Re: [parisc-linux] Non-HP PCI cards & x86 BOOT-ROMS

[Jan-Benedict Glaw <jbglaw@lug-owl.de>]

[-- Attachment #1: Type: text/plain, Size: 1428 bytes --]

On Fri, 2004-01-16 11:12:19 -0500, Carlos O'Donell <carlos@baldric.uwo.ca>
wrote in message <20040116161219.GT5550@systemhalted>:
> On Fri, Jan 16, 2004 at 12:59:40PM +0100, Jan-Benedict Glaw wrote:
> > Thinking again about that, it maybe isn't all *that* hard. Bochs already
> > does know about the concept of shared DRAM and ROM BIOSes etc. Maybe
> > it's really just enough to hack it's memory class (yup, it's written in
> > C++) to not back the memory by RAM or on-disk storage, but to redirect
> > all accesses to physical RAM.
> > 
> > The same is to be done with I/O ports too, of course... For complex card
> > (SCSI RAID host adaptors), you may need to service interrupts, too:-(
> 
> How can you use Bochs? It runs a virtualized machine. The purpose of
> running card firmware is to effect the *current* machine with some
> configuration process.

Boch is vitual. *But* it actually got quite real objects to back all the
physical things (RAM, ROM, etc). If you hack these containers to roll
their changes back into *real* RAM/ROM/whatever, you can affect quite
real hardware.

MfG, JBG

-- 
   Jan-Benedict Glaw       jbglaw@lug-owl.de    . +49-172-7608481
   "Eine Freie Meinung in  einem Freien Kopf    | Gegen Zensur | Gegen Krieg
    fuer einen Freien Staat voll Freier Bürger" | im Internet! |   im Irak!
   ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]