[PATCH 0/3] Enable EHCI port on rk3288-evb

[PATCH 0/3] Enable EHCI port on rk3288-evb

[Doug Anderson]

This series enables the EHCI port on the rk3288-evb boards.  A later
patch series (from Kever) will enable the dwc2 ports.


Doug Anderson (3):
  ARM: dts: Turn on USB host vbus on rk3288-evb
  ARM: dts: add rk3288 ehci usb devices
  ARM: dts: Enable USB host0 (EHCI) on rk3288-evb

 arch/arm/boot/dts/rk3288-evb.dtsi | 22 ++++++++++++++++++++++
 arch/arm/boot/dts/rk3288.dtsi     | 20 ++++++++++++++++++++
 2 files changed, 42 insertions(+)

-- 
2.0.0.526.g5318336

[PATCH 1/3] ARM: dts: Turn on USB host vbus on rk3288-evb

[Doug Anderson]

There is no phy driver that works on the Rockchip board for either USB
host port yet.  For now just hardcode the vbus signal to be on all the
time which makes both the dwc2 host and the EHCI port work.

Signed-off-by: Doug Anderson <dianders@chromium.org>
---
 arch/arm/boot/dts/rk3288-evb.dtsi | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/arch/arm/boot/dts/rk3288-evb.dtsi b/arch/arm/boot/dts/rk3288-evb.dtsi
index 749e20d..efd625e 100644
--- a/arch/arm/boot/dts/rk3288-evb.dtsi
+++ b/arch/arm/boot/dts/rk3288-evb.dtsi
@@ -35,6 +35,18 @@
 			debounce-interval = <100>;
 		};
 	};
+
+	/* This turns on vbus for both host0 (ehci) and host1 (dwc2) */
+	usb_host_vbus_regulator: usb-host-vbus-regulator {
+		compatible = "regulator-fixed";
+		enable-active-high;
+		gpio = <&gpio0 14 GPIO_ACTIVE_HIGH>;
+		pinctrl-names = "default";
+		pinctrl-0 = <&usb_host_vbus>;
+		regulator-name = "usb-host-vbus";
+		regulator-always-on;
+		regulator-boot-on;
+	};
 };
 
 &i2c0 {
@@ -71,4 +83,10 @@
 			rockchip,pins = <0 5 RK_FUNC_GPIO &pcfg_pull_up>;
 		};
 	};
+
+	usb-host {
+		usb_host_vbus: usb-host-vbus {
+			rockchip,pins = <0 14 RK_FUNC_GPIO &pcfg_pull_none>;
+		};
+	};
 };
-- 
2.0.0.526.g5318336

[PATCH 2/3] ARM: dts: add rk3288 ehci usb devices

[Doug Anderson]

rk3288 has two kind of usb controller; this adds the ehci variant for
host0 and hsic.

At the moment we don't add any phys for these controllers, but the
default settings seem to work OK.

There is a hardware problem in ohci controller which make it
unavailable and host0 controller can only support high-speed devices.

Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Signed-off-by: Doug Anderson <dianders@chromium.org>
---
 arch/arm/boot/dts/rk3288.dtsi | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)

diff --git a/arch/arm/boot/dts/rk3288.dtsi b/arch/arm/boot/dts/rk3288.dtsi
index 3ef8951..0bb4480 100644
--- a/arch/arm/boot/dts/rk3288.dtsi
+++ b/arch/arm/boot/dts/rk3288.dtsi
@@ -251,6 +251,26 @@
 		status = "disabled";
 	};
 
+	usb_host0_ehci: ehci at ff500000 {
+		compatible = "generic-ehci";
+		reg = <0xff500000 0x100>;
+		interrupts = <GIC_SPI 24 IRQ_TYPE_LEVEL_HIGH>;
+		clocks = <&cru HCLK_USBHOST0>;
+		clock-names = "usbhost";
+		status = "disabled";
+	};
+
+	/* NOTE: ohci at ff520000 doesn't actually work on hardware */
+
+	usb_hsic: ehci at ff5c0000 {
+		compatible = "generic-ehci";
+		reg = <0xff5c0000 0x100>;
+		interrupts = <GIC_SPI 26 IRQ_TYPE_LEVEL_HIGH>;
+		clocks = <&cru HCLK_HSIC>;
+		clock-names = "usbhost";
+		status = "disabled";
+	};
+
 	gic: interrupt-controller at ffc01000 {
 		compatible = "arm,gic-400";
 		interrupt-controller;
-- 
2.0.0.526.g5318336

[PATCH 3/3] ARM: dts: Enable USB host0 (EHCI) on rk3288-evb

[Doug Anderson]

This is the top USB port on the evb (the one closest to the Ethernet
connector).

Signed-off-by: Doug Anderson <dianders@chromium.org>
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
---
 arch/arm/boot/dts/rk3288-evb.dtsi | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/arch/arm/boot/dts/rk3288-evb.dtsi b/arch/arm/boot/dts/rk3288-evb.dtsi
index efd625e..fcdc64e 100644
--- a/arch/arm/boot/dts/rk3288-evb.dtsi
+++ b/arch/arm/boot/dts/rk3288-evb.dtsi
@@ -90,3 +90,7 @@
 		};
 	};
 };
+
+&usb_host0_ehci {
+	status = "okay";
+};
-- 
2.0.0.526.g5318336

[PATCH 1/3] ARM: dts: Turn on USB host vbus on rk3288-evb

[Heiko Stübner]

Am Dienstag, 29. Juli 2014, 16:24:31 schrieb Doug Anderson:
> There is no phy driver that works on the Rockchip board for either USB
> host port yet.  For now just hardcode the vbus signal to be on all the
> time which makes both the dwc2 host and the EHCI port work.
> 
> Signed-off-by: Doug Anderson <dianders@chromium.org>
> ---
>  arch/arm/boot/dts/rk3288-evb.dtsi | 18 ++++++++++++++++++
>  1 file changed, 18 insertions(+)
> 
> diff --git a/arch/arm/boot/dts/rk3288-evb.dtsi
> b/arch/arm/boot/dts/rk3288-evb.dtsi index 749e20d..efd625e 100644
> --- a/arch/arm/boot/dts/rk3288-evb.dtsi
> +++ b/arch/arm/boot/dts/rk3288-evb.dtsi
> @@ -35,6 +35,18 @@
>  			debounce-interval = <100>;
>  		};
>  	};
> +
> +	/* This turns on vbus for both host0 (ehci) and host1 (dwc2) */
> +	usb_host_vbus_regulator: usb-host-vbus-regulator {
> +		compatible = "regulator-fixed";
> +		enable-active-high;
> +		gpio = <&gpio0 14 GPIO_ACTIVE_HIGH>;
> +		pinctrl-names = "default";
> +		pinctrl-0 = <&usb_host_vbus>;
> +		regulator-name = "usb-host-vbus";
> +		regulator-always-on;
> +		regulator-boot-on;
> +	};
>  };

It seems I have a slightly outdated schematics pdf for the evb ... and only 
see the OTG vbus pin, on <&gpio0 12>, but am missing the whole host vbus.

Could you think about finding another name for the handle? For example, in my 
incomplete evb-schematics the supply coming from the otg regulator is called 
vcc50_usb and there should be something similar for the host supply, so I'd 
like something like

	vcc50_usbhost: usb-host-vbus-regulator { /* or whatever it gets called */
	...
	};

simply to keep with the supply names defined in the schematics - makes reading 
easier.


Heiko

> 
>  &i2c0 {
> @@ -71,4 +83,10 @@
>  			rockchip,pins = <0 5 RK_FUNC_GPIO &pcfg_pull_up>;
>  		};
>  	};
> +
> +	usb-host {
> +		usb_host_vbus: usb-host-vbus {
> +			rockchip,pins = <0 14 RK_FUNC_GPIO &pcfg_pull_none>;
> +		};
> +	};
>  };

[PATCH 3/3] ARM: dts: Enable USB host0 (EHCI) on rk3288-evb

[Heiko Stübner]

Am Dienstag, 29. Juli 2014, 16:24:33 schrieb Doug Anderson:
> This is the top USB port on the evb (the one closest to the Ethernet
> connector).
> 
> Signed-off-by: Doug Anderson <dianders@chromium.org>
> Signed-off-by: Kever Yang <kever.yang@rock-chips.com>

shouldn't the signed-offs be the other way around, like in patch 2/3?

> ---
>  arch/arm/boot/dts/rk3288-evb.dtsi | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/arch/arm/boot/dts/rk3288-evb.dtsi
> b/arch/arm/boot/dts/rk3288-evb.dtsi index efd625e..fcdc64e 100644
> --- a/arch/arm/boot/dts/rk3288-evb.dtsi
> +++ b/arch/arm/boot/dts/rk3288-evb.dtsi
> @@ -90,3 +90,7 @@
>  		};
>  	};
>  };
> +
> +&usb_host0_ehci {
> +	status = "okay";
> +};

[PATCH 1/3] ARM: dts: Turn on USB host vbus on rk3288-evb

[Doug Anderson]

Heiko,

On Wed, Jul 30, 2014 at 4:24 AM, Heiko St?bner <heiko@sntech.de> wrote:
> Am Dienstag, 29. Juli 2014, 16:24:31 schrieb Doug Anderson:
>> There is no phy driver that works on the Rockchip board for either USB
>> host port yet.  For now just hardcode the vbus signal to be on all the
>> time which makes both the dwc2 host and the EHCI port work.
>>
>> Signed-off-by: Doug Anderson <dianders@chromium.org>
>> ---
>>  arch/arm/boot/dts/rk3288-evb.dtsi | 18 ++++++++++++++++++
>>  1 file changed, 18 insertions(+)
>>
>> diff --git a/arch/arm/boot/dts/rk3288-evb.dtsi
>> b/arch/arm/boot/dts/rk3288-evb.dtsi index 749e20d..efd625e 100644
>> --- a/arch/arm/boot/dts/rk3288-evb.dtsi
>> +++ b/arch/arm/boot/dts/rk3288-evb.dtsi
>> @@ -35,6 +35,18 @@
>>                       debounce-interval = <100>;
>>               };
>>       };
>> +
>> +     /* This turns on vbus for both host0 (ehci) and host1 (dwc2) */
>> +     usb_host_vbus_regulator: usb-host-vbus-regulator {
>> +             compatible = "regulator-fixed";
>> +             enable-active-high;
>> +             gpio = <&gpio0 14 GPIO_ACTIVE_HIGH>;
>> +             pinctrl-names = "default";
>> +             pinctrl-0 = <&usb_host_vbus>;
>> +             regulator-name = "usb-host-vbus";
>> +             regulator-always-on;
>> +             regulator-boot-on;
>> +     };
>>  };
>
> It seems I have a slightly outdated schematics pdf for the evb ... and only
> see the OTG vbus pin, on <&gpio0 12>, but am missing the whole host vbus.

I have schematics that claim to be from January 25, 2014 and claim to
be rev 1.0.  On my schematics:

* GPIO0_B4 (12) = OTG_VBUS_DRV = pin 233 of the mainboard connector
* GPIO0_B6 (14) = HOST_VBUS_DRV = pin 239 of the mainboard connector

On the mainboard schematics I have the OTG signal (233) doesn't
actually go to the OTG port.  It goes to a debug header and nowhere
else.  The HOST VBUS controls VBUS on both of the two "host" ports.


> Could you think about finding another name for the handle? For example, in my
> incomplete evb-schematics the supply coming from the otg regulator is called
> vcc50_usb and there should be something similar for the host supply, so I'd
> like something like
>
>         vcc50_usbhost: usb-host-vbus-regulator { /* or whatever it gets called */
>         ...
>         };
>
> simply to keep with the supply names defined in the schematics - makes reading
> easier.

I did!  ;)  ...but I matched my schematics, not yours.  Can you
provide the date / version number from your schematics and we can see
which is newer?  Just for reference I was emailed schematics last week
but that doesn't necessarily guarantee that they're the newest ones.


Given the above, I'm not planning to spin this patch unless you
confirm you want me to.  Thanks!  :)

-Doug

[PATCH 3/3] ARM: dts: Enable USB host0 (EHCI) on rk3288-evb

[Doug Anderson]

Heiko,

On Wed, Jul 30, 2014 at 4:26 AM, Heiko St?bner <heiko@sntech.de> wrote:
> Am Dienstag, 29. Juli 2014, 16:24:33 schrieb Doug Anderson:
>> This is the top USB port on the evb (the one closest to the Ethernet
>> connector).
>>
>> Signed-off-by: Doug Anderson <dianders@chromium.org>
>> Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
>
> shouldn't the signed-offs be the other way around, like in patch 2/3?

I wasn't aware that the order was too relevant, but I'll happily swap
them.  I generally try to list them based on contribution.  Things get
complicated when lots of WIP patches are sent around and then someone
eventually puts things together.  My work enabling the EHCI port was
based on something Kever sent me in email, so I felt he should have
SoB.  ...but cleaning it up / organizing was done by me.  Probably
should be this to get it totally right:

* Patch #1: Author and SoB are me, since Kever just pointed me to the
GPIO number
* Patch #2: Probably should be author Kever, first SoB Kever, second SoB me
* Patch #3: Probably should be author me, first SoB me, second SoB Kever

...in that case I just need to tweak the author of Patch #2 and leave
everything else the same.  I'll plan to do that when I repost.

...but I don't really care a whole lot and I'm OK with any authorship
of patch set #2/#3 and any SoB order of patch set #2/#3.  Just let me
know which you prefer!  :)

-Doug

[PATCH 2/3] ARM: dts: add rk3288 ehci usb devices

[Doug Anderson]

Hi,

On Tue, Jul 29, 2014 at 4:24 PM, Doug Anderson <dianders@chromium.org> wrote:
> rk3288 has two kind of usb controller; this adds the ehci variant for
> host0 and hsic.
>
> At the moment we don't add any phys for these controllers, but the
> default settings seem to work OK.
>
> There is a hardware problem in ohci controller which make it
> unavailable and host0 controller can only support high-speed devices.
>
> Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
> Signed-off-by: Doug Anderson <dianders@chromium.org>
> ---
>  arch/arm/boot/dts/rk3288.dtsi | 20 ++++++++++++++++++++
>  1 file changed, 20 insertions(+)
>
> diff --git a/arch/arm/boot/dts/rk3288.dtsi b/arch/arm/boot/dts/rk3288.dtsi
> index 3ef8951..0bb4480 100644
> --- a/arch/arm/boot/dts/rk3288.dtsi
> +++ b/arch/arm/boot/dts/rk3288.dtsi
> @@ -251,6 +251,26 @@
>                 status = "disabled";
>         };
>
> +       usb_host0_ehci: ehci at ff500000 {

According to sergei.shtylyov at cogentembedded.com in another thread,
this should be "usb at ff500000".

I will fix (and the hsic one) this when I spin the series shortly.

-Doug

[PATCH 1/3] ARM: dts: Turn on USB host vbus on rk3288-evb

[Heiko Stübner]

Hi Doug,

Am Mittwoch, 30. Juli 2014, 08:13:52 schrieb Doug Anderson:
> On Wed, Jul 30, 2014 at 4:24 AM, Heiko St?bner <heiko@sntech.de> wrote:
> > Am Dienstag, 29. Juli 2014, 16:24:31 schrieb Doug Anderson:
> >> There is no phy driver that works on the Rockchip board for either USB
> >> host port yet.  For now just hardcode the vbus signal to be on all the
> >> time which makes both the dwc2 host and the EHCI port work.
> >> 
> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
> >> ---
> >> 
> >>  arch/arm/boot/dts/rk3288-evb.dtsi | 18 ++++++++++++++++++
> >>  1 file changed, 18 insertions(+)
> >> 
> >> diff --git a/arch/arm/boot/dts/rk3288-evb.dtsi
> >> b/arch/arm/boot/dts/rk3288-evb.dtsi index 749e20d..efd625e 100644
> >> --- a/arch/arm/boot/dts/rk3288-evb.dtsi
> >> +++ b/arch/arm/boot/dts/rk3288-evb.dtsi
> >> @@ -35,6 +35,18 @@
> >> 
> >>                       debounce-interval = <100>;
> >>               
> >>               };
> >>       
> >>       };
> >> 
> >> +
> >> +     /* This turns on vbus for both host0 (ehci) and host1 (dwc2) */
> >> +     usb_host_vbus_regulator: usb-host-vbus-regulator {
> >> +             compatible = "regulator-fixed";
> >> +             enable-active-high;
> >> +             gpio = <&gpio0 14 GPIO_ACTIVE_HIGH>;
> >> +             pinctrl-names = "default";
> >> +             pinctrl-0 = <&usb_host_vbus>;
> >> +             regulator-name = "usb-host-vbus";
> >> +             regulator-always-on;
> >> +             regulator-boot-on;
> >> +     };
> >> 
> >>  };
> > 
> > It seems I have a slightly outdated schematics pdf for the evb ... and
> > only
> > see the OTG vbus pin, on <&gpio0 12>, but am missing the whole host vbus.
> 
> I have schematics that claim to be from January 25, 2014 and claim to
> be rev 1.0.  On my schematics:
> 
> * GPIO0_B4 (12) = OTG_VBUS_DRV = pin 233 of the mainboard connector
> * GPIO0_B6 (14) = HOST_VBUS_DRV = pin 239 of the mainboard connector
> 
> On the mainboard schematics I have the OTG signal (233) doesn't
> actually go to the OTG port.  It goes to a debug header and nowhere
> else.  The HOST VBUS controls VBUS on both of the two "host" ports.
> 
> > Could you think about finding another name for the handle? For example, in
> > my incomplete evb-schematics the supply coming from the otg regulator is
> > called vcc50_usb and there should be something similar for the host
> > supply, so I'd like something like
> > 
> >         vcc50_usbhost: usb-host-vbus-regulator { /* or whatever it gets
> >         called */
> >         ...
> >         };
> > 
> > simply to keep with the supply names defined in the schematics - makes
> > reading easier.
> 
> I did!  ;)  ...but I matched my schematics, not yours.  Can you
> provide the date / version number from your schematics and we can see
> which is newer?  Just for reference I was emailed schematics last week
> but that doesn't necessarily guarantee that they're the newest ones.

the schematics I have is "RK3288_BETA", REV 0.2, created in 2014/02/12, last 
changed on 2014/03/04.

At least in my schematics on page 16 of 44, the OTG_VBUS_DRV pin leads to a 
switch, that gets supplied by VCC50_BOOST and emits the VCC50_USB .
So, a later phy node should in the otg case probably have a
	whatever-supply = <&vcc50_usb>;
and not
	whatever-supply = <&usb_otg_vbus_regulator>


And there I'd guess the host supply will probably be structured similarly - 
even if I can't see it right now :-) .


> Given the above, I'm not planning to spin this patch unless you
> confirm you want me to.  Thanks!  :)

I'd like the regulator handle to be named after the supply name, not after the 
pin-name :-) .


Heiko

[PATCH 1/3] ARM: dts: Turn on USB host vbus on rk3288-evb

[Doug Anderson]

Heiko,

On Wed, Jul 30, 2014 at 11:24 AM, Heiko St?bner <heiko@sntech.de> wrote:
> Hi Doug,
>
> Am Mittwoch, 30. Juli 2014, 08:13:52 schrieb Doug Anderson:
>> On Wed, Jul 30, 2014 at 4:24 AM, Heiko St?bner <heiko@sntech.de> wrote:
>> > Am Dienstag, 29. Juli 2014, 16:24:31 schrieb Doug Anderson:
>> >> There is no phy driver that works on the Rockchip board for either USB
>> >> host port yet.  For now just hardcode the vbus signal to be on all the
>> >> time which makes both the dwc2 host and the EHCI port work.
>> >>
>> >> Signed-off-by: Doug Anderson <dianders@chromium.org>
>> >> ---
>> >>
>> >>  arch/arm/boot/dts/rk3288-evb.dtsi | 18 ++++++++++++++++++
>> >>  1 file changed, 18 insertions(+)
>> >>
>> >> diff --git a/arch/arm/boot/dts/rk3288-evb.dtsi
>> >> b/arch/arm/boot/dts/rk3288-evb.dtsi index 749e20d..efd625e 100644
>> >> --- a/arch/arm/boot/dts/rk3288-evb.dtsi
>> >> +++ b/arch/arm/boot/dts/rk3288-evb.dtsi
>> >> @@ -35,6 +35,18 @@
>> >>
>> >>                       debounce-interval = <100>;
>> >>
>> >>               };
>> >>
>> >>       };
>> >>
>> >> +
>> >> +     /* This turns on vbus for both host0 (ehci) and host1 (dwc2) */
>> >> +     usb_host_vbus_regulator: usb-host-vbus-regulator {
>> >> +             compatible = "regulator-fixed";
>> >> +             enable-active-high;
>> >> +             gpio = <&gpio0 14 GPIO_ACTIVE_HIGH>;
>> >> +             pinctrl-names = "default";
>> >> +             pinctrl-0 = <&usb_host_vbus>;
>> >> +             regulator-name = "usb-host-vbus";
>> >> +             regulator-always-on;
>> >> +             regulator-boot-on;
>> >> +     };
>> >>
>> >>  };
>> >
>> > It seems I have a slightly outdated schematics pdf for the evb ... and
>> > only
>> > see the OTG vbus pin, on <&gpio0 12>, but am missing the whole host vbus.
>>
>> I have schematics that claim to be from January 25, 2014 and claim to
>> be rev 1.0.  On my schematics:
>>
>> * GPIO0_B4 (12) = OTG_VBUS_DRV = pin 233 of the mainboard connector
>> * GPIO0_B6 (14) = HOST_VBUS_DRV = pin 239 of the mainboard connector
>>
>> On the mainboard schematics I have the OTG signal (233) doesn't
>> actually go to the OTG port.  It goes to a debug header and nowhere
>> else.  The HOST VBUS controls VBUS on both of the two "host" ports.
>>
>> > Could you think about finding another name for the handle? For example, in
>> > my incomplete evb-schematics the supply coming from the otg regulator is
>> > called vcc50_usb and there should be something similar for the host
>> > supply, so I'd like something like
>> >
>> >         vcc50_usbhost: usb-host-vbus-regulator { /* or whatever it gets
>> >         called */
>> >         ...
>> >         };
>> >
>> > simply to keep with the supply names defined in the schematics - makes
>> > reading easier.
>>
>> I did!  ;)  ...but I matched my schematics, not yours.  Can you
>> provide the date / version number from your schematics and we can see
>> which is newer?  Just for reference I was emailed schematics last week
>> but that doesn't necessarily guarantee that they're the newest ones.
>
> the schematics I have is "RK3288_BETA", REV 0.2, created in 2014/02/12, last
> changed on 2014/03/04.

OK, yours is clearly newer.


> At least in my schematics on page 16 of 44, the OTG_VBUS_DRV pin leads to a
> switch, that gets supplied by VCC50_BOOST and emits the VCC50_USB .
> So, a later phy node should in the otg case probably have a
>         whatever-supply = <&vcc50_usb>;
> and not
>         whatever-supply = <&usb_otg_vbus_regulator>

Ah, nice.  I wonder if that's the "2.0" mainboard?  I've seen some
people with that.  I have a "1.0 mainboard".


> And there I'd guess the host supply will probably be structured similarly -
> even if I can't see it right now :-) .
>
>
>> Given the above, I'm not planning to spin this patch unless you
>> confirm you want me to.  Thanks!  :)
>
> I'd like the regulator handle to be named after the supply name, not after the
> pin-name :-) .

Got it.  That means I can use the pin name for the pinctrl, which is nice.

One thing I'm always wary of: hardware designers tend to like to
change the names of things from revision to revision.

-Doug