[PATCH] aarch64: Optimize algorithm in users of SVE expf helper.

Wilco Dijkstra Wilco.Dijkstra@arm.com
Tue Mar 18 17:17:24 GMT 2025


Hi Pierre,

> Polynomial order was unnecessarily high, unlocking multiple
> optimizations.
> Max error for new SVE expf is 0.88 +0.5ULP.
> Max error for new SVE coshf is 2.56 +0.5ULP.
> Performance improvement on Neoverse V1: expf (30%), coshf (26%).
> ---
> OK for master? If so please commit for as I don't have commit rights.

OK, committed.

Cheers,
Wilco


More information about the Libc-alpha mailing list