On Wed, 18 Jan 2023 11:04:57 GMT, Claes Redestad <redes...@openjdk.org> wrote:
> This patch makes use of `ArraysSupport.vectorizedHashCode` and helps verify > the performance win also for range-based hash calculations. Also modernized > and cleaned up the surrounding code somewhat. > > > Benchmark Mode Cnt Score Error Units > Characters.CodePoints.codePointOf avgt 15 153.753 ± 6.119 ns/op > Characters.CodePoints.codePointOf avgt 15 134.753 ± 4.386 ns/op # after, > 1.14x > > > For C1 and C2 on non-x86 platforms the performance is neutral. Interpreter > sees a 1-2% regression due a few added calls. This pull request has now been integrated. Changeset: 3ea0b8bc Author: Claes Redestad <redes...@openjdk.org> URL: https://git.openjdk.org/jdk/commit/3ea0b8bc253f1498895a879681406ecc15f37afb Stats: 30 lines in 2 files changed: 15 ins; 5 del; 10 mod 8300489: Use ArraysSupport.vectorizedHashCode in j.l.CharacterName Reviewed-by: alanb, naoto ------------- PR: https://git.openjdk.org/jdk/pull/12068
