RFR: 8350463: AArch64: Add vector rearrange support for small lane count vectors

[Emanuel Peter]

On Thu, 13 Mar 2025 09:28:58 GMT, Emanuel Peter <epeter at openjdk.org> wrote:

>> The AArch64 vector rearrange implementation currently lacks support for vector types with lane counts < 4 (see [1]). This limitation results in significant performance gaps when running Long/Double vector benchmarks on NVIDIA Grace (SVE2 architecture with 128-bit vectors) compared to other SVE and x86 platforms.
>> 
>> Vector rearrange operations depend on vector shuffle inputs, which used byte array as payload previously. The minimum vector lane count of 4 for byte type on AArch64 imposed this limitation on rearrange operations. However, vector shuffle payload has been updated to use vector-specific data types (e.g., `int` for `IntVector`) (see [2]). This change enables us to remove the lane count restriction for vector rearrange operations.
>> 
>> This patch added the rearrange support for vector types with small lane count. Here are the main changes:
>>  - Added AArch64 match rule support for `VectorRearrange` with smaller lane counts (e.g., `2D/2S`)
>>  - Relocated NEON implementation from ad file to c2 macro assembler file for better handling of complex implementation
>>  - Optimized temporary register usage in NEON implementation for short/int/float types from two registers to one
>> 
>> Following is the performance improvement data of several Vector API JMH benchmarks, on a NVIDIA Grace CPU with NEON and SVE. Performance of the same JMH with other vector types remains unchanged.
>> 
>> 1) NEON
>> 
>> JMH on panama-vector:vectorIntrinsics:
>> 
>> Benchmark                    (size) Mode   Cnt Units   Before    After   Gain
>> Double128Vector.rearrange     1024  thrpt  30  ops/ms  78.060   578.859  7.42x
>> Double128Vector.sliceUnary    1024  thrpt  30  ops/ms  72.332  1811.664  25.05x
>> Double128Vector.unsliceUnary  1024  thrpt  30  ops/ms  72.256  1812.344  25.08x
>> Float64Vector.rearrange       1024  thrpt  30  ops/ms  77.879   558.797  7.18x
>> Float64Vector.sliceUnary      1024  thrpt  30  ops/ms  70.528  1981.304  28.09x
>> Float64Vector.unsliceUnary    1024  thrpt  30  ops/ms  71.735  1994.168  27.79x
>> Int64Vector.rearrange         1024  thrpt  30  ops/ms  76.374   562.106  7.36x
>> Int64Vector.sliceUnary        1024  thrpt  30  ops/ms  71.680  1190.127  16.60x
>> Int64Vector.unsliceUnary      1024  thrpt  30  ops/ms  71.895  1185.094  16.48x
>> Long128Vector.rearrange       1024  thrpt  30  ops/ms  78.902   579.250  7.34x
>> Long128Vector.sliceUnary      1024  thrpt  30  ops/ms  72.389   747.794  10.33x
>> Long128Vector.unsliceUnary    1024  thrpt  30  ops/ms  71....
>
> But the testing on my side so far looks good. I'll rerun once you add your IR tests.

> Thanks for looking at this PR again @eme64 ! Vector API has its own jtreg tests under `test/jdk/jdk/incubator/vector/`. I double checked that it has the `rearrange` test for all vector species. Please see one of the test here: https://github.com/openjdk/jdk/blob/master/test/jdk/jdk/incubator/vector/Long128VectorTests.java#L4954 That's also way I did not add the correct tests in the IR test file.

Alright. I think result verification would still be good practice, and not too difficult to do using a `@Check` method and `Verify.java` for comparing the resulting arrays. But I leave that up to you. In my experience, the VectorAPI test coverage is not as good as I first thought, see the list of bugs I recently found:
https://bugs.openjdk.org/issues/?jql=labels%20%3D%20template-framework%20ORDER%20BY%20created%20DESC%2C%20summary%20DESC

So adding a little more rigor to your IR test could catch possible bugs that the existing tests simply do not cover.

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

PR Comment: https://git.openjdk.org/jdk/pull/23790#issuecomment-2728450332