[EMAIL PROTECTED] wrote:
> If this is true then you might be able to connect A18 (I think that
> is the right one) from the ISA bus to the correct pin on the jumper to
> directly access all 512K.
>
> I have a Gigabyte board but it doesn't have this feature. To run a
> 512K BIOS we made a little socket adaptor that let us intercept the A18
> pin of the flash and connect it to A18 of the ISA bus with a jumper.
>
> I think I previously heard that someone on this list had made this sort
> of thing easier. ...it might have had a latch on it so you could
> control the upper address bits from software and didn't need the
> jumper. ...though you would also only get 256K at a time with this
> method.
>
The A18 method, along with programming south bridge might make
this seamless (ie 512k Flash gets mapped from 0x80000 to 0xfffff).
Intex PIIX3 and up have various small defaults for BIOS_CS's
decode range, but can be programmed to decode 512k at 0x80000.
if enough address lines are connected it might be real nice.
Esp. if your video card is set to use linear frame buffer, and doesn't
map 0xa0000-0xc0000 etc.
>
> Ty
>
> On 8 Mar, Joe Cooper wrote:
> > If it's like the AOpen boards that have dual BIOS storage, it is 'hard'
> > switchable via a jumper switch. Meaning both are not simultaneously
> > accessible or usable, only one at a time.
> >
> > Alain Greppin wrote:
> >
> >> Hello,
> >>
> >> Has anyone some information about the DualBIOS feature
> >> on the Gigabyte boards ?
> >>
> >> Do you think it is possible to load linuxbios
> >> and a minimal kernel in this DualBIOS wich is 2 x 256k ?
> >>
> >> And, in general, when the BIOS is flashed with linuxbios,
> >> how can I reflash it to upgrade ?
> >>
> >> What i want, is to build a cluster of diskless Linux boxes.
> >> What is best for this case ? a) linuxbios b) nic + boot rom
> >>
