cpupri.c 6.8 KB
Newer Older
1
/*
2
 *  kernel/sched/cpupri.c
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
 *
 *  CPU priority management
 *
 *  Copyright (C) 2007-2008 Novell
 *
 *  Author: Gregory Haskins <ghaskins@novell.com>
 *
 *  This code tracks the priority of each CPU so that global migration
 *  decisions are easy to calculate.  Each CPU can be in a state as follows:
 *
 *                 (INVALID), IDLE, NORMAL, RT1, ... RT99
 *
 *  going from the lowest priority to the highest.  CPUs in the INVALID state
 *  are not eligible for routing.  The system maintains this state with
 *  a 2 dimensional bitmap (the first for priority class, the second for cpus
 *  in that class).  Therefore a typical application without affinity
 *  restrictions can find a suitable CPU with O(1) complexity (e.g. two bit
 *  searches).  For tasks with affinity restrictions, the algorithm has a
 *  worst case complexity of O(min(102, nr_domcpus)), though the scenario that
 *  yields the worst case search is fairly contrived.
 *
 *  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; version 2
 *  of the License.
 */

30
#include <linux/gfp.h>
31 32
#include <linux/sched.h>
#include <linux/sched/rt.h>
33
#include <linux/slab.h>
34
#include "cpupri.h"
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56

/* Convert between a 140 based task->prio, and our 102 based cpupri */
static int convert_prio(int prio)
{
	int cpupri;

	if (prio == CPUPRI_INVALID)
		cpupri = CPUPRI_INVALID;
	else if (prio == MAX_PRIO)
		cpupri = CPUPRI_IDLE;
	else if (prio >= MAX_RT_PRIO)
		cpupri = CPUPRI_NORMAL;
	else
		cpupri = MAX_RT_PRIO - prio + 1;

	return cpupri;
}

/**
 * cpupri_find - find the best (lowest-pri) CPU in the system
 * @cp: The cpupri context
 * @p: The task
57
 * @lowest_mask: A mask to fill in with selected CPUs (or NULL)
58 59
 *
 * Note: This function returns the recommended CPUs as calculated during the
60
 * current invocation.  By the time the call returns, the CPUs may have in
61 62 63 64 65
 * fact changed priorities any number of times.  While not ideal, it is not
 * an issue of correctness since the normal rebalancer logic will correct
 * any discrepancies created by racing against the uncertainty of the current
 * priority configuration.
 *
66
 * Return: (int)bool - CPUs were found
67 68
 */
int cpupri_find(struct cpupri *cp, struct task_struct *p,
69
		struct cpumask *lowest_mask)
70
{
71 72
	int idx = 0;
	int task_pri = convert_prio(p->prio);
73

74
	BUG_ON(task_pri >= CPUPRI_NR_PRIORITIES);
75 76

	for (idx = 0; idx < task_pri; idx++) {
77
		struct cpupri_vec *vec  = &cp->pri_to_cpu[idx];
78
		int skip = 0;
79

80
		if (!atomic_read(&(vec)->count))
81
			skip = 1;
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
		/*
		 * When looking at the vector, we need to read the counter,
		 * do a memory barrier, then read the mask.
		 *
		 * Note: This is still all racey, but we can deal with it.
		 *  Ideally, we only want to look at masks that are set.
		 *
		 *  If a mask is not set, then the only thing wrong is that we
		 *  did a little more work than necessary.
		 *
		 *  If we read a zero count but the mask is set, because of the
		 *  memory barriers, that can only happen when the highest prio
		 *  task for a run queue has left the run queue, in which case,
		 *  it will be followed by a pull. If the task we are processing
		 *  fails to find a proper place to go, that pull request will
		 *  pull this task if the run queue is running at a lower
		 *  priority.
		 */
		smp_rmb();
101

102 103 104 105
		/* Need to do the rmb for every iteration */
		if (skip)
			continue;

106
		if (cpumask_any_and(tsk_cpus_allowed(p), vec->mask) >= nr_cpu_ids)
107 108
			continue;

109
		if (lowest_mask) {
110
			cpumask_and(lowest_mask, tsk_cpus_allowed(p), vec->mask);
111 112 113 114 115 116 117 118 119 120 121 122 123

			/*
			 * We have to ensure that we have at least one bit
			 * still set in the array, since the map could have
			 * been concurrently emptied between the first and
			 * second reads of vec->mask.  If we hit this
			 * condition, simply act as though we never hit this
			 * priority level and continue on.
			 */
			if (cpumask_any(lowest_mask) >= nr_cpu_ids)
				continue;
		}

124 125 126 127 128 129 130 131 132 133
		return 1;
	}

	return 0;
}

/**
 * cpupri_set - update the cpu priority setting
 * @cp: The cpupri context
 * @cpu: The target cpu
134
 * @newpri: The priority (INVALID-RT99) to assign to this CPU
135 136 137 138 139 140 141
 *
 * Note: Assumes cpu_rq(cpu)->lock is locked
 *
 * Returns: (void)
 */
void cpupri_set(struct cpupri *cp, int cpu, int newpri)
{
142 143 144
	int *currpri = &cp->cpu_to_pri[cpu];
	int oldpri = *currpri;
	int do_mb = 0;
145 146 147 148 149 150 151 152 153 154

	newpri = convert_prio(newpri);

	BUG_ON(newpri >= CPUPRI_NR_PRIORITIES);

	if (newpri == oldpri)
		return;

	/*
	 * If the cpu was currently mapped to a different value, we
155 156
	 * need to map it to the new value then remove the old value.
	 * Note, we must add the new value first, otherwise we risk the
157
	 * cpu being missed by the priority loop in cpupri_find.
158 159 160 161
	 */
	if (likely(newpri != CPUPRI_INVALID)) {
		struct cpupri_vec *vec = &cp->pri_to_cpu[newpri];

162
		cpumask_set_cpu(cpu, vec->mask);
163 164 165 166 167
		/*
		 * When adding a new vector, we update the mask first,
		 * do a write memory barrier, and then update the count, to
		 * make sure the vector is visible when count is set.
		 */
168
		smp_mb__before_atomic();
169
		atomic_inc(&(vec)->count);
170
		do_mb = 1;
171
	}
172 173 174
	if (likely(oldpri != CPUPRI_INVALID)) {
		struct cpupri_vec *vec  = &cp->pri_to_cpu[oldpri];

175 176 177 178 179 180 181 182 183 184 185 186 187
		/*
		 * Because the order of modification of the vec->count
		 * is important, we must make sure that the update
		 * of the new prio is seen before we decrement the
		 * old prio. This makes sure that the loop sees
		 * one or the other when we raise the priority of
		 * the run queue. We don't care about when we lower the
		 * priority, as that will trigger an rt pull anyway.
		 *
		 * We only need to do a memory barrier if we updated
		 * the new priority vec.
		 */
		if (do_mb)
188
			smp_mb__after_atomic();
189

190 191 192 193 194
		/*
		 * When removing from the vector, we decrement the counter first
		 * do a memory barrier and then clear the mask.
		 */
		atomic_dec(&(vec)->count);
195
		smp_mb__after_atomic();
196 197
		cpumask_clear_cpu(cpu, vec->mask);
	}
198 199 200 201 202 203 204 205

	*currpri = newpri;
}

/**
 * cpupri_init - initialize the cpupri structure
 * @cp: The cpupri context
 *
206
 * Return: -ENOMEM on memory allocation failure.
207
 */
208
int cpupri_init(struct cpupri *cp)
209 210 211 212 213 214 215 216
{
	int i;

	memset(cp, 0, sizeof(*cp));

	for (i = 0; i < CPUPRI_NR_PRIORITIES; i++) {
		struct cpupri_vec *vec = &cp->pri_to_cpu[i];

217
		atomic_set(&vec->count, 0);
218
		if (!zalloc_cpumask_var(&vec->mask, GFP_KERNEL))
219
			goto cleanup;
220 221
	}

222 223 224 225
	cp->cpu_to_pri = kcalloc(nr_cpu_ids, sizeof(int), GFP_KERNEL);
	if (!cp->cpu_to_pri)
		goto cleanup;

226 227
	for_each_possible_cpu(i)
		cp->cpu_to_pri[i] = CPUPRI_INVALID;
228

229 230 231 232 233 234
	return 0;

cleanup:
	for (i--; i >= 0; i--)
		free_cpumask_var(cp->pri_to_cpu[i].mask);
	return -ENOMEM;
235 236
}

237 238 239 240 241 242 243
/**
 * cpupri_cleanup - clean up the cpupri structure
 * @cp: The cpupri context
 */
void cpupri_cleanup(struct cpupri *cp)
{
	int i;
244

245
	kfree(cp->cpu_to_pri);
246 247 248
	for (i = 0; i < CPUPRI_NR_PRIORITIES; i++)
		free_cpumask_var(cp->pri_to_cpu[i].mask);
}