[PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Dhaval Giani]

This patch checks if we can set the rt_runtime_us to 0. If there is a
realtime task in the group, we don't want to set the rt_runtime_us to 0
otherwise bad things will happen.

Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
---
 kernel/sched.c |   19 +++++++++++++++++++
 1 files changed, 19 insertions(+)

Index: linux-2.6.25-rc3/kernel/sched.c
===================================================================
--- linux-2.6.25-rc3.orig/kernel/sched.c
+++ linux-2.6.25-rc3/kernel/sched.c
@@ -7960,6 +7960,17 @@ static int __rt_schedulable(struct task_
 	return total + to_ratio(period, runtime) < global_ratio;
 }
 
+/* Must be called with tasklist_lock held */
+static inline int tg_has_rt_tasks(struct task_group *tg)
+{
+	struct task_struct *p;
+	for_each_process(p) {
+		if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
+			return 1;
+	}
+	return 0;
+}
+
 int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
 {
 	u64 rt_runtime, rt_period;
@@ -7971,6 +7982,13 @@ int sched_group_set_rt_runtime(struct ta
 		rt_runtime = rt_period;
 
 	mutex_lock(&rt_constraints_mutex);
+	read_lock(&tasklist_lock);
+	if (rt_runtime_us == 0) {
+		if (tg_has_rt_tasks(tg)) {
+			err = -EINVAL;
+			goto unlock;
+		}
+	}
 	if (!__rt_schedulable(tg, rt_period, rt_runtime)) {
 		err = -EINVAL;
 		goto unlock;
@@ -7979,6 +7997,7 @@ int sched_group_set_rt_runtime(struct ta
 		rt_runtime = RUNTIME_INF;
 	tg->rt_runtime = rt_runtime;
  unlock:
+	read_unlock(&tasklist_lock);
 	mutex_unlock(&rt_constraints_mutex);
 
 	return err;
-- 
regards,
Dhaval

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Peter Zijlstra]

On Wed, 2008-02-27 at 16:38 +0530, Dhaval Giani wrote:
> This patch checks if we can set the rt_runtime_us to 0. If there is a
> realtime task in the group, we don't want to set the rt_runtime_us to 0
> otherwise bad things will happen.

I had considered this a: don't do that then, thing. But sure, helping
the admin seems like a valid option.

> Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
> ---
>  kernel/sched.c |   19 +++++++++++++++++++
>  1 files changed, 19 insertions(+)
> 
> Index: linux-2.6.25-rc3/kernel/sched.c
> ===================================================================
> --- linux-2.6.25-rc3.orig/kernel/sched.c
> +++ linux-2.6.25-rc3/kernel/sched.c
> @@ -7960,6 +7960,17 @@ static int __rt_schedulable(struct task_
>  	return total + to_ratio(period, runtime) < global_ratio;
>  }
>  
> +/* Must be called with tasklist_lock held */
> +static inline int tg_has_rt_tasks(struct task_group *tg)
> +{
> +	struct task_struct *p;
> +	for_each_process(p) {
> +		if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
> +			return 1;
> +	}
> +	return 0;
> +}
> +
>  int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
>  {
>  	u64 rt_runtime, rt_period;
> @@ -7971,6 +7982,13 @@ int sched_group_set_rt_runtime(struct ta
>  		rt_runtime = rt_period;
>  
>  	mutex_lock(&rt_constraints_mutex);
> +	read_lock(&tasklist_lock);
> +	if (rt_runtime_us == 0) {
> +		if (tg_has_rt_tasks(tg)) {

&&

> +			err = -EINVAL;

-EBUSY perhaps?

> +			goto unlock;
> +		}
> +	}
>  	if (!__rt_schedulable(tg, rt_period, rt_runtime)) {
>  		err = -EINVAL;
>  		goto unlock;
> @@ -7979,6 +7997,7 @@ int sched_group_set_rt_runtime(struct ta
>  		rt_runtime = RUNTIME_INF;
>  	tg->rt_runtime = rt_runtime;
>   unlock:
> +	read_unlock(&tasklist_lock);
>  	mutex_unlock(&rt_constraints_mutex);
>  
>  	return err;

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Dhaval Giani]

On Wed, Feb 27, 2008 at 01:46:56PM +0100, Peter Zijlstra wrote:
> 
> On Wed, 2008-02-27 at 16:38 +0530, Dhaval Giani wrote:
> > This patch checks if we can set the rt_runtime_us to 0. If there is a
> > realtime task in the group, we don't want to set the rt_runtime_us to 0
> > otherwise bad things will happen.
> 
> I had considered this a: don't do that then, thing. But sure, helping
> the admin seems like a valid option.
> 

Well, if it can be done, it shall be done :)

<snip>

> 
> > +			err = -EINVAL;
> 
> -EBUSY perhaps?
>

Well, the group is certainly not busy. I will go with invalid as it is
definitly an  invalid value when the groups have rt tasks. But if someone
disagrees, I have no issue with it being something else.

Corrected patch as follows,

--
This patch checks if we can set the rt_runtime_us to 0. If there is a
realtime task in the group, we don't want to set the rt_runtime_us as 0
or bad things will happen.

Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
---
 kernel/sched.c |   17 +++++++++++++++++
 1 files changed, 17 insertions(+)

Index: linux-2.6.25-rc3/kernel/sched.c
===================================================================
--- linux-2.6.25-rc3.orig/kernel/sched.c
+++ linux-2.6.25-rc3/kernel/sched.c
@@ -7960,6 +7960,17 @@ static int __rt_schedulable(struct task_
 	return total + to_ratio(period, runtime) < global_ratio;
 }
 
+/* Must be called with tasklist_lock held */
+static inline int tg_has_rt_tasks(struct task_group *tg)
+{
+	struct task_struct *p;
+	for_each_process(p) {
+		if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
+			return 1;
+	}
+	return 0;
+}
+
 int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
 {
 	u64 rt_runtime, rt_period;
@@ -7971,6 +7982,11 @@ int sched_group_set_rt_runtime(struct ta
 		rt_runtime = rt_period;
 
 	mutex_lock(&rt_constraints_mutex);
+	read_lock(&tasklist_lock);
+	if (rt_runtime_us == 0 && tg_has_rt_tasks(tg)) {
+		err = -EINVAL;
+		goto unlock;
+	}
 	if (!__rt_schedulable(tg, rt_period, rt_runtime)) {
 		err = -EINVAL;
 		goto unlock;
@@ -7979,6 +7995,7 @@ int sched_group_set_rt_runtime(struct ta
 		rt_runtime = RUNTIME_INF;
 	tg->rt_runtime = rt_runtime;
  unlock:
+	read_unlock(&tasklist_lock);
 	mutex_unlock(&rt_constraints_mutex);
 
 	return err;
-- 
regards,
Dhaval

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Peter Zijlstra]

On Wed, 2008-02-27 at 18:54 +0530, Dhaval Giani wrote:

> Well, if it can be done, it shall be done :)

May I suggest: rm -rf /

/me runs

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Peter Zijlstra]

On Wed, 2008-02-27 at 18:54 +0530, Dhaval Giani wrote:

> This patch checks if we can set the rt_runtime_us to 0. If there is a
> realtime task in the group, we don't want to set the rt_runtime_us as 0
> or bad things will happen.
> 
> Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>

Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>

> ---
>  kernel/sched.c |   17 +++++++++++++++++
>  1 files changed, 17 insertions(+)
> 
> Index: linux-2.6.25-rc3/kernel/sched.c
> ===================================================================
> --- linux-2.6.25-rc3.orig/kernel/sched.c
> +++ linux-2.6.25-rc3/kernel/sched.c
> @@ -7960,6 +7960,17 @@ static int __rt_schedulable(struct task_
>  	return total + to_ratio(period, runtime) < global_ratio;
>  }
>  
> +/* Must be called with tasklist_lock held */
> +static inline int tg_has_rt_tasks(struct task_group *tg)
> +{
> +	struct task_struct *p;
> +	for_each_process(p) {
> +		if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
> +			return 1;
> +	}
> +	return 0;
> +}
> +
>  int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
>  {
>  	u64 rt_runtime, rt_period;
> @@ -7971,6 +7982,11 @@ int sched_group_set_rt_runtime(struct ta
>  		rt_runtime = rt_period;
>  
>  	mutex_lock(&rt_constraints_mutex);
> +	read_lock(&tasklist_lock);
> +	if (rt_runtime_us == 0 && tg_has_rt_tasks(tg)) {
> +		err = -EINVAL;
> +		goto unlock;
> +	}
>  	if (!__rt_schedulable(tg, rt_period, rt_runtime)) {
>  		err = -EINVAL;
>  		goto unlock;
> @@ -7979,6 +7995,7 @@ int sched_group_set_rt_runtime(struct ta
>  		rt_runtime = RUNTIME_INF;
>  	tg->rt_runtime = rt_runtime;
>   unlock:
> +	read_unlock(&tasklist_lock);
>  	mutex_unlock(&rt_constraints_mutex);
>  
>  	return err;

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Ingo Molnar]

* Peter Zijlstra <a.p.zijlstra@chello.nl> wrote:

> 
> On Wed, 2008-02-27 at 18:54 +0530, Dhaval Giani wrote:
> 
> > This patch checks if we can set the rt_runtime_us to 0. If there is a
> > realtime task in the group, we don't want to set the rt_runtime_us as 0
> > or bad things will happen.
> > 
> > Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
> 
> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>

doesnt apply to sched-devel.git - what's the plan here?

	Ingo

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Peter Zijlstra]

On Wed, 2008-02-27 at 16:03 +0100, Ingo Molnar wrote:
> * Peter Zijlstra <a.p.zijlstra@chello.nl> wrote:
> 
> > 
> > On Wed, 2008-02-27 at 18:54 +0530, Dhaval Giani wrote:
> > 
> > > This patch checks if we can set the rt_runtime_us to 0. If there is a
> > > realtime task in the group, we don't want to set the rt_runtime_us as 0
> > > or bad things will happen.
> > > 
> > > Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
> > 
> > Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
> 
> doesnt apply to sched-devel.git - what's the plan here?

stick in front, drop the other two sched-rt-group patches for which I'll
send you an update somewhere today.

These two patches would be candidates to push to linus..

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Dhaval Giani]

On Wed, Feb 27, 2008 at 04:03:50PM +0100, Ingo Molnar wrote:
> 
> * Peter Zijlstra <a.p.zijlstra@chello.nl> wrote:
> 
> > 
> > On Wed, 2008-02-27 at 18:54 +0530, Dhaval Giani wrote:
> > 
> > > This patch checks if we can set the rt_runtime_us to 0. If there is a
> > > realtime task in the group, we don't want to set the rt_runtime_us as 0
> > > or bad things will happen.
> > > 
> > > Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
> > 
> > Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
> 
> doesnt apply to sched-devel.git - what's the plan here?
> 

Belongs to mainline.

-- 
regards,
Dhaval

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Peter Zijlstra]

On Wed, 2008-02-27 at 18:54 +0530, Dhaval Giani wrote:

> This patch checks if we can set the rt_runtime_us to 0. If there is a
> realtime task in the group, we don't want to set the rt_runtime_us as 0
> or bad things will happen.
> 
> Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
> ---
>  kernel/sched.c |   17 +++++++++++++++++
>  1 files changed, 17 insertions(+)
> 
> Index: linux-2.6.25-rc3/kernel/sched.c
> ===================================================================
> --- linux-2.6.25-rc3.orig/kernel/sched.c
> +++ linux-2.6.25-rc3/kernel/sched.c
> @@ -7960,6 +7960,17 @@ static int __rt_schedulable(struct task_
>  	return total + to_ratio(period, runtime) < global_ratio;
>  }
>  
> +/* Must be called with tasklist_lock held */
> +static inline int tg_has_rt_tasks(struct task_group *tg)
> +{
> +	struct task_struct *p;
> +	for_each_process(p) {
> +		if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
> +			return 1;
> +	}

This should be a do_each_thread() { } while_each_thread(); iteration
loop;

cgroups are thread oriented, not task.

Sorry for missing this earlier.

> +	return 0;
> +}
> +
>  int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
>  {
>  	u64 rt_runtime, rt_period;
> @@ -7971,6 +7982,11 @@ int sched_group_set_rt_runtime(struct ta
>  		rt_runtime = rt_period;
>  
>  	mutex_lock(&rt_constraints_mutex);
> +	read_lock(&tasklist_lock);
> +	if (rt_runtime_us == 0 && tg_has_rt_tasks(tg)) {
> +		err = -EINVAL;
> +		goto unlock;
> +	}
>  	if (!__rt_schedulable(tg, rt_period, rt_runtime)) {
>  		err = -EINVAL;
>  		goto unlock;
> @@ -7979,6 +7995,7 @@ int sched_group_set_rt_runtime(struct ta
>  		rt_runtime = RUNTIME_INF;
>  	tg->rt_runtime = rt_runtime;
>   unlock:
> +	read_unlock(&tasklist_lock);
>  	mutex_unlock(&rt_constraints_mutex);
>  
>  	return err;

Re: [PATCH] sched: don't allow rt_runtime_us to be zero for groups having rt tasks

[Dhaval Giani]

On Wed, Feb 27, 2008 at 10:47:19PM +0100, Peter Zijlstra wrote:
> 
> On Wed, 2008-02-27 at 18:54 +0530, Dhaval Giani wrote:
> > +	struct task_struct *p;
> > +	for_each_process(p) {
> > +		if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
> > +			return 1;
> > +	}
> 
> This should be a do_each_thread() { } while_each_thread(); iteration
> loop;
> 
> cgroups are thread oriented, not task.
> 
> Sorry for missing this earlier.
> 

Nice catch! Sorry for the mistake, corrected patch follows.
---
This patch checks if we can set the rt_runtime_us to 0. If there is a
realtime task in the group, we don't want to set the rt_runtime_us as 0
or bad things will happen.

Signed-off-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
---
 kernel/sched.c |   17 +++++++++++++++++
 1 files changed, 17 insertions(+)

Index: linux-2.6.25-rc3/kernel/sched.c
===================================================================
--- linux-2.6.25-rc3.orig/kernel/sched.c
+++ linux-2.6.25-rc3/kernel/sched.c
@@ -7960,6 +7960,17 @@ static int __rt_schedulable(struct task_
 	return total + to_ratio(period, runtime) < global_ratio;
 }
 
+/* Must be called with tasklist_lock held */
+static inline int tg_has_rt_tasks(struct task_group *tg)
+{
+	struct task_struct *g, *p;
+	do_each_thread(g, p) {
+		if (rt_task(p) && rt_rq_of_se(&p->rt)->tg == tg)
+			return 1;
+	} while_each_thread(g, p);
+	return 0;
+}
+
 int sched_group_set_rt_runtime(struct task_group *tg, long rt_runtime_us)
 {
 	u64 rt_runtime, rt_period;
@@ -7971,6 +7982,11 @@ int sched_group_set_rt_runtime(struct ta
 		rt_runtime = rt_period;
 
 	mutex_lock(&rt_constraints_mutex);
+	read_lock(&tasklist_lock);
+	if (rt_runtime_us == 0 && tg_has_rt_tasks(tg)) {
+		err = -EINVAL;
+		goto unlock;
+	}
 	if (!__rt_schedulable(tg, rt_period, rt_runtime)) {
 		err = -EINVAL;
 		goto unlock;
@@ -7979,6 +7995,7 @@ int sched_group_set_rt_runtime(struct ta
 		rt_runtime = RUNTIME_INF;
 	tg->rt_runtime = rt_runtime;
  unlock:
+	read_unlock(&tasklist_lock);
 	mutex_unlock(&rt_constraints_mutex);
 
 	return err;
-- 
regards,
Dhaval