net: sched: Introduce act_ctinfo action
ctinfo is a new tc filter action module. It is designed to restore
information contained in firewall conntrack marks to other packet fields
and is typically used on packet ingress paths. At present it has two
independent sub-functions or operating modes, DSCP restoration mode &
skb mark restoration mode.
The DSCP restore mode:
This mode copies DSCP values that have been placed in the firewall
conntrack mark back into the IPv4/v6 diffserv fields of relevant
packets.
The DSCP restoration is intended for use and has been found useful for
restoring ingress classifications based on egress classifications across
links that bleach or otherwise change DSCP, typically home ISP Internet
links. Restoring DSCP on ingress on the WAN link allows qdiscs such as
but by no means limited to CAKE to shape inbound packets according to
policies that are easier to set & mark on egress.
Ingress classification is traditionally a challenging task since
iptables rules haven't yet run and tc filter/eBPF programs are pre-NAT
lookups, hence are unable to see internal IPv4 addresses as used on the
typical home masquerading gateway. Thus marking the connection in some
manner on egress for later restoration of classification on ingress is
easier to implement.
Parameters related to DSCP restore mode:
dscpmask - a 32 bit mask of 6 contiguous bits and indicate bits of the
conntrack mark field contain the DSCP value to be restored.
statemask - a 32 bit mask of (usually) 1 bit length, outside the area
specified by dscpmask. This represents a conditional operation flag
whereby the DSCP is only restored if the flag is set. This is useful to
implement a 'one shot' iptables based classification where the
'complicated' iptables rules are only run once to classify the
connection on initial (egress) packet and subsequent packets are all
marked/restored with the same DSCP. A mask of zero disables the
conditional behaviour ie. the conntrack mark DSCP bits are always
restored to the ip diffserv field (assuming the conntrack entry is found
& the skb is an ipv4/ipv6 type)
e.g. dscpmask 0xfc000000 statemask 0x01000000
|----0xFC----conntrack mark----000000---|
| Bits 31-26 | bit 25 | bit24 |~~~ Bit 0|
| DSCP | unused | flag |unused |
|-----------------------0x01---000000---|
| |
| |
---| Conditional flag
v only restore if set
|-ip diffserv-|
| 6 bits |
|-------------|
The skb mark restore mode (cpmark):
This mode copies the firewall conntrack mark to the skb's mark field.
It is completely the functional equivalent of the existing act_connmark
action with the additional feature of being able to apply a mask to the
restored value.
Parameters related to skb mark restore mode:
mask - a 32 bit mask applied to the firewall conntrack mark to mask out
bits unwanted for restoration. This can be useful where the conntrack
mark is being used for different purposes by different applications. If
not specified and by default the whole mark field is copied (i.e.
default mask of 0xffffffff)
e.g. mask 0x00ffffff to mask out the top 8 bits being used by the
aforementioned DSCP restore mode.
|----0x00----conntrack mark----ffffff---|
| Bits 31-24 | |
| DSCP & flag| some value here |
|---------------------------------------|
|
|
v
|------------skb mark-------------------|
| | |
| zeroed | |
|---------------------------------------|
Overall parameters:
zone - conntrack zone
control - action related control (reclassify | pipe | drop | continue |
ok | goto chain <CHAIN_INDEX>)
Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
Reviewed-by: Toke Høiland-Jørgensen <toke@redhat.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/sched/act_ctinfo.c b/net/sched/act_ctinfo.c
new file mode 100644
index 0000000..9261091
--- /dev/null
+++ b/net/sched/act_ctinfo.c
@@ -0,0 +1,396 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* net/sched/act_ctinfo.c netfilter ctinfo connmark actions
+ *
+ * Copyright (c) 2019 Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
+ */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/skbuff.h>
+#include <linux/rtnetlink.h>
+#include <linux/pkt_cls.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include <net/netlink.h>
+#include <net/pkt_sched.h>
+#include <net/act_api.h>
+#include <net/pkt_cls.h>
+#include <uapi/linux/tc_act/tc_ctinfo.h>
+#include <net/tc_act/tc_ctinfo.h>
+
+#include <net/netfilter/nf_conntrack.h>
+#include <net/netfilter/nf_conntrack_core.h>
+#include <net/netfilter/nf_conntrack_ecache.h>
+#include <net/netfilter/nf_conntrack_zones.h>
+
+static struct tc_action_ops act_ctinfo_ops;
+static unsigned int ctinfo_net_id;
+
+static void tcf_ctinfo_dscp_set(struct nf_conn *ct, struct tcf_ctinfo *ca,
+ struct tcf_ctinfo_params *cp,
+ struct sk_buff *skb, int wlen, int proto)
+{
+ u8 dscp, newdscp;
+
+ newdscp = (((ct->mark & cp->dscpmask) >> cp->dscpmaskshift) << 2) &
+ ~INET_ECN_MASK;
+
+ switch (proto) {
+ case NFPROTO_IPV4:
+ dscp = ipv4_get_dsfield(ip_hdr(skb)) & ~INET_ECN_MASK;
+ if (dscp != newdscp) {
+ if (likely(!skb_try_make_writable(skb, wlen))) {
+ ipv4_change_dsfield(ip_hdr(skb),
+ INET_ECN_MASK,
+ newdscp);
+ ca->stats_dscp_set++;
+ } else {
+ ca->stats_dscp_error++;
+ }
+ }
+ break;
+ case NFPROTO_IPV6:
+ dscp = ipv6_get_dsfield(ipv6_hdr(skb)) & ~INET_ECN_MASK;
+ if (dscp != newdscp) {
+ if (likely(!skb_try_make_writable(skb, wlen))) {
+ ipv6_change_dsfield(ipv6_hdr(skb),
+ INET_ECN_MASK,
+ newdscp);
+ ca->stats_dscp_set++;
+ } else {
+ ca->stats_dscp_error++;
+ }
+ }
+ break;
+ default:
+ break;
+ }
+}
+
+static void tcf_ctinfo_cpmark_set(struct nf_conn *ct, struct tcf_ctinfo *ca,
+ struct tcf_ctinfo_params *cp,
+ struct sk_buff *skb)
+{
+ ca->stats_cpmark_set++;
+ skb->mark = ct->mark & cp->cpmarkmask;
+}
+
+static int tcf_ctinfo_act(struct sk_buff *skb, const struct tc_action *a,
+ struct tcf_result *res)
+{
+ const struct nf_conntrack_tuple_hash *thash = NULL;
+ struct tcf_ctinfo *ca = to_ctinfo(a);
+ struct nf_conntrack_tuple tuple;
+ struct nf_conntrack_zone zone;
+ enum ip_conntrack_info ctinfo;
+ struct tcf_ctinfo_params *cp;
+ struct nf_conn *ct;
+ int proto, wlen;
+ int action;
+
+ cp = rcu_dereference_bh(ca->params);
+
+ tcf_lastuse_update(&ca->tcf_tm);
+ bstats_update(&ca->tcf_bstats, skb);
+ action = READ_ONCE(ca->tcf_action);
+
+ wlen = skb_network_offset(skb);
+ if (tc_skb_protocol(skb) == htons(ETH_P_IP)) {
+ wlen += sizeof(struct iphdr);
+ if (!pskb_may_pull(skb, wlen))
+ goto out;
+
+ proto = NFPROTO_IPV4;
+ } else if (tc_skb_protocol(skb) == htons(ETH_P_IPV6)) {
+ wlen += sizeof(struct ipv6hdr);
+ if (!pskb_may_pull(skb, wlen))
+ goto out;
+
+ proto = NFPROTO_IPV6;
+ } else {
+ goto out;
+ }
+
+ ct = nf_ct_get(skb, &ctinfo);
+ if (!ct) { /* look harder, usually ingress */
+ if (!nf_ct_get_tuplepr(skb, skb_network_offset(skb),
+ proto, cp->net, &tuple))
+ goto out;
+ zone.id = cp->zone;
+ zone.dir = NF_CT_DEFAULT_ZONE_DIR;
+
+ thash = nf_conntrack_find_get(cp->net, &zone, &tuple);
+ if (!thash)
+ goto out;
+
+ ct = nf_ct_tuplehash_to_ctrack(thash);
+ }
+
+ if (cp->mode & CTINFO_MODE_DSCP)
+ if (!cp->dscpstatemask || (ct->mark & cp->dscpstatemask))
+ tcf_ctinfo_dscp_set(ct, ca, cp, skb, wlen, proto);
+
+ if (cp->mode & CTINFO_MODE_CPMARK)
+ tcf_ctinfo_cpmark_set(ct, ca, cp, skb);
+
+ if (thash)
+ nf_ct_put(ct);
+out:
+ return action;
+}
+
+static const struct nla_policy ctinfo_policy[TCA_CTINFO_MAX + 1] = {
+ [TCA_CTINFO_ACT] = { .len = sizeof(struct
+ tc_ctinfo) },
+ [TCA_CTINFO_ZONE] = { .type = NLA_U16 },
+ [TCA_CTINFO_PARMS_DSCP_MASK] = { .type = NLA_U32 },
+ [TCA_CTINFO_PARMS_DSCP_STATEMASK] = { .type = NLA_U32 },
+ [TCA_CTINFO_PARMS_CPMARK_MASK] = { .type = NLA_U32 },
+};
+
+static int tcf_ctinfo_init(struct net *net, struct nlattr *nla,
+ struct nlattr *est, struct tc_action **a,
+ int ovr, int bind, bool rtnl_held,
+ struct tcf_proto *tp,
+ struct netlink_ext_ack *extack)
+{
+ struct tc_action_net *tn = net_generic(net, ctinfo_net_id);
+ struct nlattr *tb[TCA_CTINFO_MAX + 1];
+ struct tcf_ctinfo_params *cp_new;
+ struct tcf_chain *goto_ch = NULL;
+ u32 dscpmask = 0, dscpstatemask;
+ struct tc_ctinfo *actparm;
+ struct tcf_ctinfo *ci;
+ u8 dscpmaskshift;
+ int ret = 0, err;
+
+ if (!nla)
+ return -EINVAL;
+
+ err = nla_parse_nested(tb, TCA_CTINFO_MAX, nla, ctinfo_policy, NULL);
+ if (err < 0)
+ return err;
+
+ if (!tb[TCA_CTINFO_ACT])
+ return -EINVAL;
+ actparm = nla_data(tb[TCA_CTINFO_ACT]);
+
+ /* do some basic validation here before dynamically allocating things */
+ /* that we would otherwise have to clean up. */
+ if (tb[TCA_CTINFO_PARMS_DSCP_MASK]) {
+ dscpmask = nla_get_u32(tb[TCA_CTINFO_PARMS_DSCP_MASK]);
+ /* need contiguous 6 bit mask */
+ dscpmaskshift = dscpmask ? __ffs(dscpmask) : 0;
+ if ((~0 & (dscpmask >> dscpmaskshift)) != 0x3f)
+ return -EINVAL;
+ dscpstatemask = tb[TCA_CTINFO_PARMS_DSCP_STATEMASK] ?
+ nla_get_u32(tb[TCA_CTINFO_PARMS_DSCP_STATEMASK]) : 0;
+ /* mask & statemask must not overlap */
+ if (dscpmask & dscpstatemask)
+ return -EINVAL;
+ }
+
+ /* done the validation:now to the actual action allocation */
+ err = tcf_idr_check_alloc(tn, &actparm->index, a, bind);
+ if (!err) {
+ ret = tcf_idr_create(tn, actparm->index, est, a,
+ &act_ctinfo_ops, bind, false);
+ if (ret) {
+ tcf_idr_cleanup(tn, actparm->index);
+ return ret;
+ }
+ } else if (err > 0) {
+ if (bind) /* don't override defaults */
+ return 0;
+ if (!ovr) {
+ tcf_idr_release(*a, bind);
+ return -EEXIST;
+ }
+ } else {
+ return err;
+ }
+
+ err = tcf_action_check_ctrlact(actparm->action, tp, &goto_ch, extack);
+ if (err < 0)
+ goto release_idr;
+
+ ci = to_ctinfo(*a);
+
+ cp_new = kzalloc(sizeof(*cp_new), GFP_KERNEL);
+ if (unlikely(!cp_new)) {
+ err = -ENOMEM;
+ goto put_chain;
+ }
+
+ cp_new->net = net;
+ cp_new->zone = tb[TCA_CTINFO_ZONE] ?
+ nla_get_u16(tb[TCA_CTINFO_ZONE]) : 0;
+ if (dscpmask) {
+ cp_new->dscpmask = dscpmask;
+ cp_new->dscpmaskshift = dscpmaskshift;
+ cp_new->dscpstatemask = dscpstatemask;
+ cp_new->mode |= CTINFO_MODE_DSCP;
+ } else {
+ cp_new->mode &= ~CTINFO_MODE_DSCP;
+ }
+
+ if (tb[TCA_CTINFO_PARMS_CPMARK_MASK]) {
+ cp_new->cpmarkmask =
+ nla_get_u32(tb[TCA_CTINFO_PARMS_CPMARK_MASK]);
+ cp_new->mode |= CTINFO_MODE_CPMARK;
+ } else {
+ cp_new->mode &= ~CTINFO_MODE_CPMARK;
+ }
+
+ spin_lock_bh(&ci->tcf_lock);
+ goto_ch = tcf_action_set_ctrlact(*a, actparm->action, goto_ch);
+ rcu_swap_protected(ci->params, cp_new,
+ lockdep_is_held(&ci->tcf_lock));
+ spin_unlock_bh(&ci->tcf_lock);
+
+ if (goto_ch)
+ tcf_chain_put_by_act(goto_ch);
+ if (cp_new)
+ kfree_rcu(cp_new, rcu);
+
+ if (ret == ACT_P_CREATED)
+ tcf_idr_insert(tn, *a);
+
+ return ret;
+
+put_chain:
+ if (goto_ch)
+ tcf_chain_put_by_act(goto_ch);
+release_idr:
+ tcf_idr_release(*a, bind);
+ return err;
+}
+
+static int tcf_ctinfo_dump(struct sk_buff *skb, struct tc_action *a,
+ int bind, int ref)
+{
+ struct tcf_ctinfo *ci = to_ctinfo(a);
+ struct tc_ctinfo opt = {
+ .index = ci->tcf_index,
+ .refcnt = refcount_read(&ci->tcf_refcnt) - ref,
+ .bindcnt = atomic_read(&ci->tcf_bindcnt) - bind,
+ };
+ unsigned char *b = skb_tail_pointer(skb);
+ struct tcf_ctinfo_params *cp;
+ struct tcf_t t;
+
+ spin_lock_bh(&ci->tcf_lock);
+ cp = rcu_dereference_protected(ci->params,
+ lockdep_is_held(&ci->tcf_lock));
+
+ tcf_tm_dump(&t, &ci->tcf_tm);
+ if (nla_put_64bit(skb, TCA_CTINFO_TM, sizeof(t), &t, TCA_CTINFO_PAD))
+ goto nla_put_failure;
+
+ opt.action = ci->tcf_action;
+ if (nla_put(skb, TCA_CTINFO_ACT, sizeof(opt), &opt))
+ goto nla_put_failure;
+
+ if (nla_put_u16(skb, TCA_CTINFO_ZONE, cp->zone))
+ goto nla_put_failure;
+
+ if (cp->mode & CTINFO_MODE_DSCP) {
+ if (nla_put_u32(skb, TCA_CTINFO_PARMS_DSCP_MASK,
+ cp->dscpmask))
+ goto nla_put_failure;
+ if (nla_put_u32(skb, TCA_CTINFO_PARMS_DSCP_STATEMASK,
+ cp->dscpstatemask))
+ goto nla_put_failure;
+ }
+
+ if (cp->mode & CTINFO_MODE_CPMARK) {
+ if (nla_put_u32(skb, TCA_CTINFO_PARMS_CPMARK_MASK,
+ cp->cpmarkmask))
+ goto nla_put_failure;
+ }
+
+ if (nla_put_u64_64bit(skb, TCA_CTINFO_STATS_DSCP_SET,
+ ci->stats_dscp_set, TCA_CTINFO_PAD))
+ goto nla_put_failure;
+
+ if (nla_put_u64_64bit(skb, TCA_CTINFO_STATS_DSCP_ERROR,
+ ci->stats_dscp_error, TCA_CTINFO_PAD))
+ goto nla_put_failure;
+
+ if (nla_put_u64_64bit(skb, TCA_CTINFO_STATS_CPMARK_SET,
+ ci->stats_cpmark_set, TCA_CTINFO_PAD))
+ goto nla_put_failure;
+
+ spin_unlock_bh(&ci->tcf_lock);
+ return skb->len;
+
+nla_put_failure:
+ spin_unlock_bh(&ci->tcf_lock);
+ nlmsg_trim(skb, b);
+ return -1;
+}
+
+static int tcf_ctinfo_walker(struct net *net, struct sk_buff *skb,
+ struct netlink_callback *cb, int type,
+ const struct tc_action_ops *ops,
+ struct netlink_ext_ack *extack)
+{
+ struct tc_action_net *tn = net_generic(net, ctinfo_net_id);
+
+ return tcf_generic_walker(tn, skb, cb, type, ops, extack);
+}
+
+static int tcf_ctinfo_search(struct net *net, struct tc_action **a, u32 index)
+{
+ struct tc_action_net *tn = net_generic(net, ctinfo_net_id);
+
+ return tcf_idr_search(tn, a, index);
+}
+
+static struct tc_action_ops act_ctinfo_ops = {
+ .kind = "ctinfo",
+ .id = TCA_ID_CTINFO,
+ .owner = THIS_MODULE,
+ .act = tcf_ctinfo_act,
+ .dump = tcf_ctinfo_dump,
+ .init = tcf_ctinfo_init,
+ .walk = tcf_ctinfo_walker,
+ .lookup = tcf_ctinfo_search,
+ .size = sizeof(struct tcf_ctinfo),
+};
+
+static __net_init int ctinfo_init_net(struct net *net)
+{
+ struct tc_action_net *tn = net_generic(net, ctinfo_net_id);
+
+ return tc_action_net_init(tn, &act_ctinfo_ops);
+}
+
+static void __net_exit ctinfo_exit_net(struct list_head *net_list)
+{
+ tc_action_net_exit(net_list, ctinfo_net_id);
+}
+
+static struct pernet_operations ctinfo_net_ops = {
+ .init = ctinfo_init_net,
+ .exit_batch = ctinfo_exit_net,
+ .id = &ctinfo_net_id,
+ .size = sizeof(struct tc_action_net),
+};
+
+static int __init ctinfo_init_module(void)
+{
+ return tcf_register_action(&act_ctinfo_ops, &ctinfo_net_ops);
+}
+
+static void __exit ctinfo_cleanup_module(void)
+{
+ tcf_unregister_action(&act_ctinfo_ops, &ctinfo_net_ops);
+}
+
+module_init(ctinfo_init_module);
+module_exit(ctinfo_cleanup_module);
+MODULE_AUTHOR("Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>");
+MODULE_DESCRIPTION("Connection tracking mark actions");
+MODULE_LICENSE("GPL");