RFR: 8277997: Intrinsic creation for VectorMask.fromLong API [v6]
Jatin Bhateja
jbhateja at openjdk.java.net
Thu Dec 9 18:19:18 UTC 2021
On Thu, 9 Dec 2021 16:15:17 GMT, Vladimir Kozlov <kvn at openjdk.org> wrote:
>> Jatin Bhateja has updated the pull request incrementally with one additional commit since the last revision:
>>
>> 8277997: Review comments resolution.
>
> Looks like the issue only with AVX512. I don't see failures with AVX2 or AVX1.
Hi @vnkozlov , thanks for reporting this, it look like a bug unrelated to this patch, its occurring because MaskAll instruction pattern is not supported for 32 bit JVM, due to which operations fall backs over replicateB operation which broadcasts the mask value in a vector and after unboxing-boxing optimization this vector eventually reaches to XorVMask which see one operand in opmask register and other in vector.
Current patch extends the C2 inline expansions broadcast Coerced routine to support VectorMask.fromLong operation.
I have create separate [bug entry](https://bugs.openjdk.java.net/browse/JDK-8278508) for it targeted for JDK-18
Please let me know if there are other comments on this patch.
-------------
PR: https://git.openjdk.java.net/jdk/pull/6646
More information about the hotspot-compiler-dev
mailing list