RFR: 8292587: AArch64: Support SVE fabd instruction

[Fei Gao]

On Thu, 25 Aug 2022 01:52:41 GMT, Hao Sun <haosun at openjdk.org> wrote:

> Scalar and NEON fabd instructions were initially supported in
> JDK-8256318. In this patch, we support SVE fabd instruction [1] and add
> one Jtreg test case as well.
> 
> With this patch, two instructions `fsub + fabs` would be combined into
> one single `fabd` instruction.
> 
> 
>   fsub    z16.s, z16.s, z17.s
>   fabs    z16.s, p7/m, z16.s
> 
>   -->
> 
>   fabd    z16.s, p7/m, z16.s, z17.s
> 
> 
> In the initial evaluation of JMH case, i.e.
> FloatingScalarVectorAbsDiff.java, we found the performance uplift done
> by this optimization was easily hidden by the heavy memory load/store
> instructions. To avoid that, we updated the JMH case a bit, adding one
> more group of subtraction and Math.abs operations in the loop body.
> 
> Here shows the data with the new JMH case on one 256-bit SVE machine. We
> can observe about 39% and 35% improvements for the two functions
> respectively.
> 
> 
> Benchmark                                             Before    After  Units
> FloatingScalarVectorAbsDiff.testVectorAbsDiffDouble  260.468  160.965  ns/op
> FloatingScalarVectorAbsDiff.testVectorAbsDiffFloat   133.963   87.292  ns/op
> 
> 
> Jtreg testing: tier1~3 passed on one NEON-only machine and one 256-bit SVE machine.
> 
> [1] https://developer.arm.com/documentation/ddi0596/2021-12/SVE-Instructions/FABD--Floating-point-absolute-difference--predicated--

test/hotspot/jtreg/compiler/vectorapi/VectorAbsDiffTest.java line 97:

> 95:     public static void testFloatAbsDiff_runner() {
> 96:         testFloatAbsDiff();
> 97:         for (int i = 0; i < F_SPECIES.length(); i++) {

I suppose it should be `for (int i = 0; i < LENGTH; i++) {` here. You can check all similar code lines in the following  functions for verification.

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

PR: https://git.openjdk.org/jdk/pull/10011