180daa560SRoman Zippelconfig ARCH 280daa560SRoman Zippel string 380daa560SRoman Zippel option env="ARCH" 480daa560SRoman Zippel 580daa560SRoman Zippelconfig KERNELVERSION 680daa560SRoman Zippel string 780daa560SRoman Zippel option env="KERNELVERSION" 880daa560SRoman Zippel 9face4374SRoman Zippelconfig DEFCONFIG_LIST 10face4374SRoman Zippel string 11b2670eacSPaolo 'Blaisorblade' Giarrusso depends on !UML 12face4374SRoman Zippel option defconfig_list 13face4374SRoman Zippel default "/lib/modules/$UNAME_RELEASE/.config" 14face4374SRoman Zippel default "/etc/kernel-config" 15face4374SRoman Zippel default "/boot/config-$UNAME_RELEASE" 1673531905SSam Ravnborg default "$ARCH_DEFCONFIG" 17face4374SRoman Zippel default "arch/$ARCH/defconfig" 18face4374SRoman Zippel 19b99b87f7SPeter Oberparleiterconfig CONSTRUCTORS 20b99b87f7SPeter Oberparleiter bool 21b99b87f7SPeter Oberparleiter depends on !UML 22b99b87f7SPeter Oberparleiter 23e360adbeSPeter Zijlstraconfig IRQ_WORK 24e360adbeSPeter Zijlstra bool 25e360adbeSPeter Zijlstra 261dbdc6f1SDavid Daneyconfig BUILDTIME_EXTABLE_SORT 271dbdc6f1SDavid Daney bool 281dbdc6f1SDavid Daney 29c65eacbeSAndy Lutomirskiconfig THREAD_INFO_IN_TASK 30c65eacbeSAndy Lutomirski bool 31c65eacbeSAndy Lutomirski help 32c65eacbeSAndy Lutomirski Select this to move thread_info off the stack into task_struct. To 33c65eacbeSAndy Lutomirski make this work, an arch will need to remove all thread_info fields 34c65eacbeSAndy Lutomirski except flags and fix any runtime bugs. 35c65eacbeSAndy Lutomirski 36c6c314a6SAndy Lutomirski One subtle change that will be needed is to use try_get_task_stack() 37c6c314a6SAndy Lutomirski and put_task_stack() in save_thread_stack_tsk() and get_wchan(). 38c6c314a6SAndy Lutomirski 39ff0cfc66SAl Boldimenu "General setup" 401da177e4SLinus Torvalds 411da177e4SLinus Torvaldsconfig BROKEN 421da177e4SLinus Torvalds bool 431da177e4SLinus Torvalds 441da177e4SLinus Torvaldsconfig BROKEN_ON_SMP 451da177e4SLinus Torvalds bool 461da177e4SLinus Torvalds depends on BROKEN || !SMP 471da177e4SLinus Torvalds default y 481da177e4SLinus Torvalds 491da177e4SLinus Torvaldsconfig INIT_ENV_ARG_LIMIT 501da177e4SLinus Torvalds int 51dd673bcaSAdrian Bunk default 32 if !UML 52dd673bcaSAdrian Bunk default 128 if UML 531da177e4SLinus Torvalds help 5434ad92c2SRandy Dunlap Maximum of each of the number of arguments and environment 5534ad92c2SRandy Dunlap variables passed to init from the kernel command line. 561da177e4SLinus Torvalds 571da177e4SLinus Torvalds 5884336466SRoland McGrathconfig CROSS_COMPILE 5984336466SRoland McGrath string "Cross-compiler tool prefix" 6084336466SRoland McGrath help 6184336466SRoland McGrath Same as running 'make CROSS_COMPILE=prefix-' but stored for 6284336466SRoland McGrath default make runs in this kernel build directory. You don't 6384336466SRoland McGrath need to set this unless you want the configured kernel build 6484336466SRoland McGrath directory to select the cross-compiler automatically. 6584336466SRoland McGrath 664bb16672SJiri Slabyconfig COMPILE_TEST 674bb16672SJiri Slaby bool "Compile also drivers which will not load" 68bc083a64SRichard Weinberger depends on !UML 694bb16672SJiri Slaby default n 704bb16672SJiri Slaby help 714bb16672SJiri Slaby Some drivers can be compiled on a different platform than they are 724bb16672SJiri Slaby intended to be run on. Despite they cannot be loaded there (or even 734bb16672SJiri Slaby when they load they cannot be used due to missing HW support), 744bb16672SJiri Slaby developers still, opposing to distributors, might want to build such 754bb16672SJiri Slaby drivers to compile-test them. 764bb16672SJiri Slaby 774bb16672SJiri Slaby If you are a developer and want to build everything available, say Y 784bb16672SJiri Slaby here. If you are a user/distributor, say N here to exclude useless 794bb16672SJiri Slaby drivers to be distributed. 804bb16672SJiri Slaby 811da177e4SLinus Torvaldsconfig LOCALVERSION 821da177e4SLinus Torvalds string "Local version - append to kernel release" 831da177e4SLinus Torvalds help 841da177e4SLinus Torvalds Append an extra string to the end of your kernel version. 851da177e4SLinus Torvalds This will show up when you type uname, for example. 861da177e4SLinus Torvalds The string you set here will be appended after the contents of 871da177e4SLinus Torvalds any files with a filename matching localversion* in your 881da177e4SLinus Torvalds object and source tree, in that order. Your total string can 891da177e4SLinus Torvalds be a maximum of 64 characters. 901da177e4SLinus Torvalds 91aaebf433SRyan Andersonconfig LOCALVERSION_AUTO 92aaebf433SRyan Anderson bool "Automatically append version information to the version string" 93aaebf433SRyan Anderson default y 94ac3339baSAlexey Dobriyan depends on !COMPILE_TEST 95aaebf433SRyan Anderson help 96aaebf433SRyan Anderson This will try to automatically determine if the current tree is a 976e5a5420SRobert P. J. Day release tree by looking for git tags that belong to the current 986e5a5420SRobert P. J. Day top of tree revision. 99aaebf433SRyan Anderson 100aaebf433SRyan Anderson A string of the format -gxxxxxxxx will be added to the localversion 1016e5a5420SRobert P. J. Day if a git-based tree is found. The string generated by this will be 102aaebf433SRyan Anderson appended after any matching localversion* files, and after the value 1036e5a5420SRobert P. J. Day set in CONFIG_LOCALVERSION. 104aaebf433SRyan Anderson 1056e5a5420SRobert P. J. Day (The actual string used here is the first eight characters produced 1066e5a5420SRobert P. J. Day by running the command: 1076e5a5420SRobert P. J. Day 1086e5a5420SRobert P. J. Day $ git rev-parse --verify HEAD 1096e5a5420SRobert P. J. Day 1106e5a5420SRobert P. J. Day which is done within the script "scripts/setlocalversion".) 111aaebf433SRyan Anderson 1122e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_GZIP 1132e9f3bddSH. Peter Anvin bool 1142e9f3bddSH. Peter Anvin 1152e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_BZIP2 1162e9f3bddSH. Peter Anvin bool 1172e9f3bddSH. Peter Anvin 1182e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_LZMA 1192e9f3bddSH. Peter Anvin bool 1202e9f3bddSH. Peter Anvin 1213ebe1243SLasse Collinconfig HAVE_KERNEL_XZ 1223ebe1243SLasse Collin bool 1233ebe1243SLasse Collin 1247dd65febSAlbin Tonnerreconfig HAVE_KERNEL_LZO 1257dd65febSAlbin Tonnerre bool 1267dd65febSAlbin Tonnerre 127e76e1fdfSKyungsik Leeconfig HAVE_KERNEL_LZ4 128e76e1fdfSKyungsik Lee bool 129e76e1fdfSKyungsik Lee 13030d65dbfSAlain Knaffchoice 13130d65dbfSAlain Knaff prompt "Kernel compression mode" 13230d65dbfSAlain Knaff default KERNEL_GZIP 1332d3c6275SH. Peter Anvin depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 13430d65dbfSAlain Knaff help 13530d65dbfSAlain Knaff The linux kernel is a kind of self-extracting executable. 13630d65dbfSAlain Knaff Several compression algorithms are available, which differ 13730d65dbfSAlain Knaff in efficiency, compression and decompression speed. 13830d65dbfSAlain Knaff Compression speed is only relevant when building a kernel. 13930d65dbfSAlain Knaff Decompression speed is relevant at each boot. 14030d65dbfSAlain Knaff 14130d65dbfSAlain Knaff If you have any problems with bzip2 or lzma compressed 14230d65dbfSAlain Knaff kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older 14330d65dbfSAlain Knaff version of this functionality (bzip2 only), for 2.4, was 14430d65dbfSAlain Knaff supplied by Christian Ludwig) 14530d65dbfSAlain Knaff 14630d65dbfSAlain Knaff High compression options are mostly useful for users, who 14730d65dbfSAlain Knaff are low on disk space (embedded systems), but for whom ram 14830d65dbfSAlain Knaff size matters less. 14930d65dbfSAlain Knaff 15030d65dbfSAlain Knaff If in doubt, select 'gzip' 15130d65dbfSAlain Knaff 15230d65dbfSAlain Knaffconfig KERNEL_GZIP 15330d65dbfSAlain Knaff bool "Gzip" 1542e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_GZIP 15530d65dbfSAlain Knaff help 1567dd65febSAlbin Tonnerre The old and tried gzip compression. It provides a good balance 1577dd65febSAlbin Tonnerre between compression ratio and decompression speed. 15830d65dbfSAlain Knaff 15930d65dbfSAlain Knaffconfig KERNEL_BZIP2 16030d65dbfSAlain Knaff bool "Bzip2" 1612e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_BZIP2 16230d65dbfSAlain Knaff help 16330d65dbfSAlain Knaff Its compression ratio and speed is intermediate. 1640a4dd35cSRandy Dunlap Decompression speed is slowest among the choices. The kernel 1652e9f3bddSH. Peter Anvin size is about 10% smaller with bzip2, in comparison to gzip. 1662e9f3bddSH. Peter Anvin Bzip2 uses a large amount of memory. For modern kernels you 1672e9f3bddSH. Peter Anvin will need at least 8MB RAM or more for booting. 16830d65dbfSAlain Knaff 16930d65dbfSAlain Knaffconfig KERNEL_LZMA 17030d65dbfSAlain Knaff bool "LZMA" 1712e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_LZMA 17230d65dbfSAlain Knaff help 1730a4dd35cSRandy Dunlap This compression algorithm's ratio is best. Decompression speed 1740a4dd35cSRandy Dunlap is between gzip and bzip2. Compression is slowest. 1750a4dd35cSRandy Dunlap The kernel size is about 33% smaller with LZMA in comparison to gzip. 17630d65dbfSAlain Knaff 1773ebe1243SLasse Collinconfig KERNEL_XZ 1783ebe1243SLasse Collin bool "XZ" 1793ebe1243SLasse Collin depends on HAVE_KERNEL_XZ 1803ebe1243SLasse Collin help 1813ebe1243SLasse Collin XZ uses the LZMA2 algorithm and instruction set specific 1823ebe1243SLasse Collin BCJ filters which can improve compression ratio of executable 1833ebe1243SLasse Collin code. The size of the kernel is about 30% smaller with XZ in 1843ebe1243SLasse Collin comparison to gzip. On architectures for which there is a BCJ 1853ebe1243SLasse Collin filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ 1863ebe1243SLasse Collin will create a few percent smaller kernel than plain LZMA. 1873ebe1243SLasse Collin 1883ebe1243SLasse Collin The speed is about the same as with LZMA: The decompression 1893ebe1243SLasse Collin speed of XZ is better than that of bzip2 but worse than gzip 1903ebe1243SLasse Collin and LZO. Compression is slow. 1913ebe1243SLasse Collin 1927dd65febSAlbin Tonnerreconfig KERNEL_LZO 1937dd65febSAlbin Tonnerre bool "LZO" 1947dd65febSAlbin Tonnerre depends on HAVE_KERNEL_LZO 1957dd65febSAlbin Tonnerre help 1960a4dd35cSRandy Dunlap Its compression ratio is the poorest among the choices. The kernel 197681b3049SStephan Sperber size is about 10% bigger than gzip; however its speed 1987dd65febSAlbin Tonnerre (both compression and decompression) is the fastest. 1997dd65febSAlbin Tonnerre 200e76e1fdfSKyungsik Leeconfig KERNEL_LZ4 201e76e1fdfSKyungsik Lee bool "LZ4" 202e76e1fdfSKyungsik Lee depends on HAVE_KERNEL_LZ4 203e76e1fdfSKyungsik Lee help 204e76e1fdfSKyungsik Lee LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding. 205e76e1fdfSKyungsik Lee A preliminary version of LZ4 de/compression tool is available at 206e76e1fdfSKyungsik Lee <https://code.google.com/p/lz4/>. 207e76e1fdfSKyungsik Lee 208e76e1fdfSKyungsik Lee Its compression ratio is worse than LZO. The size of the kernel 209e76e1fdfSKyungsik Lee is about 8% bigger than LZO. But the decompression speed is 210e76e1fdfSKyungsik Lee faster than LZO. 211e76e1fdfSKyungsik Lee 21230d65dbfSAlain Knaffendchoice 21330d65dbfSAlain Knaff 214bd5dc17bSJosh Triplettconfig DEFAULT_HOSTNAME 215bd5dc17bSJosh Triplett string "Default hostname" 216bd5dc17bSJosh Triplett default "(none)" 217bd5dc17bSJosh Triplett help 218bd5dc17bSJosh Triplett This option determines the default system hostname before userspace 219bd5dc17bSJosh Triplett calls sethostname(2). The kernel traditionally uses "(none)" here, 220bd5dc17bSJosh Triplett but you may wish to use a different default here to make a minimal 221bd5dc17bSJosh Triplett system more usable with less configuration. 222bd5dc17bSJosh Triplett 2231da177e4SLinus Torvaldsconfig SWAP 2241da177e4SLinus Torvalds bool "Support for paging of anonymous memory (swap)" 2259361401eSDavid Howells depends on MMU && BLOCK 2261da177e4SLinus Torvalds default y 2271da177e4SLinus Torvalds help 2281da177e4SLinus Torvalds This option allows you to choose whether you want to have support 2291da177e4SLinus Torvalds for so called swap devices or swap files in your kernel that are 2301da177e4SLinus Torvalds used to provide more virtual memory than the actual RAM present 2311da177e4SLinus Torvalds in your computer. If unsure say Y. 2321da177e4SLinus Torvalds 2331da177e4SLinus Torvaldsconfig SYSVIPC 2341da177e4SLinus Torvalds bool "System V IPC" 2351da177e4SLinus Torvalds ---help--- 2361da177e4SLinus Torvalds Inter Process Communication is a suite of library functions and 2371da177e4SLinus Torvalds system calls which let processes (running programs) synchronize and 2381da177e4SLinus Torvalds exchange information. It is generally considered to be a good thing, 2391da177e4SLinus Torvalds and some programs won't run unless you say Y here. In particular, if 2401da177e4SLinus Torvalds you want to run the DOS emulator dosemu under Linux (read the 2411da177e4SLinus Torvalds DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>), 2421da177e4SLinus Torvalds you'll need to say Y here. 2431da177e4SLinus Torvalds 2441da177e4SLinus Torvalds You can find documentation about IPC with "info ipc" and also in 2451da177e4SLinus Torvalds section 6.4 of the Linux Programmer's Guide, available from 2461da177e4SLinus Torvalds <http://www.tldp.org/guides.html>. 2471da177e4SLinus Torvalds 248a5494dcdSEric W. Biedermanconfig SYSVIPC_SYSCTL 249a5494dcdSEric W. Biederman bool 250a5494dcdSEric W. Biederman depends on SYSVIPC 251a5494dcdSEric W. Biederman depends on SYSCTL 252a5494dcdSEric W. Biederman default y 253a5494dcdSEric W. Biederman 2541da177e4SLinus Torvaldsconfig POSIX_MQUEUE 2551da177e4SLinus Torvalds bool "POSIX Message Queues" 25619c92399SKees Cook depends on NET 2571da177e4SLinus Torvalds ---help--- 2581da177e4SLinus Torvalds POSIX variant of message queues is a part of IPC. In POSIX message 2591da177e4SLinus Torvalds queues every message has a priority which decides about succession 2601da177e4SLinus Torvalds of receiving it by a process. If you want to compile and run 2611da177e4SLinus Torvalds programs written e.g. for Solaris with use of its POSIX message 262b0e37650SRobert P. J. Day queues (functions mq_*) say Y here. 2631da177e4SLinus Torvalds 2641da177e4SLinus Torvalds POSIX message queues are visible as a filesystem called 'mqueue' 2651da177e4SLinus Torvalds and can be mounted somewhere if you want to do filesystem 2661da177e4SLinus Torvalds operations on message queues. 2671da177e4SLinus Torvalds 2681da177e4SLinus Torvalds If unsure, say Y. 2691da177e4SLinus Torvalds 270bdc8e5f8SSerge E. Hallynconfig POSIX_MQUEUE_SYSCTL 271bdc8e5f8SSerge E. Hallyn bool 272bdc8e5f8SSerge E. Hallyn depends on POSIX_MQUEUE 273bdc8e5f8SSerge E. Hallyn depends on SYSCTL 274bdc8e5f8SSerge E. Hallyn default y 275bdc8e5f8SSerge E. Hallyn 276226b4ccdSKonstantin Khlebnikovconfig CROSS_MEMORY_ATTACH 277226b4ccdSKonstantin Khlebnikov bool "Enable process_vm_readv/writev syscalls" 278226b4ccdSKonstantin Khlebnikov depends on MMU 279226b4ccdSKonstantin Khlebnikov default y 280226b4ccdSKonstantin Khlebnikov help 281226b4ccdSKonstantin Khlebnikov Enabling this option adds the system calls process_vm_readv and 282226b4ccdSKonstantin Khlebnikov process_vm_writev which allow a process with the correct privileges 283a2a368d9SGeert Uytterhoeven to directly read from or write to another process' address space. 284226b4ccdSKonstantin Khlebnikov See the man page for more details. 285226b4ccdSKonstantin Khlebnikov 28669369a70SJosh Triplettconfig USELIB 28769369a70SJosh Triplett bool "uselib syscall" 288b2113a41SRiku Voipio def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION 28969369a70SJosh Triplett help 29069369a70SJosh Triplett This option enables the uselib syscall, a system call used in the 29169369a70SJosh Triplett dynamic linker from libc5 and earlier. glibc does not use this 29269369a70SJosh Triplett system call. If you intend to run programs built on libc5 or 29369369a70SJosh Triplett earlier, you may need to enable this syscall. Current systems 29469369a70SJosh Triplett running glibc can safely disable this. 29569369a70SJosh Triplett 2961da177e4SLinus Torvaldsconfig AUDIT 2971da177e4SLinus Torvalds bool "Auditing support" 298804a6a49SChris Wright depends on NET 2991da177e4SLinus Torvalds help 3001da177e4SLinus Torvalds Enable auditing infrastructure that can be used with another 3011da177e4SLinus Torvalds kernel subsystem, such as SELinux (which requires this for 302cb74ed27SPaul Moore logging of avc messages output). System call auditing is included 303cb74ed27SPaul Moore on architectures which support it. 3041da177e4SLinus Torvalds 3057a017721SAKASHI Takahiroconfig HAVE_ARCH_AUDITSYSCALL 3067a017721SAKASHI Takahiro bool 3077a017721SAKASHI Takahiro 3081da177e4SLinus Torvaldsconfig AUDITSYSCALL 309cb74ed27SPaul Moore def_bool y 3107a017721SAKASHI Takahiro depends on AUDIT && HAVE_ARCH_AUDITSYSCALL 3111da177e4SLinus Torvalds 312939a67fcSEric Parisconfig AUDIT_WATCH 313939a67fcSEric Paris def_bool y 314939a67fcSEric Paris depends on AUDITSYSCALL 315939a67fcSEric Paris select FSNOTIFY 3161da177e4SLinus Torvalds 31774c3cbe3SAl Viroconfig AUDIT_TREE 31874c3cbe3SAl Viro def_bool y 31963c882a0SEric Paris depends on AUDITSYSCALL 32028a3a7ebSEric Paris select FSNOTIFY 32174c3cbe3SAl Viro 322d9817ebeSThomas Gleixnersource "kernel/irq/Kconfig" 323764e0da1SThomas Gleixnersource "kernel/time/Kconfig" 324d9817ebeSThomas Gleixner 325391dc69cSFrederic Weisbeckermenu "CPU/Task time and stats accounting" 326391dc69cSFrederic Weisbecker 327abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING 328abf917cdSFrederic Weisbecker bool 329abf917cdSFrederic Weisbecker 330fdf9c356SFrederic Weisbeckerchoice 331fdf9c356SFrederic Weisbecker prompt "Cputime accounting" 332fdf9c356SFrederic Weisbecker default TICK_CPU_ACCOUNTING if !PPC64 33302fc8d37SStephen Rothwell default VIRT_CPU_ACCOUNTING_NATIVE if PPC64 334fdf9c356SFrederic Weisbecker 335fdf9c356SFrederic Weisbecker# Kind of a stub config for the pure tick based cputime accounting 336fdf9c356SFrederic Weisbeckerconfig TICK_CPU_ACCOUNTING 337fdf9c356SFrederic Weisbecker bool "Simple tick based cputime accounting" 338c58b0df1SFrederic Weisbecker depends on !S390 && !NO_HZ_FULL 339fdf9c356SFrederic Weisbecker help 340fdf9c356SFrederic Weisbecker This is the basic tick based cputime accounting that maintains 341fdf9c356SFrederic Weisbecker statistics about user, system and idle time spent on per jiffies 342fdf9c356SFrederic Weisbecker granularity. 343fdf9c356SFrederic Weisbecker 344fdf9c356SFrederic Weisbecker If unsure, say Y. 345fdf9c356SFrederic Weisbecker 346abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING_NATIVE 347391dc69cSFrederic Weisbecker bool "Deterministic task and CPU time accounting" 348c58b0df1SFrederic Weisbecker depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL 349abf917cdSFrederic Weisbecker select VIRT_CPU_ACCOUNTING 350391dc69cSFrederic Weisbecker help 351391dc69cSFrederic Weisbecker Select this option to enable more accurate task and CPU time 352391dc69cSFrederic Weisbecker accounting. This is done by reading a CPU counter on each 353391dc69cSFrederic Weisbecker kernel entry and exit and on transitions within the kernel 354391dc69cSFrederic Weisbecker between system, softirq and hardirq state, so there is a 355391dc69cSFrederic Weisbecker small performance impact. In the case of s390 or IBM POWER > 5, 356391dc69cSFrederic Weisbecker this also enables accounting of stolen time on logically-partitioned 357391dc69cSFrederic Weisbecker systems. 358391dc69cSFrederic Weisbecker 359abf917cdSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING_GEN 360abf917cdSFrederic Weisbecker bool "Full dynticks CPU time accounting" 361ff3fb254SKevin Hilman depends on HAVE_CONTEXT_TRACKING 362554b0004SKevin Hilman depends on HAVE_VIRT_CPU_ACCOUNTING_GEN 363abf917cdSFrederic Weisbecker select VIRT_CPU_ACCOUNTING 364abf917cdSFrederic Weisbecker select CONTEXT_TRACKING 365abf917cdSFrederic Weisbecker help 366abf917cdSFrederic Weisbecker Select this option to enable task and CPU time accounting on full 367abf917cdSFrederic Weisbecker dynticks systems. This accounting is implemented by watching every 368abf917cdSFrederic Weisbecker kernel-user boundaries using the context tracking subsystem. 369abf917cdSFrederic Weisbecker The accounting is thus performed at the expense of some significant 370abf917cdSFrederic Weisbecker overhead. 371abf917cdSFrederic Weisbecker 372abf917cdSFrederic Weisbecker For now this is only useful if you are working on the full 373abf917cdSFrederic Weisbecker dynticks subsystem development. 374abf917cdSFrederic Weisbecker 375abf917cdSFrederic Weisbecker If unsure, say N. 376abf917cdSFrederic Weisbecker 377b58c3584SRik van Rielendchoice 378b58c3584SRik van Riel 379fdf9c356SFrederic Weisbeckerconfig IRQ_TIME_ACCOUNTING 380fdf9c356SFrederic Weisbecker bool "Fine granularity task level IRQ time accounting" 381b58c3584SRik van Riel depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE 382fdf9c356SFrederic Weisbecker help 383fdf9c356SFrederic Weisbecker Select this option to enable fine granularity task irq time 384fdf9c356SFrederic Weisbecker accounting. This is done by reading a timestamp on each 385fdf9c356SFrederic Weisbecker transitions between softirq and hardirq state, so there can be a 386fdf9c356SFrederic Weisbecker small performance impact. 387fdf9c356SFrederic Weisbecker 388fdf9c356SFrederic Weisbecker If in doubt, say N here. 389fdf9c356SFrederic Weisbecker 390391dc69cSFrederic Weisbeckerconfig BSD_PROCESS_ACCT 391391dc69cSFrederic Weisbecker bool "BSD Process Accounting" 3922813893fSIulia Manda depends on MULTIUSER 393391dc69cSFrederic Weisbecker help 394391dc69cSFrederic Weisbecker If you say Y here, a user level program will be able to instruct the 395391dc69cSFrederic Weisbecker kernel (via a special system call) to write process accounting 396391dc69cSFrederic Weisbecker information to a file: whenever a process exits, information about 397391dc69cSFrederic Weisbecker that process will be appended to the file by the kernel. The 398391dc69cSFrederic Weisbecker information includes things such as creation time, owning user, 399391dc69cSFrederic Weisbecker command name, memory usage, controlling terminal etc. (the complete 400391dc69cSFrederic Weisbecker list is in the struct acct in <file:include/linux/acct.h>). It is 401391dc69cSFrederic Weisbecker up to the user level program to do useful things with this 402391dc69cSFrederic Weisbecker information. This is generally a good idea, so say Y. 403391dc69cSFrederic Weisbecker 404391dc69cSFrederic Weisbeckerconfig BSD_PROCESS_ACCT_V3 405391dc69cSFrederic Weisbecker bool "BSD Process Accounting version 3 file format" 406391dc69cSFrederic Weisbecker depends on BSD_PROCESS_ACCT 407391dc69cSFrederic Weisbecker default n 408391dc69cSFrederic Weisbecker help 409391dc69cSFrederic Weisbecker If you say Y here, the process accounting information is written 410391dc69cSFrederic Weisbecker in a new file format that also logs the process IDs of each 411391dc69cSFrederic Weisbecker process and it's parent. Note that this file format is incompatible 412391dc69cSFrederic Weisbecker with previous v0/v1/v2 file formats, so you will need updated tools 413391dc69cSFrederic Weisbecker for processing it. A preliminary version of these tools is available 414391dc69cSFrederic Weisbecker at <http://www.gnu.org/software/acct/>. 415391dc69cSFrederic Weisbecker 416391dc69cSFrederic Weisbeckerconfig TASKSTATS 41719c92399SKees Cook bool "Export task/process statistics through netlink" 418391dc69cSFrederic Weisbecker depends on NET 4192813893fSIulia Manda depends on MULTIUSER 420391dc69cSFrederic Weisbecker default n 421391dc69cSFrederic Weisbecker help 422391dc69cSFrederic Weisbecker Export selected statistics for tasks/processes through the 423391dc69cSFrederic Weisbecker generic netlink interface. Unlike BSD process accounting, the 424391dc69cSFrederic Weisbecker statistics are available during the lifetime of tasks/processes as 425391dc69cSFrederic Weisbecker responses to commands. Like BSD accounting, they are sent to user 426391dc69cSFrederic Weisbecker space on task exit. 427391dc69cSFrederic Weisbecker 428391dc69cSFrederic Weisbecker Say N if unsure. 429391dc69cSFrederic Weisbecker 430391dc69cSFrederic Weisbeckerconfig TASK_DELAY_ACCT 43119c92399SKees Cook bool "Enable per-task delay accounting" 432391dc69cSFrederic Weisbecker depends on TASKSTATS 433f6db8347SNaveen N. Rao select SCHED_INFO 434391dc69cSFrederic Weisbecker help 435391dc69cSFrederic Weisbecker Collect information on time spent by a task waiting for system 436391dc69cSFrederic Weisbecker resources like cpu, synchronous block I/O completion and swapping 437391dc69cSFrederic Weisbecker in pages. Such statistics can help in setting a task's priorities 438391dc69cSFrederic Weisbecker relative to other tasks for cpu, io, rss limits etc. 439391dc69cSFrederic Weisbecker 440391dc69cSFrederic Weisbecker Say N if unsure. 441391dc69cSFrederic Weisbecker 442391dc69cSFrederic Weisbeckerconfig TASK_XACCT 44319c92399SKees Cook bool "Enable extended accounting over taskstats" 444391dc69cSFrederic Weisbecker depends on TASKSTATS 445391dc69cSFrederic Weisbecker help 446391dc69cSFrederic Weisbecker Collect extended task accounting data and send the data 447391dc69cSFrederic Weisbecker to userland for processing over the taskstats interface. 448391dc69cSFrederic Weisbecker 449391dc69cSFrederic Weisbecker Say N if unsure. 450391dc69cSFrederic Weisbecker 451391dc69cSFrederic Weisbeckerconfig TASK_IO_ACCOUNTING 45219c92399SKees Cook bool "Enable per-task storage I/O accounting" 453391dc69cSFrederic Weisbecker depends on TASK_XACCT 454391dc69cSFrederic Weisbecker help 455391dc69cSFrederic Weisbecker Collect information on the number of bytes of storage I/O which this 456391dc69cSFrederic Weisbecker task has caused. 457391dc69cSFrederic Weisbecker 458391dc69cSFrederic Weisbecker Say N if unsure. 459391dc69cSFrederic Weisbecker 460391dc69cSFrederic Weisbeckerendmenu # "CPU/Task time and stats accounting" 461391dc69cSFrederic Weisbecker 4625c4991e2SFrederic Weisbeckerconfig CPU_ISOLATION 4635c4991e2SFrederic Weisbecker bool "CPU isolation" 464414a2dc1SGeert Uytterhoeven depends on SMP || COMPILE_TEST 4652c43838cSFrederic Weisbecker default y 4665c4991e2SFrederic Weisbecker help 4675c4991e2SFrederic Weisbecker Make sure that CPUs running critical tasks are not disturbed by 4685c4991e2SFrederic Weisbecker any source of "noise" such as unbound workqueues, timers, kthreads... 4692c43838cSFrederic Weisbecker Unbound jobs get offloaded to housekeeping CPUs. This is driven by 4702c43838cSFrederic Weisbecker the "isolcpus=" boot parameter. 4712c43838cSFrederic Weisbecker 4722c43838cSFrederic Weisbecker Say Y if unsure. 4735c4991e2SFrederic Weisbecker 4740af92d46SPaul E. McKenneysource "kernel/rcu/Kconfig" 475c903ff83SMike Travis 476de5b56baSVivek Goyalconfig BUILD_BIN2C 477de5b56baSVivek Goyal bool 478de5b56baSVivek Goyal default n 479de5b56baSVivek Goyal 4801da177e4SLinus Torvaldsconfig IKCONFIG 481f2443ab6SRoss Biro tristate "Kernel .config support" 482de5b56baSVivek Goyal select BUILD_BIN2C 4831da177e4SLinus Torvalds ---help--- 4841da177e4SLinus Torvalds This option enables the complete Linux kernel ".config" file 4851da177e4SLinus Torvalds contents to be saved in the kernel. It provides documentation 4861da177e4SLinus Torvalds of which kernel options are used in a running kernel or in an 4871da177e4SLinus Torvalds on-disk kernel. This information can be extracted from the kernel 4881da177e4SLinus Torvalds image file with the script scripts/extract-ikconfig and used as 4891da177e4SLinus Torvalds input to rebuild the current kernel or to build another kernel. 4901da177e4SLinus Torvalds It can also be extracted from a running kernel by reading 4911da177e4SLinus Torvalds /proc/config.gz if enabled (below). 4921da177e4SLinus Torvalds 4931da177e4SLinus Torvaldsconfig IKCONFIG_PROC 4941da177e4SLinus Torvalds bool "Enable access to .config through /proc/config.gz" 4951da177e4SLinus Torvalds depends on IKCONFIG && PROC_FS 4961da177e4SLinus Torvalds ---help--- 4971da177e4SLinus Torvalds This option enables access to the kernel configuration file 4981da177e4SLinus Torvalds through /proc/config.gz. 4991da177e4SLinus Torvalds 500794543a2SAlistair John Strachanconfig LOG_BUF_SHIFT 501794543a2SAlistair John Strachan int "Kernel log buffer size (16 => 64KB, 17 => 128KB)" 502fb39f98dSIngo Molnar range 12 25 503f17a32e9SAdrian Bunk default 17 504361e9dfbSJosh Triplett depends on PRINTK 505794543a2SAlistair John Strachan help 50623b2899fSLuis R. Rodriguez Select the minimal kernel log buffer size as a power of 2. 50723b2899fSLuis R. Rodriguez The final size is affected by LOG_CPU_MAX_BUF_SHIFT config 50823b2899fSLuis R. Rodriguez parameter, see below. Any higher size also might be forced 50923b2899fSLuis R. Rodriguez by "log_buf_len" boot parameter. 51023b2899fSLuis R. Rodriguez 511f17a32e9SAdrian Bunk Examples: 512f17a32e9SAdrian Bunk 17 => 128 KB 513f17a32e9SAdrian Bunk 16 => 64 KB 514f17a32e9SAdrian Bunk 15 => 32 KB 515f17a32e9SAdrian Bunk 14 => 16 KB 516794543a2SAlistair John Strachan 13 => 8 KB 517794543a2SAlistair John Strachan 12 => 4 KB 518794543a2SAlistair John Strachan 51923b2899fSLuis R. Rodriguezconfig LOG_CPU_MAX_BUF_SHIFT 52023b2899fSLuis R. Rodriguez int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)" 5212240a31dSGeert Uytterhoeven depends on SMP 52223b2899fSLuis R. Rodriguez range 0 21 52323b2899fSLuis R. Rodriguez default 12 if !BASE_SMALL 52423b2899fSLuis R. Rodriguez default 0 if BASE_SMALL 525361e9dfbSJosh Triplett depends on PRINTK 52623b2899fSLuis R. Rodriguez help 52723b2899fSLuis R. Rodriguez This option allows to increase the default ring buffer size 52823b2899fSLuis R. Rodriguez according to the number of CPUs. The value defines the contribution 52923b2899fSLuis R. Rodriguez of each CPU as a power of 2. The used space is typically only few 53023b2899fSLuis R. Rodriguez lines however it might be much more when problems are reported, 53123b2899fSLuis R. Rodriguez e.g. backtraces. 53223b2899fSLuis R. Rodriguez 53323b2899fSLuis R. Rodriguez The increased size means that a new buffer has to be allocated and 53423b2899fSLuis R. Rodriguez the original static one is unused. It makes sense only on systems 53523b2899fSLuis R. Rodriguez with more CPUs. Therefore this value is used only when the sum of 53623b2899fSLuis R. Rodriguez contributions is greater than the half of the default kernel ring 53723b2899fSLuis R. Rodriguez buffer as defined by LOG_BUF_SHIFT. The default values are set 53823b2899fSLuis R. Rodriguez so that more than 64 CPUs are needed to trigger the allocation. 53923b2899fSLuis R. Rodriguez 54023b2899fSLuis R. Rodriguez Also this option is ignored when "log_buf_len" kernel parameter is 54123b2899fSLuis R. Rodriguez used as it forces an exact (power of two) size of the ring buffer. 54223b2899fSLuis R. Rodriguez 54323b2899fSLuis R. Rodriguez The number of possible CPUs is used for this computation ignoring 5445e0d8d59SGeert Uytterhoeven hotplugging making the computation optimal for the worst case 5455e0d8d59SGeert Uytterhoeven scenario while allowing a simple algorithm to be used from bootup. 54623b2899fSLuis R. Rodriguez 54723b2899fSLuis R. Rodriguez Examples shift values and their meaning: 54823b2899fSLuis R. Rodriguez 17 => 128 KB for each CPU 54923b2899fSLuis R. Rodriguez 16 => 64 KB for each CPU 55023b2899fSLuis R. Rodriguez 15 => 32 KB for each CPU 55123b2899fSLuis R. Rodriguez 14 => 16 KB for each CPU 55223b2899fSLuis R. Rodriguez 13 => 8 KB for each CPU 55323b2899fSLuis R. Rodriguez 12 => 4 KB for each CPU 55423b2899fSLuis R. Rodriguez 555f92bac3bSSergey Senozhatskyconfig PRINTK_SAFE_LOG_BUF_SHIFT 556f92bac3bSSergey Senozhatsky int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)" 557427934b8SPetr Mladek range 10 21 558427934b8SPetr Mladek default 13 559f92bac3bSSergey Senozhatsky depends on PRINTK 560427934b8SPetr Mladek help 561f92bac3bSSergey Senozhatsky Select the size of an alternate printk per-CPU buffer where messages 562f92bac3bSSergey Senozhatsky printed from usafe contexts are temporary stored. One example would 563f92bac3bSSergey Senozhatsky be NMI messages, another one - printk recursion. The messages are 564f92bac3bSSergey Senozhatsky copied to the main log buffer in a safe context to avoid a deadlock. 565f92bac3bSSergey Senozhatsky The value defines the size as a power of 2. 566427934b8SPetr Mladek 567f92bac3bSSergey Senozhatsky Those messages are rare and limited. The largest one is when 568427934b8SPetr Mladek a backtrace is printed. It usually fits into 4KB. Select 569427934b8SPetr Mladek 8KB if you want to be on the safe side. 570427934b8SPetr Mladek 571427934b8SPetr Mladek Examples: 572427934b8SPetr Mladek 17 => 128 KB for each CPU 573427934b8SPetr Mladek 16 => 64 KB for each CPU 574427934b8SPetr Mladek 15 => 32 KB for each CPU 575427934b8SPetr Mladek 14 => 16 KB for each CPU 576427934b8SPetr Mladek 13 => 8 KB for each CPU 577427934b8SPetr Mladek 12 => 4 KB for each CPU 578427934b8SPetr Mladek 5795cdc38f9SKAMEZAWA Hiroyuki# 5805cdc38f9SKAMEZAWA Hiroyuki# Architectures with an unreliable sched_clock() should select this: 5815cdc38f9SKAMEZAWA Hiroyuki# 5825cdc38f9SKAMEZAWA Hiroyukiconfig HAVE_UNSTABLE_SCHED_CLOCK 5835cdc38f9SKAMEZAWA Hiroyuki bool 5845cdc38f9SKAMEZAWA Hiroyuki 58538ff87f7SStephen Boydconfig GENERIC_SCHED_CLOCK 58638ff87f7SStephen Boyd bool 58738ff87f7SStephen Boyd 588be3a7284SAndrea Arcangeli# 589be3a7284SAndrea Arcangeli# For architectures that want to enable the support for NUMA-affine scheduler 590be3a7284SAndrea Arcangeli# balancing logic: 591be3a7284SAndrea Arcangeli# 592be3a7284SAndrea Arcangeliconfig ARCH_SUPPORTS_NUMA_BALANCING 593be3a7284SAndrea Arcangeli bool 594be3a7284SAndrea Arcangeli 595be5e610cSPeter Zijlstra# 59672b252aeSMel Gorman# For architectures that prefer to flush all TLBs after a number of pages 59772b252aeSMel Gorman# are unmapped instead of sending one IPI per page to flush. The architecture 59872b252aeSMel Gorman# must provide guarantees on what happens if a clean TLB cache entry is 59972b252aeSMel Gorman# written after the unmap. Details are in mm/rmap.c near the check for 60072b252aeSMel Gorman# should_defer_flush. The architecture should also consider if the full flush 60172b252aeSMel Gorman# and the refill costs are offset by the savings of sending fewer IPIs. 60272b252aeSMel Gormanconfig ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH 60372b252aeSMel Gorman bool 60472b252aeSMel Gorman 60572b252aeSMel Gorman# 606be5e610cSPeter Zijlstra# For architectures that know their GCC __int128 support is sound 607be5e610cSPeter Zijlstra# 608be5e610cSPeter Zijlstraconfig ARCH_SUPPORTS_INT128 609be5e610cSPeter Zijlstra bool 610be5e610cSPeter Zijlstra 611be3a7284SAndrea Arcangeli# For architectures that (ab)use NUMA to represent different memory regions 612be3a7284SAndrea Arcangeli# all cpu-local but of different latencies, such as SuperH. 613be3a7284SAndrea Arcangeli# 614be3a7284SAndrea Arcangeliconfig ARCH_WANT_NUMA_VARIABLE_LOCALITY 615be3a7284SAndrea Arcangeli bool 616be3a7284SAndrea Arcangeli 617be3a7284SAndrea Arcangeliconfig NUMA_BALANCING 618be3a7284SAndrea Arcangeli bool "Memory placement aware NUMA scheduler" 619be3a7284SAndrea Arcangeli depends on ARCH_SUPPORTS_NUMA_BALANCING 620be3a7284SAndrea Arcangeli depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY 621be3a7284SAndrea Arcangeli depends on SMP && NUMA && MIGRATION 622be3a7284SAndrea Arcangeli help 623be3a7284SAndrea Arcangeli This option adds support for automatic NUMA aware memory/task placement. 624be3a7284SAndrea Arcangeli The mechanism is quite primitive and is based on migrating memory when 6256d56a410SPaul Gortmaker it has references to the node the task is running on. 626be3a7284SAndrea Arcangeli 627be3a7284SAndrea Arcangeli This system will be inactive on UMA systems. 628be3a7284SAndrea Arcangeli 6296f7c97e8SAneesh Kumar K.Vconfig NUMA_BALANCING_DEFAULT_ENABLED 6306f7c97e8SAneesh Kumar K.V bool "Automatically enable NUMA aware memory/task placement" 6316f7c97e8SAneesh Kumar K.V default y 6326f7c97e8SAneesh Kumar K.V depends on NUMA_BALANCING 6336f7c97e8SAneesh Kumar K.V help 6346f7c97e8SAneesh Kumar K.V If set, automatic NUMA balancing will be enabled if running on a NUMA 6356f7c97e8SAneesh Kumar K.V machine. 6366f7c97e8SAneesh Kumar K.V 63723964d2dSLi Zefanmenuconfig CGROUPS 6386341e62bSChristoph Jaeger bool "Control Group support" 6392bd59d48STejun Heo select KERNFS 640ddbcc7e8SPaul Menage help 64123964d2dSLi Zefan This option adds support for grouping sets of processes together, for 6425cdc38f9SKAMEZAWA Hiroyuki use with process control subsystems such as Cpusets, CFS, memory 6435cdc38f9SKAMEZAWA Hiroyuki controls or device isolation. 6445cdc38f9SKAMEZAWA Hiroyuki See 6455cdc38f9SKAMEZAWA Hiroyuki - Documentation/scheduler/sched-design-CFS.txt (CFS) 6469991a9c8Sseokhoon.yoon - Documentation/cgroup-v1/ (features for grouping, isolation 64745ce80fbSLi Zefan and resource control) 648ddbcc7e8SPaul Menage 649ddbcc7e8SPaul Menage Say N if unsure. 650ddbcc7e8SPaul Menage 65123964d2dSLi Zefanif CGROUPS 65223964d2dSLi Zefan 6533e32cb2eSJohannes Weinerconfig PAGE_COUNTER 6543e32cb2eSJohannes Weiner bool 6553e32cb2eSJohannes Weiner 656c255a458SAndrew Mortonconfig MEMCG 657a0166ec4SJohannes Weiner bool "Memory controller" 6583e32cb2eSJohannes Weiner select PAGE_COUNTER 65979bd9814STejun Heo select EVENTFD 66000f0b825SBalbir Singh help 661a0166ec4SJohannes Weiner Provides control over the memory footprint of tasks in a cgroup. 66200f0b825SBalbir Singh 663c255a458SAndrew Mortonconfig MEMCG_SWAP 664a0166ec4SJohannes Weiner bool "Swap controller" 665c255a458SAndrew Morton depends on MEMCG && SWAP 666c077719bSKAMEZAWA Hiroyuki help 667a0166ec4SJohannes Weiner Provides control over the swap space consumed by tasks in a cgroup. 668a0166ec4SJohannes Weiner 669c255a458SAndrew Mortonconfig MEMCG_SWAP_ENABLED 670a0166ec4SJohannes Weiner bool "Swap controller enabled by default" 671c255a458SAndrew Morton depends on MEMCG_SWAP 672a42c390cSMichal Hocko default y 673a42c390cSMichal Hocko help 674a42c390cSMichal Hocko Memory Resource Controller Swap Extension comes with its price in 675a42c390cSMichal Hocko a bigger memory consumption. General purpose distribution kernels 67643d547f9SJim Cromie which want to enable the feature but keep it disabled by default 67707555ac1SMichal Hocko and let the user enable it by swapaccount=1 boot command line 678a42c390cSMichal Hocko parameter should have this option unselected. 679a42c390cSMichal Hocko For those who want to have the feature enabled by default should 680a42c390cSMichal Hocko select this option (if, for some reason, they need to disable it 68100a66d29SWANG Cong then swapaccount=0 does the trick). 682c077719bSKAMEZAWA Hiroyuki 6836bf024e6SJohannes Weinerconfig BLK_CGROUP 6846bf024e6SJohannes Weiner bool "IO controller" 6856bf024e6SJohannes Weiner depends on BLOCK 6862bc64a20SAneesh Kumar K.V default n 6876bf024e6SJohannes Weiner ---help--- 6886bf024e6SJohannes Weiner Generic block IO controller cgroup interface. This is the common 6896bf024e6SJohannes Weiner cgroup interface which should be used by various IO controlling 6906bf024e6SJohannes Weiner policies. 6912bc64a20SAneesh Kumar K.V 6926bf024e6SJohannes Weiner Currently, CFQ IO scheduler uses it to recognize task groups and 6936bf024e6SJohannes Weiner control disk bandwidth allocation (proportional time slice allocation) 6946bf024e6SJohannes Weiner to such task groups. It is also used by bio throttling logic in 6956bf024e6SJohannes Weiner block layer to implement upper limit in IO rates on a device. 696e5d1367fSStephane Eranian 6976bf024e6SJohannes Weiner This option only enables generic Block IO controller infrastructure. 6986bf024e6SJohannes Weiner One needs to also enable actual IO controlling logic/policy. For 6996bf024e6SJohannes Weiner enabling proportional weight division of disk bandwidth in CFQ, set 7006bf024e6SJohannes Weiner CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set 7016bf024e6SJohannes Weiner CONFIG_BLK_DEV_THROTTLING=y. 7026bf024e6SJohannes Weiner 7039991a9c8Sseokhoon.yoon See Documentation/cgroup-v1/blkio-controller.txt for more information. 7046bf024e6SJohannes Weiner 7056bf024e6SJohannes Weinerconfig DEBUG_BLK_CGROUP 7066bf024e6SJohannes Weiner bool "IO controller debugging" 7076bf024e6SJohannes Weiner depends on BLK_CGROUP 7086bf024e6SJohannes Weiner default n 7096bf024e6SJohannes Weiner ---help--- 7106bf024e6SJohannes Weiner Enable some debugging help. Currently it exports additional stat 7116bf024e6SJohannes Weiner files in a cgroup which can be useful for debugging. 7126bf024e6SJohannes Weiner 7136bf024e6SJohannes Weinerconfig CGROUP_WRITEBACK 7146bf024e6SJohannes Weiner bool 7156bf024e6SJohannes Weiner depends on MEMCG && BLK_CGROUP 7166bf024e6SJohannes Weiner default y 717e5d1367fSStephane Eranian 7187c941438SDhaval Gianimenuconfig CGROUP_SCHED 719a0166ec4SJohannes Weiner bool "CPU controller" 7207c941438SDhaval Giani default n 7217c941438SDhaval Giani help 7227c941438SDhaval Giani This feature lets CPU scheduler recognize task groups and control CPU 7237c941438SDhaval Giani bandwidth allocation to such task groups. It uses cgroups to group 7247c941438SDhaval Giani tasks. 7257c941438SDhaval Giani 7267c941438SDhaval Gianiif CGROUP_SCHED 7277c941438SDhaval Gianiconfig FAIR_GROUP_SCHED 7287c941438SDhaval Giani bool "Group scheduling for SCHED_OTHER" 7297c941438SDhaval Giani depends on CGROUP_SCHED 7307c941438SDhaval Giani default CGROUP_SCHED 7317c941438SDhaval Giani 732ab84d31eSPaul Turnerconfig CFS_BANDWIDTH 733ab84d31eSPaul Turner bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED" 734ab84d31eSPaul Turner depends on FAIR_GROUP_SCHED 735ab84d31eSPaul Turner default n 736ab84d31eSPaul Turner help 737ab84d31eSPaul Turner This option allows users to define CPU bandwidth rates (limits) for 738ab84d31eSPaul Turner tasks running within the fair group scheduler. Groups with no limit 739ab84d31eSPaul Turner set are considered to be unconstrained and will run with no 740ab84d31eSPaul Turner restriction. 741ab84d31eSPaul Turner See tip/Documentation/scheduler/sched-bwc.txt for more information. 742ab84d31eSPaul Turner 7437c941438SDhaval Gianiconfig RT_GROUP_SCHED 7447c941438SDhaval Giani bool "Group scheduling for SCHED_RR/FIFO" 7457c941438SDhaval Giani depends on CGROUP_SCHED 7467c941438SDhaval Giani default n 7477c941438SDhaval Giani help 7487c941438SDhaval Giani This feature lets you explicitly allocate real CPU bandwidth 74932bd7eb5SLi Zefan to task groups. If enabled, it will also make it impossible to 7507c941438SDhaval Giani schedule realtime tasks for non-root users until you allocate 7517c941438SDhaval Giani realtime bandwidth for them. 7527c941438SDhaval Giani See Documentation/scheduler/sched-rt-group.txt for more information. 7537c941438SDhaval Giani 7547c941438SDhaval Gianiendif #CGROUP_SCHED 7557c941438SDhaval Giani 7566bf024e6SJohannes Weinerconfig CGROUP_PIDS 7576bf024e6SJohannes Weiner bool "PIDs controller" 7586bf024e6SJohannes Weiner help 7596bf024e6SJohannes Weiner Provides enforcement of process number limits in the scope of a 7606bf024e6SJohannes Weiner cgroup. Any attempt to fork more processes than is allowed in the 7616bf024e6SJohannes Weiner cgroup will fail. PIDs are fundamentally a global resource because it 7626bf024e6SJohannes Weiner is fairly trivial to reach PID exhaustion before you reach even a 7636bf024e6SJohannes Weiner conservative kmemcg limit. As a result, it is possible to grind a 7646bf024e6SJohannes Weiner system to halt without being limited by other cgroup policies. The 7656cc578dfSParav Pandit PIDs controller is designed to stop this from happening. 7666bf024e6SJohannes Weiner 7676bf024e6SJohannes Weiner It should be noted that organisational operations (such as attaching 7686cc578dfSParav Pandit to a cgroup hierarchy will *not* be blocked by the PIDs controller), 7696bf024e6SJohannes Weiner since the PIDs limit only affects a process's ability to fork, not to 7706bf024e6SJohannes Weiner attach to a cgroup. 7716bf024e6SJohannes Weiner 77239d3e758SParav Panditconfig CGROUP_RDMA 77339d3e758SParav Pandit bool "RDMA controller" 77439d3e758SParav Pandit help 77539d3e758SParav Pandit Provides enforcement of RDMA resources defined by IB stack. 77639d3e758SParav Pandit It is fairly easy for consumers to exhaust RDMA resources, which 77739d3e758SParav Pandit can result into resource unavailability to other consumers. 77839d3e758SParav Pandit RDMA controller is designed to stop this from happening. 77939d3e758SParav Pandit Attaching processes with active RDMA resources to the cgroup 78039d3e758SParav Pandit hierarchy is allowed even if can cross the hierarchy's limit. 78139d3e758SParav Pandit 7826bf024e6SJohannes Weinerconfig CGROUP_FREEZER 7836bf024e6SJohannes Weiner bool "Freezer controller" 7846bf024e6SJohannes Weiner help 7856bf024e6SJohannes Weiner Provides a way to freeze and unfreeze all tasks in a 7866bf024e6SJohannes Weiner cgroup. 7876bf024e6SJohannes Weiner 788489c2a20SJohannes Weiner This option affects the ORIGINAL cgroup interface. The cgroup2 memory 789489c2a20SJohannes Weiner controller includes important in-kernel memory consumers per default. 790489c2a20SJohannes Weiner 791489c2a20SJohannes Weiner If you're using cgroup2, say N. 792489c2a20SJohannes Weiner 7936bf024e6SJohannes Weinerconfig CGROUP_HUGETLB 7946bf024e6SJohannes Weiner bool "HugeTLB controller" 7956bf024e6SJohannes Weiner depends on HUGETLB_PAGE 7966bf024e6SJohannes Weiner select PAGE_COUNTER 797afc24d49SVivek Goyal default n 7986bf024e6SJohannes Weiner help 7996bf024e6SJohannes Weiner Provides a cgroup controller for HugeTLB pages. 8006bf024e6SJohannes Weiner When you enable this, you can put a per cgroup limit on HugeTLB usage. 8016bf024e6SJohannes Weiner The limit is enforced during page fault. Since HugeTLB doesn't 8026bf024e6SJohannes Weiner support page reclaim, enforcing the limit at page fault time implies 8036bf024e6SJohannes Weiner that, the application will get SIGBUS signal if it tries to access 8046bf024e6SJohannes Weiner HugeTLB pages beyond its limit. This requires the application to know 8056bf024e6SJohannes Weiner beforehand how much HugeTLB pages it would require for its use. The 8066bf024e6SJohannes Weiner control group is tracked in the third page lru pointer. This means 8076bf024e6SJohannes Weiner that we cannot use the controller with huge page less than 3 pages. 808afc24d49SVivek Goyal 8096bf024e6SJohannes Weinerconfig CPUSETS 8106bf024e6SJohannes Weiner bool "Cpuset controller" 811e1d4eeecSNicolas Pitre depends on SMP 8126bf024e6SJohannes Weiner help 8136bf024e6SJohannes Weiner This option will let you create and manage CPUSETs which 8146bf024e6SJohannes Weiner allow dynamically partitioning a system into sets of CPUs and 8156bf024e6SJohannes Weiner Memory Nodes and assigning tasks to run only within those sets. 8166bf024e6SJohannes Weiner This is primarily useful on large SMP or NUMA systems. 817afc24d49SVivek Goyal 8186bf024e6SJohannes Weiner Say N if unsure. 819afc24d49SVivek Goyal 8206bf024e6SJohannes Weinerconfig PROC_PID_CPUSET 8216bf024e6SJohannes Weiner bool "Include legacy /proc/<pid>/cpuset file" 8226bf024e6SJohannes Weiner depends on CPUSETS 82389e9b9e0STejun Heo default y 82489e9b9e0STejun Heo 8256bf024e6SJohannes Weinerconfig CGROUP_DEVICE 8266bf024e6SJohannes Weiner bool "Device controller" 8276bf024e6SJohannes Weiner help 8286bf024e6SJohannes Weiner Provides a cgroup controller implementing whitelists for 8296bf024e6SJohannes Weiner devices which a process in the cgroup can mknod or open. 8306bf024e6SJohannes Weiner 8316bf024e6SJohannes Weinerconfig CGROUP_CPUACCT 8326bf024e6SJohannes Weiner bool "Simple CPU accounting controller" 8336bf024e6SJohannes Weiner help 8346bf024e6SJohannes Weiner Provides a simple controller for monitoring the 8356bf024e6SJohannes Weiner total CPU consumed by the tasks in a cgroup. 8366bf024e6SJohannes Weiner 8376bf024e6SJohannes Weinerconfig CGROUP_PERF 8386bf024e6SJohannes Weiner bool "Perf controller" 8396bf024e6SJohannes Weiner depends on PERF_EVENTS 8406bf024e6SJohannes Weiner help 8416bf024e6SJohannes Weiner This option extends the perf per-cpu mode to restrict monitoring 8426bf024e6SJohannes Weiner to threads which belong to the cgroup specified and run on the 8436bf024e6SJohannes Weiner designated cpu. 8446bf024e6SJohannes Weiner 8456bf024e6SJohannes Weiner Say N if unsure. 8466bf024e6SJohannes Weiner 84730070984SDaniel Mackconfig CGROUP_BPF 84830070984SDaniel Mack bool "Support for eBPF programs attached to cgroups" 849483c4933SAndy Lutomirski depends on BPF_SYSCALL 850483c4933SAndy Lutomirski select SOCK_CGROUP_DATA 85130070984SDaniel Mack help 85230070984SDaniel Mack Allow attaching eBPF programs to a cgroup using the bpf(2) 85330070984SDaniel Mack syscall command BPF_PROG_ATTACH. 85430070984SDaniel Mack 85530070984SDaniel Mack In which context these programs are accessed depends on the type 85630070984SDaniel Mack of attachment. For instance, programs that are attached using 85730070984SDaniel Mack BPF_CGROUP_INET_INGRESS will be executed on the ingress path of 85830070984SDaniel Mack inet sockets. 85930070984SDaniel Mack 8606bf024e6SJohannes Weinerconfig CGROUP_DEBUG 86123b0be48SWaiman Long bool "Debug controller" 8626bf024e6SJohannes Weiner default n 86323b0be48SWaiman Long depends on DEBUG_KERNEL 8646bf024e6SJohannes Weiner help 8656bf024e6SJohannes Weiner This option enables a simple controller that exports 86623b0be48SWaiman Long debugging information about the cgroups framework. This 86723b0be48SWaiman Long controller is for control cgroup debugging only. Its 86823b0be48SWaiman Long interfaces are not stable. 8696bf024e6SJohannes Weiner 8706bf024e6SJohannes Weiner Say N. 8716bf024e6SJohannes Weiner 87273b35147SArnd Bergmannconfig SOCK_CGROUP_DATA 87373b35147SArnd Bergmann bool 87473b35147SArnd Bergmann default n 87573b35147SArnd Bergmann 87623964d2dSLi Zefanendif # CGROUPS 877c077719bSKAMEZAWA Hiroyuki 8788dd2a82cSDaniel Lezcanomenuconfig NAMESPACES 8796a108a14SDavid Rientjes bool "Namespaces support" if EXPERT 8802813893fSIulia Manda depends on MULTIUSER 8816a108a14SDavid Rientjes default !EXPERT 882c5289a69SPavel Emelyanov help 883c5289a69SPavel Emelyanov Provides the way to make tasks work with different objects using 884c5289a69SPavel Emelyanov the same id. For example same IPC id may refer to different objects 885c5289a69SPavel Emelyanov or same user id or pid may refer to different tasks when used in 886c5289a69SPavel Emelyanov different namespaces. 887c5289a69SPavel Emelyanov 8888dd2a82cSDaniel Lezcanoif NAMESPACES 8898dd2a82cSDaniel Lezcano 89058bfdd6dSPavel Emelyanovconfig UTS_NS 89158bfdd6dSPavel Emelyanov bool "UTS namespace" 89217a6d441SDaniel Lezcano default y 89358bfdd6dSPavel Emelyanov help 89458bfdd6dSPavel Emelyanov In this namespace tasks see different info provided with the 89558bfdd6dSPavel Emelyanov uname() system call 89658bfdd6dSPavel Emelyanov 897ae5e1b22SPavel Emelyanovconfig IPC_NS 898ae5e1b22SPavel Emelyanov bool "IPC namespace" 8998dd2a82cSDaniel Lezcano depends on (SYSVIPC || POSIX_MQUEUE) 90017a6d441SDaniel Lezcano default y 901ae5e1b22SPavel Emelyanov help 902ae5e1b22SPavel Emelyanov In this namespace tasks work with IPC ids which correspond to 903614b84cfSSerge E. Hallyn different IPC objects in different namespaces. 904ae5e1b22SPavel Emelyanov 905aee16ce7SPavel Emelyanovconfig USER_NS 90619c92399SKees Cook bool "User namespace" 9075673a94cSEric W. Biederman default n 908aee16ce7SPavel Emelyanov help 909aee16ce7SPavel Emelyanov This allows containers, i.e. vservers, to use user namespaces 910aee16ce7SPavel Emelyanov to provide different user info for different servers. 911e11f0ae3SEric W. Biederman 912e11f0ae3SEric W. Biederman When user namespaces are enabled in the kernel it is 913d886f4e4SJohannes Weiner recommended that the MEMCG option also be enabled and that 914d886f4e4SJohannes Weiner user-space use the memory control groups to limit the amount 915d886f4e4SJohannes Weiner of memory a memory unprivileged users can use. 916e11f0ae3SEric W. Biederman 917aee16ce7SPavel Emelyanov If unsure, say N. 918aee16ce7SPavel Emelyanov 91974bd59bbSPavel Emelyanovconfig PID_NS 9209bd38c2cSDaniel Lezcano bool "PID Namespaces" 92117a6d441SDaniel Lezcano default y 92274bd59bbSPavel Emelyanov help 92312d2b8f9SHeikki Orsila Support process id namespaces. This allows having multiple 924692105b8SMatt LaPlante processes with the same pid as long as they are in different 92574bd59bbSPavel Emelyanov pid namespaces. This is a building block of containers. 92674bd59bbSPavel Emelyanov 927d6eb633fSMatt Helsleyconfig NET_NS 928d6eb633fSMatt Helsley bool "Network namespace" 9298dd2a82cSDaniel Lezcano depends on NET 93017a6d441SDaniel Lezcano default y 931d6eb633fSMatt Helsley help 932d6eb633fSMatt Helsley Allow user space to create what appear to be multiple instances 933d6eb633fSMatt Helsley of the network stack. 934d6eb633fSMatt Helsley 9358dd2a82cSDaniel Lezcanoendif # NAMESPACES 9368dd2a82cSDaniel Lezcano 9375091faa4SMike Galbraithconfig SCHED_AUTOGROUP 9385091faa4SMike Galbraith bool "Automatic process group scheduling" 9395091faa4SMike Galbraith select CGROUPS 9405091faa4SMike Galbraith select CGROUP_SCHED 9415091faa4SMike Galbraith select FAIR_GROUP_SCHED 9425091faa4SMike Galbraith help 9435091faa4SMike Galbraith This option optimizes the scheduler for common desktop workloads by 9445091faa4SMike Galbraith automatically creating and populating task groups. This separation 9455091faa4SMike Galbraith of workloads isolates aggressive CPU burners (like build jobs) from 9465091faa4SMike Galbraith desktop applications. Task group autogeneration is currently based 9475091faa4SMike Galbraith upon task session. 9485091faa4SMike Galbraith 9497af37becSDaniel Lezcanoconfig SYSFS_DEPRECATED 9505d6a4ea5SFerenc Wagner bool "Enable deprecated sysfs features to support old userspace tools" 9517af37becSDaniel Lezcano depends on SYSFS 9527af37becSDaniel Lezcano default n 9537af37becSDaniel Lezcano help 9547af37becSDaniel Lezcano This option adds code that switches the layout of the "block" class 9557af37becSDaniel Lezcano devices, to not show up in /sys/class/block/, but only in 9567af37becSDaniel Lezcano /sys/block/. 9577af37becSDaniel Lezcano 9587af37becSDaniel Lezcano This switch is only active when the sysfs.deprecated=1 boot option is 9597af37becSDaniel Lezcano passed or the SYSFS_DEPRECATED_V2 option is set. 9607af37becSDaniel Lezcano 9617af37becSDaniel Lezcano This option allows new kernels to run on old distributions and tools, 9627af37becSDaniel Lezcano which might get confused by /sys/class/block/. Since 2007/2008 all 9637af37becSDaniel Lezcano major distributions and tools handle this just fine. 9647af37becSDaniel Lezcano 9657af37becSDaniel Lezcano Recent distributions and userspace tools after 2009/2010 depend on 9667af37becSDaniel Lezcano the existence of /sys/class/block/, and will not work with this 9677af37becSDaniel Lezcano option enabled. 9687af37becSDaniel Lezcano 9697af37becSDaniel Lezcano Only if you are using a new kernel on an old distribution, you might 9707af37becSDaniel Lezcano need to say Y here. 9717af37becSDaniel Lezcano 9727af37becSDaniel Lezcanoconfig SYSFS_DEPRECATED_V2 9735d6a4ea5SFerenc Wagner bool "Enable deprecated sysfs features by default" 9747af37becSDaniel Lezcano default n 9757af37becSDaniel Lezcano depends on SYSFS 9767af37becSDaniel Lezcano depends on SYSFS_DEPRECATED 9777af37becSDaniel Lezcano help 9787af37becSDaniel Lezcano Enable deprecated sysfs by default. 9797af37becSDaniel Lezcano 9807af37becSDaniel Lezcano See the CONFIG_SYSFS_DEPRECATED option for more details about this 9817af37becSDaniel Lezcano option. 9827af37becSDaniel Lezcano 9837af37becSDaniel Lezcano Only if you are using a new kernel on an old distribution, you might 9847af37becSDaniel Lezcano need to say Y here. Even then, odds are you would not need it 9857af37becSDaniel Lezcano enabled, you can always pass the boot option if absolutely necessary. 9867af37becSDaniel Lezcano 9877af37becSDaniel Lezcanoconfig RELAY 9887af37becSDaniel Lezcano bool "Kernel->user space relay support (formerly relayfs)" 98926b5679eSPeter Zijlstra select IRQ_WORK 9907af37becSDaniel Lezcano help 9917af37becSDaniel Lezcano This option enables support for relay interface support in 9927af37becSDaniel Lezcano certain file systems (such as debugfs). 9937af37becSDaniel Lezcano It is designed to provide an efficient mechanism for tools and 9947af37becSDaniel Lezcano facilities to relay large amounts of data from kernel space to 9957af37becSDaniel Lezcano user space. 9967af37becSDaniel Lezcano 9977af37becSDaniel Lezcano If unsure, say N. 9987af37becSDaniel Lezcano 999f991633dSDimitri Gorokhovikconfig BLK_DEV_INITRD 1000f991633dSDimitri Gorokhovik bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support" 1001f991633dSDimitri Gorokhovik help 1002f991633dSDimitri Gorokhovik The initial RAM filesystem is a ramfs which is loaded by the 1003f991633dSDimitri Gorokhovik boot loader (loadlin or lilo) and that is mounted as root 1004f991633dSDimitri Gorokhovik before the normal boot procedure. It is typically used to 1005f991633dSDimitri Gorokhovik load modules needed to mount the "real" root file system, 10068c27ceffSMauro Carvalho Chehab etc. See <file:Documentation/admin-guide/initrd.rst> for details. 1007f991633dSDimitri Gorokhovik 1008f991633dSDimitri Gorokhovik If RAM disk support (BLK_DEV_RAM) is also included, this 1009f991633dSDimitri Gorokhovik also enables initial RAM disk (initrd) support and adds 1010f991633dSDimitri Gorokhovik 15 Kbytes (more on some other architectures) to the kernel size. 1011f991633dSDimitri Gorokhovik 1012f991633dSDimitri Gorokhovik If unsure say Y. 1013f991633dSDimitri Gorokhovik 1014c33df4eaSJean-Paul Samanif BLK_DEV_INITRD 1015c33df4eaSJean-Paul Saman 1016dbec4866SSam Ravnborgsource "usr/Kconfig" 1017dbec4866SSam Ravnborg 1018c33df4eaSJean-Paul Samanendif 1019c33df4eaSJean-Paul Saman 1020877417e6SArnd Bergmannchoice 1021877417e6SArnd Bergmann prompt "Compiler optimization level" 10222cc3ce24SUlf Magnusson default CC_OPTIMIZE_FOR_PERFORMANCE 1023877417e6SArnd Bergmann 1024877417e6SArnd Bergmannconfig CC_OPTIMIZE_FOR_PERFORMANCE 1025877417e6SArnd Bergmann bool "Optimize for performance" 1026877417e6SArnd Bergmann help 1027877417e6SArnd Bergmann This is the default optimization level for the kernel, building 1028877417e6SArnd Bergmann with the "-O2" compiler flag for best performance and most 1029877417e6SArnd Bergmann helpful compile-time warnings. 1030877417e6SArnd Bergmann 1031c45b4f1fSLinus Torvaldsconfig CC_OPTIMIZE_FOR_SIZE 103296fffeb4SIngo Molnar bool "Optimize for size" 1033c45b4f1fSLinus Torvalds help 103431a4af7fSMasahiro Yamada Enabling this option will pass "-Os" instead of "-O2" to 103531a4af7fSMasahiro Yamada your compiler resulting in a smaller kernel. 1036c45b4f1fSLinus Torvalds 10373a55fb0dSKirill Smelkov If unsure, say N. 1038c45b4f1fSLinus Torvalds 1039877417e6SArnd Bergmannendchoice 1040877417e6SArnd Bergmann 1041*5d20ee31SNicholas Pigginconfig HAVE_LD_DEAD_CODE_DATA_ELIMINATION 1042*5d20ee31SNicholas Piggin bool 1043*5d20ee31SNicholas Piggin help 1044*5d20ee31SNicholas Piggin This requires that the arch annotates or otherwise protects 1045*5d20ee31SNicholas Piggin its external entry points from being discarded. Linker scripts 1046*5d20ee31SNicholas Piggin must also merge .text.*, .data.*, and .bss.* correctly into 1047*5d20ee31SNicholas Piggin output sections. Care must be taken not to pull in unrelated 1048*5d20ee31SNicholas Piggin sections (e.g., '.text.init'). Typically '.' in section names 1049*5d20ee31SNicholas Piggin is used to distinguish them from label names / C identifiers. 1050*5d20ee31SNicholas Piggin 1051*5d20ee31SNicholas Pigginconfig LD_DEAD_CODE_DATA_ELIMINATION 1052*5d20ee31SNicholas Piggin bool "Dead code and data elimination (EXPERIMENTAL)" 1053*5d20ee31SNicholas Piggin depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION 1054*5d20ee31SNicholas Piggin depends on EXPERT 1055*5d20ee31SNicholas Piggin help 1056*5d20ee31SNicholas Piggin Select this if the architecture wants to do dead code and 1057*5d20ee31SNicholas Piggin data elimination with the linker by compiling with 1058*5d20ee31SNicholas Piggin -ffunction-sections -fdata-sections, and linking with 1059*5d20ee31SNicholas Piggin --gc-sections. 1060*5d20ee31SNicholas Piggin 1061*5d20ee31SNicholas Piggin This can reduce on disk and in-memory size of the kernel 1062*5d20ee31SNicholas Piggin code and static data, particularly for small configs and 1063*5d20ee31SNicholas Piggin on small systems. This has the possibility of introducing 1064*5d20ee31SNicholas Piggin silently broken kernel if the required annotations are not 1065*5d20ee31SNicholas Piggin present. This option is not well tested yet, so use at your 1066*5d20ee31SNicholas Piggin own risk. 1067*5d20ee31SNicholas Piggin 10680847062aSRandy Dunlapconfig SYSCTL 10690847062aSRandy Dunlap bool 10700847062aSRandy Dunlap 1071b943c460SRandy Dunlapconfig ANON_INODES 1072b943c460SRandy Dunlap bool 1073b943c460SRandy Dunlap 1074657a5209SMike Frysingerconfig HAVE_UID16 1075657a5209SMike Frysinger bool 1076657a5209SMike Frysinger 1077657a5209SMike Frysingerconfig SYSCTL_EXCEPTION_TRACE 1078657a5209SMike Frysinger bool 1079657a5209SMike Frysinger help 1080657a5209SMike Frysinger Enable support for /proc/sys/debug/exception-trace. 1081657a5209SMike Frysinger 1082657a5209SMike Frysingerconfig SYSCTL_ARCH_UNALIGN_NO_WARN 1083657a5209SMike Frysinger bool 1084657a5209SMike Frysinger help 1085657a5209SMike Frysinger Enable support for /proc/sys/kernel/ignore-unaligned-usertrap 1086657a5209SMike Frysinger Allows arch to define/use @no_unaligned_warning to possibly warn 1087657a5209SMike Frysinger about unaligned access emulation going on under the hood. 1088657a5209SMike Frysinger 1089657a5209SMike Frysingerconfig SYSCTL_ARCH_UNALIGN_ALLOW 1090657a5209SMike Frysinger bool 1091657a5209SMike Frysinger help 1092657a5209SMike Frysinger Enable support for /proc/sys/kernel/unaligned-trap 1093657a5209SMike Frysinger Allows arches to define/use @unaligned_enabled to runtime toggle 1094657a5209SMike Frysinger the unaligned access emulation. 1095657a5209SMike Frysinger see arch/parisc/kernel/unaligned.c for reference 1096657a5209SMike Frysinger 1097657a5209SMike Frysingerconfig HAVE_PCSPKR_PLATFORM 1098657a5209SMike Frysinger bool 1099657a5209SMike Frysinger 1100f89b7755SAlexei Starovoitov# interpreter that classic socket filters depend on 1101f89b7755SAlexei Starovoitovconfig BPF 1102f89b7755SAlexei Starovoitov bool 1103f89b7755SAlexei Starovoitov 11046a108a14SDavid Rientjesmenuconfig EXPERT 11056a108a14SDavid Rientjes bool "Configure standard kernel features (expert users)" 1106f505c553SJosh Triplett # Unhide debug options, to make the on-by-default options visible 1107f505c553SJosh Triplett select DEBUG_KERNEL 11081da177e4SLinus Torvalds help 11091da177e4SLinus Torvalds This option allows certain base kernel options and settings 11101da177e4SLinus Torvalds to be disabled or tweaked. This is for specialized 11111da177e4SLinus Torvalds environments which can tolerate a "non-standard" kernel. 11121da177e4SLinus Torvalds Only use this if you really know what you are doing. 11131da177e4SLinus Torvalds 1114ae81f9e3SChuck Ebbertconfig UID16 11156a108a14SDavid Rientjes bool "Enable 16-bit UID system calls" if EXPERT 11162813893fSIulia Manda depends on HAVE_UID16 && MULTIUSER 1117ae81f9e3SChuck Ebbert default y 1118ae81f9e3SChuck Ebbert help 1119ae81f9e3SChuck Ebbert This enables the legacy 16-bit UID syscall wrappers. 1120ae81f9e3SChuck Ebbert 11212813893fSIulia Mandaconfig MULTIUSER 11222813893fSIulia Manda bool "Multiple users, groups and capabilities support" if EXPERT 11232813893fSIulia Manda default y 11242813893fSIulia Manda help 11252813893fSIulia Manda This option enables support for non-root users, groups and 11262813893fSIulia Manda capabilities. 11272813893fSIulia Manda 11282813893fSIulia Manda If you say N here, all processes will run with UID 0, GID 0, and all 11292813893fSIulia Manda possible capabilities. Saying N here also compiles out support for 11302813893fSIulia Manda system calls related to UIDs, GIDs, and capabilities, such as setuid, 11312813893fSIulia Manda setgid, and capset. 11322813893fSIulia Manda 11332813893fSIulia Manda If unsure, say Y here. 11342813893fSIulia Manda 1135f6187769SFabian Frederickconfig SGETMASK_SYSCALL 1136f6187769SFabian Frederick bool "sgetmask/ssetmask syscalls support" if EXPERT 1137a687a533SArnd Bergmann def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH 1138f6187769SFabian Frederick ---help--- 1139f6187769SFabian Frederick sys_sgetmask and sys_ssetmask are obsolete system calls 1140f6187769SFabian Frederick no longer supported in libc but still enabled by default in some 1141f6187769SFabian Frederick architectures. 1142f6187769SFabian Frederick 1143f6187769SFabian Frederick If unsure, leave the default option here. 1144f6187769SFabian Frederick 11456af9f7bfSFabian Frederickconfig SYSFS_SYSCALL 11466af9f7bfSFabian Frederick bool "Sysfs syscall support" if EXPERT 11476af9f7bfSFabian Frederick default y 11486af9f7bfSFabian Frederick ---help--- 11496af9f7bfSFabian Frederick sys_sysfs is an obsolete system call no longer supported in libc. 11506af9f7bfSFabian Frederick Note that disabling this option is more secure but might break 11516af9f7bfSFabian Frederick compatibility with some systems. 11526af9f7bfSFabian Frederick 11536af9f7bfSFabian Frederick If unsure say Y here. 11546af9f7bfSFabian Frederick 1155b89a8171SEric W. Biedermanconfig SYSCTL_SYSCALL 11566a108a14SDavid Rientjes bool "Sysctl syscall support" if EXPERT 115726a7034bSEric W. Biederman depends on PROC_SYSCTL 1158c736de60SWANG Cong default n 1159b89a8171SEric W. Biederman select SYSCTL 1160b89a8171SEric W. Biederman ---help--- 116113bb7e37SEric W. Biederman sys_sysctl uses binary paths that have been found challenging 116213bb7e37SEric W. Biederman to properly maintain and use. The interface in /proc/sys 116313bb7e37SEric W. Biederman using paths with ascii names is now the primary path to this 116413bb7e37SEric W. Biederman information. 1165b89a8171SEric W. Biederman 116613bb7e37SEric W. Biederman Almost nothing using the binary sysctl interface so if you are 116713bb7e37SEric W. Biederman trying to save some space it is probably safe to disable this, 116813bb7e37SEric W. Biederman making your kernel marginally smaller. 1169b89a8171SEric W. Biederman 1170c736de60SWANG Cong If unsure say N here. 1171ae81f9e3SChuck Ebbert 1172d1b069f5SRandy Dunlapconfig FHANDLE 1173d1b069f5SRandy Dunlap bool "open by fhandle syscalls" if EXPERT 1174d1b069f5SRandy Dunlap select EXPORTFS 1175d1b069f5SRandy Dunlap default y 1176d1b069f5SRandy Dunlap help 1177d1b069f5SRandy Dunlap If you say Y here, a user level program will be able to map 1178d1b069f5SRandy Dunlap file names to handle and then later use the handle for 1179d1b069f5SRandy Dunlap different file system operations. This is useful in implementing 1180d1b069f5SRandy Dunlap userspace file servers, which now track files using handles instead 1181d1b069f5SRandy Dunlap of names. The handle would remain the same even if file names 1182d1b069f5SRandy Dunlap get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2) 1183d1b069f5SRandy Dunlap syscalls. 1184d1b069f5SRandy Dunlap 1185baa73d9eSNicolas Pitreconfig POSIX_TIMERS 1186baa73d9eSNicolas Pitre bool "Posix Clocks & timers" if EXPERT 1187baa73d9eSNicolas Pitre default y 1188baa73d9eSNicolas Pitre help 1189baa73d9eSNicolas Pitre This includes native support for POSIX timers to the kernel. 1190baa73d9eSNicolas Pitre Some embedded systems have no use for them and therefore they 1191baa73d9eSNicolas Pitre can be configured out to reduce the size of the kernel image. 1192baa73d9eSNicolas Pitre 1193baa73d9eSNicolas Pitre When this option is disabled, the following syscalls won't be 1194baa73d9eSNicolas Pitre available: timer_create, timer_gettime: timer_getoverrun, 1195baa73d9eSNicolas Pitre timer_settime, timer_delete, clock_adjtime, getitimer, 1196baa73d9eSNicolas Pitre setitimer, alarm. Furthermore, the clock_settime, clock_gettime, 1197baa73d9eSNicolas Pitre clock_getres and clock_nanosleep syscalls will be limited to 1198baa73d9eSNicolas Pitre CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only. 1199baa73d9eSNicolas Pitre 1200baa73d9eSNicolas Pitre If unsure say y. 1201baa73d9eSNicolas Pitre 1202d59745ceSMatt Mackallconfig PRINTK 1203d59745ceSMatt Mackall default y 12046a108a14SDavid Rientjes bool "Enable support for printk" if EXPERT 120574876a98SFrederic Weisbecker select IRQ_WORK 1206d59745ceSMatt Mackall help 1207d59745ceSMatt Mackall This option enables normal printk support. Removing it 1208d59745ceSMatt Mackall eliminates most of the message strings from the kernel image 1209d59745ceSMatt Mackall and makes the kernel more or less silent. As this makes it 1210d59745ceSMatt Mackall very difficult to diagnose system problems, saying N here is 1211d59745ceSMatt Mackall strongly discouraged. 1212d59745ceSMatt Mackall 121342a0bb3fSPetr Mladekconfig PRINTK_NMI 121442a0bb3fSPetr Mladek def_bool y 121542a0bb3fSPetr Mladek depends on PRINTK 121642a0bb3fSPetr Mladek depends on HAVE_NMI 121742a0bb3fSPetr Mladek 1218c8538a7aSMatt Mackallconfig BUG 12196a108a14SDavid Rientjes bool "BUG() support" if EXPERT 1220c8538a7aSMatt Mackall default y 1221c8538a7aSMatt Mackall help 1222c8538a7aSMatt Mackall Disabling this option eliminates support for BUG and WARN, reducing 1223c8538a7aSMatt Mackall the size of your kernel image and potentially quietly ignoring 1224c8538a7aSMatt Mackall numerous fatal conditions. You should only consider disabling this 1225c8538a7aSMatt Mackall option for embedded systems with no facilities for reporting errors. 1226c8538a7aSMatt Mackall Just say Y. 1227c8538a7aSMatt Mackall 1228708e9a79SMatt Mackallconfig ELF_CORE 1229046d662fSAlex Kelly depends on COREDUMP 1230708e9a79SMatt Mackall default y 12316a108a14SDavid Rientjes bool "Enable ELF core dumps" if EXPERT 1232708e9a79SMatt Mackall help 1233708e9a79SMatt Mackall Enable support for generating core dumps. Disabling saves about 4k. 1234708e9a79SMatt Mackall 12358761f1abSRalf Baechle 1236e5e1d3cbSStas Sergeevconfig PCSPKR_PLATFORM 12376a108a14SDavid Rientjes bool "Enable PC-Speaker support" if EXPERT 12388761f1abSRalf Baechle depends on HAVE_PCSPKR_PLATFORM 123915f304b6SRalf Baechle select I8253_LOCK 1240e5e1d3cbSStas Sergeev default y 1241e5e1d3cbSStas Sergeev help 1242e5e1d3cbSStas Sergeev This option allows to disable the internal PC-Speaker 1243e5e1d3cbSStas Sergeev support, saving some memory. 1244e5e1d3cbSStas Sergeev 12451da177e4SLinus Torvaldsconfig BASE_FULL 12461da177e4SLinus Torvalds default y 12476a108a14SDavid Rientjes bool "Enable full-sized data structures for core" if EXPERT 12481da177e4SLinus Torvalds help 12491da177e4SLinus Torvalds Disabling this option reduces the size of miscellaneous core 12501da177e4SLinus Torvalds kernel data structures. This saves memory on small machines, 12511da177e4SLinus Torvalds but may reduce performance. 12521da177e4SLinus Torvalds 12531da177e4SLinus Torvaldsconfig FUTEX 12546a108a14SDavid Rientjes bool "Enable futex support" if EXPERT 12551da177e4SLinus Torvalds default y 1256bc2eecd7SNicolas Pitre imply RT_MUTEXES 12571da177e4SLinus Torvalds help 12581da177e4SLinus Torvalds Disabling this option will cause the kernel to be built without 12591da177e4SLinus Torvalds support for "fast userspace mutexes". The resulting kernel may not 12601da177e4SLinus Torvalds run glibc-based applications correctly. 12611da177e4SLinus Torvalds 1262bc2eecd7SNicolas Pitreconfig FUTEX_PI 1263bc2eecd7SNicolas Pitre bool 1264bc2eecd7SNicolas Pitre depends on FUTEX && RT_MUTEXES 1265bc2eecd7SNicolas Pitre default y 1266bc2eecd7SNicolas Pitre 126703b8c7b6SHeiko Carstensconfig HAVE_FUTEX_CMPXCHG 126803b8c7b6SHeiko Carstens bool 126962b4d204SJosh Triplett depends on FUTEX 127003b8c7b6SHeiko Carstens help 127103b8c7b6SHeiko Carstens Architectures should select this if futex_atomic_cmpxchg_inatomic() 127203b8c7b6SHeiko Carstens is implemented and always working. This removes a couple of runtime 127303b8c7b6SHeiko Carstens checks. 127403b8c7b6SHeiko Carstens 12751da177e4SLinus Torvaldsconfig EPOLL 12766a108a14SDavid Rientjes bool "Enable eventpoll support" if EXPERT 12771da177e4SLinus Torvalds default y 1278448e3ceeSAdrian Bunk select ANON_INODES 12791da177e4SLinus Torvalds help 12801da177e4SLinus Torvalds Disabling this option will cause the kernel to be built without 12811da177e4SLinus Torvalds support for epoll family of system calls. 12821da177e4SLinus Torvalds 1283fba2afaaSDavide Libenziconfig SIGNALFD 12846a108a14SDavid Rientjes bool "Enable signalfd() system call" if EXPERT 1285448e3ceeSAdrian Bunk select ANON_INODES 1286fba2afaaSDavide Libenzi default y 1287fba2afaaSDavide Libenzi help 1288fba2afaaSDavide Libenzi Enable the signalfd() system call that allows to receive signals 1289fba2afaaSDavide Libenzi on a file descriptor. 1290fba2afaaSDavide Libenzi 1291fba2afaaSDavide Libenzi If unsure, say Y. 1292fba2afaaSDavide Libenzi 1293b215e283SDavide Libenziconfig TIMERFD 12946a108a14SDavid Rientjes bool "Enable timerfd() system call" if EXPERT 1295448e3ceeSAdrian Bunk select ANON_INODES 1296b215e283SDavide Libenzi default y 1297b215e283SDavide Libenzi help 1298b215e283SDavide Libenzi Enable the timerfd() system call that allows to receive timer 1299b215e283SDavide Libenzi events on a file descriptor. 1300b215e283SDavide Libenzi 1301b215e283SDavide Libenzi If unsure, say Y. 1302b215e283SDavide Libenzi 1303e1ad7468SDavide Libenziconfig EVENTFD 13046a108a14SDavid Rientjes bool "Enable eventfd() system call" if EXPERT 1305448e3ceeSAdrian Bunk select ANON_INODES 1306e1ad7468SDavide Libenzi default y 1307e1ad7468SDavide Libenzi help 1308e1ad7468SDavide Libenzi Enable the eventfd() system call that allows to receive both 1309e1ad7468SDavide Libenzi kernel notification (ie. KAIO) or userspace notifications. 1310e1ad7468SDavide Libenzi 1311e1ad7468SDavide Libenzi If unsure, say Y. 1312e1ad7468SDavide Libenzi 13131da177e4SLinus Torvaldsconfig SHMEM 13146a108a14SDavid Rientjes bool "Use full shmem filesystem" if EXPERT 13151da177e4SLinus Torvalds default y 13161da177e4SLinus Torvalds depends on MMU 13171da177e4SLinus Torvalds help 13181da177e4SLinus Torvalds The shmem is an internal filesystem used to manage shared memory. 13191da177e4SLinus Torvalds It is backed by swap and manages resource limits. It is also exported 13201da177e4SLinus Torvalds to userspace as tmpfs if TMPFS is enabled. Disabling this 13211da177e4SLinus Torvalds option replaces shmem and tmpfs with the much simpler ramfs code, 13221da177e4SLinus Torvalds which may be appropriate on small systems without swap. 13231da177e4SLinus Torvalds 1324ebf3f09cSThomas Petazzoniconfig AIO 13256a108a14SDavid Rientjes bool "Enable AIO support" if EXPERT 1326ebf3f09cSThomas Petazzoni default y 1327ebf3f09cSThomas Petazzoni help 1328ebf3f09cSThomas Petazzoni This option enables POSIX asynchronous I/O which may by used 1329ebf3f09cSThomas Petazzoni by some high performance threaded applications. Disabling 1330ebf3f09cSThomas Petazzoni this option saves about 7k. 1331ebf3f09cSThomas Petazzoni 1332d3ac21caSJosh Triplettconfig ADVISE_SYSCALLS 1333d3ac21caSJosh Triplett bool "Enable madvise/fadvise syscalls" if EXPERT 1334d3ac21caSJosh Triplett default y 1335d3ac21caSJosh Triplett help 1336d3ac21caSJosh Triplett This option enables the madvise and fadvise syscalls, used by 1337d3ac21caSJosh Triplett applications to advise the kernel about their future memory or file 1338d3ac21caSJosh Triplett usage, improving performance. If building an embedded system where no 1339d3ac21caSJosh Triplett applications use these syscalls, you can disable this option to save 1340d3ac21caSJosh Triplett space. 1341d3ac21caSJosh Triplett 13425b25b13aSMathieu Desnoyersconfig MEMBARRIER 13435b25b13aSMathieu Desnoyers bool "Enable membarrier() system call" if EXPERT 13445b25b13aSMathieu Desnoyers default y 13455b25b13aSMathieu Desnoyers help 13465b25b13aSMathieu Desnoyers Enable the membarrier() system call that allows issuing memory 13475b25b13aSMathieu Desnoyers barriers across all running threads, which can be used to distribute 13485b25b13aSMathieu Desnoyers the cost of user-space memory barriers asymmetrically by transforming 13495b25b13aSMathieu Desnoyers pairs of memory barriers into pairs consisting of membarrier() and a 13505b25b13aSMathieu Desnoyers compiler barrier. 13515b25b13aSMathieu Desnoyers 13525b25b13aSMathieu Desnoyers If unsure, say Y. 13535b25b13aSMathieu Desnoyers 1354d1b069f5SRandy Dunlapconfig CHECKPOINT_RESTORE 1355d1b069f5SRandy Dunlap bool "Checkpoint/restore support" if EXPERT 1356d1b069f5SRandy Dunlap select PROC_CHILDREN 1357d1b069f5SRandy Dunlap default n 1358d1b069f5SRandy Dunlap help 1359d1b069f5SRandy Dunlap Enables additional kernel features in a sake of checkpoint/restore. 1360d1b069f5SRandy Dunlap In particular it adds auxiliary prctl codes to setup process text, 1361d1b069f5SRandy Dunlap data and heap segment sizes, and a few additional /proc filesystem 1362d1b069f5SRandy Dunlap entries. 1363d1b069f5SRandy Dunlap 1364d1b069f5SRandy Dunlap If unsure, say N here. 1365d1b069f5SRandy Dunlap 1366d1b069f5SRandy Dunlapconfig KALLSYMS 1367d1b069f5SRandy Dunlap bool "Load all symbols for debugging/ksymoops" if EXPERT 1368d1b069f5SRandy Dunlap default y 1369d1b069f5SRandy Dunlap help 1370d1b069f5SRandy Dunlap Say Y here to let the kernel print out symbolic crash information and 1371d1b069f5SRandy Dunlap symbolic stack backtraces. This increases the size of the kernel 1372d1b069f5SRandy Dunlap somewhat, as all symbols have to be loaded into the kernel image. 1373d1b069f5SRandy Dunlap 1374d1b069f5SRandy Dunlapconfig KALLSYMS_ALL 1375d1b069f5SRandy Dunlap bool "Include all symbols in kallsyms" 1376d1b069f5SRandy Dunlap depends on DEBUG_KERNEL && KALLSYMS 1377d1b069f5SRandy Dunlap help 1378d1b069f5SRandy Dunlap Normally kallsyms only contains the symbols of functions for nicer 1379d1b069f5SRandy Dunlap OOPS messages and backtraces (i.e., symbols from the text and inittext 1380d1b069f5SRandy Dunlap sections). This is sufficient for most cases. And only in very rare 1381d1b069f5SRandy Dunlap cases (e.g., when a debugger is used) all symbols are required (e.g., 1382d1b069f5SRandy Dunlap names of variables from the data sections, etc). 1383d1b069f5SRandy Dunlap 1384d1b069f5SRandy Dunlap This option makes sure that all symbols are loaded into the kernel 1385d1b069f5SRandy Dunlap image (i.e., symbols from all sections) in cost of increased kernel 1386d1b069f5SRandy Dunlap size (depending on the kernel configuration, it may be 300KiB or 1387d1b069f5SRandy Dunlap something like this). 1388d1b069f5SRandy Dunlap 1389d1b069f5SRandy Dunlap Say N unless you really need all symbols. 1390d1b069f5SRandy Dunlap 1391d1b069f5SRandy Dunlapconfig KALLSYMS_ABSOLUTE_PERCPU 1392d1b069f5SRandy Dunlap bool 1393d1b069f5SRandy Dunlap depends on KALLSYMS 1394d1b069f5SRandy Dunlap default X86_64 && SMP 1395d1b069f5SRandy Dunlap 1396d1b069f5SRandy Dunlapconfig KALLSYMS_BASE_RELATIVE 1397d1b069f5SRandy Dunlap bool 1398d1b069f5SRandy Dunlap depends on KALLSYMS 1399a687a533SArnd Bergmann default !IA64 1400d1b069f5SRandy Dunlap help 1401d1b069f5SRandy Dunlap Instead of emitting them as absolute values in the native word size, 1402d1b069f5SRandy Dunlap emit the symbol references in the kallsyms table as 32-bit entries, 1403d1b069f5SRandy Dunlap each containing a relative value in the range [base, base + U32_MAX] 1404d1b069f5SRandy Dunlap or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either 1405d1b069f5SRandy Dunlap an absolute value in the range [0, S32_MAX] or a relative value in the 1406d1b069f5SRandy Dunlap range [base, base + S32_MAX], where base is the lowest relative symbol 1407d1b069f5SRandy Dunlap address encountered in the image. 1408d1b069f5SRandy Dunlap 1409d1b069f5SRandy Dunlap On 64-bit builds, this reduces the size of the address table by 50%, 1410d1b069f5SRandy Dunlap but more importantly, it results in entries whose values are build 1411d1b069f5SRandy Dunlap time constants, and no relocation pass is required at runtime to fix 1412d1b069f5SRandy Dunlap up the entries based on the runtime load address of the kernel. 1413d1b069f5SRandy Dunlap 1414d1b069f5SRandy Dunlap# end of the "standard kernel features (expert users)" menu 1415d1b069f5SRandy Dunlap 1416d1b069f5SRandy Dunlap# syscall, maps, verifier 1417d1b069f5SRandy Dunlapconfig BPF_SYSCALL 1418d1b069f5SRandy Dunlap bool "Enable bpf() system call" 1419d1b069f5SRandy Dunlap select ANON_INODES 1420d1b069f5SRandy Dunlap select BPF 1421d1b069f5SRandy Dunlap default n 1422d1b069f5SRandy Dunlap help 1423d1b069f5SRandy Dunlap Enable the bpf() system call that allows to manipulate eBPF 1424d1b069f5SRandy Dunlap programs and maps via file descriptors. 1425d1b069f5SRandy Dunlap 1426290af866SAlexei Starovoitovconfig BPF_JIT_ALWAYS_ON 1427290af866SAlexei Starovoitov bool "Permanently enable BPF JIT and remove BPF interpreter" 1428290af866SAlexei Starovoitov depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT 1429290af866SAlexei Starovoitov help 1430290af866SAlexei Starovoitov Enables BPF JIT and removes BPF interpreter to avoid 1431290af866SAlexei Starovoitov speculative execution of BPF instructions by the interpreter 1432290af866SAlexei Starovoitov 1433d1b069f5SRandy Dunlapconfig USERFAULTFD 1434d1b069f5SRandy Dunlap bool "Enable userfaultfd() system call" 1435d1b069f5SRandy Dunlap select ANON_INODES 1436d1b069f5SRandy Dunlap depends on MMU 1437d1b069f5SRandy Dunlap help 1438d1b069f5SRandy Dunlap Enable the userfaultfd() system call that allows to intercept and 1439d1b069f5SRandy Dunlap handle page faults in userland. 1440d1b069f5SRandy Dunlap 14413ccfebedSMathieu Desnoyersconfig ARCH_HAS_MEMBARRIER_CALLBACKS 14423ccfebedSMathieu Desnoyers bool 14433ccfebedSMathieu Desnoyers 144470216e18SMathieu Desnoyersconfig ARCH_HAS_MEMBARRIER_SYNC_CORE 144570216e18SMathieu Desnoyers bool 144670216e18SMathieu Desnoyers 14476befe5f6SRandy Dunlapconfig EMBEDDED 14486befe5f6SRandy Dunlap bool "Embedded system" 14495d2acfc7SJosh Triplett option allnoconfig_y 14506befe5f6SRandy Dunlap select EXPERT 14516befe5f6SRandy Dunlap help 14526befe5f6SRandy Dunlap This option should be enabled if compiling the kernel for 14536befe5f6SRandy Dunlap an embedded system so certain expert options are available 14546befe5f6SRandy Dunlap for configuration. 14556befe5f6SRandy Dunlap 1456cdd6c482SIngo Molnarconfig HAVE_PERF_EVENTS 14570793a61dSThomas Gleixner bool 1458018df72dSMike Frysinger help 1459018df72dSMike Frysinger See tools/perf/design.txt for details. 14600793a61dSThomas Gleixner 1461906010b2SPeter Zijlstraconfig PERF_USE_VMALLOC 1462906010b2SPeter Zijlstra bool 1463906010b2SPeter Zijlstra help 1464906010b2SPeter Zijlstra See tools/perf/design.txt for details 1465906010b2SPeter Zijlstra 1466ad90a3deSWilliam Breathitt Grayconfig PC104 1467424529fbSWilliam Breathitt Gray bool "PC/104 support" if EXPERT 1468ad90a3deSWilliam Breathitt Gray help 1469ad90a3deSWilliam Breathitt Gray Expose PC/104 form factor device drivers and options available for 1470ad90a3deSWilliam Breathitt Gray selection and configuration. Enable this option if your target 1471ad90a3deSWilliam Breathitt Gray machine has a PC/104 bus. 1472ad90a3deSWilliam Breathitt Gray 147357c0c15bSIngo Molnarmenu "Kernel Performance Events And Counters" 14740793a61dSThomas Gleixner 1475cdd6c482SIngo Molnarconfig PERF_EVENTS 147657c0c15bSIngo Molnar bool "Kernel performance events and counters" 1477392d65a9SRobert Richter default y if PROFILING 1478cdd6c482SIngo Molnar depends on HAVE_PERF_EVENTS 14794c59e467SIngo Molnar select ANON_INODES 1480e360adbeSPeter Zijlstra select IRQ_WORK 148183fe27eaSPranith Kumar select SRCU 14820793a61dSThomas Gleixner help 148357c0c15bSIngo Molnar Enable kernel support for various performance events provided 148457c0c15bSIngo Molnar by software and hardware. 14850793a61dSThomas Gleixner 1486dd77038dSThadeu Lima de Souza Cascardo Software events are supported either built-in or via the 148757c0c15bSIngo Molnar use of generic tracepoints. 148857c0c15bSIngo Molnar 148957c0c15bSIngo Molnar Most modern CPUs support performance events via performance 149057c0c15bSIngo Molnar counter registers. These registers count the number of certain 14910793a61dSThomas Gleixner types of hw events: such as instructions executed, cachemisses 14920793a61dSThomas Gleixner suffered, or branches mis-predicted - without slowing down the 14930793a61dSThomas Gleixner kernel or applications. These registers can also trigger interrupts 14940793a61dSThomas Gleixner when a threshold number of events have passed - and can thus be 14950793a61dSThomas Gleixner used to profile the code that runs on that CPU. 14960793a61dSThomas Gleixner 149757c0c15bSIngo Molnar The Linux Performance Event subsystem provides an abstraction of 1498dd77038dSThadeu Lima de Souza Cascardo these software and hardware event capabilities, available via a 149957c0c15bSIngo Molnar system call and used by the "perf" utility in tools/perf/. It 15000793a61dSThomas Gleixner provides per task and per CPU counters, and it provides event 15010793a61dSThomas Gleixner capabilities on top of those. 15020793a61dSThomas Gleixner 15030793a61dSThomas Gleixner Say Y if unsure. 15040793a61dSThomas Gleixner 1505906010b2SPeter Zijlstraconfig DEBUG_PERF_USE_VMALLOC 1506906010b2SPeter Zijlstra default n 1507906010b2SPeter Zijlstra bool "Debug: use vmalloc to back perf mmap() buffers" 1508cb307113SMichael Ellerman depends on PERF_EVENTS && DEBUG_KERNEL && !PPC 1509906010b2SPeter Zijlstra select PERF_USE_VMALLOC 1510906010b2SPeter Zijlstra help 1511906010b2SPeter Zijlstra Use vmalloc memory to back perf mmap() buffers. 1512906010b2SPeter Zijlstra 1513906010b2SPeter Zijlstra Mostly useful for debugging the vmalloc code on platforms 1514906010b2SPeter Zijlstra that don't require it. 1515906010b2SPeter Zijlstra 1516906010b2SPeter Zijlstra Say N if unsure. 1517906010b2SPeter Zijlstra 15180793a61dSThomas Gleixnerendmenu 15190793a61dSThomas Gleixner 1520f8891e5eSChristoph Lameterconfig VM_EVENT_COUNTERS 1521f8891e5eSChristoph Lameter default y 15226a108a14SDavid Rientjes bool "Enable VM event counters for /proc/vmstat" if EXPERT 1523f8891e5eSChristoph Lameter help 15242aea4fb6SPaul Jackson VM event counters are needed for event counts to be shown. 15252aea4fb6SPaul Jackson This option allows the disabling of the VM event counters 15266a108a14SDavid Rientjes on EXPERT systems. /proc/vmstat will only show page counts 15272aea4fb6SPaul Jackson if VM event counters are disabled. 1528f8891e5eSChristoph Lameter 152941ecc55bSChristoph Lameterconfig SLUB_DEBUG 153041ecc55bSChristoph Lameter default y 15316a108a14SDavid Rientjes bool "Enable SLUB debugging support" if EXPERT 1532f6acb635SChristoph Lameter depends on SLUB && SYSFS 153341ecc55bSChristoph Lameter help 153441ecc55bSChristoph Lameter SLUB has extensive debug support features. Disabling these can 153541ecc55bSChristoph Lameter result in significant savings in code size. This also disables 153641ecc55bSChristoph Lameter SLUB sysfs support. /sys/slab will not exist and there will be 153741ecc55bSChristoph Lameter no support for cache validation etc. 153841ecc55bSChristoph Lameter 15391663f26dSTejun Heoconfig SLUB_MEMCG_SYSFS_ON 15401663f26dSTejun Heo default n 15411663f26dSTejun Heo bool "Enable memcg SLUB sysfs support by default" if EXPERT 15421663f26dSTejun Heo depends on SLUB && SYSFS && MEMCG 15431663f26dSTejun Heo help 15441663f26dSTejun Heo SLUB creates a directory under /sys/kernel/slab for each 15451663f26dSTejun Heo allocation cache to host info and debug files. If memory 15461663f26dSTejun Heo cgroup is enabled, each cache can have per memory cgroup 15471663f26dSTejun Heo caches. SLUB can create the same sysfs directories for these 15481663f26dSTejun Heo caches under /sys/kernel/slab/CACHE/cgroup but it can lead 15491663f26dSTejun Heo to a very high number of debug files being created. This is 15501663f26dSTejun Heo controlled by slub_memcg_sysfs boot parameter and this 15511663f26dSTejun Heo config option determines the parameter's default value. 15521663f26dSTejun Heo 1553b943c460SRandy Dunlapconfig COMPAT_BRK 1554b943c460SRandy Dunlap bool "Disable heap randomization" 1555b943c460SRandy Dunlap default y 1556b943c460SRandy Dunlap help 1557b943c460SRandy Dunlap Randomizing heap placement makes heap exploits harder, but it 1558b943c460SRandy Dunlap also breaks ancient binaries (including anything libc5 based). 1559b943c460SRandy Dunlap This option changes the bootup default to heap randomization 1560692105b8SMatt LaPlante disabled, and can be overridden at runtime by setting 1561b943c460SRandy Dunlap /proc/sys/kernel/randomize_va_space to 2. 1562b943c460SRandy Dunlap 1563b943c460SRandy Dunlap On non-ancient distros (post-2000 ones) N is usually a safe choice. 1564b943c460SRandy Dunlap 156581819f0fSChristoph Lameterchoice 156681819f0fSChristoph Lameter prompt "Choose SLAB allocator" 1567a0acd820SChristoph Lameter default SLUB 156881819f0fSChristoph Lameter help 156981819f0fSChristoph Lameter This option allows to select a slab allocator. 157081819f0fSChristoph Lameter 157181819f0fSChristoph Lameterconfig SLAB 157281819f0fSChristoph Lameter bool "SLAB" 157304385fc5SKees Cook select HAVE_HARDENED_USERCOPY_ALLOCATOR 157481819f0fSChristoph Lameter help 157581819f0fSChristoph Lameter The regular slab allocator that is established and known to work 157634013886SChristoph Lameter well in all environments. It organizes cache hot objects in 157702f56210SSimon Arlott per cpu and per node queues. 157881819f0fSChristoph Lameter 157981819f0fSChristoph Lameterconfig SLUB 158081819f0fSChristoph Lameter bool "SLUB (Unqueued Allocator)" 1581ed18adc1SKees Cook select HAVE_HARDENED_USERCOPY_ALLOCATOR 158281819f0fSChristoph Lameter help 158381819f0fSChristoph Lameter SLUB is a slab allocator that minimizes cache line usage 158481819f0fSChristoph Lameter instead of managing queues of cached objects (SLAB approach). 158581819f0fSChristoph Lameter Per cpu caching is realized using slabs of objects instead 158681819f0fSChristoph Lameter of queues of objects. SLUB can use memory efficiently 158702f56210SSimon Arlott and has enhanced diagnostics. SLUB is the default choice for 158802f56210SSimon Arlott a slab allocator. 158981819f0fSChristoph Lameter 159081819f0fSChristoph Lameterconfig SLOB 15916a108a14SDavid Rientjes depends on EXPERT 159281819f0fSChristoph Lameter bool "SLOB (Simple Allocator)" 159381819f0fSChristoph Lameter help 159437291458SMatt Mackall SLOB replaces the stock allocator with a drastically simpler 159537291458SMatt Mackall allocator. SLOB is generally more space efficient but 159637291458SMatt Mackall does not perform as well on large systems. 159781819f0fSChristoph Lameter 159881819f0fSChristoph Lameterendchoice 159981819f0fSChristoph Lameter 16007660a6fdSKees Cookconfig SLAB_MERGE_DEFAULT 16017660a6fdSKees Cook bool "Allow slab caches to be merged" 16027660a6fdSKees Cook default y 16037660a6fdSKees Cook help 16047660a6fdSKees Cook For reduced kernel memory fragmentation, slab caches can be 16057660a6fdSKees Cook merged when they share the same size and other characteristics. 16067660a6fdSKees Cook This carries a risk of kernel heap overflows being able to 16077660a6fdSKees Cook overwrite objects from merged caches (and more easily control 16087660a6fdSKees Cook cache layout), which makes such heap attacks easier to exploit 16097660a6fdSKees Cook by attackers. By keeping caches unmerged, these kinds of exploits 16107660a6fdSKees Cook can usually only damage objects in the same cache. To disable 16117660a6fdSKees Cook merging at runtime, "slab_nomerge" can be passed on the kernel 16127660a6fdSKees Cook command line. 16137660a6fdSKees Cook 1614c7ce4f60SThomas Garnierconfig SLAB_FREELIST_RANDOM 1615c7ce4f60SThomas Garnier default n 1616210e7a43SThomas Garnier depends on SLAB || SLUB 1617c7ce4f60SThomas Garnier bool "SLAB freelist randomization" 1618c7ce4f60SThomas Garnier help 1619210e7a43SThomas Garnier Randomizes the freelist order used on creating new pages. This 1620c7ce4f60SThomas Garnier security feature reduces the predictability of the kernel slab 1621c7ce4f60SThomas Garnier allocator against heap overflows. 1622c7ce4f60SThomas Garnier 16232482ddecSKees Cookconfig SLAB_FREELIST_HARDENED 16242482ddecSKees Cook bool "Harden slab freelist metadata" 16252482ddecSKees Cook depends on SLUB 16262482ddecSKees Cook help 16272482ddecSKees Cook Many kernel heap attacks try to target slab cache metadata and 16282482ddecSKees Cook other infrastructure. This options makes minor performance 16292482ddecSKees Cook sacrifies to harden the kernel slab allocator against common 16302482ddecSKees Cook freelist exploit methods. 16312482ddecSKees Cook 1632345c905dSJoonsoo Kimconfig SLUB_CPU_PARTIAL 1633345c905dSJoonsoo Kim default y 1634b39ffbf8SUwe Kleine-König depends on SLUB && SMP 1635345c905dSJoonsoo Kim bool "SLUB per cpu partial cache" 1636345c905dSJoonsoo Kim help 1637345c905dSJoonsoo Kim Per cpu partial caches accellerate objects allocation and freeing 1638345c905dSJoonsoo Kim that is local to a processor at the price of more indeterminism 1639345c905dSJoonsoo Kim in the latency of the free. On overflow these caches will be cleared 1640345c905dSJoonsoo Kim which requires the taking of locks that may cause latency spikes. 1641345c905dSJoonsoo Kim Typically one would choose no for a realtime system. 1642345c905dSJoonsoo Kim 1643ea637639SJie Zhangconfig MMAP_ALLOW_UNINITIALIZED 1644ea637639SJie Zhang bool "Allow mmapped anonymous memory to be uninitialized" 16456a108a14SDavid Rientjes depends on EXPERT && !MMU 1646ea637639SJie Zhang default n 1647ea637639SJie Zhang help 1648ea637639SJie Zhang Normally, and according to the Linux spec, anonymous memory obtained 1649ea637639SJie Zhang from mmap() has it's contents cleared before it is passed to 1650ea637639SJie Zhang userspace. Enabling this config option allows you to request that 1651ea637639SJie Zhang mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 1652ea637639SJie Zhang providing a huge performance boost. If this option is not enabled, 1653ea637639SJie Zhang then the flag will be ignored. 1654ea637639SJie Zhang 1655ea637639SJie Zhang This is taken advantage of by uClibc's malloc(), and also by 1656ea637639SJie Zhang ELF-FDPIC binfmt's brk and stack allocator. 1657ea637639SJie Zhang 1658ea637639SJie Zhang Because of the obvious security issues, this option should only be 1659ea637639SJie Zhang enabled on embedded devices where you control what is run in 1660ea637639SJie Zhang userspace. Since that isn't generally a problem on no-MMU systems, 1661ea637639SJie Zhang it is normally safe to say Y here. 1662ea637639SJie Zhang 1663ea637639SJie Zhang See Documentation/nommu-mmap.txt for more information. 1664ea637639SJie Zhang 1665091f6e26SDavid Howellsconfig SYSTEM_DATA_VERIFICATION 1666091f6e26SDavid Howells def_bool n 1667091f6e26SDavid Howells select SYSTEM_TRUSTED_KEYRING 1668091f6e26SDavid Howells select KEYS 1669091f6e26SDavid Howells select CRYPTO 1670d43de6c7SDavid Howells select CRYPTO_RSA 1671091f6e26SDavid Howells select ASYMMETRIC_KEY_TYPE 1672091f6e26SDavid Howells select ASYMMETRIC_PUBLIC_KEY_SUBTYPE 1673091f6e26SDavid Howells select ASN1 1674091f6e26SDavid Howells select OID_REGISTRY 1675091f6e26SDavid Howells select X509_CERTIFICATE_PARSER 1676091f6e26SDavid Howells select PKCS7_MESSAGE_PARSER 167782c04ff8SPeter Foley help 1678091f6e26SDavid Howells Provide PKCS#7 message verification using the contents of the system 1679091f6e26SDavid Howells trusted keyring to provide public keys. This then can be used for 1680091f6e26SDavid Howells module verification, kexec image verification and firmware blob 1681091f6e26SDavid Howells verification. 168282c04ff8SPeter Foley 1683125e5645SMathieu Desnoyersconfig PROFILING 1684b309a294SRobert Richter bool "Profiling support" 1685125e5645SMathieu Desnoyers help 1686125e5645SMathieu Desnoyers Say Y here to enable the extended profiling support mechanisms used 1687125e5645SMathieu Desnoyers by profilers such as OProfile. 1688125e5645SMathieu Desnoyers 16895f87f112SIngo Molnar# 16905f87f112SIngo Molnar# Place an empty function call at each tracepoint site. Can be 16915f87f112SIngo Molnar# dynamically changed for a probe function. 16925f87f112SIngo Molnar# 169397e1c18eSMathieu Desnoyersconfig TRACEPOINTS 16945f87f112SIngo Molnar bool 169597e1c18eSMathieu Desnoyers 1696fb32e03fSMathieu Desnoyerssource "arch/Kconfig" 1697fb32e03fSMathieu Desnoyers 16981da177e4SLinus Torvaldsendmenu # General setup 16991da177e4SLinus Torvalds 1700ee7e5516SDmitry Baryshkovconfig HAVE_GENERIC_DMA_COHERENT 1701ee7e5516SDmitry Baryshkov bool 1702ee7e5516SDmitry Baryshkov default n 1703ee7e5516SDmitry Baryshkov 1704ae81f9e3SChuck Ebbertconfig RT_MUTEXES 17056341e62bSChristoph Jaeger bool 1706ae81f9e3SChuck Ebbert 17071da177e4SLinus Torvaldsconfig BASE_SMALL 17081da177e4SLinus Torvalds int 17091da177e4SLinus Torvalds default 0 if BASE_FULL 17101da177e4SLinus Torvalds default 1 if !BASE_FULL 17111da177e4SLinus Torvalds 171266da5733SJan Engelhardtmenuconfig MODULES 17131da177e4SLinus Torvalds bool "Enable loadable module support" 171411097a03SYann E. MORIN option modules 17151da177e4SLinus Torvalds help 17161da177e4SLinus Torvalds Kernel modules are small pieces of compiled code which can 17171da177e4SLinus Torvalds be inserted in the running kernel, rather than being 17181da177e4SLinus Torvalds permanently built into the kernel. You use the "modprobe" 17191da177e4SLinus Torvalds tool to add (and sometimes remove) them. If you say Y here, 17201da177e4SLinus Torvalds many parts of the kernel can be built as modules (by 17211da177e4SLinus Torvalds answering M instead of Y where indicated): this is most 17221da177e4SLinus Torvalds useful for infrequently used options which are not required 17231da177e4SLinus Torvalds for booting. For more information, see the man pages for 17241da177e4SLinus Torvalds modprobe, lsmod, modinfo, insmod and rmmod. 17251da177e4SLinus Torvalds 17261da177e4SLinus Torvalds If you say Y here, you will need to run "make 17271da177e4SLinus Torvalds modules_install" to put the modules under /lib/modules/ 17281da177e4SLinus Torvalds where modprobe can find them (you may need to be root to do 17291da177e4SLinus Torvalds this). 17301da177e4SLinus Torvalds 17311da177e4SLinus Torvalds If unsure, say Y. 17321da177e4SLinus Torvalds 17330b0de144SRobert P. J. Dayif MODULES 17340b0de144SRobert P. J. Day 1735826e4506SLinus Torvaldsconfig MODULE_FORCE_LOAD 1736826e4506SLinus Torvalds bool "Forced module loading" 1737826e4506SLinus Torvalds default n 1738826e4506SLinus Torvalds help 173991e37a79SRusty Russell Allow loading of modules without version information (ie. modprobe 174091e37a79SRusty Russell --force). Forced module loading sets the 'F' (forced) taint flag and 174191e37a79SRusty Russell is usually a really bad idea. 1742826e4506SLinus Torvalds 17431da177e4SLinus Torvaldsconfig MODULE_UNLOAD 17441da177e4SLinus Torvalds bool "Module unloading" 17451da177e4SLinus Torvalds help 17461da177e4SLinus Torvalds Without this option you will not be able to unload any 17471da177e4SLinus Torvalds modules (note that some modules may not be unloadable 1748f7f5b675SDenys Vlasenko anyway), which makes your kernel smaller, faster 1749f7f5b675SDenys Vlasenko and simpler. If unsure, say Y. 17501da177e4SLinus Torvalds 17511da177e4SLinus Torvaldsconfig MODULE_FORCE_UNLOAD 17521da177e4SLinus Torvalds bool "Forced module unloading" 175319c92399SKees Cook depends on MODULE_UNLOAD 17541da177e4SLinus Torvalds help 17551da177e4SLinus Torvalds This option allows you to force a module to unload, even if the 17561da177e4SLinus Torvalds kernel believes it is unsafe: the kernel will remove the module 17571da177e4SLinus Torvalds without waiting for anyone to stop using it (using the -f option to 17581da177e4SLinus Torvalds rmmod). This is mainly for kernel developers and desperate users. 17591da177e4SLinus Torvalds If unsure, say N. 17601da177e4SLinus Torvalds 17611da177e4SLinus Torvaldsconfig MODVERSIONS 17620d541643SSam Ravnborg bool "Module versioning support" 17631da177e4SLinus Torvalds help 17641da177e4SLinus Torvalds Usually, you have to use modules compiled with your kernel. 17651da177e4SLinus Torvalds Saying Y here makes it sometimes possible to use modules 17661da177e4SLinus Torvalds compiled for different kernels, by adding enough information 17671da177e4SLinus Torvalds to the modules to (hopefully) spot any changes which would 17681da177e4SLinus Torvalds make them incompatible with the kernel you are running. If 17691da177e4SLinus Torvalds unsure, say N. 17701da177e4SLinus Torvalds 177156067812SArd Biesheuvelconfig MODULE_REL_CRCS 177256067812SArd Biesheuvel bool 177356067812SArd Biesheuvel depends on MODVERSIONS 177456067812SArd Biesheuvel 17751da177e4SLinus Torvaldsconfig MODULE_SRCVERSION_ALL 17761da177e4SLinus Torvalds bool "Source checksum for all modules" 17771da177e4SLinus Torvalds help 17781da177e4SLinus Torvalds Modules which contain a MODULE_VERSION get an extra "srcversion" 17791da177e4SLinus Torvalds field inserted into their modinfo section, which contains a 17801da177e4SLinus Torvalds sum of the source files which made it. This helps maintainers 17811da177e4SLinus Torvalds see exactly which source was used to build a module (since 17821da177e4SLinus Torvalds others sometimes change the module source without updating 17831da177e4SLinus Torvalds the version). With this option, such a "srcversion" field 17841da177e4SLinus Torvalds will be created for all modules. If unsure, say N. 17851da177e4SLinus Torvalds 1786106a4ee2SRusty Russellconfig MODULE_SIG 1787106a4ee2SRusty Russell bool "Module signature verification" 1788106a4ee2SRusty Russell depends on MODULES 1789091f6e26SDavid Howells select SYSTEM_DATA_VERIFICATION 1790106a4ee2SRusty Russell help 1791106a4ee2SRusty Russell Check modules for valid signatures upon load: the signature 1792106a4ee2SRusty Russell is simply appended to the module. For more information see 1793cbdc8217SNathan Chancellor <file:Documentation/admin-guide/module-signing.rst>. 1794106a4ee2SRusty Russell 1795228c37ffSDavid Howells Note that this option adds the OpenSSL development packages as a 1796228c37ffSDavid Howells kernel build dependency so that the signing tool can use its crypto 1797228c37ffSDavid Howells library. 1798228c37ffSDavid Howells 1799ea0b6dcfSDavid Howells !!!WARNING!!! If you enable this option, you MUST make sure that the 1800ea0b6dcfSDavid Howells module DOES NOT get stripped after being signed. This includes the 1801ea0b6dcfSDavid Howells debuginfo strip done by some packagers (such as rpmbuild) and 1802ea0b6dcfSDavid Howells inclusion into an initramfs that wants the module size reduced. 1803ea0b6dcfSDavid Howells 1804106a4ee2SRusty Russellconfig MODULE_SIG_FORCE 1805106a4ee2SRusty Russell bool "Require modules to be validly signed" 1806106a4ee2SRusty Russell depends on MODULE_SIG 1807106a4ee2SRusty Russell help 1808106a4ee2SRusty Russell Reject unsigned modules or signed modules for which we don't have a 1809106a4ee2SRusty Russell key. Without this, such modules will simply taint the kernel. 1810ea0b6dcfSDavid Howells 1811d9d8d7edSMichal Marekconfig MODULE_SIG_ALL 1812d9d8d7edSMichal Marek bool "Automatically sign all modules" 1813d9d8d7edSMichal Marek default y 1814d9d8d7edSMichal Marek depends on MODULE_SIG 1815d9d8d7edSMichal Marek help 1816d9d8d7edSMichal Marek Sign all modules during make modules_install. Without this option, 1817d9d8d7edSMichal Marek modules must be signed manually, using the scripts/sign-file tool. 1818d9d8d7edSMichal Marek 1819d9d8d7edSMichal Marekcomment "Do not forget to sign required modules with scripts/sign-file" 1820d9d8d7edSMichal Marek depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL 1821d9d8d7edSMichal Marek 1822ea0b6dcfSDavid Howellschoice 1823ea0b6dcfSDavid Howells prompt "Which hash algorithm should modules be signed with?" 1824ea0b6dcfSDavid Howells depends on MODULE_SIG 1825ea0b6dcfSDavid Howells help 1826ea0b6dcfSDavid Howells This determines which sort of hashing algorithm will be used during 1827ea0b6dcfSDavid Howells signature generation. This algorithm _must_ be built into the kernel 1828ea0b6dcfSDavid Howells directly so that signature verification can take place. It is not 1829ea0b6dcfSDavid Howells possible to load a signed module containing the algorithm to check 1830ea0b6dcfSDavid Howells the signature on that module. 1831ea0b6dcfSDavid Howells 1832ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA1 1833ea0b6dcfSDavid Howells bool "Sign modules with SHA-1" 1834ea0b6dcfSDavid Howells select CRYPTO_SHA1 1835ea0b6dcfSDavid Howells 1836ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA224 1837ea0b6dcfSDavid Howells bool "Sign modules with SHA-224" 1838ea0b6dcfSDavid Howells select CRYPTO_SHA256 1839ea0b6dcfSDavid Howells 1840ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA256 1841ea0b6dcfSDavid Howells bool "Sign modules with SHA-256" 1842ea0b6dcfSDavid Howells select CRYPTO_SHA256 1843ea0b6dcfSDavid Howells 1844ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA384 1845ea0b6dcfSDavid Howells bool "Sign modules with SHA-384" 1846ea0b6dcfSDavid Howells select CRYPTO_SHA512 1847ea0b6dcfSDavid Howells 1848ea0b6dcfSDavid Howellsconfig MODULE_SIG_SHA512 1849ea0b6dcfSDavid Howells bool "Sign modules with SHA-512" 1850ea0b6dcfSDavid Howells select CRYPTO_SHA512 1851ea0b6dcfSDavid Howells 1852ea0b6dcfSDavid Howellsendchoice 1853ea0b6dcfSDavid Howells 185422753674SMichal Marekconfig MODULE_SIG_HASH 185522753674SMichal Marek string 185622753674SMichal Marek depends on MODULE_SIG 185722753674SMichal Marek default "sha1" if MODULE_SIG_SHA1 185822753674SMichal Marek default "sha224" if MODULE_SIG_SHA224 185922753674SMichal Marek default "sha256" if MODULE_SIG_SHA256 186022753674SMichal Marek default "sha384" if MODULE_SIG_SHA384 186122753674SMichal Marek default "sha512" if MODULE_SIG_SHA512 186222753674SMichal Marek 1863beb50df3SBertrand Jacquinconfig MODULE_COMPRESS 1864beb50df3SBertrand Jacquin bool "Compress modules on installation" 1865beb50df3SBertrand Jacquin depends on MODULES 1866beb50df3SBertrand Jacquin help 1867beb50df3SBertrand Jacquin 1868b6c09b51SRusty Russell Compresses kernel modules when 'make modules_install' is run; gzip or 1869b6c09b51SRusty Russell xz depending on "Compression algorithm" below. 1870beb50df3SBertrand Jacquin 1871b6c09b51SRusty Russell module-init-tools MAY support gzip, and kmod MAY support gzip and xz. 1872beb50df3SBertrand Jacquin 1873b6c09b51SRusty Russell Out-of-tree kernel modules installed using Kbuild will also be 1874b6c09b51SRusty Russell compressed upon installation. 1875beb50df3SBertrand Jacquin 1876b6c09b51SRusty Russell Note: for modules inside an initrd or initramfs, it's more efficient 1877b6c09b51SRusty Russell to compress the whole initrd or initramfs instead. 1878beb50df3SBertrand Jacquin 1879b6c09b51SRusty Russell Note: This is fully compatible with signed modules. 1880b6c09b51SRusty Russell 1881b6c09b51SRusty Russell If in doubt, say N. 1882beb50df3SBertrand Jacquin 1883beb50df3SBertrand Jacquinchoice 1884beb50df3SBertrand Jacquin prompt "Compression algorithm" 1885beb50df3SBertrand Jacquin depends on MODULE_COMPRESS 1886beb50df3SBertrand Jacquin default MODULE_COMPRESS_GZIP 1887beb50df3SBertrand Jacquin help 1888beb50df3SBertrand Jacquin This determines which sort of compression will be used during 1889beb50df3SBertrand Jacquin 'make modules_install'. 1890beb50df3SBertrand Jacquin 1891beb50df3SBertrand Jacquin GZIP (default) and XZ are supported. 1892beb50df3SBertrand Jacquin 1893beb50df3SBertrand Jacquinconfig MODULE_COMPRESS_GZIP 1894beb50df3SBertrand Jacquin bool "GZIP" 1895beb50df3SBertrand Jacquin 1896beb50df3SBertrand Jacquinconfig MODULE_COMPRESS_XZ 1897beb50df3SBertrand Jacquin bool "XZ" 1898beb50df3SBertrand Jacquin 1899beb50df3SBertrand Jacquinendchoice 1900beb50df3SBertrand Jacquin 1901dbacb0efSNicolas Pitreconfig TRIM_UNUSED_KSYMS 1902dbacb0efSNicolas Pitre bool "Trim unused exported kernel symbols" 1903dbacb0efSNicolas Pitre depends on MODULES && !UNUSED_SYMBOLS 1904dbacb0efSNicolas Pitre help 1905dbacb0efSNicolas Pitre The kernel and some modules make many symbols available for 1906dbacb0efSNicolas Pitre other modules to use via EXPORT_SYMBOL() and variants. Depending 1907dbacb0efSNicolas Pitre on the set of modules being selected in your kernel configuration, 1908dbacb0efSNicolas Pitre many of those exported symbols might never be used. 1909dbacb0efSNicolas Pitre 1910dbacb0efSNicolas Pitre This option allows for unused exported symbols to be dropped from 1911dbacb0efSNicolas Pitre the build. In turn, this provides the compiler more opportunities 1912dbacb0efSNicolas Pitre (especially when using LTO) for optimizing the code and reducing 1913dbacb0efSNicolas Pitre binary size. This might have some security advantages as well. 1914dbacb0efSNicolas Pitre 1915f1cb637eSValdis Kletnieks If unsure, or if you need to build out-of-tree modules, say N. 1916dbacb0efSNicolas Pitre 19170b0de144SRobert P. J. Dayendif # MODULES 19180b0de144SRobert P. J. Day 19196c9692e2SPeter Zijlstraconfig MODULES_TREE_LOOKUP 19206c9692e2SPeter Zijlstra def_bool y 19216c9692e2SPeter Zijlstra depends on PERF_EVENTS || TRACING 19226c9692e2SPeter Zijlstra 192398a79d6aSRusty Russellconfig INIT_ALL_POSSIBLE 192498a79d6aSRusty Russell bool 192598a79d6aSRusty Russell help 19265f054e31SRusty Russell Back when each arch used to define their own cpu_online_mask and 19275f054e31SRusty Russell cpu_possible_mask, some of them chose to initialize cpu_possible_mask 192898a79d6aSRusty Russell with all 1s, and others with all 0s. When they were centralised, 192998a79d6aSRusty Russell it was better to provide this option than to break all the archs 1930692105b8SMatt LaPlante and have several arch maintainers pursuing me down dark alleys. 193198a79d6aSRusty Russell 19323a65dfe8SJens Axboesource "block/Kconfig" 1933e98c3202SAvi Kivity 1934e98c3202SAvi Kivityconfig PREEMPT_NOTIFIERS 1935e98c3202SAvi Kivity bool 1936e260be67SPaul E. McKenney 193716295becSSteffen Klassertconfig PADATA 193816295becSSteffen Klassert depends on SMP 193916295becSSteffen Klassert bool 194016295becSSteffen Klassert 19414520c6a4SDavid Howellsconfig ASN1 19424520c6a4SDavid Howells tristate 19434520c6a4SDavid Howells help 19444520c6a4SDavid Howells Build a simple ASN.1 grammar compiler that produces a bytecode output 19454520c6a4SDavid Howells that can be interpreted by the ASN.1 stream decoder and used to 19464520c6a4SDavid Howells inform it as to what tags are to be expected in a stream and what 19474520c6a4SDavid Howells functions to call on what tags. 19484520c6a4SDavid Howells 19496beb0009SThomas Gleixnersource "kernel/Kconfig.locks" 1950e61938a9SMathieu Desnoyers 1951e61938a9SMathieu Desnoyersconfig ARCH_HAS_SYNC_CORE_BEFORE_USERMODE 1952e61938a9SMathieu Desnoyers bool 19531bd21c6cSDominik Brodowski 19541bd21c6cSDominik Brodowski# It may be useful for an architecture to override the definitions of the 19557303e30eSDominik Brodowski# SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h> 19567303e30eSDominik Brodowski# and the COMPAT_ variants in <linux/compat.h>, in particular to use a 19577303e30eSDominik Brodowski# different calling convention for syscalls. They can also override the 19587303e30eSDominik Brodowski# macros for not-implemented syscalls in kernel/sys_ni.c and 19597303e30eSDominik Brodowski# kernel/time/posix-stubs.c. All these overrides need to be available in 19607303e30eSDominik Brodowski# <asm/syscall_wrapper.h>. 19611bd21c6cSDominik Brodowskiconfig ARCH_HAS_SYSCALL_WRAPPER 19621bd21c6cSDominik Brodowski def_bool n 1963