[linux-block.git] / include / linux / sched / topology.h

/* SPDX-License-Identifier: GPL-2.0 */
#ifndef _LINUX_SCHED_TOPOLOGY_H
#define _LINUX_SCHED_TOPOLOGY_H

#include <linux/topology.h>

#include <linux/sched/idle.h>

/*
 * sched-domains (multiprocessor balancing) declarations:
 */
#ifdef CONFIG_SMP

#define SD_LOAD_BALANCE		0x0001	/* Do load balancing on this domain. */
#define SD_BALANCE_NEWIDLE	0x0002	/* Balance when about to become idle */
#define SD_BALANCE_EXEC		0x0004	/* Balance on exec */
#define SD_BALANCE_FORK		0x0008	/* Balance on fork, clone */
#define SD_BALANCE_WAKE		0x0010  /* Balance on wakeup */
#define SD_WAKE_AFFINE		0x0020	/* Wake task to waking CPU */
#define SD_ASYM_CPUCAPACITY	0x0040  /* Groups have different max cpu capacities */
#define SD_SHARE_CPUCAPACITY	0x0080	/* Domain members share cpu capacity */
#define SD_SHARE_POWERDOMAIN	0x0100	/* Domain members share power domain */
#define SD_SHARE_PKG_RESOURCES	0x0200	/* Domain members share cpu pkg resources */
#define SD_SERIALIZE		0x0400	/* Only a single load balancing instance */
#define SD_ASYM_PACKING		0x0800  /* Place busy groups earlier in the domain */
#define SD_PREFER_SIBLING	0x1000	/* Prefer to place tasks in a sibling domain */
#define SD_OVERLAP		0x2000	/* sched_domains of this level overlap */
#define SD_NUMA			0x4000	/* cross-node balancing */

/*
 * Increase resolution of cpu_capacity calculations
 */
#define SCHED_CAPACITY_SHIFT	SCHED_FIXEDPOINT_SHIFT
#define SCHED_CAPACITY_SCALE	(1L << SCHED_CAPACITY_SHIFT)

#ifdef CONFIG_SCHED_SMT
static inline int cpu_smt_flags(void)
{
	return SD_SHARE_CPUCAPACITY | SD_SHARE_PKG_RESOURCES;
}
#endif

#ifdef CONFIG_SCHED_MC
static inline int cpu_core_flags(void)
{
	return SD_SHARE_PKG_RESOURCES;
}
#endif

#ifdef CONFIG_NUMA
static inline int cpu_numa_flags(void)
{
	return SD_NUMA;
}
#endif

extern int arch_asym_cpu_priority(int cpu);

struct sched_domain_attr {
	int relax_domain_level;
};

#define SD_ATTR_INIT	(struct sched_domain_attr) {	\
	.relax_domain_level = -1,			\
}

extern int sched_domain_level_max;

struct sched_group;

struct sched_domain_shared {
	atomic_t	ref;
	atomic_t	nr_busy_cpus;
	int		has_idle_cores;
};

struct sched_domain {
	/* These fields must be setup */
	struct sched_domain *parent;	/* top domain must be null terminated */
	struct sched_domain *child;	/* bottom domain must be null terminated */
	struct sched_group *groups;	/* the balancing groups of the domain */
	unsigned long min_interval;	/* Minimum balance interval ms */
	unsigned long max_interval;	/* Maximum balance interval ms */
	unsigned int busy_factor;	/* less balancing by factor if busy */
	unsigned int imbalance_pct;	/* No balance until over watermark */
	unsigned int cache_nice_tries;	/* Leave cache hot tasks for # tries */
	unsigned int busy_idx;
	unsigned int idle_idx;
	unsigned int newidle_idx;
	unsigned int wake_idx;
	unsigned int forkexec_idx;
	unsigned int smt_gain;

	int nohz_idle;			/* NOHZ IDLE status */
	int flags;			/* See SD_* */
	int level;

	/* Runtime fields. */
	unsigned long last_balance;	/* init to jiffies. units in jiffies */
	unsigned int balance_interval;	/* initialise to 1. units in ms. */
	unsigned int nr_balance_failed; /* initialise to 0 */

	/* idle_balance() stats */
	u64 max_newidle_lb_cost;
	unsigned long next_decay_max_lb_cost;

	u64 avg_scan_cost;		/* select_idle_sibling */

#ifdef CONFIG_SCHEDSTATS
	/* load_balance() stats */
	unsigned int lb_count[CPU_MAX_IDLE_TYPES];
	unsigned int lb_failed[CPU_MAX_IDLE_TYPES];
	unsigned int lb_balanced[CPU_MAX_IDLE_TYPES];
	unsigned int lb_imbalance[CPU_MAX_IDLE_TYPES];
	unsigned int lb_gained[CPU_MAX_IDLE_TYPES];
	unsigned int lb_hot_gained[CPU_MAX_IDLE_TYPES];
	unsigned int lb_nobusyg[CPU_MAX_IDLE_TYPES];
	unsigned int lb_nobusyq[CPU_MAX_IDLE_TYPES];

	/* Active load balancing */
	unsigned int alb_count;
	unsigned int alb_failed;
	unsigned int alb_pushed;

	/* SD_BALANCE_EXEC stats */
	unsigned int sbe_count;
	unsigned int sbe_balanced;
	unsigned int sbe_pushed;

	/* SD_BALANCE_FORK stats */
	unsigned int sbf_count;
	unsigned int sbf_balanced;
	unsigned int sbf_pushed;

	/* try_to_wake_up() stats */
	unsigned int ttwu_wake_remote;
	unsigned int ttwu_move_affine;
	unsigned int ttwu_move_balance;
#endif
#ifdef CONFIG_SCHED_DEBUG
	char *name;
#endif
	union {
		void *private;		/* used during construction */
		struct rcu_head rcu;	/* used during destruction */
	};
	struct sched_domain_shared *shared;

	unsigned int span_weight;
	/*
	 * Span of all CPUs in this domain.
	 *
	 * NOTE: this field is variable length. (Allocated dynamically
	 * by attaching extra space to the end of the structure,
	 * depending on how many CPUs the kernel has booted up with)
	 */
	unsigned long span[0];
};

static inline struct cpumask *sched_domain_span(struct sched_domain *sd)
{
	return to_cpumask(sd->span);
}

extern void partition_sched_domains(int ndoms_new, cpumask_var_t doms_new[],
				    struct sched_domain_attr *dattr_new);

/* Allocate an array of sched domains, for partition_sched_domains(). */
cpumask_var_t *alloc_sched_domains(unsigned int ndoms);
void free_sched_domains(cpumask_var_t doms[], unsigned int ndoms);

bool cpus_share_cache(int this_cpu, int that_cpu);

typedef const struct cpumask *(*sched_domain_mask_f)(int cpu);
typedef int (*sched_domain_flags_f)(void);

#define SDTL_OVERLAP	0x01

struct sd_data {
	struct sched_domain **__percpu sd;
	struct sched_domain_shared **__percpu sds;
	struct sched_group **__percpu sg;
	struct sched_group_capacity **__percpu sgc;
};

struct sched_domain_topology_level {
	sched_domain_mask_f mask;
	sched_domain_flags_f sd_flags;
	int		    flags;
	int		    numa_level;
	struct sd_data      data;
#ifdef CONFIG_SCHED_DEBUG
	char                *name;
#endif
};

extern void set_sched_topology(struct sched_domain_topology_level *tl);

#ifdef CONFIG_SCHED_DEBUG
# define SD_INIT_NAME(type)		.name = #type
#else
# define SD_INIT_NAME(type)
#endif

#else /* CONFIG_SMP */

struct sched_domain_attr;

static inline void
partition_sched_domains(int ndoms_new, cpumask_var_t doms_new[],
			struct sched_domain_attr *dattr_new)
{
}

static inline bool cpus_share_cache(int this_cpu, int that_cpu)
{
	return true;
}

#endif	/* !CONFIG_SMP */

static inline int task_node(const struct task_struct *p)
{
	return cpu_to_node(task_cpu(p));
}

#endif /* _LINUX_SCHED_TOPOLOGY_H */
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
105ab3d8 IM	2	#ifndef _LINUX_SCHED_TOPOLOGY_H
	3	#define _LINUX_SCHED_TOPOLOGY_H
	4
ee6a3d19 IM	5	#include <linux/topology.h>
ee6a3d19 IM	6
4c822698 IM	7	#include <linux/sched/idle.h>
4c822698 IM	8
a60b9eda IM	9	/*
	10	* sched-domains (multiprocessor balancing) declarations:
	11	*/
	12	#ifdef CONFIG_SMP
	13
	14	#define SD_LOAD_BALANCE 0x0001 /* Do load balancing on this domain. */
	15	#define SD_BALANCE_NEWIDLE 0x0002 /* Balance when about to become idle */
	16	#define SD_BALANCE_EXEC 0x0004 /* Balance on exec */
	17	#define SD_BALANCE_FORK 0x0008 /* Balance on fork, clone */
	18	#define SD_BALANCE_WAKE 0x0010 /* Balance on wakeup */
	19	#define SD_WAKE_AFFINE 0x0020 /* Wake task to waking CPU */
	20	#define SD_ASYM_CPUCAPACITY 0x0040 /* Groups have different max cpu capacities */
	21	#define SD_SHARE_CPUCAPACITY 0x0080 /* Domain members share cpu capacity */
	22	#define SD_SHARE_POWERDOMAIN 0x0100 /* Domain members share power domain */
	23	#define SD_SHARE_PKG_RESOURCES 0x0200 /* Domain members share cpu pkg resources */
	24	#define SD_SERIALIZE 0x0400 /* Only a single load balancing instance */
	25	#define SD_ASYM_PACKING 0x0800 /* Place busy groups earlier in the domain */
	26	#define SD_PREFER_SIBLING 0x1000 /* Prefer to place tasks in a sibling domain */
	27	#define SD_OVERLAP 0x2000 /* sched_domains of this level overlap */
	28	#define SD_NUMA 0x4000 /* cross-node balancing */
	29
	30	/*
	31	* Increase resolution of cpu_capacity calculations
	32	*/
	33	#define SCHED_CAPACITY_SHIFT SCHED_FIXEDPOINT_SHIFT
	34	#define SCHED_CAPACITY_SCALE (1L << SCHED_CAPACITY_SHIFT)
	35
	36	#ifdef CONFIG_SCHED_SMT
	37	static inline int cpu_smt_flags(void)
	38	{
	39	return SD_SHARE_CPUCAPACITY \| SD_SHARE_PKG_RESOURCES;
	40	}
	41	#endif
	42
	43	#ifdef CONFIG_SCHED_MC
	44	static inline int cpu_core_flags(void)
	45	{
	46	return SD_SHARE_PKG_RESOURCES;
	47	}
	48	#endif
	49
	50	#ifdef CONFIG_NUMA
	51	static inline int cpu_numa_flags(void)
	52	{
	53	return SD_NUMA;
	54	}
	55	#endif
	56
	57	extern int arch_asym_cpu_priority(int cpu);
	58
	59	struct sched_domain_attr {
	60	int relax_domain_level;
	61	};
	62
	63	#define SD_ATTR_INIT (struct sched_domain_attr) { \
	64	.relax_domain_level = -1, \
	65	}
	66
	67	extern int sched_domain_level_max;
	68
	69	struct sched_group;
	70
	71	struct sched_domain_shared {
	72	atomic_t ref;
73	atomic_t nr_busy_cpus;
74	int has_idle_cores;
75	};
76
77	struct sched_domain {
78	/* These fields must be setup */
79	struct sched_domain parent; / top domain must be null terminated */
80	struct sched_domain child; / bottom domain must be null terminated */
81	struct sched_group groups; / the balancing groups of the domain */
82	unsigned long min_interval; /* Minimum balance interval ms */
83	unsigned long max_interval; /* Maximum balance interval ms */
84	unsigned int busy_factor; /* less balancing by factor if busy */
85	unsigned int imbalance_pct; /* No balance until over watermark */
86	unsigned int cache_nice_tries; /* Leave cache hot tasks for # tries */
87	unsigned int busy_idx;
88	unsigned int idle_idx;
89	unsigned int newidle_idx;
90	unsigned int wake_idx;
91	unsigned int forkexec_idx;
92	unsigned int smt_gain;
93
94	int nohz_idle; /* NOHZ IDLE status */
95	int flags; /* See SD_* */
96	int level;
97
98	/* Runtime fields. */
99	unsigned long last_balance; /* init to jiffies. units in jiffies */
100	unsigned int balance_interval; /* initialise to 1. units in ms. */
101	unsigned int nr_balance_failed; /* initialise to 0 */
102
103	/* idle_balance() stats */
104	u64 max_newidle_lb_cost;
105	unsigned long next_decay_max_lb_cost;
106
107	u64 avg_scan_cost; /* select_idle_sibling */
108
109	#ifdef CONFIG_SCHEDSTATS
110	/* load_balance() stats */
111	unsigned int lb_count[CPU_MAX_IDLE_TYPES];
112	unsigned int lb_failed[CPU_MAX_IDLE_TYPES];
113	unsigned int lb_balanced[CPU_MAX_IDLE_TYPES];
114	unsigned int lb_imbalance[CPU_MAX_IDLE_TYPES];
115	unsigned int lb_gained[CPU_MAX_IDLE_TYPES];
116	unsigned int lb_hot_gained[CPU_MAX_IDLE_TYPES];
117	unsigned int lb_nobusyg[CPU_MAX_IDLE_TYPES];
118	unsigned int lb_nobusyq[CPU_MAX_IDLE_TYPES];
119
120	/* Active load balancing */
121	unsigned int alb_count;
122	unsigned int alb_failed;
123	unsigned int alb_pushed;
124
125	/* SD_BALANCE_EXEC stats */
126	unsigned int sbe_count;
127	unsigned int sbe_balanced;
128	unsigned int sbe_pushed;
129
130	/* SD_BALANCE_FORK stats */
131	unsigned int sbf_count;
132	unsigned int sbf_balanced;
133	unsigned int sbf_pushed;
134
135	/* try_to_wake_up() stats */
136	unsigned int ttwu_wake_remote;
137	unsigned int ttwu_move_affine;
138	unsigned int ttwu_move_balance;
139	#endif
140	#ifdef CONFIG_SCHED_DEBUG
141	char *name;
142	#endif
143	union {
144	void private; / used during construction */
145	struct rcu_head rcu; /* used during destruction */
146	};
147	struct sched_domain_shared *shared;
148
149	unsigned int span_weight;
150	/*
151	* Span of all CPUs in this domain.
152	*
153	* NOTE: this field is variable length. (Allocated dynamically
154	* by attaching extra space to the end of the structure,
155	* depending on how many CPUs the kernel has booted up with)
156	*/
157	unsigned long span[0];
158	};
159
160	static inline struct cpumask sched_domain_span(struct sched_domain sd)
161	{
162	return to_cpumask(sd->span);
163	}
164
165	extern void partition_sched_domains(int ndoms_new, cpumask_var_t doms_new[],
166	struct sched_domain_attr *dattr_new);
167
168	/* Allocate an array of sched domains, for partition_sched_domains(). */
169	cpumask_var_t *alloc_sched_domains(unsigned int ndoms);
170	void free_sched_domains(cpumask_var_t doms[], unsigned int ndoms);
171
172	bool cpus_share_cache(int this_cpu, int that_cpu);
173
174	typedef const struct cpumask (sched_domain_mask_f)(int cpu);
175	typedef int (*sched_domain_flags_f)(void);
176
177	#define SDTL_OVERLAP 0x01
178
179	struct sd_data {
180	struct sched_domain **__percpu sd;
181	struct sched_domain_shared **__percpu sds;
182	struct sched_group **__percpu sg;
183	struct sched_group_capacity **__percpu sgc;
184	};
185
186	struct sched_domain_topology_level {
187	sched_domain_mask_f mask;
188	sched_domain_flags_f sd_flags;
189	int flags;
190	int numa_level;
191	struct sd_data data;
192	#ifdef CONFIG_SCHED_DEBUG
193	char *name;
194	#endif
195	};
196
197	extern void set_sched_topology(struct sched_domain_topology_level *tl);
198
199	#ifdef CONFIG_SCHED_DEBUG
200	# define SD_INIT_NAME(type) .name = #type
201	#else
202	# define SD_INIT_NAME(type)
203	#endif
204
205	#else /* CONFIG_SMP */
206
207	struct sched_domain_attr;
208
209	static inline void
210	partition_sched_domains(int ndoms_new, cpumask_var_t doms_new[],
211	struct sched_domain_attr *dattr_new)
212	{
213	}
214
215	static inline bool cpus_share_cache(int this_cpu, int that_cpu)
216	{
217	return true;
218	}
219
220	#endif /* !CONFIG_SMP */
221
ee6a3d19 IM	222	static inline int task_node(const struct task_struct *p)
	223	{
	224	return cpu_to_node(task_cpu(p));
	225	}
	226
105ab3d8	227	#endif /* _LINUX_SCHED_TOPOLOGY_H */