test(rp2040): controlled PICOBOOT-only hardware validation with the #1161 pinned artifact

[zackees]

Goal

Once the #1161 board re-enumerates (currently USB-bricked; requires physical BOOTSEL/replug), run a controlled PICOBOOT-only load to cleanly isolate the observed write-stall failure to the mass-storage deployment path — and to validate the PICOBOOT-primary transport flip (sibling sub-issue) before it ships as the default.

Pinned artifact (do not rebuild between runs)

• Size: 882,688 bytes
• SHA-256: 1C770999678BB7563A81D36C696DC44C074A712E25F0DE68A8A83F859E8AF0E0
• 1,724 UF2 blocks of 512 bytes, RP2040 family ID 0xE48BFF56, target 0x10000000–0x1006BB00

Preconditions

• PICOBOOT MI_01 has a working WinUSB binding (see the driver-preflight sub-issue; last observed state was problem code 28).
• Record before every attempt, per the investigate(rp2040): readable BOOTSEL volume stalls on first write; make watchdog cleanup handle-safe #1161 proposal: exact executable path + hash, source commit, daemon port, daemon binary (the host has shown version skew: global 2.5.2, venv exe 2.5.3, venv metadata 2.5.2).

Procedure

1. Board in BOOTSEL, enumerated as 2E8A:0003.
2. picotool info probe — record output.
3. picotool load <pinned.uf2> -x (add -f only if starting from application mode) — via the fbuild-managed picotool so the tested binary matches what fbuild ships.
4. During the load: confirm no access to the RPI-RP2 volume (no open handles on G:), and monitor the System event log for new Disk 51/153 events.
5. Verify application re-enumeration and running firmware.

Interpretation

• Success with no volume access and no Disk 51/153 events → the investigate(rp2040): readable BOOTSEL volume stalls on first write; make watchdog cleanup handle-safe #1161 failure is isolated to the mass-storage path; PICOBOOT-primary is validated as the fix direction.
• Failure over PICOBOOT too → the fault is deeper (device QSPI programming, USB transport, or host controller) and the transport flip alone won't resolve investigate(rp2040): readable BOOTSEL volume stalls on first write; make watchdog cleanup handle-safe #1161; a USB trace becomes the next step.

Part of the meta issue tracking the #1161 transport strategy change. Defaults chosen by the drafting agent are listed inline — edit as needed.