RFR: 8269725: AArch64: Add VectorMask query implementation for NEON [v5]

[Xiaohong Gong]

On Wed, 14 Jul 2021 10:01:23 GMT, Andrew Haley <aph at openjdk.org> wrote:

>> Xiaohong Gong has updated the pull request with a new target base due to a merge or a rebase. The pull request now contains five commits:
>> 
>>  - Merge jdk:master into JDK-8269725
>>  - Add more comments
>>  - Remove the begining "negr" for "firsttrue,lasttrue"
>>  - Merge branch 'jdk:master' into JDK-8269725
>>  - 8269725: AArch64: Add VectorMask query implementation for NEON
>
> src/hotspot/cpu/aarch64/aarch64_neon.ad line 5355:
> 
>> 5353:     __ lsrw($dst$$Register, $dst$$Register, 3);
>> 5354:     __ movw(rscratch1, vector_length(this, $src));
>> 5355:     __ cmpw($dst$$Register, rscratch1);
> 
> You should be able to use `cmpw($dst$$Register, vector_length(this, $src));` here if `operand_valid_for_add_sub_immediate(vector_length(this, $src))`

@theRealAph , seems the `__ movw` in line-5354 is still needed since the `vector_length` is also used in the next instruction `__ cselw($dst$$Register, rscratch1, $dst$$Register, Assembler::GE);`. It is used to decide the final result.

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

PR: https://git.openjdk.java.net/jdk/pull/4699