On Tue, 25 Aug 2026 07:18:35 GMT, Shawn Emery <[email protected]> wrote:
>> This fix allows for more accurate comparisons between intrinsics and Java >> performance for P-256 conditional set operations by emulating production >> lookup EC operations. Before this fix, benchmarks showed a 53% regression >> for intrinsics vs. Java due to constant propagation and dead-code removal by >> C2. After this fix, a 27% performance increase is observed with intrinsics >> vs. Java. >> >> --------- >> - [X] I confirm that I make this contribution in accordance with the >> [OpenJDK Interim AI Policy](https://openjdk.org/legal/ai). > > Shawn Emery has updated the pull request incrementally with one additional > commit since the last revision: > > Implement comment from merykitty test/micro/org/openjdk/bench/javax/crypto/full/PolynomialP256Bench.java line 59: > 57: final ImmutableIntegerModuloP x = residueField.getElement(refx); > 58: final ImmutableIntegerModuloP X = montField.getElement(refx); > 59: final ImmutableIntegerModuloP one = residueField.get1(); Since you are here, I would suggest having better names for these fields. ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/32047#discussion_r3859943302
