Re: port-i386/46247: 6.0 Beta boot fails on generic i386 architecture machine

[David Laight <david%l8s.co.uk@localhost>]

The following reply was made to PR port-i386/46247; it has been noted by GNATS.

From: David Laight <david%l8s.co.uk@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: port-i386/46247: 6.0 Beta boot fails on generic i386 architecture 
machine
Date: Mon, 26 Mar 2012 19:27:07 +0100

 On Mon, Mar 26, 2012 at 05:40:04PM +0000, David Holland wrote:
 > The following reply was made to PR port-i386/46247; it has been noted by 
 > GNATS.
 > 
 > From: David Holland <dholland-bugs%netbsd.org@localhost>
 > To: gnats-bugs%NetBSD.org@localhost
 > Cc: 
 > Subject: Re: port-i386/46247: 6.0 Beta boot fails on generic i386
 >  architecture machine
 > Date: Mon, 26 Mar 2012 17:35:42 +0000
 > 
 >  On Fri, Mar 23, 2012 at 08:00:07AM +0000, David Laight wrote:
 >   >  > NetBSD MBR boot
 >   >  > NetBSD/x86 ffsv2 Primary Bootstrap
 >   >  > [silence]
 >   >  > 
 >   >  > The install complained once that the disk geometry could not be 
 > uniquely matched:
 >   >  > Bios: 1024/16/63
 >   >  > Prob: 1023/255/63
 >   >  
 >   > If it got as far as showing the ffs2 message then the chs values
 >   > won't matter.
 >  
 >  Are you sure of that? It is still doing I/O via the BIOS until the
 >  kernel gets started; while I'd expect it to be using LBA for that, the
 >  entire problem could be that it isn't.
 
 It will be doing bios requests - probably as CHS requests, but the
 data it has requests a logical sector number which is converted
 to CHS using the parameters obtained from the bios that boot.
 Those ought to match!
 'sysctl machdep.diskinfo' will show the bios geometries.
 
 The mbr will read the pbr using the CHS values saved by fdisk.
 If fdisk were run on a different system they are very likely to
 be wrong - which is when mbr_lba can help somewhat.
 
        David
 
 -- 
 David Laight: david%l8s.co.uk@localhost