Messages in this thread | ![/](/images/icornerl.gif) | | From | Cong Wang <> | Date | Fri, 30 Aug 2019 13:09:29 -0700 | Subject | Re: [v2] net_sched: act_police: add 2 new attributes to support police 64bit rate and peakrate |
| |
On Fri, Aug 30, 2019 at 1:03 PM David Z. Dai <zdai@linux.vnet.ibm.com> wrote: > > On Fri, 2019-08-30 at 12:11 -0700, Cong Wang wrote: > > On Fri, Aug 30, 2019 at 12:06 PM David Dai <zdai@linux.vnet.ibm.com> wrote: > > > - if (p->peak_present) > > > + if ((police->params->rate.rate_bytes_ps >= (1ULL << 32)) && > > > + nla_put_u64_64bit(skb, TCA_POLICE_RATE64, > > > + police->params->rate.rate_bytes_ps, > > > + __TCA_POLICE_MAX)) > > > > I think the last parameter should be TCA_POLICE_PAD. > Thanks for reviewing it! > I have the impression that last parameter num value should be larger > than the attribute num value in 2nd parameter (TC_POLICE_RATE64 in this
Why do you have this impression?
> case). This is the reason I changed the last parameter value to > __TCA_POLICE_MAX after I moved the new attributes after TC_POLICE_PAD in > pkt_cls.h header.
The prototype clearly shows it must be a padding attribute:
static inline int nla_put_u64_64bit(struct sk_buff *skb, int attrtype, u64 value, int padattr)
| ![\](/images/icornerr.gif) |