1 # SPDX-License-Identifier: GPL-2.0-only
2 config CC_VERSION_TEXT
3 string
4 default "$(CC_VERSION_TEXT)"
5 help
6 This is used in unclear ways:
7
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.
12
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.
18
19 config CC_IS_GCC
20 def_bool $(success,test "$(cc-name)" = GCC)
21
22 config GCC_VERSION
23 int
24 default $(cc-version) if CC_IS_GCC
25 default 0
26
27 config CC_IS_CLANG
28 def_bool $(success,test "$(cc-name)" = Clang)
29
30 config CLANG_VERSION
31 int
32 default $(cc-version) if CC_IS_CLANG
33 default 0
34
35 config AS_IS_GNU
36 def_bool $(success,test "$(as-name)" = GNU)
37
38 config AS_IS_LLVM
39 def_bool $(success,test "$(as-name)" = LLVM)
40
41 config AS_VERSION
42 int
43 # Use clang version if this is the integrated assembler
44 default CLANG_VERSION if AS_IS_LLVM
45 default $(as-version)
46
47 config LD_IS_BFD
48 def_bool $(success,test "$(ld-name)" = BFD)
49
50 config LD_VERSION
51 int
52 default $(ld-version) if LD_IS_BFD
53 default 0
54
55 config LD_IS_LLD
56 def_bool $(success,test "$(ld-name)" = LLD)
57
58 config LLD_VERSION
59 int
60 default $(ld-version) if LD_IS_LLD
61 default 0
62
63 config RUST_IS_AVAILABLE
64 def_bool $(success,$(srctree)/scripts/rust_is_available.sh)
65 help
66 This shows whether a suitable Rust toolchain is available (found).
67
68 Please see Documentation/rust/quick-start.rst for instructions on how
69 to satisfy the build requirements of Rust support.
70
71 In particular, the Makefile target 'rustavailable' is useful to check
72 why the Rust toolchain is not being detected.
73
74 config CC_CAN_LINK
75 bool
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))
78
79 config CC_CAN_LINK_STATIC
80 bool
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)
83
84 # Fixed in GCC 14, 13.3, 12.4 and 11.5
85 # https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113921
86 config GCC_ASM_GOTO_OUTPUT_BROKEN
87 bool
88 depends on CC_IS_GCC
89 default y if GCC_VERSION < 110500
90 default y if GCC_VERSION >= 120000 && GCC_VERSION < 120400
91 default y if GCC_VERSION >= 130000 && GCC_VERSION < 130300
92
93 config CC_HAS_ASM_GOTO_OUTPUT
94 def_bool y
95 depends on !GCC_ASM_GOTO_OUTPUT_BROKEN
96 depends on $(success,echo 'int foo(int x) { asm goto ("": "=r"(x) ::: bar); return x; bar: return 0; }' | $(CC) -x c - -c -o /dev/null)
97
98 config CC_HAS_ASM_GOTO_TIED_OUTPUT
99 depends on CC_HAS_ASM_GOTO_OUTPUT
100 # Detect buggy gcc and clang, fixed in gcc-11 clang-14.
101 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)
102
103 config TOOLS_SUPPORT_RELR
104 def_bool $(success,env "CC=$(CC)" "LD=$(LD)" "NM=$(NM)" "OBJCOPY=$(OBJCOPY)" $(srctree)/scripts/tools-support-relr.sh)
105
106 config CC_HAS_ASM_INLINE
107 def_bool $(success,echo 'void foo(void) { asm inline (""); }' | $(CC) -x c - -c -o /dev/null)
108
109 config CC_HAS_NO_PROFILE_FN_ATTR
110 def_bool $(success,echo '__attribute__((no_profile_instrument_function)) int x();' | $(CC) -x c - -c -o /dev/null -Werror)
111
112 config PAHOLE_VERSION
113 int
114 default $(shell,$(srctree)/scripts/pahole-version.sh $(PAHOLE))
115
116 config CONSTRUCTORS
117 bool
118
119 config IRQ_WORK
120 def_bool y if SMP
121
122 config BUILDTIME_TABLE_SORT
123 bool
124
125 config THREAD_INFO_IN_TASK
126 bool
127 help
128 Select this to move thread_info off the stack into task_struct. To
129 make this work, an arch will need to remove all thread_info fields
130 except flags and fix any runtime bugs.
131
132 One subtle change that will be needed is to use try_get_task_stack()
133 and put_task_stack() in save_thread_stack_tsk() and get_wchan().
134
135 menu "General setup"
136
137 config BROKEN
138 bool
139
140 config BROKEN_ON_SMP
141 bool
142 depends on BROKEN || !SMP
143 default y
144
145 config INIT_ENV_ARG_LIMIT
146 int
147 default 32 if !UML
148 default 128 if UML
149 help
150 Maximum of each of the number of arguments and environment
151 variables passed to init from the kernel command line.
152
153 config COMPILE_TEST
154 bool "Compile also drivers which will not load"
155 depends on HAS_IOMEM
156 help
157 Some drivers can be compiled on a different platform than they are
158 intended to be run on. Despite they cannot be loaded there (or even
159 when they load they cannot be used due to missing HW support),
160 developers still, opposing to distributors, might want to build such
161 drivers to compile-test them.
162
163 If you are a developer and want to build everything available, say Y
164 here. If you are a user/distributor, say N here to exclude useless
165 drivers to be distributed.
166
167 config WERROR
168 bool "Compile the kernel with warnings as errors"
169 default COMPILE_TEST
170 help
171 A kernel build should not cause any compiler warnings, and this
172 enables the '-Werror' (for C) and '-Dwarnings' (for Rust) flags
173 to enforce that rule by default. Certain warnings from other tools
174 such as the linker may be upgraded to errors with this option as
175 well.
176
177 However, if you have a new (or very old) compiler or linker with odd
178 and unusual warnings, or you have some architecture with problems,
179 you may need to disable this config option in order to
180 successfully build the kernel.
181
182 If in doubt, say Y.
183
184 config UAPI_HEADER_TEST
185 bool "Compile test UAPI headers"
186 depends on HEADERS_INSTALL && CC_CAN_LINK
187 help
188 Compile test headers exported to user-space to ensure they are
189 self-contained, i.e. compilable as standalone units.
190
191 If you are a developer or tester and want to ensure the exported
192 headers are self-contained, say Y here. Otherwise, choose N.
193
194 config LOCALVERSION
195 string "Local version - append to kernel release"
196 help
197 Append an extra string to the end of your kernel version.
198 This will show up when you type uname, for example.
199 The string you set here will be appended after the contents of
200 any files with a filename matching localversion* in your
201 object and source tree, in that order. Your total string can
202 be a maximum of 64 characters.
203
204 config LOCALVERSION_AUTO
205 bool "Automatically append version information to the version string"
206 default y
207 depends on !COMPILE_TEST
208 help
209 This will try to automatically determine if the current tree is a
210 release tree by looking for git tags that belong to the current
211 top of tree revision.
212
213 A string of the format -gxxxxxxxx will be added to the localversion
214 if a git-based tree is found. The string generated by this will be
215 appended after any matching localversion* files, and after the value
216 set in CONFIG_LOCALVERSION.
217
218 (The actual string used here is the first 12 characters produced
219 by running the command:
220
221 $ git rev-parse --verify HEAD
222
223 which is done within the script "scripts/setlocalversion".)
224
225 config BUILD_SALT
226 string "Build ID Salt"
227 default ""
228 help
229 The build ID is used to link binaries and their debug info. Setting
230 this option will use the value in the calculation of the build id.
231 This is mostly useful for distributions which want to ensure the
232 build is unique between builds. It's safe to leave the default.
233
234 config HAVE_KERNEL_GZIP
235 bool
236
237 config HAVE_KERNEL_BZIP2
238 bool
239
240 config HAVE_KERNEL_LZMA
241 bool
242
243 config HAVE_KERNEL_XZ
244 bool
245
246 config HAVE_KERNEL_LZO
247 bool
248
249 config HAVE_KERNEL_LZ4
250 bool
251
252 config HAVE_KERNEL_ZSTD
253 bool
254
255 config HAVE_KERNEL_UNCOMPRESSED
256 bool
257
258 choice
259 prompt "Kernel compression mode"
260 default KERNEL_GZIP
261 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
262 help
263 The linux kernel is a kind of self-extracting executable.
264 Several compression algorithms are available, which differ
265 in efficiency, compression and decompression speed.
266 Compression speed is only relevant when building a kernel.
267 Decompression speed is relevant at each boot.
268
269 If you have any problems with bzip2 or lzma compressed
270 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
271 version of this functionality (bzip2 only), for 2.4, was
272 supplied by Christian Ludwig)
273
274 High compression options are mostly useful for users, who
275 are low on disk space (embedded systems), but for whom ram
276 size matters less.
277
278 If in doubt, select 'gzip'
279
280 config KERNEL_GZIP
281 bool "Gzip"
282 depends on HAVE_KERNEL_GZIP
283 help
284 The old and tried gzip compression. It provides a good balance
285 between compression ratio and decompression speed.
286
287 config KERNEL_BZIP2
288 bool "Bzip2"
289 depends on HAVE_KERNEL_BZIP2
290 help
291 Its compression ratio and speed is intermediate.
292 Decompression speed is slowest among the choices. The kernel
293 size is about 10% smaller with bzip2, in comparison to gzip.
294 Bzip2 uses a large amount of memory. For modern kernels you
295 will need at least 8MB RAM or more for booting.
296
297 config KERNEL_LZMA
298 bool "LZMA"
299 depends on HAVE_KERNEL_LZMA
300 help
301 This compression algorithm's ratio is best. Decompression speed
302 is between gzip and bzip2. Compression is slowest.
303 The kernel size is about 33% smaller with LZMA in comparison to gzip.
304
305 config KERNEL_XZ
306 bool "XZ"
307 depends on HAVE_KERNEL_XZ
308 help
309 XZ uses the LZMA2 algorithm and instruction set specific
310 BCJ filters which can improve compression ratio of executable
311 code. The size of the kernel is about 30% smaller with XZ in
312 comparison to gzip. On architectures for which there is a BCJ
313 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
314 will create a few percent smaller kernel than plain LZMA.
315
316 The speed is about the same as with LZMA: The decompression
317 speed of XZ is better than that of bzip2 but worse than gzip
318 and LZO. Compression is slow.
319
320 config KERNEL_LZO
321 bool "LZO"
322 depends on HAVE_KERNEL_LZO
323 help
324 Its compression ratio is the poorest among the choices. The kernel
325 size is about 10% bigger than gzip; however its speed
326 (both compression and decompression) is the fastest.
327
328 config KERNEL_LZ4
329 bool "LZ4"
330 depends on HAVE_KERNEL_LZ4
331 help
332 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
333 A preliminary version of LZ4 de/compression tool is available at
334 <https://code.google.com/p/lz4/>.
335
336 Its compression ratio is worse than LZO. The size of the kernel
337 is about 8% bigger than LZO. But the decompression speed is
338 faster than LZO.
339
340 config KERNEL_ZSTD
341 bool "ZSTD"
342 depends on HAVE_KERNEL_ZSTD
343 help
344 ZSTD is a compression algorithm targeting intermediate compression
345 with fast decompression speed. It will compress better than GZIP and
346 decompress around the same speed as LZO, but slower than LZ4. You
347 will need at least 192 KB RAM or more for booting. The zstd command
348 line tool is required for compression.
349
350 config KERNEL_UNCOMPRESSED
351 bool "None"
352 depends on HAVE_KERNEL_UNCOMPRESSED
353 help
354 Produce uncompressed kernel image. This option is usually not what
355 you want. It is useful for debugging the kernel in slow simulation
356 environments, where decompressing and moving the kernel is awfully
357 slow. This option allows early boot code to skip the decompressor
358 and jump right at uncompressed kernel image.
359
360 endchoice
361
362 config DEFAULT_INIT
363 string "Default init path"
364 default ""
365 help
366 This option determines the default init for the system if no init=
367 option is passed on the kernel command line. If the requested path is
368 not present, we will still then move on to attempting further
369 locations (e.g. /sbin/init, etc). If this is empty, we will just use
370 the fallback list when init= is not passed.
371
372 config DEFAULT_HOSTNAME
373 string "Default hostname"
374 default "(none)"
375 help
376 This option determines the default system hostname before userspace
377 calls sethostname(2). The kernel traditionally uses "(none)" here,
378 but you may wish to use a different default here to make a minimal
379 system more usable with less configuration.
380
381 config SYSVIPC
382 bool "System V IPC"
383 help
384 Inter Process Communication is a suite of library functions and
385 system calls which let processes (running programs) synchronize and
386 exchange information. It is generally considered to be a good thing,
387 and some programs won't run unless you say Y here. In particular, if
388 you want to run the DOS emulator dosemu under Linux (read the
389 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
390 you'll need to say Y here.
391
392 You can find documentation about IPC with "info ipc" and also in
393 section 6.4 of the Linux Programmer's Guide, available from
394 <http://www.tldp.org/guides.html>.
395
396 config SYSVIPC_SYSCTL
397 bool
398 depends on SYSVIPC
399 depends on SYSCTL
400 default y
401
402 config SYSVIPC_COMPAT
403 def_bool y
404 depends on COMPAT && SYSVIPC
405
406 config POSIX_MQUEUE
407 bool "POSIX Message Queues"
408 depends on NET
409 help
410 POSIX variant of message queues is a part of IPC. In POSIX message
411 queues every message has a priority which decides about succession
412 of receiving it by a process. If you want to compile and run
413 programs written e.g. for Solaris with use of its POSIX message
414 queues (functions mq_*) say Y here.
415
416 POSIX message queues are visible as a filesystem called 'mqueue'
417 and can be mounted somewhere if you want to do filesystem
418 operations on message queues.
419
420 If unsure, say Y.
421
422 config POSIX_MQUEUE_SYSCTL
423 bool
424 depends on POSIX_MQUEUE
425 depends on SYSCTL
426 default y
427
428 config WATCH_QUEUE
429 bool "General notification queue"
430 default n
431 help
432
433 This is a general notification queue for the kernel to pass events to
434 userspace by splicing them into pipes. It can be used in conjunction
435 with watches for key/keyring change notifications and device
436 notifications.
437
438 See Documentation/core-api/watch_queue.rst
439
440 config CROSS_MEMORY_ATTACH
441 bool "Enable process_vm_readv/writev syscalls"
442 depends on MMU
443 default y
444 help
445 Enabling this option adds the system calls process_vm_readv and
446 process_vm_writev which allow a process with the correct privileges
447 to directly read from or write to another process' address space.
448 See the man page for more details.
449
450 config USELIB
451 bool "uselib syscall (for libc5 and earlier)"
452 default ALPHA || M68K || SPARC
453 help
454 This option enables the uselib syscall, a system call used in the
455 dynamic linker from libc5 and earlier. glibc does not use this
456 system call. If you intend to run programs built on libc5 or
457 earlier, you may need to enable this syscall. Current systems
458 running glibc can safely disable this.
459
460 config AUDIT
461 bool "Auditing support"
462 depends on NET
463 help
464 Enable auditing infrastructure that can be used with another
465 kernel subsystem, such as SELinux (which requires this for
466 logging of avc messages output). System call auditing is included
467 on architectures which support it.
468
469 config HAVE_ARCH_AUDITSYSCALL
470 bool
471
472 config AUDITSYSCALL
473 def_bool y
474 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL
475 select FSNOTIFY
476
477 source "kernel/irq/Kconfig"
478 source "kernel/time/Kconfig"
479 source "kernel/bpf/Kconfig"
480 source "kernel/Kconfig.preempt"
481
482 menu "CPU/Task time and stats accounting"
483
484 config VIRT_CPU_ACCOUNTING
485 bool
486
487 choice
488 prompt "Cputime accounting"
489 default TICK_CPU_ACCOUNTING
490
491 # Kind of a stub config for the pure tick based cputime accounting
492 config TICK_CPU_ACCOUNTING
493 bool "Simple tick based cputime accounting"
494 depends on !S390 && !NO_HZ_FULL
495 help
496 This is the basic tick based cputime accounting that maintains
497 statistics about user, system and idle time spent on per jiffies
498 granularity.
499
500 If unsure, say Y.
501
502 config VIRT_CPU_ACCOUNTING_NATIVE
503 bool "Deterministic task and CPU time accounting"
504 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL
505 select VIRT_CPU_ACCOUNTING
506 help
507 Select this option to enable more accurate task and CPU time
508 accounting. This is done by reading a CPU counter on each
509 kernel entry and exit and on transitions within the kernel
510 between system, softirq and hardirq state, so there is a
511 small performance impact. In the case of s390 or IBM POWER > 5,
512 this also enables accounting of stolen time on logically-partitioned
513 systems.
514
515 config VIRT_CPU_ACCOUNTING_GEN
516 bool "Full dynticks CPU time accounting"
517 depends on HAVE_CONTEXT_TRACKING_USER
518 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN
519 depends on GENERIC_CLOCKEVENTS
520 select VIRT_CPU_ACCOUNTING
521 select CONTEXT_TRACKING_USER
522 help
523 Select this option to enable task and CPU time accounting on full
524 dynticks systems. This accounting is implemented by watching every
525 kernel-user boundaries using the context tracking subsystem.
526 The accounting is thus performed at the expense of some significant
527 overhead.
528
529 For now this is only useful if you are working on the full
530 dynticks subsystem development.
531
532 If unsure, say N.
533
534 endchoice
535
536 config IRQ_TIME_ACCOUNTING
537 bool "Fine granularity task level IRQ time accounting"
538 depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE
539 help
540 Select this option to enable fine granularity task irq time
541 accounting. This is done by reading a timestamp on each
542 transitions between softirq and hardirq state, so there can be a
543 small performance impact.
544
545 If in doubt, say N here.
546
547 config HAVE_SCHED_AVG_IRQ
548 def_bool y
549 depends on IRQ_TIME_ACCOUNTING || PARAVIRT_TIME_ACCOUNTING
550 depends on SMP
551
552 config SCHED_HW_PRESSURE
553 bool
554 default y if ARM && ARM_CPU_TOPOLOGY
555 default y if ARM64
556 depends on SMP
557 depends on CPU_FREQ_THERMAL
558 help
559 Select this option to enable HW pressure accounting in the
560 scheduler. HW pressure is the value conveyed to the scheduler
561 that reflects the reduction in CPU compute capacity resulted from
562 HW throttling. HW throttling occurs when the performance of
563 a CPU is capped due to high operating temperatures as an example.
564
565 If selected, the scheduler will be able to balance tasks accordingly,
566 i.e. put less load on throttled CPUs than on non/less throttled ones.
567
568 This requires the architecture to implement
569 arch_update_hw_pressure() and arch_scale_thermal_pressure().
570
571 config BSD_PROCESS_ACCT
572 bool "BSD Process Accounting"
573 depends on MULTIUSER
574 help
575 If you say Y here, a user level program will be able to instruct the
576 kernel (via a special system call) to write process accounting
577 information to a file: whenever a process exits, information about
578 that process will be appended to the file by the kernel. The
579 information includes things such as creation time, owning user,
580 command name, memory usage, controlling terminal etc. (the complete
581 list is in the struct acct in <file:include/linux/acct.h>). It is
582 up to the user level program to do useful things with this
583 information. This is generally a good idea, so say Y.
584
585 config BSD_PROCESS_ACCT_V3
586 bool "BSD Process Accounting version 3 file format"
587 depends on BSD_PROCESS_ACCT
588 default n
589 help
590 If you say Y here, the process accounting information is written
591 in a new file format that also logs the process IDs of each
592 process and its parent. Note that this file format is incompatible
593 with previous v0/v1/v2 file formats, so you will need updated tools
594 for processing it. A preliminary version of these tools is available
595 at <http://www.gnu.org/software/acct/>.
596
597 config TASKSTATS
598 bool "Export task/process statistics through netlink"
599 depends on NET
600 depends on MULTIUSER
601 default n
602 help
603 Export selected statistics for tasks/processes through the
604 generic netlink interface. Unlike BSD process accounting, the
605 statistics are available during the lifetime of tasks/processes as
606 responses to commands. Like BSD accounting, they are sent to user
607 space on task exit.
608
609 Say N if unsure.
610
611 config TASK_DELAY_ACCT
612 bool "Enable per-task delay accounting"
613 depends on TASKSTATS
614 select SCHED_INFO
615 help
616 Collect information on time spent by a task waiting for system
617 resources like cpu, synchronous block I/O completion and swapping
618 in pages. Such statistics can help in setting a task's priorities
619 relative to other tasks for cpu, io, rss limits etc.
620
621 Say N if unsure.
622
623 config TASK_XACCT
624 bool "Enable extended accounting over taskstats"
625 depends on TASKSTATS
626 help
627 Collect extended task accounting data and send the data
628 to userland for processing over the taskstats interface.
629
630 Say N if unsure.
631
632 config TASK_IO_ACCOUNTING
633 bool "Enable per-task storage I/O accounting"
634 depends on TASK_XACCT
635 help
636 Collect information on the number of bytes of storage I/O which this
637 task has caused.
638
639 Say N if unsure.
640
641 config PSI
642 bool "Pressure stall information tracking"
643 select KERNFS
644 help
645 Collect metrics that indicate how overcommitted the CPU, memory,
646 and IO capacity are in the system.
647
648 If you say Y here, the kernel will create /proc/pressure/ with the
649 pressure statistics files cpu, memory, and io. These will indicate
650 the share of walltime in which some or all tasks in the system are
651 delayed due to contention of the respective resource.
652
653 In kernels with cgroup support, cgroups (cgroup2 only) will
654 have cpu.pressure, memory.pressure, and io.pressure files,
655 which aggregate pressure stalls for the grouped tasks only.
656
657 For more details see Documentation/accounting/psi.rst.
658
659 Say N if unsure.
660
661 config PSI_DEFAULT_DISABLED
662 bool "Require boot parameter to enable pressure stall information tracking"
663 default n
664 depends on PSI
665 help
666 If set, pressure stall information tracking will be disabled
667 per default but can be enabled through passing psi=1 on the
668 kernel commandline during boot.
669
670 This feature adds some code to the task wakeup and sleep
671 paths of the scheduler. The overhead is too low to affect
672 common scheduling-intense workloads in practice (such as
673 webservers, memcache), but it does show up in artificial
674 scheduler stress tests, such as hackbench.
675
676 If you are paranoid and not sure what the kernel will be
677 used for, say Y.
678
679 Say N if unsure.
680
681 endmenu # "CPU/Task time and stats accounting"
682
683 config CPU_ISOLATION
684 bool "CPU isolation"
685 depends on SMP || COMPILE_TEST
686 default y
687 help
688 Make sure that CPUs running critical tasks are not disturbed by
689 any source of "noise" such as unbound workqueues, timers, kthreads...
690 Unbound jobs get offloaded to housekeeping CPUs. This is driven by
691 the "isolcpus=" boot parameter.
692
693 Say Y if unsure.
694
695 source "kernel/rcu/Kconfig"
696
697 config IKCONFIG
698 tristate "Kernel .config support"
699 help
700 This option enables the complete Linux kernel ".config" file
701 contents to be saved in the kernel. It provides documentation
702 of which kernel options are used in a running kernel or in an
703 on-disk kernel. This information can be extracted from the kernel
704 image file with the script scripts/extract-ikconfig and used as
705 input to rebuild the current kernel or to build another kernel.
706 It can also be extracted from a running kernel by reading
707 /proc/config.gz if enabled (below).
708
709 config IKCONFIG_PROC
710 bool "Enable access to .config through /proc/config.gz"
711 depends on IKCONFIG && PROC_FS
712 help
713 This option enables access to the kernel configuration file
714 through /proc/config.gz.
715
716 config IKHEADERS
717 tristate "Enable kernel headers through /sys/kernel/kheaders.tar.xz"
718 depends on SYSFS
719 help
720 This option enables access to the in-kernel headers that are generated during
721 the build process. These can be used to build eBPF tracing programs,
722 or similar programs. If you build the headers as a module, a module called
723 kheaders.ko is built which can be loaded on-demand to get access to headers.
724
725 config LOG_BUF_SHIFT
726 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
727 range 12 25
728 default 17
729 depends on PRINTK
730 help
731 Select the minimal kernel log buffer size as a power of 2.
732 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config
733 parameter, see below. Any higher size also might be forced
734 by "log_buf_len" boot parameter.
735
736 Examples:
737 17 => 128 KB
738 16 => 64 KB
739 15 => 32 KB
740 14 => 16 KB
741 13 => 8 KB
742 12 => 4 KB
743
744 config LOG_CPU_MAX_BUF_SHIFT
745 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)"
746 depends on SMP
747 range 0 21
748 default 0 if BASE_SMALL
749 default 12
750 depends on PRINTK
751 help
752 This option allows to increase the default ring buffer size
753 according to the number of CPUs. The value defines the contribution
754 of each CPU as a power of 2. The used space is typically only few
755 lines however it might be much more when problems are reported,
756 e.g. backtraces.
757
758 The increased size means that a new buffer has to be allocated and
759 the original static one is unused. It makes sense only on systems
760 with more CPUs. Therefore this value is used only when the sum of
761 contributions is greater than the half of the default kernel ring
762 buffer as defined by LOG_BUF_SHIFT. The default values are set
763 so that more than 16 CPUs are needed to trigger the allocation.
764
765 Also this option is ignored when "log_buf_len" kernel parameter is
766 used as it forces an exact (power of two) size of the ring buffer.
767
768 The number of possible CPUs is used for this computation ignoring
769 hotplugging making the computation optimal for the worst case
770 scenario while allowing a simple algorithm to be used from bootup.
771
772 Examples shift values and their meaning:
773 17 => 128 KB for each CPU
774 16 => 64 KB for each CPU
775 15 => 32 KB for each CPU
776 14 => 16 KB for each CPU
777 13 => 8 KB for each CPU
778 12 => 4 KB for each CPU
779
780 config PRINTK_INDEX
781 bool "Printk indexing debugfs interface"
782 depends on PRINTK && DEBUG_FS
783 help
784 Add support for indexing of all printk formats known at compile time
785 at <debugfs>/printk/index/<module>.
786
787 This can be used as part of maintaining daemons which monitor
788 /dev/kmsg, as it permits auditing the printk formats present in a
789 kernel, allowing detection of cases where monitored printks are
790 changed or no longer present.
791
792 There is no additional runtime cost to printk with this enabled.
793
794 #
795 # Architectures with an unreliable sched_clock() should select this:
796 #
797 config HAVE_UNSTABLE_SCHED_CLOCK
798 bool
799
800 config GENERIC_SCHED_CLOCK
801 bool
802
803 menu "Scheduler features"
804
805 config UCLAMP_TASK
806 bool "Enable utilization clamping for RT/FAIR tasks"
807 depends on CPU_FREQ_GOV_SCHEDUTIL
808 help
809 This feature enables the scheduler to track the clamped utilization
810 of each CPU based on RUNNABLE tasks scheduled on that CPU.
811
812 With this option, the user can specify the min and max CPU
813 utilization allowed for RUNNABLE tasks. The max utilization defines
814 the maximum frequency a task should use while the min utilization
815 defines the minimum frequency it should use.
816
817 Both min and max utilization clamp values are hints to the scheduler,
818 aiming at improving its frequency selection policy, but they do not
819 enforce or grant any specific bandwidth for tasks.
820
821 If in doubt, say N.
822
823 config UCLAMP_BUCKETS_COUNT
824 int "Number of supported utilization clamp buckets"
825 range 5 20
826 default 5
827 depends on UCLAMP_TASK
828 help
829 Defines the number of clamp buckets to use. The range of each bucket
830 will be SCHED_CAPACITY_SCALE/UCLAMP_BUCKETS_COUNT. The higher the
831 number of clamp buckets the finer their granularity and the higher
832 the precision of clamping aggregation and tracking at run-time.
833
834 For example, with the minimum configuration value we will have 5
835 clamp buckets tracking 20% utilization each. A 25% boosted tasks will
836 be refcounted in the [20..39]% bucket and will set the bucket clamp
837 effective value to 25%.
838 If a second 30% boosted task should be co-scheduled on the same CPU,
839 that task will be refcounted in the same bucket of the first task and
840 it will boost the bucket clamp effective value to 30%.
841 The clamp effective value of a bucket is reset to its nominal value
842 (20% in the example above) when there are no more tasks refcounted in
843 that bucket.
844
845 An additional boost/capping margin can be added to some tasks. In the
846 example above the 25% task will be boosted to 30% until it exits the
847 CPU. If that should be considered not acceptable on certain systems,
848 it's always possible to reduce the margin by increasing the number of
849 clamp buckets to trade off used memory for run-time tracking
850 precision.
851
852 If in doubt, use the default value.
853
854 endmenu
855
856 #
857 # For architectures that want to enable the support for NUMA-affine scheduler
858 # balancing logic:
859 #
860 config ARCH_SUPPORTS_NUMA_BALANCING
861 bool
862
863 #
864 # For architectures that prefer to flush all TLBs after a number of pages
865 # are unmapped instead of sending one IPI per page to flush. The architecture
866 # must provide guarantees on what happens if a clean TLB cache entry is
867 # written after the unmap. Details are in mm/rmap.c near the check for
868 # should_defer_flush. The architecture should also consider if the full flush
869 # and the refill costs are offset by the savings of sending fewer IPIs.
870 config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
871 bool
872
873 config CC_HAS_INT128
874 def_bool !$(cc-option,$(m64-flag) -D__SIZEOF_INT128__=0) && 64BIT
875
876 config CC_IMPLICIT_FALLTHROUGH
877 string
878 default "-Wimplicit-fallthrough=5" if CC_IS_GCC && $(cc-option,-Wimplicit-fallthrough=5)
879 default "-Wimplicit-fallthrough" if CC_IS_CLANG && $(cc-option,-Wunreachable-code-fallthrough)
880
881 # Currently, disable gcc-10+ array-bounds globally.
882 # It's still broken in gcc-13, so no upper bound yet.
883 config GCC10_NO_ARRAY_BOUNDS
884 def_bool y
885
886 config CC_NO_ARRAY_BOUNDS
887 bool
888 default y if CC_IS_GCC && GCC_VERSION >= 90000 && GCC10_NO_ARRAY_BOUNDS
889
890 # Currently, disable -Wstringop-overflow for GCC globally.
891 config GCC_NO_STRINGOP_OVERFLOW
892 def_bool y
893
894 config CC_NO_STRINGOP_OVERFLOW
895 bool
896 default y if CC_IS_GCC && GCC_NO_STRINGOP_OVERFLOW
897
898 config CC_STRINGOP_OVERFLOW
899 bool
900 default y if CC_IS_GCC && !CC_NO_STRINGOP_OVERFLOW
901
902 #
903 # For architectures that know their GCC __int128 support is sound
904 #
905 config ARCH_SUPPORTS_INT128
906 bool
907
908 # For architectures that (ab)use NUMA to represent different memory regions
909 # all cpu-local but of different latencies, such as SuperH.
910 #
911 config ARCH_WANT_NUMA_VARIABLE_LOCALITY
912 bool
913
914 config NUMA_BALANCING
915 bool "Memory placement aware NUMA scheduler"
916 depends on ARCH_SUPPORTS_NUMA_BALANCING
917 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
918 depends on SMP && NUMA && MIGRATION && !PREEMPT_RT
919 help
920 This option adds support for automatic NUMA aware memory/task placement.
921 The mechanism is quite primitive and is based on migrating memory when
922 it has references to the node the task is running on.
923
924 This system will be inactive on UMA systems.
925
926 config NUMA_BALANCING_DEFAULT_ENABLED
927 bool "Automatically enable NUMA aware memory/task placement"
928 default y
929 depends on NUMA_BALANCING
930 help
931 If set, automatic NUMA balancing will be enabled if running on a NUMA
932 machine.
933
934 config SLAB_OBJ_EXT
935 bool
936
937 menuconfig CGROUPS
938 bool "Control Group support"
939 select KERNFS
940 help
941 This option adds support for grouping sets of processes together, for
942 use with process control subsystems such as Cpusets, CFS, memory
943 controls or device isolation.
944 See
945 - Documentation/scheduler/sched-design-CFS.rst (CFS)
946 - Documentation/admin-guide/cgroup-v1/ (features for grouping, isolation
947 and resource control)
948
949 Say N if unsure.
950
951 if CGROUPS
952
953 config PAGE_COUNTER
954 bool
955
956 config CGROUP_FAVOR_DYNMODS
957 bool "Favor dynamic modification latency reduction by default"
958 help
959 This option enables the "favordynmods" mount option by default
960 which reduces the latencies of dynamic cgroup modifications such
961 as task migrations and controller on/offs at the cost of making
962 hot path operations such as forks and exits more expensive.
963
964 Say N if unsure.
965
966 config MEMCG
967 bool "Memory controller"
968 select PAGE_COUNTER
969 select EVENTFD
970 select SLAB_OBJ_EXT
971 help
972 Provides control over the memory footprint of tasks in a cgroup.
973
974 config MEMCG_V1
975 bool "Legacy cgroup v1 memory controller"
976 depends on MEMCG
977 default n
978 help
979 Legacy cgroup v1 memory controller which has been deprecated by
980 cgroup v2 implementation. The v1 is there for legacy applications
981 which haven't migrated to the new cgroup v2 interface yet. If you
982 do not have any such application then you are completely fine leaving
983 this option disabled.
984
985 Please note that feature set of the legacy memory controller is likely
986 going to shrink due to deprecation process. New deployments with v1
987 controller are highly discouraged.
988
989 San N is unsure.
990
991 config BLK_CGROUP
992 bool "IO controller"
993 depends on BLOCK
994 default n
995 help
996 Generic block IO controller cgroup interface. This is the common
997 cgroup interface which should be used by various IO controlling
998 policies.
999
1000 Currently, CFQ IO scheduler uses it to recognize task groups and
1001 control disk bandwidth allocation (proportional time slice allocation)
1002 to such task groups. It is also used by bio throttling logic in
1003 block layer to implement upper limit in IO rates on a device.
1004
1005 This option only enables generic Block IO controller infrastructure.
1006 One needs to also enable actual IO controlling logic/policy. For
1007 enabling proportional weight division of disk bandwidth in CFQ, set
1008 CONFIG_BFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
1009 CONFIG_BLK_DEV_THROTTLING=y.
1010
1011 See Documentation/admin-guide/cgroup-v1/blkio-controller.rst for more information.
1012
1013 config CGROUP_WRITEBACK
1014 bool
1015 depends on MEMCG && BLK_CGROUP
1016 default y
1017
1018 menuconfig CGROUP_SCHED
1019 bool "CPU controller"
1020 default n
1021 help
1022 This feature lets CPU scheduler recognize task groups and control CPU
1023 bandwidth allocation to such task groups. It uses cgroups to group
1024 tasks.
1025
1026 if CGROUP_SCHED
1027 config FAIR_GROUP_SCHED
1028 bool "Group scheduling for SCHED_OTHER"
1029 depends on CGROUP_SCHED
1030 default CGROUP_SCHED
1031
1032 config CFS_BANDWIDTH
1033 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
1034 depends on FAIR_GROUP_SCHED
1035 default n
1036 help
1037 This option allows users to define CPU bandwidth rates (limits) for
1038 tasks running within the fair group scheduler. Groups with no limit
1039 set are considered to be unconstrained and will run with no
1040 restriction.
1041 See Documentation/scheduler/sched-bwc.rst for more information.
1042
1043 config RT_GROUP_SCHED
1044 bool "Group scheduling for SCHED_RR/FIFO"
1045 depends on CGROUP_SCHED
1046 default n
1047 help
1048 This feature lets you explicitly allocate real CPU bandwidth
1049 to task groups. If enabled, it will also make it impossible to
1050 schedule realtime tasks for non-root users until you allocate
1051 realtime bandwidth for them.
1052 See Documentation/scheduler/sched-rt-group.rst for more information.
1053
1054 endif #CGROUP_SCHED
1055
1056 config SCHED_MM_CID
1057 def_bool y
1058 depends on SMP && RSEQ
1059
1060 config UCLAMP_TASK_GROUP
1061 bool "Utilization clamping per group of tasks"
1062 depends on CGROUP_SCHED
1063 depends on UCLAMP_TASK
1064 default n
1065 help
1066 This feature enables the scheduler to track the clamped utilization
1067 of each CPU based on RUNNABLE tasks currently scheduled on that CPU.
1068
1069 When this option is enabled, the user can specify a min and max
1070 CPU bandwidth which is allowed for each single task in a group.
1071 The max bandwidth allows to clamp the maximum frequency a task
1072 can use, while the min bandwidth allows to define a minimum
1073 frequency a task will always use.
1074
1075 When task group based utilization clamping is enabled, an eventually
1076 specified task-specific clamp value is constrained by the cgroup
1077 specified clamp value. Both minimum and maximum task clamping cannot
1078 be bigger than the corresponding clamping defined at task group level.
1079
1080 If in doubt, say N.
1081
1082 config CGROUP_PIDS
1083 bool "PIDs controller"
1084 help
1085 Provides enforcement of process number limits in the scope of a
1086 cgroup. Any attempt to fork more processes than is allowed in the
1087 cgroup will fail. PIDs are fundamentally a global resource because it
1088 is fairly trivial to reach PID exhaustion before you reach even a
1089 conservative kmemcg limit. As a result, it is possible to grind a
1090 system to halt without being limited by other cgroup policies. The
1091 PIDs controller is designed to stop this from happening.
1092
1093 It should be noted that organisational operations (such as attaching
1094 to a cgroup hierarchy) will *not* be blocked by the PIDs controller,
1095 since the PIDs limit only affects a process's ability to fork, not to
1096 attach to a cgroup.
1097
1098 config CGROUP_RDMA
1099 bool "RDMA controller"
1100 help
1101 Provides enforcement of RDMA resources defined by IB stack.
1102 It is fairly easy for consumers to exhaust RDMA resources, which
1103 can result into resource unavailability to other consumers.
1104 RDMA controller is designed to stop this from happening.
1105 Attaching processes with active RDMA resources to the cgroup
1106 hierarchy is allowed even if can cross the hierarchy's limit.
1107
1108 config CGROUP_FREEZER
1109 bool "Freezer controller"
1110 help
1111 Provides a way to freeze and unfreeze all tasks in a
1112 cgroup.
1113
1114 This option affects the ORIGINAL cgroup interface. The cgroup2 memory
1115 controller includes important in-kernel memory consumers per default.
1116
1117 If you're using cgroup2, say N.
1118
1119 config CGROUP_HUGETLB
1120 bool "HugeTLB controller"
1121 depends on HUGETLB_PAGE
1122 select PAGE_COUNTER
1123 default n
1124 help
1125 Provides a cgroup controller for HugeTLB pages.
1126 When you enable this, you can put a per cgroup limit on HugeTLB usage.
1127 The limit is enforced during page fault. Since HugeTLB doesn't
1128 support page reclaim, enforcing the limit at page fault time implies
1129 that, the application will get SIGBUS signal if it tries to access
1130 HugeTLB pages beyond its limit. This requires the application to know
1131 beforehand how much HugeTLB pages it would require for its use. The
1132 control group is tracked in the third page lru pointer. This means
1133 that we cannot use the controller with huge page less than 3 pages.
1134
1135 config CPUSETS
1136 bool "Cpuset controller"
1137 depends on SMP
1138 help
1139 This option will let you create and manage CPUSETs which
1140 allow dynamically partitioning a system into sets of CPUs and
1141 Memory Nodes and assigning tasks to run only within those sets.
1142 This is primarily useful on large SMP or NUMA systems.
1143
1144 Say N if unsure.
1145
1146 config PROC_PID_CPUSET
1147 bool "Include legacy /proc/<pid>/cpuset file"
1148 depends on CPUSETS
1149 default y
1150
1151 config CGROUP_DEVICE
1152 bool "Device controller"
1153 help
1154 Provides a cgroup controller implementing whitelists for
1155 devices which a process in the cgroup can mknod or open.
1156
1157 config CGROUP_CPUACCT
1158 bool "Simple CPU accounting controller"
1159 help
1160 Provides a simple controller for monitoring the
1161 total CPU consumed by the tasks in a cgroup.
1162
1163 config CGROUP_PERF
1164 bool "Perf controller"
1165 depends on PERF_EVENTS
1166 help
1167 This option extends the perf per-cpu mode to restrict monitoring
1168 to threads which belong to the cgroup specified and run on the
1169 designated cpu. Or this can be used to have cgroup ID in samples
1170 so that it can monitor performance events among cgroups.
1171
1172 Say N if unsure.
1173
1174 config CGROUP_BPF
1175 bool "Support for eBPF programs attached to cgroups"
1176 depends on BPF_SYSCALL
1177 select SOCK_CGROUP_DATA
1178 help
1179 Allow attaching eBPF programs to a cgroup using the bpf(2)
1180 syscall command BPF_PROG_ATTACH.
1181
1182 In which context these programs are accessed depends on the type
1183 of attachment. For instance, programs that are attached using
1184 BPF_CGROUP_INET_INGRESS will be executed on the ingress path of
1185 inet sockets.
1186
1187 config CGROUP_MISC
1188 bool "Misc resource controller"
1189 default n
1190 help
1191 Provides a controller for miscellaneous resources on a host.
1192
1193 Miscellaneous scalar resources are the resources on the host system
1194 which cannot be abstracted like the other cgroups. This controller
1195 tracks and limits the miscellaneous resources used by a process
1196 attached to a cgroup hierarchy.
1197
1198 For more information, please check misc cgroup section in
1199 /Documentation/admin-guide/cgroup-v2.rst.
1200
1201 config CGROUP_DEBUG
1202 bool "Debug controller"
1203 default n
1204 depends on DEBUG_KERNEL
1205 help
1206 This option enables a simple controller that exports
1207 debugging information about the cgroups framework. This
1208 controller is for control cgroup debugging only. Its
1209 interfaces are not stable.
1210
1211 Say N.
1212
1213 config SOCK_CGROUP_DATA
1214 bool
1215 default n
1216
1217 endif # CGROUPS
1218
1219 menuconfig NAMESPACES
1220 bool "Namespaces support" if EXPERT
1221 depends on MULTIUSER
1222 default !EXPERT
1223 help
1224 Provides the way to make tasks work with different objects using
1225 the same id. For example same IPC id may refer to different objects
1226 or same user id or pid may refer to different tasks when used in
1227 different namespaces.
1228
1229 if NAMESPACES
1230
1231 config UTS_NS
1232 bool "UTS namespace"
1233 default y
1234 help
1235 In this namespace tasks see different info provided with the
1236 uname() system call
1237
1238 config TIME_NS
1239 bool "TIME namespace"
1240 depends on GENERIC_VDSO_TIME_NS
1241 default y
1242 help
1243 In this namespace boottime and monotonic clocks can be set.
1244 The time will keep going with the same pace.
1245
1246 config IPC_NS
1247 bool "IPC namespace"
1248 depends on (SYSVIPC || POSIX_MQUEUE)
1249 default y
1250 help
1251 In this namespace tasks work with IPC ids which correspond to
1252 different IPC objects in different namespaces.
1253
1254 config USER_NS
1255 bool "User namespace"
1256 default n
1257 help
1258 This allows containers, i.e. vservers, to use user namespaces
1259 to provide different user info for different servers.
1260
1261 When user namespaces are enabled in the kernel it is
1262 recommended that the MEMCG option also be enabled and that
1263 user-space use the memory control groups to limit the amount
1264 of memory a memory unprivileged users can use.
1265
1266 If unsure, say N.
1267
1268 config PID_NS
1269 bool "PID Namespaces"
1270 default y
1271 help
1272 Support process id namespaces. This allows having multiple
1273 processes with the same pid as long as they are in different
1274 pid namespaces. This is a building block of containers.
1275
1276 config NET_NS
1277 bool "Network namespace"
1278 depends on NET
1279 default y
1280 help
1281 Allow user space to create what appear to be multiple instances
1282 of the network stack.
1283
1284 endif # NAMESPACES
1285
1286 config CHECKPOINT_RESTORE
1287 bool "Checkpoint/restore support"
1288 depends on PROC_FS
1289 select PROC_CHILDREN
1290 select KCMP
1291 default n
1292 help
1293 Enables additional kernel features in a sake of checkpoint/restore.
1294 In particular it adds auxiliary prctl codes to setup process text,
1295 data and heap segment sizes, and a few additional /proc filesystem
1296 entries.
1297
1298 If unsure, say N here.
1299
1300 config SCHED_AUTOGROUP
1301 bool "Automatic process group scheduling"
1302 select CGROUPS
1303 select CGROUP_SCHED
1304 select FAIR_GROUP_SCHED
1305 help
1306 This option optimizes the scheduler for common desktop workloads by
1307 automatically creating and populating task groups. This separation
1308 of workloads isolates aggressive CPU burners (like build jobs) from
1309 desktop applications. Task group autogeneration is currently based
1310 upon task session.
1311
1312 config RELAY
1313 bool "Kernel->user space relay support (formerly relayfs)"
1314 select IRQ_WORK
1315 help
1316 This option enables support for relay interface support in
1317 certain file systems (such as debugfs).
1318 It is designed to provide an efficient mechanism for tools and
1319 facilities to relay large amounts of data from kernel space to
1320 user space.
1321
1322 If unsure, say N.
1323
1324 config BLK_DEV_INITRD
1325 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
1326 help
1327 The initial RAM filesystem is a ramfs which is loaded by the
1328 boot loader (loadlin or lilo) and that is mounted as root
1329 before the normal boot procedure. It is typically used to
1330 load modules needed to mount the "real" root file system,
1331 etc. See <file:Documentation/admin-guide/initrd.rst> for details.
1332
1333 If RAM disk support (BLK_DEV_RAM) is also included, this
1334 also enables initial RAM disk (initrd) support and adds
1335 15 Kbytes (more on some other architectures) to the kernel size.
1336
1337 If unsure say Y.
1338
1339 if BLK_DEV_INITRD
1340
1341 source "usr/Kconfig"
1342
1343 endif
1344
1345 config BOOT_CONFIG
1346 bool "Boot config support"
1347 select BLK_DEV_INITRD if !BOOT_CONFIG_EMBED
1348 help
1349 Extra boot config allows system admin to pass a config file as
1350 complemental extension of kernel cmdline when booting.
1351 The boot config file must be attached at the end of initramfs
1352 with checksum, size and magic word.
1353 See <file:Documentation/admin-guide/bootconfig.rst> for details.
1354
1355 If unsure, say Y.
1356
1357 config BOOT_CONFIG_FORCE
1358 bool "Force unconditional bootconfig processing"
1359 depends on BOOT_CONFIG
1360 default y if BOOT_CONFIG_EMBED
1361 help
1362 With this Kconfig option set, BOOT_CONFIG processing is carried
1363 out even when the "bootconfig" kernel-boot parameter is omitted.
1364 In fact, with this Kconfig option set, there is no way to
1365 make the kernel ignore the BOOT_CONFIG-supplied kernel-boot
1366 parameters.
1367
1368 If unsure, say N.
1369
1370 config BOOT_CONFIG_EMBED
1371 bool "Embed bootconfig file in the kernel"
1372 depends on BOOT_CONFIG
1373 help
1374 Embed a bootconfig file given by BOOT_CONFIG_EMBED_FILE in the
1375 kernel. Usually, the bootconfig file is loaded with the initrd
1376 image. But if the system doesn't support initrd, this option will
1377 help you by embedding a bootconfig file while building the kernel.
1378
1379 If unsure, say N.
1380
1381 config BOOT_CONFIG_EMBED_FILE
1382 string "Embedded bootconfig file path"
1383 depends on BOOT_CONFIG_EMBED
1384 help
1385 Specify a bootconfig file which will be embedded to the kernel.
1386 This bootconfig will be used if there is no initrd or no other
1387 bootconfig in the initrd.
1388
1389 config INITRAMFS_PRESERVE_MTIME
1390 bool "Preserve cpio archive mtimes in initramfs"
1391 default y
1392 help
1393 Each entry in an initramfs cpio archive carries an mtime value. When
1394 enabled, extracted cpio items take this mtime, with directory mtime
1395 setting deferred until after creation of any child entries.
1396
1397 If unsure, say Y.
1398
1399 choice
1400 prompt "Compiler optimization level"
1401 default CC_OPTIMIZE_FOR_PERFORMANCE
1402
1403 config CC_OPTIMIZE_FOR_PERFORMANCE
1404 bool "Optimize for performance (-O2)"
1405 help
1406 This is the default optimization level for the kernel, building
1407 with the "-O2" compiler flag for best performance and most
1408 helpful compile-time warnings.
1409
1410 config CC_OPTIMIZE_FOR_SIZE
1411 bool "Optimize for size (-Os)"
1412 help
1413 Choosing this option will pass "-Os" to your compiler resulting
1414 in a smaller kernel.
1415
1416 endchoice
1417
1418 config HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1419 bool
1420 help
1421 This requires that the arch annotates or otherwise protects
1422 its external entry points from being discarded. Linker scripts
1423 must also merge .text.*, .data.*, and .bss.* correctly into
1424 output sections. Care must be taken not to pull in unrelated
1425 sections (e.g., '.text.init'). Typically '.' in section names
1426 is used to distinguish them from label names / C identifiers.
1427
1428 config LD_DEAD_CODE_DATA_ELIMINATION
1429 bool "Dead code and data elimination (EXPERIMENTAL)"
1430 depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1431 depends on EXPERT
1432 depends on $(cc-option,-ffunction-sections -fdata-sections)
1433 depends on $(ld-option,--gc-sections)
1434 help
1435 Enable this if you want to do dead code and data elimination with
1436 the linker by compiling with -ffunction-sections -fdata-sections,
1437 and linking with --gc-sections.
1438
1439 This can reduce on disk and in-memory size of the kernel
1440 code and static data, particularly for small configs and
1441 on small systems. This has the possibility of introducing
1442 silently broken kernel if the required annotations are not
1443 present. This option is not well tested yet, so use at your
1444 own risk.
1445
1446 config LD_ORPHAN_WARN
1447 def_bool y
1448 depends on ARCH_WANT_LD_ORPHAN_WARN
1449 depends on $(ld-option,--orphan-handling=warn)
1450 depends on $(ld-option,--orphan-handling=error)
1451
1452 config LD_ORPHAN_WARN_LEVEL
1453 string
1454 depends on LD_ORPHAN_WARN
1455 default "error" if WERROR
1456 default "warn"
1457
1458 config SYSCTL
1459 bool
1460
1461 config HAVE_UID16
1462 bool
1463
1464 config SYSCTL_EXCEPTION_TRACE
1465 bool
1466 help
1467 Enable support for /proc/sys/debug/exception-trace.
1468
1469 config SYSCTL_ARCH_UNALIGN_NO_WARN
1470 bool
1471 help
1472 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1473 Allows arch to define/use @no_unaligned_warning to possibly warn
1474 about unaligned access emulation going on under the hood.
1475
1476 config SYSCTL_ARCH_UNALIGN_ALLOW
1477 bool
1478 help
1479 Enable support for /proc/sys/kernel/unaligned-trap
1480 Allows arches to define/use @unaligned_enabled to runtime toggle
1481 the unaligned access emulation.
1482 see arch/parisc/kernel/unaligned.c for reference
1483
1484 config HAVE_PCSPKR_PLATFORM
1485 bool
1486
1487 menuconfig EXPERT
1488 bool "Configure standard kernel features (expert users)"
1489 # Unhide debug options, to make the on-by-default options visible
1490 select DEBUG_KERNEL
1491 help
1492 This option allows certain base kernel options and settings
1493 to be disabled or tweaked. This is for specialized
1494 environments which can tolerate a "non-standard" kernel.
1495 Only use this if you really know what you are doing.
1496
1497 config UID16
1498 bool "Enable 16-bit UID system calls" if EXPERT
1499 depends on HAVE_UID16 && MULTIUSER
1500 default y
1501 help
1502 This enables the legacy 16-bit UID syscall wrappers.
1503
1504 config MULTIUSER
1505 bool "Multiple users, groups and capabilities support" if EXPERT
1506 default y
1507 help
1508 This option enables support for non-root users, groups and
1509 capabilities.
1510
1511 If you say N here, all processes will run with UID 0, GID 0, and all
1512 possible capabilities. Saying N here also compiles out support for
1513 system calls related to UIDs, GIDs, and capabilities, such as setuid,
1514 setgid, and capset.
1515
1516 If unsure, say Y here.
1517
1518 config SGETMASK_SYSCALL
1519 bool "sgetmask/ssetmask syscalls support" if EXPERT
1520 default PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH
1521 help
1522 sys_sgetmask and sys_ssetmask are obsolete system calls
1523 no longer supported in libc but still enabled by default in some
1524 architectures.
1525
1526 If unsure, leave the default option here.
1527
1528 config SYSFS_SYSCALL
1529 bool "Sysfs syscall support" if EXPERT
1530 default y
1531 help
1532 sys_sysfs is an obsolete system call no longer supported in libc.
1533 Note that disabling this option is more secure but might break
1534 compatibility with some systems.
1535
1536 If unsure say Y here.
1537
1538 config FHANDLE
1539 bool "open by fhandle syscalls" if EXPERT
1540 select EXPORTFS
1541 default y
1542 help
1543 If you say Y here, a user level program will be able to map
1544 file names to handle and then later use the handle for
1545 different file system operations. This is useful in implementing
1546 userspace file servers, which now track files using handles instead
1547 of names. The handle would remain the same even if file names
1548 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
1549 syscalls.
1550
1551 config POSIX_TIMERS
1552 bool "Posix Clocks & timers" if EXPERT
1553 default y
1554 help
1555 This includes native support for POSIX timers to the kernel.
1556 Some embedded systems have no use for them and therefore they
1557 can be configured out to reduce the size of the kernel image.
1558
1559 When this option is disabled, the following syscalls won't be
1560 available: timer_create, timer_gettime: timer_getoverrun,
1561 timer_settime, timer_delete, clock_adjtime, getitimer,
1562 setitimer, alarm. Furthermore, the clock_settime, clock_gettime,
1563 clock_getres and clock_nanosleep syscalls will be limited to
1564 CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only.
1565
1566 If unsure say y.
1567
1568 config PRINTK
1569 default y
1570 bool "Enable support for printk" if EXPERT
1571 select IRQ_WORK
1572 help
1573 This option enables normal printk support. Removing it
1574 eliminates most of the message strings from the kernel image
1575 and makes the kernel more or less silent. As this makes it
1576 very difficult to diagnose system problems, saying N here is
1577 strongly discouraged.
1578
1579 config BUG
1580 bool "BUG() support" if EXPERT
1581 default y
1582 help
1583 Disabling this option eliminates support for BUG and WARN, reducing
1584 the size of your kernel image and potentially quietly ignoring
1585 numerous fatal conditions. You should only consider disabling this
1586 option for embedded systems with no facilities for reporting errors.
1587 Just say Y.
1588
1589 config ELF_CORE
1590 depends on COREDUMP
1591 default y
1592 bool "Enable ELF core dumps" if EXPERT
1593 help
1594 Enable support for generating core dumps. Disabling saves about 4k.
1595
1596
1597 config PCSPKR_PLATFORM
1598 bool "Enable PC-Speaker support" if EXPERT
1599 depends on HAVE_PCSPKR_PLATFORM
1600 select I8253_LOCK
1601 default y
1602 help
1603 This option allows to disable the internal PC-Speaker
1604 support, saving some memory.
1605
1606 config BASE_SMALL
1607 bool "Enable smaller-sized data structures for core" if EXPERT
1608 help
1609 Enabling this option reduces the size of miscellaneous core
1610 kernel data structures. This saves memory on small machines,
1611 but may reduce performance.
1612
1613 config FUTEX
1614 bool "Enable futex support" if EXPERT
1615 depends on !(SPARC32 && SMP)
1616 default y
1617 imply RT_MUTEXES
1618 help
1619 Disabling this option will cause the kernel to be built without
1620 support for "fast userspace mutexes". The resulting kernel may not
1621 run glibc-based applications correctly.
1622
1623 config FUTEX_PI
1624 bool
1625 depends on FUTEX && RT_MUTEXES
1626 default y
1627
1628 config EPOLL
1629 bool "Enable eventpoll support" if EXPERT
1630 default y
1631 help
1632 Disabling this option will cause the kernel to be built without
1633 support for epoll family of system calls.
1634
1635 config SIGNALFD
1636 bool "Enable signalfd() system call" if EXPERT
1637 default y
1638 help
1639 Enable the signalfd() system call that allows to receive signals
1640 on a file descriptor.
1641
1642 If unsure, say Y.
1643
1644 config TIMERFD
1645 bool "Enable timerfd() system call" if EXPERT
1646 default y
1647 help
1648 Enable the timerfd() system call that allows to receive timer
1649 events on a file descriptor.
1650
1651 If unsure, say Y.
1652
1653 config EVENTFD
1654 bool "Enable eventfd() system call" if EXPERT
1655 default y
1656 help
1657 Enable the eventfd() system call that allows to receive both
1658 kernel notification (ie. KAIO) or userspace notifications.
1659
1660 If unsure, say Y.
1661
1662 config SHMEM
1663 bool "Use full shmem filesystem" if EXPERT
1664 default y
1665 depends on MMU
1666 help
1667 The shmem is an internal filesystem used to manage shared memory.
1668 It is backed by swap and manages resource limits. It is also exported
1669 to userspace as tmpfs if TMPFS is enabled. Disabling this
1670 option replaces shmem and tmpfs with the much simpler ramfs code,
1671 which may be appropriate on small systems without swap.
1672
1673 config AIO
1674 bool "Enable AIO support" if EXPERT
1675 default y
1676 help
1677 This option enables POSIX asynchronous I/O which may by used
1678 by some high performance threaded applications. Disabling
1679 this option saves about 7k.
1680
1681 config IO_URING
1682 bool "Enable IO uring support" if EXPERT
1683 select IO_WQ
1684 default y
1685 help
1686 This option enables support for the io_uring interface, enabling
1687 applications to submit and complete IO through submission and
1688 completion rings that are shared between the kernel and application.
1689
1690 config ADVISE_SYSCALLS
1691 bool "Enable madvise/fadvise syscalls" if EXPERT
1692 default y
1693 help
1694 This option enables the madvise and fadvise syscalls, used by
1695 applications to advise the kernel about their future memory or file
1696 usage, improving performance. If building an embedded system where no
1697 applications use these syscalls, you can disable this option to save
1698 space.
1699
1700 config MEMBARRIER
1701 bool "Enable membarrier() system call" if EXPERT
1702 default y
1703 help
1704 Enable the membarrier() system call that allows issuing memory
1705 barriers across all running threads, which can be used to distribute
1706 the cost of user-space memory barriers asymmetrically by transforming
1707 pairs of memory barriers into pairs consisting of membarrier() and a
1708 compiler barrier.
1709
1710 If unsure, say Y.
1711
1712 config KCMP
1713 bool "Enable kcmp() system call" if EXPERT
1714 help
1715 Enable the kernel resource comparison system call. It provides
1716 user-space with the ability to compare two processes to see if they
1717 share a common resource, such as a file descriptor or even virtual
1718 memory space.
1719
1720 If unsure, say N.
1721
1722 config RSEQ
1723 bool "Enable rseq() system call" if EXPERT
1724 default y
1725 depends on HAVE_RSEQ
1726 select MEMBARRIER
1727 help
1728 Enable the restartable sequences system call. It provides a
1729 user-space cache for the current CPU number value, which
1730 speeds up getting the current CPU number from user-space,
1731 as well as an ABI to speed up user-space operations on
1732 per-CPU data.
1733
1734 If unsure, say Y.
1735
1736 config DEBUG_RSEQ
1737 default n
1738 bool "Enable debugging of rseq() system call" if EXPERT
1739 depends on RSEQ && DEBUG_KERNEL
1740 help
1741 Enable extra debugging checks for the rseq system call.
1742
1743 If unsure, say N.
1744
1745 config CACHESTAT_SYSCALL
1746 bool "Enable cachestat() system call" if EXPERT
1747 default y
1748 help
1749 Enable the cachestat system call, which queries the page cache
1750 statistics of a file (number of cached pages, dirty pages,
1751 pages marked for writeback, (recently) evicted pages).
1752
1753 If unsure say Y here.
1754
1755 config PC104
1756 bool "PC/104 support" if EXPERT
1757 help
1758 Expose PC/104 form factor device drivers and options available for
1759 selection and configuration. Enable this option if your target
1760 machine has a PC/104 bus.
1761
1762 config KALLSYMS
1763 bool "Load all symbols for debugging/ksymoops" if EXPERT
1764 default y
1765 help
1766 Say Y here to let the kernel print out symbolic crash information and
1767 symbolic stack backtraces. This increases the size of the kernel
1768 somewhat, as all symbols have to be loaded into the kernel image.
1769
1770 config KALLSYMS_SELFTEST
1771 bool "Test the basic functions and performance of kallsyms"
1772 depends on KALLSYMS
1773 default n
1774 help
1775 Test the basic functions and performance of some interfaces, such as
1776 kallsyms_lookup_name. It also calculates the compression rate of the
1777 kallsyms compression algorithm for the current symbol set.
1778
1779 Start self-test automatically after system startup. Suggest executing
1780 "dmesg | grep kallsyms_selftest" to collect test results. "finish" is
1781 displayed in the last line, indicating that the test is complete.
1782
1783 config KALLSYMS_ALL
1784 bool "Include all symbols in kallsyms"
1785 depends on DEBUG_KERNEL && KALLSYMS
1786 help
1787 Normally kallsyms only contains the symbols of functions for nicer
1788 OOPS messages and backtraces (i.e., symbols from the text and inittext
1789 sections). This is sufficient for most cases. And only if you want to
1790 enable kernel live patching, or other less common use cases (e.g.,
1791 when a debugger is used) all symbols are required (i.e., names of
1792 variables from the data sections, etc).
1793
1794 This option makes sure that all symbols are loaded into the kernel
1795 image (i.e., symbols from all sections) in cost of increased kernel
1796 size (depending on the kernel configuration, it may be 300KiB or
1797 something like this).
1798
1799 Say N unless you really need all symbols, or kernel live patching.
1800
1801 config KALLSYMS_ABSOLUTE_PERCPU
1802 bool
1803 depends on KALLSYMS
1804 default X86_64 && SMP
1805
1806 # end of the "standard kernel features (expert users)" menu
1807
1808 config ARCH_HAS_MEMBARRIER_CALLBACKS
1809 bool
1810
1811 config ARCH_HAS_MEMBARRIER_SYNC_CORE
1812 bool
1813
1814 config HAVE_PERF_EVENTS
1815 bool
1816 help
1817 See tools/perf/design.txt for details.
1818
1819 config GUEST_PERF_EVENTS
1820 bool
1821 depends on HAVE_PERF_EVENTS
1822
1823 config PERF_USE_VMALLOC
1824 bool
1825 help
1826 See tools/perf/design.txt for details
1827
1828 menu "Kernel Performance Events And Counters"
1829
1830 config PERF_EVENTS
1831 bool "Kernel performance events and counters"
1832 default y if PROFILING
1833 depends on HAVE_PERF_EVENTS
1834 select IRQ_WORK
1835 help
1836 Enable kernel support for various performance events provided
1837 by software and hardware.
1838
1839 Software events are supported either built-in or via the
1840 use of generic tracepoints.
1841
1842 Most modern CPUs support performance events via performance
1843 counter registers. These registers count the number of certain
1844 types of hw events: such as instructions executed, cachemisses
1845 suffered, or branches mis-predicted - without slowing down the
1846 kernel or applications. These registers can also trigger interrupts
1847 when a threshold number of events have passed - and can thus be
1848 used to profile the code that runs on that CPU.
1849
1850 The Linux Performance Event subsystem provides an abstraction of
1851 these software and hardware event capabilities, available via a
1852 system call and used by the "perf" utility in tools/perf/. It
1853 provides per task and per CPU counters, and it provides event
1854 capabilities on top of those.
1855
1856 Say Y if unsure.
1857
1858 config DEBUG_PERF_USE_VMALLOC
1859 default n
1860 bool "Debug: use vmalloc to back perf mmap() buffers"
1861 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC
1862 select PERF_USE_VMALLOC
1863 help
1864 Use vmalloc memory to back perf mmap() buffers.
1865
1866 Mostly useful for debugging the vmalloc code on platforms
1867 that don't require it.
1868
1869 Say N if unsure.
1870
1871 endmenu
1872
1873 config SYSTEM_DATA_VERIFICATION
1874 def_bool n
1875 select SYSTEM_TRUSTED_KEYRING
1876 select KEYS
1877 select CRYPTO
1878 select CRYPTO_RSA
1879 select ASYMMETRIC_KEY_TYPE
1880 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
1881 select ASN1
1882 select OID_REGISTRY
1883 select X509_CERTIFICATE_PARSER
1884 select PKCS7_MESSAGE_PARSER
1885 help
1886 Provide PKCS#7 message verification using the contents of the system
1887 trusted keyring to provide public keys. This then can be used for
1888 module verification, kexec image verification and firmware blob
1889 verification.
1890
1891 config PROFILING
1892 bool "Profiling support"
1893 help
1894 Say Y here to enable the extended profiling support mechanisms used
1895 by profilers.
1896
1897 config RUST
1898 bool "Rust support"
1899 depends on HAVE_RUST
1900 depends on RUST_IS_AVAILABLE
1901 depends on !CFI_CLANG
1902 depends on !MODVERSIONS
1903 depends on !GCC_PLUGINS
1904 depends on !RANDSTRUCT
1905 depends on !SHADOW_CALL_STACK
1906 depends on !DEBUG_INFO_BTF || PAHOLE_HAS_LANG_EXCLUDE
1907 help
1908 Enables Rust support in the kernel.
1909
1910 This allows other Rust-related options, like drivers written in Rust,
1911 to be selected.
1912
1913 It is also required to be able to load external kernel modules
1914 written in Rust.
1915
1916 See Documentation/rust/ for more information.
1917
1918 If unsure, say N.
1919
1920 config RUSTC_VERSION_TEXT
1921 string
1922 depends on RUST
1923 default "$(shell,$(RUSTC) --version 2>/dev/null)"
1924
1925 config BINDGEN_VERSION_TEXT
1926 string
1927 depends on RUST
1928 # The dummy parameter `workaround-for-0.69.0` is required to support 0.69.0
1929 # (https://github.com/rust-lang/rust-bindgen/pull/2678). It can be removed when
1930 # the minimum version is upgraded past that (0.69.1 already fixed the issue).
1931 default "$(shell,$(BINDGEN) --version workaround-for-0.69.0 2>/dev/null)"
1932
1933 #
1934 # Place an empty function call at each tracepoint site. Can be
1935 # dynamically changed for a probe function.
1936 #
1937 config TRACEPOINTS
1938 bool
1939
1940 source "kernel/Kconfig.kexec"
1941
1942 endmenu # General setup
1943
1944 source "arch/Kconfig"
1945
1946 config RT_MUTEXES
1947 bool
1948 default y if PREEMPT_RT
1949
1950 config MODULE_SIG_FORMAT
1951 def_bool n
1952 select SYSTEM_DATA_VERIFICATION
1953
1954 source "kernel/module/Kconfig"
1955
1956 config INIT_ALL_POSSIBLE
1957 bool
1958 help
1959 Back when each arch used to define their own cpu_online_mask and
1960 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
1961 with all 1s, and others with all 0s. When they were centralised,
1962 it was better to provide this option than to break all the archs
1963 and have several arch maintainers pursuing me down dark alleys.
1964
1965 source "block/Kconfig"
1966
1967 config PREEMPT_NOTIFIERS
1968 bool
1969
1970 config PADATA
1971 depends on SMP
1972 bool
1973
1974 config ASN1
1975 tristate
1976 help
1977 Build a simple ASN.1 grammar compiler that produces a bytecode output
1978 that can be interpreted by the ASN.1 stream decoder and used to
1979 inform it as to what tags are to be expected in a stream and what
1980 functions to call on what tags.
1981
1982 source "kernel/Kconfig.locks"
1983
1984 config ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE
1985 bool
1986
1987 config ARCH_HAS_PREPARE_SYNC_CORE_CMD
1988 bool
1989
1990 config ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
1991 bool
1992
1993 # It may be useful for an architecture to override the definitions of the
1994 # SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h>
1995 # and the COMPAT_ variants in <linux/compat.h>, in particular to use a
1996 # different calling convention for syscalls. They can also override the
1997 # macros for not-implemented syscalls in kernel/sys_ni.c and
1998 # kernel/time/posix-stubs.c. All these overrides need to be available in
1999 # <asm/syscall_wrapper.h>.
2000 config ARCH_HAS_SYSCALL_WRAPPER
2001 def_bool n
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.