Re: [PATCH 2/2] arm64: dts: qcom: x1p42100-lenovo-thinkbook-16: add hdmi bridge with enable pin
From: Jens Glathe
Date: Sat Dec 20 2025 - 15:23:27 EST
On 20.12.25 19:52, Dmitry Baryshkov wrote:
On Sat, Dec 20, 2025 at 06:47:10PM +0100, Jens Glathe via B4 Relay wrote:
From: Jens Glathe <jens.glathe@xxxxxxxxxxxxxxxxxxxxxx>Whaat is AeoB? I don't see the GPIO being used in other place.
Add TLMM 119 for hdmi bridge enable. AeoB has this gpio.
AeoB is a resource format used by the QCom Windows drivers to add / overwrite definitions coming from the ACPI tables. [1] It had the info that was not to be found in the DSDT. Actually, x1e80100.dtsi has the edp0_hpd_default pinctrl which is required for the display to work with the function edp_hot. I verified this, you need both. I can use that definition in this patchset, too, but its not for HDMI.
[1]: https://github.com/alexVinarskis/qcom-aeob-dumps/tree/master/lenovo-thinkbook-16-g7-qoy
I don't know. Haven't looked inside.In general this is the same setup as on the T14s. Since its using simple-bridgeIs the chip marked anyhow?
and also is Lenovo, we also use the same realtek,rtd2171 compatible.
The real chip identity is not known yet. The bridge is wired with un-swapped
lanes, though.
As for gpio119 being used twice: It doesn't work if you don't, so it may be a caseWhat is the relationship between enable and HPD?
of TLMM multiplexing hpd for both ports.
Yeah, good question. Actually HDMI hpd is wired to gpio126, and works without the enable pin. But no link training succeeds, no display coming up. With enable set on probe() this works. If you have the lane order right. The HDMI socket is "above" the type-c ports on the TB16, compared to the T14s where it is "below" the type-c ports, so swapped lanes sounds reasonable.
[...]