TAILQ 队列之一二事

TAILQ队列是 FreeBSD内核中的一种队列数据结构，在一些著名的开源库中(如 DPDK, libevent)有广泛的应用。

TAILQ队列的定义

TAILQ队列有HEAD和ENTRY两种基本的数据结构

#define    TAILQ_HEAD(name, type)                        \
struct name {                                \
    struct type *tqh_first;    /* first element */            \
    struct type **tqh_last;    /* addr of last next element */        \
}

#define TAILQ_ENTRY(type)                                            \
struct {                                                             \
    struct type *tqe_next;  /* next element */                       \
    struct type **tqe_prev;/* addr of previous next element*/        \
}   

注意：数据结构中的filed都是type类型的指针(或者是二级指针)，这里的type是用户的队列元素类型，，将ENTRY结构内嵌到用户的QUEUE_ITEM结构中：

struct QUEUE_ITEM{  
    int value;  
    TAILQ_ENTRY(QUEUE_ITEM) entries;  
};  

TAILQ_HEAD(headname,QUEUE_ITEM) queue_head; 

这和Linux中list的组织方式不一样，后者是单纯地将struct list_head作为链表的一个挂接点，并没有用户的信息，具体差别可以看下图：

TAILQ队列的操作

TAILQ提供了多种操作队列的API,比如：

TAILQ_HEAD(name, type)
TAILQ_ENTRY(type)
TAILQ_EMPTY(head)
TAILQ_FIRST(head)
TAILQ_FOREACH(var, head, field)    
TAILQ_INIT(head)
TAILQ_INSERT_AFTER(head, listelm, elm, field)
TAILQ_INSERT_BEFORE(listelm, elm, field)
TAILQ_INSERT_TAIL(head, elm, field)
.....

这些接口的实现和更多的操作接口可以参考 FreeBSD queue

TAILQ队列中为什么tqh_prev和tqh_last要使用二级指针

要搞清楚这个问题，我们可以考虑如果不使用二级指针会怎么样？ 就像定义成下面这样。

#define    FAKE_TAILQ_HEAD(name, type)                        \
struct name {                                \
    struct type *tqh_first;    /* first element */            \
    struct type *tqh_last;    /* last element */        \
}

#define FAKE_TAILQ_ENTRY(type)                                            \
struct {                                                             \
    struct type *tqe_next;  /* next element */                       \
    struct type *tqe_prev;  /*   previous element*/        \
}   

对比一下TAILQ_HEAD和FAKE_TAILQ_HEAD (注意其中的红线和绿线的区别)

如果我们想要删除队列的任意一个元素，对FAKE_TAILQ，我们需要特殊处理该元素是第一个元素的情况(第一个元素的tqe_prev指针为空)，而TAILQ就没有这个烦恼！

TAILQ队列的遍历性能

Linux中的list只将struct list_head作为用户元素的挂接点，因此在正向遍历链表时，需要使用container_of这类接口才能获取用户的数据，而TAILQ由于tqe_next指针直接指向用户元素的类型，所以理论上，正向遍历TAILQ比list更快.但逆向遍历时,由于TAILQ的取用prev元素的操作比next麻烦的多，因此逆向遍历是比正向慢的：

#define    TAILQ_PREV(elm, headname, field)                \
    (*(((struct headname *)((elm)->field.tqe_prev))->tqh_last))

以下是用附件中的代码测试的结果：

遍历TAILQ:

TAILQ traversal time is 31955 us
TAILQ reverse traversal time is 38699 us

遍历list

list traversal time is 33062 us
list list traversal time is 35864 us

附录

测试代码 bsd.c

#include 
#include  
#include 

#define TAILQ_ENTRY(type)                                            \
struct {                                                             \
    struct type *tqe_next;  /* next element */                       \
    struct type **tqe_prev;/* addr of previous next element*/        \
}   

#define    TAILQ_HEAD(name, type)                        \
struct name {                                \
    struct type *tqh_first;    /* first element */            \
    struct type **tqh_last;    /* addr of last next element */        \
}

#define    TAILQ_FIRST(head)    ((head)->tqh_first)
#define    TAILQ_NEXT(elm, field) ((elm)->field.tqe_next)
#define    TAILQ_PREV(elm, headname, field)                \
    (*(((struct headname *)((elm)->field.tqe_prev))->tqh_last))
    
#define    TAILQ_LAST(head, headname)                    \
    (*(((struct headname *)((head)->tqh_last))->tqh_last))
    
    
#define    TAILQ_INIT(head) do {                        \
    TAILQ_FIRST((head)) = NULL;                    \
    (head)->tqh_last = &TAILQ_FIRST((head));            \
} while (0)

#define TAILQ_INSERT_TAIL(head, elm, field) do {            \
    TAILQ_NEXT((elm), field) = NULL;                \
    (elm)->field.tqe_prev = (head)->tqh_last;            \
    *(head)->tqh_last = (elm);                    \
    (head)->tqh_last = &TAILQ_NEXT((elm), field);            \
} while (0)

#define    TAILQ_INSERT_BEFORE(listelm, elm, field) do {            \
    (elm)->field.tqe_prev = (listelm)->field.tqe_prev;        \
    TAILQ_NEXT((elm), field) = (listelm);                \
    *(listelm)->field.tqe_prev = (elm);                \
    (listelm)->field.tqe_prev = &TAILQ_NEXT((elm), field);        \
} while (0)

#define    TAILQ_FOREACH(var, head, field)                    \
    for ((var) = TAILQ_FIRST((head));                \
        (var);                            \
        (var) = TAILQ_NEXT((var), field))

#define    TAILQ_FOREACH_REVERSE(var, head, headname, field)        \
    for ((var) = TAILQ_LAST((head), headname);            \
        (var);                            \
        (var) = TAILQ_PREV((var), headname, field))
        
struct QUEUE_ITEM{  
    int value;  
    TAILQ_ENTRY(QUEUE_ITEM) entries;  
};  
TAILQ_HEAD(headname,QUEUE_ITEM) queue_head;  

#define ITEM_NUM 5000000
#define TRAVERSAL 20

int main(int argc,char **argv){  
    struct QUEUE_ITEM *item;   
    long long totaltime = 0;
    struct timeval start,end;
    long long metric[TRAVERSAL];
    int i = 0;
    
    TAILQ_INIT(&queue_head);  
    for(i=1;ivalue=i;  
        TAILQ_INSERT_TAIL(&queue_head, item, entries);  
    }  
    
    for (i = 0; i < TRAVERSAL; i++)
    {
        gettimeofday(&start,NULL);
        TAILQ_FOREACH(item, &queue_head, entries)
        {
            item->value++;
        }   
        gettimeofday(&end,NULL);
        metric[i] = (end.tv_sec - start.tv_sec) * 1000000 + (end.tv_usec - start.tv_usec); // get the run time by microsecond
    }
   
    totaltime = 0;
    for (i=0;ivalue++;
        }   
        gettimeofday(&end,NULL);
        metric[i] = (end.tv_sec - start.tv_sec) * 1000000 + (end.tv_usec - start.tv_usec); // get the run time by microsecond
    }
    
    totaltime = 0;
    for (i=0;i

测试代码 list.c

#include 
#include 
#include     /* for offsetof */
#include 

#define container_of(ptr, type, member) ({            \
    const typeof( ((type *)0)->member ) *__mptr = (ptr);    \
    (type *)( (char *)__mptr - offsetof(type,member) );})


#define list_entry(ptr, type, member) \
    container_of(ptr, type, member)


#define list_first_entry(ptr, type, member) \
    list_entry((ptr)->next, type, member)

#define list_last_entry(ptr, type, member) \
    list_entry((ptr)->prev, type, member)

#define list_next_entry(pos, member) \
    list_entry((pos)->member.next, typeof(*(pos)), member)

#define list_prev_entry(pos, member) \
    list_entry((pos)->member.prev, typeof(*(pos)), member)
    
#define list_for_each_entry(pos, head, member)                \
    for (pos = list_first_entry(head, typeof(*pos), member);    \
         &pos->member != (head);                    \
         pos = list_next_entry(pos, member))

#define list_for_each_entry_reverse(pos, head, member)            \
    for (pos = list_last_entry(head, typeof(*pos), member);        \
         &pos->member != (head);                     \
         pos = list_prev_entry(pos, member))
         
#define LIST_HEAD_INIT(name) { &(name), &(name) }

#define LIST_HEAD(name) \
    struct list_head name = LIST_HEAD_INIT(name)

struct list_head {
    struct list_head *next, *prev;
};
static inline void INIT_LIST_HEAD(struct list_head *list)
{
    list->next = list;
    list->prev = list;
}

static inline void __list_add(struct list_head *new,
                  struct list_head *prev,
                  struct list_head *next)
{
    next->prev = new;
    new->next = next;
    new->prev = prev;
    prev->next = new;
}

static inline void list_add(struct list_head *new, struct list_head *head)
{
    __list_add(new, head, head->next);
}

struct QUEUE_ITEM{
    int value;
    struct list_head node;
};

LIST_HEAD(queue_head);

#define ITEM_NUM 5000000
#define TRAVERSAL 20

int main()
{
    int i = 0;
    struct QUEUE_ITEM *item;
    long long totaltime = 0;
    struct timeval start,end;
    long long metric[TRAVERSAL];
    
    for(i=1;ivalue = i;
        INIT_LIST_HEAD(&item->node);
        list_add(&item->node, &queue_head);
    }
    
    for (i = 0; i < TRAVERSAL; i++)
    {
        gettimeofday(&start,NULL);
        list_for_each_entry_reverse(item, &queue_head, node)
        {
            item->value++;
        }   

        gettimeofday(&end,NULL);
        metric[i] = (end.tv_sec - start.tv_sec) * 1000000 + (end.tv_usec - start.tv_usec); // get the run time by microsecond
    }
   
    totaltime = 0;
    for (i=0;ivalue++;
        }   

        gettimeofday(&end,NULL);
        metric[i] = (end.tv_sec - start.tv_sec) * 1000000 + (end.tv_usec - start.tv_usec); // get the run time by microsecond
    }
   
    totaltime = 0;
    for (i=0;i