Re: [gcv v3 06/35] scheduler: Replace __get_cpu_var uses

[Ingo Molnar <mingo@kernel.org>]

* Christoph Lameter <cl@linux.com> wrote:

> On Thu, 29 Aug 2013, Steven Rostedt wrote:
> 
> > On Thu, Aug 29, 2013 at 04:57:43PM +0000, Christoph Lameter wrote:
> > >
> > > We could add a ____this_cpu variant that would be used in the cases we do
> > > not want preemption checks? There should not be too many but it will
> > > mean a whole lot of new definitions in percpu.h.
> >
> > Let's get away from underscores as they are meaningless.
> >
> > A this_cpu_atomic() or other descriptive name would be much more
> > appropriate.
> 
> Its not really an atomic operation in the classic sense.
> 
> this_cpu_no_preempt_check_read ?
> 
> The problem that I have is also that a kernel with preemption is not 
> something that see anywhere these days. Looks more like an academic 
> exercise? Does this really matter? All the distro I see use 
> PREEMPT_VOLUNTARY. Performance degradation is significant if massive 
> amounts of checks and preempt disable/enable points are added to the 
> kernel.
> 
> Do we agree that it is necessary and useful to add another variant of 
> this_cpu ops for this? The concern of having too many variants is no 
> longer there? Adding another variant is not that difficult just code 
> intensive.

Just stop the lame excuses and fix it already. This has come up in the 
past and you know it: you were told to fix the this_cpu debug checks by 
Linus as well, yet you didn't ... Don't send crap you know is broken.

Thanks,

	Ingo