RFR: 8318222: RISC-V: C2 CmpU3

[Fei Yang]

On Mon, 23 Oct 2023 15:45:39 GMT, Hamlin Li <mli at openjdk.org> wrote:

> Hi,
> Can you review the change to add intrinsic for CmpU3 and CmpUL3?
> Thanks!
> 
> ## Test
> 
> ### functionality
> pass jtreg test:
> jdk/java/lang/Long/Unsigned.java, jdk/java/lang/Integer/Unsigned.java
> 
> ### performance
> #### Long
> **before**:
> Benchmark                      (size)  Mode  Cnt       Score       Error  Units
> Longs.compareUnsignedDirect       500  avgt    5    1454.789 ±   129.557  ns/op
> Longs.compareUnsignedIndirect     500  avgt    5    1410.146 ±   120.017  ns/op
> 
> **after**:
> Benchmark                      (size)  Mode  Cnt       Score       Error  Units
> Longs.compareUnsignedDirect       500  avgt    5    1286.129 ±     8.441  ns/op
> Longs.compareUnsignedIndirect     500  avgt    5     993.490 ±     0.840  ns/op
> 
> #### Integer
> **before**:
> Benchmark                         (size)  Mode  Cnt       Score       Error  Units
> Integers.compareUnsignedDirect       500  avgt    5    1611.753 ±     0.700  ns/op
> Integers.compareUnsignedIndirect     500  avgt    5    1775.093 ±     1.520  ns/op
> 
> **after**:
> Benchmark                         (size)  Mode  Cnt       Score       Error  Units
> Integers.compareUnsignedDirect       500  avgt    5    1159.351 ±     0.601  ns/op
> Integers.compareUnsignedIndirect     500  avgt    5     776.185 ±     0.924  ns/op

LGTM. May I ask on which platform was the JMH tested?
Also I think it's safer to perform some regression tests like tier1-3.

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

Marked as reviewed by fyang (Reviewer).

PR Review: https://git.openjdk.org/jdk/pull/16314#pullrequestreview-1693866847