On Thu, 25 Jan 2018 13:41:58 -0200, Marcelo Ricardo Leitner wrote:
> On Wed, Jan 24, 2018 at 04:17:46PM -0800, Jakub Kicinski wrote:
> ...
> > +static inline bool
> > +tc_cls_can_offload_and_chain0(const struct net_device *dev,
> > + struct tc_cls_common_offload *common)
> > +{
> > + if (common->chain_index) {
> > + NL_SET_ERR_MSG(common->extack,
> > + "Driver supports only offload of chain 0");
> > + return false;
> > + }
> > + return tc_can_offload_extack(dev, common->extack);
>
> I know that most of the drivers updated in this patchset checks it
> this way, but considering both checks end up being done anyway in the
> success case and that performance POV on error path is irrelevant
> here, wouldn't it be better to swap both conditions here? I.e., first
> check if the device can offload, to only then check what is being
> offloaded?
>
> Otherwise the first error would be implying that the device can
> offload, just not the specified chain.
Sure, can do.
