Re: [RFC PATCH bpf-next v5 1/2] net: Rename mono_delivery_time to tstamp_type for scalabilty

[Martin KaFai Lau]

On 4/25/24 12:02 PM, Abhishek Chauhan (ABC) wrote:
> > > > @@ -9444,7 +9444,7 @@ static struct bpf_insn *bpf_convert_tstamp_read(const struct bpf_prog *prog,
> > >   					TC_AT_INGRESS_MASK | SKB_MONO_DELIVERY_TIME_MASK);
> > >   		*insn++ = BPF_JMP32_IMM(BPF_JNE, tmp_reg,
> > >   					TC_AT_INGRESS_MASK | SKB_MONO_DELIVERY_TIME_MASK, 2);
> > > -		/* skb->tc_at_ingress && skb->mono_delivery_time,
> > > +		/* skb->tc_at_ingress && skb->tstamp_type:1,
> > Is the :1 a stale comment after we discussed how to handle the 2-bit
> This is first patch which does not add tstamp_type:2 at the moment.
> This series is divided into two patches
> 1. One patchset => Just rename (So the comment is still skb->tstamp_type:1)
> 2. Second patchset => add another bit (comment is changed to skb->tstamp_type:2)

I would suggest to completely avoid the ":1" or ":2" part in patch 1. Just use 
"... && skb->tstamp_type". The number of bits does not matter. The tstamp_type 
will still be considered as a whole even if it would become 3 bits (unlikely) in 
the future.