[PATCH 0/8] Add sinf/cosf/sincosf CORE-MATH implementations

[Adhemerval Zanella Netto]

On 30/03/26 14:41, Wilco Dijkstra wrote:
> Hi Adhemerval,
> 
> I ran these on Neoverse V2 and added my old patches with traces from a game engine.
> I'm seeing major slowdowns:
> 
>              sinf	cosf
>      random  45.4%	45.2%
>       large  68.0%	53.4%
>        huge  19.4%	9.8%
> box2d-small  58.2%	100.2%
>  box2d-hard  56.7%	72.7%
> box2d-mixed  52.5%	54.4%
> 
> The box2d results are far worse since they are traces rather than randomized,
> so they run much faster due to fewer mispredictions. Latency is 2-3x lower for the
> same reason.

Do you have a link how to run box2d?

> 
> So improving the benchtests with traces would be a good idea, however it seems
> the CORE-MATH sinf/cosf implementation is significantly slower on modern cores
> even with random inputs.

CCing Paul.  I used perf.sh from CORE-MATH and I am seeing the opposite, where
CORE-MATH shows 'less' branch mispredictions than current glibc one.  But it
also does not apply to the benchtests that I have added (where the performance
is neutral).

I will try to come up with a better trace from speccpu2017.