On Fri, 6 Mar 2026 12:08:55 GMT, Galder Zamarreño <[email protected]> wrote:

> Fixed benchmark code to avoid overflowing when calculating `highestInt` and 
> `highestLong`.
> 
> I observed this during early benchmark runs of 
> https://github.com/openjdk/jdk/pull/30070. I observed some oddities in the 
> results with the clipping range benchmarks, and when I inspected the 
> assembly, the values that were being used to clip didn't make sense. After 
> fixing this the odd results went away.
> 
> ---------
> - [x] I confirm that I make this contribution in accordance with the [OpenJDK 
> Interim AI Policy](https://openjdk.org/legal/ai).

This pull request has now been integrated.

Changeset: 67afba7c
Author:    Galder Zamarreño <[email protected]>
URL:       
https://git.openjdk.org/jdk/commit/67afba7cbeeffcdd7d945d35ae72faa7d615db40
Stats:     2 lines in 1 file changed: 0 ins; 0 del; 2 mod

8364025: MinMaxVector clipping range benchmark highestInt/Long can overflow

Reviewed-by: roland, kxu

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

PR: https://git.openjdk.org/jdk/pull/30115

Reply via email to