PATCH: optimized libm single precision routines: erfcf, erff, expf for x86_64.

[Ulrich Drepper]

On Mon, Feb 27, 2012 at 13:14, Dmitrieva Liubov
<liubov.dmitrieva@gmail.com> wrote:
> Results:
> Istanbul/Atom/Nehalem/AVX
> Attached version:   51.63     97.23     31.12     25.34   (__AVX__
> enabling gives 24.98 cycles on AVX.)

OK, looks good. Those who care can easily rewrite the code in C but I
agree that gcc likely cannot generate the code as tight as this.