linux内核 semaphore,2.4内核里semaphore源码的一个疑问

博主你好, 请教一个问题.

__down()里面有一段代码,  我觉得不那么保险.我先把__down的源码贴出来:

=============================================

void __down(struct semaphore * sem)

{

struct task_struct *tsk = current;

DECLARE_WAITQUEUE(wait, tsk);                        //定义一个"队列项", 等待者是当前进程

tsk->state = TASK_UNINTERRUPTIBLE;

add_wait_queue_exclusive(&sem->wait, &wait); //把当前进程添加到该信号量的wait queue里.

spin_lock_irq(&semaphore_lock);                           //抓取"大锁"

sem->sleepers++;

for (;;) {

int sleepers = sem->sleepers;

/*

* Add "everybody else" into it. They aren't

* playing, because we own the spinlock.

*/

if (!atomic_add_negative(sleepers - 1, &sem->count)) {    //临睡前最后一次尝试

sem->sleepers = 0;

break;

}

sem->sleepers = 1;    /* us - see -1 above */

spin_unlock_irq(&semaphore_lock);

schedule();                             //睡眠

tsk->state = TASK_UNINTERRUPTIBLE;

spin_lock_irq(&semaphore_lock);

}

spin_unlock_irq(&semaphore_lock);

remove_wait_queue(&sem->wait, &wait);   //取得信号量后, 退出该信号量的等待队列

tsk->state = TASK_RUNNING;

wake_up(&sem->wait);

}

我也是这两天才开始读linux的源码. 我先说说我读到的一点经验. 2.4内核里的semaphore结构体里面没有lock字段, 整个semaphor.c里是共用一个文件域的大锁, 就是semaphore.c里定义的semaphore_lock. 每当要操作semaphore结构体之前, 就先抓取这个"全局锁".

但是up()操作的全程都没有理睬这把锁, 我很好奇, 会不会出现这样一种bug呢:

为了方便分析, 假设除了当前进程, 没有别的进程在竞争这个信号量.

刚才说到up()的全程都没理会"大锁", 所以在整个__down()的过程中, 别的cpu上, 随时可能会有一个up()平行的运行. up()最终调用的是wake_up_process().

=====================================

inline void wake_up_process(struct task_struct * p)

{

unsigned long flags;

/*

* We want the common case fall through straight, thus the goto.

*/

spin_lock_irqsave(&runqueue_lock, flags);

p->state = TASK_RUNNING;    if (task_on_runqueue(p))        goto out;

add_to_runqueue(p);

reschedule_idle(p);

out:

spin_unlock_irqrestore(&runqueue_lock, flags);

}

===============================================

那么在__down()里面的这段区间,

add_wait_queue_exclusive(&sem->wait, &wait);

...

...

schedule();

也就是, current进入信号量排队之后, 调用schedule()之前, 我们随时可能遭受"wake_up_process()".

如果在if (!atomic_add_negative(sleepers - 1, &sem->count)) 这句之前被wake_up_process(), 倒也无所谓, 因为反正我们能通过这个if拿到信号量, (既然有人up, 肯定就是有门票了).

但是如果这个if失败, 我们就要睡眠了. 在我们调用schdule入睡之前的这个空隙里, 即执行这几行代码的时候:

sem->sleepers = 1;    /* us - see -1 above */

spin_unlock_irq(&semaphore_lock);

schedule();

我们遭到了wake_up_process.

会发生什么呢? 其实看wake_up_process()的源码,它也做不了什么( 因为我们已经在运行队列里了 ),  但它把我们的状态设置成TASK_RUNNING了.

就是说, 接下来, 我们是以"TASK_RUNNING"的身份调用schedule的.

更坏的是, 我们等于说是错过了这次up(), 再没有人来唤醒我们了.

我上面说的很麻烦, 简单的说, 就是, up()为什么不理睬semaphore_lock这个锁? 明明会出bug.

我想是我哪里错了, linux肯定不会有这种bug.

恳请指教.

------------------------------------------------------------------------------

*

为了方便一些, 我把相关的源码都贴上:

== down()和up()的入口函数来自 include/asm-i386/semaphore.h

static inline void down(struct semaphore * sem)

{

#if WAITQUEUE_DEBUG

CHECK_MAGIC(sem->__magic);

#endif

__asm__ __volatile__(

"# atomic down operation\n\t"

LOCK "decl %0\n\t" /* --sem->count */

"js 2f\n"

"1:\n"

".section .text.lock,\"ax\"\n"

"2:\tcall __down_failed\n\t"

"jmp 1b\n"

".previous"

:"=m" (sem->count)

:"c" (sem)

:"memory");

}

static inline void up(struct semaphore * sem)

{

#if WAITQUEUE_DEBUG

CHECK_MAGIC(sem->__magic);

#endif

__asm__ __volatile__(

"# atomic up operation\n\t"

LOCK "incl %0\n\t" /* ++sem->count */

"jle 2f\n"

"1:\n"

".section .text.lock,\"ax\"\n"

"2:\tcall __up_wakeup\n\t"

"jmp 1b\n"

".previous"

:"=m" (sem->count)

:"c" (sem)

:"memory");

}

===fall through失败后的操作, 都在arch/i386/kernel/semaphore.c

asm(

".align 4\n"

".globl __down_failed\n"

"__down_failed:\n\t"

"pushl %eax\n\t"

"pushl %edx\n\t"

"pushl %ecx\n\t"

"call __down\n\t"

"popl %ecx\n\t"

"popl %edx\n\t"

"popl %eax\n\t"

"ret"

);

void __down(struct semaphore * sem)

{

struct task_struct *tsk = current;

DECLARE_WAITQUEUE(wait, tsk);

tsk->state = TASK_UNINTERRUPTIBLE;

add_wait_queue_exclusive(&sem->wait, &wait);

spin_lock_irq(&semaphore_lock);

sem->sleepers++;

for (;;) {

int sleepers = sem->sleepers;

/*

* Add "everybody else" into it. They aren't

* playing, because we own the spinlock.

*/

if (!atomic_add_negative(sleepers - 1, &sem->count)) {

sem->sleepers = 0;

break;

}

sem->sleepers = 1; /* us - see -1 above */

spin_unlock_irq(&semaphore_lock);

schedule();

tsk->state = TASK_UNINTERRUPTIBLE;

spin_lock_irq(&semaphore_lock);

}

spin_unlock_irq(&semaphore_lock);

remove_wait_queue(&sem->wait, &wait);

tsk->state = TASK_RUNNING;

wake_up(&sem->wait);

}

void __up(struct semaphore *sem)

{

wake_up(&sem->wait);

}

asm(

".align 4\n"

".globl __up_wakeup\n"

"__up_wakeup:\n\t"

"pushl %eax\n\t"

"pushl %edx\n\t"

"pushl %ecx\n\t"

"call __up\n\t"

"popl %ecx\n\t"

"popl %edx\n\t"

"popl %eax\n\t"

"ret"

);

*

这部分的内容在403页前后.