2024-02-26 21:12:52

by levi.yun

Subject: [PATCH v2] psi: Fix avg trigger being fired unexpectedly.

commit 915a087e4c473("psi: Fix trigger being fired unexpectedly at initial")
fixes unexpected event fired when group->total accumulated for PSI_POLL.
But, for PSI_AVGS, win->value should be initialized with group->total[PSI_AVGS].
Moreover, to get exact initial value for win->value, it should be set
under each trigger lock to avoid concurrent changes to group->total[].

Signed-off-by: Levi Yun <[email protected]>
Acked-by: Suren Baghdasaryan <[email protected]>
---
v2: Fix commit message.
---
kernel/sched/psi.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
index 7b4aa5809c0f..e7f66ab2ad3e 100644
--- a/kernel/sched/psi.c
+++ b/kernel/sched/psi.c
@@ -1323,9 +1323,6 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
t->state = state;
t->threshold = threshold_us * NSEC_PER_USEC;
t->win.size = window_us * NSEC_PER_USEC;
- window_reset(&t->win, sched_clock(),
- group->total[PSI_POLL][t->state], 0);
-
t->event = 0;
t->last_event_time = 0;
t->of = of;
@@ -1336,6 +1333,8 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,

if (privileged) {
mutex_lock(&group->rtpoll_trigger_lock);
+ window_reset(&t->win, sched_clock(),
+ group->total[PSI_POLL][t->state], 0);

if (!rcu_access_pointer(group->rtpoll_task)) {
struct task_struct *task;
@@ -1361,6 +1360,9 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
} else {
mutex_lock(&group->avgs_lock);

+ window_reset(&t->win, sched_clock(),
+ group->total[PSI_AVGS][t->state], 0);
+
list_add(&t->node, &group->avg_triggers);
group->avg_nr_triggers[t->state]++;

--
2.39.2

2024-03-12 06:10:51

by levi.yun

Subject: Re: [PATCH v2] psi: Fix avg trigger being fired unexpectedly.

Gentle ping..

On Mon, Feb 26, 2024 at 09:12:37PM +0000, Levi Yun wrote:
> commit 915a087e4c473("psi: Fix trigger being fired unexpectedly at initial")
> fixes unexpected event fired when group->total accumulated for PSI_POLL.
> But, for PSI_AVGS, win->value should be initialized with group->total[PSI_AVGS].
> Moreover, to get exact initial value for win->value, it should be set
> under each trigger lock to avoid concurrent changes to group->total[].
>
> Signed-off-by: Levi Yun <[email protected]>
> Acked-by: Suren Baghdasaryan <[email protected]>
> ---
> v2: Fix commit message.
> ---
> kernel/sched/psi.c | 8 +++++---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
> index 7b4aa5809c0f..e7f66ab2ad3e 100644
> --- a/kernel/sched/psi.c
> +++ b/kernel/sched/psi.c
> @@ -1323,9 +1323,6 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> t->state = state;
> t->threshold = threshold_us * NSEC_PER_USEC;
> t->win.size = window_us * NSEC_PER_USEC;
> - window_reset(&t->win, sched_clock(),
> - group->total[PSI_POLL][t->state], 0);
> -
> t->event = 0;
> t->last_event_time = 0;
> t->of = of;
> @@ -1336,6 +1333,8 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
>
> if (privileged) {
> mutex_lock(&group->rtpoll_trigger_lock);
> + window_reset(&t->win, sched_clock(),
> + group->total[PSI_POLL][t->state], 0);
>
> if (!rcu_access_pointer(group->rtpoll_task)) {
> struct task_struct *task;
> @@ -1361,6 +1360,9 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> } else {
> mutex_lock(&group->avgs_lock);
>
> + window_reset(&t->win, sched_clock(),
> + group->total[PSI_AVGS][t->state], 0);
> +
> list_add(&t->node, &group->avg_triggers);
> group->avg_nr_triggers[t->state]++;
>
> --
> 2.39.2

2024-03-26 17:50:10

by Suren Baghdasaryan

Subject: Re: [PATCH v2] psi: Fix avg trigger being fired unexpectedly.

On Tue, Mar 12, 2024 at 6:10 AM levi.yun <[email protected]> wrote:
>
> Gentle ping..
>
> On Mon, Feb 26, 2024 at 09:12:37PM +0000, Levi Yun wrote:
> > commit 915a087e4c473("psi: Fix trigger being fired unexpectedly at initial")
> > fixes unexpected event fired when group->total accumulated for PSI_POLL.
> > But, for PSI_AVGS, win->value should be initialized with group->total[PSI_AVGS].
> > Moreover, to get exact initial value for win->value, it should be set
> > under each trigger lock to avoid concurrent changes to group->total[].
> >
> > Signed-off-by: Levi Yun <[email protected]>
> > Acked-by: Suren Baghdasaryan <[email protected]>

Hi Peter,
Could you please pull this patch into your tree? I don't see it in any
tree, so I think you missed it.
Thanks,
Suren.

> > ---
> > v2: Fix commit message.
> > ---
> > kernel/sched/psi.c | 8 +++++---
> > 1 file changed, 5 insertions(+), 3 deletions(-)
> >
> > diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
> > index 7b4aa5809c0f..e7f66ab2ad3e 100644
> > --- a/kernel/sched/psi.c
> > +++ b/kernel/sched/psi.c
> > @@ -1323,9 +1323,6 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> > t->state = state;
> > t->threshold = threshold_us * NSEC_PER_USEC;
> > t->win.size = window_us * NSEC_PER_USEC;
> > - window_reset(&t->win, sched_clock(),
> > - group->total[PSI_POLL][t->state], 0);
> > -
> > t->event = 0;
> > t->last_event_time = 0;
> > t->of = of;
> > @@ -1336,6 +1333,8 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> >
> > if (privileged) {
> > mutex_lock(&group->rtpoll_trigger_lock);
> > + window_reset(&t->win, sched_clock(),
> > + group->total[PSI_POLL][t->state], 0);
> >
> > if (!rcu_access_pointer(group->rtpoll_task)) {
> > struct task_struct *task;
> > @@ -1361,6 +1360,9 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> > } else {
> > mutex_lock(&group->avgs_lock);
> >
> > + window_reset(&t->win, sched_clock(),
> > + group->total[PSI_AVGS][t->state], 0);
> > +
> > list_add(&t->node, &group->avg_triggers);
> > group->avg_nr_triggers[t->state]++;
> >
> > --
> > 2.39.2

2024-03-26 18:01:30

by Suren Baghdasaryan

Subject: Re: [PATCH v2] psi: Fix avg trigger being fired unexpectedly.

On Tue, Mar 26, 2024 at 10:49 AM Suren Baghdasaryan <[email protected]> wrote:
>
> On Tue, Mar 12, 2024 at 6:10 AM levi.yun <[email protected]> wrote:
> >
> > Gentle ping..
> >
> > On Mon, Feb 26, 2024 at 09:12:37PM +0000, Levi Yun wrote:
> > > commit 915a087e4c473("psi: Fix trigger being fired unexpectedly at initial")
> > > fixes unexpected event fired when group->total accumulated for PSI_POLL.
> > > But, for PSI_AVGS, win->value should be initialized with group->total[PSI_AVGS].
> > > Moreover, to get exact initial value for win->value, it should be set
> > > under each trigger lock to avoid concurrent changes to group->total[].
> > >
> > > Signed-off-by: Levi Yun <[email protected]>
> > > Acked-by: Suren Baghdasaryan <[email protected]>
>
> Hi Peter,
> Could you please pull this patch into your tree? I don't see it in any
> tree, so I think you missed it.

Forgot to set the right recipient. Fixed now.


> Thanks,
> Suren.
>
> > > ---
> > > v2: Fix commit message.
> > > ---
> > > kernel/sched/psi.c | 8 +++++---
> > > 1 file changed, 5 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
> > > index 7b4aa5809c0f..e7f66ab2ad3e 100644
> > > --- a/kernel/sched/psi.c
> > > +++ b/kernel/sched/psi.c
> > > @@ -1323,9 +1323,6 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> > > t->state = state;
> > > t->threshold = threshold_us * NSEC_PER_USEC;
> > > t->win.size = window_us * NSEC_PER_USEC;
> > > - window_reset(&t->win, sched_clock(),
> > > - group->total[PSI_POLL][t->state], 0);
> > > -
> > > t->event = 0;
> > > t->last_event_time = 0;
> > > t->of = of;
> > > @@ -1336,6 +1333,8 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> > >
> > > if (privileged) {
> > > mutex_lock(&group->rtpoll_trigger_lock);
> > > + window_reset(&t->win, sched_clock(),
> > > + group->total[PSI_POLL][t->state], 0);
> > >
> > > if (!rcu_access_pointer(group->rtpoll_task)) {
> > > struct task_struct *task;
> > > @@ -1361,6 +1360,9 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> > > } else {
> > > mutex_lock(&group->avgs_lock);
> > >
> > > + window_reset(&t->win, sched_clock(),
> > > + group->total[PSI_AVGS][t->state], 0);
> > > +
> > > list_add(&t->node, &group->avg_triggers);
> > > group->avg_nr_triggers[t->state]++;
> > >
> > > --
> > > 2.39.2

2024-04-18 17:15:44

by Suren Baghdasaryan

Subject: Re: [PATCH v2] psi: Fix avg trigger being fired unexpectedly.

On Tue, Mar 26, 2024 at 10:49 AM Suren Baghdasaryan <[email protected]> wrote:
>
> On Tue, Mar 12, 2024 at 6:10 AM levi.yun <[email protected]> wrote:
> >
> > Gentle ping..
> >
> > On Mon, Feb 26, 2024 at 09:12:37PM +0000, Levi Yun wrote:
> > > commit 915a087e4c473("psi: Fix trigger being fired unexpectedly at initial")
> > > fixes unexpected event fired when group->total accumulated for PSI_POLL.
> > > But, for PSI_AVGS, win->value should be initialized with group->total[PSI_AVGS].
> > > Moreover, to get exact initial value for win->value, it should be set
> > > under each trigger lock to avoid concurrent changes to group->total[].
> > >
> > > Signed-off-by: Levi Yun <[email protected]>
> > > Acked-by: Suren Baghdasaryan <[email protected]>
>
> Hi Peter,
> Could you please pull this patch into your tree? I don't see it in any
> tree, so I think you missed it.

I heard Peter is currently unavailable. Ingo, would you please take
this patch into the scheduler tree?
Thanks,
Suren.

> Thanks,
> Suren.
>
> > > ---
> > > v2: Fix commit message.
> > > ---
> > > kernel/sched/psi.c | 8 +++++---
> > > 1 file changed, 5 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/kernel/sched/psi.c b/kernel/sched/psi.c
> > > index 7b4aa5809c0f..e7f66ab2ad3e 100644
> > > --- a/kernel/sched/psi.c
> > > +++ b/kernel/sched/psi.c
> > > @@ -1323,9 +1323,6 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> > > t->state = state;
> > > t->threshold = threshold_us * NSEC_PER_USEC;
> > > t->win.size = window_us * NSEC_PER_USEC;
> > > - window_reset(&t->win, sched_clock(),
> > > - group->total[PSI_POLL][t->state], 0);
> > > -
> > > t->event = 0;
> > > t->last_event_time = 0;
> > > t->of = of;
> > > @@ -1336,6 +1333,8 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> > >
> > > if (privileged) {
> > > mutex_lock(&group->rtpoll_trigger_lock);
> > > + window_reset(&t->win, sched_clock(),
> > > + group->total[PSI_POLL][t->state], 0);
> > >
> > > if (!rcu_access_pointer(group->rtpoll_task)) {
> > > struct task_struct *task;
> > > @@ -1361,6 +1360,9 @@ struct psi_trigger *psi_trigger_create(struct psi_group *group, char *buf,
> > > } else {
> > > mutex_lock(&group->avgs_lock);
> > >
> > > + window_reset(&t->win, sched_clock(),
> > > + group->total[PSI_AVGS][t->state], 0);
> > > +
> > > list_add(&t->node, &group->avg_triggers);
> > > group->avg_nr_triggers[t->state]++;
> > >
> > > --
> > > 2.39.2