[tip: sched/core] psi: Optimize task switch inside shared cgroups
The following commit has been merged into the sched/core branch of tip:
Commit-ID: 4117cebf1a9fcbf35b9aabf0e37b6c5eea296798
Gitweb:
https://git.kernel.org/tip/4117cebf1a9fcbf35b9aabf0e37b6c5eea296798
Author:Chengming Zhou
AuthorDate:Wed, 03 Mar 2021 11:46:59 +08:00
Committer: Ingo Molnar
CommitterDate: Sat, 06 Mar 2021 12:40:23 +01:00
psi: Optimize task switch inside shared cgroups
The commit 36b238d57172 ("psi: Optimize switching tasks inside shared
cgroups") only update cgroups whose state actually changes during a
task switch only in task preempt case, not in task sleep case.
We actually don't need to clear and set TSK_ONCPU state for common cgroups
of next and prev task in sleep case, that can save many psi_group_change
especially when most activity comes from one leaf cgroup.
sleep before:
psi_dequeue()
while ((group = iterate_groups(prev))) # all ancestors
psi_group_change(prev, .clear=TSK_RUNNING|TSK_ONCPU)
psi_task_switch()
while ((group = iterate_groups(next))) # all ancestors
psi_group_change(next, .set=TSK_ONCPU)
sleep after:
psi_dequeue()
nop
psi_task_switch()
while ((group = iterate_groups(next))) # until (prev & next)
psi_group_change(next, .set=TSK_ONCPU)
while ((group = iterate_groups(prev))) # all ancestors
psi_group_change(prev, .clear=common?TSK_RUNNING:TSK_RUNNING|TSK_ONCPU)
When a voluntary sleep switches to another task, we remove one call of
psi_group_change() for every common cgroup ancestor of the two tasks.
Co-developed-by: Muchun Song
Signed-off-by: Muchun Song
Signed-off-by: Chengming Zhou
Signed-off-by: Peter Zijlstra (Intel)
Signed-off-by: Ingo Molnar
Acked-by: Johannes Weiner
Link:
https://lkml.kernel.org/r/20210303034659.91735-5-zhouchengm...@bytedance.com
---
kernel/sched/psi.c | 35 +--
kernel/sched/stats.h | 28
2 files changed, 37 insertions(+), 26 deletions(-)
diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
index 3907a6b..ee3c5b4 100644
--- a/kernel/sched/psi.c
+++ b/kernel/sched/psi.c
@@ -840,20 +840,35 @@ void psi_task_switch(struct task_struct *prev, struct
task_struct *next,
}
}
- /*
-* If this is a voluntary sleep, dequeue will have taken care
-* of the outgoing TSK_ONCPU alongside TSK_RUNNING already. We
-* only need to deal with it during preemption.
-*/
- if (sleep)
- return;
-
if (prev->pid) {
- psi_flags_change(prev, TSK_ONCPU, 0);
+ int clear = TSK_ONCPU, set = 0;
+
+ /*
+* When we're going to sleep, psi_dequeue() lets us handle
+* TSK_RUNNING and TSK_IOWAIT here, where we can combine it
+* with TSK_ONCPU and save walking common ancestors twice.
+*/
+ if (sleep) {
+ clear |= TSK_RUNNING;
+ if (prev->in_iowait)
+ set |= TSK_IOWAIT;
+ }
+
+ psi_flags_change(prev, clear, set);
iter = NULL;
while ((group = iterate_groups(prev, )) && group != common)
- psi_group_change(group, cpu, TSK_ONCPU, 0, true);
+ psi_group_change(group, cpu, clear, set, true);
+
+ /*
+* TSK_ONCPU is handled up to the common ancestor. If we're
tasked
+* with dequeuing too, finish that for the rest of the
hierarchy.
+*/
+ if (sleep) {
+ clear &= ~TSK_ONCPU;
+ for (; group; group = iterate_groups(prev, ))
+ psi_group_change(group, cpu, clear, set, true);
+ }
}
}
diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h
index 9e4e67a..dc218e9 100644
--- a/kernel/sched/stats.h
+++ b/kernel/sched/stats.h
@@ -84,28 +84,24 @@ static inline void psi_enqueue(struct task_struct *p, bool
wakeup)
static inline void psi_dequeue(struct task_struct *p, bool sleep)
{
- int clear = TSK_RUNNING, set = 0;
+ int clear = TSK_RUNNING;
if (static_branch_likely(_disabled))
return;
- if (!sleep) {
- if (p->in_memstall)
- clear |= TSK_MEMSTALL;
- } else {
- /*
-* When a task sleeps, schedule() dequeues it before
-* switching to the next one. Merge the clearing of
-* TSK_RUNNING and TSK_ONCPU to save an unnecessary
-* psi_task_change() call in psi_sched_switch().
-*/
- clear |= TSK_ONCPU;
+ /*
+* A voluntary sleep is a dequeue followed by a task switch. To
+* avoid walking all ancestors twice, psi_task_switch() handles
+* TSK_RUNNING and TSK_IOWAIT for us when it moves TSK_ONCPU.
+*
[tip: sched/core] psi: Optimize task switch inside shared cgroups
The following commit has been merged into the sched/core branch of tip:
Commit-ID: e6560d58334ca463061ade733674abc8dd0df9bd
Gitweb:
https://git.kernel.org/tip/e6560d58334ca463061ade733674abc8dd0df9bd
Author:Chengming Zhou
AuthorDate:Wed, 03 Mar 2021 11:46:59 +08:00
Committer: Peter Zijlstra
CommitterDate: Thu, 04 Mar 2021 09:56:02 +01:00
psi: Optimize task switch inside shared cgroups
The commit 36b238d57172 ("psi: Optimize switching tasks inside shared
cgroups") only update cgroups whose state actually changes during a
task switch only in task preempt case, not in task sleep case.
We actually don't need to clear and set TSK_ONCPU state for common cgroups
of next and prev task in sleep case, that can save many psi_group_change
especially when most activity comes from one leaf cgroup.
sleep before:
psi_dequeue()
while ((group = iterate_groups(prev))) # all ancestors
psi_group_change(prev, .clear=TSK_RUNNING|TSK_ONCPU)
psi_task_switch()
while ((group = iterate_groups(next))) # all ancestors
psi_group_change(next, .set=TSK_ONCPU)
sleep after:
psi_dequeue()
nop
psi_task_switch()
while ((group = iterate_groups(next))) # until (prev & next)
psi_group_change(next, .set=TSK_ONCPU)
while ((group = iterate_groups(prev))) # all ancestors
psi_group_change(prev, .clear=common?TSK_RUNNING:TSK_RUNNING|TSK_ONCPU)
When a voluntary sleep switches to another task, we remove one call of
psi_group_change() for every common cgroup ancestor of the two tasks.
Co-developed-by: Muchun Song
Signed-off-by: Muchun Song
Signed-off-by: Chengming Zhou
Signed-off-by: Peter Zijlstra (Intel)
Acked-by: Johannes Weiner
Link:
https://lkml.kernel.org/r/20210303034659.91735-5-zhouchengm...@bytedance.com
---
kernel/sched/psi.c | 35 +--
kernel/sched/stats.h | 28
2 files changed, 37 insertions(+), 26 deletions(-)
diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
index 3907a6b..ee3c5b4 100644
--- a/kernel/sched/psi.c
+++ b/kernel/sched/psi.c
@@ -840,20 +840,35 @@ void psi_task_switch(struct task_struct *prev, struct
task_struct *next,
}
}
- /*
-* If this is a voluntary sleep, dequeue will have taken care
-* of the outgoing TSK_ONCPU alongside TSK_RUNNING already. We
-* only need to deal with it during preemption.
-*/
- if (sleep)
- return;
-
if (prev->pid) {
- psi_flags_change(prev, TSK_ONCPU, 0);
+ int clear = TSK_ONCPU, set = 0;
+
+ /*
+* When we're going to sleep, psi_dequeue() lets us handle
+* TSK_RUNNING and TSK_IOWAIT here, where we can combine it
+* with TSK_ONCPU and save walking common ancestors twice.
+*/
+ if (sleep) {
+ clear |= TSK_RUNNING;
+ if (prev->in_iowait)
+ set |= TSK_IOWAIT;
+ }
+
+ psi_flags_change(prev, clear, set);
iter = NULL;
while ((group = iterate_groups(prev, )) && group != common)
- psi_group_change(group, cpu, TSK_ONCPU, 0, true);
+ psi_group_change(group, cpu, clear, set, true);
+
+ /*
+* TSK_ONCPU is handled up to the common ancestor. If we're
tasked
+* with dequeuing too, finish that for the rest of the
hierarchy.
+*/
+ if (sleep) {
+ clear &= ~TSK_ONCPU;
+ for (; group; group = iterate_groups(prev, ))
+ psi_group_change(group, cpu, clear, set, true);
+ }
}
}
diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h
index 9e4e67a..dc218e9 100644
--- a/kernel/sched/stats.h
+++ b/kernel/sched/stats.h
@@ -84,28 +84,24 @@ static inline void psi_enqueue(struct task_struct *p, bool
wakeup)
static inline void psi_dequeue(struct task_struct *p, bool sleep)
{
- int clear = TSK_RUNNING, set = 0;
+ int clear = TSK_RUNNING;
if (static_branch_likely(_disabled))
return;
- if (!sleep) {
- if (p->in_memstall)
- clear |= TSK_MEMSTALL;
- } else {
- /*
-* When a task sleeps, schedule() dequeues it before
-* switching to the next one. Merge the clearing of
-* TSK_RUNNING and TSK_ONCPU to save an unnecessary
-* psi_task_change() call in psi_sched_switch().
-*/
- clear |= TSK_ONCPU;
+ /*
+* A voluntary sleep is a dequeue followed by a task switch. To
+* avoid walking all ancestors twice, psi_task_switch() handles
+* TSK_RUNNING and TSK_IOWAIT for us when it moves TSK_ONCPU.
+* Do nothing here.
+
