The 'cfs_rq' in these helpers
update_stats_{wait_start, wait_end, enqueue_sleeper} is only used to get
the rq_clock, so we can pass the rq directly. Then these helpers can be
used by all sched class after being moved into stats.h.
After that change, the size of vmlinux is increased around 824Bytes.
w/o this patch, with this patch
Size of vmlinux: 78443832 78444656
Signed-off-by: Yafang Shao <[email protected]>
---
kernel/sched/fair.c | 148 ++-----------------------------------------
kernel/sched/stats.h | 144 +++++++++++++++++++++++++++++++++++++++++
2 files changed, 149 insertions(+), 143 deletions(-)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 59e454cae3be..946b60f586e4 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -869,124 +869,6 @@ static void update_curr_fair(struct rq *rq)
update_curr(cfs_rq_of(&rq->curr->se));
}
-static inline void
-update_stats_wait_start(struct cfs_rq *cfs_rq, struct sched_entity *se)
-{
- u64 wait_start, prev_wait_start;
-
- if (!schedstat_enabled())
- return;
-
- wait_start = rq_clock(rq_of(cfs_rq));
- prev_wait_start = schedstat_val(se->statistics.wait_start);
-
- if (entity_is_task(se) && task_on_rq_migrating(task_of(se)) &&
- likely(wait_start > prev_wait_start))
- wait_start -= prev_wait_start;
-
- __schedstat_set(se->statistics.wait_start, wait_start);
-}
-
-static inline void
-update_stats_wait_end(struct cfs_rq *cfs_rq, struct sched_entity *se)
-{
- struct task_struct *p;
- u64 delta;
-
- if (!schedstat_enabled())
- return;
-
- delta = rq_clock(rq_of(cfs_rq)) - schedstat_val(se->statistics.wait_start);
-
- if (entity_is_task(se)) {
- p = task_of(se);
- if (task_on_rq_migrating(p)) {
- /*
- * Preserve migrating task's wait time so wait_start
- * time stamp can be adjusted to accumulate wait time
- * prior to migration.
- */
- __schedstat_set(se->statistics.wait_start, delta);
- return;
- }
- trace_sched_stat_wait(p, delta);
- }
-
- __schedstat_set(se->statistics.wait_max,
- max(schedstat_val(se->statistics.wait_max), delta));
- __schedstat_inc(se->statistics.wait_count);
- __schedstat_add(se->statistics.wait_sum, delta);
- __schedstat_set(se->statistics.wait_start, 0);
-}
-
-static inline void
-update_stats_enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se)
-{
- struct task_struct *tsk = NULL;
- u64 sleep_start, block_start;
-
- if (!schedstat_enabled())
- return;
-
- sleep_start = schedstat_val(se->statistics.sleep_start);
- block_start = schedstat_val(se->statistics.block_start);
-
- if (entity_is_task(se))
- tsk = task_of(se);
-
- if (sleep_start) {
- u64 delta = rq_clock(rq_of(cfs_rq)) - sleep_start;
-
- if ((s64)delta < 0)
- delta = 0;
-
- if (unlikely(delta > schedstat_val(se->statistics.sleep_max)))
- __schedstat_set(se->statistics.sleep_max, delta);
-
- __schedstat_set(se->statistics.sleep_start, 0);
- __schedstat_add(se->statistics.sum_sleep_runtime, delta);
-
- if (tsk) {
- account_scheduler_latency(tsk, delta >> 10, 1);
- trace_sched_stat_sleep(tsk, delta);
- }
- }
- if (block_start) {
- u64 delta = rq_clock(rq_of(cfs_rq)) - block_start;
-
- if ((s64)delta < 0)
- delta = 0;
-
- if (unlikely(delta > schedstat_val(se->statistics.block_max)))
- __schedstat_set(se->statistics.block_max, delta);
-
- __schedstat_set(se->statistics.block_start, 0);
- __schedstat_add(se->statistics.sum_sleep_runtime, delta);
-
- if (tsk) {
- if (tsk->in_iowait) {
- __schedstat_add(se->statistics.iowait_sum, delta);
- __schedstat_inc(se->statistics.iowait_count);
- trace_sched_stat_iowait(tsk, delta);
- }
-
- trace_sched_stat_blocked(tsk, delta);
-
- /*
- * Blocking time is in units of nanosecs, so shift by
- * 20 to get a milliseconds-range estimation of the
- * amount of time that the task spent sleeping:
- */
- if (unlikely(prof_on == SLEEP_PROFILING)) {
- profile_hits(SLEEP_PROFILING,
- (void *)get_wchan(tsk),
- delta >> 20);
- }
- account_scheduler_latency(tsk, delta >> 10, 0);
- }
- }
-}
-
/*
* Task is being enqueued - update stats:
*/
@@ -1001,10 +883,10 @@ update_stats_enqueue(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)
* a dequeue/enqueue event is a NOP)
*/
if (se != cfs_rq->curr)
- update_stats_wait_start(cfs_rq, se);
+ update_stats_wait_start(rq_of(cfs_rq), se);
if (flags & ENQUEUE_WAKEUP)
- update_stats_enqueue_sleeper(cfs_rq, se);
+ update_stats_enqueue_sleeper(rq_of(cfs_rq), se);
}
static inline void
@@ -1019,7 +901,7 @@ update_stats_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)
* waiting task:
*/
if (se != cfs_rq->curr)
- update_stats_wait_end(cfs_rq, se);
+ update_stats_wait_end(rq_of(cfs_rq), se);
if ((flags & DEQUEUE_SLEEP) && entity_is_task(se)) {
struct task_struct *tsk = task_of(se);
@@ -4128,26 +4010,6 @@ place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial)
static void check_enqueue_throttle(struct cfs_rq *cfs_rq);
-static inline void check_schedstat_required(void)
-{
-#ifdef CONFIG_SCHEDSTATS
- if (schedstat_enabled())
- return;
-
- /* Force schedstat enabled if a dependent tracepoint is active */
- if (trace_sched_stat_wait_enabled() ||
- trace_sched_stat_sleep_enabled() ||
- trace_sched_stat_iowait_enabled() ||
- trace_sched_stat_blocked_enabled() ||
- trace_sched_stat_runtime_enabled()) {
- printk_deferred_once("Scheduler tracepoints stat_sleep, stat_iowait, "
- "stat_blocked and stat_runtime require the "
- "kernel parameter schedstats=enable or "
- "kernel.sched_schedstats=1\n");
- }
-#endif
-}
-
static inline bool cfs_bandwidth_used(void);
/*
@@ -4388,7 +4250,7 @@ set_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *se)
* a CPU. So account for the time it spent waiting on the
* runqueue.
*/
- update_stats_wait_end(cfs_rq, se);
+ update_stats_wait_end(rq_of(cfs_rq), se);
__dequeue_entity(cfs_rq, se);
update_load_avg(cfs_rq, se, UPDATE_TG);
}
@@ -4489,7 +4351,7 @@ static void put_prev_entity(struct cfs_rq *cfs_rq, struct sched_entity *prev)
check_spread(cfs_rq, prev);
if (prev->on_rq) {
- update_stats_wait_start(cfs_rq, prev);
+ update_stats_wait_start(rq_of(cfs_rq), prev);
/* Put 'current' back into the tree. */
__enqueue_entity(cfs_rq, prev);
/* in !on_rq case, update occurred at dequeue */
diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h
index c23b653ffc53..966cc408bd8b 100644
--- a/kernel/sched/stats.h
+++ b/kernel/sched/stats.h
@@ -42,6 +42,144 @@ rq_sched_info_dequeued(struct rq *rq, unsigned long long delta)
#define schedstat_val(var) (var)
#define schedstat_val_or_zero(var) ((schedstat_enabled()) ? (var) : 0)
+static inline void
+update_stats_wait_start(struct rq *rq, struct sched_entity *se)
+{
+ u64 wait_start, prev_wait_start;
+
+ if (!schedstat_enabled())
+ return;
+
+ wait_start = rq_clock(rq);
+ prev_wait_start = schedstat_val(se->statistics.wait_start);
+
+ if (entity_is_task(se) && task_on_rq_migrating(task_of(se)) &&
+ likely(wait_start > prev_wait_start))
+ wait_start -= prev_wait_start;
+
+ __schedstat_set(se->statistics.wait_start, wait_start);
+}
+
+static inline void
+update_stats_wait_end(struct rq *rq, struct sched_entity *se)
+{
+ struct task_struct *p;
+ u64 delta;
+
+ if (!schedstat_enabled())
+ return;
+
+ delta = rq_clock(rq) - schedstat_val(se->statistics.wait_start);
+
+ if (entity_is_task(se)) {
+ p = task_of(se);
+ if (task_on_rq_migrating(p)) {
+ /*
+ * Preserve migrating task's wait time so wait_start
+ * time stamp can be adjusted to accumulate wait time
+ * prior to migration.
+ */
+ __schedstat_set(se->statistics.wait_start, delta);
+ return;
+ }
+ trace_sched_stat_wait(p, delta);
+ }
+
+ __schedstat_set(se->statistics.wait_max,
+ max(schedstat_val(se->statistics.wait_max), delta));
+ __schedstat_inc(se->statistics.wait_count);
+ __schedstat_add(se->statistics.wait_sum, delta);
+ __schedstat_set(se->statistics.wait_start, 0);
+}
+
+static inline void
+update_stats_enqueue_sleeper(struct rq *rq, struct sched_entity *se)
+{
+ struct task_struct *tsk = NULL;
+ u64 sleep_start, block_start;
+
+ if (!schedstat_enabled())
+ return;
+
+ sleep_start = schedstat_val(se->statistics.sleep_start);
+ block_start = schedstat_val(se->statistics.block_start);
+
+ if (entity_is_task(se))
+ tsk = task_of(se);
+
+ if (sleep_start) {
+ u64 delta = rq_clock(rq) - sleep_start;
+
+ if ((s64)delta < 0)
+ delta = 0;
+
+ if (unlikely(delta > schedstat_val(se->statistics.sleep_max)))
+ __schedstat_set(se->statistics.sleep_max, delta);
+
+ __schedstat_set(se->statistics.sleep_start, 0);
+ __schedstat_add(se->statistics.sum_sleep_runtime, delta);
+
+ if (tsk) {
+ account_scheduler_latency(tsk, delta >> 10, 1);
+ trace_sched_stat_sleep(tsk, delta);
+ }
+ }
+
+ if (block_start) {
+ u64 delta = rq_clock(rq) - block_start;
+
+ if ((s64)delta < 0)
+ delta = 0;
+
+ if (unlikely(delta > schedstat_val(se->statistics.block_max)))
+ __schedstat_set(se->statistics.block_max, delta);
+
+ __schedstat_set(se->statistics.block_start, 0);
+ __schedstat_add(se->statistics.sum_sleep_runtime, delta);
+
+ if (tsk) {
+ if (tsk->in_iowait) {
+ __schedstat_add(se->statistics.iowait_sum, delta);
+ __schedstat_inc(se->statistics.iowait_count);
+ trace_sched_stat_iowait(tsk, delta);
+ }
+
+ trace_sched_stat_blocked(tsk, delta);
+
+ /*
+ * Blocking time is in units of nanosecs, so shift by
+ * 20 to get a milliseconds-range estimation of the
+ * amount of time that the task spent sleeping:
+ */
+ if (unlikely(prof_on == SLEEP_PROFILING)) {
+ profile_hits(SLEEP_PROFILING,
+ (void *)get_wchan(tsk),
+ delta >> 20);
+ }
+ account_scheduler_latency(tsk, delta >> 10, 0);
+ }
+ }
+}
+
+static inline void
+check_schedstat_required(void)
+{
+ if (schedstat_enabled())
+ return;
+
+ /* Force schedstat enabled if a dependent tracepoint is active */
+ if (trace_sched_stat_wait_enabled() ||
+ trace_sched_stat_sleep_enabled() ||
+ trace_sched_stat_iowait_enabled() ||
+ trace_sched_stat_blocked_enabled() ||
+ trace_sched_stat_runtime_enabled()) {
+ printk_deferred_once("Scheduler tracepoints stat_sleep, stat_iowait, "
+ "stat_blocked and stat_runtime require the "
+ "kernel parameter schedstats=enable or "
+ "kernel.sched_schedstats=1\n");
+ }
+}
+
#else /* !CONFIG_SCHEDSTATS: */
static inline void rq_sched_info_arrive (struct rq *rq, unsigned long long delta) { }
static inline void rq_sched_info_dequeued(struct rq *rq, unsigned long long delta) { }
@@ -55,6 +193,12 @@ static inline void rq_sched_info_depart (struct rq *rq, unsigned long long delt
# define schedstat_set(var, val) do { } while (0)
# define schedstat_val(var) 0
# define schedstat_val_or_zero(var) 0
+
+# define update_stats_wait_start(rq, se) do { } while (0)
+# define update_stats_wait_end(rq, se) do { } while (0)
+# define update_stats_enqueue_sleeper(rq, se) do { } while (0)
+# define check_schedstat_required() do { } while (0)
+
#endif /* CONFIG_SCHEDSTATS */
#ifdef CONFIG_PSI
--
2.18.4
On Tue, Nov 24, 2020 at 7:40 PM Mel Gorman <[email protected]> wrote:
>
> On Mon, Nov 23, 2020 at 08:58:06PM +0800, Yafang Shao wrote:
> > The 'cfs_rq' in these helpers
> > update_stats_{wait_start, wait_end, enqueue_sleeper} is only used to get
> > the rq_clock, so we can pass the rq directly. Then these helpers can be
> > used by all sched class after being moved into stats.h.
> >
> > After that change, the size of vmlinux is increased around 824Bytes.
> > w/o this patch, with this patch
> > Size of vmlinux: 78443832 78444656
> >
> > Signed-off-by: Yafang Shao <[email protected]>
>
> The inline helpers are quite large. When I was suggesting that the overhead
> was minimal, what I expected what that the inline functions would be a
> schedstat_enabled() followed by a real function call. It would introduce
> a small additional overhead when schedstats are enabled but avoid vmlinux
> growing too large
>
> e.g.
>
> static inline void
> update_stats_wait_start(struct cfs_rq *cfs_rq, struct sched_entity *se)
> {
> if (!schedstat_enabled())
> return;
>
> __update_stats_wait_start(cfs_rq, se);
> }
>
> where __update_stats_wait_start then lives in kernel/sched/stats.c
>
Good idea! Now I understand what you mean. Thanks for the detailed explanation.
I will update it in the next version.
--
Thanks
Yafang
On Mon, Nov 23, 2020 at 08:58:06PM +0800, Yafang Shao wrote:
> The 'cfs_rq' in these helpers
> update_stats_{wait_start, wait_end, enqueue_sleeper} is only used to get
> the rq_clock, so we can pass the rq directly. Then these helpers can be
> used by all sched class after being moved into stats.h.
>
> After that change, the size of vmlinux is increased around 824Bytes.
> w/o this patch, with this patch
> Size of vmlinux: 78443832 78444656
>
> Signed-off-by: Yafang Shao <[email protected]>
The inline helpers are quite large. When I was suggesting that the overhead
was minimal, what I expected what that the inline functions would be a
schedstat_enabled() followed by a real function call. It would introduce
a small additional overhead when schedstats are enabled but avoid vmlinux
growing too large
e.g.
static inline void
update_stats_wait_start(struct cfs_rq *cfs_rq, struct sched_entity *se)
{
if (!schedstat_enabled())
return;
__update_stats_wait_start(cfs_rq, se);
}
where __update_stats_wait_start then lives in kernel/sched/stats.c
--
Mel Gorman
SUSE Labs