|
|
|
|
|
# 47 | 接收网络包（上）：如何搞明白合作伙伴让我们做什么？
|
|
|
|
|
|
|
|
|
|
|
|
前面两节，我们分析了发送网络包的整个过程。这一节，我们来解析接收网络包的过程。
|
|
|
|
|
|
|
|
|
|
|
|
如果说网络包的发送是从应用层开始，层层调用，一直到网卡驱动程序的话，网络包的结束过程，就是一个反过来的过程，我们不能从应用层的读取开始，而应该从网卡接收到一个网络包开始。我们用两节来解析这个过程，这一节我们从硬件网卡解析到IP层，下一节，我们从IP层解析到Socket层。
|
|
|
|
|
|
|
|
|
|
|
|
## 设备驱动层
|
|
|
|
|
|
|
|
|
|
|
|
网卡作为一个硬件，接收到网络包，应该怎么通知操作系统，这个网络包到达了呢？咱们学习过输入输出设备和中断。没错，我们可以触发一个中断。但是这里有个问题，就是网络包的到来，往往是很难预期的。网络吞吐量比较大的时候，网络包的到达会十分频繁。这个时候，如果非常频繁地去触发中断，想想就觉得是个灾难。
|
|
|
|
|
|
|
|
|
|
|
|
比如说，CPU正在做某个事情，一些网络包来了，触发了中断，CPU停下手里的事情，去处理这些网络包，处理完毕按照中断处理的逻辑，应该回去继续处理其他事情。这个时候，另一些网络包又来了，又触发了中断，CPU手里的事情还没捂热，又要停下来去处理网络包。能不能大家要来的一起来，把网络包好好处理一把，然后再回去集中处理其他事情呢？
|
|
|
|
|
|
|
|
|
|
|
|
网络包能不能一起来，这个我们没法儿控制，但是我们可以有一种机制，就是当一些网络包到来触发了中断，内核处理完这些网络包之后，我们可以先进入主动轮询poll网卡的方式，主动去接收到来的网络包。如果一直有，就一直处理，等处理告一段落，就返回干其他的事情。当再有下一批网络包到来的时候，再中断，再轮询poll。这样就会大大减少中断的数量，提升网络处理的效率，这种处理方式我们称为**NAPI**。
|
|
|
|
|
|
|
|
|
|
|
|
为了帮你了解设备驱动层的工作机制，我们还是以上一节发送网络包时的网卡drivers/net/ethernet/intel/ixgb/ixgb\_main.c为例子，来进行解析。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
static struct pci_driver ixgb_driver = {
|
|
|
|
|
|
.name = ixgb_driver_name,
|
|
|
|
|
|
.id_table = ixgb_pci_tbl,
|
|
|
|
|
|
.probe = ixgb_probe,
|
|
|
|
|
|
.remove = ixgb_remove,
|
|
|
|
|
|
.err_handler = &ixgb_err_handler
|
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
MODULE_AUTHOR("Intel Corporation, <linux.nics@intel.com>");
|
|
|
|
|
|
MODULE_DESCRIPTION("Intel(R) PRO/10GbE Network Driver");
|
|
|
|
|
|
MODULE_LICENSE("GPL");
|
|
|
|
|
|
MODULE_VERSION(DRV_VERSION);
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
|
* ixgb_init_module - Driver Registration Routine
|
|
|
|
|
|
*
|
|
|
|
|
|
* ixgb_init_module is the first routine called when the driver is
|
|
|
|
|
|
* loaded. All it does is register with the PCI subsystem.
|
|
|
|
|
|
**/
|
|
|
|
|
|
|
|
|
|
|
|
static int __init
|
|
|
|
|
|
ixgb_init_module(void)
|
|
|
|
|
|
{
|
|
|
|
|
|
pr_info("%s - version %s\n", ixgb_driver_string, ixgb_driver_version);
|
|
|
|
|
|
pr_info("%s\n", ixgb_copyright);
|
|
|
|
|
|
|
|
|
|
|
|
return pci_register_driver(&ixgb_driver);
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
module_init(ixgb_init_module);
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在网卡驱动程序初始化的时候，我们会调用ixgb\_init\_module，注册一个驱动ixgb\_driver，并且调用它的probe函数ixgb\_probe。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
static int
|
|
|
|
|
|
ixgb_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
|
|
|
|
|
|
{
|
|
|
|
|
|
struct net_device *netdev = NULL;
|
|
|
|
|
|
struct ixgb_adapter *adapter;
|
|
|
|
|
|
......
|
|
|
|
|
|
netdev = alloc_etherdev(sizeof(struct ixgb_adapter));
|
|
|
|
|
|
SET_NETDEV_DEV(netdev, &pdev->dev);
|
|
|
|
|
|
|
|
|
|
|
|
pci_set_drvdata(pdev, netdev);
|
|
|
|
|
|
adapter = netdev_priv(netdev);
|
|
|
|
|
|
adapter->netdev = netdev;
|
|
|
|
|
|
adapter->pdev = pdev;
|
|
|
|
|
|
adapter->hw.back = adapter;
|
|
|
|
|
|
adapter->msg_enable = netif_msg_init(debug, DEFAULT_MSG_ENABLE);
|
|
|
|
|
|
|
|
|
|
|
|
adapter->hw.hw_addr = pci_ioremap_bar(pdev, BAR_0);
|
|
|
|
|
|
......
|
|
|
|
|
|
netdev->netdev_ops = &ixgb_netdev_ops;
|
|
|
|
|
|
ixgb_set_ethtool_ops(netdev);
|
|
|
|
|
|
netdev->watchdog_timeo = 5 * HZ;
|
|
|
|
|
|
netif_napi_add(netdev, &adapter->napi, ixgb_clean, 64);
|
|
|
|
|
|
|
|
|
|
|
|
strncpy(netdev->name, pci_name(pdev), sizeof(netdev->name) - 1);
|
|
|
|
|
|
|
|
|
|
|
|
adapter->bd_number = cards_found;
|
|
|
|
|
|
adapter->link_speed = 0;
|
|
|
|
|
|
adapter->link_duplex = 0;
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在ixgb\_probe中，我们会创建一个struct net\_device表示这个网络设备，并且netif\_napi\_add函数为这个网络设备注册一个轮询poll函数ixgb\_clean，将来一旦出现网络包的时候，就是要通过它来轮询了。
|
|
|
|
|
|
|
|
|
|
|
|
当一个网卡被激活的时候，我们会调用函数ixgb\_open->ixgb\_up，在这里面注册一个硬件的中断处理函数。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
int
|
|
|
|
|
|
ixgb_up(struct ixgb_adapter *adapter)
|
|
|
|
|
|
{
|
|
|
|
|
|
struct net_device *netdev = adapter->netdev;
|
|
|
|
|
|
......
|
|
|
|
|
|
err = request_irq(adapter->pdev->irq, ixgb_intr, irq_flags,
|
|
|
|
|
|
netdev->name, netdev);
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
|
* ixgb_intr - Interrupt Handler
|
|
|
|
|
|
* @irq: interrupt number
|
|
|
|
|
|
* @data: pointer to a network interface device structure
|
|
|
|
|
|
**/
|
|
|
|
|
|
|
|
|
|
|
|
static irqreturn_t
|
|
|
|
|
|
ixgb_intr(int irq, void *data)
|
|
|
|
|
|
{
|
|
|
|
|
|
struct net_device *netdev = data;
|
|
|
|
|
|
struct ixgb_adapter *adapter = netdev_priv(netdev);
|
|
|
|
|
|
struct ixgb_hw *hw = &adapter->hw;
|
|
|
|
|
|
......
|
|
|
|
|
|
if (napi_schedule_prep(&adapter->napi)) {
|
|
|
|
|
|
IXGB_WRITE_REG(&adapter->hw, IMC, ~0);
|
|
|
|
|
|
__napi_schedule(&adapter->napi);
|
|
|
|
|
|
}
|
|
|
|
|
|
return IRQ_HANDLED;
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
如果一个网络包到来，触发了硬件中断，就会调用ixgb\_intr，这里面会调用\_\_napi\_schedule。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
/**
|
|
|
|
|
|
* __napi_schedule - schedule for receive
|
|
|
|
|
|
* @n: entry to schedule
|
|
|
|
|
|
*
|
|
|
|
|
|
* The entry's receive function will be scheduled to run.
|
|
|
|
|
|
* Consider using __napi_schedule_irqoff() if hard irqs are masked.
|
|
|
|
|
|
*/
|
|
|
|
|
|
void __napi_schedule(struct napi_struct *n)
|
|
|
|
|
|
{
|
|
|
|
|
|
unsigned long flags;
|
|
|
|
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
|
|
|
|
|
____napi_schedule(this_cpu_ptr(&softnet_data), n);
|
|
|
|
|
|
local_irq_restore(flags);
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static inline void ____napi_schedule(struct softnet_data *sd,
|
|
|
|
|
|
struct napi_struct *napi)
|
|
|
|
|
|
{
|
|
|
|
|
|
list_add_tail(&napi->poll_list, &sd->poll_list);
|
|
|
|
|
|
__raise_softirq_irqoff(NET_RX_SOFTIRQ);
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
\_\_napi\_schedule是处于中断处理的关键部分，在他被调用的时候，中断是暂时关闭的，但是处理网络包是个复杂的过程，需要到延迟处理部分，所以\_\_\_\_napi\_schedule将当前设备放到struct softnet\_data结构的poll\_list里面，说明在延迟处理部分可以接着处理这个poll\_list里面的网络设备。
|
|
|
|
|
|
|
|
|
|
|
|
然后\_\_\_\_napi\_schedule触发一个软中断NET\_RX\_SOFTIRQ，通过软中断触发中断处理的延迟处理部分，也是常用的手段。
|
|
|
|
|
|
|
|
|
|
|
|
上一节，我们知道，软中断NET\_RX\_SOFTIRQ对应的中断处理函数是net\_rx\_action。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
static __latent_entropy void net_rx_action(struct softirq_action *h)
|
|
|
|
|
|
{
|
|
|
|
|
|
struct softnet_data *sd = this_cpu_ptr(&softnet_data);
|
|
|
|
|
|
LIST_HEAD(list);
|
|
|
|
|
|
list_splice_init(&sd->poll_list, &list);
|
|
|
|
|
|
......
|
|
|
|
|
|
for (;;) {
|
|
|
|
|
|
struct napi_struct *n;
|
|
|
|
|
|
......
|
|
|
|
|
|
n = list_first_entry(&list, struct napi_struct, poll_list);
|
|
|
|
|
|
budget -= napi_poll(n, &repoll);
|
|
|
|
|
|
}
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在net\_rx\_action中，会得到struct softnet\_data结构，这个结构在发送的时候我们也遇到过。当时它的output\_queue用于网络包的发送，这里的poll\_list用于网络包的接收。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
struct softnet_data {
|
|
|
|
|
|
struct list_head poll_list;
|
|
|
|
|
|
......
|
|
|
|
|
|
struct Qdisc *output_queue;
|
|
|
|
|
|
struct Qdisc **output_queue_tailp;
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在net\_rx\_action中，接下来是一个循环，在poll\_list里面取出网络包到达的设备，然后调用napi\_poll来轮询这些设备，napi\_poll会调用最初设备初始化的时候，注册的poll函数，对于ixgb\_driver，对应的函数是ixgb\_clean。
|
|
|
|
|
|
|
|
|
|
|
|
ixgb\_clean会调用ixgb\_clean\_rx\_irq。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
static bool
|
|
|
|
|
|
ixgb_clean_rx_irq(struct ixgb_adapter *adapter, int *work_done, int work_to_do)
|
|
|
|
|
|
{
|
|
|
|
|
|
struct ixgb_desc_ring *rx_ring = &adapter->rx_ring;
|
|
|
|
|
|
struct net_device *netdev = adapter->netdev;
|
|
|
|
|
|
struct pci_dev *pdev = adapter->pdev;
|
|
|
|
|
|
struct ixgb_rx_desc *rx_desc, *next_rxd;
|
|
|
|
|
|
struct ixgb_buffer *buffer_info, *next_buffer, *next2_buffer;
|
|
|
|
|
|
u32 length;
|
|
|
|
|
|
unsigned int i, j;
|
|
|
|
|
|
int cleaned_count = 0;
|
|
|
|
|
|
bool cleaned = false;
|
|
|
|
|
|
|
|
|
|
|
|
i = rx_ring->next_to_clean;
|
|
|
|
|
|
rx_desc = IXGB_RX_DESC(*rx_ring, i);
|
|
|
|
|
|
buffer_info = &rx_ring->buffer_info[i];
|
|
|
|
|
|
|
|
|
|
|
|
while (rx_desc->status & IXGB_RX_DESC_STATUS_DD) {
|
|
|
|
|
|
struct sk_buff *skb;
|
|
|
|
|
|
u8 status;
|
|
|
|
|
|
|
|
|
|
|
|
status = rx_desc->status;
|
|
|
|
|
|
skb = buffer_info->skb;
|
|
|
|
|
|
buffer_info->skb = NULL;
|
|
|
|
|
|
|
|
|
|
|
|
prefetch(skb->data - NET_IP_ALIGN);
|
|
|
|
|
|
|
|
|
|
|
|
if (++i == rx_ring->count)
|
|
|
|
|
|
i = 0;
|
|
|
|
|
|
next_rxd = IXGB_RX_DESC(*rx_ring, i);
|
|
|
|
|
|
prefetch(next_rxd);
|
|
|
|
|
|
|
|
|
|
|
|
j = i + 1;
|
|
|
|
|
|
if (j == rx_ring->count)
|
|
|
|
|
|
j = 0;
|
|
|
|
|
|
next2_buffer = &rx_ring->buffer_info[j];
|
|
|
|
|
|
prefetch(next2_buffer);
|
|
|
|
|
|
|
|
|
|
|
|
next_buffer = &rx_ring->buffer_info[i];
|
|
|
|
|
|
......
|
|
|
|
|
|
length = le16_to_cpu(rx_desc->length);
|
|
|
|
|
|
rx_desc->length = 0;
|
|
|
|
|
|
......
|
|
|
|
|
|
ixgb_check_copybreak(&adapter->napi, buffer_info, length, &skb);
|
|
|
|
|
|
|
|
|
|
|
|
/* Good Receive */
|
|
|
|
|
|
skb_put(skb, length);
|
|
|
|
|
|
|
|
|
|
|
|
/* Receive Checksum Offload */
|
|
|
|
|
|
ixgb_rx_checksum(adapter, rx_desc, skb);
|
|
|
|
|
|
|
|
|
|
|
|
skb->protocol = eth_type_trans(skb, netdev);
|
|
|
|
|
|
|
|
|
|
|
|
netif_receive_skb(skb);
|
|
|
|
|
|
......
|
|
|
|
|
|
/* use prefetched values */
|
|
|
|
|
|
rx_desc = next_rxd;
|
|
|
|
|
|
buffer_info = next_buffer;
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
rx_ring->next_to_clean = i;
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在网络设备的驱动层，有一个用于接收网络包的rx\_ring。它是一个环，从网卡硬件接收的包会放在这个环里面。这个环里面的buffer\_info\[\]是一个数组，存放的是网络包的内容。i和j是这个数组的下标，在ixgb\_clean\_rx\_irq里面的while循环中，依次处理环里面的数据。在这里面，我们看到了i和j加一之后，如果超过了数组的大小，就跳回下标0，就说明这是一个环。
|
|
|
|
|
|
|
|
|
|
|
|
ixgb\_check\_copybreak函数将buffer\_info里面的内容，拷贝到struct sk\_buff \*skb，从而可以作为一个网络包进行后续的处理，然后调用netif\_receive\_skb。
|
|
|
|
|
|
|
|
|
|
|
|
## 网络协议栈的二层逻辑
|
|
|
|
|
|
|
|
|
|
|
|
从netif\_receive\_skb函数开始，我们就进入了内核的网络协议栈。
|
|
|
|
|
|
|
|
|
|
|
|
接下来的调用链为：netif\_receive\_skb->netif\_receive\_skb\_internal->\_\_netif\_receive\_skb->\_\_netif\_receive\_skb\_core。
|
|
|
|
|
|
|
|
|
|
|
|
在\_\_netif\_receive\_skb\_core中，我们先是处理了二层的一些逻辑。例如，对于VLAN的处理，接下来要想办法交给第三层。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
static int __netif_receive_skb_core(struct sk_buff *skb, bool pfmemalloc)
|
|
|
|
|
|
{
|
|
|
|
|
|
struct packet_type *ptype, *pt_prev;
|
|
|
|
|
|
......
|
|
|
|
|
|
type = skb->protocol;
|
|
|
|
|
|
......
|
|
|
|
|
|
deliver_ptype_list_skb(skb, &pt_prev, orig_dev, type,
|
|
|
|
|
|
&orig_dev->ptype_specific);
|
|
|
|
|
|
if (pt_prev) {
|
|
|
|
|
|
ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev);
|
|
|
|
|
|
}
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static inline void deliver_ptype_list_skb(struct sk_buff *skb,
|
|
|
|
|
|
struct packet_type **pt,
|
|
|
|
|
|
struct net_device *orig_dev,
|
|
|
|
|
|
__be16 type,
|
|
|
|
|
|
struct list_head *ptype_list)
|
|
|
|
|
|
{
|
|
|
|
|
|
struct packet_type *ptype, *pt_prev = *pt;
|
|
|
|
|
|
|
|
|
|
|
|
list_for_each_entry_rcu(ptype, ptype_list, list) {
|
|
|
|
|
|
if (ptype->type != type)
|
|
|
|
|
|
continue;
|
|
|
|
|
|
if (pt_prev)
|
|
|
|
|
|
deliver_skb(skb, pt_prev, orig_dev);
|
|
|
|
|
|
pt_prev = ptype;
|
|
|
|
|
|
}
|
|
|
|
|
|
*pt = pt_prev;
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在网络包struct sk\_buff里面，二层的头里面有一个protocol，表示里面一层，也即三层是什么协议。deliver\_ptype\_list\_skb在一个协议列表中逐个匹配。如果能够匹配到，就返回。
|
|
|
|
|
|
|
|
|
|
|
|
这些协议的注册在网络协议栈初始化的时候， inet\_init函数调用dev\_add\_pack(&ip\_packet\_type)，添加IP协议。协议被放在一个链表里面。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
void dev_add_pack(struct packet_type *pt)
|
|
|
|
|
|
{
|
|
|
|
|
|
struct list_head *head = ptype_head(pt);
|
|
|
|
|
|
list_add_rcu(&pt->list, head);
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static inline struct list_head *ptype_head(const struct packet_type *pt)
|
|
|
|
|
|
{
|
|
|
|
|
|
if (pt->type == htons(ETH_P_ALL))
|
|
|
|
|
|
return pt->dev ? &pt->dev->ptype_all : &ptype_all;
|
|
|
|
|
|
else
|
|
|
|
|
|
return pt->dev ? &pt->dev->ptype_specific : &ptype_base[ntohs(pt->type) & PTYPE_HASH_MASK];
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
假设这个时候的网络包是一个IP包，则在这个链表里面一定能够找到ip\_packet\_type，在\_\_netif\_receive\_skb\_core中会调用ip\_packet\_type的func函数。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
static struct packet_type ip_packet_type __read_mostly = {
|
|
|
|
|
|
.type = cpu_to_be16(ETH_P_IP),
|
|
|
|
|
|
.func = ip_rcv,
|
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
从上面的定义我们可以看出，接下来，ip\_rcv会被调用。
|
|
|
|
|
|
|
|
|
|
|
|
## 网络协议栈的IP层
|
|
|
|
|
|
|
|
|
|
|
|
从ip\_rcv函数开始，我们的处理逻辑就从二层到了三层，IP层。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
int ip_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev)
|
|
|
|
|
|
{
|
|
|
|
|
|
const struct iphdr *iph;
|
|
|
|
|
|
struct net *net;
|
|
|
|
|
|
u32 len;
|
|
|
|
|
|
......
|
|
|
|
|
|
net = dev_net(dev);
|
|
|
|
|
|
......
|
|
|
|
|
|
iph = ip_hdr(skb);
|
|
|
|
|
|
len = ntohs(iph->tot_len);
|
|
|
|
|
|
skb->transport_header = skb->network_header + iph->ihl*4;
|
|
|
|
|
|
......
|
|
|
|
|
|
return NF_HOOK(NFPROTO_IPV4, NF_INET_PRE_ROUTING,
|
|
|
|
|
|
net, NULL, skb, dev, NULL,
|
|
|
|
|
|
ip_rcv_finish);
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在ip\_rcv中，得到IP头，然后又遇到了我们见过多次的NF\_HOOK，这次因为是接收网络包，第一个hook点是NF\_INET\_PRE\_ROUTING，也就是iptables的PREROUTING链。如果里面有规则，则执行规则，然后调用ip\_rcv\_finish。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
static int ip_rcv_finish(struct net *net, struct sock *sk, struct sk_buff *skb)
|
|
|
|
|
|
{
|
|
|
|
|
|
const struct iphdr *iph = ip_hdr(skb);
|
|
|
|
|
|
struct net_device *dev = skb->dev;
|
|
|
|
|
|
struct rtable *rt;
|
|
|
|
|
|
int err;
|
|
|
|
|
|
......
|
|
|
|
|
|
rt = skb_rtable(skb);
|
|
|
|
|
|
.....
|
|
|
|
|
|
return dst_input(skb);
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static inline int dst_input(struct sk_buff *skb)
|
|
|
|
|
|
{
|
|
|
|
|
|
return skb_dst(skb)->input(skb);
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
ip\_rcv\_finish得到网络包对应的路由表，然后调用dst\_input，在dst\_input中，调用的是struct rtable的成员的dst的input函数。在rt\_dst\_alloc中，我们可以看到，input函数指向的是ip\_local\_deliver。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
int ip_local_deliver(struct sk_buff *skb)
|
|
|
|
|
|
{
|
|
|
|
|
|
/*
|
|
|
|
|
|
* Reassemble IP fragments.
|
|
|
|
|
|
*/
|
|
|
|
|
|
struct net *net = dev_net(skb->dev);
|
|
|
|
|
|
|
|
|
|
|
|
if (ip_is_fragment(ip_hdr(skb))) {
|
|
|
|
|
|
if (ip_defrag(net, skb, IP_DEFRAG_LOCAL_DELIVER))
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
return NF_HOOK(NFPROTO_IPV4, NF_INET_LOCAL_IN,
|
|
|
|
|
|
net, NULL, skb, skb->dev, NULL,
|
|
|
|
|
|
ip_local_deliver_finish);
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在ip\_local\_deliver函数中，如果IP层进行了分段，则进行重新的组合。接下来就是我们熟悉的NF\_HOOK。hook点在NF\_INET\_LOCAL\_IN，对应iptables里面的INPUT链。在经过iptables规则处理完毕后，我们调用ip\_local\_deliver\_finish。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
static int ip_local_deliver_finish(struct net *net, struct sock *sk, struct sk_buff *skb)
|
|
|
|
|
|
{
|
|
|
|
|
|
__skb_pull(skb, skb_network_header_len(skb));
|
|
|
|
|
|
|
|
|
|
|
|
int protocol = ip_hdr(skb)->protocol;
|
|
|
|
|
|
const struct net_protocol *ipprot;
|
|
|
|
|
|
|
|
|
|
|
|
ipprot = rcu_dereference(inet_protos[protocol]);
|
|
|
|
|
|
if (ipprot) {
|
|
|
|
|
|
int ret;
|
|
|
|
|
|
ret = ipprot->handler(skb);
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在IP头中，有一个字段protocol用于指定里面一层的协议，在这里应该是TCP协议。于是，从inet\_protos数组中，找出TCP协议对应的处理函数。这个数组的定义如下，里面的内容是struct net\_protocol。
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
struct net_protocol __rcu *inet_protos[MAX_INET_PROTOS] __read_mostly;
|
|
|
|
|
|
|
|
|
|
|
|
int inet_add_protocol(const struct net_protocol *prot, unsigned char protocol)
|
|
|
|
|
|
{
|
|
|
|
|
|
......
|
|
|
|
|
|
return !cmpxchg((const struct net_protocol **)&inet_protos[protocol],
|
|
|
|
|
|
NULL, prot) ? 0 : -1;
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static int __init inet_init(void)
|
|
|
|
|
|
{
|
|
|
|
|
|
......
|
|
|
|
|
|
if (inet_add_protocol(&udp_protocol, IPPROTO_UDP) < 0)
|
|
|
|
|
|
pr_crit("%s: Cannot add UDP protocol\n", __func__);
|
|
|
|
|
|
if (inet_add_protocol(&tcp_protocol, IPPROTO_TCP) < 0)
|
|
|
|
|
|
pr_crit("%s: Cannot add TCP protocol\n", __func__);
|
|
|
|
|
|
......
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static struct net_protocol tcp_protocol = {
|
|
|
|
|
|
.early_demux = tcp_v4_early_demux,
|
|
|
|
|
|
.early_demux_handler = tcp_v4_early_demux,
|
|
|
|
|
|
.handler = tcp_v4_rcv,
|
|
|
|
|
|
.err_handler = tcp_v4_err,
|
|
|
|
|
|
.no_policy = 1,
|
|
|
|
|
|
.netns_ok = 1,
|
|
|
|
|
|
.icmp_strict_tag_validation = 1,
|
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
static struct net_protocol udp_protocol = {
|
|
|
|
|
|
.early_demux = udp_v4_early_demux,
|
|
|
|
|
|
.early_demux_handler = udp_v4_early_demux,
|
|
|
|
|
|
.handler = udp_rcv,
|
|
|
|
|
|
.err_handler = udp_err,
|
|
|
|
|
|
.no_policy = 1,
|
|
|
|
|
|
.netns_ok = 1,
|
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
在系统初始化的时候，网络协议栈的初始化调用的是inet\_init，它会调用inet\_add\_protocol，将TCP协议对应的处理函数tcp\_protocol、UDP协议对应的处理函数udp\_protocol，放到inet\_protos数组中。
|
|
|
|
|
|
|
|
|
|
|
|
在上面的网络包的接收过程中，会取出TCP协议对应的处理函数tcp\_protocol，然后调用handler函数，也即tcp\_v4\_rcv函数。
|
|
|
|
|
|
|
|
|
|
|
|
## 总结时刻
|
|
|
|
|
|
|
|
|
|
|
|
这一节我们讲了接收网络包的上半部分，分以下几个层次。
|
|
|
|
|
|
|
|
|
|
|
|
* 硬件网卡接收到网络包之后，通过DMA技术，将网络包放入Ring Buffer。
|
|
|
|
|
|
* 硬件网卡通过中断通知CPU新的网络包的到来。
|
|
|
|
|
|
* 网卡驱动程序会注册中断处理函数ixgb\_intr。
|
|
|
|
|
|
* 中断处理函数处理完需要暂时屏蔽中断的核心流程之后，通过软中断NET\_RX\_SOFTIRQ触发接下来的处理过程。
|
|
|
|
|
|
* NET\_RX\_SOFTIRQ软中断处理函数net\_rx\_action，net\_rx\_action会调用napi\_poll，进而调用ixgb\_clean\_rx\_irq，从Ring Buffer中读取数据到内核struct sk\_buff。
|
|
|
|
|
|
* 调用netif\_receive\_skb进入内核网络协议栈，进行一些关于VLAN的二层逻辑处理后，调用ip\_rcv进入三层IP层。
|
|
|
|
|
|
* 在IP层，会处理iptables规则，然后调用ip\_local\_deliver，交给更上层TCP层。
|
|
|
|
|
|
* 在TCP层调用tcp\_v4\_rcv。
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
## 课堂练习
|
|
|
|
|
|
|
|
|
|
|
|
我们没有仔细分析对于二层VLAN的处理，请你研究一下VLAN的原理，然后在代码中看一下对于VLAN的处理过程，这是一项重要的网络基础知识。
|
|
|
|
|
|
|
|
|
|
|
|
欢迎留言和我分享你的疑惑和见解 ，也欢迎可以收藏本节内容，反复研读。你也可以把今天的内容分享给你的朋友，和他一起学习和进步。
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
