RFR: 8287925: AArch64: intrinsics for compareUnsigned method in Integer and Long

[Andrew Haley]

On Mon, 28 Nov 2022 02:31:25 GMT, Hao Sun <haosun at openjdk.org> wrote:

> x86 implemented the intrinsics for compareUnsigned() method in Integer and Long. See JDK-8283726. We add the corresponding AArch64 backend support in this patch.
> 
> Note-1: minor style issues are fixed for CmpL3 related rules.
> 
> Note-2: Jtreg case TestCompareUnsigned.java is updated to cover the matching rules for "comparing reg with imm" case.
> 
> Testing: tier1~3 passed on Linux/AArch64 platform with no new failures.
> 
> Following is the performance data for the JMH case:
> 
> 
>                                                        Before          After
> Benchmark                         (size) Mode  Cnt   Score   Error  Score   Error  Units
> Integers.compareUnsignedDirect      500  avgt    5   0.994 ± 0.001  0.872 ± 0.015  us/op
> Integers.compareUnsignedIndirect    500  avgt    5   0.991 ± 0.001  0.833 ± 0.055  us/op
> Longs.compareUnsignedDirect         500  avgt    5   1.052 ± 0.001  0.974 ± 0.057  us/op
> Longs.compareUnsignedIndirect       500  avgt    5   1.053 ± 0.001  0.916 ± 0.038  us/op

I've been trying for an hour or so to write a benchmark that is actually improved by this patch. I have been unable to do so. Part of the problem is that using `cmp; cset; cneg` doesn't take advantage of branch prediction. And more often than not, the result of a comparison is predictable. Also, the code isn't necessarily shorter either.
I do not believe this patch should be committed unless someone manages to write a benchmark that demonstrates some advantage.

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

PR: https://git.openjdk.org/jdk/pull/11383