[PATCH v4 0/20] MMP platform fixes

[PATCH v4 0/20] MMP platform fixes

[Lubomir Rintel]

Hi,

this series contains contains a bunch of MMP2 platform fixes.

Previous spins of the patch set were sent out over the course of last
three months to the MMP platform maintainers, with Arnd and
linux-arm-kernel@lists.infradead.org on copy.

Unfortunatelly, MMP maintainers (Eric Miao and Haojian Zhuang) don't
seem to respond anymore. That's a shame, because the MMP2 support seems to
in need for some love. The DT/multiplatform kernels can't even boot without
14/20, 15/20 and perhaps more.

I'm wondering if this cat gen in via the arm-soc tree? Would it be
appropriate if I followed up with a MAINTAINERS update in that case?

The patch set has been tested on an OLPC XO-1.75 laptop.

Thanks
Lubo

[PATCH v4 01/20] dt-bindings: mrvl,mmp-timer: add clock

[Lubomir Rintel]

The timer needs the timer clock to be enabled, otherwise it stops
ticking.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Reviewed-by: Rob Herring <robh@kernel.org>
Acked-by: Pavel Machek <pavel@ucw.cz>

---
Changes since v1:
- Made the clock optional, to keep compatibility with DTs without it

 Documentation/devicetree/bindings/timer/mrvl,mmp-timer.txt | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/Documentation/devicetree/bindings/timer/mrvl,mmp-timer.txt b/Documentation/devicetree/bindings/timer/mrvl,mmp-timer.txt
index 9a6e251462e7..b8f02c663521 100644
--- a/Documentation/devicetree/bindings/timer/mrvl,mmp-timer.txt
+++ b/Documentation/devicetree/bindings/timer/mrvl,mmp-timer.txt
@@ -5,9 +5,13 @@ Required properties:
 - reg : Address and length of the register set of timer controller.
 - interrupts : Should be the interrupt number.
 
+Optional properties:
+- clocks : Should contain a single entry describing the clock input.
+
 Example:
 	timer0: timer@d4014000 {
 		compatible = "mrvl,mmp-timer";
 		reg = <0xd4014000 0x100>;
 		interrupts = <13>;
+		clocks = <&coreclk 2>;
 	};
-- 
2.19.1

Re: [PATCH v4 01/20] dt-bindings: mrvl,mmp-timer: add clock

[Olof Johansson]

On Wed, Nov 28, 2018 at 06:53:05PM +0100, Lubomir Rintel wrote:
> The timer needs the timer clock to be enabled, otherwise it stops
> ticking.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> Reviewed-by: Rob Herring <robh@kernel.org>
> Acked-by: Pavel Machek <pavel@ucw.cz>
> 
> ---
> Changes since v1:
> - Made the clock optional, to keep compatibility with DTs without it

Applied, thanks.


-Olof

[PATCH v4 02/20] DT: marvell,mmp2: fix the gpio interrupt cell number

[Lubomir Rintel]

gpio-pxa uses two cell to encode the interrupt source: the pin number
and the trigger type. Adjust the device node accordingly.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/boot/dts/mmp2.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index 766bbb8495b6..db15d1186cd0 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -180,7 +180,7 @@
 				clocks = <&soc_clocks MMP2_CLK_GPIO>;
 				resets = <&soc_clocks MMP2_CLK_GPIO>;
 				interrupt-controller;
-				#interrupt-cells = <1>;
+				#interrupt-cells = <2>;
 				ranges;
 
 				gcb0: gpio@d4019000 {
-- 
2.19.1

Re: [PATCH v4 02/20] DT: marvell,mmp2: fix the gpio interrupt cell number

[Olof Johansson]

On Wed, Nov 28, 2018 at 06:53:06PM +0100, Lubomir Rintel wrote:
> gpio-pxa uses two cell to encode the interrupt source: the pin number
> and the trigger type. Adjust the device node accordingly.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> Acked-by: Pavel Machek <pavel@ucw.cz>
> ---

Thanks for these patches!

Applied patches 2-10 to next/dt for 4.21, using different patch prefix:

ARM: dts: mmp2: ...

Some of the others (gpio, irqchip) should be reviewed and/or applied by
the corresonding subsystem maintainers.


-Olof

[PATCH v4 03/20] DT: marvell,mmp2: give gpio node a name

[Lubomir Rintel]

This will be useful for boards that actually use GPIO pins.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/boot/dts/mmp2.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index db15d1186cd0..f2a18779de7c 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -168,7 +168,7 @@
 				status = "disabled";
 			};
 
-			gpio@d4019000 {
+			gpio: gpio@d4019000 {
 				compatible = "marvell,mmp2-gpio";
 				#address-cells = <1>;
 				#size-cells = <1>;
-- 
2.19.1

[PATCH v4 04/20] DT: marvell,mmp2: add clock to the timer

[Lubomir Rintel]

The timer needs the timer clock to be enabled, otherwise it stops
ticking.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/boot/dts/mmp2.dtsi | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index f2a18779de7c..4743a1288280 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -130,6 +130,7 @@
 				compatible = "mrvl,mmp-timer";
 				reg = <0xd4014000 0x100>;
 				interrupts = <13>;
+				clocks = <&soc_clocks MMP2_CLK_TIMER>;
 			};
 
 			uart1: uart@d4030000 {
-- 
2.19.1

[PATCH v4 05/20] DT: marvell,mmp2: add MMC controllers

[Lubomir Rintel]

There's apparently four of them on a MMP2.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>

---
Changes since v3:
- Fixed the mmc4 node name

Changes since v1:
- The controllers seem to be on AXI bus, not APB. Move them.
- Remove aliases.

 arch/arm/boot/dts/mmp2.dtsi | 36 ++++++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index 4743a1288280..1120fe6abbdc 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -117,6 +117,42 @@
 				reg-names = "mux status", "mux mask";
 				mrvl,intc-nr-irqs = <2>;
 			};
+
+			mmc1: mmc@d4280000 {
+				compatible = "mrvl,pxav3-mmc";
+				reg = <0xd4280000 0x120>;
+				clocks = <&soc_clocks MMP2_CLK_SDH0>;
+				clock-names = "io";
+				interrupts = <39>;
+				status = "disabled";
+			};
+
+			mmc2: mmc@d4280800 {
+				compatible = "mrvl,pxav3-mmc";
+				reg = <0xd4280800 0x120>;
+				clocks = <&soc_clocks MMP2_CLK_SDH1>;
+				clock-names = "io";
+				interrupts = <52>;
+				status = "disabled";
+			};
+
+			mmc3: mmc@d4281000 {
+				compatible = "mrvl,pxav3-mmc";
+				reg = <0xd4281000 0x120>;
+				clocks = <&soc_clocks MMP2_CLK_SDH2>;
+				clock-names = "io";
+				interrupts = <53>;
+				status = "disabled";
+			};
+
+			mmc4: mmc@d4281800 {
+				compatible = "mrvl,pxav3-mmc";
+				reg = <0xd4281800 0x120>;
+				clocks = <&soc_clocks MMP2_CLK_SDH3>;
+				clock-names = "io";
+				interrupts = <54>;
+				status = "disabled";
+			};
 		};
 
 		apb@d4000000 {	/* APB */
-- 
2.19.1

[PATCH v4 06/20] DT: marvell,mmp2: fix TWSI2

[Lubomir Rintel]

Marvell keeps their MMP2 datasheet secret, but there are good clues
that TWSI2 is not on 0xd4025000 on that platform, not does it use
IRQ 58. In fact, the IRQ 58 on MMP2 seems to be a signal processor:

   arch/arm/mach-mmp/irqs.h:#define IRQ_MMP2_MSP  58

I'm taking a somewhat educated guess that is probably a copy & paste
error from PXA168 or PXA910 and that the real controller in fact hides
at address 0xd4031000 and uses an interrupt line multiplexed via IRQ 17.

I'm also copying some properties from TWSI1 that were missing or
incorrect.

Tested on a OLPC XO 1.75 machine, where the RTC is on TWSI2.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Tested-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/boot/dts/mmp2.dtsi | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index 1120fe6abbdc..c5787eea57c7 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -257,12 +257,15 @@
 				status = "disabled";
 			};
 
-			twsi2: i2c@d4025000 {
+			twsi2: i2c@d4031000 {
 				compatible = "mrvl,mmp-twsi";
-				reg = <0xd4025000 0x1000>;
-				interrupts = <58>;
+				reg = <0xd4031000 0x1000>;
+				interrupt-parent = <&intcmux17>;
+				interrupts = <0>;
 				clocks = <&soc_clocks MMP2_CLK_TWSI1>;
 				resets = <&soc_clocks MMP2_CLK_TWSI1>;
+				#address-cells = <1>;
+				#size-cells = <0>;
 				status = "disabled";
 			};
 
-- 
2.19.1

[PATCH v4 07/20] DT: marvell,mmp2: add more TWSI controllers

[Lubomir Rintel]

I've gotten the base addresses, clocks and interrupts from an rusty and old
out-of-tree driver. I haven't actually checked against the datasheet, since
that one is reserved for the Marvell inner circle.

Tested with an accelerometer on TWSI6 on an OLPC XO 1.75 machine.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/boot/dts/mmp2.dtsi | 49 +++++++++++++++++++++++++++++++++++++
 1 file changed, 49 insertions(+)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index c5787eea57c7..c48d17a38d6b 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -269,6 +269,55 @@
 				status = "disabled";
 			};
 
+			twsi3: i2c@d4032000 {
+				compatible = "mrvl,mmp-twsi";
+				reg = <0xd4032000 0x1000>;
+				interrupt-parent = <&intcmux17>;
+				interrupts = <1>;
+				clocks = <&soc_clocks MMP2_CLK_TWSI2>;
+				resets = <&soc_clocks MMP2_CLK_TWSI2>;
+				#address-cells = <1>;
+				#size-cells = <0>;
+				status = "disabled";
+			};
+
+			twsi4: i2c@d4033000 {
+				compatible = "mrvl,mmp-twsi";
+				reg = <0xd4033000 0x1000>;
+				interrupt-parent = <&intcmux17>;
+				interrupts = <2>;
+				clocks = <&soc_clocks MMP2_CLK_TWSI3>;
+				resets = <&soc_clocks MMP2_CLK_TWSI3>;
+				#address-cells = <1>;
+				#size-cells = <0>;
+				status = "disabled";
+			};
+
+
+			twsi5: i2c@d4033800 {
+				compatible = "mrvl,mmp-twsi";
+				reg = <0xd4033800 0x1000>;
+				interrupt-parent = <&intcmux17>;
+				interrupts = <3>;
+				clocks = <&soc_clocks MMP2_CLK_TWSI4>;
+				resets = <&soc_clocks MMP2_CLK_TWSI4>;
+				#address-cells = <1>;
+				#size-cells = <0>;
+				status = "disabled";
+			};
+
+			twsi6: i2c@d4034000 {
+				compatible = "mrvl,mmp-twsi";
+				reg = <0xd4034000 0x1000>;
+				interrupt-parent = <&intcmux17>;
+				interrupts = <4>;
+				clocks = <&soc_clocks MMP2_CLK_TWSI5>;
+				resets = <&soc_clocks MMP2_CLK_TWSI5>;
+				#address-cells = <1>;
+				#size-cells = <0>;
+				status = "disabled";
+			};
+
 			rtc: rtc@d4010000 {
 				compatible = "mrvl,mmp-rtc";
 				reg = <0xd4010000 0x1000>;
-- 
2.19.1

[PATCH v4 08/20] DT: marvell,mmp2: add OTG PHY

[Lubomir Rintel]

The USB OTG PHY chip. To be used by the OTG controller.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/boot/dts/mmp2.dtsi | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index c48d17a38d6b..57f6248f17cd 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -118,6 +118,13 @@
 				mrvl,intc-nr-irqs = <2>;
 			};
 
+			usb_otg_phy0: usb-otg-phy@d4207000 {
+				compatible = "marvell,mmp2-usb-phy";
+				reg = <0xd4207000 0x40>;
+				#phy-cells = <0>;
+				status = "disabled";
+			};
+
 			mmc1: mmc@d4280000 {
 				compatible = "mrvl,pxav3-mmc";
 				reg = <0xd4280000 0x120>;
-- 
2.19.1

[PATCH v4 09/20] DT: marvell,mmp2: add USB OTG host controller

[Lubomir Rintel]

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/boot/dts/mmp2.dtsi | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index 57f6248f17cd..0c5a51b98c3f 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -125,6 +125,17 @@
 				status = "disabled";
 			};
 
+			usb_otg0: usb-otg@d4208000 {
+				compatible = "marvell,pxau2o-ehci";
+				reg = <0xd4208000 0x200>;
+				interrupts = <44>;
+				clocks = <&soc_clocks MMP2_CLK_USB>;
+				clock-names = "USBCLK";
+				phys = <&usb_otg_phy0>;
+				phy-names = "usb";
+				status = "disabled";
+			};
+
 			mmc1: mmc@d4280000 {
 				compatible = "mrvl,pxav3-mmc";
 				reg = <0xd4280000 0x120>;
-- 
2.19.1

[PATCH v4 10/20] DT: marvell,mmp2: Add SSP controllers

[Lubomir Rintel]

Despite Marvel keeps their base addresses secret there's a good chance
they're actually correct.

SSP1 and SSP3 bases were taken from OLPC 1.75: OpenFirmware and kernel
respectively. SSP2 and SSP4 addresses are from James Cameron who actually
has a copy of the data sheet.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>

---
Changes since v2:
- Added SSP2 and SSP4 (thanks James Cameron <quozl@laptop.org>)

Changes since v1:
- Dropped the aliases

 arch/arm/boot/dts/mmp2.dtsi | 32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)

diff --git a/arch/arm/boot/dts/mmp2.dtsi b/arch/arm/boot/dts/mmp2.dtsi
index 0c5a51b98c3f..ee03e0846740 100644
--- a/arch/arm/boot/dts/mmp2.dtsi
+++ b/arch/arm/boot/dts/mmp2.dtsi
@@ -346,6 +346,38 @@
 				resets = <&soc_clocks MMP2_CLK_RTC>;
 				status = "disabled";
 			};
+
+			ssp1: ssp@d4035000 {
+				compatible = "marvell,mmp2-ssp";
+				reg = <0xd4035000 0x1000>;
+				clocks = <&soc_clocks MMP2_CLK_SSP0>;
+				interrupts = <0>;
+				status = "disabled";
+			};
+
+			ssp2: ssp@d4036000 {
+				compatible = "marvell,mmp2-ssp";
+				reg = <0xd4036000 0x1000>;
+				clocks = <&soc_clocks MMP2_CLK_SSP1>;
+				interrupts = <1>;
+				status = "disabled";
+			};
+
+			ssp3: ssp@d4037000 {
+				compatible = "marvell,mmp2-ssp";
+				reg = <0xd4037000 0x1000>;
+				clocks = <&soc_clocks MMP2_CLK_SSP2>;
+				interrupts = <20>;
+				status = "disabled";
+			};
+
+			ssp4: ssp@d4039000 {
+				compatible = "marvell,mmp2-ssp";
+				reg = <0xd4039000 0x1000>;
+				clocks = <&soc_clocks MMP2_CLK_SSP3>;
+				interrupts = <21>;
+				status = "disabled";
+			};
 		};
 
 		soc_clocks: clocks{
-- 
2.19.1

[PATCH v4 11/20] irqchip/mmp: only touch the PJ4 & FIQ bits on enable/disable

[Lubomir Rintel]

On an OLPC XO 1.75 machine, the "security processor" handles the GPIO 71
and 72 interrupts. Don't reset the "route to SP" bit (4).

I'm just assuming the bit 4 is the "route to SP" bit -- it fixes the
SP-based keyboard for me and <mach-mmp/regs-icu.h> defines
ICU_INT_ROUTE_SP_IRQ to be 1 << 4. When asked for a data sheet, Marvell
was not helpful.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>

---
Changes since v1:
- Adjusted wording & ack from Pavel

 drivers/irqchip/irq-mmp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/irqchip/irq-mmp.c b/drivers/irqchip/irq-mmp.c
index 25f32e1d7764..1ed38f9f1d0a 100644
--- a/drivers/irqchip/irq-mmp.c
+++ b/drivers/irqchip/irq-mmp.c
@@ -190,7 +190,7 @@ static const struct mmp_intc_conf mmp_conf = {
 static const struct mmp_intc_conf mmp2_conf = {
 	.conf_enable	= 0x20,
 	.conf_disable	= 0x0,
-	.conf_mask	= 0x7f,
+	.conf_mask	= 0x60,
 };
 
 static void __exception_irq_entry mmp_handle_irq(struct pt_regs *regs)
-- 
2.19.1

[PATCH v4 12/20] irqchip/mmp: do not use of_address_to_resource() to get mux regs

[Lubomir Rintel]

The "regs" property of the "mrvl,mmp2-mux-intc" devices are silly. They
are offsets from intc's base, not addresses on the parent bus. At this
point it probably can't be fixed.

On an OLPC XO-1.75 machine, the muxes are children of the intc, not the
axi bus, and thus of_address_to_resource() won't work. We should treat
the values as mere integers as opposed to bus addresses.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
---
Changes since v3:
- Added this patch

 drivers/irqchip/irq-mmp.c | 13 ++++---------
 1 file changed, 4 insertions(+), 9 deletions(-)

diff --git a/drivers/irqchip/irq-mmp.c b/drivers/irqchip/irq-mmp.c
index 1ed38f9f1d0a..e34b57f006e3 100644
--- a/drivers/irqchip/irq-mmp.c
+++ b/drivers/irqchip/irq-mmp.c
@@ -423,9 +423,9 @@ IRQCHIP_DECLARE(mmp2_intc, "mrvl,mmp2-intc", mmp2_of_init);
 static int __init mmp2_mux_of_init(struct device_node *node,
 				   struct device_node *parent)
 {
-	struct resource res;
 	int i, ret, irq, j = 0;
 	u32 nr_irqs, mfp_irq;
+	u32 reg[4];
 
 	if (!parent)
 		return -ENODEV;
@@ -437,18 +437,13 @@ static int __init mmp2_mux_of_init(struct device_node *node,
 		pr_err("Not found mrvl,intc-nr-irqs property\n");
 		return -EINVAL;
 	}
-	ret = of_address_to_resource(node, 0, &res);
+	ret = of_property_read_u32_array(node, "reg", reg, ARRAY_SIZE(reg));
 	if (ret < 0) {
 		pr_err("Not found reg property\n");
 		return -EINVAL;
 	}
-	icu_data[i].reg_status = mmp_icu_base + res.start;
-	ret = of_address_to_resource(node, 1, &res);
-	if (ret < 0) {
-		pr_err("Not found reg property\n");
-		return -EINVAL;
-	}
-	icu_data[i].reg_mask = mmp_icu_base + res.start;
+	icu_data[i].reg_status = mmp_icu_base + reg[0];
+	icu_data[i].reg_mask = mmp_icu_base + reg[2];
 	icu_data[i].cascade_irq = irq_of_parse_and_map(node, 0);
 	if (!icu_data[i].cascade_irq)
 		return -EINVAL;
-- 
2.19.1

Re: [PATCH v4 12/20] irqchip/mmp: do not use of_address_to_resource() to get mux regs

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 780 bytes --]

On Wed 2018-11-28 18:53:16, Lubomir Rintel wrote:
> The "regs" property of the "mrvl,mmp2-mux-intc" devices are silly. They
> are offsets from intc's base, not addresses on the parent bus. At this
> point it probably can't be fixed.
> 
> On an OLPC XO-1.75 machine, the muxes are children of the intc, not the
> axi bus, and thus of_address_to_resource() won't work. We should treat
> the values as mere integers as opposed to bus addresses.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>

Comment in the code and big warning in the dts description would be
nice.

Otherwise...

Acked-by: Pavel Machek <pavel@ucw.cz>

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

[PATCH v4 13/20] gpio: pxa: avoid attempting to set pin direction via pinctrl on MMP2

[Lubomir Rintel]

Similarly to PXA3xx pinctrl-single can't set pin direction on MMP2 either.
See also: commit 9dabfdd84bdfa ("gpio: pxa: disable pinctrl calls for
PXA3xx")

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 drivers/gpio/gpio-pxa.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpio/gpio-pxa.c b/drivers/gpio/gpio-pxa.c
index bfe4c5c9f41c..1b437bbe5cf5 100644
--- a/drivers/gpio/gpio-pxa.c
+++ b/drivers/gpio/gpio-pxa.c
@@ -245,6 +245,7 @@ static bool pxa_gpio_has_pinctrl(void)
 {
 	switch (gpio_type) {
 	case PXA3XX_GPIO:
+	case MMP2_GPIO:
 		return false;
 
 	default:
-- 
2.19.1

[PATCH v4 14/20] ARM: mmp/mmp2: use cpu_is_pj4() instead of cpu_is_mmp2()

[Lubomir Rintel]

The MMP2 platform uses the PJ4 CPU. The cpu_is_mmp2() macro is thus
actually not useful at all and moreover gives the wrong result on
MACH_MMP2_DT.

The actual problem I aim to fix is that on a device-tree enabled system,
the timer ends up being initialized incorrectly. In fact, it ticks like
at rate that's 1/100 slower or so.

Perhaps the other cpu_is_mmp2() uses are more benign, but still useless.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/mach-mmp/cputype.h | 9 ---------
 arch/arm/mach-mmp/mmp2.c    | 2 +-
 arch/arm/mach-mmp/pm-mmp2.c | 2 +-
 arch/arm/mach-mmp/time.c    | 2 +-
 4 files changed, 3 insertions(+), 12 deletions(-)

diff --git a/arch/arm/mach-mmp/cputype.h b/arch/arm/mach-mmp/cputype.h
index 446edaeb78a7..405dbbbbbcdd 100644
--- a/arch/arm/mach-mmp/cputype.h
+++ b/arch/arm/mach-mmp/cputype.h
@@ -44,13 +44,4 @@ static inline int cpu_is_pxa910(void)
 #define cpu_is_pxa910()	(0)
 #endif
 
-#ifdef CONFIG_CPU_MMP2
-static inline int cpu_is_mmp2(void)
-{
-	return (((read_cpuid_id() >> 8) & 0xff) == 0x58);
-}
-#else
-#define cpu_is_mmp2()	(0)
-#endif
-
 #endif /* __ASM_MACH_CPUTYPE_H */
diff --git a/arch/arm/mach-mmp/mmp2.c b/arch/arm/mach-mmp/mmp2.c
index afba5460cdaf..b670fafedeac 100644
--- a/arch/arm/mach-mmp/mmp2.c
+++ b/arch/arm/mach-mmp/mmp2.c
@@ -104,7 +104,7 @@ void __init mmp2_init_irq(void)
 
 static int __init mmp2_init(void)
 {
-	if (cpu_is_mmp2()) {
+	if (cpu_is_pj4()) {
 #ifdef CONFIG_CACHE_TAUROS2
 		tauros2_init(0);
 #endif
diff --git a/arch/arm/mach-mmp/pm-mmp2.c b/arch/arm/mach-mmp/pm-mmp2.c
index 17699be3bc3d..bcd5111ffb37 100644
--- a/arch/arm/mach-mmp/pm-mmp2.c
+++ b/arch/arm/mach-mmp/pm-mmp2.c
@@ -220,7 +220,7 @@ static int __init mmp2_pm_init(void)
 {
 	uint32_t apcr;
 
-	if (!cpu_is_mmp2())
+	if (!cpu_is_pj4())
 		return -EIO;
 
 	suspend_set_ops(&mmp2_pm_ops);
diff --git a/arch/arm/mach-mmp/time.c b/arch/arm/mach-mmp/time.c
index 96ad1db0b04b..0f49ac579a17 100644
--- a/arch/arm/mach-mmp/time.c
+++ b/arch/arm/mach-mmp/time.c
@@ -163,7 +163,7 @@ static void __init timer_config(void)
 
 	__raw_writel(0x0, mmp_timer_base + TMR_CER); /* disable */
 
-	ccr &= (cpu_is_mmp2()) ? (TMR_CCR_CS_0(0) | TMR_CCR_CS_1(0)) :
+	ccr &= (cpu_is_pj4()) ? (TMR_CCR_CS_0(0) | TMR_CCR_CS_1(0)) :
 		(TMR_CCR_CS_0(3) | TMR_CCR_CS_1(3));
 	__raw_writel(ccr, mmp_timer_base + TMR_CCR);
 
-- 
2.19.1

Re: [PATCH v4 14/20] ARM: mmp/mmp2: use cpu_is_pj4() instead of cpu_is_mmp2()

[Arnd Bergmann]

On Wed, Nov 28, 2018 at 6:54 PM Lubomir Rintel <lkundrak@v3.sk> wrote:
>
> The MMP2 platform uses the PJ4 CPU. The cpu_is_mmp2() macro is thus
> actually not useful at all and moreover gives the wrong result on
> MACH_MMP2_DT.
>
> The actual problem I aim to fix is that on a device-tree enabled system,
> the timer ends up being initialized incorrectly. In fact, it ticks like
> at rate that's 1/100 slower or so.
>
> Perhaps the other cpu_is_mmp2() uses are more benign, but still useless.

I think we still need the helper. If it's broken on some platforms,
it has to be fixed instead.

> @@ -104,7 +104,7 @@ void __init mmp2_init_irq(void)
>
>  static int __init mmp2_init(void)
>  {
> -       if (cpu_is_mmp2()) {
> +       if (cpu_is_pj4()) {
>  #ifdef CONFIG_CACHE_TAUROS2
>                 tauros2_init(0);
>  #endif
> diff --git a/arch/arm/mach-mmp/pm-mmp2.c b/arch/arm/mach-mmp/pm-mmp2.c
> index 17699be3bc3d..bcd5111ffb37 100644
> --- a/arch/arm/mach-mmp/pm-mmp2.c
> +++ b/arch/arm/mach-mmp/pm-mmp2.c
> @@ -220,7 +220,7 @@ static int __init mmp2_pm_init(void)
>  {
>         uint32_t apcr;
>
> -       if (!cpu_is_mmp2())
> +       if (!cpu_is_pj4())
>                 return -EIO;
>
>         suspend_set_ops(&mmp2_pm_ops);

These are both regular initcalls. On a multiplatform kernel, they
will be called unconditionally, and the cpu_is_mmp2() has to
guard code that would be wrong on other SoCs, including those
with a pj4 CPU (Armada XP, Dove, Berlin).

> diff --git a/arch/arm/mach-mmp/time.c b/arch/arm/mach-mmp/time.c
> index 96ad1db0b04b..0f49ac579a17 100644
> --- a/arch/arm/mach-mmp/time.c
> +++ b/arch/arm/mach-mmp/time.c
> @@ -163,7 +163,7 @@ static void __init timer_config(void)
>
>         __raw_writel(0x0, mmp_timer_base + TMR_CER); /* disable */
>
> -       ccr &= (cpu_is_mmp2()) ? (TMR_CCR_CS_0(0) | TMR_CCR_CS_1(0)) :
> +       ccr &= (cpu_is_pj4()) ? (TMR_CCR_CS_0(0) | TMR_CCR_CS_1(0)) :
>                 (TMR_CCR_CS_0(3) | TMR_CCR_CS_1(3));
>         __raw_writel(ccr, mmp_timer_base + TMR_CCR);
>

This looks correct, here we just need to ensure we are not on pj1.
Note: we should probably rename 'timer_init' to something that
that has 'mmp' in the name, the current name conflicts with
a function of the same name in mach-davinci, which is not yet
multiplatform, but hopefully will be one day.

        Arnd

[PATCH v4 15/20] ARM: mmp2: initialize clocks before the timer

[Lubomir Rintel]

The timer shall enable its clock.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/mach-mmp/mmp2-dt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/mach-mmp/mmp2-dt.c b/arch/arm/mach-mmp/mmp2-dt.c
index 0341359b24a4..50c5e8b5be3d 100644
--- a/arch/arm/mach-mmp/mmp2-dt.c
+++ b/arch/arm/mach-mmp/mmp2-dt.c
@@ -26,8 +26,8 @@ static void __init mmp_init_time(void)
 #ifdef CONFIG_CACHE_TAUROS2
 	tauros2_init(0);
 #endif
-	mmp_dt_init_timer();
 	of_clk_init(NULL);
+	mmp_dt_init_timer();
 }
 
 static const char *const mmp2_dt_board_compat[] __initconst = {
-- 
2.19.1

Re: [PATCH v4 15/20] ARM: mmp2: initialize clocks before the timer

[Olof Johansson]

On Wed, Nov 28, 2018 at 06:53:19PM +0100, Lubomir Rintel wrote:
> The timer shall enable its clock.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> Acked-by: Pavel Machek <pavel@ucw.cz>
> ---

Applied to next/soc, thanks.


-Olof

[PATCH v4 16/20] ARM: mmp/mmp2: dt: enable the clock

[Lubomir Rintel]

The device-tree booted MMP2 needs to enable the timer clock, otherwise
it would stop ticking when the boot finishes.

It can also use the clock rate from the clk, the non-DT boards need to
keep using the hardcoded rates.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Acked-by: Pavel Machek <pavel@ucw.cz>

---
Changes since v2:
- Wording improvement, whitespace fix and an ack from Pavel

Changes since v1:
- Made the clock optional, to keep compatibility with DTs without it

 arch/arm/mach-mmp/common.h |  2 +-
 arch/arm/mach-mmp/mmp2.c   |  2 +-
 arch/arm/mach-mmp/pxa168.c |  2 +-
 arch/arm/mach-mmp/time.c   | 32 ++++++++++++++++++++------------
 4 files changed, 23 insertions(+), 15 deletions(-)

diff --git a/arch/arm/mach-mmp/common.h b/arch/arm/mach-mmp/common.h
index 7e284d9c429f..5ac2851ef5d3 100644
--- a/arch/arm/mach-mmp/common.h
+++ b/arch/arm/mach-mmp/common.h
@@ -2,7 +2,7 @@
 #include <linux/reboot.h>
 #define ARRAY_AND_SIZE(x)	(x), ARRAY_SIZE(x)
 
-extern void timer_init(int irq);
+extern void timer_init(int irq, unsigned long rate);
 
 extern void __init mmp_map_io(void);
 extern void mmp_restart(enum reboot_mode, const char *);
diff --git a/arch/arm/mach-mmp/mmp2.c b/arch/arm/mach-mmp/mmp2.c
index b670fafedeac..8717de7d0757 100644
--- a/arch/arm/mach-mmp/mmp2.c
+++ b/arch/arm/mach-mmp/mmp2.c
@@ -134,7 +134,7 @@ void __init mmp2_timer_init(void)
 	clk_rst = APBC_APBCLK | APBC_FNCLK | APBC_FNCLKSEL(1);
 	__raw_writel(clk_rst, APBC_TIMERS);
 
-	timer_init(IRQ_MMP2_TIMER1);
+	timer_init(IRQ_MMP2_TIMER1, 6500000);
 }
 
 /* on-chip devices */
diff --git a/arch/arm/mach-mmp/pxa168.c b/arch/arm/mach-mmp/pxa168.c
index 0f5f16fb8c66..77a358165a56 100644
--- a/arch/arm/mach-mmp/pxa168.c
+++ b/arch/arm/mach-mmp/pxa168.c
@@ -79,7 +79,7 @@ void __init pxa168_timer_init(void)
 	/* 3.25MHz, bus/functional clock enabled, release reset */
 	__raw_writel(TIMER_CLK_RST, APBC_TIMERS);
 
-	timer_init(IRQ_PXA168_TIMER1);
+	timer_init(IRQ_PXA168_TIMER1, 6500000);
 }
 
 void pxa168_clear_keypad_wakeup(void)
diff --git a/arch/arm/mach-mmp/time.c b/arch/arm/mach-mmp/time.c
index 0f49ac579a17..007b94a92923 100644
--- a/arch/arm/mach-mmp/time.c
+++ b/arch/arm/mach-mmp/time.c
@@ -22,6 +22,7 @@
 #include <linux/kernel.h>
 #include <linux/interrupt.h>
 #include <linux/clockchips.h>
+#include <linux/clk.h>
 
 #include <linux/io.h>
 #include <linux/irq.h>
@@ -38,12 +39,6 @@
 #include "cputype.h"
 #include "clock.h"
 
-#ifdef CONFIG_CPU_MMP2
-#define MMP_CLOCK_FREQ		6500000
-#else
-#define MMP_CLOCK_FREQ		3250000
-#endif
-
 #define TIMERS_VIRT_BASE	TIMERS1_VIRT_BASE
 
 #define MAX_DELTA		(0xfffffffe)
@@ -189,19 +184,18 @@ static struct irqaction timer_irq = {
 	.dev_id		= &ckevt,
 };
 
-void __init timer_init(int irq)
+void __init timer_init(int irq, unsigned long rate)
 {
 	timer_config();
 
-	sched_clock_register(mmp_read_sched_clock, 32, MMP_CLOCK_FREQ);
+	sched_clock_register(mmp_read_sched_clock, 32, rate);
 
 	ckevt.cpumask = cpumask_of(0);
 
 	setup_irq(irq, &timer_irq);
 
-	clocksource_register_hz(&cksrc, MMP_CLOCK_FREQ);
-	clockevents_config_and_register(&ckevt, MMP_CLOCK_FREQ,
-					MIN_DELTA, MAX_DELTA);
+	clocksource_register_hz(&cksrc, rate);
+	clockevents_config_and_register(&ckevt, rate, MIN_DELTA, MAX_DELTA);
 }
 
 #ifdef CONFIG_OF
@@ -213,7 +207,9 @@ static const struct of_device_id mmp_timer_dt_ids[] = {
 void __init mmp_dt_init_timer(void)
 {
 	struct device_node *np;
+	struct clk *clk;
 	int irq, ret;
+	unsigned long rate;
 
 	np = of_find_matching_node(NULL, mmp_timer_dt_ids);
 	if (!np) {
@@ -221,6 +217,18 @@ void __init mmp_dt_init_timer(void)
 		goto out;
 	}
 
+	clk = of_clk_get(np, 0);
+	if (!IS_ERR(clk)) {
+		ret = clk_prepare_enable(clk);
+		if (ret)
+			goto out;
+		rate = clk_get_rate(clk) / 2;
+	} else if (cpu_is_pj4()) {
+		rate = 6500000;
+	} else {
+		rate = 3250000;
+	}
+
 	irq = irq_of_parse_and_map(np, 0);
 	if (!irq) {
 		ret = -EINVAL;
@@ -231,7 +239,7 @@ void __init mmp_dt_init_timer(void)
 		ret = -ENOMEM;
 		goto out;
 	}
-	timer_init(irq);
+	timer_init(irq, rate);
 	return;
 out:
 	pr_err("Failed to get timer from device tree with error:%d\n", ret);
-- 
2.19.1

Re: [PATCH v4 16/20] ARM: mmp/mmp2: dt: enable the clock

[Olof Johansson]

On Wed, Nov 28, 2018 at 06:53:20PM +0100, Lubomir Rintel wrote:
> The device-tree booted MMP2 needs to enable the timer clock, otherwise
> it would stop ticking when the boot finishes.
> 
> It can also use the clock rate from the clk, the non-DT boards need to
> keep using the hardcoded rates.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> Acked-by: Pavel Machek <pavel@ucw.cz>

Applied to next/soc, thanks!


-Olof

[PATCH v4 17/20] ARM: mmp: add a pxa-usb-phy device

[Lubomir Rintel]

This is to replace the USB PHY initialization code (pxa_usb_phy_init(),
pxa_usb_phy_deinit()) with a proper PHY driver.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
---
 arch/arm/mach-mmp/devices.c | 21 +++++++++++++++++++++
 arch/arm/mach-mmp/pxa910.h  |  1 +
 2 files changed, 22 insertions(+)

diff --git a/arch/arm/mach-mmp/devices.c b/arch/arm/mach-mmp/devices.c
index 0fca63c80e1a..822b8be042b9 100644
--- a/arch/arm/mach-mmp/devices.c
+++ b/arch/arm/mach-mmp/devices.c
@@ -240,6 +240,27 @@ void pxa_usb_phy_deinit(void __iomem *phy_reg)
 #if IS_ENABLED(CONFIG_USB_SUPPORT)
 static u64 __maybe_unused usb_dma_mask = ~(u32)0;
 
+#if IS_ENABLED(CONFIG_PHY_PXA_USB)
+struct resource pxa168_usb_phy_resources[] = {
+	[0] = {
+		.start	= PXA168_U2O_PHYBASE,
+		.end	= PXA168_U2O_PHYBASE + USB_PHY_RANGE,
+		.flags	= IORESOURCE_MEM,
+	},
+};
+
+struct platform_device pxa168_device_usb_phy = {
+	.name		= "pxa-usb-phy",
+	.id		= -1,
+	.resource	= pxa168_usb_phy_resources,
+	.num_resources	= ARRAY_SIZE(pxa168_usb_phy_resources),
+	.dev		=  {
+		.dma_mask	= &usb_dma_mask,
+		.coherent_dma_mask = 0xffffffff,
+	}
+};
+#endif /* CONFIG_PHY_PXA_USB */
+
 #if IS_ENABLED(CONFIG_USB_MV_UDC)
 struct resource pxa168_u2o_resources[] = {
 	/* regbase */
diff --git a/arch/arm/mach-mmp/pxa910.h b/arch/arm/mach-mmp/pxa910.h
index 42009c349eae..2dfe38e4acc1 100644
--- a/arch/arm/mach-mmp/pxa910.h
+++ b/arch/arm/mach-mmp/pxa910.h
@@ -22,6 +22,7 @@ extern struct pxa_device_desc pxa910_device_pwm2;
 extern struct pxa_device_desc pxa910_device_pwm3;
 extern struct pxa_device_desc pxa910_device_pwm4;
 extern struct pxa_device_desc pxa910_device_nand;
+extern struct platform_device pxa168_device_usb_phy;
 extern struct platform_device pxa168_device_u2o;
 extern struct platform_device pxa168_device_u2ootg;
 extern struct platform_device pxa168_device_u2oehci;
-- 
2.19.1

Re: [PATCH v4 17/20] ARM: mmp: add a pxa-usb-phy device

[Olof Johansson]

On Wed, Nov 28, 2018 at 06:53:21PM +0100, Lubomir Rintel wrote:
> This is to replace the USB PHY initialization code (pxa_usb_phy_init(),
> pxa_usb_phy_deinit()) with a proper PHY driver.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> ---
>  arch/arm/mach-mmp/devices.c | 21 +++++++++++++++++++++
>  arch/arm/mach-mmp/pxa910.h  |  1 +
>  2 files changed, 22 insertions(+)

Applied to next/soc. Thanks!


-Olof

[PATCH v4 18/20] ARM: mmp: add an instance of pxa-usb-phy to ttc_dkb and aspenite

[Lubomir Rintel]

This will replace the *_pdata.phy_{de,}init()

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>

---
Changes since v1:
- Squashed patches with ttc_dkb and aspenite changes together.

 arch/arm/mach-mmp/aspenite.c | 4 ++++
 arch/arm/mach-mmp/ttc_dkb.c  | 4 ++++
 2 files changed, 8 insertions(+)

diff --git a/arch/arm/mach-mmp/aspenite.c b/arch/arm/mach-mmp/aspenite.c
index 6c2ebf01893a..23f99976b5f5 100644
--- a/arch/arm/mach-mmp/aspenite.c
+++ b/arch/arm/mach-mmp/aspenite.c
@@ -256,6 +256,10 @@ static void __init common_init(void)
 	/* off-chip devices */
 	platform_device_register(&smc91x_device);
 
+#if IS_ENABLED(CONFIG_PHY_PXA_USB)
+	platform_device_register(&pxa168_device_usb_phy);
+#endif
+
 #if IS_ENABLED(CONFIG_USB_EHCI_MV)
 	pxa168_add_usb_host(&pxa168_sph_pdata);
 #endif
diff --git a/arch/arm/mach-mmp/ttc_dkb.c b/arch/arm/mach-mmp/ttc_dkb.c
index c7897fb2b6da..767dcb23ee1c 100644
--- a/arch/arm/mach-mmp/ttc_dkb.c
+++ b/arch/arm/mach-mmp/ttc_dkb.c
@@ -282,6 +282,10 @@ static void __init ttc_dkb_init(void)
 				 sizeof(struct pxa_gpio_platform_data));
 	platform_add_devices(ARRAY_AND_SIZE(ttc_dkb_devices));
 
+#if IS_ENABLED(CONFIG_PHY_PXA_USB)
+	platform_device_register(&pxa168_device_usb_phy);
+#endif
+
 #if IS_ENABLED(CONFIG_USB_MV_UDC)
 	pxa168_device_u2o.dev.platform_data = &ttc_usb_pdata;
 	platform_device_register(&pxa168_device_u2o);
-- 
2.19.1

Re: [PATCH v4 18/20] ARM: mmp: add an instance of pxa-usb-phy to ttc_dkb and aspenite

[Olof Johansson]

On Wed, Nov 28, 2018 at 06:53:22PM +0100, Lubomir Rintel wrote:
> This will replace the *_pdata.phy_{de,}init()
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> 
> ---
> Changes since v1:
> - Squashed patches with ttc_dkb and aspenite changes together.
> 
>  arch/arm/mach-mmp/aspenite.c | 4 ++++
>  arch/arm/mach-mmp/ttc_dkb.c  | 4 ++++
>  2 files changed, 8 insertions(+)

Applied to next/soc. Thanks!


-Olof

[PATCH v4 19/20] USB: phy-mv-usb: use phy-pxa-usb

[Lubomir Rintel]

Use a proper PHY driver, instead of hooks to a board support package.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Tested-by: Pavel Machek <pavel@ucw.cz>
---
 arch/arm/mach-mmp/devices.c  |  11 +---
 drivers/usb/phy/phy-mv-usb.c | 123 ++++++++++++++++-------------------
 drivers/usb/phy/phy-mv-usb.h |   8 ++-
 3 files changed, 62 insertions(+), 80 deletions(-)

diff --git a/arch/arm/mach-mmp/devices.c b/arch/arm/mach-mmp/devices.c
index 822b8be042b9..eb9b3c34e90a 100644
--- a/arch/arm/mach-mmp/devices.c
+++ b/arch/arm/mach-mmp/devices.c
@@ -325,21 +325,12 @@ struct platform_device pxa168_device_u2oehci = {
 
 #if IS_ENABLED(CONFIG_USB_MV_OTG)
 struct resource pxa168_u2ootg_resources[] = {
-	/* regbase */
 	[0] = {
-		.start	= PXA168_U2O_REGBASE + U2x_CAPREGS_OFFSET,
+		.start	= PXA168_U2O_REGBASE,
 		.end	= PXA168_U2O_REGBASE + USB_REG_RANGE,
 		.flags	= IORESOURCE_MEM,
-		.name	= "capregs",
 	},
-	/* phybase */
 	[1] = {
-		.start	= PXA168_U2O_PHYBASE,
-		.end	= PXA168_U2O_PHYBASE + USB_PHY_RANGE,
-		.flags	= IORESOURCE_MEM,
-		.name	= "phyregs",
-	},
-	[2] = {
 		.start	= IRQ_PXA168_USB1,
 		.end	= IRQ_PXA168_USB1,
 		.flags	= IORESOURCE_IRQ,
diff --git a/drivers/usb/phy/phy-mv-usb.c b/drivers/usb/phy/phy-mv-usb.c
index cfd9add10bf4..0f67c5b65fe9 100644
--- a/drivers/usb/phy/phy-mv-usb.c
+++ b/drivers/usb/phy/phy-mv-usb.c
@@ -50,7 +50,7 @@ static char *state_string[] = {
 
 static int mv_otg_set_vbus(struct usb_otg *otg, bool on)
 {
-	struct mv_otg *mvotg = container_of(otg->usb_phy, struct mv_otg, phy);
+	struct mv_otg *mvotg = container_of(otg->usb_phy, struct mv_otg, usb_phy);
 	if (mvotg->pdata->set_vbus == NULL)
 		return -ENODEV;
 
@@ -193,7 +193,7 @@ static void mv_otg_init_irq(struct mv_otg *mvotg)
 static void mv_otg_start_host(struct mv_otg *mvotg, int on)
 {
 #ifdef CONFIG_USB
-	struct usb_otg *otg = mvotg->phy.otg;
+	struct usb_otg *otg = mvotg->usb_phy.otg;
 	struct usb_hcd *hcd;
 
 	if (!otg->host)
@@ -214,12 +214,12 @@ static void mv_otg_start_host(struct mv_otg *mvotg, int on)
 
 static void mv_otg_start_periphrals(struct mv_otg *mvotg, int on)
 {
-	struct usb_otg *otg = mvotg->phy.otg;
+	struct usb_otg *otg = mvotg->usb_phy.otg;
 
 	if (!otg->gadget)
 		return;
 
-	dev_info(mvotg->phy.dev, "gadget %s\n", on ? "on" : "off");
+	dev_info(mvotg->usb_phy.dev, "gadget %s\n", on ? "on" : "off");
 
 	if (on)
 		usb_gadget_vbus_connect(otg->gadget);
@@ -247,14 +247,11 @@ static int mv_otg_enable_internal(struct mv_otg *mvotg)
 	dev_dbg(&mvotg->pdev->dev, "otg enabled\n");
 
 	otg_clock_enable(mvotg);
-	if (mvotg->pdata->phy_init) {
-		retval = mvotg->pdata->phy_init(mvotg->phy_regs);
-		if (retval) {
-			dev_err(&mvotg->pdev->dev,
-				"init phy error %d\n", retval);
-			otg_clock_disable(mvotg);
-			return retval;
-		}
+	retval = phy_init(mvotg->phy);
+	if (retval) {
+		dev_err(&mvotg->pdev->dev, "init phy error %d\n", retval);
+		otg_clock_disable(mvotg);
+		return retval;
 	}
 	mvotg->active = 1;
 
@@ -274,8 +271,7 @@ static void mv_otg_disable_internal(struct mv_otg *mvotg)
 {
 	if (mvotg->active) {
 		dev_dbg(&mvotg->pdev->dev, "otg disabled\n");
-		if (mvotg->pdata->phy_deinit)
-			mvotg->pdata->phy_deinit(mvotg->phy_regs);
+		phy_exit(mvotg->phy);
 		otg_clock_disable(mvotg);
 		mvotg->active = 0;
 	}
@@ -329,68 +325,68 @@ static void mv_otg_update_inputs(struct mv_otg *mvotg)
 static void mv_otg_update_state(struct mv_otg *mvotg)
 {
 	struct mv_otg_ctrl *otg_ctrl = &mvotg->otg_ctrl;
-	int old_state = mvotg->phy.otg->state;
+	int old_state = mvotg->usb_phy.otg->state;
 
 	switch (old_state) {
 	case OTG_STATE_UNDEFINED:
-		mvotg->phy.otg->state = OTG_STATE_B_IDLE;
+		mvotg->usb_phy.otg->state = OTG_STATE_B_IDLE;
 		/* FALL THROUGH */
 	case OTG_STATE_B_IDLE:
 		if (otg_ctrl->id == 0)
-			mvotg->phy.otg->state = OTG_STATE_A_IDLE;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_IDLE;
 		else if (otg_ctrl->b_sess_vld)
-			mvotg->phy.otg->state = OTG_STATE_B_PERIPHERAL;
+			mvotg->usb_phy.otg->state = OTG_STATE_B_PERIPHERAL;
 		break;
 	case OTG_STATE_B_PERIPHERAL:
 		if (!otg_ctrl->b_sess_vld || otg_ctrl->id == 0)
-			mvotg->phy.otg->state = OTG_STATE_B_IDLE;
+			mvotg->usb_phy.otg->state = OTG_STATE_B_IDLE;
 		break;
 	case OTG_STATE_A_IDLE:
 		if (otg_ctrl->id)
-			mvotg->phy.otg->state = OTG_STATE_B_IDLE;
+			mvotg->usb_phy.otg->state = OTG_STATE_B_IDLE;
 		else if (!(otg_ctrl->a_bus_drop) &&
 			 (otg_ctrl->a_bus_req || otg_ctrl->a_srp_det))
-			mvotg->phy.otg->state = OTG_STATE_A_WAIT_VRISE;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_WAIT_VRISE;
 		break;
 	case OTG_STATE_A_WAIT_VRISE:
 		if (otg_ctrl->a_vbus_vld)
-			mvotg->phy.otg->state = OTG_STATE_A_WAIT_BCON;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_WAIT_BCON;
 		break;
 	case OTG_STATE_A_WAIT_BCON:
 		if (otg_ctrl->id || otg_ctrl->a_bus_drop
 		    || otg_ctrl->a_wait_bcon_timeout) {
 			mv_otg_cancel_timer(mvotg, A_WAIT_BCON_TIMER);
 			mvotg->otg_ctrl.a_wait_bcon_timeout = 0;
-			mvotg->phy.otg->state = OTG_STATE_A_WAIT_VFALL;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_WAIT_VFALL;
 			otg_ctrl->a_bus_req = 0;
 		} else if (!otg_ctrl->a_vbus_vld) {
 			mv_otg_cancel_timer(mvotg, A_WAIT_BCON_TIMER);
 			mvotg->otg_ctrl.a_wait_bcon_timeout = 0;
-			mvotg->phy.otg->state = OTG_STATE_A_VBUS_ERR;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_VBUS_ERR;
 		} else if (otg_ctrl->b_conn) {
 			mv_otg_cancel_timer(mvotg, A_WAIT_BCON_TIMER);
 			mvotg->otg_ctrl.a_wait_bcon_timeout = 0;
-			mvotg->phy.otg->state = OTG_STATE_A_HOST;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_HOST;
 		}
 		break;
 	case OTG_STATE_A_HOST:
 		if (otg_ctrl->id || !otg_ctrl->b_conn
 		    || otg_ctrl->a_bus_drop)
-			mvotg->phy.otg->state = OTG_STATE_A_WAIT_BCON;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_WAIT_BCON;
 		else if (!otg_ctrl->a_vbus_vld)
-			mvotg->phy.otg->state = OTG_STATE_A_VBUS_ERR;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_VBUS_ERR;
 		break;
 	case OTG_STATE_A_WAIT_VFALL:
 		if (otg_ctrl->id
 		    || (!otg_ctrl->b_conn && otg_ctrl->a_sess_vld)
 		    || otg_ctrl->a_bus_req)
-			mvotg->phy.otg->state = OTG_STATE_A_IDLE;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_IDLE;
 		break;
 	case OTG_STATE_A_VBUS_ERR:
 		if (otg_ctrl->id || otg_ctrl->a_clr_err
 		    || otg_ctrl->a_bus_drop) {
 			otg_ctrl->a_clr_err = 0;
-			mvotg->phy.otg->state = OTG_STATE_A_WAIT_VFALL;
+			mvotg->usb_phy.otg->state = OTG_STATE_A_WAIT_VFALL;
 		}
 		break;
 	default:
@@ -409,8 +405,8 @@ static void mv_otg_work(struct work_struct *work)
 
 run:
 	/* work queue is single thread, or we need spin_lock to protect */
-	phy = &mvotg->phy;
-	otg = mvotg->phy.otg;
+	phy = &mvotg->usb_phy;
+	otg = mvotg->usb_phy.otg;
 	old_state = otg->state;
 
 	if (!mvotg->active)
@@ -419,24 +415,24 @@ static void mv_otg_work(struct work_struct *work)
 	mv_otg_update_inputs(mvotg);
 	mv_otg_update_state(mvotg);
 
-	if (old_state != mvotg->phy.otg->state) {
+	if (old_state != mvotg->usb_phy.otg->state) {
 		dev_info(&mvotg->pdev->dev, "change from state %s to %s\n",
 			 state_string[old_state],
-			 state_string[mvotg->phy.otg->state]);
+			 state_string[mvotg->usb_phy.otg->state]);
 
-		switch (mvotg->phy.otg->state) {
+		switch (mvotg->usb_phy.otg->state) {
 		case OTG_STATE_B_IDLE:
 			otg->default_a = 0;
 			if (old_state == OTG_STATE_B_PERIPHERAL)
 				mv_otg_start_periphrals(mvotg, 0);
 			mv_otg_reset(mvotg);
 			mv_otg_disable(mvotg);
-			usb_phy_set_event(&mvotg->phy, USB_EVENT_NONE);
+			usb_phy_set_event(&mvotg->usb_phy, USB_EVENT_NONE);
 			break;
 		case OTG_STATE_B_PERIPHERAL:
 			mv_otg_enable(mvotg);
 			mv_otg_start_periphrals(mvotg, 1);
-			usb_phy_set_event(&mvotg->phy, USB_EVENT_ENUMERATED);
+			usb_phy_set_event(&mvotg->usb_phy, USB_EVENT_ENUMERATED);
 			break;
 		case OTG_STATE_A_IDLE:
 			otg->default_a = 1;
@@ -536,8 +532,8 @@ a_bus_req_store(struct device *dev, struct device_attribute *attr,
 		return -1;
 
 	/* We will use this interface to change to A device */
-	if (mvotg->phy.otg->state != OTG_STATE_B_IDLE
-	    && mvotg->phy.otg->state != OTG_STATE_A_IDLE)
+	if (mvotg->usb_phy.otg->state != OTG_STATE_B_IDLE
+	    && mvotg->usb_phy.otg->state != OTG_STATE_A_IDLE)
 		return -1;
 
 	/* The clock may disabled and we need to set irq for ID detected */
@@ -566,7 +562,7 @@ a_clr_err_store(struct device *dev, struct device_attribute *attr,
 	      const char *buf, size_t count)
 {
 	struct mv_otg *mvotg = dev_get_drvdata(dev);
-	if (!mvotg->phy.otg->default_a)
+	if (!mvotg->usb_phy.otg->default_a)
 		return -1;
 
 	if (count > 2)
@@ -602,7 +598,7 @@ a_bus_drop_store(struct device *dev, struct device_attribute *attr,
 	       const char *buf, size_t count)
 {
 	struct mv_otg *mvotg = dev_get_drvdata(dev);
-	if (!mvotg->phy.otg->default_a)
+	if (!mvotg->usb_phy.otg->default_a)
 		return -1;
 
 	if (count > 2)
@@ -656,7 +652,7 @@ static int mv_otg_remove(struct platform_device *pdev)
 
 	mv_otg_disable(mvotg);
 
-	usb_remove_phy(&mvotg->phy);
+	usb_remove_phy(&mvotg->usb_phy);
 
 	return 0;
 }
@@ -687,6 +683,10 @@ static int mv_otg_probe(struct platform_device *pdev)
 	mvotg->pdev = pdev;
 	mvotg->pdata = pdata;
 
+	mvotg->phy = devm_phy_get(&pdev->dev, "usb");
+	if (IS_ERR(mvotg->phy))
+		return PTR_ERR(mvotg->phy);
+
 	mvotg->clk = devm_clk_get(&pdev->dev, NULL);
 	if (IS_ERR(mvotg->clk))
 		return PTR_ERR(mvotg->clk);
@@ -701,12 +701,12 @@ static int mv_otg_probe(struct platform_device *pdev)
 
 	/* OTG common part */
 	mvotg->pdev = pdev;
-	mvotg->phy.dev = &pdev->dev;
-	mvotg->phy.otg = otg;
-	mvotg->phy.label = driver_name;
+	mvotg->usb_phy.dev = &pdev->dev;
+	mvotg->usb_phy.otg = otg;
+	mvotg->usb_phy.label = driver_name;
 
 	otg->state = OTG_STATE_UNDEFINED;
-	otg->usb_phy = &mvotg->phy;
+	otg->usb_phy = &mvotg->usb_phy;
 	otg->set_host = mv_otg_set_host;
 	otg->set_peripheral = mv_otg_set_peripheral;
 	otg->set_vbus = mv_otg_set_vbus;
@@ -715,36 +715,23 @@ static int mv_otg_probe(struct platform_device *pdev)
 		timer_setup(&mvotg->otg_ctrl.timer[i],
 			    mv_otg_timer_await_bcon, 0);
 
-	r = platform_get_resource_byname(mvotg->pdev,
-					 IORESOURCE_MEM, "phyregs");
-	if (r == NULL) {
-		dev_err(&pdev->dev, "no phy I/O memory resource defined\n");
-		retval = -ENODEV;
-		goto err_destroy_workqueue;
-	}
-
-	mvotg->phy_regs = devm_ioremap(&pdev->dev, r->start, resource_size(r));
-	if (mvotg->phy_regs == NULL) {
-		dev_err(&pdev->dev, "failed to map phy I/O memory\n");
-		retval = -EFAULT;
-		goto err_destroy_workqueue;
-	}
-
-	r = platform_get_resource_byname(mvotg->pdev,
-					 IORESOURCE_MEM, "capregs");
+	r = platform_get_resource(mvotg->pdev, IORESOURCE_MEM, 0);
 	if (r == NULL) {
 		dev_err(&pdev->dev, "no I/O memory resource defined\n");
 		retval = -ENODEV;
 		goto err_destroy_workqueue;
 	}
 
-	mvotg->cap_regs = devm_ioremap(&pdev->dev, r->start, resource_size(r));
-	if (mvotg->cap_regs == NULL) {
+	mvotg->base = devm_ioremap(&pdev->dev, r->start, resource_size(r));
+	if (mvotg->base == NULL) {
 		dev_err(&pdev->dev, "failed to map I/O memory\n");
 		retval = -EFAULT;
 		goto err_destroy_workqueue;
 	}
 
+	mvotg->cap_regs =
+		(void __iomem *) ((unsigned long)mvotg->base + U2x_CAPREGS_OFFSET);
+
 	/* we will acces controller register, so enable the udc controller */
 	retval = mv_otg_enable_internal(mvotg);
 	if (retval) {
@@ -804,7 +791,7 @@ static int mv_otg_probe(struct platform_device *pdev)
 		goto err_disable_clk;
 	}
 
-	retval = usb_add_phy(&mvotg->phy, USB_PHY_TYPE_USB2);
+	retval = usb_add_phy(&mvotg->usb_phy, USB_PHY_TYPE_USB2);
 	if (retval < 0) {
 		dev_err(&pdev->dev, "can't register transceiver, %d\n",
 			retval);
@@ -831,7 +818,7 @@ static int mv_otg_probe(struct platform_device *pdev)
 	return 0;
 
 err_remove_phy:
-	usb_remove_phy(&mvotg->phy);
+	usb_remove_phy(&mvotg->usb_phy);
 err_disable_clk:
 	mv_otg_disable_internal(mvotg);
 err_destroy_workqueue:
@@ -846,10 +833,10 @@ static int mv_otg_suspend(struct platform_device *pdev, pm_message_t state)
 {
 	struct mv_otg *mvotg = platform_get_drvdata(pdev);
 
-	if (mvotg->phy.otg->state != OTG_STATE_B_IDLE) {
+	if (mvotg->usb_phy.otg->state != OTG_STATE_B_IDLE) {
 		dev_info(&pdev->dev,
 			 "OTG state is not B_IDLE, it is %d!\n",
-			 mvotg->phy.otg->state);
+			 mvotg->usb_phy.otg->state);
 		return -EAGAIN;
 	}
 
diff --git a/drivers/usb/phy/phy-mv-usb.h b/drivers/usb/phy/phy-mv-usb.h
index 96701a1229ad..9b7bc6d958a8 100644
--- a/drivers/usb/phy/phy-mv-usb.h
+++ b/drivers/usb/phy/phy-mv-usb.h
@@ -8,6 +8,9 @@
 
 #include <linux/types.h>
 
+/* registers */
+#define U2x_CAPREGS_OFFSET       0x100
+
 /* Command Register Bit Masks */
 #define USBCMD_RUN_STOP			(0x00000001)
 #define USBCMD_CTRL_RESET		(0x00000002)
@@ -132,11 +135,11 @@ struct mv_otg_regs {
 };
 
 struct mv_otg {
-	struct usb_phy phy;
+	struct usb_phy usb_phy;
 	struct mv_otg_ctrl otg_ctrl;
 
 	/* base address */
-	void __iomem *phy_regs;
+	void __iomem *base;
 	void __iomem *cap_regs;
 	struct mv_otg_regs __iomem *op_regs;
 
@@ -155,6 +158,7 @@ struct mv_otg {
 	unsigned int active;
 	unsigned int clock_gating;
 	struct clk *clk;
+	struct phy *phy;
 };
 
 #endif
-- 
2.19.1

Re: [PATCH v4 19/20] USB: phy-mv-usb: use phy-pxa-usb

[Olof Johansson]

On Wed, Nov 28, 2018 at 06:53:23PM +0100, Lubomir Rintel wrote:
> Use a proper PHY driver, instead of hooks to a board support package.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> Tested-by: Pavel Machek <pavel@ucw.cz>

These should have an ack from the USB PHY maintainer before we apply them.
I don't see him cc:d on these patches though (Felipe)?


-Olof

Re: [PATCH v4 19/20] USB: phy-mv-usb: use phy-pxa-usb

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 881 bytes --]

On Fri 2018-11-30 15:42:05, Olof Johansson wrote:
> On Wed, Nov 28, 2018 at 06:53:23PM +0100, Lubomir Rintel wrote:
> > Use a proper PHY driver, instead of hooks to a board support package.
> > 
> > Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> > Tested-by: Pavel Machek <pavel@ucw.cz>
> 
> These should have an ack from the USB PHY maintainer before we apply them.
> I don't see him cc:d on these patches though (Felipe)?

I put him into cc now.

USB PHY LAYER
M:      Felipe Balbi <balbi@kernel.org>
L:      linux-usb@vger.kernel.org
T:      git
git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git
S:      Maintained
F:      drivers/usb/phy/

I guess linux-usb list should be cc-ed, too...
									Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: [PATCH v4 19/20] USB: phy-mv-usb: use phy-pxa-usb

[Olof Johansson]

On Sat, Dec 1, 2018 at 5:10 AM Pavel Machek <pavel@ucw.cz> wrote:
>
> On Fri 2018-11-30 15:42:05, Olof Johansson wrote:
> > On Wed, Nov 28, 2018 at 06:53:23PM +0100, Lubomir Rintel wrote:
> > > Use a proper PHY driver, instead of hooks to a board support package.
> > >
> > > Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
> > > Tested-by: Pavel Machek <pavel@ucw.cz>
> >
> > These should have an ack from the USB PHY maintainer before we apply them.
> > I don't see him cc:d on these patches though (Felipe)?
>
> I put him into cc now.
>
> USB PHY LAYER
> M:      Felipe Balbi <balbi@kernel.org>
> L:      linux-usb@vger.kernel.org
> T:      git
> git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git
> S:      Maintained
> F:      drivers/usb/phy/
>
> I guess linux-usb list should be cc-ed, too...

Yeah, and depending on how his mail subscriptions work, he might just
get the above email and not the patch.

Lubo, I suggest resending the remaining patches and making sure that
the maintainers are on cc (and the corresponding lists). If you ask
for their ack instead of to pick up the patch, we can merge it through
our tree.


-Olof

[PATCH v4 20/20] USB: gadget: mv-udc: use phy-pxa-usb

[Lubomir Rintel]

Use a proper PHY driver, instead of hooks to a board support package.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
---
 arch/arm/mach-mmp/devices.c          | 11 +-------
 drivers/usb/gadget/udc/mv_udc.h      |  7 ++++-
 drivers/usb/gadget/udc/mv_udc_core.c | 38 ++++++++--------------------
 3 files changed, 17 insertions(+), 39 deletions(-)

diff --git a/arch/arm/mach-mmp/devices.c b/arch/arm/mach-mmp/devices.c
index eb9b3c34e90a..d925be9f14a9 100644
--- a/arch/arm/mach-mmp/devices.c
+++ b/arch/arm/mach-mmp/devices.c
@@ -263,21 +263,12 @@ struct platform_device pxa168_device_usb_phy = {
 
 #if IS_ENABLED(CONFIG_USB_MV_UDC)
 struct resource pxa168_u2o_resources[] = {
-	/* regbase */
 	[0] = {
-		.start	= PXA168_U2O_REGBASE + U2x_CAPREGS_OFFSET,
+		.start	= PXA168_U2O_REGBASE,
 		.end	= PXA168_U2O_REGBASE + USB_REG_RANGE,
 		.flags	= IORESOURCE_MEM,
-		.name	= "capregs",
 	},
-	/* phybase */
 	[1] = {
-		.start	= PXA168_U2O_PHYBASE,
-		.end	= PXA168_U2O_PHYBASE + USB_PHY_RANGE,
-		.flags	= IORESOURCE_MEM,
-		.name	= "phyregs",
-	},
-	[2] = {
 		.start	= IRQ_PXA168_USB1,
 		.end	= IRQ_PXA168_USB1,
 		.flags	= IORESOURCE_IRQ,
diff --git a/drivers/usb/gadget/udc/mv_udc.h b/drivers/usb/gadget/udc/mv_udc.h
index b3f759c0962c..6f04f432964d 100644
--- a/drivers/usb/gadget/udc/mv_udc.h
+++ b/drivers/usb/gadget/udc/mv_udc.h
@@ -6,6 +6,9 @@
 #ifndef __MV_UDC_H
 #define __MV_UDC_H
 
+/* registers */
+#define U2x_CAPREGS_OFFSET	0x100
+
 #define VUSBHS_MAX_PORTS	8
 
 #define DQH_ALIGNMENT		2048
@@ -174,9 +177,9 @@ struct mv_udc {
 	struct platform_device		*dev;
 	int				irq;
 
+	void __iomem                    *base;
 	struct mv_cap_regs __iomem	*cap_regs;
 	struct mv_op_regs __iomem	*op_regs;
-	void __iomem                    *phy_regs;
 	unsigned int			max_eps;
 	struct mv_dqh			*ep_dqh;
 	size_t				ep_dqh_size;
@@ -219,6 +222,8 @@ struct mv_udc {
 
 	/* some SOC has mutiple clock sources for USB*/
 	struct clk      *clk;
+
+	struct phy	*phy;
 };
 
 /* endpoint data structure */
diff --git a/drivers/usb/gadget/udc/mv_udc_core.c b/drivers/usb/gadget/udc/mv_udc_core.c
index cafde053788b..35b0bde2d46f 100644
--- a/drivers/usb/gadget/udc/mv_udc_core.c
+++ b/drivers/usb/gadget/udc/mv_udc_core.c
@@ -1069,14 +1069,11 @@ static int mv_udc_enable_internal(struct mv_udc *udc)
 	if (retval)
 		return retval;
 
-	if (udc->pdata->phy_init) {
-		retval = udc->pdata->phy_init(udc->phy_regs);
-		if (retval) {
-			dev_err(&udc->dev->dev,
-				"init phy error %d\n", retval);
-			udc_clock_disable(udc);
-			return retval;
-		}
+	retval = phy_init(udc->phy);
+	if (retval) {
+		dev_err(&udc->dev->dev, "init phy error %d\n", retval);
+		udc_clock_disable(udc);
+		return retval;
 	}
 	udc->active = 1;
 
@@ -1095,8 +1092,7 @@ static void mv_udc_disable_internal(struct mv_udc *udc)
 {
 	if (udc->active) {
 		dev_dbg(&udc->dev->dev, "disable udc\n");
-		if (udc->pdata->phy_deinit)
-			udc->pdata->phy_deinit(udc->phy_regs);
+		phy_exit(udc->phy);
 		udc_clock_disable(udc);
 		udc->active = 0;
 	}
@@ -2147,30 +2143,16 @@ static int mv_udc_probe(struct platform_device *pdev)
 	if (IS_ERR(udc->clk))
 		return PTR_ERR(udc->clk);
 
-	r = platform_get_resource_byname(udc->dev, IORESOURCE_MEM, "capregs");
+	r = platform_get_resource(udc->dev, IORESOURCE_MEM, 0);
 	if (r == NULL) {
 		dev_err(&pdev->dev, "no I/O memory resource defined\n");
 		return -ENODEV;
 	}
 
-	udc->cap_regs = (struct mv_cap_regs __iomem *)
-		devm_ioremap(&pdev->dev, r->start, resource_size(r));
-	if (udc->cap_regs == NULL) {
-		dev_err(&pdev->dev, "failed to map I/O memory\n");
-		return -EBUSY;
-	}
-
-	r = platform_get_resource_byname(udc->dev, IORESOURCE_MEM, "phyregs");
-	if (r == NULL) {
-		dev_err(&pdev->dev, "no phy I/O memory resource defined\n");
-		return -ENODEV;
-	}
+	udc->base = devm_ioremap(&pdev->dev, r->start, resource_size(r));
 
-	udc->phy_regs = devm_ioremap(&pdev->dev, r->start, resource_size(r));
-	if (udc->phy_regs == NULL) {
-		dev_err(&pdev->dev, "failed to map phy I/O memory\n");
-		return -EBUSY;
-	}
+	udc->cap_regs =
+		(void __iomem *) ((unsigned long)udc->base + U2x_CAPREGS_OFFSET);
 
 	/* we will acces controller register, so enable the clk */
 	retval = mv_udc_enable_internal(udc);
-- 
2.19.1

Re: [PATCH v4 20/20] USB: gadget: mv-udc: use phy-pxa-usb

[Olof Johansson]

On Wed, Nov 28, 2018 at 06:53:24PM +0100, Lubomir Rintel wrote:
> Use a proper PHY driver, instead of hooks to a board support package.
> 
> Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>

Same here, please get Felipe to ack them if you can. Thanks!


-Olof

Re: [PATCH v4 0/20] MMP platform fixes

[Arnd Bergmann]

On Wed, Nov 28, 2018 at 6:53 PM Lubomir Rintel <lkundrak@v3.sk> wrote:
>
> Hi,
>
> this series contains contains a bunch of MMP2 platform fixes.
>
> Previous spins of the patch set were sent out over the course of last
> three months to the MMP platform maintainers, with Arnd and
> linux-arm-kernel@lists.infradead.org on copy.
>
> Unfortunatelly, MMP maintainers (Eric Miao and Haojian Zhuang) don't
> seem to respond anymore. That's a shame, because the MMP2 support seems to
> in need for some love. The DT/multiplatform kernels can't even boot without
> 14/20, 15/20 and perhaps more.
>
> I'm wondering if this cat gen in via the arm-soc tree? Would it be
> appropriate if I followed up with a MAINTAINERS update in that case?

Yes, sounds good to me. I looked through the series again, and
found one patch that seems wrong to me. Everything else should
just go in.

Patches that are purely bugfixes should probably also get backported
to stable kernels, so please add a 'Cc: stable@vger.kernel.org' tag
in the changelog for things that are required to make older kernels
work correctly (as opposed to adding features that were never supposed
to work), and describe in the changelog which kernel versions should
see that backport.

> The patch set has been tested on an OLPC XO-1.75 laptop.

Excellent!

        Arnd

Re: [PATCH v4 0/20] MMP platform fixes

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 1624 bytes --]

Hi!

> > this series contains contains a bunch of MMP2 platform fixes.
> >
> > Previous spins of the patch set were sent out over the course of last
> > three months to the MMP platform maintainers, with Arnd and
> > linux-arm-kernel@lists.infradead.org on copy.
> >
> > Unfortunatelly, MMP maintainers (Eric Miao and Haojian Zhuang) don't
> > seem to respond anymore. That's a shame, because the MMP2 support seems to
> > in need for some love. The DT/multiplatform kernels can't even boot without
> > 14/20, 15/20 and perhaps more.
> >
> > I'm wondering if this cat gen in via the arm-soc tree? Would it be
> > appropriate if I followed up with a MAINTAINERS update in that case?
> 
> Yes, sounds good to me. I looked through the series again, and
> found one patch that seems wrong to me. Everything else should
> just go in.

Sounds good to me.

> Patches that are purely bugfixes should probably also get backported
> to stable kernels, so please add a 'Cc: stable@vger.kernel.org' tag
> in the changelog for things that are required to make older kernels
> work correctly (as opposed to adding features that were never supposed
> to work), and describe in the changelog which kernel versions should
> see that backport.

Dunno, I'd be careful there. MMP2 was broken long ago and mainline
never booted on olpc-1.75 AFAICT. Seems like noone cares about
MMP2. I'd just keep it out of stable; stable is for bugs when someone
cares.

									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: [PATCH v4 0/20] MMP platform fixes

[Olof Johansson]

Hi,

On Wed, Nov 28, 2018 at 10:39:06PM +0100, Arnd Bergmann wrote:
> On Wed, Nov 28, 2018 at 6:53 PM Lubomir Rintel <lkundrak@v3.sk> wrote:
> >
> > Hi,
> >
> > this series contains contains a bunch of MMP2 platform fixes.
> >
> > Previous spins of the patch set were sent out over the course of last
> > three months to the MMP platform maintainers, with Arnd and
> > linux-arm-kernel@lists.infradead.org on copy.
> >
> > Unfortunatelly, MMP maintainers (Eric Miao and Haojian Zhuang) don't
> > seem to respond anymore. That's a shame, because the MMP2 support seems to
> > in need for some love. The DT/multiplatform kernels can't even boot without
> > 14/20, 15/20 and perhaps more.
> >
> > I'm wondering if this cat gen in via the arm-soc tree? Would it be
> > appropriate if I followed up with a MAINTAINERS update in that case?
> 
> Yes, sounds good to me. I looked through the series again, and
> found one patch that seems wrong to me. Everything else should
> just go in.

I've applied the DT patches and the ones that touch only arch/arm/mach-*.
There are a few driver patches here that should be reviewed/acked by the
corresponding maintainers as well, so I didn't apply most of them yet.

> > The patch set has been tested on an OLPC XO-1.75 laptop.
> 
> Excellent!

OOooh, I have one of those... somewhere.


-Olof

Re: [PATCH v4 0/20] MMP platform fixes

[Lubomir Rintel]

On Fri, 2018-11-30 at 15:44 -0800, Olof Johansson wrote:
> Hi,
> 
> On Wed, Nov 28, 2018 at 10:39:06PM +0100, Arnd Bergmann wrote:
> > On Wed, Nov 28, 2018 at 6:53 PM Lubomir Rintel <lkundrak@v3.sk> wrote:
> > > Hi,
> > > 
> > > this series contains contains a bunch of MMP2 platform fixes.
> > > 
> > > Previous spins of the patch set were sent out over the course of last
> > > three months to the MMP platform maintainers, with Arnd and
> > > linux-arm-kernel@lists.infradead.org on copy.
> > > 
> > > Unfortunatelly, MMP maintainers (Eric Miao and Haojian Zhuang) don't
> > > seem to respond anymore. That's a shame, because the MMP2 support seems to
> > > in need for some love. The DT/multiplatform kernels can't even boot without
> > > 14/20, 15/20 and perhaps more.
> > > 
> > > I'm wondering if this cat gen in via the arm-soc tree? Would it be
> > > appropriate if I followed up with a MAINTAINERS update in that case?
> > 
> > Yes, sounds good to me. I looked through the series again, and
> > found one patch that seems wrong to me. Everything else should
> > just go in.
> 
> I've applied the DT patches and the ones that touch only arch/arm/mach-*.
> There are a few driver patches here that should be reviewed/acked by the
> corresponding maintainers as well, so I didn't apply most of them yet.

Thank you.

> > > The patch set has been tested on an OLPC XO-1.75 laptop.
> > 
> > Excellent!
> 
> OOooh, I have one of those... somewhere.

I'm working on adjusting the machine's OFW to provide a devicetree that
matches what mainline Linux expects. I'll follow up with a boot script
in a few days -- once it's good enough to be actually useful.

There also seems to be a good chance that we'll get an updated firmware
image at some point.

Until then, CONFIG_ARM_APPENDED_DTB can be used, though a zImage header
needs to be modified to account for the DTB at the end:

https://github.com/hackerspace/olpc-xo175-linux/wiki/How-to-run-an-up-to-date-Linux-on-a-XO-1.75

> -Olof

Lubo

Re: [PATCH v4 0/20] MMP platform fixes

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 526 bytes --]

Hi!

> > > The patch set has been tested on an OLPC XO-1.75 laptop.
> > 
> > Excellent!
> 
> OOooh, I have one of those... somewhere.

Time to find it :-). It is still pretty great machine, can last 8?
hours on battery, sunlight readable screen, and if you meet a tiger
you can hit him on the head with it ... and it is still likely to work
afterwards :-).

									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: [PATCH v4 0/20] MMP platform fixes

[James Cameron]

On Sun, Dec 02, 2018 at 11:29:30PM +0100, Pavel Machek wrote:
> Hi!
> 
> > > > The patch set has been tested on an OLPC XO-1.75 laptop.
> > > 
> > > Excellent!
> > 
> > OOooh, I have one of those... somewhere.
> 
> Time to find it :-). It is still pretty great machine, can last 8?
> hours on battery, sunlight readable screen, and if you meet a tiger
> you can hit him on the head with it ... and it is still likely to work
> afterwards :-).

Yes.  Best configuration is with laptop closed, ears locked, firmly
held by handle, and use the narrow side edge to hit with.  Users have
tried the large flat surface behind the display, but the display tends
to fail after five to ten hits.  It will depend on the tiger.

-- 
James Cameron
http://quozl.netrek.org/