linux-kernel - Re: [PATCH v1 02/10] cpufreq: Add a flag to auto-register a cooling device

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20190117044246.jcwhb5bxb6he33lp@vireshk-i7>
Date:   Thu, 17 Jan 2019 10:12:46 +0530
From:   Viresh Kumar <viresh.kumar@...aro.org>
To:     "Rafael J. Wysocki" <rjw@...ysocki.net>
Cc:     Amit Kucheria <amit.kucheria@...aro.org>,
        linux-kernel@...r.kernel.org, linux-arm-msm@...r.kernel.org,
        edubezval@...il.com, swboyd@...omium.org, dianders@...omium.org,
        mka@...omium.org, linux-pm@...r.kernel.org
Subject: Re: [PATCH v1 02/10] cpufreq: Add a flag to auto-register a cooling
 device

On 17-01-19, 00:03, Rafael J. Wysocki wrote:
> On Monday, January 14, 2019 5:34:54 PM CET Amit Kucheria wrote:
> > All cpufreq drivers do similar things to register as a cooling device.
> > Provide a cpufreq driver flag so drivers can just ask the cpufreq core
> > to register the cooling device on their behalf. This allows us to get
> > rid of duplicated code in the drivers.
> > 
> > Suggested-by: Stephen Boyd <swboyd@...omium.org>
> > Suggested-by: Viresh Kumar <viresh.kumar@...aro.org>
> > Signed-off-by: Amit Kucheria <amit.kucheria@...aro.org>
> > ---
> >  drivers/cpufreq/cpufreq.c | 17 +++++++++++++++++
> >  include/linux/cpufreq.h   |  6 ++++++
> >  2 files changed, 23 insertions(+)
> > 
> > diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> > index 6f23ebb395f1..cd6e750d3d82 100644
> > --- a/drivers/cpufreq/cpufreq.c
> > +++ b/drivers/cpufreq/cpufreq.c
> > @@ -30,6 +30,7 @@
> >  #include <linux/syscore_ops.h>
> >  #include <linux/tick.h>
> >  #include <trace/events/power.h>
> > +#include <linux/cpu_cooling.h>
> >  
> >  static LIST_HEAD(cpufreq_policy_list);
> >  
> > @@ -1318,6 +1319,14 @@ static int cpufreq_online(unsigned int cpu)
> >  	if (cpufreq_driver->ready)
> >  		cpufreq_driver->ready(policy);
> >  
> > +#ifdef CONFIG_CPU_THERMAL
> > +	if (cpufreq_driver->flags & CPUFREQ_AUTO_REGISTER_COOLING_DEV) {
> > +		struct thermal_cooling_device **cdev = &policy->cooldev;

We use cdev for the cooling device everywhere in the kernel, so please
do s/cooldev/cdev/ in your patches.

> > +
> > +		*cdev = of_cpufreq_cooling_register(policy);
> 
> What would be wrong with
> 
> 		policy->cooldev = of_cpufreq_cooling_register(policy);
> 
> > +	}
> > +#endif
> 
> Please remove the #ifdefs from cpufreq_online() and cpufreq_offline().
> 
> Use wrappers that would become empty stubs for CONFIG_CPU_THERMAL unset.
> 
> > +
> >  	pr_debug("initialization complete\n");
> >  
> >  	return 0;
> > @@ -1411,6 +1420,14 @@ static int cpufreq_offline(unsigned int cpu)
> >  	if (has_target())
> >  		cpufreq_exit_governor(policy);
> >  
> > +#ifdef CONFIG_CPU_THERMAL
> > +	if (cpufreq_driver->flags & CPUFREQ_AUTO_REGISTER_COOLING_DEV) {
> > +		struct thermal_cooling_device **cdev = &policy->cooldev;
> > +
> > +		cpufreq_cooling_unregister(*cdev);
> 
> Again, why don't you simply pass policy->cooldev here?

I also had the same comments when I looked at your patch :)

I also think we must do the unregistering before calling stop_cpu()
callback.

> Also, would it make sense to clear policy->cooldev at this point?  It points
> to freed memory after cpufreq_cooling_unregister().

Since the core doesn't refer to this field at all and uses it only
while registering/unregistering as a cooling device, there is no
technical issue that we will have today. If someone uses the dangling
pointer later on in future, it will be a bug. So I wouldn't care much
about resetting it here.

> > +	}
> > +#endif
> > +
> >  	/*
> >  	 * Perform the ->exit() even during light-weight tear-down,
> >  	 * since this is a core component, and is essential for the
> > diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h
> > index 7d0cf54125fa..70ad02088825 100644
> > --- a/include/linux/cpufreq.h
> > +++ b/include/linux/cpufreq.h
> > @@ -390,6 +390,12 @@ struct cpufreq_driver {
> >   */
> >  #define CPUFREQ_NO_AUTO_DYNAMIC_SWITCHING (1 << 6)
> >  
> > +/*
> > + * Set by drivers that want the core to automatically register the cpufreq
> > + * driver as a thermal cooling device

Add a full-stop here please.

> > + */
> > +#define CPUFREQ_AUTO_REGISTER_COOLING_DEV (1 << 7)
> > +
> >  int cpufreq_register_driver(struct cpufreq_driver *driver_data);
> >  int cpufreq_unregister_driver(struct cpufreq_driver *driver_data);
> >  
> > 
> 

-- 
viresh