ADDITIONAL SYSTEM INFORMATION :
    java.runtime.name = OpenJDK Runtime Environment
    java.runtime.version = 25.0.1+8-27
    java.specification.name = Java Platform API Specification
    java.specification.vendor = Oracle Corporation
    java.specification.version = 25
    java.vendor = Oracle Corporation
    java.vendor.url = https://java.oracle.com/
    java.vendor.url.bug = https://bugreport.java.com/bugreport/
    java.version = 25.0.1
    java.version.date = 2025-10-21
    java.vm.compressedOopsMode = Zero based
    java.vm.info = mixed mode, sharing
    java.vm.name = OpenJDK 64-Bit Server VM
    java.vm.specification.name = Java Virtual Machine Specification
    java.vm.specification.vendor = Oracle Corporation
    java.vm.specification.version = 25
    java.vm.vendor = Oracle Corporation
    java.vm.version = 25.0.1+8-27
    jdk.debug = release
    line.separator = \n
    native.encoding = UTF-8
    os.arch = aarch64
    os.name = Mac OS X
    os.version = 15.7.2
    path.separator = :
    stderr.encoding = UTF-8

A DESCRIPTION OF THE PROBLEM :
Recently I noticed an IDE warning where I was using Arrays::asList to construct a List with a single item. This lead me to check whether List.of or Collections.singletonList would be the best option to replace it. According to available documentation/literature online Collections.singletonList _should_ be faster than List.of but any difference is very marginal and both are preferable to Arrays:asList for constructing single item liss.

On x64 / Windows this does seem to be true:

```
Benchmark Mode Cnt Score Error Units
Benchmark.asList thrpt 30 280207.761 ± 3616.566 ops/ms
Benchmark.asList:·gc.alloc.rate thrpt 30 12826.813 ± 165.553 MB/sec
Benchmark.asList:·gc.alloc.rate.norm thrpt 30 48.000 ± 0.001 B/op
Benchmark.asList:·gc.count thrpt 30 18778.000 counts
Benchmark.asList:·gc.time thrpt 30 7555.000 ms
Benchmark.of thrpt 30 551714.274 ± 1430.914 ops/ms
Benchmark.of:·gc.alloc.rate thrpt 30 12627.667 ± 32.751 MB/sec
Benchmark.of:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.of:·gc.count thrpt 30 18702.000 counts
Benchmark.of:·gc.time thrpt 30 7378.000 ms
Benchmark.singletonList thrpt 30 552778.754 ± 986.674 ops/ms
Benchmark.singletonList:·gc.alloc.rate thrpt 30 12652.031 ± 22.581 MB/sec
Benchmark.singletonList:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.singletonList:·gc.count thrpt 30 18585.000 counts
Benchmark.singletonList:·gc.time thrpt 30 7212.000 ms
```

However, on OSX/ARM64 Collections::singletonList is _significantly_ slower than both (according to my simple benchmark).

```
Benchmark Mode Cnt Score Error Units
Benchmark.asList thrpt 30 292303.175 ± 2441.718 ops/ms
Benchmark.asList:·gc.alloc.rate thrpt 30 13380.515 ± 111.773 MB/sec
Benchmark.asList:·gc.alloc.rate.norm thrpt 30 48.000 ± 0.001 B/op
Benchmark.asList:·gc.count thrpt 30 12866.000 counts
Benchmark.asList:·gc.time thrpt 30 6820.000 ms
Benchmark.of thrpt 30 519191.871 ± 12149.670 ops/ms
Benchmark.of:·gc.alloc.rate thrpt 30 11883.295 ± 278.079 MB/sec
Benchmark.of:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.of:·gc.count thrpt 30 12054.000 counts
Benchmark.of:·gc.time thrpt 30 6342.000 ms
Benchmark.singletonList thrpt 30 62301.767 ± 909.762 ops/ms
Benchmark.singletonList:·gc.alloc.rate thrpt 30 1425.962 ± 20.823 MB/sec
Benchmark.singletonList:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.singletonList:·gc.count thrpt 30 2118.000 counts
Benchmark.singletonList:·gc.time thrpt 30 1833.000 ms
Benchmark Mode Cnt Score Error Units
Benchmark.asList thrpt 30 292303.175 ± 2441.718 ops/ms
Benchmark.asList:·gc.alloc.rate thrpt 30 13380.515 ± 111.773 MB/sec
Benchmark.asList:·gc.alloc.rate.norm thrpt 30 48.000 ± 0.001 B/op
Benchmark.asList:·gc.count thrpt 30 12866.000 counts
Benchmark.asList:·gc.time thrpt 30 6820.000 ms
Benchmark.of thrpt 30 519191.871 ± 12149.670 ops/ms
Benchmark.of:·gc.alloc.rate thrpt 30 11883.295 ± 278.079 MB/sec
Benchmark.of:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.of:·gc.count thrpt 30 12054.000 counts
Benchmark.of:·gc.time thrpt 30 6342.000 ms
Benchmark.singletonList thrpt 30 62301.767 ± 909.762 ops/ms
Benchmark.singletonList:·gc.alloc.rate thrpt 30 1425.962 ± 20.823 MB/sec
Benchmark.singletonList:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.singletonList:·gc.count thrpt 30 2118.000 counts
Benchmark.singletonList:·gc.time thrpt 30 1833.000 msBenchmark Mode Cnt Score Error Units
Benchmark.asList thrpt 30 292303.175 ± 2441.718 ops/ms
Benchmark.asList:·gc.alloc.rate thrpt 30 13380.515 ± 111.773 MB/sec
Benchmark.asList:·gc.alloc.rate.norm thrpt 30 48.000 ± 0.001 B/op
Benchmark.asList:·gc.count thrpt 30 12866.000 counts
Benchmark.asList:·gc.time thrpt 30 6820.000 ms
Benchmark.of thrpt 30 519191.871 ± 12149.670 ops/ms
Benchmark.of:·gc.alloc.rate thrpt 30 11883.295 ± 278.079 MB/sec
Benchmark.of:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.of:·gc.count thrpt 30 12054.000 counts
Benchmark.of:·gc.time thrpt 30 6342.000 ms
Benchmark.singletonList thrpt 30 62301.767 ± 909.762 ops/ms
Benchmark.singletonList:·gc.alloc.rate thrpt 30 1425.962 ± 20.823 MB/sec
Benchmark.singletonList:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.singletonList:·gc.count thrpt 30 2118.000 counts
Benchmark.singletonList:·gc.time thrpt 30 1833.000 msBenchmark Mode Cnt Score Error Units
Benchmark.asList thrpt 30 292303.175 ± 2441.718 ops/ms
Benchmark.asList:·gc.alloc.rate thrpt 30 13380.515 ± 111.773 MB/sec
Benchmark.asList:·gc.alloc.rate.norm thrpt 30 48.000 ± 0.001 B/op
Benchmark.asList:·gc.count thrpt 30 12866.000 counts
Benchmark.asList:·gc.time thrpt 30 6820.000 ms
Benchmark.of thrpt 30 519191.871 ± 12149.670 ops/ms
Benchmark.of:·gc.alloc.rate thrpt 30 11883.295 ± 278.079 MB/sec
Benchmark.of:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.of:·gc.count thrpt 30 12054.000 counts
Benchmark.of:·gc.time thrpt 30 6342.000 ms
Benchmark.singletonList thrpt 30 62301.767 ± 909.762 ops/ms
Benchmark.singletonList:·gc.alloc.rate thrpt 30 1425.962 ± 20.823 MB/sec
Benchmark.singletonList:·gc.alloc.rate.norm thrpt 30 24.000 ± 0.001 B/op
Benchmark.singletonList:·gc.count thrpt 30 2118.000 counts
Benchmark.singletonList:·gc.time thrpt 30 1833.000 ms
```

According to this benchmark Collections::singletonList is ~8x slower than List::of.


STEPS TO FOLLOW TO REPRODUCE THE PROBLEM :
```
package brutus;


import java.util.Arrays;
import java.util.Collections;
import java.util.List;
import java.util.concurrent.TimeUnit;
import org.openjdk.jmh.annotations.BenchmarkMode;
import org.openjdk.jmh.annotations.Fork;
import org.openjdk.jmh.annotations.Measurement;
import org.openjdk.jmh.annotations.Mode;
import org.openjdk.jmh.annotations.OutputTimeUnit;
import org.openjdk.jmh.annotations.Scope;
import org.openjdk.jmh.annotations.State;
import org.openjdk.jmh.annotations.Warmup;
import org.openjdk.jmh.infra.Blackhole;

@Fork(3)
@Warmup(iterations = 5, time = 20, timeUnit = TimeUnit.SECONDS)
@Measurement(iterations = 10, time = 30, timeUnit = TimeUnit.SECONDS)
@BenchmarkMode(Mode.Throughput)
@OutputTimeUnit(TimeUnit.MILLISECONDS)
@State(Scope.Benchmark)
public class Benchmark {

  private static final Object THING = new Object();

  @org.openjdk.jmh.annotations.Benchmark
  public void of(final Blackhole blackhole) {
    blackhole.consume(List.of(THING));
  }

  @org.openjdk.jmh.annotations.Benchmark
  public void asList(final Blackhole blackhole) {
    blackhole.consume(Arrays.asList(THING));
  }

  @org.openjdk.jmh.annotations.Benchmark
  public void singletonList(final Blackhole blackhole) {
    blackhole.consume(Collections.singletonList(THING));
  }
}
```

EXPECTED VERSUS ACTUAL BEHAVIOR :
EXPECTED -
For one item lists, Arrays::asList is slower than Collections::singletonList and List::of and the later two are approximately as performant as eachother on x64 and ARM64
ACTUAL -
Collections::singletonList is ~4x slower than Arrays::asList and ~8x slower than List::of on ARM64

---------- BEGIN SOURCE ----------
package brutus;


import java.util.Arrays;
import java.util.Collections;
import java.util.List;
import java.util.concurrent.TimeUnit;
import org.openjdk.jmh.annotations.BenchmarkMode;
import org.openjdk.jmh.annotations.Fork;
import org.openjdk.jmh.annotations.Measurement;
import org.openjdk.jmh.annotations.Mode;
import org.openjdk.jmh.annotations.OutputTimeUnit;
import org.openjdk.jmh.annotations.Scope;
import org.openjdk.jmh.annotations.State;
import org.openjdk.jmh.annotations.Warmup;
import org.openjdk.jmh.infra.Blackhole;

@Fork(3)
@Warmup(iterations = 5, time = 20, timeUnit = TimeUnit.SECONDS)
@Measurement(iterations = 10, time = 30, timeUnit = TimeUnit.SECONDS)
@BenchmarkMode(Mode.Throughput)
@OutputTimeUnit(TimeUnit.MILLISECONDS)
@State(Scope.Benchmark)
public class Benchmark {

  private static final Object THING = new Object();

  @org.openjdk.jmh.annotations.Benchmark
  public void of(final Blackhole blackhole) {
    blackhole.consume(List.of(THING));
  }

  @org.openjdk.jmh.annotations.Benchmark
  public void asList(final Blackhole blackhole) {
    blackhole.consume(Arrays.asList(THING));
  }

  @org.openjdk.jmh.annotations.Benchmark
  public void singletonList(final Blackhole blackhole) {
    blackhole.consume(Collections.singletonList(THING));
  }
}

---------- END SOURCE ----------