1 # SPDX-License-Identifier: GPL-2.0-only
4 default "$(CC_VERSION_TEXT)"
6 This is used in unclear ways:
8 - Re-run Kconfig when the compiler is updated
9 The 'default' property references the environment variable,
10 CC_VERSION_TEXT so it is recorded in include/config/auto.conf.cmd.
11 When the compiler is updated, Kconfig will be invoked.
13 - Ensure full rebuild when the compiler is updated
14 include/linux/compiler-version.h contains this option in the comment
15 line so fixdep adds include/config/CC_VERSION_TEXT into the
16 auto-generated dependency. When the compiler is updated, syncconfig
17 will touch it and then every file will be rebuilt.
20 def_bool $(success,test "$(cc-name)" = GCC)
24 default $(cc-version) if CC_IS_GCC
28 def_bool $(success,test "$(cc-name)" = Clang)
32 default $(cc-version) if CC_IS_CLANG
36 def_bool $(success,test "$(as-name)" = GNU)
39 def_bool $(success,test "$(as-name)" = LLVM)
43 # Use clang version if this is the integrated assembler
44 default CLANG_VERSION if AS_IS_LLVM
48 def_bool $(success,test "$(ld-name)" = BFD)
52 default $(ld-version) if LD_IS_BFD
56 def_bool $(success,test "$(ld-name)" = LLD)
60 default $(ld-version) if LD_IS_LLD
63 config RUST_IS_AVAILABLE
64 def_bool $(success,$(srctree)/scripts/rust_is_available.sh)
66 This shows whether a suitable Rust toolchain is available (found).
68 Please see Documentation/rust/quick-start.rst for instructions on how
69 to satisfy the build requirements of Rust support.
71 In particular, the Makefile target 'rustavailable' is useful to check
72 why the Rust toolchain is not being detected.
76 default $(success,$(srctree)/scripts/cc-can-link.sh $(CC) $(CLANG_FLAGS) $(USERCFLAGS) $(USERLDFLAGS) $(m64-flag)) if 64BIT
77 default $(success,$(srctree)/scripts/cc-can-link.sh $(CC) $(CLANG_FLAGS) $(USERCFLAGS) $(USERLDFLAGS) $(m32-flag))
79 config CC_CAN_LINK_STATIC
81 default $(success,$(srctree)/scripts/cc-can-link.sh $(CC) $(CLANG_FLAGS) $(USERCFLAGS) $(USERLDFLAGS) $(m64-flag) -static) if 64BIT
82 default $(success,$(srctree)/scripts/cc-can-link.sh $(CC) $(CLANG_FLAGS) $(USERCFLAGS) $(USERLDFLAGS) $(m32-flag) -static)
84 config CC_HAS_ASM_GOTO_OUTPUT
85 def_bool $(success,echo 'int foo(int x) { asm goto ("": "=r"(x) ::: bar); return x; bar: return 0; }' | $(CC) -x c - -c -o /dev/null)
87 config CC_HAS_ASM_GOTO_TIED_OUTPUT
88 depends on CC_HAS_ASM_GOTO_OUTPUT
89 # Detect buggy gcc and clang, fixed in gcc-11 clang-14.
90 def_bool $(success,echo 'int foo(int *x) { asm goto (".long (%l[bar]) - .": "+m"(*x) ::: bar); return *x; bar: return 0; }' | $CC -x c - -c -o /dev/null)
92 config TOOLS_SUPPORT_RELR
93 def_bool $(success,env "CC=$(CC)" "LD=$(LD)" "NM=$(NM)" "OBJCOPY=$(OBJCOPY)" $(srctree)/scripts/tools-support-relr.sh)
95 config CC_HAS_ASM_INLINE
96 def_bool $(success,echo 'void foo(void) { asm inline (""); }' | $(CC) -x c - -c -o /dev/null)
98 config CC_HAS_NO_PROFILE_FN_ATTR
99 def_bool $(success,echo '__attribute__((no_profile_instrument_function)) int x();' | $(CC) -x c - -c -o /dev/null -Werror)
101 config PAHOLE_VERSION
103 default $(shell,$(srctree)/scripts/pahole-version.sh $(PAHOLE))
111 config BUILDTIME_TABLE_SORT
114 config THREAD_INFO_IN_TASK
117 Select this to move thread_info off the stack into task_struct. To
118 make this work, an arch will need to remove all thread_info fields
119 except flags and fix any runtime bugs.
121 One subtle change that will be needed is to use try_get_task_stack()
122 and put_task_stack() in save_thread_stack_tsk() and get_wchan().
131 depends on BROKEN || !SMP
134 config INIT_ENV_ARG_LIMIT
139 Maximum of each of the number of arguments and environment
140 variables passed to init from the kernel command line.
143 bool "Compile also drivers which will not load"
146 Some drivers can be compiled on a different platform than they are
147 intended to be run on. Despite they cannot be loaded there (or even
148 when they load they cannot be used due to missing HW support),
149 developers still, opposing to distributors, might want to build such
150 drivers to compile-test them.
152 If you are a developer and want to build everything available, say Y
153 here. If you are a user/distributor, say N here to exclude useless
154 drivers to be distributed.
157 bool "Compile the kernel with warnings as errors"
160 A kernel build should not cause any compiler warnings, and this
161 enables the '-Werror' (for C) and '-Dwarnings' (for Rust) flags
162 to enforce that rule by default. Certain warnings from other tools
163 such as the linker may be upgraded to errors with this option as
166 However, if you have a new (or very old) compiler or linker with odd
167 and unusual warnings, or you have some architecture with problems,
168 you may need to disable this config option in order to
169 successfully build the kernel.
173 config UAPI_HEADER_TEST
174 bool "Compile test UAPI headers"
175 depends on HEADERS_INSTALL && CC_CAN_LINK
177 Compile test headers exported to user-space to ensure they are
178 self-contained, i.e. compilable as standalone units.
180 If you are a developer or tester and want to ensure the exported
181 headers are self-contained, say Y here. Otherwise, choose N.
184 string "Local version - append to kernel release"
186 Append an extra string to the end of your kernel version.
187 This will show up when you type uname, for example.
188 The string you set here will be appended after the contents of
189 any files with a filename matching localversion* in your
190 object and source tree, in that order. Your total string can
191 be a maximum of 64 characters.
193 config LOCALVERSION_AUTO
194 bool "Automatically append version information to the version string"
196 depends on !COMPILE_TEST
198 This will try to automatically determine if the current tree is a
199 release tree by looking for git tags that belong to the current
200 top of tree revision.
202 A string of the format -gxxxxxxxx will be added to the localversion
203 if a git-based tree is found. The string generated by this will be
204 appended after any matching localversion* files, and after the value
205 set in CONFIG_LOCALVERSION.
207 (The actual string used here is the first 12 characters produced
208 by running the command:
210 $ git rev-parse --verify HEAD
212 which is done within the script "scripts/setlocalversion".)
215 string "Build ID Salt"
218 The build ID is used to link binaries and their debug info. Setting
219 this option will use the value in the calculation of the build id.
220 This is mostly useful for distributions which want to ensure the
221 build is unique between builds. It's safe to leave the default.
223 config HAVE_KERNEL_GZIP
226 config HAVE_KERNEL_BZIP2
229 config HAVE_KERNEL_LZMA
232 config HAVE_KERNEL_XZ
235 config HAVE_KERNEL_LZO
238 config HAVE_KERNEL_LZ4
241 config HAVE_KERNEL_ZSTD
244 config HAVE_KERNEL_UNCOMPRESSED
248 prompt "Kernel compression mode"
250 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_ZSTD || HAVE_KERNEL_UNCOMPRESSED
252 The linux kernel is a kind of self-extracting executable.
253 Several compression algorithms are available, which differ
254 in efficiency, compression and decompression speed.
255 Compression speed is only relevant when building a kernel.
256 Decompression speed is relevant at each boot.
258 If you have any problems with bzip2 or lzma compressed
260 version of this functionality (bzip2 only), for 2.4, was
261 supplied by Christian Ludwig)
263 High compression options are mostly useful for users, who
264 are low on disk space (embedded systems), but for whom ram
267 If in doubt, select 'gzip'
271 depends on HAVE_KERNEL_GZIP
273 The old and tried gzip compression. It provides a good balance
274 between compression ratio and decompression speed.
278 depends on HAVE_KERNEL_BZIP2
280 Its compression ratio and speed is intermediate.
281 Decompression speed is slowest among the choices. The kernel
282 size is about 10% smaller with bzip2, in comparison to gzip.
283 Bzip2 uses a large amount of memory. For modern kernels you
284 will need at least 8MB RAM or more for booting.
288 depends on HAVE_KERNEL_LZMA
290 This compression algorithm's ratio is best. Decompression speed
291 is between gzip and bzip2. Compression is slowest.
292 The kernel size is about 33% smaller with LZMA in comparison to gzip.
296 depends on HAVE_KERNEL_XZ
298 XZ uses the LZMA2 algorithm and instruction set specific
299 BCJ filters which can improve compression ratio of executable
300 code. The size of the kernel is about 30% smaller with XZ in
301 comparison to gzip. On architectures for which there is a BCJ
302 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
303 will create a few percent smaller kernel than plain LZMA.
305 The speed is about the same as with LZMA: The decompression
306 speed of XZ is better than that of bzip2 but worse than gzip
307 and LZO. Compression is slow.
311 depends on HAVE_KERNEL_LZO
313 Its compression ratio is the poorest among the choices. The kernel
314 size is about 10% bigger than gzip; however its speed
315 (both compression and decompression) is the fastest.
319 depends on HAVE_KERNEL_LZ4
321 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
322 A preliminary version of LZ4 de/compression tool is available at
323 <https://code.google.com/p/lz4/>.
325 Its compression ratio is worse than LZO. The size of the kernel
326 is about 8% bigger than LZO. But the decompression speed is
331 depends on HAVE_KERNEL_ZSTD
333 ZSTD is a compression algorithm targeting intermediate compression
334 with fast decompression speed. It will compress better than GZIP and
335 decompress around the same speed as LZO, but slower than LZ4. You
336 will need at least 192 KB RAM or more for booting. The zstd command
337 line tool is required for compression.
339 config KERNEL_UNCOMPRESSED
341 depends on HAVE_KERNEL_UNCOMPRESSED
343 Produce uncompressed kernel image. This option is usually not what
344 you want. It is useful for debugging the kernel in slow simulation
345 environments, where decompressing and moving the kernel is awfully
346 slow. This option allows early boot code to skip the decompressor
347 and jump right at uncompressed kernel image.
352 string "Default init path"
355 This option determines the default init for the system if no init=
356 option is passed on the kernel command line. If the requested path is
357 not present, we will still then move on to attempting further
358 locations (e.g. /sbin/init, etc). If this is empty, we will just use
359 the fallback list when init= is not passed.
361 config DEFAULT_HOSTNAME
362 string "Default hostname"
365 This option determines the default system hostname before userspace
366 calls sethostname(2). The kernel traditionally uses "(none)" here,
367 but you may wish to use a different default here to make a minimal
368 system more usable with less configuration.
373 Inter Process Communication is a suite of library functions and
374 system calls which let processes (running programs) synchronize and
375 exchange information. It is generally considered to be a good thing,
376 and some programs won't run unless you say Y here. In particular, if
377 you want to run the DOS emulator dosemu under Linux (read the
378 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
379 you'll need to say Y here.
381 You can find documentation about IPC with "info ipc" and also in
382 section 6.4 of the Linux Programmer's Guide, available from
383 <http://www.tldp.org/guides.html>.
385 config SYSVIPC_SYSCTL
391 config SYSVIPC_COMPAT
393 depends on COMPAT && SYSVIPC
396 bool "POSIX Message Queues"
399 POSIX variant of message queues is a part of IPC. In POSIX message
400 queues every message has a priority which decides about succession
401 of receiving it by a process. If you want to compile and run
402 programs written e.g. for Solaris with use of its POSIX message
403 queues (functions mq_*) say Y here.
405 POSIX message queues are visible as a filesystem called 'mqueue'
406 and can be mounted somewhere if you want to do filesystem
407 operations on message queues.
411 config POSIX_MQUEUE_SYSCTL
413 depends on POSIX_MQUEUE
418 bool "General notification queue"
422 This is a general notification queue for the kernel to pass events to
423 userspace by splicing them into pipes. It can be used in conjunction
424 with watches for key/keyring change notifications and device
427 See Documentation/core-api/watch_queue.rst
429 config CROSS_MEMORY_ATTACH
430 bool "Enable process_vm_readv/writev syscalls"
434 Enabling this option adds the system calls process_vm_readv and
435 process_vm_writev which allow a process with the correct privileges
436 to directly read from or write to another process' address space.
437 See the man page for more details.
440 bool "uselib syscall (for libc5 and earlier)"
441 default ALPHA || M68K || SPARC
443 This option enables the uselib syscall, a system call used in the
444 dynamic linker from libc5 and earlier. glibc does not use this
445 system call. If you intend to run programs built on libc5 or
446 earlier, you may need to enable this syscall. Current systems
447 running glibc can safely disable this.
450 bool "Auditing support"
453 Enable auditing infrastructure that can be used with another
454 kernel subsystem, such as SELinux (which requires this for
455 logging of avc messages output). System call auditing is included
456 on architectures which support it.
458 config HAVE_ARCH_AUDITSYSCALL
463 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL
466 source "kernel/irq/Kconfig"
467 source "kernel/time/Kconfig"
468 source "kernel/bpf/Kconfig"
469 source "kernel/Kconfig.preempt"
471 menu "CPU/Task time and stats accounting"
473 config VIRT_CPU_ACCOUNTING
477 prompt "Cputime accounting"
478 default TICK_CPU_ACCOUNTING
480 # Kind of a stub config for the pure tick based cputime accounting
481 config TICK_CPU_ACCOUNTING
482 bool "Simple tick based cputime accounting"
483 depends on !S390 && !NO_HZ_FULL
485 This is the basic tick based cputime accounting that maintains
486 statistics about user, system and idle time spent on per jiffies
491 config VIRT_CPU_ACCOUNTING_NATIVE
492 bool "Deterministic task and CPU time accounting"
493 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL
494 select VIRT_CPU_ACCOUNTING
496 Select this option to enable more accurate task and CPU time
497 accounting. This is done by reading a CPU counter on each
498 kernel entry and exit and on transitions within the kernel
499 between system, softirq and hardirq state, so there is a
500 small performance impact. In the case of s390 or IBM POWER > 5,
501 this also enables accounting of stolen time on logically-partitioned
504 config VIRT_CPU_ACCOUNTING_GEN
505 bool "Full dynticks CPU time accounting"
506 depends on HAVE_CONTEXT_TRACKING_USER
507 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN
508 depends on GENERIC_CLOCKEVENTS
509 select VIRT_CPU_ACCOUNTING
510 select CONTEXT_TRACKING_USER
512 Select this option to enable task and CPU time accounting on full
513 dynticks systems. This accounting is implemented by watching every
514 kernel-user boundaries using the context tracking subsystem.
515 The accounting is thus performed at the expense of some significant
518 For now this is only useful if you are working on the full
519 dynticks subsystem development.
525 config IRQ_TIME_ACCOUNTING
526 bool "Fine granularity task level IRQ time accounting"
527 depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE
529 Select this option to enable fine granularity task irq time
530 accounting. This is done by reading a timestamp on each
531 transitions between softirq and hardirq state, so there can be a
532 small performance impact.
534 If in doubt, say N here.
536 config HAVE_SCHED_AVG_IRQ
538 depends on IRQ_TIME_ACCOUNTING || PARAVIRT_TIME_ACCOUNTING
541 config SCHED_THERMAL_PRESSURE
543 default y if ARM && ARM_CPU_TOPOLOGY
546 depends on CPU_FREQ_THERMAL
548 Select this option to enable thermal pressure accounting in the
549 scheduler. Thermal pressure is the value conveyed to the scheduler
550 that reflects the reduction in CPU compute capacity resulted from
551 thermal throttling. Thermal throttling occurs when the performance of
552 a CPU is capped due to high operating temperatures.
554 If selected, the scheduler will be able to balance tasks accordingly,
555 i.e. put less load on throttled CPUs than on non/less throttled ones.
557 This requires the architecture to implement
558 arch_update_thermal_pressure() and arch_scale_thermal_pressure().
560 config BSD_PROCESS_ACCT
561 bool "BSD Process Accounting"
564 If you say Y here, a user level program will be able to instruct the
565 kernel (via a special system call) to write process accounting
566 information to a file: whenever a process exits, information about
567 that process will be appended to the file by the kernel. The
568 information includes things such as creation time, owning user,
569 command name, memory usage, controlling terminal etc. (the complete
570 list is in the struct acct in <file:include/linux/acct.h>). It is
571 up to the user level program to do useful things with this
572 information. This is generally a good idea, so say Y.
574 config BSD_PROCESS_ACCT_V3
575 bool "BSD Process Accounting version 3 file format"
576 depends on BSD_PROCESS_ACCT
579 If you say Y here, the process accounting information is written
580 in a new file format that also logs the process IDs of each
581 process and its parent. Note that this file format is incompatible
582 with previous v0/v1/v2 file formats, so you will need updated tools
583 for processing it. A preliminary version of these tools is available
584 at <http://www.gnu.org/software/acct/>.
587 bool "Export task/process statistics through netlink"
592 Export selected statistics for tasks/processes through the
593 generic netlink interface. Unlike BSD process accounting, the
594 statistics are available during the lifetime of tasks/processes as
595 responses to commands. Like BSD accounting, they are sent to user
600 config TASK_DELAY_ACCT
601 bool "Enable per-task delay accounting"
605 Collect information on time spent by a task waiting for system
606 resources like cpu, synchronous block I/O completion and swapping
607 in pages. Such statistics can help in setting a task's priorities
608 relative to other tasks for cpu, io, rss limits etc.
613 bool "Enable extended accounting over taskstats"
616 Collect extended task accounting data and send the data
617 to userland for processing over the taskstats interface.
621 config TASK_IO_ACCOUNTING
622 bool "Enable per-task storage I/O accounting"
623 depends on TASK_XACCT
625 Collect information on the number of bytes of storage I/O which this
631 bool "Pressure stall information tracking"
633 Collect metrics that indicate how overcommitted the CPU, memory,
634 and IO capacity are in the system.
636 If you say Y here, the kernel will create /proc/pressure/ with the
637 pressure statistics files cpu, memory, and io. These will indicate
638 the share of walltime in which some or all tasks in the system are
639 delayed due to contention of the respective resource.
641 In kernels with cgroup support, cgroups (cgroup2 only) will
642 have cpu.pressure, memory.pressure, and io.pressure files,
643 which aggregate pressure stalls for the grouped tasks only.
645 For more details see Documentation/accounting/psi.rst.
649 config PSI_DEFAULT_DISABLED
650 bool "Require boot parameter to enable pressure stall information tracking"
654 If set, pressure stall information tracking will be disabled
655 per default but can be enabled through passing psi=1 on the
656 kernel commandline during boot.
658 This feature adds some code to the task wakeup and sleep
659 paths of the scheduler. The overhead is too low to affect
660 common scheduling-intense workloads in practice (such as
661 webservers, memcache), but it does show up in artificial
662 scheduler stress tests, such as hackbench.
664 If you are paranoid and not sure what the kernel will be
669 endmenu # "CPU/Task time and stats accounting"
673 depends on SMP || COMPILE_TEST
676 Make sure that CPUs running critical tasks are not disturbed by
677 any source of "noise" such as unbound workqueues, timers, kthreads...
678 Unbound jobs get offloaded to housekeeping CPUs. This is driven by
679 the "isolcpus=" boot parameter.
683 source "kernel/rcu/Kconfig"
690 tristate "Kernel .config support"
692 This option enables the complete Linux kernel ".config" file
693 contents to be saved in the kernel. It provides documentation
694 of which kernel options are used in a running kernel or in an
695 on-disk kernel. This information can be extracted from the kernel
696 image file with the script scripts/extract-ikconfig and used as
697 input to rebuild the current kernel or to build another kernel.
698 It can also be extracted from a running kernel by reading
699 /proc/config.gz if enabled (below).
702 bool "Enable access to .config through /proc/config.gz"
703 depends on IKCONFIG && PROC_FS
705 This option enables access to the kernel configuration file
706 through /proc/config.gz.
709 tristate "Enable kernel headers through /sys/kernel/kheaders.tar.xz"
712 This option enables access to the in-kernel headers that are generated during
713 the build process. These can be used to build eBPF tracing programs,
714 or similar programs. If you build the headers as a module, a module called
715 kheaders.ko is built which can be loaded on-demand to get access to headers.
718 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
723 Select the minimal kernel log buffer size as a power of 2.
724 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config
725 parameter, see below. Any higher size also might be forced
726 by "log_buf_len" boot parameter.
736 config LOG_CPU_MAX_BUF_SHIFT
737 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)"
740 default 12 if !BASE_SMALL
741 default 0 if BASE_SMALL
744 This option allows to increase the default ring buffer size
745 according to the number of CPUs. The value defines the contribution
746 of each CPU as a power of 2. The used space is typically only few
747 lines however it might be much more when problems are reported,
750 The increased size means that a new buffer has to be allocated and
751 the original static one is unused. It makes sense only on systems
752 with more CPUs. Therefore this value is used only when the sum of
753 contributions is greater than the half of the default kernel ring
754 buffer as defined by LOG_BUF_SHIFT. The default values are set
755 so that more than 16 CPUs are needed to trigger the allocation.
757 Also this option is ignored when "log_buf_len" kernel parameter is
758 used as it forces an exact (power of two) size of the ring buffer.
760 The number of possible CPUs is used for this computation ignoring
761 hotplugging making the computation optimal for the worst case
762 scenario while allowing a simple algorithm to be used from bootup.
764 Examples shift values and their meaning:
765 17 => 128 KB for each CPU
766 16 => 64 KB for each CPU
767 15 => 32 KB for each CPU
768 14 => 16 KB for each CPU
769 13 => 8 KB for each CPU
770 12 => 4 KB for each CPU
772 config PRINTK_SAFE_LOG_BUF_SHIFT
773 int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)"
778 Select the size of an alternate printk per-CPU buffer where messages
779 printed from unsafe contexts are temporary stored. One example would
780 be NMI messages, another one - printk recursion. The messages are
781 copied to the main log buffer in a safe context to avoid a deadlock.
782 The value defines the size as a power of 2.
784 Those messages are rare and limited. The largest one is when
785 a backtrace is printed. It usually fits into 4KB. Select
786 8KB if you want to be on the safe side.
789 17 => 128 KB for each CPU
790 16 => 64 KB for each CPU
791 15 => 32 KB for each CPU
792 14 => 16 KB for each CPU
793 13 => 8 KB for each CPU
794 12 => 4 KB for each CPU
797 bool "Printk indexing debugfs interface"
798 depends on PRINTK && DEBUG_FS
800 Add support for indexing of all printk formats known at compile time
801 at <debugfs>/printk/index/<module>.
803 This can be used as part of maintaining daemons which monitor
804 /dev/kmsg, as it permits auditing the printk formats present in a
805 kernel, allowing detection of cases where monitored printks are
806 changed or no longer present.
808 There is no additional runtime cost to printk with this enabled.
811 # Architectures with an unreliable sched_clock() should select this:
813 config HAVE_UNSTABLE_SCHED_CLOCK
816 config GENERIC_SCHED_CLOCK
819 menu "Scheduler features"
822 bool "Enable utilization clamping for RT/FAIR tasks"
823 depends on CPU_FREQ_GOV_SCHEDUTIL
825 This feature enables the scheduler to track the clamped utilization
826 of each CPU based on RUNNABLE tasks scheduled on that CPU.
828 With this option, the user can specify the min and max CPU
829 utilization allowed for RUNNABLE tasks. The max utilization defines
830 the maximum frequency a task should use while the min utilization
831 defines the minimum frequency it should use.
833 Both min and max utilization clamp values are hints to the scheduler,
834 aiming at improving its frequency selection policy, but they do not
835 enforce or grant any specific bandwidth for tasks.
839 config UCLAMP_BUCKETS_COUNT
840 int "Number of supported utilization clamp buckets"
843 depends on UCLAMP_TASK
845 Defines the number of clamp buckets to use. The range of each bucket
846 will be SCHED_CAPACITY_SCALE/UCLAMP_BUCKETS_COUNT. The higher the
847 number of clamp buckets the finer their granularity and the higher
848 the precision of clamping aggregation and tracking at run-time.
850 For example, with the minimum configuration value we will have 5
851 clamp buckets tracking 20% utilization each. A 25% boosted tasks will
852 be refcounted in the [20..39]% bucket and will set the bucket clamp
853 effective value to 25%.
854 If a second 30% boosted task should be co-scheduled on the same CPU,
855 that task will be refcounted in the same bucket of the first task and
856 it will boost the bucket clamp effective value to 30%.
857 The clamp effective value of a bucket is reset to its nominal value
858 (20% in the example above) when there are no more tasks refcounted in
861 An additional boost/capping margin can be added to some tasks. In the
862 example above the 25% task will be boosted to 30% until it exits the
863 CPU. If that should be considered not acceptable on certain systems,
864 it's always possible to reduce the margin by increasing the number of
865 clamp buckets to trade off used memory for run-time tracking
868 If in doubt, use the default value.
873 # For architectures that want to enable the support for NUMA-affine scheduler
876 config ARCH_SUPPORTS_NUMA_BALANCING
880 # For architectures that prefer to flush all TLBs after a number of pages
881 # are unmapped instead of sending one IPI per page to flush. The architecture
882 # must provide guarantees on what happens if a clean TLB cache entry is
883 # written after the unmap. Details are in mm/rmap.c near the check for
884 # should_defer_flush. The architecture should also consider if the full flush
885 # and the refill costs are offset by the savings of sending fewer IPIs.
886 config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
890 def_bool !$(cc-option,$(m64-flag) -D__SIZEOF_INT128__=0) && 64BIT
892 config CC_IMPLICIT_FALLTHROUGH
894 default "-Wimplicit-fallthrough=5" if CC_IS_GCC && $(cc-option,-Wimplicit-fallthrough=5)
895 default "-Wimplicit-fallthrough" if CC_IS_CLANG && $(cc-option,-Wunreachable-code-fallthrough)
897 # Currently, disable gcc-11,12 array-bounds globally.
898 # We may want to target only particular configurations some day.
899 config GCC11_NO_ARRAY_BOUNDS
902 config GCC12_NO_ARRAY_BOUNDS
905 config CC_NO_ARRAY_BOUNDS
907 default y if CC_IS_GCC && GCC_VERSION >= 110000 && GCC_VERSION < 120000 && GCC11_NO_ARRAY_BOUNDS
908 default y if CC_IS_GCC && GCC_VERSION >= 120000 && GCC_VERSION < 130000 && GCC12_NO_ARRAY_BOUNDS
911 # For architectures that know their GCC __int128 support is sound
913 config ARCH_SUPPORTS_INT128
916 # For architectures that (ab)use NUMA to represent different memory regions
917 # all cpu-local but of different latencies, such as SuperH.
919 config ARCH_WANT_NUMA_VARIABLE_LOCALITY
922 config NUMA_BALANCING
923 bool "Memory placement aware NUMA scheduler"
924 depends on ARCH_SUPPORTS_NUMA_BALANCING
925 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
926 depends on SMP && NUMA && MIGRATION && !PREEMPT_RT
928 This option adds support for automatic NUMA aware memory/task placement.
929 The mechanism is quite primitive and is based on migrating memory when
930 it has references to the node the task is running on.
932 This system will be inactive on UMA systems.
934 config NUMA_BALANCING_DEFAULT_ENABLED
935 bool "Automatically enable NUMA aware memory/task placement"
937 depends on NUMA_BALANCING
939 If set, automatic NUMA balancing will be enabled if running on a NUMA
943 bool "Control Group support"
946 This option adds support for grouping sets of processes together, for
947 use with process control subsystems such as Cpusets, CFS, memory
948 controls or device isolation.
950 - Documentation/scheduler/sched-design-CFS.rst (CFS)
951 - Documentation/admin-guide/cgroup-v1/ (features for grouping, isolation
952 and resource control)
961 config CGROUP_FAVOR_DYNMODS
962 bool "Favor dynamic modification latency reduction by default"
964 This option enables the "favordynmods" mount option by default
965 which reduces the latencies of dynamic cgroup modifications such
966 as task migrations and controller on/offs at the cost of making
967 hot path operations such as forks and exits more expensive.
972 bool "Memory controller"
976 Provides control over the memory footprint of tasks in a cgroup.
980 depends on MEMCG && !SLOB
988 Generic block IO controller cgroup interface. This is the common
989 cgroup interface which should be used by various IO controlling
992 Currently, CFQ IO scheduler uses it to recognize task groups and
993 control disk bandwidth allocation (proportional time slice allocation)
994 to such task groups. It is also used by bio throttling logic in
995 block layer to implement upper limit in IO rates on a device.
997 This option only enables generic Block IO controller infrastructure.
998 One needs to also enable actual IO controlling logic/policy. For
999 enabling proportional weight division of disk bandwidth in CFQ, set
1000 CONFIG_BFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
1001 CONFIG_BLK_DEV_THROTTLING=y.
1003 See Documentation/admin-guide/cgroup-v1/blkio-controller.rst for more information.
1005 config CGROUP_WRITEBACK
1007 depends on MEMCG && BLK_CGROUP
1010 menuconfig CGROUP_SCHED
1011 bool "CPU controller"
1014 This feature lets CPU scheduler recognize task groups and control CPU
1015 bandwidth allocation to such task groups. It uses cgroups to group
1019 config FAIR_GROUP_SCHED
1020 bool "Group scheduling for SCHED_OTHER"
1021 depends on CGROUP_SCHED
1022 default CGROUP_SCHED
1024 config CFS_BANDWIDTH
1025 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
1026 depends on FAIR_GROUP_SCHED
1029 This option allows users to define CPU bandwidth rates (limits) for
1030 tasks running within the fair group scheduler. Groups with no limit
1031 set are considered to be unconstrained and will run with no
1033 See Documentation/scheduler/sched-bwc.rst for more information.
1035 config RT_GROUP_SCHED
1036 bool "Group scheduling for SCHED_RR/FIFO"
1037 depends on CGROUP_SCHED
1040 This feature lets you explicitly allocate real CPU bandwidth
1041 to task groups. If enabled, it will also make it impossible to
1042 schedule realtime tasks for non-root users until you allocate
1043 realtime bandwidth for them.
1044 See Documentation/scheduler/sched-rt-group.rst for more information.
1050 depends on SMP && RSEQ
1052 config UCLAMP_TASK_GROUP
1053 bool "Utilization clamping per group of tasks"
1054 depends on CGROUP_SCHED
1055 depends on UCLAMP_TASK
1058 This feature enables the scheduler to track the clamped utilization
1059 of each CPU based on RUNNABLE tasks currently scheduled on that CPU.
1061 When this option is enabled, the user can specify a min and max
1062 CPU bandwidth which is allowed for each single task in a group.
1063 The max bandwidth allows to clamp the maximum frequency a task
1064 can use, while the min bandwidth allows to define a minimum
1065 frequency a task will always use.
1067 When task group based utilization clamping is enabled, an eventually
1068 specified task-specific clamp value is constrained by the cgroup
1069 specified clamp value. Both minimum and maximum task clamping cannot
1070 be bigger than the corresponding clamping defined at task group level.
1075 bool "PIDs controller"
1077 Provides enforcement of process number limits in the scope of a
1078 cgroup. Any attempt to fork more processes than is allowed in the
1079 cgroup will fail. PIDs are fundamentally a global resource because it
1080 is fairly trivial to reach PID exhaustion before you reach even a
1081 conservative kmemcg limit. As a result, it is possible to grind a
1082 system to halt without being limited by other cgroup policies. The
1083 PIDs controller is designed to stop this from happening.
1085 It should be noted that organisational operations (such as attaching
1086 to a cgroup hierarchy) will *not* be blocked by the PIDs controller,
1087 since the PIDs limit only affects a process's ability to fork, not to
1091 bool "RDMA controller"
1093 Provides enforcement of RDMA resources defined by IB stack.
1094 It is fairly easy for consumers to exhaust RDMA resources, which
1095 can result into resource unavailability to other consumers.
1096 RDMA controller is designed to stop this from happening.
1097 Attaching processes with active RDMA resources to the cgroup
1098 hierarchy is allowed even if can cross the hierarchy's limit.
1100 config CGROUP_FREEZER
1101 bool "Freezer controller"
1103 Provides a way to freeze and unfreeze all tasks in a
1106 This option affects the ORIGINAL cgroup interface. The cgroup2 memory
1107 controller includes important in-kernel memory consumers per default.
1109 If you're using cgroup2, say N.
1111 config CGROUP_HUGETLB
1112 bool "HugeTLB controller"
1113 depends on HUGETLB_PAGE
1117 Provides a cgroup controller for HugeTLB pages.
1118 When you enable this, you can put a per cgroup limit on HugeTLB usage.
1119 The limit is enforced during page fault. Since HugeTLB doesn't
1120 support page reclaim, enforcing the limit at page fault time implies
1121 that, the application will get SIGBUS signal if it tries to access
1122 HugeTLB pages beyond its limit. This requires the application to know
1123 beforehand how much HugeTLB pages it would require for its use. The
1124 control group is tracked in the third page lru pointer. This means
1125 that we cannot use the controller with huge page less than 3 pages.
1128 bool "Cpuset controller"
1131 This option will let you create and manage CPUSETs which
1132 allow dynamically partitioning a system into sets of CPUs and
1133 Memory Nodes and assigning tasks to run only within those sets.
1134 This is primarily useful on large SMP or NUMA systems.
1138 config PROC_PID_CPUSET
1139 bool "Include legacy /proc/<pid>/cpuset file"
1143 config CGROUP_DEVICE
1144 bool "Device controller"
1146 Provides a cgroup controller implementing whitelists for
1147 devices which a process in the cgroup can mknod or open.
1149 config CGROUP_CPUACCT
1150 bool "Simple CPU accounting controller"
1152 Provides a simple controller for monitoring the
1153 total CPU consumed by the tasks in a cgroup.
1156 bool "Perf controller"
1157 depends on PERF_EVENTS
1159 This option extends the perf per-cpu mode to restrict monitoring
1160 to threads which belong to the cgroup specified and run on the
1161 designated cpu. Or this can be used to have cgroup ID in samples
1162 so that it can monitor performance events among cgroups.
1167 bool "Support for eBPF programs attached to cgroups"
1168 depends on BPF_SYSCALL
1169 select SOCK_CGROUP_DATA
1171 Allow attaching eBPF programs to a cgroup using the bpf(2)
1172 syscall command BPF_PROG_ATTACH.
1174 In which context these programs are accessed depends on the type
1175 of attachment. For instance, programs that are attached using
1176 BPF_CGROUP_INET_INGRESS will be executed on the ingress path of
1180 bool "Misc resource controller"
1183 Provides a controller for miscellaneous resources on a host.
1185 Miscellaneous scalar resources are the resources on the host system
1186 which cannot be abstracted like the other cgroups. This controller
1187 tracks and limits the miscellaneous resources used by a process
1188 attached to a cgroup hierarchy.
1190 For more information, please check misc cgroup section in
1191 /Documentation/admin-guide/cgroup-v2.rst.
1194 bool "Debug controller"
1196 depends on DEBUG_KERNEL
1198 This option enables a simple controller that exports
1199 debugging information about the cgroups framework. This
1200 controller is for control cgroup debugging only. Its
1201 interfaces are not stable.
1205 config SOCK_CGROUP_DATA
1211 menuconfig NAMESPACES
1212 bool "Namespaces support" if EXPERT
1213 depends on MULTIUSER
1216 Provides the way to make tasks work with different objects using
1217 the same id. For example same IPC id may refer to different objects
1218 or same user id or pid may refer to different tasks when used in
1219 different namespaces.
1224 bool "UTS namespace"
1227 In this namespace tasks see different info provided with the
1231 bool "TIME namespace"
1232 depends on GENERIC_VDSO_TIME_NS
1235 In this namespace boottime and monotonic clocks can be set.
1236 The time will keep going with the same pace.
1239 bool "IPC namespace"
1240 depends on (SYSVIPC || POSIX_MQUEUE)
1243 In this namespace tasks work with IPC ids which correspond to
1244 different IPC objects in different namespaces.
1247 bool "User namespace"
1250 This allows containers, i.e. vservers, to use user namespaces
1251 to provide different user info for different servers.
1253 When user namespaces are enabled in the kernel it is
1254 recommended that the MEMCG option also be enabled and that
1255 user-space use the memory control groups to limit the amount
1256 of memory a memory unprivileged users can use.
1261 bool "PID Namespaces"
1264 Support process id namespaces. This allows having multiple
1265 processes with the same pid as long as they are in different
1266 pid namespaces. This is a building block of containers.
1269 bool "Network namespace"
1273 Allow user space to create what appear to be multiple instances
1274 of the network stack.
1278 config CHECKPOINT_RESTORE
1279 bool "Checkpoint/restore support"
1281 select PROC_CHILDREN
1285 Enables additional kernel features in a sake of checkpoint/restore.
1286 In particular it adds auxiliary prctl codes to setup process text,
1287 data and heap segment sizes, and a few additional /proc filesystem
1290 If unsure, say N here.
1292 config SCHED_AUTOGROUP
1293 bool "Automatic process group scheduling"
1296 select FAIR_GROUP_SCHED
1298 This option optimizes the scheduler for common desktop workloads by
1299 automatically creating and populating task groups. This separation
1300 of workloads isolates aggressive CPU burners (like build jobs) from
1301 desktop applications. Task group autogeneration is currently based
1304 config SYSFS_DEPRECATED
1305 bool "Enable deprecated sysfs features to support old userspace tools"
1309 This option adds code that switches the layout of the "block" class
1310 devices, to not show up in /sys/class/block/, but only in
1313 This switch is only active when the sysfs.deprecated=1 boot option is
1314 passed or the SYSFS_DEPRECATED_V2 option is set.
1316 This option allows new kernels to run on old distributions and tools,
1317 which might get confused by /sys/class/block/. Since 2007/2008 all
1318 major distributions and tools handle this just fine.
1320 Recent distributions and userspace tools after 2009/2010 depend on
1321 the existence of /sys/class/block/, and will not work with this
1324 Only if you are using a new kernel on an old distribution, you might
1327 config SYSFS_DEPRECATED_V2
1328 bool "Enable deprecated sysfs features by default"
1331 depends on SYSFS_DEPRECATED
1333 Enable deprecated sysfs by default.
1335 See the CONFIG_SYSFS_DEPRECATED option for more details about this
1338 Only if you are using a new kernel on an old distribution, you might
1339 need to say Y here. Even then, odds are you would not need it
1340 enabled, you can always pass the boot option if absolutely necessary.
1343 bool "Kernel->user space relay support (formerly relayfs)"
1346 This option enables support for relay interface support in
1347 certain file systems (such as debugfs).
1348 It is designed to provide an efficient mechanism for tools and
1349 facilities to relay large amounts of data from kernel space to
1354 config BLK_DEV_INITRD
1355 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
1357 The initial RAM filesystem is a ramfs which is loaded by the
1358 boot loader (loadlin or lilo) and that is mounted as root
1359 before the normal boot procedure. It is typically used to
1360 load modules needed to mount the "real" root file system,
1361 etc. See <file:Documentation/admin-guide/initrd.rst> for details.
1363 If RAM disk support (BLK_DEV_RAM) is also included, this
1364 also enables initial RAM disk (initrd) support and adds
1365 15 Kbytes (more on some other architectures) to the kernel size.
1371 source "usr/Kconfig"
1376 bool "Boot config support"
1377 select BLK_DEV_INITRD if !BOOT_CONFIG_EMBED
1379 Extra boot config allows system admin to pass a config file as
1380 complemental extension of kernel cmdline when booting.
1381 The boot config file must be attached at the end of initramfs
1382 with checksum, size and magic word.
1383 See <file:Documentation/admin-guide/bootconfig.rst> for details.
1387 config BOOT_CONFIG_EMBED
1388 bool "Embed bootconfig file in the kernel"
1389 depends on BOOT_CONFIG
1391 Embed a bootconfig file given by BOOT_CONFIG_EMBED_FILE in the
1392 kernel. Usually, the bootconfig file is loaded with the initrd
1393 image. But if the system doesn't support initrd, this option will
1394 help you by embedding a bootconfig file while building the kernel.
1398 config BOOT_CONFIG_EMBED_FILE
1399 string "Embedded bootconfig file path"
1400 depends on BOOT_CONFIG_EMBED
1402 Specify a bootconfig file which will be embedded to the kernel.
1403 This bootconfig will be used if there is no initrd or no other
1404 bootconfig in the initrd.
1406 config INITRAMFS_PRESERVE_MTIME
1407 bool "Preserve cpio archive mtimes in initramfs"
1410 Each entry in an initramfs cpio archive carries an mtime value. When
1411 enabled, extracted cpio items take this mtime, with directory mtime
1412 setting deferred until after creation of any child entries.
1417 prompt "Compiler optimization level"
1418 default CC_OPTIMIZE_FOR_PERFORMANCE
1420 config CC_OPTIMIZE_FOR_PERFORMANCE
1421 bool "Optimize for performance (-O2)"
1423 This is the default optimization level for the kernel, building
1424 with the "-O2" compiler flag for best performance and most
1425 helpful compile-time warnings.
1427 config CC_OPTIMIZE_FOR_SIZE
1428 bool "Optimize for size (-Os)"
1430 Choosing this option will pass "-Os" to your compiler resulting
1431 in a smaller kernel.
1435 config HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1438 This requires that the arch annotates or otherwise protects
1439 its external entry points from being discarded. Linker scripts
1440 must also merge .text.*, .data.*, and .bss.* correctly into
1441 output sections. Care must be taken not to pull in unrelated
1442 sections (e.g., '.text.init'). Typically '.' in section names
1443 is used to distinguish them from label names / C identifiers.
1445 config LD_DEAD_CODE_DATA_ELIMINATION
1446 bool "Dead code and data elimination (EXPERIMENTAL)"
1447 depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1449 depends on $(cc-option,-ffunction-sections -fdata-sections)
1450 depends on $(ld-option,--gc-sections)
1452 Enable this if you want to do dead code and data elimination with
1453 the linker by compiling with -ffunction-sections -fdata-sections,
1454 and linking with --gc-sections.
1456 This can reduce on disk and in-memory size of the kernel
1457 code and static data, particularly for small configs and
1458 on small systems. This has the possibility of introducing
1459 silently broken kernel if the required annotations are not
1460 present. This option is not well tested yet, so use at your
1463 config LD_ORPHAN_WARN
1465 depends on ARCH_WANT_LD_ORPHAN_WARN
1466 depends on $(ld-option,--orphan-handling=warn)
1467 depends on $(ld-option,--orphan-handling=error)
1469 config LD_ORPHAN_WARN_LEVEL
1471 depends on LD_ORPHAN_WARN
1472 default "error" if WERROR
1481 config SYSCTL_EXCEPTION_TRACE
1484 Enable support for /proc/sys/debug/exception-trace.
1486 config SYSCTL_ARCH_UNALIGN_NO_WARN
1489 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1490 Allows arch to define/use @no_unaligned_warning to possibly warn
1491 about unaligned access emulation going on under the hood.
1493 config SYSCTL_ARCH_UNALIGN_ALLOW
1496 Enable support for /proc/sys/kernel/unaligned-trap
1497 Allows arches to define/use @unaligned_enabled to runtime toggle
1498 the unaligned access emulation.
1499 see arch/parisc/kernel/unaligned.c for reference
1501 config HAVE_PCSPKR_PLATFORM
1504 # interpreter that classic socket filters depend on
1507 select CRYPTO_LIB_SHA1
1510 bool "Configure standard kernel features (expert users)"
1511 # Unhide debug options, to make the on-by-default options visible
1514 This option allows certain base kernel options and settings
1515 to be disabled or tweaked. This is for specialized
1516 environments which can tolerate a "non-standard" kernel.
1517 Only use this if you really know what you are doing.
1520 bool "Enable 16-bit UID system calls" if EXPERT
1521 depends on HAVE_UID16 && MULTIUSER
1524 This enables the legacy 16-bit UID syscall wrappers.
1527 bool "Multiple users, groups and capabilities support" if EXPERT
1530 This option enables support for non-root users, groups and
1533 If you say N here, all processes will run with UID 0, GID 0, and all
1534 possible capabilities. Saying N here also compiles out support for
1535 system calls related to UIDs, GIDs, and capabilities, such as setuid,
1538 If unsure, say Y here.
1540 config SGETMASK_SYSCALL
1541 bool "sgetmask/ssetmask syscalls support" if EXPERT
1542 def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH
1544 sys_sgetmask and sys_ssetmask are obsolete system calls
1545 no longer supported in libc but still enabled by default in some
1548 If unsure, leave the default option here.
1550 config SYSFS_SYSCALL
1551 bool "Sysfs syscall support" if EXPERT
1554 sys_sysfs is an obsolete system call no longer supported in libc.
1555 Note that disabling this option is more secure but might break
1556 compatibility with some systems.
1558 If unsure say Y here.
1561 bool "open by fhandle syscalls" if EXPERT
1565 If you say Y here, a user level program will be able to map
1566 file names to handle and then later use the handle for
1567 different file system operations. This is useful in implementing
1568 userspace file servers, which now track files using handles instead
1569 of names. The handle would remain the same even if file names
1570 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
1574 bool "Posix Clocks & timers" if EXPERT
1577 This includes native support for POSIX timers to the kernel.
1578 Some embedded systems have no use for them and therefore they
1579 can be configured out to reduce the size of the kernel image.
1581 When this option is disabled, the following syscalls won't be
1582 available: timer_create, timer_gettime: timer_getoverrun,
1583 timer_settime, timer_delete, clock_adjtime, getitimer,
1584 setitimer, alarm. Furthermore, the clock_settime, clock_gettime,
1585 clock_getres and clock_nanosleep syscalls will be limited to
1586 CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only.
1592 bool "Enable support for printk" if EXPERT
1595 This option enables normal printk support. Removing it
1596 eliminates most of the message strings from the kernel image
1597 and makes the kernel more or less silent. As this makes it
1598 very difficult to diagnose system problems, saying N here is
1599 strongly discouraged.
1602 bool "BUG() support" if EXPERT
1605 Disabling this option eliminates support for BUG and WARN, reducing
1606 the size of your kernel image and potentially quietly ignoring
1607 numerous fatal conditions. You should only consider disabling this
1608 option for embedded systems with no facilities for reporting errors.
1614 bool "Enable ELF core dumps" if EXPERT
1616 Enable support for generating core dumps. Disabling saves about 4k.
1619 config PCSPKR_PLATFORM
1620 bool "Enable PC-Speaker support" if EXPERT
1621 depends on HAVE_PCSPKR_PLATFORM
1625 This option allows to disable the internal PC-Speaker
1626 support, saving some memory.
1630 bool "Enable full-sized data structures for core" if EXPERT
1632 Disabling this option reduces the size of miscellaneous core
1633 kernel data structures. This saves memory on small machines,
1634 but may reduce performance.
1637 bool "Enable futex support" if EXPERT
1638 depends on !(SPARC32 && SMP)
1642 Disabling this option will cause the kernel to be built without
1643 support for "fast userspace mutexes". The resulting kernel may not
1644 run glibc-based applications correctly.
1648 depends on FUTEX && RT_MUTEXES
1652 bool "Enable eventpoll support" if EXPERT
1655 Disabling this option will cause the kernel to be built without
1656 support for epoll family of system calls.
1659 bool "Enable signalfd() system call" if EXPERT
1662 Enable the signalfd() system call that allows to receive signals
1663 on a file descriptor.
1668 bool "Enable timerfd() system call" if EXPERT
1671 Enable the timerfd() system call that allows to receive timer
1672 events on a file descriptor.
1677 bool "Enable eventfd() system call" if EXPERT
1680 Enable the eventfd() system call that allows to receive both
1681 kernel notification (ie. KAIO) or userspace notifications.
1686 bool "Use full shmem filesystem" if EXPERT
1690 The shmem is an internal filesystem used to manage shared memory.
1691 It is backed by swap and manages resource limits. It is also exported
1692 to userspace as tmpfs if TMPFS is enabled. Disabling this
1693 option replaces shmem and tmpfs with the much simpler ramfs code,
1694 which may be appropriate on small systems without swap.
1697 bool "Enable AIO support" if EXPERT
1700 This option enables POSIX asynchronous I/O which may by used
1701 by some high performance threaded applications. Disabling
1702 this option saves about 7k.
1705 bool "Enable IO uring support" if EXPERT
1709 This option enables support for the io_uring interface, enabling
1710 applications to submit and complete IO through submission and
1711 completion rings that are shared between the kernel and application.
1713 config ADVISE_SYSCALLS
1714 bool "Enable madvise/fadvise syscalls" if EXPERT
1717 This option enables the madvise and fadvise syscalls, used by
1718 applications to advise the kernel about their future memory or file
1719 usage, improving performance. If building an embedded system where no
1720 applications use these syscalls, you can disable this option to save
1724 bool "Enable membarrier() system call" if EXPERT
1727 Enable the membarrier() system call that allows issuing memory
1728 barriers across all running threads, which can be used to distribute
1729 the cost of user-space memory barriers asymmetrically by transforming
1730 pairs of memory barriers into pairs consisting of membarrier() and a
1736 bool "Load all symbols for debugging/ksymoops" if EXPERT
1739 Say Y here to let the kernel print out symbolic crash information and
1740 symbolic stack backtraces. This increases the size of the kernel
1741 somewhat, as all symbols have to be loaded into the kernel image.
1743 config KALLSYMS_SELFTEST
1744 bool "Test the basic functions and performance of kallsyms"
1748 Test the basic functions and performance of some interfaces, such as
1749 kallsyms_lookup_name. It also calculates the compression rate of the
1750 kallsyms compression algorithm for the current symbol set.
1752 Start self-test automatically after system startup. Suggest executing
1753 "dmesg | grep kallsyms_selftest" to collect test results. "finish" is
1754 displayed in the last line, indicating that the test is complete.
1757 bool "Include all symbols in kallsyms"
1758 depends on DEBUG_KERNEL && KALLSYMS
1760 Normally kallsyms only contains the symbols of functions for nicer
1761 OOPS messages and backtraces (i.e., symbols from the text and inittext
1762 sections). This is sufficient for most cases. And only if you want to
1763 enable kernel live patching, or other less common use cases (e.g.,
1764 when a debugger is used) all symbols are required (i.e., names of
1765 variables from the data sections, etc).
1767 This option makes sure that all symbols are loaded into the kernel
1768 image (i.e., symbols from all sections) in cost of increased kernel
1769 size (depending on the kernel configuration, it may be 300KiB or
1770 something like this).
1772 Say N unless you really need all symbols, or kernel live patching.
1774 config KALLSYMS_ABSOLUTE_PERCPU
1777 default X86_64 && SMP
1779 config KALLSYMS_BASE_RELATIVE
1784 Instead of emitting them as absolute values in the native word size,
1785 emit the symbol references in the kallsyms table as 32-bit entries,
1786 each containing a relative value in the range [base, base + U32_MAX]
1787 or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either
1788 an absolute value in the range [0, S32_MAX] or a relative value in the
1789 range [base, base + S32_MAX], where base is the lowest relative symbol
1790 address encountered in the image.
1792 On 64-bit builds, this reduces the size of the address table by 50%,
1793 but more importantly, it results in entries whose values are build
1794 time constants, and no relocation pass is required at runtime to fix
1795 up the entries based on the runtime load address of the kernel.
1797 # end of the "standard kernel features (expert users)" menu
1799 # syscall, maps, verifier
1801 config ARCH_HAS_MEMBARRIER_CALLBACKS
1804 config ARCH_HAS_MEMBARRIER_SYNC_CORE
1808 bool "Enable kcmp() system call" if EXPERT
1810 Enable the kernel resource comparison system call. It provides
1811 user-space with the ability to compare two processes to see if they
1812 share a common resource, such as a file descriptor or even virtual
1818 bool "Enable rseq() system call" if EXPERT
1820 depends on HAVE_RSEQ
1823 Enable the restartable sequences system call. It provides a
1824 user-space cache for the current CPU number value, which
1825 speeds up getting the current CPU number from user-space,
1826 as well as an ABI to speed up user-space operations on
1833 bool "Enabled debugging of rseq() system call" if EXPERT
1834 depends on RSEQ && DEBUG_KERNEL
1836 Enable extra debugging checks for the rseq system call.
1841 bool "Embedded system"
1844 This option should be enabled if compiling the kernel for
1845 an embedded system so certain expert options are available
1848 config HAVE_PERF_EVENTS
1851 See tools/perf/design.txt for details.
1853 config GUEST_PERF_EVENTS
1855 depends on HAVE_PERF_EVENTS
1857 config PERF_USE_VMALLOC
1860 See tools/perf/design.txt for details
1863 bool "PC/104 support" if EXPERT
1865 Expose PC/104 form factor device drivers and options available for
1866 selection and configuration. Enable this option if your target
1867 machine has a PC/104 bus.
1869 menu "Kernel Performance Events And Counters"
1872 bool "Kernel performance events and counters"
1873 default y if PROFILING
1874 depends on HAVE_PERF_EVENTS
1877 Enable kernel support for various performance events provided
1878 by software and hardware.
1880 Software events are supported either built-in or via the
1881 use of generic tracepoints.
1883 Most modern CPUs support performance events via performance
1884 counter registers. These registers count the number of certain
1885 types of hw events: such as instructions executed, cachemisses
1886 suffered, or branches mis-predicted - without slowing down the
1887 kernel or applications. These registers can also trigger interrupts
1888 when a threshold number of events have passed - and can thus be
1889 used to profile the code that runs on that CPU.
1891 The Linux Performance Event subsystem provides an abstraction of
1892 these software and hardware event capabilities, available via a
1893 system call and used by the "perf" utility in tools/perf/. It
1894 provides per task and per CPU counters, and it provides event
1895 capabilities on top of those.
1899 config DEBUG_PERF_USE_VMALLOC
1901 bool "Debug: use vmalloc to back perf mmap() buffers"
1902 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC
1903 select PERF_USE_VMALLOC
1905 Use vmalloc memory to back perf mmap() buffers.
1907 Mostly useful for debugging the vmalloc code on platforms
1908 that don't require it.
1914 config SYSTEM_DATA_VERIFICATION
1916 select SYSTEM_TRUSTED_KEYRING
1920 select ASYMMETRIC_KEY_TYPE
1921 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
1924 select X509_CERTIFICATE_PARSER
1925 select PKCS7_MESSAGE_PARSER
1927 Provide PKCS#7 message verification using the contents of the system
1928 trusted keyring to provide public keys. This then can be used for
1929 module verification, kexec image verification and firmware blob
1933 bool "Profiling support"
1935 Say Y here to enable the extended profiling support mechanisms used
1940 depends on HAVE_RUST
1941 depends on RUST_IS_AVAILABLE
1942 depends on !MODVERSIONS
1943 depends on !GCC_PLUGINS
1944 depends on !RANDSTRUCT
1945 depends on !DEBUG_INFO_BTF || PAHOLE_HAS_LANG_EXCLUDE
1948 Enables Rust support in the kernel.
1950 This allows other Rust-related options, like drivers written in Rust,
1953 It is also required to be able to load external kernel modules
1956 See Documentation/rust/ for more information.
1960 config RUSTC_VERSION_TEXT
1963 default $(shell,command -v $(RUSTC) >/dev/null 2>&1 && $(RUSTC) --version || echo n)
1965 config BINDGEN_VERSION_TEXT
1968 default $(shell,command -v $(BINDGEN) >/dev/null 2>&1 && $(BINDGEN) --version || echo n)
1971 # Place an empty function call at each tracepoint site. Can be
1972 # dynamically changed for a probe function.
1977 endmenu # General setup
1979 source "arch/Kconfig"
1983 default y if PREEMPT_RT
1987 default 0 if BASE_FULL
1988 default 1 if !BASE_FULL
1990 config MODULE_SIG_FORMAT
1992 select SYSTEM_DATA_VERIFICATION
1994 source "kernel/module/Kconfig"
1996 config INIT_ALL_POSSIBLE
1999 Back when each arch used to define their own cpu_online_mask and
2000 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
2001 with all 1s, and others with all 0s. When they were centralised,
2002 it was better to provide this option than to break all the archs
2003 and have several arch maintainers pursuing me down dark alleys.
2005 source "block/Kconfig"
2007 config PREEMPT_NOTIFIERS
2017 Build a simple ASN.1 grammar compiler that produces a bytecode output
2018 that can be interpreted by the ASN.1 stream decoder and used to
2019 inform it as to what tags are to be expected in a stream and what
2020 functions to call on what tags.
2022 source "kernel/Kconfig.locks"
2024 config ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE
2027 config ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
2030 # It may be useful for an architecture to override the definitions of the
2031 # SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h>
2032 # and the COMPAT_ variants in <linux/compat.h>, in particular to use a
2033 # different calling convention for syscalls. They can also override the
2034 # macros for not-implemented syscalls in kernel/sys_ni.c and
2035 # kernel/time/posix-stubs.c. All these overrides need to be available in
2036 # <asm/syscall_wrapper.h>.
2037 config ARCH_HAS_SYSCALL_WRAPPER
projects / linux.git / blob