Re: t_INT API

[Bill Allombert <Bill.Allombert@math.u-bordeaux.fr>]

On Sat, Sep 28, 2002 at 12:58:00PM -0700, Ilya Zakharevich wrote:
> On Sat, Sep 28, 2002 at 10:56:24AM +0200, Bill Allombert wrote:
> > One question though:
> > this system is fine if we only have to support PARI and current GMP format.
> > However should we had a provision for other formats, e.g. formats
> > that use less than 32 bits per words like "GMP with nails" ???
> > In this case we could only provide a macro to access bits individualy.
> > But that maybe slower and an overkill.
> 
> Since it is going to be very much slower anyway, the by-word-access
> macro can do word-access emulation on such architectures.

That should not be taken for granted. GMP want to use optionnaly only 
30 bits per (32bits)words to allow for carry-save addition scheme, and expects
to be faster on average.

Instead of prec/next we could have a macro that make a loop over bits...

Well, I suppose I will implement the prvious scheme, and if need arise, we will
change.

Cheers,

Bill