RFR: 8322174: RISC-V: C2 VectorizedHashCode RVV Version [v18]

[Yuri Gaevsky]

On Wed, 6 Aug 2025 12:31:26 GMT, Yuri Gaevsky <duke at openjdk.org> wrote:

>> The patch adds possibility to use RVV instructions for faster vectorizedHashCode calculations on RVV v1.0.0 capable hardware.
>> 
>> Testing: hotspot/jtreg/compiler/ under QEMU-8.1 with RVV v1.0.0.
>
> Yuri Gaevsky has updated the pull request incrementally with one additional commit since the last revision:
> 
>   make powmax calculations scalar; re-use v_tmp for sum reduction operation.

`a85abe7`:

--- -XX:+UseRVV ---
Benchmark            (size)  Mode  Cnt     Score    Error  Units
ArraysHashCode.ints       1  avgt   30    11.285 ±  0.011  ns/op
ArraysHashCode.ints       5  avgt   30    21.297 ±  0.005  ns/op
ArraysHashCode.ints      10  avgt   30    33.821 ±  0.007  ns/op
ArraysHashCode.ints      20  avgt   30    58.884 ±  0.024  ns/op
ArraysHashCode.ints      30  avgt   30    84.013 ±  0.094  ns/op
ArraysHashCode.ints      40  avgt   30   109.038 ±  0.075  ns/op
ArraysHashCode.ints      50  avgt   30   134.041 ±  0.038  ns/op
ArraysHashCode.ints      60  avgt   30   159.348 ±  0.252  ns/op
ArraysHashCode.ints      70  avgt   30    87.701 ±  0.049  ns/op
ArraysHashCode.ints      80  avgt   30   109.611 ±  0.032  ns/op
ArraysHashCode.ints      90  avgt   30   134.703 ±  0.078  ns/op
ArraysHashCode.ints     100  avgt   30   159.367 ±  0.232  ns/op
ArraysHashCode.ints     200  avgt   30   120.708 ±  0.223  ns/op
ArraysHashCode.ints     300  avgt   30   229.257 ±  0.037  ns/op
ArraysHashCode.ints    1000  avgt   30   397.151 ±  4.697  ns/op
ArraysHashCode.ints   10000  avgt   30  5362.472 ± 19.957  ns/op

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

PR Comment: https://git.openjdk.org/jdk/pull/17413#issuecomment-3160331579