On Fri, 2014-09-19 at 14:38 +0800, Jason Wang wrote:
> Commit ce93718fb7cdbc064c3000ff59e4d3200bdfa744 ("net: Don't keep
> around original SKB when we software segment GSO frames") frees the
> original skb after software GSO even for dodgy gso skbs. This breaks
> the stream throughput from untrusted sources, since only header
> checking was done during software GSO instead of a true
> segmentation. This patch fixes this by freeing the original gso skb
> only when it was really segmented by software.
>
> Fixes ce93718fb7cdbc064c3000ff59e4d3200bdfa744 ("net: Don't keep
> around original SKB when we software segment GSO frames.")
>
> CC: David S. Miller <[email protected]>
> Signed-off-by: Jason Wang <[email protected]>
> ---
> net/core/dev.c | 6 ++++--
> 1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/net/core/dev.c b/net/core/dev.c
> index e916ba8..b7a0e1d 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -2694,10 +2694,12 @@ struct sk_buff *validate_xmit_skb(struct sk_buff
> *skb, struct net_device *dev)
> struct sk_buff *segs;
>
> segs = skb_gso_segment(skb, features);
> - kfree_skb(skb);
> if (IS_ERR(segs))
>
> - skb = segs;
> + else if (segs) {
> + kfree_skb(skb);
> + skb = segs;
> + }
> } else {
> if (skb_needs_linearize(skb, features) &&
> __skb_linearize(skb))
Good catch !
While we are at it, could you use consume_skb() instead of kfree_skb(),
and add missing {} (CodingStyle) ?
if (IS_ERR(segs)) {
segs = NULL;
} else if (segs) {
consume_skb(skb);
skb = segs;
}
Thanks !
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
