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)"
10 default "arch/$(ARCH)/defconfig"
13 def_bool $(success,$(CC) --version | head -n 1 | grep -q gcc)
17 default $(shell,$(srctree)/scripts/gcc-version.sh $(CC)) if CC_IS_GCC
21 def_bool $(success,$(CC) --version | head -n 1 | grep -q clang)
25 default $(shell,$(srctree)/scripts/clang-version.sh $(CC))
28 def_bool $(success,$(srctree)/scripts/cc-can-link.sh $(CC))
30 config CC_HAS_ASM_GOTO
31 def_bool $(success,$(srctree)/scripts/gcc-goto.sh $(CC))
33 config TOOLS_SUPPORT_RELR
34 def_bool $(success,env "CC=$(CC)" "LD=$(LD)" "NM=$(NM)" "OBJCOPY=$(OBJCOPY)" $(srctree)/scripts/tools-support-relr.sh)
36 config CC_HAS_ASM_INLINE
37 def_bool $(success,echo 'void foo(void) { asm inline (""); }' | $(CC) -x c - -c -o /dev/null)
39 config CC_HAS_WARN_MAYBE_UNINITIALIZED
40 def_bool $(cc-option,-Wmaybe-uninitialized)
42 GCC >= 4.7 supports this option.
44 config CC_DISABLE_WARN_MAYBE_UNINITIALIZED
46 depends on CC_HAS_WARN_MAYBE_UNINITIALIZED
47 default CC_IS_GCC && GCC_VERSION < 40900 # unreliable for GCC < 4.9
49 GCC's -Wmaybe-uninitialized is not reliable by definition.
50 Lots of false positive warnings are produced in some cases.
52 If this option is enabled, -Wno-maybe-uninitialzed is passed
53 to the compiler to suppress maybe-uninitialized warnings.
62 config BUILDTIME_TABLE_SORT
65 config THREAD_INFO_IN_TASK
68 Select this to move thread_info off the stack into task_struct. To
69 make this work, an arch will need to remove all thread_info fields
70 except flags and fix any runtime bugs.
72 One subtle change that will be needed is to use try_get_task_stack()
73 and put_task_stack() in save_thread_stack_tsk() and get_wchan().
82 depends on BROKEN || !SMP
85 config INIT_ENV_ARG_LIMIT
90 Maximum of each of the number of arguments and environment
91 variables passed to init from the kernel command line.
94 bool "Compile also drivers which will not load"
98 Some drivers can be compiled on a different platform than they are
99 intended to be run on. Despite they cannot be loaded there (or even
100 when they load they cannot be used due to missing HW support),
101 developers still, opposing to distributors, might want to build such
102 drivers to compile-test them.
104 If you are a developer and want to build everything available, say Y
105 here. If you are a user/distributor, say N here to exclude useless
106 drivers to be distributed.
108 config UAPI_HEADER_TEST
109 bool "Compile test UAPI headers"
110 depends on HEADERS_INSTALL && CC_CAN_LINK
112 Compile test headers exported to user-space to ensure they are
113 self-contained, i.e. compilable as standalone units.
115 If you are a developer or tester and want to ensure the exported
116 headers are self-contained, say Y here. Otherwise, choose N.
119 string "Local version - append to kernel release"
121 Append an extra string to the end of your kernel version.
122 This will show up when you type uname, for example.
123 The string you set here will be appended after the contents of
124 any files with a filename matching localversion* in your
125 object and source tree, in that order. Your total string can
126 be a maximum of 64 characters.
128 config LOCALVERSION_AUTO
129 bool "Automatically append version information to the version string"
131 depends on !COMPILE_TEST
133 This will try to automatically determine if the current tree is a
134 release tree by looking for git tags that belong to the current
135 top of tree revision.
137 A string of the format -gxxxxxxxx will be added to the localversion
138 if a git-based tree is found. The string generated by this will be
139 appended after any matching localversion* files, and after the value
140 set in CONFIG_LOCALVERSION.
142 (The actual string used here is the first eight characters produced
143 by running the command:
145 $ git rev-parse --verify HEAD
147 which is done within the script "scripts/setlocalversion".)
150 string "Build ID Salt"
153 The build ID is used to link binaries and their debug info. Setting
154 this option will use the value in the calculation of the build id.
155 This is mostly useful for distributions which want to ensure the
156 build is unique between builds. It's safe to leave the default.
158 config HAVE_KERNEL_GZIP
161 config HAVE_KERNEL_BZIP2
164 config HAVE_KERNEL_LZMA
167 config HAVE_KERNEL_XZ
170 config HAVE_KERNEL_LZO
173 config HAVE_KERNEL_LZ4
176 config HAVE_KERNEL_UNCOMPRESSED
180 prompt "Kernel compression mode"
182 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_UNCOMPRESSED
184 The linux kernel is a kind of self-extracting executable.
185 Several compression algorithms are available, which differ
186 in efficiency, compression and decompression speed.
187 Compression speed is only relevant when building a kernel.
188 Decompression speed is relevant at each boot.
190 If you have any problems with bzip2 or lzma compressed
192 version of this functionality (bzip2 only), for 2.4, was
193 supplied by Christian Ludwig)
195 High compression options are mostly useful for users, who
196 are low on disk space (embedded systems), but for whom ram
199 If in doubt, select 'gzip'
203 depends on HAVE_KERNEL_GZIP
205 The old and tried gzip compression. It provides a good balance
206 between compression ratio and decompression speed.
210 depends on HAVE_KERNEL_BZIP2
212 Its compression ratio and speed is intermediate.
213 Decompression speed is slowest among the choices. The kernel
214 size is about 10% smaller with bzip2, in comparison to gzip.
215 Bzip2 uses a large amount of memory. For modern kernels you
216 will need at least 8MB RAM or more for booting.
220 depends on HAVE_KERNEL_LZMA
222 This compression algorithm's ratio is best. Decompression speed
223 is between gzip and bzip2. Compression is slowest.
224 The kernel size is about 33% smaller with LZMA in comparison to gzip.
228 depends on HAVE_KERNEL_XZ
230 XZ uses the LZMA2 algorithm and instruction set specific
231 BCJ filters which can improve compression ratio of executable
232 code. The size of the kernel is about 30% smaller with XZ in
233 comparison to gzip. On architectures for which there is a BCJ
234 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
235 will create a few percent smaller kernel than plain LZMA.
237 The speed is about the same as with LZMA: The decompression
238 speed of XZ is better than that of bzip2 but worse than gzip
239 and LZO. Compression is slow.
243 depends on HAVE_KERNEL_LZO
245 Its compression ratio is the poorest among the choices. The kernel
246 size is about 10% bigger than gzip; however its speed
247 (both compression and decompression) is the fastest.
251 depends on HAVE_KERNEL_LZ4
253 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
254 A preliminary version of LZ4 de/compression tool is available at
255 <https://code.google.com/p/lz4/>.
257 Its compression ratio is worse than LZO. The size of the kernel
258 is about 8% bigger than LZO. But the decompression speed is
261 config KERNEL_UNCOMPRESSED
263 depends on HAVE_KERNEL_UNCOMPRESSED
265 Produce uncompressed kernel image. This option is usually not what
266 you want. It is useful for debugging the kernel in slow simulation
267 environments, where decompressing and moving the kernel is awfully
268 slow. This option allows early boot code to skip the decompressor
269 and jump right at uncompressed kernel image.
273 config DEFAULT_HOSTNAME
274 string "Default hostname"
277 This option determines the default system hostname before userspace
278 calls sethostname(2). The kernel traditionally uses "(none)" here,
279 but you may wish to use a different default here to make a minimal
280 system more usable with less configuration.
283 # For some reason microblaze and nios2 hard code SWAP=n. Hopefully we can
284 # add proper SWAP support to them, in which case this can be remove.
290 bool "Support for paging of anonymous memory (swap)"
291 depends on MMU && BLOCK && !ARCH_NO_SWAP
294 This option allows you to choose whether you want to have support
295 for so called swap devices or swap files in your kernel that are
296 used to provide more virtual memory than the actual RAM present
297 in your computer. If unsure say Y.
302 Inter Process Communication is a suite of library functions and
303 system calls which let processes (running programs) synchronize and
304 exchange information. It is generally considered to be a good thing,
305 and some programs won't run unless you say Y here. In particular, if
306 you want to run the DOS emulator dosemu under Linux (read the
307 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
308 you'll need to say Y here.
310 You can find documentation about IPC with "info ipc" and also in
311 section 6.4 of the Linux Programmer's Guide, available from
312 <http://www.tldp.org/guides.html>.
314 config SYSVIPC_SYSCTL
321 bool "POSIX Message Queues"
324 POSIX variant of message queues is a part of IPC. In POSIX message
325 queues every message has a priority which decides about succession
326 of receiving it by a process. If you want to compile and run
327 programs written e.g. for Solaris with use of its POSIX message
328 queues (functions mq_*) say Y here.
330 POSIX message queues are visible as a filesystem called 'mqueue'
331 and can be mounted somewhere if you want to do filesystem
332 operations on message queues.
336 config POSIX_MQUEUE_SYSCTL
338 depends on POSIX_MQUEUE
342 config CROSS_MEMORY_ATTACH
343 bool "Enable process_vm_readv/writev syscalls"
347 Enabling this option adds the system calls process_vm_readv and
348 process_vm_writev which allow a process with the correct privileges
349 to directly read from or write to another process' address space.
350 See the man page for more details.
353 bool "uselib syscall"
354 def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION
356 This option enables the uselib syscall, a system call used in the
357 dynamic linker from libc5 and earlier. glibc does not use this
358 system call. If you intend to run programs built on libc5 or
359 earlier, you may need to enable this syscall. Current systems
360 running glibc can safely disable this.
363 bool "Auditing support"
366 Enable auditing infrastructure that can be used with another
367 kernel subsystem, such as SELinux (which requires this for
368 logging of avc messages output). System call auditing is included
369 on architectures which support it.
371 config HAVE_ARCH_AUDITSYSCALL
376 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL
379 source "kernel/irq/Kconfig"
380 source "kernel/time/Kconfig"
381 source "kernel/Kconfig.preempt"
383 menu "CPU/Task time and stats accounting"
385 config VIRT_CPU_ACCOUNTING
389 prompt "Cputime accounting"
390 default TICK_CPU_ACCOUNTING if !PPC64
391 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64
393 # Kind of a stub config for the pure tick based cputime accounting
394 config TICK_CPU_ACCOUNTING
395 bool "Simple tick based cputime accounting"
396 depends on !S390 && !NO_HZ_FULL
398 This is the basic tick based cputime accounting that maintains
399 statistics about user, system and idle time spent on per jiffies
404 config VIRT_CPU_ACCOUNTING_NATIVE
405 bool "Deterministic task and CPU time accounting"
406 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL
407 select VIRT_CPU_ACCOUNTING
409 Select this option to enable more accurate task and CPU time
410 accounting. This is done by reading a CPU counter on each
411 kernel entry and exit and on transitions within the kernel
412 between system, softirq and hardirq state, so there is a
413 small performance impact. In the case of s390 or IBM POWER > 5,
414 this also enables accounting of stolen time on logically-partitioned
417 config VIRT_CPU_ACCOUNTING_GEN
418 bool "Full dynticks CPU time accounting"
419 depends on HAVE_CONTEXT_TRACKING
420 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN
421 depends on GENERIC_CLOCKEVENTS
422 select VIRT_CPU_ACCOUNTING
423 select CONTEXT_TRACKING
425 Select this option to enable task and CPU time accounting on full
426 dynticks systems. This accounting is implemented by watching every
427 kernel-user boundaries using the context tracking subsystem.
428 The accounting is thus performed at the expense of some significant
431 For now this is only useful if you are working on the full
432 dynticks subsystem development.
438 config IRQ_TIME_ACCOUNTING
439 bool "Fine granularity task level IRQ time accounting"
440 depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE
442 Select this option to enable fine granularity task irq time
443 accounting. This is done by reading a timestamp on each
444 transitions between softirq and hardirq state, so there can be a
445 small performance impact.
447 If in doubt, say N here.
449 config HAVE_SCHED_AVG_IRQ
451 depends on IRQ_TIME_ACCOUNTING || PARAVIRT_TIME_ACCOUNTING
454 config SCHED_THERMAL_PRESSURE
455 bool "Enable periodic averaging of thermal pressure"
458 config BSD_PROCESS_ACCT
459 bool "BSD Process Accounting"
462 If you say Y here, a user level program will be able to instruct the
463 kernel (via a special system call) to write process accounting
464 information to a file: whenever a process exits, information about
465 that process will be appended to the file by the kernel. The
466 information includes things such as creation time, owning user,
467 command name, memory usage, controlling terminal etc. (the complete
468 list is in the struct acct in <file:include/linux/acct.h>). It is
469 up to the user level program to do useful things with this
470 information. This is generally a good idea, so say Y.
472 config BSD_PROCESS_ACCT_V3
473 bool "BSD Process Accounting version 3 file format"
474 depends on BSD_PROCESS_ACCT
477 If you say Y here, the process accounting information is written
478 in a new file format that also logs the process IDs of each
479 process and its parent. Note that this file format is incompatible
480 with previous v0/v1/v2 file formats, so you will need updated tools
481 for processing it. A preliminary version of these tools is available
482 at <http://www.gnu.org/software/acct/>.
485 bool "Export task/process statistics through netlink"
490 Export selected statistics for tasks/processes through the
491 generic netlink interface. Unlike BSD process accounting, the
492 statistics are available during the lifetime of tasks/processes as
493 responses to commands. Like BSD accounting, they are sent to user
498 config TASK_DELAY_ACCT
499 bool "Enable per-task delay accounting"
503 Collect information on time spent by a task waiting for system
504 resources like cpu, synchronous block I/O completion and swapping
505 in pages. Such statistics can help in setting a task's priorities
506 relative to other tasks for cpu, io, rss limits etc.
511 bool "Enable extended accounting over taskstats"
514 Collect extended task accounting data and send the data
515 to userland for processing over the taskstats interface.
519 config TASK_IO_ACCOUNTING
520 bool "Enable per-task storage I/O accounting"
521 depends on TASK_XACCT
523 Collect information on the number of bytes of storage I/O which this
529 bool "Pressure stall information tracking"
531 Collect metrics that indicate how overcommitted the CPU, memory,
532 and IO capacity are in the system.
534 If you say Y here, the kernel will create /proc/pressure/ with the
535 pressure statistics files cpu, memory, and io. These will indicate
536 the share of walltime in which some or all tasks in the system are
537 delayed due to contention of the respective resource.
539 In kernels with cgroup support, cgroups (cgroup2 only) will
540 have cpu.pressure, memory.pressure, and io.pressure files,
541 which aggregate pressure stalls for the grouped tasks only.
543 For more details see Documentation/accounting/psi.rst.
547 config PSI_DEFAULT_DISABLED
548 bool "Require boot parameter to enable pressure stall information tracking"
552 If set, pressure stall information tracking will be disabled
553 per default but can be enabled through passing psi=1 on the
554 kernel commandline during boot.
556 This feature adds some code to the task wakeup and sleep
557 paths of the scheduler. The overhead is too low to affect
558 common scheduling-intense workloads in practice (such as
559 webservers, memcache), but it does show up in artificial
560 scheduler stress tests, such as hackbench.
562 If you are paranoid and not sure what the kernel will be
567 endmenu # "CPU/Task time and stats accounting"
571 depends on SMP || COMPILE_TEST
574 Make sure that CPUs running critical tasks are not disturbed by
575 any source of "noise" such as unbound workqueues, timers, kthreads...
576 Unbound jobs get offloaded to housekeeping CPUs. This is driven by
577 the "isolcpus=" boot parameter.
581 source "kernel/rcu/Kconfig"
588 tristate "Kernel .config support"
590 This option enables the complete Linux kernel ".config" file
591 contents to be saved in the kernel. It provides documentation
592 of which kernel options are used in a running kernel or in an
593 on-disk kernel. This information can be extracted from the kernel
594 image file with the script scripts/extract-ikconfig and used as
595 input to rebuild the current kernel or to build another kernel.
596 It can also be extracted from a running kernel by reading
597 /proc/config.gz if enabled (below).
600 bool "Enable access to .config through /proc/config.gz"
601 depends on IKCONFIG && PROC_FS
603 This option enables access to the kernel configuration file
604 through /proc/config.gz.
607 tristate "Enable kernel headers through /sys/kernel/kheaders.tar.xz"
610 This option enables access to the in-kernel headers that are generated during
611 the build process. These can be used to build eBPF tracing programs,
612 or similar programs. If you build the headers as a module, a module called
613 kheaders.ko is built which can be loaded on-demand to get access to headers.
616 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
621 Select the minimal kernel log buffer size as a power of 2.
622 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config
623 parameter, see below. Any higher size also might be forced
624 by "log_buf_len" boot parameter.
634 config LOG_CPU_MAX_BUF_SHIFT
635 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)"
638 default 12 if !BASE_SMALL
639 default 0 if BASE_SMALL
642 This option allows to increase the default ring buffer size
643 according to the number of CPUs. The value defines the contribution
644 of each CPU as a power of 2. The used space is typically only few
645 lines however it might be much more when problems are reported,
648 The increased size means that a new buffer has to be allocated and
649 the original static one is unused. It makes sense only on systems
650 with more CPUs. Therefore this value is used only when the sum of
651 contributions is greater than the half of the default kernel ring
652 buffer as defined by LOG_BUF_SHIFT. The default values are set
653 so that more than 64 CPUs are needed to trigger the allocation.
655 Also this option is ignored when "log_buf_len" kernel parameter is
656 used as it forces an exact (power of two) size of the ring buffer.
658 The number of possible CPUs is used for this computation ignoring
659 hotplugging making the computation optimal for the worst case
660 scenario while allowing a simple algorithm to be used from bootup.
662 Examples shift values and their meaning:
663 17 => 128 KB for each CPU
664 16 => 64 KB for each CPU
665 15 => 32 KB for each CPU
666 14 => 16 KB for each CPU
667 13 => 8 KB for each CPU
668 12 => 4 KB for each CPU
670 config PRINTK_SAFE_LOG_BUF_SHIFT
671 int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)"
676 Select the size of an alternate printk per-CPU buffer where messages
677 printed from usafe contexts are temporary stored. One example would
678 be NMI messages, another one - printk recursion. The messages are
679 copied to the main log buffer in a safe context to avoid a deadlock.
680 The value defines the size as a power of 2.
682 Those messages are rare and limited. The largest one is when
683 a backtrace is printed. It usually fits into 4KB. Select
684 8KB if you want to be on the safe side.
687 17 => 128 KB for each CPU
688 16 => 64 KB for each CPU
689 15 => 32 KB for each CPU
690 14 => 16 KB for each CPU
691 13 => 8 KB for each CPU
692 12 => 4 KB for each CPU
695 # Architectures with an unreliable sched_clock() should select this:
697 config HAVE_UNSTABLE_SCHED_CLOCK
700 config GENERIC_SCHED_CLOCK
703 menu "Scheduler features"
706 bool "Enable utilization clamping for RT/FAIR tasks"
707 depends on CPU_FREQ_GOV_SCHEDUTIL
709 This feature enables the scheduler to track the clamped utilization
710 of each CPU based on RUNNABLE tasks scheduled on that CPU.
712 With this option, the user can specify the min and max CPU
713 utilization allowed for RUNNABLE tasks. The max utilization defines
714 the maximum frequency a task should use while the min utilization
715 defines the minimum frequency it should use.
717 Both min and max utilization clamp values are hints to the scheduler,
718 aiming at improving its frequency selection policy, but they do not
719 enforce or grant any specific bandwidth for tasks.
723 config UCLAMP_BUCKETS_COUNT
724 int "Number of supported utilization clamp buckets"
727 depends on UCLAMP_TASK
729 Defines the number of clamp buckets to use. The range of each bucket
730 will be SCHED_CAPACITY_SCALE/UCLAMP_BUCKETS_COUNT. The higher the
731 number of clamp buckets the finer their granularity and the higher
732 the precision of clamping aggregation and tracking at run-time.
734 For example, with the minimum configuration value we will have 5
735 clamp buckets tracking 20% utilization each. A 25% boosted tasks will
736 be refcounted in the [20..39]% bucket and will set the bucket clamp
737 effective value to 25%.
738 If a second 30% boosted task should be co-scheduled on the same CPU,
739 that task will be refcounted in the same bucket of the first task and
740 it will boost the bucket clamp effective value to 30%.
741 The clamp effective value of a bucket is reset to its nominal value
742 (20% in the example above) when there are no more tasks refcounted in
745 An additional boost/capping margin can be added to some tasks. In the
746 example above the 25% task will be boosted to 30% until it exits the
747 CPU. If that should be considered not acceptable on certain systems,
748 it's always possible to reduce the margin by increasing the number of
749 clamp buckets to trade off used memory for run-time tracking
752 If in doubt, use the default value.
757 # For architectures that want to enable the support for NUMA-affine scheduler
760 config ARCH_SUPPORTS_NUMA_BALANCING
764 # For architectures that prefer to flush all TLBs after a number of pages
765 # are unmapped instead of sending one IPI per page to flush. The architecture
766 # must provide guarantees on what happens if a clean TLB cache entry is
767 # written after the unmap. Details are in mm/rmap.c near the check for
768 # should_defer_flush. The architecture should also consider if the full flush
769 # and the refill costs are offset by the savings of sending fewer IPIs.
770 config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
774 def_bool !$(cc-option,$(m64-flag) -D__SIZEOF_INT128__=0) && 64BIT
777 # For architectures that know their GCC __int128 support is sound
779 config ARCH_SUPPORTS_INT128
782 # For architectures that (ab)use NUMA to represent different memory regions
783 # all cpu-local but of different latencies, such as SuperH.
785 config ARCH_WANT_NUMA_VARIABLE_LOCALITY
788 config NUMA_BALANCING
789 bool "Memory placement aware NUMA scheduler"
790 depends on ARCH_SUPPORTS_NUMA_BALANCING
791 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
792 depends on SMP && NUMA && MIGRATION
794 This option adds support for automatic NUMA aware memory/task placement.
795 The mechanism is quite primitive and is based on migrating memory when
796 it has references to the node the task is running on.
798 This system will be inactive on UMA systems.
800 config NUMA_BALANCING_DEFAULT_ENABLED
801 bool "Automatically enable NUMA aware memory/task placement"
803 depends on NUMA_BALANCING
805 If set, automatic NUMA balancing will be enabled if running on a NUMA
809 bool "Control Group support"
812 This option adds support for grouping sets of processes together, for
813 use with process control subsystems such as Cpusets, CFS, memory
814 controls or device isolation.
816 - Documentation/scheduler/sched-design-CFS.rst (CFS)
817 - Documentation/admin-guide/cgroup-v1/ (features for grouping, isolation
818 and resource control)
828 bool "Memory controller"
832 Provides control over the memory footprint of tasks in a cgroup.
835 bool "Swap controller"
836 depends on MEMCG && SWAP
838 Provides control over the swap space consumed by tasks in a cgroup.
840 config MEMCG_SWAP_ENABLED
841 bool "Swap controller enabled by default"
842 depends on MEMCG_SWAP
845 Memory Resource Controller Swap Extension comes with its price in
846 a bigger memory consumption. General purpose distribution kernels
847 which want to enable the feature but keep it disabled by default
848 and let the user enable it by swapaccount=1 boot command line
849 parameter should have this option unselected.
850 For those who want to have the feature enabled by default should
851 select this option (if, for some reason, they need to disable it
852 then swapaccount=0 does the trick).
856 depends on MEMCG && !SLOB
864 Generic block IO controller cgroup interface. This is the common
865 cgroup interface which should be used by various IO controlling
868 Currently, CFQ IO scheduler uses it to recognize task groups and
869 control disk bandwidth allocation (proportional time slice allocation)
870 to such task groups. It is also used by bio throttling logic in
871 block layer to implement upper limit in IO rates on a device.
873 This option only enables generic Block IO controller infrastructure.
874 One needs to also enable actual IO controlling logic/policy. For
875 enabling proportional weight division of disk bandwidth in CFQ, set
876 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
877 CONFIG_BLK_DEV_THROTTLING=y.
879 See Documentation/admin-guide/cgroup-v1/blkio-controller.rst for more information.
881 config CGROUP_WRITEBACK
883 depends on MEMCG && BLK_CGROUP
886 menuconfig CGROUP_SCHED
887 bool "CPU controller"
890 This feature lets CPU scheduler recognize task groups and control CPU
891 bandwidth allocation to such task groups. It uses cgroups to group
895 config FAIR_GROUP_SCHED
896 bool "Group scheduling for SCHED_OTHER"
897 depends on CGROUP_SCHED
901 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
902 depends on FAIR_GROUP_SCHED
905 This option allows users to define CPU bandwidth rates (limits) for
906 tasks running within the fair group scheduler. Groups with no limit
907 set are considered to be unconstrained and will run with no
909 See Documentation/scheduler/sched-bwc.rst for more information.
911 config RT_GROUP_SCHED
912 bool "Group scheduling for SCHED_RR/FIFO"
913 depends on CGROUP_SCHED
916 This feature lets you explicitly allocate real CPU bandwidth
917 to task groups. If enabled, it will also make it impossible to
918 schedule realtime tasks for non-root users until you allocate
919 realtime bandwidth for them.
920 See Documentation/scheduler/sched-rt-group.rst for more information.
924 config UCLAMP_TASK_GROUP
925 bool "Utilization clamping per group of tasks"
926 depends on CGROUP_SCHED
927 depends on UCLAMP_TASK
930 This feature enables the scheduler to track the clamped utilization
931 of each CPU based on RUNNABLE tasks currently scheduled on that CPU.
933 When this option is enabled, the user can specify a min and max
934 CPU bandwidth which is allowed for each single task in a group.
935 The max bandwidth allows to clamp the maximum frequency a task
936 can use, while the min bandwidth allows to define a minimum
937 frequency a task will always use.
939 When task group based utilization clamping is enabled, an eventually
940 specified task-specific clamp value is constrained by the cgroup
941 specified clamp value. Both minimum and maximum task clamping cannot
942 be bigger than the corresponding clamping defined at task group level.
947 bool "PIDs controller"
949 Provides enforcement of process number limits in the scope of a
950 cgroup. Any attempt to fork more processes than is allowed in the
951 cgroup will fail. PIDs are fundamentally a global resource because it
952 is fairly trivial to reach PID exhaustion before you reach even a
953 conservative kmemcg limit. As a result, it is possible to grind a
954 system to halt without being limited by other cgroup policies. The
955 PIDs controller is designed to stop this from happening.
957 It should be noted that organisational operations (such as attaching
958 to a cgroup hierarchy) will *not* be blocked by the PIDs controller,
959 since the PIDs limit only affects a process's ability to fork, not to
963 bool "RDMA controller"
965 Provides enforcement of RDMA resources defined by IB stack.
966 It is fairly easy for consumers to exhaust RDMA resources, which
967 can result into resource unavailability to other consumers.
968 RDMA controller is designed to stop this from happening.
969 Attaching processes with active RDMA resources to the cgroup
970 hierarchy is allowed even if can cross the hierarchy's limit.
972 config CGROUP_FREEZER
973 bool "Freezer controller"
975 Provides a way to freeze and unfreeze all tasks in a
978 This option affects the ORIGINAL cgroup interface. The cgroup2 memory
979 controller includes important in-kernel memory consumers per default.
981 If you're using cgroup2, say N.
983 config CGROUP_HUGETLB
984 bool "HugeTLB controller"
985 depends on HUGETLB_PAGE
989 Provides a cgroup controller for HugeTLB pages.
990 When you enable this, you can put a per cgroup limit on HugeTLB usage.
991 The limit is enforced during page fault. Since HugeTLB doesn't
992 support page reclaim, enforcing the limit at page fault time implies
993 that, the application will get SIGBUS signal if it tries to access
994 HugeTLB pages beyond its limit. This requires the application to know
995 beforehand how much HugeTLB pages it would require for its use. The
996 control group is tracked in the third page lru pointer. This means
997 that we cannot use the controller with huge page less than 3 pages.
1000 bool "Cpuset controller"
1003 This option will let you create and manage CPUSETs which
1004 allow dynamically partitioning a system into sets of CPUs and
1005 Memory Nodes and assigning tasks to run only within those sets.
1006 This is primarily useful on large SMP or NUMA systems.
1010 config PROC_PID_CPUSET
1011 bool "Include legacy /proc/<pid>/cpuset file"
1015 config CGROUP_DEVICE
1016 bool "Device controller"
1018 Provides a cgroup controller implementing whitelists for
1019 devices which a process in the cgroup can mknod or open.
1021 config CGROUP_CPUACCT
1022 bool "Simple CPU accounting controller"
1024 Provides a simple controller for monitoring the
1025 total CPU consumed by the tasks in a cgroup.
1028 bool "Perf controller"
1029 depends on PERF_EVENTS
1031 This option extends the perf per-cpu mode to restrict monitoring
1032 to threads which belong to the cgroup specified and run on the
1038 bool "Support for eBPF programs attached to cgroups"
1039 depends on BPF_SYSCALL
1040 select SOCK_CGROUP_DATA
1042 Allow attaching eBPF programs to a cgroup using the bpf(2)
1043 syscall command BPF_PROG_ATTACH.
1045 In which context these programs are accessed depends on the type
1046 of attachment. For instance, programs that are attached using
1047 BPF_CGROUP_INET_INGRESS will be executed on the ingress path of
1051 bool "Debug controller"
1053 depends on DEBUG_KERNEL
1055 This option enables a simple controller that exports
1056 debugging information about the cgroups framework. This
1057 controller is for control cgroup debugging only. Its
1058 interfaces are not stable.
1062 config SOCK_CGROUP_DATA
1068 menuconfig NAMESPACES
1069 bool "Namespaces support" if EXPERT
1070 depends on MULTIUSER
1073 Provides the way to make tasks work with different objects using
1074 the same id. For example same IPC id may refer to different objects
1075 or same user id or pid may refer to different tasks when used in
1076 different namespaces.
1081 bool "UTS namespace"
1084 In this namespace tasks see different info provided with the
1088 bool "TIME namespace"
1089 depends on GENERIC_VDSO_TIME_NS
1092 In this namespace boottime and monotonic clocks can be set.
1093 The time will keep going with the same pace.
1096 bool "IPC namespace"
1097 depends on (SYSVIPC || POSIX_MQUEUE)
1100 In this namespace tasks work with IPC ids which correspond to
1101 different IPC objects in different namespaces.
1104 bool "User namespace"
1107 This allows containers, i.e. vservers, to use user namespaces
1108 to provide different user info for different servers.
1110 When user namespaces are enabled in the kernel it is
1111 recommended that the MEMCG option also be enabled and that
1112 user-space use the memory control groups to limit the amount
1113 of memory a memory unprivileged users can use.
1118 bool "PID Namespaces"
1121 Support process id namespaces. This allows having multiple
1122 processes with the same pid as long as they are in different
1123 pid namespaces. This is a building block of containers.
1126 bool "Network namespace"
1130 Allow user space to create what appear to be multiple instances
1131 of the network stack.
1135 config CHECKPOINT_RESTORE
1136 bool "Checkpoint/restore support"
1137 select PROC_CHILDREN
1140 Enables additional kernel features in a sake of checkpoint/restore.
1141 In particular it adds auxiliary prctl codes to setup process text,
1142 data and heap segment sizes, and a few additional /proc filesystem
1145 If unsure, say N here.
1147 config SCHED_AUTOGROUP
1148 bool "Automatic process group scheduling"
1151 select FAIR_GROUP_SCHED
1153 This option optimizes the scheduler for common desktop workloads by
1154 automatically creating and populating task groups. This separation
1155 of workloads isolates aggressive CPU burners (like build jobs) from
1156 desktop applications. Task group autogeneration is currently based
1159 config SYSFS_DEPRECATED
1160 bool "Enable deprecated sysfs features to support old userspace tools"
1164 This option adds code that switches the layout of the "block" class
1165 devices, to not show up in /sys/class/block/, but only in
1168 This switch is only active when the sysfs.deprecated=1 boot option is
1169 passed or the SYSFS_DEPRECATED_V2 option is set.
1171 This option allows new kernels to run on old distributions and tools,
1172 which might get confused by /sys/class/block/. Since 2007/2008 all
1173 major distributions and tools handle this just fine.
1175 Recent distributions and userspace tools after 2009/2010 depend on
1176 the existence of /sys/class/block/, and will not work with this
1179 Only if you are using a new kernel on an old distribution, you might
1182 config SYSFS_DEPRECATED_V2
1183 bool "Enable deprecated sysfs features by default"
1186 depends on SYSFS_DEPRECATED
1188 Enable deprecated sysfs by default.
1190 See the CONFIG_SYSFS_DEPRECATED option for more details about this
1193 Only if you are using a new kernel on an old distribution, you might
1194 need to say Y here. Even then, odds are you would not need it
1195 enabled, you can always pass the boot option if absolutely necessary.
1198 bool "Kernel->user space relay support (formerly relayfs)"
1201 This option enables support for relay interface support in
1202 certain file systems (such as debugfs).
1203 It is designed to provide an efficient mechanism for tools and
1204 facilities to relay large amounts of data from kernel space to
1209 config BLK_DEV_INITRD
1210 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
1212 The initial RAM filesystem is a ramfs which is loaded by the
1213 boot loader (loadlin or lilo) and that is mounted as root
1214 before the normal boot procedure. It is typically used to
1215 load modules needed to mount the "real" root file system,
1216 etc. See <file:Documentation/admin-guide/initrd.rst> for details.
1218 If RAM disk support (BLK_DEV_RAM) is also included, this
1219 also enables initial RAM disk (initrd) support and adds
1220 15 Kbytes (more on some other architectures) to the kernel size.
1226 source "usr/Kconfig"
1231 bool "Boot config support"
1232 select BLK_DEV_INITRD
1234 Extra boot config allows system admin to pass a config file as
1235 complemental extension of kernel cmdline when booting.
1236 The boot config file must be attached at the end of initramfs
1237 with checksum, size and magic word.
1238 See <file:Documentation/admin-guide/bootconfig.rst> for details.
1243 prompt "Compiler optimization level"
1244 default CC_OPTIMIZE_FOR_PERFORMANCE
1246 config CC_OPTIMIZE_FOR_PERFORMANCE
1247 bool "Optimize for performance (-O2)"
1249 This is the default optimization level for the kernel, building
1250 with the "-O2" compiler flag for best performance and most
1251 helpful compile-time warnings.
1253 config CC_OPTIMIZE_FOR_PERFORMANCE_O3
1254 bool "Optimize more for performance (-O3)"
1256 imply CC_DISABLE_WARN_MAYBE_UNINITIALIZED # avoid false positives
1258 Choosing this option will pass "-O3" to your compiler to optimize
1259 the kernel yet more for performance.
1261 config CC_OPTIMIZE_FOR_SIZE
1262 bool "Optimize for size (-Os)"
1263 imply CC_DISABLE_WARN_MAYBE_UNINITIALIZED # avoid false positives
1265 Choosing this option will pass "-Os" to your compiler resulting
1266 in a smaller kernel.
1270 config HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1273 This requires that the arch annotates or otherwise protects
1274 its external entry points from being discarded. Linker scripts
1275 must also merge .text.*, .data.*, and .bss.* correctly into
1276 output sections. Care must be taken not to pull in unrelated
1277 sections (e.g., '.text.init'). Typically '.' in section names
1278 is used to distinguish them from label names / C identifiers.
1280 config LD_DEAD_CODE_DATA_ELIMINATION
1281 bool "Dead code and data elimination (EXPERIMENTAL)"
1282 depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1284 depends on !(FUNCTION_TRACER && CC_IS_GCC && GCC_VERSION < 40800)
1285 depends on $(cc-option,-ffunction-sections -fdata-sections)
1286 depends on $(ld-option,--gc-sections)
1288 Enable this if you want to do dead code and data elimination with
1289 the linker by compiling with -ffunction-sections -fdata-sections,
1290 and linking with --gc-sections.
1292 This can reduce on disk and in-memory size of the kernel
1293 code and static data, particularly for small configs and
1294 on small systems. This has the possibility of introducing
1295 silently broken kernel if the required annotations are not
1296 present. This option is not well tested yet, so use at your
1305 config SYSCTL_EXCEPTION_TRACE
1308 Enable support for /proc/sys/debug/exception-trace.
1310 config SYSCTL_ARCH_UNALIGN_NO_WARN
1313 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1314 Allows arch to define/use @no_unaligned_warning to possibly warn
1315 about unaligned access emulation going on under the hood.
1317 config SYSCTL_ARCH_UNALIGN_ALLOW
1320 Enable support for /proc/sys/kernel/unaligned-trap
1321 Allows arches to define/use @unaligned_enabled to runtime toggle
1322 the unaligned access emulation.
1323 see arch/parisc/kernel/unaligned.c for reference
1325 config HAVE_PCSPKR_PLATFORM
1328 # interpreter that classic socket filters depend on
1333 bool "Configure standard kernel features (expert users)"
1334 # Unhide debug options, to make the on-by-default options visible
1337 This option allows certain base kernel options and settings
1338 to be disabled or tweaked. This is for specialized
1339 environments which can tolerate a "non-standard" kernel.
1340 Only use this if you really know what you are doing.
1343 bool "Enable 16-bit UID system calls" if EXPERT
1344 depends on HAVE_UID16 && MULTIUSER
1347 This enables the legacy 16-bit UID syscall wrappers.
1350 bool "Multiple users, groups and capabilities support" if EXPERT
1353 This option enables support for non-root users, groups and
1356 If you say N here, all processes will run with UID 0, GID 0, and all
1357 possible capabilities. Saying N here also compiles out support for
1358 system calls related to UIDs, GIDs, and capabilities, such as setuid,
1361 If unsure, say Y here.
1363 config SGETMASK_SYSCALL
1364 bool "sgetmask/ssetmask syscalls support" if EXPERT
1365 def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH
1367 sys_sgetmask and sys_ssetmask are obsolete system calls
1368 no longer supported in libc but still enabled by default in some
1371 If unsure, leave the default option here.
1373 config SYSFS_SYSCALL
1374 bool "Sysfs syscall support" if EXPERT
1377 sys_sysfs is an obsolete system call no longer supported in libc.
1378 Note that disabling this option is more secure but might break
1379 compatibility with some systems.
1381 If unsure say Y here.
1384 bool "open by fhandle syscalls" if EXPERT
1388 If you say Y here, a user level program will be able to map
1389 file names to handle and then later use the handle for
1390 different file system operations. This is useful in implementing
1391 userspace file servers, which now track files using handles instead
1392 of names. The handle would remain the same even if file names
1393 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
1397 bool "Posix Clocks & timers" if EXPERT
1400 This includes native support for POSIX timers to the kernel.
1401 Some embedded systems have no use for them and therefore they
1402 can be configured out to reduce the size of the kernel image.
1404 When this option is disabled, the following syscalls won't be
1405 available: timer_create, timer_gettime: timer_getoverrun,
1406 timer_settime, timer_delete, clock_adjtime, getitimer,
1407 setitimer, alarm. Furthermore, the clock_settime, clock_gettime,
1408 clock_getres and clock_nanosleep syscalls will be limited to
1409 CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only.
1415 bool "Enable support for printk" if EXPERT
1418 This option enables normal printk support. Removing it
1419 eliminates most of the message strings from the kernel image
1420 and makes the kernel more or less silent. As this makes it
1421 very difficult to diagnose system problems, saying N here is
1422 strongly discouraged.
1430 bool "BUG() support" if EXPERT
1433 Disabling this option eliminates support for BUG and WARN, reducing
1434 the size of your kernel image and potentially quietly ignoring
1435 numerous fatal conditions. You should only consider disabling this
1436 option for embedded systems with no facilities for reporting errors.
1442 bool "Enable ELF core dumps" if EXPERT
1444 Enable support for generating core dumps. Disabling saves about 4k.
1447 config PCSPKR_PLATFORM
1448 bool "Enable PC-Speaker support" if EXPERT
1449 depends on HAVE_PCSPKR_PLATFORM
1453 This option allows to disable the internal PC-Speaker
1454 support, saving some memory.
1458 bool "Enable full-sized data structures for core" if EXPERT
1460 Disabling this option reduces the size of miscellaneous core
1461 kernel data structures. This saves memory on small machines,
1462 but may reduce performance.
1465 bool "Enable futex support" if EXPERT
1469 Disabling this option will cause the kernel to be built without
1470 support for "fast userspace mutexes". The resulting kernel may not
1471 run glibc-based applications correctly.
1475 depends on FUTEX && RT_MUTEXES
1478 config HAVE_FUTEX_CMPXCHG
1482 Architectures should select this if futex_atomic_cmpxchg_inatomic()
1483 is implemented and always working. This removes a couple of runtime
1487 bool "Enable eventpoll support" if EXPERT
1490 Disabling this option will cause the kernel to be built without
1491 support for epoll family of system calls.
1494 bool "Enable signalfd() system call" if EXPERT
1497 Enable the signalfd() system call that allows to receive signals
1498 on a file descriptor.
1503 bool "Enable timerfd() system call" if EXPERT
1506 Enable the timerfd() system call that allows to receive timer
1507 events on a file descriptor.
1512 bool "Enable eventfd() system call" if EXPERT
1515 Enable the eventfd() system call that allows to receive both
1516 kernel notification (ie. KAIO) or userspace notifications.
1521 bool "Use full shmem filesystem" if EXPERT
1525 The shmem is an internal filesystem used to manage shared memory.
1526 It is backed by swap and manages resource limits. It is also exported
1527 to userspace as tmpfs if TMPFS is enabled. Disabling this
1528 option replaces shmem and tmpfs with the much simpler ramfs code,
1529 which may be appropriate on small systems without swap.
1532 bool "Enable AIO support" if EXPERT
1535 This option enables POSIX asynchronous I/O which may by used
1536 by some high performance threaded applications. Disabling
1537 this option saves about 7k.
1540 bool "Enable IO uring support" if EXPERT
1545 This option enables support for the io_uring interface, enabling
1546 applications to submit and complete IO through submission and
1547 completion rings that are shared between the kernel and application.
1549 config ADVISE_SYSCALLS
1550 bool "Enable madvise/fadvise syscalls" if EXPERT
1553 This option enables the madvise and fadvise syscalls, used by
1554 applications to advise the kernel about their future memory or file
1555 usage, improving performance. If building an embedded system where no
1556 applications use these syscalls, you can disable this option to save
1560 bool "Enable membarrier() system call" if EXPERT
1563 Enable the membarrier() system call that allows issuing memory
1564 barriers across all running threads, which can be used to distribute
1565 the cost of user-space memory barriers asymmetrically by transforming
1566 pairs of memory barriers into pairs consisting of membarrier() and a
1572 bool "Load all symbols for debugging/ksymoops" if EXPERT
1575 Say Y here to let the kernel print out symbolic crash information and
1576 symbolic stack backtraces. This increases the size of the kernel
1577 somewhat, as all symbols have to be loaded into the kernel image.
1580 bool "Include all symbols in kallsyms"
1581 depends on DEBUG_KERNEL && KALLSYMS
1583 Normally kallsyms only contains the symbols of functions for nicer
1584 OOPS messages and backtraces (i.e., symbols from the text and inittext
1585 sections). This is sufficient for most cases. And only in very rare
1586 cases (e.g., when a debugger is used) all symbols are required (e.g.,
1587 names of variables from the data sections, etc).
1589 This option makes sure that all symbols are loaded into the kernel
1590 image (i.e., symbols from all sections) in cost of increased kernel
1591 size (depending on the kernel configuration, it may be 300KiB or
1592 something like this).
1594 Say N unless you really need all symbols.
1596 config KALLSYMS_ABSOLUTE_PERCPU
1599 default X86_64 && SMP
1601 config KALLSYMS_BASE_RELATIVE
1606 Instead of emitting them as absolute values in the native word size,
1607 emit the symbol references in the kallsyms table as 32-bit entries,
1608 each containing a relative value in the range [base, base + U32_MAX]
1609 or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either
1610 an absolute value in the range [0, S32_MAX] or a relative value in the
1611 range [base, base + S32_MAX], where base is the lowest relative symbol
1612 address encountered in the image.
1614 On 64-bit builds, this reduces the size of the address table by 50%,
1615 but more importantly, it results in entries whose values are build
1616 time constants, and no relocation pass is required at runtime to fix
1617 up the entries based on the runtime load address of the kernel.
1619 # end of the "standard kernel features (expert users)" menu
1621 # syscall, maps, verifier
1623 bool "Enable bpf() system call"
1628 Enable the bpf() system call that allows to manipulate eBPF
1629 programs and maps via file descriptors.
1631 config ARCH_WANT_DEFAULT_BPF_JIT
1634 config BPF_JIT_ALWAYS_ON
1635 bool "Permanently enable BPF JIT and remove BPF interpreter"
1636 depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT
1638 Enables BPF JIT and removes BPF interpreter to avoid
1639 speculative execution of BPF instructions by the interpreter
1641 config BPF_JIT_DEFAULT_ON
1642 def_bool ARCH_WANT_DEFAULT_BPF_JIT || BPF_JIT_ALWAYS_ON
1643 depends on HAVE_EBPF_JIT && BPF_JIT
1646 bool "Enable userfaultfd() system call"
1649 Enable the userfaultfd() system call that allows to intercept and
1650 handle page faults in userland.
1652 config ARCH_HAS_MEMBARRIER_CALLBACKS
1655 config ARCH_HAS_MEMBARRIER_SYNC_CORE
1659 bool "Enable rseq() system call" if EXPERT
1661 depends on HAVE_RSEQ
1664 Enable the restartable sequences system call. It provides a
1665 user-space cache for the current CPU number value, which
1666 speeds up getting the current CPU number from user-space,
1667 as well as an ABI to speed up user-space operations on
1674 bool "Enabled debugging of rseq() system call" if EXPERT
1675 depends on RSEQ && DEBUG_KERNEL
1677 Enable extra debugging checks for the rseq system call.
1682 bool "Embedded system"
1683 option allnoconfig_y
1686 This option should be enabled if compiling the kernel for
1687 an embedded system so certain expert options are available
1690 config HAVE_PERF_EVENTS
1693 See tools/perf/design.txt for details.
1695 config PERF_USE_VMALLOC
1698 See tools/perf/design.txt for details
1701 bool "PC/104 support" if EXPERT
1703 Expose PC/104 form factor device drivers and options available for
1704 selection and configuration. Enable this option if your target
1705 machine has a PC/104 bus.
1707 menu "Kernel Performance Events And Counters"
1710 bool "Kernel performance events and counters"
1711 default y if PROFILING
1712 depends on HAVE_PERF_EVENTS
1716 Enable kernel support for various performance events provided
1717 by software and hardware.
1719 Software events are supported either built-in or via the
1720 use of generic tracepoints.
1722 Most modern CPUs support performance events via performance
1723 counter registers. These registers count the number of certain
1724 types of hw events: such as instructions executed, cachemisses
1725 suffered, or branches mis-predicted - without slowing down the
1726 kernel or applications. These registers can also trigger interrupts
1727 when a threshold number of events have passed - and can thus be
1728 used to profile the code that runs on that CPU.
1730 The Linux Performance Event subsystem provides an abstraction of
1731 these software and hardware event capabilities, available via a
1732 system call and used by the "perf" utility in tools/perf/. It
1733 provides per task and per CPU counters, and it provides event
1734 capabilities on top of those.
1738 config DEBUG_PERF_USE_VMALLOC
1740 bool "Debug: use vmalloc to back perf mmap() buffers"
1741 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC
1742 select PERF_USE_VMALLOC
1744 Use vmalloc memory to back perf mmap() buffers.
1746 Mostly useful for debugging the vmalloc code on platforms
1747 that don't require it.
1753 config VM_EVENT_COUNTERS
1755 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1757 VM event counters are needed for event counts to be shown.
1758 This option allows the disabling of the VM event counters
1759 on EXPERT systems. /proc/vmstat will only show page counts
1760 if VM event counters are disabled.
1764 bool "Enable SLUB debugging support" if EXPERT
1765 depends on SLUB && SYSFS
1767 SLUB has extensive debug support features. Disabling these can
1768 result in significant savings in code size. This also disables
1769 SLUB sysfs support. /sys/slab will not exist and there will be
1770 no support for cache validation etc.
1772 config SLUB_MEMCG_SYSFS_ON
1774 bool "Enable memcg SLUB sysfs support by default" if EXPERT
1775 depends on SLUB && SYSFS && MEMCG
1777 SLUB creates a directory under /sys/kernel/slab for each
1778 allocation cache to host info and debug files. If memory
1779 cgroup is enabled, each cache can have per memory cgroup
1780 caches. SLUB can create the same sysfs directories for these
1781 caches under /sys/kernel/slab/CACHE/cgroup but it can lead
1782 to a very high number of debug files being created. This is
1783 controlled by slub_memcg_sysfs boot parameter and this
1784 config option determines the parameter's default value.
1787 bool "Disable heap randomization"
1790 Randomizing heap placement makes heap exploits harder, but it
1791 also breaks ancient binaries (including anything libc5 based).
1792 This option changes the bootup default to heap randomization
1793 disabled, and can be overridden at runtime by setting
1794 /proc/sys/kernel/randomize_va_space to 2.
1796 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1799 prompt "Choose SLAB allocator"
1802 This option allows to select a slab allocator.
1806 select HAVE_HARDENED_USERCOPY_ALLOCATOR
1808 The regular slab allocator that is established and known to work
1809 well in all environments. It organizes cache hot objects in
1810 per cpu and per node queues.
1813 bool "SLUB (Unqueued Allocator)"
1814 select HAVE_HARDENED_USERCOPY_ALLOCATOR
1816 SLUB is a slab allocator that minimizes cache line usage
1817 instead of managing queues of cached objects (SLAB approach).
1818 Per cpu caching is realized using slabs of objects instead
1819 of queues of objects. SLUB can use memory efficiently
1820 and has enhanced diagnostics. SLUB is the default choice for
1825 bool "SLOB (Simple Allocator)"
1827 SLOB replaces the stock allocator with a drastically simpler
1828 allocator. SLOB is generally more space efficient but
1829 does not perform as well on large systems.
1833 config SLAB_MERGE_DEFAULT
1834 bool "Allow slab caches to be merged"
1837 For reduced kernel memory fragmentation, slab caches can be
1838 merged when they share the same size and other characteristics.
1839 This carries a risk of kernel heap overflows being able to
1840 overwrite objects from merged caches (and more easily control
1841 cache layout), which makes such heap attacks easier to exploit
1842 by attackers. By keeping caches unmerged, these kinds of exploits
1843 can usually only damage objects in the same cache. To disable
1844 merging at runtime, "slab_nomerge" can be passed on the kernel
1847 config SLAB_FREELIST_RANDOM
1849 depends on SLAB || SLUB
1850 bool "SLAB freelist randomization"
1852 Randomizes the freelist order used on creating new pages. This
1853 security feature reduces the predictability of the kernel slab
1854 allocator against heap overflows.
1856 config SLAB_FREELIST_HARDENED
1857 bool "Harden slab freelist metadata"
1860 Many kernel heap attacks try to target slab cache metadata and
1861 other infrastructure. This options makes minor performance
1862 sacrifices to harden the kernel slab allocator against common
1863 freelist exploit methods.
1865 config SHUFFLE_PAGE_ALLOCATOR
1866 bool "Page allocator randomization"
1867 default SLAB_FREELIST_RANDOM && ACPI_NUMA
1869 Randomization of the page allocator improves the average
1870 utilization of a direct-mapped memory-side-cache. See section
1871 5.2.27 Heterogeneous Memory Attribute Table (HMAT) in the ACPI
1872 6.2a specification for an example of how a platform advertises
1873 the presence of a memory-side-cache. There are also incidental
1874 security benefits as it reduces the predictability of page
1875 allocations to compliment SLAB_FREELIST_RANDOM, but the
1876 default granularity of shuffling on the "MAX_ORDER - 1" i.e,
1877 10th order of pages is selected based on cache utilization
1880 While the randomization improves cache utilization it may
1881 negatively impact workloads on platforms without a cache. For
1882 this reason, by default, the randomization is enabled only
1883 after runtime detection of a direct-mapped memory-side-cache.
1884 Otherwise, the randomization may be force enabled with the
1885 'page_alloc.shuffle' kernel command line parameter.
1889 config SLUB_CPU_PARTIAL
1891 depends on SLUB && SMP
1892 bool "SLUB per cpu partial cache"
1894 Per cpu partial caches accelerate objects allocation and freeing
1895 that is local to a processor at the price of more indeterminism
1896 in the latency of the free. On overflow these caches will be cleared
1897 which requires the taking of locks that may cause latency spikes.
1898 Typically one would choose no for a realtime system.
1900 config MMAP_ALLOW_UNINITIALIZED
1901 bool "Allow mmapped anonymous memory to be uninitialized"
1902 depends on EXPERT && !MMU
1905 Normally, and according to the Linux spec, anonymous memory obtained
1906 from mmap() has its contents cleared before it is passed to
1907 userspace. Enabling this config option allows you to request that
1908 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1909 providing a huge performance boost. If this option is not enabled,
1910 then the flag will be ignored.
1912 This is taken advantage of by uClibc's malloc(), and also by
1913 ELF-FDPIC binfmt's brk and stack allocator.
1915 Because of the obvious security issues, this option should only be
1916 enabled on embedded devices where you control what is run in
1917 userspace. Since that isn't generally a problem on no-MMU systems,
1918 it is normally safe to say Y here.
1920 See Documentation/nommu-mmap.txt for more information.
1922 config SYSTEM_DATA_VERIFICATION
1924 select SYSTEM_TRUSTED_KEYRING
1928 select ASYMMETRIC_KEY_TYPE
1929 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
1932 select X509_CERTIFICATE_PARSER
1933 select PKCS7_MESSAGE_PARSER
1935 Provide PKCS#7 message verification using the contents of the system
1936 trusted keyring to provide public keys. This then can be used for
1937 module verification, kexec image verification and firmware blob
1941 bool "Profiling support"
1943 Say Y here to enable the extended profiling support mechanisms used
1944 by profilers such as OProfile.
1947 # Place an empty function call at each tracepoint site. Can be
1948 # dynamically changed for a probe function.
1953 endmenu # General setup
1955 source "arch/Kconfig"
1962 default 0 if BASE_FULL
1963 default 1 if !BASE_FULL
1965 config MODULE_SIG_FORMAT
1967 select SYSTEM_DATA_VERIFICATION
1970 bool "Enable loadable module support"
1973 Kernel modules are small pieces of compiled code which can
1974 be inserted in the running kernel, rather than being
1975 permanently built into the kernel. You use the "modprobe"
1976 tool to add (and sometimes remove) them. If you say Y here,
1977 many parts of the kernel can be built as modules (by
1978 answering M instead of Y where indicated): this is most
1979 useful for infrequently used options which are not required
1980 for booting. For more information, see the man pages for
1981 modprobe, lsmod, modinfo, insmod and rmmod.
1983 If you say Y here, you will need to run "make
1984 modules_install" to put the modules under /lib/modules/
1985 where modprobe can find them (you may need to be root to do
1992 config MODULE_FORCE_LOAD
1993 bool "Forced module loading"
1996 Allow loading of modules without version information (ie. modprobe
1997 --force). Forced module loading sets the 'F' (forced) taint flag and
1998 is usually a really bad idea.
2000 config MODULE_UNLOAD
2001 bool "Module unloading"
2003 Without this option you will not be able to unload any
2004 modules (note that some modules may not be unloadable
2005 anyway), which makes your kernel smaller, faster
2006 and simpler. If unsure, say Y.
2008 config MODULE_FORCE_UNLOAD
2009 bool "Forced module unloading"
2010 depends on MODULE_UNLOAD
2012 This option allows you to force a module to unload, even if the
2013 kernel believes it is unsafe: the kernel will remove the module
2014 without waiting for anyone to stop using it (using the -f option to
2015 rmmod). This is mainly for kernel developers and desperate users.
2019 bool "Module versioning support"
2021 Usually, you have to use modules compiled with your kernel.
2022 Saying Y here makes it sometimes possible to use modules
2023 compiled for different kernels, by adding enough information
2024 to the modules to (hopefully) spot any changes which would
2025 make them incompatible with the kernel you are running. If
2028 config ASM_MODVERSIONS
2030 default HAVE_ASM_MODVERSIONS && MODVERSIONS
2032 This enables module versioning for exported symbols also from
2033 assembly. This can be enabled only when the target architecture
2036 config MODULE_REL_CRCS
2038 depends on MODVERSIONS
2040 config MODULE_SRCVERSION_ALL
2041 bool "Source checksum for all modules"
2043 Modules which contain a MODULE_VERSION get an extra "srcversion"
2044 field inserted into their modinfo section, which contains a
2045 sum of the source files which made it. This helps maintainers
2046 see exactly which source was used to build a module (since
2047 others sometimes change the module source without updating
2048 the version). With this option, such a "srcversion" field
2049 will be created for all modules. If unsure, say N.
2052 bool "Module signature verification"
2053 select MODULE_SIG_FORMAT
2055 Check modules for valid signatures upon load: the signature
2056 is simply appended to the module. For more information see
2057 <file:Documentation/admin-guide/module-signing.rst>.
2059 Note that this option adds the OpenSSL development packages as a
2060 kernel build dependency so that the signing tool can use its crypto
2063 You should enable this option if you wish to use either
2064 CONFIG_SECURITY_LOCKDOWN_LSM or lockdown functionality imposed via
2065 another LSM - otherwise unsigned modules will be loadable regardless
2066 of the lockdown policy.
2068 !!!WARNING!!! If you enable this option, you MUST make sure that the
2069 module DOES NOT get stripped after being signed. This includes the
2070 debuginfo strip done by some packagers (such as rpmbuild) and
2071 inclusion into an initramfs that wants the module size reduced.
2073 config MODULE_SIG_FORCE
2074 bool "Require modules to be validly signed"
2075 depends on MODULE_SIG
2077 Reject unsigned modules or signed modules for which we don't have a
2078 key. Without this, such modules will simply taint the kernel.
2080 config MODULE_SIG_ALL
2081 bool "Automatically sign all modules"
2083 depends on MODULE_SIG
2085 Sign all modules during make modules_install. Without this option,
2086 modules must be signed manually, using the scripts/sign-file tool.
2088 comment "Do not forget to sign required modules with scripts/sign-file"
2089 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL
2092 prompt "Which hash algorithm should modules be signed with?"
2093 depends on MODULE_SIG
2095 This determines which sort of hashing algorithm will be used during
2096 signature generation. This algorithm _must_ be built into the kernel
2097 directly so that signature verification can take place. It is not
2098 possible to load a signed module containing the algorithm to check
2099 the signature on that module.
2101 config MODULE_SIG_SHA1
2102 bool "Sign modules with SHA-1"
2105 config MODULE_SIG_SHA224
2106 bool "Sign modules with SHA-224"
2107 select CRYPTO_SHA256
2109 config MODULE_SIG_SHA256
2110 bool "Sign modules with SHA-256"
2111 select CRYPTO_SHA256
2113 config MODULE_SIG_SHA384
2114 bool "Sign modules with SHA-384"
2115 select CRYPTO_SHA512
2117 config MODULE_SIG_SHA512
2118 bool "Sign modules with SHA-512"
2119 select CRYPTO_SHA512
2123 config MODULE_SIG_HASH
2125 depends on MODULE_SIG
2126 default "sha1" if MODULE_SIG_SHA1
2127 default "sha224" if MODULE_SIG_SHA224
2128 default "sha256" if MODULE_SIG_SHA256
2129 default "sha384" if MODULE_SIG_SHA384
2130 default "sha512" if MODULE_SIG_SHA512
2132 config MODULE_COMPRESS
2133 bool "Compress modules on installation"
2136 Compresses kernel modules when 'make modules_install' is run; gzip or
2137 xz depending on "Compression algorithm" below.
2139 module-init-tools MAY support gzip, and kmod MAY support gzip and xz.
2141 Out-of-tree kernel modules installed using Kbuild will also be
2142 compressed upon installation.
2144 Note: for modules inside an initrd or initramfs, it's more efficient
2145 to compress the whole initrd or initramfs instead.
2147 Note: This is fully compatible with signed modules.
2152 prompt "Compression algorithm"
2153 depends on MODULE_COMPRESS
2154 default MODULE_COMPRESS_GZIP
2156 This determines which sort of compression will be used during
2157 'make modules_install'.
2159 GZIP (default) and XZ are supported.
2161 config MODULE_COMPRESS_GZIP
2164 config MODULE_COMPRESS_XZ
2169 config MODULE_ALLOW_MISSING_NAMESPACE_IMPORTS
2170 bool "Allow loading of modules with missing namespace imports"
2172 Symbols exported with EXPORT_SYMBOL_NS*() are considered exported in
2173 a namespace. A module that makes use of a symbol exported with such a
2174 namespace is required to import the namespace via MODULE_IMPORT_NS().
2175 There is no technical reason to enforce correct namespace imports,
2176 but it creates consistency between symbols defining namespaces and
2177 users importing namespaces they make use of. This option relaxes this
2178 requirement and lifts the enforcement when loading a module.
2182 config UNUSED_SYMBOLS
2183 bool "Enable unused/obsolete exported symbols"
2186 Unused but exported symbols make the kernel needlessly bigger. For
2187 that reason most of these unused exports will soon be removed. This
2188 option is provided temporarily to provide a transition period in case
2189 some external kernel module needs one of these symbols anyway. If you
2190 encounter such a case in your module, consider if you are actually
2191 using the right API. (rationale: since nobody in the kernel is using
2192 this in a module, there is a pretty good chance it's actually the
2193 wrong interface to use). If you really need the symbol, please send a
2194 mail to the linux kernel mailing list mentioning the symbol and why
2195 you really need it, and what the merge plan to the mainline kernel for
2198 config TRIM_UNUSED_KSYMS
2199 bool "Trim unused exported kernel symbols"
2200 depends on !UNUSED_SYMBOLS
2202 The kernel and some modules make many symbols available for
2203 other modules to use via EXPORT_SYMBOL() and variants. Depending
2204 on the set of modules being selected in your kernel configuration,
2205 many of those exported symbols might never be used.
2207 This option allows for unused exported symbols to be dropped from
2208 the build. In turn, this provides the compiler more opportunities
2209 (especially when using LTO) for optimizing the code and reducing
2210 binary size. This might have some security advantages as well.
2212 If unsure, or if you need to build out-of-tree modules, say N.
2216 config MODULES_TREE_LOOKUP
2218 depends on PERF_EVENTS || TRACING
2220 config INIT_ALL_POSSIBLE
2223 Back when each arch used to define their own cpu_online_mask and
2224 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
2225 with all 1s, and others with all 0s. When they were centralised,
2226 it was better to provide this option than to break all the archs
2227 and have several arch maintainers pursuing me down dark alleys.
2229 source "block/Kconfig"
2231 config PREEMPT_NOTIFIERS
2241 Build a simple ASN.1 grammar compiler that produces a bytecode output
2242 that can be interpreted by the ASN.1 stream decoder and used to
2243 inform it as to what tags are to be expected in a stream and what
2244 functions to call on what tags.
2246 source "kernel/Kconfig.locks"
2248 config ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
2251 # It may be useful for an architecture to override the definitions of the
2252 # SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h>
2253 # and the COMPAT_ variants in <linux/compat.h>, in particular to use a
2254 # different calling convention for syscalls. They can also override the
2255 # macros for not-implemented syscalls in kernel/sys_ni.c and
2256 # kernel/time/posix-stubs.c. All these overrides need to be available in
2257 # <asm/syscall_wrapper.h>.
2258 config ARCH_HAS_SYSCALL_WRAPPER
projects / linux.git / blob