diff options
| author | Alexander Duyck <aduyck@mirantis.com> | 2016-02-25 01:46:21 +0100 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2016-02-26 20:23:35 +0100 |
| commit | 224638766235ba82c53b4216e4dabc510701fbf2 (patch) | |
| tree | 593d446859b883db6cfb9f8926a28efc141e0a99 /net/ipv4/gre_offload.c | |
| parent | Merge branch 'vrf-saddr-selection' (diff) | |
| download | linux-224638766235ba82c53b4216e4dabc510701fbf2.tar.xz linux-224638766235ba82c53b4216e4dabc510701fbf2.zip | |
GSO: Provide software checksum of tunneled UDP fragmentation offload
On reviewing the code I realized that GRE and UDP tunnels could cause a
kernel panic if we used GSO to segment a large UDP frame that was sent
through the tunnel with an outer checksum and hardware offloads were not
available.
In order to correct this we need to update the feature flags that are
passed to the skb_segment function so that in the event of UDP
fragmentation being requested for the inner header the segmentation
function will correctly generate the checksum for the payload if we cannot
segment the outer header.
Signed-off-by: Alexander Duyck <aduyck@mirantis.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net/ipv4/gre_offload.c')
| -rw-r--r-- | net/ipv4/gre_offload.c | 21 |
1 files changed, 18 insertions, 3 deletions
diff --git a/net/ipv4/gre_offload.c b/net/ipv4/gre_offload.c index 003b0ebbcfdd..47f4c544c916 100644 --- a/net/ipv4/gre_offload.c +++ b/net/ipv4/gre_offload.c @@ -24,7 +24,7 @@ static struct sk_buff *gre_gso_segment(struct sk_buff *skb, __be16 protocol = skb->protocol; u16 mac_len = skb->mac_len; int gre_offset, outer_hlen; - bool need_csum; + bool need_csum, ufo; if (unlikely(skb_shinfo(skb)->gso_type & ~(SKB_GSO_TCPV4 | @@ -58,8 +58,20 @@ static struct sk_buff *gre_gso_segment(struct sk_buff *skb, need_csum = !!(skb_shinfo(skb)->gso_type & SKB_GSO_GRE_CSUM); skb->encap_hdr_csum = need_csum; + ufo = !!(skb_shinfo(skb)->gso_type & SKB_GSO_UDP); + features &= skb->dev->hw_enc_features; + /* The only checksum offload we care about from here on out is the + * outer one so strip the existing checksum feature flags based + * on the fact that we will be computing our checksum in software. + */ + if (ufo) { + features &= ~NETIF_F_CSUM_MASK; + if (!need_csum) + features |= NETIF_F_HW_CSUM; + } + /* segment inner packet. */ segs = skb_mac_gso_segment(skb, features); if (IS_ERR_OR_NULL(segs)) { @@ -75,8 +87,11 @@ static struct sk_buff *gre_gso_segment(struct sk_buff *skb, struct gre_base_hdr *greh; __be32 *pcsum; - skb_reset_inner_headers(skb); - skb->encapsulation = 1; + /* Set up inner headers if we are offloading inner checksum */ + if (skb->ip_summed == CHECKSUM_PARTIAL) { + skb_reset_inner_headers(skb); + skb->encapsulation = 1; + } skb->mac_len = mac_len; skb->protocol = protocol; |