From Fedora Project Wiki

< QA

No edit summary
(→‎How do I enable the magic SysRq key?: /etc/sysctl.conf is deprecated, make sure override file is alphabetically last)
(14 intermediate revisions by 4 users not shown)
Line 10: Line 10:


You need to say "yes" to 'Magic SysRq key (CONFIG_MAGIC_SYSRQ)' when
You need to say "yes" to 'Magic SysRq key (CONFIG_MAGIC_SYSRQ)' when
configuring the kernel. When running a kernel with SysRq compiled in,
configuring the kernel. Stock Fedora and RHEL kernels do have this functionality enabled at compile-time, but the distributions disable it at boot time, by default, using sysctl.conf.
/proc/sys/kernel/sysrq controls the functions allowed to be invoked via
 
the SysRq key. By default the file contains 1 which means that every
To re-enable it at boot time create config file in the <code>sysctl.d</code> directory (e.g. <code>/etc/sysctl.d/90-sysrq.conf</code>) with this line:
possible SysRq request is allowed (in older versions SysRq was disabled
 
by default, and you were required to specifically enable it at run-time
<pre>kernel.sysrq = 1</pre>
but this is not the case any more). Here is the list of possible values
 
in /proc/sys/kernel/sysrq:
When running a kernel with SysRq compiled in, /proc/sys/kernel/sysrq controls the functions allowed to be invoked via the SysRq key. Here is the list of possible values in /proc/sys/kernel/sysrq:
 
* 0 - disable sysrq completely
* 0 - disable sysrq completely
* 1 - enable all functions of sysrq
* 1 - enable all functions of sysrq
Line 29: Line 30:
**256 - allow nicing of all RT tasks
**256 - allow nicing of all RT tasks


You can set the value in the file by the following command:
You can set the value in the file by the following command.
echo "number" >/proc/sys/kernel/sysrq


{{admon/warning|NOTE|The value of /proc/sys/kernel/sysrq influences only the invocation via a keyboard. Invocation of any operation via /proc/sysrq-trigger is always allowed (by a user with admin privileges).}}
<pre>echo "number" >/proc/sys/kernel/sysrq </pre>
 
So to enable it would be.
 
<pre>echo "1" > /proc/sys/kernel/sysrq</pre>
 
Or also can enable it by doing.
 
<pre>sysctl -w kernel.sysrq=1</pre>
 
{{admon/note|The value of /proc/sys/kernel/sysrq influences only the invocation via a keyboard. Invocation of any operation via /proc/sysrq-trigger is always allowed (by a user with admin privileges).}}


== How do I use the magic SysRq key? ==
== How do I use the magic SysRq key? ==
Line 48: Line 58:
You send a BREAK, then within 5 seconds a command key.  
You send a BREAK, then within 5 seconds a command key.  


{{admon/warning|NOTE| Sending BREAK twice is interpreted as a normal BREAK. }}
{{admon/note| Sending BREAK twice is interpreted as a normal BREAK. }}


'''PowerPC'''
'''PowerPC'''
Line 60: Line 70:
<pre>echo t > /proc/sysrq-trigger </pre>
<pre>echo t > /proc/sysrq-trigger </pre>


{{admon/warning|NOTE|Some keyboards may not have a key labeled 'SysRq'. The 'SysRq' key is
{{admon/note|Some keyboards may not have a key labeled 'SysRq'. The 'SysRq' key is
also known as the 'Print Screen' key. Also some keyboards cannot handle so many keys being pressed at the same time, so you might have better luck with "press Alt", "press SysRq", "release SysRq", "press <command key>", release everything.}}
also known as the 'Print Screen' key. Also some keyboards cannot handle so many keys being pressed at the same time, so you might have better luck with "press Alt", "press SysRq", "release SysRq", "press <command key>", release everything.}}


== What are the 'command' keys? ==
== What are the 'command' keys? ==


'b'     - Will immediately reboot the system without syncing or unmounting
*'b' - Will immediately reboot the system without syncing or unmounting your disks.
          your disks.
*'c' - Will perform a kexec reboot in order to take a crashdump.
*'d' - Shows all locks that are held.
*'e' - Send a SIGTERM to all processes, except for init.
*'f' - Will call oom_kill to kill a memory hog process.
*'g' - Used by kgdb on ppc and sh platforms.
*'h' - Will display help (any key that is not listed here will bring forth help )
*'i' - Send a SIGKILL to all processes, except for init.
*'k' - Secure Access Key (SAK) Kills all programs on the current virtual terminal.
{{admon/note|See important comments below in SAK section.}}
*'l' - Shows a stack backtrace for all active CPUs.
*'m' - Will dump current memory info to your console.
*'n' - Used to make RT tasks nice-able
*'o' - Will shut your system off (if configured and supported).
*'p' - Will dump the current registers and flags to your console.
*'q' - Will dump a list of all running timers.
*'r' - Turns off keyboard raw mode and sets it to XLATE.
*'s' - Will attempt to sync all mounted filesystems.
*'t' - Will dump a list of current tasks and their information to your console.
*'u' - Will attempt to remount all mounted filesystems read-only.
*'v' - Dumps Voyager SMP processor info to your console.
*'w' - Dumps tasks that are in uninterruptable (blocked) state.
*'x' - Used by xmon interface on ppc/powerpc platforms.
*'0'-'9' - Sets the console log level, controlling which kernel messages will be printed to your console. ('0', for example would make it so that only emergency messages like PANICs or OOPSes would make it to your console.)


'c'    - Will perform a kexec reboot in order to take a crashdump.
== Okay, so what can I use them for? ==
 
'd'    - Shows all locks that are held.
 
'e'    - Send a SIGTERM to all processes, except for init.
 
'f'    - Will call oom_kill to kill a memory hog process.
 
'g'    - Used by kgdb on ppc and sh platforms.
 
'h'    - Will display help (actually any other key than those listed
          here will display help. but 'h' is easy to remember :-)
 
'i'    - Send a SIGKILL to all processes, except for init.


'k'     - Secure Access Key (SAK) Kills all programs on the current virtual
Un'r'aw is very handy when your X server or a svgalib program crashes.
          console. NOTE: See important comments below in SAK section.


'l'     - Shows a stack backtrace for all active CPUs.
Sa'k' (Secure Access Key) is useful when you want to be sure there is no trojan program running at console which could grab your password when you would try to login. It will kill all programs on given console, thus letting you make sure that the login prompt you see is actually the one from init, not some trojan program.
Others find it useful as (System Attention Key) which is useful when you want to exit a program that will not let you switch consoles. (For example, X or a svgalib program.)


'm'    - Will dump current memory info to your console.
{{admon/note|In its true form it is not a true SAK like the one in a c2 compliant system, and it should not be mistaken as such. }}


'n'     - Used to make RT tasks nice-able
Re'b'oot is good when you're unable to shut down.


'o'     - Will shut your system off (if configured and supported).
{{admon/note| It's general considered a good practice to 'u'mount first }}


'p'     - Will dump the current registers and flags to your console.
Crashdump can be used to manually trigger a 'c'rashdump when the system is hung.


'q'    - Will dump a list of all running timers.
{{admon/note| The kernel needs to have been built with CONFIG_KEXEC enabled! }}
 
'r'    - Turns off keyboard raw mode and sets it to XLATE.
 
's'    - Will attempt to sync all mounted filesystems.
 
't'    - Will dump a list of current tasks and their information to your
          console.
 
'u'    - Will attempt to remount all mounted filesystems read-only.
 
'v'    - Dumps Voyager SMP processor info to your console.
 
'w'    - Dumps tasks that are in uninterruptable (blocked) state.
 
'x'    - Used by xmon interface on ppc/powerpc platforms.
 
'0'-'9' - Sets the console log level, controlling which kernel messages
          will be printed to your console. ('0', for example would make
          it so that only emergency messages like PANICs or OOPSes would
          make it to your console.)
 
== Okay, so what can I use them for? ==


un'R'aw is very handy when your X server or a svgalib program crashes.
Sync is great when your system is locked up, it allows you to 's'ync your disks and will certainly lessen the chance of data loss and fscking.


sa'K' (Secure Access Key) is useful when you want to be sure there is no
{{admon/warning| The sync hasn't taken place until you see the "OK" and "Done" appear on the screen. (If the kernel is really in strife, you may not ever get the OK or Done message. }}
trojan program running at console which could grab your password
when you would try to login. It will kill all programs on given console,
thus letting you make sure that the login prompt you see is actually
the one from init, not some trojan program.
IMPORTANT: In its true form it is not a true SAK like the one in a :IMPORTANT
IMPORTANT: c2 compliant system, and it should not be mistaken as  :IMPORTANT
IMPORTANT: such.                                                  :IMPORTANT
      It seems others find it useful as (System Attention Key) which is
useful when you want to exit a program that will not let you switch consoles.
(For example, X or a svgalib program.)


re'B'oot is good when you're unable to shut down. But you should also 'S'ync
'U'mount is basically useful in the same ways as 'S'ync.
and 'U'mount first.


'C'rashdump can be used to manually trigger a crashdump when the system is hung.
The loglevels '0'-'9' are useful when your console is being flooded with kernel messages you do not want to see. Selecting '0' will prevent all but the most urgent kernel messages from reaching your console.
The kernel needs to have been built with CONFIG_KEXEC enabled.


'S'ync is great when your system is locked up, it allows you to sync your
{{admon/note| They will still be logged if syslogd/klogd are alive }}
disks and will certainly lessen the chance of data loss and fscking. Note
that the sync hasn't taken place until you see the "OK" and "Done" appear
on the screen. (If the kernel is really in strife, you may not ever get the
OK or Done message...)


'U'mount is basically useful in the same ways as 'S'ync. I generally 'S'ync,
T'e'rm and k'i'll are useful if you have some sort of runaway process you are unable to kill any other way, especially if it's spawning other processes.
'U'mount, then re'B'oot when my system locks. It's saved me many a fsck.
Again, the unmount (remount read-only) hasn't taken place until you see the
"OK" and "Done" message appear on the screen.


The loglevels '0'-'9' are useful when your console is being flooded with
{{admon/note| When experiencing bad kernel panic do <Alt>+<Sysrg>+e then <Alt>+<Sysrg>+u then <Alt>+<Sysrg>+i and finally <Alt>+<Sysrg>+b }}
kernel messages you do not want to see. Selecting '0' will prevent all but
the most urgent kernel messages from reaching your console. (They will
still be logged if syslogd/klogd are alive, though.)


t'E'rm and k'I'll are useful if you have some sort of runaway process you
== Troubleshooting ==
are unable to kill any other way, especially if it's spawning other
processes.


== Troubleshoot ==
===Hanging before initscripts get run===


*  Sometimes SysRq seems to get 'stuck' after using it, what can I do?
If the machine is hanging before the initscripts get to run, boot with sysrq_always_enabled=1.


That happens to me, also. I've found that tapping shift, alt, and control
=== Sometimes SysRq seems to get 'stuck' after using it, what can I do? ===
on both sides of the keyboard, and hitting an invalid sysrq sequence again
will fix the problem. (i.e., something like alt-sysrq-z). Switching to another
virtual console (ALT+Fn) and then back again should also help.


* I hit SysRq, but nothing seems to happen, what's wrong?
Tapping shift, alt, and control on both sides of the keyboard, and hitting an invalid sysrq sequence again
will fix the problem. (i.e., something like alt-sysrq-z).


There are some keyboards that send different scancodes for SysRq than the
Switching to another virtual console (ALT+Fn) and then back again should also help.
pre-defined 0x54. So if SysRq doesn't work out of the box for a certain
keyboard, run 'showkey -s' to find out the proper scancode sequence. Then
use 'setkeycodes <sequence> 84' to define this sequence to the usual SysRq
code (84 is decimal for 0x54). It's probably best to put this command in a
boot script. Oh, and by the way, you exit 'showkey' by not typing anything
for ten seconds.


I want to add SysRQ key events to a module, how does it work?
=== I hit SysRq, but nothing seems to happen, what's wrong? ===


n order to register a basic function with the table, you must first include
There are some keyboards that send different scancodes for SysRq than the pre-defined 0x54. So if SysRq doesn't work out of the box for a certain keyboard, run 'showkey -s' to find out the proper scancode sequence. Then use 'setkeycodes <sequence> 84' to define this sequence to the usual SysRq code (84 is decimal for 0x54). It's probably best to put this command in a boot script.  
the header 'include/linux/sysrq.h', this will define everything else you need.
Next, you must create a sysrq_key_op struct, and populate it with A) the key
handler function you will use, B) a help_msg string, that will print when SysRQ
prints help, and C) an action_msg string, that will print right before your
handler is called. Your handler must conform to the prototype in 'sysrq.h'.


After the sysrq_key_op is created, you can call the kernel function
{{admon/warning| You exit 'showkey' by not typing anything for ten seconds.}}
register_sysrq_key(int key, 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, 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.


The Magic SysRQ system works by registering key operations against a key op
=== I want to add SysRQ key events to a module, how does it work? ===
lookup table, which is defined in 'drivers/char/sysrq.c'. This key table has
a number of operations registered into it at compile time, but is mutable,
and 2 functions are exported for interface to it:
        register_sysrq_key and unregister_sysrq_key.
Of course, never ever leave an invalid pointer in the table. I.e., when
your module that called register_sysrq_key() exits, it must call
unregister_sysrq_key() to clean up the sysrq key table entry that it used.
Null pointers in the table are always safe. :)


If for some reason you feel the need to call the handle_sysrq function from
In order to register a basic function with the table, you must first include the header 'include/linux/sysrq.h', this will define everything else you need. Next, you must create a sysrq_key_op struct, and populate it with...  
within a function called by handle_sysrq, you must be aware that you are in
a lock (you are also in an interrupt handler, which means don't sleep!), so
you must call __handle_sysrq_nolock instead.


== Credits ==
*The key handler function you will use.
*A help_msg string, that will print when SysRQ prints help
*An action_msg string, that will print right before your handler is called. Your handler must conform to the prototype in 'sysrq.h'.


Written by Mydraal
After the sysrq_key_op is created, you can call the kernel function register_sysrq_key(int key, 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, 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.


Updated by Adam Sulmicki
The Magic SysRQ system works by registering key operations against a key op lookup table, which is defined in 'drivers/char/sysrq.c'. This key table has a number of operations registered into it at compile time, but is mutable, and 2 functions are exported for interface to it the register_sysrq_key and unregister_sysrq_key. Of course, never ever leave an invalid pointer in the table. I.e., when your module that called register_sysrq_key() exits, it must call unregister_sysrq_key() to clean up the sysrq key table entry that it used.


Updated by Jeremy M. Dolan
{{admon/note| Null pointers in the table are always safe. }}


Added to by Crutcher Dunnavant
If for some reason you feel the need to call the handle_sysrq function from within a function called by handle_sysrq, you must be aware that you are in a lock (you are also in an interrupt handler, which means don't sleep!), so you must call __handle_sysrq_nolock instead.

Revision as of 00:08, 27 February 2015

QA.png


Sysrq

What is the magic SysRq key?

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.

How do I enable the magic SysRq key?

You need to say "yes" to 'Magic SysRq key (CONFIG_MAGIC_SYSRQ)' when configuring the kernel. Stock Fedora and RHEL kernels do have this functionality enabled at compile-time, but the distributions disable it at boot time, by default, using sysctl.conf.

To re-enable it at boot time create config file in the sysctl.d directory (e.g. /etc/sysctl.d/90-sysrq.conf) with this line:

kernel.sysrq = 1

When running a kernel with SysRq compiled in, /proc/sys/kernel/sysrq controls the functions allowed to be invoked via the SysRq key. Here is the list of possible values in /proc/sys/kernel/sysrq:

  • 0 - disable sysrq completely
  • 1 - enable all functions of sysrq
  • >1 - bitmask of allowed sysrq functions (see below for detailed function description):
    • 2 - enable control of console logging level
    • 4 - enable control of keyboard (SAK, unraw)
    • 8 - enable debugging dumps of processes etc.
    • 16 - enable sync command
    • 32 - enable remount read-only
    • 64 - enable signalling of processes (term, kill, oom-kill)
    • 128 - allow reboot/poweroff
    • 256 - allow nicing of all RT tasks

You can set the value in the file by the following command.

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

So to enable it would be.

echo "1" > /proc/sys/kernel/sysrq

Or also can enable it by doing.

sysctl -w kernel.sysrq=1
Note.png
The value of /proc/sys/kernel/sysrq influences only the invocation via a keyboard. Invocation of any operation via /proc/sysrq-trigger is always allowed (by a user with admin privileges).

How do I use the magic SysRq key?

x86

You press the key combo '<Alt><SysRq><command key>'.

SPARC

You press '<Alt><Stop><command key>'

Serial Console (PC style standard serial ports only )

You send a BREAK, then within 5 seconds a command key.

Note.png
Sending BREAK twice is interpreted as a normal BREAK.

PowerPC

Press '<Alt><Print Screen>(or <F13>)<command key>

All Architectures

Write a character to /proc/sysrq-trigger:

echo t > /proc/sysrq-trigger 
Note.png
Some keyboards may not have a key labeled 'SysRq'. The 'SysRq' key is also known as the 'Print Screen' key. Also some keyboards cannot handle so many keys being pressed at the same time, so you might have better luck with "press Alt", "press SysRq", "release SysRq", "press <command key>", release everything.

What are the 'command' keys?

  • 'b' - Will immediately reboot the system without syncing or unmounting your disks.
  • 'c' - Will perform a kexec reboot in order to take a crashdump.
  • 'd' - Shows all locks that are held.
  • 'e' - Send a SIGTERM to all processes, except for init.
  • 'f' - Will call oom_kill to kill a memory hog process.
  • 'g' - Used by kgdb on ppc and sh platforms.
  • 'h' - Will display help (any key that is not listed here will bring forth help )
  • 'i' - Send a SIGKILL to all processes, except for init.
  • 'k' - Secure Access Key (SAK) Kills all programs on the current virtual terminal.
Note.png
See important comments below in SAK section.
  • 'l' - Shows a stack backtrace for all active CPUs.
  • 'm' - Will dump current memory info to your console.
  • 'n' - Used to make RT tasks nice-able
  • 'o' - Will shut your system off (if configured and supported).
  • 'p' - Will dump the current registers and flags to your console.
  • 'q' - Will dump a list of all running timers.
  • 'r' - Turns off keyboard raw mode and sets it to XLATE.
  • 's' - Will attempt to sync all mounted filesystems.
  • 't' - Will dump a list of current tasks and their information to your console.
  • 'u' - Will attempt to remount all mounted filesystems read-only.
  • 'v' - Dumps Voyager SMP processor info to your console.
  • 'w' - Dumps tasks that are in uninterruptable (blocked) state.
  • 'x' - Used by xmon interface on ppc/powerpc platforms.
  • '0'-'9' - Sets the console log level, controlling which kernel messages will be printed to your console. ('0', for example would make it so that only emergency messages like PANICs or OOPSes would make it to your console.)

Okay, so what can I use them for?

Un'r'aw is very handy when your X server or a svgalib program crashes.

Sa'k' (Secure Access Key) is useful when you want to be sure there is no trojan program running at console which could grab your password when you would try to login. It will kill all programs on given console, thus letting you make sure that the login prompt you see is actually the one from init, not some trojan program. Others find it useful as (System Attention Key) which is useful when you want to exit a program that will not let you switch consoles. (For example, X or a svgalib program.)

Note.png
In its true form it is not a true SAK like the one in a c2 compliant system, and it should not be mistaken as such.

Re'b'oot is good when you're unable to shut down.

Note.png
It's general considered a good practice to 'u'mount first

Crashdump can be used to manually trigger a 'c'rashdump when the system is hung.

Note.png
The kernel needs to have been built with CONFIG_KEXEC enabled!

Sync is great when your system is locked up, it allows you to 's'ync your disks and will certainly lessen the chance of data loss and fscking.

Warning.png
The sync hasn't taken place until you see the "OK" and "Done" appear on the screen. (If the kernel is really in strife, you may not ever get the OK or Done message.

'U'mount is basically useful in the same ways as 'S'ync.

The loglevels '0'-'9' are useful when your console is being flooded with kernel messages you do not want to see. Selecting '0' will prevent all but the most urgent kernel messages from reaching your console.

Note.png
They will still be logged if syslogd/klogd are alive

T'e'rm and k'i'll are useful if you have some sort of runaway process you are unable to kill any other way, especially if it's spawning other processes.

Note.png
When experiencing bad kernel panic do <Alt>+<Sysrg>+e then <Alt>+<Sysrg>+u then <Alt>+<Sysrg>+i and finally <Alt>+<Sysrg>+b

Troubleshooting

Hanging before initscripts get run

If the machine is hanging before the initscripts get to run, boot with sysrq_always_enabled=1.

Sometimes SysRq seems to get 'stuck' after using it, what can I do?

Tapping shift, alt, and control on both sides of the keyboard, and hitting an invalid sysrq sequence again will fix the problem. (i.e., something like alt-sysrq-z).

Switching to another virtual console (ALT+Fn) and then back again should also help.

I hit SysRq, but nothing seems to happen, what's wrong?

There are some keyboards that send different scancodes for SysRq than the pre-defined 0x54. So if SysRq doesn't work out of the box for a certain keyboard, run 'showkey -s' to find out the proper scancode sequence. Then use 'setkeycodes <sequence> 84' to define this sequence to the usual SysRq code (84 is decimal for 0x54). It's probably best to put this command in a boot script.

Warning.png
You exit 'showkey' by not typing anything for ten seconds.

I want to add SysRQ key events to a module, how does it work?

In order to register a basic function with the table, you must first include the header 'include/linux/sysrq.h', this will define everything else you need. Next, you must create a sysrq_key_op struct, and populate it with...

  • The key handler function you will use.
  • A help_msg string, that will print when SysRQ prints help
  • An action_msg string, that will print right before your handler is called. Your handler must conform to the prototype in 'sysrq.h'.

After the sysrq_key_op is created, you can call the kernel function register_sysrq_key(int key, 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, 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.

The Magic SysRQ system works by registering key operations against a key op lookup table, which is defined in 'drivers/char/sysrq.c'. This key table has a number of operations registered into it at compile time, but is mutable, and 2 functions are exported for interface to it the register_sysrq_key and unregister_sysrq_key. Of course, never ever leave an invalid pointer in the table. I.e., when your module that called register_sysrq_key() exits, it must call unregister_sysrq_key() to clean up the sysrq key table entry that it used.

Note.png
Null pointers in the table are always safe.

If for some reason you feel the need to call the handle_sysrq function from within a function called by handle_sysrq, you must be aware that you are in a lock (you are also in an interrupt handler, which means don't sleep!), so you must call __handle_sysrq_nolock instead.