net/tls: don't copy negative amounts of data in reencrypt

There is no guarantee the record starts before the skb frags.
If we don't check for this condition copy amount will get
negative, leading to reads and writes to random memory locations.
Familiar hilarity ensues.

Fixes: 4799ac81 ("tls: Add rx inline crypto offload")
Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: NJohn Hurley <john.hurley@netronome.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

net/tls/tls_device.c

@@ -628,6 +628,7 @@ static int tls_device_reencrypt(struct sock *sk, struct sk_buff *skb)
 	else
 		err = 0;
 
+	if (skb_pagelen(skb) > offset) {
 		copy = min_t(int, skb_pagelen(skb) - offset,
 			     rxm->full_len - TLS_CIPHER_AES_GCM_128_TAG_SIZE);
 
@@ -636,6 +637,7 @@ static int tls_device_reencrypt(struct sock *sk, struct sk_buff *skb)
 
 		offset += copy;
 		buf += copy;
+	}
 
 	skb_walk_frags(skb, skb_iter) {
 		copy = min_t(int, skb_iter->len,