Hi, The following patchset contains Netfilter fixes for net: 1) Missing netns context in arp_tables, from Florian Westphal. 2) Underflow in flowtable reference counter, from wenxu. 3) Fix incorrect ethernet destination address in flowtable offload, from wenxu. 4) Check for status of neighbour entry, from wenxu. 5) Fix NAT port mangling, from wenxu. 6) Unbind callbacks from destroy path to cleanup hardware properly on flowtable removal. 7) Fix missing casting statistics timestamp, add nf_flowtable_time_stamp and use it. 8) NULL pointer exception when timeout argument is null in conntrack dccp and sctp protocol helpers, from Florian Westphal. 9) Possible nul-dereference in ipset with IPSET_ATTR_LINENO, also from Florian. You can pull these changes from: git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git Thanks. ---------------------------------------------------------------- The following changes since commit bd6f48546b9cb7a785344fc78058c420923d7ed8: net: stmmac: dwmac-meson8b: Fix the RGMII TX delay on Meson8b/8m2 SoCs (2019-12-27 16:37:07 -0800) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git HEAD for you to fetch changes up to 22dad713b8a5ff488e07b821195270672f486eb2: netfilter: ipset: avoid null deref when IPSET_ATTR_LINENO is present (2020-01-08 23:31:46 +0100) ---------------------------------------------------------------- Florian Westphal (3): netfilter: arp_tables: init netns pointer in xt_tgchk_param struct netfilter: conntrack: dccp, sctp: handle null timeout argument netfilter: ipset: avoid null deref when IPSET_ATTR_LINENO is present Pablo Neira Ayuso (2): netfilter: nf_tables: unbind callbacks from flowtable destroy path netfilter: flowtable: add nf_flowtable_time_stamp wenxu (4): netfilter: nft_flow_offload: fix underflow in flowtable reference counter netfilter: nf_flow_table_offload: fix incorrect ethernet dst address netfilter: nf_flow_table_offload: check the status of dst_neigh netfilter: nf_flow_table_offload: fix the nat port mangle. include/net/netfilter/nf_flow_table.h | 6 ++++ net/ipv4/netfilter/arp_tables.c | 27 ++++++++++-------- net/netfilter/ipset/ip_set_core.c | 3 +- net/netfilter/nf_conntrack_proto_dccp.c | 3 ++ net/netfilter/nf_conntrack_proto_sctp.c | 3 ++ net/netfilter/nf_flow_table_core.c | 7 +---- net/netfilter/nf_flow_table_ip.c | 4 +-- net/netfilter/nf_flow_table_offload.c | 50 ++++++++++++++++++++++++--------- net/netfilter/nf_tables_api.c | 8 ++++-- net/netfilter/nft_flow_offload.c | 3 -- 10 files changed, 75 insertions(+), 39 deletions(-)
From: Florian Westphal <fw@strlen.de> We get crash when the targets checkentry function tries to make use of the network namespace pointer for arptables. When the net pointer got added back in 2010, only ip/ip6/ebtables were changed to initialize it, so arptables has this set to NULL. This isn't a problem for normal arptables because no existing arptables target has a checkentry function that makes use of par->net. However, direct users of the setsockopt interface can provide any target they want as long as its registered for ARP or UNPSEC protocols. syzkaller managed to send a semi-valid arptables rule for RATEEST target which is enough to trigger NULL deref: kasan: GPF could be caused by NULL-ptr deref or user memory access general protection fault: 0000 [#1] PREEMPT SMP KASAN RIP: xt_rateest_tg_checkentry+0x11d/0xb40 net/netfilter/xt_RATEEST.c:109 [..] xt_check_target+0x283/0x690 net/netfilter/x_tables.c:1019 check_target net/ipv4/netfilter/arp_tables.c:399 [inline] find_check_entry net/ipv4/netfilter/arp_tables.c:422 [inline] translate_table+0x1005/0x1d70 net/ipv4/netfilter/arp_tables.c:572 do_replace net/ipv4/netfilter/arp_tables.c:977 [inline] do_arpt_set_ctl+0x310/0x640 net/ipv4/netfilter/arp_tables.c:1456 Fixes: add67461240c1d ("netfilter: add struct net * to target parameters") Reported-by: syzbot+d7358a458d8a81aee898@syzkaller.appspotmail.com Signed-off-by: Florian Westphal <fw@strlen.de> Acked-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> --- net/ipv4/netfilter/arp_tables.c | 27 ++++++++++++++++----------- 1 file changed, 16 insertions(+), 11 deletions(-) diff --git a/net/ipv4/netfilter/arp_tables.c b/net/ipv4/netfilter/arp_tables.c index 214154b47d56..069f72edb264 100644 --- a/net/ipv4/netfilter/arp_tables.c +++ b/net/ipv4/netfilter/arp_tables.c @@ -384,10 +384,11 @@ next: ; return 1; } -static inline int check_target(struct arpt_entry *e, const char *name) +static int check_target(struct arpt_entry *e, struct net *net, const char *name) { struct xt_entry_target *t = arpt_get_target(e); struct xt_tgchk_param par = { + .net = net, .table = name, .entryinfo = e, .target = t->u.kernel.target, @@ -399,8 +400,9 @@ static inline int check_target(struct arpt_entry *e, const char *name) return xt_check_target(&par, t->u.target_size - sizeof(*t), 0, false); } -static inline int -find_check_entry(struct arpt_entry *e, const char *name, unsigned int size, +static int +find_check_entry(struct arpt_entry *e, struct net *net, const char *name, + unsigned int size, struct xt_percpu_counter_alloc_state *alloc_state) { struct xt_entry_target *t; @@ -419,7 +421,7 @@ find_check_entry(struct arpt_entry *e, const char *name, unsigned int size, } t->u.kernel.target = target; - ret = check_target(e, name); + ret = check_target(e, net, name); if (ret) goto err; return 0; @@ -512,7 +514,9 @@ static inline void cleanup_entry(struct arpt_entry *e) /* Checks and translates the user-supplied table segment (held in * newinfo). */ -static int translate_table(struct xt_table_info *newinfo, void *entry0, +static int translate_table(struct net *net, + struct xt_table_info *newinfo, + void *entry0, const struct arpt_replace *repl) { struct xt_percpu_counter_alloc_state alloc_state = { 0 }; @@ -569,7 +573,7 @@ static int translate_table(struct xt_table_info *newinfo, void *entry0, /* Finally, each sanity check must pass */ i = 0; xt_entry_foreach(iter, entry0, newinfo->size) { - ret = find_check_entry(iter, repl->name, repl->size, + ret = find_check_entry(iter, net, repl->name, repl->size, &alloc_state); if (ret != 0) break; @@ -974,7 +978,7 @@ static int do_replace(struct net *net, const void __user *user, goto free_newinfo; } - ret = translate_table(newinfo, loc_cpu_entry, &tmp); + ret = translate_table(net, newinfo, loc_cpu_entry, &tmp); if (ret != 0) goto free_newinfo; @@ -1149,7 +1153,8 @@ compat_copy_entry_from_user(struct compat_arpt_entry *e, void **dstptr, } } -static int translate_compat_table(struct xt_table_info **pinfo, +static int translate_compat_table(struct net *net, + struct xt_table_info **pinfo, void **pentry0, const struct compat_arpt_replace *compatr) { @@ -1217,7 +1222,7 @@ static int translate_compat_table(struct xt_table_info **pinfo, repl.num_counters = 0; repl.counters = NULL; repl.size = newinfo->size; - ret = translate_table(newinfo, entry1, &repl); + ret = translate_table(net, newinfo, entry1, &repl); if (ret) goto free_newinfo; @@ -1270,7 +1275,7 @@ static int compat_do_replace(struct net *net, void __user *user, goto free_newinfo; } - ret = translate_compat_table(&newinfo, &loc_cpu_entry, &tmp); + ret = translate_compat_table(net, &newinfo, &loc_cpu_entry, &tmp); if (ret != 0) goto free_newinfo; @@ -1546,7 +1551,7 @@ int arpt_register_table(struct net *net, loc_cpu_entry = newinfo->entries; memcpy(loc_cpu_entry, repl->entries, repl->size); - ret = translate_table(newinfo, loc_cpu_entry, repl); + ret = translate_table(net, newinfo, loc_cpu_entry, repl); if (ret != 0) goto out_free; -- 2.11.0
From: wenxu <wenxu@ucloud.cn> The .deactivate and .activate interfaces already deal with the reference counter. Otherwise, this results in spurious "Device is busy" errors. Fixes: a3c90f7a2323 ("netfilter: nf_tables: flow offload expression") Signed-off-by: wenxu <wenxu@ucloud.cn> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> --- net/netfilter/nft_flow_offload.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/net/netfilter/nft_flow_offload.c b/net/netfilter/nft_flow_offload.c index dd82ff2ee19f..b70b48996801 100644 --- a/net/netfilter/nft_flow_offload.c +++ b/net/netfilter/nft_flow_offload.c @@ -200,9 +200,6 @@ static void nft_flow_offload_activate(const struct nft_ctx *ctx, static void nft_flow_offload_destroy(const struct nft_ctx *ctx, const struct nft_expr *expr) { - struct nft_flow_offload *priv = nft_expr_priv(expr); - - priv->flowtable->use--; nf_ct_netns_put(ctx->net, ctx->family); } -- 2.11.0
From: wenxu <wenxu@ucloud.cn> Ethernet destination for original traffic takes the source ethernet address in the reply direction. For reply traffic, this takes the source ethernet address of the original direction. Fixes: c29f74e0df7a ("netfilter: nf_flow_table: hardware offload support") Signed-off-by: wenxu <wenxu@ucloud.cn> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> --- net/netfilter/nf_flow_table_offload.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index 0d72e5ccb47b..ee9edbe50d4f 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -166,14 +166,16 @@ static int flow_offload_eth_dst(struct net *net, enum flow_offload_tuple_dir dir, struct nf_flow_rule *flow_rule) { - const struct flow_offload_tuple *tuple = &flow->tuplehash[dir].tuple; struct flow_action_entry *entry0 = flow_action_entry_next(flow_rule); struct flow_action_entry *entry1 = flow_action_entry_next(flow_rule); + const void *daddr = &flow->tuplehash[!dir].tuple.src_v4; + const struct dst_entry *dst_cache; struct neighbour *n; u32 mask, val; u16 val16; - n = dst_neigh_lookup(tuple->dst_cache, &tuple->dst_v4); + dst_cache = flow->tuplehash[dir].tuple.dst_cache; + n = dst_neigh_lookup(dst_cache, daddr); if (!n) return -ENOENT; -- 2.11.0
From: wenxu <wenxu@ucloud.cn> It is better to get the dst_neigh with neigh->lock and check the nud_state is VALID. If there is not neigh previous, the lookup will Create a non NUD_VALID with 00:00:00:00:00:00 mac. Fixes: c29f74e0df7a ("netfilter: nf_flow_table: hardware offload support") Signed-off-by: wenxu <wenxu@ucloud.cn> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> --- net/netfilter/nf_flow_table_offload.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index ee9edbe50d4f..92b0bd241073 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -170,8 +170,10 @@ static int flow_offload_eth_dst(struct net *net, struct flow_action_entry *entry1 = flow_action_entry_next(flow_rule); const void *daddr = &flow->tuplehash[!dir].tuple.src_v4; const struct dst_entry *dst_cache; + unsigned char ha[ETH_ALEN]; struct neighbour *n; u32 mask, val; + u8 nud_state; u16 val16; dst_cache = flow->tuplehash[dir].tuple.dst_cache; @@ -179,13 +181,23 @@ static int flow_offload_eth_dst(struct net *net, if (!n) return -ENOENT; + read_lock_bh(&n->lock); + nud_state = n->nud_state; + ether_addr_copy(ha, n->ha); + read_unlock_bh(&n->lock); + + if (!(nud_state & NUD_VALID)) { + neigh_release(n); + return -ENOENT; + } + mask = ~0xffffffff; - memcpy(&val, n->ha, 4); + memcpy(&val, ha, 4); flow_offload_mangle(entry0, FLOW_ACT_MANGLE_HDR_TYPE_ETH, 0, &val, &mask); mask = ~0x0000ffff; - memcpy(&val16, n->ha + 4, 2); + memcpy(&val16, ha + 4, 2); val = val16; flow_offload_mangle(entry1, FLOW_ACT_MANGLE_HDR_TYPE_ETH, 4, &val, &mask); -- 2.11.0
From: wenxu <wenxu@ucloud.cn> Shift on 32-bit word to define the port number depends on the flow direction. Fixes: c29f74e0df7a ("netfilter: nf_flow_table: hardware offload support") Fixes: 7acd9378dc652 ("netfilter: nf_flow_table_offload: Correct memcpy size for flow_overload_mangle()") Signed-off-by: wenxu <wenxu@ucloud.cn> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> --- net/netfilter/nf_flow_table_offload.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index 92b0bd241073..6c162c954c4f 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -349,22 +349,26 @@ static void flow_offload_port_snat(struct net *net, struct nf_flow_rule *flow_rule) { struct flow_action_entry *entry = flow_action_entry_next(flow_rule); - u32 mask = ~htonl(0xffff0000), port; + u32 mask, port; u32 offset; switch (dir) { case FLOW_OFFLOAD_DIR_ORIGINAL: port = ntohs(flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.dst_port); offset = 0; /* offsetof(struct tcphdr, source); */ + port = htonl(port << 16); + mask = ~htonl(0xffff0000); break; case FLOW_OFFLOAD_DIR_REPLY: port = ntohs(flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.src_port); offset = 0; /* offsetof(struct tcphdr, dest); */ + port = htonl(port); + mask = ~htonl(0xffff); break; default: return; } - port = htonl(port << 16); + flow_offload_mangle(entry, flow_offload_l4proto(flow), offset, &port, &mask); } @@ -375,22 +379,26 @@ static void flow_offload_port_dnat(struct net *net, struct nf_flow_rule *flow_rule) { struct flow_action_entry *entry = flow_action_entry_next(flow_rule); - u32 mask = ~htonl(0xffff), port; + u32 mask, port; u32 offset; switch (dir) { case FLOW_OFFLOAD_DIR_ORIGINAL: - port = ntohs(flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.dst_port); - offset = 0; /* offsetof(struct tcphdr, source); */ + port = ntohs(flow->tuplehash[FLOW_OFFLOAD_DIR_REPLY].tuple.src_port); + offset = 0; /* offsetof(struct tcphdr, dest); */ + port = htonl(port); + mask = ~htonl(0xffff); break; case FLOW_OFFLOAD_DIR_REPLY: - port = ntohs(flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.src_port); - offset = 0; /* offsetof(struct tcphdr, dest); */ + port = ntohs(flow->tuplehash[FLOW_OFFLOAD_DIR_ORIGINAL].tuple.dst_port); + offset = 0; /* offsetof(struct tcphdr, source); */ + port = htonl(port << 16); + mask = ~htonl(0xffff0000); break; default: return; } - port = htonl(port); + flow_offload_mangle(entry, flow_offload_l4proto(flow), offset, &port, &mask); } -- 2.11.0
Callback unbinding needs to be done after nf_flow_table_free(), otherwise entries are not removed from the hardware. Update nft_unregister_flowtable_net_hooks() to call nf_unregister_net_hook() instead since the commit/abort paths do not deal with the callback unbinding anymore. Add a comment to nft_flowtable_event() to clarify that flow_offload_netdev_event() already removes the entries before the callback unbinding. Fixes: 8bb69f3b2918 ("netfilter: nf_tables: add flowtable offload control plane") Fixes ff4bf2f42a40 ("netfilter: nf_tables: add nft_unregister_flowtable_hook()") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Acked-by: wenxu <wenxu@ucloud.cn> --- net/netfilter/nf_tables_api.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 273f3838318b..43f05b3acd60 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -5984,6 +5984,7 @@ nft_flowtable_type_get(struct net *net, u8 family) return ERR_PTR(-ENOENT); } +/* Only called from error and netdev event paths. */ static void nft_unregister_flowtable_hook(struct net *net, struct nft_flowtable *flowtable, struct nft_hook *hook) @@ -5999,7 +6000,7 @@ static void nft_unregister_flowtable_net_hooks(struct net *net, struct nft_hook *hook; list_for_each_entry(hook, &flowtable->hook_list, list) - nft_unregister_flowtable_hook(net, flowtable, hook); + nf_unregister_net_hook(net, &hook->ops); } static int nft_register_flowtable_net_hooks(struct net *net, @@ -6448,12 +6449,14 @@ static void nf_tables_flowtable_destroy(struct nft_flowtable *flowtable) { struct nft_hook *hook, *next; + flowtable->data.type->free(&flowtable->data); list_for_each_entry_safe(hook, next, &flowtable->hook_list, list) { + flowtable->data.type->setup(&flowtable->data, hook->ops.dev, + FLOW_BLOCK_UNBIND); list_del_rcu(&hook->list); kfree(hook); } kfree(flowtable->name); - flowtable->data.type->free(&flowtable->data); module_put(flowtable->data.type->owner); kfree(flowtable); } @@ -6497,6 +6500,7 @@ static void nft_flowtable_event(unsigned long event, struct net_device *dev, if (hook->ops.dev != dev) continue; + /* flow_offload_netdev_event() cleans up entries for us. */ nft_unregister_flowtable_hook(dev_net(dev), flowtable, hook); list_del_rcu(&hook->list); kfree_rcu(hook, rcu); -- 2.11.0
This patch adds nf_flowtable_time_stamp and updates the existing code to use it. This patch is also implicitly fixing up hardware statistic fetching via nf_flow_offload_stats() where casting to u32 is missing. Use nf_flow_timeout_delta() to fix this. Fixes: c29f74e0df7a ("netfilter: nf_flow_table: hardware offload support") Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Acked-by: wenxu <wenxu@ucloud.cn> --- include/net/netfilter/nf_flow_table.h | 6 ++++++ net/netfilter/nf_flow_table_core.c | 7 +------ net/netfilter/nf_flow_table_ip.c | 4 ++-- net/netfilter/nf_flow_table_offload.c | 4 ++-- 4 files changed, 11 insertions(+), 10 deletions(-) diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h index f0897b3c97fb..415b8f49d150 100644 --- a/include/net/netfilter/nf_flow_table.h +++ b/include/net/netfilter/nf_flow_table.h @@ -106,6 +106,12 @@ struct flow_offload { }; #define NF_FLOW_TIMEOUT (30 * HZ) +#define nf_flowtable_time_stamp (u32)jiffies + +static inline __s32 nf_flow_timeout_delta(unsigned int timeout) +{ + return (__s32)(timeout - nf_flowtable_time_stamp); +} struct nf_flow_route { struct { diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c index 9889d52eda82..e33a73cb1f42 100644 --- a/net/netfilter/nf_flow_table_core.c +++ b/net/netfilter/nf_flow_table_core.c @@ -134,11 +134,6 @@ static void flow_offload_fixup_tcp(struct ip_ct_tcp *tcp) #define NF_FLOWTABLE_TCP_PICKUP_TIMEOUT (120 * HZ) #define NF_FLOWTABLE_UDP_PICKUP_TIMEOUT (30 * HZ) -static inline __s32 nf_flow_timeout_delta(unsigned int timeout) -{ - return (__s32)(timeout - (u32)jiffies); -} - static void flow_offload_fixup_ct_timeout(struct nf_conn *ct) { const struct nf_conntrack_l4proto *l4proto; @@ -232,7 +227,7 @@ int flow_offload_add(struct nf_flowtable *flow_table, struct flow_offload *flow) { int err; - flow->timeout = (u32)jiffies + NF_FLOW_TIMEOUT; + flow->timeout = nf_flowtable_time_stamp + NF_FLOW_TIMEOUT; err = rhashtable_insert_fast(&flow_table->rhashtable, &flow->tuplehash[0].node, diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index b9e7dd6e60ce..7ea2ddc2aa93 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -280,7 +280,7 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, if (nf_flow_nat_ip(flow, skb, thoff, dir) < 0) return NF_DROP; - flow->timeout = (u32)jiffies + NF_FLOW_TIMEOUT; + flow->timeout = nf_flowtable_time_stamp + NF_FLOW_TIMEOUT; iph = ip_hdr(skb); ip_decrease_ttl(iph); skb->tstamp = 0; @@ -509,7 +509,7 @@ nf_flow_offload_ipv6_hook(void *priv, struct sk_buff *skb, if (nf_flow_nat_ipv6(flow, skb, dir) < 0) return NF_DROP; - flow->timeout = (u32)jiffies + NF_FLOW_TIMEOUT; + flow->timeout = nf_flowtable_time_stamp + NF_FLOW_TIMEOUT; ip6h = ipv6_hdr(skb); ip6h->hop_limit--; skb->tstamp = 0; diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index 6c162c954c4f..d06969af1085 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -781,9 +781,9 @@ void nf_flow_offload_stats(struct nf_flowtable *flowtable, struct flow_offload *flow) { struct flow_offload_work *offload; - s64 delta; + __s32 delta; - delta = flow->timeout - jiffies; + delta = nf_flow_timeout_delta(flow->timeout); if ((delta >= (9 * NF_FLOW_TIMEOUT) / 10) || flow->flags & FLOW_OFFLOAD_HW_DYING) return; -- 2.11.0
From: Florian Westphal <fw@strlen.de> The timeout pointer can be NULL which means we should modify the per-nets timeout instead. All do this, except sctp and dccp which instead give: general protection fault: 0000 [#1] PREEMPT SMP KASAN net/netfilter/nf_conntrack_proto_dccp.c:682 ctnl_timeout_parse_policy+0x150/0x1d0 net/netfilter/nfnetlink_cttimeout.c:67 cttimeout_default_set+0x150/0x1c0 net/netfilter/nfnetlink_cttimeout.c:368 nfnetlink_rcv_msg+0xcf2/0xfb0 net/netfilter/nfnetlink.c:229 netlink_rcv_skb+0x177/0x450 net/netlink/af_netlink.c:2477 Reported-by: syzbot+46a4ad33f345d1dd346e@syzkaller.appspotmail.com Fixes: c779e849608a8 ("netfilter: conntrack: remove get_timeout() indirection") Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> --- net/netfilter/nf_conntrack_proto_dccp.c | 3 +++ net/netfilter/nf_conntrack_proto_sctp.c | 3 +++ 2 files changed, 6 insertions(+) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index b6b14db3955b..b3f4a334f9d7 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -677,6 +677,9 @@ static int dccp_timeout_nlattr_to_obj(struct nlattr *tb[], unsigned int *timeouts = data; int i; + if (!timeouts) + timeouts = dn->dccp_timeout; + /* set default DCCP timeouts. */ for (i=0; i<CT_DCCP_MAX; i++) timeouts[i] = dn->dccp_timeout[i]; diff --git a/net/netfilter/nf_conntrack_proto_sctp.c b/net/netfilter/nf_conntrack_proto_sctp.c index fce3d93f1541..0399ae8f1188 100644 --- a/net/netfilter/nf_conntrack_proto_sctp.c +++ b/net/netfilter/nf_conntrack_proto_sctp.c @@ -594,6 +594,9 @@ static int sctp_timeout_nlattr_to_obj(struct nlattr *tb[], struct nf_sctp_net *sn = nf_sctp_pernet(net); int i; + if (!timeouts) + timeouts = sn->timeouts; + /* set default SCTP timeouts. */ for (i=0; i<SCTP_CONNTRACK_MAX; i++) timeouts[i] = sn->timeouts[i]; -- 2.11.0
From: Florian Westphal <fw@strlen.de> The set uadt functions assume lineno is never NULL, but it is in case of ip_set_utest(). syzkaller managed to generate a netlink message that calls this with LINENO attr present: general protection fault: 0000 [#1] PREEMPT SMP KASAN RIP: 0010:hash_mac4_uadt+0x1bc/0x470 net/netfilter/ipset/ip_set_hash_mac.c:104 Call Trace: ip_set_utest+0x55b/0x890 net/netfilter/ipset/ip_set_core.c:1867 nfnetlink_rcv_msg+0xcf2/0xfb0 net/netfilter/nfnetlink.c:229 netlink_rcv_skb+0x177/0x450 net/netlink/af_netlink.c:2477 nfnetlink_rcv+0x1ba/0x460 net/netfilter/nfnetlink.c:563 pass a dummy lineno storage, its easier than patching all set implementations. This seems to be a day-0 bug. Cc: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu> Reported-by: syzbot+34bd2369d38707f3f4a7@syzkaller.appspotmail.com Fixes: a7b4f989a6294 ("netfilter: ipset: IP set core support") Signed-off-by: Florian Westphal <fw@strlen.de> Acked-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> --- net/netfilter/ipset/ip_set_core.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index 169e0a04f814..cf895bc80871 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -1848,6 +1848,7 @@ static int ip_set_utest(struct net *net, struct sock *ctnl, struct sk_buff *skb, struct ip_set *set; struct nlattr *tb[IPSET_ATTR_ADT_MAX + 1] = {}; int ret = 0; + u32 lineno; if (unlikely(protocol_min_failed(attr) || !attr[IPSET_ATTR_SETNAME] || @@ -1864,7 +1865,7 @@ static int ip_set_utest(struct net *net, struct sock *ctnl, struct sk_buff *skb, return -IPSET_ERR_PROTOCOL; rcu_read_lock_bh(); - ret = set->variant->uadt(set, tb, IPSET_TEST, NULL, 0, 0); + ret = set->variant->uadt(set, tb, IPSET_TEST, &lineno, 0, 0); rcu_read_unlock_bh(); /* Userspace can't trigger element to be re-added */ if (ret == -EAGAIN) -- 2.11.0
From: Pablo Neira Ayuso <pablo@netfilter.org>
Date: Thu, 9 Jan 2020 00:17:04 +0100
> The following patchset contains Netfilter fixes for net:
>
> 1) Missing netns context in arp_tables, from Florian Westphal.
>
> 2) Underflow in flowtable reference counter, from wenxu.
>
> 3) Fix incorrect ethernet destination address in flowtable offload,
> from wenxu.
>
> 4) Check for status of neighbour entry, from wenxu.
>
> 5) Fix NAT port mangling, from wenxu.
>
> 6) Unbind callbacks from destroy path to cleanup hardware properly
> on flowtable removal.
>
> 7) Fix missing casting statistics timestamp, add nf_flowtable_time_stamp
> and use it.
>
> 8) NULL pointer exception when timeout argument is null in conntrack
> dccp and sctp protocol helpers, from Florian Westphal.
>
> 9) Possible nul-dereference in ipset with IPSET_ATTR_LINENO, also from
> Florian.
>
> You can pull these changes from:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git
Pulled, thanks Pablo.