https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60421
--- Comment #5 from Jaak Ristioja <jaak at ristioja dot ee> ---
(In reply to Jonathan Wakely from comment #4)
> Fixed for gcc5.
Looking at the diff <thread> of revision 221708, I fail to see how the
if (__rtime <= __rtime.zero())
return;
check in sleep_for() prevents the EINVAL in nanosleep(3p), as is the original
issue.
I mean aren't we dealing here with unsigned (decltype(_rtime)) to signed
(std::time_t) conversion?
