[PATCH] AArch64: Optimize strlen

[Wilco Dijkstra]

Hi Cupertino,

> I am attempting to reproduce the presented performance improvements on a
> Ampere Altra processor.
> Can you please detail some more what was your setup and how did you
> measured it.

I measured it on large strings (eg. 1024 bytes) on Neoverse V1. You can run
benchtests/bench-strlen.c and look at results for larger strings. However Altra
is half as wide as V1, so may not see much (or any) benefit. Also note by default
it uses the multiarch/strlen_asimd.S version, not strlen.S.

> BTW, I am not looking to discredit the work, but rather to be able
> to replicate the results on our end to evaluate backporting your patches.

Yes it has been a while since we backported string improvements, so it's a good
idea. Are you talking about the official GLIBC release branches or a private branch?

Cheers,
Wilco