Re: [PATCH 2/3] [BUGFIX] x86/x86_64: fix CPU offlining triggered inactive device IRQ interrruption

["Pallipadi, Venkatesh" <venkatesh.pallipadi@intel.com>]

On Sat, 2009-04-11 at 04:01 -0700, Eric W. Biederman wrote:
> Yinghai Lu <yhlu.kernel@gmail.com> writes:
> 
> > commit 932775a4ab622e3c99bd59f14cc7d96722f79501
> > Author: venkatesh.pallipadi@intel.com <venkatesh.pallipadi@intel.com>
> > Date:   Fri Sep 5 18:02:15 2008 -0700
> >
> >     x86: HPET_MSI change IRQ affinity in process context when it is disabled
> >
> >     Change the IRQ affinity in the process context when the IRQ is disabled.
> >
> >     Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
> >     Signed-off-by: Shaohua Li <shaohua.li@intel.com>
> >     Signed-off-by: Ingo Molnar <mingo@elte.hu>
> >
> > diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
> > index ddc9568..ad2ce72 100644
> > --- a/kernel/irq/manage.c
> > +++ b/kernel/irq/manage.c
> > @@ -87,10 +87,11 @@ int irq_set_affinity(unsigned int irq, cpumask_t cpumask)
> >                 return -EINVAL;
> >
> >  #ifdef CONFIG_GENERIC_PENDING_IRQ
> > -       if (desc->status & IRQ_MOVE_PCNTXT) {
> > +       if (desc->status & IRQ_MOVE_PCNTXT || desc->status & IRQ_DISABLED) {
> >                 unsigned long flags;
> >
> >                 spin_lock_irqsave(&desc->lock, flags);
> > +               desc->affinity = cpumask;
> >                 desc->chip->set_affinity(irq, cpumask);
> >                 spin_unlock_irqrestore(&desc->lock, flags);
> >         } else
> 
> If the goal is moving MSIs, we should modify the msi code to be safe
> in process context and to set IRQ_MOVE_PCNTXT.
> 
> The only reason we migrate MSIs in interrupt context today is that there
> wasn't infrastructure for support migration both in interrupt context
> and outside of it.
> 

Yes. The idea here was to force the MSI migration to happen in process
context. One of the patches in the series did

	disable_irq(dev->irq);
	irq_set_affinity(dev->irq, cpumask_of(dev->cpu));
	enable_irq(dev->irq);

with the above patch adding irq/manage code check for interrupt disabled
and moving the interrupt in process context.

IIRC, there was no IRQ_MOVE_PCNTXT when we were developing this HPET
code and we ended up having this ugly hack. IRQ_MOVE_PCNTXT was there
when we eventually submitted the patch upstream. But, looks like I did a
blind rebasing instead of using IRQ_MOVE_PCNTXT in hpet MSI code. That
was my fault. Will send a patch to fix this ugliness.

Thanks,
Venki