On Thu, 13 Feb 2025 07:52:30 GMT, Nicole Xu <d...@openjdk.org> wrote:
>> Oh, the OCA-verify is still stuck. I'm sorry about that 🙈 >> I pinged my manager @TobiHartmann , he will reach out to see what's the >> issue. > > Hi @eme64, do you see any risks here? Would you please help to review the > patch? Thanks. > @xyyNicole @jatin-bhateja I think it is reasonable to just fix the benchmark > so that it still has the same behaviour, just without the out-of-bounds > exception. @jatin-bhateja you originally wrote the benchmark, and it could > make sense if you fixed it up to what it should be more ideally. @xyyNicole I > propose that we file a follow-up RFE to fix the benchmark, and just mention > that issue in the benchmark. > > What do you think? Hi @xyyNicole , May I request you to kindly file a follow up RFE mentioning the discussed issues. Performance of lanewise operations works best without any noise if its fed contiguous memory locations. Best Regards, Jatin ------------- PR Comment: https://git.openjdk.org/jdk/pull/22963#issuecomment-2663639602
