Re: [Cake] [PATCH net v2] sched: consistently handle layer3 header accesses in the presence of VLANs

[Roman Mashak <mrv@mojatatu.com>]

Toke Høiland-Jørgensen <toke@redhat.com> writes:


[...]

> +/* A getter for the SKB protocol field which will handle VLAN tags consistently
> + * whether VLAN acceleration is enabled or not.
> + */
> +static inline __be16 skb_protocol(const struct sk_buff *skb, bool skip_vlan)
> +{
> +	unsigned int offset = skb_mac_offset(skb) + sizeof(struct ethhdr);
> +	__be16 proto = skb->protocol;
> +	struct vlan_hdr vhdr, *vh;

Nit: you could move vhdr and *vh definitions inside the while loop,
because of their inner scope use.

> +
> +	if (!skip_vlan)
> +		/* VLAN acceleration strips the VLAN header from the skb and
> +		 * moves it to skb->vlan_proto
> +		 */
> +		return skb_vlan_tag_present(skb) ? skb->vlan_proto : proto;
> +
> +	while (eth_type_vlan(proto)) {
> +		vh = skb_header_pointer(skb, offset, sizeof(vhdr), &vhdr);
> +		if (!vh)
> +			break;
> +
> +		proto = vh->h_vlan_encapsulated_proto;
> +		offset += sizeof(vhdr);
> +	}

[...]