From 0e4532a3d40b5329f562d4767791e0d99c968d06 Mon Sep 17 00:00:00 2001 From: Andy Ross Date: Mon, 24 Sep 2018 09:55:55 -0700 Subject: [PATCH] sys_clock.h: Remove variance of _TICK_ALIGN with TICKLESS_KERNEL Not sure why this was here. The point to this API (which is poorly explained) is to "round up" requested timeout values to an integer number of ticks in the future, so the timeouts don't expire too soon. There's no change of that requirement in tickless mode. While the "tick" unit will typicaly be a much smaller time (and thus much less likely to have this kind of aliasing bug), we STILL don't want early expiration. And as with everything else in tickless, changing this breaks no tests. So remove it as a needless TICKLESS dependency. Signed-off-by: Andy Ross --- include/sys_clock.h | 4 ---- 1 file changed, 4 deletions(-) diff --git a/include/sys_clock.h b/include/sys_clock.h index abcda46504d..2fe25014158 100644 --- a/include/sys_clock.h +++ b/include/sys_clock.h @@ -146,11 +146,7 @@ static inline s64_t __ticks_to_ms(s64_t ticks) } /* added tick needed to account for tick in progress */ -#ifdef CONFIG_TICKLESS_KERNEL -#define _TICK_ALIGN 0 -#else #define _TICK_ALIGN 1 -#endif /* SYS_CLOCK_HW_CYCLES_TO_NS64 converts CPU clock cycles to nanoseconds */ #define SYS_CLOCK_HW_CYCLES_TO_NS64(X) \