tcp: cache result of earlier divides when mss-aligning things

The results is very unlikely change every so often so we
hardly need to divide again after doing that once for a
connection. Yet, if divide still becomes necessary we
detect that and do the right thing and again settle for
non-divide state. Takes the u16 space which was previously
taken by the plain xmit_size_goal.

This should take care part of the tso vs non-tso difference
we found earlier.

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 886596f..0db9f3b 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -665,7 +665,7 @@
 				       int large_allowed)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
-	u32 xmit_size_goal;
+	u32 xmit_size_goal, old_size_goal;
 
 	xmit_size_goal = mss_now;
 
@@ -676,7 +676,17 @@
 				  tp->tcp_header_len);
 
 		xmit_size_goal = tcp_bound_to_half_wnd(tp, xmit_size_goal);
-		xmit_size_goal -= (xmit_size_goal % mss_now);
+
+		/* We try hard to avoid divides here */
+		old_size_goal = tp->xmit_size_goal_segs * mss_now;
+
+		if (likely(old_size_goal <= xmit_size_goal &&
+			   old_size_goal + mss_now > xmit_size_goal)) {
+			xmit_size_goal = old_size_goal;
+		} else {
+			tp->xmit_size_goal_segs = xmit_size_goal / mss_now;
+			xmit_size_goal = tp->xmit_size_goal_segs * mss_now;
+		}
 	}
 
 	return xmit_size_goal;