Skip to content

omamobile-device-tegu 0.2.9, device/tegu: the radio partitions tools/nv-backup reads - #407

Merged
derekross merged 3 commits into
derekross:mainfrom
W4HO-ham:tegu-nv-parts
Nov 4, 2026
Merged

derekross merged 3 commits into
derekross:mainfrom
W4HO-ham:tegu-nv-parts

Conversation

@W4HO-ham

@W4HO-ham W4HO-ham commented Nov 3, 2026

Copy link
Copy Markdown
Contributor

tools/nv-backup refused the 9a ("names no NV_PARTS"), so its NV check had to be done by hand. This adds NV_PARTS and NV_BACKUP_EXTRA to device/tegu/device.conf: the 7 Pro's partitions, the ones the 9a's hand-made backups have held since 2026-10-21.

On the palmtop (2026-11-03, the daily kernel, the modem up since boot and run for days before), tools/nv-backup -d tegu --check against the 2026-10-28 backup taken after the last stock boot reads efs, efs_backup, modem_userdata, persist, modem_a and modem_b the same. devinfo differs, as expected.

tools/pr-check passes.

…nv-backup reads

tools/nv-backup refused the 9a ("names no NV_PARTS"), so its NV check had to be done by hand. The 9a has the 7 Pro's partitions, the ones its hand-made backups have held since 2026-10-21. With them, --check against the 2026-10-28 backup reads every NV partition the same after days of the modem running. The package's copy of the descriptor follows it, as its test asks.
@W4HO-ham W4HO-ham changed the title device/tegu: the radio partitions tools/nv-backup reads omamobile-device-tegu 0.2.9, device/tegu: the radio partitions tools/nv-backup reads Nov 3, 2026
@W4HO-ham

W4HO-ham commented Nov 3, 2026

Copy link
Copy Markdown
Contributor Author

CI caught that the device package's copy of device.conf didn't follow the descriptor. It does now, as omamobile-device-tegu 0.2.9.

@derekross

Copy link
Copy Markdown
Owner

Thanks, @W4HO-ham. This is a small change with a real payoff: the 9a gets the same read-only
backup and check as the 7 Pro, and nothing in it writes. I read tools/nv-backup again for the
new names: it finds each one by PARTNAME, refuses anything writable, mounted or on the storage
guard's allowlist, and only reads, so the descriptor values are all it needed.

One thing before merge: #403 also bumps omamobile-device-tegu to 0.2.9. Because both PRs write
the same version line, git merges them without a conflict, and the second one would ship different
contents under a version that's already published. I'll set the next free version on whichever
lands second, so there's nothing for you to do unless you'd rather renumber yourself.

Two small things, either here or later:

  • Your hand-made backups on 10-26 and 10-28 held eight partitions, metadata included. Could you
    either add metadata to NV_BACKUP_EXTRA (it changes, so it belongs there with devinfo), or
    reword the comment so it doesn't say these are the same partitions?
  • A short tegu NOTES line with today's --check result (date, kernel release, which partitions
    read the same; no checksums) would keep the evidence in the tree.

The hand-made backups of 10-26 and 10-28 held metadata too. It changes (the bootloader and stock write it), so it goes with devinfo in NV_BACKUP_EXTRA: copied for a full way back, not checked.
@W4HO-ham

W4HO-ham commented Nov 3, 2026

Copy link
Copy Markdown
Contributor Author

Thanks. I added metadata to NV_BACKUP_EXTRA (aedf0b5), since the hand-made backups held it and it changes. The tegu NOTES line with a fresh --check (date, release, which partitions read the same) comes after the palmtop's sleep series, which holds the USB link until about 18:00 MST. Renumbering at merge is fine with me.

Today's --check against the 2026-10-28 backup, after the day's reboots: every NV partition and metadata read the same, and devinfo differs as expected.
@W4HO-ham

W4HO-ham commented Nov 3, 2026 •

Copy link
Copy Markdown
Contributor Author

The tegu NOTES line is in (52ad7ab): tools/nv-backup -d tegu --check on this branch against the 2026-10-28 backup, run today after the day's reboots. All six NV partitions and metadata read the same, and devinfo differs as expected. Release and partitions are named there, no checksums.

@derekross
derekross merged commit e95b477 into derekross:main Nov 4, 2026
2 checks passed
@derekross

Copy link
Copy Markdown
Owner

Thanks, @W4HO-ham, both follow-ups are in and they read well: metadata sits with the
extras for the right reason, and the NOTES entry gives the date, the kernel release and which
partitions read the same, with no checksums. That's all I asked for.

0.2.9 was still free on main, so this went in as it is, and the other pull requests that claimed
it (#403, #423, #371) move up a number at their merge. Merged, thank you.

robertzas pushed a commit to robertzas/omarchy-mobile that referenced this pull request Nov 14, 2026
…nv-backup reads (derekross#407)

tools/nv-backup refused the 9a ("names no NV_PARTS"), so its NV check had
to be done by hand. The 9a has the 7 Pro's six radio partitions, the ones
its hand-made backups have held since 2026-10-21; metadata joins devinfo
in NV_BACKUP_EXTRA (the bootloader and stock write it, so it is copied for
a full way back, not checked). Both copies of the descriptor carry it.
With them, --check against the 2026-10-28 backup read every NV partition
and metadata the same after days of the modem running (tegu NOTES).
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants