absl::MutexLock locker(&mutex_);保证各个线程按序排队领任务,领完任务后,就并行计算各自的任务了(Execute(task.get());)
线程中执行的函数好像是进行了拷贝,所以线程之间调用同一个函数,看似是应该排队,但实际是相互不影响的。
void ThreadPool::DoWork() {
#ifdef __linux__
// This changes the per-thread nice level of the current thread on Linux. We
// do this so that the background work done by the thread pool is not taking
//