netdev - Re: [PATCH 1/1] net: tun: add XDP metadata support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <Z5wIZ2LAjz0wTWg5@mini-arch>
Date: Thu, 30 Jan 2025 15:16:55 -0800
From: Stanislav Fomichev <stfomichev@...il.com>
To: Marcus Wichelmann <marcus.wichelmann@...zner-cloud.de>
Cc: netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
	bpf@...r.kernel.org, willemdebruijn.kernel@...il.com,
	jasowang@...hat.com, andrew+netdev@...n.ch, davem@...emloft.net,
	edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com,
	ast@...nel.org, daniel@...earbox.net, hawk@...nel.org,
	john.fastabend@...il.com
Subject: Re: [PATCH 1/1] net: tun: add XDP metadata support

On 01/30, Marcus Wichelmann wrote:
> Enable the support for bpf_xdp_adjust_meta for XDP buffers initialized
> by the tun driver. This is useful to pass metadata from an XDP program
> that's attached to a tap device to following XDP/TC programs.
> 
> When used together with vhost_net, the batched XDP buffers were already
> initialized with metadata support by the vhost_net driver, but the
> metadata was not yet passed to the skb on XDP_PASS. So this also adds
> the required skb_metadata_set calls.

Can you expand more on what kind of metadata is present with vhost_net
and who fills it in? Is it virtio header stuff? I wonder how you
want to consume it..

Can you also add a selftest to use this new functionality?

> Signed-off-by: Marcus Wichelmann <marcus.wichelmann@...zner-cloud.de>
> ---
>  drivers/net/tun.c | 23 ++++++++++++++++++-----
>  1 file changed, 18 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/net/tun.c b/drivers/net/tun.c
> index e816aaba8..d3cfea40a 100644
> --- a/drivers/net/tun.c
> +++ b/drivers/net/tun.c
> @@ -1600,7 +1600,8 @@ static bool tun_can_build_skb(struct tun_struct *tun, struct tun_file *tfile,
>  
>  static struct sk_buff *__tun_build_skb(struct tun_file *tfile,
>  				       struct page_frag *alloc_frag, char *buf,
> -				       int buflen, int len, int pad)
> +				       int buflen, int len, int pad,
> +				       int metasize)
>  {
>  	struct sk_buff *skb = build_skb(buf, buflen);
>  
> @@ -1609,6 +1610,8 @@ static struct sk_buff *__tun_build_skb(struct tun_file *tfile,
>  
>  	skb_reserve(skb, pad);
>  	skb_put(skb, len);
> +	if (metasize)
> +		skb_metadata_set(skb, metasize);
>  	skb_set_owner_w(skb, tfile->socket.sk);
>  
>  	get_page(alloc_frag->page);
> @@ -1668,6 +1671,7 @@ static struct sk_buff *tun_build_skb(struct tun_struct *tun,
>  	char *buf;
>  	size_t copied;
>  	int pad = TUN_RX_PAD;
> +	int metasize = 0;
>  	int err = 0;
>  
>  	rcu_read_lock();
> @@ -1695,7 +1699,7 @@ static struct sk_buff *tun_build_skb(struct tun_struct *tun,
>  	if (hdr->gso_type || !xdp_prog) {
>  		*skb_xdp = 1;
>  		return __tun_build_skb(tfile, alloc_frag, buf, buflen, len,
> -				       pad);
> +				       pad, metasize);
>  	}
>  
>  	*skb_xdp = 0;
> @@ -1709,7 +1713,7 @@ static struct sk_buff *tun_build_skb(struct tun_struct *tun,
>  		u32 act;
>  
>  		xdp_init_buff(&xdp, buflen, &tfile->xdp_rxq);
> -		xdp_prepare_buff(&xdp, buf, pad, len, false);
> +		xdp_prepare_buff(&xdp, buf, pad, len, true);
>  
>  		act = bpf_prog_run_xdp(xdp_prog, &xdp);
>  		if (act == XDP_REDIRECT || act == XDP_TX) {
> @@ -1730,12 +1734,16 @@ static struct sk_buff *tun_build_skb(struct tun_struct *tun,
>  
>  		pad = xdp.data - xdp.data_hard_start;
>  		len = xdp.data_end - xdp.data;
> +
> +		metasize = xdp.data - xdp.data_meta;

[..]

> +		metasize = metasize > 0 ? metasize : 0;

Why is this part needed?