On Wed, 9 Oct 2019 at 02:58, Vladimir Oltean <olte...@gmail.com> wrote:
>
> Hi Vinicius,
>
> On Wed, 9 Oct 2019 at 02:19, Vinicius Costa Gomes
> <vinicius.go...@intel.com> wrote:
> >
> > When configuring a taprio instance if "flags" is not specified (or
> > it's zero), taprio currently replies with an "Invalid argument" error.
> >
> > So, set the return value to zero after we are done with all the
> > checks.
> >
> > Fixes: 9c66d1564676 ("taprio: Add support for hardware offloading")
> > Signed-off-by: Vinicius Costa Gomes <vinicius.go...@intel.com>
> > ---
>
> You mean clockid, not flags, right?
> Otherwise the patch looks correct, sorry for the bug.
> Once you fix the commit message:
>
> Acked-by: Vladimir Oltean <olte...@gmail.com>
>No, you are actually right to phrase it this way, but the description is still confusing. When 'flags' is zero, the driver takes a code path which is buggy and always returns a negative error code. Although I'm not really sure how much better this can be phrased. I'm fine with the description now too. > > net/sched/sch_taprio.c | 4 ++++ > > 1 file changed, 4 insertions(+) > > > > diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c > > index 68b543f85a96..6719a65169d4 100644 > > --- a/net/sched/sch_taprio.c > > +++ b/net/sched/sch_taprio.c > > @@ -1341,6 +1341,10 @@ static int taprio_parse_clockid(struct Qdisc *sch, > > struct nlattr **tb, > > NL_SET_ERR_MSG(extack, "Specifying a 'clockid' is > > mandatory"); > > goto out; > > } > > + > > + /* Everything went ok, return success. */ > > + err = 0; > > + > > out: > > return err; > > } > > -- > > 2.23.0 > >
