Port-i386 archive

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

Re: wm(4) problems




To: Manuel Bouyer <bouyer%antioche.eu.org@localhost>

Subject: Re: wm(4) problems

From: Thor Lancelot Simon <tls%rek.tjls.com@localhost>

Date: Sun, 13 Jan 2008 15:09:39 -0500


On Sun, Jan 13, 2008 at 12:51:19PM +0100, Manuel Bouyer wrote:
> On Sat, Jan 12, 2008 at 07:57:09PM -0800, Brian Buhrow wrote:
> >     Is it possible you have a bad ethernet cable?  That kind of error
> > sounds like either a bad cable or a bad port on the switch.  I'm not saying
> > it's not a driver bug, but if it's generating CRC errors on the other side,
> > that's sounding very hardwareish!
> 
> I suspect CRC errors can be generated if the PHY isn't set up properly.

Yes, and there's something terribly wrong with at least one of our PHY
drivers for this chip.  Since we (still, AFAICT) do not have a dmesg
from the system on which this particular problem is reported, we can't
actually debug this much further, I'm afraid -- we can't see which PHY
is present and which PHY driver is in use.

On newer wm hardware with -current drivers, it sometimes helps to disable
ikphy and use ukphy, which appears to work correctly in many cases where
ikphy does not.

Thor



Follow-Ups:

Re: wm(4) problems
From: Rich Neswold


References:

Re: wm(4) problems
From: Brian Buhrow

Re: wm(4) problems
From: Manuel Bouyer




Prev by Date: Re: MATH_EMULATE

Next by Date: Re: MATH_EMULATE

Previous by Thread: Re: wm(4) problems

Next by Thread: Re: wm(4) problems

Indexes:

reverse Date

reverse Thread

Old Index



Home | Main Index | Thread Index | Old Index