[Perf] Windows/arm64: 2 Regressions on 8/2/2026 8:51:09 AM +00:00

### Run Information

Name | Value
-- | --
Architecture | arm64
OS | Windows 10.0.22621
Queue | Pixel
Baseline | [fdccdc6954791fcde7ffa2834d75930c0efa5456](https://github.com/dotnet/runtime/commit/fdccdc6954791fcde7ffa2834d75930c0efa5456)
Compare | [c25173808a41517b547a0880a424d64aff9c7e80](https://github.com/dotnet/runtime/commit/c25173808a41517b547a0880a424d64aff9c7e80)
Diff | [Diff](https://github.com/dotnet/runtime/compare/fdccdc6954791fcde7ffa2834d75930c0efa5456...c25173808a41517b547a0880a424d64aff9c7e80)
Configs | CodeGenType:JIT, CompilationMode:tiered, LinkingType:static, RunKind:android_scenarios, RuntimeType:coreclr

### Regressions in SOD - Android HelloWorld Extracted Size

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

• [SOD - Android HelloWorld Extracted Size - pub\META-INF\ANDROIDD.RSA]()
• [ADX - Test Multi Config Graph]()

• [SOD - Android HelloWorld Extracted Size - Aggregate - .RSA]()
• [ADX - Test Multi Config Graph]()

![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 'SOD - Android HelloWorld Extracted Size*'
```

### SOD - Android HelloWorld Extracted Size

#### 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 performance/scripts/benchmarks_ci.py and the benchmarking workflow linked in the issue. Run the provided SOD - Android HelloWorld Extracted Size benchmark, then compare the baseline and test commits and inspect the linked report. Done means the arm64 Windows regression is explained and the benchmark no longer shows the reported size increase.

Written by the indexing model from the issue text.