On Tue, 12 May 2026 09:02:15 GMT, Andrew Dinn <[email protected]> wrote:

> I left a few comments regarding 'housekeeping' issues for the new intrinsic 
> implementations. Thank you for correctly updating the registration of 
> external addresses. n.b. I have left review of the generation steps to the 
> x86 vectorization experts (I am happy to review any proposed aarch64 
> equivalent).

aarch64 already has the double_keccak intrinsic, which exhausts its SIMD 
register set, so it will not benefit from the new quad_keccak() method until 
wider or more SIMD registers become available.

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

PR Comment: https://git.openjdk.org/jdk/pull/31125#issuecomment-4450382824

Reply via email to