Linux內核同步機制mutex

mutex鎖概述

在linux內核中，互斥量mutex是一種保證CPU串行運行的睡眠鎖機制。和spinlock類似，都是同一個時刻只有一個線程進入臨界資源，不同的是，當無法獲取鎖的時候，spinlock原地自旋，而mutex則是選擇掛起當前線程，進入阻塞狀態(tài)。所以，mutex無法在中斷上下文中使用。

mutex鎖使用注意事項

• mutex一次只能有一個進程或線程持有該鎖
• mutex只有它的擁有者可以釋放該鎖
• 不能多次釋放同一把鎖
• 不可以重復獲取同一把鎖，否則會造成死鎖
• 必須使用mutex提供的專用初始化函數(shù)初始化該鎖
• 不能重復初始化同一把鎖
• 不能使用memset或memcpy等內存處理函數(shù)初始化mutex鎖
• 線程退出時要釋放自己持有的所有mutex鎖
• 不能用于設備中斷或軟中斷上下文中

mutex鎖結構體定義

• owner：記錄mutex的持有者
• wait_lock：spinlock自旋鎖
• soq：MCS鎖隊列，用于支持mutex樂觀自旋機制
• wait_list：當無法獲取鎖的時候掛起在此
• magic：用于debug調試
• dep_map：用于debug調試

struct mutex {
 atomic_long_t  owner;
 spinlock_t  wait_lock;
#ifdef CONFIG_MUTEX_SPIN_ON_OWNER
 struct optimistic_spin_queue osq; /* Spinner MCS lock */
#endif
 struct list_head wait_list;
#ifdef CONFIG_DEBUG_MUTEXES
 void   *magic;
#endif
#ifdef CONFIG_DEBUG_LOCK_ALLOC
 struct lockdep_map dep_map;
#endif
};

mutex鎖主要接口函數(shù)

獲取鎖流程分析

mutex_lock()函數(shù)調用might_sleep()函數(shù)判斷鎖的狀態(tài)，調用__mutex_trylock_fast()函數(shù)嘗試快速獲取mutex鎖，如果失敗，則調用__mutex_lock_slowpath()函數(shù)獲取mutex鎖

void __sched mutex_lock(struct mutex *lock)
{
 might_sleep();

 if (!__mutex_trylock_fast(lock))
  __mutex_lock_slowpath(lock);
}

如果沒有定義CONFIG_DEBUG_ATOMIC_SLEEP宏，might_sleep 函數(shù)退化為 might_resched() 函數(shù)。

# define might_sleep() \\
 do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0)
# define sched_annotate_sleep() (current- >task_state_change = 0)
#else
  static inline void ___might_sleep(const char *file, int line,
       int preempt_offset) { }
  static inline void __might_sleep(const char *file, int line,
       int preempt_offset) { }
# define might_sleep() do { might_resched(); } while (0)
# define sched_annotate_sleep() do { } while (0)

在配置了搶占式內核或者非搶占式內核的情況下，might_resched() 函數(shù)最終都是空函數(shù)。如果配置了主動搶占式內核CONFIG_PREEMPT_VOLUNTARY，則might_resched()函數(shù)會調用 _cond_resched() 函數(shù)來主動觸發(fā)一次搶占。

#ifdef CONFIG_PREEMPT_VOLUNTARY
extern int _cond_resched(void);
# define might_resched() _cond_resched()
#else
# define might_resched() do { } while (0)
#endif

#ifndef CONFIG_PREEMPT
extern int _cond_resched(void);
#else
static inline int _cond_resched(void) { return 0; }
#endif

——cond_resched()函數(shù)調用should_resched()函數(shù)判斷搶占計數(shù)器是否為0，如果搶占計數(shù)器為0并且設置了重新調度標記則調用preempt_schedule_common()函數(shù)進行搶占式調度

#ifndef CONFIG_PREEMPT
int __sched _cond_resched(void)
{
 if (should_resched(0)) {
  preempt_schedule_common();
  return 1;
 }
 return 0;
}
EXPORT_SYMBOL(_cond_resched);
#endif

__mutex_trylock_fast()函數(shù)調用atomic_long_cmpxchg_acquire()函數(shù)判斷lock->owner的值是否等于0，如果等于0，則直接將當前線程的task struct的指針賦值給lock->owner，表示該mutex鎖已經(jīng)被當前線程持有。如果lock->owner的值不等于0，則表示該mutex鎖已經(jīng)被其他線程持有或者鎖正在傳遞給top waiter線程，當前線程需要阻塞等待。上面描述的操作（比較和賦值）都是原子操作，不會有任何指令插入。

static __always_inline bool __mutex_trylock_fast(struct mutex *lock)
{
 unsigned long curr = (unsigned long)current;

 if (!atomic_long_cmpxchg_acquire(&lock- >owner, 0UL, curr))
  return true;

 return false;
}

慢速獲取mutex鎖的路徑就是__mutex_lock_common()函數(shù)，所謂慢速其實就是阻塞當前線程，將current task掛入mutex的等待隊列的尾部。讓所有等待mutex的任務按照時間的先后順序排列起來，當mutex被釋放的時候，會首先喚醒隊首的任務，即最先等待的任務最先被喚醒。此外，在向空隊列插入第一個任務的時候，會給mutex flag設置上MUTEX_FLAG_WAITERS標記，表示已經(jīng)有任務在等待這個mutex鎖了。

static noinline void __sched
__mutex_lock_slowpath(struct mutex *lock)
{
 __mutex_lock(lock, TASK_UNINTERRUPTIBLE, 0, NULL, _RET_IP_);
}
static int __sched
__mutex_lock(struct mutex *lock, long state, unsigned int subclass,
      struct lockdep_map *nest_lock, unsigned long ip)
{
 return __mutex_lock_common(lock, state, subclass, nest_lock, ip, NULL, false);
}
static __always_inline int __sched
__mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
      struct lockdep_map *nest_lock, unsigned long ip,
      struct ww_acquire_ctx *ww_ctx, const bool use_ww_ctx)
{
 struct mutex_waiter waiter;
 bool first = false;
 struct ww_mutex *ww;
 int ret;

 might_sleep();

 ww = container_of(lock, struct ww_mutex, base);
 if (use_ww_ctx && ww_ctx) {
  if (unlikely(ww_ctx == READ_ONCE(ww- >ctx)))
   return -EALREADY;
 }

 preempt_disable();
 mutex_acquire_nest(&lock- >dep_map, subclass, 0, nest_lock, ip);

 if (__mutex_trylock(lock) ||
     mutex_optimistic_spin(lock, ww_ctx, use_ww_ctx, NULL)) {
  /* got the lock, yay! */
  lock_acquired(&lock- >dep_map, ip);
  if (use_ww_ctx && ww_ctx)
   ww_mutex_set_context_fastpath(ww, ww_ctx);
  preempt_enable();
  return 0;
 }

 spin_lock(&lock- >wait_lock);
 /*
  * After waiting to acquire the wait_lock, try again.
  */
 if (__mutex_trylock(lock)) {
  if (use_ww_ctx && ww_ctx)
   __ww_mutex_wakeup_for_backoff(lock, ww_ctx);

  goto skip_wait;
 }

 debug_mutex_lock_common(lock, &waiter);
 debug_mutex_add_waiter(lock, &waiter, current);

 lock_contended(&lock- >dep_map, ip);

 if (!use_ww_ctx) {
  /* add waiting tasks to the end of the waitqueue (FIFO): */
  list_add_tail(&waiter.list, &lock- >wait_list);

#ifdef CONFIG_DEBUG_MUTEXES
  waiter.ww_ctx = MUTEX_POISON_WW_CTX;
#endif
 } else {
  /* Add in stamp order, waking up waiters that must back off. */
  ret = __ww_mutex_add_waiter(&waiter, lock, ww_ctx);
  if (ret)
   goto err_early_backoff;

  waiter.ww_ctx = ww_ctx;
 }

 waiter.task = current;

 if (__mutex_waiter_is_first(lock, &waiter))
  __mutex_set_flag(lock, MUTEX_FLAG_WAITERS);

 set_current_state(state);
 for (;;) {
  /*
   * Once we hold wait_lock, we're serialized against
   * mutex_unlock() handing the lock off to us, do a trylock
   * before testing the error conditions to make sure we pick up
   * the handoff.
   */
  if (__mutex_trylock(lock))
   goto acquired;

  /*
   * Check for signals and wound conditions while holding
   * wait_lock. This ensures the lock cancellation is ordered
   * against mutex_unlock() and wake-ups do not go missing.
   */
  if (unlikely(signal_pending_state(state, current))) {
   ret = -EINTR;
   goto err;
  }

  if (use_ww_ctx && ww_ctx && ww_ctx- >acquired > 0) {
   ret = __ww_mutex_lock_check_stamp(lock, &waiter, ww_ctx);
   if (ret)
    goto err;
  }

  spin_unlock(&lock- >wait_lock);
  schedule_preempt_disabled();

  /*
   * ww_mutex needs to always recheck its position since its waiter
   * list is not FIFO ordered.
   */
  if ((use_ww_ctx && ww_ctx) || !first) {
   first = __mutex_waiter_is_first(lock, &waiter);
   if (first)
    __mutex_set_flag(lock, MUTEX_FLAG_HANDOFF);
  }

  set_current_state(state);
  /*
   * Here we order against unlock; we must either see it change
   * state back to RUNNING and fall through the next schedule(),
   * or we must see its unlock and acquire.
   */
  if (__mutex_trylock(lock) ||
      (first && mutex_optimistic_spin(lock, ww_ctx, use_ww_ctx, &waiter)))
   break;

  spin_lock(&lock- >wait_lock);
 }
 spin_lock(&lock- >wait_lock);
acquired:
 __set_current_state(TASK_RUNNING);

 mutex_remove_waiter(lock, &waiter, current);
 if (likely(list_empty(&lock- >wait_list)))
  __mutex_clear_flag(lock, MUTEX_FLAGS);

 debug_mutex_free_waiter(&waiter);

skip_wait:
 /* got the lock - cleanup and rejoice! */
 lock_acquired(&lock- >dep_map, ip);

 if (use_ww_ctx && ww_ctx)
  ww_mutex_set_context_slowpath(ww, ww_ctx);

 spin_unlock(&lock- >wait_lock);
 preempt_enable();
 return 0;

err:
 __set_current_state(TASK_RUNNING);
 mutex_remove_waiter(lock, &waiter, current);
err_early_backoff:
 spin_unlock(&lock- >wait_lock);
 debug_mutex_free_waiter(&waiter);
 mutex_release(&lock- >dep_map, 1, ip);
 preempt_enable();
 return ret;
}

釋放鎖流程分析

釋放鎖的流程也分快速釋放和慢速釋放兩種路徑

void __sched mutex_unlock(struct mutex *lock)
{
#ifndef CONFIG_DEBUG_LOCK_ALLOC
 if (__mutex_unlock_fast(lock))
  return;
#endif
 __mutex_unlock_slowpath(lock, _RET_IP_);
}

如果一個線程獲取了mutex鎖之后，沒有其他的線程試圖獲取，此時的mutex的owner成員就是該線程的task struct地址，并且所有的mutex flag都沒有被設置。這時候只需要將mutex的owner成員清零即可，不需要其它操作，這就是快速釋放鎖的路徑。

static __always_inline bool __mutex_unlock_fast(struct mutex *lock)
{
 unsigned long curr = (unsigned long)current;

 if (atomic_long_cmpxchg_release(&lock- >owner, curr, 0UL) == curr)
  return true;

 return false;
}

如果有其他線程在競爭該mutex鎖，這時候就會進入慢速釋放鎖路徑，慢速釋放鎖路徑的邏輯分成兩段：一段是釋放mutex鎖，另外一段是喚醒top waiter線程。

static noinline void __sched __mutex_unlock_slowpath(struct mutex *lock, unsigned long ip)
{
 struct task_struct *next = NULL;
 DEFINE_WAKE_Q(wake_q);
 unsigned long owner;

 mutex_release(&lock- >dep_map, 1, ip);

 /*
  * Release the lock before (potentially) taking the spinlock such that
  * other contenders can get on with things ASAP.
  *
  * Except when HANDOFF, in that case we must not clear the owner field,
  * but instead set it to the top waiter.
  */
 owner = atomic_long_read(&lock- >owner);
 for (;;) {
  unsigned long old;

#ifdef CONFIG_DEBUG_MUTEXES
  DEBUG_LOCKS_WARN_ON(__owner_task(owner) != current);
  DEBUG_LOCKS_WARN_ON(owner & MUTEX_FLAG_PICKUP);
#endif

  if (owner & MUTEX_FLAG_HANDOFF)
   break;

  old = atomic_long_cmpxchg_release(&lock- >owner, owner,
        __owner_flags(owner));
  if (old == owner) {
   if (owner & MUTEX_FLAG_WAITERS)
    break;

   return;
  }

  owner = old;
 }

 spin_lock(&lock- >wait_lock);
 debug_mutex_unlock(lock);
 if (!list_empty(&lock- >wait_list)) {
  /* get the first entry from the wait-list: */
  struct mutex_waiter *waiter =
   list_first_entry(&lock- >wait_list,
      struct mutex_waiter, list);

  next = waiter- >task;

  debug_mutex_wake_waiter(lock, waiter);
  wake_q_add(&wake_q, next);
 }

 if (owner & MUTEX_FLAG_HANDOFF)
  __mutex_handoff(lock, next);

 spin_unlock(&lock- >wait_lock);

 wake_up_q(&wake_q);
}

總結

本篇主要介紹了mutex互斥鎖的使用注意事項，介紹了mutex的主要函數(shù)接口以及獲取鎖的流程分析和釋放鎖的流程分析。通過本文的學習，我們基本可以了解了mutex的實現(xiàn)機制和使用方法。