Re: A missing (?) function: exponentfp(x)

[Ilya Zakharevich <nospam-abuse@ilyaz.org>]

On Mon, Sep 16, 2024 at 07:13:39PM +0200, Bill Allombert wrote:
> > > ? install(dbllog2,lG);
> > > ? install(dbltor,L);
> > > ? flog2(x)=dbltor(dbllog2(x));
> > > ? my(il2=1/log(2));setrand(1);for(i=1,10^6,log(random(1.)<<4)*il2)
> > > cpu time = 3,868 ms, real time = 3,868 ms.
> > > ? setrand(1);for(i=1,10^6,flog2(random(1.)<<4))
> > > cpu time = 655 ms, real time = 656 ms.
> > 
> > On my Intel x64 CPU, this gives
> > 
> >   ***   in function flog2: dbltor(dbllog2(x))
> >   ***                      ^------------------
> >   ***   attempt to change built-in dbltor.
> > 
> > Do not know whether it is related to the ABI problems Bill wrote
> > about…
> 
> This is unrelated.  You probably did the install() after having defined
> flog2(), so they were not taken into account by the bytecode compiler 
> when compiling flog2

This hits the bullet: repeating the definition of flog2() fixes this!

  And the result on my CPU is: 11 times slower than exponent()…  (In
  the spirit of my other message on this thread, this would be a very
  good “second step” of “gradual improvement” of exponent() — on top
  of the step I advocate…)

Thanks,
Ilya

  P.S.  Can the error message be made more clear?