Subject: Re: wi0: record read mismatch, rid=fd42, got=fd41
To: None <current-users@NetBSD.org>
From: David Young <email@example.com>
Date: 09/27/2004 21:51:56
On Tue, Sep 28, 2004 at 01:58:55AM +0000, Charles M. Hannum wrote:
> On Tuesday 28 September 2004 00:50, David Young wrote:
> > On Mon, Sep 27, 2004 at 03:48:44PM +0200, Kurt Schreiner wrote:
> > > Sep 27 15:13:59 imega-bsd /netbsd: wi0: record read mismatch, rid=fd42,
> > > got=fd41 Sep 27 15:13:59 imega-bsd /netbsd: wi0: record read mismatch,
> > > rid=fdc1, got=fd42 Sep 27 15:13:59 imega-bsd /netbsd: panic: wi0: invalid
> > > channel 49170 Sep 27 15:14:00 imega-bsd /netbsd:
> > > Sep 27 15:14:00 imega-bsd /netbsd: syncing disks... wi0: record read
> > > mismatch, rid=fd42, got=fdc1 Sep 27 15:14:00 imega-bsd /netbsd: wi0:
> > > record read mismatch, rid=fdc1, got=fd42 Sep 27 15:14:00 imega-bsd
> > > /netbsd: panic: wi0: invalid channel 0
> > Sometimes I see this, too. Does it happen very often?
> This is almost certainly caused by me not getting around to fixing the race
> condition between user-level code and the "new & improved" transmit path.
I don't know. I have been seeing Lucent cards do this for a very
long time. Maybe it happens more often, now.
wi should check error conditions, anyway.
David Young OJC Technologies
firstname.lastname@example.org Urbana, IL * (217) 278-3933