Re: about cci400 pmu

[Willy Wolff]

Many thanks, it seems to work.

Regards,
Willy

On Fri, Apr 12, 2019 at 05:12:59PM +0100, Robin Murphy wrote:
> On 12/04/2019 17:01, Willy Wolff wrote:
> > Hi,
> > 
> > I'm trying to add pmu support for CCI400 on an Odroid-XU3/4.
> > Following https://www.kernel.org/doc/Documentation/devicetree/bindings/arm/cci.txt,
> > and other dtsi, I end up with this patch:
> > 
> > diff --git a/arch/arm/boot/dts/exynos5420.dtsi b/arch/arm/boot/dts/exynos5420.dtsi
> > index f4e8c5823bc2..8432f01fb463 100644
> > --- a/arch/arm/boot/dts/exynos5420.dtsi
> > +++ b/arch/arm/boot/dts/exynos5420.dtsi
> > @@ -170,6 +170,16 @@
> >                                  interface-type = "ace";
> >                                  reg = <0x5000 0x1000>;
> >                          };
> > +
> > +                       pmu@9000 {
> > +                               compatible = "arm,cci-400-pmu,r0";
> > +                               reg = <0x9000 0x5000>;
> > +                               interrupts = <0 105 4>,
> > +                                            <0 101 4>,
> > +                                            <0 102 4>,
> > +                                            <0 103 4>,
> > +                                            <0 104 4>;
> > +                       };
> >                  };
> >                  clock: clock-controller@10010000 {
> > 		
> > Saddly, I got:
> > [    5.375913] ARM-CCI PMU 10d20000.cci:pmu@9000: invalid resource
> 
> That message looks to come from devm_ioremap_resource(), which implies that
> platform_get_resource() is returning NULL in cci_pmu_probe(). Looking at the
> DT, you have a thing at offset 0x9000 under a parent "ranges" which is only
> 0x6000 long, so I'd guess the OF platform layer probably gave up on creating
> the resources when it couldn't translate the address.
> 
> Robin.
> 
> > [    5.389595] ARM-CCI PMU: probe of 10d20000.cci:pmu@9000 failed with error -12
> > 
> > By pluging some printk here and there in arm-cci.c in init and probe functions,
> > all return 0. I assume that the driver initialisation is succesful.
> > 
> > I trace a bit more in really_probe function, however, I can't find the real
> > cause of this error.
> > 
> > $ dmesg | grep cci
> > [    0.070299] drivers/bus/arm-cci.c:569:cci_init()
> > [    0.071484] drivers/bus/arm-cci.c:580:cci_init() cci_init_status = 0
> > [    0.071610] drivers/bus/arm-cci.c:595:cci_probed() cci_init_status = 0
> > [    0.113575] drivers/bus/arm-cci.c:569:cci_init()
> > [    0.113684] drivers/bus/arm-cci.c:572:cci_init() cci_init_status = 0
> > [    0.770287] drivers/bus/arm-cci.c:94:cci_platform_init() 0
> > [    0.811655] drivers/base/dd.c:516:really_probe() ARM-CCI 10d20000.cci try probe drv = c1058908, drv->probe = c0593498, dev = e8b2f010
> > [    0.811788] drivers/bus/arm-cci.c:72:cci_platform_probe()
> > [    0.811866] drivers/bus/arm-cci.c:569:cci_init()
> > [    0.811937] drivers/bus/arm-cci.c:572:cci_init() cci_init_status = 0
> > [    0.812023] drivers/bus/arm-cci.c:595:cci_probed() cci_init_status = 0
> > [    0.814718] drivers/bus/arm-cci.c:79:cci_platform_probe() 0
> > [    6.738502] drivers/base/dd.c:516:really_probe() ARM-CCI PMU 10d20000.cci:pmu@9000 try probe drv = c10859f4, drv->probe = c0593498, dev = e8b2fc10
> > [    6.747909] ARM-CCI PMU 10d20000.cci:pmu@9000: invalid resource
> > [    6.753711] drivers/base/dd.c:522:really_probe() ARM-CCI PMU 10d20000.cci:pmu@9000 fail probe
> > [    6.762416] drivers/base/dd.c:561:really_probe() ARM-CCI PMU 10d20000.cci:pmu@9000 probe_failed
> > [    6.770940] ARM-CCI PMU: probe of 10d20000.cci:pmu@9000 failed with error -12
> > 
> > I don't know if the provided values for dtsi is correct either.
> > 
> > I would be very happy to have some help to make these pmu working.
> > Many thanks in advance.
> > 
> > Best Regards,
> > Willy
> > 
> > _______________________________________________
> > linux-arm-kernel mailing list
> > linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> >