[GitHub/mt8127/android_kernel_alcatel_ttab.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 has not been tested, so there is probably some
 * bug somewhere.
 */
#define MAX_RCU_LVLS 3
#define RCU_FANOUT	      (CONFIG_RCU_FANOUT)
#define RCU_FANOUT_SQ	      (RCU_FANOUT * RCU_FANOUT)
#define RCU_FANOUT_CUBE	      (RCU_FANOUT_SQ * RCU_FANOUT)

#if NR_CPUS <= RCU_FANOUT
#  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
#elif NR_CPUS <= RCU_FANOUT_SQ
#  define NUM_RCU_LVLS	      2
#  define NUM_RCU_LVL_0	      1
#  define NUM_RCU_LVL_1	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT)
#  define NUM_RCU_LVL_2	      (NR_CPUS)
#  define NUM_RCU_LVL_3	      0
#elif NR_CPUS <= RCU_FANOUT_CUBE
#  define NUM_RCU_LVLS	      3
#  define NUM_RCU_LVL_0	      1
#  define NUM_RCU_LVL_1	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_SQ)
#  define NUM_RCU_LVL_2	      DIV_ROUND_UP(NR_CPUS, RCU_FANOUT)
#  define NUM_RCU_LVL_3	      NR_CPUS
#else
# error "CONFIG_RCU_FANOUT insufficient for NR_CPUS"
#endif /* #if (NR_CPUS) <= RCU_FANOUT */

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

/*
 * Dynticks per-CPU state.
 */
struct rcu_dynticks {
	int dynticks_nesting;	/* Track nesting level, sort of. */
	int dynticks;		/* Even value for dynticks-idle, else odd. */
	int dynticks_nmi;	/* Even value for either dynticks-idle or */
				/*  not in nmi handler, else odd.  So this */
				/*  remains even for nmi from irq handler. */
};

/*
 * Definition for node within the RCU grace-period-detection hierarchy.
 */
struct rcu_node {
	spinlock_t lock;	/* Root rcu_node's lock protects some */
				/*  rcu_state fields as well as following. */
	long	gpnum;		/* Current grace period for this node. */
				/*  This will either be equal to or one */
				/*  behind the root rcu_node's gpnum. */
	long	completed;	/* Last grace period 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 qsmaskinit;
				/* Per-GP initialization for qsmask. */
	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 blocked_tasks[2];
				/* Tasks blocked in RCU read-side critsect. */
				/*  Grace period number (->gpnum) x blocked */
				/*  by tasks on the (x & 0x1) element of the */
				/*  blocked_tasks[] array. */
} ____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)++)

#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 : */
	long		completed;	/* Track rsp->completed gp number */
					/*  in order to detect GP end. */
	long		gpnum;		/* Highest gp number that this CPU */
					/*  is aware of having started. */
	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_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. */
	int dynticks_nmi_snap;		/* Per-GP tracking for dynticks_nmi. */
#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. */
	long n_rcu_pending;		/* rcu_pending() calls since boot. */
	long n_rp_qs_pending;
	long n_rp_cb_ready;
	long n_rp_cpu_needs_gp;
	long n_rp_gp_completed;
	long n_rp_gp_started;
	long n_rp_need_fqs;
	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_SAVE_COMPLETED	3	/* Need to save rsp->completed. */
#define RCU_FORCE_QS		4	/* 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_SAVE_COMPLETED
#endif /* #else #ifdef CONFIG_NO_HZ */

#define RCU_JIFFIES_TILL_FORCE_QS	 3	/* for rsp->jiffies_force_qs */
#ifdef CONFIG_RCU_CPU_STALL_DETECTOR
#define RCU_SECONDS_TILL_STALL_CHECK   (10 * HZ)  /* for rsp->jiffies_stall */
#define RCU_SECONDS_TILL_STALL_RECHECK (30 * HZ)  /* 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. */

#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */

/*
 * 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 *rda[NR_CPUS];		/* array of rdp pointers. */

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

	u8	signaled ____cacheline_internodealigned_in_smp;
						/* Force QS state. */
	long	gpnum;				/* Current gp number. */
	long	completed;			/* # of last completed gp. */

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

	spinlock_t onofflock;			/* exclude on/offline and */
						/*  starting new GP.  Also */
						/*  protects the following */
						/*  orphan_cbs fields. */
	struct rcu_head *orphan_cbs_list;	/* list of rcu_head structs */
						/*  orphaned by all CPUs in */
						/*  a given leaf rcu_node */
						/*  going offline. */
	struct rcu_head **orphan_cbs_tail;	/* And tail pointer. */
	long orphan_qlen;			/* Number of orphaned cbs. */
	spinlock_t fqslock;			/* Only one task forcing */
						/*  quiescent states. */
	long	completed_fqs;			/* Value of completed @ snap. */
						/*  Protected by fqslock. */
	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. */
#ifdef CONFIG_RCU_CPU_STALL_DETECTOR
	unsigned long gp_start;			/* Time at which GP started, */
						/*  but in jiffies. */
	unsigned long jiffies_stall;		/* Time at which to check */
						/*  for CPU stalls. */
#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */
};

#ifdef RCU_TREE_NONCORE

/*
 * 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 */

#else /* #ifdef 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_preempted_readers(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 */
#ifdef CONFIG_RCU_CPU_STALL_DETECTOR
static void rcu_print_task_stall(struct rcu_node *rnp);
#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */
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));
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_orphanage(void);
static void __init __rcu_init_preempt(void);

#endif /* #else #ifdef 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.
	34	* In practice, this has not been tested, so there is probably some
	35	* bug somewhere.
	36	*/
	37	#define MAX_RCU_LVLS 3
	38	#define RCU_FANOUT (CONFIG_RCU_FANOUT)
	39	#define RCU_FANOUT_SQ (RCU_FANOUT * RCU_FANOUT)
	40	#define RCU_FANOUT_CUBE (RCU_FANOUT_SQ * RCU_FANOUT)
	41
	42	#if NR_CPUS <= RCU_FANOUT
	43	# define NUM_RCU_LVLS 1
	44	# define NUM_RCU_LVL_0 1
	45	# define NUM_RCU_LVL_1 (NR_CPUS)
	46	# define NUM_RCU_LVL_2 0
	47	# define NUM_RCU_LVL_3 0
	48	#elif NR_CPUS <= RCU_FANOUT_SQ
	49	# define NUM_RCU_LVLS 2
	50	# define NUM_RCU_LVL_0 1
fc2219d4	51	# define NUM_RCU_LVL_1 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT)
9f77da9f PM	52	# define NUM_RCU_LVL_2 (NR_CPUS)
	53	# define NUM_RCU_LVL_3 0
	54	#elif NR_CPUS <= RCU_FANOUT_CUBE
	55	# define NUM_RCU_LVLS 3
	56	# define NUM_RCU_LVL_0 1
fc2219d4 PM	57	# define NUM_RCU_LVL_1 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_SQ)
fc2219d4 PM	58	# define NUM_RCU_LVL_2 DIV_ROUND_UP(NR_CPUS, RCU_FANOUT)
9f77da9f PM	59	# define NUM_RCU_LVL_3 NR_CPUS
	60	#else
	61	# error "CONFIG_RCU_FANOUT insufficient for NR_CPUS"
	62	#endif /* #if (NR_CPUS) <= RCU_FANOUT */
	63
	64	#define RCU_SUM (NUM_RCU_LVL_0 + NUM_RCU_LVL_1 + NUM_RCU_LVL_2 + NUM_RCU_LVL_3)
	65	#define NUM_RCU_NODES (RCU_SUM - NR_CPUS)
	66
	67	/*
	68	* Dynticks per-CPU state.
	69	*/
	70	struct rcu_dynticks {
	71	int dynticks_nesting; /* Track nesting level, sort of. */
	72	int dynticks; /* Even value for dynticks-idle, else odd. */
	73	int dynticks_nmi; /* Even value for either dynticks-idle or */
	74	/* not in nmi handler, else odd. So this */
	75	/* remains even for nmi from irq handler. */
	76	};
	77
	78	/*
	79	* Definition for node within the RCU grace-period-detection hierarchy.
	80	*/
	81	struct rcu_node {
1eba8f84 PM	82	spinlock_t lock; /* Root rcu_node's lock protects some */
1eba8f84 PM	83	/* rcu_state fields as well as following. */
f41d911f	84	long gpnum; /* Current grace period for this node. */
86848966 PM	85	/* This will either be equal to or one */
86848966 PM	86	/* behind the root rcu_node's gpnum. */
d09b62df PM	87	long completed; /* Last grace period completed for this node. */
	88	/* This will either be equal to or one */
	89	/* behind the root rcu_node's gpnum. */
9f77da9f PM	90	unsigned long qsmask; /* CPUs or groups that need to switch in */
9f77da9f PM	91	/* order for current grace period to proceed.*/
1eba8f84 PM	92	/* In leaf rcu_node, each bit corresponds to */
	93	/* an rcu_data structure, otherwise, each */
	94	/* bit corresponds to a child rcu_node */
	95	/* structure. */
9f77da9f PM	96	unsigned long qsmaskinit;
	97	/* Per-GP initialization for qsmask. */
	98	unsigned long grpmask; /* Mask to apply to parent qsmask. */
1eba8f84	99	/* Only one bit will be set in this mask. */
9f77da9f PM	100	int grplo; /* lowest-numbered CPU or group here. */
	101	int grphi; /* highest-numbered CPU or group here. */
	102	u8 grpnum; /* CPU/group number for next level up. */
	103	u8 level; /* root is at level 0. */
	104	struct rcu_node *parent;
f41d911f PM	105	struct list_head blocked_tasks[2];
f41d911f PM	106	/* Tasks blocked in RCU read-side critsect. */
1eba8f84 PM	107	/* Grace period number (->gpnum) x blocked */
	108	/* by tasks on the (x & 0x1) element of the */
	109	/* blocked_tasks[] array. */
9f77da9f PM	110	} ____cacheline_internodealigned_in_smp;
9f77da9f PM	111
a0b6c9a7 PM	112	/*
	113	* Do a full breadth-first scan of the rcu_node structures for the
	114	* specified rcu_state structure.
	115	*/
	116	#define rcu_for_each_node_breadth_first(rsp, rnp) \
	117	for ((rnp) = &(rsp)->node[0]; \
	118	(rnp) < &(rsp)->node[NUM_RCU_NODES]; (rnp)++)
	119
	120	#define rcu_for_each_leaf_node(rsp, rnp) \
	121	for ((rnp) = (rsp)->level[NUM_RCU_LVLS - 1]; \
	122	(rnp) < &(rsp)->node[NUM_RCU_NODES]; (rnp)++)
	123
9f77da9f PM	124	/* Index values for nxttail array in struct rcu_data. */
	125	#define RCU_DONE_TAIL 0 /* Also RCU_WAIT head. */
	126	#define RCU_WAIT_TAIL 1 /* Also RCU_NEXT_READY head. */
	127	#define RCU_NEXT_READY_TAIL 2 /* Also RCU_NEXT head. */
	128	#define RCU_NEXT_TAIL 3
	129	#define RCU_NEXT_SIZE 4
	130
	131	/* Per-CPU data for read-copy update. */
	132	struct rcu_data {
	133	/* 1) quiescent-state and grace-period handling : */
	134	long completed; /* Track rsp->completed gp number */
	135	/* in order to detect GP end. */
	136	long gpnum; /* Highest gp number that this CPU */
	137	/* is aware of having started. */
	138	long passed_quiesc_completed;
	139	/* Value of completed at time of qs. */
	140	bool passed_quiesc; /* User-mode/idle loop etc. */
	141	bool qs_pending; /* Core waits for quiesc state. */
	142	bool beenonline; /* CPU online at least once. */
f41d911f	143	bool preemptable; /* Preemptable RCU? */
9f77da9f PM	144	struct rcu_node mynode; / This CPU's leaf of hierarchy */
	145	unsigned long grpmask; /* Mask to apply to leaf qsmask. */
	146
	147	/* 2) batch handling */
	148	/*
	149	* If nxtlist is not NULL, it is partitioned as follows.
	150	* Any of the partitions might be empty, in which case the
	151	* pointer to that partition will be equal to the pointer for
	152	* the following partition. When the list is empty, all of
1eba8f84 PM	153	* the nxttail elements point to the ->nxtlist pointer itself,
1eba8f84 PM	154	* which in that case is NULL.
9f77da9f	155	*
9f77da9f PM	156	* [nxtlist, *nxttail[RCU_DONE_TAIL]):
	157	* Entries that batch # <= ->completed
	158	* The grace period for these entries has completed, and
	159	* the other grace-period-completed entries may be moved
	160	* here temporarily in rcu_process_callbacks().
1eba8f84 PM	161	* [nxttail[RCU_DONE_TAIL], nxttail[RCU_WAIT_TAIL]):
	162	* Entries that batch # <= ->completed - 1: waiting for current GP
	163	* [nxttail[RCU_WAIT_TAIL], nxttail[RCU_NEXT_READY_TAIL]):
	164	* Entries known to have arrived before current GP ended
	165	* [nxttail[RCU_NEXT_READY_TAIL], nxttail[RCU_NEXT_TAIL]):
	166	* Entries that might have arrived after current GP ended
	167	* Note that the value of *nxttail[RCU_NEXT_TAIL] will
	168	* always be NULL, as this is the end of the list.
9f77da9f PM	169	*/
	170	struct rcu_head *nxtlist;
	171	struct rcu_head **nxttail[RCU_NEXT_SIZE];
a71fca58	172	long qlen; /* # of queued callbacks */
37c72e56 PM	173	long qlen_last_fqs_check;
	174	/* qlen at last check for QS forcing */
	175	unsigned long n_force_qs_snap;
	176	/* did other CPU force QS recently? */
9f77da9f PM	177	long blimit; /* Upper limit on a processed batch */
	178
	179	#ifdef CONFIG_NO_HZ
	180	/* 3) dynticks interface. */
	181	struct rcu_dynticks dynticks; / Shared per-CPU dynticks state. */
	182	int dynticks_snap; /* Per-GP tracking for dynticks. */
	183	int dynticks_nmi_snap; /* Per-GP tracking for dynticks_nmi. */
	184	#endif /* #ifdef CONFIG_NO_HZ */
	185
	186	/* 4) reasons this CPU needed to be kicked by force_quiescent_state */
	187	#ifdef CONFIG_NO_HZ
	188	unsigned long dynticks_fqs; /* Kicked due to dynticks idle. */
	189	#endif /* #ifdef CONFIG_NO_HZ */
	190	unsigned long offline_fqs; /* Kicked due to being offline. */
	191	unsigned long resched_ipi; /* Sent a resched IPI. */
	192
	193	/* 5) __rcu_pending() statistics. */
	194	long n_rcu_pending; /* rcu_pending() calls since boot. */
	195	long n_rp_qs_pending;
	196	long n_rp_cb_ready;
	197	long n_rp_cpu_needs_gp;
	198	long n_rp_gp_completed;
	199	long n_rp_gp_started;
	200	long n_rp_need_fqs;
	201	long n_rp_need_nothing;
	202
	203	int cpu;
	204	};
	205
	206	/* Values for signaled field in struct rcu_state. */
83f5b01f PM	207	#define RCU_GP_IDLE 0 /* No grace period in progress. */
	208	#define RCU_GP_INIT 1 /* Grace period being initialized. */
	209	#define RCU_SAVE_DYNTICK 2 /* Need to scan dyntick state. */
281d150c PM	210	#define RCU_SAVE_COMPLETED 3 /* Need to save rsp->completed. */
281d150c PM	211	#define RCU_FORCE_QS 4 /* Need to force quiescent state. */
9f77da9f PM	212	#ifdef CONFIG_NO_HZ
	213	#define RCU_SIGNAL_INIT RCU_SAVE_DYNTICK
	214	#else /* #ifdef CONFIG_NO_HZ */
281d150c	215	#define RCU_SIGNAL_INIT RCU_SAVE_COMPLETED
9f77da9f PM	216	#endif /* #else #ifdef CONFIG_NO_HZ */
	217
	218	#define RCU_JIFFIES_TILL_FORCE_QS 3 /* for rsp->jiffies_force_qs */
	219	#ifdef CONFIG_RCU_CPU_STALL_DETECTOR
	220	#define RCU_SECONDS_TILL_STALL_CHECK (10 * HZ) /* for rsp->jiffies_stall */
	221	#define RCU_SECONDS_TILL_STALL_RECHECK (30 * HZ) /* for rsp->jiffies_stall */
	222	#define RCU_STALL_RAT_DELAY 2 /* Allow other CPUs time */
	223	/* to take at least one */
	224	/* scheduling clock irq */
	225	/* before ratting on them. */
	226
	227	#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */
	228
	229	/*
	230	* RCU global state, including node hierarchy. This hierarchy is
	231	* represented in "heap" form in a dense array. The root (first level)
	232	* of the hierarchy is in ->node[0] (referenced by ->level[0]), the second
	233	* level in ->node[1] through ->node[m] (->node[1] referenced by ->level[1]),
	234	* and the third level in ->node[m+1] and following (->node[m+1] referenced
	235	* by ->level[2]). The number of levels is determined by the number of
	236	* CPUs and by CONFIG_RCU_FANOUT. Small systems will have a "hierarchy"
	237	* consisting of a single rcu_node.
	238	*/
	239	struct rcu_state {
	240	struct rcu_node node[NUM_RCU_NODES]; /* Hierarchy. */
	241	struct rcu_node level[NUM_RCU_LVLS]; / Hierarchy levels. */
	242	u32 levelcnt[MAX_RCU_LVLS + 1]; /* # nodes in each level. */
	243	u8 levelspread[NUM_RCU_LVLS]; /* kids/node in each level. */
	244	struct rcu_data rda[NR_CPUS]; / array of rdp pointers. */
	245
	246	/* The following fields are guarded by the root rcu_node's lock. */
	247
	248	u8 signaled ____cacheline_internodealigned_in_smp;
	249	/* Force QS state. */
	250	long gpnum; /* Current gp number. */
	251	long completed; /* # of last completed gp. */
1eba8f84 PM	252
	253	/* End of fields guarded by root rcu_node's lock. */
	254
9f77da9f	255	spinlock_t onofflock; /* exclude on/offline and */
e74f4c45 PM	256	/* starting new GP. Also */
	257	/* protects the following */
	258	/* orphan_cbs fields. */
	259	struct rcu_head orphan_cbs_list; / list of rcu_head structs */
	260	/* orphaned by all CPUs in */
	261	/* a given leaf rcu_node */
	262	/* going offline. */
	263	struct rcu_head *orphan_cbs_tail; / And tail pointer. */
	264	long orphan_qlen; /* Number of orphaned cbs. */
9f77da9f PM	265	spinlock_t fqslock; /* Only one task forcing */
9f77da9f PM	266	/* quiescent states. */
4bcfe055 PM	267	long completed_fqs; /* Value of completed @ snap. */
4bcfe055 PM	268	/* Protected by fqslock. */
9f77da9f PM	269	unsigned long jiffies_force_qs; /* Time at which to invoke */
	270	/* force_quiescent_state(). */
	271	unsigned long n_force_qs; /* Number of calls to */
	272	/* force_quiescent_state(). */
	273	unsigned long n_force_qs_lh; /* ~Number of calls leaving */
	274	/* due to lock unavailable. */
	275	unsigned long n_force_qs_ngp; /* Number of calls leaving */
	276	/* due to no GP active. */
	277	#ifdef CONFIG_RCU_CPU_STALL_DETECTOR
	278	unsigned long gp_start; /* Time at which GP started, */
	279	/* but in jiffies. */
	280	unsigned long jiffies_stall; /* Time at which to check */
	281	/* for CPU stalls. */
	282	#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */
9f77da9f PM	283	};
	284
	285	#ifdef RCU_TREE_NONCORE
6258c4fb IM	286
	287	/*
	288	* RCU implementation internal declarations:
	289	*/
d6714c22 PM	290	extern struct rcu_state rcu_sched_state;
d6714c22 PM	291	DECLARE_PER_CPU(struct rcu_data, rcu_sched_data);
6258c4fb IM	292
	293	extern struct rcu_state rcu_bh_state;
	294	DECLARE_PER_CPU(struct rcu_data, rcu_bh_data);
	295
f41d911f PM	296	#ifdef CONFIG_TREE_PREEMPT_RCU
	297	extern struct rcu_state rcu_preempt_state;
	298	DECLARE_PER_CPU(struct rcu_data, rcu_preempt_data);
	299	#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */
	300
9b2619af	301	#else /* #ifdef RCU_TREE_NONCORE */
9f77da9f	302
9b2619af	303	/* Forward declarations for rcutree_plugin.h */
dbe01350	304	static void rcu_bootup_announce(void);
9b2619af PM	305	long rcu_batches_completed(void);
	306	static void rcu_preempt_note_context_switch(int cpu);
	307	static int rcu_preempted_readers(struct rcu_node *rnp);
b668c9cf	308	#ifdef CONFIG_HOTPLUG_CPU
d3f6bad3 PM	309	static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp,
d3f6bad3 PM	310	unsigned long flags);
b668c9cf	311	#endif /* #ifdef CONFIG_HOTPLUG_CPU */
9b2619af PM	312	#ifdef CONFIG_RCU_CPU_STALL_DETECTOR
	313	static void rcu_print_task_stall(struct rcu_node *rnp);
	314	#endif /* #ifdef CONFIG_RCU_CPU_STALL_DETECTOR */
	315	static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp);
	316	#ifdef CONFIG_HOTPLUG_CPU
237c80c5 PM	317	static int rcu_preempt_offline_tasks(struct rcu_state *rsp,
	318	struct rcu_node *rnp,
	319	struct rcu_data *rdp);
9b2619af PM	320	static void rcu_preempt_offline_cpu(int cpu);
	321	#endif /* #ifdef CONFIG_HOTPLUG_CPU */
	322	static void rcu_preempt_check_callbacks(int cpu);
	323	static void rcu_preempt_process_callbacks(void);
	324	void call_rcu(struct rcu_head head, void (func)(struct rcu_head *rcu));
	325	static int rcu_preempt_pending(int cpu);
	326	static int rcu_preempt_needs_cpu(int cpu);
	327	static void __cpuinit rcu_preempt_init_percpu_data(int cpu);
e74f4c45	328	static void rcu_preempt_send_cbs_to_orphanage(void);
9b2619af PM	329	static void __init __rcu_init_preempt(void);
	330
	331	#endif /* #else #ifdef RCU_TREE_NONCORE */