Re: [PATCH 3/6] sched: Extract __schedule_loop()

[Peter Zijlstra]

On Tue, Aug 15, 2023 at 06:33:01PM -0400, Phil Auld wrote:
> Hi Peter,
> 
> On Tue, Aug 15, 2023 at 01:01:24PM +0200 Peter Zijlstra wrote:
> > From: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> > 
> > There are currently two implementations of this basic __schedule()
> > loop, and there is soon to be a third.
> > 
> > Signed-off-by: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
> > Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
> > Link: https://lkml.kernel.org/r/20230427111937.2745231-2-bigeasy@xxxxxxxxxxxxx
> > ---
> >  kernel/sched/core.c |   21 +++++++++++----------
> >  1 file changed, 11 insertions(+), 10 deletions(-)
> > 
> > --- a/kernel/sched/core.c
> > +++ b/kernel/sched/core.c
> > @@ -6787,16 +6787,21 @@ static void sched_update_worker(struct t
> >  	}
> >  }
> >  
> > -asmlinkage __visible void __sched schedule(void)
> > +static __always_inline void __schedule_loop(unsigned int sched_mode)
> 
> I think this needs __sched or it's the only thing that ever shows up
> in wchan. E.g.
> 
>   16995 0 bash     S __schedule_loop.constprop.0
>   17036 1 kworker/ I __schedule_loop.constprop.0
>   17151 1 kworker/ I __schedule_loop.constprop.0
>   17235 0 sleep    S __schedule_loop.constprop.0
>   17236 4 ps       R -

But but but __always_inline... ?!?