Netfilter fixes for net
Subsystems (3)
Pablo Neira Ayuso
maintainercover Netfilter fixes for net
Hi,
The following patchset contains Netfilter fixes for net:
1) Nicolas Dichtel updates MAINTAINERS file to add Netfilter IRC channel.
2) Skip non-IPv6 packets in nft_exthdr.
3) Skip non-TCP packets in nft_osf.
4) Skip non-TCP/UDP packets in nft_tproxy.
5) Memleak in hardware offload infrastructure when counters are used for first time in a rule.
6) The VLAN transfer routine must use FLOW_DISSECTOR_KEY_BASIC instead of FLOW_DISSECTOR_KEY_CONTROL. Moreover, make a more robust check for 802.1q and 802.1ad to restore simple matching on transport protocols.
7) Fix bogus EPERM when listing a ruleset when table ownership flag is set on.
8) Honor table ownership flag when table is referenced by handle.
Please, pull these changes from:
git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git
Thank you!
----------------------------------------------------------------
The following changes since commit a4f0377db1254373513b992ff31a351a7111f0fd:
Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf (2021-06-15 15:26:07 -0700)
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 e31f072ffab0397a328b31a9589dcf9733dc9c72:
netfilter: nf_tables: do not allow to delete table with owner by handle (2021-06-22 12:15:05 +0200)
----------------------------------------------------------------
Nicolas Dichtel (1):
MAINTAINERS: netfilter: add irc channel
Pablo Neira Ayuso (7):
netfilter: nft_exthdr: check for IPv6 packet before further processing
netfilter: nft_osf: check for TCP packet before further processing
netfilter: nft_tproxy: restrict support to TCP and UDP transport protocols
netfilter: nf_tables: memleak in hw offload abort path
netfilter: nf_tables_offload: check FLOW_DISSECTOR_KEY_BASIC in VLAN transfer logic
netfilter: nf_tables: skip netlink portID validation if zero
netfilter: nf_tables: do not allow to delete table with owner by handle
MAINTAINERS | 1 + net/netfilter/nf_tables_api.c | 65 ++++++++++++++++++++++++--------------- net/netfilter/nf_tables_offload.c | 34 +++++--------------- net/netfilter/nft_exthdr.c | 3 ++ net/netfilter/nft_osf.c | 5 +++ net/netfilter/nft_tproxy.c | 9 +++++- 6 files changed, 65 insertions(+), 52 deletions(-)
Pablo Neira Ayuso
maintainerPATCH 1/8 MAINTAINERS: netfilter: add irc channel
From: Nicolas Dichtel <[email protected]>
The community #netfilter IRC channel is now live on the libera.chat network
(https://libera.chat/).
--- MAINTAINERS | 1 + 1 file changed, 1 insertion(+)
MAINTAINERS +1 -0
@@ -12657,6 +12657,7 @@ W: http://www.netfilter.org/ W: http://www.iptables.org/ W: http://www.nftables.org/ Q: http://patchwork.ozlabs.org/project/netfilter-devel/list/ +C: irc://irc.libera.chat/netfilter T: git git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git T: git git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next.git F: include/linux/netfilter*
Signature · 1 line
2.30.2
Re: [PATCH net 1/8] MAINTAINERS: netfilter: add irc channel
This series was applied to netdev/net.git (refs/heads/master):
Hello:
This series was applied to netdev/net.git (refs/heads/master):
Quoted 11 lines · On Tue, 22 Jun 2021 23:59:54 +0200 you wrote:
From: Nicolas Dichtel <[email protected]>
The community #netfilter IRC channel is now live on the libera.chat network
(https://libera.chat/).CC Arturo Borrero Gonzalez <[email protected]>Signed-off-by Nicolas Dichtel <[email protected]>Signed-off-by Pablo Neira Ayuso <[email protected]>[...]
Here is the summary with links:
- [net,1/8] MAINTAINERS: netfilter: add irc channel
https://git.kernel.org/netdev/net/c/8744365e2584 - [net,2/8] netfilter: nft_exthdr: check for IPv6 packet before further processing https://git.kernel.org/netdev/net/c/cdd73cc545c0 - [net,3/8] netfilter: nft_osf: check for TCP packet before further processing https://git.kernel.org/netdev/net/c/8f518d43f89a - [net,4/8] netfilter: nft_tproxy: restrict support to TCP and UDP transport protocols https://git.kernel.org/netdev/net/c/52f0f4e178c7 - [net,5/8] netfilter: nf_tables: memleak in hw offload abort path https://git.kernel.org/netdev/net/c/3c5e44622011 - [net,6/8] netfilter: nf_tables_offload: check FLOW_DISSECTOR_KEY_BASIC in VLAN transfer logic https://git.kernel.org/netdev/net/c/ea45fdf82cc9 - [net,7/8] netfilter: nf_tables: skip netlink portID validation if zero https://git.kernel.org/netdev/net/c/534799097a77 - [net,8/8] netfilter: nf_tables: do not allow to delete table with owner by handle https://git.kernel.org/netdev/net/c/e31f072ffab0
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
Pablo Neira Ayuso
maintainerPATCH 2/8 netfilter: nft_exthdr: check for IPv6 packet before further processing
In reply to Pablo Neira Ayusoipv6_find_hdr() does not validate that this is an IPv6 packet. Add a sanity check for calling ipv6_find_hdr() to make sure an IPv6 packet is passed for parsing.
--- net/netfilter/nft_exthdr.c | 3 +++ 1 file changed, 3 insertions(+)
net/netfilter/nft_exthdr.c +3 -0
@@ -42,6 +42,9 @@ static void nft_exthdr_ipv6_eval(const struct nft_expr *expr,unsigned int offset =0 ;int err; +if (pkt->skb->protocol !=htons (ETH_P_IPV6 )) + goto err; + err =ipv6_find_hdr (pkt->skb, &offset, priv->type,NULL ,NULL );if (priv->flags &NFT_EXTHDR_F_PRESENT ) {nft_reg_store8 (dest, err >=0 );
Signature · 1 line
2.30.2
Pablo Neira Ayuso
maintainerPATCH 3/8 netfilter: nft_osf: check for TCP packet before further processing
In reply to Pablo Neira AyusoThe osf expression only supports for TCP packets, add a upfront sanity check to skip packet parsing if this is not a TCP packet.
--- net/netfilter/nft_osf.c | 5 +++++ 1 file changed, 5 insertions(+)
net/netfilter/nft_osf.c +5 -0
@@ -28,6 +28,11 @@ static void nft_osf_eval(const struct nft_expr *expr, struct nft_regs *regs,struct nf_osf_data data;struct tcphdr _tcph; +if (pkt->tprot !=IPPROTO_TCP ) { + regs->verdict.code =NFT_BREAK ; +return ; + } + tcp =skb_header_pointer (skb,ip_hdrlen (skb),sizeof (struct tcphdr ), &_tcph);if (!tcp) {
Signature · 1 line
2.30.2
Pablo Neira Ayuso
maintainerPATCH 4/8 netfilter: nft_tproxy: restrict support to TCP and UDP transport protocols
In reply to Pablo Neira AyusoAdd unfront check for TCP and UDP packets before performing further processing.
--- net/netfilter/nft_tproxy.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
net/netfilter/nft_tproxy.c +8 -1
@@ -30,6 +30,12 @@ static void nft_tproxy_eval_v4(const struct nft_expr *expr,__be16 tport =0 ;struct sock *sk; +if (pkt->tprot !=IPPROTO_TCP && + pkt->tprot !=IPPROTO_UDP ) { + regs->verdict.code =NFT_BREAK ; +return ; + } + hp =skb_header_pointer (skb,ip_hdrlen (skb),sizeof (_hdr), &_hdr);if (!hp) { regs->verdict.code =NFT_BREAK ;
@@ -91,7 +97,8 @@ static void nft_tproxy_eval_v6(const struct nft_expr *expr,memset (&taddr,0 ,sizeof (taddr)); -if (!pkt->tprot_set) { +if (pkt->tprot !=IPPROTO_TCP && + pkt->tprot !=IPPROTO_UDP ) { regs->verdict.code =NFT_BREAK ;return ; }
Signature · 1 line
2.30.2
Pablo Neira Ayuso
maintainerPATCH 5/8 netfilter: nf_tables: memleak in hw offload abort path
In reply to Pablo Neira AyusoRelease flow from the abort path, this is easy to reproduce since b72920f6e4a9 ("netfilter: nftables: counter hardware offload support"). If the preparation phase fails, then the abort path is exercised without releasing the flow rule object.
unreferenced object 0xffff8881f0fa7700 (size 128):
comm "nft", pid 1335, jiffies 4294931120 (age 4163.740s) hex dump (first 32 bytes):
08 e4 de 13 82 88 ff ff 98 e4 de 13 82 88 ff ff ................
48 e4 de 13 82 88 ff ff 01 00 00 00 00 00 00 00 H...............
backtrace:
[<00000000634547e7>] flow_rule_alloc+0x26/0x80
[<00000000c8426156>] nft_flow_rule_create+0xc9/0x3f0 [nf_tables]
[<0000000075ff8e46>] nf_tables_newrule+0xc79/0x10a0 [nf_tables]
[<00000000ba65e40e>] nfnetlink_rcv_batch+0xaac/0xf90 [nfnetlink]
[<00000000505c614a>] nfnetlink_rcv+0x1bb/0x1f0 [nfnetlink]
[<00000000eb78e1fe>] netlink_unicast+0x34b/0x480
[<00000000a8f72c94>] netlink_sendmsg+0x3af/0x690
[<000000009cb1ddf4>] sock_sendmsg+0x96/0xa0
[<0000000039d06e44>] ____sys_sendmsg+0x3fe/0x440
[<00000000137e82ca>] ___sys_sendmsg+0xd8/0x140
[<000000000c6bf6a6>] __sys_sendmsg+0xb3/0x130
[<0000000043bd6268>] do_syscall_64+0x40/0xb0
[<00000000afdebc2d>] entry_SYSCALL_64_after_hwframe+0x44/0xaeRemove flow rule release from the offload commit path, otherwise error from the offload commit phase might trigger a double-free due to the execution of the abort_offload -> abort. After this patch, the abort path takes care of releasing the flow rule.
This fix also needs to move the nft_flow_rule_create() call before the transaction object is added otherwise the abort path might find a NULL pointer to the flow rule object for the NFT_CHAIN_HW_OFFLOAD case.
While at it, rename BASIC-like goto tags to slightly more meaningful names rather than adding a new "err3" tag.
--- net/netfilter/nf_tables_api.c | 51 +++++++++++++++++++------------ net/netfilter/nf_tables_offload.c | 17 ----------- 2 files changed, 31 insertions(+), 37 deletions(-)
net/netfilter/nf_tables_api.c +31 -20
@@ -3243,9 +3243,9 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info,u8 genmask =nft_genmask_next (info->net);struct nft_rule *rule, *old_rule =NULL ;struct nft_expr_info *expr_info =NULL ; +struct nft_flow_rule *flow =NULL ;int family = nfmsg->nfgen_family;struct net *net = info->net; -struct nft_flow_rule *flow;struct nft_userdata *udata;struct nft_table *table;struct nft_chain *chain;
@@ -3340,13 +3340,13 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info,nla_for_each_nested (tmp, nla[NFTA_RULE_EXPRESSIONS ], rem) { err = -EINVAL ;if (nla_type (tmp) !=NFTA_LIST_ELEM ) - goto err1; + goto err_release_expr;if (n ==NFT_RULE_MAXEXPRS ) - goto err1; + goto err_release_expr; err =nf_tables_expr_parse (&ctx, tmp, &expr_info[n]);if (err <0 ) {NL_SET_BAD_ATTR (extack, tmp); - goto err1; + goto err_release_expr; } size += expr_info[n].ops->size; n++;
@@ -3355,7 +3355,7 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info,/* Check for overflow of dlen field */ err = -EFBIG ;if (size >=1 <<12 ) - goto err1; + goto err_release_expr;if (nla[NFTA_RULE_USERDATA ]) { ulen =nla_len (nla[NFTA_RULE_USERDATA ]);
@@ -3366,7 +3366,7 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info, err = -ENOMEM ; rule =kzalloc (sizeof (*rule) + size + usize,GFP_KERNEL );if (rule ==NULL ) - goto err1; + goto err_release_expr;nft_activate_next (net, rule);
@@ -3385,7 +3385,7 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info, err =nf_tables_newexpr (&ctx, &expr_info[i], expr);if (err <0 ) {NL_SET_BAD_ATTR (extack, expr_info[i].attr); - goto err2; + goto err_release_rule; }if (expr_info[i].ops->validate)
@@ -3395,16 +3395,24 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info, expr =nft_expr_next (expr); } +if (chain->flags &NFT_CHAIN_HW_OFFLOAD ) { + flow =nft_flow_rule_create (net, rule); +if (IS_ERR (flow)) { + err =PTR_ERR (flow); + goto err_release_rule; + } + } +if (info->nlh->nlmsg_flags &NLM_F_REPLACE ) { trans =nft_trans_rule_add (&ctx,NFT_MSG_NEWRULE , rule);if (trans ==NULL ) { err = -ENOMEM ; - goto err2; + goto err_destroy_flow_rule; } err =nft_delrule (&ctx, old_rule);if (err <0 ) {nft_trans_destroy (trans); - goto err2; + goto err_destroy_flow_rule; }list_add_tail_rcu (&rule->list, &old_rule->list);
@@ -3412,7 +3420,7 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info, trans =nft_trans_rule_add (&ctx,NFT_MSG_NEWRULE , rule);if (!trans) { err = -ENOMEM ; - goto err2; + goto err_destroy_flow_rule; }if (info->nlh->nlmsg_flags &NLM_F_APPEND ) {
@@ -3430,21 +3438,19 @@ static int nf_tables_newrule(struct sk_buff *skb, const struct nfnl_info *info,kvfree (expr_info); chain->use++; +if (flow) +nft_trans_flow_rule (trans) = flow; +if (nft_net->validate_state ==NFT_VALIDATE_DO )return nft_table_validate (net, table); -if (chain->flags &NFT_CHAIN_HW_OFFLOAD ) { - flow =nft_flow_rule_create (net, rule); -if (IS_ERR (flow)) -return PTR_ERR (flow); - -nft_trans_flow_rule (trans) = flow; - } -return 0 ; -err2: + +err_destroy_flow_rule: +nft_flow_rule_destroy (flow); +err_release_rule:nf_tables_rule_release (&ctx, rule); -err1: +err_release_expr:for (i =0 ; i < n; i++) {if (expr_info[i].ops) {module_put (expr_info[i].ops->type->owner);
@@ -8839,11 +8845,16 @@ static int __nf_tables_abort(struct net *net, enum nfnl_abort_action action)nft_rule_expr_deactivate (&trans->ctx,nft_trans_rule (trans),NFT_TRANS_ABORT ); +if (trans->ctx.chain->flags &NFT_CHAIN_HW_OFFLOAD ) +nft_flow_rule_destroy (nft_trans_flow_rule (trans));break ;case NFT_MSG_DELRULE : trans->ctx.chain->use++;nft_clear (trans->ctx.net,nft_trans_rule (trans));nft_rule_expr_activate (&trans->ctx,nft_trans_rule (trans)); +if (trans->ctx.chain->flags &NFT_CHAIN_HW_OFFLOAD ) +nft_flow_rule_destroy (nft_trans_flow_rule (trans)); +nft_trans_destroy (trans);break ;case NFT_MSG_NEWSET :
net/netfilter/nf_tables_offload.c +0 -17
@@ -594,23 +594,6 @@ int nft_flow_rule_offload_commit(struct net *net) } } -list_for_each_entry (trans, &nft_net->commit_list, list) { -if (trans->ctx.family !=NFPROTO_NETDEV ) -continue ; - -switch (trans->msg_type) { -case NFT_MSG_NEWRULE : -case NFT_MSG_DELRULE : -if (!(trans->ctx.chain->flags &NFT_CHAIN_HW_OFFLOAD )) -continue ; - -nft_flow_rule_destroy (nft_trans_flow_rule (trans)); -break ; -default : -break ; - } - } -return err; }
Signature · 1 line
2.30.2
Pablo Neira Ayuso
maintainerPATCH 6/8 netfilter: nf_tables_offload: check FLOW_DISSECTOR_KEY_BASIC in VLAN transfer logic
In reply to Pablo Neira AyusoThe VLAN transfer logic should actually check for FLOW_DISSECTOR_KEY_BASIC, not FLOW_DISSECTOR_KEY_CONTROL. Moreover, do not fallback to case 2) .n_proto is set to 802.1q or 802.1ad, if FLOW_DISSECTOR_KEY_BASIC is unset.
--- net/netfilter/nf_tables_offload.c | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-)
net/netfilter/nf_tables_offload.c +7 -10
@@ -54,15 +54,10 @@ static void nft_flow_rule_transfer_vlan(struct nft_offload_ctx *ctx, struct nft_flow_rule *flow) {struct nft_flow_match *match = &flow->match; -struct nft_offload_ethertype ethertype; - -if (match->dissector.used_keys &BIT (FLOW_DISSECTOR_KEY_CONTROL ) && - match->key.basic.n_proto !=htons (ETH_P_8021Q ) && - match->key.basic.n_proto !=htons (ETH_P_8021AD )) -return ; - - ethertype.value = match->key.basic.n_proto; - ethertype.mask = match->mask.basic.n_proto; +struct nft_offload_ethertype ethertype = { + .value = match->key.basic.n_proto, + .mask = match->mask.basic.n_proto, + }; if (match->dissector.used_keys & BIT(FLOW_DISSECTOR_KEY_VLAN) && (match->key.vlan.vlan_tpid == htons(ETH_P_8021Q) ||
@@ -76,7 +71,9 @@ static void nft_flow_rule_transfer_vlan(struct nft_offload_ctx *ctx, match->dissector.offset[FLOW_DISSECTOR_KEY_CVLAN ] = offsetof(struct nft_flow_key , cvlan); match->dissector.used_keys |=BIT (FLOW_DISSECTOR_KEY_CVLAN ); - }else { + }else if (match->dissector.used_keys &BIT (FLOW_DISSECTOR_KEY_BASIC ) && + (match->key.basic.n_proto ==htons (ETH_P_8021Q ) || + match->key.basic.n_proto ==htons (ETH_P_8021AD ))) { match->key.basic.n_proto = match->key.vlan.vlan_tpid; match->mask.basic.n_proto = match->mask.vlan.vlan_tpid; match->key.vlan.vlan_tpid = ethertype.value;
Signature · 1 line
2.30.2
Pablo Neira Ayuso
maintainerPATCH 7/8 netfilter: nf_tables: skip netlink portID validation if zero
In reply to Pablo Neira Ayusonft_table_lookup() allows us to obtain the table object by the name and the family. The netlink portID validation needs to be skipped for the dump path, since the ownership only applies to commands to update the given table. Skip validation if the specified netlink PortID is zero when calling nft_table_lookup().
--- net/netfilter/nf_tables_api.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
net/netfilter/nf_tables_api.c +1 -1
@@ -571,7 +571,7 @@ static struct nft_table *nft_table_lookup(const struct net *net, table->family == family && nft_active_genmask(table, genmask)) {if (nft_table_has_owner (table) && - table->nlpid != nlpid) + nlpid && table->nlpid != nlpid)return ERR_PTR (-EPERM );return table;
Signature · 1 line
2.30.2
Pablo Neira Ayuso
maintainerPATCH 8/8 netfilter: nf_tables: do not allow to delete table with owner by handle
In reply to Pablo Neira Ayusonft_table_lookup_byhandle() also needs to validate the netlink PortID owner when deleting a table by handle.
--- net/netfilter/nf_tables_api.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-)
net/netfilter/nf_tables_api.c +9 -3
@@ -583,7 +583,7 @@ static struct nft_table *nft_table_lookup(const struct net *net,static struct nft_table *nft_table_lookup_byhandle (const struct net *net,const struct nlattr *nla, -u8 genmask) +u8 genmask,u32 nlpid) {struct nftables_pernet *nft_net;struct nft_table *table;
@@ -591,8 +591,13 @@ static struct nft_table *nft_table_lookup_byhandle(const struct net *net, nft_net =nft_pernet (net);list_for_each_entry (table, &nft_net->tables, list) {if (be64_to_cpu (nla_get_be64 (nla)) == table->handle && -nft_active_genmask (table, genmask)) +nft_active_genmask (table, genmask)) { +if (nft_table_has_owner (table) && + nlpid && table->nlpid != nlpid) +return ERR_PTR (-EPERM ); +return table; + } }return ERR_PTR (-ENOENT );
@@ -1279,7 +1284,8 @@ static int nf_tables_deltable(struct sk_buff *skb, const struct nfnl_info *info,if (nla[NFTA_TABLE_HANDLE ]) { attr = nla[NFTA_TABLE_HANDLE ]; - table =nft_table_lookup_byhandle (net, attr, genmask); + table =nft_table_lookup_byhandle (net, attr, genmask, +NETLINK_CB (skb).portid); }else { attr = nla[NFTA_TABLE_NAME ]; table = nft_table_lookup(net, attr, family, genmask,
Signature · 1 line
2.30.2
David Miller
maintainer[davem@localhost net]$ git pull --no-ff git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git
From git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf
* branch HEAD -> FETCH_HEAD
Already up to date.
[davem@localhost net]$
???
Pablo Neira Ayuso
maintainerQuoted 8 lines · David Miller
[davem@localhost net]$ git pull --no-ff
git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git
Fromgit://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf
* branch HEAD ->FETCH_HEAD
Already up to date.
[davem@localhost net]$???
A robot got stuck here and it did not push out to origin for some reason, sorry. Please retry, it should be there now.