Re: [PATCH] sched/core: initialize resched_latency in sched_tick()

[Peter Zijlstra]

On Wed, Mar 18, 2026 at 10:48:26AM -0400, Steven Rostedt wrote:
> On Wed, 18 Mar 2026 12:00:58 +0530
> Shrikanth Hegde <sshegde@xxxxxxxxxxxxx> wrote:
> 
> > Whats the relevance of resched_latency warnings in the age of preempt
> > kernels being the default for majority of the archs.
> > 
> > In case of full/lazy preemption, at the end of sched_tick itself we would call
> > schedule if need_resched is set provided it was preemptible.
> > 
> > It kind of makes sense for non preemptive kernels. But we have preemptoff and
> > preemptirqoff tracers which one could use. (although may need a kernel build)
> > 
> > What bugs is it catching these days?
> 
> Good point. I've never used it. I guess the question is, does anyone?

Google added that, Ben are you guys still needing that?