Re: [PATCH v3 1/2] net: create skb_gso_validate_mac_len()

[Eric Dumazet <eric.dumazet@gmail.com>]

On Mon, 2018-01-29 at 17:46 -0800, Eric Dumazet wrote:
> On Tue, 2018-01-30 at 12:14 +1100, Daniel Axtens wrote:
> > If you take a GSO skb, and split it into packets, will the MAC
> > length (L2 + L3 + L4 headers + payload) of those packets be small
> > enough to fit within a given length?
> > 
> > Move skb_gso_mac_seglen() to skbuff.h with other related functions
> > like skb_gso_network_seglen() so we can use it, and then create
> > skb_gso_validate_mac_len to do the full calculation.
> > 
> > Signed-off-by: Daniel Axtens <dja@axtens.net>
> > ---
> >  include/linux/skbuff.h | 16 +++++++++++++
> >  net/core/skbuff.c      | 65 +++++++++++++++++++++++++++++++++++++++-----------
> >  net/sched/sch_tbf.c    | 10 --------
> >  3 files changed, 67 insertions(+), 24 deletions(-)
> > 
> > diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
> > index b8e0da6c27d6..242d6773c7c2 100644
> > --- a/include/linux/skbuff.h
> > +++ b/include/linux/skbuff.h
> > @@ -3287,6 +3287,7 @@ int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen);
> >  void skb_scrub_packet(struct sk_buff *skb, bool xnet);
> >  unsigned int skb_gso_transport_seglen(const struct sk_buff *skb);
> >  bool skb_gso_validate_mtu(const struct sk_buff *skb, unsigned int mtu);
> > +bool skb_gso_validate_mac_len(const struct sk_buff *skb, unsigned int len);
> >  struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features);
> >  struct sk_buff *skb_vlan_untag(struct sk_buff *skb);
> >  int skb_ensure_writable(struct sk_buff *skb, int write_len);
> > @@ -4120,6 +4121,21 @@ static inline unsigned int skb_gso_network_seglen(const struct sk_buff *skb)
> >  	return hdr_len + skb_gso_transport_seglen(skb);
> >  }
> >  
> > +/**
> > + * skb_gso_mac_seglen - Return length of individual segments of a gso packet
> > + *
> > + * @skb: GSO skb
> > + *
> > + * skb_gso_mac_seglen is used to determine the real size of the
> > + * individual segments, including MAC/L2, Layer3 (IP, IPv6) and L4
> > + * headers (TCP/UDP).
> > + */
> > +static inline unsigned int skb_gso_mac_seglen(const struct sk_buff *skb)
> > +{
> > +	unsigned int hdr_len = skb_transport_header(skb) - skb_mac_header(skb);
> > +	return hdr_len + skb_gso_transport_seglen(skb);
> > +}
> 
> skb_gso_transport_seglen(skb) is quite expensive (out of line)
> 
> It is unfortunate bnx2x seems to support 9600 MTU (
> ETH_MAX_JUMBO_PACKET_SIZE ), because 100 bytes of headers can be too
> small in some cases.
> 
> Presumably we could avoid calling the function for standard MTU <= 9000

Also are we sure about these bnx2x crashes being limited to TSO ?

Maybe it will crash the same after GSO has segmented the packet and we
provide a big (like 10,000 bytes) packet ? I do not believe upper stack
will prevent this.