Re: [PATCH] usb: misc: add a missing continue and refactor code

[Gustavo A. R. Silva]

Quoting "Gustavo A. R. Silva" <garsilva@xxxxxxxxxxxxxx>:

> Hi Alan,
>
> Quoting Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>:
>
> > On Tue, 21 Feb 2017, Gustavo A. R. Silva wrote:
> >
> > > Code refactoring to make the flow easier to follow and add missing
> > > 'continue' for case USB_ENDPOINT_XFER_INT.
> > >
> > > Addresses-Coverity-ID: 1248733
> > > Cc: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx>
> > > Signed-off-by: Gustavo A. R. Silva <garsilva@xxxxxxxxxxxxxx>
> > > ---
> > > drivers/usb/misc/usbtest.c | 50  
> > > +++++++++++++++++++++++++++-------------------
> > > 1 file changed, 30 insertions(+), 20 deletions(-)
> > >
> > > diff --git a/drivers/usb/misc/usbtest.c b/drivers/usb/misc/usbtest.c
> > > index 3525626..8723e33 100644
> > > --- a/drivers/usb/misc/usbtest.c
> > > +++ b/drivers/usb/misc/usbtest.c
> > > @@ -124,6 +124,32 @@ static struct usb_device  
> > > *testdev_to_usbdev(struct usbtest_dev *test)
> > >
> > > /*-------------------------------------------------------------------------*/
> > >
> > > +static inline void try_intr(struct usb_host_endpoint *e,
> > > +			    struct usb_host_endpoint *int_in,
> > > +			    struct usb_host_endpoint *int_out)
> > > +{
> > > +	if (usb_endpoint_dir_in(&e->desc)) {
> > > +		if (!int_in)
> > > +			int_in = e;
> > > +	} else {
> > > +		if (!int_out)
> > > +			int_out = e;
> > > +	}
> > > +}
> > > +
> > > +static inline void try_iso(struct usb_host_endpoint *e,
> > > +			   struct usb_host_endpoint *iso_in,
> > > +			   struct usb_host_endpoint *iso_out)
> > > +{
> > > +	if (usb_endpoint_dir_in(&e->desc)) {
> > > +		if (!iso_in)
> > > +			iso_in = e;
> > > +	} else {
> > > +		if (!iso_out)
> > > +			iso_out = e;
> > > +	}
> > > +}
> > > +
> >
> > This is not at all what I had in mind.  First, it's incorrect (can you
> > see why?).  Second, by "inline" I meant moving the code to be actually
> > in-line next to the conditional, not some place else in a separate
> > subroutine (even if the subroutine is declared inline).
>
> Interesting... let me double check.
>
> I thought it would've been better to have separate inline  
> subroutines for those "goto".
>
> > Also, the code for the USB_ENDPOINT_XFER_BULK case should look like the
> > other two.
>
> Do you mean a 'continue' instead of the 'break'?

Oh I see, the following piece of code should be part of the  
USB_ENDPOINT_XFER_BULK case:

if (usb_endpoint_dir_in(&e->desc)) {
        if (!in)
                in = e;
} else {
        if (!out)
                out = e;
}
continue;

--
Gustavo A. R. Silva