8d7721c775
B7A: QMI DMS time vs VNIIFTRI NTP (DMS = UTC - 0.28 s, sigma 1 ms; NITZ = DMS truncated to 1 s). B7B: a 1970 -> 2026 date -s step is harmless for the modem stack. B7C: aurora-modem probes DMS/NITZ along the start flow and steps CLOCK_REALTIME once at the first valid sample (REG registered/attached), never RTC; background SNTP check (query only) after START OK. 3/3 class-A cold boots PASS (residual +0.04/+0.36/+0.84 s), cache 14fe453a written (B7C_WRITE_VERIFIED) and verified by a normal power-on (V1, +0.26 s). Also publishes the prerequisite R2 (boot-hang / lk eMMC investigation, T4 telnet baseline that B7C builds on) and R3 (autonomous cold boot, NCM loss) material, and extends tools/publish-sanitize.py to r2/, r3/, b7/. Binary images, initramfs, busybox and raw logs stay out (see b7/*/SHA256SUMS).
1.7 KiB
1.7 KiB
B7C c2 — ANOMALY: unexpected board reset during start (not a time-bootstrap result). Nothing touched on the board.
- Off-time: last UART line of the previous boot 13:14:36.2 (local) → SBL1 Start 13:16:34.05 = ~118 s. RESET-held power-on → lk1st fastboot;
fastboot boot10:16:46 UTC rc=0. - Linux: RTC at boot 15 s (PMIC RTC reset; ~12 s of it spent in lk fastboot) → class A. EMMC-RO 30/30, telnetd, B1_OK=1, aurora-modem start at 9.5 s.
- Time probes 1-2 (QMI_READY, SIM_READY): DMS 3232/3399 ms → INVALID year<2026 (same as c1). Last line: uptime 13.59
SET: qmicli … --dms-set-operating-mode=online. - Then 18.4 s of UART silence, no panic/Oops/watchdog text, and a new SBL1 Start at 13:17:20.09 (pm_driver_init 7564 µs, batt 4194 mV).
lk #2:
pm8x41_get_is_cold_boot: cold boot, then fastboot again (noTrying to boot), still waiting in fastboot (18d1:d00d) — this lk enters fastboot only with RESIN (RESET) asserted. - Host: NCM never came up (board reset before/while NCM came up); telnet timeouts; driver stopped by PID. Host driver bug "NCM … ok" after timeout fixed (now aborts).
- Open: why the reset. Facts: silent (no kernel message), cold-boot PON, RESIN asserted at the second lk. Candidates: RESET held/pressed again ~46 s after power-on (PMIC RESIN long-press reset), or a power interruption at radio power-up (SET online) with RESET pressed. Ask the operator how RESET was handled.
- Operator (2026-10-02): RESET held only ~2 s; "может аккум выпал" → most likely cause = battery fell out / lost contact (known flaky battery contact) ~46 s after power-on. c2 = INVALID run (external power interruption), not counted, not a B7C failure. Repeated as c2b with the same procedure.