/* Process an incoming IP datagram fragment. */
static char *
ip_defrag(struct ip *iph, struct sk_buff *skb)
{
struct ipfrag *prev, *next, *tmp;
struct ipfrag *tfp;
struct ipq *qp;
char *skb2;
unsigned char *ptr;
int flags, offset;
int i, ihl, end;
if (!hostfrag_find(iph) && skb)
hostfrag_create(iph);
/* Start by cleaning up the memory. */
if (this_host)
if (this_host->ip_frag_mem > IPFRAG_HIGH_THRESH)
ip_evictor();
/* Find the entry of this IP datagram in the "incomplete datagrams" queue. */
if (this_host)
qp = ip_find(iph);
else
qp = 0;
/* Is this a non-fragmented datagram? */
offset = ntohs(iph->ip_off);
flags = offset & ~IP_OFFSET;
offset &= IP_OFFSET;
if (((flags & IP_MF) == 0) && (offset == 0)) {
if (qp != NULL)
ip_free(qp); /* Fragmented frame replaced by full
unfragmented copy */
return 0;
}
/* ip_evictor() could have removed all queues for the current host */
if (!this_host)
hostfrag_create(iph);
offset <<= 3; /* offset is in 8-byte chunks */
ihl = iph->ip_hl * 4;
/*
If the queue already existed, keep restarting its timer as long as
we still are receiving fragments. Otherwise, create a fresh queue
entry.
*/
if (qp != NULL) {
/* ANK. If the first fragment is received, we should remember the correct
IP header (with options) */
if (offset == 0) {
qp->ihlen = ihl;
memcpy(qp->iph, iph, ihl + 8);
}
del_timer(&qp->timer);
qp->timer.expires = jiffies() + IP_FRAG_TIME; /* about 30 seconds */
qp->timer.data = (unsigned long) qp; /* pointer to queue */
qp->timer.function = ip_expire; /* expire function */
add_timer(&qp->timer);
}
else {
/* If we failed to create it, then discard the frame. */
if ((qp = ip_create(iph)) == NULL) {
kfree_skb(skb, FREE_READ);
return NULL;
}
}
/* Attempt to construct an oversize packet. */
if (ntohs(iph->ip_len) + (int) offset > 65535) {
// NETDEBUG(printk("Oversized packet received from %s\n", int_ntoa(iph->ip_src.s_addr)));
nids_params.syslog(NIDS_WARN_IP, NIDS_WARN_IP_OVERSIZED, iph, 0);
kfree_skb(skb, FREE_READ);
return NULL;
}
/* Determine the position of this fragment. */
end = offset + ntohs(iph->ip_len) - ihl;
/* Point into the IP datagram 'data' part. */
ptr = (unsigned char *)(skb->data + ihl);
/* Is this the final fragment? */
if ((flags & IP_MF) == 0)
qp->len = end;
/*
Find out which fragments are in front and at the back of us in the
chain of fragments so far. We must know where to put this
fragment, right?
*/
prev = NULL;
for (next = qp->fragments; next != NULL; next = next->next) {
if (next->offset >= offset)
break; /* bingo! */
prev = next;
}
/*
We found where to put this one. Check for overlap with preceding
fragment, and, if needed, align things so that any overlaps are
eliminated.
*/
if (prev != NULL && offset < prev->end) {
nids_params.syslog(NIDS_WARN_IP, NIDS_WARN_IP_OVERLAP, iph, 0);
i = prev->end - offset;
offset += i; /* ptr into datagram */
ptr += i; /* ptr into fragment data */
}
/*
Look for overlap with succeeding segments.
If we can merge fragments, do it.
*/
for (tmp = next; tmp != NULL; tmp = tfp) {
tfp = tmp->next;
if (tmp->offset >= end)
break; /* no overlaps at all */
nids_params.syslog(NIDS_WARN_IP, NIDS_WARN_IP_OVERLAP, iph, 0);
i = end - next->offset; /* overlap is 'i' bytes */
tmp->len -= i; /* so reduce size of */
tmp->offset += i; /* next fragment */
tmp->ptr += i;
/*
If we get a frag size of <= 0, remove it and the packet that it
goes with. We never throw the new frag away, so the frag being
dumped has always been charged for.
*/
if (tmp->len <= 0) {
if (tmp->prev != NULL)
tmp->prev->next = tmp->next;
else
qp->fragments = tmp->next;
if (tmp->next != NULL)
tmp->next->prev = tmp->prev;
next = tfp; /* We have killed the original next frame */
frag_kfree_skb(tmp->skb, FREE_READ);
frag_kfree_s(tmp, sizeof(struct ipfrag));
}
}
/* Insert this fragment in the chain of fragments. */
tfp = NULL;
tfp = ip_frag_create(offset, end, skb, ptr);
/*
No memory to save the fragment - so throw the lot. If we failed
the frag_create we haven't charged the queue.
*/
if (!tfp) {
nids_params.no_mem("ip_defrag");
kfree_skb(skb, FREE_READ);
return NULL;
}
/* From now on our buffer is charged to the queues. */
tfp->prev = prev;
tfp->next = next;
if (prev != NULL)
prev->next = tfp;
else
qp->fragments = tfp;
if (next != NULL)
next->prev = tfp;
/*
OK, so we inserted this new fragment into the chain. Check if we
now have a full IP datagram which we can bump up to the IP
layer...
*/
if (ip_done(qp)) {
skb2 = ip_glue(qp); /* glue together the fragments */
fprintf(stderr,"this is ipfragment.sd");
return (skb2);
}
return (NULL);
}
// 每重组一个ip碎片，就会更新对应ipq的timer
//
//
static char *
ip_defrag(struct ip *iph, struct sk_buff *skb)
{
struct ipfrag *prev, *next, *tmp;
// 指向一个碎片
struct ipfrag *tfp;
// 指向一个队列
struct ipq *qp;
// 用来放返回值
char *skb2;
// 用来进行字节操作
unsigned char *ptr;
int flags, offset;
int i, ihl, end;
// 如果成功更新全局变量this_host, 并且skb是有内容的
if (!hostfrag_find(iph) && skb)
// 生成一个碎片
hostfrag_create(iph);
/* Start by cleaning up the memory. */
// 如果当前host不为空
if (this_host)
// 如果大于上限
if (this_host->ip_frag_mem > IPFRAG_HIGH_THRESH)
// 裁剪掉一些ip碎片，直到 ip_frag_mem < IPFRAG_LOW_THRESH
ip_evictor();
/* Find the entry of this IP datagram in the "incomplete datagrams" queue. */
// 如果host存在
if (this_host)
// 找到与这个ip头相关的ip队列
qp = ip_find(iph);
else
// 否则设置队列为空
qp = 0;
/* Is this a non-fragmented datagram? */
// ip_off是一个16位的字段，高3位用来保存标志信息，低12位用来保存当前碎片的偏移
offset = ntohs(iph->ip_off); /* 先把ip_offset这个字段取出来 */
// 把高3位取出来
flags = offset & ~IP_OFFSET;
// 把低13位取出来
offset &= IP_OFFSET;
// IP_MF==0表示当前收到是碎片后面没有碎片了，并且当前收到碎片是第一个碎片，
// 显然当前碎片虽在的ip报文仅仅有一个碎片
// 那么当前碎片(刚刚收到的碎片)并不需要重组，因此可以返回了
if (((flags & IP_MF) == 0) && (offset == 0))
{
// 如果队列不为空就释放掉
if (qp != NULL)
ip_free(qp); /* Fragmented frame replaced by full unfragmented copy */
return 0;
}
/* ip_evictor() could have removed all queues for the current host */
// 如果host全部被移出了，那么重新创建一个，针对当前ip头的host
// 但是，这个host并不会包括任何东西，它是一个空的，没有ip队列
if (!this_host)
hostfrag_create(iph);
// 计算offset和头长度
// 这个offset是刚刚收到的这个碎片的offset
offset <<= 3; /* offset is in 8-byte chunks */
ihl = iph->ip_hl * 4;
/*
If the queue already existed, keep restarting its timer as long as
we still are receiving fragments. Otherwise, create a fresh queue
entry.
*/
// 如果队列存在
if (qp != NULL)
{
/* ANK. If the first fragment is received, we should remember the correct
IP header (with options) */
// 如果偏移量为0，可能是该pi报文中的第一个碎片，因此要把前8字节保存下来
if (offset == 0)
{
// 保存头长度
qp->ihlen = ihl;
// 拷贝头信息+8字节
memcpy(qp->iph, iph, ihl + 8);
}
// 停止计时
del_timer(&qp->timer);
// 重新计时
qp->timer.expires = jiffies() + IP_FRAG_TIME; /* about 30 seconds */
// 设置关联
qp->timer.data = (unsigned long) qp; /* pointer to queue */
// 注册回调函数
qp->timer.function = ip_expire; /* expire function */
// 添加计数器
add_timer(&qp->timer);
}
// 否则队列不存在
else
{
/* If we failed to create it, then discard the frame. */
// 试图创建一个
if ((qp = ip_create(iph)) == NULL)
{
// 如果创建队列失败，那么释放当前碎片的空间并返回
kfree_skb(skb, FREE_READ);
return NULL;
}
}
/* Attempt to construct an oversize packet. */
// 如果头+长度 超长， 释放空间
if (ntohs(iph->ip_len) + (int) offset > 65535)
{
// NETDEBUG(printk("Oversized packet received from %s\n", int_ntoa(iph->ip_src.s_addr)));
nids_params.syslog(NIDS_WARN_IP, NIDS_WARN_IP_OVERSIZED, iph, 0);
kfree_skb(skb, FREE_READ);
return NULL;
}
/* Determine the position of this fragment. */
// 刚刚收到的碎片分组 + 刚刚收到的碎片ip包大小 - 刚刚收到的碎片ip包头大小
// = 刚刚收到的分组的结尾
end = offset + ntohs(iph->ip_len) - ihl;
/* Point into the IP datagram 'data' part. */
// 将指针，指向刚刚收到的碎片ip包的数据开头部分
// prt 指向肉
ptr = (unsigned char *)(skb->data + ihl);
/* Is this the final fragment? */
// 如果这是最后一个碎片，那么，整个队列的长度，就是当前的end
// 否则qp->len会在后面更改，因为会有重叠部分
if ((flags & IP_MF) == 0)
qp->len = end;
/*
Find out which fragments are in front and at the back of us in the
chain of fragments so far. We must know where to put this
fragment, right?
*/
prev = NULL;
// 给定一个offset,在所有的fragments中找到第一个
// 拥有不小于给定offset的offset的碎片，然后终止循环
// 由此可以猜测，这一个函数，其实是将某一个新来的碎片插入到
// 队列合适的位置，保证offset升序排列
// next 指向的是第一个不小于当前offset的碎片
// pre指向的是前一个碎片
for (next = qp->fragments; next != NULL; next = next->next)
{
if (next->offset >= offset)
break; /* bingo! */
prev = next;
}
/*--------------------------------------------------------------
注意:
next 指向的是第一个排在当前碎片后面的 碎片；
pre 指向的是第一个排在当前碎片前面的 碎片。
next 有可能offset与当前碎片的offset一样
----------------------------------------------------------------*/
/*
We found where to put this one. Check for overlap with preceding
fragment, and, if needed, align things so that any overlaps are
eliminated.
*/
// 如果有排在当前碎片前面的分组，并且， 该分组的结束比当前分组的offset大
// 说明有重叠，应噶修正当前offset，以先来的为准
if (prev != NULL && offset < prev->end)
{
nids_params.syslog(NIDS_WARN_IP, NIDS_WARN_IP_OVERLAP, iph, 0);
i = prev->end - offset;
// 将当前收到的碎片的offset增加i
offset += i; /* ptr into datagram */
// 将当前收到的碎片的指正向后移动i
ptr += i; /* ptr into fragment data */
}
/*
Look for overlap with succeeding segments.
If we can merge fragments, do it.
*/
// 现在往后查看是否有重叠的
// 从next开始
for (tmp = next; tmp != NULL; tmp = tfp)
{
// temp总是等于next
tfp = tmp->next;
// 如果next的 offset >= 当前分组的end，那就没有问题
if (tmp->offset >= end)
break; /* no overlaps at all */
// 否则警报
nids_params.syslog(NIDS_WARN_IP, NIDS_WARN_IP_OVERLAP, iph, 0);
// 记录当前的结束与下一个碎片的开始重叠多少
i = end - next->offset; /* overlap is 'i' bytes */
// 将next碎片的长度减少i
tmp->len -= i; /* so reduce size of */
// 将next碎片的开始增加i
tmp->offset += i; /* next fragment */
// 将指针也要向后移动i
tmp->ptr += i;
/*
If we get a frag size of <= 0, remove it and the packet that it
goes with. We never throw the new frag away, so the frag being
dumped has always been charged for.
*/
// 如果此时next碎片的长度小于0， 那么摘掉next节点
if (tmp->len <= 0)
{
if (tmp->prev != NULL)
tmp->prev->next = tmp->next;
else
qp->fragments = tmp->next;
if (tmp->next != NULL)
tmp->next->prev = tmp->prev;
// tfp原来就等于next->next，所以原来的next节点被摘掉了
next = tfp; /* We have killed the original "next" frame */
// 释放掉frag节点对应的内存
frag_kfree_skb(tmp->skb, FREE_READ);
// 释放掉frag节点
frag_kfree_s(tmp, sizeof(struct ipfrag));
}
}
/* Insert this fragment in the chain of fragments. */
tfp = NULL;
// offset是刚刚接收到的碎片偏移，按字节
// end 是刚刚接收到的碎片的结束字节
// skb 是传进来的一个内存空间，应该事先分配好
// prt 是指向这个碎片第一个数据的指针
tfp = ip_frag_create(offset, end, skb, ptr);
/*---------------------------------------------------
注意:
ip_frag_create函数只是创建一个ip_frag，还没有吧
它挂载到队列中
-----------------------------------------------------*/
/*
No memory to save the fragment - so throw the lot. If we failed
the frag_create we haven't charged the queue.
*/
if (!tfp)
{
nids_params.no_mem("ip_defrag");
kfree_skb(skb, FREE_READ);
return NULL;
}
/* From now on our buffer is charged to the queues. */
// 将刚刚创建的ip_frag挂载到队列中去了
tfp->prev = prev;
tfp->next = next;
if (prev != NULL)
prev->next = tfp;
else
qp->fragments = tfp;
if (next != NULL)
next->prev = tfp;
/*
OK, so we inserted this new fragment into the chain. Check if we
now have a full IP datagram which we can bump up to the IP
layer...
*/
// 检查是否完整
if (ip_done(qp))
{
skb2 = ip_glue(qp); /* glue together the fragments */
return (skb2);
}
// 如果没有完整，那么返回空，继续执行
return (NULL);
}
