RFR: 8365017: The SegmentBulkOperations::copy method can be improved using overlaps [v2]

[Per Minborg]

On Fri, 8 Aug 2025 09:36:14 GMT, Per Minborg <pminborg at openjdk.org> wrote:

>> This PR proposes to use overlapping memory areas in `SegmentBulkOperations::copy`, similar to what is proposed for `SegmentBulkOperations::fill` in https://github.com/openjdk/jdk/pull/25383.
>> 
>> This PR passes `tier1`, `tier2`, and `tier3`testing on multiple platforms.
>
> Per Minborg has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Optimize copy for certain lengths

The latest version improves performance further:

Bytes | Base | Patch | Iprovement
-- | -- | -- | --
2 | 3.184 | 2.836 | 10.93%
3 | 3.162 | 2.833 | 10.40%
4 | 3.344 | 2.839 | 15.10%
5 | 3.31 | 2.83 | 14.50%
6 | 3.328 | 2.846 | 14.48%
7 | 3.289 | 2.841 | 13.62%
8 | 4.139 | 2.834 | 31.53%
12 | 4.709 | 2.837 | 39.75%
16 | 4.173 | 3.749 | 10.16%
24 | 4.239 | 4.097 | 3.35%
64 | 4.86 | 4.716 | 2.96%

<img width="828" height="519" alt="image" src="https://github.com/user-attachments/assets/74d4aa1b-17f7-4ae6-9ff0-8e68f243be10" />

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

PR Comment: https://git.openjdk.org/jdk/pull/26672#issuecomment-3167220028