7 option env="KERNELVERSION"
13 default "/lib/modules/$UNAME_RELEASE/.config"
14 default "/etc/kernel-config"
15 default "/boot/config-$UNAME_RELEASE"
16 default "arch/$ARCH/defconfig"
21 bool "Prompt for development and/or incomplete code/drivers"
23 Some of the various things that Linux supports (such as network
24 drivers, file systems, network protocols, etc.) can be in a state
25 of development where the functionality, stability, or the level of
26 testing is not yet high enough for general use. This is usually
27 known as the "alpha-test" phase among developers. If a feature is
28 currently in alpha-test, then the developers usually discourage
29 uninformed widespread use of this feature by the general public to
30 avoid "Why doesn't this work?" type mail messages. However, active
31 testing and use of these systems is welcomed. Just be aware that it
32 may not meet the normal level of reliability or it may fail to work
33 in some special cases. Detailed bug reports from people familiar
34 with the kernel internals are usually welcomed by the developers
35 (before submitting bug reports, please read the documents
36 <file:README>, <file:MAINTAINERS>, <file:REPORTING-BUGS>,
37 <file:Documentation/BUG-HUNTING>, and
38 <file:Documentation/oops-tracing.txt> in the kernel source).
40 This option will also make obsoleted drivers available. These are
41 drivers that have been replaced by something else, and/or are
42 scheduled to be removed in a future kernel release.
44 Unless you intend to help test and develop a feature or driver that
45 falls into this category, or you have a situation that requires
46 using these features, you should probably say N here, which will
47 cause the configurator to present you with fewer choices. If
48 you say Y here, you will be offered the choice of using features or
49 drivers that are currently considered to be in the alpha-test phase.
56 depends on BROKEN || !SMP
61 depends on SMP || PREEMPT
64 config INIT_ENV_ARG_LIMIT
69 Maximum of each of the number of arguments and environment
70 variables passed to init from the kernel command line.
74 string "Local version - append to kernel release"
76 Append an extra string to the end of your kernel version.
77 This will show up when you type uname, for example.
78 The string you set here will be appended after the contents of
79 any files with a filename matching localversion* in your
80 object and source tree, in that order. Your total string can
81 be a maximum of 64 characters.
83 config LOCALVERSION_AUTO
84 bool "Automatically append version information to the version string"
87 This will try to automatically determine if the current tree is a
88 release tree by looking for git tags that belong to the current
91 A string of the format -gxxxxxxxx will be added to the localversion
92 if a git-based tree is found. The string generated by this will be
93 appended after any matching localversion* files, and after the value
94 set in CONFIG_LOCALVERSION.
96 (The actual string used here is the first eight characters produced
97 by running the command:
99 $ git rev-parse --verify HEAD
101 which is done within the script "scripts/setlocalversion".)
104 bool "Support for paging of anonymous memory (swap)"
105 depends on MMU && BLOCK
108 This option allows you to choose whether you want to have support
109 for so called swap devices or swap files in your kernel that are
110 used to provide more virtual memory than the actual RAM present
111 in your computer. If unsure say Y.
116 Inter Process Communication is a suite of library functions and
117 system calls which let processes (running programs) synchronize and
118 exchange information. It is generally considered to be a good thing,
119 and some programs won't run unless you say Y here. In particular, if
120 you want to run the DOS emulator dosemu under Linux (read the
121 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
122 you'll need to say Y here.
124 You can find documentation about IPC with "info ipc" and also in
125 section 6.4 of the Linux Programmer's Guide, available from
126 <http://www.tldp.org/guides.html>.
128 config SYSVIPC_SYSCTL
135 bool "POSIX Message Queues"
136 depends on NET && EXPERIMENTAL
138 POSIX variant of message queues is a part of IPC. In POSIX message
139 queues every message has a priority which decides about succession
140 of receiving it by a process. If you want to compile and run
141 programs written e.g. for Solaris with use of its POSIX message
142 queues (functions mq_*) say Y here.
144 POSIX message queues are visible as a filesystem called 'mqueue'
145 and can be mounted somewhere if you want to do filesystem
146 operations on message queues.
150 config BSD_PROCESS_ACCT
151 bool "BSD Process Accounting"
153 If you say Y here, a user level program will be able to instruct the
154 kernel (via a special system call) to write process accounting
155 information to a file: whenever a process exits, information about
156 that process will be appended to the file by the kernel. The
157 information includes things such as creation time, owning user,
158 command name, memory usage, controlling terminal etc. (the complete
159 list is in the struct acct in <file:include/linux/acct.h>). It is
160 up to the user level program to do useful things with this
161 information. This is generally a good idea, so say Y.
163 config BSD_PROCESS_ACCT_V3
164 bool "BSD Process Accounting version 3 file format"
165 depends on BSD_PROCESS_ACCT
168 If you say Y here, the process accounting information is written
169 in a new file format that also logs the process IDs of each
170 process and it's parent. Note that this file format is incompatible
171 with previous v0/v1/v2 file formats, so you will need updated tools
172 for processing it. A preliminary version of these tools is available
173 at <http://www.physik3.uni-rostock.de/tim/kernel/utils/acct/>.
176 bool "Export task/process statistics through netlink (EXPERIMENTAL)"
180 Export selected statistics for tasks/processes through the
181 generic netlink interface. Unlike BSD process accounting, the
182 statistics are available during the lifetime of tasks/processes as
183 responses to commands. Like BSD accounting, they are sent to user
188 config TASK_DELAY_ACCT
189 bool "Enable per-task delay accounting (EXPERIMENTAL)"
192 Collect information on time spent by a task waiting for system
193 resources like cpu, synchronous block I/O completion and swapping
194 in pages. Such statistics can help in setting a task's priorities
195 relative to other tasks for cpu, io, rss limits etc.
200 bool "Enable extended accounting over taskstats (EXPERIMENTAL)"
203 Collect extended task accounting data and send the data
204 to userland for processing over the taskstats interface.
208 config TASK_IO_ACCOUNTING
209 bool "Enable per-task storage I/O accounting (EXPERIMENTAL)"
210 depends on TASK_XACCT
212 Collect information on the number of bytes of storage I/O which this
218 bool "User Namespaces (EXPERIMENTAL)"
220 depends on EXPERIMENTAL
222 Support user namespaces. This allows containers, i.e.
223 vservers, to use user namespaces to provide different
224 user info for different servers. If unsure, say N.
227 bool "PID Namespaces (EXPERIMENTAL)"
229 depends on EXPERIMENTAL
231 Suport process id namespaces. This allows having multiple
232 process with the same pid as long as they are in different
233 pid namespaces. This is a building block of containers.
235 Unless you want to work with an experimental feature
239 bool "Auditing support"
242 Enable auditing infrastructure that can be used with another
243 kernel subsystem, such as SELinux (which requires this for
244 logging of avc messages output). Does not do system-call
245 auditing without CONFIG_AUDITSYSCALL.
248 bool "Enable system-call auditing support"
249 depends on AUDIT && (X86 || PPC || PPC64 || S390 || IA64 || UML || SPARC64|| SUPERH)
250 default y if SECURITY_SELINUX
252 Enable low-overhead system-call auditing infrastructure that
253 can be used independently or with another kernel subsystem,
254 such as SELinux. To use audit's filesystem watch feature, please
255 ensure that INOTIFY is configured.
259 depends on AUDITSYSCALL && INOTIFY
262 tristate "Kernel .config support"
264 This option enables the complete Linux kernel ".config" file
265 contents to be saved in the kernel. It provides documentation
266 of which kernel options are used in a running kernel or in an
267 on-disk kernel. This information can be extracted from the kernel
268 image file with the script scripts/extract-ikconfig and used as
269 input to rebuild the current kernel or to build another kernel.
270 It can also be extracted from a running kernel by reading
271 /proc/config.gz if enabled (below).
274 bool "Enable access to .config through /proc/config.gz"
275 depends on IKCONFIG && PROC_FS
277 This option enables access to the kernel configuration file
278 through /proc/config.gz.
281 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
283 default 17 if S390 || LOCKDEP
284 default 16 if X86_NUMAQ || IA64
288 Select kernel log buffer size as a power of 2.
289 Defaults and Examples:
290 17 => 128 KB for S/390
291 16 => 64 KB for x86 NUMAQ or IA-64
293 14 => 16 KB for uniprocessor
298 bool "Control Group support"
300 This option will let you use process cgroup subsystems
306 bool "Example debug cgroup subsystem"
309 This option enables a simple cgroup subsystem that
310 exports useful debugging information about the cgroups
316 bool "Namespace cgroup subsystem"
319 Provides a simple namespace cgroup subsystem to
320 provide hierarchical naming of sets of namespaces,
321 for instance virtual servers and checkpoint/restart
325 bool "Cpuset support"
326 depends on SMP && CGROUPS
328 This option will let you create and manage CPUSETs which
329 allow dynamically partitioning a system into sets of CPUs and
330 Memory Nodes and assigning tasks to run only within those sets.
331 This is primarily useful on large SMP or NUMA systems.
335 config FAIR_GROUP_SCHED
336 bool "Fair group CPU scheduler"
339 This feature lets CPU scheduler recognize task groups and control CPU
340 bandwidth allocation to such task groups.
343 depends on FAIR_GROUP_SCHED
344 prompt "Basis for grouping tasks"
345 default FAIR_USER_SCHED
347 config FAIR_USER_SCHED
350 This option will choose userid as the basis for grouping
351 tasks, thus providing equal CPU bandwidth to each user.
353 config FAIR_CGROUP_SCHED
354 bool "Control groups"
357 This option allows you to create arbitrary task groups
358 using the "cgroup" pseudo filesystem and control
359 the cpu bandwidth allocated to each such task group.
360 Refer to Documentation/cgroups.txt for more information
361 on "cgroup" pseudo filesystem.
365 config CGROUP_CPUACCT
366 bool "Simple CPU accounting cgroup subsystem"
369 Provides a simple Resource Controller for monitoring the
370 total CPU consumed by the tasks in a cgroup
372 config RESOURCE_COUNTERS
373 bool "Resource counters"
375 This option enables controller independent resource accounting
376 infrastructure that works with cgroups
379 config SYSFS_DEPRECATED
380 bool "Create deprecated sysfs files"
384 This option creates deprecated symlinks such as the
385 "device"-link, the <subsystem>:<name>-link, and the
386 "bus"-link. It may also add deprecated key in the
388 None of these features or values should be used today, as
389 they export driver core implementation details to userspace
390 or export properties which can't be kept stable across kernel
393 If enabled, this option will also move any device structures
394 that belong to a class, back into the /sys/class hierarchy, in
395 order to support older versions of udev.
397 If you are using a distro that was released in 2006 or later,
398 it should be safe to say N here.
400 config CGROUP_MEM_CONT
401 bool "Memory controller for cgroups"
402 depends on CGROUPS && RESOURCE_COUNTERS
404 Provides a memory controller that manages both page cache and
407 config PROC_PID_CPUSET
408 bool "Include legacy /proc/<pid>/cpuset file"
413 bool "Kernel->user space relay support (formerly relayfs)"
415 This option enables support for relay interface support in
416 certain file systems (such as debugfs).
417 It is designed to provide an efficient mechanism for tools and
418 facilities to relay large amounts of data from kernel space to
424 bool "Namespaces support" if EMBEDDED
427 Provides the way to make tasks work with different objects using
428 the same id. For example same IPC id may refer to different objects
429 or same user id or pid may refer to different tasks when used in
430 different namespaces.
434 depends on NAMESPACES
436 In this namespace tasks see different info provided with the
441 depends on NAMESPACES && SYSVIPC
443 In this namespace tasks work with IPC ids which correspond to
444 different IPC objects in different namespaces
446 config BLK_DEV_INITRD
447 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
448 depends on BROKEN || !FRV
450 The initial RAM filesystem is a ramfs which is loaded by the
451 boot loader (loadlin or lilo) and that is mounted as root
452 before the normal boot procedure. It is typically used to
453 load modules needed to mount the "real" root file system,
454 etc. See <file:Documentation/initrd.txt> for details.
456 If RAM disk support (BLK_DEV_RAM) is also included, this
457 also enables initial RAM disk (initrd) support and adds
458 15 Kbytes (more on some other architectures) to the kernel size.
468 config CC_OPTIMIZE_FOR_SIZE
469 bool "Optimize for size (Look out for broken compilers!)"
471 depends on ARM || H8300 || SUPERH || EXPERIMENTAL
473 Enabling this option will pass "-Os" instead of "-O2" to gcc
474 resulting in a smaller kernel.
476 WARNING: some versions of gcc may generate incorrect code with this
477 option. If problems are observed, a gcc upgrade may be needed.
485 bool "Configure standard kernel features (for small systems)"
487 This option allows certain base kernel options and settings
488 to be disabled or tweaked. This is for specialized
489 environments which can tolerate a "non-standard" kernel.
490 Only use this if you really know what you are doing.
493 bool "Enable 16-bit UID system calls" if EMBEDDED
494 depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && SPARC32_COMPAT) || UML || (X86_64 && IA32_EMULATION)
497 This enables the legacy 16-bit UID syscall wrappers.
499 config SYSCTL_SYSCALL
500 bool "Sysctl syscall support" if EMBEDDED
504 sys_sysctl uses binary paths that have been found challenging
505 to properly maintain and use. The interface in /proc/sys
506 using paths with ascii names is now the primary path to this
509 Almost nothing using the binary sysctl interface so if you are
510 trying to save some space it is probably safe to disable this,
511 making your kernel marginally smaller.
513 If unsure say Y here.
516 bool "Load all symbols for debugging/ksymoops" if EMBEDDED
519 Say Y here to let the kernel print out symbolic crash information and
520 symbolic stack backtraces. This increases the size of the kernel
521 somewhat, as all symbols have to be loaded into the kernel image.
524 bool "Include all symbols in kallsyms"
525 depends on DEBUG_KERNEL && KALLSYMS
527 Normally kallsyms only contains the symbols of functions, for nicer
528 OOPS messages. Some debuggers can use kallsyms for other
529 symbols too: say Y here to include all symbols, if you need them
530 and you don't care about adding 300k to the size of your kernel.
534 config KALLSYMS_EXTRA_PASS
535 bool "Do an extra kallsyms pass"
538 If kallsyms is not working correctly, the build will fail with
539 inconsistent kallsyms data. If that occurs, log a bug report and
540 turn on KALLSYMS_EXTRA_PASS which should result in a stable build.
541 Always say N here unless you find a bug in kallsyms, which must be
542 reported. KALLSYMS_EXTRA_PASS is only a temporary workaround while
543 you wait for kallsyms to be fixed.
547 bool "Support for hot-pluggable devices" if EMBEDDED
550 This option is provided for the case where no hotplug or uevent
551 capabilities is wanted by the kernel. You should only consider
552 disabling this option for embedded systems that do not use modules, a
553 dynamic /dev tree, or dynamic device discovery. Just say Y.
557 bool "Enable support for printk" if EMBEDDED
559 This option enables normal printk support. Removing it
560 eliminates most of the message strings from the kernel image
561 and makes the kernel more or less silent. As this makes it
562 very difficult to diagnose system problems, saying N here is
563 strongly discouraged.
566 bool "BUG() support" if EMBEDDED
569 Disabling this option eliminates support for BUG and WARN, reducing
570 the size of your kernel image and potentially quietly ignoring
571 numerous fatal conditions. You should only consider disabling this
572 option for embedded systems with no facilities for reporting errors.
577 bool "Enable ELF core dumps" if EMBEDDED
579 Enable support for generating core dumps. Disabling saves about 4k.
582 bool "Disable heap randomization"
585 Randomizing heap placement makes heap exploits harder, but it
586 also breaks ancient binaries (including anything libc5 based).
587 This option changes the bootup default to heap randomization
588 disabled, and can be overriden runtime by setting
589 /proc/sys/kernel/randomize_va_space to 2.
591 On non-ancient distros (post-2000 ones) Y is usually a safe choice.
595 bool "Enable full-sized data structures for core" if EMBEDDED
597 Disabling this option reduces the size of miscellaneous core
598 kernel data structures. This saves memory on small machines,
599 but may reduce performance.
602 bool "Enable futex support" if EMBEDDED
606 Disabling this option will cause the kernel to be built without
607 support for "fast userspace mutexes". The resulting kernel may not
608 run glibc-based applications correctly.
614 bool "Enable eventpoll support" if EMBEDDED
618 Disabling this option will cause the kernel to be built without
619 support for epoll family of system calls.
622 bool "Enable signalfd() system call" if EMBEDDED
626 Enable the signalfd() system call that allows to receive signals
627 on a file descriptor.
632 bool "Enable timerfd() system call" if EMBEDDED
636 Enable the timerfd() system call that allows to receive timer
637 events on a file descriptor.
642 bool "Enable eventfd() system call" if EMBEDDED
646 Enable the eventfd() system call that allows to receive both
647 kernel notification (ie. KAIO) or userspace notifications.
652 bool "Use full shmem filesystem" if EMBEDDED
656 The shmem is an internal filesystem used to manage shared memory.
657 It is backed by swap and manages resource limits. It is also exported
658 to userspace as tmpfs if TMPFS is enabled. Disabling this
659 option replaces shmem and tmpfs with the much simpler ramfs code,
660 which may be appropriate on small systems without swap.
662 config VM_EVENT_COUNTERS
664 bool "Enable VM event counters for /proc/vmstat" if EMBEDDED
666 VM event counters are needed for event counts to be shown.
667 This option allows the disabling of the VM event counters
668 on EMBEDDED systems. /proc/vmstat will only show page counts
669 if VM event counters are disabled.
673 bool "Enable SLUB debugging support" if EMBEDDED
676 SLUB has extensive debug support features. Disabling these can
677 result in significant savings in code size. This also disables
678 SLUB sysfs support. /sys/slab will not exist and there will be
679 no support for cache validation etc.
682 prompt "Choose SLAB allocator"
685 This option allows to select a slab allocator.
690 The regular slab allocator that is established and known to work
691 well in all environments. It organizes cache hot objects in
692 per cpu and per node queues. SLAB is the default choice for
696 bool "SLUB (Unqueued Allocator)"
698 SLUB is a slab allocator that minimizes cache line usage
699 instead of managing queues of cached objects (SLAB approach).
700 Per cpu caching is realized using slabs of objects instead
701 of queues of objects. SLUB can use memory efficiently
702 and has enhanced diagnostics.
706 bool "SLOB (Simple Allocator)"
708 SLOB replaces the stock allocator with a drastically simpler
709 allocator. SLOB is generally more space efficient but
710 does not perform as well on large systems.
715 bool "Profiling support (EXPERIMENTAL)"
717 Say Y here to enable the extended profiling support mechanisms used
718 by profilers such as OProfile.
721 bool "Activate markers"
723 Place an empty function call at each marker site. Can be
724 dynamically changed for a probe function.
726 source "arch/Kconfig"
728 config PROC_PAGE_MONITOR
730 depends on PROC_FS && MMU
731 bool "Enable /proc page monitoring" if EMBEDDED
733 Various /proc files exist to monitor process memory utilization:
734 /proc/pid/smaps, /proc/pid/clear_refs, /proc/pid/pagemap,
735 /proc/kpagecount, and /proc/kpageflags. Disabling these
736 interfaces will reduce the size of the kernel by approximately 4kb.
738 endmenu # General setup
743 depends on SLAB || SLUB
756 default 0 if BASE_FULL
757 default 1 if !BASE_FULL
760 bool "Enable loadable module support"
762 Kernel modules are small pieces of compiled code which can
763 be inserted in the running kernel, rather than being
764 permanently built into the kernel. You use the "modprobe"
765 tool to add (and sometimes remove) them. If you say Y here,
766 many parts of the kernel can be built as modules (by
767 answering M instead of Y where indicated): this is most
768 useful for infrequently used options which are not required
769 for booting. For more information, see the man pages for
770 modprobe, lsmod, modinfo, insmod and rmmod.
772 If you say Y here, you will need to run "make
773 modules_install" to put the modules under /lib/modules/
774 where modprobe can find them (you may need to be root to do
780 bool "Module unloading"
783 Without this option you will not be able to unload any
784 modules (note that some modules may not be unloadable
785 anyway), which makes your kernel slightly smaller and
786 simpler. If unsure, say Y.
788 config MODULE_FORCE_UNLOAD
789 bool "Forced module unloading"
790 depends on MODULE_UNLOAD && EXPERIMENTAL
792 This option allows you to force a module to unload, even if the
793 kernel believes it is unsafe: the kernel will remove the module
794 without waiting for anyone to stop using it (using the -f option to
795 rmmod). This is mainly for kernel developers and desperate users.
799 bool "Module versioning support"
802 Usually, you have to use modules compiled with your kernel.
803 Saying Y here makes it sometimes possible to use modules
804 compiled for different kernels, by adding enough information
805 to the modules to (hopefully) spot any changes which would
806 make them incompatible with the kernel you are running. If
809 config MODULE_SRCVERSION_ALL
810 bool "Source checksum for all modules"
813 Modules which contain a MODULE_VERSION get an extra "srcversion"
814 field inserted into their modinfo section, which contains a
815 sum of the source files which made it. This helps maintainers
816 see exactly which source was used to build a module (since
817 others sometimes change the module source without updating
818 the version). With this option, such a "srcversion" field
819 will be created for all modules. If unsure, say N.
822 bool "Automatic kernel module loading"
825 Normally when you have selected some parts of the kernel to
826 be created as kernel modules, you must load them (using the
827 "modprobe" command) before you can use them. If you say Y
828 here, some parts of the kernel will be able to load modules
829 automatically: when a part of the kernel needs a module, it
830 runs modprobe with the appropriate arguments, thereby
831 loading the module if it is available. If unsure, say Y.
836 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU
838 Need stop_machine() primitive.
840 source "block/Kconfig"
842 config PREEMPT_NOTIFIERS
846 prompt "RCU implementation type:"
849 This allows you to choose either the classic RCU implementation
850 that is designed for best read-side performance on non-realtime
851 systems, or the preemptible RCU implementation for best latency
852 on realtime systems. Note that some kernel preemption modes
853 will restrict your choice.
855 Select the default if you are unsure.
860 This option selects the classic RCU implementation that is
861 designed for best read-side performance on non-realtime
864 Say Y if you are unsure.
867 bool "Preemptible RCU"
870 This option reduces the latency of the kernel by making certain
871 RCU sections preemptible. Normally RCU code is non-preemptible, if
872 this option is selected then read-only RCU sections become
873 preemptible. This helps latency, but may expose bugs due to
874 now-naive assumptions about each RCU read-side critical section
875 remaining on a given CPU through its execution.
877 Say N if you are unsure.