[PATCH] powerpc: Optimized strncat for POWER7/PPC64
Ondřej Bílka
neleai@seznam.cz
Sat Jan 3 10:07:00 GMT 2015
On Wed, Dec 24, 2014 at 10:17:17AM -0200, Adhemerval Zanella wrote:
> Hi
>
> The commit 3eb38795dbbbd816 (Simplify strncat) implements a new
> strncat algorithm that uses strlen, strnlen, and memcpy. This is
> faster than POWER7 current implementation, especially for unaligned
> strings (where POWER7 code uses byte-byte operations).
>
> This patch removes the assembly implementation and uses a multiarch
> especialization based on default algorithm calling optimized POWER7
> symbols. Benchtests results in attachments (ran on a POWER8 machine).
>
> Tested on powerpc64 and powerpc64le.
>
Should be ok, assembly implementation is lot slower in practice as its
rarely called at all and so is unlikely in cache.
More information about the Libc-alpha
mailing list