Re: [PATCH v8 1/2] sched/deadline: Add support for SD_PREFER_SIBLING on find_later_rq()

[Byungchul Park <byungchul.park@lge.com>]

On Tue, Aug 22, 2017 at 02:53:25PM +0900, Byungchul Park wrote:
> On Mon, Aug 21, 2017 at 02:44:58PM +0100, Juri Lelli wrote:
> > Hi,
> > On 18/08/17 17:21, Byungchul Park wrote:
> > > It would be better to try to check other siblings first if
> > > SD_PREFER_SIBLING is flaged when pushing tasks - migration.
> > > 
> > > Signed-off-by: Byungchul Park <byungchul.park@lge.com>
> > 
> > Mmm, this looks like Peter's proposed patch, maybe add (at least) a
> > Suggested-by: him ?
> 
> Hi Juri,
> 
> Why not. I will add it from the next spin.
> 
> BTW, is it enough? I don't know the way I should do, whenever I got
> thankful suggestions. I really want to add them as a separate patch
> which can be stacked on my patches _if possible_. But in case that
> it's better to merge them into one like this, I don't know how.
> 
> I mean I will add 'Suggested-by' from now on - I learned what I should
> do (at least) in this case thanks to Juri, but I'm still not sure if
> it's enough.
> 
> Speaking of which, I have something to ask Peterz and Ingo for. I really
> want to interact with maintainers actively e.g. asking ways they prefer.
> But it takes too much long to get responses from them e.g. at most 2
> monthes in case rushing them. I should have decided and done what the
> best I think is, than asking.
> 
> It would be very appriciated if you pay more attention.
> 
> > > @@ -1376,8 +1399,7 @@ static int find_later_rq(struct task_struct *task)
> > >  				return this_cpu;
> > >  			}
> > >  
> > > -			best_cpu = cpumask_first_and(later_mask,
> > > -							sched_domain_span(sd));
> > > +			best_cpu = find_cpu(later_mask, sd, prefer);
> > >  			/*
> > >  			 * Last chance: if a cpu being in both later_mask
> > >  			 * and current sd span is valid, that becomes our
> > > @@ -1385,6 +1407,26 @@ static int find_later_rq(struct task_struct *task)
> > >  			 * already under consideration through later_mask.
> > >  			 */
> > 
> > It seems that the comment above should be updated as well.
> 
> How? Could you explain it more?

Let me try it by myself.. Please fix me at the next spin if needed.

Thank you,
Byungchul