On Tue, 18 Nov 2025 12:16:15 GMT, Maurizio Cimadamore <[email protected]>
wrote:
>> Igor Rudenko has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> improvements according to reviewer comments
>
> src/java.base/share/classes/jdk/internal/foreign/AbstractMemorySegmentImpl.java
> line 413:
>
>> 411: private void checkBounds(long offset, long length, BoundPolicy
>> policy) {
>> 412: if (length > 0) {
>> 413: Preconditions.checkIndex(offset, this.length - length + 1,
>> new BoundsCheckHandler(this, policy));
>
> This is a very very hot path. Allocating here is gonna have negative effect,
> especially when this code cannot be inlined (in which case the allocation
> can't be scalar replaced).
Thank you for pointing this out, I believe we can move on with this solution if
and only if the escape analysis eliminates instance creation. I'll run
benchmarks you mentioned in order to check if it's so.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/28124#discussion_r2538008716
