Openvswitch原理与代码分析：用户态流表flow table的操作

Posted 2020-08-09 刘超的通俗云计算

当内核无法查找到流表项的时候，则会通过upcall来调用用户态ovs-vswtichd中的flow table。

会调用ofproto-dpif-upcall.c中的udpif_upcall_handler函数。

static void * udpif_upcall_handler(void *arg) { ????struct handler *handler = arg; ????struct udpif *udpif = handler->udpif; ? ????while (!latch_is_set(&handler->udpif->exit_latch)) { ????????if (recv_upcalls(handler)) { ????????????poll_immediate_wake(); ????????} else { ????????????dpif_recv_wait(udpif->dpif, handler->handler_id); ????????????latch_wait(&udpif->exit_latch); ????????} ????????poll_block(); ????} ? ????return NULL; }

?

会调用static size_t recv_upcalls(struct handler *handler)

在这个函数里面

(1) 首先读取upcall调用static int upcall_receive(struct upcall *upcall, const struct dpif_backer *backer, const struct dp_packet *packet, enum dpif_upcall_type type, const struct nlattr *userdata, const struct flow *flow, const unsigned int mru, const ovs_u128 *ufid, const unsigned pmd_id)

(2) 其次提取包头调用void flow_extract(struct dp_packet *packet, struct flow *flow)，提取出的flow如下：

struct flow { ????/* Metadata */ ????struct flow_tnl tunnel; /* Encapsulating tunnel parameters. */ ????ovs_be64 metadata; /* OpenFlow Metadata. */ ????uint32_t regs[FLOW_N_REGS]; /* Registers. */ ????uint32_t skb_priority; /* Packet priority for QoS. */ ????uint32_t pkt_mark; /* Packet mark. */ ????uint32_t dp_hash; /* Datapath computed hash value. The exact ?????????????????????????????????* computation is opaque to the user space. */ ????union flow_in_port in_port; /* Input port.*/ ????uint32_t recirc_id; /* Must be exact match. */ ????uint16_t ct_state; /* Connection tracking state. */ ????uint16_t ct_zone; /* Connection tracking zone. */ ????uint32_t ct_mark; /* Connection mark.*/ ????uint8_t pad1[4]; /* Pad to 64 bits. */ ????ovs_u128 ct_label; /* Connection label. */ ????uint32_t conj_id; /* Conjunction ID. */ ????ofp_port_t actset_output; /* Output port in action set. */ ????uint8_t pad2[2]; /* Pad to 64 bits. */ ? ????/* L2, Order the same as in the Ethernet header! (64-bit aligned) */ ????struct eth_addr dl_dst; /* Ethernet destination address. */ ????struct eth_addr dl_src; /* Ethernet source address. */ ????ovs_be16 dl_type; /* Ethernet frame type. */ ????ovs_be16 vlan_tci; /* If 802.1Q, TCI | VLAN_CFI; otherwise 0. */ ????ovs_be32 mpls_lse[ROUND_UP(FLOW_MAX_MPLS_LABELS, 2)]; /* MPLS label stack ?????????????????????????????????????????????????????????????(with padding). */ ????/* L3 (64-bit aligned) */ ????ovs_be32 nw_src; /* IPv4 source address. */ ????ovs_be32 nw_dst; /* IPv4 destination address. */ ????struct in6_addr ipv6_src; /* IPv6 source address. */ ????struct in6_addr ipv6_dst; /* IPv6 destination address. */ ????ovs_be32 ipv6_label; /* IPv6 flow label. */ ????uint8_t nw_frag; /* FLOW_FRAG_* flags. */ ????uint8_t nw_tos; /* IP ToS (including DSCP and ECN). */ ????uint8_t nw_ttl; /* IP TTL/Hop Limit. */ ????uint8_t nw_proto; /* IP protocol or low 8 bits of ARP opcode. */ ????struct in6_addr nd_target; /* IPv6 neighbor discovery (ND) target. */ ????struct eth_addr arp_sha; /* ARP/ND source hardware address. */ ????struct eth_addr arp_tha; /* ARP/ND target hardware address. */ ????ovs_be16 tcp_flags; /* TCP flags. With L3 to avoid matching L4. */ ????ovs_be16 pad3; /* Pad to 64 bits. */ ? ????/* L4 (64-bit aligned) */ ????ovs_be16 tp_src; /* TCP/UDP/SCTP source port/ICMP type. */ ????ovs_be16 tp_dst; /* TCP/UDP/SCTP destination port/ICMP code. */ ????ovs_be32 igmp_group_ip4; /* IGMP group IPv4 address. ?????????????????????????????????* Keep last for BUILD_ASSERT_DECL below. */ };

?

(3) 然后调用static int process_upcall(struct udpif *udpif, struct upcall *upcall, struct ofpbuf *odp_actions, struct flow_wildcards *wc)来处理upcall。

?

对于MISS_UPCALL，调用static void upcall_xlate(struct udpif *udpif, struct upcall *upcall, struct ofpbuf *odp_actions, struct flow_wildcards *wc)

switch (classify_upcall(upcall->type, userdata)) { case MISS_UPCALL: ????upcall_xlate(udpif, upcall, odp_actions, wc); ????return 0;

?

会调用enum xlate_error xlate_actions(struct xlate_in *xin, struct xlate_out *xout)

在这个函数里面，会在flow table里面查找rule

ctx.rule = rule_dpif_lookup_from_table( ctx.xbridge->ofproto, ctx.tables_version, flow, xin->wc, ctx.xin->resubmit_stats, &ctx.table_id, flow->in_port.ofp_port, true, true);

找到rule之后，调用static void do_xlate_actions(const struct ofpact *ofpacts, size_t ofpacts_len, struct xlate_ctx *ctx)在这个函数里面，根据action的不同，修改flow的内容。

switch (a->type) { case OFPACT_OUTPUT: ????xlate_output_action(ctx, ofpact_get_OUTPUT(a)->port, ????????????????????????ofpact_get_OUTPUT(a)->max_len, true); ????break; ? case OFPACT_SET_VLAN_VID: ????wc->masks.vlan_tci |= htons(VLAN_VID_MASK | VLAN_CFI); ????if (flow->vlan_tci & htons(VLAN_CFI) || ????????ofpact_get_SET_VLAN_VID(a)->push_vlan_if_needed) { ????????flow->vlan_tci &= ~htons(VLAN_VID_MASK); ????????flow->vlan_tci |= (htons(ofpact_get_SET_VLAN_VID(a)->vlan_vid) ???????????????????????????| htons(VLAN_CFI)); ????} ????break; ? case OFPACT_SET_ETH_SRC: ????WC_MASK_FIELD(wc, dl_src); ????flow->dl_src = ofpact_get_SET_ETH_SRC(a)->mac; ????break; ? case OFPACT_SET_ETH_DST: ????WC_MASK_FIELD(wc, dl_dst); ????flow->dl_dst = ofpact_get_SET_ETH_DST(a)->mac; ????break; ? case OFPACT_SET_IPV4_SRC: ????CHECK_MPLS_RECIRCULATION(); ????if (flow->dl_type == htons(ETH_TYPE_IP)) { ????????memset(&wc->masks.nw_src, 0xff, sizeof wc->masks.nw_src); ????????flow->nw_src = ofpact_get_SET_IPV4_SRC(a)->ipv4; ????} ????break; ? case OFPACT_SET_IPV4_DST: ????CHECK_MPLS_RECIRCULATION(); ????if (flow->dl_type == htons(ETH_TYPE_IP)) { ????????memset(&wc->masks.nw_dst, 0xff, sizeof wc->masks.nw_dst); ????????flow->nw_dst = ofpact_get_SET_IPV4_DST(a)->ipv4; ????} ????break; ? case OFPACT_SET_L4_SRC_PORT: ????CHECK_MPLS_RECIRCULATION(); ????if (is_ip_any(flow) && !(flow->nw_frag & FLOW_NW_FRAG_LATER)) { ????????memset(&wc->masks.nw_proto, 0xff, sizeof wc->masks.nw_proto); ????????memset(&wc->masks.tp_src, 0xff, sizeof wc->masks.tp_src); ????????flow->tp_src = htons(ofpact_get_SET_L4_SRC_PORT(a)->port); ????} ????break; ? case OFPACT_SET_L4_DST_PORT: ????CHECK_MPLS_RECIRCULATION(); ????if (is_ip_any(flow) && !(flow->nw_frag & FLOW_NW_FRAG_LATER)) { ????????memset(&wc->masks.nw_proto, 0xff, sizeof wc->masks.nw_proto); ????????memset(&wc->masks.tp_dst, 0xff, sizeof wc->masks.tp_dst); ????????flow->tp_dst = htons(ofpact_get_SET_L4_DST_PORT(a)->port); ????} ????break;

?

?

(4) 最后调用static void handle_upcalls(struct udpif *udpif, struct upcall *upcalls, size_t n_upcalls)将flow rule添加到内核中的datapath

他会调用void dpif_operate(struct dpif *dpif, struct dpif_op **ops, size_t n_ops)，他会调用dpif->dpif_class->operate(dpif, ops, chunk);

会调用dpif_netlink_operate()

static void dpif_netlink_operate(struct dpif *dpif_, struct dpif_op **ops, size_t n_ops) { ????struct dpif_netlink *dpif = dpif_netlink_cast(dpif_); ? ????while (n_ops > 0) { ????????size_t chunk = dpif_netlink_operate__(dpif, ops, n_ops); ????????ops += chunk; ????????n_ops -= chunk; ????} }

?

在static size_t dpif_netlink_operate__(struct dpif_netlink *dpif, struct dpif_op **ops, size_t n_ops)中，有以下的代码：

switch (op->type) { case DPIF_OP_FLOW_PUT: ????put = &op->u.flow_put; ????dpif_netlink_init_flow_put(dpif, put, &flow); ????if (put->stats) { ????????flow.nlmsg_flags |= NLM_F_ECHO; ????????aux->txn.reply = &aux->reply; ????} ????dpif_netlink_flow_to_ofpbuf(&flow, &aux->request); ????break; ? case DPIF_OP_FLOW_DEL: ????del = &op->u.flow_del; ????dpif_netlink_init_flow_del(dpif, del, &flow); ????if (del->stats) { ????????flow.nlmsg_flags |= NLM_F_ECHO; ????????aux->txn.reply = &aux->reply; ????} ????dpif_netlink_flow_to_ofpbuf(&flow, &aux->request); ????break;

?

会调用netlink修改内核中datapath的规则。