[PATCH 5.10 53/82] net: do not delay dst_entries_add() in dst_release()

[Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>]

5.10-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Eric Dumazet <edumazet@xxxxxxxxxx>

commit ac888d58869bb99753e7652be19a151df9ecb35d upstream.

dst_entries_add() uses per-cpu data that might be freed at netns
dismantle from ip6_route_net_exit() calling dst_entries_destroy()

Before ip6_route_net_exit() can be called, we release all
the dsts associated with this netns, via calls to dst_release(),
which waits an rcu grace period before calling dst_destroy()

dst_entries_add() use in dst_destroy() is racy, because
dst_entries_destroy() could have been called already.

Decrementing the number of dsts must happen sooner.

Notes:

1) in CONFIG_XFRM case, dst_destroy() can call
   dst_release_immediate(child), this might also cause UAF
   if the child does not have DST_NOCOUNT set.
   IPSEC maintainers might take a look and see how to address this.

2) There is also discussion about removing this count of dst,
   which might happen in future kernels.

Fixes: f88649721268 ("ipv4: fix dst race in sk_dst_get()")
Closes: https://lore.kernel.org/lkml/CANn89iLCCGsP7SFn9HKpvnKu96Td4KD08xf7aGtiYgZnkjaL=w@xxxxxxxxxxxxxx/T/
Reported-by: Naresh Kamboju <naresh.kamboju@xxxxxxxxxx>
Tested-by: Linux Kernel Functional Testing <lkft@xxxxxxxxxx>
Tested-by: Naresh Kamboju <naresh.kamboju@xxxxxxxxxx>
Signed-off-by: Eric Dumazet <edumazet@xxxxxxxxxx>
Cc: Xin Long <lucien.xin@xxxxxxxxx>
Cc: Steffen Klassert <steffen.klassert@xxxxxxxxxxx>
Reviewed-by: Xin Long <lucien.xin@xxxxxxxxx>
Link: https://patch.msgid.link/20241008143110.1064899-1-edumazet@xxxxxxxxxx
Signed-off-by: Paolo Abeni <pabeni@xxxxxxxxxx>
[ resolved conflict due to bc9d3a9f2afc ("net: dst: Switch to rcuref_t
  reference counting") is not in the tree ]
Signed-off-by: Abdelkareem Abdelsaamad <kareemem@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
 net/core/dst.c |   17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

--- a/net/core/dst.c
+++ b/net/core/dst.c
@@ -109,9 +109,6 @@ struct dst_entry *dst_destroy(struct dst
 		child = xdst->child;
 	}
 #endif
-	if (!(dst->flags & DST_NOCOUNT))
-		dst_entries_add(dst->ops, -1);
-
 	if (dst->ops->destroy)
 		dst->ops->destroy(dst);
 	if (dst->dev)
@@ -162,6 +159,12 @@ void dst_dev_put(struct dst_entry *dst)
 }
 EXPORT_SYMBOL(dst_dev_put);
 
+static void dst_count_dec(struct dst_entry *dst)
+{
+	if (!(dst->flags & DST_NOCOUNT))
+		dst_entries_add(dst->ops, -1);
+}
+
 void dst_release(struct dst_entry *dst)
 {
 	if (dst) {
@@ -171,8 +174,10 @@ void dst_release(struct dst_entry *dst)
 		if (WARN_ONCE(newrefcnt < 0, "dst_release underflow"))
 			net_warn_ratelimited("%s: dst:%p refcnt:%d\n",
 					     __func__, dst, newrefcnt);
-		if (!newrefcnt)
+		if (!newrefcnt){
+			dst_count_dec(dst);
 			call_rcu(&dst->rcu_head, dst_destroy_rcu);
+		}
 	}
 }
 EXPORT_SYMBOL(dst_release);
@@ -186,8 +191,10 @@ void dst_release_immediate(struct dst_en
 		if (WARN_ONCE(newrefcnt < 0, "dst_release_immediate underflow"))
 			net_warn_ratelimited("%s: dst:%p refcnt:%d\n",
 					     __func__, dst, newrefcnt);
-		if (!newrefcnt)
+		if (!newrefcnt){
+			dst_count_dec(dst);
 			dst_destroy(dst);
+		}
 	}
 }
 EXPORT_SYMBOL(dst_release_immediate);