redbear-power: multi-threaded collector + Tier 1/2 display enhancements
Tier 1 (display Phase 0 infrastructure):
- sched.rs: read /scheme/sys/stat and /scheme/sys/sched for per-CPU
scheduler stats (switches, steals, queue_depth, IRQs)
- msr.rs: HWP capabilities/requests, RAPL domain energy readings
- process.rs: derive_sched_policy() for SCHED column
- render.rs: System tab shows RAPL power, scheduler stats; Per-CPU
table has IRQs/steals columns; HWP expansion rows
- acpi.rs: ACPI throttle status reading
Multi-threaded collector (exercises Phase 0 threading):
- collector.rs: parallel collect() using thread::scope + Barrier,
exercises pthread_create, futex barriers, sched_setaffinity
- app.rs: sequential per-CPU loop replaced with parallel collector
- render.rs: System tab shows Collector stats (threads, pinned, barrier)
Tier 2 kernel wiring (submodule pointer updates):
- kernel: per-CPU sched stats (/scheme/sys/sched), NUMA-aware
scheduling, numa::init_default() at boot
- relibc: robust mutex cleanup in exit_current_thread()
vasilito
eb53e8190a