Re: Nvidia MCP67 SATA controller

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Sun, Jun 22, 2008 at 06:01:14PM +0000, David Holland wrote:
> On Wed, Jun 18, 2008 at 07:30:41PM +0200, Dieter Baron wrote:
>  > the Nvidia MCP67 SATA controller does not work using the viaide driver
>  > (in native pci compatibility mode), as various users have reported.
> 
> MCP65 should also use this patch (I have one, same behavior), and
> probably all the later nvidia versions as well.
> 
>  >   This works quite well most of the time, but sometimes the chip seems
>  > to get in a strange state after the switch.  It then returns bogus
>  > data when identifying the drive:
>  >    wd0 at atabus1 drive 0: <ST506>
>  > (or similar, written from memory) instead of 
>  >    wd0 at atabus1 drive 0: <Hitachi HDT725032VLA360>
>  > and hangs when trying to read from the disk (still during boot).
> 
> IME, it will also probably probe a ridiculously small drive size. I
> believe this is a known issue with ahcisata.c, not limited to nvidia
> cards, and so it's not something you've triggered.

AFAIK it's a timing issue with some drives. The PHY goes UP after a reset
and reports the drive is ready, while it's not.

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--