Re: rtVAX300 .. need help..

[Roger Ivie <rivie%ridgenet.net@localhost>]

On Wed, 9 Jan 2013, Holm Tiffe wrote:
> Yea, I've read something from the 4000/90, NetBSD and empty Flash in the
> past, ...seems the same as you described here and someone has changed the
> Mapping of that VS_REGS from 0x2008000 to 0x25c00000 because of exactly
> this issue.. Who was this?

In the shower this morning, I realized why we're talking past each other
here.

People coming from a PC background don't understand how unusual the PC
is in having *absolutely* *every* *thing* nailed down. It is not
normally the case that a wide range of hardware has the same registers
in the same places. It only happened in the PC because PC-DOS was so
extraordinarily weak.

Out of the box, PC-DOS provided no way to do even simple tasks such as
moving the cursor around on the screen. This could only be done by
calling low-level ROM BIOS services that sit *underneath* PC-DOS or by
directly accessing the hardware. So that's what the PC software folks
did.

Consequently, if you wanted to build a box to sell to people running PC
software, it had to be *absolutely* compatible with whatever schlock IBM
decided to put in, because all the software that people wanted to run had to
make up for the deficiencies of PC-DOS by going around the operating
system to talk directly to the hardware.

This is not the way things work in *any* other segment of the industry.
Normally, software does its thing by using services offered by the
operating system. The operating system takes care of the details of
talking to the hardware.

The ARM Linux crowd occasionally complains about how little
standardization there is in the ARM world. Well, outside of the PC's
small corner of the industry, that is how *everything* works. And this
is a good thing, since you aren't stuck with the mistakes made decades
ago.
--
roger ivie
rivie%ridgenet.net@localhost