* [PATCHv1 1/7] TAP: Refactoring macvtap.c
2017-01-06 22:33 [PATCHv1 0/7] Refactor macvtap to re-use tap functionality by other virtual intefaces Sainath Grandhi
@ 2017-01-06 22:33 ` Sainath Grandhi
2017-01-06 23:12 ` Andy Shevchenko
2017-01-06 22:33 ` [PATCHv1 2/7] TAP: Renaming tap related APIs, data structures, macros Sainath Grandhi
` (5 subsequent siblings)
6 siblings, 1 reply; 21+ messages in thread
From: Sainath Grandhi @ 2017-01-06 22:33 UTC (permalink / raw)
To: netdev; +Cc: davem, mahesh, linux-kernel, Sainath Grandhi
macvtap module has code for tap/queue management and link management. This patch splits
the code into macvtap_main.c for link management and tap.c for tap/queue management.
Functionality in tap.c can be re-used for implementing tap on other virtual interfaces.
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
---
drivers/net/Makefile | 2 +
drivers/net/macvtap.c | 1374 --------------------------------------------
drivers/net/macvtap_main.c | 218 +++++++
drivers/net/tap.c | 1186 ++++++++++++++++++++++++++++++++++++++
include/linux/if_macvtap.h | 10 +
5 files changed, 1416 insertions(+), 1374 deletions(-)
delete mode 100644 drivers/net/macvtap.c
create mode 100644 drivers/net/macvtap_main.c
create mode 100644 drivers/net/tap.c
create mode 100644 include/linux/if_macvtap.h
diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index 7336cbd..19b03a9 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -29,6 +29,8 @@ obj-$(CONFIG_GTP) += gtp.o
obj-$(CONFIG_NLMON) += nlmon.o
obj-$(CONFIG_NET_VRF) += vrf.o
+macvtap-objs := macvtap_main.o tap.o
+
#
# Networking Drivers
#
diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
deleted file mode 100644
index 5c26653..0000000
--- a/drivers/net/macvtap.c
+++ /dev/null
@@ -1,1374 +0,0 @@
-#include <linux/etherdevice.h>
-#include <linux/if_macvlan.h>
-#include <linux/if_vlan.h>
-#include <linux/interrupt.h>
-#include <linux/nsproxy.h>
-#include <linux/compat.h>
-#include <linux/if_tun.h>
-#include <linux/module.h>
-#include <linux/skbuff.h>
-#include <linux/cache.h>
-#include <linux/sched.h>
-#include <linux/types.h>
-#include <linux/slab.h>
-#include <linux/wait.h>
-#include <linux/cdev.h>
-#include <linux/idr.h>
-#include <linux/fs.h>
-#include <linux/uio.h>
-
-#include <net/net_namespace.h>
-#include <net/rtnetlink.h>
-#include <net/sock.h>
-#include <linux/virtio_net.h>
-#include <linux/skb_array.h>
-
-/*
- * A macvtap queue is the central object of this driver, it connects
- * an open character device to a macvlan interface. There can be
- * multiple queues on one interface, which map back to queues
- * implemented in hardware on the underlying device.
- *
- * macvtap_proto is used to allocate queues through the sock allocation
- * mechanism.
- *
- */
-struct macvtap_queue {
- struct sock sk;
- struct socket sock;
- struct socket_wq wq;
- int vnet_hdr_sz;
- struct macvlan_dev __rcu *vlan;
- struct file *file;
- unsigned int flags;
- u16 queue_index;
- bool enabled;
- struct list_head next;
- struct skb_array skb_array;
-};
-
-#define MACVTAP_FEATURES (IFF_VNET_HDR | IFF_MULTI_QUEUE)
-
-#define MACVTAP_VNET_LE 0x80000000
-#define MACVTAP_VNET_BE 0x40000000
-
-#ifdef CONFIG_TUN_VNET_CROSS_LE
-static inline bool macvtap_legacy_is_little_endian(struct macvtap_queue *q)
-{
- return q->flags & MACVTAP_VNET_BE ? false :
- virtio_legacy_is_little_endian();
-}
-
-static long macvtap_get_vnet_be(struct macvtap_queue *q, int __user *sp)
-{
- int s = !!(q->flags & MACVTAP_VNET_BE);
-
- if (put_user(s, sp))
- return -EFAULT;
-
- return 0;
-}
-
-static long macvtap_set_vnet_be(struct macvtap_queue *q, int __user *sp)
-{
- int s;
-
- if (get_user(s, sp))
- return -EFAULT;
-
- if (s)
- q->flags |= MACVTAP_VNET_BE;
- else
- q->flags &= ~MACVTAP_VNET_BE;
-
- return 0;
-}
-#else
-static inline bool macvtap_legacy_is_little_endian(struct macvtap_queue *q)
-{
- return virtio_legacy_is_little_endian();
-}
-
-static long macvtap_get_vnet_be(struct macvtap_queue *q, int __user *argp)
-{
- return -EINVAL;
-}
-
-static long macvtap_set_vnet_be(struct macvtap_queue *q, int __user *argp)
-{
- return -EINVAL;
-}
-#endif /* CONFIG_TUN_VNET_CROSS_LE */
-
-static inline bool macvtap_is_little_endian(struct macvtap_queue *q)
-{
- return q->flags & MACVTAP_VNET_LE ||
- macvtap_legacy_is_little_endian(q);
-}
-
-static inline u16 macvtap16_to_cpu(struct macvtap_queue *q, __virtio16 val)
-{
- return __virtio16_to_cpu(macvtap_is_little_endian(q), val);
-}
-
-static inline __virtio16 cpu_to_macvtap16(struct macvtap_queue *q, u16 val)
-{
- return __cpu_to_virtio16(macvtap_is_little_endian(q), val);
-}
-
-static struct proto macvtap_proto = {
- .name = "macvtap",
- .owner = THIS_MODULE,
- .obj_size = sizeof (struct macvtap_queue),
-};
-
-/*
- * Variables for dealing with macvtaps device numbers.
- */
-static dev_t macvtap_major;
-#define MACVTAP_NUM_DEVS (1U << MINORBITS)
-static DEFINE_MUTEX(minor_lock);
-static DEFINE_IDR(minor_idr);
-
-#define GOODCOPY_LEN 128
-static const void *macvtap_net_namespace(struct device *d)
-{
- struct net_device *dev = to_net_dev(d->parent);
- return dev_net(dev);
-}
-
-static struct class macvtap_class = {
- .name = "macvtap",
- .owner = THIS_MODULE,
- .ns_type = &net_ns_type_operations,
- .namespace = macvtap_net_namespace,
-};
-static struct cdev macvtap_cdev;
-
-static const struct proto_ops macvtap_socket_ops;
-
-#define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN | NETIF_F_TSO | \
- NETIF_F_TSO6 | NETIF_F_UFO)
-#define RX_OFFLOADS (NETIF_F_GRO | NETIF_F_LRO)
-#define TAP_FEATURES (NETIF_F_GSO | NETIF_F_SG | NETIF_F_FRAGLIST)
-
-static struct macvlan_dev *macvtap_get_vlan_rcu(const struct net_device *dev)
-{
- return rcu_dereference(dev->rx_handler_data);
-}
-
-/*
- * RCU usage:
- * The macvtap_queue and the macvlan_dev are loosely coupled, the
- * pointers from one to the other can only be read while rcu_read_lock
- * or rtnl is held.
- *
- * Both the file and the macvlan_dev hold a reference on the macvtap_queue
- * through sock_hold(&q->sk). When the macvlan_dev goes away first,
- * q->vlan becomes inaccessible. When the files gets closed,
- * macvtap_get_queue() fails.
- *
- * There may still be references to the struct sock inside of the
- * queue from outbound SKBs, but these never reference back to the
- * file or the dev. The data structure is freed through __sk_free
- * when both our references and any pending SKBs are gone.
- */
-
-static int macvtap_enable_queue(struct net_device *dev, struct file *file,
- struct macvtap_queue *q)
-{
- struct macvlan_dev *vlan = netdev_priv(dev);
- int err = -EINVAL;
-
- ASSERT_RTNL();
-
- if (q->enabled)
- goto out;
-
- err = 0;
- rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
- q->queue_index = vlan->numvtaps;
- q->enabled = true;
-
- vlan->numvtaps++;
-out:
- return err;
-}
-
-/* Requires RTNL */
-static int macvtap_set_queue(struct net_device *dev, struct file *file,
- struct macvtap_queue *q)
-{
- struct macvlan_dev *vlan = netdev_priv(dev);
-
- if (vlan->numqueues == MAX_MACVTAP_QUEUES)
- return -EBUSY;
-
- rcu_assign_pointer(q->vlan, vlan);
- rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
- sock_hold(&q->sk);
-
- q->file = file;
- q->queue_index = vlan->numvtaps;
- q->enabled = true;
- file->private_data = q;
- list_add_tail(&q->next, &vlan->queue_list);
-
- vlan->numvtaps++;
- vlan->numqueues++;
-
- return 0;
-}
-
-static int macvtap_disable_queue(struct macvtap_queue *q)
-{
- struct macvlan_dev *vlan;
- struct macvtap_queue *nq;
-
- ASSERT_RTNL();
- if (!q->enabled)
- return -EINVAL;
-
- vlan = rtnl_dereference(q->vlan);
-
- if (vlan) {
- int index = q->queue_index;
- BUG_ON(index >= vlan->numvtaps);
- nq = rtnl_dereference(vlan->taps[vlan->numvtaps - 1]);
- nq->queue_index = index;
-
- rcu_assign_pointer(vlan->taps[index], nq);
- RCU_INIT_POINTER(vlan->taps[vlan->numvtaps - 1], NULL);
- q->enabled = false;
-
- vlan->numvtaps--;
- }
-
- return 0;
-}
-
-/*
- * The file owning the queue got closed, give up both
- * the reference that the files holds as well as the
- * one from the macvlan_dev if that still exists.
- *
- * Using the spinlock makes sure that we don't get
- * to the queue again after destroying it.
- */
-static void macvtap_put_queue(struct macvtap_queue *q)
-{
- struct macvlan_dev *vlan;
-
- rtnl_lock();
- vlan = rtnl_dereference(q->vlan);
-
- if (vlan) {
- if (q->enabled)
- BUG_ON(macvtap_disable_queue(q));
-
- vlan->numqueues--;
- RCU_INIT_POINTER(q->vlan, NULL);
- sock_put(&q->sk);
- list_del_init(&q->next);
- }
-
- rtnl_unlock();
-
- synchronize_rcu();
- sock_put(&q->sk);
-}
-
-/*
- * Select a queue based on the rxq of the device on which this packet
- * arrived. If the incoming device is not mq, calculate a flow hash
- * to select a queue. If all fails, find the first available queue.
- * Cache vlan->numvtaps since it can become zero during the execution
- * of this function.
- */
-static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
- struct sk_buff *skb)
-{
- struct macvlan_dev *vlan = netdev_priv(dev);
- struct macvtap_queue *tap = NULL;
- /* Access to taps array is protected by rcu, but access to numvtaps
- * isn't. Below we use it to lookup a queue, but treat it as a hint
- * and validate that the result isn't NULL - in case we are
- * racing against queue removal.
- */
- int numvtaps = ACCESS_ONCE(vlan->numvtaps);
- __u32 rxq;
-
- if (!numvtaps)
- goto out;
-
- if (numvtaps == 1)
- goto single;
-
- /* Check if we can use flow to select a queue */
- rxq = skb_get_hash(skb);
- if (rxq) {
- tap = rcu_dereference(vlan->taps[rxq % numvtaps]);
- goto out;
- }
-
- if (likely(skb_rx_queue_recorded(skb))) {
- rxq = skb_get_rx_queue(skb);
-
- while (unlikely(rxq >= numvtaps))
- rxq -= numvtaps;
-
- tap = rcu_dereference(vlan->taps[rxq]);
- goto out;
- }
-
-single:
- tap = rcu_dereference(vlan->taps[0]);
-out:
- return tap;
-}
-
-/*
- * The net_device is going away, give up the reference
- * that it holds on all queues and safely set the pointer
- * from the queues to NULL.
- */
-static void macvtap_del_queues(struct net_device *dev)
-{
- struct macvlan_dev *vlan = netdev_priv(dev);
- struct macvtap_queue *q, *tmp;
-
- ASSERT_RTNL();
- list_for_each_entry_safe(q, tmp, &vlan->queue_list, next) {
- list_del_init(&q->next);
- RCU_INIT_POINTER(q->vlan, NULL);
- if (q->enabled)
- vlan->numvtaps--;
- vlan->numqueues--;
- sock_put(&q->sk);
- }
- BUG_ON(vlan->numvtaps);
- BUG_ON(vlan->numqueues);
- /* guarantee that any future macvtap_set_queue will fail */
- vlan->numvtaps = MAX_MACVTAP_QUEUES;
-}
-
-static rx_handler_result_t macvtap_handle_frame(struct sk_buff **pskb)
-{
- struct sk_buff *skb = *pskb;
- struct net_device *dev = skb->dev;
- struct macvlan_dev *vlan;
- struct macvtap_queue *q;
- netdev_features_t features = TAP_FEATURES;
-
- vlan = macvtap_get_vlan_rcu(dev);
- if (!vlan)
- return RX_HANDLER_PASS;
-
- q = macvtap_get_queue(dev, skb);
- if (!q)
- return RX_HANDLER_PASS;
-
- if (__skb_array_full(&q->skb_array))
- goto drop;
-
- skb_push(skb, ETH_HLEN);
-
- /* Apply the forward feature mask so that we perform segmentation
- * according to users wishes. This only works if VNET_HDR is
- * enabled.
- */
- if (q->flags & IFF_VNET_HDR)
- features |= vlan->tap_features;
- if (netif_needs_gso(skb, features)) {
- struct sk_buff *segs = __skb_gso_segment(skb, features, false);
-
- if (IS_ERR(segs))
- goto drop;
-
- if (!segs) {
- if (skb_array_produce(&q->skb_array, skb))
- goto drop;
- goto wake_up;
- }
-
- consume_skb(skb);
- while (segs) {
- struct sk_buff *nskb = segs->next;
-
- segs->next = NULL;
- if (skb_array_produce(&q->skb_array, segs)) {
- kfree_skb(segs);
- kfree_skb_list(nskb);
- break;
- }
- segs = nskb;
- }
- } else {
- /* If we receive a partial checksum and the tap side
- * doesn't support checksum offload, compute the checksum.
- * Note: it doesn't matter which checksum feature to
- * check, we either support them all or none.
- */
- if (skb->ip_summed == CHECKSUM_PARTIAL &&
- !(features & NETIF_F_CSUM_MASK) &&
- skb_checksum_help(skb))
- goto drop;
- if (skb_array_produce(&q->skb_array, skb))
- goto drop;
- }
-
-wake_up:
- wake_up_interruptible_poll(sk_sleep(&q->sk), POLLIN | POLLRDNORM | POLLRDBAND);
- return RX_HANDLER_CONSUMED;
-
-drop:
- /* Count errors/drops only here, thus don't care about args. */
- macvlan_count_rx(vlan, 0, 0, 0);
- kfree_skb(skb);
- return RX_HANDLER_CONSUMED;
-}
-
-static int macvtap_get_minor(struct macvlan_dev *vlan)
-{
- int retval = -ENOMEM;
-
- mutex_lock(&minor_lock);
- retval = idr_alloc(&minor_idr, vlan, 1, MACVTAP_NUM_DEVS, GFP_KERNEL);
- if (retval >= 0) {
- vlan->minor = retval;
- } else if (retval == -ENOSPC) {
- netdev_err(vlan->dev, "Too many macvtap devices\n");
- retval = -EINVAL;
- }
- mutex_unlock(&minor_lock);
- return retval < 0 ? retval : 0;
-}
-
-static void macvtap_free_minor(struct macvlan_dev *vlan)
-{
- mutex_lock(&minor_lock);
- if (vlan->minor) {
- idr_remove(&minor_idr, vlan->minor);
- vlan->minor = 0;
- }
- mutex_unlock(&minor_lock);
-}
-
-static struct net_device *dev_get_by_macvtap_minor(int minor)
-{
- struct net_device *dev = NULL;
- struct macvlan_dev *vlan;
-
- mutex_lock(&minor_lock);
- vlan = idr_find(&minor_idr, minor);
- if (vlan) {
- dev = vlan->dev;
- dev_hold(dev);
- }
- mutex_unlock(&minor_lock);
- return dev;
-}
-
-static int macvtap_newlink(struct net *src_net,
- struct net_device *dev,
- struct nlattr *tb[],
- struct nlattr *data[])
-{
- struct macvlan_dev *vlan = netdev_priv(dev);
- int err;
-
- INIT_LIST_HEAD(&vlan->queue_list);
-
- /* Since macvlan supports all offloads by default, make
- * tap support all offloads also.
- */
- vlan->tap_features = TUN_OFFLOADS;
-
- err = netdev_rx_handler_register(dev, macvtap_handle_frame, vlan);
- if (err)
- return err;
-
- /* Don't put anything that may fail after macvlan_common_newlink
- * because we can't undo what it does.
- */
- err = macvlan_common_newlink(src_net, dev, tb, data);
- if (err) {
- netdev_rx_handler_unregister(dev);
- return err;
- }
-
- return 0;
-}
-
-static void macvtap_dellink(struct net_device *dev,
- struct list_head *head)
-{
- netdev_rx_handler_unregister(dev);
- macvtap_del_queues(dev);
- macvlan_dellink(dev, head);
-}
-
-static void macvtap_setup(struct net_device *dev)
-{
- macvlan_common_setup(dev);
- dev->tx_queue_len = TUN_READQ_SIZE;
-}
-
-static struct rtnl_link_ops macvtap_link_ops __read_mostly = {
- .kind = "macvtap",
- .setup = macvtap_setup,
- .newlink = macvtap_newlink,
- .dellink = macvtap_dellink,
-};
-
-
-static void macvtap_sock_write_space(struct sock *sk)
-{
- wait_queue_head_t *wqueue;
-
- if (!sock_writeable(sk) ||
- !test_and_clear_bit(SOCKWQ_ASYNC_NOSPACE, &sk->sk_socket->flags))
- return;
-
- wqueue = sk_sleep(sk);
- if (wqueue && waitqueue_active(wqueue))
- wake_up_interruptible_poll(wqueue, POLLOUT | POLLWRNORM | POLLWRBAND);
-}
-
-static void macvtap_sock_destruct(struct sock *sk)
-{
- struct macvtap_queue *q = container_of(sk, struct macvtap_queue, sk);
-
- skb_array_cleanup(&q->skb_array);
-}
-
-static int macvtap_open(struct inode *inode, struct file *file)
-{
- struct net *net = current->nsproxy->net_ns;
- struct net_device *dev;
- struct macvtap_queue *q;
- int err = -ENODEV;
-
- rtnl_lock();
- dev = dev_get_by_macvtap_minor(iminor(inode));
- if (!dev)
- goto err;
-
- err = -ENOMEM;
- q = (struct macvtap_queue *)sk_alloc(net, AF_UNSPEC, GFP_KERNEL,
- &macvtap_proto, 0);
- if (!q)
- goto err;
-
- RCU_INIT_POINTER(q->sock.wq, &q->wq);
- init_waitqueue_head(&q->wq.wait);
- q->sock.type = SOCK_RAW;
- q->sock.state = SS_CONNECTED;
- q->sock.file = file;
- q->sock.ops = &macvtap_socket_ops;
- sock_init_data(&q->sock, &q->sk);
- q->sk.sk_write_space = macvtap_sock_write_space;
- q->sk.sk_destruct = macvtap_sock_destruct;
- q->flags = IFF_VNET_HDR | IFF_NO_PI | IFF_TAP;
- q->vnet_hdr_sz = sizeof(struct virtio_net_hdr);
-
- /*
- * so far only KVM virtio_net uses macvtap, enable zero copy between
- * guest kernel and host kernel when lower device supports zerocopy
- *
- * The macvlan supports zerocopy iff the lower device supports zero
- * copy so we don't have to look at the lower device directly.
- */
- if ((dev->features & NETIF_F_HIGHDMA) && (dev->features & NETIF_F_SG))
- sock_set_flag(&q->sk, SOCK_ZEROCOPY);
-
- err = -ENOMEM;
- if (skb_array_init(&q->skb_array, dev->tx_queue_len, GFP_KERNEL))
- goto err_array;
-
- err = macvtap_set_queue(dev, file, q);
- if (err)
- goto err_queue;
-
- dev_put(dev);
-
- rtnl_unlock();
- return err;
-
-err_queue:
- skb_array_cleanup(&q->skb_array);
-err_array:
- sock_put(&q->sk);
-err:
- if (dev)
- dev_put(dev);
-
- rtnl_unlock();
- return err;
-}
-
-static int macvtap_release(struct inode *inode, struct file *file)
-{
- struct macvtap_queue *q = file->private_data;
- macvtap_put_queue(q);
- return 0;
-}
-
-static unsigned int macvtap_poll(struct file *file, poll_table * wait)
-{
- struct macvtap_queue *q = file->private_data;
- unsigned int mask = POLLERR;
-
- if (!q)
- goto out;
-
- mask = 0;
- poll_wait(file, &q->wq.wait, wait);
-
- if (!skb_array_empty(&q->skb_array))
- mask |= POLLIN | POLLRDNORM;
-
- if (sock_writeable(&q->sk) ||
- (!test_and_set_bit(SOCKWQ_ASYNC_NOSPACE, &q->sock.flags) &&
- sock_writeable(&q->sk)))
- mask |= POLLOUT | POLLWRNORM;
-
-out:
- return mask;
-}
-
-static inline struct sk_buff *macvtap_alloc_skb(struct sock *sk, size_t prepad,
- size_t len, size_t linear,
- int noblock, int *err)
-{
- struct sk_buff *skb;
-
- /* Under a page? Don't bother with paged skb. */
- if (prepad + len < PAGE_SIZE || !linear)
- linear = len;
-
- skb = sock_alloc_send_pskb(sk, prepad + linear, len - linear, noblock,
- err, 0);
- if (!skb)
- return NULL;
-
- skb_reserve(skb, prepad);
- skb_put(skb, linear);
- skb->data_len = len - linear;
- skb->len += len - linear;
-
- return skb;
-}
-
-/* Neighbour code has some assumptions on HH_DATA_MOD alignment */
-#define MACVTAP_RESERVE HH_DATA_OFF(ETH_HLEN)
-
-/* Get packet from user space buffer */
-static ssize_t macvtap_get_user(struct macvtap_queue *q, struct msghdr *m,
- struct iov_iter *from, int noblock)
-{
- int good_linear = SKB_MAX_HEAD(MACVTAP_RESERVE);
- struct sk_buff *skb;
- struct macvlan_dev *vlan;
- unsigned long total_len = iov_iter_count(from);
- unsigned long len = total_len;
- int err;
- struct virtio_net_hdr vnet_hdr = { 0 };
- int vnet_hdr_len = 0;
- int copylen = 0;
- int depth;
- bool zerocopy = false;
- size_t linear;
-
- if (q->flags & IFF_VNET_HDR) {
- vnet_hdr_len = q->vnet_hdr_sz;
-
- err = -EINVAL;
- if (len < vnet_hdr_len)
- goto err;
- len -= vnet_hdr_len;
-
- err = -EFAULT;
- if (!copy_from_iter_full(&vnet_hdr, sizeof(vnet_hdr), from))
- goto err;
- iov_iter_advance(from, vnet_hdr_len - sizeof(vnet_hdr));
- if ((vnet_hdr.flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) &&
- macvtap16_to_cpu(q, vnet_hdr.csum_start) +
- macvtap16_to_cpu(q, vnet_hdr.csum_offset) + 2 >
- macvtap16_to_cpu(q, vnet_hdr.hdr_len))
- vnet_hdr.hdr_len = cpu_to_macvtap16(q,
- macvtap16_to_cpu(q, vnet_hdr.csum_start) +
- macvtap16_to_cpu(q, vnet_hdr.csum_offset) + 2);
- err = -EINVAL;
- if (macvtap16_to_cpu(q, vnet_hdr.hdr_len) > len)
- goto err;
- }
-
- err = -EINVAL;
- if (unlikely(len < ETH_HLEN))
- goto err;
-
- if (m && m->msg_control && sock_flag(&q->sk, SOCK_ZEROCOPY)) {
- struct iov_iter i;
-
- copylen = vnet_hdr.hdr_len ?
- macvtap16_to_cpu(q, vnet_hdr.hdr_len) : GOODCOPY_LEN;
- if (copylen > good_linear)
- copylen = good_linear;
- else if (copylen < ETH_HLEN)
- copylen = ETH_HLEN;
- linear = copylen;
- i = *from;
- iov_iter_advance(&i, copylen);
- if (iov_iter_npages(&i, INT_MAX) <= MAX_SKB_FRAGS)
- zerocopy = true;
- }
-
- if (!zerocopy) {
- copylen = len;
- linear = macvtap16_to_cpu(q, vnet_hdr.hdr_len);
- if (linear > good_linear)
- linear = good_linear;
- else if (linear < ETH_HLEN)
- linear = ETH_HLEN;
- }
-
- skb = macvtap_alloc_skb(&q->sk, MACVTAP_RESERVE, copylen,
- linear, noblock, &err);
- if (!skb)
- goto err;
-
- if (zerocopy)
- err = zerocopy_sg_from_iter(skb, from);
- else
- err = skb_copy_datagram_from_iter(skb, 0, from, len);
-
- if (err)
- goto err_kfree;
-
- skb_set_network_header(skb, ETH_HLEN);
- skb_reset_mac_header(skb);
- skb->protocol = eth_hdr(skb)->h_proto;
-
- if (vnet_hdr_len) {
- err = virtio_net_hdr_to_skb(skb, &vnet_hdr,
- macvtap_is_little_endian(q));
- if (err)
- goto err_kfree;
- }
-
- skb_probe_transport_header(skb, ETH_HLEN);
-
- /* Move network header to the right position for VLAN tagged packets */
- if ((skb->protocol == htons(ETH_P_8021Q) ||
- skb->protocol == htons(ETH_P_8021AD)) &&
- __vlan_get_protocol(skb, skb->protocol, &depth) != 0)
- skb_set_network_header(skb, depth);
-
- rcu_read_lock();
- vlan = rcu_dereference(q->vlan);
- /* copy skb_ubuf_info for callback when skb has no error */
- if (zerocopy) {
- skb_shinfo(skb)->destructor_arg = m->msg_control;
- skb_shinfo(skb)->tx_flags |= SKBTX_DEV_ZEROCOPY;
- skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG;
- } else if (m && m->msg_control) {
- struct ubuf_info *uarg = m->msg_control;
- uarg->callback(uarg, false);
- }
-
- if (vlan) {
- skb->dev = vlan->dev;
- dev_queue_xmit(skb);
- } else {
- kfree_skb(skb);
- }
- rcu_read_unlock();
-
- return total_len;
-
-err_kfree:
- kfree_skb(skb);
-
-err:
- rcu_read_lock();
- vlan = rcu_dereference(q->vlan);
- if (vlan)
- this_cpu_inc(vlan->pcpu_stats->tx_dropped);
- rcu_read_unlock();
-
- return err;
-}
-
-static ssize_t macvtap_write_iter(struct kiocb *iocb, struct iov_iter *from)
-{
- struct file *file = iocb->ki_filp;
- struct macvtap_queue *q = file->private_data;
-
- return macvtap_get_user(q, NULL, from, file->f_flags & O_NONBLOCK);
-}
-
-/* Put packet to the user space buffer */
-static ssize_t macvtap_put_user(struct macvtap_queue *q,
- const struct sk_buff *skb,
- struct iov_iter *iter)
-{
- int ret;
- int vnet_hdr_len = 0;
- int vlan_offset = 0;
- int total;
-
- if (q->flags & IFF_VNET_HDR) {
- struct virtio_net_hdr vnet_hdr;
- vnet_hdr_len = q->vnet_hdr_sz;
- if (iov_iter_count(iter) < vnet_hdr_len)
- return -EINVAL;
-
- if (virtio_net_hdr_from_skb(skb, &vnet_hdr,
- macvtap_is_little_endian(q)))
- BUG();
-
- if (copy_to_iter(&vnet_hdr, sizeof(vnet_hdr), iter) !=
- sizeof(vnet_hdr))
- return -EFAULT;
-
- iov_iter_advance(iter, vnet_hdr_len - sizeof(vnet_hdr));
- }
- total = vnet_hdr_len;
- total += skb->len;
-
- if (skb_vlan_tag_present(skb)) {
- struct {
- __be16 h_vlan_proto;
- __be16 h_vlan_TCI;
- } veth;
- veth.h_vlan_proto = skb->vlan_proto;
- veth.h_vlan_TCI = htons(skb_vlan_tag_get(skb));
-
- vlan_offset = offsetof(struct vlan_ethhdr, h_vlan_proto);
- total += VLAN_HLEN;
-
- ret = skb_copy_datagram_iter(skb, 0, iter, vlan_offset);
- if (ret || !iov_iter_count(iter))
- goto done;
-
- ret = copy_to_iter(&veth, sizeof(veth), iter);
- if (ret != sizeof(veth) || !iov_iter_count(iter))
- goto done;
- }
-
- ret = skb_copy_datagram_iter(skb, vlan_offset, iter,
- skb->len - vlan_offset);
-
-done:
- return ret ? ret : total;
-}
-
-static ssize_t macvtap_do_read(struct macvtap_queue *q,
- struct iov_iter *to,
- int noblock)
-{
- DEFINE_WAIT(wait);
- struct sk_buff *skb;
- ssize_t ret = 0;
-
- if (!iov_iter_count(to))
- return 0;
-
- while (1) {
- if (!noblock)
- prepare_to_wait(sk_sleep(&q->sk), &wait,
- TASK_INTERRUPTIBLE);
-
- /* Read frames from the queue */
- skb = skb_array_consume(&q->skb_array);
- if (skb)
- break;
- if (noblock) {
- ret = -EAGAIN;
- break;
- }
- if (signal_pending(current)) {
- ret = -ERESTARTSYS;
- break;
- }
- /* Nothing to read, let's sleep */
- schedule();
- }
- if (!noblock)
- finish_wait(sk_sleep(&q->sk), &wait);
-
- if (skb) {
- ret = macvtap_put_user(q, skb, to);
- if (unlikely(ret < 0))
- kfree_skb(skb);
- else
- consume_skb(skb);
- }
- return ret;
-}
-
-static ssize_t macvtap_read_iter(struct kiocb *iocb, struct iov_iter *to)
-{
- struct file *file = iocb->ki_filp;
- struct macvtap_queue *q = file->private_data;
- ssize_t len = iov_iter_count(to), ret;
-
- ret = macvtap_do_read(q, to, file->f_flags & O_NONBLOCK);
- ret = min_t(ssize_t, ret, len);
- if (ret > 0)
- iocb->ki_pos = ret;
- return ret;
-}
-
-static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
-{
- struct macvlan_dev *vlan;
-
- ASSERT_RTNL();
- vlan = rtnl_dereference(q->vlan);
- if (vlan)
- dev_hold(vlan->dev);
-
- return vlan;
-}
-
-static void macvtap_put_vlan(struct macvlan_dev *vlan)
-{
- dev_put(vlan->dev);
-}
-
-static int macvtap_ioctl_set_queue(struct file *file, unsigned int flags)
-{
- struct macvtap_queue *q = file->private_data;
- struct macvlan_dev *vlan;
- int ret;
-
- vlan = macvtap_get_vlan(q);
- if (!vlan)
- return -EINVAL;
-
- if (flags & IFF_ATTACH_QUEUE)
- ret = macvtap_enable_queue(vlan->dev, file, q);
- else if (flags & IFF_DETACH_QUEUE)
- ret = macvtap_disable_queue(q);
- else
- ret = -EINVAL;
-
- macvtap_put_vlan(vlan);
- return ret;
-}
-
-static int set_offload(struct macvtap_queue *q, unsigned long arg)
-{
- struct macvlan_dev *vlan;
- netdev_features_t features;
- netdev_features_t feature_mask = 0;
-
- vlan = rtnl_dereference(q->vlan);
- if (!vlan)
- return -ENOLINK;
-
- features = vlan->dev->features;
-
- if (arg & TUN_F_CSUM) {
- feature_mask = NETIF_F_HW_CSUM;
-
- if (arg & (TUN_F_TSO4 | TUN_F_TSO6)) {
- if (arg & TUN_F_TSO_ECN)
- feature_mask |= NETIF_F_TSO_ECN;
- if (arg & TUN_F_TSO4)
- feature_mask |= NETIF_F_TSO;
- if (arg & TUN_F_TSO6)
- feature_mask |= NETIF_F_TSO6;
- }
-
- if (arg & TUN_F_UFO)
- feature_mask |= NETIF_F_UFO;
- }
-
- /* tun/tap driver inverts the usage for TSO offloads, where
- * setting the TSO bit means that the userspace wants to
- * accept TSO frames and turning it off means that user space
- * does not support TSO.
- * For macvtap, we have to invert it to mean the same thing.
- * When user space turns off TSO, we turn off GSO/LRO so that
- * user-space will not receive TSO frames.
- */
- if (feature_mask & (NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_UFO))
- features |= RX_OFFLOADS;
- else
- features &= ~RX_OFFLOADS;
-
- /* tap_features are the same as features on tun/tap and
- * reflect user expectations.
- */
- vlan->tap_features = feature_mask;
- vlan->set_features = features;
- netdev_update_features(vlan->dev);
-
- return 0;
-}
-
-/*
- * provide compatibility with generic tun/tap interface
- */
-static long macvtap_ioctl(struct file *file, unsigned int cmd,
- unsigned long arg)
-{
- struct macvtap_queue *q = file->private_data;
- struct macvlan_dev *vlan;
- void __user *argp = (void __user *)arg;
- struct ifreq __user *ifr = argp;
- unsigned int __user *up = argp;
- unsigned short u;
- int __user *sp = argp;
- struct sockaddr sa;
- int s;
- int ret;
-
- switch (cmd) {
- case TUNSETIFF:
- /* ignore the name, just look at flags */
- if (get_user(u, &ifr->ifr_flags))
- return -EFAULT;
-
- ret = 0;
- if ((u & ~MACVTAP_FEATURES) != (IFF_NO_PI | IFF_TAP))
- ret = -EINVAL;
- else
- q->flags = (q->flags & ~MACVTAP_FEATURES) | u;
-
- return ret;
-
- case TUNGETIFF:
- rtnl_lock();
- vlan = macvtap_get_vlan(q);
- if (!vlan) {
- rtnl_unlock();
- return -ENOLINK;
- }
-
- ret = 0;
- u = q->flags;
- if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
- put_user(u, &ifr->ifr_flags))
- ret = -EFAULT;
- macvtap_put_vlan(vlan);
- rtnl_unlock();
- return ret;
-
- case TUNSETQUEUE:
- if (get_user(u, &ifr->ifr_flags))
- return -EFAULT;
- rtnl_lock();
- ret = macvtap_ioctl_set_queue(file, u);
- rtnl_unlock();
- return ret;
-
- case TUNGETFEATURES:
- if (put_user(IFF_TAP | IFF_NO_PI | MACVTAP_FEATURES, up))
- return -EFAULT;
- return 0;
-
- case TUNSETSNDBUF:
- if (get_user(s, sp))
- return -EFAULT;
-
- q->sk.sk_sndbuf = s;
- return 0;
-
- case TUNGETVNETHDRSZ:
- s = q->vnet_hdr_sz;
- if (put_user(s, sp))
- return -EFAULT;
- return 0;
-
- case TUNSETVNETHDRSZ:
- if (get_user(s, sp))
- return -EFAULT;
- if (s < (int)sizeof(struct virtio_net_hdr))
- return -EINVAL;
-
- q->vnet_hdr_sz = s;
- return 0;
-
- case TUNGETVNETLE:
- s = !!(q->flags & MACVTAP_VNET_LE);
- if (put_user(s, sp))
- return -EFAULT;
- return 0;
-
- case TUNSETVNETLE:
- if (get_user(s, sp))
- return -EFAULT;
- if (s)
- q->flags |= MACVTAP_VNET_LE;
- else
- q->flags &= ~MACVTAP_VNET_LE;
- return 0;
-
- case TUNGETVNETBE:
- return macvtap_get_vnet_be(q, sp);
-
- case TUNSETVNETBE:
- return macvtap_set_vnet_be(q, sp);
-
- case TUNSETOFFLOAD:
- /* let the user check for future flags */
- if (arg & ~(TUN_F_CSUM | TUN_F_TSO4 | TUN_F_TSO6 |
- TUN_F_TSO_ECN | TUN_F_UFO))
- return -EINVAL;
-
- rtnl_lock();
- ret = set_offload(q, arg);
- rtnl_unlock();
- return ret;
-
- case SIOCGIFHWADDR:
- rtnl_lock();
- vlan = macvtap_get_vlan(q);
- if (!vlan) {
- rtnl_unlock();
- return -ENOLINK;
- }
- ret = 0;
- u = vlan->dev->type;
- if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
- copy_to_user(&ifr->ifr_hwaddr.sa_data, vlan->dev->dev_addr, ETH_ALEN) ||
- put_user(u, &ifr->ifr_hwaddr.sa_family))
- ret = -EFAULT;
- macvtap_put_vlan(vlan);
- rtnl_unlock();
- return ret;
-
- case SIOCSIFHWADDR:
- if (copy_from_user(&sa, &ifr->ifr_hwaddr, sizeof(sa)))
- return -EFAULT;
- rtnl_lock();
- vlan = macvtap_get_vlan(q);
- if (!vlan) {
- rtnl_unlock();
- return -ENOLINK;
- }
- ret = dev_set_mac_address(vlan->dev, &sa);
- macvtap_put_vlan(vlan);
- rtnl_unlock();
- return ret;
-
- default:
- return -EINVAL;
- }
-}
-
-#ifdef CONFIG_COMPAT
-static long macvtap_compat_ioctl(struct file *file, unsigned int cmd,
- unsigned long arg)
-{
- return macvtap_ioctl(file, cmd, (unsigned long)compat_ptr(arg));
-}
-#endif
-
-static const struct file_operations macvtap_fops = {
- .owner = THIS_MODULE,
- .open = macvtap_open,
- .release = macvtap_release,
- .read_iter = macvtap_read_iter,
- .write_iter = macvtap_write_iter,
- .poll = macvtap_poll,
- .llseek = no_llseek,
- .unlocked_ioctl = macvtap_ioctl,
-#ifdef CONFIG_COMPAT
- .compat_ioctl = macvtap_compat_ioctl,
-#endif
-};
-
-static int macvtap_sendmsg(struct socket *sock, struct msghdr *m,
- size_t total_len)
-{
- struct macvtap_queue *q = container_of(sock, struct macvtap_queue, sock);
- return macvtap_get_user(q, m, &m->msg_iter, m->msg_flags & MSG_DONTWAIT);
-}
-
-static int macvtap_recvmsg(struct socket *sock, struct msghdr *m,
- size_t total_len, int flags)
-{
- struct macvtap_queue *q = container_of(sock, struct macvtap_queue, sock);
- int ret;
- if (flags & ~(MSG_DONTWAIT|MSG_TRUNC))
- return -EINVAL;
- ret = macvtap_do_read(q, &m->msg_iter, flags & MSG_DONTWAIT);
- if (ret > total_len) {
- m->msg_flags |= MSG_TRUNC;
- ret = flags & MSG_TRUNC ? ret : total_len;
- }
- return ret;
-}
-
-static int macvtap_peek_len(struct socket *sock)
-{
- struct macvtap_queue *q = container_of(sock, struct macvtap_queue,
- sock);
- return skb_array_peek_len(&q->skb_array);
-}
-
-/* Ops structure to mimic raw sockets with tun */
-static const struct proto_ops macvtap_socket_ops = {
- .sendmsg = macvtap_sendmsg,
- .recvmsg = macvtap_recvmsg,
- .peek_len = macvtap_peek_len,
-};
-
-/* Get an underlying socket object from tun file. Returns error unless file is
- * attached to a device. The returned object works like a packet socket, it
- * can be used for sock_sendmsg/sock_recvmsg. The caller is responsible for
- * holding a reference to the file for as long as the socket is in use. */
-struct socket *macvtap_get_socket(struct file *file)
-{
- struct macvtap_queue *q;
- if (file->f_op != &macvtap_fops)
- return ERR_PTR(-EINVAL);
- q = file->private_data;
- if (!q)
- return ERR_PTR(-EBADFD);
- return &q->sock;
-}
-EXPORT_SYMBOL_GPL(macvtap_get_socket);
-
-static int macvtap_queue_resize(struct macvlan_dev *vlan)
-{
- struct net_device *dev = vlan->dev;
- struct macvtap_queue *q;
- struct skb_array **arrays;
- int n = vlan->numqueues;
- int ret, i = 0;
-
- arrays = kmalloc(sizeof *arrays * n, GFP_KERNEL);
- if (!arrays)
- return -ENOMEM;
-
- list_for_each_entry(q, &vlan->queue_list, next)
- arrays[i++] = &q->skb_array;
-
- ret = skb_array_resize_multiple(arrays, n,
- dev->tx_queue_len, GFP_KERNEL);
-
- kfree(arrays);
- return ret;
-}
-
-static int macvtap_device_event(struct notifier_block *unused,
- unsigned long event, void *ptr)
-{
- struct net_device *dev = netdev_notifier_info_to_dev(ptr);
- struct macvlan_dev *vlan;
- struct device *classdev;
- dev_t devt;
- int err;
- char tap_name[IFNAMSIZ];
-
- if (dev->rtnl_link_ops != &macvtap_link_ops)
- return NOTIFY_DONE;
-
- snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
- vlan = netdev_priv(dev);
-
- switch (event) {
- case NETDEV_REGISTER:
- /* Create the device node here after the network device has
- * been registered but before register_netdevice has
- * finished running.
- */
- err = macvtap_get_minor(vlan);
- if (err)
- return notifier_from_errno(err);
-
- devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
- classdev = device_create(&macvtap_class, &dev->dev, devt,
- dev, tap_name);
- if (IS_ERR(classdev)) {
- macvtap_free_minor(vlan);
- return notifier_from_errno(PTR_ERR(classdev));
- }
- err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
- tap_name);
- if (err)
- return notifier_from_errno(err);
- break;
- case NETDEV_UNREGISTER:
- /* vlan->minor == 0 if NETDEV_REGISTER above failed */
- if (vlan->minor == 0)
- break;
- sysfs_remove_link(&dev->dev.kobj, tap_name);
- devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
- device_destroy(&macvtap_class, devt);
- macvtap_free_minor(vlan);
- break;
- case NETDEV_CHANGE_TX_QUEUE_LEN:
- if (macvtap_queue_resize(vlan))
- return NOTIFY_BAD;
- break;
- }
-
- return NOTIFY_DONE;
-}
-
-static struct notifier_block macvtap_notifier_block __read_mostly = {
- .notifier_call = macvtap_device_event,
-};
-
-static int macvtap_init(void)
-{
- int err;
-
- err = alloc_chrdev_region(&macvtap_major, 0,
- MACVTAP_NUM_DEVS, "macvtap");
- if (err)
- goto out1;
-
- cdev_init(&macvtap_cdev, &macvtap_fops);
- err = cdev_add(&macvtap_cdev, macvtap_major, MACVTAP_NUM_DEVS);
- if (err)
- goto out2;
-
- err = class_register(&macvtap_class);
- if (err)
- goto out3;
-
- err = register_netdevice_notifier(&macvtap_notifier_block);
- if (err)
- goto out4;
-
- err = macvlan_link_register(&macvtap_link_ops);
- if (err)
- goto out5;
-
- return 0;
-
-out5:
- unregister_netdevice_notifier(&macvtap_notifier_block);
-out4:
- class_unregister(&macvtap_class);
-out3:
- cdev_del(&macvtap_cdev);
-out2:
- unregister_chrdev_region(macvtap_major, MACVTAP_NUM_DEVS);
-out1:
- return err;
-}
-module_init(macvtap_init);
-
-static void macvtap_exit(void)
-{
- rtnl_link_unregister(&macvtap_link_ops);
- unregister_netdevice_notifier(&macvtap_notifier_block);
- class_unregister(&macvtap_class);
- cdev_del(&macvtap_cdev);
- unregister_chrdev_region(macvtap_major, MACVTAP_NUM_DEVS);
- idr_destroy(&minor_idr);
-}
-module_exit(macvtap_exit);
-
-MODULE_ALIAS_RTNL_LINK("macvtap");
-MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
-MODULE_LICENSE("GPL");
diff --git a/drivers/net/macvtap_main.c b/drivers/net/macvtap_main.c
new file mode 100644
index 0000000..96ffa60
--- /dev/null
+++ b/drivers/net/macvtap_main.c
@@ -0,0 +1,218 @@
+#include <linux/etherdevice.h>
+#include <linux/if_macvlan.h>
+#include <linux/if_macvtap.h>
+#include <linux/if_vlan.h>
+#include <linux/interrupt.h>
+#include <linux/nsproxy.h>
+#include <linux/compat.h>
+#include <linux/if_tun.h>
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <linux/cache.h>
+#include <linux/sched.h>
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/wait.h>
+#include <linux/cdev.h>
+#include <linux/idr.h>
+#include <linux/fs.h>
+#include <linux/uio.h>
+
+#include <net/net_namespace.h>
+#include <net/rtnetlink.h>
+#include <net/sock.h>
+#include <linux/virtio_net.h>
+#include <linux/skb_array.h>
+
+/*
+ * Variables for dealing with macvtaps device numbers.
+ */
+static dev_t macvtap_major;
+#define MACVTAP_NUM_DEVS (1U << MINORBITS)
+
+static const void *macvtap_net_namespace(struct device *d)
+{
+ struct net_device *dev = to_net_dev(d->parent);
+ return dev_net(dev);
+}
+
+static struct class macvtap_class = {
+ .name = "macvtap",
+ .owner = THIS_MODULE,
+ .ns_type = &net_ns_type_operations,
+ .namespace = macvtap_net_namespace,
+};
+static struct cdev macvtap_cdev;
+
+#define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN | NETIF_F_TSO | \
+ NETIF_F_TSO6 | NETIF_F_UFO)
+
+static int macvtap_newlink(struct net *src_net,
+ struct net_device *dev,
+ struct nlattr *tb[],
+ struct nlattr *data[])
+{
+ struct macvlan_dev *vlan = netdev_priv(dev);
+ int err;
+
+ INIT_LIST_HEAD(&vlan->queue_list);
+
+ /* Since macvlan supports all offloads by default, make
+ * tap support all offloads also.
+ */
+ vlan->tap_features = TUN_OFFLOADS;
+
+ err = netdev_rx_handler_register(dev, macvtap_handle_frame, vlan);
+ if (err)
+ return err;
+
+ /* Don't put anything that may fail after macvlan_common_newlink
+ * because we can't undo what it does.
+ */
+ err = macvlan_common_newlink(src_net, dev, tb, data);
+ if (err) {
+ netdev_rx_handler_unregister(dev);
+ return err;
+ }
+
+ return 0;
+}
+
+static void macvtap_dellink(struct net_device *dev,
+ struct list_head *head)
+{
+ netdev_rx_handler_unregister(dev);
+ macvtap_del_queues(dev);
+ macvlan_dellink(dev, head);
+}
+
+static void macvtap_setup(struct net_device *dev)
+{
+ macvlan_common_setup(dev);
+ dev->tx_queue_len = TUN_READQ_SIZE;
+}
+
+static struct rtnl_link_ops macvtap_link_ops __read_mostly = {
+ .kind = "macvtap",
+ .setup = macvtap_setup,
+ .newlink = macvtap_newlink,
+ .dellink = macvtap_dellink,
+};
+
+static int macvtap_device_event(struct notifier_block *unused,
+ unsigned long event, void *ptr)
+{
+ struct net_device *dev = netdev_notifier_info_to_dev(ptr);
+ struct macvlan_dev *vlan;
+ struct device *classdev;
+ dev_t devt;
+ int err;
+ char tap_name[IFNAMSIZ];
+
+ if (dev->rtnl_link_ops != &macvtap_link_ops)
+ return NOTIFY_DONE;
+
+ snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
+ vlan = netdev_priv(dev);
+
+ switch (event) {
+ case NETDEV_REGISTER:
+ /* Create the device node here after the network device has
+ * been registered but before register_netdevice has
+ * finished running.
+ */
+ err = macvtap_get_minor(vlan);
+ if (err)
+ return notifier_from_errno(err);
+
+ devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
+ classdev = device_create(&macvtap_class, &dev->dev, devt,
+ dev, tap_name);
+ if (IS_ERR(classdev)) {
+ macvtap_free_minor(vlan);
+ return notifier_from_errno(PTR_ERR(classdev));
+ }
+ err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
+ tap_name);
+ if (err)
+ return notifier_from_errno(err);
+ break;
+ case NETDEV_UNREGISTER:
+ /* vlan->minor == 0 if NETDEV_REGISTER above failed */
+ if (vlan->minor == 0)
+ break;
+ sysfs_remove_link(&dev->dev.kobj, tap_name);
+ devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
+ device_destroy(&macvtap_class, devt);
+ macvtap_free_minor(vlan);
+ break;
+ case NETDEV_CHANGE_TX_QUEUE_LEN:
+ if (macvtap_queue_resize(vlan))
+ return NOTIFY_BAD;
+ break;
+ }
+
+ return NOTIFY_DONE;
+}
+
+static struct notifier_block macvtap_notifier_block __read_mostly = {
+ .notifier_call = macvtap_device_event,
+};
+
+extern struct file_operations macvtap_fops;
+static int macvtap_init(void)
+{
+ int err;
+
+ err = alloc_chrdev_region(&macvtap_major, 0,
+ MACVTAP_NUM_DEVS, "macvtap");
+ if (err)
+ goto out1;
+
+ cdev_init(&macvtap_cdev, &macvtap_fops);
+ err = cdev_add(&macvtap_cdev, macvtap_major, MACVTAP_NUM_DEVS);
+ if (err)
+ goto out2;
+
+ err = class_register(&macvtap_class);
+ if (err)
+ goto out3;
+
+ err = register_netdevice_notifier(&macvtap_notifier_block);
+ if (err)
+ goto out4;
+
+ err = macvlan_link_register(&macvtap_link_ops);
+ if (err)
+ goto out5;
+
+ return 0;
+
+out5:
+ unregister_netdevice_notifier(&macvtap_notifier_block);
+out4:
+ class_unregister(&macvtap_class);
+out3:
+ cdev_del(&macvtap_cdev);
+out2:
+ unregister_chrdev_region(macvtap_major, MACVTAP_NUM_DEVS);
+out1:
+ return err;
+}
+module_init(macvtap_init);
+
+extern struct idr minor_idr;
+static void macvtap_exit(void)
+{
+ rtnl_link_unregister(&macvtap_link_ops);
+ unregister_netdevice_notifier(&macvtap_notifier_block);
+ class_unregister(&macvtap_class);
+ cdev_del(&macvtap_cdev);
+ unregister_chrdev_region(macvtap_major, MACVTAP_NUM_DEVS);
+ idr_destroy(&minor_idr);
+}
+module_exit(macvtap_exit);
+
+MODULE_ALIAS_RTNL_LINK("macvtap");
+MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/net/tap.c b/drivers/net/tap.c
new file mode 100644
index 0000000..8f12a39
--- /dev/null
+++ b/drivers/net/tap.c
@@ -0,0 +1,1186 @@
+#include <linux/etherdevice.h>
+#include <linux/if_macvlan.h>
+#include <linux/if_vlan.h>
+#include <linux/interrupt.h>
+#include <linux/nsproxy.h>
+#include <linux/compat.h>
+#include <linux/if_tun.h>
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <linux/cache.h>
+#include <linux/sched.h>
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/wait.h>
+#include <linux/cdev.h>
+#include <linux/idr.h>
+#include <linux/fs.h>
+#include <linux/uio.h>
+
+#include <net/net_namespace.h>
+#include <net/rtnetlink.h>
+#include <net/sock.h>
+#include <linux/virtio_net.h>
+#include <linux/skb_array.h>
+
+/*
+ * A macvtap queue is the central object of this driver, it connects
+ * an open character device to a macvlan interface. There can be
+ * multiple queues on one interface, which map back to queues
+ * implemented in hardware on the underlying device.
+ *
+ * macvtap_proto is used to allocate queues through the sock allocation
+ * mechanism.
+ *
+ */
+struct macvtap_queue {
+ struct sock sk;
+ struct socket sock;
+ struct socket_wq wq;
+ int vnet_hdr_sz;
+ struct macvlan_dev __rcu *vlan;
+ struct file *file;
+ unsigned int flags;
+ u16 queue_index;
+ bool enabled;
+ struct list_head next;
+ struct skb_array skb_array;
+};
+
+#define MACVTAP_FEATURES (IFF_VNET_HDR | IFF_MULTI_QUEUE)
+
+#define MACVTAP_VNET_LE 0x80000000
+#define MACVTAP_VNET_BE 0x40000000
+
+#ifdef CONFIG_TUN_VNET_CROSS_LE
+static inline bool macvtap_legacy_is_little_endian(struct macvtap_queue *q)
+{
+ return q->flags & MACVTAP_VNET_BE ? false :
+ virtio_legacy_is_little_endian();
+}
+
+static long macvtap_get_vnet_be(struct macvtap_queue *q, int __user *sp)
+{
+ int s = !!(q->flags & MACVTAP_VNET_BE);
+
+ if (put_user(s, sp))
+ return -EFAULT;
+
+ return 0;
+}
+
+static long macvtap_set_vnet_be(struct macvtap_queue *q, int __user *sp)
+{
+ int s;
+
+ if (get_user(s, sp))
+ return -EFAULT;
+
+ if (s)
+ q->flags |= MACVTAP_VNET_BE;
+ else
+ q->flags &= ~MACVTAP_VNET_BE;
+
+ return 0;
+}
+#else
+static inline bool macvtap_legacy_is_little_endian(struct macvtap_queue *q)
+{
+ return virtio_legacy_is_little_endian();
+}
+
+static long macvtap_get_vnet_be(struct macvtap_queue *q, int __user *argp)
+{
+ return -EINVAL;
+}
+
+static long macvtap_set_vnet_be(struct macvtap_queue *q, int __user *argp)
+{
+ return -EINVAL;
+}
+#endif /* CONFIG_TUN_VNET_CROSS_LE */
+
+static inline bool macvtap_is_little_endian(struct macvtap_queue *q)
+{
+ return q->flags & MACVTAP_VNET_LE ||
+ macvtap_legacy_is_little_endian(q);
+}
+
+static inline u16 macvtap16_to_cpu(struct macvtap_queue *q, __virtio16 val)
+{
+ return __virtio16_to_cpu(macvtap_is_little_endian(q), val);
+}
+
+static inline __virtio16 cpu_to_macvtap16(struct macvtap_queue *q, u16 val)
+{
+ return __cpu_to_virtio16(macvtap_is_little_endian(q), val);
+}
+
+static struct proto macvtap_proto = {
+ .name = "macvtap",
+ .owner = THIS_MODULE,
+ .obj_size = sizeof (struct macvtap_queue),
+};
+
+#define MACVTAP_NUM_DEVS (1U << MINORBITS)
+static DEFINE_MUTEX(minor_lock);
+DEFINE_IDR(minor_idr);
+
+#define GOODCOPY_LEN 128
+
+static const struct proto_ops macvtap_socket_ops;
+
+#define RX_OFFLOADS (NETIF_F_GRO | NETIF_F_LRO)
+#define TAP_FEATURES (NETIF_F_GSO | NETIF_F_SG | NETIF_F_FRAGLIST)
+
+static struct macvlan_dev *macvtap_get_vlan_rcu(const struct net_device *dev)
+{
+ return rcu_dereference(dev->rx_handler_data);
+}
+
+/*
+ * RCU usage:
+ * The macvtap_queue and the macvlan_dev are loosely coupled, the
+ * pointers from one to the other can only be read while rcu_read_lock
+ * or rtnl is held.
+ *
+ * Both the file and the macvlan_dev hold a reference on the macvtap_queue
+ * through sock_hold(&q->sk). When the macvlan_dev goes away first,
+ * q->vlan becomes inaccessible. When the files gets closed,
+ * macvtap_get_queue() fails.
+ *
+ * There may still be references to the struct sock inside of the
+ * queue from outbound SKBs, but these never reference back to the
+ * file or the dev. The data structure is freed through __sk_free
+ * when both our references and any pending SKBs are gone.
+ */
+
+static int macvtap_enable_queue(struct net_device *dev, struct file *file,
+ struct macvtap_queue *q)
+{
+ struct macvlan_dev *vlan = netdev_priv(dev);
+ int err = -EINVAL;
+
+ ASSERT_RTNL();
+
+ if (q->enabled)
+ goto out;
+
+ err = 0;
+ rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
+ q->queue_index = vlan->numvtaps;
+ q->enabled = true;
+
+ vlan->numvtaps++;
+out:
+ return err;
+}
+
+/* Requires RTNL */
+static int macvtap_set_queue(struct net_device *dev, struct file *file,
+ struct macvtap_queue *q)
+{
+ struct macvlan_dev *vlan = netdev_priv(dev);
+
+ if (vlan->numqueues == MAX_MACVTAP_QUEUES)
+ return -EBUSY;
+
+ rcu_assign_pointer(q->vlan, vlan);
+ rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
+ sock_hold(&q->sk);
+
+ q->file = file;
+ q->queue_index = vlan->numvtaps;
+ q->enabled = true;
+ file->private_data = q;
+ list_add_tail(&q->next, &vlan->queue_list);
+
+ vlan->numvtaps++;
+ vlan->numqueues++;
+
+ return 0;
+}
+
+static int macvtap_disable_queue(struct macvtap_queue *q)
+{
+ struct macvlan_dev *vlan;
+ struct macvtap_queue *nq;
+
+ ASSERT_RTNL();
+ if (!q->enabled)
+ return -EINVAL;
+
+ vlan = rtnl_dereference(q->vlan);
+
+ if (vlan) {
+ int index = q->queue_index;
+ BUG_ON(index >= vlan->numvtaps);
+ nq = rtnl_dereference(vlan->taps[vlan->numvtaps - 1]);
+ nq->queue_index = index;
+
+ rcu_assign_pointer(vlan->taps[index], nq);
+ RCU_INIT_POINTER(vlan->taps[vlan->numvtaps - 1], NULL);
+ q->enabled = false;
+
+ vlan->numvtaps--;
+ }
+
+ return 0;
+}
+
+/*
+ * The file owning the queue got closed, give up both
+ * the reference that the files holds as well as the
+ * one from the macvlan_dev if that still exists.
+ *
+ * Using the spinlock makes sure that we don't get
+ * to the queue again after destroying it.
+ */
+static void macvtap_put_queue(struct macvtap_queue *q)
+{
+ struct macvlan_dev *vlan;
+
+ rtnl_lock();
+ vlan = rtnl_dereference(q->vlan);
+
+ if (vlan) {
+ if (q->enabled)
+ BUG_ON(macvtap_disable_queue(q));
+
+ vlan->numqueues--;
+ RCU_INIT_POINTER(q->vlan, NULL);
+ sock_put(&q->sk);
+ list_del_init(&q->next);
+ }
+
+ rtnl_unlock();
+
+ synchronize_rcu();
+ sock_put(&q->sk);
+}
+
+/*
+ * Select a queue based on the rxq of the device on which this packet
+ * arrived. If the incoming device is not mq, calculate a flow hash
+ * to select a queue. If all fails, find the first available queue.
+ * Cache vlan->numvtaps since it can become zero during the execution
+ * of this function.
+ */
+static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
+ struct sk_buff *skb)
+{
+ struct macvlan_dev *vlan = netdev_priv(dev);
+ struct macvtap_queue *tap = NULL;
+ /* Access to taps array is protected by rcu, but access to numvtaps
+ * isn't. Below we use it to lookup a queue, but treat it as a hint
+ * and validate that the result isn't NULL - in case we are
+ * racing against queue removal.
+ */
+ int numvtaps = ACCESS_ONCE(vlan->numvtaps);
+ __u32 rxq;
+
+ if (!numvtaps)
+ goto out;
+
+ if (numvtaps == 1)
+ goto single;
+
+ /* Check if we can use flow to select a queue */
+ rxq = skb_get_hash(skb);
+ if (rxq) {
+ tap = rcu_dereference(vlan->taps[rxq % numvtaps]);
+ goto out;
+ }
+
+ if (likely(skb_rx_queue_recorded(skb))) {
+ rxq = skb_get_rx_queue(skb);
+
+ while (unlikely(rxq >= numvtaps))
+ rxq -= numvtaps;
+
+ tap = rcu_dereference(vlan->taps[rxq]);
+ goto out;
+ }
+
+single:
+ tap = rcu_dereference(vlan->taps[0]);
+out:
+ return tap;
+}
+
+/*
+ * The net_device is going away, give up the reference
+ * that it holds on all queues and safely set the pointer
+ * from the queues to NULL.
+ */
+void macvtap_del_queues(struct net_device *dev)
+{
+ struct macvlan_dev *vlan = netdev_priv(dev);
+ struct macvtap_queue *q, *tmp;
+
+ ASSERT_RTNL();
+ list_for_each_entry_safe(q, tmp, &vlan->queue_list, next) {
+ list_del_init(&q->next);
+ RCU_INIT_POINTER(q->vlan, NULL);
+ if (q->enabled)
+ vlan->numvtaps--;
+ vlan->numqueues--;
+ sock_put(&q->sk);
+ }
+ BUG_ON(vlan->numvtaps);
+ BUG_ON(vlan->numqueues);
+ /* guarantee that any future macvtap_set_queue will fail */
+ vlan->numvtaps = MAX_MACVTAP_QUEUES;
+}
+
+rx_handler_result_t macvtap_handle_frame(struct sk_buff **pskb)
+{
+ struct sk_buff *skb = *pskb;
+ struct net_device *dev = skb->dev;
+ struct macvlan_dev *vlan;
+ struct macvtap_queue *q;
+ netdev_features_t features = TAP_FEATURES;
+
+ vlan = macvtap_get_vlan_rcu(dev);
+ if (!vlan)
+ return RX_HANDLER_PASS;
+
+ q = macvtap_get_queue(dev, skb);
+ if (!q)
+ return RX_HANDLER_PASS;
+
+ if (__skb_array_full(&q->skb_array))
+ goto drop;
+
+ skb_push(skb, ETH_HLEN);
+
+ /* Apply the forward feature mask so that we perform segmentation
+ * according to users wishes. This only works if VNET_HDR is
+ * enabled.
+ */
+ if (q->flags & IFF_VNET_HDR)
+ features |= vlan->tap_features;
+ if (netif_needs_gso(skb, features)) {
+ struct sk_buff *segs = __skb_gso_segment(skb, features, false);
+
+ if (IS_ERR(segs))
+ goto drop;
+
+ if (!segs) {
+ if (skb_array_produce(&q->skb_array, skb))
+ goto drop;
+ goto wake_up;
+ }
+
+ consume_skb(skb);
+ while (segs) {
+ struct sk_buff *nskb = segs->next;
+
+ segs->next = NULL;
+ if (skb_array_produce(&q->skb_array, segs)) {
+ kfree_skb(segs);
+ kfree_skb_list(nskb);
+ break;
+ }
+ segs = nskb;
+ }
+ } else {
+ /* If we receive a partial checksum and the tap side
+ * doesn't support checksum offload, compute the checksum.
+ * Note: it doesn't matter which checksum feature to
+ * check, we either support them all or none.
+ */
+ if (skb->ip_summed == CHECKSUM_PARTIAL &&
+ !(features & NETIF_F_CSUM_MASK) &&
+ skb_checksum_help(skb))
+ goto drop;
+ if (skb_array_produce(&q->skb_array, skb))
+ goto drop;
+ }
+
+wake_up:
+ wake_up_interruptible_poll(sk_sleep(&q->sk), POLLIN | POLLRDNORM | POLLRDBAND);
+ return RX_HANDLER_CONSUMED;
+
+drop:
+ /* Count errors/drops only here, thus don't care about args. */
+ macvlan_count_rx(vlan, 0, 0, 0);
+ kfree_skb(skb);
+ return RX_HANDLER_CONSUMED;
+}
+
+int macvtap_get_minor(struct macvlan_dev *vlan)
+{
+ int retval = -ENOMEM;
+
+ mutex_lock(&minor_lock);
+ retval = idr_alloc(&minor_idr, vlan, 1, MACVTAP_NUM_DEVS, GFP_KERNEL);
+ if (retval >= 0) {
+ vlan->minor = retval;
+ } else if (retval == -ENOSPC) {
+ netdev_err(vlan->dev, "Too many macvtap devices\n");
+ retval = -EINVAL;
+ }
+ mutex_unlock(&minor_lock);
+ return retval < 0 ? retval : 0;
+}
+
+void macvtap_free_minor(struct macvlan_dev *vlan)
+{
+ mutex_lock(&minor_lock);
+ if (vlan->minor) {
+ idr_remove(&minor_idr, vlan->minor);
+ vlan->minor = 0;
+ }
+ mutex_unlock(&minor_lock);
+}
+
+static struct net_device *dev_get_by_macvtap_minor(int minor)
+{
+ struct net_device *dev = NULL;
+ struct macvlan_dev *vlan;
+
+ mutex_lock(&minor_lock);
+ vlan = idr_find(&minor_idr, minor);
+ if (vlan) {
+ dev = vlan->dev;
+ dev_hold(dev);
+ }
+ mutex_unlock(&minor_lock);
+ return dev;
+}
+
+static void macvtap_sock_write_space(struct sock *sk)
+{
+ wait_queue_head_t *wqueue;
+
+ if (!sock_writeable(sk) ||
+ !test_and_clear_bit(SOCKWQ_ASYNC_NOSPACE, &sk->sk_socket->flags))
+ return;
+
+ wqueue = sk_sleep(sk);
+ if (wqueue && waitqueue_active(wqueue))
+ wake_up_interruptible_poll(wqueue, POLLOUT | POLLWRNORM | POLLWRBAND);
+}
+
+static void macvtap_sock_destruct(struct sock *sk)
+{
+ struct macvtap_queue *q = container_of(sk, struct macvtap_queue, sk);
+
+ skb_array_cleanup(&q->skb_array);
+}
+
+static int macvtap_open(struct inode *inode, struct file *file)
+{
+ struct net *net = current->nsproxy->net_ns;
+ struct net_device *dev;
+ struct macvtap_queue *q;
+ int err = -ENODEV;
+
+ rtnl_lock();
+ dev = dev_get_by_macvtap_minor(iminor(inode));
+ if (!dev)
+ goto err;
+
+ err = -ENOMEM;
+ q = (struct macvtap_queue *)sk_alloc(net, AF_UNSPEC, GFP_KERNEL,
+ &macvtap_proto, 0);
+ if (!q)
+ goto err;
+
+ RCU_INIT_POINTER(q->sock.wq, &q->wq);
+ init_waitqueue_head(&q->wq.wait);
+ q->sock.type = SOCK_RAW;
+ q->sock.state = SS_CONNECTED;
+ q->sock.file = file;
+ q->sock.ops = &macvtap_socket_ops;
+ sock_init_data(&q->sock, &q->sk);
+ q->sk.sk_write_space = macvtap_sock_write_space;
+ q->sk.sk_destruct = macvtap_sock_destruct;
+ q->flags = IFF_VNET_HDR | IFF_NO_PI | IFF_TAP;
+ q->vnet_hdr_sz = sizeof(struct virtio_net_hdr);
+
+ /*
+ * so far only KVM virtio_net uses macvtap, enable zero copy between
+ * guest kernel and host kernel when lower device supports zerocopy
+ *
+ * The macvlan supports zerocopy iff the lower device supports zero
+ * copy so we don't have to look at the lower device directly.
+ */
+ if ((dev->features & NETIF_F_HIGHDMA) && (dev->features & NETIF_F_SG))
+ sock_set_flag(&q->sk, SOCK_ZEROCOPY);
+
+ err = -ENOMEM;
+ if (skb_array_init(&q->skb_array, dev->tx_queue_len, GFP_KERNEL))
+ goto err_array;
+
+ err = macvtap_set_queue(dev, file, q);
+ if (err)
+ goto err_queue;
+
+ dev_put(dev);
+
+ rtnl_unlock();
+ return err;
+
+err_queue:
+ skb_array_cleanup(&q->skb_array);
+err_array:
+ sock_put(&q->sk);
+err:
+ if (dev)
+ dev_put(dev);
+
+ rtnl_unlock();
+ return err;
+}
+
+static int macvtap_release(struct inode *inode, struct file *file)
+{
+ struct macvtap_queue *q = file->private_data;
+ macvtap_put_queue(q);
+ return 0;
+}
+
+static unsigned int macvtap_poll(struct file *file, poll_table * wait)
+{
+ struct macvtap_queue *q = file->private_data;
+ unsigned int mask = POLLERR;
+
+ if (!q)
+ goto out;
+
+ mask = 0;
+ poll_wait(file, &q->wq.wait, wait);
+
+ if (!skb_array_empty(&q->skb_array))
+ mask |= POLLIN | POLLRDNORM;
+
+ if (sock_writeable(&q->sk) ||
+ (!test_and_set_bit(SOCKWQ_ASYNC_NOSPACE, &q->sock.flags) &&
+ sock_writeable(&q->sk)))
+ mask |= POLLOUT | POLLWRNORM;
+
+out:
+ return mask;
+}
+
+static inline struct sk_buff *macvtap_alloc_skb(struct sock *sk, size_t prepad,
+ size_t len, size_t linear,
+ int noblock, int *err)
+{
+ struct sk_buff *skb;
+
+ /* Under a page? Don't bother with paged skb. */
+ if (prepad + len < PAGE_SIZE || !linear)
+ linear = len;
+
+ skb = sock_alloc_send_pskb(sk, prepad + linear, len - linear, noblock,
+ err, 0);
+ if (!skb)
+ return NULL;
+
+ skb_reserve(skb, prepad);
+ skb_put(skb, linear);
+ skb->data_len = len - linear;
+ skb->len += len - linear;
+
+ return skb;
+}
+
+/* Neighbour code has some assumptions on HH_DATA_MOD alignment */
+#define MACVTAP_RESERVE HH_DATA_OFF(ETH_HLEN)
+
+/* Get packet from user space buffer */
+static ssize_t macvtap_get_user(struct macvtap_queue *q, struct msghdr *m,
+ struct iov_iter *from, int noblock)
+{
+ int good_linear = SKB_MAX_HEAD(MACVTAP_RESERVE);
+ struct sk_buff *skb;
+ struct macvlan_dev *vlan;
+ unsigned long total_len = iov_iter_count(from);
+ unsigned long len = total_len;
+ int err;
+ struct virtio_net_hdr vnet_hdr = { 0 };
+ int vnet_hdr_len = 0;
+ int copylen = 0;
+ int depth;
+ bool zerocopy = false;
+ size_t linear;
+
+ if (q->flags & IFF_VNET_HDR) {
+ vnet_hdr_len = q->vnet_hdr_sz;
+
+ err = -EINVAL;
+ if (len < vnet_hdr_len)
+ goto err;
+ len -= vnet_hdr_len;
+
+ err = -EFAULT;
+ if (!copy_from_iter_full(&vnet_hdr, sizeof(vnet_hdr), from))
+ goto err;
+ iov_iter_advance(from, vnet_hdr_len - sizeof(vnet_hdr));
+ if ((vnet_hdr.flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) &&
+ macvtap16_to_cpu(q, vnet_hdr.csum_start) +
+ macvtap16_to_cpu(q, vnet_hdr.csum_offset) + 2 >
+ macvtap16_to_cpu(q, vnet_hdr.hdr_len))
+ vnet_hdr.hdr_len = cpu_to_macvtap16(q,
+ macvtap16_to_cpu(q, vnet_hdr.csum_start) +
+ macvtap16_to_cpu(q, vnet_hdr.csum_offset) + 2);
+ err = -EINVAL;
+ if (macvtap16_to_cpu(q, vnet_hdr.hdr_len) > len)
+ goto err;
+ }
+
+ err = -EINVAL;
+ if (unlikely(len < ETH_HLEN))
+ goto err;
+
+ if (m && m->msg_control && sock_flag(&q->sk, SOCK_ZEROCOPY)) {
+ struct iov_iter i;
+
+ copylen = vnet_hdr.hdr_len ?
+ macvtap16_to_cpu(q, vnet_hdr.hdr_len) : GOODCOPY_LEN;
+ if (copylen > good_linear)
+ copylen = good_linear;
+ else if (copylen < ETH_HLEN)
+ copylen = ETH_HLEN;
+ linear = copylen;
+ i = *from;
+ iov_iter_advance(&i, copylen);
+ if (iov_iter_npages(&i, INT_MAX) <= MAX_SKB_FRAGS)
+ zerocopy = true;
+ }
+
+ if (!zerocopy) {
+ copylen = len;
+ linear = macvtap16_to_cpu(q, vnet_hdr.hdr_len);
+ if (linear > good_linear)
+ linear = good_linear;
+ else if (linear < ETH_HLEN)
+ linear = ETH_HLEN;
+ }
+
+ skb = macvtap_alloc_skb(&q->sk, MACVTAP_RESERVE, copylen,
+ linear, noblock, &err);
+ if (!skb)
+ goto err;
+
+ if (zerocopy)
+ err = zerocopy_sg_from_iter(skb, from);
+ else
+ err = skb_copy_datagram_from_iter(skb, 0, from, len);
+
+ if (err)
+ goto err_kfree;
+
+ skb_set_network_header(skb, ETH_HLEN);
+ skb_reset_mac_header(skb);
+ skb->protocol = eth_hdr(skb)->h_proto;
+
+ if (vnet_hdr_len) {
+ err = virtio_net_hdr_to_skb(skb, &vnet_hdr,
+ macvtap_is_little_endian(q));
+ if (err)
+ goto err_kfree;
+ }
+
+ skb_probe_transport_header(skb, ETH_HLEN);
+
+ /* Move network header to the right position for VLAN tagged packets */
+ if ((skb->protocol == htons(ETH_P_8021Q) ||
+ skb->protocol == htons(ETH_P_8021AD)) &&
+ __vlan_get_protocol(skb, skb->protocol, &depth) != 0)
+ skb_set_network_header(skb, depth);
+
+ rcu_read_lock();
+ vlan = rcu_dereference(q->vlan);
+ /* copy skb_ubuf_info for callback when skb has no error */
+ if (zerocopy) {
+ skb_shinfo(skb)->destructor_arg = m->msg_control;
+ skb_shinfo(skb)->tx_flags |= SKBTX_DEV_ZEROCOPY;
+ skb_shinfo(skb)->tx_flags |= SKBTX_SHARED_FRAG;
+ } else if (m && m->msg_control) {
+ struct ubuf_info *uarg = m->msg_control;
+ uarg->callback(uarg, false);
+ }
+
+ if (vlan) {
+ skb->dev = vlan->dev;
+ dev_queue_xmit(skb);
+ } else {
+ kfree_skb(skb);
+ }
+ rcu_read_unlock();
+
+ return total_len;
+
+err_kfree:
+ kfree_skb(skb);
+
+err:
+ rcu_read_lock();
+ vlan = rcu_dereference(q->vlan);
+ if (vlan)
+ this_cpu_inc(vlan->pcpu_stats->tx_dropped);
+ rcu_read_unlock();
+
+ return err;
+}
+
+static ssize_t macvtap_write_iter(struct kiocb *iocb, struct iov_iter *from)
+{
+ struct file *file = iocb->ki_filp;
+ struct macvtap_queue *q = file->private_data;
+
+ return macvtap_get_user(q, NULL, from, file->f_flags & O_NONBLOCK);
+}
+
+/* Put packet to the user space buffer */
+static ssize_t macvtap_put_user(struct macvtap_queue *q,
+ const struct sk_buff *skb,
+ struct iov_iter *iter)
+{
+ int ret;
+ int vnet_hdr_len = 0;
+ int vlan_offset = 0;
+ int total;
+
+ if (q->flags & IFF_VNET_HDR) {
+ struct virtio_net_hdr vnet_hdr;
+ vnet_hdr_len = q->vnet_hdr_sz;
+ if (iov_iter_count(iter) < vnet_hdr_len)
+ return -EINVAL;
+
+ if (virtio_net_hdr_from_skb(skb, &vnet_hdr,
+ macvtap_is_little_endian(q)))
+ BUG();
+
+ if (copy_to_iter(&vnet_hdr, sizeof(vnet_hdr), iter) !=
+ sizeof(vnet_hdr))
+ return -EFAULT;
+
+ iov_iter_advance(iter, vnet_hdr_len - sizeof(vnet_hdr));
+ }
+ total = vnet_hdr_len;
+ total += skb->len;
+
+ if (skb_vlan_tag_present(skb)) {
+ struct {
+ __be16 h_vlan_proto;
+ __be16 h_vlan_TCI;
+ } veth;
+ veth.h_vlan_proto = skb->vlan_proto;
+ veth.h_vlan_TCI = htons(skb_vlan_tag_get(skb));
+
+ vlan_offset = offsetof(struct vlan_ethhdr, h_vlan_proto);
+ total += VLAN_HLEN;
+
+ ret = skb_copy_datagram_iter(skb, 0, iter, vlan_offset);
+ if (ret || !iov_iter_count(iter))
+ goto done;
+
+ ret = copy_to_iter(&veth, sizeof(veth), iter);
+ if (ret != sizeof(veth) || !iov_iter_count(iter))
+ goto done;
+ }
+
+ ret = skb_copy_datagram_iter(skb, vlan_offset, iter,
+ skb->len - vlan_offset);
+
+done:
+ return ret ? ret : total;
+}
+
+static ssize_t macvtap_do_read(struct macvtap_queue *q,
+ struct iov_iter *to,
+ int noblock)
+{
+ DEFINE_WAIT(wait);
+ struct sk_buff *skb;
+ ssize_t ret = 0;
+
+ if (!iov_iter_count(to))
+ return 0;
+
+ while (1) {
+ if (!noblock)
+ prepare_to_wait(sk_sleep(&q->sk), &wait,
+ TASK_INTERRUPTIBLE);
+
+ /* Read frames from the queue */
+ skb = skb_array_consume(&q->skb_array);
+ if (skb)
+ break;
+ if (noblock) {
+ ret = -EAGAIN;
+ break;
+ }
+ if (signal_pending(current)) {
+ ret = -ERESTARTSYS;
+ break;
+ }
+ /* Nothing to read, let's sleep */
+ schedule();
+ }
+ if (!noblock)
+ finish_wait(sk_sleep(&q->sk), &wait);
+
+ if (skb) {
+ ret = macvtap_put_user(q, skb, to);
+ if (unlikely(ret < 0))
+ kfree_skb(skb);
+ else
+ consume_skb(skb);
+ }
+ return ret;
+}
+
+static ssize_t macvtap_read_iter(struct kiocb *iocb, struct iov_iter *to)
+{
+ struct file *file = iocb->ki_filp;
+ struct macvtap_queue *q = file->private_data;
+ ssize_t len = iov_iter_count(to), ret;
+
+ ret = macvtap_do_read(q, to, file->f_flags & O_NONBLOCK);
+ ret = min_t(ssize_t, ret, len);
+ if (ret > 0)
+ iocb->ki_pos = ret;
+ return ret;
+}
+
+static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
+{
+ struct macvlan_dev *vlan;
+
+ ASSERT_RTNL();
+ vlan = rtnl_dereference(q->vlan);
+ if (vlan)
+ dev_hold(vlan->dev);
+
+ return vlan;
+}
+
+static void macvtap_put_vlan(struct macvlan_dev *vlan)
+{
+ dev_put(vlan->dev);
+}
+
+static int macvtap_ioctl_set_queue(struct file *file, unsigned int flags)
+{
+ struct macvtap_queue *q = file->private_data;
+ struct macvlan_dev *vlan;
+ int ret;
+
+ vlan = macvtap_get_vlan(q);
+ if (!vlan)
+ return -EINVAL;
+
+ if (flags & IFF_ATTACH_QUEUE)
+ ret = macvtap_enable_queue(vlan->dev, file, q);
+ else if (flags & IFF_DETACH_QUEUE)
+ ret = macvtap_disable_queue(q);
+ else
+ ret = -EINVAL;
+
+ macvtap_put_vlan(vlan);
+ return ret;
+}
+
+static int set_offload(struct macvtap_queue *q, unsigned long arg)
+{
+ struct macvlan_dev *vlan;
+ netdev_features_t features;
+ netdev_features_t feature_mask = 0;
+
+ vlan = rtnl_dereference(q->vlan);
+ if (!vlan)
+ return -ENOLINK;
+
+ features = vlan->dev->features;
+
+ if (arg & TUN_F_CSUM) {
+ feature_mask = NETIF_F_HW_CSUM;
+
+ if (arg & (TUN_F_TSO4 | TUN_F_TSO6)) {
+ if (arg & TUN_F_TSO_ECN)
+ feature_mask |= NETIF_F_TSO_ECN;
+ if (arg & TUN_F_TSO4)
+ feature_mask |= NETIF_F_TSO;
+ if (arg & TUN_F_TSO6)
+ feature_mask |= NETIF_F_TSO6;
+ }
+
+ if (arg & TUN_F_UFO)
+ feature_mask |= NETIF_F_UFO;
+ }
+
+ /* tun/tap driver inverts the usage for TSO offloads, where
+ * setting the TSO bit means that the userspace wants to
+ * accept TSO frames and turning it off means that user space
+ * does not support TSO.
+ * For macvtap, we have to invert it to mean the same thing.
+ * When user space turns off TSO, we turn off GSO/LRO so that
+ * user-space will not receive TSO frames.
+ */
+ if (feature_mask & (NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_UFO))
+ features |= RX_OFFLOADS;
+ else
+ features &= ~RX_OFFLOADS;
+
+ /* tap_features are the same as features on tun/tap and
+ * reflect user expectations.
+ */
+ vlan->tap_features = feature_mask;
+ vlan->set_features = features;
+ netdev_update_features(vlan->dev);
+
+ return 0;
+}
+
+/*
+ * provide compatibility with generic tun/tap interface
+ */
+static long macvtap_ioctl(struct file *file, unsigned int cmd,
+ unsigned long arg)
+{
+ struct macvtap_queue *q = file->private_data;
+ struct macvlan_dev *vlan;
+ void __user *argp = (void __user *)arg;
+ struct ifreq __user *ifr = argp;
+ unsigned int __user *up = argp;
+ unsigned short u;
+ int __user *sp = argp;
+ struct sockaddr sa;
+ int s;
+ int ret;
+
+ switch (cmd) {
+ case TUNSETIFF:
+ /* ignore the name, just look at flags */
+ if (get_user(u, &ifr->ifr_flags))
+ return -EFAULT;
+
+ ret = 0;
+ if ((u & ~MACVTAP_FEATURES) != (IFF_NO_PI | IFF_TAP))
+ ret = -EINVAL;
+ else
+ q->flags = (q->flags & ~MACVTAP_FEATURES) | u;
+
+ return ret;
+
+ case TUNGETIFF:
+ rtnl_lock();
+ vlan = macvtap_get_vlan(q);
+ if (!vlan) {
+ rtnl_unlock();
+ return -ENOLINK;
+ }
+
+ ret = 0;
+ u = q->flags;
+ if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
+ put_user(u, &ifr->ifr_flags))
+ ret = -EFAULT;
+ macvtap_put_vlan(vlan);
+ rtnl_unlock();
+ return ret;
+
+ case TUNSETQUEUE:
+ if (get_user(u, &ifr->ifr_flags))
+ return -EFAULT;
+ rtnl_lock();
+ ret = macvtap_ioctl_set_queue(file, u);
+ rtnl_unlock();
+ return ret;
+
+ case TUNGETFEATURES:
+ if (put_user(IFF_TAP | IFF_NO_PI | MACVTAP_FEATURES, up))
+ return -EFAULT;
+ return 0;
+
+ case TUNSETSNDBUF:
+ if (get_user(s, sp))
+ return -EFAULT;
+
+ q->sk.sk_sndbuf = s;
+ return 0;
+
+ case TUNGETVNETHDRSZ:
+ s = q->vnet_hdr_sz;
+ if (put_user(s, sp))
+ return -EFAULT;
+ return 0;
+
+ case TUNSETVNETHDRSZ:
+ if (get_user(s, sp))
+ return -EFAULT;
+ if (s < (int)sizeof(struct virtio_net_hdr))
+ return -EINVAL;
+
+ q->vnet_hdr_sz = s;
+ return 0;
+
+ case TUNGETVNETLE:
+ s = !!(q->flags & MACVTAP_VNET_LE);
+ if (put_user(s, sp))
+ return -EFAULT;
+ return 0;
+
+ case TUNSETVNETLE:
+ if (get_user(s, sp))
+ return -EFAULT;
+ if (s)
+ q->flags |= MACVTAP_VNET_LE;
+ else
+ q->flags &= ~MACVTAP_VNET_LE;
+ return 0;
+
+ case TUNGETVNETBE:
+ return macvtap_get_vnet_be(q, sp);
+
+ case TUNSETVNETBE:
+ return macvtap_set_vnet_be(q, sp);
+
+ case TUNSETOFFLOAD:
+ /* let the user check for future flags */
+ if (arg & ~(TUN_F_CSUM | TUN_F_TSO4 | TUN_F_TSO6 |
+ TUN_F_TSO_ECN | TUN_F_UFO))
+ return -EINVAL;
+
+ rtnl_lock();
+ ret = set_offload(q, arg);
+ rtnl_unlock();
+ return ret;
+
+ case SIOCGIFHWADDR:
+ rtnl_lock();
+ vlan = macvtap_get_vlan(q);
+ if (!vlan) {
+ rtnl_unlock();
+ return -ENOLINK;
+ }
+ ret = 0;
+ u = vlan->dev->type;
+ if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
+ copy_to_user(&ifr->ifr_hwaddr.sa_data, vlan->dev->dev_addr, ETH_ALEN) ||
+ put_user(u, &ifr->ifr_hwaddr.sa_family))
+ ret = -EFAULT;
+ macvtap_put_vlan(vlan);
+ rtnl_unlock();
+ return ret;
+
+ case SIOCSIFHWADDR:
+ if (copy_from_user(&sa, &ifr->ifr_hwaddr, sizeof(sa)))
+ return -EFAULT;
+ rtnl_lock();
+ vlan = macvtap_get_vlan(q);
+ if (!vlan) {
+ rtnl_unlock();
+ return -ENOLINK;
+ }
+ ret = dev_set_mac_address(vlan->dev, &sa);
+ macvtap_put_vlan(vlan);
+ rtnl_unlock();
+ return ret;
+
+ default:
+ return -EINVAL;
+ }
+}
+
+#ifdef CONFIG_COMPAT
+static long macvtap_compat_ioctl(struct file *file, unsigned int cmd,
+ unsigned long arg)
+{
+ return macvtap_ioctl(file, cmd, (unsigned long)compat_ptr(arg));
+}
+#endif
+
+const struct file_operations macvtap_fops = {
+ .owner = THIS_MODULE,
+ .open = macvtap_open,
+ .release = macvtap_release,
+ .read_iter = macvtap_read_iter,
+ .write_iter = macvtap_write_iter,
+ .poll = macvtap_poll,
+ .llseek = no_llseek,
+ .unlocked_ioctl = macvtap_ioctl,
+#ifdef CONFIG_COMPAT
+ .compat_ioctl = macvtap_compat_ioctl,
+#endif
+};
+
+static int macvtap_sendmsg(struct socket *sock, struct msghdr *m,
+ size_t total_len)
+{
+ struct macvtap_queue *q = container_of(sock, struct macvtap_queue, sock);
+ return macvtap_get_user(q, m, &m->msg_iter, m->msg_flags & MSG_DONTWAIT);
+}
+
+static int macvtap_recvmsg(struct socket *sock, struct msghdr *m,
+ size_t total_len, int flags)
+{
+ struct macvtap_queue *q = container_of(sock, struct macvtap_queue, sock);
+ int ret;
+ if (flags & ~(MSG_DONTWAIT|MSG_TRUNC))
+ return -EINVAL;
+ ret = macvtap_do_read(q, &m->msg_iter, flags & MSG_DONTWAIT);
+ if (ret > total_len) {
+ m->msg_flags |= MSG_TRUNC;
+ ret = flags & MSG_TRUNC ? ret : total_len;
+ }
+ return ret;
+}
+
+static int macvtap_peek_len(struct socket *sock)
+{
+ struct macvtap_queue *q = container_of(sock, struct macvtap_queue,
+ sock);
+ return skb_array_peek_len(&q->skb_array);
+}
+
+/* Ops structure to mimic raw sockets with tun */
+static const struct proto_ops macvtap_socket_ops = {
+ .sendmsg = macvtap_sendmsg,
+ .recvmsg = macvtap_recvmsg,
+ .peek_len = macvtap_peek_len,
+};
+
+/* Get an underlying socket object from tun file. Returns error unless file is
+ * attached to a device. The returned object works like a packet socket, it
+ * can be used for sock_sendmsg/sock_recvmsg. The caller is responsible for
+ * holding a reference to the file for as long as the socket is in use. */
+struct socket *macvtap_get_socket(struct file *file)
+{
+ struct macvtap_queue *q;
+ if (file->f_op != &macvtap_fops)
+ return ERR_PTR(-EINVAL);
+ q = file->private_data;
+ if (!q)
+ return ERR_PTR(-EBADFD);
+ return &q->sock;
+}
+EXPORT_SYMBOL_GPL(macvtap_get_socket);
+
+int macvtap_queue_resize(struct macvlan_dev *vlan)
+{
+ struct net_device *dev = vlan->dev;
+ struct macvtap_queue *q;
+ struct skb_array **arrays;
+ int n = vlan->numqueues;
+ int ret, i = 0;
+
+ arrays = kmalloc(sizeof *arrays * n, GFP_KERNEL);
+ if (!arrays)
+ return -ENOMEM;
+
+ list_for_each_entry(q, &vlan->queue_list, next)
+ arrays[i++] = &q->skb_array;
+
+ ret = skb_array_resize_multiple(arrays, n,
+ dev->tx_queue_len, GFP_KERNEL);
+
+ kfree(arrays);
+ return ret;
+}
diff --git a/include/linux/if_macvtap.h b/include/linux/if_macvtap.h
new file mode 100644
index 0000000..c9bf84b
--- /dev/null
+++ b/include/linux/if_macvtap.h
@@ -0,0 +1,10 @@
+#ifndef _LINUX_IF_MACVTAP_H_
+#define _LINUX_IF_MACVTAP_H_
+
+rx_handler_result_t macvtap_handle_frame(struct sk_buff **pskb);
+void macvtap_del_queues(struct net_device *dev);
+int macvtap_get_minor(struct macvlan_dev *vlan);
+void macvtap_free_minor(struct macvlan_dev *vlan);
+int macvtap_queue_resize(struct macvlan_dev *vlan);
+
+#endif /*_LINUX_IF_MACVTAP_H_*/
--
2.7.4
^ permalink raw reply related [flat|nested] 21+ messages in thread* Re: [PATCHv1 1/7] TAP: Refactoring macvtap.c
2017-01-06 22:33 ` [PATCHv1 1/7] TAP: Refactoring macvtap.c Sainath Grandhi
@ 2017-01-06 23:12 ` Andy Shevchenko
0 siblings, 0 replies; 21+ messages in thread
From: Andy Shevchenko @ 2017-01-06 23:12 UTC (permalink / raw)
To: Sainath Grandhi
Cc: netdev, David S. Miller, mahesh, linux-kernel@vger.kernel.org
On Sat, Jan 7, 2017 at 12:33 AM, Sainath Grandhi
<sainath.grandhi@intel.com> wrote:
> macvtap module has code for tap/queue management and link management. This patch splits
> the code into macvtap_main.c for link management and tap.c for tap/queue management.
> Functionality in tap.c can be re-used for implementing tap on other virtual interfaces.
>
> Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
> Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
> ---
> drivers/net/Makefile | 2 +
> drivers/net/macvtap.c | 1374 --------------------------------------------
> drivers/net/macvtap_main.c | 218 +++++++
> drivers/net/tap.c | 1186 ++++++++++++++++++++++++++++++++++++++
> include/linux/if_macvtap.h | 10 +
> 5 files changed, 1416 insertions(+), 1374 deletions(-)
> delete mode 100644 drivers/net/macvtap.c
> create mode 100644 drivers/net/macvtap_main.c
> create mode 100644 drivers/net/tap.c
> create mode 100644 include/linux/if_macvtap.h
Can you use -M -C for better view of this change? It's really hard to review.
--
With Best Regards,
Andy Shevchenko
^ permalink raw reply [flat|nested] 21+ messages in thread
* [PATCHv1 2/7] TAP: Renaming tap related APIs, data structures, macros
2017-01-06 22:33 [PATCHv1 0/7] Refactor macvtap to re-use tap functionality by other virtual intefaces Sainath Grandhi
2017-01-06 22:33 ` [PATCHv1 1/7] TAP: Refactoring macvtap.c Sainath Grandhi
@ 2017-01-06 22:33 ` Sainath Grandhi
2017-01-07 10:44 ` kbuild test robot
2017-01-07 12:00 ` kbuild test robot
2017-01-06 22:33 ` [PATCHv1 3/7] TAP: Tap character device creation/destroy API Sainath Grandhi
` (4 subsequent siblings)
6 siblings, 2 replies; 21+ messages in thread
From: Sainath Grandhi @ 2017-01-06 22:33 UTC (permalink / raw)
To: netdev; +Cc: davem, mahesh, linux-kernel, Sainath Grandhi
Renaming tap related APIs, data structures and macros in tap.c from macvtap_.* to tap_.*
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
---
drivers/net/macvtap_main.c | 18 +--
drivers/net/tap.c | 332 ++++++++++++++++++++++-----------------------
drivers/vhost/net.c | 3 +-
include/linux/if_macvlan.h | 4 +-
include/linux/if_macvtap.h | 10 --
include/linux/if_tap.h | 11 ++
6 files changed, 190 insertions(+), 188 deletions(-)
delete mode 100644 include/linux/if_macvtap.h
create mode 100644 include/linux/if_tap.h
diff --git a/drivers/net/macvtap_main.c b/drivers/net/macvtap_main.c
index 96ffa60..548f339 100644
--- a/drivers/net/macvtap_main.c
+++ b/drivers/net/macvtap_main.c
@@ -1,6 +1,6 @@
#include <linux/etherdevice.h>
#include <linux/if_macvlan.h>
-#include <linux/if_macvtap.h>
+#include <linux/if_tap.h>
#include <linux/if_vlan.h>
#include <linux/interrupt.h>
#include <linux/nsproxy.h>
@@ -62,7 +62,7 @@ static int macvtap_newlink(struct net *src_net,
*/
vlan->tap_features = TUN_OFFLOADS;
- err = netdev_rx_handler_register(dev, macvtap_handle_frame, vlan);
+ err = netdev_rx_handler_register(dev, tap_handle_frame, vlan);
if (err)
return err;
@@ -82,7 +82,7 @@ static void macvtap_dellink(struct net_device *dev,
struct list_head *head)
{
netdev_rx_handler_unregister(dev);
- macvtap_del_queues(dev);
+ tap_del_queues(dev);
macvlan_dellink(dev, head);
}
@@ -121,7 +121,7 @@ static int macvtap_device_event(struct notifier_block *unused,
* been registered but before register_netdevice has
* finished running.
*/
- err = macvtap_get_minor(vlan);
+ err = tap_get_minor(vlan);
if (err)
return notifier_from_errno(err);
@@ -129,7 +129,7 @@ static int macvtap_device_event(struct notifier_block *unused,
classdev = device_create(&macvtap_class, &dev->dev, devt,
dev, tap_name);
if (IS_ERR(classdev)) {
- macvtap_free_minor(vlan);
+ tap_free_minor(vlan);
return notifier_from_errno(PTR_ERR(classdev));
}
err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
@@ -144,10 +144,10 @@ static int macvtap_device_event(struct notifier_block *unused,
sysfs_remove_link(&dev->dev.kobj, tap_name);
devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
device_destroy(&macvtap_class, devt);
- macvtap_free_minor(vlan);
+ tap_free_minor(vlan);
break;
case NETDEV_CHANGE_TX_QUEUE_LEN:
- if (macvtap_queue_resize(vlan))
+ if (tap_queue_resize(vlan))
return NOTIFY_BAD;
break;
}
@@ -159,7 +159,7 @@ static struct notifier_block macvtap_notifier_block __read_mostly = {
.notifier_call = macvtap_device_event,
};
-extern struct file_operations macvtap_fops;
+extern struct file_operations tap_fops;
static int macvtap_init(void)
{
int err;
@@ -169,7 +169,7 @@ static int macvtap_init(void)
if (err)
goto out1;
- cdev_init(&macvtap_cdev, &macvtap_fops);
+ cdev_init(&macvtap_cdev, &tap_fops);
err = cdev_add(&macvtap_cdev, macvtap_major, MACVTAP_NUM_DEVS);
if (err)
goto out2;
diff --git a/drivers/net/tap.c b/drivers/net/tap.c
index 8f12a39..d0807c2 100644
--- a/drivers/net/tap.c
+++ b/drivers/net/tap.c
@@ -24,16 +24,16 @@
#include <linux/skb_array.h>
/*
- * A macvtap queue is the central object of this driver, it connects
+ * A tap queue is the central object of this driver, it connects
* an open character device to a macvlan interface. There can be
* multiple queues on one interface, which map back to queues
* implemented in hardware on the underlying device.
*
- * macvtap_proto is used to allocate queues through the sock allocation
+ * tap_proto is used to allocate queues through the sock allocation
* mechanism.
*
*/
-struct macvtap_queue {
+struct tap_queue {
struct sock sk;
struct socket sock;
struct socket_wq wq;
@@ -47,21 +47,21 @@ struct macvtap_queue {
struct skb_array skb_array;
};
-#define MACVTAP_FEATURES (IFF_VNET_HDR | IFF_MULTI_QUEUE)
+#define TAP_IFFEATURES (IFF_VNET_HDR | IFF_MULTI_QUEUE)
-#define MACVTAP_VNET_LE 0x80000000
-#define MACVTAP_VNET_BE 0x40000000
+#define TAP_VNET_LE 0x80000000
+#define TAP_VNET_BE 0x40000000
#ifdef CONFIG_TUN_VNET_CROSS_LE
-static inline bool macvtap_legacy_is_little_endian(struct macvtap_queue *q)
+static inline bool tap_legacy_is_little_endian(struct tap_queue *q)
{
- return q->flags & MACVTAP_VNET_BE ? false :
+ return q->flags & TAP_VNET_BE ? false :
virtio_legacy_is_little_endian();
}
-static long macvtap_get_vnet_be(struct macvtap_queue *q, int __user *sp)
+static long tap_get_vnet_be(struct tap_queue *q, int __user *sp)
{
- int s = !!(q->flags & MACVTAP_VNET_BE);
+ int s = !!(q->flags & TAP_VNET_BE);
if (put_user(s, sp))
return -EFAULT;
@@ -69,7 +69,7 @@ static long macvtap_get_vnet_be(struct macvtap_queue *q, int __user *sp)
return 0;
}
-static long macvtap_set_vnet_be(struct macvtap_queue *q, int __user *sp)
+static long tap_set_vnet_be(struct tap_queue *q, int __user *sp)
{
int s;
@@ -77,77 +77,77 @@ static long macvtap_set_vnet_be(struct macvtap_queue *q, int __user *sp)
return -EFAULT;
if (s)
- q->flags |= MACVTAP_VNET_BE;
+ q->flags |= TAP_VNET_BE;
else
- q->flags &= ~MACVTAP_VNET_BE;
+ q->flags &= ~TAP_VNET_BE;
return 0;
}
#else
-static inline bool macvtap_legacy_is_little_endian(struct macvtap_queue *q)
+static inline bool tap_legacy_is_little_endian(struct tap_queue *q)
{
return virtio_legacy_is_little_endian();
}
-static long macvtap_get_vnet_be(struct macvtap_queue *q, int __user *argp)
+static long tap_get_vnet_be(struct tap_queue *q, int __user *argp)
{
return -EINVAL;
}
-static long macvtap_set_vnet_be(struct macvtap_queue *q, int __user *argp)
+static long tap_set_vnet_be(struct tap_queue *q, int __user *argp)
{
return -EINVAL;
}
#endif /* CONFIG_TUN_VNET_CROSS_LE */
-static inline bool macvtap_is_little_endian(struct macvtap_queue *q)
+static inline bool tap_is_little_endian(struct tap_queue *q)
{
- return q->flags & MACVTAP_VNET_LE ||
- macvtap_legacy_is_little_endian(q);
+ return q->flags & TAP_VNET_LE ||
+ tap_legacy_is_little_endian(q);
}
-static inline u16 macvtap16_to_cpu(struct macvtap_queue *q, __virtio16 val)
+static inline u16 tap16_to_cpu(struct tap_queue *q, __virtio16 val)
{
- return __virtio16_to_cpu(macvtap_is_little_endian(q), val);
+ return __virtio16_to_cpu(tap_is_little_endian(q), val);
}
-static inline __virtio16 cpu_to_macvtap16(struct macvtap_queue *q, u16 val)
+static inline __virtio16 cpu_to_tap16(struct tap_queue *q, u16 val)
{
- return __cpu_to_virtio16(macvtap_is_little_endian(q), val);
+ return __cpu_to_virtio16(tap_is_little_endian(q), val);
}
-static struct proto macvtap_proto = {
- .name = "macvtap",
+static struct proto tap_proto = {
+ .name = "tap",
.owner = THIS_MODULE,
- .obj_size = sizeof (struct macvtap_queue),
+ .obj_size = sizeof(struct tap_queue),
};
-#define MACVTAP_NUM_DEVS (1U << MINORBITS)
+#define TAP_NUM_DEVS (1U << MINORBITS)
static DEFINE_MUTEX(minor_lock);
DEFINE_IDR(minor_idr);
#define GOODCOPY_LEN 128
-static const struct proto_ops macvtap_socket_ops;
+static const struct proto_ops tap_socket_ops;
#define RX_OFFLOADS (NETIF_F_GRO | NETIF_F_LRO)
#define TAP_FEATURES (NETIF_F_GSO | NETIF_F_SG | NETIF_F_FRAGLIST)
-static struct macvlan_dev *macvtap_get_vlan_rcu(const struct net_device *dev)
+static struct macvlan_dev *tap_get_vlan_rcu(const struct net_device *dev)
{
return rcu_dereference(dev->rx_handler_data);
}
/*
* RCU usage:
- * The macvtap_queue and the macvlan_dev are loosely coupled, the
+ * The tap_queue and the macvlan_dev are loosely coupled, the
* pointers from one to the other can only be read while rcu_read_lock
* or rtnl is held.
*
- * Both the file and the macvlan_dev hold a reference on the macvtap_queue
+ * Both the file and the macvlan_dev hold a reference on the tap_queue
* through sock_hold(&q->sk). When the macvlan_dev goes away first,
* q->vlan becomes inaccessible. When the files gets closed,
- * macvtap_get_queue() fails.
+ * tap_get_queue() fails.
*
* There may still be references to the struct sock inside of the
* queue from outbound SKBs, but these never reference back to the
@@ -155,8 +155,8 @@ static struct macvlan_dev *macvtap_get_vlan_rcu(const struct net_device *dev)
* when both our references and any pending SKBs are gone.
*/
-static int macvtap_enable_queue(struct net_device *dev, struct file *file,
- struct macvtap_queue *q)
+static int tap_enable_queue(struct net_device *dev, struct file *file,
+ struct tap_queue *q)
{
struct macvlan_dev *vlan = netdev_priv(dev);
int err = -EINVAL;
@@ -177,12 +177,12 @@ static int macvtap_enable_queue(struct net_device *dev, struct file *file,
}
/* Requires RTNL */
-static int macvtap_set_queue(struct net_device *dev, struct file *file,
- struct macvtap_queue *q)
+static int tap_set_queue(struct net_device *dev, struct file *file,
+ struct tap_queue *q)
{
struct macvlan_dev *vlan = netdev_priv(dev);
- if (vlan->numqueues == MAX_MACVTAP_QUEUES)
+ if (vlan->numqueues == MAX_TAP_QUEUES)
return -EBUSY;
rcu_assign_pointer(q->vlan, vlan);
@@ -201,10 +201,10 @@ static int macvtap_set_queue(struct net_device *dev, struct file *file,
return 0;
}
-static int macvtap_disable_queue(struct macvtap_queue *q)
+static int tap_disable_queue(struct tap_queue *q)
{
struct macvlan_dev *vlan;
- struct macvtap_queue *nq;
+ struct tap_queue *nq;
ASSERT_RTNL();
if (!q->enabled)
@@ -236,7 +236,7 @@ static int macvtap_disable_queue(struct macvtap_queue *q)
* Using the spinlock makes sure that we don't get
* to the queue again after destroying it.
*/
-static void macvtap_put_queue(struct macvtap_queue *q)
+static void tap_put_queue(struct tap_queue *q)
{
struct macvlan_dev *vlan;
@@ -245,7 +245,7 @@ static void macvtap_put_queue(struct macvtap_queue *q)
if (vlan) {
if (q->enabled)
- BUG_ON(macvtap_disable_queue(q));
+ BUG_ON(tap_disable_queue(q));
vlan->numqueues--;
RCU_INIT_POINTER(q->vlan, NULL);
@@ -266,11 +266,11 @@ static void macvtap_put_queue(struct macvtap_queue *q)
* Cache vlan->numvtaps since it can become zero during the execution
* of this function.
*/
-static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
- struct sk_buff *skb)
+static struct tap_queue *tap_get_queue(struct net_device *dev,
+ struct sk_buff *skb)
{
struct macvlan_dev *vlan = netdev_priv(dev);
- struct macvtap_queue *tap = NULL;
+ struct tap_queue *tap = NULL;
/* Access to taps array is protected by rcu, but access to numvtaps
* isn't. Below we use it to lookup a queue, but treat it as a hint
* and validate that the result isn't NULL - in case we are
@@ -313,10 +313,10 @@ static struct macvtap_queue *macvtap_get_queue(struct net_device *dev,
* that it holds on all queues and safely set the pointer
* from the queues to NULL.
*/
-void macvtap_del_queues(struct net_device *dev)
+void tap_del_queues(struct net_device *dev)
{
struct macvlan_dev *vlan = netdev_priv(dev);
- struct macvtap_queue *q, *tmp;
+ struct tap_queue *q, *tmp;
ASSERT_RTNL();
list_for_each_entry_safe(q, tmp, &vlan->queue_list, next) {
@@ -329,23 +329,23 @@ void macvtap_del_queues(struct net_device *dev)
}
BUG_ON(vlan->numvtaps);
BUG_ON(vlan->numqueues);
- /* guarantee that any future macvtap_set_queue will fail */
- vlan->numvtaps = MAX_MACVTAP_QUEUES;
+ /* guarantee that any future tap_set_queue will fail */
+ vlan->numvtaps = MAX_TAP_QUEUES;
}
-rx_handler_result_t macvtap_handle_frame(struct sk_buff **pskb)
+rx_handler_result_t tap_handle_frame(struct sk_buff **pskb)
{
struct sk_buff *skb = *pskb;
struct net_device *dev = skb->dev;
struct macvlan_dev *vlan;
- struct macvtap_queue *q;
+ struct tap_queue *q;
netdev_features_t features = TAP_FEATURES;
- vlan = macvtap_get_vlan_rcu(dev);
+ vlan = tap_get_vlan_rcu(dev);
if (!vlan)
return RX_HANDLER_PASS;
- q = macvtap_get_queue(dev, skb);
+ q = tap_get_queue(dev, skb);
if (!q)
return RX_HANDLER_PASS;
@@ -409,23 +409,23 @@ rx_handler_result_t macvtap_handle_frame(struct sk_buff **pskb)
return RX_HANDLER_CONSUMED;
}
-int macvtap_get_minor(struct macvlan_dev *vlan)
+int tap_get_minor(struct macvlan_dev *vlan)
{
int retval = -ENOMEM;
mutex_lock(&minor_lock);
- retval = idr_alloc(&minor_idr, vlan, 1, MACVTAP_NUM_DEVS, GFP_KERNEL);
+ retval = idr_alloc(&minor_idr, vlan, 1, TAP_NUM_DEVS, GFP_KERNEL);
if (retval >= 0) {
vlan->minor = retval;
} else if (retval == -ENOSPC) {
- netdev_err(vlan->dev, "Too many macvtap devices\n");
+ netdev_err(vlan->dev, "Too many tap devices\n");
retval = -EINVAL;
}
mutex_unlock(&minor_lock);
return retval < 0 ? retval : 0;
}
-void macvtap_free_minor(struct macvlan_dev *vlan)
+void tap_free_minor(struct macvlan_dev *vlan)
{
mutex_lock(&minor_lock);
if (vlan->minor) {
@@ -435,7 +435,7 @@ void macvtap_free_minor(struct macvlan_dev *vlan)
mutex_unlock(&minor_lock);
}
-static struct net_device *dev_get_by_macvtap_minor(int minor)
+static struct net_device *dev_get_by_tap_minor(int minor)
{
struct net_device *dev = NULL;
struct macvlan_dev *vlan;
@@ -450,7 +450,7 @@ static struct net_device *dev_get_by_macvtap_minor(int minor)
return dev;
}
-static void macvtap_sock_write_space(struct sock *sk)
+static void tap_sock_write_space(struct sock *sk)
{
wait_queue_head_t *wqueue;
@@ -463,28 +463,28 @@ static void macvtap_sock_write_space(struct sock *sk)
wake_up_interruptible_poll(wqueue, POLLOUT | POLLWRNORM | POLLWRBAND);
}
-static void macvtap_sock_destruct(struct sock *sk)
+static void tap_sock_destruct(struct sock *sk)
{
- struct macvtap_queue *q = container_of(sk, struct macvtap_queue, sk);
+ struct tap_queue *q = container_of(sk, struct tap_queue, sk);
skb_array_cleanup(&q->skb_array);
}
-static int macvtap_open(struct inode *inode, struct file *file)
+static int tap_open(struct inode *inode, struct file *file)
{
struct net *net = current->nsproxy->net_ns;
struct net_device *dev;
- struct macvtap_queue *q;
+ struct tap_queue *q;
int err = -ENODEV;
rtnl_lock();
- dev = dev_get_by_macvtap_minor(iminor(inode));
+ dev = dev_get_by_tap_minor(iminor(inode));
if (!dev)
goto err;
err = -ENOMEM;
- q = (struct macvtap_queue *)sk_alloc(net, AF_UNSPEC, GFP_KERNEL,
- &macvtap_proto, 0);
+ q = (struct tap_queue *)sk_alloc(net, AF_UNSPEC, GFP_KERNEL,
+ &tap_proto, 0);
if (!q)
goto err;
@@ -493,15 +493,15 @@ static int macvtap_open(struct inode *inode, struct file *file)
q->sock.type = SOCK_RAW;
q->sock.state = SS_CONNECTED;
q->sock.file = file;
- q->sock.ops = &macvtap_socket_ops;
+ q->sock.ops = &tap_socket_ops;
sock_init_data(&q->sock, &q->sk);
- q->sk.sk_write_space = macvtap_sock_write_space;
- q->sk.sk_destruct = macvtap_sock_destruct;
+ q->sk.sk_write_space = tap_sock_write_space;
+ q->sk.sk_destruct = tap_sock_destruct;
q->flags = IFF_VNET_HDR | IFF_NO_PI | IFF_TAP;
q->vnet_hdr_sz = sizeof(struct virtio_net_hdr);
/*
- * so far only KVM virtio_net uses macvtap, enable zero copy between
+ * so far only KVM virtio_net uses tap, enable zero copy between
* guest kernel and host kernel when lower device supports zerocopy
*
* The macvlan supports zerocopy iff the lower device supports zero
@@ -514,7 +514,7 @@ static int macvtap_open(struct inode *inode, struct file *file)
if (skb_array_init(&q->skb_array, dev->tx_queue_len, GFP_KERNEL))
goto err_array;
- err = macvtap_set_queue(dev, file, q);
+ err = tap_set_queue(dev, file, q);
if (err)
goto err_queue;
@@ -535,16 +535,16 @@ static int macvtap_open(struct inode *inode, struct file *file)
return err;
}
-static int macvtap_release(struct inode *inode, struct file *file)
+static int tap_release(struct inode *inode, struct file *file)
{
- struct macvtap_queue *q = file->private_data;
- macvtap_put_queue(q);
+ struct tap_queue *q = file->private_data;
+ tap_put_queue(q);
return 0;
}
-static unsigned int macvtap_poll(struct file *file, poll_table * wait)
+static unsigned int tap_poll(struct file *file, poll_table *wait)
{
- struct macvtap_queue *q = file->private_data;
+ struct tap_queue *q = file->private_data;
unsigned int mask = POLLERR;
if (!q)
@@ -565,8 +565,8 @@ static unsigned int macvtap_poll(struct file *file, poll_table * wait)
return mask;
}
-static inline struct sk_buff *macvtap_alloc_skb(struct sock *sk, size_t prepad,
- size_t len, size_t linear,
+static inline struct sk_buff *tap_alloc_skb(struct sock *sk, size_t prepad,
+ size_t len, size_t linear,
int noblock, int *err)
{
struct sk_buff *skb;
@@ -589,13 +589,13 @@ static inline struct sk_buff *macvtap_alloc_skb(struct sock *sk, size_t prepad,
}
/* Neighbour code has some assumptions on HH_DATA_MOD alignment */
-#define MACVTAP_RESERVE HH_DATA_OFF(ETH_HLEN)
+#define TAP_RESERVE HH_DATA_OFF(ETH_HLEN)
/* Get packet from user space buffer */
-static ssize_t macvtap_get_user(struct macvtap_queue *q, struct msghdr *m,
- struct iov_iter *from, int noblock)
+static ssize_t tap_get_user(struct tap_queue *q, struct msghdr *m,
+ struct iov_iter *from, int noblock)
{
- int good_linear = SKB_MAX_HEAD(MACVTAP_RESERVE);
+ int good_linear = SKB_MAX_HEAD(TAP_RESERVE);
struct sk_buff *skb;
struct macvlan_dev *vlan;
unsigned long total_len = iov_iter_count(from);
@@ -621,14 +621,14 @@ static ssize_t macvtap_get_user(struct macvtap_queue *q, struct msghdr *m,
goto err;
iov_iter_advance(from, vnet_hdr_len - sizeof(vnet_hdr));
if ((vnet_hdr.flags & VIRTIO_NET_HDR_F_NEEDS_CSUM) &&
- macvtap16_to_cpu(q, vnet_hdr.csum_start) +
- macvtap16_to_cpu(q, vnet_hdr.csum_offset) + 2 >
- macvtap16_to_cpu(q, vnet_hdr.hdr_len))
- vnet_hdr.hdr_len = cpu_to_macvtap16(q,
- macvtap16_to_cpu(q, vnet_hdr.csum_start) +
- macvtap16_to_cpu(q, vnet_hdr.csum_offset) + 2);
+ tap16_to_cpu(q, vnet_hdr.csum_start) +
+ tap16_to_cpu(q, vnet_hdr.csum_offset) + 2 >
+ tap16_to_cpu(q, vnet_hdr.hdr_len))
+ vnet_hdr.hdr_len = cpu_to_tap16(q,
+ tap16_to_cpu(q, vnet_hdr.csum_start) +
+ tap16_to_cpu(q, vnet_hdr.csum_offset) + 2);
err = -EINVAL;
- if (macvtap16_to_cpu(q, vnet_hdr.hdr_len) > len)
+ if (tap16_to_cpu(q, vnet_hdr.hdr_len) > len)
goto err;
}
@@ -640,7 +640,7 @@ static ssize_t macvtap_get_user(struct macvtap_queue *q, struct msghdr *m,
struct iov_iter i;
copylen = vnet_hdr.hdr_len ?
- macvtap16_to_cpu(q, vnet_hdr.hdr_len) : GOODCOPY_LEN;
+ tap16_to_cpu(q, vnet_hdr.hdr_len) : GOODCOPY_LEN;
if (copylen > good_linear)
copylen = good_linear;
else if (copylen < ETH_HLEN)
@@ -654,15 +654,15 @@ static ssize_t macvtap_get_user(struct macvtap_queue *q, struct msghdr *m,
if (!zerocopy) {
copylen = len;
- linear = macvtap16_to_cpu(q, vnet_hdr.hdr_len);
+ linear = tap16_to_cpu(q, vnet_hdr.hdr_len);
if (linear > good_linear)
linear = good_linear;
else if (linear < ETH_HLEN)
linear = ETH_HLEN;
}
- skb = macvtap_alloc_skb(&q->sk, MACVTAP_RESERVE, copylen,
- linear, noblock, &err);
+ skb = tap_alloc_skb(&q->sk, TAP_RESERVE, copylen,
+ linear, noblock, &err);
if (!skb)
goto err;
@@ -680,7 +680,7 @@ static ssize_t macvtap_get_user(struct macvtap_queue *q, struct msghdr *m,
if (vnet_hdr_len) {
err = virtio_net_hdr_to_skb(skb, &vnet_hdr,
- macvtap_is_little_endian(q));
+ tap_is_little_endian(q));
if (err)
goto err_kfree;
}
@@ -728,18 +728,18 @@ static ssize_t macvtap_get_user(struct macvtap_queue *q, struct msghdr *m,
return err;
}
-static ssize_t macvtap_write_iter(struct kiocb *iocb, struct iov_iter *from)
+static ssize_t tap_write_iter(struct kiocb *iocb, struct iov_iter *from)
{
struct file *file = iocb->ki_filp;
- struct macvtap_queue *q = file->private_data;
+ struct tap_queue *q = file->private_data;
- return macvtap_get_user(q, NULL, from, file->f_flags & O_NONBLOCK);
+ return tap_get_user(q, NULL, from, file->f_flags & O_NONBLOCK);
}
/* Put packet to the user space buffer */
-static ssize_t macvtap_put_user(struct macvtap_queue *q,
- const struct sk_buff *skb,
- struct iov_iter *iter)
+static ssize_t tap_put_user(struct tap_queue *q,
+ const struct sk_buff *skb,
+ struct iov_iter *iter)
{
int ret;
int vnet_hdr_len = 0;
@@ -753,7 +753,7 @@ static ssize_t macvtap_put_user(struct macvtap_queue *q,
return -EINVAL;
if (virtio_net_hdr_from_skb(skb, &vnet_hdr,
- macvtap_is_little_endian(q)))
+ tap_is_little_endian(q)))
BUG();
if (copy_to_iter(&vnet_hdr, sizeof(vnet_hdr), iter) !=
@@ -792,9 +792,9 @@ static ssize_t macvtap_put_user(struct macvtap_queue *q,
return ret ? ret : total;
}
-static ssize_t macvtap_do_read(struct macvtap_queue *q,
- struct iov_iter *to,
- int noblock)
+static ssize_t tap_do_read(struct tap_queue *q,
+ struct iov_iter *to,
+ int noblock)
{
DEFINE_WAIT(wait);
struct sk_buff *skb;
@@ -827,7 +827,7 @@ static ssize_t macvtap_do_read(struct macvtap_queue *q,
finish_wait(sk_sleep(&q->sk), &wait);
if (skb) {
- ret = macvtap_put_user(q, skb, to);
+ ret = tap_put_user(q, skb, to);
if (unlikely(ret < 0))
kfree_skb(skb);
else
@@ -836,20 +836,20 @@ static ssize_t macvtap_do_read(struct macvtap_queue *q,
return ret;
}
-static ssize_t macvtap_read_iter(struct kiocb *iocb, struct iov_iter *to)
+static ssize_t tap_read_iter(struct kiocb *iocb, struct iov_iter *to)
{
struct file *file = iocb->ki_filp;
- struct macvtap_queue *q = file->private_data;
+ struct tap_queue *q = file->private_data;
ssize_t len = iov_iter_count(to), ret;
- ret = macvtap_do_read(q, to, file->f_flags & O_NONBLOCK);
+ ret = tap_do_read(q, to, file->f_flags & O_NONBLOCK);
ret = min_t(ssize_t, ret, len);
if (ret > 0)
iocb->ki_pos = ret;
return ret;
}
-static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
+static struct macvlan_dev *tap_get_vlan(struct tap_queue *q)
{
struct macvlan_dev *vlan;
@@ -861,33 +861,33 @@ static struct macvlan_dev *macvtap_get_vlan(struct macvtap_queue *q)
return vlan;
}
-static void macvtap_put_vlan(struct macvlan_dev *vlan)
+static void tap_put_vlan(struct macvlan_dev *vlan)
{
dev_put(vlan->dev);
}
-static int macvtap_ioctl_set_queue(struct file *file, unsigned int flags)
+static int tap_ioctl_set_queue(struct file *file, unsigned int flags)
{
- struct macvtap_queue *q = file->private_data;
+ struct tap_queue *q = file->private_data;
struct macvlan_dev *vlan;
int ret;
- vlan = macvtap_get_vlan(q);
+ vlan = tap_get_vlan(q);
if (!vlan)
return -EINVAL;
if (flags & IFF_ATTACH_QUEUE)
- ret = macvtap_enable_queue(vlan->dev, file, q);
+ ret = tap_enable_queue(vlan->dev, file, q);
else if (flags & IFF_DETACH_QUEUE)
- ret = macvtap_disable_queue(q);
+ ret = tap_disable_queue(q);
else
ret = -EINVAL;
- macvtap_put_vlan(vlan);
+ tap_put_vlan(vlan);
return ret;
}
-static int set_offload(struct macvtap_queue *q, unsigned long arg)
+static int set_offload(struct tap_queue *q, unsigned long arg)
{
struct macvlan_dev *vlan;
netdev_features_t features;
@@ -919,7 +919,7 @@ static int set_offload(struct macvtap_queue *q, unsigned long arg)
* setting the TSO bit means that the userspace wants to
* accept TSO frames and turning it off means that user space
* does not support TSO.
- * For macvtap, we have to invert it to mean the same thing.
+ * For tap, we have to invert it to mean the same thing.
* When user space turns off TSO, we turn off GSO/LRO so that
* user-space will not receive TSO frames.
*/
@@ -941,10 +941,10 @@ static int set_offload(struct macvtap_queue *q, unsigned long arg)
/*
* provide compatibility with generic tun/tap interface
*/
-static long macvtap_ioctl(struct file *file, unsigned int cmd,
- unsigned long arg)
+static long tap_ioctl(struct file *file, unsigned int cmd,
+ unsigned long arg)
{
- struct macvtap_queue *q = file->private_data;
+ struct tap_queue *q = file->private_data;
struct macvlan_dev *vlan;
void __user *argp = (void __user *)arg;
struct ifreq __user *ifr = argp;
@@ -962,16 +962,16 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
return -EFAULT;
ret = 0;
- if ((u & ~MACVTAP_FEATURES) != (IFF_NO_PI | IFF_TAP))
+ if ((u & ~TAP_IFFEATURES) != (IFF_NO_PI | IFF_TAP))
ret = -EINVAL;
else
- q->flags = (q->flags & ~MACVTAP_FEATURES) | u;
+ q->flags = (q->flags & ~TAP_IFFEATURES) | u;
return ret;
case TUNGETIFF:
rtnl_lock();
- vlan = macvtap_get_vlan(q);
+ vlan = tap_get_vlan(q);
if (!vlan) {
rtnl_unlock();
return -ENOLINK;
@@ -982,7 +982,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
put_user(u, &ifr->ifr_flags))
ret = -EFAULT;
- macvtap_put_vlan(vlan);
+ tap_put_vlan(vlan);
rtnl_unlock();
return ret;
@@ -990,12 +990,12 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
if (get_user(u, &ifr->ifr_flags))
return -EFAULT;
rtnl_lock();
- ret = macvtap_ioctl_set_queue(file, u);
+ ret = tap_ioctl_set_queue(file, u);
rtnl_unlock();
return ret;
case TUNGETFEATURES:
- if (put_user(IFF_TAP | IFF_NO_PI | MACVTAP_FEATURES, up))
+ if (put_user(IFF_TAP | IFF_NO_PI | TAP_IFFEATURES, up))
return -EFAULT;
return 0;
@@ -1022,7 +1022,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
return 0;
case TUNGETVNETLE:
- s = !!(q->flags & MACVTAP_VNET_LE);
+ s = !!(q->flags & TAP_VNET_LE);
if (put_user(s, sp))
return -EFAULT;
return 0;
@@ -1031,16 +1031,16 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
if (get_user(s, sp))
return -EFAULT;
if (s)
- q->flags |= MACVTAP_VNET_LE;
+ q->flags |= TAP_VNET_LE;
else
- q->flags &= ~MACVTAP_VNET_LE;
+ q->flags &= ~TAP_VNET_LE;
return 0;
case TUNGETVNETBE:
- return macvtap_get_vnet_be(q, sp);
+ return tap_get_vnet_be(q, sp);
case TUNSETVNETBE:
- return macvtap_set_vnet_be(q, sp);
+ return tap_set_vnet_be(q, sp);
case TUNSETOFFLOAD:
/* let the user check for future flags */
@@ -1055,7 +1055,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
case SIOCGIFHWADDR:
rtnl_lock();
- vlan = macvtap_get_vlan(q);
+ vlan = tap_get_vlan(q);
if (!vlan) {
rtnl_unlock();
return -ENOLINK;
@@ -1066,7 +1066,7 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
copy_to_user(&ifr->ifr_hwaddr.sa_data, vlan->dev->dev_addr, ETH_ALEN) ||
put_user(u, &ifr->ifr_hwaddr.sa_family))
ret = -EFAULT;
- macvtap_put_vlan(vlan);
+ tap_put_vlan(vlan);
rtnl_unlock();
return ret;
@@ -1074,13 +1074,13 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
if (copy_from_user(&sa, &ifr->ifr_hwaddr, sizeof(sa)))
return -EFAULT;
rtnl_lock();
- vlan = macvtap_get_vlan(q);
+ vlan = tap_get_vlan(q);
if (!vlan) {
rtnl_unlock();
return -ENOLINK;
}
ret = dev_set_mac_address(vlan->dev, &sa);
- macvtap_put_vlan(vlan);
+ tap_put_vlan(vlan);
rtnl_unlock();
return ret;
@@ -1090,42 +1090,42 @@ static long macvtap_ioctl(struct file *file, unsigned int cmd,
}
#ifdef CONFIG_COMPAT
-static long macvtap_compat_ioctl(struct file *file, unsigned int cmd,
- unsigned long arg)
+static long tap_compat_ioctl(struct file *file, unsigned int cmd,
+ unsigned long arg)
{
- return macvtap_ioctl(file, cmd, (unsigned long)compat_ptr(arg));
+ return tap_ioctl(file, cmd, (unsigned long)compat_ptr(arg));
}
#endif
-const struct file_operations macvtap_fops = {
+const struct file_operations tap_fops = {
.owner = THIS_MODULE,
- .open = macvtap_open,
- .release = macvtap_release,
- .read_iter = macvtap_read_iter,
- .write_iter = macvtap_write_iter,
- .poll = macvtap_poll,
+ .open = tap_open,
+ .release = tap_release,
+ .read_iter = tap_read_iter,
+ .write_iter = tap_write_iter,
+ .poll = tap_poll,
.llseek = no_llseek,
- .unlocked_ioctl = macvtap_ioctl,
+ .unlocked_ioctl = tap_ioctl,
#ifdef CONFIG_COMPAT
- .compat_ioctl = macvtap_compat_ioctl,
+ .compat_ioctl = tap_compat_ioctl,
#endif
};
-static int macvtap_sendmsg(struct socket *sock, struct msghdr *m,
- size_t total_len)
+static int tap_sendmsg(struct socket *sock, struct msghdr *m,
+ size_t total_len)
{
- struct macvtap_queue *q = container_of(sock, struct macvtap_queue, sock);
- return macvtap_get_user(q, m, &m->msg_iter, m->msg_flags & MSG_DONTWAIT);
+ struct tap_queue *q = container_of(sock, struct tap_queue, sock);
+ return tap_get_user(q, m, &m->msg_iter, m->msg_flags & MSG_DONTWAIT);
}
-static int macvtap_recvmsg(struct socket *sock, struct msghdr *m,
- size_t total_len, int flags)
+static int tap_recvmsg(struct socket *sock, struct msghdr *m,
+ size_t total_len, int flags)
{
- struct macvtap_queue *q = container_of(sock, struct macvtap_queue, sock);
+ struct tap_queue *q = container_of(sock, struct tap_queue, sock);
int ret;
if (flags & ~(MSG_DONTWAIT|MSG_TRUNC))
return -EINVAL;
- ret = macvtap_do_read(q, &m->msg_iter, flags & MSG_DONTWAIT);
+ ret = tap_do_read(q, &m->msg_iter, flags & MSG_DONTWAIT);
if (ret > total_len) {
m->msg_flags |= MSG_TRUNC;
ret = flags & MSG_TRUNC ? ret : total_len;
@@ -1133,40 +1133,40 @@ static int macvtap_recvmsg(struct socket *sock, struct msghdr *m,
return ret;
}
-static int macvtap_peek_len(struct socket *sock)
+static int tap_peek_len(struct socket *sock)
{
- struct macvtap_queue *q = container_of(sock, struct macvtap_queue,
+ struct tap_queue *q = container_of(sock, struct tap_queue,
sock);
return skb_array_peek_len(&q->skb_array);
}
/* Ops structure to mimic raw sockets with tun */
-static const struct proto_ops macvtap_socket_ops = {
- .sendmsg = macvtap_sendmsg,
- .recvmsg = macvtap_recvmsg,
- .peek_len = macvtap_peek_len,
+static const struct proto_ops tap_socket_ops = {
+ .sendmsg = tap_sendmsg,
+ .recvmsg = tap_recvmsg,
+ .peek_len = tap_peek_len,
};
/* Get an underlying socket object from tun file. Returns error unless file is
* attached to a device. The returned object works like a packet socket, it
* can be used for sock_sendmsg/sock_recvmsg. The caller is responsible for
* holding a reference to the file for as long as the socket is in use. */
-struct socket *macvtap_get_socket(struct file *file)
+struct socket *tap_get_socket(struct file *file)
{
- struct macvtap_queue *q;
- if (file->f_op != &macvtap_fops)
+ struct tap_queue *q;
+ if (file->f_op != &tap_fops)
return ERR_PTR(-EINVAL);
q = file->private_data;
if (!q)
return ERR_PTR(-EBADFD);
return &q->sock;
}
-EXPORT_SYMBOL_GPL(macvtap_get_socket);
+EXPORT_SYMBOL_GPL(tap_get_socket);
-int macvtap_queue_resize(struct macvlan_dev *vlan)
+int tap_queue_resize(struct macvlan_dev *vlan)
{
struct net_device *dev = vlan->dev;
- struct macvtap_queue *q;
+ struct tap_queue *q;
struct skb_array **arrays;
int n = vlan->numqueues;
int ret, i = 0;
diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 5dc3465..4875efd 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -24,6 +24,7 @@
#include <linux/if_arp.h>
#include <linux/if_tun.h>
#include <linux/if_macvlan.h>
+#include <linux/if_tap.h>
#include <linux/if_vlan.h>
#include <net/sock.h>
@@ -943,7 +944,7 @@ static struct socket *get_tap_socket(int fd)
sock = tun_get_socket(file);
if (!IS_ERR(sock))
return sock;
- sock = macvtap_get_socket(file);
+ sock = tap_get_socket(file);
if (IS_ERR(sock))
fput(file);
return sock;
diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
index a4ccc31..5233898 100644
--- a/include/linux/if_macvlan.h
+++ b/include/linux/if_macvlan.h
@@ -29,7 +29,7 @@ struct macvtap_queue;
* Maximum times a macvtap device can be opened. This can be used to
* configure the number of receive queue, e.g. for multiqueue virtio.
*/
-#define MAX_MACVTAP_QUEUES 256
+#define MAX_TAP_QUEUES 256
#define MACVLAN_MC_FILTER_BITS 8
#define MACVLAN_MC_FILTER_SZ (1 << MACVLAN_MC_FILTER_BITS)
@@ -49,7 +49,7 @@ struct macvlan_dev {
enum macvlan_mode mode;
u16 flags;
/* This array tracks active taps. */
- struct macvtap_queue __rcu *taps[MAX_MACVTAP_QUEUES];
+ struct tap_queue __rcu *taps[MAX_TAP_QUEUES];
/* This list tracks all taps (both enabled and disabled) */
struct list_head queue_list;
int numvtaps;
diff --git a/include/linux/if_macvtap.h b/include/linux/if_macvtap.h
deleted file mode 100644
index c9bf84b..0000000
--- a/include/linux/if_macvtap.h
+++ /dev/null
@@ -1,10 +0,0 @@
-#ifndef _LINUX_IF_MACVTAP_H_
-#define _LINUX_IF_MACVTAP_H_
-
-rx_handler_result_t macvtap_handle_frame(struct sk_buff **pskb);
-void macvtap_del_queues(struct net_device *dev);
-int macvtap_get_minor(struct macvlan_dev *vlan);
-void macvtap_free_minor(struct macvlan_dev *vlan);
-int macvtap_queue_resize(struct macvlan_dev *vlan);
-
-#endif /*_LINUX_IF_MACVTAP_H_*/
diff --git a/include/linux/if_tap.h b/include/linux/if_tap.h
new file mode 100644
index 0000000..d9ecc15
--- /dev/null
+++ b/include/linux/if_tap.h
@@ -0,0 +1,11 @@
+#ifndef _LINUX_IF_TAP_H_
+#define _LINUX_IF_TAP_H_
+
+rx_handler_result_t tap_handle_frame(struct sk_buff **pskb);
+void tap_del_queues(struct net_device *dev);
+int tap_get_minor(struct macvlan_dev *vlan);
+void tap_free_minor(struct macvlan_dev *vlan);
+int tap_queue_resize(struct macvlan_dev *vlan);
+struct socket *tap_get_socket(struct file *file);
+
+#endif /*_LINUX_IF_TAP_H_*/
--
2.7.4
^ permalink raw reply related [flat|nested] 21+ messages in thread* Re: [PATCHv1 2/7] TAP: Renaming tap related APIs, data structures, macros
2017-01-06 22:33 ` [PATCHv1 2/7] TAP: Renaming tap related APIs, data structures, macros Sainath Grandhi
@ 2017-01-07 10:44 ` kbuild test robot
2017-01-07 12:00 ` kbuild test robot
1 sibling, 0 replies; 21+ messages in thread
From: kbuild test robot @ 2017-01-07 10:44 UTC (permalink / raw)
To: Sainath Grandhi
Cc: kbuild-all, netdev, davem, mahesh, linux-kernel, Sainath Grandhi
[-- Attachment #1: Type: text/plain, Size: 1024 bytes --]
Hi Sainath,
[auto build test ERROR on net/master]
[also build test ERROR on v4.10-rc2 next-20170106]
[cannot apply to net-next/master]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]
url: https://github.com/0day-ci/linux/commits/Sainath-Grandhi/Refactor-macvtap-to-re-use-tap-functionality-by-other-virtual-intefaces/20170107-165707
config: powerpc-defconfig (attached as .config)
compiler: powerpc64-linux-gnu-gcc (Debian 6.1.1-9) 6.1.1 20160705
reproduce:
wget https://git.kernel.org/cgit/linux/kernel/git/wfg/lkp-tests.git/plain/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# save the attached .config to linux build tree
make.cross ARCH=powerpc
All errors (new ones prefixed by >>):
>> ERROR: ".tap_get_socket" [drivers/vhost/vhost_net.ko] undefined!
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 22529 bytes --]
^ permalink raw reply [flat|nested] 21+ messages in thread* Re: [PATCHv1 2/7] TAP: Renaming tap related APIs, data structures, macros
2017-01-06 22:33 ` [PATCHv1 2/7] TAP: Renaming tap related APIs, data structures, macros Sainath Grandhi
2017-01-07 10:44 ` kbuild test robot
@ 2017-01-07 12:00 ` kbuild test robot
1 sibling, 0 replies; 21+ messages in thread
From: kbuild test robot @ 2017-01-07 12:00 UTC (permalink / raw)
To: Sainath Grandhi
Cc: kbuild-all, netdev, davem, mahesh, linux-kernel, Sainath Grandhi
[-- Attachment #1: Type: text/plain, Size: 900 bytes --]
Hi Sainath,
[auto build test ERROR on net/master]
[also build test ERROR on v4.10-rc2 next-20170106]
[cannot apply to net-next/master]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]
url: https://github.com/0day-ci/linux/commits/Sainath-Grandhi/Refactor-macvtap-to-re-use-tap-functionality-by-other-virtual-intefaces/20170107-165707
config: x86_64-acpi-redef (attached as .config)
compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
reproduce:
# save the attached .config to linux build tree
make ARCH=x86_64
All errors (new ones prefixed by >>):
drivers/built-in.o: In function `vhost_net_ioctl':
>> net.c:(.text+0x68e0b1): undefined reference to `tap_get_socket'
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 28924 bytes --]
^ permalink raw reply [flat|nested] 21+ messages in thread
* [PATCHv1 3/7] TAP: Tap character device creation/destroy API
2017-01-06 22:33 [PATCHv1 0/7] Refactor macvtap to re-use tap functionality by other virtual intefaces Sainath Grandhi
2017-01-06 22:33 ` [PATCHv1 1/7] TAP: Refactoring macvtap.c Sainath Grandhi
2017-01-06 22:33 ` [PATCHv1 2/7] TAP: Renaming tap related APIs, data structures, macros Sainath Grandhi
@ 2017-01-06 22:33 ` Sainath Grandhi
2017-01-06 22:33 ` [PATCHv1 4/7] TAP: Abstract type of virtual interface from tap implementation Sainath Grandhi
` (3 subsequent siblings)
6 siblings, 0 replies; 21+ messages in thread
From: Sainath Grandhi @ 2017-01-06 22:33 UTC (permalink / raw)
To: netdev; +Cc: davem, mahesh, linux-kernel, Sainath Grandhi
This patch provides tap device create/destroy APIs in tap.c.
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
---
drivers/net/macvtap_main.c | 29 +++++++--------------
drivers/net/tap.c | 64 ++++++++++++++++++++++++++++++++++++++--------
include/linux/if_tap.h | 3 +++
3 files changed, 65 insertions(+), 31 deletions(-)
diff --git a/drivers/net/macvtap_main.c b/drivers/net/macvtap_main.c
index 548f339..32ad560 100644
--- a/drivers/net/macvtap_main.c
+++ b/drivers/net/macvtap_main.c
@@ -28,7 +28,6 @@
* Variables for dealing with macvtaps device numbers.
*/
static dev_t macvtap_major;
-#define MACVTAP_NUM_DEVS (1U << MINORBITS)
static const void *macvtap_net_namespace(struct device *d)
{
@@ -159,43 +158,35 @@ static struct notifier_block macvtap_notifier_block __read_mostly = {
.notifier_call = macvtap_device_event,
};
-extern struct file_operations tap_fops;
static int macvtap_init(void)
{
int err;
- err = alloc_chrdev_region(&macvtap_major, 0,
- MACVTAP_NUM_DEVS, "macvtap");
- if (err)
- goto out1;
+ err = tap_create_cdev(&macvtap_cdev, &macvtap_major, "macvtap");
- cdev_init(&macvtap_cdev, &tap_fops);
- err = cdev_add(&macvtap_cdev, macvtap_major, MACVTAP_NUM_DEVS);
if (err)
- goto out2;
+ goto out1;
err = class_register(&macvtap_class);
if (err)
- goto out3;
+ goto out2;
err = register_netdevice_notifier(&macvtap_notifier_block);
if (err)
- goto out4;
+ goto out3;
err = macvlan_link_register(&macvtap_link_ops);
if (err)
- goto out5;
+ goto out4;
return 0;
-out5:
- unregister_netdevice_notifier(&macvtap_notifier_block);
out4:
- class_unregister(&macvtap_class);
+ unregister_netdevice_notifier(&macvtap_notifier_block);
out3:
- cdev_del(&macvtap_cdev);
+ class_unregister(&macvtap_class);
out2:
- unregister_chrdev_region(macvtap_major, MACVTAP_NUM_DEVS);
+ cdev_del(&macvtap_cdev);
out1:
return err;
}
@@ -207,9 +198,7 @@ static void macvtap_exit(void)
rtnl_link_unregister(&macvtap_link_ops);
unregister_netdevice_notifier(&macvtap_notifier_block);
class_unregister(&macvtap_class);
- cdev_del(&macvtap_cdev);
- unregister_chrdev_region(macvtap_major, MACVTAP_NUM_DEVS);
- idr_destroy(&minor_idr);
+ tap_destroy_cdev(macvtap_major, &macvtap_cdev);
}
module_exit(macvtap_exit);
diff --git a/drivers/net/tap.c b/drivers/net/tap.c
index d0807c2..52692d2 100644
--- a/drivers/net/tap.c
+++ b/drivers/net/tap.c
@@ -123,8 +123,12 @@ static struct proto tap_proto = {
};
#define TAP_NUM_DEVS (1U << MINORBITS)
-static DEFINE_MUTEX(minor_lock);
-DEFINE_IDR(minor_idr);
+struct major_info {
+ dev_t major;
+ struct idr minor_idr;
+ struct mutex minor_lock;
+ const char *device_name;
+} macvtap_major;
#define GOODCOPY_LEN 128
@@ -413,26 +417,26 @@ int tap_get_minor(struct macvlan_dev *vlan)
{
int retval = -ENOMEM;
- mutex_lock(&minor_lock);
- retval = idr_alloc(&minor_idr, vlan, 1, TAP_NUM_DEVS, GFP_KERNEL);
+ mutex_lock(&macvtap_major.minor_lock);
+ retval = idr_alloc(&macvtap_major.minor_idr, vlan, 1, TAP_NUM_DEVS, GFP_KERNEL);
if (retval >= 0) {
vlan->minor = retval;
} else if (retval == -ENOSPC) {
netdev_err(vlan->dev, "Too many tap devices\n");
retval = -EINVAL;
}
- mutex_unlock(&minor_lock);
+ mutex_unlock(&macvtap_major.minor_lock);
return retval < 0 ? retval : 0;
}
void tap_free_minor(struct macvlan_dev *vlan)
{
- mutex_lock(&minor_lock);
+ mutex_lock(&macvtap_major.minor_lock);
if (vlan->minor) {
- idr_remove(&minor_idr, vlan->minor);
+ idr_remove(&macvtap_major.minor_idr, vlan->minor);
vlan->minor = 0;
}
- mutex_unlock(&minor_lock);
+ mutex_unlock(&macvtap_major.minor_lock);
}
static struct net_device *dev_get_by_tap_minor(int minor)
@@ -440,13 +444,13 @@ static struct net_device *dev_get_by_tap_minor(int minor)
struct net_device *dev = NULL;
struct macvlan_dev *vlan;
- mutex_lock(&minor_lock);
- vlan = idr_find(&minor_idr, minor);
+ mutex_lock(&macvtap_major.minor_lock);
+ vlan = idr_find(&macvtap_major.minor_idr, minor);
if (vlan) {
dev = vlan->dev;
dev_hold(dev);
}
- mutex_unlock(&minor_lock);
+ mutex_unlock(&macvtap_major.minor_lock);
return dev;
}
@@ -1184,3 +1188,41 @@ int tap_queue_resize(struct macvlan_dev *vlan)
kfree(arrays);
return ret;
}
+
+int tap_create_cdev(struct cdev *tap_cdev,
+ dev_t *tap_major, const char *device_name)
+{
+ int err;
+
+ err = alloc_chrdev_region(tap_major, 0, TAP_NUM_DEVS, device_name);
+
+ if (err)
+ goto out1;
+
+ cdev_init(tap_cdev, &tap_fops);
+ err = cdev_add(tap_cdev, *tap_major, TAP_NUM_DEVS);
+ if (err)
+ goto out2;
+
+ macvtap_major.major = MAJOR(*tap_major);
+
+ idr_init(&macvtap_major.minor_idr);
+ mutex_init(&macvtap_major.minor_lock);
+
+ macvtap_major.device_name = device_name;
+
+ return err;
+
+out2:
+ unregister_chrdev_region(*tap_major, TAP_NUM_DEVS);
+out1:
+ return err;
+}
+
+void tap_destroy_cdev(dev_t major, struct cdev *tap_cdev)
+{
+ cdev_del(tap_cdev);
+ unregister_chrdev_region(major, TAP_NUM_DEVS);
+ idr_destroy(&macvtap_major.minor_idr);
+}
+
diff --git a/include/linux/if_tap.h b/include/linux/if_tap.h
index d9ecc15..89bcd42 100644
--- a/include/linux/if_tap.h
+++ b/include/linux/if_tap.h
@@ -6,6 +6,9 @@ void tap_del_queues(struct net_device *dev);
int tap_get_minor(struct macvlan_dev *vlan);
void tap_free_minor(struct macvlan_dev *vlan);
int tap_queue_resize(struct macvlan_dev *vlan);
+int tap_create_cdev(struct cdev *tap_cdev,
+ dev_t *tap_major, const char *device_name);
+void tap_destroy_cdev(dev_t major, struct cdev *tap_cdev);
struct socket *tap_get_socket(struct file *file);
#endif /*_LINUX_IF_TAP_H_*/
--
2.7.4
^ permalink raw reply related [flat|nested] 21+ messages in thread* [PATCHv1 4/7] TAP: Abstract type of virtual interface from tap implementation
2017-01-06 22:33 [PATCHv1 0/7] Refactor macvtap to re-use tap functionality by other virtual intefaces Sainath Grandhi
` (2 preceding siblings ...)
2017-01-06 22:33 ` [PATCHv1 3/7] TAP: Tap character device creation/destroy API Sainath Grandhi
@ 2017-01-06 22:33 ` Sainath Grandhi
2017-01-06 22:33 ` [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs Sainath Grandhi
` (2 subsequent siblings)
6 siblings, 0 replies; 21+ messages in thread
From: Sainath Grandhi @ 2017-01-06 22:33 UTC (permalink / raw)
To: netdev; +Cc: davem, mahesh, linux-kernel, Sainath Grandhi
macvlan object is re-structured to hold tap related elements in a separate
entity, tap_dev. Upon NETDEV_REGISTER device_event, tap_dev is registered with
idr and fetched again on tap_open. Few of the tap functions are modified to
accepted tap_dev as argument. tap_dev object includes callbacks to be used by
underlying virtual interface to take care of tx and rx accounting.
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
---
drivers/net/macvlan.c | 2 +-
drivers/net/macvtap_main.c | 68 +++++++++---
drivers/net/tap.c | 264 ++++++++++++++++++++-------------------------
include/linux/if_tap.h | 59 +++++++++-
4 files changed, 227 insertions(+), 166 deletions(-)
diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c
index 20b3fdf2..79383f9 100644
--- a/drivers/net/macvlan.c
+++ b/drivers/net/macvlan.c
@@ -1526,7 +1526,6 @@ static const struct nla_policy macvlan_policy[IFLA_MACVLAN_MAX + 1] = {
int macvlan_link_register(struct rtnl_link_ops *ops)
{
/* common fields */
- ops->priv_size = sizeof(struct macvlan_dev);
ops->validate = macvlan_validate;
ops->maxtype = IFLA_MACVLAN_MAX;
ops->policy = macvlan_policy;
@@ -1549,6 +1548,7 @@ static struct rtnl_link_ops macvlan_link_ops = {
.newlink = macvlan_newlink,
.dellink = macvlan_dellink,
.get_link_net = macvlan_get_link_net,
+ .priv_size = sizeof(struct macvlan_dev),
};
static int macvlan_device_event(struct notifier_block *unused,
diff --git a/drivers/net/macvtap_main.c b/drivers/net/macvtap_main.c
index 32ad560..6326a82 100644
--- a/drivers/net/macvtap_main.c
+++ b/drivers/net/macvtap_main.c
@@ -24,6 +24,11 @@
#include <linux/virtio_net.h>
#include <linux/skb_array.h>
+struct macvtap_dev {
+ struct macvlan_dev vlan;
+ struct tap_dev tap;
+};
+
/*
* Variables for dealing with macvtaps device numbers.
*/
@@ -46,22 +51,52 @@ static struct cdev macvtap_cdev;
#define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN | NETIF_F_TSO | \
NETIF_F_TSO6 | NETIF_F_UFO)
+static void macvtap_count_tx_dropped(struct tap_dev *tap)
+{
+ struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
+
+ this_cpu_inc(vlan->pcpu_stats->tx_dropped);
+}
+
+static void macvtap_count_rx_dropped(struct tap_dev *tap)
+{
+ struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
+
+ macvlan_count_rx(vlan, 0, 0, 0);
+}
+
+static void macvtap_update_features(struct tap_dev *tap,
+ netdev_features_t features)
+{
+ struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
+
+ vlan->set_features = features;
+ netdev_update_features(vlan->dev);
+}
+
static int macvtap_newlink(struct net *src_net,
struct net_device *dev,
struct nlattr *tb[],
struct nlattr *data[])
{
- struct macvlan_dev *vlan = netdev_priv(dev);
+ struct macvtap_dev *vlantap = netdev_priv(dev);
int err;
- INIT_LIST_HEAD(&vlan->queue_list);
+ INIT_LIST_HEAD(&vlantap->tap.queue_list);
/* Since macvlan supports all offloads by default, make
* tap support all offloads also.
*/
- vlan->tap_features = TUN_OFFLOADS;
+ vlantap->tap.tap_features = TUN_OFFLOADS;
- err = netdev_rx_handler_register(dev, tap_handle_frame, vlan);
+ /* Register callbacks for rx/tx drops accounting and updating
+ * net_device features
+ */
+ vlantap->tap.count_tx_dropped = macvtap_count_tx_dropped;
+ vlantap->tap.count_rx_dropped = macvtap_count_rx_dropped;
+ vlantap->tap.update_features = macvtap_update_features;
+
+ err = netdev_rx_handler_register(dev, tap_handle_frame, &vlantap->tap);
if (err)
return err;
@@ -74,14 +109,18 @@ static int macvtap_newlink(struct net *src_net,
return err;
}
+ vlantap->tap.dev = vlantap->vlan.dev;
+
return 0;
}
static void macvtap_dellink(struct net_device *dev,
struct list_head *head)
{
+ struct macvtap_dev *vlantap = netdev_priv(dev);
+
netdev_rx_handler_unregister(dev);
- tap_del_queues(dev);
+ tap_del_queues(&vlantap->tap);
macvlan_dellink(dev, head);
}
@@ -96,13 +135,14 @@ static struct rtnl_link_ops macvtap_link_ops __read_mostly = {
.setup = macvtap_setup,
.newlink = macvtap_newlink,
.dellink = macvtap_dellink,
+ .priv_size = sizeof(struct macvtap_dev),
};
static int macvtap_device_event(struct notifier_block *unused,
unsigned long event, void *ptr)
{
struct net_device *dev = netdev_notifier_info_to_dev(ptr);
- struct macvlan_dev *vlan;
+ struct macvtap_dev *vlantap;
struct device *classdev;
dev_t devt;
int err;
@@ -112,7 +152,7 @@ static int macvtap_device_event(struct notifier_block *unused,
return NOTIFY_DONE;
snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
- vlan = netdev_priv(dev);
+ vlantap = netdev_priv(dev);
switch (event) {
case NETDEV_REGISTER:
@@ -120,15 +160,15 @@ static int macvtap_device_event(struct notifier_block *unused,
* been registered but before register_netdevice has
* finished running.
*/
- err = tap_get_minor(vlan);
+ err = tap_get_minor(&vlantap->tap);
if (err)
return notifier_from_errno(err);
- devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
+ devt = MKDEV(MAJOR(macvtap_major), vlantap->tap.minor);
classdev = device_create(&macvtap_class, &dev->dev, devt,
dev, tap_name);
if (IS_ERR(classdev)) {
- tap_free_minor(vlan);
+ tap_free_minor(&vlantap->tap);
return notifier_from_errno(PTR_ERR(classdev));
}
err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
@@ -138,15 +178,15 @@ static int macvtap_device_event(struct notifier_block *unused,
break;
case NETDEV_UNREGISTER:
/* vlan->minor == 0 if NETDEV_REGISTER above failed */
- if (vlan->minor == 0)
+ if (vlantap->tap.minor == 0)
break;
sysfs_remove_link(&dev->dev.kobj, tap_name);
- devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
+ devt = MKDEV(MAJOR(macvtap_major), vlantap->tap.minor);
device_destroy(&macvtap_class, devt);
- tap_free_minor(vlan);
+ tap_free_minor(&vlantap->tap);
break;
case NETDEV_CHANGE_TX_QUEUE_LEN:
- if (tap_queue_resize(vlan))
+ if (tap_queue_resize(&vlantap->tap))
return NOTIFY_BAD;
break;
}
diff --git a/drivers/net/tap.c b/drivers/net/tap.c
index 52692d2..6306ab9 100644
--- a/drivers/net/tap.c
+++ b/drivers/net/tap.c
@@ -1,5 +1,5 @@
#include <linux/etherdevice.h>
-#include <linux/if_macvlan.h>
+#include <linux/if_tap.h>
#include <linux/if_vlan.h>
#include <linux/interrupt.h>
#include <linux/nsproxy.h>
@@ -23,30 +23,6 @@
#include <linux/virtio_net.h>
#include <linux/skb_array.h>
-/*
- * A tap queue is the central object of this driver, it connects
- * an open character device to a macvlan interface. There can be
- * multiple queues on one interface, which map back to queues
- * implemented in hardware on the underlying device.
- *
- * tap_proto is used to allocate queues through the sock allocation
- * mechanism.
- *
- */
-struct tap_queue {
- struct sock sk;
- struct socket sock;
- struct socket_wq wq;
- int vnet_hdr_sz;
- struct macvlan_dev __rcu *vlan;
- struct file *file;
- unsigned int flags;
- u16 queue_index;
- bool enabled;
- struct list_head next;
- struct skb_array skb_array;
-};
-
#define TAP_IFFEATURES (IFF_VNET_HDR | IFF_MULTI_QUEUE)
#define TAP_VNET_LE 0x80000000
@@ -137,7 +113,7 @@ static const struct proto_ops tap_socket_ops;
#define RX_OFFLOADS (NETIF_F_GRO | NETIF_F_LRO)
#define TAP_FEATURES (NETIF_F_GSO | NETIF_F_SG | NETIF_F_FRAGLIST)
-static struct macvlan_dev *tap_get_vlan_rcu(const struct net_device *dev)
+static struct tap_dev *tap_dev_get_rcu(const struct net_device *dev)
{
return rcu_dereference(dev->rx_handler_data);
}
@@ -159,10 +135,9 @@ static struct macvlan_dev *tap_get_vlan_rcu(const struct net_device *dev)
* when both our references and any pending SKBs are gone.
*/
-static int tap_enable_queue(struct net_device *dev, struct file *file,
+static int tap_enable_queue(struct tap_dev *tap, struct file *file,
struct tap_queue *q)
{
- struct macvlan_dev *vlan = netdev_priv(dev);
int err = -EINVAL;
ASSERT_RTNL();
@@ -171,62 +146,60 @@ static int tap_enable_queue(struct net_device *dev, struct file *file,
goto out;
err = 0;
- rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
- q->queue_index = vlan->numvtaps;
+ rcu_assign_pointer(tap->taps[tap->numvtaps], q);
+ q->queue_index = tap->numvtaps;
q->enabled = true;
- vlan->numvtaps++;
+ tap->numvtaps++;
out:
return err;
}
/* Requires RTNL */
-static int tap_set_queue(struct net_device *dev, struct file *file,
+static int tap_set_queue(struct tap_dev *tap, struct file *file,
struct tap_queue *q)
{
- struct macvlan_dev *vlan = netdev_priv(dev);
-
- if (vlan->numqueues == MAX_TAP_QUEUES)
+ if (tap->numqueues == MAX_TAP_QUEUES)
return -EBUSY;
- rcu_assign_pointer(q->vlan, vlan);
- rcu_assign_pointer(vlan->taps[vlan->numvtaps], q);
+ rcu_assign_pointer(q->tap, tap);
+ rcu_assign_pointer(tap->taps[tap->numvtaps], q);
sock_hold(&q->sk);
q->file = file;
- q->queue_index = vlan->numvtaps;
+ q->queue_index = tap->numvtaps;
q->enabled = true;
file->private_data = q;
- list_add_tail(&q->next, &vlan->queue_list);
+ list_add_tail(&q->next, &tap->queue_list);
- vlan->numvtaps++;
- vlan->numqueues++;
+ tap->numvtaps++;
+ tap->numqueues++;
return 0;
}
static int tap_disable_queue(struct tap_queue *q)
{
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
struct tap_queue *nq;
ASSERT_RTNL();
if (!q->enabled)
return -EINVAL;
- vlan = rtnl_dereference(q->vlan);
+ tap = rtnl_dereference(q->tap);
- if (vlan) {
+ if (tap) {
int index = q->queue_index;
- BUG_ON(index >= vlan->numvtaps);
- nq = rtnl_dereference(vlan->taps[vlan->numvtaps - 1]);
+ BUG_ON(index >= tap->numvtaps);
+ nq = rtnl_dereference(tap->taps[tap->numvtaps - 1]);
nq->queue_index = index;
- rcu_assign_pointer(vlan->taps[index], nq);
- RCU_INIT_POINTER(vlan->taps[vlan->numvtaps - 1], NULL);
+ rcu_assign_pointer(tap->taps[index], nq);
+ RCU_INIT_POINTER(tap->taps[tap->numvtaps - 1], NULL);
q->enabled = false;
- vlan->numvtaps--;
+ tap->numvtaps--;
}
return 0;
@@ -242,17 +215,17 @@ static int tap_disable_queue(struct tap_queue *q)
*/
static void tap_put_queue(struct tap_queue *q)
{
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
rtnl_lock();
- vlan = rtnl_dereference(q->vlan);
+ tap = rtnl_dereference(q->tap);
- if (vlan) {
+ if (tap) {
if (q->enabled)
BUG_ON(tap_disable_queue(q));
- vlan->numqueues--;
- RCU_INIT_POINTER(q->vlan, NULL);
+ tap->numqueues--;
+ RCU_INIT_POINTER(q->tap, NULL);
sock_put(&q->sk);
list_del_init(&q->next);
}
@@ -270,17 +243,16 @@ static void tap_put_queue(struct tap_queue *q)
* Cache vlan->numvtaps since it can become zero during the execution
* of this function.
*/
-static struct tap_queue *tap_get_queue(struct net_device *dev,
+static struct tap_queue *tap_get_queue(struct tap_dev *tap,
struct sk_buff *skb)
{
- struct macvlan_dev *vlan = netdev_priv(dev);
- struct tap_queue *tap = NULL;
+ struct tap_queue *queue = NULL;
/* Access to taps array is protected by rcu, but access to numvtaps
* isn't. Below we use it to lookup a queue, but treat it as a hint
* and validate that the result isn't NULL - in case we are
* racing against queue removal.
*/
- int numvtaps = ACCESS_ONCE(vlan->numvtaps);
+ int numvtaps = ACCESS_ONCE(tap->numvtaps);
__u32 rxq;
if (!numvtaps)
@@ -292,7 +264,7 @@ static struct tap_queue *tap_get_queue(struct net_device *dev,
/* Check if we can use flow to select a queue */
rxq = skb_get_hash(skb);
if (rxq) {
- tap = rcu_dereference(vlan->taps[rxq % numvtaps]);
+ queue = rcu_dereference(tap->taps[rxq % numvtaps]);
goto out;
}
@@ -302,14 +274,14 @@ static struct tap_queue *tap_get_queue(struct net_device *dev,
while (unlikely(rxq >= numvtaps))
rxq -= numvtaps;
- tap = rcu_dereference(vlan->taps[rxq]);
+ queue = rcu_dereference(tap->taps[rxq]);
goto out;
}
single:
- tap = rcu_dereference(vlan->taps[0]);
+ queue = rcu_dereference(tap->taps[0]);
out:
- return tap;
+ return queue;
}
/*
@@ -317,39 +289,38 @@ static struct tap_queue *tap_get_queue(struct net_device *dev,
* that it holds on all queues and safely set the pointer
* from the queues to NULL.
*/
-void tap_del_queues(struct net_device *dev)
+void tap_del_queues(struct tap_dev *tap)
{
- struct macvlan_dev *vlan = netdev_priv(dev);
struct tap_queue *q, *tmp;
ASSERT_RTNL();
- list_for_each_entry_safe(q, tmp, &vlan->queue_list, next) {
+ list_for_each_entry_safe(q, tmp, &tap->queue_list, next) {
list_del_init(&q->next);
- RCU_INIT_POINTER(q->vlan, NULL);
+ RCU_INIT_POINTER(q->tap, NULL);
if (q->enabled)
- vlan->numvtaps--;
- vlan->numqueues--;
+ tap->numvtaps--;
+ tap->numqueues--;
sock_put(&q->sk);
}
- BUG_ON(vlan->numvtaps);
- BUG_ON(vlan->numqueues);
+ BUG_ON(tap->numvtaps);
+ BUG_ON(tap->numqueues);
/* guarantee that any future tap_set_queue will fail */
- vlan->numvtaps = MAX_TAP_QUEUES;
+ tap->numvtaps = MAX_TAP_QUEUES;
}
rx_handler_result_t tap_handle_frame(struct sk_buff **pskb)
{
struct sk_buff *skb = *pskb;
struct net_device *dev = skb->dev;
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
struct tap_queue *q;
netdev_features_t features = TAP_FEATURES;
- vlan = tap_get_vlan_rcu(dev);
- if (!vlan)
+ tap = tap_dev_get_rcu(dev);
+ if (!tap)
return RX_HANDLER_PASS;
- q = tap_get_queue(dev, skb);
+ q = tap_get_queue(tap, skb);
if (!q)
return RX_HANDLER_PASS;
@@ -363,7 +334,7 @@ rx_handler_result_t tap_handle_frame(struct sk_buff **pskb)
* enabled.
*/
if (q->flags & IFF_VNET_HDR)
- features |= vlan->tap_features;
+ features |= tap->tap_features;
if (netif_needs_gso(skb, features)) {
struct sk_buff *segs = __skb_gso_segment(skb, features, false);
@@ -408,50 +379,51 @@ rx_handler_result_t tap_handle_frame(struct sk_buff **pskb)
drop:
/* Count errors/drops only here, thus don't care about args. */
- macvlan_count_rx(vlan, 0, 0, 0);
+ if (tap->count_rx_dropped)
+ tap->count_rx_dropped(tap);
kfree_skb(skb);
return RX_HANDLER_CONSUMED;
}
-int tap_get_minor(struct macvlan_dev *vlan)
+int tap_get_minor(struct tap_dev *tap)
{
int retval = -ENOMEM;
mutex_lock(&macvtap_major.minor_lock);
- retval = idr_alloc(&macvtap_major.minor_idr, vlan, 1, TAP_NUM_DEVS, GFP_KERNEL);
+ retval = idr_alloc(&macvtap_major.minor_idr, tap, 1, TAP_NUM_DEVS, GFP_KERNEL);
if (retval >= 0) {
- vlan->minor = retval;
+ tap->minor = retval;
} else if (retval == -ENOSPC) {
- netdev_err(vlan->dev, "Too many tap devices\n");
+ netdev_err(tap->dev, "Too many tap devices\n");
retval = -EINVAL;
}
mutex_unlock(&macvtap_major.minor_lock);
return retval < 0 ? retval : 0;
}
-void tap_free_minor(struct macvlan_dev *vlan)
+void tap_free_minor(struct tap_dev *tap)
{
mutex_lock(&macvtap_major.minor_lock);
- if (vlan->minor) {
- idr_remove(&macvtap_major.minor_idr, vlan->minor);
- vlan->minor = 0;
+ if (tap->minor) {
+ idr_remove(&macvtap_major.minor_idr, tap->minor);
+ tap->minor = 0;
}
mutex_unlock(&macvtap_major.minor_lock);
}
-static struct net_device *dev_get_by_tap_minor(int minor)
+static struct tap_dev *dev_get_by_tap_minor(int minor)
{
struct net_device *dev = NULL;
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
mutex_lock(&macvtap_major.minor_lock);
- vlan = idr_find(&macvtap_major.minor_idr, minor);
- if (vlan) {
- dev = vlan->dev;
+ tap = idr_find(&macvtap_major.minor_idr, minor);
+ if (tap) {
+ dev = tap->dev;
dev_hold(dev);
}
mutex_unlock(&macvtap_major.minor_lock);
- return dev;
+ return tap;
}
static void tap_sock_write_space(struct sock *sk)
@@ -477,13 +449,13 @@ static void tap_sock_destruct(struct sock *sk)
static int tap_open(struct inode *inode, struct file *file)
{
struct net *net = current->nsproxy->net_ns;
- struct net_device *dev;
+ struct tap_dev *tap;
struct tap_queue *q;
int err = -ENODEV;
rtnl_lock();
- dev = dev_get_by_tap_minor(iminor(inode));
- if (!dev)
+ tap = dev_get_by_tap_minor(iminor(inode));
+ if (!tap)
goto err;
err = -ENOMEM;
@@ -511,18 +483,18 @@ static int tap_open(struct inode *inode, struct file *file)
* The macvlan supports zerocopy iff the lower device supports zero
* copy so we don't have to look at the lower device directly.
*/
- if ((dev->features & NETIF_F_HIGHDMA) && (dev->features & NETIF_F_SG))
+ if ((tap->dev->features & NETIF_F_HIGHDMA) && (tap->dev->features & NETIF_F_SG))
sock_set_flag(&q->sk, SOCK_ZEROCOPY);
err = -ENOMEM;
- if (skb_array_init(&q->skb_array, dev->tx_queue_len, GFP_KERNEL))
+ if (skb_array_init(&q->skb_array, tap->dev->tx_queue_len, GFP_KERNEL))
goto err_array;
- err = tap_set_queue(dev, file, q);
+ err = tap_set_queue(tap, file, q);
if (err)
goto err_queue;
- dev_put(dev);
+ dev_put(tap->dev);
rtnl_unlock();
return err;
@@ -532,8 +504,8 @@ static int tap_open(struct inode *inode, struct file *file)
err_array:
sock_put(&q->sk);
err:
- if (dev)
- dev_put(dev);
+ if (tap)
+ dev_put(tap->dev);
rtnl_unlock();
return err;
@@ -601,7 +573,7 @@ static ssize_t tap_get_user(struct tap_queue *q, struct msghdr *m,
{
int good_linear = SKB_MAX_HEAD(TAP_RESERVE);
struct sk_buff *skb;
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
unsigned long total_len = iov_iter_count(from);
unsigned long len = total_len;
int err;
@@ -698,7 +670,7 @@ static ssize_t tap_get_user(struct tap_queue *q, struct msghdr *m,
skb_set_network_header(skb, depth);
rcu_read_lock();
- vlan = rcu_dereference(q->vlan);
+ tap = rcu_dereference(q->tap);
/* copy skb_ubuf_info for callback when skb has no error */
if (zerocopy) {
skb_shinfo(skb)->destructor_arg = m->msg_control;
@@ -709,8 +681,8 @@ static ssize_t tap_get_user(struct tap_queue *q, struct msghdr *m,
uarg->callback(uarg, false);
}
- if (vlan) {
- skb->dev = vlan->dev;
+ if (tap) {
+ skb->dev = tap->dev;
dev_queue_xmit(skb);
} else {
kfree_skb(skb);
@@ -724,9 +696,9 @@ static ssize_t tap_get_user(struct tap_queue *q, struct msghdr *m,
err:
rcu_read_lock();
- vlan = rcu_dereference(q->vlan);
- if (vlan)
- this_cpu_inc(vlan->pcpu_stats->tx_dropped);
+ tap = rcu_dereference(q->tap);
+ if (tap && tap->count_tx_dropped)
+ tap->count_tx_dropped(tap);
rcu_read_unlock();
return err;
@@ -853,55 +825,55 @@ static ssize_t tap_read_iter(struct kiocb *iocb, struct iov_iter *to)
return ret;
}
-static struct macvlan_dev *tap_get_vlan(struct tap_queue *q)
+static struct tap_dev *tap_get_tap_dev(struct tap_queue *q)
{
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
ASSERT_RTNL();
- vlan = rtnl_dereference(q->vlan);
- if (vlan)
- dev_hold(vlan->dev);
+ tap = rtnl_dereference(q->tap);
+ if (tap)
+ dev_hold(tap->dev);
- return vlan;
+ return tap;
}
-static void tap_put_vlan(struct macvlan_dev *vlan)
+static void tap_put_tap_dev(struct tap_dev *tap)
{
- dev_put(vlan->dev);
+ dev_put(tap->dev);
}
static int tap_ioctl_set_queue(struct file *file, unsigned int flags)
{
struct tap_queue *q = file->private_data;
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
int ret;
- vlan = tap_get_vlan(q);
- if (!vlan)
+ tap = tap_get_tap_dev(q);
+ if (!tap)
return -EINVAL;
if (flags & IFF_ATTACH_QUEUE)
- ret = tap_enable_queue(vlan->dev, file, q);
+ ret = tap_enable_queue(tap, file, q);
else if (flags & IFF_DETACH_QUEUE)
ret = tap_disable_queue(q);
else
ret = -EINVAL;
- tap_put_vlan(vlan);
+ tap_put_tap_dev(tap);
return ret;
}
static int set_offload(struct tap_queue *q, unsigned long arg)
{
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
netdev_features_t features;
netdev_features_t feature_mask = 0;
- vlan = rtnl_dereference(q->vlan);
- if (!vlan)
+ tap = rtnl_dereference(q->tap);
+ if (!tap)
return -ENOLINK;
- features = vlan->dev->features;
+ features = tap->dev->features;
if (arg & TUN_F_CSUM) {
feature_mask = NETIF_F_HW_CSUM;
@@ -935,9 +907,9 @@ static int set_offload(struct tap_queue *q, unsigned long arg)
/* tap_features are the same as features on tun/tap and
* reflect user expectations.
*/
- vlan->tap_features = feature_mask;
- vlan->set_features = features;
- netdev_update_features(vlan->dev);
+ tap->tap_features = feature_mask;
+ if (tap->update_features)
+ tap->update_features(tap, features);
return 0;
}
@@ -949,7 +921,7 @@ static long tap_ioctl(struct file *file, unsigned int cmd,
unsigned long arg)
{
struct tap_queue *q = file->private_data;
- struct macvlan_dev *vlan;
+ struct tap_dev *tap;
void __user *argp = (void __user *)arg;
struct ifreq __user *ifr = argp;
unsigned int __user *up = argp;
@@ -975,18 +947,18 @@ static long tap_ioctl(struct file *file, unsigned int cmd,
case TUNGETIFF:
rtnl_lock();
- vlan = tap_get_vlan(q);
- if (!vlan) {
+ tap = tap_get_tap_dev(q);
+ if (!tap) {
rtnl_unlock();
return -ENOLINK;
}
ret = 0;
u = q->flags;
- if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
+ if (copy_to_user(&ifr->ifr_name, tap->dev->name, IFNAMSIZ) ||
put_user(u, &ifr->ifr_flags))
ret = -EFAULT;
- tap_put_vlan(vlan);
+ tap_put_tap_dev(tap);
rtnl_unlock();
return ret;
@@ -1059,18 +1031,18 @@ static long tap_ioctl(struct file *file, unsigned int cmd,
case SIOCGIFHWADDR:
rtnl_lock();
- vlan = tap_get_vlan(q);
- if (!vlan) {
+ tap = tap_get_tap_dev(q);
+ if (!tap) {
rtnl_unlock();
return -ENOLINK;
}
ret = 0;
- u = vlan->dev->type;
- if (copy_to_user(&ifr->ifr_name, vlan->dev->name, IFNAMSIZ) ||
- copy_to_user(&ifr->ifr_hwaddr.sa_data, vlan->dev->dev_addr, ETH_ALEN) ||
+ u = tap->dev->type;
+ if (copy_to_user(&ifr->ifr_name, tap->dev->name, IFNAMSIZ) ||
+ copy_to_user(&ifr->ifr_hwaddr.sa_data, tap->dev->dev_addr, ETH_ALEN) ||
put_user(u, &ifr->ifr_hwaddr.sa_family))
ret = -EFAULT;
- tap_put_vlan(vlan);
+ tap_put_tap_dev(tap);
rtnl_unlock();
return ret;
@@ -1078,13 +1050,13 @@ static long tap_ioctl(struct file *file, unsigned int cmd,
if (copy_from_user(&sa, &ifr->ifr_hwaddr, sizeof(sa)))
return -EFAULT;
rtnl_lock();
- vlan = tap_get_vlan(q);
- if (!vlan) {
+ tap = tap_get_tap_dev(q);
+ if (!tap) {
rtnl_unlock();
return -ENOLINK;
}
- ret = dev_set_mac_address(vlan->dev, &sa);
- tap_put_vlan(vlan);
+ ret = dev_set_mac_address(tap->dev, &sa);
+ tap_put_tap_dev(tap);
rtnl_unlock();
return ret;
@@ -1167,19 +1139,19 @@ struct socket *tap_get_socket(struct file *file)
}
EXPORT_SYMBOL_GPL(tap_get_socket);
-int tap_queue_resize(struct macvlan_dev *vlan)
+int tap_queue_resize(struct tap_dev *tap)
{
- struct net_device *dev = vlan->dev;
+ struct net_device *dev = tap->dev;
struct tap_queue *q;
struct skb_array **arrays;
- int n = vlan->numqueues;
+ int n = tap->numqueues;
int ret, i = 0;
arrays = kmalloc(sizeof *arrays * n, GFP_KERNEL);
if (!arrays)
return -ENOMEM;
- list_for_each_entry(q, &vlan->queue_list, next)
+ list_for_each_entry(q, &tap->queue_list, next)
arrays[i++] = &q->skb_array;
ret = skb_array_resize_multiple(arrays, n,
diff --git a/include/linux/if_tap.h b/include/linux/if_tap.h
index 89bcd42..e6b2a88 100644
--- a/include/linux/if_tap.h
+++ b/include/linux/if_tap.h
@@ -1,13 +1,62 @@
#ifndef _LINUX_IF_TAP_H_
#define _LINUX_IF_TAP_H_
+#include <net/sock.h>
+#include <linux/skb_array.h>
+
+#define MAX_TAP_QUEUES 256
+
+struct tap_queue;
+
+struct tap_dev {
+ struct net_device *dev;
+ u16 flags;
+ /* This array tracks active taps. */
+ struct tap_queue __rcu *taps[MAX_TAP_QUEUES];
+ /* This list tracks all taps (both enabled and disabled) */
+ struct list_head queue_list;
+ int numvtaps;
+ int numqueues;
+ netdev_features_t tap_features;
+ int minor;
+
+ void (*update_features)(struct tap_dev *tap, netdev_features_t features);
+ void (*count_tx_dropped)(struct tap_dev *tap);
+ void (*count_rx_dropped)(struct tap_dev *tap);
+};
+
+/*
+ * A tap queue is the central object of tap module, it connects
+ * an open character device to virtual interface. There can be
+ * multiple queues on one interface, which map back to queues
+ * implemented in hardware on the underlying device.
+ *
+ * tap_proto is used to allocate queues through the sock allocation
+ * mechanism.
+ *
+ */
+
+struct tap_queue {
+ struct sock sk;
+ struct socket sock;
+ struct socket_wq wq;
+ int vnet_hdr_sz;
+ struct tap_dev __rcu *tap;
+ struct file *file;
+ unsigned int flags;
+ u16 queue_index;
+ bool enabled;
+ struct list_head next;
+ struct skb_array skb_array;
+};
+
rx_handler_result_t tap_handle_frame(struct sk_buff **pskb);
-void tap_del_queues(struct net_device *dev);
-int tap_get_minor(struct macvlan_dev *vlan);
-void tap_free_minor(struct macvlan_dev *vlan);
-int tap_queue_resize(struct macvlan_dev *vlan);
+void tap_del_queues(struct tap_dev *tap);
+int tap_get_minor(struct tap_dev *tap);
+void tap_free_minor(struct tap_dev *tap);
+int tap_queue_resize(struct tap_dev *tap);
int tap_create_cdev(struct cdev *tap_cdev,
- dev_t *tap_major, const char *device_name);
+ dev_t *tap_major, const char *device_name);
void tap_destroy_cdev(dev_t major, struct cdev *tap_cdev);
struct socket *tap_get_socket(struct file *file);
--
2.7.4
^ permalink raw reply related [flat|nested] 21+ messages in thread* [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs
2017-01-06 22:33 [PATCHv1 0/7] Refactor macvtap to re-use tap functionality by other virtual intefaces Sainath Grandhi
` (3 preceding siblings ...)
2017-01-06 22:33 ` [PATCHv1 4/7] TAP: Abstract type of virtual interface from tap implementation Sainath Grandhi
@ 2017-01-06 22:33 ` Sainath Grandhi
2017-01-06 23:15 ` Eric Dumazet
2017-01-06 23:20 ` Andy Shevchenko
2017-01-06 22:33 ` [PATCHv1 6/7] TAP: tap as an independent module Sainath Grandhi
2017-01-06 22:33 ` [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver Sainath Grandhi
6 siblings, 2 replies; 21+ messages in thread
From: Sainath Grandhi @ 2017-01-06 22:33 UTC (permalink / raw)
To: netdev; +Cc: davem, mahesh, linux-kernel, Sainath Grandhi
Extending tap APIs get/free_minor and create/destroy_cdev to handle more than one
type of virtual interface.
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
---
drivers/net/macvtap_main.c | 6 +--
drivers/net/tap.c | 110 ++++++++++++++++++++++++++++++++++++---------
include/linux/if_tap.h | 4 +-
3 files changed, 93 insertions(+), 27 deletions(-)
diff --git a/drivers/net/macvtap_main.c b/drivers/net/macvtap_main.c
index 6326a82..3f047b4 100644
--- a/drivers/net/macvtap_main.c
+++ b/drivers/net/macvtap_main.c
@@ -160,7 +160,7 @@ static int macvtap_device_event(struct notifier_block *unused,
* been registered but before register_netdevice has
* finished running.
*/
- err = tap_get_minor(&vlantap->tap);
+ err = tap_get_minor(macvtap_major, &vlantap->tap);
if (err)
return notifier_from_errno(err);
@@ -168,7 +168,7 @@ static int macvtap_device_event(struct notifier_block *unused,
classdev = device_create(&macvtap_class, &dev->dev, devt,
dev, tap_name);
if (IS_ERR(classdev)) {
- tap_free_minor(&vlantap->tap);
+ tap_free_minor(macvtap_major, &vlantap->tap);
return notifier_from_errno(PTR_ERR(classdev));
}
err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
@@ -183,7 +183,7 @@ static int macvtap_device_event(struct notifier_block *unused,
sysfs_remove_link(&dev->dev.kobj, tap_name);
devt = MKDEV(MAJOR(macvtap_major), vlantap->tap.minor);
device_destroy(&macvtap_class, devt);
- tap_free_minor(&vlantap->tap);
+ tap_free_minor(macvtap_major, &vlantap->tap);
break;
case NETDEV_CHANGE_TX_QUEUE_LEN:
if (tap_queue_resize(&vlantap->tap))
diff --git a/drivers/net/tap.c b/drivers/net/tap.c
index 6306ab9..1d5bcf3 100644
--- a/drivers/net/tap.c
+++ b/drivers/net/tap.c
@@ -99,12 +99,16 @@ static struct proto tap_proto = {
};
#define TAP_NUM_DEVS (1U << MINORBITS)
+
+LIST_HEAD(major_list);
+
struct major_info {
dev_t major;
struct idr minor_idr;
struct mutex minor_lock;
const char *device_name;
-} macvtap_major;
+ struct list_head next;
+};
#define GOODCOPY_LEN 128
@@ -385,44 +389,81 @@ rx_handler_result_t tap_handle_frame(struct sk_buff **pskb)
return RX_HANDLER_CONSUMED;
}
-int tap_get_minor(struct tap_dev *tap)
+int tap_get_minor(dev_t major, struct tap_dev *tap)
{
int retval = -ENOMEM;
+ struct major_info *tap_major, *tmp;
+ bool found = false;
- mutex_lock(&macvtap_major.minor_lock);
- retval = idr_alloc(&macvtap_major.minor_idr, tap, 1, TAP_NUM_DEVS, GFP_KERNEL);
+ list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
+ if (tap_major->major == MAJOR(major)) {
+ found = true;
+ break;
+ }
+ }
+
+ if (!found)
+ return -EINVAL;
+
+ mutex_lock(&tap_major->minor_lock);
+ retval = idr_alloc(&tap_major->minor_idr, tap, 1, TAP_NUM_DEVS, GFP_KERNEL);
if (retval >= 0) {
tap->minor = retval;
} else if (retval == -ENOSPC) {
netdev_err(tap->dev, "Too many tap devices\n");
retval = -EINVAL;
}
- mutex_unlock(&macvtap_major.minor_lock);
+ mutex_unlock(&tap_major->minor_lock);
return retval < 0 ? retval : 0;
}
-void tap_free_minor(struct tap_dev *tap)
+void tap_free_minor(dev_t major, struct tap_dev *tap)
{
- mutex_lock(&macvtap_major.minor_lock);
+ struct major_info *tap_major, *tmp;
+ bool found = false;
+
+ list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
+ if (tap_major->major == MAJOR(major)) {
+ found = true;
+ break;
+ }
+ }
+
+ if (!found)
+ return;
+
+ mutex_lock(&tap_major->minor_lock);
if (tap->minor) {
- idr_remove(&macvtap_major.minor_idr, tap->minor);
+ idr_remove(&tap_major->minor_idr, tap->minor);
tap->minor = 0;
}
- mutex_unlock(&macvtap_major.minor_lock);
+ mutex_unlock(&tap_major->minor_lock);
}
-static struct tap_dev *dev_get_by_tap_minor(int minor)
+static struct tap_dev *dev_get_by_tap_file(int major, int minor)
{
struct net_device *dev = NULL;
struct tap_dev *tap;
+ struct major_info *tap_major, *tmp;
+ bool found = false;
- mutex_lock(&macvtap_major.minor_lock);
- tap = idr_find(&macvtap_major.minor_idr, minor);
+ list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
+ if (tap_major->major == major) {
+ found = true;
+ break;
+ }
+ }
+
+ if (!found)
+ return NULL;
+
+ mutex_lock(&tap_major->minor_lock);
+ tap = idr_find(&tap_major->minor_idr, minor);
if (tap) {
dev = tap->dev;
dev_hold(dev);
}
- mutex_unlock(&macvtap_major.minor_lock);
+ mutex_unlock(&tap_major->minor_lock);
return tap;
}
@@ -454,7 +495,7 @@ static int tap_open(struct inode *inode, struct file *file)
int err = -ENODEV;
rtnl_lock();
- tap = dev_get_by_tap_minor(iminor(inode));
+ tap = dev_get_by_tap_file(imajor(inode), iminor(inode));
if (!tap)
goto err;
@@ -1161,6 +1202,24 @@ int tap_queue_resize(struct tap_dev *tap)
return ret;
}
+static int tap_list_add(dev_t major, const char *device_name)
+{
+ int err = 0;
+ struct major_info *tap_major;
+
+ tap_major = kzalloc(sizeof(*tap_major), GFP_ATOMIC);
+
+ tap_major->major = MAJOR(major);
+
+ idr_init(&tap_major->minor_idr);
+ mutex_init(&tap_major->minor_lock);
+
+ tap_major->device_name = device_name;
+
+ list_add_tail(&tap_major->next, &major_list);
+ return err;
+}
+
int tap_create_cdev(struct cdev *tap_cdev,
dev_t *tap_major, const char *device_name)
{
@@ -1176,12 +1235,7 @@ int tap_create_cdev(struct cdev *tap_cdev,
if (err)
goto out2;
- macvtap_major.major = MAJOR(*tap_major);
-
- idr_init(&macvtap_major.minor_idr);
- mutex_init(&macvtap_major.minor_lock);
-
- macvtap_major.device_name = device_name;
+ err = tap_list_add(*tap_major, device_name);
return err;
@@ -1193,8 +1247,20 @@ int tap_create_cdev(struct cdev *tap_cdev,
void tap_destroy_cdev(dev_t major, struct cdev *tap_cdev)
{
+ struct major_info *tap_major, *tmp;
+ bool found = false;
+
+ list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
+ if (tap_major->major == MAJOR(major)) {
+ found = true;
+ break;
+ }
+ }
+
+ if (!found)
+ return;
+
cdev_del(tap_cdev);
unregister_chrdev_region(major, TAP_NUM_DEVS);
- idr_destroy(&macvtap_major.minor_idr);
+ idr_destroy(&tap_major->minor_idr);
}
-
diff --git a/include/linux/if_tap.h b/include/linux/if_tap.h
index e6b2a88..a8a8e16 100644
--- a/include/linux/if_tap.h
+++ b/include/linux/if_tap.h
@@ -52,8 +52,8 @@ struct tap_queue {
rx_handler_result_t tap_handle_frame(struct sk_buff **pskb);
void tap_del_queues(struct tap_dev *tap);
-int tap_get_minor(struct tap_dev *tap);
-void tap_free_minor(struct tap_dev *tap);
+int tap_get_minor(dev_t major, struct tap_dev *tap);
+void tap_free_minor(dev_t major, struct tap_dev *tap);
int tap_queue_resize(struct tap_dev *tap);
int tap_create_cdev(struct cdev *tap_cdev,
dev_t *tap_major, const char *device_name);
--
2.7.4
^ permalink raw reply related [flat|nested] 21+ messages in thread* Re: [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs
2017-01-06 22:33 ` [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs Sainath Grandhi
@ 2017-01-06 23:15 ` Eric Dumazet
2017-01-17 23:07 ` Grandhi, Sainath
2017-01-06 23:20 ` Andy Shevchenko
1 sibling, 1 reply; 21+ messages in thread
From: Eric Dumazet @ 2017-01-06 23:15 UTC (permalink / raw)
To: Sainath Grandhi; +Cc: netdev, davem, mahesh, linux-kernel
On Fri, 2017-01-06 at 14:33 -0800, Sainath Grandhi wrote:
> +static int tap_list_add(dev_t major, const char *device_name)
> +{
> + int err = 0;
> + struct major_info *tap_major;
> +
> + tap_major = kzalloc(sizeof(*tap_major), GFP_ATOMIC);
> +
> + tap_major->major = MAJOR(major);
> +
kzalloc() can perfectly return NULL.
You do not want to crash it that happens.
^ permalink raw reply [flat|nested] 21+ messages in thread* RE: [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs
2017-01-06 23:15 ` Eric Dumazet
@ 2017-01-17 23:07 ` Grandhi, Sainath
0 siblings, 0 replies; 21+ messages in thread
From: Grandhi, Sainath @ 2017-01-17 23:07 UTC (permalink / raw)
To: Eric Dumazet
Cc: netdev@vger.kernel.org, davem@davemloft.net, mahesh@bandewar.net,
linux-kernel@vger.kernel.org
> -----Original Message-----
> From: Eric Dumazet [mailto:eric.dumazet@gmail.com]
> Sent: Friday, January 06, 2017 3:16 PM
> To: Grandhi, Sainath <sainath.grandhi@intel.com>
> Cc: netdev@vger.kernel.org; davem@davemloft.net;
> mahesh@bandewar.net; linux-kernel@vger.kernel.org
> Subject: Re: [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs
>
> On Fri, 2017-01-06 at 14:33 -0800, Sainath Grandhi wrote:
>
> > +static int tap_list_add(dev_t major, const char *device_name) {
> > + int err = 0;
> > + struct major_info *tap_major;
> > +
> > + tap_major = kzalloc(sizeof(*tap_major), GFP_ATOMIC);
> > +
> > + tap_major->major = MAJOR(major);
> > +
>
>
> kzalloc() can perfectly return NULL.
>
> You do not want to crash it that happens.
>
Thanks for pointing out. Will send out next version that takes care of null pointer
^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs
2017-01-06 22:33 ` [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs Sainath Grandhi
2017-01-06 23:15 ` Eric Dumazet
@ 2017-01-06 23:20 ` Andy Shevchenko
2017-01-17 23:51 ` Grandhi, Sainath
1 sibling, 1 reply; 21+ messages in thread
From: Andy Shevchenko @ 2017-01-06 23:20 UTC (permalink / raw)
To: Sainath Grandhi
Cc: netdev, David S. Miller, mahesh, linux-kernel@vger.kernel.org
On Sat, Jan 7, 2017 at 12:33 AM, Sainath Grandhi
<sainath.grandhi@intel.com> wrote:
> Extending tap APIs get/free_minor and create/destroy_cdev to handle more than one
> type of virtual interface.
>
> Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
> Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
Usually it implies that commiter has tested the stuff.
> --- a/drivers/net/tap.c
> +++ b/drivers/net/tap.c
> @@ -99,12 +99,16 @@ static struct proto tap_proto = {
> };
>
> #define TAP_NUM_DEVS (1U << MINORBITS)
> +
> +LIST_HEAD(major_list);
> +
static ?
> -int tap_get_minor(struct tap_dev *tap)
> +int tap_get_minor(dev_t major, struct tap_dev *tap)
> {
> int retval = -ENOMEM;
> + struct major_info *tap_major, *tmp;
> + bool found = false;
>
> - mutex_lock(&macvtap_major.minor_lock);
> - retval = idr_alloc(&macvtap_major.minor_idr, tap, 1, TAP_NUM_DEVS, GFP_KERNEL);
> + list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
> + if (tap_major->major == MAJOR(major)) {
> + found = true;
> + break;
> + }
> + }
> +
> + if (!found)
> + return -EINVAL;
This is candidate to be a separate helper function. See also below.
> -void tap_free_minor(struct tap_dev *tap)
> +void tap_free_minor(dev_t major, struct tap_dev *tap)
> {
> - mutex_lock(&macvtap_major.minor_lock);
> + struct major_info *tap_major, *tmp;
> + bool found = false;
> +
> + list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
> + if (tap_major->major == MAJOR(major)) {
> + found = true;
> + break;
> + }
> + }
> +
> + if (!found)
> + return;
Here is quite the same code (as above).
> -static struct tap_dev *dev_get_by_tap_minor(int minor)
> +static struct tap_dev *dev_get_by_tap_file(int major, int minor)
> {
> struct net_device *dev = NULL;
> struct tap_dev *tap;
> + struct major_info *tap_major, *tmp;
> + bool found = false;
>
> - mutex_lock(&macvtap_major.minor_lock);
> - tap = idr_find(&macvtap_major.minor_idr, minor);
> + list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
> + if (tap_major->major == major) {
> + found = true;
> + break;
> + }
> + }
> +
> + if (!found)
> + return NULL;
And here.
> +static int tap_list_add(dev_t major, const char *device_name)
> +{
> + int err = 0;
> + struct major_info *tap_major;
Perhaps
+ struct major_info *tap_major;
+ int err = 0;
> +
> + tap_major = kzalloc(sizeof(*tap_major), GFP_ATOMIC);
> +
> + tap_major->major = MAJOR(major);
> +
> + idr_init(&tap_major->minor_idr);
> + mutex_init(&tap_major->minor_lock);
> +
> + tap_major->device_name = device_name;
> +
> + list_add_tail(&tap_major->next, &major_list);
> + return err;
> + err = tap_list_add(*tap_major, device_name);
>
> return err;
return tap_list_add();
> void tap_destroy_cdev(dev_t major, struct cdev *tap_cdev)
> {
> + struct major_info *tap_major, *tmp;
> + bool found = false;
> +
> + list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
> + if (tap_major->major == MAJOR(major)) {
> + found = true;
> + break;
> + }
> + }
> +
> + if (!found)
> + return;
And here.
--
With Best Regards,
Andy Shevchenko
^ permalink raw reply [flat|nested] 21+ messages in thread* RE: [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs
2017-01-06 23:20 ` Andy Shevchenko
@ 2017-01-17 23:51 ` Grandhi, Sainath
0 siblings, 0 replies; 21+ messages in thread
From: Grandhi, Sainath @ 2017-01-17 23:51 UTC (permalink / raw)
To: Andy Shevchenko
Cc: netdev, David S. Miller, mahesh@bandewar.net,
linux-kernel@vger.kernel.org
Please find reply inline.
> -----Original Message-----
> From: Andy Shevchenko [mailto:andy.shevchenko@gmail.com]
> Sent: Friday, January 06, 2017 3:21 PM
> To: Grandhi, Sainath <sainath.grandhi@intel.com>
> Cc: netdev <netdev@vger.kernel.org>; David S. Miller
> <davem@davemloft.net>; mahesh@bandewar.net; linux-
> kernel@vger.kernel.org
> Subject: Re: [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs
>
> On Sat, Jan 7, 2017 at 12:33 AM, Sainath Grandhi
> <sainath.grandhi@intel.com> wrote:
> > Extending tap APIs get/free_minor and create/destroy_cdev to handle
> > more than one type of virtual interface.
> >
> > Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
> > Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
>
> Usually it implies that commiter has tested the stuff.
>
> > --- a/drivers/net/tap.c
> > +++ b/drivers/net/tap.c
> > @@ -99,12 +99,16 @@ static struct proto tap_proto = { };
> >
> > #define TAP_NUM_DEVS (1U << MINORBITS)
>
> > +
> > +LIST_HEAD(major_list);
> > +
>
> static ?
Makes sense. Would take care of it.
>
> > -int tap_get_minor(struct tap_dev *tap)
> > +int tap_get_minor(dev_t major, struct tap_dev *tap)
> > {
> > int retval = -ENOMEM;
> > + struct major_info *tap_major, *tmp;
> > + bool found = false;
> >
> > - mutex_lock(&macvtap_major.minor_lock);
> > - retval = idr_alloc(&macvtap_major.minor_idr, tap, 1, TAP_NUM_DEVS,
> GFP_KERNEL);
>
> > + list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
> > + if (tap_major->major == MAJOR(major)) {
> > + found = true;
> > + break;
> > + }
> > + }
> > +
> > + if (!found)
> > + return -EINVAL;
>
> This is candidate to be a separate helper function. See also below.
Would define a helper function.
>
>
> > -void tap_free_minor(struct tap_dev *tap)
> > +void tap_free_minor(dev_t major, struct tap_dev *tap)
> > {
> > - mutex_lock(&macvtap_major.minor_lock);
> > + struct major_info *tap_major, *tmp;
>
> > + bool found = false;
> > +
> > + list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
> > + if (tap_major->major == MAJOR(major)) {
> > + found = true;
> > + break;
> > + }
> > + }
> > +
> > + if (!found)
> > + return;
>
> Here is quite the same code (as above).
>
> > -static struct tap_dev *dev_get_by_tap_minor(int minor)
> > +static struct tap_dev *dev_get_by_tap_file(int major, int minor)
> > {
> > struct net_device *dev = NULL;
> > struct tap_dev *tap;
> > + struct major_info *tap_major, *tmp;
> > + bool found = false;
> >
> > - mutex_lock(&macvtap_major.minor_lock);
> > - tap = idr_find(&macvtap_major.minor_idr, minor);
>
> > + list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
> > + if (tap_major->major == major) {
> > + found = true;
> > + break;
> > + }
> > + }
> > +
> > + if (!found)
> > + return NULL;
>
> And here.
>
> > +static int tap_list_add(dev_t major, const char *device_name) {
>
> > + int err = 0;
> > + struct major_info *tap_major;
>
> Perhaps
> + struct major_info *tap_major;
> + int err = 0;
>
> > +
> > + tap_major = kzalloc(sizeof(*tap_major), GFP_ATOMIC);
> > +
> > + tap_major->major = MAJOR(major);
> > +
> > + idr_init(&tap_major->minor_idr);
> > + mutex_init(&tap_major->minor_lock);
> > +
> > + tap_major->device_name = device_name;
> > +
> > + list_add_tail(&tap_major->next, &major_list);
> > + return err;
>
>
> > + err = tap_list_add(*tap_major, device_name);
> >
> > return err;
>
> return tap_list_add();
>
> > void tap_destroy_cdev(dev_t major, struct cdev *tap_cdev) {
> > + struct major_info *tap_major, *tmp;
> > + bool found = false;
> > +
> > + list_for_each_entry_safe(tap_major, tmp, &major_list, next) {
> > + if (tap_major->major == MAJOR(major)) {
> > + found = true;
> > + break;
> > + }
> > + }
> > +
> > + if (!found)
> > + return;
>
> And here.
>
> --
> With Best Regards,
> Andy Shevchenko
^ permalink raw reply [flat|nested] 21+ messages in thread
* [PATCHv1 6/7] TAP: tap as an independent module
2017-01-06 22:33 [PATCHv1 0/7] Refactor macvtap to re-use tap functionality by other virtual intefaces Sainath Grandhi
` (4 preceding siblings ...)
2017-01-06 22:33 ` [PATCHv1 5/7] TAP: Extending tap device create/destroy APIs Sainath Grandhi
@ 2017-01-06 22:33 ` Sainath Grandhi
2017-01-07 10:46 ` kbuild test robot
2017-01-07 12:32 ` kbuild test robot
2017-01-06 22:33 ` [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver Sainath Grandhi
6 siblings, 2 replies; 21+ messages in thread
From: Sainath Grandhi @ 2017-01-06 22:33 UTC (permalink / raw)
To: netdev; +Cc: davem, mahesh, linux-kernel, Sainath Grandhi
This patch makes tap a separate module for other types of virtual interfaces, for example,
ipvlan to use.
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
---
drivers/net/Kconfig | 14 +++
drivers/net/Makefile | 3 +-
drivers/net/macvtap.c | 247 +++++++++++++++++++++++++++++++++++++++++++++
drivers/net/macvtap_main.c | 247 ---------------------------------------------
drivers/net/tap.c | 10 ++
5 files changed, 272 insertions(+), 249 deletions(-)
create mode 100644 drivers/net/macvtap.c
delete mode 100644 drivers/net/macvtap_main.c
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index 95c32f2..280380d 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -284,6 +284,20 @@ config TUN
If you don't know what to use this for, you don't need it.
+config TAP
+ tristate "TAP module support for virtual interfaces"
+ ---help---
+ TAP module serves two purposes. This can be used as library of functions
+ for virtual interfaces to implement tap functionality.
+
+ This module also includes character device file and socket operations
+ that can be used by virtual interface implementing tap.
+
+ To compile this driver as a module, choose M here: the module
+ will be called tap.
+
+ If you don't know what to use this for, you don't need it.
+
config TUN_VNET_CROSS_LE
bool "Support for cross-endian vnet headers on little-endian kernels"
default n
diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index 19b03a9..7dd86ca 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -21,6 +21,7 @@ obj-$(CONFIG_PHYLIB) += phy/
obj-$(CONFIG_RIONET) += rionet.o
obj-$(CONFIG_NET_TEAM) += team/
obj-$(CONFIG_TUN) += tun.o
+obj-$(CONFIG_TAP) += tap.o
obj-$(CONFIG_VETH) += veth.o
obj-$(CONFIG_VIRTIO_NET) += virtio_net.o
obj-$(CONFIG_VXLAN) += vxlan.o
@@ -29,8 +30,6 @@ obj-$(CONFIG_GTP) += gtp.o
obj-$(CONFIG_NLMON) += nlmon.o
obj-$(CONFIG_NET_VRF) += vrf.o
-macvtap-objs := macvtap_main.o tap.o
-
#
# Networking Drivers
#
diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
new file mode 100644
index 0000000..3f047b4
--- /dev/null
+++ b/drivers/net/macvtap.c
@@ -0,0 +1,247 @@
+#include <linux/etherdevice.h>
+#include <linux/if_macvlan.h>
+#include <linux/if_tap.h>
+#include <linux/if_vlan.h>
+#include <linux/interrupt.h>
+#include <linux/nsproxy.h>
+#include <linux/compat.h>
+#include <linux/if_tun.h>
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <linux/cache.h>
+#include <linux/sched.h>
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/wait.h>
+#include <linux/cdev.h>
+#include <linux/idr.h>
+#include <linux/fs.h>
+#include <linux/uio.h>
+
+#include <net/net_namespace.h>
+#include <net/rtnetlink.h>
+#include <net/sock.h>
+#include <linux/virtio_net.h>
+#include <linux/skb_array.h>
+
+struct macvtap_dev {
+ struct macvlan_dev vlan;
+ struct tap_dev tap;
+};
+
+/*
+ * Variables for dealing with macvtaps device numbers.
+ */
+static dev_t macvtap_major;
+
+static const void *macvtap_net_namespace(struct device *d)
+{
+ struct net_device *dev = to_net_dev(d->parent);
+ return dev_net(dev);
+}
+
+static struct class macvtap_class = {
+ .name = "macvtap",
+ .owner = THIS_MODULE,
+ .ns_type = &net_ns_type_operations,
+ .namespace = macvtap_net_namespace,
+};
+static struct cdev macvtap_cdev;
+
+#define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN | NETIF_F_TSO | \
+ NETIF_F_TSO6 | NETIF_F_UFO)
+
+static void macvtap_count_tx_dropped(struct tap_dev *tap)
+{
+ struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
+
+ this_cpu_inc(vlan->pcpu_stats->tx_dropped);
+}
+
+static void macvtap_count_rx_dropped(struct tap_dev *tap)
+{
+ struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
+
+ macvlan_count_rx(vlan, 0, 0, 0);
+}
+
+static void macvtap_update_features(struct tap_dev *tap,
+ netdev_features_t features)
+{
+ struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
+
+ vlan->set_features = features;
+ netdev_update_features(vlan->dev);
+}
+
+static int macvtap_newlink(struct net *src_net,
+ struct net_device *dev,
+ struct nlattr *tb[],
+ struct nlattr *data[])
+{
+ struct macvtap_dev *vlantap = netdev_priv(dev);
+ int err;
+
+ INIT_LIST_HEAD(&vlantap->tap.queue_list);
+
+ /* Since macvlan supports all offloads by default, make
+ * tap support all offloads also.
+ */
+ vlantap->tap.tap_features = TUN_OFFLOADS;
+
+ /* Register callbacks for rx/tx drops accounting and updating
+ * net_device features
+ */
+ vlantap->tap.count_tx_dropped = macvtap_count_tx_dropped;
+ vlantap->tap.count_rx_dropped = macvtap_count_rx_dropped;
+ vlantap->tap.update_features = macvtap_update_features;
+
+ err = netdev_rx_handler_register(dev, tap_handle_frame, &vlantap->tap);
+ if (err)
+ return err;
+
+ /* Don't put anything that may fail after macvlan_common_newlink
+ * because we can't undo what it does.
+ */
+ err = macvlan_common_newlink(src_net, dev, tb, data);
+ if (err) {
+ netdev_rx_handler_unregister(dev);
+ return err;
+ }
+
+ vlantap->tap.dev = vlantap->vlan.dev;
+
+ return 0;
+}
+
+static void macvtap_dellink(struct net_device *dev,
+ struct list_head *head)
+{
+ struct macvtap_dev *vlantap = netdev_priv(dev);
+
+ netdev_rx_handler_unregister(dev);
+ tap_del_queues(&vlantap->tap);
+ macvlan_dellink(dev, head);
+}
+
+static void macvtap_setup(struct net_device *dev)
+{
+ macvlan_common_setup(dev);
+ dev->tx_queue_len = TUN_READQ_SIZE;
+}
+
+static struct rtnl_link_ops macvtap_link_ops __read_mostly = {
+ .kind = "macvtap",
+ .setup = macvtap_setup,
+ .newlink = macvtap_newlink,
+ .dellink = macvtap_dellink,
+ .priv_size = sizeof(struct macvtap_dev),
+};
+
+static int macvtap_device_event(struct notifier_block *unused,
+ unsigned long event, void *ptr)
+{
+ struct net_device *dev = netdev_notifier_info_to_dev(ptr);
+ struct macvtap_dev *vlantap;
+ struct device *classdev;
+ dev_t devt;
+ int err;
+ char tap_name[IFNAMSIZ];
+
+ if (dev->rtnl_link_ops != &macvtap_link_ops)
+ return NOTIFY_DONE;
+
+ snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
+ vlantap = netdev_priv(dev);
+
+ switch (event) {
+ case NETDEV_REGISTER:
+ /* Create the device node here after the network device has
+ * been registered but before register_netdevice has
+ * finished running.
+ */
+ err = tap_get_minor(macvtap_major, &vlantap->tap);
+ if (err)
+ return notifier_from_errno(err);
+
+ devt = MKDEV(MAJOR(macvtap_major), vlantap->tap.minor);
+ classdev = device_create(&macvtap_class, &dev->dev, devt,
+ dev, tap_name);
+ if (IS_ERR(classdev)) {
+ tap_free_minor(macvtap_major, &vlantap->tap);
+ return notifier_from_errno(PTR_ERR(classdev));
+ }
+ err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
+ tap_name);
+ if (err)
+ return notifier_from_errno(err);
+ break;
+ case NETDEV_UNREGISTER:
+ /* vlan->minor == 0 if NETDEV_REGISTER above failed */
+ if (vlantap->tap.minor == 0)
+ break;
+ sysfs_remove_link(&dev->dev.kobj, tap_name);
+ devt = MKDEV(MAJOR(macvtap_major), vlantap->tap.minor);
+ device_destroy(&macvtap_class, devt);
+ tap_free_minor(macvtap_major, &vlantap->tap);
+ break;
+ case NETDEV_CHANGE_TX_QUEUE_LEN:
+ if (tap_queue_resize(&vlantap->tap))
+ return NOTIFY_BAD;
+ break;
+ }
+
+ return NOTIFY_DONE;
+}
+
+static struct notifier_block macvtap_notifier_block __read_mostly = {
+ .notifier_call = macvtap_device_event,
+};
+
+static int macvtap_init(void)
+{
+ int err;
+
+ err = tap_create_cdev(&macvtap_cdev, &macvtap_major, "macvtap");
+
+ if (err)
+ goto out1;
+
+ err = class_register(&macvtap_class);
+ if (err)
+ goto out2;
+
+ err = register_netdevice_notifier(&macvtap_notifier_block);
+ if (err)
+ goto out3;
+
+ err = macvlan_link_register(&macvtap_link_ops);
+ if (err)
+ goto out4;
+
+ return 0;
+
+out4:
+ unregister_netdevice_notifier(&macvtap_notifier_block);
+out3:
+ class_unregister(&macvtap_class);
+out2:
+ cdev_del(&macvtap_cdev);
+out1:
+ return err;
+}
+module_init(macvtap_init);
+
+extern struct idr minor_idr;
+static void macvtap_exit(void)
+{
+ rtnl_link_unregister(&macvtap_link_ops);
+ unregister_netdevice_notifier(&macvtap_notifier_block);
+ class_unregister(&macvtap_class);
+ tap_destroy_cdev(macvtap_major, &macvtap_cdev);
+}
+module_exit(macvtap_exit);
+
+MODULE_ALIAS_RTNL_LINK("macvtap");
+MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/net/macvtap_main.c b/drivers/net/macvtap_main.c
deleted file mode 100644
index 3f047b4..0000000
--- a/drivers/net/macvtap_main.c
+++ /dev/null
@@ -1,247 +0,0 @@
-#include <linux/etherdevice.h>
-#include <linux/if_macvlan.h>
-#include <linux/if_tap.h>
-#include <linux/if_vlan.h>
-#include <linux/interrupt.h>
-#include <linux/nsproxy.h>
-#include <linux/compat.h>
-#include <linux/if_tun.h>
-#include <linux/module.h>
-#include <linux/skbuff.h>
-#include <linux/cache.h>
-#include <linux/sched.h>
-#include <linux/types.h>
-#include <linux/slab.h>
-#include <linux/wait.h>
-#include <linux/cdev.h>
-#include <linux/idr.h>
-#include <linux/fs.h>
-#include <linux/uio.h>
-
-#include <net/net_namespace.h>
-#include <net/rtnetlink.h>
-#include <net/sock.h>
-#include <linux/virtio_net.h>
-#include <linux/skb_array.h>
-
-struct macvtap_dev {
- struct macvlan_dev vlan;
- struct tap_dev tap;
-};
-
-/*
- * Variables for dealing with macvtaps device numbers.
- */
-static dev_t macvtap_major;
-
-static const void *macvtap_net_namespace(struct device *d)
-{
- struct net_device *dev = to_net_dev(d->parent);
- return dev_net(dev);
-}
-
-static struct class macvtap_class = {
- .name = "macvtap",
- .owner = THIS_MODULE,
- .ns_type = &net_ns_type_operations,
- .namespace = macvtap_net_namespace,
-};
-static struct cdev macvtap_cdev;
-
-#define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN | NETIF_F_TSO | \
- NETIF_F_TSO6 | NETIF_F_UFO)
-
-static void macvtap_count_tx_dropped(struct tap_dev *tap)
-{
- struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
-
- this_cpu_inc(vlan->pcpu_stats->tx_dropped);
-}
-
-static void macvtap_count_rx_dropped(struct tap_dev *tap)
-{
- struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
-
- macvlan_count_rx(vlan, 0, 0, 0);
-}
-
-static void macvtap_update_features(struct tap_dev *tap,
- netdev_features_t features)
-{
- struct macvlan_dev *vlan = (struct macvlan_dev *)container_of(tap, struct macvtap_dev, tap);
-
- vlan->set_features = features;
- netdev_update_features(vlan->dev);
-}
-
-static int macvtap_newlink(struct net *src_net,
- struct net_device *dev,
- struct nlattr *tb[],
- struct nlattr *data[])
-{
- struct macvtap_dev *vlantap = netdev_priv(dev);
- int err;
-
- INIT_LIST_HEAD(&vlantap->tap.queue_list);
-
- /* Since macvlan supports all offloads by default, make
- * tap support all offloads also.
- */
- vlantap->tap.tap_features = TUN_OFFLOADS;
-
- /* Register callbacks for rx/tx drops accounting and updating
- * net_device features
- */
- vlantap->tap.count_tx_dropped = macvtap_count_tx_dropped;
- vlantap->tap.count_rx_dropped = macvtap_count_rx_dropped;
- vlantap->tap.update_features = macvtap_update_features;
-
- err = netdev_rx_handler_register(dev, tap_handle_frame, &vlantap->tap);
- if (err)
- return err;
-
- /* Don't put anything that may fail after macvlan_common_newlink
- * because we can't undo what it does.
- */
- err = macvlan_common_newlink(src_net, dev, tb, data);
- if (err) {
- netdev_rx_handler_unregister(dev);
- return err;
- }
-
- vlantap->tap.dev = vlantap->vlan.dev;
-
- return 0;
-}
-
-static void macvtap_dellink(struct net_device *dev,
- struct list_head *head)
-{
- struct macvtap_dev *vlantap = netdev_priv(dev);
-
- netdev_rx_handler_unregister(dev);
- tap_del_queues(&vlantap->tap);
- macvlan_dellink(dev, head);
-}
-
-static void macvtap_setup(struct net_device *dev)
-{
- macvlan_common_setup(dev);
- dev->tx_queue_len = TUN_READQ_SIZE;
-}
-
-static struct rtnl_link_ops macvtap_link_ops __read_mostly = {
- .kind = "macvtap",
- .setup = macvtap_setup,
- .newlink = macvtap_newlink,
- .dellink = macvtap_dellink,
- .priv_size = sizeof(struct macvtap_dev),
-};
-
-static int macvtap_device_event(struct notifier_block *unused,
- unsigned long event, void *ptr)
-{
- struct net_device *dev = netdev_notifier_info_to_dev(ptr);
- struct macvtap_dev *vlantap;
- struct device *classdev;
- dev_t devt;
- int err;
- char tap_name[IFNAMSIZ];
-
- if (dev->rtnl_link_ops != &macvtap_link_ops)
- return NOTIFY_DONE;
-
- snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
- vlantap = netdev_priv(dev);
-
- switch (event) {
- case NETDEV_REGISTER:
- /* Create the device node here after the network device has
- * been registered but before register_netdevice has
- * finished running.
- */
- err = tap_get_minor(macvtap_major, &vlantap->tap);
- if (err)
- return notifier_from_errno(err);
-
- devt = MKDEV(MAJOR(macvtap_major), vlantap->tap.minor);
- classdev = device_create(&macvtap_class, &dev->dev, devt,
- dev, tap_name);
- if (IS_ERR(classdev)) {
- tap_free_minor(macvtap_major, &vlantap->tap);
- return notifier_from_errno(PTR_ERR(classdev));
- }
- err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
- tap_name);
- if (err)
- return notifier_from_errno(err);
- break;
- case NETDEV_UNREGISTER:
- /* vlan->minor == 0 if NETDEV_REGISTER above failed */
- if (vlantap->tap.minor == 0)
- break;
- sysfs_remove_link(&dev->dev.kobj, tap_name);
- devt = MKDEV(MAJOR(macvtap_major), vlantap->tap.minor);
- device_destroy(&macvtap_class, devt);
- tap_free_minor(macvtap_major, &vlantap->tap);
- break;
- case NETDEV_CHANGE_TX_QUEUE_LEN:
- if (tap_queue_resize(&vlantap->tap))
- return NOTIFY_BAD;
- break;
- }
-
- return NOTIFY_DONE;
-}
-
-static struct notifier_block macvtap_notifier_block __read_mostly = {
- .notifier_call = macvtap_device_event,
-};
-
-static int macvtap_init(void)
-{
- int err;
-
- err = tap_create_cdev(&macvtap_cdev, &macvtap_major, "macvtap");
-
- if (err)
- goto out1;
-
- err = class_register(&macvtap_class);
- if (err)
- goto out2;
-
- err = register_netdevice_notifier(&macvtap_notifier_block);
- if (err)
- goto out3;
-
- err = macvlan_link_register(&macvtap_link_ops);
- if (err)
- goto out4;
-
- return 0;
-
-out4:
- unregister_netdevice_notifier(&macvtap_notifier_block);
-out3:
- class_unregister(&macvtap_class);
-out2:
- cdev_del(&macvtap_cdev);
-out1:
- return err;
-}
-module_init(macvtap_init);
-
-extern struct idr minor_idr;
-static void macvtap_exit(void)
-{
- rtnl_link_unregister(&macvtap_link_ops);
- unregister_netdevice_notifier(&macvtap_notifier_block);
- class_unregister(&macvtap_class);
- tap_destroy_cdev(macvtap_major, &macvtap_cdev);
-}
-module_exit(macvtap_exit);
-
-MODULE_ALIAS_RTNL_LINK("macvtap");
-MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
-MODULE_LICENSE("GPL");
diff --git a/drivers/net/tap.c b/drivers/net/tap.c
index 1d5bcf3..ee0d49a 100644
--- a/drivers/net/tap.c
+++ b/drivers/net/tap.c
@@ -311,6 +311,7 @@ void tap_del_queues(struct tap_dev *tap)
/* guarantee that any future tap_set_queue will fail */
tap->numvtaps = MAX_TAP_QUEUES;
}
+EXPORT_SYMBOL_GPL(tap_del_queues);
rx_handler_result_t tap_handle_frame(struct sk_buff **pskb)
{
@@ -388,6 +389,7 @@ rx_handler_result_t tap_handle_frame(struct sk_buff **pskb)
kfree_skb(skb);
return RX_HANDLER_CONSUMED;
}
+EXPORT_SYMBOL_GPL(tap_handle_frame);
int tap_get_minor(dev_t major, struct tap_dev *tap)
{
@@ -416,6 +418,7 @@ int tap_get_minor(dev_t major, struct tap_dev *tap)
mutex_unlock(&tap_major->minor_lock);
return retval < 0 ? retval : 0;
}
+EXPORT_SYMBOL_GPL(tap_get_minor);
void tap_free_minor(dev_t major, struct tap_dev *tap)
{
@@ -439,6 +442,7 @@ void tap_free_minor(dev_t major, struct tap_dev *tap)
}
mutex_unlock(&tap_major->minor_lock);
}
+EXPORT_SYMBOL_GPL(tap_free_minor);
static struct tap_dev *dev_get_by_tap_file(int major, int minor)
{
@@ -1201,6 +1205,7 @@ int tap_queue_resize(struct tap_dev *tap)
kfree(arrays);
return ret;
}
+EXPORT_SYMBOL_GPL(tap_queue_resize);
static int tap_list_add(dev_t major, const char *device_name)
{
@@ -1244,6 +1249,7 @@ int tap_create_cdev(struct cdev *tap_cdev,
out1:
return err;
}
+EXPORT_SYMBOL_GPL(tap_create_cdev);
void tap_destroy_cdev(dev_t major, struct cdev *tap_cdev)
{
@@ -1264,3 +1270,7 @@ void tap_destroy_cdev(dev_t major, struct cdev *tap_cdev)
unregister_chrdev_region(major, TAP_NUM_DEVS);
idr_destroy(&tap_major->minor_idr);
}
+EXPORT_SYMBOL_GPL(tap_destroy_cdev);
+
+MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
+MODULE_LICENSE("GPL");
--
2.7.4
^ permalink raw reply related [flat|nested] 21+ messages in thread* Re: [PATCHv1 6/7] TAP: tap as an independent module
2017-01-06 22:33 ` [PATCHv1 6/7] TAP: tap as an independent module Sainath Grandhi
@ 2017-01-07 10:46 ` kbuild test robot
2017-01-07 12:32 ` kbuild test robot
1 sibling, 0 replies; 21+ messages in thread
From: kbuild test robot @ 2017-01-07 10:46 UTC (permalink / raw)
To: Sainath Grandhi
Cc: kbuild-all, netdev, davem, mahesh, linux-kernel, Sainath Grandhi
[-- Attachment #1: Type: text/plain, Size: 1468 bytes --]
Hi Sainath,
[auto build test ERROR on net/master]
[also build test ERROR on v4.10-rc2 next-20170106]
[cannot apply to net-next/master]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]
url: https://github.com/0day-ci/linux/commits/Sainath-Grandhi/Refactor-macvtap-to-re-use-tap-functionality-by-other-virtual-intefaces/20170107-165707
config: s390-default_defconfig (attached as .config)
compiler: s390x-linux-gnu-gcc (Debian 6.1.1-9) 6.1.1 20160705
reproduce:
wget https://git.kernel.org/cgit/linux/kernel/git/wfg/lkp-tests.git/plain/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# save the attached .config to linux build tree
make.cross ARCH=s390
All errors (new ones prefixed by >>):
ERROR: "tap_get_socket" [drivers/vhost/vhost_net.ko] undefined!
>> ERROR: "tap_create_cdev" [drivers/net/macvtap.ko] undefined!
>> ERROR: "tap_queue_resize" [drivers/net/macvtap.ko] undefined!
>> ERROR: "tap_get_minor" [drivers/net/macvtap.ko] undefined!
>> ERROR: "tap_destroy_cdev" [drivers/net/macvtap.ko] undefined!
>> ERROR: "tap_handle_frame" [drivers/net/macvtap.ko] undefined!
>> ERROR: "tap_free_minor" [drivers/net/macvtap.ko] undefined!
>> ERROR: "tap_del_queues" [drivers/net/macvtap.ko] undefined!
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 16728 bytes --]
^ permalink raw reply [flat|nested] 21+ messages in thread* Re: [PATCHv1 6/7] TAP: tap as an independent module
2017-01-06 22:33 ` [PATCHv1 6/7] TAP: tap as an independent module Sainath Grandhi
2017-01-07 10:46 ` kbuild test robot
@ 2017-01-07 12:32 ` kbuild test robot
1 sibling, 0 replies; 21+ messages in thread
From: kbuild test robot @ 2017-01-07 12:32 UTC (permalink / raw)
To: Sainath Grandhi
Cc: kbuild-all, netdev, davem, mahesh, linux-kernel, Sainath Grandhi
[-- Attachment #1: Type: text/plain, Size: 1746 bytes --]
Hi Sainath,
[auto build test ERROR on net/master]
[also build test ERROR on v4.10-rc2 next-20170106]
[cannot apply to net-next/master]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]
url: https://github.com/0day-ci/linux/commits/Sainath-Grandhi/Refactor-macvtap-to-re-use-tap-functionality-by-other-virtual-intefaces/20170107-165707
config: x86_64-allyesdebian (attached as .config)
compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
reproduce:
# save the attached .config to linux build tree
make ARCH=x86_64
All errors (new ones prefixed by >>):
drivers/built-in.o: In function `macvtap_dellink':
>> macvtap.c:(.text+0x7d3255): undefined reference to `tap_del_queues'
drivers/built-in.o: In function `macvtap_exit':
>> macvtap.c:(.text+0x7d32e2): undefined reference to `tap_destroy_cdev'
drivers/built-in.o: In function `macvtap_init':
>> macvtap.c:(.text+0x7d3308): undefined reference to `tap_create_cdev'
drivers/built-in.o: In function `macvtap_newlink':
>> macvtap.c:(.text+0x7d33df): undefined reference to `tap_handle_frame'
drivers/built-in.o: In function `macvtap_device_event':
>> macvtap.c:(.text+0x7d34b5): undefined reference to `tap_get_minor'
>> macvtap.c:(.text+0x7d34fe): undefined reference to `tap_free_minor'
macvtap.c:(.text+0x7d3586): undefined reference to `tap_free_minor'
>> macvtap.c:(.text+0x7d3594): undefined reference to `tap_queue_resize'
drivers/built-in.o: In function `vhost_net_ioctl':
net.c:(.text+0x11835a3): undefined reference to `tap_get_socket'
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 38068 bytes --]
^ permalink raw reply [flat|nested] 21+ messages in thread
* [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver
2017-01-06 22:33 [PATCHv1 0/7] Refactor macvtap to re-use tap functionality by other virtual intefaces Sainath Grandhi
` (5 preceding siblings ...)
2017-01-06 22:33 ` [PATCHv1 6/7] TAP: tap as an independent module Sainath Grandhi
@ 2017-01-06 22:33 ` Sainath Grandhi
2017-01-06 23:14 ` Eric Dumazet
2017-01-06 23:46 ` Mahesh Bandewar (महेश बंडेवार)
6 siblings, 2 replies; 21+ messages in thread
From: Sainath Grandhi @ 2017-01-06 22:33 UTC (permalink / raw)
To: netdev; +Cc: davem, mahesh, linux-kernel, Sainath Grandhi
This patch adds a tap character device driver that is based on the
IP-VLAN network interface, called ipvtap. An ipvtap device can be created
in the same way as an ipvlan device, using 'type ipvtap', and then accessed
using the tap user space interface.
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
---
drivers/net/Kconfig | 12 ++
drivers/net/Makefile | 1 +
drivers/net/ipvlan/Makefile | 1 +
drivers/net/ipvlan/ipvlan.h | 7 ++
drivers/net/ipvlan/ipvlan_core.c | 5 +-
drivers/net/ipvlan/ipvlan_main.c | 37 +++---
drivers/net/ipvlan/ipvtap.c | 238 +++++++++++++++++++++++++++++++++++++++
7 files changed, 282 insertions(+), 19 deletions(-)
create mode 100644 drivers/net/ipvlan/ipvtap.c
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index 280380d..ddfb30a 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -165,6 +165,18 @@ config IPVLAN
To compile this driver as a module, choose M here: the module
will be called ipvlan.
+config IPVTAP
+ tristate "IP-VLAN based tap driver"
+ depends on IPVLAN
+ depends on INET
+ help
+ This adds a specialized tap character device driver that is based
+ on the IP-VLAN network interface, called ipvtap. An ipvtap device
+ can be added in the same way as a ipvlan device, using 'type
+ ipvtap', and then be accessed through the tap user space interface.
+
+ To compile this driver as a module, choose M here: the module
+ will be called macvtap.
config VXLAN
tristate "Virtual eXtensible Local Area Network (VXLAN)"
diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index 7dd86ca..98ed4d9 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -7,6 +7,7 @@
#
obj-$(CONFIG_BONDING) += bonding/
obj-$(CONFIG_IPVLAN) += ipvlan/
+obj-$(CONFIG_IPVTAP) += ipvlan/
obj-$(CONFIG_DUMMY) += dummy.o
obj-$(CONFIG_EQUALIZER) += eql.o
obj-$(CONFIG_IFB) += ifb.o
diff --git a/drivers/net/ipvlan/Makefile b/drivers/net/ipvlan/Makefile
index df79910..8a2c64d 100644
--- a/drivers/net/ipvlan/Makefile
+++ b/drivers/net/ipvlan/Makefile
@@ -3,5 +3,6 @@
#
obj-$(CONFIG_IPVLAN) += ipvlan.o
+obj-$(CONFIG_IPVTAP) += ipvtap.o
ipvlan-objs := ipvlan_core.o ipvlan_main.o
diff --git a/drivers/net/ipvlan/ipvlan.h b/drivers/net/ipvlan/ipvlan.h
index dbfbb33..4362d88 100644
--- a/drivers/net/ipvlan/ipvlan.h
+++ b/drivers/net/ipvlan/ipvlan.h
@@ -133,4 +133,11 @@ struct sk_buff *ipvlan_l3_rcv(struct net_device *dev, struct sk_buff *skb,
u16 proto);
unsigned int ipvlan_nf_input(void *priv, struct sk_buff *skb,
const struct nf_hook_state *state);
+void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
+ unsigned int len, bool success, bool mcast);
+int ipvlan_link_new(struct net *src_net, struct net_device *dev,
+ struct nlattr *tb[], struct nlattr *data[]);
+void ipvlan_link_delete(struct net_device *dev, struct list_head *head);
+void ipvlan_link_setup(struct net_device *dev);
+int ipvlan_link_register(struct rtnl_link_ops *ops);
#endif /* __IPVLAN_H */
diff --git a/drivers/net/ipvlan/ipvlan_core.c b/drivers/net/ipvlan/ipvlan_core.c
index 83ce74a..9af16ab 100644
--- a/drivers/net/ipvlan/ipvlan_core.c
+++ b/drivers/net/ipvlan/ipvlan_core.c
@@ -16,8 +16,8 @@ void ipvlan_init_secret(void)
net_get_random_once(&ipvlan_jhash_secret, sizeof(ipvlan_jhash_secret));
}
-static void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
- unsigned int len, bool success, bool mcast)
+void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
+ unsigned int len, bool success, bool mcast)
{
if (!ipvlan)
return;
@@ -36,6 +36,7 @@ static void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
this_cpu_inc(ipvlan->pcpu_stats->rx_errs);
}
}
+EXPORT_SYMBOL_GPL(ipvlan_count_rx);
static u8 ipvlan_get_v6_hash(const void *iaddr)
{
diff --git a/drivers/net/ipvlan/ipvlan_main.c b/drivers/net/ipvlan/ipvlan_main.c
index 8b0f993..666a05d 100644
--- a/drivers/net/ipvlan/ipvlan_main.c
+++ b/drivers/net/ipvlan/ipvlan_main.c
@@ -13,14 +13,14 @@ static u32 ipvl_nf_hook_refcnt = 0;
static struct nf_hook_ops ipvl_nfops[] __read_mostly = {
{
- .hook = ipvlan_nf_input,
- .pf = NFPROTO_IPV4,
+ .hook = ipvlan_nf_input,
+ .pf = NFPROTO_IPV4,
.hooknum = NF_INET_LOCAL_IN,
.priority = INT_MAX,
},
{
- .hook = ipvlan_nf_input,
- .pf = NFPROTO_IPV6,
+ .hook = ipvlan_nf_input,
+ .pf = NFPROTO_IPV6,
.hooknum = NF_INET_LOCAL_IN,
.priority = INT_MAX,
},
@@ -398,7 +398,7 @@ static int ipvlan_hard_header(struct sk_buff *skb, struct net_device *dev,
}
static const struct header_ops ipvlan_header_ops = {
- .create = ipvlan_hard_header,
+ .create = ipvlan_hard_header,
.parse = eth_header_parse,
.cache = eth_header_cache,
.cache_update = eth_header_cache_update,
@@ -494,8 +494,8 @@ static int ipvlan_nl_fillinfo(struct sk_buff *skb,
return ret;
}
-static int ipvlan_link_new(struct net *src_net, struct net_device *dev,
- struct nlattr *tb[], struct nlattr *data[])
+int ipvlan_link_new(struct net *src_net, struct net_device *dev,
+ struct nlattr *tb[], struct nlattr *data[])
{
struct ipvl_dev *ipvlan = netdev_priv(dev);
struct ipvl_port *port;
@@ -567,8 +567,9 @@ static int ipvlan_link_new(struct net *src_net, struct net_device *dev,
ipvlan_port_destroy(phy_dev);
return err;
}
+EXPORT_SYMBOL_GPL(ipvlan_link_new);
-static void ipvlan_link_delete(struct net_device *dev, struct list_head *head)
+void ipvlan_link_delete(struct net_device *dev, struct list_head *head)
{
struct ipvl_dev *ipvlan = netdev_priv(dev);
struct ipvl_addr *addr, *next;
@@ -583,8 +584,9 @@ static void ipvlan_link_delete(struct net_device *dev, struct list_head *head)
unregister_netdevice_queue(dev, head);
netdev_upper_dev_unlink(ipvlan->phy_dev, dev);
}
+EXPORT_SYMBOL_GPL(ipvlan_link_delete);
-static void ipvlan_link_setup(struct net_device *dev)
+void ipvlan_link_setup(struct net_device *dev)
{
ether_setup(dev);
@@ -595,6 +597,7 @@ static void ipvlan_link_setup(struct net_device *dev)
dev->header_ops = &ipvlan_header_ops;
dev->ethtool_ops = &ipvlan_ethtool_ops;
}
+EXPORT_SYMBOL_GPL(ipvlan_link_setup);
static const struct nla_policy ipvlan_nl_policy[IFLA_IPVLAN_MAX + 1] =
{
@@ -605,22 +608,22 @@ static struct rtnl_link_ops ipvlan_link_ops = {
.kind = "ipvlan",
.priv_size = sizeof(struct ipvl_dev),
- .get_size = ipvlan_nl_getsize,
- .policy = ipvlan_nl_policy,
- .validate = ipvlan_nl_validate,
- .fill_info = ipvlan_nl_fillinfo,
- .changelink = ipvlan_nl_changelink,
- .maxtype = IFLA_IPVLAN_MAX,
-
.setup = ipvlan_link_setup,
.newlink = ipvlan_link_new,
.dellink = ipvlan_link_delete,
};
-static int ipvlan_link_register(struct rtnl_link_ops *ops)
+int ipvlan_link_register(struct rtnl_link_ops *ops)
{
+ ops->get_size = ipvlan_nl_getsize;
+ ops->policy = ipvlan_nl_policy;
+ ops->validate = ipvlan_nl_validate;
+ ops->fill_info = ipvlan_nl_fillinfo;
+ ops->changelink = ipvlan_nl_changelink;
+ ops->maxtype = IFLA_IPVLAN_MAX;
return rtnl_link_register(ops);
}
+EXPORT_SYMBOL_GPL(ipvlan_link_register);
static int ipvlan_device_event(struct notifier_block *unused,
unsigned long event, void *ptr)
diff --git a/drivers/net/ipvlan/ipvtap.c b/drivers/net/ipvlan/ipvtap.c
new file mode 100644
index 0000000..0422cdf
--- /dev/null
+++ b/drivers/net/ipvlan/ipvtap.c
@@ -0,0 +1,238 @@
+#include <linux/etherdevice.h>
+#include "ipvlan.h"
+#include <linux/if_vlan.h>
+#include <linux/if_tap.h>
+#include <linux/interrupt.h>
+#include <linux/nsproxy.h>
+#include <linux/compat.h>
+#include <linux/if_tun.h>
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <linux/cache.h>
+#include <linux/sched.h>
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/wait.h>
+#include <linux/cdev.h>
+#include <linux/idr.h>
+#include <linux/fs.h>
+#include <linux/uio.h>
+
+#include <net/net_namespace.h>
+#include <net/rtnetlink.h>
+#include <net/sock.h>
+#include <linux/virtio_net.h>
+
+#define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN | NETIF_F_TSO | \
+ NETIF_F_TSO6 | NETIF_F_UFO)
+
+static dev_t ipvtap_major;
+static struct cdev ipvtap_cdev;
+
+static const void *ipvtap_net_namespace(struct device *d)
+{
+ struct net_device *dev = to_net_dev(d->parent);
+ return dev_net(dev);
+}
+
+static struct class ipvtap_class = {
+ .name = "ipvtap",
+ .owner = THIS_MODULE,
+ .ns_type = &net_ns_type_operations,
+ .namespace = ipvtap_net_namespace,
+};
+
+struct ipvtap_dev {
+ struct ipvl_dev vlan;
+ struct tap_dev tap;
+};
+
+static void ipvtap_count_tx_dropped(struct tap_dev *tap)
+{
+ struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct ipvtap_dev, tap);
+
+ this_cpu_inc(vlan->pcpu_stats->tx_drps);
+}
+
+static void ipvtap_count_rx_dropped(struct tap_dev *tap)
+{
+ struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct ipvtap_dev, tap);
+
+ ipvlan_count_rx(vlan, 0, 0, 0);
+}
+
+static void ipvtap_update_features(struct tap_dev *tap,
+ netdev_features_t features)
+{
+ struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct ipvtap_dev, tap);
+
+ vlan->sfeatures = features;
+ netdev_update_features(vlan->dev);
+}
+
+static int ipvtap_newlink(struct net *src_net,
+ struct net_device *dev,
+ struct nlattr *tb[],
+ struct nlattr *data[])
+{
+ struct ipvtap_dev *vlantap = netdev_priv(dev);
+ int err;
+
+ INIT_LIST_HEAD(&vlantap->tap.queue_list);
+
+ /* Since macvlan supports all offloads by default, make
+ * tap support all offloads also.
+ */
+ vlantap->tap.tap_features = TUN_OFFLOADS;
+ vlantap->tap.count_tx_dropped = ipvtap_count_tx_dropped;
+ vlantap->tap.update_features = ipvtap_update_features;
+ vlantap->tap.count_rx_dropped = ipvtap_count_rx_dropped;
+
+ err = netdev_rx_handler_register(dev, tap_handle_frame, &vlantap->tap);
+ if (err)
+ return err;
+
+ /* Don't put anything that may fail after macvlan_common_newlink
+ * because we can't undo what it does.
+ */
+ err = ipvlan_link_new(src_net, dev, tb, data);
+ if (err) {
+ netdev_rx_handler_unregister(dev);
+ return err;
+ }
+
+ vlantap->tap.dev = vlantap->vlan.dev;
+
+ return err;
+}
+
+static void ipvtap_dellink(struct net_device *dev,
+ struct list_head *head)
+{
+ struct ipvtap_dev *vlan = netdev_priv(dev);
+
+ netdev_rx_handler_unregister(dev);
+ tap_del_queues(&vlan->tap);
+ ipvlan_link_delete(dev, head);
+}
+
+static void ipvtap_setup(struct net_device *dev)
+{
+ ipvlan_link_setup(dev);
+ dev->tx_queue_len = TUN_READQ_SIZE;
+ dev->priv_flags &= ~IFF_NO_QUEUE;
+}
+
+static struct rtnl_link_ops ipvtap_link_ops __read_mostly = {
+ .kind = "ipvtap",
+ .setup = ipvtap_setup,
+ .newlink = ipvtap_newlink,
+ .dellink = ipvtap_dellink,
+ .priv_size = sizeof(struct ipvtap_dev),
+};
+
+static int ipvtap_device_event(struct notifier_block *unused,
+ unsigned long event, void *ptr)
+{
+ struct net_device *dev = netdev_notifier_info_to_dev(ptr);
+ struct ipvtap_dev *vlantap;
+ struct device *classdev;
+ dev_t devt;
+ int err;
+ char tap_name[IFNAMSIZ];
+
+ if (dev->rtnl_link_ops != &ipvtap_link_ops)
+ return NOTIFY_DONE;
+
+ snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
+ vlantap = netdev_priv(dev);
+
+ switch (event) {
+ case NETDEV_REGISTER:
+ /* Create the device node here after the network device has
+ * been registered but before register_netdevice has
+ * finished running.
+ */
+ err = tap_get_minor(ipvtap_major, &vlantap->tap);
+ if (err)
+ return notifier_from_errno(err);
+
+ devt = MKDEV(MAJOR(ipvtap_major), vlantap->tap.minor);
+ classdev = device_create(&ipvtap_class, &dev->dev, devt,
+ dev, tap_name);
+ if (IS_ERR(classdev)) {
+ tap_free_minor(ipvtap_major, &vlantap->tap);
+ return notifier_from_errno(PTR_ERR(classdev));
+ }
+ err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
+ tap_name);
+ if (err)
+ return notifier_from_errno(err);
+ break;
+ case NETDEV_UNREGISTER:
+ /* vlan->minor == 0 if NETDEV_REGISTER above failed */
+ if (vlantap->tap.minor == 0)
+ break;
+ sysfs_remove_link(&dev->dev.kobj, tap_name);
+ devt = MKDEV(MAJOR(ipvtap_major), vlantap->tap.minor);
+ device_destroy(&ipvtap_class, devt);
+ tap_free_minor(ipvtap_major, &vlantap->tap);
+ break;
+ case NETDEV_CHANGE_TX_QUEUE_LEN:
+ if (tap_queue_resize(&vlantap->tap))
+ return NOTIFY_BAD;
+ break;
+ }
+
+ return NOTIFY_DONE;
+}
+
+static struct notifier_block ipvtap_notifier_block __read_mostly = {
+ .notifier_call = ipvtap_device_event,
+};
+
+static int ipvtap_init(void)
+{
+ int err;
+
+ err = tap_create_cdev(&ipvtap_cdev, &ipvtap_major, "ipvtap");
+
+ if (err)
+ goto out1;
+
+ err = class_register(&ipvtap_class);
+ if (err)
+ goto out2;
+
+ err = register_netdevice_notifier(&ipvtap_notifier_block);
+ if (err)
+ goto out3;
+
+ err = ipvlan_link_register(&ipvtap_link_ops);
+ if (err)
+ goto out4;
+
+ return 0;
+
+out4:
+ unregister_netdevice_notifier(&ipvtap_notifier_block);
+out3:
+ class_unregister(&ipvtap_class);
+out2:
+ cdev_del(&ipvtap_cdev);
+out1:
+ return err;
+}
+module_init(ipvtap_init);
+
+static void ipvtap_exit(void)
+{
+ rtnl_link_unregister(&ipvtap_link_ops);
+ unregister_netdevice_notifier(&ipvtap_notifier_block);
+ class_unregister(&ipvtap_class);
+ tap_destroy_cdev(ipvtap_major, &ipvtap_cdev);
+}
+module_exit(ipvtap_exit);
+MODULE_ALIAS_RTNL_LINK("ipvtap");
+MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
+MODULE_LICENSE("GPL");
--
2.7.4
^ permalink raw reply related [flat|nested] 21+ messages in thread* Re: [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver
2017-01-06 22:33 ` [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver Sainath Grandhi
@ 2017-01-06 23:14 ` Eric Dumazet
2017-01-17 23:06 ` Grandhi, Sainath
2017-01-06 23:46 ` Mahesh Bandewar (महेश बंडेवार)
1 sibling, 1 reply; 21+ messages in thread
From: Eric Dumazet @ 2017-01-06 23:14 UTC (permalink / raw)
To: Sainath Grandhi; +Cc: netdev, davem, mahesh, linux-kernel
On Fri, 2017-01-06 at 14:33 -0800, Sainath Grandhi wrote:
> This patch adds a tap character device driver that is based on the
> IP-VLAN network interface, called ipvtap. An ipvtap device can be created
> in the same way as an ipvlan device, using 'type ipvtap', and then accessed
> using the tap user space interface.
>
> Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
> Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
> ---
> +module_exit(ipvtap_exit);
> +MODULE_ALIAS_RTNL_LINK("ipvtap");
> +MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
> +MODULE_LICENSE("GPL");
Who wrote this driver exactly ???
^ permalink raw reply [flat|nested] 21+ messages in thread* RE: [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver
2017-01-06 23:14 ` Eric Dumazet
@ 2017-01-17 23:06 ` Grandhi, Sainath
0 siblings, 0 replies; 21+ messages in thread
From: Grandhi, Sainath @ 2017-01-17 23:06 UTC (permalink / raw)
To: Eric Dumazet
Cc: netdev@vger.kernel.org, davem@davemloft.net, mahesh@bandewar.net,
linux-kernel@vger.kernel.org
> -----Original Message-----
> From: Eric Dumazet [mailto:eric.dumazet@gmail.com]
> Sent: Friday, January 06, 2017 3:14 PM
> To: Grandhi, Sainath <sainath.grandhi@intel.com>
> Cc: netdev@vger.kernel.org; davem@davemloft.net;
> mahesh@bandewar.net; linux-kernel@vger.kernel.org
> Subject: Re: [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver
>
> On Fri, 2017-01-06 at 14:33 -0800, Sainath Grandhi wrote:
> > This patch adds a tap character device driver that is based on the
> > IP-VLAN network interface, called ipvtap. An ipvtap device can be
> > created in the same way as an ipvlan device, using 'type ipvtap', and
> > then accessed using the tap user space interface.
> >
> > Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
> > Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
> > ---
>
>
> > +module_exit(ipvtap_exit);
> > +MODULE_ALIAS_RTNL_LINK("ipvtap");
> > +MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
> > +MODULE_LICENSE("GPL");
>
> Who wrote this driver exactly ???
>
>
Sending out next version, modifying this.
^ permalink raw reply [flat|nested] 21+ messages in thread
* Re: [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver
2017-01-06 22:33 ` [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver Sainath Grandhi
2017-01-06 23:14 ` Eric Dumazet
@ 2017-01-06 23:46 ` Mahesh Bandewar (महेश बंडेवार)
2017-01-17 23:49 ` Grandhi, Sainath
1 sibling, 1 reply; 21+ messages in thread
From: Mahesh Bandewar (महेश बंडेवार) @ 2017-01-06 23:46 UTC (permalink / raw)
To: Sainath Grandhi; +Cc: linux-netdev, David Miller, mahesh, linux-kernel
few superficial comments inline.
On Fri, Jan 6, 2017 at 2:33 PM, Sainath Grandhi
<sainath.grandhi@intel.com> wrote:
> This patch adds a tap character device driver that is based on the
> IP-VLAN network interface, called ipvtap. An ipvtap device can be created
> in the same way as an ipvlan device, using 'type ipvtap', and then accessed
> using the tap user space interface.
>
> Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
> Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
> ---
> drivers/net/Kconfig | 12 ++
> drivers/net/Makefile | 1 +
> drivers/net/ipvlan/Makefile | 1 +
> drivers/net/ipvlan/ipvlan.h | 7 ++
> drivers/net/ipvlan/ipvlan_core.c | 5 +-
> drivers/net/ipvlan/ipvlan_main.c | 37 +++---
> drivers/net/ipvlan/ipvtap.c | 238 +++++++++++++++++++++++++++++++++++++++
> 7 files changed, 282 insertions(+), 19 deletions(-)
> create mode 100644 drivers/net/ipvlan/ipvtap.c
>
> diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
> index 280380d..ddfb30a 100644
> --- a/drivers/net/Kconfig
> +++ b/drivers/net/Kconfig
> @@ -165,6 +165,18 @@ config IPVLAN
> To compile this driver as a module, choose M here: the module
> will be called ipvlan.
>
> +config IPVTAP
> + tristate "IP-VLAN based tap driver"
> + depends on IPVLAN
> + depends on INET
> + help
> + This adds a specialized tap character device driver that is based
> + on the IP-VLAN network interface, called ipvtap. An ipvtap device
> + can be added in the same way as a ipvlan device, using 'type
> + ipvtap', and then be accessed through the tap user space interface.
> +
> + To compile this driver as a module, choose M here: the module
> + will be called macvtap.
>
> config VXLAN
> tristate "Virtual eXtensible Local Area Network (VXLAN)"
> diff --git a/drivers/net/Makefile b/drivers/net/Makefile
> index 7dd86ca..98ed4d9 100644
> --- a/drivers/net/Makefile
> +++ b/drivers/net/Makefile
> @@ -7,6 +7,7 @@
> #
> obj-$(CONFIG_BONDING) += bonding/
> obj-$(CONFIG_IPVLAN) += ipvlan/
> +obj-$(CONFIG_IPVTAP) += ipvlan/
> obj-$(CONFIG_DUMMY) += dummy.o
> obj-$(CONFIG_EQUALIZER) += eql.o
> obj-$(CONFIG_IFB) += ifb.o
> diff --git a/drivers/net/ipvlan/Makefile b/drivers/net/ipvlan/Makefile
> index df79910..8a2c64d 100644
> --- a/drivers/net/ipvlan/Makefile
> +++ b/drivers/net/ipvlan/Makefile
> @@ -3,5 +3,6 @@
> #
>
> obj-$(CONFIG_IPVLAN) += ipvlan.o
> +obj-$(CONFIG_IPVTAP) += ipvtap.o
>
> ipvlan-objs := ipvlan_core.o ipvlan_main.o
> diff --git a/drivers/net/ipvlan/ipvlan.h b/drivers/net/ipvlan/ipvlan.h
> index dbfbb33..4362d88 100644
> --- a/drivers/net/ipvlan/ipvlan.h
> +++ b/drivers/net/ipvlan/ipvlan.h
> @@ -133,4 +133,11 @@ struct sk_buff *ipvlan_l3_rcv(struct net_device *dev, struct sk_buff *skb,
> u16 proto);
> unsigned int ipvlan_nf_input(void *priv, struct sk_buff *skb,
> const struct nf_hook_state *state);
> +void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
> + unsigned int len, bool success, bool mcast);
> +int ipvlan_link_new(struct net *src_net, struct net_device *dev,
> + struct nlattr *tb[], struct nlattr *data[]);
> +void ipvlan_link_delete(struct net_device *dev, struct list_head *head);
> +void ipvlan_link_setup(struct net_device *dev);
> +int ipvlan_link_register(struct rtnl_link_ops *ops);
> #endif /* __IPVLAN_H */
> diff --git a/drivers/net/ipvlan/ipvlan_core.c b/drivers/net/ipvlan/ipvlan_core.c
> index 83ce74a..9af16ab 100644
> --- a/drivers/net/ipvlan/ipvlan_core.c
> +++ b/drivers/net/ipvlan/ipvlan_core.c
> @@ -16,8 +16,8 @@ void ipvlan_init_secret(void)
> net_get_random_once(&ipvlan_jhash_secret, sizeof(ipvlan_jhash_secret));
> }
>
> -static void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
> - unsigned int len, bool success, bool mcast)
> +void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
> + unsigned int len, bool success, bool mcast)
> {
> if (!ipvlan)
> return;
> @@ -36,6 +36,7 @@ static void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
> this_cpu_inc(ipvlan->pcpu_stats->rx_errs);
> }
> }
> +EXPORT_SYMBOL_GPL(ipvlan_count_rx);
Why export, isn't just removing 'static' enough?
>
> static u8 ipvlan_get_v6_hash(const void *iaddr)
> {
> diff --git a/drivers/net/ipvlan/ipvlan_main.c b/drivers/net/ipvlan/ipvlan_main.c
> index 8b0f993..666a05d 100644
> --- a/drivers/net/ipvlan/ipvlan_main.c
> +++ b/drivers/net/ipvlan/ipvlan_main.c
> @@ -13,14 +13,14 @@ static u32 ipvl_nf_hook_refcnt = 0;
>
> static struct nf_hook_ops ipvl_nfops[] __read_mostly = {
> {
> - .hook = ipvlan_nf_input,
> - .pf = NFPROTO_IPV4,
> + .hook = ipvlan_nf_input,
> + .pf = NFPROTO_IPV4,
spurious?
> .hooknum = NF_INET_LOCAL_IN,
> .priority = INT_MAX,
> },
> {
> - .hook = ipvlan_nf_input,
> - .pf = NFPROTO_IPV6,
> + .hook = ipvlan_nf_input,
> + .pf = NFPROTO_IPV6,
spurious??
> .hooknum = NF_INET_LOCAL_IN,
> .priority = INT_MAX,
> },
> @@ -398,7 +398,7 @@ static int ipvlan_hard_header(struct sk_buff *skb, struct net_device *dev,
> }
>
> static const struct header_ops ipvlan_header_ops = {
> - .create = ipvlan_hard_header,
> + .create = ipvlan_hard_header,
spurious???
> .parse = eth_header_parse,
> .cache = eth_header_cache,
> .cache_update = eth_header_cache_update,
> @@ -494,8 +494,8 @@ static int ipvlan_nl_fillinfo(struct sk_buff *skb,
> return ret;
> }
>
> -static int ipvlan_link_new(struct net *src_net, struct net_device *dev,
> - struct nlattr *tb[], struct nlattr *data[])
> +int ipvlan_link_new(struct net *src_net, struct net_device *dev,
> + struct nlattr *tb[], struct nlattr *data[])
> {
> struct ipvl_dev *ipvlan = netdev_priv(dev);
> struct ipvl_port *port;
> @@ -567,8 +567,9 @@ static int ipvlan_link_new(struct net *src_net, struct net_device *dev,
> ipvlan_port_destroy(phy_dev);
> return err;
> }
> +EXPORT_SYMBOL_GPL(ipvlan_link_new);
>
> -static void ipvlan_link_delete(struct net_device *dev, struct list_head *head)
> +void ipvlan_link_delete(struct net_device *dev, struct list_head *head)
> {
> struct ipvl_dev *ipvlan = netdev_priv(dev);
> struct ipvl_addr *addr, *next;
> @@ -583,8 +584,9 @@ static void ipvlan_link_delete(struct net_device *dev, struct list_head *head)
> unregister_netdevice_queue(dev, head);
> netdev_upper_dev_unlink(ipvlan->phy_dev, dev);
> }
> +EXPORT_SYMBOL_GPL(ipvlan_link_delete);
>
> -static void ipvlan_link_setup(struct net_device *dev)
> +void ipvlan_link_setup(struct net_device *dev)
> {
> ether_setup(dev);
>
> @@ -595,6 +597,7 @@ static void ipvlan_link_setup(struct net_device *dev)
> dev->header_ops = &ipvlan_header_ops;
> dev->ethtool_ops = &ipvlan_ethtool_ops;
> }
> +EXPORT_SYMBOL_GPL(ipvlan_link_setup);
>
> static const struct nla_policy ipvlan_nl_policy[IFLA_IPVLAN_MAX + 1] =
> {
> @@ -605,22 +608,22 @@ static struct rtnl_link_ops ipvlan_link_ops = {
> .kind = "ipvlan",
> .priv_size = sizeof(struct ipvl_dev),
>
> - .get_size = ipvlan_nl_getsize,
> - .policy = ipvlan_nl_policy,
> - .validate = ipvlan_nl_validate,
> - .fill_info = ipvlan_nl_fillinfo,
> - .changelink = ipvlan_nl_changelink,
> - .maxtype = IFLA_IPVLAN_MAX,
> -
> .setup = ipvlan_link_setup,
> .newlink = ipvlan_link_new,
> .dellink = ipvlan_link_delete,
> };
>
> -static int ipvlan_link_register(struct rtnl_link_ops *ops)
> +int ipvlan_link_register(struct rtnl_link_ops *ops)
> {
> + ops->get_size = ipvlan_nl_getsize;
> + ops->policy = ipvlan_nl_policy;
> + ops->validate = ipvlan_nl_validate;
> + ops->fill_info = ipvlan_nl_fillinfo;
> + ops->changelink = ipvlan_nl_changelink;
> + ops->maxtype = IFLA_IPVLAN_MAX;
> return rtnl_link_register(ops);
> }
> +EXPORT_SYMBOL_GPL(ipvlan_link_register);
>
> static int ipvlan_device_event(struct notifier_block *unused,
> unsigned long event, void *ptr)
> diff --git a/drivers/net/ipvlan/ipvtap.c b/drivers/net/ipvlan/ipvtap.c
> new file mode 100644
> index 0000000..0422cdf
> --- /dev/null
> +++ b/drivers/net/ipvlan/ipvtap.c
> @@ -0,0 +1,238 @@
> +#include <linux/etherdevice.h>
> +#include "ipvlan.h"
> +#include <linux/if_vlan.h>
> +#include <linux/if_tap.h>
> +#include <linux/interrupt.h>
> +#include <linux/nsproxy.h>
> +#include <linux/compat.h>
> +#include <linux/if_tun.h>
> +#include <linux/module.h>
> +#include <linux/skbuff.h>
> +#include <linux/cache.h>
> +#include <linux/sched.h>
> +#include <linux/types.h>
> +#include <linux/slab.h>
> +#include <linux/wait.h>
> +#include <linux/cdev.h>
> +#include <linux/idr.h>
> +#include <linux/fs.h>
> +#include <linux/uio.h>
> +
> +#include <net/net_namespace.h>
> +#include <net/rtnetlink.h>
> +#include <net/sock.h>
> +#include <linux/virtio_net.h>
> +
> +#define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN | NETIF_F_TSO | \
> + NETIF_F_TSO6 | NETIF_F_UFO)
> +
> +static dev_t ipvtap_major;
> +static struct cdev ipvtap_cdev;
> +
> +static const void *ipvtap_net_namespace(struct device *d)
> +{
> + struct net_device *dev = to_net_dev(d->parent);
> + return dev_net(dev);
> +}
> +
> +static struct class ipvtap_class = {
> + .name = "ipvtap",
> + .owner = THIS_MODULE,
> + .ns_type = &net_ns_type_operations,
> + .namespace = ipvtap_net_namespace,
> +};
> +
> +struct ipvtap_dev {
> + struct ipvl_dev vlan;
> + struct tap_dev tap;
> +};
> +
> +static void ipvtap_count_tx_dropped(struct tap_dev *tap)
> +{
> + struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct ipvtap_dev, tap);
> +
> + this_cpu_inc(vlan->pcpu_stats->tx_drps);
> +}
> +
> +static void ipvtap_count_rx_dropped(struct tap_dev *tap)
> +{
> + struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct ipvtap_dev, tap);
> +
> + ipvlan_count_rx(vlan, 0, 0, 0);
> +}
> +
> +static void ipvtap_update_features(struct tap_dev *tap,
> + netdev_features_t features)
> +{
> + struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct ipvtap_dev, tap);
> +
> + vlan->sfeatures = features;
> + netdev_update_features(vlan->dev);
> +}
> +
> +static int ipvtap_newlink(struct net *src_net,
> + struct net_device *dev,
> + struct nlattr *tb[],
> + struct nlattr *data[])
> +{
> + struct ipvtap_dev *vlantap = netdev_priv(dev);
> + int err;
> +
> + INIT_LIST_HEAD(&vlantap->tap.queue_list);
> +
> + /* Since macvlan supports all offloads by default, make
> + * tap support all offloads also.
> + */
> + vlantap->tap.tap_features = TUN_OFFLOADS;
> + vlantap->tap.count_tx_dropped = ipvtap_count_tx_dropped;
> + vlantap->tap.update_features = ipvtap_update_features;
> + vlantap->tap.count_rx_dropped = ipvtap_count_rx_dropped;
> +
> + err = netdev_rx_handler_register(dev, tap_handle_frame, &vlantap->tap);
> + if (err)
> + return err;
> +
> + /* Don't put anything that may fail after macvlan_common_newlink
> + * because we can't undo what it does.
> + */
> + err = ipvlan_link_new(src_net, dev, tb, data);
> + if (err) {
> + netdev_rx_handler_unregister(dev);
> + return err;
> + }
> +
> + vlantap->tap.dev = vlantap->vlan.dev;
> +
> + return err;
> +}
> +
> +static void ipvtap_dellink(struct net_device *dev,
> + struct list_head *head)
> +{
> + struct ipvtap_dev *vlan = netdev_priv(dev);
> +
> + netdev_rx_handler_unregister(dev);
> + tap_del_queues(&vlan->tap);
> + ipvlan_link_delete(dev, head);
> +}
> +
> +static void ipvtap_setup(struct net_device *dev)
> +{
> + ipvlan_link_setup(dev);
> + dev->tx_queue_len = TUN_READQ_SIZE;
> + dev->priv_flags &= ~IFF_NO_QUEUE;
> +}
> +
> +static struct rtnl_link_ops ipvtap_link_ops __read_mostly = {
> + .kind = "ipvtap",
> + .setup = ipvtap_setup,
> + .newlink = ipvtap_newlink,
> + .dellink = ipvtap_dellink,
> + .priv_size = sizeof(struct ipvtap_dev),
> +};
> +
> +static int ipvtap_device_event(struct notifier_block *unused,
> + unsigned long event, void *ptr)
> +{
> + struct net_device *dev = netdev_notifier_info_to_dev(ptr);
> + struct ipvtap_dev *vlantap;
> + struct device *classdev;
> + dev_t devt;
> + int err;
> + char tap_name[IFNAMSIZ];
> +
> + if (dev->rtnl_link_ops != &ipvtap_link_ops)
> + return NOTIFY_DONE;
> +
> + snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
> + vlantap = netdev_priv(dev);
> +
> + switch (event) {
> + case NETDEV_REGISTER:
> + /* Create the device node here after the network device has
> + * been registered but before register_netdevice has
> + * finished running.
> + */
> + err = tap_get_minor(ipvtap_major, &vlantap->tap);
> + if (err)
> + return notifier_from_errno(err);
> +
> + devt = MKDEV(MAJOR(ipvtap_major), vlantap->tap.minor);
> + classdev = device_create(&ipvtap_class, &dev->dev, devt,
> + dev, tap_name);
> + if (IS_ERR(classdev)) {
> + tap_free_minor(ipvtap_major, &vlantap->tap);
> + return notifier_from_errno(PTR_ERR(classdev));
> + }
> + err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
> + tap_name);
> + if (err)
> + return notifier_from_errno(err);
> + break;
> + case NETDEV_UNREGISTER:
> + /* vlan->minor == 0 if NETDEV_REGISTER above failed */
> + if (vlantap->tap.minor == 0)
> + break;
> + sysfs_remove_link(&dev->dev.kobj, tap_name);
> + devt = MKDEV(MAJOR(ipvtap_major), vlantap->tap.minor);
> + device_destroy(&ipvtap_class, devt);
> + tap_free_minor(ipvtap_major, &vlantap->tap);
> + break;
> + case NETDEV_CHANGE_TX_QUEUE_LEN:
> + if (tap_queue_resize(&vlantap->tap))
> + return NOTIFY_BAD;
> + break;
> + }
> +
> + return NOTIFY_DONE;
> +}
> +
> +static struct notifier_block ipvtap_notifier_block __read_mostly = {
> + .notifier_call = ipvtap_device_event,
> +};
> +
> +static int ipvtap_init(void)
> +{
> + int err;
> +
> + err = tap_create_cdev(&ipvtap_cdev, &ipvtap_major, "ipvtap");
> +
> + if (err)
> + goto out1;
> +
> + err = class_register(&ipvtap_class);
> + if (err)
> + goto out2;
> +
> + err = register_netdevice_notifier(&ipvtap_notifier_block);
> + if (err)
> + goto out3;
> +
> + err = ipvlan_link_register(&ipvtap_link_ops);
> + if (err)
> + goto out4;
> +
> + return 0;
> +
> +out4:
> + unregister_netdevice_notifier(&ipvtap_notifier_block);
> +out3:
> + class_unregister(&ipvtap_class);
> +out2:
> + cdev_del(&ipvtap_cdev);
> +out1:
> + return err;
> +}
> +module_init(ipvtap_init);
> +
> +static void ipvtap_exit(void)
> +{
> + rtnl_link_unregister(&ipvtap_link_ops);
> + unregister_netdevice_notifier(&ipvtap_notifier_block);
> + class_unregister(&ipvtap_class);
> + tap_destroy_cdev(ipvtap_major, &ipvtap_cdev);
> +}
> +module_exit(ipvtap_exit);
> +MODULE_ALIAS_RTNL_LINK("ipvtap");
> +MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
> +MODULE_LICENSE("GPL");
> --
> 2.7.4
>
^ permalink raw reply [flat|nested] 21+ messages in thread* RE: [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver
2017-01-06 23:46 ` Mahesh Bandewar (महेश बंडेवार)
@ 2017-01-17 23:49 ` Grandhi, Sainath
0 siblings, 0 replies; 21+ messages in thread
From: Grandhi, Sainath @ 2017-01-17 23:49 UTC (permalink / raw)
To: Mahesh Bandewar (महेश बंडेवार)
Cc: linux-netdev, David Miller, mahesh@bandewar.net,
linux-kernel@vger.kernel.org
> -----Original Message-----
> From: Mahesh Bandewar (महेश बंडेवार)
> [mailto:maheshb@google.com]
> Sent: Friday, January 06, 2017 3:47 PM
> To: Grandhi, Sainath <sainath.grandhi@intel.com>
> Cc: linux-netdev <netdev@vger.kernel.org>; David Miller
> <davem@davemloft.net>; mahesh@bandewar.net; linux-
> kernel@vger.kernel.org
> Subject: Re: [PATCHv1 7/7] IPVTAP: IP-VLAN based tap driver
>
> few superficial comments inline.
>
> On Fri, Jan 6, 2017 at 2:33 PM, Sainath Grandhi <sainath.grandhi@intel.com>
> wrote:
> > This patch adds a tap character device driver that is based on the
> > IP-VLAN network interface, called ipvtap. An ipvtap device can be
> > created in the same way as an ipvlan device, using 'type ipvtap', and
> > then accessed using the tap user space interface.
> >
> > Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
> > Tested-by: Sainath Grandhi <sainath.grandhi@intel.com>
> > ---
> > drivers/net/Kconfig | 12 ++
> > drivers/net/Makefile | 1 +
> > drivers/net/ipvlan/Makefile | 1 +
> > drivers/net/ipvlan/ipvlan.h | 7 ++
> > drivers/net/ipvlan/ipvlan_core.c | 5 +-
> > drivers/net/ipvlan/ipvlan_main.c | 37 +++---
> > drivers/net/ipvlan/ipvtap.c | 238
> +++++++++++++++++++++++++++++++++++++++
> > 7 files changed, 282 insertions(+), 19 deletions(-) create mode
> > 100644 drivers/net/ipvlan/ipvtap.c
> >
> > diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig index
> > 280380d..ddfb30a 100644
> > --- a/drivers/net/Kconfig
> > +++ b/drivers/net/Kconfig
> > @@ -165,6 +165,18 @@ config IPVLAN
> > To compile this driver as a module, choose M here: the module
> > will be called ipvlan.
> >
> > +config IPVTAP
> > + tristate "IP-VLAN based tap driver"
> > + depends on IPVLAN
> > + depends on INET
> > + help
> > + This adds a specialized tap character device driver that is based
> > + on the IP-VLAN network interface, called ipvtap. An ipvtap device
> > + can be added in the same way as a ipvlan device, using 'type
> > + ipvtap', and then be accessed through the tap user space interface.
> > +
> > + To compile this driver as a module, choose M here: the module
> > + will be called macvtap.
> >
> > config VXLAN
> > tristate "Virtual eXtensible Local Area Network (VXLAN)"
> > diff --git a/drivers/net/Makefile b/drivers/net/Makefile index
> > 7dd86ca..98ed4d9 100644
> > --- a/drivers/net/Makefile
> > +++ b/drivers/net/Makefile
> > @@ -7,6 +7,7 @@
> > #
> > obj-$(CONFIG_BONDING) += bonding/
> > obj-$(CONFIG_IPVLAN) += ipvlan/
> > +obj-$(CONFIG_IPVTAP) += ipvlan/
> > obj-$(CONFIG_DUMMY) += dummy.o
> > obj-$(CONFIG_EQUALIZER) += eql.o
> > obj-$(CONFIG_IFB) += ifb.o
> > diff --git a/drivers/net/ipvlan/Makefile b/drivers/net/ipvlan/Makefile
> > index df79910..8a2c64d 100644
> > --- a/drivers/net/ipvlan/Makefile
> > +++ b/drivers/net/ipvlan/Makefile
> > @@ -3,5 +3,6 @@
> > #
> >
> > obj-$(CONFIG_IPVLAN) += ipvlan.o
> > +obj-$(CONFIG_IPVTAP) += ipvtap.o
> >
> > ipvlan-objs := ipvlan_core.o ipvlan_main.o diff --git
> > a/drivers/net/ipvlan/ipvlan.h b/drivers/net/ipvlan/ipvlan.h index
> > dbfbb33..4362d88 100644
> > --- a/drivers/net/ipvlan/ipvlan.h
> > +++ b/drivers/net/ipvlan/ipvlan.h
> > @@ -133,4 +133,11 @@ struct sk_buff *ipvlan_l3_rcv(struct net_device
> *dev, struct sk_buff *skb,
> > u16 proto); unsigned int
> > ipvlan_nf_input(void *priv, struct sk_buff *skb,
> > const struct nf_hook_state *state);
> > +void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
> > + unsigned int len, bool success, bool mcast); int
> > +ipvlan_link_new(struct net *src_net, struct net_device *dev,
> > + struct nlattr *tb[], struct nlattr *data[]); void
> > +ipvlan_link_delete(struct net_device *dev, struct list_head *head);
> > +void ipvlan_link_setup(struct net_device *dev); int
> > +ipvlan_link_register(struct rtnl_link_ops *ops);
> > #endif /* __IPVLAN_H */
> > diff --git a/drivers/net/ipvlan/ipvlan_core.c
> > b/drivers/net/ipvlan/ipvlan_core.c
> > index 83ce74a..9af16ab 100644
> > --- a/drivers/net/ipvlan/ipvlan_core.c
> > +++ b/drivers/net/ipvlan/ipvlan_core.c
> > @@ -16,8 +16,8 @@ void ipvlan_init_secret(void)
> > net_get_random_once(&ipvlan_jhash_secret,
> > sizeof(ipvlan_jhash_secret)); }
> >
> > -static void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
> > - unsigned int len, bool success, bool mcast)
> > +void ipvlan_count_rx(const struct ipvl_dev *ipvlan,
> > + unsigned int len, bool success, bool mcast)
> > {
> > if (!ipvlan)
> > return;
> > @@ -36,6 +36,7 @@ static void ipvlan_count_rx(const struct ipvl_dev
> *ipvlan,
> > this_cpu_inc(ipvlan->pcpu_stats->rx_errs);
> > }
> > }
> > +EXPORT_SYMBOL_GPL(ipvlan_count_rx);
> Why export, isn't just removing 'static' enough?
This function becomes part of "ipvlan" module.
"ipvtap" module depends on this function exported by "ipvlan" module.
> >
> > static u8 ipvlan_get_v6_hash(const void *iaddr) { diff --git
> > a/drivers/net/ipvlan/ipvlan_main.c b/drivers/net/ipvlan/ipvlan_main.c
> > index 8b0f993..666a05d 100644
> > --- a/drivers/net/ipvlan/ipvlan_main.c
> > +++ b/drivers/net/ipvlan/ipvlan_main.c
> > @@ -13,14 +13,14 @@ static u32 ipvl_nf_hook_refcnt = 0;
> >
> > static struct nf_hook_ops ipvl_nfops[] __read_mostly = {
> > {
> > - .hook = ipvlan_nf_input,
> > - .pf = NFPROTO_IPV4,
> > + .hook = ipvlan_nf_input,
> > + .pf = NFPROTO_IPV4,
> spurious?
> > .hooknum = NF_INET_LOCAL_IN,
> > .priority = INT_MAX,
> > },
> > {
> > - .hook = ipvlan_nf_input,
> > - .pf = NFPROTO_IPV6,
> > + .hook = ipvlan_nf_input,
> > + .pf = NFPROTO_IPV6,
> spurious??
> > .hooknum = NF_INET_LOCAL_IN,
> > .priority = INT_MAX,
> > },
> > @@ -398,7 +398,7 @@ static int ipvlan_hard_header(struct sk_buff *skb,
> > struct net_device *dev, }
> >
> > static const struct header_ops ipvlan_header_ops = {
> > - .create = ipvlan_hard_header,
> > + .create = ipvlan_hard_header,
> spurious???
Would take care of it in the next version.
> > .parse = eth_header_parse,
> > .cache = eth_header_cache,
> > .cache_update = eth_header_cache_update,
> > @@ -494,8 +494,8 @@ static int ipvlan_nl_fillinfo(struct sk_buff *skb,
> > return ret;
> > }
> >
> > -static int ipvlan_link_new(struct net *src_net, struct net_device *dev,
> > - struct nlattr *tb[], struct nlattr *data[])
> > +int ipvlan_link_new(struct net *src_net, struct net_device *dev,
> > + struct nlattr *tb[], struct nlattr *data[])
> > {
> > struct ipvl_dev *ipvlan = netdev_priv(dev);
> > struct ipvl_port *port;
> > @@ -567,8 +567,9 @@ static int ipvlan_link_new(struct net *src_net,
> struct net_device *dev,
> > ipvlan_port_destroy(phy_dev);
> > return err;
> > }
> > +EXPORT_SYMBOL_GPL(ipvlan_link_new);
> >
> > -static void ipvlan_link_delete(struct net_device *dev, struct
> > list_head *head)
> > +void ipvlan_link_delete(struct net_device *dev, struct list_head
> > +*head)
> > {
> > struct ipvl_dev *ipvlan = netdev_priv(dev);
> > struct ipvl_addr *addr, *next; @@ -583,8 +584,9 @@ static void
> > ipvlan_link_delete(struct net_device *dev, struct list_head *head)
> > unregister_netdevice_queue(dev, head);
> > netdev_upper_dev_unlink(ipvlan->phy_dev, dev); }
> > +EXPORT_SYMBOL_GPL(ipvlan_link_delete);
> >
> > -static void ipvlan_link_setup(struct net_device *dev)
> > +void ipvlan_link_setup(struct net_device *dev)
> > {
> > ether_setup(dev);
> >
> > @@ -595,6 +597,7 @@ static void ipvlan_link_setup(struct net_device
> *dev)
> > dev->header_ops = &ipvlan_header_ops;
> > dev->ethtool_ops = &ipvlan_ethtool_ops; }
> > +EXPORT_SYMBOL_GPL(ipvlan_link_setup);
> >
> > static const struct nla_policy ipvlan_nl_policy[IFLA_IPVLAN_MAX + 1] =
> > {
> > @@ -605,22 +608,22 @@ static struct rtnl_link_ops ipvlan_link_ops = {
> > .kind = "ipvlan",
> > .priv_size = sizeof(struct ipvl_dev),
> >
> > - .get_size = ipvlan_nl_getsize,
> > - .policy = ipvlan_nl_policy,
> > - .validate = ipvlan_nl_validate,
> > - .fill_info = ipvlan_nl_fillinfo,
> > - .changelink = ipvlan_nl_changelink,
> > - .maxtype = IFLA_IPVLAN_MAX,
> > -
> > .setup = ipvlan_link_setup,
> > .newlink = ipvlan_link_new,
> > .dellink = ipvlan_link_delete,
> > };
> >
> > -static int ipvlan_link_register(struct rtnl_link_ops *ops)
> > +int ipvlan_link_register(struct rtnl_link_ops *ops)
> > {
> > + ops->get_size = ipvlan_nl_getsize;
> > + ops->policy = ipvlan_nl_policy;
> > + ops->validate = ipvlan_nl_validate;
> > + ops->fill_info = ipvlan_nl_fillinfo;
> > + ops->changelink = ipvlan_nl_changelink;
> > + ops->maxtype = IFLA_IPVLAN_MAX;
> > return rtnl_link_register(ops);
> > }
> > +EXPORT_SYMBOL_GPL(ipvlan_link_register);
> >
> > static int ipvlan_device_event(struct notifier_block *unused,
> > unsigned long event, void *ptr)
> > diff --git a/drivers/net/ipvlan/ipvtap.c b/drivers/net/ipvlan/ipvtap.c
> > new file mode 100644
> > index 0000000..0422cdf
> > --- /dev/null
> > +++ b/drivers/net/ipvlan/ipvtap.c
> > @@ -0,0 +1,238 @@
> > +#include <linux/etherdevice.h>
> > +#include "ipvlan.h"
> > +#include <linux/if_vlan.h>
> > +#include <linux/if_tap.h>
> > +#include <linux/interrupt.h>
> > +#include <linux/nsproxy.h>
> > +#include <linux/compat.h>
> > +#include <linux/if_tun.h>
> > +#include <linux/module.h>
> > +#include <linux/skbuff.h>
> > +#include <linux/cache.h>
> > +#include <linux/sched.h>
> > +#include <linux/types.h>
> > +#include <linux/slab.h>
> > +#include <linux/wait.h>
> > +#include <linux/cdev.h>
> > +#include <linux/idr.h>
> > +#include <linux/fs.h>
> > +#include <linux/uio.h>
> > +
> > +#include <net/net_namespace.h>
> > +#include <net/rtnetlink.h>
> > +#include <net/sock.h>
> > +#include <linux/virtio_net.h>
> > +
> > +#define TUN_OFFLOADS (NETIF_F_HW_CSUM | NETIF_F_TSO_ECN |
> NETIF_F_TSO | \
> > + NETIF_F_TSO6 | NETIF_F_UFO)
> > +
> > +static dev_t ipvtap_major;
> > +static struct cdev ipvtap_cdev;
> > +
> > +static const void *ipvtap_net_namespace(struct device *d)
> > +{
> > + struct net_device *dev = to_net_dev(d->parent);
> > + return dev_net(dev);
> > +}
> > +
> > +static struct class ipvtap_class = {
> > + .name = "ipvtap",
> > + .owner = THIS_MODULE,
> > + .ns_type = &net_ns_type_operations,
> > + .namespace = ipvtap_net_namespace,
> > +};
> > +
> > +struct ipvtap_dev {
> > + struct ipvl_dev vlan;
> > + struct tap_dev tap;
> > +};
> > +
> > +static void ipvtap_count_tx_dropped(struct tap_dev *tap)
> > +{
> > + struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct
> ipvtap_dev, tap);
> > +
> > + this_cpu_inc(vlan->pcpu_stats->tx_drps);
> > +}
> > +
> > +static void ipvtap_count_rx_dropped(struct tap_dev *tap)
> > +{
> > + struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct
> ipvtap_dev, tap);
> > +
> > + ipvlan_count_rx(vlan, 0, 0, 0);
> > +}
> > +
> > +static void ipvtap_update_features(struct tap_dev *tap,
> > + netdev_features_t features)
> > +{
> > + struct ipvl_dev *vlan = (struct ipvl_dev *)container_of(tap, struct
> ipvtap_dev, tap);
> > +
> > + vlan->sfeatures = features;
> > + netdev_update_features(vlan->dev);
> > +}
> > +
> > +static int ipvtap_newlink(struct net *src_net,
> > + struct net_device *dev,
> > + struct nlattr *tb[],
> > + struct nlattr *data[])
> > +{
> > + struct ipvtap_dev *vlantap = netdev_priv(dev);
> > + int err;
> > +
> > + INIT_LIST_HEAD(&vlantap->tap.queue_list);
> > +
> > + /* Since macvlan supports all offloads by default, make
> > + * tap support all offloads also.
> > + */
> > + vlantap->tap.tap_features = TUN_OFFLOADS;
> > + vlantap->tap.count_tx_dropped = ipvtap_count_tx_dropped;
> > + vlantap->tap.update_features = ipvtap_update_features;
> > + vlantap->tap.count_rx_dropped = ipvtap_count_rx_dropped;
> > +
> > + err = netdev_rx_handler_register(dev, tap_handle_frame, &vlantap-
> >tap);
> > + if (err)
> > + return err;
> > +
> > + /* Don't put anything that may fail after macvlan_common_newlink
> > + * because we can't undo what it does.
> > + */
> > + err = ipvlan_link_new(src_net, dev, tb, data);
> > + if (err) {
> > + netdev_rx_handler_unregister(dev);
> > + return err;
> > + }
> > +
> > + vlantap->tap.dev = vlantap->vlan.dev;
> > +
> > + return err;
> > +}
> > +
> > +static void ipvtap_dellink(struct net_device *dev,
> > + struct list_head *head)
> > +{
> > + struct ipvtap_dev *vlan = netdev_priv(dev);
> > +
> > + netdev_rx_handler_unregister(dev);
> > + tap_del_queues(&vlan->tap);
> > + ipvlan_link_delete(dev, head);
> > +}
> > +
> > +static void ipvtap_setup(struct net_device *dev)
> > +{
> > + ipvlan_link_setup(dev);
> > + dev->tx_queue_len = TUN_READQ_SIZE;
> > + dev->priv_flags &= ~IFF_NO_QUEUE;
> > +}
> > +
> > +static struct rtnl_link_ops ipvtap_link_ops __read_mostly = {
> > + .kind = "ipvtap",
> > + .setup = ipvtap_setup,
> > + .newlink = ipvtap_newlink,
> > + .dellink = ipvtap_dellink,
> > + .priv_size = sizeof(struct ipvtap_dev),
> > +};
> > +
> > +static int ipvtap_device_event(struct notifier_block *unused,
> > + unsigned long event, void *ptr)
> > +{
> > + struct net_device *dev = netdev_notifier_info_to_dev(ptr);
> > + struct ipvtap_dev *vlantap;
> > + struct device *classdev;
> > + dev_t devt;
> > + int err;
> > + char tap_name[IFNAMSIZ];
> > +
> > + if (dev->rtnl_link_ops != &ipvtap_link_ops)
> > + return NOTIFY_DONE;
> > +
> > + snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
> > + vlantap = netdev_priv(dev);
> > +
> > + switch (event) {
> > + case NETDEV_REGISTER:
> > + /* Create the device node here after the network device has
> > + * been registered but before register_netdevice has
> > + * finished running.
> > + */
> > + err = tap_get_minor(ipvtap_major, &vlantap->tap);
> > + if (err)
> > + return notifier_from_errno(err);
> > +
> > + devt = MKDEV(MAJOR(ipvtap_major), vlantap->tap.minor);
> > + classdev = device_create(&ipvtap_class, &dev->dev, devt,
> > + dev, tap_name);
> > + if (IS_ERR(classdev)) {
> > + tap_free_minor(ipvtap_major, &vlantap->tap);
> > + return notifier_from_errno(PTR_ERR(classdev));
> > + }
> > + err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
> > + tap_name);
> > + if (err)
> > + return notifier_from_errno(err);
> > + break;
> > + case NETDEV_UNREGISTER:
> > + /* vlan->minor == 0 if NETDEV_REGISTER above failed */
> > + if (vlantap->tap.minor == 0)
> > + break;
> > + sysfs_remove_link(&dev->dev.kobj, tap_name);
> > + devt = MKDEV(MAJOR(ipvtap_major), vlantap->tap.minor);
> > + device_destroy(&ipvtap_class, devt);
> > + tap_free_minor(ipvtap_major, &vlantap->tap);
> > + break;
> > + case NETDEV_CHANGE_TX_QUEUE_LEN:
> > + if (tap_queue_resize(&vlantap->tap))
> > + return NOTIFY_BAD;
> > + break;
> > + }
> > +
> > + return NOTIFY_DONE;
> > +}
> > +
> > +static struct notifier_block ipvtap_notifier_block __read_mostly = {
> > + .notifier_call = ipvtap_device_event,
> > +};
> > +
> > +static int ipvtap_init(void)
> > +{
> > + int err;
> > +
> > + err = tap_create_cdev(&ipvtap_cdev, &ipvtap_major, "ipvtap");
> > +
> > + if (err)
> > + goto out1;
> > +
> > + err = class_register(&ipvtap_class);
> > + if (err)
> > + goto out2;
> > +
> > + err = register_netdevice_notifier(&ipvtap_notifier_block);
> > + if (err)
> > + goto out3;
> > +
> > + err = ipvlan_link_register(&ipvtap_link_ops);
> > + if (err)
> > + goto out4;
> > +
> > + return 0;
> > +
> > +out4:
> > + unregister_netdevice_notifier(&ipvtap_notifier_block);
> > +out3:
> > + class_unregister(&ipvtap_class);
> > +out2:
> > + cdev_del(&ipvtap_cdev);
> > +out1:
> > + return err;
> > +}
> > +module_init(ipvtap_init);
> > +
> > +static void ipvtap_exit(void)
> > +{
> > + rtnl_link_unregister(&ipvtap_link_ops);
> > + unregister_netdevice_notifier(&ipvtap_notifier_block);
> > + class_unregister(&ipvtap_class);
> > + tap_destroy_cdev(ipvtap_major, &ipvtap_cdev);
> > +}
> > +module_exit(ipvtap_exit);
> > +MODULE_ALIAS_RTNL_LINK("ipvtap");
> > +MODULE_AUTHOR("Arnd Bergmann <arnd@arndb.de>");
> > +MODULE_LICENSE("GPL");
> > --
> > 2.7.4
> >
^ permalink raw reply [flat|nested] 21+ messages in thread