Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 17 Sep 2003 11:38:45 +0200 (CEST)
From:      Soren Schmidt <sos@spider.deepcore.dk>
To:        =?ISO-8859-1?Q?Dag-Erling_Sm=F8rgrav?= <des@des.no>
Cc:        current@FreeBSD.ORG
Subject:   Re: ATAng hangs with kernel from september 15
Message-ID:  <200309170938.h8H9cjqL010444@spider.deepcore.dk>
In-Reply-To: <xzpr82fg3sm.fsf@dwp.des.no>

next in thread | previous in thread | raw e-mail | index | archive | help
It seems Dag-Erling Smørgrav wrote:
> Soren Schmidt <sos@spider.deepcore.dk> writes:
> > Well, the ATA driver has just grown more standard compliant :)
> > You *must* hang around for 31secs to wait for slow devices to come ready,
> > according to the ATA specs. Now I've gone to great length before to
> > get around this by using clever heuristics, and I'm getting there again,
> > but there are *so* many crappy devices out there that it takes time
> > to accomodate them all. 
> 
> Is there any way you can postpone the device initialization so you can
> do them in paralell?

That wont help anything here, this is pre device init stuff..

> Or make the length of the wait configurable, like SCSI_DELAY?

That would be a gross hack, the spec says to wait 31 secs and thats it,
if you want to wait shorter go ahead and change your local sources, but we
need to find a real solution (and I will get there, I just need enough
datapoints to find the right solution)...

-Søren



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200309170938.h8H9cjqL010444>