Merge tag 'net-5.18-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Including fixes from bluetooth, bpf and netfilter.

  Current release - new code bugs:

   - bridge: switchdev: check br_vlan_group() return value

   - use this_cpu_inc() to increment net->core_stats, fix preempt-rt

  Previous releases - regressions:

   - eth: stmmac: fix write to sgmii_adapter_base

  Previous releases - always broken:

   - netfilter: nf_conntrack_tcp: re-init for syn packets only,
     resolving issues with TCP fastopen

   - tcp: md5: fix incorrect tcp_header_len for incoming connections

   - tcp: fix F-RTO may not work correctly when receiving DSACK

   - tcp: ensure use of most recently sent skb when filling rate samples

   - tcp: fix potential xmit stalls caused by TCP_NOTSENT_LOWAT

   - virtio_net: fix wrong buf address calculation when using xdp

   - xsk: fix forwarding when combining copy mode with busy poll

   - xsk: fix possible crash when multiple sockets are created

   - bpf: lwt: fix crash when using bpf_skb_set_tunnel_key() from
     bpf_xmit lwt hook

   - sctp: null-check asoc strreset_chunk in sctp_generate_reconf_event

   - wireguard: device: check for metadata_dst with skb_valid_dst()

   - netfilter: update ip6_route_me_harder to consider L3 domain

   - gre: make o_seqno start from 0 in native mode

   - gre: switch o_seqno to atomic to prevent races in collect_md mode

  Misc:

   - add Eric Dumazet to networking maintainers

   - dt: dsa: realtek: remove realtek,rtl8367s string

   - netfilter: flowtable: Remove the empty file"

* tag 'net-5.18-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (65 commits)
  tcp: fix F-RTO may not work correctly when receiving DSACK
  Revert "ibmvnic: Add ethtool private flag for driver-defined queue limits"
  net: enetc: allow tc-etf offload even with NETIF_F_CSUM_MASK
  ixgbe: ensure IPsec VF<->PF compatibility
  MAINTAINERS: Update BNXT entry with firmware files
  netfilter: nft_socket: only do sk lookups when indev is available
  net: fec: add missing of_node_put() in fec_enet_init_stop_mode()
  bnx2x: fix napi API usage sequence
  tls: Skip tls_append_frag on zero copy size
  Add Eric Dumazet to networking maintainers
  netfilter: conntrack: fix udp offload timeout sysctl
  netfilter: nf_conntrack_tcp: re-init for syn packets only
  net: dsa: lantiq_gswip: Don't set GSWIP_MII_CFG_RMII_CLK
  net: Use this_cpu_inc() to increment net->core_stats
  Bluetooth: hci_sync: Cleanup hci_conn if it cannot be aborted
  Bluetooth: hci_event: Fix creating hci_conn object on error status
  Bluetooth: hci_event: Fix checking for invalid handle on error status
  ice: fix use-after-free when deinitializing mailbox snapshot
  ice: wait 5 s for EMP reset after firmware flash
  ice: Protect vf_state check by cfg_lock in ice_vc_process_vf_msg()
  ...

1 files changed, 19 insertions, 1 deletions

diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 87838cbe38cf..cbba9d2e8f32 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -1005,6 +1005,24 @@ static struct sk_buff *receive_mergeable(struct net_device *dev,
 			 * xdp.data_meta were adjusted
 			 */
 			len = xdp.data_end - xdp.data + vi->hdr_len + metasize;
+
+			/* recalculate headroom if xdp.data or xdp_data_meta
+			 * were adjusted, note that offset should always point
+			 * to the start of the reserved bytes for virtio_net
+			 * header which are followed by xdp.data, that means
+			 * that offset is equal to the headroom (when buf is
+			 * starting at the beginning of the page, otherwise
+			 * there is a base offset inside the page) but it's used
+			 * with a different starting point (buf start) than
+			 * xdp.data (buf start + vnet hdr size). If xdp.data or
+			 * data_meta were adjusted by the xdp prog then the
+			 * headroom size has changed and so has the offset, we
+			 * can use data_hard_start, which points at buf start +
+			 * vnet hdr size, to calculate the new headroom and use
+			 * it later to compute buf start in page_to_skb()
+			 */
+			headroom = xdp.data - xdp.data_hard_start - metasize;
+
 			/* We can only create skb based on xdp_page. */
 			if (unlikely(xdp_page != page)) {
 				rcu_read_unlock();
@@ -1012,7 +1030,7 @@ static struct sk_buff *receive_mergeable(struct net_device *dev,
 				head_skb = page_to_skb(vi, rq, xdp_page, offset,
 						       len, PAGE_SIZE, false,
 						       metasize,
-						       VIRTIO_XDP_HEADROOM);
+						       headroom);
 				return head_skb;
 			}
 			break;