dpif-netdev: Reorder elements in dp_netdev_port structure.

[mirror_ovs.git] / lib / route-table-bsd.c

/*
 * Copyright (c) 2012 Ed Maste. All rights reserved.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at:
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

#include <config.h>

#include "route-table.h"

#include <sys/socket.h>
#include <sys/types.h>

#include <net/if.h>
#include <net/route.h>
#include <net/if_dl.h>
#include <netinet/in.h>

#include <errno.h>
#include <poll.h>
#include <string.h>
#include <unistd.h>

#include "ovs-router.h"
#include "packets.h"
#include "openvswitch/vlog.h"
#include "util.h"

VLOG_DEFINE_THIS_MODULE(route_table_bsd);

/* OS X does not define RT_ROUNDUP() or equivalent macro. */
#if defined(__MACH__)
#define RT_ROUNDUP(l) ((l) > 0 ? ROUND_UP((l), sizeof(long)) : sizeof(long))
#endif

bool
route_table_fallback_lookup(const struct in6_addr *ip6_dst, char name[],
                            struct in6_addr *gw6)
{
    ovs_be32 ip;
    struct {
        struct rt_msghdr rtm;
        char space[512];
    } rtmsg;

    struct rt_msghdr *rtm = &rtmsg.rtm;
    struct sockaddr_dl *ifp = NULL;
    struct sockaddr_in *sin;
    struct sockaddr *sa;
    static int seq;
    int i, namelen, rtsock;
    ssize_t len;
    const pid_t pid = getpid();
    bool got_ifp = false;
    unsigned int retry_count = 5;  /* arbitrary */

    if (!IN6_IS_ADDR_V4MAPPED(ip6_dst)) {
        return false;
    }
    ip = in6_addr_get_mapped_ipv4(ip6_dst);

    VLOG_DBG("looking route up for " IP_FMT " pid %" PRIuMAX,
        IP_ARGS(ip), (uintmax_t)pid);

    rtsock = socket(PF_ROUTE, SOCK_RAW, 0);
    if (rtsock < 0)
        return false;

retry:
    memset(&rtmsg, 0, sizeof(rtmsg));
    rtm->rtm_msglen = sizeof(struct rt_msghdr) + sizeof(struct sockaddr_in);
    rtm->rtm_version = RTM_VERSION;
    rtm->rtm_type = RTM_GET;
    rtm->rtm_addrs = RTA_DST | RTA_IFP;
    rtm->rtm_seq = ++seq;

    sin = (struct sockaddr_in *)(rtm + 1);
    sin->sin_len = len = sizeof(struct sockaddr_in);
    sin->sin_family = AF_INET;
    sin->sin_addr.s_addr = ip;

    len = write(rtsock, (char *)&rtmsg, rtm->rtm_msglen);
    if (len == -1) {
        if (errno == ENOBUFS && retry_count-- > 0) {
            VLOG_INFO("Recoverable error writing to routing socket: %s",
                      ovs_strerror(errno));
            usleep(500 * 1000);  /* arbitrary */
            goto retry;
        }
        VLOG_ERR("Error writing to routing socket: %s", ovs_strerror(errno));
        close(rtsock);
        return false;
    }
    if (len != rtm->rtm_msglen) {
        VLOG_ERR("Short write to routing socket");
        close(rtsock);
        return false;
    }

    do {
        struct pollfd pfd;
        int ret;

        memset(&pfd, 0, sizeof(pfd));
        pfd.fd = rtsock;
        pfd.events = POLLIN;
        /*
         * The timeout value below is somehow arbitrary.
         * It's to detect the lost of routing messages due to
         * buffer exhaustion etc.  The routing socket is not
         * reliable.
         */
        ret = poll(&pfd, 1, 500);
        if (ret == -1) {
            VLOG_ERR("Error polling on routing socket: %s",
                     ovs_strerror(errno));
            close(rtsock);
            return false;
        }
        if (ret == 0) {
            if (retry_count-- > 0) {
                VLOG_INFO("Timeout; resending routing message");
                goto retry;
            }
            close(rtsock);
            return false;
        }
        len = read(rtsock, (char *)&rtmsg, sizeof(rtmsg));
        if (len > 0) {
            VLOG_DBG("got rtmsg pid %" PRIuMAX " seq %d",
                (uintmax_t)rtmsg.rtm.rtm_pid,
                rtmsg.rtm.rtm_seq);
        }
    } while (len > 0 && (rtmsg.rtm.rtm_seq != seq ||
        rtmsg.rtm.rtm_pid != pid));
    close(rtsock);
    if (len == -1) {
        VLOG_ERR("Error reading from routing socket: %s", ovs_strerror(errno));
        return false;
    }

    *gw6 = in6addr_any;
    sa = (struct sockaddr *)(rtm + 1);
    for (i = 1; i; i <<= 1) {
        if (rtm->rtm_addrs & i) {
            if (i == RTA_IFP && sa->sa_family == AF_LINK &&
              ALIGNED_CAST(struct sockaddr_dl *, sa)->sdl_nlen) {
                ifp = ALIGNED_CAST(struct sockaddr_dl *, sa);
                namelen = ifp->sdl_nlen;
                if (namelen > IFNAMSIZ - 1)
                    namelen = IFNAMSIZ - 1;
                memcpy(name, ifp->sdl_data, namelen);
                name[namelen] = '\0';
                VLOG_DBG("got ifp %s", name);
                got_ifp = true;
            } else if (i == RTA_GATEWAY && sa->sa_family == AF_INET) {
                const struct sockaddr_in *sin_dst =
                    ALIGNED_CAST(struct sockaddr_in *, sa);

                in6_addr_set_mapped_ipv4(gw6, sin_dst->sin_addr.s_addr);
                VLOG_DBG("got gateway " IP_FMT, IP_ARGS(sin_dst->sin_addr.s_addr));
            }
#if defined(__FreeBSD__)
            sa = (struct sockaddr *)((char *)sa + SA_SIZE(sa));
#elif defined(__NetBSD__)
            sa = (struct sockaddr *)((char *)sa + RT_ROUNDUP(sa->sa_len));
#elif defined(__MACH__)
            sa = (struct sockaddr *)((char *)sa + RT_ROUNDUP(sa->sa_len));
#else
#error unimplemented
#endif
        }
    }
    return got_ifp;
}

uint64_t
route_table_get_change_seq(void)
{
    return 0;
}

void
route_table_init(void)
{
    ovs_router_init();
}

void
route_table_run(void)
{
}

void
route_table_wait(void)
{
}