[Perf] Linux/arm64: 1 Improvement on 8/20/2026 1:34:45 PM +00:00

### Run Information

Name | Value
-- | --
Architecture | arm64
OS | azurelinux 3.0
Queue | CobaltAzureLinux
Baseline | [361539de30ffe5c5ccb7cbdfa230747411ec7faf](https://github.com/dotnet/runtime/commit/361539de30ffe5c5ccb7cbdfa230747411ec7faf)
Compare | [6ecee4dd49b5ab3f97c467372fee3affa17a5a9b](https://github.com/dotnet/runtime/commit/6ecee4dd49b5ab3f97c467372fee3affa17a5a9b)
Diff | [Diff](https://github.com/dotnet/runtime/compare/361539de30ffe5c5ccb7cbdfa230747411ec7faf...6ecee4dd49b5ab3f97c467372fee3affa17a5a9b)
Configs | CompilationMode:tiered, R2RType:r2r_interpreter, RunKind:micro

### Improvements in System.IO.Tests.Perf_FileStream

Benchmark | Baseline | Test | Test/Base | Test Quality | Edge Detector | Baseline IR | Compare IR | IR Ratio
-- | -- | -- | -- | -- | -- | -- | -- | --
|

• [FlushAsync - Duration of single invocation]()
• 📝 - [Benchmark Source]()
• [ADX - Test Multi Config Graph]()

![graph]()
[Test Report]()

### Repro
General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

```cmd
git clone https://github.com/dotnet/performance.git
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.IO.Tests.Perf_FileStream*'
```

### System.IO.Tests.Perf_FileStream.FlushAsync(fileSize: 1024, options: Asynchronous)

#### ETL Files

#### Histogram

#### JIT Disasms

### Docs

[Profiling workflow for dotnet/runtime repository](https://github.com/dotnet/performance/blob/master/docs/profiling-workflow-dotnet-runtime.md)
[Benchmarking workflow for dotnet/runtime repository](https://github.com/dotnet/performance/blob/master/docs/benchmarking-workflow-dotnet-runtime.md)

Start with the System.IO.Tests.Perf_FileStream benchmark source in dotnet/performance/src/benchmarks/micro/libraries/System.IO.FileSystem/Perf.FileStream.cs, especially lines 412-427, and review the linked FlushAsync report. Run the provided benchmarks_ci.py command and compare the reported results; the issue is complete only when the performance change is understood and an appropriate follow-up is identified.

Written by the indexing model from the issue text.