M2-PWRAP-01 physical result
> **Historical record.** The dates, candidate identity, "current" claims, > next steps and permissions below belong to this recorded boundary. See > [current state](../CURRENT_PLATFORM_STATE.md) for the latest physically observed result.2026-09-09, owner-supplied photographs 10.jpg and 50.jpg. Originals are under out/m1-y2b250-time32/, but both screens identify M2-PWRAP-01, kernel 6.18.0-y2-m2-pwrap1. Retained private copies/manifest: evidence-private/20260909-m2-pwrap-result/. This is real photographed hardware, not a rendered fixture. The filename 10 is not the displayed heartbeat.
| Field | 10.jpg | 50.jpg |
|---|---|---|
| BEAT / FRAME | 18 / 116 | 50 / 308 |
| Stage | WAIT 1S / LOOP LIVE | STOP: RESTORE ANDROID |
| Uptime / idle seconds | 20.20 / 18.24 | 55.72 / 50.56 |
| MemTotal | 22208 kB | 22208 kB |
| CPU part / online | 0xc07 / 0 | 0xc07 / 0 |
| DT RAM | 24M + 512K, D08 STATIC | same |
| PWRAP RC / VALID | 0 / 3 | 0 / 3 |
| MUX / WRAP / WACS / INIT | 0 / 1 / 1 / 1 | same |
| ARB / PRE | 0000007F / 003000F7 | same |
| CID / VUSB | 00002023 / 0000C000 | same |
| POST | 0030C000 | same |
| Timer IRQ / last error | TIMER IRQ ERR: -61 / TIMER -61 | same |
| Sleep / proc / sysfs return | 0 / 0 / 0 | same |
| Previous frame write | 804 | 804 |
| Watchdog / framebuffer | STOPPED / GUARD OK | same |
Photo SHA-256:
10.jpg(267217 bytes):9e8c226771d06bc7a7063be50c61eb589fb16680e8f1bb312f3cd34818c9bd21.50.jpg(346864 bytes):853c20550372a6b4e4f89db0b9f63cd76b67f30123fe8b29d55e7f6c599e6225.
What this establishes
CONFIRMED narrow prerequisite: on this boot the inherited AP PWRAP transport completed CID and VUSB reads and returned to idle; CID low byte is MT6323 0x23. VUSB 0xc000 has software enable bit 14 and reported enable bit 15 set. This supports preserving the inherited supply for the next USB study. No rail enable write is needed merely to discover its state. These fields are one cached snapshot, not two independent bus reads or evidence of continuous rail monitoring. They do not measure 3.3 V or qualify battery/charger telemetry.
The kernel/PID1 remains observable through BEAT 50, with 35.52 seconds of kernel uptime between the photographed states. Those are kernel times, not an independent host-clock accuracy measurement or repeated-boot qualification. MemTotal is 21.6875 MiB under D08; application/subsystem working-set sufficiency is untested.
Timer diagnostic cause
Pinned v6.18 drivers/clocksource/timer-of.c:66–68 registers the interrupt action with np->full_name, while timer-mediatek.c names the clock-event object mtk-clkevt. The PID1 parser incorrectly searches for that clock-event name in /proc/interrupts. The board DT node is timer@10008000; a source-backed parser fix must match the actual action token and numeric IRQ row. The raw hardware interrupt listing is not supplied, so the precise rendered action string/count is still to be confirmed. -61 is the parser's no-data result; it does not show that timer IRQs are absent. Working sleeps and progress remain confirmed.
Identity and remaining boundaries
The associated candidate is the retained M2-PWRAP-01 image, SHA-256 c51102c861d65d39f5af2425ed73e9c1ba21d8adaae1b016f450a58630946ac0. The screenshots identify the build, but do not independently confirm selected or flashed/readback SHA-256. Host elapsed time, cable state, reset history and restoration outcome have not been supplied. Do not infer them from the folder name or photographs. M1 remains complete; #23, #27 and M2 stay open.
Next: correct timer reporting and resolve normal-handoff USB clocks/PHY/DMA. Only the scoped PWRAP/VUSB prerequisite has advanced to physical evidence.