Mac v3.0.2 VM
John M McIntosh
johnmci at smalltalkconsulting.com
Tue Feb 6 22:34:34 UTC 2001
>Err, does this mean G3+ _only_ or not? I'm not hurt by this (Newer G4
>card - sweet!) but others may be, so clarification could be good.
> > The readme is below.
>(snip)
> >(k) Compiled for 750(G3) processor versus 603
>
The difference between a G3 and 603 compile on CW 5.3 seems to be how
instructions are sequenced. For a 603 it might do
pseudo code
loadregister 1,0,2 load register 1 from 2
storeregister 1,0,3 store register 1 into 3
loadregister 4,0,5 load register 4 from 5
But if you compile for G3 then it changes it to
loadregister 1,0,2 load register 1 from 2
loadregister 4,0,5 load register 4 from 5
storeregister 1,0,3 store register 1 into 3
Since the G3 may do both loads of register 1 and 4 at the same time,
versus stalling on the load/store register 1 sequence. I'm not sure
why the 603 optimizer is so stupid...
Note that the G3 might reorder the instructions on the fly too (I
think) RISC base CPUS are rather complex now...
--
--
===========================================================================
John M. McIntosh <johnmci at smalltalkconsulting.com> 1-800-477-2659
Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com
===========================================================================
Custom Macintosh programming & various Smalltalk dialects
PGP Key: DSS/Diff/46FC3BE6
Fingerprint=B22F 7D67 92B7 5D52 72D7 E94A EE69 2D21 46FC 3BE6
===========================================================================
More information about the Squeak-dev
mailing list
|