04:57:38PM +0200, Srebrenko Sehic wrote: > > > That's weird, since nfe(4) worked fine in the X2100 Sun box I tested a > > while ago. That was pre-3.9. From X2100 dmesg: > > Now that you mention it, I did try a 3.8-current on that machine. nfe0 > worked, but when booting back to Windows the card wouldn't work in > Windows. It started working after I did something which I cannot > remember, but a hard reset wasn't it. > > > Could you try with an MP kernel? > > Hmm, that one doesn't even boot: > > <lots of dmesg I couldn't capture> > wd0(pciide1:0:0): timeout > type: ata > c_bcount: 512 > c_skip: 0 > wd0(pciide1:0:0): timeout > type: ata > c_bcount: 512 > c_skip: 0
This is an IRQ routing problem. Should be fixed in -current. > > I tried booting the i386 install CD, but that didn't boot either: > > <lots of dmesg again> > bios0: ROM list: 0xc0000/0xee00! 0xd0000/0x4000! > > > On Sun, Apr 16, 2006 at 01:03:22AM +1000, Jonathan Gray wrote: > > > > nfe0: tx v2 error 0x6004 > > > nfe0: tx v2 error 0x6204 > > > You're on 10baseT Ethernet right? > > Nope, a standard 10/100 switched Ethernet. > > I did "dhclient -d nfe0" again just now, and now the first error was: > > nfe0: tx v2 error 0x6100 I think this means 0x4000 - tx error 0x2000 - last packet 0x0100 - ??? > > ... after which 0x6204 was again repeated for (almost) each DHCP query. That makes sense as the breakdown is: 0x4000 - tx error 0x2000 - last packet 0x0200 - deferred 0x0004 - retry error -current has an additional fix which may have some chance of changing this behaviour. It is hard to suggest a solution without having first figured out the problem.
