On 07/27/2015 08:31 PM, David Ahern wrote:
> This driver borrows heavily from IPvlan and teaming drivers.
>
> Routing domains (VRF-lite) are created by instantiating a VRF master
> device with an associated table and enslaving all routed interfaces that
> participate in the domain. As part of the enslavement, all connected
> routes for the enslaved devices are moved to the table associated with
> the VRF device. Outgoing sockets must bind to the VRF device to function.
>
> Standard FIB rules bind the VRF device to tables and regular fib rule
> processing is followed. Routed traffic through the box, is forwarded by
> using the VRF device as the IIF and following the IIF rule to a table
> that is mated with the VRF.
>
> Example:
>
> Create vrf 1:
> ip link add vrf1 type vrf table 5
> ip rule add iif vrf1 table 5
> ip rule add oif vrf1 table 5
> ip route add table 5 prohibit default
> ip link set vrf1 up
>
> Add interface to vrf 1:
> ip link set eth1 master vrf1
>
> Signed-off-by: Shrijeet Mukherjee <s...@cumulusnetworks.com>
> Signed-off-by: David Ahern <d...@cumulusnetworks.com>
>
> v2:
> - addressed comments from first RFC
> - significant changes to improve simplicity of implementation
> ---
> drivers/net/Kconfig | 7 +
> drivers/net/Makefile | 1 +
> drivers/net/vrf.c | 596
> +++++++++++++++++++++++++++++++++++++++++++++++++++
> 3 files changed, 604 insertions(+)
> create mode 100644 drivers/net/vrf.c
>
> diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
> index c18f9e62a9fa..e58468b02987 100644
> --- a/drivers/net/Kconfig
> +++ b/drivers/net/Kconfig
> @@ -297,6 +297,13 @@ config NLMON
> diagnostics, etc. This is mostly intended for developers or support
> to debug netlink issues. If unsure, say N.
>
> +config NET_VRF
> + tristate "Virtual Routing and Forwarding (Lite)"
> + depends on IP_MULTIPLE_TABLES && IPV6_MULTIPLE_TABLES
> + ---help---
> + This option enables the support for mapping interfaces into VRF's. The
> + support enables VRF devices.
> +
> endif # NET_CORE
>
> config SUNGEM_PHY
> diff --git a/drivers/net/Makefile b/drivers/net/Makefile
> index c12cb22478a7..ca16dd689b36 100644
> --- a/drivers/net/Makefile
> +++ b/drivers/net/Makefile
> @@ -25,6 +25,7 @@ obj-$(CONFIG_VIRTIO_NET) += virtio_net.o
> obj-$(CONFIG_VXLAN) += vxlan.o
> obj-$(CONFIG_GENEVE) += geneve.o
> obj-$(CONFIG_NLMON) += nlmon.o
> +obj-$(CONFIG_NET_VRF) += vrf.o
>
> #
> # Networking Drivers
> diff --git a/drivers/net/vrf.c b/drivers/net/vrf.c
> new file mode 100644
> index 000000000000..8669b0f9d749
> --- /dev/null
> +++ b/drivers/net/vrf.c
> @@ -0,0 +1,596 @@
> +/*
> + * vrf.c: device driver to encapsulate a VRF space
> + *
> + * Copyright (c) 2015 Cumulus Networks
> + *
> + * Based on dummy, team and ipvlan drivers
> + *
> + * 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/kernel.h>
> +#include <linux/netdevice.h>
> +#include <linux/etherdevice.h>
> +#include <linux/ip.h>
> +#include <linux/init.h>
> +#include <linux/moduleparam.h>
> +#include <linux/rtnetlink.h>
> +#include <net/rtnetlink.h>
> +#include <linux/u64_stats_sync.h>
> +#include <linux/hashtable.h>
> +
> +#include <linux/inetdevice.h>
> +#include <net/ip.h>
> +#include <net/ip_fib.h>
> +#include <net/ip6_route.h>
> +#include <net/rtnetlink.h>
> +#include <net/route.h>
> +#include <net/addrconf.h>
> +#include <net/vrf.h>
> +
> +#define DRV_NAME "vrf"
> +#define DRV_VERSION "1.0"
> +
> +#define vrf_is_slave(dev) ((dev)->flags & IFF_SLAVE)
> +#define vrf_is_master(dev) ((dev)->flags & IFF_MASTER)
> +
> +#define vrf_master_get_rcu(dev) \
> + ((struct net_device *)rcu_dereference(dev->rx_handler_data))
> +
> +struct pcpu_dstats {
> + u64 tx_pkts;
> + u64 tx_bytes;
> + u64 tx_drps;
> + u64 rx_pkts;
> + u64 rx_bytes;
> + struct u64_stats_sync syncp;
> +};
> +
> +struct slave {
> + struct list_head list;
> + struct net_device *dev;
> +};
> +
> +struct slave_queue {
> + spinlock_t lock; /* lock for slave insert/delete */
^^^^
I don't think you actually need this lock since all VRF dev operations are done
under RTNL so you already got protection against add/del running concurrently.
It would simplify the code if you can get rid of it.
> + struct list_head all_slaves;
> + int num_slaves;
> +};
> +
> +struct net_vrf {
> + struct slave_queue queue;
> + struct fib_table *tb;
> + u32 tb_id;
> +};
> +
> +static bool is_ip_rx_frame(struct sk_buff *skb)
> +{
> + switch (skb->protocol) {
> + case htons(ETH_P_IP):
> + case htons(ETH_P_IPV6):
> + return true;
> + }
> + return false;
> +}
> +
> +/* note: already called with rcu_read_lock */
> +static rx_handler_result_t vrf_handle_frame(struct sk_buff **pskb)
> +{
> + struct sk_buff *skb = *pskb;
> +
> + if (is_ip_rx_frame(skb)) {
> + struct net_device *dev = vrf_master_get_rcu(skb->dev);
> + struct pcpu_dstats *dstats = this_cpu_ptr(dev->dstats);
> +
> + u64_stats_update_begin(&dstats->syncp);
> + dstats->rx_pkts++;
> + dstats->rx_bytes += skb->len;
> + u64_stats_update_end(&dstats->syncp);
> +
> + skb->dev = dev;
> +
> + return RX_HANDLER_ANOTHER;
> + }
> + return RX_HANDLER_PASS;
> +}
> +
> +static struct rtnl_link_stats64 *vrf_get_stats64(struct net_device *dev,
> + struct rtnl_link_stats64
> *stats)
> +{
> + int i;
> +
> + for_each_possible_cpu(i) {
> + const struct pcpu_dstats *dstats;
> + u64 tbytes, tpkts, tdrops, rbytes, rpkts;
> + unsigned int start;
> +
> + dstats = per_cpu_ptr(dev->dstats, i);
> + do {
> + start = u64_stats_fetch_begin_irq(&dstats->syncp);
> + tbytes = dstats->tx_bytes;
> + tpkts = dstats->tx_pkts;
> + tdrops = dstats->tx_drps;
> + rbytes = dstats->rx_bytes;
> + rpkts = dstats->rx_pkts;
> + } while (u64_stats_fetch_retry_irq(&dstats->syncp, start));
> + stats->tx_bytes += tbytes;
> + stats->tx_packets += tpkts;
> + stats->tx_dropped += tdrops;
> + stats->rx_bytes += rbytes;
> + stats->rx_packets += rpkts;
> + }
> + return stats;
> +}
> +
> +static int vrf_process_v4_outbound(struct sk_buff *skb)
> +{
> + const struct iphdr *iph = ip_hdr(skb);
> + struct net_device *dev = skb->dev;
> + struct net *net = dev_net(dev);
> + int tb_id = vrf_dev_table(dev);
> + int err, ret = NET_XMIT_DROP;
> + struct in_device *in_dev;
> + struct flowi4 fl4 = {
> + .flowi4_tos = RT_TOS(iph->tos),
> + .daddr = iph->daddr,
> + .saddr = iph->saddr,
> + };
> + struct rtable *rth;
> +
> + struct fib_result res;
> + struct fib_table *tbl;
> +
> + rcu_read_lock();
> + in_dev = __in_dev_get_rcu(dev);
> + if (!in_dev)
> + goto out;
> +
> + if (tb_id == 0)
> + goto out;
> +
> + tbl = fib_get_table(net, tb_id);
> + if (!tbl)
> + goto out;
> +
> + res.tclassid = 0;
> + res.fi = NULL;
> + res.table = NULL;
> + if (fib_table_lookup(tbl, &fl4, &res, 0) != 0)
> + goto out;
> +
> + dev = FIB_RES_DEV(res);
> +
> + rth = ip_route_new_rtable(dev, RTCF_LOCAL, res.type,
> + IN_DEV_CONF_GET(in_dev, NOPOLICY),
> + IN_DEV_CONF_GET(in_dev, NOXFRM),
> + true);
> + if (!rth)
> + goto out;
> +
> + ip_route_set_nexthop(rth, fl4.daddr, &res);
> +
> + skb_dst_drop(skb);
> + skb_dst_set(skb, &rth->dst);
> + err = ip_local_out(skb);
> + if (!net_xmit_eval(err))
> + ret = NET_XMIT_SUCCESS;
> +
> +out:
> + rcu_read_unlock();
> +
> + if (ret != NET_XMIT_SUCCESS)
> + dev->stats.tx_errors++;
> +
> + return ret;
> +}
> +
> +static netdev_tx_t vrf_xmit(struct sk_buff *skb, struct net_device *mdev)
> +{
> + struct pcpu_dstats *dstats = this_cpu_ptr(mdev->dstats);
> + unsigned int len = skb->len;
> + int ret = NET_XMIT_DROP;
> +
> + if (skb_mac_header_was_set(skb)) {
> + skb_pull(skb, sizeof(struct ethhdr));
> + skb->mac_header = (typeof(skb->mac_header))~0U;
> + skb_reset_network_header(skb);
> + }
> +
> + if (skb->protocol == htons(ETH_P_IP))
> + ret = vrf_process_v4_outbound(skb);
> +
> + if (ret == NET_XMIT_SUCCESS) {
> + u64_stats_update_begin(&dstats->syncp);
> + dstats->tx_pkts++;
> + dstats->tx_bytes += len;
> + u64_stats_update_end(&dstats->syncp);
> + } else {
> + dstats->tx_drps++;
> + kfree_skb(skb);
> + }
> + return ret;
> +}
> +
> +/**************************** device handling ********************/
> +
> +/* cycle interface to flush neighbor cache and move routes across tables */
> +static void cycle_netdev(struct net_device *dev)
> +{
> + unsigned int flags = dev->flags;
> + int ret;
> +
> + if (!netif_running(dev))
> + return;
> +
> + ret = dev_change_flags(dev, flags & ~IFF_UP);
> + if (ret >= 0)
> + ret = dev_change_flags(dev, flags);
> +
> + if (ret < 0) {
> + netdev_err(dev,
> + "Failed to cycle device %s; route tables might be
> wrong!\n",
> + dev->name);
> + }
> +}
> +
> +/* queue->lock must be held */
> +static struct slave *__vrf_find_slave_dev(struct slave_queue *queue,
> + struct net_device *dev)
> +{
> + struct list_head *this, *head;
> +
> + head = &queue->all_slaves;
> + list_for_each(this, head) {
^^^^
list_for_each_entry()
> + struct slave *slave = list_entry(this, struct slave, list);
> +
> + if (slave->dev == dev)
> + return slave;
> + }
> +
> + return NULL;
> +}
> +
> +/* inverse of __vrf_insert_slave; queue->lock must be held */
> +static void __vrf_remove_slave(struct slave_queue *queue, struct slave
> *slave)
> +{
> + dev_put(slave->dev);
> + list_del(&slave->list);
> + queue->num_slaves--;
> +}
> +
> +/* queue->lock must be held */
> +static void __vrf_insert_slave(struct slave_queue *queue, struct slave
> *slave)
> +{
> + dev_hold(slave->dev);
> + list_add(&slave->list, &queue->all_slaves);
> + queue->num_slaves++;
> +}
> +
> +static int do_vrf_add_slave(struct net_device *dev, struct net_device
> *port_dev)
> +{
> + struct net_vrf_dev *vrf_ptr = kmalloc(sizeof(*vrf_ptr), GFP_KERNEL);
> + struct slave *slave = kzalloc(sizeof(*slave), GFP_KERNEL);
> + struct slave *duplicate_slave;
> + struct net_vrf *vrf = netdev_priv(dev);
> + struct slave_queue *queue = &vrf->queue;
> + int ret = -ENOMEM;
> +
> + if (!slave || !vrf_ptr)
> + goto out_fail;
> +
> + slave->dev = port_dev;
> +
> + vrf_ptr->ifindex = dev->ifindex;
> + vrf_ptr->tb_id = vrf->tb_id;
> +
> + spin_lock_bh(&queue->lock);
> +
> + duplicate_slave = __vrf_find_slave_dev(queue, port_dev);
> + if (duplicate_slave) {
> + spin_unlock_bh(&queue->lock);
> + ret = -EBUSY;
> + goto out_fail;
> + }
> +
> + __vrf_insert_slave(queue, slave);
> +
> + spin_unlock_bh(&queue->lock);
> +
> + /* register the packet handler for slave ports */
> + ret = netdev_rx_handler_register(port_dev, vrf_handle_frame, dev);
> + if (ret) {
> + netdev_err(port_dev,
> + "Device %s failed to register rx_handler\n",
> + port_dev->name);
> + goto out_remove;
> + }
> +
> + ret = netdev_master_upper_dev_link(port_dev, dev);
> + if (ret < 0)
> + goto out_unregister;
> +
> + port_dev->flags |= IFF_SLAVE;
> +
> + rcu_assign_pointer(port_dev->vrf_ptr, vrf_ptr);
> + cycle_netdev(port_dev);
> +
> + return 0;
> +
> +out_unregister:
> + netdev_rx_handler_unregister(port_dev);
> +out_remove:
> + __vrf_remove_slave(queue, slave);
> +out_fail:
> + kfree(vrf_ptr);
> + kfree(slave);
> + return ret;
> +}
> +
> +static int vrf_add_slave(struct net_device *dev, struct net_device *port_dev)
> +{
> + ASSERT_RTNL();
These asserts are not needed, the ndo_slave_add/del always run with rtnl and
many drivers rely on that.
> +
> + if (!dev || !port_dev || dev_net(dev) != dev_net(port_dev))
> + return -ENODEV;
I don't think the !dev or !port_dev can happen. Also the net can't be different
looking at do_set_master() in rtnetlink.c
> +
> + if (!vrf_is_master(dev) || vrf_is_master(port_dev) ||
Hmm, this means that bonds won't be able to be VRF slaves.
They have the IFF_MASTER flag set.
> + vrf_is_slave(port_dev))
> + return -EINVAL;
> +
> + return do_vrf_add_slave(dev, port_dev);
> +}
> +
> +/* inverse of do_vrf_add_slave */
> +static int do_vrf_del_slave(struct net_device *dev, struct net_device
> *port_dev)
> +{
> + struct net_vrf *vrf = netdev_priv(dev);
> + struct slave_queue *queue = &vrf->queue;
> + struct net_vrf_dev *vrf_ptr = NULL;
> + struct slave *slave;
> +
> + vrf_ptr = rcu_dereference(dev->vrf_ptr);
> + RCU_INIT_POINTER(dev->vrf_ptr, NULL);
I think this isn't safe, you should wait for a grace period before freeing the
pointer. Actually you can just move the kfree() below the
netdev_rx_handler_unregister()
since it does synchronize_rcu() anyway.
> + kfree(vrf_ptr);
> +
> + netdev_upper_dev_unlink(port_dev, dev);
> + port_dev->flags &= ~IFF_SLAVE;
> +
> + netdev_rx_handler_unregister(dev);
> +
> + cycle_netdev(port_dev);
> +
> + spin_lock_bh(&queue->lock);
> +
> + slave = __vrf_find_slave_dev(queue, port_dev);
> + if (slave)
> + __vrf_remove_slave(queue, slave);
> +
> + spin_unlock_bh(&queue->lock);
> +
> + kfree(slave);
> +
> + return 0;
> +}
> +
> +static int vrf_del_slave(struct net_device *dev, struct net_device *port_dev)
> +{
> + ASSERT_RTNL();
> +
> + if (!dev || !port_dev)
> + return -ENODEV;
^
I really don't think any of this can happen.
> +
> + if (!vrf_is_master(dev))
> + return -EINVAL;
> +
> + return do_vrf_del_slave(dev, port_dev);
> +}
> +
> +static int vrf_dev_init(struct net_device *dev)
> +{
> + struct net_vrf *vrf = netdev_priv(dev);
> +
> + spin_lock_init(&vrf->queue.lock);
> + INIT_LIST_HEAD(&vrf->queue.all_slaves);
> +
> + dev->dstats = netdev_alloc_pcpu_stats(struct pcpu_dstats);
> + if (!dev->dstats)
> + return -ENOMEM;
> +
> + dev->flags = IFF_MASTER | IFF_NOARP;
> +
> + return 0;
> +}
> +
> +static void vrf_dev_uninit(struct net_device *dev)
> +{
> + free_percpu(dev->dstats);
> +}
> +
> +static int vrf_dev_close(struct net_device *dev)
> +{
> + struct net_vrf *vrf = netdev_priv(dev);
> + struct slave_queue *queue = &vrf->queue;
> + struct list_head *this, *head;
> +
> + head = &queue->all_slaves;
> + list_for_each(this, head) {
^^^
list_for_each_entry()
> + struct slave *slave = list_entry(this, struct slave, list);
> +
> + slave->dev->vrf_ptr->ifindex = 0;
> + slave->dev->vrf_ptr->tb_id = 0;
> + }
> +
> + if (dev->flags & IFF_MASTER)
> + dev->flags &= ~IFF_UP;
> +
> + return 0;
> +}
> +
> +static int vrf_dev_open(struct net_device *dev)
> +{
> + struct net_vrf *vrf = netdev_priv(dev);
> + struct slave_queue *queue = &vrf->queue;
> + struct list_head *this, *head;
> +
> + head = &queue->all_slaves;
> + list_for_each(this, head) {
^^^
list_for_each_entry()
> + struct slave *slave = list_entry(this, struct slave, list);
> +
> + slave->dev->vrf_ptr->ifindex = dev->ifindex;
> + slave->dev->vrf_ptr->tb_id = vrf->tb_id;
> + }
> +
> + if (dev->flags & IFF_MASTER)
> + dev->flags |= IFF_UP;
> +
> + if (!vrf->tb)
> + return -EINVAL;
> +
> + return 0;
> +}
> +
> +static const struct net_device_ops vrf_netdev_ops = {
> + .ndo_init = vrf_dev_init,
> + .ndo_uninit = vrf_dev_uninit,
> + .ndo_open = vrf_dev_open,
> + .ndo_stop = vrf_dev_close,
> + .ndo_start_xmit = vrf_xmit,
> + .ndo_get_stats64 = vrf_get_stats64,
> + .ndo_add_slave = vrf_add_slave,
> + .ndo_del_slave = vrf_del_slave,
> +};
> +
> +static void vrf_get_drvinfo(struct net_device *dev,
> + struct ethtool_drvinfo *info)
> +{
> + strlcpy(info->driver, DRV_NAME, sizeof(info->driver));
> + strlcpy(info->version, DRV_VERSION, sizeof(info->version));
> +}
> +
> +static const struct ethtool_ops vrf_ethtool_ops = {
> + .get_drvinfo = vrf_get_drvinfo,
> +};
> +
> +static void vrf_setup(struct net_device *dev)
> +{
> + ether_setup(dev);
> +
> + /* Initialize the device structure. */
> + dev->netdev_ops = &vrf_netdev_ops;
> + dev->ethtool_ops = &vrf_ethtool_ops;
> + dev->destructor = free_netdev;
> +
> + /* Fill in device structure with ethernet-generic values. */
> + eth_hw_addr_random(dev);
> +}
> +
> +static int vrf_validate(struct nlattr *tb[], struct nlattr *data[])
> +{
> + if (tb[IFLA_ADDRESS]) {
> + if (nla_len(tb[IFLA_ADDRESS]) != ETH_ALEN)
> + return -EINVAL;
> + if (!is_valid_ether_addr(nla_data(tb[IFLA_ADDRESS])))
> + return -EADDRNOTAVAIL;
> + }
> + return 0;
> +}
> +
> +static int vrf_newlink(struct net *src_net, struct net_device *dev,
> + struct nlattr *tb[], struct nlattr *data[])
> +{
> + struct net_vrf *vrf = netdev_priv(dev);
> + int err;
> +
> + if (!data || !data[IFLA_VRF_TABLE])
> + return -EINVAL;
> +
> + vrf->tb_id = nla_get_u32(data[IFLA_VRF_TABLE]);
> +
> + /* reserve a table for this VRF device */
> + err = -ERANGE;
> + vrf->tb = fib_new_table(dev_net(dev), vrf->tb_id);
> + if (!vrf->tb)
> + goto out_fail;
> +
> + dev->priv_flags |= IFF_VRF_MASTER;
> +
> + err = -ENOMEM;
> + dev->vrf_ptr = kmalloc(sizeof(*dev->vrf_ptr), GFP_KERNEL);
> + if (!dev->vrf_ptr)
> + goto out_fail;
> +
> + dev->vrf_ptr->ifindex = dev->ifindex;
> + dev->vrf_ptr->tb_id = vrf->tb_id;
> +
> + err = register_netdevice(dev);
> + if (err < 0)
> + goto out_fail;
> +
> + return 0;
> +
> +out_fail:
> + kfree(dev->vrf_ptr);
> + free_netdev(dev);
> + return err;
> +}
> +
> +static void vrf_dellink(struct net_device *dev, struct list_head *head)
> +{
> + struct net_vrf *vrf = netdev_priv(dev);
> +
> + kfree(dev->vrf_ptr);
> + fib_free_table(vrf->tb);
> +}
> +
> +static size_t vrf_nl_getsize(const struct net_device *dev)
> +{
> + return nla_total_size(sizeof(u32)); /* IFLA_VRF_TABLE */
> +}
> +
> +static int vrf_fillinfo(struct sk_buff *skb,
> + const struct net_device *dev)
> +{
> + struct net_vrf *vrf = netdev_priv(dev);
> +
> + return nla_put_u32(skb, IFLA_VRF_TABLE, vrf->tb_id);
> +}
> +
> +static const struct nla_policy vrf_nl_policy[IFLA_VRF_MAX + 1] = {
> + [IFLA_VRF_TABLE] = { .type = NLA_U32 },
> +};
> +
> +static struct rtnl_link_ops vrf_link_ops __read_mostly = {
> + .kind = DRV_NAME,
> + .priv_size = sizeof(struct net_vrf),
> +
> + .get_size = vrf_nl_getsize,
> + .policy = vrf_nl_policy,
> + .validate = vrf_validate,
> + .fill_info = vrf_fillinfo,
> +
> + .newlink = vrf_newlink,
> + .dellink = vrf_dellink,
> + .setup = vrf_setup,
> + .maxtype = IFLA_VRF_MAX,
> +};
> +
> +static int __init vrf_init_module(void)
> +{
> + return rtnl_link_register(&vrf_link_ops);
> +}
> +
> +static void __exit vrf_cleanup_module(void)
> +{
> + rtnl_link_unregister(&vrf_link_ops);
> +}
> +
> +module_init(vrf_init_module);
> +module_exit(vrf_cleanup_module);
> +MODULE_AUTHOR("Shrijeet Mukherjee, David Ahern");
> +MODULE_LICENSE("GPL");
> +MODULE_ALIAS_RTNL_LINK(DRV_NAME);
> +MODULE_VERSION(DRV_VERSION);
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
