RE: [PATCH v2] net/packet: support mergeable feautre of virtio

[Willem de Bruijn <willemdebruijn.kernel@gmail.com>]

沈安琪(凛玥) wrote:
> From: Jianfeng Tan <henry.tjf@antgroup.com>
> 
> Packet sockets, like tap, can be used as the backend for kernel vhost.
> In packet sockets, virtio net header size is currently hardcoded to be
> the size of struct virtio_net_hdr, which is 10 bytes; however, it is not
> always the case: some virtio features, such as mrg_rxbuf, need virtio
> net header to be 12-byte long.
> 
> Mergeable buffers, as a virtio feature, is worthy to support: packets
> that larger than one-mbuf size will be dropped in vhost worker's
> handle_rx if mrg_rxbuf feature is not used, but large packets
> cannot be avoided and increasing mbuf's size is not economical.
> 
> With this virtio feature enabled, packet sockets with hardcoded 10-byte
> virtio net header will parse mac head incorrectly in packet_snd by taking
> the last two bytes of virtio net header as part of mac header as well.
> This incorrect mac header parsing will cause packet be dropped due to
> invalid ether head checking in later under-layer device packet receiving.
> 
> By adding extra field vnet_hdr_sz with utilizing holes in struct
> packet_sock to record current using virtio net header size and supporting
> extra sockopt PACKET_VNET_HDR_SZ to set specified vnet_hdr_sz, packet
> sockets can know the exact length of virtio net header that virtio user
> gives.
> In packet_snd, tpacket_snd and packet_recvmsg, instead of using hardcode
> virtio net header size, it can get the exact vnet_hdr_sz from corresponding
> packet_sock, and parse mac header correctly based on this information to
> avoid the packets being mistakenly dropped.
> 
> Signed-off-by: Jianfeng Tan <henry.tjf@antgroup.com>
> Co-developed-by: Anqi Shen <amy.saq@antgroup.com>
> Signed-off-by: Anqi Shen <amy.saq@antgroup.com>

net-next is closed

> @@ -2311,7 +2312,7 @@ static int tpacket_rcv(struct sk_buff *skb, struct net_device *dev,
>  				       (maclen < 16 ? 16 : maclen)) +
>  				       po->tp_reserve;
>  		if (po->has_vnet_hdr) {
> -			netoff += sizeof(struct virtio_net_hdr);
> +			netoff += po->vnet_hdr_sz;
>  			do_vnet = true;
>  		}
>  		macoff = netoff - maclen;
> @@ -2552,16 +2553,23 @@ static int __packet_snd_vnet_parse(struct virtio_net_hdr *vnet_hdr, size_t len)
>  }
>  
>  static int packet_snd_vnet_parse(struct msghdr *msg, size_t *len,
> -				 struct virtio_net_hdr *vnet_hdr)
> +				 struct virtio_net_hdr *vnet_hdr, int vnet_hdr_sz)
>  {
> -	if (*len < sizeof(*vnet_hdr))
> +	int ret;
> +
> +	if (*len < vnet_hdr_sz)
>  		return -EINVAL;
> -	*len -= sizeof(*vnet_hdr);
> +	*len -= vnet_hdr_sz;
>  
>  	if (!copy_from_iter_full(vnet_hdr, sizeof(*vnet_hdr), &msg->msg_iter))
>  		return -EFAULT;
>  
> -	return __packet_snd_vnet_parse(vnet_hdr, *len);
> +	ret = __packet_snd_vnet_parse(vnet_hdr, *len);
> +
> +	/* move iter to point to the start of mac header */
> +	if (ret == 0)
> +		iov_iter_advance(&msg->msg_iter, vnet_hdr_sz - sizeof(struct virtio_net_hdr));
> +	return ret;

Let's make the error path the exception

        if (ret)
                return ret;

And maybe avoid calling iov_iter_advance if vnet_hdr_sz == sizeof(*vnet_hdr)

>  	case PACKET_VNET_HDR:
> +	case PACKET_VNET_HDR_SZ:
>  	{
>  		int val;
> +		int hdr_len = 0;
>  
>  		if (sock->type != SOCK_RAW)
>  			return -EINVAL;
> @@ -3931,11 +3945,23 @@ static void packet_flush_mclist(struct sock *sk)
>  		if (copy_from_sockptr(&val, optval, sizeof(val)))
>  			return -EFAULT;
>  
> +		if (optname == PACKET_VNET_HDR_SZ) {
> +			if (val != sizeof(struct virtio_net_hdr) &&
> +			    val != sizeof(struct virtio_net_hdr_mrg_rxbuf))
> +				return -EINVAL;
> +			hdr_len = val;
> +		}
> +

    } else {
            hdr_len = sizeof(struct virtio_net_hdr);
    }

>  		lock_sock(sk);
>  		if (po->rx_ring.pg_vec || po->tx_ring.pg_vec) {
>  			ret = -EBUSY;
>  		} else {
> -			po->has_vnet_hdr = !!val;
> +			if (optname == PACKET_VNET_HDR) {
> +				po->has_vnet_hdr = !!val;
> +				if (po->has_vnet_hdr)
> +					hdr_len = sizeof(struct virtio_net_hdr);
> +			}
> +			po->vnet_hdr_sz = hdr_len;

then this is not needed
>  			ret = 0;
>  		}
>  		release_sock(sk);
> @@ -4070,6 +4096,9 @@ static int packet_getsockopt(struct socket *sock, int level, int optname,
>  	case PACKET_VNET_HDR:
>  		val = po->has_vnet_hdr;
>  		break;
> +	case PACKET_VNET_HDR_SZ:
> +		val = po->vnet_hdr_sz;
> +		break;
>  	case PACKET_VERSION:
>  		val = po->tp_version;
>  		break;
> diff --git a/net/packet/internal.h b/net/packet/internal.h
> index 48af35b..e27b47d 100644
> --- a/net/packet/internal.h
> +++ b/net/packet/internal.h
> @@ -121,7 +121,8 @@ struct packet_sock {
>  				origdev:1,
>  				has_vnet_hdr:1,
>  				tp_loss:1,
> -				tp_tx_has_off:1;
> +				tp_tx_has_off:1,
> +				vnet_hdr_sz:8;	/* vnet header size should use */

has_vnet_hdr is no longer needed when adding vnet_hdr_sz. removing that simplifies the code

drop the comment. That is quite self explanatory from the variable name.

>  	int			pressure;
>  	int			ifindex;	/* bound device		*/
>  	__be16			num;
> -- 
> 1.8.3.1
>