1 Documentation for /proc/sys/kernel/* kernel version 2.2.10
2 (c) 1998, 1999, Rik van Riel <riel@nl.linux.org>
3 (c) 2009, Shen Feng<shen@cn.fujitsu.com>
5 For general info and legal blurb, please look in README.
7 ==============================================================
9 This file contains documentation for the sysctl files in
10 /proc/sys/kernel/ and is valid for Linux kernel version 2.2.
12 The files in this directory can be used to tune and monitor
13 miscellaneous and general things in the operation of the Linux
14 kernel. Since some of the files _can_ be used to screw up your
15 system, it is advisable to read both documentation and source
16 before actually making adjustments.
18 Currently, these files might (depending on your configuration)
19 show up in /proc/sys/kernel:
30 - java-appletviewer [ binfmt_java, obsolete ]
31 - java-interpreter [ binfmt_java, obsolete ]
32 - kstack_depth_to_print [ X86 only ]
34 - modprobe ==> Documentation/debugging-modules.txt
45 - powersave-nap [ PPC only ]
46 - panic_on_unrecovered_nmi
49 - real-root-dev ==> Documentation/initrd.txt
50 - reboot-cmd [ SPARC only ]
54 - sg-big-buff [ generic SCSI device (sg) ]
58 - stop-a [ SPARC only ]
59 - sysrq ==> Documentation/sysrq.txt
65 ==============================================================
71 See Doc*/kernel/power/video.txt, it allows mode of video boot to be
74 ==============================================================
78 highwater lowwater frequency
80 If BSD-style process accounting is enabled these values control
81 its behaviour. If free space on filesystem where the log lives
82 goes below <lowwater>% accounting suspends. If free space gets
83 above <highwater>% accounting resumes. <Frequency> determines
84 how often do we check the amount of free space (value is in
87 That is, suspend accounting if there left <= 2% free; resume it
88 if we got >=4%; consider information about amount of free space
91 ==============================================================
95 core_pattern is used to specify a core dumpfile pattern name.
96 . max length 128 characters; default value is "core"
97 . core_pattern is used as a pattern template for the output filename;
98 certain string patterns (beginning with '%') are substituted with
100 . backward compatibility with core_uses_pid:
101 If core_pattern does not include "%p" (default does not)
102 and core_uses_pid is set, then .PID will be appended to
104 . corename format specifiers:
105 %<NUL> '%' is dropped
113 %e executable filename
114 %<OTHER> both are dropped
115 . If the first character of the pattern is a '|', the kernel will treat
116 the rest of the pattern as a command to run. The core dump will be
117 written to the standard input of that program instead of to a file.
119 ==============================================================
123 The default coredump filename is "core". By setting
124 core_uses_pid to 1, the coredump filename becomes core.PID.
125 If core_pattern does not include "%p" (default does not)
126 and core_uses_pid is set, then .PID will be appended to
129 ==============================================================
133 When the value in this file is 0, ctrl-alt-del is trapped and
134 sent to the init(1) program to handle a graceful restart.
135 When, however, the value is > 0, Linux's reaction to a Vulcan
136 Nerve Pinch (tm) will be an immediate reboot, without even
137 syncing its dirty buffers.
139 Note: when a program (like dosemu) has the keyboard in 'raw'
140 mode, the ctrl-alt-del is intercepted by the program before it
141 ever reaches the kernel tty layer, and it's up to the program
142 to decide what to do with it.
144 ==============================================================
146 domainname & hostname:
148 These files can be used to set the NIS/YP domainname and the
149 hostname of your box in exactly the same way as the commands
150 domainname and hostname, i.e.:
151 # echo "darkstar" > /proc/sys/kernel/hostname
152 # echo "mydomain" > /proc/sys/kernel/domainname
153 has the same effect as
154 # hostname "darkstar"
155 # domainname "mydomain"
157 Note, however, that the classic darkstar.frop.org has the
158 hostname "darkstar" and DNS (Internet Domain Name Server)
159 domainname "frop.org", not to be confused with the NIS (Network
160 Information Service) or YP (Yellow Pages) domainname. These two
161 domain names are in general different. For a detailed discussion
162 see the hostname(1) man page.
164 ==============================================================
168 Path for the hotplug policy agent.
169 Default value is "/sbin/hotplug".
171 ==============================================================
175 This flag controls the L2 cache of G3 processor boards. If
176 0, the cache is disabled. Enabled if nonzero.
178 ==============================================================
180 kstack_depth_to_print: (X86 only)
182 Controls the number of words to print when dumping the raw
185 ==============================================================
187 osrelease, ostype & version:
194 #5 Wed Feb 25 21:49:24 MET 1998
196 The files osrelease and ostype should be clear enough. Version
197 needs a little more clarification however. The '#5' means that
198 this is the fifth kernel built from this source base and the
199 date behind it indicates the time the kernel was built.
200 The only way to tune these values is to rebuild the kernel :-)
202 ==============================================================
204 overflowgid & overflowuid:
206 if your architecture did not always support 32-bit UIDs (i.e. arm, i386,
207 m68k, sh, and sparc32), a fixed UID and GID will be returned to
208 applications that use the old 16-bit UID/GID system calls, if the actual
209 UID or GID would exceed 65535.
211 These sysctls allow you to change the value of the fixed UID and GID.
212 The default is 65534.
214 ==============================================================
218 The value in this file represents the number of seconds the
219 kernel waits before rebooting on a panic. When you use the
220 software watchdog, the recommended setting is 60.
222 ==============================================================
226 Controls the kernel's behaviour when an oops or BUG is encountered.
228 0: try to continue operation
230 1: panic immediately. If the `panic' sysctl is also non-zero then the
231 machine will be rebooted.
233 ==============================================================
237 PID allocation wrap value. When the kernel's next PID value
238 reaches this value, it wraps back to a minimum PID value.
239 PIDs of value pid_max or larger are not allocated.
241 ==============================================================
243 powersave-nap: (PPC only)
245 If set, Linux-PPC will use the 'nap' mode of powersaving,
246 otherwise the 'doze' mode will be used.
248 ==============================================================
252 The four values in printk denote: console_loglevel,
253 default_message_loglevel, minimum_console_loglevel and
254 default_console_loglevel respectively.
256 These values influence printk() behavior when printing or
257 logging error messages. See 'man 2 syslog' for more info on
258 the different loglevels.
260 - console_loglevel: messages with a higher priority than
261 this will be printed to the console
262 - default_message_level: messages without an explicit priority
263 will be printed with this priority
264 - minimum_console_loglevel: minimum (highest) value to which
265 console_loglevel can be set
266 - default_console_loglevel: default value for console_loglevel
268 ==============================================================
272 Some warning messages are rate limited. printk_ratelimit specifies
273 the minimum length of time between these messages (in jiffies), by
274 default we allow one every 5 seconds.
276 A value of 0 will disable rate limiting.
278 ==============================================================
280 printk_ratelimit_burst:
282 While long term we enforce one message per printk_ratelimit
283 seconds, we do allow a burst of messages to pass through.
284 printk_ratelimit_burst specifies the number of messages we can
285 send before ratelimiting kicks in.
287 ==============================================================
291 This option can be used to select the type of process address
292 space randomization that is used in the system, for architectures
293 that support this feature.
295 0 - Turn the process address space randomization off by default.
297 1 - Make the addresses of mmap base, stack and VDSO page randomized.
298 This, among other things, implies that shared libraries will be
299 loaded to random addresses. Also for PIE-linked binaries, the location
300 of code start is randomized.
302 With heap randomization, the situation is a little bit more
304 There a few legacy applications out there (such as some ancient
305 versions of libc.so.5 from 1996) that assume that brk area starts
306 just after the end of the code+bss. These applications break when
307 start of the brk area is randomized. There are however no known
308 non-legacy applications that would be broken this way, so for most
309 systems it is safe to choose full randomization. However there is
310 a CONFIG_COMPAT_BRK option for systems with ancient and/or broken
311 binaries, that makes heap non-randomized, but keeps all other
312 parts of process address space randomized if randomize_va_space
315 ==============================================================
317 reboot-cmd: (Sparc only)
319 ??? This seems to be a way to give an argument to the Sparc
320 ROM/Flash boot loader. Maybe to tell it what to do after
323 ==============================================================
325 rtsig-max & rtsig-nr:
327 The file rtsig-max can be used to tune the maximum number
328 of POSIX realtime (queued) signals that can be outstanding
331 rtsig-nr shows the number of RT signals currently queued.
333 ==============================================================
337 This file shows the size of the generic SCSI (sg) buffer.
338 You can't tune it just yet, but you could change it on
339 compile time by editing include/scsi/sg.h and changing
340 the value of SG_BIG_BUFF.
342 There shouldn't be any reason to change this value. If
343 you can come up with one, you probably know what you
346 ==============================================================
350 This value can be used to query and set the run time limit
351 on the maximum shared memory segment size that can be created.
352 Shared memory segments up to 1Gb are now supported in the
353 kernel. This value defaults to SHMMAX.
355 ==============================================================
359 This value can be used to lower the softlockup tolerance threshold. The
360 default threshold is 60 seconds. If a cpu is locked up for 60 seconds,
361 the kernel complains. Valid values are 1-60 seconds. Setting this
362 tunable to zero will disable the softlockup detection altogether.
364 ==============================================================
368 Non-zero if the kernel has been tainted. Numeric values, which
369 can be ORed together:
371 1 - A module with a non-GPL license has been loaded, this
372 includes modules with no license.
373 Set by modutils >= 2.4.9 and module-init-tools.
374 2 - A module was force loaded by insmod -f.
375 Set by modutils >= 2.4.9 and module-init-tools.
376 4 - Unsafe SMP processors: SMP with CPUs not designed for SMP.
377 8 - A module was forcibly unloaded from the system by rmmod -f.
378 16 - A hardware machine check error occurred on the system.
379 32 - A bad page was discovered on the system.
380 64 - The user has asked that the system be marked "tainted". This
381 could be because they are running software that directly modifies
382 the hardware, or for other reasons.
383 128 - The system has died.
384 256 - The ACPI DSDT has been overridden with one supplied by the user
385 instead of using the one provided by the hardware.
386 512 - A kernel warning has occurred.
387 1024 - A module from drivers/staging was loaded.
389 ==============================================================
393 Enables/Disables automatic recomputing of msgmni upon memory add/remove or
394 upon ipc namespace creation/removal (see the msgmni description above).
395 Echoing "1" into this file enables msgmni automatic recomputing.
396 Echoing "0" turns it off.
397 auto_msgmni default value is 1.
399 ==============================================================
403 Enables/Disables the NMI watchdog on x86 systems. When the value is non-zero
404 the NMI watchdog is enabled and will continuously test all online cpus to
405 determine whether or not they are still functioning properly. Currently,
406 passing "nmi_watchdog=" parameter at boot time is required for this function
409 If LAPIC NMI watchdog method is in use (nmi_watchdog=2 kernel parameter), the
410 NMI watchdog shares registers with oprofile. By disabling the NMI watchdog,
411 oprofile may have more registers to utilize.
413 ==============================================================
417 The value in this file affects behavior of handling NMI. When the value is
418 non-zero, unknown NMI is trapped and then panic occurs. At that time, kernel
419 debugging information is displayed on console.
421 NMI switch that most IA32 servers have fires unknown NMI up, for example.
422 If a system hangs up, try pressing the NMI switch.
424 ==============================================================
426 panic_on_unrecovered_nmi:
428 The default Linux behaviour on an NMI of either memory or unknown is to continue
429 operation. For many environments such as scientific computing it is preferable
430 that the box is taken out and the error dealt with than an uncorrected
431 parity/ECC error get propogated.
433 A small number of systems do generate NMI's for bizarre random reasons such as
434 power management so the default is off. That sysctl works like the existing
435 panic controls already in that directory.