[3/9] usb: dwc2/gadget: fix support for soft_connect udc framework feature

Message ID 1413464285-24172-4-git-send-email-m.szyprowski@samsung.com
State New
Headers show

Commit Message

Marek Szyprowski Oct. 16, 2014, 12:57 p.m.
Enabling and disabling usb gadget by writing to
/sys/class/udc/*hsotg/soft_connect results in calling udc_start/udc_stop
functions with the same usb gadget driver, so the driver should not WARN
about such case.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 drivers/usb/dwc2/gadget.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Felipe Balbi Oct. 16, 2014, 1:36 p.m. | #1
Hi,

On Thu, Oct 16, 2014 at 02:57:59PM +0200, Marek Szyprowski wrote:
> Enabling and disabling usb gadget by writing to
> /sys/class/udc/*hsotg/soft_connect results in calling udc_start/udc_stop
> functions with the same usb gadget driver, so the driver should not WARN
> about such case.
> 
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
>  drivers/usb/dwc2/gadget.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/usb/dwc2/gadget.c b/drivers/usb/dwc2/gadget.c
> index 8870e38c1d82..37fda4c03397 100644
> --- a/drivers/usb/dwc2/gadget.c
> +++ b/drivers/usb/dwc2/gadget.c
> @@ -2892,7 +2892,7 @@ static int s3c_hsotg_udc_start(struct usb_gadget *gadget,
>  		return -EINVAL;
>  	}
>  
> -	WARN_ON(hsotg->driver);
> +	WARN_ON(hsotg->driver && hsotg->driver != driver);

the bug is in your ->udc_stop(). You should clear hsotg->driver to NULL
there.
Marek Szyprowski Oct. 17, 2014, 10:43 a.m. | #2
Hello,

On 2014-10-16 15:36, Felipe Balbi wrote:
> On Thu, Oct 16, 2014 at 02:57:59PM +0200, Marek Szyprowski wrote:
>> Enabling and disabling usb gadget by writing to
>> /sys/class/udc/*hsotg/soft_connect results in calling udc_start/udc_stop
>> functions with the same usb gadget driver, so the driver should not WARN
>> about such case.
>>
>> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
>> ---
>>   drivers/usb/dwc2/gadget.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/usb/dwc2/gadget.c b/drivers/usb/dwc2/gadget.c
>> index 8870e38c1d82..37fda4c03397 100644
>> --- a/drivers/usb/dwc2/gadget.c
>> +++ b/drivers/usb/dwc2/gadget.c
>> @@ -2892,7 +2892,7 @@ static int s3c_hsotg_udc_start(struct usb_gadget *gadget,
>>   		return -EINVAL;
>>   	}
>>   
>> -	WARN_ON(hsotg->driver);
>> +	WARN_ON(hsotg->driver && hsotg->driver != driver);
> the bug is in your ->udc_stop(). You should clear hsotg->driver to NULL
> there.

Ok, I will change udc_stop() to always zero hsotg->driver, like other udc
drivers. I was a bit confused by the fact that udc core passes driver to
udc_stop(), when called from soft_connect and NULL on gadget removal.

Best regards
Felipe Balbi Oct. 17, 2014, 3:44 p.m. | #3
On Fri, Oct 17, 2014 at 12:43:48PM +0200, Marek Szyprowski wrote:
> Hello,
> 
> On 2014-10-16 15:36, Felipe Balbi wrote:
> >On Thu, Oct 16, 2014 at 02:57:59PM +0200, Marek Szyprowski wrote:
> >>Enabling and disabling usb gadget by writing to
> >>/sys/class/udc/*hsotg/soft_connect results in calling udc_start/udc_stop
> >>functions with the same usb gadget driver, so the driver should not WARN
> >>about such case.
> >>
> >>Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> >>---
> >>  drivers/usb/dwc2/gadget.c | 2 +-
> >>  1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >>diff --git a/drivers/usb/dwc2/gadget.c b/drivers/usb/dwc2/gadget.c
> >>index 8870e38c1d82..37fda4c03397 100644
> >>--- a/drivers/usb/dwc2/gadget.c
> >>+++ b/drivers/usb/dwc2/gadget.c
> >>@@ -2892,7 +2892,7 @@ static int s3c_hsotg_udc_start(struct usb_gadget *gadget,
> >>  		return -EINVAL;
> >>  	}
> >>-	WARN_ON(hsotg->driver);
> >>+	WARN_ON(hsotg->driver && hsotg->driver != driver);
> >the bug is in your ->udc_stop(). You should clear hsotg->driver to NULL
> >there.
> 
> Ok, I will change udc_stop() to always zero hsotg->driver, like other udc
> drivers. I was a bit confused by the fact that udc core passes driver to
> udc_stop(), when called from soft_connect and NULL on gadget removal.

That can probably be cleaned up, I'll go have a look on all UDCs and
make sure I won't break anything.

Patch

diff --git a/drivers/usb/dwc2/gadget.c b/drivers/usb/dwc2/gadget.c
index 8870e38c1d82..37fda4c03397 100644
--- a/drivers/usb/dwc2/gadget.c
+++ b/drivers/usb/dwc2/gadget.c
@@ -2892,7 +2892,7 @@  static int s3c_hsotg_udc_start(struct usb_gadget *gadget,
 		return -EINVAL;
 	}
 
-	WARN_ON(hsotg->driver);
+	WARN_ON(hsotg->driver && hsotg->driver != driver);
 
 	driver->driver.bus = NULL;
 	hsotg->driver = driver;