]> git.proxmox.com Git - ceph.git/blame - ceph/src/dpdk/examples/performance-thread/common/lthread_queue.h
bump version to 12.2.12-pve1
[ceph.git] / ceph / src / dpdk / examples / performance-thread / common / lthread_queue.h
CommitLineData
7c673cae
FG
1/*
2 *-
3 * BSD LICENSE
4 *
5 * Copyright(c) 2015 Intel Corporation. All rights reserved.
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 *
12 * * Redistributions of source code must retain the above copyright
13 * notice, this list of conditions and the following disclaimer.
14 * * Redistributions in binary form must reproduce the above copyright
15 * notice, this list of conditions and the following disclaimer in
16 * the documentation and/or other materials provided with the
17 * distribution.
18 * * Neither the name of Intel Corporation nor the names of its
19 * contributors may be used to endorse or promote products derived
20 * from this software without specific prior written permission.
21 *
22 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
23 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
24 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
25 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
26 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
27 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
28 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
29 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
30 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
31 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
32 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 */
34
35/*
36 * Some portions of this software is derived from the producer
37 * consumer queues described by Dmitry Vyukov and published here
38 * http://www.1024cores.net
39 *
40 * Copyright (c) 2010-2011 Dmitry Vyukov. All rights reserved.
41 * Redistribution and use in source and binary forms, with or without
42 * modification, are permitted provided that the following conditions
43 * are met:
44 *
45 * 1. Redistributions of source code must retain the above copyright notice,
46 * this list of conditions and the following disclaimer.
47 *
48 * 2. Redistributions in binary form must reproduce the above copyright notice,
49 * this list of conditions and the following disclaimer in the documentation
50 * and/or other materials provided with the distribution.
51 *
52 * THIS SOFTWARE IS PROVIDED BY DMITRY VYUKOV "AS IS"
53 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
54 * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
55 * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL DMITRY VYUKOV OR CONTRIBUTORS
56 * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY,
57 * OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT
58 * OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
59 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
60 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE
61 * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
62 * EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
63 *
64 * The views and conclusions contained in the software and documentation are
65 * those of the authors and should not be interpreted as representing official
66 * policies, either expressed or implied, of Dmitry Vyukov.
67 */
68
69#ifndef LTHREAD_QUEUE_H_
70#define LTHREAD_QUEUE_H_
71
72#include <string.h>
73
74#include <rte_prefetch.h>
75#include <rte_per_lcore.h>
76
77#include "lthread_int.h"
78#include "lthread.h"
79#include "lthread_diag.h"
80#include "lthread_pool.h"
81
82struct lthread_queue;
83
84/*
85 * This file implements an unbounded FIFO queue based on a lock free
86 * linked list.
87 *
88 * The queue is non-intrusive in that it uses intermediate nodes, and does
89 * not require these nodes to be inserted into the object being placed
90 * in the queue.
91 *
92 * This is slightly more efficient than the very similar queue in lthread_pool
93 * in that it does not have to swing a stub node as the queue becomes empty.
94 *
95 * The queue access functions allocate and free intermediate node
96 * transparently from/to a per scheduler pool ( see lthread_pool.h ).
97 *
98 * The queue provides both MPSC and SPSC insert methods
99 */
100
101/*
102 * define a queue of lthread nodes
103 */
104struct lthread_queue {
105 struct qnode *head;
106 struct qnode *tail __rte_cache_aligned;
107 struct lthread_queue *p;
108 char name[LT_MAX_NAME_SIZE];
109
110 DIAG_COUNT_DEFINE(rd);
111 DIAG_COUNT_DEFINE(wr);
112 DIAG_COUNT_DEFINE(size);
113
114} __rte_cache_aligned;
115
116
117
118static inline struct lthread_queue *
119_lthread_queue_create(const char *name)
120{
121 struct qnode *stub;
122 struct lthread_queue *new_queue;
123
124 new_queue = rte_malloc_socket(NULL, sizeof(struct lthread_queue),
125 RTE_CACHE_LINE_SIZE,
126 rte_socket_id());
127 if (new_queue == NULL)
128 return NULL;
129
130 /* allocated stub node */
131 stub = _qnode_alloc();
132 RTE_ASSERT(stub);
133
134 if (name != NULL)
135 strncpy(new_queue->name, name, sizeof(new_queue->name));
136 new_queue->name[sizeof(new_queue->name)-1] = 0;
137
138 /* initialize queue as empty */
139 stub->next = NULL;
140 new_queue->head = stub;
141 new_queue->tail = stub;
142
143 DIAG_COUNT_INIT(new_queue, rd);
144 DIAG_COUNT_INIT(new_queue, wr);
145 DIAG_COUNT_INIT(new_queue, size);
146
147 return new_queue;
148}
149
150/**
151 * Return true if the queue is empty
152 */
153static inline int __attribute__ ((always_inline))
154_lthread_queue_empty(struct lthread_queue *q)
155{
156 return q->tail == q->head;
157}
158
159
160
161/**
162 * Destroy a queue
163 * fail if queue is not empty
164 */
165static inline int _lthread_queue_destroy(struct lthread_queue *q)
166{
167 if (q == NULL)
168 return -1;
169
170 if (!_lthread_queue_empty(q))
171 return -1;
172
173 _qnode_free(q->head);
174 rte_free(q);
175 return 0;
176}
177
178RTE_DECLARE_PER_LCORE(struct lthread_sched *, this_sched);
179
180/*
181 * Insert a node into a queue
182 * this implementation is multi producer safe
183 */
184static inline struct qnode *__attribute__ ((always_inline))
185_lthread_queue_insert_mp(struct lthread_queue
186 *q, void *data)
187{
188 struct qnode *prev;
189 struct qnode *n = _qnode_alloc();
190
191 if (n == NULL)
192 return NULL;
193
194 /* set object in node */
195 n->data = data;
196 n->next = NULL;
197
198 /* this is an MPSC method, perform a locked update */
199 prev = n;
200 prev =
201 (struct qnode *)__sync_lock_test_and_set((uint64_t *) &(q)->head,
202 (uint64_t) prev);
203 /* there is a window of inconsistency until prev next is set,
204 * which is why remove must retry
205 */
206 prev->next = n;
207
208 DIAG_COUNT_INC(q, wr);
209 DIAG_COUNT_INC(q, size);
210
211 return n;
212}
213
214/*
215 * Insert an node into a queue in single producer mode
216 * this implementation is NOT mult producer safe
217 */
218static inline struct qnode *__attribute__ ((always_inline))
219_lthread_queue_insert_sp(struct lthread_queue
220 *q, void *data)
221{
222 /* allocate a queue node */
223 struct qnode *prev;
224 struct qnode *n = _qnode_alloc();
225
226 if (n == NULL)
227 return NULL;
228
229 /* set data in node */
230 n->data = data;
231 n->next = NULL;
232
233 /* this is an SPSC method, no need for locked exchange operation */
234 prev = q->head;
235 prev->next = q->head = n;
236
237 DIAG_COUNT_INC(q, wr);
238 DIAG_COUNT_INC(q, size);
239
240 return n;
241}
242
243/*
244 * Remove a node from a queue
245 */
246static inline void *__attribute__ ((always_inline))
247_lthread_queue_poll(struct lthread_queue *q)
248{
249 void *data = NULL;
250 struct qnode *tail = q->tail;
251 struct qnode *next = (struct qnode *)tail->next;
252 /*
253 * There is a small window of inconsistency between producer and
254 * consumer whereby the queue may appear empty if consumer and
255 * producer access it at the same time.
256 * The consumer must handle this by retrying
257 */
258
259 if (likely(next != NULL)) {
260 q->tail = next;
261 tail->data = next->data;
262 data = tail->data;
263
264 /* free the node */
265 _qnode_free(tail);
266
267 DIAG_COUNT_INC(q, rd);
268 DIAG_COUNT_DEC(q, size);
269 return data;
270 }
271 return NULL;
272}
273
274/*
275 * Remove a node from a queue
276 */
277static inline void *__attribute__ ((always_inline))
278_lthread_queue_remove(struct lthread_queue *q)
279{
280 void *data = NULL;
281
282 /*
283 * There is a small window of inconsistency between producer and
284 * consumer whereby the queue may appear empty if consumer and
285 * producer access it at the same time. We handle this by retrying
286 */
287 do {
288 data = _lthread_queue_poll(q);
289
290 if (likely(data != NULL)) {
291
292 DIAG_COUNT_INC(q, rd);
293 DIAG_COUNT_DEC(q, size);
294 return data;
295 }
296 rte_compiler_barrier();
297 } while (unlikely(!_lthread_queue_empty(q)));
298 return NULL;
299}
300
301
302#endif /* LTHREAD_QUEUE_H_ */