我得到了'boost :: fibers :: lock_error'并且不知道为什么

Question

i run this code

class ttt {
public:
    ~ttt() {
        LOG(INFO);
        flush();
    }   

    bool flush() {
        //std::lock_guard<boost::fibers::mutex> lock(_mutex);
        LOG(INFO);
        _mutex.lock();
        LOG(INFO);
        auto ret = flush_nonlock();
        LOG(INFO);
        _mutex.unlock(); 
        LOG(INFO);
        return ret;
    }   
private:

    bool flush_nonlock() {
        LOG(INFO);
        return std::rand()%2;
    }   
    boost::fibers::mutex _mutex;
};
int main() {
    static ttt t;
    std::cout << t.flush() << std::endl;
    return 0;
}

and i got

terminate called after throwing an instance of 'boost::fibers::lock_error'
  what():  boost fiber: a deadlock is detected: Resource deadlock avoided

the last log it print is before the _mutex.lock(). if t is not a static variable, it won't throw any error. if i remove t.flush() in main func, it won't throw any error. use std::lock_guard as i wrote in notes, the line next to it is not printed. i can't figure out why and whats the diff about the cases i tried.

i build the code use gcc 5.4.0, with -O0

Answer 1

static ttt t;

your static instance of ttt in main might be destructed after the internal data of boost.fiber have been destructed. Accesing context::active() in mutext::lock() from ~ttt() might return a null-pointer.

edit: boost.fiber uses internally a thread-local static holding the active fiber (in order to enable suspending from a deep call stack). because ttl is declared static, the compiler can destruct the ttl instance and the boost.fiber internall static in arbitrary order.