q 8d7721c775 B7: add network time bootstrap and persistent B7C cache

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).

2026-10-02 20:18:25 +03:00

2.1 KiB

Raw Blame History

B7C cache write — preflight (prepared 2026-10-02, NOT written)

Candidate

cache-extlinux-b7c.img 14fe453a… (134217728 B, ext2, label aurora-b7c), built on 480s by mkfs-cache-b7c.sh = exact T4 recipe (r2/t4/mkfs-cache-t4.sh → 9d3bc890) with only initrd/label/UUID changed. Rebuild is bit-identical (reproducible). e2fsck clean. Content: Image.gz-dtb + aurora-mm1.dtb + cmdline = b5b/components (sha-checked, identical to T4), initramfs-b7c.cpio.gz f219b04d… = the RAM image that passed B7C 3/3 class A (aurora-mm1-b7c.img d0c4cc00 carries the same ramdisk). extlinux: default b7c / label b7c.

lk2nd emulator (emu/)

S4 regression on the current flash (T4 9d3bc890): "Trying to boot 't4'", HARNESS lines == r2/t4/emu (emulator unchanged).
S5 candidate: "Trying to boot 'b7c'", boot_linux reached, ramdisk sha f219b04d (12628896 B); kernel/dtb/cmdline HARNESS lines identical to T4.
Fault variants (no extlinux.conf, no initrd, zeroed cache): "Bootable file system not found. Reverting to android boot" → boot = emmc1 fallback (as for T4).

Write (B7C-write.sh, derived 1:1 from T4-write.sh)

Gates: loader sha; candidate sha+size; rollback image r2/t4/cache-extlinux-t4.img == 9d3bc890; T4 readback backup == 9d3bc890; 9008 present; identity (Sahara serial <SAHARA_SERIAL>), cache geometry (0x4118c000 / 0x8000000); GPT primary == A3; boot == emmc1 (e9579f33); aboot == lk1st-autoboot (3b8cd266); current cache == T4 9d3bc890. Write cache only → readback cmp → CACHE_MATCH → verify GPT p/b, tzbak, boot, aboot, sbl1/rpm/tz/hyp/recovery/modemst1/2/fsg/fsc/persist == A3, modem/system/userdata == A4pre → B7C_WRITE_VERIFIED. Logs logs/b7/b7c/W/. Nothing else is written. Entry: HW EDL (battery on, D+ → GND, plug USB, release) — operator's hands. Rollback: B7C-rollback.sh → T4 cache 9d3bc890 (also via HW EDL).

After the write (proposal)

Full power-off (USB, then battery), ≥ 90 s, normal power-on (no RESET): expect lk "Trying to boot 'b7c'", Linux, NCM, telnetd, TIME SET at REG registered/attached, START OK, SNTP check < 1 s. One verification boot; more only with a GO.

2.1 KiB Raw Blame History