RFR: 8334554: RISC-V: verify & fix perf of string comparison

[Hamlin Li]

On Tue, 25 Jun 2024 09:29:56 GMT, Fei Yang <fyang at openjdk.org> wrote:

>> Based on above observation,  I wonder if we can pass different lmul value depending on different count value? e.g. when count <= 36 we use m2, when count > 36 (maybe not 36 exactly, but some value close to it) we use m4? But that means there will be a branch in the code at runtime.
>
> Sounds reasonable. Further, I wonder if we should consider m1 for these <=36 cases.

By https://github.com/openjdk/jdk/pull/18382#discussion_r1645356197, do you mean current lastest implementation of it still brings negative impact on performance after https://github.com/openjdk/jdk/pull/18382/commits/1d4320b205c173256eb6a50cedaa0ccd9b4122d8?

-------------

PR Review Comment: https://git.openjdk.org/jdk/pull/19825#discussion_r1652420887