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 42b303c6dfSMasahiro Yamadaconfig CC_HAS_WARN_MAYBE_UNINITIALIZED 43b303c6dfSMasahiro Yamada def_bool $(cc-option,-Wmaybe-uninitialized) 44b303c6dfSMasahiro Yamada help 45b303c6dfSMasahiro Yamada GCC >= 4.7 supports this option. 46b303c6dfSMasahiro Yamada 47b303c6dfSMasahiro Yamadaconfig CC_DISABLE_WARN_MAYBE_UNINITIALIZED 48b303c6dfSMasahiro Yamada bool 49b303c6dfSMasahiro Yamada depends on CC_HAS_WARN_MAYBE_UNINITIALIZED 50b303c6dfSMasahiro Yamada default CC_IS_GCC && GCC_VERSION < 40900 # unreliable for GCC < 4.9 51b303c6dfSMasahiro Yamada help 52b303c6dfSMasahiro Yamada GCC's -Wmaybe-uninitialized is not reliable by definition. 53b303c6dfSMasahiro Yamada Lots of false positive warnings are produced in some cases. 54b303c6dfSMasahiro Yamada 55b303c6dfSMasahiro Yamada If this option is enabled, -Wno-maybe-uninitialzed is passed 56b303c6dfSMasahiro Yamada to the compiler to suppress maybe-uninitialized warnings. 57b303c6dfSMasahiro Yamada 58b99b87f7SPeter Oberparleiterconfig CONSTRUCTORS 59b99b87f7SPeter Oberparleiter bool 6087c9366eSJohannes Berg depends on !UML 61b99b87f7SPeter Oberparleiter 62e360adbeSPeter Zijlstraconfig IRQ_WORK 63e360adbeSPeter Zijlstra bool 64e360adbeSPeter Zijlstra 6510916706SShile Zhangconfig BUILDTIME_TABLE_SORT 661dbdc6f1SDavid Daney bool 671dbdc6f1SDavid Daney 68c65eacbeSAndy Lutomirskiconfig THREAD_INFO_IN_TASK 69c65eacbeSAndy Lutomirski bool 70c65eacbeSAndy Lutomirski help 71c65eacbeSAndy Lutomirski Select this to move thread_info off the stack into task_struct. To 72c65eacbeSAndy Lutomirski make this work, an arch will need to remove all thread_info fields 73c65eacbeSAndy Lutomirski except flags and fix any runtime bugs. 74c65eacbeSAndy Lutomirski 75c6c314a6SAndy Lutomirski One subtle change that will be needed is to use try_get_task_stack() 76c6c314a6SAndy Lutomirski and put_task_stack() in save_thread_stack_tsk() and get_wchan(). 77c6c314a6SAndy Lutomirski 78ff0cfc66SAl Boldimenu "General setup" 791da177e4SLinus Torvalds 801da177e4SLinus Torvaldsconfig BROKEN 811da177e4SLinus Torvalds bool 821da177e4SLinus Torvalds 831da177e4SLinus Torvaldsconfig BROKEN_ON_SMP 841da177e4SLinus Torvalds bool 851da177e4SLinus Torvalds depends on BROKEN || !SMP 861da177e4SLinus Torvalds default y 871da177e4SLinus Torvalds 881da177e4SLinus Torvaldsconfig INIT_ENV_ARG_LIMIT 891da177e4SLinus Torvalds int 90dd673bcaSAdrian Bunk default 32 if !UML 91dd673bcaSAdrian Bunk default 128 if UML 921da177e4SLinus Torvalds help 9334ad92c2SRandy Dunlap Maximum of each of the number of arguments and environment 9434ad92c2SRandy Dunlap variables passed to init from the kernel command line. 951da177e4SLinus Torvalds 964bb16672SJiri Slabyconfig COMPILE_TEST 974bb16672SJiri Slaby bool "Compile also drivers which will not load" 98bc083a64SRichard Weinberger depends on !UML 994bb16672SJiri Slaby default n 1004bb16672SJiri Slaby help 1014bb16672SJiri Slaby Some drivers can be compiled on a different platform than they are 1024bb16672SJiri Slaby intended to be run on. Despite they cannot be loaded there (or even 1034bb16672SJiri Slaby when they load they cannot be used due to missing HW support), 1044bb16672SJiri Slaby developers still, opposing to distributors, might want to build such 1054bb16672SJiri Slaby drivers to compile-test them. 1064bb16672SJiri Slaby 1074bb16672SJiri Slaby If you are a developer and want to build everything available, say Y 1084bb16672SJiri Slaby here. If you are a user/distributor, say N here to exclude useless 1094bb16672SJiri Slaby drivers to be distributed. 1104bb16672SJiri Slaby 111d6fc9fcbSMasahiro Yamadaconfig UAPI_HEADER_TEST 112d6fc9fcbSMasahiro Yamada bool "Compile test UAPI headers" 113fcbb8461SMasahiro Yamada depends on HEADERS_INSTALL && CC_CAN_LINK 114d6fc9fcbSMasahiro Yamada help 115d6fc9fcbSMasahiro Yamada Compile test headers exported to user-space to ensure they are 116d6fc9fcbSMasahiro Yamada self-contained, i.e. compilable as standalone units. 117d6fc9fcbSMasahiro Yamada 118d6fc9fcbSMasahiro Yamada If you are a developer or tester and want to ensure the exported 119d6fc9fcbSMasahiro Yamada headers are self-contained, say Y here. Otherwise, choose N. 120d6fc9fcbSMasahiro Yamada 1211da177e4SLinus Torvaldsconfig LOCALVERSION 1221da177e4SLinus Torvalds string "Local version - append to kernel release" 1231da177e4SLinus Torvalds help 1241da177e4SLinus Torvalds Append an extra string to the end of your kernel version. 1251da177e4SLinus Torvalds This will show up when you type uname, for example. 1261da177e4SLinus Torvalds The string you set here will be appended after the contents of 1271da177e4SLinus Torvalds any files with a filename matching localversion* in your 1281da177e4SLinus Torvalds object and source tree, in that order. Your total string can 1291da177e4SLinus Torvalds be a maximum of 64 characters. 1301da177e4SLinus Torvalds 131aaebf433SRyan Andersonconfig LOCALVERSION_AUTO 132aaebf433SRyan Anderson bool "Automatically append version information to the version string" 133aaebf433SRyan Anderson default y 134ac3339baSAlexey Dobriyan depends on !COMPILE_TEST 135aaebf433SRyan Anderson help 136aaebf433SRyan Anderson This will try to automatically determine if the current tree is a 1376e5a5420SRobert P. J. Day release tree by looking for git tags that belong to the current 1386e5a5420SRobert P. J. Day top of tree revision. 139aaebf433SRyan Anderson 140aaebf433SRyan Anderson A string of the format -gxxxxxxxx will be added to the localversion 1416e5a5420SRobert P. J. Day if a git-based tree is found. The string generated by this will be 142aaebf433SRyan Anderson appended after any matching localversion* files, and after the value 1436e5a5420SRobert P. J. Day set in CONFIG_LOCALVERSION. 144aaebf433SRyan Anderson 1456e5a5420SRobert P. J. Day (The actual string used here is the first eight characters produced 1466e5a5420SRobert P. J. Day by running the command: 1476e5a5420SRobert P. J. Day 1486e5a5420SRobert P. J. Day $ git rev-parse --verify HEAD 1496e5a5420SRobert P. J. Day 1506e5a5420SRobert P. J. Day which is done within the script "scripts/setlocalversion".) 151aaebf433SRyan Anderson 1529afb719eSLaura Abbottconfig BUILD_SALT 1539afb719eSLaura Abbott string "Build ID Salt" 1549afb719eSLaura Abbott default "" 1559afb719eSLaura Abbott help 1569afb719eSLaura Abbott The build ID is used to link binaries and their debug info. Setting 1579afb719eSLaura Abbott this option will use the value in the calculation of the build id. 1589afb719eSLaura Abbott This is mostly useful for distributions which want to ensure the 1599afb719eSLaura Abbott build is unique between builds. It's safe to leave the default. 1609afb719eSLaura Abbott 1612e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_GZIP 1622e9f3bddSH. Peter Anvin bool 1632e9f3bddSH. Peter Anvin 1642e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_BZIP2 1652e9f3bddSH. Peter Anvin bool 1662e9f3bddSH. Peter Anvin 1672e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_LZMA 1682e9f3bddSH. Peter Anvin bool 1692e9f3bddSH. Peter Anvin 1703ebe1243SLasse Collinconfig HAVE_KERNEL_XZ 1713ebe1243SLasse Collin bool 1723ebe1243SLasse Collin 1737dd65febSAlbin Tonnerreconfig HAVE_KERNEL_LZO 1747dd65febSAlbin Tonnerre bool 1757dd65febSAlbin Tonnerre 176e76e1fdfSKyungsik Leeconfig HAVE_KERNEL_LZ4 177e76e1fdfSKyungsik Lee bool 178e76e1fdfSKyungsik Lee 179f16466afSVasily Gorbikconfig HAVE_KERNEL_UNCOMPRESSED 180f16466afSVasily Gorbik bool 181f16466afSVasily Gorbik 18230d65dbfSAlain Knaffchoice 18330d65dbfSAlain Knaff prompt "Kernel compression mode" 18430d65dbfSAlain Knaff default KERNEL_GZIP 185f16466afSVasily Gorbik depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_UNCOMPRESSED 18630d65dbfSAlain Knaff help 18730d65dbfSAlain Knaff The linux kernel is a kind of self-extracting executable. 18830d65dbfSAlain Knaff Several compression algorithms are available, which differ 18930d65dbfSAlain Knaff in efficiency, compression and decompression speed. 19030d65dbfSAlain Knaff Compression speed is only relevant when building a kernel. 19130d65dbfSAlain Knaff Decompression speed is relevant at each boot. 19230d65dbfSAlain Knaff 19330d65dbfSAlain Knaff If you have any problems with bzip2 or lzma compressed 19430d65dbfSAlain Knaff kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older 19530d65dbfSAlain Knaff version of this functionality (bzip2 only), for 2.4, was 19630d65dbfSAlain Knaff supplied by Christian Ludwig) 19730d65dbfSAlain Knaff 19830d65dbfSAlain Knaff High compression options are mostly useful for users, who 19930d65dbfSAlain Knaff are low on disk space (embedded systems), but for whom ram 20030d65dbfSAlain Knaff size matters less. 20130d65dbfSAlain Knaff 20230d65dbfSAlain Knaff If in doubt, select 'gzip' 20330d65dbfSAlain Knaff 20430d65dbfSAlain Knaffconfig KERNEL_GZIP 20530d65dbfSAlain Knaff bool "Gzip" 2062e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_GZIP 20730d65dbfSAlain Knaff help 2087dd65febSAlbin Tonnerre The old and tried gzip compression. It provides a good balance 2097dd65febSAlbin Tonnerre between compression ratio and decompression speed. 21030d65dbfSAlain Knaff 21130d65dbfSAlain Knaffconfig KERNEL_BZIP2 21230d65dbfSAlain Knaff bool "Bzip2" 2132e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_BZIP2 21430d65dbfSAlain Knaff help 21530d65dbfSAlain Knaff Its compression ratio and speed is intermediate. 2160a4dd35cSRandy Dunlap Decompression speed is slowest among the choices. The kernel 2172e9f3bddSH. Peter Anvin size is about 10% smaller with bzip2, in comparison to gzip. 2182e9f3bddSH. Peter Anvin Bzip2 uses a large amount of memory. For modern kernels you 2192e9f3bddSH. Peter Anvin will need at least 8MB RAM or more for booting. 22030d65dbfSAlain Knaff 22130d65dbfSAlain Knaffconfig KERNEL_LZMA 22230d65dbfSAlain Knaff bool "LZMA" 2232e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_LZMA 22430d65dbfSAlain Knaff help 2250a4dd35cSRandy Dunlap This compression algorithm's ratio is best. Decompression speed 2260a4dd35cSRandy Dunlap is between gzip and bzip2. Compression is slowest. 2270a4dd35cSRandy Dunlap The kernel size is about 33% smaller with LZMA in comparison to gzip. 22830d65dbfSAlain Knaff 2293ebe1243SLasse Collinconfig KERNEL_XZ 2303ebe1243SLasse Collin bool "XZ" 2313ebe1243SLasse Collin depends on HAVE_KERNEL_XZ 2323ebe1243SLasse Collin help 2333ebe1243SLasse Collin XZ uses the LZMA2 algorithm and instruction set specific 2343ebe1243SLasse Collin BCJ filters which can improve compression ratio of executable 2353ebe1243SLasse Collin code. The size of the kernel is about 30% smaller with XZ in 2363ebe1243SLasse Collin comparison to gzip. On architectures for which there is a BCJ 2373ebe1243SLasse Collin filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ 2383ebe1243SLasse Collin will create a few percent smaller kernel than plain LZMA. 2393ebe1243SLasse Collin 2403ebe1243SLasse Collin The speed is about the same as with LZMA: The decompression 2413ebe1243SLasse Collin speed of XZ is better than that of bzip2 but worse than gzip 2423ebe1243SLasse Collin and LZO. Compression is slow. 2433ebe1243SLasse Collin 2447dd65febSAlbin Tonnerreconfig KERNEL_LZO 2457dd65febSAlbin Tonnerre bool "LZO" 2467dd65febSAlbin Tonnerre depends on HAVE_KERNEL_LZO 2477dd65febSAlbin Tonnerre help 2480a4dd35cSRandy Dunlap Its compression ratio is the poorest among the choices. The kernel 249681b3049SStephan Sperber size is about 10% bigger than gzip; however its speed 2507dd65febSAlbin Tonnerre (both compression and decompression) is the fastest. 2517dd65febSAlbin Tonnerre 252e76e1fdfSKyungsik Leeconfig KERNEL_LZ4 253e76e1fdfSKyungsik Lee bool "LZ4" 254e76e1fdfSKyungsik Lee depends on HAVE_KERNEL_LZ4 255e76e1fdfSKyungsik Lee help 256e76e1fdfSKyungsik Lee LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding. 257e76e1fdfSKyungsik Lee A preliminary version of LZ4 de/compression tool is available at 258e76e1fdfSKyungsik Lee <https://code.google.com/p/lz4/>. 259e76e1fdfSKyungsik Lee 260e76e1fdfSKyungsik Lee Its compression ratio is worse than LZO. The size of the kernel 261e76e1fdfSKyungsik Lee is about 8% bigger than LZO. But the decompression speed is 262e76e1fdfSKyungsik Lee faster than LZO. 263e76e1fdfSKyungsik Lee 264f16466afSVasily Gorbikconfig KERNEL_UNCOMPRESSED 265f16466afSVasily Gorbik bool "None" 266f16466afSVasily Gorbik depends on HAVE_KERNEL_UNCOMPRESSED 267f16466afSVasily Gorbik help 268f16466afSVasily Gorbik Produce uncompressed kernel image. This option is usually not what 269f16466afSVasily Gorbik you want. It is useful for debugging the kernel in slow simulation 270f16466afSVasily Gorbik environments, where decompressing and moving the kernel is awfully 271f16466afSVasily Gorbik slow. This option allows early boot code to skip the decompressor 272f16466afSVasily Gorbik and jump right at uncompressed kernel image. 273f16466afSVasily Gorbik 27430d65dbfSAlain Knaffendchoice 27530d65dbfSAlain Knaff 276bd5dc17bSJosh Triplettconfig DEFAULT_HOSTNAME 277bd5dc17bSJosh Triplett string "Default hostname" 278bd5dc17bSJosh Triplett default "(none)" 279bd5dc17bSJosh Triplett help 280bd5dc17bSJosh Triplett This option determines the default system hostname before userspace 281bd5dc17bSJosh Triplett calls sethostname(2). The kernel traditionally uses "(none)" here, 282bd5dc17bSJosh Triplett but you may wish to use a different default here to make a minimal 283bd5dc17bSJosh Triplett system more usable with less configuration. 284bd5dc17bSJosh Triplett 28517c46a6aSChristoph Hellwig# 28617c46a6aSChristoph Hellwig# For some reason microblaze and nios2 hard code SWAP=n. Hopefully we can 28717c46a6aSChristoph Hellwig# add proper SWAP support to them, in which case this can be remove. 28817c46a6aSChristoph Hellwig# 28917c46a6aSChristoph Hellwigconfig ARCH_NO_SWAP 29017c46a6aSChristoph Hellwig bool 29117c46a6aSChristoph Hellwig 2921da177e4SLinus Torvaldsconfig SWAP 2931da177e4SLinus Torvalds bool "Support for paging of anonymous memory (swap)" 29417c46a6aSChristoph Hellwig depends on MMU && BLOCK && !ARCH_NO_SWAP 2951da177e4SLinus Torvalds default y 2961da177e4SLinus Torvalds help 2971da177e4SLinus Torvalds This option allows you to choose whether you want to have support 2981da177e4SLinus Torvalds for so called swap devices or swap files in your kernel that are 2991da177e4SLinus Torvalds used to provide more virtual memory than the actual RAM present 3001da177e4SLinus Torvalds in your computer. If unsure say Y. 3011da177e4SLinus Torvalds 3021da177e4SLinus Torvaldsconfig SYSVIPC 3031da177e4SLinus Torvalds bool "System V IPC" 3041da177e4SLinus Torvalds ---help--- 3051da177e4SLinus Torvalds Inter Process Communication is a suite of library functions and 3061da177e4SLinus Torvalds system calls which let processes (running programs) synchronize and 3071da177e4SLinus Torvalds exchange information. It is generally considered to be a good thing, 3081da177e4SLinus Torvalds and some programs won't run unless you say Y here. In particular, if 3091da177e4SLinus Torvalds you want to run the DOS emulator dosemu under Linux (read the 3101da177e4SLinus Torvalds DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>), 3111da177e4SLinus Torvalds you'll need to say Y here. 3121da177e4SLinus Torvalds 3131da177e4SLinus Torvalds You can find documentation about IPC with "info ipc" and also in 3141da177e4SLinus Torvalds section 6.4 of the Linux Programmer's Guide, available from 3151da177e4SLinus Torvalds <http://www.tldp.org/guides.html>. 3161da177e4SLinus Torvalds 317a5494dcdSEric W. Biedermanconfig SYSVIPC_SYSCTL 318a5494dcdSEric W. Biederman bool 319a5494dcdSEric W. Biederman depends on SYSVIPC 320a5494dcdSEric W. Biederman depends on SYSCTL 321a5494dcdSEric W. Biederman default y 322a5494dcdSEric W. Biederman 3231da177e4SLinus Torvaldsconfig POSIX_MQUEUE 3241da177e4SLinus Torvalds bool "POSIX Message Queues" 32519c92399SKees Cook depends on NET 3261da177e4SLinus Torvalds ---help--- 3271da177e4SLinus Torvalds POSIX variant of message queues is a part of IPC. In POSIX message 3281da177e4SLinus Torvalds queues every message has a priority which decides about succession 3291da177e4SLinus Torvalds of receiving it by a process. If you want to compile and run 3301da177e4SLinus Torvalds programs written e.g. for Solaris with use of its POSIX message 331b0e37650SRobert P. J. Day queues (functions mq_*) say Y here. 3321da177e4SLinus Torvalds 3331da177e4SLinus Torvalds POSIX message queues are visible as a filesystem called 'mqueue' 3341da177e4SLinus Torvalds and can be mounted somewhere if you want to do filesystem 3351da177e4SLinus Torvalds operations on message queues. 3361da177e4SLinus Torvalds 3371da177e4SLinus Torvalds If unsure, say Y. 3381da177e4SLinus Torvalds 339bdc8e5f8SSerge E. Hallynconfig POSIX_MQUEUE_SYSCTL 340bdc8e5f8SSerge E. Hallyn bool 341bdc8e5f8SSerge E. Hallyn depends on POSIX_MQUEUE 342bdc8e5f8SSerge E. Hallyn depends on SYSCTL 343bdc8e5f8SSerge E. Hallyn default y 344bdc8e5f8SSerge E. Hallyn 345226b4ccdSKonstantin Khlebnikovconfig CROSS_MEMORY_ATTACH 346226b4ccdSKonstantin Khlebnikov bool "Enable process_vm_readv/writev syscalls" 347226b4ccdSKonstantin Khlebnikov depends on MMU 348226b4ccdSKonstantin Khlebnikov default y 349226b4ccdSKonstantin Khlebnikov help 350226b4ccdSKonstantin Khlebnikov Enabling this option adds the system calls process_vm_readv and 351226b4ccdSKonstantin Khlebnikov process_vm_writev which allow a process with the correct privileges 352a2a368d9SGeert Uytterhoeven to directly read from or write to another process' address space. 353226b4ccdSKonstantin Khlebnikov See the man page for more details. 354226b4ccdSKonstantin Khlebnikov 35569369a70SJosh Triplettconfig USELIB 35669369a70SJosh Triplett bool "uselib syscall" 357b2113a41SRiku Voipio def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION 35869369a70SJosh Triplett help 35969369a70SJosh Triplett This option enables the uselib syscall, a system call used in the 36069369a70SJosh Triplett dynamic linker from libc5 and earlier. glibc does not use this 36169369a70SJosh Triplett system call. If you intend to run programs built on libc5 or 36269369a70SJosh Triplett earlier, you may need to enable this syscall. Current systems 36369369a70SJosh Triplett running glibc can safely disable this. 36469369a70SJosh Triplett 3651da177e4SLinus Torvaldsconfig AUDIT 3661da177e4SLinus Torvalds bool "Auditing support" 367804a6a49SChris Wright depends on NET 3681da177e4SLinus Torvalds help 3691da177e4SLinus Torvalds Enable auditing infrastructure that can be used with another 3701da177e4SLinus Torvalds kernel subsystem, such as SELinux (which requires this for 371cb74ed27SPaul Moore logging of avc messages output). System call auditing is included 372cb74ed27SPaul Moore on architectures which support it. 3731da177e4SLinus Torvalds 3747a017721SAKASHI Takahiroconfig HAVE_ARCH_AUDITSYSCALL 3757a017721SAKASHI Takahiro bool 3767a017721SAKASHI Takahiro 3771da177e4SLinus Torvaldsconfig AUDITSYSCALL 378cb74ed27SPaul Moore def_bool y 3797a017721SAKASHI Takahiro depends on AUDIT && HAVE_ARCH_AUDITSYSCALL 38028a3a7ebSEric Paris select FSNOTIFY 38174c3cbe3SAl Viro 382d9817ebeSThomas Gleixnersource "kernel/irq/Kconfig" 383764e0da1SThomas Gleixnersource "kernel/time/Kconfig" 38487a4c375SChristoph Hellwigsource "kernel/Kconfig.preempt" 385d9817ebeSThomas Gleixner 386391dc69cSFrederic Weisbeckermenu "CPU/Task time and stats accounting" 387391dc69cSFrederic Weisbecker 388abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING 389abf917cdSFrederic Weisbecker bool 390abf917cdSFrederic Weisbecker 391fdf9c356SFrederic Weisbeckerchoice 392fdf9c356SFrederic Weisbecker prompt "Cputime accounting" 393fdf9c356SFrederic Weisbecker default TICK_CPU_ACCOUNTING if !PPC64 39402fc8d37SStephen Rothwell default VIRT_CPU_ACCOUNTING_NATIVE if PPC64 395fdf9c356SFrederic Weisbecker 396fdf9c356SFrederic Weisbecker# Kind of a stub config for the pure tick based cputime accounting 397fdf9c356SFrederic Weisbeckerconfig TICK_CPU_ACCOUNTING 398fdf9c356SFrederic Weisbecker bool "Simple tick based cputime accounting" 399c58b0df1SFrederic Weisbecker depends on !S390 && !NO_HZ_FULL 400fdf9c356SFrederic Weisbecker help 401fdf9c356SFrederic Weisbecker This is the basic tick based cputime accounting that maintains 402fdf9c356SFrederic Weisbecker statistics about user, system and idle time spent on per jiffies 403fdf9c356SFrederic Weisbecker granularity. 404fdf9c356SFrederic Weisbecker 405fdf9c356SFrederic Weisbecker If unsure, say Y. 406fdf9c356SFrederic Weisbecker 407abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING_NATIVE 408391dc69cSFrederic Weisbecker bool "Deterministic task and CPU time accounting" 409c58b0df1SFrederic Weisbecker depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL 410abf917cdSFrederic Weisbecker select VIRT_CPU_ACCOUNTING 411391dc69cSFrederic Weisbecker help 412391dc69cSFrederic Weisbecker Select this option to enable more accurate task and CPU time 413391dc69cSFrederic Weisbecker accounting. This is done by reading a CPU counter on each 414391dc69cSFrederic Weisbecker kernel entry and exit and on transitions within the kernel 415391dc69cSFrederic Weisbecker between system, softirq and hardirq state, so there is a 416391dc69cSFrederic Weisbecker small performance impact. In the case of s390 or IBM POWER > 5, 417391dc69cSFrederic Weisbecker this also enables accounting of stolen time on logically-partitioned 418391dc69cSFrederic Weisbecker systems. 419391dc69cSFrederic Weisbecker 420abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING_GEN 421abf917cdSFrederic Weisbecker bool "Full dynticks CPU time accounting" 422ff3fb254SKevin Hilman depends on HAVE_CONTEXT_TRACKING 423554b0004SKevin Hilman depends on HAVE_VIRT_CPU_ACCOUNTING_GEN 424041a1574SArnd Bergmann depends on GENERIC_CLOCKEVENTS 425abf917cdSFrederic Weisbecker select VIRT_CPU_ACCOUNTING 426abf917cdSFrederic Weisbecker select CONTEXT_TRACKING 427abf917cdSFrederic Weisbecker help 428abf917cdSFrederic Weisbecker Select this option to enable task and CPU time accounting on full 429abf917cdSFrederic Weisbecker dynticks systems. This accounting is implemented by watching every 430abf917cdSFrederic Weisbecker kernel-user boundaries using the context tracking subsystem. 431abf917cdSFrederic Weisbecker The accounting is thus performed at the expense of some significant 432abf917cdSFrederic Weisbecker overhead. 433abf917cdSFrederic Weisbecker 434abf917cdSFrederic Weisbecker For now this is only useful if you are working on the full 435abf917cdSFrederic Weisbecker dynticks subsystem development. 436abf917cdSFrederic Weisbecker 437abf917cdSFrederic Weisbecker If unsure, say N. 438abf917cdSFrederic Weisbecker 439b58c3584SRik van Rielendchoice 440b58c3584SRik van Riel 441fdf9c356SFrederic Weisbeckerconfig IRQ_TIME_ACCOUNTING 442fdf9c356SFrederic Weisbecker bool "Fine granularity task level IRQ time accounting" 443b58c3584SRik van Riel depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE 444fdf9c356SFrederic Weisbecker help 445fdf9c356SFrederic Weisbecker Select this option to enable fine granularity task irq time 446fdf9c356SFrederic Weisbecker accounting. This is done by reading a timestamp on each 447fdf9c356SFrederic Weisbecker transitions between softirq and hardirq state, so there can be a 448fdf9c356SFrederic Weisbecker small performance impact. 449fdf9c356SFrederic Weisbecker 450fdf9c356SFrederic Weisbecker If in doubt, say N here. 451fdf9c356SFrederic Weisbecker 45211d4afd4SVincent Guittotconfig HAVE_SCHED_AVG_IRQ 45311d4afd4SVincent Guittot def_bool y 45411d4afd4SVincent Guittot depends on IRQ_TIME_ACCOUNTING || PARAVIRT_TIME_ACCOUNTING 45511d4afd4SVincent Guittot depends on SMP 45611d4afd4SVincent Guittot 45776504793SThara Gopinathconfig SCHED_THERMAL_PRESSURE 45876504793SThara Gopinath bool "Enable periodic averaging of thermal pressure" 45976504793SThara Gopinath depends on SMP 46076504793SThara Gopinath 461391dc69cSFrederic Weisbeckerconfig BSD_PROCESS_ACCT 462391dc69cSFrederic Weisbecker bool "BSD Process Accounting" 4632813893fSIulia Manda depends on MULTIUSER 464391dc69cSFrederic Weisbecker help 465391dc69cSFrederic Weisbecker If you say Y here, a user level program will be able to instruct the 466391dc69cSFrederic Weisbecker kernel (via a special system call) to write process accounting 467391dc69cSFrederic Weisbecker information to a file: whenever a process exits, information about 468391dc69cSFrederic Weisbecker that process will be appended to the file by the kernel. The 469391dc69cSFrederic Weisbecker information includes things such as creation time, owning user, 470391dc69cSFrederic Weisbecker command name, memory usage, controlling terminal etc. (the complete 471391dc69cSFrederic Weisbecker list is in the struct acct in <file:include/linux/acct.h>). It is 472391dc69cSFrederic Weisbecker up to the user level program to do useful things with this 473391dc69cSFrederic Weisbecker information. This is generally a good idea, so say Y. 474391dc69cSFrederic Weisbecker 475391dc69cSFrederic Weisbeckerconfig BSD_PROCESS_ACCT_V3 476391dc69cSFrederic Weisbecker bool "BSD Process Accounting version 3 file format" 477391dc69cSFrederic Weisbecker depends on BSD_PROCESS_ACCT 478391dc69cSFrederic Weisbecker default n 479391dc69cSFrederic Weisbecker help 480391dc69cSFrederic Weisbecker If you say Y here, the process accounting information is written 481391dc69cSFrederic Weisbecker in a new file format that also logs the process IDs of each 4823903bf94SRandy Dunlap process and its parent. Note that this file format is incompatible 483391dc69cSFrederic Weisbecker with previous v0/v1/v2 file formats, so you will need updated tools 484391dc69cSFrederic Weisbecker for processing it. A preliminary version of these tools is available 485391dc69cSFrederic Weisbecker at <http://www.gnu.org/software/acct/>. 486391dc69cSFrederic Weisbecker 487391dc69cSFrederic Weisbeckerconfig TASKSTATS 48819c92399SKees Cook bool "Export task/process statistics through netlink" 489391dc69cSFrederic Weisbecker depends on NET 4902813893fSIulia Manda depends on MULTIUSER 491391dc69cSFrederic Weisbecker default n 492391dc69cSFrederic Weisbecker help 493391dc69cSFrederic Weisbecker Export selected statistics for tasks/processes through the 494391dc69cSFrederic Weisbecker generic netlink interface. Unlike BSD process accounting, the 495391dc69cSFrederic Weisbecker statistics are available during the lifetime of tasks/processes as 496391dc69cSFrederic Weisbecker responses to commands. Like BSD accounting, they are sent to user 497391dc69cSFrederic Weisbecker space on task exit. 498391dc69cSFrederic Weisbecker 499391dc69cSFrederic Weisbecker Say N if unsure. 500391dc69cSFrederic Weisbecker 501391dc69cSFrederic Weisbeckerconfig TASK_DELAY_ACCT 50219c92399SKees Cook bool "Enable per-task delay accounting" 503391dc69cSFrederic Weisbecker depends on TASKSTATS 504f6db8347SNaveen N. Rao select SCHED_INFO 505391dc69cSFrederic Weisbecker help 506391dc69cSFrederic Weisbecker Collect information on time spent by a task waiting for system 507391dc69cSFrederic Weisbecker resources like cpu, synchronous block I/O completion and swapping 508391dc69cSFrederic Weisbecker in pages. Such statistics can help in setting a task's priorities 509391dc69cSFrederic Weisbecker relative to other tasks for cpu, io, rss limits etc. 510391dc69cSFrederic Weisbecker 511391dc69cSFrederic Weisbecker Say N if unsure. 512391dc69cSFrederic Weisbecker 513391dc69cSFrederic Weisbeckerconfig TASK_XACCT 51419c92399SKees Cook bool "Enable extended accounting over taskstats" 515391dc69cSFrederic Weisbecker depends on TASKSTATS 516391dc69cSFrederic Weisbecker help 517391dc69cSFrederic Weisbecker Collect extended task accounting data and send the data 518391dc69cSFrederic Weisbecker to userland for processing over the taskstats interface. 519391dc69cSFrederic Weisbecker 520391dc69cSFrederic Weisbecker Say N if unsure. 521391dc69cSFrederic Weisbecker 522391dc69cSFrederic Weisbeckerconfig TASK_IO_ACCOUNTING 52319c92399SKees Cook bool "Enable per-task storage I/O accounting" 524391dc69cSFrederic Weisbecker depends on TASK_XACCT 525391dc69cSFrederic Weisbecker help 526391dc69cSFrederic Weisbecker Collect information on the number of bytes of storage I/O which this 527391dc69cSFrederic Weisbecker task has caused. 528391dc69cSFrederic Weisbecker 529391dc69cSFrederic Weisbecker Say N if unsure. 530391dc69cSFrederic Weisbecker 531eb414681SJohannes Weinerconfig PSI 532eb414681SJohannes Weiner bool "Pressure stall information tracking" 533eb414681SJohannes Weiner help 534eb414681SJohannes Weiner Collect metrics that indicate how overcommitted the CPU, memory, 535eb414681SJohannes Weiner and IO capacity are in the system. 536eb414681SJohannes Weiner 537eb414681SJohannes Weiner If you say Y here, the kernel will create /proc/pressure/ with the 538eb414681SJohannes Weiner pressure statistics files cpu, memory, and io. These will indicate 539eb414681SJohannes Weiner the share of walltime in which some or all tasks in the system are 540eb414681SJohannes Weiner delayed due to contention of the respective resource. 541eb414681SJohannes Weiner 5422ce7135aSJohannes Weiner In kernels with cgroup support, cgroups (cgroup2 only) will 5432ce7135aSJohannes Weiner have cpu.pressure, memory.pressure, and io.pressure files, 5442ce7135aSJohannes Weiner which aggregate pressure stalls for the grouped tasks only. 5452ce7135aSJohannes Weiner 546c3123552SMauro Carvalho Chehab For more details see Documentation/accounting/psi.rst. 547eb414681SJohannes Weiner 548eb414681SJohannes Weiner Say N if unsure. 549eb414681SJohannes Weiner 550e0c27447SJohannes Weinerconfig PSI_DEFAULT_DISABLED 551e0c27447SJohannes Weiner bool "Require boot parameter to enable pressure stall information tracking" 552e0c27447SJohannes Weiner default n 553e0c27447SJohannes Weiner depends on PSI 554e0c27447SJohannes Weiner help 555e0c27447SJohannes Weiner If set, pressure stall information tracking will be disabled 556428a1cb4SBaruch Siach per default but can be enabled through passing psi=1 on the 557428a1cb4SBaruch Siach kernel commandline during boot. 558e0c27447SJohannes Weiner 5597b2489d3SJohannes Weiner This feature adds some code to the task wakeup and sleep 5607b2489d3SJohannes Weiner paths of the scheduler. The overhead is too low to affect 5617b2489d3SJohannes Weiner common scheduling-intense workloads in practice (such as 5627b2489d3SJohannes Weiner webservers, memcache), but it does show up in artificial 5637b2489d3SJohannes Weiner scheduler stress tests, such as hackbench. 5647b2489d3SJohannes Weiner 5657b2489d3SJohannes Weiner If you are paranoid and not sure what the kernel will be 5667b2489d3SJohannes Weiner used for, say Y. 5677b2489d3SJohannes Weiner 5687b2489d3SJohannes Weiner Say N if unsure. 5697b2489d3SJohannes Weiner 570391dc69cSFrederic Weisbeckerendmenu # "CPU/Task time and stats accounting" 571391dc69cSFrederic Weisbecker 5725c4991e2SFrederic Weisbeckerconfig CPU_ISOLATION 5735c4991e2SFrederic Weisbecker bool "CPU isolation" 574414a2dc1SGeert Uytterhoeven depends on SMP || COMPILE_TEST 5752c43838cSFrederic Weisbecker default y 5765c4991e2SFrederic Weisbecker help 5775c4991e2SFrederic Weisbecker Make sure that CPUs running critical tasks are not disturbed by 5785c4991e2SFrederic Weisbecker any source of "noise" such as unbound workqueues, timers, kthreads... 5792c43838cSFrederic Weisbecker Unbound jobs get offloaded to housekeeping CPUs. This is driven by 5802c43838cSFrederic Weisbecker the "isolcpus=" boot parameter. 5812c43838cSFrederic Weisbecker 5822c43838cSFrederic Weisbecker Say Y if unsure. 5835c4991e2SFrederic Weisbecker 5840af92d46SPaul E. McKenneysource "kernel/rcu/Kconfig" 585c903ff83SMike Travis 586de5b56baSVivek Goyalconfig BUILD_BIN2C 587de5b56baSVivek Goyal bool 588de5b56baSVivek Goyal default n 589de5b56baSVivek Goyal 5901da177e4SLinus Torvaldsconfig IKCONFIG 591f2443ab6SRoss Biro tristate "Kernel .config support" 5921da177e4SLinus Torvalds ---help--- 5931da177e4SLinus Torvalds This option enables the complete Linux kernel ".config" file 5941da177e4SLinus Torvalds contents to be saved in the kernel. It provides documentation 5951da177e4SLinus Torvalds of which kernel options are used in a running kernel or in an 5961da177e4SLinus Torvalds on-disk kernel. This information can be extracted from the kernel 5971da177e4SLinus Torvalds image file with the script scripts/extract-ikconfig and used as 5981da177e4SLinus Torvalds input to rebuild the current kernel or to build another kernel. 5991da177e4SLinus Torvalds It can also be extracted from a running kernel by reading 6001da177e4SLinus Torvalds /proc/config.gz if enabled (below). 6011da177e4SLinus Torvalds 6021da177e4SLinus Torvaldsconfig IKCONFIG_PROC 6031da177e4SLinus Torvalds bool "Enable access to .config through /proc/config.gz" 6041da177e4SLinus Torvalds depends on IKCONFIG && PROC_FS 6051da177e4SLinus Torvalds ---help--- 6061da177e4SLinus Torvalds This option enables access to the kernel configuration file 6071da177e4SLinus Torvalds through /proc/config.gz. 6081da177e4SLinus Torvalds 609f7b101d3SJoel Fernandes (Google)config IKHEADERS 610f7b101d3SJoel Fernandes (Google) tristate "Enable kernel headers through /sys/kernel/kheaders.tar.xz" 611f7b101d3SJoel Fernandes (Google) depends on SYSFS 61243d8ce9dSJoel Fernandes (Google) help 613f7b101d3SJoel Fernandes (Google) This option enables access to the in-kernel headers that are generated during 614f7b101d3SJoel Fernandes (Google) the build process. These can be used to build eBPF tracing programs, 615f7b101d3SJoel Fernandes (Google) or similar programs. If you build the headers as a module, a module called 616f7b101d3SJoel Fernandes (Google) kheaders.ko is built which can be loaded on-demand to get access to headers. 61743d8ce9dSJoel Fernandes (Google) 618794543a2SAlistair John Strachanconfig LOG_BUF_SHIFT 619794543a2SAlistair John Strachan int "Kernel log buffer size (16 => 64KB, 17 => 128KB)" 620fb39f98dSIngo Molnar range 12 25 621f17a32e9SAdrian Bunk default 17 622361e9dfbSJosh Triplett depends on PRINTK 623794543a2SAlistair John Strachan help 62423b2899fSLuis R. Rodriguez Select the minimal kernel log buffer size as a power of 2. 62523b2899fSLuis R. Rodriguez The final size is affected by LOG_CPU_MAX_BUF_SHIFT config 62623b2899fSLuis R. Rodriguez parameter, see below. Any higher size also might be forced 62723b2899fSLuis R. Rodriguez by "log_buf_len" boot parameter. 62823b2899fSLuis R. Rodriguez 629f17a32e9SAdrian Bunk Examples: 630f17a32e9SAdrian Bunk 17 => 128 KB 631f17a32e9SAdrian Bunk 16 => 64 KB 632f17a32e9SAdrian Bunk 15 => 32 KB 633f17a32e9SAdrian Bunk 14 => 16 KB 634794543a2SAlistair John Strachan 13 => 8 KB 635794543a2SAlistair John Strachan 12 => 4 KB 636794543a2SAlistair John Strachan 63723b2899fSLuis R. Rodriguezconfig LOG_CPU_MAX_BUF_SHIFT 63823b2899fSLuis R. Rodriguez int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)" 6392240a31dSGeert Uytterhoeven depends on SMP 64023b2899fSLuis R. Rodriguez range 0 21 64123b2899fSLuis R. Rodriguez default 12 if !BASE_SMALL 64223b2899fSLuis R. Rodriguez default 0 if BASE_SMALL 643361e9dfbSJosh Triplett depends on PRINTK 64423b2899fSLuis R. Rodriguez help 64523b2899fSLuis R. Rodriguez This option allows to increase the default ring buffer size 64623b2899fSLuis R. Rodriguez according to the number of CPUs. The value defines the contribution 64723b2899fSLuis R. Rodriguez of each CPU as a power of 2. The used space is typically only few 64823b2899fSLuis R. Rodriguez lines however it might be much more when problems are reported, 64923b2899fSLuis R. Rodriguez e.g. backtraces. 65023b2899fSLuis R. Rodriguez 65123b2899fSLuis R. Rodriguez The increased size means that a new buffer has to be allocated and 65223b2899fSLuis R. Rodriguez the original static one is unused. It makes sense only on systems 65323b2899fSLuis R. Rodriguez with more CPUs. Therefore this value is used only when the sum of 65423b2899fSLuis R. Rodriguez contributions is greater than the half of the default kernel ring 65523b2899fSLuis R. Rodriguez buffer as defined by LOG_BUF_SHIFT. The default values are set 65623b2899fSLuis R. Rodriguez so that more than 64 CPUs are needed to trigger the allocation. 65723b2899fSLuis R. Rodriguez 65823b2899fSLuis R. Rodriguez Also this option is ignored when "log_buf_len" kernel parameter is 65923b2899fSLuis R. Rodriguez used as it forces an exact (power of two) size of the ring buffer. 66023b2899fSLuis R. Rodriguez 66123b2899fSLuis R. Rodriguez The number of possible CPUs is used for this computation ignoring 6625e0d8d59SGeert Uytterhoeven hotplugging making the computation optimal for the worst case 6635e0d8d59SGeert Uytterhoeven scenario while allowing a simple algorithm to be used from bootup. 66423b2899fSLuis R. Rodriguez 66523b2899fSLuis R. Rodriguez Examples shift values and their meaning: 66623b2899fSLuis R. Rodriguez 17 => 128 KB for each CPU 66723b2899fSLuis R. Rodriguez 16 => 64 KB for each CPU 66823b2899fSLuis R. Rodriguez 15 => 32 KB for each CPU 66923b2899fSLuis R. Rodriguez 14 => 16 KB for each CPU 67023b2899fSLuis R. Rodriguez 13 => 8 KB for each CPU 67123b2899fSLuis R. Rodriguez 12 => 4 KB for each CPU 67223b2899fSLuis R. Rodriguez 673f92bac3bSSergey Senozhatskyconfig PRINTK_SAFE_LOG_BUF_SHIFT 674f92bac3bSSergey Senozhatsky int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)" 675427934b8SPetr Mladek range 10 21 676427934b8SPetr Mladek default 13 677f92bac3bSSergey Senozhatsky depends on PRINTK 678427934b8SPetr Mladek help 679f92bac3bSSergey Senozhatsky Select the size of an alternate printk per-CPU buffer where messages 680f92bac3bSSergey Senozhatsky printed from usafe contexts are temporary stored. One example would 681f92bac3bSSergey Senozhatsky be NMI messages, another one - printk recursion. The messages are 682f92bac3bSSergey Senozhatsky copied to the main log buffer in a safe context to avoid a deadlock. 683f92bac3bSSergey Senozhatsky The value defines the size as a power of 2. 684427934b8SPetr Mladek 685f92bac3bSSergey Senozhatsky Those messages are rare and limited. The largest one is when 686427934b8SPetr Mladek a backtrace is printed. It usually fits into 4KB. Select 687427934b8SPetr Mladek 8KB if you want to be on the safe side. 688427934b8SPetr Mladek 689427934b8SPetr Mladek Examples: 690427934b8SPetr Mladek 17 => 128 KB for each CPU 691427934b8SPetr Mladek 16 => 64 KB for each CPU 692427934b8SPetr Mladek 15 => 32 KB for each CPU 693427934b8SPetr Mladek 14 => 16 KB for each CPU 694427934b8SPetr Mladek 13 => 8 KB for each CPU 695427934b8SPetr Mladek 12 => 4 KB for each CPU 696427934b8SPetr Mladek 6975cdc38f9SKAMEZAWA Hiroyuki# 6985cdc38f9SKAMEZAWA Hiroyuki# Architectures with an unreliable sched_clock() should select this: 6995cdc38f9SKAMEZAWA Hiroyuki# 7005cdc38f9SKAMEZAWA Hiroyukiconfig HAVE_UNSTABLE_SCHED_CLOCK 7015cdc38f9SKAMEZAWA Hiroyuki bool 7025cdc38f9SKAMEZAWA Hiroyuki 70338ff87f7SStephen Boydconfig GENERIC_SCHED_CLOCK 70438ff87f7SStephen Boyd bool 70538ff87f7SStephen Boyd 70669842cbaSPatrick Bellasimenu "Scheduler features" 70769842cbaSPatrick Bellasi 70869842cbaSPatrick Bellasiconfig UCLAMP_TASK 70969842cbaSPatrick Bellasi bool "Enable utilization clamping for RT/FAIR tasks" 71069842cbaSPatrick Bellasi depends on CPU_FREQ_GOV_SCHEDUTIL 71169842cbaSPatrick Bellasi help 71269842cbaSPatrick Bellasi This feature enables the scheduler to track the clamped utilization 71369842cbaSPatrick Bellasi of each CPU based on RUNNABLE tasks scheduled on that CPU. 71469842cbaSPatrick Bellasi 71569842cbaSPatrick Bellasi With this option, the user can specify the min and max CPU 71669842cbaSPatrick Bellasi utilization allowed for RUNNABLE tasks. The max utilization defines 71769842cbaSPatrick Bellasi the maximum frequency a task should use while the min utilization 71869842cbaSPatrick Bellasi defines the minimum frequency it should use. 71969842cbaSPatrick Bellasi 72069842cbaSPatrick Bellasi Both min and max utilization clamp values are hints to the scheduler, 72169842cbaSPatrick Bellasi aiming at improving its frequency selection policy, but they do not 72269842cbaSPatrick Bellasi enforce or grant any specific bandwidth for tasks. 72369842cbaSPatrick Bellasi 72469842cbaSPatrick Bellasi If in doubt, say N. 72569842cbaSPatrick Bellasi 72669842cbaSPatrick Bellasiconfig UCLAMP_BUCKETS_COUNT 72769842cbaSPatrick Bellasi int "Number of supported utilization clamp buckets" 72869842cbaSPatrick Bellasi range 5 20 72969842cbaSPatrick Bellasi default 5 73069842cbaSPatrick Bellasi depends on UCLAMP_TASK 73169842cbaSPatrick Bellasi help 73269842cbaSPatrick Bellasi Defines the number of clamp buckets to use. The range of each bucket 73369842cbaSPatrick Bellasi will be SCHED_CAPACITY_SCALE/UCLAMP_BUCKETS_COUNT. The higher the 73469842cbaSPatrick Bellasi number of clamp buckets the finer their granularity and the higher 73569842cbaSPatrick Bellasi the precision of clamping aggregation and tracking at run-time. 73669842cbaSPatrick Bellasi 73769842cbaSPatrick Bellasi For example, with the minimum configuration value we will have 5 73869842cbaSPatrick Bellasi clamp buckets tracking 20% utilization each. A 25% boosted tasks will 73969842cbaSPatrick Bellasi be refcounted in the [20..39]% bucket and will set the bucket clamp 74069842cbaSPatrick Bellasi effective value to 25%. 74169842cbaSPatrick Bellasi If a second 30% boosted task should be co-scheduled on the same CPU, 74269842cbaSPatrick Bellasi that task will be refcounted in the same bucket of the first task and 74369842cbaSPatrick Bellasi it will boost the bucket clamp effective value to 30%. 74469842cbaSPatrick Bellasi The clamp effective value of a bucket is reset to its nominal value 74569842cbaSPatrick Bellasi (20% in the example above) when there are no more tasks refcounted in 74669842cbaSPatrick Bellasi that bucket. 74769842cbaSPatrick Bellasi 74869842cbaSPatrick Bellasi An additional boost/capping margin can be added to some tasks. In the 74969842cbaSPatrick Bellasi example above the 25% task will be boosted to 30% until it exits the 75069842cbaSPatrick Bellasi CPU. If that should be considered not acceptable on certain systems, 75169842cbaSPatrick Bellasi it's always possible to reduce the margin by increasing the number of 75269842cbaSPatrick Bellasi clamp buckets to trade off used memory for run-time tracking 75369842cbaSPatrick Bellasi precision. 75469842cbaSPatrick Bellasi 75569842cbaSPatrick Bellasi If in doubt, use the default value. 75669842cbaSPatrick Bellasi 75769842cbaSPatrick Bellasiendmenu 75869842cbaSPatrick Bellasi 759be3a7284SAndrea Arcangeli# 760be3a7284SAndrea Arcangeli# For architectures that want to enable the support for NUMA-affine scheduler 761be3a7284SAndrea Arcangeli# balancing logic: 762be3a7284SAndrea Arcangeli# 763be3a7284SAndrea Arcangeliconfig ARCH_SUPPORTS_NUMA_BALANCING 764be3a7284SAndrea Arcangeli bool 765be3a7284SAndrea Arcangeli 766be5e610cSPeter Zijlstra# 76772b252aeSMel Gorman# For architectures that prefer to flush all TLBs after a number of pages 76872b252aeSMel Gorman# are unmapped instead of sending one IPI per page to flush. The architecture 76972b252aeSMel Gorman# must provide guarantees on what happens if a clean TLB cache entry is 77072b252aeSMel Gorman# written after the unmap. Details are in mm/rmap.c near the check for 77172b252aeSMel Gorman# should_defer_flush. The architecture should also consider if the full flush 77272b252aeSMel Gorman# and the refill costs are offset by the savings of sending fewer IPIs. 77372b252aeSMel Gormanconfig ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH 77472b252aeSMel Gorman bool 77572b252aeSMel Gorman 776c12d3362SArd Biesheuvelconfig CC_HAS_INT128 7773a7c7331SMasahiro Yamada def_bool !$(cc-option,$(m64-flag) -D__SIZEOF_INT128__=0) && 64BIT 778c12d3362SArd Biesheuvel 77972b252aeSMel Gorman# 780be5e610cSPeter Zijlstra# For architectures that know their GCC __int128 support is sound 781be5e610cSPeter Zijlstra# 782be5e610cSPeter Zijlstraconfig ARCH_SUPPORTS_INT128 783be5e610cSPeter Zijlstra bool 784be5e610cSPeter Zijlstra 785be3a7284SAndrea Arcangeli# For architectures that (ab)use NUMA to represent different memory regions 786be3a7284SAndrea Arcangeli# all cpu-local but of different latencies, such as SuperH. 787be3a7284SAndrea Arcangeli# 788be3a7284SAndrea Arcangeliconfig ARCH_WANT_NUMA_VARIABLE_LOCALITY 789be3a7284SAndrea Arcangeli bool 790be3a7284SAndrea Arcangeli 791be3a7284SAndrea Arcangeliconfig NUMA_BALANCING 792be3a7284SAndrea Arcangeli bool "Memory placement aware NUMA scheduler" 793be3a7284SAndrea Arcangeli depends on ARCH_SUPPORTS_NUMA_BALANCING 794be3a7284SAndrea Arcangeli depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY 795be3a7284SAndrea Arcangeli depends on SMP && NUMA && MIGRATION 796be3a7284SAndrea Arcangeli help 797be3a7284SAndrea Arcangeli This option adds support for automatic NUMA aware memory/task placement. 798be3a7284SAndrea Arcangeli The mechanism is quite primitive and is based on migrating memory when 7996d56a410SPaul Gortmaker it has references to the node the task is running on. 800be3a7284SAndrea Arcangeli 801be3a7284SAndrea Arcangeli This system will be inactive on UMA systems. 802be3a7284SAndrea Arcangeli 8036f7c97e8SAneesh Kumar K.Vconfig NUMA_BALANCING_DEFAULT_ENABLED 8046f7c97e8SAneesh Kumar K.V bool "Automatically enable NUMA aware memory/task placement" 8056f7c97e8SAneesh Kumar K.V default y 8066f7c97e8SAneesh Kumar K.V depends on NUMA_BALANCING 8076f7c97e8SAneesh Kumar K.V help 8086f7c97e8SAneesh Kumar K.V If set, automatic NUMA balancing will be enabled if running on a NUMA 8096f7c97e8SAneesh Kumar K.V machine. 8106f7c97e8SAneesh Kumar K.V 81123964d2dSLi Zefanmenuconfig CGROUPS 8126341e62bSChristoph Jaeger bool "Control Group support" 8132bd59d48STejun Heo select KERNFS 814ddbcc7e8SPaul Menage help 81523964d2dSLi Zefan This option adds support for grouping sets of processes together, for 8165cdc38f9SKAMEZAWA Hiroyuki use with process control subsystems such as Cpusets, CFS, memory 8175cdc38f9SKAMEZAWA Hiroyuki controls or device isolation. 8185cdc38f9SKAMEZAWA Hiroyuki See 819d6a3b247SMauro Carvalho Chehab - Documentation/scheduler/sched-design-CFS.rst (CFS) 820da82c92fSMauro Carvalho Chehab - Documentation/admin-guide/cgroup-v1/ (features for grouping, isolation 82145ce80fbSLi Zefan and resource control) 822ddbcc7e8SPaul Menage 823ddbcc7e8SPaul Menage Say N if unsure. 824ddbcc7e8SPaul Menage 82523964d2dSLi Zefanif CGROUPS 82623964d2dSLi Zefan 8273e32cb2eSJohannes Weinerconfig PAGE_COUNTER 8283e32cb2eSJohannes Weiner bool 8293e32cb2eSJohannes Weiner 830c255a458SAndrew Mortonconfig MEMCG 831a0166ec4SJohannes Weiner bool "Memory controller" 8323e32cb2eSJohannes Weiner select PAGE_COUNTER 83379bd9814STejun Heo select EVENTFD 83400f0b825SBalbir Singh help 835a0166ec4SJohannes Weiner Provides control over the memory footprint of tasks in a cgroup. 83600f0b825SBalbir Singh 837c255a458SAndrew Mortonconfig MEMCG_SWAP 838a0166ec4SJohannes Weiner bool "Swap controller" 839c255a458SAndrew Morton depends on MEMCG && SWAP 840c077719bSKAMEZAWA Hiroyuki help 841a0166ec4SJohannes Weiner Provides control over the swap space consumed by tasks in a cgroup. 842a0166ec4SJohannes Weiner 843c255a458SAndrew Mortonconfig MEMCG_SWAP_ENABLED 844a0166ec4SJohannes Weiner bool "Swap controller enabled by default" 845c255a458SAndrew Morton depends on MEMCG_SWAP 846a42c390cSMichal Hocko default y 847a42c390cSMichal Hocko help 848a42c390cSMichal Hocko Memory Resource Controller Swap Extension comes with its price in 849a42c390cSMichal Hocko a bigger memory consumption. General purpose distribution kernels 85043d547f9SJim Cromie which want to enable the feature but keep it disabled by default 85107555ac1SMichal Hocko and let the user enable it by swapaccount=1 boot command line 852a42c390cSMichal Hocko parameter should have this option unselected. 853a42c390cSMichal Hocko For those who want to have the feature enabled by default should 854a42c390cSMichal Hocko select this option (if, for some reason, they need to disable it 85500a66d29SWANG Cong then swapaccount=0 does the trick). 856c077719bSKAMEZAWA Hiroyuki 85784c07d11SKirill Tkhaiconfig MEMCG_KMEM 85884c07d11SKirill Tkhai bool 85984c07d11SKirill Tkhai depends on MEMCG && !SLOB 86084c07d11SKirill Tkhai default y 86184c07d11SKirill Tkhai 8626bf024e6SJohannes Weinerconfig BLK_CGROUP 8636bf024e6SJohannes Weiner bool "IO controller" 8646bf024e6SJohannes Weiner depends on BLOCK 8652bc64a20SAneesh Kumar K.V default n 8666bf024e6SJohannes Weiner ---help--- 8676bf024e6SJohannes Weiner Generic block IO controller cgroup interface. This is the common 8686bf024e6SJohannes Weiner cgroup interface which should be used by various IO controlling 8696bf024e6SJohannes Weiner policies. 8702bc64a20SAneesh Kumar K.V 8716bf024e6SJohannes Weiner Currently, CFQ IO scheduler uses it to recognize task groups and 8726bf024e6SJohannes Weiner control disk bandwidth allocation (proportional time slice allocation) 8736bf024e6SJohannes Weiner to such task groups. It is also used by bio throttling logic in 8746bf024e6SJohannes Weiner block layer to implement upper limit in IO rates on a device. 875e5d1367fSStephane Eranian 8766bf024e6SJohannes Weiner This option only enables generic Block IO controller infrastructure. 8776bf024e6SJohannes Weiner One needs to also enable actual IO controlling logic/policy. For 8786bf024e6SJohannes Weiner enabling proportional weight division of disk bandwidth in CFQ, set 8797baf2199SKrzysztof Kozlowski CONFIG_BFQ_GROUP_IOSCHED=y; for enabling throttling policy, set 8806bf024e6SJohannes Weiner CONFIG_BLK_DEV_THROTTLING=y. 8816bf024e6SJohannes Weiner 882da82c92fSMauro Carvalho Chehab See Documentation/admin-guide/cgroup-v1/blkio-controller.rst for more information. 8836bf024e6SJohannes Weiner 8846bf024e6SJohannes Weinerconfig CGROUP_WRITEBACK 8856bf024e6SJohannes Weiner bool 8866bf024e6SJohannes Weiner depends on MEMCG && BLK_CGROUP 8876bf024e6SJohannes Weiner default y 888e5d1367fSStephane Eranian 8897c941438SDhaval Gianimenuconfig CGROUP_SCHED 890a0166ec4SJohannes Weiner bool "CPU controller" 8917c941438SDhaval Giani default n 8927c941438SDhaval Giani help 8937c941438SDhaval Giani This feature lets CPU scheduler recognize task groups and control CPU 8947c941438SDhaval Giani bandwidth allocation to such task groups. It uses cgroups to group 8957c941438SDhaval Giani tasks. 8967c941438SDhaval Giani 8977c941438SDhaval Gianiif CGROUP_SCHED 8987c941438SDhaval Gianiconfig FAIR_GROUP_SCHED 8997c941438SDhaval Giani bool "Group scheduling for SCHED_OTHER" 9007c941438SDhaval Giani depends on CGROUP_SCHED 9017c941438SDhaval Giani default CGROUP_SCHED 9027c941438SDhaval Giani 903ab84d31eSPaul Turnerconfig CFS_BANDWIDTH 904ab84d31eSPaul Turner bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED" 905ab84d31eSPaul Turner depends on FAIR_GROUP_SCHED 906ab84d31eSPaul Turner default n 907ab84d31eSPaul Turner help 908ab84d31eSPaul Turner This option allows users to define CPU bandwidth rates (limits) for 909ab84d31eSPaul Turner tasks running within the fair group scheduler. Groups with no limit 910ab84d31eSPaul Turner set are considered to be unconstrained and will run with no 911ab84d31eSPaul Turner restriction. 912d6a3b247SMauro Carvalho Chehab See Documentation/scheduler/sched-bwc.rst for more information. 913ab84d31eSPaul Turner 9147c941438SDhaval Gianiconfig RT_GROUP_SCHED 9157c941438SDhaval Giani bool "Group scheduling for SCHED_RR/FIFO" 9167c941438SDhaval Giani depends on CGROUP_SCHED 9177c941438SDhaval Giani default n 9187c941438SDhaval Giani help 9197c941438SDhaval Giani This feature lets you explicitly allocate real CPU bandwidth 92032bd7eb5SLi Zefan to task groups. If enabled, it will also make it impossible to 9217c941438SDhaval Giani schedule realtime tasks for non-root users until you allocate 9227c941438SDhaval Giani realtime bandwidth for them. 923d6a3b247SMauro Carvalho Chehab See Documentation/scheduler/sched-rt-group.rst for more information. 9247c941438SDhaval Giani 9257c941438SDhaval Gianiendif #CGROUP_SCHED 9267c941438SDhaval Giani 9272480c093SPatrick Bellasiconfig UCLAMP_TASK_GROUP 9282480c093SPatrick Bellasi bool "Utilization clamping per group of tasks" 9292480c093SPatrick Bellasi depends on CGROUP_SCHED 9302480c093SPatrick Bellasi depends on UCLAMP_TASK 9312480c093SPatrick Bellasi default n 9322480c093SPatrick Bellasi help 9332480c093SPatrick Bellasi This feature enables the scheduler to track the clamped utilization 9342480c093SPatrick Bellasi of each CPU based on RUNNABLE tasks currently scheduled on that CPU. 9352480c093SPatrick Bellasi 9362480c093SPatrick Bellasi When this option is enabled, the user can specify a min and max 9372480c093SPatrick Bellasi CPU bandwidth which is allowed for each single task in a group. 9382480c093SPatrick Bellasi The max bandwidth allows to clamp the maximum frequency a task 9392480c093SPatrick Bellasi can use, while the min bandwidth allows to define a minimum 9402480c093SPatrick Bellasi frequency a task will always use. 9412480c093SPatrick Bellasi 9422480c093SPatrick Bellasi When task group based utilization clamping is enabled, an eventually 9432480c093SPatrick Bellasi specified task-specific clamp value is constrained by the cgroup 9442480c093SPatrick Bellasi specified clamp value. Both minimum and maximum task clamping cannot 9452480c093SPatrick Bellasi be bigger than the corresponding clamping defined at task group level. 9462480c093SPatrick Bellasi 9472480c093SPatrick Bellasi If in doubt, say N. 9482480c093SPatrick Bellasi 9496bf024e6SJohannes Weinerconfig CGROUP_PIDS 9506bf024e6SJohannes Weiner bool "PIDs controller" 9516bf024e6SJohannes Weiner help 9526bf024e6SJohannes Weiner Provides enforcement of process number limits in the scope of a 9536bf024e6SJohannes Weiner cgroup. Any attempt to fork more processes than is allowed in the 9546bf024e6SJohannes Weiner cgroup will fail. PIDs are fundamentally a global resource because it 9556bf024e6SJohannes Weiner is fairly trivial to reach PID exhaustion before you reach even a 9566bf024e6SJohannes Weiner conservative kmemcg limit. As a result, it is possible to grind a 9576bf024e6SJohannes Weiner system to halt without being limited by other cgroup policies. The 9586cc578dfSParav Pandit PIDs controller is designed to stop this from happening. 9596bf024e6SJohannes Weiner 9606bf024e6SJohannes Weiner It should be noted that organisational operations (such as attaching 96198076833SJonathan Neuschäfer to a cgroup hierarchy) will *not* be blocked by the PIDs controller, 9626bf024e6SJohannes Weiner since the PIDs limit only affects a process's ability to fork, not to 9636bf024e6SJohannes Weiner attach to a cgroup. 9646bf024e6SJohannes Weiner 96539d3e758SParav Panditconfig CGROUP_RDMA 96639d3e758SParav Pandit bool "RDMA controller" 96739d3e758SParav Pandit help 96839d3e758SParav Pandit Provides enforcement of RDMA resources defined by IB stack. 96939d3e758SParav Pandit It is fairly easy for consumers to exhaust RDMA resources, which 97039d3e758SParav Pandit can result into resource unavailability to other consumers. 97139d3e758SParav Pandit RDMA controller is designed to stop this from happening. 97239d3e758SParav Pandit Attaching processes with active RDMA resources to the cgroup 97339d3e758SParav Pandit hierarchy is allowed even if can cross the hierarchy's limit. 97439d3e758SParav Pandit 9756bf024e6SJohannes Weinerconfig CGROUP_FREEZER 9766bf024e6SJohannes Weiner bool "Freezer controller" 9776bf024e6SJohannes Weiner help 9786bf024e6SJohannes Weiner Provides a way to freeze and unfreeze all tasks in a 9796bf024e6SJohannes Weiner cgroup. 9806bf024e6SJohannes Weiner 981489c2a20SJohannes Weiner This option affects the ORIGINAL cgroup interface. The cgroup2 memory 982489c2a20SJohannes Weiner controller includes important in-kernel memory consumers per default. 983489c2a20SJohannes Weiner 984489c2a20SJohannes Weiner If you're using cgroup2, say N. 985489c2a20SJohannes Weiner 9866bf024e6SJohannes Weinerconfig CGROUP_HUGETLB 9876bf024e6SJohannes Weiner bool "HugeTLB controller" 9886bf024e6SJohannes Weiner depends on HUGETLB_PAGE 9896bf024e6SJohannes Weiner select PAGE_COUNTER 990afc24d49SVivek Goyal default n 9916bf024e6SJohannes Weiner help 9926bf024e6SJohannes Weiner Provides a cgroup controller for HugeTLB pages. 9936bf024e6SJohannes Weiner When you enable this, you can put a per cgroup limit on HugeTLB usage. 9946bf024e6SJohannes Weiner The limit is enforced during page fault. Since HugeTLB doesn't 9956bf024e6SJohannes Weiner support page reclaim, enforcing the limit at page fault time implies 9966bf024e6SJohannes Weiner that, the application will get SIGBUS signal if it tries to access 9976bf024e6SJohannes Weiner HugeTLB pages beyond its limit. This requires the application to know 9986bf024e6SJohannes Weiner beforehand how much HugeTLB pages it would require for its use. The 9996bf024e6SJohannes Weiner control group is tracked in the third page lru pointer. This means 10006bf024e6SJohannes Weiner that we cannot use the controller with huge page less than 3 pages. 1001afc24d49SVivek Goyal 10026bf024e6SJohannes Weinerconfig CPUSETS 10036bf024e6SJohannes Weiner bool "Cpuset controller" 1004e1d4eeecSNicolas Pitre depends on SMP 10056bf024e6SJohannes Weiner help 10066bf024e6SJohannes Weiner This option will let you create and manage CPUSETs which 10076bf024e6SJohannes Weiner allow dynamically partitioning a system into sets of CPUs and 10086bf024e6SJohannes Weiner Memory Nodes and assigning tasks to run only within those sets. 10096bf024e6SJohannes Weiner This is primarily useful on large SMP or NUMA systems. 1010afc24d49SVivek Goyal 10116bf024e6SJohannes Weiner Say N if unsure. 1012afc24d49SVivek Goyal 10136bf024e6SJohannes Weinerconfig PROC_PID_CPUSET 10146bf024e6SJohannes Weiner bool "Include legacy /proc/<pid>/cpuset file" 10156bf024e6SJohannes Weiner depends on CPUSETS 101689e9b9e0STejun Heo default y 101789e9b9e0STejun Heo 10186bf024e6SJohannes Weinerconfig CGROUP_DEVICE 10196bf024e6SJohannes Weiner bool "Device controller" 10206bf024e6SJohannes Weiner help 10216bf024e6SJohannes Weiner Provides a cgroup controller implementing whitelists for 10226bf024e6SJohannes Weiner devices which a process in the cgroup can mknod or open. 10236bf024e6SJohannes Weiner 10246bf024e6SJohannes Weinerconfig CGROUP_CPUACCT 10256bf024e6SJohannes Weiner bool "Simple CPU accounting controller" 10266bf024e6SJohannes Weiner help 10276bf024e6SJohannes Weiner Provides a simple controller for monitoring the 10286bf024e6SJohannes Weiner total CPU consumed by the tasks in a cgroup. 10296bf024e6SJohannes Weiner 10306bf024e6SJohannes Weinerconfig CGROUP_PERF 10316bf024e6SJohannes Weiner bool "Perf controller" 10326bf024e6SJohannes Weiner depends on PERF_EVENTS 10336bf024e6SJohannes Weiner help 10346bf024e6SJohannes Weiner This option extends the perf per-cpu mode to restrict monitoring 10356bf024e6SJohannes Weiner to threads which belong to the cgroup specified and run on the 10366546b19fSNamhyung Kim designated cpu. Or this can be used to have cgroup ID in samples 10376546b19fSNamhyung Kim so that it can monitor performance events among cgroups. 10386bf024e6SJohannes Weiner 10396bf024e6SJohannes Weiner Say N if unsure. 10406bf024e6SJohannes Weiner 104130070984SDaniel Mackconfig CGROUP_BPF 104230070984SDaniel Mack bool "Support for eBPF programs attached to cgroups" 1043483c4933SAndy Lutomirski depends on BPF_SYSCALL 1044483c4933SAndy Lutomirski select SOCK_CGROUP_DATA 104530070984SDaniel Mack help 104630070984SDaniel Mack Allow attaching eBPF programs to a cgroup using the bpf(2) 104730070984SDaniel Mack syscall command BPF_PROG_ATTACH. 104830070984SDaniel Mack 104930070984SDaniel Mack In which context these programs are accessed depends on the type 105030070984SDaniel Mack of attachment. For instance, programs that are attached using 105130070984SDaniel Mack BPF_CGROUP_INET_INGRESS will be executed on the ingress path of 105230070984SDaniel Mack inet sockets. 105330070984SDaniel Mack 10546bf024e6SJohannes Weinerconfig CGROUP_DEBUG 105523b0be48SWaiman Long bool "Debug controller" 10566bf024e6SJohannes Weiner default n 105723b0be48SWaiman Long depends on DEBUG_KERNEL 10586bf024e6SJohannes Weiner help 10596bf024e6SJohannes Weiner This option enables a simple controller that exports 106023b0be48SWaiman Long debugging information about the cgroups framework. This 106123b0be48SWaiman Long controller is for control cgroup debugging only. Its 106223b0be48SWaiman Long interfaces are not stable. 10636bf024e6SJohannes Weiner 10646bf024e6SJohannes Weiner Say N. 10656bf024e6SJohannes Weiner 106673b35147SArnd Bergmannconfig SOCK_CGROUP_DATA 106773b35147SArnd Bergmann bool 106873b35147SArnd Bergmann default n 106973b35147SArnd Bergmann 107023964d2dSLi Zefanendif # CGROUPS 1071c077719bSKAMEZAWA Hiroyuki 10728dd2a82cSDaniel Lezcanomenuconfig NAMESPACES 10736a108a14SDavid Rientjes bool "Namespaces support" if EXPERT 10742813893fSIulia Manda depends on MULTIUSER 10756a108a14SDavid Rientjes default !EXPERT 1076c5289a69SPavel Emelyanov help 1077c5289a69SPavel Emelyanov Provides the way to make tasks work with different objects using 1078c5289a69SPavel Emelyanov the same id. For example same IPC id may refer to different objects 1079c5289a69SPavel Emelyanov or same user id or pid may refer to different tasks when used in 1080c5289a69SPavel Emelyanov different namespaces. 1081c5289a69SPavel Emelyanov 10828dd2a82cSDaniel Lezcanoif NAMESPACES 10838dd2a82cSDaniel Lezcano 108458bfdd6dSPavel Emelyanovconfig UTS_NS 108558bfdd6dSPavel Emelyanov bool "UTS namespace" 108617a6d441SDaniel Lezcano default y 108758bfdd6dSPavel Emelyanov help 108858bfdd6dSPavel Emelyanov In this namespace tasks see different info provided with the 108958bfdd6dSPavel Emelyanov uname() system call 109058bfdd6dSPavel Emelyanov 1091769071acSAndrei Vaginconfig TIME_NS 1092769071acSAndrei Vagin bool "TIME namespace" 1093660fd04fSThomas Gleixner depends on GENERIC_VDSO_TIME_NS 1094769071acSAndrei Vagin default y 1095769071acSAndrei Vagin help 1096769071acSAndrei Vagin In this namespace boottime and monotonic clocks can be set. 1097769071acSAndrei Vagin The time will keep going with the same pace. 1098769071acSAndrei Vagin 1099ae5e1b22SPavel Emelyanovconfig IPC_NS 1100ae5e1b22SPavel Emelyanov bool "IPC namespace" 11018dd2a82cSDaniel Lezcano depends on (SYSVIPC || POSIX_MQUEUE) 110217a6d441SDaniel Lezcano default y 1103ae5e1b22SPavel Emelyanov help 1104ae5e1b22SPavel Emelyanov In this namespace tasks work with IPC ids which correspond to 1105614b84cfSSerge E. Hallyn different IPC objects in different namespaces. 1106ae5e1b22SPavel Emelyanov 1107aee16ce7SPavel Emelyanovconfig USER_NS 110819c92399SKees Cook bool "User namespace" 11095673a94cSEric W. Biederman default n 1110aee16ce7SPavel Emelyanov help 1111aee16ce7SPavel Emelyanov This allows containers, i.e. vservers, to use user namespaces 1112aee16ce7SPavel Emelyanov to provide different user info for different servers. 1113e11f0ae3SEric W. Biederman 1114e11f0ae3SEric W. Biederman When user namespaces are enabled in the kernel it is 1115d886f4e4SJohannes Weiner recommended that the MEMCG option also be enabled and that 1116d886f4e4SJohannes Weiner user-space use the memory control groups to limit the amount 1117d886f4e4SJohannes Weiner of memory a memory unprivileged users can use. 1118e11f0ae3SEric W. Biederman 1119aee16ce7SPavel Emelyanov If unsure, say N. 1120aee16ce7SPavel Emelyanov 112174bd59bbSPavel Emelyanovconfig PID_NS 11229bd38c2cSDaniel Lezcano bool "PID Namespaces" 112317a6d441SDaniel Lezcano default y 112474bd59bbSPavel Emelyanov help 112512d2b8f9SHeikki Orsila Support process id namespaces. This allows having multiple 1126692105b8SMatt LaPlante processes with the same pid as long as they are in different 112774bd59bbSPavel Emelyanov pid namespaces. This is a building block of containers. 112874bd59bbSPavel Emelyanov 1129d6eb633fSMatt Helsleyconfig NET_NS 1130d6eb633fSMatt Helsley bool "Network namespace" 11318dd2a82cSDaniel Lezcano depends on NET 113217a6d441SDaniel Lezcano default y 1133d6eb633fSMatt Helsley help 1134d6eb633fSMatt Helsley Allow user space to create what appear to be multiple instances 1135d6eb633fSMatt Helsley of the network stack. 1136d6eb633fSMatt Helsley 11378dd2a82cSDaniel Lezcanoendif # NAMESPACES 11388dd2a82cSDaniel Lezcano 11395cb366bbSAdrian Reberconfig CHECKPOINT_RESTORE 11405cb366bbSAdrian Reber bool "Checkpoint/restore support" 11415cb366bbSAdrian Reber select PROC_CHILDREN 11425cb366bbSAdrian Reber default n 11435cb366bbSAdrian Reber help 11445cb366bbSAdrian Reber Enables additional kernel features in a sake of checkpoint/restore. 11455cb366bbSAdrian Reber In particular it adds auxiliary prctl codes to setup process text, 11465cb366bbSAdrian Reber data and heap segment sizes, and a few additional /proc filesystem 11475cb366bbSAdrian Reber entries. 11485cb366bbSAdrian Reber 11495cb366bbSAdrian Reber If unsure, say N here. 11505cb366bbSAdrian Reber 11515091faa4SMike Galbraithconfig SCHED_AUTOGROUP 11525091faa4SMike Galbraith bool "Automatic process group scheduling" 11535091faa4SMike Galbraith select CGROUPS 11545091faa4SMike Galbraith select CGROUP_SCHED 11555091faa4SMike Galbraith select FAIR_GROUP_SCHED 11565091faa4SMike Galbraith help 11575091faa4SMike Galbraith This option optimizes the scheduler for common desktop workloads by 11585091faa4SMike Galbraith automatically creating and populating task groups. This separation 11595091faa4SMike Galbraith of workloads isolates aggressive CPU burners (like build jobs) from 11605091faa4SMike Galbraith desktop applications. Task group autogeneration is currently based 11615091faa4SMike Galbraith upon task session. 11625091faa4SMike Galbraith 11637af37becSDaniel Lezcanoconfig SYSFS_DEPRECATED 11645d6a4ea5SFerenc Wagner bool "Enable deprecated sysfs features to support old userspace tools" 11657af37becSDaniel Lezcano depends on SYSFS 11667af37becSDaniel Lezcano default n 11677af37becSDaniel Lezcano help 11687af37becSDaniel Lezcano This option adds code that switches the layout of the "block" class 11697af37becSDaniel Lezcano devices, to not show up in /sys/class/block/, but only in 11707af37becSDaniel Lezcano /sys/block/. 11717af37becSDaniel Lezcano 11727af37becSDaniel Lezcano This switch is only active when the sysfs.deprecated=1 boot option is 11737af37becSDaniel Lezcano passed or the SYSFS_DEPRECATED_V2 option is set. 11747af37becSDaniel Lezcano 11757af37becSDaniel Lezcano This option allows new kernels to run on old distributions and tools, 11767af37becSDaniel Lezcano which might get confused by /sys/class/block/. Since 2007/2008 all 11777af37becSDaniel Lezcano major distributions and tools handle this just fine. 11787af37becSDaniel Lezcano 11797af37becSDaniel Lezcano Recent distributions and userspace tools after 2009/2010 depend on 11807af37becSDaniel Lezcano the existence of /sys/class/block/, and will not work with this 11817af37becSDaniel Lezcano option enabled. 11827af37becSDaniel Lezcano 11837af37becSDaniel Lezcano Only if you are using a new kernel on an old distribution, you might 11847af37becSDaniel Lezcano need to say Y here. 11857af37becSDaniel Lezcano 11867af37becSDaniel Lezcanoconfig SYSFS_DEPRECATED_V2 11875d6a4ea5SFerenc Wagner bool "Enable deprecated sysfs features by default" 11887af37becSDaniel Lezcano default n 11897af37becSDaniel Lezcano depends on SYSFS 11907af37becSDaniel Lezcano depends on SYSFS_DEPRECATED 11917af37becSDaniel Lezcano help 11927af37becSDaniel Lezcano Enable deprecated sysfs by default. 11937af37becSDaniel Lezcano 11947af37becSDaniel Lezcano See the CONFIG_SYSFS_DEPRECATED option for more details about this 11957af37becSDaniel Lezcano option. 11967af37becSDaniel Lezcano 11977af37becSDaniel Lezcano Only if you are using a new kernel on an old distribution, you might 11987af37becSDaniel Lezcano need to say Y here. Even then, odds are you would not need it 11997af37becSDaniel Lezcano enabled, you can always pass the boot option if absolutely necessary. 12007af37becSDaniel Lezcano 12017af37becSDaniel Lezcanoconfig RELAY 12027af37becSDaniel Lezcano bool "Kernel->user space relay support (formerly relayfs)" 120326b5679eSPeter Zijlstra select IRQ_WORK 12047af37becSDaniel Lezcano help 12057af37becSDaniel Lezcano This option enables support for relay interface support in 12067af37becSDaniel Lezcano certain file systems (such as debugfs). 12077af37becSDaniel Lezcano It is designed to provide an efficient mechanism for tools and 12087af37becSDaniel Lezcano facilities to relay large amounts of data from kernel space to 12097af37becSDaniel Lezcano user space. 12107af37becSDaniel Lezcano 12117af37becSDaniel Lezcano If unsure, say N. 12127af37becSDaniel Lezcano 1213f991633dSDimitri Gorokhovikconfig BLK_DEV_INITRD 1214f991633dSDimitri Gorokhovik bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support" 1215f991633dSDimitri Gorokhovik help 1216f991633dSDimitri Gorokhovik The initial RAM filesystem is a ramfs which is loaded by the 1217f991633dSDimitri Gorokhovik boot loader (loadlin or lilo) and that is mounted as root 1218f991633dSDimitri Gorokhovik before the normal boot procedure. It is typically used to 1219f991633dSDimitri Gorokhovik load modules needed to mount the "real" root file system, 12208c27ceffSMauro Carvalho Chehab etc. See <file:Documentation/admin-guide/initrd.rst> for details. 1221f991633dSDimitri Gorokhovik 1222f991633dSDimitri Gorokhovik If RAM disk support (BLK_DEV_RAM) is also included, this 1223f991633dSDimitri Gorokhovik also enables initial RAM disk (initrd) support and adds 1224f991633dSDimitri Gorokhovik 15 Kbytes (more on some other architectures) to the kernel size. 1225f991633dSDimitri Gorokhovik 1226f991633dSDimitri Gorokhovik If unsure say Y. 1227f991633dSDimitri Gorokhovik 1228c33df4eaSJean-Paul Samanif BLK_DEV_INITRD 1229c33df4eaSJean-Paul Saman 1230dbec4866SSam Ravnborgsource "usr/Kconfig" 1231dbec4866SSam Ravnborg 1232c33df4eaSJean-Paul Samanendif 1233c33df4eaSJean-Paul Saman 123476db5a27SMasami Hiramatsuconfig BOOT_CONFIG 123576db5a27SMasami Hiramatsu bool "Boot config support" 12362910b5aaSMasami Hiramatsu select BLK_DEV_INITRD 123776db5a27SMasami Hiramatsu help 123876db5a27SMasami Hiramatsu Extra boot config allows system admin to pass a config file as 123976db5a27SMasami Hiramatsu complemental extension of kernel cmdline when booting. 12400947db01SMasami Hiramatsu The boot config file must be attached at the end of initramfs 124185c46b78SMasami Hiramatsu with checksum, size and magic word. 12420947db01SMasami Hiramatsu See <file:Documentation/admin-guide/bootconfig.rst> for details. 124376db5a27SMasami Hiramatsu 124476db5a27SMasami Hiramatsu If unsure, say Y. 124576db5a27SMasami Hiramatsu 1246877417e6SArnd Bergmannchoice 1247877417e6SArnd Bergmann prompt "Compiler optimization level" 12482cc3ce24SUlf Magnusson default CC_OPTIMIZE_FOR_PERFORMANCE 1249877417e6SArnd Bergmann 1250877417e6SArnd Bergmannconfig CC_OPTIMIZE_FOR_PERFORMANCE 125115f5db60SMasahiro Yamada bool "Optimize for performance (-O2)" 1252877417e6SArnd Bergmann help 1253877417e6SArnd Bergmann This is the default optimization level for the kernel, building 1254877417e6SArnd Bergmann with the "-O2" compiler flag for best performance and most 1255877417e6SArnd Bergmann helpful compile-time warnings. 1256877417e6SArnd Bergmann 125715f5db60SMasahiro Yamadaconfig CC_OPTIMIZE_FOR_PERFORMANCE_O3 125815f5db60SMasahiro Yamada bool "Optimize more for performance (-O3)" 125915f5db60SMasahiro Yamada depends on ARC 1260b303c6dfSMasahiro Yamada imply CC_DISABLE_WARN_MAYBE_UNINITIALIZED # avoid false positives 1261c45b4f1fSLinus Torvalds help 126215f5db60SMasahiro Yamada Choosing this option will pass "-O3" to your compiler to optimize 126315f5db60SMasahiro Yamada the kernel yet more for performance. 1264c45b4f1fSLinus Torvalds 12655d20ee31SNicholas Pigginconfig CC_OPTIMIZE_FOR_SIZE 126615f5db60SMasahiro Yamada bool "Optimize for size (-Os)" 12675d20ee31SNicholas Piggin imply CC_DISABLE_WARN_MAYBE_UNINITIALIZED # avoid false positives 12685d20ee31SNicholas Piggin help 1269ce3b487fSMasahiro Yamada Choosing this option will pass "-Os" to your compiler resulting 1270ce3b487fSMasahiro Yamada in a smaller kernel. 1271c45b4f1fSLinus Torvalds 1272877417e6SArnd Bergmannendchoice 1273877417e6SArnd Bergmann 12745d20ee31SNicholas Pigginconfig HAVE_LD_DEAD_CODE_DATA_ELIMINATION 12755d20ee31SNicholas Piggin bool 12765d20ee31SNicholas Piggin help 12775d20ee31SNicholas Piggin This requires that the arch annotates or otherwise protects 12785d20ee31SNicholas Piggin its external entry points from being discarded. Linker scripts 12795d20ee31SNicholas Piggin must also merge .text.*, .data.*, and .bss.* correctly into 12805d20ee31SNicholas Piggin output sections. Care must be taken not to pull in unrelated 12815d20ee31SNicholas Piggin sections (e.g., '.text.init'). Typically '.' in section names 12825d20ee31SNicholas Piggin is used to distinguish them from label names / C identifiers. 12835d20ee31SNicholas Piggin 12845d20ee31SNicholas Pigginconfig LD_DEAD_CODE_DATA_ELIMINATION 12855d20ee31SNicholas Piggin bool "Dead code and data elimination (EXPERIMENTAL)" 12865d20ee31SNicholas Piggin depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION 12875d20ee31SNicholas Piggin depends on EXPERT 128816fd20aaSPaul Burton depends on !(FUNCTION_TRACER && CC_IS_GCC && GCC_VERSION < 40800) 1289e85d1d65SMasahiro Yamada depends on $(cc-option,-ffunction-sections -fdata-sections) 1290e85d1d65SMasahiro Yamada depends on $(ld-option,--gc-sections) 12915d20ee31SNicholas Piggin help 12928b9d2712SMasahiro Yamada Enable this if you want to do dead code and data elimination with 12938b9d2712SMasahiro Yamada the linker by compiling with -ffunction-sections -fdata-sections, 12948b9d2712SMasahiro Yamada and linking with --gc-sections. 12955d20ee31SNicholas Piggin 12965d20ee31SNicholas Piggin This can reduce on disk and in-memory size of the kernel 12975d20ee31SNicholas Piggin code and static data, particularly for small configs and 12985d20ee31SNicholas Piggin on small systems. This has the possibility of introducing 12995d20ee31SNicholas Piggin silently broken kernel if the required annotations are not 13005d20ee31SNicholas Piggin present. This option is not well tested yet, so use at your 13015d20ee31SNicholas Piggin own risk. 13025d20ee31SNicholas Piggin 13030847062aSRandy Dunlapconfig SYSCTL 13040847062aSRandy Dunlap bool 13050847062aSRandy Dunlap 1306657a5209SMike Frysingerconfig HAVE_UID16 1307657a5209SMike Frysinger bool 1308657a5209SMike Frysinger 1309657a5209SMike Frysingerconfig SYSCTL_EXCEPTION_TRACE 1310657a5209SMike Frysinger bool 1311657a5209SMike Frysinger help 1312657a5209SMike Frysinger Enable support for /proc/sys/debug/exception-trace. 1313657a5209SMike Frysinger 1314657a5209SMike Frysingerconfig SYSCTL_ARCH_UNALIGN_NO_WARN 1315657a5209SMike Frysinger bool 1316657a5209SMike Frysinger help 1317657a5209SMike Frysinger Enable support for /proc/sys/kernel/ignore-unaligned-usertrap 1318657a5209SMike Frysinger Allows arch to define/use @no_unaligned_warning to possibly warn 1319657a5209SMike Frysinger about unaligned access emulation going on under the hood. 1320657a5209SMike Frysinger 1321657a5209SMike Frysingerconfig SYSCTL_ARCH_UNALIGN_ALLOW 1322657a5209SMike Frysinger bool 1323657a5209SMike Frysinger help 1324657a5209SMike Frysinger Enable support for /proc/sys/kernel/unaligned-trap 1325657a5209SMike Frysinger Allows arches to define/use @unaligned_enabled to runtime toggle 1326657a5209SMike Frysinger the unaligned access emulation. 1327657a5209SMike Frysinger see arch/parisc/kernel/unaligned.c for reference 1328657a5209SMike Frysinger 1329657a5209SMike Frysingerconfig HAVE_PCSPKR_PLATFORM 1330657a5209SMike Frysinger bool 1331657a5209SMike Frysinger 1332f89b7755SAlexei Starovoitov# interpreter that classic socket filters depend on 1333f89b7755SAlexei Starovoitovconfig BPF 1334f89b7755SAlexei Starovoitov bool 1335f89b7755SAlexei Starovoitov 13366a108a14SDavid Rientjesmenuconfig EXPERT 13376a108a14SDavid Rientjes bool "Configure standard kernel features (expert users)" 1338f505c553SJosh Triplett # Unhide debug options, to make the on-by-default options visible 1339f505c553SJosh Triplett select DEBUG_KERNEL 13401da177e4SLinus Torvalds help 13411da177e4SLinus Torvalds This option allows certain base kernel options and settings 13421da177e4SLinus Torvalds to be disabled or tweaked. This is for specialized 13431da177e4SLinus Torvalds environments which can tolerate a "non-standard" kernel. 13441da177e4SLinus Torvalds Only use this if you really know what you are doing. 13451da177e4SLinus Torvalds 1346ae81f9e3SChuck Ebbertconfig UID16 13476a108a14SDavid Rientjes bool "Enable 16-bit UID system calls" if EXPERT 13482813893fSIulia Manda depends on HAVE_UID16 && MULTIUSER 1349ae81f9e3SChuck Ebbert default y 1350ae81f9e3SChuck Ebbert help 1351ae81f9e3SChuck Ebbert This enables the legacy 16-bit UID syscall wrappers. 1352ae81f9e3SChuck Ebbert 13532813893fSIulia Mandaconfig MULTIUSER 13542813893fSIulia Manda bool "Multiple users, groups and capabilities support" if EXPERT 13552813893fSIulia Manda default y 13562813893fSIulia Manda help 13572813893fSIulia Manda This option enables support for non-root users, groups and 13582813893fSIulia Manda capabilities. 13592813893fSIulia Manda 13602813893fSIulia Manda If you say N here, all processes will run with UID 0, GID 0, and all 13612813893fSIulia Manda possible capabilities. Saying N here also compiles out support for 13622813893fSIulia Manda system calls related to UIDs, GIDs, and capabilities, such as setuid, 13632813893fSIulia Manda setgid, and capset. 13642813893fSIulia Manda 13652813893fSIulia Manda If unsure, say Y here. 13662813893fSIulia Manda 1367f6187769SFabian Frederickconfig SGETMASK_SYSCALL 1368f6187769SFabian Frederick bool "sgetmask/ssetmask syscalls support" if EXPERT 1369a687a533SArnd Bergmann def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH 1370f6187769SFabian Frederick ---help--- 1371f6187769SFabian Frederick sys_sgetmask and sys_ssetmask are obsolete system calls 1372f6187769SFabian Frederick no longer supported in libc but still enabled by default in some 1373f6187769SFabian Frederick architectures. 1374f6187769SFabian Frederick 1375f6187769SFabian Frederick If unsure, leave the default option here. 1376f6187769SFabian Frederick 13776af9f7bfSFabian Frederickconfig SYSFS_SYSCALL 13786af9f7bfSFabian Frederick bool "Sysfs syscall support" if EXPERT 13796af9f7bfSFabian Frederick default y 13806af9f7bfSFabian Frederick ---help--- 13816af9f7bfSFabian Frederick sys_sysfs is an obsolete system call no longer supported in libc. 13826af9f7bfSFabian Frederick Note that disabling this option is more secure but might break 13836af9f7bfSFabian Frederick compatibility with some systems. 13846af9f7bfSFabian Frederick 13856af9f7bfSFabian Frederick If unsure say Y here. 13866af9f7bfSFabian Frederick 1387d1b069f5SRandy Dunlapconfig FHANDLE 1388d1b069f5SRandy Dunlap bool "open by fhandle syscalls" if EXPERT 1389d1b069f5SRandy Dunlap select EXPORTFS 1390d1b069f5SRandy Dunlap default y 1391d1b069f5SRandy Dunlap help 1392d1b069f5SRandy Dunlap If you say Y here, a user level program will be able to map 1393d1b069f5SRandy Dunlap file names to handle and then later use the handle for 1394d1b069f5SRandy Dunlap different file system operations. This is useful in implementing 1395d1b069f5SRandy Dunlap userspace file servers, which now track files using handles instead 1396d1b069f5SRandy Dunlap of names. The handle would remain the same even if file names 1397d1b069f5SRandy Dunlap get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2) 1398d1b069f5SRandy Dunlap syscalls. 1399d1b069f5SRandy Dunlap 1400baa73d9eSNicolas Pitreconfig POSIX_TIMERS 1401baa73d9eSNicolas Pitre bool "Posix Clocks & timers" if EXPERT 1402baa73d9eSNicolas Pitre default y 1403baa73d9eSNicolas Pitre help 1404baa73d9eSNicolas Pitre This includes native support for POSIX timers to the kernel. 1405baa73d9eSNicolas Pitre Some embedded systems have no use for them and therefore they 1406baa73d9eSNicolas Pitre can be configured out to reduce the size of the kernel image. 1407baa73d9eSNicolas Pitre 1408baa73d9eSNicolas Pitre When this option is disabled, the following syscalls won't be 1409baa73d9eSNicolas Pitre available: timer_create, timer_gettime: timer_getoverrun, 1410baa73d9eSNicolas Pitre timer_settime, timer_delete, clock_adjtime, getitimer, 1411baa73d9eSNicolas Pitre setitimer, alarm. Furthermore, the clock_settime, clock_gettime, 1412baa73d9eSNicolas Pitre clock_getres and clock_nanosleep syscalls will be limited to 1413baa73d9eSNicolas Pitre CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only. 1414baa73d9eSNicolas Pitre 1415baa73d9eSNicolas Pitre If unsure say y. 1416baa73d9eSNicolas Pitre 1417d59745ceSMatt Mackallconfig PRINTK 1418d59745ceSMatt Mackall default y 14196a108a14SDavid Rientjes bool "Enable support for printk" if EXPERT 142074876a98SFrederic Weisbecker select IRQ_WORK 1421d59745ceSMatt Mackall help 1422d59745ceSMatt Mackall This option enables normal printk support. Removing it 1423d59745ceSMatt Mackall eliminates most of the message strings from the kernel image 1424d59745ceSMatt Mackall and makes the kernel more or less silent. As this makes it 1425d59745ceSMatt Mackall very difficult to diagnose system problems, saying N here is 1426d59745ceSMatt Mackall strongly discouraged. 1427d59745ceSMatt Mackall 142842a0bb3fSPetr Mladekconfig PRINTK_NMI 142942a0bb3fSPetr Mladek def_bool y 143042a0bb3fSPetr Mladek depends on PRINTK 143142a0bb3fSPetr Mladek depends on HAVE_NMI 143242a0bb3fSPetr Mladek 1433c8538a7aSMatt Mackallconfig BUG 14346a108a14SDavid Rientjes bool "BUG() support" if EXPERT 1435c8538a7aSMatt Mackall default y 1436c8538a7aSMatt Mackall help 1437c8538a7aSMatt Mackall Disabling this option eliminates support for BUG and WARN, reducing 1438c8538a7aSMatt Mackall the size of your kernel image and potentially quietly ignoring 1439c8538a7aSMatt Mackall numerous fatal conditions. You should only consider disabling this 1440c8538a7aSMatt Mackall option for embedded systems with no facilities for reporting errors. 1441c8538a7aSMatt Mackall Just say Y. 1442c8538a7aSMatt Mackall 1443708e9a79SMatt Mackallconfig ELF_CORE 1444046d662fSAlex Kelly depends on COREDUMP 1445708e9a79SMatt Mackall default y 14466a108a14SDavid Rientjes bool "Enable ELF core dumps" if EXPERT 1447708e9a79SMatt Mackall help 1448708e9a79SMatt Mackall Enable support for generating core dumps. Disabling saves about 4k. 1449708e9a79SMatt Mackall 14508761f1abSRalf Baechle 1451e5e1d3cbSStas Sergeevconfig PCSPKR_PLATFORM 14526a108a14SDavid Rientjes bool "Enable PC-Speaker support" if EXPERT 14538761f1abSRalf Baechle depends on HAVE_PCSPKR_PLATFORM 145415f304b6SRalf Baechle select I8253_LOCK 1455e5e1d3cbSStas Sergeev default y 1456e5e1d3cbSStas Sergeev help 1457e5e1d3cbSStas Sergeev This option allows to disable the internal PC-Speaker 1458e5e1d3cbSStas Sergeev support, saving some memory. 1459e5e1d3cbSStas Sergeev 14601da177e4SLinus Torvaldsconfig BASE_FULL 14611da177e4SLinus Torvalds default y 14626a108a14SDavid Rientjes bool "Enable full-sized data structures for core" if EXPERT 14631da177e4SLinus Torvalds help 14641da177e4SLinus Torvalds Disabling this option reduces the size of miscellaneous core 14651da177e4SLinus Torvalds kernel data structures. This saves memory on small machines, 14661da177e4SLinus Torvalds but may reduce performance. 14671da177e4SLinus Torvalds 14681da177e4SLinus Torvaldsconfig FUTEX 14696a108a14SDavid Rientjes bool "Enable futex support" if EXPERT 14701da177e4SLinus Torvalds default y 1471bc2eecd7SNicolas Pitre imply RT_MUTEXES 14721da177e4SLinus Torvalds help 14731da177e4SLinus Torvalds Disabling this option will cause the kernel to be built without 14741da177e4SLinus Torvalds support for "fast userspace mutexes". The resulting kernel may not 14751da177e4SLinus Torvalds run glibc-based applications correctly. 14761da177e4SLinus Torvalds 1477bc2eecd7SNicolas Pitreconfig FUTEX_PI 1478bc2eecd7SNicolas Pitre bool 1479bc2eecd7SNicolas Pitre depends on FUTEX && RT_MUTEXES 1480bc2eecd7SNicolas Pitre default y 1481bc2eecd7SNicolas Pitre 148203b8c7b6SHeiko Carstensconfig HAVE_FUTEX_CMPXCHG 148303b8c7b6SHeiko Carstens bool 148462b4d204SJosh Triplett depends on FUTEX 148503b8c7b6SHeiko Carstens help 148603b8c7b6SHeiko Carstens Architectures should select this if futex_atomic_cmpxchg_inatomic() 148703b8c7b6SHeiko Carstens is implemented and always working. This removes a couple of runtime 148803b8c7b6SHeiko Carstens checks. 148903b8c7b6SHeiko Carstens 14901da177e4SLinus Torvaldsconfig EPOLL 14916a108a14SDavid Rientjes bool "Enable eventpoll support" if EXPERT 14921da177e4SLinus Torvalds default y 14931da177e4SLinus Torvalds help 14941da177e4SLinus Torvalds Disabling this option will cause the kernel to be built without 14951da177e4SLinus Torvalds support for epoll family of system calls. 14961da177e4SLinus Torvalds 1497fba2afaaSDavide Libenziconfig SIGNALFD 14986a108a14SDavid Rientjes bool "Enable signalfd() system call" if EXPERT 1499fba2afaaSDavide Libenzi default y 1500fba2afaaSDavide Libenzi help 1501fba2afaaSDavide Libenzi Enable the signalfd() system call that allows to receive signals 1502fba2afaaSDavide Libenzi on a file descriptor. 1503fba2afaaSDavide Libenzi 1504fba2afaaSDavide Libenzi If unsure, say Y. 1505fba2afaaSDavide Libenzi 1506b215e283SDavide Libenziconfig TIMERFD 15076a108a14SDavid Rientjes bool "Enable timerfd() system call" if EXPERT 1508b215e283SDavide Libenzi default y 1509b215e283SDavide Libenzi help 1510b215e283SDavide Libenzi Enable the timerfd() system call that allows to receive timer 1511b215e283SDavide Libenzi events on a file descriptor. 1512b215e283SDavide Libenzi 1513b215e283SDavide Libenzi If unsure, say Y. 1514b215e283SDavide Libenzi 1515e1ad7468SDavide Libenziconfig EVENTFD 15166a108a14SDavid Rientjes bool "Enable eventfd() system call" if EXPERT 1517e1ad7468SDavide Libenzi default y 1518e1ad7468SDavide Libenzi help 1519e1ad7468SDavide Libenzi Enable the eventfd() system call that allows to receive both 1520e1ad7468SDavide Libenzi kernel notification (ie. KAIO) or userspace notifications. 1521e1ad7468SDavide Libenzi 1522e1ad7468SDavide Libenzi If unsure, say Y. 1523e1ad7468SDavide Libenzi 15241da177e4SLinus Torvaldsconfig SHMEM 15256a108a14SDavid Rientjes bool "Use full shmem filesystem" if EXPERT 15261da177e4SLinus Torvalds default y 15271da177e4SLinus Torvalds depends on MMU 15281da177e4SLinus Torvalds help 15291da177e4SLinus Torvalds The shmem is an internal filesystem used to manage shared memory. 15301da177e4SLinus Torvalds It is backed by swap and manages resource limits. It is also exported 15311da177e4SLinus Torvalds to userspace as tmpfs if TMPFS is enabled. Disabling this 15321da177e4SLinus Torvalds option replaces shmem and tmpfs with the much simpler ramfs code, 15331da177e4SLinus Torvalds which may be appropriate on small systems without swap. 15341da177e4SLinus Torvalds 1535ebf3f09cSThomas Petazzoniconfig AIO 15366a108a14SDavid Rientjes bool "Enable AIO support" if EXPERT 1537ebf3f09cSThomas Petazzoni default y 1538ebf3f09cSThomas Petazzoni help 1539ebf3f09cSThomas Petazzoni This option enables POSIX asynchronous I/O which may by used 1540ebf3f09cSThomas Petazzoni by some high performance threaded applications. Disabling 1541ebf3f09cSThomas Petazzoni this option saves about 7k. 1542ebf3f09cSThomas Petazzoni 15432b188cc1SJens Axboeconfig IO_URING 15442b188cc1SJens Axboe bool "Enable IO uring support" if EXPERT 1545561fb04aSJens Axboe select IO_WQ 15462b188cc1SJens Axboe default y 15472b188cc1SJens Axboe help 15482b188cc1SJens Axboe This option enables support for the io_uring interface, enabling 15492b188cc1SJens Axboe applications to submit and complete IO through submission and 15502b188cc1SJens Axboe completion rings that are shared between the kernel and application. 15512b188cc1SJens Axboe 1552d3ac21caSJosh Triplettconfig ADVISE_SYSCALLS 1553d3ac21caSJosh Triplett bool "Enable madvise/fadvise syscalls" if EXPERT 1554d3ac21caSJosh Triplett default y 1555d3ac21caSJosh Triplett help 1556d3ac21caSJosh Triplett This option enables the madvise and fadvise syscalls, used by 1557d3ac21caSJosh Triplett applications to advise the kernel about their future memory or file 1558d3ac21caSJosh Triplett usage, improving performance. If building an embedded system where no 1559d3ac21caSJosh Triplett applications use these syscalls, you can disable this option to save 1560d3ac21caSJosh Triplett space. 1561d3ac21caSJosh Triplett 15625a281062SAndrea Arcangeliconfig HAVE_ARCH_USERFAULTFD_WP 15635a281062SAndrea Arcangeli bool 15645a281062SAndrea Arcangeli help 15655a281062SAndrea Arcangeli Arch has userfaultfd write protection support 15665a281062SAndrea Arcangeli 15675b25b13aSMathieu Desnoyersconfig MEMBARRIER 15685b25b13aSMathieu Desnoyers bool "Enable membarrier() system call" if EXPERT 15695b25b13aSMathieu Desnoyers default y 15705b25b13aSMathieu Desnoyers help 15715b25b13aSMathieu Desnoyers Enable the membarrier() system call that allows issuing memory 15725b25b13aSMathieu Desnoyers barriers across all running threads, which can be used to distribute 15735b25b13aSMathieu Desnoyers the cost of user-space memory barriers asymmetrically by transforming 15745b25b13aSMathieu Desnoyers pairs of memory barriers into pairs consisting of membarrier() and a 15755b25b13aSMathieu Desnoyers compiler barrier. 15765b25b13aSMathieu Desnoyers 15775b25b13aSMathieu Desnoyers If unsure, say Y. 15785b25b13aSMathieu Desnoyers 1579d1b069f5SRandy Dunlapconfig KALLSYMS 1580d1b069f5SRandy Dunlap bool "Load all symbols for debugging/ksymoops" if EXPERT 1581d1b069f5SRandy Dunlap default y 1582d1b069f5SRandy Dunlap help 1583d1b069f5SRandy Dunlap Say Y here to let the kernel print out symbolic crash information and 1584d1b069f5SRandy Dunlap symbolic stack backtraces. This increases the size of the kernel 1585d1b069f5SRandy Dunlap somewhat, as all symbols have to be loaded into the kernel image. 1586d1b069f5SRandy Dunlap 1587d1b069f5SRandy Dunlapconfig KALLSYMS_ALL 1588d1b069f5SRandy Dunlap bool "Include all symbols in kallsyms" 1589d1b069f5SRandy Dunlap depends on DEBUG_KERNEL && KALLSYMS 1590d1b069f5SRandy Dunlap help 1591d1b069f5SRandy Dunlap Normally kallsyms only contains the symbols of functions for nicer 1592d1b069f5SRandy Dunlap OOPS messages and backtraces (i.e., symbols from the text and inittext 1593d1b069f5SRandy Dunlap sections). This is sufficient for most cases. And only in very rare 1594d1b069f5SRandy Dunlap cases (e.g., when a debugger is used) all symbols are required (e.g., 1595d1b069f5SRandy Dunlap names of variables from the data sections, etc). 1596d1b069f5SRandy Dunlap 1597d1b069f5SRandy Dunlap This option makes sure that all symbols are loaded into the kernel 1598d1b069f5SRandy Dunlap image (i.e., symbols from all sections) in cost of increased kernel 1599d1b069f5SRandy Dunlap size (depending on the kernel configuration, it may be 300KiB or 1600d1b069f5SRandy Dunlap something like this). 1601d1b069f5SRandy Dunlap 1602d1b069f5SRandy Dunlap Say N unless you really need all symbols. 1603d1b069f5SRandy Dunlap 1604d1b069f5SRandy Dunlapconfig KALLSYMS_ABSOLUTE_PERCPU 1605d1b069f5SRandy Dunlap bool 1606d1b069f5SRandy Dunlap depends on KALLSYMS 1607d1b069f5SRandy Dunlap default X86_64 && SMP 1608d1b069f5SRandy Dunlap 1609d1b069f5SRandy Dunlapconfig KALLSYMS_BASE_RELATIVE 1610d1b069f5SRandy Dunlap bool 1611d1b069f5SRandy Dunlap depends on KALLSYMS 1612a687a533SArnd Bergmann default !IA64 1613d1b069f5SRandy Dunlap help 1614d1b069f5SRandy Dunlap Instead of emitting them as absolute values in the native word size, 1615d1b069f5SRandy Dunlap emit the symbol references in the kallsyms table as 32-bit entries, 1616d1b069f5SRandy Dunlap each containing a relative value in the range [base, base + U32_MAX] 1617d1b069f5SRandy Dunlap or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either 1618d1b069f5SRandy Dunlap an absolute value in the range [0, S32_MAX] or a relative value in the 1619d1b069f5SRandy Dunlap range [base, base + S32_MAX], where base is the lowest relative symbol 1620d1b069f5SRandy Dunlap address encountered in the image. 1621d1b069f5SRandy Dunlap 1622d1b069f5SRandy Dunlap On 64-bit builds, this reduces the size of the address table by 50%, 1623d1b069f5SRandy Dunlap but more importantly, it results in entries whose values are build 1624d1b069f5SRandy Dunlap time constants, and no relocation pass is required at runtime to fix 1625d1b069f5SRandy Dunlap up the entries based on the runtime load address of the kernel. 1626d1b069f5SRandy Dunlap 1627d1b069f5SRandy Dunlap# end of the "standard kernel features (expert users)" menu 1628d1b069f5SRandy Dunlap 1629d1b069f5SRandy Dunlap# syscall, maps, verifier 1630fc611f47SKP Singh 1631fc611f47SKP Singhconfig BPF_LSM 1632fc611f47SKP Singh bool "LSM Instrumentation with BPF" 16334edf16b7SKP Singh depends on BPF_EVENTS 1634fc611f47SKP Singh depends on BPF_SYSCALL 1635fc611f47SKP Singh depends on SECURITY 1636fc611f47SKP Singh depends on BPF_JIT 1637fc611f47SKP Singh help 1638fc611f47SKP Singh Enables instrumentation of the security hooks with eBPF programs for 1639fc611f47SKP Singh implementing dynamic MAC and Audit Policies. 1640fc611f47SKP Singh 1641fc611f47SKP Singh If you are unsure how to answer this question, answer N. 1642fc611f47SKP Singh 1643d1b069f5SRandy Dunlapconfig BPF_SYSCALL 1644d1b069f5SRandy Dunlap bool "Enable bpf() system call" 1645d1b069f5SRandy Dunlap select BPF 1646bae77c5eSSong Liu select IRQ_WORK 1647d1b069f5SRandy Dunlap default n 1648d1b069f5SRandy Dunlap help 1649d1b069f5SRandy Dunlap Enable the bpf() system call that allows to manipulate eBPF 1650d1b069f5SRandy Dunlap programs and maps via file descriptors. 1651d1b069f5SRandy Dunlap 165281c22041SDaniel Borkmannconfig ARCH_WANT_DEFAULT_BPF_JIT 165381c22041SDaniel Borkmann bool 165481c22041SDaniel Borkmann 1655290af866SAlexei Starovoitovconfig BPF_JIT_ALWAYS_ON 1656290af866SAlexei Starovoitov bool "Permanently enable BPF JIT and remove BPF interpreter" 1657290af866SAlexei Starovoitov depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT 1658290af866SAlexei Starovoitov help 1659290af866SAlexei Starovoitov Enables BPF JIT and removes BPF interpreter to avoid 1660290af866SAlexei Starovoitov speculative execution of BPF instructions by the interpreter 1661290af866SAlexei Starovoitov 166281c22041SDaniel Borkmannconfig BPF_JIT_DEFAULT_ON 166381c22041SDaniel Borkmann def_bool ARCH_WANT_DEFAULT_BPF_JIT || BPF_JIT_ALWAYS_ON 166481c22041SDaniel Borkmann depends on HAVE_EBPF_JIT && BPF_JIT 166581c22041SDaniel Borkmann 1666d1b069f5SRandy Dunlapconfig USERFAULTFD 1667d1b069f5SRandy Dunlap bool "Enable userfaultfd() system call" 1668d1b069f5SRandy Dunlap depends on MMU 1669d1b069f5SRandy Dunlap help 1670d1b069f5SRandy Dunlap Enable the userfaultfd() system call that allows to intercept and 1671d1b069f5SRandy Dunlap handle page faults in userland. 1672d1b069f5SRandy Dunlap 16733ccfebedSMathieu Desnoyersconfig ARCH_HAS_MEMBARRIER_CALLBACKS 16743ccfebedSMathieu Desnoyers bool 16753ccfebedSMathieu Desnoyers 167670216e18SMathieu Desnoyersconfig ARCH_HAS_MEMBARRIER_SYNC_CORE 167770216e18SMathieu Desnoyers bool 167870216e18SMathieu Desnoyers 1679d7822b1eSMathieu Desnoyersconfig RSEQ 1680d7822b1eSMathieu Desnoyers bool "Enable rseq() system call" if EXPERT 1681d7822b1eSMathieu Desnoyers default y 1682d7822b1eSMathieu Desnoyers depends on HAVE_RSEQ 1683d7822b1eSMathieu Desnoyers select MEMBARRIER 1684d7822b1eSMathieu Desnoyers help 1685d7822b1eSMathieu Desnoyers Enable the restartable sequences system call. It provides a 1686d7822b1eSMathieu Desnoyers user-space cache for the current CPU number value, which 1687d7822b1eSMathieu Desnoyers speeds up getting the current CPU number from user-space, 1688d7822b1eSMathieu Desnoyers as well as an ABI to speed up user-space operations on 1689d7822b1eSMathieu Desnoyers per-CPU data. 1690d7822b1eSMathieu Desnoyers 1691d7822b1eSMathieu Desnoyers If unsure, say Y. 1692d7822b1eSMathieu Desnoyers 1693d7822b1eSMathieu Desnoyersconfig DEBUG_RSEQ 1694d7822b1eSMathieu Desnoyers default n 1695d7822b1eSMathieu Desnoyers bool "Enabled debugging of rseq() system call" if EXPERT 1696d7822b1eSMathieu Desnoyers depends on RSEQ && DEBUG_KERNEL 1697d7822b1eSMathieu Desnoyers help 1698d7822b1eSMathieu Desnoyers Enable extra debugging checks for the rseq system call. 1699d7822b1eSMathieu Desnoyers 1700d7822b1eSMathieu Desnoyers If unsure, say N. 1701d7822b1eSMathieu Desnoyers 17026befe5f6SRandy Dunlapconfig EMBEDDED 17036befe5f6SRandy Dunlap bool "Embedded system" 17045d2acfc7SJosh Triplett option allnoconfig_y 17056befe5f6SRandy Dunlap select EXPERT 17066befe5f6SRandy Dunlap help 17076befe5f6SRandy Dunlap This option should be enabled if compiling the kernel for 17086befe5f6SRandy Dunlap an embedded system so certain expert options are available 17096befe5f6SRandy Dunlap for configuration. 17106befe5f6SRandy Dunlap 1711cdd6c482SIngo Molnarconfig HAVE_PERF_EVENTS 17120793a61dSThomas Gleixner bool 1713018df72dSMike Frysinger help 1714018df72dSMike Frysinger See tools/perf/design.txt for details. 17150793a61dSThomas Gleixner 1716906010b2SPeter Zijlstraconfig PERF_USE_VMALLOC 1717906010b2SPeter Zijlstra bool 1718906010b2SPeter Zijlstra help 1719906010b2SPeter Zijlstra See tools/perf/design.txt for details 1720906010b2SPeter Zijlstra 1721ad90a3deSWilliam Breathitt Grayconfig PC104 1722424529fbSWilliam Breathitt Gray bool "PC/104 support" if EXPERT 1723ad90a3deSWilliam Breathitt Gray help 1724ad90a3deSWilliam Breathitt Gray Expose PC/104 form factor device drivers and options available for 1725ad90a3deSWilliam Breathitt Gray selection and configuration. Enable this option if your target 1726ad90a3deSWilliam Breathitt Gray machine has a PC/104 bus. 1727ad90a3deSWilliam Breathitt Gray 172857c0c15bSIngo Molnarmenu "Kernel Performance Events And Counters" 17290793a61dSThomas Gleixner 1730cdd6c482SIngo Molnarconfig PERF_EVENTS 173157c0c15bSIngo Molnar bool "Kernel performance events and counters" 1732392d65a9SRobert Richter default y if PROFILING 1733cdd6c482SIngo Molnar depends on HAVE_PERF_EVENTS 1734e360adbeSPeter Zijlstra select IRQ_WORK 173583fe27eaSPranith Kumar select SRCU 17360793a61dSThomas Gleixner help 173757c0c15bSIngo Molnar Enable kernel support for various performance events provided 173857c0c15bSIngo Molnar by software and hardware. 17390793a61dSThomas Gleixner 1740dd77038dSThadeu Lima de Souza Cascardo Software events are supported either built-in or via the 174157c0c15bSIngo Molnar use of generic tracepoints. 174257c0c15bSIngo Molnar 174357c0c15bSIngo Molnar Most modern CPUs support performance events via performance 174457c0c15bSIngo Molnar counter registers. These registers count the number of certain 17450793a61dSThomas Gleixner types of hw events: such as instructions executed, cachemisses 17460793a61dSThomas Gleixner suffered, or branches mis-predicted - without slowing down the 17470793a61dSThomas Gleixner kernel or applications. These registers can also trigger interrupts 17480793a61dSThomas Gleixner when a threshold number of events have passed - and can thus be 17490793a61dSThomas Gleixner used to profile the code that runs on that CPU. 17500793a61dSThomas Gleixner 175157c0c15bSIngo Molnar The Linux Performance Event subsystem provides an abstraction of 1752dd77038dSThadeu Lima de Souza Cascardo these software and hardware event capabilities, available via a 175357c0c15bSIngo Molnar system call and used by the "perf" utility in tools/perf/. It 17540793a61dSThomas Gleixner provides per task and per CPU counters, and it provides event 17550793a61dSThomas Gleixner capabilities on top of those. 17560793a61dSThomas Gleixner 17570793a61dSThomas Gleixner Say Y if unsure. 17580793a61dSThomas Gleixner 1759906010b2SPeter Zijlstraconfig DEBUG_PERF_USE_VMALLOC 1760906010b2SPeter Zijlstra default n 1761906010b2SPeter Zijlstra bool "Debug: use vmalloc to back perf mmap() buffers" 1762cb307113SMichael Ellerman depends on PERF_EVENTS && DEBUG_KERNEL && !PPC 1763906010b2SPeter Zijlstra select PERF_USE_VMALLOC 1764906010b2SPeter Zijlstra help 1765906010b2SPeter Zijlstra Use vmalloc memory to back perf mmap() buffers. 1766906010b2SPeter Zijlstra 1767906010b2SPeter Zijlstra Mostly useful for debugging the vmalloc code on platforms 1768906010b2SPeter Zijlstra that don't require it. 1769906010b2SPeter Zijlstra 1770906010b2SPeter Zijlstra Say N if unsure. 1771906010b2SPeter Zijlstra 17720793a61dSThomas Gleixnerendmenu 17730793a61dSThomas Gleixner 1774f8891e5eSChristoph Lameterconfig VM_EVENT_COUNTERS 1775f8891e5eSChristoph Lameter default y 17766a108a14SDavid Rientjes bool "Enable VM event counters for /proc/vmstat" if EXPERT 1777f8891e5eSChristoph Lameter help 17782aea4fb6SPaul Jackson VM event counters are needed for event counts to be shown. 17792aea4fb6SPaul Jackson This option allows the disabling of the VM event counters 17806a108a14SDavid Rientjes on EXPERT systems. /proc/vmstat will only show page counts 17812aea4fb6SPaul Jackson if VM event counters are disabled. 1782f8891e5eSChristoph Lameter 178341ecc55bSChristoph Lameterconfig SLUB_DEBUG 178441ecc55bSChristoph Lameter default y 17856a108a14SDavid Rientjes bool "Enable SLUB debugging support" if EXPERT 1786f6acb635SChristoph Lameter depends on SLUB && SYSFS 178741ecc55bSChristoph Lameter help 178841ecc55bSChristoph Lameter SLUB has extensive debug support features. Disabling these can 178941ecc55bSChristoph Lameter result in significant savings in code size. This also disables 179041ecc55bSChristoph Lameter SLUB sysfs support. /sys/slab will not exist and there will be 179141ecc55bSChristoph Lameter no support for cache validation etc. 179241ecc55bSChristoph Lameter 17931663f26dSTejun Heoconfig SLUB_MEMCG_SYSFS_ON 17941663f26dSTejun Heo default n 17951663f26dSTejun Heo bool "Enable memcg SLUB sysfs support by default" if EXPERT 17961663f26dSTejun Heo depends on SLUB && SYSFS && MEMCG 17971663f26dSTejun Heo help 17981663f26dSTejun Heo SLUB creates a directory under /sys/kernel/slab for each 17991663f26dSTejun Heo allocation cache to host info and debug files. If memory 18001663f26dSTejun Heo cgroup is enabled, each cache can have per memory cgroup 18011663f26dSTejun Heo caches. SLUB can create the same sysfs directories for these 18021663f26dSTejun Heo caches under /sys/kernel/slab/CACHE/cgroup but it can lead 18031663f26dSTejun Heo to a very high number of debug files being created. This is 18041663f26dSTejun Heo controlled by slub_memcg_sysfs boot parameter and this 18051663f26dSTejun Heo config option determines the parameter's default value. 18061663f26dSTejun Heo 1807b943c460SRandy Dunlapconfig COMPAT_BRK 1808b943c460SRandy Dunlap bool "Disable heap randomization" 1809b943c460SRandy Dunlap default y 1810b943c460SRandy Dunlap help 1811b943c460SRandy Dunlap Randomizing heap placement makes heap exploits harder, but it 1812b943c460SRandy Dunlap also breaks ancient binaries (including anything libc5 based). 1813b943c460SRandy Dunlap This option changes the bootup default to heap randomization 1814692105b8SMatt LaPlante disabled, and can be overridden at runtime by setting 1815b943c460SRandy Dunlap /proc/sys/kernel/randomize_va_space to 2. 1816b943c460SRandy Dunlap 1817b943c460SRandy Dunlap On non-ancient distros (post-2000 ones) N is usually a safe choice. 1818b943c460SRandy Dunlap 181981819f0fSChristoph Lameterchoice 182081819f0fSChristoph Lameter prompt "Choose SLAB allocator" 1821a0acd820SChristoph Lameter default SLUB 182281819f0fSChristoph Lameter help 182381819f0fSChristoph Lameter This option allows to select a slab allocator. 182481819f0fSChristoph Lameter 182581819f0fSChristoph Lameterconfig SLAB 182681819f0fSChristoph Lameter bool "SLAB" 182704385fc5SKees Cook select HAVE_HARDENED_USERCOPY_ALLOCATOR 182881819f0fSChristoph Lameter help 182981819f0fSChristoph Lameter The regular slab allocator that is established and known to work 183034013886SChristoph Lameter well in all environments. It organizes cache hot objects in 183102f56210SSimon Arlott per cpu and per node queues. 183281819f0fSChristoph Lameter 183381819f0fSChristoph Lameterconfig SLUB 183481819f0fSChristoph Lameter bool "SLUB (Unqueued Allocator)" 1835ed18adc1SKees Cook select HAVE_HARDENED_USERCOPY_ALLOCATOR 183681819f0fSChristoph Lameter help 183781819f0fSChristoph Lameter SLUB is a slab allocator that minimizes cache line usage 183881819f0fSChristoph Lameter instead of managing queues of cached objects (SLAB approach). 183981819f0fSChristoph Lameter Per cpu caching is realized using slabs of objects instead 184081819f0fSChristoph Lameter of queues of objects. SLUB can use memory efficiently 184102f56210SSimon Arlott and has enhanced diagnostics. SLUB is the default choice for 184202f56210SSimon Arlott a slab allocator. 184381819f0fSChristoph Lameter 184481819f0fSChristoph Lameterconfig SLOB 18456a108a14SDavid Rientjes depends on EXPERT 184681819f0fSChristoph Lameter bool "SLOB (Simple Allocator)" 184781819f0fSChristoph Lameter help 184837291458SMatt Mackall SLOB replaces the stock allocator with a drastically simpler 184937291458SMatt Mackall allocator. SLOB is generally more space efficient but 185037291458SMatt Mackall does not perform as well on large systems. 185181819f0fSChristoph Lameter 185281819f0fSChristoph Lameterendchoice 185381819f0fSChristoph Lameter 18547660a6fdSKees Cookconfig SLAB_MERGE_DEFAULT 18557660a6fdSKees Cook bool "Allow slab caches to be merged" 18567660a6fdSKees Cook default y 18577660a6fdSKees Cook help 18587660a6fdSKees Cook For reduced kernel memory fragmentation, slab caches can be 18597660a6fdSKees Cook merged when they share the same size and other characteristics. 18607660a6fdSKees Cook This carries a risk of kernel heap overflows being able to 18617660a6fdSKees Cook overwrite objects from merged caches (and more easily control 18627660a6fdSKees Cook cache layout), which makes such heap attacks easier to exploit 18637660a6fdSKees Cook by attackers. By keeping caches unmerged, these kinds of exploits 18647660a6fdSKees Cook can usually only damage objects in the same cache. To disable 18657660a6fdSKees Cook merging at runtime, "slab_nomerge" can be passed on the kernel 18667660a6fdSKees Cook command line. 18677660a6fdSKees Cook 1868c7ce4f60SThomas Garnierconfig SLAB_FREELIST_RANDOM 1869c7ce4f60SThomas Garnier default n 1870210e7a43SThomas Garnier depends on SLAB || SLUB 1871c7ce4f60SThomas Garnier bool "SLAB freelist randomization" 1872c7ce4f60SThomas Garnier help 1873210e7a43SThomas Garnier Randomizes the freelist order used on creating new pages. This 1874c7ce4f60SThomas Garnier security feature reduces the predictability of the kernel slab 1875c7ce4f60SThomas Garnier allocator against heap overflows. 1876c7ce4f60SThomas Garnier 18772482ddecSKees Cookconfig SLAB_FREELIST_HARDENED 18782482ddecSKees Cook bool "Harden slab freelist metadata" 18792482ddecSKees Cook depends on SLUB 18802482ddecSKees Cook help 18812482ddecSKees Cook Many kernel heap attacks try to target slab cache metadata and 18822482ddecSKees Cook other infrastructure. This options makes minor performance 188392bae787SKees Cook sacrifices to harden the kernel slab allocator against common 18842482ddecSKees Cook freelist exploit methods. 18852482ddecSKees Cook 1886e900a918SDan Williamsconfig SHUFFLE_PAGE_ALLOCATOR 1887e900a918SDan Williams bool "Page allocator randomization" 1888e900a918SDan Williams default SLAB_FREELIST_RANDOM && ACPI_NUMA 1889e900a918SDan Williams help 1890e900a918SDan Williams Randomization of the page allocator improves the average 1891e900a918SDan Williams utilization of a direct-mapped memory-side-cache. See section 1892e900a918SDan Williams 5.2.27 Heterogeneous Memory Attribute Table (HMAT) in the ACPI 1893e900a918SDan Williams 6.2a specification for an example of how a platform advertises 1894e900a918SDan Williams the presence of a memory-side-cache. There are also incidental 1895e900a918SDan Williams security benefits as it reduces the predictability of page 1896e900a918SDan Williams allocations to compliment SLAB_FREELIST_RANDOM, but the 1897e900a918SDan Williams default granularity of shuffling on the "MAX_ORDER - 1" i.e, 1898e900a918SDan Williams 10th order of pages is selected based on cache utilization 1899e900a918SDan Williams benefits on x86. 1900e900a918SDan Williams 1901e900a918SDan Williams While the randomization improves cache utilization it may 1902e900a918SDan Williams negatively impact workloads on platforms without a cache. For 1903e900a918SDan Williams this reason, by default, the randomization is enabled only 1904e900a918SDan Williams after runtime detection of a direct-mapped memory-side-cache. 1905e900a918SDan Williams Otherwise, the randomization may be force enabled with the 1906e900a918SDan Williams 'page_alloc.shuffle' kernel command line parameter. 1907e900a918SDan Williams 1908e900a918SDan Williams Say Y if unsure. 1909e900a918SDan Williams 1910345c905dSJoonsoo Kimconfig SLUB_CPU_PARTIAL 1911345c905dSJoonsoo Kim default y 1912b39ffbf8SUwe Kleine-König depends on SLUB && SMP 1913345c905dSJoonsoo Kim bool "SLUB per cpu partial cache" 1914345c905dSJoonsoo Kim help 191592bae787SKees Cook Per cpu partial caches accelerate objects allocation and freeing 1916345c905dSJoonsoo Kim that is local to a processor at the price of more indeterminism 1917345c905dSJoonsoo Kim in the latency of the free. On overflow these caches will be cleared 1918345c905dSJoonsoo Kim which requires the taking of locks that may cause latency spikes. 1919345c905dSJoonsoo Kim Typically one would choose no for a realtime system. 1920345c905dSJoonsoo Kim 1921ea637639SJie Zhangconfig MMAP_ALLOW_UNINITIALIZED 1922ea637639SJie Zhang bool "Allow mmapped anonymous memory to be uninitialized" 19236a108a14SDavid Rientjes depends on EXPERT && !MMU 1924ea637639SJie Zhang default n 1925ea637639SJie Zhang help 1926ea637639SJie Zhang Normally, and according to the Linux spec, anonymous memory obtained 19273903bf94SRandy Dunlap from mmap() has its contents cleared before it is passed to 1928ea637639SJie Zhang userspace. Enabling this config option allows you to request that 1929ea637639SJie Zhang mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 1930ea637639SJie Zhang providing a huge performance boost. If this option is not enabled, 1931ea637639SJie Zhang then the flag will be ignored. 1932ea637639SJie Zhang 1933ea637639SJie Zhang This is taken advantage of by uClibc's malloc(), and also by 1934ea637639SJie Zhang ELF-FDPIC binfmt's brk and stack allocator. 1935ea637639SJie Zhang 1936ea637639SJie Zhang Because of the obvious security issues, this option should only be 1937ea637639SJie Zhang enabled on embedded devices where you control what is run in 1938ea637639SJie Zhang userspace. Since that isn't generally a problem on no-MMU systems, 1939ea637639SJie Zhang it is normally safe to say Y here. 1940ea637639SJie Zhang 1941ea637639SJie Zhang See Documentation/nommu-mmap.txt for more information. 1942ea637639SJie Zhang 1943091f6e26SDavid Howellsconfig SYSTEM_DATA_VERIFICATION 1944091f6e26SDavid Howells def_bool n 1945091f6e26SDavid Howells select SYSTEM_TRUSTED_KEYRING 1946091f6e26SDavid Howells select KEYS 1947091f6e26SDavid Howells select CRYPTO 1948d43de6c7SDavid Howells select CRYPTO_RSA 1949091f6e26SDavid Howells select ASYMMETRIC_KEY_TYPE 1950091f6e26SDavid Howells select ASYMMETRIC_PUBLIC_KEY_SUBTYPE 1951091f6e26SDavid Howells select ASN1 1952091f6e26SDavid Howells select OID_REGISTRY 1953091f6e26SDavid Howells select X509_CERTIFICATE_PARSER 1954091f6e26SDavid Howells select PKCS7_MESSAGE_PARSER 195582c04ff8SPeter Foley help 1956091f6e26SDavid Howells Provide PKCS#7 message verification using the contents of the system 1957091f6e26SDavid Howells trusted keyring to provide public keys. This then can be used for 1958091f6e26SDavid Howells module verification, kexec image verification and firmware blob 1959091f6e26SDavid Howells verification. 196082c04ff8SPeter Foley 1961125e5645SMathieu Desnoyersconfig PROFILING 1962b309a294SRobert Richter bool "Profiling support" 1963125e5645SMathieu Desnoyers help 1964125e5645SMathieu Desnoyers Say Y here to enable the extended profiling support mechanisms used 1965125e5645SMathieu Desnoyers by profilers such as OProfile. 1966125e5645SMathieu Desnoyers 19675f87f112SIngo Molnar# 19685f87f112SIngo Molnar# Place an empty function call at each tracepoint site. Can be 19695f87f112SIngo Molnar# dynamically changed for a probe function. 19705f87f112SIngo Molnar# 197197e1c18eSMathieu Desnoyersconfig TRACEPOINTS 19725f87f112SIngo Molnar bool 197397e1c18eSMathieu Desnoyers 19741da177e4SLinus Torvaldsendmenu # General setup 19751da177e4SLinus Torvalds 19761572497cSChristoph Hellwigsource "arch/Kconfig" 19771572497cSChristoph Hellwig 1978ae81f9e3SChuck Ebbertconfig RT_MUTEXES 19796341e62bSChristoph Jaeger bool 1980ae81f9e3SChuck Ebbert 19811da177e4SLinus Torvaldsconfig BASE_SMALL 19821da177e4SLinus Torvalds int 19831da177e4SLinus Torvalds default 0 if BASE_FULL 19841da177e4SLinus Torvalds default 1 if !BASE_FULL 19851da177e4SLinus Torvalds 1986c8424e77SThiago Jung Bauermannconfig MODULE_SIG_FORMAT 1987c8424e77SThiago Jung Bauermann def_bool n 1988c8424e77SThiago Jung Bauermann select SYSTEM_DATA_VERIFICATION 1989c8424e77SThiago Jung Bauermann 199066da5733SJan Engelhardtmenuconfig MODULES 19911da177e4SLinus Torvalds bool "Enable loadable module support" 199211097a03SYann E. MORIN option modules 19931da177e4SLinus Torvalds help 19941da177e4SLinus Torvalds Kernel modules are small pieces of compiled code which can 19951da177e4SLinus Torvalds be inserted in the running kernel, rather than being 19961da177e4SLinus Torvalds permanently built into the kernel. You use the "modprobe" 19971da177e4SLinus Torvalds tool to add (and sometimes remove) them. If you say Y here, 19981da177e4SLinus Torvalds many parts of the kernel can be built as modules (by 19991da177e4SLinus Torvalds answering M instead of Y where indicated): this is most 20001da177e4SLinus Torvalds useful for infrequently used options which are not required 20011da177e4SLinus Torvalds for booting. For more information, see the man pages for 20021da177e4SLinus Torvalds modprobe, lsmod, modinfo, insmod and rmmod. 20031da177e4SLinus Torvalds 20041da177e4SLinus Torvalds If you say Y here, you will need to run "make 20051da177e4SLinus Torvalds modules_install" to put the modules under /lib/modules/ 20061da177e4SLinus Torvalds where modprobe can find them (you may need to be root to do 20071da177e4SLinus Torvalds this). 20081da177e4SLinus Torvalds 20091da177e4SLinus Torvalds If unsure, say Y. 20101da177e4SLinus Torvalds 20110b0de144SRobert P. J. Dayif MODULES 20120b0de144SRobert P. J. Day 2013826e4506SLinus Torvaldsconfig MODULE_FORCE_LOAD 2014826e4506SLinus Torvalds bool "Forced module loading" 2015826e4506SLinus Torvalds default n 2016826e4506SLinus Torvalds help 201791e37a79SRusty Russell Allow loading of modules without version information (ie. modprobe 201891e37a79SRusty Russell --force). Forced module loading sets the 'F' (forced) taint flag and 201991e37a79SRusty Russell is usually a really bad idea. 2020826e4506SLinus Torvalds 20211da177e4SLinus Torvaldsconfig MODULE_UNLOAD 20221da177e4SLinus Torvalds bool "Module unloading" 20231da177e4SLinus Torvalds help 20241da177e4SLinus Torvalds Without this option you will not be able to unload any 20251da177e4SLinus Torvalds modules (note that some modules may not be unloadable 2026f7f5b675SDenys Vlasenko anyway), which makes your kernel smaller, faster 2027f7f5b675SDenys Vlasenko and simpler. If unsure, say Y. 20281da177e4SLinus Torvalds 20291da177e4SLinus Torvaldsconfig MODULE_FORCE_UNLOAD 20301da177e4SLinus Torvalds bool "Forced module unloading" 203119c92399SKees Cook depends on MODULE_UNLOAD 20321da177e4SLinus Torvalds help 20331da177e4SLinus Torvalds This option allows you to force a module to unload, even if the 20341da177e4SLinus Torvalds kernel believes it is unsafe: the kernel will remove the module 20351da177e4SLinus Torvalds without waiting for anyone to stop using it (using the -f option to 20361da177e4SLinus Torvalds rmmod). This is mainly for kernel developers and desperate users. 20371da177e4SLinus Torvalds If unsure, say N. 20381da177e4SLinus Torvalds 20391da177e4SLinus Torvaldsconfig MODVERSIONS 20400d541643SSam Ravnborg bool "Module versioning support" 20411da177e4SLinus Torvalds help 20421da177e4SLinus Torvalds Usually, you have to use modules compiled with your kernel. 20431da177e4SLinus Torvalds Saying Y here makes it sometimes possible to use modules 20441da177e4SLinus Torvalds compiled for different kernels, by adding enough information 20451da177e4SLinus Torvalds to the modules to (hopefully) spot any changes which would 20461da177e4SLinus Torvalds make them incompatible with the kernel you are running. If 20471da177e4SLinus Torvalds unsure, say N. 20481da177e4SLinus Torvalds 20492ff2b7ecSMasahiro Yamadaconfig ASM_MODVERSIONS 20502ff2b7ecSMasahiro Yamada bool 20512ff2b7ecSMasahiro Yamada default HAVE_ASM_MODVERSIONS && MODVERSIONS 20522ff2b7ecSMasahiro Yamada help 20532ff2b7ecSMasahiro Yamada This enables module versioning for exported symbols also from 20542ff2b7ecSMasahiro Yamada assembly. This can be enabled only when the target architecture 20552ff2b7ecSMasahiro Yamada supports it. 20562ff2b7ecSMasahiro Yamada 205756067812SArd Biesheuvelconfig MODULE_REL_CRCS 205856067812SArd Biesheuvel bool 205956067812SArd Biesheuvel depends on MODVERSIONS 206056067812SArd Biesheuvel 20611da177e4SLinus Torvaldsconfig MODULE_SRCVERSION_ALL 20621da177e4SLinus Torvalds bool "Source checksum for all modules" 20631da177e4SLinus Torvalds help 20641da177e4SLinus Torvalds Modules which contain a MODULE_VERSION get an extra "srcversion" 20651da177e4SLinus Torvalds field inserted into their modinfo section, which contains a 20661da177e4SLinus Torvalds sum of the source files which made it. This helps maintainers 20671da177e4SLinus Torvalds see exactly which source was used to build a module (since 20681da177e4SLinus Torvalds others sometimes change the module source without updating 20691da177e4SLinus Torvalds the version). With this option, such a "srcversion" field 20701da177e4SLinus Torvalds will be created for all modules. If unsure, say N. 20711da177e4SLinus Torvalds 2072106a4ee2SRusty Russellconfig MODULE_SIG 2073106a4ee2SRusty Russell bool "Module signature verification" 2074c8424e77SThiago Jung Bauermann select MODULE_SIG_FORMAT 2075106a4ee2SRusty Russell help 2076106a4ee2SRusty Russell Check modules for valid signatures upon load: the signature 2077106a4ee2SRusty Russell is simply appended to the module. For more information see 2078cbdc8217SNathan Chancellor <file:Documentation/admin-guide/module-signing.rst>. 2079106a4ee2SRusty Russell 2080228c37ffSDavid Howells Note that this option adds the OpenSSL development packages as a 2081228c37ffSDavid Howells kernel build dependency so that the signing tool can use its crypto 2082228c37ffSDavid Howells library. 2083228c37ffSDavid Howells 208449fcf732SDavid Howells You should enable this option if you wish to use either 208549fcf732SDavid Howells CONFIG_SECURITY_LOCKDOWN_LSM or lockdown functionality imposed via 208649fcf732SDavid Howells another LSM - otherwise unsigned modules will be loadable regardless 208749fcf732SDavid Howells of the lockdown policy. 208849fcf732SDavid Howells 2089ea0b6dcfSDavid Howells !!!WARNING!!! If you enable this option, you MUST make sure that the 2090ea0b6dcfSDavid Howells module DOES NOT get stripped after being signed. This includes the 2091ea0b6dcfSDavid Howells debuginfo strip done by some packagers (such as rpmbuild) and 2092ea0b6dcfSDavid Howells inclusion into an initramfs that wants the module size reduced. 2093ea0b6dcfSDavid Howells 2094106a4ee2SRusty Russellconfig MODULE_SIG_FORCE 2095106a4ee2SRusty Russell bool "Require modules to be validly signed" 2096106a4ee2SRusty Russell depends on MODULE_SIG 2097106a4ee2SRusty Russell help 2098106a4ee2SRusty Russell Reject unsigned modules or signed modules for which we don't have a 2099106a4ee2SRusty Russell key. Without this, such modules will simply taint the kernel. 2100ea0b6dcfSDavid Howells 2101d9d8d7edSMichal Marekconfig MODULE_SIG_ALL 2102d9d8d7edSMichal Marek bool "Automatically sign all modules" 2103d9d8d7edSMichal Marek default y 2104d9d8d7edSMichal Marek depends on MODULE_SIG 2105d9d8d7edSMichal Marek help 2106d9d8d7edSMichal Marek Sign all modules during make modules_install. Without this option, 2107d9d8d7edSMichal Marek modules must be signed manually, using the scripts/sign-file tool. 2108d9d8d7edSMichal Marek 2109d9d8d7edSMichal Marekcomment "Do not forget to sign required modules with scripts/sign-file" 2110d9d8d7edSMichal Marek depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL 2111d9d8d7edSMichal Marek 2112ea0b6dcfSDavid Howellschoice 2113ea0b6dcfSDavid Howells prompt "Which hash algorithm should modules be signed with?" 2114ea0b6dcfSDavid Howells depends on MODULE_SIG 2115ea0b6dcfSDavid Howells help 2116ea0b6dcfSDavid Howells This determines which sort of hashing algorithm will be used during 2117ea0b6dcfSDavid Howells signature generation. This algorithm _must_ be built into the kernel 2118ea0b6dcfSDavid Howells directly so that signature verification can take place. It is not 2119ea0b6dcfSDavid Howells possible to load a signed module containing the algorithm to check 2120ea0b6dcfSDavid Howells the signature on that module. 2121ea0b6dcfSDavid Howells 2122ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA1 2123ea0b6dcfSDavid Howells bool "Sign modules with SHA-1" 2124ea0b6dcfSDavid Howells select CRYPTO_SHA1 2125ea0b6dcfSDavid Howells 2126ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA224 2127ea0b6dcfSDavid Howells bool "Sign modules with SHA-224" 2128ea0b6dcfSDavid Howells select CRYPTO_SHA256 2129ea0b6dcfSDavid Howells 2130ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA256 2131ea0b6dcfSDavid Howells bool "Sign modules with SHA-256" 2132ea0b6dcfSDavid Howells select CRYPTO_SHA256 2133ea0b6dcfSDavid Howells 2134ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA384 2135ea0b6dcfSDavid Howells bool "Sign modules with SHA-384" 2136ea0b6dcfSDavid Howells select CRYPTO_SHA512 2137ea0b6dcfSDavid Howells 2138ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA512 2139ea0b6dcfSDavid Howells bool "Sign modules with SHA-512" 2140ea0b6dcfSDavid Howells select CRYPTO_SHA512 2141ea0b6dcfSDavid Howells 2142ea0b6dcfSDavid Howellsendchoice 2143ea0b6dcfSDavid Howells 214422753674SMichal Marekconfig MODULE_SIG_HASH 214522753674SMichal Marek string 214622753674SMichal Marek depends on MODULE_SIG 214722753674SMichal Marek default "sha1" if MODULE_SIG_SHA1 214822753674SMichal Marek default "sha224" if MODULE_SIG_SHA224 214922753674SMichal Marek default "sha256" if MODULE_SIG_SHA256 215022753674SMichal Marek default "sha384" if MODULE_SIG_SHA384 215122753674SMichal Marek default "sha512" if MODULE_SIG_SHA512 215222753674SMichal Marek 2153beb50df3SBertrand Jacquinconfig MODULE_COMPRESS 2154beb50df3SBertrand Jacquin bool "Compress modules on installation" 2155beb50df3SBertrand Jacquin help 2156beb50df3SBertrand Jacquin 2157b6c09b51SRusty Russell Compresses kernel modules when 'make modules_install' is run; gzip or 2158b6c09b51SRusty Russell xz depending on "Compression algorithm" below. 2159beb50df3SBertrand Jacquin 2160b6c09b51SRusty Russell module-init-tools MAY support gzip, and kmod MAY support gzip and xz. 2161beb50df3SBertrand Jacquin 2162b6c09b51SRusty Russell Out-of-tree kernel modules installed using Kbuild will also be 2163b6c09b51SRusty Russell compressed upon installation. 2164beb50df3SBertrand Jacquin 2165b6c09b51SRusty Russell Note: for modules inside an initrd or initramfs, it's more efficient 2166b6c09b51SRusty Russell to compress the whole initrd or initramfs instead. 2167beb50df3SBertrand Jacquin 2168b6c09b51SRusty Russell Note: This is fully compatible with signed modules. 2169b6c09b51SRusty Russell 2170b6c09b51SRusty Russell If in doubt, say N. 2171beb50df3SBertrand Jacquin 2172beb50df3SBertrand Jacquinchoice 2173beb50df3SBertrand Jacquin prompt "Compression algorithm" 2174beb50df3SBertrand Jacquin depends on MODULE_COMPRESS 2175beb50df3SBertrand Jacquin default MODULE_COMPRESS_GZIP 2176beb50df3SBertrand Jacquin help 2177beb50df3SBertrand Jacquin This determines which sort of compression will be used during 2178beb50df3SBertrand Jacquin 'make modules_install'. 2179beb50df3SBertrand Jacquin 2180beb50df3SBertrand Jacquin GZIP (default) and XZ are supported. 2181beb50df3SBertrand Jacquin 2182beb50df3SBertrand Jacquinconfig MODULE_COMPRESS_GZIP 2183beb50df3SBertrand Jacquin bool "GZIP" 2184beb50df3SBertrand Jacquin 2185beb50df3SBertrand Jacquinconfig MODULE_COMPRESS_XZ 2186beb50df3SBertrand Jacquin bool "XZ" 2187beb50df3SBertrand Jacquin 2188beb50df3SBertrand Jacquinendchoice 2189beb50df3SBertrand Jacquin 21903d52ec5eSMatthias Maennichconfig MODULE_ALLOW_MISSING_NAMESPACE_IMPORTS 21913d52ec5eSMatthias Maennich bool "Allow loading of modules with missing namespace imports" 21923d52ec5eSMatthias Maennich help 21933d52ec5eSMatthias Maennich Symbols exported with EXPORT_SYMBOL_NS*() are considered exported in 21943d52ec5eSMatthias Maennich a namespace. A module that makes use of a symbol exported with such a 21953d52ec5eSMatthias Maennich namespace is required to import the namespace via MODULE_IMPORT_NS(). 21963d52ec5eSMatthias Maennich There is no technical reason to enforce correct namespace imports, 21973d52ec5eSMatthias Maennich but it creates consistency between symbols defining namespaces and 21983d52ec5eSMatthias Maennich users importing namespaces they make use of. This option relaxes this 21993d52ec5eSMatthias Maennich requirement and lifts the enforcement when loading a module. 22003d52ec5eSMatthias Maennich 22013d52ec5eSMatthias Maennich If unsure, say N. 22023d52ec5eSMatthias Maennich 2203efd9763dSMasahiro Yamadaconfig UNUSED_SYMBOLS 2204efd9763dSMasahiro Yamada bool "Enable unused/obsolete exported symbols" 2205efd9763dSMasahiro Yamada default y if X86 2206efd9763dSMasahiro Yamada help 2207efd9763dSMasahiro Yamada Unused but exported symbols make the kernel needlessly bigger. For 2208efd9763dSMasahiro Yamada that reason most of these unused exports will soon be removed. This 2209efd9763dSMasahiro Yamada option is provided temporarily to provide a transition period in case 2210efd9763dSMasahiro Yamada some external kernel module needs one of these symbols anyway. If you 2211efd9763dSMasahiro Yamada encounter such a case in your module, consider if you are actually 2212efd9763dSMasahiro Yamada using the right API. (rationale: since nobody in the kernel is using 2213efd9763dSMasahiro Yamada this in a module, there is a pretty good chance it's actually the 2214efd9763dSMasahiro Yamada wrong interface to use). If you really need the symbol, please send a 2215efd9763dSMasahiro Yamada mail to the linux kernel mailing list mentioning the symbol and why 2216efd9763dSMasahiro Yamada you really need it, and what the merge plan to the mainline kernel for 2217efd9763dSMasahiro Yamada your module is. 2218efd9763dSMasahiro Yamada 2219dbacb0efSNicolas Pitreconfig TRIM_UNUSED_KSYMS 2220dbacb0efSNicolas Pitre bool "Trim unused exported kernel symbols" 2221d189c2a4SMasahiro Yamada depends on !UNUSED_SYMBOLS 2222dbacb0efSNicolas Pitre help 2223dbacb0efSNicolas Pitre The kernel and some modules make many symbols available for 2224dbacb0efSNicolas Pitre other modules to use via EXPORT_SYMBOL() and variants. Depending 2225dbacb0efSNicolas Pitre on the set of modules being selected in your kernel configuration, 2226dbacb0efSNicolas Pitre many of those exported symbols might never be used. 2227dbacb0efSNicolas Pitre 2228dbacb0efSNicolas Pitre This option allows for unused exported symbols to be dropped from 2229dbacb0efSNicolas Pitre the build. In turn, this provides the compiler more opportunities 2230dbacb0efSNicolas Pitre (especially when using LTO) for optimizing the code and reducing 2231dbacb0efSNicolas Pitre binary size. This might have some security advantages as well. 2232dbacb0efSNicolas Pitre 2233f1cb637eSValdis Kletnieks If unsure, or if you need to build out-of-tree modules, say N. 2234dbacb0efSNicolas Pitre 22351518c633SQuentin Perretconfig UNUSED_KSYMS_WHITELIST 22361518c633SQuentin Perret string "Whitelist of symbols to keep in ksymtab" 22371518c633SQuentin Perret depends on TRIM_UNUSED_KSYMS 22381518c633SQuentin Perret help 22391518c633SQuentin Perret By default, all unused exported symbols will be un-exported from the 22401518c633SQuentin Perret build when TRIM_UNUSED_KSYMS is selected. 22411518c633SQuentin Perret 22421518c633SQuentin Perret UNUSED_KSYMS_WHITELIST allows to whitelist symbols that must be kept 22431518c633SQuentin Perret exported at all times, even in absence of in-tree users. The value to 22441518c633SQuentin Perret set here is the path to a text file containing the list of symbols, 22451518c633SQuentin Perret one per line. The path can be absolute, or relative to the kernel 22461518c633SQuentin Perret source tree. 22471518c633SQuentin Perret 22480b0de144SRobert P. J. Dayendif # MODULES 22490b0de144SRobert P. J. Day 22506c9692e2SPeter Zijlstraconfig MODULES_TREE_LOOKUP 22516c9692e2SPeter Zijlstra def_bool y 22526c9692e2SPeter Zijlstra depends on PERF_EVENTS || TRACING 22536c9692e2SPeter Zijlstra 225498a79d6aSRusty Russellconfig INIT_ALL_POSSIBLE 225598a79d6aSRusty Russell bool 225698a79d6aSRusty Russell help 22575f054e31SRusty Russell Back when each arch used to define their own cpu_online_mask and 22585f054e31SRusty Russell cpu_possible_mask, some of them chose to initialize cpu_possible_mask 225998a79d6aSRusty Russell with all 1s, and others with all 0s. When they were centralised, 226098a79d6aSRusty Russell it was better to provide this option than to break all the archs 2261692105b8SMatt LaPlante and have several arch maintainers pursuing me down dark alleys. 226298a79d6aSRusty Russell 22633a65dfe8SJens Axboesource "block/Kconfig" 2264e98c3202SAvi Kivity 2265e98c3202SAvi Kivityconfig PREEMPT_NOTIFIERS 2266e98c3202SAvi Kivity bool 2267e260be67SPaul E. McKenney 226816295becSSteffen Klassertconfig PADATA 226916295becSSteffen Klassert depends on SMP 227016295becSSteffen Klassert bool 227116295becSSteffen Klassert 22724520c6a4SDavid Howellsconfig ASN1 22734520c6a4SDavid Howells tristate 22744520c6a4SDavid Howells help 22754520c6a4SDavid Howells Build a simple ASN.1 grammar compiler that produces a bytecode output 22764520c6a4SDavid Howells that can be interpreted by the ASN.1 stream decoder and used to 22774520c6a4SDavid Howells inform it as to what tags are to be expected in a stream and what 22784520c6a4SDavid Howells functions to call on what tags. 22794520c6a4SDavid Howells 22806beb0009SThomas Gleixnersource "kernel/Kconfig.locks" 2281e61938a9SMathieu Desnoyers 2282*0ebeea8cSDaniel Borkmannconfig ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE 2283*0ebeea8cSDaniel Borkmann bool 2284*0ebeea8cSDaniel Borkmann 2285e61938a9SMathieu Desnoyersconfig ARCH_HAS_SYNC_CORE_BEFORE_USERMODE 2286e61938a9SMathieu Desnoyers bool 22871bd21c6cSDominik Brodowski 22881bd21c6cSDominik Brodowski# It may be useful for an architecture to override the definitions of the 22897303e30eSDominik Brodowski# SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h> 22907303e30eSDominik Brodowski# and the COMPAT_ variants in <linux/compat.h>, in particular to use a 22917303e30eSDominik Brodowski# different calling convention for syscalls. They can also override the 22927303e30eSDominik Brodowski# macros for not-implemented syscalls in kernel/sys_ni.c and 22937303e30eSDominik Brodowski# kernel/time/posix-stubs.c. All these overrides need to be available in 22947303e30eSDominik Brodowski# <asm/syscall_wrapper.h>. 22951bd21c6cSDominik Brodowskiconfig ARCH_HAS_SYSCALL_WRAPPER 22961bd21c6cSDominik Brodowski def_bool n 2297