Parent: #126
Hypothesis
helpers::connect_all + proxy::connect_all_default_consoles may be doing redundant connects or slow connects that account for the perf delta. Without per-phase numbers we can't tell.
Steps
- Add timing instrumentation in
patina_bin/bin/patina_x64.rs::MaaBootManager::execute() around each step (use patina_performance::Measurement or raw boot_services.get_time()).
- Capture FPDT log, parse with
Measure-BootPerf.ps1.
- Compare per-phase times against C BDS phase times from prior LKG.
- If
connect_all dominates, consider scope reduction (only connect required handles for boot, not exhaustive).
Exit
Per-phase timing data available; bottleneck phase identified; mitigation path documented.
Parent: #126
Hypothesis
helpers::connect_all+proxy::connect_all_default_consolesmay be doing redundant connects or slow connects that account for the perf delta. Without per-phase numbers we can't tell.Steps
patina_bin/bin/patina_x64.rs::MaaBootManager::execute()around each step (usepatina_performance::Measurementor rawboot_services.get_time()).Measure-BootPerf.ps1.connect_alldominates, consider scope reduction (only connect required handles for boot, not exhaustive).Exit
Per-phase timing data available; bottleneck phase identified; mitigation path documented.