Changed behaviour into implementation.

added assertion: when the queues are NULL, queuelen must be zero.
remove ACK flags from a PCB, only when we could succesfully sent
an empty ACK packet.

CHANGELOG

@@ -5,16 +5,89 @@ FUTURE
     problems with exoteric (/DSP) architectures showing these problems.
     We still have to fix some of these issues neatly.
 
+  * TODO: the ARP layer is not protected against concurrent access. If
+    you run from a multitasking OS, serialize access to ARP (called from
+    your network device driver and from a timeout thread.)
+
 HISTORY
 
 (HEAD)
 
+  2004-12-28 Leon Woestenberg <leon.woestenberg@gmx.net>
+  * etharp.*: Disabled multiple packets on the ARP queue.
+    This clashes with TCP queueing.
+
+  2004-11-28 Leon Woestenberg <leon.woestenberg@gmx.net>
+  * etharp.*: Fixed race condition from ARP request to ARP timeout.
+    Halved the ARP period, doubled the period counts.
+    ETHARP_MAX_PENDING now should be at least 2. This prevents
+    the counter from reaching 0 right away (which would allow
+    too little time for ARP responses to be received).
+    
+  2004-11-25 Leon Woestenberg <leon.woestenberg@gmx.net>
+  * dhcp.c: Decline messages were not multicast but unicast.
+  * etharp.c: ETHARP_CREATE is renamed to ETHARP_TRY_HARD.
+    Do not try hard to insert arbitrary packet's source address,
+    etharp_ip_input() now calls etharp_update() without ETHARP_TRY_HARD. 
+    etharp_query() now always DOES call ETHARP_TRY_HARD so that users
+    querying an address will see it appear in the cache (DHCP could
+    suffer from this when a server invalidly gave an in-use address.)
+  * ipv4/ip_addr.h: Renamed ip_addr_maskcmp() to _netcmp() as we are
+    comparing network addresses (identifiers), not the network masks
+    themselves.
+  * ipv4/ip_addr.c: ip_addr_isbroadcast() now checks that the given
+    IP address actually belongs to the network of the given interface.
+
+  2004-11-24 Kieran Mansley <kjm25@cam.ac.uk>
+  * tcp.c: Increment pcb->snd_buf when ACK is received in SYN_SENT state.
+
+(STABLE-1_1_0-RC1)
+
+  2004-10-16 Kieran Mansley <kjm25@cam.ac.uk>
+  * tcp.c: Add code to tcp_recved() to send an ACK (window update) immediately,
+	even if one is already pending, if the rcv_wnd is above a threshold
+	(currently TCP_WND/2). This avoids waiting for a timer to expire to send a
+	delayed ACK in order to open the window if the stack is only receiving data.
+
+  2004-09-12 Kieran Mansley <kjm25@cam.ac.uk>
+  * tcp*.*: Retransmit time-out handling improvement by Sam Jansen.
+
+  2004-08-20 Tony Mountifield <tony@softins.co.uk>
+  * etharp.c: Make sure the first pbuf queued on an ARP entry
+    is properly ref counted.
+
+  2004-07-27 Tony Mountifield <tony@softins.co.uk>
+  * debug.h: Added (int) cast in LWIP_DEBUGF() to avoid compiler
+    warnings about comparison.
+  * pbuf.c: Stopped compiler complaining of empty if statement
+    when LWIP_DEBUGF() empty.  Closed an unclosed comment.
+  * tcp.c: Stopped compiler complaining of empty if statement
+    when LWIP_DEBUGF() empty.
+  * ip.h Corrected IPH_TOS() macro: returns a byte, so doesn't need htons().
+  * inet.c: Added a couple of casts to quiet the compiler.
+    No need to test isascii(c) before isdigit(c) or isxdigit(c).
+
+  2004-07-22 Tony Mountifield <tony@softins.co.uk>
+  * inet.c: Made data types consistent in inet_ntoa().
+    Added casts for return values of checksum routines, to pacify compiler.
+  * ip_frag.c, tcp_out.c, sockets.c, pbuf.c
+    Small corrections to some debugging statements, to pacify compiler.
+
+  2004-07-21 Tony Mountifield <tony@softins.co.uk>
+  * etharp.c: Removed spurious semicolon and added missing end-of-comment.
+  * ethernetif.c Updated low_level_output() to match prototype for
+    netif->linkoutput and changed low_level_input() similarly for consistency.
+  * api_msg.c: Changed recv_raw() from int to u8_t, to match prototype
+    of raw_recv() in raw.h and so avoid compiler error.
+  * sockets.c: Added trivial (int) cast to keep compiler happier.
+  * ip.c, netif.c Changed debug statements to use the tidier ip4_addrN() macros.
+  
 (STABLE-1_0_0)
 
   ++ Changes:
 
   2004-07-05 Leon Woestenberg <leon.woestenberg@gmx.net>
-  * sockets.* Restructured LWIP_PRIVATE_TIMEVAL. Make sure
+  * sockets.*: Restructured LWIP_PRIVATE_TIMEVAL. Make sure
     your cc.h file defines this either 1 or 0. If non-defined,
     defaults to 1.
   * .c: Added <string.h> and <errno.h> includes where used.

doc/contrib.txt

@@ -27,7 +27,7 @@ features of Savannah help us not lose users' input.
 
 1. JavaDoc compliant and Doxygen compatible.
 2. Function documentation above functions in .c files, not .h files.
-   (This forces you to synchronize documentation and behaviour.)
+   (This forces you to synchronize documentation and implementation.)
 3. Use current documentation style as further reference.
  
 2.3 Bug reports and patches:

doc/rawapi.txt

@@ -1,20 +1,21 @@
-Raw TCP/IP interface for lwIP 0.5
+Raw TCP/IP interface for lwIP
 
-Author: Adam Dunkels
+Authors: Adam Dunkels, Leon Woestenberg
 
 lwIP provides two Application Program's Interfaces (APIs) for programs
-to use for communication with the TCP/IP code: the sequential API
-(often just called "the API") and the raw TCP/IP interface. This
-document is intended as a description of the latter. For lwIP versions
-lower than 0.5, this API was not documented.
+to use for communication with the TCP/IP code:
+* low-level "core" / "callback" or "raw" API.
+* higher-level "sequential" API.
 
 The sequential API provides a way for ordinary, sequential, programs
 to use the lwIP stack. It is quite similar to the BSD socket API. The
-model of execution is based on the open-read-write-close
+model of execution is based on the blocking open-read-write-close
 paradigm. Since the TCP/IP stack is event based by nature, the TCP/IP
 code and the application program must reside in different execution
 contexts (threads).
 
+** The remainder of this document discusses the "raw" API. **
+
 The raw TCP/IP interface allows the application program to integrate
 better with the TCP/IP code. Program execution is event based by
 having callback functions being called from within the TCP/IP
@@ -34,7 +35,6 @@ Both APIs can be used simultaneously by different application
 programs. In fact, the sequential API is implemented as an application
 program using the raw TCP/IP interface.
 
-
 --- Callbacks
 
 Program execution is driven by callbacks. Each callback is an ordinary

doc/savannah.txt

@@ -115,40 +115,16 @@ tar czvf lwip-0.6.3.tar.gz lwip-0.6.3
 tar cjvf lwip-0.6.3.tar.bz2 lwip-0.6.3
 zip -r lwip-0.6.3.zip lwip-0.6.3
 
-First, make a local release directory to work in, I use "lwip-releases":
+Now, sign the archives with a detached GPG binary signature as follows:
 
-mkdir lwip-releases
-cd lwip-releases
+gpg -b lwip-0.6.3.tar.gz
+gpg -b lwip-0.6.3.tar.bz2
+gpg -b lwip-0.6.3.zip
 
-Now, make a new release by creating a new directory for it (these are
-Savannah conventions so that it shows up in the Files list real nice):
+Upload these files using anonymous FTP:
+ncftp ftp://savannah.gnu.org/incoming/savannah/lwip
 
-mkdir stable.pkg
-mkdir stable.pkg 0.6.3
-
-We can now copy the tar archive we made earlier into the release directory:
-
-cp ../../../lwip-0.6.3.tar.gz .
-
-Finally, synchronize this directory upwards to Savannah:
-
-rsync -n -e "ssh -1" -t -u -v -r *.pkg likewise@savannah.nongnu.org:/upload/lwip
-
-This does a "dry run": no files are modified! After you have confirmed that
-this is what you intended to do, remove "-n" and actually synchronize for
-real. The release should now be available here:
-
-http://savannah.nongnu.org/files/?group=lwip
-
----
-Explanation of rsync options used:
-
--t: preserve file timestamps
--u: do not overwrite existing files, unless they are older
--v: be verbose (long format file attributes)
--r: recurse into directories
--n: dry-run, do not modify anything. 
----
+ncftp>mput *0.6.3.*
 
 Additionally, you may post a news item on Savannah, like this:
 

src/FILES

@@ -1,11 +1,13 @@
-api/      - The code for the API.	   
+api/      - The code for the high-level wrapper API. Not needed if
+            you use the lowel-level call-back/raw API.
 
-core/     - The core files including protocol implementations, memory
-            and buffer management etc.
+core/     - The core of the TPC/IP stack; protocol implementations,
+            memory and buffer management, and the low-level raw API.
 	    
 include/  - lwIP include files.
 
-netif/    - Generic network interface device drivers are kept here.
+netif/    - Generic network interface device drivers are kept here,
+            as well as the ARP module.
 
 For more information on the various subdirectories, check the FILES
 file in each directory.

src/api/api_lib.c

@@ -280,9 +280,10 @@ netconn_delete(struct netconn *conn)
   if (conn->recvmbox != SYS_MBOX_NULL) {
     while (sys_arch_mbox_fetch(conn->recvmbox, &mem, 1) != SYS_ARCH_TIMEOUT) {
       if (conn->type == NETCONN_TCP) {
-  pbuf_free((struct pbuf *)mem);
+        if(mem != NULL)
+          pbuf_free((struct pbuf *)mem);
       } else {
-  netbuf_delete((struct netbuf *)mem);
+        netbuf_delete((struct netbuf *)mem);
       }
     }
     sys_mbox_free(conn->recvmbox);

src/api/sockets.c

@@ -418,7 +418,7 @@ lwip_recvfrom(int s, void *mem, int len, unsigned int flags,
     ip_addr_debug_print(SOCKETS_DEBUG, addr);
     LWIP_DEBUGF(SOCKETS_DEBUG, (" port=%u len=%u\n", port, copylen));
   } else {
-#if SOCKETS_DEBUG != 0
+#if SOCKETS_DEBUG
     addr = netbuf_fromaddr(buf);
     port = netbuf_fromport(buf);
 

src/api/tcpip.c

@@ -46,11 +46,10 @@
 static void (* tcpip_init_done)(void *arg) = NULL;
 static void *tcpip_init_done_arg;
 static sys_mbox_t mbox;
+
 #if LWIP_TCP
 static int tcpip_tcp_timer_active = 0;
 
-
-
 static void
 tcpip_tcp_timer(void *arg)
 {
@@ -68,6 +67,7 @@ tcpip_tcp_timer(void *arg)
   }
 }
 
+#if !NO_SYS
 void
 tcp_timer_needed(void)
 {
@@ -78,6 +78,7 @@ tcp_timer_needed(void)
     sys_timeout(TCP_TMR_INTERVAL, tcpip_tcp_timer, NULL);
   }
 }
+#endif /* !NO_SYS */
 #endif /* LWIP_TCP */
 
 static void

src/core/dhcp.c

@@ -85,7 +85,8 @@
 #if LWIP_DHCP /* don't build if not configured for use in lwipopt.h */
 
 /** global transaction identifier, must be
- *  unique for each DHCP request. */
+ *  unique for each DHCP request. We simply increment, starting
+ *  with this value (easy to match with a packet analyzer) */
 static u32_t xid = 0xABCD0000;
 
 /** DHCP client state machine functions */
@@ -99,7 +100,6 @@ static void dhcp_check(struct netif *netif);
 static void dhcp_bind(struct netif *netif);
 static err_t dhcp_decline(struct netif *netif);
 static err_t dhcp_rebind(struct netif *netif);
-static err_t dhcp_release(struct netif *netif);
 static void dhcp_set_state(struct dhcp *dhcp, unsigned char new_state);
 
 /** receive, unfold, parse and free incoming messages */
@@ -679,8 +679,10 @@ static err_t dhcp_decline(struct netif *netif)
     pbuf_realloc(dhcp->p_out, sizeof(struct dhcp_msg) - DHCP_OPTIONS_LEN + dhcp->options_out_len);
 
     udp_bind(dhcp->pcb, IP_ADDR_ANY, DHCP_CLIENT_PORT);
-    udp_connect(dhcp->pcb, &dhcp->server_ip_addr, DHCP_SERVER_PORT);
-    udp_send(dhcp->pcb, dhcp->p_out);
+    /* @todo: should we really connect here? we are performing sendto() */
+    udp_connect(dhcp->pcb, IP_ADDR_ANY, DHCP_SERVER_PORT);
+    /* per section 4.4.4, broadcast DECLINE messages */
+    udp_sendto(dhcp->pcb, dhcp->p_out, IP_ADDR_BROADCAST, DHCP_SERVER_PORT);
     dhcp_delete_request(netif);
     LWIP_DEBUGF(DHCP_DEBUG | DBG_TRACE | DBG_STATE, ("dhcp_decline: BACKING OFF\n"));
   } else {
@@ -805,6 +807,8 @@ static void dhcp_bind(struct netif *netif)
   netif_set_netmask(netif, &sn_mask);
   LWIP_DEBUGF(DHCP_DEBUG | DBG_STATE, ("dhcp_bind(): GW: 0x%08lx\n", gw_addr.addr));
   netif_set_gw(netif, &gw_addr);
+  /* bring the interface up */
+  netif_set_up(netif);
   /* netif is now bound to DHCP leased address */
   dhcp_set_state(dhcp, DHCP_BOUND);
 }
@@ -922,7 +926,7 @@ static err_t dhcp_rebind(struct netif *netif)
  *
  * @param netif network interface which must release its lease
  */
-static err_t dhcp_release(struct netif *netif)
+err_t dhcp_release(struct netif *netif)
 {
   struct dhcp *dhcp = netif->dhcp;
   err_t result;
@@ -960,10 +964,13 @@ static err_t dhcp_release(struct netif *netif)
   msecs = dhcp->tries < 10 ? dhcp->tries * 1000 : 10 * 1000;
   dhcp->request_timeout = (msecs + DHCP_FINE_TIMER_MSECS - 1) / DHCP_FINE_TIMER_MSECS;
   LWIP_DEBUGF(DHCP_DEBUG | DBG_TRACE | DBG_STATE, ("dhcp_release(): set request timeout %u msecs\n", msecs));
+  /* bring the interface down */
+  netif_set_down(netif);
   /* remove IP address from interface */
   netif_set_ipaddr(netif, IP_ADDR_ANY);
   netif_set_gw(netif, IP_ADDR_ANY);
   netif_set_netmask(netif, IP_ADDR_ANY);
+  
   /* TODO: netif_down(netif); */
   return result;
 }

src/core/inet.c

@@ -46,10 +46,21 @@
 #include "lwip/def.h"
 #include "lwip/inet.h"
 
+#include "lwip/sys.h"
 
+/* This is a reference implementation of the checksum algorithm
 
+ - it may not work on all architectures, and all processors, particularly
+   if they have issues with alignment and 16 bit access.
+
+ - in this case you will need to port it to your architecture and 
+   #define LWIP_CHKSUM <your_checksum_routine> 
+   in your sys_arch.h
+*/
+#ifndef LWIP_CHKSUM
+#define LWIP_CHKSUM lwip_standard_chksum
 static u16_t
-lwip_chksum(void *dataptr, int len)
+lwip_standard_chksum(void *dataptr, int len)
 {
   u32_t acc;
 
@@ -75,6 +86,7 @@ lwip_chksum(void *dataptr, int len)
 
   return (u16_t)acc;
 }
+#endif
 
 /* inet_chksum_pseudo:
  *
@@ -96,7 +108,7 @@ inet_chksum_pseudo(struct pbuf *p,
   for(q = p; q != NULL; q = q->next) {
     LWIP_DEBUGF(INET_DEBUG, ("inet_chksum_pseudo(): checksumming pbuf %p (has next %p) \n",
       (void *)q, (void *)q->next));
-    acc += lwip_chksum(q->payload, q->len);
+    acc += LWIP_CHKSUM(q->payload, q->len);
     /*LWIP_DEBUGF(INET_DEBUG, ("inet_chksum_pseudo(): unwrapped lwip_chksum()=%lx \n", acc));*/
     while (acc >> 16) {
       acc = (acc & 0xffffUL) + (acc >> 16);
@@ -136,7 +148,7 @@ inet_chksum(void *dataptr, u16_t len)
 {
   u32_t acc;
 
-  acc = lwip_chksum(dataptr, len);
+  acc = LWIP_CHKSUM(dataptr, len);
   while (acc >> 16) {
     acc = (acc & 0xffff) + (acc >> 16);
   }
@@ -153,7 +165,7 @@ inet_chksum_pbuf(struct pbuf *p)
   acc = 0;
   swapped = 0;
   for(q = p; q != NULL; q = q->next) {
-    acc += lwip_chksum(q->payload, q->len);
+    acc += LWIP_CHKSUM(q->payload, q->len);
     while (acc >> 16) {
       acc = (acc & 0xffffUL) + (acc >> 16);
     }

src/core/ipv4/icmp.c

@@ -74,7 +74,7 @@ icmp_input(struct pbuf *p, struct netif *inp)
   case ICMP_ECHO:
     /* broadcast or multicast destination address? */
     if (ip_addr_isbroadcast(&iphdr->dest, inp) || ip_addr_ismulticast(&iphdr->dest)) {
-      LWIP_DEBUGF(ICMP_DEBUG, ("Smurf.\n"));
+      LWIP_DEBUGF(ICMP_DEBUG, ("icmp_input: Not echoing to multicast or broadcast pings\n"));
       ICMP_STATS_INC(icmp.err);
       pbuf_free(p);
       return;

src/core/ipv4/ip.c

@@ -85,7 +85,7 @@ ip_route(struct ip_addr *dest)
   /* iterate through netifs */
   for(netif = netif_list; netif != NULL; netif = netif->next) {
     /* network mask matches? */
-    if (ip_addr_maskcmp(dest, &(netif->ip_addr), &(netif->netmask))) {
+    if (ip_addr_netcmp(dest, &(netif->ip_addr), &(netif->netmask))) {
       /* return netif on which to forward IP packet */
       return netif;
     }
@@ -238,11 +238,8 @@ ip_input(struct pbuf *p, struct netif *inp) {
     {
       /* unicast to this interface address? */
       if (ip_addr_cmp(&(iphdr->dest), &(netif->ip_addr)) ||
-        /* or broadcast matching this interface network address? */
-        (ip_addr_isbroadcast(&(iphdr->dest), netif) &&
-         ip_addr_maskcmp(&(iphdr->dest), &(netif->ip_addr), &(netif->netmask))) ||
-         /* or restricted broadcast? */
-         ip_addr_cmp(&(iphdr->dest), IP_ADDR_BROADCAST)) {
+         /* or broadcast on this interface network address? */
+         ip_addr_isbroadcast(&(iphdr->dest), netif)) {
         LWIP_DEBUGF(IP_DEBUG, ("ip_input: packet accepted on interface %c%c\n",
           netif->name[0], netif->name[1]));
         /* break out of for loop */

src/core/ipv4/ip_addr.c

@@ -60,9 +60,12 @@ u8_t ip_addr_isbroadcast(struct ip_addr *addr, struct netif *netif)
   /* address matches network interface address exactly? => no broadcast */
   else if (addr->addr == netif->ip_addr.addr)
     return 0;
-  /* host identifier bits are all ones? => network broadcast address */
-  else if ((addr->addr & ~netif->netmask.addr) ==
-           (ip_addr_broadcast.addr & ~netif->netmask.addr))
+  /*  on the same (sub) network... */
+  else if (ip_addr_netcmp(addr, &(netif->ip_addr), &(netif->netmask))
+         /* ...and host identifier bits are all ones? =>... */
+          && ((addr->addr & ~netif->netmask.addr) ==
+           (ip_addr_broadcast.addr & ~netif->netmask.addr)))
+    /* => network broadcast address */
     return 1;
   else
     return 0;

src/core/ipv6/ip6.c

@@ -77,7 +77,7 @@ ip_route(struct ip_addr *dest)
   struct netif *netif;
 
   for(netif = netif_list; netif != NULL; netif = netif->next) {
-    if (ip_addr_maskcmp(dest, &(netif->ip_addr), &(netif->netmask))) {
+    if (ip_addr_netcmp(dest, &(netif->ip_addr), &(netif->netmask))) {
       return netif;
     }
   }

src/core/ipv6/ip6_addr.c

@@ -35,7 +35,7 @@
 
 
 int
-ip_addr_maskcmp(struct ip_addr *addr1, struct ip_addr *addr2,
+ip_addr_netcmp(struct ip_addr *addr1, struct ip_addr *addr2,
                 struct ip_addr *mask)
 {
   return((addr1->addr[0] & mask->addr[0]) == (addr2->addr[0] & mask->addr[0]) &&

src/core/pbuf.c

@@ -183,7 +183,7 @@ pbuf_pool_alloc(void)
 
 
 /**
- * Allocates a pbuf.
+ * Allocates a pbuf of the given type (possibly a chain for PBUF_POOL type).
  *
  * The actual memory allocated for the pbuf is determined by the
  * layer at which the pbuf is allocated and the requested size
@@ -319,7 +319,7 @@ pbuf_alloc(pbuf_layer l, u16_t length, pbuf_flag flag)
     LWIP_ASSERT("pbuf_alloc: pbuf->payload properly aligned",
            ((mem_ptr_t)p->payload % MEM_ALIGNMENT) == 0);
     break;
-  /* pbuf references existing (static constant) ROM payload? */
+  /* pbuf references existing (non-volatile static constant) ROM payload? */
   case PBUF_ROM:
   /* pbuf references existing (externally allocated) RAM payload? */
   case PBUF_REF:
@@ -664,8 +664,8 @@ pbuf_cat(struct pbuf *h, struct pbuf *t)
 {
   struct pbuf *p;
 
-  LWIP_ASSERT("h != NULL", h != NULL);
-  LWIP_ASSERT("t != NULL", t != NULL);
+  LWIP_ASSERT("h != NULL (programmer violates API)", h != NULL);
+  LWIP_ASSERT("t != NULL (programmer violates API)", t != NULL);
   if ((h == NULL) || (t == NULL)) return;
 
   /* proceed to last pbuf of chain */
@@ -675,10 +675,14 @@ pbuf_cat(struct pbuf *h, struct pbuf *t)
   }
   /* { p is last pbuf of first h chain, p->next == NULL } */
   LWIP_ASSERT("p->tot_len == p->len (of last pbuf in chain)", p->tot_len == p->len);
+  LWIP_ASSERT("p->next == NULL", p->next == NULL);
   /* add total length of second chain to last pbuf total of first chain */
   p->tot_len += t->tot_len;
   /* chain last pbuf of head (p) with first of tail (t) */
   p->next = t;
+  /* p->next now references t, but the caller will drop its reference to t,
+   * so netto there is no change to the reference count of t.
+   */
 }
 
 /**
@@ -706,8 +710,8 @@ pbuf_chain(struct pbuf *h, struct pbuf *t)
   LWIP_DEBUGF(PBUF_DEBUG | DBG_FRESH | 2, ("pbuf_chain: %p references %p\n", (void *)h, (void *)t));
 }
 
-/* For packet queueing. Note that queued packets must be dequeued first
- * before calling any pbuf functions. */
+/* For packet queueing. Note that queued packets MUST be dequeued first
+ * using pbuf_dequeue() before calling other pbuf_() functions. */
 #if ARP_QUEUEING
 /**
  * Add a packet to the end of a queue.
@@ -715,6 +719,7 @@ pbuf_chain(struct pbuf *h, struct pbuf *t)
  * @param q pointer to first packet on the queue
  * @param n packet to be queued
  *
+ * Both packets MUST be given, and must be different.
  */
 void
 pbuf_queue(struct pbuf *p, struct pbuf *n)
@@ -723,9 +728,13 @@ pbuf_queue(struct pbuf *p, struct pbuf *n)
   struct pbuf *q = p;
 #endif
   /* programmer stupidity checks */
-  LWIP_ASSERT("p != NULL", p != NULL);
-  LWIP_ASSERT("n != NULL", n != NULL);
-  if ((p == NULL) || (n == NULL)) return;
+  LWIP_ASSERT("p == NULL in pbuf_queue: this indicates a programmer error\n", p != NULL);
+  LWIP_ASSERT("n == NULL in pbuf_queue: this indicates a programmer error\n", n != NULL);
+  LWIP_ASSERT("p == n in pbuf_queue: this indicates a programmer error\n", p != n);
+  if ((p == NULL) || (n == NULL) || (p == n)){
+    LWIP_DEBUGF(PBUF_DEBUG | DBG_HALT | 3, ("pbuf_queue: programmer argument error\n"))
+    return;
+  }
 
   /* iterate through all packets on queue */
   while (p->next != NULL) {
@@ -787,7 +796,7 @@ pbuf_dequeue(struct pbuf *p)
   /* { p->tot_len == p->len } => p is the last pbuf of the first packet */
   /* remember next packet on queue in q */
   q = p->next;
-  /* dequeue p from queue */
+  /* dequeue packet p from queue */
   p->next = NULL;
   /* any next packet on queue? */
   if (q != NULL) {

src/core/tcp.c

@@ -444,6 +444,16 @@ tcp_recved(struct tcp_pcb *pcb, u16_t len)
      * continue to transmit.
      */
     tcp_ack(pcb);
+  } 
+  else if (pcb->flags & TF_ACK_DELAY && pcb->rcv_wnd >= TCP_WND/2) {
+    /* If we can send a window update such that there is a full
+     * segment available in the window, do so now.  This is sort of
+     * nagle-like in its goals, and tries to hit a compromise between
+     * sending acks each time the window is updated, and only sending
+     * window updates when a timer expires.  The "threshold" used
+     * above (currently TCP_WND/2) can be tuned to be more or less
+     * aggressive  */
+    tcp_ack_now(pcb);
   }
 
   LWIP_DEBUGF(TCP_DEBUG, ("tcp_recved: recveived %u bytes, wnd %u (%u).\n",
@@ -590,7 +600,6 @@ tcp_slowtmr(void)
         if (pcb->state != SYN_SENT) {
           pcb->rto = ((pcb->sa >> 3) + pcb->sv) << tcp_backoff[pcb->nrtx];
         }
-        tcp_rexmit(pcb);
         /* Reduce congestion window and ssthresh. */
         eff_wnd = LWIP_MIN(pcb->cwnd, pcb->snd_wnd);
         pcb->ssthresh = eff_wnd >> 1;
@@ -600,7 +609,10 @@ tcp_slowtmr(void)
         pcb->cwnd = pcb->mss;
         LWIP_DEBUGF(TCP_CWND_DEBUG, ("tcp_slowtmr: cwnd %u ssthresh %u\n",
                                 pcb->cwnd, pcb->ssthresh));
-      }
+ 
+        /* The following needs to be called AFTER cwnd is set to one mss - STJ */
+        tcp_rexmit_rto(pcb);
+     }
     }
     /* Check if this PCB has stayed too long in FIN-WAIT-2 */
     if (pcb->state == FIN_WAIT_2) {
@@ -846,7 +858,7 @@ tcp_kill_prio(u8_t prio)
     }
   }
   if (inactive != NULL) {
-    LWIP_DEBUGF(TCP_DEBUG, ("tcp_kill_prio: killing oldest PCB 0x%p (%ld)\n",
+    LWIP_DEBUGF(TCP_DEBUG, ("tcp_kill_prio: killing oldest PCB %p (%ld)\n",
            (void *)inactive, inactivity));
     tcp_abort(inactive);
   }      
@@ -868,7 +880,7 @@ tcp_kill_timewait(void)
     }
   }
   if (inactive != NULL) {
-    LWIP_DEBUGF(TCP_DEBUG, ("tcp_kill_timewait: killing oldest TIME-WAIT PCB 0x%p (%ld)\n",
+    LWIP_DEBUGF(TCP_DEBUG, ("tcp_kill_timewait: killing oldest TIME-WAIT PCB %p (%ld)\n",
            (void *)inactive, inactivity));
     tcp_abort(inactive);
   }      

src/core/tcp_in.c

@@ -503,12 +503,14 @@ tcp_process(struct tcp_pcb *pcb)
     /* First, determine if the reset is acceptable. */
     if (pcb->state == SYN_SENT) {
       if (ackno == pcb->snd_nxt) {
-  acceptable = 1;
+        acceptable = 1;
       }
     } else {
-      if (TCP_SEQ_GEQ(seqno, pcb->rcv_nxt) &&
-   TCP_SEQ_LEQ(seqno, pcb->rcv_nxt + pcb->rcv_wnd)) {
-  acceptable = 1;
+      /*if (TCP_SEQ_GEQ(seqno, pcb->rcv_nxt) &&
+          TCP_SEQ_LEQ(seqno, pcb->rcv_nxt + pcb->rcv_wnd)) {
+      */
+      if(TCP_SEQ_BETWEEN(seqno, pcb->rcv_nxt, pcb->rcv_nxt+pcb->rcv_wnd)){
+        acceptable = 1;
       }
     }
 
@@ -538,6 +540,7 @@ tcp_process(struct tcp_pcb *pcb)
      pcb->snd_nxt, ntohl(pcb->unacked->tcphdr->seqno)));
     if ((flags & TCP_ACK) && (flags & TCP_SYN)
         && ackno == ntohl(pcb->unacked->tcphdr->seqno) + 1) {
+      pcb->snd_buf ++;
       pcb->rcv_nxt = seqno + 1;
       pcb->lastack = ackno;
       pcb->snd_wnd = tcphdr->wnd;
@@ -562,8 +565,9 @@ tcp_process(struct tcp_pcb *pcb)
   case SYN_RCVD:
     if (flags & TCP_ACK &&
        !(flags & TCP_RST)) {
-      if (TCP_SEQ_LT(pcb->lastack, ackno) &&
-          TCP_SEQ_LEQ(ackno, pcb->snd_nxt)) {
+      /*if (TCP_SEQ_LT(pcb->lastack, ackno) &&
+        TCP_SEQ_LEQ(ackno, pcb->snd_nxt)) { */
+      if(TCP_SEQ_BETWEEN(ackno, pcb->lastack+1, pcb->snd_nxt)){
         pcb->state = ESTABLISHED;
         LWIP_DEBUGF(TCP_DEBUG, ("TCP connection established %u -> %u.\n", inseg.tcphdr->src, inseg.tcphdr->dest));
 #if LWIP_CALLBACK_API
@@ -732,8 +736,10 @@ tcp_receive(struct tcp_pcb *pcb)
         LWIP_DEBUGF(TCP_FR_DEBUG, ("tcp_receive: dupack averted %lu %lu\n",
                                    pcb->snd_wl1 + pcb->snd_wnd, right_wnd_edge));
       }
-    } else if (TCP_SEQ_LT(pcb->lastack, ackno) &&
-               TCP_SEQ_LEQ(ackno, pcb->snd_max)) {
+    } else
+      /*if (TCP_SEQ_LT(pcb->lastack, ackno) &&
+        TCP_SEQ_LEQ(ackno, pcb->snd_max)) { */
+      if(TCP_SEQ_BETWEEN(ackno, pcb->lastack+1, pcb->snd_max)){
       /* We come here when the ACK acknowledges new data. */
       
       /* Reset the "IN Fast Retransmit" flag, since we are no longer
@@ -813,17 +819,14 @@ tcp_receive(struct tcp_pcb *pcb)
        rationale is that lwIP puts all outstanding segments on the
        ->unsent list after a retransmission, so these segments may
        in fact have been sent once. */
-    /* KJM 13th July 2004
-       I don't think is is necessary as we no longer move all unacked
-       segments on the unsent queue when performing retransmit */
-#if 0
     while (pcb->unsent != NULL &&
-      TCP_SEQ_LEQ(ntohl(pcb->unsent->tcphdr->seqno) + TCP_TCPLEN(pcb->unsent),
-      ackno) &&
-      TCP_SEQ_LEQ(ackno, pcb->snd_max)) {
+           /*TCP_SEQ_LEQ(ntohl(pcb->unsent->tcphdr->seqno) + TCP_TCPLEN(pcb->unsent), ackno) &&
+             TCP_SEQ_LEQ(ackno, pcb->snd_max)*/
+           TCP_SEQ_BETWEEN(ackno, ntohl(pcb->unsent->tcphdr->seqno) + TCP_TCPLEN(pcb->unsent), pcb->snd_max)
+           ) {
       LWIP_DEBUGF(TCP_INPUT_DEBUG, ("tcp_receive: removing %lu:%lu from pcb->unsent\n",
-        ntohl(pcb->unsent->tcphdr->seqno), ntohl(pcb->unsent->tcphdr->seqno) +
-        TCP_TCPLEN(pcb->unsent)));
+                                    ntohl(pcb->unsent->tcphdr->seqno), ntohl(pcb->unsent->tcphdr->seqno) +
+                                    TCP_TCPLEN(pcb->unsent)));
 
       next = pcb->unsent;
       pcb->unsent = pcb->unsent->next;
@@ -840,7 +843,6 @@ tcp_receive(struct tcp_pcb *pcb)
         pcb->snd_nxt = htonl(pcb->unsent->tcphdr->seqno);
       }
     }
-#endif
     /* End of ACK for new data processing. */
 
     LWIP_DEBUGF(TCP_RTO_DEBUG, ("tcp_receive: pcb->rttest %u rtseq %lu ackno %lu\n",
@@ -903,55 +905,57 @@ tcp_receive(struct tcp_pcb *pcb)
        this if the sequence number of the incoming segment is less
        than rcv_nxt, and the sequence number plus the length of the
        segment is larger than rcv_nxt. */
-    if (TCP_SEQ_LT(seqno, pcb->rcv_nxt)){
-      if (TCP_SEQ_LT(pcb->rcv_nxt, seqno + tcplen)) {
-        /* Trimming the first edge is done by pushing the payload
-           pointer in the pbuf downwards. This is somewhat tricky since
-           we do not want to discard the full contents of the pbuf up to
-           the new starting point of the data since we have to keep the
-           TCP header which is present in the first pbuf in the chain.
-
-           What is done is really quite a nasty hack: the first pbuf in
-           the pbuf chain is pointed to by inseg.p. Since we need to be
-           able to deallocate the whole pbuf, we cannot change this
-           inseg.p pointer to point to any of the later pbufs in the
-           chain. Instead, we point the ->payload pointer in the first
-           pbuf to data in one of the later pbufs. We also set the
-           inseg.data pointer to point to the right place. This way, the
-           ->p pointer will still point to the first pbuf, but the
-           ->p->payload pointer will point to data in another pbuf.
-
-           After we are done with adjusting the pbuf pointers we must
-           adjust the ->data pointer in the seg and the segment
-           length.*/
-
-        off = pcb->rcv_nxt - seqno;
-        p = inseg.p;
-        if (inseg.p->len < off) {
-          new_tot_len = inseg.p->tot_len - off;
-          while (p->len < off) {
-            off -= p->len;
-            /* KJM following line changed (with addition of new_tot_len var)
-               to fix bug #9076
-               inseg.p->tot_len -= p->len; */
-            p->tot_len = new_tot_len;
-            p->len = 0;
-            p = p->next;
-          }
-          pbuf_header(p, -off);
-        } else {
-          pbuf_header(inseg.p, -off);
+    /*    if (TCP_SEQ_LT(seqno, pcb->rcv_nxt)){
+          if (TCP_SEQ_LT(pcb->rcv_nxt, seqno + tcplen)) {*/
+    if(TCP_SEQ_BETWEEN(pcb->rcv_nxt, seqno+1, seqno+tcplen-1)){
+      /* Trimming the first edge is done by pushing the payload
+         pointer in the pbuf downwards. This is somewhat tricky since
+         we do not want to discard the full contents of the pbuf up to
+         the new starting point of the data since we have to keep the
+         TCP header which is present in the first pbuf in the chain.
+         
+         What is done is really quite a nasty hack: the first pbuf in
+         the pbuf chain is pointed to by inseg.p. Since we need to be
+         able to deallocate the whole pbuf, we cannot change this
+         inseg.p pointer to point to any of the later pbufs in the
+         chain. Instead, we point the ->payload pointer in the first
+         pbuf to data in one of the later pbufs. We also set the
+         inseg.data pointer to point to the right place. This way, the
+         ->p pointer will still point to the first pbuf, but the
+         ->p->payload pointer will point to data in another pbuf.
+         
+         After we are done with adjusting the pbuf pointers we must
+         adjust the ->data pointer in the seg and the segment
+         length.*/
+      
+      off = pcb->rcv_nxt - seqno;
+      p = inseg.p;
+      if (inseg.p->len < off) {
+        new_tot_len = inseg.p->tot_len - off;
+        while (p->len < off) {
+          off -= p->len;
+          /* KJM following line changed (with addition of new_tot_len var)
+             to fix bug #9076
+             inseg.p->tot_len -= p->len; */
+          p->tot_len = new_tot_len;
+          p->len = 0;
+          p = p->next;
         }
-        /* KJM following line changed to use p->payload rather than inseg->p->payload
-           to fix bug #9076 */
-        inseg.dataptr = p->payload;
-        inseg.len -= pcb->rcv_nxt - seqno;
-        inseg.tcphdr->seqno = seqno = pcb->rcv_nxt;
+        pbuf_header(p, -off);
+      } else {
+        pbuf_header(inseg.p, -off);
       }
-      else{
+      /* KJM following line changed to use p->payload rather than inseg->p->payload
+         to fix bug #9076 */
+      inseg.dataptr = p->payload;
+      inseg.len -= pcb->rcv_nxt - seqno;
+      inseg.tcphdr->seqno = seqno = pcb->rcv_nxt;
+    }
+    else{
+      if(TCP_SEQ_LT(seqno, pcb->rcv_nxt)){
         /* the whole segment is < rcv_nxt */
         /* must be a duplicate of a packet that has already been correctly handled */
-
+        
         LWIP_DEBUGF(TCP_INPUT_DEBUG, ("tcp_receive: duplicate seqno %lu\n", seqno));
         tcp_ack_now(pcb);
       }
@@ -960,8 +964,9 @@ tcp_receive(struct tcp_pcb *pcb)
     /* The sequence number must be within the window (above rcv_nxt
        and below rcv_nxt + rcv_wnd) in order to be further
        processed. */
-    if (TCP_SEQ_GEQ(seqno, pcb->rcv_nxt) &&
-        TCP_SEQ_LT(seqno, pcb->rcv_nxt + pcb->rcv_wnd)) {
+    /*if (TCP_SEQ_GEQ(seqno, pcb->rcv_nxt) &&
+      TCP_SEQ_LT(seqno, pcb->rcv_nxt + pcb->rcv_wnd)) {*/
+    if(TCP_SEQ_BETWEEN(seqno, pcb->rcv_nxt, pcb->rcv_nxt + pcb->rcv_wnd - 1)){
       if (pcb->rcv_nxt == seqno) {
         /* The incoming segment is the next in sequence. We check if
            we have to trim the end of the segment and update rcv_nxt
@@ -1115,8 +1120,10 @@ tcp_receive(struct tcp_pcb *pcb)
                   }
                   break;
                 }
-              } else if (TCP_SEQ_LT(prev->tcphdr->seqno, seqno) &&
-                         TCP_SEQ_LT(seqno, next->tcphdr->seqno)) {
+              } else 
+                /*if (TCP_SEQ_LT(prev->tcphdr->seqno, seqno) &&
+                  TCP_SEQ_LT(seqno, next->tcphdr->seqno)) {*/
+                if(TCP_SEQ_BETWEEN(seqno, prev->tcphdr->seqno+1, next->tcphdr->seqno-1)){
                 /* The sequence number of the incoming segment is in
                    between the sequence numbers of the previous and
                    the next segment on ->ooseq. We trim and insert the
@@ -1162,12 +1169,19 @@ tcp_receive(struct tcp_pcb *pcb)
 #endif /* TCP_QUEUE_OOSEQ */
 
       }
+    } else {
+      /*if (TCP_SEQ_GT(pcb->rcv_nxt, seqno) ||
+        TCP_SEQ_GEQ(seqno, pcb->rcv_nxt + pcb->rcv_wnd)) {*/
+      if(!TCP_SEQ_BETWEEN(seqno, pcb->rcv_nxt, pcb->rcv_nxt + pcb->rcv_wnd-1)){
+        tcp_ack_now(pcb);
+      }
     }
   } else {
     /* Segments with length 0 is taken care of here. Segments that
        fall out of the window are ACKed. */
-    if (TCP_SEQ_GT(pcb->rcv_nxt, seqno) ||
-        TCP_SEQ_GEQ(seqno, pcb->rcv_nxt + pcb->rcv_wnd)) {
+    /*if (TCP_SEQ_GT(pcb->rcv_nxt, seqno) ||
+      TCP_SEQ_GEQ(seqno, pcb->rcv_nxt + pcb->rcv_wnd)) {*/
+    if(!TCP_SEQ_BETWEEN(seqno, pcb->rcv_nxt, pcb->rcv_nxt + pcb->rcv_wnd-1)){
       tcp_ack_now(pcb);
     }
   }
@@ -1201,7 +1215,7 @@ tcp_parseopt(struct tcp_pcb *pcb)
         ++c;
         /* NOP option. */
       } else if (opt == 0x02 &&
-                opts[c + 1] == 0x04) {
+        opts[c + 1] == 0x04) {
         /* An MSS option with the right option length. */
         mss = (opts[c + 2] << 8) | opts[c + 3];
         pcb->mss = mss > TCP_MSS? TCP_MSS: mss;

src/core/tcp_out.c

@@ -62,16 +62,21 @@ static void tcp_output_segment(struct tcp_seg *seg, struct tcp_pcb *pcb);
 err_t
 tcp_send_ctrl(struct tcp_pcb *pcb, u8_t flags)
 {
+  /* no data, no length, flags, copy=1, no optdata, no optdatalen */
   return tcp_enqueue(pcb, NULL, 0, flags, 1, NULL, 0);
-
 }
 
-/*
- * NB. tcp_write() enqueues data for sending, but does not send it
- * straight away.  It waits in the expectation of more data being sent
- * soon (as it can send them more efficiently by combining them
- * together).  To prompt the system to send data now, call
- * tcp_output() after calling tcp_write().
+/**
+ * Write data for sending (but does not send it immediately).
+ *
+ * It waits in the expectation of more data being sent soon (as
+ * it can send them more efficiently by combining them together).
+ * To prompt the system to send data now, call tcp_output() after
+ * calling tcp_write().
+ * 
+ * @arg pcb Protocol control block of the TCP connection to enqueue data for. 
+ * 
+ * @see tcp_write()
  */
 
 err_t
@@ -79,10 +84,11 @@ tcp_write(struct tcp_pcb *pcb, const void *arg, u16_t len, u8_t copy)
 {
   LWIP_DEBUGF(TCP_OUTPUT_DEBUG, ("tcp_write(pcb=%p, arg=%p, len=%u, copy=%d)\n", (void *)pcb,
     arg, len, (unsigned int)copy));
-  if (pcb->state == SYN_SENT ||
-     pcb->state == SYN_RCVD ||
-     pcb->state == ESTABLISHED ||
-     pcb->state == CLOSE_WAIT) {
+  /* connection is in valid state for data transmission? */
+  if (pcb->state == ESTABLISHED ||
+     pcb->state == CLOSE_WAIT ||
+     pcb->state == SYN_SENT ||
+     pcb->state == SYN_RCVD) {
     if (len > 0) {
       return tcp_enqueue(pcb, (void *)arg, len, 0, copy, NULL, 0);
     }
@@ -93,10 +99,24 @@ tcp_write(struct tcp_pcb *pcb, const void *arg, u16_t len, u8_t copy)
   }
 }
 
+/**
+ * Enqueue either data or TCP options (but not both) for tranmission
+ * 
+ * 
+ * 
+ * @arg pcb Protocol control block for the TCP connection to enqueue data for.
+ * @arg arg Pointer to the data to be enqueued for sending.
+ * @arg len Data length in bytes
+ * @arg flags
+ * @arg copy 1 if data must be copied, 0 if data is non-volatile and can be
+ * referenced.
+ * @arg optdata
+ * @arg optlen
+ */
 err_t
 tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
-      u8_t flags, u8_t copy,
-            u8_t *optdata, u8_t optlen)
+  u8_t flags, u8_t copy,
+  u8_t *optdata, u8_t optlen)
 {
   struct pbuf *p;
   struct tcp_seg *seg, *useg, *queue;
@@ -107,39 +127,43 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
 
   LWIP_DEBUGF(TCP_OUTPUT_DEBUG, ("tcp_enqueue(pcb=%p, arg=%p, len=%u, flags=%x, copy=%u)\n",
     (void *)pcb, arg, len, (unsigned int)flags, (unsigned int)copy));
-  left = len;
-  ptr = arg;
+  LWIP_ASSERT("tcp_enqueue: len == 0 || optlen == 0 (programmer violates API)",
+      len == 0 || optlen == 0);
+  LWIP_ASSERT("tcp_enqueue: arg == NULL || optdata == NULL (programmer violates API)",
+      arg == NULL || optdata == NULL);
   /* fail on too much data */
   if (len > pcb->snd_buf) {
     LWIP_DEBUGF(TCP_OUTPUT_DEBUG | 3, ("tcp_enqueue: too much data (len=%u > snd_buf=%u)\n", len, pcb->snd_buf));
     return ERR_MEM;
   }
+  left = len;
+  ptr = arg;
 
   /* seqno will be the sequence number of the first segment enqueued
    * by the call to this function. */
   seqno = pcb->snd_lbb;
 
-  queue = NULL;
   LWIP_DEBUGF(TCP_QLEN_DEBUG, ("tcp_enqueue: queuelen: %u\n", (unsigned int)pcb->snd_queuelen));
 
-  /* Check if the queue length exceeds the configured maximum queue
-   * length. If so, we return an error. */
+  /* If total number of pbufs on the unsent/unacked queues exceeds the
+   * configured maximum, return an error */
   queuelen = pcb->snd_queuelen;
   if (queuelen >= TCP_SND_QUEUELEN) {
     LWIP_DEBUGF(TCP_OUTPUT_DEBUG | 3, ("tcp_enqueue: too long queue %u (max %u)\n", queuelen, TCP_SND_QUEUELEN));
     goto memerr;
   }
-
-  if (pcb->snd_queuelen != 0) {
-    LWIP_ASSERT("tcp_enqueue: valid queue length", pcb->unacked != NULL ||
-    pcb->unsent != NULL);
+  if (queuelen != 0) {
+    LWIP_ASSERT("tcp_enqueue: pbufs on queue => at least one queue non-empty",
+      pcb->unacked != NULL || pcb->unsent != NULL);
+  } else {
+    LWIP_ASSERT("tcp_enqueue: no pbufs on queue => both queues empty",
+      pcb->unacked == NULL && pcb->unsent == NULL);
   }
 
-  seg = useg = NULL;
-  seglen = 0;
-
   /* First, break up the data into segments and tuck them together in
    * the local "queue" variable. */
+  useg = queue = seg = NULL;
+  seglen = 0;
   while (queue == NULL || left > 0) {
 
     /* The segment length should be the MSS if the data to be enqueued
@@ -155,20 +179,25 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
     seg->next = NULL;
     seg->p = NULL;
 
+    /* first segment of to-be-queued data? */
     if (queue == NULL) {
-      useg = queue = seg;
+      queue = seg;
     }
+    /* subsequent segments of to-be-queued data */
     else {
-      /* Attach the segment to the end of the queued segments. */
+      /* Attach the segment to the end of the queued segments */
       LWIP_ASSERT("useg != NULL", useg != NULL);
       useg->next = seg;
-      useg = seg;
     }
+    /* remember last segment of to-be-queued data for next iteration */
+    useg = seg;
 
     /* If copy is set, memory should be allocated
      * and data copied into pbuf, otherwise data comes from
      * ROM or other static memory, and need not be copied. If
      * optdata is != NULL, we have options instead of data. */
+     
+    /* options? */
     if (optdata != NULL) {
       if ((seg->p = pbuf_alloc(PBUF_TRANSPORT, optlen, PBUF_RAM)) == NULL) {
         goto memerr;
@@ -176,6 +205,7 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
       ++queuelen;
       seg->dataptr = seg->p->payload;
     }
+    /* copy from volatile memory? */
     else if (copy) {
       if ((seg->p = pbuf_alloc(PBUF_TRANSPORT, seglen, PBUF_RAM)) == NULL) {
         LWIP_DEBUGF(TCP_OUTPUT_DEBUG | 2, ("tcp_enqueue : could not allocate memory for pbuf copy size %u\n", seglen));
@@ -189,8 +219,7 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
     }
     /* do not copy data */
     else {
-
-      /* first, allocate a pbuf for holding the data.
+      /* First, allocate a pbuf for holding the data.
        * since the referenced data is available at least until it is sent out on the
        * link (as it has to be ACKed by the remote party) we can safely use PBUF_ROM
        * instead of PBUF_REF here.
@@ -200,6 +229,7 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
         goto memerr;
       }
       ++queuelen;
+      /* reference the non-volatile payload data */
       p->payload = ptr;
       seg->dataptr = ptr;
 
@@ -214,7 +244,7 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
       ++queuelen;
 
       /* Concatenate the headers and data pbufs together. */
-      pbuf_cat(seg->p, p);
+      pbuf_cat(seg->p/*header*/, p/*data*/);
       p = NULL;
     }
 
@@ -226,16 +256,10 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
     }
 
     seg->len = seglen;
-#if 0 /* Was commented out. TODO: can someone say why this is here? */
-    if ((flags & TCP_SYN) || (flags & TCP_FIN)) {
-      ++seg->len;
-    }
-#endif
-    /* Build TCP header. */
+
+    /* build TCP header */
     if (pbuf_header(seg->p, TCP_HLEN)) {
-
       LWIP_DEBUGF(TCP_OUTPUT_DEBUG | 2, ("tcp_enqueue: no room for TCP header in pbuf.\n"));
-
       TCP_STATS_INC(tcp.err);
       goto memerr;
     }
@@ -268,7 +292,6 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
     ptr = (void *)((char *)ptr + seglen);
   }
 
-
   /* Now that the data to be enqueued has been broken up into TCP
   segments in the queue variable, we add them to the end of the
   pcb->unsent queue. */
@@ -278,6 +301,7 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
   else {
     for (useg = pcb->unsent; useg->next != NULL; useg = useg->next);
   }
+  /* { useg is last segment on the unsent queue, NULL if list is empty } */
 
   /* If there is room in the last pbuf on the unsent queue,
   chain the first pbuf on the queue together with that. */
@@ -285,24 +309,27 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
     TCP_TCPLEN(useg) != 0 &&
     !(TCPH_FLAGS(useg->tcphdr) & (TCP_SYN | TCP_FIN)) &&
     !(flags & (TCP_SYN | TCP_FIN)) &&
+    /* fit within max seg size */
     useg->len + queue->len <= pcb->mss) {
-    /* Remove TCP header from first segment. */
+    /* Remove TCP header from first segment of our to-be-queued list */
     pbuf_header(queue->p, -TCP_HLEN);
     pbuf_cat(useg->p, queue->p);
     useg->len += queue->len;
     useg->next = queue->next;
 
-    LWIP_DEBUGF(TCP_OUTPUT_DEBUG | DBG_TRACE | DBG_STATE, ("tcp_enqueue: chaining, new len %u\n", useg->len));
+    LWIP_DEBUGF(TCP_OUTPUT_DEBUG | DBG_TRACE | DBG_STATE, ("tcp_enqueue: chaining segments, new len %u\n", useg->len));
     if (seg == queue) {
       seg = NULL;
     }
     memp_free(MEMP_TCP_SEG, queue);
   }
   else {
+    /* empty list */
     if (useg == NULL) {
+      /* initialize list with this segment */
       pcb->unsent = queue;
-
     }
+    /* enqueue segment */
     else {
       useg->next = queue;
     }
@@ -312,12 +339,12 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
   }
   pcb->snd_lbb += len;
   pcb->snd_buf -= len;
+  /* update number of segments on the queues */
   pcb->snd_queuelen = queuelen;
   LWIP_DEBUGF(TCP_QLEN_DEBUG, ("tcp_enqueue: %d (after enqueued)\n", pcb->snd_queuelen));
   if (pcb->snd_queuelen != 0) {
-    LWIP_ASSERT("tcp_enqueue: valid queue length", pcb->unacked != NULL ||
-      pcb->unsent != NULL);
-
+    LWIP_ASSERT("tcp_enqueue: valid queue length",
+      pcb->unacked != NULL || pcb->unsent != NULL);
   }
 
   /* Set the PSH flag in the last segment that we enqueued, but only
@@ -336,7 +363,6 @@ tcp_enqueue(struct tcp_pcb *pcb, void *arg, u16_t len,
   if (pcb->snd_queuelen != 0) {
     LWIP_ASSERT("tcp_enqueue: valid queue length", pcb->unacked != NULL ||
       pcb->unsent != NULL);
-
   }
   LWIP_DEBUGF(TCP_QLEN_DEBUG | DBG_STATE, ("tcp_enqueue: %d (with mem err)\n", pcb->snd_queuelen));
   return ERR_MEM;
@@ -364,7 +390,6 @@ tcp_output(struct tcp_pcb *pcb)
 
   wnd = LWIP_MIN(pcb->snd_wnd, pcb->cwnd);
 
-
   seg = pcb->unsent;
 
   /* useg should point to last segment on unacked queue */
@@ -372,24 +397,24 @@ tcp_output(struct tcp_pcb *pcb)
   if (useg != NULL) {
     for (; useg->next != NULL; useg = useg->next);
   }                                                                             
-
    
-  /* If the TF_ACK_NOW flag is set, we check if there is data that is
-     to be sent. If data is to be sent out, we'll just piggyback our
-     acknowledgement with the outgoing segment. If no data will be
-     sent (either because the ->unsent queue is empty or because the
-     window doesn't allow it) we'll have to construct an empty ACK
-     segment and send it. */
+  /* If the TF_ACK_NOW flag is set and no data will be sent (either
+   * because the ->unsent queue is empty or because the window does
+   * not allow it), construct an empty ACK segment and send it.
+   *
+   * If data is to be sent, we will just piggyback the ACK (see below).
+   */
   if (pcb->flags & TF_ACK_NOW &&
      (seg == NULL ||
       ntohl(seg->tcphdr->seqno) - pcb->lastack + seg->len > wnd)) {
-    pcb->flags &= ~(TF_ACK_DELAY | TF_ACK_NOW);
     p = pbuf_alloc(PBUF_IP, TCP_HLEN, PBUF_RAM);
     if (p == NULL) {
       LWIP_DEBUGF(TCP_OUTPUT_DEBUG, ("tcp_output: (ACK) could not allocate pbuf\n"));
       return ERR_BUF;
     }
     LWIP_DEBUGF(TCP_OUTPUT_DEBUG, ("tcp_output: sending ACK for %lu\n", pcb->rcv_nxt));
+    /* remove ACK flags from the PCB, as we send an empty ACK now */
+    pcb->flags &= ~(TF_ACK_DELAY | TF_ACK_NOW);
 
     tcphdr = p->payload;
     tcphdr->src = htons(pcb->local_port);
@@ -406,7 +431,6 @@ tcp_output(struct tcp_pcb *pcb)
     tcphdr->chksum = inet_chksum_pseudo(p, &(pcb->local_ip), &(pcb->remote_ip),
           IP_PROTO_TCP, p->tot_len);
 #endif
-
     ip_output(p, &(pcb->local_ip), &(pcb->remote_ip), pcb->ttl, pcb->tos,
         IP_PROTO_TCP);
     pbuf_free(p);
@@ -431,7 +455,7 @@ tcp_output(struct tcp_pcb *pcb)
                             ntohl(seg->tcphdr->seqno), pcb->lastack));
   }
 #endif /* TCP_CWND_DEBUG */
-
+  /* data available and window allows it to be sent? */
   while (seg != NULL &&
   ntohl(seg->tcphdr->seqno) - pcb->lastack + seg->len <= wnd) {
 #if TCP_CWND_DEBUG
@@ -458,13 +482,26 @@ tcp_output(struct tcp_pcb *pcb)
     /* put segment on unacknowledged list if length > 0 */
     if (TCP_TCPLEN(seg) > 0) {
       seg->next = NULL;
+      /* unacked list is empty? */
       if (pcb->unacked == NULL) {
         pcb->unacked = seg;
         useg = seg;
+      /* unacked list is not empty? */
       } else {
-        useg->next = seg;
-        useg = useg->next;
+        /* In the case of fast retransmit, the packet should not go to the tail
+         * of the unacked queue, but rather at the head. We need to check for
+         * this case. -STJ Jul 27, 2004 */
+        if (TCP_SEQ_LT(ntohl(seg->tcphdr->seqno), ntohl(useg->tcphdr->seqno))){
+          /* add segment to head of unacked list */
+          seg->next = pcb->unacked;
+          pcb->unacked = seg;
+        } else {
+          /* add segment to tail of unacked list */
+          useg->next = seg;
+          useg = useg->next;
+        }
       }
+    /* do not queue empty segments on the unacked list */
     } else {
       tcp_seg_free(seg);
     }
@@ -473,6 +510,9 @@ tcp_output(struct tcp_pcb *pcb)
   return ERR_OK;
 }
 
+/**
+ * Actually send a TCP segment over IP
+ */
 static void
 tcp_output_segment(struct tcp_seg *seg, struct tcp_pcb *pcb)
 {
@@ -568,6 +608,36 @@ tcp_rst(u32_t seqno, u32_t ackno,
   LWIP_DEBUGF(TCP_RST_DEBUG, ("tcp_rst: seqno %lu ackno %lu.\n", seqno, ackno));
 }
 
+/* requeue all unacked segments for retransmission */
+void
+tcp_rexmit_rto(struct tcp_pcb *pcb)
+{
+  struct tcp_seg *seg;
+
+  if (pcb->unacked == NULL) {
+    return;
+  }
+
+  /* Move all unacked segments to the head of the unsent queue */
+  for (seg = pcb->unacked; seg->next != NULL; seg = seg->next);
+  /* concatenate unsent queue after unacked queue */
+  seg->next = pcb->unsent;
+  /* unsent queue is the concatenated queue (of unacked, unsent) */
+  pcb->unsent = pcb->unacked;
+  /* unacked queue is now empty */
+  pcb->unacked = NULL;
+
+  pcb->snd_nxt = ntohl(pcb->unsent->tcphdr->seqno);
+  /* increment number of retransmissions */
+  ++pcb->nrtx;
+
+  /* Don't take any RTT measurements after retransmitting. */
+  pcb->rttest = 0;
+
+  /* Do the actual retransmission */
+  tcp_output(pcb);
+}
+
 void
 tcp_rexmit(struct tcp_pcb *pcb)
 {
@@ -595,6 +665,7 @@ tcp_rexmit(struct tcp_pcb *pcb)
 
 }
 
+
 void
 tcp_keepalive(struct tcp_pcb *pcb)
 {

src/core/udp.c

@@ -356,9 +356,10 @@ udp_sendto(struct udp_pcb *pcb, struct pbuf *p,
   struct ip_addr *dst_ip, u16_t dst_port)
 {
   err_t err;
+  /* temporary space for current PCB remote address */
   struct ip_addr pcb_remote_ip;
   u16_t pcb_remote_port;
-  /* remember remote peer address of PCB */
+  /* remember current remote peer address of PCB */
   pcb_remote_ip.addr = pcb->remote_ip.addr;
   pcb_remote_port = pcb->remote_port;
   /* copy packet destination address to PCB remote peer address */
@@ -366,7 +367,7 @@ udp_sendto(struct udp_pcb *pcb, struct pbuf *p,
   pcb->remote_port = dst_port;
   /* send to the packet destination address */
   err = udp_send(pcb, p);
-  /* reset PCB remote peer address */
+  /* restore PCB remote peer address */
   pcb->remote_ip.addr = pcb_remote_ip.addr;
   pcb->remote_port = pcb_remote_port;
   return err;
@@ -433,7 +434,9 @@ udp_send(struct udp_pcb *pcb, struct pbuf *p)
   udphdr->chksum = 0x0000; 
 
   /* find the outgoing network interface for this packet */
-  if ((netif = ip_route(&(pcb->remote_ip))) == NULL) {
+  netif = ip_route(&(pcb->remote_ip));
+  /* no outgoing network interface could be found? */
+  if (netif == NULL) {
     LWIP_DEBUGF(UDP_DEBUG | 1, ("udp_send: No route to 0x%lx\n", pcb->remote_ip.addr));
     UDP_STATS_INC(udp.rterr);
     return ERR_RTE;

src/include/ipv4/lwip/ip.h

@@ -40,7 +40,6 @@
 
 #include "lwip/err.h"
 
-struct netif;
 
 void ip_init(void);
 struct netif *ip_route(struct ip_addr *dest);

src/include/ipv4/lwip/ip_addr.h

@@ -117,7 +117,15 @@ extern const struct ip_addr ip_addr_broadcast;
 #define ip_addr_set(dest, src) (dest)->addr = \
                                ((src) == NULL? 0:\
                                (src)->addr)
-#define ip_addr_maskcmp(addr1, addr2, mask) (((addr1)->addr & \
+/**
+ * Determine if two address are on the same network.
+ *
+ * @arg addr1 IP address 1
+ * @arg addr2 IP address 2
+ * @arg mask network identifier mask
+ * @return !0 if the network identifiers of both address match
+ */
+#define ip_addr_netcmp(addr1, addr2, mask) (((addr1)->addr & \
                                               (mask)->addr) == \
                                              ((addr2)->addr & \
                                               (mask)->addr))

src/include/ipv6/lwip/ip_addr.h

@@ -45,7 +45,7 @@ struct ip_addr {
                                                (ipaddr)->addr[2] = htonl(((e & 0xffff) << 16) | (f & 0xffff)); \
                                                (ipaddr)->addr[3] = htonl(((g & 0xffff) << 16) | (h & 0xffff)); } while(0)
 
-int ip_addr_maskcmp(struct ip_addr *addr1, struct ip_addr *addr2,
+int ip_addr_netcmp(struct ip_addr *addr1, struct ip_addr *addr2,
         struct ip_addr *mask);
 int ip_addr_cmp(struct ip_addr *addr1, struct ip_addr *addr2);
 void ip_addr_set(struct ip_addr *dest, struct ip_addr *src);

src/include/lwip/dhcp.h

@@ -86,8 +86,12 @@ struct dhcp_msg
   PACK_STRUCT_FIELD(u8_t file[DHCP_FILE_LEN]);
   PACK_STRUCT_FIELD(u32_t cookie);
 #define DHCP_MIN_OPTIONS_LEN 68U
+/** make sure user does not configure this too small */
+#if ((defined(DHCP_OPTIONS_LEN)) && (DHCP_OPTIONS_LEN < DHCP_MIN_OPTIONS_LEN))
+#  undef DHCP_OPTIONS_LEN
+#endif
 /** allow this to be configured in lwipopts.h, but not too small */
-#if ((!defined(DHCP_OPTIONS_LEN)) || (DHCP_OPTIONS_LEN < DHCP_MIN_OPTIONS_LEN))
+#if (!defined(DHCP_OPTIONS_LEN))
 /** set this to be sufficient for your options in outgoing DHCP msgs */
 #  define DHCP_OPTIONS_LEN DHCP_MIN_OPTIONS_LEN
 #endif
@@ -100,11 +104,13 @@ PACK_STRUCT_END
 
 /** start DHCP configuration */
 err_t dhcp_start(struct netif *netif);
+/** enforce early lease renewal (not needed normally)*/
+err_t dhcp_renew(struct netif *netif);
+/** release the DHCP lease, usually called before dhcp_stop()*/
+err_t dhcp_release(struct netif *netif);
 /** stop DHCP configuration */
 void dhcp_stop(struct netif *netif);
-/** enforce lease renewal */
-err_t dhcp_renew(struct netif *netif);
-/** inform server of our IP address */
+/** inform server of our manual IP address */
 void dhcp_inform(struct netif *netif);
 
 /** if enabled, check whether the offered IP address is not in use, using ARP */

src/include/lwip/opt.h

@@ -146,10 +146,10 @@ a lot of data that needs to be copied, this should be set high. */
 #endif
 
 /* PBUF_LINK_HLEN: the number of bytes that should be allocated for a
-   link level header. */
+   link level header. Defaults to 14 for Ethernet. */
 
 #ifndef PBUF_LINK_HLEN
-#define PBUF_LINK_HLEN                  0
+#define PBUF_LINK_HLEN                  14
 #endif
 
 
@@ -163,7 +163,15 @@ a lot of data that needs to be copied, this should be set high. */
 
 /**
  * If enabled, outgoing packets are queued during hardware address
- * resolution. The etharp.c implementation queues 1 packet only.
+ * resolution.
+ *
+ * This feature has not stabilized yet. Single-packet queueing is
+ * believed to be stable, multi-packet queueing is believed to
+ * clash with the TCP segment queueing.
+ * 
+ * As multi-packet-queueing is currently disabled, enabling this
+ * _should_ work, but we need your testing feedback on lwip-users.
+ *
  */
 #ifndef ARP_QUEUEING
 #define ARP_QUEUEING                    1
@@ -339,7 +347,7 @@ a lot of data that needs to be copied, this should be set high. */
 /* ---------- Socket Options ---------- */
 /* Enable SO_REUSEADDR and SO_REUSEPORT options */ 
 #ifndef SO_REUSE
-# define SO_REUSE 1
+# define SO_REUSE 0
 #endif                                                                        
 
 

src/include/lwip/tcp.h

@@ -105,6 +105,7 @@ void             tcp_input   (struct pbuf *p, struct netif *inp);
 /* Used within the TCP code only: */
 err_t            tcp_output  (struct tcp_pcb *pcb);
 void             tcp_rexmit  (struct tcp_pcb *pcb);
+void             tcp_rexmit_rto  (struct tcp_pcb *pcb);
 
 
 
@@ -112,7 +113,11 @@ void             tcp_rexmit  (struct tcp_pcb *pcb);
 #define TCP_SEQ_LEQ(a,b)    ((s32_t)((a)-(b)) <= 0)
 #define TCP_SEQ_GT(a,b)     ((s32_t)((a)-(b)) > 0)
 #define TCP_SEQ_GEQ(a,b)    ((s32_t)((a)-(b)) >= 0)
-
+/* is b<=a<=c? */
+#if 0 /* see bug #10548 */
+#define TCP_SEQ_BETWEEN(a,b,c) ((c)-(b) >= (a)-(b))
+#endif
+#define TCP_SEQ_BETWEEN(a,b,c) (TCP_SEQ_GEQ(a,b) && TCP_SEQ_LEQ(a,c))
 #define TCP_FIN 0x01U
 #define TCP_SYN 0x02U
 #define TCP_RST 0x04U
@@ -208,18 +213,13 @@ enum tcp_state {
   TIME_WAIT   = 10
 };
 
-
 /* the TCP protocol control block */
 struct tcp_pcb {
-/* Common members of all PCB types */
+/** common PCB members */
   IP_PCB;
-
-/* Protocol specific PCB members */
-
-  struct tcp_pcb *next;   /* for the linked list */
-
-  enum tcp_state state;   /* TCP state */
-
+/** protocol specific PCB members */
+  struct tcp_pcb *next; /* for the linked list */
+  enum tcp_state state; /* TCP state */
   u8_t prio;
   void *callback_arg;
 
@@ -235,7 +235,7 @@ struct tcp_pcb {
 #define TF_GOT_FIN   (u8_t)0x20U   /* Connection was closed by the remote end. */
 #define TF_NODELAY   (u8_t)0x40U   /* Disable Nagle algorithm */
 
-  /* receiver varables */
+  /* receiver variables */
   u32_t rcv_nxt;   /* next seqno expected */
   u16_t rcv_wnd;   /* receiver window */
   
@@ -248,10 +248,10 @@ struct tcp_pcb {
   
   u16_t mss;   /* maximum segment size */
   
-  /* RTT estimation variables. */
-  u16_t rttest; /* RTT estimate in 500ms ticks */
+  /* RTT (round trip time) estimation variables */
+  u32_t rttest; /* RTT estimate in 500ms ticks */
   u32_t rtseq;  /* sequence number being timed */
-  s16_t sa, sv;
+  s16_t sa, sv; /* @todo document this */
 
   u16_t rto;    /* retransmission time-out */
   u8_t nrtx;    /* number of retransmissions */
@@ -375,7 +375,7 @@ err_t lwip_tcp_event(void *arg, struct tcp_pcb *pcb,
 #define TCP_EVENT_RECV(pcb,p,err,ret) \
                         if((pcb)->recv != NULL) \
                         { ret = (pcb)->recv((pcb)->callback_arg,(pcb),(p),(err)); } else { \
-            pbuf_free(p); }
+                          if (p) pbuf_free(p); }
 #define TCP_EVENT_CONNECTED(pcb,err,ret) \
                         if((pcb)->connected != NULL) \
                         (ret = (pcb)->connected((pcb)->callback_arg,(pcb),(err)))
@@ -387,7 +387,7 @@ err_t lwip_tcp_event(void *arg, struct tcp_pcb *pcb,
                         (errf)((arg),(err))
 #endif /* LWIP_EVENT_API */
 
-/* This structure is used to repressent TCP segments when queued. */
+/* This structure represents a TCP segment on the unsent and unacked queues */
 struct tcp_seg {
   struct tcp_seg *next;    /* used when putting segements on a queue */
   struct pbuf *p;          /* buffer containing data + TCP header */

src/include/netif/etharp.h

@@ -52,7 +52,13 @@ struct eth_addr {
   PACK_STRUCT_FIELD(u8_t addr[6]);
 } PACK_STRUCT_STRUCT;
 PACK_STRUCT_END
+#ifdef PACK_STRUCT_USE_INCLUDES
+#  include "arch/epstruct.h"
+#endif
 
+#ifdef PACK_STRUCT_USE_INCLUDES
+#  include "arch/bpstruct.h"
+#endif
 PACK_STRUCT_BEGIN
 struct eth_hdr {
 #if ETH_PAD_SIZE
@@ -63,7 +69,13 @@ struct eth_hdr {
   PACK_STRUCT_FIELD(u16_t type);
 } PACK_STRUCT_STRUCT;
 PACK_STRUCT_END
+#ifdef PACK_STRUCT_USE_INCLUDES
+#  include "arch/epstruct.h"
+#endif
 
+#ifdef PACK_STRUCT_USE_INCLUDES
+#  include "arch/bpstruct.h"
+#endif
 PACK_STRUCT_BEGIN
 /** the ARP message */
 struct etharp_hdr {
@@ -78,17 +90,25 @@ struct etharp_hdr {
   PACK_STRUCT_FIELD(struct ip_addr2 dipaddr);
 } PACK_STRUCT_STRUCT;
 PACK_STRUCT_END
-
-PACK_STRUCT_BEGIN
-struct ethip_hdr {
-  PACK_STRUCT_FIELD(struct eth_hdr eth);
-  PACK_STRUCT_FIELD(struct ip_hdr ip);
-};
 #ifdef PACK_STRUCT_USE_INCLUDES
 #  include "arch/epstruct.h"
 #endif
 
-#define ARP_TMR_INTERVAL 10000
+#ifdef PACK_STRUCT_USE_INCLUDES
+#  include "arch/bpstruct.h"
+#endif
+PACK_STRUCT_BEGIN
+struct ethip_hdr {
+  PACK_STRUCT_FIELD(struct eth_hdr eth);
+  PACK_STRUCT_FIELD(struct ip_hdr ip);
+} PACK_STRUCT_STRUCT;
+PACK_STRUCT_END
+#ifdef PACK_STRUCT_USE_INCLUDES
+#  include "arch/epstruct.h"
+#endif
+
+/** 5 seconds period */
+#define ARP_TMR_INTERVAL 5000
 
 #define ETHTYPE_ARP 0x0806
 #define ETHTYPE_IP  0x0800

src/netif/FILES

@@ -1,17 +1,18 @@
 This directory contains generic network interface device drivers that
-don't contain any hardware or architecture specific code. The files
+do not contain any hardware or architecture specific code. The files
 are:
 
 etharp.c
           Implements the ARP (Address Resolution Protocol) over
           Ethernet. The code in this file should be used together with
-          Ethernet device drivers.
+          Ethernet device drivers. Note that this module has been
+          largely made Ethernet independent so you should be able to
+          adapt this for other link layers (such as Firewire).
 
 ethernetif.c
           An example of how an Ethernet device driver could look. This
           file can be used as a "skeleton" for developing new Ethernet
           network device drivers. It uses the etharp.c ARP code.
-	  !!!THIS FILE IS NOT IN SYNC WITH CURRENT CODE !!!!
 
 loopif.c
           An example network interface that shows how a "loopback"

src/netif/etharp.c

@@ -55,10 +55,17 @@
 #  include "lwip/dhcp.h"
 #endif
 
-/** the time an ARP entry stays valid after its last update, (120 * 10) seconds = 20 minutes. */
-#define ARP_MAXAGE 120
-/** the time an ARP entry stays pending after first request, (1 * 10) seconds = 10 seconds. */
-#define ARP_MAXPENDING 1
+/** the time an ARP entry stays valid after its last update,
+ * (240 * 5) seconds = 20 minutes.
+ */
+#define ARP_MAXAGE 240
+/** the time an ARP entry stays pending after first request,
+ * (2 * 5) seconds = 10 seconds.
+ * 
+ * @internal Keep this number at least 2, otherwise it might
+ * run out instantly if the timeout occurs directly after a request.
+ */
+#define ARP_MAXPENDING 2
 
 #define HWTYPE_ETHERNET 1
 
@@ -96,9 +103,11 @@ struct etharp_entry {
 static const struct eth_addr ethbroadcast = {{0xff,0xff,0xff,0xff,0xff,0xff}};
 static struct etharp_entry arp_table[ARP_TABLE_SIZE];
 
-/** ask update_arp_entry() to create new entry instead of merely update existing */
-/** ask find_entry() to create new entry instead of merely finding existing */
-#define ETHARP_CREATE 1
+/**
+ * Try hard to create a new entry - we want the IP address to appear in
+ * the cache (even if this means removing an active entry or so). */
+#define ETHARP_TRY_HARD 1
+
 static s8_t find_entry(struct ip_addr *ipaddr, u8_t flags);
 static err_t update_arp_entry(struct netif *netif, struct ip_addr *ipaddr, struct eth_addr *ethaddr, u8_t flags);
 /**
@@ -121,7 +130,7 @@ etharp_init(void)
 /**
  * Clears expired entries in the ARP table.
  *
- * This function should be called every ETHARP_TMR_INTERVAL microseconds (10 seconds),
+ * This function should be called every ETHARP_TMR_INTERVAL microseconds (5 seconds),
  * in order to expire entries in the ARP table.
  */
 void
@@ -169,7 +178,7 @@ etharp_tmr(void)
 }
 
 /**
- * Search the ARP table for a specific entry.
+ * Search the ARP table for a matching or new entry.
  * 
  * If an IP address is given, return a pending or stable ARP entry that matches
  * the address. If no match is found, create a new entry with this address set,
@@ -179,28 +188,29 @@ etharp_tmr(void)
  * If ipaddr is NULL, return a initialized new entry in state ETHARP_EMPTY.
  * 
  * In all cases, attempt to create new entries from an empty entry. If no
- * empty entries are available and ETHARP_CREATE flag is set, recycle
+ * empty entries are available and ETHARP_TRY_HARD flag is set, recycle
  * old entries. Heuristic choose the least important entry for recycling.
  *
  * @param ipaddr IP address to find in ARP cache, or to add if not found.
  * @param flags
- * - ETHARP_CREATE: Try hard to create a entry by allowing recycling.
+ * - ETHARP_TRY_HARD: Try hard to create a entry by allowing recycling of
+ * active (stable or pending) entries.
  *  
  * @return The ARP entry index that matched or is created, ERR_MEM if no
  * entry is found or could be recycled.
  */
 static s8_t find_entry(struct ip_addr *ipaddr, u8_t flags)
 {
-  s8_t old_pending, old_stable, empty, i;
-  u8_t age_pending, age_stable;
+  s8_t old_pending = ARP_TABLE_SIZE, old_stable = ARP_TABLE_SIZE;
+  s8_t empty = ARP_TABLE_SIZE;
+  u8_t i = 0, age_pending = 0, age_stable = 0;
 #if ARP_QUEUEING
+  /* oldest entry with packets on queue */
   s8_t old_queue = ARP_TABLE_SIZE;
+  /* its age */
   u8_t age_queue = 0;
 #endif
 
-  old_pending = old_stable = empty = ARP_TABLE_SIZE;
-  age_pending = age_stable = 0;
-
   /**
    * a) do a search through the cache, remember candidates
    * b) select candidate entry
@@ -228,7 +238,7 @@ static s8_t find_entry(struct ip_addr *ipaddr, u8_t flags)
       /* if given, does IP address match IP address in ARP entry? */
       if (ipaddr && ip_addr_cmp(ipaddr, &arp_table[i].ipaddr)) {
         LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("find_entry: found matching pending entry %d\n", i));
-        /* found match, simply bail out */
+        /* found exact IP address match, simply bail out */
         return i;
 #if ARP_QUEUEING
       /* pending with queued packets? */
@@ -251,7 +261,7 @@ static s8_t find_entry(struct ip_addr *ipaddr, u8_t flags)
       /* if given, does IP address match IP address in ARP entry? */
       if (ipaddr && ip_addr_cmp(ipaddr, &arp_table[i].ipaddr)) {
         LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("find_entry: found matching stable entry %d\n", i));
-        /* found match, simply bail out */
+        /* found exact IP address match, simply bail out */
         return i;
       /* remember entry with oldest stable entry in oldest, its age in maxtime */
       } else if (arp_table[i].ctime >= age_stable) {
@@ -260,12 +270,21 @@ static s8_t find_entry(struct ip_addr *ipaddr, u8_t flags)
       }
     }
   }
+  /* { we have no match } => try to create a new entry */
+   
+  /* no empty entry found and not allowed to recycle? */
+  if ((empty == ARP_TABLE_SIZE) && ((flags & ETHARP_TRY_HARD) == 0))
+  {
+  	return (s8_t)ERR_MEM;
+  }
   
   /* b) choose the least destructive entry to recycle:
    * 1) empty entry
    * 2) oldest stable entry
    * 3) oldest pending entry without queued packets
    * 4) oldest pending entry without queued packets
+   * 
+   * { ETHARP_TRY_HARD is set at this point }
    */ 
 
   /* 1) empty entry available? */
@@ -293,41 +312,26 @@ static s8_t find_entry(struct ip_addr *ipaddr, u8_t flags)
     /* recycle oldest pending */
     i = old_queue;
     LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("find_entry: selecting oldest pending entry %d, freeing packet queue %p\n", i, (void *)(arp_table[i].p)));
-    /* no empty or recyclable entries found */
+    pbuf_free(arp_table[i].p);
+    arp_table[i].p = NULL;
 #endif
+    /* no empty or recyclable entries found */
   } else {
-    return ERR_MEM;
+    return (s8_t)ERR_MEM;
   }
 
   /* { empty or recyclable entry found } */
-  LWIP_ASSERT("i >= 0", i >= 0);
   LWIP_ASSERT("i < ARP_TABLE_SIZE", i < ARP_TABLE_SIZE);
 
-  /* allowed to recycle a entry? */
-  if (flags & ETHARP_CREATE) {
-    /* recycle (no-op for an already empty entry) */
-    arp_table[i].state = ETHARP_STATE_EMPTY;
-  }
+  /* recycle entry (no-op for an already empty entry) */
+  arp_table[i].state = ETHARP_STATE_EMPTY;
 
-  /* empty entry found or created? */
-  if (arp_table[i].state == ETHARP_STATE_EMPTY) {
-    /* IP address given? */
-    if (ipaddr != NULL) {
-      /* set IP address */
-      ip_addr_set(&arp_table[i].ipaddr, ipaddr);
-    }
-    arp_table[i].ctime = 0;
-#if ARP_QUEUEING
-    /* remove any queued packets */
-    if (arp_table[i].p != NULL) pbuf_free(arp_table[i].p);
-    arp_table[i].p = NULL;
-#endif
-  /* no entry available */
-  } else {
-    /* return failure */
-    i = (s8_t)ERR_MEM;
+  /* IP address given? */
+  if (ipaddr != NULL) {
+    /* set IP address */
+    ip_addr_set(&arp_table[i].ipaddr, ipaddr);
   }
-        
+  arp_table[i].ctime = 0;
   return (err_t)i;
 }
 
@@ -340,17 +344,17 @@ static s8_t find_entry(struct ip_addr *ipaddr, u8_t flags)
  * @param ipaddr IP address of the inserted ARP entry.
  * @param ethaddr Ethernet address of the inserted ARP entry.
  * @param flags Defines behaviour:
- * - ETHARP_CREATE Allows ARP to insert this as a new item. If not specified,
+ * - ETHARP_TRY_HARD Allows ARP to insert this as a new item. If not specified,
  * only existing ARP entries will be updated.
  *
  * @return
  * - ERR_OK Succesfully updated ARP cache.
- * - ERR_MEM If we could not add a new ARP entry when ETHARP_CREATE was set.
+ * - ERR_MEM If we could not add a new ARP entry when ETHARP_TRY_HARD was set.
  * - ERR_ARG Non-unicast address given, those will not appear in ARP cache.
  *
  * @see pbuf_free()
  */
-err_t
+static err_t
 update_arp_entry(struct netif *netif, struct ip_addr *ipaddr, struct eth_addr *ethaddr, u8_t flags)
 {
   s8_t i, k;
@@ -385,10 +389,10 @@ update_arp_entry(struct netif *netif, struct ip_addr *ipaddr, struct eth_addr *e
 /* this is where we will send out queued packets! */
 #if ARP_QUEUEING
   while (arp_table[i].p != NULL) {
-    /* get the first packet on the queue (if any) */
+    /* get the first packet on the queue */
     struct pbuf *p = arp_table[i].p;
     /* Ethernet header */
-    struct eth_hdr *ethhdr = p->payload;;
+    struct eth_hdr *ethhdr = p->payload;
     /* remember (and reference) remainder of queue */
     /* note: this will also terminate the p pbuf chain */
     arp_table[i].p = pbuf_dequeue(p);
@@ -431,15 +435,17 @@ etharp_ip_input(struct netif *netif, struct pbuf *p)
   /* Only insert an entry if the source IP address of the
      incoming IP packet comes from a host on the local network. */
   hdr = p->payload;
-  /* source is on local network? */
-  if (!ip_addr_maskcmp(&(hdr->ip.src), &(netif->ip_addr), &(netif->netmask))) {
+  /* source is not on the local network? */
+  if (!ip_addr_netcmp(&(hdr->ip.src), &(netif->ip_addr), &(netif->netmask))) {
     /* do nothing */
     return;
   }
 
   LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("etharp_ip_input: updating ETHARP table.\n"));
-  /* update ARP table, ask to insert entry */
-  update_arp_entry(netif, &(hdr->ip.src), &(hdr->eth.src), ETHARP_CREATE);
+  /* update ARP table */
+  /* @todo We could use ETHARP_TRY_HARD if we think we are going to talk
+   * back soon (for example, if the destination IP address is ours. */
+  update_arp_entry(netif, &(hdr->ip.src), &(hdr->eth.src), 0);
 }
 
 
@@ -492,7 +498,7 @@ etharp_arp_input(struct netif *netif, struct eth_addr *ethaddr, struct pbuf *p)
   if (for_us) {
     /* add IP address in ARP cache; assume requester wants to talk to us.
      * can result in directly sending the queued packets for this host. */
-    update_arp_entry(netif, &sipaddr, &(hdr->shwaddr), ETHARP_CREATE);
+    update_arp_entry(netif, &sipaddr, &(hdr->shwaddr), ETHARP_TRY_HARD);
   /* ARP message not directed to us? */
   } else {
     /* update the source IP address in the cache, if present */
@@ -548,8 +554,9 @@ etharp_arp_input(struct netif *netif, struct eth_addr *ethaddr, struct pbuf *p)
     /* ARP reply. We already updated the ARP cache earlier. */
     LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("etharp_arp_input: incoming ARP reply\n"));
 #if (LWIP_DHCP && DHCP_DOES_ARP_CHECK)
-    /* DHCP wants to know about ARP replies to our wanna-have-address */
-    if (for_us) dhcp_arp_reply(netif, &sipaddr);
+    /* When unconfigured, DHCP wants to know about ARP replies from the
+     * address offered to us, as that means someone else uses it already! */
+    if (netif->ip_addr.addr == 0) dhcp_arp_reply(netif, &sipaddr);
 #endif
     break;
   default:
@@ -563,36 +570,27 @@ etharp_arp_input(struct netif *netif, struct eth_addr *ethaddr, struct pbuf *p)
 /**
  * Resolve and fill-in Ethernet address header for outgoing packet.
  *
- * If ARP has the Ethernet address in cache, the given packet is
- * returned, ready to be sent.
+ * For IP multicast and broadcast, corresponding Ethernet addresses
+ * are selected and the packet is transmitted on the link.
  *
- * If ARP does not have the Ethernet address in cache the packet is
- * queued (if enabled and space available) and a ARP request is sent.
- * This ARP request is returned as a pbuf, which should be sent by
- * the caller.
- *
- * A returned non-NULL packet should be sent by the caller.
- *
- * If ARP failed to allocate resources, NULL is returned.
+ * For unicast addresses, the packet is submitted to etharp_query(). In
+ * case the IP address is outside the local network, the IP address of
+ * the gateway is used.
  *
  * @param netif The lwIP network interface which the IP packet will be sent on.
  * @param ipaddr The IP address of the packet destination.
  * @param pbuf The pbuf(s) containing the IP packet to be sent.
  *
- * @return If non-NULL, a packet ready to be sent by caller.
- *
  * @return
- * - ERR_BUF Could not make room for Ethernet header.
- * - ERR_MEM Hardware address unknown, and no more ARP entries available
- *   to query for address or queue the packet.
- * - ERR_RTE No route to destination (no gateway to external networks).
+ * - ERR_RTE No route to destination (no gateway to external networks),
+ * or the return type of either etharp_query() or netif->linkoutput().
  */
 err_t
 etharp_output(struct netif *netif, struct ip_addr *ipaddr, struct pbuf *q)
 {
   struct eth_addr *dest, *srcaddr, mcastaddr;
   struct eth_hdr *ethhdr;
-  err_t result = ERR_OK;
+  u8_t i;
 
   /* make room for Ethernet header - should not fail */
   if (pbuf_header(q, sizeof(struct eth_hdr)) != 0) {
@@ -607,13 +605,13 @@ etharp_output(struct netif *netif, struct ip_addr *ipaddr, struct pbuf *q)
   /* Determine on destination hardware address. Broadcasts and multicasts
    * are special, other IP addresses are looked up in the ARP table. */
 
-  /* destination IP address is an IP broadcast address? */
-  if (ip_addr_isany(ipaddr) || ip_addr_isbroadcast(ipaddr, netif)) {
+  /* broadcast destination IP address? */
+  if (ip_addr_isbroadcast(ipaddr, netif)) {
     /* broadcast on Ethernet also */
     dest = (struct eth_addr *)&ethbroadcast;
-  /* destination IP address is an IP multicast address? */
+  /* multicast destination IP address? */
   } else if (ip_addr_ismulticast(ipaddr)) {
-    /* Hash IP multicast address to MAC address. */
+    /* Hash IP multicast address to MAC address.*/
     mcastaddr.addr[0] = 0x01;
     mcastaddr.addr[1] = 0x00;
     mcastaddr.addr[2] = 0x5e;
@@ -622,17 +620,17 @@ etharp_output(struct netif *netif, struct ip_addr *ipaddr, struct pbuf *q)
     mcastaddr.addr[5] = ip4_addr4(ipaddr);
     /* destination Ethernet address is multicast */
     dest = &mcastaddr;
-  /* destination IP address is an IP unicast address */
+  /* unicast destination IP address? */
   } else {
     /* outside local network? */
-    if (!ip_addr_maskcmp(ipaddr, &(netif->ip_addr), &(netif->netmask))) {
+    if (!ip_addr_netcmp(ipaddr, &(netif->ip_addr), &(netif->netmask))) {
       /* interface has default gateway? */
       if (netif->gw.addr != 0) {
         /* send to hardware address of default gateway IP address */
         ipaddr = &(netif->gw);
       /* no default gateway available */
       } else {
-        /* no route to destination error */
+        /* no route to destination error (default gateway missing) */
         return ERR_RTE;
       }
     }
@@ -640,23 +638,17 @@ etharp_output(struct netif *netif, struct ip_addr *ipaddr, struct pbuf *q)
     return etharp_query(netif, ipaddr, q);
   }
 
-  /* destination Ethernet address known */
-  if (dest != NULL) {
-    u8_t i;
-    /* obtain source Ethernet address of the given interface */
-    srcaddr = (struct eth_addr *)netif->hwaddr;
-    /* A valid IP->MAC address mapping was found, fill in the
-     * Ethernet header for the outgoing packet */
-    ethhdr = q->payload;
-    for (i = 0; i < netif->hwaddr_len; i++) {
-      ethhdr->dest.addr[i] = dest->addr[i];
-      ethhdr->src.addr[i] = srcaddr->addr[i];
-    }
-    ethhdr->type = htons(ETHTYPE_IP);
-    /* send packet */
-    result = netif->linkoutput(netif, q);
+  /* continuation for multicast/broadcast destinations */
+  /* obtain source Ethernet address of the given interface */
+  srcaddr = (struct eth_addr *)netif->hwaddr;
+  ethhdr = q->payload;
+  for (i = 0; i < netif->hwaddr_len; i++) {
+    ethhdr->dest.addr[i] = dest->addr[i];
+    ethhdr->src.addr[i] = srcaddr->addr[i];
   }
-  return result;
+  ethhdr->type = htons(ETHTYPE_IP);
+  /* send packet directly on the link */
+  return netif->linkoutput(netif, q);
 }
 
 /**
@@ -699,18 +691,23 @@ err_t etharp_query(struct netif *netif, struct ip_addr *ipaddr, struct pbuf *q)
   u8_t k; /* Ethernet address octet index */
 
   /* non-unicast address? */
-  if (ip_addr_isany(ipaddr) ||
-      ip_addr_isbroadcast(ipaddr, netif) ||
-      ip_addr_ismulticast(ipaddr)) {
+  if (ip_addr_isbroadcast(ipaddr, netif) ||
+      ip_addr_ismulticast(ipaddr) ||
+      ip_addr_isany(ipaddr)) {
     LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("etharp_query: will not add non-unicast IP address to ARP cache\n"));
     return ERR_ARG;
   }
 
   /* find entry in ARP cache, ask to create entry if queueing packet */
-  i = find_entry(ipaddr, (q != NULL) ? ETHARP_CREATE : 0);
+  i = find_entry(ipaddr, ETHARP_TRY_HARD);
 
   /* could not find or create entry? */
-  if (i < 0) return (err_t)i;
+  if (i < 0)
+  {
+    LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("etharp_query: could not create ARP entry\n"));
+    if (q) LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("etharp_query: packet dropped\n"));
+    return (err_t)i;
+  }
 
   /* mark a fresh entry as pending (we just sent a request) */
   if (arp_table[i].state == ETHARP_STATE_EMPTY) {
@@ -722,7 +719,7 @@ err_t etharp_query(struct netif *netif, struct ip_addr *ipaddr, struct pbuf *q)
   ((arp_table[i].state == ETHARP_STATE_PENDING) ||
    (arp_table[i].state == ETHARP_STATE_STABLE)));
 
-  /* do we have a pending entry? */
+  /* do we have a pending entry? or an implicit query request? */
   if ((arp_table[i].state == ETHARP_STATE_PENDING) || (q == NULL)) {
     /* try to resolve it; send out ARP request */
     result = etharp_request(netif, ipaddr);
@@ -752,8 +749,17 @@ err_t etharp_query(struct netif *netif, struct ip_addr *ipaddr, struct pbuf *q)
       p = pbuf_take(q);
       /* packet could be taken over? */
       if (p != NULL) {
-        /* queue packet */
-        pbuf_queue(arp_table[i].p, p);
+        /* queue packet ... */
+        if (arp_table[i].p == NULL) {
+        	/* ... in the empty queue */
+        	pbuf_ref(p);
+        	arp_table[i].p = p;
+#if 0 /* multi-packet-queueing disabled, see bug #11400 */
+        } else {
+        	/* ... at tail of non-empty queue */
+          pbuf_queue(arp_table[i].p, p);
+#endif
+        }
         LWIP_DEBUGF(ETHARP_DEBUG | DBG_TRACE, ("etharp_query: queued packet %p on ARP entry %d\n", (void *)q, i));
         result = ERR_OK;
       } else {

src/netif/ethernetif.c

@@ -42,6 +42,7 @@
 #include "lwip/mem.h"
 #include "lwip/pbuf.h"
 #include "lwip/sys.h"
+#include <lwip/stats.h>
 
 #include "netif/etharp.h"
 
@@ -117,7 +118,7 @@ low_level_output(struct netif *netif, struct pbuf *p)
   pbuf_header(p, ETH_PAD_SIZE);			/* reclaim the padding word */
 #endif
   
-#ifdef LINK_STATS
+#if LINK_STATS
   lwip_stats.link.xmit++;
 #endif /* LINK_STATS */      
 
@@ -170,12 +171,12 @@ low_level_input(struct netif *netif)
     pbuf_header(p, ETH_PAD_SIZE);			/* reclaim the padding word */
 #endif
 
-#ifdef LINK_STATS
+#if LINK_STATS
     lwip_stats.link.recv++;
 #endif /* LINK_STATS */      
   } else {
     drop packet();
-#ifdef LINK_STATS
+#if LINK_STATS
     lwip_stats.link.memerr++;
     lwip_stats.link.drop++;
 #endif /* LINK_STATS */      
@@ -229,7 +230,7 @@ ethernetif_input(struct netif *netif)
   /* points to packet payload, which starts with an Ethernet header */
   ethhdr = p->payload;
 
-#ifdef LINK_STATS
+#if LINK_STATS
   lwip_stats.link.recv++;
 #endif /* LINK_STATS */
 
@@ -273,7 +274,7 @@ arp_timer(void *arg)
  *
  */
 
-void
+err_t
 ethernetif_init(struct netif *netif)
 {
   struct ethernetif *ethernetif;
@@ -299,5 +300,7 @@ ethernetif_init(struct netif *netif)
   etharp_init();
 
   sys_timeout(ARP_TMR_INTERVAL, arp_timer, NULL);
+
+  return ERR_OK;
 }
 

src/netif/ppp/ppp.c

@@ -1236,7 +1236,7 @@ static void pppMain(void *arg)
 			pppInProc(pd, p->payload, c);
 		} else {
 		    PPPDEBUG((LOG_DEBUG, "pppMainWakeup: unit %d sio_read len=%d returned %d\n", pd, p->len, c));
-		    sys_msleep(250); /* give other tasks a chance to run */
+		    sys_msleep(1); /* give other tasks a chance to run */
 		}
         }
     }