RFR: 8339677: [vectorapi] YYYXXXVector::withLaneHelper and laneHelper should use Double::doubleToRawLongBits/Float::floatToRawIntBits

Fri Sep 6 18:41:49 UTC 2024

Hi,

This patch fixes an issue where we mistakenly use `Double::doubleToLongBits` in `DoubleXXXVector::withLaneHelper` and `DoubleXXXVector::laneHelper`. We should use the raw bit version instead.

Please kindly review, thanks very much.

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

Commit messages:
 - fix withLane and lane of floating points

Changes: https://git.openjdk.org/jdk/pull/20895/files
  Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=20895&range=00
  Issue: https://bugs.openjdk.org/browse/JDK-8339677
  Stats: 1997 lines in 65 files changed: 373 ins; 1302 del; 322 mod
  Patch: https://git.openjdk.org/jdk/pull/20895.diff
  Fetch: git fetch https://git.openjdk.org/jdk.git pull/20895/head:pull/20895

PR: https://git.openjdk.org/jdk/pull/20895