Re: [PATCH v3 0/4] arm64: dts: qcom: nord: Add USB support

From: Bartosz Golaszewski

Date: Thu Nov 12 2026 - 04:26:13 EST


On Sun, Nov 08, 2026 at 01:23:18PM -0600, Bjorn Andersson wrote:
> On Tue, Oct 13, 2026 at 01:34:45PM +0200, Bartosz Golaszewski wrote:
> > Note: This is v7.4 material and depends on several patches that are
> > still not in linux-next. Sending it out to solicit reviews.
> >
> > This series enables USB on the Qualcomm Nord platform which has three
> > USB controllers:
> >
> > usb_0 (0xa600000) eUSB2 + USB3/DP combo PHY
> > usb_1 (0xa400000) eUSB2 + USB3/DP combo PHY
> > usb_2 (0xa200000) eUSB2 only, high-speed
> >
> > The first patch is a fix for the i2c9/i2c10 pin functions, which is a
> > prerequisite because the eUSB2 repeaters and the Type-C controller all
> > sit on i2c9.
> >
> > Patch 2 adds the three DWC3 controller nodes to nord.dtsi. They are
> > kept as far as possible free of anything variant specific, so the
> > clocks, resets, power domains, interconnects and PHY references are
> > added separately in patch 3 for the "embedded" variant, alongside the
> > five PHY nodes (three eUSB2, two QMP combo). Adding the QMP PHYs also
> > lets the two pipe clock placeholders in negcc be filled in.
> >
> > Patches 4 and 5 wire the ports up per board.
> >
> > Nord RRD:
> >
> > usb_0 OTG, USB-C, FUSB15201 Type-C/PD controller
> > usb_1 host, Genesys GL3590 HS/SS onboard hub pair
> > usb_2 left disabled, not routed on this board
> >
> > Nord Ride Embedded:
> >
> > usb_0 OTG, USB-C port 0 of a dual-port FUSB15201
> > usb_1 OTG, USB-C port 1 of the same FUSB15201
> > usb_2 host, USB Type-A, VBUS from a PMIC GPIO controlled
> > fixed regulator
> >
> > Both boards use NXP PTN3222 eUSB2 repeaters, one per enabled port.
> >
> > Dependencies:
> > - PMIC DTS: https://lore.kernel.org/all/20260929155152.1898666-1-shengchao.guo@xxxxxxxxxxxxxxxx/
> > - FUSB15201: https://lore.kernel.org/all/20260916135918.1017527-1-shengchao.guo@xxxxxxxxxxxxxxxx/
> > - USB PHY clock fix: https://lore.kernel.org/all/20261008015017.3359804-1-shengchao.guo@xxxxxxxxxxxxxxxx/
> >
>
> Figured you resubmitted this because the external dependencies was
> settled. Apparently not, so I had to back this out again after applying
> it.
>
> Please resubmit once the dependencies are in linux-next.
>

The only thing left is the FUSB15201 series and it's only a run-time and
dtbs_check-time dependency. Can we proceed with this or do the repeater
bindings need to go first?

Thanks,
Bartosz