Re: [PATCH bpf v4 1/2] bpf: Run generic devmap egress prog on private skb

["Toke Høiland-Jørgensen" <toke@redhat.com>]

Sun Jian <sun.jian.kdev@gmail.com> writes:

> Generic XDP devmap multi redirect uses skb_clone() for intermediate
> destinations and sends the last destination with the original skb. This
> can leave multiple destinations sharing the same packet data.
>
> This becomes visible after generic devmap egress-program support was
> added: a devmap egress program may mutate packet data, and another
> destination sharing the same data can observe that mutation.
>
> Native XDP broadcast redirect does not have this issue because
> xdpf_clone() copies the frame data for each destination. Generic XDP
> should provide the same per-destination isolation before running a
> devmap egress program.
>
> Fix this by making cloned skbs private before running the generic devmap
> egress program. Use skb_copy() instead of skb_unshare() so allocation
> failure does not consume the skb and the existing caller error paths keep
> their ownership semantics.
>
> Fixes: 2ea5eabaf04a ("bpf: devmap: Implement devmap prog execution for generic XDP")
> Suggested-by: Jiayuan Chen <jiayuan.chen@linux.dev>
> Suggested-by: Jakub Kicinski <kuba@kernel.org>
> Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
> Signed-off-by: Sun Jian <sun.jian.kdev@gmail.com>
> ---
>  kernel/bpf/devmap.c | 41 +++++++++++++++++++++++++++++++----------
>  1 file changed, 31 insertions(+), 10 deletions(-)
>
> diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c
> index cc0a43ebab6b..a3d6c60dbddb 100644
> --- a/kernel/bpf/devmap.c
> +++ b/kernel/bpf/devmap.c
> @@ -512,35 +512,52 @@ static inline int __xdp_enqueue(struct net_device *dev, struct xdp_frame *xdpf,
>  	return 0;
>  }
>  
> -static u32 dev_map_bpf_prog_run_skb(struct sk_buff *skb, struct bpf_dtab_netdev *dst)
> +static int dev_map_bpf_prog_run_skb(struct sk_buff **pskb,
> +				    struct bpf_dtab_netdev *dst,
> +				    u32 *act)
>  {
> +	struct sk_buff *skb = *pskb;
>  	struct xdp_txq_info txq = { .dev = dst->dev };
>  	struct xdp_buff xdp;
> -	u32 act;
>  
> -	if (!dst->xdp_prog)
> -		return XDP_PASS;
> +	if (!dst->xdp_prog) {
> +		*act = XDP_PASS;
> +		return 0;
> +	}
> +
> +	if (skb_cloned(skb)) {
> +		struct sk_buff *nskb;
> +
> +		nskb = skb_copy(skb, GFP_ATOMIC);
> +		if (!nskb)
> +			return -ENOMEM;
> +
> +		nskb->mac_len = skb->mac_len;
> +		consume_skb(skb);
> +		skb = nskb;
> +		*pskb = nskb;
> +	}

So with all this pointer soup and back and forth, the version you had in
v2 (with the check and skb_copy() in dev_map_generic_redirect()) was
much cleaner :/

-Toke