Root/Documentation/RCU/listRCU.txt

1Using RCU to Protect Read-Mostly Linked Lists
2
3
4One of the best applications of RCU is to protect read-mostly linked lists
5("struct list_head" in list.h). One big advantage of this approach
6is that all of the required memory barriers are included for you in
7the list macros. This document describes several applications of RCU,
8with the best fits first.
9
10
11Example 1: Read-Side Action Taken Outside of Lock, No In-Place Updates
12
13The best applications are cases where, if reader-writer locking were
14used, the read-side lock would be dropped before taking any action
15based on the results of the search. The most celebrated example is
16the routing table. Because the routing table is tracking the state of
17equipment outside of the computer, it will at times contain stale data.
18Therefore, once the route has been computed, there is no need to hold
19the routing table static during transmission of the packet. After all,
20you can hold the routing table static all you want, but that won't keep
21the external Internet from changing, and it is the state of the external
22Internet that really matters. In addition, routing entries are typically
23added or deleted, rather than being modified in place.
24
25A straightforward example of this use of RCU may be found in the
26system-call auditing support. For example, a reader-writer locked
27implementation of audit_filter_task() might be as follows:
28
29    static enum audit_state audit_filter_task(struct task_struct *tsk)
30    {
31        struct audit_entry *e;
32        enum audit_state state;
33
34        read_lock(&auditsc_lock);
35        /* Note: audit_netlink_sem held by caller. */
36        list_for_each_entry(e, &audit_tsklist, list) {
37            if (audit_filter_rules(tsk, &e->rule, NULL, &state)) {
38                read_unlock(&auditsc_lock);
39                return state;
40            }
41        }
42        read_unlock(&auditsc_lock);
43        return AUDIT_BUILD_CONTEXT;
44    }
45
46Here the list is searched under the lock, but the lock is dropped before
47the corresponding value is returned. By the time that this value is acted
48on, the list may well have been modified. This makes sense, since if
49you are turning auditing off, it is OK to audit a few extra system calls.
50
51This means that RCU can be easily applied to the read side, as follows:
52
53    static enum audit_state audit_filter_task(struct task_struct *tsk)
54    {
55        struct audit_entry *e;
56        enum audit_state state;
57
58        rcu_read_lock();
59        /* Note: audit_netlink_sem held by caller. */
60        list_for_each_entry_rcu(e, &audit_tsklist, list) {
61            if (audit_filter_rules(tsk, &e->rule, NULL, &state)) {
62                rcu_read_unlock();
63                return state;
64            }
65        }
66        rcu_read_unlock();
67        return AUDIT_BUILD_CONTEXT;
68    }
69
70The read_lock() and read_unlock() calls have become rcu_read_lock()
71and rcu_read_unlock(), respectively, and the list_for_each_entry() has
72become list_for_each_entry_rcu(). The _rcu() list-traversal primitives
73insert the read-side memory barriers that are required on DEC Alpha CPUs.
74
75The changes to the update side are also straightforward. A reader-writer
76lock might be used as follows for deletion and insertion:
77
78    static inline int audit_del_rule(struct audit_rule *rule,
79                     struct list_head *list)
80    {
81        struct audit_entry *e;
82
83        write_lock(&auditsc_lock);
84        list_for_each_entry(e, list, list) {
85            if (!audit_compare_rule(rule, &e->rule)) {
86                list_del(&e->list);
87                write_unlock(&auditsc_lock);
88                return 0;
89            }
90        }
91        write_unlock(&auditsc_lock);
92        return -EFAULT; /* No matching rule */
93    }
94
95    static inline int audit_add_rule(struct audit_entry *entry,
96                     struct list_head *list)
97    {
98        write_lock(&auditsc_lock);
99        if (entry->rule.flags & AUDIT_PREPEND) {
100            entry->rule.flags &= ~AUDIT_PREPEND;
101            list_add(&entry->list, list);
102        } else {
103            list_add_tail(&entry->list, list);
104        }
105        write_unlock(&auditsc_lock);
106        return 0;
107    }
108
109Following are the RCU equivalents for these two functions:
110
111    static inline int audit_del_rule(struct audit_rule *rule,
112                     struct list_head *list)
113    {
114        struct audit_entry *e;
115
116        /* Do not use the _rcu iterator here, since this is the only
117         * deletion routine. */
118        list_for_each_entry(e, list, list) {
119            if (!audit_compare_rule(rule, &e->rule)) {
120                list_del_rcu(&e->list);
121                call_rcu(&e->rcu, audit_free_rule);
122                return 0;
123            }
124        }
125        return -EFAULT; /* No matching rule */
126    }
127
128    static inline int audit_add_rule(struct audit_entry *entry,
129                     struct list_head *list)
130    {
131        if (entry->rule.flags & AUDIT_PREPEND) {
132            entry->rule.flags &= ~AUDIT_PREPEND;
133            list_add_rcu(&entry->list, list);
134        } else {
135            list_add_tail_rcu(&entry->list, list);
136        }
137        return 0;
138    }
139
140Normally, the write_lock() and write_unlock() would be replaced by
141a spin_lock() and a spin_unlock(), but in this case, all callers hold
142audit_netlink_sem, so no additional locking is required. The auditsc_lock
143can therefore be eliminated, since use of RCU eliminates the need for
144writers to exclude readers. Normally, the write_lock() calls would
145be converted into spin_lock() calls.
146
147The list_del(), list_add(), and list_add_tail() primitives have been
148replaced by list_del_rcu(), list_add_rcu(), and list_add_tail_rcu().
149The _rcu() list-manipulation primitives add memory barriers that are
150needed on weakly ordered CPUs (most of them!). The list_del_rcu()
151primitive omits the pointer poisoning debug-assist code that would
152otherwise cause concurrent readers to fail spectacularly.
153
154So, when readers can tolerate stale data and when entries are either added
155or deleted, without in-place modification, it is very easy to use RCU!
156
157
158Example 2: Handling In-Place Updates
159
160The system-call auditing code does not update auditing rules in place.
161However, if it did, reader-writer-locked code to do so might look as
162follows (presumably, the field_count is only permitted to decrease,
163otherwise, the added fields would need to be filled in):
164
165    static inline int audit_upd_rule(struct audit_rule *rule,
166                     struct list_head *list,
167                     __u32 newaction,
168                     __u32 newfield_count)
169    {
170        struct audit_entry *e;
171        struct audit_newentry *ne;
172
173        write_lock(&auditsc_lock);
174        /* Note: audit_netlink_sem held by caller. */
175        list_for_each_entry(e, list, list) {
176            if (!audit_compare_rule(rule, &e->rule)) {
177                e->rule.action = newaction;
178                e->rule.file_count = newfield_count;
179                write_unlock(&auditsc_lock);
180                return 0;
181            }
182        }
183        write_unlock(&auditsc_lock);
184        return -EFAULT; /* No matching rule */
185    }
186
187The RCU version creates a copy, updates the copy, then replaces the old
188entry with the newly updated entry. This sequence of actions, allowing
189concurrent reads while doing a copy to perform an update, is what gives
190RCU ("read-copy update") its name. The RCU code is as follows:
191
192    static inline int audit_upd_rule(struct audit_rule *rule,
193                     struct list_head *list,
194                     __u32 newaction,
195                     __u32 newfield_count)
196    {
197        struct audit_entry *e;
198        struct audit_newentry *ne;
199
200        list_for_each_entry(e, list, list) {
201            if (!audit_compare_rule(rule, &e->rule)) {
202                ne = kmalloc(sizeof(*entry), GFP_ATOMIC);
203                if (ne == NULL)
204                    return -ENOMEM;
205                audit_copy_rule(&ne->rule, &e->rule);
206                ne->rule.action = newaction;
207                ne->rule.file_count = newfield_count;
208                list_replace_rcu(e, ne);
209                call_rcu(&e->rcu, audit_free_rule);
210                return 0;
211            }
212        }
213        return -EFAULT; /* No matching rule */
214    }
215
216Again, this assumes that the caller holds audit_netlink_sem. Normally,
217the reader-writer lock would become a spinlock in this sort of code.
218
219
220Example 3: Eliminating Stale Data
221
222The auditing examples above tolerate stale data, as do most algorithms
223that are tracking external state. Because there is a delay from the
224time the external state changes before Linux becomes aware of the change,
225additional RCU-induced staleness is normally not a problem.
226
227However, there are many examples where stale data cannot be tolerated.
228One example in the Linux kernel is the System V IPC (see the ipc_lock()
229function in ipc/util.c). This code checks a "deleted" flag under a
230per-entry spinlock, and, if the "deleted" flag is set, pretends that the
231entry does not exist. For this to be helpful, the search function must
232return holding the per-entry spinlock, as ipc_lock() does in fact do.
233
234Quick Quiz: Why does the search function need to return holding the
235    per-entry lock for this deleted-flag technique to be helpful?
236
237If the system-call audit module were to ever need to reject stale data,
238one way to accomplish this would be to add a "deleted" flag and a "lock"
239spinlock to the audit_entry structure, and modify audit_filter_task()
240as follows:
241
242    static enum audit_state audit_filter_task(struct task_struct *tsk)
243    {
244        struct audit_entry *e;
245        enum audit_state state;
246
247        rcu_read_lock();
248        list_for_each_entry_rcu(e, &audit_tsklist, list) {
249            if (audit_filter_rules(tsk, &e->rule, NULL, &state)) {
250                spin_lock(&e->lock);
251                if (e->deleted) {
252                    spin_unlock(&e->lock);
253                    rcu_read_unlock();
254                    return AUDIT_BUILD_CONTEXT;
255                }
256                rcu_read_unlock();
257                return state;
258            }
259        }
260        rcu_read_unlock();
261        return AUDIT_BUILD_CONTEXT;
262    }
263
264Note that this example assumes that entries are only added and deleted.
265Additional mechanism is required to deal correctly with the
266update-in-place performed by audit_upd_rule(). For one thing,
267audit_upd_rule() would need additional memory barriers to ensure
268that the list_add_rcu() was really executed before the list_del_rcu().
269
270The audit_del_rule() function would need to set the "deleted"
271flag under the spinlock as follows:
272
273    static inline int audit_del_rule(struct audit_rule *rule,
274                     struct list_head *list)
275    {
276        struct audit_entry *e;
277
278        /* Do not need to use the _rcu iterator here, since this
279         * is the only deletion routine. */
280        list_for_each_entry(e, list, list) {
281            if (!audit_compare_rule(rule, &e->rule)) {
282                spin_lock(&e->lock);
283                list_del_rcu(&e->list);
284                e->deleted = 1;
285                spin_unlock(&e->lock);
286                call_rcu(&e->rcu, audit_free_rule);
287                return 0;
288            }
289        }
290        return -EFAULT; /* No matching rule */
291    }
292
293
294Summary
295
296Read-mostly list-based data structures that can tolerate stale data are
297the most amenable to use of RCU. The simplest case is where entries are
298either added or deleted from the data structure (or atomically modified
299in place), but non-atomic in-place modifications can be handled by making
300a copy, updating the copy, then replacing the original with the copy.
301If stale data cannot be tolerated, then a "deleted" flag may be used
302in conjunction with a per-entry spinlock in order to allow the search
303function to reject newly deleted data.
304
305
306Answer to Quick Quiz
307    Why does the search function need to return holding the per-entry
308    lock for this deleted-flag technique to be helpful?
309
310    If the search function drops the per-entry lock before returning,
311    then the caller will be processing stale data in any case. If it
312    is really OK to be processing stale data, then you don't need a
313    "deleted" flag. If processing stale data really is a problem,
314    then you need to hold the per-entry lock across all of the code
315    that uses the value that was returned.
316

Archive Download this file



interactive