Re: [RFC PATCH 1/2] usb: hcd: Remove USB phy if needed

[Valentine <valentine.barshak@cogentembedded.com>]

On 11/06/2013 09:04 PM, Felipe Balbi wrote:
> Hi,
>
> On Wed, Nov 06, 2013 at 08:43:36PM +0400, Valentine wrote:
>> On 11/06/2013 07:45 PM, Felipe Balbi wrote:
>>> Hi,
>>>
>>> On Wed, Nov 06, 2013 at 12:33:26AM +0400, Valentine Barshak wrote:
>>>> This adds remove_phy flag to the HCD structure. If the flag is
>>>> set and if hcd->phy is valid, the phy is shutdown and released
>>>> whenever usb_add_hcd fails or usb_hcd_remove is called.
>>>> This can be used by the HCD drivers to auto-remove
>>>> the external USB phy when it is no longer needed.
>>>>
>>>> Signed-off-by: Valentine Barshak <valentine.barshak@cogentembedded.com>
>>>> ---
>>>>   drivers/usb/core/hcd.c  | 14 +++++++++++++-
>>>>   include/linux/usb/hcd.h |  1 +
>>>>   2 files changed, 14 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c
>>>> index d6a8d23..d939521 100644
>>>> --- a/drivers/usb/core/hcd.c
>>>> +++ b/drivers/usb/core/hcd.c
>>>> @@ -43,6 +43,7 @@
>>>>
>>>>   #include <linux/usb.h>
>>>>   #include <linux/usb/hcd.h>
>>>> +#include <linux/usb/phy.h>
>>>>
>>>>   #include "usb.h"
>>>>
>>>> @@ -2611,7 +2612,7 @@ int usb_add_hcd(struct usb_hcd *hcd,
>>>>   	 */
>>>>   	if ((retval = hcd_buffer_create(hcd)) != 0) {
>>>>   		dev_dbg(hcd->self.controller, "pool alloc failed\n");
>>>> -		return retval;
>>>> +		goto err_remove_phy;
>>>>   	}
>>>>
>>>>   	if ((retval = usb_register_bus(&hcd->self)) < 0)
>>>> @@ -2742,6 +2743,12 @@ err_allocate_root_hub:
>>>>   	usb_deregister_bus(&hcd->self);
>>>>   err_register_bus:
>>>>   	hcd_buffer_destroy(hcd);
>>>> +err_remove_phy:
>>>> +	if (hcd->remove_phy && hcd->phy) {
>>>> +		usb_phy_shutdown(hcd->phy);
>>>> +		usb_put_phy(hcd->phy);
>>>> +		hcd->phy = NULL;
>>>> +	}
>>>
>>> why do you need the flag at all ? If hcd->phy is valid, just casll
>>> usb_phy_shutdown() followed by usb_put_phy(). Did you find any issues
>>> with that ?
>>>
>>
>> I haven't been able to test it with other platforms.
>> However, some drivers use devm_usb_get_phy_dev() and we may face refcounter issues
>> if we call usb_put_phy unconditionally.
>> Adding this flag seems safe enough and we doesn't affect other drivers.
>
> then use devm_usb_get_phy_dev() here and remove the call from all other
> drivers ;-)
>

The glue drivers may use different ways of getting the USB phy:
usb_get_phy; usb_get_phy_dev; devm_usb_get_phy_dev.
We can't consolidate phy getting here since usb_get_phy and usb_get_phy_dev are not equivalent.
So we let the glue drivers decide which way of getting/putting USB phy to use.

Thanks,
Val.