Hi, On Tue, Jun 28, 2011 at 08:35:47AM -0700, Greg KH wrote: > On Tue, Jun 28, 2011 at 04:34:25PM +0300, Felipe Balbi wrote: > > UDC controllers still need to call gadget > > driver's methods. For now, let's continue > > holding a gadget driver pointer on musb's > > structure until we finish the conversion > > to the udc class. > > > > Signed-off-by: Felipe Balbi <balbi@xxxxxx> > > --- > > drivers/usb/musb/musb_gadget.c | 1 + > > 1 files changed, 1 insertions(+), 0 deletions(-) > > > > diff --git a/drivers/usb/musb/musb_gadget.c b/drivers/usb/musb/musb_gadget.c > > index 3a0c465..8b62654 100644 > > --- a/drivers/usb/musb/musb_gadget.c > > +++ b/drivers/usb/musb/musb_gadget.c > > @@ -1881,6 +1881,7 @@ static int musb_gadget_start(struct usb_gadget *g, > > dev_dbg(musb->controller, "registering driver %s\n", driver->function); > > > > musb->softconnect = 0; > > + musb->gadget_driver = driver; > > How can you "hold" a pointer safely without incrementing its reference > count? done by udc-core. This is just not to break the driver completely while we're still converting to the new API. > And when is the "conversion" going to be finished? Next merge window, it's mentioned in the GIT PULL mail. -- balbi
Attachment:
signature.asc
Description: Digital signature
