netfilter: nf_tables: reject duplicate device on updates
Subsystems (2)
Pablo Neira Ayuso
maintainerPATCH 1/1 netfilter: nf_tables: reject duplicate device on updates
A chain/flowtable update with duplicated devices in the same batch is possible. Unfortunately, netdev event path only removes the first device that is found, leaving unregistered the hook of the duplicated device.
Check if a duplicated device exists in the transaction batch, bail out with EEXIST in such case.
WARNING is hit when unregistering the hook:
[49042.221275] WARNING: CPU: 4 PID: 8425 at net/netfilter/core.c:340 nf_hook_entry_head+0xaa/0x150 [49042.221375] CPU: 4 UID: 0 PID: 8425 Comm: nft Tainted: G S 6.16.0+ #170 PREEMPT(full) [...] [49042.221382] RIP: 0010:nf_hook_entry_head+0xaa/0x150
--- net/netfilter/nf_tables_api.c | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+)
net/netfilter/nf_tables_api.c +30 -0
@@ -2803,6 +2803,7 @@ static int nf_tables_updchain(struct nft_ctx *ctx, u8 genmask, u8 policy,struct nft_chain *chain = ctx->chain;struct nft_chain_hook hook = {};struct nft_stats __percpu *stats =NULL ; +struct nftables_pernet *nft_net;struct nft_hook *h, *next;struct nf_hook_ops *ops;struct nft_trans *trans;
@@ -2845,6 +2846,20 @@ static int nf_tables_updchain(struct nft_ctx *ctx, u8 genmask, u8 policy,if (nft_hook_list_find (&basechain->hook_list, h)) {list_del (&h->list);nft_netdev_hook_free (h); +continue ; + } + + nft_net =nft_pernet (ctx->net); +list_for_each_entry (trans, &nft_net->commit_list, list) { +if (trans->msg_type !=NFT_MSG_NEWCHAIN || + trans->table != ctx->table || + !nft_trans_chain_update (trans)) +continue ; + +if (nft_hook_list_find (&nft_trans_chain_hooks (trans), h)) { +nft_chain_release_hook (&hook); +return -EEXIST ; + } } } }else {
@@ -9060,6 +9075,7 @@ static int nft_flowtable_update(struct nft_ctx *ctx, const struct nlmsghdr *nlh, {const struct nlattr *const *nla = ctx->nla;struct nft_flowtable_hook flowtable_hook; +struct nftables_pernet *nft_net;struct nft_hook *hook, *next;struct nf_hook_ops *ops;struct nft_trans *trans;
@@ -9076,6 +9092,20 @@ static int nft_flowtable_update(struct nft_ctx *ctx, const struct nlmsghdr *nlh,if (nft_hook_list_find (&flowtable->hook_list, hook)) {list_del (&hook->list);nft_netdev_hook_free (hook); +continue ; + } + + nft_net =nft_pernet (ctx->net); +list_for_each_entry (trans, &nft_net->commit_list, list) { +if (trans->msg_type !=NFT_MSG_NEWFLOWTABLE || + trans->table != ctx->table || + !nft_trans_flowtable_update (trans)) +continue ; + +if (nft_hook_list_find (&nft_trans_flowtable_hooks (trans), hook)) { + err = -EEXIST ; + goto err_flowtable_update_hook; + } } }
Signature · 1 line
2.30.2
Florian Westphal
maintainerQuoted 8 lines · Pablo Neira Ayuso
A chain/flowtable update with duplicated devices in the same batch is possible. Unfortunately, netdev event path only removes the first device that is found, leaving unregistered the hook of the duplicated device.
Check if a duplicated device exists in the transaction batch, bail out with EEXIST in such case.
Quoted 6 lines
WARNING is hit when unregistering the hook:
[49042.221275] WARNING: CPU: 4 PID: 8425 at net/netfilter/core.c:340 nf_hook_entry_head+0xaa/0x150 [49042.221375] CPU: 4 UID: 0 PID: 8425 Comm: nft Tainted: G S 6.16.0+ #170 PREEMPT(full) [...] [49042.221382] RIP: 0010:nf_hook_entry_head+0xaa/0x150
Thanks Pablo.
Just to confirm: this doesn't result in anything other than the unreg splat, correct?
Or does this leak memory too?
FTR, i placed this in nf.git:testing.
Pablo Neira Ayuso
maintainerQuoted 16 lines · Florian Westphal
Quoted 8 lines · Pablo Neira Ayuso
A chain/flowtable update with duplicated devices in the same batch is possible. Unfortunately, netdev event path only removes the first device that is found, leaving unregistered the hook of the duplicated device.
Check if a duplicated device exists in the transaction batch, bail out with EEXIST in such case.
Quoted 6 lines
WARNING is hit when unregistering the hook:
[49042.221275] WARNING: CPU: 4 PID: 8425 at net/netfilter/core.c:340 nf_hook_entry_head+0xaa/0x150 [49042.221375] CPU: 4 UID: 0 PID: 8425 Comm: nft Tainted: G S 6.16.0+ #170 PREEMPT(full) [...] [49042.221382] RIP: 0010:nf_hook_entry_head+0xaa/0x150
Quoted 6 lines
Thanks Pablo.
Just to confirm: this doesn't result in anything other than the unreg splat, correct?
Or does this leak memory too?
It seems I tested on a kernel without CONFIG_KASAN, with it, it reports UaF.
[ 97.140749] ================================================================== [ 97.140762] BUG: KASAN: slab-use-after-free in nf_hook_entry_head+0xd9/0x140 [ 97.140774] Read of size 8 at addr ffff88814feba108 by task nft/1097
Quoted 1 line
FTR, i placed this in nf.git:testing.
Thanks.