On Wed, Feb 19, 2014 at 09:34:21PM +0000, Rob Herring wrote:
> On Wed, Feb 19, 2014 at 9:41 AM, Lorenzo Pieralisi
[...]
> > diff --git a/drivers/cpuidle/cpuidle-calxeda.c b/drivers/cpuidle/cpuidle-calxeda.c
> > index 6e51114..5034f7a 100644
> > --- a/drivers/cpuidle/cpuidle-calxeda.c
> > +++ b/drivers/cpuidle/cpuidle-calxeda.c
> > @@ -41,9 +41,12 @@ static int calxeda_pwrdown_idle(struct cpuidle_device *dev,
> > struct cpuidle_driver *drv,
> > int index)
> > {
> > + int ret;
> > +
> > cpu_pm_enter();
> > - cpu_suspend(0, calxeda_idle_finish);
> > - cpu_pm_exit();
> > + ret = cpu_suspend(0, calxeda_idle_finish);
> > + if (!ret)
> > + cpu_pm_exit();
>
> It seems a little strange that the enter does not have to be balanced
> with an exit call. Couldn't the enter tear down things that need to be
> re-enabled?
That's why I mentioned it is not a proper fix, just wanted to check my
assumptions. There are multiple solutions to this problem, but the only
one which is proper IMHO is for KVM to either "reset registers" on
cpu_pm_enter (CPU_PM_ENTER in hyp_init_cpu_pm_notifier), or to avoid
executing the hyp init code if a proper registers settings is detected
when the cpu_pm_exit is triggered (CPU_PM_EXIT).
Is there a way for KVM to simply detect it has nothing to do (ie, detect
hyp mode is already initialized and it has nothing to do) ?
There is no way in the kernel to guarantee the processor has been reset,
I could infer that from return paths (ie cpu_suspend returning 0) but
that's still not correct.
Marc, Christoffer, please let me know what you think, thanks, sorry for
the noise.
Lorenzo
_______________________________________________
kvmarm mailing list
kvmarm@xxxxxxxxxxxxxxxxxxxxx
https://lists.cs.columbia.edu/cucslists/listinfo/kvmarm
