On Wed, 8 Jul 2026 14:48:28 GMT, Viktor Klang <[email protected]> wrote:
>> I don't think we are bound to use a particular behavior. The original >> behavior does not introduce an inconsistency so I think it's fine to leave >> it as-is. We might change these calculations if there are more compelling >> reasons for subsequent previews, such as if another way to accumulate the >> result is more hardware-friendly. > > It would, to me, make most sense to align with Long.hashCode, unless there's > a valid reason not to. I think reusing 4-byte-wise accumulation might be easier for compilers to reuse hardware instructions? ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/31123#discussion_r3545261094
