dpdk中kni模块

一，什么是kni，为什么要有kni

Kni(Kernel NIC Interface)内核网卡接口，是DPDK允许用户态和内核态交换报文的解决方案，模拟了一个虚拟的网口，提供dpdk的应用程序和linux内核之间通讯。kni接口允许报文从用户态接收后转发到linu协议栈去。

为什么要弄一个kni接口，虽然dpdk的高速转发性能很出色，但是也有自己的一些缺点，比如没有协议栈就是其中一项缺陷，当然也可能当时设计时就将没有将协议栈考虑进去，毕竟协议栈需要将报文转发处理，可能会使

处理报文的能力大大降低。

直接上图：

上图是kni的mbuf使用流程图，也可以看出报文的流向，因为报文在代码中其实就是一个个内存指针。其中rx_q右边是用户态，左边是内核态。最后通过调用netif_rx()将报文送入linux协议栈，这其中需要将dpdk的mbuf转换成skb_buf。

当linux向kni端口发送报文时，调用回调函数kni_net_tx()，然后报文经过转换之后发送到端口上。

二：主要代码分析：

1，和igb uio模块一样，kni模块分成内核以及用户态代码，内核模块在编译出来之后为rte_kni.ko，首先插入内核，dpdk提供了一个用户态的例子。首先看下kni内核模块代码：

在kni_misc.c中，ko代码入口为

module_init(kni_init);

可以看到函数从kni_init进入：

 static int __init
 kni_init(void)
 {
     int rc;

     KNI_PRINT("######## DPDK kni module loading ########\n");

     if (kni_parse_kthread_mode() < ) {    //kni的线程模式、单线程还是多线程
         KNI_ERR("Invalid parameter for kthread_mode\n");
         return -EINVAL;
     }

 #ifdef HAVE_SIMPLIFIED_PERNET_OPERATIONS
     rc = register_pernet_subsys(&kni_net_ops);
 #else
     rc = register_pernet_gen_subsys(&kni_net_id, &kni_net_ops);
 #endif
     if (rc)
         return -EPERM;

     rc = misc_register(&kni_misc);
     if (rc != ) {
         KNI_ERR("Misc registration failed\n");
         goto out;
     }

     /* Configure the lo mode according to the input parameter */
     kni_net_config_lo_mode(lo_mode);

     KNI_PRINT("######## DPDK kni module loaded  ########\n");

     return ;

 out:
 #ifdef HAVE_SIMPLIFIED_PERNET_OPERATIONS
     unregister_pernet_subsys(&kni_net_ops);
 #else
     register_pernet_gen_subsys(&kni_net_id, &kni_net_ops);
 #endif
     return rc;
 }

代码比较简单，首先选择kni的线程模式，分为单线程还是多线程，所谓单线程是指所有的kni端口收发都由一个线程守护，多线程只是每一个kni端口分为由一个线程守护，这部分是在插入模块时带入参数选择。

接着调用注册函数misc_register，将kni注册为一个混杂设备。其中kni_misc结构体里面定义了该混杂设备的一些操作

 static struct miscdevice kni_misc = {
     .minor = MISC_DYNAMIC_MINOR,
     .name = KNI_DEVICE,
     .fops = &kni_fops,
 };

这里主要看.fops里面的结构体

 static struct file_operations kni_fops = {
     .owner = THIS_MODULE,
     .open = kni_open,
     .release = kni_release,
     .unlocked_ioctl = (void *)kni_ioctl,
     .compat_ioctl = (void *)kni_compat_ioctl,
 };

这里涉及的主要操作有kni_open，kni_release，以及kni_ioctl，分别对应几个函数

 static int
 kni_open(struct inode *inode, struct file *file)
 {
     struct net *net = current->nsproxy->net_ns;
     struct kni_net *knet = net_generic(net, kni_net_id);

     /* kni device can be opened by one user only per netns */
     if (test_and_set_bit(KNI_DEV_IN_USE_BIT_NUM, &knet->device_in_use))
         return -EBUSY;

     /* Create kernel thread for single mode */
     if (multiple_kthread_on == ) {
         KNI_PRINT("Single kernel thread for all KNI devices\n");
         /* Create kernel thread for RX */
         knet->kni_kthread = kthread_run(kni_thread_single, (void *)knet,
                         "kni_single");
         if (IS_ERR(knet->kni_kthread)) {
             KNI_ERR("Unable to create kernel threaed\n");
             return PTR_ERR(knet->kni_kthread);
         }
     } else
         KNI_PRINT("Multiple kernel thread mode enabled\n");

     file->private_data = get_net(net);
     KNI_PRINT("/dev/kni opened\n");

     return ;
 }

kni_open时如果是单线程模式则会创建一个内核线程，并打开dev/kni，这个时候在host的dev下能看到kni文件夹

 static int
 kni_ioctl(struct inode *inode,
     unsigned int ioctl_num,
     unsigned long ioctl_param)
 {
     int ret = -EINVAL;
     struct net *net = current->nsproxy->net_ns;

     KNI_DBG("IOCTL num=0x%0x param=0x%0lx\n", ioctl_num, ioctl_param);

     /*
      * Switch according to the ioctl called
      */
     switch (_IOC_NR(ioctl_num)) {
     case _IOC_NR(RTE_KNI_IOCTL_TEST):
         /* For test only, not used */
         break;
     case _IOC_NR(RTE_KNI_IOCTL_CREATE):
         ret = kni_ioctl_create(net, ioctl_num, ioctl_param);
         break;
     case _IOC_NR(RTE_KNI_IOCTL_RELEASE):
         ret = kni_ioctl_release(net, ioctl_num, ioctl_param);
         break;
     default:
         KNI_DBG("IOCTL default\n");
         break;
     }

     return ret;
 }

kni_ioctl函数是与用户态通信的一个接口，主要是的是kni_ioctl_create函数：

 static int
 kni_ioctl_create(struct net *net,
         unsigned int ioctl_num, unsigned long ioctl_param)
 {
     struct kni_net *knet = net_generic(net, kni_net_id);
     int ret;
     struct rte_kni_device_info dev_info;
     struct pci_dev *pci = NULL;
     struct pci_dev *found_pci = NULL;
     struct net_device *net_dev = NULL;
     struct net_device *lad_dev = NULL;
     struct kni_dev *kni, *dev, *n;

     printk(KERN_INFO "KNI: Creating kni...\n");
     /* Check the buffer size, to avoid warning */
     if (_IOC_SIZE(ioctl_num) > sizeof(dev_info))
         return -EINVAL;

     /* Copy kni info from user space */
     ret = copy_from_user(&dev_info, (void *)ioctl_param, sizeof(dev_info));
     if (ret) {
         KNI_ERR("copy_from_user in kni_ioctl_create");
         return -EIO;
     }

     /**
      * Check if the cpu core id is valid for binding,
      * for multiple kernel thread mode.
      */
     if (multiple_kthread_on && dev_info.force_bind &&
                 !cpu_online(dev_info.core_id)) {
         KNI_ERR("cpu %u is not online\n", dev_info.core_id);
         return -EINVAL;
     }

     /* Check if it has been created */
     down_read(&knet->kni_list_lock);
     list_for_each_entry_safe(dev, n, &knet->kni_list_head, list) {
         if (kni_check_param(dev, &dev_info) < ) {
             up_read(&knet->kni_list_lock);
             return -EINVAL;
         }
     }
     up_read(&knet->kni_list_lock);

     net_dev = alloc_netdev(sizeof(struct kni_dev), dev_info.name,
 #ifdef NET_NAME_UNKNOWN
                             NET_NAME_UNKNOWN,
 #endif
                             kni_net_init);
     if (net_dev == NULL) {
         KNI_ERR("error allocating device \"%s\"\n", dev_info.name);
         return -EBUSY;
     }

     dev_net_set(net_dev, net);

     kni = netdev_priv(net_dev);

     kni->net_dev = net_dev;
     kni->group_id = dev_info.group_id;
     kni->core_id = dev_info.core_id;
     strncpy(kni->name, dev_info.name, RTE_KNI_NAMESIZE);

     /* Translate user space info into kernel space info */
     kni->tx_q = phys_to_virt(dev_info.tx_phys);
     kni->rx_q = phys_to_virt(dev_info.rx_phys);
     kni->alloc_q = phys_to_virt(dev_info.alloc_phys);
     kni->free_q = phys_to_virt(dev_info.free_phys);

     kni->req_q = phys_to_virt(dev_info.req_phys);
     kni->resp_q = phys_to_virt(dev_info.resp_phys);
     kni->sync_va = dev_info.sync_va;
     kni->sync_kva = phys_to_virt(dev_info.sync_phys);

     kni->mbuf_kva = phys_to_virt(dev_info.mbuf_phys);
     kni->mbuf_va = dev_info.mbuf_va;

 #ifdef RTE_KNI_VHOST
     kni->vhost_queue = NULL;
     kni->vq_status = BE_STOP;
 #endif
     kni->mbuf_size = dev_info.mbuf_size;

     KNI_PRINT("tx_phys:      0x%016llx, tx_q addr:      0x%p\n",
         (unsigned long long) dev_info.tx_phys, kni->tx_q);
     KNI_PRINT("rx_phys:      0x%016llx, rx_q addr:      0x%p\n",
         (unsigned long long) dev_info.rx_phys, kni->rx_q);
     KNI_PRINT("alloc_phys:   0x%016llx, alloc_q addr:   0x%p\n",
         (unsigned long long) dev_info.alloc_phys, kni->alloc_q);
     KNI_PRINT("free_phys:    0x%016llx, free_q addr:    0x%p\n",
         (unsigned long long) dev_info.free_phys, kni->free_q);
     KNI_PRINT("req_phys:     0x%016llx, req_q addr:     0x%p\n",
         (unsigned long long) dev_info.req_phys, kni->req_q);
     KNI_PRINT("resp_phys:    0x%016llx, resp_q addr:    0x%p\n",
         (unsigned long long) dev_info.resp_phys, kni->resp_q);
     KNI_PRINT("mbuf_phys:    0x%016llx, mbuf_kva:       0x%p\n",
         (unsigned long long) dev_info.mbuf_phys, kni->mbuf_kva);
     KNI_PRINT("mbuf_va:      0x%p\n", dev_info.mbuf_va);
     KNI_PRINT("mbuf_size:    %u\n", kni->mbuf_size);

     KNI_DBG("PCI: %02x:%02x.%02x %04x:%04x\n",
                     dev_info.bus,
                     dev_info.devid,
                     dev_info.function,
                     dev_info.vendor_id,
                     dev_info.device_id);

     pci = pci_get_device(dev_info.vendor_id, dev_info.device_id, NULL);

     /* Support Ethtool */
     while (pci) {
         KNI_PRINT("pci_bus: %02x:%02x:%02x \n",
                     pci->bus->number,
                     PCI_SLOT(pci->devfn),
                     PCI_FUNC(pci->devfn));

         if ((pci->bus->number == dev_info.bus) &&
             (PCI_SLOT(pci->devfn) == dev_info.devid) &&
             (PCI_FUNC(pci->devfn) == dev_info.function)) {
             found_pci = pci;
             switch (dev_info.device_id) {
             #define RTE_PCI_DEV_ID_DECL_IGB(vend, dev) case (dev):
             #include <rte_pci_dev_ids.h>
                 ret = igb_kni_probe(found_pci, &lad_dev);
                 break;
             #define RTE_PCI_DEV_ID_DECL_IXGBE(vend, dev) \
                             case (dev):
             #include <rte_pci_dev_ids.h>
                 ret = ixgbe_kni_probe(found_pci, &lad_dev);
                 break;
             default:
                 ret = -;
                 break;
             }

             KNI_DBG("PCI found: pci=0x%p, lad_dev=0x%p\n",
                             pci, lad_dev);
             if (ret == ) {
                 kni->lad_dev = lad_dev;
                 kni_set_ethtool_ops(kni->net_dev);
             } else {
                 KNI_ERR("Device not supported by ethtool");
                 kni->lad_dev = NULL;
             }

             kni->pci_dev = found_pci;
             kni->device_id = dev_info.device_id;
             break;
         }
         pci = pci_get_device(dev_info.vendor_id,
                 dev_info.device_id, pci);
     }
     if (pci)
         pci_dev_put(pci);

     if (kni->lad_dev)
         memcpy(net_dev->dev_addr, kni->lad_dev->dev_addr, ETH_ALEN);
     else
         /*
          * Generate random mac address. eth_random_addr() is the newer
          * version of generating mac address in linux kernel.
          */
         //random_ether_addr(net_dev->dev_addr);
         memcpy(net_dev->dev_addr, &dev_info.kni_mac,ETH_ALEN);

     ret = register_netdev(net_dev);
     if (ret) {
         KNI_ERR("error %i registering device \"%s\"\n",
                     ret, dev_info.name);
         kni_dev_remove(kni);
         return -ENODEV;
     }

 #ifdef RTE_KNI_VHOST
     kni_vhost_init(kni);
 #endif

     /**
      * Create a new kernel thread for multiple mode, set its core affinity,
      * and finally wake it up.
      */
     if (multiple_kthread_on) {
         kni->pthread = kthread_create(kni_thread_multiple,
                           (void *)kni,
                           "kni_%s", kni->name);
         if (IS_ERR(kni->pthread)) {
             kni_dev_remove(kni);
             return -ECANCELED;
         }
         if (dev_info.force_bind)
             kthread_bind(kni->pthread, kni->core_id);
         wake_up_process(kni->pthread);
     }

     down_write(&knet->kni_list_lock);
     list_add(&kni->list, &knet->kni_list_head);
     up_write(&knet->kni_list_lock);

     return ;
 }

ret = copy_from_user(&dev_info, (void *)ioctl_param, sizeof(dev_info));这条语句会拷贝从用户态传过来的消息，dev_info主要存放了虚拟kni网口的相关参数，接下来就会根据dev_info中的参数注册一个kni网口ret = register_netdev(net_dev);

这个函数完成创建，这样就虚拟出一个网口出来。其中165行是自己修改的，因为我发现按照文档提供的方法根本不能ping通报文，我将生成kni的mac地址修改成dpdk接管的网口mac即可贯通。原生态代码是随时生成一个mac。

2，用户态代码主要分析dpdk提供的example，

 int
 main(int argc, char** argv)
 {
     int ret;
     uint8_t nb_sys_ports, port;
     unsigned i;

     /* Associate signal_hanlder function with USR signals */
     signal(SIGUSR1, signal_handler);
     signal(SIGUSR2, signal_handler);
     signal(SIGRTMIN, signal_handler);
     signal(SIGINT, signal_handler);

     /* Initialise EAL */
     ret = rte_eal_init(argc, argv);
     if (ret < )
         rte_exit(EXIT_FAILURE, "Could not initialise EAL (%d)\n", ret);
     argc -= ret;
     argv += ret;

     /* Parse application arguments (after the EAL ones) */
     ret = parse_args(argc, argv);
     if (ret < )
         rte_exit(EXIT_FAILURE, "Could not parse input parameters\n");

     /* Create the mbuf pool */
     pktmbuf_pool = rte_pktmbuf_pool_create("mbuf_pool", NB_MBUF,
         MEMPOOL_CACHE_SZ, , MBUF_DATA_SZ, rte_socket_id());
     if (pktmbuf_pool == NULL) {
         rte_exit(EXIT_FAILURE, "Could not initialise mbuf pool\n");
         return -;
     }

     /* Get number of ports found in scan */
     nb_sys_ports = rte_eth_dev_count();
     if (nb_sys_ports == )
         rte_exit(EXIT_FAILURE, "No supported Ethernet device found\n");

     /* Check if the configured port ID is valid */
     for (i = ; i < RTE_MAX_ETHPORTS; i++)
         if (kni_port_params_array[i] && i >= nb_sys_ports)
             rte_exit(EXIT_FAILURE, "Configured invalid "
                         "port ID %u\n", i);

     /* Initialize KNI subsystem */
     init_kni();

     /* Initialise each port */
     for (port = ; port < nb_sys_ports; port++) {
         /* Skip ports that are not enabled */
         if (!(ports_mask & ( << port)))
             continue;
         init_port(port);

         if (port >= RTE_MAX_ETHPORTS)
             rte_exit(EXIT_FAILURE, "Can not use more than "
                 "%d ports for kni\n", RTE_MAX_ETHPORTS);

         kni_alloc(port);
     }
     check_all_ports_link_status(nb_sys_ports, ports_mask);

     /* Launch per-lcore function on every lcore */
     rte_eal_mp_remote_launch(main_loop, NULL, CALL_MASTER);
     RTE_LCORE_FOREACH_SLAVE(i) {
         if (rte_eal_wait_lcore(i) < )
             return -;
     }

     /* Release resources */
     for (port = ; port < nb_sys_ports; port++) {
         if (!(ports_mask & ( << port)))
             continue;
         kni_free_kni(port);
     }
 #ifdef RTE_LIBRTE_XEN_DOM0
     rte_kni_close();
 #endif
     for (i = ; i < RTE_MAX_ETHPORTS; i++)
         if (kni_port_params_array[i]) {
             rte_free(kni_port_params_array[i]);
             kni_port_params_array[i] = NULL;
         }

     return ;
 }

main函数进来进行一些eal的初始化，随后创建一个pktmbuf_pool，重点看一下init_kni();以及kni_alloc(port);rte_eal_mp_remote_launch(main_loop, NULL, CALL_MASTER);函数。其中init_kni()函数是初始化kni子系统

 static void
 init_kni(void)
 {
     unsigned int num_of_kni_ports = , i;
     struct kni_port_params **params = kni_port_params_array;

     /* Calculate the maximum number of KNI interfaces that will be used */
     for (i = ; i < RTE_MAX_ETHPORTS; i++) {
         if (kni_port_params_array[i]) {
             num_of_kni_ports += (params[i]->nb_lcore_k ?
                 params[i]->nb_lcore_k : );
         }
     }

     /* Invoke rte KNI init to preallocate the ports */
     rte_kni_init(num_of_kni_ports);
 }

主要代码在rte_kni_init里面

 void
 rte_kni_init(unsigned int max_kni_ifaces)
 {
     uint32_t i;
     struct rte_kni_memzone_slot *it;
     const struct rte_memzone *mz;
 #define OBJNAMSIZ 32
     char obj_name[OBJNAMSIZ];
     char mz_name[RTE_MEMZONE_NAMESIZE];

     /* Immediately return if KNI is already initialized */
     if (kni_memzone_pool.initialized) {
         RTE_LOG(WARNING, KNI, "Double call to rte_kni_init()");
         return;
     }

     if (max_kni_ifaces == ) {
         RTE_LOG(ERR, KNI, "Invalid number of max_kni_ifaces %d\n",
                             max_kni_ifaces);
         rte_panic("Unable to initialize KNI\n");
     }

     /* Check FD and open */
     if (kni_fd < ) {
         kni_fd = open("/dev/" KNI_DEVICE, O_RDWR);
         if (kni_fd < )
             rte_panic("Can not open /dev/%s\n", KNI_DEVICE);
     }

     /* Allocate slot objects */
     kni_memzone_pool.slots = (struct rte_kni_memzone_slot *)
                     rte_malloc(NULL,
                     sizeof(struct rte_kni_memzone_slot) *
                     max_kni_ifaces,
                     );
     KNI_MEM_CHECK(kni_memzone_pool.slots == NULL);

     /* Initialize general pool variables */
     kni_memzone_pool.initialized = ;
     kni_memzone_pool.max_ifaces = max_kni_ifaces;
     kni_memzone_pool.free = &kni_memzone_pool.slots[];
     rte_spinlock_init(&kni_memzone_pool.mutex);

     /* Pre-allocate all memzones of all the slots; panic on error */
     for (i = ; i < max_kni_ifaces; i++) {

         /* Recover current slot */
         it = &kni_memzone_pool.slots[i];
         it->id = i;

         /* Allocate KNI context */
         snprintf(mz_name, RTE_MEMZONE_NAMESIZE, "KNI_INFO_%d", i);
         mz = kni_memzone_reserve(mz_name, sizeof(struct rte_kni),
                     SOCKET_ID_ANY, );
         KNI_MEM_CHECK(mz == NULL);
         it->m_ctx = mz;

         /* TX RING */
         snprintf(obj_name, OBJNAMSIZ, "kni_tx_%d", i);
         mz = kni_memzone_reserve(obj_name, KNI_FIFO_SIZE,
                             SOCKET_ID_ANY, );
         KNI_MEM_CHECK(mz == NULL);
         it->m_tx_q = mz;

         /* RX RING */
         snprintf(obj_name, OBJNAMSIZ, "kni_rx_%d", i);
         mz = kni_memzone_reserve(obj_name, KNI_FIFO_SIZE,
                             SOCKET_ID_ANY, );
         KNI_MEM_CHECK(mz == NULL);
         it->m_rx_q = mz;

         /* ALLOC RING */
         snprintf(obj_name, OBJNAMSIZ, "kni_alloc_%d", i);
         mz = kni_memzone_reserve(obj_name, KNI_FIFO_SIZE,
                             SOCKET_ID_ANY, );
         KNI_MEM_CHECK(mz == NULL);
         it->m_alloc_q = mz;

         /* FREE RING */
         snprintf(obj_name, OBJNAMSIZ, "kni_free_%d", i);
         mz = kni_memzone_reserve(obj_name, KNI_FIFO_SIZE,
                             SOCKET_ID_ANY, );
         KNI_MEM_CHECK(mz == NULL);
         it->m_free_q = mz;

         /* Request RING */
         snprintf(obj_name, OBJNAMSIZ, "kni_req_%d", i);
         mz = kni_memzone_reserve(obj_name, KNI_FIFO_SIZE,
                             SOCKET_ID_ANY, );
         KNI_MEM_CHECK(mz == NULL);
         it->m_req_q = mz;

         /* Response RING */
         snprintf(obj_name, OBJNAMSIZ, "kni_resp_%d", i);
         mz = kni_memzone_reserve(obj_name, KNI_FIFO_SIZE,
                             SOCKET_ID_ANY, );
         KNI_MEM_CHECK(mz == NULL);
         it->m_resp_q = mz;

         /* Req/Resp sync mem area */
         snprintf(obj_name, OBJNAMSIZ, "kni_sync_%d", i);
         mz = kni_memzone_reserve(obj_name, KNI_FIFO_SIZE,
                             SOCKET_ID_ANY, );
         KNI_MEM_CHECK(mz == NULL);
         it->m_sync_addr = mz;

         if ((i+) == max_kni_ifaces) {
             it->next = NULL;
             kni_memzone_pool.free_tail = it;
         } else
             it->next = &kni_memzone_pool.slots[i+];
     }

     return;

 kni_fail:
     rte_panic("Unable to allocate memory for max_kni_ifaces:%d. Increase the amount of hugepages memory\n",
              max_kni_ifaces);
 }

对上图中所有的fifo分配内存。

 static int
 kni_alloc(uint8_t port_id)
 {
     uint8_t i;
     struct rte_kni *kni;
     struct rte_kni_conf conf;
     struct kni_port_params **params = kni_port_params_array;

     if (port_id >= RTE_MAX_ETHPORTS || !params[port_id])
         return -;

     params[port_id]->nb_kni = params[port_id]->nb_lcore_k ?
                 params[port_id]->nb_lcore_k : ;

     for (i = ; i < params[port_id]->nb_kni; i++) {
         /* Clear conf at first */
         memset(&conf, , sizeof(conf));
         if (params[port_id]->nb_lcore_k) {
             snprintf(conf.name, RTE_KNI_NAMESIZE,
                     "vEth%u_%u", port_id, i);
             conf.core_id = params[port_id]->lcore_k[i];
             conf.force_bind = ;
         } else
             snprintf(conf.name, RTE_KNI_NAMESIZE,
                         "vEth%u", port_id);
         conf.group_id = (uint16_t)port_id;
         conf.mbuf_size = MAX_PACKET_SZ;
         rte_eth_macaddr_get(port_id, (struct ether_addr *)&conf.kni_mac);
         /*
          * The first KNI device associated to a port
          * is the master, for multiple kernel thread
          * environment.
          */
         if (i == ) {
             struct rte_kni_ops ops;
             struct rte_eth_dev_info dev_info;

             memset(&dev_info, , sizeof(dev_info));
             rte_eth_dev_info_get(port_id, &dev_info);
             conf.addr = dev_info.pci_dev->addr;
             conf.id = dev_info.pci_dev->id;

             memset(&ops, , sizeof(ops));
             ops.port_id = port_id;
             ops.change_mtu = kni_change_mtu;
             ops.config_network_if = kni_config_network_interface;

             kni = rte_kni_alloc(pktmbuf_pool, &conf, &ops);
         } else
             kni = rte_kni_alloc(pktmbuf_pool, &conf, NULL);

         if (!kni)
             rte_exit(EXIT_FAILURE, "Fail to create kni for "
                         "port: %d\n", port_id);
         params[port_id]->kni[i] = kni;
     }

     return ;
 }

 struct rte_kni *
 rte_kni_alloc(struct rte_mempool *pktmbuf_pool,
           const struct rte_kni_conf *conf,
           struct rte_kni_ops *ops)
 {
     int ret;
     struct rte_kni_device_info dev_info;
     struct rte_kni *ctx;
     char intf_name[RTE_KNI_NAMESIZE];
     char mz_name[RTE_MEMZONE_NAMESIZE];
     const struct rte_memzone *mz;
     const struct rte_mempool *mp;
     struct rte_kni_memzone_slot *slot = NULL;

     if (!pktmbuf_pool || !conf || !conf->name[])
         return NULL;

     /* Check if KNI subsystem has been initialized */
     if (kni_memzone_pool.initialized != ) {
         RTE_LOG(ERR, KNI, "KNI subsystem has not been initialized. Invoke rte_kni_init() first\n");
         return NULL;
     }

     /* Get an available slot from the pool */
     slot = kni_memzone_pool_alloc();
     if (!slot) {
         RTE_LOG(ERR, KNI, "Cannot allocate more KNI interfaces; increase the number of max_kni_ifaces(current %d) or release unusued ones.\n",
             kni_memzone_pool.max_ifaces);
         return NULL;
     }

     /* Recover ctx */
     ctx = slot->m_ctx->addr;
     snprintf(intf_name, RTE_KNI_NAMESIZE, "%s", conf->name);

     if (ctx->in_use) {
         RTE_LOG(ERR, KNI, "KNI %s is in use\n", ctx->name);
         return NULL;
     }
     memset(ctx, , sizeof(struct rte_kni));
     if (ops)
         memcpy(&ctx->ops, ops, sizeof(struct rte_kni_ops));

     memset(&dev_info, , sizeof(dev_info));
     dev_info.bus = conf->addr.bus;
     dev_info.devid = conf->addr.devid;
     dev_info.function = conf->addr.function;
     dev_info.vendor_id = conf->id.vendor_id;
     dev_info.device_id = conf->id.device_id;
     dev_info.core_id = conf->core_id;
     dev_info.force_bind = conf->force_bind;
     dev_info.group_id = conf->group_id;
     dev_info.mbuf_size = conf->mbuf_size;

     snprintf(ctx->name, RTE_KNI_NAMESIZE, "%s", intf_name);
     snprintf(dev_info.name, RTE_KNI_NAMESIZE, "%s", intf_name);

     RTE_LOG(INFO, KNI, "pci: %02x:%02x:%02x \t %02x:%02x\n",
         dev_info.bus, dev_info.devid, dev_info.function,
             dev_info.vendor_id, dev_info.device_id);
     /* TX RING */
     mz = slot->m_tx_q;
     ctx->tx_q = mz->addr;
     kni_fifo_init(ctx->tx_q, KNI_FIFO_COUNT_MAX);
     dev_info.tx_phys = mz->phys_addr;

     /* RX RING */
     mz = slot->m_rx_q;
     ctx->rx_q = mz->addr;
     kni_fifo_init(ctx->rx_q, KNI_FIFO_COUNT_MAX);
     dev_info.rx_phys = mz->phys_addr;

     /* ALLOC RING */
     mz = slot->m_alloc_q;
     ctx->alloc_q = mz->addr;
     kni_fifo_init(ctx->alloc_q, KNI_FIFO_COUNT_MAX);
     dev_info.alloc_phys = mz->phys_addr;

     /* FREE RING */
     mz = slot->m_free_q;
     ctx->free_q = mz->addr;
     kni_fifo_init(ctx->free_q, KNI_FIFO_COUNT_MAX);
     dev_info.free_phys = mz->phys_addr;

     /* Request RING */
     mz = slot->m_req_q;
     ctx->req_q = mz->addr;
     kni_fifo_init(ctx->req_q, KNI_FIFO_COUNT_MAX);
     dev_info.req_phys = mz->phys_addr;

     /* Response RING */
     mz = slot->m_resp_q;
     ctx->resp_q = mz->addr;
     kni_fifo_init(ctx->resp_q, KNI_FIFO_COUNT_MAX);
     dev_info.resp_phys = mz->phys_addr;

     /* Req/Resp sync mem area */
     mz = slot->m_sync_addr;
     ctx->sync_addr = mz->addr;
     dev_info.sync_va = mz->addr;
     dev_info.sync_phys = mz->phys_addr;

     /* MBUF mempool */
     snprintf(mz_name, sizeof(mz_name), RTE_MEMPOOL_MZ_FORMAT,
         pktmbuf_pool->name);
     mz = rte_memzone_lookup(mz_name);
     KNI_MEM_CHECK(mz == NULL);
     mp = (struct rte_mempool *)mz->addr;
     /* KNI currently requires to have only one memory chunk */
     if (mp->nb_mem_chunks != )
         goto kni_fail;

     dev_info.mbuf_va = STAILQ_FIRST(&mp->mem_list)->addr;
     dev_info.mbuf_phys = STAILQ_FIRST(&mp->mem_list)->phys_addr;
     ctx->pktmbuf_pool = pktmbuf_pool;
     ctx->group_id = conf->group_id;
     ctx->slot_id = slot->id;
     ctx->mbuf_size = conf->mbuf_size;

     dev_info.kni_mac = conf->kni_mac;

     ret = ioctl(kni_fd, RTE_KNI_IOCTL_CREATE, &dev_info);
     KNI_MEM_CHECK(ret < );

     ctx->in_use = ;

     /* Allocate mbufs and then put them into alloc_q */
     kni_allocate_mbufs(ctx);

     return ctx;

 kni_fail:
     if (slot)
         kni_memzone_pool_release(&kni_memzone_pool.slots[slot->id]);

     return NULL;
 }

其中ret = ioctl(kni_fd, RTE_KNI_IOCTL_CREATE, &dev_info);就是讲dev_info传入内核。

 static int
 main_loop(__rte_unused void *arg)
 {
     uint8_t i, nb_ports = rte_eth_dev_count();
     int32_t f_stop;
     const unsigned lcore_id = rte_lcore_id();
     enum lcore_rxtx {
         LCORE_NONE,
         LCORE_RX,
         LCORE_TX,
         LCORE_MAX
     };
     enum lcore_rxtx flag = LCORE_NONE;

     for (i = ; i < nb_ports; i++) {
         if (!kni_port_params_array[i])
             continue;
         if (kni_port_params_array[i]->lcore_rx == (uint8_t)lcore_id) {
             flag = LCORE_RX;
             break;
         } else if (kni_port_params_array[i]->lcore_tx ==
                         (uint8_t)lcore_id) {
             flag = LCORE_TX;
             break;
         }
     }

     if (flag == LCORE_RX) {
         RTE_LOG(INFO, APP, "Lcore %u is reading from port %d\n",
                     kni_port_params_array[i]->lcore_rx,
                     kni_port_params_array[i]->port_id);
         while () {
             f_stop = rte_atomic32_read(&kni_stop);
             if (f_stop)
                 break;
             kni_ingress(kni_port_params_array[i]);
         }
     } else if (flag == LCORE_TX) {
         RTE_LOG(INFO, APP, "Lcore %u is writing to port %d\n",
                     kni_port_params_array[i]->lcore_tx,
                     kni_port_params_array[i]->port_id);
         while () {
             f_stop = rte_atomic32_read(&kni_stop);
             if (f_stop)
                 break;
             kni_egress(kni_port_params_array[i]);
         }
     } else
         RTE_LOG(INFO, APP, "Lcore %u has nothing to do\n", lcore_id);

     return ;
 }

进入循环收发包，

 static void
 kni_ingress(struct kni_port_params *p)
 {
     uint8_t i, port_id;
     unsigned nb_rx, num;
     uint32_t nb_kni;
     struct rte_mbuf *pkts_burst[PKT_BURST_SZ];

     if (p == NULL)
         return;

     nb_kni = p->nb_kni;
     port_id = p->port_id;
     for (i = ; i < nb_kni; i++) {
         /* Burst rx from eth */
         nb_rx = rte_eth_rx_burst(port_id, , pkts_burst, PKT_BURST_SZ);
         if (unlikely(nb_rx > PKT_BURST_SZ)) {
             RTE_LOG(ERR, APP, "Error receiving from eth\n");
             return;
         }
         /* Burst tx to kni */
         num = rte_kni_tx_burst(p->kni[i], pkts_burst, nb_rx);
         kni_stats[port_id].rx_packets += num;
         //if(kni_stats[port_id].rx_packets != 0 && kni_stats[port_id].rx_packets%20 == 0 && num > 0)
          //   printf("recv packet num : %"PRIu64"\n",kni_stats[port_id].rx_packets);
         rte_kni_handle_request(p->kni[i]);
         if (unlikely(num < nb_rx)) {
             /* Free mbufs not tx to kni interface */
             kni_burst_free_mbufs(&pkts_burst[num], nb_rx - num);
             kni_stats[port_id].rx_dropped += nb_rx - num;
         }
     }
 }

 static void
 kni_egress(struct kni_port_params *p)
 {
     uint8_t i, port_id;
     unsigned nb_tx, num;
     uint32_t nb_kni;
     struct rte_mbuf *pkts_burst[PKT_BURST_SZ];

     if (p == NULL)
         return;

     nb_kni = p->nb_kni;
     port_id = p->port_id;
     for (i = ; i < nb_kni; i++) {
         /* Burst rx from kni */
         num = rte_kni_rx_burst(p->kni[i], pkts_burst, PKT_BURST_SZ);
         if (unlikely(num > PKT_BURST_SZ)) {
             RTE_LOG(ERR, APP, "Error receiving from KNI\n");
             return;
         }
         /* Burst tx to eth */
         nb_tx = rte_eth_tx_burst(port_id, , pkts_burst, (uint16_t)num);
         kni_stats[port_id].tx_packets += nb_tx;
         if (unlikely(nb_tx < num)) {
             /* Free mbufs not tx to NIC */
             kni_burst_free_mbufs(&pkts_burst[nb_tx], num - nb_tx);
             kni_stats[port_id].tx_dropped += num - nb_tx;
         }
     }
 }

代码就守护在这个kni网口进行收发包。篇幅有限，后面再整理。