1ec8f24b7SThomas Gleixner# SPDX-License-Identifier: GPL-2.0-only 2face4374SRoman Zippelconfig DEFCONFIG_LIST 3face4374SRoman Zippel string 4b2670eacSPaolo 'Blaisorblade' Giarrusso depends on !UML 5face4374SRoman Zippel option defconfig_list 647f38ae0SRob Landley default "/lib/modules/$(shell,uname -r)/.config" 7face4374SRoman Zippel default "/etc/kernel-config" 847f38ae0SRob Landley default "/boot/config-$(shell,uname -r)" 92a86f661SMasahiro Yamada default "arch/$(SRCARCH)/configs/$(KBUILD_DEFCONFIG)" 10face4374SRoman Zippel 11a4353898SMasahiro Yamadaconfig CC_IS_GCC 12a4353898SMasahiro Yamada def_bool $(success,$(CC) --version | head -n 1 | grep -q gcc) 13a4353898SMasahiro Yamada 14a4353898SMasahiro Yamadaconfig GCC_VERSION 15a4353898SMasahiro Yamada int 16fa7295abSMasahiro Yamada default $(shell,$(srctree)/scripts/gcc-version.sh $(CC)) if CC_IS_GCC 17a4353898SMasahiro Yamada default 0 18a4353898SMasahiro Yamada 199553d16fSAmit Daniel Kachhapconfig LD_VERSION 209553d16fSAmit Daniel Kachhap int 219553d16fSAmit Daniel Kachhap default $(shell,$(LD) --version | $(srctree)/scripts/ld-version.sh) 229553d16fSAmit Daniel Kachhap 23469cb737SMasahiro Yamadaconfig CC_IS_CLANG 24469cb737SMasahiro Yamada def_bool $(success,$(CC) --version | head -n 1 | grep -q clang) 25469cb737SMasahiro Yamada 26469cb737SMasahiro Yamadaconfig CLANG_VERSION 27469cb737SMasahiro Yamada int 28469cb737SMasahiro Yamada default $(shell,$(srctree)/scripts/clang-version.sh $(CC)) 29469cb737SMasahiro Yamada 301a927fd3SMasahiro Yamadaconfig CC_CAN_LINK 311a927fd3SMasahiro Yamada def_bool $(success,$(srctree)/scripts/cc-can-link.sh $(CC)) 321a927fd3SMasahiro Yamada 33e9666d10SMasahiro Yamadaconfig CC_HAS_ASM_GOTO 34e9666d10SMasahiro Yamada def_bool $(success,$(srctree)/scripts/gcc-goto.sh $(CC)) 35e9666d10SMasahiro Yamada 365cf896fbSPeter Collingbourneconfig TOOLS_SUPPORT_RELR 372d122942SWill Deacon def_bool $(success,env "CC=$(CC)" "LD=$(LD)" "NM=$(NM)" "OBJCOPY=$(OBJCOPY)" $(srctree)/scripts/tools-support-relr.sh) 385cf896fbSPeter Collingbourne 39eb111869SRasmus Villemoesconfig CC_HAS_ASM_INLINE 40eb111869SRasmus Villemoes def_bool $(success,echo 'void foo(void) { asm inline (""); }' | $(CC) -x c - -c -o /dev/null) 41eb111869SRasmus Villemoes 42b99b87f7SPeter Oberparleiterconfig CONSTRUCTORS 43b99b87f7SPeter Oberparleiter bool 4487c9366eSJohannes Berg depends on !UML 45b99b87f7SPeter Oberparleiter 46e360adbeSPeter Zijlstraconfig IRQ_WORK 47e360adbeSPeter Zijlstra bool 48e360adbeSPeter Zijlstra 4910916706SShile Zhangconfig BUILDTIME_TABLE_SORT 501dbdc6f1SDavid Daney bool 511dbdc6f1SDavid Daney 52c65eacbeSAndy Lutomirskiconfig THREAD_INFO_IN_TASK 53c65eacbeSAndy Lutomirski bool 54c65eacbeSAndy Lutomirski help 55c65eacbeSAndy Lutomirski Select this to move thread_info off the stack into task_struct. To 56c65eacbeSAndy Lutomirski make this work, an arch will need to remove all thread_info fields 57c65eacbeSAndy Lutomirski except flags and fix any runtime bugs. 58c65eacbeSAndy Lutomirski 59c6c314a6SAndy Lutomirski One subtle change that will be needed is to use try_get_task_stack() 60c6c314a6SAndy Lutomirski and put_task_stack() in save_thread_stack_tsk() and get_wchan(). 61c6c314a6SAndy Lutomirski 62ff0cfc66SAl Boldimenu "General setup" 631da177e4SLinus Torvalds 641da177e4SLinus Torvaldsconfig BROKEN 651da177e4SLinus Torvalds bool 661da177e4SLinus Torvalds 671da177e4SLinus Torvaldsconfig BROKEN_ON_SMP 681da177e4SLinus Torvalds bool 691da177e4SLinus Torvalds depends on BROKEN || !SMP 701da177e4SLinus Torvalds default y 711da177e4SLinus Torvalds 721da177e4SLinus Torvaldsconfig INIT_ENV_ARG_LIMIT 731da177e4SLinus Torvalds int 74dd673bcaSAdrian Bunk default 32 if !UML 75dd673bcaSAdrian Bunk default 128 if UML 761da177e4SLinus Torvalds help 7734ad92c2SRandy Dunlap Maximum of each of the number of arguments and environment 7834ad92c2SRandy Dunlap variables passed to init from the kernel command line. 791da177e4SLinus Torvalds 804bb16672SJiri Slabyconfig COMPILE_TEST 814bb16672SJiri Slaby bool "Compile also drivers which will not load" 82bc083a64SRichard Weinberger depends on !UML 834bb16672SJiri Slaby default n 844bb16672SJiri Slaby help 854bb16672SJiri Slaby Some drivers can be compiled on a different platform than they are 864bb16672SJiri Slaby intended to be run on. Despite they cannot be loaded there (or even 874bb16672SJiri Slaby when they load they cannot be used due to missing HW support), 884bb16672SJiri Slaby developers still, opposing to distributors, might want to build such 894bb16672SJiri Slaby drivers to compile-test them. 904bb16672SJiri Slaby 914bb16672SJiri Slaby If you are a developer and want to build everything available, say Y 924bb16672SJiri Slaby here. If you are a user/distributor, say N here to exclude useless 934bb16672SJiri Slaby drivers to be distributed. 944bb16672SJiri Slaby 95d6fc9fcbSMasahiro Yamadaconfig UAPI_HEADER_TEST 96d6fc9fcbSMasahiro Yamada bool "Compile test UAPI headers" 97fcbb8461SMasahiro Yamada depends on HEADERS_INSTALL && CC_CAN_LINK 98d6fc9fcbSMasahiro Yamada help 99d6fc9fcbSMasahiro Yamada Compile test headers exported to user-space to ensure they are 100d6fc9fcbSMasahiro Yamada self-contained, i.e. compilable as standalone units. 101d6fc9fcbSMasahiro Yamada 102d6fc9fcbSMasahiro Yamada If you are a developer or tester and want to ensure the exported 103d6fc9fcbSMasahiro Yamada headers are self-contained, say Y here. Otherwise, choose N. 104d6fc9fcbSMasahiro Yamada 1051da177e4SLinus Torvaldsconfig LOCALVERSION 1061da177e4SLinus Torvalds string "Local version - append to kernel release" 1071da177e4SLinus Torvalds help 1081da177e4SLinus Torvalds Append an extra string to the end of your kernel version. 1091da177e4SLinus Torvalds This will show up when you type uname, for example. 1101da177e4SLinus Torvalds The string you set here will be appended after the contents of 1111da177e4SLinus Torvalds any files with a filename matching localversion* in your 1121da177e4SLinus Torvalds object and source tree, in that order. Your total string can 1131da177e4SLinus Torvalds be a maximum of 64 characters. 1141da177e4SLinus Torvalds 115aaebf433SRyan Andersonconfig LOCALVERSION_AUTO 116aaebf433SRyan Anderson bool "Automatically append version information to the version string" 117aaebf433SRyan Anderson default y 118ac3339baSAlexey Dobriyan depends on !COMPILE_TEST 119aaebf433SRyan Anderson help 120aaebf433SRyan Anderson This will try to automatically determine if the current tree is a 1216e5a5420SRobert P. J. Day release tree by looking for git tags that belong to the current 1226e5a5420SRobert P. J. Day top of tree revision. 123aaebf433SRyan Anderson 124aaebf433SRyan Anderson A string of the format -gxxxxxxxx will be added to the localversion 1256e5a5420SRobert P. J. Day if a git-based tree is found. The string generated by this will be 126aaebf433SRyan Anderson appended after any matching localversion* files, and after the value 1276e5a5420SRobert P. J. Day set in CONFIG_LOCALVERSION. 128aaebf433SRyan Anderson 1296e5a5420SRobert P. J. Day (The actual string used here is the first eight characters produced 1306e5a5420SRobert P. J. Day by running the command: 1316e5a5420SRobert P. J. Day 1326e5a5420SRobert P. J. Day $ git rev-parse --verify HEAD 1336e5a5420SRobert P. J. Day 1346e5a5420SRobert P. J. Day which is done within the script "scripts/setlocalversion".) 135aaebf433SRyan Anderson 1369afb719eSLaura Abbottconfig BUILD_SALT 1379afb719eSLaura Abbott string "Build ID Salt" 1389afb719eSLaura Abbott default "" 1399afb719eSLaura Abbott help 1409afb719eSLaura Abbott The build ID is used to link binaries and their debug info. Setting 1419afb719eSLaura Abbott this option will use the value in the calculation of the build id. 1429afb719eSLaura Abbott This is mostly useful for distributions which want to ensure the 1439afb719eSLaura Abbott build is unique between builds. It's safe to leave the default. 1449afb719eSLaura Abbott 1452e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_GZIP 1462e9f3bddSH. Peter Anvin bool 1472e9f3bddSH. Peter Anvin 1482e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_BZIP2 1492e9f3bddSH. Peter Anvin bool 1502e9f3bddSH. Peter Anvin 1512e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_LZMA 1522e9f3bddSH. Peter Anvin bool 1532e9f3bddSH. Peter Anvin 1543ebe1243SLasse Collinconfig HAVE_KERNEL_XZ 1553ebe1243SLasse Collin bool 1563ebe1243SLasse Collin 1577dd65febSAlbin Tonnerreconfig HAVE_KERNEL_LZO 1587dd65febSAlbin Tonnerre bool 1597dd65febSAlbin Tonnerre 160e76e1fdfSKyungsik Leeconfig HAVE_KERNEL_LZ4 161e76e1fdfSKyungsik Lee bool 162e76e1fdfSKyungsik Lee 163f16466afSVasily Gorbikconfig HAVE_KERNEL_UNCOMPRESSED 164f16466afSVasily Gorbik bool 165f16466afSVasily Gorbik 16630d65dbfSAlain Knaffchoice 16730d65dbfSAlain Knaff prompt "Kernel compression mode" 16830d65dbfSAlain Knaff default KERNEL_GZIP 169f16466afSVasily Gorbik depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_UNCOMPRESSED 17030d65dbfSAlain Knaff help 17130d65dbfSAlain Knaff The linux kernel is a kind of self-extracting executable. 17230d65dbfSAlain Knaff Several compression algorithms are available, which differ 17330d65dbfSAlain Knaff in efficiency, compression and decompression speed. 17430d65dbfSAlain Knaff Compression speed is only relevant when building a kernel. 17530d65dbfSAlain Knaff Decompression speed is relevant at each boot. 17630d65dbfSAlain Knaff 17730d65dbfSAlain Knaff If you have any problems with bzip2 or lzma compressed 17830d65dbfSAlain Knaff kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older 17930d65dbfSAlain Knaff version of this functionality (bzip2 only), for 2.4, was 18030d65dbfSAlain Knaff supplied by Christian Ludwig) 18130d65dbfSAlain Knaff 18230d65dbfSAlain Knaff High compression options are mostly useful for users, who 18330d65dbfSAlain Knaff are low on disk space (embedded systems), but for whom ram 18430d65dbfSAlain Knaff size matters less. 18530d65dbfSAlain Knaff 18630d65dbfSAlain Knaff If in doubt, select 'gzip' 18730d65dbfSAlain Knaff 18830d65dbfSAlain Knaffconfig KERNEL_GZIP 18930d65dbfSAlain Knaff bool "Gzip" 1902e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_GZIP 19130d65dbfSAlain Knaff help 1927dd65febSAlbin Tonnerre The old and tried gzip compression. It provides a good balance 1937dd65febSAlbin Tonnerre between compression ratio and decompression speed. 19430d65dbfSAlain Knaff 19530d65dbfSAlain Knaffconfig KERNEL_BZIP2 19630d65dbfSAlain Knaff bool "Bzip2" 1972e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_BZIP2 19830d65dbfSAlain Knaff help 19930d65dbfSAlain Knaff Its compression ratio and speed is intermediate. 2000a4dd35cSRandy Dunlap Decompression speed is slowest among the choices. The kernel 2012e9f3bddSH. Peter Anvin size is about 10% smaller with bzip2, in comparison to gzip. 2022e9f3bddSH. Peter Anvin Bzip2 uses a large amount of memory. For modern kernels you 2032e9f3bddSH. Peter Anvin will need at least 8MB RAM or more for booting. 20430d65dbfSAlain Knaff 20530d65dbfSAlain Knaffconfig KERNEL_LZMA 20630d65dbfSAlain Knaff bool "LZMA" 2072e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_LZMA 20830d65dbfSAlain Knaff help 2090a4dd35cSRandy Dunlap This compression algorithm's ratio is best. Decompression speed 2100a4dd35cSRandy Dunlap is between gzip and bzip2. Compression is slowest. 2110a4dd35cSRandy Dunlap The kernel size is about 33% smaller with LZMA in comparison to gzip. 21230d65dbfSAlain Knaff 2133ebe1243SLasse Collinconfig KERNEL_XZ 2143ebe1243SLasse Collin bool "XZ" 2153ebe1243SLasse Collin depends on HAVE_KERNEL_XZ 2163ebe1243SLasse Collin help 2173ebe1243SLasse Collin XZ uses the LZMA2 algorithm and instruction set specific 2183ebe1243SLasse Collin BCJ filters which can improve compression ratio of executable 2193ebe1243SLasse Collin code. The size of the kernel is about 30% smaller with XZ in 2203ebe1243SLasse Collin comparison to gzip. On architectures for which there is a BCJ 2213ebe1243SLasse Collin filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ 2223ebe1243SLasse Collin will create a few percent smaller kernel than plain LZMA. 2233ebe1243SLasse Collin 2243ebe1243SLasse Collin The speed is about the same as with LZMA: The decompression 2253ebe1243SLasse Collin speed of XZ is better than that of bzip2 but worse than gzip 2263ebe1243SLasse Collin and LZO. Compression is slow. 2273ebe1243SLasse Collin 2287dd65febSAlbin Tonnerreconfig KERNEL_LZO 2297dd65febSAlbin Tonnerre bool "LZO" 2307dd65febSAlbin Tonnerre depends on HAVE_KERNEL_LZO 2317dd65febSAlbin Tonnerre help 2320a4dd35cSRandy Dunlap Its compression ratio is the poorest among the choices. The kernel 233681b3049SStephan Sperber size is about 10% bigger than gzip; however its speed 2347dd65febSAlbin Tonnerre (both compression and decompression) is the fastest. 2357dd65febSAlbin Tonnerre 236e76e1fdfSKyungsik Leeconfig KERNEL_LZ4 237e76e1fdfSKyungsik Lee bool "LZ4" 238e76e1fdfSKyungsik Lee depends on HAVE_KERNEL_LZ4 239e76e1fdfSKyungsik Lee help 240e76e1fdfSKyungsik Lee LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding. 241e76e1fdfSKyungsik Lee A preliminary version of LZ4 de/compression tool is available at 242e76e1fdfSKyungsik Lee <https://code.google.com/p/lz4/>. 243e76e1fdfSKyungsik Lee 244e76e1fdfSKyungsik Lee Its compression ratio is worse than LZO. The size of the kernel 245e76e1fdfSKyungsik Lee is about 8% bigger than LZO. But the decompression speed is 246e76e1fdfSKyungsik Lee faster than LZO. 247e76e1fdfSKyungsik Lee 248f16466afSVasily Gorbikconfig KERNEL_UNCOMPRESSED 249f16466afSVasily Gorbik bool "None" 250f16466afSVasily Gorbik depends on HAVE_KERNEL_UNCOMPRESSED 251f16466afSVasily Gorbik help 252f16466afSVasily Gorbik Produce uncompressed kernel image. This option is usually not what 253f16466afSVasily Gorbik you want. It is useful for debugging the kernel in slow simulation 254f16466afSVasily Gorbik environments, where decompressing and moving the kernel is awfully 255f16466afSVasily Gorbik slow. This option allows early boot code to skip the decompressor 256f16466afSVasily Gorbik and jump right at uncompressed kernel image. 257f16466afSVasily Gorbik 25830d65dbfSAlain Knaffendchoice 25930d65dbfSAlain Knaff 260bd5dc17bSJosh Triplettconfig DEFAULT_HOSTNAME 261bd5dc17bSJosh Triplett string "Default hostname" 262bd5dc17bSJosh Triplett default "(none)" 263bd5dc17bSJosh Triplett help 264bd5dc17bSJosh Triplett This option determines the default system hostname before userspace 265bd5dc17bSJosh Triplett calls sethostname(2). The kernel traditionally uses "(none)" here, 266bd5dc17bSJosh Triplett but you may wish to use a different default here to make a minimal 267bd5dc17bSJosh Triplett system more usable with less configuration. 268bd5dc17bSJosh Triplett 26917c46a6aSChristoph Hellwig# 27017c46a6aSChristoph Hellwig# For some reason microblaze and nios2 hard code SWAP=n. Hopefully we can 27117c46a6aSChristoph Hellwig# add proper SWAP support to them, in which case this can be remove. 27217c46a6aSChristoph Hellwig# 27317c46a6aSChristoph Hellwigconfig ARCH_NO_SWAP 27417c46a6aSChristoph Hellwig bool 27517c46a6aSChristoph Hellwig 2761da177e4SLinus Torvaldsconfig SWAP 2771da177e4SLinus Torvalds bool "Support for paging of anonymous memory (swap)" 27817c46a6aSChristoph Hellwig depends on MMU && BLOCK && !ARCH_NO_SWAP 2791da177e4SLinus Torvalds default y 2801da177e4SLinus Torvalds help 2811da177e4SLinus Torvalds This option allows you to choose whether you want to have support 2821da177e4SLinus Torvalds for so called swap devices or swap files in your kernel that are 2831da177e4SLinus Torvalds used to provide more virtual memory than the actual RAM present 2841da177e4SLinus Torvalds in your computer. If unsure say Y. 2851da177e4SLinus Torvalds 2861da177e4SLinus Torvaldsconfig SYSVIPC 2871da177e4SLinus Torvalds bool "System V IPC" 2881da177e4SLinus Torvalds ---help--- 2891da177e4SLinus Torvalds Inter Process Communication is a suite of library functions and 2901da177e4SLinus Torvalds system calls which let processes (running programs) synchronize and 2911da177e4SLinus Torvalds exchange information. It is generally considered to be a good thing, 2921da177e4SLinus Torvalds and some programs won't run unless you say Y here. In particular, if 2931da177e4SLinus Torvalds you want to run the DOS emulator dosemu under Linux (read the 2941da177e4SLinus Torvalds DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>), 2951da177e4SLinus Torvalds you'll need to say Y here. 2961da177e4SLinus Torvalds 2971da177e4SLinus Torvalds You can find documentation about IPC with "info ipc" and also in 2981da177e4SLinus Torvalds section 6.4 of the Linux Programmer's Guide, available from 2991da177e4SLinus Torvalds <http://www.tldp.org/guides.html>. 3001da177e4SLinus Torvalds 301a5494dcdSEric W. Biedermanconfig SYSVIPC_SYSCTL 302a5494dcdSEric W. Biederman bool 303a5494dcdSEric W. Biederman depends on SYSVIPC 304a5494dcdSEric W. Biederman depends on SYSCTL 305a5494dcdSEric W. Biederman default y 306a5494dcdSEric W. Biederman 3071da177e4SLinus Torvaldsconfig POSIX_MQUEUE 3081da177e4SLinus Torvalds bool "POSIX Message Queues" 30919c92399SKees Cook depends on NET 3101da177e4SLinus Torvalds ---help--- 3111da177e4SLinus Torvalds POSIX variant of message queues is a part of IPC. In POSIX message 3121da177e4SLinus Torvalds queues every message has a priority which decides about succession 3131da177e4SLinus Torvalds of receiving it by a process. If you want to compile and run 3141da177e4SLinus Torvalds programs written e.g. for Solaris with use of its POSIX message 315b0e37650SRobert P. J. Day queues (functions mq_*) say Y here. 3161da177e4SLinus Torvalds 3171da177e4SLinus Torvalds POSIX message queues are visible as a filesystem called 'mqueue' 3181da177e4SLinus Torvalds and can be mounted somewhere if you want to do filesystem 3191da177e4SLinus Torvalds operations on message queues. 3201da177e4SLinus Torvalds 3211da177e4SLinus Torvalds If unsure, say Y. 3221da177e4SLinus Torvalds 323bdc8e5f8SSerge E. Hallynconfig POSIX_MQUEUE_SYSCTL 324bdc8e5f8SSerge E. Hallyn bool 325bdc8e5f8SSerge E. Hallyn depends on POSIX_MQUEUE 326bdc8e5f8SSerge E. Hallyn depends on SYSCTL 327bdc8e5f8SSerge E. Hallyn default y 328bdc8e5f8SSerge E. Hallyn 329*c73be61cSDavid Howellsconfig WATCH_QUEUE 330*c73be61cSDavid Howells bool "General notification queue" 331*c73be61cSDavid Howells default n 332*c73be61cSDavid Howells help 333*c73be61cSDavid Howells 334*c73be61cSDavid Howells This is a general notification queue for the kernel to pass events to 335*c73be61cSDavid Howells userspace by splicing them into pipes. It can be used in conjunction 336*c73be61cSDavid Howells with watches for key/keyring change notifications and device 337*c73be61cSDavid Howells notifications. 338*c73be61cSDavid Howells 339*c73be61cSDavid Howells See Documentation/watch_queue.rst 340*c73be61cSDavid Howells 341226b4ccdSKonstantin Khlebnikovconfig CROSS_MEMORY_ATTACH 342226b4ccdSKonstantin Khlebnikov bool "Enable process_vm_readv/writev syscalls" 343226b4ccdSKonstantin Khlebnikov depends on MMU 344226b4ccdSKonstantin Khlebnikov default y 345226b4ccdSKonstantin Khlebnikov help 346226b4ccdSKonstantin Khlebnikov Enabling this option adds the system calls process_vm_readv and 347226b4ccdSKonstantin Khlebnikov process_vm_writev which allow a process with the correct privileges 348a2a368d9SGeert Uytterhoeven to directly read from or write to another process' address space. 349226b4ccdSKonstantin Khlebnikov See the man page for more details. 350226b4ccdSKonstantin Khlebnikov 35169369a70SJosh Triplettconfig USELIB 35269369a70SJosh Triplett bool "uselib syscall" 353b2113a41SRiku Voipio def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION 35469369a70SJosh Triplett help 35569369a70SJosh Triplett This option enables the uselib syscall, a system call used in the 35669369a70SJosh Triplett dynamic linker from libc5 and earlier. glibc does not use this 35769369a70SJosh Triplett system call. If you intend to run programs built on libc5 or 35869369a70SJosh Triplett earlier, you may need to enable this syscall. Current systems 35969369a70SJosh Triplett running glibc can safely disable this. 36069369a70SJosh Triplett 3611da177e4SLinus Torvaldsconfig AUDIT 3621da177e4SLinus Torvalds bool "Auditing support" 363804a6a49SChris Wright depends on NET 3641da177e4SLinus Torvalds help 3651da177e4SLinus Torvalds Enable auditing infrastructure that can be used with another 3661da177e4SLinus Torvalds kernel subsystem, such as SELinux (which requires this for 367cb74ed27SPaul Moore logging of avc messages output). System call auditing is included 368cb74ed27SPaul Moore on architectures which support it. 3691da177e4SLinus Torvalds 3707a017721SAKASHI Takahiroconfig HAVE_ARCH_AUDITSYSCALL 3717a017721SAKASHI Takahiro bool 3727a017721SAKASHI Takahiro 3731da177e4SLinus Torvaldsconfig AUDITSYSCALL 374cb74ed27SPaul Moore def_bool y 3757a017721SAKASHI Takahiro depends on AUDIT && HAVE_ARCH_AUDITSYSCALL 37628a3a7ebSEric Paris select FSNOTIFY 37774c3cbe3SAl Viro 378d9817ebeSThomas Gleixnersource "kernel/irq/Kconfig" 379764e0da1SThomas Gleixnersource "kernel/time/Kconfig" 38087a4c375SChristoph Hellwigsource "kernel/Kconfig.preempt" 381d9817ebeSThomas Gleixner 382391dc69cSFrederic Weisbeckermenu "CPU/Task time and stats accounting" 383391dc69cSFrederic Weisbecker 384abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING 385abf917cdSFrederic Weisbecker bool 386abf917cdSFrederic Weisbecker 387fdf9c356SFrederic Weisbeckerchoice 388fdf9c356SFrederic Weisbecker prompt "Cputime accounting" 389fdf9c356SFrederic Weisbecker default TICK_CPU_ACCOUNTING if !PPC64 39002fc8d37SStephen Rothwell default VIRT_CPU_ACCOUNTING_NATIVE if PPC64 391fdf9c356SFrederic Weisbecker 392fdf9c356SFrederic Weisbecker# Kind of a stub config for the pure tick based cputime accounting 393fdf9c356SFrederic Weisbeckerconfig TICK_CPU_ACCOUNTING 394fdf9c356SFrederic Weisbecker bool "Simple tick based cputime accounting" 395c58b0df1SFrederic Weisbecker depends on !S390 && !NO_HZ_FULL 396fdf9c356SFrederic Weisbecker help 397fdf9c356SFrederic Weisbecker This is the basic tick based cputime accounting that maintains 398fdf9c356SFrederic Weisbecker statistics about user, system and idle time spent on per jiffies 399fdf9c356SFrederic Weisbecker granularity. 400fdf9c356SFrederic Weisbecker 401fdf9c356SFrederic Weisbecker If unsure, say Y. 402fdf9c356SFrederic Weisbecker 403abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING_NATIVE 404391dc69cSFrederic Weisbecker bool "Deterministic task and CPU time accounting" 405c58b0df1SFrederic Weisbecker depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL 406abf917cdSFrederic Weisbecker select VIRT_CPU_ACCOUNTING 407391dc69cSFrederic Weisbecker help 408391dc69cSFrederic Weisbecker Select this option to enable more accurate task and CPU time 409391dc69cSFrederic Weisbecker accounting. This is done by reading a CPU counter on each 410391dc69cSFrederic Weisbecker kernel entry and exit and on transitions within the kernel 411391dc69cSFrederic Weisbecker between system, softirq and hardirq state, so there is a 412391dc69cSFrederic Weisbecker small performance impact. In the case of s390 or IBM POWER > 5, 413391dc69cSFrederic Weisbecker this also enables accounting of stolen time on logically-partitioned 414391dc69cSFrederic Weisbecker systems. 415391dc69cSFrederic Weisbecker 416abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING_GEN 417abf917cdSFrederic Weisbecker bool "Full dynticks CPU time accounting" 418ff3fb254SKevin Hilman depends on HAVE_CONTEXT_TRACKING 419554b0004SKevin Hilman depends on HAVE_VIRT_CPU_ACCOUNTING_GEN 420041a1574SArnd Bergmann depends on GENERIC_CLOCKEVENTS 421abf917cdSFrederic Weisbecker select VIRT_CPU_ACCOUNTING 422abf917cdSFrederic Weisbecker select CONTEXT_TRACKING 423abf917cdSFrederic Weisbecker help 424abf917cdSFrederic Weisbecker Select this option to enable task and CPU time accounting on full 425abf917cdSFrederic Weisbecker dynticks systems. This accounting is implemented by watching every 426abf917cdSFrederic Weisbecker kernel-user boundaries using the context tracking subsystem. 427abf917cdSFrederic Weisbecker The accounting is thus performed at the expense of some significant 428abf917cdSFrederic Weisbecker overhead. 429abf917cdSFrederic Weisbecker 430abf917cdSFrederic Weisbecker For now this is only useful if you are working on the full 431abf917cdSFrederic Weisbecker dynticks subsystem development. 432abf917cdSFrederic Weisbecker 433abf917cdSFrederic Weisbecker If unsure, say N. 434abf917cdSFrederic Weisbecker 435b58c3584SRik van Rielendchoice 436b58c3584SRik van Riel 437fdf9c356SFrederic Weisbeckerconfig IRQ_TIME_ACCOUNTING 438fdf9c356SFrederic Weisbecker bool "Fine granularity task level IRQ time accounting" 439b58c3584SRik van Riel depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE 440fdf9c356SFrederic Weisbecker help 441fdf9c356SFrederic Weisbecker Select this option to enable fine granularity task irq time 442fdf9c356SFrederic Weisbecker accounting. This is done by reading a timestamp on each 443fdf9c356SFrederic Weisbecker transitions between softirq and hardirq state, so there can be a 444fdf9c356SFrederic Weisbecker small performance impact. 445fdf9c356SFrederic Weisbecker 446fdf9c356SFrederic Weisbecker If in doubt, say N here. 447fdf9c356SFrederic Weisbecker 44811d4afd4SVincent Guittotconfig HAVE_SCHED_AVG_IRQ 44911d4afd4SVincent Guittot def_bool y 45011d4afd4SVincent Guittot depends on IRQ_TIME_ACCOUNTING || PARAVIRT_TIME_ACCOUNTING 45111d4afd4SVincent Guittot depends on SMP 45211d4afd4SVincent Guittot 45376504793SThara Gopinathconfig SCHED_THERMAL_PRESSURE 45476504793SThara Gopinath bool "Enable periodic averaging of thermal pressure" 45576504793SThara Gopinath depends on SMP 45676504793SThara Gopinath 457391dc69cSFrederic Weisbeckerconfig BSD_PROCESS_ACCT 458391dc69cSFrederic Weisbecker bool "BSD Process Accounting" 4592813893fSIulia Manda depends on MULTIUSER 460391dc69cSFrederic Weisbecker help 461391dc69cSFrederic Weisbecker If you say Y here, a user level program will be able to instruct the 462391dc69cSFrederic Weisbecker kernel (via a special system call) to write process accounting 463391dc69cSFrederic Weisbecker information to a file: whenever a process exits, information about 464391dc69cSFrederic Weisbecker that process will be appended to the file by the kernel. The 465391dc69cSFrederic Weisbecker information includes things such as creation time, owning user, 466391dc69cSFrederic Weisbecker command name, memory usage, controlling terminal etc. (the complete 467391dc69cSFrederic Weisbecker list is in the struct acct in <file:include/linux/acct.h>). It is 468391dc69cSFrederic Weisbecker up to the user level program to do useful things with this 469391dc69cSFrederic Weisbecker information. This is generally a good idea, so say Y. 470391dc69cSFrederic Weisbecker 471391dc69cSFrederic Weisbeckerconfig BSD_PROCESS_ACCT_V3 472391dc69cSFrederic Weisbecker bool "BSD Process Accounting version 3 file format" 473391dc69cSFrederic Weisbecker depends on BSD_PROCESS_ACCT 474391dc69cSFrederic Weisbecker default n 475391dc69cSFrederic Weisbecker help 476391dc69cSFrederic Weisbecker If you say Y here, the process accounting information is written 477391dc69cSFrederic Weisbecker in a new file format that also logs the process IDs of each 4783903bf94SRandy Dunlap process and its parent. Note that this file format is incompatible 479391dc69cSFrederic Weisbecker with previous v0/v1/v2 file formats, so you will need updated tools 480391dc69cSFrederic Weisbecker for processing it. A preliminary version of these tools is available 481391dc69cSFrederic Weisbecker at <http://www.gnu.org/software/acct/>. 482391dc69cSFrederic Weisbecker 483391dc69cSFrederic Weisbeckerconfig TASKSTATS 48419c92399SKees Cook bool "Export task/process statistics through netlink" 485391dc69cSFrederic Weisbecker depends on NET 4862813893fSIulia Manda depends on MULTIUSER 487391dc69cSFrederic Weisbecker default n 488391dc69cSFrederic Weisbecker help 489391dc69cSFrederic Weisbecker Export selected statistics for tasks/processes through the 490391dc69cSFrederic Weisbecker generic netlink interface. Unlike BSD process accounting, the 491391dc69cSFrederic Weisbecker statistics are available during the lifetime of tasks/processes as 492391dc69cSFrederic Weisbecker responses to commands. Like BSD accounting, they are sent to user 493391dc69cSFrederic Weisbecker space on task exit. 494391dc69cSFrederic Weisbecker 495391dc69cSFrederic Weisbecker Say N if unsure. 496391dc69cSFrederic Weisbecker 497391dc69cSFrederic Weisbeckerconfig TASK_DELAY_ACCT 49819c92399SKees Cook bool "Enable per-task delay accounting" 499391dc69cSFrederic Weisbecker depends on TASKSTATS 500f6db8347SNaveen N. Rao select SCHED_INFO 501391dc69cSFrederic Weisbecker help 502391dc69cSFrederic Weisbecker Collect information on time spent by a task waiting for system 503391dc69cSFrederic Weisbecker resources like cpu, synchronous block I/O completion and swapping 504391dc69cSFrederic Weisbecker in pages. Such statistics can help in setting a task's priorities 505391dc69cSFrederic Weisbecker relative to other tasks for cpu, io, rss limits etc. 506391dc69cSFrederic Weisbecker 507391dc69cSFrederic Weisbecker Say N if unsure. 508391dc69cSFrederic Weisbecker 509391dc69cSFrederic Weisbeckerconfig TASK_XACCT 51019c92399SKees Cook bool "Enable extended accounting over taskstats" 511391dc69cSFrederic Weisbecker depends on TASKSTATS 512391dc69cSFrederic Weisbecker help 513391dc69cSFrederic Weisbecker Collect extended task accounting data and send the data 514391dc69cSFrederic Weisbecker to userland for processing over the taskstats interface. 515391dc69cSFrederic Weisbecker 516391dc69cSFrederic Weisbecker Say N if unsure. 517391dc69cSFrederic Weisbecker 518391dc69cSFrederic Weisbeckerconfig TASK_IO_ACCOUNTING 51919c92399SKees Cook bool "Enable per-task storage I/O accounting" 520391dc69cSFrederic Weisbecker depends on TASK_XACCT 521391dc69cSFrederic Weisbecker help 522391dc69cSFrederic Weisbecker Collect information on the number of bytes of storage I/O which this 523391dc69cSFrederic Weisbecker task has caused. 524391dc69cSFrederic Weisbecker 525391dc69cSFrederic Weisbecker Say N if unsure. 526391dc69cSFrederic Weisbecker 527eb414681SJohannes Weinerconfig PSI 528eb414681SJohannes Weiner bool "Pressure stall information tracking" 529eb414681SJohannes Weiner help 530eb414681SJohannes Weiner Collect metrics that indicate how overcommitted the CPU, memory, 531eb414681SJohannes Weiner and IO capacity are in the system. 532eb414681SJohannes Weiner 533eb414681SJohannes Weiner If you say Y here, the kernel will create /proc/pressure/ with the 534eb414681SJohannes Weiner pressure statistics files cpu, memory, and io. These will indicate 535eb414681SJohannes Weiner the share of walltime in which some or all tasks in the system are 536eb414681SJohannes Weiner delayed due to contention of the respective resource. 537eb414681SJohannes Weiner 5382ce7135aSJohannes Weiner In kernels with cgroup support, cgroups (cgroup2 only) will 5392ce7135aSJohannes Weiner have cpu.pressure, memory.pressure, and io.pressure files, 5402ce7135aSJohannes Weiner which aggregate pressure stalls for the grouped tasks only. 5412ce7135aSJohannes Weiner 542c3123552SMauro Carvalho Chehab For more details see Documentation/accounting/psi.rst. 543eb414681SJohannes Weiner 544eb414681SJohannes Weiner Say N if unsure. 545eb414681SJohannes Weiner 546e0c27447SJohannes Weinerconfig PSI_DEFAULT_DISABLED 547e0c27447SJohannes Weiner bool "Require boot parameter to enable pressure stall information tracking" 548e0c27447SJohannes Weiner default n 549e0c27447SJohannes Weiner depends on PSI 550e0c27447SJohannes Weiner help 551e0c27447SJohannes Weiner If set, pressure stall information tracking will be disabled 552428a1cb4SBaruch Siach per default but can be enabled through passing psi=1 on the 553428a1cb4SBaruch Siach kernel commandline during boot. 554e0c27447SJohannes Weiner 5557b2489d3SJohannes Weiner This feature adds some code to the task wakeup and sleep 5567b2489d3SJohannes Weiner paths of the scheduler. The overhead is too low to affect 5577b2489d3SJohannes Weiner common scheduling-intense workloads in practice (such as 5587b2489d3SJohannes Weiner webservers, memcache), but it does show up in artificial 5597b2489d3SJohannes Weiner scheduler stress tests, such as hackbench. 5607b2489d3SJohannes Weiner 5617b2489d3SJohannes Weiner If you are paranoid and not sure what the kernel will be 5627b2489d3SJohannes Weiner used for, say Y. 5637b2489d3SJohannes Weiner 5647b2489d3SJohannes Weiner Say N if unsure. 5657b2489d3SJohannes Weiner 566391dc69cSFrederic Weisbeckerendmenu # "CPU/Task time and stats accounting" 567391dc69cSFrederic Weisbecker 5685c4991e2SFrederic Weisbeckerconfig CPU_ISOLATION 5695c4991e2SFrederic Weisbecker bool "CPU isolation" 570414a2dc1SGeert Uytterhoeven depends on SMP || COMPILE_TEST 5712c43838cSFrederic Weisbecker default y 5725c4991e2SFrederic Weisbecker help 5735c4991e2SFrederic Weisbecker Make sure that CPUs running critical tasks are not disturbed by 5745c4991e2SFrederic Weisbecker any source of "noise" such as unbound workqueues, timers, kthreads... 5752c43838cSFrederic Weisbecker Unbound jobs get offloaded to housekeeping CPUs. This is driven by 5762c43838cSFrederic Weisbecker the "isolcpus=" boot parameter. 5772c43838cSFrederic Weisbecker 5782c43838cSFrederic Weisbecker Say Y if unsure. 5795c4991e2SFrederic Weisbecker 5800af92d46SPaul E. McKenneysource "kernel/rcu/Kconfig" 581c903ff83SMike Travis 582de5b56baSVivek Goyalconfig BUILD_BIN2C 583de5b56baSVivek Goyal bool 584de5b56baSVivek Goyal default n 585de5b56baSVivek Goyal 5861da177e4SLinus Torvaldsconfig IKCONFIG 587f2443ab6SRoss Biro tristate "Kernel .config support" 5881da177e4SLinus Torvalds ---help--- 5891da177e4SLinus Torvalds This option enables the complete Linux kernel ".config" file 5901da177e4SLinus Torvalds contents to be saved in the kernel. It provides documentation 5911da177e4SLinus Torvalds of which kernel options are used in a running kernel or in an 5921da177e4SLinus Torvalds on-disk kernel. This information can be extracted from the kernel 5931da177e4SLinus Torvalds image file with the script scripts/extract-ikconfig and used as 5941da177e4SLinus Torvalds input to rebuild the current kernel or to build another kernel. 5951da177e4SLinus Torvalds It can also be extracted from a running kernel by reading 5961da177e4SLinus Torvalds /proc/config.gz if enabled (below). 5971da177e4SLinus Torvalds 5981da177e4SLinus Torvaldsconfig IKCONFIG_PROC 5991da177e4SLinus Torvalds bool "Enable access to .config through /proc/config.gz" 6001da177e4SLinus Torvalds depends on IKCONFIG && PROC_FS 6011da177e4SLinus Torvalds ---help--- 6021da177e4SLinus Torvalds This option enables access to the kernel configuration file 6031da177e4SLinus Torvalds through /proc/config.gz. 6041da177e4SLinus Torvalds 605f7b101d3SJoel Fernandes (Google)config IKHEADERS 606f7b101d3SJoel Fernandes (Google) tristate "Enable kernel headers through /sys/kernel/kheaders.tar.xz" 607f7b101d3SJoel Fernandes (Google) depends on SYSFS 60843d8ce9dSJoel Fernandes (Google) help 609f7b101d3SJoel Fernandes (Google) This option enables access to the in-kernel headers that are generated during 610f7b101d3SJoel Fernandes (Google) the build process. These can be used to build eBPF tracing programs, 611f7b101d3SJoel Fernandes (Google) or similar programs. If you build the headers as a module, a module called 612f7b101d3SJoel Fernandes (Google) kheaders.ko is built which can be loaded on-demand to get access to headers. 61343d8ce9dSJoel Fernandes (Google) 614794543a2SAlistair John Strachanconfig LOG_BUF_SHIFT 615794543a2SAlistair John Strachan int "Kernel log buffer size (16 => 64KB, 17 => 128KB)" 616fb39f98dSIngo Molnar range 12 25 617f17a32e9SAdrian Bunk default 17 618361e9dfbSJosh Triplett depends on PRINTK 619794543a2SAlistair John Strachan help 62023b2899fSLuis R. Rodriguez Select the minimal kernel log buffer size as a power of 2. 62123b2899fSLuis R. Rodriguez The final size is affected by LOG_CPU_MAX_BUF_SHIFT config 62223b2899fSLuis R. Rodriguez parameter, see below. Any higher size also might be forced 62323b2899fSLuis R. Rodriguez by "log_buf_len" boot parameter. 62423b2899fSLuis R. Rodriguez 625f17a32e9SAdrian Bunk Examples: 626f17a32e9SAdrian Bunk 17 => 128 KB 627f17a32e9SAdrian Bunk 16 => 64 KB 628f17a32e9SAdrian Bunk 15 => 32 KB 629f17a32e9SAdrian Bunk 14 => 16 KB 630794543a2SAlistair John Strachan 13 => 8 KB 631794543a2SAlistair John Strachan 12 => 4 KB 632794543a2SAlistair John Strachan 63323b2899fSLuis R. Rodriguezconfig LOG_CPU_MAX_BUF_SHIFT 63423b2899fSLuis R. Rodriguez int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)" 6352240a31dSGeert Uytterhoeven depends on SMP 63623b2899fSLuis R. Rodriguez range 0 21 63723b2899fSLuis R. Rodriguez default 12 if !BASE_SMALL 63823b2899fSLuis R. Rodriguez default 0 if BASE_SMALL 639361e9dfbSJosh Triplett depends on PRINTK 64023b2899fSLuis R. Rodriguez help 64123b2899fSLuis R. Rodriguez This option allows to increase the default ring buffer size 64223b2899fSLuis R. Rodriguez according to the number of CPUs. The value defines the contribution 64323b2899fSLuis R. Rodriguez of each CPU as a power of 2. The used space is typically only few 64423b2899fSLuis R. Rodriguez lines however it might be much more when problems are reported, 64523b2899fSLuis R. Rodriguez e.g. backtraces. 64623b2899fSLuis R. Rodriguez 64723b2899fSLuis R. Rodriguez The increased size means that a new buffer has to be allocated and 64823b2899fSLuis R. Rodriguez the original static one is unused. It makes sense only on systems 64923b2899fSLuis R. Rodriguez with more CPUs. Therefore this value is used only when the sum of 65023b2899fSLuis R. Rodriguez contributions is greater than the half of the default kernel ring 65123b2899fSLuis R. Rodriguez buffer as defined by LOG_BUF_SHIFT. The default values are set 65223b2899fSLuis R. Rodriguez so that more than 64 CPUs are needed to trigger the allocation. 65323b2899fSLuis R. Rodriguez 65423b2899fSLuis R. Rodriguez Also this option is ignored when "log_buf_len" kernel parameter is 65523b2899fSLuis R. Rodriguez used as it forces an exact (power of two) size of the ring buffer. 65623b2899fSLuis R. Rodriguez 65723b2899fSLuis R. Rodriguez The number of possible CPUs is used for this computation ignoring 6585e0d8d59SGeert Uytterhoeven hotplugging making the computation optimal for the worst case 6595e0d8d59SGeert Uytterhoeven scenario while allowing a simple algorithm to be used from bootup. 66023b2899fSLuis R. Rodriguez 66123b2899fSLuis R. Rodriguez Examples shift values and their meaning: 66223b2899fSLuis R. Rodriguez 17 => 128 KB for each CPU 66323b2899fSLuis R. Rodriguez 16 => 64 KB for each CPU 66423b2899fSLuis R. Rodriguez 15 => 32 KB for each CPU 66523b2899fSLuis R. Rodriguez 14 => 16 KB for each CPU 66623b2899fSLuis R. Rodriguez 13 => 8 KB for each CPU 66723b2899fSLuis R. Rodriguez 12 => 4 KB for each CPU 66823b2899fSLuis R. Rodriguez 669f92bac3bSSergey Senozhatskyconfig PRINTK_SAFE_LOG_BUF_SHIFT 670f92bac3bSSergey Senozhatsky int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)" 671427934b8SPetr Mladek range 10 21 672427934b8SPetr Mladek default 13 673f92bac3bSSergey Senozhatsky depends on PRINTK 674427934b8SPetr Mladek help 675f92bac3bSSergey Senozhatsky Select the size of an alternate printk per-CPU buffer where messages 676f92bac3bSSergey Senozhatsky printed from usafe contexts are temporary stored. One example would 677f92bac3bSSergey Senozhatsky be NMI messages, another one - printk recursion. The messages are 678f92bac3bSSergey Senozhatsky copied to the main log buffer in a safe context to avoid a deadlock. 679f92bac3bSSergey Senozhatsky The value defines the size as a power of 2. 680427934b8SPetr Mladek 681f92bac3bSSergey Senozhatsky Those messages are rare and limited. The largest one is when 682427934b8SPetr Mladek a backtrace is printed. It usually fits into 4KB. Select 683427934b8SPetr Mladek 8KB if you want to be on the safe side. 684427934b8SPetr Mladek 685427934b8SPetr Mladek Examples: 686427934b8SPetr Mladek 17 => 128 KB for each CPU 687427934b8SPetr Mladek 16 => 64 KB for each CPU 688427934b8SPetr Mladek 15 => 32 KB for each CPU 689427934b8SPetr Mladek 14 => 16 KB for each CPU 690427934b8SPetr Mladek 13 => 8 KB for each CPU 691427934b8SPetr Mladek 12 => 4 KB for each CPU 692427934b8SPetr Mladek 6935cdc38f9SKAMEZAWA Hiroyuki# 6945cdc38f9SKAMEZAWA Hiroyuki# Architectures with an unreliable sched_clock() should select this: 6955cdc38f9SKAMEZAWA Hiroyuki# 6965cdc38f9SKAMEZAWA Hiroyukiconfig HAVE_UNSTABLE_SCHED_CLOCK 6975cdc38f9SKAMEZAWA Hiroyuki bool 6985cdc38f9SKAMEZAWA Hiroyuki 69938ff87f7SStephen Boydconfig GENERIC_SCHED_CLOCK 70038ff87f7SStephen Boyd bool 70138ff87f7SStephen Boyd 70269842cbaSPatrick Bellasimenu "Scheduler features" 70369842cbaSPatrick Bellasi 70469842cbaSPatrick Bellasiconfig UCLAMP_TASK 70569842cbaSPatrick Bellasi bool "Enable utilization clamping for RT/FAIR tasks" 70669842cbaSPatrick Bellasi depends on CPU_FREQ_GOV_SCHEDUTIL 70769842cbaSPatrick Bellasi help 70869842cbaSPatrick Bellasi This feature enables the scheduler to track the clamped utilization 70969842cbaSPatrick Bellasi of each CPU based on RUNNABLE tasks scheduled on that CPU. 71069842cbaSPatrick Bellasi 71169842cbaSPatrick Bellasi With this option, the user can specify the min and max CPU 71269842cbaSPatrick Bellasi utilization allowed for RUNNABLE tasks. The max utilization defines 71369842cbaSPatrick Bellasi the maximum frequency a task should use while the min utilization 71469842cbaSPatrick Bellasi defines the minimum frequency it should use. 71569842cbaSPatrick Bellasi 71669842cbaSPatrick Bellasi Both min and max utilization clamp values are hints to the scheduler, 71769842cbaSPatrick Bellasi aiming at improving its frequency selection policy, but they do not 71869842cbaSPatrick Bellasi enforce or grant any specific bandwidth for tasks. 71969842cbaSPatrick Bellasi 72069842cbaSPatrick Bellasi If in doubt, say N. 72169842cbaSPatrick Bellasi 72269842cbaSPatrick Bellasiconfig UCLAMP_BUCKETS_COUNT 72369842cbaSPatrick Bellasi int "Number of supported utilization clamp buckets" 72469842cbaSPatrick Bellasi range 5 20 72569842cbaSPatrick Bellasi default 5 72669842cbaSPatrick Bellasi depends on UCLAMP_TASK 72769842cbaSPatrick Bellasi help 72869842cbaSPatrick Bellasi Defines the number of clamp buckets to use. The range of each bucket 72969842cbaSPatrick Bellasi will be SCHED_CAPACITY_SCALE/UCLAMP_BUCKETS_COUNT. The higher the 73069842cbaSPatrick Bellasi number of clamp buckets the finer their granularity and the higher 73169842cbaSPatrick Bellasi the precision of clamping aggregation and tracking at run-time. 73269842cbaSPatrick Bellasi 73369842cbaSPatrick Bellasi For example, with the minimum configuration value we will have 5 73469842cbaSPatrick Bellasi clamp buckets tracking 20% utilization each. A 25% boosted tasks will 73569842cbaSPatrick Bellasi be refcounted in the [20..39]% bucket and will set the bucket clamp 73669842cbaSPatrick Bellasi effective value to 25%. 73769842cbaSPatrick Bellasi If a second 30% boosted task should be co-scheduled on the same CPU, 73869842cbaSPatrick Bellasi that task will be refcounted in the same bucket of the first task and 73969842cbaSPatrick Bellasi it will boost the bucket clamp effective value to 30%. 74069842cbaSPatrick Bellasi The clamp effective value of a bucket is reset to its nominal value 74169842cbaSPatrick Bellasi (20% in the example above) when there are no more tasks refcounted in 74269842cbaSPatrick Bellasi that bucket. 74369842cbaSPatrick Bellasi 74469842cbaSPatrick Bellasi An additional boost/capping margin can be added to some tasks. In the 74569842cbaSPatrick Bellasi example above the 25% task will be boosted to 30% until it exits the 74669842cbaSPatrick Bellasi CPU. If that should be considered not acceptable on certain systems, 74769842cbaSPatrick Bellasi it's always possible to reduce the margin by increasing the number of 74869842cbaSPatrick Bellasi clamp buckets to trade off used memory for run-time tracking 74969842cbaSPatrick Bellasi precision. 75069842cbaSPatrick Bellasi 75169842cbaSPatrick Bellasi If in doubt, use the default value. 75269842cbaSPatrick Bellasi 75369842cbaSPatrick Bellasiendmenu 75469842cbaSPatrick Bellasi 755be3a7284SAndrea Arcangeli# 756be3a7284SAndrea Arcangeli# For architectures that want to enable the support for NUMA-affine scheduler 757be3a7284SAndrea Arcangeli# balancing logic: 758be3a7284SAndrea Arcangeli# 759be3a7284SAndrea Arcangeliconfig ARCH_SUPPORTS_NUMA_BALANCING 760be3a7284SAndrea Arcangeli bool 761be3a7284SAndrea Arcangeli 762be5e610cSPeter Zijlstra# 76372b252aeSMel Gorman# For architectures that prefer to flush all TLBs after a number of pages 76472b252aeSMel Gorman# are unmapped instead of sending one IPI per page to flush. The architecture 76572b252aeSMel Gorman# must provide guarantees on what happens if a clean TLB cache entry is 76672b252aeSMel Gorman# written after the unmap. Details are in mm/rmap.c near the check for 76772b252aeSMel Gorman# should_defer_flush. The architecture should also consider if the full flush 76872b252aeSMel Gorman# and the refill costs are offset by the savings of sending fewer IPIs. 76972b252aeSMel Gormanconfig ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH 77072b252aeSMel Gorman bool 77172b252aeSMel Gorman 772c12d3362SArd Biesheuvelconfig CC_HAS_INT128 7733a7c7331SMasahiro Yamada def_bool !$(cc-option,$(m64-flag) -D__SIZEOF_INT128__=0) && 64BIT 774c12d3362SArd Biesheuvel 77572b252aeSMel Gorman# 776be5e610cSPeter Zijlstra# For architectures that know their GCC __int128 support is sound 777be5e610cSPeter Zijlstra# 778be5e610cSPeter Zijlstraconfig ARCH_SUPPORTS_INT128 779be5e610cSPeter Zijlstra bool 780be5e610cSPeter Zijlstra 781be3a7284SAndrea Arcangeli# For architectures that (ab)use NUMA to represent different memory regions 782be3a7284SAndrea Arcangeli# all cpu-local but of different latencies, such as SuperH. 783be3a7284SAndrea Arcangeli# 784be3a7284SAndrea Arcangeliconfig ARCH_WANT_NUMA_VARIABLE_LOCALITY 785be3a7284SAndrea Arcangeli bool 786be3a7284SAndrea Arcangeli 787be3a7284SAndrea Arcangeliconfig NUMA_BALANCING 788be3a7284SAndrea Arcangeli bool "Memory placement aware NUMA scheduler" 789be3a7284SAndrea Arcangeli depends on ARCH_SUPPORTS_NUMA_BALANCING 790be3a7284SAndrea Arcangeli depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY 791be3a7284SAndrea Arcangeli depends on SMP && NUMA && MIGRATION 792be3a7284SAndrea Arcangeli help 793be3a7284SAndrea Arcangeli This option adds support for automatic NUMA aware memory/task placement. 794be3a7284SAndrea Arcangeli The mechanism is quite primitive and is based on migrating memory when 7956d56a410SPaul Gortmaker it has references to the node the task is running on. 796be3a7284SAndrea Arcangeli 797be3a7284SAndrea Arcangeli This system will be inactive on UMA systems. 798be3a7284SAndrea Arcangeli 7996f7c97e8SAneesh Kumar K.Vconfig NUMA_BALANCING_DEFAULT_ENABLED 8006f7c97e8SAneesh Kumar K.V bool "Automatically enable NUMA aware memory/task placement" 8016f7c97e8SAneesh Kumar K.V default y 8026f7c97e8SAneesh Kumar K.V depends on NUMA_BALANCING 8036f7c97e8SAneesh Kumar K.V help 8046f7c97e8SAneesh Kumar K.V If set, automatic NUMA balancing will be enabled if running on a NUMA 8056f7c97e8SAneesh Kumar K.V machine. 8066f7c97e8SAneesh Kumar K.V 80723964d2dSLi Zefanmenuconfig CGROUPS 8086341e62bSChristoph Jaeger bool "Control Group support" 8092bd59d48STejun Heo select KERNFS 810ddbcc7e8SPaul Menage help 81123964d2dSLi Zefan This option adds support for grouping sets of processes together, for 8125cdc38f9SKAMEZAWA Hiroyuki use with process control subsystems such as Cpusets, CFS, memory 8135cdc38f9SKAMEZAWA Hiroyuki controls or device isolation. 8145cdc38f9SKAMEZAWA Hiroyuki See 815d6a3b247SMauro Carvalho Chehab - Documentation/scheduler/sched-design-CFS.rst (CFS) 816da82c92fSMauro Carvalho Chehab - Documentation/admin-guide/cgroup-v1/ (features for grouping, isolation 81745ce80fbSLi Zefan and resource control) 818ddbcc7e8SPaul Menage 819ddbcc7e8SPaul Menage Say N if unsure. 820ddbcc7e8SPaul Menage 82123964d2dSLi Zefanif CGROUPS 82223964d2dSLi Zefan 8233e32cb2eSJohannes Weinerconfig PAGE_COUNTER 8243e32cb2eSJohannes Weiner bool 8253e32cb2eSJohannes Weiner 826c255a458SAndrew Mortonconfig MEMCG 827a0166ec4SJohannes Weiner bool "Memory controller" 8283e32cb2eSJohannes Weiner select PAGE_COUNTER 82979bd9814STejun Heo select EVENTFD 83000f0b825SBalbir Singh help 831a0166ec4SJohannes Weiner Provides control over the memory footprint of tasks in a cgroup. 83200f0b825SBalbir Singh 833c255a458SAndrew Mortonconfig MEMCG_SWAP 834a0166ec4SJohannes Weiner bool "Swap controller" 835c255a458SAndrew Morton depends on MEMCG && SWAP 836c077719bSKAMEZAWA Hiroyuki help 837a0166ec4SJohannes Weiner Provides control over the swap space consumed by tasks in a cgroup. 838a0166ec4SJohannes Weiner 839c255a458SAndrew Mortonconfig MEMCG_SWAP_ENABLED 840a0166ec4SJohannes Weiner bool "Swap controller enabled by default" 841c255a458SAndrew Morton depends on MEMCG_SWAP 842a42c390cSMichal Hocko default y 843a42c390cSMichal Hocko help 844a42c390cSMichal Hocko Memory Resource Controller Swap Extension comes with its price in 845a42c390cSMichal Hocko a bigger memory consumption. General purpose distribution kernels 84643d547f9SJim Cromie which want to enable the feature but keep it disabled by default 84707555ac1SMichal Hocko and let the user enable it by swapaccount=1 boot command line 848a42c390cSMichal Hocko parameter should have this option unselected. 849a42c390cSMichal Hocko For those who want to have the feature enabled by default should 850a42c390cSMichal Hocko select this option (if, for some reason, they need to disable it 85100a66d29SWANG Cong then swapaccount=0 does the trick). 852c077719bSKAMEZAWA Hiroyuki 85384c07d11SKirill Tkhaiconfig MEMCG_KMEM 85484c07d11SKirill Tkhai bool 85584c07d11SKirill Tkhai depends on MEMCG && !SLOB 85684c07d11SKirill Tkhai default y 85784c07d11SKirill Tkhai 8586bf024e6SJohannes Weinerconfig BLK_CGROUP 8596bf024e6SJohannes Weiner bool "IO controller" 8606bf024e6SJohannes Weiner depends on BLOCK 8612bc64a20SAneesh Kumar K.V default n 8626bf024e6SJohannes Weiner ---help--- 8636bf024e6SJohannes Weiner Generic block IO controller cgroup interface. This is the common 8646bf024e6SJohannes Weiner cgroup interface which should be used by various IO controlling 8656bf024e6SJohannes Weiner policies. 8662bc64a20SAneesh Kumar K.V 8676bf024e6SJohannes Weiner Currently, CFQ IO scheduler uses it to recognize task groups and 8686bf024e6SJohannes Weiner control disk bandwidth allocation (proportional time slice allocation) 8696bf024e6SJohannes Weiner to such task groups. It is also used by bio throttling logic in 8706bf024e6SJohannes Weiner block layer to implement upper limit in IO rates on a device. 871e5d1367fSStephane Eranian 8726bf024e6SJohannes Weiner This option only enables generic Block IO controller infrastructure. 8736bf024e6SJohannes Weiner One needs to also enable actual IO controlling logic/policy. For 8746bf024e6SJohannes Weiner enabling proportional weight division of disk bandwidth in CFQ, set 8757baf2199SKrzysztof Kozlowski CONFIG_BFQ_GROUP_IOSCHED=y; for enabling throttling policy, set 8766bf024e6SJohannes Weiner CONFIG_BLK_DEV_THROTTLING=y. 8776bf024e6SJohannes Weiner 878da82c92fSMauro Carvalho Chehab See Documentation/admin-guide/cgroup-v1/blkio-controller.rst for more information. 8796bf024e6SJohannes Weiner 8806bf024e6SJohannes Weinerconfig CGROUP_WRITEBACK 8816bf024e6SJohannes Weiner bool 8826bf024e6SJohannes Weiner depends on MEMCG && BLK_CGROUP 8836bf024e6SJohannes Weiner default y 884e5d1367fSStephane Eranian 8857c941438SDhaval Gianimenuconfig CGROUP_SCHED 886a0166ec4SJohannes Weiner bool "CPU controller" 8877c941438SDhaval Giani default n 8887c941438SDhaval Giani help 8897c941438SDhaval Giani This feature lets CPU scheduler recognize task groups and control CPU 8907c941438SDhaval Giani bandwidth allocation to such task groups. It uses cgroups to group 8917c941438SDhaval Giani tasks. 8927c941438SDhaval Giani 8937c941438SDhaval Gianiif CGROUP_SCHED 8947c941438SDhaval Gianiconfig FAIR_GROUP_SCHED 8957c941438SDhaval Giani bool "Group scheduling for SCHED_OTHER" 8967c941438SDhaval Giani depends on CGROUP_SCHED 8977c941438SDhaval Giani default CGROUP_SCHED 8987c941438SDhaval Giani 899ab84d31eSPaul Turnerconfig CFS_BANDWIDTH 900ab84d31eSPaul Turner bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED" 901ab84d31eSPaul Turner depends on FAIR_GROUP_SCHED 902ab84d31eSPaul Turner default n 903ab84d31eSPaul Turner help 904ab84d31eSPaul Turner This option allows users to define CPU bandwidth rates (limits) for 905ab84d31eSPaul Turner tasks running within the fair group scheduler. Groups with no limit 906ab84d31eSPaul Turner set are considered to be unconstrained and will run with no 907ab84d31eSPaul Turner restriction. 908d6a3b247SMauro Carvalho Chehab See Documentation/scheduler/sched-bwc.rst for more information. 909ab84d31eSPaul Turner 9107c941438SDhaval Gianiconfig RT_GROUP_SCHED 9117c941438SDhaval Giani bool "Group scheduling for SCHED_RR/FIFO" 9127c941438SDhaval Giani depends on CGROUP_SCHED 9137c941438SDhaval Giani default n 9147c941438SDhaval Giani help 9157c941438SDhaval Giani This feature lets you explicitly allocate real CPU bandwidth 91632bd7eb5SLi Zefan to task groups. If enabled, it will also make it impossible to 9177c941438SDhaval Giani schedule realtime tasks for non-root users until you allocate 9187c941438SDhaval Giani realtime bandwidth for them. 919d6a3b247SMauro Carvalho Chehab See Documentation/scheduler/sched-rt-group.rst for more information. 9207c941438SDhaval Giani 9217c941438SDhaval Gianiendif #CGROUP_SCHED 9227c941438SDhaval Giani 9232480c093SPatrick Bellasiconfig UCLAMP_TASK_GROUP 9242480c093SPatrick Bellasi bool "Utilization clamping per group of tasks" 9252480c093SPatrick Bellasi depends on CGROUP_SCHED 9262480c093SPatrick Bellasi depends on UCLAMP_TASK 9272480c093SPatrick Bellasi default n 9282480c093SPatrick Bellasi help 9292480c093SPatrick Bellasi This feature enables the scheduler to track the clamped utilization 9302480c093SPatrick Bellasi of each CPU based on RUNNABLE tasks currently scheduled on that CPU. 9312480c093SPatrick Bellasi 9322480c093SPatrick Bellasi When this option is enabled, the user can specify a min and max 9332480c093SPatrick Bellasi CPU bandwidth which is allowed for each single task in a group. 9342480c093SPatrick Bellasi The max bandwidth allows to clamp the maximum frequency a task 9352480c093SPatrick Bellasi can use, while the min bandwidth allows to define a minimum 9362480c093SPatrick Bellasi frequency a task will always use. 9372480c093SPatrick Bellasi 9382480c093SPatrick Bellasi When task group based utilization clamping is enabled, an eventually 9392480c093SPatrick Bellasi specified task-specific clamp value is constrained by the cgroup 9402480c093SPatrick Bellasi specified clamp value. Both minimum and maximum task clamping cannot 9412480c093SPatrick Bellasi be bigger than the corresponding clamping defined at task group level. 9422480c093SPatrick Bellasi 9432480c093SPatrick Bellasi If in doubt, say N. 9442480c093SPatrick Bellasi 9456bf024e6SJohannes Weinerconfig CGROUP_PIDS 9466bf024e6SJohannes Weiner bool "PIDs controller" 9476bf024e6SJohannes Weiner help 9486bf024e6SJohannes Weiner Provides enforcement of process number limits in the scope of a 9496bf024e6SJohannes Weiner cgroup. Any attempt to fork more processes than is allowed in the 9506bf024e6SJohannes Weiner cgroup will fail. PIDs are fundamentally a global resource because it 9516bf024e6SJohannes Weiner is fairly trivial to reach PID exhaustion before you reach even a 9526bf024e6SJohannes Weiner conservative kmemcg limit. As a result, it is possible to grind a 9536bf024e6SJohannes Weiner system to halt without being limited by other cgroup policies. The 9546cc578dfSParav Pandit PIDs controller is designed to stop this from happening. 9556bf024e6SJohannes Weiner 9566bf024e6SJohannes Weiner It should be noted that organisational operations (such as attaching 95798076833SJonathan Neuschäfer to a cgroup hierarchy) will *not* be blocked by the PIDs controller, 9586bf024e6SJohannes Weiner since the PIDs limit only affects a process's ability to fork, not to 9596bf024e6SJohannes Weiner attach to a cgroup. 9606bf024e6SJohannes Weiner 96139d3e758SParav Panditconfig CGROUP_RDMA 96239d3e758SParav Pandit bool "RDMA controller" 96339d3e758SParav Pandit help 96439d3e758SParav Pandit Provides enforcement of RDMA resources defined by IB stack. 96539d3e758SParav Pandit It is fairly easy for consumers to exhaust RDMA resources, which 96639d3e758SParav Pandit can result into resource unavailability to other consumers. 96739d3e758SParav Pandit RDMA controller is designed to stop this from happening. 96839d3e758SParav Pandit Attaching processes with active RDMA resources to the cgroup 96939d3e758SParav Pandit hierarchy is allowed even if can cross the hierarchy's limit. 97039d3e758SParav Pandit 9716bf024e6SJohannes Weinerconfig CGROUP_FREEZER 9726bf024e6SJohannes Weiner bool "Freezer controller" 9736bf024e6SJohannes Weiner help 9746bf024e6SJohannes Weiner Provides a way to freeze and unfreeze all tasks in a 9756bf024e6SJohannes Weiner cgroup. 9766bf024e6SJohannes Weiner 977489c2a20SJohannes Weiner This option affects the ORIGINAL cgroup interface. The cgroup2 memory 978489c2a20SJohannes Weiner controller includes important in-kernel memory consumers per default. 979489c2a20SJohannes Weiner 980489c2a20SJohannes Weiner If you're using cgroup2, say N. 981489c2a20SJohannes Weiner 9826bf024e6SJohannes Weinerconfig CGROUP_HUGETLB 9836bf024e6SJohannes Weiner bool "HugeTLB controller" 9846bf024e6SJohannes Weiner depends on HUGETLB_PAGE 9856bf024e6SJohannes Weiner select PAGE_COUNTER 986afc24d49SVivek Goyal default n 9876bf024e6SJohannes Weiner help 9886bf024e6SJohannes Weiner Provides a cgroup controller for HugeTLB pages. 9896bf024e6SJohannes Weiner When you enable this, you can put a per cgroup limit on HugeTLB usage. 9906bf024e6SJohannes Weiner The limit is enforced during page fault. Since HugeTLB doesn't 9916bf024e6SJohannes Weiner support page reclaim, enforcing the limit at page fault time implies 9926bf024e6SJohannes Weiner that, the application will get SIGBUS signal if it tries to access 9936bf024e6SJohannes Weiner HugeTLB pages beyond its limit. This requires the application to know 9946bf024e6SJohannes Weiner beforehand how much HugeTLB pages it would require for its use. The 9956bf024e6SJohannes Weiner control group is tracked in the third page lru pointer. This means 9966bf024e6SJohannes Weiner that we cannot use the controller with huge page less than 3 pages. 997afc24d49SVivek Goyal 9986bf024e6SJohannes Weinerconfig CPUSETS 9996bf024e6SJohannes Weiner bool "Cpuset controller" 1000e1d4eeecSNicolas Pitre depends on SMP 10016bf024e6SJohannes Weiner help 10026bf024e6SJohannes Weiner This option will let you create and manage CPUSETs which 10036bf024e6SJohannes Weiner allow dynamically partitioning a system into sets of CPUs and 10046bf024e6SJohannes Weiner Memory Nodes and assigning tasks to run only within those sets. 10056bf024e6SJohannes Weiner This is primarily useful on large SMP or NUMA systems. 1006afc24d49SVivek Goyal 10076bf024e6SJohannes Weiner Say N if unsure. 1008afc24d49SVivek Goyal 10096bf024e6SJohannes Weinerconfig PROC_PID_CPUSET 10106bf024e6SJohannes Weiner bool "Include legacy /proc/<pid>/cpuset file" 10116bf024e6SJohannes Weiner depends on CPUSETS 101289e9b9e0STejun Heo default y 101389e9b9e0STejun Heo 10146bf024e6SJohannes Weinerconfig CGROUP_DEVICE 10156bf024e6SJohannes Weiner bool "Device controller" 10166bf024e6SJohannes Weiner help 10176bf024e6SJohannes Weiner Provides a cgroup controller implementing whitelists for 10186bf024e6SJohannes Weiner devices which a process in the cgroup can mknod or open. 10196bf024e6SJohannes Weiner 10206bf024e6SJohannes Weinerconfig CGROUP_CPUACCT 10216bf024e6SJohannes Weiner bool "Simple CPU accounting controller" 10226bf024e6SJohannes Weiner help 10236bf024e6SJohannes Weiner Provides a simple controller for monitoring the 10246bf024e6SJohannes Weiner total CPU consumed by the tasks in a cgroup. 10256bf024e6SJohannes Weiner 10266bf024e6SJohannes Weinerconfig CGROUP_PERF 10276bf024e6SJohannes Weiner bool "Perf controller" 10286bf024e6SJohannes Weiner depends on PERF_EVENTS 10296bf024e6SJohannes Weiner help 10306bf024e6SJohannes Weiner This option extends the perf per-cpu mode to restrict monitoring 10316bf024e6SJohannes Weiner to threads which belong to the cgroup specified and run on the 10326546b19fSNamhyung Kim designated cpu. Or this can be used to have cgroup ID in samples 10336546b19fSNamhyung Kim so that it can monitor performance events among cgroups. 10346bf024e6SJohannes Weiner 10356bf024e6SJohannes Weiner Say N if unsure. 10366bf024e6SJohannes Weiner 103730070984SDaniel Mackconfig CGROUP_BPF 103830070984SDaniel Mack bool "Support for eBPF programs attached to cgroups" 1039483c4933SAndy Lutomirski depends on BPF_SYSCALL 1040483c4933SAndy Lutomirski select SOCK_CGROUP_DATA 104130070984SDaniel Mack help 104230070984SDaniel Mack Allow attaching eBPF programs to a cgroup using the bpf(2) 104330070984SDaniel Mack syscall command BPF_PROG_ATTACH. 104430070984SDaniel Mack 104530070984SDaniel Mack In which context these programs are accessed depends on the type 104630070984SDaniel Mack of attachment. For instance, programs that are attached using 104730070984SDaniel Mack BPF_CGROUP_INET_INGRESS will be executed on the ingress path of 104830070984SDaniel Mack inet sockets. 104930070984SDaniel Mack 10506bf024e6SJohannes Weinerconfig CGROUP_DEBUG 105123b0be48SWaiman Long bool "Debug controller" 10526bf024e6SJohannes Weiner default n 105323b0be48SWaiman Long depends on DEBUG_KERNEL 10546bf024e6SJohannes Weiner help 10556bf024e6SJohannes Weiner This option enables a simple controller that exports 105623b0be48SWaiman Long debugging information about the cgroups framework. This 105723b0be48SWaiman Long controller is for control cgroup debugging only. Its 105823b0be48SWaiman Long interfaces are not stable. 10596bf024e6SJohannes Weiner 10606bf024e6SJohannes Weiner Say N. 10616bf024e6SJohannes Weiner 106273b35147SArnd Bergmannconfig SOCK_CGROUP_DATA 106373b35147SArnd Bergmann bool 106473b35147SArnd Bergmann default n 106573b35147SArnd Bergmann 106623964d2dSLi Zefanendif # CGROUPS 1067c077719bSKAMEZAWA Hiroyuki 10688dd2a82cSDaniel Lezcanomenuconfig NAMESPACES 10696a108a14SDavid Rientjes bool "Namespaces support" if EXPERT 10702813893fSIulia Manda depends on MULTIUSER 10716a108a14SDavid Rientjes default !EXPERT 1072c5289a69SPavel Emelyanov help 1073c5289a69SPavel Emelyanov Provides the way to make tasks work with different objects using 1074c5289a69SPavel Emelyanov the same id. For example same IPC id may refer to different objects 1075c5289a69SPavel Emelyanov or same user id or pid may refer to different tasks when used in 1076c5289a69SPavel Emelyanov different namespaces. 1077c5289a69SPavel Emelyanov 10788dd2a82cSDaniel Lezcanoif NAMESPACES 10798dd2a82cSDaniel Lezcano 108058bfdd6dSPavel Emelyanovconfig UTS_NS 108158bfdd6dSPavel Emelyanov bool "UTS namespace" 108217a6d441SDaniel Lezcano default y 108358bfdd6dSPavel Emelyanov help 108458bfdd6dSPavel Emelyanov In this namespace tasks see different info provided with the 108558bfdd6dSPavel Emelyanov uname() system call 108658bfdd6dSPavel Emelyanov 1087769071acSAndrei Vaginconfig TIME_NS 1088769071acSAndrei Vagin bool "TIME namespace" 1089660fd04fSThomas Gleixner depends on GENERIC_VDSO_TIME_NS 1090769071acSAndrei Vagin default y 1091769071acSAndrei Vagin help 1092769071acSAndrei Vagin In this namespace boottime and monotonic clocks can be set. 1093769071acSAndrei Vagin The time will keep going with the same pace. 1094769071acSAndrei Vagin 1095ae5e1b22SPavel Emelyanovconfig IPC_NS 1096ae5e1b22SPavel Emelyanov bool "IPC namespace" 10978dd2a82cSDaniel Lezcano depends on (SYSVIPC || POSIX_MQUEUE) 109817a6d441SDaniel Lezcano default y 1099ae5e1b22SPavel Emelyanov help 1100ae5e1b22SPavel Emelyanov In this namespace tasks work with IPC ids which correspond to 1101614b84cfSSerge E. Hallyn different IPC objects in different namespaces. 1102ae5e1b22SPavel Emelyanov 1103aee16ce7SPavel Emelyanovconfig USER_NS 110419c92399SKees Cook bool "User namespace" 11055673a94cSEric W. Biederman default n 1106aee16ce7SPavel Emelyanov help 1107aee16ce7SPavel Emelyanov This allows containers, i.e. vservers, to use user namespaces 1108aee16ce7SPavel Emelyanov to provide different user info for different servers. 1109e11f0ae3SEric W. Biederman 1110e11f0ae3SEric W. Biederman When user namespaces are enabled in the kernel it is 1111d886f4e4SJohannes Weiner recommended that the MEMCG option also be enabled and that 1112d886f4e4SJohannes Weiner user-space use the memory control groups to limit the amount 1113d886f4e4SJohannes Weiner of memory a memory unprivileged users can use. 1114e11f0ae3SEric W. Biederman 1115aee16ce7SPavel Emelyanov If unsure, say N. 1116aee16ce7SPavel Emelyanov 111774bd59bbSPavel Emelyanovconfig PID_NS 11189bd38c2cSDaniel Lezcano bool "PID Namespaces" 111917a6d441SDaniel Lezcano default y 112074bd59bbSPavel Emelyanov help 112112d2b8f9SHeikki Orsila Support process id namespaces. This allows having multiple 1122692105b8SMatt LaPlante processes with the same pid as long as they are in different 112374bd59bbSPavel Emelyanov pid namespaces. This is a building block of containers. 112474bd59bbSPavel Emelyanov 1125d6eb633fSMatt Helsleyconfig NET_NS 1126d6eb633fSMatt Helsley bool "Network namespace" 11278dd2a82cSDaniel Lezcano depends on NET 112817a6d441SDaniel Lezcano default y 1129d6eb633fSMatt Helsley help 1130d6eb633fSMatt Helsley Allow user space to create what appear to be multiple instances 1131d6eb633fSMatt Helsley of the network stack. 1132d6eb633fSMatt Helsley 11338dd2a82cSDaniel Lezcanoendif # NAMESPACES 11348dd2a82cSDaniel Lezcano 11355cb366bbSAdrian Reberconfig CHECKPOINT_RESTORE 11365cb366bbSAdrian Reber bool "Checkpoint/restore support" 11375cb366bbSAdrian Reber select PROC_CHILDREN 11385cb366bbSAdrian Reber default n 11395cb366bbSAdrian Reber help 11405cb366bbSAdrian Reber Enables additional kernel features in a sake of checkpoint/restore. 11415cb366bbSAdrian Reber In particular it adds auxiliary prctl codes to setup process text, 11425cb366bbSAdrian Reber data and heap segment sizes, and a few additional /proc filesystem 11435cb366bbSAdrian Reber entries. 11445cb366bbSAdrian Reber 11455cb366bbSAdrian Reber If unsure, say N here. 11465cb366bbSAdrian Reber 11475091faa4SMike Galbraithconfig SCHED_AUTOGROUP 11485091faa4SMike Galbraith bool "Automatic process group scheduling" 11495091faa4SMike Galbraith select CGROUPS 11505091faa4SMike Galbraith select CGROUP_SCHED 11515091faa4SMike Galbraith select FAIR_GROUP_SCHED 11525091faa4SMike Galbraith help 11535091faa4SMike Galbraith This option optimizes the scheduler for common desktop workloads by 11545091faa4SMike Galbraith automatically creating and populating task groups. This separation 11555091faa4SMike Galbraith of workloads isolates aggressive CPU burners (like build jobs) from 11565091faa4SMike Galbraith desktop applications. Task group autogeneration is currently based 11575091faa4SMike Galbraith upon task session. 11585091faa4SMike Galbraith 11597af37becSDaniel Lezcanoconfig SYSFS_DEPRECATED 11605d6a4ea5SFerenc Wagner bool "Enable deprecated sysfs features to support old userspace tools" 11617af37becSDaniel Lezcano depends on SYSFS 11627af37becSDaniel Lezcano default n 11637af37becSDaniel Lezcano help 11647af37becSDaniel Lezcano This option adds code that switches the layout of the "block" class 11657af37becSDaniel Lezcano devices, to not show up in /sys/class/block/, but only in 11667af37becSDaniel Lezcano /sys/block/. 11677af37becSDaniel Lezcano 11687af37becSDaniel Lezcano This switch is only active when the sysfs.deprecated=1 boot option is 11697af37becSDaniel Lezcano passed or the SYSFS_DEPRECATED_V2 option is set. 11707af37becSDaniel Lezcano 11717af37becSDaniel Lezcano This option allows new kernels to run on old distributions and tools, 11727af37becSDaniel Lezcano which might get confused by /sys/class/block/. Since 2007/2008 all 11737af37becSDaniel Lezcano major distributions and tools handle this just fine. 11747af37becSDaniel Lezcano 11757af37becSDaniel Lezcano Recent distributions and userspace tools after 2009/2010 depend on 11767af37becSDaniel Lezcano the existence of /sys/class/block/, and will not work with this 11777af37becSDaniel Lezcano option enabled. 11787af37becSDaniel Lezcano 11797af37becSDaniel Lezcano Only if you are using a new kernel on an old distribution, you might 11807af37becSDaniel Lezcano need to say Y here. 11817af37becSDaniel Lezcano 11827af37becSDaniel Lezcanoconfig SYSFS_DEPRECATED_V2 11835d6a4ea5SFerenc Wagner bool "Enable deprecated sysfs features by default" 11847af37becSDaniel Lezcano default n 11857af37becSDaniel Lezcano depends on SYSFS 11867af37becSDaniel Lezcano depends on SYSFS_DEPRECATED 11877af37becSDaniel Lezcano help 11887af37becSDaniel Lezcano Enable deprecated sysfs by default. 11897af37becSDaniel Lezcano 11907af37becSDaniel Lezcano See the CONFIG_SYSFS_DEPRECATED option for more details about this 11917af37becSDaniel Lezcano option. 11927af37becSDaniel Lezcano 11937af37becSDaniel Lezcano Only if you are using a new kernel on an old distribution, you might 11947af37becSDaniel Lezcano need to say Y here. Even then, odds are you would not need it 11957af37becSDaniel Lezcano enabled, you can always pass the boot option if absolutely necessary. 11967af37becSDaniel Lezcano 11977af37becSDaniel Lezcanoconfig RELAY 11987af37becSDaniel Lezcano bool "Kernel->user space relay support (formerly relayfs)" 119926b5679eSPeter Zijlstra select IRQ_WORK 12007af37becSDaniel Lezcano help 12017af37becSDaniel Lezcano This option enables support for relay interface support in 12027af37becSDaniel Lezcano certain file systems (such as debugfs). 12037af37becSDaniel Lezcano It is designed to provide an efficient mechanism for tools and 12047af37becSDaniel Lezcano facilities to relay large amounts of data from kernel space to 12057af37becSDaniel Lezcano user space. 12067af37becSDaniel Lezcano 12077af37becSDaniel Lezcano If unsure, say N. 12087af37becSDaniel Lezcano 1209f991633dSDimitri Gorokhovikconfig BLK_DEV_INITRD 1210f991633dSDimitri Gorokhovik bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support" 1211f991633dSDimitri Gorokhovik help 1212f991633dSDimitri Gorokhovik The initial RAM filesystem is a ramfs which is loaded by the 1213f991633dSDimitri Gorokhovik boot loader (loadlin or lilo) and that is mounted as root 1214f991633dSDimitri Gorokhovik before the normal boot procedure. It is typically used to 1215f991633dSDimitri Gorokhovik load modules needed to mount the "real" root file system, 12168c27ceffSMauro Carvalho Chehab etc. See <file:Documentation/admin-guide/initrd.rst> for details. 1217f991633dSDimitri Gorokhovik 1218f991633dSDimitri Gorokhovik If RAM disk support (BLK_DEV_RAM) is also included, this 1219f991633dSDimitri Gorokhovik also enables initial RAM disk (initrd) support and adds 1220f991633dSDimitri Gorokhovik 15 Kbytes (more on some other architectures) to the kernel size. 1221f991633dSDimitri Gorokhovik 1222f991633dSDimitri Gorokhovik If unsure say Y. 1223f991633dSDimitri Gorokhovik 1224c33df4eaSJean-Paul Samanif BLK_DEV_INITRD 1225c33df4eaSJean-Paul Saman 1226dbec4866SSam Ravnborgsource "usr/Kconfig" 1227dbec4866SSam Ravnborg 1228c33df4eaSJean-Paul Samanendif 1229c33df4eaSJean-Paul Saman 123076db5a27SMasami Hiramatsuconfig BOOT_CONFIG 123176db5a27SMasami Hiramatsu bool "Boot config support" 12322910b5aaSMasami Hiramatsu select BLK_DEV_INITRD 123376db5a27SMasami Hiramatsu help 123476db5a27SMasami Hiramatsu Extra boot config allows system admin to pass a config file as 123576db5a27SMasami Hiramatsu complemental extension of kernel cmdline when booting. 12360947db01SMasami Hiramatsu The boot config file must be attached at the end of initramfs 123785c46b78SMasami Hiramatsu with checksum, size and magic word. 12380947db01SMasami Hiramatsu See <file:Documentation/admin-guide/bootconfig.rst> for details. 123976db5a27SMasami Hiramatsu 124076db5a27SMasami Hiramatsu If unsure, say Y. 124176db5a27SMasami Hiramatsu 1242877417e6SArnd Bergmannchoice 1243877417e6SArnd Bergmann prompt "Compiler optimization level" 12442cc3ce24SUlf Magnusson default CC_OPTIMIZE_FOR_PERFORMANCE 1245877417e6SArnd Bergmann 1246877417e6SArnd Bergmannconfig CC_OPTIMIZE_FOR_PERFORMANCE 124715f5db60SMasahiro Yamada bool "Optimize for performance (-O2)" 1248877417e6SArnd Bergmann help 1249877417e6SArnd Bergmann This is the default optimization level for the kernel, building 1250877417e6SArnd Bergmann with the "-O2" compiler flag for best performance and most 1251877417e6SArnd Bergmann helpful compile-time warnings. 1252877417e6SArnd Bergmann 125315f5db60SMasahiro Yamadaconfig CC_OPTIMIZE_FOR_PERFORMANCE_O3 125415f5db60SMasahiro Yamada bool "Optimize more for performance (-O3)" 125515f5db60SMasahiro Yamada depends on ARC 1256c45b4f1fSLinus Torvalds help 125715f5db60SMasahiro Yamada Choosing this option will pass "-O3" to your compiler to optimize 125815f5db60SMasahiro Yamada the kernel yet more for performance. 1259c45b4f1fSLinus Torvalds 12605d20ee31SNicholas Pigginconfig CC_OPTIMIZE_FOR_SIZE 126115f5db60SMasahiro Yamada bool "Optimize for size (-Os)" 1262c45b4f1fSLinus Torvalds help 1263ce3b487fSMasahiro Yamada Choosing this option will pass "-Os" to your compiler resulting 1264ce3b487fSMasahiro Yamada in a smaller kernel. 1265c45b4f1fSLinus Torvalds 1266877417e6SArnd Bergmannendchoice 1267877417e6SArnd Bergmann 12685d20ee31SNicholas Pigginconfig HAVE_LD_DEAD_CODE_DATA_ELIMINATION 12695d20ee31SNicholas Piggin bool 12705d20ee31SNicholas Piggin help 12715d20ee31SNicholas Piggin This requires that the arch annotates or otherwise protects 12725d20ee31SNicholas Piggin its external entry points from being discarded. Linker scripts 12735d20ee31SNicholas Piggin must also merge .text.*, .data.*, and .bss.* correctly into 12745d20ee31SNicholas Piggin output sections. Care must be taken not to pull in unrelated 12755d20ee31SNicholas Piggin sections (e.g., '.text.init'). Typically '.' in section names 12765d20ee31SNicholas Piggin is used to distinguish them from label names / C identifiers. 12775d20ee31SNicholas Piggin 12785d20ee31SNicholas Pigginconfig LD_DEAD_CODE_DATA_ELIMINATION 12795d20ee31SNicholas Piggin bool "Dead code and data elimination (EXPERIMENTAL)" 12805d20ee31SNicholas Piggin depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION 12815d20ee31SNicholas Piggin depends on EXPERT 128216fd20aaSPaul Burton depends on !(FUNCTION_TRACER && CC_IS_GCC && GCC_VERSION < 40800) 1283e85d1d65SMasahiro Yamada depends on $(cc-option,-ffunction-sections -fdata-sections) 1284e85d1d65SMasahiro Yamada depends on $(ld-option,--gc-sections) 12855d20ee31SNicholas Piggin help 12868b9d2712SMasahiro Yamada Enable this if you want to do dead code and data elimination with 12878b9d2712SMasahiro Yamada the linker by compiling with -ffunction-sections -fdata-sections, 12888b9d2712SMasahiro Yamada and linking with --gc-sections. 12895d20ee31SNicholas Piggin 12905d20ee31SNicholas Piggin This can reduce on disk and in-memory size of the kernel 12915d20ee31SNicholas Piggin code and static data, particularly for small configs and 12925d20ee31SNicholas Piggin on small systems. This has the possibility of introducing 12935d20ee31SNicholas Piggin silently broken kernel if the required annotations are not 12945d20ee31SNicholas Piggin present. This option is not well tested yet, so use at your 12955d20ee31SNicholas Piggin own risk. 12965d20ee31SNicholas Piggin 12970847062aSRandy Dunlapconfig SYSCTL 12980847062aSRandy Dunlap bool 12990847062aSRandy Dunlap 1300657a5209SMike Frysingerconfig HAVE_UID16 1301657a5209SMike Frysinger bool 1302657a5209SMike Frysinger 1303657a5209SMike Frysingerconfig SYSCTL_EXCEPTION_TRACE 1304657a5209SMike Frysinger bool 1305657a5209SMike Frysinger help 1306657a5209SMike Frysinger Enable support for /proc/sys/debug/exception-trace. 1307657a5209SMike Frysinger 1308657a5209SMike Frysingerconfig SYSCTL_ARCH_UNALIGN_NO_WARN 1309657a5209SMike Frysinger bool 1310657a5209SMike Frysinger help 1311657a5209SMike Frysinger Enable support for /proc/sys/kernel/ignore-unaligned-usertrap 1312657a5209SMike Frysinger Allows arch to define/use @no_unaligned_warning to possibly warn 1313657a5209SMike Frysinger about unaligned access emulation going on under the hood. 1314657a5209SMike Frysinger 1315657a5209SMike Frysingerconfig SYSCTL_ARCH_UNALIGN_ALLOW 1316657a5209SMike Frysinger bool 1317657a5209SMike Frysinger help 1318657a5209SMike Frysinger Enable support for /proc/sys/kernel/unaligned-trap 1319657a5209SMike Frysinger Allows arches to define/use @unaligned_enabled to runtime toggle 1320657a5209SMike Frysinger the unaligned access emulation. 1321657a5209SMike Frysinger see arch/parisc/kernel/unaligned.c for reference 1322657a5209SMike Frysinger 1323657a5209SMike Frysingerconfig HAVE_PCSPKR_PLATFORM 1324657a5209SMike Frysinger bool 1325657a5209SMike Frysinger 1326f89b7755SAlexei Starovoitov# interpreter that classic socket filters depend on 1327f89b7755SAlexei Starovoitovconfig BPF 1328f89b7755SAlexei Starovoitov bool 1329f89b7755SAlexei Starovoitov 13306a108a14SDavid Rientjesmenuconfig EXPERT 13316a108a14SDavid Rientjes bool "Configure standard kernel features (expert users)" 1332f505c553SJosh Triplett # Unhide debug options, to make the on-by-default options visible 1333f505c553SJosh Triplett select DEBUG_KERNEL 13341da177e4SLinus Torvalds help 13351da177e4SLinus Torvalds This option allows certain base kernel options and settings 13361da177e4SLinus Torvalds to be disabled or tweaked. This is for specialized 13371da177e4SLinus Torvalds environments which can tolerate a "non-standard" kernel. 13381da177e4SLinus Torvalds Only use this if you really know what you are doing. 13391da177e4SLinus Torvalds 1340ae81f9e3SChuck Ebbertconfig UID16 13416a108a14SDavid Rientjes bool "Enable 16-bit UID system calls" if EXPERT 13422813893fSIulia Manda depends on HAVE_UID16 && MULTIUSER 1343ae81f9e3SChuck Ebbert default y 1344ae81f9e3SChuck Ebbert help 1345ae81f9e3SChuck Ebbert This enables the legacy 16-bit UID syscall wrappers. 1346ae81f9e3SChuck Ebbert 13472813893fSIulia Mandaconfig MULTIUSER 13482813893fSIulia Manda bool "Multiple users, groups and capabilities support" if EXPERT 13492813893fSIulia Manda default y 13502813893fSIulia Manda help 13512813893fSIulia Manda This option enables support for non-root users, groups and 13522813893fSIulia Manda capabilities. 13532813893fSIulia Manda 13542813893fSIulia Manda If you say N here, all processes will run with UID 0, GID 0, and all 13552813893fSIulia Manda possible capabilities. Saying N here also compiles out support for 13562813893fSIulia Manda system calls related to UIDs, GIDs, and capabilities, such as setuid, 13572813893fSIulia Manda setgid, and capset. 13582813893fSIulia Manda 13592813893fSIulia Manda If unsure, say Y here. 13602813893fSIulia Manda 1361f6187769SFabian Frederickconfig SGETMASK_SYSCALL 1362f6187769SFabian Frederick bool "sgetmask/ssetmask syscalls support" if EXPERT 1363a687a533SArnd Bergmann def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH 1364f6187769SFabian Frederick ---help--- 1365f6187769SFabian Frederick sys_sgetmask and sys_ssetmask are obsolete system calls 1366f6187769SFabian Frederick no longer supported in libc but still enabled by default in some 1367f6187769SFabian Frederick architectures. 1368f6187769SFabian Frederick 1369f6187769SFabian Frederick If unsure, leave the default option here. 1370f6187769SFabian Frederick 13716af9f7bfSFabian Frederickconfig SYSFS_SYSCALL 13726af9f7bfSFabian Frederick bool "Sysfs syscall support" if EXPERT 13736af9f7bfSFabian Frederick default y 13746af9f7bfSFabian Frederick ---help--- 13756af9f7bfSFabian Frederick sys_sysfs is an obsolete system call no longer supported in libc. 13766af9f7bfSFabian Frederick Note that disabling this option is more secure but might break 13776af9f7bfSFabian Frederick compatibility with some systems. 13786af9f7bfSFabian Frederick 13796af9f7bfSFabian Frederick If unsure say Y here. 13806af9f7bfSFabian Frederick 1381d1b069f5SRandy Dunlapconfig FHANDLE 1382d1b069f5SRandy Dunlap bool "open by fhandle syscalls" if EXPERT 1383d1b069f5SRandy Dunlap select EXPORTFS 1384d1b069f5SRandy Dunlap default y 1385d1b069f5SRandy Dunlap help 1386d1b069f5SRandy Dunlap If you say Y here, a user level program will be able to map 1387d1b069f5SRandy Dunlap file names to handle and then later use the handle for 1388d1b069f5SRandy Dunlap different file system operations. This is useful in implementing 1389d1b069f5SRandy Dunlap userspace file servers, which now track files using handles instead 1390d1b069f5SRandy Dunlap of names. The handle would remain the same even if file names 1391d1b069f5SRandy Dunlap get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2) 1392d1b069f5SRandy Dunlap syscalls. 1393d1b069f5SRandy Dunlap 1394baa73d9eSNicolas Pitreconfig POSIX_TIMERS 1395baa73d9eSNicolas Pitre bool "Posix Clocks & timers" if EXPERT 1396baa73d9eSNicolas Pitre default y 1397baa73d9eSNicolas Pitre help 1398baa73d9eSNicolas Pitre This includes native support for POSIX timers to the kernel. 1399baa73d9eSNicolas Pitre Some embedded systems have no use for them and therefore they 1400baa73d9eSNicolas Pitre can be configured out to reduce the size of the kernel image. 1401baa73d9eSNicolas Pitre 1402baa73d9eSNicolas Pitre When this option is disabled, the following syscalls won't be 1403baa73d9eSNicolas Pitre available: timer_create, timer_gettime: timer_getoverrun, 1404baa73d9eSNicolas Pitre timer_settime, timer_delete, clock_adjtime, getitimer, 1405baa73d9eSNicolas Pitre setitimer, alarm. Furthermore, the clock_settime, clock_gettime, 1406baa73d9eSNicolas Pitre clock_getres and clock_nanosleep syscalls will be limited to 1407baa73d9eSNicolas Pitre CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only. 1408baa73d9eSNicolas Pitre 1409baa73d9eSNicolas Pitre If unsure say y. 1410baa73d9eSNicolas Pitre 1411d59745ceSMatt Mackallconfig PRINTK 1412d59745ceSMatt Mackall default y 14136a108a14SDavid Rientjes bool "Enable support for printk" if EXPERT 141474876a98SFrederic Weisbecker select IRQ_WORK 1415d59745ceSMatt Mackall help 1416d59745ceSMatt Mackall This option enables normal printk support. Removing it 1417d59745ceSMatt Mackall eliminates most of the message strings from the kernel image 1418d59745ceSMatt Mackall and makes the kernel more or less silent. As this makes it 1419d59745ceSMatt Mackall very difficult to diagnose system problems, saying N here is 1420d59745ceSMatt Mackall strongly discouraged. 1421d59745ceSMatt Mackall 142242a0bb3fSPetr Mladekconfig PRINTK_NMI 142342a0bb3fSPetr Mladek def_bool y 142442a0bb3fSPetr Mladek depends on PRINTK 142542a0bb3fSPetr Mladek depends on HAVE_NMI 142642a0bb3fSPetr Mladek 1427c8538a7aSMatt Mackallconfig BUG 14286a108a14SDavid Rientjes bool "BUG() support" if EXPERT 1429c8538a7aSMatt Mackall default y 1430c8538a7aSMatt Mackall help 1431c8538a7aSMatt Mackall Disabling this option eliminates support for BUG and WARN, reducing 1432c8538a7aSMatt Mackall the size of your kernel image and potentially quietly ignoring 1433c8538a7aSMatt Mackall numerous fatal conditions. You should only consider disabling this 1434c8538a7aSMatt Mackall option for embedded systems with no facilities for reporting errors. 1435c8538a7aSMatt Mackall Just say Y. 1436c8538a7aSMatt Mackall 1437708e9a79SMatt Mackallconfig ELF_CORE 1438046d662fSAlex Kelly depends on COREDUMP 1439708e9a79SMatt Mackall default y 14406a108a14SDavid Rientjes bool "Enable ELF core dumps" if EXPERT 1441708e9a79SMatt Mackall help 1442708e9a79SMatt Mackall Enable support for generating core dumps. Disabling saves about 4k. 1443708e9a79SMatt Mackall 14448761f1abSRalf Baechle 1445e5e1d3cbSStas Sergeevconfig PCSPKR_PLATFORM 14466a108a14SDavid Rientjes bool "Enable PC-Speaker support" if EXPERT 14478761f1abSRalf Baechle depends on HAVE_PCSPKR_PLATFORM 144815f304b6SRalf Baechle select I8253_LOCK 1449e5e1d3cbSStas Sergeev default y 1450e5e1d3cbSStas Sergeev help 1451e5e1d3cbSStas Sergeev This option allows to disable the internal PC-Speaker 1452e5e1d3cbSStas Sergeev support, saving some memory. 1453e5e1d3cbSStas Sergeev 14541da177e4SLinus Torvaldsconfig BASE_FULL 14551da177e4SLinus Torvalds default y 14566a108a14SDavid Rientjes bool "Enable full-sized data structures for core" if EXPERT 14571da177e4SLinus Torvalds help 14581da177e4SLinus Torvalds Disabling this option reduces the size of miscellaneous core 14591da177e4SLinus Torvalds kernel data structures. This saves memory on small machines, 14601da177e4SLinus Torvalds but may reduce performance. 14611da177e4SLinus Torvalds 14621da177e4SLinus Torvaldsconfig FUTEX 14636a108a14SDavid Rientjes bool "Enable futex support" if EXPERT 14641da177e4SLinus Torvalds default y 1465bc2eecd7SNicolas Pitre imply RT_MUTEXES 14661da177e4SLinus Torvalds help 14671da177e4SLinus Torvalds Disabling this option will cause the kernel to be built without 14681da177e4SLinus Torvalds support for "fast userspace mutexes". The resulting kernel may not 14691da177e4SLinus Torvalds run glibc-based applications correctly. 14701da177e4SLinus Torvalds 1471bc2eecd7SNicolas Pitreconfig FUTEX_PI 1472bc2eecd7SNicolas Pitre bool 1473bc2eecd7SNicolas Pitre depends on FUTEX && RT_MUTEXES 1474bc2eecd7SNicolas Pitre default y 1475bc2eecd7SNicolas Pitre 147603b8c7b6SHeiko Carstensconfig HAVE_FUTEX_CMPXCHG 147703b8c7b6SHeiko Carstens bool 147862b4d204SJosh Triplett depends on FUTEX 147903b8c7b6SHeiko Carstens help 148003b8c7b6SHeiko Carstens Architectures should select this if futex_atomic_cmpxchg_inatomic() 148103b8c7b6SHeiko Carstens is implemented and always working. This removes a couple of runtime 148203b8c7b6SHeiko Carstens checks. 148303b8c7b6SHeiko Carstens 14841da177e4SLinus Torvaldsconfig EPOLL 14856a108a14SDavid Rientjes bool "Enable eventpoll support" if EXPERT 14861da177e4SLinus Torvalds default y 14871da177e4SLinus Torvalds help 14881da177e4SLinus Torvalds Disabling this option will cause the kernel to be built without 14891da177e4SLinus Torvalds support for epoll family of system calls. 14901da177e4SLinus Torvalds 1491fba2afaaSDavide Libenziconfig SIGNALFD 14926a108a14SDavid Rientjes bool "Enable signalfd() system call" if EXPERT 1493fba2afaaSDavide Libenzi default y 1494fba2afaaSDavide Libenzi help 1495fba2afaaSDavide Libenzi Enable the signalfd() system call that allows to receive signals 1496fba2afaaSDavide Libenzi on a file descriptor. 1497fba2afaaSDavide Libenzi 1498fba2afaaSDavide Libenzi If unsure, say Y. 1499fba2afaaSDavide Libenzi 1500b215e283SDavide Libenziconfig TIMERFD 15016a108a14SDavid Rientjes bool "Enable timerfd() system call" if EXPERT 1502b215e283SDavide Libenzi default y 1503b215e283SDavide Libenzi help 1504b215e283SDavide Libenzi Enable the timerfd() system call that allows to receive timer 1505b215e283SDavide Libenzi events on a file descriptor. 1506b215e283SDavide Libenzi 1507b215e283SDavide Libenzi If unsure, say Y. 1508b215e283SDavide Libenzi 1509e1ad7468SDavide Libenziconfig EVENTFD 15106a108a14SDavid Rientjes bool "Enable eventfd() system call" if EXPERT 1511e1ad7468SDavide Libenzi default y 1512e1ad7468SDavide Libenzi help 1513e1ad7468SDavide Libenzi Enable the eventfd() system call that allows to receive both 1514e1ad7468SDavide Libenzi kernel notification (ie. KAIO) or userspace notifications. 1515e1ad7468SDavide Libenzi 1516e1ad7468SDavide Libenzi If unsure, say Y. 1517e1ad7468SDavide Libenzi 15181da177e4SLinus Torvaldsconfig SHMEM 15196a108a14SDavid Rientjes bool "Use full shmem filesystem" if EXPERT 15201da177e4SLinus Torvalds default y 15211da177e4SLinus Torvalds depends on MMU 15221da177e4SLinus Torvalds help 15231da177e4SLinus Torvalds The shmem is an internal filesystem used to manage shared memory. 15241da177e4SLinus Torvalds It is backed by swap and manages resource limits. It is also exported 15251da177e4SLinus Torvalds to userspace as tmpfs if TMPFS is enabled. Disabling this 15261da177e4SLinus Torvalds option replaces shmem and tmpfs with the much simpler ramfs code, 15271da177e4SLinus Torvalds which may be appropriate on small systems without swap. 15281da177e4SLinus Torvalds 1529ebf3f09cSThomas Petazzoniconfig AIO 15306a108a14SDavid Rientjes bool "Enable AIO support" if EXPERT 1531ebf3f09cSThomas Petazzoni default y 1532ebf3f09cSThomas Petazzoni help 1533ebf3f09cSThomas Petazzoni This option enables POSIX asynchronous I/O which may by used 1534ebf3f09cSThomas Petazzoni by some high performance threaded applications. Disabling 1535ebf3f09cSThomas Petazzoni this option saves about 7k. 1536ebf3f09cSThomas Petazzoni 15372b188cc1SJens Axboeconfig IO_URING 15382b188cc1SJens Axboe bool "Enable IO uring support" if EXPERT 1539561fb04aSJens Axboe select IO_WQ 15402b188cc1SJens Axboe default y 15412b188cc1SJens Axboe help 15422b188cc1SJens Axboe This option enables support for the io_uring interface, enabling 15432b188cc1SJens Axboe applications to submit and complete IO through submission and 15442b188cc1SJens Axboe completion rings that are shared between the kernel and application. 15452b188cc1SJens Axboe 1546d3ac21caSJosh Triplettconfig ADVISE_SYSCALLS 1547d3ac21caSJosh Triplett bool "Enable madvise/fadvise syscalls" if EXPERT 1548d3ac21caSJosh Triplett default y 1549d3ac21caSJosh Triplett help 1550d3ac21caSJosh Triplett This option enables the madvise and fadvise syscalls, used by 1551d3ac21caSJosh Triplett applications to advise the kernel about their future memory or file 1552d3ac21caSJosh Triplett usage, improving performance. If building an embedded system where no 1553d3ac21caSJosh Triplett applications use these syscalls, you can disable this option to save 1554d3ac21caSJosh Triplett space. 1555d3ac21caSJosh Triplett 15565a281062SAndrea Arcangeliconfig HAVE_ARCH_USERFAULTFD_WP 15575a281062SAndrea Arcangeli bool 15585a281062SAndrea Arcangeli help 15595a281062SAndrea Arcangeli Arch has userfaultfd write protection support 15605a281062SAndrea Arcangeli 15615b25b13aSMathieu Desnoyersconfig MEMBARRIER 15625b25b13aSMathieu Desnoyers bool "Enable membarrier() system call" if EXPERT 15635b25b13aSMathieu Desnoyers default y 15645b25b13aSMathieu Desnoyers help 15655b25b13aSMathieu Desnoyers Enable the membarrier() system call that allows issuing memory 15665b25b13aSMathieu Desnoyers barriers across all running threads, which can be used to distribute 15675b25b13aSMathieu Desnoyers the cost of user-space memory barriers asymmetrically by transforming 15685b25b13aSMathieu Desnoyers pairs of memory barriers into pairs consisting of membarrier() and a 15695b25b13aSMathieu Desnoyers compiler barrier. 15705b25b13aSMathieu Desnoyers 15715b25b13aSMathieu Desnoyers If unsure, say Y. 15725b25b13aSMathieu Desnoyers 1573d1b069f5SRandy Dunlapconfig KALLSYMS 1574d1b069f5SRandy Dunlap bool "Load all symbols for debugging/ksymoops" if EXPERT 1575d1b069f5SRandy Dunlap default y 1576d1b069f5SRandy Dunlap help 1577d1b069f5SRandy Dunlap Say Y here to let the kernel print out symbolic crash information and 1578d1b069f5SRandy Dunlap symbolic stack backtraces. This increases the size of the kernel 1579d1b069f5SRandy Dunlap somewhat, as all symbols have to be loaded into the kernel image. 1580d1b069f5SRandy Dunlap 1581d1b069f5SRandy Dunlapconfig KALLSYMS_ALL 1582d1b069f5SRandy Dunlap bool "Include all symbols in kallsyms" 1583d1b069f5SRandy Dunlap depends on DEBUG_KERNEL && KALLSYMS 1584d1b069f5SRandy Dunlap help 1585d1b069f5SRandy Dunlap Normally kallsyms only contains the symbols of functions for nicer 1586d1b069f5SRandy Dunlap OOPS messages and backtraces (i.e., symbols from the text and inittext 1587d1b069f5SRandy Dunlap sections). This is sufficient for most cases. And only in very rare 1588d1b069f5SRandy Dunlap cases (e.g., when a debugger is used) all symbols are required (e.g., 1589d1b069f5SRandy Dunlap names of variables from the data sections, etc). 1590d1b069f5SRandy Dunlap 1591d1b069f5SRandy Dunlap This option makes sure that all symbols are loaded into the kernel 1592d1b069f5SRandy Dunlap image (i.e., symbols from all sections) in cost of increased kernel 1593d1b069f5SRandy Dunlap size (depending on the kernel configuration, it may be 300KiB or 1594d1b069f5SRandy Dunlap something like this). 1595d1b069f5SRandy Dunlap 1596d1b069f5SRandy Dunlap Say N unless you really need all symbols. 1597d1b069f5SRandy Dunlap 1598d1b069f5SRandy Dunlapconfig KALLSYMS_ABSOLUTE_PERCPU 1599d1b069f5SRandy Dunlap bool 1600d1b069f5SRandy Dunlap depends on KALLSYMS 1601d1b069f5SRandy Dunlap default X86_64 && SMP 1602d1b069f5SRandy Dunlap 1603d1b069f5SRandy Dunlapconfig KALLSYMS_BASE_RELATIVE 1604d1b069f5SRandy Dunlap bool 1605d1b069f5SRandy Dunlap depends on KALLSYMS 1606a687a533SArnd Bergmann default !IA64 1607d1b069f5SRandy Dunlap help 1608d1b069f5SRandy Dunlap Instead of emitting them as absolute values in the native word size, 1609d1b069f5SRandy Dunlap emit the symbol references in the kallsyms table as 32-bit entries, 1610d1b069f5SRandy Dunlap each containing a relative value in the range [base, base + U32_MAX] 1611d1b069f5SRandy Dunlap or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either 1612d1b069f5SRandy Dunlap an absolute value in the range [0, S32_MAX] or a relative value in the 1613d1b069f5SRandy Dunlap range [base, base + S32_MAX], where base is the lowest relative symbol 1614d1b069f5SRandy Dunlap address encountered in the image. 1615d1b069f5SRandy Dunlap 1616d1b069f5SRandy Dunlap On 64-bit builds, this reduces the size of the address table by 50%, 1617d1b069f5SRandy Dunlap but more importantly, it results in entries whose values are build 1618d1b069f5SRandy Dunlap time constants, and no relocation pass is required at runtime to fix 1619d1b069f5SRandy Dunlap up the entries based on the runtime load address of the kernel. 1620d1b069f5SRandy Dunlap 1621d1b069f5SRandy Dunlap# end of the "standard kernel features (expert users)" menu 1622d1b069f5SRandy Dunlap 1623d1b069f5SRandy Dunlap# syscall, maps, verifier 1624fc611f47SKP Singh 1625fc611f47SKP Singhconfig BPF_LSM 1626fc611f47SKP Singh bool "LSM Instrumentation with BPF" 16274edf16b7SKP Singh depends on BPF_EVENTS 1628fc611f47SKP Singh depends on BPF_SYSCALL 1629fc611f47SKP Singh depends on SECURITY 1630fc611f47SKP Singh depends on BPF_JIT 1631fc611f47SKP Singh help 1632fc611f47SKP Singh Enables instrumentation of the security hooks with eBPF programs for 1633fc611f47SKP Singh implementing dynamic MAC and Audit Policies. 1634fc611f47SKP Singh 1635fc611f47SKP Singh If you are unsure how to answer this question, answer N. 1636fc611f47SKP Singh 1637d1b069f5SRandy Dunlapconfig BPF_SYSCALL 1638d1b069f5SRandy Dunlap bool "Enable bpf() system call" 1639d1b069f5SRandy Dunlap select BPF 1640bae77c5eSSong Liu select IRQ_WORK 1641d1b069f5SRandy Dunlap default n 1642d1b069f5SRandy Dunlap help 1643d1b069f5SRandy Dunlap Enable the bpf() system call that allows to manipulate eBPF 1644d1b069f5SRandy Dunlap programs and maps via file descriptors. 1645d1b069f5SRandy Dunlap 164681c22041SDaniel Borkmannconfig ARCH_WANT_DEFAULT_BPF_JIT 164781c22041SDaniel Borkmann bool 164881c22041SDaniel Borkmann 1649290af866SAlexei Starovoitovconfig BPF_JIT_ALWAYS_ON 1650290af866SAlexei Starovoitov bool "Permanently enable BPF JIT and remove BPF interpreter" 1651290af866SAlexei Starovoitov depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT 1652290af866SAlexei Starovoitov help 1653290af866SAlexei Starovoitov Enables BPF JIT and removes BPF interpreter to avoid 1654290af866SAlexei Starovoitov speculative execution of BPF instructions by the interpreter 1655290af866SAlexei Starovoitov 165681c22041SDaniel Borkmannconfig BPF_JIT_DEFAULT_ON 165781c22041SDaniel Borkmann def_bool ARCH_WANT_DEFAULT_BPF_JIT || BPF_JIT_ALWAYS_ON 165881c22041SDaniel Borkmann depends on HAVE_EBPF_JIT && BPF_JIT 165981c22041SDaniel Borkmann 1660d1b069f5SRandy Dunlapconfig USERFAULTFD 1661d1b069f5SRandy Dunlap bool "Enable userfaultfd() system call" 1662d1b069f5SRandy Dunlap depends on MMU 1663d1b069f5SRandy Dunlap help 1664d1b069f5SRandy Dunlap Enable the userfaultfd() system call that allows to intercept and 1665d1b069f5SRandy Dunlap handle page faults in userland. 1666d1b069f5SRandy Dunlap 16673ccfebedSMathieu Desnoyersconfig ARCH_HAS_MEMBARRIER_CALLBACKS 16683ccfebedSMathieu Desnoyers bool 16693ccfebedSMathieu Desnoyers 167070216e18SMathieu Desnoyersconfig ARCH_HAS_MEMBARRIER_SYNC_CORE 167170216e18SMathieu Desnoyers bool 167270216e18SMathieu Desnoyers 1673d7822b1eSMathieu Desnoyersconfig RSEQ 1674d7822b1eSMathieu Desnoyers bool "Enable rseq() system call" if EXPERT 1675d7822b1eSMathieu Desnoyers default y 1676d7822b1eSMathieu Desnoyers depends on HAVE_RSEQ 1677d7822b1eSMathieu Desnoyers select MEMBARRIER 1678d7822b1eSMathieu Desnoyers help 1679d7822b1eSMathieu Desnoyers Enable the restartable sequences system call. It provides a 1680d7822b1eSMathieu Desnoyers user-space cache for the current CPU number value, which 1681d7822b1eSMathieu Desnoyers speeds up getting the current CPU number from user-space, 1682d7822b1eSMathieu Desnoyers as well as an ABI to speed up user-space operations on 1683d7822b1eSMathieu Desnoyers per-CPU data. 1684d7822b1eSMathieu Desnoyers 1685d7822b1eSMathieu Desnoyers If unsure, say Y. 1686d7822b1eSMathieu Desnoyers 1687d7822b1eSMathieu Desnoyersconfig DEBUG_RSEQ 1688d7822b1eSMathieu Desnoyers default n 1689d7822b1eSMathieu Desnoyers bool "Enabled debugging of rseq() system call" if EXPERT 1690d7822b1eSMathieu Desnoyers depends on RSEQ && DEBUG_KERNEL 1691d7822b1eSMathieu Desnoyers help 1692d7822b1eSMathieu Desnoyers Enable extra debugging checks for the rseq system call. 1693d7822b1eSMathieu Desnoyers 1694d7822b1eSMathieu Desnoyers If unsure, say N. 1695d7822b1eSMathieu Desnoyers 16966befe5f6SRandy Dunlapconfig EMBEDDED 16976befe5f6SRandy Dunlap bool "Embedded system" 16985d2acfc7SJosh Triplett option allnoconfig_y 16996befe5f6SRandy Dunlap select EXPERT 17006befe5f6SRandy Dunlap help 17016befe5f6SRandy Dunlap This option should be enabled if compiling the kernel for 17026befe5f6SRandy Dunlap an embedded system so certain expert options are available 17036befe5f6SRandy Dunlap for configuration. 17046befe5f6SRandy Dunlap 1705cdd6c482SIngo Molnarconfig HAVE_PERF_EVENTS 17060793a61dSThomas Gleixner bool 1707018df72dSMike Frysinger help 1708018df72dSMike Frysinger See tools/perf/design.txt for details. 17090793a61dSThomas Gleixner 1710906010b2SPeter Zijlstraconfig PERF_USE_VMALLOC 1711906010b2SPeter Zijlstra bool 1712906010b2SPeter Zijlstra help 1713906010b2SPeter Zijlstra See tools/perf/design.txt for details 1714906010b2SPeter Zijlstra 1715ad90a3deSWilliam Breathitt Grayconfig PC104 1716424529fbSWilliam Breathitt Gray bool "PC/104 support" if EXPERT 1717ad90a3deSWilliam Breathitt Gray help 1718ad90a3deSWilliam Breathitt Gray Expose PC/104 form factor device drivers and options available for 1719ad90a3deSWilliam Breathitt Gray selection and configuration. Enable this option if your target 1720ad90a3deSWilliam Breathitt Gray machine has a PC/104 bus. 1721ad90a3deSWilliam Breathitt Gray 172257c0c15bSIngo Molnarmenu "Kernel Performance Events And Counters" 17230793a61dSThomas Gleixner 1724cdd6c482SIngo Molnarconfig PERF_EVENTS 172557c0c15bSIngo Molnar bool "Kernel performance events and counters" 1726392d65a9SRobert Richter default y if PROFILING 1727cdd6c482SIngo Molnar depends on HAVE_PERF_EVENTS 1728e360adbeSPeter Zijlstra select IRQ_WORK 172983fe27eaSPranith Kumar select SRCU 17300793a61dSThomas Gleixner help 173157c0c15bSIngo Molnar Enable kernel support for various performance events provided 173257c0c15bSIngo Molnar by software and hardware. 17330793a61dSThomas Gleixner 1734dd77038dSThadeu Lima de Souza Cascardo Software events are supported either built-in or via the 173557c0c15bSIngo Molnar use of generic tracepoints. 173657c0c15bSIngo Molnar 173757c0c15bSIngo Molnar Most modern CPUs support performance events via performance 173857c0c15bSIngo Molnar counter registers. These registers count the number of certain 17390793a61dSThomas Gleixner types of hw events: such as instructions executed, cachemisses 17400793a61dSThomas Gleixner suffered, or branches mis-predicted - without slowing down the 17410793a61dSThomas Gleixner kernel or applications. These registers can also trigger interrupts 17420793a61dSThomas Gleixner when a threshold number of events have passed - and can thus be 17430793a61dSThomas Gleixner used to profile the code that runs on that CPU. 17440793a61dSThomas Gleixner 174557c0c15bSIngo Molnar The Linux Performance Event subsystem provides an abstraction of 1746dd77038dSThadeu Lima de Souza Cascardo these software and hardware event capabilities, available via a 174757c0c15bSIngo Molnar system call and used by the "perf" utility in tools/perf/. It 17480793a61dSThomas Gleixner provides per task and per CPU counters, and it provides event 17490793a61dSThomas Gleixner capabilities on top of those. 17500793a61dSThomas Gleixner 17510793a61dSThomas Gleixner Say Y if unsure. 17520793a61dSThomas Gleixner 1753906010b2SPeter Zijlstraconfig DEBUG_PERF_USE_VMALLOC 1754906010b2SPeter Zijlstra default n 1755906010b2SPeter Zijlstra bool "Debug: use vmalloc to back perf mmap() buffers" 1756cb307113SMichael Ellerman depends on PERF_EVENTS && DEBUG_KERNEL && !PPC 1757906010b2SPeter Zijlstra select PERF_USE_VMALLOC 1758906010b2SPeter Zijlstra help 1759906010b2SPeter Zijlstra Use vmalloc memory to back perf mmap() buffers. 1760906010b2SPeter Zijlstra 1761906010b2SPeter Zijlstra Mostly useful for debugging the vmalloc code on platforms 1762906010b2SPeter Zijlstra that don't require it. 1763906010b2SPeter Zijlstra 1764906010b2SPeter Zijlstra Say N if unsure. 1765906010b2SPeter Zijlstra 17660793a61dSThomas Gleixnerendmenu 17670793a61dSThomas Gleixner 1768f8891e5eSChristoph Lameterconfig VM_EVENT_COUNTERS 1769f8891e5eSChristoph Lameter default y 17706a108a14SDavid Rientjes bool "Enable VM event counters for /proc/vmstat" if EXPERT 1771f8891e5eSChristoph Lameter help 17722aea4fb6SPaul Jackson VM event counters are needed for event counts to be shown. 17732aea4fb6SPaul Jackson This option allows the disabling of the VM event counters 17746a108a14SDavid Rientjes on EXPERT systems. /proc/vmstat will only show page counts 17752aea4fb6SPaul Jackson if VM event counters are disabled. 1776f8891e5eSChristoph Lameter 177741ecc55bSChristoph Lameterconfig SLUB_DEBUG 177841ecc55bSChristoph Lameter default y 17796a108a14SDavid Rientjes bool "Enable SLUB debugging support" if EXPERT 1780f6acb635SChristoph Lameter depends on SLUB && SYSFS 178141ecc55bSChristoph Lameter help 178241ecc55bSChristoph Lameter SLUB has extensive debug support features. Disabling these can 178341ecc55bSChristoph Lameter result in significant savings in code size. This also disables 178441ecc55bSChristoph Lameter SLUB sysfs support. /sys/slab will not exist and there will be 178541ecc55bSChristoph Lameter no support for cache validation etc. 178641ecc55bSChristoph Lameter 17871663f26dSTejun Heoconfig SLUB_MEMCG_SYSFS_ON 17881663f26dSTejun Heo default n 17891663f26dSTejun Heo bool "Enable memcg SLUB sysfs support by default" if EXPERT 17901663f26dSTejun Heo depends on SLUB && SYSFS && MEMCG 17911663f26dSTejun Heo help 17921663f26dSTejun Heo SLUB creates a directory under /sys/kernel/slab for each 17931663f26dSTejun Heo allocation cache to host info and debug files. If memory 17941663f26dSTejun Heo cgroup is enabled, each cache can have per memory cgroup 17951663f26dSTejun Heo caches. SLUB can create the same sysfs directories for these 17961663f26dSTejun Heo caches under /sys/kernel/slab/CACHE/cgroup but it can lead 17971663f26dSTejun Heo to a very high number of debug files being created. This is 17981663f26dSTejun Heo controlled by slub_memcg_sysfs boot parameter and this 17991663f26dSTejun Heo config option determines the parameter's default value. 18001663f26dSTejun Heo 1801b943c460SRandy Dunlapconfig COMPAT_BRK 1802b943c460SRandy Dunlap bool "Disable heap randomization" 1803b943c460SRandy Dunlap default y 1804b943c460SRandy Dunlap help 1805b943c460SRandy Dunlap Randomizing heap placement makes heap exploits harder, but it 1806b943c460SRandy Dunlap also breaks ancient binaries (including anything libc5 based). 1807b943c460SRandy Dunlap This option changes the bootup default to heap randomization 1808692105b8SMatt LaPlante disabled, and can be overridden at runtime by setting 1809b943c460SRandy Dunlap /proc/sys/kernel/randomize_va_space to 2. 1810b943c460SRandy Dunlap 1811b943c460SRandy Dunlap On non-ancient distros (post-2000 ones) N is usually a safe choice. 1812b943c460SRandy Dunlap 181381819f0fSChristoph Lameterchoice 181481819f0fSChristoph Lameter prompt "Choose SLAB allocator" 1815a0acd820SChristoph Lameter default SLUB 181681819f0fSChristoph Lameter help 181781819f0fSChristoph Lameter This option allows to select a slab allocator. 181881819f0fSChristoph Lameter 181981819f0fSChristoph Lameterconfig SLAB 182081819f0fSChristoph Lameter bool "SLAB" 182104385fc5SKees Cook select HAVE_HARDENED_USERCOPY_ALLOCATOR 182281819f0fSChristoph Lameter help 182381819f0fSChristoph Lameter The regular slab allocator that is established and known to work 182434013886SChristoph Lameter well in all environments. It organizes cache hot objects in 182502f56210SSimon Arlott per cpu and per node queues. 182681819f0fSChristoph Lameter 182781819f0fSChristoph Lameterconfig SLUB 182881819f0fSChristoph Lameter bool "SLUB (Unqueued Allocator)" 1829ed18adc1SKees Cook select HAVE_HARDENED_USERCOPY_ALLOCATOR 183081819f0fSChristoph Lameter help 183181819f0fSChristoph Lameter SLUB is a slab allocator that minimizes cache line usage 183281819f0fSChristoph Lameter instead of managing queues of cached objects (SLAB approach). 183381819f0fSChristoph Lameter Per cpu caching is realized using slabs of objects instead 183481819f0fSChristoph Lameter of queues of objects. SLUB can use memory efficiently 183502f56210SSimon Arlott and has enhanced diagnostics. SLUB is the default choice for 183602f56210SSimon Arlott a slab allocator. 183781819f0fSChristoph Lameter 183881819f0fSChristoph Lameterconfig SLOB 18396a108a14SDavid Rientjes depends on EXPERT 184081819f0fSChristoph Lameter bool "SLOB (Simple Allocator)" 184181819f0fSChristoph Lameter help 184237291458SMatt Mackall SLOB replaces the stock allocator with a drastically simpler 184337291458SMatt Mackall allocator. SLOB is generally more space efficient but 184437291458SMatt Mackall does not perform as well on large systems. 184581819f0fSChristoph Lameter 184681819f0fSChristoph Lameterendchoice 184781819f0fSChristoph Lameter 18487660a6fdSKees Cookconfig SLAB_MERGE_DEFAULT 18497660a6fdSKees Cook bool "Allow slab caches to be merged" 18507660a6fdSKees Cook default y 18517660a6fdSKees Cook help 18527660a6fdSKees Cook For reduced kernel memory fragmentation, slab caches can be 18537660a6fdSKees Cook merged when they share the same size and other characteristics. 18547660a6fdSKees Cook This carries a risk of kernel heap overflows being able to 18557660a6fdSKees Cook overwrite objects from merged caches (and more easily control 18567660a6fdSKees Cook cache layout), which makes such heap attacks easier to exploit 18577660a6fdSKees Cook by attackers. By keeping caches unmerged, these kinds of exploits 18587660a6fdSKees Cook can usually only damage objects in the same cache. To disable 18597660a6fdSKees Cook merging at runtime, "slab_nomerge" can be passed on the kernel 18607660a6fdSKees Cook command line. 18617660a6fdSKees Cook 1862c7ce4f60SThomas Garnierconfig SLAB_FREELIST_RANDOM 1863c7ce4f60SThomas Garnier default n 1864210e7a43SThomas Garnier depends on SLAB || SLUB 1865c7ce4f60SThomas Garnier bool "SLAB freelist randomization" 1866c7ce4f60SThomas Garnier help 1867210e7a43SThomas Garnier Randomizes the freelist order used on creating new pages. This 1868c7ce4f60SThomas Garnier security feature reduces the predictability of the kernel slab 1869c7ce4f60SThomas Garnier allocator against heap overflows. 1870c7ce4f60SThomas Garnier 18712482ddecSKees Cookconfig SLAB_FREELIST_HARDENED 18722482ddecSKees Cook bool "Harden slab freelist metadata" 18732482ddecSKees Cook depends on SLUB 18742482ddecSKees Cook help 18752482ddecSKees Cook Many kernel heap attacks try to target slab cache metadata and 18762482ddecSKees Cook other infrastructure. This options makes minor performance 187792bae787SKees Cook sacrifices to harden the kernel slab allocator against common 18782482ddecSKees Cook freelist exploit methods. 18792482ddecSKees Cook 1880e900a918SDan Williamsconfig SHUFFLE_PAGE_ALLOCATOR 1881e900a918SDan Williams bool "Page allocator randomization" 1882e900a918SDan Williams default SLAB_FREELIST_RANDOM && ACPI_NUMA 1883e900a918SDan Williams help 1884e900a918SDan Williams Randomization of the page allocator improves the average 1885e900a918SDan Williams utilization of a direct-mapped memory-side-cache. See section 1886e900a918SDan Williams 5.2.27 Heterogeneous Memory Attribute Table (HMAT) in the ACPI 1887e900a918SDan Williams 6.2a specification for an example of how a platform advertises 1888e900a918SDan Williams the presence of a memory-side-cache. There are also incidental 1889e900a918SDan Williams security benefits as it reduces the predictability of page 1890e900a918SDan Williams allocations to compliment SLAB_FREELIST_RANDOM, but the 1891e900a918SDan Williams default granularity of shuffling on the "MAX_ORDER - 1" i.e, 1892e900a918SDan Williams 10th order of pages is selected based on cache utilization 1893e900a918SDan Williams benefits on x86. 1894e900a918SDan Williams 1895e900a918SDan Williams While the randomization improves cache utilization it may 1896e900a918SDan Williams negatively impact workloads on platforms without a cache. For 1897e900a918SDan Williams this reason, by default, the randomization is enabled only 1898e900a918SDan Williams after runtime detection of a direct-mapped memory-side-cache. 1899e900a918SDan Williams Otherwise, the randomization may be force enabled with the 1900e900a918SDan Williams 'page_alloc.shuffle' kernel command line parameter. 1901e900a918SDan Williams 1902e900a918SDan Williams Say Y if unsure. 1903e900a918SDan Williams 1904345c905dSJoonsoo Kimconfig SLUB_CPU_PARTIAL 1905345c905dSJoonsoo Kim default y 1906b39ffbf8SUwe Kleine-König depends on SLUB && SMP 1907345c905dSJoonsoo Kim bool "SLUB per cpu partial cache" 1908345c905dSJoonsoo Kim help 190992bae787SKees Cook Per cpu partial caches accelerate objects allocation and freeing 1910345c905dSJoonsoo Kim that is local to a processor at the price of more indeterminism 1911345c905dSJoonsoo Kim in the latency of the free. On overflow these caches will be cleared 1912345c905dSJoonsoo Kim which requires the taking of locks that may cause latency spikes. 1913345c905dSJoonsoo Kim Typically one would choose no for a realtime system. 1914345c905dSJoonsoo Kim 1915ea637639SJie Zhangconfig MMAP_ALLOW_UNINITIALIZED 1916ea637639SJie Zhang bool "Allow mmapped anonymous memory to be uninitialized" 19176a108a14SDavid Rientjes depends on EXPERT && !MMU 1918ea637639SJie Zhang default n 1919ea637639SJie Zhang help 1920ea637639SJie Zhang Normally, and according to the Linux spec, anonymous memory obtained 19213903bf94SRandy Dunlap from mmap() has its contents cleared before it is passed to 1922ea637639SJie Zhang userspace. Enabling this config option allows you to request that 1923ea637639SJie Zhang mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 1924ea637639SJie Zhang providing a huge performance boost. If this option is not enabled, 1925ea637639SJie Zhang then the flag will be ignored. 1926ea637639SJie Zhang 1927ea637639SJie Zhang This is taken advantage of by uClibc's malloc(), and also by 1928ea637639SJie Zhang ELF-FDPIC binfmt's brk and stack allocator. 1929ea637639SJie Zhang 1930ea637639SJie Zhang Because of the obvious security issues, this option should only be 1931ea637639SJie Zhang enabled on embedded devices where you control what is run in 1932ea637639SJie Zhang userspace. Since that isn't generally a problem on no-MMU systems, 1933ea637639SJie Zhang it is normally safe to say Y here. 1934ea637639SJie Zhang 1935ea637639SJie Zhang See Documentation/nommu-mmap.txt for more information. 1936ea637639SJie Zhang 1937091f6e26SDavid Howellsconfig SYSTEM_DATA_VERIFICATION 1938091f6e26SDavid Howells def_bool n 1939091f6e26SDavid Howells select SYSTEM_TRUSTED_KEYRING 1940091f6e26SDavid Howells select KEYS 1941091f6e26SDavid Howells select CRYPTO 1942d43de6c7SDavid Howells select CRYPTO_RSA 1943091f6e26SDavid Howells select ASYMMETRIC_KEY_TYPE 1944091f6e26SDavid Howells select ASYMMETRIC_PUBLIC_KEY_SUBTYPE 1945091f6e26SDavid Howells select ASN1 1946091f6e26SDavid Howells select OID_REGISTRY 1947091f6e26SDavid Howells select X509_CERTIFICATE_PARSER 1948091f6e26SDavid Howells select PKCS7_MESSAGE_PARSER 194982c04ff8SPeter Foley help 1950091f6e26SDavid Howells Provide PKCS#7 message verification using the contents of the system 1951091f6e26SDavid Howells trusted keyring to provide public keys. This then can be used for 1952091f6e26SDavid Howells module verification, kexec image verification and firmware blob 1953091f6e26SDavid Howells verification. 195482c04ff8SPeter Foley 1955125e5645SMathieu Desnoyersconfig PROFILING 1956b309a294SRobert Richter bool "Profiling support" 1957125e5645SMathieu Desnoyers help 1958125e5645SMathieu Desnoyers Say Y here to enable the extended profiling support mechanisms used 1959125e5645SMathieu Desnoyers by profilers such as OProfile. 1960125e5645SMathieu Desnoyers 19615f87f112SIngo Molnar# 19625f87f112SIngo Molnar# Place an empty function call at each tracepoint site. Can be 19635f87f112SIngo Molnar# dynamically changed for a probe function. 19645f87f112SIngo Molnar# 196597e1c18eSMathieu Desnoyersconfig TRACEPOINTS 19665f87f112SIngo Molnar bool 196797e1c18eSMathieu Desnoyers 19681da177e4SLinus Torvaldsendmenu # General setup 19691da177e4SLinus Torvalds 19701572497cSChristoph Hellwigsource "arch/Kconfig" 19711572497cSChristoph Hellwig 1972ae81f9e3SChuck Ebbertconfig RT_MUTEXES 19736341e62bSChristoph Jaeger bool 1974ae81f9e3SChuck Ebbert 19751da177e4SLinus Torvaldsconfig BASE_SMALL 19761da177e4SLinus Torvalds int 19771da177e4SLinus Torvalds default 0 if BASE_FULL 19781da177e4SLinus Torvalds default 1 if !BASE_FULL 19791da177e4SLinus Torvalds 1980c8424e77SThiago Jung Bauermannconfig MODULE_SIG_FORMAT 1981c8424e77SThiago Jung Bauermann def_bool n 1982c8424e77SThiago Jung Bauermann select SYSTEM_DATA_VERIFICATION 1983c8424e77SThiago Jung Bauermann 198466da5733SJan Engelhardtmenuconfig MODULES 19851da177e4SLinus Torvalds bool "Enable loadable module support" 198611097a03SYann E. MORIN option modules 19871da177e4SLinus Torvalds help 19881da177e4SLinus Torvalds Kernel modules are small pieces of compiled code which can 19891da177e4SLinus Torvalds be inserted in the running kernel, rather than being 19901da177e4SLinus Torvalds permanently built into the kernel. You use the "modprobe" 19911da177e4SLinus Torvalds tool to add (and sometimes remove) them. If you say Y here, 19921da177e4SLinus Torvalds many parts of the kernel can be built as modules (by 19931da177e4SLinus Torvalds answering M instead of Y where indicated): this is most 19941da177e4SLinus Torvalds useful for infrequently used options which are not required 19951da177e4SLinus Torvalds for booting. For more information, see the man pages for 19961da177e4SLinus Torvalds modprobe, lsmod, modinfo, insmod and rmmod. 19971da177e4SLinus Torvalds 19981da177e4SLinus Torvalds If you say Y here, you will need to run "make 19991da177e4SLinus Torvalds modules_install" to put the modules under /lib/modules/ 20001da177e4SLinus Torvalds where modprobe can find them (you may need to be root to do 20011da177e4SLinus Torvalds this). 20021da177e4SLinus Torvalds 20031da177e4SLinus Torvalds If unsure, say Y. 20041da177e4SLinus Torvalds 20050b0de144SRobert P. J. Dayif MODULES 20060b0de144SRobert P. J. Day 2007826e4506SLinus Torvaldsconfig MODULE_FORCE_LOAD 2008826e4506SLinus Torvalds bool "Forced module loading" 2009826e4506SLinus Torvalds default n 2010826e4506SLinus Torvalds help 201191e37a79SRusty Russell Allow loading of modules without version information (ie. modprobe 201291e37a79SRusty Russell --force). Forced module loading sets the 'F' (forced) taint flag and 201391e37a79SRusty Russell is usually a really bad idea. 2014826e4506SLinus Torvalds 20151da177e4SLinus Torvaldsconfig MODULE_UNLOAD 20161da177e4SLinus Torvalds bool "Module unloading" 20171da177e4SLinus Torvalds help 20181da177e4SLinus Torvalds Without this option you will not be able to unload any 20191da177e4SLinus Torvalds modules (note that some modules may not be unloadable 2020f7f5b675SDenys Vlasenko anyway), which makes your kernel smaller, faster 2021f7f5b675SDenys Vlasenko and simpler. If unsure, say Y. 20221da177e4SLinus Torvalds 20231da177e4SLinus Torvaldsconfig MODULE_FORCE_UNLOAD 20241da177e4SLinus Torvalds bool "Forced module unloading" 202519c92399SKees Cook depends on MODULE_UNLOAD 20261da177e4SLinus Torvalds help 20271da177e4SLinus Torvalds This option allows you to force a module to unload, even if the 20281da177e4SLinus Torvalds kernel believes it is unsafe: the kernel will remove the module 20291da177e4SLinus Torvalds without waiting for anyone to stop using it (using the -f option to 20301da177e4SLinus Torvalds rmmod). This is mainly for kernel developers and desperate users. 20311da177e4SLinus Torvalds If unsure, say N. 20321da177e4SLinus Torvalds 20331da177e4SLinus Torvaldsconfig MODVERSIONS 20340d541643SSam Ravnborg bool "Module versioning support" 20351da177e4SLinus Torvalds help 20361da177e4SLinus Torvalds Usually, you have to use modules compiled with your kernel. 20371da177e4SLinus Torvalds Saying Y here makes it sometimes possible to use modules 20381da177e4SLinus Torvalds compiled for different kernels, by adding enough information 20391da177e4SLinus Torvalds to the modules to (hopefully) spot any changes which would 20401da177e4SLinus Torvalds make them incompatible with the kernel you are running. If 20411da177e4SLinus Torvalds unsure, say N. 20421da177e4SLinus Torvalds 20432ff2b7ecSMasahiro Yamadaconfig ASM_MODVERSIONS 20442ff2b7ecSMasahiro Yamada bool 20452ff2b7ecSMasahiro Yamada default HAVE_ASM_MODVERSIONS && MODVERSIONS 20462ff2b7ecSMasahiro Yamada help 20472ff2b7ecSMasahiro Yamada This enables module versioning for exported symbols also from 20482ff2b7ecSMasahiro Yamada assembly. This can be enabled only when the target architecture 20492ff2b7ecSMasahiro Yamada supports it. 20502ff2b7ecSMasahiro Yamada 205156067812SArd Biesheuvelconfig MODULE_REL_CRCS 205256067812SArd Biesheuvel bool 205356067812SArd Biesheuvel depends on MODVERSIONS 205456067812SArd Biesheuvel 20551da177e4SLinus Torvaldsconfig MODULE_SRCVERSION_ALL 20561da177e4SLinus Torvalds bool "Source checksum for all modules" 20571da177e4SLinus Torvalds help 20581da177e4SLinus Torvalds Modules which contain a MODULE_VERSION get an extra "srcversion" 20591da177e4SLinus Torvalds field inserted into their modinfo section, which contains a 20601da177e4SLinus Torvalds sum of the source files which made it. This helps maintainers 20611da177e4SLinus Torvalds see exactly which source was used to build a module (since 20621da177e4SLinus Torvalds others sometimes change the module source without updating 20631da177e4SLinus Torvalds the version). With this option, such a "srcversion" field 20641da177e4SLinus Torvalds will be created for all modules. If unsure, say N. 20651da177e4SLinus Torvalds 2066106a4ee2SRusty Russellconfig MODULE_SIG 2067106a4ee2SRusty Russell bool "Module signature verification" 2068c8424e77SThiago Jung Bauermann select MODULE_SIG_FORMAT 2069106a4ee2SRusty Russell help 2070106a4ee2SRusty Russell Check modules for valid signatures upon load: the signature 2071106a4ee2SRusty Russell is simply appended to the module. For more information see 2072cbdc8217SNathan Chancellor <file:Documentation/admin-guide/module-signing.rst>. 2073106a4ee2SRusty Russell 2074228c37ffSDavid Howells Note that this option adds the OpenSSL development packages as a 2075228c37ffSDavid Howells kernel build dependency so that the signing tool can use its crypto 2076228c37ffSDavid Howells library. 2077228c37ffSDavid Howells 207849fcf732SDavid Howells You should enable this option if you wish to use either 207949fcf732SDavid Howells CONFIG_SECURITY_LOCKDOWN_LSM or lockdown functionality imposed via 208049fcf732SDavid Howells another LSM - otherwise unsigned modules will be loadable regardless 208149fcf732SDavid Howells of the lockdown policy. 208249fcf732SDavid Howells 2083ea0b6dcfSDavid Howells !!!WARNING!!! If you enable this option, you MUST make sure that the 2084ea0b6dcfSDavid Howells module DOES NOT get stripped after being signed. This includes the 2085ea0b6dcfSDavid Howells debuginfo strip done by some packagers (such as rpmbuild) and 2086ea0b6dcfSDavid Howells inclusion into an initramfs that wants the module size reduced. 2087ea0b6dcfSDavid Howells 2088106a4ee2SRusty Russellconfig MODULE_SIG_FORCE 2089106a4ee2SRusty Russell bool "Require modules to be validly signed" 2090106a4ee2SRusty Russell depends on MODULE_SIG 2091106a4ee2SRusty Russell help 2092106a4ee2SRusty Russell Reject unsigned modules or signed modules for which we don't have a 2093106a4ee2SRusty Russell key. Without this, such modules will simply taint the kernel. 2094ea0b6dcfSDavid Howells 2095d9d8d7edSMichal Marekconfig MODULE_SIG_ALL 2096d9d8d7edSMichal Marek bool "Automatically sign all modules" 2097d9d8d7edSMichal Marek default y 2098d9d8d7edSMichal Marek depends on MODULE_SIG 2099d9d8d7edSMichal Marek help 2100d9d8d7edSMichal Marek Sign all modules during make modules_install. Without this option, 2101d9d8d7edSMichal Marek modules must be signed manually, using the scripts/sign-file tool. 2102d9d8d7edSMichal Marek 2103d9d8d7edSMichal Marekcomment "Do not forget to sign required modules with scripts/sign-file" 2104d9d8d7edSMichal Marek depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL 2105d9d8d7edSMichal Marek 2106ea0b6dcfSDavid Howellschoice 2107ea0b6dcfSDavid Howells prompt "Which hash algorithm should modules be signed with?" 2108ea0b6dcfSDavid Howells depends on MODULE_SIG 2109ea0b6dcfSDavid Howells help 2110ea0b6dcfSDavid Howells This determines which sort of hashing algorithm will be used during 2111ea0b6dcfSDavid Howells signature generation. This algorithm _must_ be built into the kernel 2112ea0b6dcfSDavid Howells directly so that signature verification can take place. It is not 2113ea0b6dcfSDavid Howells possible to load a signed module containing the algorithm to check 2114ea0b6dcfSDavid Howells the signature on that module. 2115ea0b6dcfSDavid Howells 2116ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA1 2117ea0b6dcfSDavid Howells bool "Sign modules with SHA-1" 2118ea0b6dcfSDavid Howells select CRYPTO_SHA1 2119ea0b6dcfSDavid Howells 2120ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA224 2121ea0b6dcfSDavid Howells bool "Sign modules with SHA-224" 2122ea0b6dcfSDavid Howells select CRYPTO_SHA256 2123ea0b6dcfSDavid Howells 2124ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA256 2125ea0b6dcfSDavid Howells bool "Sign modules with SHA-256" 2126ea0b6dcfSDavid Howells select CRYPTO_SHA256 2127ea0b6dcfSDavid Howells 2128ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA384 2129ea0b6dcfSDavid Howells bool "Sign modules with SHA-384" 2130ea0b6dcfSDavid Howells select CRYPTO_SHA512 2131ea0b6dcfSDavid Howells 2132ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA512 2133ea0b6dcfSDavid Howells bool "Sign modules with SHA-512" 2134ea0b6dcfSDavid Howells select CRYPTO_SHA512 2135ea0b6dcfSDavid Howells 2136ea0b6dcfSDavid Howellsendchoice 2137ea0b6dcfSDavid Howells 213822753674SMichal Marekconfig MODULE_SIG_HASH 213922753674SMichal Marek string 214022753674SMichal Marek depends on MODULE_SIG 214122753674SMichal Marek default "sha1" if MODULE_SIG_SHA1 214222753674SMichal Marek default "sha224" if MODULE_SIG_SHA224 214322753674SMichal Marek default "sha256" if MODULE_SIG_SHA256 214422753674SMichal Marek default "sha384" if MODULE_SIG_SHA384 214522753674SMichal Marek default "sha512" if MODULE_SIG_SHA512 214622753674SMichal Marek 2147beb50df3SBertrand Jacquinconfig MODULE_COMPRESS 2148beb50df3SBertrand Jacquin bool "Compress modules on installation" 2149beb50df3SBertrand Jacquin help 2150beb50df3SBertrand Jacquin 2151b6c09b51SRusty Russell Compresses kernel modules when 'make modules_install' is run; gzip or 2152b6c09b51SRusty Russell xz depending on "Compression algorithm" below. 2153beb50df3SBertrand Jacquin 2154b6c09b51SRusty Russell module-init-tools MAY support gzip, and kmod MAY support gzip and xz. 2155beb50df3SBertrand Jacquin 2156b6c09b51SRusty Russell Out-of-tree kernel modules installed using Kbuild will also be 2157b6c09b51SRusty Russell compressed upon installation. 2158beb50df3SBertrand Jacquin 2159b6c09b51SRusty Russell Note: for modules inside an initrd or initramfs, it's more efficient 2160b6c09b51SRusty Russell to compress the whole initrd or initramfs instead. 2161beb50df3SBertrand Jacquin 2162b6c09b51SRusty Russell Note: This is fully compatible with signed modules. 2163b6c09b51SRusty Russell 2164b6c09b51SRusty Russell If in doubt, say N. 2165beb50df3SBertrand Jacquin 2166beb50df3SBertrand Jacquinchoice 2167beb50df3SBertrand Jacquin prompt "Compression algorithm" 2168beb50df3SBertrand Jacquin depends on MODULE_COMPRESS 2169beb50df3SBertrand Jacquin default MODULE_COMPRESS_GZIP 2170beb50df3SBertrand Jacquin help 2171beb50df3SBertrand Jacquin This determines which sort of compression will be used during 2172beb50df3SBertrand Jacquin 'make modules_install'. 2173beb50df3SBertrand Jacquin 2174beb50df3SBertrand Jacquin GZIP (default) and XZ are supported. 2175beb50df3SBertrand Jacquin 2176beb50df3SBertrand Jacquinconfig MODULE_COMPRESS_GZIP 2177beb50df3SBertrand Jacquin bool "GZIP" 2178beb50df3SBertrand Jacquin 2179beb50df3SBertrand Jacquinconfig MODULE_COMPRESS_XZ 2180beb50df3SBertrand Jacquin bool "XZ" 2181beb50df3SBertrand Jacquin 2182beb50df3SBertrand Jacquinendchoice 2183beb50df3SBertrand Jacquin 21843d52ec5eSMatthias Maennichconfig MODULE_ALLOW_MISSING_NAMESPACE_IMPORTS 21853d52ec5eSMatthias Maennich bool "Allow loading of modules with missing namespace imports" 21863d52ec5eSMatthias Maennich help 21873d52ec5eSMatthias Maennich Symbols exported with EXPORT_SYMBOL_NS*() are considered exported in 21883d52ec5eSMatthias Maennich a namespace. A module that makes use of a symbol exported with such a 21893d52ec5eSMatthias Maennich namespace is required to import the namespace via MODULE_IMPORT_NS(). 21903d52ec5eSMatthias Maennich There is no technical reason to enforce correct namespace imports, 21913d52ec5eSMatthias Maennich but it creates consistency between symbols defining namespaces and 21923d52ec5eSMatthias Maennich users importing namespaces they make use of. This option relaxes this 21933d52ec5eSMatthias Maennich requirement and lifts the enforcement when loading a module. 21943d52ec5eSMatthias Maennich 21953d52ec5eSMatthias Maennich If unsure, say N. 21963d52ec5eSMatthias Maennich 2197efd9763dSMasahiro Yamadaconfig UNUSED_SYMBOLS 2198efd9763dSMasahiro Yamada bool "Enable unused/obsolete exported symbols" 2199efd9763dSMasahiro Yamada default y if X86 2200efd9763dSMasahiro Yamada help 2201efd9763dSMasahiro Yamada Unused but exported symbols make the kernel needlessly bigger. For 2202efd9763dSMasahiro Yamada that reason most of these unused exports will soon be removed. This 2203efd9763dSMasahiro Yamada option is provided temporarily to provide a transition period in case 2204efd9763dSMasahiro Yamada some external kernel module needs one of these symbols anyway. If you 2205efd9763dSMasahiro Yamada encounter such a case in your module, consider if you are actually 2206efd9763dSMasahiro Yamada using the right API. (rationale: since nobody in the kernel is using 2207efd9763dSMasahiro Yamada this in a module, there is a pretty good chance it's actually the 2208efd9763dSMasahiro Yamada wrong interface to use). If you really need the symbol, please send a 2209efd9763dSMasahiro Yamada mail to the linux kernel mailing list mentioning the symbol and why 2210efd9763dSMasahiro Yamada you really need it, and what the merge plan to the mainline kernel for 2211efd9763dSMasahiro Yamada your module is. 2212efd9763dSMasahiro Yamada 2213dbacb0efSNicolas Pitreconfig TRIM_UNUSED_KSYMS 2214dbacb0efSNicolas Pitre bool "Trim unused exported kernel symbols" 2215d189c2a4SMasahiro Yamada depends on !UNUSED_SYMBOLS 2216dbacb0efSNicolas Pitre help 2217dbacb0efSNicolas Pitre The kernel and some modules make many symbols available for 2218dbacb0efSNicolas Pitre other modules to use via EXPORT_SYMBOL() and variants. Depending 2219dbacb0efSNicolas Pitre on the set of modules being selected in your kernel configuration, 2220dbacb0efSNicolas Pitre many of those exported symbols might never be used. 2221dbacb0efSNicolas Pitre 2222dbacb0efSNicolas Pitre This option allows for unused exported symbols to be dropped from 2223dbacb0efSNicolas Pitre the build. In turn, this provides the compiler more opportunities 2224dbacb0efSNicolas Pitre (especially when using LTO) for optimizing the code and reducing 2225dbacb0efSNicolas Pitre binary size. This might have some security advantages as well. 2226dbacb0efSNicolas Pitre 2227f1cb637eSValdis Kletnieks If unsure, or if you need to build out-of-tree modules, say N. 2228dbacb0efSNicolas Pitre 22291518c633SQuentin Perretconfig UNUSED_KSYMS_WHITELIST 22301518c633SQuentin Perret string "Whitelist of symbols to keep in ksymtab" 22311518c633SQuentin Perret depends on TRIM_UNUSED_KSYMS 22321518c633SQuentin Perret help 22331518c633SQuentin Perret By default, all unused exported symbols will be un-exported from the 22341518c633SQuentin Perret build when TRIM_UNUSED_KSYMS is selected. 22351518c633SQuentin Perret 22361518c633SQuentin Perret UNUSED_KSYMS_WHITELIST allows to whitelist symbols that must be kept 22371518c633SQuentin Perret exported at all times, even in absence of in-tree users. The value to 22381518c633SQuentin Perret set here is the path to a text file containing the list of symbols, 22391518c633SQuentin Perret one per line. The path can be absolute, or relative to the kernel 22401518c633SQuentin Perret source tree. 22411518c633SQuentin Perret 22420b0de144SRobert P. J. Dayendif # MODULES 22430b0de144SRobert P. J. Day 22446c9692e2SPeter Zijlstraconfig MODULES_TREE_LOOKUP 22456c9692e2SPeter Zijlstra def_bool y 22466c9692e2SPeter Zijlstra depends on PERF_EVENTS || TRACING 22476c9692e2SPeter Zijlstra 224898a79d6aSRusty Russellconfig INIT_ALL_POSSIBLE 224998a79d6aSRusty Russell bool 225098a79d6aSRusty Russell help 22515f054e31SRusty Russell Back when each arch used to define their own cpu_online_mask and 22525f054e31SRusty Russell cpu_possible_mask, some of them chose to initialize cpu_possible_mask 225398a79d6aSRusty Russell with all 1s, and others with all 0s. When they were centralised, 225498a79d6aSRusty Russell it was better to provide this option than to break all the archs 2255692105b8SMatt LaPlante and have several arch maintainers pursuing me down dark alleys. 225698a79d6aSRusty Russell 22573a65dfe8SJens Axboesource "block/Kconfig" 2258e98c3202SAvi Kivity 2259e98c3202SAvi Kivityconfig PREEMPT_NOTIFIERS 2260e98c3202SAvi Kivity bool 2261e260be67SPaul E. McKenney 226216295becSSteffen Klassertconfig PADATA 226316295becSSteffen Klassert depends on SMP 226416295becSSteffen Klassert bool 226516295becSSteffen Klassert 22664520c6a4SDavid Howellsconfig ASN1 22674520c6a4SDavid Howells tristate 22684520c6a4SDavid Howells help 22694520c6a4SDavid Howells Build a simple ASN.1 grammar compiler that produces a bytecode output 22704520c6a4SDavid Howells that can be interpreted by the ASN.1 stream decoder and used to 22714520c6a4SDavid Howells inform it as to what tags are to be expected in a stream and what 22724520c6a4SDavid Howells functions to call on what tags. 22734520c6a4SDavid Howells 22746beb0009SThomas Gleixnersource "kernel/Kconfig.locks" 2275e61938a9SMathieu Desnoyers 22760ebeea8cSDaniel Borkmannconfig ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE 22770ebeea8cSDaniel Borkmann bool 22780ebeea8cSDaniel Borkmann 2279e61938a9SMathieu Desnoyersconfig ARCH_HAS_SYNC_CORE_BEFORE_USERMODE 2280e61938a9SMathieu Desnoyers bool 22811bd21c6cSDominik Brodowski 22821bd21c6cSDominik Brodowski# It may be useful for an architecture to override the definitions of the 22837303e30eSDominik Brodowski# SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h> 22847303e30eSDominik Brodowski# and the COMPAT_ variants in <linux/compat.h>, in particular to use a 22857303e30eSDominik Brodowski# different calling convention for syscalls. They can also override the 22867303e30eSDominik Brodowski# macros for not-implemented syscalls in kernel/sys_ni.c and 22877303e30eSDominik Brodowski# kernel/time/posix-stubs.c. All these overrides need to be available in 22887303e30eSDominik Brodowski# <asm/syscall_wrapper.h>. 22891bd21c6cSDominik Brodowskiconfig ARCH_HAS_SYSCALL_WRAPPER 22901bd21c6cSDominik Brodowski def_bool n 2291