5
By: David Howells <dhowells@redhat.com>
6
Paul E. McKenney <paulmck@linux.vnet.ibm.com>
9
Linux provides a number of features that can be used to implement circular
10
buffering. There are two sets of such features:
12
(1) Convenience functions for determining information about power-of-2 sized
15
(2) Memory barriers for when the producer and the consumer of objects in the
16
buffer don't want to share a lock.
18
To use these facilities, as discussed below, there needs to be just one
19
producer and just one consumer. It is possible to handle multiple producers by
20
serialising them, and to handle multiple consumers by serialising them.
25
(*) What is a circular buffer?
27
(*) Measuring power-of-2 buffers.
29
(*) Using memory barriers with circular buffers.
34
==========================
35
WHAT IS A CIRCULAR BUFFER?
36
==========================
38
First of all, what is a circular buffer? A circular buffer is a buffer of
39
fixed, finite size into which there are two indices:
41
(1) A 'head' index - the point at which the producer inserts items into the
44
(2) A 'tail' index - the point at which the consumer finds the next item in
47
Typically when the tail pointer is equal to the head pointer, the buffer is
48
empty; and the buffer is full when the head pointer is one less than the tail
51
The head index is incremented when items are added, and the tail index when
52
items are removed. The tail index should never jump the head index, and both
53
indices should be wrapped to 0 when they reach the end of the buffer, thus
54
allowing an infinite amount of data to flow through the buffer.
56
Typically, items will all be of the same unit size, but this isn't strictly
57
required to use the techniques below. The indices can be increased by more
58
than 1 if multiple items or variable-sized items are to be included in the
59
buffer, provided that neither index overtakes the other. The implementer must
60
be careful, however, as a region more than one unit in size may wrap the end of
61
the buffer and be broken into two segments.
64
============================
65
MEASURING POWER-OF-2 BUFFERS
66
============================
68
Calculation of the occupancy or the remaining capacity of an arbitrarily sized
69
circular buffer would normally be a slow operation, requiring the use of a
70
modulus (divide) instruction. However, if the buffer is of a power-of-2 size,
71
then a much quicker bitwise-AND instruction can be used instead.
73
Linux provides a set of macros for handling power-of-2 circular buffers. These
74
can be made use of by:
76
#include <linux/circ_buf.h>
80
(*) Measure the remaining capacity of a buffer:
82
CIRC_SPACE(head_index, tail_index, buffer_size);
84
This returns the amount of space left in the buffer[1] into which items
88
(*) Measure the maximum consecutive immediate space in a buffer:
90
CIRC_SPACE_TO_END(head_index, tail_index, buffer_size);
92
This returns the amount of consecutive space left in the buffer[1] into
93
which items can be immediately inserted without having to wrap back to the
94
beginning of the buffer.
97
(*) Measure the occupancy of a buffer:
99
CIRC_CNT(head_index, tail_index, buffer_size);
101
This returns the number of items currently occupying a buffer[2].
104
(*) Measure the non-wrapping occupancy of a buffer:
106
CIRC_CNT_TO_END(head_index, tail_index, buffer_size);
108
This returns the number of consecutive items[2] that can be extracted from
109
the buffer without having to wrap back to the beginning of the buffer.
112
Each of these macros will nominally return a value between 0 and buffer_size-1,
115
[1] CIRC_SPACE*() are intended to be used in the producer. To the producer
116
they will return a lower bound as the producer controls the head index,
117
but the consumer may still be depleting the buffer on another CPU and
118
moving the tail index.
120
To the consumer it will show an upper bound as the producer may be busy
123
[2] CIRC_CNT*() are intended to be used in the consumer. To the consumer they
124
will return a lower bound as the consumer controls the tail index, but the
125
producer may still be filling the buffer on another CPU and moving the
128
To the producer it will show an upper bound as the consumer may be busy
131
[3] To a third party, the order in which the writes to the indices by the
132
producer and consumer become visible cannot be guaranteed as they are
133
independent and may be made on different CPUs - so the result in such a
134
situation will merely be a guess, and may even be negative.
137
===========================================
138
USING MEMORY BARRIERS WITH CIRCULAR BUFFERS
139
===========================================
141
By using memory barriers in conjunction with circular buffers, you can avoid
144
(1) use a single lock to govern access to both ends of the buffer, thus
145
allowing the buffer to be filled and emptied at the same time; and
147
(2) use atomic counter operations.
149
There are two sides to this: the producer that fills the buffer, and the
150
consumer that empties it. Only one thing should be filling a buffer at any one
151
time, and only one thing should be emptying a buffer at any one time, but the
152
two sides can operate simultaneously.
158
The producer will look something like this:
160
spin_lock(&producer_lock);
162
unsigned long head = buffer->head;
163
unsigned long tail = ACCESS_ONCE(buffer->tail);
165
if (CIRC_SPACE(head, tail, buffer->size) >= 1) {
166
/* insert one item into the buffer */
167
struct item *item = buffer[head];
171
smp_wmb(); /* commit the item before incrementing the head */
173
buffer->head = (head + 1) & (buffer->size - 1);
175
/* wake_up() will make sure that the head is committed before
176
* waking anyone up */
180
spin_unlock(&producer_lock);
182
This will instruct the CPU that the contents of the new item must be written
183
before the head index makes it available to the consumer and then instructs the
184
CPU that the revised head index must be written before the consumer is woken.
186
Note that wake_up() doesn't have to be the exact mechanism used, but whatever
187
is used must guarantee a (write) memory barrier between the update of the head
188
index and the change of state of the consumer, if a change of state occurs.
194
The consumer will look something like this:
196
spin_lock(&consumer_lock);
198
unsigned long head = ACCESS_ONCE(buffer->head);
199
unsigned long tail = buffer->tail;
201
if (CIRC_CNT(head, tail, buffer->size) >= 1) {
202
/* read index before reading contents at that index */
203
smp_read_barrier_depends();
205
/* extract one item from the buffer */
206
struct item *item = buffer[tail];
210
smp_mb(); /* finish reading descriptor before incrementing tail */
212
buffer->tail = (tail + 1) & (buffer->size - 1);
215
spin_unlock(&consumer_lock);
217
This will instruct the CPU to make sure the index is up to date before reading
218
the new item, and then it shall make sure the CPU has finished reading the item
219
before it writes the new tail pointer, which will erase the item.
222
Note the use of ACCESS_ONCE() in both algorithms to read the opposition index.
223
This prevents the compiler from discarding and reloading its cached value -
224
which some compilers will do across smp_read_barrier_depends(). This isn't
225
strictly needed if you can be sure that the opposition index will _only_ be
233
See also Documentation/memory-barriers.txt for a description of Linux's memory