1 # SPDX-License-Identifier: GPL-2.0-only
6 default "/lib/modules/$(shell,uname -r)/.config"
7 default "/etc/kernel-config"
8 default "/boot/config-$(shell,uname -r)"
9 default "arch/$(SRCARCH)/configs/$(KBUILD_DEFCONFIG)"
12 def_bool $(success,$(CC) --version | head -n 1 | grep -q gcc)
16 default $(shell,$(srctree)/scripts/gcc-version.sh $(CC)) if CC_IS_GCC
21 default $(shell,$(LD) --version | $(srctree)/scripts/ld-version.sh)
24 def_bool $(success,$(CC) --version | head -n 1 | grep -q clang)
27 def_bool $(success,$(LD) -v | head -n 1 | grep -q LLD)
31 default $(shell,$(srctree)/scripts/clang-version.sh $(CC))
34 def_bool $(success,$(srctree)/scripts/cc-can-link.sh $(CC))
36 config CC_HAS_ASM_GOTO
37 def_bool $(success,$(srctree)/scripts/gcc-goto.sh $(CC))
39 config TOOLS_SUPPORT_RELR
40 def_bool $(success,env "CC=$(CC)" "LD=$(LD)" "NM=$(NM)" "OBJCOPY=$(OBJCOPY)" $(srctree)/scripts/tools-support-relr.sh)
42 config CC_HAS_ASM_INLINE
43 def_bool $(success,echo 'void foo(void) { asm inline (""); }' | $(CC) -x c - -c -o /dev/null)
52 config BUILDTIME_TABLE_SORT
55 config THREAD_INFO_IN_TASK
58 Select this to move thread_info off the stack into task_struct. To
59 make this work, an arch will need to remove all thread_info fields
60 except flags and fix any runtime bugs.
62 One subtle change that will be needed is to use try_get_task_stack()
63 and put_task_stack() in save_thread_stack_tsk() and get_wchan().
72 depends on BROKEN || !SMP
75 config INIT_ENV_ARG_LIMIT
80 Maximum of each of the number of arguments and environment
81 variables passed to init from the kernel command line.
84 bool "Compile also drivers which will not load"
88 Some drivers can be compiled on a different platform than they are
89 intended to be run on. Despite they cannot be loaded there (or even
90 when they load they cannot be used due to missing HW support),
91 developers still, opposing to distributors, might want to build such
92 drivers to compile-test them.
94 If you are a developer and want to build everything available, say Y
95 here. If you are a user/distributor, say N here to exclude useless
96 drivers to be distributed.
98 config UAPI_HEADER_TEST
99 bool "Compile test UAPI headers"
100 depends on HEADERS_INSTALL && CC_CAN_LINK
102 Compile test headers exported to user-space to ensure they are
103 self-contained, i.e. compilable as standalone units.
105 If you are a developer or tester and want to ensure the exported
106 headers are self-contained, say Y here. Otherwise, choose N.
109 string "Local version - append to kernel release"
111 Append an extra string to the end of your kernel version.
112 This will show up when you type uname, for example.
113 The string you set here will be appended after the contents of
114 any files with a filename matching localversion* in your
115 object and source tree, in that order. Your total string can
116 be a maximum of 64 characters.
118 config LOCALVERSION_AUTO
119 bool "Automatically append version information to the version string"
121 depends on !COMPILE_TEST
123 This will try to automatically determine if the current tree is a
124 release tree by looking for git tags that belong to the current
125 top of tree revision.
127 A string of the format -gxxxxxxxx will be added to the localversion
128 if a git-based tree is found. The string generated by this will be
129 appended after any matching localversion* files, and after the value
130 set in CONFIG_LOCALVERSION.
132 (The actual string used here is the first eight characters produced
133 by running the command:
135 $ git rev-parse --verify HEAD
137 which is done within the script "scripts/setlocalversion".)
140 string "Build ID Salt"
143 The build ID is used to link binaries and their debug info. Setting
144 this option will use the value in the calculation of the build id.
145 This is mostly useful for distributions which want to ensure the
146 build is unique between builds. It's safe to leave the default.
148 config HAVE_KERNEL_GZIP
151 config HAVE_KERNEL_BZIP2
154 config HAVE_KERNEL_LZMA
157 config HAVE_KERNEL_XZ
160 config HAVE_KERNEL_LZO
163 config HAVE_KERNEL_LZ4
166 config HAVE_KERNEL_UNCOMPRESSED
170 prompt "Kernel compression mode"
172 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_UNCOMPRESSED
174 The linux kernel is a kind of self-extracting executable.
175 Several compression algorithms are available, which differ
176 in efficiency, compression and decompression speed.
177 Compression speed is only relevant when building a kernel.
178 Decompression speed is relevant at each boot.
180 If you have any problems with bzip2 or lzma compressed
181 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
182 version of this functionality (bzip2 only), for 2.4, was
183 supplied by Christian Ludwig)
185 High compression options are mostly useful for users, who
186 are low on disk space (embedded systems), but for whom ram
189 If in doubt, select 'gzip'
193 depends on HAVE_KERNEL_GZIP
195 The old and tried gzip compression. It provides a good balance
196 between compression ratio and decompression speed.
200 depends on HAVE_KERNEL_BZIP2
202 Its compression ratio and speed is intermediate.
203 Decompression speed is slowest among the choices. The kernel
204 size is about 10% smaller with bzip2, in comparison to gzip.
205 Bzip2 uses a large amount of memory. For modern kernels you
206 will need at least 8MB RAM or more for booting.
210 depends on HAVE_KERNEL_LZMA
212 This compression algorithm's ratio is best. Decompression speed
213 is between gzip and bzip2. Compression is slowest.
214 The kernel size is about 33% smaller with LZMA in comparison to gzip.
218 depends on HAVE_KERNEL_XZ
220 XZ uses the LZMA2 algorithm and instruction set specific
221 BCJ filters which can improve compression ratio of executable
222 code. The size of the kernel is about 30% smaller with XZ in
223 comparison to gzip. On architectures for which there is a BCJ
224 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
225 will create a few percent smaller kernel than plain LZMA.
227 The speed is about the same as with LZMA: The decompression
228 speed of XZ is better than that of bzip2 but worse than gzip
229 and LZO. Compression is slow.
233 depends on HAVE_KERNEL_LZO
235 Its compression ratio is the poorest among the choices. The kernel
236 size is about 10% bigger than gzip; however its speed
237 (both compression and decompression) is the fastest.
241 depends on HAVE_KERNEL_LZ4
243 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
244 A preliminary version of LZ4 de/compression tool is available at
245 <https://code.google.com/p/lz4/>.
247 Its compression ratio is worse than LZO. The size of the kernel
248 is about 8% bigger than LZO. But the decompression speed is
251 config KERNEL_UNCOMPRESSED
253 depends on HAVE_KERNEL_UNCOMPRESSED
255 Produce uncompressed kernel image. This option is usually not what
256 you want. It is useful for debugging the kernel in slow simulation
257 environments, where decompressing and moving the kernel is awfully
258 slow. This option allows early boot code to skip the decompressor
259 and jump right at uncompressed kernel image.
263 config DEFAULT_HOSTNAME
264 string "Default hostname"
267 This option determines the default system hostname before userspace
268 calls sethostname(2). The kernel traditionally uses "(none)" here,
269 but you may wish to use a different default here to make a minimal
270 system more usable with less configuration.
273 # For some reason microblaze and nios2 hard code SWAP=n. Hopefully we can
274 # add proper SWAP support to them, in which case this can be remove.
280 bool "Support for paging of anonymous memory (swap)"
281 depends on MMU && BLOCK && !ARCH_NO_SWAP
284 This option allows you to choose whether you want to have support
285 for so called swap devices or swap files in your kernel that are
286 used to provide more virtual memory than the actual RAM present
287 in your computer. If unsure say Y.
292 Inter Process Communication is a suite of library functions and
293 system calls which let processes (running programs) synchronize and
294 exchange information. It is generally considered to be a good thing,
295 and some programs won't run unless you say Y here. In particular, if
296 you want to run the DOS emulator dosemu under Linux (read the
297 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
298 you'll need to say Y here.
300 You can find documentation about IPC with "info ipc" and also in
301 section 6.4 of the Linux Programmer's Guide, available from
302 <http://www.tldp.org/guides.html>.
304 config SYSVIPC_SYSCTL
311 bool "POSIX Message Queues"
314 POSIX variant of message queues is a part of IPC. In POSIX message
315 queues every message has a priority which decides about succession
316 of receiving it by a process. If you want to compile and run
317 programs written e.g. for Solaris with use of its POSIX message
318 queues (functions mq_*) say Y here.
320 POSIX message queues are visible as a filesystem called 'mqueue'
321 and can be mounted somewhere if you want to do filesystem
322 operations on message queues.
326 config POSIX_MQUEUE_SYSCTL
328 depends on POSIX_MQUEUE
332 config CROSS_MEMORY_ATTACH
333 bool "Enable process_vm_readv/writev syscalls"
337 Enabling this option adds the system calls process_vm_readv and
338 process_vm_writev which allow a process with the correct privileges
339 to directly read from or write to another process' address space.
340 See the man page for more details.
343 bool "uselib syscall"
344 def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION
346 This option enables the uselib syscall, a system call used in the
347 dynamic linker from libc5 and earlier. glibc does not use this
348 system call. If you intend to run programs built on libc5 or
349 earlier, you may need to enable this syscall. Current systems
350 running glibc can safely disable this.
353 bool "Auditing support"
356 Enable auditing infrastructure that can be used with another
357 kernel subsystem, such as SELinux (which requires this for
358 logging of avc messages output). System call auditing is included
359 on architectures which support it.
361 config HAVE_ARCH_AUDITSYSCALL
366 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL
369 source "kernel/irq/Kconfig"
370 source "kernel/time/Kconfig"
371 source "kernel/Kconfig.preempt"
373 menu "CPU/Task time and stats accounting"
375 config VIRT_CPU_ACCOUNTING
379 prompt "Cputime accounting"
380 default TICK_CPU_ACCOUNTING if !PPC64
381 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64
383 # Kind of a stub config for the pure tick based cputime accounting
384 config TICK_CPU_ACCOUNTING
385 bool "Simple tick based cputime accounting"
386 depends on !S390 && !NO_HZ_FULL
388 This is the basic tick based cputime accounting that maintains
389 statistics about user, system and idle time spent on per jiffies
394 config VIRT_CPU_ACCOUNTING_NATIVE
395 bool "Deterministic task and CPU time accounting"
396 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL
397 select VIRT_CPU_ACCOUNTING
399 Select this option to enable more accurate task and CPU time
400 accounting. This is done by reading a CPU counter on each
401 kernel entry and exit and on transitions within the kernel
402 between system, softirq and hardirq state, so there is a
403 small performance impact. In the case of s390 or IBM POWER > 5,
404 this also enables accounting of stolen time on logically-partitioned
407 config VIRT_CPU_ACCOUNTING_GEN
408 bool "Full dynticks CPU time accounting"
409 depends on HAVE_CONTEXT_TRACKING
410 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN
411 depends on GENERIC_CLOCKEVENTS
412 select VIRT_CPU_ACCOUNTING
413 select CONTEXT_TRACKING
415 Select this option to enable task and CPU time accounting on full
416 dynticks systems. This accounting is implemented by watching every
417 kernel-user boundaries using the context tracking subsystem.
418 The accounting is thus performed at the expense of some significant
421 For now this is only useful if you are working on the full
422 dynticks subsystem development.
428 config IRQ_TIME_ACCOUNTING
429 bool "Fine granularity task level IRQ time accounting"
430 depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE
432 Select this option to enable fine granularity task irq time
433 accounting. This is done by reading a timestamp on each
434 transitions between softirq and hardirq state, so there can be a
435 small performance impact.
437 If in doubt, say N here.
439 config HAVE_SCHED_AVG_IRQ
441 depends on IRQ_TIME_ACCOUNTING || PARAVIRT_TIME_ACCOUNTING
444 config SCHED_THERMAL_PRESSURE
445 bool "Enable periodic averaging of thermal pressure"
448 config BSD_PROCESS_ACCT
449 bool "BSD Process Accounting"
452 If you say Y here, a user level program will be able to instruct the
453 kernel (via a special system call) to write process accounting
454 information to a file: whenever a process exits, information about
455 that process will be appended to the file by the kernel. The
456 information includes things such as creation time, owning user,
457 command name, memory usage, controlling terminal etc. (the complete
458 list is in the struct acct in <file:include/linux/acct.h>). It is
459 up to the user level program to do useful things with this
460 information. This is generally a good idea, so say Y.
462 config BSD_PROCESS_ACCT_V3
463 bool "BSD Process Accounting version 3 file format"
464 depends on BSD_PROCESS_ACCT
467 If you say Y here, the process accounting information is written
468 in a new file format that also logs the process IDs of each
469 process and its parent. Note that this file format is incompatible
470 with previous v0/v1/v2 file formats, so you will need updated tools
471 for processing it. A preliminary version of these tools is available
472 at <http://www.gnu.org/software/acct/>.
475 bool "Export task/process statistics through netlink"
480 Export selected statistics for tasks/processes through the
481 generic netlink interface. Unlike BSD process accounting, the
482 statistics are available during the lifetime of tasks/processes as
483 responses to commands. Like BSD accounting, they are sent to user
488 config TASK_DELAY_ACCT
489 bool "Enable per-task delay accounting"
493 Collect information on time spent by a task waiting for system
494 resources like cpu, synchronous block I/O completion and swapping
495 in pages. Such statistics can help in setting a task's priorities
496 relative to other tasks for cpu, io, rss limits etc.
501 bool "Enable extended accounting over taskstats"
504 Collect extended task accounting data and send the data
505 to userland for processing over the taskstats interface.
509 config TASK_IO_ACCOUNTING
510 bool "Enable per-task storage I/O accounting"
511 depends on TASK_XACCT
513 Collect information on the number of bytes of storage I/O which this
519 bool "Pressure stall information tracking"
521 Collect metrics that indicate how overcommitted the CPU, memory,
522 and IO capacity are in the system.
524 If you say Y here, the kernel will create /proc/pressure/ with the
525 pressure statistics files cpu, memory, and io. These will indicate
526 the share of walltime in which some or all tasks in the system are
527 delayed due to contention of the respective resource.
529 In kernels with cgroup support, cgroups (cgroup2 only) will
530 have cpu.pressure, memory.pressure, and io.pressure files,
531 which aggregate pressure stalls for the grouped tasks only.
533 For more details see Documentation/accounting/psi.rst.
537 config PSI_DEFAULT_DISABLED
538 bool "Require boot parameter to enable pressure stall information tracking"
542 If set, pressure stall information tracking will be disabled
543 per default but can be enabled through passing psi=1 on the
544 kernel commandline during boot.
546 This feature adds some code to the task wakeup and sleep
547 paths of the scheduler. The overhead is too low to affect
548 common scheduling-intense workloads in practice (such as
549 webservers, memcache), but it does show up in artificial
550 scheduler stress tests, such as hackbench.
552 If you are paranoid and not sure what the kernel will be
557 endmenu # "CPU/Task time and stats accounting"
561 depends on SMP || COMPILE_TEST
564 Make sure that CPUs running critical tasks are not disturbed by
565 any source of "noise" such as unbound workqueues, timers, kthreads...
566 Unbound jobs get offloaded to housekeeping CPUs. This is driven by
567 the "isolcpus=" boot parameter.
571 source "kernel/rcu/Kconfig"
578 tristate "Kernel .config support"
580 This option enables the complete Linux kernel ".config" file
581 contents to be saved in the kernel. It provides documentation
582 of which kernel options are used in a running kernel or in an
583 on-disk kernel. This information can be extracted from the kernel
584 image file with the script scripts/extract-ikconfig and used as
585 input to rebuild the current kernel or to build another kernel.
586 It can also be extracted from a running kernel by reading
587 /proc/config.gz if enabled (below).
590 bool "Enable access to .config through /proc/config.gz"
591 depends on IKCONFIG && PROC_FS
593 This option enables access to the kernel configuration file
594 through /proc/config.gz.
597 tristate "Enable kernel headers through /sys/kernel/kheaders.tar.xz"
600 This option enables access to the in-kernel headers that are generated during
601 the build process. These can be used to build eBPF tracing programs,
602 or similar programs. If you build the headers as a module, a module called
603 kheaders.ko is built which can be loaded on-demand to get access to headers.
606 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
611 Select the minimal kernel log buffer size as a power of 2.
612 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config
613 parameter, see below. Any higher size also might be forced
614 by "log_buf_len" boot parameter.
624 config LOG_CPU_MAX_BUF_SHIFT
625 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)"
628 default 12 if !BASE_SMALL
629 default 0 if BASE_SMALL
632 This option allows to increase the default ring buffer size
633 according to the number of CPUs. The value defines the contribution
634 of each CPU as a power of 2. The used space is typically only few
635 lines however it might be much more when problems are reported,
638 The increased size means that a new buffer has to be allocated and
639 the original static one is unused. It makes sense only on systems
640 with more CPUs. Therefore this value is used only when the sum of
641 contributions is greater than the half of the default kernel ring
642 buffer as defined by LOG_BUF_SHIFT. The default values are set
643 so that more than 64 CPUs are needed to trigger the allocation.
645 Also this option is ignored when "log_buf_len" kernel parameter is
646 used as it forces an exact (power of two) size of the ring buffer.
648 The number of possible CPUs is used for this computation ignoring
649 hotplugging making the computation optimal for the worst case
650 scenario while allowing a simple algorithm to be used from bootup.
652 Examples shift values and their meaning:
653 17 => 128 KB for each CPU
654 16 => 64 KB for each CPU
655 15 => 32 KB for each CPU
656 14 => 16 KB for each CPU
657 13 => 8 KB for each CPU
658 12 => 4 KB for each CPU
660 config PRINTK_SAFE_LOG_BUF_SHIFT
661 int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)"
666 Select the size of an alternate printk per-CPU buffer where messages
667 printed from usafe contexts are temporary stored. One example would
668 be NMI messages, another one - printk recursion. The messages are
669 copied to the main log buffer in a safe context to avoid a deadlock.
670 The value defines the size as a power of 2.
672 Those messages are rare and limited. The largest one is when
673 a backtrace is printed. It usually fits into 4KB. Select
674 8KB if you want to be on the safe side.
677 17 => 128 KB for each CPU
678 16 => 64 KB for each CPU
679 15 => 32 KB for each CPU
680 14 => 16 KB for each CPU
681 13 => 8 KB for each CPU
682 12 => 4 KB for each CPU
685 # Architectures with an unreliable sched_clock() should select this:
687 config HAVE_UNSTABLE_SCHED_CLOCK
690 config GENERIC_SCHED_CLOCK
693 menu "Scheduler features"
696 bool "Enable utilization clamping for RT/FAIR tasks"
697 depends on CPU_FREQ_GOV_SCHEDUTIL
699 This feature enables the scheduler to track the clamped utilization
700 of each CPU based on RUNNABLE tasks scheduled on that CPU.
702 With this option, the user can specify the min and max CPU
703 utilization allowed for RUNNABLE tasks. The max utilization defines
704 the maximum frequency a task should use while the min utilization
705 defines the minimum frequency it should use.
707 Both min and max utilization clamp values are hints to the scheduler,
708 aiming at improving its frequency selection policy, but they do not
709 enforce or grant any specific bandwidth for tasks.
713 config UCLAMP_BUCKETS_COUNT
714 int "Number of supported utilization clamp buckets"
717 depends on UCLAMP_TASK
719 Defines the number of clamp buckets to use. The range of each bucket
720 will be SCHED_CAPACITY_SCALE/UCLAMP_BUCKETS_COUNT. The higher the
721 number of clamp buckets the finer their granularity and the higher
722 the precision of clamping aggregation and tracking at run-time.
724 For example, with the minimum configuration value we will have 5
725 clamp buckets tracking 20% utilization each. A 25% boosted tasks will
726 be refcounted in the [20..39]% bucket and will set the bucket clamp
727 effective value to 25%.
728 If a second 30% boosted task should be co-scheduled on the same CPU,
729 that task will be refcounted in the same bucket of the first task and
730 it will boost the bucket clamp effective value to 30%.
731 The clamp effective value of a bucket is reset to its nominal value
732 (20% in the example above) when there are no more tasks refcounted in
735 An additional boost/capping margin can be added to some tasks. In the
736 example above the 25% task will be boosted to 30% until it exits the
737 CPU. If that should be considered not acceptable on certain systems,
738 it's always possible to reduce the margin by increasing the number of
739 clamp buckets to trade off used memory for run-time tracking
742 If in doubt, use the default value.
747 # For architectures that want to enable the support for NUMA-affine scheduler
750 config ARCH_SUPPORTS_NUMA_BALANCING
754 # For architectures that prefer to flush all TLBs after a number of pages
755 # are unmapped instead of sending one IPI per page to flush. The architecture
756 # must provide guarantees on what happens if a clean TLB cache entry is
757 # written after the unmap. Details are in mm/rmap.c near the check for
758 # should_defer_flush. The architecture should also consider if the full flush
759 # and the refill costs are offset by the savings of sending fewer IPIs.
760 config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
764 def_bool !$(cc-option,$(m64-flag) -D__SIZEOF_INT128__=0) && 64BIT
767 # For architectures that know their GCC __int128 support is sound
769 config ARCH_SUPPORTS_INT128
772 # For architectures that (ab)use NUMA to represent different memory regions
773 # all cpu-local but of different latencies, such as SuperH.
775 config ARCH_WANT_NUMA_VARIABLE_LOCALITY
778 config NUMA_BALANCING
779 bool "Memory placement aware NUMA scheduler"
780 depends on ARCH_SUPPORTS_NUMA_BALANCING
781 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
782 depends on SMP && NUMA && MIGRATION
784 This option adds support for automatic NUMA aware memory/task placement.
785 The mechanism is quite primitive and is based on migrating memory when
786 it has references to the node the task is running on.
788 This system will be inactive on UMA systems.
790 config NUMA_BALANCING_DEFAULT_ENABLED
791 bool "Automatically enable NUMA aware memory/task placement"
793 depends on NUMA_BALANCING
795 If set, automatic NUMA balancing will be enabled if running on a NUMA
799 bool "Control Group support"
802 This option adds support for grouping sets of processes together, for
803 use with process control subsystems such as Cpusets, CFS, memory
804 controls or device isolation.
806 - Documentation/scheduler/sched-design-CFS.rst (CFS)
807 - Documentation/admin-guide/cgroup-v1/ (features for grouping, isolation
808 and resource control)
818 bool "Memory controller"
822 Provides control over the memory footprint of tasks in a cgroup.
826 depends on MEMCG && SWAP
831 depends on MEMCG && !SLOB
839 Generic block IO controller cgroup interface. This is the common
840 cgroup interface which should be used by various IO controlling
843 Currently, CFQ IO scheduler uses it to recognize task groups and
844 control disk bandwidth allocation (proportional time slice allocation)
845 to such task groups. It is also used by bio throttling logic in
846 block layer to implement upper limit in IO rates on a device.
848 This option only enables generic Block IO controller infrastructure.
849 One needs to also enable actual IO controlling logic/policy. For
850 enabling proportional weight division of disk bandwidth in CFQ, set
851 CONFIG_BFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
852 CONFIG_BLK_DEV_THROTTLING=y.
854 See Documentation/admin-guide/cgroup-v1/blkio-controller.rst for more information.
856 config CGROUP_WRITEBACK
858 depends on MEMCG && BLK_CGROUP
861 menuconfig CGROUP_SCHED
862 bool "CPU controller"
865 This feature lets CPU scheduler recognize task groups and control CPU
866 bandwidth allocation to such task groups. It uses cgroups to group
870 config FAIR_GROUP_SCHED
871 bool "Group scheduling for SCHED_OTHER"
872 depends on CGROUP_SCHED
876 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
877 depends on FAIR_GROUP_SCHED
880 This option allows users to define CPU bandwidth rates (limits) for
881 tasks running within the fair group scheduler. Groups with no limit
882 set are considered to be unconstrained and will run with no
884 See Documentation/scheduler/sched-bwc.rst for more information.
886 config RT_GROUP_SCHED
887 bool "Group scheduling for SCHED_RR/FIFO"
888 depends on CGROUP_SCHED
891 This feature lets you explicitly allocate real CPU bandwidth
892 to task groups. If enabled, it will also make it impossible to
893 schedule realtime tasks for non-root users until you allocate
894 realtime bandwidth for them.
895 See Documentation/scheduler/sched-rt-group.rst for more information.
899 config UCLAMP_TASK_GROUP
900 bool "Utilization clamping per group of tasks"
901 depends on CGROUP_SCHED
902 depends on UCLAMP_TASK
905 This feature enables the scheduler to track the clamped utilization
906 of each CPU based on RUNNABLE tasks currently scheduled on that CPU.
908 When this option is enabled, the user can specify a min and max
909 CPU bandwidth which is allowed for each single task in a group.
910 The max bandwidth allows to clamp the maximum frequency a task
911 can use, while the min bandwidth allows to define a minimum
912 frequency a task will always use.
914 When task group based utilization clamping is enabled, an eventually
915 specified task-specific clamp value is constrained by the cgroup
916 specified clamp value. Both minimum and maximum task clamping cannot
917 be bigger than the corresponding clamping defined at task group level.
922 bool "PIDs controller"
924 Provides enforcement of process number limits in the scope of a
925 cgroup. Any attempt to fork more processes than is allowed in the
926 cgroup will fail. PIDs are fundamentally a global resource because it
927 is fairly trivial to reach PID exhaustion before you reach even a
928 conservative kmemcg limit. As a result, it is possible to grind a
929 system to halt without being limited by other cgroup policies. The
930 PIDs controller is designed to stop this from happening.
932 It should be noted that organisational operations (such as attaching
933 to a cgroup hierarchy) will *not* be blocked by the PIDs controller,
934 since the PIDs limit only affects a process's ability to fork, not to
938 bool "RDMA controller"
940 Provides enforcement of RDMA resources defined by IB stack.
941 It is fairly easy for consumers to exhaust RDMA resources, which
942 can result into resource unavailability to other consumers.
943 RDMA controller is designed to stop this from happening.
944 Attaching processes with active RDMA resources to the cgroup
945 hierarchy is allowed even if can cross the hierarchy's limit.
947 config CGROUP_FREEZER
948 bool "Freezer controller"
950 Provides a way to freeze and unfreeze all tasks in a
953 This option affects the ORIGINAL cgroup interface. The cgroup2 memory
954 controller includes important in-kernel memory consumers per default.
956 If you're using cgroup2, say N.
958 config CGROUP_HUGETLB
959 bool "HugeTLB controller"
960 depends on HUGETLB_PAGE
964 Provides a cgroup controller for HugeTLB pages.
965 When you enable this, you can put a per cgroup limit on HugeTLB usage.
966 The limit is enforced during page fault. Since HugeTLB doesn't
967 support page reclaim, enforcing the limit at page fault time implies
968 that, the application will get SIGBUS signal if it tries to access
969 HugeTLB pages beyond its limit. This requires the application to know
970 beforehand how much HugeTLB pages it would require for its use. The
971 control group is tracked in the third page lru pointer. This means
972 that we cannot use the controller with huge page less than 3 pages.
975 bool "Cpuset controller"
978 This option will let you create and manage CPUSETs which
979 allow dynamically partitioning a system into sets of CPUs and
980 Memory Nodes and assigning tasks to run only within those sets.
981 This is primarily useful on large SMP or NUMA systems.
985 config PROC_PID_CPUSET
986 bool "Include legacy /proc/<pid>/cpuset file"
991 bool "Device controller"
993 Provides a cgroup controller implementing whitelists for
994 devices which a process in the cgroup can mknod or open.
996 config CGROUP_CPUACCT
997 bool "Simple CPU accounting controller"
999 Provides a simple controller for monitoring the
1000 total CPU consumed by the tasks in a cgroup.
1003 bool "Perf controller"
1004 depends on PERF_EVENTS
1006 This option extends the perf per-cpu mode to restrict monitoring
1007 to threads which belong to the cgroup specified and run on the
1008 designated cpu. Or this can be used to have cgroup ID in samples
1009 so that it can monitor performance events among cgroups.
1014 bool "Support for eBPF programs attached to cgroups"
1015 depends on BPF_SYSCALL
1016 select SOCK_CGROUP_DATA
1018 Allow attaching eBPF programs to a cgroup using the bpf(2)
1019 syscall command BPF_PROG_ATTACH.
1021 In which context these programs are accessed depends on the type
1022 of attachment. For instance, programs that are attached using
1023 BPF_CGROUP_INET_INGRESS will be executed on the ingress path of
1027 bool "Debug controller"
1029 depends on DEBUG_KERNEL
1031 This option enables a simple controller that exports
1032 debugging information about the cgroups framework. This
1033 controller is for control cgroup debugging only. Its
1034 interfaces are not stable.
1038 config SOCK_CGROUP_DATA
1044 menuconfig NAMESPACES
1045 bool "Namespaces support" if EXPERT
1046 depends on MULTIUSER
1049 Provides the way to make tasks work with different objects using
1050 the same id. For example same IPC id may refer to different objects
1051 or same user id or pid may refer to different tasks when used in
1052 different namespaces.
1057 bool "UTS namespace"
1060 In this namespace tasks see different info provided with the
1064 bool "TIME namespace"
1065 depends on GENERIC_VDSO_TIME_NS
1068 In this namespace boottime and monotonic clocks can be set.
1069 The time will keep going with the same pace.
1072 bool "IPC namespace"
1073 depends on (SYSVIPC || POSIX_MQUEUE)
1076 In this namespace tasks work with IPC ids which correspond to
1077 different IPC objects in different namespaces.
1080 bool "User namespace"
1083 This allows containers, i.e. vservers, to use user namespaces
1084 to provide different user info for different servers.
1086 When user namespaces are enabled in the kernel it is
1087 recommended that the MEMCG option also be enabled and that
1088 user-space use the memory control groups to limit the amount
1089 of memory a memory unprivileged users can use.
1094 bool "PID Namespaces"
1097 Support process id namespaces. This allows having multiple
1098 processes with the same pid as long as they are in different
1099 pid namespaces. This is a building block of containers.
1102 bool "Network namespace"
1106 Allow user space to create what appear to be multiple instances
1107 of the network stack.
1111 config CHECKPOINT_RESTORE
1112 bool "Checkpoint/restore support"
1113 select PROC_CHILDREN
1116 Enables additional kernel features in a sake of checkpoint/restore.
1117 In particular it adds auxiliary prctl codes to setup process text,
1118 data and heap segment sizes, and a few additional /proc filesystem
1121 If unsure, say N here.
1123 config SCHED_AUTOGROUP
1124 bool "Automatic process group scheduling"
1127 select FAIR_GROUP_SCHED
1129 This option optimizes the scheduler for common desktop workloads by
1130 automatically creating and populating task groups. This separation
1131 of workloads isolates aggressive CPU burners (like build jobs) from
1132 desktop applications. Task group autogeneration is currently based
1135 config SYSFS_DEPRECATED
1136 bool "Enable deprecated sysfs features to support old userspace tools"
1140 This option adds code that switches the layout of the "block" class
1141 devices, to not show up in /sys/class/block/, but only in
1144 This switch is only active when the sysfs.deprecated=1 boot option is
1145 passed or the SYSFS_DEPRECATED_V2 option is set.
1147 This option allows new kernels to run on old distributions and tools,
1148 which might get confused by /sys/class/block/. Since 2007/2008 all
1149 major distributions and tools handle this just fine.
1151 Recent distributions and userspace tools after 2009/2010 depend on
1152 the existence of /sys/class/block/, and will not work with this
1155 Only if you are using a new kernel on an old distribution, you might
1158 config SYSFS_DEPRECATED_V2
1159 bool "Enable deprecated sysfs features by default"
1162 depends on SYSFS_DEPRECATED
1164 Enable deprecated sysfs by default.
1166 See the CONFIG_SYSFS_DEPRECATED option for more details about this
1169 Only if you are using a new kernel on an old distribution, you might
1170 need to say Y here. Even then, odds are you would not need it
1171 enabled, you can always pass the boot option if absolutely necessary.
1174 bool "Kernel->user space relay support (formerly relayfs)"
1177 This option enables support for relay interface support in
1178 certain file systems (such as debugfs).
1179 It is designed to provide an efficient mechanism for tools and
1180 facilities to relay large amounts of data from kernel space to
1185 config BLK_DEV_INITRD
1186 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
1188 The initial RAM filesystem is a ramfs which is loaded by the
1189 boot loader (loadlin or lilo) and that is mounted as root
1190 before the normal boot procedure. It is typically used to
1191 load modules needed to mount the "real" root file system,
1192 etc. See <file:Documentation/admin-guide/initrd.rst> for details.
1194 If RAM disk support (BLK_DEV_RAM) is also included, this
1195 also enables initial RAM disk (initrd) support and adds
1196 15 Kbytes (more on some other architectures) to the kernel size.
1202 source "usr/Kconfig"
1207 bool "Boot config support"
1208 select BLK_DEV_INITRD
1210 Extra boot config allows system admin to pass a config file as
1211 complemental extension of kernel cmdline when booting.
1212 The boot config file must be attached at the end of initramfs
1213 with checksum, size and magic word.
1214 See <file:Documentation/admin-guide/bootconfig.rst> for details.
1219 prompt "Compiler optimization level"
1220 default CC_OPTIMIZE_FOR_PERFORMANCE
1222 config CC_OPTIMIZE_FOR_PERFORMANCE
1223 bool "Optimize for performance (-O2)"
1225 This is the default optimization level for the kernel, building
1226 with the "-O2" compiler flag for best performance and most
1227 helpful compile-time warnings.
1229 config CC_OPTIMIZE_FOR_PERFORMANCE_O3
1230 bool "Optimize more for performance (-O3)"
1233 Choosing this option will pass "-O3" to your compiler to optimize
1234 the kernel yet more for performance.
1236 config CC_OPTIMIZE_FOR_SIZE
1237 bool "Optimize for size (-Os)"
1239 Choosing this option will pass "-Os" to your compiler resulting
1240 in a smaller kernel.
1244 config HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1247 This requires that the arch annotates or otherwise protects
1248 its external entry points from being discarded. Linker scripts
1249 must also merge .text.*, .data.*, and .bss.* correctly into
1250 output sections. Care must be taken not to pull in unrelated
1251 sections (e.g., '.text.init'). Typically '.' in section names
1252 is used to distinguish them from label names / C identifiers.
1254 config LD_DEAD_CODE_DATA_ELIMINATION
1255 bool "Dead code and data elimination (EXPERIMENTAL)"
1256 depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1258 depends on !(FUNCTION_TRACER && CC_IS_GCC && GCC_VERSION < 40800)
1259 depends on $(cc-option,-ffunction-sections -fdata-sections)
1260 depends on $(ld-option,--gc-sections)
1262 Enable this if you want to do dead code and data elimination with
1263 the linker by compiling with -ffunction-sections -fdata-sections,
1264 and linking with --gc-sections.
1266 This can reduce on disk and in-memory size of the kernel
1267 code and static data, particularly for small configs and
1268 on small systems. This has the possibility of introducing
1269 silently broken kernel if the required annotations are not
1270 present. This option is not well tested yet, so use at your
1279 config SYSCTL_EXCEPTION_TRACE
1282 Enable support for /proc/sys/debug/exception-trace.
1284 config SYSCTL_ARCH_UNALIGN_NO_WARN
1287 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1288 Allows arch to define/use @no_unaligned_warning to possibly warn
1289 about unaligned access emulation going on under the hood.
1291 config SYSCTL_ARCH_UNALIGN_ALLOW
1294 Enable support for /proc/sys/kernel/unaligned-trap
1295 Allows arches to define/use @unaligned_enabled to runtime toggle
1296 the unaligned access emulation.
1297 see arch/parisc/kernel/unaligned.c for reference
1299 config HAVE_PCSPKR_PLATFORM
1302 # interpreter that classic socket filters depend on
1307 bool "Configure standard kernel features (expert users)"
1308 # Unhide debug options, to make the on-by-default options visible
1311 This option allows certain base kernel options and settings
1312 to be disabled or tweaked. This is for specialized
1313 environments which can tolerate a "non-standard" kernel.
1314 Only use this if you really know what you are doing.
1317 bool "Enable 16-bit UID system calls" if EXPERT
1318 depends on HAVE_UID16 && MULTIUSER
1321 This enables the legacy 16-bit UID syscall wrappers.
1324 bool "Multiple users, groups and capabilities support" if EXPERT
1327 This option enables support for non-root users, groups and
1330 If you say N here, all processes will run with UID 0, GID 0, and all
1331 possible capabilities. Saying N here also compiles out support for
1332 system calls related to UIDs, GIDs, and capabilities, such as setuid,
1335 If unsure, say Y here.
1337 config SGETMASK_SYSCALL
1338 bool "sgetmask/ssetmask syscalls support" if EXPERT
1339 def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH
1341 sys_sgetmask and sys_ssetmask are obsolete system calls
1342 no longer supported in libc but still enabled by default in some
1345 If unsure, leave the default option here.
1347 config SYSFS_SYSCALL
1348 bool "Sysfs syscall support" if EXPERT
1351 sys_sysfs is an obsolete system call no longer supported in libc.
1352 Note that disabling this option is more secure but might break
1353 compatibility with some systems.
1355 If unsure say Y here.
1358 bool "open by fhandle syscalls" if EXPERT
1362 If you say Y here, a user level program will be able to map
1363 file names to handle and then later use the handle for
1364 different file system operations. This is useful in implementing
1365 userspace file servers, which now track files using handles instead
1366 of names. The handle would remain the same even if file names
1367 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
1371 bool "Posix Clocks & timers" if EXPERT
1374 This includes native support for POSIX timers to the kernel.
1375 Some embedded systems have no use for them and therefore they
1376 can be configured out to reduce the size of the kernel image.
1378 When this option is disabled, the following syscalls won't be
1379 available: timer_create, timer_gettime: timer_getoverrun,
1380 timer_settime, timer_delete, clock_adjtime, getitimer,
1381 setitimer, alarm. Furthermore, the clock_settime, clock_gettime,
1382 clock_getres and clock_nanosleep syscalls will be limited to
1383 CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only.
1389 bool "Enable support for printk" if EXPERT
1392 This option enables normal printk support. Removing it
1393 eliminates most of the message strings from the kernel image
1394 and makes the kernel more or less silent. As this makes it
1395 very difficult to diagnose system problems, saying N here is
1396 strongly discouraged.
1404 bool "BUG() support" if EXPERT
1407 Disabling this option eliminates support for BUG and WARN, reducing
1408 the size of your kernel image and potentially quietly ignoring
1409 numerous fatal conditions. You should only consider disabling this
1410 option for embedded systems with no facilities for reporting errors.
1416 bool "Enable ELF core dumps" if EXPERT
1418 Enable support for generating core dumps. Disabling saves about 4k.
1421 config PCSPKR_PLATFORM
1422 bool "Enable PC-Speaker support" if EXPERT
1423 depends on HAVE_PCSPKR_PLATFORM
1427 This option allows to disable the internal PC-Speaker
1428 support, saving some memory.
1432 bool "Enable full-sized data structures for core" if EXPERT
1434 Disabling this option reduces the size of miscellaneous core
1435 kernel data structures. This saves memory on small machines,
1436 but may reduce performance.
1439 bool "Enable futex support" if EXPERT
1443 Disabling this option will cause the kernel to be built without
1444 support for "fast userspace mutexes". The resulting kernel may not
1445 run glibc-based applications correctly.
1449 depends on FUTEX && RT_MUTEXES
1452 config HAVE_FUTEX_CMPXCHG
1456 Architectures should select this if futex_atomic_cmpxchg_inatomic()
1457 is implemented and always working. This removes a couple of runtime
1461 bool "Enable eventpoll support" if EXPERT
1464 Disabling this option will cause the kernel to be built without
1465 support for epoll family of system calls.
1468 bool "Enable signalfd() system call" if EXPERT
1471 Enable the signalfd() system call that allows to receive signals
1472 on a file descriptor.
1477 bool "Enable timerfd() system call" if EXPERT
1480 Enable the timerfd() system call that allows to receive timer
1481 events on a file descriptor.
1486 bool "Enable eventfd() system call" if EXPERT
1489 Enable the eventfd() system call that allows to receive both
1490 kernel notification (ie. KAIO) or userspace notifications.
1495 bool "Use full shmem filesystem" if EXPERT
1499 The shmem is an internal filesystem used to manage shared memory.
1500 It is backed by swap and manages resource limits. It is also exported
1501 to userspace as tmpfs if TMPFS is enabled. Disabling this
1502 option replaces shmem and tmpfs with the much simpler ramfs code,
1503 which may be appropriate on small systems without swap.
1506 bool "Enable AIO support" if EXPERT
1509 This option enables POSIX asynchronous I/O which may by used
1510 by some high performance threaded applications. Disabling
1511 this option saves about 7k.
1514 bool "Enable IO uring support" if EXPERT
1518 This option enables support for the io_uring interface, enabling
1519 applications to submit and complete IO through submission and
1520 completion rings that are shared between the kernel and application.
1522 config ADVISE_SYSCALLS
1523 bool "Enable madvise/fadvise syscalls" if EXPERT
1526 This option enables the madvise and fadvise syscalls, used by
1527 applications to advise the kernel about their future memory or file
1528 usage, improving performance. If building an embedded system where no
1529 applications use these syscalls, you can disable this option to save
1532 config HAVE_ARCH_USERFAULTFD_WP
1535 Arch has userfaultfd write protection support
1538 bool "Enable membarrier() system call" if EXPERT
1541 Enable the membarrier() system call that allows issuing memory
1542 barriers across all running threads, which can be used to distribute
1543 the cost of user-space memory barriers asymmetrically by transforming
1544 pairs of memory barriers into pairs consisting of membarrier() and a
1550 bool "Load all symbols for debugging/ksymoops" if EXPERT
1553 Say Y here to let the kernel print out symbolic crash information and
1554 symbolic stack backtraces. This increases the size of the kernel
1555 somewhat, as all symbols have to be loaded into the kernel image.
1558 bool "Include all symbols in kallsyms"
1559 depends on DEBUG_KERNEL && KALLSYMS
1561 Normally kallsyms only contains the symbols of functions for nicer
1562 OOPS messages and backtraces (i.e., symbols from the text and inittext
1563 sections). This is sufficient for most cases. And only in very rare
1564 cases (e.g., when a debugger is used) all symbols are required (e.g.,
1565 names of variables from the data sections, etc).
1567 This option makes sure that all symbols are loaded into the kernel
1568 image (i.e., symbols from all sections) in cost of increased kernel
1569 size (depending on the kernel configuration, it may be 300KiB or
1570 something like this).
1572 Say N unless you really need all symbols.
1574 config KALLSYMS_ABSOLUTE_PERCPU
1577 default X86_64 && SMP
1579 config KALLSYMS_BASE_RELATIVE
1584 Instead of emitting them as absolute values in the native word size,
1585 emit the symbol references in the kallsyms table as 32-bit entries,
1586 each containing a relative value in the range [base, base + U32_MAX]
1587 or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either
1588 an absolute value in the range [0, S32_MAX] or a relative value in the
1589 range [base, base + S32_MAX], where base is the lowest relative symbol
1590 address encountered in the image.
1592 On 64-bit builds, this reduces the size of the address table by 50%,
1593 but more importantly, it results in entries whose values are build
1594 time constants, and no relocation pass is required at runtime to fix
1595 up the entries based on the runtime load address of the kernel.
1597 # end of the "standard kernel features (expert users)" menu
1599 # syscall, maps, verifier
1602 bool "LSM Instrumentation with BPF"
1603 depends on BPF_EVENTS
1604 depends on BPF_SYSCALL
1608 Enables instrumentation of the security hooks with eBPF programs for
1609 implementing dynamic MAC and Audit Policies.
1611 If you are unsure how to answer this question, answer N.
1614 bool "Enable bpf() system call"
1619 Enable the bpf() system call that allows to manipulate eBPF
1620 programs and maps via file descriptors.
1622 config ARCH_WANT_DEFAULT_BPF_JIT
1625 config BPF_JIT_ALWAYS_ON
1626 bool "Permanently enable BPF JIT and remove BPF interpreter"
1627 depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT
1629 Enables BPF JIT and removes BPF interpreter to avoid
1630 speculative execution of BPF instructions by the interpreter
1632 config BPF_JIT_DEFAULT_ON
1633 def_bool ARCH_WANT_DEFAULT_BPF_JIT || BPF_JIT_ALWAYS_ON
1634 depends on HAVE_EBPF_JIT && BPF_JIT
1637 bool "Enable userfaultfd() system call"
1640 Enable the userfaultfd() system call that allows to intercept and
1641 handle page faults in userland.
1643 config ARCH_HAS_MEMBARRIER_CALLBACKS
1646 config ARCH_HAS_MEMBARRIER_SYNC_CORE
1650 bool "Enable rseq() system call" if EXPERT
1652 depends on HAVE_RSEQ
1655 Enable the restartable sequences system call. It provides a
1656 user-space cache for the current CPU number value, which
1657 speeds up getting the current CPU number from user-space,
1658 as well as an ABI to speed up user-space operations on
1665 bool "Enabled debugging of rseq() system call" if EXPERT
1666 depends on RSEQ && DEBUG_KERNEL
1668 Enable extra debugging checks for the rseq system call.
1673 bool "Embedded system"
1674 option allnoconfig_y
1677 This option should be enabled if compiling the kernel for
1678 an embedded system so certain expert options are available
1681 config HAVE_PERF_EVENTS
1684 See tools/perf/design.txt for details.
1686 config PERF_USE_VMALLOC
1689 See tools/perf/design.txt for details
1692 bool "PC/104 support" if EXPERT
1694 Expose PC/104 form factor device drivers and options available for
1695 selection and configuration. Enable this option if your target
1696 machine has a PC/104 bus.
1698 menu "Kernel Performance Events And Counters"
1701 bool "Kernel performance events and counters"
1702 default y if PROFILING
1703 depends on HAVE_PERF_EVENTS
1707 Enable kernel support for various performance events provided
1708 by software and hardware.
1710 Software events are supported either built-in or via the
1711 use of generic tracepoints.
1713 Most modern CPUs support performance events via performance
1714 counter registers. These registers count the number of certain
1715 types of hw events: such as instructions executed, cachemisses
1716 suffered, or branches mis-predicted - without slowing down the
1717 kernel or applications. These registers can also trigger interrupts
1718 when a threshold number of events have passed - and can thus be
1719 used to profile the code that runs on that CPU.
1721 The Linux Performance Event subsystem provides an abstraction of
1722 these software and hardware event capabilities, available via a
1723 system call and used by the "perf" utility in tools/perf/. It
1724 provides per task and per CPU counters, and it provides event
1725 capabilities on top of those.
1729 config DEBUG_PERF_USE_VMALLOC
1731 bool "Debug: use vmalloc to back perf mmap() buffers"
1732 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC
1733 select PERF_USE_VMALLOC
1735 Use vmalloc memory to back perf mmap() buffers.
1737 Mostly useful for debugging the vmalloc code on platforms
1738 that don't require it.
1744 config VM_EVENT_COUNTERS
1746 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1748 VM event counters are needed for event counts to be shown.
1749 This option allows the disabling of the VM event counters
1750 on EXPERT systems. /proc/vmstat will only show page counts
1751 if VM event counters are disabled.
1755 bool "Enable SLUB debugging support" if EXPERT
1756 depends on SLUB && SYSFS
1758 SLUB has extensive debug support features. Disabling these can
1759 result in significant savings in code size. This also disables
1760 SLUB sysfs support. /sys/slab will not exist and there will be
1761 no support for cache validation etc.
1763 config SLUB_MEMCG_SYSFS_ON
1765 bool "Enable memcg SLUB sysfs support by default" if EXPERT
1766 depends on SLUB && SYSFS && MEMCG
1768 SLUB creates a directory under /sys/kernel/slab for each
1769 allocation cache to host info and debug files. If memory
1770 cgroup is enabled, each cache can have per memory cgroup
1771 caches. SLUB can create the same sysfs directories for these
1772 caches under /sys/kernel/slab/CACHE/cgroup but it can lead
1773 to a very high number of debug files being created. This is
1774 controlled by slub_memcg_sysfs boot parameter and this
1775 config option determines the parameter's default value.
1778 bool "Disable heap randomization"
1781 Randomizing heap placement makes heap exploits harder, but it
1782 also breaks ancient binaries (including anything libc5 based).
1783 This option changes the bootup default to heap randomization
1784 disabled, and can be overridden at runtime by setting
1785 /proc/sys/kernel/randomize_va_space to 2.
1787 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1790 prompt "Choose SLAB allocator"
1793 This option allows to select a slab allocator.
1797 select HAVE_HARDENED_USERCOPY_ALLOCATOR
1799 The regular slab allocator that is established and known to work
1800 well in all environments. It organizes cache hot objects in
1801 per cpu and per node queues.
1804 bool "SLUB (Unqueued Allocator)"
1805 select HAVE_HARDENED_USERCOPY_ALLOCATOR
1807 SLUB is a slab allocator that minimizes cache line usage
1808 instead of managing queues of cached objects (SLAB approach).
1809 Per cpu caching is realized using slabs of objects instead
1810 of queues of objects. SLUB can use memory efficiently
1811 and has enhanced diagnostics. SLUB is the default choice for
1816 bool "SLOB (Simple Allocator)"
1818 SLOB replaces the stock allocator with a drastically simpler
1819 allocator. SLOB is generally more space efficient but
1820 does not perform as well on large systems.
1824 config SLAB_MERGE_DEFAULT
1825 bool "Allow slab caches to be merged"
1828 For reduced kernel memory fragmentation, slab caches can be
1829 merged when they share the same size and other characteristics.
1830 This carries a risk of kernel heap overflows being able to
1831 overwrite objects from merged caches (and more easily control
1832 cache layout), which makes such heap attacks easier to exploit
1833 by attackers. By keeping caches unmerged, these kinds of exploits
1834 can usually only damage objects in the same cache. To disable
1835 merging at runtime, "slab_nomerge" can be passed on the kernel
1838 config SLAB_FREELIST_RANDOM
1840 depends on SLAB || SLUB
1841 bool "SLAB freelist randomization"
1843 Randomizes the freelist order used on creating new pages. This
1844 security feature reduces the predictability of the kernel slab
1845 allocator against heap overflows.
1847 config SLAB_FREELIST_HARDENED
1848 bool "Harden slab freelist metadata"
1851 Many kernel heap attacks try to target slab cache metadata and
1852 other infrastructure. This options makes minor performance
1853 sacrifices to harden the kernel slab allocator against common
1854 freelist exploit methods.
1856 config SHUFFLE_PAGE_ALLOCATOR
1857 bool "Page allocator randomization"
1858 default SLAB_FREELIST_RANDOM && ACPI_NUMA
1860 Randomization of the page allocator improves the average
1861 utilization of a direct-mapped memory-side-cache. See section
1862 5.2.27 Heterogeneous Memory Attribute Table (HMAT) in the ACPI
1863 6.2a specification for an example of how a platform advertises
1864 the presence of a memory-side-cache. There are also incidental
1865 security benefits as it reduces the predictability of page
1866 allocations to compliment SLAB_FREELIST_RANDOM, but the
1867 default granularity of shuffling on the "MAX_ORDER - 1" i.e,
1868 10th order of pages is selected based on cache utilization
1871 While the randomization improves cache utilization it may
1872 negatively impact workloads on platforms without a cache. For
1873 this reason, by default, the randomization is enabled only
1874 after runtime detection of a direct-mapped memory-side-cache.
1875 Otherwise, the randomization may be force enabled with the
1876 'page_alloc.shuffle' kernel command line parameter.
1880 config SLUB_CPU_PARTIAL
1882 depends on SLUB && SMP
1883 bool "SLUB per cpu partial cache"
1885 Per cpu partial caches accelerate objects allocation and freeing
1886 that is local to a processor at the price of more indeterminism
1887 in the latency of the free. On overflow these caches will be cleared
1888 which requires the taking of locks that may cause latency spikes.
1889 Typically one would choose no for a realtime system.
1891 config MMAP_ALLOW_UNINITIALIZED
1892 bool "Allow mmapped anonymous memory to be uninitialized"
1893 depends on EXPERT && !MMU
1896 Normally, and according to the Linux spec, anonymous memory obtained
1897 from mmap() has its contents cleared before it is passed to
1898 userspace. Enabling this config option allows you to request that
1899 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1900 providing a huge performance boost. If this option is not enabled,
1901 then the flag will be ignored.
1903 This is taken advantage of by uClibc's malloc(), and also by
1904 ELF-FDPIC binfmt's brk and stack allocator.
1906 Because of the obvious security issues, this option should only be
1907 enabled on embedded devices where you control what is run in
1908 userspace. Since that isn't generally a problem on no-MMU systems,
1909 it is normally safe to say Y here.
1911 See Documentation/nommu-mmap.txt for more information.
1913 config SYSTEM_DATA_VERIFICATION
1915 select SYSTEM_TRUSTED_KEYRING
1919 select ASYMMETRIC_KEY_TYPE
1920 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
1923 select X509_CERTIFICATE_PARSER
1924 select PKCS7_MESSAGE_PARSER
1926 Provide PKCS#7 message verification using the contents of the system
1927 trusted keyring to provide public keys. This then can be used for
1928 module verification, kexec image verification and firmware blob
1932 bool "Profiling support"
1934 Say Y here to enable the extended profiling support mechanisms used
1935 by profilers such as OProfile.
1938 # Place an empty function call at each tracepoint site. Can be
1939 # dynamically changed for a probe function.
1944 endmenu # General setup
1946 source "arch/Kconfig"
1953 default 0 if BASE_FULL
1954 default 1 if !BASE_FULL
1956 config MODULE_SIG_FORMAT
1958 select SYSTEM_DATA_VERIFICATION
1961 bool "Enable loadable module support"
1964 Kernel modules are small pieces of compiled code which can
1965 be inserted in the running kernel, rather than being
1966 permanently built into the kernel. You use the "modprobe"
1967 tool to add (and sometimes remove) them. If you say Y here,
1968 many parts of the kernel can be built as modules (by
1969 answering M instead of Y where indicated): this is most
1970 useful for infrequently used options which are not required
1971 for booting. For more information, see the man pages for
1972 modprobe, lsmod, modinfo, insmod and rmmod.
1974 If you say Y here, you will need to run "make
1975 modules_install" to put the modules under /lib/modules/
1976 where modprobe can find them (you may need to be root to do
1983 config MODULE_FORCE_LOAD
1984 bool "Forced module loading"
1987 Allow loading of modules without version information (ie. modprobe
1988 --force). Forced module loading sets the 'F' (forced) taint flag and
1989 is usually a really bad idea.
1991 config MODULE_UNLOAD
1992 bool "Module unloading"
1994 Without this option you will not be able to unload any
1995 modules (note that some modules may not be unloadable
1996 anyway), which makes your kernel smaller, faster
1997 and simpler. If unsure, say Y.
1999 config MODULE_FORCE_UNLOAD
2000 bool "Forced module unloading"
2001 depends on MODULE_UNLOAD
2003 This option allows you to force a module to unload, even if the
2004 kernel believes it is unsafe: the kernel will remove the module
2005 without waiting for anyone to stop using it (using the -f option to
2006 rmmod). This is mainly for kernel developers and desperate users.
2010 bool "Module versioning support"
2012 Usually, you have to use modules compiled with your kernel.
2013 Saying Y here makes it sometimes possible to use modules
2014 compiled for different kernels, by adding enough information
2015 to the modules to (hopefully) spot any changes which would
2016 make them incompatible with the kernel you are running. If
2019 config ASM_MODVERSIONS
2021 default HAVE_ASM_MODVERSIONS && MODVERSIONS
2023 This enables module versioning for exported symbols also from
2024 assembly. This can be enabled only when the target architecture
2027 config MODULE_REL_CRCS
2029 depends on MODVERSIONS
2031 config MODULE_SRCVERSION_ALL
2032 bool "Source checksum for all modules"
2034 Modules which contain a MODULE_VERSION get an extra "srcversion"
2035 field inserted into their modinfo section, which contains a
2036 sum of the source files which made it. This helps maintainers
2037 see exactly which source was used to build a module (since
2038 others sometimes change the module source without updating
2039 the version). With this option, such a "srcversion" field
2040 will be created for all modules. If unsure, say N.
2043 bool "Module signature verification"
2044 select MODULE_SIG_FORMAT
2046 Check modules for valid signatures upon load: the signature
2047 is simply appended to the module. For more information see
2048 <file:Documentation/admin-guide/module-signing.rst>.
2050 Note that this option adds the OpenSSL development packages as a
2051 kernel build dependency so that the signing tool can use its crypto
2054 You should enable this option if you wish to use either
2055 CONFIG_SECURITY_LOCKDOWN_LSM or lockdown functionality imposed via
2056 another LSM - otherwise unsigned modules will be loadable regardless
2057 of the lockdown policy.
2059 !!!WARNING!!! If you enable this option, you MUST make sure that the
2060 module DOES NOT get stripped after being signed. This includes the
2061 debuginfo strip done by some packagers (such as rpmbuild) and
2062 inclusion into an initramfs that wants the module size reduced.
2064 config MODULE_SIG_FORCE
2065 bool "Require modules to be validly signed"
2066 depends on MODULE_SIG
2068 Reject unsigned modules or signed modules for which we don't have a
2069 key. Without this, such modules will simply taint the kernel.
2071 config MODULE_SIG_ALL
2072 bool "Automatically sign all modules"
2074 depends on MODULE_SIG
2076 Sign all modules during make modules_install. Without this option,
2077 modules must be signed manually, using the scripts/sign-file tool.
2079 comment "Do not forget to sign required modules with scripts/sign-file"
2080 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL
2083 prompt "Which hash algorithm should modules be signed with?"
2084 depends on MODULE_SIG
2086 This determines which sort of hashing algorithm will be used during
2087 signature generation. This algorithm _must_ be built into the kernel
2088 directly so that signature verification can take place. It is not
2089 possible to load a signed module containing the algorithm to check
2090 the signature on that module.
2092 config MODULE_SIG_SHA1
2093 bool "Sign modules with SHA-1"
2096 config MODULE_SIG_SHA224
2097 bool "Sign modules with SHA-224"
2098 select CRYPTO_SHA256
2100 config MODULE_SIG_SHA256
2101 bool "Sign modules with SHA-256"
2102 select CRYPTO_SHA256
2104 config MODULE_SIG_SHA384
2105 bool "Sign modules with SHA-384"
2106 select CRYPTO_SHA512
2108 config MODULE_SIG_SHA512
2109 bool "Sign modules with SHA-512"
2110 select CRYPTO_SHA512
2114 config MODULE_SIG_HASH
2116 depends on MODULE_SIG
2117 default "sha1" if MODULE_SIG_SHA1
2118 default "sha224" if MODULE_SIG_SHA224
2119 default "sha256" if MODULE_SIG_SHA256
2120 default "sha384" if MODULE_SIG_SHA384
2121 default "sha512" if MODULE_SIG_SHA512
2123 config MODULE_COMPRESS
2124 bool "Compress modules on installation"
2127 Compresses kernel modules when 'make modules_install' is run; gzip or
2128 xz depending on "Compression algorithm" below.
2130 module-init-tools MAY support gzip, and kmod MAY support gzip and xz.
2132 Out-of-tree kernel modules installed using Kbuild will also be
2133 compressed upon installation.
2135 Note: for modules inside an initrd or initramfs, it's more efficient
2136 to compress the whole initrd or initramfs instead.
2138 Note: This is fully compatible with signed modules.
2143 prompt "Compression algorithm"
2144 depends on MODULE_COMPRESS
2145 default MODULE_COMPRESS_GZIP
2147 This determines which sort of compression will be used during
2148 'make modules_install'.
2150 GZIP (default) and XZ are supported.
2152 config MODULE_COMPRESS_GZIP
2155 config MODULE_COMPRESS_XZ
2160 config MODULE_ALLOW_MISSING_NAMESPACE_IMPORTS
2161 bool "Allow loading of modules with missing namespace imports"
2163 Symbols exported with EXPORT_SYMBOL_NS*() are considered exported in
2164 a namespace. A module that makes use of a symbol exported with such a
2165 namespace is required to import the namespace via MODULE_IMPORT_NS().
2166 There is no technical reason to enforce correct namespace imports,
2167 but it creates consistency between symbols defining namespaces and
2168 users importing namespaces they make use of. This option relaxes this
2169 requirement and lifts the enforcement when loading a module.
2173 config UNUSED_SYMBOLS
2174 bool "Enable unused/obsolete exported symbols"
2177 Unused but exported symbols make the kernel needlessly bigger. For
2178 that reason most of these unused exports will soon be removed. This
2179 option is provided temporarily to provide a transition period in case
2180 some external kernel module needs one of these symbols anyway. If you
2181 encounter such a case in your module, consider if you are actually
2182 using the right API. (rationale: since nobody in the kernel is using
2183 this in a module, there is a pretty good chance it's actually the
2184 wrong interface to use). If you really need the symbol, please send a
2185 mail to the linux kernel mailing list mentioning the symbol and why
2186 you really need it, and what the merge plan to the mainline kernel for
2189 config TRIM_UNUSED_KSYMS
2190 bool "Trim unused exported kernel symbols"
2191 depends on !UNUSED_SYMBOLS
2193 The kernel and some modules make many symbols available for
2194 other modules to use via EXPORT_SYMBOL() and variants. Depending
2195 on the set of modules being selected in your kernel configuration,
2196 many of those exported symbols might never be used.
2198 This option allows for unused exported symbols to be dropped from
2199 the build. In turn, this provides the compiler more opportunities
2200 (especially when using LTO) for optimizing the code and reducing
2201 binary size. This might have some security advantages as well.
2203 If unsure, or if you need to build out-of-tree modules, say N.
2205 config UNUSED_KSYMS_WHITELIST
2206 string "Whitelist of symbols to keep in ksymtab"
2207 depends on TRIM_UNUSED_KSYMS
2209 By default, all unused exported symbols will be un-exported from the
2210 build when TRIM_UNUSED_KSYMS is selected.
2212 UNUSED_KSYMS_WHITELIST allows to whitelist symbols that must be kept
2213 exported at all times, even in absence of in-tree users. The value to
2214 set here is the path to a text file containing the list of symbols,
2215 one per line. The path can be absolute, or relative to the kernel
2220 config MODULES_TREE_LOOKUP
2222 depends on PERF_EVENTS || TRACING
2224 config INIT_ALL_POSSIBLE
2227 Back when each arch used to define their own cpu_online_mask and
2228 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
2229 with all 1s, and others with all 0s. When they were centralised,
2230 it was better to provide this option than to break all the archs
2231 and have several arch maintainers pursuing me down dark alleys.
2233 source "block/Kconfig"
2235 config PREEMPT_NOTIFIERS
2245 Build a simple ASN.1 grammar compiler that produces a bytecode output
2246 that can be interpreted by the ASN.1 stream decoder and used to
2247 inform it as to what tags are to be expected in a stream and what
2248 functions to call on what tags.
2250 source "kernel/Kconfig.locks"
2252 config ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE
2255 config ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
2258 # It may be useful for an architecture to override the definitions of the
2259 # SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h>
2260 # and the COMPAT_ variants in <linux/compat.h>, in particular to use a
2261 # different calling convention for syscalls. They can also override the
2262 # macros for not-implemented syscalls in kernel/sys_ni.c and
2263 # kernel/time/posix-stubs.c. All these overrides need to be available in
2264 # <asm/syscall_wrapper.h>.
2265 config ARCH_HAS_SYSCALL_WRAPPER