mtcp: add dpdk 19.11 support.

- dpdk 19.11 added rte_ to some variables and struct.
- change it to support this.

Author: qingmin-liu

dpdk-iface-kmod/dpdk_iface_main.c

@@ -22,7 +22,11 @@
 typedef struct {
 	PciDevice pd;
 	struct rte_eth_dev_info dev_details;
+#if RTE_VERSION < RTE_VERSION_NUM(19, 8, 0, 0)
 	struct ether_addr ports_eth_addr;
+#else
+	struct rte_ether_addr ports_eth_addr;
+#endif
 } DevInfo;
 
 static DevInfo di[RTE_MAX_ETHPORTS];	

mtcp/src/dpdk_module.c

@@ -87,7 +87,7 @@
 
 #define ETHER_IFG			12
 #define	ETHER_PREAMBLE			8
-#define ETHER_OVR			(ETHER_CRC_LEN + ETHER_PREAMBLE + ETHER_IFG)
+#define ETHER_OVR			(MTCP_ETHER_CRC_LEN + ETHER_PREAMBLE + ETHER_IFG)
 
 static const uint16_t nb_rxd = 		RTE_TEST_RX_DESC_DEFAULT;
 static const uint16_t nb_txd = 		RTE_TEST_TX_DESC_DEFAULT;
@@ -98,15 +98,15 @@ static struct rte_mempool *pktmbuf_pool[MAX_CPUS] = {NULL};
 //#define DEBUG				1
 #ifdef DEBUG
 /* ethernet addresses of ports */
-static struct ether_addr ports_eth_addr[RTE_MAX_ETHPORTS];
+static MTCP_ETHER_ADDR ports_eth_addr[RTE_MAX_ETHPORTS];
 #endif
 
 static struct rte_eth_dev_info dev_info[RTE_MAX_ETHPORTS];
 
 static struct rte_eth_conf port_conf = {
 	.rxmode = {
 		.mq_mode	= 	ETH_MQ_RX_RSS,
-		.max_rx_pkt_len = 	ETHER_MAX_LEN,
+		.max_rx_pkt_len = 	MTCP_ETHER_MAX_LEN,
 #if RTE_VERSION > RTE_VERSION_NUM(17, 8, 0, 0)
 		.offloads	=	(
 #if RTE_VERSION < RTE_VERSION_NUM(18, 5, 0, 0)
@@ -412,7 +412,7 @@ dpdk_get_wptr(struct mtcp_thread_context *ctxt, int ifidx, uint16_t pktsize)
 	m = dpc->wmbufs[ifidx].m_table[len_of_mbuf];
 
 	/* retrieve the right write offset */
-	ptr = (void *)rte_pktmbuf_mtod(m, struct ether_hdr *);
+	ptr = (void *)rte_pktmbuf_mtod(m, MTCP_ETHER_HDR *);
 	m->pkt_len = m->data_len = pktsize;
 	m->nb_segs = 1;
 	m->next = NULL;
@@ -467,16 +467,16 @@ dpdk_recv_pkts(struct mtcp_thread_context *ctxt, int ifidx)
 struct rte_mbuf *
 ip_reassemble(struct dpdk_private_context *dpc, struct rte_mbuf *m)
 {
-	struct ether_hdr *eth_hdr;
+	MTCP_ETHER_HDR *eth_hdr;
 	struct rte_ip_frag_tbl *tbl;
 	struct rte_ip_frag_death_row *dr;
 
 	/* if packet is IPv4 */
 	if (RTE_ETH_IS_IPV4_HDR(m->packet_type)) {
-		struct ipv4_hdr *ip_hdr;
+		MTCP_IPV4_HDR *ip_hdr;
 
-		eth_hdr = rte_pktmbuf_mtod(m, struct ether_hdr *);
-		ip_hdr = (struct ipv4_hdr *)(eth_hdr + 1);
+		eth_hdr = rte_pktmbuf_mtod(m, MTCP_ETHER_HDR *);
+		ip_hdr = (MTCP_IPV4_HDR *)(eth_hdr + 1);
 
 		/* if it is a fragmented packet, then try to reassemble. */
 		if (rte_ipv4_frag_pkt_is_fragmented(ip_hdr)) {
@@ -826,7 +826,7 @@ dpdk_dev_ioctl(struct mtcp_thread_context *ctx, int nif, int cmd, void *argp)
 			goto dev_ioctl_err;
 		m = dpc->wmbufs[eidx].m_table[len_of_mbuf - 1];
 		m->ol_flags = PKT_TX_IP_CKSUM | PKT_TX_IPV4;
-		m->l2_len = sizeof(struct ether_hdr);
+		m->l2_len = sizeof(MTCP_ETHER_HDR);
 		m->l3_len = (iph->ihl<<2);
 		break;
 	case PKT_TX_TCP_CSUM:
@@ -835,19 +835,19 @@ dpdk_dev_ioctl(struct mtcp_thread_context *ctx, int nif, int cmd, void *argp)
 		m = dpc->wmbufs[eidx].m_table[len_of_mbuf - 1];
 		tcph = (struct tcphdr *)((unsigned char *)iph + (iph->ihl<<2));
 		m->ol_flags |= PKT_TX_TCP_CKSUM;
-		tcph->check = rte_ipv4_phdr_cksum((struct ipv4_hdr *)iph, m->ol_flags);
+		tcph->check = rte_ipv4_phdr_cksum((MTCP_IPV4_HDR *)iph, m->ol_flags);
 		break;
 #ifdef ENABLELRO
 	case PKT_RX_TCP_LROSEG:
 		m = dpc->cur_rx_m;
 		//if (m->next != NULL)
 		//	rte_prefetch0(rte_pktmbuf_mtod(m->next, void *));
-		iph = rte_pktmbuf_mtod_offset(m, struct iphdr *, sizeof(struct ether_hdr));
+		iph = rte_pktmbuf_mtod_offset(m, struct iphdr *, sizeof(MTCP_ETHER_HDR));
 		tcph = (struct tcphdr *)((u_char *)iph + (iph->ihl << 2));
 		payload = (uint8_t *)tcph + (tcph->doff << 2);
 
 		seg_off = m->data_len -
-			sizeof(struct ether_hdr) - (iph->ihl << 2) -
+			sizeof(MTCP_ETHER_HDR) - (iph->ihl << 2) -
 			(tcph->doff << 2);
 
 		to = (uint8_t *) argp;
@@ -870,13 +870,13 @@ dpdk_dev_ioctl(struct mtcp_thread_context *ctx, int nif, int cmd, void *argp)
 		if ((dev_info[nif].tx_offload_capa & DEV_TX_OFFLOAD_TCP_CKSUM) == 0)
 			goto dev_ioctl_err;
 		m = dpc->wmbufs[eidx].m_table[len_of_mbuf - 1];
-		iph = rte_pktmbuf_mtod_offset(m, struct iphdr *, sizeof(struct ether_hdr));
+		iph = rte_pktmbuf_mtod_offset(m, struct iphdr *, sizeof(MTCP_ETHER_HDR));
 		tcph = (struct tcphdr *)((uint8_t *)iph + (iph->ihl<<2));
-		m->l2_len = sizeof(struct ether_hdr);
+		m->l2_len = sizeof(MTCP_ETHER_HDR);
 		m->l3_len = (iph->ihl<<2);
 		m->l4_len = (tcph->doff<<2);
 		m->ol_flags = PKT_TX_TCP_CKSUM | PKT_TX_IP_CKSUM | PKT_TX_IPV4;
-		tcph->check = rte_ipv4_phdr_cksum((struct ipv4_hdr *)iph, m->ol_flags);
+		tcph->check = rte_ipv4_phdr_cksum((MTCP_IPV4_HDR *)iph, m->ol_flags);
 		break;
 	case PKT_RX_IP_CSUM:
 		if ((dev_info[nif].rx_offload_capa & DEV_RX_OFFLOAD_IPV4_CKSUM) == 0)

mtcp/src/include/mtcp.h

@@ -39,6 +39,22 @@
 #define ERROR (-1)
 #endif
 
+#if RTE_VERSION < RTE_VERSION_NUM(19, 8, 0, 0)
+typedef struct ether_addr MTCP_ETHER_ADDR;
+typedef struct ipv4_hdr MTCP_IPV4_HDR;
+typedef struct ether_hdr MTCP_ETHER_HDR;
+
+#define MTCP_ETHER_MAX_LEN  ETHER_MAX_LEN
+#define MTCP_ETHER_CRC_LEN  ETHER_CRC_LEN
+#else
+typedef struct rte_ether_addr MTCP_ETHER_ADDR;
+typedef struct rte_ipv4_hdr MTCP_IPV4_HDR;
+typedef struct rte_ether_hdr MTCP_ETHER_HDR;
+
+#define MTCP_ETHER_MAX_LEN  RTE_ETHER_MAX_LEN
+#define MTCP_ETHER_CRC_LEN  RTE_ETHER_CRC_LEN
+#endif
+
 #define ETHERNET_HEADER_LEN		14	// sizeof(struct ethhdr)
 #define IP_HEADER_LEN			20	// sizeof(struct iphdr)
 #define TCP_HEADER_LEN			20	// sizeof(struct tcphdr)

mtcp/src/io_module.c

@@ -260,7 +260,11 @@ SetNetEnv(char *dev_name_list, char *port_stat_list)
 		char socket_mem_str[32] = "";
 		// int i;
 		int ret, socket_mem;
+#if RTE_VERSION < RTE_VERSION_NUM(19, 8, 0, 0)
 		static struct ether_addr ports_eth_addr[RTE_MAX_ETHPORTS];
+#else
+		static struct rte_ether_addr ports_eth_addr[RTE_MAX_ETHPORTS];
+#endif
 
 		/* STEP 1: first determine CPU mask */
 		mpz_init(_cpumask);

mtcp/src/onvm_module.c

@@ -63,7 +63,11 @@ static struct rte_mempool *pktmbuf_pool = NULL;
 //#define DEBUG				1
 #ifdef DEBUG
 /* ethernet addresses of ports */
+#if RTE_VERSION < RTE_VERSION_NUM(19, 8, 0, 0)
 static struct ether_addr ports_eth_addr[RTE_MAX_ETHPORTS];
+#else
+static struct rte_ether_addr ports_eth_addr[RTE_MAX_ETHPORTS];
+#endif
 #endif
 
 static struct rte_eth_dev_info dev_info[RTE_MAX_ETHPORTS];