X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Fmlx5%2Fmlx5_flow_tcf.c;h=e021ea5e043827e589e22d167530090bf2a4b2c3;hb=a7c528e5d71ff3f569898d268f9de129fdfc152b;hp=4d541123c8e9573d8e332ab132a069b6de2316b5;hpb=a57f692be036a986e2fda719e707a88c0f4e33c3;p=dpdk.git diff --git a/drivers/net/mlx5/mlx5_flow_tcf.c b/drivers/net/mlx5/mlx5_flow_tcf.c index 4d541123c8..e021ea5e04 100644 --- a/drivers/net/mlx5/mlx5_flow_tcf.c +++ b/drivers/net/mlx5/mlx5_flow_tcf.c @@ -28,6 +28,7 @@ #include #include #include +#include #include "mlx5.h" #include "mlx5_flow.h" @@ -125,6 +126,14 @@ struct tc_pedit_sel { #define TCA_TUNNEL_KEY_NO_CSUM 10 #endif +#ifndef HAVE_TCA_TUNNEL_KEY_ENC_TOS +#define TCA_TUNNEL_KEY_ENC_TOS 12 +#endif + +#ifndef HAVE_TCA_TUNNEL_KEY_ENC_TTL +#define TCA_TUNNEL_KEY_ENC_TTL 13 +#endif + #else /* HAVE_TC_ACT_TUNNEL_KEY */ #define TCA_ACT_TUNNEL_KEY 17 @@ -138,6 +147,8 @@ struct tc_pedit_sel { #define TCA_TUNNEL_KEY_ENC_KEY_ID 7 #define TCA_TUNNEL_KEY_ENC_DST_PORT 9 #define TCA_TUNNEL_KEY_NO_CSUM 10 +#define TCA_TUNNEL_KEY_ENC_TOS 12 +#define TCA_TUNNEL_KEY_ENC_TTL 13 struct tc_tunnel_key { tc_gen; @@ -160,6 +171,9 @@ struct tc_tunnel_key { #ifndef TCA_CLS_FLAGS_SKIP_SW #define TCA_CLS_FLAGS_SKIP_SW (1 << 1) #endif +#ifndef TCA_CLS_FLAGS_IN_HW +#define TCA_CLS_FLAGS_IN_HW (1 << 2) +#endif #ifndef HAVE_TCA_CHAIN #define TCA_CHAIN 11 #endif @@ -289,6 +303,31 @@ struct tc_tunnel_key { #ifndef HAVE_TCA_FLOWER_KEY_TCP_FLAGS_MASK #define TCA_FLOWER_KEY_TCP_FLAGS_MASK 72 #endif +#ifndef HAVE_TCA_FLOWER_KEY_IP_TOS +#define TCA_FLOWER_KEY_IP_TOS 73 +#endif +#ifndef HAVE_TCA_FLOWER_KEY_IP_TOS_MASK +#define TCA_FLOWER_KEY_IP_TOS_MASK 74 +#endif +#ifndef HAVE_TCA_FLOWER_KEY_IP_TTL +#define TCA_FLOWER_KEY_IP_TTL 75 +#endif +#ifndef HAVE_TCA_FLOWER_KEY_IP_TTL_MASK +#define TCA_FLOWER_KEY_IP_TTL_MASK 76 +#endif +#ifndef HAVE_TCA_FLOWER_KEY_ENC_IP_TOS +#define TCA_FLOWER_KEY_ENC_IP_TOS 80 +#endif +#ifndef HAVE_TCA_FLOWER_KEY_ENC_IP_TOS_MASK +#define TCA_FLOWER_KEY_ENC_IP_TOS_MASK 81 +#endif +#ifndef HAVE_TCA_FLOWER_KEY_ENC_IP_TTL +#define TCA_FLOWER_KEY_ENC_IP_TTL 82 +#endif +#ifndef HAVE_TCA_FLOWER_KEY_ENC_IP_TTL_MASK +#define TCA_FLOWER_KEY_ENC_IP_TTL_MASK 83 +#endif + #ifndef HAVE_TC_ACT_GOTO_CHAIN #define TC_ACT_GOTO_CHAIN 0x20000000 #endif @@ -313,10 +352,14 @@ struct tc_tunnel_key { #define TCA_ACT_MAX_PRIO 32 #endif -/** UDP port range of VXLAN devices created by driver. */ -#define MLX5_VXLAN_PORT_MIN 30000 -#define MLX5_VXLAN_PORT_MAX 60000 +/** Parameters of VXLAN devices created by driver. */ +#define MLX5_VXLAN_DEFAULT_VNI 1 #define MLX5_VXLAN_DEVICE_PFX "vmlx_" +/** + * Timeout in milliseconds to wait VXLAN UDP offloaded port + * registration completed within the mlx5 driver. + */ +#define MLX5_VXLAN_WAIT_PORT_REG_MS 250 /** Tunnel action type, used for @p type in header structure. */ enum flow_tcf_tunact_type { @@ -334,6 +377,8 @@ enum flow_tcf_tunact_type { #define FLOW_TCF_ENCAP_UDP_SRC (1u << 6) #define FLOW_TCF_ENCAP_UDP_DST (1u << 7) #define FLOW_TCF_ENCAP_VXLAN_VNI (1u << 8) +#define FLOW_TCF_ENCAP_IP_TTL (1u << 9) +#define FLOW_TCF_ENCAP_IP_TOS (1u << 10) /** * Structure for holding netlink context. @@ -348,6 +393,110 @@ struct mlx5_flow_tcf_context { uint8_t *buf; /* Message buffer. */ }; +/** + * Neigh rule structure. The neigh rule is applied via Netlink to + * outer tunnel iface in order to provide destination MAC address + * for the VXLAN encapsultion. The neigh rule is implicitly related + * to the Flow itself and can be shared by multiple Flows. + */ +struct tcf_neigh_rule { + LIST_ENTRY(tcf_neigh_rule) next; + uint32_t refcnt; + struct rte_ether_addr eth; + uint16_t mask; + union { + struct { + rte_be32_t dst; + } ipv4; + struct { + uint8_t dst[IPV6_ADDR_LEN]; + } ipv6; + }; +}; + +/** + * Local rule structure. The local rule is applied via Netlink to + * outer tunnel iface in order to provide local and peer IP addresses + * of the VXLAN tunnel for encapsulation. The local rule is implicitly + * related to the Flow itself and can be shared by multiple Flows. + */ +struct tcf_local_rule { + LIST_ENTRY(tcf_local_rule) next; + uint32_t refcnt; + uint16_t mask; + union { + struct { + rte_be32_t dst; + rte_be32_t src; + } ipv4; + struct { + uint8_t dst[IPV6_ADDR_LEN]; + uint8_t src[IPV6_ADDR_LEN]; + } ipv6; + }; +}; + +/** Outer interface VXLAN encapsulation rules container. */ +struct tcf_irule { + LIST_ENTRY(tcf_irule) next; + LIST_HEAD(, tcf_neigh_rule) neigh; + LIST_HEAD(, tcf_local_rule) local; + uint32_t refcnt; + unsigned int ifouter; /**< Own interface index. */ +}; + +/** VXLAN virtual netdev. */ +struct tcf_vtep { + LIST_ENTRY(tcf_vtep) next; + uint32_t refcnt; + unsigned int ifindex; /**< Own interface index. */ + uint16_t port; + uint32_t created:1; /**< Actually created by PMD. */ + uint32_t waitreg:1; /**< Wait for VXLAN UDP port registration. */ +}; + +/** Tunnel descriptor header, common for all tunnel types. */ +struct flow_tcf_tunnel_hdr { + uint32_t type; /**< Tunnel action type. */ + struct tcf_vtep *vtep; /**< Virtual tunnel endpoint device. */ + unsigned int ifindex_org; /**< Original dst/src interface */ + unsigned int *ifindex_ptr; /**< Interface ptr in message. */ +}; + +struct flow_tcf_vxlan_decap { + struct flow_tcf_tunnel_hdr hdr; + uint16_t udp_port; +}; + +struct flow_tcf_vxlan_encap { + struct flow_tcf_tunnel_hdr hdr; + struct tcf_irule *iface; + uint32_t mask; + uint8_t ip_tos; + uint8_t ip_ttl_hop; + struct { + struct rte_ether_addr dst; + struct rte_ether_addr src; + } eth; + union { + struct { + rte_be32_t dst; + rte_be32_t src; + } ipv4; + struct { + uint8_t dst[IPV6_ADDR_LEN]; + uint8_t src[IPV6_ADDR_LEN]; + } ipv6; + }; + struct { + rte_be16_t src; + rte_be16_t dst; + } udp; + struct { + uint8_t vni[3]; + } vxlan; +}; + /** Structure used when extracting the values of a flow counters * from a netlink message. */ @@ -365,7 +514,10 @@ static const union { struct rte_flow_item_ipv6 ipv6; struct rte_flow_item_tcp tcp; struct rte_flow_item_udp udp; -} flow_tcf_mask_empty; + struct rte_flow_item_vxlan vxlan; +} flow_tcf_mask_empty = { + {0}, +}; /** Supported masks for known item types. */ static const struct { @@ -376,6 +528,7 @@ static const struct { struct rte_flow_item_ipv6 ipv6; struct rte_flow_item_tcp tcp; struct rte_flow_item_udp udp; + struct rte_flow_item_vxlan vxlan; } flow_tcf_mask_supported = { .port_id = { .id = 0xffffffff, @@ -392,11 +545,15 @@ static const struct { }, .ipv4.hdr = { .next_proto_id = 0xff, + .time_to_live = 0xff, + .type_of_service = 0xff, .src_addr = RTE_BE32(0xffffffff), .dst_addr = RTE_BE32(0xffffffff), }, .ipv6.hdr = { .proto = 0xff, + .vtc_flow = RTE_BE32(0xfful << IPV6_HDR_FL_SHIFT), + .hop_limits = 0xff, .src_addr = "\xff\xff\xff\xff\xff\xff\xff\xff" "\xff\xff\xff\xff\xff\xff\xff\xff", @@ -413,6 +570,9 @@ static const struct { .src_port = RTE_BE16(0xffff), .dst_port = RTE_BE16(0xffff), }, + .vxlan = { + .vni = "\xff\xff\xff", + }, }; #define SZ_NLATTR_HDR MNL_ALIGN(sizeof(struct nlattr)) @@ -431,7 +591,15 @@ struct flow_tcf_ptoi { /* Due to a limitation on driver/FW. */ #define MLX5_TCF_GROUP_ID_MAX 3 -#define MLX5_TCF_GROUP_PRIORITY_MAX 14 + +/* + * Due to a limitation on driver/FW, priority ranges from 1 to 16 in kernel. + * Priority in rte_flow attribute starts from 0 and is added by 1 in + * translation. This is subject to be changed to determine the max priority + * based on trial-and-error like Verbs driver once the restriction is lifted or + * the range is extended. + */ +#define MLX5_TCF_GROUP_PRIORITY_MAX 15 #define MLX5_TCF_FATE_ACTIONS \ (MLX5_FLOW_ACTION_DROP | MLX5_FLOW_ACTION_PORT_ID | \ @@ -521,8 +689,8 @@ flow_tcf_pedit_key_set_mac(const struct rte_flow_action *actions, { int idx = p_parser->sel.nkeys; uint32_t off = actions->type == RTE_FLOW_ACTION_TYPE_SET_MAC_SRC ? - offsetof(struct ether_hdr, s_addr) : - offsetof(struct ether_hdr, d_addr); + offsetof(struct rte_ether_hdr, s_addr) : + offsetof(struct rte_ether_hdr, d_addr); const struct rte_flow_action_set_mac *conf = (const struct rte_flow_action_set_mac *)actions->conf; @@ -539,7 +707,7 @@ flow_tcf_pedit_key_set_mac(const struct rte_flow_action *actions, p_parser->keys_ex[idx].cmd = TCA_PEDIT_KEY_EX_CMD_SET; memcpy(&p_parser->keys[idx].val, conf->mac_addr + SZ_PEDIT_KEY_VAL, - ETHER_ADDR_LEN - SZ_PEDIT_KEY_VAL); + RTE_ETHER_ADDR_LEN - SZ_PEDIT_KEY_VAL); p_parser->sel.nkeys = (++idx); } @@ -564,12 +732,12 @@ flow_tcf_pedit_key_set_dec_ttl(const struct rte_flow_action *actions, if (item_flags & MLX5_FLOW_LAYER_OUTER_L3_IPV4) { p_parser->keys_ex[idx].htype = TCA_PEDIT_KEY_EX_HDR_TYPE_IP4; p_parser->keys[idx].off = - offsetof(struct ipv4_hdr, time_to_live); + offsetof(struct rte_ipv4_hdr, time_to_live); } if (item_flags & MLX5_FLOW_LAYER_OUTER_L3_IPV6) { p_parser->keys_ex[idx].htype = TCA_PEDIT_KEY_EX_HDR_TYPE_IP6; p_parser->keys[idx].off = - offsetof(struct ipv6_hdr, hop_limits); + offsetof(struct rte_ipv6_hdr, hop_limits); } if (actions->type == RTE_FLOW_ACTION_TYPE_DEC_TTL) { p_parser->keys_ex[idx].cmd = TCA_PEDIT_KEY_EX_CMD_ADD; @@ -633,8 +801,8 @@ flow_tcf_pedit_key_set_ipv6_addr(const struct rte_flow_action *actions, int keys = NUM_OF_PEDIT_KEYS(IPV6_ADDR_LEN); int off_base = actions->type == RTE_FLOW_ACTION_TYPE_SET_IPV6_SRC ? - offsetof(struct ipv6_hdr, src_addr) : - offsetof(struct ipv6_hdr, dst_addr); + offsetof(struct rte_ipv6_hdr, src_addr) : + offsetof(struct rte_ipv6_hdr, dst_addr); const struct rte_flow_action_set_ipv6 *conf = (const struct rte_flow_action_set_ipv6 *)actions->conf; @@ -668,8 +836,8 @@ flow_tcf_pedit_key_set_ipv4_addr(const struct rte_flow_action *actions, p_parser->keys_ex[idx].cmd = TCA_PEDIT_KEY_EX_CMD_SET; p_parser->keys[idx].off = actions->type == RTE_FLOW_ACTION_TYPE_SET_IPV4_SRC ? - offsetof(struct ipv4_hdr, src_addr) : - offsetof(struct ipv4_hdr, dst_addr); + offsetof(struct rte_ipv4_hdr, src_addr) : + offsetof(struct rte_ipv4_hdr, dst_addr); p_parser->keys[idx].mask = ~UINT32_MAX; p_parser->keys[idx].val = ((const struct rte_flow_action_set_ipv4 *) @@ -816,11 +984,11 @@ flow_tcf_get_pedit_actions_size(const struct rte_flow_action **actions, flags |= MLX5_FLOW_ACTION_DEC_TTL; break; case RTE_FLOW_ACTION_TYPE_SET_MAC_SRC: - keys += NUM_OF_PEDIT_KEYS(ETHER_ADDR_LEN); + keys += NUM_OF_PEDIT_KEYS(RTE_ETHER_ADDR_LEN); flags |= MLX5_FLOW_ACTION_SET_MAC_SRC; break; case RTE_FLOW_ACTION_TYPE_SET_MAC_DST: - keys += NUM_OF_PEDIT_KEYS(ETHER_ADDR_LEN); + keys += NUM_OF_PEDIT_KEYS(RTE_ETHER_ADDR_LEN); flags |= MLX5_FLOW_ACTION_SET_MAC_DST; break; default: @@ -992,19 +1160,13 @@ flow_tcf_validate_attributes(const struct rte_flow_attr *attr, "group ID larger than " RTE_STR(MLX5_TCF_GROUP_ID_MAX) " isn't supported"); - else if (attr->group > 0 && - attr->priority > MLX5_TCF_GROUP_PRIORITY_MAX) + else if (attr->priority > MLX5_TCF_GROUP_PRIORITY_MAX) return rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ATTR_PRIORITY, attr, - "lowest priority level is " + "priority more than " RTE_STR(MLX5_TCF_GROUP_PRIORITY_MAX) - " when group is configured"); - else if (attr->priority > 0xfffe) - return rte_flow_error_set(error, ENOTSUP, - RTE_FLOW_ERROR_TYPE_ATTR_PRIORITY, - attr, - "lowest priority level is 0xfffe"); + " is not supported"); if (!attr->ingress) return rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ATTR_INGRESS, @@ -1016,6 +1178,563 @@ flow_tcf_validate_attributes(const struct rte_flow_attr *attr, return 0; } +/** + * Validate VXLAN_ENCAP action RTE_FLOW_ITEM_TYPE_ETH item for E-Switch. + * The routine checks the L2 fields to be used in encapsulation header. + * + * @param[in] item + * Pointer to the item structure. + * @param[out] error + * Pointer to the error structure. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + **/ +static int +flow_tcf_validate_vxlan_encap_eth(const struct rte_flow_item *item, + struct rte_flow_error *error) +{ + const struct rte_flow_item_eth *spec = item->spec; + const struct rte_flow_item_eth *mask = item->mask; + + if (!spec) { + /* + * Specification for L2 addresses can be empty + * because these ones are optional and not + * required directly by tc rule. Kernel tries + * to resolve these ones on its own + */ + return 0; + } + if (!mask) { + /* If mask is not specified use the default one. */ + mask = &rte_flow_item_eth_mask; + } + if (memcmp(&mask->dst, + &flow_tcf_mask_empty.eth.dst, + sizeof(flow_tcf_mask_empty.eth.dst))) { + if (memcmp(&mask->dst, + &rte_flow_item_eth_mask.dst, + sizeof(rte_flow_item_eth_mask.dst))) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"eth.dst\" field"); + } + if (memcmp(&mask->src, + &flow_tcf_mask_empty.eth.src, + sizeof(flow_tcf_mask_empty.eth.src))) { + if (memcmp(&mask->src, + &rte_flow_item_eth_mask.src, + sizeof(rte_flow_item_eth_mask.src))) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"eth.src\" field"); + } + if (mask->type != RTE_BE16(0x0000)) { + if (mask->type != RTE_BE16(0xffff)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"eth.type\" field"); + DRV_LOG(WARNING, + "outer ethernet type field" + " cannot be forced for vxlan" + " encapsulation, parameter ignored"); + } + return 0; +} + +/** + * Validate VXLAN_ENCAP action RTE_FLOW_ITEM_TYPE_IPV4 item for E-Switch. + * The routine checks the IPv4 fields to be used in encapsulation header. + * + * @param[in] item + * Pointer to the item structure. + * @param[out] error + * Pointer to the error structure. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + **/ +static int +flow_tcf_validate_vxlan_encap_ipv4(const struct rte_flow_item *item, + struct rte_flow_error *error) +{ + const struct rte_flow_item_ipv4 *spec = item->spec; + const struct rte_flow_item_ipv4 *mask = item->mask; + + if (!spec) { + /* + * Specification for IP addresses cannot be empty + * because it is required by tunnel_key parameter. + */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "NULL outer ipv4 address" + " specification for vxlan" + " encapsulation"); + } + if (!mask) + mask = &rte_flow_item_ipv4_mask; + if (mask->hdr.dst_addr != RTE_BE32(0x00000000)) { + if (mask->hdr.dst_addr != RTE_BE32(0xffffffff)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"ipv4.hdr.dst_addr\" field" + " for vxlan encapsulation"); + /* More IPv4 address validations can be put here. */ + } else { + /* + * Kernel uses the destination IP address to determine + * the routing path and obtain the MAC destination + * address, so IP destination address must be + * specified in the tc rule. + */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "outer ipv4 destination address" + " must be specified for" + " vxlan encapsulation"); + } + if (mask->hdr.src_addr != RTE_BE32(0x00000000)) { + if (mask->hdr.src_addr != RTE_BE32(0xffffffff)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"ipv4.hdr.src_addr\" field" + " for vxlan encapsulation"); + /* More IPv4 address validations can be put here. */ + } else { + /* + * Kernel uses the source IP address to select the + * interface for egress encapsulated traffic, so + * it must be specified in the tc rule. + */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "outer ipv4 source address" + " must be specified for" + " vxlan encapsulation"); + } + if (mask->hdr.type_of_service && + mask->hdr.type_of_service != 0xff) + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"ipv4.hdr.type_of_service\" field" + " for vxlan encapsulation"); + if (mask->hdr.time_to_live && + mask->hdr.time_to_live != 0xff) + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"ipv4.hdr.time_to_live\" field" + " for vxlan encapsulation"); + return 0; +} + +/** + * Validate VXLAN_ENCAP action RTE_FLOW_ITEM_TYPE_IPV6 item for E-Switch. + * The routine checks the IPv6 fields to be used in encapsulation header. + * + * @param[in] item + * Pointer to the item structure. + * @param[out] error + * Pointer to the error structure. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + **/ +static int +flow_tcf_validate_vxlan_encap_ipv6(const struct rte_flow_item *item, + struct rte_flow_error *error) +{ + const struct rte_flow_item_ipv6 *spec = item->spec; + const struct rte_flow_item_ipv6 *mask = item->mask; + uint8_t msk6; + + if (!spec) { + /* + * Specification for IP addresses cannot be empty + * because it is required by tunnel_key parameter. + */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "NULL outer ipv6 address" + " specification for" + " vxlan encapsulation"); + } + if (!mask) + mask = &rte_flow_item_ipv6_mask; + if (memcmp(&mask->hdr.dst_addr, + &flow_tcf_mask_empty.ipv6.hdr.dst_addr, + IPV6_ADDR_LEN)) { + if (memcmp(&mask->hdr.dst_addr, + &rte_flow_item_ipv6_mask.hdr.dst_addr, + IPV6_ADDR_LEN)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"ipv6.hdr.dst_addr\" field" + " for vxlan encapsulation"); + /* More IPv6 address validations can be put here. */ + } else { + /* + * Kernel uses the destination IP address to determine + * the routing path and obtain the MAC destination + * address (heigh or gate), so IP destination address + * must be specified within the tc rule. + */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "outer ipv6 destination address" + " must be specified for" + " vxlan encapsulation"); + } + if (memcmp(&mask->hdr.src_addr, + &flow_tcf_mask_empty.ipv6.hdr.src_addr, + IPV6_ADDR_LEN)) { + if (memcmp(&mask->hdr.src_addr, + &rte_flow_item_ipv6_mask.hdr.src_addr, + IPV6_ADDR_LEN)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"ipv6.hdr.src_addr\" field" + " for vxlan encapsulation"); + /* More L3 address validation can be put here. */ + } else { + /* + * Kernel uses the source IP address to select the + * interface for egress encapsulated traffic, so + * it must be specified in the tc rule. + */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "outer L3 source address" + " must be specified for" + " vxlan encapsulation"); + } + msk6 = (rte_be_to_cpu_32(mask->hdr.vtc_flow) >> + IPV6_HDR_TC_SHIFT) & 0xff; + if (msk6 && msk6 != 0xff) + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"ipv6.hdr.vtc_flow.tos\" field" + " for vxlan encapsulation"); + if (mask->hdr.hop_limits && mask->hdr.hop_limits != 0xff) + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"ipv6.hdr.hop_limits\" field" + " for vxlan encapsulation"); + return 0; +} + +/** + * Validate VXLAN_ENCAP action RTE_FLOW_ITEM_TYPE_UDP item for E-Switch. + * The routine checks the UDP fields to be used in encapsulation header. + * + * @param[in] item + * Pointer to the item structure. + * @param[out] error + * Pointer to the error structure. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + **/ +static int +flow_tcf_validate_vxlan_encap_udp(const struct rte_flow_item *item, + struct rte_flow_error *error) +{ + const struct rte_flow_item_udp *spec = item->spec; + const struct rte_flow_item_udp *mask = item->mask; + + if (!spec) { + /* + * Specification for UDP ports cannot be empty + * because it is required by tunnel_key parameter. + */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "NULL UDP port specification " + " for vxlan encapsulation"); + } + if (!mask) + mask = &rte_flow_item_udp_mask; + if (mask->hdr.dst_port != RTE_BE16(0x0000)) { + if (mask->hdr.dst_port != RTE_BE16(0xffff)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"udp.hdr.dst_port\" field" + " for vxlan encapsulation"); + if (!spec->hdr.dst_port) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "outer UDP remote port cannot be" + " 0 for vxlan encapsulation"); + } else { + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "outer UDP remote port" + " must be specified for" + " vxlan encapsulation"); + } + if (mask->hdr.src_port != RTE_BE16(0x0000)) { + if (mask->hdr.src_port != RTE_BE16(0xffff)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"udp.hdr.src_port\" field" + " for vxlan encapsulation"); + DRV_LOG(WARNING, + "outer UDP source port cannot be" + " forced for vxlan encapsulation," + " parameter ignored"); + } + return 0; +} + +/** + * Validate VXLAN_ENCAP action RTE_FLOW_ITEM_TYPE_VXLAN item for E-Switch. + * The routine checks the VNIP fields to be used in encapsulation header. + * + * @param[in] item + * Pointer to the item structure. + * @param[out] error + * Pointer to the error structure. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + **/ +static int +flow_tcf_validate_vxlan_encap_vni(const struct rte_flow_item *item, + struct rte_flow_error *error) +{ + const struct rte_flow_item_vxlan *spec = item->spec; + const struct rte_flow_item_vxlan *mask = item->mask; + + if (!spec) { + /* Outer VNI is required by tunnel_key parameter. */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "NULL VNI specification" + " for vxlan encapsulation"); + } + if (!mask) + mask = &rte_flow_item_vxlan_mask; + if (!mask->vni[0] && !mask->vni[1] && !mask->vni[2]) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "outer VNI must be specified " + "for vxlan encapsulation"); + if (mask->vni[0] != 0xff || + mask->vni[1] != 0xff || + mask->vni[2] != 0xff) + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"vxlan.vni\" field"); + + if (!spec->vni[0] && !spec->vni[1] && !spec->vni[2]) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, item, + "vxlan vni cannot be 0"); + return 0; +} + +/** + * Validate VXLAN_ENCAP action item list for E-Switch. + * The routine checks items to be used in encapsulation header. + * + * @param[in] action + * Pointer to the VXLAN_ENCAP action structure. + * @param[out] error + * Pointer to the error structure. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + **/ +static int +flow_tcf_validate_vxlan_encap(const struct rte_flow_action *action, + struct rte_flow_error *error) +{ + const struct rte_flow_item *items; + int ret; + uint32_t item_flags = 0; + + if (!action->conf) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, action, + "Missing vxlan tunnel" + " action configuration"); + items = ((const struct rte_flow_action_vxlan_encap *) + action->conf)->definition; + if (!items) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, action, + "Missing vxlan tunnel" + " encapsulation parameters"); + for (; items->type != RTE_FLOW_ITEM_TYPE_END; items++) { + switch (items->type) { + case RTE_FLOW_ITEM_TYPE_VOID: + break; + case RTE_FLOW_ITEM_TYPE_ETH: + ret = mlx5_flow_validate_item_eth(items, item_flags, + error); + if (ret < 0) + return ret; + ret = flow_tcf_validate_vxlan_encap_eth(items, error); + if (ret < 0) + return ret; + item_flags |= MLX5_FLOW_LAYER_OUTER_L2; + break; + break; + case RTE_FLOW_ITEM_TYPE_IPV4: + ret = mlx5_flow_validate_item_ipv4 + (items, item_flags, + &flow_tcf_mask_supported.ipv4, error); + if (ret < 0) + return ret; + ret = flow_tcf_validate_vxlan_encap_ipv4(items, error); + if (ret < 0) + return ret; + item_flags |= MLX5_FLOW_LAYER_OUTER_L3_IPV4; + break; + case RTE_FLOW_ITEM_TYPE_IPV6: + ret = mlx5_flow_validate_item_ipv6 + (items, item_flags, + &flow_tcf_mask_supported.ipv6, error); + if (ret < 0) + return ret; + ret = flow_tcf_validate_vxlan_encap_ipv6(items, error); + if (ret < 0) + return ret; + item_flags |= MLX5_FLOW_LAYER_OUTER_L3_IPV6; + break; + case RTE_FLOW_ITEM_TYPE_UDP: + ret = mlx5_flow_validate_item_udp(items, item_flags, + 0xFF, error); + if (ret < 0) + return ret; + ret = flow_tcf_validate_vxlan_encap_udp(items, error); + if (ret < 0) + return ret; + item_flags |= MLX5_FLOW_LAYER_OUTER_L4_UDP; + break; + case RTE_FLOW_ITEM_TYPE_VXLAN: + ret = mlx5_flow_validate_item_vxlan(items, + item_flags, error); + if (ret < 0) + return ret; + ret = flow_tcf_validate_vxlan_encap_vni(items, error); + if (ret < 0) + return ret; + item_flags |= MLX5_FLOW_LAYER_VXLAN; + break; + default: + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM, items, + "vxlan encap item not supported"); + } + } + if (!(item_flags & MLX5_FLOW_LAYER_OUTER_L3)) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, action, + "no outer IP layer found" + " for vxlan encapsulation"); + if (!(item_flags & MLX5_FLOW_LAYER_OUTER_L4_UDP)) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, action, + "no outer UDP layer found" + " for vxlan encapsulation"); + if (!(item_flags & MLX5_FLOW_LAYER_VXLAN)) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, action, + "no VXLAN VNI found" + " for vxlan encapsulation"); + return 0; +} + +/** + * Validate outer RTE_FLOW_ITEM_TYPE_UDP item if tunnel item + * RTE_FLOW_ITEM_TYPE_VXLAN is present in item list. + * + * @param[in] udp + * Outer UDP layer item (if any, NULL otherwise). + * @param[out] error + * Pointer to the error structure. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + **/ +static int +flow_tcf_validate_vxlan_decap_udp(const struct rte_flow_item *udp, + struct rte_flow_error *error) +{ + const struct rte_flow_item_udp *spec = udp->spec; + const struct rte_flow_item_udp *mask = udp->mask; + + if (!spec) + /* + * Specification for UDP ports cannot be empty + * because it is required as decap parameter. + */ + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, udp, + "NULL UDP port specification" + " for VXLAN decapsulation"); + if (!mask) + mask = &rte_flow_item_udp_mask; + if (mask->hdr.dst_port != RTE_BE16(0x0000)) { + if (mask->hdr.dst_port != RTE_BE16(0xffff)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"udp.hdr.dst_port\" field"); + if (!spec->hdr.dst_port) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, udp, + "zero decap local UDP port"); + } else { + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, udp, + "outer UDP destination port must be " + "specified for vxlan decapsulation"); + } + if (mask->hdr.src_port != RTE_BE16(0x0000)) { + if (mask->hdr.src_port != RTE_BE16(0xffff)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM_MASK, mask, + "no support for partial mask on" + " \"udp.hdr.src_port\" field"); + DRV_LOG(WARNING, + "outer UDP local port cannot be " + "forced for VXLAN encapsulation, " + "parameter ignored"); + } + return 0; +} + /** * Validate flow for E-Switch. * @@ -1031,7 +1750,7 @@ flow_tcf_validate_attributes(const struct rte_flow_attr *attr, * Pointer to the error structure. * * @return - * 0 on success, a negative errno value otherwise and rte_ernno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ static int flow_tcf_validate(struct rte_eth_dev *dev, @@ -1048,6 +1767,7 @@ flow_tcf_validate(struct rte_eth_dev *dev, const struct rte_flow_item_ipv6 *ipv6; const struct rte_flow_item_tcp *tcp; const struct rte_flow_item_udp *udp; + const struct rte_flow_item_vxlan *vxlan; } spec, mask; union { const struct rte_flow_action_port_id *port_id; @@ -1057,12 +1777,17 @@ flow_tcf_validate(struct rte_eth_dev *dev, of_set_vlan_vid; const struct rte_flow_action_of_set_vlan_pcp * of_set_vlan_pcp; + const struct rte_flow_action_vxlan_encap *vxlan_encap; const struct rte_flow_action_set_ipv4 *set_ipv4; const struct rte_flow_action_set_ipv6 *set_ipv6; } conf; + const struct rte_flow_item *outer_udp = NULL; + rte_be16_t inner_etype = RTE_BE16(ETH_P_ALL); + rte_be16_t outer_etype = RTE_BE16(ETH_P_ALL); + rte_be16_t vlan_etype = RTE_BE16(ETH_P_ALL); uint64_t item_flags = 0; uint64_t action_flags = 0; - uint8_t next_protocol = -1; + uint8_t next_protocol = 0xff; unsigned int tcm_ifindex = 0; uint8_t pedit_validated = 0; struct flow_tcf_ptoi ptoi[PTOI_TABLE_SZ_MAX(dev)]; @@ -1075,13 +1800,187 @@ flow_tcf_validate(struct rte_eth_dev *dev, ret = flow_tcf_validate_attributes(attr, error); if (ret < 0) return ret; - for (; items->type != RTE_FLOW_ITEM_TYPE_END; items++) { + for (; actions->type != RTE_FLOW_ACTION_TYPE_END; actions++) { unsigned int i; + uint64_t current_action_flag = 0; - switch (items->type) { - case RTE_FLOW_ITEM_TYPE_VOID: + switch (actions->type) { + case RTE_FLOW_ACTION_TYPE_VOID: + break; + case RTE_FLOW_ACTION_TYPE_PORT_ID: + current_action_flag = MLX5_FLOW_ACTION_PORT_ID; + if (!actions->conf) + break; + conf.port_id = actions->conf; + if (conf.port_id->original) + i = 0; + else + for (i = 0; ptoi[i].ifindex; ++i) + if (ptoi[i].port_id == conf.port_id->id) + break; + if (!ptoi[i].ifindex) + return rte_flow_error_set + (error, ENODEV, + RTE_FLOW_ERROR_TYPE_ACTION_CONF, + conf.port_id, + "missing data to convert port ID to" + " ifindex"); + port_id_dev = &rte_eth_devices[conf.port_id->id]; + break; + case RTE_FLOW_ACTION_TYPE_JUMP: + current_action_flag = MLX5_FLOW_ACTION_JUMP; + if (!actions->conf) + break; + conf.jump = actions->conf; + if (attr->group >= conf.jump->group) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ACTION, + actions, + "can jump only to a group forward"); + break; + case RTE_FLOW_ACTION_TYPE_DROP: + current_action_flag = MLX5_FLOW_ACTION_DROP; + break; + case RTE_FLOW_ACTION_TYPE_COUNT: + break; + case RTE_FLOW_ACTION_TYPE_OF_POP_VLAN: + current_action_flag = MLX5_FLOW_ACTION_OF_POP_VLAN; + break; + case RTE_FLOW_ACTION_TYPE_OF_PUSH_VLAN: { + rte_be16_t ethertype; + + current_action_flag = MLX5_FLOW_ACTION_OF_PUSH_VLAN; + if (!actions->conf) + break; + conf.of_push_vlan = actions->conf; + ethertype = conf.of_push_vlan->ethertype; + if (ethertype != RTE_BE16(ETH_P_8021Q) && + ethertype != RTE_BE16(ETH_P_8021AD)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, actions, + "vlan push TPID must be " + "802.1Q or 802.1AD"); + break; + } + case RTE_FLOW_ACTION_TYPE_OF_SET_VLAN_VID: + if (!(action_flags & MLX5_FLOW_ACTION_OF_PUSH_VLAN)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ACTION, actions, + "vlan modify is not supported," + " set action must follow push action"); + current_action_flag = MLX5_FLOW_ACTION_OF_SET_VLAN_VID; + break; + case RTE_FLOW_ACTION_TYPE_OF_SET_VLAN_PCP: + if (!(action_flags & MLX5_FLOW_ACTION_OF_PUSH_VLAN)) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ACTION, actions, + "vlan modify is not supported," + " set action must follow push action"); + current_action_flag = MLX5_FLOW_ACTION_OF_SET_VLAN_PCP; + break; + case RTE_FLOW_ACTION_TYPE_VXLAN_DECAP: + current_action_flag = MLX5_FLOW_ACTION_VXLAN_DECAP; + break; + case RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP: + ret = flow_tcf_validate_vxlan_encap(actions, error); + if (ret < 0) + return ret; + current_action_flag = MLX5_FLOW_ACTION_VXLAN_ENCAP; + break; + case RTE_FLOW_ACTION_TYPE_SET_IPV4_SRC: + current_action_flag = MLX5_FLOW_ACTION_SET_IPV4_SRC; + break; + case RTE_FLOW_ACTION_TYPE_SET_IPV4_DST: + current_action_flag = MLX5_FLOW_ACTION_SET_IPV4_DST; + break; + case RTE_FLOW_ACTION_TYPE_SET_IPV6_SRC: + current_action_flag = MLX5_FLOW_ACTION_SET_IPV6_SRC; + break; + case RTE_FLOW_ACTION_TYPE_SET_IPV6_DST: + current_action_flag = MLX5_FLOW_ACTION_SET_IPV6_DST; + break; + case RTE_FLOW_ACTION_TYPE_SET_TP_SRC: + current_action_flag = MLX5_FLOW_ACTION_SET_TP_SRC; + break; + case RTE_FLOW_ACTION_TYPE_SET_TP_DST: + current_action_flag = MLX5_FLOW_ACTION_SET_TP_DST; + break; + case RTE_FLOW_ACTION_TYPE_SET_TTL: + current_action_flag = MLX5_FLOW_ACTION_SET_TTL; + break; + case RTE_FLOW_ACTION_TYPE_DEC_TTL: + current_action_flag = MLX5_FLOW_ACTION_DEC_TTL; + break; + case RTE_FLOW_ACTION_TYPE_SET_MAC_SRC: + current_action_flag = MLX5_FLOW_ACTION_SET_MAC_SRC; + break; + case RTE_FLOW_ACTION_TYPE_SET_MAC_DST: + current_action_flag = MLX5_FLOW_ACTION_SET_MAC_DST; + break; + default: + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ACTION, + actions, + "action not supported"); + } + if (current_action_flag & MLX5_TCF_CONFIG_ACTIONS) { + if (!actions->conf) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION_CONF, + actions, + "action configuration not set"); + } + if ((current_action_flag & MLX5_TCF_PEDIT_ACTIONS) && + pedit_validated) + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ACTION, + actions, + "set actions should be " + "listed successively"); + if ((current_action_flag & ~MLX5_TCF_PEDIT_ACTIONS) && + (action_flags & MLX5_TCF_PEDIT_ACTIONS)) + pedit_validated = 1; + if ((current_action_flag & MLX5_TCF_FATE_ACTIONS) && + (action_flags & MLX5_TCF_FATE_ACTIONS)) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, + actions, + "can't have multiple fate" + " actions"); + if ((current_action_flag & MLX5_TCF_VXLAN_ACTIONS) && + (action_flags & MLX5_TCF_VXLAN_ACTIONS)) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, + actions, + "can't have multiple vxlan" + " actions"); + if ((current_action_flag & MLX5_TCF_VXLAN_ACTIONS) && + (action_flags & MLX5_TCF_VLAN_ACTIONS)) + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ACTION, + actions, + "can't have vxlan and vlan" + " actions in the same rule"); + action_flags |= current_action_flag; + } + for (; items->type != RTE_FLOW_ITEM_TYPE_END; items++) { + unsigned int i; + + switch (items->type) { + case RTE_FLOW_ITEM_TYPE_VOID: break; case RTE_FLOW_ITEM_TYPE_PORT_ID: + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM, items, + "inner tunnel port id" + " item is not supported"); mask.port_id = flow_tcf_item_mask (items, &rte_flow_item_port_id_mask, &flow_tcf_mask_supported.port_id, @@ -1131,7 +2030,9 @@ flow_tcf_validate(struct rte_eth_dev *dev, error); if (ret < 0) return ret; - item_flags |= MLX5_FLOW_LAYER_OUTER_L2; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L2 : + MLX5_FLOW_LAYER_OUTER_L2; /* TODO: * Redundant check due to different supported mask. * Same for the rest of items. @@ -1152,8 +2053,40 @@ flow_tcf_validate(struct rte_eth_dev *dev, mask.eth, "no support for partial mask on" " \"type\" field"); + assert(items->spec); + spec.eth = items->spec; + if (mask.eth->type && + (item_flags & MLX5_FLOW_LAYER_TUNNEL) && + inner_etype != RTE_BE16(ETH_P_ALL) && + inner_etype != spec.eth->type) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "inner eth_type conflict"); + if (mask.eth->type && + !(item_flags & MLX5_FLOW_LAYER_TUNNEL) && + outer_etype != RTE_BE16(ETH_P_ALL) && + outer_etype != spec.eth->type) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "outer eth_type conflict"); + if (mask.eth->type) { + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) + inner_etype = spec.eth->type; + else + outer_etype = spec.eth->type; + } break; case RTE_FLOW_ITEM_TYPE_VLAN: + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) + return rte_flow_error_set + (error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ITEM, items, + "inner tunnel VLAN" + " is not supported"); ret = mlx5_flow_validate_item_vlan(items, item_flags, error); if (ret < 0) @@ -1182,13 +2115,37 @@ flow_tcf_validate(struct rte_eth_dev *dev, "no support for partial masks on" " \"tci\" (PCP and VID parts) and" " \"inner_type\" fields"); + if (outer_etype != RTE_BE16(ETH_P_ALL) && + outer_etype != RTE_BE16(ETH_P_8021Q)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "outer eth_type conflict," + " must be 802.1Q"); + outer_etype = RTE_BE16(ETH_P_8021Q); + assert(items->spec); + spec.vlan = items->spec; + if (mask.vlan->inner_type && + vlan_etype != RTE_BE16(ETH_P_ALL) && + vlan_etype != spec.vlan->inner_type) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "vlan eth_type conflict"); + if (mask.vlan->inner_type) + vlan_etype = spec.vlan->inner_type; break; case RTE_FLOW_ITEM_TYPE_IPV4: - ret = mlx5_flow_validate_item_ipv4(items, item_flags, - error); + ret = mlx5_flow_validate_item_ipv4 + (items, item_flags, + &flow_tcf_mask_supported.ipv4, error); if (ret < 0) return ret; - item_flags |= MLX5_FLOW_LAYER_OUTER_L3_IPV4; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L3_IPV4 : + MLX5_FLOW_LAYER_OUTER_L3_IPV4; mask.ipv4 = flow_tcf_item_mask (items, &rte_flow_item_ipv4_mask, &flow_tcf_mask_supported.ipv4, @@ -1209,13 +2166,47 @@ flow_tcf_validate(struct rte_eth_dev *dev, next_protocol = ((const struct rte_flow_item_ipv4 *) (items->spec))->hdr.next_proto_id; + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) { + if (inner_etype != RTE_BE16(ETH_P_ALL) && + inner_etype != RTE_BE16(ETH_P_IP)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "inner eth_type conflict," + " IPv4 is required"); + inner_etype = RTE_BE16(ETH_P_IP); + } else if (item_flags & MLX5_FLOW_LAYER_OUTER_VLAN) { + if (vlan_etype != RTE_BE16(ETH_P_ALL) && + vlan_etype != RTE_BE16(ETH_P_IP)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "vlan eth_type conflict," + " IPv4 is required"); + vlan_etype = RTE_BE16(ETH_P_IP); + } else { + if (outer_etype != RTE_BE16(ETH_P_ALL) && + outer_etype != RTE_BE16(ETH_P_IP)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "eth_type conflict," + " IPv4 is required"); + outer_etype = RTE_BE16(ETH_P_IP); + } break; case RTE_FLOW_ITEM_TYPE_IPV6: - ret = mlx5_flow_validate_item_ipv6(items, item_flags, - error); + ret = mlx5_flow_validate_item_ipv6 + (items, item_flags, + &flow_tcf_mask_supported.ipv6, error); if (ret < 0) return ret; - item_flags |= MLX5_FLOW_LAYER_OUTER_L3_IPV6; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L3_IPV6 : + MLX5_FLOW_LAYER_OUTER_L3_IPV6; mask.ipv6 = flow_tcf_item_mask (items, &rte_flow_item_ipv6_mask, &flow_tcf_mask_supported.ipv6, @@ -1236,13 +2227,46 @@ flow_tcf_validate(struct rte_eth_dev *dev, next_protocol = ((const struct rte_flow_item_ipv6 *) (items->spec))->hdr.proto; + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) { + if (inner_etype != RTE_BE16(ETH_P_ALL) && + inner_etype != RTE_BE16(ETH_P_IPV6)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "inner eth_type conflict," + " IPv6 is required"); + inner_etype = RTE_BE16(ETH_P_IPV6); + } else if (item_flags & MLX5_FLOW_LAYER_OUTER_VLAN) { + if (vlan_etype != RTE_BE16(ETH_P_ALL) && + vlan_etype != RTE_BE16(ETH_P_IPV6)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "vlan eth_type conflict," + " IPv6 is required"); + vlan_etype = RTE_BE16(ETH_P_IPV6); + } else { + if (outer_etype != RTE_BE16(ETH_P_ALL) && + outer_etype != RTE_BE16(ETH_P_IPV6)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ITEM, + items, + "eth_type conflict," + " IPv6 is required"); + outer_etype = RTE_BE16(ETH_P_IPV6); + } break; case RTE_FLOW_ITEM_TYPE_UDP: ret = mlx5_flow_validate_item_udp(items, item_flags, next_protocol, error); if (ret < 0) return ret; - item_flags |= MLX5_FLOW_LAYER_OUTER_L4_UDP; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L4_UDP : + MLX5_FLOW_LAYER_OUTER_L4_UDP; mask.udp = flow_tcf_item_mask (items, &rte_flow_item_udp_mask, &flow_tcf_mask_supported.udp, @@ -1251,6 +2275,12 @@ flow_tcf_validate(struct rte_eth_dev *dev, error); if (!mask.udp) return -rte_errno; + /* + * Save the presumed outer UDP item for extra check + * if the tunnel item will be found later in the list. + */ + if (!(item_flags & MLX5_FLOW_LAYER_TUNNEL)) + outer_udp = items; break; case RTE_FLOW_ITEM_TYPE_TCP: ret = mlx5_flow_validate_item_tcp @@ -1260,7 +2290,9 @@ flow_tcf_validate(struct rte_eth_dev *dev, error); if (ret < 0) return ret; - item_flags |= MLX5_FLOW_LAYER_OUTER_L4_TCP; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L4_TCP : + MLX5_FLOW_LAYER_OUTER_L4_TCP; mask.tcp = flow_tcf_item_mask (items, &rte_flow_item_tcp_mask, &flow_tcf_mask_supported.tcp, @@ -1270,141 +2302,79 @@ flow_tcf_validate(struct rte_eth_dev *dev, if (!mask.tcp) return -rte_errno; break; - default: - return rte_flow_error_set(error, ENOTSUP, - RTE_FLOW_ERROR_TYPE_ITEM, - NULL, "item not supported"); - } - } - for (; actions->type != RTE_FLOW_ACTION_TYPE_END; actions++) { - unsigned int i; - uint64_t current_action_flag = 0; - - switch (actions->type) { - case RTE_FLOW_ACTION_TYPE_VOID: - break; - case RTE_FLOW_ACTION_TYPE_PORT_ID: - current_action_flag = MLX5_FLOW_ACTION_PORT_ID; - if (!actions->conf) - break; - conf.port_id = actions->conf; - if (conf.port_id->original) - i = 0; - else - for (i = 0; ptoi[i].ifindex; ++i) - if (ptoi[i].port_id == conf.port_id->id) - break; - if (!ptoi[i].ifindex) - return rte_flow_error_set - (error, ENODEV, - RTE_FLOW_ERROR_TYPE_ACTION_CONF, - conf.port_id, - "missing data to convert port ID to" - " ifindex"); - port_id_dev = &rte_eth_devices[conf.port_id->id]; - break; - case RTE_FLOW_ACTION_TYPE_JUMP: - current_action_flag = MLX5_FLOW_ACTION_JUMP; - if (!actions->conf) - break; - conf.jump = actions->conf; - if (attr->group >= conf.jump->group) + case RTE_FLOW_ITEM_TYPE_VXLAN: + if (item_flags & MLX5_FLOW_LAYER_OUTER_VLAN) return rte_flow_error_set (error, ENOTSUP, - RTE_FLOW_ERROR_TYPE_ACTION, - actions, - "can jump only to a group forward"); - break; - case RTE_FLOW_ACTION_TYPE_DROP: - current_action_flag = MLX5_FLOW_ACTION_DROP; - break; - case RTE_FLOW_ACTION_TYPE_COUNT: - break; - case RTE_FLOW_ACTION_TYPE_OF_POP_VLAN: - current_action_flag = MLX5_FLOW_ACTION_OF_POP_VLAN; - break; - case RTE_FLOW_ACTION_TYPE_OF_PUSH_VLAN: - current_action_flag = MLX5_FLOW_ACTION_OF_PUSH_VLAN; - break; - case RTE_FLOW_ACTION_TYPE_OF_SET_VLAN_VID: - if (!(action_flags & MLX5_FLOW_ACTION_OF_PUSH_VLAN)) + RTE_FLOW_ERROR_TYPE_ITEM, items, + "vxlan tunnel over vlan" + " is not supported"); + ret = mlx5_flow_validate_item_vxlan(items, + item_flags, error); + if (ret < 0) + return ret; + item_flags |= MLX5_FLOW_LAYER_VXLAN; + mask.vxlan = flow_tcf_item_mask + (items, &rte_flow_item_vxlan_mask, + &flow_tcf_mask_supported.vxlan, + &flow_tcf_mask_empty.vxlan, + sizeof(flow_tcf_mask_supported.vxlan), error); + if (!mask.vxlan) + return -rte_errno; + if (mask.vxlan->vni[0] != 0xff || + mask.vxlan->vni[1] != 0xff || + mask.vxlan->vni[2] != 0xff) return rte_flow_error_set (error, ENOTSUP, - RTE_FLOW_ERROR_TYPE_ACTION, actions, - "vlan modify is not supported," - " set action must follow push action"); - current_action_flag = MLX5_FLOW_ACTION_OF_SET_VLAN_VID; - break; - case RTE_FLOW_ACTION_TYPE_OF_SET_VLAN_PCP: - if (!(action_flags & MLX5_FLOW_ACTION_OF_PUSH_VLAN)) + RTE_FLOW_ERROR_TYPE_ITEM_MASK, + mask.vxlan, + "no support for partial or " + "empty mask on \"vxlan.vni\" field"); + /* + * The VNI item assumes the VXLAN tunnel, it requires + * at least the outer destination UDP port must be + * specified without wildcards to allow kernel select + * the virtual VXLAN device by port. Also outer IPv4 + * or IPv6 item must be specified (wilcards or even + * zero mask are allowed) to let driver know the tunnel + * IP version and process UDP traffic correctly. + */ + if (!(item_flags & + (MLX5_FLOW_LAYER_OUTER_L3_IPV4 | + MLX5_FLOW_LAYER_OUTER_L3_IPV6))) return rte_flow_error_set - (error, ENOTSUP, - RTE_FLOW_ERROR_TYPE_ACTION, actions, - "vlan modify is not supported," - " set action must follow push action"); - current_action_flag = MLX5_FLOW_ACTION_OF_SET_VLAN_PCP; - break; - case RTE_FLOW_ACTION_TYPE_SET_IPV4_SRC: - current_action_flag = MLX5_FLOW_ACTION_SET_IPV4_SRC; - break; - case RTE_FLOW_ACTION_TYPE_SET_IPV4_DST: - current_action_flag = MLX5_FLOW_ACTION_SET_IPV4_DST; - break; - case RTE_FLOW_ACTION_TYPE_SET_IPV6_SRC: - current_action_flag = MLX5_FLOW_ACTION_SET_IPV6_SRC; - break; - case RTE_FLOW_ACTION_TYPE_SET_IPV6_DST: - current_action_flag = MLX5_FLOW_ACTION_SET_IPV6_DST; - break; - case RTE_FLOW_ACTION_TYPE_SET_TP_SRC: - current_action_flag = MLX5_FLOW_ACTION_SET_TP_SRC; - break; - case RTE_FLOW_ACTION_TYPE_SET_TP_DST: - current_action_flag = MLX5_FLOW_ACTION_SET_TP_DST; - break; - case RTE_FLOW_ACTION_TYPE_SET_TTL: - current_action_flag = MLX5_FLOW_ACTION_SET_TTL; - break; - case RTE_FLOW_ACTION_TYPE_DEC_TTL: - current_action_flag = MLX5_FLOW_ACTION_DEC_TTL; - break; - case RTE_FLOW_ACTION_TYPE_SET_MAC_SRC: - current_action_flag = MLX5_FLOW_ACTION_SET_MAC_SRC; - break; - case RTE_FLOW_ACTION_TYPE_SET_MAC_DST: - current_action_flag = MLX5_FLOW_ACTION_SET_MAC_DST; + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, + NULL, + "no outer IP pattern found" + " for vxlan tunnel"); + if (!(item_flags & MLX5_FLOW_LAYER_OUTER_L4_UDP)) + return rte_flow_error_set + (error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, + NULL, + "no outer UDP pattern found" + " for vxlan tunnel"); + /* + * All items preceding the tunnel item become outer + * ones and we should do extra validation for them + * due to tc limitations for tunnel outer parameters. + * Currently only outer UDP item requres extra check, + * use the saved pointer instead of item list rescan. + */ + assert(outer_udp); + ret = flow_tcf_validate_vxlan_decap_udp + (outer_udp, error); + if (ret < 0) + return ret; + /* Reset L4 protocol for inner parameters. */ + next_protocol = 0xff; break; default: return rte_flow_error_set(error, ENOTSUP, - RTE_FLOW_ERROR_TYPE_ACTION, - actions, - "action not supported"); - } - if (current_action_flag & MLX5_TCF_CONFIG_ACTIONS) { - if (!actions->conf) - return rte_flow_error_set(error, EINVAL, - RTE_FLOW_ERROR_TYPE_ACTION_CONF, - actions, - "action configuration not set"); + RTE_FLOW_ERROR_TYPE_ITEM, + items, "item not supported"); } - if ((current_action_flag & MLX5_TCF_PEDIT_ACTIONS) && - pedit_validated) - return rte_flow_error_set(error, ENOTSUP, - RTE_FLOW_ERROR_TYPE_ACTION, - actions, - "set actions should be " - "listed successively"); - if ((current_action_flag & ~MLX5_TCF_PEDIT_ACTIONS) && - (action_flags & MLX5_TCF_PEDIT_ACTIONS)) - pedit_validated = 1; - if ((current_action_flag & MLX5_TCF_FATE_ACTIONS) && - (action_flags & MLX5_TCF_FATE_ACTIONS)) - return rte_flow_error_set(error, EINVAL, - RTE_FLOW_ERROR_TYPE_ACTION, - actions, - "can't have multiple fate" - " actions"); - action_flags |= current_action_flag; } if ((action_flags & MLX5_TCF_PEDIT_ACTIONS) && (action_flags & MLX5_FLOW_ACTION_DROP)) @@ -1456,7 +2426,7 @@ flow_tcf_validate(struct rte_eth_dev *dev, */ if ((action_flags & MLX5_FLOW_ACTION_OF_PUSH_VLAN) && (action_flags & MLX5_FLOW_ACTION_PORT_ID) && - ((struct priv *)port_id_dev->data->dev_private)->representor) + ((struct mlx5_priv *)port_id_dev->data->dev_private)->representor) return rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ACTION, actions, "vlan push can only be applied" @@ -1472,6 +2442,12 @@ flow_tcf_validate(struct rte_eth_dev *dev, RTE_FLOW_ERROR_TYPE_ACTION, actions, "vlan actions are supported" " only with port_id action"); + if ((action_flags & MLX5_TCF_VXLAN_ACTIONS) && + !(action_flags & MLX5_FLOW_ACTION_PORT_ID)) + return rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_ACTION, NULL, + "vxlan actions are supported" + " only with port_id action"); if (!(action_flags & MLX5_TCF_FATE_ACTIONS)) return rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_ACTION, actions, @@ -1495,30 +2471,45 @@ flow_tcf_validate(struct rte_eth_dev *dev, "no ethernet found in" " pattern"); } + if ((action_flags & MLX5_FLOW_ACTION_VXLAN_DECAP) && + !(item_flags & MLX5_FLOW_LAYER_VXLAN)) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, + NULL, + "no VNI pattern found" + " for vxlan decap action"); + if ((action_flags & MLX5_FLOW_ACTION_VXLAN_ENCAP) && + (item_flags & MLX5_FLOW_LAYER_TUNNEL)) + return rte_flow_error_set(error, EINVAL, + RTE_FLOW_ERROR_TYPE_ACTION, + NULL, + "vxlan encap not supported" + " for tunneled traffic"); return 0; } /** - * Calculate maximum size of memory for flow items of Linux TC flower and - * extract specified items. + * Calculate maximum size of memory for flow items of Linux TC flower. * + * @param[in] attr + * Pointer to the flow attributes. * @param[in] items * Pointer to the list of items. - * @param[out] item_flags - * Pointer to the detected items. + * @param[out] action_flags + * Pointer to the detected actions. * * @return * Maximum size of memory for items. */ static int -flow_tcf_get_items_and_size(const struct rte_flow_attr *attr, - const struct rte_flow_item items[], - uint64_t *item_flags) +flow_tcf_get_items_size(const struct rte_flow_attr *attr, + const struct rte_flow_item items[], + uint64_t *action_flags) { int size = 0; - uint64_t flags = 0; size += SZ_NLATTR_STRZ_OF("flower") + + SZ_NLATTR_TYPE_OF(uint16_t) + /* Outer ether type. */ SZ_NLATTR_NEST + /* TCA_OPTIONS. */ SZ_NLATTR_TYPE_OF(uint32_t); /* TCA_CLS_FLAGS_SKIP_SW. */ if (attr->group > 0) @@ -1530,44 +2521,62 @@ flow_tcf_get_items_and_size(const struct rte_flow_attr *attr, case RTE_FLOW_ITEM_TYPE_PORT_ID: break; case RTE_FLOW_ITEM_TYPE_ETH: - size += SZ_NLATTR_TYPE_OF(uint16_t) + /* Ether type. */ - SZ_NLATTR_DATA_OF(ETHER_ADDR_LEN) * 4; + size += SZ_NLATTR_DATA_OF(RTE_ETHER_ADDR_LEN) * 4; /* dst/src MAC addr and mask. */ - flags |= MLX5_FLOW_LAYER_OUTER_L2; break; case RTE_FLOW_ITEM_TYPE_VLAN: - size += SZ_NLATTR_TYPE_OF(uint16_t) + /* Ether type. */ - SZ_NLATTR_TYPE_OF(uint16_t) + + size += SZ_NLATTR_TYPE_OF(uint16_t) + /* VLAN Ether type. */ SZ_NLATTR_TYPE_OF(uint8_t) + /* VLAN prio. */ SZ_NLATTR_TYPE_OF(uint16_t); /* VLAN ID. */ - flags |= MLX5_FLOW_LAYER_OUTER_VLAN; break; - case RTE_FLOW_ITEM_TYPE_IPV4: - size += SZ_NLATTR_TYPE_OF(uint16_t) + /* Ether type. */ - SZ_NLATTR_TYPE_OF(uint8_t) + /* IP proto. */ + case RTE_FLOW_ITEM_TYPE_IPV4: { + const struct rte_flow_item_ipv4 *ipv4 = items->mask; + + size += SZ_NLATTR_TYPE_OF(uint8_t) + /* IP proto. */ SZ_NLATTR_TYPE_OF(uint32_t) * 4; /* dst/src IP addr and mask. */ - flags |= MLX5_FLOW_LAYER_OUTER_L3_IPV4; + if (ipv4 && ipv4->hdr.time_to_live) + size += SZ_NLATTR_TYPE_OF(uint8_t) * 2; + if (ipv4 && ipv4->hdr.type_of_service) + size += SZ_NLATTR_TYPE_OF(uint8_t) * 2; break; - case RTE_FLOW_ITEM_TYPE_IPV6: - size += SZ_NLATTR_TYPE_OF(uint16_t) + /* Ether type. */ - SZ_NLATTR_TYPE_OF(uint8_t) + /* IP proto. */ - SZ_NLATTR_TYPE_OF(IPV6_ADDR_LEN) * 4; + } + case RTE_FLOW_ITEM_TYPE_IPV6: { + const struct rte_flow_item_ipv6 *ipv6 = items->mask; + + size += SZ_NLATTR_TYPE_OF(uint8_t) + /* IP proto. */ + SZ_NLATTR_DATA_OF(IPV6_ADDR_LEN) * 4; /* dst/src IP addr and mask. */ - flags |= MLX5_FLOW_LAYER_OUTER_L3_IPV6; + if (ipv6 && ipv6->hdr.hop_limits) + size += SZ_NLATTR_TYPE_OF(uint8_t) * 2; + if (ipv6 && (rte_be_to_cpu_32(ipv6->hdr.vtc_flow) & + (0xfful << IPV6_HDR_TC_SHIFT))) + size += SZ_NLATTR_TYPE_OF(uint8_t) * 2; break; + } case RTE_FLOW_ITEM_TYPE_UDP: size += SZ_NLATTR_TYPE_OF(uint8_t) + /* IP proto. */ SZ_NLATTR_TYPE_OF(uint16_t) * 4; /* dst/src port and mask. */ - flags |= MLX5_FLOW_LAYER_OUTER_L4_UDP; break; case RTE_FLOW_ITEM_TYPE_TCP: size += SZ_NLATTR_TYPE_OF(uint8_t) + /* IP proto. */ SZ_NLATTR_TYPE_OF(uint16_t) * 4; /* dst/src port and mask. */ - flags |= MLX5_FLOW_LAYER_OUTER_L4_TCP; + break; + case RTE_FLOW_ITEM_TYPE_VXLAN: + size += SZ_NLATTR_TYPE_OF(uint32_t); + /* + * There might be no VXLAN decap action in the action + * list, nonetheless the VXLAN tunnel flow requires + * the decap structure to be correctly applied to + * VXLAN device, set the flag to create the structure. + * Translation routine will not put the decap action + * in tne Netlink message if there is no actual action + * in the list. + */ + *action_flags |= MLX5_FLOW_ACTION_VXLAN_DECAP; break; default: DRV_LOG(WARNING, @@ -1577,7 +2586,84 @@ flow_tcf_get_items_and_size(const struct rte_flow_attr *attr, break; } } - *item_flags = flags; + return size; +} + +/** + * Calculate size of memory to store the VXLAN encapsultion + * related items in the Netlink message buffer. Items list + * is specified by RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP action. + * The item list should be validated. + * + * @param[in] action + * RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP action object. + * List of pattern items to scan data from. + * + * @return + * The size the part of Netlink message buffer to store the + * VXLAN encapsulation item attributes. + */ +static int +flow_tcf_vxlan_encap_size(const struct rte_flow_action *action) +{ + const struct rte_flow_item *items; + int size = 0; + + assert(action->type == RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP); + assert(action->conf); + + items = ((const struct rte_flow_action_vxlan_encap *) + action->conf)->definition; + assert(items); + for (; items->type != RTE_FLOW_ITEM_TYPE_END; items++) { + switch (items->type) { + case RTE_FLOW_ITEM_TYPE_VOID: + break; + case RTE_FLOW_ITEM_TYPE_ETH: + /* This item does not require message buffer. */ + break; + case RTE_FLOW_ITEM_TYPE_IPV4: { + const struct rte_flow_item_ipv4 *ipv4 = items->mask; + + size += SZ_NLATTR_DATA_OF(IPV4_ADDR_LEN) * 2; + if (ipv4 && ipv4->hdr.time_to_live) + size += SZ_NLATTR_TYPE_OF(uint8_t) * 2; + if (ipv4 && ipv4->hdr.type_of_service) + size += SZ_NLATTR_TYPE_OF(uint8_t) * 2; + break; + } + case RTE_FLOW_ITEM_TYPE_IPV6: { + const struct rte_flow_item_ipv6 *ipv6 = items->mask; + + size += SZ_NLATTR_DATA_OF(IPV6_ADDR_LEN) * 2; + if (ipv6 && ipv6->hdr.hop_limits) + size += SZ_NLATTR_TYPE_OF(uint8_t) * 2; + if (ipv6 && (rte_be_to_cpu_32(ipv6->hdr.vtc_flow) & + (0xfful << IPV6_HDR_TC_SHIFT))) + size += SZ_NLATTR_TYPE_OF(uint8_t) * 2; + break; + } + case RTE_FLOW_ITEM_TYPE_UDP: { + const struct rte_flow_item_udp *udp = items->mask; + + size += SZ_NLATTR_TYPE_OF(uint16_t); + if (!udp || udp->hdr.src_port != RTE_BE16(0x0000)) + size += SZ_NLATTR_TYPE_OF(uint16_t); + break; + } + case RTE_FLOW_ITEM_TYPE_VXLAN: + size += SZ_NLATTR_TYPE_OF(uint32_t); + break; + default: + assert(false); + DRV_LOG(WARNING, + "unsupported item %p type %d," + " items must be validated" + " before flow creation", + (const void *)items, items->type); + return 0; + } + } return size; } @@ -1598,7 +2684,7 @@ flow_tcf_get_actions_and_size(const struct rte_flow_action actions[], uint64_t *action_flags) { int size = 0; - uint64_t flags = 0; + uint64_t flags = *action_flags; size += SZ_NLATTR_NEST; /* TCA_FLOWER_ACT. */ for (; actions->type != RTE_FLOW_ACTION_TYPE_END; actions++) { @@ -1650,6 +2736,29 @@ action_of_vlan: SZ_NLATTR_TYPE_OF(uint16_t) + /* VLAN ID. */ SZ_NLATTR_TYPE_OF(uint8_t); /* VLAN prio. */ break; + case RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP: + size += SZ_NLATTR_NEST + /* na_act_index. */ + SZ_NLATTR_STRZ_OF("tunnel_key") + + SZ_NLATTR_NEST + /* TCA_ACT_OPTIONS. */ + SZ_NLATTR_TYPE_OF(uint8_t); + size += SZ_NLATTR_TYPE_OF(struct tc_tunnel_key); + size += flow_tcf_vxlan_encap_size(actions) + + RTE_ALIGN_CEIL /* preceding encap params. */ + (sizeof(struct flow_tcf_vxlan_encap), + MNL_ALIGNTO); + flags |= MLX5_FLOW_ACTION_VXLAN_ENCAP; + break; + case RTE_FLOW_ACTION_TYPE_VXLAN_DECAP: + size += SZ_NLATTR_NEST + /* na_act_index. */ + SZ_NLATTR_STRZ_OF("tunnel_key") + + SZ_NLATTR_NEST + /* TCA_ACT_OPTIONS. */ + SZ_NLATTR_TYPE_OF(uint8_t); + size += SZ_NLATTR_TYPE_OF(struct tc_tunnel_key); + size += RTE_ALIGN_CEIL /* preceding decap params. */ + (sizeof(struct flow_tcf_vxlan_decap), + MNL_ALIGNTO); + flags |= MLX5_FLOW_ACTION_VXLAN_DECAP; + break; case RTE_FLOW_ACTION_TYPE_SET_IPV4_SRC: case RTE_FLOW_ACTION_TYPE_SET_IPV4_DST: case RTE_FLOW_ACTION_TYPE_SET_IPV6_SRC: @@ -1675,27 +2784,6 @@ action_of_vlan: return size; } -/** - * Brand rtnetlink buffer with unique handle. - * - * This handle should be unique for a given network interface to avoid - * collisions. - * - * @param nlh - * Pointer to Netlink message. - * @param handle - * Unique 32-bit handle to use. - */ -static void -flow_tcf_nl_brand(struct nlmsghdr *nlh, uint32_t handle) -{ - struct tcmsg *tcm = mnl_nlmsg_get_payload(nlh); - - tcm->tcm_handle = handle; - DRV_LOG(DEBUG, "Netlink msg %p is branded with handle %x", - (void *)nlh, handle); -} - /** * Prepare a flow object for Linux TC flower. It calculates the maximum size of * memory required, allocates the memory, initializes Netlink message headers @@ -1707,33 +2795,32 @@ flow_tcf_nl_brand(struct nlmsghdr *nlh, uint32_t handle) * Pointer to the list of items. * @param[in] actions * Pointer to the list of actions. - * @param[out] item_flags - * Pointer to bit mask of all items detected. - * @param[out] action_flags - * Pointer to bit mask of all actions detected. * @param[out] error * Pointer to the error structure. * * @return * Pointer to mlx5_flow object on success, - * otherwise NULL and rte_ernno is set. + * otherwise NULL and rte_errno is set. */ static struct mlx5_flow * flow_tcf_prepare(const struct rte_flow_attr *attr, const struct rte_flow_item items[], const struct rte_flow_action actions[], - uint64_t *item_flags, uint64_t *action_flags, struct rte_flow_error *error) { - size_t size = sizeof(struct mlx5_flow) + + size_t size = RTE_ALIGN_CEIL + (sizeof(struct mlx5_flow), + alignof(struct flow_tcf_tunnel_hdr)) + MNL_ALIGN(sizeof(struct nlmsghdr)) + MNL_ALIGN(sizeof(struct tcmsg)); struct mlx5_flow *dev_flow; + uint64_t action_flags = 0; struct nlmsghdr *nlh; struct tcmsg *tcm; + uint8_t *sp, *tun = NULL; - size += flow_tcf_get_items_and_size(attr, items, item_flags); - size += flow_tcf_get_actions_and_size(actions, action_flags); + size += flow_tcf_get_items_size(attr, items, &action_flags); + size += flow_tcf_get_actions_and_size(actions, &action_flags); dev_flow = rte_zmalloc(__func__, size, MNL_ALIGNTO); if (!dev_flow) { rte_flow_error_set(error, ENOMEM, @@ -1741,28 +2828,52 @@ flow_tcf_prepare(const struct rte_flow_attr *attr, "not enough memory to create E-Switch flow"); return NULL; } - nlh = mnl_nlmsg_put_header((void *)(dev_flow + 1)); + sp = (uint8_t *)(dev_flow + 1); + if (action_flags & MLX5_FLOW_ACTION_VXLAN_ENCAP) { + sp = RTE_PTR_ALIGN + (sp, alignof(struct flow_tcf_tunnel_hdr)); + tun = sp; + sp += RTE_ALIGN_CEIL + (sizeof(struct flow_tcf_vxlan_encap), + MNL_ALIGNTO); +#ifndef NDEBUG + size -= RTE_ALIGN_CEIL + (sizeof(struct flow_tcf_vxlan_encap), + MNL_ALIGNTO); +#endif + } else if (action_flags & MLX5_FLOW_ACTION_VXLAN_DECAP) { + sp = RTE_PTR_ALIGN + (sp, alignof(struct flow_tcf_tunnel_hdr)); + tun = sp; + sp += RTE_ALIGN_CEIL + (sizeof(struct flow_tcf_vxlan_decap), + MNL_ALIGNTO); +#ifndef NDEBUG + size -= RTE_ALIGN_CEIL + (sizeof(struct flow_tcf_vxlan_decap), + MNL_ALIGNTO); +#endif + } else { + sp = RTE_PTR_ALIGN(sp, MNL_ALIGNTO); + } + nlh = mnl_nlmsg_put_header(sp); tcm = mnl_nlmsg_put_extra_header(nlh, sizeof(*tcm)); *dev_flow = (struct mlx5_flow){ .tcf = (struct mlx5_flow_tcf){ +#ifndef NDEBUG + .nlsize = size - RTE_ALIGN_CEIL + (sizeof(struct mlx5_flow), + alignof(struct flow_tcf_tunnel_hdr)), +#endif + .tunnel = (struct flow_tcf_tunnel_hdr *)tun, .nlh = nlh, .tcm = tcm, }, }; - /* - * Generate a reasonably unique handle based on the address of the - * target buffer. - * - * This is straightforward on 32-bit systems where the flow pointer can - * be used directly. Otherwise, its least significant part is taken - * after shifting it by the previous power of two of the pointed buffer - * size. - */ - if (sizeof(dev_flow) <= 4) - flow_tcf_nl_brand(nlh, (uintptr_t)dev_flow); - else - flow_tcf_nl_brand(nlh, (uintptr_t)dev_flow >> - rte_log2_u32(rte_align32prevpow2(size))); + if (action_flags & MLX5_FLOW_ACTION_VXLAN_DECAP) + dev_flow->tcf.tunnel->type = FLOW_TCF_TUNACT_VXLAN_DECAP; + else if (action_flags & MLX5_FLOW_ACTION_VXLAN_ENCAP) + dev_flow->tcf.tunnel->type = FLOW_TCF_TUNACT_VXLAN_ENCAP; return dev_flow; } @@ -1798,6 +2909,272 @@ flow_tcf_translate_action_count(struct rte_eth_dev *dev __rte_unused, return 0; } +/** + * Convert VXLAN VNI to 32-bit integer. + * + * @param[in] vni + * VXLAN VNI in 24-bit wire format. + * + * @return + * VXLAN VNI as a 32-bit integer value in network endianness. + */ +static inline rte_be32_t +vxlan_vni_as_be32(const uint8_t vni[3]) +{ + union { + uint8_t vni[4]; + rte_be32_t dword; + } ret = { + .vni = { 0, vni[0], vni[1], vni[2] }, + }; + return ret.dword; +} + +/** + * Helper function to process RTE_FLOW_ITEM_TYPE_ETH entry in configuration + * of action RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP. Fills the MAC address fields + * in the encapsulation parameters structure. The item must be prevalidated, + * no any validation checks performed by function. + * + * @param[in] spec + * RTE_FLOW_ITEM_TYPE_ETH entry specification. + * @param[in] mask + * RTE_FLOW_ITEM_TYPE_ETH entry mask. + * @param[out] encap + * Structure to fill the gathered MAC address data. + */ +static void +flow_tcf_parse_vxlan_encap_eth(const struct rte_flow_item_eth *spec, + const struct rte_flow_item_eth *mask, + struct flow_tcf_vxlan_encap *encap) +{ + /* Item must be validated before. No redundant checks. */ + assert(spec); + if (!mask || !memcmp(&mask->dst, + &rte_flow_item_eth_mask.dst, + sizeof(rte_flow_item_eth_mask.dst))) { + /* + * Ethernet addresses are not supported by + * tc as tunnel_key parameters. Destination + * address is needed to form encap packet + * header and retrieved by kernel from + * implicit sources (ARP table, etc), + * address masks are not supported at all. + */ + encap->eth.dst = spec->dst; + encap->mask |= FLOW_TCF_ENCAP_ETH_DST; + } + if (!mask || !memcmp(&mask->src, + &rte_flow_item_eth_mask.src, + sizeof(rte_flow_item_eth_mask.src))) { + /* + * Ethernet addresses are not supported by + * tc as tunnel_key parameters. Source ethernet + * address is ignored anyway. + */ + encap->eth.src = spec->src; + encap->mask |= FLOW_TCF_ENCAP_ETH_SRC; + } +} + +/** + * Helper function to process RTE_FLOW_ITEM_TYPE_IPV4 entry in configuration + * of action RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP. Fills the IPV4 address fields + * in the encapsulation parameters structure. The item must be prevalidated, + * no any validation checks performed by function. + * + * @param[in] spec + * RTE_FLOW_ITEM_TYPE_IPV4 entry specification. + * @param[in] mask + * RTE_FLOW_ITEM_TYPE_IPV4 entry mask. + * @param[out] encap + * Structure to fill the gathered IPV4 address data. + */ +static void +flow_tcf_parse_vxlan_encap_ipv4(const struct rte_flow_item_ipv4 *spec, + const struct rte_flow_item_ipv4 *mask, + struct flow_tcf_vxlan_encap *encap) +{ + /* Item must be validated before. No redundant checks. */ + assert(spec); + encap->ipv4.dst = spec->hdr.dst_addr; + encap->ipv4.src = spec->hdr.src_addr; + encap->mask |= FLOW_TCF_ENCAP_IPV4_SRC | + FLOW_TCF_ENCAP_IPV4_DST; + if (mask && mask->hdr.type_of_service) { + encap->mask |= FLOW_TCF_ENCAP_IP_TOS; + encap->ip_tos = spec->hdr.type_of_service; + } + if (mask && mask->hdr.time_to_live) { + encap->mask |= FLOW_TCF_ENCAP_IP_TTL; + encap->ip_ttl_hop = spec->hdr.time_to_live; + } +} + +/** + * Helper function to process RTE_FLOW_ITEM_TYPE_IPV6 entry in configuration + * of action RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP. Fills the IPV6 address fields + * in the encapsulation parameters structure. The item must be prevalidated, + * no any validation checks performed by function. + * + * @param[in] spec + * RTE_FLOW_ITEM_TYPE_IPV6 entry specification. + * @param[in] mask + * RTE_FLOW_ITEM_TYPE_IPV6 entry mask. + * @param[out] encap + * Structure to fill the gathered IPV6 address data. + */ +static void +flow_tcf_parse_vxlan_encap_ipv6(const struct rte_flow_item_ipv6 *spec, + const struct rte_flow_item_ipv6 *mask, + struct flow_tcf_vxlan_encap *encap) +{ + /* Item must be validated before. No redundant checks. */ + assert(spec); + memcpy(encap->ipv6.dst, spec->hdr.dst_addr, IPV6_ADDR_LEN); + memcpy(encap->ipv6.src, spec->hdr.src_addr, IPV6_ADDR_LEN); + encap->mask |= FLOW_TCF_ENCAP_IPV6_SRC | + FLOW_TCF_ENCAP_IPV6_DST; + if (mask) { + if ((rte_be_to_cpu_32(mask->hdr.vtc_flow) >> + IPV6_HDR_TC_SHIFT) & 0xff) { + encap->mask |= FLOW_TCF_ENCAP_IP_TOS; + encap->ip_tos = (rte_be_to_cpu_32 + (spec->hdr.vtc_flow) >> + IPV6_HDR_TC_SHIFT) & 0xff; + } + if (mask->hdr.hop_limits) { + encap->mask |= FLOW_TCF_ENCAP_IP_TTL; + encap->ip_ttl_hop = spec->hdr.hop_limits; + } + } +} + +/** + * Helper function to process RTE_FLOW_ITEM_TYPE_UDP entry in configuration + * of action RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP. Fills the UDP port fields + * in the encapsulation parameters structure. The item must be prevalidated, + * no any validation checks performed by function. + * + * @param[in] spec + * RTE_FLOW_ITEM_TYPE_UDP entry specification. + * @param[in] mask + * RTE_FLOW_ITEM_TYPE_UDP entry mask. + * @param[out] encap + * Structure to fill the gathered UDP port data. + */ +static void +flow_tcf_parse_vxlan_encap_udp(const struct rte_flow_item_udp *spec, + const struct rte_flow_item_udp *mask, + struct flow_tcf_vxlan_encap *encap) +{ + assert(spec); + encap->udp.dst = spec->hdr.dst_port; + encap->mask |= FLOW_TCF_ENCAP_UDP_DST; + if (!mask || mask->hdr.src_port != RTE_BE16(0x0000)) { + encap->udp.src = spec->hdr.src_port; + encap->mask |= FLOW_TCF_ENCAP_IPV4_SRC; + } +} + +/** + * Helper function to process RTE_FLOW_ITEM_TYPE_VXLAN entry in configuration + * of action RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP. Fills the VNI fields + * in the encapsulation parameters structure. The item must be prevalidated, + * no any validation checks performed by function. + * + * @param[in] spec + * RTE_FLOW_ITEM_TYPE_VXLAN entry specification. + * @param[out] encap + * Structure to fill the gathered VNI address data. + */ +static void +flow_tcf_parse_vxlan_encap_vni(const struct rte_flow_item_vxlan *spec, + struct flow_tcf_vxlan_encap *encap) +{ + /* Item must be validated before. Do not redundant checks. */ + assert(spec); + memcpy(encap->vxlan.vni, spec->vni, sizeof(encap->vxlan.vni)); + encap->mask |= FLOW_TCF_ENCAP_VXLAN_VNI; +} + +/** + * Populate consolidated encapsulation object from list of pattern items. + * + * Helper function to process configuration of action such as + * RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP. The item list should be + * validated, there is no way to return an meaningful error. + * + * @param[in] action + * RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP action object. + * List of pattern items to gather data from. + * @param[out] src + * Structure to fill gathered data. + */ +static void +flow_tcf_vxlan_encap_parse(const struct rte_flow_action *action, + struct flow_tcf_vxlan_encap *encap) +{ + union { + const struct rte_flow_item_eth *eth; + const struct rte_flow_item_ipv4 *ipv4; + const struct rte_flow_item_ipv6 *ipv6; + const struct rte_flow_item_udp *udp; + const struct rte_flow_item_vxlan *vxlan; + } spec, mask; + const struct rte_flow_item *items; + + assert(action->type == RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP); + assert(action->conf); + + items = ((const struct rte_flow_action_vxlan_encap *) + action->conf)->definition; + assert(items); + for (; items->type != RTE_FLOW_ITEM_TYPE_END; items++) { + switch (items->type) { + case RTE_FLOW_ITEM_TYPE_VOID: + break; + case RTE_FLOW_ITEM_TYPE_ETH: + mask.eth = items->mask; + spec.eth = items->spec; + flow_tcf_parse_vxlan_encap_eth(spec.eth, mask.eth, + encap); + break; + case RTE_FLOW_ITEM_TYPE_IPV4: + spec.ipv4 = items->spec; + mask.ipv4 = items->mask; + flow_tcf_parse_vxlan_encap_ipv4(spec.ipv4, mask.ipv4, + encap); + break; + case RTE_FLOW_ITEM_TYPE_IPV6: + spec.ipv6 = items->spec; + mask.ipv6 = items->mask; + flow_tcf_parse_vxlan_encap_ipv6(spec.ipv6, mask.ipv6, + encap); + break; + case RTE_FLOW_ITEM_TYPE_UDP: + mask.udp = items->mask; + spec.udp = items->spec; + flow_tcf_parse_vxlan_encap_udp(spec.udp, mask.udp, + encap); + break; + case RTE_FLOW_ITEM_TYPE_VXLAN: + spec.vxlan = items->spec; + flow_tcf_parse_vxlan_encap_vni(spec.vxlan, encap); + break; + default: + assert(false); + DRV_LOG(WARNING, + "unsupported item %p type %d," + " items must be validated" + " before flow creation", + (const void *)items, items->type); + encap->mask = 0; + return; + } + } +} + /** * Translate flow for Linux TC flower and construct Netlink message. * @@ -1815,7 +3192,7 @@ flow_tcf_translate_action_count(struct rte_eth_dev *dev __rte_unused, * Pointer to the error structure. * * @return - * 0 on success, a negative errno value otherwise and rte_ernno is set. + * 0 on success, a negative errno value otherwise and rte_errno is set. */ static int flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, @@ -1832,6 +3209,7 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, const struct rte_flow_item_ipv6 *ipv6; const struct rte_flow_item_tcp *tcp; const struct rte_flow_item_udp *udp; + const struct rte_flow_item_vxlan *vxlan; } spec, mask; union { const struct rte_flow_action_port_id *port_id; @@ -1842,14 +3220,27 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, const struct rte_flow_action_of_set_vlan_pcp * of_set_vlan_pcp; } conf; + union { + struct flow_tcf_tunnel_hdr *hdr; + struct flow_tcf_vxlan_decap *vxlan; + } decap = { + .hdr = NULL, + }; + union { + struct flow_tcf_tunnel_hdr *hdr; + struct flow_tcf_vxlan_encap *vxlan; + } encap = { + .hdr = NULL, + }; struct flow_tcf_ptoi ptoi[PTOI_TABLE_SZ_MAX(dev)]; struct nlmsghdr *nlh = dev_flow->tcf.nlh; struct tcmsg *tcm = dev_flow->tcf.tcm; uint32_t na_act_index_cur; - bool eth_type_set = 0; - bool vlan_present = 0; - bool vlan_eth_type_set = 0; + rte_be16_t inner_etype = RTE_BE16(ETH_P_ALL); + rte_be16_t outer_etype = RTE_BE16(ETH_P_ALL); + rte_be16_t vlan_etype = RTE_BE16(ETH_P_ALL); bool ip_proto_set = 0; + bool tunnel_outer = 0; struct nlattr *na_flower; struct nlattr *na_flower_act; struct nlattr *na_vlan_id = NULL; @@ -1859,6 +3250,21 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, claim_nonzero(flow_tcf_build_ptoi_table(dev, ptoi, PTOI_TABLE_SZ_MAX(dev))); + if (dev_flow->tcf.tunnel) { + switch (dev_flow->tcf.tunnel->type) { + case FLOW_TCF_TUNACT_VXLAN_DECAP: + decap.vxlan = dev_flow->tcf.vxlan_decap; + tunnel_outer = 1; + break; + case FLOW_TCF_TUNACT_VXLAN_ENCAP: + encap.vxlan = dev_flow->tcf.vxlan_encap; + break; + /* New tunnel actions can be added here. */ + default: + assert(false); + break; + } + } nlh = dev_flow->tcf.nlh; tcm = dev_flow->tcf.tcm; /* Prepare API must have been called beforehand. */ @@ -1870,13 +3276,11 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, * Priority cannot be zero to prevent the kernel from picking one * automatically. */ - tcm->tcm_info = TC_H_MAKE((attr->priority + 1) << 16, - RTE_BE16(ETH_P_ALL)); + tcm->tcm_info = TC_H_MAKE((attr->priority + 1) << 16, outer_etype); if (attr->group > 0) mnl_attr_put_u32(nlh, TCA_CHAIN, attr->group); mnl_attr_put_strz(nlh, TCA_KIND, "flower"); na_flower = mnl_attr_nest_start(nlh, TCA_OPTIONS); - mnl_attr_put_u32(nlh, TCA_FLOWER_FLAGS, TCA_CLS_FLAGS_SKIP_SW); for (; items->type != RTE_FLOW_ITEM_TYPE_END; items++) { unsigned int i; @@ -1904,7 +3308,9 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, tcm->tcm_ifindex = ptoi[i].ifindex; break; case RTE_FLOW_ITEM_TYPE_ETH: - item_flags |= MLX5_FLOW_LAYER_OUTER_L2; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L2 : + MLX5_FLOW_LAYER_OUTER_L2; mask.eth = flow_tcf_item_mask (items, &rte_flow_item_eth_mask, &flow_tcf_mask_supported.eth, @@ -1916,28 +3322,40 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, break; spec.eth = items->spec; if (mask.eth->type) { - mnl_attr_put_u16(nlh, TCA_FLOWER_KEY_ETH_TYPE, - spec.eth->type); - eth_type_set = 1; + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) + inner_etype = spec.eth->type; + else + outer_etype = spec.eth->type; + } + if (tunnel_outer) { + DRV_LOG(WARNING, + "outer L2 addresses cannot be" + " forced is outer ones for tunnel," + " parameter is ignored"); + break; } - if (!is_zero_ether_addr(&mask.eth->dst)) { + if (!rte_is_zero_ether_addr(&mask.eth->dst)) { mnl_attr_put(nlh, TCA_FLOWER_KEY_ETH_DST, - ETHER_ADDR_LEN, + RTE_ETHER_ADDR_LEN, spec.eth->dst.addr_bytes); mnl_attr_put(nlh, TCA_FLOWER_KEY_ETH_DST_MASK, - ETHER_ADDR_LEN, + RTE_ETHER_ADDR_LEN, mask.eth->dst.addr_bytes); } - if (!is_zero_ether_addr(&mask.eth->src)) { + if (!rte_is_zero_ether_addr(&mask.eth->src)) { mnl_attr_put(nlh, TCA_FLOWER_KEY_ETH_SRC, - ETHER_ADDR_LEN, + RTE_ETHER_ADDR_LEN, spec.eth->src.addr_bytes); mnl_attr_put(nlh, TCA_FLOWER_KEY_ETH_SRC_MASK, - ETHER_ADDR_LEN, + RTE_ETHER_ADDR_LEN, mask.eth->src.addr_bytes); } + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); break; case RTE_FLOW_ITEM_TYPE_VLAN: + assert(!encap.hdr); + assert(!decap.hdr); + assert(!tunnel_outer); item_flags |= MLX5_FLOW_LAYER_OUTER_VLAN; mask.vlan = flow_tcf_item_mask (items, &rte_flow_item_vlan_mask, @@ -1946,20 +3364,14 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, sizeof(flow_tcf_mask_supported.vlan), error); assert(mask.vlan); - if (!eth_type_set) - mnl_attr_put_u16(nlh, TCA_FLOWER_KEY_ETH_TYPE, - RTE_BE16(ETH_P_8021Q)); - eth_type_set = 1; - vlan_present = 1; if (mask.vlan == &flow_tcf_mask_empty.vlan) break; spec.vlan = items->spec; - if (mask.vlan->inner_type) { - mnl_attr_put_u16(nlh, - TCA_FLOWER_KEY_VLAN_ETH_TYPE, - spec.vlan->inner_type); - vlan_eth_type_set = 1; - } + assert(outer_etype == RTE_BE16(ETH_P_ALL) || + outer_etype == RTE_BE16(ETH_P_8021Q)); + outer_etype = RTE_BE16(ETH_P_8021Q); + if (mask.vlan->inner_type) + vlan_etype = spec.vlan->inner_type; if (mask.vlan->tci & RTE_BE16(0xe000)) mnl_attr_put_u8(nlh, TCA_FLOWER_KEY_VLAN_PRIO, (rte_be_to_cpu_16 @@ -1969,9 +3381,12 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, rte_be_to_cpu_16 (spec.vlan->tci & RTE_BE16(0x0fff))); + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); break; case RTE_FLOW_ITEM_TYPE_IPV4: - item_flags |= MLX5_FLOW_LAYER_OUTER_L3_IPV4; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L3_IPV4 : + MLX5_FLOW_LAYER_OUTER_L3_IPV4; mask.ipv4 = flow_tcf_item_mask (items, &rte_flow_item_ipv4_mask, &flow_tcf_mask_supported.ipv4, @@ -1979,39 +3394,108 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, sizeof(flow_tcf_mask_supported.ipv4), error); assert(mask.ipv4); - if (!eth_type_set || !vlan_eth_type_set) - mnl_attr_put_u16(nlh, - vlan_present ? - TCA_FLOWER_KEY_VLAN_ETH_TYPE : - TCA_FLOWER_KEY_ETH_TYPE, - RTE_BE16(ETH_P_IP)); - eth_type_set = 1; - vlan_eth_type_set = 1; - if (mask.ipv4 == &flow_tcf_mask_empty.ipv4) - break; + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) { + assert(inner_etype == RTE_BE16(ETH_P_ALL) || + inner_etype == RTE_BE16(ETH_P_IP)); + inner_etype = RTE_BE16(ETH_P_IP); + } else if (outer_etype == RTE_BE16(ETH_P_8021Q)) { + assert(vlan_etype == RTE_BE16(ETH_P_ALL) || + vlan_etype == RTE_BE16(ETH_P_IP)); + vlan_etype = RTE_BE16(ETH_P_IP); + } else { + assert(outer_etype == RTE_BE16(ETH_P_ALL) || + outer_etype == RTE_BE16(ETH_P_IP)); + outer_etype = RTE_BE16(ETH_P_IP); + } spec.ipv4 = items->spec; - if (mask.ipv4->hdr.next_proto_id) { + if (!tunnel_outer && mask.ipv4->hdr.next_proto_id) { + /* + * No way to set IP protocol for outer tunnel + * layers. Usually it is fixed, for example, + * to UDP for VXLAN/GPE. + */ + assert(spec.ipv4); /* Mask is not empty. */ mnl_attr_put_u8(nlh, TCA_FLOWER_KEY_IP_PROTO, spec.ipv4->hdr.next_proto_id); ip_proto_set = 1; } + if (mask.ipv4 == &flow_tcf_mask_empty.ipv4 || + (!mask.ipv4->hdr.src_addr && + !mask.ipv4->hdr.dst_addr)) { + if (!tunnel_outer) + break; + /* + * For tunnel outer we must set outer IP key + * anyway, even if the specification/mask is + * empty. There is no another way to tell + * kernel about he outer layer protocol. + */ + mnl_attr_put_u32 + (nlh, TCA_FLOWER_KEY_ENC_IPV4_SRC, + mask.ipv4->hdr.src_addr); + mnl_attr_put_u32 + (nlh, TCA_FLOWER_KEY_ENC_IPV4_SRC_MASK, + mask.ipv4->hdr.src_addr); + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); + break; + } if (mask.ipv4->hdr.src_addr) { - mnl_attr_put_u32(nlh, TCA_FLOWER_KEY_IPV4_SRC, - spec.ipv4->hdr.src_addr); - mnl_attr_put_u32(nlh, - TCA_FLOWER_KEY_IPV4_SRC_MASK, - mask.ipv4->hdr.src_addr); + mnl_attr_put_u32 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IPV4_SRC : + TCA_FLOWER_KEY_IPV4_SRC, + spec.ipv4->hdr.src_addr); + mnl_attr_put_u32 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IPV4_SRC_MASK : + TCA_FLOWER_KEY_IPV4_SRC_MASK, + mask.ipv4->hdr.src_addr); } if (mask.ipv4->hdr.dst_addr) { - mnl_attr_put_u32(nlh, TCA_FLOWER_KEY_IPV4_DST, - spec.ipv4->hdr.dst_addr); - mnl_attr_put_u32(nlh, - TCA_FLOWER_KEY_IPV4_DST_MASK, - mask.ipv4->hdr.dst_addr); + mnl_attr_put_u32 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IPV4_DST : + TCA_FLOWER_KEY_IPV4_DST, + spec.ipv4->hdr.dst_addr); + mnl_attr_put_u32 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IPV4_DST_MASK : + TCA_FLOWER_KEY_IPV4_DST_MASK, + mask.ipv4->hdr.dst_addr); + } + if (mask.ipv4->hdr.time_to_live) { + mnl_attr_put_u8 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IP_TTL : + TCA_FLOWER_KEY_IP_TTL, + spec.ipv4->hdr.time_to_live); + mnl_attr_put_u8 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IP_TTL_MASK : + TCA_FLOWER_KEY_IP_TTL_MASK, + mask.ipv4->hdr.time_to_live); } + if (mask.ipv4->hdr.type_of_service) { + mnl_attr_put_u8 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IP_TOS : + TCA_FLOWER_KEY_IP_TOS, + spec.ipv4->hdr.type_of_service); + mnl_attr_put_u8 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IP_TOS_MASK : + TCA_FLOWER_KEY_IP_TOS_MASK, + mask.ipv4->hdr.type_of_service); + } + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); break; - case RTE_FLOW_ITEM_TYPE_IPV6: - item_flags |= MLX5_FLOW_LAYER_OUTER_L3_IPV6; + case RTE_FLOW_ITEM_TYPE_IPV6: { + bool ipv6_src, ipv6_dst; + uint8_t msk6, tos6; + + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L3_IPV6 : + MLX5_FLOW_LAYER_OUTER_L3_IPV6; mask.ipv6 = flow_tcf_item_mask (items, &rte_flow_item_ipv6_mask, &flow_tcf_mask_supported.ipv6, @@ -2019,41 +3503,114 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, sizeof(flow_tcf_mask_supported.ipv6), error); assert(mask.ipv6); - if (!eth_type_set || !vlan_eth_type_set) - mnl_attr_put_u16(nlh, - vlan_present ? - TCA_FLOWER_KEY_VLAN_ETH_TYPE : - TCA_FLOWER_KEY_ETH_TYPE, - RTE_BE16(ETH_P_IPV6)); - eth_type_set = 1; - vlan_eth_type_set = 1; - if (mask.ipv6 == &flow_tcf_mask_empty.ipv6) - break; + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) { + assert(inner_etype == RTE_BE16(ETH_P_ALL) || + inner_etype == RTE_BE16(ETH_P_IPV6)); + inner_etype = RTE_BE16(ETH_P_IPV6); + } else if (outer_etype == RTE_BE16(ETH_P_8021Q)) { + assert(vlan_etype == RTE_BE16(ETH_P_ALL) || + vlan_etype == RTE_BE16(ETH_P_IPV6)); + vlan_etype = RTE_BE16(ETH_P_IPV6); + } else { + assert(outer_etype == RTE_BE16(ETH_P_ALL) || + outer_etype == RTE_BE16(ETH_P_IPV6)); + outer_etype = RTE_BE16(ETH_P_IPV6); + } spec.ipv6 = items->spec; - if (mask.ipv6->hdr.proto) { + if (!tunnel_outer && mask.ipv6->hdr.proto) { + /* + * No way to set IP protocol for outer tunnel + * layers. Usually it is fixed, for example, + * to UDP for VXLAN/GPE. + */ + assert(spec.ipv6); /* Mask is not empty. */ mnl_attr_put_u8(nlh, TCA_FLOWER_KEY_IP_PROTO, spec.ipv6->hdr.proto); ip_proto_set = 1; } - if (!IN6_IS_ADDR_UNSPECIFIED(mask.ipv6->hdr.src_addr)) { - mnl_attr_put(nlh, TCA_FLOWER_KEY_IPV6_SRC, - sizeof(spec.ipv6->hdr.src_addr), + ipv6_dst = !IN6_IS_ADDR_UNSPECIFIED + (mask.ipv6->hdr.dst_addr); + ipv6_src = !IN6_IS_ADDR_UNSPECIFIED + (mask.ipv6->hdr.src_addr); + if (mask.ipv6 == &flow_tcf_mask_empty.ipv6 || + (!ipv6_dst && !ipv6_src)) { + if (!tunnel_outer) + break; + /* + * For tunnel outer we must set outer IP key + * anyway, even if the specification/mask is + * empty. There is no another way to tell + * kernel about he outer layer protocol. + */ + mnl_attr_put(nlh, + TCA_FLOWER_KEY_ENC_IPV6_SRC, + IPV6_ADDR_LEN, + mask.ipv6->hdr.src_addr); + mnl_attr_put(nlh, + TCA_FLOWER_KEY_ENC_IPV6_SRC_MASK, + IPV6_ADDR_LEN, + mask.ipv6->hdr.src_addr); + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); + break; + } + if (ipv6_src) { + mnl_attr_put(nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IPV6_SRC : + TCA_FLOWER_KEY_IPV6_SRC, + IPV6_ADDR_LEN, spec.ipv6->hdr.src_addr); - mnl_attr_put(nlh, TCA_FLOWER_KEY_IPV6_SRC_MASK, - sizeof(mask.ipv6->hdr.src_addr), + mnl_attr_put(nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IPV6_SRC_MASK : + TCA_FLOWER_KEY_IPV6_SRC_MASK, + IPV6_ADDR_LEN, mask.ipv6->hdr.src_addr); } - if (!IN6_IS_ADDR_UNSPECIFIED(mask.ipv6->hdr.dst_addr)) { - mnl_attr_put(nlh, TCA_FLOWER_KEY_IPV6_DST, - sizeof(spec.ipv6->hdr.dst_addr), + if (ipv6_dst) { + mnl_attr_put(nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IPV6_DST : + TCA_FLOWER_KEY_IPV6_DST, + IPV6_ADDR_LEN, spec.ipv6->hdr.dst_addr); - mnl_attr_put(nlh, TCA_FLOWER_KEY_IPV6_DST_MASK, - sizeof(mask.ipv6->hdr.dst_addr), + mnl_attr_put(nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IPV6_DST_MASK : + TCA_FLOWER_KEY_IPV6_DST_MASK, + IPV6_ADDR_LEN, mask.ipv6->hdr.dst_addr); } + if (mask.ipv6->hdr.hop_limits) { + mnl_attr_put_u8 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IP_TTL : + TCA_FLOWER_KEY_IP_TTL, + spec.ipv6->hdr.hop_limits); + mnl_attr_put_u8 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IP_TTL_MASK : + TCA_FLOWER_KEY_IP_TTL_MASK, + mask.ipv6->hdr.hop_limits); + } + msk6 = (rte_be_to_cpu_32(mask.ipv6->hdr.vtc_flow) >> + IPV6_HDR_TC_SHIFT) & 0xff; + if (msk6) { + tos6 = (rte_be_to_cpu_32 + (spec.ipv6->hdr.vtc_flow) >> + IPV6_HDR_TC_SHIFT) & 0xff; + mnl_attr_put_u8 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IP_TOS : + TCA_FLOWER_KEY_IP_TOS, tos6); + mnl_attr_put_u8 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_IP_TOS_MASK : + TCA_FLOWER_KEY_IP_TOS_MASK, msk6); + } + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); break; + } case RTE_FLOW_ITEM_TYPE_UDP: - item_flags |= MLX5_FLOW_LAYER_OUTER_L4_UDP; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L4_UDP : + MLX5_FLOW_LAYER_OUTER_L4_UDP; mask.udp = flow_tcf_item_mask (items, &rte_flow_item_udp_mask, &flow_tcf_mask_supported.udp, @@ -2061,29 +3618,50 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, sizeof(flow_tcf_mask_supported.udp), error); assert(mask.udp); - if (!ip_proto_set) - mnl_attr_put_u8(nlh, TCA_FLOWER_KEY_IP_PROTO, - IPPROTO_UDP); - if (mask.udp == &flow_tcf_mask_empty.udp) - break; spec.udp = items->spec; + if (!tunnel_outer) { + if (!ip_proto_set) + mnl_attr_put_u8 + (nlh, TCA_FLOWER_KEY_IP_PROTO, + IPPROTO_UDP); + if (mask.udp == &flow_tcf_mask_empty.udp) + break; + } else { + assert(mask.udp != &flow_tcf_mask_empty.udp); + decap.vxlan->udp_port = + rte_be_to_cpu_16 + (spec.udp->hdr.dst_port); + } if (mask.udp->hdr.src_port) { - mnl_attr_put_u16(nlh, TCA_FLOWER_KEY_UDP_SRC, - spec.udp->hdr.src_port); - mnl_attr_put_u16(nlh, - TCA_FLOWER_KEY_UDP_SRC_MASK, - mask.udp->hdr.src_port); + mnl_attr_put_u16 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_UDP_SRC_PORT : + TCA_FLOWER_KEY_UDP_SRC, + spec.udp->hdr.src_port); + mnl_attr_put_u16 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_UDP_SRC_PORT_MASK : + TCA_FLOWER_KEY_UDP_SRC_MASK, + mask.udp->hdr.src_port); } if (mask.udp->hdr.dst_port) { - mnl_attr_put_u16(nlh, TCA_FLOWER_KEY_UDP_DST, - spec.udp->hdr.dst_port); - mnl_attr_put_u16(nlh, - TCA_FLOWER_KEY_UDP_DST_MASK, - mask.udp->hdr.dst_port); + mnl_attr_put_u16 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_UDP_DST_PORT : + TCA_FLOWER_KEY_UDP_DST, + spec.udp->hdr.dst_port); + mnl_attr_put_u16 + (nlh, tunnel_outer ? + TCA_FLOWER_KEY_ENC_UDP_DST_PORT_MASK : + TCA_FLOWER_KEY_UDP_DST_MASK, + mask.udp->hdr.dst_port); } + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); break; case RTE_FLOW_ITEM_TYPE_TCP: - item_flags |= MLX5_FLOW_LAYER_OUTER_L4_TCP; + item_flags |= (item_flags & MLX5_FLOW_LAYER_TUNNEL) ? + MLX5_FLOW_LAYER_INNER_L4_TCP : + MLX5_FLOW_LAYER_OUTER_L4_TCP; mask.tcp = flow_tcf_item_mask (items, &rte_flow_item_tcp_mask, &flow_tcf_mask_supported.tcp, @@ -2123,6 +3701,17 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, rte_cpu_to_be_16 (mask.tcp->hdr.tcp_flags)); } + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); + break; + case RTE_FLOW_ITEM_TYPE_VXLAN: + assert(decap.vxlan); + tunnel_outer = 0; + item_flags |= MLX5_FLOW_LAYER_VXLAN; + spec.vxlan = items->spec; + mnl_attr_put_u32(nlh, + TCA_FLOWER_KEY_ENC_KEY_ID, + vxlan_vni_as_be32(spec.vxlan->vni)); + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); break; default: return rte_flow_error_set(error, ENOTSUP, @@ -2130,6 +3719,34 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, NULL, "item not supported"); } } + /* + * Set the ether_type flower key and tc rule protocol: + * - if there is nor VLAN neither VXLAN the key is taken from + * eth item directly or deduced from L3 items. + * - if there is vlan item then key is fixed to 802.1q. + * - if there is vxlan item then key is set to inner tunnel type. + * - simultaneous vlan and vxlan items are prohibited. + */ + if (outer_etype != RTE_BE16(ETH_P_ALL)) { + tcm->tcm_info = TC_H_MAKE((attr->priority + 1) << 16, + outer_etype); + if (item_flags & MLX5_FLOW_LAYER_TUNNEL) { + if (inner_etype != RTE_BE16(ETH_P_ALL)) + mnl_attr_put_u16(nlh, + TCA_FLOWER_KEY_ETH_TYPE, + inner_etype); + } else { + mnl_attr_put_u16(nlh, + TCA_FLOWER_KEY_ETH_TYPE, + outer_etype); + if (outer_etype == RTE_BE16(ETH_P_8021Q) && + vlan_etype != RTE_BE16(ETH_P_ALL)) + mnl_attr_put_u16(nlh, + TCA_FLOWER_KEY_VLAN_ETH_TYPE, + vlan_etype); + } + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); + } na_flower_act = mnl_attr_nest_start(nlh, TCA_FLOWER_ACT); na_act_index_cur = 1; for (; actions->type != RTE_FLOW_ACTION_TYPE_END; actions++) { @@ -2156,6 +3773,18 @@ flow_tcf_translate(struct rte_eth_dev *dev, struct mlx5_flow *dev_flow, mnl_attr_put_strz(nlh, TCA_ACT_KIND, "mirred"); na_act = mnl_attr_nest_start(nlh, TCA_ACT_OPTIONS); assert(na_act); + if (encap.hdr) { + assert(dev_flow->tcf.tunnel); + dev_flow->tcf.tunnel->ifindex_ptr = + &((struct tc_mirred *) + mnl_attr_get_payload + (mnl_nlmsg_get_payload_tail + (nlh)))->ifindex; + } else if (decap.hdr) { + assert(dev_flow->tcf.tunnel); + dev_flow->tcf.tunnel->ifindex_ptr = + (unsigned int *)&tcm->tcm_ifindex; + } mnl_attr_put(nlh, TCA_MIRRED_PARMS, sizeof(struct tc_mirred), &(struct tc_mirred){ @@ -2273,6 +3902,82 @@ override_na_vlan_priority: conf.of_set_vlan_pcp->vlan_pcp; } break; + case RTE_FLOW_ACTION_TYPE_VXLAN_DECAP: + assert(decap.vxlan); + assert(dev_flow->tcf.tunnel); + dev_flow->tcf.tunnel->ifindex_ptr = + (unsigned int *)&tcm->tcm_ifindex; + na_act_index = + mnl_attr_nest_start(nlh, na_act_index_cur++); + assert(na_act_index); + mnl_attr_put_strz(nlh, TCA_ACT_KIND, "tunnel_key"); + na_act = mnl_attr_nest_start(nlh, TCA_ACT_OPTIONS); + assert(na_act); + mnl_attr_put(nlh, TCA_TUNNEL_KEY_PARMS, + sizeof(struct tc_tunnel_key), + &(struct tc_tunnel_key){ + .action = TC_ACT_PIPE, + .t_action = TCA_TUNNEL_KEY_ACT_RELEASE, + }); + mnl_attr_nest_end(nlh, na_act); + mnl_attr_nest_end(nlh, na_act_index); + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); + break; + case RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP: + assert(encap.vxlan); + flow_tcf_vxlan_encap_parse(actions, encap.vxlan); + na_act_index = + mnl_attr_nest_start(nlh, na_act_index_cur++); + assert(na_act_index); + mnl_attr_put_strz(nlh, TCA_ACT_KIND, "tunnel_key"); + na_act = mnl_attr_nest_start(nlh, TCA_ACT_OPTIONS); + assert(na_act); + mnl_attr_put(nlh, TCA_TUNNEL_KEY_PARMS, + sizeof(struct tc_tunnel_key), + &(struct tc_tunnel_key){ + .action = TC_ACT_PIPE, + .t_action = TCA_TUNNEL_KEY_ACT_SET, + }); + if (encap.vxlan->mask & FLOW_TCF_ENCAP_UDP_DST) + mnl_attr_put_u16(nlh, + TCA_TUNNEL_KEY_ENC_DST_PORT, + encap.vxlan->udp.dst); + if (encap.vxlan->mask & FLOW_TCF_ENCAP_IPV4_SRC) + mnl_attr_put_u32(nlh, + TCA_TUNNEL_KEY_ENC_IPV4_SRC, + encap.vxlan->ipv4.src); + if (encap.vxlan->mask & FLOW_TCF_ENCAP_IPV4_DST) + mnl_attr_put_u32(nlh, + TCA_TUNNEL_KEY_ENC_IPV4_DST, + encap.vxlan->ipv4.dst); + if (encap.vxlan->mask & FLOW_TCF_ENCAP_IPV6_SRC) + mnl_attr_put(nlh, + TCA_TUNNEL_KEY_ENC_IPV6_SRC, + sizeof(encap.vxlan->ipv6.src), + &encap.vxlan->ipv6.src); + if (encap.vxlan->mask & FLOW_TCF_ENCAP_IPV6_DST) + mnl_attr_put(nlh, + TCA_TUNNEL_KEY_ENC_IPV6_DST, + sizeof(encap.vxlan->ipv6.dst), + &encap.vxlan->ipv6.dst); + if (encap.vxlan->mask & FLOW_TCF_ENCAP_IP_TTL) + mnl_attr_put_u8(nlh, + TCA_TUNNEL_KEY_ENC_TTL, + encap.vxlan->ip_ttl_hop); + if (encap.vxlan->mask & FLOW_TCF_ENCAP_IP_TOS) + mnl_attr_put_u8(nlh, + TCA_TUNNEL_KEY_ENC_TOS, + encap.vxlan->ip_tos); + if (encap.vxlan->mask & FLOW_TCF_ENCAP_VXLAN_VNI) + mnl_attr_put_u32(nlh, + TCA_TUNNEL_KEY_ENC_KEY_ID, + vxlan_vni_as_be32 + (encap.vxlan->vxlan.vni)); + mnl_attr_put_u8(nlh, TCA_TUNNEL_KEY_NO_CSUM, 0); + mnl_attr_nest_end(nlh, na_act); + mnl_attr_nest_end(nlh, na_act_index); + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); + break; case RTE_FLOW_ACTION_TYPE_SET_IPV4_SRC: case RTE_FLOW_ACTION_TYPE_SET_IPV4_DST: case RTE_FLOW_ACTION_TYPE_SET_IPV6_SRC: @@ -2299,132 +4004,1794 @@ override_na_vlan_priority: assert(na_flower); assert(na_flower_act); mnl_attr_nest_end(nlh, na_flower_act); + dev_flow->tcf.ptc_flags = mnl_attr_get_payload + (mnl_nlmsg_get_payload_tail(nlh)); + mnl_attr_put_u32(nlh, TCA_FLOWER_FLAGS, decap.vxlan ? + 0 : TCA_CLS_FLAGS_SKIP_SW); mnl_attr_nest_end(nlh, na_flower); + if (dev_flow->tcf.tunnel && dev_flow->tcf.tunnel->ifindex_ptr) + dev_flow->tcf.tunnel->ifindex_org = + *dev_flow->tcf.tunnel->ifindex_ptr; + assert(dev_flow->tcf.nlsize >= nlh->nlmsg_len); return 0; } /** * Send Netlink message with acknowledgment. * - * @param ctx + * @param tcf * Flow context to use. * @param nlh * Message to send. This function always raises the NLM_F_ACK flag before * sending. + * @param[in] cb + * Callback handler for received message. + * @param[in] arg + * Context pointer for callback handler. * * @return * 0 on success, a negative errno value otherwise and rte_errno is set. */ static int -flow_tcf_nl_ack(struct mlx5_flow_tcf_context *ctx, struct nlmsghdr *nlh) +flow_tcf_nl_ack(struct mlx5_flow_tcf_context *tcf, + struct nlmsghdr *nlh, + mnl_cb_t cb, void *arg) { - alignas(struct nlmsghdr) - uint8_t ans[mnl_nlmsg_size(sizeof(struct nlmsgerr)) + - nlh->nlmsg_len - sizeof(*nlh)]; - uint32_t seq = ctx->seq++; - struct mnl_socket *nl = ctx->nl; - int ret; - - nlh->nlmsg_flags |= NLM_F_ACK; + unsigned int portid = mnl_socket_get_portid(tcf->nl); + uint32_t seq = tcf->seq++; + int ret, err = 0; + + assert(tcf->nl); + assert(tcf->buf); + if (!seq) { + /* seq 0 is reserved for kernel event-driven notifications. */ + seq = tcf->seq++; + } nlh->nlmsg_seq = seq; - ret = mnl_socket_sendto(nl, nlh, nlh->nlmsg_len); - if (ret != -1) - ret = mnl_socket_recvfrom(nl, ans, sizeof(ans)); - if (ret != -1) - ret = mnl_cb_run - (ans, ret, seq, mnl_socket_get_portid(nl), NULL, NULL); - if (ret > 0) + nlh->nlmsg_flags |= NLM_F_ACK; + ret = mnl_socket_sendto(tcf->nl, nlh, nlh->nlmsg_len); + if (ret <= 0) { + /* Message send error occurred. */ + rte_errno = errno; + return -rte_errno; + } + nlh = (struct nlmsghdr *)(tcf->buf); + /* + * The following loop postpones non-fatal errors until multipart + * messages are complete. + */ + while (true) { + ret = mnl_socket_recvfrom(tcf->nl, tcf->buf, tcf->buf_size); + if (ret < 0) { + err = errno; + /* + * In case of overflow Will receive till + * end of multipart message. We may lost part + * of reply messages but mark and return an error. + */ + if (err != ENOSPC || + !(nlh->nlmsg_flags & NLM_F_MULTI) || + nlh->nlmsg_type == NLMSG_DONE) + break; + } else { + ret = mnl_cb_run(nlh, ret, seq, portid, cb, arg); + if (!ret) { + /* + * libmnl returns 0 if DONE or + * success ACK message found. + */ + break; + } + if (ret < 0) { + /* + * ACK message with error found + * or some error occurred. + */ + err = errno; + break; + } + /* We should continue receiving. */ + } + } + if (!err) return 0; - rte_errno = errno; - return -rte_errno; + rte_errno = err; + return -err; } +#define MNL_BUF_EXTRA_SPACE 16 +#define MNL_REQUEST_SIZE_MIN 256 +#define MNL_REQUEST_SIZE_MAX 2048 +#define MNL_REQUEST_SIZE RTE_MIN(RTE_MAX(sysconf(_SC_PAGESIZE), \ + MNL_REQUEST_SIZE_MIN), MNL_REQUEST_SIZE_MAX) + +/* Data structures used by flow_tcf_xxx_cb() routines. */ +struct tcf_nlcb_buf { + LIST_ENTRY(tcf_nlcb_buf) next; + uint32_t size; + alignas(struct nlmsghdr) + uint8_t msg[]; /**< Netlink message data. */ +}; + +struct tcf_nlcb_context { + unsigned int ifindex; /**< Base interface index. */ + uint32_t bufsize; + LIST_HEAD(, tcf_nlcb_buf) nlbuf; +}; + /** - * Apply flow to E-Switch by sending Netlink message. + * Allocate space for netlink command in buffer list * - * @param[in] dev - * Pointer to Ethernet device. - * @param[in, out] flow - * Pointer to the sub flow. - * @param[out] error - * Pointer to the error structure. + * @param[in, out] ctx + * Pointer to callback context with command buffers list. + * @param[in] size + * Required size of data buffer to be allocated. + * + * @return + * Pointer to allocated memory, aligned as message header. + * NULL if some error occurred. + */ +static struct nlmsghdr * +flow_tcf_alloc_nlcmd(struct tcf_nlcb_context *ctx, uint32_t size) +{ + struct tcf_nlcb_buf *buf; + struct nlmsghdr *nlh; + + size = NLMSG_ALIGN(size); + buf = LIST_FIRST(&ctx->nlbuf); + if (buf && (buf->size + size) <= ctx->bufsize) { + nlh = (struct nlmsghdr *)&buf->msg[buf->size]; + buf->size += size; + return nlh; + } + if (size > ctx->bufsize) { + DRV_LOG(WARNING, "netlink: too long command buffer requested"); + return NULL; + } + buf = rte_malloc(__func__, + ctx->bufsize + sizeof(struct tcf_nlcb_buf), + alignof(struct tcf_nlcb_buf)); + if (!buf) { + DRV_LOG(WARNING, "netlink: no memory for command buffer"); + return NULL; + } + LIST_INSERT_HEAD(&ctx->nlbuf, buf, next); + buf->size = size; + nlh = (struct nlmsghdr *)&buf->msg[0]; + return nlh; +} + +/** + * Send the buffers with prepared netlink commands. Scans the list and + * sends all found buffers. Buffers are sent and freed anyway in order + * to prevent memory leakage if some every message in received packet. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in, out] ctx + * Pointer to callback context with command buffers list. * * @return - * 0 on success, a negative errno value otherwise and rte_ernno is set. + * Zero value on success, negative errno value otherwise + * and rte_errno is set. */ static int -flow_tcf_apply(struct rte_eth_dev *dev, struct rte_flow *flow, - struct rte_flow_error *error) +flow_tcf_send_nlcmd(struct mlx5_flow_tcf_context *tcf, + struct tcf_nlcb_context *ctx) +{ + struct tcf_nlcb_buf *bc = LIST_FIRST(&ctx->nlbuf); + int ret = 0; + + while (bc) { + struct tcf_nlcb_buf *bn = LIST_NEXT(bc, next); + struct nlmsghdr *nlh; + uint32_t msg = 0; + int rc; + + while (msg < bc->size) { + /* + * Send Netlink commands from buffer in one by one + * fashion. If we send multiple rule deletion commands + * in one Netlink message and some error occurs it may + * cause multiple ACK error messages and break sequence + * numbers of Netlink communication, because we expect + * the only one ACK reply. + */ + assert((bc->size - msg) >= sizeof(struct nlmsghdr)); + nlh = (struct nlmsghdr *)&bc->msg[msg]; + assert((bc->size - msg) >= nlh->nlmsg_len); + msg += nlh->nlmsg_len; + rc = flow_tcf_nl_ack(tcf, nlh, NULL, NULL); + if (rc) { + DRV_LOG(WARNING, + "netlink: cleanup error %d", rc); + if (!ret) + ret = rc; + } + } + rte_free(bc); + bc = bn; + } + LIST_INIT(&ctx->nlbuf); + return ret; +} + +/** + * Collect local IP address rules with scope link attribute on specified + * network device. This is callback routine called by libmnl mnl_cb_run() + * in loop for every message in received packet. + * + * @param[in] nlh + * Pointer to reply header. + * @param[in, out] arg + * Opaque data pointer for this callback. + * + * @return + * A positive, nonzero value on success, negative errno value otherwise + * and rte_errno is set. + */ +static int +flow_tcf_collect_local_cb(const struct nlmsghdr *nlh, void *arg) +{ + struct tcf_nlcb_context *ctx = arg; + struct nlmsghdr *cmd; + struct ifaddrmsg *ifa; + struct nlattr *na; + struct nlattr *na_local = NULL; + struct nlattr *na_peer = NULL; + unsigned char family; + uint32_t size; + + if (nlh->nlmsg_type != RTM_NEWADDR) { + rte_errno = EINVAL; + return -rte_errno; + } + ifa = mnl_nlmsg_get_payload(nlh); + family = ifa->ifa_family; + if (ifa->ifa_index != ctx->ifindex || + ifa->ifa_scope != RT_SCOPE_LINK || + !(ifa->ifa_flags & IFA_F_PERMANENT) || + (family != AF_INET && family != AF_INET6)) + return 1; + mnl_attr_for_each(na, nlh, sizeof(*ifa)) { + switch (mnl_attr_get_type(na)) { + case IFA_LOCAL: + na_local = na; + break; + case IFA_ADDRESS: + na_peer = na; + break; + } + if (na_local && na_peer) + break; + } + if (!na_local || !na_peer) + return 1; + /* Local rule found with scope link, permanent and assigned peer. */ + size = MNL_ALIGN(sizeof(struct nlmsghdr)) + + MNL_ALIGN(sizeof(struct ifaddrmsg)) + + (family == AF_INET6 ? 2 * SZ_NLATTR_DATA_OF(IPV6_ADDR_LEN) + : 2 * SZ_NLATTR_TYPE_OF(uint32_t)); + cmd = flow_tcf_alloc_nlcmd(ctx, size); + if (!cmd) { + rte_errno = ENOMEM; + return -rte_errno; + } + cmd = mnl_nlmsg_put_header(cmd); + cmd->nlmsg_type = RTM_DELADDR; + cmd->nlmsg_flags = NLM_F_REQUEST; + ifa = mnl_nlmsg_put_extra_header(cmd, sizeof(*ifa)); + ifa->ifa_flags = IFA_F_PERMANENT; + ifa->ifa_scope = RT_SCOPE_LINK; + ifa->ifa_index = ctx->ifindex; + if (family == AF_INET) { + ifa->ifa_family = AF_INET; + ifa->ifa_prefixlen = 32; + mnl_attr_put_u32(cmd, IFA_LOCAL, mnl_attr_get_u32(na_local)); + mnl_attr_put_u32(cmd, IFA_ADDRESS, mnl_attr_get_u32(na_peer)); + } else { + ifa->ifa_family = AF_INET6; + ifa->ifa_prefixlen = 128; + mnl_attr_put(cmd, IFA_LOCAL, IPV6_ADDR_LEN, + mnl_attr_get_payload(na_local)); + mnl_attr_put(cmd, IFA_ADDRESS, IPV6_ADDR_LEN, + mnl_attr_get_payload(na_peer)); + } + assert(size == cmd->nlmsg_len); + return 1; +} + +/** + * Cleanup the local IP addresses on outer interface. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] ifindex + * Network interface index to perform cleanup. + */ +static void +flow_tcf_encap_local_cleanup(struct mlx5_flow_tcf_context *tcf, + unsigned int ifindex) { - struct priv *priv = dev->data->dev_private; - struct mlx5_flow_tcf_context *ctx = priv->tcf_context; - struct mlx5_flow *dev_flow; struct nlmsghdr *nlh; + struct ifaddrmsg *ifa; + struct tcf_nlcb_context ctx = { + .ifindex = ifindex, + .bufsize = MNL_REQUEST_SIZE, + .nlbuf = LIST_HEAD_INITIALIZER(), + }; + int ret; - dev_flow = LIST_FIRST(&flow->dev_flows); - /* E-Switch flow can't be expanded. */ - assert(!LIST_NEXT(dev_flow, next)); - nlh = dev_flow->tcf.nlh; - nlh->nlmsg_type = RTM_NEWTFILTER; - nlh->nlmsg_flags = NLM_F_REQUEST | NLM_F_CREATE | NLM_F_EXCL; - if (!flow_tcf_nl_ack(ctx, nlh)) - return 0; - return rte_flow_error_set(error, rte_errno, - RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, - "netlink: failed to create TC flow rule"); + assert(ifindex); + /* + * Seek and destroy leftovers of local IP addresses with + * matching properties "scope link". + */ + nlh = mnl_nlmsg_put_header(tcf->buf); + nlh->nlmsg_type = RTM_GETADDR; + nlh->nlmsg_flags = NLM_F_REQUEST | NLM_F_DUMP; + ifa = mnl_nlmsg_put_extra_header(nlh, sizeof(*ifa)); + ifa->ifa_family = AF_UNSPEC; + ifa->ifa_index = ifindex; + ifa->ifa_scope = RT_SCOPE_LINK; + ret = flow_tcf_nl_ack(tcf, nlh, flow_tcf_collect_local_cb, &ctx); + if (ret) + DRV_LOG(WARNING, "netlink: query device list error %d", ret); + ret = flow_tcf_send_nlcmd(tcf, &ctx); + if (ret) + DRV_LOG(WARNING, "netlink: device delete error %d", ret); } /** - * Remove flow from E-Switch by sending Netlink message. + * Collect neigh permanent rules on specified network device. + * This is callback routine called by libmnl mnl_cb_run() in loop for + * every message in received packet. * - * @param[in] dev - * Pointer to Ethernet device. - * @param[in, out] flow - * Pointer to the sub flow. + * @param[in] nlh + * Pointer to reply header. + * @param[in, out] arg + * Opaque data pointer for this callback. + * + * @return + * A positive, nonzero value on success, negative errno value otherwise + * and rte_errno is set. + */ +static int +flow_tcf_collect_neigh_cb(const struct nlmsghdr *nlh, void *arg) +{ + struct tcf_nlcb_context *ctx = arg; + struct nlmsghdr *cmd; + struct ndmsg *ndm; + struct nlattr *na; + struct nlattr *na_ip = NULL; + struct nlattr *na_mac = NULL; + unsigned char family; + uint32_t size; + + if (nlh->nlmsg_type != RTM_NEWNEIGH) { + rte_errno = EINVAL; + return -rte_errno; + } + ndm = mnl_nlmsg_get_payload(nlh); + family = ndm->ndm_family; + if (ndm->ndm_ifindex != (int)ctx->ifindex || + !(ndm->ndm_state & NUD_PERMANENT) || + (family != AF_INET && family != AF_INET6)) + return 1; + mnl_attr_for_each(na, nlh, sizeof(*ndm)) { + switch (mnl_attr_get_type(na)) { + case NDA_DST: + na_ip = na; + break; + case NDA_LLADDR: + na_mac = na; + break; + } + if (na_mac && na_ip) + break; + } + if (!na_mac || !na_ip) + return 1; + /* Neigh rule with permanent attribute found. */ + size = MNL_ALIGN(sizeof(struct nlmsghdr)) + + MNL_ALIGN(sizeof(struct ndmsg)) + + SZ_NLATTR_DATA_OF(RTE_ETHER_ADDR_LEN) + + (family == AF_INET6 ? SZ_NLATTR_DATA_OF(IPV6_ADDR_LEN) + : SZ_NLATTR_TYPE_OF(uint32_t)); + cmd = flow_tcf_alloc_nlcmd(ctx, size); + if (!cmd) { + rte_errno = ENOMEM; + return -rte_errno; + } + cmd = mnl_nlmsg_put_header(cmd); + cmd->nlmsg_type = RTM_DELNEIGH; + cmd->nlmsg_flags = NLM_F_REQUEST; + ndm = mnl_nlmsg_put_extra_header(cmd, sizeof(*ndm)); + ndm->ndm_ifindex = ctx->ifindex; + ndm->ndm_state = NUD_PERMANENT; + ndm->ndm_flags = 0; + ndm->ndm_type = 0; + if (family == AF_INET) { + ndm->ndm_family = AF_INET; + mnl_attr_put_u32(cmd, NDA_DST, mnl_attr_get_u32(na_ip)); + } else { + ndm->ndm_family = AF_INET6; + mnl_attr_put(cmd, NDA_DST, IPV6_ADDR_LEN, + mnl_attr_get_payload(na_ip)); + } + mnl_attr_put(cmd, NDA_LLADDR, RTE_ETHER_ADDR_LEN, + mnl_attr_get_payload(na_mac)); + assert(size == cmd->nlmsg_len); + return 1; +} + +/** + * Cleanup the neigh rules on outer interface. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] ifindex + * Network interface index to perform cleanup. */ static void -flow_tcf_remove(struct rte_eth_dev *dev, struct rte_flow *flow) +flow_tcf_encap_neigh_cleanup(struct mlx5_flow_tcf_context *tcf, + unsigned int ifindex) { - struct priv *priv = dev->data->dev_private; - struct mlx5_flow_tcf_context *ctx = priv->tcf_context; - struct mlx5_flow *dev_flow; struct nlmsghdr *nlh; + struct ndmsg *ndm; + struct tcf_nlcb_context ctx = { + .ifindex = ifindex, + .bufsize = MNL_REQUEST_SIZE, + .nlbuf = LIST_HEAD_INITIALIZER(), + }; + int ret; - if (!flow) - return; - if (flow->counter) { - if (--flow->counter->ref_cnt == 0) { - rte_free(flow->counter); - flow->counter = NULL; + assert(ifindex); + /* Seek and destroy leftovers of neigh rules. */ + nlh = mnl_nlmsg_put_header(tcf->buf); + nlh->nlmsg_type = RTM_GETNEIGH; + nlh->nlmsg_flags = NLM_F_REQUEST | NLM_F_DUMP; + ndm = mnl_nlmsg_put_extra_header(nlh, sizeof(*ndm)); + ndm->ndm_family = AF_UNSPEC; + ndm->ndm_ifindex = ifindex; + ndm->ndm_state = NUD_PERMANENT; + ret = flow_tcf_nl_ack(tcf, nlh, flow_tcf_collect_neigh_cb, &ctx); + if (ret) + DRV_LOG(WARNING, "netlink: query device list error %d", ret); + ret = flow_tcf_send_nlcmd(tcf, &ctx); + if (ret) + DRV_LOG(WARNING, "netlink: device delete error %d", ret); +} + +/** + * Collect indices of VXLAN encap/decap interfaces associated with device. + * This is callback routine called by libmnl mnl_cb_run() in loop for + * every message in received packet. + * + * @param[in] nlh + * Pointer to reply header. + * @param[in, out] arg + * Opaque data pointer for this callback. + * + * @return + * A positive, nonzero value on success, negative errno value otherwise + * and rte_errno is set. + */ +static int +flow_tcf_collect_vxlan_cb(const struct nlmsghdr *nlh, void *arg) +{ + struct tcf_nlcb_context *ctx = arg; + struct nlmsghdr *cmd; + struct ifinfomsg *ifm; + struct nlattr *na; + struct nlattr *na_info = NULL; + struct nlattr *na_vxlan = NULL; + bool found = false; + unsigned int vxindex; + uint32_t size; + + if (nlh->nlmsg_type != RTM_NEWLINK) { + rte_errno = EINVAL; + return -rte_errno; + } + ifm = mnl_nlmsg_get_payload(nlh); + if (!ifm->ifi_index) { + rte_errno = EINVAL; + return -rte_errno; + } + mnl_attr_for_each(na, nlh, sizeof(*ifm)) + if (mnl_attr_get_type(na) == IFLA_LINKINFO) { + na_info = na; + break; + } + if (!na_info) + return 1; + mnl_attr_for_each_nested(na, na_info) { + switch (mnl_attr_get_type(na)) { + case IFLA_INFO_KIND: + if (!strncmp("vxlan", mnl_attr_get_str(na), + mnl_attr_get_len(na))) + found = true; + break; + case IFLA_INFO_DATA: + na_vxlan = na; + break; } + if (found && na_vxlan) + break; } - dev_flow = LIST_FIRST(&flow->dev_flows); - if (!dev_flow) - return; - /* E-Switch flow can't be expanded. */ - assert(!LIST_NEXT(dev_flow, next)); - nlh = dev_flow->tcf.nlh; - nlh->nlmsg_type = RTM_DELTFILTER; - nlh->nlmsg_flags = NLM_F_REQUEST; - flow_tcf_nl_ack(ctx, nlh); + if (!found || !na_vxlan) + return 1; + found = false; + mnl_attr_for_each_nested(na, na_vxlan) { + if (mnl_attr_get_type(na) == IFLA_VXLAN_LINK && + mnl_attr_get_u32(na) == ctx->ifindex) { + found = true; + break; + } + } + if (!found) + return 1; + /* Attached VXLAN device found, store the command to delete. */ + vxindex = ifm->ifi_index; + size = MNL_ALIGN(sizeof(struct nlmsghdr)) + + MNL_ALIGN(sizeof(struct ifinfomsg)); + cmd = flow_tcf_alloc_nlcmd(ctx, size); + if (!cmd) { + rte_errno = ENOMEM; + return -rte_errno; + } + cmd = mnl_nlmsg_put_header(cmd); + cmd->nlmsg_type = RTM_DELLINK; + cmd->nlmsg_flags = NLM_F_REQUEST; + ifm = mnl_nlmsg_put_extra_header(cmd, sizeof(*ifm)); + ifm->ifi_family = AF_UNSPEC; + ifm->ifi_index = vxindex; + assert(size == cmd->nlmsg_len); + return 1; } /** - * Remove flow from E-Switch and release resources of the device flow. + * Cleanup the outer interface. Removes all found vxlan devices + * attached to specified index, flushes the neigh and local IP + * database. * - * @param[in] dev - * Pointer to Ethernet device. - * @param[in, out] flow - * Pointer to the sub flow. + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] ifindex + * Network inferface index to perform cleanup. */ static void -flow_tcf_destroy(struct rte_eth_dev *dev, struct rte_flow *flow) +flow_tcf_encap_iface_cleanup(struct mlx5_flow_tcf_context *tcf, + unsigned int ifindex) { - struct mlx5_flow *dev_flow; + struct nlmsghdr *nlh; + struct ifinfomsg *ifm; + struct tcf_nlcb_context ctx = { + .ifindex = ifindex, + .bufsize = MNL_REQUEST_SIZE, + .nlbuf = LIST_HEAD_INITIALIZER(), + }; + int ret; - if (!flow) - return; - flow_tcf_remove(dev, flow); + assert(ifindex); + /* + * Seek and destroy leftover VXLAN encap/decap interfaces with + * matching properties. + */ + nlh = mnl_nlmsg_put_header(tcf->buf); + nlh->nlmsg_type = RTM_GETLINK; + nlh->nlmsg_flags = NLM_F_REQUEST | NLM_F_DUMP; + ifm = mnl_nlmsg_put_extra_header(nlh, sizeof(*ifm)); + ifm->ifi_family = AF_UNSPEC; + ret = flow_tcf_nl_ack(tcf, nlh, flow_tcf_collect_vxlan_cb, &ctx); + if (ret) + DRV_LOG(WARNING, "netlink: query device list error %d", ret); + ret = flow_tcf_send_nlcmd(tcf, &ctx); + if (ret) + DRV_LOG(WARNING, "netlink: device delete error %d", ret); +} + +/** + * Emit Netlink message to add/remove local address to the outer device. + * The address being added is visible within the link only (scope link). + * + * Note that an implicit route is maintained by the kernel due to the + * presence of a peer address (IFA_ADDRESS). + * + * These rules are used for encapsulation only and allow to assign + * the outer tunnel source IP address. + * + * @param[in] tcf + * Libmnl socket context object. + * @param[in] encap + * Encapsulation properties (source address and its peer). + * @param[in] ifindex + * Network interface to apply rule. + * @param[in] enable + * Toggle between add and remove. + * @param[out] error + * Perform verbose error reporting if not NULL. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + */ +static int +flow_tcf_rule_local(struct mlx5_flow_tcf_context *tcf, + const struct flow_tcf_vxlan_encap *encap, + unsigned int ifindex, + bool enable, + struct rte_flow_error *error) +{ + struct nlmsghdr *nlh; + struct ifaddrmsg *ifa; + alignas(struct nlmsghdr) + uint8_t buf[mnl_nlmsg_size(sizeof(*ifa) + 128)]; + + nlh = mnl_nlmsg_put_header(buf); + nlh->nlmsg_type = enable ? RTM_NEWADDR : RTM_DELADDR; + nlh->nlmsg_flags = + NLM_F_REQUEST | (enable ? NLM_F_CREATE | NLM_F_REPLACE : 0); + nlh->nlmsg_seq = 0; + ifa = mnl_nlmsg_put_extra_header(nlh, sizeof(*ifa)); + ifa->ifa_flags = IFA_F_PERMANENT; + ifa->ifa_scope = RT_SCOPE_LINK; + ifa->ifa_index = ifindex; + if (encap->mask & FLOW_TCF_ENCAP_IPV4_SRC) { + ifa->ifa_family = AF_INET; + ifa->ifa_prefixlen = 32; + mnl_attr_put_u32(nlh, IFA_LOCAL, encap->ipv4.src); + if (encap->mask & FLOW_TCF_ENCAP_IPV4_DST) + mnl_attr_put_u32(nlh, IFA_ADDRESS, + encap->ipv4.dst); + } else { + assert(encap->mask & FLOW_TCF_ENCAP_IPV6_SRC); + ifa->ifa_family = AF_INET6; + ifa->ifa_prefixlen = 128; + mnl_attr_put(nlh, IFA_LOCAL, + sizeof(encap->ipv6.src), + &encap->ipv6.src); + if (encap->mask & FLOW_TCF_ENCAP_IPV6_DST) + mnl_attr_put(nlh, IFA_ADDRESS, + sizeof(encap->ipv6.dst), + &encap->ipv6.dst); + } + if (!flow_tcf_nl_ack(tcf, nlh, NULL, NULL)) + return 0; + return rte_flow_error_set(error, rte_errno, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "netlink: cannot complete IFA request" + " (ip addr add)"); +} + +/** + * Emit Netlink message to add/remove neighbor. + * + * @param[in] tcf + * Libmnl socket context object. + * @param[in] encap + * Encapsulation properties (destination address). + * @param[in] ifindex + * Network interface. + * @param[in] enable + * Toggle between add and remove. + * @param[out] error + * Perform verbose error reporting if not NULL. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + */ +static int +flow_tcf_rule_neigh(struct mlx5_flow_tcf_context *tcf, + const struct flow_tcf_vxlan_encap *encap, + unsigned int ifindex, + bool enable, + struct rte_flow_error *error) +{ + struct nlmsghdr *nlh; + struct ndmsg *ndm; + alignas(struct nlmsghdr) + uint8_t buf[mnl_nlmsg_size(sizeof(*ndm) + 128)]; + + nlh = mnl_nlmsg_put_header(buf); + nlh->nlmsg_type = enable ? RTM_NEWNEIGH : RTM_DELNEIGH; + nlh->nlmsg_flags = + NLM_F_REQUEST | (enable ? NLM_F_CREATE | NLM_F_REPLACE : 0); + nlh->nlmsg_seq = 0; + ndm = mnl_nlmsg_put_extra_header(nlh, sizeof(*ndm)); + ndm->ndm_ifindex = ifindex; + ndm->ndm_state = NUD_PERMANENT; + ndm->ndm_flags = 0; + ndm->ndm_type = 0; + if (encap->mask & FLOW_TCF_ENCAP_IPV4_DST) { + ndm->ndm_family = AF_INET; + mnl_attr_put_u32(nlh, NDA_DST, encap->ipv4.dst); + } else { + assert(encap->mask & FLOW_TCF_ENCAP_IPV6_DST); + ndm->ndm_family = AF_INET6; + mnl_attr_put(nlh, NDA_DST, sizeof(encap->ipv6.dst), + &encap->ipv6.dst); + } + if (encap->mask & FLOW_TCF_ENCAP_ETH_SRC && enable) + DRV_LOG(WARNING, + "outer ethernet source address cannot be " + "forced for VXLAN encapsulation"); + if (encap->mask & FLOW_TCF_ENCAP_ETH_DST) + mnl_attr_put(nlh, NDA_LLADDR, sizeof(encap->eth.dst), + &encap->eth.dst); + if (!flow_tcf_nl_ack(tcf, nlh, NULL, NULL)) + return 0; + return rte_flow_error_set(error, rte_errno, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "netlink: cannot complete ND request" + " (ip neigh)"); +} + +/** + * Manage the local IP addresses and their peers IP addresses on the + * outer interface for encapsulation purposes. The kernel searches the + * appropriate device for tunnel egress traffic using the outer source + * IP, this IP should be assigned to the outer network device, otherwise + * kernel rejects the rule. + * + * Adds or removes the addresses using the Netlink command like this: + * ip addr add peer scope link dev + * + * The addresses are local to the netdev ("scope link"), this reduces + * the risk of conflicts. Note that an implicit route is maintained by + * the kernel due to the presence of a peer address (IFA_ADDRESS). + * + * @param[in] tcf + * Libmnl socket context object. + * @param[in] iface + * Object, contains rule database and ifouter index. + * @param[in] dev_flow + * Flow object, contains the tunnel parameters (for encap only). + * @param[in] enable + * Toggle between add and remove. + * @param[out] error + * Perform verbose error reporting if not NULL. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + */ +static int +flow_tcf_encap_local(struct mlx5_flow_tcf_context *tcf, + struct tcf_irule *iface, + struct mlx5_flow *dev_flow, + bool enable, + struct rte_flow_error *error) +{ + const struct flow_tcf_vxlan_encap *encap = dev_flow->tcf.vxlan_encap; + struct tcf_local_rule *rule = NULL; + int ret; + + assert(encap); + assert(encap->hdr.type == FLOW_TCF_TUNACT_VXLAN_ENCAP); + if (encap->mask & FLOW_TCF_ENCAP_IPV4_SRC) { + assert(encap->mask & FLOW_TCF_ENCAP_IPV4_DST); + LIST_FOREACH(rule, &iface->local, next) { + if (rule->mask & FLOW_TCF_ENCAP_IPV4_SRC && + encap->ipv4.src == rule->ipv4.src && + encap->ipv4.dst == rule->ipv4.dst) { + break; + } + } + } else { + assert(encap->mask & FLOW_TCF_ENCAP_IPV6_SRC); + assert(encap->mask & FLOW_TCF_ENCAP_IPV6_DST); + LIST_FOREACH(rule, &iface->local, next) { + if (rule->mask & FLOW_TCF_ENCAP_IPV6_SRC && + !memcmp(&encap->ipv6.src, &rule->ipv6.src, + sizeof(encap->ipv6.src)) && + !memcmp(&encap->ipv6.dst, &rule->ipv6.dst, + sizeof(encap->ipv6.dst))) { + break; + } + } + } + if (rule) { + if (enable) { + rule->refcnt++; + return 0; + } + if (!rule->refcnt || !--rule->refcnt) { + LIST_REMOVE(rule, next); + return flow_tcf_rule_local(tcf, encap, + iface->ifouter, false, error); + } + return 0; + } + if (!enable) { + DRV_LOG(WARNING, "disabling not existing local rule"); + rte_flow_error_set(error, ENOENT, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "disabling not existing local rule"); + return -ENOENT; + } + rule = rte_zmalloc(__func__, sizeof(struct tcf_local_rule), + alignof(struct tcf_local_rule)); + if (!rule) { + rte_flow_error_set(error, ENOMEM, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "unable to allocate memory for local rule"); + return -rte_errno; + } + *rule = (struct tcf_local_rule){.refcnt = 0, + .mask = 0, + }; + if (encap->mask & FLOW_TCF_ENCAP_IPV4_SRC) { + rule->mask = FLOW_TCF_ENCAP_IPV4_SRC + | FLOW_TCF_ENCAP_IPV4_DST; + rule->ipv4.src = encap->ipv4.src; + rule->ipv4.dst = encap->ipv4.dst; + } else { + rule->mask = FLOW_TCF_ENCAP_IPV6_SRC + | FLOW_TCF_ENCAP_IPV6_DST; + memcpy(&rule->ipv6.src, &encap->ipv6.src, IPV6_ADDR_LEN); + memcpy(&rule->ipv6.dst, &encap->ipv6.dst, IPV6_ADDR_LEN); + } + ret = flow_tcf_rule_local(tcf, encap, iface->ifouter, true, error); + if (ret) { + rte_free(rule); + return ret; + } + rule->refcnt++; + LIST_INSERT_HEAD(&iface->local, rule, next); + return 0; +} + +/** + * Manage the destination MAC/IP addresses neigh database, kernel uses + * this one to determine the destination MAC address within encapsulation + * header. Adds or removes the entries using the Netlink command like this: + * ip neigh add dev lladdr to nud permanent + * + * @param[in] tcf + * Libmnl socket context object. + * @param[in] iface + * Object, contains rule database and ifouter index. + * @param[in] dev_flow + * Flow object, contains the tunnel parameters (for encap only). + * @param[in] enable + * Toggle between add and remove. + * @param[out] error + * Perform verbose error reporting if not NULL. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + */ +static int +flow_tcf_encap_neigh(struct mlx5_flow_tcf_context *tcf, + struct tcf_irule *iface, + struct mlx5_flow *dev_flow, + bool enable, + struct rte_flow_error *error) +{ + const struct flow_tcf_vxlan_encap *encap = dev_flow->tcf.vxlan_encap; + struct tcf_neigh_rule *rule = NULL; + int ret; + + assert(encap); + assert(encap->hdr.type == FLOW_TCF_TUNACT_VXLAN_ENCAP); + if (encap->mask & FLOW_TCF_ENCAP_IPV4_DST) { + assert(encap->mask & FLOW_TCF_ENCAP_IPV4_SRC); + LIST_FOREACH(rule, &iface->neigh, next) { + if (rule->mask & FLOW_TCF_ENCAP_IPV4_DST && + encap->ipv4.dst == rule->ipv4.dst) { + break; + } + } + } else { + assert(encap->mask & FLOW_TCF_ENCAP_IPV6_SRC); + assert(encap->mask & FLOW_TCF_ENCAP_IPV6_DST); + LIST_FOREACH(rule, &iface->neigh, next) { + if (rule->mask & FLOW_TCF_ENCAP_IPV6_DST && + !memcmp(&encap->ipv6.dst, &rule->ipv6.dst, + sizeof(encap->ipv6.dst))) { + break; + } + } + } + if (rule) { + if (memcmp(&encap->eth.dst, &rule->eth, + sizeof(encap->eth.dst))) { + DRV_LOG(WARNING, "Destination MAC differs" + " in neigh rule"); + rte_flow_error_set(error, EEXIST, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, + NULL, "Different MAC address" + " neigh rule for the same" + " destination IP"); + return -EEXIST; + } + if (enable) { + rule->refcnt++; + return 0; + } + if (!rule->refcnt || !--rule->refcnt) { + LIST_REMOVE(rule, next); + return flow_tcf_rule_neigh(tcf, encap, + iface->ifouter, + false, error); + } + return 0; + } + if (!enable) { + DRV_LOG(WARNING, "Disabling not existing neigh rule"); + rte_flow_error_set(error, ENOENT, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "unable to allocate memory for neigh rule"); + return -ENOENT; + } + rule = rte_zmalloc(__func__, sizeof(struct tcf_neigh_rule), + alignof(struct tcf_neigh_rule)); + if (!rule) { + rte_flow_error_set(error, ENOMEM, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "unable to allocate memory for neigh rule"); + return -rte_errno; + } + *rule = (struct tcf_neigh_rule){.refcnt = 0, + .mask = 0, + }; + if (encap->mask & FLOW_TCF_ENCAP_IPV4_DST) { + rule->mask = FLOW_TCF_ENCAP_IPV4_DST; + rule->ipv4.dst = encap->ipv4.dst; + } else { + rule->mask = FLOW_TCF_ENCAP_IPV6_DST; + memcpy(&rule->ipv6.dst, &encap->ipv6.dst, IPV6_ADDR_LEN); + } + memcpy(&rule->eth, &encap->eth.dst, sizeof(rule->eth)); + ret = flow_tcf_rule_neigh(tcf, encap, iface->ifouter, true, error); + if (ret) { + rte_free(rule); + return ret; + } + rule->refcnt++; + LIST_INSERT_HEAD(&iface->neigh, rule, next); + return 0; +} + +/* VXLAN encap rule database for outer interfaces. */ +static LIST_HEAD(, tcf_irule) iface_list_vxlan = LIST_HEAD_INITIALIZER(); + +/* VTEP device list is shared between PMD port instances. */ +static LIST_HEAD(, tcf_vtep) vtep_list_vxlan = LIST_HEAD_INITIALIZER(); +static pthread_mutex_t vtep_list_mutex = PTHREAD_MUTEX_INITIALIZER; + +/** + * Acquire the VXLAN encap rules container for specified interface. + * First looks for the container in the existing ones list, creates + * and initializes the new container if existing not found. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] ifouter + * Network interface index to create VXLAN encap rules on. + * @param[out] error + * Perform verbose error reporting if not NULL. + * @return + * Rule container pointer on success, + * NULL otherwise and rte_errno is set. + */ +static struct tcf_irule* +flow_tcf_encap_irule_acquire(struct mlx5_flow_tcf_context *tcf, + unsigned int ifouter, + struct rte_flow_error *error) +{ + struct tcf_irule *iface; + + /* Look whether the container for encap rules is created. */ + assert(ifouter); + LIST_FOREACH(iface, &iface_list_vxlan, next) { + if (iface->ifouter == ifouter) + break; + } + if (iface) { + /* Container already exists, just increment the reference. */ + iface->refcnt++; + return iface; + } + /* Not found, we should create the new container. */ + iface = rte_zmalloc(__func__, sizeof(*iface), + alignof(struct tcf_irule)); + if (!iface) { + rte_flow_error_set(error, ENOMEM, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "unable to allocate memory for container"); + return NULL; + } + *iface = (struct tcf_irule){ + .local = LIST_HEAD_INITIALIZER(), + .neigh = LIST_HEAD_INITIALIZER(), + .ifouter = ifouter, + .refcnt = 1, + }; + /* Interface cleanup for new container created. */ + flow_tcf_encap_iface_cleanup(tcf, ifouter); + flow_tcf_encap_local_cleanup(tcf, ifouter); + flow_tcf_encap_neigh_cleanup(tcf, ifouter); + LIST_INSERT_HEAD(&iface_list_vxlan, iface, next); + return iface; +} + +/** + * Releases VXLAN encap rules container by pointer. Decrements the + * reference counter and deletes the container if counter is zero. + * + * @param[in] irule + * VXLAN rule container pointer to release. + */ +static void +flow_tcf_encap_irule_release(struct tcf_irule *iface) +{ + assert(iface->refcnt); + if (--iface->refcnt == 0) { + /* Reference counter is zero, delete the container. */ + assert(LIST_EMPTY(&iface->local)); + assert(LIST_EMPTY(&iface->neigh)); + LIST_REMOVE(iface, next); + rte_free(iface); + } +} + +/** + * Deletes VTEP network device. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] vtep + * Object representing the network device to delete. Memory + * allocated for this object is freed by routine. + */ +static void +flow_tcf_vtep_delete(struct mlx5_flow_tcf_context *tcf, + struct tcf_vtep *vtep) +{ + struct nlmsghdr *nlh; + struct ifinfomsg *ifm; + alignas(struct nlmsghdr) + uint8_t buf[mnl_nlmsg_size(MNL_ALIGN(sizeof(*ifm))) + + MNL_BUF_EXTRA_SPACE]; + int ret; + + assert(!vtep->refcnt); + /* Delete only ifaces those we actually created. */ + if (vtep->created && vtep->ifindex) { + DRV_LOG(INFO, "VTEP delete (%d)", vtep->ifindex); + nlh = mnl_nlmsg_put_header(buf); + nlh->nlmsg_type = RTM_DELLINK; + nlh->nlmsg_flags = NLM_F_REQUEST; + ifm = mnl_nlmsg_put_extra_header(nlh, sizeof(*ifm)); + ifm->ifi_family = AF_UNSPEC; + ifm->ifi_index = vtep->ifindex; + assert(sizeof(buf) >= nlh->nlmsg_len); + ret = flow_tcf_nl_ack(tcf, nlh, NULL, NULL); + if (ret) + DRV_LOG(WARNING, "netlink: error deleting vxlan" + " encap/decap ifindex %u", + ifm->ifi_index); + } + rte_free(vtep); +} + +/** + * Creates VTEP network device. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] port + * UDP port of created VTEP device. + * @param[out] error + * Perform verbose error reporting if not NULL. + * + * @return + * Pointer to created device structure on success, + * NULL otherwise and rte_errno is set. + */ +static struct tcf_vtep* +flow_tcf_vtep_create(struct mlx5_flow_tcf_context *tcf, + uint16_t port, struct rte_flow_error *error) +{ + struct tcf_vtep *vtep; + struct nlmsghdr *nlh; + struct ifinfomsg *ifm; + char name[sizeof(MLX5_VXLAN_DEVICE_PFX) + 24]; + alignas(struct nlmsghdr) + uint8_t buf[mnl_nlmsg_size(sizeof(*ifm)) + + SZ_NLATTR_DATA_OF(sizeof(name)) + + SZ_NLATTR_NEST * 2 + + SZ_NLATTR_STRZ_OF("vxlan") + + SZ_NLATTR_DATA_OF(sizeof(uint32_t)) + + SZ_NLATTR_DATA_OF(sizeof(uint16_t)) + + SZ_NLATTR_DATA_OF(sizeof(uint8_t)) * 3 + + MNL_BUF_EXTRA_SPACE]; + struct nlattr *na_info; + struct nlattr *na_vxlan; + rte_be16_t vxlan_port = rte_cpu_to_be_16(port); + int ret; + + vtep = rte_zmalloc(__func__, sizeof(*vtep), alignof(struct tcf_vtep)); + if (!vtep) { + rte_flow_error_set(error, ENOMEM, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "unable to allocate memory for VTEP"); + return NULL; + } + *vtep = (struct tcf_vtep){ + .port = port, + }; + memset(buf, 0, sizeof(buf)); + nlh = mnl_nlmsg_put_header(buf); + nlh->nlmsg_type = RTM_NEWLINK; + nlh->nlmsg_flags = NLM_F_REQUEST | NLM_F_CREATE | NLM_F_EXCL; + ifm = mnl_nlmsg_put_extra_header(nlh, sizeof(*ifm)); + ifm->ifi_family = AF_UNSPEC; + ifm->ifi_type = 0; + ifm->ifi_index = 0; + ifm->ifi_flags = IFF_UP; + ifm->ifi_change = 0xffffffff; + snprintf(name, sizeof(name), "%s%u", MLX5_VXLAN_DEVICE_PFX, port); + mnl_attr_put_strz(nlh, IFLA_IFNAME, name); + na_info = mnl_attr_nest_start(nlh, IFLA_LINKINFO); + assert(na_info); + mnl_attr_put_strz(nlh, IFLA_INFO_KIND, "vxlan"); + na_vxlan = mnl_attr_nest_start(nlh, IFLA_INFO_DATA); + assert(na_vxlan); +#ifdef HAVE_IFLA_VXLAN_COLLECT_METADATA + /* + * RH 7.2 does not support metadata for tunnel device. + * It does not matter because we are going to use the + * hardware offload by mlx5 driver. + */ + mnl_attr_put_u8(nlh, IFLA_VXLAN_COLLECT_METADATA, 1); +#endif + mnl_attr_put_u8(nlh, IFLA_VXLAN_UDP_ZERO_CSUM6_RX, 1); + mnl_attr_put_u8(nlh, IFLA_VXLAN_LEARNING, 0); + mnl_attr_put_u16(nlh, IFLA_VXLAN_PORT, vxlan_port); +#ifndef HAVE_IFLA_VXLAN_COLLECT_METADATA + /* + * We must specify VNI explicitly if metadata not supported. + * Note, VNI is transferred with native endianness format. + */ + mnl_attr_put_u16(nlh, IFLA_VXLAN_ID, MLX5_VXLAN_DEFAULT_VNI); +#endif + mnl_attr_nest_end(nlh, na_vxlan); + mnl_attr_nest_end(nlh, na_info); + assert(sizeof(buf) >= nlh->nlmsg_len); + ret = flow_tcf_nl_ack(tcf, nlh, NULL, NULL); + if (ret) { + DRV_LOG(WARNING, + "netlink: VTEP %s create failure (%d)", + name, rte_errno); + if (rte_errno != EEXIST) + /* + * Some unhandled error occurred or device is + * for encapsulation and cannot be shared. + */ + goto error; + } else { + /* + * Mark device we actually created. + * We should explicitly delete + * when we do not need it anymore. + */ + vtep->created = 1; + vtep->waitreg = 1; + } + /* Try to get ifindex of created of pre-existing device. */ + ret = if_nametoindex(name); + if (!ret) { + DRV_LOG(WARNING, + "VTEP %s failed to get index (%d)", name, errno); + rte_flow_error_set + (error, -errno, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "netlink: failed to retrieve VTEP ifindex"); + goto error; + } + vtep->ifindex = ret; + memset(buf, 0, sizeof(buf)); + nlh = mnl_nlmsg_put_header(buf); + nlh->nlmsg_type = RTM_NEWLINK; + nlh->nlmsg_flags = NLM_F_REQUEST; + ifm = mnl_nlmsg_put_extra_header(nlh, sizeof(*ifm)); + ifm->ifi_family = AF_UNSPEC; + ifm->ifi_type = 0; + ifm->ifi_index = vtep->ifindex; + ifm->ifi_flags = IFF_UP; + ifm->ifi_change = IFF_UP; + ret = flow_tcf_nl_ack(tcf, nlh, NULL, NULL); + if (ret) { + rte_flow_error_set(error, -errno, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "netlink: failed to set VTEP link up"); + DRV_LOG(WARNING, "netlink: VTEP %s set link up failure (%d)", + name, rte_errno); + goto clean; + } + ret = mlx5_flow_tcf_init(tcf, vtep->ifindex, error); + if (ret) { + DRV_LOG(WARNING, "VTEP %s init failure (%d)", name, rte_errno); + goto clean; + } + DRV_LOG(INFO, "VTEP create (%d, %d)", vtep->port, vtep->ifindex); + vtep->refcnt = 1; + return vtep; +clean: + flow_tcf_vtep_delete(tcf, vtep); + return NULL; +error: + rte_free(vtep); + return NULL; +} + +/** + * Acquire target interface index for VXLAN tunneling decapsulation. + * In order to share the UDP port within the other interfaces the + * VXLAN device created as not attached to any interface (if created). + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] dev_flow + * Flow tcf object with tunnel structure pointer set. + * @param[out] error + * Perform verbose error reporting if not NULL. + * @return + * Interface descriptor pointer on success, + * NULL otherwise and rte_errno is set. + */ +static struct tcf_vtep* +flow_tcf_decap_vtep_acquire(struct mlx5_flow_tcf_context *tcf, + struct mlx5_flow *dev_flow, + struct rte_flow_error *error) +{ + struct tcf_vtep *vtep; + uint16_t port = dev_flow->tcf.vxlan_decap->udp_port; + + LIST_FOREACH(vtep, &vtep_list_vxlan, next) { + if (vtep->port == port) + break; + } + if (vtep) { + /* Device exists, just increment the reference counter. */ + vtep->refcnt++; + assert(vtep->ifindex); + return vtep; + } + /* No decapsulation device exists, try to create the new one. */ + vtep = flow_tcf_vtep_create(tcf, port, error); + if (vtep) + LIST_INSERT_HEAD(&vtep_list_vxlan, vtep, next); + return vtep; +} + +/** + * Acquire target interface index for VXLAN tunneling encapsulation. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] ifouter + * Network interface index to attach VXLAN encap device to. + * @param[in] dev_flow + * Flow tcf object with tunnel structure pointer set. + * @param[out] error + * Perform verbose error reporting if not NULL. + * @return + * Interface descriptor pointer on success, + * NULL otherwise and rte_errno is set. + */ +static struct tcf_vtep* +flow_tcf_encap_vtep_acquire(struct mlx5_flow_tcf_context *tcf, + unsigned int ifouter, + struct mlx5_flow *dev_flow, + struct rte_flow_error *error) +{ + static uint16_t port; + struct tcf_vtep *vtep; + struct tcf_irule *iface; + int ret; + + assert(ifouter); + /* Look whether the VTEP for specified port is created. */ + port = rte_be_to_cpu_16(dev_flow->tcf.vxlan_encap->udp.dst); + LIST_FOREACH(vtep, &vtep_list_vxlan, next) { + if (vtep->port == port) + break; + } + if (vtep) { + /* VTEP already exists, just increment the reference. */ + vtep->refcnt++; + } else { + /* Not found, we should create the new VTEP. */ + vtep = flow_tcf_vtep_create(tcf, port, error); + if (!vtep) + return NULL; + LIST_INSERT_HEAD(&vtep_list_vxlan, vtep, next); + } + assert(vtep->ifindex); + iface = flow_tcf_encap_irule_acquire(tcf, ifouter, error); + if (!iface) { + if (--vtep->refcnt == 0) + flow_tcf_vtep_delete(tcf, vtep); + return NULL; + } + dev_flow->tcf.vxlan_encap->iface = iface; + /* Create local ipaddr with peer to specify the outer IPs. */ + ret = flow_tcf_encap_local(tcf, iface, dev_flow, true, error); + if (!ret) { + /* Create neigh rule to specify outer destination MAC. */ + ret = flow_tcf_encap_neigh(tcf, iface, dev_flow, true, error); + if (ret) + flow_tcf_encap_local(tcf, iface, + dev_flow, false, error); + } + if (ret) { + dev_flow->tcf.vxlan_encap->iface = NULL; + flow_tcf_encap_irule_release(iface); + if (--vtep->refcnt == 0) + flow_tcf_vtep_delete(tcf, vtep); + return NULL; + } + return vtep; +} + +/** + * Acquires target interface index for tunneling of any type. + * Creates the new VTEP if needed. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] ifouter + * Network interface index to create VXLAN encap rules on. + * @param[in] dev_flow + * Flow tcf object with tunnel structure pointer set. + * @param[out] error + * Perform verbose error reporting if not NULL. + * @return + * Interface descriptor pointer on success, + * NULL otherwise and rte_errno is set. + */ +static struct tcf_vtep* +flow_tcf_vtep_acquire(struct mlx5_flow_tcf_context *tcf, + unsigned int ifouter, + struct mlx5_flow *dev_flow, + struct rte_flow_error *error) +{ + struct tcf_vtep *vtep = NULL; + + assert(dev_flow->tcf.tunnel); + pthread_mutex_lock(&vtep_list_mutex); + switch (dev_flow->tcf.tunnel->type) { + case FLOW_TCF_TUNACT_VXLAN_ENCAP: + vtep = flow_tcf_encap_vtep_acquire(tcf, ifouter, + dev_flow, error); + break; + case FLOW_TCF_TUNACT_VXLAN_DECAP: + vtep = flow_tcf_decap_vtep_acquire(tcf, dev_flow, error); + break; + default: + rte_flow_error_set(error, ENOTSUP, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "unsupported tunnel type"); + break; + } + pthread_mutex_unlock(&vtep_list_mutex); + return vtep; +} + +/** + * Release tunneling interface by ifindex. Decrements reference + * counter and actually removes the device if counter is zero. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] vtep + * VTEP device descriptor structure. + * @param[in] dev_flow + * Flow tcf object with tunnel structure pointer set. + */ +static void +flow_tcf_vtep_release(struct mlx5_flow_tcf_context *tcf, + struct tcf_vtep *vtep, + struct mlx5_flow *dev_flow) +{ + assert(dev_flow->tcf.tunnel); + pthread_mutex_lock(&vtep_list_mutex); + switch (dev_flow->tcf.tunnel->type) { + case FLOW_TCF_TUNACT_VXLAN_DECAP: + break; + case FLOW_TCF_TUNACT_VXLAN_ENCAP: { + struct tcf_irule *iface; + + /* Remove the encap ancillary rules first. */ + iface = dev_flow->tcf.vxlan_encap->iface; + assert(iface); + flow_tcf_encap_neigh(tcf, iface, dev_flow, false, NULL); + flow_tcf_encap_local(tcf, iface, dev_flow, false, NULL); + flow_tcf_encap_irule_release(iface); + dev_flow->tcf.vxlan_encap->iface = NULL; + break; + } + default: + assert(false); + DRV_LOG(WARNING, "Unsupported tunnel type"); + break; + } + assert(vtep->refcnt); + if (--vtep->refcnt == 0) { + LIST_REMOVE(vtep, next); + flow_tcf_vtep_delete(tcf, vtep); + } + pthread_mutex_unlock(&vtep_list_mutex); +} + +struct tcf_nlcb_query { + uint32_t handle; + uint32_t tc_flags; + uint32_t flags_valid:1; +}; + +/** + * Collect queried rule attributes. This is callback routine called by + * libmnl mnl_cb_run() in loop for every message in received packet. + * Current implementation collects the flower flags only. + * + * @param[in] nlh + * Pointer to reply header. + * @param[in, out] arg + * Context pointer for this callback. + * + * @return + * A positive, nonzero value on success (required by libmnl + * to continue messages processing). + */ +static int +flow_tcf_collect_query_cb(const struct nlmsghdr *nlh, void *arg) +{ + struct tcf_nlcb_query *query = arg; + struct tcmsg *tcm = mnl_nlmsg_get_payload(nlh); + struct nlattr *na, *na_opt; + bool flower = false; + + if (nlh->nlmsg_type != RTM_NEWTFILTER || + tcm->tcm_handle != query->handle) + return 1; + mnl_attr_for_each(na, nlh, sizeof(*tcm)) { + switch (mnl_attr_get_type(na)) { + case TCA_KIND: + if (strcmp(mnl_attr_get_payload(na), "flower")) { + /* Not flower filter, drop entire message. */ + return 1; + } + flower = true; + break; + case TCA_OPTIONS: + if (!flower) { + /* Not flower options, drop entire message. */ + return 1; + } + /* Check nested flower options. */ + mnl_attr_for_each_nested(na_opt, na) { + switch (mnl_attr_get_type(na_opt)) { + case TCA_FLOWER_FLAGS: + query->flags_valid = 1; + query->tc_flags = + mnl_attr_get_u32(na_opt); + break; + } + } + break; + } + } + return 1; +} + +/** + * Query a TC flower rule flags via netlink. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] dev_flow + * Pointer to the flow. + * @param[out] pflags + * pointer to the data retrieved by the query. + * + * @return + * 0 on success, a negative errno value otherwise. + */ +static int +flow_tcf_query_flags(struct mlx5_flow_tcf_context *tcf, + struct mlx5_flow *dev_flow, + uint32_t *pflags) +{ + struct nlmsghdr *nlh; + struct tcmsg *tcm; + struct tcf_nlcb_query query = { + .handle = dev_flow->tcf.tcm->tcm_handle, + }; + + nlh = mnl_nlmsg_put_header(tcf->buf); + nlh->nlmsg_type = RTM_GETTFILTER; + nlh->nlmsg_flags = NLM_F_REQUEST; + tcm = mnl_nlmsg_put_extra_header(nlh, sizeof(*tcm)); + memcpy(tcm, dev_flow->tcf.tcm, sizeof(*tcm)); + /* + * Ignore Netlink error for filter query operations. + * The reply length is sent by kernel as errno. + * Just check we got the flags option. + */ + flow_tcf_nl_ack(tcf, nlh, flow_tcf_collect_query_cb, &query); + if (!query.flags_valid) { + *pflags = 0; + return -ENOENT; + } + *pflags = query.tc_flags; + return 0; +} + +/** + * Query and check the in_hw set for specified rule. + * + * @param[in] tcf + * Context object initialized by mlx5_flow_tcf_context_create(). + * @param[in] dev_flow + * Pointer to the flow to check. + * + * @return + * 0 on success, a negative errno value otherwise. + */ +static int +flow_tcf_check_inhw(struct mlx5_flow_tcf_context *tcf, + struct mlx5_flow *dev_flow) +{ + uint32_t flags; + int ret; + + ret = flow_tcf_query_flags(tcf, dev_flow, &flags); + if (ret) + return ret; + return (flags & TCA_CLS_FLAGS_IN_HW) ? 0 : -ENOENT; +} + +/** + * Remove flow from E-Switch by sending Netlink message. + * + * @param[in] dev + * Pointer to Ethernet device. + * @param[in, out] flow + * Pointer to the sub flow. + */ +static void +flow_tcf_remove(struct rte_eth_dev *dev, struct rte_flow *flow) +{ + struct mlx5_priv *priv = dev->data->dev_private; + struct mlx5_flow_tcf_context *ctx = priv->tcf_context; + struct mlx5_flow *dev_flow; + struct nlmsghdr *nlh; + struct tcmsg *tcm; + + if (!flow) + return; + dev_flow = LIST_FIRST(&flow->dev_flows); + if (!dev_flow) + return; + /* E-Switch flow can't be expanded. */ + assert(!LIST_NEXT(dev_flow, next)); + if (dev_flow->tcf.applied) { + nlh = dev_flow->tcf.nlh; + nlh->nlmsg_type = RTM_DELTFILTER; + nlh->nlmsg_flags = NLM_F_REQUEST; + flow_tcf_nl_ack(ctx, nlh, NULL, NULL); + if (dev_flow->tcf.tunnel) { + assert(dev_flow->tcf.tunnel->vtep); + flow_tcf_vtep_release(ctx, + dev_flow->tcf.tunnel->vtep, + dev_flow); + dev_flow->tcf.tunnel->vtep = NULL; + } + /* Cleanup the rule handle value. */ + tcm = mnl_nlmsg_get_payload(nlh); + tcm->tcm_handle = 0; + dev_flow->tcf.applied = 0; + } +} + +/** + * Fetch the applied rule handle. This is callback routine called by + * libmnl mnl_cb_run() in loop for every message in received packet. + * When the NLM_F_ECHO flag is specified the kernel sends the created + * rule descriptor back to the application and we can retrieve the + * actual rule handle from updated descriptor. + * + * @param[in] nlh + * Pointer to reply header. + * @param[in, out] arg + * Context pointer for this callback. + * + * @return + * A positive, nonzero value on success (required by libmnl + * to continue messages processing). + */ +static int +flow_tcf_collect_apply_cb(const struct nlmsghdr *nlh, void *arg) +{ + struct nlmsghdr *nlhrq = arg; + struct tcmsg *tcmrq = mnl_nlmsg_get_payload(nlhrq); + struct tcmsg *tcm = mnl_nlmsg_get_payload(nlh); + struct nlattr *na; + + if (nlh->nlmsg_type != RTM_NEWTFILTER || + nlh->nlmsg_seq != nlhrq->nlmsg_seq) + return 1; + mnl_attr_for_each(na, nlh, sizeof(*tcm)) { + switch (mnl_attr_get_type(na)) { + case TCA_KIND: + if (strcmp(mnl_attr_get_payload(na), "flower")) { + /* Not flower filter, drop entire message. */ + return 1; + } + tcmrq->tcm_handle = tcm->tcm_handle; + return 1; + } + } + return 1; +} +/** + * Apply flow to E-Switch by sending Netlink message. + * + * @param[in] dev + * Pointer to Ethernet device. + * @param[in, out] flow + * Pointer to the sub flow. + * @param[out] error + * Pointer to the error structure. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + */ +static int +flow_tcf_apply(struct rte_eth_dev *dev, struct rte_flow *flow, + struct rte_flow_error *error) +{ + struct mlx5_priv *priv = dev->data->dev_private; + struct mlx5_flow_tcf_context *ctx = priv->tcf_context; + struct mlx5_flow *dev_flow; + struct nlmsghdr *nlh; + struct tcmsg *tcm; + uint64_t start = 0; + uint64_t twait = 0; + int ret; + + dev_flow = LIST_FIRST(&flow->dev_flows); + /* E-Switch flow can't be expanded. */ + assert(!LIST_NEXT(dev_flow, next)); + if (dev_flow->tcf.applied) + return 0; + nlh = dev_flow->tcf.nlh; + nlh->nlmsg_type = RTM_NEWTFILTER; + nlh->nlmsg_flags = NLM_F_REQUEST | NLM_F_CREATE | + NLM_F_EXCL | NLM_F_ECHO; + tcm = mnl_nlmsg_get_payload(nlh); + /* Allow kernel to assign handle on its own. */ + tcm->tcm_handle = 0; + if (dev_flow->tcf.tunnel) { + /* + * Replace the interface index, target for + * encapsulation, source for decapsulation. + */ + assert(!dev_flow->tcf.tunnel->vtep); + assert(dev_flow->tcf.tunnel->ifindex_ptr); + /* Acquire actual VTEP device when rule is being applied. */ + dev_flow->tcf.tunnel->vtep = + flow_tcf_vtep_acquire(ctx, + dev_flow->tcf.tunnel->ifindex_org, + dev_flow, error); + if (!dev_flow->tcf.tunnel->vtep) + return -rte_errno; + DRV_LOG(INFO, "Replace ifindex: %d->%d", + dev_flow->tcf.tunnel->vtep->ifindex, + dev_flow->tcf.tunnel->ifindex_org); + *dev_flow->tcf.tunnel->ifindex_ptr = + dev_flow->tcf.tunnel->vtep->ifindex; + if (dev_flow->tcf.tunnel->vtep->waitreg) { + /* Clear wait flag for VXLAN port registration. */ + dev_flow->tcf.tunnel->vtep->waitreg = 0; + twait = rte_get_timer_hz(); + assert(twait > MS_PER_S); + twait = twait * MLX5_VXLAN_WAIT_PORT_REG_MS; + twait = twait / MS_PER_S; + start = rte_get_timer_cycles(); + } + } + /* + * Kernel creates the VXLAN devices and registers UDP ports to + * be hardware offloaded within the NIC kernel drivers. The + * registration process is being performed into context of + * working kernel thread and the race conditions might happen. + * The VXLAN device is created and success is returned to + * calling application, but the UDP port registration process + * is not completed yet. The next applied rule may be rejected + * by the driver with ENOSUP code. We are going to wait a bit, + * allowing registration process to be completed. The waiting + * is performed once after device been created. + */ + do { + struct timespec onems; + + ret = flow_tcf_nl_ack(ctx, nlh, + flow_tcf_collect_apply_cb, nlh); + if (!ret || ret != -ENOTSUP || !twait) + break; + /* Wait one millisecond and try again till timeout. */ + onems.tv_sec = 0; + onems.tv_nsec = NS_PER_S / MS_PER_S; + nanosleep(&onems, 0); + if ((rte_get_timer_cycles() - start) > twait) { + /* Timeout elapsed, try once more and exit. */ + twait = 0; + } + } while (true); + if (!ret) { + if (!tcm->tcm_handle) { + flow_tcf_remove(dev, flow); + return rte_flow_error_set + (error, ENOENT, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "netlink: rule zero handle returned"); + } + dev_flow->tcf.applied = 1; + if (*dev_flow->tcf.ptc_flags & TCA_CLS_FLAGS_SKIP_SW) + return 0; + /* + * Rule was applied without skip_sw flag set. + * We should check whether the rule was acctually + * accepted by hardware (have look at in_hw flag). + */ + if (flow_tcf_check_inhw(ctx, dev_flow)) { + flow_tcf_remove(dev, flow); + return rte_flow_error_set + (error, ENOENT, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "netlink: rule has no in_hw flag set"); + } + return 0; + } + if (dev_flow->tcf.tunnel) { + /* Rollback the VTEP configuration if rule apply failed. */ + assert(dev_flow->tcf.tunnel->vtep); + flow_tcf_vtep_release(ctx, dev_flow->tcf.tunnel->vtep, + dev_flow); + dev_flow->tcf.tunnel->vtep = NULL; + } + return rte_flow_error_set(error, rte_errno, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, + "netlink: failed to create TC flow rule"); +} + +/** + * Remove flow from E-Switch and release resources of the device flow. + * + * @param[in] dev + * Pointer to Ethernet device. + * @param[in, out] flow + * Pointer to the sub flow. + */ +static void +flow_tcf_destroy(struct rte_eth_dev *dev, struct rte_flow *flow) +{ + struct mlx5_flow *dev_flow; + + if (!flow) + return; + flow_tcf_remove(dev, flow); + if (flow->counter) { + if (--flow->counter->ref_cnt == 0) { + rte_free(flow->counter); + flow->counter = NULL; + } + } dev_flow = LIST_FIRST(&flow->dev_flows); if (!dev_flow) return; @@ -2722,7 +6089,7 @@ flow_tcf_nl_filter_parse_and_get(struct nlmsghdr *cnlh, * Message received from Netlink. * @param[out] data * Pointer to data area to be filled by the parsing routine. - * assumed to be a pinter to struct flow_tcf_stats_basic. + * assumed to be a pointer to struct flow_tcf_stats_basic. * * @return * MNL_CB_OK value. @@ -2770,9 +6137,9 @@ flow_tcf_query_count(struct rte_eth_dev *dev, void *data, struct rte_flow_error *error) { - struct flow_tcf_stats_basic sb_data = { 0 }; + struct flow_tcf_stats_basic sb_data; struct rte_flow_query_count *qc = data; - struct priv *priv = dev->data->dev_private; + struct mlx5_priv *priv = dev->data->dev_private; struct mlx5_flow_tcf_context *ctx = priv->tcf_context; struct mnl_socket *nl = ctx->nl; struct mlx5_flow *dev_flow; @@ -2781,6 +6148,7 @@ flow_tcf_query_count(struct rte_eth_dev *dev, ssize_t ret; assert(qc); + memset(&sb_data, 0, sizeof(sb_data)); dev_flow = LIST_FIRST(&flow->dev_flows); /* E-Switch flow can't be expanded. */ assert(!LIST_NEXT(dev_flow, next)); @@ -2927,7 +6295,9 @@ mlx5_flow_tcf_init(struct mlx5_flow_tcf_context *ctx, struct nlmsghdr *nlh; struct tcmsg *tcm; alignas(struct nlmsghdr) - uint8_t buf[mnl_nlmsg_size(sizeof(*tcm) + 128)]; + uint8_t buf[mnl_nlmsg_size(sizeof(*tcm)) + + SZ_NLATTR_STRZ_OF("ingress") + + MNL_BUF_EXTRA_SPACE]; /* Destroy existing ingress qdisc and everything attached to it. */ nlh = mnl_nlmsg_put_header(buf); @@ -2938,8 +6308,9 @@ mlx5_flow_tcf_init(struct mlx5_flow_tcf_context *ctx, tcm->tcm_ifindex = ifindex; tcm->tcm_handle = TC_H_MAKE(TC_H_INGRESS, 0); tcm->tcm_parent = TC_H_INGRESS; + assert(sizeof(buf) >= nlh->nlmsg_len); /* Ignore errors when qdisc is already absent. */ - if (flow_tcf_nl_ack(ctx, nlh) && + if (flow_tcf_nl_ack(ctx, nlh, NULL, NULL) && rte_errno != EINVAL && rte_errno != ENOENT) return rte_flow_error_set(error, rte_errno, RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, @@ -2955,7 +6326,8 @@ mlx5_flow_tcf_init(struct mlx5_flow_tcf_context *ctx, tcm->tcm_handle = TC_H_MAKE(TC_H_INGRESS, 0); tcm->tcm_parent = TC_H_INGRESS; mnl_attr_put_strz_check(nlh, sizeof(buf), TCA_KIND, "ingress"); - if (flow_tcf_nl_ack(ctx, nlh)) + assert(sizeof(buf) >= nlh->nlmsg_len); + if (flow_tcf_nl_ack(ctx, nlh, NULL, NULL)) return rte_flow_error_set(error, rte_errno, RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, "netlink: failed to create ingress"