Messages in this thread | | | From | Thomas Gleixner <> | Subject | Re: [patch 2/8] hrtimer: Force clock_was_set() handling for the HIGHRES=n, NOHZ=y case | Date | Wed, 12 May 2021 18:40:49 +0200 |
| |
On Wed, May 12 2021 at 16:59, Peter Zijlstra wrote: > On Tue, Apr 27, 2021 at 10:25:39AM +0200, Thomas Gleixner wrote: >> void clock_was_set(void) >> { >> + if (!hrtimer_hres_active() && !tick_nohz_active) >> + goto out_timerfd; >> + >> /* Retrigger the CPU local events everywhere */ >> on_each_cpu(retrigger_next_event, NULL, 1); >> + >> +out_timerfd: >> timerfd_clock_was_set(); >> } > > Isn't that simpler when written like: > > if (hrtimer_hres_active() || tick_nohz_active()) > on_each_cpu(retrigger_next_event, NULL, 1); > > timerfd_clock_was_set(); > > ?
Yes, but look at the later patches. Then we'll introduce it there. :)
| |