lkml.org 
[lkml]   [1999]   [Sep]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: IDE Driver in Linux Reports different geometry that Netware during Volume and Partition Create, causing loss of user data under Linux
Date
Andries,

I changed the drive setting to "AUTO" in the CMOS for this drive, and now
the drive reports 850/63/64. It is interesting to note that there was no
setting for this disk device that matched any of the configurations you gave
me. Long ago I wrote an IDE driver on Netware (Wolf Mountain), so I am
aware of the ability to translate drives into other logical geometries and
am very famliar with low level IDE stuff (unfortunately). The one thing
that sucks about IDE and ATA is that both use IO port reads and writes to
transfer data to and from the disk (which sucks up a lot of bus cycles).
It's better, however, to get this type of info to the right folks who
maintain the driver. There are several IDE drivers out there on other OS's
that for some reason did not implement the new commands for EIDE when they
came out. I try to focus on the stuff I am supposed to be working on and
not trying to fix everyone else's bugs, but report then to them instead. I
did not know if there were underlying hardware issues that I was unaware of
since it's been about two years since I did any IDE work. I thought someone
else may be more recently on top of it (like you).

Thanks for the help. I am relieved to know that the Linux IDE drivers are
not busted and don''t have any cross platform issues that will cuase
different OS's to report different geometry.

Thanks,

Jeff

----- Original Message -----
From: Guest section DW <dwguest@win.tue.nl>
To: <dwguest@win.tue.nl>; <jmerkey@timpanogas.com>;
<linux-kernel@vger.rutgers.edu>
Sent: Monday, September 13, 1999 4:37 PM
Subject: Re: IDE Driver in Linux Reports different geometry that Netware
during Volume and Partition Create, causing loss of user data under Linux


> > > With a MAXTOR 81750-A4 disk device, Netware reports that this disk
> > > has a geometry of 850 cylinders, 64 heads, and 63 sectors per
track.
> >
> > That sounds correct.
> >
> > > Under Linux, the same device reports a geometry of 780 cylinders,
> > > 64 heads, and 63 sectors per track.
> >
> > I always repeat the same question: what are the boot messages?
> > And in case you answer you might as well add the output of
hdparm -I.
> > My first conjecture in such cases is user error.
> > For example, the disk may be entered incorrectly in BIOS/CMOS.
>
> From jmerkey@timpanogas.com Mon Sep 13 23:13:41 1999
>
> Boot messages report the wrong geometry as well. These newer drives
use a
> EIDE command to get the correct geometry and typically, at least in
modern
> times, don't rely on the information in CMOS since it can be
incorrect.
>
> Jeff
>
> Ah - you are naive. There is no `correct geometry'.
> But there is a geometry reported by the disk (two geometries, in fact).
> And there is a correct geometry for doing I/O (if LBA is not used).
> And there is a correct geometry for booting, for LILO to use.
> And there is a correct geometry for interaction with other
> operating systems - the one that everybody agrees upon for the
> interpretation of the partition table.
> These geometries differ in general.
>
> The disk here reports 3400 cylinders, 16 heads and 63 sectors per track.
> For the partition table you would have liked 850 cylinders, 64 heads,
> and 63 sectors per track.
> It sounds as if you told the computer in the CMOS setup not to use
> more than 780 cylinders.
>
> If that is the case, then just correct that, and the problem is gone.
> Otherwise I would like to see more precise data.
>
> Andries - aeb@cwi.nl
>


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:53    [W:0.045 / U:0.008 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site