Re: [PATCH 2/4] perf: add OMAP support for the new power events

[Thomas Renninger <trenn@suse.de>]

On Monday 04 October 2010 05:20:59 pm Jean Pihet wrote:
> The patch adds the new power management trace points for
> the OMAP architecture.
>
> The trace points are for cpuidle, cpufreq (DVFS), the clocks
> changes (enable, disable, set_rate) and the power domains
> transitions.
...
> diff --git a/arch/arm/plat-omap/cpu-omap.c b/arch/arm/plat-omap/cpu-omap.c
> index 6d3d333..19d5a6c 100644
> --- a/arch/arm/plat-omap/cpu-omap.c
> +++ b/arch/arm/plat-omap/cpu-omap.c
> @@ -21,6 +21,7 @@
>  #include <linux/err.h>
>  #include <linux/clk.h>
>  #include <linux/io.h>
> +#include <trace/events/power.h>
>
>  #include <mach/hardware.h>
>  #include <plat/clock.h>
> @@ -95,6 +96,7 @@ static int omap_target(struct cpufreq_policy *policy,
>  	printk(KERN_DEBUG "cpufreq-omap: transition: %u --> %u\n",
>  	       freqs.old, freqs.new);
>  #endif
> +	trace_processor_frequency(freqs.new, freqs.cpu);
>  	ret = clk_set_rate(mpu_clk, freqs.new * 1000);
>  	cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);

This one is not needed, frequency power events are all fired in
drivers/cpufreq/cpufreq.c
for cpufreq drivers:
static struct cpufreq_driver omap_driver = {
        .target         = omap_target,
}