X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=examples%2Fipsec-secgw%2Fipsec-secgw.c;h=66e92165df419f18514765828d3b0638e076546c;hb=6d13ea8e8e49ab957deae2bba5ecf4a4bfe747d1;hp=685feec1fdca253c25613f3daa3608de85b9b720;hpb=9d5ca5323910591f2cafbac5070fcb5b6d02d022;p=dpdk.git diff --git a/examples/ipsec-secgw/ipsec-secgw.c b/examples/ipsec-secgw/ipsec-secgw.c index 685feec1fd..66e92165df 100644 --- a/examples/ipsec-secgw/ipsec-secgw.c +++ b/examples/ipsec-secgw/ipsec-secgw.c @@ -1,34 +1,5 @@ -/*- - * BSD LICENSE - * - * Copyright(c) 2016 Intel Corporation. All rights reserved. - * All rights reserved. - * - * Redistribution and use in source and binary forms, with or without - * modification, are permitted provided that the following conditions - * are met: - * - * * Redistributions of source code must retain the above copyright - * notice, this list of conditions and the following disclaimer. - * * Redistributions in binary form must reproduce the above copyright - * notice, this list of conditions and the following disclaimer in - * the documentation and/or other materials provided with the - * distribution. - * * Neither the name of Intel Corporation nor the names of its - * contributors may be used to endorse or promote products derived - * from this software without specific prior written permission. - * - * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS - * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT - * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR - * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT - * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, - * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT - * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, - * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY - * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT - * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE - * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +/* SPDX-License-Identifier: BSD-3-Clause + * Copyright(c) 2016 Intel Corporation */ #include @@ -57,7 +28,6 @@ #include #include #include -#include #include #include #include @@ -70,6 +40,7 @@ #include #include #include +#include #include "ipsec.h" #include "parser.h" @@ -83,14 +54,11 @@ #define NB_MBUF (32000) #define CDEV_QUEUE_DESC 2048 -#define CDEV_MAP_ENTRIES 1024 -#define CDEV_MP_NB_OBJS 2048 +#define CDEV_MAP_ENTRIES 16384 +#define CDEV_MP_NB_OBJS 1024 #define CDEV_MP_CACHE_SZ 64 #define MAX_QUEUE_PAIRS 1 -#define OPTION_CONFIG "config" -#define OPTION_SINGLE_SA "single-sa" - #define BURST_TX_DRAIN_US 100 /* TX drain every ~100us */ #define NB_SOCKETS 4 @@ -107,8 +75,8 @@ /* * Configurable number of RX/TX ring descriptors */ -#define IPSEC_SECGW_RX_DESC_DEFAULT 128 -#define IPSEC_SECGW_TX_DESC_DEFAULT 512 +#define IPSEC_SECGW_RX_DESC_DEFAULT 1024 +#define IPSEC_SECGW_TX_DESC_DEFAULT 1024 static uint16_t nb_rxd = IPSEC_SECGW_RX_DESC_DEFAULT; static uint16_t nb_txd = IPSEC_SECGW_TX_DESC_DEFAULT; @@ -136,9 +104,9 @@ static uint16_t nb_txd = IPSEC_SECGW_TX_DESC_DEFAULT; #define ETHADDR(a, b, c, d, e, f) (__BYTES_TO_UINT64(a, b, c, d, e, f, 0, 0)) #define ETHADDR_TO_UINT64(addr) __BYTES_TO_UINT64( \ - addr.addr_bytes[0], addr.addr_bytes[1], \ - addr.addr_bytes[2], addr.addr_bytes[3], \ - addr.addr_bytes[4], addr.addr_bytes[5], \ + (addr)->addr_bytes[0], (addr)->addr_bytes[1], \ + (addr)->addr_bytes[2], (addr)->addr_bytes[3], \ + (addr)->addr_bytes[4], (addr)->addr_bytes[5], \ 0, 0) /* port/source ethernet addr and destination ethernet addr */ @@ -153,22 +121,63 @@ struct ethaddr_info ethaddr_tbl[RTE_MAX_ETHPORTS] = { { 0, ETHADDR(0x00, 0x16, 0x3e, 0x49, 0x9e, 0xdd) } }; +#define CMD_LINE_OPT_CONFIG "config" +#define CMD_LINE_OPT_SINGLE_SA "single-sa" +#define CMD_LINE_OPT_CRYPTODEV_MASK "cryptodev_mask" +#define CMD_LINE_OPT_RX_OFFLOAD "rxoffload" +#define CMD_LINE_OPT_TX_OFFLOAD "txoffload" + +enum { + /* long options mapped to a short option */ + + /* first long only option value must be >= 256, so that we won't + * conflict with short options + */ + CMD_LINE_OPT_MIN_NUM = 256, + CMD_LINE_OPT_CONFIG_NUM, + CMD_LINE_OPT_SINGLE_SA_NUM, + CMD_LINE_OPT_CRYPTODEV_MASK_NUM, + CMD_LINE_OPT_RX_OFFLOAD_NUM, + CMD_LINE_OPT_TX_OFFLOAD_NUM, +}; + +static const struct option lgopts[] = { + {CMD_LINE_OPT_CONFIG, 1, 0, CMD_LINE_OPT_CONFIG_NUM}, + {CMD_LINE_OPT_SINGLE_SA, 1, 0, CMD_LINE_OPT_SINGLE_SA_NUM}, + {CMD_LINE_OPT_CRYPTODEV_MASK, 1, 0, CMD_LINE_OPT_CRYPTODEV_MASK_NUM}, + {CMD_LINE_OPT_RX_OFFLOAD, 1, 0, CMD_LINE_OPT_RX_OFFLOAD_NUM}, + {CMD_LINE_OPT_TX_OFFLOAD, 1, 0, CMD_LINE_OPT_TX_OFFLOAD_NUM}, + {NULL, 0, 0, 0} +}; + /* mask of enabled ports */ static uint32_t enabled_port_mask; +static uint64_t enabled_cryptodev_mask = UINT64_MAX; static uint32_t unprotected_port_mask; static int32_t promiscuous_on = 1; static int32_t numa_on = 1; /**< NUMA is enabled by default. */ static uint32_t nb_lcores; static uint32_t single_sa; static uint32_t single_sa_idx; +static uint32_t frame_size; + +/* + * RX/TX HW offload capabilities to enable/use on ethernet ports. + * By default all capabilities are enabled. + */ +static uint64_t dev_rx_offload = UINT64_MAX; +static uint64_t dev_tx_offload = UINT64_MAX; + +/* application wide librte_ipsec/SA parameters */ +struct app_sa_prm app_sa_prm = {.enable = 0}; struct lcore_rx_queue { - uint8_t port_id; + uint16_t port_id; uint8_t queue_id; } __rte_cache_aligned; struct lcore_params { - uint8_t port_id; + uint16_t port_id; uint8_t queue_id; uint8_t lcore_id; } __rte_cache_aligned; @@ -204,11 +213,7 @@ static struct rte_eth_conf port_conf = { .mq_mode = ETH_MQ_RX_RSS, .max_rx_pkt_len = ETHER_MAX_LEN, .split_hdr_size = 0, - .header_split = 0, /**< Header Split disabled */ - .hw_ip_checksum = 1, /**< IP checksum offload enabled */ - .hw_vlan_filter = 0, /**< VLAN filtering disabled */ - .jumbo_frame = 0, /**< Jumbo Frame Support disabled */ - .hw_strip_crc = 0, /**< CRC stripped by hardware */ + .offloads = DEV_RX_OFFLOAD_CHECKSUM, }, .rx_adv_conf = { .rss_conf = { @@ -224,26 +229,13 @@ static struct rte_eth_conf port_conf = { static struct socket_ctx socket_ctx[NB_SOCKETS]; -struct traffic_type { - const uint8_t *data[MAX_PKT_BURST * 2]; - struct rte_mbuf *pkts[MAX_PKT_BURST * 2]; - uint32_t res[MAX_PKT_BURST * 2]; - uint32_t num; -}; - -struct ipsec_traffic { - struct traffic_type ipsec; - struct traffic_type ip4; - struct traffic_type ip6; -}; - static inline void prepare_one_packet(struct rte_mbuf *pkt, struct ipsec_traffic *t) { uint8_t *nlp; - struct ether_hdr *eth; + struct rte_ether_hdr *eth; - eth = rte_pktmbuf_mtod(pkt, struct ether_hdr *); + eth = rte_pktmbuf_mtod(pkt, struct rte_ether_hdr *); if (eth->ether_type == rte_cpu_to_be_16(ETHER_TYPE_IPv4)) { nlp = (uint8_t *)rte_pktmbuf_adj(pkt, ETHER_HDR_LEN); nlp = RTE_PTR_ADD(nlp, offsetof(struct ip, ip_p)); @@ -253,6 +245,8 @@ prepare_one_packet(struct rte_mbuf *pkt, struct ipsec_traffic *t) t->ip4.data[t->ip4.num] = nlp; t->ip4.pkts[(t->ip4.num)++] = pkt; } + pkt->l2_len = 0; + pkt->l3_len = sizeof(struct ip); } else if (eth->ether_type == rte_cpu_to_be_16(ETHER_TYPE_IPv6)) { nlp = (uint8_t *)rte_pktmbuf_adj(pkt, ETHER_HDR_LEN); nlp = RTE_PTR_ADD(nlp, offsetof(struct ip6_hdr, ip6_nxt)); @@ -262,11 +256,48 @@ prepare_one_packet(struct rte_mbuf *pkt, struct ipsec_traffic *t) t->ip6.data[t->ip6.num] = nlp; t->ip6.pkts[(t->ip6.num)++] = pkt; } + pkt->l2_len = 0; + pkt->l3_len = sizeof(struct ip6_hdr); } else { /* Unknown/Unsupported type, drop the packet */ - RTE_LOG(ERR, IPSEC, "Unsupported packet type\n"); + RTE_LOG(ERR, IPSEC, "Unsupported packet type 0x%x\n", + rte_be_to_cpu_16(eth->ether_type)); rte_pktmbuf_free(pkt); } + + /* Check if the packet has been processed inline. For inline protocol + * processed packets, the metadata in the mbuf can be used to identify + * the security processing done on the packet. The metadata will be + * used to retrieve the application registered userdata associated + * with the security session. + */ + + if (pkt->ol_flags & PKT_RX_SEC_OFFLOAD) { + struct ipsec_sa *sa; + struct ipsec_mbuf_metadata *priv; + struct rte_security_ctx *ctx = (struct rte_security_ctx *) + rte_eth_dev_get_sec_ctx( + pkt->port); + + /* Retrieve the userdata registered. Here, the userdata + * registered is the SA pointer. + */ + + sa = (struct ipsec_sa *) + rte_security_get_userdata(ctx, pkt->udata64); + + if (sa == NULL) { + /* userdata could not be retrieved */ + return; + } + + /* Save SA as priv member in mbuf. This will be used in the + * IPsec selector(SP-SA) check. + */ + + priv = get_priv(pkt); + priv->sa = sa; + } } static inline void @@ -290,23 +321,31 @@ prepare_traffic(struct rte_mbuf **pkts, struct ipsec_traffic *t, } static inline void -prepare_tx_pkt(struct rte_mbuf *pkt, uint8_t port) +prepare_tx_pkt(struct rte_mbuf *pkt, uint16_t port, + const struct lcore_conf *qconf) { struct ip *ip; - struct ether_hdr *ethhdr; + struct rte_ether_hdr *ethhdr; ip = rte_pktmbuf_mtod(pkt, struct ip *); - ethhdr = (struct ether_hdr *)rte_pktmbuf_prepend(pkt, ETHER_HDR_LEN); + ethhdr = (struct rte_ether_hdr *) + rte_pktmbuf_prepend(pkt, ETHER_HDR_LEN); if (ip->ip_v == IPVERSION) { - pkt->ol_flags |= PKT_TX_IP_CKSUM | PKT_TX_IPV4; + pkt->ol_flags |= qconf->outbound.ipv4_offloads; pkt->l3_len = sizeof(struct ip); pkt->l2_len = ETHER_HDR_LEN; + ip->ip_sum = 0; + + /* calculate IPv4 cksum in SW */ + if ((pkt->ol_flags & PKT_TX_IP_CKSUM) == 0) + ip->ip_sum = rte_ipv4_cksum((struct ipv4_hdr *)ip); + ethhdr->ether_type = rte_cpu_to_be_16(ETHER_TYPE_IPv4); } else { - pkt->ol_flags |= PKT_TX_IPV6; + pkt->ol_flags |= qconf->outbound.ipv6_offloads; pkt->l3_len = sizeof(struct ip6_hdr); pkt->l2_len = ETHER_HDR_LEN; @@ -314,29 +353,30 @@ prepare_tx_pkt(struct rte_mbuf *pkt, uint8_t port) } memcpy(ðhdr->s_addr, ðaddr_tbl[port].src, - sizeof(struct ether_addr)); + sizeof(struct rte_ether_addr)); memcpy(ðhdr->d_addr, ðaddr_tbl[port].dst, - sizeof(struct ether_addr)); + sizeof(struct rte_ether_addr)); } static inline void -prepare_tx_burst(struct rte_mbuf *pkts[], uint16_t nb_pkts, uint8_t port) +prepare_tx_burst(struct rte_mbuf *pkts[], uint16_t nb_pkts, uint16_t port, + const struct lcore_conf *qconf) { int32_t i; const int32_t prefetch_offset = 2; for (i = 0; i < (nb_pkts - prefetch_offset); i++) { rte_mbuf_prefetch_part2(pkts[i + prefetch_offset]); - prepare_tx_pkt(pkts[i], port); + prepare_tx_pkt(pkts[i], port, qconf); } /* Process left packets */ for (; i < nb_pkts; i++) - prepare_tx_pkt(pkts[i], port); + prepare_tx_pkt(pkts[i], port, qconf); } /* Send burst of packets on an output interface */ static inline int32_t -send_burst(struct lcore_conf *qconf, uint16_t n, uint8_t port) +send_burst(struct lcore_conf *qconf, uint16_t n, uint16_t port) { struct rte_mbuf **m_table; int32_t ret; @@ -345,7 +385,7 @@ send_burst(struct lcore_conf *qconf, uint16_t n, uint8_t port) queueid = qconf->tx_queue_id[port]; m_table = (struct rte_mbuf **)qconf->tx_mbufs[port].m_table; - prepare_tx_burst(m_table, n, port); + prepare_tx_burst(m_table, n, port, qconf); ret = rte_eth_tx_burst(port, queueid, m_table, n); if (unlikely(ret < n)) { @@ -359,7 +399,7 @@ send_burst(struct lcore_conf *qconf, uint16_t n, uint8_t port) /* Enqueue a single packet, and send burst if queue is filled */ static inline int32_t -send_single_packet(struct rte_mbuf *m, uint8_t port) +send_single_packet(struct rte_mbuf *m, uint16_t port) { uint32_t lcore_id; uint16_t len; @@ -399,17 +439,23 @@ inbound_sp_sa(struct sp_ctx *sp, struct sa_ctx *sa, struct traffic_type *ip, for (i = 0; i < ip->num; i++) { m = ip->pkts[i]; res = ip->res[i]; - if (res & BYPASS) { + if (res == BYPASS) { ip->pkts[j++] = m; continue; } - if (res & DISCARD || i < lim) { + if (res == DISCARD) { rte_pktmbuf_free(m); continue; } + /* Only check SPI match for processed IPSec packets */ - sa_idx = ip->res[i] & PROTECT_MASK; - if (sa_idx == 0 || !inbound_sa_check(sa, m, sa_idx)) { + if (i < lim && ((m->ol_flags & PKT_RX_SEC_OFFLOAD) == 0)) { + rte_pktmbuf_free(m); + continue; + } + + sa_idx = SPI2IDX(res); + if (!inbound_sa_check(sa, m, sa_idx)) { rte_pktmbuf_free(m); continue; } @@ -418,38 +464,60 @@ inbound_sp_sa(struct sp_ctx *sp, struct sa_ctx *sa, struct traffic_type *ip, ip->num = j; } -static inline void -process_pkts_inbound(struct ipsec_ctx *ipsec_ctx, - struct ipsec_traffic *traffic) +static void +split46_traffic(struct ipsec_traffic *trf, struct rte_mbuf *mb[], uint32_t num) { + uint32_t i, n4, n6; + struct ip *ip; struct rte_mbuf *m; - uint16_t idx, nb_pkts_in, i, n_ip4, n_ip6; - nb_pkts_in = ipsec_inbound(ipsec_ctx, traffic->ipsec.pkts, - traffic->ipsec.num, MAX_PKT_BURST); + n4 = trf->ip4.num; + n6 = trf->ip6.num; - n_ip4 = traffic->ip4.num; - n_ip6 = traffic->ip6.num; + for (i = 0; i < num; i++) { + + m = mb[i]; + ip = rte_pktmbuf_mtod(m, struct ip *); - /* SP/ACL Inbound check ipsec and ip4 */ - for (i = 0; i < nb_pkts_in; i++) { - m = traffic->ipsec.pkts[i]; - struct ip *ip = rte_pktmbuf_mtod(m, struct ip *); if (ip->ip_v == IPVERSION) { - idx = traffic->ip4.num++; - traffic->ip4.pkts[idx] = m; - traffic->ip4.data[idx] = rte_pktmbuf_mtod_offset(m, + trf->ip4.pkts[n4] = m; + trf->ip4.data[n4] = rte_pktmbuf_mtod_offset(m, uint8_t *, offsetof(struct ip, ip_p)); + n4++; } else if (ip->ip_v == IP6_VERSION) { - idx = traffic->ip6.num++; - traffic->ip6.pkts[idx] = m; - traffic->ip6.data[idx] = rte_pktmbuf_mtod_offset(m, + trf->ip6.pkts[n6] = m; + trf->ip6.data[n6] = rte_pktmbuf_mtod_offset(m, uint8_t *, offsetof(struct ip6_hdr, ip6_nxt)); + n6++; } else rte_pktmbuf_free(m); } + trf->ip4.num = n4; + trf->ip6.num = n6; +} + + +static inline void +process_pkts_inbound(struct ipsec_ctx *ipsec_ctx, + struct ipsec_traffic *traffic) +{ + uint16_t nb_pkts_in, n_ip4, n_ip6; + + n_ip4 = traffic->ip4.num; + n_ip6 = traffic->ip6.num; + + if (app_sa_prm.enable == 0) { + nb_pkts_in = ipsec_inbound(ipsec_ctx, traffic->ipsec.pkts, + traffic->ipsec.num, MAX_PKT_BURST); + split46_traffic(traffic, traffic->ipsec.pkts, nb_pkts_in); + } else { + inbound_sa_lookup(ipsec_ctx->sa_ctx, traffic->ipsec.pkts, + traffic->ipsec.saptr, traffic->ipsec.num); + ipsec_process(ipsec_ctx, traffic); + } + inbound_sp_sa(ipsec_ctx->sp4_ctx, ipsec_ctx->sa_ctx, &traffic->ip4, n_ip4); @@ -473,14 +541,15 @@ outbound_sp(struct sp_ctx *sp, struct traffic_type *ip, j = 0; for (i = 0; i < ip->num; i++) { m = ip->pkts[i]; - sa_idx = ip->res[i] & PROTECT_MASK; - if ((ip->res[i] == 0) || (ip->res[i] & DISCARD)) + sa_idx = SPI2IDX(ip->res[i]); + if (ip->res[i] == DISCARD) rte_pktmbuf_free(m); - else if (sa_idx != 0) { + else if (ip->res[i] == BYPASS) + ip->pkts[j++] = m; + else { ipsec->res[ipsec->num] = sa_idx; ipsec->pkts[ipsec->num++] = m; - } else /* BYPASS */ - ip->pkts[j++] = m; + } } ip->num = j; } @@ -502,20 +571,27 @@ process_pkts_outbound(struct ipsec_ctx *ipsec_ctx, outbound_sp(ipsec_ctx->sp6_ctx, &traffic->ip6, &traffic->ipsec); - nb_pkts_out = ipsec_outbound(ipsec_ctx, traffic->ipsec.pkts, - traffic->ipsec.res, traffic->ipsec.num, - MAX_PKT_BURST); - - for (i = 0; i < nb_pkts_out; i++) { - m = traffic->ipsec.pkts[i]; - struct ip *ip = rte_pktmbuf_mtod(m, struct ip *); - if (ip->ip_v == IPVERSION) { - idx = traffic->ip4.num++; - traffic->ip4.pkts[idx] = m; - } else { - idx = traffic->ip6.num++; - traffic->ip6.pkts[idx] = m; + if (app_sa_prm.enable == 0) { + + nb_pkts_out = ipsec_outbound(ipsec_ctx, traffic->ipsec.pkts, + traffic->ipsec.res, traffic->ipsec.num, + MAX_PKT_BURST); + + for (i = 0; i < nb_pkts_out; i++) { + m = traffic->ipsec.pkts[i]; + struct ip *ip = rte_pktmbuf_mtod(m, struct ip *); + if (ip->ip_v == IPVERSION) { + idx = traffic->ip4.num++; + traffic->ip4.pkts[idx] = m; + } else { + idx = traffic->ip6.num++; + traffic->ip6.pkts[idx] = m; + } } + } else { + outbound_sa_lookup(ipsec_ctx->sa_ctx, traffic->ipsec.res, + traffic->ipsec.saptr, traffic->ipsec.num); + ipsec_process(ipsec_ctx, traffic); } } @@ -538,19 +614,26 @@ process_pkts_inbound_nosp(struct ipsec_ctx *ipsec_ctx, traffic->ip6.num = 0; - nb_pkts_in = ipsec_inbound(ipsec_ctx, traffic->ipsec.pkts, - traffic->ipsec.num, MAX_PKT_BURST); + if (app_sa_prm.enable == 0) { - for (i = 0; i < nb_pkts_in; i++) { - m = traffic->ipsec.pkts[i]; - struct ip *ip = rte_pktmbuf_mtod(m, struct ip *); - if (ip->ip_v == IPVERSION) { - idx = traffic->ip4.num++; - traffic->ip4.pkts[idx] = m; - } else { - idx = traffic->ip6.num++; - traffic->ip6.pkts[idx] = m; + nb_pkts_in = ipsec_inbound(ipsec_ctx, traffic->ipsec.pkts, + traffic->ipsec.num, MAX_PKT_BURST); + + for (i = 0; i < nb_pkts_in; i++) { + m = traffic->ipsec.pkts[i]; + struct ip *ip = rte_pktmbuf_mtod(m, struct ip *); + if (ip->ip_v == IPVERSION) { + idx = traffic->ip4.num++; + traffic->ip4.pkts[idx] = m; + } else { + idx = traffic->ip6.num++; + traffic->ip6.pkts[idx] = m; + } } + } else { + inbound_sa_lookup(ipsec_ctx->sa_ctx, traffic->ipsec.pkts, + traffic->ipsec.saptr, traffic->ipsec.num); + ipsec_process(ipsec_ctx, traffic); } } @@ -559,32 +642,80 @@ process_pkts_outbound_nosp(struct ipsec_ctx *ipsec_ctx, struct ipsec_traffic *traffic) { struct rte_mbuf *m; - uint32_t nb_pkts_out, i; + uint32_t nb_pkts_out, i, n; struct ip *ip; /* Drop any IPsec traffic from protected ports */ for (i = 0; i < traffic->ipsec.num; i++) rte_pktmbuf_free(traffic->ipsec.pkts[i]); - traffic->ipsec.num = 0; + n = 0; - for (i = 0; i < traffic->ip4.num; i++) - traffic->ip4.res[i] = single_sa_idx; + for (i = 0; i < traffic->ip4.num; i++) { + traffic->ipsec.pkts[n] = traffic->ip4.pkts[i]; + traffic->ipsec.res[n++] = single_sa_idx; + } - for (i = 0; i < traffic->ip6.num; i++) - traffic->ip6.res[i] = single_sa_idx; + for (i = 0; i < traffic->ip6.num; i++) { + traffic->ipsec.pkts[n] = traffic->ip6.pkts[i]; + traffic->ipsec.res[n++] = single_sa_idx; + } - nb_pkts_out = ipsec_outbound(ipsec_ctx, traffic->ip4.pkts, - traffic->ip4.res, traffic->ip4.num, - MAX_PKT_BURST); + traffic->ip4.num = 0; + traffic->ip6.num = 0; + traffic->ipsec.num = n; - /* They all sue the same SA (ip4 or ip6 tunnel) */ - m = traffic->ipsec.pkts[i]; - ip = rte_pktmbuf_mtod(m, struct ip *); - if (ip->ip_v == IPVERSION) - traffic->ip4.num = nb_pkts_out; - else - traffic->ip6.num = nb_pkts_out; + if (app_sa_prm.enable == 0) { + + nb_pkts_out = ipsec_outbound(ipsec_ctx, traffic->ipsec.pkts, + traffic->ipsec.res, traffic->ipsec.num, + MAX_PKT_BURST); + + /* They all sue the same SA (ip4 or ip6 tunnel) */ + m = traffic->ipsec.pkts[0]; + ip = rte_pktmbuf_mtod(m, struct ip *); + if (ip->ip_v == IPVERSION) { + traffic->ip4.num = nb_pkts_out; + for (i = 0; i < nb_pkts_out; i++) + traffic->ip4.pkts[i] = traffic->ipsec.pkts[i]; + } else { + traffic->ip6.num = nb_pkts_out; + for (i = 0; i < nb_pkts_out; i++) + traffic->ip6.pkts[i] = traffic->ipsec.pkts[i]; + } + } else { + outbound_sa_lookup(ipsec_ctx->sa_ctx, traffic->ipsec.res, + traffic->ipsec.saptr, traffic->ipsec.num); + ipsec_process(ipsec_ctx, traffic); + } +} + +static inline int32_t +get_hop_for_offload_pkt(struct rte_mbuf *pkt, int is_ipv6) +{ + struct ipsec_mbuf_metadata *priv; + struct ipsec_sa *sa; + + priv = get_priv(pkt); + + sa = priv->sa; + if (unlikely(sa == NULL)) { + RTE_LOG(ERR, IPSEC, "SA not saved in private data\n"); + goto fail; + } + + if (is_ipv6) + return sa->portid; + + /* else */ + return (sa->portid | RTE_LPM_LOOKUP_SUCCESS); + +fail: + if (is_ipv6) + return -1; + + /* else */ + return 0; } static inline void @@ -592,61 +723,106 @@ route4_pkts(struct rt_ctx *rt_ctx, struct rte_mbuf *pkts[], uint8_t nb_pkts) { uint32_t hop[MAX_PKT_BURST * 2]; uint32_t dst_ip[MAX_PKT_BURST * 2]; + int32_t pkt_hop = 0; uint16_t i, offset; + uint16_t lpm_pkts = 0; if (nb_pkts == 0) return; + /* Need to do an LPM lookup for non-inline packets. Inline packets will + * have port ID in the SA + */ + for (i = 0; i < nb_pkts; i++) { - offset = offsetof(struct ip, ip_dst); - dst_ip[i] = *rte_pktmbuf_mtod_offset(pkts[i], - uint32_t *, offset); - dst_ip[i] = rte_be_to_cpu_32(dst_ip[i]); + if (!(pkts[i]->ol_flags & PKT_TX_SEC_OFFLOAD)) { + /* Security offload not enabled. So an LPM lookup is + * required to get the hop + */ + offset = offsetof(struct ip, ip_dst); + dst_ip[lpm_pkts] = *rte_pktmbuf_mtod_offset(pkts[i], + uint32_t *, offset); + dst_ip[lpm_pkts] = rte_be_to_cpu_32(dst_ip[lpm_pkts]); + lpm_pkts++; + } } - rte_lpm_lookup_bulk((struct rte_lpm *)rt_ctx, dst_ip, hop, nb_pkts); + rte_lpm_lookup_bulk((struct rte_lpm *)rt_ctx, dst_ip, hop, lpm_pkts); + + lpm_pkts = 0; for (i = 0; i < nb_pkts; i++) { - if ((hop[i] & RTE_LPM_LOOKUP_SUCCESS) == 0) { + if (pkts[i]->ol_flags & PKT_TX_SEC_OFFLOAD) { + /* Read hop from the SA */ + pkt_hop = get_hop_for_offload_pkt(pkts[i], 0); + } else { + /* Need to use hop returned by lookup */ + pkt_hop = hop[lpm_pkts++]; + } + + if ((pkt_hop & RTE_LPM_LOOKUP_SUCCESS) == 0) { rte_pktmbuf_free(pkts[i]); continue; } - send_single_packet(pkts[i], hop[i] & 0xff); + send_single_packet(pkts[i], pkt_hop & 0xff); } } static inline void route6_pkts(struct rt_ctx *rt_ctx, struct rte_mbuf *pkts[], uint8_t nb_pkts) { - int16_t hop[MAX_PKT_BURST * 2]; + int32_t hop[MAX_PKT_BURST * 2]; uint8_t dst_ip[MAX_PKT_BURST * 2][16]; uint8_t *ip6_dst; + int32_t pkt_hop = 0; uint16_t i, offset; + uint16_t lpm_pkts = 0; if (nb_pkts == 0) return; + /* Need to do an LPM lookup for non-inline packets. Inline packets will + * have port ID in the SA + */ + for (i = 0; i < nb_pkts; i++) { - offset = offsetof(struct ip6_hdr, ip6_dst); - ip6_dst = rte_pktmbuf_mtod_offset(pkts[i], uint8_t *, offset); - memcpy(&dst_ip[i][0], ip6_dst, 16); + if (!(pkts[i]->ol_flags & PKT_TX_SEC_OFFLOAD)) { + /* Security offload not enabled. So an LPM lookup is + * required to get the hop + */ + offset = offsetof(struct ip6_hdr, ip6_dst); + ip6_dst = rte_pktmbuf_mtod_offset(pkts[i], uint8_t *, + offset); + memcpy(&dst_ip[lpm_pkts][0], ip6_dst, 16); + lpm_pkts++; + } } - rte_lpm6_lookup_bulk_func((struct rte_lpm6 *)rt_ctx, dst_ip, - hop, nb_pkts); + rte_lpm6_lookup_bulk_func((struct rte_lpm6 *)rt_ctx, dst_ip, hop, + lpm_pkts); + + lpm_pkts = 0; for (i = 0; i < nb_pkts; i++) { - if (hop[i] == -1) { + if (pkts[i]->ol_flags & PKT_TX_SEC_OFFLOAD) { + /* Read hop from the SA */ + pkt_hop = get_hop_for_offload_pkt(pkts[i], 1); + } else { + /* Need to use hop returned by lookup */ + pkt_hop = hop[lpm_pkts++]; + } + + if (pkt_hop == -1) { rte_pktmbuf_free(pkts[i]); continue; } - send_single_packet(pkts[i], hop[i] & 0xff); + send_single_packet(pkts[i], pkt_hop & 0xff); } } static inline void process_pkts(struct lcore_conf *qconf, struct rte_mbuf **pkts, - uint8_t nb_pkts, uint8_t portid) + uint8_t nb_pkts, uint16_t portid) { struct ipsec_traffic traffic; @@ -669,7 +845,7 @@ process_pkts(struct lcore_conf *qconf, struct rte_mbuf **pkts, } static inline void -drain_buffers(struct lcore_conf *qconf) +drain_tx_buffers(struct lcore_conf *qconf) { struct buffer *buf; uint32_t portid; @@ -683,6 +859,91 @@ drain_buffers(struct lcore_conf *qconf) } } +static inline void +drain_crypto_buffers(struct lcore_conf *qconf) +{ + uint32_t i; + struct ipsec_ctx *ctx; + + /* drain inbound buffers*/ + ctx = &qconf->inbound; + for (i = 0; i != ctx->nb_qps; i++) { + if (ctx->tbl[i].len != 0) + enqueue_cop_burst(ctx->tbl + i); + } + + /* drain outbound buffers*/ + ctx = &qconf->outbound; + for (i = 0; i != ctx->nb_qps; i++) { + if (ctx->tbl[i].len != 0) + enqueue_cop_burst(ctx->tbl + i); + } +} + +static void +drain_inbound_crypto_queues(const struct lcore_conf *qconf, + struct ipsec_ctx *ctx) +{ + uint32_t n; + struct ipsec_traffic trf; + + if (app_sa_prm.enable == 0) { + + /* dequeue packets from crypto-queue */ + n = ipsec_inbound_cqp_dequeue(ctx, trf.ipsec.pkts, + RTE_DIM(trf.ipsec.pkts)); + + trf.ip4.num = 0; + trf.ip6.num = 0; + + /* split traffic by ipv4-ipv6 */ + split46_traffic(&trf, trf.ipsec.pkts, n); + } else + ipsec_cqp_process(ctx, &trf); + + /* process ipv4 packets */ + if (trf.ip4.num != 0) { + inbound_sp_sa(ctx->sp4_ctx, ctx->sa_ctx, &trf.ip4, 0); + route4_pkts(qconf->rt4_ctx, trf.ip4.pkts, trf.ip4.num); + } + + /* process ipv6 packets */ + if (trf.ip6.num != 0) { + inbound_sp_sa(ctx->sp6_ctx, ctx->sa_ctx, &trf.ip6, 0); + route6_pkts(qconf->rt6_ctx, trf.ip6.pkts, trf.ip6.num); + } +} + +static void +drain_outbound_crypto_queues(const struct lcore_conf *qconf, + struct ipsec_ctx *ctx) +{ + uint32_t n; + struct ipsec_traffic trf; + + if (app_sa_prm.enable == 0) { + + /* dequeue packets from crypto-queue */ + n = ipsec_outbound_cqp_dequeue(ctx, trf.ipsec.pkts, + RTE_DIM(trf.ipsec.pkts)); + + trf.ip4.num = 0; + trf.ip6.num = 0; + + /* split traffic by ipv4-ipv6 */ + split46_traffic(&trf, trf.ipsec.pkts, n); + } else + ipsec_cqp_process(ctx, &trf); + + /* process ipv4 packets */ + if (trf.ip4.num != 0) + route4_pkts(qconf->rt4_ctx, trf.ip4.pkts, trf.ip4.num); + + /* process ipv6 packets */ + if (trf.ip6.num != 0) + route6_pkts(qconf->rt6_ctx, trf.ip6.pkts, trf.ip6.num); +} + /* main processing loop */ static int32_t main_loop(__attribute__((unused)) void *dummy) @@ -691,7 +952,8 @@ main_loop(__attribute__((unused)) void *dummy) uint32_t lcore_id; uint64_t prev_tsc, diff_tsc, cur_tsc; int32_t i, nb_rx; - uint8_t portid, queueid; + uint16_t portid; + uint8_t queueid; struct lcore_conf *qconf; int32_t socket_id; const uint64_t drain_tsc = (rte_get_tsc_hz() + US_PER_S - 1) @@ -710,13 +972,20 @@ main_loop(__attribute__((unused)) void *dummy) qconf->inbound.sp6_ctx = socket_ctx[socket_id].sp_ip6_in; qconf->inbound.sa_ctx = socket_ctx[socket_id].sa_in; qconf->inbound.cdev_map = cdev_map_in; + qconf->inbound.session_pool = socket_ctx[socket_id].session_pool; + qconf->inbound.session_priv_pool = + socket_ctx[socket_id].session_priv_pool; qconf->outbound.sp4_ctx = socket_ctx[socket_id].sp_ip4_out; qconf->outbound.sp6_ctx = socket_ctx[socket_id].sp_ip6_out; qconf->outbound.sa_ctx = socket_ctx[socket_id].sa_out; qconf->outbound.cdev_map = cdev_map_out; + qconf->outbound.session_pool = socket_ctx[socket_id].session_pool; + qconf->outbound.session_priv_pool = + socket_ctx[socket_id].session_priv_pool; if (qconf->nb_rx_queue == 0) { - RTE_LOG(INFO, IPSEC, "lcore %u has nothing to do\n", lcore_id); + RTE_LOG(DEBUG, IPSEC, "lcore %u has nothing to do\n", + lcore_id); return 0; } @@ -726,7 +995,7 @@ main_loop(__attribute__((unused)) void *dummy) portid = rxql[i].port_id; queueid = rxql[i].queue_id; RTE_LOG(INFO, IPSEC, - " -- lcoreid=%u portid=%hhu rxqueueid=%hhu\n", + " -- lcoreid=%u portid=%u rxqueueid=%hhu\n", lcore_id, portid, queueid); } @@ -737,12 +1006,14 @@ main_loop(__attribute__((unused)) void *dummy) diff_tsc = cur_tsc - prev_tsc; if (unlikely(diff_tsc > drain_tsc)) { - drain_buffers(qconf); + drain_tx_buffers(qconf); + drain_crypto_buffers(qconf); prev_tsc = cur_tsc; } - /* Read packet from RX queues */ for (i = 0; i < qconf->nb_rx_queue; ++i) { + + /* Read packets from RX queues */ portid = rxql[i].port_id; queueid = rxql[i].queue_id; nb_rx = rte_eth_rx_burst(portid, queueid, @@ -750,6 +1021,14 @@ main_loop(__attribute__((unused)) void *dummy) if (nb_rx > 0) process_pkts(qconf, pkts, nb_rx, portid); + + /* dequeue and process completed crypto-ops */ + if (UNPROTECTED_PORT(portid)) + drain_inbound_crypto_queues(qconf, + &qconf->inbound); + else + drain_outbound_crypto_queues(qconf, + &qconf->outbound); } } } @@ -757,7 +1036,8 @@ main_loop(__attribute__((unused)) void *dummy) static int32_t check_params(void) { - uint8_t lcore, portid, nb_ports; + uint8_t lcore; + uint16_t portid; uint16_t i; int32_t socket_id; @@ -766,8 +1046,6 @@ check_params(void) return -1; } - nb_ports = rte_eth_dev_count(); - for (i = 0; i < nb_lcore_params; ++i) { lcore = lcore_params[i].lcore_id; if (!rte_lcore_is_enabled(lcore)) { @@ -786,7 +1064,7 @@ check_params(void) printf("port %u is not enabled in port mask\n", portid); return -1; } - if (portid >= nb_ports) { + if (!rte_eth_dev_is_valid_port(portid)) { printf("port %u is not present on the board\n", portid); return -1; } @@ -795,7 +1073,7 @@ check_params(void) } static uint8_t -get_port_nb_rx_queues(const uint8_t port) +get_port_nb_rx_queues(const uint16_t port) { int32_t queue = -1; uint16_t i; @@ -835,20 +1113,63 @@ init_lcore_rx_queues(void) static void print_usage(const char *prgname) { - printf("%s [EAL options] -- -p PORTMASK -P -u PORTMASK" - " --"OPTION_CONFIG" (port,queue,lcore)[,(port,queue,lcore]" - " --single-sa SAIDX -f CONFIG_FILE\n" - " -p PORTMASK: hexadecimal bitmask of ports to configure\n" - " -P : enable promiscuous mode\n" - " -u PORTMASK: hexadecimal bitmask of unprotected ports\n" - " --"OPTION_CONFIG": (port,queue,lcore): " - "rx queues configuration\n" - " --single-sa SAIDX: use single SA index for outbound, " - "bypassing the SP\n" - " -f CONFIG_FILE: Configuration file path\n", + fprintf(stderr, "%s [EAL options] --" + " -p PORTMASK" + " [-P]" + " [-u PORTMASK]" + " [-j FRAMESIZE]" + " [-l]" + " [-w REPLAY_WINDOW_SIZE]" + " [-e]" + " [-a]" + " -f CONFIG_FILE" + " --config (port,queue,lcore)[,(port,queue,lcore)]" + " [--single-sa SAIDX]" + " [--cryptodev_mask MASK]" + " [--" CMD_LINE_OPT_RX_OFFLOAD " RX_OFFLOAD_MASK]" + " [--" CMD_LINE_OPT_TX_OFFLOAD " TX_OFFLOAD_MASK]" + "\n\n" + " -p PORTMASK: Hexadecimal bitmask of ports to configure\n" + " -P : Enable promiscuous mode\n" + " -u PORTMASK: Hexadecimal bitmask of unprotected ports\n" + " -j FRAMESIZE: Enable jumbo frame with 'FRAMESIZE' as maximum\n" + " packet size\n" + " -l enables code-path that uses librte_ipsec\n" + " -w REPLAY_WINDOW_SIZE specifies IPsec SQN replay window\n" + " size for each SA\n" + " -e enables ESN\n" + " -a enables SA SQN atomic behaviour\n" + " -f CONFIG_FILE: Configuration file\n" + " --config (port,queue,lcore): Rx queue configuration\n" + " --single-sa SAIDX: Use single SA index for outbound traffic,\n" + " bypassing the SP\n" + " --cryptodev_mask MASK: Hexadecimal bitmask of the crypto\n" + " devices to configure\n" + " --" CMD_LINE_OPT_RX_OFFLOAD + ": bitmask of the RX HW offload capabilities to enable/use\n" + " (DEV_RX_OFFLOAD_*)\n" + " --" CMD_LINE_OPT_TX_OFFLOAD + ": bitmask of the TX HW offload capabilities to enable/use\n" + " (DEV_TX_OFFLOAD_*)\n" + "\n", prgname); } +static int +parse_mask(const char *str, uint64_t *val) +{ + char *end; + unsigned long t; + + errno = 0; + t = strtoul(str, &end, 0); + if (errno != 0 || end[0] != 0) + return -EINVAL; + + *val = t; + return 0; +} + static int32_t parse_portmask(const char *portmask) { @@ -935,33 +1256,19 @@ parse_config(const char *q_arg) return 0; } -#define __STRNCMP(name, opt) (!strncmp(name, opt, sizeof(opt))) -static int32_t -parse_args_long_options(struct option *lgopts, int32_t option_index) +static void +print_app_sa_prm(const struct app_sa_prm *prm) { - int32_t ret = -1; - const char *optname = lgopts[option_index].name; - - if (__STRNCMP(optname, OPTION_CONFIG)) { - ret = parse_config(optarg); - if (ret) - printf("invalid config\n"); - } + printf("librte_ipsec usage: %s\n", + (prm->enable == 0) ? "disabled" : "enabled"); - if (__STRNCMP(optname, OPTION_SINGLE_SA)) { - ret = parse_decimal(optarg); - if (ret != -1) { - single_sa = 1; - single_sa_idx = ret; - printf("Configured with single SA index %u\n", - single_sa_idx); - ret = 0; - } - } + if (prm->enable == 0) + return; - return ret; + printf("replay window size: %u\n", prm->window_size); + printf("ESN: %s\n", (prm->enable_esn == 0) ? "disabled" : "enabled"); + printf("SA flags: %#" PRIx64 "\n", prm->flags); } -#undef __STRNCMP static int32_t parse_args(int32_t argc, char **argv) @@ -970,16 +1277,11 @@ parse_args(int32_t argc, char **argv) char **argvopt; int32_t option_index; char *prgname = argv[0]; - static struct option lgopts[] = { - {OPTION_CONFIG, 1, 0, 0}, - {OPTION_SINGLE_SA, 1, 0, 0}, - {NULL, 0, 0, 0} - }; int32_t f_present = 0; argvopt = argv; - while ((opt = getopt_long(argc, argvopt, "p:Pu:f:", + while ((opt = getopt_long(argc, argvopt, "aelp:Pu:f:j:w:", lgopts, &option_index)) != EOF) { switch (opt) { @@ -1018,8 +1320,85 @@ parse_args(int32_t argc, char **argv) } f_present = 1; break; - case 0: - if (parse_args_long_options(lgopts, option_index)) { + case 'j': + { + int32_t size = parse_decimal(optarg); + if (size <= 1518) { + printf("Invalid jumbo frame size\n"); + if (size < 0) { + print_usage(prgname); + return -1; + } + printf("Using default value 9000\n"); + frame_size = 9000; + } else { + frame_size = size; + } + } + printf("Enabled jumbo frames size %u\n", frame_size); + break; + case 'l': + app_sa_prm.enable = 1; + break; + case 'w': + app_sa_prm.enable = 1; + app_sa_prm.window_size = parse_decimal(optarg); + break; + case 'e': + app_sa_prm.enable = 1; + app_sa_prm.enable_esn = 1; + break; + case 'a': + app_sa_prm.enable = 1; + app_sa_prm.flags |= RTE_IPSEC_SAFLAG_SQN_ATOM; + break; + case CMD_LINE_OPT_CONFIG_NUM: + ret = parse_config(optarg); + if (ret) { + printf("Invalid config\n"); + print_usage(prgname); + return -1; + } + break; + case CMD_LINE_OPT_SINGLE_SA_NUM: + ret = parse_decimal(optarg); + if (ret == -1) { + printf("Invalid argument[sa_idx]\n"); + print_usage(prgname); + return -1; + } + + /* else */ + single_sa = 1; + single_sa_idx = ret; + printf("Configured with single SA index %u\n", + single_sa_idx); + break; + case CMD_LINE_OPT_CRYPTODEV_MASK_NUM: + ret = parse_portmask(optarg); + if (ret == -1) { + printf("Invalid argument[portmask]\n"); + print_usage(prgname); + return -1; + } + + /* else */ + enabled_cryptodev_mask = ret; + break; + case CMD_LINE_OPT_RX_OFFLOAD_NUM: + ret = parse_mask(optarg, &dev_rx_offload); + if (ret != 0) { + printf("Invalid argument for \'%s\': %s\n", + CMD_LINE_OPT_RX_OFFLOAD, optarg); + print_usage(prgname); + return -1; + } + break; + case CMD_LINE_OPT_TX_OFFLOAD_NUM: + ret = parse_mask(optarg, &dev_tx_offload); + if (ret != 0) { + printf("Invalid argument for \'%s\': %s\n", + CMD_LINE_OPT_TX_OFFLOAD, optarg); print_usage(prgname); return -1; } @@ -1035,6 +1414,8 @@ parse_args(int32_t argc, char **argv) return -1; } + print_app_sa_prm(&app_sa_prm); + if (optind >= 0) argv[optind-1] = prgname; @@ -1044,27 +1425,41 @@ parse_args(int32_t argc, char **argv) } static void -print_ethaddr(const char *name, const struct ether_addr *eth_addr) +print_ethaddr(const char *name, const struct rte_ether_addr *eth_addr) { char buf[ETHER_ADDR_FMT_SIZE]; ether_format_addr(buf, ETHER_ADDR_FMT_SIZE, eth_addr); printf("%s%s", name, buf); } +/* + * Update destination ethaddr for the port. + */ +int +add_dst_ethaddr(uint16_t port, const struct rte_ether_addr *addr) +{ + if (port >= RTE_DIM(ethaddr_tbl)) + return -EINVAL; + + ethaddr_tbl[port].dst = ETHADDR_TO_UINT64(addr); + return 0; +} + /* Check the link status of all ports in up to 9s, and print them finally */ static void -check_all_ports_link_status(uint8_t port_num, uint32_t port_mask) +check_all_ports_link_status(uint32_t port_mask) { #define CHECK_INTERVAL 100 /* 100ms */ #define MAX_CHECK_TIME 90 /* 9s (90 * 100ms) in total */ - uint8_t portid, count, all_ports_up, print_flag = 0; + uint16_t portid; + uint8_t count, all_ports_up, print_flag = 0; struct rte_eth_link link; printf("\nChecking link status"); fflush(stdout); for (count = 0; count <= MAX_CHECK_TIME; count++) { all_ports_up = 1; - for (portid = 0; portid < port_num; portid++) { + RTE_ETH_FOREACH_DEV(portid) { if ((port_mask & (1 << portid)) == 0) continue; memset(&link, 0, sizeof(link)); @@ -1072,14 +1467,13 @@ check_all_ports_link_status(uint8_t port_num, uint32_t port_mask) /* print link status if flag set */ if (print_flag == 1) { if (link.link_status) - printf("Port %d Link Up - speed %u " - "Mbps - %s\n", (uint8_t)portid, - (uint32_t)link.link_speed, + printf( + "Port%d Link Up - speed %u Mbps -%s\n", + portid, link.link_speed, (link.link_duplex == ETH_LINK_FULL_DUPLEX) ? ("full-duplex") : ("half-duplex\n")); else - printf("Port %d Link Down\n", - (uint8_t)portid); + printf("Port %d Link Down\n", portid); continue; } /* clear all_ports_up flag if any link down */ @@ -1111,7 +1505,8 @@ add_mapping(struct rte_hash *map, const char *str, uint16_t cdev_id, uint16_t qp, struct lcore_params *params, struct ipsec_ctx *ipsec_ctx, const struct rte_cryptodev_capabilities *cipher, - const struct rte_cryptodev_capabilities *auth) + const struct rte_cryptodev_capabilities *auth, + const struct rte_cryptodev_capabilities *aead) { int32_t ret = 0; unsigned long i; @@ -1122,6 +1517,8 @@ add_mapping(struct rte_hash *map, const char *str, uint16_t cdev_id, key.cipher_algo = cipher->sym.cipher.algo; if (auth) key.auth_algo = auth->sym.auth.algo; + if (aead) + key.aead_algo = aead->sym.aead.algo; ret = rte_hash_lookup(map, &key); if (ret != -ENOENT) @@ -1190,6 +1587,12 @@ add_cdev_mapping(struct rte_cryptodev_info *dev_info, uint16_t cdev_id, if (i->op != RTE_CRYPTO_OP_TYPE_SYMMETRIC) continue; + if (i->sym.xform_type == RTE_CRYPTO_SYM_XFORM_AEAD) { + ret |= add_mapping(map, str, cdev_id, qp, params, + ipsec_ctx, NULL, NULL, i); + continue; + } + if (i->sym.xform_type != RTE_CRYPTO_SYM_XFORM_CIPHER) continue; @@ -1202,20 +1605,30 @@ add_cdev_mapping(struct rte_cryptodev_info *dev_info, uint16_t cdev_id, continue; ret |= add_mapping(map, str, cdev_id, qp, params, - ipsec_ctx, i, j); + ipsec_ctx, i, j, NULL); } } return ret; } +/* Check if the device is enabled by cryptodev_mask */ +static int +check_cryptodev_mask(uint8_t cdev_id) +{ + if (enabled_cryptodev_mask & (1 << cdev_id)) + return 0; + + return -1; +} + static int32_t cryptodevs_init(void) { struct rte_cryptodev_config dev_conf; struct rte_cryptodev_qp_conf qp_conf; uint16_t idx, max_nb_qps, qp, i; - int16_t cdev_id; + int16_t cdev_id, port_id; struct rte_hash_parameters params = { 0 }; params.entries = CDEV_MAP_ENTRIES; @@ -1238,11 +1651,52 @@ cryptodevs_init(void) printf("lcore/cryptodev/qp mappings:\n"); + uint32_t max_sess_sz = 0, sess_sz; + for (cdev_id = 0; cdev_id < rte_cryptodev_count(); cdev_id++) { + void *sec_ctx; + + /* Get crypto priv session size */ + sess_sz = rte_cryptodev_sym_get_private_session_size(cdev_id); + if (sess_sz > max_sess_sz) + max_sess_sz = sess_sz; + + /* + * If crypto device is security capable, need to check the + * size of security session as well. + */ + + /* Get security context of the crypto device */ + sec_ctx = rte_cryptodev_get_sec_ctx(cdev_id); + if (sec_ctx == NULL) + continue; + + /* Get size of security session */ + sess_sz = rte_security_session_get_size(sec_ctx); + if (sess_sz > max_sess_sz) + max_sess_sz = sess_sz; + } + RTE_ETH_FOREACH_DEV(port_id) { + void *sec_ctx; + + if ((enabled_port_mask & (1 << port_id)) == 0) + continue; + + sec_ctx = rte_eth_dev_get_sec_ctx(port_id); + if (sec_ctx == NULL) + continue; + + sess_sz = rte_security_session_get_size(sec_ctx); + if (sess_sz > max_sess_sz) + max_sess_sz = sess_sz; + } + idx = 0; - /* Start from last cdev id to give HW priority */ - for (cdev_id = rte_cryptodev_count() - 1; cdev_id >= 0; cdev_id--) { + for (cdev_id = 0; cdev_id < rte_cryptodev_count(); cdev_id++) { struct rte_cryptodev_info cdev_info; + if (check_cryptodev_mask((uint8_t)cdev_id)) + continue; + rte_cryptodev_info_get(cdev_id, &cdev_info); if (nb_lcore_params > cdev_info.max_nb_queue_pairs) @@ -1266,17 +1720,64 @@ cryptodevs_init(void) dev_conf.socket_id = rte_cryptodev_socket_id(cdev_id); dev_conf.nb_queue_pairs = qp; - dev_conf.session_mp.nb_objs = CDEV_MP_NB_OBJS; - dev_conf.session_mp.cache_size = CDEV_MP_CACHE_SZ; + + uint32_t dev_max_sess = cdev_info.sym.max_nb_sessions; + if (dev_max_sess != 0 && dev_max_sess < CDEV_MP_NB_OBJS) + rte_exit(EXIT_FAILURE, + "Device does not support at least %u " + "sessions", CDEV_MP_NB_OBJS); + + if (!socket_ctx[dev_conf.socket_id].session_pool) { + char mp_name[RTE_MEMPOOL_NAMESIZE]; + struct rte_mempool *sess_mp; + + snprintf(mp_name, RTE_MEMPOOL_NAMESIZE, + "sess_mp_%u", dev_conf.socket_id); + sess_mp = rte_cryptodev_sym_session_pool_create( + mp_name, CDEV_MP_NB_OBJS, + 0, CDEV_MP_CACHE_SZ, 0, + dev_conf.socket_id); + socket_ctx[dev_conf.socket_id].session_pool = sess_mp; + } + + if (!socket_ctx[dev_conf.socket_id].session_priv_pool) { + char mp_name[RTE_MEMPOOL_NAMESIZE]; + struct rte_mempool *sess_mp; + + snprintf(mp_name, RTE_MEMPOOL_NAMESIZE, + "sess_mp_priv_%u", dev_conf.socket_id); + sess_mp = rte_mempool_create(mp_name, + CDEV_MP_NB_OBJS, + max_sess_sz, + CDEV_MP_CACHE_SZ, + 0, NULL, NULL, NULL, + NULL, dev_conf.socket_id, + 0); + socket_ctx[dev_conf.socket_id].session_priv_pool = + sess_mp; + } + + if (!socket_ctx[dev_conf.socket_id].session_priv_pool || + !socket_ctx[dev_conf.socket_id].session_pool) + rte_exit(EXIT_FAILURE, + "Cannot create session pool on socket %d\n", + dev_conf.socket_id); + else + printf("Allocated session pool on socket %d\n", + dev_conf.socket_id); if (rte_cryptodev_configure(cdev_id, &dev_conf)) - rte_panic("Failed to initialize crypodev %u\n", + rte_panic("Failed to initialize cryptodev %u\n", cdev_id); qp_conf.nb_descriptors = CDEV_QUEUE_DESC; + qp_conf.mp_session = + socket_ctx[dev_conf.socket_id].session_pool; + qp_conf.mp_session_private = + socket_ctx[dev_conf.socket_id].session_priv_pool; for (qp = 0; qp < dev_conf.nb_queue_pairs; qp++) if (rte_cryptodev_queue_pair_setup(cdev_id, qp, - &qp_conf, dev_conf.socket_id)) + &qp_conf, dev_conf.socket_id)) rte_panic("Failed to setup queue %u for " "cdev_id %u\n", 0, cdev_id); @@ -1285,13 +1786,46 @@ cryptodevs_init(void) cdev_id); } + /* create session pools for eth devices that implement security */ + RTE_ETH_FOREACH_DEV(port_id) { + if ((enabled_port_mask & (1 << port_id)) && + rte_eth_dev_get_sec_ctx(port_id)) { + int socket_id = rte_eth_dev_socket_id(port_id); + + if (!socket_ctx[socket_id].session_priv_pool) { + char mp_name[RTE_MEMPOOL_NAMESIZE]; + struct rte_mempool *sess_mp; + + snprintf(mp_name, RTE_MEMPOOL_NAMESIZE, + "sess_mp_%u", socket_id); + sess_mp = rte_mempool_create(mp_name, + (CDEV_MP_NB_OBJS * 2), + max_sess_sz, + CDEV_MP_CACHE_SZ, + 0, NULL, NULL, NULL, + NULL, socket_id, + 0); + if (sess_mp == NULL) + rte_exit(EXIT_FAILURE, + "Cannot create session pool " + "on socket %d\n", socket_id); + else + printf("Allocated session pool " + "on socket %d\n", socket_id); + socket_ctx[socket_id].session_priv_pool = + sess_mp; + } + } + } + + printf("\n"); return 0; } static void -port_init(uint8_t portid) +port_init(uint16_t portid, uint64_t req_rx_offloads, uint64_t req_tx_offloads) { struct rte_eth_dev_info dev_info; struct rte_eth_txconf *txconf; @@ -1299,14 +1833,19 @@ port_init(uint8_t portid) uint16_t tx_queueid, rx_queueid, queue, lcore_id; int32_t ret, socket_id; struct lcore_conf *qconf; - struct ether_addr ethaddr; + struct rte_ether_addr ethaddr; + struct rte_eth_conf local_port_conf = port_conf; rte_eth_dev_info_get(portid, &dev_info); + /* limit allowed HW offloafs, as user requested */ + dev_info.rx_offload_capa &= dev_rx_offload; + dev_info.tx_offload_capa &= dev_tx_offload; + printf("Configuring device port %u:\n", portid); rte_eth_macaddr_get(portid, ðaddr); - ethaddr_tbl[portid].src = ETHADDR_TO_UINT64(ethaddr); + ethaddr_tbl[portid].src = ETHADDR_TO_UINT64(ðaddr); print_ethaddr("Address: ", ðaddr); printf("\n"); @@ -1326,12 +1865,65 @@ port_init(uint8_t portid) printf("Creating queues: nb_rx_queue=%d nb_tx_queue=%u...\n", nb_rx_queue, nb_tx_queue); + if (frame_size) { + local_port_conf.rxmode.max_rx_pkt_len = frame_size; + local_port_conf.rxmode.offloads |= DEV_RX_OFFLOAD_JUMBO_FRAME; + } + + local_port_conf.rxmode.offloads |= req_rx_offloads; + local_port_conf.txmode.offloads |= req_tx_offloads; + + /* Check that all required capabilities are supported */ + if ((local_port_conf.rxmode.offloads & dev_info.rx_offload_capa) != + local_port_conf.rxmode.offloads) + rte_exit(EXIT_FAILURE, + "Error: port %u required RX offloads: 0x%" PRIx64 + ", avaialbe RX offloads: 0x%" PRIx64 "\n", + portid, local_port_conf.rxmode.offloads, + dev_info.rx_offload_capa); + + if ((local_port_conf.txmode.offloads & dev_info.tx_offload_capa) != + local_port_conf.txmode.offloads) + rte_exit(EXIT_FAILURE, + "Error: port %u required TX offloads: 0x%" PRIx64 + ", avaialbe TX offloads: 0x%" PRIx64 "\n", + portid, local_port_conf.txmode.offloads, + dev_info.tx_offload_capa); + + if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_MBUF_FAST_FREE) + local_port_conf.txmode.offloads |= + DEV_TX_OFFLOAD_MBUF_FAST_FREE; + + if (dev_info.tx_offload_capa & DEV_TX_OFFLOAD_IPV4_CKSUM) + local_port_conf.txmode.offloads |= DEV_TX_OFFLOAD_IPV4_CKSUM; + + printf("port %u configurng rx_offloads=0x%" PRIx64 + ", tx_offloads=0x%" PRIx64 "\n", + portid, local_port_conf.rxmode.offloads, + local_port_conf.txmode.offloads); + + local_port_conf.rx_adv_conf.rss_conf.rss_hf &= + dev_info.flow_type_rss_offloads; + if (local_port_conf.rx_adv_conf.rss_conf.rss_hf != + port_conf.rx_adv_conf.rss_conf.rss_hf) { + printf("Port %u modified RSS hash function based on hardware support," + "requested:%#"PRIx64" configured:%#"PRIx64"\n", + portid, + port_conf.rx_adv_conf.rss_conf.rss_hf, + local_port_conf.rx_adv_conf.rss_conf.rss_hf); + } + ret = rte_eth_dev_configure(portid, nb_rx_queue, nb_tx_queue, - &port_conf); + &local_port_conf); if (ret < 0) rte_exit(EXIT_FAILURE, "Cannot configure device: " "err=%d, port=%d\n", ret, portid); + ret = rte_eth_dev_adjust_nb_rx_tx_desc(portid, &nb_rxd, &nb_txd); + if (ret < 0) + rte_exit(EXIT_FAILURE, "Cannot adjust number of descriptors: " + "err=%d, port=%d\n", ret, portid); + /* init one TX queue per lcore */ tx_queueid = 0; for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { @@ -1347,7 +1939,7 @@ port_init(uint8_t portid) printf("Setup txq=%u,%d,%d\n", lcore_id, tx_queueid, socket_id); txconf = &dev_info.default_txconf; - txconf->txq_flags = 0; + txconf->offloads = local_port_conf.txmode.offloads; ret = rte_eth_tx_queue_setup(portid, tx_queueid, nb_txd, socket_id, txconf); @@ -1357,10 +1949,19 @@ port_init(uint8_t portid) qconf = &lcore_conf[lcore_id]; qconf->tx_queue_id[portid] = tx_queueid; + + /* Pre-populate pkt offloads based on capabilities */ + qconf->outbound.ipv4_offloads = PKT_TX_IPV4; + qconf->outbound.ipv6_offloads = PKT_TX_IPV6; + if (local_port_conf.txmode.offloads & DEV_TX_OFFLOAD_IPV4_CKSUM) + qconf->outbound.ipv4_offloads |= PKT_TX_IP_CKSUM; + tx_queueid++; /* init RX queues */ for (queue = 0; queue < qconf->nb_rx_queue; ++queue) { + struct rte_eth_rxconf rxq_conf; + if (portid != qconf->rx_queue_list[queue].port_id) continue; @@ -1369,8 +1970,10 @@ port_init(uint8_t portid) printf("Setup rxq=%d,%d,%d\n", portid, rx_queueid, socket_id); + rxq_conf = dev_info.default_rxconf; + rxq_conf.offloads = local_port_conf.rxmode.offloads; ret = rte_eth_rx_queue_setup(portid, rx_queueid, - nb_rxd, socket_id, NULL, + nb_rxd, socket_id, &rxq_conf, socket_ctx[socket_id].mbuf_pool); if (ret < 0) rte_exit(EXIT_FAILURE, @@ -1385,11 +1988,14 @@ static void pool_init(struct socket_ctx *ctx, int32_t socket_id, uint32_t nb_mbuf) { char s[64]; + uint32_t buff_size = frame_size ? (frame_size + RTE_PKTMBUF_HEADROOM) : + RTE_MBUF_DEFAULT_BUF_SIZE; + snprintf(s, sizeof(s), "mbuf_pool_%d", socket_id); ctx->mbuf_pool = rte_pktmbuf_pool_create(s, nb_mbuf, MEMPOOL_CACHE_SIZE, ipsec_metadata_size(), - RTE_MBUF_DEFAULT_BUF_SIZE, + buff_size, socket_id); if (ctx->mbuf_pool == NULL) rte_exit(EXIT_FAILURE, "Cannot init mbuf pool on socket %d\n", @@ -1398,12 +2004,69 @@ pool_init(struct socket_ctx *ctx, int32_t socket_id, uint32_t nb_mbuf) printf("Allocated mbuf pool on socket %d\n", socket_id); } +static inline int +inline_ipsec_event_esn_overflow(struct rte_security_ctx *ctx, uint64_t md) +{ + struct ipsec_sa *sa; + + /* For inline protocol processing, the metadata in the event will + * uniquely identify the security session which raised the event. + * Application would then need the userdata it had registered with the + * security session to process the event. + */ + + sa = (struct ipsec_sa *)rte_security_get_userdata(ctx, md); + + if (sa == NULL) { + /* userdata could not be retrieved */ + return -1; + } + + /* Sequence number over flow. SA need to be re-established */ + RTE_SET_USED(sa); + return 0; +} + +static int +inline_ipsec_event_callback(uint16_t port_id, enum rte_eth_event_type type, + void *param, void *ret_param) +{ + uint64_t md; + struct rte_eth_event_ipsec_desc *event_desc = NULL; + struct rte_security_ctx *ctx = (struct rte_security_ctx *) + rte_eth_dev_get_sec_ctx(port_id); + + RTE_SET_USED(param); + + if (type != RTE_ETH_EVENT_IPSEC) + return -1; + + event_desc = ret_param; + if (event_desc == NULL) { + printf("Event descriptor not set\n"); + return -1; + } + + md = event_desc->metadata; + + if (event_desc->subtype == RTE_ETH_EVENT_IPSEC_ESN_OVERFLOW) + return inline_ipsec_event_esn_overflow(ctx, md); + else if (event_desc->subtype >= RTE_ETH_EVENT_IPSEC_MAX) { + printf("Invalid IPsec event reported\n"); + return -1; + } + + return -1; +} + int32_t main(int32_t argc, char **argv) { int32_t ret; - uint32_t lcore_id, nb_ports; - uint8_t portid, socket_id; + uint32_t lcore_id; + uint8_t socket_id; + uint16_t portid; + uint64_t req_rx_offloads, req_tx_offloads; /* init EAL */ ret = rte_eal_init(argc, argv); @@ -1422,8 +2085,6 @@ main(int32_t argc, char **argv) rte_exit(EXIT_FAILURE, "Invalid unprotected portmask 0x%x\n", unprotected_port_mask); - nb_ports = rte_eth_dev_count(); - if (check_params() < 0) rte_exit(EXIT_FAILURE, "check_params failed\n"); @@ -1433,7 +2094,7 @@ main(int32_t argc, char **argv) nb_lcores = rte_lcore_count(); - /* Replicate each contex per socket */ + /* Replicate each context per socket */ for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { if (rte_lcore_is_enabled(lcore_id) == 0) continue; @@ -1446,28 +2107,31 @@ main(int32_t argc, char **argv) if (socket_ctx[socket_id].mbuf_pool) continue; - sa_init(&socket_ctx[socket_id], socket_id); - + /* initilaze SPD */ sp4_init(&socket_ctx[socket_id], socket_id); sp6_init(&socket_ctx[socket_id], socket_id); + /* initilaze SAD */ + sa_init(&socket_ctx[socket_id], socket_id); + rt_init(&socket_ctx[socket_id], socket_id); pool_init(&socket_ctx[socket_id], socket_id, NB_MBUF); } - for (portid = 0; portid < nb_ports; portid++) { + RTE_ETH_FOREACH_DEV(portid) { if ((enabled_port_mask & (1 << portid)) == 0) continue; - port_init(portid); + sa_check_offloads(portid, &req_rx_offloads, &req_tx_offloads); + port_init(portid, req_rx_offloads, req_tx_offloads); } cryptodevs_init(); /* start ports */ - for (portid = 0; portid < nb_ports; portid++) { + RTE_ETH_FOREACH_DEV(portid) { if ((enabled_port_mask & (1 << portid)) == 0) continue; @@ -1484,9 +2148,12 @@ main(int32_t argc, char **argv) */ if (promiscuous_on) rte_eth_promiscuous_enable(portid); + + rte_eth_dev_callback_register(portid, + RTE_ETH_EVENT_IPSEC, inline_ipsec_event_callback, NULL); } - check_all_ports_link_status((uint8_t)nb_ports, enabled_port_mask); + check_all_ports_link_status(enabled_port_mask); /* launch per-lcore init on every lcore */ rte_eal_mp_remote_launch(main_loop, NULL, CALL_MASTER);