Threaded code density, 68000, Alpha

[Jaap van Ganswijk <ganswijk@xxxxxxxxx>]

At 07:00 PM 4/1/96 +0200, you wrote:
>
>On Sun, 31 Mar 1996, Eugene Leitl wrote:
>
>> At 11:52 AM 3/25/96 GMT, Andrew Haley <aph@atml.co.uk> wrote:
>> >RISC instructions are often a bit less dense than CISC.
>> 
>> Often more than a factor of two.
>
>What is the relative average density of threaded code
>in comparison to that, does anybody know? (It sure
>bears programmer's fingerprint, agreed).

If you mean with threaded: Code with in-task multitasking,
I don't see why it would have another density than normal
code for that processor... (Could you explain?)

>> >Yes, but the 68000 is a completely broken CISC architecture.  Alpha is
>> >a much better example...
>
>For the time 68k came out (I saw an announcement in 1979) it was
>pretty good: the memories were slow and the assembler programmers
>still numerous. 68020, 68030 and 60040 were good, 68060 much less
>so.

I never liked the un-orthogonality of the 68000 and the two
kinds of registers (data and address) and the unability to use
a displacement of 32 bits...

>Apropos broken CISC architectures: is i80xxx any better?

When you mean, the 80x86's: They are of course far worse!

>> It's very nice indeed. ;-)
>
>Alpha stinks, imnho. It's fast, but for all the wrong
>reasons.

Can you explain?

I think, that the instruction set is fairly small, harmonious
and orthogonal.

Have you seen the instruction set of the PowerPC?
It seems, that everybody at IBM has added it's own
favorite instruction or instruction variation...

Groeten,
Jaap

-- Chip Directory
-- http://www.hitex.com/chipdir/ - USA
-- http://www.xs4all.nl/~ganswijk/chipdir/ - Europe
-- Many other international mirror sites from there...