Merge tag 'net-5.13-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Networking fixes, including fixes from bpf, wireless, netfilter and
  wireguard trees.

  The bpf vs lockdown+audit fix is the most notable.

  Things haven't slowed down just yet, both in terms of regressions in
  current release and largish fixes for older code, but we usually see a
  slowdown only after -rc5.

  Current release - regressions:

   - virtio-net: fix page faults and crashes when XDP is enabled

   - mlx5e: fix HW timestamping with CQE compression, and make sure they
     are only allowed to coexist with capable devices

   - stmmac:
      - fix kernel panic due to NULL pointer dereference of
        mdio_bus_data
      - fix double clk unprepare when no PHY device is connected

  Current release - new code bugs:

   - mt76: a few fixes for the recent MT7921 devices and runtime power
     management

  Previous releases - regressions:

   - ice:
      - track AF_XDP ZC enabled queues in bitmap to fix copy mode Tx
      - fix allowing VF to request more/less queues via virtchnl
      - correct supported and advertised autoneg by using PHY
        capabilities
      - allow all LLDP packets from PF to Tx

   - kbuild: quote OBJCOPY var to avoid a pahole call break the build

  Previous releases - always broken:

   - bpf, lockdown, audit: fix buggy SELinux lockdown permission checks

   - mt76: address the recent FragAttack vulnerabilities not covered by
     generic fixes

   - ipv6: fix KASAN: slab-out-of-bounds Read in
     fib6_nh_flush_exceptions

   - Bluetooth:
      - fix the erroneous flush_work() order, to avoid double free
      - use correct lock to prevent UAF of hdev object

   - nfc: fix NULL ptr dereference in llcp_sock_getname() after failed
     connect

   - ieee802154: multiple fixes to error checking and return values

   - igb: fix XDP with PTP enabled

   - intel: add correct exception tracing for XDP

   - tls: fix use-after-free when TLS offload device goes down and back
     up

   - ipvs: ignore IP_VS_SVC_F_HASHED flag when adding service

   - netfilter: nft_ct: skip expectations for confirmed conntrack

   - mptcp: fix falling back to TCP in presence of out of order packets
     early in connection lifetime

   - wireguard: switch from O(n) to a O(1) algorithm for maintaining
     peers, fixing stalls and a large memory leak in the process

  Misc:

   - devlink: correct VIRTUAL port to not have phys_port attributes

   - Bluetooth: fix VIRTIO_ID_BT assigned number

   - net: return the correct errno code ENOBUF -> ENOMEM

   - wireguard:
      - peer: allocate in kmem_cache saving 25% on peer memory
      - do not use -O3"

* tag 'net-5.13-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (91 commits)
  cxgb4: avoid link re-train during TC-MQPRIO configuration
  sch_htb: fix refcount leak in htb_parent_to_leaf_offload
  wireguard: allowedips: free empty intermediate nodes when removing single node
  wireguard: allowedips: allocate nodes in kmem_cache
  wireguard: allowedips: remove nodes in O(1)
  wireguard: allowedips: initialize list head in selftest
  wireguard: peer: allocate in kmem_cache
  wireguard: use synchronize_net rather than synchronize_rcu
  wireguard: do not use -O3
  wireguard: selftests: make sure rp_filter is disabled on vethc
  wireguard: selftests: remove old conntrack kconfig value
  virtchnl: Add missing padding to virtchnl_proto_hdrs
  ice: Allow all LLDP packets from PF to Tx
  ice: report supported and advertised autoneg using PHY capabilities
  ice: handle the VF VSI rebuild failure
  ice: Fix VFR issues for AVF drivers that expect ATQLEN cleared
  ice: Fix allowing VF to request more/less queues via virtchnl
  virtio-net: fix for skb_over_panic inside big mode
  ipv6: Fix KASAN: slab-out-of-bounds Read in fib6_nh_flush_exceptions
  fib: Return the correct errno code
  ...

1 files changed, 32 insertions, 23 deletions

diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c
index 038a9fd1af44..b2a042f825ff 100644
--- a/drivers/net/ethernet/intel/igb/igb_main.c
+++ b/drivers/net/ethernet/intel/igb/igb_main.c
@@ -8280,7 +8280,7 @@ static void igb_add_rx_frag(struct igb_ring *rx_ring,
 static struct sk_buff *igb_construct_skb(struct igb_ring *rx_ring,
 					 struct igb_rx_buffer *rx_buffer,
 					 struct xdp_buff *xdp,
-					 union e1000_adv_rx_desc *rx_desc)
+					 ktime_t timestamp)
 {
 #if (PAGE_SIZE < 8192)
 	unsigned int truesize = igb_rx_pg_size(rx_ring) / 2;
@@ -8300,12 +8300,8 @@ static struct sk_buff *igb_construct_skb(struct igb_ring *rx_ring,
 	if (unlikely(!skb))
 		return NULL;
 
-	if (unlikely(igb_test_staterr(rx_desc, E1000_RXDADV_STAT_TSIP))) {
-		if (!igb_ptp_rx_pktstamp(rx_ring->q_vector, xdp->data, skb)) {
-			xdp->data += IGB_TS_HDR_LEN;
-			size -= IGB_TS_HDR_LEN;
-		}
-	}
+	if (timestamp)
+		skb_hwtstamps(skb)->hwtstamp = timestamp;
 
 	/* Determine available headroom for copy */
 	headlen = size;
@@ -8336,7 +8332,7 @@ static struct sk_buff *igb_construct_skb(struct igb_ring *rx_ring,
 static struct sk_buff *igb_build_skb(struct igb_ring *rx_ring,
 				     struct igb_rx_buffer *rx_buffer,
 				     struct xdp_buff *xdp,
-				     union e1000_adv_rx_desc *rx_desc)
+				     ktime_t timestamp)
 {
 #if (PAGE_SIZE < 8192)
 	unsigned int truesize = igb_rx_pg_size(rx_ring) / 2;
@@ -8363,11 +8359,8 @@ static struct sk_buff *igb_build_skb(struct igb_ring *rx_ring,
 	if (metasize)
 		skb_metadata_set(skb, metasize);
 
-	/* pull timestamp out of packet data */
-	if (igb_test_staterr(rx_desc, E1000_RXDADV_STAT_TSIP)) {
-		if (!igb_ptp_rx_pktstamp(rx_ring->q_vector, skb->data, skb))
-			__skb_pull(skb, IGB_TS_HDR_LEN);
-	}
+	if (timestamp)
+		skb_hwtstamps(skb)->hwtstamp = timestamp;
 
 	/* update buffer offset */
 #if (PAGE_SIZE < 8192)
@@ -8401,18 +8394,20 @@ static struct sk_buff *igb_run_xdp(struct igb_adapter *adapter,
 		break;
 	case XDP_TX:
 		result = igb_xdp_xmit_back(adapter, xdp);
+		if (result == IGB_XDP_CONSUMED)
+			goto out_failure;
 		break;
 	case XDP_REDIRECT:
 		err = xdp_do_redirect(adapter->netdev, xdp, xdp_prog);
-		if (!err)
-			result = IGB_XDP_REDIR;
-		else
-			result = IGB_XDP_CONSUMED;
+		if (err)
+			goto out_failure;
+		result = IGB_XDP_REDIR;
 		break;
 	default:
 		bpf_warn_invalid_xdp_action(act);
 		fallthrough;
 	case XDP_ABORTED:
+out_failure:
 		trace_xdp_exception(rx_ring->netdev, xdp_prog, act);
 		fallthrough;
 	case XDP_DROP:
@@ -8682,7 +8677,10 @@ static int igb_clean_rx_irq(struct igb_q_vector *q_vector, const int budget)
 	while (likely(total_packets < budget)) {
 		union e1000_adv_rx_desc *rx_desc;
 		struct igb_rx_buffer *rx_buffer;
+		ktime_t timestamp = 0;
+		int pkt_offset = 0;
 		unsigned int size;
+		void *pktbuf;
 
 		/* return some buffers to hardware, one at a time is too slow */
 		if (cleaned_count >= IGB_RX_BUFFER_WRITE) {
@@ -8702,14 +8700,24 @@ static int igb_clean_rx_irq(struct igb_q_vector *q_vector, const int budget)
 		dma_rmb();
 
 		rx_buffer = igb_get_rx_buffer(rx_ring, size, &rx_buf_pgcnt);
+		pktbuf = page_address(rx_buffer->page) + rx_buffer->page_offset;
+
+		/* pull rx packet timestamp if available and valid */
+		if (igb_test_staterr(rx_desc, E1000_RXDADV_STAT_TSIP)) {
+			int ts_hdr_len;
+
+			ts_hdr_len = igb_ptp_rx_pktstamp(rx_ring->q_vector,
+							 pktbuf, &timestamp);
+
+			pkt_offset += ts_hdr_len;
+			size -= ts_hdr_len;
+		}
 
 		/* retrieve a buffer from the ring */
 		if (!skb) {
-			unsigned int offset = igb_rx_offset(rx_ring);
-			unsigned char *hard_start;
+			unsigned char *hard_start = pktbuf - igb_rx_offset(rx_ring);
+			unsigned int offset = pkt_offset + igb_rx_offset(rx_ring);
 
-			hard_start = page_address(rx_buffer->page) +
-				     rx_buffer->page_offset - offset;
 			xdp_prepare_buff(&xdp, hard_start, offset, size, true);
 #if (PAGE_SIZE > 4096)
 			/* At larger PAGE_SIZE, frame_sz depend on len size */
@@ -8732,10 +8740,11 @@ static int igb_clean_rx_irq(struct igb_q_vector *q_vector, const int budget)
 		} else if (skb)
 			igb_add_rx_frag(rx_ring, rx_buffer, skb, size);
 		else if (ring_uses_build_skb(rx_ring))
-			skb = igb_build_skb(rx_ring, rx_buffer, &xdp, rx_desc);
+			skb = igb_build_skb(rx_ring, rx_buffer, &xdp,
+					    timestamp);
 		else
 			skb = igb_construct_skb(rx_ring, rx_buffer,
-						&xdp, rx_desc);
+						&xdp, timestamp);
 
 		/* exit if we failed to retrieve a buffer */
 		if (!skb) {