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 HAVE_IRQ_WORK 24e360adbeSPeter Zijlstra bool 25e360adbeSPeter Zijlstra 26e360adbeSPeter Zijlstraconfig IRQ_WORK 27e360adbeSPeter Zijlstra bool 28e360adbeSPeter Zijlstra depends on HAVE_IRQ_WORK 29e360adbeSPeter Zijlstra 301dbdc6f1SDavid Daneyconfig BUILDTIME_EXTABLE_SORT 311dbdc6f1SDavid Daney bool 321dbdc6f1SDavid Daney 33ff0cfc66SAl Boldimenu "General setup" 341da177e4SLinus Torvalds 351da177e4SLinus Torvaldsconfig EXPERIMENTAL 361da177e4SLinus Torvalds bool "Prompt for development and/or incomplete code/drivers" 371da177e4SLinus Torvalds ---help--- 381da177e4SLinus Torvalds Some of the various things that Linux supports (such as network 391da177e4SLinus Torvalds drivers, file systems, network protocols, etc.) can be in a state 401da177e4SLinus Torvalds of development where the functionality, stability, or the level of 411da177e4SLinus Torvalds testing is not yet high enough for general use. This is usually 421da177e4SLinus Torvalds known as the "alpha-test" phase among developers. If a feature is 431da177e4SLinus Torvalds currently in alpha-test, then the developers usually discourage 441da177e4SLinus Torvalds uninformed widespread use of this feature by the general public to 451da177e4SLinus Torvalds avoid "Why doesn't this work?" type mail messages. However, active 461da177e4SLinus Torvalds testing and use of these systems is welcomed. Just be aware that it 471da177e4SLinus Torvalds may not meet the normal level of reliability or it may fail to work 481da177e4SLinus Torvalds in some special cases. Detailed bug reports from people familiar 491da177e4SLinus Torvalds with the kernel internals are usually welcomed by the developers 501da177e4SLinus Torvalds (before submitting bug reports, please read the documents 511da177e4SLinus Torvalds <file:README>, <file:MAINTAINERS>, <file:REPORTING-BUGS>, 521da177e4SLinus Torvalds <file:Documentation/BUG-HUNTING>, and 531da177e4SLinus Torvalds <file:Documentation/oops-tracing.txt> in the kernel source). 541da177e4SLinus Torvalds 551da177e4SLinus Torvalds This option will also make obsoleted drivers available. These are 561da177e4SLinus Torvalds drivers that have been replaced by something else, and/or are 571da177e4SLinus Torvalds scheduled to be removed in a future kernel release. 581da177e4SLinus Torvalds 591da177e4SLinus Torvalds Unless you intend to help test and develop a feature or driver that 601da177e4SLinus Torvalds falls into this category, or you have a situation that requires 611da177e4SLinus Torvalds using these features, you should probably say N here, which will 621da177e4SLinus Torvalds cause the configurator to present you with fewer choices. If 631da177e4SLinus Torvalds you say Y here, you will be offered the choice of using features or 641da177e4SLinus Torvalds drivers that are currently considered to be in the alpha-test phase. 651da177e4SLinus Torvalds 661da177e4SLinus Torvaldsconfig BROKEN 671da177e4SLinus Torvalds bool 681da177e4SLinus Torvalds 691da177e4SLinus Torvaldsconfig BROKEN_ON_SMP 701da177e4SLinus Torvalds bool 711da177e4SLinus Torvalds depends on BROKEN || !SMP 721da177e4SLinus Torvalds default y 731da177e4SLinus Torvalds 741da177e4SLinus Torvaldsconfig INIT_ENV_ARG_LIMIT 751da177e4SLinus Torvalds int 76dd673bcaSAdrian Bunk default 32 if !UML 77dd673bcaSAdrian Bunk default 128 if UML 781da177e4SLinus Torvalds help 7934ad92c2SRandy Dunlap Maximum of each of the number of arguments and environment 8034ad92c2SRandy Dunlap variables passed to init from the kernel command line. 811da177e4SLinus Torvalds 821da177e4SLinus Torvalds 8384336466SRoland McGrathconfig CROSS_COMPILE 8484336466SRoland McGrath string "Cross-compiler tool prefix" 8584336466SRoland McGrath help 8684336466SRoland McGrath Same as running 'make CROSS_COMPILE=prefix-' but stored for 8784336466SRoland McGrath default make runs in this kernel build directory. You don't 8884336466SRoland McGrath need to set this unless you want the configured kernel build 8984336466SRoland McGrath directory to select the cross-compiler automatically. 9084336466SRoland McGrath 911da177e4SLinus Torvaldsconfig LOCALVERSION 921da177e4SLinus Torvalds string "Local version - append to kernel release" 931da177e4SLinus Torvalds help 941da177e4SLinus Torvalds Append an extra string to the end of your kernel version. 951da177e4SLinus Torvalds This will show up when you type uname, for example. 961da177e4SLinus Torvalds The string you set here will be appended after the contents of 971da177e4SLinus Torvalds any files with a filename matching localversion* in your 981da177e4SLinus Torvalds object and source tree, in that order. Your total string can 991da177e4SLinus Torvalds be a maximum of 64 characters. 1001da177e4SLinus Torvalds 101aaebf433SRyan Andersonconfig LOCALVERSION_AUTO 102aaebf433SRyan Anderson bool "Automatically append version information to the version string" 103aaebf433SRyan Anderson default y 104aaebf433SRyan Anderson help 105aaebf433SRyan Anderson This will try to automatically determine if the current tree is a 1066e5a5420SRobert P. J. Day release tree by looking for git tags that belong to the current 1076e5a5420SRobert P. J. Day top of tree revision. 108aaebf433SRyan Anderson 109aaebf433SRyan Anderson A string of the format -gxxxxxxxx will be added to the localversion 1106e5a5420SRobert P. J. Day if a git-based tree is found. The string generated by this will be 111aaebf433SRyan Anderson appended after any matching localversion* files, and after the value 1126e5a5420SRobert P. J. Day set in CONFIG_LOCALVERSION. 113aaebf433SRyan Anderson 1146e5a5420SRobert P. J. Day (The actual string used here is the first eight characters produced 1156e5a5420SRobert P. J. Day by running the command: 1166e5a5420SRobert P. J. Day 1176e5a5420SRobert P. J. Day $ git rev-parse --verify HEAD 1186e5a5420SRobert P. J. Day 1196e5a5420SRobert P. J. Day which is done within the script "scripts/setlocalversion".) 120aaebf433SRyan Anderson 1212e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_GZIP 1222e9f3bddSH. Peter Anvin bool 1232e9f3bddSH. Peter Anvin 1242e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_BZIP2 1252e9f3bddSH. Peter Anvin bool 1262e9f3bddSH. Peter Anvin 1272e9f3bddSH. Peter Anvinconfig HAVE_KERNEL_LZMA 1282e9f3bddSH. Peter Anvin bool 1292e9f3bddSH. Peter Anvin 1303ebe1243SLasse Collinconfig HAVE_KERNEL_XZ 1313ebe1243SLasse Collin bool 1323ebe1243SLasse Collin 1337dd65febSAlbin Tonnerreconfig HAVE_KERNEL_LZO 1347dd65febSAlbin Tonnerre bool 1357dd65febSAlbin Tonnerre 13630d65dbfSAlain Knaffchoice 13730d65dbfSAlain Knaff prompt "Kernel compression mode" 13830d65dbfSAlain Knaff default KERNEL_GZIP 1393ebe1243SLasse Collin depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO 14030d65dbfSAlain Knaff help 14130d65dbfSAlain Knaff The linux kernel is a kind of self-extracting executable. 14230d65dbfSAlain Knaff Several compression algorithms are available, which differ 14330d65dbfSAlain Knaff in efficiency, compression and decompression speed. 14430d65dbfSAlain Knaff Compression speed is only relevant when building a kernel. 14530d65dbfSAlain Knaff Decompression speed is relevant at each boot. 14630d65dbfSAlain Knaff 14730d65dbfSAlain Knaff If you have any problems with bzip2 or lzma compressed 14830d65dbfSAlain Knaff kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older 14930d65dbfSAlain Knaff version of this functionality (bzip2 only), for 2.4, was 15030d65dbfSAlain Knaff supplied by Christian Ludwig) 15130d65dbfSAlain Knaff 15230d65dbfSAlain Knaff High compression options are mostly useful for users, who 15330d65dbfSAlain Knaff are low on disk space (embedded systems), but for whom ram 15430d65dbfSAlain Knaff size matters less. 15530d65dbfSAlain Knaff 15630d65dbfSAlain Knaff If in doubt, select 'gzip' 15730d65dbfSAlain Knaff 15830d65dbfSAlain Knaffconfig KERNEL_GZIP 15930d65dbfSAlain Knaff bool "Gzip" 1602e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_GZIP 16130d65dbfSAlain Knaff help 1627dd65febSAlbin Tonnerre The old and tried gzip compression. It provides a good balance 1637dd65febSAlbin Tonnerre between compression ratio and decompression speed. 16430d65dbfSAlain Knaff 16530d65dbfSAlain Knaffconfig KERNEL_BZIP2 16630d65dbfSAlain Knaff bool "Bzip2" 1672e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_BZIP2 16830d65dbfSAlain Knaff help 16930d65dbfSAlain Knaff Its compression ratio and speed is intermediate. 1700a4dd35cSRandy Dunlap Decompression speed is slowest among the choices. The kernel 1712e9f3bddSH. Peter Anvin size is about 10% smaller with bzip2, in comparison to gzip. 1722e9f3bddSH. Peter Anvin Bzip2 uses a large amount of memory. For modern kernels you 1732e9f3bddSH. Peter Anvin will need at least 8MB RAM or more for booting. 17430d65dbfSAlain Knaff 17530d65dbfSAlain Knaffconfig KERNEL_LZMA 17630d65dbfSAlain Knaff bool "LZMA" 1772e9f3bddSH. Peter Anvin depends on HAVE_KERNEL_LZMA 17830d65dbfSAlain Knaff help 1790a4dd35cSRandy Dunlap This compression algorithm's ratio is best. Decompression speed 1800a4dd35cSRandy Dunlap is between gzip and bzip2. Compression is slowest. 1810a4dd35cSRandy Dunlap The kernel size is about 33% smaller with LZMA in comparison to gzip. 18230d65dbfSAlain Knaff 1833ebe1243SLasse Collinconfig KERNEL_XZ 1843ebe1243SLasse Collin bool "XZ" 1853ebe1243SLasse Collin depends on HAVE_KERNEL_XZ 1863ebe1243SLasse Collin help 1873ebe1243SLasse Collin XZ uses the LZMA2 algorithm and instruction set specific 1883ebe1243SLasse Collin BCJ filters which can improve compression ratio of executable 1893ebe1243SLasse Collin code. The size of the kernel is about 30% smaller with XZ in 1903ebe1243SLasse Collin comparison to gzip. On architectures for which there is a BCJ 1913ebe1243SLasse Collin filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ 1923ebe1243SLasse Collin will create a few percent smaller kernel than plain LZMA. 1933ebe1243SLasse Collin 1943ebe1243SLasse Collin The speed is about the same as with LZMA: The decompression 1953ebe1243SLasse Collin speed of XZ is better than that of bzip2 but worse than gzip 1963ebe1243SLasse Collin and LZO. Compression is slow. 1973ebe1243SLasse Collin 1987dd65febSAlbin Tonnerreconfig KERNEL_LZO 1997dd65febSAlbin Tonnerre bool "LZO" 2007dd65febSAlbin Tonnerre depends on HAVE_KERNEL_LZO 2017dd65febSAlbin Tonnerre help 2020a4dd35cSRandy Dunlap Its compression ratio is the poorest among the choices. The kernel 203681b3049SStephan Sperber size is about 10% bigger than gzip; however its speed 2047dd65febSAlbin Tonnerre (both compression and decompression) is the fastest. 2057dd65febSAlbin Tonnerre 20630d65dbfSAlain Knaffendchoice 20730d65dbfSAlain Knaff 208bd5dc17bSJosh Triplettconfig DEFAULT_HOSTNAME 209bd5dc17bSJosh Triplett string "Default hostname" 210bd5dc17bSJosh Triplett default "(none)" 211bd5dc17bSJosh Triplett help 212bd5dc17bSJosh Triplett This option determines the default system hostname before userspace 213bd5dc17bSJosh Triplett calls sethostname(2). The kernel traditionally uses "(none)" here, 214bd5dc17bSJosh Triplett but you may wish to use a different default here to make a minimal 215bd5dc17bSJosh Triplett system more usable with less configuration. 216bd5dc17bSJosh Triplett 2171da177e4SLinus Torvaldsconfig SWAP 2181da177e4SLinus Torvalds bool "Support for paging of anonymous memory (swap)" 2199361401eSDavid Howells depends on MMU && BLOCK 2201da177e4SLinus Torvalds default y 2211da177e4SLinus Torvalds help 2221da177e4SLinus Torvalds This option allows you to choose whether you want to have support 2231da177e4SLinus Torvalds for so called swap devices or swap files in your kernel that are 2241da177e4SLinus Torvalds used to provide more virtual memory than the actual RAM present 2251da177e4SLinus Torvalds in your computer. If unsure say Y. 2261da177e4SLinus Torvalds 2271da177e4SLinus Torvaldsconfig SYSVIPC 2281da177e4SLinus Torvalds bool "System V IPC" 2291da177e4SLinus Torvalds ---help--- 2301da177e4SLinus Torvalds Inter Process Communication is a suite of library functions and 2311da177e4SLinus Torvalds system calls which let processes (running programs) synchronize and 2321da177e4SLinus Torvalds exchange information. It is generally considered to be a good thing, 2331da177e4SLinus Torvalds and some programs won't run unless you say Y here. In particular, if 2341da177e4SLinus Torvalds you want to run the DOS emulator dosemu under Linux (read the 2351da177e4SLinus Torvalds DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>), 2361da177e4SLinus Torvalds you'll need to say Y here. 2371da177e4SLinus Torvalds 2381da177e4SLinus Torvalds You can find documentation about IPC with "info ipc" and also in 2391da177e4SLinus Torvalds section 6.4 of the Linux Programmer's Guide, available from 2401da177e4SLinus Torvalds <http://www.tldp.org/guides.html>. 2411da177e4SLinus Torvalds 242a5494dcdSEric W. Biedermanconfig SYSVIPC_SYSCTL 243a5494dcdSEric W. Biederman bool 244a5494dcdSEric W. Biederman depends on SYSVIPC 245a5494dcdSEric W. Biederman depends on SYSCTL 246a5494dcdSEric W. Biederman default y 247a5494dcdSEric W. Biederman 2481da177e4SLinus Torvaldsconfig POSIX_MQUEUE 2491da177e4SLinus Torvalds bool "POSIX Message Queues" 2501da177e4SLinus Torvalds depends on NET && EXPERIMENTAL 2511da177e4SLinus Torvalds ---help--- 2521da177e4SLinus Torvalds POSIX variant of message queues is a part of IPC. In POSIX message 2531da177e4SLinus Torvalds queues every message has a priority which decides about succession 2541da177e4SLinus Torvalds of receiving it by a process. If you want to compile and run 2551da177e4SLinus Torvalds programs written e.g. for Solaris with use of its POSIX message 256b0e37650SRobert P. J. Day queues (functions mq_*) say Y here. 2571da177e4SLinus Torvalds 2581da177e4SLinus Torvalds POSIX message queues are visible as a filesystem called 'mqueue' 2591da177e4SLinus Torvalds and can be mounted somewhere if you want to do filesystem 2601da177e4SLinus Torvalds operations on message queues. 2611da177e4SLinus Torvalds 2621da177e4SLinus Torvalds If unsure, say Y. 2631da177e4SLinus Torvalds 264bdc8e5f8SSerge E. Hallynconfig POSIX_MQUEUE_SYSCTL 265bdc8e5f8SSerge E. Hallyn bool 266bdc8e5f8SSerge E. Hallyn depends on POSIX_MQUEUE 267bdc8e5f8SSerge E. Hallyn depends on SYSCTL 268bdc8e5f8SSerge E. Hallyn default y 269bdc8e5f8SSerge E. Hallyn 270990d6c2dSAneesh Kumar K.Vconfig FHANDLE 271990d6c2dSAneesh Kumar K.V bool "open by fhandle syscalls" 272990d6c2dSAneesh Kumar K.V select EXPORTFS 273990d6c2dSAneesh Kumar K.V help 274990d6c2dSAneesh Kumar K.V If you say Y here, a user level program will be able to map 275990d6c2dSAneesh Kumar K.V file names to handle and then later use the handle for 276990d6c2dSAneesh Kumar K.V different file system operations. This is useful in implementing 277990d6c2dSAneesh Kumar K.V userspace file servers, which now track files using handles instead 278990d6c2dSAneesh Kumar K.V of names. The handle would remain the same even if file names 279990d6c2dSAneesh Kumar K.V get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2) 280990d6c2dSAneesh Kumar K.V syscalls. 281990d6c2dSAneesh Kumar K.V 2821da177e4SLinus Torvaldsconfig AUDIT 2831da177e4SLinus Torvalds bool "Auditing support" 284804a6a49SChris Wright depends on NET 2851da177e4SLinus Torvalds help 2861da177e4SLinus Torvalds Enable auditing infrastructure that can be used with another 2871da177e4SLinus Torvalds kernel subsystem, such as SELinux (which requires this for 2881da177e4SLinus Torvalds logging of avc messages output). Does not do system-call 2891da177e4SLinus Torvalds auditing without CONFIG_AUDITSYSCALL. 2901da177e4SLinus Torvalds 2911da177e4SLinus Torvaldsconfig AUDITSYSCALL 2921da177e4SLinus Torvalds bool "Enable system-call auditing support" 2938f827a14SWill Deacon depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH || (ARM && AEABI && !OABI_COMPAT)) 2941da177e4SLinus Torvalds default y if SECURITY_SELINUX 2951da177e4SLinus Torvalds help 2961da177e4SLinus Torvalds Enable low-overhead system-call auditing infrastructure that 2971da177e4SLinus Torvalds can be used independently or with another kernel subsystem, 29867640b60SEric Paris such as SELinux. 2991da177e4SLinus Torvalds 300939a67fcSEric Parisconfig AUDIT_WATCH 301939a67fcSEric Paris def_bool y 302939a67fcSEric Paris depends on AUDITSYSCALL 303939a67fcSEric Paris select FSNOTIFY 3041da177e4SLinus Torvalds 30574c3cbe3SAl Viroconfig AUDIT_TREE 30674c3cbe3SAl Viro def_bool y 30763c882a0SEric Paris depends on AUDITSYSCALL 30828a3a7ebSEric Paris select FSNOTIFY 30974c3cbe3SAl Viro 310633b4545SEric Parisconfig AUDIT_LOGINUID_IMMUTABLE 311633b4545SEric Paris bool "Make audit loginuid immutable" 312633b4545SEric Paris depends on AUDIT 313633b4545SEric Paris help 314f429ee3bSLinus Torvalds The config option toggles if a task setting its loginuid requires 315633b4545SEric Paris CAP_SYS_AUDITCONTROL or if that task should require no special permissions 316633b4545SEric Paris but should instead only allow setting its loginuid if it was never 317633b4545SEric Paris previously set. On systems which use systemd or a similar central 318633b4545SEric Paris process to restart login services this should be set to true. On older 319633b4545SEric Paris systems in which an admin would typically have to directly stop and 320633b4545SEric Paris start processes this should be set to false. Setting this to true allows 321633b4545SEric Paris one to drop potentially dangerous capabilites from the login tasks, 322633b4545SEric Paris but may not be backwards compatible with older init systems. 323633b4545SEric Paris 324d9817ebeSThomas Gleixnersource "kernel/irq/Kconfig" 325764e0da1SThomas Gleixnersource "kernel/time/Kconfig" 326d9817ebeSThomas Gleixner 327391dc69cSFrederic Weisbeckermenu "CPU/Task time and stats accounting" 328391dc69cSFrederic Weisbecker 329fdf9c356SFrederic Weisbeckerchoice 330fdf9c356SFrederic Weisbecker prompt "Cputime accounting" 331fdf9c356SFrederic Weisbecker default TICK_CPU_ACCOUNTING if !PPC64 332fdf9c356SFrederic Weisbecker default VIRT_CPU_ACCOUNTING if PPC64 333fdf9c356SFrederic Weisbecker 334fdf9c356SFrederic Weisbecker# Kind of a stub config for the pure tick based cputime accounting 335fdf9c356SFrederic Weisbeckerconfig TICK_CPU_ACCOUNTING 336fdf9c356SFrederic Weisbecker bool "Simple tick based cputime accounting" 337fdf9c356SFrederic Weisbecker depends on !S390 338fdf9c356SFrederic Weisbecker help 339fdf9c356SFrederic Weisbecker This is the basic tick based cputime accounting that maintains 340fdf9c356SFrederic Weisbecker statistics about user, system and idle time spent on per jiffies 341fdf9c356SFrederic Weisbecker granularity. 342fdf9c356SFrederic Weisbecker 343fdf9c356SFrederic Weisbecker If unsure, say Y. 344fdf9c356SFrederic Weisbecker 345391dc69cSFrederic Weisbeckerconfig VIRT_CPU_ACCOUNTING 346391dc69cSFrederic Weisbecker bool "Deterministic task and CPU time accounting" 347391dc69cSFrederic Weisbecker depends on HAVE_VIRT_CPU_ACCOUNTING 348391dc69cSFrederic Weisbecker help 349391dc69cSFrederic Weisbecker Select this option to enable more accurate task and CPU time 350391dc69cSFrederic Weisbecker accounting. This is done by reading a CPU counter on each 351391dc69cSFrederic Weisbecker kernel entry and exit and on transitions within the kernel 352391dc69cSFrederic Weisbecker between system, softirq and hardirq state, so there is a 353391dc69cSFrederic Weisbecker small performance impact. In the case of s390 or IBM POWER > 5, 354391dc69cSFrederic Weisbecker this also enables accounting of stolen time on logically-partitioned 355391dc69cSFrederic Weisbecker systems. 356391dc69cSFrederic Weisbecker 357fdf9c356SFrederic Weisbeckerconfig IRQ_TIME_ACCOUNTING 358fdf9c356SFrederic Weisbecker bool "Fine granularity task level IRQ time accounting" 359fdf9c356SFrederic Weisbecker depends on HAVE_IRQ_TIME_ACCOUNTING 360fdf9c356SFrederic Weisbecker help 361fdf9c356SFrederic Weisbecker Select this option to enable fine granularity task irq time 362fdf9c356SFrederic Weisbecker accounting. This is done by reading a timestamp on each 363fdf9c356SFrederic Weisbecker transitions between softirq and hardirq state, so there can be a 364fdf9c356SFrederic Weisbecker small performance impact. 365fdf9c356SFrederic Weisbecker 366fdf9c356SFrederic Weisbecker If in doubt, say N here. 367fdf9c356SFrederic Weisbecker 368fdf9c356SFrederic Weisbeckerendchoice 369fdf9c356SFrederic Weisbecker 370391dc69cSFrederic Weisbeckerconfig BSD_PROCESS_ACCT 371391dc69cSFrederic Weisbecker bool "BSD Process Accounting" 372391dc69cSFrederic Weisbecker help 373391dc69cSFrederic Weisbecker If you say Y here, a user level program will be able to instruct the 374391dc69cSFrederic Weisbecker kernel (via a special system call) to write process accounting 375391dc69cSFrederic Weisbecker information to a file: whenever a process exits, information about 376391dc69cSFrederic Weisbecker that process will be appended to the file by the kernel. The 377391dc69cSFrederic Weisbecker information includes things such as creation time, owning user, 378391dc69cSFrederic Weisbecker command name, memory usage, controlling terminal etc. (the complete 379391dc69cSFrederic Weisbecker list is in the struct acct in <file:include/linux/acct.h>). It is 380391dc69cSFrederic Weisbecker up to the user level program to do useful things with this 381391dc69cSFrederic Weisbecker information. This is generally a good idea, so say Y. 382391dc69cSFrederic Weisbecker 383391dc69cSFrederic Weisbeckerconfig BSD_PROCESS_ACCT_V3 384391dc69cSFrederic Weisbecker bool "BSD Process Accounting version 3 file format" 385391dc69cSFrederic Weisbecker depends on BSD_PROCESS_ACCT 386391dc69cSFrederic Weisbecker default n 387391dc69cSFrederic Weisbecker help 388391dc69cSFrederic Weisbecker If you say Y here, the process accounting information is written 389391dc69cSFrederic Weisbecker in a new file format that also logs the process IDs of each 390391dc69cSFrederic Weisbecker process and it's parent. Note that this file format is incompatible 391391dc69cSFrederic Weisbecker with previous v0/v1/v2 file formats, so you will need updated tools 392391dc69cSFrederic Weisbecker for processing it. A preliminary version of these tools is available 393391dc69cSFrederic Weisbecker at <http://www.gnu.org/software/acct/>. 394391dc69cSFrederic Weisbecker 395391dc69cSFrederic Weisbeckerconfig TASKSTATS 396391dc69cSFrederic Weisbecker bool "Export task/process statistics through netlink (EXPERIMENTAL)" 397391dc69cSFrederic Weisbecker depends on NET 398391dc69cSFrederic Weisbecker default n 399391dc69cSFrederic Weisbecker help 400391dc69cSFrederic Weisbecker Export selected statistics for tasks/processes through the 401391dc69cSFrederic Weisbecker generic netlink interface. Unlike BSD process accounting, the 402391dc69cSFrederic Weisbecker statistics are available during the lifetime of tasks/processes as 403391dc69cSFrederic Weisbecker responses to commands. Like BSD accounting, they are sent to user 404391dc69cSFrederic Weisbecker space on task exit. 405391dc69cSFrederic Weisbecker 406391dc69cSFrederic Weisbecker Say N if unsure. 407391dc69cSFrederic Weisbecker 408391dc69cSFrederic Weisbeckerconfig TASK_DELAY_ACCT 409391dc69cSFrederic Weisbecker bool "Enable per-task delay accounting (EXPERIMENTAL)" 410391dc69cSFrederic Weisbecker depends on TASKSTATS 411391dc69cSFrederic Weisbecker help 412391dc69cSFrederic Weisbecker Collect information on time spent by a task waiting for system 413391dc69cSFrederic Weisbecker resources like cpu, synchronous block I/O completion and swapping 414391dc69cSFrederic Weisbecker in pages. Such statistics can help in setting a task's priorities 415391dc69cSFrederic Weisbecker relative to other tasks for cpu, io, rss limits etc. 416391dc69cSFrederic Weisbecker 417391dc69cSFrederic Weisbecker Say N if unsure. 418391dc69cSFrederic Weisbecker 419391dc69cSFrederic Weisbeckerconfig TASK_XACCT 420391dc69cSFrederic Weisbecker bool "Enable extended accounting over taskstats (EXPERIMENTAL)" 421391dc69cSFrederic Weisbecker depends on TASKSTATS 422391dc69cSFrederic Weisbecker help 423391dc69cSFrederic Weisbecker Collect extended task accounting data and send the data 424391dc69cSFrederic Weisbecker to userland for processing over the taskstats interface. 425391dc69cSFrederic Weisbecker 426391dc69cSFrederic Weisbecker Say N if unsure. 427391dc69cSFrederic Weisbecker 428391dc69cSFrederic Weisbeckerconfig TASK_IO_ACCOUNTING 429391dc69cSFrederic Weisbecker bool "Enable per-task storage I/O accounting (EXPERIMENTAL)" 430391dc69cSFrederic Weisbecker depends on TASK_XACCT 431391dc69cSFrederic Weisbecker help 432391dc69cSFrederic Weisbecker Collect information on the number of bytes of storage I/O which this 433391dc69cSFrederic Weisbecker task has caused. 434391dc69cSFrederic Weisbecker 435391dc69cSFrederic Weisbecker Say N if unsure. 436391dc69cSFrederic Weisbecker 437391dc69cSFrederic Weisbeckerendmenu # "CPU/Task time and stats accounting" 438391dc69cSFrederic Weisbecker 439c903ff83SMike Travismenu "RCU Subsystem" 440c903ff83SMike Travis 441c903ff83SMike Travischoice 442c903ff83SMike Travis prompt "RCU Implementation" 44331c9a24eSPaul E. McKenney default TREE_RCU 444c903ff83SMike Travis 445c903ff83SMike Travisconfig TREE_RCU 446c903ff83SMike Travis bool "Tree-based hierarchical RCU" 447687d7a96SPaul E. McKenney depends on !PREEMPT && SMP 448c903ff83SMike Travis help 449c903ff83SMike Travis This option selects the RCU implementation that is 450c903ff83SMike Travis designed for very large SMP system with hundreds or 451c17ef453SPaul E. McKenney thousands of CPUs. It also scales down nicely to 452c17ef453SPaul E. McKenney smaller systems. 453c903ff83SMike Travis 454f41d911fSPaul E. McKenneyconfig TREE_PREEMPT_RCU 455a57eb940SPaul E. McKenney bool "Preemptible tree-based hierarchical RCU" 4568008e129SPaul E. McKenney depends on PREEMPT && SMP 457f41d911fSPaul E. McKenney help 458f41d911fSPaul E. McKenney This option selects the RCU implementation that is 459f41d911fSPaul E. McKenney designed for very large SMP systems with hundreds or 460f41d911fSPaul E. McKenney thousands of CPUs, but for which real-time response 461bbe3eae8SPaul E. McKenney is also required. It also scales down nicely to 462bbe3eae8SPaul E. McKenney smaller systems. 463f41d911fSPaul E. McKenney 4649b1d82faSPaul E. McKenneyconfig TINY_RCU 4659b1d82faSPaul E. McKenney bool "UP-only small-memory-footprint RCU" 4668008e129SPaul E. McKenney depends on !PREEMPT && !SMP 4679b1d82faSPaul E. McKenney help 4689b1d82faSPaul E. McKenney This option selects the RCU implementation that is 4699b1d82faSPaul E. McKenney designed for UP systems from which real-time response 4709b1d82faSPaul E. McKenney is not required. This option greatly reduces the 4719b1d82faSPaul E. McKenney memory footprint of RCU. 4729b1d82faSPaul E. McKenney 473a57eb940SPaul E. McKenneyconfig TINY_PREEMPT_RCU 474a57eb940SPaul E. McKenney bool "Preemptible UP-only small-memory-footprint RCU" 4758008e129SPaul E. McKenney depends on PREEMPT && !SMP 476a57eb940SPaul E. McKenney help 477a57eb940SPaul E. McKenney This option selects the RCU implementation that is designed 478a57eb940SPaul E. McKenney for real-time UP systems. This option greatly reduces the 479a57eb940SPaul E. McKenney memory footprint of RCU. 480a57eb940SPaul E. McKenney 481c903ff83SMike Travisendchoice 482c903ff83SMike Travis 483a57eb940SPaul E. McKenneyconfig PREEMPT_RCU 484a57eb940SPaul E. McKenney def_bool ( TREE_PREEMPT_RCU || TINY_PREEMPT_RCU ) 485a57eb940SPaul E. McKenney help 486a57eb940SPaul E. McKenney This option enables preemptible-RCU code that is common between 487a57eb940SPaul E. McKenney the TREE_PREEMPT_RCU and TINY_PREEMPT_RCU implementations. 488a57eb940SPaul E. McKenney 4892b1d5024SFrederic Weisbeckerconfig RCU_USER_QS 4902b1d5024SFrederic Weisbecker bool "Consider userspace as in RCU extended quiescent state" 4912b1d5024SFrederic Weisbecker depends on HAVE_RCU_USER_QS && SMP 4922b1d5024SFrederic Weisbecker help 4932b1d5024SFrederic Weisbecker This option sets hooks on kernel / userspace boundaries and 4942b1d5024SFrederic Weisbecker puts RCU in extended quiescent state when the CPU runs in 4952b1d5024SFrederic Weisbecker userspace. It means that when a CPU runs in userspace, it is 4962b1d5024SFrederic Weisbecker excluded from the global RCU state machine and thus doesn't 4972b1d5024SFrederic Weisbecker to keep the timer tick on for RCU. 4982b1d5024SFrederic Weisbecker 4991fd2b442SFrederic Weisbeckerconfig RCU_USER_QS_FORCE 5001fd2b442SFrederic Weisbecker bool "Force userspace extended QS by default" 5011fd2b442SFrederic Weisbecker depends on RCU_USER_QS 5021fd2b442SFrederic Weisbecker help 5031fd2b442SFrederic Weisbecker Set the hooks in user/kernel boundaries by default in order to 5041fd2b442SFrederic Weisbecker test this feature that treats userspace as an extended quiescent 5051fd2b442SFrederic Weisbecker state until we have a real user like a full adaptive nohz option. 5061fd2b442SFrederic Weisbecker 507c903ff83SMike Travisconfig RCU_FANOUT 508c903ff83SMike Travis int "Tree-based hierarchical RCU fanout value" 509c903ff83SMike Travis range 2 64 if 64BIT 510c903ff83SMike Travis range 2 32 if !64BIT 511f41d911fSPaul E. McKenney depends on TREE_RCU || TREE_PREEMPT_RCU 512c903ff83SMike Travis default 64 if 64BIT 513c903ff83SMike Travis default 32 if !64BIT 514c903ff83SMike Travis help 515c903ff83SMike Travis This option controls the fanout of hierarchical implementations 516c903ff83SMike Travis of RCU, allowing RCU to work efficiently on machines with 5174d87ffadSPaul E. McKenney large numbers of CPUs. This value must be at least the fourth 5184d87ffadSPaul E. McKenney root of NR_CPUS, which allows NR_CPUS to be insanely large. 5194d87ffadSPaul E. McKenney The default value of RCU_FANOUT should be used for production 5204d87ffadSPaul E. McKenney systems, but if you are stress-testing the RCU implementation 5214d87ffadSPaul E. McKenney itself, small RCU_FANOUT values allow you to test large-system 5224d87ffadSPaul E. McKenney code paths on small(er) systems. 523c903ff83SMike Travis 524c903ff83SMike Travis Select a specific number if testing RCU itself. 525c903ff83SMike Travis Take the default if unsure. 526c903ff83SMike Travis 5278932a63dSPaul E. McKenneyconfig RCU_FANOUT_LEAF 5288932a63dSPaul E. McKenney int "Tree-based hierarchical RCU leaf-level fanout value" 5298932a63dSPaul E. McKenney range 2 RCU_FANOUT if 64BIT 5308932a63dSPaul E. McKenney range 2 RCU_FANOUT if !64BIT 5318932a63dSPaul E. McKenney depends on TREE_RCU || TREE_PREEMPT_RCU 5328932a63dSPaul E. McKenney default 16 5338932a63dSPaul E. McKenney help 5348932a63dSPaul E. McKenney This option controls the leaf-level fanout of hierarchical 5358932a63dSPaul E. McKenney implementations of RCU, and allows trading off cache misses 5368932a63dSPaul E. McKenney against lock contention. Systems that synchronize their 5378932a63dSPaul E. McKenney scheduling-clock interrupts for energy-efficiency reasons will 5388932a63dSPaul E. McKenney want the default because the smaller leaf-level fanout keeps 5398932a63dSPaul E. McKenney lock contention levels acceptably low. Very large systems 5408932a63dSPaul E. McKenney (hundreds or thousands of CPUs) will instead want to set this 5418932a63dSPaul E. McKenney value to the maximum value possible in order to reduce the 5428932a63dSPaul E. McKenney number of cache misses incurred during RCU's grace-period 5438932a63dSPaul E. McKenney initialization. These systems tend to run CPU-bound, and thus 5448932a63dSPaul E. McKenney are not helped by synchronized interrupts, and thus tend to 5458932a63dSPaul E. McKenney skew them, which reduces lock contention enough that large 5468932a63dSPaul E. McKenney leaf-level fanouts work well. 5478932a63dSPaul E. McKenney 5488932a63dSPaul E. McKenney Select a specific number if testing RCU itself. 5498932a63dSPaul E. McKenney 5508932a63dSPaul E. McKenney Select the maximum permissible value for large systems. 5518932a63dSPaul E. McKenney 5528932a63dSPaul E. McKenney Take the default if unsure. 5538932a63dSPaul E. McKenney 554c903ff83SMike Travisconfig RCU_FANOUT_EXACT 555c903ff83SMike Travis bool "Disable tree-based hierarchical RCU auto-balancing" 556f41d911fSPaul E. McKenney depends on TREE_RCU || TREE_PREEMPT_RCU 557c903ff83SMike Travis default n 558c903ff83SMike Travis help 559c903ff83SMike Travis This option forces use of the exact RCU_FANOUT value specified, 560c903ff83SMike Travis regardless of imbalances in the hierarchy. This is useful for 561c903ff83SMike Travis testing RCU itself, and might one day be useful on systems with 562c903ff83SMike Travis strong NUMA behavior. 563c903ff83SMike Travis 564c903ff83SMike Travis Without RCU_FANOUT_EXACT, the code will balance the hierarchy. 565c903ff83SMike Travis 566c903ff83SMike Travis Say N if unsure. 567c903ff83SMike Travis 5688bd93a2cSPaul E. McKenneyconfig RCU_FAST_NO_HZ 5698bd93a2cSPaul E. McKenney bool "Accelerate last non-dyntick-idle CPU's grace periods" 570b807fbffSPaul E. McKenney depends on NO_HZ && SMP 5718bd93a2cSPaul E. McKenney default n 5728bd93a2cSPaul E. McKenney help 5738bd93a2cSPaul E. McKenney This option causes RCU to attempt to accelerate grace periods 574b807fbffSPaul E. McKenney in order to allow CPUs to enter dynticks-idle state more 575b807fbffSPaul E. McKenney quickly. On the other hand, this option increases the overhead 576b807fbffSPaul E. McKenney of the dynticks-idle checking, particularly on systems with 577b807fbffSPaul E. McKenney large numbers of CPUs. 5788bd93a2cSPaul E. McKenney 5798bd93a2cSPaul E. McKenney Say Y if energy efficiency is critically important, particularly 5808bd93a2cSPaul E. McKenney if you have relatively few CPUs. 5818bd93a2cSPaul E. McKenney 5828bd93a2cSPaul E. McKenney Say N if you are unsure. 5838bd93a2cSPaul E. McKenney 584c903ff83SMike Travisconfig TREE_RCU_TRACE 585f41d911fSPaul E. McKenney def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU ) 586c903ff83SMike Travis select DEBUG_FS 587c903ff83SMike Travis help 588f41d911fSPaul E. McKenney This option provides tracing for the TREE_RCU and 589f41d911fSPaul E. McKenney TREE_PREEMPT_RCU implementations, permitting Makefile to 590f41d911fSPaul E. McKenney trivially select kernel/rcutree_trace.c. 591c903ff83SMike Travis 59224278d14SPaul E. McKenneyconfig RCU_BOOST 59324278d14SPaul E. McKenney bool "Enable RCU priority boosting" 59427f4d280SPaul E. McKenney depends on RT_MUTEXES && PREEMPT_RCU 59524278d14SPaul E. McKenney default n 59624278d14SPaul E. McKenney help 59724278d14SPaul E. McKenney This option boosts the priority of preempted RCU readers that 59824278d14SPaul E. McKenney block the current preemptible RCU grace period for too long. 59924278d14SPaul E. McKenney This option also prevents heavy loads from blocking RCU 60024278d14SPaul E. McKenney callback invocation for all flavors of RCU. 60124278d14SPaul E. McKenney 60224278d14SPaul E. McKenney Say Y here if you are working with real-time apps or heavy loads 60324278d14SPaul E. McKenney Say N here if you are unsure. 60424278d14SPaul E. McKenney 60524278d14SPaul E. McKenneyconfig RCU_BOOST_PRIO 60624278d14SPaul E. McKenney int "Real-time priority to boost RCU readers to" 60724278d14SPaul E. McKenney range 1 99 60824278d14SPaul E. McKenney depends on RCU_BOOST 60924278d14SPaul E. McKenney default 1 61024278d14SPaul E. McKenney help 611c9336643SPaul E. McKenney This option specifies the real-time priority to which long-term 612c9336643SPaul E. McKenney preempted RCU readers are to be boosted. If you are working 613c9336643SPaul E. McKenney with a real-time application that has one or more CPU-bound 614c9336643SPaul E. McKenney threads running at a real-time priority level, you should set 615c9336643SPaul E. McKenney RCU_BOOST_PRIO to a priority higher then the highest-priority 616c9336643SPaul E. McKenney real-time CPU-bound thread. The default RCU_BOOST_PRIO value 617c9336643SPaul E. McKenney of 1 is appropriate in the common case, which is real-time 618c9336643SPaul E. McKenney applications that do not have any CPU-bound threads. 619c9336643SPaul E. McKenney 620c9336643SPaul E. McKenney Some real-time applications might not have a single real-time 621c9336643SPaul E. McKenney thread that saturates a given CPU, but instead might have 622c9336643SPaul E. McKenney multiple real-time threads that, taken together, fully utilize 623c9336643SPaul E. McKenney that CPU. In this case, you should set RCU_BOOST_PRIO to 624c9336643SPaul E. McKenney a priority higher than the lowest-priority thread that is 625c9336643SPaul E. McKenney conspiring to prevent the CPU from running any non-real-time 626c9336643SPaul E. McKenney tasks. For example, if one thread at priority 10 and another 627c9336643SPaul E. McKenney thread at priority 5 are between themselves fully consuming 628c9336643SPaul E. McKenney the CPU time on a given CPU, then RCU_BOOST_PRIO should be 629c9336643SPaul E. McKenney set to priority 6 or higher. 63024278d14SPaul E. McKenney 63124278d14SPaul E. McKenney Specify the real-time priority, or take the default if unsure. 63224278d14SPaul E. McKenney 63324278d14SPaul E. McKenneyconfig RCU_BOOST_DELAY 63424278d14SPaul E. McKenney int "Milliseconds to delay boosting after RCU grace-period start" 63524278d14SPaul E. McKenney range 0 3000 63624278d14SPaul E. McKenney depends on RCU_BOOST 63724278d14SPaul E. McKenney default 500 63824278d14SPaul E. McKenney help 63924278d14SPaul E. McKenney This option specifies the time to wait after the beginning of 64024278d14SPaul E. McKenney a given grace period before priority-boosting preempted RCU 64124278d14SPaul E. McKenney readers blocking that grace period. Note that any RCU reader 64224278d14SPaul E. McKenney blocking an expedited RCU grace period is boosted immediately. 64324278d14SPaul E. McKenney 64424278d14SPaul E. McKenney Accept the default if unsure. 64524278d14SPaul E. McKenney 646c903ff83SMike Travisendmenu # "RCU Subsystem" 647c903ff83SMike Travis 6481da177e4SLinus Torvaldsconfig IKCONFIG 649f2443ab6SRoss Biro tristate "Kernel .config support" 6501da177e4SLinus Torvalds ---help--- 6511da177e4SLinus Torvalds This option enables the complete Linux kernel ".config" file 6521da177e4SLinus Torvalds contents to be saved in the kernel. It provides documentation 6531da177e4SLinus Torvalds of which kernel options are used in a running kernel or in an 6541da177e4SLinus Torvalds on-disk kernel. This information can be extracted from the kernel 6551da177e4SLinus Torvalds image file with the script scripts/extract-ikconfig and used as 6561da177e4SLinus Torvalds input to rebuild the current kernel or to build another kernel. 6571da177e4SLinus Torvalds It can also be extracted from a running kernel by reading 6581da177e4SLinus Torvalds /proc/config.gz if enabled (below). 6591da177e4SLinus Torvalds 6601da177e4SLinus Torvaldsconfig IKCONFIG_PROC 6611da177e4SLinus Torvalds bool "Enable access to .config through /proc/config.gz" 6621da177e4SLinus Torvalds depends on IKCONFIG && PROC_FS 6631da177e4SLinus Torvalds ---help--- 6641da177e4SLinus Torvalds This option enables access to the kernel configuration file 6651da177e4SLinus Torvalds through /proc/config.gz. 6661da177e4SLinus Torvalds 667794543a2SAlistair John Strachanconfig LOG_BUF_SHIFT 668794543a2SAlistair John Strachan int "Kernel log buffer size (16 => 64KB, 17 => 128KB)" 669794543a2SAlistair John Strachan range 12 21 670f17a32e9SAdrian Bunk default 17 671794543a2SAlistair John Strachan help 672794543a2SAlistair John Strachan Select kernel log buffer size as a power of 2. 673f17a32e9SAdrian Bunk Examples: 674f17a32e9SAdrian Bunk 17 => 128 KB 675f17a32e9SAdrian Bunk 16 => 64 KB 676f17a32e9SAdrian Bunk 15 => 32 KB 677f17a32e9SAdrian Bunk 14 => 16 KB 678794543a2SAlistair John Strachan 13 => 8 KB 679794543a2SAlistair John Strachan 12 => 4 KB 680794543a2SAlistair John Strachan 6815cdc38f9SKAMEZAWA Hiroyuki# 6825cdc38f9SKAMEZAWA Hiroyuki# Architectures with an unreliable sched_clock() should select this: 6835cdc38f9SKAMEZAWA Hiroyuki# 6845cdc38f9SKAMEZAWA Hiroyukiconfig HAVE_UNSTABLE_SCHED_CLOCK 6855cdc38f9SKAMEZAWA Hiroyuki bool 6865cdc38f9SKAMEZAWA Hiroyuki 68723964d2dSLi Zefanmenuconfig CGROUPS 68823964d2dSLi Zefan boolean "Control Group support" 6890dea1168SKirill A. Shutemov depends on EVENTFD 690ddbcc7e8SPaul Menage help 69123964d2dSLi Zefan This option adds support for grouping sets of processes together, for 6925cdc38f9SKAMEZAWA Hiroyuki use with process control subsystems such as Cpusets, CFS, memory 6935cdc38f9SKAMEZAWA Hiroyuki controls or device isolation. 6945cdc38f9SKAMEZAWA Hiroyuki See 6955cdc38f9SKAMEZAWA Hiroyuki - Documentation/scheduler/sched-design-CFS.txt (CFS) 69645ce80fbSLi Zefan - Documentation/cgroups/ (features for grouping, isolation 69745ce80fbSLi Zefan and resource control) 698ddbcc7e8SPaul Menage 699ddbcc7e8SPaul Menage Say N if unsure. 700ddbcc7e8SPaul Menage 70123964d2dSLi Zefanif CGROUPS 70223964d2dSLi Zefan 703006cb992SPaul Menageconfig CGROUP_DEBUG 704006cb992SPaul Menage bool "Example debug cgroup subsystem" 705418d7d87SPaul Menage default n 706006cb992SPaul Menage help 707006cb992SPaul Menage This option enables a simple cgroup subsystem that 708006cb992SPaul Menage exports useful debugging information about the cgroups 70923964d2dSLi Zefan framework. 710006cb992SPaul Menage 71123964d2dSLi Zefan Say N if unsure. 712006cb992SPaul Menage 713dc52ddc0SMatt Helsleyconfig CGROUP_FREEZER 71423964d2dSLi Zefan bool "Freezer cgroup subsystem" 715dc52ddc0SMatt Helsley help 716dc52ddc0SMatt Helsley Provides a way to freeze and unfreeze all tasks in a 717dc52ddc0SMatt Helsley cgroup. 718dc52ddc0SMatt Helsley 71908ce5f16SSerge E. Hallynconfig CGROUP_DEVICE 72008ce5f16SSerge E. Hallyn bool "Device controller for cgroups" 72108ce5f16SSerge E. Hallyn help 72208ce5f16SSerge E. Hallyn Provides a cgroup implementing whitelists for devices which 72308ce5f16SSerge E. Hallyn a process in the cgroup can mknod or open. 72408ce5f16SSerge E. Hallyn 7251da177e4SLinus Torvaldsconfig CPUSETS 7261da177e4SLinus Torvalds bool "Cpuset support" 7271da177e4SLinus Torvalds help 728d9fd8a6dSRandy Dunlap This option will let you create and manage CPUSETs which 7291da177e4SLinus Torvalds allow dynamically partitioning a system into sets of CPUs and 7301da177e4SLinus Torvalds Memory Nodes and assigning tasks to run only within those sets. 7311da177e4SLinus Torvalds This is primarily useful on large SMP or NUMA systems. 7321da177e4SLinus Torvalds 7331da177e4SLinus Torvalds Say N if unsure. 7341da177e4SLinus Torvalds 73523964d2dSLi Zefanconfig PROC_PID_CPUSET 73623964d2dSLi Zefan bool "Include legacy /proc/<pid>/cpuset file" 73723964d2dSLi Zefan depends on CPUSETS 73823964d2dSLi Zefan default y 73923964d2dSLi Zefan 740d842de87SSrivatsa Vaddagiriconfig CGROUP_CPUACCT 741d842de87SSrivatsa Vaddagiri bool "Simple CPU accounting cgroup subsystem" 742d842de87SSrivatsa Vaddagiri help 743d842de87SSrivatsa Vaddagiri Provides a simple Resource Controller for monitoring the 74423964d2dSLi Zefan total CPU consumed by the tasks in a cgroup. 745d842de87SSrivatsa Vaddagiri 746e552b661SPavel Emelianovconfig RESOURCE_COUNTERS 747e552b661SPavel Emelianov bool "Resource counters" 748e552b661SPavel Emelianov help 749e552b661SPavel Emelianov This option enables controller independent resource accounting 75023964d2dSLi Zefan infrastructure that works with cgroups. 751e552b661SPavel Emelianov 752c255a458SAndrew Mortonconfig MEMCG 75300f0b825SBalbir Singh bool "Memory Resource Controller for Control Groups" 75479ae9c29SDaniel Lezcano depends on RESOURCE_COUNTERS 755cf475ad2SBalbir Singh select MM_OWNER 75600f0b825SBalbir Singh help 75784ad6d70SKAMEZAWA Hiroyuki Provides a memory resource controller that manages both anonymous 75821acb9caSThadeu Lima de Souza Cascardo memory and page cache. (See Documentation/cgroups/memory.txt) 75900f0b825SBalbir Singh 76000f0b825SBalbir Singh Note that setting this option increases fixed memory overhead 76184ad6d70SKAMEZAWA Hiroyuki associated with each page of memory in the system. By this, 76284ad6d70SKAMEZAWA Hiroyuki 20(40)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory 76384ad6d70SKAMEZAWA Hiroyuki usage tracking struct at boot. Total amount of this is printed out 76484ad6d70SKAMEZAWA Hiroyuki at boot. 76500f0b825SBalbir Singh 76600f0b825SBalbir Singh Only enable when you're ok with these trade offs and really 76784ad6d70SKAMEZAWA Hiroyuki sure you need the memory resource controller. Even when you enable 76884ad6d70SKAMEZAWA Hiroyuki this, you can set "cgroup_disable=memory" at your boot option to 76984ad6d70SKAMEZAWA Hiroyuki disable memory resource controller and you can avoid overheads. 770c9d5409fSLi Zefan (and lose benefits of memory resource controller) 77100f0b825SBalbir Singh 772cf475ad2SBalbir Singh This config option also selects MM_OWNER config option, which 773cf475ad2SBalbir Singh could in turn add some fork/exit overhead. 774cf475ad2SBalbir Singh 775c255a458SAndrew Mortonconfig MEMCG_SWAP 77665e0e811SKAMEZAWA Hiroyuki bool "Memory Resource Controller Swap Extension" 777c255a458SAndrew Morton depends on MEMCG && SWAP 778c077719bSKAMEZAWA Hiroyuki help 779c077719bSKAMEZAWA Hiroyuki Add swap management feature to memory resource controller. When you 780c077719bSKAMEZAWA Hiroyuki enable this, you can limit mem+swap usage per cgroup. In other words, 781c077719bSKAMEZAWA Hiroyuki when you disable this, memory resource controller has no cares to 782c077719bSKAMEZAWA Hiroyuki usage of swap...a process can exhaust all of the swap. This extension 783c077719bSKAMEZAWA Hiroyuki is useful when you want to avoid exhaustion swap but this itself 784c077719bSKAMEZAWA Hiroyuki adds more overheads and consumes memory for remembering information. 785c077719bSKAMEZAWA Hiroyuki Especially if you use 32bit system or small memory system, please 786c077719bSKAMEZAWA Hiroyuki be careful about enabling this. When memory resource controller 787c077719bSKAMEZAWA Hiroyuki is disabled by boot option, this will be automatically disabled and 788c077719bSKAMEZAWA Hiroyuki there will be no overhead from this. Even when you set this config=y, 78900a66d29SWANG Cong if boot option "swapaccount=0" is set, swap will not be accounted. 790627991a2SKAMEZAWA Hiroyuki Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page 791627991a2SKAMEZAWA Hiroyuki size is 4096bytes, 512k per 1Gbytes of swap. 792c255a458SAndrew Mortonconfig MEMCG_SWAP_ENABLED 793a42c390cSMichal Hocko bool "Memory Resource Controller Swap Extension enabled by default" 794c255a458SAndrew Morton depends on MEMCG_SWAP 795a42c390cSMichal Hocko default y 796a42c390cSMichal Hocko help 797a42c390cSMichal Hocko Memory Resource Controller Swap Extension comes with its price in 798a42c390cSMichal Hocko a bigger memory consumption. General purpose distribution kernels 79943d547f9SJim Cromie which want to enable the feature but keep it disabled by default 800a42c390cSMichal Hocko and let the user enable it by swapaccount boot command line 801a42c390cSMichal Hocko parameter should have this option unselected. 802a42c390cSMichal Hocko For those who want to have the feature enabled by default should 803a42c390cSMichal Hocko select this option (if, for some reason, they need to disable it 80400a66d29SWANG Cong then swapaccount=0 does the trick). 805c255a458SAndrew Mortonconfig MEMCG_KMEM 806e5671dfaSGlauber Costa bool "Memory Resource Controller Kernel Memory accounting (EXPERIMENTAL)" 807c255a458SAndrew Morton depends on MEMCG && EXPERIMENTAL 808e5671dfaSGlauber Costa default n 809e5671dfaSGlauber Costa help 810e5671dfaSGlauber Costa The Kernel Memory extension for Memory Resource Controller can limit 811e5671dfaSGlauber Costa the amount of memory used by kernel objects in the system. Those are 812e5671dfaSGlauber Costa fundamentally different from the entities handled by the standard 813e5671dfaSGlauber Costa Memory Controller, which are page-based, and can be swapped. Users of 814e5671dfaSGlauber Costa the kmem extension can use it to guarantee that no group of processes 815e5671dfaSGlauber Costa will ever exhaust kernel resources alone. 816c077719bSKAMEZAWA Hiroyuki 8172bc64a20SAneesh Kumar K.Vconfig CGROUP_HUGETLB 8182bc64a20SAneesh Kumar K.V bool "HugeTLB Resource Controller for Control Groups" 8192bc64a20SAneesh Kumar K.V depends on RESOURCE_COUNTERS && HUGETLB_PAGE && EXPERIMENTAL 8202bc64a20SAneesh Kumar K.V default n 8212bc64a20SAneesh Kumar K.V help 8222bc64a20SAneesh Kumar K.V Provides a cgroup Resource Controller for HugeTLB pages. 8232bc64a20SAneesh Kumar K.V When you enable this, you can put a per cgroup limit on HugeTLB usage. 8242bc64a20SAneesh Kumar K.V The limit is enforced during page fault. Since HugeTLB doesn't 8252bc64a20SAneesh Kumar K.V support page reclaim, enforcing the limit at page fault time implies 8262bc64a20SAneesh Kumar K.V that, the application will get SIGBUS signal if it tries to access 8272bc64a20SAneesh Kumar K.V HugeTLB pages beyond its limit. This requires the application to know 8282bc64a20SAneesh Kumar K.V beforehand how much HugeTLB pages it would require for its use. The 8292bc64a20SAneesh Kumar K.V control group is tracked in the third page lru pointer. This means 8302bc64a20SAneesh Kumar K.V that we cannot use the controller with huge page less than 3 pages. 8312bc64a20SAneesh Kumar K.V 832e5d1367fSStephane Eranianconfig CGROUP_PERF 833e5d1367fSStephane Eranian bool "Enable perf_event per-cpu per-container group (cgroup) monitoring" 834e5d1367fSStephane Eranian depends on PERF_EVENTS && CGROUPS 835e5d1367fSStephane Eranian help 836e5d1367fSStephane Eranian This option extends the per-cpu mode to restrict monitoring to 8372d0f2520SLi Zefan threads which belong to the cgroup specified and run on the 838e5d1367fSStephane Eranian designated cpu. 839e5d1367fSStephane Eranian 840e5d1367fSStephane Eranian Say N if unsure. 841e5d1367fSStephane Eranian 8427c941438SDhaval Gianimenuconfig CGROUP_SCHED 8437c941438SDhaval Giani bool "Group CPU scheduler" 8447c941438SDhaval Giani default n 8457c941438SDhaval Giani help 8467c941438SDhaval Giani This feature lets CPU scheduler recognize task groups and control CPU 8477c941438SDhaval Giani bandwidth allocation to such task groups. It uses cgroups to group 8487c941438SDhaval Giani tasks. 8497c941438SDhaval Giani 8507c941438SDhaval Gianiif CGROUP_SCHED 8517c941438SDhaval Gianiconfig FAIR_GROUP_SCHED 8527c941438SDhaval Giani bool "Group scheduling for SCHED_OTHER" 8537c941438SDhaval Giani depends on CGROUP_SCHED 8547c941438SDhaval Giani default CGROUP_SCHED 8557c941438SDhaval Giani 856ab84d31eSPaul Turnerconfig CFS_BANDWIDTH 857ab84d31eSPaul Turner bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED" 858ab84d31eSPaul Turner depends on EXPERIMENTAL 859ab84d31eSPaul Turner depends on FAIR_GROUP_SCHED 860ab84d31eSPaul Turner default n 861ab84d31eSPaul Turner help 862ab84d31eSPaul Turner This option allows users to define CPU bandwidth rates (limits) for 863ab84d31eSPaul Turner tasks running within the fair group scheduler. Groups with no limit 864ab84d31eSPaul Turner set are considered to be unconstrained and will run with no 865ab84d31eSPaul Turner restriction. 866ab84d31eSPaul Turner See tip/Documentation/scheduler/sched-bwc.txt for more information. 867ab84d31eSPaul Turner 8687c941438SDhaval Gianiconfig RT_GROUP_SCHED 8697c941438SDhaval Giani bool "Group scheduling for SCHED_RR/FIFO" 8707c941438SDhaval Giani depends on EXPERIMENTAL 8717c941438SDhaval Giani depends on CGROUP_SCHED 8727c941438SDhaval Giani default n 8737c941438SDhaval Giani help 8747c941438SDhaval Giani This feature lets you explicitly allocate real CPU bandwidth 87532bd7eb5SLi Zefan to task groups. If enabled, it will also make it impossible to 8767c941438SDhaval Giani schedule realtime tasks for non-root users until you allocate 8777c941438SDhaval Giani realtime bandwidth for them. 8787c941438SDhaval Giani See Documentation/scheduler/sched-rt-group.txt for more information. 8797c941438SDhaval Giani 8807c941438SDhaval Gianiendif #CGROUP_SCHED 8817c941438SDhaval Giani 882afc24d49SVivek Goyalconfig BLK_CGROUP 88332e380aeSTejun Heo bool "Block IO controller" 88479ae9c29SDaniel Lezcano depends on BLOCK 885afc24d49SVivek Goyal default n 886afc24d49SVivek Goyal ---help--- 887afc24d49SVivek Goyal Generic block IO controller cgroup interface. This is the common 888afc24d49SVivek Goyal cgroup interface which should be used by various IO controlling 889afc24d49SVivek Goyal policies. 890afc24d49SVivek Goyal 891afc24d49SVivek Goyal Currently, CFQ IO scheduler uses it to recognize task groups and 892afc24d49SVivek Goyal control disk bandwidth allocation (proportional time slice allocation) 893e43473b7SVivek Goyal to such task groups. It is also used by bio throttling logic in 894e43473b7SVivek Goyal block layer to implement upper limit in IO rates on a device. 895afc24d49SVivek Goyal 896afc24d49SVivek Goyal This option only enables generic Block IO controller infrastructure. 897e43473b7SVivek Goyal One needs to also enable actual IO controlling logic/policy. For 89879e2e759SMichael Witten enabling proportional weight division of disk bandwidth in CFQ, set 89979e2e759SMichael Witten CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set 900c5e0591aSMichael Witten CONFIG_BLK_DEV_THROTTLING=y. 901afc24d49SVivek Goyal 902afc24d49SVivek Goyal See Documentation/cgroups/blkio-controller.txt for more information. 903afc24d49SVivek Goyal 904afc24d49SVivek Goyalconfig DEBUG_BLK_CGROUP 905afc24d49SVivek Goyal bool "Enable Block IO controller debugging" 906afc24d49SVivek Goyal depends on BLK_CGROUP 907afc24d49SVivek Goyal default n 908afc24d49SVivek Goyal ---help--- 909afc24d49SVivek Goyal Enable some debugging help. Currently it exports additional stat 910afc24d49SVivek Goyal files in a cgroup which can be useful for debugging. 911afc24d49SVivek Goyal 91223964d2dSLi Zefanendif # CGROUPS 913c077719bSKAMEZAWA Hiroyuki 914067bce1aSCyrill Gorcunovconfig CHECKPOINT_RESTORE 915067bce1aSCyrill Gorcunov bool "Checkpoint/restore support" if EXPERT 916067bce1aSCyrill Gorcunov default n 917067bce1aSCyrill Gorcunov help 918067bce1aSCyrill Gorcunov Enables additional kernel features in a sake of checkpoint/restore. 919067bce1aSCyrill Gorcunov In particular it adds auxiliary prctl codes to setup process text, 920067bce1aSCyrill Gorcunov data and heap segment sizes, and a few additional /proc filesystem 921067bce1aSCyrill Gorcunov entries. 922067bce1aSCyrill Gorcunov 923067bce1aSCyrill Gorcunov If unsure, say N here. 924067bce1aSCyrill Gorcunov 9258dd2a82cSDaniel Lezcanomenuconfig NAMESPACES 9266a108a14SDavid Rientjes bool "Namespaces support" if EXPERT 9276a108a14SDavid Rientjes default !EXPERT 928c5289a69SPavel Emelyanov help 929c5289a69SPavel Emelyanov Provides the way to make tasks work with different objects using 930c5289a69SPavel Emelyanov the same id. For example same IPC id may refer to different objects 931c5289a69SPavel Emelyanov or same user id or pid may refer to different tasks when used in 932c5289a69SPavel Emelyanov different namespaces. 933c5289a69SPavel Emelyanov 9348dd2a82cSDaniel Lezcanoif NAMESPACES 9358dd2a82cSDaniel Lezcano 93658bfdd6dSPavel Emelyanovconfig UTS_NS 93758bfdd6dSPavel Emelyanov bool "UTS namespace" 93817a6d441SDaniel Lezcano default y 93958bfdd6dSPavel Emelyanov help 94058bfdd6dSPavel Emelyanov In this namespace tasks see different info provided with the 94158bfdd6dSPavel Emelyanov uname() system call 94258bfdd6dSPavel Emelyanov 943ae5e1b22SPavel Emelyanovconfig IPC_NS 944ae5e1b22SPavel Emelyanov bool "IPC namespace" 9458dd2a82cSDaniel Lezcano depends on (SYSVIPC || POSIX_MQUEUE) 94617a6d441SDaniel Lezcano default y 947ae5e1b22SPavel Emelyanov help 948ae5e1b22SPavel Emelyanov In this namespace tasks work with IPC ids which correspond to 949614b84cfSSerge E. Hallyn different IPC objects in different namespaces. 950ae5e1b22SPavel Emelyanov 951aee16ce7SPavel Emelyanovconfig USER_NS 952aee16ce7SPavel Emelyanov bool "User namespace (EXPERIMENTAL)" 9538dd2a82cSDaniel Lezcano depends on EXPERIMENTAL 954e1c972b6SEric W. Biederman depends on UIDGID_CONVERTED 9555673a94cSEric W. Biederman select UIDGID_STRICT_TYPE_CHECKS 956e1c972b6SEric W. Biederman 9575673a94cSEric W. Biederman default n 958aee16ce7SPavel Emelyanov help 959aee16ce7SPavel Emelyanov This allows containers, i.e. vservers, to use user namespaces 960aee16ce7SPavel Emelyanov to provide different user info for different servers. 961aee16ce7SPavel Emelyanov If unsure, say N. 962aee16ce7SPavel Emelyanov 96374bd59bbSPavel Emelyanovconfig PID_NS 9649bd38c2cSDaniel Lezcano bool "PID Namespaces" 96517a6d441SDaniel Lezcano default y 96674bd59bbSPavel Emelyanov help 96712d2b8f9SHeikki Orsila Support process id namespaces. This allows having multiple 968692105b8SMatt LaPlante processes with the same pid as long as they are in different 96974bd59bbSPavel Emelyanov pid namespaces. This is a building block of containers. 97074bd59bbSPavel Emelyanov 971d6eb633fSMatt Helsleyconfig NET_NS 972d6eb633fSMatt Helsley bool "Network namespace" 9738dd2a82cSDaniel Lezcano depends on NET 97417a6d441SDaniel Lezcano default y 975d6eb633fSMatt Helsley help 976d6eb633fSMatt Helsley Allow user space to create what appear to be multiple instances 977d6eb633fSMatt Helsley of the network stack. 978d6eb633fSMatt Helsley 9798dd2a82cSDaniel Lezcanoendif # NAMESPACES 9808dd2a82cSDaniel Lezcano 981e1c972b6SEric W. Biedermanconfig UIDGID_CONVERTED 982e1c972b6SEric W. Biederman # True if all of the selected software conmponents are known 983e1c972b6SEric W. Biederman # to have uid_t and gid_t converted to kuid_t and kgid_t 984e1c972b6SEric W. Biederman # where appropriate and are otherwise safe to use with 985e1c972b6SEric W. Biederman # the user namespace. 986e1c972b6SEric W. Biederman bool 987e1c972b6SEric W. Biederman default y 988e1c972b6SEric W. Biederman 989e1c972b6SEric W. Biederman # Networking 990e1c972b6SEric W. Biederman depends on NET_9P = n 991e1c972b6SEric W. Biederman 992e1c972b6SEric W. Biederman # Filesystems 993e1c972b6SEric W. Biederman depends on 9P_FS = n 994e1c972b6SEric W. Biederman depends on AFS_FS = n 995e1c972b6SEric W. Biederman depends on AUTOFS4_FS = n 996e1c972b6SEric W. Biederman depends on CEPH_FS = n 997e1c972b6SEric W. Biederman depends on CIFS = n 998e1c972b6SEric W. Biederman depends on CODA_FS = n 999e1c972b6SEric W. Biederman depends on FUSE_FS = n 1000e1c972b6SEric W. Biederman depends on GFS2_FS = n 1001e1c972b6SEric W. Biederman depends on NCP_FS = n 1002e1c972b6SEric W. Biederman depends on NFSD = n 1003e1c972b6SEric W. Biederman depends on NFS_FS = n 1004e1c972b6SEric W. Biederman depends on OCFS2_FS = n 1005e1c972b6SEric W. Biederman depends on XFS_FS = n 1006e1c972b6SEric W. Biederman 10075673a94cSEric W. Biedermanconfig UIDGID_STRICT_TYPE_CHECKS 10085673a94cSEric W. Biederman bool "Require conversions between uid/gids and their internal representation" 1009e1c972b6SEric W. Biederman depends on UIDGID_CONVERTED 10105673a94cSEric W. Biederman default n 10115673a94cSEric W. Biederman help 10125673a94cSEric W. Biederman While the nececessary conversions are being added to all subsystems this option allows 10135673a94cSEric W. Biederman the code to continue to build for unconverted subsystems. 10145673a94cSEric W. Biederman 10155673a94cSEric W. Biederman Say Y here if you want the strict type checking enabled 10165673a94cSEric W. Biederman 10175091faa4SMike Galbraithconfig SCHED_AUTOGROUP 10185091faa4SMike Galbraith bool "Automatic process group scheduling" 10195091faa4SMike Galbraith select EVENTFD 10205091faa4SMike Galbraith select CGROUPS 10215091faa4SMike Galbraith select CGROUP_SCHED 10225091faa4SMike Galbraith select FAIR_GROUP_SCHED 10235091faa4SMike Galbraith help 10245091faa4SMike Galbraith This option optimizes the scheduler for common desktop workloads by 10255091faa4SMike Galbraith automatically creating and populating task groups. This separation 10265091faa4SMike Galbraith of workloads isolates aggressive CPU burners (like build jobs) from 10275091faa4SMike Galbraith desktop applications. Task group autogeneration is currently based 10285091faa4SMike Galbraith upon task session. 10295091faa4SMike Galbraith 10307af37becSDaniel Lezcanoconfig MM_OWNER 10317af37becSDaniel Lezcano bool 10327af37becSDaniel Lezcano 10337af37becSDaniel Lezcanoconfig SYSFS_DEPRECATED 10345d6a4ea5SFerenc Wagner bool "Enable deprecated sysfs features to support old userspace tools" 10357af37becSDaniel Lezcano depends on SYSFS 10367af37becSDaniel Lezcano default n 10377af37becSDaniel Lezcano help 10387af37becSDaniel Lezcano This option adds code that switches the layout of the "block" class 10397af37becSDaniel Lezcano devices, to not show up in /sys/class/block/, but only in 10407af37becSDaniel Lezcano /sys/block/. 10417af37becSDaniel Lezcano 10427af37becSDaniel Lezcano This switch is only active when the sysfs.deprecated=1 boot option is 10437af37becSDaniel Lezcano passed or the SYSFS_DEPRECATED_V2 option is set. 10447af37becSDaniel Lezcano 10457af37becSDaniel Lezcano This option allows new kernels to run on old distributions and tools, 10467af37becSDaniel Lezcano which might get confused by /sys/class/block/. Since 2007/2008 all 10477af37becSDaniel Lezcano major distributions and tools handle this just fine. 10487af37becSDaniel Lezcano 10497af37becSDaniel Lezcano Recent distributions and userspace tools after 2009/2010 depend on 10507af37becSDaniel Lezcano the existence of /sys/class/block/, and will not work with this 10517af37becSDaniel Lezcano option enabled. 10527af37becSDaniel Lezcano 10537af37becSDaniel Lezcano Only if you are using a new kernel on an old distribution, you might 10547af37becSDaniel Lezcano need to say Y here. 10557af37becSDaniel Lezcano 10567af37becSDaniel Lezcanoconfig SYSFS_DEPRECATED_V2 10575d6a4ea5SFerenc Wagner bool "Enable deprecated sysfs features by default" 10587af37becSDaniel Lezcano default n 10597af37becSDaniel Lezcano depends on SYSFS 10607af37becSDaniel Lezcano depends on SYSFS_DEPRECATED 10617af37becSDaniel Lezcano help 10627af37becSDaniel Lezcano Enable deprecated sysfs by default. 10637af37becSDaniel Lezcano 10647af37becSDaniel Lezcano See the CONFIG_SYSFS_DEPRECATED option for more details about this 10657af37becSDaniel Lezcano option. 10667af37becSDaniel Lezcano 10677af37becSDaniel Lezcano Only if you are using a new kernel on an old distribution, you might 10687af37becSDaniel Lezcano need to say Y here. Even then, odds are you would not need it 10697af37becSDaniel Lezcano enabled, you can always pass the boot option if absolutely necessary. 10707af37becSDaniel Lezcano 10717af37becSDaniel Lezcanoconfig RELAY 10727af37becSDaniel Lezcano bool "Kernel->user space relay support (formerly relayfs)" 10737af37becSDaniel Lezcano help 10747af37becSDaniel Lezcano This option enables support for relay interface support in 10757af37becSDaniel Lezcano certain file systems (such as debugfs). 10767af37becSDaniel Lezcano It is designed to provide an efficient mechanism for tools and 10777af37becSDaniel Lezcano facilities to relay large amounts of data from kernel space to 10787af37becSDaniel Lezcano user space. 10797af37becSDaniel Lezcano 10807af37becSDaniel Lezcano If unsure, say N. 10817af37becSDaniel Lezcano 1082f991633dSDimitri Gorokhovikconfig BLK_DEV_INITRD 1083f991633dSDimitri Gorokhovik bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support" 1084f991633dSDimitri Gorokhovik depends on BROKEN || !FRV 1085f991633dSDimitri Gorokhovik help 1086f991633dSDimitri Gorokhovik The initial RAM filesystem is a ramfs which is loaded by the 1087f991633dSDimitri Gorokhovik boot loader (loadlin or lilo) and that is mounted as root 1088f991633dSDimitri Gorokhovik before the normal boot procedure. It is typically used to 1089f991633dSDimitri Gorokhovik load modules needed to mount the "real" root file system, 1090f991633dSDimitri Gorokhovik etc. See <file:Documentation/initrd.txt> for details. 1091f991633dSDimitri Gorokhovik 1092f991633dSDimitri Gorokhovik If RAM disk support (BLK_DEV_RAM) is also included, this 1093f991633dSDimitri Gorokhovik also enables initial RAM disk (initrd) support and adds 1094f991633dSDimitri Gorokhovik 15 Kbytes (more on some other architectures) to the kernel size. 1095f991633dSDimitri Gorokhovik 1096f991633dSDimitri Gorokhovik If unsure say Y. 1097f991633dSDimitri Gorokhovik 1098c33df4eaSJean-Paul Samanif BLK_DEV_INITRD 1099c33df4eaSJean-Paul Saman 1100dbec4866SSam Ravnborgsource "usr/Kconfig" 1101dbec4866SSam Ravnborg 1102c33df4eaSJean-Paul Samanendif 1103c33df4eaSJean-Paul Saman 1104c45b4f1fSLinus Torvaldsconfig CC_OPTIMIZE_FOR_SIZE 110596fffeb4SIngo Molnar bool "Optimize for size" 1106c45b4f1fSLinus Torvalds help 1107c45b4f1fSLinus Torvalds Enabling this option will pass "-Os" instead of "-O2" to gcc 1108c45b4f1fSLinus Torvalds resulting in a smaller kernel. 1109c45b4f1fSLinus Torvalds 1110775a7229Sjkacur If unsure, say Y. 1111c45b4f1fSLinus Torvalds 11120847062aSRandy Dunlapconfig SYSCTL 11130847062aSRandy Dunlap bool 11140847062aSRandy Dunlap 1115b943c460SRandy Dunlapconfig ANON_INODES 1116b943c460SRandy Dunlap bool 1117b943c460SRandy Dunlap 11186a108a14SDavid Rientjesmenuconfig EXPERT 11196a108a14SDavid Rientjes bool "Configure standard kernel features (expert users)" 1120f505c553SJosh Triplett # Unhide debug options, to make the on-by-default options visible 1121f505c553SJosh Triplett select DEBUG_KERNEL 11221da177e4SLinus Torvalds help 11231da177e4SLinus Torvalds This option allows certain base kernel options and settings 11241da177e4SLinus Torvalds to be disabled or tweaked. This is for specialized 11251da177e4SLinus Torvalds environments which can tolerate a "non-standard" kernel. 11261da177e4SLinus Torvalds Only use this if you really know what you are doing. 11271da177e4SLinus Torvalds 1128*af1839ebSCatalin Marinasconfig HAVE_UID16 1129*af1839ebSCatalin Marinas bool 1130*af1839ebSCatalin Marinas 1131ae81f9e3SChuck Ebbertconfig UID16 11326a108a14SDavid Rientjes bool "Enable 16-bit UID system calls" if EXPERT 1133*af1839ebSCatalin Marinas depends on HAVE_UID16 1134ae81f9e3SChuck Ebbert default y 1135ae81f9e3SChuck Ebbert help 1136ae81f9e3SChuck Ebbert This enables the legacy 16-bit UID syscall wrappers. 1137ae81f9e3SChuck Ebbert 1138b89a8171SEric W. Biedermanconfig SYSCTL_SYSCALL 11396a108a14SDavid Rientjes bool "Sysctl syscall support" if EXPERT 114026a7034bSEric W. Biederman depends on PROC_SYSCTL 1141c736de60SWANG Cong default n 1142b89a8171SEric W. Biederman select SYSCTL 1143b89a8171SEric W. Biederman ---help--- 114413bb7e37SEric W. Biederman sys_sysctl uses binary paths that have been found challenging 114513bb7e37SEric W. Biederman to properly maintain and use. The interface in /proc/sys 114613bb7e37SEric W. Biederman using paths with ascii names is now the primary path to this 114713bb7e37SEric W. Biederman information. 1148b89a8171SEric W. Biederman 114913bb7e37SEric W. Biederman Almost nothing using the binary sysctl interface so if you are 115013bb7e37SEric W. Biederman trying to save some space it is probably safe to disable this, 115113bb7e37SEric W. Biederman making your kernel marginally smaller. 1152b89a8171SEric W. Biederman 1153c736de60SWANG Cong If unsure say N here. 1154ae81f9e3SChuck Ebbert 11551da177e4SLinus Torvaldsconfig KALLSYMS 11566a108a14SDavid Rientjes bool "Load all symbols for debugging/ksymoops" if EXPERT 11571da177e4SLinus Torvalds default y 11581da177e4SLinus Torvalds help 11591da177e4SLinus Torvalds Say Y here to let the kernel print out symbolic crash information and 11601da177e4SLinus Torvalds symbolic stack backtraces. This increases the size of the kernel 11611da177e4SLinus Torvalds somewhat, as all symbols have to be loaded into the kernel image. 11621da177e4SLinus Torvalds 11631da177e4SLinus Torvaldsconfig KALLSYMS_ALL 11641da177e4SLinus Torvalds bool "Include all symbols in kallsyms" 11651da177e4SLinus Torvalds depends on DEBUG_KERNEL && KALLSYMS 11661da177e4SLinus Torvalds help 116771a83ec7SArtem Bityutskiy Normally kallsyms only contains the symbols of functions for nicer 116871a83ec7SArtem Bityutskiy OOPS messages and backtraces (i.e., symbols from the text and inittext 116971a83ec7SArtem Bityutskiy sections). This is sufficient for most cases. And only in very rare 117071a83ec7SArtem Bityutskiy cases (e.g., when a debugger is used) all symbols are required (e.g., 117171a83ec7SArtem Bityutskiy names of variables from the data sections, etc). 11721da177e4SLinus Torvalds 117371a83ec7SArtem Bityutskiy This option makes sure that all symbols are loaded into the kernel 117471a83ec7SArtem Bityutskiy image (i.e., symbols from all sections) in cost of increased kernel 117571a83ec7SArtem Bityutskiy size (depending on the kernel configuration, it may be 300KiB or 117671a83ec7SArtem Bityutskiy something like this). 11771da177e4SLinus Torvalds 117871a83ec7SArtem Bityutskiy Say N unless you really need all symbols. 1179d59745ceSMatt Mackall 1180712f47ceSGreg Kroah-Hartmanconfig HOTPLUG 118145f035abSGreg Kroah-Hartman def_bool y 1182712f47ceSGreg Kroah-Hartman 1183d59745ceSMatt Mackallconfig PRINTK 1184d59745ceSMatt Mackall default y 11856a108a14SDavid Rientjes bool "Enable support for printk" if EXPERT 1186d59745ceSMatt Mackall help 1187d59745ceSMatt Mackall This option enables normal printk support. Removing it 1188d59745ceSMatt Mackall eliminates most of the message strings from the kernel image 1189d59745ceSMatt Mackall and makes the kernel more or less silent. As this makes it 1190d59745ceSMatt Mackall very difficult to diagnose system problems, saying N here is 1191d59745ceSMatt Mackall strongly discouraged. 1192d59745ceSMatt Mackall 1193c8538a7aSMatt Mackallconfig BUG 11946a108a14SDavid Rientjes bool "BUG() support" if EXPERT 1195c8538a7aSMatt Mackall default y 1196c8538a7aSMatt Mackall help 1197c8538a7aSMatt Mackall Disabling this option eliminates support for BUG and WARN, reducing 1198c8538a7aSMatt Mackall the size of your kernel image and potentially quietly ignoring 1199c8538a7aSMatt Mackall numerous fatal conditions. You should only consider disabling this 1200c8538a7aSMatt Mackall option for embedded systems with no facilities for reporting errors. 1201c8538a7aSMatt Mackall Just say Y. 1202c8538a7aSMatt Mackall 1203708e9a79SMatt Mackallconfig ELF_CORE 1204046d662fSAlex Kelly depends on COREDUMP 1205708e9a79SMatt Mackall default y 12066a108a14SDavid Rientjes bool "Enable ELF core dumps" if EXPERT 1207708e9a79SMatt Mackall help 1208708e9a79SMatt Mackall Enable support for generating core dumps. Disabling saves about 4k. 1209708e9a79SMatt Mackall 12108761f1abSRalf Baechle 1211e5e1d3cbSStas Sergeevconfig PCSPKR_PLATFORM 12126a108a14SDavid Rientjes bool "Enable PC-Speaker support" if EXPERT 12138761f1abSRalf Baechle depends on HAVE_PCSPKR_PLATFORM 121415f304b6SRalf Baechle select I8253_LOCK 1215e5e1d3cbSStas Sergeev default y 1216e5e1d3cbSStas Sergeev help 1217e5e1d3cbSStas Sergeev This option allows to disable the internal PC-Speaker 1218e5e1d3cbSStas Sergeev support, saving some memory. 1219e5e1d3cbSStas Sergeev 12208761f1abSRalf Baechleconfig HAVE_PCSPKR_PLATFORM 12218761f1abSRalf Baechle bool 12228761f1abSRalf Baechle 12231da177e4SLinus Torvaldsconfig BASE_FULL 12241da177e4SLinus Torvalds default y 12256a108a14SDavid Rientjes bool "Enable full-sized data structures for core" if EXPERT 12261da177e4SLinus Torvalds help 12271da177e4SLinus Torvalds Disabling this option reduces the size of miscellaneous core 12281da177e4SLinus Torvalds kernel data structures. This saves memory on small machines, 12291da177e4SLinus Torvalds but may reduce performance. 12301da177e4SLinus Torvalds 12311da177e4SLinus Torvaldsconfig FUTEX 12326a108a14SDavid Rientjes bool "Enable futex support" if EXPERT 12331da177e4SLinus Torvalds default y 123423f78d4aSIngo Molnar select RT_MUTEXES 12351da177e4SLinus Torvalds help 12361da177e4SLinus Torvalds Disabling this option will cause the kernel to be built without 12371da177e4SLinus Torvalds support for "fast userspace mutexes". The resulting kernel may not 12381da177e4SLinus Torvalds run glibc-based applications correctly. 12391da177e4SLinus Torvalds 12401da177e4SLinus Torvaldsconfig EPOLL 12416a108a14SDavid Rientjes bool "Enable eventpoll support" if EXPERT 12421da177e4SLinus Torvalds default y 1243448e3ceeSAdrian Bunk select ANON_INODES 12441da177e4SLinus Torvalds help 12451da177e4SLinus Torvalds Disabling this option will cause the kernel to be built without 12461da177e4SLinus Torvalds support for epoll family of system calls. 12471da177e4SLinus Torvalds 1248fba2afaaSDavide Libenziconfig SIGNALFD 12496a108a14SDavid Rientjes bool "Enable signalfd() system call" if EXPERT 1250448e3ceeSAdrian Bunk select ANON_INODES 1251fba2afaaSDavide Libenzi default y 1252fba2afaaSDavide Libenzi help 1253fba2afaaSDavide Libenzi Enable the signalfd() system call that allows to receive signals 1254fba2afaaSDavide Libenzi on a file descriptor. 1255fba2afaaSDavide Libenzi 1256fba2afaaSDavide Libenzi If unsure, say Y. 1257fba2afaaSDavide Libenzi 1258b215e283SDavide Libenziconfig TIMERFD 12596a108a14SDavid Rientjes bool "Enable timerfd() system call" if EXPERT 1260448e3ceeSAdrian Bunk select ANON_INODES 1261b215e283SDavide Libenzi default y 1262b215e283SDavide Libenzi help 1263b215e283SDavide Libenzi Enable the timerfd() system call that allows to receive timer 1264b215e283SDavide Libenzi events on a file descriptor. 1265b215e283SDavide Libenzi 1266b215e283SDavide Libenzi If unsure, say Y. 1267b215e283SDavide Libenzi 1268e1ad7468SDavide Libenziconfig EVENTFD 12696a108a14SDavid Rientjes bool "Enable eventfd() system call" if EXPERT 1270448e3ceeSAdrian Bunk select ANON_INODES 1271e1ad7468SDavide Libenzi default y 1272e1ad7468SDavide Libenzi help 1273e1ad7468SDavide Libenzi Enable the eventfd() system call that allows to receive both 1274e1ad7468SDavide Libenzi kernel notification (ie. KAIO) or userspace notifications. 1275e1ad7468SDavide Libenzi 1276e1ad7468SDavide Libenzi If unsure, say Y. 1277e1ad7468SDavide Libenzi 12781da177e4SLinus Torvaldsconfig SHMEM 12796a108a14SDavid Rientjes bool "Use full shmem filesystem" if EXPERT 12801da177e4SLinus Torvalds default y 12811da177e4SLinus Torvalds depends on MMU 12821da177e4SLinus Torvalds help 12831da177e4SLinus Torvalds The shmem is an internal filesystem used to manage shared memory. 12841da177e4SLinus Torvalds It is backed by swap and manages resource limits. It is also exported 12851da177e4SLinus Torvalds to userspace as tmpfs if TMPFS is enabled. Disabling this 12861da177e4SLinus Torvalds option replaces shmem and tmpfs with the much simpler ramfs code, 12871da177e4SLinus Torvalds which may be appropriate on small systems without swap. 12881da177e4SLinus Torvalds 1289ebf3f09cSThomas Petazzoniconfig AIO 12906a108a14SDavid Rientjes bool "Enable AIO support" if EXPERT 1291ebf3f09cSThomas Petazzoni default y 1292ebf3f09cSThomas Petazzoni help 1293ebf3f09cSThomas Petazzoni This option enables POSIX asynchronous I/O which may by used 1294ebf3f09cSThomas Petazzoni by some high performance threaded applications. Disabling 1295ebf3f09cSThomas Petazzoni this option saves about 7k. 1296ebf3f09cSThomas Petazzoni 12976befe5f6SRandy Dunlapconfig EMBEDDED 12986befe5f6SRandy Dunlap bool "Embedded system" 12996befe5f6SRandy Dunlap select EXPERT 13006befe5f6SRandy Dunlap help 13016befe5f6SRandy Dunlap This option should be enabled if compiling the kernel for 13026befe5f6SRandy Dunlap an embedded system so certain expert options are available 13036befe5f6SRandy Dunlap for configuration. 13046befe5f6SRandy Dunlap 1305cdd6c482SIngo Molnarconfig HAVE_PERF_EVENTS 13060793a61dSThomas Gleixner bool 1307018df72dSMike Frysinger help 1308018df72dSMike Frysinger See tools/perf/design.txt for details. 13090793a61dSThomas Gleixner 1310906010b2SPeter Zijlstraconfig PERF_USE_VMALLOC 1311906010b2SPeter Zijlstra bool 1312906010b2SPeter Zijlstra help 1313906010b2SPeter Zijlstra See tools/perf/design.txt for details 1314906010b2SPeter Zijlstra 131557c0c15bSIngo Molnarmenu "Kernel Performance Events And Counters" 13160793a61dSThomas Gleixner 1317cdd6c482SIngo Molnarconfig PERF_EVENTS 131857c0c15bSIngo Molnar bool "Kernel performance events and counters" 1319392d65a9SRobert Richter default y if PROFILING 1320cdd6c482SIngo Molnar depends on HAVE_PERF_EVENTS 13214c59e467SIngo Molnar select ANON_INODES 1322e360adbeSPeter Zijlstra select IRQ_WORK 13230793a61dSThomas Gleixner help 132457c0c15bSIngo Molnar Enable kernel support for various performance events provided 132557c0c15bSIngo Molnar by software and hardware. 13260793a61dSThomas Gleixner 1327dd77038dSThadeu Lima de Souza Cascardo Software events are supported either built-in or via the 132857c0c15bSIngo Molnar use of generic tracepoints. 132957c0c15bSIngo Molnar 133057c0c15bSIngo Molnar Most modern CPUs support performance events via performance 133157c0c15bSIngo Molnar counter registers. These registers count the number of certain 13320793a61dSThomas Gleixner types of hw events: such as instructions executed, cachemisses 13330793a61dSThomas Gleixner suffered, or branches mis-predicted - without slowing down the 13340793a61dSThomas Gleixner kernel or applications. These registers can also trigger interrupts 13350793a61dSThomas Gleixner when a threshold number of events have passed - and can thus be 13360793a61dSThomas Gleixner used to profile the code that runs on that CPU. 13370793a61dSThomas Gleixner 133857c0c15bSIngo Molnar The Linux Performance Event subsystem provides an abstraction of 1339dd77038dSThadeu Lima de Souza Cascardo these software and hardware event capabilities, available via a 134057c0c15bSIngo Molnar system call and used by the "perf" utility in tools/perf/. It 13410793a61dSThomas Gleixner provides per task and per CPU counters, and it provides event 13420793a61dSThomas Gleixner capabilities on top of those. 13430793a61dSThomas Gleixner 13440793a61dSThomas Gleixner Say Y if unsure. 13450793a61dSThomas Gleixner 1346906010b2SPeter Zijlstraconfig DEBUG_PERF_USE_VMALLOC 1347906010b2SPeter Zijlstra default n 1348906010b2SPeter Zijlstra bool "Debug: use vmalloc to back perf mmap() buffers" 1349906010b2SPeter Zijlstra depends on PERF_EVENTS && DEBUG_KERNEL 1350906010b2SPeter Zijlstra select PERF_USE_VMALLOC 1351906010b2SPeter Zijlstra help 1352906010b2SPeter Zijlstra Use vmalloc memory to back perf mmap() buffers. 1353906010b2SPeter Zijlstra 1354906010b2SPeter Zijlstra Mostly useful for debugging the vmalloc code on platforms 1355906010b2SPeter Zijlstra that don't require it. 1356906010b2SPeter Zijlstra 1357906010b2SPeter Zijlstra Say N if unsure. 1358906010b2SPeter Zijlstra 13590793a61dSThomas Gleixnerendmenu 13600793a61dSThomas Gleixner 1361f8891e5eSChristoph Lameterconfig VM_EVENT_COUNTERS 1362f8891e5eSChristoph Lameter default y 13636a108a14SDavid Rientjes bool "Enable VM event counters for /proc/vmstat" if EXPERT 1364f8891e5eSChristoph Lameter help 13652aea4fb6SPaul Jackson VM event counters are needed for event counts to be shown. 13662aea4fb6SPaul Jackson This option allows the disabling of the VM event counters 13676a108a14SDavid Rientjes on EXPERT systems. /proc/vmstat will only show page counts 13682aea4fb6SPaul Jackson if VM event counters are disabled. 1369f8891e5eSChristoph Lameter 13703d137310SThomas Petazzoniconfig PCI_QUIRKS 13713d137310SThomas Petazzoni default y 13726a108a14SDavid Rientjes bool "Enable PCI quirk workarounds" if EXPERT 137361cfc7e4SGeert Uytterhoeven depends on PCI 13743d137310SThomas Petazzoni help 13753d137310SThomas Petazzoni This enables workarounds for various PCI chipset 13763d137310SThomas Petazzoni bugs/quirks. Disable this only if your target machine is 13773d137310SThomas Petazzoni unaffected by PCI quirks. 13783d137310SThomas Petazzoni 137941ecc55bSChristoph Lameterconfig SLUB_DEBUG 138041ecc55bSChristoph Lameter default y 13816a108a14SDavid Rientjes bool "Enable SLUB debugging support" if EXPERT 1382f6acb635SChristoph Lameter depends on SLUB && SYSFS 138341ecc55bSChristoph Lameter help 138441ecc55bSChristoph Lameter SLUB has extensive debug support features. Disabling these can 138541ecc55bSChristoph Lameter result in significant savings in code size. This also disables 138641ecc55bSChristoph Lameter SLUB sysfs support. /sys/slab will not exist and there will be 138741ecc55bSChristoph Lameter no support for cache validation etc. 138841ecc55bSChristoph Lameter 1389b943c460SRandy Dunlapconfig COMPAT_BRK 1390b943c460SRandy Dunlap bool "Disable heap randomization" 1391b943c460SRandy Dunlap default y 1392b943c460SRandy Dunlap help 1393b943c460SRandy Dunlap Randomizing heap placement makes heap exploits harder, but it 1394b943c460SRandy Dunlap also breaks ancient binaries (including anything libc5 based). 1395b943c460SRandy Dunlap This option changes the bootup default to heap randomization 1396692105b8SMatt LaPlante disabled, and can be overridden at runtime by setting 1397b943c460SRandy Dunlap /proc/sys/kernel/randomize_va_space to 2. 1398b943c460SRandy Dunlap 1399b943c460SRandy Dunlap On non-ancient distros (post-2000 ones) N is usually a safe choice. 1400b943c460SRandy Dunlap 140181819f0fSChristoph Lameterchoice 140281819f0fSChristoph Lameter prompt "Choose SLAB allocator" 1403a0acd820SChristoph Lameter default SLUB 140481819f0fSChristoph Lameter help 140581819f0fSChristoph Lameter This option allows to select a slab allocator. 140681819f0fSChristoph Lameter 140781819f0fSChristoph Lameterconfig SLAB 140881819f0fSChristoph Lameter bool "SLAB" 140981819f0fSChristoph Lameter help 141081819f0fSChristoph Lameter The regular slab allocator that is established and known to work 141134013886SChristoph Lameter well in all environments. It organizes cache hot objects in 141202f56210SSimon Arlott per cpu and per node queues. 141381819f0fSChristoph Lameter 141481819f0fSChristoph Lameterconfig SLUB 141581819f0fSChristoph Lameter bool "SLUB (Unqueued Allocator)" 141681819f0fSChristoph Lameter help 141781819f0fSChristoph Lameter SLUB is a slab allocator that minimizes cache line usage 141881819f0fSChristoph Lameter instead of managing queues of cached objects (SLAB approach). 141981819f0fSChristoph Lameter Per cpu caching is realized using slabs of objects instead 142081819f0fSChristoph Lameter of queues of objects. SLUB can use memory efficiently 142102f56210SSimon Arlott and has enhanced diagnostics. SLUB is the default choice for 142202f56210SSimon Arlott a slab allocator. 142381819f0fSChristoph Lameter 142481819f0fSChristoph Lameterconfig SLOB 14256a108a14SDavid Rientjes depends on EXPERT 142681819f0fSChristoph Lameter bool "SLOB (Simple Allocator)" 142781819f0fSChristoph Lameter help 142837291458SMatt Mackall SLOB replaces the stock allocator with a drastically simpler 142937291458SMatt Mackall allocator. SLOB is generally more space efficient but 143037291458SMatt Mackall does not perform as well on large systems. 143181819f0fSChristoph Lameter 143281819f0fSChristoph Lameterendchoice 143381819f0fSChristoph Lameter 1434ea637639SJie Zhangconfig MMAP_ALLOW_UNINITIALIZED 1435ea637639SJie Zhang bool "Allow mmapped anonymous memory to be uninitialized" 14366a108a14SDavid Rientjes depends on EXPERT && !MMU 1437ea637639SJie Zhang default n 1438ea637639SJie Zhang help 1439ea637639SJie Zhang Normally, and according to the Linux spec, anonymous memory obtained 1440ea637639SJie Zhang from mmap() has it's contents cleared before it is passed to 1441ea637639SJie Zhang userspace. Enabling this config option allows you to request that 1442ea637639SJie Zhang mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 1443ea637639SJie Zhang providing a huge performance boost. If this option is not enabled, 1444ea637639SJie Zhang then the flag will be ignored. 1445ea637639SJie Zhang 1446ea637639SJie Zhang This is taken advantage of by uClibc's malloc(), and also by 1447ea637639SJie Zhang ELF-FDPIC binfmt's brk and stack allocator. 1448ea637639SJie Zhang 1449ea637639SJie Zhang Because of the obvious security issues, this option should only be 1450ea637639SJie Zhang enabled on embedded devices where you control what is run in 1451ea637639SJie Zhang userspace. Since that isn't generally a problem on no-MMU systems, 1452ea637639SJie Zhang it is normally safe to say Y here. 1453ea637639SJie Zhang 1454ea637639SJie Zhang See Documentation/nommu-mmap.txt for more information. 1455ea637639SJie Zhang 1456125e5645SMathieu Desnoyersconfig PROFILING 1457b309a294SRobert Richter bool "Profiling support" 1458125e5645SMathieu Desnoyers help 1459125e5645SMathieu Desnoyers Say Y here to enable the extended profiling support mechanisms used 1460125e5645SMathieu Desnoyers by profilers such as OProfile. 1461125e5645SMathieu Desnoyers 14625f87f112SIngo Molnar# 14635f87f112SIngo Molnar# Place an empty function call at each tracepoint site. Can be 14645f87f112SIngo Molnar# dynamically changed for a probe function. 14655f87f112SIngo Molnar# 146697e1c18eSMathieu Desnoyersconfig TRACEPOINTS 14675f87f112SIngo Molnar bool 146897e1c18eSMathieu Desnoyers 1469fb32e03fSMathieu Desnoyerssource "arch/Kconfig" 1470fb32e03fSMathieu Desnoyers 14711da177e4SLinus Torvaldsendmenu # General setup 14721da177e4SLinus Torvalds 1473ee7e5516SDmitry Baryshkovconfig HAVE_GENERIC_DMA_COHERENT 1474ee7e5516SDmitry Baryshkov bool 1475ee7e5516SDmitry Baryshkov default n 1476ee7e5516SDmitry Baryshkov 1477158a9624SLinus Torvaldsconfig SLABINFO 1478158a9624SLinus Torvalds bool 1479158a9624SLinus Torvalds depends on PROC_FS 14800f389ec6SChristoph Lameter depends on SLAB || SLUB_DEBUG 1481158a9624SLinus Torvalds default y 1482158a9624SLinus Torvalds 1483ae81f9e3SChuck Ebbertconfig RT_MUTEXES 1484ae81f9e3SChuck Ebbert boolean 1485ae81f9e3SChuck Ebbert 14861da177e4SLinus Torvaldsconfig BASE_SMALL 14871da177e4SLinus Torvalds int 14881da177e4SLinus Torvalds default 0 if BASE_FULL 14891da177e4SLinus Torvalds default 1 if !BASE_FULL 14901da177e4SLinus Torvalds 149166da5733SJan Engelhardtmenuconfig MODULES 14921da177e4SLinus Torvalds bool "Enable loadable module support" 14931da177e4SLinus Torvalds help 14941da177e4SLinus Torvalds Kernel modules are small pieces of compiled code which can 14951da177e4SLinus Torvalds be inserted in the running kernel, rather than being 14961da177e4SLinus Torvalds permanently built into the kernel. You use the "modprobe" 14971da177e4SLinus Torvalds tool to add (and sometimes remove) them. If you say Y here, 14981da177e4SLinus Torvalds many parts of the kernel can be built as modules (by 14991da177e4SLinus Torvalds answering M instead of Y where indicated): this is most 15001da177e4SLinus Torvalds useful for infrequently used options which are not required 15011da177e4SLinus Torvalds for booting. For more information, see the man pages for 15021da177e4SLinus Torvalds modprobe, lsmod, modinfo, insmod and rmmod. 15031da177e4SLinus Torvalds 15041da177e4SLinus Torvalds If you say Y here, you will need to run "make 15051da177e4SLinus Torvalds modules_install" to put the modules under /lib/modules/ 15061da177e4SLinus Torvalds where modprobe can find them (you may need to be root to do 15071da177e4SLinus Torvalds this). 15081da177e4SLinus Torvalds 15091da177e4SLinus Torvalds If unsure, say Y. 15101da177e4SLinus Torvalds 15110b0de144SRobert P. J. Dayif MODULES 15120b0de144SRobert P. J. Day 1513826e4506SLinus Torvaldsconfig MODULE_FORCE_LOAD 1514826e4506SLinus Torvalds bool "Forced module loading" 1515826e4506SLinus Torvalds default n 1516826e4506SLinus Torvalds help 151791e37a79SRusty Russell Allow loading of modules without version information (ie. modprobe 151891e37a79SRusty Russell --force). Forced module loading sets the 'F' (forced) taint flag and 151991e37a79SRusty Russell is usually a really bad idea. 1520826e4506SLinus Torvalds 15211da177e4SLinus Torvaldsconfig MODULE_UNLOAD 15221da177e4SLinus Torvalds bool "Module unloading" 15231da177e4SLinus Torvalds help 15241da177e4SLinus Torvalds Without this option you will not be able to unload any 15251da177e4SLinus Torvalds modules (note that some modules may not be unloadable 1526f7f5b675SDenys Vlasenko anyway), which makes your kernel smaller, faster 1527f7f5b675SDenys Vlasenko and simpler. If unsure, say Y. 15281da177e4SLinus Torvalds 15291da177e4SLinus Torvaldsconfig MODULE_FORCE_UNLOAD 15301da177e4SLinus Torvalds bool "Forced module unloading" 15311da177e4SLinus Torvalds depends on MODULE_UNLOAD && EXPERIMENTAL 15321da177e4SLinus Torvalds help 15331da177e4SLinus Torvalds This option allows you to force a module to unload, even if the 15341da177e4SLinus Torvalds kernel believes it is unsafe: the kernel will remove the module 15351da177e4SLinus Torvalds without waiting for anyone to stop using it (using the -f option to 15361da177e4SLinus Torvalds rmmod). This is mainly for kernel developers and desperate users. 15371da177e4SLinus Torvalds If unsure, say N. 15381da177e4SLinus Torvalds 15391da177e4SLinus Torvaldsconfig MODVERSIONS 15400d541643SSam Ravnborg bool "Module versioning support" 15411da177e4SLinus Torvalds help 15421da177e4SLinus Torvalds Usually, you have to use modules compiled with your kernel. 15431da177e4SLinus Torvalds Saying Y here makes it sometimes possible to use modules 15441da177e4SLinus Torvalds compiled for different kernels, by adding enough information 15451da177e4SLinus Torvalds to the modules to (hopefully) spot any changes which would 15461da177e4SLinus Torvalds make them incompatible with the kernel you are running. If 15471da177e4SLinus Torvalds unsure, say N. 15481da177e4SLinus Torvalds 15491da177e4SLinus Torvaldsconfig MODULE_SRCVERSION_ALL 15501da177e4SLinus Torvalds bool "Source checksum for all modules" 15511da177e4SLinus Torvalds help 15521da177e4SLinus Torvalds Modules which contain a MODULE_VERSION get an extra "srcversion" 15531da177e4SLinus Torvalds field inserted into their modinfo section, which contains a 15541da177e4SLinus Torvalds sum of the source files which made it. This helps maintainers 15551da177e4SLinus Torvalds see exactly which source was used to build a module (since 15561da177e4SLinus Torvalds others sometimes change the module source without updating 15571da177e4SLinus Torvalds the version). With this option, such a "srcversion" field 15581da177e4SLinus Torvalds will be created for all modules. If unsure, say N. 15591da177e4SLinus Torvalds 15600b0de144SRobert P. J. Dayendif # MODULES 15610b0de144SRobert P. J. Day 156298a79d6aSRusty Russellconfig INIT_ALL_POSSIBLE 156398a79d6aSRusty Russell bool 156498a79d6aSRusty Russell help 15655f054e31SRusty Russell Back when each arch used to define their own cpu_online_mask and 15665f054e31SRusty Russell cpu_possible_mask, some of them chose to initialize cpu_possible_mask 156798a79d6aSRusty Russell with all 1s, and others with all 0s. When they were centralised, 156898a79d6aSRusty Russell it was better to provide this option than to break all the archs 1569692105b8SMatt LaPlante and have several arch maintainers pursuing me down dark alleys. 157098a79d6aSRusty Russell 15711da177e4SLinus Torvaldsconfig STOP_MACHINE 15721da177e4SLinus Torvalds bool 15731da177e4SLinus Torvalds default y 15741da177e4SLinus Torvalds depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU 15751da177e4SLinus Torvalds help 15761da177e4SLinus Torvalds Need stop_machine() primitive. 15773a65dfe8SJens Axboe 15783a65dfe8SJens Axboesource "block/Kconfig" 1579e98c3202SAvi Kivity 1580e98c3202SAvi Kivityconfig PREEMPT_NOTIFIERS 1581e98c3202SAvi Kivity bool 1582e260be67SPaul E. McKenney 158316295becSSteffen Klassertconfig PADATA 158416295becSSteffen Klassert depends on SMP 158516295becSSteffen Klassert bool 158616295becSSteffen Klassert 1587754b7b63SAndi Kleen# Can be selected by architectures with broken toolchains 1588754b7b63SAndi Kleen# that get confused by correct const<->read_only section 1589754b7b63SAndi Kleen# mappings 1590754b7b63SAndi Kleenconfig BROKEN_RODATA 1591754b7b63SAndi Kleen bool 1592754b7b63SAndi Kleen 15936beb0009SThomas Gleixnersource "kernel/Kconfig.locks" 1594