环形缓冲区

#include <stdio.h>#include <stdlib.h>#include <string.h>#include <unistd.h>#include <pthread.h>#define BUFFSIZE 1024 * 1024#define min(x, y) ((x) < (y) ? (x) : (y))pthread_mutex_t lock = PTHREAD_MUTEX_INITIALIZER;struct cycle_buffer {        unsigned char *buf; /* store data */        unsigned int size; /* cycle buffer size */        unsigned int in; /* next write position in buffer */        unsigned int out; /* next read position in buffer */        pthread_mutex_t lock;};static struct cycle_buffer *fifo = NULL;static int init_cycle_buffer(void){        int size = BUFFSIZE, ret;        ret = size & (size - 1); /* size must be power of 2 */        if (ret)                goto err;        fifo = (struct cycle_buffer *) malloc(sizeof(struct cycle_buffer));        if (!fifo)                return -1;        memset(fifo, '/0', sizeof(struct cycle_buffer));        fifo->size = size;        fifo->in = fifo->out = 0;        pthread_mutex_init(&fifo->lock, NULL);        fifo->buf = (unsigned char *) malloc(size);        if (!fifo->buf)                goto malloc_err;        memset(fifo->buf, '/0', size);        return 0;malloc_err:        free(fifo);err:        return ret;}unsigned int fifo_get(unsigned char *buf, unsigned int len){        unsigned int l;        len = min(len, fifo->in - fifo->out);        l = min(len, fifo->size - (fifo->out & (fifo->size - 1)));        memcpy(buf, fifo->buf + (fifo->out & (fifo->size - 1)), l);        memcpy(buf + l, fifo->buf, len - l);        fifo->out += len;        return len;}unsigned int fifo_put(unsigned char *buf, unsigned int len){        unsigned int l;        len = min(len, fifo->size - fifo->in + fifo->out);        l = min(len, fifo->size - (fifo->in & (fifo->size - 1)));        memcpy(fifo->buf + (fifo->in & (fifo->size - 1)), buf, l);        memcpy(fifo->buf, buf + l, len - l);        fifo->in += len;        return len;}static void * thread_read(void *arg){        char buf[1024];        unsigned int n;        pthread_detach(pthread_self());        for (;;) {                memset(buf, '/0', sizeof(buf));                pthread_mutex_lock(&fifo->lock);                n = fifo_get(buf, sizeof(buf));                pthread_mutex_unlock(&fifo->lock);                write(STDOUT_FILENO, buf, n);        }        return NULL;}static void * thread_write(void *arg){        unsigned char buf[] = "hello world";        pthread_detach(pthread_self());        for (;;) {                pthread_mutex_lock(&fifo->lock);                fifo_put(buf, strlen(buf));                pthread_mutex_unlock(&fifo->lock);        }        return NULL;}int main(void){        int ret;        pthread_t wtid, rtid;        ret = init_cycle_buffer();        if (ret == -1)                goto err;        pthread_create(&wtid, NULL, thread_write, NULL);        pthread_create(&rtid, NULL, thread_read, NULL);        pthread_exit(NULL);err:        return ret;}

buffer指向存放数据的缓冲区，size是缓冲区的大小，in是写指针下标，out是读指针下标,在len和(fifo->size - fifo->in + fifo->out)之间取一个较小的值赋给len。注意，当(fifo->in == fifo->out+fifo->size)时，表示缓冲区已满，此时得到的较小值一定是0，后面实际写入的字节数也全为0。另一种边界情况是当len很大时（因为len是无符号的，负数对它来说也是一个很大的正数），这一句也能保证len取到一个较小的值，因为fifo->in总是大于等于fifo->out，所以后面的那个表达式的值不会超过fifo->size的大小把上一步决定的要写入的字节数len“切开”，这里又使用了一个技巧。注意：实际分配给fifo->buffer的字节数fifo->size，必须是2的幂，否则这里就会出错。既然fifo->size是2的幂，那么(fifo->size-1)也就是一个后面几位全为1的数，也就能保证(fifo->in & (fifo->size - 1))总为不超过(fifo->size - 1)的那一部分，和(fifo->in)% (fifo->size - 1)的效果一样。这样后面的代码就不难理解了，它先向fifo->in到缓冲区末端这一块写数据，如果还没写完，在从缓冲区头开始写入剩下的，从而实现了循环缓冲。最后，把写指针后移len个字节，并返回len。

从上面可以看出，fifo->in的值可以从0变化到超过fifo->size的数值，fifo->out也如此，但它们的差不会超过fifo->size