Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
72 changes: 72 additions & 0 deletions device/raven/AUDIO-PLAN.md
Original file line number Diff line number Diff line change
Expand Up @@ -404,6 +404,61 @@ speaker calibration is never committed.
speaker protection.
6. **R5:** sensors (auto-rotate) through the same AoC.

## R6: raven's own speaker limits (2026-10-27, `gb3d6253b0d61`, omamobile-audio 0.4.4)

**Stock's values.** Raven's `mixer_paths.xml` (its factory vendor image)
sets `AMP PCM Gain` 17 and `Digital PCM Volume` 817 on both amplifiers, as
cheetah's, so stage media (737/10) sits as far below stock on raven as on
cheetah. The one raven difference is the boost peak current: 3.50 A by
default, lowered in the `speaker` path to 1.70 A top and 1.90 A bottom
(3.50 A only in voice-speaker and speaker-safe). Mainline has no control
for it and takes it from the DT at probe, so raven played at 3.50 A.
`omamobile-gs101-audio` 9551be2e5 sets 1700/1900 mA (needs the human's
sign-off); checked as an overlay: `BSTCVRT_PEAK_CUR` (0x3808) read 0x12 and
0x16 instead of 0x36.

**C9a on raven, stage media (737/10):** passed. The sense paths during
playback (bounded read, never past 0x10100): `0x2018` 0x3721 (VMON, IMON
on), DSP inputs `0x18 0x19 0x28 0x29 0x3a`, cheetah's values. The firmware
ran throughout (halo 2, state 0, errorno 0, calibration set 2, the
heartbeat about 1520 a second, no freeze), both coils at 23.00-23.04 C at
-20 dBFS; at -10 dBFS both tracked the signal (+0.15 C top, +0.11 C
bottom per 10 s burst, back down within about 4 s). `verify` passed and
wrote the stamp. The human heard both speakers, clean. `vmon low` reads 1
at every level tried.

**Stepping up** (stage c9, `speaker_protection=1` from
`/etc/modprobe.d/omamobile-audio-c9-bringup.conf`; one -20 dBFS multitone
burst per step unless noted, stop on a jump over 2 C in a second or a rise
over 1 C):

| Digital / analog | Gain over 737/10 | Coils |
| --- | --- | --- |
| 737/10 | 0 dB | +0.04 C, sane |
| 737/13 | +3 dB | +0.09 / +0.06 C, sane |
| 737/17 | +7 dB | +0.22 / +0.16 C, sane; the human: louder, clean |
| 757/17 | +9.5 dB | +0.39 / +0.28 C, sane |
| 797/10 | +7.5 dB | music for a minute, 23.4-23.9 C, sane |
| 817/10 | +10 dB | **bottom misread**: 23 to 36 C in the first second, flat, 33 C with nothing playing |
| 777/17 | +12 dB | **both misread**: 23 to 37 / 36 C in the first second; stopped, the watchdog muted too |

So the protection firmware's coil estimate holds up to about +9.5 dB over
stage media, whether the gain is digital or analog, and breaks at about
+10 dB on raven (cheetah broke at 817/10 as well). The rises below that
scale with the power (about nine times stage media's at +9.5 dB). The last
~10 dB to stock is the gain-17 problem (cheetah C9d-7/8).

**737/17 with music:** 2 min 40 s at full volume, 121 samples, top
23.62 to 23.88 C at most, bottom 23.44 to 23.61 C, no jump. `verify`
refuses analog 17 by design (its stamp gates c9 at stage-media levels);
its "heartbeat did not move" there needs a look (a track ending?). The
human's phone now runs stage c9 at 737/17 daily, locally (its
`speaker.conf` says how to go back); not in any package.

Open: with music the first burst (-20 dBFS) at 797/10 was "still a bit
quiet" and the -10 dBFS one "quite loud"; 737/17 is about as loud as
797/10. Stock's level needs the gain-17 work.

## Without the amplifiers (2026-10-27, `gb3d6253b0d61`, device-raven 0.1.23)

Derek asked on #41 whether the sound card, and so the microphones, still
Expand All @@ -423,3 +478,20 @@ clash Derek pointed out. Ways forward, not decided: give the card a
speaker-less link set when the amplifiers are absent, or load the
amplifier driver muted with no protection firmware, so the card and the
microphones come up while the speakers stay off.

**Microphones without protection (2026-10-27, `gb3d6253b0d61`,
omamobile-audio 0.5.3, device-raven 0.1.25; branch
`omamobile-raven-mics`):** the second way. `omamobile-speaker-protection
mics` (run by `omamobile-audio-mics.service`, which device-raven enables)
loads `snd_soc_cs35l41_spi` and mutes both amplifiers, as `start` does
before anything else, and stops there: no DSP firmware, no unmute. The
unit doesn't run once speaker protection is enabled, and uses the same
hang marker. On raven, after a reboot: the unit finished at 12:01:45, the
card came up, both amplifiers read `Digital PCM Volume` 0, speaker
protection inactive, and PipeWire's default source was "Built-in Audio
Microphone". A 4 s recording in a quiet room read -64.5 dBFS RMS (not
digital silence); with the human counting out loud, -48 to -56 dBFS per
half second, falling to about -71 dBFS between words. So the microphones
work with the speakers muted. Not tested: a call (the modem doesn't start
at boot with this package set), and the speaker sink, which PipeWire still
lists but which plays into muted amplifiers.
10 changes: 6 additions & 4 deletions device/raven/device.conf
Original file line number Diff line number Diff line change
Expand Up @@ -15,10 +15,12 @@ SOC=gs101
COMPATIBLE="google,gs101-raven"
PACKAGE=omamobile-device-raven
# This phone's other packages in the root image, on top of the shared
# rootfs/packages.omamobile: the sensors through the AoC, and GNSS (off
# until `omamobile-gnss on`). Audio waits on the amplifiers' limits, the
# modem on its kernel.
ROOTFS_PACKAGES="omamobile-sensors omamobile-gnss"
# rootfs/packages.omamobile: the sensors through the AoC, GNSS (off until
# `omamobile-gnss on`), and the audio package for the microphones (its
# blacklist keeps the amplifiers' driver from loading on its own; device-
# raven's omamobile-audio-mics loads it muted). Speaker protection stays
# off until raven's limits are in a package; the modem waits on its kernel.
ROOTFS_PACKAGES="omamobile-sensors omamobile-gnss omamobile-audio"

# --- Kernel and device tree ---
DTB=exynos/google/gs101-raven.dtb
Expand Down
2 changes: 1 addition & 1 deletion docs/DEVICES.md
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,7 @@ The status of each device, as verified on real hardware. Rows in the matrix have
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Pixel 7 Pro | cheetah | gs201 (Tensor G2) | works | works | works (120 Hz) | works (DVFS) | works | works | works | works | works (audio) | works | works | works | works | works | works | works | works | works (calls, texts and mobile data on a Verizon-network SIM, 2026-11-03; see below) | works | works (telephoto fixed-focus) | WIP | works |
| Pixel Fold | felix | gs201 (Tensor G2) | WIP ([#3](https://github.com/derekross/omarchy-mobile/issues/3)) | WIP (UFS behind the read-only guard, checked on the phone) | WIP (cover: simplefb console; inner: simpledrm output listed by Hyprland, its splash reserved by [kernel #72](https://github.com/derekross/linux-omarchy-mobile/pull/72); both panels' display triggers set by the descriptor, [#82](https://github.com/derekross/omarchy-mobile/pull/82)) | testing (Panthor; glmark2 on the cover, not up on every image yet; [notes](../device/felix/NOTES.md#2026-10-26-the-power-key-and-the-gpu-in-the-published-kernel)) | WIP (cover: chip id read, no touches yet; the driver's touch reports are merged, [kernel #48](https://github.com/derekross/linux-omarchy-mobile/pull/48), untested on the Fold; [notes](../device/felix/NOTES.md#2026-10-23-cover-touch-id)) | works (volume and power; [notes](../device/felix/NOTES.md#2026-10-26-the-power-key-and-the-gpu-in-the-published-kernel)) | works | no | no | no | no | no | no | no | no | no | no | no | no | no | no | no |
| Pixel 6 Pro ([notes](../device/raven/NOTES.md)) | raven | gs101 (Tensor) | works ([#2](https://github.com/derekross/omarchy-mobile/issues/2)) | works | works (120 Hz through gs-decon and the S6E3HC3 driver; unpublished kernel, `omamobile-gs101-display`, in review as [kernel #116](https://github.com/derekross/linux-omarchy-mobile/pull/116), [plan](../device/raven/DISPLAY-PLAN.md)); the published kernel has the boot console only | partial (Panfrost; DVFS on an unpublished kernel, [plan](../device/raven/GPU-PLAN.md)) | works ([kernel #8](https://github.com/derekross/linux-omarchy-mobile/pull/8)) | works | works (a NOT FOR MERGE DT change; [plan](../device/raven/USB-PLAN.md)) | works (unpublished kernel, `omamobile-gs101-wifi`; [plan](../device/raven/WIFI-PLAN.md)) | partial ([kernel #6](https://github.com/derekross/linux-omarchy-mobile/pull/6); own address; pairing from the menu and A2DP audio work, crackling while Wi-Fi is busy; the chip came up silent after one of seven clean reboots on 2026-11-03, [#77](https://github.com/derekross/omarchy-mobile/issues/77); [plan](../device/raven/BLUETOOTH-PLAN.md)) | works (the gauge; unpublished kernel) | works (the MAX77759 charger at stock's limits with its temperature guard, loaded at boot by device-raven 0.1.26; a full charge to 100 %; the battery node is in the published kernel, [kernel #112](https://github.com/derekross/linux-omarchy-mobile/pull/112), tested on an unpublished kernel; [plan](../device/raven/USB-PLAN.md#4-charging-the-max77759-charger-ravens-b3-after-cheetahs)) | works (CS40L25A through feedbackd, device-raven 0.1.27, with the phone's own firmware from `tools/extract-vendor-firmware -d raven`; the driver and node are merged, [kernel #114](https://github.com/derekross/linux-omarchy-mobile/pull/114); [plan](../device/raven/HAPTICS-TORCH-PLAN.md)) | works (tested on `omamobile-gs101-torch`; the LED's node is now in the published kernel, [kernel #113](https://github.com/derekross/linux-omarchy-mobile/pull/113), not yet run from it; [plan](../device/raven/HAPTICS-TORCH-PLAN.md)) | works (unpublished kernel, `omamobile-gs101-gnss`; [plan](../device/raven/GNSS-PLAN.md)) | works (unpublished kernel, `omamobile-gs101-aoc`; [plan](../device/raven/AUDIO-PLAN.md)) | no (the amplifiers stay off until raven's own limits are validated, and without them the card doesn't probe, so no mics either; [plan](../device/raven/AUDIO-PLAN.md#without-the-amplifiers-2026-10-27-gb3d6253b0d61-device-raven-0123)) | works (accelerometer, light, proximity, compass through the AoC; unpublished kernel) | WIP (calls, texts and data on an unpublished kernel, `omamobile-gs101-cpif`; the package doesn't enable it; [plan](../device/raven/MODEM-PLAN.md)) | no | no | testing (s2idle when asked for, or through the shell's auto-sleep with the `omamobile-sleep` toggle on, which ships off; Power wakes it, 3 of 3 on 2026-11-03 on an unpublished kernel (the published one plus [kernel #116](https://github.com/derekross/linux-omarchy-mobile/pull/116) and raven's USB override); UFS resumes clean with [kernel #188](https://github.com/derekross/linux-omarchy-mobile/pull/188), in review; with the modem up its sleep block holds; [plan](../device/raven/SUSPEND-PLAN.md)) | no |
| Pixel 6 Pro ([notes](../device/raven/NOTES.md)) | raven | gs101 (Tensor) | works ([#2](https://github.com/derekross/omarchy-mobile/issues/2)) | works | works (120 Hz through gs-decon and the S6E3HC3 driver; unpublished kernel, `omamobile-gs101-display`, in review as [kernel #116](https://github.com/derekross/linux-omarchy-mobile/pull/116), [plan](../device/raven/DISPLAY-PLAN.md)); the published kernel has the boot console only | partial (Panfrost; DVFS on an unpublished kernel, [plan](../device/raven/GPU-PLAN.md)) | works ([kernel #8](https://github.com/derekross/linux-omarchy-mobile/pull/8)) | works | works (a NOT FOR MERGE DT change; [plan](../device/raven/USB-PLAN.md)) | works (unpublished kernel, `omamobile-gs101-wifi`; [plan](../device/raven/WIFI-PLAN.md)) | partial ([kernel #6](https://github.com/derekross/linux-omarchy-mobile/pull/6); own address; pairing from the menu and A2DP audio work, crackling while Wi-Fi is busy; the chip came up silent after one of seven clean reboots on 2026-11-03, [#77](https://github.com/derekross/omarchy-mobile/issues/77); [plan](../device/raven/BLUETOOTH-PLAN.md)) | works (the gauge; unpublished kernel) | works (the MAX77759 charger at stock's limits with its temperature guard, loaded at boot by device-raven 0.1.26; a full charge to 100 %; the battery node is in the published kernel, [kernel #112](https://github.com/derekross/linux-omarchy-mobile/pull/112), tested on an unpublished kernel; [plan](../device/raven/USB-PLAN.md#4-charging-the-max77759-charger-ravens-b3-after-cheetahs)) | works (CS40L25A through feedbackd, device-raven 0.1.27, with the phone's own firmware from `tools/extract-vendor-firmware -d raven`; the driver and node are merged, [kernel #114](https://github.com/derekross/linux-omarchy-mobile/pull/114); [plan](../device/raven/HAPTICS-TORCH-PLAN.md)) | works (tested on `omamobile-gs101-torch`; the LED's node is now in the published kernel, [kernel #113](https://github.com/derekross/linux-omarchy-mobile/pull/113), not yet run from it; [plan](../device/raven/HAPTICS-TORCH-PLAN.md)) | works (unpublished kernel, `omamobile-gs101-gnss`; [plan](../device/raven/GNSS-PLAN.md)) | works (unpublished kernel, `omamobile-gs101-aoc`; [plan](../device/raven/AUDIO-PLAN.md)) | works (both speakers through raven's protection firmware with the phone's factory calibration, at cheetah's stage media (737/10), which R6 verified on raven; the microphones with them; unpublished kernel, `omamobile-gs101-aoc`; [plan](../device/raven/AUDIO-PLAN.md#r6-ravens-own-speaker-limits-2026-10-27-gb3d6253b0d61-omamobile-audio-044)) | works (accelerometer, light, proximity, compass through the AoC; unpublished kernel) | WIP (calls, texts and data on an unpublished kernel, `omamobile-gs101-cpif`; the package doesn't enable it; [plan](../device/raven/MODEM-PLAN.md)) | no | no | testing (s2idle when asked for, or through the shell's auto-sleep with the `omamobile-sleep` toggle on, which ships off; Power wakes it, 3 of 3 on 2026-11-03 on an unpublished kernel (the published one plus [kernel #116](https://github.com/derekross/linux-omarchy-mobile/pull/116) and raven's USB override); UFS resumes clean with [kernel #188](https://github.com/derekross/linux-omarchy-mobile/pull/188), in review; with the modem up its sleep block holds; [plan](../device/raven/SUSPEND-PLAN.md)) | no |
| Pixel 7a ([notes](../device/lynx/NOTES.md)) | lynx | gs201 (Tensor G2) | WIP ([#36](https://github.com/derekross/omarchy-mobile/issues/36): Arch root on `userdata` from slot B, SSH; the device package marks the slot successful) | WIP (UFS behind the read-only guard, in the published kernel, [kernel #91](https://github.com/derekross/linux-omarchy-mobile/pull/91); [notes](../device/lynx/NOTES.md#2026-10-24-on-screen-console-and-ufs)) | WIP (simpledrm console only, [notes](../device/lynx/NOTES.md)) | no | WIP (GT9895 touches in the session, in the published kernel, [kernel #93](https://github.com/derekross/linux-omarchy-mobile/pull/93); [notes](../device/lynx/NOTES.md#2026-10-24-touchscreen-rtc-and-slot-retries)) | no | works | no (parked at the chip's boot ROM; [plan](../device/lynx/WIFI-PLAN.md)) | WIP (same chip as Wi-Fi; scans through its own UART; the device tree is in review, [kernel #98](https://github.com/derekross/linux-omarchy-mobile/pull/98), and the package side, [#98](https://github.com/derekross/omarchy-mobile/pull/98); [plan](../device/lynx/WIFI-PLAN.md)) | no | no | no | no | no | no | no | no | no | no | no | no | no |
| Pixel 6 | oriole | gs101 (Tensor G1) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); native test image on one phone; slot B marked successful and `vbmeta_b` verification turned off, both without project approval; see notes) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); guarded `userdata` root on test image) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); connected output on test image) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); Panfrost render node present) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); touchscreen detected; input not checked by smoke) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); gpio-keys detected; input not checked by smoke) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); USB network and SSH on test image) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); link up; traffic not checked by smoke) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); controller powered; pairing not checked) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); charge and temperature reported) | WIP ([unpublished kernel/packages](../device/oriole/NOTES.md#2026-10-31-build-provenance-and-recovery-limits); charging reported; control not checked) | no | no | no | no | no | no | no | no | no | no | no |
| Pixel 4 | flame | SM8150 (Snapdragon 855) | testing (temporary RAM boot; [notes](../device/flame/BRINGUP-M1.md)) | testing (RAM probe; all UFS read-only) | testing (1080x2280 console; software desktop in separate Arch payload) | testing (separate Arch payload; FD640 contexts, llvmpipe desktop) | testing (separate Arch payload; events and PIN unlock verified) | no | testing (NCM + ACM + SSH; sustained transfers can disconnect) | testing (guarded RAM scan; association untested) | no | no | no | no | no | no | - | no | no | no | no | no | no | - |
Expand Down
Loading