Re: gcc compile-time performance

[Tim Hollebeek <tim at hollebeek dot com>]

On Mon, May 20, 2002 at 03:54:20PM +0100, Richard Earnshaw wrote:
> > <<Native floating point code is problem, unfortunately since for i386 you get
> > different results in optimized and non-optimized builds breaking bootstrap.
> > Fixed point code is problem, as we are interested in comparisons relative
> > to the highest fequency in the program.  This may be the entry block for
> > tree-structured function, but it may be the internal loop and at high loop
> > nests there is more than 2^30 differences between these two we can affort
> > in integral arithmetics.
> > >>
> > 
> > Why not just use 64-bit scaled integer arithmetic, sounds like it would
> > work fine in this case, and it will be a lot faster than emulated fpt
> 
> We can't assume the presence of 64-bit integer types during the initial 
> stage of the bootstrap (ie before we have finished making stage1)

Don't need native 64-bit types.

Emulated 64-bit arithmetic is still an order of magnitude less complex
than emulated floating point.