Skip to content

arm64: dts: exynos: google: zumapro-tegu: Add the eUSB2 repeater - #118

Draft
W4HO-ham wants to merge 3 commits into
derekross:omamobilefrom
W4HO-ham:tegu-dt-repeater
Draft

W4HO-ham wants to merge 3 commits into
derekross:omamobilefrom
W4HO-ham:tegu-dt-repeater

Conversation

@W4HO-ham

Copy link
Copy Markdown
Contributor

Draft, stacked on #106 (the DT stack) and #107 (the TUSB2E11 binding and driver); only the last commit is new. Like #107, it waits for the USB host test on the palmtop.

The 9a's USB 2.0 path runs from the eUSB2 PHY through a TI TUSB2E11 repeater at 0x3e on hsi2c9, powered from the S2MPG14's LDO19M (stock: L19M_EUSB2). This adds LDO19M, the repeater node with the board's tuning from Google's GPL zuma-tegu-usb.dtsi, and phys = <&usb_repeater> on usb_hsphy. The Exynos eUSB2 PHY binding already allows a repeater there.

Checks: checkpatch --strict clean; on today's omamobile with the stack, zumapro-tegu.dtb builds and CHECK_DTBS reports nothing.

Why it isn't on the palmtop yet: once usb_hsphy has phys, the PHY waits for the repeater's driver. A kernel without it would leave USB gadget mode (our only link) waiting at probe, so this goes in together with #107 after the host test.

@derekross

Copy link
Copy Markdown
Owner

Thank you for holding this as a draft until the host test, and for pointing out that the PHY defers without the repeater driver. A few things to put into the runbook before it boots, since USB is the 9a's only link. They cover #107 too:

  • tusb2e11_exit() disables the port and the supply, and dwc3 calls phy_exit() when the gadget suspends (cable out, for example). Please include an unplug and replug in peripheral mode, with a dead-man reboot armed first and the check done over a fresh SSH login.
  • Describing ldo19m puts LDO19M under the regulator core's unused-rail cleanup. If the repeater driver isn't bound about 30 s into boot, the repeater loses power. With phys set, the eUSB2 PHY also waits for that driver, so CONFIG_PHY_TI_TUSB2E11 needs to be in our config fragment (kernel/configs/omamobile.config in omarchy-mobile) in the same change, built in or in the initramfs.
  • For upstream, one DT property per raw register value will probably get pushback. The NXP PTN3222 driver and the Qualcomm eUSB2 repeater binding are worth a look for how they describe tuning. The new Kconfig and Makefile entries also sit out of alphabetical order (TI between the two NXP entries).

No review from me until it's out of draft.

@derekross

Copy link
Copy Markdown
Owner

A short note while this is a draft, so it doesn't get lost. The last two pushes rebased it under the board commit's new message and #105's cleanup. The repeater commit and #107's two commits are otherwise unchanged, so the points from my earlier comment still apply when it's ready: the unplug/replug test behind a dead-man switch, CONFIG_PHY_TI_TUSB2E11 in the config fragment in the same change, and the tuning properties and Kconfig order for upstream.

Two things for the order:

No full review until it's out of draft.

@W4HO-ham

Copy link
Copy Markdown
Contributor Author

Rebased onto #106's new head; the three commits are unchanged (the TUSB2E11 binding and driver are #107 too). Compiles; checkpatch's one warning is the MAINTAINERS note it had before.

@derekross

Copy link
Copy Markdown
Owner

Rebased with no change, noted. My 10-28 points still apply when it comes out of draft, after #107.

The TUSB2E11 converts between an SoC's eUSB2 signalling and USB 2.0 on
one port, as on the Pixel 9a between the zumapro eUSB2 PHY and the
Type-C connector. It takes its role from the eUSB2 link, so the binding
only describes its supply and the board's signal tuning, one property
per tuning register; the register names follow Google's GPL
eusb_repeater driver.

Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
Power the repeater, wait the 3 ms it needs before it answers I2C, enable
its port and write the board's tuning from the device tree; at exit,
disable the port and the supply. The role needs no driver: the repeater
follows the eUSB2 link. Registers and timing follow Google's GPL
eusb_repeater driver for the Pixel 9a, which writes the same tuning.

Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
The Pixel 9a's USB 2.0 path runs from the eUSB2 PHY through a TI
TUSB2E11 repeater at 0x3e on hsi2c9, powered from the S2MPG14's LDO19M
(stock: L19M_EUSB2). Describe it with the board's signal tuning from
Google's GPL zuma-tegu-usb.dtsi, and give it to the eUSB2 PHY as its
repeater.

Signed-off-by: W4HO <87742609+W4HO-ham@users.noreply.github.com>
@W4HO-ham

W4HO-ham commented Nov 9, 2026

Copy link
Copy Markdown
Contributor Author

Rebased onto today's omamobile tip (1a9d1d4), now 37d69d1, 3 commits, no conflicts. W=1 on phy-ti-tusb2e11.o and zumapro-tegu.dtb: 0 warnings; checkpatch --strict: only the MAINTAINERS reminder for the new bindings file.

@derekross

Copy link
Copy Markdown
Owner

Thank you for the rebase, @W4HO-ham: all three commits are unchanged, the bottom two are the same patches as #107, and the gate and CI (with dtbs_check and the binding check) pass on today's omamobile. Nothing new from me. My 10-28 points still apply when it comes out of draft after #107 and the host test: the unplug and replug behind a dead-man switch in the runbook, CONFIG_PHY_TI_TUSB2E11 in omarchy-mobile's shared config fragment in the same change, the Kconfig and Makefile entries in alphabetical order (the TI entry still sits before the NXP ones), and the tuning properties for upstream.

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