Merge tag 'net-6.8-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net

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

  We have one outstanding issue with the stmmac driver, which may be a
  LOCKDEP false positive, not a blocker.

  Current release - regressions:

   - netfilter: nf_tables: re-allow NFPROTO_INET in
     nft_(match/target)_validate()

   - eth: ionic: fix error handling in PCI reset code

  Current release - new code bugs:

   - eth: stmmac: complete meta data only when enabled, fix null-deref

   - kunit: fix again checksum tests on big endian CPUs

  Previous releases - regressions:

   - veth: try harder when allocating queue memory

   - Bluetooth:
      - hci_bcm4377: do not mark valid bd_addr as invalid
      - hci_event: fix handling of HCI_EV_IO_CAPA_REQUEST

  Previous releases - always broken:

   - info leak in __skb_datagram_iter() on netlink socket

   - mptcp:
      - map v4 address to v6 when destroying subflow
      - fix potential wake-up event loss due to sndbuf auto-tuning
      - fix double-free on socket dismantle

   - wifi: nl80211: reject iftype change with mesh ID change

   - fix small out-of-bound read when validating netlink be16/32 types

   - rtnetlink: fix error logic of IFLA_BRIDGE_FLAGS writing back

   - ipv6: fix potential "struct net" ref-leak in inet6_rtm_getaddr()

   - ip_tunnel: prevent perpetual headroom growth with huge number of
     tunnels on top of each other

   - mctp: fix skb leaks on error paths of mctp_local_output()

   - eth: ice: fixes for DPLL state reporting

   - dpll: rely on rcu for netdev_dpll_pin() to prevent UaF

   - eth: dpaa: accept phy-interface-type = '10gbase-r' in the device
     tree"

* tag 'net-6.8-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (73 commits)
  dpll: fix build failure due to rcu_dereference_check() on unknown type
  kunit: Fix again checksum tests on big endian CPUs
  tls: fix use-after-free on failed backlog decryption
  tls: separate no-async decryption request handling from async
  tls: fix peeking with sync+async decryption
  tls: decrement decrypt_pending if no async completion will be called
  gtp: fix use-after-free and null-ptr-deref in gtp_newlink()
  net: hsr: Use correct offset for HSR TLV values in supervisory HSR frames
  igb: extend PTP timestamp adjustments to i211
  rtnetlink: fix error logic of IFLA_BRIDGE_FLAGS writing back
  tools: ynl: fix handling of multiple mcast groups
  selftests: netfilter: add bridge conntrack + multicast test case
  netfilter: bridge: confirm multicast packets before passing them up the stack
  netfilter: nf_tables: allow NFPROTO_INET in nft_(match/target)_validate()
  Bluetooth: qca: Fix triggering coredump implementation
  Bluetooth: hci_qca: Set BDA quirk bit if fwnode exists in DT
  Bluetooth: qca: Fix wrong event type for patch config command
  Bluetooth: Enforce validation on max value of connection interval
  Bluetooth: hci_event: Fix handling of HCI_EV_IO_CAPA_REQUEST
  Bluetooth: mgmt: Fix limited discoverable off timeout
  ...

2 files changed, 126 insertions, 0 deletions

diff --git a/net/bridge/br_netfilter_hooks.c b/net/bridge/br_netfilter_hooks.c
index ed1720890757..35e10c5a766d 100644
--- a/net/bridge/br_netfilter_hooks.c
+++ b/net/bridge/br_netfilter_hooks.c
@@ -43,6 +43,10 @@
 #include <linux/sysctl.h>
 #endif
 
+#if IS_ENABLED(CONFIG_NF_CONNTRACK)
+#include <net/netfilter/nf_conntrack_core.h>
+#endif
+
 static unsigned int brnf_net_id __read_mostly;
 
 struct brnf_net {
@@ -553,6 +557,90 @@ static unsigned int br_nf_pre_routing(void *priv,
 	return NF_STOLEN;
 }
 
+#if IS_ENABLED(CONFIG_NF_CONNTRACK)
+/* conntracks' nf_confirm logic cannot handle cloned skbs referencing
+ * the same nf_conn entry, which will happen for multicast (broadcast)
+ * Frames on bridges.
+ *
+ * Example:
+ *      macvlan0
+ *      br0
+ *  ethX  ethY
+ *
+ * ethX (or Y) receives multicast or broadcast packet containing
+ * an IP packet, not yet in conntrack table.
+ *
+ * 1. skb passes through bridge and fake-ip (br_netfilter)Prerouting.
+ *    -> skb->_nfct now references a unconfirmed entry
+ * 2. skb is broad/mcast packet. bridge now passes clones out on each bridge
+ *    interface.
+ * 3. skb gets passed up the stack.
+ * 4. In macvlan case, macvlan driver retains clone(s) of the mcast skb
+ *    and schedules a work queue to send them out on the lower devices.
+ *
+ *    The clone skb->_nfct is not a copy, it is the same entry as the
+ *    original skb.  The macvlan rx handler then returns RX_HANDLER_PASS.
+ * 5. Normal conntrack hooks (in NF_INET_LOCAL_IN) confirm the orig skb.
+ *
+ * The Macvlan broadcast worker and normal confirm path will race.
+ *
+ * This race will not happen if step 2 already confirmed a clone. In that
+ * case later steps perform skb_clone() with skb->_nfct already confirmed (in
+ * hash table).  This works fine.
+ *
+ * But such confirmation won't happen when eb/ip/nftables rules dropped the
+ * packets before they reached the nf_confirm step in postrouting.
+ *
+ * Work around this problem by explicit confirmation of the entry at
+ * LOCAL_IN time, before upper layer has a chance to clone the unconfirmed
+ * entry.
+ *
+ */
+static unsigned int br_nf_local_in(void *priv,
+				   struct sk_buff *skb,
+				   const struct nf_hook_state *state)
+{
+	struct nf_conntrack *nfct = skb_nfct(skb);
+	const struct nf_ct_hook *ct_hook;
+	struct nf_conn *ct;
+	int ret;
+
+	if (!nfct || skb->pkt_type == PACKET_HOST)
+		return NF_ACCEPT;
+
+	ct = container_of(nfct, struct nf_conn, ct_general);
+	if (likely(nf_ct_is_confirmed(ct)))
+		return NF_ACCEPT;
+
+	WARN_ON_ONCE(skb_shared(skb));
+	WARN_ON_ONCE(refcount_read(&nfct->use) != 1);
+
+	/* We can't call nf_confirm here, it would create a dependency
+	 * on nf_conntrack module.
+	 */
+	ct_hook = rcu_dereference(nf_ct_hook);
+	if (!ct_hook) {
+		skb->_nfct = 0ul;
+		nf_conntrack_put(nfct);
+		return NF_ACCEPT;
+	}
+
+	nf_bridge_pull_encap_header(skb);
+	ret = ct_hook->confirm(skb);
+	switch (ret & NF_VERDICT_MASK) {
+	case NF_STOLEN:
+		return NF_STOLEN;
+	default:
+		nf_bridge_push_encap_header(skb);
+		break;
+	}
+
+	ct = container_of(nfct, struct nf_conn, ct_general);
+	WARN_ON_ONCE(!nf_ct_is_confirmed(ct));
+
+	return ret;
+}
+#endif
 
 /* PF_BRIDGE/FORWARD *************************************************/
 static int br_nf_forward_finish(struct net *net, struct sock *sk, struct sk_buff *skb)
@@ -964,6 +1052,14 @@ static const struct nf_hook_ops br_nf_ops[] = {
 		.hooknum = NF_BR_PRE_ROUTING,
 		.priority = NF_BR_PRI_BRNF,
 	},
+#if IS_ENABLED(CONFIG_NF_CONNTRACK)
+	{
+		.hook = br_nf_local_in,
+		.pf = NFPROTO_BRIDGE,
+		.hooknum = NF_BR_LOCAL_IN,
+		.priority = NF_BR_PRI_LAST,
+	},
+#endif
 	{
 		.hook = br_nf_forward,
 		.pf = NFPROTO_BRIDGE,
diff --git a/net/bridge/netfilter/nf_conntrack_bridge.c b/net/bridge/netfilter/nf_conntrack_bridge.c
index abb090f94ed2..6f877e31709b 100644
--- a/net/bridge/netfilter/nf_conntrack_bridge.c
+++ b/net/bridge/netfilter/nf_conntrack_bridge.c
@@ -291,6 +291,30 @@ static unsigned int nf_ct_bridge_pre(void *priv, struct sk_buff *skb,
 	return nf_conntrack_in(skb, &bridge_state);
 }
 
+static unsigned int nf_ct_bridge_in(void *priv, struct sk_buff *skb,
+				    const struct nf_hook_state *state)
+{
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct;
+
+	if (skb->pkt_type == PACKET_HOST)
+		return NF_ACCEPT;
+
+	/* nf_conntrack_confirm() cannot handle concurrent clones,
+	 * this happens for broad/multicast frames with e.g. macvlan on top
+	 * of the bridge device.
+	 */
+	ct = nf_ct_get(skb, &ctinfo);
+	if (!ct || nf_ct_is_confirmed(ct) || nf_ct_is_template(ct))
+		return NF_ACCEPT;
+
+	/* let inet prerouting call conntrack again */
+	skb->_nfct = 0;
+	nf_ct_put(ct);
+
+	return NF_ACCEPT;
+}
+
 static void nf_ct_bridge_frag_save(struct sk_buff *skb,
 				   struct nf_bridge_frag_data *data)
 {
@@ -386,6 +410,12 @@ static struct nf_hook_ops nf_ct_bridge_hook_ops[] __read_mostly = {
 		.priority	= NF_IP_PRI_CONNTRACK,
 	},
 	{
+		.hook		= nf_ct_bridge_in,
+		.pf		= NFPROTO_BRIDGE,
+		.hooknum	= NF_BR_LOCAL_IN,
+		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM,
+	},
+	{
 		.hook		= nf_ct_bridge_post,
 		.pf		= NFPROTO_BRIDGE,
 		.hooknum	= NF_BR_POST_ROUTING,