[libcxx-dev] Questions regarding implementation of std::this_thread::sleep_until

[Erik Rigtorp via libcxx-dev]

Sorry for the late reply Louis,

I don't have any interest, I need only efficient sleep on Linux, so it's
easy to work around. Mostly wanted to let people know about this gotcha.
Probably most timers are driven from a epoll/kqueue event loop so these C++
APIs are not used much outside demo/example/student code.

On Wed, Apr 8, 2020 at 8:16 AM Louis Dionne <ldionne at apple.com> wrote:

>
> > On Mar 5, 2020, at 15:30, Erik Rigtorp via libcxx-dev <
> libcxx-dev at lists.llvm.org> wrote:
> >
> > Hi!
> >
> > libcxx seems to have a suboptimal implementation of
> > std::this_thread::sleep_until for Linux.
> >
> > Currently it's implemented using pthread_cond_timedwait:
> > template <class _Clock, class _Duration>
> > void
> > sleep_until(const chrono::time_point<_Clock, _Duration>& __t)
> > {
> >  using namespace chrono;
> >  mutex __mut;
> >  condition_variable __cv;
> >  unique_lock<mutex> __lk(__mut);
> >  while (_Clock::now() < __t)
> >    __cv.wait_until(__lk, __t);
> > }
> > (
> https://github.com/llvm/llvm-project/blob/master/libcxx/include/thread#L391
> )
> >
> > This leads to unnecessary stores to the stack and useless atomic
> > operations that can potential cause memory bus traffic. On Linux this
> > could be implemented using the clock_nanosleep() function with the
> > TIMER_ABSTIME flag.
> >
> > For std::chrono::steady_clock there is a specialization:
> > template <class _Duration>
> > inline _LIBCPP_INLINE_VISIBILITY
> > void
> > sleep_until(const chrono::time_point<chrono::steady_clock, _Duration>&
> __t)
> > {
> >  using namespace chrono;
> >  sleep_for(__t - steady_clock::now());
> > }
> > (
> https://github.com/llvm/llvm-project/blob/master/libcxx/include/thread#L404
> )
> >
> > This has a race condition where the thread could be preempted after
> > calling now() but before sleeping. The race window is extremely small,
> > but will lead to increased jitter in actual deadline. If the thread is
> > using something like linux SCHED_IDLE or competing with threads using
> > SCHED_FIFO/SCHED_RR the likelihood of hitting this race could be high.
> > Using clock_nanosleep() would also help here.
> >
> > Linux and most BSDs support the clock_nanosleep() function, it makes
> > sense to add a feature detection for it and use it if supported.
>
>
> If you still have interest in fixing that and want to submit a patch, I'd
> be happy to help review it.
>
> Louis
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/libcxx-dev/attachments/20200615/fdee61d8/attachment.html>