Re: [PATCH 6/7] cputime: Introduce vtime accounting check for readers

[Frederic Weisbecker <fweisbec@gmail.com>]

On Mon, Nov 23, 2015 at 03:35:16PM +0100, Peter Zijlstra wrote:
> On Mon, Nov 23, 2015 at 03:19:19PM +0100, Frederic Weisbecker wrote:
> > On Thu, Nov 19, 2015 at 09:28:28PM +0100, Peter Zijlstra wrote:
> > > On Thu, Nov 19, 2015 at 04:47:33PM +0100, Frederic Weisbecker wrote:
> > > > +++ b/include/linux/vtime.h
> > > > @@ -17,9 +17,20 @@ static inline bool vtime_accounting_cpu_enabled(void) { return true; }
> > > >  #endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
> > > >  
> > > >  #ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
> > > > +/*
> > > > + * Checks if vtime is enabled on some CPU. Cputime readers want to be careful
> > > > + * in that case and compute the tickless cputime.
> > > > + * For now vtime state is tied to context tracking. We might want to decouple
> > > > + * those later if necessary.
> > > > + */
> > > > +static inline bool vtime_accounting_enabled(void)
> > > > +{
> > > > +	return context_tracking_is_enabled();
> > > > +}
> > > 
> > > Should this not also include a definition of this function for
> > > CONFIG_VIRT_CPU_ACCOUNTING_NATIVE and CONFIG_VIRT_CPU_ACCOUNTING ?
> > 
> > I could but I haven't found any user of it yet for others than
> > CONFIG_VIRT_CPU_ACCOUNTING_GEN. The NATIVE version, when enabled, runs on
> > all CPUs anyway.
> 
> Aah, I see, task_gtime() etc.. have a different definition for !GEN.
> tricky.

Right, this whole CONFIG_VIRT_CPU_ACCOUNTING stuff is a bit messy. I tried
to consolidate as much code I could between NATIVE and GEN but the result
is hard to parse. I'll see if I can clean a few things up there.

To begin with, VIRT_CPU_ACCOUNTING is a horrible misnomer, as vtime.
VIRT suggests we are dealing with virtualization while it's absolutely
not the case. Perhaps something like BOUNDARY_CPU_ACCOUNTING would parse
better. Or TICKLESS_CPU_ACCOUNTING.