lkml.org 
[lkml]   [2023]   [Jan]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    From
    SubjectRE: [PATCH v9 09/13] cpufreq: amd-pstate: add driver working mode switch support
    Date
    [AMD Official Use Only - General]



    > -----Original Message-----
    > From: Karny, Wyes <Wyes.Karny@amd.com>
    > Sent: Thursday, January 5, 2023 2:02 PM
    > To: Yuan, Perry <Perry.Yuan@amd.com>; rafael.j.wysocki@intel.com;
    > Limonciello, Mario <Mario.Limonciello@amd.com>; Huang, Ray
    > <Ray.Huang@amd.com>; viresh.kumar@linaro.org
    > Cc: Sharma, Deepak <Deepak.Sharma@amd.com>; Fontenot, Nathan
    > <Nathan.Fontenot@amd.com>; Deucher, Alexander
    > <Alexander.Deucher@amd.com>; Huang, Shimmer
    > <Shimmer.Huang@amd.com>; Du, Xiaojian <Xiaojian.Du@amd.com>; Meng,
    > Li (Jassmine) <Li.Meng@amd.com>; linux-pm@vger.kernel.org; linux-
    > kernel@vger.kernel.org
    > Subject: Re: [PATCH v9 09/13] cpufreq: amd-pstate: add driver working mode
    > switch support
    >
    > Hi Perry,
    >
    > On 12/25/2022 10:04 PM, Perry Yuan wrote:
    > > From: Perry Yuan <Perry.Yuan@amd.com>
    > >
    > > While amd-pstate driver was loaded with specific driver mode, it will
    > > need to check which mode is enabled for the pstate driver,add this
    > > sysfs entry to show the current status
    > >
    > > $ cat /sys/devices/system/cpu/amd-pstate/status
    > > active
    > >
    > > Meanwhile, user can switch the pstate driver mode with writing mode
    > > string to sysfs entry as below.
    > >
    > > Enable passive mode:
    > > $ sudo bash -c "echo passive > /sys/devices/system/cpu/amd-
    > pstate/status"
    > >
    > > Enable active mode (EPP driver mode):
    > > $ sudo bash -c "echo active > /sys/devices/system/cpu/amd-pstate/status"
    > >
    > > Signed-off-by: Perry Yuan <Perry.Yuan@amd.com>
    > > ---
    > > drivers/cpufreq/amd-pstate.c | 118
    > > +++++++++++++++++++++++++++++++++++
    > > 1 file changed, 118 insertions(+)
    > >
    > > diff --git a/drivers/cpufreq/amd-pstate.c
    > > b/drivers/cpufreq/amd-pstate.c index e3676d1a85c7..8ceca4524fc1 100644
    > > --- a/drivers/cpufreq/amd-pstate.c
    > > +++ b/drivers/cpufreq/amd-pstate.c
    > > @@ -63,6 +63,7 @@ static struct cpufreq_driver *default_pstate_driver;
    > > static struct cpufreq_driver amd_pstate_driver; static struct
    > > cpufreq_driver amd_pstate_epp_driver; static int cppc_state =
    > > AMD_PSTATE_DISABLE;
    > > +struct kobject *amd_pstate_kobj;
    > >
    > > static inline int get_mode_idx_from_str(const char *str, size_t size)
    > > { @@ -632,6 +633,8 @@ static int amd_pstate_cpu_init(struct
    > > cpufreq_policy *policy)
    > > policy->driver_data = cpudata;
    > >
    > > amd_pstate_boost_init(cpudata);
    > > + if (!default_pstate_driver->adjust_perf)
    > > + default_pstate_driver->adjust_perf =
    > amd_pstate_adjust_perf;
    > >
    > > return 0;
    > >
    > > @@ -772,12 +775,99 @@ static ssize_t
    > show_energy_performance_preference(
    > > return sysfs_emit(buf, "%s\n", energy_perf_strings[preference]); }
    > >
    > > +static ssize_t amd_pstate_show_status(char *buf) {
    > > + if (!default_pstate_driver)
    > > + return sysfs_emit(buf, "off\n");
    >
    > IMO 'disable' is more consistent to cppc_state.

    Will change the string to be "disable" in v10.

    >
    > > +
    > > + return sysfs_emit(buf, "%s\n",
    > amd_pstate_mode_string[cppc_state]);
    > > +}
    > > +
    > > +static void amd_pstate_driver_cleanup(void) {
    > > + default_pstate_driver = NULL;
    > > +}
    > > +
    > > +static int amd_pstate_update_status(const char *buf, size_t size) {
    > > + int ret;
    > > + int mode_idx;
    > > +
    > > + if (size > 7 || size < 3)
    > > + return -EINVAL;
    > > + mode_idx = get_mode_idx_from_str(buf, size);
    > > +
    > > + switch(mode_idx) {
    > > + case AMD_PSTATE_DISABLE:
    > > + if (!default_pstate_driver)
    > > + return -EINVAL;
    > > + if (cppc_state == AMD_PSTATE_ACTIVE)
    > > + return -EBUSY;
    > > + ret = cpufreq_unregister_driver(default_pstate_driver);
    > > + amd_pstate_driver_cleanup();
    > > + break;
    > > + case AMD_PSTATE_PASSIVE:
    > > + if (default_pstate_driver) {
    > > + if (default_pstate_driver == &amd_pstate_driver)
    > > + return 0;
    > > + cpufreq_unregister_driver(default_pstate_driver);
    > > + cppc_state = AMD_PSTATE_PASSIVE;
    > > + default_pstate_driver = &amd_pstate_driver;
    > > + }
    > > +
    > > + ret = cpufreq_register_driver(default_pstate_driver);
    > > + break;
    > > + case AMD_PSTATE_ACTIVE:
    > > + if (default_pstate_driver) {
    > > + if (default_pstate_driver ==
    > &amd_pstate_epp_driver)
    > > + return 0;
    > > + cpufreq_unregister_driver(default_pstate_driver);
    > > + default_pstate_driver = &amd_pstate_epp_driver;
    > > + cppc_state = AMD_PSTATE_ACTIVE;
    > > + }
    > > +
    > > + ret = cpufreq_register_driver(default_pstate_driver);
    > > + break;
    > > + default:
    > > + break;
    > > + ret = -EINVAL;
    > > + }
    > > +
    > > + return ret;
    > > +}
    >
    > The implementation of amd_pstate_update_status function is good as long
    > as the possible states are less.
    > Currently prev_state and next_state has 9 combinations. But with guided
    > mode this becomes 16 combinations.
    > Do you have any concerns if I optimize this function by creating a state
    > transition table in guided patch series?

    No concern , please go ahead in your series.

    >
    > > +
    > > +static ssize_t show_status(struct kobject *kobj,
    > > + struct kobj_attribute *attr, char *buf) {
    > > + ssize_t ret;
    > > +
    > > + mutex_lock(&amd_pstate_driver_lock);
    > > + ret = amd_pstate_show_status(buf);
    > > + mutex_unlock(&amd_pstate_driver_lock);
    > > +
    > > + return ret;
    > > +}
    > > +
    > > +static ssize_t store_status(struct kobject *a, struct kobj_attribute *b,
    > > + const char *buf, size_t count) {
    > > + char *p = memchr(buf, '\n', count);
    > > + int ret;
    > > +
    > > + mutex_lock(&amd_pstate_driver_lock);
    > > + ret = amd_pstate_update_status(buf, p ? p - buf : count);
    > > + mutex_unlock(&amd_pstate_driver_lock);
    > > +
    > > + return ret < 0 ? ret : count;
    > > +}
    > > +
    > > cpufreq_freq_attr_ro(amd_pstate_max_freq);
    > > cpufreq_freq_attr_ro(amd_pstate_lowest_nonlinear_freq);
    > >
    > > cpufreq_freq_attr_ro(amd_pstate_highest_perf);
    > > cpufreq_freq_attr_rw(energy_performance_preference);
    > > cpufreq_freq_attr_ro(energy_performance_available_preferences);
    > > +define_one_global_rw(status);
    > >
    > > static struct freq_attr *amd_pstate_attr[] = {
    > > &amd_pstate_max_freq,
    > > @@ -795,6 +885,15 @@ static struct freq_attr *amd_pstate_epp_attr[] = {
    > > NULL,
    > > };
    > >
    > > +static struct attribute *pstate_global_attributes[] = {
    > > + &status.attr,
    > > + NULL
    > > +};
    > > +
    > > +static const struct attribute_group amd_pstate_global_attr_group = {
    > > + .attrs = pstate_global_attributes,
    > > +};
    > > +
    > > static int amd_pstate_epp_cpu_init(struct cpufreq_policy *policy) {
    > > int min_freq, max_freq, nominal_freq, lowest_nonlinear_freq, ret;
    > @@
    > > -1154,6 +1253,25 @@ static int __init amd_pstate_init(void)
    > > if (ret)
    > > pr_err("failed to register with return %d\n", ret);
    > >
    > > + amd_pstate_kobj = kobject_create_and_add("amd-pstate",
    > &cpu_subsys.dev_root->kobj);
    > > + if (!amd_pstate_kobj) {
    > > + ret = -EINVAL;
    > > + pr_err("global sysfs registration failed.\n");
    > > + goto kobject_free;
    > > + }
    > > +
    > > + ret = sysfs_create_group(amd_pstate_kobj,
    > &amd_pstate_global_attr_group);
    > > + if (ret) {
    > > + pr_err("sysfs attribute export failed with error %d.\n", ret);
    > > + goto global_attr_free;
    > > + }
    > > +
    > > + return ret;
    > > +
    > > +global_attr_free:
    > > + kobject_put(amd_pstate_kobj);
    > > +kobject_free:
    > > + cpufreq_unregister_driver(default_pstate_driver);
    > > return ret;
    > > }
    > > device_initcall(amd_pstate_init);
    >
    > --
    > Thanks & Regards,
    > Wyes
    \
     
     \ /
      Last update: 2023-03-26 23:28    [W:5.416 / U:0.040 seconds]
    ©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site