Skip to content

device/panther: SUSPEND-PLAN and BLUETOOTH-PLAN, the Pixel 7's sleep and Bluetooth records - #852

Merged
derekross merged 2 commits into
derekross:mainfrom
nesmeck:panther-docs
Nov 18, 2026
Merged

derekross merged 2 commits into
derekross:mainfrom
nesmeck:panther-docs

Conversation

@nesmeck

@nesmeck nesmeck commented Nov 16, 2026 •

Copy link
Copy Markdown
Contributor

The Pixel 7's two plans that until now lived only in my tree: device/panther/SUSPEND-PLAN.md and device/panther/BLUETOOTH-PLAN.md. Documentation only, no package or kernel change. I'm opening them first so the pull requests that follow can cite results on main, as you preferred on #827.

SUSPEND-PLAN, s2idle on the Pixel 7 Pro's path, stages Z1 to Z6 and P7, with their results from 2026-11-05 to 2026-11-13:

  • Z1: the RTC alarm wakes the phone, with the S2MPG12 fix that merged as linux-omarchy-mobile#201.
  • Z2: Wi-Fi's hang after a resume is the firmware trapping on the freeze path (rtcwake -m freeze, 4 of about 73); on systemctl suspend it hasn't trapped in 256 resumes. The record decoded, the driver's report, and the check that reloads the stack (Z2l, Z2m: it ran after 70 of 70 resumes and found the firmware answering). This is the evidence behind device-panther 0.1.10: Wi-Fi at boot with the Pixel 7 Pro's guarded start, unloaded around suspend (from #419) #475's default, Wi-Fi unloaded around sleep, and behind the check and the two brcmfmac patches to come.
  • Z3, Z4: the sleep hook in the package, and Power and Volume Down as the wake.
  • Z6: standby numbers: asleep with Wi-Fi kept, about 284-296 mA with the memory bus at 2028 MHz and 248-251 mA at 1014 MHz; at 1014 MHz with Wi-Fi unloaded, 184-200 mA; why deep doesn't wake; the power domains kernel that resets the phone, and the flat one that sleeps. The memory bus floor is a kernel pull request to come.
  • P7: power off powers off on a kernel with linux-omarchy-mobile#233 (in the kernel releases since kernel-2026.11.13); with the cable in, the bootloader's charger boot, which is what omamobile-device-panther: off-mode charging instead of starting the phone on its cable #788 handles.

BLUETOOTH-PLAN, the BCM4389 on the APM UART, gates BT1 to BT3 with their results from 2026-11-01 (the kernel side merged as linux-omarchy-mobile#196, the options as #304), and the headset profile against WirePlumber's switch, which is why BT_RFCOMM stayed out of #304. The package side is a pull request to come.

Both status lines say what has merged since and what is still to come. SUSPEND-PLAN cites DISPLAY-PLAN D5 (the screen after s2idle), which comes with DISPLAY-PLAN's update in the 90 Hz pull request after linux-omarchy-mobile#181. Kernel names, local build numbers and the dates are as they were on the phone; I've checked the files for identifiers.

Checks: tools/pr-check passes.

🤖 Generated with Claude Code

…and Bluetooth records

The two plans that until now lived only in a local tree, so that the
pull requests that follow (the Wi-Fi check, the brcmfmac patches, the
memory bus floor, the Bluetooth package) can cite results that are on
main.

SUSPEND-PLAN: s2idle on the Pixel 7 Pro's path, stages Z1 to Z6 and P7
with their results from 2026-11-05 to 2026-11-13: the RTC alarm, the
Wi-Fi firmware trap at a resume on the freeze path and the check that
reloads it, the sleep hook, the keys as the wake, standby numbers with
the memory bus at 2028 and 1014 MHz and with Wi-Fi kept, why deep
doesn't wake, and power off on a kernel with linux-omarchy-mobile#233.

BLUETOOTH-PLAN: the BCM4389 on the APM UART, gates BT1 to BT3 with
their results from 2026-11-01, and the headset profile against
WirePlumber's switch.

Both status lines say what has merged since and what is still to come.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
@derekross

Copy link
Copy Markdown
Owner

Thank you, Alejandro. I read both plans from top to bottom, and this is the kind of record I'd like every device to have: one change per boot, every result with its date and kernel, the two sleep paths counted apart, and the readings that turned out wrong (Z2e's PME theory, Z6g's empty pstore) corrected in place where a reader can follow them. I found no identifiers in either file, and the pull requests and merges the status lines name are as you say. The Wi-Fi trap work in Z2 and the same-evening MIF runs in Z6j are especially good.

Five small things in SUSPEND-PLAN before I merge it:

  1. In "What is known", device/tegu/SLEEP-PLAN.md is the Pixel 9a's, not the Pixel 9 Pro's.
  2. The status line says the upower drop-in waits for a kernel release with linux-omarchy-mobile#233. That release exists: kernel-2026.11.13 and kernel-2026.11.15 both carry it. What it still waits for is the charging boot (omamobile-device-panther: off-mode charging instead of starting the phone on its cable #788). In P7's last takeaway, the Pixel 7 Pro's charger-mode handling landed as omamobile-device-cheetah 0.1.55: suspend on the 7 Pro: double tap from sleep, auto-sleep by default, the charging boot, power off at a critical battery #730 (device-cheetah 0.1.55), which carried device-cheetah 0.1.55: handle the bootloader's off-mode charging boot #554's commits.
  3. The standby numbers in the status line (and in the description): 284 mA at 2028 MHz was with Wi-Fi kept, while 184-200 mA at 1014 MHz is with Wi-Fi unloaded, so side by side they make the bus rate look worth twice what Z6j measured. Could it give each number its condition, for example "about 284 mA with the bus at 2028 MHz and Wi-Fi kept; at 1014 MHz about 248 mA kept and 184-200 mA unloaded"?
  4. Z2k's boot: "the phone fell to slot A's stock recovery". Did only the recovery screen come up, or did Android itself start at any point? If only recovery, one sentence saying so is enough. If Android started, please add the read-only backup that followed, as your 2026-10-27 notes did.
  5. Two pointers lead to text that isn't on main yet: DISPLAY-PLAN D5, and "(NOTES)" for the local repository and the temporary pacman config. A "(to come with the 90 Hz pull request)" or similar beside each is enough.

One thing I'm taking from this for the Pixel 7 Pro: your Z1 explains the early alarm wakes I never tracked down there, and Z6k's 52 mA for keeping Wi-Fi loaded is a number I should measure on mine.

- The Pixel 9a, not the Pixel 9 Pro, for device/tegu/SLEEP-PLAN.md.
- The status line: the kernel releases since kernel-2026.11.13 carry
  linux-omarchy-mobile#233, so the upower drop-in waits only for
  off-mode charging (derekross#788). P7's last takeaway names derekross#730, which
  carried derekross#554's commits.
- The standby numbers in the status line each with their condition:
  Wi-Fi kept at 2028 and at 1014 MHz, Wi-Fi unloaded at 1014 MHz.
- Z2k's boot: only stock recovery came up, not Android.
- DISPLAY-PLAN D5 and the NOTES entries, which come with the 90 Hz pull
  request, say so.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@nesmeck

nesmeck commented Nov 18, 2026

Copy link
Copy Markdown
Contributor Author

Thank you for reading both plans so closely. All five are in one commit on top, 45b31a0:

  1. The Pixel 9a for device/tegu/SLEEP-PLAN.md.
  2. The status line now says the kernel releases since kernel-2026.11.13 carry omamobile-shell 0.2.53: an Updates entry in the app drawer, and update sheet fixes from its first run #233, so the upower drop-in waits only for omamobile-device-panther: off-mode charging instead of starting the phone on its cable #788. P7's last takeaway names omamobile-device-cheetah 0.1.55: suspend on the 7 Pro: double tap from sleep, auto-sleep by default, the charging boot, power off at a critical battery #730 (device-cheetah 0.1.55), which carried device-cheetah 0.1.55: handle the bootloader's off-mode charging boot #554's commits.
  3. The standby numbers each with their condition. With Wi-Fi kept, about 284-296 mA at 2028 MHz (Z6c, and Z6j's same-evening 296) and 248-251 mA at 1014 MHz (Z6j, Z6k); at 1014 MHz with Wi-Fi unloaded, 184-200 mA (Z6k). I kept Z6j's same-evening pair as the bus's measure. The description says the same now.
  4. Z2k's boot: only stock recovery came up. adb devices listed the phone as recovery, and Android itself didn't start; nothing was accepted there, and adb reboot bootloader took it to fastboot. One sentence says so now.
  5. DISPLAY-PLAN D5 and the two NOTES pointers say "to come with the 90 Hz pull request".

tools/pr-check passes.

I'm glad Z1 helps on the Pixel 7 Pro too.

🤖 Generated with Claude Code

@derekross

Copy link
Copy Markdown
Owner

Thank you, Alejandro. All five are in, and I checked the new standby numbers against Z6c, Z6j and Z6k: they agree, and the status line now says what each one was measured with. Z2k's sentence settles the backup question, since Android itself never started. Merging.

@derekross
derekross merged commit e0936eb into derekross:main Nov 18, 2026
2 checks passed
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