RFR: 8322174: RISC-V: C2 VectorizedHashCode RVV Version [v16]
Yuri Gaevsky
duke at openjdk.org
Tue Aug 5 12:53:24 UTC 2025
> 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 with a new target base due to a merge or a rebase. The pull request now contains 17 commits:
- Merge master
- replaced vmul_vv + vadd_vv by vmadd_vv
- returned lmul==m4
- fixed error made for prevoius lmul-m1 experiment
- make an experiment with lmul==1 instead of lmul==4.
- move vredsum_vs out of VEC_LOOP to improve performance
- - removed tail processing with RVV instructions as simple scalar loop provides in general better results
- simplified arrays_hashcode_v() to be closer to VLA and use less general-purpose registers; minor cosmetic changes
- change slli+add sequence to shadd
- reorder instructions to make RVV instructions contiguous
- ... and 7 more: https://git.openjdk.org/jdk/compare/ba0ae4cb...e7fac6c7
-------------
Changes: https://git.openjdk.org/jdk/pull/17413/files
Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=17413&range=15
Stats: 443 lines in 6 files changed: 441 ins; 0 del; 2 mod
Patch: https://git.openjdk.org/jdk/pull/17413.diff
Fetch: git fetch https://git.openjdk.org/jdk.git pull/17413/head:pull/17413
PR: https://git.openjdk.org/jdk/pull/17413
More information about the hotspot-compiler-dev
mailing list