Hi,
On Sat, Oct 24, 2015, at 00:48, Eric Dumazet wrote:
> On Fri, 2015-10-23 at 15:13 +0200, Hannes Frederic Sowa wrote:
> > CHECKSUM_PARTIAL should only be used on plain vanilla IPv6 + UDP packets
> > in ip6_append_data. Some drivers don't correctly handle extension headers,
> > especially not ipv6 fragmentation which could result in broken checksums.
> >
> > 1) This patch improves the test for fragmentation and extension headers
> > in ip6_append_data, so we set the ip_summed mode as early as possible
> > to the correct value to compute the checksum during memory copy-in from
> > user space.
> >
> > 2) We always call skb_checksum_help on CHECKSUM_PARTIAL fragments in
> > ip6_fragment, because we don't know if the underlying hardware can deal
> > with ip6_fragments.
> >
> > Fixes: commit 32dce968dd987 ("ipv6: Allow for partial checksums on non-ufo
> > packets")
> > See-also: commit 72e843bb09d45 ("ipv6: ip6_fragment() should check
> > CHECKSUM_PARTIAL")
> > Cc: Eric Dumazet <eduma...@google.com>
> > Cc: Vlad Yasevich <vyasev...@gmail.com>
> > Cc: Benjamin Coddington <bcodd...@redhat.com>
> > Signed-off-by: Hannes Frederic Sowa <han...@stressinduktion.org>
> > ---
> > net/ipv6/ip6_output.c | 78
> > ++++++++++++++++++++++++---------------------------
> > 1 file changed, 37 insertions(+), 41 deletions(-)
> >
> > diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
> > index 8dddb45..26d2911 100644
> > --- a/net/ipv6/ip6_output.c
> > +++ b/net/ipv6/ip6_output.c
> > @@ -593,6 +593,10 @@ int ip6_fragment(struct sock *sk, struct sk_buff *skb,
> > frag_id = ipv6_select_ident(net, &ipv6_hdr(skb)->daddr,
> > &ipv6_hdr(skb)->saddr);
> >
> > + if ((skb->ip_summed == CHECKSUM_PARTIAL) &&
> > + (err = skb_checksum_help(skb)))
> > + goto fail;
> > +
> > hroom = LL_RESERVED_SPACE(rt->dst.dev);
> > if (skb_has_frag_list(skb)) {
> > int first_len = skb_pagelen(skb);
> > @@ -721,10 +725,6 @@ slow_path_clean:
> > }
> >
> > slow_path:
> > - if ((skb->ip_summed == CHECKSUM_PARTIAL) &&
> > - skb_checksum_help(skb))
> > - goto fail;
> > -
>
>
> It looks like this patch could be split in two, to ease future bisection
> maybe ?
Sure, will split it up.
Thanks Eric!
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
