Re: int32 & float live together (Re: Question about table arrays)

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Re: int32 & float live together (Re: Question about table arrays)

Shawn Kirst
I would be very interested in taking a look at your patch!  Thanks.

On Monday 09 January 2006 19:22, Asko Kauppi wrote:

> I have an unpublished patch for making Lua 5.1-w4 run with _both_
> float and int32 number types, transparently to the application layer.
> This is imho highly useful on ARM etc. platforms that don't have an FPU.
>
> The patch adds complexity to the code, and is not much faster than
> pure-float code would be. However, it does offer "clean" integer
> range, while using 24-bit for floating points. Anyone interested? :)
>
> -asko
>
> Rici Lake kirjoitti 9.1.2006 kello 18.09:
> > On the other hand, if the Lua instance were compiled with
> > lua_Number as a (single-precision) float, then the array part could
> > grow to a larger size than could be represented as an integral
> > lua_Number, which seems like it ought to trigger an error of some
> > sort.

--
Shawn Kirst
http://nifelheim.dyndns.org/~cocidius
http://ignh.dyndns.org/