X-Git-Url: http://git.droids-corp.org/?a=blobdiff_plain;f=drivers%2Fnet%2Ftap%2Frte_eth_tap.c;h=ef33aace9167267fb4a216cbb7efabc772a66688;hb=6db80eb7c97e7db9c006641cc7ecae46c625d64e;hp=6567bba75b475d51ee15afb8508f0b596d4496d6;hpb=c0bddd3a057f1ef34d2d5aa69e7338caa82de2bf;p=dpdk.git diff --git a/drivers/net/tap/rte_eth_tap.c b/drivers/net/tap/rte_eth_tap.c index 6567bba75b..ef33aace91 100644 --- a/drivers/net/tap/rte_eth_tap.c +++ b/drivers/net/tap/rte_eth_tap.c @@ -1,45 +1,21 @@ -/*- - * 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) 2016-2017 Intel Corporation */ #include #include +#include #include #include -#include +#include +#include #include -#include +#include #include #include +#include +#include +#include #include #include @@ -49,6 +25,7 @@ #include #include #include +#include #include #include #include @@ -56,7 +33,6 @@ #include #include #include -#include #include #include @@ -67,30 +43,32 @@ /* Linux based path to the TUN device */ #define TUN_TAP_DEV_PATH "/dev/net/tun" #define DEFAULT_TAP_NAME "dtap" +#define DEFAULT_TUN_NAME "dtun" #define ETH_TAP_IFACE_ARG "iface" -#define ETH_TAP_SPEED_ARG "speed" #define ETH_TAP_REMOTE_ARG "remote" +#define ETH_TAP_MAC_ARG "mac" +#define ETH_TAP_MAC_FIXED "fixed" -#ifdef IFF_MULTI_QUEUE -#define RTE_PMD_TAP_MAX_QUEUES 16 -#else -#define RTE_PMD_TAP_MAX_QUEUES 1 -#endif - -#define FLOWER_KERNEL_VERSION KERNEL_VERSION(4, 2, 0) -#define FLOWER_VLAN_KERNEL_VERSION KERNEL_VERSION(4, 9, 0) +#define ETH_TAP_USR_MAC_FMT "xx:xx:xx:xx:xx:xx" +#define ETH_TAP_CMP_MAC_FMT "0123456789ABCDEFabcdef" +#define ETH_TAP_MAC_ARG_FMT ETH_TAP_MAC_FIXED "|" ETH_TAP_USR_MAC_FMT static struct rte_vdev_driver pmd_tap_drv; +static struct rte_vdev_driver pmd_tun_drv; static const char *valid_arguments[] = { ETH_TAP_IFACE_ARG, - ETH_TAP_SPEED_ARG, ETH_TAP_REMOTE_ARG, + ETH_TAP_MAC_ARG, NULL }; static int tap_unit; +static int tun_unit; + +static int tap_type; +static char tuntap_name[8]; static volatile uint32_t tap_trigger; /* Rx trigger */ @@ -98,7 +76,7 @@ static struct rte_eth_link pmd_link = { .link_speed = ETH_SPEED_NUM_10G, .link_duplex = ETH_LINK_FULL_DUPLEX, .link_status = ETH_LINK_DOWN, - .link_autoneg = ETH_LINK_SPEED_AUTONEG + .link_autoneg = ETH_LINK_AUTONEG }; static void @@ -108,9 +86,12 @@ tap_trigger_cb(int sig __rte_unused) tap_trigger = (tap_trigger + 1) | 0x80000000; } -static int -tap_ioctl(struct pmd_internals *pmd, unsigned long request, - struct ifreq *ifr, int set); +/* Specifies on what netdevices the ioctl should be applied */ +enum ioctl_mode { + LOCAL_AND_REMOTE, + LOCAL_ONLY, + REMOTE_ONLY, +}; static int tap_intr_handle_set(struct rte_eth_dev *dev, int set); @@ -120,7 +101,7 @@ static int tap_intr_handle_set(struct rte_eth_dev *dev, int set); * supplied name. */ static int -tun_alloc(struct pmd_internals *pmd, uint16_t qid) +tun_alloc(struct pmd_internals *pmd) { struct ifreq ifr; #ifdef IFF_MULTI_QUEUE @@ -134,24 +115,26 @@ tun_alloc(struct pmd_internals *pmd, uint16_t qid) * Do not set IFF_NO_PI as packet information header will be needed * to check if a received packet has been truncated. */ - ifr.ifr_flags = IFF_TAP; + ifr.ifr_flags = (tap_type) ? IFF_TAP : IFF_TUN | IFF_POINTOPOINT; snprintf(ifr.ifr_name, IFNAMSIZ, "%s", pmd->name); RTE_LOG(DEBUG, PMD, "ifr_name '%s'\n", ifr.ifr_name); fd = open(TUN_TAP_DEV_PATH, O_RDWR); if (fd < 0) { - RTE_LOG(ERR, PMD, "Unable to create TAP interface"); + RTE_LOG(ERR, PMD, "Unable to create %s interface\n", + tuntap_name); goto error; } #ifdef IFF_MULTI_QUEUE /* Grab the TUN features to verify we can work multi-queue */ if (ioctl(fd, TUNGETFEATURES, &features) < 0) { - RTE_LOG(ERR, PMD, "TAP unable to get TUN/TAP features\n"); + RTE_LOG(ERR, PMD, "%s unable to get TUN/TAP features\n", + tuntap_name); goto error; } - RTE_LOG(DEBUG, PMD, " TAP Features %08x\n", features); + RTE_LOG(DEBUG, PMD, "%s Features %08x\n", tuntap_name, features); if (features & IFF_MULTI_QUEUE) { RTE_LOG(DEBUG, PMD, " Multi-queue support for %d queues\n", @@ -219,70 +202,6 @@ tun_alloc(struct pmd_internals *pmd, uint16_t qid) strerror(errno)); } - if (qid == 0) { - struct ifreq ifr; - - if (tap_ioctl(pmd, SIOCGIFHWADDR, &ifr, 0) < 0) - goto error; - rte_memcpy(&pmd->eth_addr, ifr.ifr_hwaddr.sa_data, - ETHER_ADDR_LEN); - - pmd->if_index = if_nametoindex(pmd->name); - if (!pmd->if_index) { - RTE_LOG(ERR, PMD, - "Could not find ifindex for %s: rte_flow won't be usable.\n", - pmd->name); - return fd; - } - if (!pmd->flower_support) - return fd; - if (qdisc_create_multiq(pmd->nlsk_fd, pmd->if_index) < 0) { - RTE_LOG(ERR, PMD, - "Could not create multiq qdisc for %s: rte_flow won't be usable.\n", - pmd->name); - return fd; - } - if (qdisc_create_ingress(pmd->nlsk_fd, pmd->if_index) < 0) { - RTE_LOG(ERR, PMD, - "Could not create multiq qdisc for %s: rte_flow won't be usable.\n", - pmd->name); - return fd; - } - if (pmd->remote_if_index) { - /* - * Flush usually returns negative value because it tries - * to delete every QDISC (and on a running device, one - * QDISC at least is needed). Ignore negative return - * value. - */ - qdisc_flush(pmd->nlsk_fd, pmd->remote_if_index); - if (qdisc_create_ingress(pmd->nlsk_fd, - pmd->remote_if_index) < 0) - goto remote_fail; - LIST_INIT(&pmd->implicit_flows); - if (tap_flow_implicit_create( - pmd, TAP_REMOTE_LOCAL_MAC) < 0) - goto remote_fail; - if (tap_flow_implicit_create( - pmd, TAP_REMOTE_BROADCAST) < 0) - goto remote_fail; - if (tap_flow_implicit_create( - pmd, TAP_REMOTE_BROADCASTV6) < 0) - goto remote_fail; - if (tap_flow_implicit_create( - pmd, TAP_REMOTE_TX) < 0) - goto remote_fail; - } - } - - return fd; - -remote_fail: - RTE_LOG(ERR, PMD, - "Could not set up remote flow rules for %s: remote disabled.\n", - pmd->name); - pmd->remote_if_index = 0; - tap_flow_implicit_flush(pmd, NULL); return fd; error: @@ -291,6 +210,99 @@ error: return -1; } +static void +tap_verify_csum(struct rte_mbuf *mbuf) +{ + uint32_t l2 = mbuf->packet_type & RTE_PTYPE_L2_MASK; + uint32_t l3 = mbuf->packet_type & RTE_PTYPE_L3_MASK; + uint32_t l4 = mbuf->packet_type & RTE_PTYPE_L4_MASK; + unsigned int l2_len = sizeof(struct ether_hdr); + unsigned int l3_len; + uint16_t cksum = 0; + void *l3_hdr; + void *l4_hdr; + + if (l2 == RTE_PTYPE_L2_ETHER_VLAN) + l2_len += 4; + else if (l2 == RTE_PTYPE_L2_ETHER_QINQ) + l2_len += 8; + /* Don't verify checksum for packets with discontinuous L2 header */ + if (unlikely(l2_len + sizeof(struct ipv4_hdr) > + rte_pktmbuf_data_len(mbuf))) + return; + l3_hdr = rte_pktmbuf_mtod_offset(mbuf, void *, l2_len); + if (l3 == RTE_PTYPE_L3_IPV4 || l3 == RTE_PTYPE_L3_IPV4_EXT) { + struct ipv4_hdr *iph = l3_hdr; + + /* ihl contains the number of 4-byte words in the header */ + l3_len = 4 * (iph->version_ihl & 0xf); + if (unlikely(l2_len + l3_len > rte_pktmbuf_data_len(mbuf))) + return; + + cksum = ~rte_raw_cksum(iph, l3_len); + mbuf->ol_flags |= cksum ? + PKT_RX_IP_CKSUM_BAD : + PKT_RX_IP_CKSUM_GOOD; + } else if (l3 == RTE_PTYPE_L3_IPV6) { + l3_len = sizeof(struct ipv6_hdr); + } else { + /* IPv6 extensions are not supported */ + return; + } + if (l4 == RTE_PTYPE_L4_UDP || l4 == RTE_PTYPE_L4_TCP) { + l4_hdr = rte_pktmbuf_mtod_offset(mbuf, void *, l2_len + l3_len); + /* Don't verify checksum for multi-segment packets. */ + if (mbuf->nb_segs > 1) + return; + if (l3 == RTE_PTYPE_L3_IPV4) + cksum = ~rte_ipv4_udptcp_cksum(l3_hdr, l4_hdr); + else if (l3 == RTE_PTYPE_L3_IPV6) + cksum = ~rte_ipv6_udptcp_cksum(l3_hdr, l4_hdr); + mbuf->ol_flags |= cksum ? + PKT_RX_L4_CKSUM_BAD : + PKT_RX_L4_CKSUM_GOOD; + } +} + +static uint64_t +tap_rx_offload_get_port_capa(void) +{ + /* + * In order to support legacy apps, + * report capabilities also as port capabilities. + */ + return DEV_RX_OFFLOAD_SCATTER | + DEV_RX_OFFLOAD_IPV4_CKSUM | + DEV_RX_OFFLOAD_UDP_CKSUM | + DEV_RX_OFFLOAD_TCP_CKSUM | + DEV_RX_OFFLOAD_CRC_STRIP; +} + +static uint64_t +tap_rx_offload_get_queue_capa(void) +{ + return DEV_RX_OFFLOAD_SCATTER | + DEV_RX_OFFLOAD_IPV4_CKSUM | + DEV_RX_OFFLOAD_UDP_CKSUM | + DEV_RX_OFFLOAD_TCP_CKSUM | + DEV_RX_OFFLOAD_CRC_STRIP; +} + +static bool +tap_rxq_are_offloads_valid(struct rte_eth_dev *dev, uint64_t offloads) +{ + uint64_t port_offloads = dev->data->dev_conf.rxmode.offloads; + uint64_t queue_supp_offloads = tap_rx_offload_get_queue_capa(); + uint64_t port_supp_offloads = tap_rx_offload_get_port_capa(); + + if ((offloads & (queue_supp_offloads | port_supp_offloads)) != + offloads) + return false; + if ((port_offloads ^ offloads) & port_supp_offloads) + return false; + return true; +} + /* Callback to handle the rx burst of packets to the correct interface and * file descriptor(s) in a multi-queue setup. */ @@ -315,8 +327,9 @@ pmd_rx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) int len; len = readv(rxq->fd, *rxq->iovecs, - 1 + (rxq->rxmode->enable_scatter ? - rxq->nb_rx_desc : 1)); + 1 + + (rxq->rxmode->offloads & DEV_RX_OFFLOAD_SCATTER ? + rxq->nb_rx_desc : 1)); if (len < (int)sizeof(struct tun_pi)) break; @@ -371,6 +384,8 @@ pmd_rx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) seg->next = NULL; mbuf->packet_type = rte_net_get_ptype(mbuf, NULL, RTE_PTYPE_ALL_MASK); + if (rxq->rxmode->offloads & DEV_RX_OFFLOAD_CHECKSUM) + tap_verify_csum(mbuf); /* account for the receive frame */ bufs[num_rx++] = mbuf; @@ -383,6 +398,94 @@ end: return num_rx; } +static uint64_t +tap_tx_offload_get_port_capa(void) +{ + /* + * In order to support legacy apps, + * report capabilities also as port capabilities. + */ + return DEV_TX_OFFLOAD_MULTI_SEGS | + DEV_TX_OFFLOAD_IPV4_CKSUM | + DEV_TX_OFFLOAD_UDP_CKSUM | + DEV_TX_OFFLOAD_TCP_CKSUM; +} + +static uint64_t +tap_tx_offload_get_queue_capa(void) +{ + return DEV_TX_OFFLOAD_MULTI_SEGS | + DEV_TX_OFFLOAD_IPV4_CKSUM | + DEV_TX_OFFLOAD_UDP_CKSUM | + DEV_TX_OFFLOAD_TCP_CKSUM; +} + +static bool +tap_txq_are_offloads_valid(struct rte_eth_dev *dev, uint64_t offloads) +{ + uint64_t port_offloads = dev->data->dev_conf.txmode.offloads; + uint64_t queue_supp_offloads = tap_tx_offload_get_queue_capa(); + uint64_t port_supp_offloads = tap_tx_offload_get_port_capa(); + + if ((offloads & (queue_supp_offloads | port_supp_offloads)) != + offloads) + return false; + /* Verify we have no conflict with port offloads */ + if ((port_offloads ^ offloads) & port_supp_offloads) + return false; + return true; +} + +static void +tap_tx_offload(char *packet, uint64_t ol_flags, unsigned int l2_len, + unsigned int l3_len) +{ + void *l3_hdr = packet + l2_len; + + if (ol_flags & (PKT_TX_IP_CKSUM | PKT_TX_IPV4)) { + struct ipv4_hdr *iph = l3_hdr; + uint16_t cksum; + + iph->hdr_checksum = 0; + cksum = rte_raw_cksum(iph, l3_len); + iph->hdr_checksum = (cksum == 0xffff) ? cksum : ~cksum; + } + if (ol_flags & PKT_TX_L4_MASK) { + uint16_t l4_len; + uint32_t cksum; + uint16_t *l4_cksum; + void *l4_hdr; + + l4_hdr = packet + l2_len + l3_len; + if ((ol_flags & PKT_TX_L4_MASK) == PKT_TX_UDP_CKSUM) + l4_cksum = &((struct udp_hdr *)l4_hdr)->dgram_cksum; + else if ((ol_flags & PKT_TX_L4_MASK) == PKT_TX_TCP_CKSUM) + l4_cksum = &((struct tcp_hdr *)l4_hdr)->cksum; + else + return; + *l4_cksum = 0; + if (ol_flags & PKT_TX_IPV4) { + struct ipv4_hdr *iph = l3_hdr; + + l4_len = rte_be_to_cpu_16(iph->total_length) - l3_len; + cksum = rte_ipv4_phdr_cksum(l3_hdr, 0); + } else { + struct ipv6_hdr *ip6h = l3_hdr; + + /* payload_len does not include ext headers */ + l4_len = rte_be_to_cpu_16(ip6h->payload_len) - + l3_len + sizeof(struct ipv6_hdr); + cksum = rte_ipv6_phdr_cksum(l3_hdr, 0); + } + cksum += rte_raw_cksum(l4_hdr, l4_len); + cksum = ((cksum & 0xffff0000) >> 16) + (cksum & 0xffff); + cksum = (~cksum) & 0xffff; + if (cksum == 0) + cksum = 0xffff; + *l4_cksum = cksum; + } +} + /* Callback to handle sending packets from the tap interface */ static uint16_t @@ -401,8 +504,9 @@ pmd_tx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) for (i = 0; i < nb_pkts; i++) { struct rte_mbuf *mbuf = bufs[num_tx]; struct iovec iovecs[mbuf->nb_segs + 1]; - struct tun_pi pi = { .flags = 0 }; + struct tun_pi pi = { .flags = 0, .proto = 0x00 }; struct rte_mbuf *seg = mbuf; + char m_copy[mbuf->data_len]; int n; int j; @@ -410,6 +514,21 @@ pmd_tx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) if (rte_pktmbuf_pkt_len(mbuf) > max_size) break; + /* + * TUN and TAP are created with IFF_NO_PI disabled. + * For TUN PMD this mandatory as fields are used by + * Kernel tun.c to determine whether its IP or non IP + * packets. + * + * The logic fetches the first byte of data from mbuf. + * compares whether its v4 or v6. If none matches default + * value 0x00 is taken for protocol field. + */ + char *buff_data = rte_pktmbuf_mtod(seg, void *); + j = (*buff_data & 0xf0); + pi.proto = (j == 0x40) ? 0x0008 : + (j == 0x60) ? 0xdd86 : 0x00; + iovecs[0].iov_base = π iovecs[0].iov_len = sizeof(pi); for (j = 1; j <= mbuf->nb_segs; j++) { @@ -418,6 +537,20 @@ pmd_tx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) rte_pktmbuf_mtod(seg, void *); seg = seg->next; } + if (txq->csum && + ((mbuf->ol_flags & (PKT_TX_IP_CKSUM | PKT_TX_IPV4) || + (mbuf->ol_flags & PKT_TX_L4_MASK) == PKT_TX_UDP_CKSUM || + (mbuf->ol_flags & PKT_TX_L4_MASK) == PKT_TX_TCP_CKSUM))) { + /* Support only packets with all data in the same seg */ + if (mbuf->nb_segs > 1) + break; + /* To change checksums, work on a copy of data. */ + rte_memcpy(m_copy, rte_pktmbuf_mtod(mbuf, void *), + rte_pktmbuf_data_len(mbuf)); + tap_tx_offload(m_copy, mbuf->ol_flags, + mbuf->l2_len, mbuf->l3_len); + iovecs[1].iov_base = m_copy; + } /* copy the tx frame data */ n = writev(txq->fd, iovecs, mbuf->nb_segs + 1); if (n <= 0) @@ -435,29 +568,42 @@ pmd_tx_burst(void *queue, struct rte_mbuf **bufs, uint16_t nb_pkts) return num_tx; } +static const char * +tap_ioctl_req2str(unsigned long request) +{ + switch (request) { + case SIOCSIFFLAGS: + return "SIOCSIFFLAGS"; + case SIOCGIFFLAGS: + return "SIOCGIFFLAGS"; + case SIOCGIFHWADDR: + return "SIOCGIFHWADDR"; + case SIOCSIFHWADDR: + return "SIOCSIFHWADDR"; + case SIOCSIFMTU: + return "SIOCSIFMTU"; + } + return "UNKNOWN"; +} + static int tap_ioctl(struct pmd_internals *pmd, unsigned long request, - struct ifreq *ifr, int set) + struct ifreq *ifr, int set, enum ioctl_mode mode) { short req_flags = ifr->ifr_flags; - int remote = !!pmd->remote_if_index; + int remote = pmd->remote_if_index && + (mode == REMOTE_ONLY || mode == LOCAL_AND_REMOTE); + if (!pmd->remote_if_index && mode == REMOTE_ONLY) + return 0; /* * If there is a remote netdevice, apply ioctl on it, then apply it on * the tap netdevice. */ - if (request == SIOCGIFFLAGS && !set) { - /* - * Special case for getting flags. If set is given, - * then return the flags from the remote netdevice only. - * Otherwise return the flags from the tap netdevice. - */ - remote = 0; - } apply: if (remote) snprintf(ifr->ifr_name, IFNAMSIZ, "%s", pmd->remote_iface); - else + else if (mode == LOCAL_ONLY || mode == LOCAL_AND_REMOTE) snprintf(ifr->ifr_name, IFNAMSIZ, "%s", pmd->name); switch (request) { case SIOCSIFFLAGS: @@ -470,16 +616,7 @@ apply: ifr->ifr_flags &= ~req_flags; break; case SIOCGIFFLAGS: - if (remote && set) - remote = 0; /* don't loop */ - break; case SIOCGIFHWADDR: - /* Set remote MAC on the tap netdevice */ - if (!remote && pmd->remote_if_index) { - request = SIOCSIFHWADDR; - goto apply; - } - break; case SIOCSIFHWADDR: case SIOCSIFMTU: break; @@ -490,13 +627,13 @@ apply: } if (ioctl(pmd->ioctl_sock, request, ifr) < 0) goto error; - if (remote--) + if (remote-- && mode == LOCAL_AND_REMOTE) goto apply; return 0; error: - RTE_LOG(ERR, PMD, "%s: ioctl(%lu) failed with error: %s\n", - ifr->ifr_name, request, strerror(errno)); + RTE_LOG(DEBUG, PMD, "%s: %s(%s) failed: %s(%d)\n", ifr->ifr_name, + __func__, tap_ioctl_req2str(request), strerror(errno), errno); return -errno; } @@ -507,7 +644,7 @@ tap_link_set_down(struct rte_eth_dev *dev) struct ifreq ifr = { .ifr_flags = IFF_UP }; dev->data->dev_link.link_status = ETH_LINK_DOWN; - return tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0); + return tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0, LOCAL_ONLY); } static int @@ -517,7 +654,7 @@ tap_link_set_up(struct rte_eth_dev *dev) struct ifreq ifr = { .ifr_flags = IFF_UP }; dev->data->dev_link.link_status = ETH_LINK_UP; - return tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1); + return tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1, LOCAL_AND_REMOTE); } static int @@ -541,8 +678,42 @@ tap_dev_stop(struct rte_eth_dev *dev) } static int -tap_dev_configure(struct rte_eth_dev *dev __rte_unused) +tap_dev_configure(struct rte_eth_dev *dev) { + uint64_t supp_tx_offloads = tap_tx_offload_get_port_capa(); + uint64_t tx_offloads = dev->data->dev_conf.txmode.offloads; + + if ((tx_offloads & supp_tx_offloads) != tx_offloads) { + rte_errno = ENOTSUP; + RTE_LOG(ERR, PMD, + "Some Tx offloads are not supported " + "requested 0x%" PRIx64 " supported 0x%" PRIx64 "\n", + tx_offloads, supp_tx_offloads); + return -rte_errno; + } + if (dev->data->nb_rx_queues > RTE_PMD_TAP_MAX_QUEUES) { + RTE_LOG(ERR, PMD, + "%s: number of rx queues %d exceeds max num of queues %d\n", + dev->device->name, + dev->data->nb_rx_queues, + RTE_PMD_TAP_MAX_QUEUES); + return -1; + } + if (dev->data->nb_tx_queues > RTE_PMD_TAP_MAX_QUEUES) { + RTE_LOG(ERR, PMD, + "%s: number of tx queues %d exceeds max num of queues %d\n", + dev->device->name, + dev->data->nb_tx_queues, + RTE_PMD_TAP_MAX_QUEUES); + return -1; + } + + RTE_LOG(INFO, PMD, "%s: %p: TX configured queues number: %u\n", + dev->device->name, (void *)dev, dev->data->nb_tx_queues); + + RTE_LOG(INFO, PMD, "%s: %p: RX configured queues number: %u\n", + dev->device->name, (void *)dev, dev->data->nb_rx_queues); + return 0; } @@ -588,14 +759,19 @@ tap_dev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info) dev_info->if_index = internals->if_index; dev_info->max_mac_addrs = 1; dev_info->max_rx_pktlen = (uint32_t)ETHER_MAX_VLAN_FRAME_LEN; - dev_info->max_rx_queues = internals->nb_queues; - dev_info->max_tx_queues = internals->nb_queues; + dev_info->max_rx_queues = RTE_PMD_TAP_MAX_QUEUES; + dev_info->max_tx_queues = RTE_PMD_TAP_MAX_QUEUES; dev_info->min_rx_bufsize = 0; - dev_info->pci_dev = NULL; dev_info->speed_capa = tap_dev_speed_capa(); + dev_info->rx_queue_offload_capa = tap_rx_offload_get_queue_capa(); + dev_info->rx_offload_capa = tap_rx_offload_get_port_capa() | + dev_info->rx_queue_offload_capa; + dev_info->tx_queue_offload_capa = tap_tx_offload_get_queue_capa(); + dev_info->tx_offload_capa = tap_tx_offload_get_port_capa() | + dev_info->tx_queue_offload_capa; } -static void +static int tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats) { unsigned int i, imax; @@ -604,9 +780,9 @@ tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats) unsigned long rx_nombuf = 0, ierrors = 0; const struct pmd_internals *pmd = dev->data->dev_private; - imax = (pmd->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ? - pmd->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS; - + /* rx queue statistics */ + imax = (dev->data->nb_rx_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ? + dev->data->nb_rx_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS; for (i = 0; i < imax; i++) { tap_stats->q_ipackets[i] = pmd->rxq[i].stats.ipackets; tap_stats->q_ibytes[i] = pmd->rxq[i].stats.ibytes; @@ -614,7 +790,13 @@ tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats) rx_bytes_total += tap_stats->q_ibytes[i]; rx_nombuf += pmd->rxq[i].stats.rx_nombuf; ierrors += pmd->rxq[i].stats.ierrors; + } + + /* tx queue statistics */ + imax = (dev->data->nb_tx_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS) ? + dev->data->nb_tx_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS; + for (i = 0; i < imax; i++) { tap_stats->q_opackets[i] = pmd->txq[i].stats.opackets; tap_stats->q_errors[i] = pmd->txq[i].stats.errs; tap_stats->q_obytes[i] = pmd->txq[i].stats.obytes; @@ -630,6 +812,7 @@ tap_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *tap_stats) tap_stats->opackets = tx_total; tap_stats->oerrors = tx_err_total; tap_stats->obytes = tx_bytes_total; + return 0; } static void @@ -638,7 +821,7 @@ tap_stats_reset(struct rte_eth_dev *dev) int i; struct pmd_internals *pmd = dev->data->dev_private; - for (i = 0; i < pmd->nb_queues; i++) { + for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) { pmd->rxq[i].stats.ipackets = 0; pmd->rxq[i].stats.ibytes = 0; pmd->rxq[i].stats.ierrors = 0; @@ -651,7 +834,7 @@ tap_stats_reset(struct rte_eth_dev *dev) } static void -tap_dev_close(struct rte_eth_dev *dev __rte_unused) +tap_dev_close(struct rte_eth_dev *dev) { int i; struct pmd_internals *internals = dev->data->dev_private; @@ -660,11 +843,21 @@ tap_dev_close(struct rte_eth_dev *dev __rte_unused) tap_flow_flush(dev, NULL); tap_flow_implicit_flush(internals, NULL); - for (i = 0; i < internals->nb_queues; i++) { - if (internals->rxq[i].fd != -1) + for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) { + if (internals->rxq[i].fd != -1) { close(internals->rxq[i].fd); - internals->rxq[i].fd = -1; - internals->txq[i].fd = -1; + internals->rxq[i].fd = -1; + } + if (internals->txq[i].fd != -1) { + close(internals->txq[i].fd); + internals->txq[i].fd = -1; + } + } + + if (internals->remote_if_index) { + /* Restore initial remote state */ + ioctl(internals->ioctl_sock, SIOCSIFFLAGS, + &internals->remote_initial_flags); } } @@ -702,14 +895,14 @@ tap_link_update(struct rte_eth_dev *dev, int wait_to_complete __rte_unused) struct ifreq ifr = { .ifr_flags = 0 }; if (pmd->remote_if_index) { - tap_ioctl(pmd, SIOCGIFFLAGS, &ifr, 1); + tap_ioctl(pmd, SIOCGIFFLAGS, &ifr, 0, REMOTE_ONLY); if (!(ifr.ifr_flags & IFF_UP) || !(ifr.ifr_flags & IFF_RUNNING)) { dev_link->link_status = ETH_LINK_DOWN; return 0; } } - tap_ioctl(pmd, SIOCGIFFLAGS, &ifr, 0); + tap_ioctl(pmd, SIOCGIFFLAGS, &ifr, 0, LOCAL_ONLY); dev_link->link_status = ((ifr.ifr_flags & IFF_UP) && (ifr.ifr_flags & IFF_RUNNING) ? ETH_LINK_UP : @@ -724,8 +917,8 @@ tap_promisc_enable(struct rte_eth_dev *dev) struct ifreq ifr = { .ifr_flags = IFF_PROMISC }; dev->data->promiscuous = 1; - tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1); - if (pmd->remote_if_index) + tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1, LOCAL_AND_REMOTE); + if (pmd->remote_if_index && !pmd->flow_isolate) tap_flow_implicit_create(pmd, TAP_REMOTE_PROMISC); } @@ -736,8 +929,8 @@ tap_promisc_disable(struct rte_eth_dev *dev) struct ifreq ifr = { .ifr_flags = IFF_PROMISC }; dev->data->promiscuous = 0; - tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0); - if (pmd->remote_if_index) + tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0, LOCAL_AND_REMOTE); + if (pmd->remote_if_index && !pmd->flow_isolate) tap_flow_implicit_destroy(pmd, TAP_REMOTE_PROMISC); } @@ -748,8 +941,8 @@ tap_allmulti_enable(struct rte_eth_dev *dev) struct ifreq ifr = { .ifr_flags = IFF_ALLMULTI }; dev->data->all_multicast = 1; - tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1); - if (pmd->remote_if_index) + tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 1, LOCAL_AND_REMOTE); + if (pmd->remote_if_index && !pmd->flow_isolate) tap_flow_implicit_create(pmd, TAP_REMOTE_ALLMULTI); } @@ -760,91 +953,119 @@ tap_allmulti_disable(struct rte_eth_dev *dev) struct ifreq ifr = { .ifr_flags = IFF_ALLMULTI }; dev->data->all_multicast = 0; - tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0); - if (pmd->remote_if_index) + tap_ioctl(pmd, SIOCSIFFLAGS, &ifr, 0, LOCAL_AND_REMOTE); + if (pmd->remote_if_index && !pmd->flow_isolate) tap_flow_implicit_destroy(pmd, TAP_REMOTE_ALLMULTI); } - -static void +static int tap_mac_set(struct rte_eth_dev *dev, struct ether_addr *mac_addr) { struct pmd_internals *pmd = dev->data->dev_private; + enum ioctl_mode mode = LOCAL_ONLY; struct ifreq ifr; + int ret; if (is_zero_ether_addr(mac_addr)) { RTE_LOG(ERR, PMD, "%s: can't set an empty MAC address\n", - dev->data->name); - return; + dev->device->name); + return -EINVAL; } - + /* Check the actual current MAC address on the tap netdevice */ + ret = tap_ioctl(pmd, SIOCGIFHWADDR, &ifr, 0, LOCAL_ONLY); + if (ret < 0) + return ret; + if (is_same_ether_addr((struct ether_addr *)&ifr.ifr_hwaddr.sa_data, + mac_addr)) + return 0; + /* Check the current MAC address on the remote */ + ret = tap_ioctl(pmd, SIOCGIFHWADDR, &ifr, 0, REMOTE_ONLY); + if (ret < 0) + return ret; + if (!is_same_ether_addr((struct ether_addr *)&ifr.ifr_hwaddr.sa_data, + mac_addr)) + mode = LOCAL_AND_REMOTE; ifr.ifr_hwaddr.sa_family = AF_LOCAL; rte_memcpy(ifr.ifr_hwaddr.sa_data, mac_addr, ETHER_ADDR_LEN); - if (tap_ioctl(pmd, SIOCSIFHWADDR, &ifr, 1) < 0) - return; + ret = tap_ioctl(pmd, SIOCSIFHWADDR, &ifr, 1, mode); + if (ret < 0) + return ret; rte_memcpy(&pmd->eth_addr, mac_addr, ETHER_ADDR_LEN); + if (pmd->remote_if_index && !pmd->flow_isolate) { + /* Replace MAC redirection rule after a MAC change */ + ret = tap_flow_implicit_destroy(pmd, TAP_REMOTE_LOCAL_MAC); + if (ret < 0) { + RTE_LOG(ERR, PMD, + "%s: Couldn't delete MAC redirection rule\n", + dev->device->name); + return ret; + } + ret = tap_flow_implicit_create(pmd, TAP_REMOTE_LOCAL_MAC); + if (ret < 0) { + RTE_LOG(ERR, PMD, + "%s: Couldn't add MAC redirection rule\n", + dev->device->name); + return ret; + } + } + + return 0; } static int tap_setup_queue(struct rte_eth_dev *dev, struct pmd_internals *internals, - uint16_t qid) + uint16_t qid, + int is_rx) { + int *fd; + int *other_fd; + const char *dir; struct pmd_internals *pmd = dev->data->dev_private; struct rx_queue *rx = &internals->rxq[qid]; struct tx_queue *tx = &internals->txq[qid]; - int fd; - fd = rx->fd; - if (fd < 0) { - fd = tx->fd; - if (fd < 0) { - RTE_LOG(INFO, PMD, "Add queue to TAP %s for qid %d\n", - pmd->name, qid); - fd = tun_alloc(pmd, qid); - if (fd < 0) { - RTE_LOG(ERR, PMD, "tun_alloc(%s, %d) failed\n", - pmd->name, qid); - return -1; - } - if (qid == 0) { - struct ifreq ifr; - - ifr.ifr_mtu = dev->data->mtu; - if (tap_ioctl(pmd, SIOCSIFMTU, &ifr, 1) < 0) { - close(fd); - return -1; - } - } + if (is_rx) { + fd = &rx->fd; + other_fd = &tx->fd; + dir = "rx"; + } else { + fd = &tx->fd; + other_fd = &rx->fd; + dir = "tx"; + } + if (*fd != -1) { + /* fd for this queue already exists */ + RTE_LOG(DEBUG, PMD, "%s: fd %d for %s queue qid %d exists\n", + pmd->name, *fd, dir, qid); + } else if (*other_fd != -1) { + /* Only other_fd exists. dup it */ + *fd = dup(*other_fd); + if (*fd < 0) { + *fd = -1; + RTE_LOG(ERR, PMD, "%s: dup() failed.\n", + pmd->name); + return -1; } + RTE_LOG(DEBUG, PMD, "%s: dup fd %d for %s queue qid %d (%d)\n", + pmd->name, *other_fd, dir, qid, *fd); + } else { + /* Both RX and TX fds do not exist (equal -1). Create fd */ + *fd = tun_alloc(pmd); + if (*fd < 0) { + *fd = -1; /* restore original value */ + RTE_LOG(ERR, PMD, "%s: tun_alloc() failed.\n", + pmd->name); + return -1; + } + RTE_LOG(DEBUG, PMD, "%s: add %s queue for qid %d fd %d\n", + pmd->name, dir, qid, *fd); } - rx->fd = fd; - tx->fd = fd; tx->mtu = &dev->data->mtu; rx->rxmode = &dev->data->dev_conf.rxmode; - return fd; -} - -static int -rx_setup_queue(struct rte_eth_dev *dev, - struct pmd_internals *internals, - uint16_t qid) -{ - dev->data->rx_queues[qid] = &internals->rxq[qid]; - - return tap_setup_queue(dev, internals, qid); -} - -static int -tx_setup_queue(struct rte_eth_dev *dev, - struct pmd_internals *internals, - uint16_t qid) -{ - dev->data->tx_queues[qid] = &internals->txq[qid]; - - return tap_setup_queue(dev, internals, qid); + return *fd; } static int @@ -858,47 +1079,50 @@ tap_rx_queue_setup(struct rte_eth_dev *dev, struct pmd_internals *internals = dev->data->dev_private; struct rx_queue *rxq = &internals->rxq[rx_queue_id]; struct rte_mbuf **tmp = &rxq->pool; - struct iovec (*iovecs)[nb_rx_desc + 1]; + long iov_max = sysconf(_SC_IOV_MAX); + uint16_t nb_desc = RTE_MIN(nb_rx_desc, iov_max - 1); + struct iovec (*iovecs)[nb_desc + 1]; int data_off = RTE_PKTMBUF_HEADROOM; - uint16_t buf_size; int ret = 0; int fd; int i; - if ((rx_queue_id >= internals->nb_queues) || !mp) { + if (rx_queue_id >= dev->data->nb_rx_queues || !mp) { RTE_LOG(WARNING, PMD, - "nb_queues %d too small or mempool NULL\n", - internals->nb_queues); + "nb_rx_queues %d too small or mempool NULL\n", + dev->data->nb_rx_queues); return -1; } + /* Verify application offloads are valid for our port and queue. */ + if (!tap_rxq_are_offloads_valid(dev, rx_conf->offloads)) { + rte_errno = ENOTSUP; + RTE_LOG(ERR, PMD, + "%p: Rx queue offloads 0x%" PRIx64 + " don't match port offloads 0x%" PRIx64 + " or supported offloads 0x%" PRIx64 "\n", + (void *)dev, rx_conf->offloads, + dev->data->dev_conf.rxmode.offloads, + (tap_rx_offload_get_port_capa() | + tap_rx_offload_get_queue_capa())); + return -rte_errno; + } rxq->mp = mp; rxq->trigger_seen = 1; /* force initial burst */ rxq->in_port = dev->data->port_id; - rxq->nb_rx_desc = nb_rx_desc; - iovecs = rte_zmalloc_socket(dev->data->name, sizeof(*iovecs), 0, + rxq->nb_rx_desc = nb_desc; + iovecs = rte_zmalloc_socket(dev->device->name, sizeof(*iovecs), 0, socket_id); if (!iovecs) { RTE_LOG(WARNING, PMD, "%s: Couldn't allocate %d RX descriptors\n", - dev->data->name, nb_rx_desc); + dev->device->name, nb_desc); return -ENOMEM; } rxq->iovecs = iovecs; - /* Now get the space available for data in the mbuf */ - buf_size = (uint16_t)(rte_pktmbuf_data_room_size(mp) - - RTE_PKTMBUF_HEADROOM); - - if (buf_size < ETH_FRAME_LEN) { - RTE_LOG(WARNING, PMD, - "%s: %d bytes will not fit in mbuf (%d bytes)\n", - dev->data->name, ETH_FRAME_LEN, buf_size); - ret = -ENOMEM; - goto error; - } - - fd = rx_setup_queue(dev, internals, rx_queue_id); + dev->data->rx_queues[rx_queue_id] = rxq; + fd = tap_setup_queue(dev, internals, rx_queue_id, 1); if (fd == -1) { ret = fd; goto error; @@ -907,12 +1131,12 @@ tap_rx_queue_setup(struct rte_eth_dev *dev, (*rxq->iovecs)[0].iov_len = sizeof(struct tun_pi); (*rxq->iovecs)[0].iov_base = &rxq->pi; - for (i = 1; i <= nb_rx_desc; i++) { + for (i = 1; i <= nb_desc; i++) { *tmp = rte_pktmbuf_alloc(rxq->mp); if (!*tmp) { RTE_LOG(WARNING, PMD, "%s: couldn't allocate memory for queue %d\n", - dev->data->name, rx_queue_id); + dev->device->name, rx_queue_id); ret = -ENOMEM; goto error; } @@ -923,7 +1147,7 @@ tap_rx_queue_setup(struct rte_eth_dev *dev, tmp = &(*tmp)->next; } - RTE_LOG(DEBUG, PMD, " RX TAP device name %s, qid %d on fd %d\n", + RTE_LOG(DEBUG, PMD, " RX TUNTAP device name %s, qid %d on fd %d\n", internals->name, rx_queue_id, internals->rxq[rx_queue_id].fd); return 0; @@ -941,20 +1165,46 @@ tap_tx_queue_setup(struct rte_eth_dev *dev, uint16_t tx_queue_id, uint16_t nb_tx_desc __rte_unused, unsigned int socket_id __rte_unused, - const struct rte_eth_txconf *tx_conf __rte_unused) + const struct rte_eth_txconf *tx_conf) { struct pmd_internals *internals = dev->data->dev_private; + struct tx_queue *txq; int ret; - if (tx_queue_id >= internals->nb_queues) + if (tx_queue_id >= dev->data->nb_tx_queues) return -1; - - ret = tx_setup_queue(dev, internals, tx_queue_id); + dev->data->tx_queues[tx_queue_id] = &internals->txq[tx_queue_id]; + txq = dev->data->tx_queues[tx_queue_id]; + /* + * Don't verify port offloads for application which + * use the old API. + */ + if (tx_conf != NULL && + !!(tx_conf->txq_flags & ETH_TXQ_FLAGS_IGNORE)) { + if (tap_txq_are_offloads_valid(dev, tx_conf->offloads)) { + txq->csum = !!(tx_conf->offloads & + (DEV_TX_OFFLOAD_IPV4_CKSUM | + DEV_TX_OFFLOAD_UDP_CKSUM | + DEV_TX_OFFLOAD_TCP_CKSUM)); + } else { + rte_errno = ENOTSUP; + RTE_LOG(ERR, PMD, + "%p: Tx queue offloads 0x%" PRIx64 + " don't match port offloads 0x%" PRIx64 + " or supported offloads 0x%" PRIx64, + (void *)dev, tx_conf->offloads, + dev->data->dev_conf.txmode.offloads, + tap_tx_offload_get_port_capa()); + return -rte_errno; + } + } + ret = tap_setup_queue(dev, internals, tx_queue_id, 0); if (ret == -1) return -1; - - RTE_LOG(DEBUG, PMD, " TX TAP device name %s, qid %d on fd %d\n", - internals->name, tx_queue_id, internals->txq[tx_queue_id].fd); + RTE_LOG(DEBUG, PMD, + " TX TUNTAP device name %s, qid %d on fd %d csum %s\n", + internals->name, tx_queue_id, internals->txq[tx_queue_id].fd, + txq->csum ? "on" : "off"); return 0; } @@ -966,7 +1216,7 @@ tap_mtu_set(struct rte_eth_dev *dev, uint16_t mtu) struct ifreq ifr = { .ifr_mtu = mtu }; int err = 0; - err = tap_ioctl(pmd, SIOCSIFMTU, &ifr, 1); + err = tap_ioctl(pmd, SIOCSIFMTU, &ifr, 1, LOCAL_AND_REMOTE); if (!err) dev->data->mtu = mtu; @@ -1000,40 +1250,54 @@ tap_nl_msg_handler(struct nlmsghdr *nh, void *arg) } static void -tap_dev_intr_handler(struct rte_intr_handle *intr_handle __rte_unused, - void *cb_arg) +tap_dev_intr_handler(void *cb_arg) { struct rte_eth_dev *dev = cb_arg; struct pmd_internals *pmd = dev->data->dev_private; - nl_recv(pmd->intr_handle.fd, tap_nl_msg_handler, dev); + tap_nl_recv(pmd->intr_handle.fd, tap_nl_msg_handler, dev); } static int -tap_intr_handle_set(struct rte_eth_dev *dev, int set) +tap_lsc_intr_handle_set(struct rte_eth_dev *dev, int set) { struct pmd_internals *pmd = dev->data->dev_private; /* In any case, disable interrupt if the conf is no longer there. */ if (!dev->data->dev_conf.intr_conf.lsc) { - if (pmd->intr_handle.fd != -1) - nl_final(pmd->intr_handle.fd); - rte_intr_callback_unregister( - &pmd->intr_handle, tap_dev_intr_handler, dev); + if (pmd->intr_handle.fd != -1) { + tap_nl_final(pmd->intr_handle.fd); + rte_intr_callback_unregister(&pmd->intr_handle, + tap_dev_intr_handler, dev); + } return 0; } if (set) { - pmd->intr_handle.fd = nl_init(RTMGRP_LINK); + pmd->intr_handle.fd = tap_nl_init(RTMGRP_LINK); if (unlikely(pmd->intr_handle.fd == -1)) return -EBADF; return rte_intr_callback_register( &pmd->intr_handle, tap_dev_intr_handler, dev); } - nl_final(pmd->intr_handle.fd); + tap_nl_final(pmd->intr_handle.fd); return rte_intr_callback_unregister(&pmd->intr_handle, tap_dev_intr_handler, dev); } +static int +tap_intr_handle_set(struct rte_eth_dev *dev, int set) +{ + int err; + + err = tap_lsc_intr_handle_set(dev, set); + if (err) + return err; + err = tap_rx_intr_vec_set(dev, set); + if (err && set) + tap_lsc_intr_handle_set(dev, 0); + return err; +} + static const uint32_t* tap_dev_supported_ptypes_get(struct rte_eth_dev *dev __rte_unused) { @@ -1111,88 +1375,57 @@ static const struct eth_dev_ops ops = { }; static int -tap_kernel_support(struct pmd_internals *pmd) -{ - struct utsname utsname; - int ver[3]; - - if (uname(&utsname) == -1 || - sscanf(utsname.release, "%d.%d.%d", - &ver[0], &ver[1], &ver[2]) != 3) - return 0; - if (KERNEL_VERSION(ver[0], ver[1], ver[2]) >= FLOWER_KERNEL_VERSION) - pmd->flower_support = 1; - if (KERNEL_VERSION(ver[0], ver[1], ver[2]) >= - FLOWER_VLAN_KERNEL_VERSION) - pmd->flower_vlan_support = 1; - return 1; -} - -static int -eth_dev_tap_create(const char *name, char *tap_name, char *remote_iface) +eth_dev_tap_create(struct rte_vdev_device *vdev, char *tap_name, + char *remote_iface, struct ether_addr *mac_addr) { int numa_node = rte_socket_id(); - struct rte_eth_dev *dev = NULL; - struct pmd_internals *pmd = NULL; - struct rte_eth_dev_data *data = NULL; + struct rte_eth_dev *dev; + struct pmd_internals *pmd; + struct rte_eth_dev_data *data; + struct ifreq ifr; int i; - RTE_LOG(DEBUG, PMD, " TAP device on numa %u\n", rte_socket_id()); + RTE_LOG(DEBUG, PMD, "%s device on numa %u\n", + tuntap_name, rte_socket_id()); - data = rte_zmalloc_socket(tap_name, sizeof(*data), 0, numa_node); - if (!data) { - RTE_LOG(ERR, PMD, "TAP Failed to allocate data\n"); - goto error_exit; - } - - pmd = rte_zmalloc_socket(tap_name, sizeof(*pmd), 0, numa_node); - if (!pmd) { - RTE_LOG(ERR, PMD, "TAP Unable to allocate internal struct\n"); - goto error_exit; - } - - /* name in allocation and data->name must be consistent */ - snprintf(data->name, sizeof(data->name), "%s", name); - dev = rte_eth_dev_allocate(name); + dev = rte_eth_vdev_allocate(vdev, sizeof(*pmd)); if (!dev) { - RTE_LOG(ERR, PMD, "TAP Unable to allocate device struct\n"); - goto error_exit; + RTE_LOG(ERR, PMD, "%s Unable to allocate device struct\n", + tuntap_name); + goto error_exit_nodev; } + pmd = dev->data->dev_private; + pmd->dev = dev; snprintf(pmd->name, sizeof(pmd->name), "%s", tap_name); - pmd->nb_queues = RTE_PMD_TAP_MAX_QUEUES; - pmd->ioctl_sock = socket(AF_INET, SOCK_DGRAM, 0); if (pmd->ioctl_sock == -1) { RTE_LOG(ERR, PMD, - "TAP Unable to get a socket for management: %s\n", - strerror(errno)); + "%s Unable to get a socket for management: %s\n", + tuntap_name, strerror(errno)); goto error_exit; } /* Setup some default values */ + data = dev->data; data->dev_private = pmd; - data->port_id = dev->data->port_id; - data->mtu = dev->data->mtu; - data->dev_flags = RTE_ETH_DEV_DETACHABLE | RTE_ETH_DEV_INTR_LSC; - data->kdrv = RTE_KDRV_NONE; - data->drv_name = pmd_tap_drv.driver.name; + data->dev_flags = RTE_ETH_DEV_INTR_LSC; data->numa_node = numa_node; data->dev_link = pmd_link; data->mac_addrs = &pmd->eth_addr; - data->nb_rx_queues = pmd->nb_queues; - data->nb_tx_queues = pmd->nb_queues; + /* Set the number of RX and TX queues */ + data->nb_rx_queues = 0; + data->nb_tx_queues = 0; - dev->data = data; dev->dev_ops = &ops; - dev->driver = NULL; dev->rx_pkt_burst = pmd_rx_burst; dev->tx_pkt_burst = pmd_tx_burst; pmd->intr_handle.type = RTE_INTR_HANDLE_EXT; pmd->intr_handle.fd = -1; + dev->intr_handle = &pmd->intr_handle; /* Presetup the fds to -1 as being not valid */ for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) { @@ -1200,35 +1433,142 @@ eth_dev_tap_create(const char *name, char *tap_name, char *remote_iface) pmd->txq[i].fd = -1; } - tap_kernel_support(pmd); - if (!pmd->flower_support) - return 0; - LIST_INIT(&pmd->flows); + if (tap_type) { + if (is_zero_ether_addr(mac_addr)) + eth_random_addr((uint8_t *)&pmd->eth_addr); + else + rte_memcpy(&pmd->eth_addr, mac_addr, sizeof(*mac_addr)); + } + + /* Immediately create the netdevice (this will create the 1st queue). */ + /* rx queue */ + if (tap_setup_queue(dev, pmd, 0, 1) == -1) + goto error_exit; + /* tx queue */ + if (tap_setup_queue(dev, pmd, 0, 0) == -1) + goto error_exit; + + ifr.ifr_mtu = dev->data->mtu; + if (tap_ioctl(pmd, SIOCSIFMTU, &ifr, 1, LOCAL_AND_REMOTE) < 0) + goto error_exit; + + if (tap_type) { + memset(&ifr, 0, sizeof(struct ifreq)); + ifr.ifr_hwaddr.sa_family = AF_LOCAL; + rte_memcpy(ifr.ifr_hwaddr.sa_data, &pmd->eth_addr, + ETHER_ADDR_LEN); + if (tap_ioctl(pmd, SIOCSIFHWADDR, &ifr, 0, LOCAL_ONLY) < 0) + goto error_exit; + } + /* - * If no netlink socket can be created, then it will fail when - * creating/destroying flow rules. + * Set up everything related to rte_flow: + * - netlink socket + * - tap / remote if_index + * - mandatory QDISCs + * - rte_flow actual/implicit lists + * - implicit rules */ - pmd->nlsk_fd = nl_init(0); + pmd->nlsk_fd = tap_nl_init(0); + if (pmd->nlsk_fd == -1) { + RTE_LOG(WARNING, PMD, "%s: failed to create netlink socket.\n", + pmd->name); + goto disable_rte_flow; + } + pmd->if_index = if_nametoindex(pmd->name); + if (!pmd->if_index) { + RTE_LOG(ERR, PMD, "%s: failed to get if_index.\n", pmd->name); + goto disable_rte_flow; + } + if (qdisc_create_multiq(pmd->nlsk_fd, pmd->if_index) < 0) { + RTE_LOG(ERR, PMD, "%s: failed to create multiq qdisc.\n", + pmd->name); + goto disable_rte_flow; + } + if (qdisc_create_ingress(pmd->nlsk_fd, pmd->if_index) < 0) { + RTE_LOG(ERR, PMD, "%s: failed to create ingress qdisc.\n", + pmd->name); + goto disable_rte_flow; + } + LIST_INIT(&pmd->flows); + if (strlen(remote_iface)) { pmd->remote_if_index = if_nametoindex(remote_iface); + if (!pmd->remote_if_index) { + RTE_LOG(ERR, PMD, "%s: failed to get %s if_index.\n", + pmd->name, remote_iface); + goto error_remote; + } snprintf(pmd->remote_iface, RTE_ETH_NAME_MAX_LEN, "%s", remote_iface); - if (!pmd->remote_if_index) - RTE_LOG(ERR, PMD, "Could not find %s ifindex: " - "remote interface will remain unconfigured\n", - remote_iface); + + /* Save state of remote device */ + tap_ioctl(pmd, SIOCGIFFLAGS, &pmd->remote_initial_flags, 0, REMOTE_ONLY); + + /* Replicate remote MAC address */ + if (tap_ioctl(pmd, SIOCGIFHWADDR, &ifr, 0, REMOTE_ONLY) < 0) { + RTE_LOG(ERR, PMD, "%s: failed to get %s MAC address.\n", + pmd->name, pmd->remote_iface); + goto error_remote; + } + rte_memcpy(&pmd->eth_addr, ifr.ifr_hwaddr.sa_data, + ETHER_ADDR_LEN); + /* The desired MAC is already in ifreq after SIOCGIFHWADDR. */ + if (tap_ioctl(pmd, SIOCSIFHWADDR, &ifr, 0, LOCAL_ONLY) < 0) { + RTE_LOG(ERR, PMD, "%s: failed to get %s MAC address.\n", + pmd->name, remote_iface); + goto error_remote; + } + + /* + * Flush usually returns negative value because it tries to + * delete every QDISC (and on a running device, one QDISC at + * least is needed). Ignore negative return value. + */ + qdisc_flush(pmd->nlsk_fd, pmd->remote_if_index); + if (qdisc_create_ingress(pmd->nlsk_fd, + pmd->remote_if_index) < 0) { + RTE_LOG(ERR, PMD, "%s: failed to create ingress qdisc.\n", + pmd->remote_iface); + goto error_remote; + } + LIST_INIT(&pmd->implicit_flows); + if (tap_flow_implicit_create(pmd, TAP_REMOTE_TX) < 0 || + tap_flow_implicit_create(pmd, TAP_REMOTE_LOCAL_MAC) < 0 || + tap_flow_implicit_create(pmd, TAP_REMOTE_BROADCAST) < 0 || + tap_flow_implicit_create(pmd, TAP_REMOTE_BROADCASTV6) < 0) { + RTE_LOG(ERR, PMD, + "%s: failed to create implicit rules.\n", + pmd->name); + goto error_remote; + } } return 0; -error_exit: - RTE_LOG(DEBUG, PMD, "TAP Unable to initialize %s\n", name); +disable_rte_flow: + RTE_LOG(ERR, PMD, " Disabling rte flow support: %s(%d)\n", + strerror(errno), errno); + if (strlen(remote_iface)) { + RTE_LOG(ERR, PMD, "Remote feature requires flow support.\n"); + goto error_exit; + } + return 0; - rte_free(data); - rte_free(pmd); +error_remote: + RTE_LOG(ERR, PMD, " Can't set up remote feature: %s(%d)\n", + strerror(errno), errno); + tap_flow_implicit_flush(pmd, NULL); +error_exit: + if (pmd->ioctl_sock > 0) + close(pmd->ioctl_sock); rte_eth_dev_release_port(dev); +error_exit_nodev: + RTE_LOG(ERR, PMD, "%s Unable to initialize %s\n", + tuntap_name, rte_vdev_device_name(vdev)); + return -EINVAL; } @@ -1240,7 +1580,7 @@ set_interface_name(const char *key __rte_unused, char *name = (char *)extra_args; if (value) - snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s", value); + strlcpy(name, value, RTE_ETH_NAME_MAX_LEN - 1); else snprintf(name, RTE_ETH_NAME_MAX_LEN - 1, "%s%d", DEFAULT_TAP_NAME, (tap_unit - 1)); @@ -1249,38 +1589,159 @@ set_interface_name(const char *key __rte_unused, } static int -set_interface_speed(const char *key __rte_unused, - const char *value, - void *extra_args) +set_remote_iface(const char *key __rte_unused, + const char *value, + void *extra_args) { - *(int *)extra_args = (value) ? atoi(value) : ETH_SPEED_NUM_10G; + char *name = (char *)extra_args; + + if (value) + strlcpy(name, value, RTE_ETH_NAME_MAX_LEN); return 0; } +static int parse_user_mac(struct ether_addr *user_mac, + const char *value) +{ + unsigned int index = 0; + char mac_temp[strlen(ETH_TAP_USR_MAC_FMT) + 1], *mac_byte = NULL; + + if (user_mac == NULL || value == NULL) + return 0; + + strlcpy(mac_temp, value, sizeof(mac_temp)); + mac_byte = strtok(mac_temp, ":"); + + while ((mac_byte != NULL) && + (strlen(mac_byte) <= 2) && + (strlen(mac_byte) == strspn(mac_byte, + ETH_TAP_CMP_MAC_FMT))) { + user_mac->addr_bytes[index++] = strtoul(mac_byte, NULL, 16); + mac_byte = strtok(NULL, ":"); + } + + return index; +} + static int -set_remote_iface(const char *key __rte_unused, - const char *value, - void *extra_args) +set_mac_type(const char *key __rte_unused, + const char *value, + void *extra_args) { - char *name = (char *)extra_args; + struct ether_addr *user_mac = extra_args; - if (value) - snprintf(name, RTE_ETH_NAME_MAX_LEN, "%s", value); + if (!value) + return 0; + + if (!strncasecmp(ETH_TAP_MAC_FIXED, value, strlen(ETH_TAP_MAC_FIXED))) { + static int iface_idx; + + /* fixed mac = 00:64:74:61:70: */ + memcpy((char *)user_mac->addr_bytes, "\0dtap", ETHER_ADDR_LEN); + user_mac->addr_bytes[ETHER_ADDR_LEN - 1] = iface_idx++ + '0'; + goto success; + } + if (parse_user_mac(user_mac, value) != 6) + goto error; +success: + RTE_LOG(DEBUG, PMD, "TAP user MAC param (%s)\n", value); return 0; + +error: + RTE_LOG(ERR, PMD, "TAP user MAC (%s) is not in format (%s|%s)\n", + value, ETH_TAP_MAC_FIXED, ETH_TAP_USR_MAC_FMT); + return -1; +} + +/* + * Open a TUN interface device. TUN PMD + * 1) sets tap_type as false + * 2) intakes iface as argument. + * 3) as interface is virtual set speed to 10G + */ +static int +rte_pmd_tun_probe(struct rte_vdev_device *dev) +{ + const char *name, *params; + int ret; + struct rte_kvargs *kvlist = NULL; + char tun_name[RTE_ETH_NAME_MAX_LEN]; + char remote_iface[RTE_ETH_NAME_MAX_LEN]; + + tap_type = 0; + strcpy(tuntap_name, "TUN"); + + name = rte_vdev_device_name(dev); + params = rte_vdev_device_args(dev); + memset(remote_iface, 0, RTE_ETH_NAME_MAX_LEN); + + if (params && (params[0] != '\0')) { + RTE_LOG(DEBUG, PMD, "parameters (%s)\n", params); + + kvlist = rte_kvargs_parse(params, valid_arguments); + if (kvlist) { + if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) { + ret = rte_kvargs_process(kvlist, + ETH_TAP_IFACE_ARG, + &set_interface_name, + tun_name); + + if (ret == -1) + goto leave; + } + } + } + pmd_link.link_speed = ETH_SPEED_NUM_10G; + + RTE_LOG(NOTICE, PMD, "Initializing pmd_tun for %s as %s\n", + name, tun_name); + + ret = eth_dev_tap_create(dev, tun_name, remote_iface, 0); + +leave: + if (ret == -1) { + RTE_LOG(ERR, PMD, "Failed to create pmd for %s as %s\n", + name, tun_name); + tun_unit--; /* Restore the unit number */ + } + rte_kvargs_free(kvlist); + + return ret; } /* Open a TAP interface device. */ static int -rte_pmd_tap_probe(const char *name, const char *params) +rte_pmd_tap_probe(struct rte_vdev_device *dev) { + const char *name, *params; int ret; struct rte_kvargs *kvlist = NULL; int speed; char tap_name[RTE_ETH_NAME_MAX_LEN]; char remote_iface[RTE_ETH_NAME_MAX_LEN]; + struct ether_addr user_mac = { .addr_bytes = {0} }; + struct rte_eth_dev *eth_dev; + + tap_type = 1; + strcpy(tuntap_name, "TAP"); + + name = rte_vdev_device_name(dev); + params = rte_vdev_device_args(dev); + + if (rte_eal_process_type() == RTE_PROC_SECONDARY && + strlen(params) == 0) { + eth_dev = rte_eth_dev_attach_secondary(name); + if (!eth_dev) { + RTE_LOG(ERR, PMD, "Failed to probe %s\n", name); + return -1; + } + /* TODO: request info from primary to set up Rx and Tx */ + eth_dev->dev_ops = &ops; + return 0; + } speed = ETH_SPEED_NUM_10G; snprintf(tap_name, sizeof(tap_name), "%s%d", @@ -1288,19 +1749,10 @@ rte_pmd_tap_probe(const char *name, const char *params) memset(remote_iface, 0, RTE_ETH_NAME_MAX_LEN); if (params && (params[0] != '\0')) { - RTE_LOG(DEBUG, PMD, "paramaters (%s)\n", params); + RTE_LOG(DEBUG, PMD, "parameters (%s)\n", params); kvlist = rte_kvargs_parse(params, valid_arguments); if (kvlist) { - if (rte_kvargs_count(kvlist, ETH_TAP_SPEED_ARG) == 1) { - ret = rte_kvargs_process(kvlist, - ETH_TAP_SPEED_ARG, - &set_interface_speed, - &speed); - if (ret == -1) - goto leave; - } - if (rte_kvargs_count(kvlist, ETH_TAP_IFACE_ARG) == 1) { ret = rte_kvargs_process(kvlist, ETH_TAP_IFACE_ARG, @@ -1318,6 +1770,15 @@ rte_pmd_tap_probe(const char *name, const char *params) if (ret == -1) goto leave; } + + if (rte_kvargs_count(kvlist, ETH_TAP_MAC_ARG) == 1) { + ret = rte_kvargs_process(kvlist, + ETH_TAP_MAC_ARG, + &set_mac_type, + &user_mac); + if (ret == -1) + goto leave; + } } } pmd_link.link_speed = speed; @@ -1325,7 +1786,7 @@ rte_pmd_tap_probe(const char *name, const char *params) RTE_LOG(NOTICE, PMD, "Initializing pmd_tap for %s as %s\n", name, tap_name); - ret = eth_dev_tap_create(name, tap_name, remote_iface); + ret = eth_dev_tap_create(dev, tap_name, remote_iface, &user_mac); leave: if (ret == -1) { @@ -1338,10 +1799,10 @@ leave: return ret; } -/* detach a TAP device. +/* detach a TUNTAP device. */ static int -rte_pmd_tap_remove(const char *name) +rte_pmd_tap_remove(struct rte_vdev_device *dev) { struct rte_eth_dev *eth_dev = NULL; struct pmd_internals *internals; @@ -1351,33 +1812,50 @@ rte_pmd_tap_remove(const char *name) rte_socket_id()); /* find the ethdev entry */ - eth_dev = rte_eth_dev_allocated(name); + eth_dev = rte_eth_dev_allocated(rte_vdev_device_name(dev)); if (!eth_dev) return 0; internals = eth_dev->data->dev_private; - if (internals->flower_support && internals->nlsk_fd) { + if (internals->nlsk_fd) { tap_flow_flush(eth_dev, NULL); tap_flow_implicit_flush(internals, NULL); - nl_final(internals->nlsk_fd); + tap_nl_final(internals->nlsk_fd); } - for (i = 0; i < internals->nb_queues; i++) - if (internals->rxq[i].fd != -1) + for (i = 0; i < RTE_PMD_TAP_MAX_QUEUES; i++) { + if (internals->rxq[i].fd != -1) { close(internals->rxq[i].fd); + internals->rxq[i].fd = -1; + } + if (internals->txq[i].fd != -1) { + close(internals->txq[i].fd); + internals->txq[i].fd = -1; + } + } close(internals->ioctl_sock); rte_free(eth_dev->data->dev_private); - rte_free(eth_dev->data); rte_eth_dev_release_port(eth_dev); return 0; } +static struct rte_vdev_driver pmd_tun_drv = { + .probe = rte_pmd_tun_probe, + .remove = rte_pmd_tap_remove, +}; + static struct rte_vdev_driver pmd_tap_drv = { .probe = rte_pmd_tap_probe, .remove = rte_pmd_tap_remove, }; RTE_PMD_REGISTER_VDEV(net_tap, pmd_tap_drv); +RTE_PMD_REGISTER_VDEV(net_tun, pmd_tun_drv); RTE_PMD_REGISTER_ALIAS(net_tap, eth_tap); -RTE_PMD_REGISTER_PARAM_STRING(net_tap, "iface=,speed=N"); +RTE_PMD_REGISTER_PARAM_STRING(net_tun, + ETH_TAP_IFACE_ARG "= "); +RTE_PMD_REGISTER_PARAM_STRING(net_tap, + ETH_TAP_IFACE_ARG "= " + ETH_TAP_MAC_ARG "=" ETH_TAP_MAC_ARG_FMT " " + ETH_TAP_REMOTE_ARG "=");