Something went wrong. Try again.
Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Something went wrong. Try again.
53 kB · 1454 lines
at master
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986987988989990991992993994995996997998999100010011002100310041005100610071008100910101011101210131014101510161017101810191020102110221023102410251026102710281029103010311032103310341035103610371038103910401041104210431044104510461047104810491050105110521053105410551056105710581059106010611062106310641065106610671068106910701071107210731074107510761077107810791080108110821083108410851086108710881089109010911092109310941095109610971098109911001101110211031104110511061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170117111721173117411751176117711781179118011811182118311841185118611871188118911901191119211931194119511961197119811991200120112021203120412051206120712081209121012111212121312141215121612171218121912201221122212231224122512261227122812291230123112321233123412351236123712381239124012411242124312441245124612471248124912501251125212531254125512561257125812591260126112621263126412651266126712681269127012711272127312741275127612771278127912801281128212831284128512861287128812891290129112921293129412951296129712981299130013011302130313041305130613071308130913101311131213131314131513161317131813191320132113221323132413251326132713281329133013311332133313341335133613371338133913401341134213431344134513461347134813491350135113521353135413551356135713581359136013611362136313641365136613671368136913701371137213731374137513761377137813791380138113821383138413851386138713881389139013911392139313941395139613971398139914001401140214031404140514061407140814091410141114121413141414151416141714181419142014211422142314241425142614271428142914301431143214331434143514361437143814391440144114421443144414451446144714481449145014511452145314541455.. _kernel_hacking_lock:
===========================Unreliable Guide To Locking===========================
:Author: Rusty Russell
Introduction============
Welcome, to Rusty's Remarkably Unreliable Guide to Kernel Lockingissues. This document describes the locking systems in the Linux Kernelin 2.6.
With the wide availability of HyperThreading, and preemption in theLinux Kernel, everyone hacking on the kernel needs to know thefundamentals of concurrency and locking for SMP.
The Problem With Concurrency============================
(Skip this if you know what a Race Condition is).
In a normal program, you can increment a counter like so:
::
very_important_count++;
This is what they would expect to happen:
.. table:: Expected Results
+------------------------------------+------------------------------------+ | Instance 1 | Instance 2 | +====================================+====================================+ | read very_important_count (5) | | +------------------------------------+------------------------------------+ | add 1 (6) | | +------------------------------------+------------------------------------+ | write very_important_count (6) | | +------------------------------------+------------------------------------+ | | read very_important_count (6) | +------------------------------------+------------------------------------+ | | add 1 (7) | +------------------------------------+------------------------------------+ | | write very_important_count (7) | +------------------------------------+------------------------------------+
This is what might happen:
.. table:: Possible Results
+------------------------------------+------------------------------------+ | Instance 1 | Instance 2 | +====================================+====================================+ | read very_important_count (5) | | +------------------------------------+------------------------------------+ | | read very_important_count (5) | +------------------------------------+------------------------------------+ | add 1 (6) | | +------------------------------------+------------------------------------+ | | add 1 (6) | +------------------------------------+------------------------------------+ | write very_important_count (6) | | +------------------------------------+------------------------------------+ | | write very_important_count (6) | +------------------------------------+------------------------------------+
Race Conditions and Critical Regions------------------------------------
This overlap, where the result depends on the relative timing ofmultiple tasks, is called a race condition. The piece of code containingthe concurrency issue is called a critical region. And especially sinceLinux starting running on SMP machines, they became one of the majorissues in kernel design and implementation.
Preemption can have the same effect, even if there is only one CPU: bypreempting one task during the critical region, we have exactly the samerace condition. In this case the thread which preempts might run thecritical region itself.
The solution is to recognize when these simultaneous accesses occur, anduse locks to make sure that only one instance can enter the criticalregion at any time. There are many friendly primitives in the Linuxkernel to help you do this. And then there are the unfriendlyprimitives, but I'll pretend they don't exist.
Locking in the Linux Kernel===========================
If I could give you one piece of advice on locking: **keep it simple**.
Be reluctant to introduce new locks.
Two Main Types of Kernel Locks: Spinlocks and Mutexes-----------------------------------------------------
There are two main types of kernel locks. The fundamental type is thespinlock (``include/asm/spinlock.h``), which is a very simplesingle-holder lock: if you can't get the spinlock, you keep trying(spinning) until you can. Spinlocks are very small and fast, and can beused anywhere.
The second type is a mutex (``include/linux/mutex.h``): it is like aspinlock, but you may block holding a mutex. If you can't lock a mutex,your task will suspend itself, and be woken up when the mutex isreleased. This means the CPU can do something else while you arewaiting. There are many cases when you simply can't sleep (see`What Functions Are Safe To Call From Interrupts?`_),and so have to use a spinlock instead.
Neither type of lock is recursive: see`Deadlock: Simple and Advanced`_.
Locks and Uniprocessor Kernels------------------------------
For kernels compiled without ``CONFIG_SMP``, and without``CONFIG_PREEMPT`` spinlocks do not exist at all. This is an excellentdesign decision: when no-one else can run at the same time, there is noreason to have a lock.
If the kernel is compiled without ``CONFIG_SMP``, but ``CONFIG_PREEMPT``is set, then spinlocks simply disable preemption, which is sufficient toprevent any races. For most purposes, we can think of preemption asequivalent to SMP, and not worry about it separately.
You should always test your locking code with ``CONFIG_SMP`` and``CONFIG_PREEMPT`` enabled, even if you don't have an SMP test box,because it will still catch some kinds of locking bugs.
Mutexes still exist, because they are required for synchronizationbetween user contexts, as we will see below.
Locking Only In User Context----------------------------
If you have a data structure which is only ever accessed from usercontext, then you can use a simple mutex (``include/linux/mutex.h``) toprotect it. This is the most trivial case: you initialize the mutex.Then you can call mutex_lock_interruptible() to grab themutex, and mutex_unlock() to release it. There is also amutex_lock(), which should be avoided, because it willnot return if a signal is received.
Example: ``net/netfilter/nf_sockopt.c`` allows registration of newsetsockopt() and getsockopt() calls, withnf_register_sockopt(). Registration and de-registrationare only done on module load and unload (and boot time, where there isno concurrency), and the list of registrations is only consulted for anunknown setsockopt() or getsockopt() systemcall. The ``nf_sockopt_mutex`` is perfect to protect this, especiallysince the setsockopt and getsockopt calls may well sleep.
Locking Between User Context and Softirqs-----------------------------------------
If a softirq shares data with user context, you have two problems.Firstly, the current user context can be interrupted by a softirq, andsecondly, the critical region could be entered from another CPU. This iswhere spin_lock_bh() (``include/linux/spinlock.h``) isused. It disables softirqs on that CPU, then grabs the lock.spin_unlock_bh() does the reverse. (The '_bh' suffix isa historical reference to "Bottom Halves", the old name for softwareinterrupts. It should really be called spin_lock_softirq()' in aperfect world).
Note that you can also use spin_lock_irq() orspin_lock_irqsave() here, which stop hardware interruptsas well: see `Hard IRQ Context`_.
This works perfectly for UP as well: the spin lock vanishes, and thismacro simply becomes local_bh_disable()(``include/linux/interrupt.h``), which protects you from the softirqbeing run.
Locking Between User Context and Tasklets-----------------------------------------
This is exactly the same as above, because tasklets are actually runfrom a softirq.
Locking Between User Context and Timers---------------------------------------
This, too, is exactly the same as above, because timers are actually runfrom a softirq. From a locking point of view, tasklets and timers areidentical.
Locking Between Tasklets/Timers-------------------------------
Sometimes a tasklet or timer might want to share data with anothertasklet or timer.
The Same Tasklet/Timer~~~~~~~~~~~~~~~~~~~~~~
Since a tasklet is never run on two CPUs at once, you don't need toworry about your tasklet being reentrant (running twice at once), evenon SMP.
Different Tasklets/Timers~~~~~~~~~~~~~~~~~~~~~~~~~
If another tasklet/timer wants to share data with your tasklet or timer, you will both need to use spin_lock() andspin_unlock() calls. spin_lock_bh() isunnecessary here, as you are already in a tasklet, and none will be runon the same CPU.
Locking Between Softirqs------------------------
Often a softirq might want to share data with itself or a tasklet/timer.
The Same Softirq~~~~~~~~~~~~~~~~
The same softirq can run on the other CPUs: you can use a per-CPU array(see `Per-CPU Data`_) for better performance. If you'regoing so far as to use a softirq, you probably care about scalableperformance enough to justify the extra complexity.
You'll need to use spin_lock() andspin_unlock() for shared data.
Different Softirqs~~~~~~~~~~~~~~~~~~
You'll need to use spin_lock() andspin_unlock() for shared data, whether it be a timer,tasklet, different softirq or the same or another softirq: any of themcould be running on a different CPU.
Hard IRQ Context================
Hardware interrupts usually communicate with a tasklet or softirq.Frequently this involves putting work in a queue, which the softirq willtake out.
Locking Between Hard IRQ and Softirqs/Tasklets----------------------------------------------
If a hardware irq handler shares data with a softirq, you have twoconcerns. Firstly, the softirq processing can be interrupted by ahardware interrupt, and secondly, the critical region could be enteredby a hardware interrupt on another CPU. This is wherespin_lock_irq() is used. It is defined to disableinterrupts on that cpu, then grab the lock.spin_unlock_irq() does the reverse.
The irq handler does not need to use spin_lock_irq(), becausethe softirq cannot run while the irq handler is running: it can usespin_lock(), which is slightly faster. The only exceptionwould be if a different hardware irq handler uses the same lock:spin_lock_irq() will stop that from interrupting us.
This works perfectly for UP as well: the spin lock vanishes, and thismacro simply becomes local_irq_disable()(``include/asm/smp.h``), which protects you from the softirq/tasklet/BHbeing run.
spin_lock_irqsave() (``include/linux/spinlock.h``) is avariant which saves whether interrupts were on or off in a flags word,which is passed to spin_unlock_irqrestore(). This meansthat the same code can be used inside an hard irq handler (whereinterrupts are already off) and in softirqs (where the irq disabling isrequired).
Note that softirqs (and hence tasklets and timers) are run on returnfrom hardware interrupts, so spin_lock_irq() also stopsthese. In that sense, spin_lock_irqsave() is the mostgeneral and powerful locking function.
Locking Between Two Hard IRQ Handlers-------------------------------------
It is rare to have to share data between two IRQ handlers, but if youdo, spin_lock_irqsave() should be used: it isarchitecture-specific whether all interrupts are disabled inside irqhandlers themselves.
Cheat Sheet For Locking=======================
Pete Zaitcev gives the following summary:
- If you are in a process context (any syscall) and want to lock other process out, use a mutex. You can take a mutex and sleep (``copy_from_user()`` or ``kmalloc(x,GFP_KERNEL)``).
- Otherwise (== data can be touched in an interrupt), use spin_lock_irqsave() and spin_unlock_irqrestore().
- Avoid holding spinlock for more than 5 lines of code and across any function call (except accessors like readb()).
Table of Minimum Requirements-----------------------------
The following table lists the **minimum** locking requirements betweenvarious contexts. In some cases, the same context can only be running onone CPU at a time, so no locking is required for that context (eg. aparticular thread can only run on one CPU at a time, but if it needsshares data with another thread, locking is required).
Remember the advice above: you can always usespin_lock_irqsave(), which is a superset of all otherspinlock primitives.
============== ============= ============= ========= ========= ========= ========= ======= ======= ============== ==============. IRQ Handler A IRQ Handler B Softirq A Softirq B Tasklet A Tasklet B Timer A Timer B User Context A User Context B============== ============= ============= ========= ========= ========= ========= ======= ======= ============== ==============IRQ Handler A NoneIRQ Handler B SLIS NoneSoftirq A SLI SLI SLSoftirq B SLI SLI SL SLTasklet A SLI SLI SL SL NoneTasklet B SLI SLI SL SL SL NoneTimer A SLI SLI SL SL SL SL NoneTimer B SLI SLI SL SL SL SL SL NoneUser Context A SLI SLI SLBH SLBH SLBH SLBH SLBH SLBH NoneUser Context B SLI SLI SLBH SLBH SLBH SLBH SLBH SLBH MLI None============== ============= ============= ========= ========= ========= ========= ======= ======= ============== ==============
Table: Table of Locking Requirements
+--------+----------------------------+| SLIS | spin_lock_irqsave |+--------+----------------------------+| SLI | spin_lock_irq |+--------+----------------------------+| SL | spin_lock |+--------+----------------------------+| SLBH | spin_lock_bh |+--------+----------------------------+| MLI | mutex_lock_interruptible |+--------+----------------------------+
Table: Legend for Locking Requirements Table
The trylock Functions=====================
There are functions that try to acquire a lock only once and immediatelyreturn a value telling about success or failure to acquire the lock.They can be used if you need no access to the data protected with thelock when some other thread is holding the lock. You should acquire thelock later if you then need access to the data protected with the lock.
spin_trylock() does not spin but returns non-zero if itacquires the spinlock on the first try or 0 if not. This function can beused in all contexts like spin_lock(): you must havedisabled the contexts that might interrupt you and acquire the spinlock.
mutex_trylock() does not suspend your task but returnsnon-zero if it could lock the mutex on the first try or 0 if not. Thisfunction cannot be safely used in hardware or software interruptcontexts despite not sleeping.
Common Examples===============
Let's step through a simple example: a cache of number to name mappings.The cache keeps a count of how often each of the objects is used, andwhen it gets full, throws out the least used one.
All In User Context-------------------
For our first example, we assume that all operations are in user context(ie. from system calls), so we can sleep. This means we can use a mutexto protect the cache and all the objects within it. Here's the code::
#include <linux/list.h> #include <linux/slab.h> #include <linux/string.h> #include <linux/mutex.h> #include <asm/errno.h>
struct object { struct list_head list; int id; char name[32]; int popularity; };
/* Protects the cache, cache_num, and the objects within it */ static DEFINE_MUTEX(cache_lock); static LIST_HEAD(cache); static unsigned int cache_num = 0; #define MAX_CACHE_SIZE 10
/* Must be holding cache_lock */ static struct object *__cache_find(int id) { struct object *i;
list_for_each_entry(i, &cache, list) if (i->id == id) { i->popularity++; return i; } return NULL; }
/* Must be holding cache_lock */ static void __cache_delete(struct object *obj) { BUG_ON(!obj); list_del(&obj->list); kfree(obj); cache_num--; }
/* Must be holding cache_lock */ static void __cache_add(struct object *obj) { list_add(&obj->list, &cache); if (++cache_num > MAX_CACHE_SIZE) { struct object *i, *outcast = NULL; list_for_each_entry(i, &cache, list) { if (!outcast || i->popularity < outcast->popularity) outcast = i; } __cache_delete(outcast); } }
int cache_add(int id, const char *name) { struct object *obj;
if ((obj = kmalloc_obj(*obj)) == NULL) return -ENOMEM;
strscpy(obj->name, name, sizeof(obj->name)); obj->id = id; obj->popularity = 0;
mutex_lock(&cache_lock); __cache_add(obj); mutex_unlock(&cache_lock); return 0; }
void cache_delete(int id) { mutex_lock(&cache_lock); __cache_delete(__cache_find(id)); mutex_unlock(&cache_lock); }
int cache_find(int id, char *name) { struct object *obj; int ret = -ENOENT;
mutex_lock(&cache_lock); obj = __cache_find(id); if (obj) { ret = 0; strcpy(name, obj->name); } mutex_unlock(&cache_lock); return ret; }
Note that we always make sure we have the cache_lock when we add,delete, or look up the cache: both the cache infrastructure itself andthe contents of the objects are protected by the lock. In this case it'seasy, since we copy the data for the user, and never let them access theobjects directly.
There is a slight (and common) optimization here: incache_add() we set up the fields of the object beforegrabbing the lock. This is safe, as no-one else can access it until weput it in cache.
Accessing From Interrupt Context--------------------------------
Now consider the case where cache_find() can be calledfrom interrupt context: either a hardware interrupt or a softirq. Anexample would be a timer which deletes object from the cache.
The change is shown below, in standard patch format: the ``-`` are lineswhich are taken away, and the ``+`` are lines which are added.
::
--- cache.c.usercontext 2003-12-09 13:58:54.000000000 +1100 +++ cache.c.interrupt 2003-12-09 14:07:49.000000000 +1100 @@ -12,7 +12,7 @@ int popularity; };
-static DEFINE_MUTEX(cache_lock); +static DEFINE_SPINLOCK(cache_lock); static LIST_HEAD(cache); static unsigned int cache_num = 0; #define MAX_CACHE_SIZE 10 @@ -55,6 +55,7 @@ int cache_add(int id, const char *name) { struct object *obj; + unsigned long flags;
if ((obj = kmalloc_obj(*obj)) == NULL) return -ENOMEM; @@ -63,30 +64,33 @@ obj->id = id; obj->popularity = 0;
- mutex_lock(&cache_lock); + spin_lock_irqsave(&cache_lock, flags); __cache_add(obj); - mutex_unlock(&cache_lock); + spin_unlock_irqrestore(&cache_lock, flags); return 0; }
void cache_delete(int id) { - mutex_lock(&cache_lock); + unsigned long flags; + + spin_lock_irqsave(&cache_lock, flags); __cache_delete(__cache_find(id)); - mutex_unlock(&cache_lock); + spin_unlock_irqrestore(&cache_lock, flags); }
int cache_find(int id, char *name) { struct object *obj; int ret = -ENOENT; + unsigned long flags;
- mutex_lock(&cache_lock); + spin_lock_irqsave(&cache_lock, flags); obj = __cache_find(id); if (obj) { ret = 0; strcpy(name, obj->name); } - mutex_unlock(&cache_lock); + spin_unlock_irqrestore(&cache_lock, flags); return ret; }
Note that the spin_lock_irqsave() will turn offinterrupts if they are on, otherwise does nothing (if we are already inan interrupt handler), hence these functions are safe to call from anycontext.
Unfortunately, cache_add() calls kmalloc()with the ``GFP_KERNEL`` flag, which is only legal in user context. Ihave assumed that cache_add() is still only called inuser context, otherwise this should become a parameter tocache_add().
Exposing Objects Outside This File----------------------------------
If our objects contained more information, it might not be sufficient tocopy the information in and out: other parts of the code might want tokeep pointers to these objects, for example, rather than looking up theid every time. This produces two problems.
The first problem is that we use the ``cache_lock`` to protect objects:we'd need to make this non-static so the rest of the code can use it.This makes locking trickier, as it is no longer all in one place.
The second problem is the lifetime problem: if another structure keeps apointer to an object, it presumably expects that pointer to remainvalid. Unfortunately, this is only guaranteed while you hold the lock,otherwise someone might call cache_delete() and evenworse, add another object, re-using the same address.
As there is only one lock, you can't hold it forever: no-one else wouldget any work done.
The solution to this problem is to use a reference count: everyone whohas a pointer to the object increases it when they first get the object,and drops the reference count when they're finished with it. Whoeverdrops it to zero knows it is unused, and can actually delete it.
Here is the code::
--- cache.c.interrupt 2003-12-09 14:25:43.000000000 +1100 +++ cache.c.refcnt 2003-12-09 14:33:05.000000000 +1100 @@ -7,6 +7,7 @@ struct object { struct list_head list; + unsigned int refcnt; int id; char name[32]; int popularity; @@ -17,6 +18,35 @@ static unsigned int cache_num = 0; #define MAX_CACHE_SIZE 10
+static void __object_put(struct object *obj) +{ + if (--obj->refcnt == 0) + kfree(obj); +} + +static void __object_get(struct object *obj) +{ + obj->refcnt++; +} + +void object_put(struct object *obj) +{ + unsigned long flags; + + spin_lock_irqsave(&cache_lock, flags); + __object_put(obj); + spin_unlock_irqrestore(&cache_lock, flags); +} + +void object_get(struct object *obj) +{ + unsigned long flags; + + spin_lock_irqsave(&cache_lock, flags); + __object_get(obj); + spin_unlock_irqrestore(&cache_lock, flags); +} + /* Must be holding cache_lock */ static struct object *__cache_find(int id) { @@ -35,6 +65,7 @@ { BUG_ON(!obj); list_del(&obj->list); + __object_put(obj); cache_num--; }
@@ -63,6 +94,7 @@ strscpy(obj->name, name, sizeof(obj->name)); obj->id = id; obj->popularity = 0; + obj->refcnt = 1; /* The cache holds a reference */
spin_lock_irqsave(&cache_lock, flags); __cache_add(obj); @@ -79,18 +111,15 @@ spin_unlock_irqrestore(&cache_lock, flags); }
-int cache_find(int id, char *name) +struct object *cache_find(int id) { struct object *obj; - int ret = -ENOENT; unsigned long flags;
spin_lock_irqsave(&cache_lock, flags); obj = __cache_find(id); - if (obj) { - ret = 0; - strcpy(name, obj->name); - } + if (obj) + __object_get(obj); spin_unlock_irqrestore(&cache_lock, flags); - return ret; + return obj; }
We encapsulate the reference counting in the standard 'get' and 'put'functions. Now we can return the object itself fromcache_find() which has the advantage that the user cannow sleep holding the object (eg. to copy_to_user() toname to userspace).
The other point to note is that I said a reference should be held forevery pointer to the object: thus the reference count is 1 when firstinserted into the cache. In some versions the framework does not hold areference count, but they are more complicated.
Using Atomic Operations For The Reference Count~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In practice, :c:type:`atomic_t` would usually be used for refcnt. There are anumber of atomic operations defined in ``include/asm/atomic.h``: theseare guaranteed to be seen atomically from all CPUs in the system, so nolock is required. In this case, it is simpler than using spinlocks,although for anything non-trivial using spinlocks is clearer. Theatomic_inc() and atomic_dec_and_test()are used instead of the standard increment and decrement operators, andthe lock is no longer used to protect the reference count itself.
::
--- cache.c.refcnt 2003-12-09 15:00:35.000000000 +1100 +++ cache.c.refcnt-atomic 2003-12-11 15:49:42.000000000 +1100 @@ -7,7 +7,7 @@ struct object { struct list_head list; - unsigned int refcnt; + atomic_t refcnt; int id; char name[32]; int popularity; @@ -18,33 +18,15 @@ static unsigned int cache_num = 0; #define MAX_CACHE_SIZE 10
-static void __object_put(struct object *obj) -{ - if (--obj->refcnt == 0) - kfree(obj); -} - -static void __object_get(struct object *obj) -{ - obj->refcnt++; -} - void object_put(struct object *obj) { - unsigned long flags; - - spin_lock_irqsave(&cache_lock, flags); - __object_put(obj); - spin_unlock_irqrestore(&cache_lock, flags); + if (atomic_dec_and_test(&obj->refcnt)) + kfree(obj); }
void object_get(struct object *obj) { - unsigned long flags; - - spin_lock_irqsave(&cache_lock, flags); - __object_get(obj); - spin_unlock_irqrestore(&cache_lock, flags); + atomic_inc(&obj->refcnt); }
/* Must be holding cache_lock */ @@ -65,7 +47,7 @@ { BUG_ON(!obj); list_del(&obj->list); - __object_put(obj); + object_put(obj); cache_num--; }
@@ -94,7 +76,7 @@ strscpy(obj->name, name, sizeof(obj->name)); obj->id = id; obj->popularity = 0; - obj->refcnt = 1; /* The cache holds a reference */ + atomic_set(&obj->refcnt, 1); /* The cache holds a reference */
spin_lock_irqsave(&cache_lock, flags); __cache_add(obj); @@ -119,7 +101,7 @@ spin_lock_irqsave(&cache_lock, flags); obj = __cache_find(id); if (obj) - __object_get(obj); + object_get(obj); spin_unlock_irqrestore(&cache_lock, flags); return obj; }
Protecting The Objects Themselves---------------------------------
In these examples, we assumed that the objects (except the referencecounts) never changed once they are created. If we wanted to allow thename to change, there are three possibilities:
- You can make ``cache_lock`` non-static, and tell people to grab that lock before changing the name in any object.
- You can provide a cache_obj_rename() which grabs this lock and changes the name for the caller, and tell everyone to use that function.
- You can make the ``cache_lock`` protect only the cache itself, and use another lock to protect the name.
Theoretically, you can make the locks as fine-grained as one lock forevery field, for every object. In practice, the most common variantsare:
- One lock which protects the infrastructure (the ``cache`` list in this example) and all the objects. This is what we have done so far.
- One lock which protects the infrastructure (including the list pointers inside the objects), and one lock inside the object which protects the rest of that object.
- Multiple locks to protect the infrastructure (eg. one lock per hash chain), possibly with a separate per-object lock.
Here is the "lock-per-object" implementation:
::
--- cache.c.refcnt-atomic 2003-12-11 15:50:54.000000000 +1100 +++ cache.c.perobjectlock 2003-12-11 17:15:03.000000000 +1100 @@ -6,11 +6,17 @@
struct object { + /* These two protected by cache_lock. */ struct list_head list; + int popularity; + atomic_t refcnt; + + /* Doesn't change once created. */ int id; + + spinlock_t lock; /* Protects the name */ char name[32]; - int popularity; };
static DEFINE_SPINLOCK(cache_lock); @@ -77,6 +84,7 @@ obj->id = id; obj->popularity = 0; atomic_set(&obj->refcnt, 1); /* The cache holds a reference */ + spin_lock_init(&obj->lock);
spin_lock_irqsave(&cache_lock, flags); __cache_add(obj);
Note that I decide that the popularity count should be protected by the``cache_lock`` rather than the per-object lock: this is because it (likethe :c:type:`struct list_head <list_head>` inside the object)is logically part of the infrastructure. This way, I don't need to grabthe lock of every object in __cache_add() when seekingthe least popular.
I also decided that the id member is unchangeable, so I don't need tograb each object lock in __cache_find() to examine theid: the object lock is only used by a caller who wants to read or writethe name field.
Note also that I added a comment describing what data was protected bywhich locks. This is extremely important, as it describes the runtimebehavior of the code, and can be hard to gain from just reading. And asAlan Cox says, “Lock data, not code”.
Common Problems===============
Deadlock: Simple and Advanced-----------------------------
There is a coding bug where a piece of code tries to grab a spinlocktwice: it will spin forever, waiting for the lock to be released(spinlocks, rwlocks and mutexes are not recursive in Linux). This istrivial to diagnose: not astay-up-five-nights-talk-to-fluffy-code-bunnies kind of problem.
For a slightly more complex case, imagine you have a region shared by asoftirq and user context. If you use a spin_lock() callto protect it, it is possible that the user context will be interruptedby the softirq while it holds the lock, and the softirq will then spinforever trying to get the same lock.
Both of these are called deadlock, and as shown above, it can occur evenwith a single CPU (although not on UP compiles, since spinlocks vanishon kernel compiles with ``CONFIG_SMP``\ =n. You'll still get datacorruption in the second example).
This complete lockup is easy to diagnose: on SMP boxes the watchdogtimer or compiling with ``DEBUG_SPINLOCK`` set(``include/linux/spinlock.h``) will show this up immediately when ithappens.
A more complex problem is the so-called 'deadly embrace', involving twoor more locks. Say you have a hash table: each entry in the table is aspinlock, and a chain of hashed objects. Inside a softirq handler, yousometimes want to alter an object from one place in the hash to another:you grab the spinlock of the old hash chain and the spinlock of the newhash chain, and delete the object from the old one, and insert it in thenew one.
There are two problems here. First, if your code ever tries to move theobject to the same chain, it will deadlock with itself as it tries tolock it twice. Secondly, if the same softirq on another CPU is trying tomove another object in the reverse direction, the following couldhappen:
+-----------------------+-----------------------+| CPU 1 | CPU 2 |+=======================+=======================+| Grab lock A -> OK | Grab lock B -> OK |+-----------------------+-----------------------+| Grab lock B -> spin | Grab lock A -> spin |+-----------------------+-----------------------+
Table: Consequences
The two CPUs will spin forever, waiting for the other to give up theirlock. It will look, smell, and feel like a crash.
Preventing Deadlock-------------------
Textbooks will tell you that if you always lock in the same order, youwill never get this kind of deadlock. Practice will tell you that thisapproach doesn't scale: when I create a new lock, I don't understandenough of the kernel to figure out where in the 5000 lock hierarchy itwill fit.
The best locks are encapsulated: they never get exposed in headers, andare never held around calls to non-trivial functions outside the samefile. You can read through this code and see that it will neverdeadlock, because it never tries to grab another lock while it has thatone. People using your code don't even need to know you are using alock.
A classic problem here is when you provide callbacks or hooks: if youcall these with the lock held, you risk simple deadlock, or a deadlyembrace (who knows what the callback will do?).
Overzealous Prevention Of Deadlocks~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Deadlocks are problematic, but not as bad as data corruption. Code whichgrabs a read lock, searches a list, fails to find what it wants, dropsthe read lock, grabs a write lock and inserts the object has a racecondition.
Racing Timers: A Kernel Pastime-------------------------------
Timers can produce their own special problems with races. Consider acollection of objects (list, hash, etc) where each object has a timerwhich is due to destroy it.
If you want to destroy the entire collection (say on module removal),you might do the following::
/* THIS CODE BAD BAD BAD BAD: IF IT WAS ANY WORSE IT WOULD USE HUNGARIAN NOTATION */ spin_lock_bh(&list_lock);
while (list) { struct foo *next = list->next; timer_delete(&list->timer); kfree(list); list = next; }
spin_unlock_bh(&list_lock);
Sooner or later, this will crash on SMP, because a timer can have justgone off before the spin_lock_bh(), and it will only getthe lock after we spin_unlock_bh(), and then try to freethe element (which has already been freed!).
This can be avoided by checking the result oftimer_delete(): if it returns 1, the timer has been deleted.If 0, it means (in this case) that it is currently running, so we cando::
retry: spin_lock_bh(&list_lock);
while (list) { struct foo *next = list->next; if (!timer_delete(&list->timer)) { /* Give timer a chance to delete this */ spin_unlock_bh(&list_lock); goto retry; } kfree(list); list = next; }
spin_unlock_bh(&list_lock);
Another common problem is deleting timers which restart themselves (bycalling add_timer() at the end of their timer function).Because this is a fairly common case which is prone to races, you shoulduse timer_delete_sync() (``include/linux/timer.h``) to handle this case.
Before freeing a timer, timer_shutdown() or timer_shutdown_sync() should becalled which will keep it from being rearmed. Any subsequent attempt torearm the timer will be silently ignored by the core code.
Locking Speed=============
There are three main things to worry about when considering speed ofsome code which does locking. First is concurrency: how many things aregoing to be waiting while someone else is holding a lock. Second is thetime taken to actually acquire and release an uncontended lock. Third isusing fewer, or smarter locks. I'm assuming that the lock is used fairlyoften: otherwise, you wouldn't be concerned about efficiency.
Concurrency depends on how long the lock is usually held: you shouldhold the lock for as long as needed, but no longer. In the cacheexample, we always create the object without the lock held, and thengrab the lock only when we are ready to insert it in the list.
Acquisition times depend on how much damage the lock operations do tothe pipeline (pipeline stalls) and how likely it is that this CPU wasthe last one to grab the lock (ie. is the lock cache-hot for this CPU):on a machine with more CPUs, this likelihood drops fast. Consider a700MHz Intel Pentium III: an instruction takes about 0.7ns, an atomicincrement takes about 58ns, a lock which is cache-hot on this CPU takes160ns, and a cacheline transfer from another CPU takes an additional 170to 360ns. (These figures from Paul McKenney's `Linux Journal RCUarticle <http://www.linuxjournal.com/article.php?sid=6993>`__).
These two aims conflict: holding a lock for a short time might be doneby splitting locks into parts (such as in our final per-object-lockexample), but this increases the number of lock acquisitions, and theresults are often slower than having a single lock. This is anotherreason to advocate locking simplicity.
The third concern is addressed below: there are some methods to reducethe amount of locking which needs to be done.
Read/Write Lock Variants------------------------
Both spinlocks and mutexes have read/write variants: ``rwlock_t`` and:c:type:`struct rw_semaphore <rw_semaphore>`. These divideusers into two classes: the readers and the writers. If you are onlyreading the data, you can get a read lock, but to write to the data youneed the write lock. Many people can hold a read lock, but a writer mustbe sole holder.
If your code divides neatly along reader/writer lines (as our cache codedoes), and the lock is held by readers for significant lengths of time,using these locks can help. They are slightly slower than the normallocks though, so in practice ``rwlock_t`` is not usually worthwhile.
Avoiding Locks: Read Copy Update--------------------------------
There is a special method of read/write locking called Read Copy Update.Using RCU, the readers can avoid taking a lock altogether: as we expectour cache to be read more often than updated (otherwise the cache is awaste of time), it is a candidate for this optimization.
How do we get rid of read locks? Getting rid of read locks means thatwriters may be changing the list underneath the readers. That isactually quite simple: we can read a linked list while an element isbeing added if the writer adds the element very carefully. For example,adding ``new`` to a single linked list called ``list``::
new->next = list->next; wmb(); list->next = new;
The wmb() is a write memory barrier. It ensures that thefirst operation (setting the new element's ``next`` pointer) is completeand will be seen by all CPUs, before the second operation is (puttingthe new element into the list). This is important, since moderncompilers and modern CPUs can both reorder instructions unless toldotherwise: we want a reader to either not see the new element at all, orsee the new element with the ``next`` pointer correctly pointing at therest of the list.
Fortunately, there is a function to do this for standard:c:type:`struct list_head <list_head>` lists:list_add_rcu() (``include/linux/list.h``).
Removing an element from the list is even simpler: we replace thepointer to the old element with a pointer to its successor, and readerswill either see it, or skip over it.
::
list->next = old->next;
There is list_del_rcu() (``include/linux/list.h``) whichdoes this (the normal version poisons the old object, which we don'twant).
The reader must also be careful: some CPUs can look through the ``next``pointer to start reading the contents of the next element early, butdon't realize that the pre-fetched contents is wrong when the ``next``pointer changes underneath them. Once again, there is alist_for_each_entry_rcu() (``include/linux/list.h``)to help you. Of course, writers can just uselist_for_each_entry(), since there cannot be twosimultaneous writers.
Our final dilemma is this: when can we actually destroy the removedelement? Remember, a reader might be stepping through this element inthe list right now: if we free this element and the ``next`` pointerchanges, the reader will jump off into garbage and crash. We need towait until we know that all the readers who were traversing the listwhen we deleted the element are finished. We usecall_rcu() to register a callback which will actuallydestroy the object once all pre-existing readers are finished.Alternatively, synchronize_rcu() may be used to blockuntil all pre-existing are finished.
But how does Read Copy Update know when the readers are finished? Themethod is this: firstly, the readers always traverse the list insidercu_read_lock()/rcu_read_unlock() pairs:these simply disable preemption so the reader won't go to sleep whilereading the list.
RCU then waits until every other CPU has slept at least once: sincereaders cannot sleep, we know that any readers which were traversing thelist during the deletion are finished, and the callback is triggered.The real Read Copy Update code is a little more optimized than this, butthis is the fundamental idea.
::
--- cache.c.perobjectlock 2003-12-11 17:15:03.000000000 +1100 +++ cache.c.rcupdate 2003-12-11 17:55:14.000000000 +1100 @@ -1,15 +1,18 @@ #include <linux/list.h> #include <linux/slab.h> #include <linux/string.h> +#include <linux/rcupdate.h> #include <linux/mutex.h> #include <asm/errno.h>
struct object { - /* These two protected by cache_lock. */ + /* This is protected by RCU */ struct list_head list; int popularity;
+ struct rcu_head rcu; + atomic_t refcnt;
/* Doesn't change once created. */ @@ -40,7 +43,7 @@ { struct object *i;
- list_for_each_entry(i, &cache, list) { + list_for_each_entry_rcu(i, &cache, list) { if (i->id == id) { i->popularity++; return i; @@ -49,19 +52,25 @@ return NULL; }
+/* Final discard done once we know no readers are looking. */ +static void cache_delete_rcu(void *arg) +{ + object_put(arg); +} + /* Must be holding cache_lock */ static void __cache_delete(struct object *obj) { BUG_ON(!obj); - list_del(&obj->list); - object_put(obj); + list_del_rcu(&obj->list); cache_num--; + call_rcu(&obj->rcu, cache_delete_rcu); }
/* Must be holding cache_lock */ static void __cache_add(struct object *obj) { - list_add(&obj->list, &cache); + list_add_rcu(&obj->list, &cache); if (++cache_num > MAX_CACHE_SIZE) { struct object *i, *outcast = NULL; list_for_each_entry(i, &cache, list) { @@ -104,12 +114,11 @@ struct object *cache_find(int id) { struct object *obj; - unsigned long flags;
- spin_lock_irqsave(&cache_lock, flags); + rcu_read_lock(); obj = __cache_find(id); if (obj) object_get(obj); - spin_unlock_irqrestore(&cache_lock, flags); + rcu_read_unlock(); return obj; }
Note that the reader will alter the popularity member in__cache_find(), and now it doesn't hold a lock. Onesolution would be to make it an ``atomic_t``, but for this usage, wedon't really care about races: an approximate result is good enough, soI didn't change it.
The result is that cache_find() requires nosynchronization with any other functions, so is almost as fast on SMP asit would be on UP.
There is a further optimization possible here: remember our originalcache code, where there were no reference counts and the caller simplyheld the lock whenever using the object? This is still possible: if youhold the lock, no one can delete the object, so you don't need to getand put the reference count.
Now, because the 'read lock' in RCU is simply disabling preemption, acaller which always has preemption disabled between callingcache_find() and object_put() does notneed to actually get and put the reference count: we could expose__cache_find() by making it non-static, and suchcallers could simply call that.
The benefit here is that the reference count is not written to: theobject is not altered in any way, which is much faster on SMP machinesdue to caching.
Per-CPU Data------------
Another technique for avoiding locking which is used fairly widely is toduplicate information for each CPU. For example, if you wanted to keep acount of a common condition, you could use a spin lock and a singlecounter. Nice and simple.
If that was too slow (it's usually not, but if you've got a really bigmachine to test on and can show that it is), you could instead use acounter for each CPU, then none of them need an exclusive lock. SeeDEFINE_PER_CPU(), get_cpu_var() andput_cpu_var() (``include/linux/percpu.h``).
Of particular use for simple per-cpu counters is the ``local_t`` type,and the cpu_local_inc() and related functions, which aremore efficient than simple code on some architectures(``include/asm/local.h``).
Note that there is no simple, reliable way of getting an exact value ofsuch a counter, without introducing more locks. This is not a problemfor some uses.
Data Which Mostly Used By An IRQ Handler----------------------------------------
If data is always accessed from within the same IRQ handler, you don'tneed a lock at all: the kernel already guarantees that the irq handlerwill not run simultaneously on multiple CPUs.
Manfred Spraul points out that you can still do this, even if the datais very occasionally accessed in user context or softirqs/tasklets. Theirq handler doesn't use a lock, and all other accesses are done as so::
mutex_lock(&lock); disable_irq(irq); ... enable_irq(irq); mutex_unlock(&lock);
The disable_irq() prevents the irq handler from running(and waits for it to finish if it's currently running on other CPUs).The spinlock prevents any other accesses happening at the same time.Naturally, this is slower than just a spin_lock_irq()call, so it only makes sense if this type of access happens extremelyrarely.
What Functions Are Safe To Call From Interrupts?================================================
Many functions in the kernel sleep (ie. call schedule()) directly orindirectly: you can never call them while holding a spinlock, or withpreemption disabled. This also means you need to be in user context:calling them from an interrupt is illegal.
Some Functions Which Sleep--------------------------
The most common ones are listed below, but you usually have to read thecode to find out if other calls are safe. If everyone else who calls itcan sleep, you probably need to be able to sleep, too. In particular,registration and deregistration functions usually expect to be calledfrom user context, and can sleep.
- Accesses to userspace:
- copy_from_user()
- copy_to_user()
- get_user()
- put_user()
- kmalloc(GP_KERNEL) <kmalloc>`
- mutex_lock_interruptible() and mutex_lock()
There is a mutex_trylock() which does not sleep. Still, it must not be used inside interrupt context since its implementation is not safe for that. mutex_unlock() will also never sleep. It cannot be used in interrupt context either since a mutex must be released by the same task that acquired it.
Some Functions Which Don't Sleep--------------------------------
Some functions are safe to call from any context, or holding almost anylock.
- printk()
- kfree()
- add_timer() and timer_delete()
Mutex API reference===================
.. kernel-doc:: include/linux/mutex.h :internal:
.. kernel-doc:: kernel/locking/mutex.c :export:
Futex API reference===================
.. kernel-doc:: kernel/futex/core.c :internal:
.. kernel-doc:: kernel/futex/futex.h :internal:
.. kernel-doc:: kernel/futex/pi.c :internal:
.. kernel-doc:: kernel/futex/requeue.c :internal:
.. kernel-doc:: kernel/futex/waitwake.c :internal:
Further reading===============
- ``Documentation/locking/spinlocks.rst``: Linus Torvalds' spinlocking tutorial in the kernel sources.
- Unix Systems for Modern Architectures: Symmetric Multiprocessing and Caching for Kernel Programmers:
Curt Schimmel's very good introduction to kernel level locking (not written for Linux, but nearly everything applies). The book is expensive, but really worth every penny to understand SMP locking. [ISBN: 0201633388]
Thanks======
Thanks to Telsa Gwynne for DocBooking, neatening and adding style.
Thanks to Martin Pool, Philipp Rumpf, Stephen Rothwell, Paul Mackerras,Ruedi Aschwanden, Alan Cox, Manfred Spraul, Tim Waugh, Pete Zaitcev,James Morris, Robert Love, Paul McKenney, John Ashby for proofreading,correcting, flaming, commenting.
Thanks to the cabal for having no influence on this document.
Glossary========
preemption Prior to 2.5, or when ``CONFIG_PREEMPT`` is unset, processes in user context inside the kernel would not preempt each other (ie. you had that CPU until you gave it up, except for interrupts). With the addition of ``CONFIG_PREEMPT`` in 2.5.4, this changed: when in user context, higher priority tasks can "cut in": spinlocks were changed to disable preemption, even on UP.
bh Bottom Half: for historical reasons, functions with '_bh' in them often now refer to any software interrupt, e.g. spin_lock_bh() blocks any software interrupt on the current CPU. Bottom halves are deprecated, and will eventually be replaced by tasklets. Only one bottom half will be running at any time.
Hardware Interrupt / Hardware IRQ Hardware interrupt request. in_hardirq() returns true in a hardware interrupt handler.
Interrupt Context Not user context: processing a hardware irq or software irq. Indicated by the in_interrupt() macro returning true.
SMP Symmetric Multi-Processor: kernels compiled for multiple-CPU machines. (``CONFIG_SMP=y``).
Software Interrupt / softirq Software interrupt handler. in_hardirq() returns false; in_softirq() returns true. Tasklets and softirqs both fall into the category of 'software interrupts'.
Strictly speaking a softirq is one of up to 32 enumerated software interrupts which can run on multiple CPUs at once. Sometimes used to refer to tasklets as well (ie. all software interrupts).
tasklet A dynamically-registrable software interrupt, which is guaranteed to only run on one CPU at a time.
timer A dynamically-registrable software interrupt, which is run at (or close to) a given time. When running, it is just like a tasklet (in fact, they are called from the ``TIMER_SOFTIRQ``).
UP Uni-Processor: Non-SMP. (``CONFIG_SMP=n``).
User Context The kernel executing on behalf of a particular process (ie. a system call or trap) or kernel thread. You can tell which process with the ``current`` macro.) Not to be confused with userspace. Can be interrupted by software or hardware interrupts.
Userspace A process executing its own code outside the kernel.