Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 27 Dec 2001 15:35:37 +0100 (CET)
From:      Søren Schmidt <sos@freebsd.dk>
To:        sos@freebsd.dk
Cc:        Pete French <pfrench@firstcallgroup.co.uk>, freebsd-hackers@FreeBSD.ORG, freebsd-stable@FreeBSD.ORG
Subject:   Re: Need PCI/VIA chipset help (was Re: 4.4-STABLE crashes - suspects new ata-driver over wd-drivers)
Message-ID:  <200112271435.fBREZbm68994@freebsd.dk>
In-Reply-To: <200112271415.fBREFFl64118@freebsd.dk>

next in thread | previous in thread | raw e-mail | index | archive | help
It seems Søren Schmidt wrote:
> > Does this mean that if you dont have the ATA drivers in the kernel then
> > you wont get the fix ? As the bug occurs under high PCI loads then this
> > can potentially affect people with large amounts of PCI activity on
> > SCSI drives too surely ? Granted I've never seen it happen, but it still
> > bothers me that it might...
> 
> As far as I'm able to tell, the problem can appear with any PCI device 
> that can generate enough PCI traffic for the bug to appear, or
> at least my tests points in that direction. However I dont have 
> any SCSI equipment here so I cant tell if that can trigger the
> problem...

Hmm, just flipped through some of my notes on this, to be precise I
can *not* say that any PCI device can cause this problem, all cases I
have has the 686b device as part of the equation, so what we need to
show is that there is a problem (or not) when not using the 686b
part at all (ie no ATA channels, no VIA sound, no USB traffic and 
possbily more I've forgotten the 686b handles). 
If corruption can still appear the fix is "generic".

Mind you the fix VIA originally brought out was just to cope with
data corruption when using a SB live! and the onboard ATA channels
simultaniously. Intensive testing then revealed that the problem
could also appear without the SBlive! in the mix, but only banging
on the ATA channels (a simple 'dd' between two fast disks one on each
channel reveals the bug within seconds). I then did alot of testing
using various means of generating lots of PCI traffic, but I can
only say from the results I saved that the corruption appeared
when the 686b was involved, I dont have any significant data that
shows that high PCI loads without the 686b involved fails/works.

-Søren

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200112271435.fBREZbm68994>