Messages in this thread |  | | Date | Thu, 27 Jun 2013 15:06:43 +0530 | Subject | Re: [PATCH v4 5/7] cpufreq: Calculate number of busy CPUs | From | Viresh Kumar <> |
| |
On 19 June 2013 22:42, Lukasz Majewski <l.majewski@samsung.com> wrote: > In the core governor code, per cpu load value is calculated. This patch > uses it to mark processor as a "busy" one, when load value is higher than > 90%.
How can we take this assumption. What about a cpu which is only 70-80% busy? So, we would be running at around 70% of policy->max.. Now at this point you want to enable boost frequency and so that will happen for all cores... Wouldn't that burn your chip?
> New cpufreq sysfs attribute is created (busy_cpus). It is read only > and provides information about number of actually busy CPU.
Not required.
> diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c > index 077cea7..3402533 100644 > --- a/drivers/cpufreq/cpufreq_governor.c > +++ b/drivers/cpufreq/cpufreq_governor.c > @@ -148,6 +148,7 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu) > continue; > > load = 100 * (wall_time - idle_time) / wall_time; > + cpufreq_set_busy_cpu(j, load > 90 ? 1 : 0);
do this only when boost is enabled.
> if (dbs_data->cdata->governor == GOV_ONDEMAND) { > int freq_avg = __cpufreq_driver_getavg(policy, j);
|  |