RFR: 8373026: C2 SuperWord and Vector API: vector algorithms test and benchmark

[Paul Sandoz]

On Wed, 3 Dec 2025 15:11:33 GMT, Emanuel Peter <epeter at openjdk.org> wrote:

> This is an exploratory work. I wanted to use auto vectorization and the Vector API to implement some SIMD algorithms. We don't have too many IR tests and benchmarks, so I'm proposing an initial set of them, to be extended in the future.
> 
> Note: for now they are all `int` based. And some of them may not use the Vector API optimally, so feel free to propose ideas and integrate them in a follow-up RFE ;)
> 
> **Discussion**
> 
> Observations:
> - If the loop can be auto vectorized, that is the fastest. If we cannot vectorize, we at least get reasonable scalar performance.
> - If the Vector API code can be fully intrinsified, we get fast code. But somtimes, the Vector API is horribly slow, much slower than scalar loop performance.
>   - `linux_aarch64_server`: `filterI`, `scanAddI`, `reduceAddIFieldsX4` are very slow
>   - `macosx_aarch64`: `filterI`, `scanAddI`, `reduceAddIFieldsX4`, `findMinIndex` are very slow
>   - `linux_x64_oci_server`: Vector API leads to really nice speedups
>   - `windows_x64_oci_server`: the only one that gets good/better performance on all benchmarks
>   - `macosx_x64_sandybridge`: `scanAddI`!, `reduceAddIFieldsX4` are very slow. Other benchmarks benefit.
> - Compact Object Headers has some negative effect on some loop benchmarks.
>   - `linux_aarch64_server`: `reduceAddI`, `copyI`
>   - `macosx_aarch64`: `mapI`, `reduceAddI`, `copyI`
>   - `linux_x64_oci_server`: `reduceAddI`, `copyI`, `findI`?
>   - `windows_x64_oci_server`: `reduceAddI` and some others a little bit
>   - `macosx_x64_sandybridge`: `fillI`, `iotaI`, `mapI`, `reduceAddI`, `copyI`
> - Intrinsics can be much faster than auto vectoirzed or Vector API code.
>   - `linux_aarch64_server`: `copyI`
>   - `macosx_x64_sandybridge`: actually, `Arrays.fill` seems to suffer with Compact Object Headers as well.
> - `rearrange` often needs to do the `mask load` and `and` operation inside the loop. That has a slight performance impact, I filed [JDK-8373240](https://bugs.openjdk.org/browse/JDK-8373240).
> 
> **Benchmark Plots**
> 
> Units: nanoseconds per algorithm invocation.
> 
> `linux_x64_oci`
> <img width="4500" height="6000" alt="algo_linux_x64_oci_server" src="https://github.com/user-attachments/assets/f2c5bbcb-e009-4c54-a1bf-91af45326cb9" />
> 
> `windows_x64_oci`
> <img width="4500" height="6000" alt="algo_windows_x64_oci_server" src="https://github.com/user-attachments/assets/8946d248-4d75-4b16-8f17-627a90dcb6c3" />
> 
> `macosx_x64_sandybridge`
> <img width="4500" height="6000" alt="algo_macosx_x64_sandybridge" src="https://github.com/user...

test/micro/org/openjdk/bench/vm/compiler/VectorAlgorithmsImpl.java line 361:

> 359:             v = v.compress(mask);
> 360:             int trueCount = mask.trueCount();
> 361:             var prefixMask = VectorMask.fromLong(SPECIES_I, (1L << trueCount) - 1);

In case you did not already try, you can compress the mask to produce the mask prefix, as in `v.intoArray(r, j, mask.compress())`. Unsure if that will make a difference in the codegen though.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/28639#discussion_r2627986075