RFR: 8257561: Some code is not vectorized after 8251925 and 8250607

Thu Dec 3 17:05:55 UTC 2020

On Thu, 3 Dec 2020 13:39:47 GMT, Tobias Hartmann <thartmann at openjdk.org> wrote:

> Looks good to me. Hopefully such failures are caught by tests using [JDK-8254129](https://bugs.openjdk.java.net/browse/JDK-8254129) in the future.

Yes, it would be nice. Or we can analyze TraceNewVectors output for these tests - unfortunately it is not very stable (depends on AVX version support and other parameters). May be have only 'yes or 'no' for presence of new vectors and not their count.

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

PR: https://git.openjdk.java.net/jdk/pull/1584