Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1610382

Re: [PATCH 0/8] sched/deadline: Return the best satisfying affinity and dl in cpudl_find

Show all headers | View raw

On Tue, Mar 28, 2017 at 08:11:53AM +0100, Juri Lelli wrote:
> > > > For example:
> > > > 
> > > >    cpu 0 is running a task (dl: 10).
> > > >    cpu 1 is running a task (dl: 9).
> > > >    cpu 2 is running a task (dl: 8).
> > > >    cpu 3 is running a task (dl: 2).
> > > > 
> > > >    where cpu 3 want to push a task (affinity is 1 2 3 and dl is 1).
> > > 
> > > Hummm, but this should only happen if you disable admission control,
> > > right? Otherwise task's affinity can't be smaller that 0-3.
> > 
> > Hi Juri,
> > 
> > Can I ask you what is addmission control? Do you mean affinity setting?
> 
> sched_setattr() for DEADLINE tasks peforms a set of checks before
> admitting the task to the system. Please have a look at Documentation/
> scheduler/sched-deadline.txt::Section5 for what concerns affinity.

I see.

> > And do you mean s/disable/enable? Or am I misunderstanding?
> > 
> 
> No, I meant disable. The problem is that if you disable admission
> control the problem you are pointing out can happen, if admission
> control is enabled otherwise it can't, as we enforce that tasks have
> affinity equal to the root_domain span to which they belong. E.g, in
> your case the task will have affinity set to 0-3 (or it won't be able to
> enter the system), so that would make the problem go away.

I see.

> > > > In this case, the task should be migrated from cpu 3 to cpu 1, and
> > > > preempt cpu 1's task. However, current code just returns fail because
> > > > it fails at the affinity test with the maximum cpu, that is, cpu 0.
> > > > 
> > > > This patch set tries to find the best among ones satisfying task's
> > > > affinity and dl constraint until success or no more to see.
> > > > 
> > > 
> > > Anyway, do you have numbers showing how common is you fail scenario?
> > 
> > Actually, it very depends on how to set test environment. I can provide
> > you ones which generate many fails. IMHO, it's not a matter of frequency
> > but a matter of whether it works corrently. As you know, rt policy already
> > works corrently regarding this problem.
> > 
> 
> Right. But, my point is that if what you are highlighting turns out to
> be a pretty frequent situation, maybe we need to find a better data
> structure to speed up push operations or we will end up using the slow
> path most of the times, making the heap useless.

I totally agree with you. I will check it and let you know.

Thank you,
Byungchul

Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

[PATCH 0/8] sched/deadline: Return the best satisfying affinity and dl in cpudl_find Byungchul Park <byungchul.park@lge.com> - 2017-03-23 11:40 +0100
  [PATCH 4/8] sched/deadline: Factor out the selecting of suitable max cpu Byungchul Park <byungchul.park@lge.com> - 2017-03-23 11:40 +0100
  [PATCH 3/8] sched/deadline: Add cpudl_maximum_dl() for clean-up Byungchul Park <byungchul.park@lge.com> - 2017-03-23 11:40 +0100
  [PATCH 5/8] sched/deadline: Protect read of cpudl heap with a lock Byungchul Park <byungchul.park@lge.com> - 2017-03-23 11:40 +0100
  [PATCH 2/8] sched/deadline: Re-define parameters of cpudl heapify functions Byungchul Park <byungchul.park@lge.com> - 2017-03-23 11:40 +0100
  [PATCH 6/8] sched/deadline: Don't return meaningless cpu in cpudl_maximum_cpu() Byungchul Park <byungchul.park@lge.com> - 2017-03-23 11:40 +0100
  [PATCH 7/8] sched/deadline: Factor out the modifying of cpudl's heap tree Byungchul Park <byungchul.park@lge.com> - 2017-03-23 11:40 +0100
  Re: [PATCH 0/8] sched/deadline: Return the best satisfying affinity  and dl in cpudl_find Byungchul Park <byungchul.park@lge.com> - 2017-03-23 23:40 +0100
  Re: [PATCH 0/8] sched/deadline: Return the best satisfying affinity  and dl in cpudl_find Juri Lelli <juri.lelli@arm.com> - 2017-03-27 16:10 +0200
    Re: [PATCH 0/8] sched/deadline: Return the best satisfying affinity  and dl in cpudl_find Byungchul Park <byungchul.park@lge.com> - 2017-03-28 02:50 +0200
      Re: [PATCH 0/8] sched/deadline: Return the best satisfying affinity  and dl in cpudl_find Juri Lelli <juri.lelli@arm.com> - 2017-03-28 09:20 +0200
        Re: [PATCH 0/8] sched/deadline: Return the best satisfying affinity  and dl in cpudl_find Byungchul Park <byungchul.park@lge.com> - 2017-03-28 09:40 +0200

csiph-web