RFR: 8316594: C2 SuperWord: wrong result with hand unrolled loops

[Emanuel Peter]

On Tue, 3 Oct 2023 06:21:06 GMT, Tobias Hartmann <thartmann at openjdk.org> wrote:

>> Please read description in https://github.com/openjdk/jdk/pull/15864.
>> 
>> Currently, we only compare the first element of the load-pack with the StoreVector. This is not correct, we need to compare all elements of the load-pack with the StoreVector. Otherwise it may be that the first element lays below/before the memory region of the StoreVector (Less), but the last element of the load-pack for example could have overlap with the StoreVector (Equal).
>> 
>> Tier1-6 + stress-testing (Running)
>> Performance testing (Running)
>
> Looks good to me. To avoid merge conflicts when pushing https://github.com/openjdk/jdk/pull/15864, you could fix both issues in the same PR since they are in the same code.

@TobiHartmann I want to push them as separate fixes, because one can be back-ported, but not the other one:
> We could fix both issues at once. But I am fixing them separately, first JDK-8316679, which can be backported to JDK-11, and then JDK-8316594 which only seems to have caused regressions in JDK-22.

I'll unify the tests once the other PR is integrated.

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

PR Comment: https://git.openjdk.org/jdk/pull/15866#issuecomment-1744339794