Add staged MAUI desktop benchmarks

[LoopedBard3]

Summary

This is an alternative implementation to #5161 that keeps all changes in dotnet/performance while avoiding runtime rewrites of MAUI benchmark entrypoints.

• resolves the matching dotnet/maui branch to an exact SHA on the build agent and stages a sparse checkout in the correlation payload
• injects stable MSBuild imports into the staged checkout, retaining every original Program.cs
• publishes the existing BenchmarkDotNet.Extensions harness and aligns MAUI's benchmark projects to the performance repo's BDN version
• runs Core, XAML, and Graphics as isolated Helix work items through the existing benchmarks_ci.py build, collection, and upload path
• records the MAUI repository, branch, commit, and selected toolchain in PerfLab result metadata
• uses the default out-of-process toolchain for Core and Graphics; XAML uses InProcessEmit because BDN's child reference gatherer otherwise rebuilds MAUI's incompatible legacy TFMs

Validation

• exercised the real PreparePayloadWorkItems MSBuild target against the current net11.0 MAUI branch
• executed representative Core, XAML, and Graphics benchmarks through simulated Helix work-item roots
• verified public one-iteration BDN arguments are forwarded without the category filter that would exclude uncategorized MAUI benchmarks
• ran all 46 BenchmarkDotNet.Extensions.Tests
• ran the targeted Python tests and strict Pyright checks

[Copilot]

🔵 Needs a closer look

It introduces a new end-to-end Helix scenario with external repo staging and pipeline wiring where correctness depends on integration behavior that can’t be fully validated from the diff alone.

Pull request overview

Adds a new “staged MAUI desktop benchmarks” scenario that prepares a sparse MAUI checkout on the build agent, injects stable MSBuild imports/entrypoint overrides in the staged payload, and runs Core/XAML/Graphics BenchmarkDotNet suites as isolated Helix work items while recording MAUI source/toolchain metadata in PerfLab output.

Changes:

• Introduces mauidesktopbenchmarks scenario payload preparation (pre.py) and Helix execution wrapper (test.py) plus an MSBuild overlay (props/targets + custom entrypoint).
• Updates RecommendedConfig.Create to (a) allow an optional explicit IToolchain and (b) omit MandatoryCategoryValidator when no mandatory categories are provided; adds unit tests for this behavior.
• Wires the new run kind into CI (new Helix project + pipeline jobs) and enables strict Pyright checking for the new scenario folder.

File summaries

File	Description
src/tests/harness/BenchmarkDotNet.Extensions.Tests/RecommendedConfigTests.cs	Adds unit coverage for the “no mandatory categories => no validator” behavior.
src/scenarios/mauidesktopbenchmarks/test.py	Runs a staged MAUI suite in a Helix work item, forwarding BDN args and uploading expected artifacts.
src/scenarios/mauidesktopbenchmarks/pre.py	Prepares the sparse MAUI checkout + overlay, publishes the harness, and writes the source manifest into the payload.
src/scenarios/mauidesktopbenchmarks/overlay/PerfLabEntry.cs	Provides a stable benchmark entrypoint that selects toolchain mode and uses PerfLab’s recommended config/artifacts location.
src/scenarios/mauidesktopbenchmarks/overlay/MauiPerfLab.targets	Injects the PerfLab entrypoint + aligns BenchmarkDotNet package versions and references the published harness binaries.
src/scenarios/mauidesktopbenchmarks/overlay/MauiPerfLab.props	Forces desktop-only target framework selection by disabling non-Windows MAUI target frameworks.
src/harness/BenchmarkDotNet.Extensions/RecommendedConfig.cs	Adds optional toolchain support and avoids adding MandatoryCategoryValidator when mandatory categories are empty.
scripts/tests/test_maui_desktop_benchmarks.py	Adds targeted Python tests for framework->branch mapping and BDN argument shaping.
scripts/run_performance_job.py	Ensures binlogs produced by the new scenario’s prepare step are copied into CI artifacts.
pyrightconfig.json	Includes the new scenario directory as a strict Pyright execution environment.
eng/pipelines/sdk-perf-jobs.yml	Adds public and internal pipeline jobs for the new maui_desktop_benchmarks run kind.
eng/performance/maui_desktop_benchmarks.proj	Adds the Helix project wiring: prepare payload work item + 3 suite work items (core/xaml/graphics).

Review details

• Files reviewed: 12/12 changed files
• Comments generated: 0
• Review effort level: Lite

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.