1. Task node
typedef void (*cb_fun)(void *); //任务结构体 typedef struct task { void *argv; //任务函数的参数(任务执行结束前,要保证参数地址有效) cb_fun handler; //任务函数(返回值必须为0 非0值用作增加线程,和销毁线程池) struct task *next; //任务链指针 }zoey_task_t;
Handler is a function pointer, which is the actual task function, argv is the parameter of the function, and next points to the next task.
2. Task Queue
typedef struct task_queue { zoey_task_t *head; //队列头 zoey_task_t **tail; //队列尾 unsigned int maxtasknum; //最大任务限制 unsigned int curtasknum; //当前任务数 }zoey_task_queue_t;
Head is the head pointer of the task queue, tail is the tail pointer of the task queue, maxtasknum is the maximum number of tasks in the queue, and curtasknum is the current task of the queue. number.
3. Thread pool
typedef struct threadpool { pthread_mutex_t mutex; //互斥锁 pthread_cond_t cond; //条件锁 zoey_task_queue_t tasks;//任务队列 unsigned int threadnum; //线程数 unsigned int thread_stack_size; //线程堆栈大小 }zoey_threadpool_t;
mutex is a mutex lock and cond is a conditional lock. Mutex and cond jointly ensure the mutual exclusion of thread pool tasks to receive or add.
Tasks points to the task queue.
Threadnum is the number of threads in the thread pool
Thread_stack_size is the thread stack size
4. Startup configuration
//配置参数 typedef struct threadpool_conf { unsigned int threadnum; //线程数 unsigned int thread_stack_size;//线程堆栈大小 unsigned int maxtasknum;//最大任务限制 }zoey_threadpool_conf_t;
Startup configuration structure The body is some parameters when initializing the thread pool.
5. Initialize the thread pool
First check whether the parameters are legal, and then initialize mutex, cond, key (pthread_key_t). The key is used to read and write thread global variables. This global variable controls whether the thread exits.
Finally create the thread.
zoey_threadpool_t* zoey_threadpool_init(zoey_threadpool_conf_t *conf) { zoey_threadpool_t *pool = null; int error_flag_mutex = 0; int error_flag_cond = 0; pthread_attr_t attr; do{ if (z_conf_check(conf) == -1){ //检查参数是否合法 break; } pool = (zoey_threadpool_t *)malloc(sizeof(zoey_threadpool_t));//申请线程池句柄 if (pool == null){ break; } //初始化线程池基本参数 pool->threadnum = conf->threadnum; pool->thread_stack_size = conf->thread_stack_size; pool->tasks.maxtasknum = conf->maxtasknum; pool->tasks.curtasknum = 0; z_task_queue_init(&pool->tasks); if (z_thread_key_create() != 0){//创建一个pthread_key_t,用以访问线程全局变量。 free(pool); break; } if (z_thread_mutex_create(&pool->mutex) != 0){ //初始化互斥锁 z_thread_key_destroy(); free(pool); break; } if (z_thread_cond_create(&pool->cond) != 0){ //初始化条件锁 z_thread_key_destroy(); z_thread_mutex_destroy(&pool->mutex); free(pool); break; } if (z_threadpool_create(pool) != 0){ //创建线程池 z_thread_key_destroy(); z_thread_mutex_destroy(&pool->mutex); z_thread_cond_destroy(&pool->cond); free(pool); break; } return pool; }while(0); return null; }
6. Add a task
First apply for a task node, add the node to the task queue after instantiation, add the current task queue number and notify other processes of new tasks. Task. The entire process is locked.
int zoey_threadpool_add_task(zoey_threadpool_t *pool, cb_fun handler, void* argv) { zoey_task_t *task = null; //申请一个任务节点并赋值 task = (zoey_task_t *)malloc(sizeof(zoey_task_t)); if (task == null){ return -1; } task->handler = handler; task->argv = argv; task->next = null; if (pthread_mutex_lock(&pool->mutex) != 0){ //加锁 free(task); return -1; } do{ if (pool->tasks.curtasknum >= pool->tasks.maxtasknum){//判断工作队列中的任务数是否达到限制 break; } //将任务节点尾插到任务队列 *(pool->tasks.tail) = task; pool->tasks.tail = &task->next; pool->tasks.curtasknum++; //通知阻塞的线程 if (pthread_cond_signal(&pool->cond) != 0){ break; } //解锁 pthread_mutex_unlock(&pool->mutex); return 0; }while(0); pthread_mutex_unlock(&pool->mutex); free(task); return -1; }
7. Destroy the thread pool
Destroying the thread pool is actually adding tasks to the task queue, but the added task is to let the thread exit. The z_threadpool_exit_cb function will set the lock to 0 and then exit the thread. A lock of 0 means that this thread
has exited, and then exits the next thread. After exiting the thread, all resources are released.
void zoey_threadpool_destroy(zoey_threadpool_t *pool) { unsigned int n = 0; volatile unsigned int lock; //z_threadpool_exit_cb函数会使对应线程退出 for (; n < pool->threadnum; n++){ lock = 1; if (zoey_threadpool_add_task(pool, z_threadpool_exit_cb, &lock) != 0){ return; } while (lock){ usleep(1); } } z_thread_mutex_destroy(&pool->mutex); z_thread_cond_destroy(&pool->cond); z_thread_key_destroy(); free(pool); }
8. Add a thread
It’s very simple, just generate a thread and the number of threads. Lock.
int zoey_thread_add(zoey_threadpool_t *pool) { int ret = 0; if (pthread_mutex_lock(&pool->mutex) != 0){ return -1; } ret = z_thread_add(pool); pthread_mutex_unlock(&pool->mutex); return ret; }
9. Change the maximum task limit of the task queue
When num=0, set the number of threads to infinite.
void zoey_set_max_tasknum(zoey_threadpool_t *pool,unsigned int num) { if (pthread_mutex_lock(&pool->mutex) != 0){ return -1; } z_change_maxtask_num(pool, num); //改变最大任务限制 pthread_mutex_unlock(&pool->mutex); }
10.Usage examples
int main() { int array[10000] = {0}; int i = 0; zoey_threadpool_conf_t conf = {5,0,5}; //实例化启动参数 zoey_threadpool_t *pool = zoey_threadpool_init(&conf);//初始化线程池 if (pool == null){ return 0; } for (; i < 10000; i++){ array[i] = i; if (i == 80){ zoey_thread_add(pool); //增加线程 zoey_thread_add(pool); } if (i == 100){ zoey_set_max_tasknum(pool, 0); //改变最大任务数 0为不做上限 } while(1){ if (zoey_threadpool_add_task(pool, testfun, &array[i]) == 0){ break; } printf("error in i = %d\n",i); } } zoey_threadpool_destroy(pool); while(1){ sleep(5); } return 0; }
The above is the detailed content of What is the source code of nginx thread pool?. For more information, please follow other related articles on the PHP Chinese website!

NGINX is suitable for handling high concurrency and static content, while Apache is suitable for dynamic content and complex URL rewrites. 1.NGINX adopts an event-driven model, suitable for high concurrency. 2. Apache uses process or thread model, which is suitable for dynamic content. 3. NGINX configuration is simple, Apache configuration is complex but more flexible.

NGINX and Apache each have their own advantages, and the choice depends on the specific needs. 1.NGINX is suitable for high concurrency, with simple deployment, and configuration examples include virtual hosts and reverse proxy. 2. Apache is suitable for complex configurations and is equally simple to deploy. Configuration examples include virtual hosts and URL rewrites.

The purpose of NGINXUnit is to simplify the deployment and management of web applications. Its advantages include: 1) Supports multiple programming languages, such as Python, PHP, Go, Java and Node.js; 2) Provides dynamic configuration and automatic reloading functions; 3) manages application lifecycle through a unified API; 4) Adopt an asynchronous I/O model to support high concurrency and load balancing.

NGINX started in 2002 and was developed by IgorSysoev to solve the C10k problem. 1.NGINX is a high-performance web server, an event-driven asynchronous architecture, suitable for high concurrency. 2. Provide advanced functions such as reverse proxy, load balancing and caching to improve system performance and reliability. 3. Optimization techniques include adjusting the number of worker processes, enabling Gzip compression, using HTTP/2 and security configuration.

The main architecture difference between NGINX and Apache is that NGINX adopts event-driven, asynchronous non-blocking model, while Apache uses process or thread model. 1) NGINX efficiently handles high-concurrent connections through event loops and I/O multiplexing mechanisms, suitable for static content and reverse proxy. 2) Apache adopts a multi-process or multi-threaded model, which is highly stable but has high resource consumption, and is suitable for scenarios where rich module expansion is required.

NGINX is suitable for handling high concurrent and static content, while Apache is suitable for complex configurations and dynamic content. 1. NGINX efficiently handles concurrent connections, suitable for high-traffic scenarios, but requires additional configuration when processing dynamic content. 2. Apache provides rich modules and flexible configurations, which are suitable for complex needs, but have poor high concurrency performance.

NGINX and Apache each have their own advantages and disadvantages, and the choice should be based on specific needs. 1.NGINX is suitable for high concurrency scenarios because of its asynchronous non-blocking architecture. 2. Apache is suitable for low-concurrency scenarios that require complex configurations, because of its modular design.

NGINXUnit is an open source application server that supports multiple programming languages and provides functions such as dynamic configuration, zero downtime updates and built-in load balancing. 1. Dynamic configuration: You can modify the configuration without restarting. 2. Multilingual support: compatible with Python, Go, Java, PHP, etc. 3. Zero downtime update: Supports application updates that do not interrupt services. 4. Built-in load balancing: Requests can be distributed to multiple application instances.


Hot AI Tools

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Undress AI Tool
Undress images for free

Clothoff.io
AI clothes remover

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

SublimeText3 Chinese version
Chinese version, very easy to use

SAP NetWeaver Server Adapter for Eclipse
Integrate Eclipse with SAP NetWeaver application server.

SublimeText3 English version
Recommended: Win version, supports code prompts!

mPDF
mPDF is a PHP library that can generate PDF files from UTF-8 encoded HTML. The original author, Ian Back, wrote mPDF to output PDF files "on the fly" from his website and handle different languages. It is slower than original scripts like HTML2FPDF and produces larger files when using Unicode fonts, but supports CSS styles etc. and has a lot of enhancements. Supports almost all languages, including RTL (Arabic and Hebrew) and CJK (Chinese, Japanese and Korean). Supports nested block-level elements (such as P, DIV),

SecLists
SecLists is the ultimate security tester's companion. It is a collection of various types of lists that are frequently used during security assessments, all in one place. SecLists helps make security testing more efficient and productive by conveniently providing all the lists a security tester might need. List types include usernames, passwords, URLs, fuzzing payloads, sensitive data patterns, web shells, and more. The tester can simply pull this repository onto a new test machine and he will have access to every type of list he needs.
