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"
29 depends on HAVE_IRQ_WORK
34 bool "Prompt for development and/or incomplete code/drivers"
36 Some of the various things that Linux supports (such as network
37 drivers, file systems, network protocols, etc.) can be in a state
38 of development where the functionality, stability, or the level of
39 testing is not yet high enough for general use. This is usually
40 known as the "alpha-test" phase among developers. If a feature is
41 currently in alpha-test, then the developers usually discourage
42 uninformed widespread use of this feature by the general public to
43 avoid "Why doesn't this work?" type mail messages. However, active
44 testing and use of these systems is welcomed. Just be aware that it
45 may not meet the normal level of reliability or it may fail to work
46 in some special cases. Detailed bug reports from people familiar
47 with the kernel internals are usually welcomed by the developers
48 (before submitting bug reports, please read the documents
49 <file:README>, <file:MAINTAINERS>, <file:REPORTING-BUGS>,
50 <file:Documentation/BUG-HUNTING>, and
51 <file:Documentation/oops-tracing.txt> in the kernel source).
53 This option will also make obsoleted drivers available. These are
54 drivers that have been replaced by something else, and/or are
55 scheduled to be removed in a future kernel release.
57 Unless you intend to help test and develop a feature or driver that
58 falls into this category, or you have a situation that requires
59 using these features, you should probably say N here, which will
60 cause the configurator to present you with fewer choices. If
61 you say Y here, you will be offered the choice of using features or
62 drivers that are currently considered to be in the alpha-test phase.
69 depends on BROKEN || !SMP
74 depends on (SMP || PREEMPT) && BKL
77 config INIT_ENV_ARG_LIMIT
82 Maximum of each of the number of arguments and environment
83 variables passed to init from the kernel command line.
87 string "Cross-compiler tool prefix"
89 Same as running 'make CROSS_COMPILE=prefix-' but stored for
90 default make runs in this kernel build directory. You don't
91 need to set this unless you want the configured kernel build
92 directory to select the cross-compiler automatically.
95 string "Local version - append to kernel release"
97 Append an extra string to the end of your kernel version.
98 This will show up when you type uname, for example.
99 The string you set here will be appended after the contents of
100 any files with a filename matching localversion* in your
101 object and source tree, in that order. Your total string can
102 be a maximum of 64 characters.
104 config LOCALVERSION_AUTO
105 bool "Automatically append version information to the version string"
108 This will try to automatically determine if the current tree is a
109 release tree by looking for git tags that belong to the current
110 top of tree revision.
112 A string of the format -gxxxxxxxx will be added to the localversion
113 if a git-based tree is found. The string generated by this will be
114 appended after any matching localversion* files, and after the value
115 set in CONFIG_LOCALVERSION.
117 (The actual string used here is the first eight characters produced
118 by running the command:
120 $ git rev-parse --verify HEAD
122 which is done within the script "scripts/setlocalversion".)
124 config HAVE_KERNEL_GZIP
127 config HAVE_KERNEL_BZIP2
130 config HAVE_KERNEL_LZMA
133 config HAVE_KERNEL_XZ
136 config HAVE_KERNEL_LZO
140 prompt "Kernel compression mode"
142 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO
144 The linux kernel is a kind of self-extracting executable.
145 Several compression algorithms are available, which differ
146 in efficiency, compression and decompression speed.
147 Compression speed is only relevant when building a kernel.
148 Decompression speed is relevant at each boot.
150 If you have any problems with bzip2 or lzma compressed
151 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
152 version of this functionality (bzip2 only), for 2.4, was
153 supplied by Christian Ludwig)
155 High compression options are mostly useful for users, who
156 are low on disk space (embedded systems), but for whom ram
159 If in doubt, select 'gzip'
163 depends on HAVE_KERNEL_GZIP
165 The old and tried gzip compression. It provides a good balance
166 between compression ratio and decompression speed.
170 depends on HAVE_KERNEL_BZIP2
172 Its compression ratio and speed is intermediate.
173 Decompression speed is slowest among the three. The kernel
174 size is about 10% smaller with bzip2, in comparison to gzip.
175 Bzip2 uses a large amount of memory. For modern kernels you
176 will need at least 8MB RAM or more for booting.
180 depends on HAVE_KERNEL_LZMA
182 The most recent compression algorithm.
183 Its ratio is best, decompression speed is between the other
184 two. Compression is slowest. The kernel size is about 33%
185 smaller with LZMA in comparison to gzip.
189 depends on HAVE_KERNEL_XZ
191 XZ uses the LZMA2 algorithm and instruction set specific
192 BCJ filters which can improve compression ratio of executable
193 code. The size of the kernel is about 30% smaller with XZ in
194 comparison to gzip. On architectures for which there is a BCJ
195 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
196 will create a few percent smaller kernel than plain LZMA.
198 The speed is about the same as with LZMA: The decompression
199 speed of XZ is better than that of bzip2 but worse than gzip
200 and LZO. Compression is slow.
204 depends on HAVE_KERNEL_LZO
206 Its compression ratio is the poorest among the 4. The kernel
207 size is about 10% bigger than gzip; however its speed
208 (both compression and decompression) is the fastest.
213 bool "Support for paging of anonymous memory (swap)"
214 depends on MMU && BLOCK
217 This option allows you to choose whether you want to have support
218 for so called swap devices or swap files in your kernel that are
219 used to provide more virtual memory than the actual RAM present
220 in your computer. If unsure say Y.
225 Inter Process Communication is a suite of library functions and
226 system calls which let processes (running programs) synchronize and
227 exchange information. It is generally considered to be a good thing,
228 and some programs won't run unless you say Y here. In particular, if
229 you want to run the DOS emulator dosemu under Linux (read the
230 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
231 you'll need to say Y here.
233 You can find documentation about IPC with "info ipc" and also in
234 section 6.4 of the Linux Programmer's Guide, available from
235 <http://www.tldp.org/guides.html>.
237 config SYSVIPC_SYSCTL
244 bool "POSIX Message Queues"
245 depends on NET && EXPERIMENTAL
247 POSIX variant of message queues is a part of IPC. In POSIX message
248 queues every message has a priority which decides about succession
249 of receiving it by a process. If you want to compile and run
250 programs written e.g. for Solaris with use of its POSIX message
251 queues (functions mq_*) say Y here.
253 POSIX message queues are visible as a filesystem called 'mqueue'
254 and can be mounted somewhere if you want to do filesystem
255 operations on message queues.
259 config POSIX_MQUEUE_SYSCTL
261 depends on POSIX_MQUEUE
265 config BSD_PROCESS_ACCT
266 bool "BSD Process Accounting"
268 If you say Y here, a user level program will be able to instruct the
269 kernel (via a special system call) to write process accounting
270 information to a file: whenever a process exits, information about
271 that process will be appended to the file by the kernel. The
272 information includes things such as creation time, owning user,
273 command name, memory usage, controlling terminal etc. (the complete
274 list is in the struct acct in <file:include/linux/acct.h>). It is
275 up to the user level program to do useful things with this
276 information. This is generally a good idea, so say Y.
278 config BSD_PROCESS_ACCT_V3
279 bool "BSD Process Accounting version 3 file format"
280 depends on BSD_PROCESS_ACCT
283 If you say Y here, the process accounting information is written
284 in a new file format that also logs the process IDs of each
285 process and it's parent. Note that this file format is incompatible
286 with previous v0/v1/v2 file formats, so you will need updated tools
287 for processing it. A preliminary version of these tools is available
288 at <http://www.gnu.org/software/acct/>.
291 bool "open by fhandle syscalls"
294 If you say Y here, a user level program will be able to map
295 file names to handle and then later use the handle for
296 different file system operations. This is useful in implementing
297 userspace file servers, which now track files using handles instead
298 of names. The handle would remain the same even if file names
299 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
303 bool "Export task/process statistics through netlink (EXPERIMENTAL)"
307 Export selected statistics for tasks/processes through the
308 generic netlink interface. Unlike BSD process accounting, the
309 statistics are available during the lifetime of tasks/processes as
310 responses to commands. Like BSD accounting, they are sent to user
315 config TASK_DELAY_ACCT
316 bool "Enable per-task delay accounting (EXPERIMENTAL)"
319 Collect information on time spent by a task waiting for system
320 resources like cpu, synchronous block I/O completion and swapping
321 in pages. Such statistics can help in setting a task's priorities
322 relative to other tasks for cpu, io, rss limits etc.
327 bool "Enable extended accounting over taskstats (EXPERIMENTAL)"
330 Collect extended task accounting data and send the data
331 to userland for processing over the taskstats interface.
335 config TASK_IO_ACCOUNTING
336 bool "Enable per-task storage I/O accounting (EXPERIMENTAL)"
337 depends on TASK_XACCT
339 Collect information on the number of bytes of storage I/O which this
345 bool "Auditing support"
348 Enable auditing infrastructure that can be used with another
349 kernel subsystem, such as SELinux (which requires this for
350 logging of avc messages output). Does not do system-call
351 auditing without CONFIG_AUDITSYSCALL.
354 bool "Enable system-call auditing support"
355 depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH)
356 default y if SECURITY_SELINUX
358 Enable low-overhead system-call auditing infrastructure that
359 can be used independently or with another kernel subsystem,
364 depends on AUDITSYSCALL
369 depends on AUDITSYSCALL
372 source "kernel/irq/Kconfig"
377 prompt "RCU Implementation"
381 bool "Tree-based hierarchical RCU"
382 depends on !PREEMPT && SMP
384 This option selects the RCU implementation that is
385 designed for very large SMP system with hundreds or
386 thousands of CPUs. It also scales down nicely to
389 config TREE_PREEMPT_RCU
390 bool "Preemptible tree-based hierarchical RCU"
393 This option selects the RCU implementation that is
394 designed for very large SMP systems with hundreds or
395 thousands of CPUs, but for which real-time response
396 is also required. It also scales down nicely to
400 bool "UP-only small-memory-footprint RCU"
403 This option selects the RCU implementation that is
404 designed for UP systems from which real-time response
405 is not required. This option greatly reduces the
406 memory footprint of RCU.
408 config TINY_PREEMPT_RCU
409 bool "Preemptible UP-only small-memory-footprint RCU"
410 depends on !SMP && PREEMPT
412 This option selects the RCU implementation that is designed
413 for real-time UP systems. This option greatly reduces the
414 memory footprint of RCU.
419 def_bool ( TREE_PREEMPT_RCU || TINY_PREEMPT_RCU )
421 This option enables preemptible-RCU code that is common between
422 the TREE_PREEMPT_RCU and TINY_PREEMPT_RCU implementations.
425 bool "Enable tracing for RCU"
427 This option provides tracing in RCU which presents stats
428 in debugfs for debugging RCU implementation.
430 Say Y here if you want to enable RCU tracing
431 Say N if you are unsure.
434 int "Tree-based hierarchical RCU fanout value"
437 depends on TREE_RCU || TREE_PREEMPT_RCU
441 This option controls the fanout of hierarchical implementations
442 of RCU, allowing RCU to work efficiently on machines with
443 large numbers of CPUs. This value must be at least the fourth
444 root of NR_CPUS, which allows NR_CPUS to be insanely large.
445 The default value of RCU_FANOUT should be used for production
446 systems, but if you are stress-testing the RCU implementation
447 itself, small RCU_FANOUT values allow you to test large-system
448 code paths on small(er) systems.
450 Select a specific number if testing RCU itself.
451 Take the default if unsure.
453 config RCU_FANOUT_EXACT
454 bool "Disable tree-based hierarchical RCU auto-balancing"
455 depends on TREE_RCU || TREE_PREEMPT_RCU
458 This option forces use of the exact RCU_FANOUT value specified,
459 regardless of imbalances in the hierarchy. This is useful for
460 testing RCU itself, and might one day be useful on systems with
461 strong NUMA behavior.
463 Without RCU_FANOUT_EXACT, the code will balance the hierarchy.
467 config RCU_FAST_NO_HZ
468 bool "Accelerate last non-dyntick-idle CPU's grace periods"
469 depends on TREE_RCU && NO_HZ && SMP
472 This option causes RCU to attempt to accelerate grace periods
473 in order to allow the final CPU to enter dynticks-idle state
474 more quickly. On the other hand, this option increases the
475 overhead of the dynticks-idle checking, particularly on systems
476 with large numbers of CPUs.
478 Say Y if energy efficiency is critically important, particularly
479 if you have relatively few CPUs.
481 Say N if you are unsure.
483 config TREE_RCU_TRACE
484 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU )
487 This option provides tracing for the TREE_RCU and
488 TREE_PREEMPT_RCU implementations, permitting Makefile to
489 trivially select kernel/rcutree_trace.c.
492 bool "Enable RCU priority boosting"
493 depends on RT_MUTEXES && TINY_PREEMPT_RCU
496 This option boosts the priority of preempted RCU readers that
497 block the current preemptible RCU grace period for too long.
498 This option also prevents heavy loads from blocking RCU
499 callback invocation for all flavors of RCU.
501 Say Y here if you are working with real-time apps or heavy loads
502 Say N here if you are unsure.
504 config RCU_BOOST_PRIO
505 int "Real-time priority to boost RCU readers to"
510 This option specifies the real-time priority to which preempted
511 RCU readers are to be boosted. If you are working with CPU-bound
512 real-time applications, you should specify a priority higher then
513 the highest-priority CPU-bound application.
515 Specify the real-time priority, or take the default if unsure.
517 config RCU_BOOST_DELAY
518 int "Milliseconds to delay boosting after RCU grace-period start"
523 This option specifies the time to wait after the beginning of
524 a given grace period before priority-boosting preempted RCU
525 readers blocking that grace period. Note that any RCU reader
526 blocking an expedited RCU grace period is boosted immediately.
528 Accept the default if unsure.
530 endmenu # "RCU Subsystem"
533 tristate "Kernel .config support"
535 This option enables the complete Linux kernel ".config" file
536 contents to be saved in the kernel. It provides documentation
537 of which kernel options are used in a running kernel or in an
538 on-disk kernel. This information can be extracted from the kernel
539 image file with the script scripts/extract-ikconfig and used as
540 input to rebuild the current kernel or to build another kernel.
541 It can also be extracted from a running kernel by reading
542 /proc/config.gz if enabled (below).
545 bool "Enable access to .config through /proc/config.gz"
546 depends on IKCONFIG && PROC_FS
548 This option enables access to the kernel configuration file
549 through /proc/config.gz.
552 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
556 Select kernel log buffer size as a power of 2.
566 # Architectures with an unreliable sched_clock() should select this:
568 config HAVE_UNSTABLE_SCHED_CLOCK
572 boolean "Control Group support"
575 This option adds support for grouping sets of processes together, for
576 use with process control subsystems such as Cpusets, CFS, memory
577 controls or device isolation.
579 - Documentation/scheduler/sched-design-CFS.txt (CFS)
580 - Documentation/cgroups/ (features for grouping, isolation
581 and resource control)
588 bool "Example debug cgroup subsystem"
591 This option enables a simple cgroup subsystem that
592 exports useful debugging information about the cgroups
598 bool "Namespace cgroup subsystem"
600 Provides a simple namespace cgroup subsystem to
601 provide hierarchical naming of sets of namespaces,
602 for instance virtual servers and checkpoint/restart
605 config CGROUP_FREEZER
606 bool "Freezer cgroup subsystem"
608 Provides a way to freeze and unfreeze all tasks in a
612 bool "Device controller for cgroups"
614 Provides a cgroup implementing whitelists for devices which
615 a process in the cgroup can mknod or open.
618 bool "Cpuset support"
620 This option will let you create and manage CPUSETs which
621 allow dynamically partitioning a system into sets of CPUs and
622 Memory Nodes and assigning tasks to run only within those sets.
623 This is primarily useful on large SMP or NUMA systems.
627 config PROC_PID_CPUSET
628 bool "Include legacy /proc/<pid>/cpuset file"
632 config CGROUP_CPUACCT
633 bool "Simple CPU accounting cgroup subsystem"
635 Provides a simple Resource Controller for monitoring the
636 total CPU consumed by the tasks in a cgroup.
638 config RESOURCE_COUNTERS
639 bool "Resource counters"
641 This option enables controller independent resource accounting
642 infrastructure that works with cgroups.
644 config CGROUP_MEM_RES_CTLR
645 bool "Memory Resource Controller for Control Groups"
646 depends on RESOURCE_COUNTERS
649 Provides a memory resource controller that manages both anonymous
650 memory and page cache. (See Documentation/cgroups/memory.txt)
652 Note that setting this option increases fixed memory overhead
653 associated with each page of memory in the system. By this,
654 20(40)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory
655 usage tracking struct at boot. Total amount of this is printed out
658 Only enable when you're ok with these trade offs and really
659 sure you need the memory resource controller. Even when you enable
660 this, you can set "cgroup_disable=memory" at your boot option to
661 disable memory resource controller and you can avoid overheads.
662 (and lose benefits of memory resource controller)
664 This config option also selects MM_OWNER config option, which
665 could in turn add some fork/exit overhead.
667 config CGROUP_MEM_RES_CTLR_SWAP
668 bool "Memory Resource Controller Swap Extension"
669 depends on CGROUP_MEM_RES_CTLR && SWAP
671 Add swap management feature to memory resource controller. When you
672 enable this, you can limit mem+swap usage per cgroup. In other words,
673 when you disable this, memory resource controller has no cares to
674 usage of swap...a process can exhaust all of the swap. This extension
675 is useful when you want to avoid exhaustion swap but this itself
676 adds more overheads and consumes memory for remembering information.
677 Especially if you use 32bit system or small memory system, please
678 be careful about enabling this. When memory resource controller
679 is disabled by boot option, this will be automatically disabled and
680 there will be no overhead from this. Even when you set this config=y,
681 if boot option "noswapaccount" is set, swap will not be accounted.
682 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page
683 size is 4096bytes, 512k per 1Gbytes of swap.
684 config CGROUP_MEM_RES_CTLR_SWAP_ENABLED
685 bool "Memory Resource Controller Swap Extension enabled by default"
686 depends on CGROUP_MEM_RES_CTLR_SWAP
689 Memory Resource Controller Swap Extension comes with its price in
690 a bigger memory consumption. General purpose distribution kernels
691 which want to enable the feature but keep it disabled by default
692 and let the user enable it by swapaccount boot command line
693 parameter should have this option unselected.
694 For those who want to have the feature enabled by default should
695 select this option (if, for some reason, they need to disable it
696 then noswapaccount does the trick).
698 menuconfig CGROUP_SCHED
699 bool "Group CPU scheduler"
700 depends on EXPERIMENTAL
703 This feature lets CPU scheduler recognize task groups and control CPU
704 bandwidth allocation to such task groups. It uses cgroups to group
708 config FAIR_GROUP_SCHED
709 bool "Group scheduling for SCHED_OTHER"
710 depends on CGROUP_SCHED
713 config RT_GROUP_SCHED
714 bool "Group scheduling for SCHED_RR/FIFO"
715 depends on EXPERIMENTAL
716 depends on CGROUP_SCHED
719 This feature lets you explicitly allocate real CPU bandwidth
720 to task groups. If enabled, it will also make it impossible to
721 schedule realtime tasks for non-root users until you allocate
722 realtime bandwidth for them.
723 See Documentation/scheduler/sched-rt-group.txt for more information.
728 tristate "Block IO controller"
732 Generic block IO controller cgroup interface. This is the common
733 cgroup interface which should be used by various IO controlling
736 Currently, CFQ IO scheduler uses it to recognize task groups and
737 control disk bandwidth allocation (proportional time slice allocation)
738 to such task groups. It is also used by bio throttling logic in
739 block layer to implement upper limit in IO rates on a device.
741 This option only enables generic Block IO controller infrastructure.
742 One needs to also enable actual IO controlling logic/policy. For
743 enabling proportional weight division of disk bandwidth in CFQ seti
744 CONFIG_CFQ_GROUP_IOSCHED=y and for enabling throttling policy set
745 CONFIG_BLK_THROTTLE=y.
747 See Documentation/cgroups/blkio-controller.txt for more information.
749 config DEBUG_BLK_CGROUP
750 bool "Enable Block IO controller debugging"
751 depends on BLK_CGROUP
754 Enable some debugging help. Currently it exports additional stat
755 files in a cgroup which can be useful for debugging.
759 menuconfig NAMESPACES
760 bool "Namespaces support" if EXPERT
763 Provides the way to make tasks work with different objects using
764 the same id. For example same IPC id may refer to different objects
765 or same user id or pid may refer to different tasks when used in
766 different namespaces.
774 In this namespace tasks see different info provided with the
779 depends on (SYSVIPC || POSIX_MQUEUE)
782 In this namespace tasks work with IPC ids which correspond to
783 different IPC objects in different namespaces.
786 bool "User namespace (EXPERIMENTAL)"
787 depends on EXPERIMENTAL
790 This allows containers, i.e. vservers, to use user namespaces
791 to provide different user info for different servers.
795 bool "PID Namespaces"
798 Support process id namespaces. This allows having multiple
799 processes with the same pid as long as they are in different
800 pid namespaces. This is a building block of containers.
803 bool "Network namespace"
807 Allow user space to create what appear to be multiple instances
808 of the network stack.
812 config SCHED_AUTOGROUP
813 bool "Automatic process group scheduling"
817 select FAIR_GROUP_SCHED
819 This option optimizes the scheduler for common desktop workloads by
820 automatically creating and populating task groups. This separation
821 of workloads isolates aggressive CPU burners (like build jobs) from
822 desktop applications. Task group autogeneration is currently based
828 config SYSFS_DEPRECATED
829 bool "enable deprecated sysfs features to support old userspace tools"
833 This option adds code that switches the layout of the "block" class
834 devices, to not show up in /sys/class/block/, but only in
837 This switch is only active when the sysfs.deprecated=1 boot option is
838 passed or the SYSFS_DEPRECATED_V2 option is set.
840 This option allows new kernels to run on old distributions and tools,
841 which might get confused by /sys/class/block/. Since 2007/2008 all
842 major distributions and tools handle this just fine.
844 Recent distributions and userspace tools after 2009/2010 depend on
845 the existence of /sys/class/block/, and will not work with this
848 Only if you are using a new kernel on an old distribution, you might
851 config SYSFS_DEPRECATED_V2
852 bool "enabled deprecated sysfs features by default"
855 depends on SYSFS_DEPRECATED
857 Enable deprecated sysfs by default.
859 See the CONFIG_SYSFS_DEPRECATED option for more details about this
862 Only if you are using a new kernel on an old distribution, you might
863 need to say Y here. Even then, odds are you would not need it
864 enabled, you can always pass the boot option if absolutely necessary.
867 bool "Kernel->user space relay support (formerly relayfs)"
869 This option enables support for relay interface support in
870 certain file systems (such as debugfs).
871 It is designed to provide an efficient mechanism for tools and
872 facilities to relay large amounts of data from kernel space to
877 config BLK_DEV_INITRD
878 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
879 depends on BROKEN || !FRV
881 The initial RAM filesystem is a ramfs which is loaded by the
882 boot loader (loadlin or lilo) and that is mounted as root
883 before the normal boot procedure. It is typically used to
884 load modules needed to mount the "real" root file system,
885 etc. See <file:Documentation/initrd.txt> for details.
887 If RAM disk support (BLK_DEV_RAM) is also included, this
888 also enables initial RAM disk (initrd) support and adds
889 15 Kbytes (more on some other architectures) to the kernel size.
899 config CC_OPTIMIZE_FOR_SIZE
900 bool "Optimize for size"
903 Enabling this option will pass "-Os" instead of "-O2" to gcc
904 resulting in a smaller kernel.
915 bool "Configure standard kernel features (expert users)"
917 This option allows certain base kernel options and settings
918 to be disabled or tweaked. This is for specialized
919 environments which can tolerate a "non-standard" kernel.
920 Only use this if you really know what you are doing.
923 bool "Embedded system"
926 This option should be enabled if compiling the kernel for
927 an embedded system so certain expert options are available
931 bool "Enable 16-bit UID system calls" if EXPERT
932 depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && COMPAT) || UML || (X86_64 && IA32_EMULATION)
935 This enables the legacy 16-bit UID syscall wrappers.
937 config SYSCTL_SYSCALL
938 bool "Sysctl syscall support" if EXPERT
939 depends on PROC_SYSCTL
943 sys_sysctl uses binary paths that have been found challenging
944 to properly maintain and use. The interface in /proc/sys
945 using paths with ascii names is now the primary path to this
948 Almost nothing using the binary sysctl interface so if you are
949 trying to save some space it is probably safe to disable this,
950 making your kernel marginally smaller.
952 If unsure say Y here.
955 bool "Load all symbols for debugging/ksymoops" if EXPERT
958 Say Y here to let the kernel print out symbolic crash information and
959 symbolic stack backtraces. This increases the size of the kernel
960 somewhat, as all symbols have to be loaded into the kernel image.
963 bool "Include all symbols in kallsyms"
964 depends on DEBUG_KERNEL && KALLSYMS
966 Normally kallsyms only contains the symbols of functions, for nicer
967 OOPS messages. Some debuggers can use kallsyms for other
968 symbols too: say Y here to include all symbols, if you need them
969 and you don't care about adding 300k to the size of your kernel.
973 config KALLSYMS_EXTRA_PASS
974 bool "Do an extra kallsyms pass"
977 If kallsyms is not working correctly, the build will fail with
978 inconsistent kallsyms data. If that occurs, log a bug report and
979 turn on KALLSYMS_EXTRA_PASS which should result in a stable build.
980 Always say N here unless you find a bug in kallsyms, which must be
981 reported. KALLSYMS_EXTRA_PASS is only a temporary workaround while
982 you wait for kallsyms to be fixed.
986 bool "Support for hot-pluggable devices" if EXPERT
989 This option is provided for the case where no hotplug or uevent
990 capabilities is wanted by the kernel. You should only consider
991 disabling this option for embedded systems that do not use modules, a
992 dynamic /dev tree, or dynamic device discovery. Just say Y.
996 bool "Enable support for printk" if EXPERT
998 This option enables normal printk support. Removing it
999 eliminates most of the message strings from the kernel image
1000 and makes the kernel more or less silent. As this makes it
1001 very difficult to diagnose system problems, saying N here is
1002 strongly discouraged.
1005 bool "BUG() support" if EXPERT
1008 Disabling this option eliminates support for BUG and WARN, reducing
1009 the size of your kernel image and potentially quietly ignoring
1010 numerous fatal conditions. You should only consider disabling this
1011 option for embedded systems with no facilities for reporting errors.
1016 bool "Enable ELF core dumps" if EXPERT
1018 Enable support for generating core dumps. Disabling saves about 4k.
1020 config PCSPKR_PLATFORM
1021 bool "Enable PC-Speaker support" if EXPERT
1022 depends on ALPHA || X86 || MIPS || PPC_PREP || PPC_CHRP || PPC_PSERIES
1025 This option allows to disable the internal PC-Speaker
1026 support, saving some memory.
1030 bool "Enable full-sized data structures for core" if EXPERT
1032 Disabling this option reduces the size of miscellaneous core
1033 kernel data structures. This saves memory on small machines,
1034 but may reduce performance.
1037 bool "Enable futex support" if EXPERT
1041 Disabling this option will cause the kernel to be built without
1042 support for "fast userspace mutexes". The resulting kernel may not
1043 run glibc-based applications correctly.
1046 bool "Enable eventpoll support" if EXPERT
1050 Disabling this option will cause the kernel to be built without
1051 support for epoll family of system calls.
1054 bool "Enable signalfd() system call" if EXPERT
1058 Enable the signalfd() system call that allows to receive signals
1059 on a file descriptor.
1064 bool "Enable timerfd() system call" if EXPERT
1068 Enable the timerfd() system call that allows to receive timer
1069 events on a file descriptor.
1074 bool "Enable eventfd() system call" if EXPERT
1078 Enable the eventfd() system call that allows to receive both
1079 kernel notification (ie. KAIO) or userspace notifications.
1084 bool "Use full shmem filesystem" if EXPERT
1088 The shmem is an internal filesystem used to manage shared memory.
1089 It is backed by swap and manages resource limits. It is also exported
1090 to userspace as tmpfs if TMPFS is enabled. Disabling this
1091 option replaces shmem and tmpfs with the much simpler ramfs code,
1092 which may be appropriate on small systems without swap.
1095 bool "Enable AIO support" if EXPERT
1098 This option enables POSIX asynchronous I/O which may by used
1099 by some high performance threaded applications. Disabling
1100 this option saves about 7k.
1102 config HAVE_PERF_EVENTS
1105 See tools/perf/design.txt for details.
1107 config PERF_USE_VMALLOC
1110 See tools/perf/design.txt for details
1112 menu "Kernel Performance Events And Counters"
1115 bool "Kernel performance events and counters"
1116 default y if (PROFILING || PERF_COUNTERS)
1117 depends on HAVE_PERF_EVENTS
1121 Enable kernel support for various performance events provided
1122 by software and hardware.
1124 Software events are supported either built-in or via the
1125 use of generic tracepoints.
1127 Most modern CPUs support performance events via performance
1128 counter registers. These registers count the number of certain
1129 types of hw events: such as instructions executed, cachemisses
1130 suffered, or branches mis-predicted - without slowing down the
1131 kernel or applications. These registers can also trigger interrupts
1132 when a threshold number of events have passed - and can thus be
1133 used to profile the code that runs on that CPU.
1135 The Linux Performance Event subsystem provides an abstraction of
1136 these software and hardware event capabilities, available via a
1137 system call and used by the "perf" utility in tools/perf/. It
1138 provides per task and per CPU counters, and it provides event
1139 capabilities on top of those.
1143 config PERF_COUNTERS
1144 bool "Kernel performance counters (old config option)"
1145 depends on HAVE_PERF_EVENTS
1147 This config has been obsoleted by the PERF_EVENTS
1148 config option - please see that one for details.
1150 It has no effect on the kernel whether you enable
1151 it or not, it is a compatibility placeholder.
1155 config DEBUG_PERF_USE_VMALLOC
1157 bool "Debug: use vmalloc to back perf mmap() buffers"
1158 depends on PERF_EVENTS && DEBUG_KERNEL
1159 select PERF_USE_VMALLOC
1161 Use vmalloc memory to back perf mmap() buffers.
1163 Mostly useful for debugging the vmalloc code on platforms
1164 that don't require it.
1170 config VM_EVENT_COUNTERS
1172 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1174 VM event counters are needed for event counts to be shown.
1175 This option allows the disabling of the VM event counters
1176 on EXPERT systems. /proc/vmstat will only show page counts
1177 if VM event counters are disabled.
1181 bool "Enable PCI quirk workarounds" if EXPERT
1184 This enables workarounds for various PCI chipset
1185 bugs/quirks. Disable this only if your target machine is
1186 unaffected by PCI quirks.
1190 bool "Enable SLUB debugging support" if EXPERT
1191 depends on SLUB && SYSFS
1193 SLUB has extensive debug support features. Disabling these can
1194 result in significant savings in code size. This also disables
1195 SLUB sysfs support. /sys/slab will not exist and there will be
1196 no support for cache validation etc.
1199 bool "Disable heap randomization"
1202 Randomizing heap placement makes heap exploits harder, but it
1203 also breaks ancient binaries (including anything libc5 based).
1204 This option changes the bootup default to heap randomization
1205 disabled, and can be overridden at runtime by setting
1206 /proc/sys/kernel/randomize_va_space to 2.
1208 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1211 prompt "Choose SLAB allocator"
1214 This option allows to select a slab allocator.
1219 The regular slab allocator that is established and known to work
1220 well in all environments. It organizes cache hot objects in
1221 per cpu and per node queues.
1224 bool "SLUB (Unqueued Allocator)"
1226 SLUB is a slab allocator that minimizes cache line usage
1227 instead of managing queues of cached objects (SLAB approach).
1228 Per cpu caching is realized using slabs of objects instead
1229 of queues of objects. SLUB can use memory efficiently
1230 and has enhanced diagnostics. SLUB is the default choice for
1235 bool "SLOB (Simple Allocator)"
1237 SLOB replaces the stock allocator with a drastically simpler
1238 allocator. SLOB is generally more space efficient but
1239 does not perform as well on large systems.
1243 config MMAP_ALLOW_UNINITIALIZED
1244 bool "Allow mmapped anonymous memory to be uninitialized"
1245 depends on EXPERT && !MMU
1248 Normally, and according to the Linux spec, anonymous memory obtained
1249 from mmap() has it's contents cleared before it is passed to
1250 userspace. Enabling this config option allows you to request that
1251 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1252 providing a huge performance boost. If this option is not enabled,
1253 then the flag will be ignored.
1255 This is taken advantage of by uClibc's malloc(), and also by
1256 ELF-FDPIC binfmt's brk and stack allocator.
1258 Because of the obvious security issues, this option should only be
1259 enabled on embedded devices where you control what is run in
1260 userspace. Since that isn't generally a problem on no-MMU systems,
1261 it is normally safe to say Y here.
1263 See Documentation/nommu-mmap.txt for more information.
1266 bool "Profiling support"
1268 Say Y here to enable the extended profiling support mechanisms used
1269 by profilers such as OProfile.
1272 # Place an empty function call at each tracepoint site. Can be
1273 # dynamically changed for a probe function.
1278 source "arch/Kconfig"
1280 endmenu # General setup
1282 config HAVE_GENERIC_DMA_COHERENT
1289 depends on SLAB || SLUB_DEBUG
1297 default 0 if BASE_FULL
1298 default 1 if !BASE_FULL
1301 bool "Enable loadable module support"
1303 Kernel modules are small pieces of compiled code which can
1304 be inserted in the running kernel, rather than being
1305 permanently built into the kernel. You use the "modprobe"
1306 tool to add (and sometimes remove) them. If you say Y here,
1307 many parts of the kernel can be built as modules (by
1308 answering M instead of Y where indicated): this is most
1309 useful for infrequently used options which are not required
1310 for booting. For more information, see the man pages for
1311 modprobe, lsmod, modinfo, insmod and rmmod.
1313 If you say Y here, you will need to run "make
1314 modules_install" to put the modules under /lib/modules/
1315 where modprobe can find them (you may need to be root to do
1322 config MODULE_FORCE_LOAD
1323 bool "Forced module loading"
1326 Allow loading of modules without version information (ie. modprobe
1327 --force). Forced module loading sets the 'F' (forced) taint flag and
1328 is usually a really bad idea.
1330 config MODULE_UNLOAD
1331 bool "Module unloading"
1333 Without this option you will not be able to unload any
1334 modules (note that some modules may not be unloadable
1335 anyway), which makes your kernel smaller, faster
1336 and simpler. If unsure, say Y.
1338 config MODULE_FORCE_UNLOAD
1339 bool "Forced module unloading"
1340 depends on MODULE_UNLOAD && EXPERIMENTAL
1342 This option allows you to force a module to unload, even if the
1343 kernel believes it is unsafe: the kernel will remove the module
1344 without waiting for anyone to stop using it (using the -f option to
1345 rmmod). This is mainly for kernel developers and desperate users.
1349 bool "Module versioning support"
1351 Usually, you have to use modules compiled with your kernel.
1352 Saying Y here makes it sometimes possible to use modules
1353 compiled for different kernels, by adding enough information
1354 to the modules to (hopefully) spot any changes which would
1355 make them incompatible with the kernel you are running. If
1358 config MODULE_SRCVERSION_ALL
1359 bool "Source checksum for all modules"
1361 Modules which contain a MODULE_VERSION get an extra "srcversion"
1362 field inserted into their modinfo section, which contains a
1363 sum of the source files which made it. This helps maintainers
1364 see exactly which source was used to build a module (since
1365 others sometimes change the module source without updating
1366 the version). With this option, such a "srcversion" field
1367 will be created for all modules. If unsure, say N.
1371 config INIT_ALL_POSSIBLE
1374 Back when each arch used to define their own cpu_online_map and
1375 cpu_possible_map, some of them chose to initialize cpu_possible_map
1376 with all 1s, and others with all 0s. When they were centralised,
1377 it was better to provide this option than to break all the archs
1378 and have several arch maintainers pursuing me down dark alleys.
1383 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU
1385 Need stop_machine() primitive.
1387 source "block/Kconfig"
1389 config PREEMPT_NOTIFIERS
1396 source "kernel/Kconfig.locks"