[PATCH 1/1] irqchip: irq-gic: forward SGI to itself for cortex-a7 single core

[Marc Zyngier]

On Tue, 9 Aug 2016 11:46:13 +0800
Peter Chen <hzpeterchen at gmail.com> wrote:

> On Mon, Aug 08, 2016 at 02:59:16PM +0100, Marc Zyngier wrote:
> > On Mon, 8 Aug 2016 14:48:42 +0100
> > Mark Rutland <mark.rutland at arm.com> wrote:
> >   
> > > On Mon, Aug 08, 2016 at 09:28:47PM +0800, Peter Chen wrote:  
> > > > On Mon, Aug 08, 2016 at 02:07:54PM +0100, Mark Rutland wrote:    
> > > > > I see that for arm64 we have:
> > > > > 
> > > > > static inline bool arch_irq_work_has_interrupt(void)
> > > > > {
> > > > > 	return !!__smp_cross_call;
> > > > > }
> > > > > 
> > > > > Could we do similarly for ARM, and ony register gic_raise_softirq if
> > > > > we have non-zero SGI targets?
> > > > > 
> > > > > If I've understood correctly, that would make things behave as they do
> > > > > for UP on you system.    
> > > 
> > > [...]
> > >   
> > > > > If self-IPI is necessary, then this would be up to the GIC code to
> > > > > solve.
> > > > > 
> > > > > For that case, it would be nicer if we could detect whether this was
> > > > > necessary based on the GIC registers alone. That way we handle the
> > > > > various ways this can be integrated, aren't totally relient on the DT,
> > > > > work in VMs, etc.    
> > > > 
> > > > How we can detect IPI capabilities based on GIC register?    
> > > 
> > > Check the mask associated with SGIs, as we do for gic_get_cpumask(). If
> > > this is zero, we have a non-multiprocessor GIC (or one that's otherwise
> > > broken), and can't do SGI in the usual way.
> > > 
> > > However, it only makes sense to do this if self-IPI is truly a
> > > necessity. Given there are other interrupt controllers that can't do
> > > self-IPI, avoiding self-IPI in general would be a better strategy,
> > > avoiding churn in each and every driver...  
> > 
> > Indeed. And I won't take such a patch until all other avenues have been
> > explored, including fixing core code if required...
> >   
> 
> Ok, it seems both you and Mark agree with disable IPI for GIC who has only
> self-IPI capability (GICD_ITARGETSR0 to GICD_ITARGETSR7 are all
> zero), right?

Not necessarily. This can be seen a latency improvement, compared to
the timer method which should be the fallback.

> 
> But even we do that, we still have problem that the callers for
> smp_cross_call don't know well if the platform has IPI capability. Eg,
> IRQ work considers the SMP system has IPI capability, but it is not a
> must in this case (Cortex-A7 MPcore version, but cpu number is one).
> It will cause NULL pointer dereference problem as __smp_cross_call is
> NULL, and we need to make below change to let it work:
> 
> diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c
> index 937c892..276bd94 100644
> --- a/arch/arm/kernel/smp.c
> +++ b/arch/arm/kernel/smp.c
> @@ -487,7 +487,8 @@ static const char *ipi_types[NR_IPI] __tracepoint_string = {
>  static void smp_cross_call(const struct cpumask *target, unsigned int ipinr)
>  {
>  	trace_ipi_raise_rcuidle(target, ipi_types[ipinr]);
> -	__smp_cross_call(target, ipinr);
> +	if (__smp_cross_call)
> +		__smp_cross_call(target, ipinr);
>  }

I came up with a slightly different approach, which is to have
arch_irq_work_has_interrupt() to check for an IPI-capable system:

From 9be5fc16f10e4d32a8ad3d70db50d2dfb96f70a1 Mon Sep 17 00:00:00 2001
From: Marc Zyngier <marc.zyngier at arm.com>
Date: Tue, 9 Aug 2016 06:04:12 +0100
Subject: [PATCH] ARM: irq_work: Do not attempt to IPI on non IPI-capable HW

Not all of the ARM HW is IPI capable (i.e. most of the non-SMP
systems). Unfortunately, some systems do advertise being SMP
capable, even if they have a single core and do not define
a cross call method. In this case, irq_work_queue dies
as arch_irq_work_has_interrupt() fails to detect this
particular case.

Let's redefine arch_irq_work_has_interrupt() to actually check
if we're IPI capable instead of simply being SMP. This sidesteps
the issue entierely.

Signed-off-by: Marc Zyngier <marc.zyngier at arm.com>
---
 arch/arm/include/asm/irq_work.h | 2 +-
 arch/arm/include/asm/smp_plat.h | 2 ++
 arch/arm/kernel/smp.c           | 2 +-
 3 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/arch/arm/include/asm/irq_work.h b/arch/arm/include/asm/irq_work.h
index 712d03e..025420b 100644
--- a/arch/arm/include/asm/irq_work.h
+++ b/arch/arm/include/asm/irq_work.h
@@ -5,7 +5,7 @@
 
 static inline bool arch_irq_work_has_interrupt(void)
 {
-	return is_smp();
+	return !!__smp_cross_call;
 }
 
 #endif /* _ASM_ARM_IRQ_WORK_H */
diff --git a/arch/arm/include/asm/smp_plat.h b/arch/arm/include/asm/smp_plat.h
index f908071..ffee073 100644
--- a/arch/arm/include/asm/smp_plat.h
+++ b/arch/arm/include/asm/smp_plat.h
@@ -26,6 +26,8 @@ static inline bool is_smp(void)
 #endif
 }
 
+extern void (*__smp_cross_call)(const struct cpumask *, unsigned int);
+
 /**
  * smp_cpuid_part() - return part id for a given cpu
  * @cpu:	logical cpu id.
diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c
index 8615216..f9d771f 100644
--- a/arch/arm/kernel/smp.c
+++ b/arch/arm/kernel/smp.c
@@ -465,7 +465,7 @@ void __init smp_prepare_cpus(unsigned int max_cpus)
 	}
 }
 
-static void (*__smp_cross_call)(const struct cpumask *, unsigned int);
+void (*__smp_cross_call)(const struct cpumask *, unsigned int);
 
 void __init set_smp_cross_call(void (*fn)(const struct cpumask *, unsigned int))
 {
-- 
2.8.1

Does it work for you? We could then add self-IPI as a further
optimization.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny.