diff --git a/src/net_common.c b/src/net_common.c
index fdbb2fa..c4b88fe 100644
--- a/src/net_common.c
+++ b/src/net_common.c
@@ -102,13 +102,33 @@ static int gtp_jump_to_layer(const char *raw_data,  int raw_layer_type, int expe
 	}else if((*next_ip_layer_hdr & 0x60) == 0x60){
 		skip_len = ipv6_jump_to_layer((char *)next_ip_layer_hdr, __ADDR_TYPE_IP_PAIR_V6, expect_layer_type);
 	}else{
-		////sapp_runtime_log(20, "TODO: jmp unsupport type in GTP, 0x%x!\n", (*next_ip_layer_hdr));
+		printf("TODO: jmp unsupport type in GTP, 0x%x!\n", (*next_ip_layer_hdr));
 		return -1;
 	}
 
 	return opt_len + sizeof(struct gtp_hdr) + skip_len;
 }
 
+#define TEREDO_AUTH_HDR_FLAG (0x0001)
+#define TEREDO_INDICATION_HDR_FLAG (0x0000)
+
+#define TEREDO_INDICATION_HDR_LEN (8)
+
+struct teredo_auth_hdr{
+#if __BYTE_ORDER == __LITTLE_ENDIAN
+	unsigned short flags;
+	unsigned char id_len;
+	unsigned char au_len;
+#elif __BYTE_ORDER == __BIG_ENDIAN
+	unsigned char au_len;
+    unsigned char id_len;
+	unsigned short flags;
+#else
+#error "Please check <endian.h>"
+#endif
+};
+
+
 static int udp_jump_to_layer(const char *raw_data,  int raw_layer_type, int expect_layer_type)
 {
 	const struct mesa_udp_hdr *uh = (const struct mesa_udp_hdr *)raw_data;
@@ -132,11 +152,38 @@ static int udp_jump_to_layer(const char *raw_data,  int raw_layer_type, int expe
 		}
 		skip_len += 8; /* skip vxlan header */
 	}else if((3544 == usport) || (3544 == udport)){
-		/* teredo实际没有数据包头, 直接跳转到ipv6层即可 */
-		skip_len = ipv6_jump_to_layer(raw_data+sizeof(struct mesa_udp_hdr), __ADDR_TYPE_IP_PAIR_V6, expect_layer_type);
+		/* teredo实际没有数据包头, 直接跳转到ipv6层即可 , before 20181204*/
+		const char *next_layer_hdr = raw_data+sizeof(struct mesa_udp_hdr);
+        struct teredo_auth_hdr *p_teredo_hdr;
+		int teredo_layer_len = 0, tmp_hdr_len = 0;
+		while((*next_layer_hdr & 0xF0) != 0x60)
+		{
+			p_teredo_hdr = (struct teredo_auth_hdr *)next_layer_hdr;
+			if(p_teredo_hdr->flags == ntohs(TEREDO_AUTH_HDR_FLAG))
+			{
+				//rfc4380 5.1.1 teredo 负载0x0001时为Teredo authentication headers，需要跳过
+				tmp_hdr_len += sizeof(struct teredo_auth_hdr) + ntohs(p_teredo_hdr->au_len) + ntohs
+				(p_teredo_hdr->id_len) + 8 + 1;
+				next_layer_hdr += tmp_hdr_len;
+				teredo_layer_len += tmp_hdr_len;
+			}
+			else if(p_teredo_hdr->flags == ntohs(TEREDO_INDICATION_HDR_FLAG))
+			{
+				//rfc4380 teredo 负载0x0000时为Teredo indication headers，需要跳过
+				next_layer_hdr += TEREDO_INDICATION_HDR_LEN;
+				teredo_layer_len += TEREDO_INDICATION_HDR_LEN;
+			}
+            else
+            {
+                printf("udp_jump_to_layer(): unsupport teredo hdr:0x%d!\n", *(unsigned int *)(next_layer_hdr));
+                return -1;
+            }
+		}
+		skip_len = ipv6_jump_to_layer(next_layer_hdr, __ADDR_TYPE_IP_PAIR_V6, expect_layer_type);
 		if(skip_len < 0){
 			return -1;
 		}
+		skip_len += teredo_layer_len;
 	}else{
 		/* 其他UDP类型不支持再跳转 */
 		return -1;
@@ -147,21 +194,20 @@ static int udp_jump_to_layer(const char *raw_data,  int raw_layer_type, int expe
 
 static int ipv4_jump_to_layer(const char *raw_data,  int raw_layer_type, int expect_layer_type)
 {
-	struct ip *p_ip_hdr = (struct ip *)raw_data;
+	struct mesa_ip4_hdr *p_ip_hdr = (struct mesa_ip4_hdr *)raw_data;
 	int skip_len = 0;
-	int ip_hdr_len = IP_HL(p_ip_hdr) * 4;
+	int ip_hdr_len = p_ip_hdr->ip_hl * 4;
 	//const char *next_layer_data = raw_data + ip_hdr_len;
 
 	if(raw_layer_type == expect_layer_type){
 		return 0;
 	}
 
-	/* IP分片不再继续跳转 */
-	unsigned short *iphdr_off = (unsigned short *)&p_ip_hdr->ip_off[0];
-	if((htons(*iphdr_off) & IP_MF) || (htons(*iphdr_off) & IP_OFFMASK)){
+	if((ntohs(p_ip_hdr->ip_off) & IP_MF ) || (ntohs(p_ip_hdr->ip_off) & IP_OFFMASK)){
+		/* IP分片不再继续向内层跳转 */
 		return -1;
 	}
-	
+				
 	switch(p_ip_hdr->ip_p){
 		case IPPROTO_TCP:
 			if(ADDR_TYPE_TCP == expect_layer_type){
@@ -202,6 +248,23 @@ static int ipv4_jump_to_layer(const char *raw_data,  int raw_layer_type, int exp
 	return skip_len + sizeof(struct ip);
 }
 
+/*
+ *	NextHeader field of IPv6 header
+ */
+#define NEXTHDR_HOP			0	/* Hop-by-hop option header. */
+#define NEXTHDR_IPIP			4	/* IPIP header. */
+#define NEXTHDR_TCP			6	/* TCP segment. */
+#define NEXTHDR_UDP			17	/* UDP message. */
+#define NEXTHDR_IPV6			41	/* IPv6 in IPv6 */
+#define NEXTHDR_ROUTING		43	/* Routing header. */
+#define NEXTHDR_FRAGMENT		44	/* Fragmentation/reassembly header. */
+#define NEXTHDR_ESP			50	/* Encapsulating security payload. */
+#define NEXTHDR_AUTH			51	/* Authentication header. */
+#define NEXTHDR_ICMP			58	/* ICMP for IPv6. */
+#define NEXTHDR_NONE			59	/* No next header */
+#define NEXTHDR_DEST			60	/* Destination options header. */
+#define NEXTHDR_MOBILITY		135	/* Mobility header. */
+
 static int ipv6_jump_to_layer(const char *raw_data,  int raw_layer_type, int expect_layer_type)
 {
 	const struct mesa_ip6_hdr *a_packet = (const struct mesa_ip6_hdr *)raw_data;
@@ -218,14 +281,14 @@ static int ipv6_jump_to_layer(const char *raw_data,  int raw_layer_type, int exp
 		offset_to_ip6 = 0;
 		switch(next_hdr_type)
 		{
-			case 0: //NEXTHDR_HOP:
-			case 43://NEXTHDR_ROUTING:
-			case 51://NEXTHDR_AUTH:
-			case 60://NEXTHDR_DEST:
+			case NEXTHDR_HOP:
+			case NEXTHDR_ROUTING:
+			case NEXTHDR_AUTH:
+			case NEXTHDR_DEST:
 				offset_to_ip6 = (*(next_hdr_ptr + 1))*8 + 8; /* 选项长度以8字节为单位 */				
 			break;
 
-			case 4://NEXTHDR_IPIP:
+			case NEXTHDR_IPIP:
 				if(__ADDR_TYPE_IP_PAIR_V4 == expect_layer_type){
 					skip_len = next_hdr_ptr - (UINT8 *)raw_data;
 				}else{
@@ -239,17 +302,17 @@ static int ipv6_jump_to_layer(const char *raw_data,  int raw_layer_type, int exp
 				goto done;
 			break;
 
-			case 59://NEXTHDR_NONE:
+			case NEXTHDR_NONE:
 				skip_len = -1;
 				goto done;					
 			break;
 
-			case 58://NEXTHDR_ICMP: /* IMCP不再承载其他协议 */
+			case NEXTHDR_ICMP: /* IMCP不再承载其他协议 */
 				skip_len = -1;
 				goto done;
 			break;
 			
-			case 6://NEXTHDR_TCP:
+			case NEXTHDR_TCP:
 				if(ADDR_TYPE_TCP == expect_layer_type){
 					skip_len = next_hdr_ptr - (UINT8 *)raw_data;
 				}else{
@@ -258,7 +321,7 @@ static int ipv6_jump_to_layer(const char *raw_data,  int raw_layer_type, int exp
 				goto done;
 			break;
 				
-			case 17://NEXTHDR_UDP:
+			case NEXTHDR_UDP:
 				if(ADDR_TYPE_UDP == expect_layer_type){
 					skip_len = next_hdr_ptr - (UINT8 *)raw_data;
 				}else{
@@ -268,17 +331,18 @@ static int ipv6_jump_to_layer(const char *raw_data,  int raw_layer_type, int exp
 				goto done;
 			break;
 
-			case 44:///NEXTHDR_FRAGMENT:	
-				/* IP分片不再继续跳转 */
-				return -1;
+			case NEXTHDR_FRAGMENT:	
+				/* IP分片不再继续向内层跳转 */
+				skip_len = -1;
+				goto done;
 			break;
 
-			case 50://NEXTHDR_ESP:
+			case NEXTHDR_ESP:
 				skip_len = -1;
 				goto done;
 
 			default:
-				printf("TODO:jmp Unknown IPv6 header type:0x%x!\n", next_hdr_type);		
+				printf("ipv6_jump_to_layer(): unknown IPv6 header type:0x%x!\n", next_hdr_type);		
 				skip_len = -1;
 				goto done;
 			break;
@@ -330,7 +394,7 @@ static int ppp_jump_to_layer(const char *raw_data,  int raw_layer_type, int expe
 		case PPP_IPCP:
 		case PPP_PAP:
 		case PPP_CHAP:
-		case 0xC025: ///PPP_LQR:
+		case PPP_LQM:
 		case PPP_PROTOCOL_LCP: 
 		
 			/* 不承载应用层协议 */
@@ -338,7 +402,7 @@ static int ppp_jump_to_layer(const char *raw_data,  int raw_layer_type, int expe
 		break;
 
 		default:
-			printf("TODO: jmp unsupport ppp pro:0x%x!\n", ntohs(pppoe_ses_hdr->ppp_protocol));
+			printf("ppp_jump_to_layer(): unsupport ppp pro:0x%x!\n", ntohs(pppoe_ses_hdr->ppp_protocol));
 		break;
 		
 	}
@@ -349,33 +413,118 @@ static int ppp_jump_to_layer(const char *raw_data,  int raw_layer_type, int expe
 	
 	return skip_len + sizeof(struct mesa_pppoe_session_hdr);
 }
+static int guess_mpls_with_control_word(const unsigned char *maybe_eth_hdr)
+{
+	const struct mesa_ethernet_hdr *ehdr = (struct mesa_ethernet_hdr *)(maybe_eth_hdr);
 
+	/* 
+	    MPLS没有字段表示承载的协议类型, 靠猜!!
+	    RFC4623, or https://wiki.mikrotik.com/wiki/Manual:VPLS_Control_Word
+	    除标准的ipv4, ipv6之外, 还有可能是ethernet, 还有可能是带PW Ethernet Control Word的, 格式如下:
+	    
+        0               1               2                 3
+		0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2
+		+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+		|0 0 0 0| flags |frag|len(6bit) |    Sequence Number(16bit)     |
+		+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+
+		实际情况证明, 靠4bit是否==0为条件, 还是可能猜错, 再判断一下以太类型是否为常见类型.
+	*/
+
+	switch(ntohs(ehdr->ether_type)){
+		case ETH_P_IP:
+		case ETH_P_IPV6:
+		case ETH_P_8021Q:
+		case ETH_P_MPLS_UC:
+		case ETH_P_PPP_SES:
+			return 0; /* 合法的ethernet类型, 不含CW */
+		break;
+	}
+
+	ehdr = (struct mesa_ethernet_hdr *)(maybe_eth_hdr + 4);
+	switch(ntohs(ehdr->ether_type)){
+		case ETH_P_IP:
+		case ETH_P_IPV6:
+		case ETH_P_8021Q:
+		case ETH_P_MPLS_UC:
+		case ETH_P_PPP_SES:
+			return 1; /* 合法的ethernet类型, 包含CW */
+		break;
+	}
+
+	/* TODO: 以上都不是, 此处应该返回个什么值? */
+	return 0;
+}
+
+static int set_mpls_addr(struct layer_addr_mpls *addr, const unsigned char *raw_mpls_pkt_data)
+{
+	const struct mesa_mpls_hdr *this_mpls_hdr;
+	int i;
+	
+    memset(addr, 0, sizeof(struct layer_addr_mpls));
+	for(i = 0; i < MAX_MPLS_ADDR_LAYER; i++)
+	{
+		this_mpls_hdr = (const struct mesa_mpls_hdr *)raw_mpls_pkt_data;
+		memcpy(&addr->src_mpls_pkt[i], raw_mpls_pkt_data,  sizeof(struct mesa_mpls_hdr));
+		addr->src_mpls_layer_num += 1;
+		if(1 == this_mpls_hdr->mpls_bls){
+			raw_mpls_pkt_data += sizeof(struct mesa_mpls_hdr); /* 为了后面方便判断是否还有ctrl word */
+			break;
+		}
+		raw_mpls_pkt_data += sizeof(struct mesa_mpls_hdr);
+	}
+
+	if(1 != this_mpls_hdr->mpls_bls) /* 超过MAX_MPLS_ADDR_LAYER, MPLS还没有结束 */
+	{
+		printf("MPLS layer number over load, only support %d\n", MAX_MPLS_ADDR_LAYER);
+		return -1;
+	}
+
+	if(((*raw_mpls_pkt_data & 0xF0) != 0x40) && ((*raw_mpls_pkt_data & 0xF0) != 0x60)){ //VPLS, MPLS with Control Word
+		if(guess_mpls_with_control_word(raw_mpls_pkt_data) > 0){
+			memcpy(&addr->src_mpls_ctrl_word, raw_mpls_pkt_data,  sizeof(int));
+			addr->src_has_ctrl_word = 1;
+		}
+	}
+
+	return 0;
+}
 
 static int mpls_jump_to_layer(const char *raw_data,  int raw_layer_type, int expect_layer_type)
 {
 	int skip_len = 0;
-	const struct mesa_mpls_hdr *mpls_hdr = (const struct mesa_mpls_hdr *)raw_data;
-	const char *next_layer_data = raw_data + sizeof(struct mesa_mpls_hdr);
+	struct layer_addr_mpls mpls_addr = {};
+	const char *next_layer_data;
+	int mpls_layer_len;
 
 	if(raw_layer_type == expect_layer_type){
 		return 0;
 	}
 
-	if(0 == mpls_hdr->mpls_bls){ /* 非MPLS栈底, 递归调用自身 */
-		return mpls_jump_to_layer(next_layer_data, ADDR_TYPE_MPLS, expect_layer_type) + sizeof(struct mesa_mpls_hdr);
+	set_mpls_addr(&mpls_addr, (unsigned char *)raw_data);
+	mpls_layer_len = mpls_addr.src_mpls_layer_num * sizeof(int);
+	if(mpls_addr.src_has_ctrl_word){
+		mpls_layer_len += sizeof(int);
 	}
 
+	next_layer_data = raw_data + mpls_layer_len;
+
 	/* MPLS没有字段标识下一层是什么, 靠猜测下一层的IP类型 */
-	if((*next_layer_data & 0x40) == 0x40){
+	if((*next_layer_data & 0xF0) == 0x40){
 		skip_len = ipv4_jump_to_layer(next_layer_data, __ADDR_TYPE_IP_PAIR_V4, expect_layer_type);
-	}else if((*next_layer_data & 0x60) == 0x60){
+	}else if((*next_layer_data & 0xF0) == 0x60){
 		skip_len = ipv6_jump_to_layer(next_layer_data, __ADDR_TYPE_IP_PAIR_V6, expect_layer_type);
-	}else{
-		//sapp_runtime_log(20, "TODO: jmp unsupport type in MPLS, 0x%x!\n",  (unsigned char)(*next_layer_data));
-		return -1;
+	}else{	
+		/* VPLS和不带control word一并处理, 在set_mpls_addr()函数中, 如果包括Control word, next_layer_data已经跳过了四字节的control word  */
+		skip_len = eth_jump_to_layer(next_layer_data, ADDR_TYPE_MAC, expect_layer_type);
+		if(skip_len < 0){
+			printf("WARNING: jmp unsupport type in MPLS to Ethernet, 0x%x!\n",  
+							(unsigned char)(*next_layer_data));
+			return -1;
+		}
 	}
 
-	return skip_len + sizeof(struct mesa_mpls_hdr); /* mpls header is 4 byte */
+	return skip_len + mpls_layer_len;
 }
 
 static int __common_eth_type_dispatch(UINT16 eth_type, const char *next_layer_data, int raw_layer_type, int expect_layer_type)
@@ -443,17 +592,50 @@ static int __common_eth_type_dispatch(UINT16 eth_type, const char *next_layer_da
 	return skip_len;
 }
 
+static int set_vlan_addr(struct layer_addr_vlan *addr, const unsigned char *vlan_tag)
+{
+	int i;
+	const struct mesa_vlan_hdr *vhdr;
+	int vlan_layer_len = 0;
+
+	memset(addr, 0, sizeof(struct layer_addr_vlan));
+
+	addr->vlan_id = ntohs(*(unsigned short *)vlan_tag) & VLAN_ID_MASK; //compat old value
+
+	for(i = 0; i < MAX_VLAN_ADDR_LAYER; i++){
+		vhdr = (struct mesa_vlan_hdr *)vlan_tag;
+		memcpy(&addr->src_vlan_pkt[i], vlan_tag, sizeof(struct mesa_vlan_hdr));
+		vlan_tag += sizeof(struct mesa_vlan_hdr);
+		addr->src_vlan_layer_num++;
+		vlan_layer_len += sizeof(struct mesa_vlan_hdr);
+		if(ETH_P_8021Q != ntohs(vhdr->type)){
+			break;
+		}
+	}
+	
+	return vlan_layer_len;
+}
+
 static int vlan8021q_jump_to_layer(const char *raw_data,  int raw_layer_type, int expect_layer_type)
 {
 	int skip_len = 0;
-	const struct mesa_vlan_hdr *vlan_hdr = (const struct mesa_vlan_hdr *)raw_data;
-	const char *next_layer_data = raw_data + sizeof(struct mesa_vlan_hdr);
+	struct layer_addr_vlan vlan_addr;
+	const char *next_layer_data;
+	const struct mesa_vlan_hdr *vhdr;
+	unsigned short next_layer_type;
+	int vlan_layer_len;
 
 	if(raw_layer_type == expect_layer_type){
 		return 0;
 	}
+
+	set_vlan_addr(&vlan_addr, raw_data);
+	vlan_layer_len = sizeof(struct mesa_vlan_hdr) * vlan_addr.src_vlan_layer_num;
+	next_layer_data = raw_data + vlan_layer_len;
+	vhdr = (struct mesa_vlan_hdr *)&vlan_addr.src_vlan_pkt[vlan_addr.src_vlan_layer_num-1];
+	next_layer_type = ntohs(vhdr->type);
 	
-	switch(ntohs(vlan_hdr->type)){
+	switch(next_layer_type){
 		case ETH_P_ARP:
 			if(ADDR_TYPE_ARP == expect_layer_type){
 				break;
@@ -490,12 +672,18 @@ static int vlan8021q_jump_to_layer(const char *raw_data,  int raw_layer_type, in
 			skip_len = -1;
 		break;
 
-		case 0x8200: /* XinJing捕包发现此类型, 未知什么包 */
-			skip_len = -1;
+		/* QinQ */
+		case ETH_P_8021Q:
+			printf("vlan8021q_jump_to_layer(): multiple VLAN combine to one layer!\n");
+			assert(0);
+		break;
+
+		case ETH_P_MPLS_UC:
+			skip_len = mpls_jump_to_layer(next_layer_data, ADDR_TYPE_MPLS, expect_layer_type);
 		break;
 		
 		default:
-			printf("unsupport type in vlan8021q, 0x%x, You can try set the '--vlan-as-mac-in-mac' arg!\n", ntohs(vlan_hdr->type));
+			printf("vlan8021q_jump_to_layer(): unsupport type: 0x%x!\n", next_layer_type);
 			skip_len = -1;
 	}
 
@@ -503,7 +691,7 @@ static int vlan8021q_jump_to_layer(const char *raw_data,  int raw_layer_type, in
 		return -1;
 	}
 
-	return skip_len + sizeof(struct mesa_vlan_hdr);
+	return skip_len + vlan_layer_len;
 }
 
 static int eth_jump_to_layer(const char *raw_data,  int raw_layer_type, int expect_layer_type)
@@ -564,6 +752,7 @@ const void *MESA_net_jump_to_layer(const void *raw_data,  int raw_layer_type, in
 		return NULL;
 	}
 
+
 	if(ADDR_TYPE_IPV4 == expect_layer_type){
 		/* 转成纯IPv4地址类型 */
 		expect_layer_type = __ADDR_TYPE_IP_PAIR_V4;
@@ -587,7 +776,7 @@ const void *MESA_net_jump_to_layer(const void *raw_data,  int raw_layer_type, in
 			ret = arp_jump_to_layer((const char *)raw_data, raw_layer_type, expect_layer_type);
 		break;
 		case ADDR_TYPE_VLAN:
-			if(treat_vlan_as_mac_in_mac_sw){
+			if(treat_vlan_as_mac_in_mac_sw == 1){
 				ret = mac_in_mac_jump_to_layer((const char *)raw_data, raw_layer_type, expect_layer_type);
 			}else{
 				ret = vlan8021q_jump_to_layer((const char *)raw_data, raw_layer_type, expect_layer_type);
@@ -609,12 +798,15 @@ const void *MESA_net_jump_to_layer(const void *raw_data,  int raw_layer_type, in
 		case ADDR_TYPE_UDP:
 			ret = udp_jump_to_layer((const char *)raw_data, raw_layer_type, expect_layer_type);
 		break;
-				
-		case ADDR_TYPE_PPPOE_SES:
-		case ADDR_TYPE_MPLS:		
+
+		case ADDR_TYPE_MPLS:	
+			ret = mpls_jump_to_layer((const char *)raw_data, raw_layer_type, expect_layer_type);
+		break;
+		
+		case ADDR_TYPE_PPPOE_SES:	
 		case ADDR_TYPE_GRE:
 		default:
-			printf("TODO: jmp unsupport raw_layer_type:%d in MESA_net_jump_to_layer()!\n", raw_layer_type);
+			printf("MESA_net_jump_to_layer(): unsupport raw_layer_type:%d in MESA_net_jump_to_layer()!\n", raw_layer_type);
 			return NULL;
 	}
 
@@ -652,8 +844,7 @@ const void *MESA_net_jump_to_layer_greedy(const void *raw_data, int raw_layer_ty
 				if((ntohs(ip4hdr->ip_off) & IP_MF ) || (ntohs(ip4hdr->ip_off) & IP_OFFMASK)){
 					/* IP分片不再继续向内层跳转 */
 					goto done;
-				}
-				
+				}				
 				if(IPPROTO_UDP == ip4hdr->ip_p){
 					new_next_layer_data = (char *)expect_layer + ip4hdr->ip_hl * 4;
 					new_raw_layer_type = ADDR_TYPE_UDP; /* IP层如果继续向下一层偏移, 只支持UDP, IPIP, GRE, L2TPv3. */
@@ -667,13 +858,12 @@ const void *MESA_net_jump_to_layer_greedy(const void *raw_data, int raw_layer_ty
 			case __ADDR_TYPE_IP_PAIR_V6:
 			{
 				//TODO2,
-				//sapp_runtime_log(20, "MESA_net_jump_to_layer_greedy() not support IPv6 layer yet\n");
 				goto done;		
 			}
 			break;
 
 			default:
-				//sapp_runtime_log(20, "MESA_net_jump_to_layer_greedy() not support layer type:%d\n", expect_layer_type);
+				printf("MESA_net_jump_to_layer_greedy() not support layer type:%d\n", expect_layer_type);
 				goto done;
 		}
 
@@ -710,7 +900,7 @@ UINT8 net_layer_to_ipv4_protocol(int addr_type)
 		break;
 
 		default:
-			printf("unknown ip4 protocolr:%d\n", addr_type);
+			printf("net_layer_to_ipv4_protocol(): unknown ip4 protocolr:%d\n", addr_type);
 			proto = 0xFF;
 		break;
 	}
@@ -724,19 +914,19 @@ UINT8 net_layer_to_ipv6_protocol(int addr_type)
 
 	switch(addr_type){
 		case ADDR_TYPE_TCP:
-			proto = 6; //NEXTHDR_TCP;
+			proto = NEXTHDR_TCP;
 		break;
 
 		case ADDR_TYPE_UDP:
-			proto = 17;///NEXTHDR_UDP;
+			proto = NEXTHDR_UDP;
 		break;
 
 		case __ADDR_TYPE_IP_PAIR_V4:
-			proto = 4;///NEXTHDR_IPIP;
+			proto = NEXTHDR_IPIP;
 		break;
 
 		default:
-			printf("unknown ip6 next-hdr:%d\n", addr_type);
+			printf("net_layer_to_ipv6_protocol(): unknown ip6 next-hdr:%d\n", addr_type);
 			return 0xFF;
 		break;
 	}
@@ -756,10 +946,12 @@ UINT16 net_layer_to_ethernet_protocol(int addr_type)
 
 	switch(addr_type){
 		case __ADDR_TYPE_IP_PAIR_V4:
+		case ADDR_TYPE_IPV4:
 			ether_type = ETH_P_IP;
 		break;
 
 		case __ADDR_TYPE_IP_PAIR_V6:
+		case ADDR_TYPE_IPV6:
 			ether_type = ETH_P_IPV6;
 		break;
 
@@ -769,7 +961,7 @@ UINT16 net_layer_to_ethernet_protocol(int addr_type)
 
 		case ADDR_TYPE_TCP:
 		case ADDR_TYPE_UDP:
-			//sapp_runtime_log(RLOG_LV_FATAL, "%s:%d: Ethernet can't carry addr type:%d directly!\n", __FILE__, __LINE__,addr_type);
+			printf("%s:%d: Ethernet can't carry addr type:%d directly!\n", __FILE__, __LINE__,addr_type);
 			//assert(0);
 			ether_type = -1;
 		break;
@@ -785,10 +977,18 @@ UINT16 net_layer_to_ethernet_protocol(int addr_type)
 		case ADDR_TYPE_ARP:
 			ether_type = ETH_P_ARP;
 		break;
+
+		case ADDR_TYPE_MAC:
+			if(treat_vlan_as_mac_in_mac_sw==1){
+				ether_type = ETHERTYPE_VLAN;
+			}else{
+				ether_type = ETHERTYPE_PANGU_MAC_IN_MAC;
+			}
+		break;
 		
 		default:
 			/* to do, unknown type */
-			///sapp_runtime_log(RLOG_LV_FATAL, "%s:%d: Ethernet addr type:%d!\n", __FILE__, __LINE__, addr_type);
+			printf("%s:%d: unknown ethernet carry addr type:%d!\n", __FILE__, __LINE__, addr_type);
 			ether_type = -1;
 		break;
 	}
@@ -797,6 +997,53 @@ UINT16 net_layer_to_ethernet_protocol(int addr_type)
 }
 
 
+/*
+	将标准Ethernet类型转换为MESA地址类型;
+*/
+enum addr_type_t ethernet_protocol_to_net_layer(UINT16 ether_type_host)
+{
+	enum addr_type_t addrtype = __ADDR_TYPE_INIT;
+
+	switch(ether_type_host){
+		case ETH_P_IP:
+			addrtype = __ADDR_TYPE_IP_PAIR_V4;
+		break;
+
+		case ETH_P_IPV6:
+			addrtype = __ADDR_TYPE_IP_PAIR_V6;
+		break;
+
+		case ETHERTYPE_VLAN:
+			addrtype = ADDR_TYPE_VLAN;
+		break;
+
+		case ETH_P_PPP_SES:
+			addrtype =  ADDR_TYPE_PPPOE_SES;
+		break;
+
+		case ETH_P_MPLS_UC :
+			addrtype = ADDR_TYPE_MPLS;
+		break;
+
+		case ETH_P_ARP:
+			addrtype = ADDR_TYPE_ARP;
+		break;
+
+		case ETHERTYPE_PANGU_MAC_IN_MAC:
+			addrtype = ADDR_TYPE_MAC_IN_MAC;
+		break;
+		
+		default:
+			/* to do, unknown type */
+			printf("%s:%d: unknown ethernet carry addr type:0x%x\n", __FILE__, __LINE__, ether_type_host);
+			;
+		break;
+	}
+
+	return addrtype;
+}
+
+
 /* 删除末尾的换行符"\r\n" */
 void del_last_rn(char *data, int max_len)
 {
@@ -842,6 +1089,8 @@ int get_pkt_len_from_eth_hdr(const struct mesa_ethernet_hdr *ehdr)
 }
 
 
+
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/src/stream_base.h b/src/stream_base.h
index 2311cee..33a1a24 100644
--- a/src/stream_base.h
+++ b/src/stream_base.h
@@ -1,9 +1,10 @@
 #ifndef _APP_STREAM_BASE_H_
 #define _APP_STREAM_BASE_H_ 
 
-#define STREAM_BASE_H_VERSION		(20170616)
+#define STREAM_BASE_H_VERSION		(20201009)
 
 #include <sys/types.h>
+#include <linux/if_ether.h>   
 #include <stdlib.h>
 #include <string.h>
 
@@ -87,6 +88,11 @@ enum addr_type_t{
 #define APP_STATE_FAWPKT   0x00
 #define APP_STATE_DROPPKT  0x10
 
+
+#define APP_STATE_KILL_FOLLOW 0x40 /* 强制CLOSE当前插件后续的所有插件 */
+#define APP_STATE_KILL_OTHER  0x80 /* 强制CLOSE除当前插件外的所有插件 */
+
+
 /* CHN : 流的类型定义 */
 /* ENG : stream type */
 enum stream_type_t{
@@ -112,13 +118,14 @@ enum stream_type_t{
 enum stream_carry_tunnel_t{
 	STREAM_TUNNLE_NON 		= 0, 	/* default is 0, not tunnel; 默认为0, 非隧道; */
 	STREAM_TUNNLE_6OVER4 	= 1 << 0,
-	STREAM_TUNNLE_4OVER6		= 1 << 1,
-	STREAM_TUNNLE_GRE			= 1 << 2,
+	STREAM_TUNNLE_4OVER6	= 1 << 1,
+	STREAM_TUNNLE_GRE		= 1 << 2,
 	STREAM_TUNNLE_IP_IN_IP	= 1 << 3,
 	STREAM_TUNNLE_PPTP		= 1 << 4,
 	STREAM_TUNNLE_L2TP		= 1 << 5,
-	STREAM_TUNNLE_TEREDO		= 1 << 6,
+	STREAM_TUNNLE_TEREDO	= 1 << 6,
 	STREAM_TUNNEL_GPRS_TUNNEL = 1 << 7,
+	STREAM_TUNNEL_MULTI_MAC = 1 << 8, /* is obsoulte */
 };
 
 typedef struct raw_ipfrag_list{
@@ -179,15 +186,15 @@ struct layer_addr_gre
 
 #define VLAN_ID_MASK		(0x0FFF)
 #define VLAN_TAG_LEN 		(4)
+#define MAX_VLAN_ADDR_LAYER (3)
+
 struct layer_addr_vlan
 {
-	UINT16 vlan_id;	/* network order */
-};
-
-#define VLAN_ID_LEN 4
-struct tuplevlan
-{
-	UCHAR vlan_id[VLAN_ID_LEN];
+	UINT16 vlan_id;	/* network order, old value, to be obsoleted */
+	UCHAR src_vlan_layer_num;
+	UCHAR dst_vlan_layer_num;
+	UINT32 src_vlan_pkt[MAX_VLAN_ADDR_LAYER]; /* network order */
+	UINT32 dst_vlan_pkt[MAX_VLAN_ADDR_LAYER]; /* network order */
 };
 
 struct layer_addr_pppoe_session
@@ -209,8 +216,16 @@ struct layer_addr_pppoe_session
 
 struct layer_addr_mac
 {
-	UCHAR dst_mac[MAC_ADDR_LEN]; /* network order */
-	UCHAR src_mac[MAC_ADDR_LEN]; /* network order */
+	/* 
+	   C2S和S2C方向来自不同的链路和设备, 会导致两个方向的mac地址全不一样, 反向注入时不能用传统方式, 即简单的颠倒src和dst,
+	   修改定义如下, mirror模式下, 还是存储在src_addr中,
+	   API不同了, ABI还是向前兼容的, 结构体内存分布与之前的一致. 
+	*/
+	struct ethhdr src_addr;
+	struct ethhdr dst_addr;
+	
+	//UCHAR dst_mac[MAC_ADDR_LEN]; /* network order */
+	//UCHAR src_mac[MAC_ADDR_LEN]; /* network order */
 };
 
 struct layer_addr_ipv4
@@ -276,8 +291,14 @@ struct layer_addr_l2tp
 #define MAX_MPLS_ADDR_LAYER 4
 struct layer_addr_mpls
 {
-	unsigned int mpls_pkt[MAX_MPLS_ADDR_LAYER];
-    char mpls_layer_num;
+	unsigned int src_mpls_pkt[MAX_MPLS_ADDR_LAYER];
+	unsigned int dest_mpls_pkt[MAX_MPLS_ADDR_LAYER];
+    char src_mpls_layer_num;
+    char dest_mpls_layer_num;
+	char src_has_ctrl_word;
+	char dst_has_ctrl_word;
+    unsigned int src_mpls_ctrl_word; /* refer to RFC4623 */
+    unsigned int dst_mpls_ctrl_word; /* refer to RFC4623 */
 };
 
 struct layer_addr_pptp
@@ -288,9 +309,9 @@ struct layer_addr_pptp
 
 struct layer_addr_gtp
 {
-    unsigned long long source;
+    unsigned long long source;  /* From Client */
     unsigned int src_seq;
-    unsigned long long dest;
+    unsigned long long dest;	/* From Server */
     unsigned int dest_seq;
 }__attribute__ ((aligned (1)));
 
@@ -327,7 +348,9 @@ struct layer_addr
 		struct layer_addr_pptp	*pptp;
 		struct layer_addr_mac_in_mac *mimac;
         struct layer_addr_gtp *gtp;
-		void 					*paddr;
+        struct layer_addr_mpls *mpls;
+
+        void 					*paddr;
 	};
 
 };
@@ -399,6 +422,18 @@ struct streaminfo
  };
 
 
+typedef struct {
+	unsigned int type;
+	unsigned int length;
+	union{
+		char	char_value;
+		short	short_value;
+		int	int_value;
+		long	long_value;
+		char	array_value[8];
+		void	*ptr_value; /* more than 8bytes data, or complex struct. */
+	};
+}SAPP_TLV_T;
 
 #ifdef __cplusplus
 extern "C" {
@@ -517,4 +552,3 @@ const raw_ipfrag_list_t *ip_plug_get_raw_ipfrag_list(int thread_num, enum addr_t
 #endif
 
 #endif
-