RealFrac methods for Double and Float

Daniel Fischer daniel.is.fischer at web.de
Tue Oct 12 11:22:58 EDT 2010


On Tuesday 12 October 2010 17:15:14, Antoine Latter wrote:
> On Tue, Oct 12, 2010 at 10:02 AM, Tyson Whitehead <twhitehead at gmail.com> 
wrote:
> > On October 11, 2010 15:12:12 Daniel Fischer wrote:
> >> GHC's new code generator does some incredible stuff with the code for
> >> Float's RealFrac instance, properFraction has become about _seven
> >> times_ faster, floor and ceiling about _thirteen times_ and round
> >> about _10.5 times_ (values for my box, the exact numbers will differ,
> >> but the tendency will be the same).
> >
> > Out of curiosity, by new code generator, are you meaning just the
> > stock GHC 7 code path or are you meaning the new now available but
> > disabled by default unless you use some flag (which I forget right
> > now) code generation path.
> >
> > Cheers!  -Tyson
>
> When I did the GHC 7 timings I simply ran the shell script provided in
> the linked tarball - no funny business.
>
> Antoine

Yep, just a vanilla perf build of HEAD.


More information about the Libraries mailing list