[deliverable/linux.git] / kernel / rcutree.h

/*
 * Read-Copy Update mechanism for mutual exclusion (tree-based version)
 * Internal non-public definitions.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
 *
 * Copyright IBM Corporation, 2008
 *
 * Author: Ingo Molnar <mingo@elte.hu>
 *	   Paul E. McKenney <paulmck@linux.vnet.ibm.com>
 */

#include <linux/cache.h>
#include <linux/spinlock.h>
#include <linux/threads.h>
#include <linux/cpumask.h>
#include <linux/seqlock.h>

/*
 * Define shape of hierarchy based on NR_CPUS and CONFIG_RCU_FANOUT.
 * In theory, it should be possible to add more levels straightforwardly.
 * In practice, this did work well going from three levels to four.
 * Of course, your mileage may vary.
 */
#define MAX_RCU_LVLS 4
#if CONFIG_RCU_FANOUT > 16
#define RCU_FANOUT_LEAF       16
#else /* #if CONFIG_RCU_FANOUT > 16 */
#define RCU_FANOUT_LEAF       (CONFIG_RCU_FANOUT)
#endif /* #else #if CONFIG_RCU_FANOUT > 16 */
#define RCU_FANOUT_1	      (RCU_FANOUT_LEAF)
#define RCU_FANOUT_2	      (RCU_FANOUT_1 * CONFIG_RCU_FANOUT)
#define RCU_FANOUT_3	      (RCU_FANOUT_2 * CONFIG_RCU_FANOUT)
#define RCU_FANOUT_4	      (RCU_FANOUT_3 * CONFIG_RCU_FANOUT)

#if NR_CPUS <= RCU_FANOUT_1
#  define NUM_RCU_LVLS	      1
#  define NUM_RCU_LVL_0	      1
#  define NUM_RCU_LVL_1	      (NR_CPUS)
#  define NUM_RCU_LVL_2	      0
#  define NUM_RCU_LVL_3	      0
#  define NUM_RCU_LVL_4	      0
#elif NR_CPUS <= RCU_FANOUT_2
#  define NUM_RCU_LVLS	      2
#  define NUM_RCU_LVL_0	      1
#  define NUM_RCU_LVL_1	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
#  define NUM_RCU_LVL_2	      (NR_CPUS)
#  define NUM_RCU_LVL_3	      0
#  define NUM_RCU_LVL_4	      0
#elif NR_CPUS <= RCU_FANOUT_3
#  define NUM_RCU_LVLS	      3
#  define NUM_RCU_LVL_0	      1
#  define NUM_RCU_LVL_1	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_2)
#  define NUM_RCU_LVL_2	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
#  define NUM_RCU_LVL_3	      (NR_CPUS)
#  define NUM_RCU_LVL_4	      0
#elif NR_CPUS <= RCU_FANOUT_4
#  define NUM_RCU_LVLS	      4
#  define NUM_RCU_LVL_0	      1
#  define NUM_RCU_LVL_1	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_3)
#  define NUM_RCU_LVL_2	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_2)
#  define NUM_RCU_LVL_3	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
#  define NUM_RCU_LVL_4	      (NR_CPUS)
#else
# error "CONFIG_RCU_FANOUT insufficient for NR_CPUS"
#endif /* #if (NR_CPUS) <= RCU_FANOUT_1 */

#define RCU_SUM (NUM_RCU_LVL_0 + NUM_RCU_LVL_1 + NUM_RCU_LVL_2 + NUM_RCU_LVL_3 + NUM_RCU_LVL_4)
#define NUM_RCU_NODES (RCU_SUM - NR_CPUS)

/*
 * Dynticks per-CPU state.
 */
struct rcu_dynticks {
	int dynticks_nesting;	/* Track irq/process nesting level. */
	int dynticks_nmi_nesting; /* Track NMI nesting level. */
	atomic_t dynticks;	/* Even value for dynticks-idle, else odd. */
};

/*
 * Definition for node within the RCU grace-period-detection hierarchy.
 */
struct rcu_node {
	raw_spinlock_t lock;	/* Root rcu_node's lock protects some */
				/*  rcu_state fields as well as following. */
	unsigned long gpnum;	/* Current grace period for this node. */
				/*  This will either be equal to or one */
				/*  behind the root rcu_node's gpnum. */
	unsigned long completed; /* Last GP completed for this node. */
				/*  This will either be equal to or one */
				/*  behind the root rcu_node's gpnum. */
	unsigned long qsmask;	/* CPUs or groups that need to switch in */
				/*  order for current grace period to proceed.*/
				/*  In leaf rcu_node, each bit corresponds to */
				/*  an rcu_data structure, otherwise, each */
				/*  bit corresponds to a child rcu_node */
				/*  structure. */
	unsigned long expmask;	/* Groups that have ->blkd_tasks */
				/*  elements that need to drain to allow the */
				/*  current expedited grace period to */
				/*  complete (only for TREE_PREEMPT_RCU). */
	unsigned long wakemask; /* CPUs whose kthread needs to be awakened. */
	unsigned long qsmaskinit;
				/* Per-GP initial value for qsmask & expmask. */
	unsigned long grpmask;	/* Mask to apply to parent qsmask. */
				/*  Only one bit will be set in this mask. */
	int	grplo;		/* lowest-numbered CPU or group here. */
	int	grphi;		/* highest-numbered CPU or group here. */
	u8	grpnum;		/* CPU/group number for next level up. */
	u8	level;		/* root is at level 0. */
	struct rcu_node *parent;
	struct list_head blkd_tasks;
				/* Tasks blocked in RCU read-side critical */
				/*  section.  Tasks are placed at the head */
				/*  of this list and age towards the tail. */
	struct list_head *gp_tasks;
				/* Pointer to the first task blocking the */
				/*  current grace period, or NULL if there */
				/*  is no such task. */
	struct list_head *exp_tasks;
				/* Pointer to the first task blocking the */
				/*  current expedited grace period, or NULL */
				/*  if there is no such task.  If there */
				/*  is no current expedited grace period, */
				/*  then there can cannot be any such task. */
#ifdef CONFIG_RCU_BOOST
	struct list_head *boost_tasks;
				/* Pointer to first task that needs to be */
				/*  priority boosted, or NULL if no priority */
				/*  boosting is needed for this rcu_node */
				/*  structure.  If there are no tasks */
				/*  queued on this rcu_node structure that */
				/*  are blocking the current grace period, */
				/*  there can be no such task. */
	unsigned long boost_time;
				/* When to start boosting (jiffies). */
	struct task_struct *boost_kthread_task;
				/* kthread that takes care of priority */
				/*  boosting for this rcu_node structure. */
	wait_queue_head_t boost_wq;
				/* Wait queue on which to park the boost */
				/*  kthread. */
#endif /* #ifdef CONFIG_RCU_BOOST */
	struct task_struct *node_kthread_task;
				/* kthread that takes care of this rcu_node */
				/*  structure, for example, awakening the */
				/*  per-CPU kthreads as needed. */
	wait_queue_head_t node_wq;
				/* Wait queue on which to park the per-node */
				/*  kthread. */
} ____cacheline_internodealigned_in_smp;

/*
 * Do a full breadth-first scan of the rcu_node structures for the
 * specified rcu_state structure.
 */
#define rcu_for_each_node_breadth_first(rsp, rnp) \
	for ((rnp) = &(rsp)->node[0]; \
	     (rnp) < &(rsp)->node[NUM_RCU_NODES]; (rnp)++)

/*
 * Do a breadth-first scan of the non-leaf rcu_node structures for the
 * specified rcu_state structure.  Note that if there is a singleton
 * rcu_node tree with but one rcu_node structure, this loop is a no-op.
 */
#define rcu_for_each_nonleaf_node_breadth_first(rsp, rnp) \
	for ((rnp) = &(rsp)->node[0]; \
	     (rnp) < (rsp)->level[NUM_RCU_LVLS - 1]; (rnp)++)

/*
 * Scan the leaves of the rcu_node hierarchy for the specified rcu_state
 * structure.  Note that if there is a singleton rcu_node tree with but
 * one rcu_node structure, this loop -will- visit the rcu_node structure.
 * It is still a leaf node, even if it is also the root node.
 */
#define rcu_for_each_leaf_node(rsp, rnp) \
	for ((rnp) = (rsp)->level[NUM_RCU_LVLS - 1]; \
	     (rnp) < &(rsp)->node[NUM_RCU_NODES]; (rnp)++)

/* Index values for nxttail array in struct rcu_data. */
#define RCU_DONE_TAIL		0	/* Also RCU_WAIT head. */
#define RCU_WAIT_TAIL		1	/* Also RCU_NEXT_READY head. */
#define RCU_NEXT_READY_TAIL	2	/* Also RCU_NEXT head. */
#define RCU_NEXT_TAIL		3
#define RCU_NEXT_SIZE		4

/* Per-CPU data for read-copy update. */
struct rcu_data {
	/* 1) quiescent-state and grace-period handling : */
	unsigned long	completed;	/* Track rsp->completed gp number */
					/*  in order to detect GP end. */
	unsigned long	gpnum;		/* Highest gp number that this CPU */
					/*  is aware of having started. */
	unsigned long	passed_quiesc_completed;
					/* Value of completed at time of qs. */
	bool		passed_quiesc;	/* User-mode/idle loop etc. */
	bool		qs_pending;	/* Core waits for quiesc state. */
	bool		beenonline;	/* CPU online at least once. */
	bool		preemptable;	/* Preemptable RCU? */
	struct rcu_node *mynode;	/* This CPU's leaf of hierarchy */
	unsigned long grpmask;		/* Mask to apply to leaf qsmask. */

	/* 2) batch handling */
	/*
	 * If nxtlist is not NULL, it is partitioned as follows.
	 * Any of the partitions might be empty, in which case the
	 * pointer to that partition will be equal to the pointer for
	 * the following partition.  When the list is empty, all of
	 * the nxttail elements point to the ->nxtlist pointer itself,
	 * which in that case is NULL.
	 *
	 * [nxtlist, *nxttail[RCU_DONE_TAIL]):
	 *	Entries that batch # <= ->completed
	 *	The grace period for these entries has completed, and
	 *	the other grace-period-completed entries may be moved
	 *	here temporarily in rcu_process_callbacks().
	 * [*nxttail[RCU_DONE_TAIL], *nxttail[RCU_WAIT_TAIL]):
	 *	Entries that batch # <= ->completed - 1: waiting for current GP
	 * [*nxttail[RCU_WAIT_TAIL], *nxttail[RCU_NEXT_READY_TAIL]):
	 *	Entries known to have arrived before current GP ended
	 * [*nxttail[RCU_NEXT_READY_TAIL], *nxttail[RCU_NEXT_TAIL]):
	 *	Entries that might have arrived after current GP ended
	 *	Note that the value of *nxttail[RCU_NEXT_TAIL] will
	 *	always be NULL, as this is the end of the list.
	 */
	struct rcu_head *nxtlist;
	struct rcu_head **nxttail[RCU_NEXT_SIZE];
	long		qlen;		/* # of queued callbacks */
	long		qlen_last_fqs_check;
					/* qlen at last check for QS forcing */
	unsigned long	n_cbs_invoked;	/* count of RCU cbs invoked. */
	unsigned long   n_cbs_orphaned; /* RCU cbs orphaned by dying CPU */
	unsigned long   n_cbs_adopted;  /* RCU cbs adopted from dying CPU */
	unsigned long	n_force_qs_snap;
					/* did other CPU force QS recently? */
	long		blimit;		/* Upper limit on a processed batch */

#ifdef CONFIG_NO_HZ
	/* 3) dynticks interface. */
	struct rcu_dynticks *dynticks;	/* Shared per-CPU dynticks state. */
	int dynticks_snap;		/* Per-GP tracking for dynticks. */
#endif /* #ifdef CONFIG_NO_HZ */

	/* 4) reasons this CPU needed to be kicked by force_quiescent_state */
#ifdef CONFIG_NO_HZ
	unsigned long dynticks_fqs;	/* Kicked due to dynticks idle. */
#endif /* #ifdef CONFIG_NO_HZ */
	unsigned long offline_fqs;	/* Kicked due to being offline. */
	unsigned long resched_ipi;	/* Sent a resched IPI. */

	/* 5) __rcu_pending() statistics. */
	unsigned long n_rcu_pending;	/* rcu_pending() calls since boot. */
	unsigned long n_rp_qs_pending;
	unsigned long n_rp_report_qs;
	unsigned long n_rp_cb_ready;
	unsigned long n_rp_cpu_needs_gp;
	unsigned long n_rp_gp_completed;
	unsigned long n_rp_gp_started;
	unsigned long n_rp_need_fqs;
	unsigned long n_rp_need_nothing;

	int cpu;
};

/* Values for signaled field in struct rcu_state. */
#define RCU_GP_IDLE		0	/* No grace period in progress. */
#define RCU_GP_INIT		1	/* Grace period being initialized. */
#define RCU_SAVE_DYNTICK	2	/* Need to scan dyntick state. */
#define RCU_FORCE_QS		3	/* Need to force quiescent state. */
#ifdef CONFIG_NO_HZ
#define RCU_SIGNAL_INIT		RCU_SAVE_DYNTICK
#else /* #ifdef CONFIG_NO_HZ */
#define RCU_SIGNAL_INIT		RCU_FORCE_QS
#endif /* #else #ifdef CONFIG_NO_HZ */

#define RCU_JIFFIES_TILL_FORCE_QS	 3	/* for rsp->jiffies_force_qs */

#ifdef CONFIG_PROVE_RCU
#define RCU_STALL_DELAY_DELTA	       (5 * HZ)
#else
#define RCU_STALL_DELAY_DELTA	       0
#endif

#define RCU_SECONDS_TILL_STALL_CHECK   (CONFIG_RCU_CPU_STALL_TIMEOUT * HZ + \
					RCU_STALL_DELAY_DELTA)
						/* for rsp->jiffies_stall */
#define RCU_SECONDS_TILL_STALL_RECHECK (3 * RCU_SECONDS_TILL_STALL_CHECK + 30)
						/* for rsp->jiffies_stall */
#define RCU_STALL_RAT_DELAY		2	/* Allow other CPUs time */
						/*  to take at least one */
						/*  scheduling clock irq */
						/*  before ratting on them. */


/*
 * RCU global state, including node hierarchy.  This hierarchy is
 * represented in "heap" form in a dense array.  The root (first level)
 * of the hierarchy is in ->node[0] (referenced by ->level[0]), the second
 * level in ->node[1] through ->node[m] (->node[1] referenced by ->level[1]),
 * and the third level in ->node[m+1] and following (->node[m+1] referenced
 * by ->level[2]).  The number of levels is determined by the number of
 * CPUs and by CONFIG_RCU_FANOUT.  Small systems will have a "hierarchy"
 * consisting of a single rcu_node.
 */
struct rcu_state {
	struct rcu_node node[NUM_RCU_NODES];	/* Hierarchy. */
	struct rcu_node *level[NUM_RCU_LVLS];	/* Hierarchy levels. */
	u32 levelcnt[MAX_RCU_LVLS + 1];		/* # nodes in each level. */
	u8 levelspread[NUM_RCU_LVLS];		/* kids/node in each level. */
	struct rcu_data __percpu *rda;		/* pointer of percu rcu_data. */

	/* The following fields are guarded by the root rcu_node's lock. */

	u8	signaled ____cacheline_internodealigned_in_smp;
						/* Force QS state. */
	u8	fqs_active;			/* force_quiescent_state() */
						/*  is running. */
	u8	fqs_need_gp;			/* A CPU was prevented from */
						/*  starting a new grace */
						/*  period because */
						/*  force_quiescent_state() */
						/*  was running. */
	unsigned long gpnum;			/* Current gp number. */
	unsigned long completed;		/* # of last completed gp. */

	/* End of fields guarded by root rcu_node's lock. */

	raw_spinlock_t onofflock;		/* exclude on/offline and */
						/*  starting new GP. */
	raw_spinlock_t fqslock;			/* Only one task forcing */
						/*  quiescent states. */
	unsigned long jiffies_force_qs;		/* Time at which to invoke */
						/*  force_quiescent_state(). */
	unsigned long n_force_qs;		/* Number of calls to */
						/*  force_quiescent_state(). */
	unsigned long n_force_qs_lh;		/* ~Number of calls leaving */
						/*  due to lock unavailable. */
	unsigned long n_force_qs_ngp;		/* Number of calls leaving */
						/*  due to no GP active. */
	unsigned long gp_start;			/* Time at which GP started, */
						/*  but in jiffies. */
	unsigned long jiffies_stall;		/* Time at which to check */
						/*  for CPU stalls. */
	char *name;				/* Name of structure. */
};

/* Return values for rcu_preempt_offline_tasks(). */

#define RCU_OFL_TASKS_NORM_GP	0x1		/* Tasks blocking normal */
						/*  GP were moved to root. */
#define RCU_OFL_TASKS_EXP_GP	0x2		/* Tasks blocking expedited */
						/*  GP were moved to root. */

/*
 * RCU implementation internal declarations:
 */
extern struct rcu_state rcu_sched_state;
DECLARE_PER_CPU(struct rcu_data, rcu_sched_data);

extern struct rcu_state rcu_bh_state;
DECLARE_PER_CPU(struct rcu_data, rcu_bh_data);

#ifdef CONFIG_TREE_PREEMPT_RCU
extern struct rcu_state rcu_preempt_state;
DECLARE_PER_CPU(struct rcu_data, rcu_preempt_data);
#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */

#ifndef RCU_TREE_NONCORE

/* Forward declarations for rcutree_plugin.h */
static void rcu_bootup_announce(void);
long rcu_batches_completed(void);
static void rcu_preempt_note_context_switch(int cpu);
static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp);
#ifdef CONFIG_HOTPLUG_CPU
static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp,
				      unsigned long flags);
#endif /* #ifdef CONFIG_HOTPLUG_CPU */
static void rcu_print_detail_task_stall(struct rcu_state *rsp);
static void rcu_print_task_stall(struct rcu_node *rnp);
static void rcu_preempt_stall_reset(void);
static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp);
#ifdef CONFIG_HOTPLUG_CPU
static int rcu_preempt_offline_tasks(struct rcu_state *rsp,
				     struct rcu_node *rnp,
				     struct rcu_data *rdp);
static void rcu_preempt_offline_cpu(int cpu);
#endif /* #ifdef CONFIG_HOTPLUG_CPU */
static void rcu_preempt_check_callbacks(int cpu);
static void rcu_preempt_process_callbacks(void);
void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
#if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_TREE_PREEMPT_RCU)
static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp);
#endif /* #if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_TREE_PREEMPT_RCU) */
static int rcu_preempt_pending(int cpu);
static int rcu_preempt_needs_cpu(int cpu);
static void __cpuinit rcu_preempt_init_percpu_data(int cpu);
static void rcu_preempt_send_cbs_to_online(void);
static void __init __rcu_init_preempt(void);
static void rcu_needs_cpu_flush(void);
static void __init rcu_init_boost_waitqueue(struct rcu_node *rnp);
static void rcu_initiate_boost(struct rcu_node *rnp);
static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp,
					  cpumask_var_t cm);
static void rcu_preempt_boost_start_gp(struct rcu_node *rnp);
static int __cpuinit rcu_spawn_one_boost_kthread(struct rcu_state *rsp,
						 struct rcu_node *rnp,
						 int rnp_index);
#ifdef CONFIG_HOTPLUG_CPU
static void rcu_stop_boost_kthread(struct rcu_node *rnp);
#endif /* #ifdef CONFIG_HOTPLUG_CPU */

#endif /* #ifndef RCU_TREE_NONCORE */
Commit	Line	Data
9f77da9f PM	1	/*
	2	* Read-Copy Update mechanism for mutual exclusion (tree-based version)
	3	* Internal non-public definitions.
	4	*
	5	* This program is free software; you can redistribute it and/or modify
	6	* it under the terms of the GNU General Public License as published by
	7	* the Free Software Foundation; either version 2 of the License, or
	8	* (at your option) any later version.
	9	*
	10	* This program is distributed in the hope that it will be useful,
	11	* but WITHOUT ANY WARRANTY; without even the implied warranty of
	12	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
	13	* GNU General Public License for more details.
	14	*
	15	* You should have received a copy of the GNU General Public License
	16	* along with this program; if not, write to the Free Software
	17	* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
	18	*
	19	* Copyright IBM Corporation, 2008
	20	*
	21	* Author: Ingo Molnar <mingo@elte.hu>
	22	* Paul E. McKenney <paulmck@linux.vnet.ibm.com>
	23	*/
	24
	25	#include <linux/cache.h>
	26	#include <linux/spinlock.h>
	27	#include <linux/threads.h>
	28	#include <linux/cpumask.h>
	29	#include <linux/seqlock.h>
	30
	31	/*
	32	* Define shape of hierarchy based on NR_CPUS and CONFIG_RCU_FANOUT.
	33	* In theory, it should be possible to add more levels straightforwardly.
0209f649 PM	34	* In practice, this did work well going from three levels to four.
0209f649 PM	35	* Of course, your mileage may vary.
9f77da9f	36	*/
cf244dc0	37	#define MAX_RCU_LVLS 4
0209f649 PM	38	#if CONFIG_RCU_FANOUT > 16
	39	#define RCU_FANOUT_LEAF 16
	40	#else /* #if CONFIG_RCU_FANOUT > 16 */
	41	#define RCU_FANOUT_LEAF (CONFIG_RCU_FANOUT)
	42	#endif /* #else #if CONFIG_RCU_FANOUT > 16 */
	43	#define RCU_FANOUT_1 (RCU_FANOUT_LEAF)
	44	#define RCU_FANOUT_2 (RCU_FANOUT_1 * CONFIG_RCU_FANOUT)
	45	#define RCU_FANOUT_3 (RCU_FANOUT_2 * CONFIG_RCU_FANOUT)
	46	#define RCU_FANOUT_4 (RCU_FANOUT_3 * CONFIG_RCU_FANOUT)
9f77da9f	47
0209f649	48	#if NR_CPUS <= RCU_FANOUT_1
9f77da9f PM	49	# define NUM_RCU_LVLS 1
	50	# define NUM_RCU_LVL_0 1
	51	# define NUM_RCU_LVL_1 (NR_CPUS)
	52	# define NUM_RCU_LVL_2 0
	53	# define NUM_RCU_LVL_3 0
cf244dc0	54	# define NUM_RCU_LVL_4 0
0209f649	55	#elif NR_CPUS <= RCU_FANOUT_2
9f77da9f PM	56	# define NUM_RCU_LVLS 2
9f77da9f PM	57	# define NUM_RCU_LVL_0 1
0209f649	58	# define NUM_RCU_LVL_1 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
9f77da9f PM	59	# define NUM_RCU_LVL_2 (NR_CPUS)
9f77da9f PM	60	# define NUM_RCU_LVL_3 0
cf244dc0	61	# define NUM_RCU_LVL_4 0
0209f649	62	#elif NR_CPUS <= RCU_FANOUT_3
9f77da9f PM	63	# define NUM_RCU_LVLS 3
9f77da9f PM	64	# define NUM_RCU_LVL_0 1
0209f649 PM	65	# define NUM_RCU_LVL_1 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_2)
	66	# define NUM_RCU_LVL_2 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
	67	# define NUM_RCU_LVL_3 (NR_CPUS)
cf244dc0	68	# define NUM_RCU_LVL_4 0
0209f649	69	#elif NR_CPUS <= RCU_FANOUT_4
cf244dc0 PM	70	# define NUM_RCU_LVLS 4
cf244dc0 PM	71	# define NUM_RCU_LVL_0 1
0209f649 PM	72	# define NUM_RCU_LVL_1 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_3)
	73	# define NUM_RCU_LVL_2 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_2)
	74	# define NUM_RCU_LVL_3 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
	75	# define NUM_RCU_LVL_4 (NR_CPUS)
9f77da9f PM	76	#else
9f77da9f PM	77	# error "CONFIG_RCU_FANOUT insufficient for NR_CPUS"
0209f649	78	#endif /* #if (NR_CPUS) <= RCU_FANOUT_1 */
9f77da9f	79
cf244dc0	80	#define RCU_SUM (NUM_RCU_LVL_0 + NUM_RCU_LVL_1 + NUM_RCU_LVL_2 + NUM_RCU_LVL_3 + NUM_RCU_LVL_4)
9f77da9f PM	81	#define NUM_RCU_NODES (RCU_SUM - NR_CPUS)
	82
	83	/*
	84	* Dynticks per-CPU state.
	85	*/
	86	struct rcu_dynticks {
e59fb312 PM	87	int dynticks_nesting; /* Track irq/process nesting level. */
	88	int dynticks_nmi_nesting; /* Track NMI nesting level. */
	89	atomic_t dynticks; /* Even value for dynticks-idle, else odd. */
9f77da9f PM	90	};
	91
	92	/*
	93	* Definition for node within the RCU grace-period-detection hierarchy.
	94	*/
	95	struct rcu_node {
1304afb2	96	raw_spinlock_t lock; /* Root rcu_node's lock protects some */
1eba8f84	97	/* rcu_state fields as well as following. */
20133cfc	98	unsigned long gpnum; /* Current grace period for this node. */
86848966 PM	99	/* This will either be equal to or one */
86848966 PM	100	/* behind the root rcu_node's gpnum. */
20133cfc	101	unsigned long completed; /* Last GP completed for this node. */
d09b62df PM	102	/* This will either be equal to or one */
d09b62df PM	103	/* behind the root rcu_node's gpnum. */
9f77da9f PM	104	unsigned long qsmask; /* CPUs or groups that need to switch in */
9f77da9f PM	105	/* order for current grace period to proceed.*/
1eba8f84 PM	106	/* In leaf rcu_node, each bit corresponds to */
	107	/* an rcu_data structure, otherwise, each */
	108	/* bit corresponds to a child rcu_node */
	109	/* structure. */
12f5f524	110	unsigned long expmask; /* Groups that have ->blkd_tasks */
d9a3da06 PM	111	/* elements that need to drain to allow the */
	112	/* current expedited grace period to */
	113	/* complete (only for TREE_PREEMPT_RCU). */
a26ac245	114	unsigned long wakemask; /* CPUs whose kthread needs to be awakened. */
9f77da9f	115	unsigned long qsmaskinit;
d9a3da06	116	/* Per-GP initial value for qsmask & expmask. */
9f77da9f	117	unsigned long grpmask; /* Mask to apply to parent qsmask. */
1eba8f84	118	/* Only one bit will be set in this mask. */
9f77da9f PM	119	int grplo; /* lowest-numbered CPU or group here. */
	120	int grphi; /* highest-numbered CPU or group here. */
	121	u8 grpnum; /* CPU/group number for next level up. */
	122	u8 level; /* root is at level 0. */
	123	struct rcu_node *parent;
12f5f524 PM	124	struct list_head blkd_tasks;
	125	/* Tasks blocked in RCU read-side critical */
	126	/* section. Tasks are placed at the head */
	127	/* of this list and age towards the tail. */
	128	struct list_head *gp_tasks;
	129	/* Pointer to the first task blocking the */
	130	/* current grace period, or NULL if there */
	131	/* is no such task. */
	132	struct list_head *exp_tasks;
	133	/* Pointer to the first task blocking the */
	134	/* current expedited grace period, or NULL */
	135	/* if there is no such task. If there */
	136	/* is no current expedited grace period, */
	137	/* then there can cannot be any such task. */
27f4d280 PM	138	#ifdef CONFIG_RCU_BOOST
	139	struct list_head *boost_tasks;
	140	/* Pointer to first task that needs to be */
	141	/* priority boosted, or NULL if no priority */
	142	/* boosting is needed for this rcu_node */
	143	/* structure. If there are no tasks */
	144	/* queued on this rcu_node structure that */
	145	/* are blocking the current grace period, */
	146	/* there can be no such task. */
	147	unsigned long boost_time;
	148	/* When to start boosting (jiffies). */
	149	struct task_struct *boost_kthread_task;
	150	/* kthread that takes care of priority */
	151	/* boosting for this rcu_node structure. */
	152	wait_queue_head_t boost_wq;
	153	/* Wait queue on which to park the boost */
	154	/* kthread. */
	155	#endif /* #ifdef CONFIG_RCU_BOOST */
a26ac245 PM	156	struct task_struct *node_kthread_task;
	157	/* kthread that takes care of this rcu_node */
	158	/* structure, for example, awakening the */
	159	/* per-CPU kthreads as needed. */
	160	wait_queue_head_t node_wq;
	161	/* Wait queue on which to park the per-node */
	162	/* kthread. */
9f77da9f PM	163	} ____cacheline_internodealigned_in_smp;
9f77da9f PM	164
a0b6c9a7 PM	165	/*
	166	* Do a full breadth-first scan of the rcu_node structures for the
	167	* specified rcu_state structure.
	168	*/
	169	#define rcu_for_each_node_breadth_first(rsp, rnp) \
	170	for ((rnp) = &(rsp)->node[0]; \
	171	(rnp) < &(rsp)->node[NUM_RCU_NODES]; (rnp)++)
	172
d9a3da06 PM	173	/*
	174	* Do a breadth-first scan of the non-leaf rcu_node structures for the
	175	* specified rcu_state structure. Note that if there is a singleton
	176	* rcu_node tree with but one rcu_node structure, this loop is a no-op.
	177	*/
	178	#define rcu_for_each_nonleaf_node_breadth_first(rsp, rnp) \
	179	for ((rnp) = &(rsp)->node[0]; \
	180	(rnp) < (rsp)->level[NUM_RCU_LVLS - 1]; (rnp)++)
	181
	182	/*
	183	* Scan the leaves of the rcu_node hierarchy for the specified rcu_state
	184	* structure. Note that if there is a singleton rcu_node tree with but
	185	* one rcu_node structure, this loop -will- visit the rcu_node structure.
	186	* It is still a leaf node, even if it is also the root node.
	187	*/
a0b6c9a7 PM	188	#define rcu_for_each_leaf_node(rsp, rnp) \
	189	for ((rnp) = (rsp)->level[NUM_RCU_LVLS - 1]; \
	190	(rnp) < &(rsp)->node[NUM_RCU_NODES]; (rnp)++)
	191
9f77da9f PM	192	/* Index values for nxttail array in struct rcu_data. */
	193	#define RCU_DONE_TAIL 0 /* Also RCU_WAIT head. */
	194	#define RCU_WAIT_TAIL 1 /* Also RCU_NEXT_READY head. */
	195	#define RCU_NEXT_READY_TAIL 2 /* Also RCU_NEXT head. */
	196	#define RCU_NEXT_TAIL 3
	197	#define RCU_NEXT_SIZE 4
	198
	199	/* Per-CPU data for read-copy update. */
	200	struct rcu_data {
	201	/* 1) quiescent-state and grace-period handling : */
20133cfc	202	unsigned long completed; /* Track rsp->completed gp number */
9f77da9f	203	/* in order to detect GP end. */
20133cfc	204	unsigned long gpnum; /* Highest gp number that this CPU */
9f77da9f	205	/* is aware of having started. */
20133cfc	206	unsigned long passed_quiesc_completed;
9f77da9f PM	207	/* Value of completed at time of qs. */
	208	bool passed_quiesc; /* User-mode/idle loop etc. */
	209	bool qs_pending; /* Core waits for quiesc state. */
	210	bool beenonline; /* CPU online at least once. */
f41d911f	211	bool preemptable; /* Preemptable RCU? */
9f77da9f PM	212	struct rcu_node mynode; / This CPU's leaf of hierarchy */
	213	unsigned long grpmask; /* Mask to apply to leaf qsmask. */
	214
	215	/* 2) batch handling */
	216	/*
	217	* If nxtlist is not NULL, it is partitioned as follows.
	218	* Any of the partitions might be empty, in which case the
	219	* pointer to that partition will be equal to the pointer for
	220	* the following partition. When the list is empty, all of
1eba8f84 PM	221	* the nxttail elements point to the ->nxtlist pointer itself,
1eba8f84 PM	222	* which in that case is NULL.
9f77da9f	223	*
9f77da9f PM	224	* [nxtlist, *nxttail[RCU_DONE_TAIL]):
	225	* Entries that batch # <= ->completed
	226	* The grace period for these entries has completed, and
	227	* the other grace-period-completed entries may be moved
	228	* here temporarily in rcu_process_callbacks().
1eba8f84 PM	229	* [nxttail[RCU_DONE_TAIL], nxttail[RCU_WAIT_TAIL]):
	230	* Entries that batch # <= ->completed - 1: waiting for current GP
	231	* [nxttail[RCU_WAIT_TAIL], nxttail[RCU_NEXT_READY_TAIL]):
	232	* Entries known to have arrived before current GP ended
	233	* [nxttail[RCU_NEXT_READY_TAIL], nxttail[RCU_NEXT_TAIL]):
	234	* Entries that might have arrived after current GP ended
	235	* Note that the value of *nxttail[RCU_NEXT_TAIL] will
	236	* always be NULL, as this is the end of the list.
9f77da9f PM	237	*/
	238	struct rcu_head *nxtlist;
	239	struct rcu_head **nxttail[RCU_NEXT_SIZE];
a71fca58	240	long qlen; /* # of queued callbacks */
37c72e56 PM	241	long qlen_last_fqs_check;
37c72e56 PM	242	/* qlen at last check for QS forcing */
269dcc1c	243	unsigned long n_cbs_invoked; /* count of RCU cbs invoked. */
29494be7 LJ	244	unsigned long n_cbs_orphaned; /* RCU cbs orphaned by dying CPU */
29494be7 LJ	245	unsigned long n_cbs_adopted; /* RCU cbs adopted from dying CPU */
37c72e56 PM	246	unsigned long n_force_qs_snap;
37c72e56 PM	247	/* did other CPU force QS recently? */
9f77da9f PM	248	long blimit; /* Upper limit on a processed batch */
	249
	250	#ifdef CONFIG_NO_HZ
	251	/* 3) dynticks interface. */
	252	struct rcu_dynticks dynticks; / Shared per-CPU dynticks state. */
	253	int dynticks_snap; /* Per-GP tracking for dynticks. */
9f77da9f PM	254	#endif /* #ifdef CONFIG_NO_HZ */
	255
	256	/* 4) reasons this CPU needed to be kicked by force_quiescent_state */
	257	#ifdef CONFIG_NO_HZ
	258	unsigned long dynticks_fqs; /* Kicked due to dynticks idle. */
	259	#endif /* #ifdef CONFIG_NO_HZ */
	260	unsigned long offline_fqs; /* Kicked due to being offline. */
	261	unsigned long resched_ipi; /* Sent a resched IPI. */
	262
	263	/* 5) __rcu_pending() statistics. */
20133cfc PM	264	unsigned long n_rcu_pending; /* rcu_pending() calls since boot. */
20133cfc PM	265	unsigned long n_rp_qs_pending;
d21670ac	266	unsigned long n_rp_report_qs;
20133cfc PM	267	unsigned long n_rp_cb_ready;
	268	unsigned long n_rp_cpu_needs_gp;
	269	unsigned long n_rp_gp_completed;
	270	unsigned long n_rp_gp_started;
	271	unsigned long n_rp_need_fqs;
	272	unsigned long n_rp_need_nothing;
9f77da9f PM	273
	274	int cpu;
	275	};
	276
	277	/* Values for signaled field in struct rcu_state. */
83f5b01f PM	278	#define RCU_GP_IDLE 0 /* No grace period in progress. */
	279	#define RCU_GP_INIT 1 /* Grace period being initialized. */
	280	#define RCU_SAVE_DYNTICK 2 /* Need to scan dyntick state. */
ee47eb9f	281	#define RCU_FORCE_QS 3 /* Need to force quiescent state. */
9f77da9f PM	282	#ifdef CONFIG_NO_HZ
	283	#define RCU_SIGNAL_INIT RCU_SAVE_DYNTICK
	284	#else /* #ifdef CONFIG_NO_HZ */
ee47eb9f	285	#define RCU_SIGNAL_INIT RCU_FORCE_QS
9f77da9f PM	286	#endif /* #else #ifdef CONFIG_NO_HZ */
	287
	288	#define RCU_JIFFIES_TILL_FORCE_QS 3 /* for rsp->jiffies_force_qs */
007b0924 PM	289
	290	#ifdef CONFIG_PROVE_RCU
	291	#define RCU_STALL_DELAY_DELTA (5 * HZ)
	292	#else
	293	#define RCU_STALL_DELAY_DELTA 0
	294	#endif
	295
b163760e PM	296	#define RCU_SECONDS_TILL_STALL_CHECK (CONFIG_RCU_CPU_STALL_TIMEOUT * HZ + \
b163760e PM	297	RCU_STALL_DELAY_DELTA)
007b0924	298	/* for rsp->jiffies_stall */
b163760e	299	#define RCU_SECONDS_TILL_STALL_RECHECK (3 * RCU_SECONDS_TILL_STALL_CHECK + 30)
007b0924 PM	300	/* for rsp->jiffies_stall */
	301	#define RCU_STALL_RAT_DELAY 2 /* Allow other CPUs time */
	302	/* to take at least one */
	303	/* scheduling clock irq */
	304	/* before ratting on them. */
9f77da9f	305
9f77da9f PM	306
	307	/*
	308	* RCU global state, including node hierarchy. This hierarchy is
	309	* represented in "heap" form in a dense array. The root (first level)
	310	* of the hierarchy is in ->node[0] (referenced by ->level[0]), the second
	311	* level in ->node[1] through ->node[m] (->node[1] referenced by ->level[1]),
	312	* and the third level in ->node[m+1] and following (->node[m+1] referenced
	313	* by ->level[2]). The number of levels is determined by the number of
	314	* CPUs and by CONFIG_RCU_FANOUT. Small systems will have a "hierarchy"
	315	* consisting of a single rcu_node.
	316	*/
	317	struct rcu_state {
	318	struct rcu_node node[NUM_RCU_NODES]; /* Hierarchy. */
	319	struct rcu_node level[NUM_RCU_LVLS]; / Hierarchy levels. */
	320	u32 levelcnt[MAX_RCU_LVLS + 1]; /* # nodes in each level. */
	321	u8 levelspread[NUM_RCU_LVLS]; /* kids/node in each level. */
394f99a9	322	struct rcu_data __percpu rda; / pointer of percu rcu_data. */
9f77da9f PM	323
	324	/* The following fields are guarded by the root rcu_node's lock. */
	325
	326	u8 signaled ____cacheline_internodealigned_in_smp;
	327	/* Force QS state. */
07079d53 PM	328	u8 fqs_active; /* force_quiescent_state() */
07079d53 PM	329	/* is running. */
46a1e34e PM	330	u8 fqs_need_gp; /* A CPU was prevented from */
	331	/* starting a new grace */
	332	/* period because */
	333	/* force_quiescent_state() */
	334	/* was running. */
20133cfc PM	335	unsigned long gpnum; /* Current gp number. */
20133cfc PM	336	unsigned long completed; /* # of last completed gp. */
1eba8f84	337
d9a3da06	338	/* End of fields guarded by root rcu_node's lock. */
1eba8f84	339
1304afb2	340	raw_spinlock_t onofflock; /* exclude on/offline and */
29494be7	341	/* starting new GP. */
1304afb2	342	raw_spinlock_t fqslock; /* Only one task forcing */
9f77da9f PM	343	/* quiescent states. */
	344	unsigned long jiffies_force_qs; /* Time at which to invoke */
	345	/* force_quiescent_state(). */
	346	unsigned long n_force_qs; /* Number of calls to */
	347	/* force_quiescent_state(). */
	348	unsigned long n_force_qs_lh; /* ~Number of calls leaving */
	349	/* due to lock unavailable. */
	350	unsigned long n_force_qs_ngp; /* Number of calls leaving */
	351	/* due to no GP active. */
9f77da9f PM	352	unsigned long gp_start; /* Time at which GP started, */
	353	/* but in jiffies. */
	354	unsigned long jiffies_stall; /* Time at which to check */
	355	/* for CPU stalls. */
4300aa64	356	char name; / Name of structure. */
9f77da9f PM	357	};
9f77da9f PM	358
d9a3da06 PM	359	/* Return values for rcu_preempt_offline_tasks(). */
	360
	361	#define RCU_OFL_TASKS_NORM_GP 0x1 /* Tasks blocking normal */
	362	/* GP were moved to root. */
	363	#define RCU_OFL_TASKS_EXP_GP 0x2 /* Tasks blocking expedited */
	364	/* GP were moved to root. */
	365
6258c4fb IM	366	/*
	367	* RCU implementation internal declarations:
	368	*/
d6714c22 PM	369	extern struct rcu_state rcu_sched_state;
d6714c22 PM	370	DECLARE_PER_CPU(struct rcu_data, rcu_sched_data);
6258c4fb IM	371
	372	extern struct rcu_state rcu_bh_state;
	373	DECLARE_PER_CPU(struct rcu_data, rcu_bh_data);
	374
f41d911f PM	375	#ifdef CONFIG_TREE_PREEMPT_RCU
	376	extern struct rcu_state rcu_preempt_state;
	377	DECLARE_PER_CPU(struct rcu_data, rcu_preempt_data);
	378	#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */
	379
017c4261	380	#ifndef RCU_TREE_NONCORE
9f77da9f	381
9b2619af	382	/* Forward declarations for rcutree_plugin.h */
dbe01350	383	static void rcu_bootup_announce(void);
9b2619af PM	384	long rcu_batches_completed(void);
9b2619af PM	385	static void rcu_preempt_note_context_switch(int cpu);
27f4d280	386	static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp);
b668c9cf	387	#ifdef CONFIG_HOTPLUG_CPU
d3f6bad3 PM	388	static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp,
d3f6bad3 PM	389	unsigned long flags);
b668c9cf	390	#endif /* #ifdef CONFIG_HOTPLUG_CPU */
1ed509a2	391	static void rcu_print_detail_task_stall(struct rcu_state *rsp);
9b2619af	392	static void rcu_print_task_stall(struct rcu_node *rnp);
53d84e00	393	static void rcu_preempt_stall_reset(void);
9b2619af PM	394	static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp);
9b2619af PM	395	#ifdef CONFIG_HOTPLUG_CPU
237c80c5 PM	396	static int rcu_preempt_offline_tasks(struct rcu_state *rsp,
	397	struct rcu_node *rnp,
	398	struct rcu_data *rdp);
9b2619af PM	399	static void rcu_preempt_offline_cpu(int cpu);
	400	#endif /* #ifdef CONFIG_HOTPLUG_CPU */
	401	static void rcu_preempt_check_callbacks(int cpu);
	402	static void rcu_preempt_process_callbacks(void);
	403	void call_rcu(struct rcu_head head, void (func)(struct rcu_head *rcu));
d9a3da06 PM	404	#if defined(CONFIG_HOTPLUG_CPU) \|\| defined(CONFIG_TREE_PREEMPT_RCU)
	405	static void rcu_report_exp_rnp(struct rcu_state rsp, struct rcu_node rnp);
	406	#endif /* #if defined(CONFIG_HOTPLUG_CPU) \|\| defined(CONFIG_TREE_PREEMPT_RCU) */
9b2619af PM	407	static int rcu_preempt_pending(int cpu);
	408	static int rcu_preempt_needs_cpu(int cpu);
	409	static void __cpuinit rcu_preempt_init_percpu_data(int cpu);
29494be7	410	static void rcu_preempt_send_cbs_to_online(void);
9b2619af	411	static void __init __rcu_init_preempt(void);
a47cd880	412	static void rcu_needs_cpu_flush(void);
27f4d280 PM	413	static void __init rcu_init_boost_waitqueue(struct rcu_node *rnp);
	414	static void rcu_initiate_boost(struct rcu_node *rnp);
	415	static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp,
	416	cpumask_var_t cm);
	417	static void rcu_preempt_boost_start_gp(struct rcu_node *rnp);
	418	static int __cpuinit rcu_spawn_one_boost_kthread(struct rcu_state *rsp,
	419	struct rcu_node *rnp,
	420	int rnp_index);
	421	#ifdef CONFIG_HOTPLUG_CPU
	422	static void rcu_stop_boost_kthread(struct rcu_node *rnp);
	423	#endif /* #ifdef CONFIG_HOTPLUG_CPU */
9b2619af	424
017c4261	425	#endif /* #ifndef RCU_TREE_NONCORE */