From 2156d321b879cdadb95a633d046169cfebdbf784 Mon Sep 17 00:00:00 2001 From: Arturo Borrero Date: Sat, 21 Feb 2015 19:30:55 +0100 Subject: netfilter: nft_compat: don't truncate ethernet protocol type to u8 Use u16 for protocol and then cast it to __be16 >> net/netfilter/nft_compat.c:140:37: sparse: incorrect type in assignment (different base types) net/netfilter/nft_compat.c:140:37: expected restricted __be16 [usertype] ethproto net/netfilter/nft_compat.c:140:37: got unsigned char [unsigned] [usertype] proto >> net/netfilter/nft_compat.c:351:37: sparse: incorrect type in assignment (different base types) net/netfilter/nft_compat.c:351:37: expected restricted __be16 [usertype] ethproto net/netfilter/nft_compat.c:351:37: got unsigned char [unsigned] [usertype] proto Reported-by: kbuild test robot Signed-off-by: Arturo Borrero Gonzalez Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nft_compat.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'net') diff --git a/net/netfilter/nft_compat.c b/net/netfilter/nft_compat.c index 1279cd85663e..213584cf04b3 100644 --- a/net/netfilter/nft_compat.c +++ b/net/netfilter/nft_compat.c @@ -123,7 +123,7 @@ static void nft_target_set_tgchk_param(struct xt_tgchk_param *par, const struct nft_ctx *ctx, struct xt_target *target, void *info, - union nft_entry *entry, u8 proto, bool inv) + union nft_entry *entry, u16 proto, bool inv) { par->net = ctx->net; par->table = ctx->table->name; @@ -137,7 +137,7 @@ nft_target_set_tgchk_param(struct xt_tgchk_param *par, entry->e6.ipv6.invflags = inv ? IP6T_INV_PROTO : 0; break; case NFPROTO_BRIDGE: - entry->ebt.ethproto = proto; + entry->ebt.ethproto = (__force __be16)proto; entry->ebt.invflags = inv ? EBT_IPROTO : 0; break; } @@ -171,7 +171,7 @@ static const struct nla_policy nft_rule_compat_policy[NFTA_RULE_COMPAT_MAX + 1] [NFTA_RULE_COMPAT_FLAGS] = { .type = NLA_U32 }, }; -static int nft_parse_compat(const struct nlattr *attr, u8 *proto, bool *inv) +static int nft_parse_compat(const struct nlattr *attr, u16 *proto, bool *inv) { struct nlattr *tb[NFTA_RULE_COMPAT_MAX+1]; u32 flags; @@ -203,7 +203,7 @@ nft_target_init(const struct nft_ctx *ctx, const struct nft_expr *expr, struct xt_target *target = expr->ops->data; struct xt_tgchk_param par; size_t size = XT_ALIGN(nla_len(tb[NFTA_TARGET_INFO])); - u8 proto = 0; + u16 proto = 0; bool inv = false; union nft_entry e = {}; int ret; @@ -334,7 +334,7 @@ static const struct nla_policy nft_match_policy[NFTA_MATCH_MAX + 1] = { static void nft_match_set_mtchk_param(struct xt_mtchk_param *par, const struct nft_ctx *ctx, struct xt_match *match, void *info, - union nft_entry *entry, u8 proto, bool inv) + union nft_entry *entry, u16 proto, bool inv) { par->net = ctx->net; par->table = ctx->table->name; @@ -348,7 +348,7 @@ nft_match_set_mtchk_param(struct xt_mtchk_param *par, const struct nft_ctx *ctx, entry->e6.ipv6.invflags = inv ? IP6T_INV_PROTO : 0; break; case NFPROTO_BRIDGE: - entry->ebt.ethproto = proto; + entry->ebt.ethproto = (__force __be16)proto; entry->ebt.invflags = inv ? EBT_IPROTO : 0; break; } @@ -385,7 +385,7 @@ nft_match_init(const struct nft_ctx *ctx, const struct nft_expr *expr, struct xt_match *match = expr->ops->data; struct xt_mtchk_param par; size_t size = XT_ALIGN(nla_len(tb[NFTA_MATCH_INFO])); - u8 proto = 0; + u16 proto = 0; bool inv = false; union nft_entry e = {}; int ret; -- cgit v1.2.3 From 02263db00b6cb98701332aa257c07ca549c2324b Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 20 Feb 2015 17:11:10 +0100 Subject: netfilter: nf_tables: fix addition/deletion of elements from commit/abort We have several problems in this path: 1) There is a use-after-free when removing individual elements from the commit path. 2) We have to uninit() the data part of the element from the abort path to avoid a chain refcount leak. 3) We have to check for set->flags to see if there's a mapping, instead of the element flags. 4) We have to check for !(flags & NFT_SET_ELEM_INTERVAL_END) to skip elements that are part of the interval that have no data part, so they don't need to be uninit(). Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-) (limited to 'net') diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 199fd0f27b0e..a8c94620f20e 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -3612,12 +3612,11 @@ static int nf_tables_commit(struct sk_buff *skb) &te->elem, NFT_MSG_DELSETELEM, 0); te->set->ops->get(te->set, &te->elem); - te->set->ops->remove(te->set, &te->elem); nft_data_uninit(&te->elem.key, NFT_DATA_VALUE); - if (te->elem.flags & NFT_SET_MAP) { - nft_data_uninit(&te->elem.data, - te->set->dtype); - } + if (te->set->flags & NFT_SET_MAP && + !(te->elem.flags & NFT_SET_ELEM_INTERVAL_END)) + nft_data_uninit(&te->elem.data, te->set->dtype); + te->set->ops->remove(te->set, &te->elem); nft_trans_destroy(trans); break; } @@ -3658,7 +3657,7 @@ static int nf_tables_abort(struct sk_buff *skb) { struct net *net = sock_net(skb->sk); struct nft_trans *trans, *next; - struct nft_set *set; + struct nft_trans_elem *te; list_for_each_entry_safe(trans, next, &net->nft.commit_list, list) { switch (trans->msg_type) { @@ -3719,9 +3718,13 @@ static int nf_tables_abort(struct sk_buff *skb) break; case NFT_MSG_NEWSETELEM: nft_trans_elem_set(trans)->nelems--; - set = nft_trans_elem_set(trans); - set->ops->get(set, &nft_trans_elem(trans)); - set->ops->remove(set, &nft_trans_elem(trans)); + te = (struct nft_trans_elem *)trans->data; + te->set->ops->get(te->set, &te->elem); + nft_data_uninit(&te->elem.key, NFT_DATA_VALUE); + if (te->set->flags & NFT_SET_MAP && + !(te->elem.flags & NFT_SET_ELEM_INTERVAL_END)) + nft_data_uninit(&te->elem.data, te->set->dtype); + te->set->ops->remove(te->set, &te->elem); nft_trans_destroy(trans); break; case NFT_MSG_DELSETELEM: -- cgit v1.2.3 From 528c943f3bb919aef75ab2fff4f00176f09a4019 Mon Sep 17 00:00:00 2001 From: Julian Anastasov Date: Sat, 21 Feb 2015 21:03:10 +0200 Subject: ipvs: add missing ip_vs_pe_put in sync code ip_vs_conn_fill_param_sync() gets in param.pe a module reference for persistence engine from __ip_vs_pe_getbyname() but forgets to put it. Problem occurs in backup for sync protocol v1 (2.6.39). Also, pe_data usually comes in sync messages for connection templates and ip_vs_conn_new() copies the pointer only in this case. Make sure pe_data is not leaked if it comes unexpectedly for normal connections. Leak can happen only if bogus messages are sent to backup server. Fixes: fe5e7a1efb66 ("IPVS: Backup, Adding Version 1 receive capability") Signed-off-by: Julian Anastasov Signed-off-by: Simon Horman --- net/netfilter/ipvs/ip_vs_sync.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'net') diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index c47ffd7a0a70..d93ceeb3ef04 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -896,6 +896,8 @@ static void ip_vs_proc_conn(struct net *net, struct ip_vs_conn_param *param, IP_VS_DBG(2, "BACKUP, add new conn. failed\n"); return; } + if (!(flags & IP_VS_CONN_F_TEMPLATE)) + kfree(param->pe_data); } if (opt) @@ -1169,6 +1171,7 @@ static inline int ip_vs_proc_sync_conn(struct net *net, __u8 *p, __u8 *msg_end) (opt_flags & IPVS_OPT_F_SEQ_DATA ? &opt : NULL) ); #endif + ip_vs_pe_put(param.pe); return 0; /* Error exit */ out: -- cgit v1.2.3 From 8670c3a55e91cb27a4b4d4d4c4fa35b0149e1abf Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Tue, 3 Mar 2015 20:04:18 +0000 Subject: netfilter: nf_tables: fix transaction race condition A race condition exists in the rule transaction code for rules that get added and removed within the same transaction. The new rule starts out as inactive in the current and active in the next generation and is inserted into the ruleset. When it is deleted, it is additionally set to inactive in the next generation as well. On commit the next generation is begun, then the actions are finalized. For the new rule this would mean clearing out the inactive bit for the previously current, now next generation. However nft_rule_clear() clears out the bits for *both* generations, activating the rule in the current generation, where it should be deactivated due to being deleted. The rule will thus be active until the deletion is finalized, removing the rule from the ruleset. Similarly, when aborting a transaction for the same case, the undo of insertion will remove it from the RCU protected rule list, the deletion will clear out all bits. However until the next RCU synchronization after all operations have been undone, the rule is active on CPUs which can still see the rule on the list. Generally, there may never be any modifications of the current generations' inactive bit since this defeats the entire purpose of atomicity. Change nft_rule_clear() to only touch the next generations bit to fix this. Signed-off-by: Patrick McHardy Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net') diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index a8c94620f20e..6fb532bf0fdb 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -227,7 +227,7 @@ nft_rule_deactivate_next(struct net *net, struct nft_rule *rule) static inline void nft_rule_clear(struct net *net, struct nft_rule *rule) { - rule->genmask = 0; + rule->genmask &= ~(1 << gencursor_next(net)); } static int -- cgit v1.2.3 From 9889840f5988ecfd43b00c9abb83c1804e21406b Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Tue, 3 Mar 2015 20:04:19 +0000 Subject: netfilter: nf_tables: check for overflow of rule dlen field Check that the space required for the expressions doesn't exceed the size of the dlen field, which would lead to the iterators crashing. Signed-off-by: Patrick McHardy Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'net') diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 6fb532bf0fdb..7baafd5ab520 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -1968,6 +1968,10 @@ static int nf_tables_newrule(struct sock *nlsk, struct sk_buff *skb, n++; } } + /* Check for overflow of dlen field */ + err = -EFBIG; + if (size >= 1 << 12) + goto err1; if (nla[NFTA_RULE_USERDATA]) ulen = nla_len(nla[NFTA_RULE_USERDATA]); -- cgit v1.2.3 From 86f1ec32318159a24de349f0a38e79b9d2b3131a Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Tue, 3 Mar 2015 20:04:20 +0000 Subject: netfilter: nf_tables: fix userdata length overflow The NFT_USERDATA_MAXLEN is defined to 256, however we only have a u8 to store its size. Introduce a struct nft_userdata which contains a length field and indicate its presence using a single bit in the rule. The length field of struct nft_userdata is also a u8, however we don't store zero sized data, so the actual length is udata->len + 1. Signed-off-by: Patrick McHardy Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_tables.h | 22 +++++++++++++++++++--- net/netfilter/nf_tables_api.c | 28 +++++++++++++++++++--------- 2 files changed, 38 insertions(+), 12 deletions(-) (limited to 'net') diff --git a/include/net/netfilter/nf_tables.h b/include/net/netfilter/nf_tables.h index 9eaaa7884586..decb9a095ae7 100644 --- a/include/net/netfilter/nf_tables.h +++ b/include/net/netfilter/nf_tables.h @@ -119,6 +119,22 @@ int nft_validate_data_load(const struct nft_ctx *ctx, enum nft_registers reg, const struct nft_data *data, enum nft_data_types type); + +/** + * struct nft_userdata - user defined data associated with an object + * + * @len: length of the data + * @data: content + * + * The presence of user data is indicated in an object specific fashion, + * so a length of zero can't occur and the value "len" indicates data + * of length len + 1. + */ +struct nft_userdata { + u8 len; + unsigned char data[0]; +}; + /** * struct nft_set_elem - generic representation of set elements * @@ -380,7 +396,7 @@ static inline void *nft_expr_priv(const struct nft_expr *expr) * @handle: rule handle * @genmask: generation mask * @dlen: length of expression data - * @ulen: length of user data (used for comments) + * @udata: user data is appended to the rule * @data: expression data */ struct nft_rule { @@ -388,7 +404,7 @@ struct nft_rule { u64 handle:42, genmask:2, dlen:12, - ulen:8; + udata:1; unsigned char data[] __attribute__((aligned(__alignof__(struct nft_expr)))); }; @@ -476,7 +492,7 @@ static inline struct nft_expr *nft_expr_last(const struct nft_rule *rule) return (struct nft_expr *)&rule->data[rule->dlen]; } -static inline void *nft_userdata(const struct nft_rule *rule) +static inline struct nft_userdata *nft_userdata(const struct nft_rule *rule) { return (void *)&rule->data[rule->dlen]; } diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 7baafd5ab520..74e4b876c96e 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -1711,9 +1711,12 @@ static int nf_tables_fill_rule_info(struct sk_buff *skb, struct net *net, } nla_nest_end(skb, list); - if (rule->ulen && - nla_put(skb, NFTA_RULE_USERDATA, rule->ulen, nft_userdata(rule))) - goto nla_put_failure; + if (rule->udata) { + struct nft_userdata *udata = nft_userdata(rule); + if (nla_put(skb, NFTA_RULE_USERDATA, udata->len + 1, + udata->data) < 0) + goto nla_put_failure; + } nlmsg_end(skb, nlh); return 0; @@ -1896,11 +1899,12 @@ static int nf_tables_newrule(struct sock *nlsk, struct sk_buff *skb, struct nft_table *table; struct nft_chain *chain; struct nft_rule *rule, *old_rule = NULL; + struct nft_userdata *udata; struct nft_trans *trans = NULL; struct nft_expr *expr; struct nft_ctx ctx; struct nlattr *tmp; - unsigned int size, i, n, ulen = 0; + unsigned int size, i, n, ulen = 0, usize = 0; int err, rem; bool create; u64 handle, pos_handle; @@ -1973,11 +1977,14 @@ static int nf_tables_newrule(struct sock *nlsk, struct sk_buff *skb, if (size >= 1 << 12) goto err1; - if (nla[NFTA_RULE_USERDATA]) + if (nla[NFTA_RULE_USERDATA]) { ulen = nla_len(nla[NFTA_RULE_USERDATA]); + if (ulen > 0) + usize = sizeof(struct nft_userdata) + ulen; + } err = -ENOMEM; - rule = kzalloc(sizeof(*rule) + size + ulen, GFP_KERNEL); + rule = kzalloc(sizeof(*rule) + size + usize, GFP_KERNEL); if (rule == NULL) goto err1; @@ -1985,10 +1992,13 @@ static int nf_tables_newrule(struct sock *nlsk, struct sk_buff *skb, rule->handle = handle; rule->dlen = size; - rule->ulen = ulen; + rule->udata = ulen ? 1 : 0; - if (ulen) - nla_memcpy(nft_userdata(rule), nla[NFTA_RULE_USERDATA], ulen); + if (ulen) { + udata = nft_userdata(rule); + udata->len = ulen - 1; + nla_memcpy(udata->data, nla[NFTA_RULE_USERDATA], ulen); + } expr = nft_expr_first(rule); for (i = 0; i < n; i++) { -- cgit v1.2.3 From 59900e0a019e7c2bdb7809a03ed5742d311b15b3 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Wed, 4 Mar 2015 17:55:27 +0100 Subject: netfilter: nf_tables: fix error handling of rule replacement In general, if a transaction object is added to the list successfully, we can rely on the abort path to undo what we've done. This allows us to simplify the error handling of the rule replacement path in nf_tables_newrule(). This implicitly fixes an unnecessary removal of the old rule, which needs to be left in place if we fail to replace. Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'net') diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 74e4b876c96e..6ab777912237 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -2045,12 +2045,6 @@ static int nf_tables_newrule(struct sock *nlsk, struct sk_buff *skb, err3: list_del_rcu(&rule->list); - if (trans) { - list_del_rcu(&nft_trans_rule(trans)->list); - nft_rule_clear(net, nft_trans_rule(trans)); - nft_trans_destroy(trans); - chain->use++; - } err2: nf_tables_rule_destroy(&ctx, rule); err1: -- cgit v1.2.3 From 9145736d4862145684009d6a72a6e61324a9439e Mon Sep 17 00:00:00 2001 From: Lorenzo Colitti Date: Tue, 3 Mar 2015 23:16:16 +0900 Subject: net: ping: Return EAFNOSUPPORT when appropriate. 1. For an IPv4 ping socket, ping_check_bind_addr does not check the family of the socket address that's passed in. Instead, make it behave like inet_bind, which enforces either that the address family is AF_INET, or that the family is AF_UNSPEC and the address is 0.0.0.0. 2. For an IPv6 ping socket, ping_check_bind_addr returns EINVAL if the socket family is not AF_INET6. Return EAFNOSUPPORT instead, for consistency with inet6_bind. 3. Make ping_v4_sendmsg and ping_v6_sendmsg return EAFNOSUPPORT instead of EINVAL if an incorrect socket address structure is passed in. 4. Make IPv6 ping sockets be IPv6-only. The code does not support IPv4, and it cannot easily be made to support IPv4 because the protocol numbers for ICMP and ICMPv6 are different. This makes connect(::ffff:192.0.2.1) fail with EAFNOSUPPORT instead of making the socket unusable. Among other things, this fixes an oops that can be triggered by: int s = socket(AF_INET, SOCK_DGRAM, IPPROTO_ICMP); struct sockaddr_in6 sin6 = { .sin6_family = AF_INET6, .sin6_addr = in6addr_any, }; bind(s, (struct sockaddr *) &sin6, sizeof(sin6)); Change-Id: If06ca86d9f1e4593c0d6df174caca3487c57a241 Signed-off-by: Lorenzo Colitti Signed-off-by: David S. Miller --- net/ipv4/ping.c | 12 ++++++++++-- net/ipv6/ping.c | 5 +++-- 2 files changed, 13 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c index e9f66e1cda50..208d5439e59b 100644 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@ -259,6 +259,9 @@ int ping_init_sock(struct sock *sk) kgid_t low, high; int ret = 0; + if (sk->sk_family == AF_INET6) + sk->sk_ipv6only = 1; + inet_get_ping_group_range_net(net, &low, &high); if (gid_lte(low, group) && gid_lte(group, high)) return 0; @@ -305,6 +308,11 @@ static int ping_check_bind_addr(struct sock *sk, struct inet_sock *isk, if (addr_len < sizeof(*addr)) return -EINVAL; + if (addr->sin_family != AF_INET && + !(addr->sin_family == AF_UNSPEC && + addr->sin_addr.s_addr == htonl(INADDR_ANY))) + return -EAFNOSUPPORT; + pr_debug("ping_check_bind_addr(sk=%p,addr=%pI4,port=%d)\n", sk, &addr->sin_addr.s_addr, ntohs(addr->sin_port)); @@ -330,7 +338,7 @@ static int ping_check_bind_addr(struct sock *sk, struct inet_sock *isk, return -EINVAL; if (addr->sin6_family != AF_INET6) - return -EINVAL; + return -EAFNOSUPPORT; pr_debug("ping_check_bind_addr(sk=%p,addr=%pI6c,port=%d)\n", sk, addr->sin6_addr.s6_addr, ntohs(addr->sin6_port)); @@ -716,7 +724,7 @@ static int ping_v4_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *m if (msg->msg_namelen < sizeof(*usin)) return -EINVAL; if (usin->sin_family != AF_INET) - return -EINVAL; + return -EAFNOSUPPORT; daddr = usin->sin_addr.s_addr; /* no remote port */ } else { diff --git a/net/ipv6/ping.c b/net/ipv6/ping.c index bd46f736f61d..a2dfff6ff227 100644 --- a/net/ipv6/ping.c +++ b/net/ipv6/ping.c @@ -102,9 +102,10 @@ int ping_v6_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, if (msg->msg_name) { DECLARE_SOCKADDR(struct sockaddr_in6 *, u, msg->msg_name); - if (msg->msg_namelen < sizeof(struct sockaddr_in6) || - u->sin6_family != AF_INET6) { + if (msg->msg_namelen < sizeof(*u)) return -EINVAL; + if (u->sin6_family != AF_INET6) { + return -EAFNOSUPPORT; } if (sk->sk_bound_dev_if && sk->sk_bound_dev_if != u->sin6_scope_id) { -- cgit v1.2.3 From 3e32e733d1bbb3f227259dc782ef01d5706bdae0 Mon Sep 17 00:00:00 2001 From: Alexander Drozdov Date: Thu, 5 Mar 2015 10:29:39 +0300 Subject: ipv4: ip_check_defrag should not assume that skb_network_offset is zero ip_check_defrag() may be used by af_packet to defragment outgoing packets. skb_network_offset() of af_packet's outgoing packets is not zero. Signed-off-by: Alexander Drozdov Signed-off-by: David S. Miller --- net/ipv4/ip_fragment.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c index 2c8d98e728c0..145a50c4d566 100644 --- a/net/ipv4/ip_fragment.c +++ b/net/ipv4/ip_fragment.c @@ -659,27 +659,30 @@ EXPORT_SYMBOL(ip_defrag); struct sk_buff *ip_check_defrag(struct sk_buff *skb, u32 user) { struct iphdr iph; + int netoff; u32 len; if (skb->protocol != htons(ETH_P_IP)) return skb; - if (skb_copy_bits(skb, 0, &iph, sizeof(iph)) < 0) + netoff = skb_network_offset(skb); + + if (skb_copy_bits(skb, netoff, &iph, sizeof(iph)) < 0) return skb; if (iph.ihl < 5 || iph.version != 4) return skb; len = ntohs(iph.tot_len); - if (skb->len < len || len < (iph.ihl * 4)) + if (skb->len < netoff + len || len < (iph.ihl * 4)) return skb; if (ip_is_fragment(&iph)) { skb = skb_share_check(skb, GFP_ATOMIC); if (skb) { - if (!pskb_may_pull(skb, iph.ihl*4)) + if (!pskb_may_pull(skb, netoff + iph.ihl * 4)) return skb; - if (pskb_trim_rcsum(skb, len)) + if (pskb_trim_rcsum(skb, netoff + len)) return skb; memset(IPCB(skb), 0, sizeof(struct inet_skb_parm)); if (ip_defrag(skb, user)) -- cgit v1.2.3 From 6c09fa09d468d730eecd7122122175da772d3b09 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 5 Mar 2015 08:03:06 -0800 Subject: tcp: align tcp_xmit_size_goal() on tcp_tso_autosize() With some mss values, it is possible tcp_xmit_size_goal() puts one segment more in TSO packet than tcp_tso_autosize(). We send then one TSO packet followed by one single MSS. It is not a serious bug, but we can do slightly better, especially for drivers using netif_set_gso_max_size() to lower gso_max_size. Using same formula avoids these corner cases and makes tcp_xmit_size_goal() a bit faster. Signed-off-by: Eric Dumazet Fixes: 605ad7f184b6 ("tcp: refine TSO autosizing") Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) (limited to 'net') diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 9d72a0fcd928..995a2259bcfc 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -835,17 +835,13 @@ static unsigned int tcp_xmit_size_goal(struct sock *sk, u32 mss_now, int large_allowed) { struct tcp_sock *tp = tcp_sk(sk); - u32 new_size_goal, size_goal, hlen; + u32 new_size_goal, size_goal; if (!large_allowed || !sk_can_gso(sk)) return mss_now; - /* Maybe we should/could use sk->sk_prot->max_header here ? */ - hlen = inet_csk(sk)->icsk_af_ops->net_header_len + - inet_csk(sk)->icsk_ext_hdr_len + - tp->tcp_header_len; - - new_size_goal = sk->sk_gso_max_size - 1 - hlen; + /* Note : tcp_tso_autosize() will eventually split this later */ + new_size_goal = sk->sk_gso_max_size - 1 - MAX_TCP_HEADER; new_size_goal = tcp_bound_to_half_wnd(tp, new_size_goal); /* We try hard to avoid divides here */ -- cgit v1.2.3 From c247f0534cc5a5a547a343903f42295a471844e2 Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Sat, 7 Mar 2015 20:33:22 -0500 Subject: ip: fix error queue empty skb handling When reading from the error queue, msg_name and msg_control are only populated for some errors. A new exception for empty timestamp skbs added a false positive on icmp errors without payload. `traceroute -M udpconn` only displayed gateways that return payload with the icmp error: the embedded network headers are pulled before sock_queue_err_skb, leaving an skb with skb->len == 0 otherwise. Fix this regression by refining when msg_name and msg_control branches are taken. The solutions for the two fields are independent. msg_name only makes sense for errors that configure serr->port and serr->addr_offset. Test the first instead of skb->len. This also fixes another issue. saddr could hold the wrong data, as serr->addr_offset is not initialized in some code paths, pointing to the start of the network header. It is only valid when serr->port is set (non-zero). msg_control support differs between IPv4 and IPv6. IPv4 only honors requests for ICMP and timestamps with SOF_TIMESTAMPING_OPT_CMSG. The skb->len test can simply be removed, because skb->dev is also tested and never true for empty skbs. IPv6 honors requests for all errors aside from local errors and timestamps on empty skbs. In both cases, make the policy more explicit by moving this logic to a new function that decides whether to process msg_control and that optionally prepares the necessary fields in skb->cb[]. After this change, the IPv4 and IPv6 paths are more similar. The last case is rxrpc. Here, simply refine to only match timestamps. Fixes: 49ca0d8bfaf3 ("net-timestamp: no-payload option") Reported-by: Jan Niehusmann Signed-off-by: Willem de Bruijn ---- Changes v1->v2 - fix local origin test inversion in ip6_datagram_support_cmsg - make v4 and v6 code paths more similar by introducing analogous ipv4_datagram_support_cmsg - fix compile bug in rxrpc Signed-off-by: David S. Miller --- net/ipv4/ip_sockglue.c | 33 +++++++++++++++++++++++---------- net/ipv6/datagram.c | 39 ++++++++++++++++++++++++++++----------- net/rxrpc/ar-error.c | 4 ++-- 3 files changed, 53 insertions(+), 23 deletions(-) (limited to 'net') diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c index 31d8c71986b4..5cd99271d3a6 100644 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@ -432,17 +432,32 @@ void ip_local_error(struct sock *sk, int err, __be32 daddr, __be16 port, u32 inf kfree_skb(skb); } -static bool ipv4_pktinfo_prepare_errqueue(const struct sock *sk, - const struct sk_buff *skb, - int ee_origin) +/* IPv4 supports cmsg on all imcp errors and some timestamps + * + * Timestamp code paths do not initialize the fields expected by cmsg: + * the PKTINFO fields in skb->cb[]. Fill those in here. + */ +static bool ipv4_datagram_support_cmsg(const struct sock *sk, + struct sk_buff *skb, + int ee_origin) { - struct in_pktinfo *info = PKTINFO_SKB_CB(skb); + struct in_pktinfo *info; + + if (ee_origin == SO_EE_ORIGIN_ICMP) + return true; - if ((ee_origin != SO_EE_ORIGIN_TIMESTAMPING) || - (!(sk->sk_tsflags & SOF_TIMESTAMPING_OPT_CMSG)) || + if (ee_origin == SO_EE_ORIGIN_LOCAL) + return false; + + /* Support IP_PKTINFO on tstamp packets if requested, to correlate + * timestamp with egress dev. Not possible for packets without dev + * or without payload (SOF_TIMESTAMPING_OPT_TSONLY). + */ + if ((!(sk->sk_tsflags & SOF_TIMESTAMPING_OPT_CMSG)) || (!skb->dev)) return false; + info = PKTINFO_SKB_CB(skb); info->ipi_spec_dst.s_addr = ip_hdr(skb)->saddr; info->ipi_ifindex = skb->dev->ifindex; return true; @@ -483,7 +498,7 @@ int ip_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len) serr = SKB_EXT_ERR(skb); - if (sin && skb->len) { + if (sin && serr->port) { sin->sin_family = AF_INET; sin->sin_addr.s_addr = *(__be32 *)(skb_network_header(skb) + serr->addr_offset); @@ -496,9 +511,7 @@ int ip_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len) sin = &errhdr.offender; memset(sin, 0, sizeof(*sin)); - if (skb->len && - (serr->ee.ee_origin == SO_EE_ORIGIN_ICMP || - ipv4_pktinfo_prepare_errqueue(sk, skb, serr->ee.ee_origin))) { + if (ipv4_datagram_support_cmsg(sk, skb, serr->ee.ee_origin)) { sin->sin_family = AF_INET; sin->sin_addr.s_addr = ip_hdr(skb)->saddr; if (inet_sk(sk)->cmsg_flags) diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c index c215be70cac0..ace8daca5c83 100644 --- a/net/ipv6/datagram.c +++ b/net/ipv6/datagram.c @@ -325,14 +325,34 @@ void ipv6_local_rxpmtu(struct sock *sk, struct flowi6 *fl6, u32 mtu) kfree_skb(skb); } -static void ip6_datagram_prepare_pktinfo_errqueue(struct sk_buff *skb) +/* IPv6 supports cmsg on all origins aside from SO_EE_ORIGIN_LOCAL. + * + * At one point, excluding local errors was a quick test to identify icmp/icmp6 + * errors. This is no longer true, but the test remained, so the v6 stack, + * unlike v4, also honors cmsg requests on all wifi and timestamp errors. + * + * Timestamp code paths do not initialize the fields expected by cmsg: + * the PKTINFO fields in skb->cb[]. Fill those in here. + */ +static bool ip6_datagram_support_cmsg(struct sk_buff *skb, + struct sock_exterr_skb *serr) { - int ifindex = skb->dev ? skb->dev->ifindex : -1; + if (serr->ee.ee_origin == SO_EE_ORIGIN_ICMP || + serr->ee.ee_origin == SO_EE_ORIGIN_ICMP6) + return true; + + if (serr->ee.ee_origin == SO_EE_ORIGIN_LOCAL) + return false; + + if (!skb->dev) + return false; if (skb->protocol == htons(ETH_P_IPV6)) - IP6CB(skb)->iif = ifindex; + IP6CB(skb)->iif = skb->dev->ifindex; else - PKTINFO_SKB_CB(skb)->ipi_ifindex = ifindex; + PKTINFO_SKB_CB(skb)->ipi_ifindex = skb->dev->ifindex; + + return true; } /* @@ -369,7 +389,7 @@ int ipv6_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len) serr = SKB_EXT_ERR(skb); - if (sin && skb->len) { + if (sin && serr->port) { const unsigned char *nh = skb_network_header(skb); sin->sin6_family = AF_INET6; sin->sin6_flowinfo = 0; @@ -394,14 +414,11 @@ int ipv6_recv_error(struct sock *sk, struct msghdr *msg, int len, int *addr_len) memcpy(&errhdr.ee, &serr->ee, sizeof(struct sock_extended_err)); sin = &errhdr.offender; memset(sin, 0, sizeof(*sin)); - if (serr->ee.ee_origin != SO_EE_ORIGIN_LOCAL && skb->len) { + + if (ip6_datagram_support_cmsg(skb, serr)) { sin->sin6_family = AF_INET6; - if (np->rxopt.all) { - if (serr->ee.ee_origin != SO_EE_ORIGIN_ICMP && - serr->ee.ee_origin != SO_EE_ORIGIN_ICMP6) - ip6_datagram_prepare_pktinfo_errqueue(skb); + if (np->rxopt.all) ip6_datagram_recv_common_ctl(sk, msg, skb); - } if (skb->protocol == htons(ETH_P_IPV6)) { sin->sin6_addr = ipv6_hdr(skb)->saddr; if (np->rxopt.all) diff --git a/net/rxrpc/ar-error.c b/net/rxrpc/ar-error.c index 5394b6be46ec..0610efa83d72 100644 --- a/net/rxrpc/ar-error.c +++ b/net/rxrpc/ar-error.c @@ -42,7 +42,8 @@ void rxrpc_UDP_error_report(struct sock *sk) _leave("UDP socket errqueue empty"); return; } - if (!skb->len) { + serr = SKB_EXT_ERR(skb); + if (!skb->len && serr->ee.ee_origin == SO_EE_ORIGIN_TIMESTAMPING) { _leave("UDP empty message"); kfree_skb(skb); return; @@ -50,7 +51,6 @@ void rxrpc_UDP_error_report(struct sock *sk) rxrpc_new_skb(skb); - serr = SKB_EXT_ERR(skb); addr = *(__be32 *)(skb_network_header(skb) + serr->addr_offset); port = serr->port; -- cgit v1.2.3 From 969439016d2cf61fef53a973d7e6d2061c3793b1 Mon Sep 17 00:00:00 2001 From: Oliver Hartkopp Date: Mon, 23 Feb 2015 20:37:54 +0100 Subject: can: add missing initialisations in CAN related skbuffs When accessing CAN network interfaces with AF_PACKET sockets e.g. by dhclient this can lead to a skb_under_panic due to missing skb initialisations. Add the missing initialisations at the CAN skbuff creation times on driver level (rx path) and in the network layer (tx path). Reported-by: Austin Schuh Reported-by: Daniel Steer Signed-off-by: Oliver Hartkopp Cc: linux-stable Signed-off-by: Marc Kleine-Budde --- drivers/net/can/dev.c | 8 ++++++++ net/can/af_can.c | 3 +++ 2 files changed, 11 insertions(+) (limited to 'net') diff --git a/drivers/net/can/dev.c b/drivers/net/can/dev.c index 3c82e02e3dae..b0f69248cb71 100644 --- a/drivers/net/can/dev.c +++ b/drivers/net/can/dev.c @@ -579,6 +579,10 @@ struct sk_buff *alloc_can_skb(struct net_device *dev, struct can_frame **cf) skb->pkt_type = PACKET_BROADCAST; skb->ip_summed = CHECKSUM_UNNECESSARY; + skb_reset_mac_header(skb); + skb_reset_network_header(skb); + skb_reset_transport_header(skb); + can_skb_reserve(skb); can_skb_prv(skb)->ifindex = dev->ifindex; @@ -603,6 +607,10 @@ struct sk_buff *alloc_canfd_skb(struct net_device *dev, skb->pkt_type = PACKET_BROADCAST; skb->ip_summed = CHECKSUM_UNNECESSARY; + skb_reset_mac_header(skb); + skb_reset_network_header(skb); + skb_reset_transport_header(skb); + can_skb_reserve(skb); can_skb_prv(skb)->ifindex = dev->ifindex; diff --git a/net/can/af_can.c b/net/can/af_can.c index 66e08040ced7..32d710eaf1fc 100644 --- a/net/can/af_can.c +++ b/net/can/af_can.c @@ -259,6 +259,9 @@ int can_send(struct sk_buff *skb, int loop) goto inval_skb; } + skb->ip_summed = CHECKSUM_UNNECESSARY; + + skb_reset_mac_header(skb); skb_reset_network_header(skb); skb_reset_transport_header(skb); -- cgit v1.2.3 From 82f17091e68254d1612b42cf23291cad63cfaf04 Mon Sep 17 00:00:00 2001 From: Francesco Ruggeri Date: Mon, 9 Mar 2015 11:51:04 -0700 Subject: net: delete stale packet_mclist entries When an interface is deleted from a net namespace the ifindex in the corresponding entries in PF_PACKET sockets' mclists becomes stale. This can create inconsistencies if later an interface with the same ifindex is moved from a different namespace (not that unlikely since ifindexes are per-namespace). In particular we saw problems with dev->promiscuity, resulting in "promiscuity touches roof, set promiscuity failed. promiscuity feature of device might be broken" warnings and EOVERFLOW failures of setsockopt(PACKET_ADD_MEMBERSHIP). This patch deletes the mclist entries for interfaces that are deleted. Since this now causes setsockopt(PACKET_DROP_MEMBERSHIP) to fail with EADDRNOTAVAIL if called after the interface is deleted, also make packet_mc_drop not fail. Signed-off-by: Francesco Ruggeri Signed-off-by: David S. Miller --- net/packet/af_packet.c | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) (limited to 'net') diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 5bf1e968a728..f8db7064d81c 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -3123,11 +3123,18 @@ static int packet_dev_mc(struct net_device *dev, struct packet_mclist *i, return 0; } -static void packet_dev_mclist(struct net_device *dev, struct packet_mclist *i, int what) +static void packet_dev_mclist_delete(struct net_device *dev, + struct packet_mclist **mlp) { - for ( ; i; i = i->next) { - if (i->ifindex == dev->ifindex) - packet_dev_mc(dev, i, what); + struct packet_mclist *ml; + + while ((ml = *mlp) != NULL) { + if (ml->ifindex == dev->ifindex) { + packet_dev_mc(dev, ml, -1); + *mlp = ml->next; + kfree(ml); + } else + mlp = &ml->next; } } @@ -3204,12 +3211,11 @@ static int packet_mc_drop(struct sock *sk, struct packet_mreq_max *mreq) packet_dev_mc(dev, ml, -1); kfree(ml); } - rtnl_unlock(); - return 0; + break; } } rtnl_unlock(); - return -EADDRNOTAVAIL; + return 0; } static void packet_flush_mclist(struct sock *sk) @@ -3559,7 +3565,7 @@ static int packet_notifier(struct notifier_block *this, switch (msg) { case NETDEV_UNREGISTER: if (po->mclist) - packet_dev_mclist(dev, po->mclist, -1); + packet_dev_mclist_delete(dev, &po->mclist); /* fallthrough */ case NETDEV_DOWN: -- cgit v1.2.3 From e6441bae326271090755e1707196ad05aa1dc703 Mon Sep 17 00:00:00 2001 From: Jon Paul Maloy Date: Mon, 9 Mar 2015 16:16:22 -0400 Subject: tipc: fix bug in link failover handling In commit c637c1035534867b85b78b453c38c495b58e2c5a ("tipc: resolve race problem at unicast message reception") we introduced a new mechanism for delivering buffers upwards from link to socket layer. That code contains a bug in how we handle the new link input queue during failover. When a link is reset, some of its users may be blocked because of congestion, and in order to resolve this, we add any pending wakeup pseudo messages to the link's input queue, and deliver them to the socket. This misses the case where the other, remaining link also may have congested users. Currently, the owner node's reference to the remaining link's input queue is unconditionally overwritten by the reset link's input queue. This has the effect that wakeup events from the remaining link may be unduely delayed (but not lost) for a potentially long period. We fix this by adding the pending events from the reset link to the input queue that is currently referenced by the node, whichever one it is. This commit should be applied to both net and net-next. Signed-off-by: Jon Maloy Signed-off-by: David S. Miller --- net/tipc/link.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'net') diff --git a/net/tipc/link.c b/net/tipc/link.c index a4cf364316de..14f09b3cb87c 100644 --- a/net/tipc/link.c +++ b/net/tipc/link.c @@ -464,10 +464,11 @@ void tipc_link_reset(struct tipc_link *l_ptr) /* Clean up all queues, except inputq: */ __skb_queue_purge(&l_ptr->outqueue); __skb_queue_purge(&l_ptr->deferred_queue); - skb_queue_splice_init(&l_ptr->wakeupq, &l_ptr->inputq); - if (!skb_queue_empty(&l_ptr->inputq)) + if (!owner->inputq) + owner->inputq = &l_ptr->inputq; + skb_queue_splice_init(&l_ptr->wakeupq, owner->inputq); + if (!skb_queue_empty(owner->inputq)) owner->action_flags |= TIPC_MSG_EVT; - owner->inputq = &l_ptr->inputq; l_ptr->next_out = NULL; l_ptr->unacked_window = 0; l_ptr->checkpoint = 1; -- cgit v1.2.3