gustafn
committed
9 months ago
scalability improvement: keep result of lc_time_fmt_compile in (lock-free) per-thread cache rather than in the util_memoize cache