<table border="1" cellspacing="0" cellpadding="8">
    <tr>
        <th>Issue</th>
        <td>
            <a href=https://github.com/llvm/llvm-project/issues/148328>148328</a>
        </td>
    </tr>

    <tr>
        <th>Summary</th>
        <td>
            std::condition_variable and std::condition_variable_any
        </td>
    </tr>

    <tr>
      <th>Labels</th>
      <td>
            new issue
      </td>
    </tr>

    <tr>
      <th>Assignees</th>
      <td>
      </td>
    </tr>

    <tr>
      <th>Reporter</th>
      <td>
          johnsheu
      </td>
    </tr>
</table>

<pre>
    `wait_for()` and `wait_until()` for `std::condition_variable` and `std::condition_variable_any` take their wait duration or timeout parameters by const-reference.  This can lead to unexpected behavior if the referent changes while the `wait_for()` or `wait_until()` blocks.

For example:
```c++
  template <class _Clock, class _Duration>
  _LIBCPP_HIDE_FROM_ABI cv_status
 wait_until(unique_lock<mutex>& __lk, const chrono::time_point<_Clock, _Duration>& __t) {
    using namespace chrono;
    using __clock_tp_ns = time_point<_Clock, nanoseconds>;

    typename _Clock::time_point __now = _Clock::now();
    if (__t <= __now)
      return cv_status::timeout;

 __clock_tp_ns __t_ns = __clock_tp_ns(std::__safe_nanosecond_cast(__t.time_since_epoch()));

 __do_timed_wait(__lk, __t_ns);
    return _Clock::now() < __t ? cv_status::no_timeout : cv_status::timeout;
  }
```
The duration parameter `__t` is accessed by-reference, and so its value may be different before and after the wait.

For `std::condition_variable_any`, this is an actual data race if the timeout parameter is protected by the lock, as the implementation accesses the timeout outside of the lock.
</pre>
<img width="1" height="1" alt="" src="http://email.email.llvm.org/o/eJyMVU2P4zYM_TXKhdjAkePYOfiQSTboAC26KPYuyDIda9eWXIlKJv--kO3Jx3Y6KCAgCEg9vvco0tJ7fTKIJcteWHZYyECtdeUP2xrfYlhUtr6WbJNcpCbRWMd4wfiWbRKQpob3QDCku3uosS6GPNUs3bF0p6ypNWlrxFk6LasOHwA-yRLSXGMmyZ8I1KJ2EMtBHZyMiWAdkO7RBoJBOtkjofNQXUFZ4-mLwwYdGoVLgO-t9qCkgQ5lDWQhGHwbUBHWUGErz9o60E0sA_M9AtVKc0IPl1Z3IwP40IpJ7kdOVJ1VP_2SJTuW7I7WAb7JfugwKk52bJNMRzH-Ek-yAyDsh04SAkv3qpPeg9hHFMb3MP8_zPpZ-nW8In5_fdl_-yZ-ez18Fce__vxD7F5eQZ2FJ0nBx5wncsHovwOKETXd94HwLULxDQjRTYWif6BaZ42duhN9FoPVhli6vzN64jICEONbYPkkBiB4bU5gZI9-kApvmM9xIVREFDQI44GlB_iPekYa6zE-FR8rjjAzEl0HjHXe_fqFNghh7GXEfkww9jK360ZJN8B4IQTFFoz5YszaznEAhxSceXD4VssGupN6ViUEvYt7CjBe3GZACC8bFHeVQklPE5vlqMVro1DgYFU7857OQ9Haiphai9j08e7U1InAk9RZyIeGRPUwuXD8VaqZSsTBY-nuUyMAWH54fOws2X1v8T7Et8mNQxTfzyYB7UEqhd7H6bzeBznKiHvDW9Dk4Sy7gNDLK1QItW7msa2wsQ7HRNlE4Di50YyHQfxfmyeWo7g4Ih8DUlGQHdSSJLj4mOd98a8dFPMHZ2leL9cx6_0JSz_-1XEP9GhocmGW658AbSCvawTb3BCWi7pM6226lQssV3nGeZEXebZoyyQpVvl6lW0V4naTFskqW63lus7zVVbwPFvokic8S_IVT3iaZ9lyg-u6WFdN3fBVuuEFWyfYS90tu-7cL607LbT3AcvVukh5sehkhZ0fvxScG7zAGGWcxw-HK-OlL1U4ebZOOu3J32FIU4flJ35PTf28H4vgurIlGsZXxo-MH0-a2lAtle0ZP8Zq88-XwdkfqIjx48jRM36cRZxL_k8AAAD__9_3QaA">