sched/balancing: Simplify the sg_status bitmask and use separate ->overloaded and ->overutilized flags

SG_OVERLOADED and SG_OVERUTILIZED flags plus the sg_status bitmask are an
unnecessary complication that only make the code harder to read and slower.

We only ever set them separately:

 thule:~/tip> git grep SG_OVER kernel/sched/
 kernel/sched/fair.c:            set_rd_overutilized_status(rq->rd, SG_OVERUTILIZED);
 kernel/sched/fair.c:                    *sg_status |= SG_OVERLOADED;
 kernel/sched/fair.c:                    *sg_status |= SG_OVERUTILIZED;
 kernel/sched/fair.c:                            *sg_status |= SG_OVERLOADED;
 kernel/sched/fair.c:            set_rd_overloaded(env->dst_rq->rd, sg_status & SG_OVERLOADED);
 kernel/sched/fair.c:                                       sg_status & SG_OVERUTILIZED);
 kernel/sched/fair.c:    } else if (sg_status & SG_OVERUTILIZED) {
 kernel/sched/fair.c:            set_rd_overutilized_status(env->dst_rq->rd, SG_OVERUTILIZED);
 kernel/sched/sched.h:#define SG_OVERLOADED              0x1 /* More than one runnable task on a CPU. */
 kernel/sched/sched.h:#define SG_OVERUTILIZED            0x2 /* One or more CPUs are over-utilized. */
 kernel/sched/sched.h:           set_rd_overloaded(rq->rd, SG_OVERLOADED);

And use them separately, which results in suboptimal code:

                /* update overload indicator if we are at root domain */
                set_rd_overloaded(env->dst_rq->rd, sg_status & SG_OVERLOADED);

                /* Update over-utilization (tipping point, U >= 0) indicator */
                set_rd_overutilized_status(env->dst_rq->rd,

Introduce separate sg_overloaded and sg_overutilized flags in update_sd_lb_stats()
and its lower level functions, and change all of them to 'bool'.

Remove the now unused SG_OVERLOADED and SG_OVERUTILIZED flags.

Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Tested-by: Shrikanth Hegde <sshegde@linux.ibm.com>
Cc: Qais Yousef <qyousef@layalina.io>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/ZgVPhODZ8/nbsqbP@gmail.com
This commit is contained in:
Ingo Molnar 2024-03-28 12:07:48 +01:00
parent 4d0a63e5b8
commit 4475cd8bfd
2 changed files with 24 additions and 29 deletions

View File

@ -6688,19 +6688,18 @@ static inline bool cpu_overutilized(int cpu)
/* /*
* overutilized value make sense only if EAS is enabled * overutilized value make sense only if EAS is enabled
*/ */
static inline int is_rd_overutilized(struct root_domain *rd) static inline bool is_rd_overutilized(struct root_domain *rd)
{ {
return !sched_energy_enabled() || READ_ONCE(rd->overutilized); return !sched_energy_enabled() || READ_ONCE(rd->overutilized);
} }
static inline void set_rd_overutilized(struct root_domain *rd, static inline void set_rd_overutilized(struct root_domain *rd, bool flag)
unsigned int status)
{ {
if (!sched_energy_enabled()) if (!sched_energy_enabled())
return; return;
WRITE_ONCE(rd->overutilized, status); WRITE_ONCE(rd->overutilized, flag);
trace_sched_overutilized_tp(rd, !!status); trace_sched_overutilized_tp(rd, flag);
} }
static inline void check_update_overutilized_status(struct rq *rq) static inline void check_update_overutilized_status(struct rq *rq)
@ -6711,7 +6710,7 @@ static inline void check_update_overutilized_status(struct rq *rq)
*/ */
if (!is_rd_overutilized(rq->rd) && cpu_overutilized(rq->cpu)) if (!is_rd_overutilized(rq->rd) && cpu_overutilized(rq->cpu))
set_rd_overutilized(rq->rd, SG_OVERUTILIZED); set_rd_overutilized(rq->rd, 1);
} }
#else #else
static inline void check_update_overutilized_status(struct rq *rq) { } static inline void check_update_overutilized_status(struct rq *rq) { }
@ -9934,13 +9933,15 @@ sched_reduced_capacity(struct rq *rq, struct sched_domain *sd)
* @sds: Load-balancing data with statistics of the local group. * @sds: Load-balancing data with statistics of the local group.
* @group: sched_group whose statistics are to be updated. * @group: sched_group whose statistics are to be updated.
* @sgs: variable to hold the statistics for this group. * @sgs: variable to hold the statistics for this group.
* @sg_status: Holds flag indicating the status of the sched_group * @sg_overloaded: sched_group is overloaded
* @sg_overutilized: sched_group is overutilized
*/ */
static inline void update_sg_lb_stats(struct lb_env *env, static inline void update_sg_lb_stats(struct lb_env *env,
struct sd_lb_stats *sds, struct sd_lb_stats *sds,
struct sched_group *group, struct sched_group *group,
struct sg_lb_stats *sgs, struct sg_lb_stats *sgs,
int *sg_status) bool *sg_overloaded,
bool *sg_overutilized)
{ {
int i, nr_running, local_group; int i, nr_running, local_group;
@ -9961,10 +9962,10 @@ static inline void update_sg_lb_stats(struct lb_env *env,
sgs->sum_nr_running += nr_running; sgs->sum_nr_running += nr_running;
if (nr_running > 1) if (nr_running > 1)
*sg_status |= SG_OVERLOADED; *sg_overloaded = 1;
if (cpu_overutilized(i)) if (cpu_overutilized(i))
*sg_status |= SG_OVERUTILIZED; *sg_overutilized = 1;
#ifdef CONFIG_NUMA_BALANCING #ifdef CONFIG_NUMA_BALANCING
sgs->nr_numa_running += rq->nr_numa_running; sgs->nr_numa_running += rq->nr_numa_running;
@ -9986,7 +9987,7 @@ static inline void update_sg_lb_stats(struct lb_env *env,
/* Check for a misfit task on the cpu */ /* Check for a misfit task on the cpu */
if (sgs->group_misfit_task_load < rq->misfit_task_load) { if (sgs->group_misfit_task_load < rq->misfit_task_load) {
sgs->group_misfit_task_load = rq->misfit_task_load; sgs->group_misfit_task_load = rq->misfit_task_load;
*sg_status |= SG_OVERLOADED; *sg_overloaded = 1;
} }
} else if (env->idle && sched_reduced_capacity(rq, env->sd)) { } else if (env->idle && sched_reduced_capacity(rq, env->sd)) {
/* Check for a task running on a CPU with reduced capacity */ /* Check for a task running on a CPU with reduced capacity */
@ -10612,7 +10613,7 @@ static inline void update_sd_lb_stats(struct lb_env *env, struct sd_lb_stats *sd
struct sg_lb_stats *local = &sds->local_stat; struct sg_lb_stats *local = &sds->local_stat;
struct sg_lb_stats tmp_sgs; struct sg_lb_stats tmp_sgs;
unsigned long sum_util = 0; unsigned long sum_util = 0;
int sg_status = 0; bool sg_overloaded = 0, sg_overutilized = 0;
do { do {
struct sg_lb_stats *sgs = &tmp_sgs; struct sg_lb_stats *sgs = &tmp_sgs;
@ -10628,7 +10629,7 @@ static inline void update_sd_lb_stats(struct lb_env *env, struct sd_lb_stats *sd
update_group_capacity(env->sd, env->dst_cpu); update_group_capacity(env->sd, env->dst_cpu);
} }
update_sg_lb_stats(env, sds, sg, sgs, &sg_status); update_sg_lb_stats(env, sds, sg, sgs, &sg_overloaded, &sg_overutilized);
if (!local_group && update_sd_pick_busiest(env, sds, sg, sgs)) { if (!local_group && update_sd_pick_busiest(env, sds, sg, sgs)) {
sds->busiest = sg; sds->busiest = sg;
@ -10657,13 +10658,12 @@ static inline void update_sd_lb_stats(struct lb_env *env, struct sd_lb_stats *sd
if (!env->sd->parent) { if (!env->sd->parent) {
/* update overload indicator if we are at root domain */ /* update overload indicator if we are at root domain */
set_rd_overloaded(env->dst_rq->rd, sg_status & SG_OVERLOADED); set_rd_overloaded(env->dst_rq->rd, sg_overloaded);
/* Update over-utilization (tipping point, U >= 0) indicator */ /* Update over-utilization (tipping point, U >= 0) indicator */
set_rd_overutilized(env->dst_rq->rd, set_rd_overutilized(env->dst_rq->rd, sg_overloaded);
sg_status & SG_OVERUTILIZED); } else if (sg_overutilized) {
} else if (sg_status & SG_OVERUTILIZED) { set_rd_overutilized(env->dst_rq->rd, sg_overutilized);
set_rd_overutilized(env->dst_rq->rd, SG_OVERUTILIZED);
} }
update_idle_cpu_scan(env, sum_util); update_idle_cpu_scan(env, sum_util);

View File

@ -713,7 +713,7 @@ struct rt_rq {
} highest_prio; } highest_prio;
#endif #endif
#ifdef CONFIG_SMP #ifdef CONFIG_SMP
int overloaded; bool overloaded;
struct plist_head pushable_tasks; struct plist_head pushable_tasks;
#endif /* CONFIG_SMP */ #endif /* CONFIG_SMP */
@ -757,7 +757,7 @@ struct dl_rq {
u64 next; u64 next;
} earliest_dl; } earliest_dl;
int overloaded; bool overloaded;
/* /*
* Tasks on this rq that can be pushed away. They are kept in * Tasks on this rq that can be pushed away. They are kept in
@ -850,10 +850,6 @@ struct perf_domain {
struct rcu_head rcu; struct rcu_head rcu;
}; };
/* Scheduling group status flags */
#define SG_OVERLOADED 0x1 /* More than one runnable task on a CPU. */
#define SG_OVERUTILIZED 0x2 /* One or more CPUs are over-utilized. */
/* /*
* We add the notion of a root-domain which will be used to define per-domain * We add the notion of a root-domain which will be used to define per-domain
* variables. Each exclusive cpuset essentially defines an island domain by * variables. Each exclusive cpuset essentially defines an island domain by
@ -874,10 +870,10 @@ struct root_domain {
* - More than one runnable task * - More than one runnable task
* - Running task is misfit * - Running task is misfit
*/ */
int overloaded; bool overloaded;
/* Indicate one or more cpus over-utilized (tipping point) */ /* Indicate one or more cpus over-utilized (tipping point) */
int overutilized; bool overutilized;
/* /*
* The bit corresponding to a CPU gets set here if such CPU has more * The bit corresponding to a CPU gets set here if such CPU has more
@ -2540,9 +2536,8 @@ static inline void add_nr_running(struct rq *rq, unsigned count)
} }
#ifdef CONFIG_SMP #ifdef CONFIG_SMP
if (prev_nr < 2 && rq->nr_running >= 2) { if (prev_nr < 2 && rq->nr_running >= 2)
set_rd_overloaded(rq->rd, SG_OVERLOADED); set_rd_overloaded(rq->rd, 1);
}
#endif #endif
sched_update_tick_dependency(rq); sched_update_tick_dependency(rq);