Re: [PATCH v2 net-next v2 2/5] net: add support for segmenting TCP fraglist GSO packets

[Felix Fietkau]

On 25.04.24 17:17, Willem de Bruijn wrote:
> Felix Fietkau wrote:
> > Preparation for adding TCP fraglist GRO support. It expects packets to be
> > combined in a similar way as UDP fraglist GSO packets.
> > For IPv4 packets, NAT is handled in the same way as UDP fraglist GSO.
> >
> > Signed-off-by: Felix Fietkau <nbd@xxxxxxxx>
> > ---
> >  net/ipv4/tcp_offload.c   | 69 ++++++++++++++++++++++++++++++++++++++++
> >  net/ipv6/tcpv6_offload.c |  3 ++
> >  2 files changed, 72 insertions(+)
> >
> > diff --git a/net/ipv4/tcp_offload.c b/net/ipv4/tcp_offload.c
> > index fab0973f995b..e455f884190c 100644
> > --- a/net/ipv4/tcp_offload.c
> > +++ b/net/ipv4/tcp_offload.c
> > @@ -28,6 +28,72 @@ static void tcp_gso_tstamp(struct sk_buff *skb, unsigned int ts_seq,
> >  	}
> >  }
> >  
> > +static void __tcpv4_gso_segment_csum(struct sk_buff *seg,
> > +				     __be32 *oldip, __be32 *newip,
> > +				     __be16 *oldport, __be16 *newport)
> > +{
> > +	struct tcphdr *th;
> > +	struct iphdr *iph;
> > +
> > +	if (*oldip == *newip && *oldport == *newport)
> > +		return;
> > +
> > +	th = tcp_hdr(seg);
> > +	iph = ip_hdr(seg);
> > +
> > +	inet_proto_csum_replace4(&th->check, seg, *oldip, *newip, true);
> > +	inet_proto_csum_replace2(&th->check, seg, *oldport, *newport, false);
> > +	*oldport = *newport;
> > +
> > +	csum_replace4(&iph->check, *oldip, *newip);
> > +	*oldip = *newip;
> > +}
> > +
> > +static struct sk_buff *__tcpv4_gso_segment_list_csum(struct sk_buff *segs)
> > +{
> > +	struct sk_buff *seg;
> > +	struct tcphdr *th, *th2;
> > +	struct iphdr *iph, *iph2;
> > +	__be32 flags, flags2;
> > +
> > +	seg = segs;
> > +	th = tcp_hdr(seg);
> > +	iph = ip_hdr(seg);
> > +	flags = tcp_flag_word(th);
> > +	flags2 = tcp_flag_word(tcp_hdr(seg->next));
>
> Vestigial, now that flag overwrite is removed in v2?

Will fix, thanks.

> All this code is very similar to __udpv4_gso_segment_list_csum. But
> the zero checksum handling in __udpv4_gso_segment_csum makes it just
> different enough that I also do not immediately see a straightforward
> way to avoid duplicating.

Also, the checksum field is in a different location in the udp header. I 
don't think avoiding duplication makes sense here.

- Felix