Re: Cobalt Qube
On Wed, Dec 02, 1998 at 10:21:07AM -0500, Daniel Jacobowitz wrote:
> On Wed, Dec 02, 1998 at 04:17:45PM +0100, Sven LUTHER wrote:
> > On Wed, Dec 02, 1998 at 10:11:15AM -0500, Daniel Jacobowitz wrote:
> > > On Wed, Dec 02, 1998 at 04:01:23PM +0100, Sven LUTHER wrote:
> > > > The Powerpc can do that also, but i think a particular endianness is chosen
> > > > either in the hardware (hardwired, or hardware switch) or by the OS. in the
> > > > powerpc it is possible to change endiannes my changing a value in the control
> > > > register, but i don't think anyone has yet taken advantage of it, for example
> > > > to use little endian only drivers in a big endian setting.
> > >
> > > As I understand it, the processor will gladly operate in another
> > > endianness - but your firmware may object strenuously. There's really
> > > no adequate way I can see to take advantage of this, except for
> > > bragging points.
> > what firmware ? what has it to do with it ? it could possibly work on non mac
> > hardware then ? I thought the main point was that it would probably confuse a
> > lot of things, and that in the long end it would probably be easier to rewrite
> > the driver to be endianess aware than to play with the endianness issue ?
> Every machine has some sort of boot firmware. This is (am I out on a
> limb here?) probably not endian-independant! So if you switch the
> processor's endianness, it may simply fail to boot. And yes, fixing
> drivers is MUCH easier.
But you can always swap endianess after the booting is done, you can even have
a different endianness during interrupts than during normal handling, or just
change it as you need it. so there should be no problem with the firmware if
said firmware is only used at boot time. (in my case the i am bootstrapping
from m68k amigaos, so endianess is the less of my problems :))