The SoQuartz CM4-IO should support otg mode, however it currently
defaults to device mode when otg is enabled. Force it to host mode to
retain the previous behavior.
Fixes: bc405bb3eeee ("arm64: dts: rockchip: enable otg/drd operation of usb_host0_xhci in rk356x")
Signed-off-by: Peter Geis <[email protected]>
---
arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts | 1 +
1 file changed, 1 insertion(+)
diff --git a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
index e00568a6be5c..56764fae9f9d 100644
--- a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
+++ b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
@@ -169,6 +169,7 @@ &usb2phy0_otg {
};
&usb_host0_xhci {
+ dr_mode = "host";
status = "okay";
};
--
2.25.1
On Fri, Jun 10, 2022 at 9:26 AM Peter Geis <[email protected]> wrote:
>
> The SoQuartz CM4-IO should support otg mode, however it currently
> defaults to device mode when otg is enabled. Force it to host mode to
> retain the previous behavior.
It would seem the role-switch issue is more widespread than originally
thought, affecting other boards where automatic control should work.
Please hold this patch while I investigate further.
Thanks!
>
> Fixes: bc405bb3eeee ("arm64: dts: rockchip: enable otg/drd operation of usb_host0_xhci in rk356x")
>
> Signed-off-by: Peter Geis <[email protected]>
> ---
> arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> index e00568a6be5c..56764fae9f9d 100644
> --- a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> +++ b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> @@ -169,6 +169,7 @@ &usb2phy0_otg {
> };
>
> &usb_host0_xhci {
> + dr_mode = "host";
> status = "okay";
> };
>
> --
> 2.25.1
>
Am Freitag, 10. Juni 2022, 16:44:23 CEST schrieb Peter Geis:
> On Fri, Jun 10, 2022 at 9:26 AM Peter Geis <[email protected]> wrote:
> >
> > The SoQuartz CM4-IO should support otg mode, however it currently
> > defaults to device mode when otg is enabled. Force it to host mode to
> > retain the previous behavior.
>
> It would seem the role-switch issue is more widespread than originally
> thought, affecting other boards where automatic control should work.
> Please hold this patch while I investigate further.
ok!
I guess this also applies to
"arm64: dts: rockchip: Fix Quartz64-A dwc3 otg port behavior"?
Heiko
> >
> > Fixes: bc405bb3eeee ("arm64: dts: rockchip: enable otg/drd operation of usb_host0_xhci in rk356x")
> >
> > Signed-off-by: Peter Geis <[email protected]>
> > ---
> > arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts | 1 +
> > 1 file changed, 1 insertion(+)
> >
> > diff --git a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > index e00568a6be5c..56764fae9f9d 100644
> > --- a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > +++ b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > @@ -169,6 +169,7 @@ &usb2phy0_otg {
> > };
> >
> > &usb_host0_xhci {
> > + dr_mode = "host";
> > status = "okay";
> > };
> >
> > --
> > 2.25.1
> >
>
On Sat, Jun 11, 2022 at 9:50 AM Heiko Stuebner <[email protected]> wrote:
>
> Am Freitag, 10. Juni 2022, 16:44:23 CEST schrieb Peter Geis:
> > On Fri, Jun 10, 2022 at 9:26 AM Peter Geis <[email protected]> wrote:
> > >
> > > The SoQuartz CM4-IO should support otg mode, however it currently
> > > defaults to device mode when otg is enabled. Force it to host mode to
> > > retain the previous behavior.
> >
> > It would seem the role-switch issue is more widespread than originally
> > thought, affecting other boards where automatic control should work.
> > Please hold this patch while I investigate further.
>
> ok!
>
> I guess this also applies to
> "arm64: dts: rockchip: Fix Quartz64-A dwc3 otg port behavior"?
No, on Quartz64 Model A it can't work, the requisite pin isn't connected.
I think I've found the problem, I've requested a few people to test a fix.
Do you have a Model B or SoQuartz to test against (or any other rk356x
board that has the id pin hooked up)?
Thanks,
Peter
>
>
> Heiko
>
>
> > >
> > > Fixes: bc405bb3eeee ("arm64: dts: rockchip: enable otg/drd operation of usb_host0_xhci in rk356x")
> > >
> > > Signed-off-by: Peter Geis <[email protected]>
> > > ---
> > > arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts | 1 +
> > > 1 file changed, 1 insertion(+)
> > >
> > > diff --git a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > > index e00568a6be5c..56764fae9f9d 100644
> > > --- a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > > +++ b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > > @@ -169,6 +169,7 @@ &usb2phy0_otg {
> > > };
> > >
> > > &usb_host0_xhci {
> > > + dr_mode = "host";
> > > status = "okay";
> > > };
> > >
> > > --
> > > 2.25.1
> > >
> >
>
>
>
>
Am Samstag, 11. Juni 2022, 15:53:41 CEST schrieb Peter Geis:
> On Sat, Jun 11, 2022 at 9:50 AM Heiko Stuebner <[email protected]> wrote:
> >
> > Am Freitag, 10. Juni 2022, 16:44:23 CEST schrieb Peter Geis:
> > > On Fri, Jun 10, 2022 at 9:26 AM Peter Geis <[email protected]> wrote:
> > > >
> > > > The SoQuartz CM4-IO should support otg mode, however it currently
> > > > defaults to device mode when otg is enabled. Force it to host mode to
> > > > retain the previous behavior.
> > >
> > > It would seem the role-switch issue is more widespread than originally
> > > thought, affecting other boards where automatic control should work.
> > > Please hold this patch while I investigate further.
> >
> > ok!
> >
> > I guess this also applies to
> > "arm64: dts: rockchip: Fix Quartz64-A dwc3 otg port behavior"?
>
> No, on Quartz64 Model A it can't work, the requisite pin isn't connected.
ok, so I can pick that other patch up :-) .
> I think I've found the problem, I've requested a few people to test a fix.
> Do you have a Model B or SoQuartz to test against (or any other rk356x
> board that has the id pin hooked up)?
The rk356x in my boardfarm is a Quartz-B ... though an very early model,
so hopefully nothing changed on the board since then.
Although there is nothing connected to usb right now and I only have
remote access this weekend, so any usb tests would need to wait till next
week on my end.
Heiko
>
> Thanks,
> Peter
>
> >
> >
> > Heiko
> >
> >
> > > >
> > > > Fixes: bc405bb3eeee ("arm64: dts: rockchip: enable otg/drd operation of usb_host0_xhci in rk356x")
> > > >
> > > > Signed-off-by: Peter Geis <[email protected]>
> > > > ---
> > > > arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts | 1 +
> > > > 1 file changed, 1 insertion(+)
> > > >
> > > > diff --git a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > > > index e00568a6be5c..56764fae9f9d 100644
> > > > --- a/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > > > +++ b/arch/arm64/boot/dts/rockchip/rk3566-soquartz-cm4.dts
> > > > @@ -169,6 +169,7 @@ &usb2phy0_otg {
> > > > };
> > > >
> > > > &usb_host0_xhci {
> > > > + dr_mode = "host";
> > > > status = "okay";
> > > > };
> > > >
> > > > --
> > > > 2.25.1
> > > >
> > >
> >
> >
> >
> >
>