On Mon, Sep 30, 2024 at 11:36:40PM +0300, Sergey Shtylyov wrote:
> On 9/30/24 19:08, Paul Barker wrote:
>
> > From: Paul Barker <paul.barker.ct@xxxxxxxxxxxxxx>
> >
> > The GbEth IP supports offloading checksum calculation for VLAN-tagged
> > packets, provided that the EtherType is 0x8100 and only one VLAN tag is
> > present.
> >
> > Signed-off-by: Paul Barker <paul.barker.ct@xxxxxxxxxxxxxx>
> [...]
>
> > diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
> > index 832132d44fb4..eb7499d42a9b 100644
> > --- a/drivers/net/ethernet/renesas/ravb_main.c
> > +++ b/drivers/net/ethernet/renesas/ravb_main.c
> > @@ -2063,11 +2063,30 @@ static void ravb_tx_timeout_work(struct work_struct *work)
> >
> > static bool ravb_can_tx_csum_gbeth(struct sk_buff *skb)
> > {
> > - /* TODO: Need to add support for VLAN tag 802.1Q */
> > - if (skb_vlan_tag_present(skb))
> > + u16 net_protocol = ntohs(skb->protocol);
> > +
> > + /* GbEth IP can calculate the checksum if:
> > + * - there are zero or one VLAN headers with TPID=0x8100
> > + * - the network protocol is IPv4 or IPv6
> > + * - the transport protocol is TCP, UDP or ICMP
> > + * - the packet is not fragmented
> > + */
> > +
> > + if (skb_vlan_tag_present(skb) &&
> > + (skb->vlan_proto != ETH_P_8021Q || net_protocol == ETH_P_8021Q))
>
> Not sure I understand this check... Maybe s/==/!=/?
A minor nit if the check stays in some form:
vlan_proto is big endian, while ETH_P_8021Q is host byte order.
>
> > return false;
> >
> > - switch (ntohs(skb->protocol)) {
> > + if (net_protocol == ETH_P_8021Q) {
> > + struct vlan_hdr vhdr, *vh;
> > +
> > + vh = skb_header_pointer(skb, ETH_HLEN, sizeof(vhdr), &vhdr);
>
> Hm, I thought the VLAN header starts at ETH_HLEN - 2, not at ETH_HLEN...
>
> [...]
>
> MBR, Sergey
>
>
