Skip to content

arm64: dts: exynos: google: gs201-panther: enable the MAX77759's I2C bus - #167

Merged
derekross merged 1 commit into
derekross:omamobilefrom
nesmeck:omamobile-gs201-panther-usb
Nov 1, 2026
Merged

derekross merged 1 commit into
derekross:omamobilefrom
nesmeck:omamobile-gs201-panther-usb

Conversation

@nesmeck

@nesmeck nesmeck commented Oct 31, 2026

Copy link
Copy Markdown
Contributor

Enable USI13 as I2C at 400 kHz, as the Pixel 7 Pro does. The MAX77759 sits on that bus, on gpp25-0/1: the Type-C port controller at 0x25, the PMIC at 0x66, the fuel gauge at 0x36 and the charger at 0x69. The Pixel 7's live device tree has the same bus.

Only the bus, no children. A Type-C port controller node would take over VBUS and the data role from the forced-VBUS setup the USB PHY uses, and USB is the only way into the phone. Describing the chips is a later step.

Why now. The MAX77759 opens its USB2 data switch when the cable is unplugged, and nothing closes it again, so after a replug the phone stays off the bus until it reboots. With the bus up, omamobile-usb-rescue closes it from user space through TCPC vendor register 0x93. That is the script cheetah already uses, and it comes to omamobile-device-panther in a later repo PR.

Tested on a Pixel 7 (slot B, 7.3.0-rc6), with the touchscreen, the GPU and the haptics:

  • The bus probes as i2c-1.
  • The TCPC's USBSW_CTRL reads 0x09 and the charger's CHG_DETAILS_00 reads 0x60 (VBUS valid).
  • After an unplug the switch read 0x00. omamobile-usb-rescue set it back to 0x09, USB networking came back by itself in about 9 s, and there were no I2C errors.
  • Rebased onto today's omamobile (b0d62d9), it booted twice more (…-g8c9496448397, with the old Input: ft3658 - review fixes: DMA-safe buffers, fsleep(), the Google source #133 head, and …-ge57e346e9b71). Both times i2c-1 was up, the switch read 0x09 and the charger 0x60. The phone-smoke report from the second boot is in the touch PR.

Checks: tools/kernel-pr-check --build OUT --merge origin/omamobile passes everything:

  • a clean merge onto b0d62d9;
  • the sign-off;
  • checkpatch --strict with no warnings;
  • the 8 device trees with W=1 and dtbs_check, with no new warnings.

🤖 Generated with Claude Code

Enable USI13 as I2C at 400 kHz, as the Pixel 7 Pro does: the MAX77759
(Type-C port controller at 0x25, PMIC at 0x66, fuel gauge at 0x36,
charger at 0x69) sits there, on gpp25-0/1, in the Pixel 7's live device
tree too. Its devices stay undescribed for now: the Type-C port
controller would take over VBUS and the data role from the forced-VBUS
setup the USB PHY uses, and the USB link is the only way into the
phone.

With the bus up, user space can close the MAX77759's USB2 data switch
after a replug (TCPC vendor register 0x93), which the chip opens on
unplug; until a Type-C driver owns it, nothing else does, and the phone
stays off the bus.

Tested on a Pixel 7 from slot B (7.3.0-rc6), with the touchscreen, GPU
and haptics: the bus probes as i2c-1, the TCPC's USBSW_CTRL reads 0x09
and the charger's CHG_DETAILS_00 0x60 (VBUS valid). After an unplug the
switch read 0x00; omamobile-usb-rescue set it back to 0x09 and USB
networking came back by itself, with no I2C errors.

Assisted-by: Claude Opus 5.5
Signed-off-by: Alejandro Garcia <skipper.pelygro@gmail.com>
@derekross
derekross merged commit caf5f8a into derekross:omamobile Nov 1, 2026
1 check passed
@derekross

Copy link
Copy Markdown
Owner

Thank you, @nesmeck. Enabling only the bus is the careful way to do this, and the message gives exactly the right reason for leaving the port controller out while the forced-VBUS setup owns USB.

The node is the same as the Pixel 7 Pro's (I2C mode, 400 kHz), and Google's gs201-pantah-charging.dtsi runs hsi2c_13 at 400 kHz on the Pixel 7 too. It merges cleanly with #166, the checks pass, dt-validate reports nothing, the only DTB change is USI13 and its I2C controller turning on, and every other phone's DTB is byte-identical. With no child nodes nothing binds, and omamobile-usb-rescue finds the bus by its address. Merged.

Optional, for a later push: stock also has the PCA9468 direct charger at 0x57 on this bus (gs201-pantah-charging.dtsi), worth adding to the comment's address list. When the device-panther PR brings the rescue unit, please add the replug result to panther's notes so the matrix's USB cell can change.

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