Init commit

lwip/test/fuzz/README

@@ -0,0 +1,37 @@
+
+Fuzzing the lwIP stack (afl-fuzz requires linux/unix or similar)
+
+This directory contains a small app that reads Ethernet frames from stdin and
+processes them. It is used together with the 'american fuzzy lop' tool (found
+at http://lcamtuf.coredump.cx/afl/) and the sample inputs to test how
+unexpected inputs are handled. The afl tool will read the known inputs, and
+try to modify them to exercise as many code paths as possible, by instrumenting
+the code and keeping track of which code is executed.
+
+Just running make will produce the test program.
+
+Running make with parameter 'D=-DLWIP_FUZZ_MULTI_PACKET' will produce a binary
+that parses the input data as multiple packets (experimental!).
+
+Then run afl with:
+
+afl-fuzz -i inputs/<INPUT> -o output ./lwip_fuzz
+
+and it should start working. It will probably complain about CPU scheduler,
+set AFL_SKIP_CPUFREQ=1 to ignore it.
+If it complains about invalid "/proc/sys/kernel/core_pattern" setting, try
+executing "sudo bash -c 'echo core > /proc/sys/kernel/core_pattern'".
+
+The input is split into different subdirectories since they test different
+parts of the code, and since you want to run one instance of afl-fuzz on each
+core.
+
+When afl finds a crash or a hang, the input that caused it will be placed in
+the output directory. If you have hexdump and text2pcap tools installed,
+running output_to_pcap.sh <outputdir> will create pcap files for each input
+file to simplify viewing in wireshark.
+
+The lwipopts.h file needs to have checksum checking off, otherwise almost every
+packet will be discarded because of that. The other options can be tuned to
+expose different parts of the code.
+

lwip/test/fuzz/fuzz.c

@@ -0,0 +1,189 @@
+/*
+ * Copyright (c) 2001-2003 Swedish Institute of Computer Science.
+ * All rights reserved. 
+ * 
+ * Redistribution and use in source and binary forms, with or without modification, 
+ * are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission. 
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED 
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT 
+ * SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 
+ * OF SUCH DAMAGE.
+ *
+ * This file is part of the lwIP TCP/IP stack.
+ * 
+ * Author: Erik Ekman <erik@kryo.se>
+ *
+ */
+
+#include "lwip/init.h"
+#include "lwip/netif.h"
+#include "lwip/dns.h"
+#include "netif/etharp.h"
+#if LWIP_IPV6
+#include "lwip/ethip6.h"
+#include "lwip/nd6.h"
+#endif
+
+#include "lwip/apps/httpd.h"
+#include "lwip/apps/snmp.h"
+#include "lwip/apps/lwiperf.h"
+#include "lwip/apps/mdns.h"
+
+#include <string.h>
+#include <stdio.h>
+
+/* This define enables multi packet processing.
+ * For this, the input is interpreted as 2 byte length + data + 2 byte length + data...
+ * #define LWIP_FUZZ_MULTI_PACKET
+*/
+#ifdef LWIP_FUZZ_MULTI_PACKET
+u8_t pktbuf[20000];
+#else
+u8_t pktbuf[2000];
+#endif
+
+/* no-op send function */
+static err_t lwip_tx_func(struct netif *netif, struct pbuf *p)
+{
+  LWIP_UNUSED_ARG(netif);
+  LWIP_UNUSED_ARG(p);
+  return ERR_OK;
+}
+
+static err_t testif_init(struct netif *netif)
+{
+  netif->name[0] = 'f';
+  netif->name[1] = 'z';
+  netif->output = etharp_output;
+  netif->linkoutput = lwip_tx_func;
+  netif->mtu = 1500;
+  netif->hwaddr_len = 6;
+  netif->flags = NETIF_FLAG_BROADCAST | NETIF_FLAG_ETHARP | NETIF_FLAG_IGMP;
+
+  netif->hwaddr[0] = 0x00;
+  netif->hwaddr[1] = 0x23;
+  netif->hwaddr[2] = 0xC1;
+  netif->hwaddr[3] = 0xDE;
+  netif->hwaddr[4] = 0xD0;
+  netif->hwaddr[5] = 0x0D;
+
+#if LWIP_IPV6
+  netif->output_ip6 = ethip6_output;
+  netif->ip6_autoconfig_enabled = 1;
+  netif_create_ip6_linklocal_address(netif, 1);
+  netif->flags |= NETIF_FLAG_MLD6;
+#endif
+
+  return ERR_OK;
+}
+
+static void input_pkt(struct netif *netif, const u8_t *data, size_t len)
+{
+  struct pbuf *p, *q;
+  err_t err;
+
+  LWIP_ASSERT("pkt too big", len <= 0xFFFF);
+  p = pbuf_alloc(PBUF_RAW, (u16_t)len, PBUF_POOL);
+  LWIP_ASSERT("alloc failed", p);
+  for(q = p; q != NULL; q = q->next) {
+    MEMCPY(q->payload, data, q->len);
+    data += q->len;
+  }
+  err = netif->input(p, netif);
+  if (err != ERR_OK) {
+    pbuf_free(p);
+  }
+}
+
+static void input_pkts(struct netif *netif, const u8_t *data, size_t len)
+{
+#ifdef LWIP_FUZZ_MULTI_PACKET
+  const u16_t max_packet_size = 1514;
+  const u8_t *ptr = data;
+  size_t rem_len = len;
+
+  while (rem_len > sizeof(u16_t)) {
+    u16_t frame_len;
+    memcpy(&frame_len, ptr, sizeof(u16_t));
+    ptr += sizeof(u16_t);
+    rem_len -= sizeof(u16_t);
+    frame_len = htons(frame_len) & 0x7FF;
+    frame_len = LWIP_MIN(frame_len, max_packet_size);
+    if (frame_len > rem_len) {
+      frame_len = (u16_t)rem_len;
+    }
+    if (frame_len != 0) {
+      input_pkt(netif, ptr, frame_len);
+    }
+    ptr += frame_len;
+    rem_len -= frame_len;
+  }
+#else /* LWIP_FUZZ_MULTI_PACKET */
+  input_pkt(netif, data, len);
+#endif /* LWIP_FUZZ_MULTI_PACKET */
+}
+
+int main(int argc, char** argv)
+{
+  struct netif net_test;
+  ip4_addr_t addr;
+  ip4_addr_t netmask;
+  ip4_addr_t gw;
+  size_t len;
+
+  lwip_init();
+
+  IP4_ADDR(&addr, 172, 30, 115, 84);
+  IP4_ADDR(&netmask, 255, 255, 255, 0);
+  IP4_ADDR(&gw, 172, 30, 115, 1);
+
+  netif_add(&net_test, &addr, &netmask, &gw, &net_test, testif_init, ethernet_input);
+  netif_set_up(&net_test);
+  netif_set_link_up(&net_test);
+
+#if LWIP_IPV6
+  nd6_tmr(); /* tick nd to join multicast groups */
+#endif
+  dns_setserver(0, &net_test.gw);
+
+  /* initialize apps */
+  httpd_init();
+  lwiperf_start_tcp_server_default(NULL, NULL);
+  mdns_resp_init();
+  mdns_resp_add_netif(&net_test, "hostname", 255);
+  snmp_init();
+
+  if(argc > 1) {
+    FILE* f;
+    const char* filename;
+    printf("reading input from file... ");
+    fflush(stdout);
+    filename = argv[1];
+    LWIP_ASSERT("invalid filename", filename != NULL);
+    f = fopen(filename, "rb");
+    LWIP_ASSERT("open failed", f != NULL);
+    len = fread(pktbuf, 1, sizeof(pktbuf), f);
+    fclose(f);
+    printf("testing file: \"%s\"...\r\n", filename);
+  } else {
+    len = fread(pktbuf, 1, sizeof(pktbuf), stdin);
+  }
+  input_pkts(&net_test, pktbuf, len);
+
+  return 0;
+}

lwip/test/fuzz/lwipopts.h

@@ -0,0 +1,80 @@
+/*
+ * Copyright (c) 2001-2003 Swedish Institute of Computer Science.
+ * All rights reserved. 
+ * 
+ * Redistribution and use in source and binary forms, with or without modification, 
+ * are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission. 
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED 
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT 
+ * SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 
+ * OF SUCH DAMAGE.
+ *
+ * This file is part of the lwIP TCP/IP stack.
+ * 
+ * Author: Simon Goldschmidt
+ *
+ */
+#ifndef LWIP_HDR_LWIPOPTS_H__
+#define LWIP_HDR_LWIPOPTS_H__
+
+/* Prevent having to link sys_arch.c (we don't test the API layers in unit tests) */
+#define NO_SYS                          1
+#define LWIP_NETCONN                    0
+#define LWIP_SOCKET                     0
+#define SYS_LIGHTWEIGHT_PROT            0
+
+#define LWIP_IPV6                       1
+#define IPV6_FRAG_COPYHEADER            1
+#define LWIP_IPV6_DUP_DETECT_ATTEMPTS   0
+
+/* Enable some protocols to test them */
+#define LWIP_DHCP                       1
+#define LWIP_AUTOIP                     1
+
+#define LWIP_IGMP                       1
+#define LWIP_DNS                        1
+
+#define LWIP_ALTCP                      1
+
+/* Turn off checksum verification of fuzzed data */
+#define CHECKSUM_CHECK_IP               0
+#define CHECKSUM_CHECK_UDP              0
+#define CHECKSUM_CHECK_TCP              0
+#define CHECKSUM_CHECK_ICMP             0
+#define CHECKSUM_CHECK_ICMP6            0
+
+/* Minimal changes to opt.h required for tcp unit tests: */
+#define MEM_SIZE                        16000
+#define TCP_SND_QUEUELEN                40
+#define MEMP_NUM_TCP_SEG                TCP_SND_QUEUELEN
+#define TCP_OVERSIZE                    1
+#define TCP_SND_BUF                     (12 * TCP_MSS)
+#define TCP_WND                         (10 * TCP_MSS)
+#define LWIP_WND_SCALE                  1
+#define TCP_RCV_SCALE                   2
+#define PBUF_POOL_SIZE                  400 /* pbuf tests need ~200KByte */
+
+/* Minimal changes to opt.h required for etharp unit tests: */
+#define ETHARP_SUPPORT_STATIC_ENTRIES   1
+
+#define LWIP_NUM_NETIF_CLIENT_DATA      1
+#define LWIP_SNMP                       1
+#define MIB2_STATS                      1
+#define LWIP_MDNS_RESPONDER             1
+
+#endif /* LWIP_HDR_LWIPOPTS_H__ */

lwip/test/fuzz/output_to_pcap.sh

@@ -0,0 +1,31 @@
+#!/bin/bash
+
+if [ -z "$1" ]
+then
+	echo "This script will make pcap files from the afl-fuzz crash/hang files"
+	echo "It needs hexdump and text2pcap"
+	echo "Please give output directory as argument"
+	exit 2
+fi
+
+for i in `ls $1/crashes/id*`
+do
+	PCAPNAME=`echo $i | grep pcap`
+	if [ -z "$PCAPNAME" ]; then
+		hexdump -C $i > $1/$$.tmp
+		text2pcap $1/$$.tmp ${i}.pcap
+	fi
+done
+for i in `ls $1/hangs/id*`
+do
+	PCAPNAME=`echo $i | grep pcap`
+	if [ -z "$PCAPNAME" ]; then
+		hexdump -C $i > $1/$$.tmp
+		text2pcap $1/$$.tmp ${i}.pcap
+	fi
+done
+rm -f $1/$$.tmp
+
+echo
+echo "Created pcap files:"
+ls $1/*/*.pcap

lwip/test/sockets/sockets_stresstest.c

@@ -0,0 +1,726 @@
+/**
+ * @file
+ * Sockets stresstest
+ *
+ * This file uses the lwIP socket API to do stress tests that should test the
+ * stability when used in many different situations, with many concurrent
+ * sockets making concurrent transfers in different manners.
+ *
+ * - test rely on loopback sockets for now, so netif drivers are not tested
+ * - all enabled functions shall be used
+ * - parallelism of the tests depend on enough resources being available
+ *   (configure your lwipopts.h settings high enough)
+ * - test should also be able to run in a real target
+ *
+ * TODO:
+ * - full duplex
+ * - add asserts about internal socket/netconn/pcb state?
+ */
+ 
+ /*
+ * Copyright (c) 2017 Simon Goldschmidt
+ * All rights reserved.
+ * 
+ * Redistribution and use in source and binary forms, with or without modification, 
+ * are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission. 
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED 
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT 
+ * SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 
+ * OF SUCH DAMAGE.
+ *
+ * This file is part of the lwIP TCP/IP stack.
+ * 
+ * Author: Simon Goldschmidt <goldsimon@gmx.de>
+ *
+ */
+
+#include "lwip/opt.h"
+#include "sockets_stresstest.h"
+
+#include "lwip/sockets.h"
+#include "lwip/sys.h"
+
+#include "lwip/mem.h"
+
+#include <stdio.h>
+#include <string.h>
+
+#if LWIP_SOCKET && LWIP_IPV4 /* this uses IPv4 loopback sockets, currently */
+
+#ifndef TEST_SOCKETS_STRESS
+#define TEST_SOCKETS_STRESS   LWIP_DBG_OFF
+#endif
+
+#define TEST_TIME_SECONDS     10
+#define TEST_TXRX_BUFSIZE     (TCP_MSS * 2)
+#define TEST_MAX_RXWAIT_MS    50
+#define TEST_MAX_CONNECTIONS  50
+
+#define TEST_SOCK_READABLE    0x01
+#define TEST_SOCK_WRITABLE    0x02
+#define TEST_SOCK_ERR         0x04
+
+#define TEST_MODE_SELECT      0x01
+#define TEST_MODE_POLL        0x02
+#define TEST_MODE_NONBLOCKING 0x04
+#define TEST_MODE_WAIT        0x08
+#define TEST_MODE_RECVTIMEO   0x10
+#define TEST_MODE_SLEEP       0x20
+
+static int sockets_stresstest_numthreads;
+
+struct test_settings {
+  struct sockaddr_storage addr;
+  int start_client;
+  int loop_cnt;
+};
+
+struct sockets_stresstest_fullduplex {
+  int s;
+  volatile int closed;
+};
+
+static void
+fill_test_data(void *buf, size_t buf_len_bytes)
+{
+  u8_t *p = (u8_t*)buf;
+  u16_t i, chk;
+
+  LWIP_ASSERT("buffer too short", buf_len_bytes >= 4);
+  LWIP_ASSERT("buffer too big", buf_len_bytes <= 0xFFFF);
+  /* store the total number of bytes */
+  p[0] = (u8_t)(buf_len_bytes >> 8);
+  p[1] = (u8_t)buf_len_bytes;
+
+  /* fill buffer with random */
+  chk = 0;
+  for (i = 4; i < buf_len_bytes; i++) {
+    u8_t rnd = (u8_t)LWIP_RAND();
+    p[i] = rnd;
+    chk += rnd;
+  }
+  /* store checksum */
+  p[2] = (u8_t)(chk >> 8);
+  p[3] = (u8_t)chk;
+}
+
+static size_t
+check_test_data(const void *buf, size_t buf_len_bytes)
+{
+  u8_t *p = (u8_t*)buf;
+  u16_t i, chk, chk_rx, len_rx;
+
+  LWIP_ASSERT("buffer too short", buf_len_bytes >= 4);
+  len_rx = (((u16_t)p[0]) << 8) | p[1];
+  LWIP_ASSERT("len too short", len_rx >= 4);
+  if (len_rx > buf_len_bytes) {
+    /* not all data received in this segment */
+    LWIP_DEBUGF(TEST_SOCKETS_STRESS | LWIP_DBG_TRACE, ("check-\n"));
+    return buf_len_bytes;
+  }
+  chk_rx = (((u16_t)p[2]) << 8) | p[3];
+  /* calculate received checksum */
+  chk = 0;
+  for (i = 4; i < len_rx; i++) {
+    chk += p[i];
+  }
+  LWIP_ASSERT("invalid checksum", chk == chk_rx);
+  if (len_rx < buf_len_bytes) {
+    size_t data_left = buf_len_bytes - len_rx;
+    memmove(p, &p[len_rx], data_left);
+    return data_left;
+  }
+  /* if we come here, we received exactly one chunk
+     -> next offset is 0 */
+  return 0;
+}
+
+static size_t
+recv_and_check_data_return_offset(int s, char *rxbuf, size_t rxbufsize, size_t rxoff, int *closed, const char *dbg)
+{
+  ssize_t ret;
+
+  ret = lwip_read(s, &rxbuf[rxoff], rxbufsize - rxoff);
+  if (ret == 0) {
+    *closed = 1;
+    return rxoff;
+  }
+  *closed = 0;
+  LWIP_DEBUGF(TEST_SOCKETS_STRESS | LWIP_DBG_TRACE, ("%s %d rx %d\n", dbg, s, (int)ret));
+  if (ret == -1) {
+    /* TODO: for this to work, 'errno' has to support multithreading... */
+    int err = errno;
+    if (err == ENOTCONN) {
+      *closed = 1;
+      return 0;
+    }
+    LWIP_ASSERT("err == 0", err == 0);
+  }
+  LWIP_ASSERT("ret > 0", ret > 0);
+  return check_test_data(rxbuf, rxoff + ret);
+}
+
+#if LWIP_SOCKET_SELECT
+static int
+sockets_stresstest_wait_readable_select(int s, int timeout_ms)
+{
+  int ret;
+  struct timeval tv;
+  fd_set fs_r;
+  fd_set fs_w;
+  fd_set fs_e;
+
+  FD_ZERO(&fs_r);
+  FD_ZERO(&fs_w);
+  FD_ZERO(&fs_e);
+
+  FD_SET(s, &fs_r);
+  FD_SET(s, &fs_e);
+
+  tv.tv_sec = timeout_ms / 1000;
+  tv.tv_usec = (timeout_ms - (tv.tv_sec * 1000)) * 1000;
+  ret = lwip_select(s + 1, &fs_r, &fs_w, &fs_e, &tv);
+  LWIP_ASSERT("select error", ret >= 0);
+  if (ret) {
+    /* convert poll flags to our flags */
+    ret = 0;
+    if (FD_ISSET(s, &fs_r)) {
+      ret |= TEST_SOCK_READABLE;
+    }
+    if (FD_ISSET(s, &fs_w)) {
+      ret |= TEST_SOCK_WRITABLE;
+    }
+    if (FD_ISSET(s, &fs_e)) {
+      ret |= TEST_SOCK_ERR;
+    }
+    return ret;
+  }
+  return 0;
+}
+#endif
+
+#if LWIP_SOCKET_POLL
+static int
+sockets_stresstest_wait_readable_poll(int s, int timeout_ms)
+{
+  int ret;
+  struct pollfd pfd;
+
+  pfd.fd = s;
+  pfd.revents = 0;
+  pfd.events = POLLIN | POLLERR;
+
+  ret = lwip_poll(&pfd, 1, timeout_ms);
+  if (ret) {
+    /* convert poll flags to our flags */
+    ret = 0;
+    if (pfd.revents & POLLIN) {
+      ret |= TEST_SOCK_READABLE;
+    }
+    if (pfd.revents & POLLOUT) {
+      ret |= TEST_SOCK_WRITABLE;
+    }
+    if (pfd.revents & POLLERR) {
+      ret |= TEST_SOCK_ERR;
+    }
+    return ret;
+  }
+  return 0;
+}
+#endif
+
+#if LWIP_SO_RCVTIMEO
+static int
+sockets_stresstest_wait_readable_recvtimeo(int s, int timeout_ms)
+{
+  int ret;
+  char buf;
+#if LWIP_SO_SNDRCVTIMEO_NONSTANDARD
+  int opt_on = timeout_ms;
+  int opt_off = 0;
+#else
+  struct timeval opt_on, opt_off;
+  opt_on.tv_sec = timeout_ms / 1000;
+  opt_on.tv_usec = (timeout_ms - (opt_on.tv_sec * 1000)) * 1000;
+  opt_off.tv_sec = 0;
+  opt_off.tv_usec = 0;
+#endif
+
+  /* enable receive timeout */
+  ret = lwip_setsockopt(s, SOL_SOCKET, SO_RCVTIMEO, &opt_on, sizeof(opt_on));
+  LWIP_ASSERT("setsockopt error", ret == 0);
+
+  /* peek for one byte with timeout */
+  ret = lwip_recv(s, &buf, 1, MSG_PEEK);
+
+  /* disable receive timeout */
+  ret = lwip_setsockopt(s, SOL_SOCKET, SO_RCVTIMEO, &opt_off, sizeof(opt_off));
+  LWIP_ASSERT("setsockopt error", ret == 0);
+
+  if (ret == 1) {
+    return TEST_SOCK_READABLE;
+  }
+  if (ret == 0) {
+    return 0;
+  }
+  if (ret == -1) {
+    return TEST_SOCK_ERR;
+  }
+  LWIP_ASSERT("invalid return value", 0);
+  return TEST_SOCK_ERR;
+}
+#endif
+
+static int
+sockets_stresstest_wait_readable_wait_peek(int s, int timeout_ms)
+{
+  int ret;
+  char buf;
+
+  LWIP_UNUSED_ARG(timeout_ms); /* cannot time out here */
+
+  /* peek for one byte */
+  ret = lwip_recv(s, &buf, 1, MSG_PEEK);
+
+  if (ret == 1) {
+    return TEST_SOCK_READABLE;
+  }
+  if (ret == 0) {
+    return 0;
+  }
+  if (ret == -1) {
+    return TEST_SOCK_ERR;
+  }
+  LWIP_ASSERT("invalid return value", 0);
+  return TEST_SOCK_ERR;
+}
+
+static int
+sockets_stresstest_wait_readable_nonblock(int s, int timeout_ms)
+{
+  int ret;
+  char buf;
+  u32_t wait_until = sys_now() + timeout_ms;
+
+  while(sys_now() < wait_until) {
+    /* peek for one byte */
+    ret = lwip_recv(s, &buf, 1, MSG_PEEK | MSG_DONTWAIT);
+
+    if (ret == 1) {
+      return TEST_SOCK_READABLE;
+    }
+    if (ret == -1) {
+      /* TODO: for this to work, 'errno' has to support multithreading... */
+      int err = errno;
+      if (err != EWOULDBLOCK) {
+        return TEST_SOCK_ERR;
+      }
+    }
+    /* TODO: sleep? */
+  }
+  return 0;
+}
+
+static int sockets_stresstest_rand_mode(int allow_wait, int allow_rx)
+{
+  u32_t random_value = LWIP_RAND();
+#if LWIP_SOCKET_SELECT
+  if (random_value & TEST_MODE_SELECT) {
+    return TEST_MODE_SELECT;
+  }
+#endif
+#if LWIP_SOCKET_POLL
+  if (random_value & TEST_MODE_POLL) {
+    return TEST_MODE_POLL;
+  }
+#endif
+  if (!allow_rx) {
+    return TEST_MODE_SLEEP;
+  }
+#if LWIP_SO_RCVTIMEO
+  if (random_value & TEST_MODE_RECVTIMEO) {
+    return TEST_MODE_RECVTIMEO;
+  }
+#endif
+  if (allow_wait) {
+    if (random_value & TEST_MODE_RECVTIMEO) {
+      return TEST_MODE_RECVTIMEO;
+    }
+  }
+  return TEST_MODE_NONBLOCKING;
+}
+
+static int
+sockets_stresstest_wait_readable(int mode, int s, int timeout_ms)
+{
+  switch(mode)
+  {
+#if LWIP_SOCKET_SELECT
+  case TEST_MODE_SELECT:
+    return sockets_stresstest_wait_readable_select(s, timeout_ms);
+#endif
+#if LWIP_SOCKET_POLL
+  case TEST_MODE_POLL:
+    return sockets_stresstest_wait_readable_poll(s, timeout_ms);
+#endif
+#if LWIP_SO_RCVTIMEO
+  case TEST_MODE_RECVTIMEO:
+    return sockets_stresstest_wait_readable_recvtimeo(s, timeout_ms);
+#endif
+  case TEST_MODE_WAIT:
+    return sockets_stresstest_wait_readable_wait_peek(s, timeout_ms);
+  case TEST_MODE_NONBLOCKING:
+    return sockets_stresstest_wait_readable_nonblock(s, timeout_ms);
+  case TEST_MODE_SLEEP:
+    {
+      sys_msleep(timeout_ms);
+      return 1;
+    }
+  default:
+    LWIP_ASSERT("invalid mode", 0);
+    break;
+  }
+  return 0;
+}
+
+#if LWIP_NETCONN_FULLDUPLEX
+static void
+sockets_stresstest_conn_client_r(void *arg)
+{
+  struct sockets_stresstest_fullduplex *fd = (struct sockets_stresstest_fullduplex *)arg;
+  int s = fd->s;
+  size_t rxoff = 0;
+  char rxbuf[TEST_TXRX_BUFSIZE];
+
+  while (1) {
+    int closed;
+    if (fd->closed) {
+      break;
+    }
+    rxoff = recv_and_check_data_return_offset(s, rxbuf, sizeof(rxbuf), rxoff, &closed, "cli");
+    if (fd->closed) {
+      break;
+    }
+    if (closed) {
+      lwip_close(s);
+      break;
+    }
+  }
+
+  SYS_ARCH_DEC(sockets_stresstest_numthreads, 1);
+  LWIP_ASSERT("", sockets_stresstest_numthreads >= 0);
+}
+#endif
+
+static void
+sockets_stresstest_conn_client(void *arg)
+{
+  struct sockaddr_storage addr;
+  struct sockaddr_in *addr_in;
+  int s, ret;
+  char txbuf[TEST_TXRX_BUFSIZE];
+  char rxbuf[TEST_TXRX_BUFSIZE];
+  size_t rxoff = 0;
+  u32_t max_time = sys_now() + (TEST_TIME_SECONDS * 1000);
+  int do_rx = 1;
+  struct sockets_stresstest_fullduplex *data = NULL;
+
+  memcpy(&addr, arg, sizeof(addr));
+  LWIP_ASSERT("", addr.ss_family == AF_INET);
+  addr_in = (struct sockaddr_in *)&addr;
+  addr_in->sin_addr.s_addr = inet_addr("127.0.0.1");
+
+  /* sleep a random time between 1 and 2 seconds */
+  sys_msleep(1000 + (LWIP_RAND() % 1000));
+
+  /* connect to the server */
+  s = lwip_socket(addr.ss_family, SOCK_STREAM, 0);
+  LWIP_ASSERT("s >= 0", s >= 0);
+
+#if LWIP_NETCONN_FULLDUPLEX
+  if (LWIP_RAND() & 1) {
+    sys_thread_t t;
+    data = (struct sockets_stresstest_fullduplex*)mem_malloc(sizeof(struct sockets_stresstest_fullduplex));
+    LWIP_ASSERT("data != NULL", data != 0);
+    SYS_ARCH_INC(sockets_stresstest_numthreads, 1);
+    data->s = s;
+    data->closed = 0;
+    t = sys_thread_new("sockets_stresstest_conn_client_r", sockets_stresstest_conn_client_r, data, 0, 0);
+    LWIP_ASSERT("thread != NULL", t != 0);
+    do_rx = 0;
+  }
+#endif
+
+  /* @todo: nonblocking connect? */
+  ret = lwip_connect(s, (struct sockaddr *)&addr, sizeof(struct sockaddr_storage));
+  LWIP_ASSERT("ret == 0", ret == 0);
+
+  while (sys_now() < max_time) {
+    int closed;
+    int mode = sockets_stresstest_rand_mode(0, do_rx);
+    int timeout_ms = LWIP_RAND() % TEST_MAX_RXWAIT_MS;
+    ret = sockets_stresstest_wait_readable(mode, s, timeout_ms);
+    if (ret) {
+      if (do_rx) {
+        /* read some */
+        LWIP_ASSERT("readable", ret == TEST_SOCK_READABLE);
+        rxoff = recv_and_check_data_return_offset(s, rxbuf, sizeof(rxbuf), rxoff, &closed, "cli");
+        LWIP_ASSERT("client got closed", !closed);
+      }
+    } else {
+      /* timeout, send some */
+      size_t send_len = (LWIP_RAND() % (sizeof(txbuf) - 4)) + 4;
+      fill_test_data(txbuf, send_len);
+      LWIP_DEBUGF(TEST_SOCKETS_STRESS | LWIP_DBG_TRACE, ("cli %d tx %d\n", s, (int)send_len));
+      ret = lwip_write(s, txbuf, send_len);
+      if (ret == -1) {
+        /* TODO: for this to work, 'errno' has to support multithreading... */
+        int err = errno;
+        LWIP_ASSERT("err == 0", err == 0);
+      }
+      LWIP_ASSERT("ret == send_len", ret == (int)send_len);
+    }
+  }
+  if (data) {
+    data->closed = 1;
+  }
+  ret = lwip_close(s);
+  LWIP_ASSERT("ret == 0", ret == 0);
+
+  SYS_ARCH_DEC(sockets_stresstest_numthreads, 1);
+  LWIP_ASSERT("", sockets_stresstest_numthreads >= 0);
+}
+
+static void
+sockets_stresstest_conn_server(void *arg)
+{
+  int s, ret;
+  char txbuf[TEST_TXRX_BUFSIZE];
+  char rxbuf[TEST_TXRX_BUFSIZE];
+  size_t rxoff = 0;
+
+  s = (int)arg;
+
+  while (1) {
+    int closed;
+    int mode = sockets_stresstest_rand_mode(1, 1);
+    int timeout_ms = LWIP_RAND() % TEST_MAX_RXWAIT_MS;
+    ret = sockets_stresstest_wait_readable(mode, s, timeout_ms);
+    if (ret) {
+      if (ret & TEST_SOCK_ERR) {
+        /* closed? */
+        break;
+      }
+      /* read some */
+      LWIP_ASSERT("readable", ret == TEST_SOCK_READABLE);
+      rxoff = recv_and_check_data_return_offset(s, rxbuf, sizeof(rxbuf), rxoff, &closed, "srv");
+      if (closed) {
+        break;
+      }
+    } else {
+      /* timeout, send some */
+      size_t send_len = (LWIP_RAND() % (sizeof(txbuf) - 4)) + 4;
+      fill_test_data(txbuf, send_len);
+      LWIP_DEBUGF(TEST_SOCKETS_STRESS | LWIP_DBG_TRACE, ("srv %d tx %d\n", s, (int)send_len));
+      ret = lwip_write(s, txbuf, send_len);
+      if (ret == -1) {
+        /* TODO: for this to work, 'errno' has to support multithreading... */
+        int err = errno;
+        if (err == ECONNRESET) {
+          break;
+        }
+        if (err == ENOTCONN) {
+          break;
+        }
+        LWIP_ASSERT("unknown error", 0);
+      }
+      LWIP_ASSERT("ret == send_len", ret == (int)send_len);
+    }
+  }
+  ret = lwip_close(s);
+  LWIP_ASSERT("ret == 0", ret == 0);
+
+  SYS_ARCH_DEC(sockets_stresstest_numthreads, 1);
+  LWIP_ASSERT("", sockets_stresstest_numthreads >= 0);
+}
+
+static int
+sockets_stresstest_start_clients(const struct sockaddr_storage *remote_addr)
+{
+  /* limit the number of connections */
+  const int max_connections = LWIP_MIN(TEST_MAX_CONNECTIONS, MEMP_NUM_TCP_PCB/3);
+  int i;
+
+  for (i = 0; i < max_connections; i++) {
+    sys_thread_t t;
+    SYS_ARCH_INC(sockets_stresstest_numthreads, 1);
+    t = sys_thread_new("sockets_stresstest_conn_client", sockets_stresstest_conn_client, (void*)remote_addr, 0, 0);
+    LWIP_ASSERT("thread != NULL", t != 0);
+  }
+  return max_connections;
+}
+
+static void
+sockets_stresstest_listener(void *arg)
+{
+  int slisten;
+  int ret;
+  struct sockaddr_storage addr;
+  socklen_t addr_len;
+  struct test_settings *settings = (struct test_settings *)arg;
+  int num_clients, num_servers = 0;
+
+  slisten = lwip_socket(AF_INET, SOCK_STREAM, 0);
+  LWIP_ASSERT("slisten >= 0", slisten >= 0);
+
+  memcpy(&addr, &settings->addr, sizeof(struct sockaddr_storage));
+  ret = lwip_bind(slisten, (struct sockaddr *)&addr, sizeof(addr));
+  LWIP_ASSERT("ret == 0", ret == 0);
+
+  ret = lwip_listen(slisten, 0);
+  LWIP_ASSERT("ret == 0", ret == 0);
+
+  addr_len = sizeof(addr);
+  ret = lwip_getsockname(slisten, (struct sockaddr *)&addr, &addr_len);
+  LWIP_ASSERT("ret == 0", ret == 0);
+
+  num_clients = sockets_stresstest_start_clients(&addr);
+
+  while (num_servers < num_clients) {
+    struct sockaddr_storage aclient;
+    socklen_t aclient_len = sizeof(aclient);
+    int sclient = lwip_accept(slisten, (struct sockaddr *)&aclient, &aclient_len);
+#if 1
+    /* using server threads */
+    {
+      sys_thread_t t;
+      SYS_ARCH_INC(sockets_stresstest_numthreads, 1);
+      num_servers++;
+      t = sys_thread_new("sockets_stresstest_conn_server", sockets_stresstest_conn_server, (void*)sclient, 0, 0);
+      LWIP_ASSERT("thread != NULL", t != 0);
+    }
+#else
+    /* using server select */
+#endif
+  }
+  LWIP_DEBUGF(TEST_SOCKETS_STRESS | LWIP_DBG_STATE, ("sockets_stresstest_listener: all %d connections established\n", num_clients));
+
+  /* accepted all clients */
+  while (sockets_stresstest_numthreads > 0) {
+    sys_msleep(1);
+  }
+
+  ret = lwip_close(slisten);
+  LWIP_ASSERT("ret == 0", ret == 0);
+
+  LWIP_DEBUGF(TEST_SOCKETS_STRESS |LWIP_DBG_STATE, ("sockets_stresstest_listener: done\n"));
+}
+
+static void
+sockets_stresstest_listener_loop(void *arg)
+{
+  int i;
+  struct test_settings *settings = (struct test_settings *)arg;
+
+  if (settings->loop_cnt) {
+    for (i = 0; i < settings->loop_cnt; i++) {
+      LWIP_DEBUGF(TEST_SOCKETS_STRESS |LWIP_DBG_STATE, ("sockets_stresstest_listener_loop: iteration %d\n", i));
+      sockets_stresstest_listener(arg);
+      sys_msleep(2);
+    }
+    LWIP_DEBUGF(TEST_SOCKETS_STRESS |LWIP_DBG_STATE, ("sockets_stresstest_listener_loop: done\n"));
+  } else {
+    for (i = 0; ; i++) {
+      LWIP_DEBUGF(TEST_SOCKETS_STRESS |LWIP_DBG_STATE, ("sockets_stresstest_listener_loop: iteration %d\n", i));
+      sockets_stresstest_listener(arg);
+      sys_msleep(2);
+    }
+  }
+}
+
+void
+sockets_stresstest_init_loopback(int addr_family)
+{
+  sys_thread_t t;
+  struct test_settings *settings = (struct test_settings *)mem_malloc(sizeof(struct test_settings));
+
+  LWIP_ASSERT("OOM", settings != NULL);
+  memset(settings, 0, sizeof(struct test_settings));
+#if LWIP_IPV4 && LWIP_IPV6
+  LWIP_ASSERT("invalid addr_family", (addr_family == AF_INET) || (addr_family == AF_INET6));
+#endif
+  settings->addr.ss_family = (sa_family_t)addr_family;
+  LWIP_UNUSED_ARG(addr_family);
+  settings->start_client = 1;
+
+  t = sys_thread_new("sockets_stresstest_listener_loop", sockets_stresstest_listener_loop, settings, 0, 0);
+  LWIP_ASSERT("thread != NULL", t != 0);
+}
+
+void
+sockets_stresstest_init_server(int addr_family, u16_t server_port)
+{
+  sys_thread_t t;
+  struct test_settings *settings = (struct test_settings *)mem_malloc(sizeof(struct test_settings));
+
+  LWIP_ASSERT("OOM", settings != NULL);
+  memset(settings, 0, sizeof(struct test_settings));
+#if LWIP_IPV4 && LWIP_IPV6
+  LWIP_ASSERT("invalid addr_family", (addr_family == AF_INET) || (addr_family == AF_INET6));
+  settings->addr.ss_family = (sa_family_t)addr_family;
+#endif
+  LWIP_UNUSED_ARG(addr_family);
+  ((struct sockaddr_in *)(&settings->addr))->sin_port = server_port;
+
+  t = sys_thread_new("sockets_stresstest_listener", sockets_stresstest_listener, settings, 0, 0);
+  LWIP_ASSERT("thread != NULL", t != 0);
+}
+
+void
+sockets_stresstest_init_client(const char *remote_ip, u16_t remote_port)
+{
+#if LWIP_IPV4
+  ip4_addr_t ip4;
+#endif
+#if LWIP_IPV6
+  ip6_addr_t ip6;
+#endif
+  struct sockaddr_storage *addr = (struct sockaddr_storage *)mem_malloc(sizeof(struct sockaddr_storage));
+
+  LWIP_ASSERT("OOM", addr != NULL);
+  memset(addr, 0, sizeof(struct test_settings));
+#if LWIP_IPV4
+  if (ip4addr_aton(remote_ip, &ip4)) {
+    addr->ss_family = AF_INET;
+    ((struct sockaddr_in *)addr)->sin_addr.s_addr = ip4_addr_get_u32(&ip4);
+  }
+#endif
+#if LWIP_IPV4 && LWIP_IPV6
+  else
+#endif
+#if LWIP_IPV6
+  if (ip6addr_aton(remote_ip, &ip6)) {
+    addr->ss_family = AF_INET6;
+    /* todo: copy ipv6 address */
+  }
+#endif
+  ((struct sockaddr_in *)addr)->sin_port = remote_port;
+  sockets_stresstest_start_clients(addr);
+}
+
+#endif /* LWIP_SOCKET && LWIP_IPV4 */

lwip/test/sockets/sockets_stresstest.h

@@ -0,0 +1,40 @@
+/*
+ * Copyright (c) 2017 Simon Goldschmidt
+ * All rights reserved.
+ * 
+ * Redistribution and use in source and binary forms, with or without modification, 
+ * are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission. 
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED 
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT 
+ * SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 
+ * OF SUCH DAMAGE.
+ *
+ * This file is part of the lwIP TCP/IP stack.
+ * 
+ * Author: Simon Goldschmidt <goldsimon@gmx.de>
+ *
+ */
+
+#ifndef LWIP_HDR_TEST_SOCKETS_STRESSTEST
+#define LWIP_HDR_TEST_SOCKETS_STRESSTEST
+
+void sockets_stresstest_init_loopback(int addr_family);
+void sockets_stresstest_init_server(int addr_family, u16_t server_port);
+void sockets_stresstest_init_client(const char *remote_ip, u16_t remote_port);
+
+#endif /* LWIP_HDR_TEST_SOCKETS_STRESSTEST */

lwip/test/unit/Filelists.cmake

@@ -0,0 +1,31 @@
+# This file is indended to be included in end-user CMakeLists.txt
+# include(/path/to/Filelists.cmake)
+# It assumes the variable LWIP_DIR is defined pointing to the
+# root path of lwIP sources.
+#
+# This file is NOT designed (on purpose) to be used as cmake
+# subdir via add_subdirectory()
+# The intention is to provide greater flexibility to users to 
+# create their own targets using the *_SRCS variables.
+
+set(LWIP_TESTDIR ${LWIP_DIR}/test/unit)
+set(LWIP_TESTFILES
+	${LWIP_TESTDIR}/lwip_unittests.c
+	${LWIP_TESTDIR}/api/test_sockets.c
+	${LWIP_TESTDIR}/arch/sys_arch.c
+	${LWIP_TESTDIR}/core/test_def.c
+	${LWIP_TESTDIR}/core/test_mem.c
+	${LWIP_TESTDIR}/core/test_netif.c
+	${LWIP_TESTDIR}/core/test_pbuf.c
+	${LWIP_TESTDIR}/core/test_timers.c
+	${LWIP_TESTDIR}/dhcp/test_dhcp.c
+	${LWIP_TESTDIR}/etharp/test_etharp.c
+	${LWIP_TESTDIR}/ip4/test_ip4.c
+	${LWIP_TESTDIR}/ip6/test_ip6.c
+	${LWIP_TESTDIR}/mdns/test_mdns.c
+	${LWIP_TESTDIR}/mqtt/test_mqtt.c
+	${LWIP_TESTDIR}/tcp/tcp_helper.c
+	${LWIP_TESTDIR}/tcp/test_tcp_oos.c
+	${LWIP_TESTDIR}/tcp/test_tcp.c
+	${LWIP_TESTDIR}/udp/test_udp.c
+)

lwip/test/unit/Filelists.mk

@@ -0,0 +1,51 @@
+#
+# Copyright (c) 2001, 2002 Swedish Institute of Computer Science.
+# All rights reserved. 
+# 
+# Redistribution and use in source and binary forms, with or without modification, 
+# are permitted provided that the following conditions are met:
+#
+# 1. Redistributions of source code must retain the above copyright notice,
+#    this list of conditions and the following disclaimer.
+# 2. Redistributions in binary form must reproduce the above copyright notice,
+#    this list of conditions and the following disclaimer in the documentation
+#    and/or other materials provided with the distribution.
+# 3. The name of the author may not be used to endorse or promote products
+#    derived from this software without specific prior written permission. 
+#
+# THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED 
+# WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 
+# MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT 
+# SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 
+# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 
+# OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
+# INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
+# CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 
+# IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 
+# OF SUCH DAMAGE.
+#
+# This file is part of the lwIP TCP/IP stack.
+# 
+# Author: Adam Dunkels <adam@sics.se>
+#
+
+TESTDIR=$(LWIPDIR)/../test/unit
+TESTFILES=$(TESTDIR)/lwip_unittests.c \
+	$(TESTDIR)/api/test_sockets.c \
+	$(TESTDIR)/arch/sys_arch.c \
+	$(TESTDIR)/core/test_def.c \
+	$(TESTDIR)/core/test_mem.c \
+	$(TESTDIR)/core/test_netif.c \
+	$(TESTDIR)/core/test_pbuf.c \
+	$(TESTDIR)/core/test_timers.c \
+	$(TESTDIR)/dhcp/test_dhcp.c \
+	$(TESTDIR)/etharp/test_etharp.c \
+	$(TESTDIR)/ip4/test_ip4.c \
+	$(TESTDIR)/ip6/test_ip6.c \
+	$(TESTDIR)/mdns/test_mdns.c \
+	$(TESTDIR)/mqtt/test_mqtt.c \
+	$(TESTDIR)/tcp/tcp_helper.c \
+	$(TESTDIR)/tcp/test_tcp_oos.c \
+	$(TESTDIR)/tcp/test_tcp.c \
+	$(TESTDIR)/udp/test_udp.c
+

lwip/test/unit/api/test_sockets.c

@@ -0,0 +1,852 @@
+#include "test_sockets.h"
+
+#include "lwip/mem.h"
+#include "lwip/opt.h"
+#include "lwip/sockets.h"
+#include "lwip/priv/sockets_priv.h"
+#include "lwip/stats.h"
+
+#include "lwip/tcpip.h"
+#include "lwip/priv/tcp_priv.h"
+#include "lwip/api.h"
+
+
+static int
+test_sockets_get_used_count(void)
+{
+  int used = 0;
+  int i;
+
+  for (i = 0; i < NUM_SOCKETS; i++) {
+    struct lwip_sock* s = lwip_socket_dbg_get_socket(i);
+    if (s != NULL) {
+      if (s->fd_used) {
+        used++;
+      }
+    }
+  }
+  return used;
+}
+
+
+/* Setups/teardown functions */
+
+static void
+sockets_setup(void)
+{
+  /* expect full free heap */
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+sockets_teardown(void)
+{
+  fail_unless(test_sockets_get_used_count() == 0);
+  /* poll until all memory is released... */
+  tcpip_thread_poll_one();
+  while (tcp_tw_pcbs) {
+    tcp_abort(tcp_tw_pcbs);
+    tcpip_thread_poll_one();
+  }
+  tcpip_thread_poll_one();
+  /* ensure full free heap */
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+#ifndef NUM_SOCKETS
+#define NUM_SOCKETS MEMP_NUM_NETCONN
+#endif
+
+#if LWIP_SOCKET
+static int
+test_sockets_alloc_socket_nonblocking(int domain, int type)
+{
+  int s = lwip_socket(domain, type, 0);
+  if (s >= 0) {
+    int ret = lwip_fcntl(s, F_SETFL, O_NONBLOCK);
+    fail_unless(ret == 0);
+  }
+  return s;
+}
+
+/* Verify basic sockets functionality
+ */
+START_TEST(test_sockets_basics)
+{
+  int s, i, ret;
+  int s2[NUM_SOCKETS];
+  LWIP_UNUSED_ARG(_i);
+
+  s = lwip_socket(AF_INET, SOCK_STREAM, 0);
+  fail_unless(s >= 0);
+  lwip_close(s);
+
+  for (i = 0; i < NUM_SOCKETS; i++) {
+    s2[i] = lwip_socket(AF_INET, SOCK_STREAM, 0);
+    fail_unless(s2[i] >= 0);
+  }
+
+  /* all sockets used, now it should fail */
+  s = lwip_socket(AF_INET, SOCK_STREAM, 0);
+  fail_unless(s == -1);
+  /* close one socket */
+  ret = lwip_close(s2[0]);
+  fail_unless(ret == 0);
+  /* now it should succeed */
+  s2[0] = lwip_socket(AF_INET, SOCK_STREAM, 0);
+  fail_unless(s2[0] >= 0);
+
+  /* close all sockets */
+  for (i = 0; i < NUM_SOCKETS; i++) {
+    ret = lwip_close(s2[i]);
+    fail_unless(ret == 0);
+  }
+}
+END_TEST
+
+static void test_sockets_allfunctions_basic_domain(int domain)
+{
+  int s, s2, s3, ret;
+  struct sockaddr_storage addr, addr2;
+  socklen_t addrlen, addr2len;
+  char buf[4];
+  /* listen socket */
+  s = lwip_socket(domain, SOCK_STREAM, 0);
+  fail_unless(s >= 0);
+
+  ret = lwip_listen(s, 0);
+  fail_unless(ret == 0);
+
+  addrlen = sizeof(addr);
+  ret = lwip_getsockname(s, (struct sockaddr*)&addr, &addrlen);
+  fail_unless(ret == 0);
+
+  s2 = test_sockets_alloc_socket_nonblocking(domain, SOCK_STREAM);
+  fail_unless(s2 >= 0);
+  /* nonblocking connect s2 to s (but use loopback address) */
+  if (domain == AF_INET) {
+#if LWIP_IPV4
+    struct sockaddr_in *addr4 = (struct sockaddr_in *)&addr;
+    addr4->sin_addr.s_addr = PP_HTONL(INADDR_LOOPBACK);
+#endif
+  } else {
+#if LWIP_IPV6
+    struct sockaddr_in6 *addr6 = (struct sockaddr_in6 *)&addr;
+    struct in6_addr lo6 = IN6ADDR_LOOPBACK_INIT;
+    addr6->sin6_addr = lo6;
+#endif
+  }
+  ret = lwip_connect(s2, (struct sockaddr*)&addr, addrlen);
+  fail_unless(ret == -1);
+  fail_unless(errno == EINPROGRESS);
+  ret = lwip_connect(s2, (struct sockaddr*)&addr, addrlen);
+  fail_unless(ret == -1);
+  fail_unless(errno == EALREADY);
+
+  while(tcpip_thread_poll_one());
+
+  s3 = lwip_accept(s, (struct sockaddr*)&addr2, &addr2len);
+  fail_unless(s3 >= 0);
+
+  ret = lwip_connect(s2, (struct sockaddr*)&addr, addrlen);
+  fail_unless(ret == -1);
+  fail_unless(errno == EISCONN);
+
+  /* write from server to client */
+  ret = write(s3, "test", 4);
+  fail_unless(ret == 4);
+
+  ret = lwip_shutdown(s3, SHUT_WR);
+  fail_unless(ret == 0);
+
+  while(tcpip_thread_poll_one());
+
+  ret = lwip_recv(s2, buf, 3, MSG_PEEK);
+  fail_unless(ret == 3);
+
+  ret = lwip_recv(s2, buf, 3, MSG_PEEK);
+  fail_unless(ret == 3);
+
+  ret = lwip_read(s2, buf, 4);
+  fail_unless(ret == 4);
+
+  ret = lwip_read(s2, buf, 1);
+  fail_unless(ret == 0);
+
+  ret = lwip_read(s2, buf, 1);
+  fail_unless(ret == -1);
+
+  ret = lwip_write(s2, "foo", 3);
+  fail_unless(ret == 3);
+
+  ret = lwip_close(s2);
+  fail_unless(ret == 0);
+
+  while(tcpip_thread_poll_one());
+
+  /* read one byte more than available to check handling FIN */
+  ret = lwip_read(s3, buf, 4);
+  fail_unless(ret == 3);
+
+  ret = lwip_read(s3, buf, 1);
+  fail_unless(ret == 0);
+
+  ret = lwip_read(s3, buf, 1);
+  fail_unless(ret == -1);
+
+  while(tcpip_thread_poll_one());
+
+  ret = lwip_close(s);
+  fail_unless(ret == 0);
+  ret = lwip_close(s3);
+  fail_unless(ret == 0);
+}
+
+/* Try to step through all sockets functions once...
+ */
+START_TEST(test_sockets_allfunctions_basic)
+{
+  LWIP_UNUSED_ARG(_i);
+#if LWIP_IPV4
+  test_sockets_allfunctions_basic_domain(AF_INET);
+#endif
+#if LWIP_IPV6
+  test_sockets_allfunctions_basic_domain(AF_INET6);
+#endif
+}
+END_TEST
+
+static void test_sockets_init_loopback_addr(int domain, struct sockaddr_storage *addr_st, socklen_t *sz)
+{
+  memset(addr_st, 0, sizeof(*addr_st));
+  switch(domain) {
+#if LWIP_IPV6
+    case AF_INET6: {
+      struct sockaddr_in6 *addr = (struct sockaddr_in6*)addr_st;
+      struct in6_addr lo6 = IN6ADDR_LOOPBACK_INIT;
+      addr->sin6_family = AF_INET6;
+      addr->sin6_port = 0; /* use ephemeral port */
+      addr->sin6_addr = lo6;
+      *sz = sizeof(*addr);
+   }
+      break;
+#endif /* LWIP_IPV6 */
+#if LWIP_IPV4
+    case AF_INET: {
+      struct sockaddr_in *addr = (struct sockaddr_in*)addr_st;
+      addr->sin_family = AF_INET;
+      addr->sin_port = 0; /* use ephemeral port */
+      addr->sin_addr.s_addr = PP_HTONL(INADDR_LOOPBACK);
+      *sz = sizeof(*addr);
+    }
+      break;
+#endif /* LWIP_IPV4 */
+    default:
+      *sz = 0;
+      fail();
+      break;
+  }
+}
+
+static void test_sockets_msgapi_update_iovs(struct msghdr *msg, size_t bytes)
+{
+  int i;
+
+  /* note: this modifies the underyling iov_base and iov_len for a partial
+     read for an individual vector. This updates the msg->msg_iov pointer
+     to skip fully consumed vecotrs */
+  
+  /* process fully consumed vectors */
+  for (i = 0; i < msg->msg_iovlen; i++) {
+    if (msg->msg_iov[i].iov_len <= bytes) {
+      /* reduce bytes by amount of this vector */
+      bytes -= msg->msg_iov[i].iov_len;
+    } else {
+      break; /* iov not fully consumed */
+    }
+  }
+
+  /* slide down over fully consumed vectors */
+  msg->msg_iov = &msg->msg_iov[i];
+  msg->msg_iovlen -= i;
+
+  /* update new first vector with any remaining amount */
+  msg->msg_iov[0].iov_base = ((u8_t *)msg->msg_iov[0].iov_base + bytes);
+  msg->msg_iov[0].iov_len -= bytes;
+}
+
+static void test_sockets_msgapi_tcp(int domain)
+{
+  #define BUF_SZ          (TCP_SND_BUF/4)
+  #define TOTAL_DATA_SZ   (BUF_SZ*8) /* ~(TCP_SND_BUF*2) that accounts for integer rounding */
+  #define NEED_TRAILER    (BUF_SZ % 4 != 0)
+  int listnr, s1, s2, i, ret, opt;
+  int bytes_written, bytes_read;
+  struct sockaddr_storage addr_storage;
+  socklen_t addr_size;
+  struct iovec siovs[8];
+  struct msghdr smsg;
+  u8_t * snd_buf;
+  struct iovec riovs[5];
+  struct iovec riovs_tmp[5];
+  struct msghdr rmsg;
+  u8_t * rcv_buf;
+  int    rcv_off;
+  int    rcv_trailer = 0;
+  u8_t val;
+
+  test_sockets_init_loopback_addr(domain, &addr_storage, &addr_size);
+
+  listnr = test_sockets_alloc_socket_nonblocking(domain, SOCK_STREAM);
+  fail_unless(listnr >= 0);
+  s1 = test_sockets_alloc_socket_nonblocking(domain, SOCK_STREAM);
+  fail_unless(s1 >= 0);
+
+  /* setup a listener socket on loopback with ephemeral port */
+  ret = lwip_bind(listnr, (struct sockaddr*)&addr_storage, addr_size);
+  fail_unless(ret == 0);
+  ret = lwip_listen(listnr, 0);
+  fail_unless(ret == 0);
+
+  /* update address with ephemeral port */
+  ret = lwip_getsockname(listnr, (struct sockaddr*)&addr_storage, &addr_size);
+  fail_unless(ret == 0);
+
+  /* connect, won't complete until we accept it */
+  ret = lwip_connect(s1, (struct sockaddr*)&addr_storage, addr_size);
+  fail_unless(ret == -1);
+  fail_unless(errno == EINPROGRESS);
+
+  while (tcpip_thread_poll_one());
+
+  /* accept, creating the other side of the connection */
+  s2 = lwip_accept(listnr, NULL, NULL);
+  fail_unless(s2 >= 0);
+
+  /* double check s1 is connected */
+  ret = lwip_connect(s1, (struct sockaddr*)&addr_storage, addr_size);
+  fail_unless(ret == -1);
+  fail_unless(errno == EISCONN);
+
+  /* set s2 to non-blocking, not inherited from listener */
+  opt = lwip_fcntl(s2, F_GETFL, 0);
+  fail_unless(opt == 6);
+  opt = O_NONBLOCK;
+  ret = lwip_fcntl(s2, F_SETFL, opt);
+  fail_unless(ret == 0);
+
+  /* we are done with listener, close it */
+  ret = lwip_close(listnr);
+  fail_unless(ret == 0);
+
+  /* allocate a buffer for a stream of incrementing hex (0x00..0xFF) which we will use
+     to create an input vector set that is larger than the TCP's send buffer. This will
+     force execution of the partial IO vector send case */
+  snd_buf = (u8_t*)mem_malloc(BUF_SZ);
+  val = 0x00;
+  fail_unless(snd_buf != NULL);
+  for (i = 0; i < BUF_SZ; i++,val++) {
+    snd_buf[i] = val;
+  }
+
+  /* send the buffer 8 times in one message, equating to TOTAL_DATA_SZ */
+  for (i = 0; i < 8; i++) {
+    siovs[i].iov_base = snd_buf;
+    siovs[i].iov_len = BUF_SZ;
+  }
+
+  /* allocate a receive buffer, same size as snd_buf for easy verification */
+  rcv_buf = (u8_t*)mem_calloc(1, BUF_SZ);
+  fail_unless(rcv_buf != NULL);
+  /* split across iovs */
+  for (i = 0; i < 4; i++) {
+    riovs[i].iov_base = &rcv_buf[i*(BUF_SZ/4)];
+    riovs[i].iov_len = BUF_SZ/4;
+  }
+  /* handling trailing bytes if buffer doesn't evenly divide by 4 */
+#if NEED_TRAILER
+  if ((BUF_SZ % 4) != 0) {
+    riovs[5].iov_base = &rcv_buf[4*(BUF_SZ/4)];
+    riovs[5].iov_len = BUF_SZ - (4*(BUF_SZ/4));
+    rcv_trailer = 1;
+  }
+#endif /* NEED_TRAILER */
+
+  /* we use a copy of riovs since we'll be modifying base and len during
+     receiving. This gives us an easy way to reset the iovs for next recvmsg */
+  memcpy(riovs_tmp, riovs, sizeof(riovs));
+
+  memset(&smsg, 0, sizeof(smsg));
+  smsg.msg_iov = siovs;
+  smsg.msg_iovlen = 8;
+
+  memset(&rmsg, 0, sizeof(rmsg));
+  rmsg.msg_iov = riovs_tmp;
+  rmsg.msg_iovlen = (rcv_trailer ? 5 : 4);
+
+  bytes_written = 0;
+  bytes_read = 0;
+  rcv_off = 0;
+
+  while (bytes_written < TOTAL_DATA_SZ && (bytes_read < TOTAL_DATA_SZ)) {
+    /* send data */
+    if (bytes_written < TOTAL_DATA_SZ) {
+      ret = lwip_sendmsg(s1, &smsg, 0);
+      /* note: since we always receive after sending, there will be open
+         space in the send buffer */
+      fail_unless(ret > 0);
+    
+      bytes_written += ret;
+      if (bytes_written < TOTAL_DATA_SZ) {
+        test_sockets_msgapi_update_iovs(&smsg, (size_t)ret);
+      }
+    }
+
+    while (tcpip_thread_poll_one());
+
+    /* receive and verify data */
+    do {
+      if (bytes_read < TOTAL_DATA_SZ) {
+        ret = lwip_recvmsg(s2, &rmsg, 0);
+        fail_unless(ret > 0 || (ret == -1 && errno == EWOULDBLOCK));
+
+        if (ret > 0) {
+          rcv_off += ret;
+          /* we have received a full buffer */
+          if (rcv_off == BUF_SZ) {
+            /* note: since iovs are just pointers, compare underlying buf */
+            fail_unless(!memcmp(snd_buf, rcv_buf, BUF_SZ));
+            bytes_read += BUF_SZ;
+            /* reset receive state for next buffer */
+            rcv_off = 0;
+            memset(rcv_buf, 0, BUF_SZ);
+            memcpy(riovs_tmp, riovs, sizeof(riovs));
+            rmsg.msg_iov = riovs_tmp;
+            rmsg.msg_iovlen = (rcv_trailer ? 5 : 4);
+          } else { /* partial read */
+            test_sockets_msgapi_update_iovs(&rmsg, (size_t)ret);
+          }
+        }
+      } else {
+        break;
+      }
+    } while(ret > 0);
+  }
+  
+  ret = lwip_close(s1);
+  fail_unless(ret == 0);
+  ret = lwip_close(s2);
+  fail_unless(ret == 0);
+  mem_free(snd_buf);
+  mem_free(rcv_buf);
+}
+
+static void test_sockets_msgapi_udp_send_recv_loop(int s, struct msghdr *smsg, struct msghdr *rmsg)
+{
+  int i, ret;
+
+  /* send/receive our datagram of IO vectors 10 times */
+  for (i = 0; i < 10; i++) {
+    ret = lwip_sendmsg(s, smsg, 0);
+    fail_unless(ret == 4);
+
+    while (tcpip_thread_poll_one());
+
+    /* receive the datagram split across 4 buffers */
+    ret = lwip_recvmsg(s, rmsg, 0);
+    fail_unless(ret == 4);
+
+    /* verify data */
+    fail_unless(*((u8_t*)rmsg->msg_iov[0].iov_base) == 0xDE);
+    fail_unless(*((u8_t*)rmsg->msg_iov[1].iov_base) == 0xAD);
+    fail_unless(*((u8_t*)rmsg->msg_iov[2].iov_base) == 0xBE);
+    fail_unless(*((u8_t*)rmsg->msg_iov[3].iov_base) == 0xEF);
+
+    /* clear rcv_buf to ensure no data is being skipped */
+    *((u8_t*)rmsg->msg_iov[0].iov_base) = 0x00;
+    *((u8_t*)rmsg->msg_iov[1].iov_base) = 0x00;
+    *((u8_t*)rmsg->msg_iov[2].iov_base) = 0x00;
+    *((u8_t*)rmsg->msg_iov[3].iov_base) = 0x00;
+  }
+}
+
+static void test_sockets_msgapi_udp(int domain)
+{
+  int s, i, ret;
+  struct sockaddr_storage addr_storage;
+  socklen_t addr_size;
+  struct iovec riovs[4];
+  struct msghdr rmsg;
+  u8_t rcv_buf[4];
+  struct iovec siovs[4];
+  struct msghdr smsg;
+  u8_t snd_buf[4] = {0xDE, 0xAD, 0xBE, 0xEF};
+
+  /* initialize IO vectors with data */
+  for (i = 0; i < 4; i++) {
+    siovs[i].iov_base = &snd_buf[i];
+    siovs[i].iov_len = sizeof(u8_t);
+    riovs[i].iov_base = &rcv_buf[i];
+    riovs[i].iov_len = sizeof(u8_t);
+  }
+
+  test_sockets_init_loopback_addr(domain, &addr_storage, &addr_size);
+
+  s = test_sockets_alloc_socket_nonblocking(domain, SOCK_DGRAM);
+  fail_unless(s >= 0);
+
+  ret = lwip_bind(s, (struct sockaddr*)&addr_storage, addr_size);
+  fail_unless(ret == 0);
+
+  /* Update addr with epehermal port */
+  ret = lwip_getsockname(s, (struct sockaddr*)&addr_storage, &addr_size);
+  fail_unless(ret == 0);
+  switch(domain) {
+#if LWIP_IPV6
+    case AF_INET6:
+      fail_unless(addr_size == sizeof(struct sockaddr_in6));
+      break;
+#endif /* LWIP_IPV6 */
+#if LWIP_IPV4
+    case AF_INET:
+        fail_unless(addr_size == sizeof(struct sockaddr_in));
+        break;
+#endif /* LWIP_IPV6 */
+    default:
+      fail();
+      break;
+  }
+
+  /* send and receive the datagram in 4 pieces */
+  memset(&smsg, 0, sizeof(smsg));
+  smsg.msg_iov = siovs;
+  smsg.msg_iovlen = 4;
+  memset(&rmsg, 0, sizeof(rmsg));
+  rmsg.msg_iov = riovs;
+  rmsg.msg_iovlen = 4;
+
+  /* perform a sendmsg with remote host (self) */
+  smsg.msg_name = &addr_storage;
+  smsg.msg_namelen = addr_size;
+
+  test_sockets_msgapi_udp_send_recv_loop(s, &smsg, &rmsg);
+
+  /* Connect to self, allowing us to not pass message name */
+  ret = lwip_connect(s, (struct sockaddr*)&addr_storage, addr_size);
+  fail_unless(ret == 0);
+
+  smsg.msg_name = NULL;
+  smsg.msg_namelen = 0;
+
+  test_sockets_msgapi_udp_send_recv_loop(s, &smsg, &rmsg);
+
+  ret = lwip_close(s);
+  fail_unless(ret == 0);
+}
+
+#if LWIP_IPV4
+static void test_sockets_msgapi_cmsg(int domain)
+{
+  int s, ret, enable;
+  struct sockaddr_storage addr_storage;
+  socklen_t addr_size;
+  struct iovec iov;
+  struct msghdr msg;
+  struct cmsghdr *cmsg;
+  struct in_pktinfo *pktinfo;
+  u8_t rcv_buf[4];
+  u8_t snd_buf[4] = {0xDE, 0xAD, 0xBE, 0xEF};
+  u8_t cmsg_buf[CMSG_SPACE(sizeof(struct in_pktinfo))];
+
+  test_sockets_init_loopback_addr(domain, &addr_storage, &addr_size);
+
+  s = test_sockets_alloc_socket_nonblocking(domain, SOCK_DGRAM);
+  fail_unless(s >= 0);
+
+  ret = lwip_bind(s, (struct sockaddr*)&addr_storage, addr_size);
+  fail_unless(ret == 0);
+
+  /* Update addr with epehermal port */
+  ret = lwip_getsockname(s, (struct sockaddr*)&addr_storage, &addr_size);
+  fail_unless(ret == 0);
+
+  enable = 1;
+  ret = lwip_setsockopt(s, IPPROTO_IP, IP_PKTINFO, &enable, sizeof(enable));
+  fail_unless(ret == 0);
+
+  /* Receive full message, including control message */
+  iov.iov_base = rcv_buf;
+  iov.iov_len = sizeof(rcv_buf);
+  msg.msg_control = cmsg_buf;
+  msg.msg_controllen = sizeof(cmsg_buf);
+  msg.msg_flags = 0;
+  msg.msg_iov = &iov;
+  msg.msg_iovlen = 1;
+  msg.msg_name = NULL;
+  msg.msg_namelen = 0;
+
+  memset(rcv_buf, 0, sizeof(rcv_buf));
+  ret = lwip_sendto(s, snd_buf, sizeof(snd_buf), 0, (struct sockaddr*)&addr_storage, addr_size);
+  fail_unless(ret == sizeof(snd_buf));
+  
+  tcpip_thread_poll_one();
+
+  ret = lwip_recvmsg(s, &msg, 0);
+  fail_unless(ret == sizeof(rcv_buf));
+  fail_unless(!memcmp(rcv_buf, snd_buf, sizeof(rcv_buf)));
+  
+  /* Verify message header */
+  cmsg = CMSG_FIRSTHDR(&msg);
+  fail_unless(cmsg != NULL);
+  fail_unless(cmsg->cmsg_len > 0);
+  fail_unless(cmsg->cmsg_level == IPPROTO_IP);
+  fail_unless(cmsg->cmsg_type == IP_PKTINFO);
+
+  /* Verify message data */
+  pktinfo = (struct in_pktinfo*)CMSG_DATA(cmsg);
+  /* We only have loopback interface enabled */
+  fail_unless(pktinfo->ipi_ifindex == 1);
+  fail_unless(pktinfo->ipi_addr.s_addr == PP_HTONL(INADDR_LOOPBACK));
+
+  /* Verify there are no additional messages */
+  cmsg = CMSG_NXTHDR(&msg, cmsg);
+  fail_unless(cmsg == NULL);
+
+  /* Send datagram again, testing truncation */
+  memset(rcv_buf, 0, sizeof(rcv_buf));
+  ret = lwip_sendto(s, snd_buf, sizeof(snd_buf), 0, (struct sockaddr*)&addr_storage, addr_size);
+  fail_unless(ret == sizeof(snd_buf));
+
+  tcpip_thread_poll_one();
+
+  msg.msg_controllen = 1;
+  msg.msg_flags = 0;
+  ret = lwip_recvmsg(s, &msg, 0);
+  fail_unless(ret == sizeof(rcv_buf));
+  fail_unless(!memcmp(rcv_buf, snd_buf, sizeof(rcv_buf)));
+  /* Ensure truncation was returned */
+  fail_unless(msg.msg_flags & MSG_CTRUNC);
+  /* Ensure no control messages were returned */
+  fail_unless(msg.msg_controllen == 0);
+
+  ret = lwip_close(s);
+  fail_unless(ret == 0);
+}
+#endif /* LWIP_IPV4 */
+
+START_TEST(test_sockets_msgapis)
+{
+  LWIP_UNUSED_ARG(_i);
+#if LWIP_IPV4
+  test_sockets_msgapi_udp(AF_INET);
+  test_sockets_msgapi_tcp(AF_INET);
+  test_sockets_msgapi_cmsg(AF_INET);
+#endif
+#if LWIP_IPV6
+  test_sockets_msgapi_udp(AF_INET6);
+  test_sockets_msgapi_tcp(AF_INET6);
+#endif
+}
+END_TEST
+
+START_TEST(test_sockets_select)
+{
+#if LWIP_SOCKET_SELECT
+  int s;
+  int ret;
+  fd_set readset;
+  fd_set writeset;
+  fd_set errset;
+  struct timeval tv;
+
+  fail_unless(test_sockets_get_used_count() == 0);
+
+  s = lwip_socket(AF_INET, SOCK_STREAM, 0);
+  fail_unless(s >= 0);
+  fail_unless(test_sockets_get_used_count() == 0);
+
+  FD_ZERO(&readset);
+  FD_SET(s, &readset);
+  FD_ZERO(&writeset);
+  FD_SET(s, &writeset);
+  FD_ZERO(&errset);
+  FD_SET(s, &errset);
+
+  tv.tv_sec = tv.tv_usec = 0;
+  ret = lwip_select(s + 1, &readset, &writeset, &errset, &tv);
+  fail_unless(ret == 0);
+  fail_unless(test_sockets_get_used_count() == 0);
+
+  ret = lwip_close(s);
+  fail_unless(ret == 0);
+
+#endif
+  LWIP_UNUSED_ARG(_i);
+}
+END_TEST
+
+START_TEST(test_sockets_recv_after_rst)
+{
+  int sl, sact;
+  int spass = -1;
+  int ret;
+  struct sockaddr_in sa_listen;
+  const u16_t port = 1234;
+  int arg;
+  const char txbuf[] = "something";
+  char rxbuf[16];
+  struct lwip_sock *sact_sock;
+  int err;
+  LWIP_UNUSED_ARG(_i);
+
+  fail_unless(test_sockets_get_used_count() == 0);
+
+  memset(&sa_listen, 0, sizeof(sa_listen));
+  sa_listen.sin_family = AF_INET;
+  sa_listen.sin_port = PP_HTONS(port);
+  sa_listen.sin_addr.s_addr = PP_HTONL(INADDR_LOOPBACK);
+
+  /* set up the listener */
+  sl = lwip_socket(AF_INET, SOCK_STREAM, 0);
+  fail_unless(sl >= 0);
+  fail_unless(test_sockets_get_used_count() == 0);
+
+  ret = lwip_bind(sl, (struct sockaddr *)&sa_listen, sizeof(sa_listen));
+  fail_unless(ret == 0);
+  ret = lwip_listen(sl, 0);
+  fail_unless(ret == 0);
+
+  /* set up the client */
+  sact = lwip_socket(AF_INET, SOCK_STREAM, 0);
+  fail_unless(sact >= 0);
+  fail_unless(test_sockets_get_used_count() == 0);
+  /* set the client to nonblocking to simplify this test */
+  arg = 1;
+  ret = lwip_ioctl(sact, FIONBIO, &arg);
+  fail_unless(ret == 0);
+  /* connect */
+  do {
+    ret = lwip_connect(sact, (struct sockaddr *)&sa_listen, sizeof(sa_listen));
+    err = errno;
+    fail_unless((ret == 0) || (ret == -1));
+    if (ret != 0) {
+      if (err == EISCONN) {
+        /* Although this is not valid, use EISCONN as an indicator for successful connection.
+           This marks us as "connect phase is done". On error, we would either have a different
+           errno code or "send" fails later... -> good enough for this test. */
+        ret = 0;
+      } else {
+        fail_unless(err == EINPROGRESS);
+        if (err != EINPROGRESS) {
+          goto cleanup;
+        }
+        /* we're in progress: little side check: test for EALREADY */
+        ret = lwip_connect(sact, (struct sockaddr *)&sa_listen, sizeof(sa_listen));
+        err = errno;
+        fail_unless(ret == -1);
+        fail_unless(err == EALREADY);
+        if ((ret != -1) || (err != EALREADY)) {
+          goto cleanup;
+        }
+      }
+      tcpip_thread_poll_one();
+      tcpip_thread_poll_one();
+      tcpip_thread_poll_one();
+      tcpip_thread_poll_one();
+    }
+  } while (ret != 0);
+  fail_unless(ret == 0);
+
+  /* accept the server connection part */
+  spass = lwip_accept(sl, NULL, NULL);
+  fail_unless(spass >= 0);
+
+  /* write data from client */
+  ret = lwip_send(sact, txbuf, sizeof(txbuf), 0);
+  fail_unless(ret == sizeof(txbuf));
+
+  tcpip_thread_poll_one();
+  tcpip_thread_poll_one();
+
+  /* issue RST (This is a HACK, don't try this in your own app!) */
+  sact_sock = lwip_socket_dbg_get_socket(sact);
+  fail_unless(sact_sock != NULL);
+  if (sact_sock != NULL) {
+    struct netconn *sact_conn = sact_sock->conn;
+    fail_unless(sact_conn != NULL);
+    if (sact_conn != NULL) {
+      struct tcp_pcb *pcb = sact_conn->pcb.tcp;
+      fail_unless(pcb != NULL);
+      if (pcb != NULL) {
+        tcp_rst(pcb, pcb->snd_nxt, pcb->rcv_nxt, &pcb->local_ip, &pcb->remote_ip,
+                     pcb->local_port, pcb->remote_port);
+      }
+    }
+  }
+  tcpip_thread_poll_one();
+  tcpip_thread_poll_one();
+
+  /* expect to receive data first */
+  ret = lwip_recv(spass, rxbuf, sizeof(rxbuf), 0);
+  fail_unless(ret > 0);
+  tcpip_thread_poll_one();
+  tcpip_thread_poll_one();
+
+  /* expect to receive RST indication */
+  ret = lwip_recv(spass, rxbuf, sizeof(rxbuf), 0);
+  fail_unless(ret == -1);
+  err = errno;
+  fail_unless(err == ECONNRESET);
+  tcpip_thread_poll_one();
+  tcpip_thread_poll_one();
+
+  /* expect to receive ENOTCONN indication */
+  ret = lwip_recv(spass, rxbuf, sizeof(rxbuf), 0);
+  fail_unless(ret == -1);
+  err = errno;
+  fail_unless(err == ENOTCONN);
+  tcpip_thread_poll_one();
+  tcpip_thread_poll_one();
+
+  /* expect to receive ENOTCONN indication */
+  ret = lwip_recv(spass, rxbuf, sizeof(rxbuf), 0);
+  fail_unless(ret == -1);
+  err = errno;
+  fail_unless(err == ENOTCONN);
+  tcpip_thread_poll_one();
+  tcpip_thread_poll_one();
+
+cleanup:
+  ret = lwip_close(sl);
+  fail_unless(ret == 0);
+  ret = lwip_close(sact);
+  fail_unless(ret == 0);
+  if (spass >= 0) {
+    ret = lwip_close(spass);
+    fail_unless(ret == 0);
+  }
+}
+END_TEST
+
+/** Create the suite including all tests for this module */
+Suite *
+sockets_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_sockets_basics),
+    TESTFUNC(test_sockets_allfunctions_basic),
+    TESTFUNC(test_sockets_msgapis),
+    TESTFUNC(test_sockets_select),
+    TESTFUNC(test_sockets_recv_after_rst),
+  };
+  return create_suite("SOCKETS", tests, sizeof(tests)/sizeof(testfunc), sockets_setup, sockets_teardown);
+}
+
+#else /* LWIP_SOCKET */
+
+Suite *
+sockets_suite(void)
+{
+  return create_suite("SOCKETS", NULL, 0, NULL, NULL);
+}
+#endif /* LWIP_SOCKET */

lwip/test/unit/api/test_sockets.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_SOCKETS_H
+#define LWIP_HDR_TEST_SOCKETS_H
+
+#include "../lwip_check.h"
+
+Suite *sockets_suite(void);
+
+#endif

lwip/test/unit/arch/sys_arch.c

@@ -0,0 +1,371 @@
+/*
+ * Copyright (c) 2017 Simon Goldschmidt
+ * All rights reserved. 
+ * 
+ * Redistribution and use in source and binary forms, with or without modification, 
+ * are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission. 
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED 
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT 
+ * SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 
+ * OF SUCH DAMAGE.
+ *
+ * This file is part of the lwIP TCP/IP stack.
+ * 
+ * Author: Simon Goldschmidt
+ *
+ */
+
+
+#include <lwip/opt.h>
+#include <lwip/arch.h>
+#if !NO_SYS
+#include "sys_arch.h"
+#endif
+#include <lwip/stats.h>
+#include <lwip/debug.h>
+#include <lwip/sys.h>
+
+#include <string.h>
+
+u32_t lwip_sys_now;
+
+u32_t
+sys_jiffies(void)
+{
+  return lwip_sys_now;
+}
+
+u32_t
+sys_now(void)
+{
+  return lwip_sys_now;
+}
+
+void
+sys_init(void)
+{
+}
+
+#if !NO_SYS
+
+test_sys_arch_waiting_fn the_waiting_fn;
+
+void
+test_sys_arch_wait_callback(test_sys_arch_waiting_fn waiting_fn)
+{
+  the_waiting_fn = waiting_fn;
+}
+
+err_t
+sys_sem_new(sys_sem_t *sem, u8_t count)
+{
+  LWIP_ASSERT("sem != NULL", sem != NULL);
+  *sem = count + 1;
+  return ERR_OK;
+}
+
+void
+sys_sem_free(sys_sem_t *sem)
+{
+  LWIP_ASSERT("sem != NULL", sem != NULL);
+  *sem = 0;
+}
+
+void
+sys_sem_set_invalid(sys_sem_t *sem)
+{
+  LWIP_ASSERT("sem != NULL", sem != NULL);
+  *sem = 0;
+}
+
+/* semaphores are 1-based because RAM is initialized as 0, which would be valid */
+u32_t
+sys_arch_sem_wait(sys_sem_t *sem, u32_t timeout)
+{
+  u32_t ret = 0;
+  LWIP_ASSERT("sem != NULL", sem != NULL);
+  LWIP_ASSERT("*sem > 0", *sem > 0);
+  if (*sem == 1) {
+    /* need to wait */
+    if(!timeout)
+    {
+      /* wait infinite */
+      LWIP_ASSERT("cannot wait without waiting callback", the_waiting_fn != NULL);
+      do {
+        int expectSomething = the_waiting_fn(sem, NULL);
+        LWIP_ASSERT("*sem > 0", *sem > 0);
+        LWIP_ASSERT("expecting a semaphore count but it's 0", !expectSomething || (*sem > 1));
+        ret++;
+        if (ret == SYS_ARCH_TIMEOUT) {
+          ret--;
+        }
+      } while(*sem == 1);
+    }
+    else
+    {
+      if (the_waiting_fn) {
+        int expectSomething = the_waiting_fn(sem, NULL);
+        LWIP_ASSERT("expecting a semaphore count but it's 0", !expectSomething || (*sem > 1));
+      }
+      LWIP_ASSERT("*sem > 0", *sem > 0);
+      if (*sem == 1) {
+        return SYS_ARCH_TIMEOUT;
+      }
+      ret = 1;
+    }
+  }
+  LWIP_ASSERT("*sem > 0", *sem > 0);
+  (*sem)--;
+  LWIP_ASSERT("*sem > 0", *sem > 0);
+  /* return the time we waited for the sem */
+  return ret;
+}
+
+void
+sys_sem_signal(sys_sem_t *sem)
+{
+  LWIP_ASSERT("sem != NULL", sem != NULL);
+  LWIP_ASSERT("*sem > 0", *sem > 0);
+  (*sem)++;
+  LWIP_ASSERT("*sem > 0", *sem > 0);
+}
+
+err_t
+sys_mutex_new(sys_mutex_t *mutex)
+{
+  LWIP_ASSERT("mutex != NULL", mutex != NULL);
+  *mutex = 1; /* 1 allocated */
+  return ERR_OK;
+}
+
+void
+sys_mutex_free(sys_mutex_t *mutex)
+{
+  /* parameter check */
+  LWIP_ASSERT("mutex != NULL", mutex != NULL);
+  LWIP_ASSERT("*mutex >= 1", *mutex >= 1);
+  *mutex = 0;
+}
+
+void
+sys_mutex_set_invalid(sys_mutex_t *mutex)
+{
+  LWIP_ASSERT("mutex != NULL", mutex != NULL);
+  *mutex = 0;
+}
+
+void
+sys_mutex_lock(sys_mutex_t *mutex)
+{
+  /* nothing to do, no multithreading supported */
+  LWIP_ASSERT("mutex != NULL", mutex != NULL);
+  /* check that the mutext is valid and unlocked (no nested locking) */
+  LWIP_ASSERT("*mutex >= 1", *mutex == 1);
+  /* we count up just to check the correct pairing of lock/unlock */
+  (*mutex)++;
+  LWIP_ASSERT("*mutex >= 1", *mutex >= 1);
+}
+
+void
+sys_mutex_unlock(sys_mutex_t *mutex)
+{
+  /* nothing to do, no multithreading supported */
+  LWIP_ASSERT("mutex != NULL", mutex != NULL);
+  LWIP_ASSERT("*mutex >= 1", *mutex >= 1);
+  /* we count down just to check the correct pairing of lock/unlock */
+  (*mutex)--;
+  LWIP_ASSERT("*mutex >= 1", *mutex >= 1);
+}
+
+
+sys_thread_t
+sys_thread_new(const char *name, lwip_thread_fn function, void *arg, int stacksize, int prio)
+{
+  LWIP_UNUSED_ARG(name);
+  LWIP_UNUSED_ARG(function);
+  LWIP_UNUSED_ARG(arg);
+  LWIP_UNUSED_ARG(stacksize);
+  LWIP_UNUSED_ARG(prio);
+  /* threads not supported */
+  return 0;
+}
+
+err_t
+sys_mbox_new(sys_mbox_t *mbox, int size)
+{
+  int mboxsize = size;
+  LWIP_ASSERT("mbox != NULL", mbox != NULL);
+  LWIP_ASSERT("size >= 0", size >= 0);
+  if (size == 0) {
+    mboxsize = 1024;
+  }
+  mbox->head = mbox->tail = 0;
+  mbox->sem = mbox; /* just point to something for sys_mbox_valid() */
+  mbox->q_mem = (void**)malloc(sizeof(void*)*mboxsize);
+  mbox->size = mboxsize;
+  mbox->used = 0;
+
+  memset(mbox->q_mem, 0, sizeof(void*)*mboxsize);
+  return ERR_OK;
+}
+
+void
+sys_mbox_free(sys_mbox_t *mbox)
+{
+  /* parameter check */
+  LWIP_ASSERT("mbox != NULL", mbox != NULL);
+  LWIP_ASSERT("mbox->sem != NULL", mbox->sem != NULL);
+  LWIP_ASSERT("mbox->sem == mbox", mbox->sem == mbox);
+  LWIP_ASSERT("mbox->q_mem != NULL", mbox->q_mem != NULL);
+  mbox->sem = NULL;
+  free(mbox->q_mem);
+  mbox->q_mem = NULL;
+}
+
+void
+sys_mbox_set_invalid(sys_mbox_t *mbox)
+{
+  LWIP_ASSERT("mbox != NULL", mbox != NULL);
+  LWIP_ASSERT("mbox->q_mem == NULL", mbox->q_mem == NULL);
+  mbox->sem = NULL;
+  mbox->q_mem = NULL;
+}
+
+void
+sys_mbox_post(sys_mbox_t *q, void *msg)
+{
+  LWIP_ASSERT("q != SYS_MBOX_NULL", q != SYS_MBOX_NULL);
+  LWIP_ASSERT("q->sem == q", q->sem == q);
+  LWIP_ASSERT("q->q_mem != NULL", q->q_mem != NULL);
+  LWIP_ASSERT("q->used >= 0", q->used >= 0);
+  LWIP_ASSERT("q->size > 0", q->size > 0);
+
+  LWIP_ASSERT("mbox already full", q->used < q->size);
+
+  q->q_mem[q->head] = msg;
+  q->head++;
+  if (q->head >= (unsigned int)q->size) {
+    q->head = 0;
+  }
+  LWIP_ASSERT("mbox is full!", q->head != q->tail);
+  q->used++;
+}
+
+err_t
+sys_mbox_trypost(sys_mbox_t *q, void *msg)
+{
+  LWIP_ASSERT("q != SYS_MBOX_NULL", q != SYS_MBOX_NULL);
+  LWIP_ASSERT("q->sem == q", q->sem == q);
+  LWIP_ASSERT("q->q_mem != NULL", q->q_mem != NULL);
+  LWIP_ASSERT("q->used >= 0", q->used >= 0);
+  LWIP_ASSERT("q->size > 0", q->size > 0);
+  LWIP_ASSERT("q->used <= q->size", q->used <= q->size);
+
+  if (q->used == q->size) {
+    return ERR_MEM;
+  }
+  sys_mbox_post(q, msg);
+  return ERR_OK;
+}
+
+err_t
+sys_mbox_trypost_fromisr(sys_mbox_t *q, void *msg)
+{
+  return sys_mbox_trypost(q, msg);
+}
+
+u32_t
+sys_arch_mbox_fetch(sys_mbox_t *q, void **msg, u32_t timeout)
+{
+  u32_t ret = 0;
+  u32_t ret2;
+  LWIP_ASSERT("q != SYS_MBOX_NULL", q != SYS_MBOX_NULL);
+  LWIP_ASSERT("q->sem == q", q->sem == q);
+  LWIP_ASSERT("q->q_mem != NULL", q->q_mem != NULL);
+  LWIP_ASSERT("q->used >= 0", q->used >= 0);
+  LWIP_ASSERT("q->size > 0", q->size > 0);
+
+  if (q->used == 0) {
+    /* need to wait */
+    /* need to wait */
+    if(!timeout)
+    {
+      /* wait infinite */
+      LWIP_ASSERT("cannot wait without waiting callback", the_waiting_fn != NULL);
+      do {
+        int expectSomething = the_waiting_fn(NULL, q);
+        LWIP_ASSERT("q->used >= 0", q->used >= 0);
+        LWIP_ASSERT("expecting item available but it's 0", !expectSomething || (q->used > 0));
+        ret++;
+        if (ret == SYS_ARCH_TIMEOUT) {
+          ret--;
+        }
+      } while(q->used == 0);
+    }
+    else
+    {
+      if (the_waiting_fn) {
+        int expectSomething = the_waiting_fn(NULL, q);
+        LWIP_ASSERT("expecting item available count but it's 0", !expectSomething || (q->used > 0));
+      }
+      LWIP_ASSERT("q->used >= 0", q->used >= 0);
+      if (q->used == 0) {
+        if(msg) {
+          *msg = NULL;
+        }
+        return SYS_ARCH_TIMEOUT;
+      }
+      ret = 1;
+    }
+  }
+  LWIP_ASSERT("q->used > 0", q->used > 0);
+  ret2 = sys_arch_mbox_tryfetch(q, msg);
+  LWIP_ASSERT("got no message", ret2 == 0);
+  return ret;
+}
+
+u32_t
+sys_arch_mbox_tryfetch(sys_mbox_t *q, void **msg)
+{
+  LWIP_ASSERT("q != SYS_MBOX_NULL", q != SYS_MBOX_NULL);
+  LWIP_ASSERT("q->sem == q", q->sem == q);
+  LWIP_ASSERT("q->q_mem != NULL", q->q_mem != NULL);
+  LWIP_ASSERT("q->used >= 0", q->used >= 0);
+  LWIP_ASSERT("q->size > 0", q->size > 0);
+
+  if (!q->used) {
+    return SYS_ARCH_TIMEOUT;
+  }
+  if(msg) {
+    *msg = q->q_mem[q->tail];
+  }
+
+  q->tail++;
+  if (q->tail >= (unsigned int)q->size) {
+    q->tail = 0;
+  }
+  q->used--;
+  LWIP_ASSERT("q->used >= 0", q->used >= 0);
+  return 0;
+}
+
+#if LWIP_NETCONN_SEM_PER_THREAD
+#error LWIP_NETCONN_SEM_PER_THREAD==1 not supported
+#endif /* LWIP_NETCONN_SEM_PER_THREAD */
+
+#endif /* !NO_SYS */

lwip/test/unit/arch/sys_arch.h

@@ -0,0 +1,72 @@
+/*
+ * Copyright (c) 2017 Simon Goldschmidt
+ * All rights reserved. 
+ * 
+ * Redistribution and use in source and binary forms, with or without modification, 
+ * are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission. 
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED 
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT 
+ * SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 
+ * OF SUCH DAMAGE.
+ *
+ * This file is part of the lwIP TCP/IP stack.
+ * 
+ * Author: Simon Goldschmidt
+ *
+ */
+#ifndef LWIP_HDR_TEST_SYS_ARCH_H
+#define LWIP_HDR_TEST_SYS_ARCH_H
+
+typedef int sys_sem_t;
+#define sys_sem_valid(sema) ((sema) != NULL)
+
+typedef int sys_mutex_t;
+#define sys_mutex_valid(mutex) (((mutex) != NULL)
+
+struct lwip_mbox {
+  void* sem;
+  void** q_mem;
+  unsigned int head, tail;
+  int size;
+  int used;
+};
+typedef struct lwip_mbox sys_mbox_t;
+#define SYS_MBOX_NULL NULL
+#define sys_mbox_valid(mbox) ((mbox != NULL) && ((mbox)->sem != NULL)  && ((mbox)->sem != (void*)-1))
+#define sys_mbox_valid_val(mbox) (((mbox).sem != NULL)  && ((mbox).sem != (void*)-1))
+
+/* DWORD (thread id) is used for sys_thread_t but we won't include windows.h */
+typedef u32_t sys_thread_t;
+
+#define SYS_ARCH_DECL_PROTECT(lev)
+#define SYS_ARCH_PROTECT(lev)
+#define SYS_ARCH_UNPROTECT(lev)
+
+/* to implement doing something while blocking on an mbox or semaphore:
+ * pass a function to test_sys_arch_wait_callback() that returns
+ * '0' if waiting again and
+ * '1' if now there should be something to do (used for asserting)
+ */
+typedef int (*test_sys_arch_waiting_fn)(sys_sem_t* wait_sem, sys_mbox_t* wait_mbox);
+void test_sys_arch_wait_callback(test_sys_arch_waiting_fn waiting_fn);
+
+/* current time */
+extern u32_t lwip_sys_now;
+
+#endif /* LWIP_HDR_TEST_SYS_ARCH_H */
+

lwip/test/unit/core/test_def.c

@@ -0,0 +1,84 @@
+#include "test_def.h"
+
+#include "lwip/def.h"
+
+#define MAGIC_UNTOUCHED_BYTE  0x7a
+#define TEST_BUFSIZE          32
+#define GUARD_SIZE            4
+
+/* Setups/teardown functions */
+
+static void
+def_setup(void)
+{
+}
+
+static void
+def_teardown(void)
+{
+}
+
+static void
+def_check_range_untouched(const char *buf, size_t len)
+{
+  size_t i;
+
+  for (i = 0; i < len; i++) {
+    fail_unless(buf[i] == (char)MAGIC_UNTOUCHED_BYTE);
+  }
+}
+
+static void test_def_itoa(int number, const char *expected)
+{
+  char buf[TEST_BUFSIZE];
+  char *test_buf = &buf[GUARD_SIZE];
+
+  size_t exp_len = strlen(expected);
+  fail_unless(exp_len + 4 < (TEST_BUFSIZE - (2 * GUARD_SIZE)));
+
+  memset(buf, MAGIC_UNTOUCHED_BYTE, sizeof(buf));
+  lwip_itoa(test_buf, exp_len + 1, number);
+  def_check_range_untouched(buf, GUARD_SIZE);
+  fail_unless(test_buf[exp_len] == 0);
+  fail_unless(!memcmp(test_buf, expected, exp_len));
+  def_check_range_untouched(&test_buf[exp_len + 1], TEST_BUFSIZE - GUARD_SIZE - exp_len - 1);
+
+  /* check with too small buffer */
+  memset(buf, MAGIC_UNTOUCHED_BYTE, sizeof(buf));
+  lwip_itoa(test_buf, exp_len, number);
+  def_check_range_untouched(buf, GUARD_SIZE);
+  def_check_range_untouched(&test_buf[exp_len + 1], TEST_BUFSIZE - GUARD_SIZE - exp_len - 1);
+
+  /* check with too large buffer */
+  memset(buf, MAGIC_UNTOUCHED_BYTE, sizeof(buf));
+  lwip_itoa(test_buf, exp_len + 4, number);
+  def_check_range_untouched(buf, GUARD_SIZE);
+  fail_unless(test_buf[exp_len] == 0);
+  fail_unless(!memcmp(test_buf, expected, exp_len));
+  def_check_range_untouched(&test_buf[exp_len + 4], TEST_BUFSIZE - GUARD_SIZE - exp_len - 4);
+}
+
+START_TEST(test_def_lwip_itoa)
+{
+  LWIP_UNUSED_ARG(_i);
+
+  test_def_itoa(0, "0");
+  test_def_itoa(1, "1");
+  test_def_itoa(-1, "-1");
+  test_def_itoa(15, "15");
+  test_def_itoa(-15, "-15");
+  test_def_itoa(156, "156");
+  test_def_itoa(1192, "1192");
+  test_def_itoa(-156, "-156");
+}
+END_TEST
+
+/** Create the suite including all tests for this module */
+Suite *
+def_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_def_lwip_itoa)
+  };
+  return create_suite("DEF", tests, sizeof(tests)/sizeof(testfunc), def_setup, def_teardown);
+}

lwip/test/unit/core/test_def.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_DEF_H
+#define LWIP_HDR_TEST_DEF_H
+
+#include "../lwip_check.h"
+
+Suite *def_suite(void);
+
+#endif

lwip/test/unit/core/test_mem.c

@@ -0,0 +1,224 @@
+#include "test_mem.h"
+
+#include "lwip/mem.h"
+#include "lwip/stats.h"
+
+#if !LWIP_STATS || !MEM_STATS
+#error "This tests needs MEM-statistics enabled"
+#endif
+#if LWIP_DNS
+#error "This test needs DNS turned off (as it mallocs on init)"
+#endif
+
+/* Setups/teardown functions */
+
+static void
+mem_setup(void)
+{
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+mem_teardown(void)
+{
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+
+/* Test functions */
+
+/** Call mem_malloc, mem_free and mem_trim and check stats */
+START_TEST(test_mem_one)
+{
+#define SIZE1   16
+#define SIZE1_2 12
+#define SIZE2   16
+  void *p1, *p2;
+  mem_size_t s1, s2;
+  LWIP_UNUSED_ARG(_i);
+
+  fail_unless(lwip_stats.mem.used == 0);
+
+  p1 = mem_malloc(SIZE1);
+  fail_unless(p1 != NULL);
+  fail_unless(lwip_stats.mem.used >= SIZE1);
+  s1 = lwip_stats.mem.used;
+
+  p2 = mem_malloc(SIZE2);
+  fail_unless(p2 != NULL);
+  fail_unless(lwip_stats.mem.used >= SIZE2 + s1);
+  s2 = lwip_stats.mem.used;
+
+  mem_trim(p1, SIZE1_2);
+
+  mem_free(p2);
+  fail_unless(lwip_stats.mem.used <= s2 - SIZE2);
+
+  mem_free(p1);
+  fail_unless(lwip_stats.mem.used == 0);
+}
+END_TEST
+
+static void malloc_keep_x(int x, int num, int size, int freestep)
+{
+   int i;
+   void* p[16];
+   LWIP_ASSERT("invalid size", size >= 0 && size < (mem_size_t)-1);
+   memset(p, 0, sizeof(p));
+   for(i = 0; i < num && i < 16; i++) {
+      p[i] = mem_malloc((mem_size_t)size);
+      fail_unless(p[i] != NULL);
+   }
+   for(i = 0; i < num && i < 16; i += freestep) {
+      if (i == x) {
+         continue;
+      }
+      mem_free(p[i]);
+      p[i] = NULL;
+   }
+   for(i = 0; i < num && i < 16; i++) {
+      if (i == x) {
+         continue;
+      }
+      if (p[i] != NULL) {
+         mem_free(p[i]);
+         p[i] = NULL;
+      }
+   }
+   fail_unless(p[x] != NULL);
+   mem_free(p[x]);
+}
+
+START_TEST(test_mem_random)
+{
+  const int num = 16;
+  int x;
+  int size;
+  int freestep;
+  LWIP_UNUSED_ARG(_i);
+
+  fail_unless(lwip_stats.mem.used == 0);
+
+  for (x = 0; x < num; x++) {
+     for (size = 1; size < 32; size++) {
+        for (freestep = 1; freestep <= 3; freestep++) {
+          fail_unless(lwip_stats.mem.used == 0);
+          malloc_keep_x(x, num, size, freestep);
+          fail_unless(lwip_stats.mem.used == 0);
+        }
+     }
+  }
+}
+END_TEST
+
+START_TEST(test_mem_invalid_free)
+{
+  u8_t *ptr, *ptr_low, *ptr_high;
+  LWIP_UNUSED_ARG(_i);
+
+  fail_unless(lwip_stats.mem.used == 0);
+  fail_unless(lwip_stats.mem.illegal == 0);
+
+  ptr = (u8_t *)mem_malloc(1);
+  fail_unless(ptr != NULL);
+  fail_unless(lwip_stats.mem.used != 0);
+
+  ptr_low = ptr - 0x10;
+  mem_free(ptr_low);
+  fail_unless(lwip_stats.mem.illegal == 1);
+  lwip_stats.mem.illegal = 0;
+
+  ptr_high = ptr + (MEM_SIZE * 2);
+  mem_free(ptr_high);
+  fail_unless(lwip_stats.mem.illegal == 1);
+  lwip_stats.mem.illegal = 0;
+
+  mem_free(ptr);
+  fail_unless(lwip_stats.mem.illegal == 0);
+  fail_unless(lwip_stats.mem.used == 0);
+}
+END_TEST
+
+START_TEST(test_mem_double_free)
+{
+  u8_t *ptr1b, *ptr1, *ptr2, *ptr3;
+  LWIP_UNUSED_ARG(_i);
+
+  fail_unless(lwip_stats.mem.used == 0);
+  fail_unless(lwip_stats.mem.illegal == 0);
+
+  ptr1 = (u8_t *)mem_malloc(1);
+  fail_unless(ptr1 != NULL);
+  fail_unless(lwip_stats.mem.used != 0);
+
+  ptr2 = (u8_t *)mem_malloc(1);
+  fail_unless(ptr2 != NULL);
+  fail_unless(lwip_stats.mem.used != 0);
+
+  ptr3 = (u8_t *)mem_malloc(1);
+  fail_unless(ptr3 != NULL);
+  fail_unless(lwip_stats.mem.used != 0);
+
+  /* free the middle mem */
+  mem_free(ptr2);
+  fail_unless(lwip_stats.mem.illegal == 0);
+
+  /* double-free of middle mem: should fail */
+  mem_free(ptr2);
+  fail_unless(lwip_stats.mem.illegal == 1);
+  lwip_stats.mem.illegal = 0;
+
+  /* free upper memory and try again */
+  mem_free(ptr3);
+  fail_unless(lwip_stats.mem.illegal == 0);
+
+  mem_free(ptr2);
+  fail_unless(lwip_stats.mem.illegal == 1);
+  lwip_stats.mem.illegal = 0;
+
+  /* free lower memory and try again */
+  mem_free(ptr1);
+  fail_unless(lwip_stats.mem.illegal == 0);
+  fail_unless(lwip_stats.mem.used == 0);
+
+  mem_free(ptr2);
+  fail_unless(lwip_stats.mem.illegal == 1);
+  fail_unless(lwip_stats.mem.used == 0);
+  lwip_stats.mem.illegal = 0;
+
+  /* reallocate lowest memory, now overlapping already freed ptr2 */
+#ifndef MIN_SIZE
+#define MIN_SIZE 12
+#endif
+  ptr1b = (u8_t *)mem_malloc(MIN_SIZE * 2);
+  fail_unless(ptr1b != NULL);
+  fail_unless(lwip_stats.mem.used != 0);
+
+  mem_free(ptr2);
+  fail_unless(lwip_stats.mem.illegal == 1);
+  lwip_stats.mem.illegal = 0;
+
+  memset(ptr1b, 1, MIN_SIZE * 2);
+
+  mem_free(ptr2);
+  fail_unless(lwip_stats.mem.illegal == 1);
+  lwip_stats.mem.illegal = 0;
+
+  mem_free(ptr1b);
+  fail_unless(lwip_stats.mem.illegal == 0);
+  fail_unless(lwip_stats.mem.used == 0);
+}
+END_TEST
+
+/** Create the suite including all tests for this module */
+Suite *
+mem_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_mem_one),
+    TESTFUNC(test_mem_random),
+    TESTFUNC(test_mem_invalid_free),
+    TESTFUNC(test_mem_double_free)
+  };
+  return create_suite("MEM", tests, sizeof(tests)/sizeof(testfunc), mem_setup, mem_teardown);
+}

lwip/test/unit/core/test_mem.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_MEM_H
+#define LWIP_HDR_TEST_MEM_H
+
+#include "../lwip_check.h"
+
+Suite *mem_suite(void);
+
+#endif

lwip/test/unit/core/test_netif.c

@@ -0,0 +1,227 @@
+#include "test_netif.h"
+
+#include "lwip/netif.h"
+#include "lwip/stats.h"
+#include "lwip/etharp.h"
+#include "netif/ethernet.h"
+
+#if !LWIP_NETIF_EXT_STATUS_CALLBACK
+#error "This tests needs LWIP_NETIF_EXT_STATUS_CALLBACK enabled"
+#endif
+
+struct netif net_test;
+
+
+/* Setups/teardown functions */
+
+static void
+netif_setup(void)
+{
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+netif_teardown(void)
+{
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+/* test helper functions */
+
+static err_t
+testif_tx_func(struct netif *netif, struct pbuf *p)
+{
+  LWIP_UNUSED_ARG(netif);
+  LWIP_UNUSED_ARG(p);
+  return ERR_OK;
+}
+
+static err_t
+testif_init(struct netif *netif)
+{
+  netif->name[0] = 'c';
+  netif->name[1] = 'h';
+  netif->output = etharp_output;
+  netif->linkoutput = testif_tx_func;
+  netif->mtu = 1500;
+  netif->hwaddr_len = 6;
+  netif->flags = NETIF_FLAG_BROADCAST | NETIF_FLAG_ETHARP | NETIF_FLAG_ETHERNET | NETIF_FLAG_IGMP | NETIF_FLAG_MLD6;
+
+  netif->hwaddr[0] = 0x02;
+  netif->hwaddr[1] = 0x03;
+  netif->hwaddr[2] = 0x04;
+  netif->hwaddr[3] = 0x05;
+  netif->hwaddr[4] = 0x06;
+  netif->hwaddr[5] = 0x07;
+
+  return ERR_OK;
+}
+
+#define MAX_NSC_REASON_IDX 10
+static netif_nsc_reason_t expected_reasons;
+static int callback_ctr;
+
+static int dummy_active;
+
+static void
+test_netif_ext_callback_dummy(struct netif* netif, netif_nsc_reason_t reason, const netif_ext_callback_args_t* args)
+{
+  LWIP_UNUSED_ARG(netif);
+  LWIP_UNUSED_ARG(reason);
+  LWIP_UNUSED_ARG(args);
+
+  fail_unless(dummy_active);
+}
+
+static void
+test_netif_ext_callback(struct netif* netif, netif_nsc_reason_t reason, const netif_ext_callback_args_t* args)
+{
+  LWIP_UNUSED_ARG(args); /* @todo */
+  callback_ctr++;
+
+  fail_unless(netif == &net_test);
+
+  fail_unless(expected_reasons == reason);
+}
+
+/* Test functions */
+
+NETIF_DECLARE_EXT_CALLBACK(netif_callback_1)
+NETIF_DECLARE_EXT_CALLBACK(netif_callback_2)
+NETIF_DECLARE_EXT_CALLBACK(netif_callback_3)
+
+START_TEST(test_netif_extcallbacks)
+{
+  ip4_addr_t addr;
+  ip4_addr_t netmask;
+  ip4_addr_t gw;
+  LWIP_UNUSED_ARG(_i);
+
+  IP4_ADDR(&addr, 0, 0, 0, 0);
+  IP4_ADDR(&netmask, 0, 0, 0, 0);
+  IP4_ADDR(&gw, 0, 0, 0, 0);
+
+  netif_add_ext_callback(&netif_callback_3, test_netif_ext_callback_dummy);
+  netif_add_ext_callback(&netif_callback_2, test_netif_ext_callback);
+  netif_add_ext_callback(&netif_callback_1, test_netif_ext_callback_dummy);
+
+  dummy_active = 1;
+
+  /* positive tests: check that single events come as expected */
+
+  expected_reasons = LWIP_NSC_NETIF_ADDED;
+  callback_ctr = 0;
+  netif_add(&net_test, &addr, &netmask, &gw, &net_test, testif_init, ethernet_input);
+  fail_unless(callback_ctr == 1);
+
+  expected_reasons = LWIP_NSC_LINK_CHANGED;
+  callback_ctr = 0;
+  netif_set_link_up(&net_test);
+  fail_unless(callback_ctr == 1);
+
+  expected_reasons = LWIP_NSC_STATUS_CHANGED;
+  callback_ctr = 0;
+  netif_set_up(&net_test);
+  fail_unless(callback_ctr == 1);
+
+  IP4_ADDR(&addr, 1, 2, 3, 4);
+  expected_reasons = LWIP_NSC_IPV4_ADDRESS_CHANGED;
+  callback_ctr = 0;
+  netif_set_ipaddr(&net_test, &addr);
+  fail_unless(callback_ctr == 1);
+
+  IP4_ADDR(&netmask, 255, 255, 255, 0);
+  expected_reasons = LWIP_NSC_IPV4_NETMASK_CHANGED;
+  callback_ctr = 0;
+  netif_set_netmask(&net_test, &netmask);
+  fail_unless(callback_ctr == 1);
+
+  IP4_ADDR(&gw, 1, 2, 3, 254);
+  expected_reasons = LWIP_NSC_IPV4_GATEWAY_CHANGED;
+  callback_ctr = 0;
+  netif_set_gw(&net_test, &gw);
+  fail_unless(callback_ctr == 1);
+
+  IP4_ADDR(&addr, 0, 0, 0, 0);
+  expected_reasons = LWIP_NSC_IPV4_ADDRESS_CHANGED;
+  callback_ctr = 0;
+  netif_set_ipaddr(&net_test, &addr);
+  fail_unless(callback_ctr == 1);
+
+  IP4_ADDR(&netmask, 0, 0, 0, 0);
+  expected_reasons = LWIP_NSC_IPV4_NETMASK_CHANGED;
+  callback_ctr = 0;
+  netif_set_netmask(&net_test, &netmask);
+  fail_unless(callback_ctr == 1);
+
+  IP4_ADDR(&gw, 0, 0, 0, 0);
+  expected_reasons = LWIP_NSC_IPV4_GATEWAY_CHANGED;
+  callback_ctr = 0;
+  netif_set_gw(&net_test, &gw);
+  fail_unless(callback_ctr == 1);
+
+  /* check for multi-events (only one combined callback expected) */
+
+  IP4_ADDR(&addr, 1, 2, 3, 4);
+  IP4_ADDR(&netmask, 255, 255, 255, 0);
+  IP4_ADDR(&gw, 1, 2, 3, 254);
+  expected_reasons = (netif_nsc_reason_t)(LWIP_NSC_IPV4_ADDRESS_CHANGED | LWIP_NSC_IPV4_NETMASK_CHANGED |
+                                          LWIP_NSC_IPV4_GATEWAY_CHANGED | LWIP_NSC_IPV4_SETTINGS_CHANGED);
+  callback_ctr = 0;
+  netif_set_addr(&net_test, &addr, &netmask, &gw);
+  fail_unless(callback_ctr == 1);
+
+  /* check that for no-change, no callback is expected */
+  expected_reasons = LWIP_NSC_NONE;
+  callback_ctr = 0;
+  netif_set_ipaddr(&net_test, &addr);
+  fail_unless(callback_ctr == 0);
+
+  netif_set_netmask(&net_test, &netmask);
+  callback_ctr = 0;
+  fail_unless(callback_ctr == 0);
+
+  callback_ctr = 0;
+  netif_set_gw(&net_test, &gw);
+  fail_unless(callback_ctr == 0);
+
+  callback_ctr = 0;
+  netif_set_addr(&net_test, &addr, &netmask, &gw);
+  fail_unless(callback_ctr == 0);
+
+  /* check for single-events */
+  IP4_ADDR(&addr, 1, 2, 3, 5);
+  expected_reasons = (netif_nsc_reason_t)(LWIP_NSC_IPV4_ADDRESS_CHANGED | LWIP_NSC_IPV4_SETTINGS_CHANGED);
+  callback_ctr = 0;
+  netif_set_addr(&net_test, &addr, &netmask, &gw);
+  fail_unless(callback_ctr == 1);
+
+  expected_reasons = LWIP_NSC_STATUS_CHANGED;
+  callback_ctr = 0;
+  netif_set_down(&net_test);
+  fail_unless(callback_ctr == 1);
+
+  expected_reasons = LWIP_NSC_NETIF_REMOVED;
+  callback_ctr = 0;
+  netif_remove(&net_test);
+  fail_unless(callback_ctr == 1);
+
+  expected_reasons = LWIP_NSC_NONE;
+
+  netif_remove_ext_callback(&netif_callback_2);
+  netif_remove_ext_callback(&netif_callback_3);
+  netif_remove_ext_callback(&netif_callback_1);
+  dummy_active = 0;
+}
+END_TEST
+
+
+/** Create the suite including all tests for this module */
+Suite *
+netif_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_netif_extcallbacks)
+  };
+  return create_suite("NETIF", tests, sizeof(tests)/sizeof(testfunc), netif_setup, netif_teardown);
+}

lwip/test/unit/core/test_netif.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_NETIF_H
+#define LWIP_HDR_TEST_NETIF_H
+
+#include "../lwip_check.h"
+
+Suite *netif_suite(void);
+
+#endif

lwip/test/unit/core/test_pbuf.c

@@ -0,0 +1,271 @@
+#include "test_pbuf.h"
+
+#include "lwip/pbuf.h"
+#include "lwip/stats.h"
+
+#if !LWIP_STATS || !MEM_STATS ||!MEMP_STATS
+#error "This tests needs MEM- and MEMP-statistics enabled"
+#endif
+#if LWIP_DNS
+#error "This test needs DNS turned off (as it mallocs on init)"
+#endif
+#if !LWIP_TCP || !TCP_QUEUE_OOSEQ || !LWIP_WND_SCALE
+#error "This test needs TCP OOSEQ queueing and window scaling enabled"
+#endif
+
+/* Setups/teardown functions */
+
+static void
+pbuf_setup(void)
+{
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+pbuf_teardown(void)
+{
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+
+#define TESTBUFSIZE_1 65535
+#define TESTBUFSIZE_2 65530
+#define TESTBUFSIZE_3 50050
+static u8_t testbuf_1[TESTBUFSIZE_1];
+static u8_t testbuf_1a[TESTBUFSIZE_1];
+static u8_t testbuf_2[TESTBUFSIZE_2];
+static u8_t testbuf_2a[TESTBUFSIZE_2];
+static u8_t testbuf_3[TESTBUFSIZE_3];
+static u8_t testbuf_3a[TESTBUFSIZE_3];
+
+/* Test functions */
+START_TEST(test_pbuf_alloc_zero_pbufs)
+{
+  struct pbuf *p;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, 0, PBUF_ROM);
+  fail_unless(p != NULL);
+  if (p != NULL) {
+    pbuf_free(p);
+  }
+
+  p = pbuf_alloc(PBUF_RAW, 0, PBUF_RAM);
+  fail_unless(p != NULL);
+  if (p != NULL) {
+    pbuf_free(p);
+  }
+
+  p = pbuf_alloc(PBUF_RAW, 0, PBUF_REF);
+  fail_unless(p != NULL);
+  if (p != NULL) {
+    pbuf_free(p);
+  }
+
+  p = pbuf_alloc(PBUF_RAW, 0, PBUF_POOL);
+  fail_unless(p != NULL);
+  if (p != NULL) {
+    pbuf_free(p);
+  }
+}
+END_TEST
+
+/** Call pbuf_copy on a pbuf with zero length */
+START_TEST(test_pbuf_copy_zero_pbuf)
+{
+  struct pbuf *p1, *p2, *p3;
+  err_t err;
+  LWIP_UNUSED_ARG(_i);
+
+  fail_unless(lwip_stats.mem.used == 0);
+  fail_unless(MEMP_STATS_GET(used, MEMP_PBUF_POOL) == 0);
+
+  p1 = pbuf_alloc(PBUF_RAW, 1024, PBUF_RAM);
+  fail_unless(p1 != NULL);
+  fail_unless(p1->ref == 1);
+
+  p2 = pbuf_alloc(PBUF_RAW, 2, PBUF_POOL);
+  fail_unless(p2 != NULL);
+  fail_unless(p2->ref == 1);
+  p2->len = p2->tot_len = 0;
+
+  pbuf_cat(p1, p2);
+  fail_unless(p1->ref == 1);
+  fail_unless(p2->ref == 1);
+
+  p3 = pbuf_alloc(PBUF_RAW, p1->tot_len, PBUF_POOL);
+  err = pbuf_copy(p3, p1);
+  fail_unless(err == ERR_VAL);
+
+  pbuf_free(p1);
+  pbuf_free(p3);
+  fail_unless(lwip_stats.mem.used == 0);
+
+  fail_unless(lwip_stats.mem.used == 0);
+  fail_unless(MEMP_STATS_GET(used, MEMP_PBUF_POOL) == 0);
+}
+END_TEST
+
+START_TEST(test_pbuf_split_64k_on_small_pbufs)
+{
+  struct pbuf *p, *rest=NULL;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, 1, PBUF_POOL);
+  pbuf_split_64k(p, &rest);
+  fail_unless(p->tot_len == 1);
+  pbuf_free(p);
+}
+END_TEST
+
+START_TEST(test_pbuf_queueing_bigger_than_64k)
+{
+  int i;
+  err_t err;
+  struct pbuf *p1, *p2, *p3, *rest2=NULL, *rest3=NULL;
+  LWIP_UNUSED_ARG(_i);
+
+  for(i = 0; i < TESTBUFSIZE_1; i++) {
+    testbuf_1[i] = (u8_t)rand();
+  }
+  for(i = 0; i < TESTBUFSIZE_2; i++) {
+    testbuf_2[i] = (u8_t)rand();
+  }
+  for(i = 0; i < TESTBUFSIZE_3; i++) {
+    testbuf_3[i] = (u8_t)rand();
+  }
+
+  p1 = pbuf_alloc(PBUF_RAW, TESTBUFSIZE_1, PBUF_POOL);
+  fail_unless(p1 != NULL);
+  p2 = pbuf_alloc(PBUF_RAW, TESTBUFSIZE_2, PBUF_POOL);
+  fail_unless(p2 != NULL);
+  p3 = pbuf_alloc(PBUF_RAW, TESTBUFSIZE_3, PBUF_POOL);
+  fail_unless(p3 != NULL);
+  err = pbuf_take(p1, testbuf_1, TESTBUFSIZE_1);
+  fail_unless(err == ERR_OK);
+  err = pbuf_take(p2, testbuf_2, TESTBUFSIZE_2);
+  fail_unless(err == ERR_OK);
+  err = pbuf_take(p3, testbuf_3, TESTBUFSIZE_3);
+  fail_unless(err == ERR_OK);
+
+  pbuf_cat(p1, p2);
+  pbuf_cat(p1, p3);
+
+  pbuf_split_64k(p1, &rest2);
+  fail_unless(p1->tot_len == TESTBUFSIZE_1);
+  fail_unless(rest2->tot_len == (u16_t)((TESTBUFSIZE_2+TESTBUFSIZE_3) & 0xFFFF));
+  pbuf_split_64k(rest2, &rest3);
+  fail_unless(rest2->tot_len == TESTBUFSIZE_2);
+  fail_unless(rest3->tot_len == TESTBUFSIZE_3);
+
+  pbuf_copy_partial(p1, testbuf_1a, TESTBUFSIZE_1, 0);
+  pbuf_copy_partial(rest2, testbuf_2a, TESTBUFSIZE_2, 0);
+  pbuf_copy_partial(rest3, testbuf_3a, TESTBUFSIZE_3, 0);
+  fail_if(memcmp(testbuf_1, testbuf_1a, TESTBUFSIZE_1));
+  fail_if(memcmp(testbuf_2, testbuf_2a, TESTBUFSIZE_2));
+  fail_if(memcmp(testbuf_3, testbuf_3a, TESTBUFSIZE_3));
+
+  pbuf_free(p1);
+  pbuf_free(rest2);
+  pbuf_free(rest3);
+}
+END_TEST
+
+/* Test for bug that writing with pbuf_take_at() did nothing
+ * and returned ERR_OK when writing at beginning of a pbuf
+ * in the chain.
+ */
+START_TEST(test_pbuf_take_at_edge)
+{
+  err_t res;
+  u8_t *out;
+  int i;
+  u8_t testdata[] = { 0x01, 0x08, 0x82, 0x02 };
+  struct pbuf *p = pbuf_alloc(PBUF_RAW, 1024, PBUF_POOL);
+  struct pbuf *q = p->next;
+  LWIP_UNUSED_ARG(_i);
+  /* alloc big enough to get a chain of pbufs */
+  fail_if(p->tot_len == p->len);
+  memset(p->payload, 0, p->len);
+  memset(q->payload, 0, q->len);
+
+  /* copy data to the beginning of first pbuf */
+  res = pbuf_take_at(p, &testdata, sizeof(testdata), 0);
+  fail_unless(res == ERR_OK);
+
+  out = (u8_t*)p->payload;
+  for (i = 0; i < (int)sizeof(testdata); i++) {
+    fail_unless(out[i] == testdata[i],
+      "Bad data at pos %d, was %02X, expected %02X", i, out[i], testdata[i]);
+  }
+
+  /* copy data to the just before end of first pbuf */
+  res = pbuf_take_at(p, &testdata, sizeof(testdata), p->len - 1);
+  fail_unless(res == ERR_OK);
+
+  out = (u8_t*)p->payload;
+  fail_unless(out[p->len - 1] == testdata[0],
+    "Bad data at pos %d, was %02X, expected %02X", p->len - 1, out[p->len - 1], testdata[0]);
+  out = (u8_t*)q->payload;
+  for (i = 1; i < (int)sizeof(testdata); i++) {
+    fail_unless(out[i-1] == testdata[i],
+      "Bad data at pos %d, was %02X, expected %02X", p->len - 1 + i, out[i-1], testdata[i]);
+  }
+
+  /* copy data to the beginning of second pbuf */
+  res = pbuf_take_at(p, &testdata, sizeof(testdata), p->len);
+  fail_unless(res == ERR_OK);
+
+  out = (u8_t*)p->payload;
+  for (i = 0; i < (int)sizeof(testdata); i++) {
+    fail_unless(out[i] == testdata[i],
+      "Bad data at pos %d, was %02X, expected %02X", p->len+i, out[i], testdata[i]);
+  }
+  pbuf_free(p);
+}
+END_TEST
+
+/* Verify pbuf_put_at()/pbuf_get_at() when using
+ * offsets equal to beginning of new pbuf in chain
+ */
+START_TEST(test_pbuf_get_put_at_edge)
+{
+  u8_t *out;
+  u8_t testdata = 0x01;
+  u8_t getdata;
+  struct pbuf *p = pbuf_alloc(PBUF_RAW, 1024, PBUF_POOL);
+  struct pbuf *q = p->next;
+  LWIP_UNUSED_ARG(_i);
+  /* alloc big enough to get a chain of pbufs */
+  fail_if(p->tot_len == p->len);
+  memset(p->payload, 0, p->len);
+  memset(q->payload, 0, q->len);
+
+  /* put byte at the beginning of second pbuf */
+  pbuf_put_at(p, p->len, testdata);
+
+  out = (u8_t*)q->payload;
+  fail_unless(*out == testdata,
+    "Bad data at pos %d, was %02X, expected %02X", p->len, *out, testdata);
+
+  getdata = pbuf_get_at(p, p->len);
+  fail_unless(*out == getdata,
+    "pbuf_get_at() returned bad data at pos %d, was %02X, expected %02X", p->len, getdata, *out);
+  pbuf_free(p);
+}
+END_TEST
+
+/** Create the suite including all tests for this module */
+Suite *
+pbuf_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_pbuf_alloc_zero_pbufs),
+    TESTFUNC(test_pbuf_copy_zero_pbuf),
+    TESTFUNC(test_pbuf_split_64k_on_small_pbufs),
+    TESTFUNC(test_pbuf_queueing_bigger_than_64k),
+    TESTFUNC(test_pbuf_take_at_edge),
+    TESTFUNC(test_pbuf_get_put_at_edge)
+  };
+  return create_suite("PBUF", tests, sizeof(tests)/sizeof(testfunc), pbuf_setup, pbuf_teardown);
+}

lwip/test/unit/core/test_pbuf.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_PBUF_H
+#define LWIP_HDR_TEST_PBUF_H
+
+#include "../lwip_check.h"
+
+Suite *pbuf_suite(void);
+
+#endif

lwip/test/unit/core/test_timers.c

@@ -0,0 +1,233 @@
+#include "test_timers.h"
+
+#include "lwip/def.h"
+#include "lwip/timeouts.h"
+#include "arch/sys_arch.h"
+
+/* Setups/teardown functions */
+
+static struct sys_timeo* old_list_head;
+
+static void
+timers_setup(void)
+{
+  struct sys_timeo** list_head = sys_timeouts_get_next_timeout();
+  old_list_head = *list_head;
+  *list_head = NULL;
+}
+
+static void
+timers_teardown(void)
+{
+  struct sys_timeo** list_head = sys_timeouts_get_next_timeout();
+  *list_head = old_list_head;
+  lwip_sys_now = 0;
+}
+
+static int fired[3];
+static void
+dummy_handler(void* arg)
+{
+  int index = LWIP_PTR_NUMERIC_CAST(int, arg);
+  fired[index] = 1;
+}
+
+#define HANDLER_EXECUTION_TIME 5
+static int cyclic_fired;
+static void
+dummy_cyclic_handler(void)
+{
+   cyclic_fired = 1;
+   lwip_sys_now += HANDLER_EXECUTION_TIME;
+}
+
+struct lwip_cyclic_timer test_cyclic = {10, dummy_cyclic_handler};
+
+static void
+do_test_cyclic_timers(u32_t offset)
+{
+  struct sys_timeo** list_head = sys_timeouts_get_next_timeout();
+
+  /* verify normal timer expiration */
+  lwip_sys_now = offset + 0;
+  sys_timeout(test_cyclic.interval_ms, lwip_cyclic_timer, &test_cyclic);
+
+  cyclic_fired = 0;
+  sys_check_timeouts();
+  fail_unless(cyclic_fired == 0);
+
+  lwip_sys_now = offset + test_cyclic.interval_ms;
+  sys_check_timeouts();
+  fail_unless(cyclic_fired == 1);
+
+  fail_unless((*list_head)->time == (u32_t)(lwip_sys_now + test_cyclic.interval_ms - HANDLER_EXECUTION_TIME));
+  
+  sys_untimeout(lwip_cyclic_timer, &test_cyclic);
+
+
+  /* verify "overload" - next cyclic timer execution is already overdue twice */
+  lwip_sys_now = offset + 0;
+  sys_timeout(test_cyclic.interval_ms, lwip_cyclic_timer, &test_cyclic);
+
+  cyclic_fired = 0;
+  sys_check_timeouts();
+  fail_unless(cyclic_fired == 0);
+
+  lwip_sys_now = offset + 2*test_cyclic.interval_ms;
+  sys_check_timeouts();
+  fail_unless(cyclic_fired == 1);
+
+  fail_unless((*list_head)->time == (u32_t)(lwip_sys_now + test_cyclic.interval_ms));
+}
+
+START_TEST(test_cyclic_timers)
+{
+  LWIP_UNUSED_ARG(_i);
+
+  /* check without u32_t wraparound */
+  do_test_cyclic_timers(0);
+
+  /* check with u32_t wraparound */
+  do_test_cyclic_timers(0xfffffff0);
+}
+END_TEST
+
+/* reproduce bug #52748: the bug in timeouts.c */
+START_TEST(test_bug52748)
+{
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&fired, 0, sizeof(fired));
+
+  lwip_sys_now = 50;
+  sys_timeout(20, dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 0));
+  sys_timeout( 5, dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 2));
+
+  lwip_sys_now = 55;
+  sys_check_timeouts();
+  fail_unless(fired[0] == 0);
+  fail_unless(fired[1] == 0);
+  fail_unless(fired[2] == 1);
+
+  lwip_sys_now = 60;
+  sys_timeout(10, dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 1));
+  sys_check_timeouts();
+  fail_unless(fired[0] == 0);
+  fail_unless(fired[1] == 0);
+  fail_unless(fired[2] == 1);
+
+  lwip_sys_now = 70;
+  sys_check_timeouts();
+  fail_unless(fired[0] == 1);
+  fail_unless(fired[1] == 1);
+  fail_unless(fired[2] == 1);
+}
+END_TEST
+
+static void
+do_test_timers(u32_t offset)
+{
+  struct sys_timeo** list_head = sys_timeouts_get_next_timeout();
+  
+  lwip_sys_now = offset + 0;
+
+  sys_timeout(10, dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 0));
+  fail_unless(sys_timeouts_sleeptime() == 10);
+  sys_timeout(20, dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 1));
+  fail_unless(sys_timeouts_sleeptime() == 10);
+  sys_timeout( 5, dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 2));
+  fail_unless(sys_timeouts_sleeptime() == 5);
+
+  /* linked list correctly sorted? */
+  fail_unless((*list_head)->time             == (u32_t)(lwip_sys_now + 5));
+  fail_unless((*list_head)->next->time       == (u32_t)(lwip_sys_now + 10));
+  fail_unless((*list_head)->next->next->time == (u32_t)(lwip_sys_now + 20));
+  
+  /* check timers expire in correct order */
+  memset(&fired, 0, sizeof(fired));
+
+  lwip_sys_now += 4;
+  sys_check_timeouts();
+  fail_unless(fired[2] == 0);
+
+  lwip_sys_now += 1;
+  sys_check_timeouts();
+  fail_unless(fired[2] == 1);
+
+  lwip_sys_now += 4;
+  sys_check_timeouts();
+  fail_unless(fired[0] == 0);
+
+  lwip_sys_now += 1;
+  sys_check_timeouts();
+  fail_unless(fired[0] == 1);
+
+  lwip_sys_now += 9;
+  sys_check_timeouts();
+  fail_unless(fired[1] == 0);
+
+  lwip_sys_now += 1;
+  sys_check_timeouts();
+  fail_unless(fired[1] == 1);
+
+  sys_untimeout(dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 0));
+  sys_untimeout(dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 1));
+  sys_untimeout(dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 2));
+}
+
+START_TEST(test_timers)
+{
+  LWIP_UNUSED_ARG(_i);
+
+  /* check without u32_t wraparound */
+  do_test_timers(0);
+
+  /* check with u32_t wraparound */
+  do_test_timers(0xfffffff0);
+}
+END_TEST
+
+START_TEST(test_long_timer)
+{
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&fired, 0, sizeof(fired));
+  lwip_sys_now = 0;
+
+  sys_timeout(LWIP_UINT32_MAX / 4, dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 0));
+  fail_unless(sys_timeouts_sleeptime() == LWIP_UINT32_MAX / 4);
+
+  sys_check_timeouts();
+  fail_unless(fired[0] == 0);
+
+  lwip_sys_now += LWIP_UINT32_MAX / 8;
+
+  sys_check_timeouts();
+  fail_unless(fired[0] == 0);
+
+  lwip_sys_now += LWIP_UINT32_MAX / 8;
+
+  sys_check_timeouts();
+  fail_unless(fired[0] == 0);
+
+  lwip_sys_now += 1;
+
+  sys_check_timeouts();
+  fail_unless(fired[0] == 1);
+
+  sys_untimeout(dummy_handler, LWIP_PTR_NUMERIC_CAST(void*, 0));
+}
+END_TEST
+
+/** Create the suite including all tests for this module */
+Suite *
+timers_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_bug52748),
+    TESTFUNC(test_cyclic_timers),
+    TESTFUNC(test_timers),
+    TESTFUNC(test_long_timer),
+  };
+  return create_suite("TIMERS", tests, LWIP_ARRAYSIZE(tests), timers_setup, timers_teardown);
+}

lwip/test/unit/core/test_timers.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_TIMERS_H
+#define LWIP_HDR_TEST_TIMERS_H
+
+#include "../lwip_check.h"
+
+Suite *timers_suite(void);
+
+#endif

lwip/test/unit/dhcp/test_dhcp.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_DHCP_H
+#define LWIP_HDR_TEST_DHCP_H
+
+#include "../lwip_check.h"
+
+Suite* dhcp_suite(void);
+
+#endif

lwip/test/unit/etharp/test_etharp.c

@@ -0,0 +1,272 @@
+#include "test_etharp.h"
+
+#include "lwip/udp.h"
+#include "lwip/etharp.h"
+#include "netif/ethernet.h"
+#include "lwip/stats.h"
+#include "lwip/prot/iana.h"
+
+#if !LWIP_STATS || !UDP_STATS || !MEMP_STATS || !ETHARP_STATS
+#error "This tests needs UDP-, MEMP- and ETHARP-statistics enabled"
+#endif
+#if !ETHARP_SUPPORT_STATIC_ENTRIES
+#error "This test needs ETHARP_SUPPORT_STATIC_ENTRIES enabled"
+#endif
+
+static struct netif test_netif;
+static ip4_addr_t test_ipaddr, test_netmask, test_gw;
+struct eth_addr test_ethaddr =  {{1,1,1,1,1,1}};
+struct eth_addr test_ethaddr2 = {{1,1,1,1,1,2}};
+struct eth_addr test_ethaddr3 = {{1,1,1,1,1,3}};
+struct eth_addr test_ethaddr4 = {{1,1,1,1,1,4}};
+static int linkoutput_ctr;
+
+/* Helper functions */
+static void
+etharp_remove_all(void)
+{
+  int i;
+  /* call etharp_tmr often enough to have all entries cleaned */
+  for(i = 0; i < 0xff; i++) {
+    etharp_tmr();
+  }
+}
+
+static err_t
+default_netif_linkoutput(struct netif *netif, struct pbuf *p)
+{
+  fail_unless(netif == &test_netif);
+  fail_unless(p != NULL);
+  linkoutput_ctr++;
+  return ERR_OK;
+}
+
+static err_t
+default_netif_init(struct netif *netif)
+{
+  fail_unless(netif != NULL);
+  netif->linkoutput = default_netif_linkoutput;
+  netif->output = etharp_output;
+  netif->mtu = 1500;
+  netif->flags = NETIF_FLAG_BROADCAST | NETIF_FLAG_ETHARP | NETIF_FLAG_LINK_UP;
+  netif->hwaddr_len = ETHARP_HWADDR_LEN;
+  return ERR_OK;
+}
+
+static void
+default_netif_add(void)
+{
+  IP4_ADDR(&test_gw, 192,168,0,1);
+  IP4_ADDR(&test_ipaddr, 192,168,0,1);
+  IP4_ADDR(&test_netmask, 255,255,0,0);
+
+  fail_unless(netif_default == NULL);
+  netif_set_default(netif_add(&test_netif, &test_ipaddr, &test_netmask,
+                              &test_gw, NULL, default_netif_init, NULL));
+  netif_set_up(&test_netif);
+}
+
+static void
+default_netif_remove(void)
+{
+  fail_unless(netif_default == &test_netif);
+  netif_remove(&test_netif);
+}
+
+static void
+create_arp_response(ip4_addr_t *adr)
+{
+  int k;
+  struct eth_hdr *ethhdr;
+  struct etharp_hdr *etharphdr;
+  struct pbuf *p = pbuf_alloc(PBUF_RAW, sizeof(struct eth_hdr) + sizeof(struct etharp_hdr), PBUF_RAM);
+  if(p == NULL) {
+    FAIL_RET();
+  }
+  ethhdr = (struct eth_hdr*)p->payload;
+  etharphdr = (struct etharp_hdr*)(ethhdr + 1);
+
+  ethhdr->dest = test_ethaddr;
+  ethhdr->src = test_ethaddr2;
+  ethhdr->type = htons(ETHTYPE_ARP);
+
+  etharphdr->hwtype = htons(LWIP_IANA_HWTYPE_ETHERNET);
+  etharphdr->proto = htons(ETHTYPE_IP);
+  etharphdr->hwlen = ETHARP_HWADDR_LEN;
+  etharphdr->protolen = sizeof(ip4_addr_t);
+  etharphdr->opcode = htons(ARP_REPLY);
+
+  SMEMCPY(&etharphdr->sipaddr, adr, sizeof(ip4_addr_t));
+  SMEMCPY(&etharphdr->dipaddr, &test_ipaddr, sizeof(ip4_addr_t));
+
+  k = 6;
+  while(k > 0) {
+    k--;
+    /* Write the ARP MAC-Addresses */
+    etharphdr->shwaddr.addr[k] = test_ethaddr2.addr[k];
+    etharphdr->dhwaddr.addr[k] = test_ethaddr.addr[k];
+    /* Write the Ethernet MAC-Addresses */
+    ethhdr->dest.addr[k] = test_ethaddr.addr[k];
+    ethhdr->src.addr[k]  = test_ethaddr2.addr[k];
+  }
+
+  ethernet_input(p, &test_netif);
+}
+
+/* Setups/teardown functions */
+
+static void
+etharp_setup(void)
+{
+  etharp_remove_all();
+  default_netif_add();
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+etharp_teardown(void)
+{
+  etharp_remove_all();
+  default_netif_remove();
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+
+/* Test functions */
+
+START_TEST(test_etharp_table)
+{
+#if ETHARP_SUPPORT_STATIC_ENTRIES
+  err_t err;
+#endif /* ETHARP_SUPPORT_STATIC_ENTRIES */
+  ssize_t idx;
+  const ip4_addr_t *unused_ipaddr;
+  struct eth_addr *unused_ethaddr;
+  struct udp_pcb* pcb;
+  LWIP_UNUSED_ARG(_i);
+
+  if (netif_default != &test_netif) {
+    fail("This test needs a default netif");
+  }
+
+  linkoutput_ctr = 0;
+
+  pcb = udp_new();
+  fail_unless(pcb != NULL);
+  if (pcb != NULL) {
+    ip4_addr_t adrs[ARP_TABLE_SIZE + 2];
+    int i;
+    for(i = 0; i < ARP_TABLE_SIZE + 2; i++) {
+      IP4_ADDR(&adrs[i], 192,168,0,i+2);
+    }
+    /* fill ARP-table with dynamic entries */
+    for(i = 0; i < ARP_TABLE_SIZE; i++) {
+      struct pbuf *p = pbuf_alloc(PBUF_TRANSPORT, 10, PBUF_RAM);
+      fail_unless(p != NULL);
+      if (p != NULL) {
+        err_t err2;
+        ip_addr_t dst;
+        ip_addr_copy_from_ip4(dst, adrs[i]);
+        err2 = udp_sendto(pcb, p, &dst, 123);
+        fail_unless(err2 == ERR_OK);
+        /* etharp request sent? */
+        fail_unless(linkoutput_ctr == (2*i) + 1);
+        pbuf_free(p);
+
+        /* create an ARP response */
+        create_arp_response(&adrs[i]);
+        /* queued UDP packet sent? */
+        fail_unless(linkoutput_ctr == (2*i) + 2);
+
+        idx = etharp_find_addr(NULL, &adrs[i], &unused_ethaddr, &unused_ipaddr);
+        fail_unless(idx == i);
+        etharp_tmr();
+      }
+    }
+    linkoutput_ctr = 0;
+#if ETHARP_SUPPORT_STATIC_ENTRIES
+    /* create one static entry */
+    err = etharp_add_static_entry(&adrs[ARP_TABLE_SIZE], &test_ethaddr3);
+    fail_unless(err == ERR_OK);
+    idx = etharp_find_addr(NULL, &adrs[ARP_TABLE_SIZE], &unused_ethaddr, &unused_ipaddr);
+    fail_unless(idx == 0);
+    fail_unless(linkoutput_ctr == 0);
+#endif /* ETHARP_SUPPORT_STATIC_ENTRIES */
+
+    linkoutput_ctr = 0;
+    /* fill ARP-table with dynamic entries */
+    for(i = 0; i < ARP_TABLE_SIZE; i++) {
+      struct pbuf *p = pbuf_alloc(PBUF_TRANSPORT, 10, PBUF_RAM);
+      fail_unless(p != NULL);
+      if (p != NULL) {
+        err_t err2;
+        ip_addr_t dst;
+        ip_addr_copy_from_ip4(dst, adrs[i]);
+        err2 = udp_sendto(pcb, p, &dst, 123);
+        fail_unless(err2 == ERR_OK);
+        /* etharp request sent? */
+        fail_unless(linkoutput_ctr == (2*i) + 1);
+        pbuf_free(p);
+
+        /* create an ARP response */
+        create_arp_response(&adrs[i]);
+        /* queued UDP packet sent? */
+        fail_unless(linkoutput_ctr == (2*i) + 2);
+
+        idx = etharp_find_addr(NULL, &adrs[i], &unused_ethaddr, &unused_ipaddr);
+        if (i < ARP_TABLE_SIZE - 1) {
+          fail_unless(idx == i+1);
+        } else {
+          /* the last entry must not overwrite the static entry! */
+          fail_unless(idx == 1);
+        }
+        etharp_tmr();
+      }
+    }
+#if ETHARP_SUPPORT_STATIC_ENTRIES
+    /* create a second static entry */
+    err = etharp_add_static_entry(&adrs[ARP_TABLE_SIZE+1], &test_ethaddr4);
+    fail_unless(err == ERR_OK);
+    idx = etharp_find_addr(NULL, &adrs[ARP_TABLE_SIZE], &unused_ethaddr, &unused_ipaddr);
+    fail_unless(idx == 0);
+    idx = etharp_find_addr(NULL, &adrs[ARP_TABLE_SIZE+1], &unused_ethaddr, &unused_ipaddr);
+    fail_unless(idx == 2);
+    /* and remove it again */
+    err = etharp_remove_static_entry(&adrs[ARP_TABLE_SIZE+1]);
+    fail_unless(err == ERR_OK);
+    idx = etharp_find_addr(NULL, &adrs[ARP_TABLE_SIZE], &unused_ethaddr, &unused_ipaddr);
+    fail_unless(idx == 0);
+    idx = etharp_find_addr(NULL, &adrs[ARP_TABLE_SIZE+1], &unused_ethaddr, &unused_ipaddr);
+    fail_unless(idx == -1);
+#endif /* ETHARP_SUPPORT_STATIC_ENTRIES */
+
+    /* check that static entries don't time out */
+    etharp_remove_all();
+    idx = etharp_find_addr(NULL, &adrs[ARP_TABLE_SIZE], &unused_ethaddr, &unused_ipaddr);
+    fail_unless(idx == 0);
+
+#if ETHARP_SUPPORT_STATIC_ENTRIES
+    /* remove the first static entry */
+    err = etharp_remove_static_entry(&adrs[ARP_TABLE_SIZE]);
+    fail_unless(err == ERR_OK);
+    idx = etharp_find_addr(NULL, &adrs[ARP_TABLE_SIZE], &unused_ethaddr, &unused_ipaddr);
+    fail_unless(idx == -1);
+    idx = etharp_find_addr(NULL, &adrs[ARP_TABLE_SIZE+1], &unused_ethaddr, &unused_ipaddr);
+    fail_unless(idx == -1);
+#endif /* ETHARP_SUPPORT_STATIC_ENTRIES */
+
+    udp_remove(pcb);
+  }
+}
+END_TEST
+
+
+/** Create the suite including all tests for this module */
+Suite *
+etharp_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_etharp_table)
+  };
+  return create_suite("ETHARP", tests, sizeof(tests)/sizeof(testfunc), etharp_setup, etharp_teardown);
+}

lwip/test/unit/etharp/test_etharp.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_ETHARP_H
+#define LWIP_HDR_TEST_ETHARP_H
+
+#include "../lwip_check.h"
+
+Suite* etharp_suite(void);
+
+#endif

lwip/test/unit/ip4/test_ip4.c

@@ -0,0 +1,160 @@
+#include "test_ip4.h"
+
+#include "lwip/ip4.h"
+#include "lwip/inet_chksum.h"
+#include "lwip/stats.h"
+#include "lwip/prot/ip.h"
+#include "lwip/prot/ip4.h"
+
+#include "lwip/tcpip.h"
+
+#if !LWIP_IPV4 || !IP_REASSEMBLY || !MIB2_STATS || !IPFRAG_STATS
+#error "This tests needs LWIP_IPV4, IP_REASSEMBLY; MIB2- and IPFRAG-statistics enabled"
+#endif
+
+/* Helper functions */
+static void
+create_ip4_input_fragment(u16_t ip_id, u16_t start, u16_t len, int last)
+{
+  struct pbuf *p;
+  struct netif *input_netif = netif_list; /* just use any netif */
+  fail_unless((start & 7) == 0);
+  fail_unless(((len & 7) == 0) || last);
+  fail_unless(input_netif != NULL);
+
+  p = pbuf_alloc(PBUF_RAW, len + sizeof(struct ip_hdr), PBUF_RAM);
+  fail_unless(p != NULL);
+  if (p != NULL) {
+    err_t err;
+    struct ip_hdr *iphdr = (struct ip_hdr *)p->payload;
+    IPH_VHL_SET(iphdr, 4, sizeof(struct ip_hdr) / 4);
+    IPH_TOS_SET(iphdr, 0);
+    IPH_LEN_SET(iphdr, lwip_htons(p->tot_len));
+    IPH_ID_SET(iphdr, lwip_htons(ip_id));
+    if (last) {
+      IPH_OFFSET_SET(iphdr, lwip_htons(start / 8));
+    } else {
+      IPH_OFFSET_SET(iphdr, lwip_htons((start / 8) | IP_MF));
+    }
+    IPH_TTL_SET(iphdr, 5);
+    IPH_PROTO_SET(iphdr, IP_PROTO_UDP);
+    IPH_CHKSUM_SET(iphdr, 0);
+    ip4_addr_copy(iphdr->src, *netif_ip4_addr(input_netif));
+    iphdr->src.addr = lwip_htonl(lwip_htonl(iphdr->src.addr) + 1);
+    ip4_addr_copy(iphdr->dest, *netif_ip4_addr(input_netif));
+    IPH_CHKSUM_SET(iphdr, inet_chksum(iphdr, sizeof(struct ip_hdr)));
+
+    err = ip4_input(p, input_netif);
+    if (err != ERR_OK) {
+      pbuf_free(p);
+    }
+    fail_unless(err == ERR_OK);
+  }
+}
+
+/* Setups/teardown functions */
+
+static void
+ip4_setup(void)
+{
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+ip4_teardown(void)
+{
+  if (netif_list->loop_first != NULL) {
+    pbuf_free(netif_list->loop_first);
+    netif_list->loop_first = NULL;
+  }
+  netif_list->loop_last = NULL;
+  /* poll until all memory is released... */
+  tcpip_thread_poll_one();
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+
+/* Test functions */
+
+START_TEST(test_ip4_reass)
+{
+  const u16_t ip_id = 128;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&lwip_stats.mib2, 0, sizeof(lwip_stats.mib2));
+
+  create_ip4_input_fragment(ip_id, 8*200, 200, 1);
+  fail_unless(lwip_stats.ip_frag.recv == 1);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 0);
+
+  create_ip4_input_fragment(ip_id, 0*200, 200, 0);
+  fail_unless(lwip_stats.ip_frag.recv == 2);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 0);
+
+  create_ip4_input_fragment(ip_id, 1*200, 200, 0);
+  fail_unless(lwip_stats.ip_frag.recv == 3);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 0);
+
+  create_ip4_input_fragment(ip_id, 2*200, 200, 0);
+  fail_unless(lwip_stats.ip_frag.recv == 4);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 0);
+
+  create_ip4_input_fragment(ip_id, 3*200, 200, 0);
+  fail_unless(lwip_stats.ip_frag.recv == 5);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 0);
+
+  create_ip4_input_fragment(ip_id, 4*200, 200, 0);
+  fail_unless(lwip_stats.ip_frag.recv == 6);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 0);
+
+  create_ip4_input_fragment(ip_id, 7*200, 200, 0);
+  fail_unless(lwip_stats.ip_frag.recv == 7);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 0);
+
+  create_ip4_input_fragment(ip_id, 6*200, 200, 0);
+  fail_unless(lwip_stats.ip_frag.recv == 8);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 0);
+
+  create_ip4_input_fragment(ip_id, 5*200, 200, 0);
+  fail_unless(lwip_stats.ip_frag.recv == 9);
+  fail_unless(lwip_stats.ip_frag.err == 0);
+  fail_unless(lwip_stats.ip_frag.memerr == 0);
+  fail_unless(lwip_stats.ip_frag.drop == 0);
+  fail_unless(lwip_stats.mib2.ipreasmoks == 1);
+}
+END_TEST
+
+
+/** Create the suite including all tests for this module */
+Suite *
+ip4_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_ip4_reass),
+  };
+  return create_suite("IPv4", tests, sizeof(tests)/sizeof(testfunc), ip4_setup, ip4_teardown);
+}

lwip/test/unit/ip4/test_ip4.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_IP4_H
+#define LWIP_HDR_TEST_IP4_H
+
+#include "../lwip_check.h"
+
+Suite* ip4_suite(void);
+
+#endif

lwip/test/unit/ip6/test_ip6.c

@@ -0,0 +1,321 @@
+#include "test_ip6.h"
+
+#include "lwip/ethip6.h"
+#include "lwip/ip6.h"
+#include "lwip/inet_chksum.h"
+#include "lwip/nd6.h"
+#include "lwip/stats.h"
+#include "lwip/prot/ethernet.h"
+#include "lwip/prot/ip.h"
+#include "lwip/prot/ip6.h"
+
+#include "lwip/tcpip.h"
+
+#if LWIP_IPV6 /* allow to build the unit tests without IPv6 support */
+
+static struct netif test_netif6;
+static int linkoutput_ctr;
+
+static err_t
+default_netif_linkoutput(struct netif *netif, struct pbuf *p)
+{
+  fail_unless(netif == &test_netif6);
+  fail_unless(p != NULL);
+  linkoutput_ctr++;
+  return ERR_OK;
+}
+
+static err_t
+default_netif_init(struct netif *netif)
+{
+  fail_unless(netif != NULL);
+  netif->linkoutput = default_netif_linkoutput;
+  netif->output_ip6 = ethip6_output;
+  netif->mtu = 1500;
+  netif->flags = NETIF_FLAG_BROADCAST | NETIF_FLAG_ETHERNET | NETIF_FLAG_MLD6;
+  netif->hwaddr_len = ETH_HWADDR_LEN;
+  return ERR_OK;
+}
+
+static void
+default_netif_add(void)
+{
+  struct netif *n;
+  fail_unless(netif_default == NULL);
+  n = netif_add_noaddr(&test_netif6, NULL, default_netif_init, NULL);
+  fail_unless(n == &test_netif6);
+  netif_set_default(&test_netif6);
+}
+
+static void
+default_netif_remove(void)
+{
+  fail_unless(netif_default == &test_netif6);
+  netif_remove(&test_netif6);
+}
+
+static void
+ip6_test_handle_timers(int count)
+{
+  int i;
+  for (i = 0; i < count; i++) {
+    nd6_tmr();
+  }
+}
+
+/* Setups/teardown functions */
+
+static void
+ip6_setup(void)
+{
+  default_netif_add();
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+ip6_teardown(void)
+{
+  if (netif_list->loop_first != NULL) {
+    pbuf_free(netif_list->loop_first);
+    netif_list->loop_first = NULL;
+  }
+  netif_list->loop_last = NULL;
+  /* poll until all memory is released... */
+  tcpip_thread_poll_one();
+  default_netif_remove();
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+
+/* Test functions */
+
+static void
+test_ip6_ll_addr_iter(int expected_ctr1, int expected_ctr2)
+{
+  fail_unless(linkoutput_ctr == 0);
+
+  /* test that nothing is sent with link uo but netif down */
+  netif_set_link_up(&test_netif6);
+  ip6_test_handle_timers(500);
+  fail_unless(linkoutput_ctr == 0);
+  netif_set_link_down(&test_netif6);
+  fail_unless(linkoutput_ctr == 0);
+
+  /* test that nothing is sent with link down but netif up */
+  netif_set_up(&test_netif6);
+  ip6_test_handle_timers(500);
+  fail_unless(linkoutput_ctr == 0);
+  netif_set_down(&test_netif6);
+  fail_unless(linkoutput_ctr == 0);
+
+  /* test what is sent with link up + netif up */
+  netif_set_link_up(&test_netif6);
+  netif_set_up(&test_netif6);
+  ip6_test_handle_timers(500);
+  fail_unless(linkoutput_ctr == expected_ctr1);
+  netif_set_down(&test_netif6);
+  netif_set_link_down(&test_netif6);
+  fail_unless(linkoutput_ctr == expected_ctr1);
+  linkoutput_ctr = 0;
+
+  netif_set_up(&test_netif6);
+  netif_set_link_up(&test_netif6);
+  ip6_test_handle_timers(500);
+  fail_unless(linkoutput_ctr == expected_ctr2);
+  netif_set_link_down(&test_netif6);
+  netif_set_down(&test_netif6);
+  fail_unless(linkoutput_ctr == expected_ctr2);
+  linkoutput_ctr = 0;
+}
+
+START_TEST(test_ip6_ll_addr)
+{
+  LWIP_UNUSED_ARG(_i);
+
+  /* test without link-local address */
+  test_ip6_ll_addr_iter(0, 0);
+
+  /* test with link-local address */
+  netif_create_ip6_linklocal_address(&test_netif6, 1);
+  test_ip6_ll_addr_iter(3 + LWIP_IPV6_DUP_DETECT_ATTEMPTS + LWIP_IPV6_MLD, 3);
+}
+END_TEST
+
+START_TEST(test_ip6_aton_ipv4mapped)
+{
+  int ret;
+  ip_addr_t addr;
+  ip6_addr_t addr6;
+  const ip_addr_t addr_expected = IPADDR6_INIT_HOST(0, 0, 0xFFFF, 0xD4CC65D2);
+  const char *full_ipv6_addr = "0:0:0:0:0:FFFF:D4CC:65D2";
+  const char *shortened_ipv6_addr = "::FFFF:D4CC:65D2";
+  const char *full_ipv4_mapped_addr = "0:0:0:0:0:FFFF:212.204.101.210";
+  const char *shortened_ipv4_mapped_addr = "::FFFF:212.204.101.210";
+  const char *bogus_ipv4_mapped_addr = "::FFFF:212.204.101.2101";
+  LWIP_UNUSED_ARG(_i);
+
+  /* check IPv6 representation */
+  memset(&addr6, 0, sizeof(addr6));
+  ret = ip6addr_aton(full_ipv6_addr, &addr6);
+  fail_unless(ret == 1);
+  fail_unless(memcmp(&addr6, &addr_expected, 16) == 0);
+  memset(&addr, 0, sizeof(addr));
+  ret = ipaddr_aton(full_ipv6_addr, &addr);
+  fail_unless(ret == 1);
+  fail_unless(memcmp(&addr, &addr_expected, 16) == 0);
+
+  /* check shortened IPv6 representation */
+  memset(&addr6, 0, sizeof(addr6));
+  ret = ip6addr_aton(shortened_ipv6_addr, &addr6);
+  fail_unless(ret == 1);
+  fail_unless(memcmp(&addr6, &addr_expected, 16) == 0);
+  memset(&addr, 0, sizeof(addr));
+  ret = ipaddr_aton(shortened_ipv6_addr, &addr);
+  fail_unless(ret == 1);
+  fail_unless(memcmp(&addr, &addr_expected, 16) == 0);
+
+  /* checked shortened mixed representation */
+  memset(&addr6, 0, sizeof(addr6));
+  ret = ip6addr_aton(shortened_ipv4_mapped_addr, &addr6);
+  fail_unless(ret == 1);
+  fail_unless(memcmp(&addr6, &addr_expected, 16) == 0);
+  memset(&addr, 0, sizeof(addr));
+  ret = ipaddr_aton(shortened_ipv4_mapped_addr, &addr);
+  fail_unless(ret == 1);
+  fail_unless(memcmp(&addr, &addr_expected, 16) == 0);
+
+  /* checked mixed representation */
+  memset(&addr6, 0, sizeof(addr6));
+  ret = ip6addr_aton(full_ipv4_mapped_addr, &addr6);
+  fail_unless(ret == 1);
+  fail_unless(memcmp(&addr6, &addr_expected, 16) == 0);
+  memset(&addr, 0, sizeof(addr));
+  ret = ipaddr_aton(full_ipv4_mapped_addr, &addr);
+  fail_unless(ret == 1);
+  fail_unless(memcmp(&addr, &addr_expected, 16) == 0);
+
+  /* checked bogus mixed representation */
+  memset(&addr6, 0, sizeof(addr6));
+  ret = ip6addr_aton(bogus_ipv4_mapped_addr, &addr6);
+  fail_unless(ret == 0);
+  memset(&addr, 0, sizeof(addr));
+  ret = ipaddr_aton(bogus_ipv4_mapped_addr, &addr);
+  fail_unless(ret == 0);
+}
+END_TEST
+
+START_TEST(test_ip6_ntoa_ipv4mapped)
+{
+  const ip_addr_t addr = IPADDR6_INIT_HOST(0, 0, 0xFFFF, 0xD4CC65D2);
+  char buf[128];
+  char *str;
+  LWIP_UNUSED_ARG(_i);
+
+  str = ip6addr_ntoa_r(ip_2_ip6(&addr), buf, sizeof(buf));
+  fail_unless(str == buf);
+  fail_unless(!strcmp(str, "::FFFF:212.204.101.210"));
+}
+END_TEST
+
+struct test_addr_and_str {
+  ip_addr_t addr;
+  const char *str;
+};
+
+START_TEST(test_ip6_ntoa)
+{
+  struct test_addr_and_str tests[] = {
+    {IPADDR6_INIT_HOST(0xfe800000, 0x00000000, 0xb2a1a2ff, 0xfea3a4a5), "FE80::B2A1:A2FF:FEA3:A4A5"}, /* test shortened zeros */
+    {IPADDR6_INIT_HOST(0xfe800000, 0xff000000, 0xb2a1a2ff, 0xfea3a4a5), "FE80:0:FF00:0:B2A1:A2FF:FEA3:A4A5"}, /* don't omit single zero blocks */
+    {IPADDR6_INIT_HOST(0xfe800000, 0xff000000, 0xb2000000, 0x0000a4a5), "FE80:0:FF00:0:B200::A4A5"}, /* omit longest zero block */
+  };
+  char buf[128];
+  char *str;
+  size_t i;
+  LWIP_UNUSED_ARG(_i);
+
+  for (i = 0; i < LWIP_ARRAYSIZE(tests); i++) {
+    str = ip6addr_ntoa_r(ip_2_ip6(&tests[i].addr), buf, sizeof(buf));
+    fail_unless(str == buf);
+    fail_unless(!strcmp(str, tests[i].str));
+  }
+}
+END_TEST
+
+START_TEST(test_ip6_lladdr)
+{
+  u8_t zeros[128];
+  const u8_t test_mac_addr[6] = {0xb0, 0xa1, 0xa2, 0xa3, 0xa4, 0xa5};
+  const u32_t expected_ip6_addr_1[4] = {PP_HTONL(0xfe800000), 0, PP_HTONL(0xb2a1a2ff), PP_HTONL(0xfea3a4a5)};
+  const u32_t expected_ip6_addr_2[4] = {PP_HTONL(0xfe800000), 0, PP_HTONL(0x0000b0a1), PP_HTONL(0xa2a3a4a5)};
+  LWIP_UNUSED_ARG(_i);
+  memset(zeros, 0, sizeof(zeros));
+
+  fail_unless(test_netif6.hwaddr_len == 6);
+  fail_unless(!memcmp(test_netif6.hwaddr, zeros, 6));
+
+  fail_unless(test_netif6.ip6_addr_state[0] == 0);
+  fail_unless(!memcmp(netif_ip6_addr(&test_netif6, 0), zeros, sizeof(ip6_addr_t)));
+
+  /* set specific mac addr */
+  memcpy(test_netif6.hwaddr, test_mac_addr, 6);
+
+  /* create link-local addr based on mac (EUI-48) */
+  netif_create_ip6_linklocal_address(&test_netif6, 1);
+  fail_unless(IP_IS_V6(&test_netif6.ip6_addr[0]));
+  fail_unless(!memcmp(&netif_ip6_addr(&test_netif6, 0)->addr, expected_ip6_addr_1, 16));
+#if LWIP_IPV6_SCOPES
+  fail_unless(netif_ip6_addr(&test_netif6, 0)->zone == (test_netif6.num + 1));
+#endif
+  /* reset address */
+  memset(&test_netif6.ip6_addr[0], 0, sizeof(ip6_addr_t));
+  test_netif6.ip6_addr_state[0] = 0;
+
+  /* create link-local addr based interface ID */
+  netif_create_ip6_linklocal_address(&test_netif6, 0);
+  fail_unless(IP_IS_V6(&test_netif6.ip6_addr[0]));
+  fail_unless(!memcmp(&netif_ip6_addr(&test_netif6, 0)->addr, expected_ip6_addr_2, 16));
+#if LWIP_IPV6_SCOPES
+  fail_unless(netif_ip6_addr(&test_netif6, 0)->zone == (test_netif6.num + 1));
+#endif
+  /* reset address */
+  memset(&test_netif6.ip6_addr[0], 0, sizeof(ip6_addr_t));
+  test_netif6.ip6_addr_state[0] = 0;
+
+  /* reset mac address */
+  memset(&test_netif6.hwaddr, 0, sizeof(test_netif6.hwaddr));
+}
+END_TEST
+
+/** Create the suite including all tests for this module */
+Suite *
+ip6_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_ip6_ll_addr),
+    TESTFUNC(test_ip6_aton_ipv4mapped),
+    TESTFUNC(test_ip6_ntoa_ipv4mapped),
+    TESTFUNC(test_ip6_ntoa),
+    TESTFUNC(test_ip6_lladdr)
+  };
+  return create_suite("IPv6", tests, sizeof(tests)/sizeof(testfunc), ip6_setup, ip6_teardown);
+}
+
+#else /* LWIP_IPV6 */
+
+/* allow to build the unit tests without IPv6 support */
+START_TEST(test_ip6_dummy)
+{
+  LWIP_UNUSED_ARG(_i);
+}
+END_TEST
+
+Suite *
+ip6_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_ip6_dummy),
+  };
+  return create_suite("IPv6", tests, sizeof(tests)/sizeof(testfunc), NULL, NULL);
+}
+#endif /* LWIP_IPV6 */

lwip/test/unit/ip6/test_ip6.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_IP6_H
+#define LWIP_HDR_TEST_IP6_H
+
+#include "../lwip_check.h"
+
+Suite* ip6_suite(void);
+
+#endif

lwip/test/unit/lwip_check.h

@@ -0,0 +1,42 @@
+#ifndef LWIP_HDR_LWIP_CHECK_H
+#define LWIP_HDR_LWIP_CHECK_H
+
+/* Common header file for lwIP unit tests using the check framework */
+
+#include <config.h>
+#include <check.h>
+#include <stdlib.h>
+
+#define FAIL_RET() do { fail(); return; } while(0)
+#define EXPECT(x) fail_unless(x)
+#define EXPECT_RET(x) do { fail_unless(x); if(!(x)) { return; }} while(0)
+#define EXPECT_RETX(x, y) do { fail_unless(x); if(!(x)) { return y; }} while(0)
+#define EXPECT_RETNULL(x) EXPECT_RETX(x, NULL)
+
+typedef struct {
+  TFun func;
+  const char *name;
+} testfunc;
+
+#define TESTFUNC(x) {(x), "" # x "" }
+
+/* Modified function from check.h, supplying function name */
+#define tcase_add_named_test(tc,tf) \
+   _tcase_add_test((tc),(tf).func,(tf).name,0, 0, 0, 1)
+
+/** typedef for a function returning a test suite */
+typedef Suite* (suite_getter_fn)(void);
+
+/** Create a test suite */
+Suite* create_suite(const char* name, testfunc *tests, size_t num_tests, SFun setup, SFun teardown);
+
+#ifdef LWIP_UNITTESTS_LIB
+int lwip_unittests_run(void)
+#endif
+
+/* helper functions */
+#define SKIP_POOL(x) (1 << x)
+#define SKIP_HEAP    (1 << MEMP_MAX)
+void lwip_check_ensure_no_alloc(unsigned int skip);
+
+#endif /* LWIP_HDR_LWIP_CHECK_H */

lwip/test/unit/lwip_unittests.c

@@ -0,0 +1,107 @@
+#include "lwip_check.h"
+
+#include "ip4/test_ip4.h"
+#include "ip6/test_ip6.h"
+#include "udp/test_udp.h"
+#include "tcp/test_tcp.h"
+#include "tcp/test_tcp_oos.h"
+#include "core/test_def.h"
+#include "core/test_mem.h"
+#include "core/test_netif.h"
+#include "core/test_pbuf.h"
+#include "core/test_timers.h"
+#include "etharp/test_etharp.h"
+#include "dhcp/test_dhcp.h"
+#include "mdns/test_mdns.h"
+#include "mqtt/test_mqtt.h"
+#include "api/test_sockets.h"
+
+#include "lwip/init.h"
+#if !NO_SYS
+#include "lwip/tcpip.h"
+#endif
+
+Suite* create_suite(const char* name, testfunc *tests, size_t num_tests, SFun setup, SFun teardown)
+{
+  size_t i;
+  Suite *s = suite_create(name);
+
+  for(i = 0; i < num_tests; i++) {
+    TCase *tc_core = tcase_create(name);
+    if ((setup != NULL) || (teardown != NULL)) {
+      tcase_add_checked_fixture(tc_core, setup, teardown);
+    }
+    tcase_add_named_test(tc_core, tests[i]);
+    suite_add_tcase(s, tc_core);
+  }
+  return s;
+}
+
+void lwip_check_ensure_no_alloc(unsigned int skip)
+{
+  int i;
+  unsigned int mask;
+
+  if (!(skip & SKIP_HEAP)) {
+    fail_unless(lwip_stats.mem.used == 0);
+  }
+  for (i = 0, mask = 1; i < MEMP_MAX; i++, mask <<= 1) {
+    if (!(skip & mask)) {
+      fail_unless(lwip_stats.memp[i]->used == 0);
+    }
+  }
+}
+
+#ifdef LWIP_UNITTESTS_LIB
+int lwip_unittests_run(void)
+#else
+int main(void)
+#endif
+{
+  int number_failed;
+  SRunner *sr;
+  size_t i;
+  suite_getter_fn* suites[] = {
+    ip4_suite,
+    ip6_suite,
+    udp_suite,
+    tcp_suite,
+    tcp_oos_suite,
+    def_suite,
+    mem_suite,
+    netif_suite,
+    pbuf_suite,
+    timers_suite,
+    etharp_suite,
+    dhcp_suite,
+    mdns_suite,
+    mqtt_suite,
+    sockets_suite
+  };
+  size_t num = sizeof(suites)/sizeof(void*);
+  LWIP_ASSERT("No suites defined", num > 0);
+
+#if NO_SYS
+  lwip_init();
+#else
+  tcpip_init(NULL, NULL);
+#endif
+
+  sr = srunner_create((suites[0])());
+  srunner_set_xml(sr, "lwip_unittests.xml");
+  for(i = 1; i < num; i++) {
+    srunner_add_suite(sr, ((suite_getter_fn*)suites[i])());
+  }
+
+#ifdef LWIP_UNITTESTS_NOFORK
+  srunner_set_fork_status(sr, CK_NOFORK);
+#endif
+#ifdef LWIP_UNITTESTS_FORK
+  srunner_set_fork_status(sr, CK_FORK);
+#endif
+
+  srunner_run_all(sr, CK_NORMAL);
+  number_failed = srunner_ntests_failed(sr);
+  srunner_free(sr);
+  return (number_failed == 0) ? EXIT_SUCCESS : EXIT_FAILURE;
+}

lwip/test/unit/lwipopts.h

@@ -0,0 +1,85 @@
+/*
+ * Copyright (c) 2001-2003 Swedish Institute of Computer Science.
+ * All rights reserved. 
+ * 
+ * Redistribution and use in source and binary forms, with or without modification, 
+ * are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission. 
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED 
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT 
+ * SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT 
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS 
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 
+ * OF SUCH DAMAGE.
+ *
+ * This file is part of the lwIP TCP/IP stack.
+ * 
+ * Author: Simon Goldschmidt
+ *
+ */
+#ifndef LWIP_HDR_LWIPOPTS_H
+#define LWIP_HDR_LWIPOPTS_H
+
+#define LWIP_TESTMODE                   1
+
+#define LWIP_IPV6                       1
+
+#define LWIP_CHECKSUM_ON_COPY           1
+#define TCP_CHECKSUM_ON_COPY_SANITY_CHECK 1
+#define TCP_CHECKSUM_ON_COPY_SANITY_CHECK_FAIL(printfmsg) LWIP_ASSERT("TCP_CHECKSUM_ON_COPY_SANITY_CHECK_FAIL", 0)
+
+/* We link to special sys_arch.c (for basic non-waiting API layers unit tests) */
+#define NO_SYS                          0
+#define SYS_LIGHTWEIGHT_PROT            0
+#define LWIP_NETCONN                    !NO_SYS
+#define LWIP_SOCKET                     !NO_SYS
+#define LWIP_NETCONN_FULLDUPLEX         LWIP_SOCKET
+#define LWIP_NETBUF_RECVINFO            1
+#define LWIP_HAVE_LOOPIF                1
+#define TCPIP_THREAD_TEST
+
+/* Enable DHCP to test it, disable UDP checksum to easier inject packets */
+#define LWIP_DHCP                       1
+
+/* Minimal changes to opt.h required for tcp unit tests: */
+#define MEM_SIZE                        16000
+#define TCP_SND_QUEUELEN                40
+#define MEMP_NUM_TCP_SEG                TCP_SND_QUEUELEN
+#define TCP_SND_BUF                     (12 * TCP_MSS)
+#define TCP_WND                         (10 * TCP_MSS)
+#define LWIP_WND_SCALE                  1
+#define TCP_RCV_SCALE                   0
+#define PBUF_POOL_SIZE                  400 /* pbuf tests need ~200KByte */
+
+/* Enable IGMP and MDNS for MDNS tests */
+#define LWIP_IGMP                       1
+#define LWIP_MDNS_RESPONDER             1
+#define LWIP_NUM_NETIF_CLIENT_DATA      (LWIP_MDNS_RESPONDER)
+
+/* Minimal changes to opt.h required for etharp unit tests: */
+#define ETHARP_SUPPORT_STATIC_ENTRIES   1
+
+#define MEMP_NUM_SYS_TIMEOUT            (LWIP_NUM_SYS_TIMEOUT_INTERNAL + 8)
+
+/* MIB2 stats are required to check IPv4 reassembly results */
+#define MIB2_STATS                      1
+
+/* netif tests want to test this, so enable: */
+#define LWIP_NETIF_EXT_STATUS_CALLBACK  1
+
+/* Check lwip_stats.mem.illegal instead of asserting */
+#define LWIP_MEM_ILLEGAL_FREE(msg)      /* to nothing */
+
+#endif /* LWIP_HDR_LWIPOPTS_H */

lwip/test/unit/mdns/test_mdns.c

@@ -0,0 +1,915 @@
+/*
+ * Copyright (c) 2015 Verisure Innovation AB
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without modification,
+ * are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ * 3. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT
+ * SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+ * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT
+ * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
+ * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY
+ * OF SUCH DAMAGE.
+ *
+ * This file is part of the lwIP TCP/IP stack.
+ *
+ * Author: Erik Ekman <erik@kryo.se>
+ *
+ */
+
+#include "test_mdns.h"
+
+#include "lwip/pbuf.h"
+#include "lwip/apps/mdns.h"
+#include "lwip/apps/mdns_priv.h"
+
+START_TEST(readname_basic)
+{
+  static const u8_t data[] = { 0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0x00 };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0, &domain);
+  pbuf_free(p);
+  fail_unless(offset == sizeof(data));
+  fail_unless(domain.length == sizeof(data));
+  fail_if(memcmp(&domain.name, data, sizeof(data)));
+}
+END_TEST
+
+START_TEST(readname_anydata)
+{
+  static const u8_t data[] = { 0x05, 0x00, 0xFF, 0x08, 0xc0, 0x0f, 0x04, 0x7f, 0x80, 0x82, 0x88, 0x00 };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0, &domain);
+  pbuf_free(p);
+  fail_unless(offset == sizeof(data));
+  fail_unless(domain.length == sizeof(data));
+  fail_if(memcmp(&domain.name, data, sizeof(data)));
+}
+END_TEST
+
+START_TEST(readname_short_buf)
+{
+  static const u8_t data[] = { 0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a' };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0, &domain);
+  pbuf_free(p);
+  fail_unless(offset == MDNS_READNAME_ERROR);
+}
+END_TEST
+
+START_TEST(readname_long_label)
+{
+  static const u8_t data[] = {
+      0x05, 'm', 'u', 'l', 't', 'i',
+      0x52, 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a',
+      'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a',
+      'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a',
+      'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a',
+      'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a',
+      'a', 'a', 'a', 'a', 'a', 'a', 'a', 'a', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0, &domain);
+  pbuf_free(p);
+  fail_unless(offset == MDNS_READNAME_ERROR);
+}
+END_TEST
+
+START_TEST(readname_overflow)
+{
+  static const u8_t data[] = {
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0, &domain);
+  pbuf_free(p);
+  fail_unless(offset == MDNS_READNAME_ERROR);
+}
+END_TEST
+
+START_TEST(readname_jump_earlier)
+{
+  static const u8_t data[] = {
+      /* Some padding needed, not supported to jump to bytes containing dns header */
+      /*  0 */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+      /* 10 */ 0x0f, 0x0e, 0x05, 'l', 'o', 'c', 'a', 'l', 0x00, 0xab,
+      /* 20 */ 0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0xc0, 0x0c
+  };
+  static const u8_t fullname[] = {
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0x05, 'l', 'o', 'c', 'a', 'l', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 20, &domain);
+  pbuf_free(p);
+  fail_unless(offset == sizeof(data));
+  fail_unless(domain.length == sizeof(fullname));
+
+  fail_if(memcmp(&domain.name, fullname, sizeof(fullname)));
+}
+END_TEST
+
+START_TEST(readname_jump_earlier_jump)
+{
+  static const u8_t data[] = {
+      /* Some padding needed, not supported to jump to bytes containing dns header */
+      /* 0x00 */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+      /* 0x08 */ 0x00, 0x00, 0x00, 0x00, 0x03, 0x0b, 0x0a, 0xf2,
+      /* 0x10 */ 0x04, 'c', 'a', 's', 't', 0x00, 0xc0, 0x10,
+      /* 0x18 */ 0x05, 'm', 'u', 'l', 't', 'i', 0xc0, 0x16
+  };
+  static const u8_t fullname[] = {
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0x18, &domain);
+  pbuf_free(p);
+  fail_unless(offset == sizeof(data));
+  fail_unless(domain.length == sizeof(fullname));
+
+  fail_if(memcmp(&domain.name, fullname, sizeof(fullname)));
+}
+END_TEST
+
+START_TEST(readname_jump_maxdepth)
+{
+  static const u8_t data[] = {
+      /* Some padding needed, not supported to jump to bytes containing dns header */
+      /* 0x00 */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+      /* 0x08 */ 0x00, 0x00, 0x00, 0x00, 0x03, 0x0b, 0x0a, 0xf2,
+      /* 0x10 */ 0x04, 'n', 'a', 'm', 'e', 0xc0, 0x27, 0x03,
+      /* 0x18 */ 0x03, 'd', 'n', 's', 0xc0, 0x10, 0xc0, 0x10,
+      /* 0x20 */ 0x04, 'd', 'e', 'e', 'p', 0xc0, 0x18, 0x00,
+      /* 0x28 */ 0x04, 'c', 'a', 's', 't', 0xc0, 0x20, 0xb0,
+      /* 0x30 */ 0x05, 'm', 'u', 'l', 't', 'i', 0xc0, 0x28
+  };
+  static const u8_t fullname[] = {
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't',
+      0x04, 'd', 'e', 'e', 'p', 0x03, 'd', 'n', 's',
+      0x04, 'n', 'a', 'm', 'e', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0x30, &domain);
+  pbuf_free(p);
+  fail_unless(offset == sizeof(data));
+  fail_unless(domain.length == sizeof(fullname));
+
+  fail_if(memcmp(&domain.name, fullname, sizeof(fullname)));
+}
+END_TEST
+
+START_TEST(readname_jump_later)
+{
+  static const u8_t data[] = {
+      /* 0x00 */ 0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0xc0, 0x10, 0x00, 0x01, 0x40,
+      /* 0x10 */ 0x05, 'l', 'o', 'c', 'a', 'l', 0x00, 0xab
+  };
+  static const u8_t fullname[] = {
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0x05, 'l', 'o', 'c', 'a', 'l', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0, &domain);
+  pbuf_free(p);
+  fail_unless(offset == 13);
+  fail_unless(domain.length == sizeof(fullname));
+
+  fail_if(memcmp(&domain.name, fullname, sizeof(fullname)));
+}
+END_TEST
+
+START_TEST(readname_half_jump)
+{
+  static const u8_t data[] = {
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0xc0
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0, &domain);
+  pbuf_free(p);
+  fail_unless(offset == MDNS_READNAME_ERROR);
+}
+END_TEST
+
+START_TEST(readname_jump_toolong)
+{
+  static const u8_t data[] = {
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0xc2, 0x10, 0x00, 0x01, 0x40
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 0, &domain);
+  pbuf_free(p);
+  fail_unless(offset == MDNS_READNAME_ERROR);
+}
+END_TEST
+
+START_TEST(readname_jump_loop_label)
+{
+  static const u8_t data[] = {
+      /*  0 */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+      /* 10 */ 0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0xc0, 0x10
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 10, &domain);
+  pbuf_free(p);
+  fail_unless(offset == MDNS_READNAME_ERROR);
+}
+END_TEST
+
+START_TEST(readname_jump_loop_jump)
+{
+  static const u8_t data[] = {
+      /*  0 */ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+      /* 10 */ 0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0xc0, 0x15
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+  offset = mdns_readname(p, 10, &domain);
+  pbuf_free(p);
+  fail_unless(offset == MDNS_READNAME_ERROR);
+}
+END_TEST
+
+START_TEST(add_label_basic)
+{
+  static const u8_t data[] = { 0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0x00 };
+  struct mdns_domain domain;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "cast", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == sizeof(data));
+  fail_if(memcmp(&domain.name, data, sizeof(data)));
+}
+END_TEST
+
+START_TEST(add_label_long_label)
+{
+  static const char *toolong = "abcdefghijklmnopqrstuvwxyz0123456789-abcdefghijklmnopqrstuvwxyz0123456789-abcdefghijklmnopqrstuvwxyz0123456789-";
+  struct mdns_domain domain;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, toolong, (u8_t)strlen(toolong));
+  fail_unless(res == ERR_VAL);
+}
+END_TEST
+
+START_TEST(add_label_full)
+{
+  static const char *label = "0123456789abcdef0123456789abcdef";
+  struct mdns_domain domain;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, label, (u8_t)strlen(label));
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 33);
+  res = mdns_domain_add_label(&domain, label, (u8_t)strlen(label));
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 66);
+  res = mdns_domain_add_label(&domain, label, (u8_t)strlen(label));
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 99);
+  res = mdns_domain_add_label(&domain, label, (u8_t)strlen(label));
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 132);
+  res = mdns_domain_add_label(&domain, label, (u8_t)strlen(label));
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 165);
+  res = mdns_domain_add_label(&domain, label, (u8_t)strlen(label));
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 198);
+  res = mdns_domain_add_label(&domain, label, (u8_t)strlen(label));
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 231);
+  res = mdns_domain_add_label(&domain, label, (u8_t)strlen(label));
+  fail_unless(res == ERR_VAL);
+  fail_unless(domain.length == 231);
+  res = mdns_domain_add_label(&domain, label, 25);
+  fail_unless(res == ERR_VAL);
+  fail_unless(domain.length == 231);
+  res = mdns_domain_add_label(&domain, label, 24);
+  fail_unless(res == ERR_VAL);
+  fail_unless(domain.length == 231);
+  res = mdns_domain_add_label(&domain, label, 23);
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 255);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+  fail_unless(domain.length == 256);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_VAL);
+  fail_unless(domain.length == 256);
+}
+END_TEST
+
+START_TEST(domain_eq_basic)
+{
+  static const u8_t data[] = {
+      0x05, 'm', 'u', 'l', 't', 'i', 0x04, 'c', 'a', 's', 't', 0x00
+  };
+  struct mdns_domain domain1, domain2;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&domain1, 0, sizeof(domain1));
+  res = mdns_domain_add_label(&domain1, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, "cast", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, NULL, 0);
+  fail_unless(res == ERR_OK);
+  fail_unless(domain1.length == sizeof(data));
+
+  memset(&domain2, 0, sizeof(domain2));
+  res = mdns_domain_add_label(&domain2, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, "cast", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  fail_unless(mdns_domain_eq(&domain1, &domain2));
+}
+END_TEST
+
+START_TEST(domain_eq_diff)
+{
+  struct mdns_domain domain1, domain2;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&domain1, 0, sizeof(domain1));
+  res = mdns_domain_add_label(&domain1, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, "base", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  memset(&domain2, 0, sizeof(domain2));
+  res = mdns_domain_add_label(&domain2, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, "cast", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  fail_if(mdns_domain_eq(&domain1, &domain2));
+}
+END_TEST
+
+START_TEST(domain_eq_case)
+{
+  struct mdns_domain domain1, domain2;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&domain1, 0, sizeof(domain1));
+  res = mdns_domain_add_label(&domain1, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, "cast", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  memset(&domain2, 0, sizeof(domain2));
+  res = mdns_domain_add_label(&domain2, "MulTI", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, "casT", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  fail_unless(mdns_domain_eq(&domain1, &domain2));
+}
+END_TEST
+
+START_TEST(domain_eq_anydata)
+{
+  static const u8_t data1[] = { 0x05, 0xcc, 0xdc, 0x00, 0xa0 };
+  static const u8_t data2[] = { 0x7f, 0x8c, 0x01, 0xff, 0xcf };
+  struct mdns_domain domain1, domain2;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&domain1, 0, sizeof(domain1));
+  res = mdns_domain_add_label(&domain1, (const char*)data1, sizeof(data1));
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, "cast", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, (const char*)data2, sizeof(data2));
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  memset(&domain2, 0, sizeof(domain2));
+  res = mdns_domain_add_label(&domain2, (const char*)data1, sizeof(data1));
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, "casT", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, (const char*)data2, sizeof(data2));
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  fail_unless(mdns_domain_eq(&domain1, &domain2));
+}
+END_TEST
+
+START_TEST(domain_eq_length)
+{
+  struct mdns_domain domain1, domain2;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  memset(&domain1, 0, sizeof(domain1));
+  memset(domain1.name, 0xAA, sizeof(MDNS_DOMAIN_MAXLEN));
+  res = mdns_domain_add_label(&domain1, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain1, "cast", 4);
+  fail_unless(res == ERR_OK);
+
+  memset(&domain2, 0, sizeof(domain2));
+  memset(domain2.name, 0xBB, sizeof(MDNS_DOMAIN_MAXLEN));
+  res = mdns_domain_add_label(&domain2, "multi", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain2, "cast", 4);
+  fail_unless(res == ERR_OK);
+
+  fail_unless(mdns_domain_eq(&domain1, &domain2));
+}
+END_TEST
+
+START_TEST(compress_full_match)
+{
+  static const u8_t data[] = {
+      0x00, 0x00,
+      0x06, 'f', 'o', 'o', 'b', 'a', 'r', 0x05, 'l', 'o', 'c', 'a', 'l', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  u16_t length;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "foobar", 6);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "local", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  offset = 2;
+  length = mdns_compress_domain(p, &offset, &domain);
+  /* Write 0 bytes, then a jump to addr 2 */
+  fail_unless(length == 0);
+  fail_unless(offset == 2);
+
+  pbuf_free(p);
+}
+END_TEST
+
+START_TEST(compress_full_match_subset)
+{
+  static const u8_t data[] = {
+      0x00, 0x00,
+      0x02, 'g', 'o', 0x06, 'f', 'o', 'o', 'b', 'a', 'r', 0x05, 'l', 'o', 'c', 'a', 'l', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  u16_t length;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "foobar", 6);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "local", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  offset = 2;
+  length = mdns_compress_domain(p, &offset, &domain);
+  /* Write 0 bytes, then a jump to addr 5 */
+  fail_unless(length == 0);
+  fail_unless(offset == 5);
+
+  pbuf_free(p);
+}
+END_TEST
+
+START_TEST(compress_full_match_jump)
+{
+  static const u8_t data[] = {
+    /* 0x00 */ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
+               0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
+    /* 0x10 */ 0x04, 'l', 'w', 'i', 'p', 0x05, 'l', 'o', 'c', 'a', 'l', 0x00, 0xc0, 0x00, 0x02, 0x00,
+    /* 0x20 */ 0x06, 'f', 'o', 'o', 'b', 'a', 'r', 0xc0, 0x15
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  u16_t length;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "foobar", 6);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "local", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  offset = 0x20;
+  length = mdns_compress_domain(p, &offset, &domain);
+  /* Write 0 bytes, then a jump to addr 0x20 */
+  fail_unless(length == 0);
+  fail_unless(offset == 0x20);
+
+  pbuf_free(p);
+}
+END_TEST
+
+START_TEST(compress_no_match)
+{
+  static const u8_t data[] = {
+      0x00, 0x00,
+      0x04, 'l', 'w', 'i', 'p', 0x05, 'w', 'i', 'k', 'i', 'a', 0x03, 'c', 'o', 'm', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  u16_t length;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "foobar", 6);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "local", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  offset = 2;
+  length = mdns_compress_domain(p, &offset, &domain);
+  /* Write all bytes, no jump */
+  fail_unless(length == domain.length);
+
+  pbuf_free(p);
+}
+END_TEST
+
+START_TEST(compress_2nd_label)
+{
+  static const u8_t data[] = {
+      0x00, 0x00,
+      0x06, 'f', 'o', 'o', 'b', 'a', 'r', 0x05, 'l', 'o', 'c', 'a', 'l', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  u16_t length;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "lwip", 4);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "local", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  offset = 2;
+  length = mdns_compress_domain(p, &offset, &domain);
+  /* Write 5 bytes, then a jump to addr 9 */
+  fail_unless(length == 5);
+  fail_unless(offset == 9);
+
+  pbuf_free(p);
+}
+END_TEST
+
+START_TEST(compress_2nd_label_short)
+{
+  static const u8_t data[] = {
+      0x00, 0x00,
+      0x04, 'l', 'w', 'i', 'p', 0x05, 'l', 'o', 'c', 'a', 'l', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  u16_t length;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "foobar", 6);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "local", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  offset = 2;
+  length = mdns_compress_domain(p, &offset, &domain);
+  /* Write 5 bytes, then a jump to addr 7 */
+  fail_unless(length == 7);
+  fail_unless(offset == 7);
+
+  pbuf_free(p);
+}
+END_TEST
+
+START_TEST(compress_jump_to_jump)
+{
+  static const u8_t data[] = {
+      /* 0x00 */ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
+                 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
+      /* 0x10 */ 0x04, 'l', 'w', 'i', 'p', 0x05, 'l', 'o', 'c', 'a', 'l', 0x00, 0xc0, 0x00, 0x02, 0x00,
+      /* 0x20 */ 0x07, 'b', 'a', 'n', 'a', 'n', 'a', 's', 0xc0, 0x15
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  u16_t length;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "foobar", 6);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "local", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  offset = 0x20;
+  length = mdns_compress_domain(p, &offset, &domain);
+  /* Dont compress if jump would be to a jump */
+  fail_unless(length == domain.length);
+
+  offset = 0x10;
+  length = mdns_compress_domain(p, &offset, &domain);
+  /* Write 7 bytes, then a jump to addr 0x15 */
+  fail_unless(length == 7);
+  fail_unless(offset == 0x15);
+
+  pbuf_free(p);
+}
+END_TEST
+
+START_TEST(compress_long_match)
+{
+  static const u8_t data[] = {
+      0x00, 0x00,
+      0x06, 'f', 'o', 'o', 'b', 'a', 'r', 0x05, 'l', 'o', 'c', 'a', 'l', 0x03, 'c', 'o', 'm', 0x00
+  };
+  struct pbuf *p;
+  struct mdns_domain domain;
+  u16_t offset;
+  u16_t length;
+  err_t res;
+  LWIP_UNUSED_ARG(_i);
+
+  p = pbuf_alloc(PBUF_RAW, sizeof(data), PBUF_ROM);
+  fail_if(p == NULL);
+  p->payload = (void *)(size_t)data;
+
+  memset(&domain, 0, sizeof(domain));
+  res = mdns_domain_add_label(&domain, "foobar", 6);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, "local", 5);
+  fail_unless(res == ERR_OK);
+  res = mdns_domain_add_label(&domain, NULL, 0);
+  fail_unless(res == ERR_OK);
+
+  offset = 2;
+  length = mdns_compress_domain(p, &offset, &domain);
+  fail_unless(length == domain.length);
+
+  pbuf_free(p);
+}
+END_TEST
+
+Suite* mdns_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(readname_basic),
+    TESTFUNC(readname_anydata),
+    TESTFUNC(readname_short_buf),
+    TESTFUNC(readname_long_label),
+    TESTFUNC(readname_overflow),
+    TESTFUNC(readname_jump_earlier),
+    TESTFUNC(readname_jump_earlier_jump),
+    TESTFUNC(readname_jump_maxdepth),
+    TESTFUNC(readname_jump_later),
+    TESTFUNC(readname_half_jump),
+    TESTFUNC(readname_jump_toolong),
+    TESTFUNC(readname_jump_loop_label),
+    TESTFUNC(readname_jump_loop_jump),
+
+    TESTFUNC(add_label_basic),
+    TESTFUNC(add_label_long_label),
+    TESTFUNC(add_label_full),
+
+    TESTFUNC(domain_eq_basic),
+    TESTFUNC(domain_eq_diff),
+    TESTFUNC(domain_eq_case),
+    TESTFUNC(domain_eq_anydata),
+    TESTFUNC(domain_eq_length),
+
+    TESTFUNC(compress_full_match),
+    TESTFUNC(compress_full_match_subset),
+    TESTFUNC(compress_full_match_jump),
+    TESTFUNC(compress_no_match),
+    TESTFUNC(compress_2nd_label),
+    TESTFUNC(compress_2nd_label_short),
+    TESTFUNC(compress_jump_to_jump),
+    TESTFUNC(compress_long_match),
+  };
+  return create_suite("MDNS", tests, sizeof(tests)/sizeof(testfunc), NULL, NULL);
+}

lwip/test/unit/mdns/test_mdns.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_MDNS_H__
+#define LWIP_HDR_TEST_MDNS_H__
+
+#include "../lwip_check.h"
+
+Suite* mdns_suite(void);
+
+#endif

lwip/test/unit/mqtt/test_mqtt.c

@@ -0,0 +1,115 @@
+#include "test_mqtt.h"
+
+#include "lwip/pbuf.h"
+#include "lwip/apps/mqtt.h"
+#include "lwip/apps/mqtt_priv.h"
+#include "lwip/netif.h"
+
+const ip_addr_t test_mqtt_local_ip = IPADDR4_INIT_BYTES(192, 168, 1, 1);
+const ip_addr_t test_mqtt_remote_ip = IPADDR4_INIT_BYTES(192, 168, 1, 2);
+const ip_addr_t test_mqtt_netmask = IPADDR4_INIT_BYTES(255, 255, 255, 0);
+
+static err_t test_mqtt_netif_output(struct netif *netif, struct pbuf *p,
+       const ip4_addr_t *ipaddr)
+{
+  LWIP_UNUSED_ARG(netif);
+  LWIP_UNUSED_ARG(ipaddr);
+  LWIP_UNUSED_ARG(p);
+  return ERR_OK;
+}
+
+static void
+test_mqtt_init_netif(struct netif *netif, const ip_addr_t *ip_addr, const ip_addr_t *netmask)
+{
+  struct netif *n;
+  memset(netif, 0, sizeof(struct netif));
+  netif->output = test_mqtt_netif_output;
+  netif->flags |= NETIF_FLAG_UP | NETIF_FLAG_LINK_UP;
+  ip_addr_copy_from_ip4(netif->netmask, *ip_2_ip4(netmask));
+  ip_addr_copy_from_ip4(netif->ip_addr, *ip_2_ip4(ip_addr));
+  for (n = netif_list; n != NULL; n = n->next) {
+    if (n == netif) {
+      return;
+    }
+  }
+  netif->next = NULL;
+  netif_list = netif;
+}
+
+/* Setups/teardown functions */
+static struct netif *old_netif_list;
+static struct netif *old_netif_default;
+
+static void
+mqtt_setup(void)
+{
+  old_netif_list = netif_list;
+  old_netif_default = netif_default;
+  netif_list = NULL;
+  netif_default = NULL;
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+mqtt_teardown(void)
+{
+  netif_list = NULL;
+  netif_default = NULL;
+  /* restore netif_list for next tests (e.g. loopif) */
+  netif_list = old_netif_list;
+  netif_default = old_netif_default;
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void test_mqtt_connection_cb(mqtt_client_t *client, void *arg, mqtt_connection_status_t status)
+{
+  LWIP_UNUSED_ARG(client);
+  LWIP_UNUSED_ARG(arg);
+  LWIP_UNUSED_ARG(status);
+}
+
+START_TEST(basic_connect)
+{
+  mqtt_client_t* client;
+  struct netif netif;
+  err_t err;
+  struct mqtt_connect_client_info_t client_info = {
+    "dumm",
+    NULL, NULL,
+    10,
+    NULL, NULL, 0, 0
+  };
+  struct pbuf *p;
+  unsigned char rxbuf[] = {0x20, 0x02, 0x00, 0x00};
+  LWIP_UNUSED_ARG(_i);
+
+  test_mqtt_init_netif(&netif, &test_mqtt_local_ip, &test_mqtt_netmask);
+
+  client = mqtt_client_new();
+  fail_unless(client != NULL);
+  err = mqtt_client_connect(client, &test_mqtt_remote_ip, 1234, test_mqtt_connection_cb, NULL, &client_info);
+  fail_unless(err == ERR_OK);
+
+  client->conn->connected(client->conn->callback_arg, client->conn, ERR_OK);
+  p = pbuf_alloc(PBUF_RAW, sizeof(rxbuf), PBUF_REF);
+  fail_unless(p != NULL);
+  p->payload = rxbuf;
+  /* since we hack the rx path, we have to hack the rx window, too: */
+  client->conn->rcv_wnd -= p->tot_len;
+  if (client->conn->recv(client->conn->callback_arg, client->conn, p, ERR_OK) != ERR_OK) {
+    pbuf_free(p);
+  }
+
+  mqtt_disconnect(client);
+  /* fixme: mqtt_client_fre() is missing... */
+  mem_free(client);
+}
+END_TEST
+
+Suite* mqtt_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(basic_connect),
+  };
+  return create_suite("MQTT", tests, sizeof(tests)/sizeof(testfunc), mqtt_setup, mqtt_teardown);
+}

lwip/test/unit/mqtt/test_mqtt.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_MQTT_H__
+#define LWIP_HDR_TEST_MQTT_H__
+
+#include "../lwip_check.h"
+
+Suite* mqtt_suite(void);
+
+#endif

lwip/test/unit/tcp/tcp_helper.c

@@ -0,0 +1,319 @@
+#include "tcp_helper.h"
+
+#include "lwip/priv/tcp_priv.h"
+#include "lwip/stats.h"
+#include "lwip/pbuf.h"
+#include "lwip/inet_chksum.h"
+#include "lwip/ip_addr.h"
+
+#if !LWIP_STATS || !TCP_STATS || !MEMP_STATS
+#error "This tests needs TCP- and MEMP-statistics enabled"
+#endif
+
+const ip_addr_t test_local_ip = IPADDR4_INIT_BYTES(192, 168, 1, 1);
+const ip_addr_t test_remote_ip = IPADDR4_INIT_BYTES(192, 168, 1, 2);
+const ip_addr_t test_netmask = IPADDR4_INIT_BYTES(255, 255, 255, 0);
+
+/** Remove all pcbs on the given list. */
+static void
+tcp_remove(struct tcp_pcb* pcb_list)
+{
+  struct tcp_pcb *pcb = pcb_list;
+  struct tcp_pcb *pcb2;
+
+  while(pcb != NULL) {
+    pcb2 = pcb;
+    pcb = pcb->next;
+    tcp_abort(pcb2);
+  }
+}
+
+/** Remove all pcbs on listen-, active- and time-wait-list (bound- isn't exported). */
+void
+tcp_remove_all(void)
+{
+  tcp_remove(tcp_listen_pcbs.pcbs);
+  tcp_remove(tcp_bound_pcbs);
+  tcp_remove(tcp_active_pcbs);
+  tcp_remove(tcp_tw_pcbs);
+  fail_unless(MEMP_STATS_GET(used, MEMP_TCP_PCB) == 0);
+  fail_unless(MEMP_STATS_GET(used, MEMP_TCP_PCB_LISTEN) == 0);
+  fail_unless(MEMP_STATS_GET(used, MEMP_TCP_SEG) == 0);
+  fail_unless(MEMP_STATS_GET(used, MEMP_PBUF_POOL) == 0);
+}
+
+/** Create a TCP segment usable for passing to tcp_input */
+static struct pbuf*
+tcp_create_segment_wnd(ip_addr_t* src_ip, ip_addr_t* dst_ip,
+                   u16_t src_port, u16_t dst_port, void* data, size_t data_len,
+                   u32_t seqno, u32_t ackno, u8_t headerflags, u16_t wnd)
+{
+  struct pbuf *p, *q;
+  struct ip_hdr* iphdr;
+  struct tcp_hdr* tcphdr;
+  u16_t pbuf_len = (u16_t)(sizeof(struct ip_hdr) + sizeof(struct tcp_hdr) + data_len);
+  LWIP_ASSERT("data_len too big", data_len <= 0xFFFF);
+
+  p = pbuf_alloc(PBUF_RAW, pbuf_len, PBUF_POOL);
+  EXPECT_RETNULL(p != NULL);
+  /* first pbuf must be big enough to hold the headers */
+  EXPECT_RETNULL(p->len >= (sizeof(struct ip_hdr) + sizeof(struct tcp_hdr)));
+  if (data_len > 0) {
+    /* first pbuf must be big enough to hold at least 1 data byte, too */
+    EXPECT_RETNULL(p->len > (sizeof(struct ip_hdr) + sizeof(struct tcp_hdr)));
+  }
+
+  for(q = p; q != NULL; q = q->next) {
+    memset(q->payload, 0, q->len);
+  }
+
+  iphdr = (struct ip_hdr*)p->payload;
+  /* fill IP header */
+  iphdr->dest.addr = ip_2_ip4(dst_ip)->addr;
+  iphdr->src.addr = ip_2_ip4(src_ip)->addr;
+  IPH_VHL_SET(iphdr, 4, IP_HLEN / 4);
+  IPH_TOS_SET(iphdr, 0);
+  IPH_LEN_SET(iphdr, htons(p->tot_len));
+  IPH_CHKSUM_SET(iphdr, inet_chksum(iphdr, IP_HLEN));
+
+  /* let p point to TCP header */
+  pbuf_header(p, -(s16_t)sizeof(struct ip_hdr));
+
+  tcphdr = (struct tcp_hdr*)p->payload;
+  tcphdr->src   = htons(src_port);
+  tcphdr->dest  = htons(dst_port);
+  tcphdr->seqno = htonl(seqno);
+  tcphdr->ackno = htonl(ackno);
+  TCPH_HDRLEN_SET(tcphdr, sizeof(struct tcp_hdr)/4);
+  TCPH_FLAGS_SET(tcphdr, headerflags);
+  tcphdr->wnd   = htons(wnd);
+
+  if (data_len > 0) {
+    /* let p point to TCP data */
+    pbuf_header(p, -(s16_t)sizeof(struct tcp_hdr));
+    /* copy data */
+    pbuf_take(p, data, (u16_t)data_len);
+    /* let p point to TCP header again */
+    pbuf_header(p, sizeof(struct tcp_hdr));
+  }
+
+  /* calculate checksum */
+
+  tcphdr->chksum = ip_chksum_pseudo(p,
+          IP_PROTO_TCP, p->tot_len, src_ip, dst_ip);
+
+  pbuf_header(p, sizeof(struct ip_hdr));
+
+  return p;
+}
+
+/** Create a TCP segment usable for passing to tcp_input */
+struct pbuf*
+tcp_create_segment(ip_addr_t* src_ip, ip_addr_t* dst_ip,
+                   u16_t src_port, u16_t dst_port, void* data, size_t data_len,
+                   u32_t seqno, u32_t ackno, u8_t headerflags)
+{
+  return tcp_create_segment_wnd(src_ip, dst_ip, src_port, dst_port, data,
+    data_len, seqno, ackno, headerflags, TCP_WND);
+}
+
+/** Create a TCP segment usable for passing to tcp_input
+ * - IP-addresses, ports, seqno and ackno are taken from pcb
+ * - seqno and ackno can be altered with an offset
+ */
+struct pbuf*
+tcp_create_rx_segment(struct tcp_pcb* pcb, void* data, size_t data_len, u32_t seqno_offset,
+                      u32_t ackno_offset, u8_t headerflags)
+{
+  return tcp_create_segment(&pcb->remote_ip, &pcb->local_ip, pcb->remote_port, pcb->local_port,
+    data, data_len, pcb->rcv_nxt + seqno_offset, pcb->lastack + ackno_offset, headerflags);
+}
+
+/** Create a TCP segment usable for passing to tcp_input
+ * - IP-addresses, ports, seqno and ackno are taken from pcb
+ * - seqno and ackno can be altered with an offset
+ * - TCP window can be adjusted
+ */
+struct pbuf* tcp_create_rx_segment_wnd(struct tcp_pcb* pcb, void* data, size_t data_len,
+                   u32_t seqno_offset, u32_t ackno_offset, u8_t headerflags, u16_t wnd)
+{
+  return tcp_create_segment_wnd(&pcb->remote_ip, &pcb->local_ip, pcb->remote_port, pcb->local_port,
+    data, data_len, pcb->rcv_nxt + seqno_offset, pcb->lastack + ackno_offset, headerflags, wnd);
+}
+
+/** Safely bring a tcp_pcb into the requested state */
+void
+tcp_set_state(struct tcp_pcb* pcb, enum tcp_state state, const ip_addr_t* local_ip,
+                   const ip_addr_t* remote_ip, u16_t local_port, u16_t remote_port)
+{
+  u32_t iss;
+
+  /* @todo: are these all states? */
+  /* @todo: remove from previous list */
+  pcb->state = state;
+  
+  iss = tcp_next_iss(pcb);
+  pcb->snd_wl2 = iss;
+  pcb->snd_nxt = iss;
+  pcb->lastack = iss;
+  pcb->snd_lbb = iss;
+  
+  if (state == ESTABLISHED) {
+    TCP_REG(&tcp_active_pcbs, pcb);
+    ip_addr_copy(pcb->local_ip, *local_ip);
+    pcb->local_port = local_port;
+    ip_addr_copy(pcb->remote_ip, *remote_ip);
+    pcb->remote_port = remote_port;
+  } else if(state == LISTEN) {
+    TCP_REG(&tcp_listen_pcbs.pcbs, pcb);
+    ip_addr_copy(pcb->local_ip, *local_ip);
+    pcb->local_port = local_port;
+  } else if(state == TIME_WAIT) {
+    TCP_REG(&tcp_tw_pcbs, pcb);
+    ip_addr_copy(pcb->local_ip, *local_ip);
+    pcb->local_port = local_port;
+    ip_addr_copy(pcb->remote_ip, *remote_ip);
+    pcb->remote_port = remote_port;
+  } else {
+    fail();
+  }
+}
+
+void
+test_tcp_counters_err(void* arg, err_t err)
+{
+  struct test_tcp_counters* counters = (struct test_tcp_counters*)arg;
+  EXPECT_RET(arg != NULL);
+  counters->err_calls++;
+  counters->last_err = err;
+}
+
+static void
+test_tcp_counters_check_rxdata(struct test_tcp_counters* counters, struct pbuf* p)
+{
+  struct pbuf* q;
+  u32_t i, received;
+  if(counters->expected_data == NULL) {
+    /* no data to compare */
+    return;
+  }
+  EXPECT_RET(counters->recved_bytes + p->tot_len <= counters->expected_data_len);
+  received = counters->recved_bytes;
+  for(q = p; q != NULL; q = q->next) {
+    char *data = (char*)q->payload;
+    for(i = 0; i < q->len; i++) {
+      EXPECT_RET(data[i] == counters->expected_data[received]);
+      received++;
+    }
+  }
+  EXPECT(received == counters->recved_bytes + p->tot_len);
+}
+
+err_t
+test_tcp_counters_recv(void* arg, struct tcp_pcb* pcb, struct pbuf* p, err_t err)
+{
+  struct test_tcp_counters* counters = (struct test_tcp_counters*)arg;
+  EXPECT_RETX(arg != NULL, ERR_OK);
+  EXPECT_RETX(pcb != NULL, ERR_OK);
+  EXPECT_RETX(err == ERR_OK, ERR_OK);
+
+  if (p != NULL) {
+    if (counters->close_calls == 0) {
+      counters->recv_calls++;
+      test_tcp_counters_check_rxdata(counters, p);
+      counters->recved_bytes += p->tot_len;
+    } else {
+      counters->recv_calls_after_close++;
+      counters->recved_bytes_after_close += p->tot_len;
+    }
+    pbuf_free(p);
+  } else {
+    counters->close_calls++;
+  }
+  EXPECT(counters->recv_calls_after_close == 0 && counters->recved_bytes_after_close == 0);
+  return ERR_OK;
+}
+
+/** Allocate a pcb and set up the test_tcp_counters_* callbacks */
+struct tcp_pcb*
+test_tcp_new_counters_pcb(struct test_tcp_counters* counters)
+{
+  struct tcp_pcb* pcb = tcp_new();
+  if (pcb != NULL) {
+    /* set up args and callbacks */
+    tcp_arg(pcb, counters);
+    tcp_recv(pcb, test_tcp_counters_recv);
+    tcp_err(pcb, test_tcp_counters_err);
+    pcb->snd_wnd = TCP_WND;
+    pcb->snd_wnd_max = TCP_WND;
+  }
+  return pcb;
+}
+
+/** Calls tcp_input() after adjusting current_iphdr_dest */
+void test_tcp_input(struct pbuf *p, struct netif *inp)
+{
+  struct ip_hdr *iphdr = (struct ip_hdr*)p->payload;
+  /* these lines are a hack, don't use them as an example :-) */
+  ip_addr_copy_from_ip4(*ip_current_dest_addr(), iphdr->dest);
+  ip_addr_copy_from_ip4(*ip_current_src_addr(), iphdr->src);
+  ip_current_netif() = inp;
+  ip_data.current_ip4_header = iphdr;
+
+  /* since adding IPv6, p->payload must point to tcp header, not ip header */
+  pbuf_header(p, -(s16_t)sizeof(struct ip_hdr));
+
+  tcp_input(p, inp);
+
+  ip_addr_set_zero(ip_current_dest_addr());
+  ip_addr_set_zero(ip_current_src_addr());
+  ip_current_netif() = NULL;
+  ip_data.current_ip4_header = NULL;
+}
+
+static err_t test_tcp_netif_output(struct netif *netif, struct pbuf *p,
+       const ip4_addr_t *ipaddr)
+{
+  struct test_tcp_txcounters *txcounters = (struct test_tcp_txcounters*)netif->state;
+  LWIP_UNUSED_ARG(ipaddr);
+  if (txcounters != NULL)
+  {
+    txcounters->num_tx_calls++;
+    txcounters->num_tx_bytes += p->tot_len;
+    if (txcounters->copy_tx_packets) {
+      struct pbuf *p_copy = pbuf_alloc(PBUF_LINK, p->tot_len, PBUF_RAM);
+      err_t err;
+      EXPECT(p_copy != NULL);
+      err = pbuf_copy(p_copy, p);
+      EXPECT(err == ERR_OK);
+      if (txcounters->tx_packets == NULL) {
+        txcounters->tx_packets = p_copy;
+      } else {
+        pbuf_cat(txcounters->tx_packets, p_copy);
+      }
+    }
+  }
+  return ERR_OK;
+}
+
+void test_tcp_init_netif(struct netif *netif, struct test_tcp_txcounters *txcounters,
+                         const ip_addr_t *ip_addr, const ip_addr_t *netmask)
+{
+  struct netif *n;
+  memset(netif, 0, sizeof(struct netif));
+  if (txcounters != NULL) {
+    memset(txcounters, 0, sizeof(struct test_tcp_txcounters));
+    netif->state = txcounters;
+  }
+  netif->output = test_tcp_netif_output;
+  netif->flags |= NETIF_FLAG_UP | NETIF_FLAG_LINK_UP;
+  ip_addr_copy_from_ip4(netif->netmask, *ip_2_ip4(netmask));
+  ip_addr_copy_from_ip4(netif->ip_addr, *ip_2_ip4(ip_addr));
+  for (n = netif_list; n != NULL; n = n->next) {
+    if (n == netif) {
+      return;
+    }
+  }
+  netif->next = NULL;
+  netif_list = netif;
+}

lwip/test/unit/tcp/tcp_helper.h

@@ -0,0 +1,58 @@
+#ifndef LWIP_HDR_TCP_HELPER_H
+#define LWIP_HDR_TCP_HELPER_H
+
+#include "../lwip_check.h"
+#include "lwip/arch.h"
+#include "lwip/tcp.h"
+#include "lwip/netif.h"
+
+/* counters used for test_tcp_counters_* callback functions */
+struct test_tcp_counters {
+  u32_t recv_calls;
+  u32_t recved_bytes;
+  u32_t recv_calls_after_close;
+  u32_t recved_bytes_after_close;
+  u32_t close_calls;
+  u32_t err_calls;
+  err_t last_err;
+  char* expected_data;
+  u32_t expected_data_len;
+};
+
+struct test_tcp_txcounters {
+  u32_t num_tx_calls;
+  u32_t num_tx_bytes;
+  u8_t  copy_tx_packets;
+  struct pbuf *tx_packets;
+};
+
+extern const ip_addr_t test_local_ip;
+extern const ip_addr_t test_remote_ip;
+extern const ip_addr_t test_netmask;
+#define TEST_REMOTE_PORT    0x100
+#define TEST_LOCAL_PORT     0x101
+
+/* Helper functions */
+void tcp_remove_all(void);
+
+struct pbuf* tcp_create_segment(ip_addr_t* src_ip, ip_addr_t* dst_ip,
+                   u16_t src_port, u16_t dst_port, void* data, size_t data_len,
+                   u32_t seqno, u32_t ackno, u8_t headerflags);
+struct pbuf* tcp_create_rx_segment(struct tcp_pcb* pcb, void* data, size_t data_len,
+                   u32_t seqno_offset, u32_t ackno_offset, u8_t headerflags);
+struct pbuf* tcp_create_rx_segment_wnd(struct tcp_pcb* pcb, void* data, size_t data_len,
+                   u32_t seqno_offset, u32_t ackno_offset, u8_t headerflags, u16_t wnd);
+void tcp_set_state(struct tcp_pcb* pcb, enum tcp_state state, const ip_addr_t* local_ip,
+                   const ip_addr_t* remote_ip, u16_t local_port, u16_t remote_port);
+void test_tcp_counters_err(void* arg, err_t err);
+err_t test_tcp_counters_recv(void* arg, struct tcp_pcb* pcb, struct pbuf* p, err_t err);
+
+struct tcp_pcb* test_tcp_new_counters_pcb(struct test_tcp_counters* counters);
+
+void test_tcp_input(struct pbuf *p, struct netif *inp);
+
+void test_tcp_init_netif(struct netif *netif, struct test_tcp_txcounters *txcounters,
+                         const ip_addr_t *ip_addr, const ip_addr_t *netmask);
+
+
+#endif

lwip/test/unit/tcp/test_tcp.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_TCP_H
+#define LWIP_HDR_TEST_TCP_H
+
+#include "../lwip_check.h"
+
+Suite *tcp_suite(void);
+
+#endif

lwip/test/unit/tcp/test_tcp_oos.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_TCP_OOS_H
+#define LWIP_HDR_TEST_TCP_OOS_H
+
+#include "../lwip_check.h"
+
+Suite *tcp_oos_suite(void);
+
+#endif

lwip/test/unit/udp/test_udp.c

@@ -0,0 +1,347 @@
+#include "test_udp.h"
+
+#include "lwip/udp.h"
+#include "lwip/stats.h"
+#include "lwip/inet_chksum.h"
+
+#if !LWIP_STATS || !UDP_STATS || !MEMP_STATS
+#error "This tests needs UDP- and MEMP-statistics enabled"
+#endif
+
+struct test_udp_rxdata {
+  u32_t rx_cnt;
+  u32_t rx_bytes;
+  struct udp_pcb *pcb;
+};
+
+static struct netif test_netif1, test_netif2;
+static ip4_addr_t test_gw1, test_ipaddr1, test_netmask1;
+static ip4_addr_t test_gw2, test_ipaddr2, test_netmask2;
+static int output_ctr, linkoutput_ctr;
+
+/* Helper functions */
+static void
+udp_remove_all(void)
+{
+  struct udp_pcb *pcb = udp_pcbs;
+  struct udp_pcb *pcb2;
+
+  while(pcb != NULL) {
+    pcb2 = pcb;
+    pcb = pcb->next;
+    udp_remove(pcb2);
+  }
+  fail_unless(MEMP_STATS_GET(used, MEMP_UDP_PCB) == 0);
+}
+
+static err_t
+default_netif_output(struct netif *netif, struct pbuf *p, const ip4_addr_t *ipaddr)
+{
+  fail_unless((netif == &test_netif1) || (netif == &test_netif2));
+  fail_unless(p != NULL);
+  fail_unless(ipaddr != NULL);
+  output_ctr++;
+  return ERR_OK;
+}
+
+static err_t
+default_netif_linkoutput(struct netif *netif, struct pbuf *p)
+{
+  fail_unless((netif == &test_netif1) || (netif == &test_netif2));
+  fail_unless(p != NULL);
+  linkoutput_ctr++;
+  return ERR_OK;
+}
+
+static err_t
+default_netif_init(struct netif *netif)
+{
+  fail_unless(netif != NULL);
+  netif->output = default_netif_output;
+  netif->linkoutput = default_netif_linkoutput;
+  netif->mtu = 1500;
+  netif->flags = NETIF_FLAG_BROADCAST | NETIF_FLAG_ETHARP | NETIF_FLAG_LINK_UP;
+  netif->hwaddr_len = 6;
+  return ERR_OK;
+}
+
+static void
+default_netif_add(void)
+{
+  struct netif *n;
+
+#if LWIP_HAVE_LOOPIF
+  fail_unless(netif_list != NULL); /* the loopif */
+  fail_unless(netif_list->next == NULL);
+#else
+  fail_unless(netif_list == NULL);
+#endif
+  fail_unless(netif_default == NULL);
+
+  IP4_ADDR(&test_ipaddr1, 192,168,0,1);
+  IP4_ADDR(&test_netmask1, 255,255,255,0);
+  IP4_ADDR(&test_gw1, 192,168,0,254);
+  n = netif_add(&test_netif1, &test_ipaddr1, &test_netmask1,
+                &test_gw1, NULL, default_netif_init, NULL);
+  fail_unless(n == &test_netif1);
+
+  IP4_ADDR(&test_ipaddr2, 192,168,1,1);
+  IP4_ADDR(&test_netmask2, 255,255,255,0);
+  IP4_ADDR(&test_gw2, 192,168,1,254);
+  n = netif_add(&test_netif2, &test_ipaddr2, &test_netmask2,
+                &test_gw2, NULL, default_netif_init, NULL);
+  fail_unless(n == &test_netif2);
+
+  netif_set_default(&test_netif1);
+  netif_set_up(&test_netif1);
+  netif_set_up(&test_netif2);
+}
+
+static void
+default_netif_remove(void)
+{
+  fail_unless(netif_default == &test_netif1);
+  netif_remove(&test_netif1);
+  netif_remove(&test_netif2);
+  fail_unless(netif_default == NULL);
+#if LWIP_HAVE_LOOPIF
+  fail_unless(netif_list != NULL); /* the loopif */
+  fail_unless(netif_list->next == NULL);
+#else
+  fail_unless(netif_list == NULL);
+#endif
+}
+/* Setups/teardown functions */
+
+static void
+udp_setup(void)
+{
+  udp_remove_all();
+  default_netif_add();
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+static void
+udp_teardown(void)
+{
+  udp_remove_all();
+  default_netif_remove();
+  lwip_check_ensure_no_alloc(SKIP_POOL(MEMP_SYS_TIMEOUT));
+}
+
+
+/* Test functions */
+
+START_TEST(test_udp_new_remove)
+{
+  struct udp_pcb* pcb;
+  LWIP_UNUSED_ARG(_i);
+
+  fail_unless(MEMP_STATS_GET(used, MEMP_UDP_PCB) == 0);
+
+  pcb = udp_new();
+  fail_unless(pcb != NULL);
+  if (pcb != NULL) {
+    fail_unless(MEMP_STATS_GET(used, MEMP_UDP_PCB) == 1);
+    udp_remove(pcb);
+    fail_unless(MEMP_STATS_GET(used, MEMP_UDP_PCB) == 0);
+  }
+}
+END_TEST
+
+static void test_recv(void *arg, struct udp_pcb *pcb, struct pbuf *p,
+    const ip_addr_t *addr, u16_t port)
+{
+  struct test_udp_rxdata *ctr = (struct test_udp_rxdata *)arg;
+
+  LWIP_UNUSED_ARG(addr);
+  LWIP_UNUSED_ARG(port);
+
+  fail_unless(arg != NULL);
+  fail_unless(ctr->pcb == pcb);
+
+  ctr->rx_cnt++;
+  ctr->rx_bytes += p->tot_len;
+
+  if (p != NULL) {
+    pbuf_free(p);
+  }
+}
+
+static struct pbuf *
+test_udp_create_test_packet(u16_t length, u16_t port, u32_t dst_addr)
+{
+  err_t err;
+  u8_t ret;
+  struct udp_hdr *uh;
+  struct ip_hdr *ih;
+  struct pbuf *p;
+  const u8_t test_data[16] = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0xa, 0xb, 0xc, 0xd, 0xe, 0xf};
+
+  p = pbuf_alloc(PBUF_TRANSPORT, length, PBUF_POOL);
+  fail_unless(p != NULL);
+  if (p == NULL) {
+    return NULL;
+  }
+  fail_unless(p->next == NULL);
+  err = pbuf_take(p, test_data, length);
+  fail_unless(err == ERR_OK);
+
+  /* add UDP header */
+  ret = pbuf_add_header(p, sizeof(struct udp_hdr));
+  fail_unless(!ret);
+  uh = (struct udp_hdr *)p->payload;
+  uh->chksum = 0;
+  uh->dest = uh->src = lwip_htons(port);
+  uh->len = lwip_htons(p->tot_len);
+  /* add IPv4 header */
+  ret = pbuf_add_header(p, sizeof(struct ip_hdr));
+  fail_unless(!ret);
+  ih = (struct ip_hdr *)p->payload;
+  memset(ih, 0, sizeof(*ih));
+  ih->dest.addr = dst_addr;
+  ih->_len = lwip_htons(p->tot_len);
+  ih->_ttl = 32;
+  ih->_proto = IP_PROTO_UDP;
+  IPH_VHL_SET(ih, 4, sizeof(struct ip_hdr) / 4);
+  IPH_CHKSUM_SET(ih, inet_chksum(ih, sizeof(struct ip_hdr)));
+  return p;
+}
+
+/* bind 2 pcbs to specific netif IP and test which one gets broadcasts */
+START_TEST(test_udp_broadcast_rx_with_2_netifs)
+{
+  err_t err;
+  struct udp_pcb *pcb1, *pcb2;
+  const u16_t port = 12345;
+  struct test_udp_rxdata ctr1, ctr2;
+  struct pbuf *p;
+#if SO_REUSE
+  struct udp_pcb *pcb_any;
+  struct test_udp_rxdata ctr_any;
+#endif
+  LWIP_UNUSED_ARG(_i);
+
+  pcb1 = udp_new();
+  fail_unless(pcb1 != NULL);
+  pcb2 = udp_new();
+  fail_unless(pcb2 != NULL);
+
+#if SO_REUSE
+  pcb_any = udp_new();
+  fail_unless(pcb_any != NULL);
+
+  ip_set_option(pcb1, SOF_REUSEADDR);
+  ip_set_option(pcb2, SOF_REUSEADDR);
+  ip_set_option(pcb_any, SOF_REUSEADDR);
+
+  err = udp_bind(pcb_any, NULL, port);
+  fail_unless(err == ERR_OK);
+  memset(&ctr_any, 0, sizeof(ctr_any));
+  ctr_any.pcb = pcb_any;
+  udp_recv(pcb_any, test_recv, &ctr_any);
+#endif
+
+  err = udp_bind(pcb1, &test_netif1.ip_addr, port);
+  fail_unless(err == ERR_OK);
+  err = udp_bind(pcb2, &test_netif2.ip_addr, port);
+  fail_unless(err == ERR_OK);
+
+  memset(&ctr1, 0, sizeof(ctr1));
+  ctr1.pcb = pcb1;
+  memset(&ctr2, 0, sizeof(ctr2));
+  ctr2.pcb = pcb2;
+
+  udp_recv(pcb1, test_recv, &ctr1);
+  udp_recv(pcb2, test_recv, &ctr2);
+
+  /* unicast to netif1 */
+  p = test_udp_create_test_packet(16, port, test_ipaddr1.addr);
+  EXPECT_RET(p != NULL);
+  err = ip4_input(p, &test_netif1);
+  fail_unless(err == ERR_OK);
+  fail_unless(ctr1.rx_cnt == 1);
+  fail_unless(ctr1.rx_bytes == 16);
+  fail_unless(ctr2.rx_cnt == 0);
+#if SO_REUSE
+  fail_unless(ctr_any.rx_cnt == 0);
+#endif
+  ctr1.rx_cnt = ctr1.rx_bytes = 0;
+
+  /* unicast to netif2 */
+  p = test_udp_create_test_packet(16, port, test_ipaddr2.addr);
+  EXPECT_RET(p != NULL);
+  err = ip4_input(p, &test_netif2);
+  fail_unless(err == ERR_OK);
+  fail_unless(ctr2.rx_cnt == 1);
+  fail_unless(ctr2.rx_bytes == 16);
+  fail_unless(ctr1.rx_cnt == 0);
+#if SO_REUSE
+  fail_unless(ctr_any.rx_cnt == 0);
+#endif
+  ctr2.rx_cnt = ctr2.rx_bytes = 0;
+
+  /* broadcast to netif1-broadcast, input to netif2 */
+  p = test_udp_create_test_packet(16, port, test_ipaddr1.addr | ~test_netmask1.addr);
+  EXPECT_RET(p != NULL);
+  err = ip4_input(p, &test_netif2);
+  fail_unless(err == ERR_OK);
+  fail_unless(ctr1.rx_cnt == 1);
+  fail_unless(ctr1.rx_bytes == 16);
+  fail_unless(ctr2.rx_cnt == 0);
+#if SO_REUSE
+  fail_unless(ctr_any.rx_cnt == 0);
+#endif
+  ctr1.rx_cnt = ctr1.rx_bytes = 0;
+
+  /* broadcast to netif2-broadcast, input to netif1 */
+  p = test_udp_create_test_packet(16, port, test_ipaddr2.addr | ~test_netmask2.addr);
+  EXPECT_RET(p != NULL);
+  err = ip4_input(p, &test_netif1);
+  fail_unless(err == ERR_OK);
+  fail_unless(ctr2.rx_cnt == 1);
+  fail_unless(ctr2.rx_bytes == 16);
+  fail_unless(ctr1.rx_cnt == 0);
+#if SO_REUSE
+  fail_unless(ctr_any.rx_cnt == 0);
+#endif
+  ctr2.rx_cnt = ctr2.rx_bytes = 0;
+
+  /* broadcast to global-broadcast, input to netif1 */
+  p = test_udp_create_test_packet(16, port, 0xffffffff);
+  EXPECT_RET(p != NULL);
+  err = ip4_input(p, &test_netif1);
+  fail_unless(err == ERR_OK);
+  fail_unless(ctr1.rx_cnt == 1);
+  fail_unless(ctr1.rx_bytes == 16);
+  fail_unless(ctr2.rx_cnt == 0);
+#if SO_REUSE
+  fail_unless(ctr_any.rx_cnt == 0);
+#endif
+  ctr1.rx_cnt = ctr1.rx_bytes = 0;
+
+  /* broadcast to global-broadcast, input to netif2 */
+  p = test_udp_create_test_packet(16, port, 0xffffffff);
+  EXPECT_RET(p != NULL);
+  err = ip4_input(p, &test_netif2);
+  fail_unless(err == ERR_OK);
+  fail_unless(ctr2.rx_cnt == 1);
+  fail_unless(ctr2.rx_bytes == 16);
+  fail_unless(ctr1.rx_cnt == 0);
+#if SO_REUSE
+  fail_unless(ctr_any.rx_cnt == 0);
+#endif
+  ctr2.rx_cnt = ctr2.rx_bytes = 0;
+}
+END_TEST
+
+/** Create the suite including all tests for this module */
+Suite *
+udp_suite(void)
+{
+  testfunc tests[] = {
+    TESTFUNC(test_udp_new_remove),
+    TESTFUNC(test_udp_broadcast_rx_with_2_netifs)
+  };
+  return create_suite("UDP", tests, sizeof(tests)/sizeof(testfunc), udp_setup, udp_teardown);
+}

lwip/test/unit/udp/test_udp.h

@@ -0,0 +1,8 @@
+#ifndef LWIP_HDR_TEST_UDP_H
+#define LWIP_HDR_TEST_UDP_H
+
+#include "../lwip_check.h"
+
+Suite* udp_suite(void);
+
+#endif