net/tcp: Set the default TCP MSS to the value required by RFC 879 and…

… never change it under any circumstance unless the remote host requests a larger MSS via an option the TCP header.

include/nuttx/net/tcp.h

@@ -107,40 +107,51 @@
 #define TCP_MAX_HDRLEN    60                         /* Maximum size of TCP header */
 
 #ifdef CONFIG_NET_IPv4
-#  define IPv4TCP_HDRLEN (TCP_HDRLEN + IPv4_HDRLEN) /* Size of IPv4 + TCP header */
+#  define IPv4TCP_HDRLEN  (TCP_HDRLEN + IPv4_HDRLEN) /* Size of IPv4 + TCP header */
 #endif
 
 #ifdef CONFIG_NET_IPv6
-#  define IPv6TCP_HDRLEN (TCP_HDRLEN + IPv6_HDRLEN) /* Size of IPv4 + TCP header */
+#  define IPv6TCP_HDRLEN  (TCP_HDRLEN + IPv6_HDRLEN) /* Size of IPv4 + TCP header */
 #endif
 
 /* Initial minimum MSS according to RFC 879
  *
- * There have been some assumptions made about using other than the
- * default size for datagrams with some unfortunate results.
+ * The default TCP Maximum Segment Sizes (MSS) are defined below.is 536.  If
+ * a host wishes to set the maximum segment size to a value other than the
+ * default, the maximum segment size is specified as a TCP option, initially
+ * in the TCP SYN packet during the TCP handshake.  The value cannot be
+ * changed after the connection is established.
  *
- *     HOSTS MUST NOT SEND DATAGRAMS LARGER THAN 576 OCTETS UNLESS THEY
- *     HAVE SPECIFIC KNOWLEDGE THAT THE DESTINATION HOST IS PREPARED TO
- *     ACCEPT LARGER DATAGRAMS.
+ * These defaults correspond to the minimum MTU values:
  *
- * This is a long established rule.
+ *   IPv4:  MTU=576;  MSS=536  (MTU - IPv4_HDRLEN - TCP_HDRLEN)
+ *   IPv6:  MTU=1280; MSS=1200 (MTU - IPv5_HDRLEN - TCP_HDRLEN)
  */
 
-#define TCP_INITIAL_MSS(d,h)       (TCP_MSS(d,h) > 576 ? 576 : TCP_MSS(d,h))
+#define TCP_DEFAULT_IPv4_MSS  536
+#define TCP_DEFAULT_IPv6_MSS  1200
 
-#define MIN_TCP_INITIAL_MSS(h)     (__MIN_TCP_MSS(h) > 576 ? 576 : __MIN_TCP_MSS(h))
-#define MAX_TCP_INITIAL_MSS(h)     (__MAX_TCP_MSS(h) > 576 ? 576 : __MAX_TCP_MSS(h))
+/* However, we do need to make allowance for certain links such as SLIP that
+ * have unusually small MTUs.
+ */
 
 #ifdef CONFIG_NET_IPv4
-#  define TCP_IPv4_INITIAL_MSS(d)  TCP_INITIAL_MSS(d,IPv4_HDRLEN)
-#  define MIN_IPv4_TCP_INITIAL_MSS MIN_TCP_INITIAL_MSS(IPv4_HDRLEN)
-#  define MAX_IPv4_TCP_INITIAL_MSS MAX_TCP_INITIAL_MSS(IPv4_HDRLEN)
+#  define TCP_IPv4_INITIAL_MSS(d) \
+     (TCP_MSS(d,IPv4_HDRLEN) > 536 ? 536 : TCP_MSS(d,IPv4_HDRLEN))
+
+#  define MIN_IPv4_TCP_INITIAL_MSS \
+     (__MIN_TCP_MSS(IPv4_HDRLEN) > 536 ? 536 : __MIN_TCP_MSS(IPv4_HDRLEN))
+#  define MAX_IPv4_TCP_INITIAL_MSS  \
+     (__MAX_TCP_MSS(IPv4_HDRLEN) > 536 ? 536 : __MAX_TCP_MSS(h))
 #endif
 
 #ifdef CONFIG_NET_IPv6
-#  define TCP_IPv6_INITIAL_MSS(d)  TCP_INITIAL_MSS(d,IPv6_HDRLEN)
-#  define MIN_IPv6_TCP_INITIAL_MSS MIN_TCP_INITIAL_MSS(IPv6_HDRLEN)
-#  define MAX_IPv6_TCP_INITIAL_MSS MAX_TCP_INITIAL_MSS(IPv6_HDRLEN)
+#  define TCP_IPv6_INITIAL_MSS(d) \
+     (TCP_MSS(d,IPv6_HDRLEN) > 1200 ? 1200 : TCP_MSS(d,IPv6_HDRLEN))
+#  define MIN_IPv6_TCP_INITIAL_MSS \
+     (__MIN_TCP_MSS(IPv6_HDRLEN) > 1200 ? 1200 : __MIN_TCP_MSS(IPv6_HDRLEN))
+#  define MAX_IPv6_TCP_INITIAL_MSS  \
+     (__MAX_TCP_MSS(IPv6_HDRLEN) > 1200 ? 1200 : __MAX_TCP_MSS(IPv6_HDRLEN))
 #endif
 
 /****************************************************************************

net/tcp/tcp.h

@@ -343,8 +343,8 @@ void tcp_initialize(void);
  * Description:
  *   Find a free TCP/IP connection structure and allocate it
  *   for use.  This is normally something done by the implementation of the
- *   socket() API but is also called from the driver level when a TCP
- *   packet is received while "listening"
+ *   socket() API but is also called from the event processing logic when a
+ *   TCP packet is received while "listening"
  *
  ****************************************************************************/
 

net/tcp/tcp_conn.c

@@ -613,7 +613,7 @@ void tcp_initialize(void)
  * Description:
  *   Find a free TCP/IP connection structure and allocate it
  *   for use.  This is normally something done by the implementation of the
- *   socket() API but is also called from the even procressing lock when a
+ *   socket() API but is also called from the event processing logic when a
  *   TCP packet is received while "listening"
  *
  ****************************************************************************/

net/tcp/tcp_connect.c

@@ -264,36 +264,11 @@ static uint16_t psock_connect_eventhandler(FAR struct net_driver_s *dev,
 
       psock_teardown_callbacks(pstate, pstate->tc_result);
 
-      /* When we set up the connection structure, we did not know the size
-       * of the initial MSS.  Now that the connection is associated with a
-       * network device, we now know the size of link layer header and can
-       * determine the correct initial MSS.
-       */
-
-      DEBUGASSERT(pstate->tc_conn);
-
-#ifdef CONFIG_NET_IPv4
-#ifdef CONFIG_NET_IPv6
-      if (pstate->tc_conn->domain == PF_INET)
-#endif
-        {
-          pstate->tc_conn->mss = TCP_IPv4_INITIAL_MSS(dev);
-        }
-#endif /* CONFIG_NET_IPv4 */
-
-#ifdef CONFIG_NET_IPv6
-#ifdef CONFIG_NET_IPv4
-      else
-#endif
-        {
-          pstate->tc_conn->mss = TCP_IPv6_INITIAL_MSS(dev);
-        }
-#endif /* CONFIG_NET_IPv6 */
-
       /* We now have to filter all outgoing transfers so that they use only
        * the MSS of this device.
        */
 
+      DEBUGASSERT(pstate->tc_conn != NULL);
       DEBUGASSERT(pstate->tc_conn->dev == NULL ||
                   pstate->tc_conn->dev == dev);
       pstate->tc_conn->dev = dev;