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:
24 - bootloader_type [ X86 only ]
25 - bootloader_version [ X86 only ]
26 - callhome [ S390 only ]
37 - kstack_depth_to_print [ X86 only ]
39 - modprobe ==> Documentation/debugging-modules.txt
41 - msg_next_id [ sysv ipc ]
52 - panic_on_unrecovered_nmi
53 - panic_on_stackoverflow
55 - powersave-nap [ PPC only ]
59 - printk_ratelimit_burst
61 - real-root-dev ==> Documentation/initrd.txt
62 - reboot-cmd [ SPARC only ]
66 - sem_next_id [ sysv ipc ]
67 - sg-big-buff [ generic SCSI device (sg) ]
68 - shm_next_id [ sysv ipc ]
74 - stop-a [ SPARC only ]
75 - sysrq ==> Documentation/sysrq.txt
81 ==============================================================
85 highwater lowwater frequency
87 If BSD-style process accounting is enabled these values control
88 its behaviour. If free space on filesystem where the log lives
89 goes below <lowwater>% accounting suspends. If free space gets
90 above <highwater>% accounting resumes. <Frequency> determines
91 how often do we check the amount of free space (value is in
94 That is, suspend accounting if there left <= 2% free; resume it
95 if we got >=4%; consider information about amount of free space
98 ==============================================================
104 See Doc*/kernel/power/video.txt, it allows mode of video boot to be
107 ==============================================================
111 Enables/Disables automatic recomputing of msgmni upon memory add/remove
112 or upon ipc namespace creation/removal (see the msgmni description
113 above). Echoing "1" into this file enables msgmni automatic recomputing.
114 Echoing "0" turns it off. auto_msgmni default value is 1.
117 ==============================================================
121 x86 bootloader identification
123 This gives the bootloader type number as indicated by the bootloader,
124 shifted left by 4, and OR'd with the low four bits of the bootloader
125 version. The reason for this encoding is that this used to match the
126 type_of_loader field in the kernel header; the encoding is kept for
127 backwards compatibility. That is, if the full bootloader type number
128 is 0x15 and the full version number is 0x234, this file will contain
129 the value 340 = 0x154.
131 See the type_of_loader and ext_loader_type fields in
132 Documentation/x86/boot.txt for additional information.
134 ==============================================================
138 x86 bootloader version
140 The complete bootloader version number. In the example above, this
141 file will contain the value 564 = 0x234.
143 See the type_of_loader and ext_loader_ver fields in
144 Documentation/x86/boot.txt for additional information.
146 ==============================================================
150 Controls the kernel's callhome behavior in case of a kernel panic.
152 The s390 hardware allows an operating system to send a notification
153 to a service organization (callhome) in case of an operating system panic.
155 When the value in this file is 0 (which is the default behavior)
156 nothing happens in case of a kernel panic. If this value is set to "1"
157 the complete kernel oops message is send to the IBM customer service
158 organization in case the mainframe the Linux operating system is running
159 on has a service contract with IBM.
161 ==============================================================
165 Highest valid capability of the running kernel. Exports
166 CAP_LAST_CAP from the kernel.
168 ==============================================================
172 core_pattern is used to specify a core dumpfile pattern name.
173 . max length 128 characters; default value is "core"
174 . core_pattern is used as a pattern template for the output filename;
175 certain string patterns (beginning with '%') are substituted with
177 . backward compatibility with core_uses_pid:
178 If core_pattern does not include "%p" (default does not)
179 and core_uses_pid is set, then .PID will be appended to
181 . corename format specifiers:
182 %<NUL> '%' is dropped
187 %d dump mode, matches PR_SET_DUMPABLE and
188 /proc/sys/fs/suid_dumpable
192 %e executable filename (may be shortened)
194 %<OTHER> both are dropped
195 . If the first character of the pattern is a '|', the kernel will treat
196 the rest of the pattern as a command to run. The core dump will be
197 written to the standard input of that program instead of to a file.
199 ==============================================================
203 This sysctl is only applicable when core_pattern is configured to pipe
204 core files to a user space helper (when the first character of
205 core_pattern is a '|', see above). When collecting cores via a pipe
206 to an application, it is occasionally useful for the collecting
207 application to gather data about the crashing process from its
208 /proc/pid directory. In order to do this safely, the kernel must wait
209 for the collecting process to exit, so as not to remove the crashing
210 processes proc files prematurely. This in turn creates the
211 possibility that a misbehaving userspace collecting process can block
212 the reaping of a crashed process simply by never exiting. This sysctl
213 defends against that. It defines how many concurrent crashing
214 processes may be piped to user space applications in parallel. If
215 this value is exceeded, then those crashing processes above that value
216 are noted via the kernel log and their cores are skipped. 0 is a
217 special value, indicating that unlimited processes may be captured in
218 parallel, but that no waiting will take place (i.e. the collecting
219 process is not guaranteed access to /proc/<crashing pid>/). This
222 ==============================================================
226 The default coredump filename is "core". By setting
227 core_uses_pid to 1, the coredump filename becomes core.PID.
228 If core_pattern does not include "%p" (default does not)
229 and core_uses_pid is set, then .PID will be appended to
232 ==============================================================
236 When the value in this file is 0, ctrl-alt-del is trapped and
237 sent to the init(1) program to handle a graceful restart.
238 When, however, the value is > 0, Linux's reaction to a Vulcan
239 Nerve Pinch (tm) will be an immediate reboot, without even
240 syncing its dirty buffers.
242 Note: when a program (like dosemu) has the keyboard in 'raw'
243 mode, the ctrl-alt-del is intercepted by the program before it
244 ever reaches the kernel tty layer, and it's up to the program
245 to decide what to do with it.
247 ==============================================================
251 This toggle indicates whether unprivileged users are prevented
252 from using dmesg(8) to view messages from the kernel's log buffer.
253 When dmesg_restrict is set to (0) there are no restrictions. When
254 dmesg_restrict is set set to (1), users must have CAP_SYSLOG to use
257 The kernel config option CONFIG_SECURITY_DMESG_RESTRICT sets the
258 default value of dmesg_restrict.
260 ==============================================================
262 domainname & hostname:
264 These files can be used to set the NIS/YP domainname and the
265 hostname of your box in exactly the same way as the commands
266 domainname and hostname, i.e.:
267 # echo "darkstar" > /proc/sys/kernel/hostname
268 # echo "mydomain" > /proc/sys/kernel/domainname
269 has the same effect as
270 # hostname "darkstar"
271 # domainname "mydomain"
273 Note, however, that the classic darkstar.frop.org has the
274 hostname "darkstar" and DNS (Internet Domain Name Server)
275 domainname "frop.org", not to be confused with the NIS (Network
276 Information Service) or YP (Yellow Pages) domainname. These two
277 domain names are in general different. For a detailed discussion
278 see the hostname(1) man page.
280 ==============================================================
284 Path for the hotplug policy agent.
285 Default value is "/sbin/hotplug".
287 ==============================================================
291 This toggle indicates whether restrictions are placed on
292 exposing kernel addresses via /proc and other interfaces.
294 When kptr_restrict is set to (0), the default, there are no restrictions.
296 When kptr_restrict is set to (1), kernel pointers printed using the %pK
297 format specifier will be replaced with 0's unless the user has CAP_SYSLOG
298 and effective user and group ids are equal to the real ids. This is
299 because %pK checks are done at read() time rather than open() time, so
300 if permissions are elevated between the open() and the read() (e.g via
301 a setuid binary) then %pK will not leak kernel pointers to unprivileged
302 users. Note, this is a temporary solution only. The correct long-term
303 solution is to do the permission checks at open() time. Consider removing
304 world read permissions from files that use %pK, and using dmesg_restrict
305 to protect against uses of %pK in dmesg(8) if leaking kernel pointer
306 values to unprivileged users is a concern.
308 When kptr_restrict is set to (2), kernel pointers printed using
309 %pK will be replaced with 0's regardless of privileges.
311 ==============================================================
313 kstack_depth_to_print: (X86 only)
315 Controls the number of words to print when dumping the raw
318 ==============================================================
322 This flag controls the L2 cache of G3 processor boards. If
323 0, the cache is disabled. Enabled if nonzero.
325 ==============================================================
329 A toggle value indicating if modules are allowed to be loaded
330 in an otherwise modular kernel. This toggle defaults to off
331 (0), but can be set true (1). Once true, modules can be
332 neither loaded nor unloaded, and the toggle cannot be set back
335 ==============================================================
337 msg_next_id, sem_next_id, and shm_next_id:
339 These three toggles allows to specify desired id for next allocated IPC
340 object: message, semaphore or shared memory respectively.
342 By default they are equal to -1, which means generic allocation logic.
343 Possible values to set are in range {0..INT_MAX}.
346 1) kernel doesn't guarantee, that new object will have desired id. So,
347 it's up to userspace, how to handle an object with "wrong" id.
348 2) Toggle with non-default value will be set back to -1 by kernel after
349 successful IPC object allocation.
351 ==============================================================
355 Enables/Disables the NMI watchdog on x86 systems. When the value is
356 non-zero the NMI watchdog is enabled and will continuously test all
357 online cpus to determine whether or not they are still functioning
358 properly. Currently, passing "nmi_watchdog=" parameter at boot time is
359 required for this function to work.
361 If LAPIC NMI watchdog method is in use (nmi_watchdog=2 kernel
362 parameter), the NMI watchdog shares registers with oprofile. By
363 disabling the NMI watchdog, oprofile may have more registers to
366 ==============================================================
368 osrelease, ostype & version:
375 #5 Wed Feb 25 21:49:24 MET 1998
377 The files osrelease and ostype should be clear enough. Version
378 needs a little more clarification however. The '#5' means that
379 this is the fifth kernel built from this source base and the
380 date behind it indicates the time the kernel was built.
381 The only way to tune these values is to rebuild the kernel :-)
383 ==============================================================
385 overflowgid & overflowuid:
387 if your architecture did not always support 32-bit UIDs (i.e. arm,
388 i386, m68k, sh, and sparc32), a fixed UID and GID will be returned to
389 applications that use the old 16-bit UID/GID system calls, if the
390 actual UID or GID would exceed 65535.
392 These sysctls allow you to change the value of the fixed UID and GID.
393 The default is 65534.
395 ==============================================================
399 The value in this file represents the number of seconds the kernel
400 waits before rebooting on a panic. When you use the software watchdog,
401 the recommended setting is 60.
403 ==============================================================
405 panic_on_unrecovered_nmi:
407 The default Linux behaviour on an NMI of either memory or unknown is
408 to continue operation. For many environments such as scientific
409 computing it is preferable that the box is taken out and the error
410 dealt with than an uncorrected parity/ECC error get propagated.
412 A small number of systems do generate NMI's for bizarre random reasons
413 such as power management so the default is off. That sysctl works like
414 the existing panic controls already in that directory.
416 ==============================================================
420 Controls the kernel's behaviour when an oops or BUG is encountered.
422 0: try to continue operation
424 1: panic immediately. If the `panic' sysctl is also non-zero then the
425 machine will be rebooted.
427 ==============================================================
429 panic_on_stackoverflow:
431 Controls the kernel's behavior when detecting the overflows of
432 kernel, IRQ and exception stacks except a user stack.
433 This file shows up if CONFIG_DEBUG_STACKOVERFLOW is enabled.
435 0: try to continue operation.
437 1: panic immediately.
439 ==============================================================
441 perf_cpu_time_max_percent:
443 Hints to the kernel how much CPU time it should be allowed to
444 use to handle perf sampling events. If the perf subsystem
445 is informed that its samples are exceeding this limit, it
446 will drop its sampling frequency to attempt to reduce its CPU
449 Some perf sampling happens in NMIs. If these samples
450 unexpectedly take too long to execute, the NMIs can become
451 stacked up next to each other so much that nothing else is
454 0: disable the mechanism. Do not monitor or correct perf's
455 sampling rate no matter how CPU time it takes.
457 1-100: attempt to throttle perf's sample rate to this
458 percentage of CPU. Note: the kernel calculates an
459 "expected" length of each sample event. 100 here means
460 100% of that expected length. Even if this is set to
461 100, you may still see sample throttling if this
462 length is exceeded. Set to 0 if you truly do not care
463 how much CPU is consumed.
465 ==============================================================
470 PID allocation wrap value. When the kernel's next PID value
471 reaches this value, it wraps back to a minimum PID value.
472 PIDs of value pid_max or larger are not allocated.
474 ==============================================================
478 The last pid allocated in the current (the one task using this sysctl
479 lives in) pid namespace. When selecting a pid for a next task on fork
480 kernel tries to allocate a number starting from this one.
482 ==============================================================
484 powersave-nap: (PPC only)
486 If set, Linux-PPC will use the 'nap' mode of powersaving,
487 otherwise the 'doze' mode will be used.
489 ==============================================================
493 The four values in printk denote: console_loglevel,
494 default_message_loglevel, minimum_console_loglevel and
495 default_console_loglevel respectively.
497 These values influence printk() behavior when printing or
498 logging error messages. See 'man 2 syslog' for more info on
499 the different loglevels.
501 - console_loglevel: messages with a higher priority than
502 this will be printed to the console
503 - default_message_loglevel: messages without an explicit priority
504 will be printed with this priority
505 - minimum_console_loglevel: minimum (highest) value to which
506 console_loglevel can be set
507 - default_console_loglevel: default value for console_loglevel
509 ==============================================================
513 Delay each printk message in printk_delay milliseconds
515 Value from 0 - 10000 is allowed.
517 ==============================================================
521 Some warning messages are rate limited. printk_ratelimit specifies
522 the minimum length of time between these messages (in jiffies), by
523 default we allow one every 5 seconds.
525 A value of 0 will disable rate limiting.
527 ==============================================================
529 printk_ratelimit_burst:
531 While long term we enforce one message per printk_ratelimit
532 seconds, we do allow a burst of messages to pass through.
533 printk_ratelimit_burst specifies the number of messages we can
534 send before ratelimiting kicks in.
536 ==============================================================
540 This option can be used to select the type of process address
541 space randomization that is used in the system, for architectures
542 that support this feature.
544 0 - Turn the process address space randomization off. This is the
545 default for architectures that do not support this feature anyways,
546 and kernels that are booted with the "norandmaps" parameter.
548 1 - Make the addresses of mmap base, stack and VDSO page randomized.
549 This, among other things, implies that shared libraries will be
550 loaded to random addresses. Also for PIE-linked binaries, the
551 location of code start is randomized. This is the default if the
552 CONFIG_COMPAT_BRK option is enabled.
554 2 - Additionally enable heap randomization. This is the default if
555 CONFIG_COMPAT_BRK is disabled.
557 There are a few legacy applications out there (such as some ancient
558 versions of libc.so.5 from 1996) that assume that brk area starts
559 just after the end of the code+bss. These applications break when
560 start of the brk area is randomized. There are however no known
561 non-legacy applications that would be broken this way, so for most
562 systems it is safe to choose full randomization.
564 Systems with ancient and/or broken binaries should be configured
565 with CONFIG_COMPAT_BRK enabled, which excludes the heap from process
566 address space randomization.
568 ==============================================================
570 reboot-cmd: (Sparc only)
572 ??? This seems to be a way to give an argument to the Sparc
573 ROM/Flash boot loader. Maybe to tell it what to do after
576 ==============================================================
578 rtsig-max & rtsig-nr:
580 The file rtsig-max can be used to tune the maximum number
581 of POSIX realtime (queued) signals that can be outstanding
584 rtsig-nr shows the number of RT signals currently queued.
586 ==============================================================
590 This file shows the size of the generic SCSI (sg) buffer.
591 You can't tune it just yet, but you could change it on
592 compile time by editing include/scsi/sg.h and changing
593 the value of SG_BIG_BUFF.
595 There shouldn't be any reason to change this value. If
596 you can come up with one, you probably know what you
599 ==============================================================
603 This parameter sets the total amount of shared memory pages that
604 can be used system wide. Hence, SHMALL should always be at least
605 ceil(shmmax/PAGE_SIZE).
607 If you are not sure what the default PAGE_SIZE is on your Linux
608 system, you can run the following command:
612 ==============================================================
616 This value can be used to query and set the run time limit
617 on the maximum shared memory segment size that can be created.
618 Shared memory segments up to 1Gb are now supported in the
619 kernel. This value defaults to SHMMAX.
621 ==============================================================
625 Linux lets you set resource limits, including how much memory one
626 process can consume, via setrlimit(2). Unfortunately, shared memory
627 segments are allowed to exist without association with any process, and
628 thus might not be counted against any resource limits. If enabled,
629 shared memory segments are automatically destroyed when their attach
630 count becomes zero after a detach or a process termination. It will
631 also destroy segments that were created, but never attached to, on exit
632 from the process. The only use left for IPC_RMID is to immediately
633 destroy an unattached segment. Of course, this breaks the way things are
634 defined, so some applications might stop working. Note that this
635 feature will do you no good unless you also configure your resource
636 limits (in particular, RLIMIT_AS and RLIMIT_NPROC). Most systems don't
639 Note that if you change this from 0 to 1, already created segments
640 without users and with a dead originative process will be destroyed.
642 ==============================================================
646 This value can be used to lower the softlockup tolerance threshold. The
647 default threshold is 60 seconds. If a cpu is locked up for 60 seconds,
648 the kernel complains. Valid values are 1-60 seconds. Setting this
649 tunable to zero will disable the softlockup detection altogether.
651 ==============================================================
655 Non-zero if the kernel has been tainted. Numeric values, which
656 can be ORed together:
658 1 - A module with a non-GPL license has been loaded, this
659 includes modules with no license.
660 Set by modutils >= 2.4.9 and module-init-tools.
661 2 - A module was force loaded by insmod -f.
662 Set by modutils >= 2.4.9 and module-init-tools.
663 4 - Unsafe SMP processors: SMP with CPUs not designed for SMP.
664 8 - A module was forcibly unloaded from the system by rmmod -f.
665 16 - A hardware machine check error occurred on the system.
666 32 - A bad page was discovered on the system.
667 64 - The user has asked that the system be marked "tainted". This
668 could be because they are running software that directly modifies
669 the hardware, or for other reasons.
670 128 - The system has died.
671 256 - The ACPI DSDT has been overridden with one supplied by the user
672 instead of using the one provided by the hardware.
673 512 - A kernel warning has occurred.
674 1024 - A module from drivers/staging was loaded.
675 2048 - The system is working around a severe firmware bug.
676 4096 - An out-of-tree module has been loaded.
678 ==============================================================
682 The value in this file affects behavior of handling NMI. When the
683 value is non-zero, unknown NMI is trapped and then panic occurs. At
684 that time, kernel debugging information is displayed on console.
686 NMI switch that most IA32 servers have fires unknown NMI up, for
687 example. If a system hangs up, try pressing the NMI switch.