[OT] But important to me. Appropiate Mailing Lists ??

William Case billlinux at rogers.com
Mon Apr 17 01:54:34 UTC 2006


Hi;
Sorry for taking up your time, but you have set me on the trail to a
couple of things I would like to understand.

> > Don't mistake shortcuts used for an example for stupidity.

> OK.

To be more accurate about the BIOS question; one of the things that has
always bothered me -- and it may be a result of stupidity -- is when I
turn on my computer in a cold boot, power first flows to the main BIOS
(read Firmware Hub for Intel) and then to the other auxiliary BIOSes.
E.g. something has to turn on, nudge or elbow the video BIOS, probably
even before POST is begun (I will check the exact timing), so that video
BIOS can start the video card in order for the Monitor to show the LOGO
screen or if interrupted the BIOS setup screen.  The LOGO screen and
setup screen are in main BIOS flash memory -- aren't they?

How does main BIOS, at this early stage, know were the auxiliary BIOSes
are and which ones to elbow into action  -- this is after all before PnP
has kicked in?
> 

> >> how addresses are parsed
> > 
> > I suspect that that was what I was asking about.  I know port numbers
> > have special designations so that the CPU or other devices know where
> > they are.  (It is more complicated than that but you get the idea)
> 
> The unfortunate thing here is that the PCI bus standard is expensive.
> You might glean something by reading the kernel code partaining to the
> PCI bus and accessing it.

Very good suggestion.  I have a manual explaining the kernel code for
2.6.  It should be in there.

> >> (The venerable
> >> old 68000 CPU might help understand the answers to the second batch. It's
> >> command set was very "regular". It's manuals were relatively easy to
> >> understand, as well.) And interpreters simply add an execution step and
> >> remove the linkage step.
> > 
> > It is more complex than that, I am sure.
> 
> The registers can be thought of as tiny customized chunks of memory that
> have their own internal addresses within the CPUs. Their access is coded
> within the instructions and the "microcode" that might exist to interpret
> those instructions. The 68000 made it particularly "obvious".
> 
> The interpreters are very simple, too. Check out a copy of the source for
> the VERY old "TinyBASIC" interpreter. Or find one of the copies for the
> disassembled and recreated source for Commodore BASIC. It takes time to
> get a handle on it. But you can see how it parses and executes rather
> than parses and translates. A BASIC interpreter sometimes includes specific
> lexical passes, translation to an internal code representation, and
> execution, via translation, of the internal code representation. Often the
> translation is a jump table. (I learned by creating my own source for an
> HP BASIC interpreter that was handy. Darned fine code in it, too.)
> 
You know, to get a real answer I am probably going to have to slog my
way through the BASH source code.  Ruefully; I would have thought that
someone would have already done the slogging and left a nice neat
readable explanation for the world.  Oh well.


> Simpler CPUs are easier to grasp at first. That's why textbooks with their
> simplifications can help. Lacking textbooks the 68000 manuals, hardware
> and software, can illuminate a LOT of important concepts.
> 
> {^_^}

Yea, my neighbour has a basement full of 68000 textbooks and manuals.

Pardon the sarcasm.  Its getting late.

Regards Bill




More information about the users mailing list