diff options
Diffstat (limited to 'sysdep/linux/netlink.c')
| -rw-r--r-- | sysdep/linux/netlink.c | 808 |
1 files changed, 565 insertions, 243 deletions
diff --git a/sysdep/linux/netlink.c b/sysdep/linux/netlink.c index 22313f43..083a8d90 100644 --- a/sysdep/linux/netlink.c +++ b/sysdep/linux/netlink.c @@ -6,6 +6,7 @@ * Can be freely distributed and used under the terms of the GNU GPL. */ +#include <alloca.h> #include <stdio.h> #include <unistd.h> #include <fcntl.h> @@ -19,9 +20,10 @@ #include "nest/route.h" #include "nest/protocol.h" #include "nest/iface.h" -#include "lib/timer.h" -#include "lib/unix.h" -#include "lib/krt.h" +#include "lib/alloca.h" +#include "sysdep/unix/timer.h" +#include "sysdep/unix/unix.h" +#include "sysdep/unix/krt.h" #include "lib/socket.h" #include "lib/string.h" #include "lib/hash.h" @@ -29,6 +31,11 @@ #include <asm/types.h> #include <linux/if.h> +#ifdef HAVE_LWTUNNEL +#include <linux/lwtunnel.h> +#else +#include "sysdep/linux/lwtunnel.h" +#endif #include <linux/netlink.h> #include <linux/rtnetlink.h> @@ -49,13 +56,31 @@ #define RTA_TABLE 15 #endif +#ifndef RTA_VIA +#define RTA_VIA 18 +#endif -#ifdef IPV6 -#define krt_ecmp6(X) 1 -#else -#define krt_ecmp6(X) 0 +#ifndef HAVE_STRUCT_RTVIA +struct rtvia { + __kernel_sa_family_t rtvia_family; + __u8 rtvia_addr[0]; +}; +#endif + +#ifndef RTA_NEWDST +#define RTA_NEWDST 19 +#endif + +#ifndef RTA_ENCAP_TYPE +#define RTA_ENCAP_TYPE 21 +#endif + +#ifndef RTA_ENCAP +#define RTA_ENCAP 22 #endif +#define krt_ecmp6(p) ((p)->af == AF_INET6) + /* * Structure nl_parse_state keeps state of received route processing. Ideally, * we could just independently parse received Netlink messages and immediately @@ -302,29 +327,32 @@ static struct nl_want_attrs ifla_attr_want[BIRD_IFLA_MAX] = { #define BIRD_IFA_MAX (IFA_FLAGS+1) -#ifndef IPV6 static struct nl_want_attrs ifa_attr_want4[BIRD_IFA_MAX] = { [IFA_ADDRESS] = { 1, 1, sizeof(ip4_addr) }, [IFA_LOCAL] = { 1, 1, sizeof(ip4_addr) }, [IFA_BROADCAST] = { 1, 1, sizeof(ip4_addr) }, - [IFA_FLAGS] = { 1, 1, sizeof(u32) }, + [IFA_FLAGS] = { 1, 1, sizeof(u32) }, }; -#else + static struct nl_want_attrs ifa_attr_want6[BIRD_IFA_MAX] = { [IFA_ADDRESS] = { 1, 1, sizeof(ip6_addr) }, [IFA_LOCAL] = { 1, 1, sizeof(ip6_addr) }, [IFA_FLAGS] = { 1, 1, sizeof(u32) }, }; -#endif -#define BIRD_RTA_MAX (RTA_TABLE+1) +#define BIRD_RTA_MAX (RTA_ENCAP+1) -static struct nl_want_attrs mpnh_attr_want4[BIRD_RTA_MAX] = { +static struct nl_want_attrs nexthop_attr_want4[BIRD_RTA_MAX] = { [RTA_GATEWAY] = { 1, 1, sizeof(ip4_addr) }, + [RTA_ENCAP_TYPE]= { 1, 1, sizeof(u16) }, + [RTA_ENCAP] = { 1, 0, 0 }, +}; + +static struct nl_want_attrs encap_mpls_want[BIRD_RTA_MAX] = { + [RTA_DST] = { 1, 0, 0 }, }; -#ifndef IPV6 static struct nl_want_attrs rtm_attr_want4[BIRD_RTA_MAX] = { [RTA_DST] = { 1, 1, sizeof(ip4_addr) }, [RTA_OIF] = { 1, 1, sizeof(u32) }, @@ -335,8 +363,10 @@ static struct nl_want_attrs rtm_attr_want4[BIRD_RTA_MAX] = { [RTA_MULTIPATH] = { 1, 0, 0 }, [RTA_FLOW] = { 1, 1, sizeof(u32) }, [RTA_TABLE] = { 1, 1, sizeof(u32) }, + [RTA_ENCAP_TYPE]= { 1, 1, sizeof(u16) }, + [RTA_ENCAP] = { 1, 0, 0 }, }; -#else + static struct nl_want_attrs rtm_attr_want6[BIRD_RTA_MAX] = { [RTA_DST] = { 1, 1, sizeof(ip6_addr) }, [RTA_IIF] = { 1, 1, sizeof(u32) }, @@ -347,8 +377,21 @@ static struct nl_want_attrs rtm_attr_want6[BIRD_RTA_MAX] = { [RTA_METRICS] = { 1, 0, 0 }, [RTA_FLOW] = { 1, 1, sizeof(u32) }, [RTA_TABLE] = { 1, 1, sizeof(u32) }, + [RTA_ENCAP_TYPE]= { 1, 1, sizeof(u16) }, + [RTA_ENCAP] = { 1, 0, 0 }, +}; + +static struct nl_want_attrs rtm_attr_want_mpls[BIRD_RTA_MAX] = { + [RTA_DST] = { 1, 1, sizeof(u32) }, + [RTA_IIF] = { 1, 1, sizeof(u32) }, + [RTA_OIF] = { 1, 1, sizeof(u32) }, + [RTA_PRIORITY] = { 1, 1, sizeof(u32) }, + [RTA_METRICS] = { 1, 0, 0 }, + [RTA_FLOW] = { 1, 1, sizeof(u32) }, + [RTA_TABLE] = { 1, 1, sizeof(u32) }, + [RTA_VIA] = { 1, 0, 0 }, + [RTA_NEWDST] = { 1, 0, 0 }, }; -#endif static int @@ -364,7 +407,7 @@ nl_parse_attrs(struct rtattr *a, struct nl_want_attrs *want, struct rtattr **k, if (want[a->rta_type].checksize && (RTA_PAYLOAD(a) != want[a->rta_type].size)) { - log(L_ERR "nl_parse_attrs: Malformed message received"); + log(L_ERR "nl_parse_attrs: Malformed attribute received"); return 0; } @@ -380,6 +423,9 @@ nl_parse_attrs(struct rtattr *a, struct nl_want_attrs *want, struct rtattr **k, return 1; } +static inline u16 rta_get_u16(struct rtattr *a) +{ return *(u16 *) RTA_DATA(a); } + static inline u32 rta_get_u32(struct rtattr *a) { return *(u32 *) RTA_DATA(a); } @@ -389,6 +435,32 @@ static inline ip4_addr rta_get_ip4(struct rtattr *a) static inline ip6_addr rta_get_ip6(struct rtattr *a) { return ip6_ntoh(*(ip6_addr *) RTA_DATA(a)); } +static inline ip_addr rta_get_ipa(struct rtattr *a) +{ + if (RTA_PAYLOAD(a) == sizeof(ip4_addr)) + return ipa_from_ip4(rta_get_ip4(a)); + else + return ipa_from_ip6(rta_get_ip6(a)); +} + +static inline ip_addr rta_get_via(struct rtattr *a) +{ + struct rtvia *v = RTA_DATA(a); + switch(v->rtvia_family) { + case AF_INET: return ipa_from_ip4(ip4_ntoh(*(ip4_addr *) v->rtvia_addr)); + case AF_INET6: return ipa_from_ip6(ip6_ntoh(*(ip6_addr *) v->rtvia_addr)); + } + return IPA_NONE; +} + +static u32 rta_mpls_stack[MPLS_MAX_LABEL_STACK]; +static inline int rta_get_mpls(struct rtattr *a, u32 *stack) +{ + if (RTA_PAYLOAD(a) % 4) + log(L_WARN "KRT: Strange length of received MPLS stack: %u", RTA_PAYLOAD(a)); + + return mpls_get(RTA_DATA(a), RTA_PAYLOAD(a) & ~0x3, stack); +} struct rtattr * nl_add_attr(struct nlmsghdr *h, uint bufsize, uint code, const void *data, uint dlen) @@ -410,29 +482,93 @@ nl_add_attr(struct nlmsghdr *h, uint bufsize, uint code, const void *data, uint return a; } +static inline struct rtattr * +nl_open_attr(struct nlmsghdr *h, uint bufsize, uint code) +{ + return nl_add_attr(h, bufsize, code, NULL, 0); +} + static inline void -nl_add_attr_u32(struct nlmsghdr *h, unsigned bufsize, int code, u32 data) +nl_close_attr(struct nlmsghdr *h, struct rtattr *a) +{ + a->rta_len = (void *)h + NLMSG_ALIGN(h->nlmsg_len) - (void *)a; +} + +static inline void +nl_add_attr_u16(struct nlmsghdr *h, uint bufsize, int code, u16 data) +{ + nl_add_attr(h, bufsize, code, &data, 2); +} + +static inline void +nl_add_attr_u32(struct nlmsghdr *h, uint bufsize, int code, u32 data) { nl_add_attr(h, bufsize, code, &data, 4); } static inline void -nl_add_attr_ipa(struct nlmsghdr *h, unsigned bufsize, int code, ip_addr ipa) +nl_add_attr_ip4(struct nlmsghdr *h, uint bufsize, int code, ip4_addr ip4) { - ipa_hton(ipa); - nl_add_attr(h, bufsize, code, &ipa, sizeof(ipa)); + ip4 = ip4_hton(ip4); + nl_add_attr(h, bufsize, code, &ip4, sizeof(ip4)); } -static inline struct rtattr * -nl_open_attr(struct nlmsghdr *h, uint bufsize, uint code) +static inline void +nl_add_attr_ip6(struct nlmsghdr *h, uint bufsize, int code, ip6_addr ip6) { - return nl_add_attr(h, bufsize, code, NULL, 0); + ip6 = ip6_hton(ip6); + nl_add_attr(h, bufsize, code, &ip6, sizeof(ip6)); } static inline void -nl_close_attr(struct nlmsghdr *h, struct rtattr *a) +nl_add_attr_ipa(struct nlmsghdr *h, uint bufsize, int code, ip_addr ipa) { - a->rta_len = (void *)h + NLMSG_ALIGN(h->nlmsg_len) - (void *)a; + if (ipa_is_ip4(ipa)) + nl_add_attr_ip4(h, bufsize, code, ipa_to_ip4(ipa)); + else + nl_add_attr_ip6(h, bufsize, code, ipa_to_ip6(ipa)); +} + +static inline void +nl_add_attr_mpls(struct nlmsghdr *h, uint bufsize, int code, int len, u32 *stack) +{ + char buf[len*4]; + mpls_put(buf, len, stack); + nl_add_attr(h, bufsize, code, buf, len*4); +} + +static inline void +nl_add_attr_mpls_encap(struct nlmsghdr *h, uint bufsize, int len, u32 *stack) +{ + nl_add_attr_u16(h, bufsize, RTA_ENCAP_TYPE, LWTUNNEL_ENCAP_MPLS); + + struct rtattr *nest = nl_open_attr(h, bufsize, RTA_ENCAP); + nl_add_attr_mpls(h, bufsize, RTA_DST, len, stack); + nl_close_attr(h, nest); +} + +static inline void +nl_add_attr_via(struct nlmsghdr *h, uint bufsize, ip_addr ipa) +{ + struct rtattr *nest = nl_open_attr(h, bufsize, RTA_VIA); + struct rtvia *via = RTA_DATA(nest); + + h->nlmsg_len += sizeof(*via); + + if (ipa_is_ip4(ipa)) + { + via->rtvia_family = AF_INET; + put_ip4(via->rtvia_addr, ipa_to_ip4(ipa)); + h->nlmsg_len += sizeof(ip4_addr); + } + else + { + via->rtvia_family = AF_INET6; + put_ip6(via->rtvia_addr, ipa_to_ip6(ipa)); + h->nlmsg_len += sizeof(ip6_addr); + } + + nl_close_attr(h, nest); } static inline struct rtnexthop * @@ -455,8 +591,24 @@ nl_close_nexthop(struct nlmsghdr *h, struct rtnexthop *nh) nh->rtnh_len = (void *)h + NLMSG_ALIGN(h->nlmsg_len) - (void *)nh; } +static inline void +nl_add_nexthop(struct nlmsghdr *h, uint bufsize, struct nexthop *nh, int af) +{ + if (nh->labels > 0) + if (af == AF_MPLS) + nl_add_attr_mpls(h, bufsize, RTA_NEWDST, nh->labels, nh->label); + else + nl_add_attr_mpls_encap(h, bufsize, nh->labels, nh->label); + + if (ipa_nonzero(nh->gw)) + if (af == AF_MPLS) + nl_add_attr_via(h, bufsize, nh->gw); + else + nl_add_attr_ipa(h, bufsize, RTA_GATEWAY, nh->gw); +} + static void -nl_add_multipath(struct nlmsghdr *h, unsigned bufsize, struct mpnh *nh) +nl_add_multipath(struct nlmsghdr *h, uint bufsize, struct nexthop *nh, int af) { struct rtattr *a = nl_open_attr(h, bufsize, RTA_MULTIPATH); @@ -468,7 +620,7 @@ nl_add_multipath(struct nlmsghdr *h, unsigned bufsize, struct mpnh *nh) rtnh->rtnh_hops = nh->weight; rtnh->rtnh_ifindex = nh->iface->index; - nl_add_attr_ipa(h, bufsize, RTA_GATEWAY, nh->gw); + nl_add_nexthop(h, bufsize, nh, af); nl_close_nexthop(h, rtnh); } @@ -476,17 +628,17 @@ nl_add_multipath(struct nlmsghdr *h, unsigned bufsize, struct mpnh *nh) nl_close_attr(h, a); } -static struct mpnh * +static struct nexthop * nl_parse_multipath(struct krt_proto *p, struct rtattr *ra) { /* Temporary buffer for multicast nexthops */ - static struct mpnh *nh_buffer; + static struct nexthop *nh_buffer; static int nh_buf_size; /* in number of structures */ static int nh_buf_used; struct rtattr *a[BIRD_RTA_MAX]; struct rtnexthop *nh = RTA_DATA(ra); - struct mpnh *rv, *first, **last; + struct nexthop *rv, *first, **last; unsigned len = RTA_PAYLOAD(ra); first = NULL; @@ -502,7 +654,7 @@ nl_parse_multipath(struct krt_proto *p, struct rtattr *ra) if (nh_buf_used == nh_buf_size) { nh_buf_size = nh_buf_size ? (nh_buf_size * 2) : 4; - nh_buffer = xrealloc(nh_buffer, nh_buf_size * sizeof(struct mpnh)); + nh_buffer = xrealloc(nh_buffer, nh_buf_size * NEXTHOP_MAX_SIZE); } *last = rv = nh_buffer + nh_buf_used++; rv->next = NULL; @@ -515,19 +667,34 @@ nl_parse_multipath(struct krt_proto *p, struct rtattr *ra) /* Nonexistent RTNH_PAYLOAD ?? */ nl_attr_len = nh->rtnh_len - RTNH_LENGTH(0); - nl_parse_attrs(RTNH_DATA(nh), mpnh_attr_want4, a, sizeof(a)); + nl_parse_attrs(RTNH_DATA(nh), nexthop_attr_want4, a, sizeof(a)); if (a[RTA_GATEWAY]) { - memcpy(&rv->gw, RTA_DATA(a[RTA_GATEWAY]), sizeof(ip_addr)); - ipa_ntoh(rv->gw); + rv->gw = rta_get_ipa(a[RTA_GATEWAY]); - neighbor *ng = neigh_find2(&p->p, &rv->gw, rv->iface, - (nh->rtnh_flags & RTNH_F_ONLINK) ? NEF_ONLINK : 0); - if (!ng || (ng->scope == SCOPE_HOST)) + neighbor *nbr; + nbr = neigh_find2(&p->p, &rv->gw, rv->iface, + (nh->rtnh_flags & RTNH_F_ONLINK) ? NEF_ONLINK : 0); + if (!nbr || (nbr->scope == SCOPE_HOST)) return NULL; } else - return NULL; + rv->gw = IPA_NONE; + + if (a[RTA_ENCAP_TYPE]) + { + if (rta_get_u16(a[RTA_ENCAP_TYPE]) != LWTUNNEL_ENCAP_MPLS) { + log(L_WARN "KRT: Unknown encapsulation method %d in multipath", rta_get_u16(a[RTA_ENCAP_TYPE])); + return NULL; + } + + struct rtattr *enca[BIRD_RTA_MAX]; + nl_attr_len = RTA_PAYLOAD(a[RTA_ENCAP]); + nl_parse_attrs(RTA_DATA(a[RTA_ENCAP]), encap_mpls_want, enca, sizeof(enca)); + rv->labels = rta_get_mpls(enca[RTA_DST], rv->label); + break; + } + len -= NLMSG_ALIGN(nh->rtnh_len); nh = RTNH_NEXT(nh); @@ -658,52 +825,131 @@ nl_parse_link(struct nlmsghdr *h, int scan) } static void -nl_parse_addr(struct nlmsghdr *h, int scan) +nl_parse_addr4(struct ifaddrmsg *i, int scan, int new) { - struct ifaddrmsg *i; struct rtattr *a[BIRD_IFA_MAX]; - int new = h->nlmsg_type == RTM_NEWADDR; - struct ifa ifa; struct iface *ifi; - int scope; u32 ifa_flags; + int scope; - if (!(i = nl_checkin(h, sizeof(*i)))) + if (!nl_parse_attrs(IFA_RTA(i), ifa_attr_want4, a, sizeof(a))) return; - switch (i->ifa_family) + if (!a[IFA_LOCAL]) { -#ifndef IPV6 - case AF_INET: - if (!nl_parse_attrs(IFA_RTA(i), ifa_attr_want4, a, sizeof(a))) - return; - if (!a[IFA_LOCAL]) - { - log(L_ERR "KIF: Malformed message received (missing IFA_LOCAL)"); - return; - } - break; -#else - case AF_INET6: - if (!nl_parse_attrs(IFA_RTA(i), ifa_attr_want6, a, sizeof(a))) - return; - break; -#endif - default: - return; + log(L_ERR "KIF: Malformed message received (missing IFA_LOCAL)"); + return; } - if (!a[IFA_ADDRESS]) { log(L_ERR "KIF: Malformed message received (missing IFA_ADDRESS)"); return; } + ifi = if_find_by_index(i->ifa_index); + if (!ifi) + { + log(L_ERR "KIF: Received address message for unknown interface %d", i->ifa_index); + return; + } + if (a[IFA_FLAGS]) ifa_flags = rta_get_u32(a[IFA_FLAGS]); else ifa_flags = i->ifa_flags; + struct ifa ifa; + bzero(&ifa, sizeof(ifa)); + ifa.iface = ifi; + if (ifa_flags & IFA_F_SECONDARY) + ifa.flags |= IA_SECONDARY; + + ifa.ip = rta_get_ipa(a[IFA_LOCAL]); + + if (i->ifa_prefixlen > IP4_MAX_PREFIX_LENGTH) + { + log(L_ERR "KIF: Invalid prefix length for interface %s: %d", ifi->name, i->ifa_prefixlen); + new = 0; + } + if (i->ifa_prefixlen == IP4_MAX_PREFIX_LENGTH) + { + ifa.brd = rta_get_ipa(a[IFA_ADDRESS]); + net_fill_ip4(&ifa.prefix, rta_get_ip4(a[IFA_ADDRESS]), i->ifa_prefixlen); + + /* It is either a host address or a peer address */ + if (ipa_equal(ifa.ip, ifa.brd)) + ifa.flags |= IA_HOST; + else + { + ifa.flags |= IA_PEER; + ifa.opposite = ifa.brd; + } + } + else + { + net_fill_ip4(&ifa.prefix, ipa_to_ip4(ifa.ip), i->ifa_prefixlen); + net_normalize(&ifa.prefix); + + if (i->ifa_prefixlen == IP4_MAX_PREFIX_LENGTH - 1) + ifa.opposite = ipa_opposite_m1(ifa.ip); + + if (i->ifa_prefixlen == IP4_MAX_PREFIX_LENGTH - 2) + ifa.opposite = ipa_opposite_m2(ifa.ip); + + if ((ifi->flags & IF_BROADCAST) && a[IFA_BROADCAST]) + { + ip4_addr xbrd = rta_get_ip4(a[IFA_BROADCAST]); + ip4_addr ybrd = ip4_or(ipa_to_ip4(ifa.ip), ip4_not(ip4_mkmask(i->ifa_prefixlen))); + + if (ip4_equal(xbrd, net4_prefix(&ifa.prefix)) || ip4_equal(xbrd, ybrd)) + ifa.brd = ipa_from_ip4(xbrd); + else if (ifi->flags & IF_TMP_DOWN) /* Complain only during the first scan */ + { + log(L_ERR "KIF: Invalid broadcast address %I4 for %s", xbrd, ifi->name); + ifa.brd = ipa_from_ip4(ybrd); + } + } + } + + scope = ipa_classify(ifa.ip); + if (scope < 0) + { + log(L_ERR "KIF: Invalid interface address %I for %s", ifa.ip, ifi->name); + return; + } + ifa.scope = scope & IADDR_SCOPE_MASK; + + DBG("KIF: IF%d(%s): %s IPA %I, flg %x, net %N, brd %I, opp %I\n", + ifi->index, ifi->name, + new ? "added" : "removed", + ifa.ip, ifa.flags, ifa.prefix, ifa.brd, ifa.opposite); + + if (new) + ifa_update(&ifa); + else + ifa_delete(&ifa); + + if (!scan) + if_end_partial_update(ifi); +} + +static void +nl_parse_addr6(struct ifaddrmsg *i, int scan, int new) +{ + struct rtattr *a[BIRD_IFA_MAX]; + struct iface *ifi; + u32 ifa_flags; + int scope; + + if (!nl_parse_attrs(IFA_RTA(i), ifa_attr_want6, a, sizeof(a))) + return; + + if (!a[IFA_ADDRESS]) + { + log(L_ERR "KIF: Malformed message received (missing IFA_ADDRESS)"); + return; + } + ifi = if_find_by_index(i->ifa_index); if (!ifi) { @@ -711,65 +957,50 @@ nl_parse_addr(struct nlmsghdr *h, int scan) return; } + if (a[IFA_FLAGS]) + ifa_flags = rta_get_u32(a[IFA_FLAGS]); + else + ifa_flags = i->ifa_flags; + + struct ifa ifa; bzero(&ifa, sizeof(ifa)); ifa.iface = ifi; if (ifa_flags & IFA_F_SECONDARY) ifa.flags |= IA_SECONDARY; -#ifdef IPV6 /* Ignore tentative addresses silently */ if (ifa_flags & IFA_F_TENTATIVE) return; -#endif /* IFA_LOCAL can be unset for IPv6 interfaces */ - memcpy(&ifa.ip, RTA_DATA(a[IFA_LOCAL] ? : a[IFA_ADDRESS]), sizeof(ifa.ip)); - ipa_ntoh(ifa.ip); - ifa.pxlen = i->ifa_prefixlen; - if (i->ifa_prefixlen > BITS_PER_IP_ADDRESS) + ifa.ip = rta_get_ipa(a[IFA_LOCAL] ? : a[IFA_ADDRESS]); + + if (i->ifa_prefixlen > IP6_MAX_PREFIX_LENGTH) { log(L_ERR "KIF: Invalid prefix length for interface %s: %d", ifi->name, i->ifa_prefixlen); new = 0; } - if (i->ifa_prefixlen == BITS_PER_IP_ADDRESS) + if (i->ifa_prefixlen == IP6_MAX_PREFIX_LENGTH) { - ip_addr addr; - memcpy(&addr, RTA_DATA(a[IFA_ADDRESS]), sizeof(addr)); - ipa_ntoh(addr); - ifa.prefix = ifa.brd = addr; + ifa.brd = rta_get_ipa(a[IFA_ADDRESS]); + net_fill_ip6(&ifa.prefix, rta_get_ip6(a[IFA_ADDRESS]), i->ifa_prefixlen); /* It is either a host address or a peer address */ - if (ipa_equal(ifa.ip, addr)) + if (ipa_equal(ifa.ip, ifa.brd)) ifa.flags |= IA_HOST; else { ifa.flags |= IA_PEER; - ifa.opposite = addr; + ifa.opposite = ifa.brd; } } else { - ip_addr netmask = ipa_mkmask(ifa.pxlen); - ifa.prefix = ipa_and(ifa.ip, netmask); - ifa.brd = ipa_or(ifa.ip, ipa_not(netmask)); - if (i->ifa_prefixlen == BITS_PER_IP_ADDRESS - 1) - ifa.opposite = ipa_opposite_m1(ifa.ip); - -#ifndef IPV6 - if (i->ifa_prefixlen == BITS_PER_IP_ADDRESS - 2) - ifa.opposite = ipa_opposite_m2(ifa.ip); + net_fill_ip6(&ifa.prefix, ipa_to_ip6(ifa.ip), i->ifa_prefixlen); + net_normalize(&ifa.prefix); - if ((ifi->flags & IF_BROADCAST) && a[IFA_BROADCAST]) - { - ip_addr xbrd; - memcpy(&xbrd, RTA_DATA(a[IFA_BROADCAST]), sizeof(xbrd)); - ipa_ntoh(xbrd); - if (ipa_equal(xbrd, ifa.prefix) || ipa_equal(xbrd, ifa.brd)) - ifa.brd = xbrd; - else if (ifi->flags & IF_TMP_DOWN) /* Complain only during the first scan */ - log(L_ERR "KIF: Invalid broadcast address %I for %s", xbrd, ifi->name); - } -#endif + if (i->ifa_prefixlen == IP6_MAX_PREFIX_LENGTH - 1) + ifa.opposite = ipa_opposite_m1(ifa.ip); } scope = ipa_classify(ifa.ip); @@ -780,10 +1011,10 @@ nl_parse_addr(struct nlmsghdr *h, int scan) } ifa.scope = scope & IADDR_SCOPE_MASK; - DBG("KIF: IF%d(%s): %s IPA %I, flg %x, net %I/%d, brd %I, opp %I\n", + DBG("KIF: IF%d(%s): %s IPA %I, flg %x, net %N, brd %I, opp %I\n", ifi->index, ifi->name, new ? "added" : "removed", - ifa.ip, ifa.flags, ifa.prefix, ifa.pxlen, ifa.brd, ifa.opposite); + ifa.ip, ifa.flags, ifa.prefix, ifa.brd, ifa.opposite); if (new) ifa_update(&ifa); @@ -794,6 +1025,26 @@ nl_parse_addr(struct nlmsghdr *h, int scan) if_end_partial_update(ifi); } +static void +nl_parse_addr(struct nlmsghdr *h, int scan) +{ + struct ifaddrmsg *i; + + if (!(i = nl_checkin(h, sizeof(*i)))) + return; + + int new = (h->nlmsg_type == RTM_NEWADDR); + + switch (i->ifa_family) + { + case AF_INET: + return nl_parse_addr4(i, scan, new); + + case AF_INET6: + return nl_parse_addr6(i, scan, new); + } +} + void kif_do_scan(struct kif_proto *p UNUSED) { @@ -808,7 +1059,14 @@ kif_do_scan(struct kif_proto *p UNUSED) else log(L_DEBUG "nl_scan_ifaces: Unknown packet received (type=%d)", h->nlmsg_type); - nl_request_dump(BIRD_AF, RTM_GETADDR); + nl_request_dump(AF_INET, RTM_GETADDR); + while (h = nl_get_scan()) + if (h->nlmsg_type == RTM_NEWADDR || h->nlmsg_type == RTM_DELADDR) + nl_parse_addr(h, 1); + else + log(L_DEBUG "nl_scan_ifaces: Unknown packet received (type=%d)", h->nlmsg_type); + + nl_request_dump(AF_INET6, RTM_GETADDR); while (h = nl_get_scan()) if (h->nlmsg_type == RTM_NEWADDR || h->nlmsg_type == RTM_DELADDR) nl_parse_addr(h, 1); @@ -830,10 +1088,10 @@ krt_table_id(struct krt_proto *p) static HASH(struct krt_proto) nl_table_map; -#define RTH_FN(k) u32_hash(k) -#define RTH_EQ(k1,k2) k1 == k2 -#define RTH_KEY(p) krt_table_id(p) -#define RTH_NEXT(p) p->sys.hash_next +#define RTH_KEY(p) p->af, krt_table_id(p) +#define RTH_NEXT(p) p->sys.hash_next +#define RTH_EQ(a1,i1,a2,i2) a1 == a2 && i1 == i2 +#define RTH_FN(a,i) a ^ u32_hash(i) #define RTH_REHASH rth_rehash #define RTH_PARAMS /8, *2, 2, 2, 6, 20 @@ -845,28 +1103,21 @@ krt_capable(rte *e) { rta *a = e->attrs; - if (a->cast != RTC_UNICAST) - return 0; - switch (a->dest) - { - case RTD_ROUTER: - case RTD_DEVICE: - if (a->iface == NULL) - return 0; + { + case RTD_UNICAST: case RTD_BLACKHOLE: case RTD_UNREACHABLE: case RTD_PROHIBIT: - case RTD_MULTIPATH: - break; + return 1; + default: return 0; - } - return 1; + } } static inline int -nh_bufsize(struct mpnh *nh) +nh_bufsize(struct nexthop *nh) { int rv = 0; for (; nh != NULL; nh = nh->next) @@ -875,32 +1126,42 @@ nh_bufsize(struct mpnh *nh) } static int -nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int dest, ip_addr gw, struct iface *iface) +nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int dest, struct nexthop *nh) { eattr *ea; net *net = e->net; rta *a = e->attrs; + int bufsize = 128 + KRT_METRICS_MAX*8 + nh_bufsize(&(a->nh)); u32 priority = 0; struct { struct nlmsghdr h; struct rtmsg r; - char buf[128 + KRT_METRICS_MAX*8 + nh_bufsize(a->nexthops)]; - } r; + char buf[0]; + } *r; - DBG("nl_send_route(%I/%d,op=%x)\n", net->n.prefix, net->n.pxlen, op); + int rsize = sizeof(*r) + bufsize; + r = alloca(rsize); - bzero(&r.h, sizeof(r.h)); - bzero(&r.r, sizeof(r.r)); - r.h.nlmsg_type = op ? RTM_NEWROUTE : RTM_DELROUTE; - r.h.nlmsg_len = NLMSG_LENGTH(sizeof(struct rtmsg)); - r.h.nlmsg_flags = op | NLM_F_REQUEST | NLM_F_ACK; + DBG("nl_send_route(%N,op=%x)\n", net->n.addr, op); - r.r.rtm_family = BIRD_AF; - r.r.rtm_dst_len = net->n.pxlen; - r.r.rtm_protocol = RTPROT_BIRD; - r.r.rtm_scope = RT_SCOPE_NOWHERE; - nl_add_attr_ipa(&r.h, sizeof(r), RTA_DST, net->n.prefix); + bzero(&r->h, sizeof(r->h)); + bzero(&r->r, sizeof(r->r)); + r->h.nlmsg_type = op ? RTM_NEWROUTE : RTM_DELROUTE; + r->h.nlmsg_len = NLMSG_LENGTH(sizeof(struct rtmsg)); + r->h.nlmsg_flags = op | NLM_F_REQUEST | NLM_F_ACK; + + r->r.rtm_family = p->af; + r->r.rtm_dst_len = net_pxlen(net->n.addr); + r->r.rtm_protocol = RTPROT_BIRD; + r->r.rtm_scope = RT_SCOPE_NOWHERE; + if (p->af == AF_MPLS) + { + u32 label = net_mpls(net->n.addr); + nl_add_attr_mpls(&r->h, rsize, RTA_DST, 1, &label); + } + else + nl_add_attr_ipa(&r->h, rsize, RTA_DST, net_prefix(net->n.addr)); /* * Strange behavior for RTM_DELROUTE: @@ -910,9 +1171,9 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d */ if (krt_table_id(p) < 256) - r.r.rtm_table = krt_table_id(p); + r->r.rtm_table = krt_table_id(p); else - nl_add_attr_u32(&r.h, sizeof(r), RTA_TABLE, krt_table_id(p)); + nl_add_attr_u32(&r->h, rsize, RTA_TABLE, krt_table_id(p)); if (a->source == RTS_DUMMY) priority = e->u.krt.metric; @@ -922,7 +1183,7 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d priority = ea->u.data; if (priority) - nl_add_attr_u32(&r.h, sizeof(r), RTA_PRIORITY, priority); + nl_add_attr_u32(&r->h, rsize, RTA_PRIORITY, priority); /* For route delete, we do not specify remaining route attributes */ if (op == NL_OP_DELETE) @@ -930,15 +1191,15 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d /* Default scope is LINK for device routes, UNIVERSE otherwise */ if (ea = ea_find(eattrs, EA_KRT_SCOPE)) - r.r.rtm_scope = ea->u.data; + r->r.rtm_scope = ea->u.data; else - r.r.rtm_scope = (dest == RTD_DEVICE) ? RT_SCOPE_LINK : RT_SCOPE_UNIVERSE; + r->r.rtm_scope = (dest == RTD_UNICAST && ipa_zero(nh->gw)) ? RT_SCOPE_LINK : RT_SCOPE_UNIVERSE; if (ea = ea_find(eattrs, EA_KRT_PREFSRC)) - nl_add_attr_ipa(&r.h, sizeof(r), RTA_PREFSRC, *(ip_addr *)ea->u.ptr->data); + nl_add_attr_ipa(&r->h, rsize, RTA_PREFSRC, *(ip_addr *)ea->u.ptr->data); if (ea = ea_find(eattrs, EA_KRT_REALM)) - nl_add_attr_u32(&r.h, sizeof(r), RTA_FLOW, ea->u.data); + nl_add_attr_u32(&r->h, rsize, RTA_FLOW, ea->u.data); u32 metrics[KRT_METRICS_MAX]; @@ -953,34 +1214,30 @@ nl_send_route(struct krt_proto *p, rte *e, struct ea_list *eattrs, int op, int d } if (metrics[0]) - nl_add_metrics(&r.h, sizeof(r), metrics, KRT_METRICS_MAX); + nl_add_metrics(&r->h, rsize, metrics, KRT_METRICS_MAX); dest: - /* a->iface != NULL checked in krt_capable() for router and device routes */ switch (dest) { - case RTD_ROUTER: - r.r.rtm_type = RTN_UNICAST; - nl_add_attr_u32(&r.h, sizeof(r), RTA_OIF, iface->index); - nl_add_attr_ipa(&r.h, sizeof(r), RTA_GATEWAY, gw); - break; - case RTD_DEVICE: - r.r.rtm_type = RTN_UNICAST; - nl_add_attr_u32(&r.h, sizeof(r), RTA_OIF, iface->index); + case RTD_UNICAST: + r->r.rtm_type = RTN_UNICAST; + if (nh->next && !krt_ecmp6(p)) + nl_add_multipath(&r->h, rsize, nh, p->af); + else + { + nl_add_attr_u32(&r->h, rsize, RTA_OIF, nh->iface->index); + nl_add_nexthop(&r->h, rsize, nh, p->af); + } break; case RTD_BLACKHOLE: - r.r.rtm_type = RTN_BLACKHOLE; + r->r.rtm_type = RTN_BLACKHOLE; break; case RTD_UNREACHABLE: - r.r.rtm_type = RTN_UNREACHABLE; + r->r.rtm_type = RTN_UNREACHABLE; break; case RTD_PROHIBIT: - r.r.rtm_type = RTN_PROHIBIT; - break; - case RTD_MULTIPATH: - r.r.rtm_type = RTN_UNICAST; - nl_add_multipath(&r.h, sizeof(r), a->nexthops); + r->r.rtm_type = RTN_PROHIBIT; break; case RTD_NONE: break; @@ -989,7 +1246,7 @@ dest: } /* Ignore missing for DELETE */ - return nl_exchange(&r.h, (op == NL_OP_DELETE)); + return nl_exchange(&r->h, (op == NL_OP_DELETE)); } static inline int @@ -998,21 +1255,21 @@ nl_add_rte(struct krt_proto *p, rte *e, struct ea_list *eattrs) rta *a = e->attrs; int err = 0; - if (krt_ecmp6(p) && (a->dest == RTD_MULTIPATH)) + if (krt_ecmp6(p) && a->nh.next) { - struct mpnh *nh = a->nexthops; + struct nexthop *nh = &(a->nh); - err = nl_send_route(p, e, eattrs, NL_OP_ADD, RTD_ROUTER, nh->gw, nh->iface); + err = nl_send_route(p, e, eattrs, NL_OP_ADD, RTD_UNICAST, nh); if (err < 0) return err; for (nh = nh->next; nh; nh = nh->next) - err += nl_send_route(p, e, eattrs, NL_OP_APPEND, RTD_ROUTER, nh->gw, nh->iface); + err += nl_send_route(p, e, eattrs, NL_OP_APPEND, RTD_UNICAST, nh); return err; } - return nl_send_route(p, e, eattrs, NL_OP_ADD, a->dest, a->gw, a->iface); + return nl_send_route(p, e, eattrs, NL_OP_ADD, a->dest, &(a->nh)); } static inline int @@ -1022,7 +1279,7 @@ nl_delete_rte(struct krt_proto *p, rte *e, struct ea_list *eattrs) /* For IPv6, we just repeatedly request DELETE until we get error */ do - err = nl_send_route(p, e, eattrs, NL_OP_DELETE, RTD_NONE, IPA_NONE, NULL); + err = nl_send_route(p, e, eattrs, NL_OP_DELETE, RTD_NONE, NULL); while (krt_ecmp6(p) && !err); return err; @@ -1057,10 +1314,10 @@ krt_replace_rte(struct krt_proto *p, net *n, rte *new, rte *old, struct ea_list } -static inline struct mpnh * -nl_alloc_mpnh(struct nl_parse_state *s, ip_addr gw, struct iface *iface, byte weight) +static inline struct nexthop * +nl_alloc_nexthop(struct nl_parse_state *s, ip_addr gw, struct iface *iface, byte weight) { - struct mpnh *nh = lp_alloc(s->pool, sizeof(struct mpnh)); + struct nexthop *nh = lp_alloc(s->pool, sizeof(struct nexthop)); nh->gw = gw; nh->iface = iface; @@ -1137,9 +1394,9 @@ nl_parse_route(struct nl_parse_state *s, struct nlmsghdr *h) struct rtattr *a[BIRD_RTA_MAX]; int new = h->nlmsg_type == RTM_NEWROUTE; - ip_addr dst = IPA_NONE; + net_addr dst; u32 oif = ~0; - u32 table; + u32 table_id; u32 priority = 0; u32 def_scope = RT_SCOPE_UNIVERSE; int src; @@ -1149,47 +1406,61 @@ nl_parse_route(struct nl_parse_state *s, struct nlmsghdr *h) switch (i->rtm_family) { -#ifndef IPV6 - case AF_INET: - if (!nl_parse_attrs(RTM_RTA(i), rtm_attr_want4, a, sizeof(a))) - return; - break; -#else - case AF_INET6: - if (!nl_parse_attrs(RTM_RTA(i), rtm_attr_want6, a, sizeof(a))) - return; - break; -#endif - default: + case AF_INET: + if (!nl_parse_attrs(RTM_RTA(i), rtm_attr_want4, a, sizeof(a))) return; - } - if (a[RTA_DST]) - { - memcpy(&dst, RTA_DATA(a[RTA_DST]), sizeof(dst)); - ipa_ntoh(dst); + if (a[RTA_DST]) + net_fill_ip4(&dst, rta_get_ip4(a[RTA_DST]), i->rtm_dst_len); + else + net_fill_ip4(&dst, IP4_NONE, 0); + break; + + case AF_INET6: + if (!nl_parse_attrs(RTM_RTA(i), rtm_attr_want6, a, sizeof(a))) + return; + + if (a[RTA_DST]) + net_fill_ip6(&dst, rta_get_ip6(a[RTA_DST]), i->rtm_dst_len); + else + net_fill_ip6(&dst, IP6_NONE, 0); + break; + + case AF_MPLS: + if (!nl_parse_attrs(RTM_RTA(i), rtm_attr_want_mpls, a, sizeof(a))) + return; + + if (!a[RTA_DST]) + SKIP("MPLS route without RTA_DST"); + + if (rta_get_mpls(a[RTA_DST], rta_mpls_stack) != 1) + SKIP("MPLS route with multi-label RTA_DST"); + + net_fill_mpls(&dst, rta_mpls_stack[0]); + break; + + default: + return; } if (a[RTA_OIF]) oif = rta_get_u32(a[RTA_OIF]); if (a[RTA_TABLE]) - table = rta_get_u32(a[RTA_TABLE]); + table_id = rta_get_u32(a[RTA_TABLE]); else - table = i->rtm_table; + table_id = i->rtm_table; - p = HASH_FIND(nl_table_map, RTH, table); /* Do we know this table? */ - DBG("KRT: Got %I/%d, type=%d, oif=%d, table=%d, prid=%d, proto=%s\n", dst, i->rtm_dst_len, i->rtm_type, oif, table, i->rtm_protocol, p ? p->p.name : "(none)"); + /* Do we know this table? */ + p = HASH_FIND(nl_table_map, RTH, i->rtm_family, table_id); if (!p) SKIP("unknown table %d\n", table); -#ifdef IPV6 if (a[RTA_IIF]) SKIP("IIF set\n"); -#else + if (i->rtm_tos != 0) /* We don't support TOS */ SKIP("TOS %02x\n", i->rtm_tos); -#endif if (s->scan && !new) SKIP("RTM_DELROUTE in scan\n"); @@ -1197,7 +1468,7 @@ nl_parse_route(struct nl_parse_state *s, struct nlmsghdr *h) if (a[RTA_PRIORITY]) priority = rta_get_u32(a[RTA_PRIORITY]); - int c = ipa_classify_net(dst); + int c = net_classify(&dst); if ((c < 0) || !(c & IADDR_HOST) || ((c & IADDR_SCOPE_MASK) <= SCOPE_LINK)) SKIP("strange class/scope\n"); @@ -1225,70 +1496,63 @@ nl_parse_route(struct nl_parse_state *s, struct nlmsghdr *h) src = KRT_SRC_ALIEN; } - net *net = net_get(p->p.table, dst, i->rtm_dst_len); + net *net = net_get(p->p.main_channel->table, &dst); if (s->net && !nl_mergable_route(s, net, p, priority, i->rtm_type)) nl_announce_route(s); - rta *ra = lp_allocz(s->pool, sizeof(rta)); + rta *ra = lp_allocz(s->pool, RTA_MAX_SIZE); ra->src = p->p.main_source; ra->source = RTS_INHERIT; ra->scope = SCOPE_UNIVERSE; - ra->cast = RTC_UNICAST; switch (i->rtm_type) { case RTN_UNICAST: + ra->dest = RTD_UNICAST; if (a[RTA_MULTIPATH] && (i->rtm_family == AF_INET)) { - ra->dest = RTD_MULTIPATH; - ra->nexthops = nl_parse_multipath(p, a[RTA_MULTIPATH]); - if (!ra->nexthops) + struct nexthop *nh = nl_parse_multipath(p, a[RTA_MULTIPATH]); + if (!nh) { - log(L_ERR "KRT: Received strange multipath route %I/%d", - net->n.prefix, net->n.pxlen); + log(L_ERR "KRT: Received strange multipath route %N", net->n.addr); return; } + ra->nh = *nh; break; } - ra->iface = if_find_by_index(oif); - if (!ra->iface) + ra->nh.iface = if_find_by_index(oif); + if (!ra->nh.iface) { - log(L_ERR "KRT: Received route %I/%d with unknown ifindex %u", - net->n.prefix, net->n.pxlen, oif); + log(L_ERR "KRT: Received route %N with unknown ifindex %u", net->n.addr, oif); return; } - if (a[RTA_GATEWAY]) + if ((i->rtm_family != AF_MPLS) && a[RTA_GATEWAY] || (i->rtm_family == AF_MPLS) && a[RTA_VIA]) { - neighbor *ng; - ra->dest = RTD_ROUTER; - memcpy(&ra->gw, RTA_DATA(a[RTA_GATEWAY]), sizeof(ra->gw)); - ipa_ntoh(ra->gw); + if (i->rtm_family == AF_MPLS) + ra->nh.gw = rta_get_via(a[RTA_VIA]); + else + ra->nh.gw = rta_get_ipa(a[RTA_GATEWAY]); -#ifdef IPV6 /* Silently skip strange 6to4 routes */ - if (ipa_in_net(ra->gw, IPA_NONE, 96)) + const net_addr_ip6 sit = NET_ADDR_IP6(IP6_NONE, 96); + if ((i->rtm_family == AF_INET6) && ipa_in_netX(ra->nh.gw, (net_addr *) &sit)) return; -#endif - ng = neigh_find2(&p->p, &ra->gw, ra->iface, - (i->rtm_flags & RTNH_F_ONLINK) ? NEF_ONLINK : 0); - if (!ng || (ng->scope == SCOPE_HOST)) + neighbor *nbr; + nbr = neigh_find2(&p->p, &(ra->nh.gw), ra->nh.iface, + (i->rtm_flags & RTNH_F_ONLINK) ? NEF_ONLINK : 0); + if (!nbr || (nbr->scope == SCOPE_HOST)) { - log(L_ERR "KRT: Received route %I/%d with strange next-hop %I", - net->n.prefix, net->n.pxlen, ra->gw); + log(L_ERR "KRT: Received route %N with strange next-hop %I", net->n.addr, + ra->nh.gw); return; } } - else - { - ra->dest = RTD_DEVICE; - def_scope = RT_SCOPE_LINK; - } break; case RTN_BLACKHOLE: @@ -1306,6 +1570,44 @@ nl_parse_route(struct nl_parse_state *s, struct nlmsghdr *h) return; } + int labels = 0; + if ((i->rtm_family == AF_MPLS) && a[RTA_NEWDST] && !ra->nh.next) + labels = rta_get_mpls(a[RTA_NEWDST], ra->nh.label); + + if (a[RTA_ENCAP] && a[RTA_ENCAP_TYPE] && !ra->nh.next) + { + switch (rta_get_u16(a[RTA_ENCAP_TYPE])) + { + case LWTUNNEL_ENCAP_MPLS: + { + struct rtattr *enca[BIRD_RTA_MAX]; + nl_attr_len = RTA_PAYLOAD(a[RTA_ENCAP]); + nl_parse_attrs(RTA_DATA(a[RTA_ENCAP]), encap_mpls_want, enca, sizeof(enca)); + labels = rta_get_mpls(enca[RTA_DST], ra->nh.label); + break; + } + default: + SKIP("unknown encapsulation method %d\n", rta_get_u16(a[RTA_ENCAP_TYPE])); + break; + } + } + + if (labels < 0) + { + log(L_WARN "KRT: Too long MPLS stack received, ignoring."); + ra->nh.labels = 0; + } + else + ra->nh.labels = labels; + + rte *e = rte_get_temp(ra); + e->net = net; + e->u.krt.src = src; + e->u.krt.proto = i->rtm_protocol; + e->u.krt.seen = 0; + e->u.krt.best = 0; + e->u.krt.metric = 0; + if (i->rtm_scope != def_scope) { ea_list *ea = lp_alloc(s->pool, sizeof(ea_list) + sizeof(eattr)); @@ -1319,11 +1621,12 @@ nl_parse_route(struct nl_parse_state *s, struct nlmsghdr *h) ea->attrs[0].u.data = i->rtm_scope; } + if (a[RTA_PRIORITY]) + e->u.krt.metric = rta_get_u32(a[RTA_PRIORITY]); + if (a[RTA_PREFSRC]) { - ip_addr ps; - memcpy(&ps, RTA_DATA(a[RTA_PREFSRC]), sizeof(ps)); - ipa_ntoh(ps); + ip_addr ps = rta_get_ipa(a[RTA_PREFSRC]); ea_list *ea = lp_alloc(s->pool, sizeof(ea_list) + sizeof(eattr)); ea->next = ra->eattrs; @@ -1359,8 +1662,7 @@ nl_parse_route(struct nl_parse_state *s, struct nlmsghdr *h) if (nl_parse_metrics(a[RTA_METRICS], metrics, ARRAY_SIZE(metrics)) < 0) { - log(L_ERR "KRT: Received route %I/%d with strange RTA_METRICS attribute", - net->n.prefix, net->n.pxlen); + log(L_ERR "KRT: Received route %N with strange RTA_METRICS attribute", net->n.addr); return; } @@ -1404,15 +1706,20 @@ nl_parse_route(struct nl_parse_state *s, struct nlmsghdr *h) else { /* Merge next hops with the stored route */ - rta *a = s->attrs; + rta *oa = s->attrs; + + struct nexthop *nhs = &oa->nh; + nexthop_insert(&nhs, &ra->nh); - if (a->dest != RTD_MULTIPATH) + /* Perhaps new nexthop is inserted at the first position */ + if (nhs == &ra->nh) { - a->dest = RTD_MULTIPATH; - a->nexthops = nl_alloc_mpnh(s, a->gw, a->iface, 0); - } + /* Swap rtas */ + s->attrs = ra; - mpnh_insert(&a->nexthops, nl_alloc_mpnh(s, ra->gw, ra->iface, 0)); + /* Keep old eattrs */ + ra->eattrs = oa->eattrs; + } } } @@ -1422,15 +1729,31 @@ krt_do_scan(struct krt_proto *p UNUSED) /* CONFIG_ALL_TABLES_AT_ONCE => p is NUL struct nlmsghdr *h; struct nl_parse_state s; - nl_parse_begin(&s, 1, krt_ecmp6(p)); + nl_parse_begin(&s, 1, 0); + nl_request_dump(AF_INET, RTM_GETROUTE); + while (h = nl_get_scan()) + if (h->nlmsg_type == RTM_NEWROUTE || h->nlmsg_type == RTM_DELROUTE) + nl_parse_route(&s, h); + else + log(L_DEBUG "nl_scan_fire: Unknown packet received (type=%d)", h->nlmsg_type); + nl_parse_end(&s); - nl_request_dump(BIRD_AF, RTM_GETROUTE); + nl_parse_begin(&s, 1, 1); + nl_request_dump(AF_INET6, RTM_GETROUTE); while (h = nl_get_scan()) if (h->nlmsg_type == RTM_NEWROUTE || h->nlmsg_type == RTM_DELROUTE) nl_parse_route(&s, h); else log(L_DEBUG "nl_scan_fire: Unknown packet received (type=%d)", h->nlmsg_type); + nl_parse_end(&s); + nl_parse_begin(&s, 1, 1); + nl_request_dump(AF_MPLS, RTM_GETROUTE); + while (h = nl_get_scan()) + if (h->nlmsg_type == RTM_NEWROUTE || h->nlmsg_type == RTM_DELROUTE) + nl_parse_route(&s, h); + else + log(L_DEBUG "nl_scan_fire: Unknown packet received (type=%d)", h->nlmsg_type); nl_parse_end(&s); } @@ -1553,11 +1876,10 @@ nl_open_async(void) bzero(&sa, sizeof(sa)); sa.nl_family = AF_NETLINK; -#ifdef IPV6 - sa.nl_groups = RTMGRP_LINK | RTMGRP_IPV6_IFADDR | RTMGRP_IPV6_ROUTE; -#else - sa.nl_groups = RTMGRP_LINK | RTMGRP_IPV4_IFADDR | RTMGRP_IPV4_ROUTE; -#endif + sa.nl_groups = RTMGRP_LINK | + RTMGRP_IPV4_IFADDR | RTMGRP_IPV4_ROUTE | + RTMGRP_IPV6_IFADDR | RTMGRP_IPV6_ROUTE; + if (bind(fd, (struct sockaddr *) &sa, sizeof(sa)) < 0) { log(L_ERR "Unable to bind asynchronous rtnetlink socket: %m"); @@ -1591,7 +1913,7 @@ krt_sys_io_init(void) int krt_sys_start(struct krt_proto *p) { - struct krt_proto *old = HASH_FIND(nl_table_map, RTH, krt_table_id(p)); + struct krt_proto *old = HASH_FIND(nl_table_map, RTH, p->af, krt_table_id(p)); if (old) { @@ -1624,7 +1946,7 @@ void krt_sys_init_config(struct krt_config *cf) { cf->sys.table_id = RT_TABLE_MAIN; - cf->sys.metric = 0; + cf->sys.metric = 32; } void |