Re: [PATCH net-next v7 1/2] macsec: add support for IFLA_MACSEC_OFFLOAD in macsec_changelink

[Sabrina Dubroca <sd@queasysnail.net>]

2023-01-10, 09:43:37 +0100, Antoine Tenart wrote:
> Quoting Sabrina Dubroca (2023-01-09 16:14:32)
> > 2023-01-09, 10:55:56 +0200, ehakim@nvidia.com wrote:
> > > @@ -3840,6 +3835,12 @@ static int macsec_changelink(struct net_device *dev, struct nlattr *tb[],
> > >       if (ret)
> > >               goto cleanup;
> > >  
> > > +     if (data[IFLA_MACSEC_OFFLOAD]) {
> > > +             ret = macsec_update_offload(dev, nla_get_u8(data[IFLA_MACSEC_OFFLOAD]));
> > > +             if (ret)
> > > +                     goto cleanup;
> > > +     }
> > > +
> > >       /* If h/w offloading is available, propagate to the device */
> > >       if (macsec_is_offloaded(macsec)) {
> > >               const struct macsec_ops *ops;
> > 
> > There's a missing rollback of the offloading status in the (probably
> > quite unlikely) case that mdo_upd_secy fails, no? We can't fail
> > macsec_get_ops because macsec_update_offload would have failed
> > already, but I guess the driver could fail in mdo_upd_secy, and then
> > "goto cleanup" doesn't restore the offloading state.  Sorry I didn't
> > notice this earlier.
> > 
> > In case the IFLA_MACSEC_OFFLOAD attribute is provided and we're
> > enabling offload, we also end up calling the driver's mdo_add_secy,
> > and then immediately afterwards mdo_upd_secy, which probably doesn't
> > make much sense.
> > 
> > Maybe we could turn that into:
> > 
> >     if (data[IFLA_MACSEC_OFFLOAD]) {
> 
> If data[IFLA_MACSEC_OFFLOAD] is provided but doesn't change the
> offloading state, then macsec_update_offload will return early and
> mdo_upd_secy won't be called.

Ouch, thanks for catching this.

> 
> >         ... macsec_update_offload
> >     } else if (macsec_is_offloaded(macsec)) {
> >         /* If h/w offloading is available, propagate to the device */
> >         ... mdo_upd_secy
> >     }
> > 
> > Antoine, does that look reasonable to you?
> 
> But yes I agree we can improve the logic. Maybe something like:
> 
>   prev_offload = macsec->offload;
>   offload = data[IFLA_MACSEC_OFFLOAD];

That needs to be under if (data[IFLA_MACSEC_OFFLOAD]) and then the
rest gets a bit messy.

> 
>   if (prev_offload != offload) {
>       macsec_update_offload(...)
>   } else if (macsec_is_offloaded(macsec)) {
>       ...
>       prev_offload can be used to restore the offloading state on
>       failure here.
>   }

We also have a prev != new test at the start of macsec_update_offload,
the duplication is a bit ugly. We could move it out and then only call
macsec_update_offload when there is a change to do, both from
macsec_changelink and macsec_upd_offload.

Since we don't need to restore in the second branch, and we can only
fetch IFLA_MACSEC_OFFLOAD when it's present, maybe:

    change = false;
    if (data[IFLA_MACSEC_OFFLOAD]) {
        offload = nla_get_u8(data[IFLA_MACSEC_OFFLOAD]);
        if (macsec->offload != offload) {
            change = true;
            macsec_update_offload ...cleanup
        }
    }
    
    if (!change && macsec_is_offloaded(macsec)) {
        ...
    }

Or let macsec_update_offload do the macsec->offload != offload test
and pass &change so that changelink can know what to do next.

-- 
Sabrina