[PATCH v2 4/5] can: do not increase rx_bytes statistics for RTR frames

Vincent MAILHOL mailhol.vincent at wanadoo.fr
Thu Nov 25 15:28:56 PST 2021


Hi Oliver,

On. 26 Nov. 2021 at 03:50, Oliver Hartkopp <socketcan at hartkopp.net> wrote:
> Hi Vincent,
>
> On 25.11.21 18:20, Vincent Mailhol wrote:
>
> > diff --git a/drivers/net/can/usb/ucan.c b/drivers/net/can/usb/ucan.c
> > index d582c39fc8d0..717d4925fdb0 100644
> > --- a/drivers/net/can/usb/ucan.c
> > +++ b/drivers/net/can/usb/ucan.c
> > @@ -619,12 +619,13 @@ static void ucan_rx_can_msg(struct ucan_priv *up, struct ucan_message_in *m)
> >       /* copy the payload of non RTR frames */
> >       if (!(cf->can_id & CAN_RTR_FLAG) || (cf->can_id & CAN_ERR_FLAG))
> >               memcpy(cf->data, m->msg.can_msg.data, cf->len);
> > +     /* only frames which are neither RTR nor ERR have a payload */
> > +     else
> > +             stats->rx_bytes += cf->len;
>
> This 'else' path looks wrong ...

Ack. I will send a v3.
And thanks a lot for the review!

> >
> >       /* don't count error frames as real packets */
> > -     if (!(cf->can_id & CAN_ERR_FLAG)) {
> > +     if (!(cf->can_id & CAN_ERR_FLAG))
> >               stats->rx_packets++;
> > -             stats->rx_bytes += cf->len;
> > -     }
> >
> >       /* pass it to Linux */
> >       netif_rx(skb);
>
> Regards,
> Oliver



More information about the linux-arm-kernel mailing list