Skip site navigation (1)Skip section navigation (2) 
Date:      Thu, 4 Aug 2016 08:51:02 +0000
From:      "sepherosa_gmail.com (Sepherosa Ziehau)" <phabric-noreply@FreeBSD.org>
To:        freebsd-net@freebsd.org
Subject:   [Differential] D7415: tcp/lro: If timestamps mismatch or it's a FIN, force flush.
Message-ID:  <differential-rev-PHID-DREV-nxbjynsxkd5zh6l3wyn3-req@FreeBSD.org>
next in thread | raw e-mail | index | archive | help 

[-- Attachment #1 --]
sepherosa_gmail.com created this revision.
sepherosa_gmail.com added reviewers: rrs, gallatin, hselasky, np, glebius.
sepherosa_gmail.com added a subscriber: freebsd-net-list.
Herald added a reviewer: transport.

REVISION SUMMARY
    This keeps the segments/ACK/FIN delivery order.
    
    Before this patch, it was observed: if A sent FIN immediately after
    an ACK, B would deliver FIN first to the TCP stack, then the ACK.
    This out-of-order delivery causes one unnecessary ACK sent from B.
    
    Obtained from:  gallatin, rrs

REVISION DETAIL
  https://reviews.freebsd.org/D7415

AFFECTED FILES
  sys/netinet/tcp_lro.c

EMAIL PREFERENCES
  https://reviews.freebsd.org/settings/panel/emailpreferences/

To: sepherosa_gmail.com, #transport, rrs, gallatin, hselasky, np, glebius
Cc: freebsd-net-list

[-- Attachment #2 --]
diff --git a/sys/netinet/tcp_lro.c b/sys/netinet/tcp_lro.c
--- a/sys/netinet/tcp_lro.c
+++ b/sys/netinet/tcp_lro.c
@@ -603,6 +603,7 @@
 	int error, ip_len, l;
 	uint16_t eh_type, tcp_data_len;
 	struct lro_head *bucket;
+	int force_flush = 0;
 
 	/* We expect a contiguous header [eh, ip, tcp]. */
 
@@ -669,8 +670,15 @@
 	 * Check TCP header constraints.
 	 */
 	/* Ensure no bits set besides ACK or PSH. */
-	if ((th->th_flags & ~(TH_ACK | TH_PUSH)) != 0)
-		return (TCP_LRO_CANNOT);
+	if ((th->th_flags & ~(TH_ACK | TH_PUSH)) != 0) {
+		if (th->th_flags & TH_SYN)
+			return (TCP_LRO_CANNOT);
+		/*
+		 * Make sure that previously seen segements/ACKs are delivered
+		 * before this segement, e.g. FIN.
+		 */
+		force_flush = 1;
+	}
 
 	/* XXX-BZ We lose a ACK|PUSH flag concatenating multiple segments. */
 	/* XXX-BZ Ideally we'd flush on PUSH? */
@@ -686,8 +694,13 @@
 	ts_ptr = (uint32_t *)(th + 1);
 	if (l != 0 && (__predict_false(l != TCPOLEN_TSTAMP_APPA) ||
 	    (*ts_ptr != ntohl(TCPOPT_NOP<<24|TCPOPT_NOP<<16|
-	    TCPOPT_TIMESTAMP<<8|TCPOLEN_TIMESTAMP))))
-		return (TCP_LRO_CANNOT);
+	    TCPOPT_TIMESTAMP<<8|TCPOLEN_TIMESTAMP)))) {
+		/*
+		 * Make sure that previously seen segements/ACKs are delivered
+		 * before this segement.
+		 */
+		force_flush = 1;
+	}
 
 	/* If the driver did not pass in the checksum, set it now. */
 	if (csum == 0x0000)
@@ -754,6 +767,13 @@
 #endif
 		}
 
+		if (force_flush) {
+			/* Timestamps mismatch; this is a FIN, etc */
+			tcp_lro_active_remove(le);
+			tcp_lro_flush(lc, le);
+			return (TCP_LRO_CANNOT);
+		}
+
 		/* Flush now if appending will result in overflow. */
 		if (le->p_len > (lc->lro_length_lim - tcp_data_len)) {
 			tcp_lro_active_remove(le);

Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?differential-rev-PHID-DREV-nxbjynsxkd5zh6l3wyn3-req>