Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
tjh.dev
kernel
os
linux
1// SPDX-License-Identifier: GPL-2.0
2/*
3 * Floating proportions with flexible aging period
4 *
5 * Copyright (C) 2011, SUSE, Jan Kara <jack@suse.cz>
6 *
7 * The goal of this code is: Given different types of event, measure proportion
8 * of each type of event over time. The proportions are measured with
9 * exponentially decaying history to give smooth transitions. A formula
10 * expressing proportion of event of type 'j' is:
11 *
12 * p_{j} = (\Sum_{i>=0} x_{i,j}/2^{i+1})/(\Sum_{i>=0} x_i/2^{i+1})
13 *
14 * Where x_{i,j} is j's number of events in i-th last time period and x_i is
15 * total number of events in i-th last time period.
16 *
17 * Note that p_{j}'s are normalised, i.e.
18 *
19 * \Sum_{j} p_{j} = 1,
20 *
21 * This formula can be straightforwardly computed by maintaining denominator
22 * (let's call it 'd') and for each event type its numerator (let's call it
23 * 'n_j'). When an event of type 'j' happens, we simply need to do:
24 * n_j++; d++;
25 *
26 * When a new period is declared, we could do:
27 * d /= 2
28 * for each j
29 * n_j /= 2
30 *
31 * To avoid iteration over all event types, we instead shift numerator of event
32 * j lazily when someone asks for a proportion of event j or when event j
33 * occurs. This can bit trivially implemented by remembering last period in
34 * which something happened with proportion of type j.
35 */
36#include <linux/flex_proportions.h>
37
38int fprop_global_init(struct fprop_global *p, gfp_t gfp)
39{
40 int err;
41
42 p->period = 0;
43 /* Use 1 to avoid dealing with periods with 0 events... */
44 err = percpu_counter_init(&p->events, 1, gfp);
45 if (err)
46 return err;
47 seqcount_init(&p->sequence);
48 return 0;
49}
50
51void fprop_global_destroy(struct fprop_global *p)
52{
53 percpu_counter_destroy(&p->events);
54}
55
56/*
57 * Declare @periods new periods. It is upto the caller to make sure period
58 * transitions cannot happen in parallel.
59 *
60 * The function returns true if the proportions are still defined and false
61 * if aging zeroed out all events. This can be used to detect whether declaring
62 * further periods has any effect.
63 */
64bool fprop_new_period(struct fprop_global *p, int periods)
65{
66 s64 events = percpu_counter_sum(&p->events);
67 unsigned long flags;
68
69 /*
70 * Don't do anything if there are no events.
71 */
72 if (events <= 1)
73 return false;
74 local_irq_save(flags);
75 write_seqcount_begin(&p->sequence);
76 if (periods < 64)
77 events -= events >> periods;
78 /* Use addition to avoid losing events happening between sum and set */
79 percpu_counter_add(&p->events, -events);
80 p->period += periods;
81 write_seqcount_end(&p->sequence);
82 local_irq_restore(flags);
83
84 return true;
85}
86
87/*
88 * ---- PERCPU ----
89 */
90#define PROP_BATCH (8*(1+ilog2(nr_cpu_ids)))
91
92int fprop_local_init_percpu(struct fprop_local_percpu *pl, gfp_t gfp)
93{
94 int err;
95
96 err = percpu_counter_init(&pl->events, 0, gfp);
97 if (err)
98 return err;
99 pl->period = 0;
100 raw_spin_lock_init(&pl->lock);
101 return 0;
102}
103
104void fprop_local_destroy_percpu(struct fprop_local_percpu *pl)
105{
106 percpu_counter_destroy(&pl->events);
107}
108
109static void fprop_reflect_period_percpu(struct fprop_global *p,
110 struct fprop_local_percpu *pl)
111{
112 unsigned int period = p->period;
113 unsigned long flags;
114
115 /* Fast path - period didn't change */
116 if (pl->period == period)
117 return;
118 raw_spin_lock_irqsave(&pl->lock, flags);
119 /* Someone updated pl->period while we were spinning? */
120 if (pl->period >= period) {
121 raw_spin_unlock_irqrestore(&pl->lock, flags);
122 return;
123 }
124 /* Aging zeroed our fraction? */
125 if (period - pl->period < BITS_PER_LONG) {
126 s64 val = percpu_counter_read(&pl->events);
127
128 if (val < (nr_cpu_ids * PROP_BATCH))
129 val = percpu_counter_sum(&pl->events);
130
131 percpu_counter_add_batch(&pl->events,
132 -val + (val >> (period-pl->period)), PROP_BATCH);
133 } else
134 percpu_counter_set(&pl->events, 0);
135 pl->period = period;
136 raw_spin_unlock_irqrestore(&pl->lock, flags);
137}
138
139/* Event of type pl happened */
140void __fprop_add_percpu(struct fprop_global *p, struct fprop_local_percpu *pl,
141 long nr)
142{
143 fprop_reflect_period_percpu(p, pl);
144 percpu_counter_add_batch(&pl->events, nr, PROP_BATCH);
145 percpu_counter_add(&p->events, nr);
146}
147
148void fprop_fraction_percpu(struct fprop_global *p,
149 struct fprop_local_percpu *pl,
150 unsigned long *numerator, unsigned long *denominator)
151{
152 unsigned int seq;
153 s64 num, den;
154
155 do {
156 seq = read_seqcount_begin(&p->sequence);
157 fprop_reflect_period_percpu(p, pl);
158 num = percpu_counter_read_positive(&pl->events);
159 den = percpu_counter_read_positive(&p->events);
160 } while (read_seqcount_retry(&p->sequence, seq));
161
162 /*
163 * Make fraction <= 1 and denominator > 0 even in presence of percpu
164 * counter errors
165 */
166 if (den <= num) {
167 if (num)
168 den = num;
169 else
170 den = 1;
171 }
172 *denominator = den;
173 *numerator = num;
174}
175
176/*
177 * Like __fprop_add_percpu() except that event is counted only if the given
178 * type has fraction smaller than @max_frac/FPROP_FRAC_BASE
179 */
180void __fprop_add_percpu_max(struct fprop_global *p,
181 struct fprop_local_percpu *pl, int max_frac, long nr)
182{
183 if (unlikely(max_frac < FPROP_FRAC_BASE)) {
184 unsigned long numerator, denominator;
185 s64 tmp;
186
187 fprop_fraction_percpu(p, pl, &numerator, &denominator);
188 /* Adding 'nr' to fraction exceeds max_frac/FPROP_FRAC_BASE? */
189 tmp = (u64)denominator * max_frac -
190 ((u64)numerator << FPROP_FRAC_SHIFT);
191 if (tmp < 0) {
192 /* Maximum fraction already exceeded? */
193 return;
194 } else if (tmp < nr * (FPROP_FRAC_BASE - max_frac)) {
195 /* Add just enough for the fraction to saturate */
196 nr = div_u64(tmp + FPROP_FRAC_BASE - max_frac - 1,
197 FPROP_FRAC_BASE - max_frac);
198 }
199 }
200
201 __fprop_add_percpu(p, pl, nr);
202}