This is the mail archive of the
mailing list for the glibc project.
Re: [RFC][PATCH 0/2] aarch64: Add optimized ASIMD versions of sinf/cosf
On Fri, 2017-06-23 at 15:13 +0530, Siddhesh Poyarekar wrote:
> On Friday 23 June 2017 02:26 PM, Sekhar, Ashwin wrote:
> > I am going forward with the C implementation of the sinf/cosf.
> > Along
> > with it I am also working on the above changes you suggested for
> > benchtest framework. Also I am trying to add sinf and cosf inputs
> > in
> > the benchtest.
> > I need your inputs on how you arrived at the inputs for the sin and
> > cos
> > benchmarks (benchtests/cos-inputs and benchtests/sin-inputs). The
> > commit message and as well as the mail archive doesn't provide much
> > information on this.
> Those inputs basically just exercise all branches, with the obvious
> inouts (0, NaN, etc.) getting less weight since we shouldn't really
> about optimizing them.
Were the weights for different input ranges based on some benchmark?
And does the chosen values within an input range have some significance
or were they random values within that different input range ?
For sinf/cosf, I am planning to use the same weights as sin/cos and
random values within different input ranges. Thats why the question.