1 files changed, 22 insertions, 12 deletions

diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 8f060d7f9a0e..0a49a8c7c564 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -2785,6 +2785,7 @@ static int netlink_dump(struct sock *sk)
 	struct sk_buff *skb = NULL;
 	struct nlmsghdr *nlh;
 	int len, err = -ENOBUFS;
+	int alloc_min_size;
 	int alloc_size;
 
 	mutex_lock(nlk->cb_mutex);
@@ -2793,9 +2794,6 @@ static int netlink_dump(struct sock *sk)
 		goto errout_skb;
 	}
 
-	cb = &nlk->cb;
-	alloc_size = max_t(int, cb->min_dump_alloc, NLMSG_GOODSIZE);
-
 	if (!netlink_rx_is_mmaped(sk) &&
 	    atomic_read(&sk->sk_rmem_alloc) >= sk->sk_rcvbuf)
 		goto errout_skb;
@@ -2805,23 +2803,35 @@ static int netlink_dump(struct sock *sk)
 	 * to reduce number of system calls on dump operations, if user
 	 * ever provided a big enough buffer.
 	 */
-	if (alloc_size < nlk->max_recvmsg_len) {
-		skb = netlink_alloc_skb(sk,
-					nlk->max_recvmsg_len,
-					nlk->portid,
+	cb = &nlk->cb;
+	alloc_min_size = max_t(int, cb->min_dump_alloc, NLMSG_GOODSIZE);
+
+	if (alloc_min_size < nlk->max_recvmsg_len) {
+		alloc_size = nlk->max_recvmsg_len;
+		skb = netlink_alloc_skb(sk, alloc_size, nlk->portid,
 					GFP_KERNEL |
 					__GFP_NOWARN |
 					__GFP_NORETRY);
-		/* available room should be exact amount to avoid MSG_TRUNC */
-		if (skb)
-			skb_reserve(skb, skb_tailroom(skb) -
-					 nlk->max_recvmsg_len);
 	}
-	if (!skb)
+	if (!skb) {
+		alloc_size = alloc_min_size;
 		skb = netlink_alloc_skb(sk, alloc_size, nlk->portid,
 					GFP_KERNEL);
+	}
 	if (!skb)
 		goto errout_skb;
+
+	/* Trim skb to allocated size. User is expected to provide buffer as
+	 * large as max(min_dump_alloc, 16KiB (mac_recvmsg_len capped at
+	 * netlink_recvmsg())). dump will pack as many smaller messages as
+	 * could fit within the allocated skb. skb is typically allocated
+	 * with larger space than required (could be as much as near 2x the
+	 * requested size with align to next power of 2 approach). Allowing
+	 * dump to use the excess space makes it difficult for a user to have a
+	 * reasonable static buffer based on the expected largest dump of a
+	 * single netdev. The outcome is MSG_TRUNC error.
+	 */
+	skb_reserve(skb, skb_tailroom(skb) - alloc_size);
 	netlink_skb_set_owner_r(skb, sk);
 
 	len = cb->dump(skb, cb);