diff --git a/root/target/linux/generic/hack-5.4/601-netfilter-export-udp_get_timeouts-function.patch b/root/target/linux/generic/hack-5.4/601-netfilter-export-udp_get_timeouts-function.patch new file mode 100644 index 00000000..ef3fba22 --- /dev/null +++ b/root/target/linux/generic/hack-5.4/601-netfilter-export-udp_get_timeouts-function.patch @@ -0,0 +1,34 @@ +From ba041eb8000e4a8f556ca641335894f7e7429dbb Mon Sep 17 00:00:00 2001 +From: Ailick <277498654@qq.com> +Date: Sat, 21 Aug 2021 17:12:22 +0800 +Subject: [PATCH] net: patch linux kernel to support shortcut-fe-cm + +--- + nf_conntrack_proto_udp.c | 3 ++- + nf_conntrack_timeout.h | 2 + + 1 file changed, 2 insertion(+) + +--- a/include/net/netfilter/nf_conntrack_timeout.h ++++ b/include/net/netfilter/nf_conntrack_timeout.h +@@ -124,4 +124,6 @@ extern struct nf_ct_timeout *(*nf_ct_tim + extern void (*nf_ct_timeout_put_hook)(struct nf_ct_timeout *timeout); + #endif + ++extern unsigned int *udp_get_timeouts(struct net *net); ++ + #endif /* _NF_CONNTRACK_TIMEOUT_H */ +--- a/net/netfilter/nf_conntrack_proto_udp.c ++++ b/net/netfilter/nf_conntrack_proto_udp.c +@@ -29,10 +29,11 @@ static const unsigned int udp_timeouts[U + [UDP_CT_REPLIED] = 120*HZ, + }; + +-static unsigned int *udp_get_timeouts(struct net *net) ++unsigned int *udp_get_timeouts(struct net *net) + { + return nf_udp_pernet(net)->timeouts; + } ++EXPORT_SYMBOL(udp_get_timeouts); + + static void udp_error_log(const struct sk_buff *skb, + const struct nf_hook_state *state, diff --git a/root/target/linux/generic/hack-5.4/645-netfilter-connmark-introduce-set-dscpmark.patch b/root/target/linux/generic/hack-5.4/645-netfilter-connmark-introduce-set-dscpmark.patch new file mode 100644 index 00000000..2d3fe01a --- /dev/null +++ b/root/target/linux/generic/hack-5.4/645-netfilter-connmark-introduce-set-dscpmark.patch @@ -0,0 +1,212 @@ +From eda40b8c8c82e0f2789d6bc8bf63846dce2e8f32 Mon Sep 17 00:00:00 2001 +From: Kevin Darbyshire-Bryant +Date: Sat, 23 Mar 2019 09:29:49 +0000 +Subject: [PATCH] netfilter: connmark: introduce set-dscpmark + +set-dscpmark is a method of storing the DSCP of an ip packet into +conntrack mark. In combination with a suitable tc filter action +(act_ctinfo) DSCP values are able to be stored in the mark on egress and +restored on ingress across links that otherwise alter or bleach DSCP. + +This is useful for qdiscs such as CAKE which are able to shape according +to policies based on DSCP. + +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. + +x_tables CONNMARK set-dscpmark target solves the problem of storing the +DSCP to the conntrack mark in a way suitable for the new act_ctinfo tc +action to restore. + +The set-dscpmark option accepts 2 parameters, a 32bit 'dscpmask' and a +32bit 'statemask'. The dscp mask must be 6 contiguous bits and +represents the area where the DSCP will be stored in the connmark. The +state mask is a minimum 1 bit length mask that must not overlap with the +dscpmask. It represents a flag which is set when the DSCP has been +stored in the conntrack mark. 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 state +mask of zero disables the setting of a status bit/s. + +example syntax with a suitably modified iptables user space application: + +iptables -A QOS_MARK_eth0 -t mangle -j CONNMARK --set-dscpmark 0xfc000000/0x01000000 + +Would store the DSCP in the top 6 bits of the 32bit mark field, and use +the LSB of the top byte as the 'DSCP has been stored' marker. + +|----0xFC----conntrack mark----000000---| +| Bits 31-26 | bit 25 | bit24 |~~~ Bit 0| +| DSCP | unused | flag |unused | +|-----------------------0x01---000000---| + ^ ^ + | | + ---| Conditional flag + | set this when dscp +|-ip diffserv-| stored in mark +| 6 bits | +|-------------| + +an identically configured tc action to restore looks like: + +tc filter show dev eth0 ingress +filter parent ffff: protocol all pref 10 u32 chain 0 +filter parent ffff: protocol all pref 10 u32 chain 0 fh 800: ht divisor 1 +filter parent ffff: protocol all pref 10 u32 chain 0 fh 800::800 order 2048 key ht 800 bkt 0 flowid 1: not_in_hw + match 00000000/00000000 at 0 + action order 1: ctinfo zone 0 pipe + index 2 ref 1 bind 1 dscp 0xfc000000/0x1000000 + + action order 2: mirred (Egress Redirect to device ifb4eth0) stolen + index 1 ref 1 bind 1 + +|----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 | +|-------------| + +Signed-off-by: Kevin Darbyshire-Bryant +--- + include/uapi/linux/netfilter/xt_connmark.h | 10 ++++ + net/netfilter/xt_connmark.c | 55 ++++++++++++++++++---- + 2 files changed, 57 insertions(+), 8 deletions(-) + +--- a/include/uapi/linux/netfilter/xt_connmark.h ++++ b/include/uapi/linux/netfilter/xt_connmark.h +@@ -20,6 +20,11 @@ enum { + }; + + enum { ++ XT_CONNMARK_VALUE = (1 << 0), ++ XT_CONNMARK_DSCP = (1 << 1) ++}; ++ ++enum { + D_SHIFT_LEFT = 0, + D_SHIFT_RIGHT, + }; +@@ -34,6 +39,11 @@ struct xt_connmark_tginfo2 { + __u8 shift_dir, shift_bits, mode; + }; + ++struct xt_connmark_tginfo3 { ++ __u32 ctmark, ctmask, nfmask; ++ __u8 shift_dir, shift_bits, mode, func; ++}; ++ + struct xt_connmark_mtinfo1 { + __u32 mark, mask; + __u8 invert; +--- a/net/netfilter/xt_connmark.c ++++ b/net/netfilter/xt_connmark.c +@@ -24,12 +24,13 @@ MODULE_ALIAS("ipt_connmark"); + MODULE_ALIAS("ip6t_connmark"); + + static unsigned int +-connmark_tg_shift(struct sk_buff *skb, const struct xt_connmark_tginfo2 *info) ++connmark_tg_shift(struct sk_buff *skb, const struct xt_connmark_tginfo3 *info) + { + enum ip_conntrack_info ctinfo; + u_int32_t new_targetmark; + struct nf_conn *ct; + u_int32_t newmark; ++ u_int8_t dscp; + + ct = nf_ct_get(skb, &ctinfo); + if (ct == NULL) +@@ -37,12 +38,24 @@ connmark_tg_shift(struct sk_buff *skb, c + + switch (info->mode) { + case XT_CONNMARK_SET: +- newmark = (ct->mark & ~info->ctmask) ^ info->ctmark; +- if (info->shift_dir == D_SHIFT_RIGHT) +- newmark >>= info->shift_bits; +- else +- newmark <<= info->shift_bits; ++ newmark = ct->mark; ++ if (info->func & XT_CONNMARK_VALUE) { ++ newmark = (newmark & ~info->ctmask) ^ info->ctmark; ++ if (info->shift_dir == D_SHIFT_RIGHT) ++ newmark >>= info->shift_bits; ++ else ++ newmark <<= info->shift_bits; ++ } else if (info->func & XT_CONNMARK_DSCP) { ++ if (skb->protocol == htons(ETH_P_IP)) ++ dscp = ipv4_get_dsfield(ip_hdr(skb)) >> 2; ++ else if (skb->protocol == htons(ETH_P_IPV6)) ++ dscp = ipv6_get_dsfield(ipv6_hdr(skb)) >> 2; ++ else /* protocol doesn't have diffserv */ ++ break; + ++ newmark = (newmark & ~info->ctmark) | ++ (info->ctmask | (dscp << info->shift_bits)); ++ } + if (ct->mark != newmark) { + ct->mark = newmark; + nf_conntrack_event_cache(IPCT_MARK, ct); +@@ -81,20 +94,36 @@ static unsigned int + connmark_tg(struct sk_buff *skb, const struct xt_action_param *par) + { + const struct xt_connmark_tginfo1 *info = par->targinfo; +- const struct xt_connmark_tginfo2 info2 = { ++ const struct xt_connmark_tginfo3 info3 = { + .ctmark = info->ctmark, + .ctmask = info->ctmask, + .nfmask = info->nfmask, + .mode = info->mode, ++ .func = XT_CONNMARK_VALUE + }; + +- return connmark_tg_shift(skb, &info2); ++ return connmark_tg_shift(skb, &info3); + } + + static unsigned int + connmark_tg_v2(struct sk_buff *skb, const struct xt_action_param *par) + { + const struct xt_connmark_tginfo2 *info = par->targinfo; ++ const struct xt_connmark_tginfo3 info3 = { ++ .ctmark = info->ctmark, ++ .ctmask = info->ctmask, ++ .nfmask = info->nfmask, ++ .mode = info->mode, ++ .func = XT_CONNMARK_VALUE ++ }; ++ ++ return connmark_tg_shift(skb, &info3); ++} ++ ++static unsigned int ++connmark_tg_v3(struct sk_buff *skb, const struct xt_action_param *par) ++{ ++ const struct xt_connmark_tginfo3 *info = par->targinfo; + + return connmark_tg_shift(skb, info); + } +@@ -165,6 +194,16 @@ static struct xt_target connmark_tg_reg[ + .targetsize = sizeof(struct xt_connmark_tginfo2), + .destroy = connmark_tg_destroy, + .me = THIS_MODULE, ++ }, ++ { ++ .name = "CONNMARK", ++ .revision = 3, ++ .family = NFPROTO_UNSPEC, ++ .checkentry = connmark_tg_check, ++ .target = connmark_tg_v3, ++ .targetsize = sizeof(struct xt_connmark_tginfo3), ++ .destroy = connmark_tg_destroy, ++ .me = THIS_MODULE, + } + }; + diff --git a/root/target/linux/generic/hack-5.4/647-netfilter-flow-acct.patch b/root/target/linux/generic/hack-5.4/647-netfilter-flow-acct.patch new file mode 100644 index 00000000..9f7ca615 --- /dev/null +++ b/root/target/linux/generic/hack-5.4/647-netfilter-flow-acct.patch @@ -0,0 +1,70 @@ +--- a/include/net/netfilter/nf_flow_table.h ++++ b/include/net/netfilter/nf_flow_table.h +@@ -163,6 +163,8 @@ struct nf_flow_table_hw { + int nf_flow_table_hw_register(const struct nf_flow_table_hw *offload); + void nf_flow_table_hw_unregister(const struct nf_flow_table_hw *offload); + ++void nf_flow_table_acct(struct flow_offload *flow, struct sk_buff *skb, int dir); ++ + extern struct work_struct nf_flow_offload_hw_work; + + #define MODULE_ALIAS_NF_FLOWTABLE(family) \ +--- a/net/netfilter/nf_flow_table_core.c ++++ b/net/netfilter/nf_flow_table_core.c +@@ -13,6 +13,7 @@ + #include + #include + #include ++#include + + struct flow_offload_entry { + struct flow_offload flow; +@@ -177,6 +178,22 @@ void flow_offload_free(struct flow_offlo + } + EXPORT_SYMBOL_GPL(flow_offload_free); + ++void nf_flow_table_acct(struct flow_offload *flow, struct sk_buff *skb, int dir) ++{ ++ struct flow_offload_entry *entry; ++ struct nf_conn_acct *acct; ++ ++ entry = container_of(flow, struct flow_offload_entry, flow); ++ acct = nf_conn_acct_find(entry->ct); ++ if (acct) { ++ struct nf_conn_counter *counter = acct->counter; ++ ++ atomic64_inc(&counter[dir].packets); ++ atomic64_add(skb->len, &counter[dir].bytes); ++ } ++} ++EXPORT_SYMBOL_GPL(nf_flow_table_acct); ++ + static u32 flow_offload_hash(const void *data, u32 len, u32 seed) + { + const struct flow_offload_tuple *tuple = data; +--- a/net/netfilter/nf_flow_table_ip.c ++++ b/net/netfilter/nf_flow_table_ip.c +@@ -12,6 +12,7 @@ + #include + #include + #include ++ + /* For layer 4 checksum field offset. */ + #include + #include +@@ -288,6 +289,7 @@ nf_flow_offload_ip_hook(void *priv, stru + skb->dev = outdev; + nexthop = rt_nexthop(rt, flow->tuplehash[!dir].tuple.src_v4.s_addr); + skb_dst_set_noref(skb, &rt->dst); ++ nf_flow_table_acct(flow, skb, dir); + neigh_xmit(NEIGH_ARP_TABLE, outdev, &nexthop, skb); + + return NF_STOLEN; +@@ -518,6 +520,7 @@ nf_flow_offload_ipv6_hook(void *priv, st + skb->dev = outdev; + nexthop = rt6_nexthop(rt, &flow->tuplehash[!dir].tuple.src_v6); + skb_dst_set_noref(skb, &rt->dst); ++ nf_flow_table_acct(flow, skb, dir); + neigh_xmit(NEIGH_ND_TABLE, outdev, nexthop, skb); + + return NF_STOLEN; diff --git a/root/target/linux/generic/hack-5.4/650-netfilter-add-xt_OFFLOAD-target.patch b/root/target/linux/generic/hack-5.4/650-netfilter-add-xt_OFFLOAD-target.patch new file mode 100644 index 00000000..8d9fec06 --- /dev/null +++ b/root/target/linux/generic/hack-5.4/650-netfilter-add-xt_OFFLOAD-target.patch @@ -0,0 +1,589 @@ +From: Felix Fietkau +Date: Tue, 20 Feb 2018 15:56:02 +0100 +Subject: [PATCH] netfilter: add xt_OFFLOAD target + +Signed-off-by: Felix Fietkau +--- + create mode 100644 net/netfilter/xt_OFFLOAD.c + +--- a/net/ipv4/netfilter/Kconfig ++++ b/net/ipv4/netfilter/Kconfig +@@ -56,8 +56,6 @@ config NF_TABLES_ARP + help + This option enables the ARP support for nf_tables. + +-endif # NF_TABLES +- + config NF_FLOW_TABLE_IPV4 + tristate "Netfilter flow table IPv4 module" + depends on NF_FLOW_TABLE +@@ -66,6 +64,8 @@ config NF_FLOW_TABLE_IPV4 + + To compile it as a module, choose M here. + ++endif # NF_TABLES ++ + config NF_DUP_IPV4 + tristate "Netfilter IPv4 packet duplication to alternate destination" + depends on !NF_CONNTRACK || NF_CONNTRACK +--- a/net/ipv6/netfilter/Kconfig ++++ b/net/ipv6/netfilter/Kconfig +@@ -45,7 +45,6 @@ config NFT_FIB_IPV6 + multicast or blackhole. + + endif # NF_TABLES_IPV6 +-endif # NF_TABLES + + config NF_FLOW_TABLE_IPV6 + tristate "Netfilter flow table IPv6 module" +@@ -55,6 +54,8 @@ config NF_FLOW_TABLE_IPV6 + + To compile it as a module, choose M here. + ++endif # NF_TABLES ++ + config NF_DUP_IPV6 + tristate "Netfilter IPv6 packet duplication to alternate destination" + depends on !NF_CONNTRACK || NF_CONNTRACK +--- a/net/netfilter/Kconfig ++++ b/net/netfilter/Kconfig +@@ -689,8 +689,6 @@ config NFT_FIB_NETDEV + + endif # NF_TABLES_NETDEV + +-endif # NF_TABLES +- + config NF_FLOW_TABLE_INET + tristate "Netfilter flow table mixed IPv4/IPv6 module" + depends on NF_FLOW_TABLE +@@ -699,11 +697,12 @@ config NF_FLOW_TABLE_INET + + To compile it as a module, choose M here. + ++endif # NF_TABLES ++ + config NF_FLOW_TABLE + tristate "Netfilter flow table module" + depends on NETFILTER_INGRESS + depends on NF_CONNTRACK +- depends on NF_TABLES + help + This option adds the flow table core infrastructure. + +@@ -992,6 +991,15 @@ config NETFILTER_XT_TARGET_NOTRACK + depends on NETFILTER_ADVANCED + select NETFILTER_XT_TARGET_CT + ++config NETFILTER_XT_TARGET_FLOWOFFLOAD ++ tristate '"FLOWOFFLOAD" target support' ++ depends on NF_FLOW_TABLE ++ depends on NETFILTER_INGRESS ++ help ++ This option adds a `FLOWOFFLOAD' target, which uses the nf_flow_offload ++ module to speed up processing of packets by bypassing the usual ++ netfilter chains ++ + config NETFILTER_XT_TARGET_RATEEST + tristate '"RATEEST" target support' + depends on NETFILTER_ADVANCED +--- a/net/netfilter/Makefile ++++ b/net/netfilter/Makefile +@@ -141,6 +141,7 @@ obj-$(CONFIG_NETFILTER_XT_TARGET_CLASSIF + obj-$(CONFIG_NETFILTER_XT_TARGET_CONNSECMARK) += xt_CONNSECMARK.o + obj-$(CONFIG_NETFILTER_XT_TARGET_CT) += xt_CT.o + obj-$(CONFIG_NETFILTER_XT_TARGET_DSCP) += xt_DSCP.o ++obj-$(CONFIG_NETFILTER_XT_TARGET_FLOWOFFLOAD) += xt_FLOWOFFLOAD.o + obj-$(CONFIG_NETFILTER_XT_TARGET_HL) += xt_HL.o + obj-$(CONFIG_NETFILTER_XT_TARGET_HMARK) += xt_HMARK.o + obj-$(CONFIG_NETFILTER_XT_TARGET_LED) += xt_LED.o +--- /dev/null ++++ b/net/netfilter/xt_FLOWOFFLOAD.c +@@ -0,0 +1,427 @@ ++/* ++ * Copyright (C) 2018 Felix Fietkau ++ * ++ * This program is free software; you can redistribute it and/or modify ++ * it under the terms of the GNU General Public License version 2 as ++ * published by the Free Software Foundation. ++ */ ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++#include ++ ++static struct nf_flowtable nf_flowtable; ++static HLIST_HEAD(hooks); ++static DEFINE_SPINLOCK(hooks_lock); ++static struct delayed_work hook_work; ++ ++struct xt_flowoffload_hook { ++ struct hlist_node list; ++ struct nf_hook_ops ops; ++ struct net *net; ++ bool registered; ++ bool used; ++}; ++ ++static unsigned int ++xt_flowoffload_net_hook(void *priv, struct sk_buff *skb, ++ const struct nf_hook_state *state) ++{ ++ switch (skb->protocol) { ++ case htons(ETH_P_IP): ++ return nf_flow_offload_ip_hook(priv, skb, state); ++ case htons(ETH_P_IPV6): ++ return nf_flow_offload_ipv6_hook(priv, skb, state); ++ } ++ ++ return NF_ACCEPT; ++} ++ ++int nf_flow_table_iterate(struct nf_flowtable *flow_table, ++ void (*iter)(struct flow_offload *flow, void *data), ++ void *data); ++ ++static int ++xt_flowoffload_create_hook(struct net_device *dev) ++{ ++ struct xt_flowoffload_hook *hook; ++ struct nf_hook_ops *ops; ++ ++ hook = kzalloc(sizeof(*hook), GFP_ATOMIC); ++ if (!hook) ++ return -ENOMEM; ++ ++ ops = &hook->ops; ++ ops->pf = NFPROTO_NETDEV; ++ ops->hooknum = NF_NETDEV_INGRESS; ++ ops->priority = 10; ++ ops->priv = &nf_flowtable; ++ ops->hook = xt_flowoffload_net_hook; ++ ops->dev = dev; ++ ++ hlist_add_head(&hook->list, &hooks); ++ mod_delayed_work(system_power_efficient_wq, &hook_work, 0); ++ ++ return 0; ++} ++ ++static struct xt_flowoffload_hook * ++flow_offload_lookup_hook(struct net_device *dev) ++{ ++ struct xt_flowoffload_hook *hook; ++ ++ hlist_for_each_entry(hook, &hooks, list) { ++ if (hook->ops.dev == dev) ++ return hook; ++ } ++ ++ return NULL; ++} ++ ++static void ++xt_flowoffload_check_device(struct net_device *dev) ++{ ++ struct xt_flowoffload_hook *hook; ++ ++ spin_lock_bh(&hooks_lock); ++ hook = flow_offload_lookup_hook(dev); ++ if (hook) ++ hook->used = true; ++ else ++ xt_flowoffload_create_hook(dev); ++ spin_unlock_bh(&hooks_lock); ++} ++ ++static void ++xt_flowoffload_register_hooks(void) ++{ ++ struct xt_flowoffload_hook *hook; ++ ++restart: ++ hlist_for_each_entry(hook, &hooks, list) { ++ if (hook->registered) ++ continue; ++ ++ hook->registered = true; ++ hook->net = dev_net(hook->ops.dev); ++ spin_unlock_bh(&hooks_lock); ++ nf_register_net_hook(hook->net, &hook->ops); ++ spin_lock_bh(&hooks_lock); ++ goto restart; ++ } ++ ++} ++ ++static void ++xt_flowoffload_cleanup_hooks(void) ++{ ++ struct xt_flowoffload_hook *hook; ++ ++restart: ++ hlist_for_each_entry(hook, &hooks, list) { ++ if (hook->used || !hook->registered) ++ continue; ++ ++ hlist_del(&hook->list); ++ spin_unlock_bh(&hooks_lock); ++ nf_unregister_net_hook(hook->net, &hook->ops); ++ kfree(hook); ++ spin_lock_bh(&hooks_lock); ++ goto restart; ++ } ++ ++} ++ ++static void ++xt_flowoffload_check_hook(struct flow_offload *flow, void *data) ++{ ++ struct flow_offload_tuple *tuple = &flow->tuplehash[0].tuple; ++ struct xt_flowoffload_hook *hook; ++ bool *found = data; ++ struct rtable *rt = (struct rtable *)tuple->dst_cache; ++ ++ spin_lock_bh(&hooks_lock); ++ hlist_for_each_entry(hook, &hooks, list) { ++ if (hook->ops.dev->ifindex != tuple->iifidx && ++ hook->ops.dev->ifindex != rt->dst.dev->ifindex) ++ continue; ++ ++ hook->used = true; ++ *found = true; ++ } ++ spin_unlock_bh(&hooks_lock); ++} ++ ++static void ++xt_flowoffload_hook_work(struct work_struct *work) ++{ ++ struct xt_flowoffload_hook *hook; ++ bool found = false; ++ int err; ++ ++ spin_lock_bh(&hooks_lock); ++ xt_flowoffload_register_hooks(); ++ hlist_for_each_entry(hook, &hooks, list) ++ hook->used = false; ++ spin_unlock_bh(&hooks_lock); ++ ++ err = nf_flow_table_iterate(&nf_flowtable, xt_flowoffload_check_hook, ++ &found); ++ if (err && err != -EAGAIN) ++ goto out; ++ ++ spin_lock_bh(&hooks_lock); ++ xt_flowoffload_cleanup_hooks(); ++ spin_unlock_bh(&hooks_lock); ++ ++out: ++ if (found) ++ queue_delayed_work(system_power_efficient_wq, &hook_work, HZ); ++} ++ ++static bool ++xt_flowoffload_skip(struct sk_buff *skb, int family) ++{ ++ if (skb_sec_path(skb)) ++ return true; ++ ++ if (family == NFPROTO_IPV4) { ++ const struct ip_options *opt = &(IPCB(skb)->opt); ++ ++ if (unlikely(opt->optlen)) ++ return true; ++ } ++ ++ return false; ++} ++ ++static struct dst_entry * ++xt_flowoffload_dst(const struct nf_conn *ct, enum ip_conntrack_dir dir, ++ const struct xt_action_param *par, int ifindex) ++{ ++ struct dst_entry *dst = NULL; ++ struct flowi fl; ++ ++ memset(&fl, 0, sizeof(fl)); ++ switch (xt_family(par)) { ++ case NFPROTO_IPV4: ++ fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip; ++ fl.u.ip4.flowi4_oif = ifindex; ++ break; ++ case NFPROTO_IPV6: ++ fl.u.ip6.saddr = ct->tuplehash[dir].tuple.dst.u3.in6; ++ fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6; ++ fl.u.ip6.flowi6_oif = ifindex; ++ break; ++ } ++ ++ nf_route(xt_net(par), &dst, &fl, false, xt_family(par)); ++ ++ return dst; ++} ++ ++static int ++xt_flowoffload_route(struct sk_buff *skb, const struct nf_conn *ct, ++ const struct xt_action_param *par, ++ struct nf_flow_route *route, enum ip_conntrack_dir dir) ++{ ++ struct dst_entry *this_dst, *other_dst; ++ ++ this_dst = xt_flowoffload_dst(ct, !dir, par, xt_out(par)->ifindex); ++ other_dst = xt_flowoffload_dst(ct, dir, par, xt_in(par)->ifindex); ++ ++ route->tuple[dir].dst = this_dst; ++ route->tuple[!dir].dst = other_dst; ++ ++ if (!this_dst || !other_dst) ++ return -ENOENT; ++ ++ if (dst_xfrm(this_dst) || dst_xfrm(other_dst)) ++ return -EINVAL; ++ ++ return 0; ++} ++ ++static unsigned int ++flowoffload_tg(struct sk_buff *skb, const struct xt_action_param *par) ++{ ++ const struct xt_flowoffload_target_info *info = par->targinfo; ++ struct tcphdr _tcph, *tcph = NULL; ++ enum ip_conntrack_info ctinfo; ++ enum ip_conntrack_dir dir; ++ struct nf_flow_route route; ++ struct flow_offload *flow = NULL; ++ struct nf_conn *ct; ++ struct net *net; ++ ++ if (xt_flowoffload_skip(skb, xt_family(par))) ++ return XT_CONTINUE; ++ ++ ct = nf_ct_get(skb, &ctinfo); ++ if (ct == NULL) ++ return XT_CONTINUE; ++ ++ switch (ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum) { ++ case IPPROTO_TCP: ++ if (ct->proto.tcp.state != TCP_CONNTRACK_ESTABLISHED) ++ return XT_CONTINUE; ++ ++ tcph = skb_header_pointer(skb, par->thoff, ++ sizeof(_tcph), &_tcph); ++ if (unlikely(!tcph || tcph->fin || tcph->rst)) ++ return XT_CONTINUE; ++ break; ++ case IPPROTO_UDP: ++ break; ++ default: ++ return XT_CONTINUE; ++ } ++ ++ if (nf_ct_ext_exist(ct, NF_CT_EXT_HELPER) || ++ ct->status & IPS_SEQ_ADJUST) ++ return XT_CONTINUE; ++ ++ if (!nf_ct_is_confirmed(ct)) ++ return XT_CONTINUE; ++ ++ if (!xt_in(par) || !xt_out(par)) ++ return XT_CONTINUE; ++ ++ if (test_and_set_bit(IPS_OFFLOAD_BIT, &ct->status)) ++ return XT_CONTINUE; ++ ++ dir = CTINFO2DIR(ctinfo); ++ ++ if (xt_flowoffload_route(skb, ct, par, &route, dir) == 0) ++ flow = flow_offload_alloc(ct, &route); ++ ++ dst_release(route.tuple[dir].dst); ++ dst_release(route.tuple[!dir].dst); ++ ++ if (!flow) ++ goto err_flow_route; ++ ++ if (tcph) { ++ ct->proto.tcp.seen[0].flags |= IP_CT_TCP_FLAG_BE_LIBERAL; ++ ct->proto.tcp.seen[1].flags |= IP_CT_TCP_FLAG_BE_LIBERAL; ++ } ++ ++ if (flow_offload_add(&nf_flowtable, flow) < 0) ++ goto err_flow_add; ++ ++ xt_flowoffload_check_device(xt_in(par)); ++ xt_flowoffload_check_device(xt_out(par)); ++ ++ net = read_pnet(&nf_flowtable.ft_net); ++ if (!net) ++ write_pnet(&nf_flowtable.ft_net, xt_net(par)); ++ ++ if (info->flags & XT_FLOWOFFLOAD_HW) ++ nf_flow_offload_hw_add(xt_net(par), flow, ct); ++ ++ return XT_CONTINUE; ++ ++err_flow_add: ++ flow_offload_free(flow); ++err_flow_route: ++ clear_bit(IPS_OFFLOAD_BIT, &ct->status); ++ return XT_CONTINUE; ++} ++ ++ ++static int flowoffload_chk(const struct xt_tgchk_param *par) ++{ ++ struct xt_flowoffload_target_info *info = par->targinfo; ++ ++ if (info->flags & ~XT_FLOWOFFLOAD_MASK) ++ return -EINVAL; ++ ++ return 0; ++} ++ ++static struct xt_target offload_tg_reg __read_mostly = { ++ .family = NFPROTO_UNSPEC, ++ .name = "FLOWOFFLOAD", ++ .revision = 0, ++ .targetsize = sizeof(struct xt_flowoffload_target_info), ++ .usersize = sizeof(struct xt_flowoffload_target_info), ++ .checkentry = flowoffload_chk, ++ .target = flowoffload_tg, ++ .me = THIS_MODULE, ++}; ++ ++static int xt_flowoffload_table_init(struct nf_flowtable *table) ++{ ++ table->flags = NF_FLOWTABLE_F_HW; ++ nf_flow_table_init(table); ++ return 0; ++} ++ ++static void xt_flowoffload_table_cleanup(struct nf_flowtable *table) ++{ ++ nf_flow_table_free(table); ++} ++ ++static int flow_offload_netdev_event(struct notifier_block *this, ++ unsigned long event, void *ptr) ++{ ++ struct xt_flowoffload_hook *hook = NULL; ++ struct net_device *dev = netdev_notifier_info_to_dev(ptr); ++ ++ if (event != NETDEV_UNREGISTER) ++ return NOTIFY_DONE; ++ ++ spin_lock_bh(&hooks_lock); ++ hook = flow_offload_lookup_hook(dev); ++ if (hook) { ++ hlist_del(&hook->list); ++ } ++ spin_unlock_bh(&hooks_lock); ++ if (hook) { ++ nf_unregister_net_hook(hook->net, &hook->ops); ++ kfree(hook); ++ } ++ ++ nf_flow_table_cleanup(dev); ++ ++ return NOTIFY_DONE; ++} ++ ++static struct notifier_block flow_offload_netdev_notifier = { ++ .notifier_call = flow_offload_netdev_event, ++}; ++ ++static int __init xt_flowoffload_tg_init(void) ++{ ++ int ret; ++ ++ register_netdevice_notifier(&flow_offload_netdev_notifier); ++ ++ INIT_DELAYED_WORK(&hook_work, xt_flowoffload_hook_work); ++ ++ ret = xt_flowoffload_table_init(&nf_flowtable); ++ if (ret) ++ return ret; ++ ++ ret = xt_register_target(&offload_tg_reg); ++ if (ret) ++ xt_flowoffload_table_cleanup(&nf_flowtable); ++ ++ return ret; ++} ++ ++static void __exit xt_flowoffload_tg_exit(void) ++{ ++ xt_unregister_target(&offload_tg_reg); ++ xt_flowoffload_table_cleanup(&nf_flowtable); ++ unregister_netdevice_notifier(&flow_offload_netdev_notifier); ++} ++ ++MODULE_LICENSE("GPL"); ++module_init(xt_flowoffload_tg_init); ++module_exit(xt_flowoffload_tg_exit); +--- a/net/netfilter/nf_flow_table_core.c ++++ b/net/netfilter/nf_flow_table_core.c +@@ -7,7 +7,6 @@ + #include + #include + #include +-#include + #include + #include + #include +@@ -351,8 +350,7 @@ flow_offload_lookup(struct nf_flowtable + } + EXPORT_SYMBOL_GPL(flow_offload_lookup); + +-static int +-nf_flow_table_iterate(struct nf_flowtable *flow_table, ++int nf_flow_table_iterate(struct nf_flowtable *flow_table, + void (*iter)(struct flow_offload *flow, void *data), + void *data) + { +@@ -385,6 +383,7 @@ nf_flow_table_iterate(struct nf_flowtabl + + return err; + } ++EXPORT_SYMBOL_GPL(nf_flow_table_iterate); + + static void nf_flow_offload_gc_step(struct flow_offload *flow, void *data) + { +--- /dev/null ++++ b/include/uapi/linux/netfilter/xt_FLOWOFFLOAD.h +@@ -0,0 +1,17 @@ ++/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ ++#ifndef _XT_FLOWOFFLOAD_H ++#define _XT_FLOWOFFLOAD_H ++ ++#include ++ ++enum { ++ XT_FLOWOFFLOAD_HW = 1 << 0, ++ ++ XT_FLOWOFFLOAD_MASK = XT_FLOWOFFLOAD_HW ++}; ++ ++struct xt_flowoffload_target_info { ++ __u32 flags; ++}; ++ ++#endif /* _XT_FLOWOFFLOAD_H */ +--- a/include/net/netfilter/nf_flow_table.h ++++ b/include/net/netfilter/nf_flow_table.h +@@ -133,6 +133,10 @@ static inline void flow_offload_dead(str + flow->flags |= FLOW_OFFLOAD_DYING; + } + ++int nf_flow_table_iterate(struct nf_flowtable *flow_table, ++ void (*iter)(struct flow_offload *flow, void *data), ++ void *data); ++ + int nf_flow_snat_port(const struct flow_offload *flow, + struct sk_buff *skb, unsigned int thoff, + u8 protocol, enum flow_offload_tuple_dir dir);