On Tue, 25 Jul 2000, R Joseph Wright wrote:
> On Tue, Jul 25, 2000 at 03:29:08PM -0600, Fred Clift wrote:
> > On Sun, 23 Jul 2000, John Baldwin wrote:
> >
> > > the geometry of a disk. At the very least, dangerously dedicated mode
> > > should specify a valid length for the slice the way that truly dedicated
> Would you mind posting that script? I'd like to see it.
I'm told that at least for some people that disklabel auto actually does
this as recently as 4.0-RELEASE -- I've not had a chance to verify this
but I know that using the da driver in 3.4 that this doesn't happen so I
do approximately the following (from a C program, that runs from a
pxe-like boot-rom booted kernel) note that only the important bits are
shown and some pseudo-code for clarity
fo = fopen(.......
...
fprintf(fo, "\n\n\n\ny\n\n\n%d\n\ny\n\ny\n",driveInfo[disk].size);
fclose(fo);
/* driveInfo[disk].size is the number of sectors/unit as reported by
disklabel */
/* and then this is done.... */
/sbin/fdisk -u da0 < /tmp/fdisk.in
This gets the right numbers in the disk label for me to make my Intel
ISP2150 boxes to work right (LB440GX+ based I believe).
An interesting side note is that if I dd the stock boot0 (with the bogus
partition table entry) onto _any_ disk in the system, regardless of what
disk is the boot disk, then the machine wont boot at all, not a floppy,
not any of the boot devices, nothing. I have to either low-level scsi
format the disk or boot from our custom bootrom and nuke it from
there... I would say that any machine that can be crippled by having
corrupted data on a non-boot disk is by design broken. But, thats another
discussion.
Fred
--
Fred Clift - [EMAIL PROTECTED] -- Remember: If brute
force doesn't work, you're just not using enough.
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-stable" in the body of the message
