i386/109200: [ata] READ_UDMA UDMA ICRC error cause notdetecting cable type

[Joris Dobbelsteen]

The following reply was made to PR i386/109200; it has been noted by GNATS.

From: "Joris Dobbelsteen" <Joris at familiedobbelsteen.nl>
To: "Gavin Atkinson" <gavin.atkinson at ury.york.ac.uk>
Cc: <bug-followup at freebsd.org>
Subject: RE: i386/109200: [ata] READ_UDMA UDMA ICRC error cause notdetecting cable type
Date: Fri, 25 May 2007 13:31:41 +0200

 >-----Original Message-----
 >From: Gavin Atkinson [mailto:gavin.atkinson at ury.york.ac.uk]=20
 >Sent: vrijdag 25 mei 2007 13:18
 >To: Joris Dobbelsteen; bug-followup at freebsd.org
 >Subject: Re: i386/109200: [ata] READ_UDMA UDMA ICRC error=20
 >cause notdetecting cable type
 >
 >Hi,
 >
 >Can you please provide a full verbose dmesg for this system? =20
 
 Sorry, at this point FreeBSD was dropped. And the system is mostly
 inoperational. Will perhaps try later on. Will also try to lookup
 componentes.
 
 The bios string is 09/27/2000-693A-686-6a6LGA1AC-UX
 It is a VIA chipset.
 
 Unfortunally cannot corfirm the brand, but most likely either abit,
 chaintech, gigabyte, asus or msi.
 
 >Also, is it possible to confirm that the IDE cable has been=20
 >connected the correct way around (e.g. the (usually blue or=20
 >red) connector is connected to the motherboard, and the=20
 >furthest (usually black) connector is plugged into the master=20
 >device)? =20
 
 The issue is not the cable in any way. It operates under both Windows
 (for years) and Linux without problems.
 The issue is that FreeBSD uses UDMA66 mode. However the cable is NOT an
 80-conductor and hence interference causes a lot of CRC errors. The
 system should detect whether a 80-conductor cable is, or is not, in
 place and possiblity limit the speed to UDMA33.
 
 Joris Dobbelsteen B.ICT.