* [PATCHSET] IPv4 FIB configuration rework
@ 2006-08-08 19:38 Thomas Graf
2006-08-07 22:00 ` [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config Thomas Graf
` (2 more replies)
0 siblings, 3 replies; 7+ messages in thread
From: Thomas Graf @ 2006-08-08 19:38 UTC (permalink / raw)
To: davem; +Cc: netdev
Replaces the rather ugly struct kern_rta + rtmsg with a new
struct fib_config while converting everything to the new netlink
api. Simplifies the FIB module interface quite a bit and allows
changing internals while keeping a stable netlink interface.
Gets rid of things like passing on netlink_skb_parms all around
the fib code etc.
Also fixes some memory corruptions that could be triggered via
netlink messages, not sure which would actually lead to an oops
though.
^ permalink raw reply [flat|nested] 7+ messages in thread* [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config 2006-08-08 19:38 [PATCHSET] IPv4 FIB configuration rework Thomas Graf @ 2006-08-07 22:00 ` Thomas Graf 2006-08-08 23:11 ` Thomas Graf 2006-08-07 22:00 ` [PATCH 2/3] [IPv4]: Convert FIB dumping to use new netlink api Thomas Graf 2006-08-07 22:00 ` [PATCH 3/3] [IPv4]: Convert route get to " Thomas Graf 2 siblings, 1 reply; 7+ messages in thread From: Thomas Graf @ 2006-08-07 22:00 UTC (permalink / raw) To: davem; +Cc: netdev [-- Attachment #1: nl_ipv4_route_fib_config --] [-- Type: text/plain, Size: 43292 bytes --] Introduces struct fib_config replacing the ugly struct kern_rta prone to ordering issues. Avoids creating faked netlink messages for auto generated routes or requests via ioctl. A new interface net/nexthop.h is added to help navigate through nexthop configuration arrays. Since the netlink source pid of the requesting process is stored in fib_config, it is no longer required to pass netlink_skb_parms all along just for event notification. Signed-off-by: Thomas Graf <tgraf@suug.ch> Index: net-2.6.19.git/include/net/ip_fib.h =================================================================== --- net-2.6.19.git.orig/include/net/ip_fib.h +++ net-2.6.19.git/include/net/ip_fib.h @@ -20,25 +20,46 @@ #include <linux/seq_file.h> #include <net/fib_rules.h> -/* WARNING: The ordering of these elements must match ordering - * of RTA_* rtnetlink attribute numbers. - */ -struct kern_rta { - void *rta_dst; - void *rta_src; - int *rta_iif; - int *rta_oif; - void *rta_gw; - u32 *rta_priority; - void *rta_prefsrc; - struct rtattr *rta_mx; - struct rtattr *rta_mp; - unsigned char *rta_protoinfo; - u32 *rta_flow; - struct rta_cacheinfo *rta_ci; - struct rta_session *rta_sess; - u32 *rta_mp_alg; -}; +struct fib_config { + u8 fc_family; + u8 fc_dst_len; + u8 fc_src_len; + u8 fc_tos; + u8 fc_protocol; + u8 fc_scope; + u8 fc_type; + /* 1 byte unused */ + u32 fc_table; + u32 fc_flags; + u32 fc_dst; + u32 fc_src; + u32 fc_gw; + int fc_oif; + u32 fc_priority; + u32 fc_prefsrc; + struct nlattr *fc_mx; + struct rtnexthop *fc_mp; + int fc_mx_len; + int fc_mp_len; + u32 fc_flow; + u32 fc_mp_alg; + + /* + * Request was issued via netlink, points to netlink header + * to answer request + */ + struct nlmsghdr *fc_nlh; + + /* + * Use separate field for netlink flags so non-netlink + * sources are able to emulate it without providing a + * complete header. + */ + u32 fc_nlmsg_flags; + + /* Netlink PID of the process requesting the change */ + u32 fc_pid; + }; struct fib_info; @@ -153,12 +174,8 @@ struct fib_table { unsigned char tb_id; unsigned tb_stamp; int (*tb_lookup)(struct fib_table *tb, const struct flowi *flp, struct fib_result *res); - int (*tb_insert)(struct fib_table *table, struct rtmsg *r, - struct kern_rta *rta, struct nlmsghdr *n, - struct netlink_skb_parms *req); - int (*tb_delete)(struct fib_table *table, struct rtmsg *r, - struct kern_rta *rta, struct nlmsghdr *n, - struct netlink_skb_parms *req); + int (*tb_insert)(struct fib_table *, struct fib_config *); + int (*tb_delete)(struct fib_table *, struct fib_config *); int (*tb_dump)(struct fib_table *table, struct sk_buff *skb, struct netlink_callback *cb); int (*tb_flush)(struct fib_table *table); @@ -243,8 +260,6 @@ struct rtentry; extern int ip_fib_check_default(u32 gw, struct net_device *dev); extern int fib_sync_down(u32 local, struct net_device *dev, int force); extern int fib_sync_up(struct net_device *dev); -extern int fib_convert_rtentry(int cmd, struct nlmsghdr *nl, struct rtmsg *rtm, - struct kern_rta *rta, struct rtentry *r); extern u32 __fib_res_prefsrc(struct fib_result *res); /* Exported by fib_hash.c */ Index: net-2.6.19.git/net/ipv4/fib_frontend.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_frontend.c +++ net-2.6.19.git/net/ipv4/fib_frontend.c @@ -233,42 +233,190 @@ e_inval: #ifndef CONFIG_IP_NOSIOCRT +static inline u32 sk_extract_addr(struct sockaddr *addr) +{ + return ((struct sockaddr_in *) addr)->sin_addr.s_addr; +} + +static int put_rtax(struct nlattr *mx, int len, int type, u32 value) +{ + struct nlattr *nla; + + nla = (struct nlattr *) ((char *) mx + len); + nla->nla_type = type; + nla->nla_len = nla_attr_size(4); + *(u32 *) nla_data(nla) = value; + + return len + nla_total_size(4); +} + +static int rtentry_to_fib_config(int cmd, struct rtentry *rt, + struct fib_config *cfg) +{ + u32 addr; + int plen; + + memset(cfg, 0, sizeof(*cfg)); + + if (rt->rt_dst.sa_family != AF_INET) + return -EAFNOSUPPORT; + + /* + * Check mask for validity: + * a) it must be contiguous. + * b) destination must have all host bits clear. + * c) if application forgot to set correct family (AF_INET), + * reject request unless it is absolutely clear i.e. + * both family and mask are zero. + */ + plen = 32; + addr = sk_extract_addr(&rt->rt_dst); + if (!(rt->rt_flags & RTF_HOST)) { + u32 mask = sk_extract_addr(&rt->rt_genmask); + + if (rt->rt_genmask.sa_family != AF_INET) { + if (mask || rt->rt_genmask.sa_family) + return -EAFNOSUPPORT; + } + + if (bad_mask(mask, addr)) + return -EINVAL; + + plen = inet_mask_len(mask); + } + + cfg->fc_dst_len = plen; + cfg->fc_dst = addr; + + if (cmd != SIOCDELRT) { + cfg->fc_nlmsg_flags = NLM_F_CREATE; + cfg->fc_protocol = RTPROT_BOOT; + } + + if (rt->rt_metric) + cfg->fc_priority = rt->rt_metric - 1; + + if (rt->rt_flags & RTF_REJECT) { + cfg->fc_scope = RT_SCOPE_HOST; + cfg->fc_type = RTN_UNREACHABLE; + return 0; + } + + cfg->fc_scope = RT_SCOPE_NOWHERE; + cfg->fc_type = RTN_UNICAST; + + if (rt->rt_dev) { + char *colon; + struct net_device *dev; + char devname[IFNAMSIZ]; + + if (copy_from_user(devname, rt->rt_dev, IFNAMSIZ-1)) + return -EFAULT; + + devname[IFNAMSIZ-1] = 0; + colon = strchr(devname, ':'); + if (colon) + *colon = 0; + dev = __dev_get_by_name(devname); + if (!dev) + return -ENODEV; + cfg->fc_oif = dev->ifindex; + if (colon) { + struct in_ifaddr *ifa; + struct in_device *in_dev = __in_dev_get_rtnl(dev); + if (!in_dev) + return -ENODEV; + *colon = ':'; + for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next) + if (strcmp(ifa->ifa_label, devname) == 0) + break; + if (ifa == NULL) + return -ENODEV; + cfg->fc_prefsrc = ifa->ifa_local; + } + } + + addr = sk_extract_addr(&rt->rt_gateway); + if (rt->rt_gateway.sa_family == AF_INET && addr) { + cfg->fc_gw = addr; + if (rt->rt_flags & RTF_GATEWAY && + inet_addr_type(addr) == RTN_UNICAST) + cfg->fc_scope = RT_SCOPE_UNIVERSE; + } + + if (cmd == SIOCDELRT) + return 0; + + if (rt->rt_flags & RTF_GATEWAY && !cfg->fc_gw) + return -EINVAL; + + if (cfg->fc_scope == RT_SCOPE_NOWHERE) + cfg->fc_scope = RT_SCOPE_LINK; + + if (rt->rt_flags & (RTF_MTU | RTF_WINDOW | RTF_IRTT)) { + struct nlattr *mx; + int len = 0; + + mx = kzalloc(3 * nla_total_size(4), GFP_KERNEL); + if (mx == NULL) + return -ENOMEM; + + if (rt->rt_flags & RTF_MTU) + len = put_rtax(mx, len, RTAX_ADVMSS, rt->rt_mtu - 40); + + if (rt->rt_flags & RTF_WINDOW) + len = put_rtax(mx, len, RTAX_WINDOW, rt->rt_window); + + if (rt->rt_flags & RTF_IRTT) + len = put_rtax(mx, len, RTAX_RTT, rt->rt_irtt << 3); + + cfg->fc_mx = mx; + cfg->fc_mx_len = len; + } + + return 0; +} + /* * Handle IP routing ioctl calls. These are used to manipulate the routing tables */ int ip_rt_ioctl(unsigned int cmd, void __user *arg) { + struct fib_config cfg; + struct rtentry rt; int err; - struct kern_rta rta; - struct rtentry r; - struct { - struct nlmsghdr nlh; - struct rtmsg rtm; - } req; switch (cmd) { case SIOCADDRT: /* Add a route */ case SIOCDELRT: /* Delete a route */ if (!capable(CAP_NET_ADMIN)) return -EPERM; - if (copy_from_user(&r, arg, sizeof(struct rtentry))) + + if (copy_from_user(&rt, arg, sizeof(rt))) return -EFAULT; + rtnl_lock(); - err = fib_convert_rtentry(cmd, &req.nlh, &req.rtm, &rta, &r); + err = rtentry_to_fib_config(cmd, &rt, &cfg); if (err == 0) { + struct fib_table *tb; + if (cmd == SIOCDELRT) { - struct fib_table *tb = fib_get_table(req.rtm.rtm_table); - err = -ESRCH; + tb = fib_get_table(cfg.fc_table); if (tb) - err = tb->tb_delete(tb, &req.rtm, &rta, &req.nlh, NULL); + err = tb->tb_delete(tb, &cfg); + else + err = -ESRCH; } else { - struct fib_table *tb = fib_new_table(req.rtm.rtm_table); - err = -ENOBUFS; + tb = fib_new_table(cfg.fc_table); if (tb) - err = tb->tb_insert(tb, &req.rtm, &rta, &req.nlh, NULL); + err = tb->tb_insert(tb, &cfg); + else + err = -ENOBUFS; } - kfree(rta.rta_mx); + + /* allocated by rtentry_to_fib_config() */ + kfree(cfg.fc_mx); } rtnl_unlock(); return err; @@ -285,50 +433,128 @@ int ip_rt_ioctl(unsigned int cmd, void * #endif -static int inet_check_attr(struct rtmsg *r, struct rtattr **rta) -{ - int i; +static struct nla_policy rtm_ipv4_policy[RTA_MAX+1] __read_mostly = { + [RTA_DST] = { .type = NLA_U32 }, + [RTA_SRC] = { .type = NLA_U32 }, + [RTA_IIF] = { .type = NLA_U32 }, + [RTA_OIF] = { .type = NLA_U32 }, + [RTA_GATEWAY] = { .type = NLA_U32 }, + [RTA_PRIORITY] = { .type = NLA_U32 }, + [RTA_PREFSRC] = { .type = NLA_U32 }, + [RTA_METRICS] = { .type = NLA_NESTED }, + [RTA_MULTIPATH] = { .minlen = sizeof(struct rtnexthop) }, + [RTA_PROTOINFO] = { .type = NLA_U32 }, + [RTA_FLOW] = { .type = NLA_U32 }, + [RTA_MP_ALGO] = { .type = NLA_U32 }, +}; - for (i=1; i<=RTA_MAX; i++, rta++) { - struct rtattr *attr = *rta; - if (attr) { - if (RTA_PAYLOAD(attr) < 4) - return -EINVAL; - if (i != RTA_MULTIPATH && i != RTA_METRICS) - *rta = (struct rtattr*)RTA_DATA(attr); +static int rtm_to_fib_config(struct nlmsghdr *nlh, struct fib_config *cfg) +{ + struct nlattr *attr; + int err, remaining; + struct rtmsg *rtm; + + err = nlmsg_validate(nlh, sizeof(*rtm), RTA_MAX, rtm_ipv4_policy); + if (err < 0) + goto errout; + + memset(cfg, 0, sizeof(*cfg)); + + rtm = nlmsg_data(nlh); + cfg->fc_family = rtm->rtm_family; + cfg->fc_dst_len = rtm->rtm_dst_len; + cfg->fc_src_len = rtm->rtm_src_len; + cfg->fc_tos = rtm->rtm_tos; + cfg->fc_table = rtm->rtm_table; + cfg->fc_protocol = rtm->rtm_protocol; + cfg->fc_scope = rtm->rtm_scope; + cfg->fc_type = rtm->rtm_type; + cfg->fc_nlh = nlh; + cfg->fc_nlmsg_flags = nlh->nlmsg_flags; + cfg->fc_pid = nlh->nlmsg_pid; + + nlmsg_for_each_attr(attr, nlh, sizeof(struct rtmsg), remaining) { + switch (attr->nla_type) { + case RTA_DST: + cfg->fc_dst = nla_get_u32(attr); + break; + case RTA_SRC: + cfg->fc_src = nla_get_u32(attr); + break; + case RTA_OIF: + cfg->fc_oif = nla_get_u32(attr); + break; + case RTA_GATEWAY: + cfg->fc_gw = nla_get_u32(attr); + break; + case RTA_PRIORITY: + cfg->fc_priority = nla_get_u32(attr); + break; + case RTA_PREFSRC: + cfg->fc_prefsrc = nla_get_u32(attr); + break; + case RTA_METRICS: + cfg->fc_mx = nla_data(attr); + cfg->fc_mx_len = nla_len(attr); + break; + case RTA_MULTIPATH: + cfg->fc_mp = nla_data(attr); + cfg->fc_mp_len = nla_len(attr); + break; + case RTA_FLOW: + cfg->fc_flow = nla_get_u32(attr); + break; + case RTA_MP_ALGO: + cfg->fc_mp_alg = nla_get_u32(attr); + break; } } + return 0; +errout: + return err; } int inet_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg) { - struct fib_table * tb; - struct rtattr **rta = arg; - struct rtmsg *r = NLMSG_DATA(nlh); + struct fib_config cfg; + struct fib_table *tb; + int err; - if (inet_check_attr(r, rta)) - return -EINVAL; + err = rtm_to_fib_config(nlh, &cfg); + if (err < 0) + goto errout; + + tb = fib_get_table(cfg.fc_table); + if (tb == NULL) { + err = -ESRCH; + goto errout; + } - tb = fib_get_table(r->rtm_table); - if (tb) - return tb->tb_delete(tb, r, (struct kern_rta*)rta, nlh, &NETLINK_CB(skb)); - return -ESRCH; + err = tb->tb_delete(tb, &cfg); +errout: + return err; } int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg) { - struct fib_table * tb; - struct rtattr **rta = arg; - struct rtmsg *r = NLMSG_DATA(nlh); + struct fib_config cfg; + struct fib_table *tb; + int err; - if (inet_check_attr(r, rta)) - return -EINVAL; + err = rtm_to_fib_config(nlh, &cfg); + if (err < 0) + goto errout; + + tb = fib_new_table(cfg.fc_table); + if (tb == NULL) { + err = -ENOBUFS; + goto errout; + } - tb = fib_new_table(r->rtm_table); - if (tb) - return tb->tb_insert(tb, r, (struct kern_rta*)rta, nlh, &NETLINK_CB(skb)); - return -ENOBUFS; + err = tb->tb_insert(tb, &cfg); +errout: + return err; } int inet_dump_fib(struct sk_buff *skb, struct netlink_callback *cb) @@ -367,17 +593,19 @@ int inet_dump_fib(struct sk_buff *skb, s only when netlink is already locked. */ -static void fib_magic(int cmd, int type, u32 dst, int dst_len, struct in_ifaddr *ifa) +static void fib_magic(int cmd, int type, u32 dst, int dst_len, + struct in_ifaddr *ifa) { - struct fib_table * tb; - struct { - struct nlmsghdr nlh; - struct rtmsg rtm; - } req; - struct kern_rta rta; - - memset(&req.rtm, 0, sizeof(req.rtm)); - memset(&rta, 0, sizeof(rta)); + struct fib_table *tb; + struct fib_config cfg = { + .fc_protocol = RTPROT_KERNEL, + .fc_type = type, + .fc_dst = dst, + .fc_dst_len = dst_len, + .fc_prefsrc = ifa->ifa_local, + .fc_oif = ifa->ifa_dev->dev->ifindex, + .fc_nlmsg_flags = NLM_F_CREATE | NLM_F_APPEND, + }; if (type == RTN_UNICAST) tb = fib_new_table(RT_TABLE_MAIN); @@ -387,26 +615,17 @@ static void fib_magic(int cmd, int type, if (tb == NULL) return; - req.nlh.nlmsg_len = sizeof(req); - req.nlh.nlmsg_type = cmd; - req.nlh.nlmsg_flags = NLM_F_REQUEST|NLM_F_CREATE|NLM_F_APPEND; - req.nlh.nlmsg_pid = 0; - req.nlh.nlmsg_seq = 0; - - req.rtm.rtm_dst_len = dst_len; - req.rtm.rtm_table = tb->tb_id; - req.rtm.rtm_protocol = RTPROT_KERNEL; - req.rtm.rtm_scope = (type != RTN_LOCAL ? RT_SCOPE_LINK : RT_SCOPE_HOST); - req.rtm.rtm_type = type; - - rta.rta_dst = &dst; - rta.rta_prefsrc = &ifa->ifa_local; - rta.rta_oif = &ifa->ifa_dev->dev->ifindex; + cfg.fc_table = tb->tb_id; + + if (type != RTN_LOCAL) + cfg.fc_scope = RT_SCOPE_LINK; + else + cfg.fc_scope = RT_SCOPE_HOST; if (cmd == RTM_NEWROUTE) - tb->tb_insert(tb, &req.rtm, &rta, &req.nlh, NULL); + tb->tb_insert(tb, &cfg); else - tb->tb_delete(tb, &req.rtm, &rta, &req.nlh, NULL); + tb->tb_delete(tb, &cfg); } void fib_add_ifaddr(struct in_ifaddr *ifa) Index: net-2.6.19.git/net/ipv4/fib_semantics.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_semantics.c +++ net-2.6.19.git/net/ipv4/fib_semantics.c @@ -44,6 +44,7 @@ #include <net/sock.h> #include <net/ip_fib.h> #include <net/ip_mp_alg.h> +#include <net/nexthop.h> #include "fib_lookup.h" @@ -273,29 +274,31 @@ int ip_fib_check_default(u32 gw, struct } void rtmsg_fib(int event, u32 key, struct fib_alias *fa, - int z, int tb_id, - struct nlmsghdr *n, struct netlink_skb_parms *req) + int dst_len, int tb_id, struct fib_config *cfg) { struct sk_buff *skb; - u32 pid = req ? req->pid : n->nlmsg_pid; - int size = NLMSG_SPACE(sizeof(struct rtmsg)+256); + u32 pid = cfg->fc_pid; + u32 seq = cfg->fc_nlh ? cfg->fc_nlh->nlmsg_seq : 0; + int payload = sizeof(struct rtmsg) + 256; - skb = alloc_skb(size, GFP_KERNEL); - if (!skb) + skb = nlmsg_new(nlmsg_total_size(payload), GFP_KERNEL); + if (skb == NULL) return; - if (fib_dump_info(skb, pid, n->nlmsg_seq, event, tb_id, - fa->fa_type, fa->fa_scope, &key, z, - fa->fa_tos, - fa->fa_info, 0) < 0) { + if (fib_dump_info(skb, pid, seq, event, tb_id, fa->fa_type, + fa->fa_scope, &key, dst_len, fa->fa_tos, + fa->fa_info, 0) <= 0) { kfree_skb(skb); return; } - NETLINK_CB(skb).dst_group = RTNLGRP_IPV4_ROUTE; - if (n->nlmsg_flags&NLM_F_ECHO) + + if (cfg->fc_nlmsg_flags & NLM_F_ECHO) atomic_inc(&skb->users); + + NETLINK_CB(skb).dst_group = RTNLGRP_IPV4_ROUTE; netlink_broadcast(rtnl, skb, pid, RTNLGRP_IPV4_ROUTE, GFP_KERNEL); - if (n->nlmsg_flags&NLM_F_ECHO) + + if (cfg->fc_nlmsg_flags & NLM_F_ECHO) netlink_unicast(rtnl, skb, pid, MSG_DONTWAIT); } @@ -342,102 +345,100 @@ int fib_detect_death(struct fib_info *fi #ifdef CONFIG_IP_ROUTE_MULTIPATH -static u32 fib_get_attr32(struct rtattr *attr, int attrlen, int type) -{ - while (RTA_OK(attr,attrlen)) { - if (attr->rta_type == type) - return *(u32*)RTA_DATA(attr); - attr = RTA_NEXT(attr, attrlen); - } - return 0; -} - -static int -fib_count_nexthops(struct rtattr *rta) +static int fib_count_nexthops(struct rtnexthop *rtnh, int remaining) { int nhs = 0; - struct rtnexthop *nhp = RTA_DATA(rta); - int nhlen = RTA_PAYLOAD(rta); - while (nhlen >= (int)sizeof(struct rtnexthop)) { - if ((nhlen -= nhp->rtnh_len) < 0) - return 0; + while (rtnh_ok(rtnh, remaining)) { nhs++; - nhp = RTNH_NEXT(nhp); - }; - return nhs; + rtnh = rtnh_next(rtnh, &remaining); + } + + /* leftover implies invalid nexthop configuration, discard it */ + return remaining > 0 ? 0 : nhs; } -static int -fib_get_nhs(struct fib_info *fi, const struct rtattr *rta, const struct rtmsg *r) +static int fib_get_nhs(struct fib_info *fi, struct rtnexthop *rtnh, + int remaining, struct fib_config *cfg) { - struct rtnexthop *nhp = RTA_DATA(rta); - int nhlen = RTA_PAYLOAD(rta); - change_nexthops(fi) { - int attrlen = nhlen - sizeof(struct rtnexthop); - if (attrlen < 0 || (nhlen -= nhp->rtnh_len) < 0) + int attrlen; + + if (!rtnh_ok(rtnh, remaining)) return -EINVAL; - nh->nh_flags = (r->rtm_flags&~0xFF) | nhp->rtnh_flags; - nh->nh_oif = nhp->rtnh_ifindex; - nh->nh_weight = nhp->rtnh_hops + 1; - if (attrlen) { - nh->nh_gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_GATEWAY); + + nh->nh_flags = (cfg->fc_flags & ~0xFF) | rtnh->rtnh_flags; + nh->nh_oif = rtnh->rtnh_ifindex; + nh->nh_weight = rtnh->rtnh_hops + 1; + + attrlen = rtnh_attrlen(rtnh); + if (attrlen > 0) { + struct nlattr *nla, *attrs = rtnh_attrs(rtnh); + + nla = nla_find(attrs, attrlen, RTA_GATEWAY); + nh->nh_gw = nla ? nla_get_u32(nla) : 0; #ifdef CONFIG_NET_CLS_ROUTE - nh->nh_tclassid = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_FLOW); + nla = nla_find(attrs, attrlen, RTA_FLOW); + nh->nh_tclassid = nla ? nla_get_u32(nla) : 0; #endif } - nhp = RTNH_NEXT(nhp); + + rtnh = rtnh_next(rtnh, &remaining); } endfor_nexthops(fi); + return 0; } #endif -int fib_nh_match(struct rtmsg *r, struct nlmsghdr *nlh, struct kern_rta *rta, - struct fib_info *fi) +int fib_nh_match(struct fib_config *cfg, struct fib_info *fi) { #ifdef CONFIG_IP_ROUTE_MULTIPATH - struct rtnexthop *nhp; - int nhlen; + struct rtnexthop *rtnh; + int remaining; #endif - if (rta->rta_priority && - *rta->rta_priority != fi->fib_priority) + if (cfg->fc_priority && cfg->fc_priority != fi->fib_priority) return 1; - if (rta->rta_oif || rta->rta_gw) { - if ((!rta->rta_oif || *rta->rta_oif == fi->fib_nh->nh_oif) && - (!rta->rta_gw || memcmp(rta->rta_gw, &fi->fib_nh->nh_gw, 4) == 0)) + if (cfg->fc_oif || cfg->fc_gw) { + if ((!cfg->fc_oif || cfg->fc_oif == fi->fib_nh->nh_oif) && + (!cfg->fc_gw || cfg->fc_gw == fi->fib_nh->nh_gw)) return 0; return 1; } #ifdef CONFIG_IP_ROUTE_MULTIPATH - if (rta->rta_mp == NULL) + if (cfg->fc_mp == NULL) return 0; - nhp = RTA_DATA(rta->rta_mp); - nhlen = RTA_PAYLOAD(rta->rta_mp); + + rtnh = cfg->fc_mp; + remaining = cfg->fc_mp_len; for_nexthops(fi) { - int attrlen = nhlen - sizeof(struct rtnexthop); - u32 gw; + int attrlen; - if (attrlen < 0 || (nhlen -= nhp->rtnh_len) < 0) + if (!rtnh_ok(rtnh, remaining)) return -EINVAL; - if (nhp->rtnh_ifindex && nhp->rtnh_ifindex != nh->nh_oif) + + if (rtnh->rtnh_ifindex && rtnh->rtnh_ifindex != nh->nh_oif) return 1; - if (attrlen) { - gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_GATEWAY); - if (gw && gw != nh->nh_gw) + + attrlen = rtnh_attrlen(rtnh); + if (attrlen < 0) { + struct nlattr *nla, *attrs = rtnh_attrs(rtnh); + + nla = nla_find(attrs, attrlen, RTA_GATEWAY); + if (nla && nla_get_u32(nla) != nh->nh_gw) return 1; #ifdef CONFIG_NET_CLS_ROUTE - gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_FLOW); - if (gw && gw != nh->nh_tclassid) + nla = nla_find(attrs, attrlen, RTA_FLOW); + if (nla && nla_get_u32(nla) != nh->nh_tclassid) return 1; #endif } - nhp = RTNH_NEXT(nhp); + + rtnh = rtnh_next(rtnh, &remaining); } endfor_nexthops(fi); #endif return 0; @@ -488,7 +489,8 @@ int fib_nh_match(struct rtmsg *r, struct |-> {local prefix} (terminal node) */ -static int fib_check_nh(const struct rtmsg *r, struct fib_info *fi, struct fib_nh *nh) +static int fib_check_nh(struct fib_config *cfg, struct fib_info *fi, + struct fib_nh *nh) { int err; @@ -502,7 +504,7 @@ static int fib_check_nh(const struct rtm if (nh->nh_flags&RTNH_F_ONLINK) { struct net_device *dev; - if (r->rtm_scope >= RT_SCOPE_LINK) + if (cfg->fc_scope >= RT_SCOPE_LINK) return -EINVAL; if (inet_addr_type(nh->nh_gw) != RTN_UNICAST) return -EINVAL; @@ -516,10 +518,15 @@ static int fib_check_nh(const struct rtm return 0; } { - struct flowi fl = { .nl_u = { .ip4_u = - { .daddr = nh->nh_gw, - .scope = r->rtm_scope + 1 } }, - .oif = nh->nh_oif }; + struct flowi fl = { + .nl_u = { + .ip4_u = { + .daddr = nh->nh_gw, + .scope = cfg->fc_scope + 1, + }, + }, + .oif = nh->nh_oif, + }; /* It is not necessary, but requires a bit of thinking */ if (fl.fl4_scope < RT_SCOPE_LINK) @@ -646,39 +653,28 @@ static void fib_hash_move(struct hlist_h fib_hash_free(old_laddrhash, bytes); } -struct fib_info * -fib_create_info(const struct rtmsg *r, struct kern_rta *rta, - const struct nlmsghdr *nlh, int *errp) +struct fib_info *fib_create_info(struct fib_config *cfg) { int err; struct fib_info *fi = NULL; struct fib_info *ofi; -#ifdef CONFIG_IP_ROUTE_MULTIPATH int nhs = 1; -#else - const int nhs = 1; -#endif -#ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED - u32 mp_alg = IP_MP_ALG_NONE; -#endif /* Fast check to catch the most weird cases */ - if (fib_props[r->rtm_type].scope > r->rtm_scope) + if (fib_props[cfg->fc_type].scope > cfg->fc_scope) goto err_inval; #ifdef CONFIG_IP_ROUTE_MULTIPATH - if (rta->rta_mp) { - nhs = fib_count_nexthops(rta->rta_mp); + if (cfg->fc_mp) { + nhs = fib_count_nexthops(cfg->fc_mp, cfg->fc_mp_len); if (nhs == 0) goto err_inval; } #endif #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED - if (rta->rta_mp_alg) { - mp_alg = *rta->rta_mp_alg; - - if (mp_alg < IP_MP_ALG_NONE || - mp_alg > IP_MP_ALG_MAX) + if (cfg->fc_mp_alg) { + if (cfg->fc_mp_alg < IP_MP_ALG_NONE || + cfg->fc_mp_alg > IP_MP_ALG_MAX) goto err_inval; } #endif @@ -714,43 +710,45 @@ fib_create_info(const struct rtmsg *r, s goto failure; fib_info_cnt++; - fi->fib_protocol = r->rtm_protocol; - + fi->fib_protocol = cfg->fc_protocol; fi->fib_nhs = nhs; change_nexthops(fi) { nh->nh_parent = fi; } endfor_nexthops(fi) - fi->fib_flags = r->rtm_flags; - if (rta->rta_priority) - fi->fib_priority = *rta->rta_priority; - if (rta->rta_mx) { - int attrlen = RTA_PAYLOAD(rta->rta_mx); - struct rtattr *attr = RTA_DATA(rta->rta_mx); - - while (RTA_OK(attr, attrlen)) { - unsigned flavor = attr->rta_type; - if (flavor) { - if (flavor > RTAX_MAX) + fi->fib_flags = cfg->fc_flags; + if (cfg->fc_priority) + fi->fib_priority = cfg->fc_priority; + + if (cfg->fc_mx) { + struct nlattr *nla; + int remaining; + + nla_for_each_attr(nla, cfg->fc_mx, cfg->fc_mx_len, remaining) { + int type = nla->nla_type; + + if (type) { + if (type > RTAX_MAX) goto err_inval; - fi->fib_metrics[flavor-1] = *(unsigned*)RTA_DATA(attr); + fi->fib_metrics[type - 1] = nla_get_u32(nla); } - attr = RTA_NEXT(attr, attrlen); } } - if (rta->rta_prefsrc) - memcpy(&fi->fib_prefsrc, rta->rta_prefsrc, 4); - if (rta->rta_mp) { + if (cfg->fc_prefsrc) + fi->fib_prefsrc = cfg->fc_prefsrc; + + if (cfg->fc_mp) { #ifdef CONFIG_IP_ROUTE_MULTIPATH - if ((err = fib_get_nhs(fi, rta->rta_mp, r)) != 0) + err = fib_get_nhs(fi, cfg->fc_mp, cfg->fc_mp_len, cfg); + if (err != 0) goto failure; - if (rta->rta_oif && fi->fib_nh->nh_oif != *rta->rta_oif) + if (cfg->fc_oif && fi->fib_nh->nh_oif != cfg->fc_oif) goto err_inval; - if (rta->rta_gw && memcmp(&fi->fib_nh->nh_gw, rta->rta_gw, 4)) + if (cfg->fc_gw && fi->fib_nh->nh_gw != cfg->fc_gw) goto err_inval; #ifdef CONFIG_NET_CLS_ROUTE - if (rta->rta_flow && memcmp(&fi->fib_nh->nh_tclassid, rta->rta_flow, 4)) + if (cfg->fc_flow && fi->fib_nh->nh_tclassid != cfg->fc_flow) goto err_inval; #endif #else @@ -758,34 +756,34 @@ fib_create_info(const struct rtmsg *r, s #endif } else { struct fib_nh *nh = fi->fib_nh; - if (rta->rta_oif) - nh->nh_oif = *rta->rta_oif; - if (rta->rta_gw) - memcpy(&nh->nh_gw, rta->rta_gw, 4); + if (cfg->fc_oif) + nh->nh_oif = cfg->fc_oif; + if (cfg->fc_gw) + nh->nh_gw = cfg->fc_gw; #ifdef CONFIG_NET_CLS_ROUTE - if (rta->rta_flow) - memcpy(&nh->nh_tclassid, rta->rta_flow, 4); + if (cfg->fc_flow) + nh->nh_tclassid = cfg->fc_flow; #endif - nh->nh_flags = r->rtm_flags; + nh->nh_flags = cfg->fc_flags; #ifdef CONFIG_IP_ROUTE_MULTIPATH nh->nh_weight = 1; #endif } #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED - fi->fib_mp_alg = mp_alg; + fi->fib_mp_alg = cfg->fc_mp_alg; #endif - if (fib_props[r->rtm_type].error) { - if (rta->rta_gw || rta->rta_oif || rta->rta_mp) + if (fib_props[cfg->fc_type].error) { + if (cfg->fc_gw || cfg->fc_oif || cfg->fc_mp) goto err_inval; goto link_it; } - if (r->rtm_scope > RT_SCOPE_HOST) + if (cfg->fc_scope > RT_SCOPE_HOST) goto err_inval; - if (r->rtm_scope == RT_SCOPE_HOST) { + if (cfg->fc_scope == RT_SCOPE_HOST) { struct fib_nh *nh = fi->fib_nh; /* Local address is added. */ @@ -798,14 +796,14 @@ fib_create_info(const struct rtmsg *r, s goto failure; } else { change_nexthops(fi) { - if ((err = fib_check_nh(r, fi, nh)) != 0) + if ((err = fib_check_nh(cfg, fi, nh)) != 0) goto failure; } endfor_nexthops(fi) } if (fi->fib_prefsrc) { - if (r->rtm_type != RTN_LOCAL || rta->rta_dst == NULL || - memcmp(&fi->fib_prefsrc, rta->rta_dst, 4)) + if (cfg->fc_type != RTN_LOCAL || !cfg->fc_dst || + fi->fib_prefsrc != cfg->fc_dst) if (inet_addr_type(fi->fib_prefsrc) != RTN_LOCAL) goto err_inval; } @@ -846,12 +844,12 @@ err_inval: err = -EINVAL; failure: - *errp = err; if (fi) { fi->fib_dead = 1; free_fib_info(fi); } - return NULL; + + return ERR_PTR(err); } /* Note! fib_semantic_match intentionally uses RCU list functions. */ @@ -1011,150 +1009,6 @@ rtattr_failure: return -1; } -#ifndef CONFIG_IP_NOSIOCRT - -int -fib_convert_rtentry(int cmd, struct nlmsghdr *nl, struct rtmsg *rtm, - struct kern_rta *rta, struct rtentry *r) -{ - int plen; - u32 *ptr; - - memset(rtm, 0, sizeof(*rtm)); - memset(rta, 0, sizeof(*rta)); - - if (r->rt_dst.sa_family != AF_INET) - return -EAFNOSUPPORT; - - /* Check mask for validity: - a) it must be contiguous. - b) destination must have all host bits clear. - c) if application forgot to set correct family (AF_INET), - reject request unless it is absolutely clear i.e. - both family and mask are zero. - */ - plen = 32; - ptr = &((struct sockaddr_in*)&r->rt_dst)->sin_addr.s_addr; - if (!(r->rt_flags&RTF_HOST)) { - u32 mask = ((struct sockaddr_in*)&r->rt_genmask)->sin_addr.s_addr; - if (r->rt_genmask.sa_family != AF_INET) { - if (mask || r->rt_genmask.sa_family) - return -EAFNOSUPPORT; - } - if (bad_mask(mask, *ptr)) - return -EINVAL; - plen = inet_mask_len(mask); - } - - nl->nlmsg_flags = NLM_F_REQUEST; - nl->nlmsg_pid = 0; - nl->nlmsg_seq = 0; - nl->nlmsg_len = NLMSG_LENGTH(sizeof(*rtm)); - if (cmd == SIOCDELRT) { - nl->nlmsg_type = RTM_DELROUTE; - nl->nlmsg_flags = 0; - } else { - nl->nlmsg_type = RTM_NEWROUTE; - nl->nlmsg_flags = NLM_F_REQUEST|NLM_F_CREATE; - rtm->rtm_protocol = RTPROT_BOOT; - } - - rtm->rtm_dst_len = plen; - rta->rta_dst = ptr; - - if (r->rt_metric) { - *(u32*)&r->rt_pad3 = r->rt_metric - 1; - rta->rta_priority = (u32*)&r->rt_pad3; - } - if (r->rt_flags&RTF_REJECT) { - rtm->rtm_scope = RT_SCOPE_HOST; - rtm->rtm_type = RTN_UNREACHABLE; - return 0; - } - rtm->rtm_scope = RT_SCOPE_NOWHERE; - rtm->rtm_type = RTN_UNICAST; - - if (r->rt_dev) { - char *colon; - struct net_device *dev; - char devname[IFNAMSIZ]; - - if (copy_from_user(devname, r->rt_dev, IFNAMSIZ-1)) - return -EFAULT; - devname[IFNAMSIZ-1] = 0; - colon = strchr(devname, ':'); - if (colon) - *colon = 0; - dev = __dev_get_by_name(devname); - if (!dev) - return -ENODEV; - rta->rta_oif = &dev->ifindex; - if (colon) { - struct in_ifaddr *ifa; - struct in_device *in_dev = __in_dev_get_rtnl(dev); - if (!in_dev) - return -ENODEV; - *colon = ':'; - for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next) - if (strcmp(ifa->ifa_label, devname) == 0) - break; - if (ifa == NULL) - return -ENODEV; - rta->rta_prefsrc = &ifa->ifa_local; - } - } - - ptr = &((struct sockaddr_in*)&r->rt_gateway)->sin_addr.s_addr; - if (r->rt_gateway.sa_family == AF_INET && *ptr) { - rta->rta_gw = ptr; - if (r->rt_flags&RTF_GATEWAY && inet_addr_type(*ptr) == RTN_UNICAST) - rtm->rtm_scope = RT_SCOPE_UNIVERSE; - } - - if (cmd == SIOCDELRT) - return 0; - - if (r->rt_flags&RTF_GATEWAY && rta->rta_gw == NULL) - return -EINVAL; - - if (rtm->rtm_scope == RT_SCOPE_NOWHERE) - rtm->rtm_scope = RT_SCOPE_LINK; - - if (r->rt_flags&(RTF_MTU|RTF_WINDOW|RTF_IRTT)) { - struct rtattr *rec; - struct rtattr *mx = kmalloc(RTA_LENGTH(3*RTA_LENGTH(4)), GFP_KERNEL); - if (mx == NULL) - return -ENOMEM; - rta->rta_mx = mx; - mx->rta_type = RTA_METRICS; - mx->rta_len = RTA_LENGTH(0); - if (r->rt_flags&RTF_MTU) { - rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len)); - rec->rta_type = RTAX_ADVMSS; - rec->rta_len = RTA_LENGTH(4); - mx->rta_len += RTA_LENGTH(4); - *(u32*)RTA_DATA(rec) = r->rt_mtu - 40; - } - if (r->rt_flags&RTF_WINDOW) { - rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len)); - rec->rta_type = RTAX_WINDOW; - rec->rta_len = RTA_LENGTH(4); - mx->rta_len += RTA_LENGTH(4); - *(u32*)RTA_DATA(rec) = r->rt_window; - } - if (r->rt_flags&RTF_IRTT) { - rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len)); - rec->rta_type = RTAX_RTT; - rec->rta_len = RTA_LENGTH(4); - mx->rta_len += RTA_LENGTH(4); - *(u32*)RTA_DATA(rec) = r->rt_irtt<<3; - } - } - return 0; -} - -#endif - /* Update FIB if: - local address disappeared -> we must delete all the entries Index: net-2.6.19.git/include/net/nexthop.h =================================================================== --- /dev/null +++ net-2.6.19.git/include/net/nexthop.h @@ -0,0 +1,33 @@ +#ifndef __NET_NEXTHOP_H +#define __NET_NEXTHOP_H + +#include <linux/rtnetlink.h> +#include <net/netlink.h> + +static inline int rtnh_ok(const struct rtnexthop *rtnh, int remaining) +{ + return remaining >= sizeof(*rtnh) && + rtnh->rtnh_len >= sizeof(*rtnh) && + rtnh->rtnh_len <= remaining; +} + +static inline struct rtnexthop *rtnh_next(const struct rtnexthop *rtnh, + int *remaining) +{ + int totlen = NLA_ALIGN(rtnh->rtnh_len); + + *remaining -= totlen; + return (struct rtnexthop *) ((char *) rtnh + totlen); +} + +static inline struct nlattr *rtnh_attrs(const struct rtnexthop *rtnh) +{ + return (struct nlattr *) ((char *) rtnh + NLA_ALIGN(sizeof(*rtnh))); +} + +static inline int rtnh_attrlen(const struct rtnexthop *rtnh) +{ + return rtnh->rtnh_len - NLA_ALIGN(sizeof(*rtnh)); +} + +#endif Index: net-2.6.19.git/net/ipv4/fib_hash.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_hash.c +++ net-2.6.19.git/net/ipv4/fib_hash.c @@ -379,42 +379,39 @@ static struct fib_node *fib_find_node(st return NULL; } -static int -fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *n, struct netlink_skb_parms *req) +static int fn_hash_insert(struct fib_table *tb, struct fib_config *cfg) { struct fn_hash *table = (struct fn_hash *) tb->tb_data; struct fib_node *new_f, *f; struct fib_alias *fa, *new_fa; struct fn_zone *fz; struct fib_info *fi; - int z = r->rtm_dst_len; - int type = r->rtm_type; - u8 tos = r->rtm_tos; + u8 tos = cfg->fc_tos; u32 key; int err; - if (z > 32) + if (cfg->fc_dst_len > 32) return -EINVAL; - fz = table->fn_zones[z]; - if (!fz && !(fz = fn_new_zone(table, z))) + + fz = table->fn_zones[cfg->fc_dst_len]; + if (!fz && !(fz = fn_new_zone(table, cfg->fc_dst_len))) return -ENOBUFS; key = 0; - if (rta->rta_dst) { - u32 dst; - memcpy(&dst, rta->rta_dst, 4); - if (dst & ~FZ_MASK(fz)) + if (cfg->fc_dst) { + if (cfg->fc_dst & ~FZ_MASK(fz)) return -EINVAL; - key = fz_key(dst, fz); + key = fz_key(cfg->fc_dst, fz); } - if ((fi = fib_create_info(r, rta, n, &err)) == NULL) - return err; + fi = fib_create_info(cfg); + if (IS_ERR(fi)) + return PTR_ERR(fi); if (fz->fz_nent > (fz->fz_divisor<<1) && fz->fz_divisor < FZ_MAX_DIVISOR && - (z==32 || (1<<z) > fz->fz_divisor)) + (cfg->fc_dst_len == 32 || + (1 << cfg->fc_dst_len) > fz->fz_divisor)) fn_rehash_zone(fz); f = fib_find_node(fz, key); @@ -440,18 +437,18 @@ fn_hash_insert(struct fib_table *tb, str struct fib_alias *fa_orig; err = -EEXIST; - if (n->nlmsg_flags & NLM_F_EXCL) + if (cfg->fc_nlmsg_flags & NLM_F_EXCL) goto out; - if (n->nlmsg_flags & NLM_F_REPLACE) { + if (cfg->fc_nlmsg_flags & NLM_F_REPLACE) { struct fib_info *fi_drop; u8 state; write_lock_bh(&fib_hash_lock); fi_drop = fa->fa_info; fa->fa_info = fi; - fa->fa_type = type; - fa->fa_scope = r->rtm_scope; + fa->fa_type = cfg->fc_type; + fa->fa_scope = cfg->fc_scope; state = fa->fa_state; fa->fa_state &= ~FA_S_ACCESSED; fib_hash_genid++; @@ -474,17 +471,17 @@ fn_hash_insert(struct fib_table *tb, str break; if (fa->fa_info->fib_priority != fi->fib_priority) break; - if (fa->fa_type == type && - fa->fa_scope == r->rtm_scope && + if (fa->fa_type == cfg->fc_type && + fa->fa_scope == cfg->fc_scope && fa->fa_info == fi) goto out; } - if (!(n->nlmsg_flags & NLM_F_APPEND)) + if (!(cfg->fc_nlmsg_flags & NLM_F_APPEND)) fa = fa_orig; } err = -ENOENT; - if (!(n->nlmsg_flags&NLM_F_CREATE)) + if (!(cfg->fc_nlmsg_flags & NLM_F_CREATE)) goto out; err = -ENOBUFS; @@ -506,8 +503,8 @@ fn_hash_insert(struct fib_table *tb, str new_fa->fa_info = fi; new_fa->fa_tos = tos; - new_fa->fa_type = type; - new_fa->fa_scope = r->rtm_scope; + new_fa->fa_type = cfg->fc_type; + new_fa->fa_scope = cfg->fc_scope; new_fa->fa_state = 0; /* @@ -526,7 +523,7 @@ fn_hash_insert(struct fib_table *tb, str fz->fz_nent++; rt_cache_flush(-1); - rtmsg_fib(RTM_NEWROUTE, key, new_fa, z, tb->tb_id, n, req); + rtmsg_fib(RTM_NEWROUTE, key, new_fa, cfg->fc_dst_len, tb->tb_id, cfg); return 0; out_free_new_fa: @@ -537,30 +534,25 @@ out: } -static int -fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *n, struct netlink_skb_parms *req) +static int fn_hash_delete(struct fib_table *tb, struct fib_config *cfg) { struct fn_hash *table = (struct fn_hash*)tb->tb_data; struct fib_node *f; struct fib_alias *fa, *fa_to_delete; - int z = r->rtm_dst_len; struct fn_zone *fz; u32 key; - u8 tos = r->rtm_tos; - if (z > 32) + if (cfg->fc_dst_len > 32) return -EINVAL; - if ((fz = table->fn_zones[z]) == NULL) + + if ((fz = table->fn_zones[cfg->fc_dst_len]) == NULL) return -ESRCH; key = 0; - if (rta->rta_dst) { - u32 dst; - memcpy(&dst, rta->rta_dst, 4); - if (dst & ~FZ_MASK(fz)) + if (cfg->fc_dst) { + if (cfg->fc_dst & ~FZ_MASK(fz)) return -EINVAL; - key = fz_key(dst, fz); + key = fz_key(cfg->fc_dst, fz); } f = fib_find_node(fz, key); @@ -568,7 +560,7 @@ fn_hash_delete(struct fib_table *tb, str if (!f) fa = NULL; else - fa = fib_find_alias(&f->fn_alias, tos, 0); + fa = fib_find_alias(&f->fn_alias, cfg->fc_tos, 0); if (!fa) return -ESRCH; @@ -577,16 +569,16 @@ fn_hash_delete(struct fib_table *tb, str list_for_each_entry_continue(fa, &f->fn_alias, fa_list) { struct fib_info *fi = fa->fa_info; - if (fa->fa_tos != tos) + if (fa->fa_tos != cfg->fc_tos) break; - if ((!r->rtm_type || - fa->fa_type == r->rtm_type) && - (r->rtm_scope == RT_SCOPE_NOWHERE || - fa->fa_scope == r->rtm_scope) && - (!r->rtm_protocol || - fi->fib_protocol == r->rtm_protocol) && - fib_nh_match(r, n, rta, fi) == 0) { + if ((!cfg->fc_type || + fa->fa_type == cfg->fc_type) && + (cfg->fc_scope == RT_SCOPE_NOWHERE || + fa->fa_scope == cfg->fc_scope) && + (!cfg->fc_protocol || + fi->fib_protocol == cfg->fc_protocol) && + fib_nh_match(cfg, fi) == 0) { fa_to_delete = fa; break; } @@ -596,7 +588,8 @@ fn_hash_delete(struct fib_table *tb, str int kill_fn; fa = fa_to_delete; - rtmsg_fib(RTM_DELROUTE, key, fa, z, tb->tb_id, n, req); + rtmsg_fib(RTM_DELROUTE, key, fa, cfg->fc_dst_len, + tb->tb_id, cfg); kill_fn = 0; write_lock_bh(&fib_hash_lock); Index: net-2.6.19.git/net/ipv4/fib_lookup.h =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_lookup.h +++ net-2.6.19.git/net/ipv4/fib_lookup.h @@ -23,19 +23,14 @@ extern int fib_semantic_match(struct lis struct fib_result *res, __u32 zone, __u32 mask, int prefixlen); extern void fib_release_info(struct fib_info *); -extern struct fib_info *fib_create_info(const struct rtmsg *r, - struct kern_rta *rta, - const struct nlmsghdr *, - int *err); -extern int fib_nh_match(struct rtmsg *r, struct nlmsghdr *, - struct kern_rta *rta, struct fib_info *fi); +extern struct fib_info *fib_create_info(struct fib_config *cfg); +extern int fib_nh_match(struct fib_config *cfg, struct fib_info *fi); extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event, u8 tb_id, u8 type, u8 scope, void *dst, int dst_len, u8 tos, struct fib_info *fi, unsigned int); extern void rtmsg_fib(int event, u32 key, struct fib_alias *fa, - int z, int tb_id, - struct nlmsghdr *n, struct netlink_skb_parms *req); + int dst_len, int tb_id, struct fib_config *cfg); extern struct fib_alias *fib_find_alias(struct list_head *fah, u8 tos, u32 prio); extern int fib_detect_death(struct fib_info *fi, int order, Index: net-2.6.19.git/net/ipv4/fib_trie.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_trie.c +++ net-2.6.19.git/net/ipv4/fib_trie.c @@ -1124,17 +1124,14 @@ err: return fa_head; } -static int -fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *nlhdr, struct netlink_skb_parms *req) +static int fn_trie_insert(struct fib_table *tb, struct fib_config *cfg) { struct trie *t = (struct trie *) tb->tb_data; struct fib_alias *fa, *new_fa; struct list_head *fa_head = NULL; struct fib_info *fi; - int plen = r->rtm_dst_len; - int type = r->rtm_type; - u8 tos = r->rtm_tos; + int plen = cfg->fc_dst_len; + u8 tos = cfg->fc_tos; u32 key, mask; int err; struct leaf *l; @@ -1142,11 +1139,7 @@ fn_trie_insert(struct fib_table *tb, str if (plen > 32) return -EINVAL; - key = 0; - if (rta->rta_dst) - memcpy(&key, rta->rta_dst, 4); - - key = ntohl(key); + key = ntohl(cfg->fc_dst); pr_debug("Insert table=%d %08x/%d\n", tb->tb_id, key, plen); @@ -1157,10 +1150,11 @@ fn_trie_insert(struct fib_table *tb, str key = key & mask; - fi = fib_create_info(r, rta, nlhdr, &err); - - if (!fi) + fi = fib_create_info(cfg); + if (IS_ERR(fi)) { + err = PTR_ERR(fi); goto err; + } l = fib_find_node(t, key); fa = NULL; @@ -1185,10 +1179,10 @@ fn_trie_insert(struct fib_table *tb, str struct fib_alias *fa_orig; err = -EEXIST; - if (nlhdr->nlmsg_flags & NLM_F_EXCL) + if (cfg->fc_nlmsg_flags & NLM_F_EXCL) goto out; - if (nlhdr->nlmsg_flags & NLM_F_REPLACE) { + if (cfg->fc_nlmsg_flags & NLM_F_REPLACE) { struct fib_info *fi_drop; u8 state; @@ -1200,8 +1194,8 @@ fn_trie_insert(struct fib_table *tb, str fi_drop = fa->fa_info; new_fa->fa_tos = fa->fa_tos; new_fa->fa_info = fi; - new_fa->fa_type = type; - new_fa->fa_scope = r->rtm_scope; + new_fa->fa_type = cfg->fc_type; + new_fa->fa_scope = cfg->fc_scope; state = fa->fa_state; new_fa->fa_state &= ~FA_S_ACCESSED; @@ -1224,8 +1218,8 @@ fn_trie_insert(struct fib_table *tb, str break; if (fa->fa_info->fib_priority != fi->fib_priority) break; - if (fa->fa_type == type && - fa->fa_scope == r->rtm_scope && + if (fa->fa_type == cfg->fc_type && + fa->fa_scope == cfg->fc_scope && fa->fa_info == fi) { goto out; } @@ -1244,8 +1238,8 @@ fn_trie_insert(struct fib_table *tb, str new_fa->fa_info = fi; new_fa->fa_tos = tos; - new_fa->fa_type = type; - new_fa->fa_scope = r->rtm_scope; + new_fa->fa_type = cfg->fc_type; + new_fa->fa_scope = cfg->fc_scope; new_fa->fa_state = 0; /* * Insert new entry to the list. @@ -1262,7 +1256,7 @@ fn_trie_insert(struct fib_table *tb, str (fa ? &fa->fa_list : fa_head)); rt_cache_flush(-1); - rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id, nlhdr, req); + rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id, cfg); succeeded: return 0; @@ -1548,28 +1542,21 @@ static int trie_leaf_remove(struct trie return 1; } -static int -fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *nlhdr, struct netlink_skb_parms *req) +static int fn_trie_delete(struct fib_table *tb, struct fib_config *cfg) { struct trie *t = (struct trie *) tb->tb_data; u32 key, mask; - int plen = r->rtm_dst_len; - u8 tos = r->rtm_tos; + int plen = cfg->fc_dst_len; + u8 tos = cfg->fc_tos; struct fib_alias *fa, *fa_to_delete; struct list_head *fa_head; struct leaf *l; struct leaf_info *li; - if (plen > 32) return -EINVAL; - key = 0; - if (rta->rta_dst) - memcpy(&key, rta->rta_dst, 4); - - key = ntohl(key); + key = ntohl(cfg->fc_dst); mask = ntohl(inet_make_mask(plen)); if (key & ~mask) @@ -1598,13 +1585,12 @@ fn_trie_delete(struct fib_table *tb, str if (fa->fa_tos != tos) break; - if ((!r->rtm_type || - fa->fa_type == r->rtm_type) && - (r->rtm_scope == RT_SCOPE_NOWHERE || - fa->fa_scope == r->rtm_scope) && - (!r->rtm_protocol || - fi->fib_protocol == r->rtm_protocol) && - fib_nh_match(r, nlhdr, rta, fi) == 0) { + if ((!cfg->fc_type || fa->fa_type == cfg->fc_type) && + (cfg->fc_scope == RT_SCOPE_NOWHERE || + fa->fa_scope == cfg->fc_scope) && + (!cfg->fc_protocol || + fi->fib_protocol == cfg->fc_protocol) && + fib_nh_match(cfg, fi) == 0) { fa_to_delete = fa; break; } @@ -1614,7 +1600,7 @@ fn_trie_delete(struct fib_table *tb, str return -ESRCH; fa = fa_to_delete; - rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, nlhdr, req); + rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, cfg); l = fib_find_node(t, key); li = find_leaf_info(l, plen); ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config 2006-08-07 22:00 ` [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config Thomas Graf @ 2006-08-08 23:11 ` Thomas Graf 2006-08-08 23:12 ` David Miller 0 siblings, 1 reply; 7+ messages in thread From: Thomas Graf @ 2006-08-08 23:11 UTC (permalink / raw) To: davem; +Cc: netdev * Thomas Graf <tgraf@suug.ch> 2006-08-08 00:00 > Introduces struct fib_config replacing the ugly struct kern_rta > prone to ordering issues. Avoids creating faked netlink messages > for auto generated routes or requests via ioctl. > > A new interface net/nexthop.h is added to help navigate through > nexthop configuration arrays. > > Since the netlink source pid of the requesting process is stored > in fib_config, it is no longer required to pass netlink_skb_parms > all along just for event notification. Hold on with this patch, netlink pid handling for netlink events is completely inconsistent. I want to figure out the right way to do it first. ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config 2006-08-08 23:11 ` Thomas Graf @ 2006-08-08 23:12 ` David Miller 0 siblings, 0 replies; 7+ messages in thread From: David Miller @ 2006-08-08 23:12 UTC (permalink / raw) To: tgraf; +Cc: netdev From: Thomas Graf <tgraf@suug.ch> Date: Wed, 9 Aug 2006 01:11:03 +0200 > Hold on with this patch, netlink pid handling for netlink events > is completely inconsistent. I want to figure out the right way > to do it first. Ok. ^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH 2/3] [IPv4]: Convert FIB dumping to use new netlink api 2006-08-08 19:38 [PATCHSET] IPv4 FIB configuration rework Thomas Graf 2006-08-07 22:00 ` [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config Thomas Graf @ 2006-08-07 22:00 ` Thomas Graf 2006-08-07 22:00 ` [PATCH 3/3] [IPv4]: Convert route get to " Thomas Graf 2 siblings, 0 replies; 7+ messages in thread From: Thomas Graf @ 2006-08-07 22:00 UTC (permalink / raw) To: davem; +Cc: netdev [-- Attachment #1: nl_ipv4_fib_dump --] [-- Type: text/plain, Size: 10053 bytes --] Signed-off-by: Thomas Graf <tgraf@suug.ch> Index: net-2.6.19.git/net/ipv4/fib_semantics.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_semantics.c +++ net-2.6.19.git/net/ipv4/fib_semantics.c @@ -286,7 +286,7 @@ void rtmsg_fib(int event, u32 key, struc return; if (fib_dump_info(skb, pid, seq, event, tb_id, fa->fa_type, - fa->fa_scope, &key, dst_len, fa->fa_tos, + fa->fa_scope, key, dst_len, fa->fa_tos, fa->fa_info, 0) <= 0) { kfree_skb(skb); return; @@ -935,17 +935,18 @@ u32 __fib_res_prefsrc(struct fib_result return inet_select_addr(FIB_RES_DEV(*res), FIB_RES_GW(*res), res->scope); } -int -fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event, - u8 tb_id, u8 type, u8 scope, void *dst, int dst_len, u8 tos, - struct fib_info *fi, unsigned int flags) +int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event, + u8 tb_id, u8 type, u8 scope, u32 dst, int dst_len, u8 tos, + struct fib_info *fi, unsigned int flags) { + struct nlmsghdr *nlh; struct rtmsg *rtm; - struct nlmsghdr *nlh; - unsigned char *b = skb->tail; - nlh = NLMSG_NEW(skb, pid, seq, event, sizeof(*rtm), flags); - rtm = NLMSG_DATA(nlh); + nlh = nlmsg_put(skb, pid, seq, event, sizeof(*rtm), flags); + if (nlh == NULL) + return -ENOBUFS; + + rtm = nlmsg_data(nlh); rtm->rtm_family = AF_INET; rtm->rtm_dst_len = dst_len; rtm->rtm_src_len = 0; @@ -954,59 +955,66 @@ fib_dump_info(struct sk_buff *skb, u32 p rtm->rtm_type = type; rtm->rtm_flags = fi->fib_flags; rtm->rtm_scope = scope; - if (rtm->rtm_dst_len) - RTA_PUT(skb, RTA_DST, 4, dst); rtm->rtm_protocol = fi->fib_protocol; + + if (rtm->rtm_dst_len) + NLA_PUT_U32(skb, RTA_DST, dst); + if (fi->fib_priority) - RTA_PUT(skb, RTA_PRIORITY, 4, &fi->fib_priority); + NLA_PUT_U32(skb, RTA_PRIORITY, fi->fib_priority); + if (rtnetlink_put_metrics(skb, fi->fib_metrics) < 0) - goto rtattr_failure; + goto nla_put_failure; + if (fi->fib_prefsrc) - RTA_PUT(skb, RTA_PREFSRC, 4, &fi->fib_prefsrc); + NLA_PUT_U32(skb, RTA_PREFSRC, fi->fib_prefsrc); + if (fi->fib_nhs == 1) { if (fi->fib_nh->nh_gw) - RTA_PUT(skb, RTA_GATEWAY, 4, &fi->fib_nh->nh_gw); + NLA_PUT_U32(skb, RTA_GATEWAY, fi->fib_nh->nh_gw); + if (fi->fib_nh->nh_oif) - RTA_PUT(skb, RTA_OIF, sizeof(int), &fi->fib_nh->nh_oif); + NLA_PUT_U32(skb, RTA_OIF, fi->fib_nh->nh_oif); #ifdef CONFIG_NET_CLS_ROUTE if (fi->fib_nh[0].nh_tclassid) - RTA_PUT(skb, RTA_FLOW, 4, &fi->fib_nh[0].nh_tclassid); + NLA_PUT_U32(skb, RTA_FLOW, fi->fib_nh[0].nh_tclassid); #endif } #ifdef CONFIG_IP_ROUTE_MULTIPATH if (fi->fib_nhs > 1) { - struct rtnexthop *nhp; - struct rtattr *mp_head; - if (skb_tailroom(skb) <= RTA_SPACE(0)) - goto rtattr_failure; - mp_head = (struct rtattr*)skb_put(skb, RTA_SPACE(0)); + struct rtnexthop *rtnh; + struct nlattr *mp; + + mp = nla_nest_start(skb, RTA_MULTIPATH); + if (mp == NULL) + goto nla_put_failure; for_nexthops(fi) { - if (skb_tailroom(skb) < RTA_ALIGN(RTA_ALIGN(sizeof(*nhp)) + 4)) - goto rtattr_failure; - nhp = (struct rtnexthop*)skb_put(skb, RTA_ALIGN(sizeof(*nhp))); - nhp->rtnh_flags = nh->nh_flags & 0xFF; - nhp->rtnh_hops = nh->nh_weight-1; - nhp->rtnh_ifindex = nh->nh_oif; + rtnh = nla_reserve_nohdr(skb, sizeof(*rtnh)); + if (rtnh == NULL) + goto nla_put_failure; + + rtnh->rtnh_flags = nh->nh_flags & 0xFF; + rtnh->rtnh_hops = nh->nh_weight - 1; + rtnh->rtnh_ifindex = nh->nh_oif; + if (nh->nh_gw) - RTA_PUT(skb, RTA_GATEWAY, 4, &nh->nh_gw); + NLA_PUT_U32(skb, RTA_GATEWAY, nh->nh_gw); #ifdef CONFIG_NET_CLS_ROUTE if (nh->nh_tclassid) - RTA_PUT(skb, RTA_FLOW, 4, &nh->nh_tclassid); + NLA_PUT_U32(skb, RTA_FLOW, nh->nh_tclassid); #endif - nhp->rtnh_len = skb->tail - (unsigned char*)nhp; + /* length of rtnetlink header + attributes */ + rtnh->rtnh_len = nlmsg_get_pos(skb) - (void *) rtnh; } endfor_nexthops(fi); - mp_head->rta_type = RTA_MULTIPATH; - mp_head->rta_len = skb->tail - (u8*)mp_head; + + nla_nest_end(skb, mp); } #endif - nlh->nlmsg_len = skb->tail - b; - return skb->len; + return nlmsg_end(skb, nlh); -nlmsg_failure: -rtattr_failure: - skb_trim(skb, b - skb->data); - return -1; +nla_put_failure: + return nlmsg_cancel(skb, nlh); } /* Index: net-2.6.19.git/net/ipv4/fib_hash.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_hash.c +++ net-2.6.19.git/net/ipv4/fib_hash.c @@ -692,7 +692,7 @@ fn_hash_dump_bucket(struct sk_buff *skb, tb->tb_id, fa->fa_type, fa->fa_scope, - &f->fn_key, + f->fn_key, fz->fz_order, fa->fa_tos, fa->fa_info, Index: net-2.6.19.git/net/ipv4/fib_trie.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_trie.c +++ net-2.6.19.git/net/ipv4/fib_trie.c @@ -1852,7 +1852,7 @@ static int fn_trie_dump_fa(t_key key, in tb->tb_id, fa->fa_type, fa->fa_scope, - &xkey, + xkey, plen, fa->fa_tos, fa->fa_info, 0) < 0) { Index: net-2.6.19.git/net/ipv4/fib_lookup.h =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_lookup.h +++ net-2.6.19.git/net/ipv4/fib_lookup.h @@ -26,7 +26,7 @@ extern void fib_release_info(struct fib_ extern struct fib_info *fib_create_info(struct fib_config *cfg); extern int fib_nh_match(struct fib_config *cfg, struct fib_info *fi); extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event, - u8 tb_id, u8 type, u8 scope, void *dst, + u8 tb_id, u8 type, u8 scope, u32 dst, int dst_len, u8 tos, struct fib_info *fi, unsigned int); extern void rtmsg_fib(int event, u32 key, struct fib_alias *fa, Index: net-2.6.19.git/net/ipv4/route.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/route.c +++ net-2.6.19.git/net/ipv4/route.c @@ -2639,14 +2639,14 @@ static int rt_fill_info(struct sk_buff * { struct rtable *rt = (struct rtable*)skb->dst; struct rtmsg *r; - struct nlmsghdr *nlh; - unsigned char *b = skb->tail; + struct nlmsghdr *nlh; struct rta_cacheinfo ci; -#ifdef CONFIG_IP_MROUTE - struct rtattr *eptr; -#endif - nlh = NLMSG_NEW(skb, pid, seq, event, sizeof(*r), flags); - r = NLMSG_DATA(nlh); + + nlh = nlmsg_put(skb, pid, seq, event, sizeof(*r), flags); + if (nlh == NULL) + return -ENOBUFS; + + r = nlmsg_data(nlh); r->rtm_family = AF_INET; r->rtm_dst_len = 32; r->rtm_src_len = 0; @@ -2658,32 +2658,34 @@ static int rt_fill_info(struct sk_buff * r->rtm_flags = (rt->rt_flags & ~0xFFFF) | RTM_F_CLONED; if (rt->rt_flags & RTCF_NOTIFY) r->rtm_flags |= RTM_F_NOTIFY; - RTA_PUT(skb, RTA_DST, 4, &rt->rt_dst); + + NLA_PUT_U32(skb, RTA_DST, rt->rt_dst); + if (rt->fl.fl4_src) { r->rtm_src_len = 32; - RTA_PUT(skb, RTA_SRC, 4, &rt->fl.fl4_src); + NLA_PUT_U32(skb, RTA_SRC, rt->fl.fl4_src); } if (rt->u.dst.dev) - RTA_PUT(skb, RTA_OIF, sizeof(int), &rt->u.dst.dev->ifindex); + NLA_PUT_U32(skb, RTA_OIF, rt->u.dst.dev->ifindex); #ifdef CONFIG_NET_CLS_ROUTE if (rt->u.dst.tclassid) - RTA_PUT(skb, RTA_FLOW, 4, &rt->u.dst.tclassid); + NLA_PUT_U32(skb, RTA_FLOW, rt->u.dst.tclassid); #endif #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED - if (rt->rt_multipath_alg != IP_MP_ALG_NONE) { - __u32 alg = rt->rt_multipath_alg; - - RTA_PUT(skb, RTA_MP_ALGO, 4, &alg); - } + if (rt->rt_multipath_alg != IP_MP_ALG_NONE) + NLA_PUT_U32(skb, RTA_MP_ALGO, rt->rt_multipath_alg); #endif if (rt->fl.iif) - RTA_PUT(skb, RTA_PREFSRC, 4, &rt->rt_spec_dst); + NLA_PUT_U32(skb, RTA_PREFSRC, rt->rt_spec_dst); else if (rt->rt_src != rt->fl.fl4_src) - RTA_PUT(skb, RTA_PREFSRC, 4, &rt->rt_src); + NLA_PUT_U32(skb, RTA_PREFSRC, rt->rt_src); + if (rt->rt_dst != rt->rt_gateway) - RTA_PUT(skb, RTA_GATEWAY, 4, &rt->rt_gateway); + NLA_PUT_U32(skb, RTA_GATEWAY, rt->rt_gateway); + if (rtnetlink_put_metrics(skb, rt->u.dst.metrics) < 0) - goto rtattr_failure; + goto nla_put_failure; + ci.rta_lastuse = jiffies_to_clock_t(jiffies - rt->u.dst.lastuse); ci.rta_used = rt->u.dst.__use; ci.rta_clntref = atomic_read(&rt->u.dst.__refcnt); @@ -2700,10 +2702,7 @@ static int rt_fill_info(struct sk_buff * ci.rta_tsage = xtime.tv_sec - rt->peer->tcp_ts_stamp; } } -#ifdef CONFIG_IP_MROUTE - eptr = (struct rtattr*)skb->tail; -#endif - RTA_PUT(skb, RTA_CACHEINFO, sizeof(ci), &ci); + if (rt->fl.iif) { #ifdef CONFIG_IP_MROUTE u32 dst = rt->rt_dst; @@ -2715,25 +2714,24 @@ static int rt_fill_info(struct sk_buff * if (!nowait) { if (err == 0) return 0; - goto nlmsg_failure; + goto nla_put_failure; } else { if (err == -EMSGSIZE) - goto nlmsg_failure; - ((struct rta_cacheinfo*)RTA_DATA(eptr))->rta_error = err; + goto nla_put_failure; + ci.rta_error = err; } } } else #endif - RTA_PUT(skb, RTA_IIF, sizeof(int), &rt->fl.iif); + NLA_PUT_U32(skb, RTA_IIF, rt->fl.iif); } - nlh->nlmsg_len = skb->tail - b; - return skb->len; + NLA_PUT(skb, RTA_CACHEINFO, sizeof(ci), &ci); + + return nlmsg_end(skb, nlh); -nlmsg_failure: -rtattr_failure: - skb_trim(skb, b - skb->data); - return -1; +nla_put_failure: + return nlmsg_cancel(skb, nlh); } int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg) Index: net-2.6.19.git/net/ipv4/fib_frontend.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_frontend.c +++ net-2.6.19.git/net/ipv4/fib_frontend.c @@ -563,8 +563,8 @@ int inet_dump_fib(struct sk_buff *skb, s int s_t; struct fib_table *tb; - if (NLMSG_PAYLOAD(cb->nlh, 0) >= sizeof(struct rtmsg) && - ((struct rtmsg*)NLMSG_DATA(cb->nlh))->rtm_flags&RTM_F_CLONED) + if (nlmsg_len(cb->nlh) >= sizeof(struct rtmsg) && + ((struct rtmsg *) nlmsg_data(cb->nlh))->rtm_flags & RTM_F_CLONED) return ip_rt_dump(skb, cb); s_t = cb->args[0]; ^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH 3/3] [IPv4]: Convert route get to new netlink api 2006-08-08 19:38 [PATCHSET] IPv4 FIB configuration rework Thomas Graf 2006-08-07 22:00 ` [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config Thomas Graf 2006-08-07 22:00 ` [PATCH 2/3] [IPv4]: Convert FIB dumping to use new netlink api Thomas Graf @ 2006-08-07 22:00 ` Thomas Graf 2 siblings, 0 replies; 7+ messages in thread From: Thomas Graf @ 2006-08-07 22:00 UTC (permalink / raw) To: davem; +Cc: netdev [-- Attachment #1: nl_ipv4_get_route --] [-- Type: text/plain, Size: 4526 bytes --] Fixes various unvalidated netlink attributes causing memory corruptions when left empty by userspace applications. Signed-off-by: Thomas Graf <tgraf@suug.ch> Index: net-2.6.19.git/include/net/ip_fib.h =================================================================== --- net-2.6.19.git.orig/include/net/ip_fib.h +++ net-2.6.19.git/include/net/ip_fib.h @@ -245,6 +245,7 @@ extern void fib_select_default(const str #endif /* CONFIG_IP_MULTIPLE_TABLES */ /* Exported by fib_frontend.c */ +extern struct nla_policy rtm_ipv4_policy[]; extern void ip_fib_init(void); extern int inet_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg); extern int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg); Index: net-2.6.19.git/net/ipv4/fib_frontend.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_frontend.c +++ net-2.6.19.git/net/ipv4/fib_frontend.c @@ -433,7 +433,7 @@ int ip_rt_ioctl(unsigned int cmd, void * #endif -static struct nla_policy rtm_ipv4_policy[RTA_MAX+1] __read_mostly = { +struct nla_policy rtm_ipv4_policy[RTA_MAX+1] __read_mostly = { [RTA_DST] = { .type = NLA_U32 }, [RTA_SRC] = { .type = NLA_U32 }, [RTA_IIF] = { .type = NLA_U32 }, Index: net-2.6.19.git/net/ipv4/route.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/route.c +++ net-2.6.19.git/net/ipv4/route.c @@ -2736,18 +2736,24 @@ nla_put_failure: int inet_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg) { - struct rtattr **rta = arg; - struct rtmsg *rtm = NLMSG_DATA(nlh); + struct rtmsg *rtm; + struct nlattr *tb[RTA_MAX+1]; struct rtable *rt = NULL; - u32 dst = 0; - u32 src = 0; - int iif = 0; - int err = -ENOBUFS; + u32 dst, src, iif; + int err; struct sk_buff *skb; + err = nlmsg_parse(nlh, sizeof(*rtm), tb, RTA_MAX, rtm_ipv4_policy); + if (err < 0) + goto errout; + + rtm = nlmsg_data(nlh); + skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL); - if (!skb) - goto out; + if (skb == NULL) { + err = -ENOBUFS; + goto errout; + } /* Reserve room for dummy headers, this skb can pass through good chunk of routing engine. @@ -2758,62 +2764,64 @@ int inet_rtm_getroute(struct sk_buff *in skb->nh.iph->protocol = IPPROTO_ICMP; skb_reserve(skb, MAX_HEADER + sizeof(struct iphdr)); - if (rta[RTA_SRC - 1]) - memcpy(&src, RTA_DATA(rta[RTA_SRC - 1]), 4); - if (rta[RTA_DST - 1]) - memcpy(&dst, RTA_DATA(rta[RTA_DST - 1]), 4); - if (rta[RTA_IIF - 1]) - memcpy(&iif, RTA_DATA(rta[RTA_IIF - 1]), sizeof(int)); + src = tb[RTA_SRC] ? nla_get_u32(tb[RTA_SRC]) : 0; + dst = tb[RTA_DST] ? nla_get_u32(tb[RTA_DST]) : 0; + iif = tb[RTA_IIF] ? nla_get_u32(tb[RTA_IIF]) : 0; if (iif) { - struct net_device *dev = __dev_get_by_index(iif); - err = -ENODEV; - if (!dev) - goto out_free; + struct net_device *dev; + + dev = __dev_get_by_index(iif); + if (dev == NULL) { + err = -ENODEV; + goto errout_free; + } + skb->protocol = htons(ETH_P_IP); skb->dev = dev; + local_bh_disable(); err = ip_route_input(skb, dst, src, rtm->rtm_tos, dev); local_bh_enable(); - rt = (struct rtable*)skb->dst; - if (!err && rt->u.dst.error) + + rt = (struct rtable*) skb->dst; + if (err == 0 && rt->u.dst.error) err = -rt->u.dst.error; } else { - struct flowi fl = { .nl_u = { .ip4_u = { .daddr = dst, - .saddr = src, - .tos = rtm->rtm_tos } } }; - int oif = 0; - if (rta[RTA_OIF - 1]) - memcpy(&oif, RTA_DATA(rta[RTA_OIF - 1]), sizeof(int)); - fl.oif = oif; + struct flowi fl = { + .nl_u = { + .ip4_u = { + .daddr = dst, + .saddr = src, + .tos = rtm->rtm_tos, + }, + }, + .oif = tb[RTA_OIF] ? nla_get_u32(tb[RTA_OIF]) : 0, + }; + err = ip_route_output_key(&rt, &fl); } if (err) - goto out_free; + goto errout_free; skb->dst = &rt->u.dst; if (rtm->rtm_flags & RTM_F_NOTIFY) rt->rt_flags |= RTCF_NOTIFY; - NETLINK_CB(skb).dst_pid = NETLINK_CB(in_skb).pid; - err = rt_fill_info(skb, NETLINK_CB(in_skb).pid, nlh->nlmsg_seq, RTM_NEWROUTE, 0, 0); - if (!err) - goto out_free; - if (err < 0) { - err = -EMSGSIZE; - goto out_free; - } + if (err <= 0) + goto errout_free; err = netlink_unicast(rtnl, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT); if (err > 0) err = 0; -out: return err; +errout: + return err; -out_free: +errout_free: kfree_skb(skb); - goto out; + goto errout; } int ip_rt_dump(struct sk_buff *skb, struct netlink_callback *cb) ^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCHSET] IPv4 FIB configuration conversions @ 2006-08-15 14:57 Thomas Graf 2006-08-14 22:00 ` [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config Thomas Graf 0 siblings, 1 reply; 7+ messages in thread From: Thomas Graf @ 2006-08-15 14:57 UTC (permalink / raw) To: davem; +Cc: netdev Replaces the rather ugly struct kern_rta + rtmsg with a new struct fib_config while converting everything to the new netlink api. Simplifies the FIB module interface quite a bit and allows changing internals while keeping a stable netlink interface. Gets rid of things like passing on netlink_skb_parms all around the fib code etc. Also fixes some memory corruptions that could be triggered via netlink messages, not sure which would actually lead to an oops though. ^ permalink raw reply [flat|nested] 7+ messages in thread
* [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config 2006-08-15 14:57 [PATCHSET] IPv4 FIB configuration conversions Thomas Graf @ 2006-08-14 22:00 ` Thomas Graf 0 siblings, 0 replies; 7+ messages in thread From: Thomas Graf @ 2006-08-14 22:00 UTC (permalink / raw) To: davem; +Cc: netdev [-- Attachment #1: nl_ipv4_route_fib_config --] [-- Type: text/plain, Size: 43626 bytes --] Introduces struct fib_config replacing the ugly struct kern_rta prone to ordering issues. Avoids creating faked netlink messages for auto generated routes or requests via ioctl. A new interface net/nexthop.h is added to help navigate through nexthop configuration arrays. A new struct nl_info will be used to carry the necessary netlink information to be used for notifications later on. Signed-off-by: Thomas Graf <tgraf@suug.ch> Index: net-2.6.19.git/include/net/ip_fib.h =================================================================== --- net-2.6.19.git.orig/include/net/ip_fib.h +++ net-2.6.19.git/include/net/ip_fib.h @@ -20,25 +20,32 @@ #include <linux/seq_file.h> #include <net/fib_rules.h> -/* WARNING: The ordering of these elements must match ordering - * of RTA_* rtnetlink attribute numbers. - */ -struct kern_rta { - void *rta_dst; - void *rta_src; - int *rta_iif; - int *rta_oif; - void *rta_gw; - u32 *rta_priority; - void *rta_prefsrc; - struct rtattr *rta_mx; - struct rtattr *rta_mp; - unsigned char *rta_protoinfo; - u32 *rta_flow; - struct rta_cacheinfo *rta_ci; - struct rta_session *rta_sess; - u32 *rta_mp_alg; -}; +struct fib_config { + u8 fc_family; + u8 fc_dst_len; + u8 fc_src_len; + u8 fc_tos; + u8 fc_protocol; + u8 fc_scope; + u8 fc_type; + /* 1 byte unused */ + u32 fc_table; + u32 fc_dst; + u32 fc_src; + u32 fc_gw; + int fc_oif; + u32 fc_flags; + u32 fc_priority; + u32 fc_prefsrc; + struct nlattr *fc_mx; + struct rtnexthop *fc_mp; + int fc_mx_len; + int fc_mp_len; + u32 fc_flow; + u32 fc_mp_alg; + u32 fc_nlflags; + struct nl_info fc_nlinfo; + }; struct fib_info; @@ -154,12 +161,8 @@ struct fib_table { u32 tb_id; unsigned tb_stamp; int (*tb_lookup)(struct fib_table *tb, const struct flowi *flp, struct fib_result *res); - int (*tb_insert)(struct fib_table *table, struct rtmsg *r, - struct kern_rta *rta, struct nlmsghdr *n, - struct netlink_skb_parms *req); - int (*tb_delete)(struct fib_table *table, struct rtmsg *r, - struct kern_rta *rta, struct nlmsghdr *n, - struct netlink_skb_parms *req); + int (*tb_insert)(struct fib_table *, struct fib_config *); + int (*tb_delete)(struct fib_table *, struct fib_config *); int (*tb_dump)(struct fib_table *table, struct sk_buff *skb, struct netlink_callback *cb); int (*tb_flush)(struct fib_table *table); @@ -228,8 +231,6 @@ struct rtentry; extern int ip_fib_check_default(u32 gw, struct net_device *dev); extern int fib_sync_down(u32 local, struct net_device *dev, int force); extern int fib_sync_up(struct net_device *dev); -extern int fib_convert_rtentry(int cmd, struct nlmsghdr *nl, struct rtmsg *rtm, - struct kern_rta *rta, struct rtentry *r); extern u32 __fib_res_prefsrc(struct fib_result *res); /* Exported by fib_hash.c */ Index: net-2.6.19.git/net/ipv4/fib_frontend.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_frontend.c +++ net-2.6.19.git/net/ipv4/fib_frontend.c @@ -253,42 +253,190 @@ e_inval: #ifndef CONFIG_IP_NOSIOCRT +static inline u32 sk_extract_addr(struct sockaddr *addr) +{ + return ((struct sockaddr_in *) addr)->sin_addr.s_addr; +} + +static int put_rtax(struct nlattr *mx, int len, int type, u32 value) +{ + struct nlattr *nla; + + nla = (struct nlattr *) ((char *) mx + len); + nla->nla_type = type; + nla->nla_len = nla_attr_size(4); + *(u32 *) nla_data(nla) = value; + + return len + nla_total_size(4); +} + +static int rtentry_to_fib_config(int cmd, struct rtentry *rt, + struct fib_config *cfg) +{ + u32 addr; + int plen; + + memset(cfg, 0, sizeof(*cfg)); + + if (rt->rt_dst.sa_family != AF_INET) + return -EAFNOSUPPORT; + + /* + * Check mask for validity: + * a) it must be contiguous. + * b) destination must have all host bits clear. + * c) if application forgot to set correct family (AF_INET), + * reject request unless it is absolutely clear i.e. + * both family and mask are zero. + */ + plen = 32; + addr = sk_extract_addr(&rt->rt_dst); + if (!(rt->rt_flags & RTF_HOST)) { + u32 mask = sk_extract_addr(&rt->rt_genmask); + + if (rt->rt_genmask.sa_family != AF_INET) { + if (mask || rt->rt_genmask.sa_family) + return -EAFNOSUPPORT; + } + + if (bad_mask(mask, addr)) + return -EINVAL; + + plen = inet_mask_len(mask); + } + + cfg->fc_dst_len = plen; + cfg->fc_dst = addr; + + if (cmd != SIOCDELRT) { + cfg->fc_nlflags = NLM_F_CREATE; + cfg->fc_protocol = RTPROT_BOOT; + } + + if (rt->rt_metric) + cfg->fc_priority = rt->rt_metric - 1; + + if (rt->rt_flags & RTF_REJECT) { + cfg->fc_scope = RT_SCOPE_HOST; + cfg->fc_type = RTN_UNREACHABLE; + return 0; + } + + cfg->fc_scope = RT_SCOPE_NOWHERE; + cfg->fc_type = RTN_UNICAST; + + if (rt->rt_dev) { + char *colon; + struct net_device *dev; + char devname[IFNAMSIZ]; + + if (copy_from_user(devname, rt->rt_dev, IFNAMSIZ-1)) + return -EFAULT; + + devname[IFNAMSIZ-1] = 0; + colon = strchr(devname, ':'); + if (colon) + *colon = 0; + dev = __dev_get_by_name(devname); + if (!dev) + return -ENODEV; + cfg->fc_oif = dev->ifindex; + if (colon) { + struct in_ifaddr *ifa; + struct in_device *in_dev = __in_dev_get_rtnl(dev); + if (!in_dev) + return -ENODEV; + *colon = ':'; + for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next) + if (strcmp(ifa->ifa_label, devname) == 0) + break; + if (ifa == NULL) + return -ENODEV; + cfg->fc_prefsrc = ifa->ifa_local; + } + } + + addr = sk_extract_addr(&rt->rt_gateway); + if (rt->rt_gateway.sa_family == AF_INET && addr) { + cfg->fc_gw = addr; + if (rt->rt_flags & RTF_GATEWAY && + inet_addr_type(addr) == RTN_UNICAST) + cfg->fc_scope = RT_SCOPE_UNIVERSE; + } + + if (cmd == SIOCDELRT) + return 0; + + if (rt->rt_flags & RTF_GATEWAY && !cfg->fc_gw) + return -EINVAL; + + if (cfg->fc_scope == RT_SCOPE_NOWHERE) + cfg->fc_scope = RT_SCOPE_LINK; + + if (rt->rt_flags & (RTF_MTU | RTF_WINDOW | RTF_IRTT)) { + struct nlattr *mx; + int len = 0; + + mx = kzalloc(3 * nla_total_size(4), GFP_KERNEL); + if (mx == NULL) + return -ENOMEM; + + if (rt->rt_flags & RTF_MTU) + len = put_rtax(mx, len, RTAX_ADVMSS, rt->rt_mtu - 40); + + if (rt->rt_flags & RTF_WINDOW) + len = put_rtax(mx, len, RTAX_WINDOW, rt->rt_window); + + if (rt->rt_flags & RTF_IRTT) + len = put_rtax(mx, len, RTAX_RTT, rt->rt_irtt << 3); + + cfg->fc_mx = mx; + cfg->fc_mx_len = len; + } + + return 0; +} + /* * Handle IP routing ioctl calls. These are used to manipulate the routing tables */ int ip_rt_ioctl(unsigned int cmd, void __user *arg) { + struct fib_config cfg; + struct rtentry rt; int err; - struct kern_rta rta; - struct rtentry r; - struct { - struct nlmsghdr nlh; - struct rtmsg rtm; - } req; switch (cmd) { case SIOCADDRT: /* Add a route */ case SIOCDELRT: /* Delete a route */ if (!capable(CAP_NET_ADMIN)) return -EPERM; - if (copy_from_user(&r, arg, sizeof(struct rtentry))) + + if (copy_from_user(&rt, arg, sizeof(rt))) return -EFAULT; + rtnl_lock(); - err = fib_convert_rtentry(cmd, &req.nlh, &req.rtm, &rta, &r); + err = rtentry_to_fib_config(cmd, &rt, &cfg); if (err == 0) { + struct fib_table *tb; + if (cmd == SIOCDELRT) { - struct fib_table *tb = fib_get_table(req.rtm.rtm_table); - err = -ESRCH; + tb = fib_get_table(cfg.fc_table); if (tb) - err = tb->tb_delete(tb, &req.rtm, &rta, &req.nlh, NULL); + err = tb->tb_delete(tb, &cfg); + else + err = -ESRCH; } else { - struct fib_table *tb = fib_new_table(req.rtm.rtm_table); - err = -ENOBUFS; + tb = fib_new_table(cfg.fc_table); if (tb) - err = tb->tb_insert(tb, &req.rtm, &rta, &req.nlh, NULL); + err = tb->tb_insert(tb, &cfg); + else + err = -ENOBUFS; } - kfree(rta.rta_mx); + + /* allocated by rtentry_to_fib_config() */ + kfree(cfg.fc_mx); } rtnl_unlock(); return err; @@ -305,51 +453,134 @@ int ip_rt_ioctl(unsigned int cmd, void * #endif -static int inet_check_attr(struct rtmsg *r, struct rtattr **rta) -{ - int i; +static struct nla_policy rtm_ipv4_policy[RTA_MAX+1] __read_mostly = { + [RTA_DST] = { .type = NLA_U32 }, + [RTA_SRC] = { .type = NLA_U32 }, + [RTA_IIF] = { .type = NLA_U32 }, + [RTA_OIF] = { .type = NLA_U32 }, + [RTA_GATEWAY] = { .type = NLA_U32 }, + [RTA_PRIORITY] = { .type = NLA_U32 }, + [RTA_PREFSRC] = { .type = NLA_U32 }, + [RTA_METRICS] = { .type = NLA_NESTED }, + [RTA_MULTIPATH] = { .minlen = sizeof(struct rtnexthop) }, + [RTA_PROTOINFO] = { .type = NLA_U32 }, + [RTA_FLOW] = { .type = NLA_U32 }, + [RTA_MP_ALGO] = { .type = NLA_U32 }, +}; - for (i=1; i<=RTA_MAX; i++, rta++) { - struct rtattr *attr = *rta; - if (attr) { - if (RTA_PAYLOAD(attr) < 4) - return -EINVAL; - if (i != RTA_MULTIPATH && i != RTA_METRICS && - i != RTA_TABLE) - *rta = (struct rtattr*)RTA_DATA(attr); +static int rtm_to_fib_config(struct sk_buff *skb, struct nlmsghdr *nlh, + struct fib_config *cfg) +{ + struct nlattr *attr; + int err, remaining; + struct rtmsg *rtm; + + err = nlmsg_validate(nlh, sizeof(*rtm), RTA_MAX, rtm_ipv4_policy); + if (err < 0) + goto errout; + + memset(cfg, 0, sizeof(*cfg)); + + rtm = nlmsg_data(nlh); + cfg->fc_family = rtm->rtm_family; + cfg->fc_dst_len = rtm->rtm_dst_len; + cfg->fc_src_len = rtm->rtm_src_len; + cfg->fc_tos = rtm->rtm_tos; + cfg->fc_table = rtm->rtm_table; + cfg->fc_protocol = rtm->rtm_protocol; + cfg->fc_scope = rtm->rtm_scope; + cfg->fc_type = rtm->rtm_type; + cfg->fc_flags = rtm->rtm_flags; + cfg->fc_nlflags = nlh->nlmsg_flags; + + cfg->fc_nlinfo.pid = NETLINK_CB(skb).pid; + cfg->fc_nlinfo.nlh = nlh; + + nlmsg_for_each_attr(attr, nlh, sizeof(struct rtmsg), remaining) { + switch (attr->nla_type) { + case RTA_DST: + cfg->fc_dst = nla_get_u32(attr); + break; + case RTA_SRC: + cfg->fc_src = nla_get_u32(attr); + break; + case RTA_OIF: + cfg->fc_oif = nla_get_u32(attr); + break; + case RTA_GATEWAY: + cfg->fc_gw = nla_get_u32(attr); + break; + case RTA_PRIORITY: + cfg->fc_priority = nla_get_u32(attr); + break; + case RTA_PREFSRC: + cfg->fc_prefsrc = nla_get_u32(attr); + break; + case RTA_METRICS: + cfg->fc_mx = nla_data(attr); + cfg->fc_mx_len = nla_len(attr); + break; + case RTA_MULTIPATH: + cfg->fc_mp = nla_data(attr); + cfg->fc_mp_len = nla_len(attr); + break; + case RTA_FLOW: + cfg->fc_flow = nla_get_u32(attr); + break; + case RTA_MP_ALGO: + cfg->fc_mp_alg = nla_get_u32(attr); + break; + case RTA_TABLE: + cfg->fc_table = nla_get_u32(attr); + break; } } + return 0; +errout: + return err; } int inet_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg) { - struct fib_table * tb; - struct rtattr **rta = arg; - struct rtmsg *r = NLMSG_DATA(nlh); + struct fib_config cfg; + struct fib_table *tb; + int err; - if (inet_check_attr(r, rta)) - return -EINVAL; + err = rtm_to_fib_config(skb, nlh, &cfg); + if (err < 0) + goto errout; + + tb = fib_get_table(cfg.fc_table); + if (tb == NULL) { + err = -ESRCH; + goto errout; + } - tb = fib_get_table(rtm_get_table(rta, r->rtm_table)); - if (tb) - return tb->tb_delete(tb, r, (struct kern_rta*)rta, nlh, &NETLINK_CB(skb)); - return -ESRCH; + err = tb->tb_delete(tb, &cfg); +errout: + return err; } int inet_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg) { - struct fib_table * tb; - struct rtattr **rta = arg; - struct rtmsg *r = NLMSG_DATA(nlh); + struct fib_config cfg; + struct fib_table *tb; + int err; - if (inet_check_attr(r, rta)) - return -EINVAL; + err = rtm_to_fib_config(skb, nlh, &cfg); + if (err < 0) + goto errout; + + tb = fib_new_table(cfg.fc_table); + if (tb == NULL) { + err = -ENOBUFS; + goto errout; + } - tb = fib_new_table(rtm_get_table(rta, r->rtm_table)); - if (tb) - return tb->tb_insert(tb, r, (struct kern_rta*)rta, nlh, &NETLINK_CB(skb)); - return -ENOBUFS; + err = tb->tb_insert(tb, &cfg); +errout: + return err; } int inet_dump_fib(struct sk_buff *skb, struct netlink_callback *cb) @@ -396,17 +627,19 @@ out: only when netlink is already locked. */ -static void fib_magic(int cmd, int type, u32 dst, int dst_len, struct in_ifaddr *ifa) +static void fib_magic(int cmd, int type, u32 dst, int dst_len, + struct in_ifaddr *ifa) { - struct fib_table * tb; - struct { - struct nlmsghdr nlh; - struct rtmsg rtm; - } req; - struct kern_rta rta; - - memset(&req.rtm, 0, sizeof(req.rtm)); - memset(&rta, 0, sizeof(rta)); + struct fib_table *tb; + struct fib_config cfg = { + .fc_protocol = RTPROT_KERNEL, + .fc_type = type, + .fc_dst = dst, + .fc_dst_len = dst_len, + .fc_prefsrc = ifa->ifa_local, + .fc_oif = ifa->ifa_dev->dev->ifindex, + .fc_nlflags = NLM_F_CREATE | NLM_F_APPEND, + }; if (type == RTN_UNICAST) tb = fib_new_table(RT_TABLE_MAIN); @@ -416,26 +649,17 @@ static void fib_magic(int cmd, int type, if (tb == NULL) return; - req.nlh.nlmsg_len = sizeof(req); - req.nlh.nlmsg_type = cmd; - req.nlh.nlmsg_flags = NLM_F_REQUEST|NLM_F_CREATE|NLM_F_APPEND; - req.nlh.nlmsg_pid = 0; - req.nlh.nlmsg_seq = 0; - - req.rtm.rtm_dst_len = dst_len; - req.rtm.rtm_table = tb->tb_id; - req.rtm.rtm_protocol = RTPROT_KERNEL; - req.rtm.rtm_scope = (type != RTN_LOCAL ? RT_SCOPE_LINK : RT_SCOPE_HOST); - req.rtm.rtm_type = type; - - rta.rta_dst = &dst; - rta.rta_prefsrc = &ifa->ifa_local; - rta.rta_oif = &ifa->ifa_dev->dev->ifindex; + cfg.fc_table = tb->tb_id; + + if (type != RTN_LOCAL) + cfg.fc_scope = RT_SCOPE_LINK; + else + cfg.fc_scope = RT_SCOPE_HOST; if (cmd == RTM_NEWROUTE) - tb->tb_insert(tb, &req.rtm, &rta, &req.nlh, NULL); + tb->tb_insert(tb, &cfg); else - tb->tb_delete(tb, &req.rtm, &rta, &req.nlh, NULL); + tb->tb_delete(tb, &cfg); } void fib_add_ifaddr(struct in_ifaddr *ifa) Index: net-2.6.19.git/net/ipv4/fib_semantics.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_semantics.c +++ net-2.6.19.git/net/ipv4/fib_semantics.c @@ -44,6 +44,7 @@ #include <net/ip_fib.h> #include <net/ip_mp_alg.h> #include <net/netlink.h> +#include <net/nexthop.h> #include "fib_lookup.h" @@ -273,27 +274,27 @@ int ip_fib_check_default(u32 gw, struct } void rtmsg_fib(int event, u32 key, struct fib_alias *fa, - int z, u32 tb_id, - struct nlmsghdr *n, struct netlink_skb_parms *req) + int dst_len, u32 tb_id, struct nl_info *info) { struct sk_buff *skb; - u32 pid = req ? req->pid : n->nlmsg_pid; int payload = sizeof(struct rtmsg) + 256; + u32 seq = info->nlh ? info->nlh->nlmsg_seq : 0; int err = -ENOBUFS; skb = nlmsg_new(nlmsg_total_size(payload), GFP_KERNEL); if (skb == NULL) goto errout; - err = fib_dump_info(skb, pid, n->nlmsg_seq, event, tb_id, - fa->fa_type, fa->fa_scope, &key, z, fa->fa_tos, - fa->fa_info, 0); + err = fib_dump_info(skb, info->pid, seq, event, tb_id, + fa->fa_type, fa->fa_scope, &key, dst_len, + fa->fa_tos, fa->fa_info, 0); if (err < 0) { kfree_skb(skb); goto errout; } - err = rtnl_notify(skb, pid, RTNLGRP_IPV4_ROUTE, n, GFP_KERNEL); + err = rtnl_notify(skb, info->pid, RTNLGRP_IPV4_ROUTE, + info->nlh, GFP_KERNEL); errout: if (err < 0) rtnl_set_sk_err(RTNLGRP_IPV4_ROUTE, err); @@ -342,102 +343,100 @@ int fib_detect_death(struct fib_info *fi #ifdef CONFIG_IP_ROUTE_MULTIPATH -static u32 fib_get_attr32(struct rtattr *attr, int attrlen, int type) -{ - while (RTA_OK(attr,attrlen)) { - if (attr->rta_type == type) - return *(u32*)RTA_DATA(attr); - attr = RTA_NEXT(attr, attrlen); - } - return 0; -} - -static int -fib_count_nexthops(struct rtattr *rta) +static int fib_count_nexthops(struct rtnexthop *rtnh, int remaining) { int nhs = 0; - struct rtnexthop *nhp = RTA_DATA(rta); - int nhlen = RTA_PAYLOAD(rta); - while (nhlen >= (int)sizeof(struct rtnexthop)) { - if ((nhlen -= nhp->rtnh_len) < 0) - return 0; + while (rtnh_ok(rtnh, remaining)) { nhs++; - nhp = RTNH_NEXT(nhp); - }; - return nhs; + rtnh = rtnh_next(rtnh, &remaining); + } + + /* leftover implies invalid nexthop configuration, discard it */ + return remaining > 0 ? 0 : nhs; } -static int -fib_get_nhs(struct fib_info *fi, const struct rtattr *rta, const struct rtmsg *r) +static int fib_get_nhs(struct fib_info *fi, struct rtnexthop *rtnh, + int remaining, struct fib_config *cfg) { - struct rtnexthop *nhp = RTA_DATA(rta); - int nhlen = RTA_PAYLOAD(rta); - change_nexthops(fi) { - int attrlen = nhlen - sizeof(struct rtnexthop); - if (attrlen < 0 || (nhlen -= nhp->rtnh_len) < 0) + int attrlen; + + if (!rtnh_ok(rtnh, remaining)) return -EINVAL; - nh->nh_flags = (r->rtm_flags&~0xFF) | nhp->rtnh_flags; - nh->nh_oif = nhp->rtnh_ifindex; - nh->nh_weight = nhp->rtnh_hops + 1; - if (attrlen) { - nh->nh_gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_GATEWAY); + + nh->nh_flags = (cfg->fc_flags & ~0xFF) | rtnh->rtnh_flags; + nh->nh_oif = rtnh->rtnh_ifindex; + nh->nh_weight = rtnh->rtnh_hops + 1; + + attrlen = rtnh_attrlen(rtnh); + if (attrlen > 0) { + struct nlattr *nla, *attrs = rtnh_attrs(rtnh); + + nla = nla_find(attrs, attrlen, RTA_GATEWAY); + nh->nh_gw = nla ? nla_get_u32(nla) : 0; #ifdef CONFIG_NET_CLS_ROUTE - nh->nh_tclassid = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_FLOW); + nla = nla_find(attrs, attrlen, RTA_FLOW); + nh->nh_tclassid = nla ? nla_get_u32(nla) : 0; #endif } - nhp = RTNH_NEXT(nhp); + + rtnh = rtnh_next(rtnh, &remaining); } endfor_nexthops(fi); + return 0; } #endif -int fib_nh_match(struct rtmsg *r, struct nlmsghdr *nlh, struct kern_rta *rta, - struct fib_info *fi) +int fib_nh_match(struct fib_config *cfg, struct fib_info *fi) { #ifdef CONFIG_IP_ROUTE_MULTIPATH - struct rtnexthop *nhp; - int nhlen; + struct rtnexthop *rtnh; + int remaining; #endif - if (rta->rta_priority && - *rta->rta_priority != fi->fib_priority) + if (cfg->fc_priority && cfg->fc_priority != fi->fib_priority) return 1; - if (rta->rta_oif || rta->rta_gw) { - if ((!rta->rta_oif || *rta->rta_oif == fi->fib_nh->nh_oif) && - (!rta->rta_gw || memcmp(rta->rta_gw, &fi->fib_nh->nh_gw, 4) == 0)) + if (cfg->fc_oif || cfg->fc_gw) { + if ((!cfg->fc_oif || cfg->fc_oif == fi->fib_nh->nh_oif) && + (!cfg->fc_gw || cfg->fc_gw == fi->fib_nh->nh_gw)) return 0; return 1; } #ifdef CONFIG_IP_ROUTE_MULTIPATH - if (rta->rta_mp == NULL) + if (cfg->fc_mp == NULL) return 0; - nhp = RTA_DATA(rta->rta_mp); - nhlen = RTA_PAYLOAD(rta->rta_mp); + + rtnh = cfg->fc_mp; + remaining = cfg->fc_mp_len; for_nexthops(fi) { - int attrlen = nhlen - sizeof(struct rtnexthop); - u32 gw; + int attrlen; - if (attrlen < 0 || (nhlen -= nhp->rtnh_len) < 0) + if (!rtnh_ok(rtnh, remaining)) return -EINVAL; - if (nhp->rtnh_ifindex && nhp->rtnh_ifindex != nh->nh_oif) + + if (rtnh->rtnh_ifindex && rtnh->rtnh_ifindex != nh->nh_oif) return 1; - if (attrlen) { - gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_GATEWAY); - if (gw && gw != nh->nh_gw) + + attrlen = rtnh_attrlen(rtnh); + if (attrlen < 0) { + struct nlattr *nla, *attrs = rtnh_attrs(rtnh); + + nla = nla_find(attrs, attrlen, RTA_GATEWAY); + if (nla && nla_get_u32(nla) != nh->nh_gw) return 1; #ifdef CONFIG_NET_CLS_ROUTE - gw = fib_get_attr32(RTNH_DATA(nhp), attrlen, RTA_FLOW); - if (gw && gw != nh->nh_tclassid) + nla = nla_find(attrs, attrlen, RTA_FLOW); + if (nla && nla_get_u32(nla) != nh->nh_tclassid) return 1; #endif } - nhp = RTNH_NEXT(nhp); + + rtnh = rtnh_next(rtnh, &remaining); } endfor_nexthops(fi); #endif return 0; @@ -488,7 +487,8 @@ int fib_nh_match(struct rtmsg *r, struct |-> {local prefix} (terminal node) */ -static int fib_check_nh(const struct rtmsg *r, struct fib_info *fi, struct fib_nh *nh) +static int fib_check_nh(struct fib_config *cfg, struct fib_info *fi, + struct fib_nh *nh) { int err; @@ -502,7 +502,7 @@ static int fib_check_nh(const struct rtm if (nh->nh_flags&RTNH_F_ONLINK) { struct net_device *dev; - if (r->rtm_scope >= RT_SCOPE_LINK) + if (cfg->fc_scope >= RT_SCOPE_LINK) return -EINVAL; if (inet_addr_type(nh->nh_gw) != RTN_UNICAST) return -EINVAL; @@ -516,10 +516,15 @@ static int fib_check_nh(const struct rtm return 0; } { - struct flowi fl = { .nl_u = { .ip4_u = - { .daddr = nh->nh_gw, - .scope = r->rtm_scope + 1 } }, - .oif = nh->nh_oif }; + struct flowi fl = { + .nl_u = { + .ip4_u = { + .daddr = nh->nh_gw, + .scope = cfg->fc_scope + 1, + }, + }, + .oif = nh->nh_oif, + }; /* It is not necessary, but requires a bit of thinking */ if (fl.fl4_scope < RT_SCOPE_LINK) @@ -646,39 +651,28 @@ static void fib_hash_move(struct hlist_h fib_hash_free(old_laddrhash, bytes); } -struct fib_info * -fib_create_info(const struct rtmsg *r, struct kern_rta *rta, - const struct nlmsghdr *nlh, int *errp) +struct fib_info *fib_create_info(struct fib_config *cfg) { int err; struct fib_info *fi = NULL; struct fib_info *ofi; -#ifdef CONFIG_IP_ROUTE_MULTIPATH int nhs = 1; -#else - const int nhs = 1; -#endif -#ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED - u32 mp_alg = IP_MP_ALG_NONE; -#endif /* Fast check to catch the most weird cases */ - if (fib_props[r->rtm_type].scope > r->rtm_scope) + if (fib_props[cfg->fc_type].scope > cfg->fc_scope) goto err_inval; #ifdef CONFIG_IP_ROUTE_MULTIPATH - if (rta->rta_mp) { - nhs = fib_count_nexthops(rta->rta_mp); + if (cfg->fc_mp) { + nhs = fib_count_nexthops(cfg->fc_mp, cfg->fc_mp_len); if (nhs == 0) goto err_inval; } #endif #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED - if (rta->rta_mp_alg) { - mp_alg = *rta->rta_mp_alg; - - if (mp_alg < IP_MP_ALG_NONE || - mp_alg > IP_MP_ALG_MAX) + if (cfg->fc_mp_alg) { + if (cfg->fc_mp_alg < IP_MP_ALG_NONE || + cfg->fc_mp_alg > IP_MP_ALG_MAX) goto err_inval; } #endif @@ -714,43 +708,42 @@ fib_create_info(const struct rtmsg *r, s goto failure; fib_info_cnt++; - fi->fib_protocol = r->rtm_protocol; + fi->fib_protocol = cfg->fc_protocol; + fi->fib_flags = cfg->fc_flags; + fi->fib_priority = cfg->fc_priority; + fi->fib_prefsrc = cfg->fc_prefsrc; fi->fib_nhs = nhs; change_nexthops(fi) { nh->nh_parent = fi; } endfor_nexthops(fi) - fi->fib_flags = r->rtm_flags; - if (rta->rta_priority) - fi->fib_priority = *rta->rta_priority; - if (rta->rta_mx) { - int attrlen = RTA_PAYLOAD(rta->rta_mx); - struct rtattr *attr = RTA_DATA(rta->rta_mx); - - while (RTA_OK(attr, attrlen)) { - unsigned flavor = attr->rta_type; - if (flavor) { - if (flavor > RTAX_MAX) + if (cfg->fc_mx) { + struct nlattr *nla; + int remaining; + + nla_for_each_attr(nla, cfg->fc_mx, cfg->fc_mx_len, remaining) { + int type = nla->nla_type; + + if (type) { + if (type > RTAX_MAX) goto err_inval; - fi->fib_metrics[flavor-1] = *(unsigned*)RTA_DATA(attr); + fi->fib_metrics[type - 1] = nla_get_u32(nla); } - attr = RTA_NEXT(attr, attrlen); } } - if (rta->rta_prefsrc) - memcpy(&fi->fib_prefsrc, rta->rta_prefsrc, 4); - if (rta->rta_mp) { + if (cfg->fc_mp) { #ifdef CONFIG_IP_ROUTE_MULTIPATH - if ((err = fib_get_nhs(fi, rta->rta_mp, r)) != 0) + err = fib_get_nhs(fi, cfg->fc_mp, cfg->fc_mp_len, cfg); + if (err != 0) goto failure; - if (rta->rta_oif && fi->fib_nh->nh_oif != *rta->rta_oif) + if (cfg->fc_oif && fi->fib_nh->nh_oif != cfg->fc_oif) goto err_inval; - if (rta->rta_gw && memcmp(&fi->fib_nh->nh_gw, rta->rta_gw, 4)) + if (cfg->fc_gw && fi->fib_nh->nh_gw != cfg->fc_gw) goto err_inval; #ifdef CONFIG_NET_CLS_ROUTE - if (rta->rta_flow && memcmp(&fi->fib_nh->nh_tclassid, rta->rta_flow, 4)) + if (cfg->fc_flow && fi->fib_nh->nh_tclassid != cfg->fc_flow) goto err_inval; #endif #else @@ -758,34 +751,32 @@ fib_create_info(const struct rtmsg *r, s #endif } else { struct fib_nh *nh = fi->fib_nh; - if (rta->rta_oif) - nh->nh_oif = *rta->rta_oif; - if (rta->rta_gw) - memcpy(&nh->nh_gw, rta->rta_gw, 4); + + nh->nh_oif = cfg->fc_oif; + nh->nh_gw = cfg->fc_gw; + nh->nh_flags = cfg->fc_flags; #ifdef CONFIG_NET_CLS_ROUTE - if (rta->rta_flow) - memcpy(&nh->nh_tclassid, rta->rta_flow, 4); + nh->nh_tclassid = cfg->fc_flow; #endif - nh->nh_flags = r->rtm_flags; #ifdef CONFIG_IP_ROUTE_MULTIPATH nh->nh_weight = 1; #endif } #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED - fi->fib_mp_alg = mp_alg; + fi->fib_mp_alg = cfg->fc_mp_alg; #endif - if (fib_props[r->rtm_type].error) { - if (rta->rta_gw || rta->rta_oif || rta->rta_mp) + if (fib_props[cfg->fc_type].error) { + if (cfg->fc_gw || cfg->fc_oif || cfg->fc_mp) goto err_inval; goto link_it; } - if (r->rtm_scope > RT_SCOPE_HOST) + if (cfg->fc_scope > RT_SCOPE_HOST) goto err_inval; - if (r->rtm_scope == RT_SCOPE_HOST) { + if (cfg->fc_scope == RT_SCOPE_HOST) { struct fib_nh *nh = fi->fib_nh; /* Local address is added. */ @@ -798,14 +789,14 @@ fib_create_info(const struct rtmsg *r, s goto failure; } else { change_nexthops(fi) { - if ((err = fib_check_nh(r, fi, nh)) != 0) + if ((err = fib_check_nh(cfg, fi, nh)) != 0) goto failure; } endfor_nexthops(fi) } if (fi->fib_prefsrc) { - if (r->rtm_type != RTN_LOCAL || rta->rta_dst == NULL || - memcmp(&fi->fib_prefsrc, rta->rta_dst, 4)) + if (cfg->fc_type != RTN_LOCAL || !cfg->fc_dst || + fi->fib_prefsrc != cfg->fc_dst) if (inet_addr_type(fi->fib_prefsrc) != RTN_LOCAL) goto err_inval; } @@ -846,12 +837,12 @@ err_inval: err = -EINVAL; failure: - *errp = err; if (fi) { fi->fib_dead = 1; free_fib_info(fi); } - return NULL; + + return ERR_PTR(err); } /* Note! fib_semantic_match intentionally uses RCU list functions. */ @@ -1012,150 +1003,6 @@ rtattr_failure: return -1; } -#ifndef CONFIG_IP_NOSIOCRT - -int -fib_convert_rtentry(int cmd, struct nlmsghdr *nl, struct rtmsg *rtm, - struct kern_rta *rta, struct rtentry *r) -{ - int plen; - u32 *ptr; - - memset(rtm, 0, sizeof(*rtm)); - memset(rta, 0, sizeof(*rta)); - - if (r->rt_dst.sa_family != AF_INET) - return -EAFNOSUPPORT; - - /* Check mask for validity: - a) it must be contiguous. - b) destination must have all host bits clear. - c) if application forgot to set correct family (AF_INET), - reject request unless it is absolutely clear i.e. - both family and mask are zero. - */ - plen = 32; - ptr = &((struct sockaddr_in*)&r->rt_dst)->sin_addr.s_addr; - if (!(r->rt_flags&RTF_HOST)) { - u32 mask = ((struct sockaddr_in*)&r->rt_genmask)->sin_addr.s_addr; - if (r->rt_genmask.sa_family != AF_INET) { - if (mask || r->rt_genmask.sa_family) - return -EAFNOSUPPORT; - } - if (bad_mask(mask, *ptr)) - return -EINVAL; - plen = inet_mask_len(mask); - } - - nl->nlmsg_flags = NLM_F_REQUEST; - nl->nlmsg_pid = 0; - nl->nlmsg_seq = 0; - nl->nlmsg_len = NLMSG_LENGTH(sizeof(*rtm)); - if (cmd == SIOCDELRT) { - nl->nlmsg_type = RTM_DELROUTE; - nl->nlmsg_flags = 0; - } else { - nl->nlmsg_type = RTM_NEWROUTE; - nl->nlmsg_flags = NLM_F_REQUEST|NLM_F_CREATE; - rtm->rtm_protocol = RTPROT_BOOT; - } - - rtm->rtm_dst_len = plen; - rta->rta_dst = ptr; - - if (r->rt_metric) { - *(u32*)&r->rt_pad3 = r->rt_metric - 1; - rta->rta_priority = (u32*)&r->rt_pad3; - } - if (r->rt_flags&RTF_REJECT) { - rtm->rtm_scope = RT_SCOPE_HOST; - rtm->rtm_type = RTN_UNREACHABLE; - return 0; - } - rtm->rtm_scope = RT_SCOPE_NOWHERE; - rtm->rtm_type = RTN_UNICAST; - - if (r->rt_dev) { - char *colon; - struct net_device *dev; - char devname[IFNAMSIZ]; - - if (copy_from_user(devname, r->rt_dev, IFNAMSIZ-1)) - return -EFAULT; - devname[IFNAMSIZ-1] = 0; - colon = strchr(devname, ':'); - if (colon) - *colon = 0; - dev = __dev_get_by_name(devname); - if (!dev) - return -ENODEV; - rta->rta_oif = &dev->ifindex; - if (colon) { - struct in_ifaddr *ifa; - struct in_device *in_dev = __in_dev_get_rtnl(dev); - if (!in_dev) - return -ENODEV; - *colon = ':'; - for (ifa = in_dev->ifa_list; ifa; ifa = ifa->ifa_next) - if (strcmp(ifa->ifa_label, devname) == 0) - break; - if (ifa == NULL) - return -ENODEV; - rta->rta_prefsrc = &ifa->ifa_local; - } - } - - ptr = &((struct sockaddr_in*)&r->rt_gateway)->sin_addr.s_addr; - if (r->rt_gateway.sa_family == AF_INET && *ptr) { - rta->rta_gw = ptr; - if (r->rt_flags&RTF_GATEWAY && inet_addr_type(*ptr) == RTN_UNICAST) - rtm->rtm_scope = RT_SCOPE_UNIVERSE; - } - - if (cmd == SIOCDELRT) - return 0; - - if (r->rt_flags&RTF_GATEWAY && rta->rta_gw == NULL) - return -EINVAL; - - if (rtm->rtm_scope == RT_SCOPE_NOWHERE) - rtm->rtm_scope = RT_SCOPE_LINK; - - if (r->rt_flags&(RTF_MTU|RTF_WINDOW|RTF_IRTT)) { - struct rtattr *rec; - struct rtattr *mx = kmalloc(RTA_LENGTH(3*RTA_LENGTH(4)), GFP_KERNEL); - if (mx == NULL) - return -ENOMEM; - rta->rta_mx = mx; - mx->rta_type = RTA_METRICS; - mx->rta_len = RTA_LENGTH(0); - if (r->rt_flags&RTF_MTU) { - rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len)); - rec->rta_type = RTAX_ADVMSS; - rec->rta_len = RTA_LENGTH(4); - mx->rta_len += RTA_LENGTH(4); - *(u32*)RTA_DATA(rec) = r->rt_mtu - 40; - } - if (r->rt_flags&RTF_WINDOW) { - rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len)); - rec->rta_type = RTAX_WINDOW; - rec->rta_len = RTA_LENGTH(4); - mx->rta_len += RTA_LENGTH(4); - *(u32*)RTA_DATA(rec) = r->rt_window; - } - if (r->rt_flags&RTF_IRTT) { - rec = (void*)((char*)mx + RTA_ALIGN(mx->rta_len)); - rec->rta_type = RTAX_RTT; - rec->rta_len = RTA_LENGTH(4); - mx->rta_len += RTA_LENGTH(4); - *(u32*)RTA_DATA(rec) = r->rt_irtt<<3; - } - } - return 0; -} - -#endif - /* Update FIB if: - local address disappeared -> we must delete all the entries Index: net-2.6.19.git/include/net/nexthop.h =================================================================== --- /dev/null +++ net-2.6.19.git/include/net/nexthop.h @@ -0,0 +1,33 @@ +#ifndef __NET_NEXTHOP_H +#define __NET_NEXTHOP_H + +#include <linux/rtnetlink.h> +#include <net/netlink.h> + +static inline int rtnh_ok(const struct rtnexthop *rtnh, int remaining) +{ + return remaining >= sizeof(*rtnh) && + rtnh->rtnh_len >= sizeof(*rtnh) && + rtnh->rtnh_len <= remaining; +} + +static inline struct rtnexthop *rtnh_next(const struct rtnexthop *rtnh, + int *remaining) +{ + int totlen = NLA_ALIGN(rtnh->rtnh_len); + + *remaining -= totlen; + return (struct rtnexthop *) ((char *) rtnh + totlen); +} + +static inline struct nlattr *rtnh_attrs(const struct rtnexthop *rtnh) +{ + return (struct nlattr *) ((char *) rtnh + NLA_ALIGN(sizeof(*rtnh))); +} + +static inline int rtnh_attrlen(const struct rtnexthop *rtnh) +{ + return rtnh->rtnh_len - NLA_ALIGN(sizeof(*rtnh)); +} + +#endif Index: net-2.6.19.git/net/ipv4/fib_hash.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_hash.c +++ net-2.6.19.git/net/ipv4/fib_hash.c @@ -379,42 +379,39 @@ static struct fib_node *fib_find_node(st return NULL; } -static int -fn_hash_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *n, struct netlink_skb_parms *req) +static int fn_hash_insert(struct fib_table *tb, struct fib_config *cfg) { struct fn_hash *table = (struct fn_hash *) tb->tb_data; struct fib_node *new_f, *f; struct fib_alias *fa, *new_fa; struct fn_zone *fz; struct fib_info *fi; - int z = r->rtm_dst_len; - int type = r->rtm_type; - u8 tos = r->rtm_tos; + u8 tos = cfg->fc_tos; u32 key; int err; - if (z > 32) + if (cfg->fc_dst_len > 32) return -EINVAL; - fz = table->fn_zones[z]; - if (!fz && !(fz = fn_new_zone(table, z))) + + fz = table->fn_zones[cfg->fc_dst_len]; + if (!fz && !(fz = fn_new_zone(table, cfg->fc_dst_len))) return -ENOBUFS; key = 0; - if (rta->rta_dst) { - u32 dst; - memcpy(&dst, rta->rta_dst, 4); - if (dst & ~FZ_MASK(fz)) + if (cfg->fc_dst) { + if (cfg->fc_dst & ~FZ_MASK(fz)) return -EINVAL; - key = fz_key(dst, fz); + key = fz_key(cfg->fc_dst, fz); } - if ((fi = fib_create_info(r, rta, n, &err)) == NULL) - return err; + fi = fib_create_info(cfg); + if (IS_ERR(fi)) + return PTR_ERR(fi); if (fz->fz_nent > (fz->fz_divisor<<1) && fz->fz_divisor < FZ_MAX_DIVISOR && - (z==32 || (1<<z) > fz->fz_divisor)) + (cfg->fc_dst_len == 32 || + (1 << cfg->fc_dst_len) > fz->fz_divisor)) fn_rehash_zone(fz); f = fib_find_node(fz, key); @@ -440,18 +437,18 @@ fn_hash_insert(struct fib_table *tb, str struct fib_alias *fa_orig; err = -EEXIST; - if (n->nlmsg_flags & NLM_F_EXCL) + if (cfg->fc_nlflags & NLM_F_EXCL) goto out; - if (n->nlmsg_flags & NLM_F_REPLACE) { + if (cfg->fc_nlflags & NLM_F_REPLACE) { struct fib_info *fi_drop; u8 state; write_lock_bh(&fib_hash_lock); fi_drop = fa->fa_info; fa->fa_info = fi; - fa->fa_type = type; - fa->fa_scope = r->rtm_scope; + fa->fa_type = cfg->fc_type; + fa->fa_scope = cfg->fc_scope; state = fa->fa_state; fa->fa_state &= ~FA_S_ACCESSED; fib_hash_genid++; @@ -474,17 +471,17 @@ fn_hash_insert(struct fib_table *tb, str break; if (fa->fa_info->fib_priority != fi->fib_priority) break; - if (fa->fa_type == type && - fa->fa_scope == r->rtm_scope && + if (fa->fa_type == cfg->fc_type && + fa->fa_scope == cfg->fc_scope && fa->fa_info == fi) goto out; } - if (!(n->nlmsg_flags & NLM_F_APPEND)) + if (!(cfg->fc_nlflags & NLM_F_APPEND)) fa = fa_orig; } err = -ENOENT; - if (!(n->nlmsg_flags&NLM_F_CREATE)) + if (!(cfg->fc_nlflags & NLM_F_CREATE)) goto out; err = -ENOBUFS; @@ -506,8 +503,8 @@ fn_hash_insert(struct fib_table *tb, str new_fa->fa_info = fi; new_fa->fa_tos = tos; - new_fa->fa_type = type; - new_fa->fa_scope = r->rtm_scope; + new_fa->fa_type = cfg->fc_type; + new_fa->fa_scope = cfg->fc_scope; new_fa->fa_state = 0; /* @@ -526,7 +523,8 @@ fn_hash_insert(struct fib_table *tb, str fz->fz_nent++; rt_cache_flush(-1); - rtmsg_fib(RTM_NEWROUTE, key, new_fa, z, tb->tb_id, n, req); + rtmsg_fib(RTM_NEWROUTE, key, new_fa, cfg->fc_dst_len, tb->tb_id, + &cfg->fc_nlinfo); return 0; out_free_new_fa: @@ -537,30 +535,25 @@ out: } -static int -fn_hash_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *n, struct netlink_skb_parms *req) +static int fn_hash_delete(struct fib_table *tb, struct fib_config *cfg) { struct fn_hash *table = (struct fn_hash*)tb->tb_data; struct fib_node *f; struct fib_alias *fa, *fa_to_delete; - int z = r->rtm_dst_len; struct fn_zone *fz; u32 key; - u8 tos = r->rtm_tos; - if (z > 32) + if (cfg->fc_dst_len > 32) return -EINVAL; - if ((fz = table->fn_zones[z]) == NULL) + + if ((fz = table->fn_zones[cfg->fc_dst_len]) == NULL) return -ESRCH; key = 0; - if (rta->rta_dst) { - u32 dst; - memcpy(&dst, rta->rta_dst, 4); - if (dst & ~FZ_MASK(fz)) + if (cfg->fc_dst) { + if (cfg->fc_dst & ~FZ_MASK(fz)) return -EINVAL; - key = fz_key(dst, fz); + key = fz_key(cfg->fc_dst, fz); } f = fib_find_node(fz, key); @@ -568,7 +561,7 @@ fn_hash_delete(struct fib_table *tb, str if (!f) fa = NULL; else - fa = fib_find_alias(&f->fn_alias, tos, 0); + fa = fib_find_alias(&f->fn_alias, cfg->fc_tos, 0); if (!fa) return -ESRCH; @@ -577,16 +570,16 @@ fn_hash_delete(struct fib_table *tb, str list_for_each_entry_continue(fa, &f->fn_alias, fa_list) { struct fib_info *fi = fa->fa_info; - if (fa->fa_tos != tos) + if (fa->fa_tos != cfg->fc_tos) break; - if ((!r->rtm_type || - fa->fa_type == r->rtm_type) && - (r->rtm_scope == RT_SCOPE_NOWHERE || - fa->fa_scope == r->rtm_scope) && - (!r->rtm_protocol || - fi->fib_protocol == r->rtm_protocol) && - fib_nh_match(r, n, rta, fi) == 0) { + if ((!cfg->fc_type || + fa->fa_type == cfg->fc_type) && + (cfg->fc_scope == RT_SCOPE_NOWHERE || + fa->fa_scope == cfg->fc_scope) && + (!cfg->fc_protocol || + fi->fib_protocol == cfg->fc_protocol) && + fib_nh_match(cfg, fi) == 0) { fa_to_delete = fa; break; } @@ -596,7 +589,8 @@ fn_hash_delete(struct fib_table *tb, str int kill_fn; fa = fa_to_delete; - rtmsg_fib(RTM_DELROUTE, key, fa, z, tb->tb_id, n, req); + rtmsg_fib(RTM_DELROUTE, key, fa, cfg->fc_dst_len, + tb->tb_id, &cfg->fc_nlinfo); kill_fn = 0; write_lock_bh(&fib_hash_lock); Index: net-2.6.19.git/net/ipv4/fib_trie.c =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_trie.c +++ net-2.6.19.git/net/ipv4/fib_trie.c @@ -1124,17 +1124,14 @@ err: return fa_head; } -static int -fn_trie_insert(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *nlhdr, struct netlink_skb_parms *req) +static int fn_trie_insert(struct fib_table *tb, struct fib_config *cfg) { struct trie *t = (struct trie *) tb->tb_data; struct fib_alias *fa, *new_fa; struct list_head *fa_head = NULL; struct fib_info *fi; - int plen = r->rtm_dst_len; - int type = r->rtm_type; - u8 tos = r->rtm_tos; + int plen = cfg->fc_dst_len; + u8 tos = cfg->fc_tos; u32 key, mask; int err; struct leaf *l; @@ -1142,11 +1139,7 @@ fn_trie_insert(struct fib_table *tb, str if (plen > 32) return -EINVAL; - key = 0; - if (rta->rta_dst) - memcpy(&key, rta->rta_dst, 4); - - key = ntohl(key); + key = ntohl(cfg->fc_dst); pr_debug("Insert table=%u %08x/%d\n", tb->tb_id, key, plen); @@ -1157,10 +1150,11 @@ fn_trie_insert(struct fib_table *tb, str key = key & mask; - fi = fib_create_info(r, rta, nlhdr, &err); - - if (!fi) + fi = fib_create_info(cfg); + if (IS_ERR(fi)) { + err = PTR_ERR(fi); goto err; + } l = fib_find_node(t, key); fa = NULL; @@ -1185,10 +1179,10 @@ fn_trie_insert(struct fib_table *tb, str struct fib_alias *fa_orig; err = -EEXIST; - if (nlhdr->nlmsg_flags & NLM_F_EXCL) + if (cfg->fc_nlflags & NLM_F_EXCL) goto out; - if (nlhdr->nlmsg_flags & NLM_F_REPLACE) { + if (cfg->fc_nlflags & NLM_F_REPLACE) { struct fib_info *fi_drop; u8 state; @@ -1200,8 +1194,8 @@ fn_trie_insert(struct fib_table *tb, str fi_drop = fa->fa_info; new_fa->fa_tos = fa->fa_tos; new_fa->fa_info = fi; - new_fa->fa_type = type; - new_fa->fa_scope = r->rtm_scope; + new_fa->fa_type = cfg->fc_type; + new_fa->fa_scope = cfg->fc_scope; state = fa->fa_state; new_fa->fa_state &= ~FA_S_ACCESSED; @@ -1224,17 +1218,17 @@ fn_trie_insert(struct fib_table *tb, str break; if (fa->fa_info->fib_priority != fi->fib_priority) break; - if (fa->fa_type == type && - fa->fa_scope == r->rtm_scope && + if (fa->fa_type == cfg->fc_type && + fa->fa_scope == cfg->fc_scope && fa->fa_info == fi) { goto out; } } - if (!(nlhdr->nlmsg_flags & NLM_F_APPEND)) + if (!(cfg->fc_nlflags & NLM_F_APPEND)) fa = fa_orig; } err = -ENOENT; - if (!(nlhdr->nlmsg_flags & NLM_F_CREATE)) + if (!(cfg->fc_nlflags & NLM_F_CREATE)) goto out; err = -ENOBUFS; @@ -1244,8 +1238,8 @@ fn_trie_insert(struct fib_table *tb, str new_fa->fa_info = fi; new_fa->fa_tos = tos; - new_fa->fa_type = type; - new_fa->fa_scope = r->rtm_scope; + new_fa->fa_type = cfg->fc_type; + new_fa->fa_scope = cfg->fc_scope; new_fa->fa_state = 0; /* * Insert new entry to the list. @@ -1262,7 +1256,8 @@ fn_trie_insert(struct fib_table *tb, str (fa ? &fa->fa_list : fa_head)); rt_cache_flush(-1); - rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id, nlhdr, req); + rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id, + &cfg->fc_nlinfo); succeeded: return 0; @@ -1548,28 +1543,21 @@ static int trie_leaf_remove(struct trie return 1; } -static int -fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, - struct nlmsghdr *nlhdr, struct netlink_skb_parms *req) +static int fn_trie_delete(struct fib_table *tb, struct fib_config *cfg) { struct trie *t = (struct trie *) tb->tb_data; u32 key, mask; - int plen = r->rtm_dst_len; - u8 tos = r->rtm_tos; + int plen = cfg->fc_dst_len; + u8 tos = cfg->fc_tos; struct fib_alias *fa, *fa_to_delete; struct list_head *fa_head; struct leaf *l; struct leaf_info *li; - if (plen > 32) return -EINVAL; - key = 0; - if (rta->rta_dst) - memcpy(&key, rta->rta_dst, 4); - - key = ntohl(key); + key = ntohl(cfg->fc_dst); mask = ntohl(inet_make_mask(plen)); if (key & ~mask) @@ -1598,13 +1586,12 @@ fn_trie_delete(struct fib_table *tb, str if (fa->fa_tos != tos) break; - if ((!r->rtm_type || - fa->fa_type == r->rtm_type) && - (r->rtm_scope == RT_SCOPE_NOWHERE || - fa->fa_scope == r->rtm_scope) && - (!r->rtm_protocol || - fi->fib_protocol == r->rtm_protocol) && - fib_nh_match(r, nlhdr, rta, fi) == 0) { + if ((!cfg->fc_type || fa->fa_type == cfg->fc_type) && + (cfg->fc_scope == RT_SCOPE_NOWHERE || + fa->fa_scope == cfg->fc_scope) && + (!cfg->fc_protocol || + fi->fib_protocol == cfg->fc_protocol) && + fib_nh_match(cfg, fi) == 0) { fa_to_delete = fa; break; } @@ -1614,7 +1601,8 @@ fn_trie_delete(struct fib_table *tb, str return -ESRCH; fa = fa_to_delete; - rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, nlhdr, req); + rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, + &cfg->fc_nlinfo); l = fib_find_node(t, key); li = find_leaf_info(l, plen); Index: net-2.6.19.git/net/ipv4/fib_lookup.h =================================================================== --- net-2.6.19.git.orig/net/ipv4/fib_lookup.h +++ net-2.6.19.git/net/ipv4/fib_lookup.h @@ -23,19 +23,14 @@ extern int fib_semantic_match(struct lis struct fib_result *res, __u32 zone, __u32 mask, int prefixlen); extern void fib_release_info(struct fib_info *); -extern struct fib_info *fib_create_info(const struct rtmsg *r, - struct kern_rta *rta, - const struct nlmsghdr *, - int *err); -extern int fib_nh_match(struct rtmsg *r, struct nlmsghdr *, - struct kern_rta *rta, struct fib_info *fi); +extern struct fib_info *fib_create_info(struct fib_config *cfg); +extern int fib_nh_match(struct fib_config *cfg, struct fib_info *fi); extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event, u32 tb_id, u8 type, u8 scope, void *dst, int dst_len, u8 tos, struct fib_info *fi, unsigned int); extern void rtmsg_fib(int event, u32 key, struct fib_alias *fa, - int z, u32 tb_id, - struct nlmsghdr *n, struct netlink_skb_parms *req); + int dst_len, u32 tb_id, struct nl_info *info); extern struct fib_alias *fib_find_alias(struct list_head *fah, u8 tos, u32 prio); extern int fib_detect_death(struct fib_info *fi, int order, Index: net-2.6.19.git/include/net/netlink.h =================================================================== --- net-2.6.19.git.orig/include/net/netlink.h +++ net-2.6.19.git/include/net/netlink.h @@ -192,6 +192,16 @@ struct nla_policy { u16 minlen; }; +/** + * struct nl_info - netlink source information + * @nlh: Netlink message header of original request + * @pid: Netlink PID of requesting application + */ +struct nl_info { + struct nlmsghdr *nlh; + u32 pid; +}; + extern void netlink_run_queue(struct sock *sk, unsigned int *qlen, int (*cb)(struct sk_buff *, struct nlmsghdr *, int *)); ^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2006-08-15 15:00 UTC | newest] Thread overview: 7+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2006-08-08 19:38 [PATCHSET] IPv4 FIB configuration rework Thomas Graf 2006-08-07 22:00 ` [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config Thomas Graf 2006-08-08 23:11 ` Thomas Graf 2006-08-08 23:12 ` David Miller 2006-08-07 22:00 ` [PATCH 2/3] [IPv4]: Convert FIB dumping to use new netlink api Thomas Graf 2006-08-07 22:00 ` [PATCH 3/3] [IPv4]: Convert route get to " Thomas Graf -- strict thread matches above, loose matches on Subject: below -- 2006-08-15 14:57 [PATCHSET] IPv4 FIB configuration conversions Thomas Graf 2006-08-14 22:00 ` [PATCH 1/3] [IPv4]: FIB configuration using struct fib_config Thomas Graf
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.