[aarch64-port-dev ] RFR(M): 8212043: Add floating-point Math.min/max intrinsics

Andrew Haley aph at redhat.com
Wed Dec 5 10:29:24 UTC 2018

On 12/5/18 4:19 AM, Pengfei Li (Arm Technology China) wrote:
>> On 12/4/18 10:22 AM, Andrew Haley wrote:
>>> It can go in as it is. It doesn't help performance, but neither does
>>> it make it much worse. The gain will come with vectorization.
>> But let us not wait! Pengfeng Li, will you start on the vectorization patch?

> Yes. I've started working on the vectorization today and found it not that difficult as I thought before.

That's excellent. Thanks you.

max and min are the same shape as any of the commutative arithmetic and
boolean operations, so it should be easy to copy that code.

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

More information about the hotspot-compiler-dev mailing list