blob: e2e99544da8d5bf2518e1a9d942f1c1257ed4d1a [file] [log] [blame]
Thomas Gleixnerec8f24b2019-05-19 13:07:45 +01001# SPDX-License-Identifier: GPL-2.0-only
Roman Zippelface4372006-06-08 22:12:45 -07002config DEFCONFIG_LIST
3 string
Paolo 'Blaisorblade' Giarrussob2670eac2006-10-19 23:28:23 -07004 depends on !UML
Roman Zippelface4372006-06-08 22:12:45 -07005 option defconfig_list
Rob Landley47f38ae2018-08-08 13:06:43 +09006 default "/lib/modules/$(shell,uname -r)/.config"
Roman Zippelface4372006-06-08 22:12:45 -07007 default "/etc/kernel-config"
Rob Landley47f38ae2018-08-08 13:06:43 +09008 default "/boot/config-$(shell,uname -r)"
Masahiro Yamada104daea2018-05-28 18:21:40 +09009 default ARCH_DEFCONFIG
10 default "arch/$(ARCH)/defconfig"
Roman Zippelface4372006-06-08 22:12:45 -070011
Masahiro Yamadaa4353892018-05-28 18:22:01 +090012config CC_IS_GCC
13 def_bool $(success,$(CC) --version | head -n 1 | grep -q gcc)
14
15config GCC_VERSION
16 int
Masahiro Yamadafa7295a2019-03-01 16:10:22 +090017 default $(shell,$(srctree)/scripts/gcc-version.sh $(CC)) if CC_IS_GCC
Masahiro Yamadaa4353892018-05-28 18:22:01 +090018 default 0
19
Masahiro Yamada469cb732018-05-28 18:22:02 +090020config CC_IS_CLANG
21 def_bool $(success,$(CC) --version | head -n 1 | grep -q clang)
22
23config CLANG_VERSION
24 int
25 default $(shell,$(srctree)/scripts/clang-version.sh $(CC))
26
Masahiro Yamada1a927fd2019-07-01 09:58:39 +090027config CC_CAN_LINK
28 def_bool $(success,$(srctree)/scripts/cc-can-link.sh $(CC))
29
Masahiro Yamadae9666d12018-12-31 00:14:15 +090030config CC_HAS_ASM_GOTO
31 def_bool $(success,$(srctree)/scripts/gcc-goto.sh $(CC))
32
Masahiro Yamadab303c6d2019-02-21 13:13:38 +090033config CC_HAS_WARN_MAYBE_UNINITIALIZED
34 def_bool $(cc-option,-Wmaybe-uninitialized)
35 help
36 GCC >= 4.7 supports this option.
37
38config CC_DISABLE_WARN_MAYBE_UNINITIALIZED
39 bool
40 depends on CC_HAS_WARN_MAYBE_UNINITIALIZED
41 default CC_IS_GCC && GCC_VERSION < 40900 # unreliable for GCC < 4.9
42 help
43 GCC's -Wmaybe-uninitialized is not reliable by definition.
44 Lots of false positive warnings are produced in some cases.
45
46 If this option is enabled, -Wno-maybe-uninitialzed is passed
47 to the compiler to suppress maybe-uninitialized warnings.
48
Peter Oberparleiterb99b87f2009-06-17 16:28:03 -070049config CONSTRUCTORS
50 bool
51 depends on !UML
Peter Oberparleiterb99b87f2009-06-17 16:28:03 -070052
Peter Zijlstrae360adb2010-10-14 14:01:34 +080053config IRQ_WORK
54 bool
Peter Zijlstrae360adb2010-10-14 14:01:34 +080055
David Daney1dbdc6f2012-04-19 14:59:57 -070056config BUILDTIME_EXTABLE_SORT
57 bool
58
Andy Lutomirskic65eacb2016-09-13 14:29:24 -070059config THREAD_INFO_IN_TASK
60 bool
61 help
62 Select this to move thread_info off the stack into task_struct. To
63 make this work, an arch will need to remove all thread_info fields
64 except flags and fix any runtime bugs.
65
Andy Lutomirskic6c314a2016-09-15 22:45:43 -070066 One subtle change that will be needed is to use try_get_task_stack()
67 and put_task_stack() in save_thread_stack_tsk() and get_wchan().
68
Al Boldiff0cfc62007-07-31 00:39:23 -070069menu "General setup"
Linus Torvalds1da177e2005-04-16 15:20:36 -070070
Linus Torvalds1da177e2005-04-16 15:20:36 -070071config BROKEN
72 bool
Linus Torvalds1da177e2005-04-16 15:20:36 -070073
74config BROKEN_ON_SMP
75 bool
76 depends on BROKEN || !SMP
77 default y
78
Linus Torvalds1da177e2005-04-16 15:20:36 -070079config INIT_ENV_ARG_LIMIT
80 int
Adrian Bunkdd673bc2006-06-30 01:55:51 -070081 default 32 if !UML
82 default 128 if UML
Linus Torvalds1da177e2005-04-16 15:20:36 -070083 help
Randy Dunlap34ad92c2005-10-30 15:01:46 -080084 Maximum of each of the number of arguments and environment
85 variables passed to init from the kernel command line.
Linus Torvalds1da177e2005-04-16 15:20:36 -070086
Jiri Slaby4bb16672013-05-22 10:56:24 +020087config COMPILE_TEST
88 bool "Compile also drivers which will not load"
Richard Weinbergerbc083a62016-08-02 14:03:27 -070089 depends on !UML
Jiri Slaby4bb16672013-05-22 10:56:24 +020090 default n
91 help
92 Some drivers can be compiled on a different platform than they are
93 intended to be run on. Despite they cannot be loaded there (or even
94 when they load they cannot be used due to missing HW support),
95 developers still, opposing to distributors, might want to build such
96 drivers to compile-test them.
97
98 If you are a developer and want to build everything available, say Y
99 here. If you are a user/distributor, say N here to exclude useless
100 drivers to be distributed.
101
Jani Nikulae846f0d2019-06-04 15:42:48 +0300102config HEADER_TEST
103 bool "Compile test headers that should be standalone compilable"
104 help
105 Compile test headers listed in header-test-y target to ensure they are
106 self-contained, i.e. compilable as standalone units.
107
108 If you are a developer or tester and want to ensure the requested
109 headers are self-contained, say Y here. Otherwise, choose N.
110
Masahiro Yamada43c78d82019-07-01 09:58:45 +0900111config KERNEL_HEADER_TEST
112 bool "Compile test kernel headers"
113 depends on HEADER_TEST
114 help
115 Headers in include/ are used to build external moduls.
116 Compile test them to ensure they are self-contained, i.e.
117 compilable as standalone units.
118
119 If you are a developer or tester and want to ensure the headers
120 in include/ are self-contained, say Y here. Otherwise, choose N.
121
Masahiro Yamadad6fc9fc2019-07-01 09:58:40 +0900122config UAPI_HEADER_TEST
123 bool "Compile test UAPI headers"
124 depends on HEADER_TEST && HEADERS_INSTALL && CC_CAN_LINK
125 help
126 Compile test headers exported to user-space to ensure they are
127 self-contained, i.e. compilable as standalone units.
128
129 If you are a developer or tester and want to ensure the exported
130 headers are self-contained, say Y here. Otherwise, choose N.
131
Linus Torvalds1da177e2005-04-16 15:20:36 -0700132config LOCALVERSION
133 string "Local version - append to kernel release"
134 help
135 Append an extra string to the end of your kernel version.
136 This will show up when you type uname, for example.
137 The string you set here will be appended after the contents of
138 any files with a filename matching localversion* in your
139 object and source tree, in that order. Your total string can
140 be a maximum of 64 characters.
141
Ryan Andersonaaebf432005-07-31 04:57:49 -0400142config LOCALVERSION_AUTO
143 bool "Automatically append version information to the version string"
144 default y
Alexey Dobriyanac3339b2016-08-02 14:07:21 -0700145 depends on !COMPILE_TEST
Ryan Andersonaaebf432005-07-31 04:57:49 -0400146 help
147 This will try to automatically determine if the current tree is a
Robert P. J. Day6e5a5422007-05-01 23:08:11 +0200148 release tree by looking for git tags that belong to the current
149 top of tree revision.
Ryan Andersonaaebf432005-07-31 04:57:49 -0400150
151 A string of the format -gxxxxxxxx will be added to the localversion
Robert P. J. Day6e5a5422007-05-01 23:08:11 +0200152 if a git-based tree is found. The string generated by this will be
Ryan Andersonaaebf432005-07-31 04:57:49 -0400153 appended after any matching localversion* files, and after the value
Robert P. J. Day6e5a5422007-05-01 23:08:11 +0200154 set in CONFIG_LOCALVERSION.
Ryan Andersonaaebf432005-07-31 04:57:49 -0400155
Robert P. J. Day6e5a5422007-05-01 23:08:11 +0200156 (The actual string used here is the first eight characters produced
157 by running the command:
158
159 $ git rev-parse --verify HEAD
160
161 which is done within the script "scripts/setlocalversion".)
Ryan Andersonaaebf432005-07-31 04:57:49 -0400162
Laura Abbott9afb7192018-07-05 17:49:37 -0700163config BUILD_SALT
164 string "Build ID Salt"
165 default ""
166 help
167 The build ID is used to link binaries and their debug info. Setting
168 this option will use the value in the calculation of the build id.
169 This is mostly useful for distributions which want to ensure the
170 build is unique between builds. It's safe to leave the default.
171
H. Peter Anvin2e9f3bd2009-01-04 15:41:25 -0800172config HAVE_KERNEL_GZIP
173 bool
174
175config HAVE_KERNEL_BZIP2
176 bool
177
178config HAVE_KERNEL_LZMA
179 bool
180
Lasse Collin3ebe1242011-01-12 17:01:23 -0800181config HAVE_KERNEL_XZ
182 bool
183
Albin Tonnerre7dd65fe2010-01-08 14:42:42 -0800184config HAVE_KERNEL_LZO
185 bool
186
Kyungsik Leee76e1fd2013-07-08 16:01:46 -0700187config HAVE_KERNEL_LZ4
188 bool
189
Vasily Gorbikf16466a2018-06-12 21:26:35 +0200190config HAVE_KERNEL_UNCOMPRESSED
191 bool
192
Alain Knaff30d65db2009-01-04 22:46:17 +0100193choice
H. Peter Anvin2e9f3bd2009-01-04 15:41:25 -0800194 prompt "Kernel compression mode"
195 default KERNEL_GZIP
Vasily Gorbikf16466a2018-06-12 21:26:35 +0200196 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_UNCOMPRESSED
H. Peter Anvin2e9f3bd2009-01-04 15:41:25 -0800197 help
Alain Knaff30d65db2009-01-04 22:46:17 +0100198 The linux kernel is a kind of self-extracting executable.
199 Several compression algorithms are available, which differ
200 in efficiency, compression and decompression speed.
201 Compression speed is only relevant when building a kernel.
202 Decompression speed is relevant at each boot.
203
204 If you have any problems with bzip2 or lzma compressed
205 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
206 version of this functionality (bzip2 only), for 2.4, was
207 supplied by Christian Ludwig)
208
209 High compression options are mostly useful for users, who
210 are low on disk space (embedded systems), but for whom ram
211 size matters less.
212
213 If in doubt, select 'gzip'
214
215config KERNEL_GZIP
H. Peter Anvin2e9f3bd2009-01-04 15:41:25 -0800216 bool "Gzip"
217 depends on HAVE_KERNEL_GZIP
218 help
Albin Tonnerre7dd65fe2010-01-08 14:42:42 -0800219 The old and tried gzip compression. It provides a good balance
220 between compression ratio and decompression speed.
Alain Knaff30d65db2009-01-04 22:46:17 +0100221
222config KERNEL_BZIP2
223 bool "Bzip2"
H. Peter Anvin2e9f3bd2009-01-04 15:41:25 -0800224 depends on HAVE_KERNEL_BZIP2
Alain Knaff30d65db2009-01-04 22:46:17 +0100225 help
226 Its compression ratio and speed is intermediate.
Randy Dunlap0a4dd352012-05-31 16:26:46 -0700227 Decompression speed is slowest among the choices. The kernel
H. Peter Anvin2e9f3bd2009-01-04 15:41:25 -0800228 size is about 10% smaller with bzip2, in comparison to gzip.
229 Bzip2 uses a large amount of memory. For modern kernels you
230 will need at least 8MB RAM or more for booting.
Alain Knaff30d65db2009-01-04 22:46:17 +0100231
232config KERNEL_LZMA
H. Peter Anvin2e9f3bd2009-01-04 15:41:25 -0800233 bool "LZMA"
234 depends on HAVE_KERNEL_LZMA
235 help
Randy Dunlap0a4dd352012-05-31 16:26:46 -0700236 This compression algorithm's ratio is best. Decompression speed
237 is between gzip and bzip2. Compression is slowest.
238 The kernel size is about 33% smaller with LZMA in comparison to gzip.
Alain Knaff30d65db2009-01-04 22:46:17 +0100239
Lasse Collin3ebe1242011-01-12 17:01:23 -0800240config KERNEL_XZ
241 bool "XZ"
242 depends on HAVE_KERNEL_XZ
243 help
244 XZ uses the LZMA2 algorithm and instruction set specific
245 BCJ filters which can improve compression ratio of executable
246 code. The size of the kernel is about 30% smaller with XZ in
247 comparison to gzip. On architectures for which there is a BCJ
248 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
249 will create a few percent smaller kernel than plain LZMA.
250
251 The speed is about the same as with LZMA: The decompression
252 speed of XZ is better than that of bzip2 but worse than gzip
253 and LZO. Compression is slow.
254
Albin Tonnerre7dd65fe2010-01-08 14:42:42 -0800255config KERNEL_LZO
256 bool "LZO"
257 depends on HAVE_KERNEL_LZO
258 help
Randy Dunlap0a4dd352012-05-31 16:26:46 -0700259 Its compression ratio is the poorest among the choices. The kernel
Stephan Sperber681b3042010-07-14 11:23:08 +0200260 size is about 10% bigger than gzip; however its speed
Albin Tonnerre7dd65fe2010-01-08 14:42:42 -0800261 (both compression and decompression) is the fastest.
262
Kyungsik Leee76e1fd2013-07-08 16:01:46 -0700263config KERNEL_LZ4
264 bool "LZ4"
265 depends on HAVE_KERNEL_LZ4
266 help
267 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
268 A preliminary version of LZ4 de/compression tool is available at
269 <https://code.google.com/p/lz4/>.
270
271 Its compression ratio is worse than LZO. The size of the kernel
272 is about 8% bigger than LZO. But the decompression speed is
273 faster than LZO.
274
Vasily Gorbikf16466a2018-06-12 21:26:35 +0200275config KERNEL_UNCOMPRESSED
276 bool "None"
277 depends on HAVE_KERNEL_UNCOMPRESSED
278 help
279 Produce uncompressed kernel image. This option is usually not what
280 you want. It is useful for debugging the kernel in slow simulation
281 environments, where decompressing and moving the kernel is awfully
282 slow. This option allows early boot code to skip the decompressor
283 and jump right at uncompressed kernel image.
284
Alain Knaff30d65db2009-01-04 22:46:17 +0100285endchoice
286
Josh Triplettbd5dc172011-06-15 15:08:28 -0700287config DEFAULT_HOSTNAME
288 string "Default hostname"
289 default "(none)"
290 help
291 This option determines the default system hostname before userspace
292 calls sethostname(2). The kernel traditionally uses "(none)" here,
293 but you may wish to use a different default here to make a minimal
294 system more usable with less configuration.
295
Christoph Hellwig17c46a62018-07-31 13:39:29 +0200296#
297# For some reason microblaze and nios2 hard code SWAP=n. Hopefully we can
298# add proper SWAP support to them, in which case this can be remove.
299#
300config ARCH_NO_SWAP
301 bool
302
Linus Torvalds1da177e2005-04-16 15:20:36 -0700303config SWAP
304 bool "Support for paging of anonymous memory (swap)"
Christoph Hellwig17c46a62018-07-31 13:39:29 +0200305 depends on MMU && BLOCK && !ARCH_NO_SWAP
Linus Torvalds1da177e2005-04-16 15:20:36 -0700306 default y
307 help
308 This option allows you to choose whether you want to have support
Jesper Juhl92c35042006-01-15 02:40:08 +0100309 for so called swap devices or swap files in your kernel that are
Linus Torvalds1da177e2005-04-16 15:20:36 -0700310 used to provide more virtual memory than the actual RAM present
311 in your computer. If unsure say Y.
312
313config SYSVIPC
314 bool "System V IPC"
Linus Torvalds1da177e2005-04-16 15:20:36 -0700315 ---help---
316 Inter Process Communication is a suite of library functions and
317 system calls which let processes (running programs) synchronize and
318 exchange information. It is generally considered to be a good thing,
319 and some programs won't run unless you say Y here. In particular, if
320 you want to run the DOS emulator dosemu under Linux (read the
321 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
322 you'll need to say Y here.
323
324 You can find documentation about IPC with "info ipc" and also in
325 section 6.4 of the Linux Programmer's Guide, available from
326 <http://www.tldp.org/guides.html>.
327
Eric W. Biedermana5494dc2007-02-14 00:34:06 -0800328config SYSVIPC_SYSCTL
329 bool
330 depends on SYSVIPC
331 depends on SYSCTL
332 default y
333
Linus Torvalds1da177e2005-04-16 15:20:36 -0700334config POSIX_MQUEUE
335 bool "POSIX Message Queues"
Kees Cook19c92392012-10-02 11:19:29 -0700336 depends on NET
Linus Torvalds1da177e2005-04-16 15:20:36 -0700337 ---help---
338 POSIX variant of message queues is a part of IPC. In POSIX message
339 queues every message has a priority which decides about succession
340 of receiving it by a process. If you want to compile and run
341 programs written e.g. for Solaris with use of its POSIX message
Robert P. J. Dayb0e37652007-05-09 07:25:13 +0200342 queues (functions mq_*) say Y here.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700343
344 POSIX message queues are visible as a filesystem called 'mqueue'
345 and can be mounted somewhere if you want to do filesystem
346 operations on message queues.
347
348 If unsure, say Y.
349
Serge E. Hallynbdc8e5f2009-04-06 19:01:11 -0700350config POSIX_MQUEUE_SYSCTL
351 bool
352 depends on POSIX_MQUEUE
353 depends on SYSCTL
354 default y
355
Konstantin Khlebnikov226b4cc2014-06-04 16:10:50 -0700356config CROSS_MEMORY_ATTACH
357 bool "Enable process_vm_readv/writev syscalls"
358 depends on MMU
359 default y
360 help
361 Enabling this option adds the system calls process_vm_readv and
362 process_vm_writev which allow a process with the correct privileges
Geert Uytterhoevena2a368d2014-08-12 13:46:11 -0700363 to directly read from or write to another process' address space.
Konstantin Khlebnikov226b4cc2014-06-04 16:10:50 -0700364 See the man page for more details.
365
Josh Triplett69369a72014-04-03 14:48:27 -0700366config USELIB
367 bool "uselib syscall"
Riku Voipiob2113a42016-01-15 16:58:13 -0800368 def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION
Josh Triplett69369a72014-04-03 14:48:27 -0700369 help
370 This option enables the uselib syscall, a system call used in the
371 dynamic linker from libc5 and earlier. glibc does not use this
372 system call. If you intend to run programs built on libc5 or
373 earlier, you may need to enable this syscall. Current systems
374 running glibc can safely disable this.
375
Linus Torvalds1da177e2005-04-16 15:20:36 -0700376config AUDIT
377 bool "Auditing support"
Chris Wright804a6a492005-05-11 10:52:45 +0100378 depends on NET
Linus Torvalds1da177e2005-04-16 15:20:36 -0700379 help
380 Enable auditing infrastructure that can be used with another
381 kernel subsystem, such as SELinux (which requires this for
Paul Moorecb74ed22016-01-13 09:18:55 -0500382 logging of avc messages output). System call auditing is included
383 on architectures which support it.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700384
AKASHI Takahiro7a017722014-02-25 18:16:24 +0900385config HAVE_ARCH_AUDITSYSCALL
386 bool
387
Linus Torvalds1da177e2005-04-16 15:20:36 -0700388config AUDITSYSCALL
Paul Moorecb74ed22016-01-13 09:18:55 -0500389 def_bool y
AKASHI Takahiro7a017722014-02-25 18:16:24 +0900390 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL
Eric Paris28a3a7e2009-12-17 20:12:05 -0500391 select FSNOTIFY
Al Viro74c3cbe2007-07-22 08:04:18 -0400392
Thomas Gleixnerd9817eb2010-09-27 12:45:59 +0000393source "kernel/irq/Kconfig"
Thomas Gleixner764e0da2012-05-21 23:16:18 +0200394source "kernel/time/Kconfig"
Christoph Hellwig87a4c372018-07-31 13:39:32 +0200395source "kernel/Kconfig.preempt"
Thomas Gleixnerd9817eb2010-09-27 12:45:59 +0000396
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200397menu "CPU/Task time and stats accounting"
398
Frederic Weisbeckerabf917c2012-07-25 07:56:04 +0200399config VIRT_CPU_ACCOUNTING
400 bool
401
Frederic Weisbeckerfdf9c352012-09-09 14:56:31 +0200402choice
403 prompt "Cputime accounting"
404 default TICK_CPU_ACCOUNTING if !PPC64
Stephen Rothwell02fc8d32013-02-08 14:19:38 +1100405 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64
Frederic Weisbeckerfdf9c352012-09-09 14:56:31 +0200406
407# Kind of a stub config for the pure tick based cputime accounting
408config TICK_CPU_ACCOUNTING
409 bool "Simple tick based cputime accounting"
Frederic Weisbeckerc58b0df2013-04-26 15:16:31 +0200410 depends on !S390 && !NO_HZ_FULL
Frederic Weisbeckerfdf9c352012-09-09 14:56:31 +0200411 help
412 This is the basic tick based cputime accounting that maintains
413 statistics about user, system and idle time spent on per jiffies
414 granularity.
415
416 If unsure, say Y.
417
Frederic Weisbeckerabf917c2012-07-25 07:56:04 +0200418config VIRT_CPU_ACCOUNTING_NATIVE
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200419 bool "Deterministic task and CPU time accounting"
Frederic Weisbeckerc58b0df2013-04-26 15:16:31 +0200420 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL
Frederic Weisbeckerabf917c2012-07-25 07:56:04 +0200421 select VIRT_CPU_ACCOUNTING
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200422 help
423 Select this option to enable more accurate task and CPU time
424 accounting. This is done by reading a CPU counter on each
425 kernel entry and exit and on transitions within the kernel
426 between system, softirq and hardirq state, so there is a
427 small performance impact. In the case of s390 or IBM POWER > 5,
428 this also enables accounting of stolen time on logically-partitioned
429 systems.
430
Frederic Weisbeckerabf917c2012-07-25 07:56:04 +0200431config VIRT_CPU_ACCOUNTING_GEN
432 bool "Full dynticks CPU time accounting"
Kevin Hilmanff3fb252013-09-16 15:28:19 -0700433 depends on HAVE_CONTEXT_TRACKING
Kevin Hilman554b0002013-09-16 15:28:21 -0700434 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN
Arnd Bergmann041a1572019-03-04 21:01:31 +0100435 depends on GENERIC_CLOCKEVENTS
Frederic Weisbeckerabf917c2012-07-25 07:56:04 +0200436 select VIRT_CPU_ACCOUNTING
437 select CONTEXT_TRACKING
438 help
439 Select this option to enable task and CPU time accounting on full
440 dynticks systems. This accounting is implemented by watching every
441 kernel-user boundaries using the context tracking subsystem.
442 The accounting is thus performed at the expense of some significant
443 overhead.
444
445 For now this is only useful if you are working on the full
446 dynticks subsystem development.
447
448 If unsure, say N.
449
Rik van Rielb58c3582016-07-13 16:50:02 +0200450endchoice
451
Frederic Weisbeckerfdf9c352012-09-09 14:56:31 +0200452config IRQ_TIME_ACCOUNTING
453 bool "Fine granularity task level IRQ time accounting"
Rik van Rielb58c3582016-07-13 16:50:02 +0200454 depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE
Frederic Weisbeckerfdf9c352012-09-09 14:56:31 +0200455 help
456 Select this option to enable fine granularity task irq time
457 accounting. This is done by reading a timestamp on each
458 transitions between softirq and hardirq state, so there can be a
459 small performance impact.
460
461 If in doubt, say N here.
462
Vincent Guittot11d4afd2018-09-25 11:17:42 +0200463config HAVE_SCHED_AVG_IRQ
464 def_bool y
465 depends on IRQ_TIME_ACCOUNTING || PARAVIRT_TIME_ACCOUNTING
466 depends on SMP
467
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200468config BSD_PROCESS_ACCT
469 bool "BSD Process Accounting"
Iulia Manda28138932015-04-15 16:16:41 -0700470 depends on MULTIUSER
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200471 help
472 If you say Y here, a user level program will be able to instruct the
473 kernel (via a special system call) to write process accounting
474 information to a file: whenever a process exits, information about
475 that process will be appended to the file by the kernel. The
476 information includes things such as creation time, owning user,
477 command name, memory usage, controlling terminal etc. (the complete
478 list is in the struct acct in <file:include/linux/acct.h>). It is
479 up to the user level program to do useful things with this
480 information. This is generally a good idea, so say Y.
481
482config BSD_PROCESS_ACCT_V3
483 bool "BSD Process Accounting version 3 file format"
484 depends on BSD_PROCESS_ACCT
485 default n
486 help
487 If you say Y here, the process accounting information is written
488 in a new file format that also logs the process IDs of each
Randy Dunlap3903bf92018-08-21 21:58:34 -0700489 process and its parent. Note that this file format is incompatible
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200490 with previous v0/v1/v2 file formats, so you will need updated tools
491 for processing it. A preliminary version of these tools is available
492 at <http://www.gnu.org/software/acct/>.
493
494config TASKSTATS
Kees Cook19c92392012-10-02 11:19:29 -0700495 bool "Export task/process statistics through netlink"
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200496 depends on NET
Iulia Manda28138932015-04-15 16:16:41 -0700497 depends on MULTIUSER
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200498 default n
499 help
500 Export selected statistics for tasks/processes through the
501 generic netlink interface. Unlike BSD process accounting, the
502 statistics are available during the lifetime of tasks/processes as
503 responses to commands. Like BSD accounting, they are sent to user
504 space on task exit.
505
506 Say N if unsure.
507
508config TASK_DELAY_ACCT
Kees Cook19c92392012-10-02 11:19:29 -0700509 bool "Enable per-task delay accounting"
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200510 depends on TASKSTATS
Naveen N. Raof6db8342015-06-25 23:53:37 +0530511 select SCHED_INFO
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200512 help
513 Collect information on time spent by a task waiting for system
514 resources like cpu, synchronous block I/O completion and swapping
515 in pages. Such statistics can help in setting a task's priorities
516 relative to other tasks for cpu, io, rss limits etc.
517
518 Say N if unsure.
519
520config TASK_XACCT
Kees Cook19c92392012-10-02 11:19:29 -0700521 bool "Enable extended accounting over taskstats"
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200522 depends on TASKSTATS
523 help
524 Collect extended task accounting data and send the data
525 to userland for processing over the taskstats interface.
526
527 Say N if unsure.
528
529config TASK_IO_ACCOUNTING
Kees Cook19c92392012-10-02 11:19:29 -0700530 bool "Enable per-task storage I/O accounting"
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200531 depends on TASK_XACCT
532 help
533 Collect information on the number of bytes of storage I/O which this
534 task has caused.
535
536 Say N if unsure.
537
Johannes Weinereb414682018-10-26 15:06:27 -0700538config PSI
539 bool "Pressure stall information tracking"
540 help
541 Collect metrics that indicate how overcommitted the CPU, memory,
542 and IO capacity are in the system.
543
544 If you say Y here, the kernel will create /proc/pressure/ with the
545 pressure statistics files cpu, memory, and io. These will indicate
546 the share of walltime in which some or all tasks in the system are
547 delayed due to contention of the respective resource.
548
Johannes Weiner2ce71352018-10-26 15:06:31 -0700549 In kernels with cgroup support, cgroups (cgroup2 only) will
550 have cpu.pressure, memory.pressure, and io.pressure files,
551 which aggregate pressure stalls for the grouped tasks only.
552
Johannes Weinereb414682018-10-26 15:06:27 -0700553 For more details see Documentation/accounting/psi.txt.
554
555 Say N if unsure.
556
Johannes Weinere0c27442018-11-30 14:09:58 -0800557config PSI_DEFAULT_DISABLED
558 bool "Require boot parameter to enable pressure stall information tracking"
559 default n
560 depends on PSI
561 help
562 If set, pressure stall information tracking will be disabled
Baruch Siach428a1cb2018-12-14 14:17:03 -0800563 per default but can be enabled through passing psi=1 on the
564 kernel commandline during boot.
Johannes Weinere0c27442018-11-30 14:09:58 -0800565
Johannes Weiner7b2489d2019-02-01 14:21:15 -0800566 This feature adds some code to the task wakeup and sleep
567 paths of the scheduler. The overhead is too low to affect
568 common scheduling-intense workloads in practice (such as
569 webservers, memcache), but it does show up in artificial
570 scheduler stress tests, such as hackbench.
571
572 If you are paranoid and not sure what the kernel will be
573 used for, say Y.
574
575 Say N if unsure.
576
Frederic Weisbecker391dc692012-09-09 14:22:07 +0200577endmenu # "CPU/Task time and stats accounting"
578
Frederic Weisbecker5c4991e2017-10-27 04:42:34 +0200579config CPU_ISOLATION
580 bool "CPU isolation"
Geert Uytterhoeven414a2dc2018-01-02 12:13:10 +0100581 depends on SMP || COMPILE_TEST
Frederic Weisbecker2c438382017-12-14 19:18:26 +0100582 default y
Frederic Weisbecker5c4991e2017-10-27 04:42:34 +0200583 help
584 Make sure that CPUs running critical tasks are not disturbed by
585 any source of "noise" such as unbound workqueues, timers, kthreads...
Frederic Weisbecker2c438382017-12-14 19:18:26 +0100586 Unbound jobs get offloaded to housekeeping CPUs. This is driven by
587 the "isolcpus=" boot parameter.
588
589 Say Y if unsure.
Frederic Weisbecker5c4991e2017-10-27 04:42:34 +0200590
Paul E. McKenney0af92d42017-05-17 08:43:40 -0700591source "kernel/rcu/Kconfig"
Mike Travisc903ff82009-01-15 12:28:29 -0800592
Vivek Goyalde5b56b2014-08-08 14:25:41 -0700593config BUILD_BIN2C
594 bool
595 default n
596
Linus Torvalds1da177e2005-04-16 15:20:36 -0700597config IKCONFIG
Ross Birof2443ab2006-09-30 23:27:25 -0700598 tristate "Kernel .config support"
Linus Torvalds1da177e2005-04-16 15:20:36 -0700599 ---help---
600 This option enables the complete Linux kernel ".config" file
601 contents to be saved in the kernel. It provides documentation
602 of which kernel options are used in a running kernel or in an
603 on-disk kernel. This information can be extracted from the kernel
604 image file with the script scripts/extract-ikconfig and used as
605 input to rebuild the current kernel or to build another kernel.
606 It can also be extracted from a running kernel by reading
607 /proc/config.gz if enabled (below).
608
609config IKCONFIG_PROC
610 bool "Enable access to .config through /proc/config.gz"
611 depends on IKCONFIG && PROC_FS
612 ---help---
613 This option enables access to the kernel configuration file
614 through /proc/config.gz.
615
Joel Fernandes (Google)f7b101d2019-05-15 17:35:51 -0400616config IKHEADERS
617 tristate "Enable kernel headers through /sys/kernel/kheaders.tar.xz"
618 depends on SYSFS
Joel Fernandes (Google)43d8ce92019-04-26 15:04:29 -0400619 help
Joel Fernandes (Google)f7b101d2019-05-15 17:35:51 -0400620 This option enables access to the in-kernel headers that are generated during
621 the build process. These can be used to build eBPF tracing programs,
622 or similar programs. If you build the headers as a module, a module called
623 kheaders.ko is built which can be loaded on-demand to get access to headers.
Joel Fernandes (Google)43d8ce92019-04-26 15:04:29 -0400624
Alistair John Strachan794543a2007-05-08 00:31:15 -0700625config LOG_BUF_SHIFT
626 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
Ingo Molnarfb39f982015-07-01 10:19:11 +0200627 range 12 25
Adrian Bunkf17a32e2008-04-29 00:58:58 -0700628 default 17
Josh Triplett361e9df2014-10-03 16:00:54 -0700629 depends on PRINTK
Alistair John Strachan794543a2007-05-08 00:31:15 -0700630 help
Luis R. Rodriguez23b28992014-08-06 16:08:56 -0700631 Select the minimal kernel log buffer size as a power of 2.
632 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config
633 parameter, see below. Any higher size also might be forced
634 by "log_buf_len" boot parameter.
635
Adrian Bunkf17a32e2008-04-29 00:58:58 -0700636 Examples:
Luis R. Rodriguez23b28992014-08-06 16:08:56 -0700637 17 => 128 KB
Adrian Bunkf17a32e2008-04-29 00:58:58 -0700638 16 => 64 KB
Luis R. Rodriguez23b28992014-08-06 16:08:56 -0700639 15 => 32 KB
640 14 => 16 KB
Alistair John Strachan794543a2007-05-08 00:31:15 -0700641 13 => 8 KB
642 12 => 4 KB
643
Luis R. Rodriguez23b28992014-08-06 16:08:56 -0700644config LOG_CPU_MAX_BUF_SHIFT
645 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)"
Geert Uytterhoeven2240a312014-10-13 15:51:11 -0700646 depends on SMP
Luis R. Rodriguez23b28992014-08-06 16:08:56 -0700647 range 0 21
648 default 12 if !BASE_SMALL
649 default 0 if BASE_SMALL
Josh Triplett361e9df2014-10-03 16:00:54 -0700650 depends on PRINTK
Luis R. Rodriguez23b28992014-08-06 16:08:56 -0700651 help
652 This option allows to increase the default ring buffer size
653 according to the number of CPUs. The value defines the contribution
654 of each CPU as a power of 2. The used space is typically only few
655 lines however it might be much more when problems are reported,
656 e.g. backtraces.
657
658 The increased size means that a new buffer has to be allocated and
659 the original static one is unused. It makes sense only on systems
660 with more CPUs. Therefore this value is used only when the sum of
661 contributions is greater than the half of the default kernel ring
662 buffer as defined by LOG_BUF_SHIFT. The default values are set
663 so that more than 64 CPUs are needed to trigger the allocation.
664
665 Also this option is ignored when "log_buf_len" kernel parameter is
666 used as it forces an exact (power of two) size of the ring buffer.
667
668 The number of possible CPUs is used for this computation ignoring
Geert Uytterhoeven5e0d8d52016-06-05 10:47:02 +0200669 hotplugging making the computation optimal for the worst case
670 scenario while allowing a simple algorithm to be used from bootup.
Luis R. Rodriguez23b28992014-08-06 16:08:56 -0700671
672 Examples shift values and their meaning:
673 17 => 128 KB for each CPU
674 16 => 64 KB for each CPU
675 15 => 32 KB for each CPU
676 14 => 16 KB for each CPU
677 13 => 8 KB for each CPU
678 12 => 4 KB for each CPU
679
Sergey Senozhatskyf92bac32016-12-27 23:16:05 +0900680config PRINTK_SAFE_LOG_BUF_SHIFT
681 int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)"
Petr Mladek427934b2016-05-20 17:00:39 -0700682 range 10 21
683 default 13
Sergey Senozhatskyf92bac32016-12-27 23:16:05 +0900684 depends on PRINTK
Petr Mladek427934b2016-05-20 17:00:39 -0700685 help
Sergey Senozhatskyf92bac32016-12-27 23:16:05 +0900686 Select the size of an alternate printk per-CPU buffer where messages
687 printed from usafe contexts are temporary stored. One example would
688 be NMI messages, another one - printk recursion. The messages are
689 copied to the main log buffer in a safe context to avoid a deadlock.
690 The value defines the size as a power of 2.
Petr Mladek427934b2016-05-20 17:00:39 -0700691
Sergey Senozhatskyf92bac32016-12-27 23:16:05 +0900692 Those messages are rare and limited. The largest one is when
Petr Mladek427934b2016-05-20 17:00:39 -0700693 a backtrace is printed. It usually fits into 4KB. Select
694 8KB if you want to be on the safe side.
695
696 Examples:
697 17 => 128 KB for each CPU
698 16 => 64 KB for each CPU
699 15 => 32 KB for each CPU
700 14 => 16 KB for each CPU
701 13 => 8 KB for each CPU
702 12 => 4 KB for each CPU
703
KAMEZAWA Hiroyuki5cdc38f2009-01-07 18:07:30 -0800704#
705# Architectures with an unreliable sched_clock() should select this:
706#
707config HAVE_UNSTABLE_SCHED_CLOCK
708 bool
709
Stephen Boyd38ff87f2013-06-01 23:39:40 -0700710config GENERIC_SCHED_CLOCK
711 bool
712
Andrea Arcangelibe3a7282012-10-04 01:50:47 +0200713#
714# For architectures that want to enable the support for NUMA-affine scheduler
715# balancing logic:
716#
717config ARCH_SUPPORTS_NUMA_BALANCING
718 bool
719
Peter Zijlstrabe5e6102013-11-18 18:27:06 +0100720#
Mel Gorman72b252a2015-09-04 15:47:32 -0700721# For architectures that prefer to flush all TLBs after a number of pages
722# are unmapped instead of sending one IPI per page to flush. The architecture
723# must provide guarantees on what happens if a clean TLB cache entry is
724# written after the unmap. Details are in mm/rmap.c near the check for
725# should_defer_flush. The architecture should also consider if the full flush
726# and the refill costs are offset by the savings of sending fewer IPIs.
727config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
728 bool
729
730#
Peter Zijlstrabe5e6102013-11-18 18:27:06 +0100731# For architectures that know their GCC __int128 support is sound
732#
733config ARCH_SUPPORTS_INT128
734 bool
735
Andrea Arcangelibe3a7282012-10-04 01:50:47 +0200736# For architectures that (ab)use NUMA to represent different memory regions
737# all cpu-local but of different latencies, such as SuperH.
738#
739config ARCH_WANT_NUMA_VARIABLE_LOCALITY
740 bool
741
Andrea Arcangelibe3a7282012-10-04 01:50:47 +0200742config NUMA_BALANCING
743 bool "Memory placement aware NUMA scheduler"
Andrea Arcangelibe3a7282012-10-04 01:50:47 +0200744 depends on ARCH_SUPPORTS_NUMA_BALANCING
745 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
746 depends on SMP && NUMA && MIGRATION
747 help
748 This option adds support for automatic NUMA aware memory/task placement.
749 The mechanism is quite primitive and is based on migrating memory when
Paul Gortmaker6d56a412013-08-13 11:06:50 -0400750 it has references to the node the task is running on.
Andrea Arcangelibe3a7282012-10-04 01:50:47 +0200751
752 This system will be inactive on UMA systems.
753
Aneesh Kumar K.V6f7c97e2014-12-10 15:43:37 -0800754config NUMA_BALANCING_DEFAULT_ENABLED
755 bool "Automatically enable NUMA aware memory/task placement"
756 default y
757 depends on NUMA_BALANCING
758 help
759 If set, automatic NUMA balancing will be enabled if running on a NUMA
760 machine.
761
Li Zefan23964d22009-01-15 13:50:58 -0800762menuconfig CGROUPS
Christoph Jaeger6341e622014-12-20 15:41:11 -0500763 bool "Control Group support"
Tejun Heo2bd59d42014-02-11 11:52:49 -0500764 select KERNFS
Paul Menageddbcc7e2007-10-18 23:39:30 -0700765 help
Li Zefan23964d22009-01-15 13:50:58 -0800766 This option adds support for grouping sets of processes together, for
KAMEZAWA Hiroyuki5cdc38f2009-01-07 18:07:30 -0800767 use with process control subsystems such as Cpusets, CFS, memory
768 controls or device isolation.
769 See
KAMEZAWA Hiroyuki5cdc38f2009-01-07 18:07:30 -0800770 - Documentation/scheduler/sched-design-CFS.txt (CFS)
seokhoon.yoon9991a9c2016-08-02 14:03:13 -0700771 - Documentation/cgroup-v1/ (features for grouping, isolation
Li Zefan45ce80f2009-01-15 13:50:59 -0800772 and resource control)
Paul Menageddbcc7e2007-10-18 23:39:30 -0700773
774 Say N if unsure.
775
Li Zefan23964d22009-01-15 13:50:58 -0800776if CGROUPS
777
Johannes Weiner3e32cb22014-12-10 15:42:31 -0800778config PAGE_COUNTER
779 bool
780
Andrew Mortonc255a452012-07-31 16:43:02 -0700781config MEMCG
Johannes Weinera0166ec2015-12-17 17:19:56 -0500782 bool "Memory controller"
Johannes Weiner3e32cb22014-12-10 15:42:31 -0800783 select PAGE_COUNTER
Tejun Heo79bd9812013-11-22 18:20:42 -0500784 select EVENTFD
Balbir Singh00f0b822008-03-04 14:28:39 -0800785 help
Johannes Weinera0166ec2015-12-17 17:19:56 -0500786 Provides control over the memory footprint of tasks in a cgroup.
Balbir Singh00f0b822008-03-04 14:28:39 -0800787
Andrew Mortonc255a452012-07-31 16:43:02 -0700788config MEMCG_SWAP
Johannes Weinera0166ec2015-12-17 17:19:56 -0500789 bool "Swap controller"
Andrew Mortonc255a452012-07-31 16:43:02 -0700790 depends on MEMCG && SWAP
KAMEZAWA Hiroyukic0777192009-01-07 18:07:57 -0800791 help
Johannes Weinera0166ec2015-12-17 17:19:56 -0500792 Provides control over the swap space consumed by tasks in a cgroup.
793
Andrew Mortonc255a452012-07-31 16:43:02 -0700794config MEMCG_SWAP_ENABLED
Johannes Weinera0166ec2015-12-17 17:19:56 -0500795 bool "Swap controller enabled by default"
Andrew Mortonc255a452012-07-31 16:43:02 -0700796 depends on MEMCG_SWAP
Michal Hockoa42c3902010-11-24 12:57:08 -0800797 default y
798 help
799 Memory Resource Controller Swap Extension comes with its price in
800 a bigger memory consumption. General purpose distribution kernels
Jim Cromie43d547f2010-12-17 14:32:36 -0700801 which want to enable the feature but keep it disabled by default
Michal Hocko07555ac2013-08-22 16:35:46 -0700802 and let the user enable it by swapaccount=1 boot command line
Michal Hockoa42c3902010-11-24 12:57:08 -0800803 parameter should have this option unselected.
804 For those who want to have the feature enabled by default should
805 select this option (if, for some reason, they need to disable it
WANG Cong00a66d22011-07-25 17:12:12 -0700806 then swapaccount=0 does the trick).
KAMEZAWA Hiroyukic0777192009-01-07 18:07:57 -0800807
Kirill Tkhai84c07d12018-08-17 15:47:25 -0700808config MEMCG_KMEM
809 bool
810 depends on MEMCG && !SLOB
811 default y
812
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500813config BLK_CGROUP
814 bool "IO controller"
815 depends on BLOCK
Aneesh Kumar K.V2bc64a22012-07-31 16:42:12 -0700816 default n
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500817 ---help---
818 Generic block IO controller cgroup interface. This is the common
819 cgroup interface which should be used by various IO controlling
820 policies.
Aneesh Kumar K.V2bc64a22012-07-31 16:42:12 -0700821
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500822 Currently, CFQ IO scheduler uses it to recognize task groups and
823 control disk bandwidth allocation (proportional time slice allocation)
824 to such task groups. It is also used by bio throttling logic in
825 block layer to implement upper limit in IO rates on a device.
Stephane Eraniane5d13672011-02-14 11:20:01 +0200826
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500827 This option only enables generic Block IO controller infrastructure.
828 One needs to also enable actual IO controlling logic/policy. For
829 enabling proportional weight division of disk bandwidth in CFQ, set
830 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
831 CONFIG_BLK_DEV_THROTTLING=y.
832
seokhoon.yoon9991a9c2016-08-02 14:03:13 -0700833 See Documentation/cgroup-v1/blkio-controller.txt for more information.
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500834
835config DEBUG_BLK_CGROUP
836 bool "IO controller debugging"
837 depends on BLK_CGROUP
838 default n
839 ---help---
840 Enable some debugging help. Currently it exports additional stat
841 files in a cgroup which can be useful for debugging.
842
843config CGROUP_WRITEBACK
844 bool
845 depends on MEMCG && BLK_CGROUP
846 default y
Stephane Eraniane5d13672011-02-14 11:20:01 +0200847
Dhaval Giani7c941432010-01-20 13:26:18 +0100848menuconfig CGROUP_SCHED
Johannes Weinera0166ec2015-12-17 17:19:56 -0500849 bool "CPU controller"
Dhaval Giani7c941432010-01-20 13:26:18 +0100850 default n
851 help
852 This feature lets CPU scheduler recognize task groups and control CPU
853 bandwidth allocation to such task groups. It uses cgroups to group
854 tasks.
855
856if CGROUP_SCHED
857config FAIR_GROUP_SCHED
858 bool "Group scheduling for SCHED_OTHER"
859 depends on CGROUP_SCHED
860 default CGROUP_SCHED
861
Paul Turnerab84d312011-07-21 09:43:28 -0700862config CFS_BANDWIDTH
863 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
Paul Turnerab84d312011-07-21 09:43:28 -0700864 depends on FAIR_GROUP_SCHED
865 default n
866 help
867 This option allows users to define CPU bandwidth rates (limits) for
868 tasks running within the fair group scheduler. Groups with no limit
869 set are considered to be unconstrained and will run with no
870 restriction.
Sebastian Andrzej Siewiorcd33d882018-05-15 18:53:28 +0200871 See Documentation/scheduler/sched-bwc.txt for more information.
Paul Turnerab84d312011-07-21 09:43:28 -0700872
Dhaval Giani7c941432010-01-20 13:26:18 +0100873config RT_GROUP_SCHED
874 bool "Group scheduling for SCHED_RR/FIFO"
Dhaval Giani7c941432010-01-20 13:26:18 +0100875 depends on CGROUP_SCHED
876 default n
877 help
878 This feature lets you explicitly allocate real CPU bandwidth
Li Zefan32bd7eb2010-03-24 13:17:19 +0800879 to task groups. If enabled, it will also make it impossible to
Dhaval Giani7c941432010-01-20 13:26:18 +0100880 schedule realtime tasks for non-root users until you allocate
881 realtime bandwidth for them.
882 See Documentation/scheduler/sched-rt-group.txt for more information.
883
884endif #CGROUP_SCHED
885
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500886config CGROUP_PIDS
887 bool "PIDs controller"
888 help
889 Provides enforcement of process number limits in the scope of a
890 cgroup. Any attempt to fork more processes than is allowed in the
891 cgroup will fail. PIDs are fundamentally a global resource because it
892 is fairly trivial to reach PID exhaustion before you reach even a
893 conservative kmemcg limit. As a result, it is possible to grind a
894 system to halt without being limited by other cgroup policies. The
Parav Pandit6cc578d2016-03-05 11:30:56 +0530895 PIDs controller is designed to stop this from happening.
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500896
897 It should be noted that organisational operations (such as attaching
Jonathan Neuschäfer98076832019-02-01 14:21:01 -0800898 to a cgroup hierarchy) will *not* be blocked by the PIDs controller,
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500899 since the PIDs limit only affects a process's ability to fork, not to
900 attach to a cgroup.
901
Parav Pandit39d3e752017-01-10 00:02:13 +0000902config CGROUP_RDMA
903 bool "RDMA controller"
904 help
905 Provides enforcement of RDMA resources defined by IB stack.
906 It is fairly easy for consumers to exhaust RDMA resources, which
907 can result into resource unavailability to other consumers.
908 RDMA controller is designed to stop this from happening.
909 Attaching processes with active RDMA resources to the cgroup
910 hierarchy is allowed even if can cross the hierarchy's limit.
911
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500912config CGROUP_FREEZER
913 bool "Freezer controller"
914 help
915 Provides a way to freeze and unfreeze all tasks in a
916 cgroup.
917
Johannes Weiner489c2a22016-01-20 15:02:41 -0800918 This option affects the ORIGINAL cgroup interface. The cgroup2 memory
919 controller includes important in-kernel memory consumers per default.
920
921 If you're using cgroup2, say N.
922
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500923config CGROUP_HUGETLB
924 bool "HugeTLB controller"
925 depends on HUGETLB_PAGE
926 select PAGE_COUNTER
Vivek Goyalafc24d42010-04-26 19:27:56 +0200927 default n
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500928 help
929 Provides a cgroup controller for HugeTLB pages.
930 When you enable this, you can put a per cgroup limit on HugeTLB usage.
931 The limit is enforced during page fault. Since HugeTLB doesn't
932 support page reclaim, enforcing the limit at page fault time implies
933 that, the application will get SIGBUS signal if it tries to access
934 HugeTLB pages beyond its limit. This requires the application to know
935 beforehand how much HugeTLB pages it would require for its use. The
936 control group is tracked in the third page lru pointer. This means
937 that we cannot use the controller with huge page less than 3 pages.
Vivek Goyalafc24d42010-04-26 19:27:56 +0200938
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500939config CPUSETS
940 bool "Cpuset controller"
Nicolas Pitree1d4eee2017-06-14 13:19:23 -0400941 depends on SMP
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500942 help
943 This option will let you create and manage CPUSETs which
944 allow dynamically partitioning a system into sets of CPUs and
945 Memory Nodes and assigning tasks to run only within those sets.
946 This is primarily useful on large SMP or NUMA systems.
Vivek Goyalafc24d42010-04-26 19:27:56 +0200947
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500948 Say N if unsure.
Vivek Goyalafc24d42010-04-26 19:27:56 +0200949
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500950config PROC_PID_CPUSET
951 bool "Include legacy /proc/<pid>/cpuset file"
952 depends on CPUSETS
Tejun Heo89e9b9e2015-05-22 17:13:36 -0400953 default y
954
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500955config CGROUP_DEVICE
956 bool "Device controller"
957 help
958 Provides a cgroup controller implementing whitelists for
959 devices which a process in the cgroup can mknod or open.
960
961config CGROUP_CPUACCT
962 bool "Simple CPU accounting controller"
963 help
964 Provides a simple controller for monitoring the
965 total CPU consumed by the tasks in a cgroup.
966
967config CGROUP_PERF
968 bool "Perf controller"
969 depends on PERF_EVENTS
970 help
971 This option extends the perf per-cpu mode to restrict monitoring
972 to threads which belong to the cgroup specified and run on the
973 designated cpu.
974
975 Say N if unsure.
976
Daniel Mack30070982016-11-23 16:52:26 +0100977config CGROUP_BPF
978 bool "Support for eBPF programs attached to cgroups"
Andy Lutomirski483c4932016-12-16 08:33:45 -0800979 depends on BPF_SYSCALL
980 select SOCK_CGROUP_DATA
Daniel Mack30070982016-11-23 16:52:26 +0100981 help
982 Allow attaching eBPF programs to a cgroup using the bpf(2)
983 syscall command BPF_PROG_ATTACH.
984
985 In which context these programs are accessed depends on the type
986 of attachment. For instance, programs that are attached using
987 BPF_CGROUP_INET_INGRESS will be executed on the ingress path of
988 inet sockets.
989
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500990config CGROUP_DEBUG
Waiman Long23b0be42017-06-13 17:18:03 -0400991 bool "Debug controller"
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500992 default n
Waiman Long23b0be42017-06-13 17:18:03 -0400993 depends on DEBUG_KERNEL
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500994 help
995 This option enables a simple controller that exports
Waiman Long23b0be42017-06-13 17:18:03 -0400996 debugging information about the cgroups framework. This
997 controller is for control cgroup debugging only. Its
998 interfaces are not stable.
Johannes Weiner6bf024e2015-12-17 17:19:57 -0500999
1000 Say N.
1001
Arnd Bergmann73b35142017-01-10 13:08:06 +01001002config SOCK_CGROUP_DATA
1003 bool
1004 default n
1005
Li Zefan23964d22009-01-15 13:50:58 -08001006endif # CGROUPS
KAMEZAWA Hiroyukic0777192009-01-07 18:07:57 -08001007
Daniel Lezcano8dd2a822010-10-27 15:34:38 -07001008menuconfig NAMESPACES
David Rientjes6a108a12011-01-20 14:44:16 -08001009 bool "Namespaces support" if EXPERT
Iulia Manda28138932015-04-15 16:16:41 -07001010 depends on MULTIUSER
David Rientjes6a108a12011-01-20 14:44:16 -08001011 default !EXPERT
Pavel Emelyanovc5289a62008-02-08 04:18:19 -08001012 help
1013 Provides the way to make tasks work with different objects using
1014 the same id. For example same IPC id may refer to different objects
1015 or same user id or pid may refer to different tasks when used in
1016 different namespaces.
1017
Daniel Lezcano8dd2a822010-10-27 15:34:38 -07001018if NAMESPACES
1019
Pavel Emelyanov58bfdd6d2008-02-08 04:18:21 -08001020config UTS_NS
1021 bool "UTS namespace"
Daniel Lezcano17a6d442010-10-27 15:34:37 -07001022 default y
Pavel Emelyanov58bfdd6d2008-02-08 04:18:21 -08001023 help
1024 In this namespace tasks see different info provided with the
1025 uname() system call
1026
Pavel Emelyanovae5e1b22008-02-08 04:18:22 -08001027config IPC_NS
1028 bool "IPC namespace"
Daniel Lezcano8dd2a822010-10-27 15:34:38 -07001029 depends on (SYSVIPC || POSIX_MQUEUE)
Daniel Lezcano17a6d442010-10-27 15:34:37 -07001030 default y
Pavel Emelyanovae5e1b22008-02-08 04:18:22 -08001031 help
1032 In this namespace tasks work with IPC ids which correspond to
Serge E. Hallyn614b84c2009-04-06 19:01:08 -07001033 different IPC objects in different namespaces.
Pavel Emelyanovae5e1b22008-02-08 04:18:22 -08001034
Pavel Emelyanovaee16ce2008-02-08 04:18:23 -08001035config USER_NS
Kees Cook19c92392012-10-02 11:19:29 -07001036 bool "User namespace"
Eric W. Biederman5673a942011-11-17 10:23:55 -08001037 default n
Pavel Emelyanovaee16ce2008-02-08 04:18:23 -08001038 help
1039 This allows containers, i.e. vservers, to use user namespaces
1040 to provide different user info for different servers.
Eric W. Biedermane11f0ae2013-01-25 16:48:31 -08001041
1042 When user namespaces are enabled in the kernel it is
Johannes Weinerd886f4e2016-01-20 15:02:47 -08001043 recommended that the MEMCG option also be enabled and that
1044 user-space use the memory control groups to limit the amount
1045 of memory a memory unprivileged users can use.
Eric W. Biedermane11f0ae2013-01-25 16:48:31 -08001046
Pavel Emelyanovaee16ce2008-02-08 04:18:23 -08001047 If unsure, say N.
1048
Pavel Emelyanov74bd59b2008-02-08 04:18:24 -08001049config PID_NS
Daniel Lezcano9bd38c22010-10-27 15:34:37 -07001050 bool "PID Namespaces"
Daniel Lezcano17a6d442010-10-27 15:34:37 -07001051 default y
Pavel Emelyanov74bd59b2008-02-08 04:18:24 -08001052 help
Heikki Orsila12d2b8f2008-07-06 15:48:02 +03001053 Support process id namespaces. This allows having multiple
Matt LaPlante692105b2009-01-26 11:12:25 +01001054 processes with the same pid as long as they are in different
Pavel Emelyanov74bd59b2008-02-08 04:18:24 -08001055 pid namespaces. This is a building block of containers.
1056
Matt Helsleyd6eb6332009-01-26 12:25:55 -08001057config NET_NS
1058 bool "Network namespace"
Daniel Lezcano8dd2a822010-10-27 15:34:38 -07001059 depends on NET
Daniel Lezcano17a6d442010-10-27 15:34:37 -07001060 default y
Matt Helsleyd6eb6332009-01-26 12:25:55 -08001061 help
1062 Allow user space to create what appear to be multiple instances
1063 of the network stack.
1064
Daniel Lezcano8dd2a822010-10-27 15:34:38 -07001065endif # NAMESPACES
1066
Adrian Reber5cb366b2018-08-21 22:01:17 -07001067config CHECKPOINT_RESTORE
1068 bool "Checkpoint/restore support"
1069 select PROC_CHILDREN
1070 default n
1071 help
1072 Enables additional kernel features in a sake of checkpoint/restore.
1073 In particular it adds auxiliary prctl codes to setup process text,
1074 data and heap segment sizes, and a few additional /proc filesystem
1075 entries.
1076
1077 If unsure, say N here.
1078
Mike Galbraith5091faa2010-11-30 14:18:03 +01001079config SCHED_AUTOGROUP
1080 bool "Automatic process group scheduling"
Mike Galbraith5091faa2010-11-30 14:18:03 +01001081 select CGROUPS
1082 select CGROUP_SCHED
1083 select FAIR_GROUP_SCHED
1084 help
1085 This option optimizes the scheduler for common desktop workloads by
1086 automatically creating and populating task groups. This separation
1087 of workloads isolates aggressive CPU burners (like build jobs) from
1088 desktop applications. Task group autogeneration is currently based
1089 upon task session.
1090
Daniel Lezcano7af37be2010-10-27 15:34:41 -07001091config SYSFS_DEPRECATED
Ferenc Wagner5d6a4ea2011-01-10 19:04:22 +01001092 bool "Enable deprecated sysfs features to support old userspace tools"
Daniel Lezcano7af37be2010-10-27 15:34:41 -07001093 depends on SYSFS
1094 default n
1095 help
1096 This option adds code that switches the layout of the "block" class
1097 devices, to not show up in /sys/class/block/, but only in
1098 /sys/block/.
1099
1100 This switch is only active when the sysfs.deprecated=1 boot option is
1101 passed or the SYSFS_DEPRECATED_V2 option is set.
1102
1103 This option allows new kernels to run on old distributions and tools,
1104 which might get confused by /sys/class/block/. Since 2007/2008 all
1105 major distributions and tools handle this just fine.
1106
1107 Recent distributions and userspace tools after 2009/2010 depend on
1108 the existence of /sys/class/block/, and will not work with this
1109 option enabled.
1110
1111 Only if you are using a new kernel on an old distribution, you might
1112 need to say Y here.
1113
1114config SYSFS_DEPRECATED_V2
Ferenc Wagner5d6a4ea2011-01-10 19:04:22 +01001115 bool "Enable deprecated sysfs features by default"
Daniel Lezcano7af37be2010-10-27 15:34:41 -07001116 default n
1117 depends on SYSFS
1118 depends on SYSFS_DEPRECATED
1119 help
1120 Enable deprecated sysfs by default.
1121
1122 See the CONFIG_SYSFS_DEPRECATED option for more details about this
1123 option.
1124
1125 Only if you are using a new kernel on an old distribution, you might
1126 need to say Y here. Even then, odds are you would not need it
1127 enabled, you can always pass the boot option if absolutely necessary.
1128
1129config RELAY
1130 bool "Kernel->user space relay support (formerly relayfs)"
Peter Zijlstra26b56792016-10-11 13:54:33 -07001131 select IRQ_WORK
Daniel Lezcano7af37be2010-10-27 15:34:41 -07001132 help
1133 This option enables support for relay interface support in
1134 certain file systems (such as debugfs).
1135 It is designed to provide an efficient mechanism for tools and
1136 facilities to relay large amounts of data from kernel space to
1137 user space.
1138
1139 If unsure, say N.
1140
Dimitri Gorokhovikf9916332007-03-06 01:42:17 -08001141config BLK_DEV_INITRD
1142 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
Dimitri Gorokhovikf9916332007-03-06 01:42:17 -08001143 help
1144 The initial RAM filesystem is a ramfs which is loaded by the
1145 boot loader (loadlin or lilo) and that is mounted as root
1146 before the normal boot procedure. It is typically used to
1147 load modules needed to mount the "real" root file system,
Mauro Carvalho Chehab8c27ceff32016-10-18 10:12:27 -02001148 etc. See <file:Documentation/admin-guide/initrd.rst> for details.
Dimitri Gorokhovikf9916332007-03-06 01:42:17 -08001149
1150 If RAM disk support (BLK_DEV_RAM) is also included, this
1151 also enables initial RAM disk (initrd) support and adds
1152 15 Kbytes (more on some other architectures) to the kernel size.
1153
1154 If unsure say Y.
1155
Jean-Paul Samanc33df4e2007-02-10 01:44:43 -08001156if BLK_DEV_INITRD
1157
Sam Ravnborgdbec4862005-08-10 20:44:50 +02001158source "usr/Kconfig"
1159
Jean-Paul Samanc33df4e2007-02-10 01:44:43 -08001160endif
1161
Arnd Bergmann877417e2016-04-25 17:35:27 +02001162choice
1163 prompt "Compiler optimization level"
Ulf Magnusson2cc3ce22017-10-04 01:53:26 +02001164 default CC_OPTIMIZE_FOR_PERFORMANCE
Arnd Bergmann877417e2016-04-25 17:35:27 +02001165
1166config CC_OPTIMIZE_FOR_PERFORMANCE
1167 bool "Optimize for performance"
1168 help
1169 This is the default optimization level for the kernel, building
1170 with the "-O2" compiler flag for best performance and most
1171 helpful compile-time warnings.
1172
Linus Torvaldsc45b4f12005-12-14 18:52:21 -08001173config CC_OPTIMIZE_FOR_SIZE
Ingo Molnar96fffeb2008-04-28 01:39:43 +02001174 bool "Optimize for size"
Masahiro Yamadab303c6d2019-02-21 13:13:38 +09001175 imply CC_DISABLE_WARN_MAYBE_UNINITIALIZED # avoid false positives
Linus Torvaldsc45b4f12005-12-14 18:52:21 -08001176 help
Masahiro Yamada31a4af72014-08-05 14:43:07 +09001177 Enabling this option will pass "-Os" instead of "-O2" to
1178 your compiler resulting in a smaller kernel.
Linus Torvaldsc45b4f12005-12-14 18:52:21 -08001179
Kirill Smelkov3a55fb02012-11-02 15:41:01 +04001180 If unsure, say N.
Linus Torvaldsc45b4f12005-12-14 18:52:21 -08001181
Arnd Bergmann877417e2016-04-25 17:35:27 +02001182endchoice
1183
Nicholas Piggin5d20ee32018-05-09 23:00:00 +10001184config HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1185 bool
1186 help
1187 This requires that the arch annotates or otherwise protects
1188 its external entry points from being discarded. Linker scripts
1189 must also merge .text.*, .data.*, and .bss.* correctly into
1190 output sections. Care must be taken not to pull in unrelated
1191 sections (e.g., '.text.init'). Typically '.' in section names
1192 is used to distinguish them from label names / C identifiers.
1193
1194config LD_DEAD_CODE_DATA_ELIMINATION
1195 bool "Dead code and data elimination (EXPERIMENTAL)"
1196 depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1197 depends on EXPERT
Paul Burton16fd20a2019-01-11 19:06:44 +00001198 depends on !(FUNCTION_TRACER && CC_IS_GCC && GCC_VERSION < 40800)
Masahiro Yamadae85d1d62018-08-22 22:51:09 +09001199 depends on $(cc-option,-ffunction-sections -fdata-sections)
1200 depends on $(ld-option,--gc-sections)
Nicholas Piggin5d20ee32018-05-09 23:00:00 +10001201 help
Masahiro Yamada8b9d2712018-06-24 01:41:51 +09001202 Enable this if you want to do dead code and data elimination with
1203 the linker by compiling with -ffunction-sections -fdata-sections,
1204 and linking with --gc-sections.
Nicholas Piggin5d20ee32018-05-09 23:00:00 +10001205
1206 This can reduce on disk and in-memory size of the kernel
1207 code and static data, particularly for small configs and
1208 on small systems. This has the possibility of introducing
1209 silently broken kernel if the required annotations are not
1210 present. This option is not well tested yet, so use at your
1211 own risk.
1212
Randy Dunlap08470622006-09-30 23:28:13 -07001213config SYSCTL
1214 bool
1215
Mike Frysinger657a5202013-04-30 15:28:45 -07001216config HAVE_UID16
1217 bool
1218
1219config SYSCTL_EXCEPTION_TRACE
1220 bool
1221 help
1222 Enable support for /proc/sys/debug/exception-trace.
1223
1224config SYSCTL_ARCH_UNALIGN_NO_WARN
1225 bool
1226 help
1227 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1228 Allows arch to define/use @no_unaligned_warning to possibly warn
1229 about unaligned access emulation going on under the hood.
1230
1231config SYSCTL_ARCH_UNALIGN_ALLOW
1232 bool
1233 help
1234 Enable support for /proc/sys/kernel/unaligned-trap
1235 Allows arches to define/use @unaligned_enabled to runtime toggle
1236 the unaligned access emulation.
1237 see arch/parisc/kernel/unaligned.c for reference
1238
Mike Frysinger657a5202013-04-30 15:28:45 -07001239config HAVE_PCSPKR_PLATFORM
1240 bool
1241
Alexei Starovoitovf89b7752014-10-23 18:41:08 -07001242# interpreter that classic socket filters depend on
1243config BPF
1244 bool
1245
David Rientjes6a108a12011-01-20 14:44:16 -08001246menuconfig EXPERT
1247 bool "Configure standard kernel features (expert users)"
Josh Triplettf505c552011-06-05 18:23:58 -07001248 # Unhide debug options, to make the on-by-default options visible
1249 select DEBUG_KERNEL
Linus Torvalds1da177e2005-04-16 15:20:36 -07001250 help
1251 This option allows certain base kernel options and settings
1252 to be disabled or tweaked. This is for specialized
1253 environments which can tolerate a "non-standard" kernel.
1254 Only use this if you really know what you are doing.
1255
Chuck Ebbertae81f9e2006-09-16 12:15:53 -07001256config UID16
David Rientjes6a108a12011-01-20 14:44:16 -08001257 bool "Enable 16-bit UID system calls" if EXPERT
Iulia Manda28138932015-04-15 16:16:41 -07001258 depends on HAVE_UID16 && MULTIUSER
Chuck Ebbertae81f9e2006-09-16 12:15:53 -07001259 default y
1260 help
1261 This enables the legacy 16-bit UID syscall wrappers.
1262
Iulia Manda28138932015-04-15 16:16:41 -07001263config MULTIUSER
1264 bool "Multiple users, groups and capabilities support" if EXPERT
1265 default y
1266 help
1267 This option enables support for non-root users, groups and
1268 capabilities.
1269
1270 If you say N here, all processes will run with UID 0, GID 0, and all
1271 possible capabilities. Saying N here also compiles out support for
1272 system calls related to UIDs, GIDs, and capabilities, such as setuid,
1273 setgid, and capset.
1274
1275 If unsure, say Y here.
1276
Fabian Frederickf6187762014-06-04 16:11:12 -07001277config SGETMASK_SYSCALL
1278 bool "sgetmask/ssetmask syscalls support" if EXPERT
Arnd Bergmanna687a532018-03-07 23:30:54 +01001279 def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH
Fabian Frederickf6187762014-06-04 16:11:12 -07001280 ---help---
1281 sys_sgetmask and sys_ssetmask are obsolete system calls
1282 no longer supported in libc but still enabled by default in some
1283 architectures.
1284
1285 If unsure, leave the default option here.
1286
Fabian Frederick6af9f7b2014-04-03 14:48:25 -07001287config SYSFS_SYSCALL
1288 bool "Sysfs syscall support" if EXPERT
1289 default y
1290 ---help---
1291 sys_sysfs is an obsolete system call no longer supported in libc.
1292 Note that disabling this option is more secure but might break
1293 compatibility with some systems.
1294
1295 If unsure say Y here.
1296
Eric W. Biedermanb89a8172006-09-27 01:51:04 -07001297config SYSCTL_SYSCALL
David Rientjes6a108a12011-01-20 14:44:16 -08001298 bool "Sysctl syscall support" if EXPERT
Eric W. Biederman26a70342009-11-05 05:26:41 -08001299 depends on PROC_SYSCTL
WANG Congc736de62011-11-02 13:39:25 -07001300 default n
Eric W. Biedermanb89a8172006-09-27 01:51:04 -07001301 select SYSCTL
1302 ---help---
Eric W. Biederman13bb7e32006-11-08 17:44:51 -08001303 sys_sysctl uses binary paths that have been found challenging
1304 to properly maintain and use. The interface in /proc/sys
1305 using paths with ascii names is now the primary path to this
1306 information.
Eric W. Biedermanb89a8172006-09-27 01:51:04 -07001307
Eric W. Biederman13bb7e32006-11-08 17:44:51 -08001308 Almost nothing using the binary sysctl interface so if you are
1309 trying to save some space it is probably safe to disable this,
1310 making your kernel marginally smaller.
Eric W. Biedermanb89a8172006-09-27 01:51:04 -07001311
WANG Congc736de62011-11-02 13:39:25 -07001312 If unsure say N here.
Chuck Ebbertae81f9e2006-09-16 12:15:53 -07001313
Randy Dunlapd1b069f2017-11-17 15:31:47 -08001314config FHANDLE
1315 bool "open by fhandle syscalls" if EXPERT
1316 select EXPORTFS
1317 default y
1318 help
1319 If you say Y here, a user level program will be able to map
1320 file names to handle and then later use the handle for
1321 different file system operations. This is useful in implementing
1322 userspace file servers, which now track files using handles instead
1323 of names. The handle would remain the same even if file names
1324 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
1325 syscalls.
1326
Nicolas Pitrebaa73d92016-11-11 00:10:10 -05001327config POSIX_TIMERS
1328 bool "Posix Clocks & timers" if EXPERT
1329 default y
1330 help
1331 This includes native support for POSIX timers to the kernel.
1332 Some embedded systems have no use for them and therefore they
1333 can be configured out to reduce the size of the kernel image.
1334
1335 When this option is disabled, the following syscalls won't be
1336 available: timer_create, timer_gettime: timer_getoverrun,
1337 timer_settime, timer_delete, clock_adjtime, getitimer,
1338 setitimer, alarm. Furthermore, the clock_settime, clock_gettime,
1339 clock_getres and clock_nanosleep syscalls will be limited to
1340 CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only.
1341
1342 If unsure say y.
1343
Matt Mackalld59745c2005-05-01 08:59:02 -07001344config PRINTK
1345 default y
David Rientjes6a108a12011-01-20 14:44:16 -08001346 bool "Enable support for printk" if EXPERT
Frederic Weisbecker74876a92012-10-12 18:00:23 +02001347 select IRQ_WORK
Matt Mackalld59745c2005-05-01 08:59:02 -07001348 help
1349 This option enables normal printk support. Removing it
1350 eliminates most of the message strings from the kernel image
1351 and makes the kernel more or less silent. As this makes it
1352 very difficult to diagnose system problems, saying N here is
1353 strongly discouraged.
1354
Petr Mladek42a0bb32016-05-20 17:00:33 -07001355config PRINTK_NMI
1356 def_bool y
1357 depends on PRINTK
1358 depends on HAVE_NMI
1359
Matt Mackallc8538a72005-05-01 08:59:01 -07001360config BUG
David Rientjes6a108a12011-01-20 14:44:16 -08001361 bool "BUG() support" if EXPERT
Matt Mackallc8538a72005-05-01 08:59:01 -07001362 default y
1363 help
1364 Disabling this option eliminates support for BUG and WARN, reducing
1365 the size of your kernel image and potentially quietly ignoring
1366 numerous fatal conditions. You should only consider disabling this
1367 option for embedded systems with no facilities for reporting errors.
1368 Just say Y.
1369
Matt Mackall708e9a72006-01-08 01:05:25 -08001370config ELF_CORE
Alex Kelly046d6622012-10-04 17:15:23 -07001371 depends on COREDUMP
Matt Mackall708e9a72006-01-08 01:05:25 -08001372 default y
David Rientjes6a108a12011-01-20 14:44:16 -08001373 bool "Enable ELF core dumps" if EXPERT
Matt Mackall708e9a72006-01-08 01:05:25 -08001374 help
1375 Enable support for generating core dumps. Disabling saves about 4k.
1376
Ralf Baechle8761f1a2011-06-01 19:05:09 +01001377
Stas Sergeeve5e1d3c2008-05-07 12:39:56 +02001378config PCSPKR_PLATFORM
David Rientjes6a108a12011-01-20 14:44:16 -08001379 bool "Enable PC-Speaker support" if EXPERT
Ralf Baechle8761f1a2011-06-01 19:05:09 +01001380 depends on HAVE_PCSPKR_PLATFORM
Ralf Baechle15f304b2011-06-01 19:04:59 +01001381 select I8253_LOCK
Stas Sergeeve5e1d3c2008-05-07 12:39:56 +02001382 default y
1383 help
1384 This option allows to disable the internal PC-Speaker
1385 support, saving some memory.
1386
Linus Torvalds1da177e2005-04-16 15:20:36 -07001387config BASE_FULL
1388 default y
David Rientjes6a108a12011-01-20 14:44:16 -08001389 bool "Enable full-sized data structures for core" if EXPERT
Linus Torvalds1da177e2005-04-16 15:20:36 -07001390 help
1391 Disabling this option reduces the size of miscellaneous core
1392 kernel data structures. This saves memory on small machines,
1393 but may reduce performance.
1394
1395config FUTEX
David Rientjes6a108a12011-01-20 14:44:16 -08001396 bool "Enable futex support" if EXPERT
Linus Torvalds1da177e2005-04-16 15:20:36 -07001397 default y
Nicolas Pitrebc2eecd2017-08-01 00:31:32 -04001398 imply RT_MUTEXES
Linus Torvalds1da177e2005-04-16 15:20:36 -07001399 help
1400 Disabling this option will cause the kernel to be built without
1401 support for "fast userspace mutexes". The resulting kernel may not
1402 run glibc-based applications correctly.
1403
Nicolas Pitrebc2eecd2017-08-01 00:31:32 -04001404config FUTEX_PI
1405 bool
1406 depends on FUTEX && RT_MUTEXES
1407 default y
1408
Heiko Carstens03b8c7b2014-03-02 13:09:47 +01001409config HAVE_FUTEX_CMPXCHG
1410 bool
Josh Triplett62b4d202014-10-03 16:19:24 -07001411 depends on FUTEX
Heiko Carstens03b8c7b2014-03-02 13:09:47 +01001412 help
1413 Architectures should select this if futex_atomic_cmpxchg_inatomic()
1414 is implemented and always working. This removes a couple of runtime
1415 checks.
1416
Linus Torvalds1da177e2005-04-16 15:20:36 -07001417config EPOLL
David Rientjes6a108a12011-01-20 14:44:16 -08001418 bool "Enable eventpoll support" if EXPERT
Linus Torvalds1da177e2005-04-16 15:20:36 -07001419 default y
1420 help
1421 Disabling this option will cause the kernel to be built without
1422 support for epoll family of system calls.
1423
Davide Libenzifba2afa2007-05-10 22:23:13 -07001424config SIGNALFD
David Rientjes6a108a12011-01-20 14:44:16 -08001425 bool "Enable signalfd() system call" if EXPERT
Davide Libenzifba2afa2007-05-10 22:23:13 -07001426 default y
1427 help
1428 Enable the signalfd() system call that allows to receive signals
1429 on a file descriptor.
1430
1431 If unsure, say Y.
1432
Davide Libenzib215e282007-05-10 22:23:16 -07001433config TIMERFD
David Rientjes6a108a12011-01-20 14:44:16 -08001434 bool "Enable timerfd() system call" if EXPERT
Davide Libenzib215e282007-05-10 22:23:16 -07001435 default y
1436 help
1437 Enable the timerfd() system call that allows to receive timer
1438 events on a file descriptor.
1439
1440 If unsure, say Y.
1441
Davide Libenzie1ad7462007-05-10 22:23:19 -07001442config EVENTFD
David Rientjes6a108a12011-01-20 14:44:16 -08001443 bool "Enable eventfd() system call" if EXPERT
Davide Libenzie1ad7462007-05-10 22:23:19 -07001444 default y
1445 help
1446 Enable the eventfd() system call that allows to receive both
1447 kernel notification (ie. KAIO) or userspace notifications.
1448
1449 If unsure, say Y.
1450
Linus Torvalds1da177e2005-04-16 15:20:36 -07001451config SHMEM
David Rientjes6a108a12011-01-20 14:44:16 -08001452 bool "Use full shmem filesystem" if EXPERT
Linus Torvalds1da177e2005-04-16 15:20:36 -07001453 default y
1454 depends on MMU
1455 help
1456 The shmem is an internal filesystem used to manage shared memory.
1457 It is backed by swap and manages resource limits. It is also exported
1458 to userspace as tmpfs if TMPFS is enabled. Disabling this
1459 option replaces shmem and tmpfs with the much simpler ramfs code,
1460 which may be appropriate on small systems without swap.
1461
Thomas Petazzoniebf3f092008-10-15 22:05:12 -07001462config AIO
David Rientjes6a108a12011-01-20 14:44:16 -08001463 bool "Enable AIO support" if EXPERT
Thomas Petazzoniebf3f092008-10-15 22:05:12 -07001464 default y
1465 help
1466 This option enables POSIX asynchronous I/O which may by used
Mike Frysinger657a5202013-04-30 15:28:45 -07001467 by some high performance threaded applications. Disabling
1468 this option saves about 7k.
1469
Jens Axboe2b188cc2019-01-07 10:46:33 -07001470config IO_URING
1471 bool "Enable IO uring support" if EXPERT
1472 select ANON_INODES
1473 default y
1474 help
1475 This option enables support for the io_uring interface, enabling
1476 applications to submit and complete IO through submission and
1477 completion rings that are shared between the kernel and application.
1478
Josh Triplettd3ac21c2014-08-17 19:41:09 -05001479config ADVISE_SYSCALLS
1480 bool "Enable madvise/fadvise syscalls" if EXPERT
1481 default y
1482 help
1483 This option enables the madvise and fadvise syscalls, used by
1484 applications to advise the kernel about their future memory or file
1485 usage, improving performance. If building an embedded system where no
1486 applications use these syscalls, you can disable this option to save
1487 space.
1488
Mathieu Desnoyers5b25b132015-09-11 13:07:39 -07001489config MEMBARRIER
1490 bool "Enable membarrier() system call" if EXPERT
1491 default y
1492 help
1493 Enable the membarrier() system call that allows issuing memory
1494 barriers across all running threads, which can be used to distribute
1495 the cost of user-space memory barriers asymmetrically by transforming
1496 pairs of memory barriers into pairs consisting of membarrier() and a
1497 compiler barrier.
1498
1499 If unsure, say Y.
1500
Randy Dunlapd1b069f2017-11-17 15:31:47 -08001501config KALLSYMS
1502 bool "Load all symbols for debugging/ksymoops" if EXPERT
1503 default y
1504 help
1505 Say Y here to let the kernel print out symbolic crash information and
1506 symbolic stack backtraces. This increases the size of the kernel
1507 somewhat, as all symbols have to be loaded into the kernel image.
1508
1509config KALLSYMS_ALL
1510 bool "Include all symbols in kallsyms"
1511 depends on DEBUG_KERNEL && KALLSYMS
1512 help
1513 Normally kallsyms only contains the symbols of functions for nicer
1514 OOPS messages and backtraces (i.e., symbols from the text and inittext
1515 sections). This is sufficient for most cases. And only in very rare
1516 cases (e.g., when a debugger is used) all symbols are required (e.g.,
1517 names of variables from the data sections, etc).
1518
1519 This option makes sure that all symbols are loaded into the kernel
1520 image (i.e., symbols from all sections) in cost of increased kernel
1521 size (depending on the kernel configuration, it may be 300KiB or
1522 something like this).
1523
1524 Say N unless you really need all symbols.
1525
1526config KALLSYMS_ABSOLUTE_PERCPU
1527 bool
1528 depends on KALLSYMS
1529 default X86_64 && SMP
1530
1531config KALLSYMS_BASE_RELATIVE
1532 bool
1533 depends on KALLSYMS
Arnd Bergmanna687a532018-03-07 23:30:54 +01001534 default !IA64
Randy Dunlapd1b069f2017-11-17 15:31:47 -08001535 help
1536 Instead of emitting them as absolute values in the native word size,
1537 emit the symbol references in the kallsyms table as 32-bit entries,
1538 each containing a relative value in the range [base, base + U32_MAX]
1539 or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either
1540 an absolute value in the range [0, S32_MAX] or a relative value in the
1541 range [base, base + S32_MAX], where base is the lowest relative symbol
1542 address encountered in the image.
1543
1544 On 64-bit builds, this reduces the size of the address table by 50%,
1545 but more importantly, it results in entries whose values are build
1546 time constants, and no relocation pass is required at runtime to fix
1547 up the entries based on the runtime load address of the kernel.
1548
1549# end of the "standard kernel features (expert users)" menu
1550
1551# syscall, maps, verifier
1552config BPF_SYSCALL
1553 bool "Enable bpf() system call"
Randy Dunlapd1b069f2017-11-17 15:31:47 -08001554 select BPF
Song Liubae77c52018-05-07 10:50:48 -07001555 select IRQ_WORK
Randy Dunlapd1b069f2017-11-17 15:31:47 -08001556 default n
1557 help
1558 Enable the bpf() system call that allows to manipulate eBPF
1559 programs and maps via file descriptors.
1560
Alexei Starovoitov290af862018-01-09 10:04:29 -08001561config BPF_JIT_ALWAYS_ON
1562 bool "Permanently enable BPF JIT and remove BPF interpreter"
1563 depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT
1564 help
1565 Enables BPF JIT and removes BPF interpreter to avoid
1566 speculative execution of BPF instructions by the interpreter
1567
Randy Dunlapd1b069f2017-11-17 15:31:47 -08001568config USERFAULTFD
1569 bool "Enable userfaultfd() system call"
Randy Dunlapd1b069f2017-11-17 15:31:47 -08001570 depends on MMU
1571 help
1572 Enable the userfaultfd() system call that allows to intercept and
1573 handle page faults in userland.
1574
Mathieu Desnoyers3ccfebe2018-01-29 15:20:11 -05001575config ARCH_HAS_MEMBARRIER_CALLBACKS
1576 bool
1577
Mathieu Desnoyers70216e12018-01-29 15:20:17 -05001578config ARCH_HAS_MEMBARRIER_SYNC_CORE
1579 bool
1580
Mathieu Desnoyersd7822b12018-06-02 08:43:54 -04001581config RSEQ
1582 bool "Enable rseq() system call" if EXPERT
1583 default y
1584 depends on HAVE_RSEQ
1585 select MEMBARRIER
1586 help
1587 Enable the restartable sequences system call. It provides a
1588 user-space cache for the current CPU number value, which
1589 speeds up getting the current CPU number from user-space,
1590 as well as an ABI to speed up user-space operations on
1591 per-CPU data.
1592
1593 If unsure, say Y.
1594
1595config DEBUG_RSEQ
1596 default n
1597 bool "Enabled debugging of rseq() system call" if EXPERT
1598 depends on RSEQ && DEBUG_KERNEL
1599 help
1600 Enable extra debugging checks for the rseq system call.
1601
1602 If unsure, say N.
1603
Randy Dunlap6befe5f2011-04-26 12:33:21 -07001604config EMBEDDED
1605 bool "Embedded system"
Josh Triplett5d2acfc2014-04-07 15:39:09 -07001606 option allnoconfig_y
Randy Dunlap6befe5f2011-04-26 12:33:21 -07001607 select EXPERT
1608 help
1609 This option should be enabled if compiling the kernel for
1610 an embedded system so certain expert options are available
1611 for configuration.
1612
Ingo Molnarcdd6c482009-09-21 12:02:48 +02001613config HAVE_PERF_EVENTS
Thomas Gleixner0793a612008-12-04 20:12:29 +01001614 bool
Mike Frysinger018df722009-06-12 13:17:43 -04001615 help
1616 See tools/perf/design.txt for details.
Thomas Gleixner0793a612008-12-04 20:12:29 +01001617
Peter Zijlstra906010b2009-09-21 16:08:49 +02001618config PERF_USE_VMALLOC
1619 bool
1620 help
1621 See tools/perf/design.txt for details
1622
William Breathitt Grayad90a3d2017-01-10 13:50:54 -05001623config PC104
William Breathitt Gray424529f2017-12-29 15:14:59 -05001624 bool "PC/104 support" if EXPERT
William Breathitt Grayad90a3d2017-01-10 13:50:54 -05001625 help
1626 Expose PC/104 form factor device drivers and options available for
1627 selection and configuration. Enable this option if your target
1628 machine has a PC/104 bus.
1629
Ingo Molnar57c0c152009-09-21 12:20:38 +02001630menu "Kernel Performance Events And Counters"
Thomas Gleixner0793a612008-12-04 20:12:29 +01001631
Ingo Molnarcdd6c482009-09-21 12:02:48 +02001632config PERF_EVENTS
Ingo Molnar57c0c152009-09-21 12:20:38 +02001633 bool "Kernel performance events and counters"
Robert Richter392d65a2012-04-05 18:24:44 +02001634 default y if PROFILING
Ingo Molnarcdd6c482009-09-21 12:02:48 +02001635 depends on HAVE_PERF_EVENTS
Peter Zijlstrae360adb2010-10-14 14:01:34 +08001636 select IRQ_WORK
Pranith Kumar83fe27e2014-12-05 11:24:45 -05001637 select SRCU
Thomas Gleixner0793a612008-12-04 20:12:29 +01001638 help
Ingo Molnar57c0c152009-09-21 12:20:38 +02001639 Enable kernel support for various performance events provided
1640 by software and hardware.
Thomas Gleixner0793a612008-12-04 20:12:29 +01001641
Thadeu Lima de Souza Cascardodd770382009-10-30 19:32:25 -02001642 Software events are supported either built-in or via the
Ingo Molnar57c0c152009-09-21 12:20:38 +02001643 use of generic tracepoints.
1644
1645 Most modern CPUs support performance events via performance
1646 counter registers. These registers count the number of certain
Thomas Gleixner0793a612008-12-04 20:12:29 +01001647 types of hw events: such as instructions executed, cachemisses
1648 suffered, or branches mis-predicted - without slowing down the
1649 kernel or applications. These registers can also trigger interrupts
1650 when a threshold number of events have passed - and can thus be
1651 used to profile the code that runs on that CPU.
1652
Ingo Molnar57c0c152009-09-21 12:20:38 +02001653 The Linux Performance Event subsystem provides an abstraction of
Thadeu Lima de Souza Cascardodd770382009-10-30 19:32:25 -02001654 these software and hardware event capabilities, available via a
Ingo Molnar57c0c152009-09-21 12:20:38 +02001655 system call and used by the "perf" utility in tools/perf/. It
Thomas Gleixner0793a612008-12-04 20:12:29 +01001656 provides per task and per CPU counters, and it provides event
1657 capabilities on top of those.
1658
1659 Say Y if unsure.
1660
Peter Zijlstra906010b2009-09-21 16:08:49 +02001661config DEBUG_PERF_USE_VMALLOC
1662 default n
1663 bool "Debug: use vmalloc to back perf mmap() buffers"
Michael Ellermancb3071132015-05-04 16:26:39 +10001664 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC
Peter Zijlstra906010b2009-09-21 16:08:49 +02001665 select PERF_USE_VMALLOC
1666 help
1667 Use vmalloc memory to back perf mmap() buffers.
1668
1669 Mostly useful for debugging the vmalloc code on platforms
1670 that don't require it.
1671
1672 Say N if unsure.
1673
Thomas Gleixner0793a612008-12-04 20:12:29 +01001674endmenu
1675
Christoph Lameterf8891e52006-06-30 01:55:45 -07001676config VM_EVENT_COUNTERS
1677 default y
David Rientjes6a108a12011-01-20 14:44:16 -08001678 bool "Enable VM event counters for /proc/vmstat" if EXPERT
Christoph Lameterf8891e52006-06-30 01:55:45 -07001679 help
Paul Jackson2aea4fb2006-12-22 01:06:10 -08001680 VM event counters are needed for event counts to be shown.
1681 This option allows the disabling of the VM event counters
David Rientjes6a108a12011-01-20 14:44:16 -08001682 on EXPERT systems. /proc/vmstat will only show page counts
Paul Jackson2aea4fb2006-12-22 01:06:10 -08001683 if VM event counters are disabled.
Christoph Lameterf8891e52006-06-30 01:55:45 -07001684
Christoph Lameter41ecc552007-05-09 02:32:44 -07001685config SLUB_DEBUG
1686 default y
David Rientjes6a108a12011-01-20 14:44:16 -08001687 bool "Enable SLUB debugging support" if EXPERT
Christoph Lameterf6acb632008-04-29 16:16:06 -07001688 depends on SLUB && SYSFS
Christoph Lameter41ecc552007-05-09 02:32:44 -07001689 help
1690 SLUB has extensive debug support features. Disabling these can
1691 result in significant savings in code size. This also disables
1692 SLUB sysfs support. /sys/slab will not exist and there will be
1693 no support for cache validation etc.
1694
Tejun Heo1663f262017-02-22 15:41:39 -08001695config SLUB_MEMCG_SYSFS_ON
1696 default n
1697 bool "Enable memcg SLUB sysfs support by default" if EXPERT
1698 depends on SLUB && SYSFS && MEMCG
1699 help
1700 SLUB creates a directory under /sys/kernel/slab for each
1701 allocation cache to host info and debug files. If memory
1702 cgroup is enabled, each cache can have per memory cgroup
1703 caches. SLUB can create the same sysfs directories for these
1704 caches under /sys/kernel/slab/CACHE/cgroup but it can lead
1705 to a very high number of debug files being created. This is
1706 controlled by slub_memcg_sysfs boot parameter and this
1707 config option determines the parameter's default value.
1708
Randy Dunlapb943c462009-03-10 12:55:46 -07001709config COMPAT_BRK
1710 bool "Disable heap randomization"
1711 default y
1712 help
1713 Randomizing heap placement makes heap exploits harder, but it
1714 also breaks ancient binaries (including anything libc5 based).
1715 This option changes the bootup default to heap randomization
Matt LaPlante692105b2009-01-26 11:12:25 +01001716 disabled, and can be overridden at runtime by setting
Randy Dunlapb943c462009-03-10 12:55:46 -07001717 /proc/sys/kernel/randomize_va_space to 2.
1718
1719 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1720
Christoph Lameter81819f02007-05-06 14:49:36 -07001721choice
1722 prompt "Choose SLAB allocator"
Christoph Lametera0acd822007-07-17 04:03:32 -07001723 default SLUB
Christoph Lameter81819f02007-05-06 14:49:36 -07001724 help
1725 This option allows to select a slab allocator.
1726
1727config SLAB
1728 bool "SLAB"
Kees Cook04385fc2016-06-23 15:20:59 -07001729 select HAVE_HARDENED_USERCOPY_ALLOCATOR
Christoph Lameter81819f02007-05-06 14:49:36 -07001730 help
1731 The regular slab allocator that is established and known to work
Christoph Lameter34013882007-05-09 02:32:47 -07001732 well in all environments. It organizes cache hot objects in
Simon Arlott02f56212008-11-05 22:18:19 +00001733 per cpu and per node queues.
Christoph Lameter81819f02007-05-06 14:49:36 -07001734
1735config SLUB
Christoph Lameter81819f02007-05-06 14:49:36 -07001736 bool "SLUB (Unqueued Allocator)"
Kees Cooked18adc2016-06-23 15:24:05 -07001737 select HAVE_HARDENED_USERCOPY_ALLOCATOR
Christoph Lameter81819f02007-05-06 14:49:36 -07001738 help
1739 SLUB is a slab allocator that minimizes cache line usage
1740 instead of managing queues of cached objects (SLAB approach).
1741 Per cpu caching is realized using slabs of objects instead
1742 of queues of objects. SLUB can use memory efficiently
Simon Arlott02f56212008-11-05 22:18:19 +00001743 and has enhanced diagnostics. SLUB is the default choice for
1744 a slab allocator.
Christoph Lameter81819f02007-05-06 14:49:36 -07001745
1746config SLOB
David Rientjes6a108a12011-01-20 14:44:16 -08001747 depends on EXPERT
Christoph Lameter81819f02007-05-06 14:49:36 -07001748 bool "SLOB (Simple Allocator)"
1749 help
Matt Mackall37291452008-02-04 22:29:38 -08001750 SLOB replaces the stock allocator with a drastically simpler
1751 allocator. SLOB is generally more space efficient but
1752 does not perform as well on large systems.
Christoph Lameter81819f02007-05-06 14:49:36 -07001753
1754endchoice
1755
Kees Cook7660a6f2017-07-06 15:36:40 -07001756config SLAB_MERGE_DEFAULT
1757 bool "Allow slab caches to be merged"
1758 default y
1759 help
1760 For reduced kernel memory fragmentation, slab caches can be
1761 merged when they share the same size and other characteristics.
1762 This carries a risk of kernel heap overflows being able to
1763 overwrite objects from merged caches (and more easily control
1764 cache layout), which makes such heap attacks easier to exploit
1765 by attackers. By keeping caches unmerged, these kinds of exploits
1766 can usually only damage objects in the same cache. To disable
1767 merging at runtime, "slab_nomerge" can be passed on the kernel
1768 command line.
1769
Thomas Garnierc7ce4f602016-05-19 17:10:37 -07001770config SLAB_FREELIST_RANDOM
1771 default n
Thomas Garnier210e7a42016-07-26 15:21:59 -07001772 depends on SLAB || SLUB
Thomas Garnierc7ce4f602016-05-19 17:10:37 -07001773 bool "SLAB freelist randomization"
1774 help
Thomas Garnier210e7a42016-07-26 15:21:59 -07001775 Randomizes the freelist order used on creating new pages. This
Thomas Garnierc7ce4f602016-05-19 17:10:37 -07001776 security feature reduces the predictability of the kernel slab
1777 allocator against heap overflows.
1778
Kees Cook2482ddec2017-09-06 16:19:18 -07001779config SLAB_FREELIST_HARDENED
1780 bool "Harden slab freelist metadata"
1781 depends on SLUB
1782 help
1783 Many kernel heap attacks try to target slab cache metadata and
1784 other infrastructure. This options makes minor performance
1785 sacrifies to harden the kernel slab allocator against common
1786 freelist exploit methods.
1787
Dan Williamse900a912019-05-14 15:41:28 -07001788config SHUFFLE_PAGE_ALLOCATOR
1789 bool "Page allocator randomization"
1790 default SLAB_FREELIST_RANDOM && ACPI_NUMA
1791 help
1792 Randomization of the page allocator improves the average
1793 utilization of a direct-mapped memory-side-cache. See section
1794 5.2.27 Heterogeneous Memory Attribute Table (HMAT) in the ACPI
1795 6.2a specification for an example of how a platform advertises
1796 the presence of a memory-side-cache. There are also incidental
1797 security benefits as it reduces the predictability of page
1798 allocations to compliment SLAB_FREELIST_RANDOM, but the
1799 default granularity of shuffling on the "MAX_ORDER - 1" i.e,
1800 10th order of pages is selected based on cache utilization
1801 benefits on x86.
1802
1803 While the randomization improves cache utilization it may
1804 negatively impact workloads on platforms without a cache. For
1805 this reason, by default, the randomization is enabled only
1806 after runtime detection of a direct-mapped memory-side-cache.
1807 Otherwise, the randomization may be force enabled with the
1808 'page_alloc.shuffle' kernel command line parameter.
1809
1810 Say Y if unsure.
1811
Joonsoo Kim345c9052013-06-19 14:05:52 +09001812config SLUB_CPU_PARTIAL
1813 default y
Uwe Kleine-Königb39ffbf2013-07-17 16:54:59 +02001814 depends on SLUB && SMP
Joonsoo Kim345c9052013-06-19 14:05:52 +09001815 bool "SLUB per cpu partial cache"
1816 help
1817 Per cpu partial caches accellerate objects allocation and freeing
1818 that is local to a processor at the price of more indeterminism
1819 in the latency of the free. On overflow these caches will be cleared
1820 which requires the taking of locks that may cause latency spikes.
1821 Typically one would choose no for a realtime system.
1822
Jie Zhangea637632009-12-14 18:00:02 -08001823config MMAP_ALLOW_UNINITIALIZED
1824 bool "Allow mmapped anonymous memory to be uninitialized"
David Rientjes6a108a12011-01-20 14:44:16 -08001825 depends on EXPERT && !MMU
Jie Zhangea637632009-12-14 18:00:02 -08001826 default n
1827 help
1828 Normally, and according to the Linux spec, anonymous memory obtained
Randy Dunlap3903bf92018-08-21 21:58:34 -07001829 from mmap() has its contents cleared before it is passed to
Jie Zhangea637632009-12-14 18:00:02 -08001830 userspace. Enabling this config option allows you to request that
1831 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1832 providing a huge performance boost. If this option is not enabled,
1833 then the flag will be ignored.
1834
1835 This is taken advantage of by uClibc's malloc(), and also by
1836 ELF-FDPIC binfmt's brk and stack allocator.
1837
1838 Because of the obvious security issues, this option should only be
1839 enabled on embedded devices where you control what is run in
1840 userspace. Since that isn't generally a problem on no-MMU systems,
1841 it is normally safe to say Y here.
1842
1843 See Documentation/nommu-mmap.txt for more information.
1844
David Howells091f6e22015-07-20 21:16:28 +01001845config SYSTEM_DATA_VERIFICATION
1846 def_bool n
1847 select SYSTEM_TRUSTED_KEYRING
1848 select KEYS
1849 select CRYPTO
David Howellsd43de6c2016-03-03 21:49:27 +00001850 select CRYPTO_RSA
David Howells091f6e22015-07-20 21:16:28 +01001851 select ASYMMETRIC_KEY_TYPE
1852 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
David Howells091f6e22015-07-20 21:16:28 +01001853 select ASN1
1854 select OID_REGISTRY
1855 select X509_CERTIFICATE_PARSER
1856 select PKCS7_MESSAGE_PARSER
Peter Foley82c04ff2014-04-18 15:07:11 -07001857 help
David Howells091f6e22015-07-20 21:16:28 +01001858 Provide PKCS#7 message verification using the contents of the system
1859 trusted keyring to provide public keys. This then can be used for
1860 module verification, kexec image verification and firmware blob
1861 verification.
Peter Foley82c04ff2014-04-18 15:07:11 -07001862
Mathieu Desnoyers125e5642008-02-02 15:10:36 -05001863config PROFILING
Robert Richterb309a292010-02-26 15:01:23 +01001864 bool "Profiling support"
Mathieu Desnoyers125e5642008-02-02 15:10:36 -05001865 help
1866 Say Y here to enable the extended profiling support mechanisms used
1867 by profilers such as OProfile.
1868
Ingo Molnar5f87f112008-07-23 14:15:22 +02001869#
1870# Place an empty function call at each tracepoint site. Can be
1871# dynamically changed for a probe function.
1872#
Mathieu Desnoyers97e1c182008-07-18 12:16:16 -04001873config TRACEPOINTS
Ingo Molnar5f87f112008-07-23 14:15:22 +02001874 bool
Mathieu Desnoyers97e1c182008-07-18 12:16:16 -04001875
Linus Torvalds1da177e2005-04-16 15:20:36 -07001876endmenu # General setup
1877
Christoph Hellwig15724972018-07-31 13:39:30 +02001878source "arch/Kconfig"
1879
Chuck Ebbertae81f9e2006-09-16 12:15:53 -07001880config RT_MUTEXES
Christoph Jaeger6341e622014-12-20 15:41:11 -05001881 bool
Chuck Ebbertae81f9e2006-09-16 12:15:53 -07001882
Linus Torvalds1da177e2005-04-16 15:20:36 -07001883config BASE_SMALL
1884 int
1885 default 0 if BASE_FULL
1886 default 1 if !BASE_FULL
1887
Jan Engelhardt66da5732007-07-15 23:39:29 -07001888menuconfig MODULES
Linus Torvalds1da177e2005-04-16 15:20:36 -07001889 bool "Enable loadable module support"
Yann E. MORIN11097a02013-08-11 16:07:50 +02001890 option modules
Linus Torvalds1da177e2005-04-16 15:20:36 -07001891 help
1892 Kernel modules are small pieces of compiled code which can
1893 be inserted in the running kernel, rather than being
1894 permanently built into the kernel. You use the "modprobe"
1895 tool to add (and sometimes remove) them. If you say Y here,
1896 many parts of the kernel can be built as modules (by
1897 answering M instead of Y where indicated): this is most
1898 useful for infrequently used options which are not required
1899 for booting. For more information, see the man pages for
1900 modprobe, lsmod, modinfo, insmod and rmmod.
1901
1902 If you say Y here, you will need to run "make
1903 modules_install" to put the modules under /lib/modules/
1904 where modprobe can find them (you may need to be root to do
1905 this).
1906
1907 If unsure, say Y.
1908
Robert P. J. Day0b0de142008-08-04 13:31:32 -04001909if MODULES
1910
Linus Torvalds826e4502008-05-04 17:04:16 -07001911config MODULE_FORCE_LOAD
1912 bool "Forced module loading"
Linus Torvalds826e4502008-05-04 17:04:16 -07001913 default n
1914 help
Rusty Russell91e37a72008-05-09 16:25:28 +10001915 Allow loading of modules without version information (ie. modprobe
1916 --force). Forced module loading sets the 'F' (forced) taint flag and
1917 is usually a really bad idea.
Linus Torvalds826e4502008-05-04 17:04:16 -07001918
Linus Torvalds1da177e2005-04-16 15:20:36 -07001919config MODULE_UNLOAD
1920 bool "Module unloading"
Linus Torvalds1da177e2005-04-16 15:20:36 -07001921 help
1922 Without this option you will not be able to unload any
1923 modules (note that some modules may not be unloadable
Denys Vlasenkof7f5b672008-07-22 19:24:26 -05001924 anyway), which makes your kernel smaller, faster
1925 and simpler. If unsure, say Y.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001926
1927config MODULE_FORCE_UNLOAD
1928 bool "Forced module unloading"
Kees Cook19c92392012-10-02 11:19:29 -07001929 depends on MODULE_UNLOAD
Linus Torvalds1da177e2005-04-16 15:20:36 -07001930 help
1931 This option allows you to force a module to unload, even if the
1932 kernel believes it is unsafe: the kernel will remove the module
1933 without waiting for anyone to stop using it (using the -f option to
1934 rmmod). This is mainly for kernel developers and desperate users.
1935 If unsure, say N.
1936
Linus Torvalds1da177e2005-04-16 15:20:36 -07001937config MODVERSIONS
Sam Ravnborg0d541642005-12-26 23:04:02 +01001938 bool "Module versioning support"
Linus Torvalds1da177e2005-04-16 15:20:36 -07001939 help
1940 Usually, you have to use modules compiled with your kernel.
1941 Saying Y here makes it sometimes possible to use modules
1942 compiled for different kernels, by adding enough information
1943 to the modules to (hopefully) spot any changes which would
1944 make them incompatible with the kernel you are running. If
1945 unsure, say N.
1946
Ard Biesheuvel56067812017-02-03 09:54:05 +00001947config MODULE_REL_CRCS
1948 bool
1949 depends on MODVERSIONS
1950
Linus Torvalds1da177e2005-04-16 15:20:36 -07001951config MODULE_SRCVERSION_ALL
1952 bool "Source checksum for all modules"
Linus Torvalds1da177e2005-04-16 15:20:36 -07001953 help
1954 Modules which contain a MODULE_VERSION get an extra "srcversion"
1955 field inserted into their modinfo section, which contains a
1956 sum of the source files which made it. This helps maintainers
1957 see exactly which source was used to build a module (since
1958 others sometimes change the module source without updating
1959 the version). With this option, such a "srcversion" field
1960 will be created for all modules. If unsure, say N.
1961
Rusty Russell106a4ee2012-09-26 10:09:40 +01001962config MODULE_SIG
1963 bool "Module signature verification"
1964 depends on MODULES
David Howells091f6e22015-07-20 21:16:28 +01001965 select SYSTEM_DATA_VERIFICATION
Rusty Russell106a4ee2012-09-26 10:09:40 +01001966 help
1967 Check modules for valid signatures upon load: the signature
1968 is simply appended to the module. For more information see
Nathan Chancellorcbdc8212017-09-10 02:48:29 -07001969 <file:Documentation/admin-guide/module-signing.rst>.
Rusty Russell106a4ee2012-09-26 10:09:40 +01001970
David Howells228c37f2015-08-11 12:38:54 +01001971 Note that this option adds the OpenSSL development packages as a
1972 kernel build dependency so that the signing tool can use its crypto
1973 library.
1974
David Howellsea0b6dc2012-09-26 10:09:50 +01001975 !!!WARNING!!! If you enable this option, you MUST make sure that the
1976 module DOES NOT get stripped after being signed. This includes the
1977 debuginfo strip done by some packagers (such as rpmbuild) and
1978 inclusion into an initramfs that wants the module size reduced.
1979
Rusty Russell106a4ee2012-09-26 10:09:40 +01001980config MODULE_SIG_FORCE
1981 bool "Require modules to be validly signed"
1982 depends on MODULE_SIG
1983 help
1984 Reject unsigned modules or signed modules for which we don't have a
1985 key. Without this, such modules will simply taint the kernel.
David Howellsea0b6dc2012-09-26 10:09:50 +01001986
Michal Marekd9d8d7e2013-01-25 13:41:31 +10301987config MODULE_SIG_ALL
1988 bool "Automatically sign all modules"
1989 default y
1990 depends on MODULE_SIG
1991 help
1992 Sign all modules during make modules_install. Without this option,
1993 modules must be signed manually, using the scripts/sign-file tool.
1994
1995comment "Do not forget to sign required modules with scripts/sign-file"
1996 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL
1997
David Howellsea0b6dc2012-09-26 10:09:50 +01001998choice
1999 prompt "Which hash algorithm should modules be signed with?"
2000 depends on MODULE_SIG
2001 help
2002 This determines which sort of hashing algorithm will be used during
2003 signature generation. This algorithm _must_ be built into the kernel
2004 directly so that signature verification can take place. It is not
2005 possible to load a signed module containing the algorithm to check
2006 the signature on that module.
2007
2008config MODULE_SIG_SHA1
2009 bool "Sign modules with SHA-1"
2010 select CRYPTO_SHA1
2011
2012config MODULE_SIG_SHA224
2013 bool "Sign modules with SHA-224"
2014 select CRYPTO_SHA256
2015
2016config MODULE_SIG_SHA256
2017 bool "Sign modules with SHA-256"
2018 select CRYPTO_SHA256
2019
2020config MODULE_SIG_SHA384
2021 bool "Sign modules with SHA-384"
2022 select CRYPTO_SHA512
2023
2024config MODULE_SIG_SHA512
2025 bool "Sign modules with SHA-512"
2026 select CRYPTO_SHA512
2027
2028endchoice
2029
Michal Marek22753672013-01-25 13:41:00 +10302030config MODULE_SIG_HASH
2031 string
2032 depends on MODULE_SIG
2033 default "sha1" if MODULE_SIG_SHA1
2034 default "sha224" if MODULE_SIG_SHA224
2035 default "sha256" if MODULE_SIG_SHA256
2036 default "sha384" if MODULE_SIG_SHA384
2037 default "sha512" if MODULE_SIG_SHA512
2038
Bertrand Jacquinbeb50df2014-08-27 20:31:56 +09302039config MODULE_COMPRESS
2040 bool "Compress modules on installation"
2041 depends on MODULES
2042 help
Bertrand Jacquinbeb50df2014-08-27 20:31:56 +09302043
Rusty Russellb6c09b52015-06-16 12:16:22 +09302044 Compresses kernel modules when 'make modules_install' is run; gzip or
2045 xz depending on "Compression algorithm" below.
Bertrand Jacquinbeb50df2014-08-27 20:31:56 +09302046
Rusty Russellb6c09b52015-06-16 12:16:22 +09302047 module-init-tools MAY support gzip, and kmod MAY support gzip and xz.
Bertrand Jacquinbeb50df2014-08-27 20:31:56 +09302048
Rusty Russellb6c09b52015-06-16 12:16:22 +09302049 Out-of-tree kernel modules installed using Kbuild will also be
2050 compressed upon installation.
Bertrand Jacquinbeb50df2014-08-27 20:31:56 +09302051
Rusty Russellb6c09b52015-06-16 12:16:22 +09302052 Note: for modules inside an initrd or initramfs, it's more efficient
2053 to compress the whole initrd or initramfs instead.
Bertrand Jacquinbeb50df2014-08-27 20:31:56 +09302054
Rusty Russellb6c09b52015-06-16 12:16:22 +09302055 Note: This is fully compatible with signed modules.
2056
2057 If in doubt, say N.
Bertrand Jacquinbeb50df2014-08-27 20:31:56 +09302058
2059choice
2060 prompt "Compression algorithm"
2061 depends on MODULE_COMPRESS
2062 default MODULE_COMPRESS_GZIP
2063 help
2064 This determines which sort of compression will be used during
2065 'make modules_install'.
2066
2067 GZIP (default) and XZ are supported.
2068
2069config MODULE_COMPRESS_GZIP
2070 bool "GZIP"
2071
2072config MODULE_COMPRESS_XZ
2073 bool "XZ"
2074
2075endchoice
2076
Nicolas Pitredbacb0e2016-01-26 21:51:05 -05002077config TRIM_UNUSED_KSYMS
2078 bool "Trim unused exported kernel symbols"
2079 depends on MODULES && !UNUSED_SYMBOLS
2080 help
2081 The kernel and some modules make many symbols available for
2082 other modules to use via EXPORT_SYMBOL() and variants. Depending
2083 on the set of modules being selected in your kernel configuration,
2084 many of those exported symbols might never be used.
2085
2086 This option allows for unused exported symbols to be dropped from
2087 the build. In turn, this provides the compiler more opportunities
2088 (especially when using LTO) for optimizing the code and reducing
2089 binary size. This might have some security advantages as well.
2090
Valdis Kletnieksf1cb6372016-08-02 14:07:27 -07002091 If unsure, or if you need to build out-of-tree modules, say N.
Nicolas Pitredbacb0e2016-01-26 21:51:05 -05002092
Robert P. J. Day0b0de142008-08-04 13:31:32 -04002093endif # MODULES
2094
Peter Zijlstra6c9692e2015-05-27 11:09:37 +09302095config MODULES_TREE_LOOKUP
2096 def_bool y
2097 depends on PERF_EVENTS || TRACING
2098
Rusty Russell98a79d62008-12-13 21:19:41 +10302099config INIT_ALL_POSSIBLE
2100 bool
2101 help
Rusty Russell5f054e32012-03-29 15:38:31 +10302102 Back when each arch used to define their own cpu_online_mask and
2103 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
Rusty Russell98a79d62008-12-13 21:19:41 +10302104 with all 1s, and others with all 0s. When they were centralised,
2105 it was better to provide this option than to break all the archs
Matt LaPlante692105b2009-01-26 11:12:25 +01002106 and have several arch maintainers pursuing me down dark alleys.
Rusty Russell98a79d62008-12-13 21:19:41 +10302107
Jens Axboe3a65dfe2005-11-04 08:43:35 +01002108source "block/Kconfig"
Avi Kivitye98c3202007-10-16 23:27:31 -07002109
2110config PREEMPT_NOTIFIERS
2111 bool
Paul E. McKenneye260be62008-01-25 21:08:24 +01002112
Steffen Klassert16295be2010-01-06 19:47:10 +11002113config PADATA
2114 depends on SMP
2115 bool
2116
David Howells4520c6a2012-09-21 23:31:13 +01002117config ASN1
2118 tristate
2119 help
2120 Build a simple ASN.1 grammar compiler that produces a bytecode output
2121 that can be interpreted by the ASN.1 stream decoder and used to
2122 inform it as to what tags are to be expected in a stream and what
2123 functions to call on what tags.
2124
Thomas Gleixner6beb0002009-11-09 15:21:34 +00002125source "kernel/Kconfig.locks"
Mathieu Desnoyerse61938a2018-01-29 15:20:15 -05002126
2127config ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
2128 bool
Dominik Brodowski1bd21c62018-04-05 11:53:01 +02002129
2130# It may be useful for an architecture to override the definitions of the
Dominik Brodowski7303e302018-04-05 11:53:03 +02002131# SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h>
2132# and the COMPAT_ variants in <linux/compat.h>, in particular to use a
2133# different calling convention for syscalls. They can also override the
2134# macros for not-implemented syscalls in kernel/sys_ni.c and
2135# kernel/time/posix-stubs.c. All these overrides need to be available in
2136# <asm/syscall_wrapper.h>.
Dominik Brodowski1bd21c62018-04-05 11:53:01 +02002137config ARCH_HAS_SYSCALL_WRAPPER
2138 def_bool n