int64 patch -- help request

[Asko Kauppi <askok@...>]

I've ran to some bug in making the integer patch work with 64-bit  
integers, as well as 32-bit.

Anyone interested in that work (perhaps someone targetting PS2  
platform?) would you please take contact. Would be interesting to get  
performance figures on 64-bit optimized (non-FPU) platforms, as well.

In short, the integer patch allows for transparent optimization of  
all integer operations, with fallbacks to float/double only when  
numbers need them. So, one could i.e. do int64/float or int64/double,  
without affecting Lua compatibility in any way (byte codes still  
remain float/double only, and this seems to be part of the background  
for the bug at hand).

-asko