diff mbox

[v2,12/12] USB: ehci-omap: Fix detection in HSIC mode

Message ID 1363086540-30567-1-git-send-email-rogerq@ti.com (mailing list archive)
State New, archived
Headers show

Commit Message

Roger Quadros March 12, 2013, 11:09 a.m. UTC
The HSIC devices need to be kept in reset while the EHCI controller
is being initialized and only brought out of reset after the
initialization is complete, else HSIC devices will not be detected.

Also remove outdated TODO list from header.

Signed-off-by: Roger Quadros <rogerq@ti.com>
CC: Alan Stern <stern@rowland.harvard.edu>
---
 drivers/usb/host/ehci-omap.c |   38 +++++++++++++++-----------------------
 1 files changed, 15 insertions(+), 23 deletions(-)

Comments

Alan Stern March 12, 2013, 4:07 p.m. UTC | #1
On Tue, 12 Mar 2013, Roger Quadros wrote:

> The HSIC devices need to be kept in reset while the EHCI controller
> is being initialized and only brought out of reset after the
> initialization is complete, else HSIC devices will not be detected.

This is not a great description of what the patch does.  You should say
that it moves PHY initialization until after EHCI initialization is
complete, instead of initializing the PHYs first, shutting them down
again, and then initializing them a second time.

> Also remove outdated TODO list from header.
> 
> Signed-off-by: Roger Quadros <rogerq@ti.com>
> CC: Alan Stern <stern@rowland.harvard.edu>
> ---
>  drivers/usb/host/ehci-omap.c |   38 +++++++++++++++-----------------------
>  1 files changed, 15 insertions(+), 23 deletions(-)
> 
> diff --git a/drivers/usb/host/ehci-omap.c b/drivers/usb/host/ehci-omap.c
> index 1ba1df8..dc2de47 100644
> --- a/drivers/usb/host/ehci-omap.c
> +++ b/drivers/usb/host/ehci-omap.c
> @@ -29,12 +29,6 @@
>   * along with this program; if not, write to the Free Software
>   * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
>   *
> - * TODO (last updated Feb 27, 2010):
> - *	- add kernel-doc
> - *	- enable AUTOIDLE
> - *	- add suspend/resume
> - *	- add HSIC and TLL support
> - *	- convert to use hwmod and runtime PM
>   */
>  
>  #include <linux/kernel.h>
> @@ -90,26 +84,13 @@ static inline u32 ehci_read(void __iomem *base, u32 reg)
>  static int omap_ehci_init(struct usb_hcd *hcd)
>  {
>  	struct ehci_hcd	*ehci = hcd_to_ehci(hcd);
> -	struct omap_hcd	*omap = (struct omap_hcd *)ehci->priv;
> -	int rc, i;
> -
> -	/* Hold PHYs in reset while initializing EHCI controller */
> -	for (i = 0; i < omap->nports; i++) {
> -		if (omap->phy[i])
> -			usb_phy_shutdown(omap->phy[i]);
> -	}
> +	int rc;
>  
>  	/* we know this is the memory we want, no need to ioremap again */
>  	ehci->caps = hcd->regs;
>  
>  	rc = ehci_setup(hcd);
>  
> -	/* Bring PHYs out of reset */
> -	for (i = 0; i < omap->nports; i++) {
> -		if (omap->phy[i])
> -			usb_phy_init(omap->phy[i]);
> -	}
> -
>  	return rc;
>  }

Now that there's almost nothing left of this function, you can get rid
of it completely.  Move the assignment to ehci->caps into
ehci_hcd_omap_probe() and remove the init entry in the overrides
structure.

Alan Stern

--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Roger Quadros March 13, 2013, 11:48 a.m. UTC | #2
On 03/12/2013 06:07 PM, Alan Stern wrote:
> On Tue, 12 Mar 2013, Roger Quadros wrote:
> 
>> The HSIC devices need to be kept in reset while the EHCI controller
>> is being initialized and only brought out of reset after the
>> initialization is complete, else HSIC devices will not be detected.
> 
> This is not a great description of what the patch does.  You should say
> that it moves PHY initialization until after EHCI initialization is
> complete, instead of initializing the PHYs first, shutting them down
> again, and then initializing them a second time.
> 

Okay will update the description.

>> Also remove outdated TODO list from header.
>>
>> Signed-off-by: Roger Quadros <rogerq@ti.com>
>> CC: Alan Stern <stern@rowland.harvard.edu>
>> ---
>>  drivers/usb/host/ehci-omap.c |   38 +++++++++++++++-----------------------
>>  1 files changed, 15 insertions(+), 23 deletions(-)
>>
>> diff --git a/drivers/usb/host/ehci-omap.c b/drivers/usb/host/ehci-omap.c
>> index 1ba1df8..dc2de47 100644
>> --- a/drivers/usb/host/ehci-omap.c
>> +++ b/drivers/usb/host/ehci-omap.c
>> @@ -29,12 +29,6 @@
>>   * along with this program; if not, write to the Free Software
>>   * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
>>   *
>> - * TODO (last updated Feb 27, 2010):
>> - *	- add kernel-doc
>> - *	- enable AUTOIDLE
>> - *	- add suspend/resume
>> - *	- add HSIC and TLL support
>> - *	- convert to use hwmod and runtime PM
>>   */
>>  
>>  #include <linux/kernel.h>
>> @@ -90,26 +84,13 @@ static inline u32 ehci_read(void __iomem *base, u32 reg)
>>  static int omap_ehci_init(struct usb_hcd *hcd)
>>  {
>>  	struct ehci_hcd	*ehci = hcd_to_ehci(hcd);
>> -	struct omap_hcd	*omap = (struct omap_hcd *)ehci->priv;
>> -	int rc, i;
>> -
>> -	/* Hold PHYs in reset while initializing EHCI controller */
>> -	for (i = 0; i < omap->nports; i++) {
>> -		if (omap->phy[i])
>> -			usb_phy_shutdown(omap->phy[i]);
>> -	}
>> +	int rc;
>>  
>>  	/* we know this is the memory we want, no need to ioremap again */
>>  	ehci->caps = hcd->regs;
>>  
>>  	rc = ehci_setup(hcd);
>>  
>> -	/* Bring PHYs out of reset */
>> -	for (i = 0; i < omap->nports; i++) {
>> -		if (omap->phy[i])
>> -			usb_phy_init(omap->phy[i]);
>> -	}
>> -
>>  	return rc;
>>  }
> 
> Now that there's almost nothing left of this function, you can get rid
> of it completely.  Move the assignment to ehci->caps into
> ehci_hcd_omap_probe() and remove the init entry in the overrides
> structure.

Right. Sounds good.

cheers,
-roger
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drivers/usb/host/ehci-omap.c b/drivers/usb/host/ehci-omap.c
index 1ba1df8..dc2de47 100644
--- a/drivers/usb/host/ehci-omap.c
+++ b/drivers/usb/host/ehci-omap.c
@@ -29,12 +29,6 @@ 
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
  *
- * TODO (last updated Feb 27, 2010):
- *	- add kernel-doc
- *	- enable AUTOIDLE
- *	- add suspend/resume
- *	- add HSIC and TLL support
- *	- convert to use hwmod and runtime PM
  */
 
 #include <linux/kernel.h>
@@ -90,26 +84,13 @@  static inline u32 ehci_read(void __iomem *base, u32 reg)
 static int omap_ehci_init(struct usb_hcd *hcd)
 {
 	struct ehci_hcd	*ehci = hcd_to_ehci(hcd);
-	struct omap_hcd	*omap = (struct omap_hcd *)ehci->priv;
-	int rc, i;
-
-	/* Hold PHYs in reset while initializing EHCI controller */
-	for (i = 0; i < omap->nports; i++) {
-		if (omap->phy[i])
-			usb_phy_shutdown(omap->phy[i]);
-	}
+	int rc;
 
 	/* we know this is the memory we want, no need to ioremap again */
 	ehci->caps = hcd->regs;
 
 	rc = ehci_setup(hcd);
 
-	/* Bring PHYs out of reset */
-	for (i = 0; i < omap->nports; i++) {
-		if (omap->phy[i])
-			usb_phy_init(omap->phy[i]);
-	}
-
 	return rc;
 }
 
@@ -219,9 +200,6 @@  static int ehci_hcd_omap_probe(struct platform_device *pdev)
 		}
 
 		omap->phy[i] = phy;
-		usb_phy_init(omap->phy[i]);
-		/* bring PHY out of suspend */
-		usb_phy_set_suspend(omap->phy[i], 0);
 	}
 
 	pm_runtime_enable(dev);
@@ -245,6 +223,20 @@  static int ehci_hcd_omap_probe(struct platform_device *pdev)
 		goto err_pm_runtime;
 	}
 
+	/*
+	 * Bring PHYs out of reset.
+	 * Even though HSIC mode is a PHY-less mode, the reset
+	 * line exists between the chips and can be modelled
+	 * as a PHY device for reset control.
+	 */
+	for (i = 0; i < omap->nports; i++) {
+		if (!omap->phy[i])
+			continue;
+
+		usb_phy_init(omap->phy[i]);
+		/* bring PHY out of suspend */
+		usb_phy_set_suspend(omap->phy[i], 0);
+	}
 
 	return 0;