On Tue, 12 May 2026 14:38:09 GMT, Kangcheng Xu <[email protected]> wrote:
> Hi, > > This PR adds intrinsics to `Preconditions.checkFromToIndex()` and > `Preconditions.checkFromIndexSize()` to produced optimized IR that uses > `RangeCheck` node instead of implicit comparisons. > `Preconditions.checkIndex()` is also refactored (without additional > optimizations) to use the same helper function. > > Some common patterns where calling `checkFromToIndex` or `checkFromIndexSize` > in a loop can have range checks in main loop eliminated completely and, > therefore, enables empty loop removal. > > IR and correctness tests are included and passing. Additional tests on > `Preconditions.checkIndex()` were also added. > > Thanks! > > --------- > - [x] I confirm that I make this contribution in accordance with the [OpenJDK > Interim AI Policy](https://openjdk.org/legal/ai).  In terms of benchmark, I've observed the following: 1. a significant higher throughput for `checkFromToIndex()` and `checkFromIndexSize()` cases. Range checks are successfully hoisted out and eliminated where possible. 2. One-check-before-loop and array-copy are baseline tests. They shows intrinsification doesn't introduce regression when it can't benefit from range check removals, and both implementations are on par. 3. interestingly, `checkIndex()` is slightly worse with the refactor, but they even out as the number of operations scale. ------------- PR Comment: https://git.openjdk.org/jdk/pull/31138#issuecomment-5343235738
