Merge tag 'trace-v7.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-tracegrafted

Pull tracing fixes from Steven Rostedt:

 - Don't destroy user event fields when removal fails

   User event fields are destroyed before the event is removed from
   visibility. But that can fail leaving the still visible event with no
   fields. Move the destroying of the fields to after the event is
   successfully removed from visibility.

 - Initialize function graph state is fork before calling
   copy_exec_state()

   For non-CLONE_VM forks, copy_exec_state() allocates a new
   task_exec_state. If that allocation fails, ftrace_graph_exit_task()
   will free the tasks ret_stack pointer. Since that pointer is still
   using the parent's ret_stack, it mistakenly frees the parent's
   pointer too.

   Call ftrace_graph_init() on the task first which will NULL out the
   new tasks's ret_stack and if the copy fails, it will not free
   anything.

 - Remove FGRAPH_MAX_INDEX

   The macro FGRAPH_MAX_INDEX was added but never used. Remove it.

 - Save ent_size in function graph printing of nested functions

   The function graph tracer needs to look at the next event to see if
   the next event is the return of the current function entry. If it is,
   it prints a single line:

	ktime_get();

   Otherwise it prints it like a nested function:

	tick_nohz_irq_exit() {
	    ktime_get();
	    kcpustat_irq_exit();
	}

   In order to look at the next event, it must save the current event so
   that it has the information to print from it. It saves the event in
   the iterator descriptor called "ent". What it doesn't save is the
   ent_size of the event which is now used to know if the function graph
   arguments are to be printed. The peek doesn't save the size so the
   size used happens to be that of the size of the last event that was
   seen.

   Save the entry event size in the iterator descriptor so that the
   correct size is used.

 - Fix several errors with freeing data in the histogram code

   The histogram code had a lot of leaked or or incorrect accounting
   when failures happen. Correct them.

 - Fix histogram regression of .percent and .graph modifiers

   Up until 6.3 histogram values could have "percent" or "graph"
   modifiers that changed how they were printed. But a change that added
   restricting histograms values from being strings, stack traces and
   other modifiers inadvertently prevented them from using the percent
   and graph modifiers, which were legal use cases for values.

   Put back the percent and graph modifiers.

 - Fix various typos in the comments

 - Set the trace_clock before initializing a histogram with clock
   argument

   The histogram API allows the user to specific which trace clock to
   use via a "clock=" string. The histogram is set up first before the
   clock is checked. If the passed in clock is not valid, it exits
   without fully fixing up the histogram leaving it on the list and a
   use-after-free can trigger.

   Update the clock argument first and if it fails then exit gracefully
   before the histogram trigger is placed on any lists.

 - Restore :mod: trailer after parsing in ftrace_set_clr_event

   The function ftrace_set_clr_event() modifies the parse string and
   needs to put it back to what was passed in. It searches for ":mod:"
   via a strsep() but fails to put back the first ':' in the string.

   Add back the ':' in the passed in string.

 - Take trace_array reference when opening a tracer options file

   The options files are dynamically created and some tracers add their
   own options. When a tracer adds their own list of options, the
   trace_array holding them has an array to hold the list of options for
   each tracer. This array increases in size via a krealloc(), and the
   new entry gets a newly allocated array to hold the options of the new
   tracer being added.

   The element in each entry of the tracer's option array holds a
   pointer back to the trace_array, a pointer to the tracer it is
   associated to, a pointer to the flags of the option.

   The issue is that these arrays are freed when the trace_array is
   freed when its instance it represents is removed from the instances
   directory. There's a race that an open of one of these options files
   can happen when the instance is being removed.

   Add a new helper function to be called by the open function of the
   options file to iterate all existing trace_arrays under a lock and
   find the one that has the given option element in one of it's tracer
   arrays. If found, then update the associated trace_array's reference
   counter to keep it from being freed. If not found, have the open call
   return -ENODEV.

 - Disable interrupts when acquiring the lock in rb_wake_up_waiters()

   The function rb_wake_up_waiters() assumes it will be called in
   interrupt context and does not disable irqs when taking
   cpu_buffer->reader_lock, which can be called in hard interrupt
   context. The issue is in PREEMPT_RT, this function is called in
   thread context leaving this lock open to a deadlock.

   Take the lock with interrupts disabled.

 - Use rcu_assign_pointer() for tmp_ops filter hash

   The tmp_ops used in update_ftrace_direct_mod() assigns its
   filter_hash field directly, but that field is annotated as __rcu and
   sparse complains. Assign it with rcu_assign_pointer()

 - Fix use-after-free in enable_trigger_private_data_free()

   The trace_event_call is accessed through the event_trigger_data's
   trace_event_file pointer to put the trace_event_call on freeing. The
   issue is that the trace_event_file data may have been freed already
   causing a use-after-free. Add a field to the event_trigger_data that
   points directly to the trace_event_call so that it can decrement its
   reference directly without needing to go through the
   trace_event_file.

 - Fix accounting of buffer data remote headers

   trace_buffer_desc_size() and trace_remote_alloc_buffer() undercount
   the number of pages is needed for the asked for size as it doesn't
   take into account the meta data on each page. Add a helper function
   to do the calculation properly and use that in these functions.

 - Catch nr_page_va overflow in ring_buffer_desc sizing

   The number of pages per remote ring buffer is capped by
   ring_buffer_desc::nr_page_va (32 bits). A buffer_size large enough to
   overflow that field would silently allocate a descriptor smaller than
   what was asked for.

 - Do not resize the subbuf order if any per_cpu buffer is disabled

   The mmapping of ring buffers disables resizing the subbuffers, but it
   is done per-cpu whereas the subbuf size change is done for all the
   per_cpu buffers under the buffer->mutex. It could change the size of
   some while the mapping is happening on others. Have the resize of the
   subbuf order check all the per_cpu buffers under the lock to see if
   any of them is disabled before starting and causing an inconsistency
   between buffers that are being mapped.

* tag 'trace-v7.3-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace: (25 commits)
  ring-buffer: Check resize_disabled before publishing the new subbuf order
  tracing/remotes: Catch nr_page_va overflow in ring_buffer_desc sizing
  tracing/remotes: Account for ring buffer page header in size calculation
  tracing: Don't dereference trace_event_file in deferred trigger free
  ftrace: Use rcu_assign_pointer() for tmp_ops filter hash
  ring-buffer: Acquire the lock with irqsave in rb_wake_up_waiters()
  tracing: Take trace_array reference when opening a tracer options file
  tracing: Fix ring_buffer_read_page_size() kernel-doc
  tracing: Restore :mod: trailer after parsing in ftrace_set_clr_event()
  tracing: Fix memory corruption from a "STACKTRACE" histogram key
  tracing: Fix memory corruption from the histogram stacktrace modifier
  tracing: Undo the registration when enabling the histogram trigger fails
  tracing: Take the reference before publishing the named histogram trigger
  tracing: Set the trace clock before registering the histogram trigger
  tracing: Fix typo "preceeded" in comment
  tracing: Fix typo "availabe" in comment
  tracing: Let histogram values keep the percent and graph modifiers
  tracing: Keep the entry count when the histogram stats allocation fails
  tracing: Free histogram the field rejected for a bad modifier
  tracing: Free histogram the var ref when its initialization fails
  ...

1 files changed, 2273 insertions, 0 deletions

diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
new file mode 100644
index 000000000..7a362f2e2
--- /dev/null
+++ b/net/ipv4/fib_semantics.c
@@ -0,0 +1,2273 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * INET		An implementation of the TCP/IP protocol suite for the LINUX
+ *		operating system.  INET is implemented using the  BSD Socket
+ *		interface as the means of communication with the user level.
+ *
+ *		IPv4 Forwarding Information Base: semantics.
+ *
+ * Authors:	Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
+ */
+
+#include <linux/uaccess.h>
+#include <linux/bitops.h>
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/jiffies.h>
+#include <linux/mm.h>
+#include <linux/string.h>
+#include <linux/socket.h>
+#include <linux/sockios.h>
+#include <linux/errno.h>
+#include <linux/in.h>
+#include <linux/inet.h>
+#include <linux/inetdevice.h>
+#include <linux/netdevice.h>
+#include <linux/if_arp.h>
+#include <linux/proc_fs.h>
+#include <linux/skbuff.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/netlink.h>
+#include <linux/hash.h>
+#include <linux/nospec.h>
+
+#include <net/arp.h>
+#include <net/inet_dscp.h>
+#include <net/ip.h>
+#include <net/protocol.h>
+#include <net/route.h>
+#include <net/tcp.h>
+#include <net/sock.h>
+#include <net/ip_fib.h>
+#include <net/ip6_fib.h>
+#include <net/nexthop.h>
+#include <net/netlink.h>
+#include <net/rtnh.h>
+#include <net/lwtunnel.h>
+#include <net/fib_notifier.h>
+#include <net/addrconf.h>
+
+#include "fib_lookup.h"
+
+/* for_nexthops and change_nexthops only used when nexthop object
+ * is not set in a fib_info. The logic within can reference fib_nh.
+ */
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+
+#define for_nexthops(fi) {						\
+	int nhsel; const struct fib_nh *nh;				\
+	for (nhsel = 0, nh = (fi)->fib_nh;				\
+	     nhsel < fib_info_num_path((fi));				\
+	     nh++, nhsel++)
+
+#define change_nexthops(fi) {						\
+	int nhsel; struct fib_nh *nexthop_nh;				\
+	for (nhsel = 0,	nexthop_nh = (struct fib_nh *)((fi)->fib_nh);	\
+	     nhsel < fib_info_num_path((fi));				\
+	     nexthop_nh++, nhsel++)
+
+#else /* CONFIG_IP_ROUTE_MULTIPATH */
+
+/* Hope, that gcc will optimize it to get rid of dummy loop */
+
+#define for_nexthops(fi) {						\
+	int nhsel; const struct fib_nh *nh = (fi)->fib_nh;		\
+	for (nhsel = 0; nhsel < 1; nhsel++)
+
+#define change_nexthops(fi) {						\
+	int nhsel;							\
+	struct fib_nh *nexthop_nh = (struct fib_nh *)((fi)->fib_nh);	\
+	for (nhsel = 0; nhsel < 1; nhsel++)
+
+#endif /* CONFIG_IP_ROUTE_MULTIPATH */
+
+#define endfor_nexthops(fi) }
+
+
+const struct fib_prop fib_props[RTN_MAX + 1] = {
+	[RTN_UNSPEC] = {
+		.error	= 0,
+		.scope	= RT_SCOPE_NOWHERE,
+	},
+	[RTN_UNICAST] = {
+		.error	= 0,
+		.scope	= RT_SCOPE_UNIVERSE,
+	},
+	[RTN_LOCAL] = {
+		.error	= 0,
+		.scope	= RT_SCOPE_HOST,
+	},
+	[RTN_BROADCAST] = {
+		.error	= 0,
+		.scope	= RT_SCOPE_LINK,
+	},
+	[RTN_ANYCAST] = {
+		.error	= 0,
+		.scope	= RT_SCOPE_LINK,
+	},
+	[RTN_MULTICAST] = {
+		.error	= 0,
+		.scope	= RT_SCOPE_UNIVERSE,
+	},
+	[RTN_BLACKHOLE] = {
+		.error	= -EINVAL,
+		.scope	= RT_SCOPE_UNIVERSE,
+	},
+	[RTN_UNREACHABLE] = {
+		.error	= -EHOSTUNREACH,
+		.scope	= RT_SCOPE_UNIVERSE,
+	},
+	[RTN_PROHIBIT] = {
+		.error	= -EACCES,
+		.scope	= RT_SCOPE_UNIVERSE,
+	},
+	[RTN_THROW] = {
+		.error	= -EAGAIN,
+		.scope	= RT_SCOPE_UNIVERSE,
+	},
+	[RTN_NAT] = {
+		.error	= -EINVAL,
+		.scope	= RT_SCOPE_NOWHERE,
+	},
+	[RTN_XRESOLVE] = {
+		.error	= -EINVAL,
+		.scope	= RT_SCOPE_NOWHERE,
+	},
+};
+
+static void rt_fibinfo_free(struct rtable __rcu **rtp)
+{
+	struct rtable *rt = rcu_dereference_protected(*rtp, 1);
+
+	if (!rt)
+		return;
+
+	/* Not even needed : RCU_INIT_POINTER(*rtp, NULL);
+	 * because we waited an RCU grace period before calling
+	 * free_fib_info_rcu()
+	 */
+
+	dst_dev_put(&rt->dst);
+	dst_release_immediate(&rt->dst);
+}
+
+static void free_nh_exceptions(struct fib_nh_common *nhc)
+{
+	struct fnhe_hash_bucket *hash;
+	int i;
+
+	hash = rcu_dereference_protected(nhc->nhc_exceptions, 1);
+	if (!hash)
+		return;
+	for (i = 0; i < FNHE_HASH_SIZE; i++) {
+		struct fib_nh_exception *fnhe;
+
+		fnhe = rcu_dereference_protected(hash[i].chain, 1);
+		while (fnhe) {
+			struct fib_nh_exception *next;
+
+			next = rcu_dereference_protected(fnhe->fnhe_next, 1);
+
+			rt_fibinfo_free(&fnhe->fnhe_rth_input);
+			rt_fibinfo_free(&fnhe->fnhe_rth_output);
+
+			kfree(fnhe);
+
+			fnhe = next;
+		}
+	}
+	kfree(hash);
+}
+
+static void rt_fibinfo_free_cpus(struct rtable __rcu * __percpu *rtp)
+{
+	int cpu;
+
+	if (!rtp)
+		return;
+
+	for_each_possible_cpu(cpu) {
+		struct rtable *rt;
+
+		rt = rcu_dereference_protected(*per_cpu_ptr(rtp, cpu), 1);
+		if (rt) {
+			dst_dev_put(&rt->dst);
+			dst_release_immediate(&rt->dst);
+		}
+	}
+	free_percpu(rtp);
+}
+
+void fib_nh_common_release(struct fib_nh_common *nhc)
+{
+	netdev_put(nhc->nhc_dev, &nhc->nhc_dev_tracker);
+	lwtstate_put(nhc->nhc_lwtstate);
+	rt_fibinfo_free_cpus(nhc->nhc_pcpu_rth_output);
+	rt_fibinfo_free(&nhc->nhc_rth_input);
+	free_nh_exceptions(nhc);
+}
+
+void fib_nh_release(struct net *net, struct fib_nh *fib_nh)
+{
+#ifdef CONFIG_IP_ROUTE_CLASSID
+	if (fib_nh->nh_tclassid)
+		atomic_dec(&net->ipv4.fib_num_tclassid_users);
+#endif
+	fib_nh_common_release(&fib_nh->nh_common);
+}
+
+/* Release a nexthop info record */
+static void free_fib_info_rcu(struct rcu_head *head)
+{
+	struct fib_info *fi = container_of(head, struct fib_info, rcu);
+
+	if (fi->nh) {
+		nexthop_put(fi->nh);
+	} else {
+		change_nexthops(fi) {
+			fib_nh_release(fi->fib_net, nexthop_nh);
+		} endfor_nexthops(fi);
+	}
+
+	ip_fib_metrics_put(fi->fib_metrics);
+
+	kfree(fi);
+}
+
+void free_fib_info(struct fib_info *fi)
+{
+	if (fi->fib_dead == 0) {
+		pr_warn("Freeing alive fib_info %p\n", fi);
+		return;
+	}
+
+	call_rcu_hurry(&fi->rcu, free_fib_info_rcu);
+}
+EXPORT_SYMBOL_GPL(free_fib_info);
+
+void fib_release_info(struct fib_info *fi)
+{
+	ASSERT_RTNL();
+	if (fi && refcount_dec_and_test(&fi->fib_treeref)) {
+		hlist_del(&fi->fib_hash);
+		fi->fib_net->ipv4.fib_info_cnt--;
+
+		if (fi->fib_prefsrc)
+			hlist_del(&fi->fib_lhash);
+		if (fi->nh) {
+			list_del(&fi->nh_list);
+		} else {
+			change_nexthops(fi) {
+				if (!nexthop_nh->fib_nh_dev)
+					continue;
+				hlist_del_rcu(&nexthop_nh->nh_hash);
+			} endfor_nexthops(fi)
+		}
+		/* Paired with READ_ONCE() from fib_table_lookup() */
+		WRITE_ONCE(fi->fib_dead, 1);
+		fib_info_put(fi);
+	}
+}
+
+static inline int nh_comp(struct fib_info *fi, struct fib_info *ofi)
+{
+	const struct fib_nh *onh;
+
+	if (fi->nh || ofi->nh)
+		return nexthop_cmp(fi->nh, ofi->nh) ? 0 : -1;
+
+	if (ofi->fib_nhs == 0)
+		return 0;
+
+	for_nexthops(fi) {
+		onh = fib_info_nh(ofi, nhsel);
+
+		if (nh->fib_nh_oif != onh->fib_nh_oif ||
+		    nh->fib_nh_gw_family != onh->fib_nh_gw_family ||
+		    nh->fib_nh_scope != onh->fib_nh_scope ||
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+		    nh->fib_nh_weight != onh->fib_nh_weight ||
+#endif
+#ifdef CONFIG_IP_ROUTE_CLASSID
+		    nh->nh_tclassid != onh->nh_tclassid ||
+#endif
+		    lwtunnel_cmp_encap(nh->fib_nh_lws, onh->fib_nh_lws) ||
+		    ((nh->fib_nh_flags ^ onh->fib_nh_flags) & ~RTNH_COMPARE_MASK))
+			return -1;
+
+		if (nh->fib_nh_gw_family == AF_INET &&
+		    nh->fib_nh_gw4 != onh->fib_nh_gw4)
+			return -1;
+
+		if (nh->fib_nh_gw_family == AF_INET6 &&
+		    ipv6_addr_cmp(&nh->fib_nh_gw6, &onh->fib_nh_gw6))
+			return -1;
+	} endfor_nexthops(fi);
+	return 0;
+}
+
+static struct hlist_head *fib_nh_head(struct net_device *dev)
+{
+	return &dev->fib_nh_head;
+}
+
+static unsigned int fib_info_hashfn_1(int init_val, u8 protocol, u8 scope,
+				      u32 prefsrc, u32 priority)
+{
+	unsigned int val = init_val;
+
+	val ^= (protocol << 8) | scope;
+	val ^= prefsrc;
+	val ^= priority;
+
+	return val;
+}
+
+static unsigned int fib_info_hashfn_result(const struct net *net,
+					   unsigned int val)
+{
+	return hash_32(val ^ net_hash_mix(net), net->ipv4.fib_info_hash_bits);
+}
+
+static struct hlist_head *fib_info_hash_bucket(struct fib_info *fi)
+{
+	struct net *net = fi->fib_net;
+	unsigned int val;
+
+	val = fib_info_hashfn_1(fi->fib_nhs, fi->fib_protocol,
+				fi->fib_scope, (__force u32)fi->fib_prefsrc,
+				fi->fib_priority);
+
+	if (fi->nh) {
+		val ^= fi->nh->id;
+	} else {
+		for_nexthops(fi) {
+			val ^= nh->fib_nh_oif;
+		} endfor_nexthops(fi)
+	}
+
+	return &net->ipv4.fib_info_hash[fib_info_hashfn_result(net, val)];
+}
+
+static struct hlist_head *fib_info_laddrhash_bucket(const struct net *net,
+						    __be32 val)
+{
+	unsigned int hash_bits = net->ipv4.fib_info_hash_bits;
+	u32 slot;
+
+	slot = hash_32(net_hash_mix(net) ^ (__force u32)val, hash_bits);
+
+	return &net->ipv4.fib_info_hash[(1 << hash_bits) + slot];
+}
+
+static struct hlist_head *fib_info_hash_alloc(unsigned int hash_bits)
+{
+	/* The second half is used for prefsrc */
+	return kvzalloc_objs(struct hlist_head, (1 << hash_bits) * 2);
+}
+
+static void fib_info_hash_free(struct hlist_head *head)
+{
+	kvfree(head);
+}
+
+static void fib_info_hash_grow(struct net *net)
+{
+	unsigned int old_size = 1 << net->ipv4.fib_info_hash_bits;
+	struct hlist_head *new_info_hash, *old_info_hash;
+	unsigned int i;
+
+	if (net->ipv4.fib_info_cnt < old_size)
+		return;
+
+	new_info_hash = fib_info_hash_alloc(net->ipv4.fib_info_hash_bits + 1);
+	if (!new_info_hash)
+		return;
+
+	old_info_hash = net->ipv4.fib_info_hash;
+	net->ipv4.fib_info_hash = new_info_hash;
+	net->ipv4.fib_info_hash_bits += 1;
+
+	for (i = 0; i < old_size; i++) {
+		struct hlist_head *head = &old_info_hash[i];
+		struct hlist_node *n;
+		struct fib_info *fi;
+
+		hlist_for_each_entry_safe(fi, n, head, fib_hash)
+			hlist_add_head(&fi->fib_hash, fib_info_hash_bucket(fi));
+	}
+
+	for (i = 0; i < old_size; i++) {
+		struct hlist_head *lhead = &old_info_hash[old_size + i];
+		struct hlist_node *n;
+		struct fib_info *fi;
+
+		hlist_for_each_entry_safe(fi, n, lhead, fib_lhash)
+			hlist_add_head(&fi->fib_lhash,
+				       fib_info_laddrhash_bucket(fi->fib_net,
+								 fi->fib_prefsrc));
+	}
+
+	fib_info_hash_free(old_info_hash);
+}
+
+/* no metrics, only nexthop id */
+static struct fib_info *fib_find_info_nh(struct net *net,
+					 const struct fib_config *cfg)
+{
+	struct hlist_head *head;
+	struct fib_info *fi;
+	unsigned int hash;
+
+	hash = fib_info_hashfn_1(cfg->fc_nh_id,
+				 cfg->fc_protocol, cfg->fc_scope,
+				 (__force u32)cfg->fc_prefsrc,
+				 cfg->fc_priority);
+	hash = fib_info_hashfn_result(net, hash);
+	head = &net->ipv4.fib_info_hash[hash];
+
+	hlist_for_each_entry(fi, head, fib_hash) {
+		if (!fi->nh || fi->nh->id != cfg->fc_nh_id)
+			continue;
+
+		if (cfg->fc_protocol == fi->fib_protocol &&
+		    cfg->fc_scope == fi->fib_scope &&
+		    cfg->fc_prefsrc == fi->fib_prefsrc &&
+		    cfg->fc_priority == fi->fib_priority &&
+		    cfg->fc_type == fi->fib_type &&
+		    cfg->fc_table == fi->fib_tb_id &&
+		    !((cfg->fc_flags ^ fi->fib_flags) & ~RTNH_COMPARE_MASK))
+			return fi;
+	}
+
+	return NULL;
+}
+
+static struct fib_info *fib_find_info(struct fib_info *nfi)
+{
+	struct hlist_head *head = fib_info_hash_bucket(nfi);
+	struct fib_info *fi;
+
+	hlist_for_each_entry(fi, head, fib_hash) {
+		if (fi->fib_nhs != nfi->fib_nhs)
+			continue;
+
+		if (nfi->fib_protocol == fi->fib_protocol &&
+		    nfi->fib_scope == fi->fib_scope &&
+		    nfi->fib_prefsrc == fi->fib_prefsrc &&
+		    nfi->fib_priority == fi->fib_priority &&
+		    nfi->fib_type == fi->fib_type &&
+		    nfi->fib_tb_id == fi->fib_tb_id &&
+		    memcmp(nfi->fib_metrics, fi->fib_metrics,
+			   sizeof(u32) * RTAX_MAX) == 0 &&
+		    !((nfi->fib_flags ^ fi->fib_flags) & ~RTNH_COMPARE_MASK) &&
+		    nh_comp(fi, nfi) == 0)
+			return fi;
+	}
+
+	return NULL;
+}
+
+/* Check, that the gateway is already configured.
+ * Used only by redirect accept routine, under rcu_read_lock();
+ */
+int ip_fib_check_default(__be32 gw, struct net_device *dev)
+{
+	struct hlist_head *head;
+	struct fib_nh *nh;
+
+	head = fib_nh_head(dev);
+
+	hlist_for_each_entry_rcu(nh, head, nh_hash) {
+		DEBUG_NET_WARN_ON_ONCE(nh->fib_nh_dev != dev);
+		if (nh->fib_nh_gw4 == gw &&
+		    !(nh->fib_nh_flags & RTNH_F_DEAD)) {
+			return 0;
+		}
+	}
+
+	return -1;
+}
+
+static size_t fib_nexthop_nlmsg_size(const struct fib_nh_common *nhc,
+				     bool skip_oif)
+{
+	size_t nhsize = 0;
+
+	switch (nhc->nhc_gw_family) {
+	case AF_INET:
+		nhsize += nla_total_size(4); /* RTA_GATEWAY */
+		break;
+	case AF_INET6:
+		nhsize += nla_total_size(sizeof(struct rtvia) +
+					 sizeof(struct in6_addr));
+		break;
+	}
+
+	if (!skip_oif && nhc->nhc_dev)
+		nhsize += nla_total_size(4); /* RTA_OIF */
+
+	if (nhc->nhc_lwtstate) {
+		/* RTA_ENCAP */
+		nhsize += lwtunnel_get_encap_size(nhc->nhc_lwtstate);
+		/* RTA_ENCAP_TYPE */
+		nhsize += nla_total_size(2);
+	}
+
+	return nhsize;
+}
+
+size_t fib_nlmsg_size(struct fib_info *fi)
+{
+	size_t payload = NLMSG_ALIGN(sizeof(struct rtmsg))
+			 + nla_total_size(4) /* RTA_TABLE */
+			 + nla_total_size(4) /* RTA_DST */
+			 + nla_total_size(4) /* RTA_PRIORITY */
+			 + nla_total_size(4) /* RTA_PREFSRC */
+			 + nla_total_size(TCP_CA_NAME_MAX); /* RTAX_CC_ALGO */
+	unsigned int nhs = fib_info_num_path(fi);
+
+	/* space for nested metrics */
+	payload += nla_total_size((RTAX_MAX * nla_total_size(4)));
+
+	if (fi->nh)
+		payload += nla_total_size(4); /* RTA_NH_ID */
+
+	if (nhs) {
+		size_t mpsize = 0;
+		unsigned int i;
+
+		for (i = 0; i < fib_info_num_path(fi); i++) {
+			struct fib_nh_common *nhc = fib_info_nhc(fi, i);
+			size_t nhsize;
+
+			nhsize = fib_nexthop_nlmsg_size(nhc, nhs != 1);
+
+			if (nhs != 1)
+				nhsize += NLA_ALIGN(sizeof(struct rtnexthop));
+
+#ifdef CONFIG_IP_ROUTE_CLASSID
+			if (nhc->nhc_family == AF_INET) {
+				struct fib_nh *nh;
+
+				nh = container_of(nhc, struct fib_nh, nh_common);
+				if (nh->nh_tclassid)
+					nhsize += nla_total_size(4);
+			}
+#endif
+			if (nhs == 1)
+				payload += nhsize;
+			else
+				mpsize += nhsize;
+		}
+
+		if (nhs != 1)
+			payload += nla_total_size(mpsize);
+	}
+
+	return payload;
+}
+
+void rtmsg_fib(int event, __be32 key, struct fib_alias *fa,
+	       int dst_len, u32 tb_id, const struct nl_info *info,
+	       unsigned int nlm_flags)
+{
+	struct fib_rt_info fri;
+	struct sk_buff *skb;
+	u32 seq = info->nlh ? info->nlh->nlmsg_seq : 0;
+	int err = -ENOBUFS;
+
+	skb = nlmsg_new(fib_nlmsg_size(fa->fa_info), GFP_KERNEL);
+	if (!skb)
+		goto errout;
+
+	fri.fi = fa->fa_info;
+	fri.tb_id = tb_id;
+	fri.dst = key;
+	fri.dst_len = dst_len;
+	fri.dscp = fa->fa_dscp;
+	fri.type = fa->fa_type;
+	fri.offload = READ_ONCE(fa->offload);
+	fri.trap = READ_ONCE(fa->trap);
+	fri.offload_failed = READ_ONCE(fa->offload_failed);
+	err = fib_dump_info(skb, info->portid, seq, event, &fri, nlm_flags);
+	if (err < 0) {
+		/* -EMSGSIZE implies BUG in fib_nlmsg_size() */
+		WARN_ON(err == -EMSGSIZE);
+		kfree_skb(skb);
+		goto errout;
+	}
+	rtnl_notify(skb, info->nl_net, info->portid, RTNLGRP_IPV4_ROUTE,
+		    info->nlh, GFP_KERNEL);
+	return;
+errout:
+	rtnl_set_sk_err(info->nl_net, RTNLGRP_IPV4_ROUTE, err);
+}
+
+static int fib_detect_death(struct fib_info *fi, int order,
+			    struct fib_info **last_resort, int *last_idx,
+			    int dflt)
+{
+	const struct fib_nh_common *nhc = fib_info_nhc(fi, 0);
+	struct neighbour *n;
+	int state = NUD_NONE;
+
+	if (likely(nhc->nhc_gw_family == AF_INET))
+		n = neigh_lookup(&arp_tbl, &nhc->nhc_gw.ipv4, nhc->nhc_dev);
+	else if (IS_ENABLED(CONFIG_IPV6) && nhc->nhc_gw_family == AF_INET6)
+		n = neigh_lookup(&nd_tbl, &nhc->nhc_gw.ipv6, nhc->nhc_dev);
+	else
+		n = NULL;
+
+	if (n) {
+		state = READ_ONCE(n->nud_state);
+		neigh_release(n);
+	} else {
+		return 0;
+	}
+	if (state == NUD_REACHABLE)
+		return 0;
+	if ((state & NUD_VALID) && order != dflt)
+		return 0;
+	if ((state & NUD_VALID) ||
+	    (*last_idx < 0 && order > dflt && state != NUD_INCOMPLETE)) {
+		*last_resort = fi;
+		*last_idx = order;
+	}
+	return 1;
+}
+
+int fib_nh_common_init(struct net *net, struct fib_nh_common *nhc,
+		       struct nlattr *encap, u16 encap_type,
+		       void *cfg, gfp_t gfp_flags,
+		       struct netlink_ext_ack *extack)
+{
+	int err;
+
+	nhc->nhc_pcpu_rth_output = alloc_percpu_gfp(struct rtable __rcu *,
+						    gfp_flags);
+	if (!nhc->nhc_pcpu_rth_output)
+		return -ENOMEM;
+
+	if (encap) {
+		struct lwtunnel_state *lwtstate;
+
+		err = lwtunnel_build_state(net, encap_type, encap,
+					   nhc->nhc_family, cfg, &lwtstate,
+					   extack);
+		if (err)
+			goto lwt_failure;
+
+		nhc->nhc_lwtstate = lwtstate_get(lwtstate);
+	}
+
+	return 0;
+
+lwt_failure:
+	rt_fibinfo_free_cpus(nhc->nhc_pcpu_rth_output);
+	nhc->nhc_pcpu_rth_output = NULL;
+	return err;
+}
+
+int fib_nh_init(struct net *net, struct fib_nh *nh,
+		struct fib_config *cfg, int nh_weight,
+		struct netlink_ext_ack *extack)
+{
+	int err;
+
+	nh->fib_nh_family = AF_INET;
+
+	err = fib_nh_common_init(net, &nh->nh_common, cfg->fc_encap,
+				 cfg->fc_encap_type, cfg, GFP_KERNEL, extack);
+	if (err)
+		return err;
+
+	nh->fib_nh_oif = cfg->fc_oif;
+	nh->fib_nh_gw_family = cfg->fc_gw_family;
+	if (cfg->fc_gw_family == AF_INET)
+		nh->fib_nh_gw4 = cfg->fc_gw4;
+	else if (cfg->fc_gw_family == AF_INET6)
+		nh->fib_nh_gw6 = cfg->fc_gw6;
+
+	nh->fib_nh_flags = cfg->fc_flags;
+
+#ifdef CONFIG_IP_ROUTE_CLASSID
+	nh->nh_tclassid = cfg->fc_flow;
+	if (nh->nh_tclassid)
+		atomic_inc(&net->ipv4.fib_num_tclassid_users);
+#endif
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+	nh->fib_nh_weight = nh_weight;
+#endif
+	return 0;
+}
+
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+
+static int fib_count_nexthops(struct rtnexthop *rtnh, int remaining,
+			      struct netlink_ext_ack *extack)
+{
+	int nhs = 0;
+
+	while (rtnh_ok(rtnh, remaining)) {
+		nhs++;
+		rtnh = rtnh_next(rtnh, &remaining);
+	}
+
+	/* leftover implies invalid nexthop configuration, discard it */
+	if (remaining > 0) {
+		NL_SET_ERR_MSG(extack,
+			       "Invalid nexthop configuration - extra data after nexthops");
+		nhs = 0;
+	}
+
+	return nhs;
+}
+
+static int fib_gw_from_attr(__be32 *gw, struct nlattr *nla,
+			    struct netlink_ext_ack *extack)
+{
+	if (nla_len(nla) < sizeof(*gw)) {
+		NL_SET_ERR_MSG(extack, "Invalid IPv4 address in RTA_GATEWAY");
+		return -EINVAL;
+	}
+
+	*gw = nla_get_in_addr(nla);
+
+	return 0;
+}
+
+/* only called when fib_nh is integrated into fib_info */
+static int fib_get_nhs(struct fib_info *fi, struct rtnexthop *rtnh,
+		       int remaining, struct fib_config *cfg,
+		       struct netlink_ext_ack *extack)
+{
+	struct net *net = fi->fib_net;
+	struct fib_config fib_cfg;
+	struct fib_nh *nh;
+	int ret;
+
+	change_nexthops(fi) {
+		int attrlen;
+
+		memset(&fib_cfg, 0, sizeof(fib_cfg));
+
+		if (!rtnh_ok(rtnh, remaining)) {
+			NL_SET_ERR_MSG(extack,
+				       "Invalid nexthop configuration - extra data after nexthop");
+			return -EINVAL;
+		}
+
+		if (rtnh->rtnh_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN)) {
+			NL_SET_ERR_MSG(extack,
+				       "Invalid flags for nexthop - can not contain DEAD or LINKDOWN");
+			return -EINVAL;
+		}
+
+		fib_cfg.fc_flags = (cfg->fc_flags & ~0xFF) | rtnh->rtnh_flags;
+		fib_cfg.fc_oif = rtnh->rtnh_ifindex;
+
+		attrlen = rtnh_attrlen(rtnh);
+		if (attrlen > 0) {
+			struct nlattr *nla, *nlav, *attrs = rtnh_attrs(rtnh);
+
+			nla = nla_find(attrs, attrlen, RTA_GATEWAY);
+			nlav = nla_find(attrs, attrlen, RTA_VIA);
+			if (nla && nlav) {
+				NL_SET_ERR_MSG(extack,
+					       "Nexthop configuration can not contain both GATEWAY and VIA");
+				return -EINVAL;
+			}
+			if (nla) {
+				ret = fib_gw_from_attr(&fib_cfg.fc_gw4, nla,
+						       extack);
+				if (ret)
+					goto errout;
+
+				if (fib_cfg.fc_gw4)
+					fib_cfg.fc_gw_family = AF_INET;
+			} else if (nlav) {
+				ret = fib_gw_from_via(&fib_cfg, nlav, extack);
+				if (ret)
+					goto errout;
+			}
+
+			nla = nla_find(attrs, attrlen, RTA_FLOW);
+			if (nla) {
+				if (nla_len(nla) < sizeof(u32)) {
+					NL_SET_ERR_MSG(extack, "Invalid RTA_FLOW");
+					return -EINVAL;
+				}
+				fib_cfg.fc_flow = nla_get_u32(nla);
+			}
+
+			fib_cfg.fc_encap = nla_find(attrs, attrlen, RTA_ENCAP);
+			/* RTA_ENCAP_TYPE length checked in
+			 * lwtunnel_valid_encap_type_attr
+			 */
+			nla = nla_find(attrs, attrlen, RTA_ENCAP_TYPE);
+			if (nla)
+				fib_cfg.fc_encap_type = nla_get_u16(nla);
+		}
+
+		ret = fib_nh_init(net, nexthop_nh, &fib_cfg,
+				  rtnh->rtnh_hops + 1, extack);
+		if (ret)
+			goto errout;
+
+		rtnh = rtnh_next(rtnh, &remaining);
+	} endfor_nexthops(fi);
+
+	ret = -EINVAL;
+	nh = fib_info_nh(fi, 0);
+	if (cfg->fc_oif && nh->fib_nh_oif != cfg->fc_oif) {
+		NL_SET_ERR_MSG(extack,
+			       "Nexthop device index does not match RTA_OIF");
+		goto errout;
+	}
+	if (cfg->fc_gw_family) {
+		if (cfg->fc_gw_family != nh->fib_nh_gw_family ||
+		    (cfg->fc_gw_family == AF_INET &&
+		     nh->fib_nh_gw4 != cfg->fc_gw4) ||
+		    (cfg->fc_gw_family == AF_INET6 &&
+		     ipv6_addr_cmp(&nh->fib_nh_gw6, &cfg->fc_gw6))) {
+			NL_SET_ERR_MSG(extack,
+				       "Nexthop gateway does not match RTA_GATEWAY or RTA_VIA");
+			goto errout;
+		}
+	}
+#ifdef CONFIG_IP_ROUTE_CLASSID
+	if (cfg->fc_flow && nh->nh_tclassid != cfg->fc_flow) {
+		NL_SET_ERR_MSG(extack,
+			       "Nexthop class id does not match RTA_FLOW");
+		goto errout;
+	}
+#endif
+	ret = 0;
+errout:
+	return ret;
+}
+
+/* only called when fib_nh is integrated into fib_info */
+static void fib_rebalance(struct fib_info *fi)
+{
+	int total;
+	int w;
+
+	if (fib_info_num_path(fi) < 2)
+		return;
+
+	total = 0;
+	for_nexthops(fi) {
+		if (nh->fib_nh_flags & RTNH_F_DEAD)
+			continue;
+
+		if (ip_ignore_linkdown(nh->fib_nh_dev) &&
+		    nh->fib_nh_flags & RTNH_F_LINKDOWN)
+			continue;
+
+		total += nh->fib_nh_weight;
+	} endfor_nexthops(fi);
+
+	w = 0;
+	change_nexthops(fi) {
+		int upper_bound;
+
+		if (!total || nexthop_nh->fib_nh_flags & RTNH_F_DEAD) {
+			upper_bound = -1;
+		} else if (ip_ignore_linkdown(nexthop_nh->fib_nh_dev) &&
+			   nexthop_nh->fib_nh_flags & RTNH_F_LINKDOWN) {
+			upper_bound = -1;
+		} else {
+			w += nexthop_nh->fib_nh_weight;
+			upper_bound = DIV_ROUND_CLOSEST_ULL((u64)w << 31,
+							    total) - 1;
+		}
+
+		atomic_set(&nexthop_nh->fib_nh_upper_bound, upper_bound);
+	} endfor_nexthops(fi);
+}
+#else /* CONFIG_IP_ROUTE_MULTIPATH */
+
+static int fib_get_nhs(struct fib_info *fi, struct rtnexthop *rtnh,
+		       int remaining, struct fib_config *cfg,
+		       struct netlink_ext_ack *extack)
+{
+	NL_SET_ERR_MSG(extack, "Multipath support not enabled in kernel");
+
+	return -EINVAL;
+}
+
+#define fib_rebalance(fi) do { } while (0)
+
+#endif /* CONFIG_IP_ROUTE_MULTIPATH */
+
+static int fib_encap_match(struct net *net, u16 encap_type,
+			   struct nlattr *encap,
+			   const struct fib_nh *nh,
+			   const struct fib_config *cfg,
+			   struct netlink_ext_ack *extack)
+{
+	struct lwtunnel_state *lwtstate;
+	int ret, result = 0;
+
+	if (encap_type == LWTUNNEL_ENCAP_NONE)
+		return 0;
+
+	ret = lwtunnel_build_state(net, encap_type, encap, AF_INET,
+				   cfg, &lwtstate, extack);
+	if (!ret) {
+		result = lwtunnel_cmp_encap(lwtstate, nh->fib_nh_lws);
+		lwtstate_free(lwtstate);
+	}
+
+	return result;
+}
+
+int fib_nh_match(struct net *net, struct fib_config *cfg, struct fib_info *fi,
+		 struct netlink_ext_ack *extack)
+{
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+	struct rtnexthop *rtnh;
+	int remaining;
+#endif
+
+	if (cfg->fc_priority && cfg->fc_priority != fi->fib_priority)
+		return 1;
+
+	if (cfg->fc_nh_id) {
+		if (fi->nh && cfg->fc_nh_id == fi->nh->id)
+			return 0;
+		return 1;
+	}
+
+	if (fi->nh) {
+		if (cfg->fc_oif || cfg->fc_gw_family || cfg->fc_mp)
+			return 1;
+		return 0;
+	}
+
+	if (cfg->fc_oif || cfg->fc_gw_family) {
+		struct fib_nh *nh;
+
+		nh = fib_info_nh(fi, 0);
+		if (cfg->fc_encap) {
+			if (fib_encap_match(net, cfg->fc_encap_type,
+					    cfg->fc_encap, nh, cfg, extack))
+				return 1;
+		}
+#ifdef CONFIG_IP_ROUTE_CLASSID
+		if (cfg->fc_flow &&
+		    cfg->fc_flow != nh->nh_tclassid)
+			return 1;
+#endif
+		if ((cfg->fc_oif && cfg->fc_oif != nh->fib_nh_oif) ||
+		    (cfg->fc_gw_family &&
+		     cfg->fc_gw_family != nh->fib_nh_gw_family))
+			return 1;
+
+		if (cfg->fc_gw_family == AF_INET &&
+		    cfg->fc_gw4 != nh->fib_nh_gw4)
+			return 1;
+
+		if (cfg->fc_gw_family == AF_INET6 &&
+		    ipv6_addr_cmp(&cfg->fc_gw6, &nh->fib_nh_gw6))
+			return 1;
+
+		return 0;
+	}
+
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+	if (!cfg->fc_mp)
+		return 0;
+
+	rtnh = cfg->fc_mp;
+	remaining = cfg->fc_mp_len;
+
+	for_nexthops(fi) {
+		int attrlen;
+
+		if (!rtnh_ok(rtnh, remaining))
+			return -EINVAL;
+
+		if (rtnh->rtnh_ifindex && rtnh->rtnh_ifindex != nh->fib_nh_oif)
+			return 1;
+
+		attrlen = rtnh_attrlen(rtnh);
+		if (attrlen > 0) {
+			struct nlattr *nla, *nlav, *attrs = rtnh_attrs(rtnh);
+			int err;
+
+			nla = nla_find(attrs, attrlen, RTA_GATEWAY);
+			nlav = nla_find(attrs, attrlen, RTA_VIA);
+			if (nla && nlav) {
+				NL_SET_ERR_MSG(extack,
+					       "Nexthop configuration can not contain both GATEWAY and VIA");
+				return -EINVAL;
+			}
+
+			if (nla) {
+				__be32 gw;
+
+				err = fib_gw_from_attr(&gw, nla, extack);
+				if (err)
+					return err;
+
+				if (nh->fib_nh_gw_family != AF_INET ||
+				    gw != nh->fib_nh_gw4)
+					return 1;
+			} else if (nlav) {
+				struct fib_config cfg2;
+
+				err = fib_gw_from_via(&cfg2, nlav, extack);
+				if (err)
+					return err;
+
+				switch (nh->fib_nh_gw_family) {
+				case AF_INET:
+					if (cfg2.fc_gw_family != AF_INET ||
+					    cfg2.fc_gw4 != nh->fib_nh_gw4)
+						return 1;
+					break;
+				case AF_INET6:
+					if (cfg2.fc_gw_family != AF_INET6 ||
+					    ipv6_addr_cmp(&cfg2.fc_gw6,
+							  &nh->fib_nh_gw6))
+						return 1;
+					break;
+				}
+			}
+
+#ifdef CONFIG_IP_ROUTE_CLASSID
+			nla = nla_find(attrs, attrlen, RTA_FLOW);
+			if (nla) {
+				if (nla_len(nla) < sizeof(u32)) {
+					NL_SET_ERR_MSG(extack, "Invalid RTA_FLOW");
+					return -EINVAL;
+				}
+				if (nla_get_u32(nla) != nh->nh_tclassid)
+					return 1;
+			}
+#endif
+		}
+
+		rtnh = rtnh_next(rtnh, &remaining);
+	} endfor_nexthops(fi);
+#endif
+	return 0;
+}
+
+bool fib_metrics_match(struct fib_config *cfg, struct fib_info *fi)
+{
+	struct nlattr *nla;
+	int remaining;
+
+	if (!cfg->fc_mx)
+		return true;
+
+	nla_for_each_attr(nla, cfg->fc_mx, cfg->fc_mx_len, remaining) {
+		int type = nla_type(nla);
+		u32 fi_val, val;
+
+		if (!type)
+			continue;
+		if (type > RTAX_MAX)
+			return false;
+
+		type = array_index_nospec(type, RTAX_MAX + 1);
+		if (type == RTAX_CC_ALGO) {
+			char tmp[TCP_CA_NAME_MAX];
+			bool ecn_ca = false;
+
+			nla_strscpy(tmp, nla, sizeof(tmp));
+			val = tcp_ca_get_key_by_name(tmp, &ecn_ca);
+		} else {
+			if (nla_len(nla) != sizeof(u32))
+				return false;
+			val = nla_get_u32(nla);
+		}
+
+		fi_val = fi->fib_metrics->metrics[type - 1];
+		if (type == RTAX_FEATURES)
+			fi_val &= ~DST_FEATURE_ECN_CA;
+
+		if (fi_val != val)
+			return false;
+	}
+
+	return true;
+}
+
+static int fib_check_nh_v6_gw(struct net *net, struct fib_nh *nh,
+			      u32 table, struct netlink_ext_ack *extack)
+{
+	struct fib6_config cfg = {
+		.fc_table = table,
+		.fc_flags = nh->fib_nh_flags | RTF_GATEWAY,
+		.fc_ifindex = nh->fib_nh_oif,
+		.fc_gateway = nh->fib_nh_gw6,
+	};
+	struct fib6_nh fib6_nh = {};
+	int err;
+
+	err = fib6_nh_init(net, &fib6_nh, &cfg, GFP_KERNEL, extack);
+	if (!err) {
+		nh->fib_nh_dev = fib6_nh.fib_nh_dev;
+		netdev_hold(nh->fib_nh_dev, &nh->fib_nh_dev_tracker,
+			    GFP_KERNEL);
+		nh->fib_nh_oif = nh->fib_nh_dev->ifindex;
+		nh->fib_nh_scope = RT_SCOPE_LINK;
+
+		fib6_nh_release(&fib6_nh);
+	}
+
+	return err;
+}
+
+/*
+ * Picture
+ * -------
+ *
+ * Semantics of nexthop is very messy by historical reasons.
+ * We have to take into account, that:
+ * a) gateway can be actually local interface address,
+ *    so that gatewayed route is direct.
+ * b) gateway must be on-link address, possibly
+ *    described not by an ifaddr, but also by a direct route.
+ * c) If both gateway and interface are specified, they should not
+ *    contradict.
+ * d) If we use tunnel routes, gateway could be not on-link.
+ *
+ * Attempt to reconcile all of these (alas, self-contradictory) conditions
+ * results in pretty ugly and hairy code with obscure logic.
+ *
+ * I chose to generalized it instead, so that the size
+ * of code does not increase practically, but it becomes
+ * much more general.
+ * Every prefix is assigned a "scope" value: "host" is local address,
+ * "link" is direct route,
+ * [ ... "site" ... "interior" ... ]
+ * and "universe" is true gateway route with global meaning.
+ *
+ * Every prefix refers to a set of "nexthop"s (gw, oif),
+ * where gw must have narrower scope. This recursion stops
+ * when gw has LOCAL scope or if "nexthop" is declared ONLINK,
+ * which means that gw is forced to be on link.
+ *
+ * Code is still hairy, but now it is apparently logically
+ * consistent and very flexible. F.e. as by-product it allows
+ * to co-exists in peace independent exterior and interior
+ * routing processes.
+ *
+ * Normally it looks as following.
+ *
+ * {universe prefix}  -> (gw, oif) [scope link]
+ *		  |
+ *		  |-> {link prefix} -> (gw, oif) [scope local]
+ *					|
+ *					|-> {local prefix} (terminal node)
+ */
+static int fib_check_nh_v4_gw(struct net *net, struct fib_nh *nh, u32 table,
+			      u8 scope, struct netlink_ext_ack *extack)
+{
+	struct net_device *dev;
+	struct fib_result res;
+	int err = 0;
+
+	if (nh->fib_nh_flags & RTNH_F_ONLINK) {
+		unsigned int addr_type;
+
+		if (scope >= RT_SCOPE_LINK) {
+			NL_SET_ERR_MSG(extack, "Nexthop has invalid scope");
+			return -EINVAL;
+		}
+		dev = __dev_get_by_index(net, nh->fib_nh_oif);
+		if (!dev) {
+			NL_SET_ERR_MSG(extack, "Nexthop device required for onlink");
+			return -ENODEV;
+		}
+		if (!(dev->flags & IFF_UP)) {
+			NL_SET_ERR_MSG(extack, "Nexthop device is not up");
+			return -ENETDOWN;
+		}
+		addr_type = inet_addr_type_dev_table(net, dev, nh->fib_nh_gw4);
+		if (addr_type != RTN_UNICAST) {
+			NL_SET_ERR_MSG(extack, "Nexthop has invalid gateway");
+			return -EINVAL;
+		}
+		if (!netif_carrier_ok(dev))
+			nh->fib_nh_flags |= RTNH_F_LINKDOWN;
+		nh->fib_nh_dev = dev;
+		netdev_hold(dev, &nh->fib_nh_dev_tracker, GFP_ATOMIC);
+		nh->fib_nh_scope = RT_SCOPE_LINK;
+		return 0;
+	}
+	rcu_read_lock();
+	{
+		struct fib_table *tbl = NULL;
+		struct flowi4 fl4 = {
+			.daddr = nh->fib_nh_gw4,
+			.flowi4_scope = scope + 1,
+			.flowi4_oif = nh->fib_nh_oif,
+			.flowi4_iif = LOOPBACK_IFINDEX,
+		};
+
+		/* It is not necessary, but requires a bit of thinking */
+		if (fl4.flowi4_scope < RT_SCOPE_LINK)
+			fl4.flowi4_scope = RT_SCOPE_LINK;
+
+		if (table && table != RT_TABLE_MAIN)
+			tbl = fib_get_table(net, table);
+
+		if (tbl)
+			err = fib_table_lookup(tbl, &fl4, &res,
+					       FIB_LOOKUP_IGNORE_LINKSTATE |
+					       FIB_LOOKUP_NOREF);
+
+		/* on error or if no table given do full lookup. This
+		 * is needed for example when nexthops are in the local
+		 * table rather than the given table
+		 */
+		if (!tbl || err) {
+			err = fib_lookup(net, &fl4, &res,
+					 FIB_LOOKUP_IGNORE_LINKSTATE);
+		}
+
+		if (err) {
+			NL_SET_ERR_MSG(extack, "Nexthop has invalid gateway");
+			goto out;
+		}
+	}
+
+	err = -EINVAL;
+	if (res.type != RTN_UNICAST && res.type != RTN_LOCAL) {
+		NL_SET_ERR_MSG(extack, "Nexthop has invalid gateway");
+		goto out;
+	}
+	nh->fib_nh_scope = res.scope;
+	nh->fib_nh_oif = FIB_RES_OIF(res);
+	nh->fib_nh_dev = dev = FIB_RES_DEV(res);
+	if (!dev) {
+		NL_SET_ERR_MSG(extack,
+			       "No egress device for nexthop gateway");
+		goto out;
+	}
+	netdev_hold(dev, &nh->fib_nh_dev_tracker, GFP_ATOMIC);
+	if (!netif_carrier_ok(dev))
+		nh->fib_nh_flags |= RTNH_F_LINKDOWN;
+	err = (dev->flags & IFF_UP) ? 0 : -ENETDOWN;
+out:
+	rcu_read_unlock();
+	return err;
+}
+
+static int fib_check_nh_nongw(struct net *net, struct fib_nh *nh,
+			      struct netlink_ext_ack *extack)
+{
+	struct in_device *in_dev;
+	int err;
+
+	if (nh->fib_nh_flags & (RTNH_F_PERVASIVE | RTNH_F_ONLINK)) {
+		NL_SET_ERR_MSG(extack,
+			       "Invalid flags for nexthop - PERVASIVE and ONLINK can not be set");
+		return -EINVAL;
+	}
+
+	rcu_read_lock();
+
+	err = -ENODEV;
+	in_dev = inetdev_by_index(net, nh->fib_nh_oif);
+	if (!in_dev)
+		goto out;
+	err = -ENETDOWN;
+	if (!(in_dev->dev->flags & IFF_UP)) {
+		NL_SET_ERR_MSG(extack, "Device for nexthop is not up");
+		goto out;
+	}
+
+	nh->fib_nh_dev = in_dev->dev;
+	netdev_hold(nh->fib_nh_dev, &nh->fib_nh_dev_tracker, GFP_ATOMIC);
+	nh->fib_nh_scope = RT_SCOPE_HOST;
+	if (!netif_carrier_ok(nh->fib_nh_dev))
+		nh->fib_nh_flags |= RTNH_F_LINKDOWN;
+	err = 0;
+out:
+	rcu_read_unlock();
+	return err;
+}
+
+int fib_check_nh(struct net *net, struct fib_nh *nh, u32 table, u8 scope,
+		 struct netlink_ext_ack *extack)
+{
+	int err;
+
+	if (nh->fib_nh_gw_family == AF_INET)
+		err = fib_check_nh_v4_gw(net, nh, table, scope, extack);
+	else if (nh->fib_nh_gw_family == AF_INET6)
+		err = fib_check_nh_v6_gw(net, nh, table, extack);
+	else
+		err = fib_check_nh_nongw(net, nh, extack);
+
+	return err;
+}
+
+__be32 fib_info_update_nhc_saddr(struct net *net, struct fib_nh_common *nhc,
+				 unsigned char scope)
+{
+	struct fib_nh *nh;
+	__be32 saddr;
+
+	if (nhc->nhc_family != AF_INET)
+		return inet_select_addr(nhc->nhc_dev, 0, scope);
+
+	nh = container_of(nhc, struct fib_nh, nh_common);
+	saddr = inet_select_addr(nh->fib_nh_dev, nh->fib_nh_gw4, scope);
+
+	WRITE_ONCE(nh->nh_saddr, saddr);
+	WRITE_ONCE(nh->nh_saddr_genid, atomic_read(&net->ipv4.dev_addr_genid));
+
+	return saddr;
+}
+
+__be32 fib_result_prefsrc(struct net *net, struct fib_result *res)
+{
+	struct fib_nh_common *nhc = res->nhc;
+
+	if (res->fi->fib_prefsrc)
+		return res->fi->fib_prefsrc;
+
+	if (nhc->nhc_family == AF_INET) {
+		struct fib_nh *nh;
+
+		nh = container_of(nhc, struct fib_nh, nh_common);
+		if (READ_ONCE(nh->nh_saddr_genid) ==
+		    atomic_read(&net->ipv4.dev_addr_genid))
+			return READ_ONCE(nh->nh_saddr);
+	}
+
+	return fib_info_update_nhc_saddr(net, nhc, res->fi->fib_scope);
+}
+
+static bool fib_valid_prefsrc(struct fib_config *cfg, __be32 fib_prefsrc)
+{
+	if (cfg->fc_type != RTN_LOCAL || !cfg->fc_dst ||
+	    fib_prefsrc != cfg->fc_dst) {
+		u32 tb_id = cfg->fc_table;
+		int rc;
+
+		if (tb_id == RT_TABLE_MAIN)
+			tb_id = RT_TABLE_LOCAL;
+
+		rc = inet_addr_type_table(cfg->fc_nlinfo.nl_net,
+					  fib_prefsrc, tb_id);
+
+		if (rc != RTN_LOCAL && tb_id != RT_TABLE_LOCAL) {
+			rc = inet_addr_type_table(cfg->fc_nlinfo.nl_net,
+						  fib_prefsrc, RT_TABLE_LOCAL);
+		}
+
+		if (rc != RTN_LOCAL)
+			return false;
+	}
+	return true;
+}
+
+struct fib_info *fib_create_info(struct fib_config *cfg,
+				 struct netlink_ext_ack *extack)
+{
+	int err;
+	struct fib_info *fi = NULL;
+	struct nexthop *nh = NULL;
+	struct fib_info *ofi;
+	int nhs = 1;
+	struct net *net = cfg->fc_nlinfo.nl_net;
+
+	ASSERT_RTNL();
+	if (cfg->fc_type > RTN_MAX)
+		goto err_inval;
+
+	/* Fast check to catch the most weird cases */
+	if (fib_props[cfg->fc_type].scope > cfg->fc_scope) {
+		NL_SET_ERR_MSG(extack, "Invalid scope");
+		goto err_inval;
+	}
+
+	if (cfg->fc_flags & (RTNH_F_DEAD | RTNH_F_LINKDOWN)) {
+		NL_SET_ERR_MSG(extack,
+			       "Invalid rtm_flags - can not contain DEAD or LINKDOWN");
+		goto err_inval;
+	}
+
+	if (cfg->fc_nh_id) {
+		if (!cfg->fc_mx) {
+			fi = fib_find_info_nh(net, cfg);
+			if (fi) {
+				refcount_inc(&fi->fib_treeref);
+				return fi;
+			}
+		}
+
+		nh = nexthop_find_by_id(net, cfg->fc_nh_id);
+		if (!nh) {
+			NL_SET_ERR_MSG(extack, "Nexthop id does not exist");
+			goto err_inval;
+		}
+		nhs = 0;
+	}
+
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+	if (cfg->fc_mp) {
+		nhs = fib_count_nexthops(cfg->fc_mp, cfg->fc_mp_len, extack);
+		if (nhs == 0)
+			goto err_inval;
+	}
+#endif
+
+	fib_info_hash_grow(net);
+
+	fi = kzalloc_flex(*fi, fib_nh, nhs);
+	if (!fi) {
+		err = -ENOBUFS;
+		goto failure;
+	}
+
+	fi->fib_metrics = ip_fib_metrics_init(cfg->fc_mx, cfg->fc_mx_len, extack);
+	if (IS_ERR(fi->fib_metrics)) {
+		err = PTR_ERR(fi->fib_metrics);
+		kfree(fi);
+		return ERR_PTR(err);
+	}
+
+	fi->fib_net = net;
+	fi->fib_protocol = cfg->fc_protocol;
+	fi->fib_scope = cfg->fc_scope;
+	fi->fib_flags = cfg->fc_flags;
+	fi->fib_priority = cfg->fc_priority;
+	fi->fib_prefsrc = cfg->fc_prefsrc;
+	fi->fib_type = cfg->fc_type;
+	fi->fib_tb_id = cfg->fc_table;
+
+	fi->fib_nhs = nhs;
+	if (nh) {
+		if (!nexthop_get(nh)) {
+			NL_SET_ERR_MSG(extack, "Nexthop has been deleted");
+			err = -EINVAL;
+		} else {
+			err = 0;
+			fi->nh = nh;
+		}
+	} else {
+		change_nexthops(fi) {
+			nexthop_nh->nh_parent = fi;
+		} endfor_nexthops(fi)
+
+		if (cfg->fc_mp)
+			err = fib_get_nhs(fi, cfg->fc_mp, cfg->fc_mp_len, cfg,
+					  extack);
+		else
+			err = fib_nh_init(net, fi->fib_nh, cfg, 1, extack);
+	}
+
+	if (err != 0)
+		goto failure;
+
+	if (fib_props[cfg->fc_type].error) {
+		if (cfg->fc_gw_family || cfg->fc_oif || cfg->fc_mp) {
+			NL_SET_ERR_MSG(extack,
+				       "Gateway, device and multipath can not be specified for this route type");
+			goto err_inval;
+		}
+		goto link_it;
+	} else {
+		switch (cfg->fc_type) {
+		case RTN_UNICAST:
+		case RTN_LOCAL:
+		case RTN_BROADCAST:
+		case RTN_ANYCAST:
+		case RTN_MULTICAST:
+			break;
+		default:
+			NL_SET_ERR_MSG(extack, "Invalid route type");
+			goto err_inval;
+		}
+	}
+
+	if (cfg->fc_scope > RT_SCOPE_HOST) {
+		NL_SET_ERR_MSG(extack, "Invalid scope");
+		goto err_inval;
+	}
+
+	if (fi->nh) {
+		err = fib_check_nexthop(fi->nh, cfg->fc_scope, extack);
+		if (err)
+			goto failure;
+	} else if (cfg->fc_scope == RT_SCOPE_HOST) {
+		struct fib_nh *nh = fi->fib_nh;
+
+		/* Local address is added. */
+		if (nhs != 1) {
+			NL_SET_ERR_MSG(extack,
+				       "Route with host scope can not have multiple nexthops");
+			goto err_inval;
+		}
+		if (nh->fib_nh_gw_family) {
+			NL_SET_ERR_MSG(extack,
+				       "Route with host scope can not have a gateway");
+			goto err_inval;
+		}
+		nh->fib_nh_scope = RT_SCOPE_NOWHERE;
+		nh->fib_nh_dev = dev_get_by_index(net, nh->fib_nh_oif);
+		err = -ENODEV;
+		if (!nh->fib_nh_dev)
+			goto failure;
+		netdev_tracker_alloc(nh->fib_nh_dev, &nh->fib_nh_dev_tracker,
+				     GFP_KERNEL);
+	} else {
+		int linkdown = 0;
+
+		change_nexthops(fi) {
+			err = fib_check_nh(cfg->fc_nlinfo.nl_net, nexthop_nh,
+					   cfg->fc_table, cfg->fc_scope,
+					   extack);
+			if (err != 0)
+				goto failure;
+			if (nexthop_nh->fib_nh_flags & RTNH_F_LINKDOWN)
+				linkdown++;
+		} endfor_nexthops(fi)
+		if (linkdown == fi->fib_nhs)
+			fi->fib_flags |= RTNH_F_LINKDOWN;
+	}
+
+	if (fi->fib_prefsrc && !fib_valid_prefsrc(cfg, fi->fib_prefsrc)) {
+		NL_SET_ERR_MSG(extack, "Invalid prefsrc address");
+		goto err_inval;
+	}
+
+	if (!fi->nh) {
+		change_nexthops(fi) {
+			fib_info_update_nhc_saddr(net, &nexthop_nh->nh_common,
+						  fi->fib_scope);
+			if (nexthop_nh->fib_nh_gw_family == AF_INET6)
+				fi->fib_nh_is_v6 = true;
+		} endfor_nexthops(fi)
+
+		fib_rebalance(fi);
+	}
+
+link_it:
+	ofi = fib_find_info(fi);
+	if (ofi) {
+		/* fib_table_lookup() should not see @fi yet. */
+		fi->fib_dead = 1;
+		free_fib_info(fi);
+		refcount_inc(&ofi->fib_treeref);
+		return ofi;
+	}
+
+	refcount_set(&fi->fib_treeref, 1);
+	refcount_set(&fi->fib_clntref, 1);
+
+	net->ipv4.fib_info_cnt++;
+	hlist_add_head(&fi->fib_hash, fib_info_hash_bucket(fi));
+
+	if (fi->fib_prefsrc) {
+		struct hlist_head *head;
+
+		head = fib_info_laddrhash_bucket(net, fi->fib_prefsrc);
+		hlist_add_head(&fi->fib_lhash, head);
+	}
+	if (fi->nh) {
+		list_add(&fi->nh_list, &nh->fi_list);
+	} else {
+		change_nexthops(fi) {
+			struct hlist_head *head;
+
+			if (!nexthop_nh->fib_nh_dev)
+				continue;
+			head = fib_nh_head(nexthop_nh->fib_nh_dev);
+			hlist_add_head_rcu(&nexthop_nh->nh_hash, head);
+		} endfor_nexthops(fi)
+	}
+	return fi;
+
+err_inval:
+	err = -EINVAL;
+
+failure:
+	if (fi) {
+		/* fib_table_lookup() should not see @fi yet. */
+		fi->fib_dead = 1;
+		free_fib_info(fi);
+	}
+
+	return ERR_PTR(err);
+}
+
+int fib_nexthop_info(struct sk_buff *skb, const struct fib_nh_common *nhc,
+		     u8 rt_family, unsigned char *flags, bool skip_oif)
+{
+	if (nhc->nhc_flags & RTNH_F_DEAD)
+		*flags |= RTNH_F_DEAD;
+
+	if (nhc->nhc_flags & RTNH_F_LINKDOWN) {
+		*flags |= RTNH_F_LINKDOWN;
+
+		rcu_read_lock();
+		switch (nhc->nhc_family) {
+		case AF_INET:
+			if (ip_ignore_linkdown(nhc->nhc_dev))
+				*flags |= RTNH_F_DEAD;
+			break;
+		case AF_INET6:
+			if (ip6_ignore_linkdown(nhc->nhc_dev))
+				*flags |= RTNH_F_DEAD;
+			break;
+		}
+		rcu_read_unlock();
+	}
+
+	switch (nhc->nhc_gw_family) {
+	case AF_INET:
+		if (nla_put_in_addr(skb, RTA_GATEWAY, nhc->nhc_gw.ipv4))
+			goto nla_put_failure;
+		break;
+	case AF_INET6:
+		/* if gateway family does not match nexthop family
+		 * gateway is encoded as RTA_VIA
+		 */
+		if (rt_family != nhc->nhc_gw_family) {
+			int alen = sizeof(struct in6_addr);
+			struct nlattr *nla;
+			struct rtvia *via;
+
+			nla = nla_reserve(skb, RTA_VIA, alen + 2);
+			if (!nla)
+				goto nla_put_failure;
+
+			via = nla_data(nla);
+			via->rtvia_family = AF_INET6;
+			memcpy(via->rtvia_addr, &nhc->nhc_gw.ipv6, alen);
+		} else if (nla_put_in6_addr(skb, RTA_GATEWAY,
+					    &nhc->nhc_gw.ipv6) < 0) {
+			goto nla_put_failure;
+		}
+		break;
+	}
+
+	*flags |= (nhc->nhc_flags &
+		   (RTNH_F_ONLINK | RTNH_F_OFFLOAD | RTNH_F_TRAP));
+
+	if (!skip_oif && nhc->nhc_dev &&
+	    nla_put_u32(skb, RTA_OIF, nhc->nhc_dev->ifindex))
+		goto nla_put_failure;
+
+	if (lwtunnel_fill_encap(skb, nhc->nhc_lwtstate,
+				RTA_ENCAP, RTA_ENCAP_TYPE) < 0)
+		goto nla_put_failure;
+
+	return 0;
+
+nla_put_failure:
+	return -EMSGSIZE;
+}
+
+#if IS_ENABLED(CONFIG_IP_ROUTE_MULTIPATH) || IS_ENABLED(CONFIG_IPV6)
+int fib_add_nexthop(struct sk_buff *skb, const struct fib_nh_common *nhc,
+		    int nh_weight, u8 rt_family, u32 nh_tclassid)
+{
+	const struct net_device *dev = nhc->nhc_dev;
+	struct rtnexthop *rtnh;
+	unsigned char flags = 0;
+
+	rtnh = nla_reserve_nohdr(skb, sizeof(*rtnh));
+	if (!rtnh)
+		goto nla_put_failure;
+
+	rtnh->rtnh_hops = nh_weight - 1;
+	rtnh->rtnh_ifindex = dev ? dev->ifindex : 0;
+
+	if (fib_nexthop_info(skb, nhc, rt_family, &flags, true) < 0)
+		goto nla_put_failure;
+
+	rtnh->rtnh_flags = flags;
+
+	if (nh_tclassid && nla_put_u32(skb, RTA_FLOW, nh_tclassid))
+		goto nla_put_failure;
+
+	/* length of rtnetlink header + attributes */
+	rtnh->rtnh_len = nlmsg_get_pos(skb) - (void *)rtnh;
+
+	return 0;
+
+nla_put_failure:
+	return -EMSGSIZE;
+}
+#endif
+
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+static int fib_add_multipath(struct sk_buff *skb, struct fib_info *fi)
+{
+	struct nlattr *mp;
+
+	mp = nla_nest_start_noflag(skb, RTA_MULTIPATH);
+	if (!mp)
+		goto nla_put_failure;
+
+	if (unlikely(fi->nh)) {
+		if (nexthop_mpath_fill_node(skb, fi->nh, AF_INET) < 0)
+			goto nla_put_failure;
+		goto mp_end;
+	}
+
+	for_nexthops(fi) {
+		u32 nh_tclassid = 0;
+#ifdef CONFIG_IP_ROUTE_CLASSID
+		nh_tclassid = nh->nh_tclassid;
+#endif
+		if (fib_add_nexthop(skb, &nh->nh_common, nh->fib_nh_weight,
+				    AF_INET, nh_tclassid) < 0)
+			goto nla_put_failure;
+	} endfor_nexthops(fi);
+
+mp_end:
+	nla_nest_end(skb, mp);
+
+	return 0;
+
+nla_put_failure:
+	return -EMSGSIZE;
+}
+#else
+static int fib_add_multipath(struct sk_buff *skb, struct fib_info *fi)
+{
+	return 0;
+}
+#endif
+
+int fib_dump_info(struct sk_buff *skb, u32 portid, u32 seq, int event,
+		  const struct fib_rt_info *fri, unsigned int flags)
+{
+	unsigned int nhs = fib_info_num_path(fri->fi);
+	struct fib_info *fi = fri->fi;
+	u32 tb_id = fri->tb_id;
+	struct nlmsghdr *nlh;
+	struct rtmsg *rtm;
+
+	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*rtm), flags);
+	if (!nlh)
+		return -EMSGSIZE;
+
+	rtm = nlmsg_data(nlh);
+	rtm->rtm_family = AF_INET;
+	rtm->rtm_dst_len = fri->dst_len;
+	rtm->rtm_src_len = 0;
+	rtm->rtm_tos = inet_dscp_to_dsfield(fri->dscp);
+	if (tb_id < 256)
+		rtm->rtm_table = tb_id;
+	else
+		rtm->rtm_table = RT_TABLE_COMPAT;
+	if (nla_put_u32(skb, RTA_TABLE, tb_id))
+		goto nla_put_failure;
+	rtm->rtm_type = fri->type;
+	rtm->rtm_flags = fi->fib_flags;
+	rtm->rtm_scope = fi->fib_scope;
+	rtm->rtm_protocol = fi->fib_protocol;
+
+	if (rtm->rtm_dst_len &&
+	    nla_put_in_addr(skb, RTA_DST, fri->dst))
+		goto nla_put_failure;
+	if (fi->fib_priority &&
+	    nla_put_u32(skb, RTA_PRIORITY, fi->fib_priority))
+		goto nla_put_failure;
+	if (rtnetlink_put_metrics(skb, fi->fib_metrics->metrics) < 0)
+		goto nla_put_failure;
+
+	if (fi->fib_prefsrc &&
+	    nla_put_in_addr(skb, RTA_PREFSRC, fi->fib_prefsrc))
+		goto nla_put_failure;
+
+	if (fi->nh) {
+		if (nla_put_u32(skb, RTA_NH_ID, fi->nh->id))
+			goto nla_put_failure;
+		if (nexthop_is_blackhole(fi->nh))
+			rtm->rtm_type = RTN_BLACKHOLE;
+		if (!READ_ONCE(fi->fib_net->ipv4.sysctl_nexthop_compat_mode))
+			goto offload;
+	}
+
+	if (nhs == 1) {
+		const struct fib_nh_common *nhc = fib_info_nhc(fi, 0);
+		unsigned char flags = 0;
+
+		if (fib_nexthop_info(skb, nhc, AF_INET, &flags, false) < 0)
+			goto nla_put_failure;
+
+		rtm->rtm_flags = flags;
+#ifdef CONFIG_IP_ROUTE_CLASSID
+		if (nhc->nhc_family == AF_INET) {
+			struct fib_nh *nh;
+
+			nh = container_of(nhc, struct fib_nh, nh_common);
+			if (nh->nh_tclassid &&
+			    nla_put_u32(skb, RTA_FLOW, nh->nh_tclassid))
+				goto nla_put_failure;
+		}
+#endif
+	} else {
+		if (fib_add_multipath(skb, fi) < 0)
+			goto nla_put_failure;
+	}
+
+offload:
+	if (fri->offload)
+		rtm->rtm_flags |= RTM_F_OFFLOAD;
+	if (fri->trap)
+		rtm->rtm_flags |= RTM_F_TRAP;
+	if (fri->offload_failed)
+		rtm->rtm_flags |= RTM_F_OFFLOAD_FAILED;
+
+	nlmsg_end(skb, nlh);
+	return 0;
+
+nla_put_failure:
+	nlmsg_cancel(skb, nlh);
+	return -EMSGSIZE;
+}
+
+/*
+ * Update FIB if:
+ * - local address disappeared -> we must delete all the entries
+ *   referring to it.
+ * - device went down -> we must shutdown all nexthops going via it.
+ */
+int fib_sync_down_addr(struct net_device *dev, __be32 local)
+{
+	int tb_id = l3mdev_fib_table(dev) ? : RT_TABLE_MAIN;
+	struct net *net = dev_net(dev);
+	struct hlist_head *head;
+	struct fib_info *fi;
+	int ret = 0;
+
+	if (!local)
+		return 0;
+
+	head = fib_info_laddrhash_bucket(net, local);
+	hlist_for_each_entry(fi, head, fib_lhash) {
+		if (!net_eq(fi->fib_net, net) ||
+		    fi->fib_tb_id != tb_id)
+			continue;
+		if (fi->fib_prefsrc == local) {
+			fi->fib_flags |= RTNH_F_DEAD;
+			fi->pfsrc_removed = true;
+			ret++;
+		}
+	}
+	return ret;
+}
+
+static int call_fib_nh_notifiers(struct fib_nh *nh,
+				 enum fib_event_type event_type)
+{
+	bool ignore_link_down = ip_ignore_linkdown(nh->fib_nh_dev);
+	struct fib_nh_notifier_info info = {
+		.fib_nh = nh,
+	};
+
+	switch (event_type) {
+	case FIB_EVENT_NH_ADD:
+		if (nh->fib_nh_flags & RTNH_F_DEAD)
+			break;
+		if (ignore_link_down && nh->fib_nh_flags & RTNH_F_LINKDOWN)
+			break;
+		return call_fib4_notifiers(dev_net(nh->fib_nh_dev), event_type,
+					   &info.info);
+	case FIB_EVENT_NH_DEL:
+		if ((ignore_link_down && nh->fib_nh_flags & RTNH_F_LINKDOWN) ||
+		    (nh->fib_nh_flags & RTNH_F_DEAD))
+			return call_fib4_notifiers(dev_net(nh->fib_nh_dev),
+						   event_type, &info.info);
+		break;
+	default:
+		break;
+	}
+
+	return NOTIFY_DONE;
+}
+
+/* Walk the exceptions of a nexthop after its first hop MTU changed. The
+ * chain is RCU protected here, while fnhe_update_pmtu() takes fnhe_lock
+ * for the update of each entry.
+ */
+void fib_nhc_update_mtu(struct fib_nh_common *nhc, u32 new, u32 orig)
+{
+	struct fnhe_hash_bucket *bucket;
+	int i;
+
+	rcu_read_lock();
+	bucket = rcu_dereference(nhc->nhc_exceptions);
+	if (!bucket)
+		goto out;
+
+	for (i = 0; i < FNHE_HASH_SIZE; i++) {
+		struct fib_nh_exception *fnhe;
+
+		for (fnhe = rcu_dereference(bucket[i].chain);
+		     fnhe;
+		     fnhe = rcu_dereference(fnhe->fnhe_next))
+			fnhe_update_pmtu(fnhe, new, orig);
+	}
+out:
+	rcu_read_unlock();
+}
+
+void fib_sync_mtu(struct net_device *dev, u32 orig_mtu)
+{
+	struct hlist_head *head = fib_nh_head(dev);
+	struct fib_nh *nh;
+
+	hlist_for_each_entry(nh, head, nh_hash) {
+		DEBUG_NET_WARN_ON_ONCE(nh->fib_nh_dev != dev);
+		fib_nhc_update_mtu(&nh->nh_common, dev->mtu, orig_mtu);
+	}
+}
+
+/* Event              force Flags           Description
+ * NETDEV_CHANGE      0     LINKDOWN        Carrier OFF, not for scope host
+ * NETDEV_DOWN        0     LINKDOWN|DEAD   Link down, not for scope host
+ * NETDEV_DOWN        1     LINKDOWN|DEAD   Last address removed
+ * NETDEV_UNREGISTER  1     LINKDOWN|DEAD   Device removed
+ *
+ * only used when fib_nh is built into fib_info
+ */
+int fib_sync_down_dev(struct net_device *dev, unsigned long event, bool force)
+{
+	struct hlist_head *head = fib_nh_head(dev);
+	struct fib_info *prev_fi = NULL;
+	int scope = RT_SCOPE_NOWHERE;
+	struct fib_nh *nh;
+	int ret = 0;
+
+	if (force)
+		scope = -1;
+
+	hlist_for_each_entry(nh, head, nh_hash) {
+		struct fib_info *fi = nh->nh_parent;
+		int dead;
+
+		BUG_ON(!fi->fib_nhs);
+		DEBUG_NET_WARN_ON_ONCE(nh->fib_nh_dev != dev);
+		if (fi == prev_fi)
+			continue;
+		prev_fi = fi;
+		dead = 0;
+		change_nexthops(fi) {
+			if (nexthop_nh->fib_nh_flags & RTNH_F_DEAD)
+				dead++;
+			else if (nexthop_nh->fib_nh_dev == dev &&
+				 nexthop_nh->fib_nh_scope != scope) {
+				switch (event) {
+				case NETDEV_DOWN:
+				case NETDEV_UNREGISTER:
+					nexthop_nh->fib_nh_flags |= RTNH_F_DEAD;
+					fallthrough;
+				case NETDEV_CHANGE:
+					nexthop_nh->fib_nh_flags |= RTNH_F_LINKDOWN;
+					break;
+				}
+				call_fib_nh_notifiers(nexthop_nh,
+						      FIB_EVENT_NH_DEL);
+				dead++;
+			}
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+			if (event == NETDEV_UNREGISTER &&
+			    nexthop_nh->fib_nh_dev == dev) {
+				dead = fi->fib_nhs;
+				break;
+			}
+#endif
+		} endfor_nexthops(fi)
+		if (dead == fi->fib_nhs) {
+			switch (event) {
+			case NETDEV_DOWN:
+			case NETDEV_UNREGISTER:
+				fi->fib_flags |= RTNH_F_DEAD;
+				fallthrough;
+			case NETDEV_CHANGE:
+				fi->fib_flags |= RTNH_F_LINKDOWN;
+				break;
+			}
+			ret++;
+		}
+
+		fib_rebalance(fi);
+	}
+
+	return ret;
+}
+
+/* Must be invoked inside of an RCU protected region.  */
+static void fib_select_default(const struct flowi4 *flp, struct fib_result *res)
+{
+	struct fib_info *fi = NULL, *last_resort = NULL;
+	struct hlist_head *fa_head = res->fa_head;
+	struct fib_table *tb = res->table;
+	u8 slen = 32 - res->prefixlen;
+	int order = -1, last_idx = -1;
+	struct fib_alias *fa, *fa1 = NULL;
+	u32 last_prio = res->fi->fib_priority;
+	dscp_t last_dscp = 0;
+
+	hlist_for_each_entry_rcu(fa, fa_head, fa_list) {
+		struct fib_info *next_fi = fa->fa_info;
+		struct fib_nh_common *nhc;
+
+		if (fa->fa_slen != slen)
+			continue;
+		if (fa->fa_dscp && !fib_dscp_masked_match(fa->fa_dscp, flp))
+			continue;
+		if (fa->tb_id != tb->tb_id)
+			continue;
+		if (next_fi->fib_priority > last_prio &&
+		    fa->fa_dscp == last_dscp) {
+			if (last_dscp)
+				continue;
+			break;
+		}
+		if (next_fi->fib_flags & RTNH_F_DEAD)
+			continue;
+		last_dscp = fa->fa_dscp;
+		last_prio = next_fi->fib_priority;
+
+		if (next_fi->fib_scope != res->scope ||
+		    fa->fa_type != RTN_UNICAST)
+			continue;
+
+		nhc = fib_info_nhc(next_fi, 0);
+		if (!nhc->nhc_gw_family || nhc->nhc_scope != RT_SCOPE_LINK)
+			continue;
+
+		fib_alias_accessed(fa);
+
+		if (!fi) {
+			if (next_fi != res->fi)
+				break;
+			fa1 = fa;
+		} else if (!fib_detect_death(fi, order, &last_resort,
+					     &last_idx, fa1->fa_default)) {
+			fib_result_assign(res, fi);
+			fa1->fa_default = order;
+			goto out;
+		}
+		fi = next_fi;
+		order++;
+	}
+
+	if (order <= 0 || !fi) {
+		if (fa1)
+			fa1->fa_default = -1;
+		goto out;
+	}
+
+	if (!fib_detect_death(fi, order, &last_resort, &last_idx,
+			      fa1->fa_default)) {
+		fib_result_assign(res, fi);
+		fa1->fa_default = order;
+		goto out;
+	}
+
+	if (last_idx >= 0)
+		fib_result_assign(res, last_resort);
+	fa1->fa_default = last_idx;
+out:
+	return;
+}
+
+/*
+ * Dead device goes up. We wake up dead nexthops.
+ * It takes sense only on multipath routes.
+ *
+ * only used when fib_nh is built into fib_info
+ */
+int fib_sync_up(struct net_device *dev, unsigned char nh_flags)
+{
+	struct fib_info *prev_fi;
+	struct hlist_head *head;
+	struct fib_nh *nh;
+	int ret;
+
+	if (!(dev->flags & IFF_UP))
+		return 0;
+
+	if (nh_flags & RTNH_F_DEAD) {
+		unsigned int flags = netif_get_flags(dev);
+
+		if (flags & (IFF_RUNNING | IFF_LOWER_UP))
+			nh_flags |= RTNH_F_LINKDOWN;
+	}
+
+	prev_fi = NULL;
+	head = fib_nh_head(dev);
+	ret = 0;
+
+	hlist_for_each_entry(nh, head, nh_hash) {
+		struct fib_info *fi = nh->nh_parent;
+		int alive;
+
+		BUG_ON(!fi->fib_nhs);
+		DEBUG_NET_WARN_ON_ONCE(nh->fib_nh_dev != dev);
+		if (fi == prev_fi)
+			continue;
+
+		prev_fi = fi;
+		alive = 0;
+		change_nexthops(fi) {
+			if (!(nexthop_nh->fib_nh_flags & nh_flags)) {
+				alive++;
+				continue;
+			}
+			if (!nexthop_nh->fib_nh_dev ||
+			    !(nexthop_nh->fib_nh_dev->flags & IFF_UP))
+				continue;
+			if (nexthop_nh->fib_nh_dev != dev ||
+			    !__in_dev_get_rtnl(dev))
+				continue;
+			alive++;
+			nexthop_nh->fib_nh_flags &= ~nh_flags;
+			call_fib_nh_notifiers(nexthop_nh, FIB_EVENT_NH_ADD);
+		} endfor_nexthops(fi)
+
+		if (alive > 0) {
+			fi->fib_flags &= ~nh_flags;
+			ret++;
+		}
+
+		fib_rebalance(fi);
+	}
+
+	return ret;
+}
+
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+static bool fib_good_nh(const struct fib_nh *nh)
+{
+	int state = NUD_REACHABLE;
+
+	if (nh->fib_nh_scope == RT_SCOPE_LINK) {
+		struct neighbour *n;
+
+		rcu_read_lock();
+
+		if (likely(nh->fib_nh_gw_family == AF_INET))
+			n = __ipv4_neigh_lookup_noref(nh->fib_nh_dev,
+						   (__force u32)nh->fib_nh_gw4);
+		else if (IS_ENABLED(CONFIG_IPV6) &&
+			 nh->fib_nh_gw_family == AF_INET6)
+			n = __ipv6_neigh_lookup_noref(nh->fib_nh_dev,
+						      &nh->fib_nh_gw6);
+		else
+			n = NULL;
+		if (n)
+			state = READ_ONCE(n->nud_state);
+
+		rcu_read_unlock();
+	}
+
+	return !!(state & NUD_VALID);
+}
+
+void fib_select_multipath(struct fib_result *res, int hash,
+			  const struct flowi4 *fl4)
+{
+	struct fib_info *fi = res->fi;
+	struct net *net = fi->fib_net;
+	bool use_neigh;
+	int score = -1;
+	__be32 saddr;
+
+	if (unlikely(res->fi->nh)) {
+		nexthop_path_fib_result(res, hash);
+		return;
+	}
+
+	use_neigh = READ_ONCE(net->ipv4.sysctl_fib_multipath_use_neigh);
+	saddr = fl4 ? fl4->saddr : 0;
+
+	change_nexthops(fi) {
+		int nh_upper_bound, nh_score = 0;
+
+		/* Nexthops without a carrier are assigned an upper bound of
+		 * minus one when "ignore_routes_with_linkdown" is set.
+		 */
+		nh_upper_bound = atomic_read(&nexthop_nh->fib_nh_upper_bound);
+		if (nh_upper_bound == -1 ||
+		    (use_neigh && !fib_good_nh(nexthop_nh)))
+			continue;
+
+		if (saddr && nexthop_nh->nh_saddr == saddr)
+			nh_score += 2;
+		if (hash <= nh_upper_bound)
+			nh_score++;
+		if (score < nh_score) {
+			res->nh_sel = nhsel;
+			res->nhc = &nexthop_nh->nh_common;
+			if (nh_score == 3 || (!saddr && nh_score == 1))
+				return;
+			score = nh_score;
+		}
+
+	} endfor_nexthops(fi);
+}
+#endif
+
+void fib_select_path(struct net *net, struct fib_result *res,
+		     struct flowi4 *fl4, const struct sk_buff *skb)
+{
+	if (fl4->flowi4_oif)
+		goto check_saddr;
+
+#ifdef CONFIG_IP_ROUTE_MULTIPATH
+	if (fib_info_num_path(res->fi) > 1) {
+		int h = fib_multipath_hash(net, fl4, skb, NULL);
+
+		fib_select_multipath(res, h, fl4);
+	}
+	else
+#endif
+	if (!res->prefixlen &&
+	    res->table->tb_num_default > 1 &&
+	    res->type == RTN_UNICAST)
+		fib_select_default(fl4, res);
+
+check_saddr:
+	if (!fl4->saddr) {
+		struct net_device *l3mdev;
+
+		l3mdev = dev_get_by_index_rcu(net, fl4->flowi4_l3mdev);
+
+		if (!l3mdev ||
+		    l3mdev_master_dev_rcu(FIB_RES_DEV(*res)) == l3mdev)
+			fl4->saddr = fib_result_prefsrc(net, res);
+		else
+			fl4->saddr = inet_select_addr(l3mdev, 0, RT_SCOPE_LINK);
+	}
+}
+
+int __net_init fib4_semantics_init(struct net *net)
+{
+	unsigned int hash_bits = 4;
+
+	net->ipv4.fib_info_hash = fib_info_hash_alloc(hash_bits);
+	if (!net->ipv4.fib_info_hash)
+		return -ENOMEM;
+
+	net->ipv4.fib_info_hash_bits = hash_bits;
+	net->ipv4.fib_info_cnt = 0;
+
+	return 0;
+}
+
+void __net_exit fib4_semantics_exit(struct net *net)
+{
+	fib_info_hash_free(net->ipv4.fib_info_hash);
+}