On Mon, 26 Aug 2024 14:25:31 GMT, Per Minborg <pminb...@openjdk.org> wrote:
>> It is true, that this is a compromise where we give up inline space,
>> code-cache space, and introduce added complexity against the prospect of
>> better small-size performance. Depending on the workload, this may or may
>> not pay off. In the (presumably common) case where we allocate/fill small
>> segments of constant sizes, this is likely a win. Writing a dynamic
>> performance test sounds like a good idea.
>
> Here is a benchmark that fills segments of various random sizes:
>
>
>
> @BenchmarkMode(Mode.AverageTime)
> @Warmup(iterations = 5, time = 500, timeUnit = TimeUnit.MILLISECONDS)
> @Measurement(iterations = 10, time = 500, timeUnit = TimeUnit.MILLISECONDS)
> @State(Scope.Thread)
> @OutputTimeUnit(TimeUnit.NANOSECONDS)
> @Fork(value = 3)
> public class TestFill {
>
> private static final int SIZE = 16;
> private static final int[] INDICES = new Random(42).ints(0, 8)
> .limit(SIZE)
> .toArray();
>
>
> private MemorySegment[] segments;
>
> @Setup
> public void setup() {
> segments = IntStream.of(INDICES)
> .mapToObj(i -> MemorySegment.ofArray(new byte[i]))
> .toArray(MemorySegment[]::new);
> }
>
> @Benchmark
> public void heap_segment_fill() {
> for (int i = 0; i < SIZE; i++) {
> segments[i].fill((byte) 0);
> }
> }
>
> }
>
>
> This produces the following on my Mac M1:
>
>
> Benchmark Mode Cnt Score Error Units
> TestFill.heap_segment_fill avgt 30 59.054 ? 3.723 ns/op
>
>
> On average, an operation will take 59/16 = ~3 ns per operation (including
> looping).
>
> A test with the same size for every benchmark looks like this on my machine:
>
>
> Benchmark (ELEM_SIZE) Mode Cnt Score Error Units
> TestFill.heap_segment_fill 0 avgt 30 1.112 ? 0.027 ns/op
> TestFill.heap_segment_fill 1 avgt 30 1.602 ? 0.060 ns/op
> TestFill.heap_segment_fill 2 avgt 30 1.583 ? 0.004 ns/op
> TestFill.heap_segment_fill 3 avgt 30 1.909 ? 0.055 ns/op
> TestFill.heap_segment_fill 4 avgt 30 1.605 ? 0.059 ns/op
> TestFill.heap_segment_fill 5 avgt 30 1.900 ? 0.064 ns/op
> TestFill.heap_segment_fill 6 avgt 30 1.891 ? 0.038 ns/op
> TestFill.heap_segment_fill 7 avgt 30 2.237 ? 0.091 ns/op
As discussed offline, can't we use a stable array of functions or something
like that which can be populated lazily? That way you can access the function
you want in a single array access, and we could put all these helper methods
somewhere else.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/20712#discussion_r1731855496
