cpufreq_cooling: Get effective CPU utilization from scheduler

There is nothing schedutil specific in schedutil_cpu_util(), move it to core.c and define it only for CONFIG_SMP.

Signed-off-by Viresh Kumar <[email protected]>

---
 kernel/sched/core.c              | 108 +++++++++++++++++++++++++++++++
 kernel/sched/cpufreq_schedutil.c | 106 ------------------------------
 kernel/sched/sched.h             |  12 +---
 3 files changed, 109 insertions(+), 117 deletions(-)

kernel/sched/core.c +108 -0

index d2003a7d5ab5..b81265aec4a0 100644
@@ -5117,6 +5117,114 @@ struct task_struct *idle_task(int cpu)
        return cpu_rq(cpu)->idle;
 }
 
+#ifdef CONFIG_SMP
+/*
+ * This function computes an effective utilization for the given CPU, to be
+ * used for frequency selection given the linear relation: f = u * f_max.
+ *
+ * The scheduler tracks the following metrics:
+ *
+ *   cpu_util_{cfs,rt,dl,irq}()
+ *   cpu_bw_dl()
+ *
+ * Where the cfs,rt and dl util numbers are tracked with the same metric and
+ * synchronized windows and are thus directly comparable.
+ *
+ * The cfs,rt,dl utilization are the running times measured with rq->clock_task
+ * which excludes things like IRQ and steal-time. These latter are then accrued
+ * in the irq utilization.
+ *
+ * The DL bandwidth number otoh is not a measured metric but a value computed
+ * based on the task model parameters and gives the minimal utilization
+ * required to meet deadlines.
+ */
+unsigned long schedutil_cpu_util(int cpu, unsigned long util_cfs,
+                                unsigned long max, enum schedutil_type type,
+                                struct task_struct *p)
+{
+       unsigned long dl_util, util, irq;
+       struct rq *rq = cpu_rq(cpu);
+
+       if (!uclamp_is_used() &&
+           type == FREQUENCY_UTIL && rt_rq_is_runnable(&rq->rt)) {
+               return max;
+       }
+
+       /*
+        * Early check to see if IRQ/steal time saturates the CPU, can be
+        * because of inaccuracies in how we track these -- see
+        * update_irq_load_avg().
+        */
+       irq = cpu_util_irq(rq);
+       if (unlikely(irq >= max))
+               return max;
+
+       /*
+        * Because the time spend on RT/DL tasks is visible as 'lost' time to
+        * CFS tasks and we use the same metric to track the effective
+        * utilization (PELT windows are synchronized) we can directly add them
+        * to obtain the CPU's actual utilization.
+        *
+        * CFS and RT utilization can be boosted or capped, depending on
+        * utilization clamp constraints requested by currently RUNNABLE
+        * tasks.
+        * When there are no CFS RUNNABLE tasks, clamps are released and
+        * frequency will be gracefully reduced with the utilization decay.
+        */
+       util = util_cfs + cpu_util_rt(rq);
+       if (type == FREQUENCY_UTIL)
+               util = uclamp_rq_util_with(rq, util, p);
+
+       dl_util = cpu_util_dl(rq);
+
+       /*
+        * For frequency selection we do not make cpu_util_dl() a permanent part
+        * of this sum because we want to use cpu_bw_dl() later on, but we need
+        * to check if the CFS+RT+DL sum is saturated (ie. no idle time) such
+        * that we select f_max when there is no idle time.
+        *
+        * NOTE: numerical errors or stop class might cause us to not quite hit
+        * saturation when we should -- something for later.
+        */
+       if (util + dl_util >= max)
+               return max;
+
+       /*
+        * OTOH, for energy computation we need the estimated running time, so
+        * include util_dl and ignore dl_bw.
+        */
+       if (type == ENERGY_UTIL)
+               util += dl_util;
+
+       /*
+        * There is still idle time; further improve the number by using the
+        * irq metric. Because IRQ/steal time is hidden from the task clock we
+        * need to scale the task numbers:
+        *
+        *              max - irq
+        *   U' = irq + --------- * U
+        *                 max
+        */
+       util = scale_irq_capacity(util, irq, max);
+       util += irq;
+
+       /*
+        * Bandwidth required by DEADLINE must always be granted while, for
+        * FAIR and RT, we use blocked utilization of IDLE CPUs as a mechanism
+        * to gracefully reduce the frequency when no tasks show up for longer
+        * periods of time.
+        *
+        * Ideally we would like to set bw_dl as min/guaranteed freq and util +
+        * bw_dl as requested freq. However, cpufreq is not yet ready for such
+        * an interface. So, we only do the latter for now.
+        */
+       if (type == FREQUENCY_UTIL)
+               util += cpu_bw_dl(rq);
+
+       return min(max, util);
+}
+#endif /* CONFIG_SMP */
+
 /**
  * find_process_by_pid - find a process with a matching PID value.
  * @pid: the pid in question.

kernel/sched/cpufreq_schedutil.c +0 -106

index e254745a82cb..2d44befb322b 100644
@@ -169,112 +169,6 @@ static unsigned int get_next_freq(struct sugov_policy *sg_policy,
        return cpufreq_driver_resolve_freq(policy, freq);
 }
 
-/*
- * This function computes an effective utilization for the given CPU, to be
- * used for frequency selection given the linear relation: f = u * f_max.
- *
- * The scheduler tracks the following metrics:
- *
- *   cpu_util_{cfs,rt,dl,irq}()
- *   cpu_bw_dl()
- *
- * Where the cfs,rt and dl util numbers are tracked with the same metric and
- * synchronized windows and are thus directly comparable.
- *
- * The cfs,rt,dl utilization are the running times measured with rq->clock_task
- * which excludes things like IRQ and steal-time. These latter are then accrued
- * in the irq utilization.
- *
- * The DL bandwidth number otoh is not a measured metric but a value computed
- * based on the task model parameters and gives the minimal utilization
- * required to meet deadlines.
- */
-unsigned long schedutil_cpu_util(int cpu, unsigned long util_cfs,
-                                unsigned long max, enum schedutil_type type,
-                                struct task_struct *p)
-{
-       unsigned long dl_util, util, irq;
-       struct rq *rq = cpu_rq(cpu);
-
-       if (!uclamp_is_used() &&
-           type == FREQUENCY_UTIL && rt_rq_is_runnable(&rq->rt)) {
-               return max;
-       }
-
-       /*
-        * Early check to see if IRQ/steal time saturates the CPU, can be
-        * because of inaccuracies in how we track these -- see
-        * update_irq_load_avg().
-        */
-       irq = cpu_util_irq(rq);
-       if (unlikely(irq >= max))
-               return max;
-
-       /*
-        * Because the time spend on RT/DL tasks is visible as 'lost' time to
-        * CFS tasks and we use the same metric to track the effective
-        * utilization (PELT windows are synchronized) we can directly add them
-        * to obtain the CPU's actual utilization.
-        *
-        * CFS and RT utilization can be boosted or capped, depending on
-        * utilization clamp constraints requested by currently RUNNABLE
-        * tasks.
-        * When there are no CFS RUNNABLE tasks, clamps are released and
-        * frequency will be gracefully reduced with the utilization decay.
-        */
-       util = util_cfs + cpu_util_rt(rq);
-       if (type == FREQUENCY_UTIL)
-               util = uclamp_rq_util_with(rq, util, p);
-
-       dl_util = cpu_util_dl(rq);
-
-       /*
-        * For frequency selection we do not make cpu_util_dl() a permanent part
-        * of this sum because we want to use cpu_bw_dl() later on, but we need
-        * to check if the CFS+RT+DL sum is saturated (ie. no idle time) such
-        * that we select f_max when there is no idle time.
-        *
-        * NOTE: numerical errors or stop class might cause us to not quite hit
-        * saturation when we should -- something for later.
-        */
-       if (util + dl_util >= max)
-               return max;
-
-       /*
-        * OTOH, for energy computation we need the estimated running time, so
-        * include util_dl and ignore dl_bw.
-        */
-       if (type == ENERGY_UTIL)
-               util += dl_util;
-
-       /*
-        * There is still idle time; further improve the number by using the
-        * irq metric. Because IRQ/steal time is hidden from the task clock we
-        * need to scale the task numbers:
-        *
-        *              max - irq
-        *   U' = irq + --------- * U
-        *                 max
-        */
-       util = scale_irq_capacity(util, irq, max);
-       util += irq;
-
-       /*
-        * Bandwidth required by DEADLINE must always be granted while, for
-        * FAIR and RT, we use blocked utilization of IDLE CPUs as a mechanism
-        * to gracefully reduce the frequency when no tasks show up for longer
-        * periods of time.
-        *
-        * Ideally we would like to set bw_dl as min/guaranteed freq and util +
-        * bw_dl as requested freq. However, cpufreq is not yet ready for such
-        * an interface. So, we only do the latter for now.
-        */
-       if (type == FREQUENCY_UTIL)
-               util += cpu_bw_dl(rq);
-
-       return min(max, util);
-}
-
 static unsigned long sugov_get_util(struct sugov_cpu *sg_cpu)
 {
        struct rq *rq = cpu_rq(sg_cpu->cpu);

kernel/sched/sched.h +1 -12

index df80bfcea92e..0db6bcf0881f 100644
@@ -2484,7 +2484,6 @@ static inline unsigned long capacity_orig_of(int cpu)
 {
        return cpu_rq(cpu)->cpu_capacity_orig;
 }
-#endif
 
 /**
  * enum schedutil_type - CPU utilization type

@@ -2501,8 +2500,6 @@ enum schedutil_type {
        ENERGY_UTIL,
 };
 
-#ifdef CONFIG_CPU_FREQ_GOV_SCHEDUTIL
-
 unsigned long schedutil_cpu_util(int cpu, unsigned long util_cfs,
                                 unsigned long max, enum schedutil_type type,
                                 struct task_struct *p);

@@ -2533,14 +2530,7 @@ static inline unsigned long cpu_util_rt(struct rq *rq)
 {
        return READ_ONCE(rq->avg_rt.util_avg);
 }
-#else /* CONFIG_CPU_FREQ_GOV_SCHEDUTIL */
-static inline unsigned long schedutil_cpu_util(int cpu, unsigned long util_cfs,
-                                unsigned long max, enum schedutil_type type,
-                                struct task_struct *p)
-{
-       return 0;
-}
-#endif /* CONFIG_CPU_FREQ_GOV_SCHEDUTIL */
+#endif
 
 #ifdef CONFIG_HAVE_SCHED_AVG_IRQ
 static inline unsigned long cpu_util_irq(struct rq *rq)
--

2.25.0.rc1.19.g042ed3e048af

There is nothing schedutil specific in schedutil_cpu_util(), rename it to effective_cpu_util(). Also create and expose another wrapper sched_cpu_util() which can be used by other parts of the kernel, like thermal core (that will be done in a later commit).

Signed-off-by Viresh Kumar <[email protected]>

---
 include/linux/sched.h            | 21 +++++++++++++++++++++
 kernel/sched/core.c              | 11 +++++++++--
 kernel/sched/cpufreq_schedutil.c |  2 +-
 kernel/sched/fair.c              |  6 +++---
 kernel/sched/sched.h             | 19 ++-----------------
 5 files changed, 36 insertions(+), 23 deletions(-)

include/linux/sched.h +9 -0

index 063cd120b459..926b944dae5e 100644
@@ -1926,6 +1926,27 @@ extern long sched_getaffinity(pid_t pid, struct cpumask *mask);
 #define TASK_SIZE_OF(tsk)	TASK_SIZE
 #endif
 
+#ifdef CONFIG_SMP
+/**
+ * enum cpu_util_type - CPU utilization type
+ * @FREQUENCY_UTIL:	Utilization used to select frequency
+ * @ENERGY_UTIL:	Utilization used during energy calculation
+ *
+ * The utilization signals of all scheduling classes (CFS/RT/DL) and IRQ time
+ * need to be aggregated differently depending on the usage made of them. This
+ * enum is used within sched_cpu_util() to differentiate the types of

Hey Viresh,

Quoted 38 lines · Viresh Kumar

There is nothing schedutil specific in schedutil_cpu_util(), rename it to effective_cpu_util(). Also create and expose another wrapper sched_cpu_util() which can be used by other parts of the kernel, like thermal core (that will be done in a later commit).

Signed-off-by Viresh Kumar <[email protected]>

---
 include/linux/sched.h            | 21 +++++++++++++++++++++
 kernel/sched/core.c              | 11 +++++++++--
 kernel/sched/cpufreq_schedutil.c |  2 +-
 kernel/sched/fair.c              |  6 +++---
 kernel/sched/sched.h             | 19 ++-----------------
 5 files changed, 36 insertions(+), 23 deletions(-)

include/linux/sched.h +16 -0

index 063cd120b459..926b944dae5e 100644
@@ -1926,6 +1926,27 @@ extern long sched_getaffinity(pid_t pid, struct cpumask *mask);
 #define TASK_SIZE_OF(tsk)	TASK_SIZE
 #endif
 
+#ifdef CONFIG_SMP
+/**
+ * enum cpu_util_type - CPU utilization type
+ * @FREQUENCY_UTIL:	Utilization used to select frequency
+ * @ENERGY_UTIL:	Utilization used during energy calculation
+ *
+ * The utilization signals of all scheduling classes (CFS/RT/DL) and IRQ time
+ * need to be aggregated differently depending on the usage made of them. This
+ * enum is used within sched_cpu_util() to differentiate the types of
+ * utilization expected by the callers, and adjust the aggregation accordingly.
+ */
+enum cpu_util_type {
+	FREQUENCY_UTIL,
+	ENERGY_UTIL,
+};
+

include/linux/sched.h extern long sched_getaffinity(pid_t pid, struct cpumask *mask); · line 1947

Quoted 3 lines

include/linux/sched.h +3 -0

+/* Returns effective CPU utilization, as seen by the scheduler */
+unsigned long sched_cpu_util(int cpu, enum cpu_util_type type,
+			     unsigned long max);

Are 'type' and 'max' useful to anybody outside of kernel/sched ? If not then how about we hide them, keep the cpu_util_type enum in kernel/sched/sched.h and evaluate arch_scale_cpu_capacity() in sched_cpu_util() directly?

Quoted 41 lines · Quentin Perret

Hey Viresh,

Quoted 38 lines · Viresh Kumar

There is nothing schedutil specific in schedutil_cpu_util(), rename it to effective_cpu_util(). Also create and expose another wrapper sched_cpu_util() which can be used by other parts of the kernel, like thermal core (that will be done in a later commit).

Signed-off-by Viresh Kumar <[email protected]>

---
 include/linux/sched.h            | 21 +++++++++++++++++++++
 kernel/sched/core.c              | 11 +++++++++--
 kernel/sched/cpufreq_schedutil.c |  2 +-
 kernel/sched/fair.c              |  6 +++---
 kernel/sched/sched.h             | 19 ++-----------------
 5 files changed, 36 insertions(+), 23 deletions(-)

include/linux/sched.h +16 -0

index 063cd120b459..926b944dae5e 100644
@@ -1926,6 +1926,27 @@ extern long sched_getaffinity(pid_t pid, struct cpumask *mask);
 #define TASK_SIZE_OF(tsk)	TASK_SIZE
 #endif
 
+#ifdef CONFIG_SMP
+/**
+ * enum cpu_util_type - CPU utilization type
+ * @FREQUENCY_UTIL:	Utilization used to select frequency
+ * @ENERGY_UTIL:	Utilization used during energy calculation
+ *
+ * The utilization signals of all scheduling classes (CFS/RT/DL) and IRQ time
+ * need to be aggregated differently depending on the usage made of them. This
+ * enum is used within sched_cpu_util() to differentiate the types of
+ * utilization expected by the callers, and adjust the aggregation accordingly.
+ */
+enum cpu_util_type {
+	FREQUENCY_UTIL,
+	ENERGY_UTIL,
+};
+

Quoted 8 lines

include/linux/sched.h extern long sched_getaffinity(pid_t pid, struct cpumask *mask); · line 1947

Quoted 3 lines

include/linux/sched.h +3 -0

+/* Returns effective CPU utilization, as seen by the scheduler */
+unsigned long sched_cpu_util(int cpu, enum cpu_util_type type,
+			     unsigned long max);

Are 'type' and 'max' useful to anybody outside of kernel/sched ? If not then how about we hide them, keep the cpu_util_type enum in kernel/sched/sched.h and evaluate arch_scale_cpu_capacity() in sched_cpu_util() directly?

cpufreq_cooling uses 'max' (as can be seen in the next patch).

There is nothing schedutil specific in schedutil_cpu_util(), rename it to effective_cpu_util(). Also create and expose another wrapper sched_cpu_util() which can be used by other parts of the kernel, like thermal core (that will be done in a later commit).

Signed-off-by Viresh Kumar <[email protected]>

---
 include/linux/sched.h            | 21 +++++++++++++++++++++
 kernel/sched/core.c              | 11 +++++++++--
 kernel/sched/cpufreq_schedutil.c |  2 +-
 kernel/sched/fair.c              |  6 +++---
 kernel/sched/sched.h             | 19 ++-----------------
 5 files changed, 36 insertions(+), 23 deletions(-)

include/linux/sched.h +21 -0

index 063cd120b459..926b944dae5e 100644
@@ -1926,6 +1926,27 @@ extern long sched_getaffinity(pid_t pid, struct cpumask *mask);
 #define TASK_SIZE_OF(tsk)      TASK_SIZE
 #endif
 
+#ifdef CONFIG_SMP
+/**
+ * enum cpu_util_type - CPU utilization type
+ * @FREQUENCY_UTIL:    Utilization used to select frequency
+ * @ENERGY_UTIL:       Utilization used during energy calculation
+ *
+ * The utilization signals of all scheduling classes (CFS/RT/DL) and IRQ time
+ * need to be aggregated differently depending on the usage made of them. This
+ * enum is used within sched_cpu_util() to differentiate the types of
+ * utilization expected by the callers, and adjust the aggregation accordingly.
+ */
+enum cpu_util_type {
+       FREQUENCY_UTIL,
+       ENERGY_UTIL,
+};
+
+/* Returns effective CPU utilization, as seen by the scheduler */
+unsigned long sched_cpu_util(int cpu, enum cpu_util_type type,
+                            unsigned long max);
+#endif /* CONFIG_SMP */
+
 #ifdef CONFIG_RSEQ
 
 /*

kernel/sched/core.c +9 -2

index b81265aec4a0..845c976ccd53 100644
@@ -5138,8 +5138,8 @@ struct task_struct *idle_task(int cpu)
  * based on the task model parameters and gives the minimal utilization
  * required to meet deadlines.
  */
-unsigned long schedutil_cpu_util(int cpu, unsigned long util_cfs,
-                                unsigned long max, enum schedutil_type type,
+unsigned long effective_cpu_util(int cpu, unsigned long util_cfs,
+                                unsigned long max, enum cpu_util_type type,
                                 struct task_struct *p)
 {
        unsigned long dl_util, util, irq;

@@ -5223,6 +5223,13 @@ unsigned long schedutil_cpu_util(int cpu, unsigned long util_cfs,
 
        return min(max, util);
 }
+
+unsigned long sched_cpu_util(int cpu, enum cpu_util_type type,
+                            unsigned long max)
+{
+       return effective_cpu_util(cpu, cpu_util_cfs(cpu_rq(cpu)), max, type,
+                                 NULL);
+}
 #endif /* CONFIG_SMP */
 
 /**

kernel/sched/cpufreq_schedutil.c +1 -1

index 2d44befb322b..e71627a3792b 100644
@@ -178,7 +178,7 @@ static unsigned long sugov_get_util(struct sugov_cpu *sg_cpu)
        sg_cpu->max = max;
        sg_cpu->bw_dl = cpu_bw_dl(rq);
 
-       return schedutil_cpu_util(sg_cpu->cpu, util, max, FREQUENCY_UTIL, NULL);
+       return effective_cpu_util(sg_cpu->cpu, util, max, FREQUENCY_UTIL, NULL);
 }
 
 /**

kernel/sched/fair.c +3 -3

index 290f9e38378c..0e1c8eb7ad53 100644
@@ -6499,7 +6499,7 @@ compute_energy(struct task_struct *p, int dst_cpu, struct perf_domain *pd)
                 * is already enough to scale the EM reported power
                 * consumption at the (eventually clamped) cpu_capacity.
                 */
-               sum_util += schedutil_cpu_util(cpu, util_cfs, cpu_cap,
+               sum_util += effective_cpu_util(cpu, util_cfs, cpu_cap,
                                               ENERGY_UTIL, NULL);
 
                /*

@@ -6509,7 +6509,7 @@ compute_energy(struct task_struct *p, int dst_cpu, struct perf_domain *pd)
                 * NOTE: in case RT tasks are running, by default the
                 * FREQUENCY_UTIL's utilization can be max OPP.
                 */
-               cpu_util = schedutil_cpu_util(cpu, util_cfs, cpu_cap,
+               cpu_util = effective_cpu_util(cpu, util_cfs, cpu_cap,
                                              FREQUENCY_UTIL, tsk);
                max_util = max(max_util, cpu_util);
        }

@@ -6607,7 +6607,7 @@ static int find_energy_efficient_cpu(struct task_struct *p, int prev_cpu)
                         * IOW, placing the task there would make the CPU
                         * overutilized. Take uclamp into account to see how
                         * much capacity we can get out of the CPU; this is
-                        * aligned with schedutil_cpu_util().
+                        * aligned with sched_cpu_util().
                         */
                        util = uclamp_rq_util_with(cpu_rq(cpu), util, p);
                        if (!fits_capacity(util, cpu_cap))

kernel/sched/sched.h +2 -18

index 0db6bcf0881f..4fab3b930ace 100644
@@ -2485,23 +2485,8 @@ static inline unsigned long capacity_orig_of(int cpu)
        return cpu_rq(cpu)->cpu_capacity_orig;
 }
 
-/**
- * enum schedutil_type - CPU utilization type
- * @FREQUENCY_UTIL:    Utilization used to select frequency
- * @ENERGY_UTIL:       Utilization used during energy calculation
- *
- * The utilization signals of all scheduling classes (CFS/RT/DL) and IRQ time
- * need to be aggregated differently depending on the usage made of them. This
- * enum is used within schedutil_freq_util() to differentiate the types of
- * utilization expected by the callers, and adjust the aggregation accordingly.
- */
-enum schedutil_type {
-       FREQUENCY_UTIL,
-       ENERGY_UTIL,
-};
-
-unsigned long schedutil_cpu_util(int cpu, unsigned long util_cfs,
-                                unsigned long max, enum schedutil_type type,
+unsigned long effective_cpu_util(int cpu, unsigned long util_cfs,
+                                unsigned long max, enum cpu_util_type type,
                                 struct task_struct *p);
 
 static inline unsigned long cpu_bw_dl(struct rq *rq)
--

2.25.0.rc1.19.g042ed3e048af

Hi Viresh,

I love your patch! Yet something to improve:

[auto build test ERROR on 3650b228f83adda7e5ee532e2b90429c03f7b9ec]

url:    https://github.com/0day-ci/linux/commits/Viresh-Kumar/cpufreq_cooling-Get-effective-CPU-utilization-from-scheduler/20201124-143027
base:    3650b228f83adda7e5ee532e2b90429c03f7b9ec
config: powerpc64-randconfig-r025-20201202 (attached as .config)
compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project 2671fccf0381769276ca8246ec0499adcb9b0355)

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# install powerpc64 cross compiling tool for clang build
# apt-get install binutils-powerpc64-linux-gnu
# https://github.com/0day-ci/linux/commit/17f0c7f4372070206925c3a10ec0e7a09d03615e
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Viresh-Kumar/cpufreq_cooling-Get-effective-CPU-utilization-from-scheduler/20201124-143027 git checkout 17f0c7f4372070206925c3a10ec0e7a09d03615e
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=powerpc64

If you fix the issue, kindly add following tag as appropriate

Reported-by kernel test robot <[email protected]>

All errors (new ones prefixed by >>):

Quoted 1 line

Quoted 1 line

drivers/thermal/cpufreq_cooling.c:562:8: error: implicit declaration of function 'allocate_idle_time' [-Werror,-Wimplicit-function-declaration]

           ret = allocate_idle_time(cpufreq_cdev);
                 ^

Quoted 1 line

Quoted 1 line

drivers/thermal/cpufreq_cooling.c:624:2: error: implicit declaration of function 'free_idle_time' [-Werror,-Wimplicit-function-declaration]

           free_idle_time(cpufreq_cdev);
           ^
   drivers/thermal/cpufreq_cooling.c:717:2: error: implicit declaration of function 'free_idle_time' [-Werror,-Wimplicit-function-declaration]
           free_idle_time(cpufreq_cdev);
           ^

3 errors generated.

vim +/allocate_idle_time +562 drivers/thermal/cpufreq_cooling.c

Quoted 9 lines · kernel test robot

Hi Viresh,

I love your patch! Yet something to improve:

[auto build test ERROR on 3650b228f83adda7e5ee532e2b90429c03f7b9ec]

url:    https://github.com/0day-ci/linux/commits/Viresh-Kumar/cpufreq_cooling-Get-effective-CPU-utilization-from-scheduler/20201124-143027
base:    3650b228f83adda7e5ee532e2b90429c03f7b9ec
config: powerpc64-randconfig-r025-20201202 (attached as .config)

Quoted 13 lines

compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project 2671fccf0381769276ca8246ec0499adcb9b0355) reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# install powerpc64 cross compiling tool for clang build
# apt-get install binutils-powerpc64-linux-gnu
# https://github.com/0day-ci/linux/commit/17f0c7f4372070206925c3a10ec0e7a09d03615e
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Viresh-Kumar/cpufreq_cooling-Get-effective-CPU-utilization-from-scheduler/20201124-143027 git checkout 17f0c7f4372070206925c3a10ec0e7a09d03615e
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=powerpc64

Quoted 8 lines

If you fix the issue, kindly add following tag as appropriate

Reported-by kernel test robot <[email protected]>

All errors (new ones prefixed by >>):

Quoted 1 line

Quoted 1 line

drivers/thermal/cpufreq_cooling.c:562:8: error: implicit declaration of function 'allocate_idle_time' [-Werror,-Wimplicit-function-declaration]

           ret = allocate_idle_time(cpufreq_cdev);
                 ^

Quoted 9 lines

Quoted 1 line

Quoted 1 line

drivers/thermal/cpufreq_cooling.c:624:2: error: implicit declaration of function 'free_idle_time' [-Werror,-Wimplicit-function-declaration]

           free_idle_time(cpufreq_cdev);
           ^
   drivers/thermal/cpufreq_cooling.c:717:2: error: implicit declaration of function 'free_idle_time' [-Werror,-Wimplicit-function-declaration]
           free_idle_time(cpufreq_cdev);
           ^

3 errors generated.

vim +/allocate_idle_time +562 drivers/thermal/cpufreq_cooling.c

I am not sure why this should happen here, I don't see any such errors on my side. Can someone please have another look ?

--
viresh

--
You received this message because you are subscribed to the Google Groups "Clang Built Linux" group.
To unsubscribe from this group and stop receiving emails from it, send an email to clang-built-linux+unsubscribe(a)googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/clang-built-linux/20201207094419.lakxblzsono3nqpf%40vireshk-i7.

I can see in the top commit
(https://github.com/0day-ci/linux/commit/17f0c7f4372070206925c3a10ec0e7a09d03615e)
you've modified __cpufreq_cooling_register to now call allocate_idle_time. -Wimplicit-function-declaration is observed when calling a function for which there has been no previous declaration. Let's look and see where or under what config allocate_idle_time is declared, and how a randconfig might expose the missing declaration. (The first thing I suspect is transitive header dependencies, where some intermediary header changes what it includes based on #define/CONFIGs, or calling a function before it's been declared).

It looks like allocate_idle_time is declared in the same commit...while it's defined twice (once for CONFIG_SMP, once without), it's not defined when CONFIG_THERMAL_GOV_POWER_ALLOCATOR is not enabled, which is probably what the randconfig has tickled.

Several parts of the kernel are already using the effective CPU utilization (as seen by the scheduler) to get the current load on the CPU, do the same here instead of depending on the idle time of the CPU, which isn't that accurate comparatively.

This is also the right thing to do as it makes the cpufreq governor (schedutil) align better with the cpufreq_cooling driver, as the power requested by cpufreq_cooling governor will exactly match the next frequency requested by the schedutil governor since they are both using the same metric to calculate load.

This was tested on ARM Hikey6220 platform with hackbench, sysbench and schbench. None of them showed any regression or significant improvements. Schbench is the most important ones out of these as it creates the scenario where the utilization numbers provide a better estimate of the future.

Scenario 1: The CPUs were mostly idle in the previous polling window of the IPA governor as the tasks were sleeping and here are the details from traces (load is in %):

Old: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=203 load={{0x35,0x1,0x0,0x31,0x0,0x0,0x64,0x0}} dynamic_power=1339
New: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=600 load={{0x60,0x46,0x45,0x45,0x48,0x3b,0x61,0x44}} dynamic_power=3960

Here, the "Old" line gives the load and requested_power (dynamic_power here) numbers calculated using the idle time based implementation, while "New" is based on the CPU utilization from scheduler.

As can be clearly seen, the load and requested_power numbers are simply incorrect in the idle time based approach and the numbers collected from CPU's utilization are much closer to the reality.

Scenario 2: The CPUs were busy in the previous polling window of the IPA governor:

Old: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=800 load={{0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64}} dynamic_power=5280
New: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=708 load={{0x4d,0x5c,0x5c,0x5b,0x5c,0x5c,0x51,0x5b}} dynamic_power=4672

As can be seen, the idle time based load is 100% for all the CPUs as it took only the last window into account, but in reality the CPUs aren't that loaded as shown by the utilization numbers.

Quoted 17 lines · Viresh Kumar

Several parts of the kernel are already using the effective CPU utilization (as seen by the scheduler) to get the current load on the CPU, do the same here instead of depending on the idle time of the CPU, which isn't that accurate comparatively.

This is also the right thing to do as it makes the cpufreq governor (schedutil) align better with the cpufreq_cooling driver, as the power requested by cpufreq_cooling governor will exactly match the next frequency requested by the schedutil governor since they are both using the same metric to calculate load.

This was tested on ARM Hikey6220 platform with hackbench, sysbench and schbench. None of them showed any regression or significant improvements. Schbench is the most important ones out of these as it creates the scenario where the utilization numbers provide a better estimate of the future.

Quoted 6 lines

Scenario 1: The CPUs were mostly idle in the previous polling window of the IPA governor as the tasks were sleeping and here are the details from traces (load is in %):

Old: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=203 load={{0x35,0x1,0x0,0x31,0x0,0x0,0x64,0x0}} dynamic_power=1339
New: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=600 load={{0x60,0x46,0x45,0x45,0x48,0x3b,0x61,0x44}} dynamic_power=3960

When I ran schbench (-t 16 -r 5) on hikey960 I get multiple (~50) instances of ~80ms task activity phase and then ~20ms idle phase on all CPUs.

So I assume that scenario 1 is at the beginning (but you mentioned the task were sleeping?)

and scenario 2 is somewhere in the middle of the
testrun?

IMHO, the util-based approach delivers really better results at the beginning and at the end of the entire testrun.
During the testrun, the util-based and the idle-based approach deliver similar results.

It's a little bit tricky to compare test results since the IPA sampling rate is 100ms and the load values you get depend on how the workload pattern and the IPA sampling align.

Quoted 7 lines

Here, the "Old" line gives the load and requested_power (dynamic_power here) numbers calculated using the idle time based implementation, while "New" is based on the CPU utilization from scheduler.

As can be clearly seen, the load and requested_power numbers are simply incorrect in the idle time based approach and the numbers collected from CPU's utilization are much closer to the reality.

I assume the IPA sampling is done after ~50ms of the first task activity phase.

Quoted 3 lines

Scenario 2: The CPUs were busy in the previous polling window of the IPA governor:

Quoted 6 lines

Old: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=800 load={{0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64}} dynamic_power=5280
New: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=708 load={{0x4d,0x5c,0x5c,0x5b,0x5c,0x5c,0x51,0x5b}} dynamic_power=4672

As can be seen, the idle time based load is 100% for all the CPUs as it took only the last window into account, but in reality the CPUs aren't that loaded as shown by the utilization numbers.

Is this an IPA sampling at the end of the ~20ms idle phase?

Quoted 1 line · Dietmar Eggemann

On 24/11/2020 07:26, Viresh Kumar wrote:

Quoted 6 lines

When I ran schbench (-t 16 -r 5) on hikey960 I get multiple (~50) instances of ~80ms task activity phase and then ~20ms idle phase on all CPUs.

So I assume that scenario 1 is at the beginning (but you mentioned the task were sleeping?)

I am not able to find the exact values I used, but I did something like this to create a scenario where the old computations shall find the CPU as idle in the last IPA window:

- schbench -m 2 -t 4 -s 25000 -c 20000 -r 60

- sampling rate of IPA to 10 ms

With this IPA wakes up many times and finds the CPUs to have been idle in the last IPA window (i.e. 10ms).

Quoted 8 lines

Quoted 6 lines

Old: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=800 load={{0x64,0x64,0x64,0x64,0x64,0x64,0x64,0x64}} dynamic_power=5280
New: thermal_power_cpu_get_power: cpus=00000000,000000ff freq=1200000 total_load=708 load={{0x4d,0x5c,0x5c,0x5b,0x5c,0x5c,0x51,0x5b}} dynamic_power=4672

As can be seen, the idle time based load is 100% for all the CPUs as it took only the last window into account, but in reality the CPUs aren't that loaded as shown by the utilization numbers.

Is this an IPA sampling at the end of the ~20ms idle phase?

This is during the phase where the CPUs were fully busy for the last period.