[PATCH/RFC 0/2] usb: host: ehci-platform: fix usb 1.1 device is not connected in system resume

[PATCH/RFC 0/2] usb: host: ehci-platform: fix usb 1.1 device is not connected in system resume

[Yoshihiro Shimoda]

This patch set is based on Greg's usb.git / usb-next branch
(the commit id = 0df8a3dbacb585bb9c8b2e55de43c6aac9d86488).

This patch set is related to the following email threads:
 http://marc.info/?t=148653514200001&r=1&w=2
 http://marc.info/?t=148712534300005&r=1&w=2

Since I'm not sure the helper functions can be in core/of.c,
I sent the patch set as RFC. I guess if these functions are not needed
to other codes, the functions should be into ehci-platform.c.

Yoshihiro Shimoda (2):
  usb: of: add functions to bind a companion controller
  usb: host: ehci-platform: fix usb 1.1 device is not connected in
    system resume

 Documentation/devicetree/bindings/usb/generic.txt |  1 +
 drivers/usb/core/of.c                             | 36 +++++++++++++++++++++++
 drivers/usb/host/ehci-platform.c                  |  8 +++++
 include/linux/usb/of.h                            | 10 +++++++
 4 files changed, 55 insertions(+)

-- 
1.9.1

[PATCH/RFC 1/2] usb: of: add functions to bind a companion controller

[Yoshihiro Shimoda]

EHCI controllers will have a companion controller. However, on platform
bus, there was difficult to bind them in previous code. So, this
patch adds helper functions to bind them using a "companion" property.

Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
---
 Documentation/devicetree/bindings/usb/generic.txt |  1 +
 drivers/usb/core/of.c                             | 36 +++++++++++++++++++++++
 include/linux/usb/of.h                            | 10 +++++++
 3 files changed, 47 insertions(+)

diff --git a/Documentation/devicetree/bindings/usb/generic.txt b/Documentation/devicetree/bindings/usb/generic.txt
index bfadeb1..0a74ab8 100644
--- a/Documentation/devicetree/bindings/usb/generic.txt
+++ b/Documentation/devicetree/bindings/usb/generic.txt
@@ -22,6 +22,7 @@ Optional properties:
 			property is used if any real OTG features(HNP/SRP/ADP)
 			is enabled, if ADP is required, otg-rev should be
 			0x0200 or above.
+ - companion: phandle of a companion
  - hnp-disable: tells OTG controllers we want to disable OTG HNP, normally HNP
 			is the basic function of real OTG except you want it
 			to be a srp-capable only B device.
diff --git a/drivers/usb/core/of.c b/drivers/usb/core/of.c
index 3de4f88..033530b 100644
--- a/drivers/usb/core/of.c
+++ b/drivers/usb/core/of.c
@@ -18,6 +18,7 @@
  */
 
 #include <linux/of.h>
+#include <linux/of_platform.h>
 #include <linux/usb/of.h>
 
 /**
@@ -46,3 +47,38 @@ struct device_node *usb_of_get_child_node(struct device_node *parent,
 }
 EXPORT_SYMBOL_GPL(usb_of_get_child_node);
 
+/**
+ * usb_of_has_companion - To get if the device has a companion device
+ * @dev: the device pointer to get if it has a companion
+ *
+ * This function gets if the device has a companion property.
+ *
+ */
+bool usb_of_has_companion(struct device *dev)
+{
+	return !!of_find_property(dev->of_node, "companion", NULL);
+}
+EXPORT_SYMBOL_GPL(usb_of_has_companion);
+
+/**
+ * usb_of_get_companion_dev - Find the companion device
+ * @dev: the device pointer to find a companion
+ *
+ * Find the companion device from platform bus.
+ *
+ * Return: On success, a pointer to the companion device, %NULL on failure.
+ */
+struct device *usb_of_get_companion_dev(struct device *dev)
+{
+	struct device_node *node;
+	struct platform_device *pdev = NULL;
+
+	node = of_parse_phandle(dev->of_node, "companion", 0);
+	if (node)
+		pdev = of_find_device_by_node(node);
+
+	of_node_put(node);
+
+	return pdev ? &pdev->dev : NULL;
+}
+EXPORT_SYMBOL_GPL(usb_of_get_companion_dev);
diff --git a/include/linux/usb/of.h b/include/linux/usb/of.h
index 5ff9032..a5c7885 100644
--- a/include/linux/usb/of.h
+++ b/include/linux/usb/of.h
@@ -18,6 +18,8 @@ int of_usb_update_otg_caps(struct device_node *np,
 			struct usb_otg_caps *otg_caps);
 struct device_node *usb_of_get_child_node(struct device_node *parent,
 			int portnum);
+bool usb_of_has_companion(struct device *dev);
+struct device *usb_of_get_companion_dev(struct device *dev);
 #else
 static inline enum usb_dr_mode
 of_usb_get_dr_mode_by_phy(struct device_node *np, int arg0)
@@ -38,6 +40,14 @@ static inline int of_usb_update_otg_caps(struct device_node *np,
 {
 	return NULL;
 }
+static inline bool usb_of_has_companion(struct device *dev)
+{
+	return false;
+}
+static inline struct device *usb_of_get_companion_dev(struct device *dev)
+{
+	return NULL;
+}
 #endif
 
 #if IS_ENABLED(CONFIG_OF) && IS_ENABLED(CONFIG_USB_SUPPORT)
-- 
1.9.1

[PATCH/RFC 2/2] usb: host: ehci-platform: fix usb 1.1 device is not connected in system resume

[Yoshihiro Shimoda]

This patch fixes an issue that a usb 1.1 device is not connected in
system resume and then the following message appeared if debug messages
are enabled:
	usb 2-1: Waited 2000ms for CONNECT

To resolve this issue, the EHCI controller must be resumed after its
companion controllers. So, this patch adds such code on the driver.

Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
---
 drivers/usb/host/ehci-platform.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/usb/host/ehci-platform.c b/drivers/usb/host/ehci-platform.c
index a268d9e..65a7725 100644
--- a/drivers/usb/host/ehci-platform.c
+++ b/drivers/usb/host/ehci-platform.c
@@ -34,6 +34,7 @@
 #include <linux/usb.h>
 #include <linux/usb/hcd.h>
 #include <linux/usb/ehci_pdriver.h>
+#include <linux/usb/of.h>
 
 #include "ehci.h"
 
@@ -297,6 +298,8 @@ static int ehci_platform_probe(struct platform_device *dev)
 		goto err_power;
 
 	device_wakeup_enable(hcd->self.controller);
+	if (usb_of_has_companion(hcd->self.controller))
+		device_enable_async_suspend(hcd->self.controller);
 	platform_set_drvdata(dev, hcd);
 
 	return err;
@@ -370,6 +373,7 @@ static int ehci_platform_resume(struct device *dev)
 	struct usb_ehci_pdata *pdata = dev_get_platdata(dev);
 	struct platform_device *pdev = to_platform_device(dev);
 	struct ehci_platform_priv *priv = hcd_to_ehci_priv(hcd);
+	struct device *companion_dev;
 
 	if (pdata->power_on) {
 		int err = pdata->power_on(pdev);
@@ -377,6 +381,10 @@ static int ehci_platform_resume(struct device *dev)
 			return err;
 	}
 
+	companion_dev = usb_of_get_companion_dev(hcd->self.controller);
+	if (companion_dev)
+		device_pm_wait_for_dev(hcd->self.controller, companion_dev);
+
 	ehci_resume(hcd, priv->reset_on_resume);
 	return 0;
 }
-- 
1.9.1

Re: [PATCH/RFC 2/2] usb: host: ehci-platform: fix usb 1.1 device is not connected in system resume

[Peter Chen]

On Mon, Feb 20, 2017 at 04:35:11PM +0900, Yoshihiro Shimoda wrote:
> This patch fixes an issue that a usb 1.1 device is not connected in
> system resume and then the following message appeared if debug messages
> are enabled:
> 	usb 2-1: Waited 2000ms for CONNECT
> 
> To resolve this issue, the EHCI controller must be resumed after its
> companion controllers. So, this patch adds such code on the driver.
> 
> Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh-zM6kxYcvzFBBDgjK7y7TUQ@public.gmane.org>
> ---
>  drivers/usb/host/ehci-platform.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/drivers/usb/host/ehci-platform.c b/drivers/usb/host/ehci-platform.c
> index a268d9e..65a7725 100644
> --- a/drivers/usb/host/ehci-platform.c
> +++ b/drivers/usb/host/ehci-platform.c
> @@ -34,6 +34,7 @@
>  #include <linux/usb.h>
>  #include <linux/usb/hcd.h>
>  #include <linux/usb/ehci_pdriver.h>
> +#include <linux/usb/of.h>
>  
>  #include "ehci.h"
>  
> @@ -297,6 +298,8 @@ static int ehci_platform_probe(struct platform_device *dev)
>  		goto err_power;
>  
>  	device_wakeup_enable(hcd->self.controller);
> +	if (usb_of_has_companion(hcd->self.controller))

According to Alan's comments, all USB ehci platform device can be async
device, so this API can be skipped.

> +		device_enable_async_suspend(hcd->self.controller);
>  	platform_set_drvdata(dev, hcd);
>  
>  	return err;
> @@ -370,6 +373,7 @@ static int ehci_platform_resume(struct device *dev)
>  	struct usb_ehci_pdata *pdata = dev_get_platdata(dev);
>  	struct platform_device *pdev = to_platform_device(dev);
>  	struct ehci_platform_priv *priv = hcd_to_ehci_priv(hcd);
> +	struct device *companion_dev;
>  
>  	if (pdata->power_on) {
>  		int err = pdata->power_on(pdev);
> @@ -377,6 +381,10 @@ static int ehci_platform_resume(struct device *dev)
>  			return err;
>  	}
>  
> +	companion_dev = usb_of_get_companion_dev(hcd->self.controller);

Maybe other EHCI controller has companion controller too, so it is ok
for you to let it as a common API.

> +	if (companion_dev)
> +		device_pm_wait_for_dev(hcd->self.controller, companion_dev);
> +
>  	ehci_resume(hcd, priv->reset_on_resume);
>  	return 0;
>  }

You can send formal patch next time.

-- 

Best Regards,
Peter Chen
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH/RFC 2/2] usb: host: ehci-platform: fix usb 1.1 device is not connected in system resume

[Yoshihiro Shimoda]

Hi Peter,

> From: Peter Chen
> Sent: Tuesday, February 21, 2017 5:51 PM
> 
> On Mon, Feb 20, 2017 at 04:35:11PM +0900, Yoshihiro Shimoda wrote:
< snip >
> > @@ -297,6 +298,8 @@ static int ehci_platform_probe(struct platform_device *dev)
> >  		goto err_power;
> >
> >  	device_wakeup_enable(hcd->self.controller);
> > +	if (usb_of_has_companion(hcd->self.controller))
> 
> According to Alan's comments, all USB ehci platform device can be async
> device, so this API can be skipped.

I got it. I will remove this API.

> > +		device_enable_async_suspend(hcd->self.controller);
> >  	platform_set_drvdata(dev, hcd);
> >
> >  	return err;
> > @@ -370,6 +373,7 @@ static int ehci_platform_resume(struct device *dev)
> >  	struct usb_ehci_pdata *pdata = dev_get_platdata(dev);
> >  	struct platform_device *pdev = to_platform_device(dev);
> >  	struct ehci_platform_priv *priv = hcd_to_ehci_priv(hcd);
> > +	struct device *companion_dev;
> >
> >  	if (pdata->power_on) {
> >  		int err = pdata->power_on(pdev);
> > @@ -377,6 +381,10 @@ static int ehci_platform_resume(struct device *dev)
> >  			return err;
> >  	}
> >
> > +	companion_dev = usb_of_get_companion_dev(hcd->self.controller);
> 
> Maybe other EHCI controller has companion controller too, so it is ok
> for you to let it as a common API.

I got it.

> > +	if (companion_dev)
> > +		device_pm_wait_for_dev(hcd->self.controller, companion_dev);
> > +
> >  	ehci_resume(hcd, priv->reset_on_resume);
> >  	return 0;
> >  }
> 
> You can send formal patch next time.

Thank you for your review! I will send v2 patch set without RFC.

Best regards,
Yoshihiro Shimoda

> --
> 
> Best Regards,
> Peter Chen
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html