[11u] RFR: 8257436: [aarch64] Regressions in ArrayCopyUnalignedDst.testByte/testChar for 65-78 bytes when UseSIMDForMemoryOps is on

[Hohensee, Paul]

Thanks, Andrew.

Paul

-----Original Message-----
From: jdk-updates-dev <jdk-updates-dev-retn at openjdk.java.net> on behalf of Andrew Haley <aph at redhat.com>
Date: Monday, December 7, 2020 at 3:51 AM
To: "jdk-updates-dev at openjdk.java.net" <jdk-updates-dev at openjdk.java.net>
Subject: RE: [11u] RFR: 8257436: [aarch64] Regressions in ArrayCopyUnalignedDst.testByte/testChar for 65-78 bytes when UseSIMDForMemoryOps is on

is lOn 04/12/2020 23:45, Hohensee, Paul wrote:
> Please review this backport to 11u. This is the second follow-on backport after 8255351: Add detection for Graviton 2 CPUs. After the first follow-on backport (8257436, review request posted), patch applies cleanly except for a context difference at the end of hunk #1.
>
> JBS: https://bugs.openjdk.java.net/browse/JDK-8257436
> Patch: https://github.com/openjdk/jdk/commit/e8363962
> Webrev: http://cr.openjdk.java.net/~phh/8257436/webrev.11u.00/

OK. It's rather a kludge, but maybe the best we can do. It looks
like future AArch64 silicon is much better equipped with read and
write ports, so such things can (hopefully!) go away over time.

--
Andrew Haley  (he/him)
Java Platform Lead Engineer
Red Hat UK Ltd. <https://www.redhat.com>
https://keybase.io/andrewhaley
EAC8 43EB D3EF DB98 CC77 2FAD A5CD 6035 332F A671