This is the mail archive of the mailing list for the glibc project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFC][PATCH 0/2] aarch64: Add optimized ASIMD versions of sinf/cosf

On Tue, 2017-06-13 at 19:49 +0530, Siddhesh Poyarekar wrote:
> Currently the microbenchmark framework tests the same input in a loop
> a
> specific number of times to get a large enough number that a single
> iteration gives a stable mean and then tests inputs in a loop - I
> agree
> that this is cheating a bit since it eliminates cache effects as well
> as
> branches.  It will need a pretty straightforward fix to run only once
> for a single input and it will do what you want, i.e. measure the
> effect
> of branches.
> Maybe Ashwin could patch the framework as well when he posts his
> patch.
Hi Siddhesh,

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.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]