ad0: TIMEOUT - READ_DMA retrying (2 retries left) LBA=207594611

[Kevin Oberman]

> Date: Fri, 17 Sep 2004 04:36:53 +0800
> From: Ariff Abdullah <skywizard at MyBSD.org.my>
> 
> On Thu, 16 Sep 2004 13:13:42 -0700
> "Kevin Oberman" <oberman at es.net> wrote:
> > > From: "Daniel Eriksson" <daniel_k_eriksson at telia.com>
> > > Date: Thu, 16 Sep 2004 21:17:45 +0200
> > > Sender: owner-freebsd-current at freebsd.org
> > > 
> > > 
> > > This is a me-too report:
> > > 
> > > After upgrading a 6-CURRENT kernel and world from
> > > 2004.09.09.08.00.00 to 2004.09.16.13.00.00, I am now getting these
> > > messages on a machine that previously worked just fine:
> > 
> > Thanks! This should greatly simplify my search for the change that
> > is causing the problem. I can start with 9/9/04 and work forward. I
> > saw the problem on 9/11/04 in RELENG_5, so I now have a window. I
> > could shrink it to less than nothing if I was REALLY confident that
> > nothing was MT5 in less than 3 days, so something the was moved to
> > RELENG_5 after a very short time is suspect.
> > 
> > I am currently building a 9/9/04 RELENG_5 kernel and I'll test it
> > later today (if nobody beats me to it.)
> 
> One more thing to consider, the default scheduler. I found that those
> errors occured with SCHED_4BSD (PREEMPTION or NOT), while SCHED_ULE
> (of course without PREEMPTION, or *else*), nothing such that.
> 
> Note that this is *my* case, your mileage may vary.

I can confirm I'm using 4BSD.

If anyone is using ULE and seeing this EXACT problem, please holler?
adX: TIMEOUT - READ_DMA retrying (2 retries left) LBA= or
adX: TIMEOUT - WRITE_DMA retrying (2 retries left) LBA=
It should be differing only in the ad unit number and the specified LBA.

-- 
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at es.net			Phone: +1 510 486-8634