3 The most basic primitive for locking is spinlock.
5 static DEFINE_SPINLOCK(xxx_lock);
9 spin_lock_irqsave(&xxx_lock, flags);
10 ... critical section here ..
11 spin_unlock_irqrestore(&xxx_lock, flags);
13 The above is always safe. It will disable interrupts _locally_, but the
14 spinlock itself will guarantee the global lock, so it will guarantee that
15 there is only one thread-of-control within the region(s) protected by that
16 lock. This works well even under UP. The above sequence under UP
17 essentially is just the same as doing
21 save_flags(flags); cli();
22 ... critical section ...
25 so the code does _not_ need to worry about UP vs SMP issues: the spinlocks
26 work correctly under both (and spinlocks are actually more efficient on
27 architectures that allow doing the "save_flags + cli" in one operation).
29 NOTE! Implications of spin_locks for memory are further described in:
31 Documentation/memory-barriers.txt
33 (6) UNLOCK operations.
35 The above is usually pretty simple (you usually need and want only one
36 spinlock for most things - using more than one spinlock can make things a
37 lot more complex and even slower and is usually worth it only for
38 sequences that you _know_ need to be split up: avoid it at all cost if you
39 aren't sure). HOWEVER, it _does_ mean that if you have some code that does
42 .. critical section ..
45 and another sequence that does
47 spin_lock_irqsave(flags);
48 .. critical section ..
49 spin_unlock_irqrestore(flags);
51 then they are NOT mutually exclusive, and the critical regions can happen
52 at the same time on two different CPU's. That's fine per se, but the
53 critical regions had better be critical for different things (ie they
54 can't stomp on each other).
56 The above is a problem mainly if you end up mixing code - for example the
57 routines in ll_rw_block() tend to use cli/sti to protect the atomicity of
58 their actions, and if a driver uses spinlocks instead then you should
59 think about issues like the above.
61 This is really the only really hard part about spinlocks: once you start
62 using spinlocks they tend to expand to areas you might not have noticed
63 before, because you have to make sure the spinlocks correctly protect the
64 shared data structures _everywhere_ they are used. The spinlocks are most
65 easily added to places that are completely independent of other code (for
66 example, internal driver data structures that nobody else ever touches).
68 NOTE! The spin-lock is safe only when you _also_ use the lock itself
69 to do locking across CPU's, which implies that EVERYTHING that
70 touches a shared variable has to agree about the spinlock they want
75 Lesson 2: reader-writer spinlocks.
77 If your data accesses have a very natural pattern where you usually tend
78 to mostly read from the shared variables, the reader-writer locks
79 (rw_lock) versions of the spinlocks are sometimes useful. They allow multiple
80 readers to be in the same critical region at once, but if somebody wants
81 to change the variables it has to get an exclusive write lock.
83 NOTE! reader-writer locks require more atomic memory operations than
84 simple spinlocks. Unless the reader critical section is long, you
85 are better off just using spinlocks.
87 The routines look the same as above:
89 rwlock_t xxx_lock = RW_LOCK_UNLOCKED;
93 read_lock_irqsave(&xxx_lock, flags);
94 .. critical section that only reads the info ...
95 read_unlock_irqrestore(&xxx_lock, flags);
97 write_lock_irqsave(&xxx_lock, flags);
98 .. read and write exclusive access to the info ...
99 write_unlock_irqrestore(&xxx_lock, flags);
101 The above kind of lock may be useful for complex data structures like
102 linked lists, especially searching for entries without changing the list
103 itself. The read lock allows many concurrent readers. Anything that
104 _changes_ the list will have to get the write lock.
106 NOTE! RCU is better for list traversal, but requires careful
107 attention to design detail (see Documentation/RCU/listRCU.txt).
109 Also, you cannot "upgrade" a read-lock to a write-lock, so if you at _any_
110 time need to do any changes (even if you don't do it every time), you have
111 to get the write-lock at the very beginning.
113 NOTE! We are working hard to remove reader-writer spinlocks in most
114 cases, so please don't add a new one without consensus. (Instead, see
115 Documentation/RCU/rcu.txt for complete information.)
119 Lesson 3: spinlocks revisited.
121 The single spin-lock primitives above are by no means the only ones. They
122 are the most safe ones, and the ones that work under all circumstances,
123 but partly _because_ they are safe they are also fairly slow. They are
124 much faster than a generic global cli/sti pair, but slower than they'd
125 need to be, because they do have to disable interrupts (which is just a
126 single instruction on a x86, but it's an expensive one - and on other
127 architectures it can be worse).
129 If you have a case where you have to protect a data structure across
130 several CPU's and you want to use spinlocks you can potentially use
131 cheaper versions of the spinlocks. IFF you know that the spinlocks are
132 never used in interrupt handlers, you can use the non-irq versions:
138 (and the equivalent read-write versions too, of course). The spinlock will
139 guarantee the same kind of exclusive access, and it will be much faster.
140 This is useful if you know that the data in question is only ever
141 manipulated from a "process context", ie no interrupts involved.
143 The reasons you mustn't use these versions if you have interrupts that
144 play with the spinlock is that you can get deadlocks:
148 <- interrupt comes in:
151 where an interrupt tries to lock an already locked variable. This is ok if
152 the other interrupt happens on another CPU, but it is _not_ ok if the
153 interrupt happens on the same CPU that already holds the lock, because the
154 lock will obviously never be released (because the interrupt is waiting
155 for the lock, and the lock-holder is interrupted by the interrupt and will
156 not continue until the interrupt has been processed).
158 (This is also the reason why the irq-versions of the spinlocks only need
159 to disable the _local_ interrupts - it's ok to use spinlocks in interrupts
160 on other CPU's, because an interrupt on another CPU doesn't interrupt the
161 CPU that holds the lock, so the lock-holder can continue and eventually
164 Note that you can be clever with read-write locks and interrupts. For
165 example, if you know that the interrupt only ever gets a read-lock, then
166 you can use a non-irq version of read locks everywhere - because they
167 don't block on each other (and thus there is no dead-lock wrt interrupts.
168 But when you do the write-lock, you have to use the irq-safe version.
170 For an example of being clever with rw-locks, see the "waitqueue_lock"
171 handling in kernel/sched.c - nothing ever _changes_ a wait-queue from
172 within an interrupt, they only read the queue in order to know whom to
173 wake up. So read-locks are safe (which is good: they are very common
174 indeed), while write-locks need to protect themselves against interrupts.
180 Reference information:
182 For dynamic initialization, use spin_lock_init() or rwlock_init() as
186 rwlock_t xxx_rw_lock;
188 static int __init xxx_init(void)
190 spin_lock_init(&xxx_lock);
191 rwlock_init(&xxx_rw_lock);
195 module_init(xxx_init);
197 For static initialization, use DEFINE_SPINLOCK() / DEFINE_RWLOCK() or
198 __SPIN_LOCK_UNLOCKED() / __RW_LOCK_UNLOCKED() as appropriate.
200 SPIN_LOCK_UNLOCKED and RW_LOCK_UNLOCKED are deprecated. These interfere
201 with lockdep state tracking.
203 Most of the time, you can simply turn:
204 static spinlock_t xxx_lock = SPIN_LOCK_UNLOCKED;
206 static DEFINE_SPINLOCK(xxx_lock);
208 Static structure member variables go from:
211 .lock = SPIN_LOCK_UNLOCKED;
217 .lock = __SPIN_LOCK_UNLOCKED(bar.lock);
220 Declaration of static rw_locks undergo a similar transformation.