Pythoniac: Thoughts on a hardware Python processor

[Paul Foley]

On Thu, 27 Jun 2002 03:32:29 -0400, Oren Tirosh wrote:

> On Wed, Jun 26, 2002 at 10:29:52AM +0000, Christopher Saunter wrote:
>> Dear All,
>> 
>> I have seen the occasional post about creating a specific hardware 'Python
>> Processor', that could then be built and used in programmable logic or the
>> like (Valves, relays, core memory - Pythoniac ;-) - in other words, create
>> a processor that executes Python bytecode directly.  There would be
>> various different reasons for doing this (the 'neat' factor etc...)  

> One of the most effective optimizations for such a machine would be to 
> support compact objects.  The machine word would be somewhat larger than the 
> actual address space. Let's say 40 bits.  If the 4 high bits are 0 this is a
> reference to an object allocated somewhere in memory.  If the high bits are 
> not zero it means that the actual object is stored inside the register.  You 

It's usual to use the low bits for this purpose, rather than the high
bits.  You can use at least 2 low bits for free, on ordinary 32 bit
hardware, if pointers are aligned to 32 bit boundaries anyway.

-- 
Majority, n.:
        That quality that distinguishes a crime from a law.

(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))