On Oct 31, OKUJI Yoshinori wrote:
> From: "Jochen Hoenicke" <[EMAIL PROTECTED]>
> Subject: EZ-BIOS support
> Date: Mon, 30 Oct 2000 12:12:03 +0100 (MET)
>
> > The attached patch will detect if EZ-BIOS is present, and refuses to
> > embed grub after the MBR in that case. The patch will look for a
> > signature at the begin of the EZ-BIOS code. Since the remapping is a
> > bit inpredictable for the first few sectors, it will look for this
> > signature in two places.
>
> I don't like your patch, since it just avoids totally breaking
> ez-bios; it's not real support.
At least it prevents users from loosing their partition table. But if
they are unlucky, they have to install ez-bios again to access their
disks under DOS/Windows.
> As you know, this isn't so easy. If
> you really want to support ez-bios, then you will have to virtualize
> all the accesses to partition tables.
Normally ez-bios or linux should do the remapping of the partition
table. Only in the case where grub is booted from floppy, without
going via ez-bios the remapping isn't done. This case is of course
not recommended.
Is it enough if grub just refuses to access a disk where it can see
the ez-bios partition in the MBR (which means that ez-bios wasn't
loaded) or should grub do the mapping itself in that case?
The first would be much easier to add and it would be safe: The worst
that can happen is that grub refuses to install itself with a
comprehendable error message.
This is all theory, I haven't played with EZ-BIOS enough to be sure.
Jochen
_______________________________________________
Bug-grub mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/bug-grub
