[vectorIntrinsics+mask] RFR: 8264563: Add masked vector intrinsics for binary/store operations [v5]

[Paul Sandoz]

On Thu, 15 Apr 2021 02:16:00 GMT, Xiaohong Gong <xgong at openjdk.org> wrote:

>> src/jdk.incubator.vector/share/classes/jdk/incubator/vector/IntVector.java line 640:
>> 
>>> 638:             }
>>> 639:             // suppress div/0 exceptions in unset lanes
>>> 640:             that = that.lanewise(NOT, eqz);
>> 
>> I guess this is required because we don't know how the intrinsic will support masking? How can a  div/0 exception can occur?
>
> Thanks for looking at this PR @PaulSandoz ! Yes, this is actually required for masked division operation. I think it has to make sure no div/0 exception happens during hotspot compiling the intrinsic? Am I missing something? Thanks!

Did you observe any issues without this? I understand how such exceptions can occur when compiling the instrinsic, but could be missing something.

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

PR: https://git.openjdk.java.net/panama-vector/pull/57