拨开荷叶行,寻梦已然成。仙女莲花里,翩翩白鹭情。
IMG-LOGO
主页 文章列表 freeswitch APR-UTIL库执行绪池实作分析

freeswitch APR-UTIL库执行绪池实作分析

白鹭 - 2022-02-13 2147 0 0

 

 

 

概述

freeswitch的核心源代码是基于apr库开发的,在不同的系统上有很好的移植性,

APR库在之前的文章中已经介绍过了,APR-UTIL库是和APR并列的工具库,它们都是由APACHE开源出来的跨平台可移植库,不同点在于库中实作的功能界面有区别,

在应用的开发程序中,多执行绪并发是提高效率的常用方案,但是多执行绪管理并不好做,

在很多大型应用中,都会引入执行绪池的框架,执行绪池是一个执行绪集合,有统一的管理,当有一个新的任务下发,执行绪池管理会按照一定的策略将任务分配给空闲的执行绪,当任务积压较多时,执行绪池会创建新的执行绪来加快处理效率,

APR-UTIL库中就提供了一套执行绪池界面,

我对几个问题比较好奇,执行绪池如何管理?执行绪池什么情况要增加执行绪?什么情况会减少执行绪?执行绪池执行绪数目如何设定才有最优的效率?

下面我们对apr-util库的执行绪池实作做一个介绍,

 

环境

centos:CentOS  release 7.0 (Final)或以上版本

APR-UTIL:1.6.1

GCC:4.8.5

本来要使用freeswitch1.8.7中带的apr-util库源代码来梳理,但是很遗憾的是这个apr-util库版本是1.2.8,里面没有apr_thread_pool界面,,,所以从APR官网上下载了最新的1.6.1版本来做分析,

 

资料结构

apr执行绪池源档案:

apr-util-1.6.1\include\apr_thread_pool.h

apr-util-1.6.1\misc\apr_thread_pool.c

 

号码池资料结构定义在apr_thread_pool.c中

typedef struct apr_thread_pool_task

{

    APR_RING_ENTRY(apr_thread_pool_task) link;

    apr_thread_start_t func;

    void *param;

    void *owner;

    union

    {

        apr_byte_t priority;

        apr_time_t time;

    } dispatch;

} apr_thread_pool_task_t;

 

APR_RING_HEAD(apr_thread_pool_tasks, apr_thread_pool_task);

 

struct apr_thread_list_elt

{

    APR_RING_ENTRY(apr_thread_list_elt) link;

    apr_thread_t *thd;

    volatile void *current_owner;

    volatile enum { TH_RUN, TH_STOP, TH_PROBATION } state;

};

 

APR_RING_HEAD(apr_thread_list, apr_thread_list_elt);

 

struct apr_thread_pool

{

    apr_pool_t *pool;

    volatile apr_size_t thd_max;

    volatile apr_size_t idle_max;

    volatile apr_interval_time_t idle_wait;

    volatile apr_size_t thd_cnt;

    volatile apr_size_t idle_cnt;

    volatile apr_size_t task_cnt;

    volatile apr_size_t scheduled_task_cnt;

    volatile apr_size_t threshold;

    volatile apr_size_t tasks_run;

    volatile apr_size_t tasks_high;

    volatile apr_size_t thd_high;

    volatile apr_size_t thd_timed_out;

    struct apr_thread_pool_tasks *tasks;

    struct apr_thread_pool_tasks *scheduled_tasks;

    struct apr_thread_list *busy_thds;

    struct apr_thread_list *idle_thds;

    apr_thread_mutex_t *lock;

    apr_thread_cond_t *cond;

    volatile int terminated;

    struct apr_thread_pool_tasks *recycled_tasks;

    struct apr_thread_list *recycled_thds;

    apr_thread_pool_task_t *task_idx[TASK_PRIORITY_SEGS];

};

 

执行绪池存储器模型总图,执行绪池,任务队列,执行绪队列,

 

 

 

 

常用函式

常用函式界面

apr_thread_pool_create       //Create a thread pool

apr_thread_pool_destroy     //Destroy the thread pool and stop all the threads

apr_thread_pool_push  //Schedule a task to the bottom of the tasks of same priority.

apr_thread_pool_schedule   //Schedule a task to be run after a delay

apr_thread_pool_top    //Schedule a task to the top of the tasks of same priority.

apr_thread_pool_tasks_cancel     //Cancel tasks submitted by the owner. If there is any task from the owner that is currently running, the function will spin until the task finished.

apr_thread_pool_tasks_count      //Get the current number of tasks waiting in the queue

apr_thread_pool_scheduled_tasks_count   //Get the current number of scheduled tasks waiting in the queue

apr_thread_pool_threads_count  //Get the current number of threads

apr_thread_pool_busy_count      //Get the current number of busy threads

apr_thread_pool_idle_count //Get the current number of idle threads

apr_thread_pool_idle_max_set    //Access function for the maximum number of idle threads. Number of current idle threads will be reduced to the new limit.

apr_thread_pool_tasks_run_count      //Get number of tasks that have run

apr_thread_pool_tasks_high_count    //Get high water mark of the number of tasks waiting to run

apr_thread_pool_threads_high_count //Get high water mark of the number of threads

apr_thread_pool_threads_idle_timeout_count   //Get the number of idle threads that were destroyed after timing out

apr_thread_pool_idle_max_get    //Access function for the maximum number of idle threads

apr_thread_pool_thread_max_set       //Access function for the maximum number of threads.

apr_thread_pool_idle_wait_set     //Access function for the maximum wait time (in microseconds) of an idling thread that exceeds the maximum number of idling threads. A non-zero value allows for the reaping of idling threads to shrink over time.  Which helps reduce thrashing.

apr_thread_pool_idle_wait_get    //Access function for the maximum wait time (in microseconds) of an idling thread that exceeds the maximum number of idling threads

apr_thread_pool_thread_max_get      //Access function for the maximum number of threads

apr_thread_pool_threshold_set   //Access function for the threshold of tasks in queue to trigger a new thread.

apr_thread_pool_threshold_get   //Access function for the threshold of tasks in queue to trigger a new thread.

apr_thread_pool_task_owner_get       //Get owner of the task currently been executed by the thread.

 

apr_thread_pool_create创建

APU_DECLARE(apr_status_t) apr_thread_pool_create(apr_thread_pool_t ** me,

                                                 apr_size_t init_threads,

                                                 apr_size_t max_threads,

                                                 apr_pool_t * pool)

 

界面逻辑:

  1. 分配一块大小为apr_thread_pool_t的存储器tp,
  2. 在传入的存储器池pool中申请一个新的存储器池tp->pool,
  3. 初始化执行绪池资料,

    a)      执行绪池资料初始化,

b)      创建执行绪互斥锁me->lock,

c)      创建条件变量me->cond,

d)      在存储器池pool上分配一块大小为“apr_thread_pool_tasks“的存储器赋值给me->tasks,

e)      在存储器池pool上分配一块大小为“apr_thread_pool_tasks“的存储器赋值给me->scheduled_tasks,

f)       在存储器池pool上分配一块大小为“apr_thread_pool_tasks“的存储器赋值给me->recycled_tasks,

g)      在存储器池pool上分配一块大小为“apr_thread_list“的存储器赋值给me->busy_thds,

h)      在存储器池pool上分配一块大小为“apr_thread_list“的存储器赋值给me->idle_thds,

i)       在存储器池pool上分配一块大小为“apr_thread_list“的存储器赋值给me->recycled_thds,

j)       执行绪池资料初始化,

  1. 在存储器池tp->pool中注册清理回呼函式,
  2. 回圈创建初始作业执行绪,并加入执行绪池的管理,作业执行绪的逻辑见“thread_pool_func作业执行绪”,
  3. 回传创建结果,

 

执行绪池初始化成功后,存储器模型如图(作业执行绪启动未完成时)

 

 

 

thread_pool_func作业执行绪

static void *APR_THREAD_FUNC thread_pool_func(apr_thread_t * t, void *param)

 

界面逻辑:

  1. 加锁me->lock
  2. 判断me->recycled_thds链表为空?为空则创建新的apr_thread_list_elt节点elt,不为空则获取recycled_thds中首节点elt并从recycled_thds中移除该节点,
  3. 回圈处理,

a)      将elt节点加入me->busy_thds链表,

b)      获取一个新任务task,TODO

c)      回圈处理,解锁me->lock,呼叫任务回呼task->func,加锁me->lock,将task加入me->recycled_tasks链表,获取新任务task,执行绪状态置为TH_STOP时跳出回圈,获取任务为空跳出回圈,

d)      执行绪从busy到stop状态,将elt加入me->recycled_thds链表尾部,解锁me->lock,退出执行绪,

e)      执行绪从busy到idle状态,将elt节点从me->busy_thds链表中移除,将elt加入me->idle_thds链表尾部,

f)       检查是否有定时任务并获取任务执行等待时间,

g)      检查当前空闲执行绪数是否大于最大空闲数,获取空闲等待时间me->idle_wait,并设定当前执行绪状态为TH_PROBATION,下一轮回圈中进入stop处理流程,

h)      执行绪阻塞,等待条件变量me->cond的通知或超时,

  1. 执行绪数me->thd_cnt自减,
  2. 解锁me->lock,
  3. 退出执行绪,

 

执行绪池初始化成功后,存储器模型如图(作业执行绪启动完成时)

 

 

 

apr_thread_pool_push添加任务

APU_DECLARE(apr_status_t) apr_thread_pool_push(apr_thread_pool_t *me,

                                               apr_thread_start_t func,

                                               void *param,

                                               apr_byte_t priority,

                                               void *owner)

 

界面逻辑:

  1. 加锁me->lock,
  2. 检查me->recycled_tasks是否为空,为空则新建任务节点t,不为空则从me->recycled_tasks获取任务节点t,
  3. 任务节点t资料初始化,
  4. 计算任务优先级,根据优先级设定me->task_idx[seg]和me->tasks,
  5. 当前作业执行绪数为0时,或者空闲执行绪数为0并且当前执行绪数未达到最大并且当前任务数超过阈值等条件,动态创建新的作业执行绪,
  6. 对条件变量me->cond发通知,
  7. 解锁me->lock,

 

执行绪池添加任务后的存储器模型图,

 

 

 

 

apr_thread_pool_tasks_cancel取消任务

APU_DECLARE(apr_status_t) apr_thread_pool_tasks_cancel(apr_thread_pool_t *me,

                                                       void *owner)

 

界面逻辑:

  1. 加锁me->lock,
  2. 如果当前任务数大于0,则清空owner的所有任务,
  3. 如果定时任务数大于0,则清空owner的所有定时任务,
  4. 解锁me->lock,
  5. 等待执行绪退出,

 

总结

APR执行绪池的几个关注点,

执行绪从busy到stop状态时,没有将elt节点从me->busy_thds链表中洗掉?

APR执行绪池没有内置的管理执行绪,根据当前执行绪数和任务数进行动态的调整,而是通过任务阈值、空闲执行绪最大值和超时时间等设定来控制执行绪数的增减,这一点和我开始想的不一样,

 

 


 

空空如常

求真得真

 

标签:

0 评论

发表评论

您的电子邮件地址不会被公开。 必填的字段已做标记 *