On Thu, Aug 30, 2012 at 11:19:15AM +0900, Chanho Park wrote:
> > -----Original Message-----
> > From: Felipe Balbi [mailto:ba...@ti.com]
> > Sent: Thursday, August 30, 2012 1:15 AM
> > To: Chanho Park
> > Cc: ba...@ti.com; linux-usb@vger.kernel.org; gre...@linuxfoundation.org;
> > Kyungmin Park
> > Subject: Re: [PATCH] usb: dwc3: correct default value of maxburst
> > 
> > Hi,
> > 
> > On Wed, Aug 29, 2012 at 10:49:58AM +0900, Chanho Park wrote:
> > > This patch changes default value of maxburst to 1. When a connection
> > > is established to highspeed, the maxburst is set to 0. So, we need to
> > > check and change default value of the maxburst before setting up burst
> > size of ep.
> > >
> > > Signed-off-by: Chanho Park <chanho61.p...@samsung.com>
> > > Signed-off-by: Kyungmin Park <kyungmin.p...@samsung.com>
> > > ---
> > >  drivers/usb/dwc3/gadget.c |    3 +++
> > >  1 file changed, 3 insertions(+)
> > >
> > > diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
> > > index 58fdfad..025a1bc 100644
> > > --- a/drivers/usb/dwc3/gadget.c
> > > +++ b/drivers/usb/dwc3/gadget.c
> > > @@ -437,6 +437,9 @@ static int dwc3_gadget_set_ep_config(struct dwc3
> > > *dwc, struct dwc3_ep *dep,
> > >
> > >   memset(&params, 0x00, sizeof(params));
> > >
> > > + if (!dep->endpoint.maxburst)
> > > +         dep->endpoint.maxburst = 1;
> > 
> > a better patch would be to only check maxburst if endpoint is ss-capable. 
> > You
> > can just check if comp_desc is a valid pointer (aka not NULL).
> 
> Thank you for your suggestion.
> I'll change the check routine like below:
> 
> +       if (!comp_desc)
> +               dep->endpoint.maxburst = 1;

not what I meant. What I mean is that maxburst doesn't exist in non-usb3
mode, so we shouldn't even write that bitfield to the HW, meaning we
shouldn't forcefully initialize dep->endpoint.maxburst either. Although
that won't cause any issues today, we cannot initialize fields we
shouldn't use in the endpoint structure.

Try to initialize that bitfield on params.param0 conditionally, instead
of hardcoding maxburst to some wrong value.

-- 
balbi

Attachment: signature.asc
Description: Digital signature

Reply via email to