Linux 调试之SysRq

文章目录

前言

system request (SysRq) keys：是预定义的(在内核中硬编码)键组合，可触发各种操作。

It is a ‘magical’ key combo you can hit which the kernel will respond to regardless of whatever else it is doing, unless it is completely locked up.

Sysrq被称为”魔术组合键”，是内建于Linux内核的调试工具，只要内核没有完全锁住，不管内核在做什么事情，内核都会响应这一系列组合键，使用这些组合键可以搜集包括系统内存使用、CPU任务处理、进程运行状态等系统运行信息。

SysRq 键在确认内核运行、调查内核死机原因等各种情况下都非常有效。

一、SysRq简介

要想使用 Magic SysRq key 打开内核配置选项 CONFIG_MAGIC_SYSRQ，一般的发行版都默认打开该配置选项：

#
# Kernel hacking
#
......
CONFIG_MAGIC_SYSRQ=y

当运行一个编译了SysRq的内核时，/proc/sys/kernel/sysrq文件的值控制着允许通过SysRq键（键盘组合键）调用的函数，下面是/proc/sys/kernel/sysrq中可能的值列表：

0 - disable sysrq completely

1 - enable all functions of sysrq

>1 - bitmask of allowed sysrq functions (see below for detailed function description):

	  2 =   0x2 - enable control of console logging level
	  4 =   0x4 - enable control of keyboard (SAK, unraw)
	  8 =   0x8 - enable debugging dumps of processes etc.
	 16 =  0x10 - enable sync command
	 32 =  0x20 - enable remount read-only
	 64 =  0x40 - enable signalling of processes (term, kill, oom-kill)
	128 =  0x80 - allow reboot/poweroff
	256 = 0x100 - allow nicing of all RT tasks

/proc/sys/kernel/sysrq文件的值控制着SysRq键（键盘组合键）的一些功能。

可以通过以下命令在文件中设置该值：

echo "number" >/proc/sys/kernel/sysrq

比如：

echo 1 >/proc/sys/kernel/sysrq

/proc/sys/kernel/sysrq的值设置为1时，使能SysRq键的所有功能。

或者通过 sysctl 命令也可以设置或者读写内核参数：

sysctl -w kernel.sysrq = 1

NAME
       sysctl - configure kernel parameters at runtime

DESCRIPTION
       sysctl  is used to modify kernel parameters at runtime.  The parameters available are those listed under /proc/sys/.  Procfs is required for sysctl support in Linux.  You
       can use sysctl to both read and write sysctl data.

       -w, --write
              Use this option when you want to change a sysctl setting.

备注：/proc/sys/kernel/sysrq的值只影响通过键盘组合键的调用。通过/proc/sysrq-trigger调用任何操作总是被允许的(由具有管理权限的用户)。
即 /proc/sys/kernel/sysrq的值只是影响键盘组合键触发内核操作，对于 /proc/sysrq-trigger触发内核操作没有影响。

内核配置选项中使能CONFIG_MAGIC_SYSRQ选项，这样系统启动之后，会生成/proc/sysrq-trigger节点用于调试。

二、SysRq的使用

SysRq的使用有两种方式：
第一种是键盘组合键：Alt+SysRq + command key 。(受到/proc/sys/kernel/sysrq值的影响)
第二种是修改/proc/sysrq-trigger文件的值。(不受/proc/sys/kernel/sysrq值的影响)

接下来主要介绍第二种方式：

/proc/sysrq-trigger文件的值不受/proc/sys/kernel/sysrq值的影响，所以我将/proc/sys/kernel/sysrq的值设置为0，依然可以触发内核的各种事件。

[root@localhost ~]# echo 0 > /proc/sys/kernel/sysrq
[root@localhost ~]# cat /proc/sys/kernel/sysrq
0

echo <command key> > /proc/sysrq-trigger

比如：

echo t > /proc/sysrq-trigger

-t 选项把当前的任务快照保存下来，将转储当前任务及其信息的列表到控制台。

SysRq可以很好地追踪系统瞬时状态，即系统快照。

列举一些用于内核调试的command key（区分字母大小写）：

Command	Function
c	Will perform a system crash and a crashdump will be taken if configured.
d	Shows all locks that are held.
l	Shows a stack backtrace for all active CPUs.
m	Will dump current memory info to your console.
p	Will dump the current registers and flags to your console.
l	Shows a stack backtrace for all active CPUs.
t	Will dump a list of current tasks and their information to your console.
w	Dumps tasks that are in uninterruptable (blocked) state.

SysRq可以观察当前的内存快照、任务快照，可以构造 vmcore 把系统的所有信息都保存下来（-c 选项），甚至还可以在内存紧张的时候用它杀掉内存开销最大的那个进程。

2.1 获取内存快照

echo m > /proc/sysrq-trigger

[580415.132207] SysRq : Show Memory
[580415.132217] Mem-Info:
[580415.132231] active_anon:70197 inactive_anon:8419 isolated_anon:0
 active_file:91044 inactive_file:128367 isolated_file:0
 unevictable:0 dirty:0 writeback:0 unstable:0
 slab_reclaimable:32560 slab_unreclaimable:16646
 mapped:24361 shmem:8655 pagetables:5377 bounce:0
 free:1569061 free_pcp:1121 free_cma:0
[580415.132243] Node 0 DMA free:15896kB min:136kB low:168kB high:204kB active_anon:0kB inactive_anon:0kB active_file:0kB inactive_file:0kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:15992kB managed:15896kB mlocked:0kB dirty:0kB writeback:0kB mapped:0kB shmem:0kB slab_reclaimable:0kB slab_unreclaimable:0kB kernel_stack:0kB pagetables:0kB unstable:0kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no
[580415.132259] lowmem_reserve[]: 0 1960 7687 7687
[580415.132267] Node 0 DMA32 free:1619040kB min:17204kB low:21504kB high:25804kB active_anon:77084kB inactive_anon:9124kB active_file:88908kB inactive_file:126020kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:2257312kB managed:2010956kB mlocked:0kB dirty:0kB writeback:0kB mapped:19716kB shmem:9336kB slab_reclaimable:33356kB slab_unreclaimable:11568kB kernel_stack:1408kB pagetables:6192kB unstable:0kB bounce:0kB free_pcp:2264kB local_pcp:308kB free_cma:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no
[580415.132282] lowmem_reserve[]: 0 0 5726 5726
[580415.132289] Node 0 Normal free:4641308kB min:50240kB low:62800kB high:75360kB active_anon:203704kB inactive_anon:24552kB active_file:275268kB inactive_file:387448kB unevictable:0kB isolated(anon):0kB isolated(file):0kB present:5996544kB managed:5863960kB mlocked:0kB dirty:0kB writeback:0kB mapped:77728kB shmem:25284kB slab_reclaimable:96884kB slab_unreclaimable:55016kB kernel_stack:4368kB pagetables:15316kB unstable:0kB bounce:0kB free_pcp:2220kB local_pcp:212kB free_cma:0kB writeback_tmp:0kB pages_scanned:0 all_unreclaimable? no
[580415.132304] lowmem_reserve[]: 0 0 0 0
[580415.132311] Node 0 DMA: 2*4kB (U) 2*8kB (U) 2*16kB (U) 1*32kB (U) 3*64kB (U) 2*128kB (U) 0*256kB 0*512kB 1*1024kB (U) 1*2048kB (M) 3*4096kB (M) = 15896kB
[580415.132339] Node 0 DMA32: 156*4kB (UEM) 140*8kB (UEM) 295*16kB (UEM) 439*32kB (UEM) 251*64kB (UEM) 115*128kB (UM) 34*256kB (UM) 11*512kB (UEM) 3*1024kB (M) 3*2048kB (UM) 377*4096kB (M) = 1619040kB
[580415.132369] Node 0 Normal: 455*4kB (UEM) 630*8kB (UEM) 1329*16kB (UEM) 1244*32kB (UEM) 719*64kB (UEM) 404*128kB (UEM) 243*256kB (UM) 110*512kB (UEM) 63*1024kB (UEM) 70*2048kB (UEM) 1013*4096kB (UM) = 4641308kB
[580415.132401] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=1048576kB
[580415.132406] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB
[580415.132409] 228066 total pagecache pages
[580415.132415] 0 pages in swap cache
[580415.132419] Swap cache stats: add 0, delete 0, find 0/0
[580415.132422] Free swap  = 8126460kB
[580415.132425] Total swap = 8126460kB
[580415.132428] 2067462 pages RAM
[580415.132431] 0 pages HighMem/MovableOnly
[580415.132434] 94759 pages reserved

2.2 获取任务快照

echo t > /proc/sysrq-trigger

显示了每个任务当前运行的情况和每个CPU当前运行任务的情况：

[580573.523247] SysRq : Show State
[580573.523256]   task                        PC stack   pid father
[580573.523262] systemd         S ffff9709fcf28000     0     1      0 0x00000000
[580573.523271] Call Trace:
[580573.523285]  [<ffffffff9d767bc9>] schedule+0x29/0x70
[580573.523294]  [<ffffffff9d766dfd>] schedule_hrtimeout_range_clock+0x12d/0x150
[580573.523305]  [<ffffffff9d28e839>] ? ep_scan_ready_list.isra.7+0x1b9/0x1f0
[580573.523312]  [<ffffffff9d766e33>] schedule_hrtimeout_range+0x13/0x20
[580573.523319]  [<ffffffff9d28eace>] ep_poll+0x23e/0x360
[580573.523328]  [<ffffffff9d0d67b0>] ? wake_up_state+0x20/0x20
[580573.523336]  [<ffffffff9d28ff9d>] SyS_epoll_wait+0xed/0x120
[580573.523345]  [<ffffffff9d774ddb>] system_call_fastpath+0x22/0x27
[580573.523351] kthreadd        S ffff9709fcf29040     0     2      0 0x00000000
[580573.523358] Call Trace:
[580573.523366]  [<ffffffff9d767bc9>] schedule+0x29/0x70
[580573.523373]  [<ffffffff9d0c2625>] kthreadd+0x2f5/0x300
[580573.523381]  [<ffffffff9d0c2330>] ? kthread_create_on_cpu+0x60/0x60
[580573.523388]  [<ffffffff9d774c1d>] ret_from_fork_nospec_begin+0x7/0x21
[580573.523395]  [<ffffffff9d0c2330>] ? kthread_create_on_cpu+0x60/0x60
[580573.523400] ksoftirqd/0     S ffff9709fcf2a080     0     3      2 0x00000000
[580573.523406] Call Trace:
[580573.523413]  [<ffffffff9d767bc9>] schedule+0x29/0x70
[580573.523420]  [<ffffffff9d0ca562>] smpboot_thread_fn+0xe2/0x1a0
[580573.523426]  [<ffffffff9d0ca480>] ? lg_double_unlock+0x40/0x40
[580573.523432]  [<ffffffff9d0c1c31>] kthread+0xd1/0xe0
[580573.523439]  [<ffffffff9d0c1b60>] ? insert_kthread_work+0x40/0x40
[580573.523446]  [<ffffffff9d774c1d>] ret_from_fork_nospec_begin+0x7/0x21
[580573.523453]  [<ffffffff9d0c1b60>] ? insert_kthread_work+0x40/0x40
[580573.523457] kworker/0:0H    S ffff9709fcf2c100     0     5      2 0x00000000
[580573.523474] Call Trace:
[580573.523483]  [<ffffffff9d0b9dea>] ? process_one_work+0x21a/0x440
[580573.523489]  [<ffffffff9d767bc9>] schedule+0x29/0x70
[580573.523496]  [<ffffffff9d0bae99>] worker_thread+0x1d9/0x3c0
[580573.523504]  [<ffffffff9d0bacc0>] ? manage_workers.isra.25+0x2a0/0x2a0
[580573.523509]  [<ffffffff9d0c1c31>] kthread+0xd1/0xe0
[580573.523516]  [<ffffffff9d0c1b60>] ? insert_kthread_work+0x40/0x40
[580573.523523]  [<ffffffff9d774c1d>] ret_from_fork_nospec_begin+0x7/0x21
[580573.523529]  [<ffffffff9d0c1b60>] ? insert_kthread_work+0x40/0x40

......
//cpu0 当前运行任务的情况
[580573.538243] cpu#0, 3600.000 MHz
[580573.538244]   .nr_running                    : 0
[580573.538245]   .load                          : 0
[580573.538246]   .nr_switches                   : 20421008
[580573.538247]   .nr_load_updates               : 11012978
[580573.538247]   .nr_uninterruptible            : -655
[580573.538248]   .next_balance                  : 4875.237938
[580573.538249]   .curr->pid                     : 0
[580573.538250]   .clock                         : 580573538.220004
[580573.538251]   .cpu_load[0]                   : 0
[580573.538252]   .cpu_load[1]                   : 237
[580573.538252]   .cpu_load[2]                   : 493
[580573.538253]   .cpu_load[3]                   : 655
[580573.538254]   .cpu_load[4]                   : 599
[580573.538254]   .avg_idle                      : 5107
[580573.538255]   .max_idle_balance_cost         : 500000

[580573.538258] cfs_rq[0]:/
[580573.538259]   .exec_clock                    : 0.000000
[580573.538260]   .MIN_vruntime                  : 0.000001
[580573.538261]   .min_vruntime                  : 2824982.413943
[580573.538262]   .max_vruntime                  : 0.000001
[580573.538263]   .spread                        : 0.000000
[580573.538263]   .spread0                       : 0.000000
[580573.538264]   .nr_spread_over                : 0
[580573.538265]   .nr_running                    : 0
[580573.538266]   .load                          : 0
[580573.538266]   .runnable_load_avg             : 0
[580573.538267]   .blocked_load_avg              : 317
[580573.538268]   .tg_load_avg                   : 0
[580573.538269]   .tg_load_contrib               : 0
[580573.538269]   .tg_runnable_contrib           : 0
[580573.538270]   .tg->runnable_avg              : 0
[580573.538271]   .tg->cfs_bandwidth.timer_active: 0
[580573.538272]   .throttled                     : 0
[580573.538273]   .throttle_count                : 0
[580573.538274]   .avg->runnable_avg_sum         : 8138
[580573.538275]   .avg->runnable_avg_period      : 46352

[580573.538283] rt_rq[0]:/
[580573.538284]   .rt_nr_running                 : 0
[580573.538285]   .rt_throttled                  : 0
[580573.538285]   .rt_time                       : 0.000000
[580573.538286]   .rt_runtime                    : 950.000000

[580573.538288] runnable tasks:
[580573.538289]             task   PID         tree-key  switches  prio     wait-time             sum-exec        sum-sleep
[580573.538290] ----------------------------------------------------------------------------------------------------------
[580573.538291]      ksoftirqd/0     3   2824973.436009    108307   120         0.000000      1680.428964         0.000000 0 /
[580573.538294]     kworker/0:0H     5      2776.777100         9   100         0.000000         0.089769         0.000000 0 /
[580573.538296]      migration/0     7         0.000000     31932     0         0.000000       272.980359         0.000000 0 /
[580573.538299]           rcu_bh     8        38.123474         2   120         0.000000         0.000709         0.000000 0 /
[580573.538301]        rcu_sched     9   2824973.474298   1248369   120         0.000000     23413.300648         0.000000 0 /
[580573.538305]    lru-add-drain    10        42.124246         2   100         0.000000         0.001218         0.000000 0 /
[580573.538307]       watchdog/0    11        -5.967430    145146     0         0.000000      2874.498528         0.000000 0 /
[580573.538312]     kmpath_rdacd    62       635.006034         2   100         0.000000         0.012378         0.000000 0 /
[580573.538315]    ipv6_addrconf    67       712.678812         2   100         0.000000         0.100476         0.000000 0 /
[580573.538318]       scsi_tmf_0   883      1679.281366         2   100         0.000000         0.003571         0.000000 0 /
[580573.538321]       scsi_tmf_1   907      1687.801687         2   100         0.000000         0.002658         0.000000 0 /
[580573.538324]       scsi_tmf_2   921      1697.538324         2   100         0.000000         0.003882         0.000000 0 /
[580573.538326]       scsi_tmf_4   951      1713.204537         2   100         0.000000         0.003118         0.000000 0 /
[580573.538329]       scsi_tmf_5   964      1721.307070         2   100         0.000000         0.002513         0.000000 0 /

//cpu当前任务运行的情况
[580573.538547] cpu#1, 3600.000 MHz
......

perf和ftrace一般都是采集一个时间段内的信息，对于追踪系统瞬时状态即系统快照sysrq更合适。

下面是一个脚本来把当前 CPU 正在做的工作记录下来，记录高于CPU运行在内核态时间较多的情况：

[root@localhost ~]#  top -bn2 | grep "Cpu(s)" | tail -1
%Cpu(s):  0.3 us,  0.2 sy,  0.0 ni, 99.4 id,  0.0 wa,  0.0 hi,  0.0 si,  0.0 st
[root@localhost ~]#  top -bn2 | grep "Cpu(s)" | tail -1 | awk '{print $1, $2, $3, $4, $5}'
%Cpu(s): 0.4 us, 0.2 sy,

#!/bin/sh

while [ 1 ]; do
     top -bn2 | grep "Cpu(s)" | tail -1 | awk '{ # $2 is usr, $4 is sys. if ($2 < 30.0 && $4 > 15.0) { # save the current usr and sys into a tmp file while ("date" | getline date) { split(date, str, " "); prefix=sprintf("%s_%s_%s_%s", str[2],str[3], str[4], str[5]); } sys_usr_file=sprintf("/tmp/%s_info.highsys", prefix); print $2 > sys_usr_file; print $4 >> sys_usr_file; # run sysrq system("echo t > /proc/sysrq-trigger"); } }'
     sleep 1m
done

这个脚本会检测 sys 利用率高于 15% 同时 usr 较低的情况，也就是说检测 CPU 是否在内核里花费了太多时间。如果出现这种情况，就会运行 sysrq 来保存当前任务快照。

上述代码来自于极客时间：Linux内核技术实战

三、SysRq源码解析

3.1 源码分析

SysRq相关的内核源码在内核驱动目录下：

/drivers/tty/sysrq.c

Magic SysRQ系统通过在key op lookup table上注册键操作来工作，key table在编译时注册了许多操作，但它是可变的，并且导出了 2 个函数作为其接口：

register_sysrq_key and unregister_sysrq_key.

static int __sysrq_swap_key_ops(int key, const struct sysrq_key_op *insert_op_p,
				const struct sysrq_key_op *remove_op_p)
{ 
   
	int retval;

	spin_lock(&sysrq_key_table_lock);
	if (__sysrq_get_key_op(key) == remove_op_p) { 
   
		__sysrq_put_key_op(key, insert_op_p);
		retval = 0;
	} else { 
   
		retval = -1;
	}
	spin_unlock(&sysrq_key_table_lock);

	/* * A concurrent __handle_sysrq either got the old op or the new op. * Wait for it to go away before returning, so the code for an old * op is not freed (eg. on module unload) while it is in use. */
	synchronize_rcu();

	return retval;
}

int register_sysrq_key(int key, const struct sysrq_key_op *op_p)
{ 
   
	return __sysrq_swap_key_ops(key, op_p, NULL);
}
EXPORT_SYMBOL(register_sysrq_key);

int unregister_sysrq_key(int key, const struct sysrq_key_op *op_p)
{ 
   
	return __sysrq_swap_key_ops(key, NULL, op_p);
}
EXPORT_SYMBOL(unregister_sysrq_key);

/* Key Operations table and lock */
static DEFINE_SPINLOCK(sysrq_key_table_lock);

static const struct sysrq_key_op *sysrq_key_table[62] = { 
   
	&sysrq_loglevel_op,		/* 0 */
	&sysrq_loglevel_op,		/* 1 */
	&sysrq_loglevel_op,		/* 2 */
	&sysrq_loglevel_op,		/* 3 */
	&sysrq_loglevel_op,		/* 4 */
	&sysrq_loglevel_op,		/* 5 */
	&sysrq_loglevel_op,		/* 6 */
	&sysrq_loglevel_op,		/* 7 */
	&sysrq_loglevel_op,		/* 8 */
	&sysrq_loglevel_op,		/* 9 */

	/* * a: Don't use for system provided sysrqs, it is handled specially on * sparc and will never arrive. */
	NULL,				/* a */
	&sysrq_reboot_op,		/* b */
	&sysrq_crash_op,		/* c */
	&sysrq_showlocks_op,		/* d */
	&sysrq_term_op,			/* e */
	&sysrq_moom_op,			/* f */
	/* g: May be registered for the kernel debugger */
	NULL,				/* g */
	NULL,				/* h - reserved for help */
	&sysrq_kill_op,			/* i */
	&sysrq_thaw_op,			/* j */
	&sysrq_SAK_op,			/* k */
	&sysrq_showallcpus_op,		/* l */
	&sysrq_showmem_op,		/* m */
	&sysrq_unrt_op,			/* n */
	/* o: This will often be registered as 'Off' at init time */
	NULL,				/* o */
	&sysrq_showregs_op,		/* p */
	&sysrq_show_timers_op,		/* q */
	&sysrq_unraw_op,		/* r */
	&sysrq_sync_op,			/* s */
	&sysrq_showstate_op,		/* t */
	&sysrq_mountro_op,		/* u */
	/* v: May be registered for frame buffer console restore */
	NULL,				/* v */
	&sysrq_showstate_blocked_op,	/* w */
	/* x: May be registered on mips for TLB dump */
	/* x: May be registered on ppc/powerpc for xmon */
	/* x: May be registered on sparc64 for global PMU dump */
	NULL,				/* x */
	/* y: May be registered on sparc64 for global register dump */
	NULL,				/* y */
	&sysrq_ftrace_dump_op,		/* z */
	NULL,				/* A */
	NULL,				/* B */
	NULL,				/* C */
	......
}

static int __init sysrq_init(void)
{ 
   
	(1) 初始化/proc/sysrq-trigger , 通过 /proc/sysrq-trigger 文件的值触发内核操作
	sysrq_init_procfs();

	(2) 初始化组合键 Alt + SysRq + key , 通过组合键 Alt + SysRq + key 触发内核操作
	if (sysrq_on())
		sysrq_register_handler();

	return 0;
}
device_initcall(sysrq_init);

（1）/proc/sysrq-trigger

#ifdef CONFIG_PROC_FS
/* * writing 'C' to /proc/sysrq-trigger is like sysrq-C */
static ssize_t write_sysrq_trigger(struct file *file, const char __user *buf,
				   size_t count, loff_t *ppos)
{ 
   
	if (count) { 
   
		char c;

		if (get_user(c, buf))
			return -EFAULT;
		__handle_sysrq(c, false);
	}

	return count;
}

static const struct proc_ops sysrq_trigger_proc_ops = { 
   
	.proc_write	= write_sysrq_trigger,
	.proc_lseek	= noop_llseek,
};

static void sysrq_init_procfs(void)
{ 
   
	if (!proc_create("sysrq-trigger", S_IWUSR, NULL,
			 &sysrq_trigger_proc_ops))
		pr_err("Failed to register proc interface\n");
}


#endif /* CONFIG_PROC_FS */

执行 echo > /proc/sysrq-trigger 就会调用 __handle_sysrq 函数，找到sysrq_key_table表中对应的操作函数。

/* * get and put functions for the table, exposed to modules. */
static const struct sysrq_key_op *__sysrq_get_key_op(int key)
{ 
   
	const struct sysrq_key_op *op_p = NULL;
	int i;

	i = sysrq_key_table_key2index(key);
	if (i != -1)
		op_p = sysrq_key_table[i];

	return op_p;
}

void __handle_sysrq(int key, bool check_mask)
{ 
   
	const struct sysrq_key_op *op_p;

	......
	
	rcu_sysrq_start();
	rcu_read_lock();


	op_p = __sysrq_get_key_op(key);
	if (op_p) { 
   
		/* * Should we check for enabled operations (/proc/sysrq-trigger * should not) and is the invoked operation enabled? */
		if (!check_mask || sysrq_on_mask(op_p->enable_mask)) { 
   
			......
			op_p->handler(key);
		} 
	}
	rcu_read_unlock();
	rcu_sysrq_end();
	......
}

(2) 初始化组合键 Alt + SysRq + key

static struct input_handler sysrq_handler = { 
   
	.filter		= sysrq_filter,
	.connect	= sysrq_connect,
	.disconnect	= sysrq_disconnect,
	.name		= "sysrq",
	.id_table	= sysrq_ids,
};

static inline void sysrq_register_handler(void)
{ 
   
	int error;

	sysrq_of_get_keyreset_config();

	error = input_register_handler(&sysrq_handler);
	if (error)
		pr_err("Failed to register input handler, error %d", error);
}

static bool sysrq_handle_keypress(struct sysrq_state *sysrq,
				  unsigned int code, int value)
{ 
   
	......

	default:
		if (sysrq->active && value && value != 2) { 
   
			unsigned char c = sysrq_xlate[code];

			sysrq->need_reinject = false;
			if (sysrq->shift_use != KEY_RESERVED)
				c = toupper(c);
			__handle_sysrq(c, true);
		}
		break;
	}

	......

	return suppress;
}

static bool sysrq_filter(struct input_handle *handle,
			 unsigned int type, unsigned int code, int value)
{ 
   
	struct sysrq_state *sysrq = handle->private;
	bool suppress;

	/* * Do not filter anything if we are in the process of re-injecting * Alt+SysRq combination. */
	if (sysrq->reinjecting)
		return false;

	switch (type) { 
   
	
	......
	case EV_KEY:
		suppress = sysrq_handle_keypress(sysrq, code, value);
		break;
	......
	
	}

	return suppress;

组合键 Alt + SysRq + key 最终也会调用和设置 /proc/sysrq-trigger 的值一样调用 __handle_sysrq 函数。

3.2 源码例程分析

比如执行 echo t > /proc/sysrq-trigger，就会在sysrq_key_table表中找到sysrq_showstate_op：

static const struct sysrq_key_op *sysrq_key_table[62] = { 
   
	......
	&sysrq_showstate_op,		/* t */
	......
}

然后执行 sysrq_showstate_op ->handler()：

static void sysrq_handle_showstate(int key)
{ 
   
	show_state();
	show_all_workqueues();
}
static const struct sysrq_key_op sysrq_showstate_op = { 
   
	.handler	= sysrq_handle_showstate,
	.help_msg	= "show-task-states(t)",
	.action_msg	= "Show State",
	.enable_mask	= SYSRQ_ENABLE_DUMP,
};

最终执行sysrq_handle_showstate函数。

3.3 add SysRQ key events to a module

// /include/linux/sysrq.h

struct sysrq_key_op { 
   
	void (* const handler)(int);
	const char * const help_msg;
	const char * const action_msg;
	const int enable_mask;
};

#ifdef CONFIG_MAGIC_SYSRQ

/* Generic SysRq interface -- you may call it from any device driver, supplying * ASCII code of the key, pointer to registers and kbd/tty structs (if they * are available -- else NULL's). */

void handle_sysrq(int key);
void __handle_sysrq(int key, bool check_mask);
int register_sysrq_key(int key, const struct sysrq_key_op *op);
int unregister_sysrq_key(int key, const struct sysrq_key_op *op);

为了在表中注册一个基本函数，你必须首先包含头include/linux/sysrq.h，这将定义你需要的所有其他东西。接下来，您必须创建一个sysrq_key_op结构体，填充sysrq_key_op结构体：
A) the key handler function you will use.
B) a help_msg string, that will print when SysRQ prints help.
C) an action_msg string, that will print right before your handler is called.

the key handler function 必须符合’ sysrq.h ‘中的原型：void (* const handler)(int)。

创建sysrq_key_op后，就可以调用 register_sysrq_key(int key, const struct sysrq_key_op *op_p) 登记 sysrq 操作。
在模块卸载时，调用 unregister_sysrq_key(int key, const struct sysrq_key_op *op_p) 注销 sysrq 操作。

当然，永远不要在表中留下无效的指针。也就是说，当调用register_sysrq_key()的模块退出时，它必须调用unregister_sysrq_key()来清理它使用的sysrq键表项。空指针在表中总是安全的。

After the sysrq_key_op is created, you can call the kernel function register_sysrq_key(int key, const struct sysrq_key_op *op_p); this will register the operation pointed to by op_p at table key ‘key’, if that slot in the table is blank. At module unload time, you must call the function unregister_sysrq_key(int key, const struct sysrq_key_op *op_p), which will remove the key op pointed to by ‘op_p’ from the key ‘key’, if and only if it is currently registered in that slot. This is in case the slot has been overwritten since you registered it.

参考资料

极客时间：Linux内核技术实战

https://static.lwn.net/kerneldoc/admin-guide/sysrq.html
https://elixir.bootlin.com/linux/v6.0/source/drivers/tty/sysrq.c

Linux 调试之SysRq
Linux 调试之SysRq
Linux 调试之SysRq

今天的文章Linux 调试之SysRq分享到此就结束了，感谢您的阅读。

版权声明：本文内容由互联网用户自发贡献，该文观点仅代表作者本人。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容，请发送邮件至举报，一经查实，本站将立刻删除。
如需转载请保留出处：https://bianchenghao.cn/74101.html