编译器对操作的重新排序

#include <stdio.h>
#include <pthread.h>

volatile int start[4];
volatile int done[4];

void *work(void *param)
{
	int id = (int)param;
	while(start[id] == 0) {}
	printf("Thread %i started\n", id);
	double total = 0;
	for(int i = 0; i < 100000000; i++) { total += i; }
	printf("Thread %i done\n", id);
	done[id] = 1;
}

int main()
{
	pthread_t thread[4];
	for(int i = 0; i < 4; i++)
	{
		pthread_create(&thread[i], 0, work, (void*)i);
		done[i] = 0;
	}
	for(int i = 0; i < 4; i++)
	{
		start[i] = 1;
	}
	for(int i = 0; i < 4; i++)
	{
		while(done[i] == 0) {}
	}
	for(int i = 0; i < 4; i++)
	{
		pthread_join(thread[i], 0);
	}
}

有优化及无优化的输出：