Message-Id: <1473713206-2450-1-git-send-email-jhs@emojatatu.com>
Date:   Mon, 12 Sep 2016 16:46:46 -0400
From:   Jamal Hadi Salim <jhs@...atatu.com>
To:     davem@...emloft.net
Cc:     netdev@...r.kernel.org, xiyou.wangcong@...il.com,
        daniel@...earbox.net, eric.dumazet@...il.com,
        john.r.fastabend@...el.com, Jamal Hadi Salim <jhs@...atatu.com>
Subject: [PATCH v4 net-next 1/1] net_sched: Introduce skbmod action

From: Jamal Hadi Salim <jhs@...atatu.com>

This action is intended to be an upgrade from a usability perspective
from pedit (as well as operational debugability).
Compare this:

sudo tc filter add dev $ETH parent 1: protocol ip prio 10 \
u32 match ip protocol 1 0xff flowid 1:2 \
action pedit munge offset -14 u8 set 0x02 \
munge offset -13 u8 set 0x15 \
munge offset -12 u8 set 0x15 \
munge offset -11 u8 set 0x15 \
munge offset -10 u16 set 0x1515 \
pipe

To: 

sudo tc filter add dev $ETH parent 1: protocol ip prio 10 \
u32 match ip protocol 1 0xff flowid 1:2 \
action skbmod dmac 02:15:15:15:15:15

Also try to do a MAC address swap with pedit or worse
try to debug a policy with destination mac, source mac and
etherype. Then make few rules out of those and you'll get my point.

In the future common use cases on pedit can be migrated to this action
(as an example different fields in ip v4/6, transports like tcp/udp/sctp
etc). For this first cut, this allows modifying basic ethernet header.

The most important ethernet use case at the moment is when redirecting or
mirroring packets to a remote machine. The dst mac address needs a re-write
so that it doesn't get dropped or confuse an interconnecting (learning) switch
or dropped by a target machine (which looks at the dst mac). And at times
when flipping back the packet a swap of the MAC addresses is needed.

Signed-off-by: Jamal Hadi Salim <jhs@...atatu.com>
---
 include/net/tc_act/tc_skbmod.h        |  30 ++++
 include/uapi/linux/tc_act/tc_skbmod.h |  39 +++++
 net/sched/Kconfig                     |  11 ++
 net/sched/Makefile                    |   1 +
 net/sched/act_skbmod.c                | 293 ++++++++++++++++++++++++++++++++++
 5 files changed, 374 insertions(+)
 create mode 100644 include/net/tc_act/tc_skbmod.h
 create mode 100644 include/uapi/linux/tc_act/tc_skbmod.h
 create mode 100644 net/sched/act_skbmod.c

diff --git a/include/net/tc_act/tc_skbmod.h b/include/net/tc_act/tc_skbmod.h
new file mode 100644
index 0000000..644a211
--- /dev/null
+++ b/include/net/tc_act/tc_skbmod.h
@@ -0,0 +1,30 @@
+/*
+ * Copyright (c) 2016, Jamal Hadi Salim
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+*/
+
+#ifndef __NET_TC_SKBMOD_H
+#define __NET_TC_SKBMOD_H
+
+#include <net/act_api.h>
+#include <linux/tc_act/tc_skbmod.h>
+
+struct tcf_skbmod_params {
+	struct rcu_head	rcu;
+	u64	flags; /*up to 64 types of operations; extend if needed */
+	u8	eth_dst[ETH_ALEN];
+	u16	eth_type;
+	u8	eth_src[ETH_ALEN];
+};
+
+struct tcf_skbmod {
+	struct tc_action	common;
+	struct tcf_skbmod_params __rcu *skbmod_p;
+};
+#define to_skbmod(a) ((struct tcf_skbmod *)a)
+
+#endif /* __NET_TC_SKBMOD_H */
diff --git a/include/uapi/linux/tc_act/tc_skbmod.h b/include/uapi/linux/tc_act/tc_skbmod.h
new file mode 100644
index 0000000..10fc07d
--- /dev/null
+++ b/include/uapi/linux/tc_act/tc_skbmod.h
@@ -0,0 +1,39 @@
+/*
+ * Copyright (c) 2016, Jamal Hadi Salim
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+*/
+
+#ifndef __LINUX_TC_SKBMOD_H
+#define __LINUX_TC_SKBMOD_H
+
+#include <linux/pkt_cls.h>
+
+#define TCA_ACT_SKBMOD 15
+
+#define SKBMOD_F_DMAC	0x1
+#define SKBMOD_F_SMAC	0x2
+#define SKBMOD_F_ETYPE	0x4
+#define SKBMOD_F_SWAPMAC 0x8
+
+struct tc_skbmod {
+	tc_gen;
+	__u64 flags;
+};
+
+enum {
+	TCA_SKBMOD_UNSPEC,
+	TCA_SKBMOD_TM,
+	TCA_SKBMOD_PARMS,
+	TCA_SKBMOD_DMAC,
+	TCA_SKBMOD_SMAC,
+	TCA_SKBMOD_ETYPE,
+	TCA_SKBMOD_PAD,
+	__TCA_SKBMOD_MAX
+};
+#define TCA_SKBMOD_MAX (__TCA_SKBMOD_MAX - 1)
+
+#endif
diff --git a/net/sched/Kconfig b/net/sched/Kconfig
index 72e3426..7795d5a 100644
--- a/net/sched/Kconfig
+++ b/net/sched/Kconfig
@@ -749,6 +749,17 @@ config NET_ACT_CONNMARK
 	  To compile this code as a module, choose M here: the
 	  module will be called act_connmark.
 
+config NET_ACT_SKBMOD
+        tristate "skb data modification action"
+        depends on NET_CLS_ACT
+        ---help---
+         Say Y here to allow modification of skb data
+
+         If unsure, say N.
+
+         To compile this code as a module, choose M here: the
+         module will be called act_skbmod.
+
 config NET_ACT_IFE
         tristate "Inter-FE action based on IETF ForCES InterFE LFB"
         depends on NET_CLS_ACT
diff --git a/net/sched/Makefile b/net/sched/Makefile
index b9d046b..148ae0d 100644
--- a/net/sched/Makefile
+++ b/net/sched/Makefile
@@ -19,6 +19,7 @@ obj-$(CONFIG_NET_ACT_CSUM)	+= act_csum.o
 obj-$(CONFIG_NET_ACT_VLAN)	+= act_vlan.o
 obj-$(CONFIG_NET_ACT_BPF)	+= act_bpf.o
 obj-$(CONFIG_NET_ACT_CONNMARK)	+= act_connmark.o
+obj-$(CONFIG_NET_ACT_SKBMOD)	+= act_skbmod.o
 obj-$(CONFIG_NET_ACT_IFE)	+= act_ife.o
 obj-$(CONFIG_NET_IFE_SKBMARK)	+= act_meta_mark.o
 obj-$(CONFIG_NET_IFE_SKBPRIO)	+= act_meta_skbprio.o
diff --git a/net/sched/act_skbmod.c b/net/sched/act_skbmod.c
new file mode 100644
index 0000000..40a2dfd
--- /dev/null
+++ b/net/sched/act_skbmod.c
@@ -0,0 +1,293 @@
+/*
+ * net/sched/act_skbmod.c  skb data modifier
+ *
+ * Copyright (c) 2016 Jamal Hadi Salim <jhs@...atatu.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+*/
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/skbuff.h>
+#include <linux/rtnetlink.h>
+#include <net/netlink.h>
+#include <net/pkt_sched.h>
+
+#include <linux/tc_act/tc_skbmod.h>
+#include <net/tc_act/tc_skbmod.h>
+
+#define SKBMOD_TAB_MASK     15
+
+static int skbmod_net_id;
+static struct tc_action_ops act_skbmod_ops;
+
+#define MAX_EDIT_LEN ETH_HLEN
+static int tcf_skbmod_run(struct sk_buff *skb, const struct tc_action *a,
+			  struct tcf_result *res)
+{
+	struct tcf_skbmod *d = to_skbmod(a);
+	int action;
+	struct tcf_skbmod_params *p;
+	u64 flags;
+	int err;
+
+	tcf_lastuse_update(&d->tcf_tm);
+	bstats_cpu_update(this_cpu_ptr(d->common.cpu_bstats), skb);
+
+	/* XXX: if you are going to edit more fields beyond ethernet header
+	 * (example when you add IP header replacement or vlan swap)
+	 * then MAX_EDIT_LEN needs to change appropriately
+	*/
+	err = skb_ensure_writable(skb, ETH_HLEN);
+	if (unlikely(err)) /* best policy is to drop on the floor */
+		action = TC_ACT_SHOT;
+
+	rcu_read_lock();
+	action = READ_ONCE(d->tcf_action);
+	if (unlikely(action == TC_ACT_SHOT)) {
+		qstats_overlimit_inc(this_cpu_ptr(d->common.cpu_qstats));
+		rcu_read_unlock();
+		return action;
+	}
+
+	p = rcu_dereference(d->skbmod_p);
+	flags = p->flags;
+	if (flags & SKBMOD_F_DMAC)
+		ether_addr_copy(eth_hdr(skb)->h_dest, p->eth_dst);
+	if (flags & SKBMOD_F_SMAC)
+		ether_addr_copy(eth_hdr(skb)->h_source, p->eth_src);
+	if (flags & SKBMOD_F_ETYPE)
+		eth_hdr(skb)->h_proto = p->eth_type;
+	rcu_read_unlock();
+
+	if (flags & SKBMOD_F_SWAPMAC) {
+		u16 tmpaddr[ETH_ALEN / 2]; /* ether_addr_copy() requirement */
+		/*XXX: I am sure we can come up with more efficient swapping*/
+		ether_addr_copy((u8 *)tmpaddr, eth_hdr(skb)->h_dest);
+		ether_addr_copy(eth_hdr(skb)->h_dest, eth_hdr(skb)->h_source);
+		ether_addr_copy(eth_hdr(skb)->h_source, (u8 *)tmpaddr);
+	}
+
+	return action;
+}
+
+static const struct nla_policy skbmod_policy[TCA_SKBMOD_MAX + 1] = {
+	[TCA_SKBMOD_PARMS]		= { .len = sizeof(struct tc_skbmod) },
+	[TCA_SKBMOD_DMAC]		= { .len = ETH_ALEN },
+	[TCA_SKBMOD_SMAC]		= { .len = ETH_ALEN },
+	[TCA_SKBMOD_ETYPE]		= { .type = NLA_U16 },
+};
+
+static int tcf_skbmod_init(struct net *net, struct nlattr *nla,
+			   struct nlattr *est, struct tc_action **a,
+			   int ovr, int bind)
+{
+	struct tc_action_net *tn = net_generic(net, skbmod_net_id);
+	struct nlattr *tb[TCA_SKBMOD_MAX + 1];
+	struct tcf_skbmod_params *p, *p_old;
+	struct tc_skbmod *parm;
+	struct tcf_skbmod *d;
+	bool exists = false;
+	u8 *daddr = NULL;
+	u8 *saddr = NULL;
+	u16 eth_type = 0;
+	u32 lflags = 0;
+	int ret = 0, err;
+
+	if (!nla)
+		return -EINVAL;
+
+	err = nla_parse_nested(tb, TCA_SKBMOD_MAX, nla, skbmod_policy);
+	if (err < 0)
+		return err;
+
+	if (!tb[TCA_SKBMOD_PARMS])
+		return -EINVAL;
+
+	if (tb[TCA_SKBMOD_DMAC]) {
+		daddr = nla_data(tb[TCA_SKBMOD_DMAC]);
+		lflags |= SKBMOD_F_DMAC;
+	}
+
+	if (tb[TCA_SKBMOD_SMAC]) {
+		saddr = nla_data(tb[TCA_SKBMOD_SMAC]);
+		lflags |= SKBMOD_F_SMAC;
+	}
+
+	if (tb[TCA_SKBMOD_ETYPE]) {
+		eth_type = nla_get_u16(tb[TCA_SKBMOD_ETYPE]);
+		lflags |= SKBMOD_F_ETYPE;
+	}
+
+	parm = nla_data(tb[TCA_SKBMOD_PARMS]);
+	if (parm->flags & SKBMOD_F_SWAPMAC)
+		lflags = SKBMOD_F_SWAPMAC;
+
+	exists = tcf_hash_check(tn, parm->index, a, bind);
+	if (exists && bind)
+		return 0;
+
+	if (!lflags)
+		return -EINVAL;
+
+	if (!exists) {
+		ret = tcf_hash_create(tn, parm->index, est, a,
+				      &act_skbmod_ops, bind, true);
+		if (ret)
+			return ret;
+
+		ret = ACT_P_CREATED;
+	} else {
+		tcf_hash_release(*a, bind);
+		if (!ovr)
+			return -EEXIST;
+	}
+
+	d = to_skbmod(*a);
+
+	ASSERT_RTNL();
+	p = kzalloc(sizeof(struct tcf_skbmod_params), GFP_KERNEL);
+	if (unlikely(!p)) {
+		if (ovr)
+			tcf_hash_release(*a, bind);
+		return -ENOMEM;
+	}
+
+	p->flags = lflags;
+	d->tcf_action = parm->action;
+
+	p_old = rtnl_dereference(d->skbmod_p);
+
+	if (ovr)
+		spin_lock_bh(&d->tcf_lock);
+
+	if (lflags & SKBMOD_F_DMAC)
+		ether_addr_copy(p->eth_dst, daddr);
+	if (lflags & SKBMOD_F_SMAC)
+		ether_addr_copy(p->eth_src, saddr);
+	if (lflags & SKBMOD_F_ETYPE)
+		p->eth_type = htons(eth_type);
+
+	rcu_assign_pointer(d->skbmod_p, p);
+	if (ovr)
+		spin_unlock_bh(&d->tcf_lock);
+
+	if (p_old)
+		kfree_rcu(p_old, rcu);
+
+	if (ret == ACT_P_CREATED)
+		tcf_hash_insert(tn, *a);
+	return ret;
+}
+
+static int tcf_skbmod_dump(struct sk_buff *skb, struct tc_action *a,
+			   int bind, int ref)
+{
+	struct tcf_skbmod *d = to_skbmod(a);
+	unsigned char *b = skb_tail_pointer(skb);
+	struct tcf_skbmod_params  *p = rtnl_dereference(d->skbmod_p);
+	struct tc_skbmod opt = {
+		.index   = d->tcf_index,
+		.refcnt  = d->tcf_refcnt - ref,
+		.bindcnt = d->tcf_bindcnt - bind,
+		.action  = d->tcf_action,
+	};
+	struct tcf_t t;
+
+	rcu_read_lock();
+
+	opt.flags  = p->flags;
+	if (nla_put(skb, TCA_SKBMOD_PARMS, sizeof(opt), &opt))
+		goto nla_put_failure;
+	if ((p->flags & SKBMOD_F_DMAC) &&
+	    nla_put(skb, TCA_SKBMOD_DMAC, ETH_ALEN, p->eth_dst))
+		goto nla_put_failure;
+	if ((p->flags & SKBMOD_F_SMAC) &&
+	    nla_put(skb, TCA_SKBMOD_SMAC, ETH_ALEN, p->eth_src))
+		goto nla_put_failure;
+	if ((p->flags & SKBMOD_F_ETYPE) &&
+	    nla_put_u16(skb, TCA_SKBMOD_ETYPE, ntohs(p->eth_type)))
+		goto nla_put_failure;
+
+	tcf_tm_dump(&t, &d->tcf_tm);
+	if (nla_put_64bit(skb, TCA_SKBMOD_TM, sizeof(t), &t, TCA_SKBMOD_PAD))
+		goto nla_put_failure;
+
+	rcu_read_unlock();
+
+	return skb->len;
+nla_put_failure:
+	rcu_read_unlock();
+	nlmsg_trim(skb, b);
+	return -1;
+}
+
+static int tcf_skbmod_walker(struct net *net, struct sk_buff *skb,
+			     struct netlink_callback *cb, int type,
+			     const struct tc_action_ops *ops)
+{
+	struct tc_action_net *tn = net_generic(net, skbmod_net_id);
+
+	return tcf_generic_walker(tn, skb, cb, type, ops);
+}
+
+static int tcf_skbmod_search(struct net *net, struct tc_action **a, u32 index)
+{
+	struct tc_action_net *tn = net_generic(net, skbmod_net_id);
+
+	return tcf_hash_search(tn, a, index);
+}
+
+static struct tc_action_ops act_skbmod_ops = {
+	.kind		=	"skbmod",
+	.type		=	TCA_ACT_SKBMOD,
+	.owner		=	THIS_MODULE,
+	.act		=	tcf_skbmod_run,
+	.dump		=	tcf_skbmod_dump,
+	.init		=	tcf_skbmod_init,
+	.walk		=	tcf_skbmod_walker,
+	.lookup		=	tcf_skbmod_search,
+	.size		=	sizeof(struct tcf_skbmod),
+};
+
+static __net_init int skbmod_init_net(struct net *net)
+{
+	struct tc_action_net *tn = net_generic(net, skbmod_net_id);
+
+	return tc_action_net_init(tn, &act_skbmod_ops, SKBMOD_TAB_MASK);
+}
+
+static void __net_exit skbmod_exit_net(struct net *net)
+{
+	struct tc_action_net *tn = net_generic(net, skbmod_net_id);
+
+	tc_action_net_exit(tn);
+}
+
+static struct pernet_operations skbmod_net_ops = {
+	.init = skbmod_init_net,
+	.exit = skbmod_exit_net,
+	.id   = &skbmod_net_id,
+	.size = sizeof(struct tc_action_net),
+};
+
+MODULE_AUTHOR("Jamal Hadi Salim, <jhs@...atatu.com>");
+MODULE_DESCRIPTION("SKB data mod-ing");
+MODULE_LICENSE("GPL");
+
+static int __init skbmod_init_module(void)
+{
+	return tcf_register_action(&act_skbmod_ops, &skbmod_net_ops);
+}
+
+static void __exit skbmod_cleanup_module(void)
+{
+	tcf_unregister_action(&act_skbmod_ops, &skbmod_net_ops);
+}
+
+module_init(skbmod_init_module);
+module_exit(skbmod_cleanup_module);
-- 
1.9.1

Powered by blists - more mailing lists