On Fri, Oct 25, 2024 at 2:15 AM Simon Horman <horms@xxxxxxxxxx> wrote:
>
> On Thu, Oct 24, 2024 at 12:52:38PM -0700, Rosen Penev wrote:
> > These are the preferred way to copy ethtool strings.
> >
> > Avoids incrementing pointers all over the place.
> >
> > Signed-off-by: Rosen Penev <rosenp@xxxxxxxxx>
> > Reviewed-by: Kurt Kanzenbach <kurt@xxxxxxxxxxxxx>
> > ---
> > v3: remove curly braces from ksz_common.c
> > v2: remove curly braces from rzn1_a5psw.c
>
> ...
>
> > diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c
> > index 9201f07839ad..2bb1832d21bc 100644
> > --- a/drivers/net/dsa/bcm_sf2.c
> > +++ b/drivers/net/dsa/bcm_sf2.c
> > @@ -1180,11 +1180,8 @@ static const struct b53_io_ops bcm_sf2_io_ops = {
> > static void bcm_sf2_sw_get_strings(struct dsa_switch *ds, int port,
> > u32 stringset, uint8_t *data)
> > {
> > - int cnt = b53_get_sset_count(ds, port, stringset);
> > -
> > b53_get_strings(ds, port, stringset, data);
>
> Do you need to update b53_get_strings() so that you pass &data
> rather than data? Otherwise, doesn't the call to bcm_sf2_cfp_get_strings()
> overwrite what has been written by bcm_sf2_cfp_get_strings() ?
good point. I can't change b53_get_strings as that's being passed to
.get_strings() which wants a single pointer.
I added back some pointer math for this reason.
>
> > - bcm_sf2_cfp_get_strings(ds, port, stringset,
> > - data + cnt * ETH_GSTRING_LEN);
> > + bcm_sf2_cfp_get_strings(ds, port, stringset, data);
> > }
> >
>
> ...
