On 07-Nov 15:55, Peter Zijlstra wrote:
> On Wed, Nov 07, 2018 at 02:48:09PM +, Patrick Bellasi wrote:
> > On 07-Nov 14:35, Peter Zijlstra wrote:
> > You mean se_count overflow ?
>
> Yah..
>
> > > And I'm not really a fan of hiding that error in a define like you keep
> > > doing.
> >
> >
On 07-Nov 15:55, Peter Zijlstra wrote:
> On Wed, Nov 07, 2018 at 02:48:09PM +, Patrick Bellasi wrote:
> > On 07-Nov 14:35, Peter Zijlstra wrote:
> > You mean se_count overflow ?
>
> Yah..
>
> > > And I'm not really a fan of hiding that error in a define like you keep
> > > doing.
> >
> >
On 07-Nov 15:44, Peter Zijlstra wrote:
> On Wed, Nov 07, 2018 at 02:24:28PM +, Patrick Bellasi wrote:
> > On 07-Nov 14:44, Peter Zijlstra wrote:
> > > On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
>
> > > > +static void uclamp_group_get(struct uclamp_se *uc_se, unsigned int
On 07-Nov 15:44, Peter Zijlstra wrote:
> On Wed, Nov 07, 2018 at 02:24:28PM +, Patrick Bellasi wrote:
> > On 07-Nov 14:44, Peter Zijlstra wrote:
> > > On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
>
> > > > +static void uclamp_group_get(struct uclamp_se *uc_se, unsigned int
On 07-Nov 15:42, Peter Zijlstra wrote:
> On Wed, Nov 07, 2018 at 02:19:49PM +, Patrick Bellasi wrote:
> > On 07-Nov 13:19, Peter Zijlstra wrote:
> > > On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > > > +struct uclamp_se {
> > > > + unsigned int value :
On 07-Nov 15:42, Peter Zijlstra wrote:
> On Wed, Nov 07, 2018 at 02:19:49PM +, Patrick Bellasi wrote:
> > On 07-Nov 13:19, Peter Zijlstra wrote:
> > > On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > > > +struct uclamp_se {
> > > > + unsigned int value :
On Wed, Nov 07, 2018 at 02:48:09PM +, Patrick Bellasi wrote:
> On 07-Nov 14:35, Peter Zijlstra wrote:
> You mean se_count overflow ?
Yah..
> > And I'm not really a fan of hiding that error in a define like you keep
> > doing.
>
> The #define is not there to mask an overflow, it's there to
On Wed, Nov 07, 2018 at 02:48:09PM +, Patrick Bellasi wrote:
> On 07-Nov 14:35, Peter Zijlstra wrote:
> You mean se_count overflow ?
Yah..
> > And I'm not really a fan of hiding that error in a define like you keep
> > doing.
>
> The #define is not there to mask an overflow, it's there to
On 07-Nov 15:37, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > +static int __setscheduler_uclamp(struct task_struct *p,
> > +const struct sched_attr *attr)
> > +{
> > + unsigned int lower_bound =
On 07-Nov 15:37, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > +static int __setscheduler_uclamp(struct task_struct *p,
> > +const struct sched_attr *attr)
> > +{
> > + unsigned int lower_bound =
On 07-Nov 14:35, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > +/**
> > + * uclamp_group_get: increase the reference count for a clamp group
> > + * @uc_se: the utilization clamp data for the task
> > + * @clamp_id: the clamp index affected by the
On 07-Nov 14:35, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > +/**
> > + * uclamp_group_get: increase the reference count for a clamp group
> > + * @uc_se: the utilization clamp data for the task
> > + * @clamp_id: the clamp index affected by the
On Wed, Nov 07, 2018 at 02:24:28PM +, Patrick Bellasi wrote:
> On 07-Nov 14:44, Peter Zijlstra wrote:
> > On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > > +static void uclamp_group_get(struct uclamp_se *uc_se, unsigned int
> > > clamp_id,
> > > +
On Wed, Nov 07, 2018 at 02:24:28PM +, Patrick Bellasi wrote:
> On 07-Nov 14:44, Peter Zijlstra wrote:
> > On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > > +static void uclamp_group_get(struct uclamp_se *uc_se, unsigned int
> > > clamp_id,
> > > +
On Wed, Nov 07, 2018 at 02:19:49PM +, Patrick Bellasi wrote:
> On 07-Nov 13:19, Peter Zijlstra wrote:
> > On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > > +struct uclamp_se {
> > > + unsigned int value : SCHED_CAPACITY_SHIFT + 1;
> > > + unsigned int group_id
On Wed, Nov 07, 2018 at 02:19:49PM +, Patrick Bellasi wrote:
> On 07-Nov 13:19, Peter Zijlstra wrote:
> > On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > > +struct uclamp_se {
> > > + unsigned int value : SCHED_CAPACITY_SHIFT + 1;
> > > + unsigned int group_id
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +static int __setscheduler_uclamp(struct task_struct *p,
> + const struct sched_attr *attr)
> +{
> + unsigned int lower_bound = p->uclamp[UCLAMP_MIN].value;
> + unsigned int upper_bound =
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +static int __setscheduler_uclamp(struct task_struct *p,
> + const struct sched_attr *attr)
> +{
> + unsigned int lower_bound = p->uclamp[UCLAMP_MIN].value;
> + unsigned int upper_bound =
On 07-Nov 14:44, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > +/**
> > + * uclamp_group_get: increase the reference count for a clamp group
> > + * @uc_se: the utilization clamp data for the task
> > + * @clamp_id: the clamp index affected by the
On 07-Nov 14:44, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > +/**
> > + * uclamp_group_get: increase the reference count for a clamp group
> > + * @uc_se: the utilization clamp data for the task
> > + * @clamp_id: the clamp index affected by the
On 07-Nov 13:19, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > +struct uclamp_se {
> > + unsigned int value : SCHED_CAPACITY_SHIFT + 1;
> > + unsigned int group_id : order_base_2(UCLAMP_GROUPS);
>
> Are you sure about ob2()
On 07-Nov 13:19, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> > +struct uclamp_se {
> > + unsigned int value : SCHED_CAPACITY_SHIFT + 1;
> > + unsigned int group_id : order_base_2(UCLAMP_GROUPS);
>
> Are you sure about ob2()
On Wed, Nov 07, 2018 at 01:57:38PM +, Patrick Bellasi wrote:
> On 07-Nov 14:16, Peter Zijlstra wrote:
> > Please write cmpxchg loops in the form:
> >
> > atomic_long_t *ptr = _maps[clamp_id][group_id].adata;
> > union uclamp_map old, new;
> >
> > old.data =
On Wed, Nov 07, 2018 at 01:57:38PM +, Patrick Bellasi wrote:
> On 07-Nov 14:16, Peter Zijlstra wrote:
> > Please write cmpxchg loops in the form:
> >
> > atomic_long_t *ptr = _maps[clamp_id][group_id].adata;
> > union uclamp_map old, new;
> >
> > old.data =
On 07-Nov 14:16, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
>
> > +static void uclamp_group_put(unsigned int clamp_id, unsigned int group_id)
> > {
> > + union uclamp_map *uc_maps = _maps[clamp_id][0];
> > + union uclamp_map uc_map_old,
On 07-Nov 14:16, Peter Zijlstra wrote:
> On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
>
> > +static void uclamp_group_put(unsigned int clamp_id, unsigned int group_id)
> > {
> > + union uclamp_map *uc_maps = _maps[clamp_id][0];
> > + union uclamp_map uc_map_old,
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +/**
> + * uclamp_group_get: increase the reference count for a clamp group
> + * @uc_se: the utilization clamp data for the task
> + * @clamp_id: the clamp index affected by the task
> + * @clamp_value: the new clamp value for the
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +/**
> + * uclamp_group_get: increase the reference count for a clamp group
> + * @uc_se: the utilization clamp data for the task
> + * @clamp_id: the clamp index affected by the task
> + * @clamp_value: the new clamp value for the
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +/**
> + * uclamp_group_get: increase the reference count for a clamp group
> + * @uc_se: the utilization clamp data for the task
> + * @clamp_id: the clamp index affected by the task
> + * @clamp_value: the new clamp value for the
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +/**
> + * uclamp_group_get: increase the reference count for a clamp group
> + * @uc_se: the utilization clamp data for the task
> + * @clamp_id: the clamp index affected by the task
> + * @clamp_value: the new clamp value for the
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +static void uclamp_group_put(unsigned int clamp_id, unsigned int group_id)
> {
> + union uclamp_map *uc_maps = _maps[clamp_id][0];
> + union uclamp_map uc_map_old, uc_map_new;
> + long res;
> +
> +retry:
> +
> +
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +static void uclamp_group_put(unsigned int clamp_id, unsigned int group_id)
> {
> + union uclamp_map *uc_maps = _maps[clamp_id][0];
> + union uclamp_map uc_map_old, uc_map_new;
> + long res;
> +
> +retry:
> +
> +
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +struct uclamp_se {
> + unsigned int value : SCHED_CAPACITY_SHIFT + 1;
> + unsigned int group_id : order_base_2(UCLAMP_GROUPS);
Are you sure about ob2() ? seems weird we'll end up with 0 for
On Mon, Oct 29, 2018 at 06:32:57PM +, Patrick Bellasi wrote:
> +struct uclamp_se {
> + unsigned int value : SCHED_CAPACITY_SHIFT + 1;
> + unsigned int group_id : order_base_2(UCLAMP_GROUPS);
Are you sure about ob2() ? seems weird we'll end up with 0 for
Utilization clamping requires each CPU to know which clamp values are
assigned to tasks that are currently RUNNABLE on that CPU: multiple
tasks can be assigned the same clamp value and tasks with different
clamp values can be concurrently active on the same CPU.
A proper data structure is
Utilization clamping requires each CPU to know which clamp values are
assigned to tasks that are currently RUNNABLE on that CPU: multiple
tasks can be assigned the same clamp value and tasks with different
clamp values can be concurrently active on the same CPU.
A proper data structure is
36 matches
Mail list logo
