X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Fbonding%2Frte_eth_bond_pmd.c;h=4ca8afb535e1789a190ac330e2e6adfb63be7d88;hb=35b2d13fd6fdcbd191f2a30d74648faeb1186c65;hp=815e0e4c70be7e56e1d33d8b04b82e1659d5360d;hpb=b9c0dea24beb4b893e88d93c1b52481003ff8999;p=dpdk.git diff --git a/drivers/net/bonding/rte_eth_bond_pmd.c b/drivers/net/bonding/rte_eth_bond_pmd.c index 815e0e4c70..4ca8afb535 100644 --- a/drivers/net/bonding/rte_eth_bond_pmd.c +++ b/drivers/net/bonding/rte_eth_bond_pmd.c @@ -25,6 +25,7 @@ #define REORDER_PERIOD_MS 10 #define DEFAULT_POLLING_INTERVAL_10_MS (10) +#define BOND_MAX_MAC_ADDRS 16 #define HASH_L4_PORTS(h) ((h)->src_port ^ (h)->dst_port) @@ -32,20 +33,22 @@ static uint64_t tlb_last_obytets[RTE_MAX_ETHPORTS]; static inline size_t -get_vlan_offset(struct ether_hdr *eth_hdr, uint16_t *proto) +get_vlan_offset(struct rte_ether_hdr *eth_hdr, uint16_t *proto) { size_t vlan_offset = 0; - if (rte_cpu_to_be_16(ETHER_TYPE_VLAN) == *proto) { - struct vlan_hdr *vlan_hdr = (struct vlan_hdr *)(eth_hdr + 1); + if (rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN) == *proto || + rte_cpu_to_be_16(RTE_ETHER_TYPE_QINQ) == *proto) { + struct rte_vlan_hdr *vlan_hdr = + (struct rte_vlan_hdr *)(eth_hdr + 1); - vlan_offset = sizeof(struct vlan_hdr); + vlan_offset = sizeof(struct rte_vlan_hdr); *proto = vlan_hdr->eth_proto; - if (rte_cpu_to_be_16(ETHER_TYPE_VLAN) == *proto) { + if (rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN) == *proto) { vlan_hdr = vlan_hdr + 1; *proto = vlan_hdr->eth_proto; - vlan_offset += sizeof(struct vlan_hdr); + vlan_offset += sizeof(struct rte_vlan_hdr); } } return vlan_offset; @@ -56,28 +59,34 @@ bond_ethdev_rx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) { struct bond_dev_private *internals; - uint16_t num_rx_slave = 0; uint16_t num_rx_total = 0; - + uint16_t slave_count; + uint16_t active_slave; int i; /* Cast to structure, containing bonded device's port id and queue id */ struct bond_rx_queue *bd_rx_q = (struct bond_rx_queue *)queue; - internals = bd_rx_q->dev_private; + slave_count = internals->active_slave_count; + active_slave = internals->active_slave; + for (i = 0; i < slave_count && nb_pkts; i++) { + uint16_t num_rx_slave; - for (i = 0; i < internals->active_slave_count && nb_pkts; i++) { /* Offset of pointer to *bufs increases as packets are received * from other slaves */ - num_rx_slave = rte_eth_rx_burst(internals->active_slaves[i], - bd_rx_q->queue_id, bufs + num_rx_total, nb_pkts); - if (num_rx_slave) { - num_rx_total += num_rx_slave; - nb_pkts -= num_rx_slave; - } + num_rx_slave = + rte_eth_rx_burst(internals->active_slaves[active_slave], + bd_rx_q->queue_id, + bufs + num_rx_total, nb_pkts); + num_rx_total += num_rx_slave; + nb_pkts -= num_rx_slave; + if (++active_slave == slave_count) + active_slave = 0; } + if (++internals->active_slave >= slave_count) + internals->active_slave = 0; return num_rx_total; } @@ -99,7 +108,8 @@ bond_ethdev_rx_burst_active_backup(void *queue, struct rte_mbuf **bufs, static inline uint8_t is_lacp_packets(uint16_t ethertype, uint8_t subtype, struct rte_mbuf *mbuf) { - const uint16_t ether_type_slow_be = rte_be_to_cpu_16(ETHER_TYPE_SLOW); + const uint16_t ether_type_slow_be = + rte_be_to_cpu_16(RTE_ETHER_TYPE_SLOW); return !((mbuf->ol_flags & PKT_RX_VLAN) ? mbuf->vlan_tci : 0) && (ethertype == ether_type_slow_be && @@ -113,7 +123,7 @@ is_lacp_packets(uint16_t ethertype, uint8_t subtype, struct rte_mbuf *mbuf) static struct rte_flow_item_eth flow_item_eth_type_8023ad = { .dst.addr_bytes = { 0 }, .src.addr_bytes = { 0 }, - .type = RTE_BE16(ETHER_TYPE_SLOW), + .type = RTE_BE16(RTE_ETHER_TYPE_SLOW), }; static struct rte_flow_item_eth flow_item_eth_mask_type_8023ad = { @@ -256,25 +266,32 @@ bond_ethdev_rx_burst_8023ad_fast_queue(void *queue, struct rte_mbuf **bufs, uint16_t num_rx_total = 0; /* Total number of received packets */ uint16_t slaves[RTE_MAX_ETHPORTS]; uint16_t slave_count; - - uint16_t i, idx; + uint16_t active_slave; + uint16_t i; /* Copy slave list to protect against slave up/down changes during tx * bursting */ slave_count = internals->active_slave_count; + active_slave = internals->active_slave; memcpy(slaves, internals->active_slaves, sizeof(internals->active_slaves[0]) * slave_count); - for (i = 0, idx = internals->active_slave; - i < slave_count && num_rx_total < nb_pkts; i++, idx++) { - idx = idx % slave_count; + for (i = 0; i < slave_count && nb_pkts; i++) { + uint16_t num_rx_slave; /* Read packets from this slave */ - num_rx_total += rte_eth_rx_burst(slaves[idx], bd_rx_q->queue_id, - &bufs[num_rx_total], nb_pkts - num_rx_total); + num_rx_slave = rte_eth_rx_burst(slaves[active_slave], + bd_rx_q->queue_id, + bufs + num_rx_total, nb_pkts); + num_rx_total += num_rx_slave; + nb_pkts -= num_rx_slave; + + if (++active_slave == slave_count) + active_slave = 0; } - internals->active_slave = idx; + if (++internals->active_slave >= slave_count) + internals->active_slave = 0; return num_rx_total; } @@ -299,10 +316,10 @@ bond_ethdev_tx_burst_8023ad_fast_queue(void *queue, struct rte_mbuf **bufs, /* Mapping array generated by hash function to map mbufs to slaves */ uint16_t bufs_slave_port_idxs[RTE_MAX_ETHPORTS] = { 0 }; - uint16_t slave_tx_count, slave_tx_fail_count[RTE_MAX_ETHPORTS] = { 0 }; + uint16_t slave_tx_count; uint16_t total_tx_count = 0, total_tx_fail_count = 0; - uint16_t i, j; + uint16_t i; if (unlikely(nb_bufs == 0)) return 0; @@ -319,7 +336,7 @@ bond_ethdev_tx_burst_8023ad_fast_queue(void *queue, struct rte_mbuf **bufs, dist_slave_count = 0; for (i = 0; i < slave_count; i++) { - struct port *port = &mode_8023ad_ports[slave_port_ids[i]]; + struct port *port = &bond_mode_8023ad_ports[slave_port_ids[i]]; if (ACTOR_STATE(port, DISTRIBUTING)) dist_slave_port_ids[dist_slave_count++] = @@ -338,7 +355,7 @@ bond_ethdev_tx_burst_8023ad_fast_queue(void *queue, struct rte_mbuf **bufs, for (i = 0; i < nb_bufs; i++) { /* Populate slave mbuf arrays with mbufs for that slave. */ - uint8_t slave_idx = bufs_slave_port_idxs[i]; + uint16_t slave_idx = bufs_slave_port_idxs[i]; slave_bufs[slave_idx][slave_nb_bufs[slave_idx]++] = bufs[i]; } @@ -357,34 +374,12 @@ bond_ethdev_tx_burst_8023ad_fast_queue(void *queue, struct rte_mbuf **bufs, /* If tx burst fails move packets to end of bufs */ if (unlikely(slave_tx_count < slave_nb_bufs[i])) { - slave_tx_fail_count[i] = slave_nb_bufs[i] - + int slave_tx_fail_count = slave_nb_bufs[i] - slave_tx_count; - total_tx_fail_count += slave_tx_fail_count[i]; - - /* - * Shift bufs to beginning of array to allow reordering - * later - */ - for (j = 0; j < slave_tx_fail_count[i]; j++) { - slave_bufs[i][j] = - slave_bufs[i][(slave_tx_count - 1) + j]; - } - } - } - - /* - * If there are tx burst failures we move packets to end of bufs to - * preserve expected PMD behaviour of all failed transmitted being - * at the end of the input mbuf array - */ - if (unlikely(total_tx_fail_count > 0)) { - int bufs_idx = nb_bufs - total_tx_fail_count - 1; - - for (i = 0; i < slave_count; i++) { - if (slave_tx_fail_count[i] > 0) { - for (j = 0; j < slave_tx_fail_count[i]; j++) - bufs[bufs_idx++] = slave_bufs[i][j]; - } + total_tx_fail_count += slave_tx_fail_count; + memcpy(&bufs[nb_bufs - total_tx_fail_count], + &slave_bufs[i][slave_tx_count], + slave_tx_fail_count * sizeof(bufs[0])); } } @@ -399,21 +394,24 @@ bond_ethdev_rx_burst_8023ad(void *queue, struct rte_mbuf **bufs, /* Cast to structure, containing bonded device's port id and queue id */ struct bond_rx_queue *bd_rx_q = (struct bond_rx_queue *)queue; struct bond_dev_private *internals = bd_rx_q->dev_private; - struct ether_addr bond_mac; - - struct ether_hdr *hdr; + struct rte_eth_dev *bonded_eth_dev = + &rte_eth_devices[internals->port_id]; + struct rte_ether_addr *bond_mac = bonded_eth_dev->data->mac_addrs; + struct rte_ether_hdr *hdr; - const uint16_t ether_type_slow_be = rte_be_to_cpu_16(ETHER_TYPE_SLOW); + const uint16_t ether_type_slow_be = + rte_be_to_cpu_16(RTE_ETHER_TYPE_SLOW); uint16_t num_rx_total = 0; /* Total number of received packets */ uint16_t slaves[RTE_MAX_ETHPORTS]; uint16_t slave_count, idx; uint8_t collecting; /* current slave collecting status */ const uint8_t promisc = internals->promiscuous_en; - uint8_t i, j, k; uint8_t subtype; + uint16_t i; + uint16_t j; + uint16_t k; - rte_eth_macaddr_get(internals->port_id, &bond_mac); /* Copy slave list to protect against slave up/down changes during tx * bursting */ slave_count = internals->active_slave_count; @@ -427,7 +425,7 @@ bond_ethdev_rx_burst_8023ad(void *queue, struct rte_mbuf **bufs, } for (i = 0; i < slave_count && num_rx_total < nb_pkts; i++) { j = num_rx_total; - collecting = ACTOR_STATE(&mode_8023ad_ports[slaves[idx]], + collecting = ACTOR_STATE(&bond_mode_8023ad_ports[slaves[idx]], COLLECTING); /* Read packets from this slave */ @@ -449,16 +447,18 @@ bond_ethdev_rx_burst_8023ad(void *queue, struct rte_mbuf **bufs, if (j + 3 < num_rx_total) rte_prefetch0(rte_pktmbuf_mtod(bufs[j + 3], void *)); - hdr = rte_pktmbuf_mtod(bufs[j], struct ether_hdr *); + hdr = rte_pktmbuf_mtod(bufs[j], struct rte_ether_hdr *); subtype = ((struct slow_protocol_frame *)hdr)->slow_protocol.subtype; /* Remove packet from array if it is slow packet or slave is not * in collecting state or bonding interface is not in promiscuous * mode and packet address does not match. */ if (unlikely(is_lacp_packets(hdr->ether_type, subtype, bufs[j]) || - !collecting || (!promisc && - !is_multicast_ether_addr(&hdr->d_addr) && - !is_same_ether_addr(&bond_mac, &hdr->d_addr)))) { + !collecting || + (!promisc && + !rte_is_multicast_ether_addr(&hdr->d_addr) && + !rte_is_same_ether_addr(bond_mac, + &hdr->d_addr)))) { if (hdr->ether_type == ether_type_slow_be) { bond_mode_8023ad_handle_slow_pkt( @@ -479,7 +479,9 @@ bond_ethdev_rx_burst_8023ad(void *queue, struct rte_mbuf **bufs, idx = 0; } - internals->active_slave = idx; + if (++internals->active_slave >= slave_count) + internals->active_slave = 0; + return num_rx_total; } @@ -490,35 +492,31 @@ uint32_t burstnumberTX; #ifdef RTE_LIBRTE_BOND_DEBUG_ALB static void -arp_op_name(uint16_t arp_op, char *buf) +arp_op_name(uint16_t arp_op, char *buf, size_t buf_len) { switch (arp_op) { - case ARP_OP_REQUEST: - snprintf(buf, sizeof("ARP Request"), "%s", "ARP Request"); + case RTE_ARP_OP_REQUEST: + strlcpy(buf, "ARP Request", buf_len); return; - case ARP_OP_REPLY: - snprintf(buf, sizeof("ARP Reply"), "%s", "ARP Reply"); + case RTE_ARP_OP_REPLY: + strlcpy(buf, "ARP Reply", buf_len); return; - case ARP_OP_REVREQUEST: - snprintf(buf, sizeof("Reverse ARP Request"), "%s", - "Reverse ARP Request"); + case RTE_ARP_OP_REVREQUEST: + strlcpy(buf, "Reverse ARP Request", buf_len); return; - case ARP_OP_REVREPLY: - snprintf(buf, sizeof("Reverse ARP Reply"), "%s", - "Reverse ARP Reply"); + case RTE_ARP_OP_REVREPLY: + strlcpy(buf, "Reverse ARP Reply", buf_len); return; - case ARP_OP_INVREQUEST: - snprintf(buf, sizeof("Peer Identify Request"), "%s", - "Peer Identify Request"); + case RTE_ARP_OP_INVREQUEST: + strlcpy(buf, "Peer Identify Request", buf_len); return; - case ARP_OP_INVREPLY: - snprintf(buf, sizeof("Peer Identify Reply"), "%s", - "Peer Identify Reply"); + case RTE_ARP_OP_INVREPLY: + strlcpy(buf, "Peer Identify Reply", buf_len); return; default: break; } - snprintf(buf, sizeof("Unknown"), "%s", "Unknown"); + strlcpy(buf, "Unknown", buf_len); return; } #endif @@ -571,43 +569,31 @@ update_client_stats(uint32_t addr, uint16_t port, uint32_t *TXorRXindicator) } #ifdef RTE_LIBRTE_BOND_DEBUG_ALB -#define MODE6_DEBUG(info, src_ip, dst_ip, eth_h, arp_op, port, burstnumber) \ - RTE_LOG(DEBUG, PMD, \ - "%s " \ - "port:%d " \ - "SrcMAC:%02X:%02X:%02X:%02X:%02X:%02X " \ - "SrcIP:%s " \ - "DstMAC:%02X:%02X:%02X:%02X:%02X:%02X " \ - "DstIP:%s " \ - "%s " \ - "%d\n", \ - info, \ - port, \ - eth_h->s_addr.addr_bytes[0], \ - eth_h->s_addr.addr_bytes[1], \ - eth_h->s_addr.addr_bytes[2], \ - eth_h->s_addr.addr_bytes[3], \ - eth_h->s_addr.addr_bytes[4], \ - eth_h->s_addr.addr_bytes[5], \ - src_ip, \ - eth_h->d_addr.addr_bytes[0], \ - eth_h->d_addr.addr_bytes[1], \ - eth_h->d_addr.addr_bytes[2], \ - eth_h->d_addr.addr_bytes[3], \ - eth_h->d_addr.addr_bytes[4], \ - eth_h->d_addr.addr_bytes[5], \ - dst_ip, \ - arp_op, \ - ++burstnumber) +#define MODE6_DEBUG(info, src_ip, dst_ip, eth_h, arp_op, port, burstnumber) \ + rte_log(RTE_LOG_DEBUG, bond_logtype, \ + "%s port:%d SrcMAC:%02X:%02X:%02X:%02X:%02X:%02X SrcIP:%s " \ + "DstMAC:%02X:%02X:%02X:%02X:%02X:%02X DstIP:%s %s %d\n", \ + info, \ + port, \ + eth_h->s_addr.addr_bytes[0], eth_h->s_addr.addr_bytes[1], \ + eth_h->s_addr.addr_bytes[2], eth_h->s_addr.addr_bytes[3], \ + eth_h->s_addr.addr_bytes[4], eth_h->s_addr.addr_bytes[5], \ + src_ip, \ + eth_h->d_addr.addr_bytes[0], eth_h->d_addr.addr_bytes[1], \ + eth_h->d_addr.addr_bytes[2], eth_h->d_addr.addr_bytes[3], \ + eth_h->d_addr.addr_bytes[4], eth_h->d_addr.addr_bytes[5], \ + dst_ip, \ + arp_op, ++burstnumber) #endif static void -mode6_debug(const char __attribute__((unused)) *info, struct ether_hdr *eth_h, - uint16_t port, uint32_t __attribute__((unused)) *burstnumber) +mode6_debug(const char __attribute__((unused)) *info, + struct rte_ether_hdr *eth_h, uint16_t port, + uint32_t __attribute__((unused)) *burstnumber) { struct ipv4_hdr *ipv4_h; #ifdef RTE_LIBRTE_BOND_DEBUG_ALB - struct arp_hdr *arp_h; + struct rte_arp_hdr *arp_h; char dst_ip[16]; char ArpOp[24]; char buf[16]; @@ -621,7 +607,7 @@ mode6_debug(const char __attribute__((unused)) *info, struct ether_hdr *eth_h, strlcpy(buf, info, 16); #endif - if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_IPv4)) { + if (ether_type == rte_cpu_to_be_16(RTE_ETHER_TYPE_IPv4)) { ipv4_h = (struct ipv4_hdr *)((char *)(eth_h + 1) + offset); ipv4_addr_to_dot(ipv4_h->src_addr, src_ip, MaxIPv4String); #ifdef RTE_LIBRTE_BOND_DEBUG_ALB @@ -631,11 +617,12 @@ mode6_debug(const char __attribute__((unused)) *info, struct ether_hdr *eth_h, update_client_stats(ipv4_h->src_addr, port, burstnumber); } #ifdef RTE_LIBRTE_BOND_DEBUG_ALB - else if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_ARP)) { - arp_h = (struct arp_hdr *)((char *)(eth_h + 1) + offset); + else if (ether_type == rte_cpu_to_be_16(RTE_ETHER_TYPE_ARP)) { + arp_h = (struct rte_arp_hdr *)((char *)(eth_h + 1) + offset); ipv4_addr_to_dot(arp_h->arp_data.arp_sip, src_ip, MaxIPv4String); ipv4_addr_to_dot(arp_h->arp_data.arp_tip, dst_ip, MaxIPv4String); - arp_op_name(rte_be_to_cpu_16(arp_h->arp_op), ArpOp); + arp_op_name(rte_be_to_cpu_16(arp_h->arp_opcode), + ArpOp, sizeof(ArpOp)); MODE6_DEBUG(buf, src_ip, dst_ip, eth_h, ArpOp, port, *burstnumber); } #endif @@ -647,7 +634,7 @@ bond_ethdev_rx_burst_alb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) { struct bond_tx_queue *bd_tx_q = (struct bond_tx_queue *)queue; struct bond_dev_private *internals = bd_tx_q->dev_private; - struct ether_hdr *eth_h; + struct rte_ether_hdr *eth_h; uint16_t ether_type, offset; uint16_t nb_recv_pkts; int i; @@ -655,18 +642,18 @@ bond_ethdev_rx_burst_alb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) nb_recv_pkts = bond_ethdev_rx_burst(queue, bufs, nb_pkts); for (i = 0; i < nb_recv_pkts; i++) { - eth_h = rte_pktmbuf_mtod(bufs[i], struct ether_hdr *); + eth_h = rte_pktmbuf_mtod(bufs[i], struct rte_ether_hdr *); ether_type = eth_h->ether_type; offset = get_vlan_offset(eth_h, ðer_type); - if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_ARP)) { + if (ether_type == rte_cpu_to_be_16(RTE_ETHER_TYPE_ARP)) { #if defined(RTE_LIBRTE_BOND_DEBUG_ALB) || defined(RTE_LIBRTE_BOND_DEBUG_ALB_L1) mode6_debug("RX ARP:", eth_h, bufs[i]->port, &burstnumberRX); #endif bond_mode_alb_arp_recv(eth_h, offset, internals); } #if defined(RTE_LIBRTE_BOND_DEBUG_ALB) || defined(RTE_LIBRTE_BOND_DEBUG_ALB_L1) - else if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_IPv4)) + else if (ether_type == rte_cpu_to_be_16(RTE_ETHER_TYPE_IPv4)) mode6_debug("RX IPv4:", eth_h, bufs[i]->port, &burstnumberRX); #endif } @@ -727,8 +714,8 @@ bond_ethdev_tx_burst_round_robin(void *queue, struct rte_mbuf **bufs, tx_fail_total += tx_fail_slave; memcpy(&bufs[nb_pkts - tx_fail_total], - &slave_bufs[i][num_tx_slave], - tx_fail_slave * sizeof(bufs[0])); + &slave_bufs[i][num_tx_slave], + tx_fail_slave * sizeof(bufs[0])); } num_tx_total += num_tx_slave; } @@ -755,7 +742,7 @@ bond_ethdev_tx_burst_active_backup(void *queue, } static inline uint16_t -ether_hash(struct ether_hdr *eth_hdr) +ether_hash(struct rte_ether_hdr *eth_hdr) { unaligned_uint16_t *word_src_addr = (unaligned_uint16_t *)eth_hdr->s_addr.addr_bytes; @@ -790,14 +777,14 @@ ipv6_hash(struct ipv6_hdr *ipv6_hdr) void burst_xmit_l2_hash(struct rte_mbuf **buf, uint16_t nb_pkts, - uint8_t slave_count, uint16_t *slaves) + uint16_t slave_count, uint16_t *slaves) { - struct ether_hdr *eth_hdr; + struct rte_ether_hdr *eth_hdr; uint32_t hash; int i; for (i = 0; i < nb_pkts; i++) { - eth_hdr = rte_pktmbuf_mtod(buf[i], struct ether_hdr *); + eth_hdr = rte_pktmbuf_mtod(buf[i], struct rte_ether_hdr *); hash = ether_hash(eth_hdr); @@ -807,16 +794,16 @@ burst_xmit_l2_hash(struct rte_mbuf **buf, uint16_t nb_pkts, void burst_xmit_l23_hash(struct rte_mbuf **buf, uint16_t nb_pkts, - uint8_t slave_count, uint16_t *slaves) + uint16_t slave_count, uint16_t *slaves) { uint16_t i; - struct ether_hdr *eth_hdr; + struct rte_ether_hdr *eth_hdr; uint16_t proto; size_t vlan_offset; uint32_t hash, l3hash; for (i = 0; i < nb_pkts; i++) { - eth_hdr = rte_pktmbuf_mtod(buf[i], struct ether_hdr *); + eth_hdr = rte_pktmbuf_mtod(buf[i], struct rte_ether_hdr *); l3hash = 0; proto = eth_hdr->ether_type; @@ -824,12 +811,12 @@ burst_xmit_l23_hash(struct rte_mbuf **buf, uint16_t nb_pkts, vlan_offset = get_vlan_offset(eth_hdr, &proto); - if (rte_cpu_to_be_16(ETHER_TYPE_IPv4) == proto) { + if (rte_cpu_to_be_16(RTE_ETHER_TYPE_IPv4) == proto) { struct ipv4_hdr *ipv4_hdr = (struct ipv4_hdr *) ((char *)(eth_hdr + 1) + vlan_offset); l3hash = ipv4_hash(ipv4_hdr); - } else if (rte_cpu_to_be_16(ETHER_TYPE_IPv6) == proto) { + } else if (rte_cpu_to_be_16(RTE_ETHER_TYPE_IPv6) == proto) { struct ipv6_hdr *ipv6_hdr = (struct ipv6_hdr *) ((char *)(eth_hdr + 1) + vlan_offset); l3hash = ipv6_hash(ipv6_hdr); @@ -845,9 +832,9 @@ burst_xmit_l23_hash(struct rte_mbuf **buf, uint16_t nb_pkts, void burst_xmit_l34_hash(struct rte_mbuf **buf, uint16_t nb_pkts, - uint8_t slave_count, uint16_t *slaves) + uint16_t slave_count, uint16_t *slaves) { - struct ether_hdr *eth_hdr; + struct rte_ether_hdr *eth_hdr; uint16_t proto; size_t vlan_offset; int i; @@ -857,13 +844,14 @@ burst_xmit_l34_hash(struct rte_mbuf **buf, uint16_t nb_pkts, uint32_t hash, l3hash, l4hash; for (i = 0; i < nb_pkts; i++) { - eth_hdr = rte_pktmbuf_mtod(buf[i], struct ether_hdr *); + eth_hdr = rte_pktmbuf_mtod(buf[i], struct rte_ether_hdr *); + size_t pkt_end = (size_t)eth_hdr + rte_pktmbuf_data_len(buf[i]); proto = eth_hdr->ether_type; vlan_offset = get_vlan_offset(eth_hdr, &proto); l3hash = 0; l4hash = 0; - if (rte_cpu_to_be_16(ETHER_TYPE_IPv4) == proto) { + if (rte_cpu_to_be_16(RTE_ETHER_TYPE_IPv4) == proto) { struct ipv4_hdr *ipv4_hdr = (struct ipv4_hdr *) ((char *)(eth_hdr + 1) + vlan_offset); size_t ip_hdr_offset; @@ -881,16 +869,20 @@ burst_xmit_l34_hash(struct rte_mbuf **buf, uint16_t nb_pkts, tcp_hdr = (struct tcp_hdr *) ((char *)ipv4_hdr + ip_hdr_offset); - l4hash = HASH_L4_PORTS(tcp_hdr); + if ((size_t)tcp_hdr + sizeof(*tcp_hdr) + < pkt_end) + l4hash = HASH_L4_PORTS(tcp_hdr); } else if (ipv4_hdr->next_proto_id == IPPROTO_UDP) { udp_hdr = (struct udp_hdr *) ((char *)ipv4_hdr + ip_hdr_offset); - l4hash = HASH_L4_PORTS(udp_hdr); + if ((size_t)udp_hdr + sizeof(*udp_hdr) + < pkt_end) + l4hash = HASH_L4_PORTS(udp_hdr); } } - } else if (rte_cpu_to_be_16(ETHER_TYPE_IPv6) == proto) { + } else if (rte_cpu_to_be_16(RTE_ETHER_TYPE_IPv6) == proto) { struct ipv6_hdr *ipv6_hdr = (struct ipv6_hdr *) ((char *)(eth_hdr + 1) + vlan_offset); l3hash = ipv6_hash(ipv6_hdr); @@ -915,7 +907,7 @@ burst_xmit_l34_hash(struct rte_mbuf **buf, uint16_t nb_pkts, struct bwg_slave { uint64_t bwg_left_int; uint64_t bwg_left_remainder; - uint8_t slave; + uint16_t slave; }; void @@ -968,11 +960,12 @@ bond_ethdev_update_tlb_slave_cb(void *arg) struct bond_dev_private *internals = arg; struct rte_eth_stats slave_stats; struct bwg_slave bwg_array[RTE_MAX_ETHPORTS]; - uint8_t slave_count; + uint16_t slave_count; uint64_t tx_bytes; uint8_t update_stats = 0; - uint8_t i, slave_id; + uint16_t slave_id; + uint16_t i; internals->slave_update_idx++; @@ -1019,9 +1012,9 @@ bond_ethdev_tx_burst_tlb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) uint16_t num_of_slaves = internals->active_slave_count; uint16_t slaves[RTE_MAX_ETHPORTS]; - struct ether_hdr *ether_hdr; - struct ether_addr primary_slave_addr; - struct ether_addr active_slave_addr; + struct rte_ether_hdr *ether_hdr; + struct rte_ether_addr primary_slave_addr; + struct rte_ether_addr active_slave_addr; if (num_of_slaves < 1) return num_tx_total; @@ -1030,7 +1023,7 @@ bond_ethdev_tx_burst_tlb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) sizeof(internals->tlb_slaves_order[0]) * num_of_slaves); - ether_addr_copy(primary_port->data->mac_addrs, &primary_slave_addr); + rte_ether_addr_copy(primary_port->data->mac_addrs, &primary_slave_addr); if (nb_pkts > 3) { for (i = 0; i < 3; i++) @@ -1043,9 +1036,12 @@ bond_ethdev_tx_burst_tlb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) if (j + 3 < nb_pkts) rte_prefetch0(rte_pktmbuf_mtod(bufs[j+3], void*)); - ether_hdr = rte_pktmbuf_mtod(bufs[j], struct ether_hdr *); - if (is_same_ether_addr(ðer_hdr->s_addr, &primary_slave_addr)) - ether_addr_copy(&active_slave_addr, ðer_hdr->s_addr); + ether_hdr = rte_pktmbuf_mtod(bufs[j], + struct rte_ether_hdr *); + if (rte_is_same_ether_addr(ðer_hdr->s_addr, + &primary_slave_addr)) + rte_ether_addr_copy(&active_slave_addr, + ðer_hdr->s_addr); #if defined(RTE_LIBRTE_BOND_DEBUG_ALB) || defined(RTE_LIBRTE_BOND_DEBUG_ALB_L1) mode6_debug("TX IPv4:", ether_hdr, slaves[i], &burstnumberTX); #endif @@ -1079,7 +1075,7 @@ bond_ethdev_tx_burst_alb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) struct bond_tx_queue *bd_tx_q = (struct bond_tx_queue *)queue; struct bond_dev_private *internals = bd_tx_q->dev_private; - struct ether_hdr *eth_h; + struct rte_ether_hdr *eth_h; uint16_t ether_type, offset; struct client_data *client_info; @@ -1109,11 +1105,11 @@ bond_ethdev_tx_burst_alb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) /* Search tx buffer for ARP packets and forward them to alb */ for (i = 0; i < nb_pkts; i++) { - eth_h = rte_pktmbuf_mtod(bufs[i], struct ether_hdr *); + eth_h = rte_pktmbuf_mtod(bufs[i], struct rte_ether_hdr *); ether_type = eth_h->ether_type; offset = get_vlan_offset(eth_h, ðer_type); - if (ether_type == rte_cpu_to_be_16(ETHER_TYPE_ARP)) { + if (ether_type == rte_cpu_to_be_16(RTE_ETHER_TYPE_ARP)) { slave_idx = bond_mode_alb_arp_xmit(eth_h, offset, internals); /* Change src mac in eth header */ @@ -1139,11 +1135,14 @@ bond_ethdev_tx_burst_alb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) /* Allocate new packet to send ARP update on current slave */ upd_pkt = rte_pktmbuf_alloc(internals->mode6.mempool); if (upd_pkt == NULL) { - RTE_LOG(ERR, PMD, "Failed to allocate ARP packet from pool\n"); + RTE_BOND_LOG(ERR, + "Failed to allocate ARP packet from pool"); continue; } - pkt_size = sizeof(struct ether_hdr) + sizeof(struct arp_hdr) - + client_info->vlan_count * sizeof(struct vlan_hdr); + pkt_size = sizeof(struct rte_ether_hdr) + + sizeof(struct rte_arp_hdr) + + client_info->vlan_count * + sizeof(struct rte_vlan_hdr); upd_pkt->data_len = pkt_size; upd_pkt->pkt_len = pkt_size; @@ -1174,7 +1173,8 @@ bond_ethdev_tx_burst_alb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) #if defined(RTE_LIBRTE_BOND_DEBUG_ALB) || defined(RTE_LIBRTE_BOND_DEBUG_ALB_L1) /* Print TX stats including update packets */ for (j = 0; j < slave_bufs_pkts[i]; j++) { - eth_h = rte_pktmbuf_mtod(slave_bufs[i][j], struct ether_hdr *); + eth_h = rte_pktmbuf_mtod(slave_bufs[i][j], + struct rte_ether_hdr *); mode6_debug("TX ARP:", eth_h, i, &burstnumberTX); } #endif @@ -1191,7 +1191,8 @@ bond_ethdev_tx_burst_alb(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) } #if defined(RTE_LIBRTE_BOND_DEBUG_ALB) || defined(RTE_LIBRTE_BOND_DEBUG_ALB_L1) for (j = 0; j < update_bufs_pkts[i]; j++) { - eth_h = rte_pktmbuf_mtod(update_bufs[i][j], struct ether_hdr *); + eth_h = rte_pktmbuf_mtod(update_bufs[i][j], + struct rte_ether_hdr *); mode6_debug("TX ARPupd:", eth_h, i, &burstnumberTX); } #endif @@ -1232,10 +1233,10 @@ bond_ethdev_tx_burst_balance(void *queue, struct rte_mbuf **bufs, /* Mapping array generated by hash function to map mbufs to slaves */ uint16_t bufs_slave_port_idxs[nb_bufs]; - uint16_t slave_tx_count, slave_tx_fail_count[RTE_MAX_ETHPORTS] = { 0 }; + uint16_t slave_tx_count; uint16_t total_tx_count = 0, total_tx_fail_count = 0; - uint16_t i, j; + uint16_t i; if (unlikely(nb_bufs == 0)) return 0; @@ -1258,7 +1259,7 @@ bond_ethdev_tx_burst_balance(void *queue, struct rte_mbuf **bufs, for (i = 0; i < nb_bufs; i++) { /* Populate slave mbuf arrays with mbufs for that slave. */ - uint8_t slave_idx = bufs_slave_port_idxs[i]; + uint16_t slave_idx = bufs_slave_port_idxs[i]; slave_bufs[slave_idx][slave_nb_bufs[slave_idx]++] = bufs[i]; } @@ -1276,34 +1277,12 @@ bond_ethdev_tx_burst_balance(void *queue, struct rte_mbuf **bufs, /* If tx burst fails move packets to end of bufs */ if (unlikely(slave_tx_count < slave_nb_bufs[i])) { - slave_tx_fail_count[i] = slave_nb_bufs[i] - + int slave_tx_fail_count = slave_nb_bufs[i] - slave_tx_count; - total_tx_fail_count += slave_tx_fail_count[i]; - - /* - * Shift bufs to beginning of array to allow reordering - * later - */ - for (j = 0; j < slave_tx_fail_count[i]; j++) { - slave_bufs[i][j] = - slave_bufs[i][(slave_tx_count - 1) + j]; - } - } - } - - /* - * If there are tx burst failures we move packets to end of bufs to - * preserve expected PMD behaviour of all failed transmitted being - * at the end of the input mbuf array - */ - if (unlikely(total_tx_fail_count > 0)) { - int bufs_idx = nb_bufs - total_tx_fail_count - 1; - - for (i = 0; i < slave_count; i++) { - if (slave_tx_fail_count[i] > 0) { - for (j = 0; j < slave_tx_fail_count[i]; j++) - bufs[bufs_idx++] = slave_bufs[i][j]; - } + total_tx_fail_count += slave_tx_fail_count; + memcpy(&bufs[nb_bufs - total_tx_fail_count], + &slave_bufs[i][slave_tx_count], + slave_tx_fail_count * sizeof(bufs[0])); } } @@ -1330,13 +1309,10 @@ bond_ethdev_tx_burst_8023ad(void *queue, struct rte_mbuf **bufs, /* Mapping array generated by hash function to map mbufs to slaves */ uint16_t bufs_slave_port_idxs[RTE_MAX_ETHPORTS] = { 0 }; - uint16_t slave_tx_count, slave_tx_fail_count[RTE_MAX_ETHPORTS] = { 0 }; + uint16_t slave_tx_count; uint16_t total_tx_count = 0, total_tx_fail_count = 0; - uint16_t i, j; - - if (unlikely(nb_bufs == 0)) - return 0; + uint16_t i; /* Copy slave list to protect against slave up/down changes during tx * bursting */ @@ -1347,16 +1323,40 @@ bond_ethdev_tx_burst_8023ad(void *queue, struct rte_mbuf **bufs, memcpy(slave_port_ids, internals->active_slaves, sizeof(slave_port_ids[0]) * slave_count); + /* Check for LACP control packets and send if available */ + for (i = 0; i < slave_count; i++) { + struct port *port = &bond_mode_8023ad_ports[slave_port_ids[i]]; + struct rte_mbuf *ctrl_pkt = NULL; + + if (likely(rte_ring_empty(port->tx_ring))) + continue; + + if (rte_ring_dequeue(port->tx_ring, + (void **)&ctrl_pkt) != -ENOENT) { + slave_tx_count = rte_eth_tx_burst(slave_port_ids[i], + bd_tx_q->queue_id, &ctrl_pkt, 1); + /* + * re-enqueue LAG control plane packets to buffering + * ring if transmission fails so the packet isn't lost. + */ + if (slave_tx_count != 1) + rte_ring_enqueue(port->tx_ring, ctrl_pkt); + } + } + + if (unlikely(nb_bufs == 0)) + return 0; + dist_slave_count = 0; for (i = 0; i < slave_count; i++) { - struct port *port = &mode_8023ad_ports[slave_port_ids[i]]; + struct port *port = &bond_mode_8023ad_ports[slave_port_ids[i]]; if (ACTOR_STATE(port, DISTRIBUTING)) dist_slave_port_ids[dist_slave_count++] = slave_port_ids[i]; } - if (likely(dist_slave_count > 1)) { + if (likely(dist_slave_count > 0)) { /* * Populate slaves mbuf with the packets which are to be sent @@ -1370,7 +1370,7 @@ bond_ethdev_tx_burst_8023ad(void *queue, struct rte_mbuf **bufs, * Populate slave mbuf arrays with mbufs for that * slave */ - uint8_t slave_idx = bufs_slave_port_idxs[i]; + uint16_t slave_idx = bufs_slave_port_idxs[i]; slave_bufs[slave_idx][slave_nb_bufs[slave_idx]++] = bufs[i]; @@ -1391,64 +1391,17 @@ bond_ethdev_tx_burst_8023ad(void *queue, struct rte_mbuf **bufs, /* If tx burst fails move packets to end of bufs */ if (unlikely(slave_tx_count < slave_nb_bufs[i])) { - slave_tx_fail_count[i] = slave_nb_bufs[i] - + int slave_tx_fail_count = slave_nb_bufs[i] - slave_tx_count; - total_tx_fail_count += slave_tx_fail_count[i]; - - /* - * Shift bufs to beginning of array to allow - * reordering later - */ - for (j = 0; j < slave_tx_fail_count[i]; j++) - slave_bufs[i][j] = - slave_bufs[i] - [(slave_tx_count - 1) - + j]; - } - } + total_tx_fail_count += slave_tx_fail_count; - /* - * If there are tx burst failures we move packets to end of - * bufs to preserve expected PMD behaviour of all failed - * transmitted being at the end of the input mbuf array - */ - if (unlikely(total_tx_fail_count > 0)) { - int bufs_idx = nb_bufs - total_tx_fail_count - 1; - - for (i = 0; i < slave_count; i++) { - if (slave_tx_fail_count[i] > 0) { - for (j = 0; - j < slave_tx_fail_count[i]; - j++) { - bufs[bufs_idx++] = - slave_bufs[i][j]; - } - } + memcpy(&bufs[nb_bufs - total_tx_fail_count], + &slave_bufs[i][slave_tx_count], + slave_tx_fail_count * sizeof(bufs[0])); } } } - /* Check for LACP control packets and send if available */ - for (i = 0; i < slave_count; i++) { - struct port *port = &mode_8023ad_ports[slave_port_ids[i]]; - struct rte_mbuf *ctrl_pkt = NULL; - - if (likely(rte_ring_empty(port->tx_ring))) - continue; - - if (rte_ring_dequeue(port->tx_ring, - (void **)&ctrl_pkt) != -ENOENT) { - slave_tx_count = rte_eth_tx_burst(slave_port_ids[i], - bd_tx_q->queue_id, &ctrl_pkt, 1); - /* - * re-enqueue LAG control plane packets to buffering - * ring if transmission fails so the packet isn't lost. - */ - if (slave_tx_count != 1) - rte_ring_enqueue(port->tx_ring, ctrl_pkt); - } - } - return total_tx_count; } @@ -1459,8 +1412,9 @@ bond_ethdev_tx_burst_broadcast(void *queue, struct rte_mbuf **bufs, struct bond_dev_private *internals; struct bond_tx_queue *bd_tx_q; - uint8_t tx_failed_flag = 0, num_of_slaves; uint16_t slaves[RTE_MAX_ETHPORTS]; + uint8_t tx_failed_flag = 0; + uint16_t num_of_slaves; uint16_t max_nb_of_tx_pkts = 0; @@ -1512,7 +1466,7 @@ bond_ethdev_tx_burst_broadcast(void *queue, struct rte_mbuf **bufs, return max_nb_of_tx_pkts; } -void +static void link_properties_set(struct rte_eth_dev *ethdev, struct rte_eth_link *slave_link) { struct bond_dev_private *bond_ctx = ethdev->data->dev_private; @@ -1537,7 +1491,7 @@ link_properties_set(struct rte_eth_dev *ethdev, struct rte_eth_link *slave_link) } } -int +static int link_properties_valid(struct rte_eth_dev *ethdev, struct rte_eth_link *slave_link) { @@ -1556,30 +1510,32 @@ link_properties_valid(struct rte_eth_dev *ethdev, } int -mac_address_get(struct rte_eth_dev *eth_dev, struct ether_addr *dst_mac_addr) +mac_address_get(struct rte_eth_dev *eth_dev, + struct rte_ether_addr *dst_mac_addr) { - struct ether_addr *mac_addr; + struct rte_ether_addr *mac_addr; if (eth_dev == NULL) { - RTE_LOG(ERR, PMD, "%s: NULL pointer eth_dev specified\n", __func__); + RTE_BOND_LOG(ERR, "NULL pointer eth_dev specified"); return -1; } if (dst_mac_addr == NULL) { - RTE_LOG(ERR, PMD, "%s: NULL pointer MAC specified\n", __func__); + RTE_BOND_LOG(ERR, "NULL pointer MAC specified"); return -1; } mac_addr = eth_dev->data->mac_addrs; - ether_addr_copy(mac_addr, dst_mac_addr); + rte_ether_addr_copy(mac_addr, dst_mac_addr); return 0; } int -mac_address_set(struct rte_eth_dev *eth_dev, struct ether_addr *new_mac_addr) +mac_address_set(struct rte_eth_dev *eth_dev, + struct rte_ether_addr *new_mac_addr) { - struct ether_addr *mac_addr; + struct rte_ether_addr *mac_addr; if (eth_dev == NULL) { RTE_BOND_LOG(ERR, "NULL pointer eth_dev specified"); @@ -1600,6 +1556,61 @@ mac_address_set(struct rte_eth_dev *eth_dev, struct ether_addr *new_mac_addr) return 0; } +static const struct rte_ether_addr null_mac_addr; + +/* + * Add additional MAC addresses to the slave + */ +int +slave_add_mac_addresses(struct rte_eth_dev *bonded_eth_dev, + uint16_t slave_port_id) +{ + int i, ret; + struct rte_ether_addr *mac_addr; + + for (i = 1; i < BOND_MAX_MAC_ADDRS; i++) { + mac_addr = &bonded_eth_dev->data->mac_addrs[i]; + if (rte_is_same_ether_addr(mac_addr, &null_mac_addr)) + break; + + ret = rte_eth_dev_mac_addr_add(slave_port_id, mac_addr, 0); + if (ret < 0) { + /* rollback */ + for (i--; i > 0; i--) + rte_eth_dev_mac_addr_remove(slave_port_id, + &bonded_eth_dev->data->mac_addrs[i]); + return ret; + } + } + + return 0; +} + +/* + * Remove additional MAC addresses from the slave + */ +int +slave_remove_mac_addresses(struct rte_eth_dev *bonded_eth_dev, + uint16_t slave_port_id) +{ + int i, rc, ret; + struct rte_ether_addr *mac_addr; + + rc = 0; + for (i = 1; i < BOND_MAX_MAC_ADDRS; i++) { + mac_addr = &bonded_eth_dev->data->mac_addrs[i]; + if (rte_is_same_ether_addr(mac_addr, &null_mac_addr)) + break; + + ret = rte_eth_dev_mac_addr_remove(slave_port_id, mac_addr); + /* save only the first error */ + if (ret < 0 && rc == 0) + rc = ret; + } + + return rc; +} + int mac_address_slaves_update(struct rte_eth_dev *bonded_eth_dev) { @@ -1687,9 +1698,9 @@ bond_ethdev_mode_set(struct rte_eth_dev *eth_dev, int mode) if (internals->mode4.dedicated_queues.enabled == 0) { eth_dev->rx_pkt_burst = bond_ethdev_rx_burst_8023ad; eth_dev->tx_pkt_burst = bond_ethdev_tx_burst_8023ad; - RTE_LOG(WARNING, PMD, + RTE_BOND_LOG(WARNING, "Using mode 4, it is necessary to do TX burst " - "and RX burst at least every 100ms.\n"); + "and RX burst at least every 100ms."); } else { /* Use flow director's optimization */ eth_dev->rx_pkt_burst = @@ -1726,7 +1737,7 @@ slave_configure_slow_queue(struct rte_eth_dev *bonded_eth_dev, int errval = 0; struct bond_dev_private *internals = (struct bond_dev_private *) bonded_eth_dev->data->dev_private; - struct port *port = &mode_8023ad_ports[slave_eth_dev->data->port_id]; + struct port *port = &bond_mode_8023ad_ports[slave_eth_dev->data->port_id]; if (port->slow_pool == NULL) { char mem_name[256]; @@ -1803,12 +1814,11 @@ slave_configure(struct rte_eth_dev *bonded_eth_dev, /* If RSS is enabled for bonding, try to enable it for slaves */ if (bonded_eth_dev->data->dev_conf.rxmode.mq_mode & ETH_MQ_RX_RSS_FLAG) { - if (bonded_eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key_len - != 0) { + if (internals->rss_key_len != 0) { slave_eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key_len = - bonded_eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key_len; + internals->rss_key_len; slave_eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key = - bonded_eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key; + internals->rss_key; } else { slave_eth_dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key = NULL; } @@ -1850,7 +1860,7 @@ slave_configure(struct rte_eth_dev *bonded_eth_dev, nb_rx_queues, nb_tx_queues, &(slave_eth_dev->data->dev_conf)); if (errval != 0) { - RTE_BOND_LOG(ERR, "Cannot configure slave device: port %u , err (%d)", + RTE_BOND_LOG(ERR, "Cannot configure slave device: port %u, err (%d)", slave_eth_dev->data->port_id, errval); return errval; } @@ -1932,10 +1942,10 @@ slave_configure(struct rte_eth_dev *bonded_eth_dev, &internals->reta_conf[0], internals->slaves[i].reta_size); if (errval != 0) { - RTE_LOG(WARNING, PMD, - "rte_eth_dev_rss_reta_update on slave port %d fails (err %d)." - " RSS Configuration for bonding may be inconsistent.\n", - slave_eth_dev->data->port_id, errval); + RTE_BOND_LOG(WARNING, + "rte_eth_dev_rss_reta_update on slave port %d fails (err %d)." + " RSS Configuration for bonding may be inconsistent.", + slave_eth_dev->data->port_id, errval); } break; } @@ -1957,17 +1967,26 @@ void slave_remove(struct bond_dev_private *internals, struct rte_eth_dev *slave_eth_dev) { - uint8_t i; + uint16_t i; for (i = 0; i < internals->slave_count; i++) if (internals->slaves[i].port_id == slave_eth_dev->data->port_id) break; - if (i < (internals->slave_count - 1)) + if (i < (internals->slave_count - 1)) { + struct rte_flow *flow; + memmove(&internals->slaves[i], &internals->slaves[i + 1], sizeof(internals->slaves[0]) * (internals->slave_count - i - 1)); + TAILQ_FOREACH(flow, &internals->flow_list, next) { + memmove(&flow->flows[i], &flow->flows[i + 1], + sizeof(flow->flows[0]) * + (internals->slave_count - i - 1)); + flow->flows[internals->slave_count - 1] = NULL; + } + } internals->slave_count--; @@ -1998,7 +2017,7 @@ slave_add(struct bond_dev_private *internals, slave_details->link_status_wait_to_complete = 0; /* clean tlb_last_obytes when adding port for bonding device */ memcpy(&(slave_details->persisted_mac_addr), slave_eth_dev->data->mac_addrs, - sizeof(struct ether_addr)); + sizeof(struct rte_ether_addr)); } void @@ -2044,7 +2063,7 @@ bond_ethdev_start(struct rte_eth_dev *eth_dev) } if (internals->user_defined_mac == 0) { - struct ether_addr *new_mac_addr = NULL; + struct rte_ether_addr *new_mac_addr = NULL; for (i = 0; i < internals->slave_count; i++) if (internals->slaves[i].port_id == internals->primary_port) @@ -2060,10 +2079,6 @@ bond_ethdev_start(struct rte_eth_dev *eth_dev) } } - /* Update all slave devices MACs*/ - if (mac_address_slaves_update(eth_dev) != 0) - goto out_err; - /* If bonded device is configure in promiscuous mode then re-apply config */ if (internals->promiscuous_en) bond_ethdev_promiscuous_enable(eth_dev); @@ -2104,6 +2119,10 @@ bond_ethdev_start(struct rte_eth_dev *eth_dev) (void *)&rte_eth_devices[internals->port_id]); } + /* Update all slave devices MACs*/ + if (mac_address_slaves_update(eth_dev) != 0) + goto out_err; + if (internals->user_defined_primary_port) bond_ethdev_primary_set(internals, internals->primary_port); @@ -2124,7 +2143,7 @@ out_err: static void bond_ethdev_free_queues(struct rte_eth_dev *dev) { - uint8_t i; + uint16_t i; if (dev->data->rx_queues != NULL) { for (i = 0; i < dev->data->nb_rx_queues; i++) { @@ -2147,7 +2166,7 @@ void bond_ethdev_stop(struct rte_eth_dev *eth_dev) { struct bond_dev_private *internals = eth_dev->data->dev_private; - uint8_t i; + uint16_t i; if (internals->mode == BONDING_MODE_8023AD) { struct port *port; @@ -2157,7 +2176,7 @@ bond_ethdev_stop(struct rte_eth_dev *eth_dev) /* Discard all messages to/from mode 4 state machines */ for (i = 0; i < internals->active_slave_count; i++) { - port = &mode_8023ad_ports[internals->active_slaves[i]]; + port = &bond_mode_8023ad_ports[internals->active_slaves[i]]; RTE_ASSERT(port->rx_ring != NULL); while (rte_ring_dequeue(port->rx_ring, &pkt) != -ENOENT) @@ -2176,35 +2195,44 @@ bond_ethdev_stop(struct rte_eth_dev *eth_dev) tlb_last_obytets[internals->active_slaves[i]] = 0; } - internals->active_slave_count = 0; - internals->link_status_polling_enabled = 0; - for (i = 0; i < internals->slave_count; i++) - internals->slaves[i].last_link_status = 0; - eth_dev->data->dev_link.link_status = ETH_LINK_DOWN; eth_dev->data->dev_started = 0; + + internals->link_status_polling_enabled = 0; + for (i = 0; i < internals->slave_count; i++) { + uint16_t slave_id = internals->slaves[i].port_id; + if (find_slave_by_id(internals->active_slaves, + internals->active_slave_count, slave_id) != + internals->active_slave_count) { + internals->slaves[i].last_link_status = 0; + rte_eth_dev_stop(slave_id); + deactivate_slave(eth_dev, slave_id); + } + } } void bond_ethdev_close(struct rte_eth_dev *dev) { struct bond_dev_private *internals = dev->data->dev_private; - uint8_t bond_port_id = internals->port_id; + uint16_t bond_port_id = internals->port_id; int skipped = 0; + struct rte_flow_error ferror; - RTE_LOG(INFO, EAL, "Closing bonded device %s\n", dev->device->name); + RTE_BOND_LOG(INFO, "Closing bonded device %s", dev->device->name); while (internals->slave_count != skipped) { uint16_t port_id = internals->slaves[skipped].port_id; rte_eth_dev_stop(port_id); if (rte_eth_bond_slave_remove(bond_port_id, port_id) != 0) { - RTE_LOG(ERR, EAL, - "Failed to remove port %d from bonded device " - "%s\n", port_id, dev->device->name); + RTE_BOND_LOG(ERR, + "Failed to remove port %d from bonded device %s", + port_id, dev->device->name); skipped++; } } + bond_flow_ops.flush(dev, &ferror); bond_ethdev_free_queues(dev); rte_bitmap_reset(internals->vlan_filter_bmp); } @@ -2219,12 +2247,14 @@ bond_ethdev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info) uint16_t max_nb_rx_queues = UINT16_MAX; uint16_t max_nb_tx_queues = UINT16_MAX; + uint16_t max_rx_desc_lim = UINT16_MAX; + uint16_t max_tx_desc_lim = UINT16_MAX; - dev_info->max_mac_addrs = 1; + dev_info->max_mac_addrs = BOND_MAX_MAC_ADDRS; dev_info->max_rx_pktlen = internals->candidate_max_rx_pktlen ? internals->candidate_max_rx_pktlen : - ETHER_MAX_JUMBO_FRAME_LEN; + RTE_ETHER_MAX_JUMBO_FRAME_LEN; /* Max number of tx/rx queues that the bonded device can support is the * minimum values of the bonded slaves, as all slaves must be capable @@ -2232,7 +2262,7 @@ bond_ethdev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info) */ if (internals->slave_count > 0) { struct rte_eth_dev_info slave_info; - uint8_t idx; + uint16_t idx; for (idx = 0; idx < internals->slave_count; idx++) { rte_eth_dev_info_get(internals->slaves[idx].port_id, @@ -2243,12 +2273,26 @@ bond_ethdev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info) if (slave_info.max_tx_queues < max_nb_tx_queues) max_nb_tx_queues = slave_info.max_tx_queues; + + if (slave_info.rx_desc_lim.nb_max < max_rx_desc_lim) + max_rx_desc_lim = slave_info.rx_desc_lim.nb_max; + + if (slave_info.tx_desc_lim.nb_max < max_tx_desc_lim) + max_tx_desc_lim = slave_info.tx_desc_lim.nb_max; } } dev_info->max_rx_queues = max_nb_rx_queues; dev_info->max_tx_queues = max_nb_tx_queues; + memcpy(&dev_info->default_rxconf, &internals->default_rxconf, + sizeof(dev_info->default_rxconf)); + memcpy(&dev_info->default_txconf, &internals->default_txconf, + sizeof(dev_info->default_txconf)); + + dev_info->rx_desc_lim.nb_max = max_rx_desc_lim; + dev_info->tx_desc_lim.nb_max = max_tx_desc_lim; + /** * If dedicated hw queues enabled for link bonding device in LACP mode * then we need to reduce the maximum number of data path queues by 1. @@ -2290,9 +2334,9 @@ bond_ethdev_vlan_filter_set(struct rte_eth_dev *dev, uint16_t vlan_id, int on) res = rte_eth_dev_vlan_filter(port_id, vlan_id, on); if (res == ENOTSUP) - RTE_LOG(WARNING, PMD, - "Setting VLAN filter on slave port %u not supported.\n", - port_id); + RTE_BOND_LOG(WARNING, + "Setting VLAN filter on slave port %u not supported.", + port_id); } rte_spinlock_unlock(&internals->lock); @@ -2574,6 +2618,9 @@ bond_ethdev_promiscuous_enable(struct rte_eth_dev *eth_dev) case BONDING_MODE_TLB: case BONDING_MODE_ALB: default: + /* Do not touch promisc when there cannot be primary ports */ + if (internals->slave_count == 0) + break; rte_eth_promiscuous_enable(internals->current_primary_port); } } @@ -2602,6 +2649,9 @@ bond_ethdev_promiscuous_disable(struct rte_eth_dev *dev) case BONDING_MODE_TLB: case BONDING_MODE_ALB: default: + /* Do not touch promisc when there cannot be primary ports */ + if (internals->slave_count == 0) + break; rte_eth_promiscuous_disable(internals->current_primary_port); } } @@ -2625,14 +2675,15 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type, struct rte_eth_link link; int rc = -1; - int i, valid_slave = 0; - uint8_t active_pos; uint8_t lsc_flag = 0; + int valid_slave = 0; + uint16_t active_pos; + uint16_t i; if (type != RTE_ETH_EVENT_INTR_LSC || param == NULL) return rc; - bonded_eth_dev = &rte_eth_devices[*(uint8_t *)param]; + bonded_eth_dev = &rte_eth_devices[*(uint16_t *)param]; if (check_for_bonded_ethdev(bonded_eth_dev)) return rc; @@ -2654,6 +2705,11 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type, if (!valid_slave) return rc; + /* Synchronize lsc callback parallel calls either by real link event + * from the slaves PMDs or by the bonding PMD itself. + */ + rte_spinlock_lock(&internals->lsc_lock); + /* Search for port in active port list */ active_pos = find_slave_by_id(internals->active_slaves, internals->active_slave_count, port_id); @@ -2661,12 +2717,26 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type, rte_eth_link_get_nowait(port_id, &link); if (link.link_status) { if (active_pos < internals->active_slave_count) - return rc; + goto link_update; + + /* check link state properties if bonded link is up*/ + if (bonded_eth_dev->data->dev_link.link_status == ETH_LINK_UP) { + if (link_properties_valid(bonded_eth_dev, &link) != 0) + RTE_BOND_LOG(ERR, "Invalid link properties " + "for slave %d in bonding mode %d", + port_id, internals->mode); + } else { + /* inherit slave link properties */ + link_properties_set(bonded_eth_dev, &link); + } - /* if no active slave ports then set this port to be primary port */ + /* If no active slave ports then set this port to be + * the primary port. + */ if (internals->active_slave_count < 1) { /* If first active slave, then change link status */ - bonded_eth_dev->data->dev_link.link_status = ETH_LINK_UP; + bonded_eth_dev->data->dev_link.link_status = + ETH_LINK_UP; internals->current_primary_port = port_id; lsc_flag = 1; @@ -2675,13 +2745,15 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type, activate_slave(bonded_eth_dev, port_id); - /* If user has defined the primary port then default to using it */ + /* If the user has defined the primary port then default to + * using it. + */ if (internals->user_defined_primary_port && internals->primary_port == port_id) bond_ethdev_primary_set(internals, port_id); } else { if (active_pos == internals->active_slave_count) - return rc; + goto link_update; /* Remove from active slave list */ deactivate_slave(bonded_eth_dev, port_id); @@ -2700,6 +2772,7 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type, } } +link_update: /** * Update bonded device link properties after any change to active * slaves @@ -2734,7 +2807,10 @@ bond_ethdev_lsc_event_callback(uint16_t port_id, enum rte_eth_event_type type, NULL); } } - return 0; + + rte_spinlock_unlock(&internals->lsc_lock); + + return rc; } static int @@ -2873,7 +2949,8 @@ bond_ethdev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu) } static int -bond_ethdev_mac_address_set(struct rte_eth_dev *dev, struct ether_addr *addr) +bond_ethdev_mac_address_set(struct rte_eth_dev *dev, + struct rte_ether_addr *addr) { if (mac_address_set(dev, addr)) { RTE_BOND_LOG(ERR, "Failed to update MAC address"); @@ -2883,6 +2960,80 @@ bond_ethdev_mac_address_set(struct rte_eth_dev *dev, struct ether_addr *addr) return 0; } +static int +bond_filter_ctrl(struct rte_eth_dev *dev __rte_unused, + enum rte_filter_type type, enum rte_filter_op op, void *arg) +{ + if (type == RTE_ETH_FILTER_GENERIC && op == RTE_ETH_FILTER_GET) { + *(const void **)arg = &bond_flow_ops; + return 0; + } + return -ENOTSUP; +} + +static int +bond_ethdev_mac_addr_add(struct rte_eth_dev *dev, + struct rte_ether_addr *mac_addr, + __rte_unused uint32_t index, uint32_t vmdq) +{ + struct rte_eth_dev *slave_eth_dev; + struct bond_dev_private *internals = dev->data->dev_private; + int ret, i; + + rte_spinlock_lock(&internals->lock); + + for (i = 0; i < internals->slave_count; i++) { + slave_eth_dev = &rte_eth_devices[internals->slaves[i].port_id]; + if (*slave_eth_dev->dev_ops->mac_addr_add == NULL || + *slave_eth_dev->dev_ops->mac_addr_remove == NULL) { + ret = -ENOTSUP; + goto end; + } + } + + for (i = 0; i < internals->slave_count; i++) { + ret = rte_eth_dev_mac_addr_add(internals->slaves[i].port_id, + mac_addr, vmdq); + if (ret < 0) { + /* rollback */ + for (i--; i >= 0; i--) + rte_eth_dev_mac_addr_remove( + internals->slaves[i].port_id, mac_addr); + goto end; + } + } + + ret = 0; +end: + rte_spinlock_unlock(&internals->lock); + return ret; +} + +static void +bond_ethdev_mac_addr_remove(struct rte_eth_dev *dev, uint32_t index) +{ + struct rte_eth_dev *slave_eth_dev; + struct bond_dev_private *internals = dev->data->dev_private; + int i; + + rte_spinlock_lock(&internals->lock); + + for (i = 0; i < internals->slave_count; i++) { + slave_eth_dev = &rte_eth_devices[internals->slaves[i].port_id]; + if (*slave_eth_dev->dev_ops->mac_addr_remove == NULL) + goto end; + } + + struct rte_ether_addr *mac_addr = &dev->data->mac_addrs[index]; + + for (i = 0; i < internals->slave_count; i++) + rte_eth_dev_mac_addr_remove(internals->slaves[i].port_id, + mac_addr); + +end: + rte_spinlock_unlock(&internals->lock); +} + const struct eth_dev_ops default_dev_ops = { .dev_start = bond_ethdev_start, .dev_stop = bond_ethdev_stop, @@ -2904,7 +3055,10 @@ const struct eth_dev_ops default_dev_ops = { .rss_hash_update = bond_ethdev_rss_hash_update, .rss_hash_conf_get = bond_ethdev_rss_hash_conf_get, .mtu_set = bond_ethdev_mtu_set, - .mac_addr_set = bond_ethdev_mac_address_set + .mac_addr_set = bond_ethdev_mac_address_set, + .mac_addr_add = bond_ethdev_mac_addr_add, + .mac_addr_remove = bond_ethdev_mac_addr_remove, + .filter_ctrl = bond_filter_ctrl }; static int @@ -2931,10 +3085,13 @@ bond_alloc(struct rte_vdev_device *dev, uint8_t mode) eth_dev->data->nb_rx_queues = (uint16_t)1; eth_dev->data->nb_tx_queues = (uint16_t)1; - eth_dev->data->mac_addrs = rte_zmalloc_socket(name, ETHER_ADDR_LEN, 0, - socket_id); + /* Allocate memory for storing MAC addresses */ + eth_dev->data->mac_addrs = rte_zmalloc_socket(name, RTE_ETHER_ADDR_LEN * + BOND_MAX_MAC_ADDRS, 0, socket_id); if (eth_dev->data->mac_addrs == NULL) { - RTE_BOND_LOG(ERR, "Unable to malloc mac_addrs"); + RTE_BOND_LOG(ERR, + "Failed to allocate %u bytes needed to store MAC addresses", + RTE_ETHER_ADDR_LEN * BOND_MAX_MAC_ADDRS); goto err; } @@ -2942,6 +3099,7 @@ bond_alloc(struct rte_vdev_device *dev, uint8_t mode) eth_dev->data->dev_flags = RTE_ETH_DEV_INTR_LSC; rte_spinlock_init(&internals->lock); + rte_spinlock_init(&internals->lsc_lock); internals->port_id = eth_dev->data->port_id; internals->mode = BONDING_MODE_INVALID; @@ -2969,33 +3127,44 @@ bond_alloc(struct rte_vdev_device *dev, uint8_t mode) /* Initially allow to choose any offload type */ internals->flow_type_rss_offloads = ETH_RSS_PROTO_MASK; + memset(&internals->default_rxconf, 0, + sizeof(internals->default_rxconf)); + memset(&internals->default_txconf, 0, + sizeof(internals->default_txconf)); + + memset(&internals->rx_desc_lim, 0, sizeof(internals->rx_desc_lim)); + memset(&internals->tx_desc_lim, 0, sizeof(internals->tx_desc_lim)); + memset(internals->active_slaves, 0, sizeof(internals->active_slaves)); memset(internals->slaves, 0, sizeof(internals->slaves)); + TAILQ_INIT(&internals->flow_list); + internals->flow_isolated_valid = 0; + /* Set mode 4 default configuration */ bond_mode_8023ad_setup(eth_dev, NULL); if (bond_ethdev_mode_set(eth_dev, mode)) { - RTE_BOND_LOG(ERR, "Failed to set bonded device %d mode too %d", + RTE_BOND_LOG(ERR, "Failed to set bonded device %d mode to %d", eth_dev->data->port_id, mode); goto err; } vlan_filter_bmp_size = - rte_bitmap_get_memory_footprint(ETHER_MAX_VLAN_ID + 1); + rte_bitmap_get_memory_footprint(RTE_ETHER_MAX_VLAN_ID + 1); internals->vlan_filter_bmpmem = rte_malloc(name, vlan_filter_bmp_size, RTE_CACHE_LINE_SIZE); if (internals->vlan_filter_bmpmem == NULL) { RTE_BOND_LOG(ERR, - "Failed to allocate vlan bitmap for bonded device %u\n", + "Failed to allocate vlan bitmap for bonded device %u", eth_dev->data->port_id); goto err; } - internals->vlan_filter_bmp = rte_bitmap_init(ETHER_MAX_VLAN_ID + 1, + internals->vlan_filter_bmp = rte_bitmap_init(RTE_ETHER_MAX_VLAN_ID + 1, internals->vlan_filter_bmpmem, vlan_filter_bmp_size); if (internals->vlan_filter_bmp == NULL) { RTE_BOND_LOG(ERR, - "Failed to init vlan bitmap for bonded device %u\n", + "Failed to init vlan bitmap for bonded device %u", eth_dev->data->port_id); rte_free(internals->vlan_filter_bmpmem); goto err; @@ -3005,10 +3174,9 @@ bond_alloc(struct rte_vdev_device *dev, uint8_t mode) err: rte_free(internals); - if (eth_dev != NULL) { - rte_free(eth_dev->data->mac_addrs); - rte_eth_dev_release_port(eth_dev); - } + if (eth_dev != NULL) + eth_dev->data->dev_private = NULL; + rte_eth_dev_release_port(eth_dev); return -1; } @@ -3021,12 +3189,26 @@ bond_probe(struct rte_vdev_device *dev) uint8_t bonding_mode, socket_id/*, agg_mode*/; int arg_count, port_id; uint8_t agg_mode; + struct rte_eth_dev *eth_dev; if (!dev) return -EINVAL; name = rte_vdev_device_name(dev); - RTE_LOG(INFO, EAL, "Initializing pmd_bond for %s\n", name); + RTE_BOND_LOG(INFO, "Initializing pmd_bond for %s", name); + + if (rte_eal_process_type() == RTE_PROC_SECONDARY) { + eth_dev = rte_eth_dev_attach_secondary(name); + if (!eth_dev) { + RTE_BOND_LOG(ERR, "Failed to probe %s", name); + return -1; + } + /* TODO: request info from primary to set up Rx and Tx */ + eth_dev->dev_ops = &default_dev_ops; + eth_dev->device = &dev->device; + rte_eth_dev_probing_finish(eth_dev); + return 0; + } kvlist = rte_kvargs_parse(rte_vdev_device_args(dev), pmd_bond_init_valid_arguments); @@ -3038,13 +3220,13 @@ bond_probe(struct rte_vdev_device *dev) if (rte_kvargs_process(kvlist, PMD_BOND_MODE_KVARG, &bond_ethdev_parse_slave_mode_kvarg, &bonding_mode) != 0) { - RTE_LOG(ERR, EAL, "Invalid mode for bonded device %s\n", + RTE_BOND_LOG(ERR, "Invalid mode for bonded device %s", name); goto parse_error; } } else { - RTE_LOG(ERR, EAL, "Mode must be specified only once for bonded " - "device %s\n", name); + RTE_BOND_LOG(ERR, "Mode must be specified only once for bonded " + "device %s", name); goto parse_error; } @@ -3054,13 +3236,13 @@ bond_probe(struct rte_vdev_device *dev) if (rte_kvargs_process(kvlist, PMD_BOND_SOCKET_ID_KVARG, &bond_ethdev_parse_socket_id_kvarg, &socket_id) != 0) { - RTE_LOG(ERR, EAL, "Invalid socket Id specified for " - "bonded device %s\n", name); + RTE_BOND_LOG(ERR, "Invalid socket Id specified for " + "bonded device %s", name); goto parse_error; } } else if (arg_count > 1) { - RTE_LOG(ERR, EAL, "Socket Id can be specified only once for " - "bonded device %s\n", name); + RTE_BOND_LOG(ERR, "Socket Id can be specified only once for " + "bonded device %s", name); goto parse_error; } else { socket_id = rte_socket_id(); @@ -3071,34 +3253,33 @@ bond_probe(struct rte_vdev_device *dev) /* Create link bonding eth device */ port_id = bond_alloc(dev, bonding_mode); if (port_id < 0) { - RTE_LOG(ERR, EAL, "Failed to create socket %s in mode %u on " - "socket %u.\n", name, bonding_mode, socket_id); + RTE_BOND_LOG(ERR, "Failed to create socket %s in mode %u on " + "socket %u.", name, bonding_mode, socket_id); goto parse_error; } internals = rte_eth_devices[port_id].data->dev_private; internals->kvlist = kvlist; - if (rte_kvargs_count(kvlist, PMD_BOND_AGG_MODE_KVARG) == 1) { if (rte_kvargs_process(kvlist, PMD_BOND_AGG_MODE_KVARG, &bond_ethdev_parse_slave_agg_mode_kvarg, &agg_mode) != 0) { - RTE_LOG(ERR, EAL, - "Failed to parse agg selection mode for bonded device %s\n", + RTE_BOND_LOG(ERR, + "Failed to parse agg selection mode for bonded device %s", name); goto parse_error; } if (internals->mode == BONDING_MODE_8023AD) - rte_eth_bond_8023ad_agg_selection_set(port_id, - agg_mode); + internals->mode4.agg_selection = agg_mode; } else { - rte_eth_bond_8023ad_agg_selection_set(port_id, AGG_STABLE); + internals->mode4.agg_selection = AGG_STABLE; } - RTE_LOG(INFO, EAL, "Create bonded device %s on port %d in mode %u on " - "socket %u.\n", name, port_id, bonding_mode, socket_id); + rte_eth_dev_probing_finish(&rte_eth_devices[port_id]); + RTE_BOND_LOG(INFO, "Create bonded device %s on port %d in mode %u on " + "socket %u.", name, port_id, bonding_mode, socket_id); return 0; parse_error: @@ -3118,7 +3299,7 @@ bond_remove(struct rte_vdev_device *dev) return -EINVAL; name = rte_vdev_device_name(dev); - RTE_LOG(INFO, EAL, "Uninitializing pmd_bond for %s\n", name); + RTE_BOND_LOG(INFO, "Uninitializing pmd_bond for %s", name); /* now free all data allocation - for eth_dev structure, * dummy pci driver and internal (private) data @@ -3129,6 +3310,9 @@ bond_remove(struct rte_vdev_device *dev) if (eth_dev == NULL) return -ENODEV; + if (rte_eal_process_type() != RTE_PROC_PRIMARY) + return rte_eth_dev_release_port(eth_dev); + RTE_ASSERT(eth_dev->device == &dev->device); internals = eth_dev->data->dev_private; @@ -3151,8 +3335,6 @@ bond_remove(struct rte_vdev_device *dev) rte_mempool_free(internals->mode6.mempool); rte_bitmap_free(internals->vlan_filter_bmp); rte_free(internals->vlan_filter_bmpmem); - rte_free(eth_dev->data->dev_private); - rte_free(eth_dev->data->mac_addrs); rte_eth_dev_release_port(eth_dev); @@ -3180,16 +3362,30 @@ bond_ethdev_configure(struct rte_eth_dev *dev) unsigned i, j; - /* If RSS is enabled, fill table and key with default values */ + /* + * If RSS is enabled, fill table with default values and + * set key to the the value specified in port RSS configuration. + * Fall back to default RSS key if the key is not specified + */ if (dev->data->dev_conf.rxmode.mq_mode & ETH_MQ_RX_RSS) { - dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key = internals->rss_key; - dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key_len = 0; - memcpy(internals->rss_key, default_rss_key, 40); + if (dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key != NULL) { + internals->rss_key_len = + dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key_len; + memcpy(internals->rss_key, + dev->data->dev_conf.rx_adv_conf.rss_conf.rss_key, + internals->rss_key_len); + } else { + internals->rss_key_len = sizeof(default_rss_key); + memcpy(internals->rss_key, default_rss_key, + internals->rss_key_len); + } for (i = 0; i < RTE_DIM(internals->reta_conf); i++) { internals->reta_conf[i].mask = ~0LL; for (j = 0; j < RTE_RETA_GROUP_SIZE; j++) - internals->reta_conf[i].reta[j] = j % dev->data->nb_rx_queues; + internals->reta_conf[i].reta[j] = + (i * RTE_RETA_GROUP_SIZE + j) % + dev->data->nb_rx_queues; } } @@ -3206,26 +3402,26 @@ bond_ethdev_configure(struct rte_eth_dev *dev) /* Parse MAC address for bonded device */ arg_count = rte_kvargs_count(kvlist, PMD_BOND_MAC_ADDR_KVARG); if (arg_count == 1) { - struct ether_addr bond_mac; + struct rte_ether_addr bond_mac; if (rte_kvargs_process(kvlist, PMD_BOND_MAC_ADDR_KVARG, - &bond_ethdev_parse_bond_mac_addr_kvarg, &bond_mac) < 0) { - RTE_LOG(INFO, EAL, "Invalid mac address for bonded device %s\n", - name); + &bond_ethdev_parse_bond_mac_addr_kvarg, &bond_mac) < 0) { + RTE_BOND_LOG(INFO, "Invalid mac address for bonded device %s", + name); return -1; } /* Set MAC address */ if (rte_eth_bond_mac_address_set(port_id, &bond_mac) != 0) { - RTE_LOG(ERR, EAL, - "Failed to set mac address on bonded device %s\n", - name); + RTE_BOND_LOG(ERR, + "Failed to set mac address on bonded device %s", + name); return -1; } } else if (arg_count > 1) { - RTE_LOG(ERR, EAL, - "MAC address can be specified only once for bonded device %s\n", - name); + RTE_BOND_LOG(ERR, + "MAC address can be specified only once for bonded device %s", + name); return -1; } @@ -3235,40 +3431,47 @@ bond_ethdev_configure(struct rte_eth_dev *dev) uint8_t xmit_policy; if (rte_kvargs_process(kvlist, PMD_BOND_XMIT_POLICY_KVARG, - &bond_ethdev_parse_balance_xmit_policy_kvarg, &xmit_policy) != - 0) { - RTE_LOG(INFO, EAL, - "Invalid xmit policy specified for bonded device %s\n", - name); + &bond_ethdev_parse_balance_xmit_policy_kvarg, &xmit_policy) != + 0) { + RTE_BOND_LOG(INFO, + "Invalid xmit policy specified for bonded device %s", + name); return -1; } /* Set balance mode transmit policy*/ if (rte_eth_bond_xmit_policy_set(port_id, xmit_policy) != 0) { - RTE_LOG(ERR, EAL, - "Failed to set balance xmit policy on bonded device %s\n", - name); + RTE_BOND_LOG(ERR, + "Failed to set balance xmit policy on bonded device %s", + name); return -1; } } else if (arg_count > 1) { - RTE_LOG(ERR, EAL, - "Transmit policy can be specified only once for bonded device" - " %s\n", name); + RTE_BOND_LOG(ERR, + "Transmit policy can be specified only once for bonded device %s", + name); return -1; } if (rte_kvargs_count(kvlist, PMD_BOND_AGG_MODE_KVARG) == 1) { if (rte_kvargs_process(kvlist, - PMD_BOND_AGG_MODE_KVARG, - &bond_ethdev_parse_slave_agg_mode_kvarg, - &agg_mode) != 0) { - RTE_LOG(ERR, EAL, - "Failed to parse agg selection mode for bonded device %s\n", + PMD_BOND_AGG_MODE_KVARG, + &bond_ethdev_parse_slave_agg_mode_kvarg, + &agg_mode) != 0) { + RTE_BOND_LOG(ERR, + "Failed to parse agg selection mode for bonded device %s", + name); + } + if (internals->mode == BONDING_MODE_8023AD) { + int ret = rte_eth_bond_8023ad_agg_selection_set(port_id, + agg_mode); + if (ret < 0) { + RTE_BOND_LOG(ERR, + "Invalid args for agg selection set for bonded device %s", name); + return -1; + } } - if (internals->mode == BONDING_MODE_8023AD) - rte_eth_bond_8023ad_agg_selection_set(port_id, - agg_mode); } /* Parse/add slave ports to bonded device */ @@ -3279,23 +3482,23 @@ bond_ethdev_configure(struct rte_eth_dev *dev) memset(&slave_ports, 0, sizeof(slave_ports)); if (rte_kvargs_process(kvlist, PMD_BOND_SLAVE_PORT_KVARG, - &bond_ethdev_parse_slave_port_kvarg, &slave_ports) != 0) { - RTE_LOG(ERR, EAL, - "Failed to parse slave ports for bonded device %s\n", - name); + &bond_ethdev_parse_slave_port_kvarg, &slave_ports) != 0) { + RTE_BOND_LOG(ERR, + "Failed to parse slave ports for bonded device %s", + name); return -1; } for (i = 0; i < slave_ports.slave_count; i++) { if (rte_eth_bond_slave_add(port_id, slave_ports.slaves[i]) != 0) { - RTE_LOG(ERR, EAL, - "Failed to add port %d as slave to bonded device %s\n", - slave_ports.slaves[i], name); + RTE_BOND_LOG(ERR, + "Failed to add port %d as slave to bonded device %s", + slave_ports.slaves[i], name); } } } else { - RTE_LOG(INFO, EAL, "No slaves specified for bonded device %s\n", name); + RTE_BOND_LOG(INFO, "No slaves specified for bonded device %s", name); return -1; } @@ -3305,27 +3508,27 @@ bond_ethdev_configure(struct rte_eth_dev *dev) uint16_t primary_slave_port_id; if (rte_kvargs_process(kvlist, - PMD_BOND_PRIMARY_SLAVE_KVARG, - &bond_ethdev_parse_primary_slave_port_id_kvarg, - &primary_slave_port_id) < 0) { - RTE_LOG(INFO, EAL, - "Invalid primary slave port id specified for bonded device" - " %s\n", name); + PMD_BOND_PRIMARY_SLAVE_KVARG, + &bond_ethdev_parse_primary_slave_port_id_kvarg, + &primary_slave_port_id) < 0) { + RTE_BOND_LOG(INFO, + "Invalid primary slave port id specified for bonded device %s", + name); return -1; } /* Set balance mode transmit policy*/ if (rte_eth_bond_primary_set(port_id, primary_slave_port_id) - != 0) { - RTE_LOG(ERR, EAL, - "Failed to set primary slave port %d on bonded device %s\n", - primary_slave_port_id, name); + != 0) { + RTE_BOND_LOG(ERR, + "Failed to set primary slave port %d on bonded device %s", + primary_slave_port_id, name); return -1; } } else if (arg_count > 1) { - RTE_LOG(INFO, EAL, - "Primary slave can be specified only once for bonded device" - " %s\n", name); + RTE_BOND_LOG(INFO, + "Primary slave can be specified only once for bonded device %s", + name); return -1; } @@ -3335,26 +3538,26 @@ bond_ethdev_configure(struct rte_eth_dev *dev) uint32_t lsc_poll_interval_ms; if (rte_kvargs_process(kvlist, - PMD_BOND_LSC_POLL_PERIOD_KVARG, - &bond_ethdev_parse_time_ms_kvarg, - &lsc_poll_interval_ms) < 0) { - RTE_LOG(INFO, EAL, - "Invalid lsc polling interval value specified for bonded" - " device %s\n", name); + PMD_BOND_LSC_POLL_PERIOD_KVARG, + &bond_ethdev_parse_time_ms_kvarg, + &lsc_poll_interval_ms) < 0) { + RTE_BOND_LOG(INFO, + "Invalid lsc polling interval value specified for bonded" + " device %s", name); return -1; } if (rte_eth_bond_link_monitoring_set(port_id, lsc_poll_interval_ms) - != 0) { - RTE_LOG(ERR, EAL, - "Failed to set lsc monitor polling interval (%u ms) on" - " bonded device %s\n", lsc_poll_interval_ms, name); + != 0) { + RTE_BOND_LOG(ERR, + "Failed to set lsc monitor polling interval (%u ms) on bonded device %s", + lsc_poll_interval_ms, name); return -1; } } else if (arg_count > 1) { - RTE_LOG(INFO, EAL, - "LSC polling interval can be specified only once for bonded" - " device %s\n", name); + RTE_BOND_LOG(INFO, + "LSC polling interval can be specified only once for bonded" + " device %s", name); return -1; } @@ -3364,27 +3567,27 @@ bond_ethdev_configure(struct rte_eth_dev *dev) uint32_t link_up_delay_ms; if (rte_kvargs_process(kvlist, - PMD_BOND_LINK_UP_PROP_DELAY_KVARG, - &bond_ethdev_parse_time_ms_kvarg, - &link_up_delay_ms) < 0) { - RTE_LOG(INFO, EAL, - "Invalid link up propagation delay value specified for" - " bonded device %s\n", name); + PMD_BOND_LINK_UP_PROP_DELAY_KVARG, + &bond_ethdev_parse_time_ms_kvarg, + &link_up_delay_ms) < 0) { + RTE_BOND_LOG(INFO, + "Invalid link up propagation delay value specified for" + " bonded device %s", name); return -1; } /* Set balance mode transmit policy*/ if (rte_eth_bond_link_up_prop_delay_set(port_id, link_up_delay_ms) - != 0) { - RTE_LOG(ERR, EAL, - "Failed to set link up propagation delay (%u ms) on bonded" - " device %s\n", link_up_delay_ms, name); + != 0) { + RTE_BOND_LOG(ERR, + "Failed to set link up propagation delay (%u ms) on bonded" + " device %s", link_up_delay_ms, name); return -1; } } else if (arg_count > 1) { - RTE_LOG(INFO, EAL, - "Link up propagation delay can be specified only once for" - " bonded device %s\n", name); + RTE_BOND_LOG(INFO, + "Link up propagation delay can be specified only once for" + " bonded device %s", name); return -1; } @@ -3394,27 +3597,27 @@ bond_ethdev_configure(struct rte_eth_dev *dev) uint32_t link_down_delay_ms; if (rte_kvargs_process(kvlist, - PMD_BOND_LINK_DOWN_PROP_DELAY_KVARG, - &bond_ethdev_parse_time_ms_kvarg, - &link_down_delay_ms) < 0) { - RTE_LOG(INFO, EAL, - "Invalid link down propagation delay value specified for" - " bonded device %s\n", name); + PMD_BOND_LINK_DOWN_PROP_DELAY_KVARG, + &bond_ethdev_parse_time_ms_kvarg, + &link_down_delay_ms) < 0) { + RTE_BOND_LOG(INFO, + "Invalid link down propagation delay value specified for" + " bonded device %s", name); return -1; } /* Set balance mode transmit policy*/ if (rte_eth_bond_link_down_prop_delay_set(port_id, link_down_delay_ms) - != 0) { - RTE_LOG(ERR, EAL, - "Failed to set link down propagation delay (%u ms) on" - " bonded device %s\n", link_down_delay_ms, name); + != 0) { + RTE_BOND_LOG(ERR, + "Failed to set link down propagation delay (%u ms) on bonded device %s", + link_down_delay_ms, name); return -1; } } else if (arg_count > 1) { - RTE_LOG(INFO, EAL, - "Link down propagation delay can be specified only once for" - " bonded device %s\n", name); + RTE_BOND_LOG(INFO, + "Link down propagation delay can be specified only once for bonded device %s", + name); return -1; } @@ -3440,3 +3643,12 @@ RTE_PMD_REGISTER_PARAM_STRING(net_bonding, "lsc_poll_period_ms= " "up_delay= " "down_delay="); + +int bond_logtype; + +RTE_INIT(bond_init_log) +{ + bond_logtype = rte_log_register("pmd.net.bond"); + if (bond_logtype >= 0) + rte_log_set_level(bond_logtype, RTE_LOG_NOTICE); +}