[PATCH] ARM: imx6q: work around faulty PMU irq routing

[shawn.guo@freescale.com (Shawn Guo)]

On Fri, Apr 25, 2014 at 11:13:25AM +0200, Lucas Stach wrote:
> Am Freitag, den 25.04.2014, 07:37 +0200 schrieb Dirk Behme:
> > On 24.04.2014 22:23, Lucas Stach wrote:
> > > The i.MX6 PMU has a design errata where the interrupts of all cores are
> > > wired together into a single irq line. To work around this we have to
> > > bounce the interrupt around all cores until we find the one where the PMU
> > > counter overflow has happened.
> > >
> > > This causes the perf measurements to be less accurate and we can't really
> > > handle the case where two cores fire a PMU irq at the same time. The
> > > implemented woraround makes perf at least somewhat useable on imx6 SoCs
> > > with more than one core.
> > >
> > > Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
> > > ---
> > >   arch/arm/mach-imx/mach-imx6q.c | 38 +++++++++++++++++++++++++++++++++++++-
> > >   1 file changed, 37 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/arch/arm/mach-imx/mach-imx6q.c b/arch/arm/mach-imx/mach-imx6q.c
> > > index e60456d85c9d..73976c484826 100644
> > > --- a/arch/arm/mach-imx/mach-imx6q.c
> > > +++ b/arch/arm/mach-imx/mach-imx6q.c
> > > @@ -16,6 +16,7 @@
> > >   #include <linux/delay.h>
> > >   #include <linux/export.h>
> > >   #include <linux/init.h>
> > > +#include "linux/interrupt.h"
> > >   #include <linux/io.h>
> > >   #include <linux/irq.h>
> > >   #include <linux/irqchip.h>
> > > @@ -33,6 +34,7 @@
> > >   #include <linux/mfd/syscon/imx6q-iomuxc-gpr.h>
> > >   #include <asm/mach/arch.h>
> > >   #include <asm/mach/map.h>
> > > +#include "asm/pmu.h"
> > >   #include <asm/system_misc.h>
> > >
> > >   #include "common.h"
> > > @@ -261,6 +263,39 @@ static void __init imx6q_axi_init(void)
> > >   	}
> > >   }
> > >
> > > +/*
> > > + * The i.MX6 PMU has a design errata where the interrupts of all cores are
> > > + * wired together into a single irq line. To work around this we have to
> > > + * bounce the interrupt around all cores until we find the one where the PMU
> > > + * counter overflow has happened.
> > > + */
> > > +static irqreturn_t imx6q_pmu_handler(int irq, void *dev, irq_handler_t handler)
> > > +{
> > > +	irqreturn_t ret = handler(irq, dev);
> > > +	int next;
> > > +
> > > +	if (ret == IRQ_NONE) {
> > > +		/*
> > > +		 * Kick the irq over to the next cpu, regardless of it's
> > > +		 * online status (it might have gone offline while we were busy
> > > +		 * bouncing the irq).
> > > +		 */
> > > +		next = (smp_processor_id() + 1) % num_present_cpus();
> > > +		irq_set_affinity(irq, cpumask_of(next));
> > > +	}
> > > +
> > > +	return ret;
> > > +}
> > > +
> > > +struct arm_pmu_platdata imx6q_pmu_platdata = {
> > > +	.handle_irq		= imx6q_pmu_handler,
> > > +};
> > > +
> > > +static struct of_dev_auxdata imx6q_auxdata_lookup[] __initdata = {
> > > +	OF_DEV_AUXDATA("arm,cortex-a9-pmu", 0, "arm-pmu", &imx6q_pmu_platdata),
> > > +	{}
> > > +};
> > > +
> > >   static void __init imx6q_init_machine(void)
> > >   {
> > >   	struct device *parent;
> > > @@ -276,7 +311,8 @@ static void __init imx6q_init_machine(void)
> > >
> > >   	imx6q_enet_phy_init();
> > >
> > > -	of_platform_populate(NULL, of_default_bus_match_table, NULL, parent);
> > > +	of_platform_populate(NULL, of_default_bus_match_table,
> > > +			imx6q_auxdata_lookup, parent);
> > >
> > >   	imx_anatop_init();
> > >   	cpu_is_imx6q() ?  imx6q_pm_init() : imx6dl_pm_init();
> > 
> > Do you have anything like a test case which shows that it works (at 
> > least better) on a !single core with this patch? Compared to a 
> > non-patched system?
> > 
> Without this patch, running perf top completely kills the system on
> i.MX6q, most likely because of the sheer number of spurious interrupts
> hitting the system from 4 cores. Even the spurious killer doesn't work
> sometimes, so perf is completely busted right now.
> 
> With this patch perf has to reduce the sample frequency in order to
> compensate the added irq latency, but at least we get some plausible
> numbers out. Though I won't take any blame if the amount of salt you
> have to apply while looking at those numbers is already a deadly
> dose. ;)
> 
> I don't yet have any numbers on how accurate the measurement is, but at
> least things didn't look completely off.

If it cannot provide correct/accurate data, I'd say let's not fake it
to, and just let it be completely broken there, so that people can be
aware of the brokenness, and not take inaccurate data as accurate one.

Shawn