> This PR proposes to improve the performance of `MemorySegment::mismatch` by > using Java code rather than transitioning to native code.
Per Minborg has updated the pull request with a new target base due to a merge or a rebase. The incremental webrev excludes the unrelated changes brought in by the merge/rebase. The pull request contains 22 additional commits since the last revision: - Always use Java for mismatch by default on a64 - Only use unrolling on a64 - Unroll loop and move method - Merge branch 'master' into mismatch-performance2 - Use unaligned ops and rename benchmarks - Fix errors in a benchmark - Lower the mismatch threshold - Move method to SegmentBulkOperations and fix benchmark - Merge branch 'master' into mismatch-performance2 - Consolidate logic in one method - ... and 12 more: https://git.openjdk.org/jdk/compare/9f3c9169...a75dc159 ------------- Changes: - all: https://git.openjdk.org/jdk/pull/20848/files - new: https://git.openjdk.org/jdk/pull/20848/files/f40573c4..a75dc159 Webrevs: - full: https://webrevs.openjdk.org/?repo=jdk&pr=20848&range=10 - incr: https://webrevs.openjdk.org/?repo=jdk&pr=20848&range=09-10 Stats: 9514 lines in 321 files changed: 5283 ins; 2651 del; 1580 mod Patch: https://git.openjdk.org/jdk/pull/20848.diff Fetch: git fetch https://git.openjdk.org/jdk.git pull/20848/head:pull/20848 PR: https://git.openjdk.org/jdk/pull/20848
