7 option env="KERNELVERSION"
13 default "/lib/modules/$UNAME_RELEASE/.config"
14 default "/etc/kernel-config"
15 default "/boot/config-$UNAME_RELEASE"
16 default "$ARCH_DEFCONFIG"
17 default "arch/$ARCH/defconfig"
27 bool "Prompt for development and/or incomplete code/drivers"
29 Some of the various things that Linux supports (such as network
30 drivers, file systems, network protocols, etc.) can be in a state
31 of development where the functionality, stability, or the level of
32 testing is not yet high enough for general use. This is usually
33 known as the "alpha-test" phase among developers. If a feature is
34 currently in alpha-test, then the developers usually discourage
35 uninformed widespread use of this feature by the general public to
36 avoid "Why doesn't this work?" type mail messages. However, active
37 testing and use of these systems is welcomed. Just be aware that it
38 may not meet the normal level of reliability or it may fail to work
39 in some special cases. Detailed bug reports from people familiar
40 with the kernel internals are usually welcomed by the developers
41 (before submitting bug reports, please read the documents
42 <file:README>, <file:MAINTAINERS>, <file:REPORTING-BUGS>,
43 <file:Documentation/BUG-HUNTING>, and
44 <file:Documentation/oops-tracing.txt> in the kernel source).
46 This option will also make obsoleted drivers available. These are
47 drivers that have been replaced by something else, and/or are
48 scheduled to be removed in a future kernel release.
50 Unless you intend to help test and develop a feature or driver that
51 falls into this category, or you have a situation that requires
52 using these features, you should probably say N here, which will
53 cause the configurator to present you with fewer choices. If
54 you say Y here, you will be offered the choice of using features or
55 drivers that are currently considered to be in the alpha-test phase.
62 depends on BROKEN || !SMP
67 depends on SMP || PREEMPT
70 config INIT_ENV_ARG_LIMIT
75 Maximum of each of the number of arguments and environment
76 variables passed to init from the kernel command line.
80 string "Cross-compiler tool prefix"
82 Same as running 'make CROSS_COMPILE=prefix-' but stored for
83 default make runs in this kernel build directory. You don't
84 need to set this unless you want the configured kernel build
85 directory to select the cross-compiler automatically.
88 string "Local version - append to kernel release"
90 Append an extra string to the end of your kernel version.
91 This will show up when you type uname, for example.
92 The string you set here will be appended after the contents of
93 any files with a filename matching localversion* in your
94 object and source tree, in that order. Your total string can
95 be a maximum of 64 characters.
97 config LOCALVERSION_AUTO
98 bool "Automatically append version information to the version string"
101 This will try to automatically determine if the current tree is a
102 release tree by looking for git tags that belong to the current
103 top of tree revision.
105 A string of the format -gxxxxxxxx will be added to the localversion
106 if a git-based tree is found. The string generated by this will be
107 appended after any matching localversion* files, and after the value
108 set in CONFIG_LOCALVERSION.
110 (The actual string used here is the first eight characters produced
111 by running the command:
113 $ git rev-parse --verify HEAD
115 which is done within the script "scripts/setlocalversion".)
117 config HAVE_KERNEL_GZIP
120 config HAVE_KERNEL_BZIP2
123 config HAVE_KERNEL_LZMA
126 config HAVE_KERNEL_LZO
130 prompt "Kernel compression mode"
132 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_LZO
134 The linux kernel is a kind of self-extracting executable.
135 Several compression algorithms are available, which differ
136 in efficiency, compression and decompression speed.
137 Compression speed is only relevant when building a kernel.
138 Decompression speed is relevant at each boot.
140 If you have any problems with bzip2 or lzma compressed
141 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
142 version of this functionality (bzip2 only), for 2.4, was
143 supplied by Christian Ludwig)
145 High compression options are mostly useful for users, who
146 are low on disk space (embedded systems), but for whom ram
149 If in doubt, select 'gzip'
153 depends on HAVE_KERNEL_GZIP
155 The old and tried gzip compression. It provides a good balance
156 between compression ratio and decompression speed.
160 depends on HAVE_KERNEL_BZIP2
162 Its compression ratio and speed is intermediate.
163 Decompression speed is slowest among the three. The kernel
164 size is about 10% smaller with bzip2, in comparison to gzip.
165 Bzip2 uses a large amount of memory. For modern kernels you
166 will need at least 8MB RAM or more for booting.
170 depends on HAVE_KERNEL_LZMA
172 The most recent compression algorithm.
173 Its ratio is best, decompression speed is between the other
174 two. Compression is slowest. The kernel size is about 33%
175 smaller with LZMA in comparison to gzip.
179 depends on HAVE_KERNEL_LZO
181 Its compression ratio is the poorest among the 4. The kernel
182 size is about about 10% bigger than gzip; however its speed
183 (both compression and decompression) is the fastest.
188 bool "Support for paging of anonymous memory (swap)"
189 depends on MMU && BLOCK
192 This option allows you to choose whether you want to have support
193 for so called swap devices or swap files in your kernel that are
194 used to provide more virtual memory than the actual RAM present
195 in your computer. If unsure say Y.
200 Inter Process Communication is a suite of library functions and
201 system calls which let processes (running programs) synchronize and
202 exchange information. It is generally considered to be a good thing,
203 and some programs won't run unless you say Y here. In particular, if
204 you want to run the DOS emulator dosemu under Linux (read the
205 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
206 you'll need to say Y here.
208 You can find documentation about IPC with "info ipc" and also in
209 section 6.4 of the Linux Programmer's Guide, available from
210 <http://www.tldp.org/guides.html>.
212 config SYSVIPC_SYSCTL
219 bool "POSIX Message Queues"
220 depends on NET && EXPERIMENTAL
222 POSIX variant of message queues is a part of IPC. In POSIX message
223 queues every message has a priority which decides about succession
224 of receiving it by a process. If you want to compile and run
225 programs written e.g. for Solaris with use of its POSIX message
226 queues (functions mq_*) say Y here.
228 POSIX message queues are visible as a filesystem called 'mqueue'
229 and can be mounted somewhere if you want to do filesystem
230 operations on message queues.
234 config POSIX_MQUEUE_SYSCTL
236 depends on POSIX_MQUEUE
240 config BSD_PROCESS_ACCT
241 bool "BSD Process Accounting"
243 If you say Y here, a user level program will be able to instruct the
244 kernel (via a special system call) to write process accounting
245 information to a file: whenever a process exits, information about
246 that process will be appended to the file by the kernel. The
247 information includes things such as creation time, owning user,
248 command name, memory usage, controlling terminal etc. (the complete
249 list is in the struct acct in <file:include/linux/acct.h>). It is
250 up to the user level program to do useful things with this
251 information. This is generally a good idea, so say Y.
253 config BSD_PROCESS_ACCT_V3
254 bool "BSD Process Accounting version 3 file format"
255 depends on BSD_PROCESS_ACCT
258 If you say Y here, the process accounting information is written
259 in a new file format that also logs the process IDs of each
260 process and it's parent. Note that this file format is incompatible
261 with previous v0/v1/v2 file formats, so you will need updated tools
262 for processing it. A preliminary version of these tools is available
263 at <http://www.gnu.org/software/acct/>.
266 bool "Export task/process statistics through netlink (EXPERIMENTAL)"
270 Export selected statistics for tasks/processes through the
271 generic netlink interface. Unlike BSD process accounting, the
272 statistics are available during the lifetime of tasks/processes as
273 responses to commands. Like BSD accounting, they are sent to user
278 config TASK_DELAY_ACCT
279 bool "Enable per-task delay accounting (EXPERIMENTAL)"
282 Collect information on time spent by a task waiting for system
283 resources like cpu, synchronous block I/O completion and swapping
284 in pages. Such statistics can help in setting a task's priorities
285 relative to other tasks for cpu, io, rss limits etc.
290 bool "Enable extended accounting over taskstats (EXPERIMENTAL)"
293 Collect extended task accounting data and send the data
294 to userland for processing over the taskstats interface.
298 config TASK_IO_ACCOUNTING
299 bool "Enable per-task storage I/O accounting (EXPERIMENTAL)"
300 depends on TASK_XACCT
302 Collect information on the number of bytes of storage I/O which this
308 bool "Auditing support"
311 Enable auditing infrastructure that can be used with another
312 kernel subsystem, such as SELinux (which requires this for
313 logging of avc messages output). Does not do system-call
314 auditing without CONFIG_AUDITSYSCALL.
317 bool "Enable system-call auditing support"
318 depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH)
319 default y if SECURITY_SELINUX
321 Enable low-overhead system-call auditing infrastructure that
322 can be used independently or with another kernel subsystem,
327 depends on AUDITSYSCALL
332 depends on AUDITSYSCALL
338 prompt "RCU Implementation"
342 bool "Tree-based hierarchical RCU"
344 This option selects the RCU implementation that is
345 designed for very large SMP system with hundreds or
346 thousands of CPUs. It also scales down nicely to
349 config TREE_PREEMPT_RCU
350 bool "Preemptable tree-based hierarchical RCU"
353 This option selects the RCU implementation that is
354 designed for very large SMP systems with hundreds or
355 thousands of CPUs, but for which real-time response
356 is also required. It also scales down nicely to
360 bool "UP-only small-memory-footprint RCU"
363 This option selects the RCU implementation that is
364 designed for UP systems from which real-time response
365 is not required. This option greatly reduces the
366 memory footprint of RCU.
371 bool "Enable tracing for RCU"
372 depends on TREE_RCU || TREE_PREEMPT_RCU
374 This option provides tracing in RCU which presents stats
375 in debugfs for debugging RCU implementation.
377 Say Y here if you want to enable RCU tracing
378 Say N if you are unsure.
381 int "Tree-based hierarchical RCU fanout value"
384 depends on TREE_RCU || TREE_PREEMPT_RCU
388 This option controls the fanout of hierarchical implementations
389 of RCU, allowing RCU to work efficiently on machines with
390 large numbers of CPUs. This value must be at least the cube
391 root of NR_CPUS, which allows NR_CPUS up to 32,768 for 32-bit
392 systems and up to 262,144 for 64-bit systems.
394 Select a specific number if testing RCU itself.
395 Take the default if unsure.
397 config RCU_FANOUT_EXACT
398 bool "Disable tree-based hierarchical RCU auto-balancing"
399 depends on TREE_RCU || TREE_PREEMPT_RCU
402 This option forces use of the exact RCU_FANOUT value specified,
403 regardless of imbalances in the hierarchy. This is useful for
404 testing RCU itself, and might one day be useful on systems with
405 strong NUMA behavior.
407 Without RCU_FANOUT_EXACT, the code will balance the hierarchy.
411 config RCU_FAST_NO_HZ
412 bool "Accelerate last non-dyntick-idle CPU's grace periods"
413 depends on TREE_RCU && NO_HZ && SMP
416 This option causes RCU to attempt to accelerate grace periods
417 in order to allow the final CPU to enter dynticks-idle state
418 more quickly. On the other hand, this option increases the
419 overhead of the dynticks-idle checking, particularly on systems
420 with large numbers of CPUs.
422 Say Y if energy efficiency is critically important, particularly
423 if you have relatively few CPUs.
425 Say N if you are unsure.
427 config TREE_RCU_TRACE
428 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU )
431 This option provides tracing for the TREE_RCU and
432 TREE_PREEMPT_RCU implementations, permitting Makefile to
433 trivially select kernel/rcutree_trace.c.
435 endmenu # "RCU Subsystem"
438 tristate "Kernel .config support"
440 This option enables the complete Linux kernel ".config" file
441 contents to be saved in the kernel. It provides documentation
442 of which kernel options are used in a running kernel or in an
443 on-disk kernel. This information can be extracted from the kernel
444 image file with the script scripts/extract-ikconfig and used as
445 input to rebuild the current kernel or to build another kernel.
446 It can also be extracted from a running kernel by reading
447 /proc/config.gz if enabled (below).
450 bool "Enable access to .config through /proc/config.gz"
451 depends on IKCONFIG && PROC_FS
453 This option enables access to the kernel configuration file
454 through /proc/config.gz.
457 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
461 Select kernel log buffer size as a power of 2.
471 # Architectures with an unreliable sched_clock() should select this:
473 config HAVE_UNSTABLE_SCHED_CLOCK
477 boolean "Control Group support"
480 This option adds support for grouping sets of processes together, for
481 use with process control subsystems such as Cpusets, CFS, memory
482 controls or device isolation.
484 - Documentation/scheduler/sched-design-CFS.txt (CFS)
485 - Documentation/cgroups/ (features for grouping, isolation
486 and resource control)
493 bool "Example debug cgroup subsystem"
497 This option enables a simple cgroup subsystem that
498 exports useful debugging information about the cgroups
504 bool "Namespace cgroup subsystem"
507 Provides a simple namespace cgroup subsystem to
508 provide hierarchical naming of sets of namespaces,
509 for instance virtual servers and checkpoint/restart
512 config CGROUP_FREEZER
513 bool "Freezer cgroup subsystem"
516 Provides a way to freeze and unfreeze all tasks in a
520 bool "Device controller for cgroups"
521 depends on CGROUPS && EXPERIMENTAL
523 Provides a cgroup implementing whitelists for devices which
524 a process in the cgroup can mknod or open.
527 bool "Cpuset support"
530 This option will let you create and manage CPUSETs which
531 allow dynamically partitioning a system into sets of CPUs and
532 Memory Nodes and assigning tasks to run only within those sets.
533 This is primarily useful on large SMP or NUMA systems.
537 config PROC_PID_CPUSET
538 bool "Include legacy /proc/<pid>/cpuset file"
542 config CGROUP_CPUACCT
543 bool "Simple CPU accounting cgroup subsystem"
546 Provides a simple Resource Controller for monitoring the
547 total CPU consumed by the tasks in a cgroup.
549 config RESOURCE_COUNTERS
550 bool "Resource counters"
552 This option enables controller independent resource accounting
553 infrastructure that works with cgroups.
556 config CGROUP_MEM_RES_CTLR
557 bool "Memory Resource Controller for Control Groups"
558 depends on CGROUPS && RESOURCE_COUNTERS
561 Provides a memory resource controller that manages both anonymous
562 memory and page cache. (See Documentation/cgroups/memory.txt)
564 Note that setting this option increases fixed memory overhead
565 associated with each page of memory in the system. By this,
566 20(40)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory
567 usage tracking struct at boot. Total amount of this is printed out
570 Only enable when you're ok with these trade offs and really
571 sure you need the memory resource controller. Even when you enable
572 this, you can set "cgroup_disable=memory" at your boot option to
573 disable memory resource controller and you can avoid overheads.
574 (and lose benefits of memory resource controller)
576 This config option also selects MM_OWNER config option, which
577 could in turn add some fork/exit overhead.
579 config CGROUP_MEM_RES_CTLR_SWAP
580 bool "Memory Resource Controller Swap Extension(EXPERIMENTAL)"
581 depends on CGROUP_MEM_RES_CTLR && SWAP && EXPERIMENTAL
583 Add swap management feature to memory resource controller. When you
584 enable this, you can limit mem+swap usage per cgroup. In other words,
585 when you disable this, memory resource controller has no cares to
586 usage of swap...a process can exhaust all of the swap. This extension
587 is useful when you want to avoid exhaustion swap but this itself
588 adds more overheads and consumes memory for remembering information.
589 Especially if you use 32bit system or small memory system, please
590 be careful about enabling this. When memory resource controller
591 is disabled by boot option, this will be automatically disabled and
592 there will be no overhead from this. Even when you set this config=y,
593 if boot option "noswapaccount" is set, swap will not be accounted.
594 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page
595 size is 4096bytes, 512k per 1Gbytes of swap.
597 menuconfig CGROUP_SCHED
598 bool "Group CPU scheduler"
599 depends on EXPERIMENTAL && CGROUPS
602 This feature lets CPU scheduler recognize task groups and control CPU
603 bandwidth allocation to such task groups. It uses cgroups to group
607 config FAIR_GROUP_SCHED
608 bool "Group scheduling for SCHED_OTHER"
609 depends on CGROUP_SCHED
612 config RT_GROUP_SCHED
613 bool "Group scheduling for SCHED_RR/FIFO"
614 depends on EXPERIMENTAL
615 depends on CGROUP_SCHED
618 This feature lets you explicitly allocate real CPU bandwidth
619 to task groups. If enabled, it will also make it impossible to
620 schedule realtime tasks for non-root users until you allocate
621 realtime bandwidth for them.
622 See Documentation/scheduler/sched-rt-group.txt for more information.
627 tristate "Block IO controller"
628 depends on CGROUPS && BLOCK
631 Generic block IO controller cgroup interface. This is the common
632 cgroup interface which should be used by various IO controlling
635 Currently, CFQ IO scheduler uses it to recognize task groups and
636 control disk bandwidth allocation (proportional time slice allocation)
639 This option only enables generic Block IO controller infrastructure.
640 One needs to also enable actual IO controlling logic in CFQ for it
641 to take effect. (CONFIG_CFQ_GROUP_IOSCHED=y).
643 See Documentation/cgroups/blkio-controller.txt for more information.
645 config DEBUG_BLK_CGROUP
646 bool "Enable Block IO controller debugging"
647 depends on BLK_CGROUP
650 Enable some debugging help. Currently it exports additional stat
651 files in a cgroup which can be useful for debugging.
658 config SYSFS_DEPRECATED
661 config SYSFS_DEPRECATED_V2
662 bool "enable deprecated sysfs features to support old userspace tools"
665 select SYSFS_DEPRECATED
667 This option switches the layout of sysfs to the deprecated
668 version. Do not use it on recent distributions.
670 The current sysfs layout features a unified device tree at
671 /sys/devices/, which is able to express a hierarchy between
672 class devices. If the deprecated option is set to Y, the
673 unified device tree is split into a bus device tree at
674 /sys/devices/ and several individual class device trees at
675 /sys/class/. The class and bus devices will be connected by
676 "<subsystem>:<name>" and the "device" links. The "block"
677 class devices, will not show up in /sys/class/block/. Some
678 subsystems will suppress the creation of some devices which
679 depend on the unified device tree.
681 This option is not a pure compatibility option that can
682 be safely enabled on newer distributions. It will change the
683 layout of sysfs to the non-extensible deprecated version,
684 and disable some features, which can not be exported without
685 confusing older userspace tools. Since 2007/2008 all major
686 distributions do not enable this option, and ship no tools which
687 depend on the deprecated layout or this option.
689 If you are using a new kernel on an older distribution, or use
690 older userspace tools, you might need to say Y here. Do not say Y,
691 if the original kernel, that came with your distribution, has
692 this option set to N.
695 bool "Kernel->user space relay support (formerly relayfs)"
697 This option enables support for relay interface support in
698 certain file systems (such as debugfs).
699 It is designed to provide an efficient mechanism for tools and
700 facilities to relay large amounts of data from kernel space to
706 bool "Namespaces support" if EMBEDDED
709 Provides the way to make tasks work with different objects using
710 the same id. For example same IPC id may refer to different objects
711 or same user id or pid may refer to different tasks when used in
712 different namespaces.
716 depends on NAMESPACES
718 In this namespace tasks see different info provided with the
723 depends on NAMESPACES && (SYSVIPC || POSIX_MQUEUE)
725 In this namespace tasks work with IPC ids which correspond to
726 different IPC objects in different namespaces.
729 bool "User namespace (EXPERIMENTAL)"
730 depends on NAMESPACES && EXPERIMENTAL
732 This allows containers, i.e. vservers, to use user namespaces
733 to provide different user info for different servers.
737 bool "PID Namespaces (EXPERIMENTAL)"
739 depends on NAMESPACES && EXPERIMENTAL
741 Support process id namespaces. This allows having multiple
742 processes with the same pid as long as they are in different
743 pid namespaces. This is a building block of containers.
745 Unless you want to work with an experimental feature
749 bool "Network namespace"
751 depends on NAMESPACES && EXPERIMENTAL && NET
753 Allow user space to create what appear to be multiple instances
754 of the network stack.
756 config BLK_DEV_INITRD
757 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
758 depends on BROKEN || !FRV
760 The initial RAM filesystem is a ramfs which is loaded by the
761 boot loader (loadlin or lilo) and that is mounted as root
762 before the normal boot procedure. It is typically used to
763 load modules needed to mount the "real" root file system,
764 etc. See <file:Documentation/initrd.txt> for details.
766 If RAM disk support (BLK_DEV_RAM) is also included, this
767 also enables initial RAM disk (initrd) support and adds
768 15 Kbytes (more on some other architectures) to the kernel size.
778 config CC_OPTIMIZE_FOR_SIZE
779 bool "Optimize for size"
782 Enabling this option will pass "-Os" instead of "-O2" to gcc
783 resulting in a smaller kernel.
794 bool "Configure standard kernel features (for small systems)"
796 This option allows certain base kernel options and settings
797 to be disabled or tweaked. This is for specialized
798 environments which can tolerate a "non-standard" kernel.
799 Only use this if you really know what you are doing.
802 bool "Enable 16-bit UID system calls" if EMBEDDED
803 depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && COMPAT) || UML || (X86_64 && IA32_EMULATION)
806 This enables the legacy 16-bit UID syscall wrappers.
808 config SYSCTL_SYSCALL
809 bool "Sysctl syscall support" if EMBEDDED
810 depends on PROC_SYSCTL
814 sys_sysctl uses binary paths that have been found challenging
815 to properly maintain and use. The interface in /proc/sys
816 using paths with ascii names is now the primary path to this
819 Almost nothing using the binary sysctl interface so if you are
820 trying to save some space it is probably safe to disable this,
821 making your kernel marginally smaller.
823 If unsure say Y here.
826 bool "Load all symbols for debugging/ksymoops" if EMBEDDED
829 Say Y here to let the kernel print out symbolic crash information and
830 symbolic stack backtraces. This increases the size of the kernel
831 somewhat, as all symbols have to be loaded into the kernel image.
834 bool "Include all symbols in kallsyms"
835 depends on DEBUG_KERNEL && KALLSYMS
837 Normally kallsyms only contains the symbols of functions, for nicer
838 OOPS messages. Some debuggers can use kallsyms for other
839 symbols too: say Y here to include all symbols, if you need them
840 and you don't care about adding 300k to the size of your kernel.
844 config KALLSYMS_EXTRA_PASS
845 bool "Do an extra kallsyms pass"
848 If kallsyms is not working correctly, the build will fail with
849 inconsistent kallsyms data. If that occurs, log a bug report and
850 turn on KALLSYMS_EXTRA_PASS which should result in a stable build.
851 Always say N here unless you find a bug in kallsyms, which must be
852 reported. KALLSYMS_EXTRA_PASS is only a temporary workaround while
853 you wait for kallsyms to be fixed.
857 bool "Support for hot-pluggable devices" if EMBEDDED
860 This option is provided for the case where no hotplug or uevent
861 capabilities is wanted by the kernel. You should only consider
862 disabling this option for embedded systems that do not use modules, a
863 dynamic /dev tree, or dynamic device discovery. Just say Y.
867 bool "Enable support for printk" if EMBEDDED
869 This option enables normal printk support. Removing it
870 eliminates most of the message strings from the kernel image
871 and makes the kernel more or less silent. As this makes it
872 very difficult to diagnose system problems, saying N here is
873 strongly discouraged.
876 bool "BUG() support" if EMBEDDED
879 Disabling this option eliminates support for BUG and WARN, reducing
880 the size of your kernel image and potentially quietly ignoring
881 numerous fatal conditions. You should only consider disabling this
882 option for embedded systems with no facilities for reporting errors.
887 bool "Enable ELF core dumps" if EMBEDDED
889 Enable support for generating core dumps. Disabling saves about 4k.
891 config PCSPKR_PLATFORM
892 bool "Enable PC-Speaker support" if EMBEDDED
893 depends on ALPHA || X86 || MIPS || PPC_PREP || PPC_CHRP || PPC_PSERIES
896 This option allows to disable the internal PC-Speaker
897 support, saving some memory.
901 bool "Enable full-sized data structures for core" if EMBEDDED
903 Disabling this option reduces the size of miscellaneous core
904 kernel data structures. This saves memory on small machines,
905 but may reduce performance.
908 bool "Enable futex support" if EMBEDDED
912 Disabling this option will cause the kernel to be built without
913 support for "fast userspace mutexes". The resulting kernel may not
914 run glibc-based applications correctly.
917 bool "Enable eventpoll support" if EMBEDDED
921 Disabling this option will cause the kernel to be built without
922 support for epoll family of system calls.
925 bool "Enable signalfd() system call" if EMBEDDED
929 Enable the signalfd() system call that allows to receive signals
930 on a file descriptor.
935 bool "Enable timerfd() system call" if EMBEDDED
939 Enable the timerfd() system call that allows to receive timer
940 events on a file descriptor.
945 bool "Enable eventfd() system call" if EMBEDDED
949 Enable the eventfd() system call that allows to receive both
950 kernel notification (ie. KAIO) or userspace notifications.
955 bool "Use full shmem filesystem" if EMBEDDED
959 The shmem is an internal filesystem used to manage shared memory.
960 It is backed by swap and manages resource limits. It is also exported
961 to userspace as tmpfs if TMPFS is enabled. Disabling this
962 option replaces shmem and tmpfs with the much simpler ramfs code,
963 which may be appropriate on small systems without swap.
966 bool "Enable AIO support" if EMBEDDED
969 This option enables POSIX asynchronous I/O which may by used
970 by some high performance threaded applications. Disabling
971 this option saves about 7k.
973 config HAVE_PERF_EVENTS
976 See tools/perf/design.txt for details.
978 config PERF_USE_VMALLOC
981 See tools/perf/design.txt for details
983 menu "Kernel Performance Events And Counters"
986 bool "Kernel performance events and counters"
987 default y if (PROFILING || PERF_COUNTERS)
988 depends on HAVE_PERF_EVENTS
991 Enable kernel support for various performance events provided
992 by software and hardware.
994 Software events are supported either built-in or via the
995 use of generic tracepoints.
997 Most modern CPUs support performance events via performance
998 counter registers. These registers count the number of certain
999 types of hw events: such as instructions executed, cachemisses
1000 suffered, or branches mis-predicted - without slowing down the
1001 kernel or applications. These registers can also trigger interrupts
1002 when a threshold number of events have passed - and can thus be
1003 used to profile the code that runs on that CPU.
1005 The Linux Performance Event subsystem provides an abstraction of
1006 these software and hardware event capabilities, available via a
1007 system call and used by the "perf" utility in tools/perf/. It
1008 provides per task and per CPU counters, and it provides event
1009 capabilities on top of those.
1013 config PERF_COUNTERS
1014 bool "Kernel performance counters (old config option)"
1015 depends on HAVE_PERF_EVENTS
1017 This config has been obsoleted by the PERF_EVENTS
1018 config option - please see that one for details.
1020 It has no effect on the kernel whether you enable
1021 it or not, it is a compatibility placeholder.
1025 config DEBUG_PERF_USE_VMALLOC
1027 bool "Debug: use vmalloc to back perf mmap() buffers"
1028 depends on PERF_EVENTS && DEBUG_KERNEL
1029 select PERF_USE_VMALLOC
1031 Use vmalloc memory to back perf mmap() buffers.
1033 Mostly useful for debugging the vmalloc code on platforms
1034 that don't require it.
1040 config VM_EVENT_COUNTERS
1042 bool "Enable VM event counters for /proc/vmstat" if EMBEDDED
1044 VM event counters are needed for event counts to be shown.
1045 This option allows the disabling of the VM event counters
1046 on EMBEDDED systems. /proc/vmstat will only show page counts
1047 if VM event counters are disabled.
1051 bool "Enable PCI quirk workarounds" if EMBEDDED
1054 This enables workarounds for various PCI chipset
1055 bugs/quirks. Disable this only if your target machine is
1056 unaffected by PCI quirks.
1060 bool "Enable SLUB debugging support" if EMBEDDED
1061 depends on SLUB && SYSFS
1063 SLUB has extensive debug support features. Disabling these can
1064 result in significant savings in code size. This also disables
1065 SLUB sysfs support. /sys/slab will not exist and there will be
1066 no support for cache validation etc.
1069 bool "Disable heap randomization"
1072 Randomizing heap placement makes heap exploits harder, but it
1073 also breaks ancient binaries (including anything libc5 based).
1074 This option changes the bootup default to heap randomization
1075 disabled, and can be overridden at runtime by setting
1076 /proc/sys/kernel/randomize_va_space to 2.
1078 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1081 prompt "Choose SLAB allocator"
1084 This option allows to select a slab allocator.
1089 The regular slab allocator that is established and known to work
1090 well in all environments. It organizes cache hot objects in
1091 per cpu and per node queues.
1094 bool "SLUB (Unqueued Allocator)"
1096 SLUB is a slab allocator that minimizes cache line usage
1097 instead of managing queues of cached objects (SLAB approach).
1098 Per cpu caching is realized using slabs of objects instead
1099 of queues of objects. SLUB can use memory efficiently
1100 and has enhanced diagnostics. SLUB is the default choice for
1105 bool "SLOB (Simple Allocator)"
1107 SLOB replaces the stock allocator with a drastically simpler
1108 allocator. SLOB is generally more space efficient but
1109 does not perform as well on large systems.
1113 config MMAP_ALLOW_UNINITIALIZED
1114 bool "Allow mmapped anonymous memory to be uninitialized"
1115 depends on EMBEDDED && !MMU
1118 Normally, and according to the Linux spec, anonymous memory obtained
1119 from mmap() has it's contents cleared before it is passed to
1120 userspace. Enabling this config option allows you to request that
1121 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1122 providing a huge performance boost. If this option is not enabled,
1123 then the flag will be ignored.
1125 This is taken advantage of by uClibc's malloc(), and also by
1126 ELF-FDPIC binfmt's brk and stack allocator.
1128 Because of the obvious security issues, this option should only be
1129 enabled on embedded devices where you control what is run in
1130 userspace. Since that isn't generally a problem on no-MMU systems,
1131 it is normally safe to say Y here.
1133 See Documentation/nommu-mmap.txt for more information.
1136 bool "Profiling support"
1138 Say Y here to enable the extended profiling support mechanisms used
1139 by profilers such as OProfile.
1142 # Place an empty function call at each tracepoint site. Can be
1143 # dynamically changed for a probe function.
1148 source "arch/Kconfig"
1150 endmenu # General setup
1152 config HAVE_GENERIC_DMA_COHERENT
1159 depends on SLAB || SLUB_DEBUG
1167 default 0 if BASE_FULL
1168 default 1 if !BASE_FULL
1171 bool "Enable loadable module support"
1173 Kernel modules are small pieces of compiled code which can
1174 be inserted in the running kernel, rather than being
1175 permanently built into the kernel. You use the "modprobe"
1176 tool to add (and sometimes remove) them. If you say Y here,
1177 many parts of the kernel can be built as modules (by
1178 answering M instead of Y where indicated): this is most
1179 useful for infrequently used options which are not required
1180 for booting. For more information, see the man pages for
1181 modprobe, lsmod, modinfo, insmod and rmmod.
1183 If you say Y here, you will need to run "make
1184 modules_install" to put the modules under /lib/modules/
1185 where modprobe can find them (you may need to be root to do
1192 config MODULE_FORCE_LOAD
1193 bool "Forced module loading"
1196 Allow loading of modules without version information (ie. modprobe
1197 --force). Forced module loading sets the 'F' (forced) taint flag and
1198 is usually a really bad idea.
1200 config MODULE_UNLOAD
1201 bool "Module unloading"
1203 Without this option you will not be able to unload any
1204 modules (note that some modules may not be unloadable
1205 anyway), which makes your kernel smaller, faster
1206 and simpler. If unsure, say Y.
1208 config MODULE_FORCE_UNLOAD
1209 bool "Forced module unloading"
1210 depends on MODULE_UNLOAD && EXPERIMENTAL
1212 This option allows you to force a module to unload, even if the
1213 kernel believes it is unsafe: the kernel will remove the module
1214 without waiting for anyone to stop using it (using the -f option to
1215 rmmod). This is mainly for kernel developers and desperate users.
1219 bool "Module versioning support"
1221 Usually, you have to use modules compiled with your kernel.
1222 Saying Y here makes it sometimes possible to use modules
1223 compiled for different kernels, by adding enough information
1224 to the modules to (hopefully) spot any changes which would
1225 make them incompatible with the kernel you are running. If
1228 config MODULE_SRCVERSION_ALL
1229 bool "Source checksum for all modules"
1231 Modules which contain a MODULE_VERSION get an extra "srcversion"
1232 field inserted into their modinfo section, which contains a
1233 sum of the source files which made it. This helps maintainers
1234 see exactly which source was used to build a module (since
1235 others sometimes change the module source without updating
1236 the version). With this option, such a "srcversion" field
1237 will be created for all modules. If unsure, say N.
1241 config INIT_ALL_POSSIBLE
1244 Back when each arch used to define their own cpu_online_map and
1245 cpu_possible_map, some of them chose to initialize cpu_possible_map
1246 with all 1s, and others with all 0s. When they were centralised,
1247 it was better to provide this option than to break all the archs
1248 and have several arch maintainers pursuing me down dark alleys.
1253 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU
1255 Need stop_machine() primitive.
1257 source "block/Kconfig"
1259 config PREEMPT_NOTIFIERS
1266 source "kernel/Kconfig.locks"