RE: [PATCH] cpufreq: Add Kryo CPU scaling driver

[ilialin]

> -----Original Message-----
> From: Sudeep Holla <sudeep.holla@xxxxxxx>
> Sent: Monday, May 21, 2018 16:05
> To: ilialin@xxxxxxxxxxxxxx; mturquette@xxxxxxxxxxxx; sboyd@xxxxxxxxxx;
> robh@xxxxxxxxxx; mark.rutland@xxxxxxx; viresh.kumar@xxxxxxxxxx;
> nm@xxxxxx; lgirdwood@xxxxxxxxx; broonie@xxxxxxxxxx;
> andy.gross@xxxxxxxxxx; david.brown@xxxxxxxxxx; catalin.marinas@xxxxxxx;
> will.deacon@xxxxxxx; rjw@xxxxxxxxxxxxx; linux-clk@xxxxxxxxxxxxxxx
> Cc: Sudeep Holla <sudeep.holla@xxxxxxx>; devicetree@xxxxxxxxxxxxxxx;
> linux-kernel@xxxxxxxxxxxxxxx; linux-pm@xxxxxxxxxxxxxxx; linux-arm-
> msm@xxxxxxxxxxxxxxx; linux-soc@xxxxxxxxxxxxxxx; linux-arm-
> kernel@xxxxxxxxxxxxxxxxxxx; rnayak@xxxxxxxxxxxxxx;
> amit.kucheria@xxxxxxxxxx; nicolas.dechesne@xxxxxxxxxx;
> celster@xxxxxxxxxxxxxx; tfinkel@xxxxxxxxxxxxxx
> Subject: Re: [PATCH] cpufreq: Add Kryo CPU scaling driver
> 
> 
> 
> On 21/05/18 13:57, ilialin@xxxxxxxxxxxxxx wrote:
> >
> [...]
> 
> >>> +#include <linux/cpu.h>
> >>> +#include <linux/err.h>
> >>> +#include <linux/init.h>
> >>> +#include <linux/kernel.h>
> >>> +#include <linux/module.h>
> >>> +#include <linux/nvmem-consumer.h>
> >>> +#include <linux/of.h>
> >>> +#include <linux/platform_device.h>
> >>> +#include <linux/pm_opp.h>
> >>> +#include <linux/slab.h>
> >>> +#include <linux/soc/qcom/smem.h>
> >>> +
> >>> +#define MSM_ID_SMEM	137
> >>> +#define SILVER_LEAD	0
> >>> +#define GOLD_LEAD	2
> >>> +
> >>
> >> So I gather form other emails, that these are physical cpu number(not
> >> even unique identifier like MPIDR). Will this work on parts or
> >> platforms that need to boot in GOLD LEAD cpus.
> >
> > The driver is for Kryo CPU, which (and AFAIK all multicore MSMs)
> > always boots on the CPU0.
> 
> 
> That may be true and I am not that bothered about it. But assuming physical
> ordering from the logical cpu number is *incorrect* and will break if kernel
> decides to change the allocation algorithm. Kernel provides no guarantee on
> that, so you need to depend on some physical ID or may be DT to achieve
> what your want. But the current code as it stands is wrong.

Got your point. In fact CPUs are numbered 0-3 and ordered into 2 clusters in the DT:

cpus {
	#address-cells = <2>;
	#size-cells = <0>;

	CPU0: cpu@0 {
		...
		reg = <0x0 0x0>;
		...
	};

	CPU1: cpu@1 {
		...
		reg = <0x0 0x1>;
		...
	};

	CPU2: cpu@100 {
		...
		reg = <0x0 0x100>;
		...
	};

	CPU3: cpu@101 {
		...
		reg = <0x0 0x101>;
		...
	};

	cpu-map {
		cluster0 {
			core0 {
				cpu = <&CPU0>;
			};

			core1 {
				cpu = <&CPU1>;
			};
		};

		cluster1 {
			core0 {
				cpu = <&CPU2>;
			};

			core1 {
				cpu = <&CPU3>;
			};
		};
	};
};

As far, as I understand, they are probed in the same order. However, to be certain that the physical CPU is the one I intend to configure, I have to fetch the device structure pointer for the cpu-map -> clusterX -> core0 -> cpu path. Could you suggest a kernel API to do that?



> 
> --
> Regards,
> Sudeep