X-Git-Url: https://git.saurik.com/apple/xnu.git/blobdiff_plain/89b3af67bb32e691275bf6fa803d1834b2284115..b0d623f7f2ae71ed96e60569f61f9a9a27016e80:/bsd/net/route.c diff --git a/bsd/net/route.c b/bsd/net/route.c index 7bb3d89be..66082731e 100644 --- a/bsd/net/route.c +++ b/bsd/net/route.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2000 Apple Computer, Inc. All rights reserved. + * Copyright (c) 2000-2008 Apple Inc. All rights reserved. * * @APPLE_OSREFERENCE_LICENSE_HEADER_START@ * @@ -68,42 +68,503 @@ #include #include #include +#include #include +#include #include #include #include +#include #include +#include #include +#include +#include + +#include + +/* + * Synchronization notes: + * + * Routing entries fall under two locking domains: the global routing table + * lock (rnh_lock) and the per-entry lock (rt_lock); the latter is a mutex that + * resides (statically defined) in the rtentry structure. + * + * The locking domains for routing are defined as follows: + * + * The global routing lock is used to serialize all accesses to the radix + * trees defined by rt_tables[], as well as the tree of masks. This includes + * lookups, insertions and removals of nodes to/from the respective tree. + * It is also used to protect certain fields in the route entry that aren't + * often modified and/or require global serialization (more details below.) + * + * The per-route entry lock is used to serialize accesses to several routing + * entry fields (more details below.) Acquiring and releasing this lock is + * done via RT_LOCK() and RT_UNLOCK() routines. + * + * In cases where both rnh_lock and rt_lock must be held, the former must be + * acquired first in order to maintain lock ordering. It is not a requirement + * that rnh_lock be acquired first before rt_lock, but in case both must be + * acquired in succession, the correct lock ordering must be followed. + * + * The fields of the rtentry structure are protected in the following way: + * + * rt_nodes[] + * + * - Routing table lock (rnh_lock). + * + * rt_parent, rt_mask, rt_llinfo_free + * + * - Set once during creation and never changes; no locks to read. + * + * rt_flags, rt_genmask, rt_llinfo, rt_rmx, rt_refcnt, rt_gwroute + * + * - Routing entry lock (rt_lock) for read/write access. + * + * - Some values of rt_flags are either set once at creation time, + * or aren't currently used, and thus checking against them can + * be done without rt_lock: RTF_GATEWAY, RTF_HOST, RTF_DYNAMIC, + * RTF_DONE, RTF_XRESOLVE, RTF_STATIC, RTF_BLACKHOLE, RTF_ANNOUNCE, + * RTF_USETRAILERS, RTF_WASCLONED, RTF_PINNED, RTF_LOCAL, + * RTF_BROADCAST, RTF_MULTICAST, RTF_IFSCOPE. + * + * rt_key, rt_gateway, rt_ifp, rt_ifa + * + * - Always written/modified with both rnh_lock and rt_lock held. + * + * - May be read freely with rnh_lock held, else must hold rt_lock + * for read access; holding both locks for read is also okay. + * + * - In the event rnh_lock is not acquired, or is not possible to be + * acquired across the operation, setting RTF_CONDEMNED on a route + * entry will prevent its rt_key, rt_gateway, rt_ifp and rt_ifa + * from being modified. This is typically done on a route that + * has been chosen for a removal (from the tree) prior to dropping + * the rt_lock, so that those values will remain the same until + * the route is freed. + * + * When rnh_lock is held rt_setgate(), rt_setif(), and rtsetifa() are + * single-threaded, thus exclusive. This flag will also prevent the + * route from being looked up via rt_lookup(). + * + * generation_id + * + * - Assumes that 32-bit writes are atomic; no locks. + * + * rt_dlt, rt_output + * + * - Currently unused; no locks. + * + * Operations on a route entry can be described as follows: + * + * CREATE an entry with reference count set to 0 as part of RTM_ADD/RESOLVE. + * + * INSERTION of an entry into the radix tree holds the rnh_lock, checks + * for duplicates and then adds the entry. rtrequest returns the entry + * after bumping up the reference count to 1 (for the caller). + * + * LOOKUP of an entry holds the rnh_lock and bumps up the reference count + * before returning; it is valid to also bump up the reference count using + * RT_ADDREF after the lookup has returned an entry. + * + * REMOVAL of an entry from the radix tree holds the rnh_lock, removes the + * entry but does not decrement the reference count. Removal happens when + * the route is explicitly deleted (RTM_DELETE) or when it is in the cached + * state and it expires. The route is said to be "down" when it is no + * longer present in the tree. Freeing the entry will happen on the last + * reference release of such a "down" route. + * + * RT_ADDREF/RT_REMREF operates on the routing entry which increments/ + * decrements the reference count, rt_refcnt, atomically on the rtentry. + * rt_refcnt is modified only using this routine. The general rule is to + * do RT_ADDREF in the function that is passing the entry as an argument, + * in order to prevent the entry from being freed by the callee. + */ + +#define equal(a1, a2) (bcmp((caddr_t)(a1), (caddr_t)(a2), (a1)->sa_len) == 0) #define SA(p) ((struct sockaddr *)(p)) +extern void kdp_set_gateway_mac (void *gatewaymac); + extern struct domain routedomain; struct route_cb route_cb; __private_extern__ struct rtstat rtstat = { 0, 0, 0, 0, 0 }; struct radix_node_head *rt_tables[AF_MAX+1]; -lck_mtx_t *rt_mtx; /*### global routing tables mutex for now */ -lck_attr_t *rt_mtx_attr; -lck_grp_t *rt_mtx_grp; -lck_grp_attr_t *rt_mtx_grp_attr; +lck_mtx_t *rnh_lock; /* global routing tables mutex */ +static lck_attr_t *rnh_lock_attr; +static lck_grp_t *rnh_lock_grp; +static lck_grp_attr_t *rnh_lock_grp_attr; + +/* Lock group and attribute for routing entry locks */ +static lck_attr_t *rte_mtx_attr; +static lck_grp_t *rte_mtx_grp; +static lck_grp_attr_t *rte_mtx_grp_attr; lck_mtx_t *route_domain_mtx; /*### global routing tables mutex for now */ -__private_extern__ int rttrash = 0; /* routes not in table but not freed */ +int rttrash = 0; /* routes not in table but not freed */ + +unsigned int rte_debug; + +/* Possible flags for rte_debug */ +#define RTD_DEBUG 0x1 /* enable or disable rtentry debug facility */ +#define RTD_TRACE 0x2 /* trace alloc, free, refcnt and lock */ +#define RTD_NO_FREE 0x4 /* don't free (good to catch corruptions) */ + +#define RTE_NAME "rtentry" /* name for zone and rt_lock */ + +static struct zone *rte_zone; /* special zone for rtentry */ +#define RTE_ZONE_MAX 65536 /* maximum elements in zone */ +#define RTE_ZONE_NAME RTE_NAME /* name of rtentry zone */ + +#define RTD_INUSE 0xFEEDFACE /* entry is in use */ +#define RTD_FREED 0xDEADBEEF /* entry is freed */ + +/* For gdb */ +__private_extern__ unsigned int ctrace_stack_size = CTRACE_STACK_SIZE; +__private_extern__ unsigned int ctrace_hist_size = CTRACE_HIST_SIZE; + +/* + * Debug variant of rtentry structure. + */ +struct rtentry_dbg { + struct rtentry rtd_entry; /* rtentry */ + struct rtentry rtd_entry_saved; /* saved rtentry */ + uint32_t rtd_inuse; /* in use pattern */ + uint16_t rtd_refhold_cnt; /* # of rtref */ + uint16_t rtd_refrele_cnt; /* # of rtunref */ + uint32_t rtd_lock_cnt; /* # of locks */ + uint32_t rtd_unlock_cnt; /* # of unlocks */ + /* + * Alloc and free callers. + */ + ctrace_t rtd_alloc; + ctrace_t rtd_free; + /* + * Circular lists of rtref and rtunref callers. + */ + ctrace_t rtd_refhold[CTRACE_HIST_SIZE]; + ctrace_t rtd_refrele[CTRACE_HIST_SIZE]; + /* + * Circular lists of locks and unlocks. + */ + ctrace_t rtd_lock[CTRACE_HIST_SIZE]; + ctrace_t rtd_unlock[CTRACE_HIST_SIZE]; + /* + * Trash list linkage + */ + TAILQ_ENTRY(rtentry_dbg) rtd_trash_link; +}; +#define atomic_add_16_ov(a, n) \ + ((uint16_t) OSAddAtomic16(n, (volatile SInt16 *)a)) +#define atomic_add_32_ov(a, n) \ + ((uint32_t) OSAddAtomic(n, a)) + +/* List of trash route entries protected by rnh_lock */ +static TAILQ_HEAD(, rtentry_dbg) rttrash_head; + +static void rte_lock_init(struct rtentry *); +static void rte_lock_destroy(struct rtentry *); +static inline struct rtentry *rte_alloc_debug(void); +static inline void rte_free_debug(struct rtentry *); +static inline void rte_lock_debug(struct rtentry_dbg *); +static inline void rte_unlock_debug(struct rtentry_dbg *); static void rt_maskedcopy(struct sockaddr *, struct sockaddr *, struct sockaddr *); static void rtable_init(void **); +static inline void rtref_audit(struct rtentry_dbg *); +static inline void rtunref_audit(struct rtentry_dbg *); +static struct rtentry *rtalloc1_common_locked(struct sockaddr *, int, uint32_t, + unsigned int); +static int rtrequest_common_locked(int, struct sockaddr *, + struct sockaddr *, struct sockaddr *, int, struct rtentry **, + unsigned int); +static void rtalloc_ign_common_locked(struct route *, uint32_t, unsigned int); +static inline void sa_set_ifscope(struct sockaddr *, unsigned int); +static struct sockaddr *sin_copy(struct sockaddr_in *, struct sockaddr_in *, + unsigned int); +static struct sockaddr *mask_copy(struct sockaddr *, struct sockaddr_in *, + unsigned int); +static struct sockaddr *sa_trim(struct sockaddr *, int); +static struct radix_node *node_lookup(struct sockaddr *, struct sockaddr *, + unsigned int); +static struct radix_node *node_lookup_default(void); +static int rn_match_ifscope(struct radix_node *, void *); +static struct ifaddr *ifa_ifwithroute_common_locked(int, + const struct sockaddr *, const struct sockaddr *, unsigned int); +static struct rtentry *rte_alloc(void); +static void rte_free(struct rtentry *); +static void rtfree_common(struct rtentry *, boolean_t); + +uint32_t route_generation = 0; + +/* + * sockaddr_in with embedded interface scope; this is used internally + * to keep track of scoped route entries in the routing table. The + * fact that such a scope is embedded in the structure is an artifact + * of the current implementation which could change in future. + */ +struct sockaddr_inifscope { + __uint8_t sin_len; + sa_family_t sin_family; + in_port_t sin_port; + struct in_addr sin_addr; + /* + * To avoid possible conflict with an overlaid sockaddr_inarp + * having sin_other set to SIN_PROXY, we use the first 4-bytes + * of sin_zero since sin_srcaddr is one of the unused fields + * in sockaddr_inarp. + */ + union { + char sin_zero[8]; + struct { + __uint32_t ifscope; + } _in_index; + } un; +#define sin_ifscope un._in_index.ifscope +}; + +#define SIN(sa) ((struct sockaddr_in *)(size_t)(sa)) +#define SINIFSCOPE(sa) ((struct sockaddr_inifscope *)(size_t)(sa)) + +#define ASSERT_SINIFSCOPE(sa) { \ + if ((sa)->sa_family != AF_INET || \ + (sa)->sa_len < sizeof (struct sockaddr_in)) \ + panic("%s: bad sockaddr_in %p\n", __func__, sa); \ +} + +/* + * Argument to leaf-matching routine; at present it is scoped routing + * specific but can be expanded in future to include other search filters. + */ +struct matchleaf_arg { + unsigned int ifscope; /* interface scope */ +}; + +/* + * For looking up the non-scoped default route (sockaddr instead + * of sockaddr_in for convenience). + */ +static struct sockaddr sin_def = { + sizeof (struct sockaddr_in), AF_INET, { 0, } +}; + +/* + * Interface index (scope) of the primary interface; determined at + * the time when the default, non-scoped route gets added, changed + * or deleted. Protected by rnh_lock. + */ +static unsigned int primary_ifscope = IFSCOPE_NONE; + +#define INET_DEFAULT(dst) \ + ((dst)->sa_family == AF_INET && SIN(dst)->sin_addr.s_addr == 0) + +#define RT(r) ((struct rtentry *)r) +#define RT_HOST(r) (RT(r)->rt_flags & RTF_HOST) + +/* + * Given a route, determine whether or not it is the non-scoped default + * route; dst typically comes from rt_key(rt) but may be coming from + * a separate place when rt is in the process of being created. + */ +boolean_t +rt_inet_default(struct rtentry *rt, struct sockaddr *dst) +{ + return (INET_DEFAULT(dst) && !(rt->rt_flags & RTF_IFSCOPE)); +} + +/* + * Set the ifscope of the primary interface; caller holds rnh_lock. + */ +void +set_primary_ifscope(unsigned int ifscope) +{ + primary_ifscope = ifscope; +} + +/* + * Return the ifscope of the primary interface; caller holds rnh_lock. + */ +unsigned int +get_primary_ifscope(void) +{ + return (primary_ifscope); +} + +/* + * Embed ifscope into a given a sockaddr_in. + */ +static inline void +sa_set_ifscope(struct sockaddr *sa, unsigned int ifscope) +{ + /* Caller must pass in sockaddr_in */ + ASSERT_SINIFSCOPE(sa); + + SINIFSCOPE(sa)->sin_ifscope = ifscope; +} -__private_extern__ u_long route_generation = 0; -extern int use_routegenid; +/* + * Given a sockaddr_in, return the embedded ifscope to the caller. + */ +unsigned int +sa_get_ifscope(struct sockaddr *sa) +{ + /* Caller must pass in sockaddr_in */ + ASSERT_SINIFSCOPE(sa); + + return (SINIFSCOPE(sa)->sin_ifscope); +} + +/* + * Copy a sockaddr_in src to dst and embed ifscope into dst. + */ +static struct sockaddr * +sin_copy(struct sockaddr_in *src, struct sockaddr_in *dst, unsigned int ifscope) +{ + *dst = *src; + sa_set_ifscope(SA(dst), ifscope); + + return (SA(dst)); +} + +/* + * Copy a mask from src to a sockaddr_in dst and embed ifscope into dst. + */ +static struct sockaddr * +mask_copy(struct sockaddr *src, struct sockaddr_in *dst, unsigned int ifscope) +{ + /* We know dst is at least the size of sockaddr{_in} */ + bzero(dst, sizeof (*dst)); + rt_maskedcopy(src, SA(dst), src); + + /* + * The length of the mask sockaddr would need to be adjusted + * to cover the additional sin_ifscope field; when ifscope is + * IFSCOPE_NONE, we'd end up clearing the embedded ifscope on + * the destination mask in addition to extending the length + * of the sockaddr, as a side effect. This is okay, as any + * trailing zeroes would be skipped by rn_addmask prior to + * inserting or looking up the mask in the mask tree. + */ + SINIFSCOPE(dst)->sin_ifscope = ifscope; + SINIFSCOPE(dst)->sin_len = + offsetof(struct sockaddr_inifscope, sin_ifscope) + + sizeof (SINIFSCOPE(dst)->sin_ifscope); + + return (SA(dst)); +} + +/* + * Trim trailing zeroes on a sockaddr and update its length. + */ +static struct sockaddr * +sa_trim(struct sockaddr *sa, int skip) +{ + caddr_t cp, base = (caddr_t)sa + skip; + + if (sa->sa_len <= skip) + return (sa); + + for (cp = base + (sa->sa_len - skip); cp > base && cp[-1] == 0;) + cp--; + + sa->sa_len = (cp - base) + skip; + if (sa->sa_len < skip) { + /* Must not happen, and if so, panic */ + panic("%s: broken logic (sa_len %d < skip %d )", __func__, + sa->sa_len, skip); + /* NOTREACHED */ + } else if (sa->sa_len == skip) { + /* If we end up with all zeroes, then there's no mask */ + sa->sa_len = 0; + } + + return (sa); +} + +/* + * Called by rtm_msg{1,2} routines to "scrub" the embedded interface scope + * away from the socket address structure, so that clients of the routing + * socket will not be confused by the presence of the embedded scope, or the + * side effect of the increased length due to that. The source sockaddr is + * not modified; instead, the scrubbing happens on the destination sockaddr + * storage that is passed in by the caller. + */ +struct sockaddr * +rtm_scrub_ifscope(int idx, struct sockaddr *hint, struct sockaddr *sa, + struct sockaddr_storage *ss) +{ + struct sockaddr *ret = sa; + + switch (idx) { + case RTAX_DST: + /* + * If this is for an AF_INET destination address, call + * sin_copy() with IFSCOPE_NONE as it does what we need. + */ + if (sa->sa_family == AF_INET && + SINIFSCOPE(sa)->sin_ifscope != IFSCOPE_NONE) { + bzero(ss, sizeof (*ss)); + ret = sin_copy(SIN(sa), SIN(ss), IFSCOPE_NONE); + } + break; + + case RTAX_NETMASK: { + /* + * If this is for a mask, we can't tell whether or not + * there is an embedded interface scope, as the span of + * bytes between sa_len and the beginning of the mask + * (offset of sin_addr in the case of AF_INET) may be + * filled with all-ones by rn_addmask(), and hence we + * cannot rely on sa_family. Because of this, we use + * the sa_family of the hint sockaddr (RTAX_{DST,IFA}) + * as indicator as to whether or not the mask is to be + * treated as one for AF_INET. Clearing the embedded + * scope involves setting it to IFSCOPE_NONE followed + * by calling sa_trim() to trim trailing zeroes from + * the storage sockaddr, which reverses what was done + * earlier by mask_copy() on the source sockaddr. + */ + int skip = offsetof(struct sockaddr_in, sin_addr); + if (sa->sa_len > skip && sa->sa_len <= sizeof (*ss) && + hint != NULL && hint->sa_family == AF_INET) { + bzero(ss, sizeof (*ss)); + bcopy(sa, ss, sa->sa_len); + SINIFSCOPE(ss)->sin_ifscope = IFSCOPE_NONE; + ret = sa_trim(SA(ss), skip); + } + break; + } + default: + break; + } + + return (ret); +} + +/* + * Callback leaf-matching routine for rn_matchaddr_args used + * for looking up an exact match for a scoped route entry. + */ +static int +rn_match_ifscope(struct radix_node *rn, void *arg) +{ + struct rtentry *rt = (struct rtentry *)rn; + struct matchleaf_arg *ma = arg; + if (!(rt->rt_flags & RTF_IFSCOPE) || rt_key(rt)->sa_family != AF_INET) + return (0); + + return (SINIFSCOPE(rt_key(rt))->sin_ifscope == ma->ifscope); +} static void -rtable_init(table) - void **table; +rtable_init(void **table) { struct domain *dom; for (dom = domains; dom; dom = dom->dom_next) @@ -113,152 +574,238 @@ rtable_init(table) } void -route_init() +route_init(void) { - rt_mtx_grp_attr = lck_grp_attr_alloc_init(); - - rt_mtx_grp = lck_grp_alloc_init("route", rt_mtx_grp_attr); - - rt_mtx_attr = lck_attr_alloc_init(); - - if ((rt_mtx = lck_mtx_alloc_init(rt_mtx_grp, rt_mtx_attr)) == NULL) { - printf("route_init: can't alloc rt_mtx\n"); + int size; + + PE_parse_boot_argn("rte_debug", &rte_debug, sizeof (rte_debug)); + if (rte_debug != 0) + rte_debug |= RTD_DEBUG; + + rnh_lock_grp_attr = lck_grp_attr_alloc_init(); + rnh_lock_grp = lck_grp_alloc_init("route", rnh_lock_grp_attr); + rnh_lock_attr = lck_attr_alloc_init(); + if ((rnh_lock = lck_mtx_alloc_init(rnh_lock_grp, + rnh_lock_attr)) == NULL) { + printf("route_init: can't alloc rnh_lock\n"); return; } - lck_mtx_lock(rt_mtx); + rte_mtx_grp_attr = lck_grp_attr_alloc_init(); + rte_mtx_grp = lck_grp_alloc_init(RTE_NAME, rte_mtx_grp_attr); + rte_mtx_attr = lck_attr_alloc_init(); + + lck_mtx_lock(rnh_lock); rn_init(); /* initialize all zeroes, all ones, mask table */ - lck_mtx_unlock(rt_mtx); + lck_mtx_unlock(rnh_lock); rtable_init((void **)rt_tables); route_domain_mtx = routedomain.dom_mtx; + + if (rte_debug & RTD_DEBUG) + size = sizeof (struct rtentry_dbg); + else + size = sizeof (struct rtentry); + + rte_zone = zinit(size, RTE_ZONE_MAX * size, 0, RTE_ZONE_NAME); + if (rte_zone == NULL) + panic("route_init: failed allocating rte_zone"); + + zone_change(rte_zone, Z_EXPAND, TRUE); + + TAILQ_INIT(&rttrash_head); +} + +/* + * Atomically increment route generation counter + */ +void +routegenid_update(void) +{ + (void) atomic_add_32_ov(&route_generation, 1); } /* * Packet routing routines. */ void -rtalloc(ro) - register struct route *ro; +rtalloc(struct route *ro) +{ + rtalloc_ign(ro, 0); +} + +void +rtalloc_ign_locked(struct route *ro, uint32_t ignore) { - rtalloc_ign(ro, 0UL); + return (rtalloc_ign_common_locked(ro, ignore, IFSCOPE_NONE)); } void -rtalloc_ign_locked(ro, ignore) - register struct route *ro; - u_long ignore; +rtalloc_scoped_ign_locked(struct route *ro, uint32_t ignore, + unsigned int ifscope) +{ + return (rtalloc_ign_common_locked(ro, ignore, ifscope)); +} + +static void +rtalloc_ign_common_locked(struct route *ro, uint32_t ignore, + unsigned int ifscope) { struct rtentry *rt; if ((rt = ro->ro_rt) != NULL) { - if (rt->rt_ifp != NULL && rt->rt_flags & RTF_UP) + RT_LOCK_SPIN(rt); + if (rt->rt_ifp != NULL && (rt->rt_flags & RTF_UP) && + rt->generation_id == route_generation) { + RT_UNLOCK(rt); return; - /* XXX - We are probably always at splnet here already. */ + } + RT_UNLOCK(rt); rtfree_locked(rt); ro->ro_rt = NULL; } - ro->ro_rt = rtalloc1_locked(&ro->ro_dst, 1, ignore); - if (ro->ro_rt) + ro->ro_rt = rtalloc1_common_locked(&ro->ro_dst, 1, ignore, ifscope); + if (ro->ro_rt != NULL) { ro->ro_rt->generation_id = route_generation; + RT_LOCK_ASSERT_NOTHELD(ro->ro_rt); + } } + void -rtalloc_ign(ro, ignore) - register struct route *ro; - u_long ignore; +rtalloc_ign(struct route *ro, uint32_t ignore) { - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_NOTOWNED); - lck_mtx_lock(rt_mtx); + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_NOTOWNED); + lck_mtx_lock(rnh_lock); rtalloc_ign_locked(ro, ignore); - lck_mtx_unlock(rt_mtx); + lck_mtx_unlock(rnh_lock); +} + +void +rtalloc_scoped_ign(struct route *ro, uint32_t ignore, unsigned int ifscope) +{ + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_NOTOWNED); + lck_mtx_lock(rnh_lock); + rtalloc_scoped_ign_locked(ro, ignore, ifscope); + lck_mtx_unlock(rnh_lock); +} + +struct rtentry * +rtalloc1_locked(struct sockaddr *dst, int report, uint32_t ignflags) +{ + return (rtalloc1_common_locked(dst, report, ignflags, IFSCOPE_NONE)); +} + +struct rtentry * +rtalloc1_scoped_locked(struct sockaddr *dst, int report, uint32_t ignflags, + unsigned int ifscope) +{ + return (rtalloc1_common_locked(dst, report, ignflags, ifscope)); } /* * Look up the route that matches the address given * Or, at least try.. Create a cloned route if needed. */ -struct rtentry * -rtalloc1_locked(dst, report, ignflags) - const struct sockaddr *dst; - int report; - u_long ignflags; +static struct rtentry * +rtalloc1_common_locked(struct sockaddr *dst, int report, uint32_t ignflags, + unsigned int ifscope) { - register struct radix_node_head *rnh = rt_tables[dst->sa_family]; - register struct rtentry *rt; - register struct radix_node *rn; - struct rtentry *newrt = 0; + struct radix_node_head *rnh = rt_tables[dst->sa_family]; + struct rtentry *rt, *newrt = NULL; struct rt_addrinfo info; - u_long nflags; + uint32_t nflags; int err = 0, msgtype = RTM_MISS; + + if (rnh == NULL) + goto unreachable; + /* - * Look up the address in the table for that Address Family + * Find the longest prefix or exact (in the scoped case) address match; + * callee adds a reference to entry and checks for root node as well */ - if (rnh && (rn = rnh->rnh_matchaddr((caddr_t)dst, rnh)) && - ((rn->rn_flags & RNF_ROOT) == 0)) { + rt = rt_lookup(FALSE, dst, NULL, rnh, ifscope); + if (rt == NULL) + goto unreachable; + + RT_LOCK_SPIN(rt); + newrt = rt; + nflags = rt->rt_flags & ~ignflags; + RT_UNLOCK(rt); + if (report && (nflags & (RTF_CLONING | RTF_PRCLONING))) { /* - * If we find it and it's not the root node, then - * get a refernce on the rtentry associated. + * We are apparently adding (report = 0 in delete). + * If it requires that it be cloned, do so. + * (This implies it wasn't a HOST route.) */ - newrt = rt = (struct rtentry *)rn; - nflags = rt->rt_flags & ~ignflags; - if (report && (nflags & (RTF_CLONING | RTF_PRCLONING))) { + err = rtrequest_locked(RTM_RESOLVE, dst, NULL, NULL, 0, &newrt); + if (err) { /* - * We are apparently adding (report = 0 in delete). - * If it requires that it be cloned, do so. - * (This implies it wasn't a HOST route.) + * If the cloning didn't succeed, maybe what we + * have from lookup above will do. Return that; + * no need to hold another reference since it's + * already done. */ - err = rtrequest_locked(RTM_RESOLVE, dst, SA(0), - SA(0), 0, &newrt); - if (err) { - /* - * If the cloning didn't succeed, maybe - * what we have will do. Return that. - */ - newrt = rt; - rtref(rt); - goto miss; - } - if ((rt = newrt) && (rt->rt_flags & RTF_XRESOLVE)) { - /* - * If the new route specifies it be - * externally resolved, then go do that. - */ - msgtype = RTM_RESOLVE; - goto miss; - } - } else - rtref(rt); - } else { + newrt = rt; + goto miss; + } + /* - * Either we hit the root or couldn't find any match, - * Which basically means - * "caint get there frm here" + * We cloned it; drop the original route found during lookup. + * The resulted cloned route (newrt) would now have an extra + * reference held during rtrequest. */ - rtstat.rts_unreach++; - miss: if (report) { + rtfree_locked(rt); + if ((rt = newrt) && (rt->rt_flags & RTF_XRESOLVE)) { /* - * If required, report the failure to the supervising - * Authorities. - * For a delete, this is not an error. (report == 0) + * If the new route specifies it be + * externally resolved, then go do that. */ - bzero((caddr_t)&info, sizeof(info)); - info.rti_info[RTAX_DST] = dst; - rt_missmsg(msgtype, &info, 0, err); + msgtype = RTM_RESOLVE; + goto miss; } } + goto done; + +unreachable: + /* + * Either we hit the root or couldn't find any match, + * Which basically means "cant get there from here" + */ + rtstat.rts_unreach++; +miss: + if (report) { + /* + * If required, report the failure to the supervising + * Authorities. + * For a delete, this is not an error. (report == 0) + */ + bzero((caddr_t)&info, sizeof(info)); + info.rti_info[RTAX_DST] = dst; + rt_missmsg(msgtype, &info, 0, err); + } +done: return (newrt); } struct rtentry * -rtalloc1(dst, report, ignflags) - register struct sockaddr *dst; - int report; - u_long ignflags; +rtalloc1(struct sockaddr *dst, int report, uint32_t ignflags) { struct rtentry * entry; - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_NOTOWNED); - lck_mtx_lock(rt_mtx); + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_NOTOWNED); + lck_mtx_lock(rnh_lock); entry = rtalloc1_locked(dst, report, ignflags); - lck_mtx_unlock(rt_mtx); + lck_mtx_unlock(rnh_lock); + return (entry); +} + +struct rtentry * +rtalloc1_scoped(struct sockaddr *dst, int report, uint32_t ignflags, + unsigned int ifscope) +{ + struct rtentry * entry; + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_NOTOWNED); + lck_mtx_lock(rnh_lock); + entry = rtalloc1_scoped_locked(dst, report, ignflags, ifscope); + lck_mtx_unlock(rnh_lock); return (entry); } @@ -267,69 +814,133 @@ rtalloc1(dst, report, ignflags) * If the count gets low enough, take it out of the routing table */ void -rtfree_locked(rt) - register struct rtentry *rt; +rtfree_locked(struct rtentry *rt) { - /* - * find the tree for that address family - * Note: in the case of igmp packets, there might not be an rnh - */ - register struct radix_node_head *rnh; + rtfree_common(rt, TRUE); +} - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_OWNED); +static void +rtfree_common(struct rtentry *rt, boolean_t locked) +{ + struct radix_node_head *rnh; - /* See 3582620 - We hit this during the transition from funnels to locks */ - if (rt == 0) { - printf("rtfree - rt is NULL\n"); + /* + * Atomically decrement the reference count and if it reaches 0, + * and there is a close function defined, call the close function. + */ + RT_LOCK_SPIN(rt); + if (rtunref(rt) > 0) { + RT_UNLOCK(rt); return; } - + + /* + * To avoid violating lock ordering, we must drop rt_lock before + * trying to acquire the global rnh_lock. If we are called with + * rnh_lock held, then we already have exclusive access; otherwise + * we do the lock dance. + */ + if (!locked) { + /* + * Note that we check it again below after grabbing rnh_lock, + * since it is possible that another thread doing a lookup wins + * the race, grabs the rnh_lock first, and bumps up the reference + * count in which case the route should be left alone as it is + * still in use. It's also possible that another thread frees + * the route after we drop rt_lock; to prevent the route from + * being freed, we hold an extra reference. + */ + RT_ADDREF_LOCKED(rt); + RT_UNLOCK(rt); + lck_mtx_lock(rnh_lock); + RT_LOCK_SPIN(rt); + RT_REMREF_LOCKED(rt); + if (rt->rt_refcnt > 0) { + /* We've lost the race, so abort */ + RT_UNLOCK(rt); + goto done; + } + } + + /* + * We may be blocked on other lock(s) as part of freeing + * the entry below, so convert from spin to full mutex. + */ + RT_CONVERT_LOCK(rt); + + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_OWNED); + + /* Negative refcnt must never happen */ + if (rt->rt_refcnt != 0) + panic("rt %p invalid refcnt %d", rt, rt->rt_refcnt); + + /* + * find the tree for that address family + * Note: in the case of igmp packets, there might not be an rnh + */ rnh = rt_tables[rt_key(rt)->sa_family]; /* - * decrement the reference count by one and if it reaches 0, - * and there is a close function defined, call the close function + * On last reference give the "close method" a chance to cleanup + * private state. This also permits (for IPv4 and IPv6) a chance + * to decide if the routing table entry should be purged immediately + * or at a later time. When an immediate purge is to happen the + * close routine typically issues RTM_DELETE which clears the RTF_UP + * flag on the entry so that the code below reclaims the storage. */ - rt->rt_refcnt--; - if(rnh && rnh->rnh_close && rt->rt_refcnt == 0) { + if (rnh != NULL && rnh->rnh_close != NULL) rnh->rnh_close((struct radix_node *)rt, rnh); - } /* - * If we are no longer "up" (and ref == 0) - * then we can free the resources associated - * with the route. + * If we are no longer "up" (and ref == 0) then we can free the + * resources associated with the route. */ - if (rt->rt_refcnt <= 0 && (rt->rt_flags & RTF_UP) == 0) { + if (!(rt->rt_flags & RTF_UP)) { if (rt->rt_nodes->rn_flags & (RNF_ACTIVE | RNF_ROOT)) - panic ("rtfree 2"); + panic("rt %p freed while in radix tree\n", rt); /* * the rtentry must have been removed from the routing table - * so it is represented in rttrash.. remove that now. + * so it is represented in rttrash; remove that now. */ - rttrash--; - -#ifdef DIAGNOSTIC - if (rt->rt_refcnt < 0) { - printf("rtfree: %p not freed (neg refs) cnt=%d\n", rt, rt->rt_refcnt); - return; + (void) OSDecrementAtomic(&rttrash); + if (rte_debug & RTD_DEBUG) { + TAILQ_REMOVE(&rttrash_head, (struct rtentry_dbg *)rt, + rtd_trash_link); } -#endif + + /* + * Route is no longer in the tree and refcnt is 0; + * we have exclusive access, so destroy it. + */ + RT_UNLOCK(rt); /* * release references on items we hold them on.. * e.g other routes and ifaddrs. */ - if (rt->rt_parent) + if (rt->rt_parent != NULL) { rtfree_locked(rt->rt_parent); + rt->rt_parent = NULL; + } - if(rt->rt_ifa) { + if (rt->rt_ifa != NULL) { ifafree(rt->rt_ifa); rt->rt_ifa = NULL; } /* - * The key is separatly alloc'd so free it (see rt_setgate()). + * Now free any attached link-layer info. + */ + if (rt->rt_llinfo != NULL) { + if (rt->rt_llinfo_free != NULL) + (*rt->rt_llinfo_free)(rt->rt_llinfo); + else + R_Free(rt->rt_llinfo); + rt->rt_llinfo = NULL; + } + + /* + * The key is separately alloc'd so free it (see rt_setgate()). * This also frees the gateway, as they are always malloc'd * together. */ @@ -338,18 +949,25 @@ rtfree_locked(rt) /* * and the rtentry itself of course */ - R_Free(rt); + rte_lock_destroy(rt); + rte_free(rt); + } else { + /* + * The "close method" has been called, but the route is + * still in the radix tree with zero refcnt, i.e. "up" + * and in the cached state. + */ + RT_UNLOCK(rt); } +done: + if (!locked) + lck_mtx_unlock(rnh_lock); } void -rtfree(rt) - register struct rtentry *rt; +rtfree(struct rtentry *rt) { - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_NOTOWNED); - lck_mtx_lock(rt_mtx); - rtfree_locked(rt); - lck_mtx_unlock(rt_mtx); + rtfree_common(rt, FALSE); } /* @@ -357,39 +975,70 @@ rtfree(rt) * the refcount reaches zero. Unless you have really good reason, * use rtfree not rtunref. */ -void -rtunref(struct rtentry* rt) +int +rtunref(struct rtentry *p) { - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_OWNED); + RT_LOCK_ASSERT_HELD(p); - if (rt == NULL) - panic("rtunref"); - rt->rt_refcnt--; -#if DEBUG - if (rt->rt_refcnt <= 0 && (rt->rt_flags & RTF_UP) == 0) - printf("rtunref - if rtfree were called, we would have freed route\n"); -#endif + if (p->rt_refcnt == 0) + panic("%s(%p) bad refcnt\n", __func__, p); + + --p->rt_refcnt; + + if (rte_debug & RTD_DEBUG) + rtunref_audit((struct rtentry_dbg *)p); + + /* Return new value */ + return (p->rt_refcnt); +} + +static inline void +rtunref_audit(struct rtentry_dbg *rte) +{ + uint16_t idx; + + if (rte->rtd_inuse != RTD_INUSE) + panic("rtunref: on freed rte=%p\n", rte); + + idx = atomic_add_16_ov(&rte->rtd_refrele_cnt, 1) % CTRACE_HIST_SIZE; + if (rte_debug & RTD_TRACE) + ctrace_record(&rte->rtd_refrele[idx]); } /* * Add a reference count from an rtentry. */ void -rtref(struct rtentry* rt) +rtref(struct rtentry *p) { - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_OWNED); + RT_LOCK_ASSERT_HELD(p); - if (rt == NULL) - panic("rtref"); + if (++p->rt_refcnt == 0) + panic("%s(%p) bad refcnt\n", __func__, p); + + if (rte_debug & RTD_DEBUG) + rtref_audit((struct rtentry_dbg *)p); +} + +static inline void +rtref_audit(struct rtentry_dbg *rte) +{ + uint16_t idx; + + if (rte->rtd_inuse != RTD_INUSE) + panic("rtref_audit: on freed rte=%p\n", rte); - rt->rt_refcnt++; + idx = atomic_add_16_ov(&rte->rtd_refhold_cnt, 1) % CTRACE_HIST_SIZE; + if (rte_debug & RTD_TRACE) + ctrace_record(&rte->rtd_refhold[idx]); } void rtsetifa(struct rtentry *rt, struct ifaddr* ifa) { - if (rt == NULL) - panic("rtsetifa"); + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_OWNED); + + RT_LOCK_ASSERT_HELD(rt); if (rt->rt_ifa == ifa) return; @@ -406,78 +1055,58 @@ rtsetifa(struct rtentry *rt, struct ifaddr* ifa) ifaref(rt->rt_ifa); } -void -ifafree(ifa) - register struct ifaddr *ifa; -{ - int i, oldval; - u_char *ptr = (u_char*)ifa; - - if (ifa == NULL) - panic("ifafree"); - - oldval = OSAddAtomic(-1, &ifa->ifa_refcnt); - - if (oldval == 0) { - if ((ifa->ifa_flags & IFA_ATTACHED) != 0) { - panic("ifa attached to ifp is being freed\n"); - } - FREE(ifa, M_IFADDR); - } -} - -void -ifaref(struct ifaddr *ifa) -{ - if (ifa == NULL) - panic("ifaref"); - - if (OSAddAtomic(1, &ifa->ifa_refcnt) == 0xffffffff) - panic("ifaref - reference count rolled over!"); -} - /* * Force a routing table entry to the specified * destination to go through the given gateway. * Normally called as a result of a routing redirect * message from the network layer. - * - * N.B.: must be called at splnet - * */ void -rtredirect(dst, gateway, netmask, flags, src, rtp) - struct sockaddr *dst, *gateway, *netmask, *src; - int flags; - struct rtentry **rtp; +rtredirect(struct ifnet *ifp, struct sockaddr *dst, struct sockaddr *gateway, + struct sockaddr *netmask, int flags, struct sockaddr *src, + struct rtentry **rtp) { - register struct rtentry *rt; + struct rtentry *rt = NULL; int error = 0; short *stat = 0; struct rt_addrinfo info; struct ifaddr *ifa = NULL; + unsigned int ifscope = (ifp != NULL) ? ifp->if_index : IFSCOPE_NONE; + struct sockaddr_in sin; - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_NOTOWNED); - lck_mtx_lock(rt_mtx); + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_NOTOWNED); + lck_mtx_lock(rnh_lock); - /* verify the gateway is directly reachable */ - if ((ifa = ifa_ifwithnet(gateway)) == 0) { + /* + * Verify the gateway is directly reachable; if scoped routing + * is enabled, verify that it is reachable from the interface + * where the ICMP redirect arrived on. + */ + if ((ifa = ifa_ifwithnet_scoped(gateway, ifscope)) == NULL) { error = ENETUNREACH; goto out; } - rt = rtalloc1_locked(dst, 0, 0UL); + /* Lookup route to the destination (from the original IP header) */ + rt = rtalloc1_scoped_locked(dst, 0, RTF_CLONING|RTF_PRCLONING, ifscope); + if (rt != NULL) + RT_LOCK(rt); + + /* Embed scope in src for comparison against rt_gateway below */ + if (ip_doscopedroute && src->sa_family == AF_INET) + src = sin_copy(SIN(src), &sin, ifscope); + /* * If the redirect isn't from our current router for this dst, * it's either old or wrong. If it redirects us to ourselves, * we have a routing loop, perhaps as a result of an interface * going down recently. */ -#define equal(a1, a2) (bcmp((caddr_t)(a1), (caddr_t)(a2), (a1)->sa_len) == 0) - if (!(flags & RTF_DONE) && rt && - (!equal(src, rt->rt_gateway) || rt->rt_ifa != ifa)) + if (!(flags & RTF_DONE) && rt != NULL && + (!equal(src, rt->rt_gateway) || !equal(rt->rt_ifa->ifa_addr, + ifa->ifa_addr))) { error = EINVAL; - else { + } else { ifafree(ifa); if ((ifa = ifa_ifwithaddr(gateway))) { ifafree(ifa); @@ -485,37 +1114,45 @@ rtredirect(dst, gateway, netmask, flags, src, rtp) error = EHOSTUNREACH; } } - + if (ifa) { ifafree(ifa); ifa = NULL; } - - if (error) + + if (error) { + if (rt != NULL) + RT_UNLOCK(rt); goto done; + } + /* * Create a new entry if we just got back a wildcard entry * or the the lookup failed. This is necessary for hosts * which use routing redirects generated by smart gateways * to dynamically build the routing tables. */ - if ((rt == 0) || (rt_mask(rt) && rt_mask(rt)->sa_len < 2)) + if ((rt == NULL) || (rt_mask(rt) != NULL && rt_mask(rt)->sa_len < 2)) goto create; /* * Don't listen to the redirect if it's * for a route to an interface. */ + RT_LOCK_ASSERT_HELD(rt); if (rt->rt_flags & RTF_GATEWAY) { if (((rt->rt_flags & RTF_HOST) == 0) && (flags & RTF_HOST)) { /* * Changing from route to net => route to host. - * Create new route, rather than smashing route to net. + * Create new route, rather than smashing route + * to net; similar to cloned routes, the newly + * created host route is scoped as well. */ - create: +create: + if (rt != NULL) + RT_UNLOCK(rt); flags |= RTF_GATEWAY | RTF_DYNAMIC; - error = rtrequest_locked((int)RTM_ADD, dst, gateway, - netmask, flags, - (struct rtentry **)0); + error = rtrequest_scoped_locked(RTM_ADD, dst, + gateway, netmask, flags, NULL, ifscope); stat = &rtstat.rts_dynamic; } else { /* @@ -528,50 +1165,51 @@ rtredirect(dst, gateway, netmask, flags, src, rtp) /* * add the key and gateway (in one malloc'd chunk). */ - rt_setgate(rt, rt_key(rt), gateway); + error = rt_setgate(rt, rt_key(rt), gateway); + RT_UNLOCK(rt); } - } else + } else { + RT_UNLOCK(rt); error = EHOSTUNREACH; + } done: - if (rt) { + if (rt != NULL) { + RT_LOCK_ASSERT_NOTHELD(rt); if (rtp && !error) *rtp = rt; else rtfree_locked(rt); } out: - if (error) + if (error) { rtstat.rts_badredirect++; - else if (stat != NULL) - (*stat)++; + } else { + if (stat != NULL) + (*stat)++; + if (use_routegenid) + routegenid_update(); + } + lck_mtx_unlock(rnh_lock); bzero((caddr_t)&info, sizeof(info)); info.rti_info[RTAX_DST] = dst; info.rti_info[RTAX_GATEWAY] = gateway; info.rti_info[RTAX_NETMASK] = netmask; info.rti_info[RTAX_AUTHOR] = src; rt_missmsg(RTM_REDIRECT, &info, flags, error); - lck_mtx_unlock(rt_mtx); } /* * Routing table ioctl interface. */ int -rtioctl(req, data, p) - int req; - caddr_t data; - struct proc *p; +rtioctl(unsigned long req, caddr_t data, struct proc *p) { -#if INET - /* Multicast goop, grrr... */ -#if MROUTING +#pragma unused(p) +#if INET && MROUTING return mrt_ioctl(req, data); #else - return mrt_ioctl(req, data, p); -#endif -#else /* INET */ return ENXIO; -#endif /* INET */ +#endif } struct ifaddr * @@ -580,11 +1218,58 @@ ifa_ifwithroute( const struct sockaddr *dst, const struct sockaddr *gateway) { + struct ifaddr *ifa; + + lck_mtx_lock(rnh_lock); + ifa = ifa_ifwithroute_locked(flags, dst, gateway); + lck_mtx_unlock(rnh_lock); + + return (ifa); +} + +struct ifaddr * +ifa_ifwithroute_locked(int flags, const struct sockaddr *dst, + const struct sockaddr *gateway) +{ + return (ifa_ifwithroute_common_locked((flags & ~RTF_IFSCOPE), dst, + gateway, IFSCOPE_NONE)); +} + +struct ifaddr * +ifa_ifwithroute_scoped_locked(int flags, const struct sockaddr *dst, + const struct sockaddr *gateway, unsigned int ifscope) +{ + if (ifscope != IFSCOPE_NONE) + flags |= RTF_IFSCOPE; + else + flags &= ~RTF_IFSCOPE; - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_OWNED); + return (ifa_ifwithroute_common_locked(flags, dst, gateway, ifscope)); +} + +static struct ifaddr * +ifa_ifwithroute_common_locked(int flags, const struct sockaddr *dst, + const struct sockaddr *gateway, unsigned int ifscope) +{ + struct ifaddr *ifa = NULL; + struct rtentry *rt = NULL; + struct sockaddr_in dst_in, gw_in; - struct ifaddr *ifa = 0; - if ((flags & RTF_GATEWAY) == 0) { + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_OWNED); + + if (ip_doscopedroute) { + /* + * Just in case the sockaddr passed in by the caller + * contains embedded scope, make sure to clear it since + * IPv4 interface addresses aren't scoped. + */ + if (dst != NULL && dst->sa_family == AF_INET) + dst = sin_copy(SIN(dst), &dst_in, IFSCOPE_NONE); + if (gateway != NULL && gateway->sa_family == AF_INET) + gateway = sin_copy(SIN(gateway), &gw_in, IFSCOPE_NONE); + } + + if (!(flags & RTF_GATEWAY)) { /* * If we are adding a route to an interface, * and the interface is a pt to pt link @@ -595,8 +1280,8 @@ ifa_ifwithroute( if (flags & RTF_HOST) { ifa = ifa_ifwithdstaddr(dst); } - if (ifa == 0) - ifa = ifa_ifwithaddr(gateway); + if (ifa == NULL) + ifa = ifa_ifwithaddr_scoped(gateway, ifscope); } else { /* * If we are adding a route to a remote net @@ -605,61 +1290,126 @@ ifa_ifwithroute( */ ifa = ifa_ifwithdstaddr(gateway); } - if (ifa == 0) - ifa = ifa_ifwithnet(gateway); - if (ifa == 0) { - struct rtentry *rt = rtalloc1_locked(dst, 0, 0UL); - if (rt == 0) - return (0); - ifa = rt->rt_ifa; - if (ifa) - ifaref(ifa); - rtunref(rt); - if (ifa == 0) - return 0; + if (ifa == NULL) + ifa = ifa_ifwithnet_scoped(gateway, ifscope); + if (ifa == NULL) { + /* Workaround to avoid gcc warning regarding const variable */ + rt = rtalloc1_scoped_locked((struct sockaddr *)(size_t)dst, + 0, 0, ifscope); + if (rt != NULL) { + RT_LOCK_SPIN(rt); + ifa = rt->rt_ifa; + if (ifa != NULL) + ifaref(ifa); + RT_REMREF_LOCKED(rt); + RT_UNLOCK(rt); + rt = NULL; + } } - if (ifa->ifa_addr->sa_family != dst->sa_family) { + if (ifa != NULL && ifa->ifa_addr->sa_family != dst->sa_family) { struct ifaddr *newifa; + /* Callee adds reference to newifa upon success */ newifa = ifaof_ifpforaddr(dst, ifa->ifa_ifp); - if (newifa != 0) { + if (newifa != NULL) { ifafree(ifa); ifa = newifa; } } + /* + * If we are adding a gateway, it is quite possible that the + * routing table has a static entry in place for the gateway, + * that may not agree with info garnered from the interfaces. + * The routing table should carry more precedence than the + * interfaces in this matter. Must be careful not to stomp + * on new entries from rtinit, hence (ifa->ifa_addr != gateway). + */ + if ((ifa == NULL || + !equal(ifa->ifa_addr, (struct sockaddr *)(size_t)gateway)) && + (rt = rtalloc1_scoped_locked((struct sockaddr *)(size_t)gateway, + 0, 0, ifscope)) != NULL) { + if (ifa != NULL) + ifafree(ifa); + RT_LOCK_SPIN(rt); + ifa = rt->rt_ifa; + if (ifa != NULL) + ifaref(ifa); + RT_REMREF_LOCKED(rt); + RT_UNLOCK(rt); + } + /* + * If an interface scope was specified, the interface index of + * the found ifaddr must be equivalent to that of the scope; + * otherwise there is no match. + */ + if ((flags & RTF_IFSCOPE) && + ifa != NULL && ifa->ifa_ifp->if_index != ifscope) { + ifafree(ifa); + ifa = NULL; + } + return (ifa); } -#define ROUNDUP(a) (a>0 ? (1 + (((a) - 1) | (sizeof(long) - 1))) : sizeof(long)) - -static int rt_fixdelete __P((struct radix_node *, void *)); -static int rt_fixchange __P((struct radix_node *, void *)); +static int rt_fixdelete(struct radix_node *, void *); +static int rt_fixchange(struct radix_node *, void *); struct rtfc_arg { struct rtentry *rt0; struct radix_node_head *rnh; }; +int +rtrequest_locked(int req, struct sockaddr *dst, struct sockaddr *gateway, + struct sockaddr *netmask, int flags, struct rtentry **ret_nrt) +{ + return (rtrequest_common_locked(req, dst, gateway, netmask, + (flags & ~RTF_IFSCOPE), ret_nrt, IFSCOPE_NONE)); +} + +int +rtrequest_scoped_locked(int req, struct sockaddr *dst, + struct sockaddr *gateway, struct sockaddr *netmask, int flags, + struct rtentry **ret_nrt, unsigned int ifscope) +{ + if (ifscope != IFSCOPE_NONE) + flags |= RTF_IFSCOPE; + else + flags &= ~RTF_IFSCOPE; + + return (rtrequest_common_locked(req, dst, gateway, netmask, + flags, ret_nrt, ifscope)); +} + /* - * Do appropriate manipulations of a routing tree given - * all the bits of info needed + * Do appropriate manipulations of a routing tree given all the bits of + * info needed. + * + * Embedding the scope in the radix key is an internal job that should be + * left to routines in this module. Callers should specify the scope value + * to the "scoped" variants of route routines instead of manipulating the + * key itself. This is typically done when creating a scoped route, e.g. + * rtrequest(RTM_ADD). Once such a route is created and marked with the + * RTF_IFSCOPE flag, callers can simply use its rt_key(rt) to clone it + * (RTM_RESOLVE) or to remove it (RTM_DELETE). An exception to this is + * during certain routing socket operations where the search key might be + * derived from the routing message itself, in which case the caller must + * specify the destination address and scope value for RTM_ADD/RTM_DELETE. */ -rtrequest_locked( - int req, - struct sockaddr *dst, - struct sockaddr *gateway, - struct sockaddr *netmask, - int flags, - struct rtentry **ret_nrt) +static int +rtrequest_common_locked(int req, struct sockaddr *dst0, + struct sockaddr *gateway, struct sockaddr *netmask, int flags, + struct rtentry **ret_nrt, unsigned int ifscope) { int error = 0; - register struct rtentry *rt; - register struct radix_node *rn; - register struct radix_node_head *rnh; + struct rtentry *rt; + struct radix_node *rn; + struct radix_node_head *rnh; struct ifaddr *ifa = NULL; - struct sockaddr *ndst; + struct sockaddr *ndst, *dst = dst0; + struct sockaddr_in sin, mask; #define senderr(x) { error = x ; goto bad; } - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_OWNED); + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_OWNED); /* * Find the correct routing tree to use for this Address Family */ @@ -671,6 +1421,37 @@ rtrequest_locked( */ if (flags & RTF_HOST) netmask = 0; + + /* + * If RTF_IFSCOPE is specified, use a local copy of the destination + * address to embed the scope into. This logic is repeated below + * in the RTM_RESOLVE handler since the caller does not normally + * specify such a flag during a resolve; instead it passes in the + * route used for cloning for which the scope info is derived from. + * Note also that in the case of RTM_DELETE, the address passed in + * by the caller might already contain the embedded scope info when + * it is the key itself, thus making RTF_IFSCOPE unnecessary; one + * instance where it is explicitly set is inside route_output() + * as part of handling a routing socket request. + */ + if (req != RTM_RESOLVE && (flags & RTF_IFSCOPE)) { + /* Scoped routing is for AF_INET only */ + if (dst->sa_family != AF_INET || + (req == RTM_ADD && !ip_doscopedroute)) + senderr(EINVAL); + + if (ifscope == IFSCOPE_NONE) { + flags &= ~RTF_IFSCOPE; + } else { + /* Embed ifscope into the key (local copy) */ + dst = sin_copy(SIN(dst), &sin, ifscope); + + /* Embed ifscope into netmask (local copy) */ + if (netmask != NULL) + netmask = mask_copy(netmask, &mask, ifscope); + } + } + switch (req) { case RTM_DELETE: /* @@ -683,14 +1464,34 @@ rtrequest_locked( panic ("rtrequest delete"); rt = (struct rtentry *)rn; + /* + * Take an extra reference to handle the deletion of a route + * entry whose reference count is already 0; e.g. an expiring + * cloned route entry or an entry that was added to the table + * with 0 reference. If the caller is interested in this route, + * we will return it with the reference intact. Otherwise we + * will decrement the reference via rtfree_locked() and then + * possibly deallocate it. + */ + RT_LOCK(rt); + RT_ADDREF_LOCKED(rt); + rt->rt_flags &= ~RTF_UP; + + /* + * For consistency, in case the caller didn't set the flag. + */ + rt->rt_flags |= RTF_CONDEMNED; + /* * Now search what's left of the subtree for any cloned * routes which might have been formed from this node. */ if ((rt->rt_flags & (RTF_CLONING | RTF_PRCLONING)) && rt_mask(rt)) { + RT_UNLOCK(rt); rnh->rnh_walktree_from(rnh, dst, rt_mask(rt), rt_fixdelete, rt); + RT_LOCK(rt); } /* @@ -698,21 +1499,11 @@ rtrequest_locked( * This might result in another rtentry being freed if * we held its last reference. */ - if (rt->rt_gwroute) { - rt = rt->rt_gwroute; - rtfree_locked(rt); - (rt = (struct rtentry *)rn)->rt_gwroute = 0; + if (rt->rt_gwroute != NULL) { + rtfree_locked(rt->rt_gwroute); + rt->rt_gwroute = NULL; } - /* - * NB: RTF_UP must be set during the search above, - * because we might delete the last ref, causing - * rt to get freed prematurely. - * eh? then why not just add a reference? - * I'm not sure how RTF_UP helps matters. (JRE) - */ - rt->rt_flags &= ~RTF_UP; - /* * give the protocol a chance to keep things in sync. */ @@ -721,20 +1512,34 @@ rtrequest_locked( ifa = NULL; /* - * one more rtentry floating around that is not - * linked to the routing table. + * one more rtentry floating around that is not + * linked to the routing table. + */ + (void) OSIncrementAtomic(&rttrash); + if (rte_debug & RTD_DEBUG) { + TAILQ_INSERT_TAIL(&rttrash_head, + (struct rtentry_dbg *)rt, rtd_trash_link); + } + + /* + * If this is the (non-scoped) default route, clear + * the interface index used for the primary ifscope. */ - rttrash++; + if (rt_inet_default(rt, rt_key(rt))) + set_primary_ifscope(IFSCOPE_NONE); + + RT_UNLOCK(rt); /* * If the caller wants it, then it can have it, * but it's up to it to free the rtentry as we won't be * doing it. */ - if (ret_nrt) + if (ret_nrt != NULL) { + /* Return the route to caller with reference intact */ *ret_nrt = rt; - else if (rt->rt_refcnt <= 0) { - rt->rt_refcnt++; /* make a 1->0 transition */ + } else { + /* Dereference or deallocate the route */ rtfree_locked(rt); } break; @@ -742,6 +1547,16 @@ rtrequest_locked( case RTM_RESOLVE: if (ret_nrt == 0 || (rt = *ret_nrt) == 0) senderr(EINVAL); + /* + * If cloning, we have the parent route given by the caller + * and will use its rt_gateway, rt_rmx as part of the cloning + * process below. Since rnh_lock is held at this point, the + * parent's rt_ifa and rt_gateway will not change, and its + * relevant rt_flags will not change as well. The only thing + * that could change are the metrics, and thus we hold the + * parent route's rt_lock later on during the actual copying + * of rt_rmx. + */ ifa = rt->rt_ifa; ifaref(ifa); flags = rt->rt_flags & @@ -750,27 +1565,64 @@ rtrequest_locked( gateway = rt->rt_gateway; if ((netmask = rt->rt_genmask) == 0) flags |= RTF_HOST; + + if (!ip_doscopedroute || dst->sa_family != AF_INET) + goto makeroute; + /* + * When scoped routing is enabled, cloned entries are + * always scoped according to the interface portion of + * the parent route. The exception to this are IPv4 + * link local addresses. + */ + if (!IN_LINKLOCAL(ntohl(SIN(dst)->sin_addr.s_addr))) { + if (flags & RTF_IFSCOPE) { + ifscope = sa_get_ifscope(rt_key(rt)); + } else { + ifscope = rt->rt_ifp->if_index; + flags |= RTF_IFSCOPE; + } + } else { + ifscope = IFSCOPE_NONE; + flags &= ~RTF_IFSCOPE; + } + + /* Embed or clear ifscope into/from the key (local copy) */ + dst = sin_copy(SIN(dst), &sin, ifscope); + + /* Embed or clear ifscope into/from netmask (local copy) */ + if (netmask != NULL) + netmask = mask_copy(netmask, &mask, ifscope); + goto makeroute; case RTM_ADD: if ((flags & RTF_GATEWAY) && !gateway) - panic("rtrequest: GATEWAY but no gateway"); + panic("rtrequest: RTF_GATEWAY but no gateway"); - if ((ifa = ifa_ifwithroute(flags, dst, gateway)) == 0) + if (flags & RTF_IFSCOPE) { + ifa = ifa_ifwithroute_scoped_locked(flags, dst0, + gateway, ifscope); + } else { + ifa = ifa_ifwithroute_locked(flags, dst0, gateway); + } + if (ifa == NULL) senderr(ENETUNREACH); - - makeroute: - R_Malloc(rt, struct rtentry *, sizeof(*rt)); - if (rt == 0) +makeroute: + if ((rt = rte_alloc()) == NULL) senderr(ENOBUFS); Bzero(rt, sizeof(*rt)); + rte_lock_init(rt); + RT_LOCK(rt); rt->rt_flags = RTF_UP | flags; + /* * Add the gateway. Possibly re-malloc-ing the storage for it * also add the rt_gwroute if possible. */ if ((error = rt_setgate(rt, dst, gateway)) != 0) { - R_Free(rt); + RT_UNLOCK(rt); + rte_lock_destroy(rt); + rte_free(rt); senderr(error); } @@ -782,9 +1634,9 @@ rtrequest_locked( /* * make sure it contains the value we want (masked if needed). */ - if (netmask) { + if (netmask) rt_maskedcopy(dst, ndst, netmask); - } else + else Bcopy(dst, ndst, dst->sa_len); /* @@ -808,12 +1660,21 @@ rtrequest_locked( * mechanism, then we just blow it away and retry * the insertion of the new one. */ - rt2 = rtalloc1_locked(dst, 0, RTF_PRCLONING); + if (flags & RTF_IFSCOPE) { + rt2 = rtalloc1_scoped_locked(dst0, 0, + RTF_CLONING | RTF_PRCLONING, ifscope); + } else { + rt2 = rtalloc1_locked(dst, 0, + RTF_CLONING | RTF_PRCLONING); + } if (rt2 && rt2->rt_parent) { - rtrequest_locked(RTM_DELETE, - (struct sockaddr *)rt_key(rt2), - rt2->rt_gateway, - rt_mask(rt2), rt2->rt_flags, 0); + /* + * rnh_lock is held here, so rt_key and + * rt_gateway of rt2 will not change. + */ + (void) rtrequest_locked(RTM_DELETE, rt_key(rt2), + rt2->rt_gateway, rt_mask(rt2), + rt2->rt_flags, 0); rtfree_locked(rt2); rn = rnh->rnh_addaddr((caddr_t)ndst, (caddr_t)netmask, @@ -829,29 +1690,40 @@ rtrequest_locked( * then un-make it (this should be a function) */ if (rn == 0) { - if (rt->rt_gwroute) + if (rt->rt_gwroute) { rtfree_locked(rt->rt_gwroute); + rt->rt_gwroute = NULL; + } if (rt->rt_ifa) { ifafree(rt->rt_ifa); + rt->rt_ifa = NULL; } R_Free(rt_key(rt)); - R_Free(rt); + RT_UNLOCK(rt); + rte_lock_destroy(rt); + rte_free(rt); senderr(EEXIST); } rt->rt_parent = 0; /* - * If we got here from RESOLVE, then we are cloning - * so clone the rest, and note that we - * are a clone (and increment the parent's references) + * If we got here from RESOLVE, then we are cloning so clone + * the rest, and note that we are a clone (and increment the + * parent's references). rnh_lock is still held, which prevents + * a lookup from returning the newly-created route. Hence + * holding and releasing the parent's rt_lock while still + * holding the route's rt_lock is safe since the new route + * is not yet externally visible. */ if (req == RTM_RESOLVE) { + RT_LOCK_SPIN(*ret_nrt); rt->rt_rmx = (*ret_nrt)->rt_rmx; /* copy metrics */ if ((*ret_nrt)->rt_flags & (RTF_CLONING | RTF_PRCLONING)) { rt->rt_parent = (*ret_nrt); - rtref(*ret_nrt); + RT_ADDREF_LOCKED(*ret_nrt); } + RT_UNLOCK(*ret_nrt); } /* @@ -863,6 +1735,22 @@ rtrequest_locked( ifafree(ifa); ifa = 0; + /* + * If this is the (non-scoped) default route, record + * the interface index used for the primary ifscope. + */ + if (rt_inet_default(rt, rt_key(rt))) + set_primary_ifscope(rt->rt_ifp->if_index); + + /* + * actually return a resultant rtentry and + * give the caller a single reference. + */ + if (ret_nrt) { + *ret_nrt = rt; + RT_ADDREF_LOCKED(rt); + } + /* * We repeat the same procedure from rt_setgate() here because * it doesn't fire when we call it there because the node @@ -872,17 +1760,11 @@ rtrequest_locked( struct rtfc_arg arg; arg.rnh = rnh; arg.rt0 = rt; + RT_UNLOCK(rt); rnh->rnh_walktree_from(rnh, rt_key(rt), rt_mask(rt), rt_fixchange, &arg); - } - - /* - * actually return a resultant rtentry and - * give the caller a single reference. - */ - if (ret_nrt) { - *ret_nrt = rt; - rtref(rt); + } else { + RT_UNLOCK(rt); } break; } @@ -902,10 +1784,10 @@ rtrequest( struct rtentry **ret_nrt) { int error; - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_NOTOWNED); - lck_mtx_lock(rt_mtx); + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_NOTOWNED); + lck_mtx_lock(rnh_lock); error = rtrequest_locked(req, dst, gateway, netmask, flags, ret_nrt); - lck_mtx_unlock(rt_mtx); + lck_mtx_unlock(rnh_lock); return (error); } /* @@ -916,20 +1798,26 @@ rtrequest( * the late parent (passed in as VP here) are themselves deleted. */ static int -rt_fixdelete(rn, vp) - struct radix_node *rn; - void *vp; +rt_fixdelete(struct radix_node *rn, void *vp) { struct rtentry *rt = (struct rtentry *)rn; struct rtentry *rt0 = vp; - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_OWNED); + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_OWNED); - if (rt->rt_parent == rt0 && !(rt->rt_flags & RTF_PINNED)) { - return rtrequest_locked(RTM_DELETE, rt_key(rt), - (struct sockaddr *)0, rt_mask(rt), - rt->rt_flags, (struct rtentry **)0); + RT_LOCK(rt); + if (rt->rt_parent == rt0 && + !(rt->rt_flags & (RTF_PINNED | RTF_CLONING | RTF_PRCLONING))) { + /* + * Safe to drop rt_lock and use rt_key, since holding + * rnh_lock here prevents another thread from calling + * rt_setgate() on this route. + */ + RT_UNLOCK(rt); + return (rtrequest_locked(RTM_DELETE, rt_key(rt), NULL, + rt_mask(rt), rt->rt_flags, NULL)); } + RT_UNLOCK(rt); return 0; } @@ -946,14 +1834,8 @@ rt_fixdelete(rn, vp) * routine just for adds. I'm not sure why I thought it was necessary to do * changes this way. */ -#ifdef DEBUG -static int rtfcdebug = 0; -#endif - static int -rt_fixchange(rn, vp) - struct radix_node *rn; - void *vp; +rt_fixchange(struct radix_node *rn, void *vp) { struct rtentry *rt = (struct rtentry *)rn; struct rtfc_arg *ap = vp; @@ -962,35 +1844,24 @@ rt_fixchange(rn, vp) u_char *xk1, *xm1, *xk2, *xmp; int i, len, mlen; -#ifdef DEBUG - if (rtfcdebug) - printf("rt_fixchange: rt %p, rt0 %p\n", rt, rt0); -#endif + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_OWNED); - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_OWNED); + RT_LOCK(rt); - if (!rt->rt_parent || (rt->rt_flags & RTF_PINNED)) { -#ifdef DEBUG - if(rtfcdebug) printf("no parent or pinned\n"); -#endif - return 0; + if (!rt->rt_parent || + (rt->rt_flags & (RTF_PINNED | RTF_CLONING | RTF_PRCLONING))) { + RT_UNLOCK(rt); + return (0); } - if (rt->rt_parent == rt0) { -#ifdef DEBUG - if(rtfcdebug) printf("parent match\n"); -#endif - return rtrequest_locked(RTM_DELETE, rt_key(rt), - (struct sockaddr *)0, rt_mask(rt), - rt->rt_flags, (struct rtentry **)0); - } + if (rt->rt_parent == rt0) + goto delete_rt; /* * There probably is a function somewhere which does this... * if not, there should be. */ - len = imin(((struct sockaddr *)rt_key(rt0))->sa_len, - ((struct sockaddr *)rt_key(rt))->sa_len); + len = imin(rt_key(rt0)->sa_len, rt_key(rt)->sa_len); xk1 = (u_char *)rt_key(rt0); xm1 = (u_char *)rt_mask(rt0); @@ -998,32 +1869,23 @@ rt_fixchange(rn, vp) /* avoid applying a less specific route */ xmp = (u_char *)rt_mask(rt->rt_parent); - mlen = ((struct sockaddr *)rt_key(rt->rt_parent))->sa_len; - if (mlen > ((struct sockaddr *)rt_key(rt0))->sa_len) { -#if DEBUG - if (rtfcdebug) - printf("rt_fixchange: inserting a less " - "specific route\n"); -#endif - return 0; + mlen = rt_key(rt->rt_parent)->sa_len; + if (mlen > rt_key(rt0)->sa_len) { + RT_UNLOCK(rt); + return (0); } + for (i = rnh->rnh_treetop->rn_offset; i < mlen; i++) { if ((xmp[i] & ~(xmp[i] ^ xm1[i])) != xmp[i]) { -#if DEBUG - if (rtfcdebug) - printf("rt_fixchange: inserting a less " - "specific route\n"); -#endif - return 0; + RT_UNLOCK(rt); + return (0); } } for (i = rnh->rnh_treetop->rn_offset; i < len; i++) { if ((xk2[i] & xm1[i]) != xk1[i]) { -#ifdef DEBUG - if(rtfcdebug) printf("no match\n"); -#endif - return 0; + RT_UNLOCK(rt); + return (0); } } @@ -1031,110 +1893,218 @@ rt_fixchange(rn, vp) * OK, this node is a clone, and matches the node currently being * changed/added under the node's mask. So, get rid of it. */ -#ifdef DEBUG - if(rtfcdebug) printf("deleting\n"); -#endif - return rtrequest_locked(RTM_DELETE, rt_key(rt), (struct sockaddr *)0, - rt_mask(rt), rt->rt_flags, (struct rtentry **)0); +delete_rt: + /* + * Safe to drop rt_lock and use rt_key, since holding rnh_lock here + * prevents another thread from calling rt_setgate() on this route. + */ + RT_UNLOCK(rt); + return (rtrequest_locked(RTM_DELETE, rt_key(rt), NULL, + rt_mask(rt), rt->rt_flags, NULL)); } +/* + * Round up sockaddr len to multiples of 32-bytes. This will reduce + * or even eliminate the need to re-allocate the chunk of memory used + * for rt_key and rt_gateway in the event the gateway portion changes. + * Certain code paths (e.g. IPSec) are notorious for caching the address + * of rt_gateway; this rounding-up would help ensure that the gateway + * portion never gets deallocated (though it may change contents) and + * thus greatly simplifies things. + */ +#define SA_SIZE(x) (-(-((uintptr_t)(x)) & -(32))) + +/* + * Sets the gateway and/or gateway route portion of a route; may be + * called on an existing route to modify the gateway portion. Both + * rt_key and rt_gateway are allocated out of the same memory chunk. + * Route entry lock must be held by caller; this routine will return + * with the lock held. + */ int -rt_setgate(rt0, dst, gate) - struct rtentry *rt0; - struct sockaddr *dst, *gate; +rt_setgate(struct rtentry *rt, struct sockaddr *dst, struct sockaddr *gate) { - caddr_t new, old; - int dlen = ROUNDUP(dst->sa_len), glen = ROUNDUP(gate->sa_len); - register struct rtentry *rt = rt0; + int dlen = SA_SIZE(dst->sa_len), glen = SA_SIZE(gate->sa_len); struct radix_node_head *rnh = rt_tables[dst->sa_family]; - extern void kdp_set_gateway_mac (void *gatewaymac); + + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_OWNED); + RT_LOCK_ASSERT_HELD(rt); + + /* + * If this is for a route that is on its way of being removed, + * or is temporarily frozen, reject the modification request. + */ + if (rt->rt_flags & RTF_CONDEMNED) + return (EBUSY); + + /* Add an extra ref for ourselves */ + RT_ADDREF_LOCKED(rt); + /* * A host route with the destination equal to the gateway * will interfere with keeping LLINFO in the routing * table, so disallow it. */ - - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_OWNED); - - if (((rt0->rt_flags & (RTF_HOST|RTF_GATEWAY|RTF_LLINFO)) == - (RTF_HOST|RTF_GATEWAY)) && - (dst->sa_len == gate->sa_len) && + if (((rt->rt_flags & (RTF_HOST|RTF_GATEWAY|RTF_LLINFO)) == + (RTF_HOST|RTF_GATEWAY)) && (dst->sa_len == gate->sa_len) && (bcmp(dst, gate, dst->sa_len) == 0)) { /* * The route might already exist if this is an RTM_CHANGE * or a routing redirect, so try to delete it. */ - if (rt_key(rt0)) - rtrequest_locked(RTM_DELETE, (struct sockaddr *)rt_key(rt0), - rt0->rt_gateway, rt_mask(rt0), rt0->rt_flags, 0); - return EADDRNOTAVAIL; + if (rt_key(rt) != NULL) { + /* + * Safe to drop rt_lock and use rt_key, rt_gateway, + * since holding rnh_lock here prevents another thread + * from calling rt_setgate() on this route. + */ + RT_UNLOCK(rt); + (void) rtrequest_locked(RTM_DELETE, rt_key(rt), + rt->rt_gateway, rt_mask(rt), rt->rt_flags, NULL); + RT_LOCK(rt); + } + /* Release extra ref */ + RT_REMREF_LOCKED(rt); + return (EADDRNOTAVAIL); } /* - * Both dst and gateway are stored in the same malloc'd chunk - * (If I ever get my hands on....) - * if we need to malloc a new chunk, then keep the old one around - * till we don't need it any more. + * The destination is not directly reachable. Get a route + * to the next-hop gateway and store it in rt_gwroute. */ - if (rt->rt_gateway == 0 || glen > ROUNDUP(rt->rt_gateway->sa_len)) { - old = (caddr_t)rt_key(rt); - R_Malloc(new, caddr_t, dlen + glen); - if (new == 0) - return ENOBUFS; - rt->rt_nodes->rn_key = new; - } else { + if (rt->rt_flags & RTF_GATEWAY) { + struct rtentry *gwrt; + unsigned int ifscope; + + ifscope = (dst->sa_family == AF_INET) ? + sa_get_ifscope(dst) : IFSCOPE_NONE; + + RT_UNLOCK(rt); + gwrt = rtalloc1_scoped_locked(gate, 1, RTF_PRCLONING, ifscope); + if (gwrt != NULL) + RT_LOCK_ASSERT_NOTHELD(gwrt); + RT_LOCK(rt); + + /* + * Cloning loop avoidance: + * + * In the presence of protocol-cloning and bad configuration, + * it is possible to get stuck in bottomless mutual recursion + * (rtrequest rt_setgate rtalloc1). We avoid this by not + * allowing protocol-cloning to operate for gateways (which + * is probably the correct choice anyway), and avoid the + * resulting reference loops by disallowing any route to run + * through itself as a gateway. This is obviously mandatory + * when we get rt->rt_output(). It implies that a route to + * the gateway must already be present in the system in order + * for the gateway to be referred to by another route. + */ + if (gwrt == rt) { + RT_REMREF_LOCKED(gwrt); + /* Release extra ref */ + RT_REMREF_LOCKED(rt); + return (EADDRINUSE); /* failure */ + } + + /* + * If scoped, the gateway route must use the same interface; + * we're holding rnh_lock now, so rt_gateway and rt_ifp of gwrt + * should not change and are freely accessible. + */ + if (ifscope != IFSCOPE_NONE && (rt->rt_flags & RTF_IFSCOPE) && + gwrt != NULL && gwrt->rt_ifp != NULL && + gwrt->rt_ifp->if_index != ifscope) { + rtfree_locked(gwrt); /* rt != gwrt, no deadlock */ + /* Release extra ref */ + RT_REMREF_LOCKED(rt); + return ((rt->rt_flags & RTF_HOST) ? + EHOSTUNREACH : ENETUNREACH); + } + + /* Check again since we dropped the lock above */ + if (rt->rt_flags & RTF_CONDEMNED) { + if (gwrt != NULL) + rtfree_locked(gwrt); + /* Release extra ref */ + RT_REMREF_LOCKED(rt); + return (EBUSY); + } + + if (rt->rt_gwroute != NULL) + rtfree_locked(rt->rt_gwroute); + rt->rt_gwroute = gwrt; + + /* + * In case the (non-scoped) default route gets modified via + * an ICMP redirect, record the interface index used for the + * primary ifscope. Also done in rt_setif() to take care + * of the non-redirect cases. + */ + if (rt_inet_default(rt, dst) && rt->rt_ifp != NULL) + set_primary_ifscope(rt->rt_ifp->if_index); + /* - * otherwise just overwrite the old one + * Tell the kernel debugger about the new default gateway + * if the gateway route uses the primary interface, or + * if we are in a transient state before the non-scoped + * default gateway is installed (similar to how the system + * was behaving in the past). In future, it would be good + * to do all this only when KDP is enabled. */ - new = rt->rt_nodes->rn_key; - old = 0; + if ((dst->sa_family == AF_INET) && + gwrt != NULL && gwrt->rt_gateway->sa_family == AF_LINK && + (gwrt->rt_ifp->if_index == get_primary_ifscope() || + get_primary_ifscope() == IFSCOPE_NONE)) + kdp_set_gateway_mac(SDL(gwrt->rt_gateway)->sdl_data); } /* - * copy the new gateway value into the memory chunk + * Prepare to store the gateway in rt_gateway. Both dst and gateway + * are stored one after the other in the same malloc'd chunk. If we + * have room, reuse the old buffer since rt_gateway already points + * to the right place. Otherwise, malloc a new block and update + * the 'dst' address and point rt_gateway to the right place. */ - Bcopy(gate, (rt->rt_gateway = (struct sockaddr *)(new + dlen)), glen); + if (rt->rt_gateway == NULL || glen > SA_SIZE(rt->rt_gateway->sa_len)) { + caddr_t new; - /* - * if we are replacing the chunk (or it's new) we need to - * replace the dst as well - */ - if (old) { - Bcopy(dst, new, dlen); - R_Free(old); + /* The underlying allocation is done with M_WAITOK set */ + R_Malloc(new, caddr_t, dlen + glen); + if (new == NULL) { + if (rt->rt_gwroute != NULL) + rtfree_locked(rt->rt_gwroute); + rt->rt_gwroute = NULL; + /* Release extra ref */ + RT_REMREF_LOCKED(rt); + return (ENOBUFS); + } + + /* + * Copy from 'dst' and not rt_key(rt) because we can get + * here to initialize a newly allocated route entry, in + * which case rt_key(rt) is NULL (and so does rt_gateway). + */ + bzero(new, dlen + glen); + Bcopy(dst, new, dst->sa_len); + R_Free(rt_key(rt)); /* free old block; NULL is okay */ + rt->rt_nodes->rn_key = new; + rt->rt_gateway = (struct sockaddr *)(new + dlen); } /* - * If there is already a gwroute, it's now almost definitly wrong - * so drop it. + * Copy the new gateway value into the memory chunk. */ - if (rt->rt_gwroute) { - rt = rt->rt_gwroute; rtfree_locked(rt); - rt = rt0; rt->rt_gwroute = 0; - } + Bcopy(gate, rt->rt_gateway, gate->sa_len); + /* - * Cloning loop avoidance: - * In the presence of protocol-cloning and bad configuration, - * it is possible to get stuck in bottomless mutual recursion - * (rtrequest rt_setgate rtalloc1). We avoid this by not allowing - * protocol-cloning to operate for gateways (which is probably the - * correct choice anyway), and avoid the resulting reference loops - * by disallowing any route to run through itself as a gateway. - * This is obviously mandatory when we get rt->rt_output(). + * For consistency between rt_gateway and rt_key(gwrt). */ - if (rt->rt_flags & RTF_GATEWAY) { - rt->rt_gwroute = rtalloc1_locked(gate, 1, RTF_PRCLONING); - if (rt->rt_gwroute == rt) { - rtfree_locked(rt->rt_gwroute); - rt->rt_gwroute = 0; - return EDQUOT; /* failure */ - } - /* Tell the kernel debugger about the new default gateway */ - if ((AF_INET == rt->rt_gateway->sa_family) && - rt->rt_gwroute && rt->rt_gwroute->rt_gateway && - (AF_LINK == rt->rt_gwroute->rt_gateway->sa_family)) { - kdp_set_gateway_mac(((struct sockaddr_dl *)rt0->rt_gwroute->rt_gateway)->sdl_data); - } + if ((rt->rt_flags & RTF_GATEWAY) && rt->rt_gwroute != NULL && + (rt->rt_gwroute->rt_flags & RTF_IFSCOPE) && + rt->rt_gateway->sa_family == AF_INET && + rt_key(rt->rt_gwroute)->sa_family == AF_INET) { + sa_set_ifscope(rt->rt_gateway, + sa_get_ifscope(rt_key(rt->rt_gwroute))); } /* @@ -1146,20 +2116,26 @@ rt_setgate(rt0, dst, gate) struct rtfc_arg arg; arg.rnh = rnh; arg.rt0 = rt; + RT_UNLOCK(rt); rnh->rnh_walktree_from(rnh, rt_key(rt), rt_mask(rt), - rt_fixchange, &arg); + rt_fixchange, &arg); + RT_LOCK(rt); } - return 0; + /* Release extra ref */ + RT_REMREF_LOCKED(rt); + return (0); } +#undef SA_SIZE + static void -rt_maskedcopy(src, dst, netmask) - struct sockaddr *src, *dst, *netmask; +rt_maskedcopy(struct sockaddr *src, struct sockaddr *dst, + struct sockaddr *netmask) { - register u_char *cp1 = (u_char *)src; - register u_char *cp2 = (u_char *)dst; - register u_char *cp3 = (u_char *)netmask; + u_char *cp1 = (u_char *)src; + u_char *cp2 = (u_char *)dst; + u_char *cp3 = (u_char *)netmask; u_char *cplim = cp2 + *cp3; u_char *cplim2 = cp2 + *cp1; @@ -1173,31 +2149,253 @@ rt_maskedcopy(src, dst, netmask) bzero((caddr_t)cp2, (unsigned)(cplim2 - cp2)); } +/* + * Lookup an AF_INET scoped or non-scoped route depending on the ifscope + * value passed in by the caller (IFSCOPE_NONE implies non-scoped). + */ +static struct radix_node * +node_lookup(struct sockaddr *dst, struct sockaddr *netmask, + unsigned int ifscope) +{ + struct radix_node_head *rnh = rt_tables[AF_INET]; + struct radix_node *rn; + struct sockaddr_in sin, mask; + struct matchleaf_arg ma = { ifscope }; + rn_matchf_t *f = rn_match_ifscope; + void *w = &ma; + + if (dst->sa_family != AF_INET) + return (NULL); + + /* + * Embed ifscope into the search key; for a non-scoped + * search this will clear out any embedded scope value. + */ + dst = sin_copy(SIN(dst), &sin, ifscope); + + /* Embed (or clear) ifscope into netmask */ + if (netmask != NULL) + netmask = mask_copy(netmask, &mask, ifscope); + + if (ifscope == IFSCOPE_NONE) + f = w = NULL; + + rn = rnh->rnh_lookup_args(dst, netmask, rnh, f, w); + if (rn != NULL && (rn->rn_flags & RNF_ROOT)) + rn = NULL; + + return (rn); +} + +/* + * Lookup the AF_INET non-scoped default route. + */ +static struct radix_node * +node_lookup_default(void) +{ + struct radix_node_head *rnh = rt_tables[AF_INET]; + return (rnh->rnh_lookup(&sin_def, NULL, rnh)); +} + +/* + * Common routine to lookup/match a route. It invokes the lookup/matchaddr + * callback which could be address family-specific. The main difference + * between the two (at least for AF_INET/AF_INET6) is that a lookup does + * not alter the expiring state of a route, whereas a match would unexpire + * or revalidate the route. + * + * The optional scope or interface index property of a route allows for a + * per-interface route instance. This permits multiple route entries having + * the same destination (but not necessarily the same gateway) to exist in + * the routing table; each of these entries is specific to the corresponding + * interface. This is made possible by embedding the scope value into the + * radix key, thus making each route entry unique. These scoped entries + * exist along with the regular, non-scoped entries in the same radix tree + * for a given address family (currently AF_INET only); the scope logically + * partitions it into multiple per-interface sub-trees. + * + * When a scoped route lookup is performed, the routing table is searched for + * the best match that would result in a route using the same interface as the + * one associated with the scope (the exception to this are routes that point + * to the loopback interface). The search rule follows the longest matching + * prefix with the additional interface constraint. + */ +struct rtentry * +rt_lookup(boolean_t lookup_only, struct sockaddr *dst, struct sockaddr *netmask, + struct radix_node_head *rnh, unsigned int ifscope) +{ + struct radix_node *rn0, *rn; + boolean_t dontcare = (ifscope == IFSCOPE_NONE); + + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_OWNED); + + if (!lookup_only) + netmask = NULL; + + /* + * Non-scoped route lookup. + */ + if (!ip_doscopedroute || dst->sa_family != AF_INET) { + if (lookup_only) + rn = rnh->rnh_lookup(dst, netmask, rnh); + else + rn = rnh->rnh_matchaddr(dst, rnh); + + /* + * Don't return a root node; also, rnh_matchaddr callback + * would have done the necessary work to clear RTPRF_OURS + * for certain protocol families. + */ + if (rn != NULL && (rn->rn_flags & RNF_ROOT)) + rn = NULL; + if (rn != NULL) { + RT_LOCK_SPIN(RT(rn)); + if (!(RT(rn)->rt_flags & RTF_CONDEMNED)) { + RT_ADDREF_LOCKED(RT(rn)); + RT_UNLOCK(RT(rn)); + } else { + RT_UNLOCK(RT(rn)); + rn = NULL; + } + } + return (RT(rn)); + } + + /* + * Scoped route lookup: + * + * We first perform a non-scoped lookup for the original result. + * Afterwards, depending on whether or not the caller has specified + * a scope, we perform a more specific scoped search and fallback + * to this original result upon failure. + */ + rn0 = rn = node_lookup(dst, netmask, IFSCOPE_NONE); + + /* + * If the caller did not specify a scope, use the primary scope + * derived from the system's non-scoped default route. If, for + * any reason, there is no primary interface, return what we have. + */ + if (dontcare && (ifscope = get_primary_ifscope()) == IFSCOPE_NONE) + goto done; + + /* + * Keep the original result if either of the following is true: + * + * 1) The interface portion of the route has the same interface + * index as the scope value and it is marked with RTF_IFSCOPE. + * 2) The route uses the loopback interface, in which case the + * destination (host/net) is local/loopback. + * + * Otherwise, do a more specified search using the scope; + * we're holding rnh_lock now, so rt_ifp should not change. + */ + if (rn != NULL) { + struct rtentry *rt = RT(rn); + if (rt->rt_ifp != lo_ifp) { + if (rt->rt_ifp->if_index != ifscope) { + /* + * Wrong interface; keep the original result + * only if the caller did not specify a scope, + * and do a more specific scoped search using + * the scope of the found route. Otherwise, + * start again from scratch. + */ + rn = NULL; + if (dontcare) + ifscope = rt->rt_ifp->if_index; + else + rn0 = NULL; + } else if (!(rt->rt_flags & RTF_IFSCOPE)) { + /* + * Right interface, except that this route + * isn't marked with RTF_IFSCOPE. Do a more + * specific scoped search. Keep the original + * result and return it it in case the scoped + * search fails. + */ + rn = NULL; + } + } + } + + /* + * Scoped search. Find the most specific entry having the same + * interface scope as the one requested. The following will result + * in searching for the longest prefix scoped match. + */ + if (rn == NULL) + rn = node_lookup(dst, netmask, ifscope); + + /* + * Use the original result if either of the following is true: + * + * 1) The scoped search did not yield any result. + * 2) The result from the scoped search is a scoped default route, + * and the original (non-scoped) result is not a default route, + * i.e. the original result is a more specific host/net route. + * 3) The scoped search yielded a net route but the original + * result is a host route, i.e. the original result is treated + * as a more specific route. + */ + if (rn == NULL || (rn0 != NULL && + ((INET_DEFAULT(rt_key(RT(rn))) && !INET_DEFAULT(rt_key(RT(rn0)))) || + (!RT_HOST(rn) && RT_HOST(rn0))))) + rn = rn0; + + /* + * If we still don't have a route, use the non-scoped default + * route as long as the interface portion satistifes the scope. + */ + if (rn == NULL && (rn = node_lookup_default()) != NULL && + RT(rn)->rt_ifp->if_index != ifscope) + rn = NULL; + +done: + if (rn != NULL) { + /* + * Manually clear RTPRF_OURS using in_validate() and + * bump up the reference count after, and not before; + * we only get here for AF_INET. node_lookup() has + * done the check against RNF_ROOT, so we can be sure + * that we're not returning a root node here. + */ + RT_LOCK_SPIN(RT(rn)); + if (!(RT(rn)->rt_flags & RTF_CONDEMNED)) { + if (!lookup_only) + (void) in_validate(rn); + RT_ADDREF_LOCKED(RT(rn)); + RT_UNLOCK(RT(rn)); + } else { + RT_UNLOCK(RT(rn)); + rn = NULL; + } + } + + return (RT(rn)); +} + /* * Set up a routing table entry, normally * for an interface. */ int -rtinit(ifa, cmd, flags) - register struct ifaddr *ifa; - int cmd, flags; +rtinit(struct ifaddr *ifa, int cmd, int flags) { int error; - lck_mtx_assert(rt_mtx, LCK_MTX_ASSERT_NOTOWNED); - lck_mtx_lock(rt_mtx); + lck_mtx_assert(rnh_lock, LCK_MTX_ASSERT_NOTOWNED); + lck_mtx_lock(rnh_lock); error = rtinit_locked(ifa, cmd, flags); - lck_mtx_unlock(rt_mtx); + lck_mtx_unlock(rnh_lock); return (error); } int -rtinit_locked(ifa, cmd, flags) - register struct ifaddr *ifa; - int cmd, flags; +rtinit_locked(struct ifaddr *ifa, int cmd, int flags) { - register struct rtentry *rt; - register struct sockaddr *dst; - register struct sockaddr *deldst; + struct rtentry *rt; + struct sockaddr *dst; + struct sockaddr *deldst; struct mbuf *m = 0; struct rtentry *nrt = 0; int error; @@ -1229,15 +2427,17 @@ rtinit_locked(ifa, cmd, flags) * We set "report" to FALSE so that if it doesn't exist, * it doesn't report an error or clone a route, etc. etc. */ - rt = rtalloc1_locked(dst, 0, 0UL); + rt = rtalloc1_locked(dst, 0, 0); if (rt) { /* * Ok so we found the rtentry. it has an extra reference * for us at this stage. we won't need that so * lop that off now. */ - rtunref(rt); + RT_LOCK_SPIN(rt); if (rt->rt_ifa != ifa) { + RT_REMREF_LOCKED(rt); + RT_UNLOCK(rt); /* * If the interface in the rtentry doesn't match * the interface we are using, then we don't @@ -1249,6 +2449,9 @@ rtinit_locked(ifa, cmd, flags) (void) m_free(m); return (flags & RTF_HOST ? EHOSTUNREACH : ENETUNREACH); + } else { + RT_REMREF_LOCKED(rt); + RT_UNLOCK(rt); } } /* XXX */ @@ -1259,7 +2462,7 @@ rtinit_locked(ifa, cmd, flags) * it doesn't exist, we could just return at this point * with an "ELSE" clause, but apparently not.. */ - lck_mtx_unlock(rt_mtx); + lck_mtx_unlock(rnh_lock); return (flags & RTF_HOST ? EHOSTUNREACH : ENETUNREACH); } @@ -1278,15 +2481,14 @@ rtinit_locked(ifa, cmd, flags) */ if (cmd == RTM_DELETE && error == 0 && (rt = nrt)) { /* - * notify any listenning routing agents of the change + * notify any listening routing agents of the change */ + RT_LOCK(rt); rt_newaddrmsg(cmd, ifa, error, nrt); if (use_routegenid) - route_generation++; - if (rt->rt_refcnt <= 0) { - rt->rt_refcnt++; /* need a 1->0 transition to free */ - rtfree_locked(rt); - } + routegenid_update(); + RT_UNLOCK(rt); + rtfree_locked(rt); } /* @@ -1294,10 +2496,7 @@ rtinit_locked(ifa, cmd, flags) * We need to sanity check the result. */ if (cmd == RTM_ADD && error == 0 && (rt = nrt)) { - /* - * We just wanted to add it.. we don't actually need a reference - */ - rtunref(rt); + RT_LOCK(rt); /* * If it came back with an unexpected interface, then it must * have already existed or something. (XXX) @@ -1336,7 +2535,143 @@ rtinit_locked(ifa, cmd, flags) */ rt_newaddrmsg(cmd, ifa, error, nrt); if (use_routegenid) - route_generation++; - } + routegenid_update(); + /* + * We just wanted to add it; we don't actually need a + * reference. This will result in a route that's added + * to the routing table without a reference count. The + * RTM_DELETE code will do the necessary step to adjust + * the reference count at deletion time. + */ + RT_REMREF_LOCKED(rt); + RT_UNLOCK(rt); + } return (error); } + +static void +rte_lock_init(struct rtentry *rt) +{ + lck_mtx_init(&rt->rt_lock, rte_mtx_grp, rte_mtx_attr); +} + +static void +rte_lock_destroy(struct rtentry *rt) +{ + RT_LOCK_ASSERT_NOTHELD(rt); + lck_mtx_destroy(&rt->rt_lock, rte_mtx_grp); +} + +void +rt_lock(struct rtentry *rt, boolean_t spin) +{ + RT_LOCK_ASSERT_NOTHELD(rt); + if (spin) + lck_mtx_lock_spin(&rt->rt_lock); + else + lck_mtx_lock(&rt->rt_lock); + if (rte_debug & RTD_DEBUG) + rte_lock_debug((struct rtentry_dbg *)rt); +} + +void +rt_unlock(struct rtentry *rt) +{ + RT_LOCK_ASSERT_HELD(rt); + if (rte_debug & RTD_DEBUG) + rte_unlock_debug((struct rtentry_dbg *)rt); + lck_mtx_unlock(&rt->rt_lock); + +} + +static inline void +rte_lock_debug(struct rtentry_dbg *rte) +{ + uint32_t idx; + + idx = atomic_add_32_ov(&rte->rtd_lock_cnt, 1) % CTRACE_HIST_SIZE; + if (rte_debug & RTD_TRACE) + ctrace_record(&rte->rtd_lock[idx]); +} + +static inline void +rte_unlock_debug(struct rtentry_dbg *rte) +{ + uint32_t idx; + + idx = atomic_add_32_ov(&rte->rtd_unlock_cnt, 1) % CTRACE_HIST_SIZE; + if (rte_debug & RTD_TRACE) + ctrace_record(&rte->rtd_unlock[idx]); +} + +static struct rtentry * +rte_alloc(void) +{ + if (rte_debug & RTD_DEBUG) + return (rte_alloc_debug()); + + return ((struct rtentry *)zalloc(rte_zone)); +} + +static void +rte_free(struct rtentry *p) +{ + if (rte_debug & RTD_DEBUG) { + rte_free_debug(p); + return; + } + + if (p->rt_refcnt != 0) + panic("rte_free: rte=%p refcnt=%d non-zero\n", p, p->rt_refcnt); + + zfree(rte_zone, p); +} + +static inline struct rtentry * +rte_alloc_debug(void) +{ + struct rtentry_dbg *rte; + + rte = ((struct rtentry_dbg *)zalloc(rte_zone)); + if (rte != NULL) { + bzero(rte, sizeof (*rte)); + if (rte_debug & RTD_TRACE) + ctrace_record(&rte->rtd_alloc); + rte->rtd_inuse = RTD_INUSE; + } + return ((struct rtentry *)rte); +} + +static inline void +rte_free_debug(struct rtentry *p) +{ + struct rtentry_dbg *rte = (struct rtentry_dbg *)p; + + if (p->rt_refcnt != 0) + panic("rte_free: rte=%p refcnt=%d\n", p, p->rt_refcnt); + + if (rte->rtd_inuse == RTD_FREED) + panic("rte_free: double free rte=%p\n", rte); + else if (rte->rtd_inuse != RTD_INUSE) + panic("rte_free: corrupted rte=%p\n", rte); + + bcopy((caddr_t)p, (caddr_t)&rte->rtd_entry_saved, sizeof (*p)); + /* Preserve rt_lock to help catch use-after-free cases */ + bzero((caddr_t)p, offsetof(struct rtentry, rt_lock)); + + rte->rtd_inuse = RTD_FREED; + + if (rte_debug & RTD_TRACE) + ctrace_record(&rte->rtd_free); + + if (!(rte_debug & RTD_NO_FREE)) + zfree(rte_zone, p); +} + +void +ctrace_record(ctrace_t *tr) +{ + tr->th = current_thread(); + bzero(tr->pc, sizeof (tr->pc)); + (void) OSBacktrace(tr->pc, CTRACE_STACK_SIZE); +}