tcp: add missing tcp_skb_can_collapse() test in tcp_shift_skb_data()

tcp_shift_skb_data() might collapse three packets into a larger one.

P_A, P_B, P_C  -> P_ABC

Historically, it used a single tcp_skb_can_collapse_to(P_A) call,
because it was enough.

In commit 8571248 ("tcp: coalesce/collapse must respect MPTCP extensions"),
this call was replaced by a call to tcp_skb_can_collapse(P_A, P_B)

But the now needed test over P_C has been missed.

This probably broke MPTCP.

Then later, commit 9b65b17 ("net: avoid double accounting for pure zerocopy skbs")
added an extra condition to tcp_skb_can_collapse(), but the missing call
from tcp_shift_skb_data() is also breaking TCP zerocopy, because P_A and P_C
might have different skb_zcopy_pure() status.

Fixes: 8571248 ("tcp: coalesce/collapse must respect MPTCP extensions")
Fixes: 9b65b17 ("net: avoid double accounting for pure zerocopy skbs")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Mat Martineau <mathew.j.martineau@linux.intel.com>
Cc: Talal Ahmad <talalahmad@google.com>
Cc: Arjun Roy <arjunroy@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Link: https://lore.kernel.org/r/20220201184640.756716-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Author: edumazet

net/ipv4/tcp_input.c

@@ -1660,6 +1660,8 @@ static struct sk_buff *tcp_shift_skb_data(struct sock *sk, struct sk_buff *skb,
 	    (mss != tcp_skb_seglen(skb)))
 		goto out;
 
+	if (!tcp_skb_can_collapse(prev, skb))
+		goto out;
 	len = skb->len;
 	pcount = tcp_skb_pcount(skb);
 	if (tcp_skb_shift(prev, skb, pcount, len))