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).
824 B
B7C c5 — class-A cold boot #1: PASS
Power-off ≥ 90 s (operator); one attempt, fastboot caught first time. SBL1 16:23:05 UTC, fastboot boot rc=0, RTC at boot 6 s → class A (PMIC POR).
(Off-time not measurable from UART: the running B7C Linux printed nothing between logger start 16:16:22 and power-off; RTC 6 s proves the POR.)
Probes: 1-4 INVALID (DMS 3221…7179 ms = modem uptime), 5 INVALID with bogus NITZ 1980-01-06 00:00:13 (rejected by year gate),
6 REG registered/attached VALID (NITZ 16:23:37, Δ +0.76 s) → TIME SET at uptime 27.46 s (33 → 2026-10-02 16:23:38), 7-8 VALID, no re-set.
START OK 78.8 s, Beeline home, LTE ping 3/3. SNTP check +0.041 s (best, ntp3, 50 ms RTT); verify ntp2 +0.048…+0.055 s. Clock kept.
UNSYNC kept, RTC untouched, eMMC writes 0, dmesg-errs 0, no FAIL/rollback.