linux内核线程死锁或死循环（soft lockup）之后如何让系统宕机重启

在开发内核模块或驱动时，如果处理失误，导致内核线程中出现死锁或者死循环，你会发现，除了重启之外，你没有任何可以做的。这时你的输入不起任何作用，终端（不是指远程的ssh工具）只会在那重复的输出类似“BUG: soft lockup - CPU#0 stuck for 67s! [fclustertool:2043]”，更无奈的是你重启之后导致系统挂起的堆栈信息也看不到，你所能做的就是一遍遍的加调试信息，一遍遍的重启机器（这是我的经历，现在想想很傻）。
  这种情况你肯定不是第一个遇到的，所以内核肯定会提供处理这种情况的一些机制。但是如何来找到这些机制在哪个地方，或者说根据什么信息去google呢？最有用的就是这句话“BUG: soft lockup - CPU#0 stuck for 67s! [fclustertool:2043]”，因为这句话提供你的信息量很大。首先，这条信息可以输出，说明即使发生死锁或者死循环，还是有代码可以执行。第二，可以通过这个日志信息，找到对应的处理函数，这个函数所在的模块就是用来处理CPU被过度使用时用到的。所以通过这个事情，可以看到内核打印出的只言片语都有可能成为你解决问题的关键，一定要从重视这些信息，从中找出有用的东西。
  我经常看的内核版本是官方的2.6.32内核，这个版本中我找到的函数是softlockup_tick()，这个函数在时钟中断的处理函数run_local_timers()中调用。这个函数会首先检查watchdog线程是否被挂起，如果不是watchdog线程，会检查当前占有CPU的线程占有的时间是否超过系统配置的阈值，即softlockup_thresh。如果当前占有CPU的时间过长，则会在系统日志中输出我们上面看到的那条日志。接下来才是最关键的，就是输出模块信息、寄存器信息和堆栈信息，检查softlockup_panic的值是否为1。如果softlockup_panic为1，则调用panic()让内核挂起，输出OOPS信息。代码如下所示：

[cpp] view
plain copy

/* 

* This callback runs from the timer interrupt, and checks 

* whether the watchdog thread has hung or not: 

*/  

void softlockup_tick(void)  

{  

    int this_cpu = smp_processor_id();  

    unsigned long touch_timestamp = per_cpu(touch_timestamp, this_cpu);  

    unsigned long print_timestamp;  

    struct pt_regs *regs = get_irq_regs();  

    unsigned long now;  

    ......  

  

    /* Warn about unreasonable delays: */  

    if (now <= (touch_timestamp + softlockup_thresh))  

        return;  

  

    per_cpu(print_timestamp, this_cpu) = touch_timestamp;  

    spin_lock(&print_lock);  

    printk(KERN_ERR "BUG: soft lockup - CPU#%d stuck for %lus! [%s:%d]\n",  

            this_cpu, now - touch_timestamp,  

            current->comm, task_pid_nr(current));  

    print_modules();  

    print_irqtrace_events(current);  

    if (regs)  

        show_regs(regs);  

    else  

        dump_stack();  

    spin_unlock(&print_lock);  

  

  

    if (softlockup_panic)  

        panic("softlockup: hung tasks");  

}  

  但是softlockup_panic的值默认竟然是0，所以在出现死锁或者死循环的时候，会一直只输出日志信息，而不会宕机，这个真是好坑啊！所以你得手动修改/proc/sys/kernel/softlockup_panic的值，让内核可以在死锁或者死循环的时候可以宕机。如果你的机器中安装了kdump，在重启之后，你会得到一份内核的core文件，这时从core文件中查找问题就方便很多了，而且再也不用手动重启机器了。如果你的内核是标准内核的话，可以通过修改/proc/sys/kernel/softlockup_thresh来修改超时的阈值，如果是CentOS内核的话，对应的文件是/proc/sys/kernel/watchdog_thresh。CentOS内核和标准内核还有一个地方不一样，就是处理CPU占用时间过长的函数，CentOS下是watchdog_timer_fn()函数。
  这里介绍下lockup的概念。lockup分为soft lockup和hard lockup。 soft lockup是指内核中有BUG导致在内核模式下一直循环的时间超过10s（根据实现和配置有所不同），而其他进程得不到运行的机会。hard softlockup是指内核已经挂起，可以通过watchdog这样的机制来获取详细信息。这两个概念比较类似。如果你想了解更多关于lockup的信息，可以参考这篇文档：http://www.mjmwired.net/kernel/Documentation/lockup-watchdogs.txt。
  注意上面说的这些，都是在内核线程中有效，对用户态的死循环没用。如果要监视用户态的死循环，或者内存不足等资源的情况，强烈推荐软件层面的watchdog。你可以自己开发基于soft watchdog的监控程序，也可以安装watchdog包通过修改配置来做，非常的方便。具体的操作可以参考下面的文章，都写的非常好，非常实用：
http://purplegrape.blog.51cto.com/1330104/1131910
http://www.ibm.com/developerworks/cn/linux/l-cn-watchdog/index.html#resources

附上一段bug信息：

[2098 : 2098 CRITICAL][utils.c:664]Erase flash at 0x00000000

[2098 : 2098 CRITICAL][utils.c:539]EraseFlashArea (0x0, 0x10000)

[2098 : 2098 CRITICAL][utils.c:561]EI Start is 0, size is 10000

[ 356.270000] BUG: soft lockup - CPU#0 stuck for 22s! [hostflasher:2098]
[ 356.270000] Modules linked in: videocap(O) hid(O) hdisk(O) floppy(O) ncsi(O) cdrom(O) usb1_hw(O) iUSB(O) usbe(O) bt_hw(O) bt(O) host_spi_flash_hw(O) host_spi_)
[ 356.270000]
[ 356.270000] CPU: 0 PID: 2098 Comm: hostflasher Tainted: P O 3.14.17-ami #1
[ 356.270000] task: c5fbbc20 ti: c4878000 task.ti: c4878000
[ 356.270000] PC is at ast_host_spi_flash_transfer+0x34/0x1dc [host_spi_flash_hw]
[ 356.270000] LR is at ast_host_spi_flash_transfer+0x28/0x1dc [host_spi_flash_hw]
[ 356.270000] pc : [<bf0e60e4>] lr : [<bf0e60d8>] psr: 60000013
[ 356.270000] sp : c4879c60 ip : 00000004 fp : c4879c84
[ 356.270000] r10: bf0e2348 r9 : 00000005 r8 : 00000000
[ 356.270000] r7 : 00000001 r6 : c4879c97 r5 : 00000001 r4 : c4879c96
[ 356.270000] r3 : cd946000 r2 : 7100d2e6 r1 : bf0e6730 r0 : 00000000
[ 356.270000] Flags: nZCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user
[ 356.270000] Control: 00c5387d Table: 84808008 DAC: 00000015
[ 356.270000] CPU: 0 PID: 2098 Comm: hostflasher Tainted: P O 3.14.17-ami #1
[ 356.270000] Backtrace:
[ 356.270000] [<c000b494>] (dump_backtrace) from [<c000b83c>] (show_stack+0x18/0x1c)
[ 356.270000] r6:c4879c18 r5:00000135 r4:c4879c18 r3:c04d2008
[ 356.270000] [<c000b824>] (show_stack) from [<c0386a84>] (dump_stack+0x20/0x28)
[ 356.270000] [<c0386a64>] (dump_stack) from [<c0009f9c>] (show_regs+0x24/0x2c)
[ 356.270000] [<c0009f78>] (show_regs) from [<c004dfc4>] (watchdog_timer_fn+0xe0/0x130)
[ 356.270000] r4:c04dc9f8 r3:a3febb35
[ 356.270000] [<c004dee4>] (<span style="color:#ff0000;">watchdog_timer_fn</span>) from [<c002da60>] (hrtimer_run_queues+0x120/0x1c0)
[ 356.270000] r9:1439d6a1 r8:3a814c28 r7:ffffffff r6:c04d9858 r5:c04d9858 r4:c04dca08
[ 356.270000] [<c002d940>] (hrtimer_run_queues) from [<c001c614>] (run_local_timers+0x10/0x1c)
[ 356.270000] r10:bf0e2348 r9:00000005 r8:00000000 r7:00000000 r6:00000000 r5:c5fbbc20
[ 356.270000] r4:c4878000
[ 356.270000] [<c001c604>] (run_local_timers) from [<c001c650>] (update_process_times+0x30/0x5c)
[ 356.270000] [<c001c620>] (update_process_times) from [<c000b3d4>] (timer_tick+0x30/0x38)
[ 356.270000] r6:00000010 r5:c04dacfc r4:c04d7fd0 r3:60000013
[ 356.270000] [<c000b3a4>] (timer_tick) from [<c0010d98>] (ast_timer_interrupt+0x20/0x2c)
[ 356.270000] [<c0010d78>] (ast_timer_interrupt) from [<c003c8d4>] (handle_irq_event_percpu+0x34/0x19c)
[ 356.270000] [<c003c8a0>] (handle_irq_event_percpu) from [<c003ca6c>] (handle_irq_event+0x30/0x40)
[ 356.270000] r10:bf0e2348 r9:00000005 r8:00000000 r7:c4879c4c r6:fe6c0000 r5:00000000
[ 356.270000] r4:c04dacfc
[ 356.270000] [<c003ca3c>] (handle_irq_event) from [<c003f198>] (handle_level_irq+0xc8/0xd8)
[ 356.270000] r4:c04dacfc r3:00020000
[ 356.270000] [<c003f0d0>] (handle_level_irq) from [<c003c2f4>] (generic_handle_irq+0x28/0x38)
[ 356.270000] r4:00000010 r3:c003f0d0
[ 356.270000] [<c003c2cc>] (generic_handle_irq) from [<c0009b30>] (handle_IRQ+0x68/0x8c)
[ 356.270000] r4:00000010 r3:00000040
[ 356.270000] [<c0009ac8>] (handle_IRQ) from [<c0008210>] (asm_do_IRQ+0x10/0x14)
[ 356.270000] r5:60000013 r4:bf0e60e4
[ 356.270000] [<c0008200>] (asm_do_IRQ) from [<c000c274>] (__irq_svc+0x34/0x104)
[ 356.270000] Exception stack(0xc4879c18 to 0xc4879c60)
[ 356.270000] 9c00: 00000000 bf0e6730
[ 356.270000] 9c20: 7100d2e6 cd946000 c4879c96 00000001 c4879c97 00000001 00000000 00000005
[ 356.270000] 9c40: bf0e2348 c4879c84 00000004 c4879c60 bf0e60d8 bf0e60e4 60000013 ffffffff
[ 356.270000] [<bf0e60b0>] (ast_host_spi_flash_transfer [host_spi_flash_hw]) from [<bf0e1250>] (host_spi_flash_unregister+0x9c/0x2f8 [host_spi_flash])
[ 356.270000] r8:00000000 r7:00000000 r6:c4879c96 r5:bf0e65e0 r4:01827fe5 r3:00000001
[ 356.270000] [<bf0e11ec>] (host_spi_flash_unregister [host_spi_flash]) from [<bf0e1668>] (host_spi_flash_erase+0x1bc/0x294 [host_spi_flash])
[ 356.270000] r8:00010000 r7:bf0e65e4 r6:00010000 r5:bf0e65e0 r4:00000000
[ 356.270000] [<bf0e14ac>] (host_spi_flash_erase [host_spi_flash]) from [<c023a4c8>] (part_erase+0x38/0x88)
[ 356.270000] r10:00000000 r9:00000000 r8:000001b8 r7:c34a68c0 r6:c89a5800 r5:00000000
[ 356.270000] r4:00000000
[ 356.270000] [<c023a490>] (part_erase) from [<c0237d48>] (mtd_erase+0x78/0x8c)
[ 356.270000] r8:00010000 r7:c4879e9c r6:c34a68c0 r5:00000000 r4:00000000 r3:00010000
[ 356.270000] [<c0237cd0>] (mtd_erase) from [<c023c560>] (mtdchar_ioctl+0x46c/0xe94)
[ 356.270000] r9:c4878000 r8:c4878000 r6:c34a68c0 r5:c89a5800 r4:becd4b80 r3:00000000
[ 356.270000] [<c023c0f4>] (mtdchar_ioctl) from [<c023cfbc>] (mtdchar_unlocked_ioctl+0x34/0x4c)
[ 356.270000] r9:c4878000 r8:00000006 r7:40084d02 r6:becd4b80 r5:c3435820 r4:c04eeaf0
[ 356.270000] [<c023cf88>] (mtdchar_unlocked_ioctl) from [<c008a558>] (vfs_ioctl+0x20/0x3c)
[ 356.270000] r7:40084d02 r6:00000006 r5:c929eaa0 r4:becd4b80
[ 356.270000] [<c008a538>] (vfs_ioctl) from [<c008aff4>] (do_vfs_ioctl+0x4f8/0x54c)
[ 356.270000] [<c008aafc>] (do_vfs_ioctl) from [<c008b088>] (SyS_ioctl+0x40/0x5c)
[ 356.270000] r9:c4878000 r8:00000006 r7:40084d02 r6:becd4b80 r5:c3435820 r4:c3435820
[ 356.270000] [<c008b048>] (SyS_ioctl) from [<c0009220>] (ret_fast_syscall+0x0/0x30)
[ 356.270000] r8:c00093a4 r7:00000036 r6:00010000 r5:00000000 r4:00010000 r3:00000001


from： http://blog.csdn.net/justlinux2010/article/details/8739395