Re: [RFC][PATCH 5/5] sched/core: Add debug code to catch missing update_rq_clock()

[Matt Fleming]

On Wed, 18 May, at 03:01:27AM, Yuyang Du wrote:
> On Tue, May 17, 2016 at 01:24:15PM +0100, Matt Fleming wrote:
> > So, if the code looks like the following, either now or in the future,
> > 
> > static void __schedule(bool preempt)
> > {
> > 	...
> > 	/* Clear RQCF_ACT_SKIP */
> > 	rq->clock_update_flags = 0;
> > 	...
> > 	delta = rq_clock();
> > }
>  
> Sigh, you even said "Clear RQCF_ACT_SKIP", but you not only clear it,
> you clear everything.

That was sloppy on my part but intentional because that's what the
code looks like in tip/sched/core today.

It was purely meant to demonstrate that setting RQCF_UPDATE just
because RQCF_ACT_SKIP is set does not make sense. You can replace the
clearing line with the correct bit masking operation.

But I get it, the pseudo-code was confusing. I'll send out a v2.

> And if you clear the RQCF_UPDATE also (maybe you shouldn't, but
> actually it does not matter), of course you will get a warning...

Right, I wouldn't actually clear RQCF_UPDATE in v2 of this patch.

> In addition, it looks like multiple skips are possible, so:
 
I'm not sure what you mean, could you elaborate?

> update_rq_clock() {
>         rq->clock_update_flags |= RQCF_UPDATE;
> 
>         ...
> }
> 
> instead of clearing the skip flag there.

Huh? RQCF_*_SKIP are not cleared in update_rq_clock().