内核中的原子上下文

内核的一个基本原则就是：在中断或者说原子上下文中，内核不能访问用户空间，而且内核是不能睡眠的。也就是说在这种情况下，内核是不能调用有可能引起睡眠的任何函数。一般来讲原子上下文指的是在中断或软中断中，以及在持有自旋锁的时候。内核提供了四个宏来判断是否处于这几种情况里：

#define  in_irq()             (hardirq_count()) //在处理硬中断中

#define  in_softirq()       (softirq_count()) //在处理软中断中

#define  in_interrupt()   (irq_count()) //在处理硬中断或软中断中

#define  in_atomic()      ((preempt_count() & ~PREEMPT_ACTIVE) != 0) //包含以上所有情况

 

这四个宏所访问的count都是thread_info->preempt_count。这个变量其实是一个位掩码。最低8位表示抢占计数，通常由spin_lock/spin_unlock修改，或程序员强制修改，同时表明内核容许的最大抢占深度是256。
8－15位表示软中断计数，通常由local_bh_disable/local_bh_enable修改，同时表明内核容许的最大软中断深度是256。
位16－27是硬中断计数，通常由enter_irq/exit_irq修改，同时表明内核容许的最大硬中断深度是4096。
第28位是PREEMPT_ACTIVE标志。用代码表示就是：

PREEMPT_MASK    0x000000ff
SOFTIRQ_MASK    0x0000ff00
HARDIRQ_MASK   0x0fff0000

凡是上面4个宏返回1得到地方都是原子上下文，是不容许内核访问用户空间，不容许内核睡眠的，不容许调用任何可能引起睡眠的函数。而且代表thread_info->preempt_count不是0，这就告诉内核，在这里面抢占被禁用。

但是，对于in_atomic()来说，在启用抢占的情况下，它工作的很好，可以告诉内核目前是否持有自旋锁，是否禁用抢占等。但是，在没有启用抢占的情况下，spin_lock根本不修改preempt_count，所以即使内核调用了spin_lock，持有了自旋锁，in_atomic()仍然会返回0，错误的告诉内核目前在非原子上下文中。所以凡是依赖in_atomic()来判断是否在原子上下文的代码，在禁抢占的情况下都是有问题的。

 

#ifndef LINUX_HARDIRQ_H
#define LINUX_HARDIRQ_H

#include <linux/preempt.h>
#ifdef CONFIG_PREEMPT
#include <linux/smp_lock.h>
#endif
#include <linux/lockdep.h>
#include <linux/ftrace_irq.h>
#include <asm/hardirq.h>
#include <asm/system.h>
/*
* We put the hardirq and softirq counter into the preemption
* counter. The bitmask has the following meaning:
*
* - bits 0-7 are the preemption count (max preemption depth: 256)
* - bits 8-15 are the softirq count (max # of softirqs: 256)
*
* The hardirq count can in theory reach the same as NR_IRQS.
* In reality, the number of nested IRQS is limited to the stack
* size as well. For archs with over 1000 IRQS it is not practical
* to expect that they will all nest. We give a max of 10 bits for
* hardirq nesting. An arch may choose to give less than 10 bits.
* m68k expects it to be 8.
*
* - bits 16-25 are the hardirq count (max # of nested hardirqs: 1024)
* - bit 26 is the NMI_MASK
* - bit 28 is the PREEMPT_ACTIVE flag
*
* PREEMPT_MASK: 0x000000ff
* SOFTIRQ_MASK: 0x0000ff00
* HARDIRQ_MASK: 0x03ff0000
*     NMI_MASK: 0x04000000
*/
#define PREEMPT_BITS    8
#define SOFTIRQ_BITS    8
#define NMI_BITS    1

#define MAX_HARDIRQ_BITS 10

#ifndef HARDIRQ_BITS
# define HARDIRQ_BITS    MAX_HARDIRQ_BITS
#endif

#if HARDIRQ_BITS > MAX_HARDIRQ_BITS
#error HARDIRQ_BITS too high!
#endif

#define PREEMPT_SHIFT    0
#define SOFTIRQ_SHIFT    (PREEMPT_SHIFT+ PREEMPT_BITS)
#define HARDIRQ_SHIFT    (SOFTIRQ_SHIFT+ SOFTIRQ_BITS)
#define NMI_SHIFT    (HARDIRQ_SHIFT+ HARDIRQ_BITS)

#define __IRQ_MASK(x)    ((1UL<< (x))-1)

#define PREEMPT_MASK    (__IRQ_MASK(PREEMPT_BITS)<< PREEMPT_SHIFT)
#define SOFTIRQ_MASK    (__IRQ_MASK(SOFTIRQ_BITS)<< SOFTIRQ_SHIFT)
#define HARDIRQ_MASK    (__IRQ_MASK(HARDIRQ_BITS)<< HARDIRQ_SHIFT)
#define NMI_MASK    (__IRQ_MASK(NMI_BITS)     << NMI_SHIFT)

#define PREEMPT_OFFSET    (1UL<< PREEMPT_SHIFT)
#define SOFTIRQ_OFFSET    (1UL<< SOFTIRQ_SHIFT)
#define HARDIRQ_OFFSET    (1UL<< HARDIRQ_SHIFT)
#define NMI_OFFSET    (1UL<< NMI_SHIFT)

#ifndef PREEMPT_ACTIVE
#define PREEMPT_ACTIVE_BITS    1
#define PREEMPT_ACTIVE_SHIFT    (NMI_SHIFT+ NMI_BITS)
#define PREEMPT_ACTIVE    (__IRQ_MASK(PREEMPT_ACTIVE_BITS)<< PREEMPT_ACTIVE_SHIFT)
#endif

#if PREEMPT_ACTIVE < (1 << (NMI_SHIFT+ NMI_BITS))
#error PREEMPT_ACTIVE is too low!
#endif

#define hardirq_count()    (preempt_count()& HARDIRQ_MASK)
#define softirq_count()    (preempt_count()& SOFTIRQ_MASK)
#define irq_count()    (preempt_count()& (HARDIRQ_MASK | SOFTIRQ_MASK \
                 | NMI_MASK))

/*
* Are we doing bottom half or hardware interrupt processing?
* Are we in a softirq context? Interrupt context?
*/
#define in_irq()        (hardirq_count())
#define in_softirq()        (softirq_count())
#define in_interrupt()        (irq_count())
/*
* Are we in NMI context?
*/
#define in_nmi()    (preempt_count()& NMI_MASK)

#if defined(CONFIG_PREEMPT)
# define PREEMPT_INATOMIC_BASE kernel_locked()
# define PREEMPT_CHECK_OFFSET1
#else
# define PREEMPT_INATOMIC_BASE0
# define PREEMPT_CHECK_OFFSET0
#endif

/*
 * Are we running in atomic context?  WARNING: this macro cannot
* always detect atomic context; in particular, it cannot know about
 * held spinlocks in non-preemptible kernels.  Thus it should not be
* used in the general case to determine whether sleeping is possible.
* Do not use in_atomic() in driver code.
*/
#define in_atomic()    ((preempt_count()& ~PREEMPT_ACTIVE)!= PREEMPT_INATOMIC_BASE)