1 ===================================
2 Documentation for /proc/sys/kernel/
3 ===================================
5 .. See scripts/check-sysctl-docs to keep this up to date
8 Copyright (c) 1998, 1999, Rik van Riel <riel@nl.linux.org>
10 Copyright (c) 2009, Shen Feng<shen@cn.fujitsu.com>
12 For general info and legal blurb, please look in
13 Documentation/admin-guide/sysctl/index.rst.
15 ------------------------------------------------------------------------------
17 This file contains documentation for the sysctl files in
18 ``/proc/sys/kernel/``.
20 The files in this directory can be used to tune and monitor
21 miscellaneous and general things in the operation of the Linux
22 kernel. Since some of the files *can* be used to screw up your
23 system, it is advisable to read both documentation and source
24 before actually making adjustments.
26 Currently, these files might (depending on your configuration)
27 show up in ``/proc/sys/kernel``:
37 highwater lowwater frequency
39 If BSD-style process accounting is enabled these values control
40 its behaviour. If free space on filesystem where the log lives
41 goes below ``lowwater``\ % accounting suspends. If free space gets
42 above ``highwater``\ % accounting resumes. ``frequency`` determines
43 how often do we check the amount of free space (value is in
50 That is, suspend accounting if free space drops below 2%; resume it
51 if it increases to at least 4%; consider information about amount of
52 free space valid for 30 seconds.
58 See Documentation/power/video.rst. This allows the video resume mode to be set,
59 in a similar fashion to the ``acpi_sleep`` kernel parameter, by
60 combining the following values:
71 The machine hardware name, the same output as ``uname -m``
72 (e.g. ``x86_64`` or ``aarch64``).
77 This variable has no effect and may be removed in future kernel
78 releases. Reading it always returns 0.
79 Up to Linux 3.17, it enabled/disabled automatic recomputing of
81 upon memory add/remove or upon IPC namespace creation/removal.
82 Echoing "1" into this file enabled msgmni automatic recomputing.
83 Echoing "0" turned it off. The default value was 1.
86 bootloader_type (x86 only)
87 ==========================
89 This gives the bootloader type number as indicated by the bootloader,
90 shifted left by 4, and OR'd with the low four bits of the bootloader
91 version. The reason for this encoding is that this used to match the
92 ``type_of_loader`` field in the kernel header; the encoding is kept for
93 backwards compatibility. That is, if the full bootloader type number
94 is 0x15 and the full version number is 0x234, this file will contain
95 the value 340 = 0x154.
97 See the ``type_of_loader`` and ``ext_loader_type`` fields in
98 Documentation/x86/boot.rst for additional information.
101 bootloader_version (x86 only)
102 =============================
104 The complete bootloader version number. In the example above, this
105 file will contain the value 564 = 0x234.
107 See the ``type_of_loader`` and ``ext_loader_ver`` fields in
108 Documentation/x86/boot.rst for additional information.
114 Controls whether the kernel should collect statistics on BPF programs
115 (total time spent running, number of times run...). Enabling
116 statistics causes a slight reduction in performance on each program
117 run. The statistics can be seen using ``bpftool``.
119 = ===================================
120 0 Don't collect statistics (default).
121 1 Collect statistics.
122 = ===================================
128 This is the pid which will be signalled on reboot (notably, by
129 Ctrl-Alt-Delete). Writing a value to this file which doesn't
130 correspond to a running process will result in ``-ESRCH``.
132 See also `ctrl-alt-del`_.
138 Highest valid capability of the running kernel. Exports
139 ``CAP_LAST_CAP`` from the kernel.
147 ``core_pattern`` is used to specify a core dumpfile pattern name.
149 * max length 127 characters; default value is "core"
150 * ``core_pattern`` is used as a pattern template for the output
151 filename; certain string patterns (beginning with '%') are
152 substituted with their actual values.
153 * backward compatibility with ``core_uses_pid``:
155 If ``core_pattern`` does not include "%p" (default does not)
156 and ``core_uses_pid`` is set, then .PID will be appended to
159 * corename format specifiers
161 ======== ==========================================
162 %<NUL> '%' is dropped
165 %P global pid (init PID namespace)
167 %I global tid (init PID namespace)
168 %u uid (in initial user namespace)
169 %g gid (in initial user namespace)
170 %d dump mode, matches ``PR_SET_DUMPABLE`` and
171 ``/proc/sys/fs/suid_dumpable``
175 %e executable filename (may be shortened, could be changed by prctl etc)
176 %f executable filename
178 %c maximum size of core file by resource limit RLIMIT_CORE
179 %C CPU the task ran on
180 %<OTHER> both are dropped
181 ======== ==========================================
183 * If the first character of the pattern is a '|', the kernel will treat
184 the rest of the pattern as a command to run. The core dump will be
185 written to the standard input of that program instead of to a file.
191 This sysctl is only applicable when `core_pattern`_ is configured to
192 pipe core files to a user space helper (when the first character of
193 ``core_pattern`` is a '|', see above).
194 When collecting cores via a pipe to an application, it is occasionally
195 useful for the collecting application to gather data about the
196 crashing process from its ``/proc/pid`` directory.
197 In order to do this safely, the kernel must wait for the collecting
198 process to exit, so as not to remove the crashing processes proc files
200 This in turn creates the possibility that a misbehaving userspace
201 collecting process can block the reaping of a crashed process simply
203 This sysctl defends against that.
204 It defines how many concurrent crashing processes may be piped to user
205 space applications in parallel.
206 If this value is exceeded, then those crashing processes above that
207 value are noted via the kernel log and their cores are skipped.
208 0 is a special value, indicating that unlimited processes may be
209 captured in parallel, but that no waiting will take place (i.e. the
210 collecting process is not guaranteed access to ``/proc/<crashing
212 This value defaults to 0.
218 The default coredump filename is "core". By setting
219 ``core_uses_pid`` to 1, the coredump filename becomes core.PID.
220 If `core_pattern`_ does not include "%p" (default does not)
221 and ``core_uses_pid`` is set, then .PID will be appended to
228 When the value in this file is 0, ctrl-alt-del is trapped and
229 sent to the ``init(1)`` program to handle a graceful restart.
230 When, however, the value is > 0, Linux's reaction to a Vulcan
231 Nerve Pinch (tm) will be an immediate reboot, without even
232 syncing its dirty buffers.
235 when a program (like dosemu) has the keyboard in 'raw'
236 mode, the ctrl-alt-del is intercepted by the program before it
237 ever reaches the kernel tty layer, and it's up to the program
238 to decide what to do with it.
244 This toggle indicates whether unprivileged users are prevented
245 from using ``dmesg(8)`` to view messages from the kernel's log
247 When ``dmesg_restrict`` is set to 0 there are no restrictions.
248 When ``dmesg_restrict`` is set to 1, users must have
249 ``CAP_SYSLOG`` to use ``dmesg(8)``.
251 The kernel config option ``CONFIG_SECURITY_DMESG_RESTRICT`` sets the
252 default value of ``dmesg_restrict``.
255 domainname & hostname
256 =====================
258 These files can be used to set the NIS/YP domainname and the
259 hostname of your box in exactly the same way as the commands
260 domainname and hostname, i.e.::
262 # echo "darkstar" > /proc/sys/kernel/hostname
263 # echo "mydomain" > /proc/sys/kernel/domainname
265 has the same effect as::
267 # hostname "darkstar"
268 # domainname "mydomain"
270 Note, however, that the classic darkstar.frop.org has the
271 hostname "darkstar" and DNS (Internet Domain Name Server)
272 domainname "frop.org", not to be confused with the NIS (Network
273 Information Service) or YP (Yellow Pages) domainname. These two
274 domain names are in general different. For a detailed discussion
275 see the ``hostname(1)`` man page.
281 See Documentation/driver-api/firmware/fallback-mechanisms.rst.
283 The entries in this directory allow the firmware loader helper
284 fallback to be controlled:
286 * ``force_sysfs_fallback``, when set to 1, forces the use of the
288 * ``ignore_sysfs_fallback``, when set to 1, ignores any fallback.
294 Determines whether ``ftrace_dump()`` should be called on an oops (or
295 kernel panic). This will output the contents of the ftrace buffers to
296 the console. This is very useful for capturing traces that lead to
297 crashes and outputting them to a serial console.
299 = ===================================================
300 0 Disabled (default).
301 1 Dump buffers of all CPUs.
302 2 Dump the buffer of the CPU that triggered the oops.
303 = ===================================================
306 ftrace_enabled, stack_tracer_enabled
307 ====================================
309 See Documentation/trace/ftrace.rst.
312 hardlockup_all_cpu_backtrace
313 ============================
315 This value controls the hard lockup detector behavior when a hard
316 lockup condition is detected as to whether or not to gather further
317 debug information. If enabled, arch-specific all-CPU stack dumping
320 = ============================================
321 0 Do nothing. This is the default behavior.
322 1 On detection capture more debug information.
323 = ============================================
329 This parameter can be used to control whether the kernel panics
330 when a hard lockup is detected.
332 = ===========================
333 0 Don't panic on hard lockup.
334 1 Panic on hard lockup.
335 = ===========================
337 See Documentation/admin-guide/lockup-watchdogs.rst for more information.
338 This can also be set using the nmi_watchdog kernel parameter.
344 Path for the hotplug policy agent.
345 Default value is ``CONFIG_UEVENT_HELPER_PATH``, which in turn defaults
348 This file only exists when ``CONFIG_UEVENT_HELPER`` is enabled. Most
349 modern systems rely exclusively on the netlink-based uevent source and
353 hung_task_all_cpu_backtrace
354 ===========================
356 If this option is set, the kernel will send an NMI to all CPUs to dump
357 their backtraces when a hung task is detected. This file shows up if
358 CONFIG_DETECT_HUNG_TASK and CONFIG_SMP are enabled.
360 0: Won't show all CPUs backtraces when a hung task is detected.
361 This is the default behavior.
363 1: Will non-maskably interrupt all CPUs and dump their backtraces when
364 a hung task is detected.
370 Controls the kernel's behavior when a hung task is detected.
371 This file shows up if ``CONFIG_DETECT_HUNG_TASK`` is enabled.
373 = =================================================
374 0 Continue operation. This is the default behavior.
376 = =================================================
379 hung_task_check_count
380 =====================
382 The upper bound on the number of tasks that are checked.
383 This file shows up if ``CONFIG_DETECT_HUNG_TASK`` is enabled.
386 hung_task_timeout_secs
387 ======================
389 When a task in D state did not get scheduled
390 for more than this value report a warning.
391 This file shows up if ``CONFIG_DETECT_HUNG_TASK`` is enabled.
393 0 means infinite timeout, no checking is done.
395 Possible values to set are in range {0:``LONG_MAX``/``HZ``}.
398 hung_task_check_interval_secs
399 =============================
401 Hung task check interval. If hung task checking is enabled
402 (see `hung_task_timeout_secs`_), the check is done every
403 ``hung_task_check_interval_secs`` seconds.
404 This file shows up if ``CONFIG_DETECT_HUNG_TASK`` is enabled.
406 0 (default) means use ``hung_task_timeout_secs`` as checking
409 Possible values to set are in range {0:``LONG_MAX``/``HZ``}.
415 The maximum number of warnings to report. During a check interval
416 if a hung task is detected, this value is decreased by 1.
417 When this value reaches 0, no more warnings will be reported.
418 This file shows up if ``CONFIG_DETECT_HUNG_TASK`` is enabled.
420 -1: report an infinite number of warnings.
423 hyperv_record_panic_msg
424 =======================
426 Controls whether the panic kmsg data should be reported to Hyper-V.
428 = =========================================================
429 0 Do not report panic kmsg data.
430 1 Report the panic kmsg data. This is the default behavior.
431 = =========================================================
434 ignore-unaligned-usertrap
435 =========================
437 On architectures where unaligned accesses cause traps, and where this
438 feature is supported (``CONFIG_SYSCTL_ARCH_UNALIGN_NO_WARN``;
439 currently, ``arc`` and ``ia64``), controls whether all unaligned traps
442 = =============================================================
443 0 Log all unaligned accesses.
444 1 Only warn the first time a process traps. This is the default
446 = =============================================================
448 See also `unaligned-trap`_ and `unaligned-dump-stack`_. On ``ia64``,
449 this allows system administrators to override the
450 ``IA64_THREAD_UAC_NOPRINT`` ``prctl`` and avoid logs being flooded.
456 A toggle indicating if the ``kexec_load`` syscall has been disabled.
457 This value defaults to 0 (false: ``kexec_load`` enabled), but can be
458 set to 1 (true: ``kexec_load`` disabled).
459 Once true, kexec can no longer be used, and the toggle cannot be set
461 This allows a kexec image to be loaded before disabling the syscall,
462 allowing a system to set up (and later use) an image without it being
464 Generally used together with the `modules_disabled`_ sysctl.
470 This toggle indicates whether restrictions are placed on
471 exposing kernel addresses via ``/proc`` and other interfaces.
473 When ``kptr_restrict`` is set to 0 (the default) the address is hashed
475 (This is the equivalent to %p.)
477 When ``kptr_restrict`` is set to 1, kernel pointers printed using the
478 %pK format specifier will be replaced with 0s unless the user has
479 ``CAP_SYSLOG`` and effective user and group ids are equal to the real
481 This is because %pK checks are done at read() time rather than open()
482 time, so if permissions are elevated between the open() and the read()
483 (e.g via a setuid binary) then %pK will not leak kernel pointers to
485 Note, this is a temporary solution only.
486 The correct long-term solution is to do the permission checks at
488 Consider removing world read permissions from files that use %pK, and
489 using `dmesg_restrict`_ to protect against uses of %pK in ``dmesg(8)``
490 if leaking kernel pointer values to unprivileged users is a concern.
492 When ``kptr_restrict`` is set to 2, kernel pointers printed using
493 %pK will be replaced with 0s regardless of privileges.
499 The full path to the usermode helper for autoloading kernel modules,
500 by default ``CONFIG_MODPROBE_PATH``, which in turn defaults to
501 "/sbin/modprobe". This binary is executed when the kernel requests a
502 module. For example, if userspace passes an unknown filesystem type
503 to mount(), then the kernel will automatically request the
504 corresponding filesystem module by executing this usermode helper.
505 This usermode helper should insert the needed module into the kernel.
507 This sysctl only affects module autoloading. It has no effect on the
508 ability to explicitly insert modules.
510 This sysctl can be used to debug module loading requests::
512 echo '#! /bin/sh' > /tmp/modprobe
513 echo 'echo "$@" >> /tmp/modprobe.log' >> /tmp/modprobe
514 echo 'exec /sbin/modprobe "$@"' >> /tmp/modprobe
515 chmod a+x /tmp/modprobe
516 echo /tmp/modprobe > /proc/sys/kernel/modprobe
518 Alternatively, if this sysctl is set to the empty string, then module
519 autoloading is completely disabled. The kernel will not try to
520 execute a usermode helper at all, nor will it call the
521 kernel_module_request LSM hook.
523 If CONFIG_STATIC_USERMODEHELPER=y is set in the kernel configuration,
524 then the configured static usermode helper overrides this sysctl,
525 except that the empty string is still accepted to completely disable
526 module autoloading as described above.
531 A toggle value indicating if modules are allowed to be loaded
532 in an otherwise modular kernel. This toggle defaults to off
533 (0), but can be set true (1). Once true, modules can be
534 neither loaded nor unloaded, and the toggle cannot be set back
535 to false. Generally used with the `kexec_load_disabled`_ toggle.
540 msgmax, msgmnb, and msgmni
541 ==========================
543 ``msgmax`` is the maximum size of an IPC message, in bytes. 8192 by
544 default (``MSGMAX``).
546 ``msgmnb`` is the maximum size of an IPC queue, in bytes. 16384 by
547 default (``MSGMNB``).
549 ``msgmni`` is the maximum number of IPC queues. 32000 by default
553 msg_next_id, sem_next_id, and shm_next_id (System V IPC)
554 ========================================================
556 These three toggles allows to specify desired id for next allocated IPC
557 object: message, semaphore or shared memory respectively.
559 By default they are equal to -1, which means generic allocation logic.
560 Possible values to set are in range {0:``INT_MAX``}.
563 1) kernel doesn't guarantee, that new object will have desired id. So,
564 it's up to userspace, how to handle an object with "wrong" id.
565 2) Toggle with non-default value will be set back to -1 by kernel after
566 successful IPC object allocation. If an IPC object allocation syscall
567 fails, it is undefined if the value remains unmodified or is reset to -1.
573 Maximum number of supplementary groups, _i.e._ the maximum size which
574 ``setgroups`` will accept. Exports ``NGROUPS_MAX`` from the kernel.
581 This parameter can be used to control the NMI watchdog
582 (i.e. the hard lockup detector) on x86 systems.
584 = =================================
585 0 Disable the hard lockup detector.
586 1 Enable the hard lockup detector.
587 = =================================
589 The hard lockup detector monitors each CPU for its ability to respond to
590 timer interrupts. The mechanism utilizes CPU performance counter registers
591 that are programmed to generate Non-Maskable Interrupts (NMIs) periodically
592 while a CPU is busy. Hence, the alternative name 'NMI watchdog'.
594 The NMI watchdog is disabled by default if the kernel is running as a guest
595 in a KVM virtual machine. This default can be overridden by adding::
599 to the guest kernel command line (see
600 Documentation/admin-guide/kernel-parameters.rst).
603 nmi_wd_lpm_factor (PPC only)
604 ============================
606 Factor to apply to the NMI watchdog timeout (only when ``nmi_watchdog`` is
607 set to 1). This factor represents the percentage added to
608 ``watchdog_thresh`` when calculating the NMI watchdog timeout during an
609 LPM. The soft lockup timeout is not impacted.
611 A value of 0 means no change. The default value is 200 meaning the NMI
612 watchdog is set to 30s (based on ``watchdog_thresh`` equal to 10).
618 Enables/disables and configures automatic page fault based NUMA memory
619 balancing. Memory is moved automatically to nodes that access it often.
620 The value to set can be the result of ORing the following:
622 = =================================
623 0 NUMA_BALANCING_DISABLED
624 1 NUMA_BALANCING_NORMAL
625 2 NUMA_BALANCING_MEMORY_TIERING
626 = =================================
628 Or NUMA_BALANCING_NORMAL to optimize page placement among different
629 NUMA nodes to reduce remote accessing. On NUMA machines, there is a
630 performance penalty if remote memory is accessed by a CPU. When this
631 feature is enabled the kernel samples what task thread is accessing
632 memory by periodically unmapping pages and later trapping a page
633 fault. At the time of the page fault, it is determined if the data
634 being accessed should be migrated to a local memory node.
636 The unmapping of pages and trapping faults incur additional overhead that
637 ideally is offset by improved memory locality but there is no universal
638 guarantee. If the target workload is already bound to NUMA nodes then this
639 feature should be disabled.
641 Or NUMA_BALANCING_MEMORY_TIERING to optimize page placement among
642 different types of memory (represented as different NUMA nodes) to
643 place the hot pages in the fast memory. This is implemented based on
644 unmapping and page fault too.
646 numa_balancing_promote_rate_limit_MBps
647 ======================================
649 Too high promotion/demotion throughput between different memory types
650 may hurt application latency. This can be used to rate limit the
651 promotion throughput. The per-node max promotion throughput in MB/s
652 will be limited to be no more than the set value.
654 A rule of thumb is to set this to less than 1/10 of the PMEM node
657 oops_all_cpu_backtrace
658 ======================
660 If this option is set, the kernel will send an NMI to all CPUs to dump
661 their backtraces when an oops event occurs. It should be used as a last
662 resort in case a panic cannot be triggered (to protect VMs running, for
663 example) or kdump can't be collected. This file shows up if CONFIG_SMP
666 0: Won't show all CPUs backtraces when an oops is detected.
667 This is the default behavior.
669 1: Will non-maskably interrupt all CPUs and dump their backtraces when
670 an oops event is detected.
676 Number of kernel oopses after which the kernel should panic when
677 ``panic_on_oops`` is not set. Setting this to 0 disables checking
678 the count. Setting this to 1 has the same effect as setting
679 ``panic_on_oops=1``. The default value is 10000.
682 osrelease, ostype & version
683 ===========================
692 #5 Wed Feb 25 21:49:24 MET 1998
694 The files ``osrelease`` and ``ostype`` should be clear enough.
696 needs a little more clarification however. The '#5' means that
697 this is the fifth kernel built from this source base and the
698 date behind it indicates the time the kernel was built.
699 The only way to tune these values is to rebuild the kernel :-)
702 overflowgid & overflowuid
703 =========================
705 if your architecture did not always support 32-bit UIDs (i.e. arm,
706 i386, m68k, sh, and sparc32), a fixed UID and GID will be returned to
707 applications that use the old 16-bit UID/GID system calls, if the
708 actual UID or GID would exceed 65535.
710 These sysctls allow you to change the value of the fixed UID and GID.
711 The default is 65534.
717 The value in this file determines the behaviour of the kernel on a
720 * if zero, the kernel will loop forever;
721 * if negative, the kernel will reboot immediately;
722 * if positive, the kernel will reboot after the corresponding number
725 When you use the software watchdog, the recommended setting is 60.
731 Controls the kernel's behavior when a CPU receives an NMI caused by
734 = ==================================================================
735 0 Try to continue operation (default).
736 1 Panic immediately. The IO error triggered an NMI. This indicates a
737 serious system condition which could result in IO data corruption.
738 Rather than continuing, panicking might be a better choice. Some
739 servers issue this sort of NMI when the dump button is pushed,
740 and you can use this option to take a crash dump.
741 = ==================================================================
747 Controls the kernel's behaviour when an oops or BUG is encountered.
749 = ===================================================================
750 0 Try to continue operation.
751 1 Panic immediately. If the `panic` sysctl is also non-zero then the
752 machine will be rebooted.
753 = ===================================================================
756 panic_on_stackoverflow
757 ======================
759 Controls the kernel's behavior when detecting the overflows of
760 kernel, IRQ and exception stacks except a user stack.
761 This file shows up if ``CONFIG_DEBUG_STACKOVERFLOW`` is enabled.
763 = ==========================
764 0 Try to continue operation.
766 = ==========================
769 panic_on_unrecovered_nmi
770 ========================
772 The default Linux behaviour on an NMI of either memory or unknown is
773 to continue operation. For many environments such as scientific
774 computing it is preferable that the box is taken out and the error
775 dealt with than an uncorrected parity/ECC error get propagated.
777 A small number of systems do generate NMIs for bizarre random reasons
778 such as power management so the default is off. That sysctl works like
779 the existing panic controls already in that directory.
785 Calls panic() in the WARN() path when set to 1. This is useful to avoid
786 a kernel rebuild when attempting to kdump at the location of a WARN().
788 = ================================================
789 0 Only WARN(), default behaviour.
790 1 Call panic() after printing out WARN() location.
791 = ================================================
797 Bitmask for printing system info when panic happens. User can chose
798 combination of the following bits:
800 ===== ============================================
801 bit 0 print all tasks info
802 bit 1 print system memory info
803 bit 2 print timer info
804 bit 3 print locks info if ``CONFIG_LOCKDEP`` is on
805 bit 4 print ftrace buffer
806 bit 5 print all printk messages in buffer
807 bit 6 print all CPUs backtrace (if available in the arch)
808 ===== ============================================
810 So for example to print tasks and memory info on panic, user can::
812 echo 3 > /proc/sys/kernel/panic_print
818 When set to 1, calls panic() after RCU stall detection messages. This
819 is useful to define the root cause of RCU stalls using a vmcore.
821 = ============================================================
822 0 Do not panic() when RCU stall takes place, default behavior.
823 1 panic() after printing RCU stall messages.
824 = ============================================================
826 max_rcu_stall_to_panic
827 ======================
829 When ``panic_on_rcu_stall`` is set to 1, this value determines the
830 number of times that RCU can stall before panic() is called.
832 When ``panic_on_rcu_stall`` is set to 0, this value is has no effect.
834 perf_cpu_time_max_percent
835 =========================
837 Hints to the kernel how much CPU time it should be allowed to
838 use to handle perf sampling events. If the perf subsystem
839 is informed that its samples are exceeding this limit, it
840 will drop its sampling frequency to attempt to reduce its CPU
843 Some perf sampling happens in NMIs. If these samples
844 unexpectedly take too long to execute, the NMIs can become
845 stacked up next to each other so much that nothing else is
848 ===== ========================================================
849 0 Disable the mechanism. Do not monitor or correct perf's
850 sampling rate no matter how CPU time it takes.
852 1-100 Attempt to throttle perf's sample rate to this
853 percentage of CPU. Note: the kernel calculates an
854 "expected" length of each sample event. 100 here means
855 100% of that expected length. Even if this is set to
856 100, you may still see sample throttling if this
857 length is exceeded. Set to 0 if you truly do not care
858 how much CPU is consumed.
859 ===== ========================================================
865 Controls use of the performance events system by unprivileged
866 users (without CAP_PERFMON). The default value is 2.
868 For backward compatibility reasons access to system performance
869 monitoring and observability remains open for CAP_SYS_ADMIN
870 privileged processes but CAP_SYS_ADMIN usage for secure system
871 performance monitoring and observability operations is discouraged
872 with respect to CAP_PERFMON use cases.
874 === ==================================================================
875 -1 Allow use of (almost) all events by all users.
877 Ignore mlock limit after perf_event_mlock_kb without
880 >=0 Disallow ftrace function tracepoint by users without
883 Disallow raw tracepoint access by users without ``CAP_PERFMON``.
885 >=1 Disallow CPU event access by users without ``CAP_PERFMON``.
887 >=2 Disallow kernel profiling by users without ``CAP_PERFMON``.
888 === ==================================================================
894 Controls maximum number of stack frames to copy for (``attr.sample_type &
895 PERF_SAMPLE_CALLCHAIN``) configured events, for instance, when using
896 '``perf record -g``' or '``perf trace --call-graph fp``'.
898 This can only be done when no events are in use that have callchains
899 enabled, otherwise writing to this file will return ``-EBUSY``.
901 The default value is 127.
907 Control size of per-cpu ring buffer not counted against mlock limit.
909 The default value is 512 + 1 page
912 perf_event_max_contexts_per_stack
913 =================================
915 Controls maximum number of stack frame context entries for
916 (``attr.sample_type & PERF_SAMPLE_CALLCHAIN``) configured events, for
917 instance, when using '``perf record -g``' or '``perf trace --call-graph fp``'.
919 This can only be done when no events are in use that have callchains
920 enabled, otherwise writing to this file will return ``-EBUSY``.
922 The default value is 8.
925 perf_user_access (arm64 only)
926 =================================
928 Controls user space access for reading perf event counters. When set to 1,
929 user space can read performance monitor counter registers directly.
931 The default value is 0 (access disabled).
933 See Documentation/arm64/perf.rst for more information.
939 PID allocation wrap value. When the kernel's next PID value
940 reaches this value, it wraps back to a minimum PID value.
941 PIDs of value ``pid_max`` or larger are not allocated.
947 The last pid allocated in the current (the one task using this sysctl
948 lives in) pid namespace. When selecting a pid for a next task on fork
949 kernel tries to allocate a number starting from this one.
952 powersave-nap (PPC only)
953 ========================
955 If set, Linux-PPC will use the 'nap' mode of powersaving,
956 otherwise the 'doze' mode will be used.
959 ==============================================================
964 The four values in printk denote: ``console_loglevel``,
965 ``default_message_loglevel``, ``minimum_console_loglevel`` and
966 ``default_console_loglevel`` respectively.
968 These values influence printk() behavior when printing or
969 logging error messages. See '``man 2 syslog``' for more info on
970 the different loglevels.
972 ======================== =====================================
973 console_loglevel messages with a higher priority than
974 this will be printed to the console
975 default_message_loglevel messages without an explicit priority
976 will be printed with this priority
977 minimum_console_loglevel minimum (highest) value to which
978 console_loglevel can be set
979 default_console_loglevel default value for console_loglevel
980 ======================== =====================================
986 Delay each printk message in ``printk_delay`` milliseconds
988 Value from 0 - 10000 is allowed.
994 Some warning messages are rate limited. ``printk_ratelimit`` specifies
995 the minimum length of time between these messages (in seconds).
996 The default value is 5 seconds.
998 A value of 0 will disable rate limiting.
1001 printk_ratelimit_burst
1002 ======================
1004 While long term we enforce one message per `printk_ratelimit`_
1005 seconds, we do allow a burst of messages to pass through.
1006 ``printk_ratelimit_burst`` specifies the number of messages we can
1007 send before ratelimiting kicks in.
1009 The default value is 10 messages.
1015 Control the logging to ``/dev/kmsg`` from userspace:
1017 ========= =============================================
1018 ratelimit default, ratelimited
1019 on unlimited logging to /dev/kmsg from userspace
1020 off logging to /dev/kmsg disabled
1021 ========= =============================================
1023 The kernel command line parameter ``printk.devkmsg=`` overrides this and is
1024 a one-time setting until next reboot: once set, it cannot be changed by
1025 this sysctl interface anymore.
1027 ==============================================================
1033 See Documentation/filesystems/devpts.rst.
1039 This is a directory, with the following entries:
1041 * ``boot_id``: a UUID generated the first time this is retrieved, and
1042 unvarying after that;
1044 * ``uuid``: a UUID generated every time this is retrieved (this can
1045 thus be used to generate UUIDs at will);
1047 * ``entropy_avail``: the pool's entropy count, in bits;
1049 * ``poolsize``: the entropy pool size, in bits;
1051 * ``urandom_min_reseed_secs``: obsolete (used to determine the minimum
1052 number of seconds between urandom pool reseeding). This file is
1053 writable for compatibility purposes, but writing to it has no effect
1054 on any RNG behavior;
1056 * ``write_wakeup_threshold``: when the entropy count drops below this
1057 (as a number of bits), processes waiting to write to ``/dev/random``
1058 are woken up. This file is writable for compatibility purposes, but
1059 writing to it has no effect on any RNG behavior.
1065 This option can be used to select the type of process address
1066 space randomization that is used in the system, for architectures
1067 that support this feature.
1069 == ===========================================================================
1070 0 Turn the process address space randomization off. This is the
1071 default for architectures that do not support this feature anyways,
1072 and kernels that are booted with the "norandmaps" parameter.
1074 1 Make the addresses of mmap base, stack and VDSO page randomized.
1075 This, among other things, implies that shared libraries will be
1076 loaded to random addresses. Also for PIE-linked binaries, the
1077 location of code start is randomized. This is the default if the
1078 ``CONFIG_COMPAT_BRK`` option is enabled.
1080 2 Additionally enable heap randomization. This is the default if
1081 ``CONFIG_COMPAT_BRK`` is disabled.
1083 There are a few legacy applications out there (such as some ancient
1084 versions of libc.so.5 from 1996) that assume that brk area starts
1085 just after the end of the code+bss. These applications break when
1086 start of the brk area is randomized. There are however no known
1087 non-legacy applications that would be broken this way, so for most
1088 systems it is safe to choose full randomization.
1090 Systems with ancient and/or broken binaries should be configured
1091 with ``CONFIG_COMPAT_BRK`` enabled, which excludes the heap from process
1092 address space randomization.
1093 == ===========================================================================
1099 See Documentation/admin-guide/initrd.rst.
1102 reboot-cmd (SPARC only)
1103 =======================
1105 ??? This seems to be a way to give an argument to the Sparc
1106 ROM/Flash boot loader. Maybe to tell it what to do after
1113 Enables/disables Energy Aware Scheduling (EAS). EAS starts
1114 automatically on platforms where it can run (that is,
1115 platforms with asymmetric CPU topologies and having an Energy
1116 Model available). If your platform happens to meet the
1117 requirements for EAS but you do not want to use it, change
1123 Enables/disables task delay accounting (see
1124 Documentation/accounting/delay-accounting.rst. Enabling this feature incurs
1125 a small amount of overhead in the scheduler but is useful for debugging
1126 and performance tuning. It is required by some tools such as iotop.
1131 Enables/disables scheduler statistics. Enabling this feature
1132 incurs a small amount of overhead in the scheduler but is
1133 useful for debugging and performance tuning.
1135 sched_util_clamp_min
1136 ====================
1138 Max allowed *minimum* utilization.
1140 Default value is 1024, which is the maximum possible value.
1142 It means that any requested uclamp.min value cannot be greater than
1143 sched_util_clamp_min, i.e., it is restricted to the range
1144 [0:sched_util_clamp_min].
1146 sched_util_clamp_max
1147 ====================
1149 Max allowed *maximum* utilization.
1151 Default value is 1024, which is the maximum possible value.
1153 It means that any requested uclamp.max value cannot be greater than
1154 sched_util_clamp_max, i.e., it is restricted to the range
1155 [0:sched_util_clamp_max].
1157 sched_util_clamp_min_rt_default
1158 ===============================
1160 By default Linux is tuned for performance. Which means that RT tasks always run
1161 at the highest frequency and most capable (highest capacity) CPU (in
1162 heterogeneous systems).
1164 Uclamp achieves this by setting the requested uclamp.min of all RT tasks to
1165 1024 by default, which effectively boosts the tasks to run at the highest
1166 frequency and biases them to run on the biggest CPU.
1168 This knob allows admins to change the default behavior when uclamp is being
1169 used. In battery powered devices particularly, running at the maximum
1170 capacity and frequency will increase energy consumption and shorten the battery
1173 This knob is only effective for RT tasks which the user hasn't modified their
1174 requested uclamp.min value via sched_setattr() syscall.
1176 This knob will not escape the range constraint imposed by sched_util_clamp_min
1181 sched_util_clamp_min_rt_default = 800
1182 sched_util_clamp_min = 600
1184 Then the boost will be clamped to 600 because 800 is outside of the permissible
1185 range of [0:600]. This could happen for instance if a powersave mode will
1186 restrict all boosts temporarily by modifying sched_util_clamp_min. As soon as
1187 this restriction is lifted, the requested sched_util_clamp_min_rt_default
1193 See Documentation/userspace-api/seccomp_filter.rst.
1199 This file shows the size of the generic SCSI (sg) buffer.
1200 You can't tune it just yet, but you could change it on
1201 compile time by editing ``include/scsi/sg.h`` and changing
1202 the value of ``SG_BIG_BUFF``.
1204 There shouldn't be any reason to change this value. If
1205 you can come up with one, you probably know what you
1212 This parameter sets the total amount of shared memory pages that
1213 can be used system wide. Hence, ``shmall`` should always be at least
1214 ``ceil(shmmax/PAGE_SIZE)``.
1216 If you are not sure what the default ``PAGE_SIZE`` is on your Linux
1217 system, you can run the following command::
1225 This value can be used to query and set the run time limit
1226 on the maximum shared memory segment size that can be created.
1227 Shared memory segments up to 1Gb are now supported in the
1228 kernel. This value defaults to ``SHMMAX``.
1234 This value determines the maximum number of shared memory segments.
1235 4096 by default (``SHMMNI``).
1241 Linux lets you set resource limits, including how much memory one
1242 process can consume, via ``setrlimit(2)``. Unfortunately, shared memory
1243 segments are allowed to exist without association with any process, and
1244 thus might not be counted against any resource limits. If enabled,
1245 shared memory segments are automatically destroyed when their attach
1246 count becomes zero after a detach or a process termination. It will
1247 also destroy segments that were created, but never attached to, on exit
1248 from the process. The only use left for ``IPC_RMID`` is to immediately
1249 destroy an unattached segment. Of course, this breaks the way things are
1250 defined, so some applications might stop working. Note that this
1251 feature will do you no good unless you also configure your resource
1252 limits (in particular, ``RLIMIT_AS`` and ``RLIMIT_NPROC``). Most systems don't
1255 Note that if you change this from 0 to 1, already created segments
1256 without users and with a dead originative process will be destroyed.
1259 sysctl_writes_strict
1260 ====================
1262 Control how file position affects the behavior of updating sysctl values
1263 via the ``/proc/sys`` interface:
1265 == ======================================================================
1266 -1 Legacy per-write sysctl value handling, with no printk warnings.
1267 Each write syscall must fully contain the sysctl value to be
1268 written, and multiple writes on the same sysctl file descriptor
1269 will rewrite the sysctl value, regardless of file position.
1270 0 Same behavior as above, but warn about processes that perform writes
1271 to a sysctl file descriptor when the file position is not 0.
1272 1 (default) Respect file position when writing sysctl strings. Multiple
1273 writes will append to the sysctl value buffer. Anything past the max
1274 length of the sysctl value buffer will be ignored. Writes to numeric
1275 sysctl entries must always be at file position 0 and the value must
1276 be fully contained in the buffer sent in the write syscall.
1277 == ======================================================================
1280 softlockup_all_cpu_backtrace
1281 ============================
1283 This value controls the soft lockup detector thread's behavior
1284 when a soft lockup condition is detected as to whether or not
1285 to gather further debug information. If enabled, each cpu will
1286 be issued an NMI and instructed to capture stack trace.
1288 This feature is only applicable for architectures which support
1291 = ============================================
1292 0 Do nothing. This is the default behavior.
1293 1 On detection capture more debug information.
1294 = ============================================
1300 This parameter can be used to control whether the kernel panics
1301 when a soft lockup is detected.
1303 = ============================================
1304 0 Don't panic on soft lockup.
1305 1 Panic on soft lockup.
1306 = ============================================
1308 This can also be set using the softlockup_panic kernel parameter.
1314 This parameter can be used to control the soft lockup detector.
1316 = =================================
1317 0 Disable the soft lockup detector.
1318 1 Enable the soft lockup detector.
1319 = =================================
1321 The soft lockup detector monitors CPUs for threads that are hogging the CPUs
1322 without rescheduling voluntarily, and thus prevent the 'migration/N' threads
1323 from running, causing the watchdog work fail to execute. The mechanism depends
1324 on the CPUs ability to respond to timer interrupts which are needed for the
1325 watchdog work to be queued by the watchdog timer function, otherwise the NMI
1326 watchdog — if enabled — can detect a hard lockup condition.
1329 split_lock_mitigate (x86 only)
1330 ==============================
1332 On x86, each "split lock" imposes a system-wide performance penalty. On larger
1333 systems, large numbers of split locks from unprivileged users can result in
1334 denials of service to well-behaved and potentially more important users.
1336 The kernel mitigates these bad users by detecting split locks and imposing
1337 penalties: forcing them to wait and only allowing one core to execute split
1340 These mitigations can make those bad applications unbearably slow. Setting
1341 split_lock_mitigate=0 may restore some application performance, but will also
1342 increase system exposure to denial of service attacks from split lock users.
1344 = ===================================================================
1345 0 Disable the mitigation mode - just warns the split lock on kernel log
1346 and exposes the system to denials of service from the split lockers.
1347 1 Enable the mitigation mode (this is the default) - penalizes the split
1348 lockers with intentional performance degradation.
1349 = ===================================================================
1355 This parameter can be used to control kernel stack erasing at the end
1356 of syscalls for kernels built with ``CONFIG_GCC_PLUGIN_STACKLEAK``.
1358 That erasing reduces the information which kernel stack leak bugs
1359 can reveal and blocks some uninitialized stack variable attacks.
1360 The tradeoff is the performance impact: on a single CPU system kernel
1361 compilation sees a 1% slowdown, other systems and workloads may vary.
1363 = ====================================================================
1364 0 Kernel stack erasing is disabled, STACKLEAK_METRICS are not updated.
1365 1 Kernel stack erasing is enabled (default), it is performed before
1366 returning to the userspace at the end of syscalls.
1367 = ====================================================================
1375 = ====================================
1376 0 Stop-A has no effect.
1377 1 Stop-A breaks to the PROM (default).
1378 = ====================================
1380 Stop-A is always enabled on a panic, so that the user can return to
1387 See Documentation/admin-guide/sysrq.rst.
1393 Non-zero if the kernel has been tainted. Numeric values, which can be
1394 ORed together. The letters are seen in "Tainted" line of Oops reports.
1396 ====== ===== ==============================================================
1397 1 `(P)` proprietary module was loaded
1398 2 `(F)` module was force loaded
1399 4 `(S)` kernel running on an out of specification system
1400 8 `(R)` module was force unloaded
1401 16 `(M)` processor reported a Machine Check Exception (MCE)
1402 32 `(B)` bad page referenced or some unexpected page flags
1403 64 `(U)` taint requested by userspace application
1404 128 `(D)` kernel died recently, i.e. there was an OOPS or BUG
1405 256 `(A)` an ACPI table was overridden by user
1406 512 `(W)` kernel issued warning
1407 1024 `(C)` staging driver was loaded
1408 2048 `(I)` workaround for bug in platform firmware applied
1409 4096 `(O)` externally-built ("out-of-tree") module was loaded
1410 8192 `(E)` unsigned module was loaded
1411 16384 `(L)` soft lockup occurred
1412 32768 `(K)` kernel has been live patched
1413 65536 `(X)` Auxiliary taint, defined and used by for distros
1414 131072 `(T)` The kernel was built with the struct randomization plugin
1415 ====== ===== ==============================================================
1417 See Documentation/admin-guide/tainted-kernels.rst for more information.
1420 writes to this sysctl interface will fail with ``EINVAL`` if the kernel is
1421 booted with the command line option ``panic_on_taint=<bitmask>,nousertaint``
1422 and any of the ORed together values being written to ``tainted`` match with
1423 the bitmask declared on panic_on_taint.
1424 See Documentation/admin-guide/kernel-parameters.rst for more details on
1425 that particular kernel command line option and its optional
1426 ``nousertaint`` switch.
1431 This value controls the maximum number of threads that can be created
1434 During initialization the kernel sets this value such that even if the
1435 maximum number of threads is created, the thread structures occupy only
1436 a part (1/8th) of the available RAM pages.
1438 The minimum value that can be written to ``threads-max`` is 1.
1440 The maximum value that can be written to ``threads-max`` is given by the
1441 constant ``FUTEX_TID_MASK`` (0x3fffffff).
1443 If a value outside of this range is written to ``threads-max`` an
1444 ``EINVAL`` error occurs.
1450 When set, disables tracing (see Documentation/trace/ftrace.rst) when a
1457 When tracepoints are sent to printk() (enabled by the ``tp_printk``
1458 boot parameter), this entry provides runtime control::
1460 echo 0 > /proc/sys/kernel/tracepoint_printk
1462 will stop tracepoints from being sent to printk(), and::
1464 echo 1 > /proc/sys/kernel/tracepoint_printk
1466 will send them to printk() again.
1468 This only works if the kernel was booted with ``tp_printk`` enabled.
1470 See Documentation/admin-guide/kernel-parameters.rst and
1471 Documentation/trace/boottime-trace.rst.
1474 .. _unaligned-dump-stack:
1476 unaligned-dump-stack (ia64)
1477 ===========================
1479 When logging unaligned accesses, controls whether the stack is
1482 = ===================================================
1483 0 Do not dump the stack. This is the default setting.
1485 = ===================================================
1487 See also `ignore-unaligned-usertrap`_.
1493 On architectures where unaligned accesses cause traps, and where this
1494 feature is supported (``CONFIG_SYSCTL_ARCH_UNALIGN_ALLOW``; currently,
1495 ``arc`` and ``parisc``), controls whether unaligned traps are caught
1496 and emulated (instead of failing).
1498 = ========================================================
1499 0 Do not emulate unaligned accesses.
1500 1 Emulate unaligned accesses. This is the default setting.
1501 = ========================================================
1503 See also `ignore-unaligned-usertrap`_.
1509 The value in this file affects behavior of handling NMI. When the
1510 value is non-zero, unknown NMI is trapped and then panic occurs. At
1511 that time, kernel debugging information is displayed on console.
1513 NMI switch that most IA32 servers have fires unknown NMI up, for
1514 example. If a system hangs up, try pressing the NMI switch.
1517 unprivileged_bpf_disabled
1518 =========================
1520 Writing 1 to this entry will disable unprivileged calls to ``bpf()``;
1521 once disabled, calling ``bpf()`` without ``CAP_SYS_ADMIN`` or ``CAP_BPF``
1522 will return ``-EPERM``. Once set to 1, this can't be cleared from the
1523 running kernel anymore.
1525 Writing 2 to this entry will also disable unprivileged calls to ``bpf()``,
1526 however, an admin can still change this setting later on, if needed, by
1527 writing 0 or 1 to this entry.
1529 If ``BPF_UNPRIV_DEFAULT_OFF`` is enabled in the kernel config, then this
1530 entry will default to 2 instead of 0.
1532 = =============================================================
1533 0 Unprivileged calls to ``bpf()`` are enabled
1534 1 Unprivileged calls to ``bpf()`` are disabled without recovery
1535 2 Unprivileged calls to ``bpf()`` are disabled
1536 = =============================================================
1542 Number of kernel warnings after which the kernel should panic when
1543 ``panic_on_warn`` is not set. Setting this to 0 disables checking
1544 the warning count. Setting this to 1 has the same effect as setting
1545 ``panic_on_warn=1``. The default value is 0.
1551 This parameter can be used to disable or enable the soft lockup detector
1552 *and* the NMI watchdog (i.e. the hard lockup detector) at the same time.
1554 = ==============================
1555 0 Disable both lockup detectors.
1556 1 Enable both lockup detectors.
1557 = ==============================
1559 The soft lockup detector and the NMI watchdog can also be disabled or
1560 enabled individually, using the ``soft_watchdog`` and ``nmi_watchdog``
1562 If the ``watchdog`` parameter is read, for example by executing::
1564 cat /proc/sys/kernel/watchdog
1566 the output of this command (0 or 1) shows the logical OR of
1567 ``soft_watchdog`` and ``nmi_watchdog``.
1573 This value can be used to control on which cpus the watchdog may run.
1574 The default cpumask is all possible cores, but if ``NO_HZ_FULL`` is
1575 enabled in the kernel config, and cores are specified with the
1576 ``nohz_full=`` boot argument, those cores are excluded by default.
1577 Offline cores can be included in this mask, and if the core is later
1578 brought online, the watchdog will be started based on the mask value.
1580 Typically this value would only be touched in the ``nohz_full`` case
1581 to re-enable cores that by default were not running the watchdog,
1582 if a kernel lockup was suspected on those cores.
1584 The argument value is the standard cpulist format for cpumasks,
1585 so for example to enable the watchdog on cores 0, 2, 3, and 4 you
1588 echo 0,2-4 > /proc/sys/kernel/watchdog_cpumask
1594 This value can be used to control the frequency of hrtimer and NMI
1595 events and the soft and hard lockup thresholds. The default threshold
1598 The softlockup threshold is (``2 * watchdog_thresh``). Setting this
1599 tunable to zero will disable lockup detection altogether.