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"
28 depends on HAVE_IRQ_WORK
33 bool "Prompt for development and/or incomplete code/drivers"
35 Some of the various things that Linux supports (such as network
36 drivers, file systems, network protocols, etc.) can be in a state
37 of development where the functionality, stability, or the level of
38 testing is not yet high enough for general use. This is usually
39 known as the "alpha-test" phase among developers. If a feature is
40 currently in alpha-test, then the developers usually discourage
41 uninformed widespread use of this feature by the general public to
42 avoid "Why doesn't this work?" type mail messages. However, active
43 testing and use of these systems is welcomed. Just be aware that it
44 may not meet the normal level of reliability or it may fail to work
45 in some special cases. Detailed bug reports from people familiar
46 with the kernel internals are usually welcomed by the developers
47 (before submitting bug reports, please read the documents
48 <file:README>, <file:MAINTAINERS>, <file:REPORTING-BUGS>,
49 <file:Documentation/BUG-HUNTING>, and
50 <file:Documentation/oops-tracing.txt> in the kernel source).
52 This option will also make obsoleted drivers available. These are
53 drivers that have been replaced by something else, and/or are
54 scheduled to be removed in a future kernel release.
56 Unless you intend to help test and develop a feature or driver that
57 falls into this category, or you have a situation that requires
58 using these features, you should probably say N here, which will
59 cause the configurator to present you with fewer choices. If
60 you say Y here, you will be offered the choice of using features or
61 drivers that are currently considered to be in the alpha-test phase.
68 depends on BROKEN || !SMP
71 config INIT_ENV_ARG_LIMIT
76 Maximum of each of the number of arguments and environment
77 variables passed to init from the kernel command line.
81 string "Cross-compiler tool prefix"
83 Same as running 'make CROSS_COMPILE=prefix-' but stored for
84 default make runs in this kernel build directory. You don't
85 need to set this unless you want the configured kernel build
86 directory to select the cross-compiler automatically.
89 string "Local version - append to kernel release"
91 Append an extra string to the end of your kernel version.
92 This will show up when you type uname, for example.
93 The string you set here will be appended after the contents of
94 any files with a filename matching localversion* in your
95 object and source tree, in that order. Your total string can
96 be a maximum of 64 characters.
98 config LOCALVERSION_AUTO
99 bool "Automatically append version information to the version string"
102 This will try to automatically determine if the current tree is a
103 release tree by looking for git tags that belong to the current
104 top of tree revision.
106 A string of the format -gxxxxxxxx will be added to the localversion
107 if a git-based tree is found. The string generated by this will be
108 appended after any matching localversion* files, and after the value
109 set in CONFIG_LOCALVERSION.
111 (The actual string used here is the first eight characters produced
112 by running the command:
114 $ git rev-parse --verify HEAD
116 which is done within the script "scripts/setlocalversion".)
118 config HAVE_KERNEL_GZIP
121 config HAVE_KERNEL_BZIP2
124 config HAVE_KERNEL_LZMA
127 config HAVE_KERNEL_XZ
130 config HAVE_KERNEL_LZO
134 prompt "Kernel compression mode"
135 default KERNEL_LZO if ARCH_RK29
137 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO
139 The linux kernel is a kind of self-extracting executable.
140 Several compression algorithms are available, which differ
141 in efficiency, compression and decompression speed.
142 Compression speed is only relevant when building a kernel.
143 Decompression speed is relevant at each boot.
145 If you have any problems with bzip2 or lzma compressed
146 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
147 version of this functionality (bzip2 only), for 2.4, was
148 supplied by Christian Ludwig)
150 High compression options are mostly useful for users, who
151 are low on disk space (embedded systems), but for whom ram
154 If in doubt, select 'gzip'
158 depends on HAVE_KERNEL_GZIP
160 The old and tried gzip compression. It provides a good balance
161 between compression ratio and decompression speed.
165 depends on HAVE_KERNEL_BZIP2
167 Its compression ratio and speed is intermediate.
168 Decompression speed is slowest among the three. The kernel
169 size is about 10% smaller with bzip2, in comparison to gzip.
170 Bzip2 uses a large amount of memory. For modern kernels you
171 will need at least 8MB RAM or more for booting.
175 depends on HAVE_KERNEL_LZMA
177 The most recent compression algorithm.
178 Its ratio is best, decompression speed is between the other
179 two. Compression is slowest. The kernel size is about 33%
180 smaller with LZMA in comparison to gzip.
184 depends on HAVE_KERNEL_XZ
186 XZ uses the LZMA2 algorithm and instruction set specific
187 BCJ filters which can improve compression ratio of executable
188 code. The size of the kernel is about 30% smaller with XZ in
189 comparison to gzip. On architectures for which there is a BCJ
190 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
191 will create a few percent smaller kernel than plain LZMA.
193 The speed is about the same as with LZMA: The decompression
194 speed of XZ is better than that of bzip2 but worse than gzip
195 and LZO. Compression is slow.
199 depends on HAVE_KERNEL_LZO
201 Its compression ratio is the poorest among the 4. The kernel
202 size is about 10% bigger than gzip; however its speed
203 (both compression and decompression) is the fastest.
207 config DEFAULT_HOSTNAME
208 string "Default hostname"
211 This option determines the default system hostname before userspace
212 calls sethostname(2). The kernel traditionally uses "(none)" here,
213 but you may wish to use a different default here to make a minimal
214 system more usable with less configuration.
217 bool "Support for paging of anonymous memory (swap)"
218 depends on MMU && BLOCK
221 This option allows you to choose whether you want to have support
222 for so called swap devices or swap files in your kernel that are
223 used to provide more virtual memory than the actual RAM present
224 in your computer. If unsure say Y.
229 Inter Process Communication is a suite of library functions and
230 system calls which let processes (running programs) synchronize and
231 exchange information. It is generally considered to be a good thing,
232 and some programs won't run unless you say Y here. In particular, if
233 you want to run the DOS emulator dosemu under Linux (read the
234 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
235 you'll need to say Y here.
237 You can find documentation about IPC with "info ipc" and also in
238 section 6.4 of the Linux Programmer's Guide, available from
239 <http://www.tldp.org/guides.html>.
241 config SYSVIPC_SYSCTL
248 bool "POSIX Message Queues"
249 depends on NET && EXPERIMENTAL
251 POSIX variant of message queues is a part of IPC. In POSIX message
252 queues every message has a priority which decides about succession
253 of receiving it by a process. If you want to compile and run
254 programs written e.g. for Solaris with use of its POSIX message
255 queues (functions mq_*) say Y here.
257 POSIX message queues are visible as a filesystem called 'mqueue'
258 and can be mounted somewhere if you want to do filesystem
259 operations on message queues.
263 config POSIX_MQUEUE_SYSCTL
265 depends on POSIX_MQUEUE
269 config BSD_PROCESS_ACCT
270 bool "BSD Process Accounting"
272 If you say Y here, a user level program will be able to instruct the
273 kernel (via a special system call) to write process accounting
274 information to a file: whenever a process exits, information about
275 that process will be appended to the file by the kernel. The
276 information includes things such as creation time, owning user,
277 command name, memory usage, controlling terminal etc. (the complete
278 list is in the struct acct in <file:include/linux/acct.h>). It is
279 up to the user level program to do useful things with this
280 information. This is generally a good idea, so say Y.
282 config BSD_PROCESS_ACCT_V3
283 bool "BSD Process Accounting version 3 file format"
284 depends on BSD_PROCESS_ACCT
287 If you say Y here, the process accounting information is written
288 in a new file format that also logs the process IDs of each
289 process and it's parent. Note that this file format is incompatible
290 with previous v0/v1/v2 file formats, so you will need updated tools
291 for processing it. A preliminary version of these tools is available
292 at <http://www.gnu.org/software/acct/>.
295 bool "open by fhandle syscalls"
298 If you say Y here, a user level program will be able to map
299 file names to handle and then later use the handle for
300 different file system operations. This is useful in implementing
301 userspace file servers, which now track files using handles instead
302 of names. The handle would remain the same even if file names
303 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
307 bool "Export task/process statistics through netlink (EXPERIMENTAL)"
311 Export selected statistics for tasks/processes through the
312 generic netlink interface. Unlike BSD process accounting, the
313 statistics are available during the lifetime of tasks/processes as
314 responses to commands. Like BSD accounting, they are sent to user
319 config TASK_DELAY_ACCT
320 bool "Enable per-task delay accounting (EXPERIMENTAL)"
323 Collect information on time spent by a task waiting for system
324 resources like cpu, synchronous block I/O completion and swapping
325 in pages. Such statistics can help in setting a task's priorities
326 relative to other tasks for cpu, io, rss limits etc.
331 bool "Enable extended accounting over taskstats (EXPERIMENTAL)"
334 Collect extended task accounting data and send the data
335 to userland for processing over the taskstats interface.
339 config TASK_IO_ACCOUNTING
340 bool "Enable per-task storage I/O accounting (EXPERIMENTAL)"
341 depends on TASK_XACCT
343 Collect information on the number of bytes of storage I/O which this
349 bool "Auditing support"
352 Enable auditing infrastructure that can be used with another
353 kernel subsystem, such as SELinux (which requires this for
354 logging of avc messages output). Does not do system-call
355 auditing without CONFIG_AUDITSYSCALL.
358 bool "Enable system-call auditing support"
359 depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH)
360 default y if SECURITY_SELINUX
362 Enable low-overhead system-call auditing infrastructure that
363 can be used independently or with another kernel subsystem,
368 depends on AUDITSYSCALL
373 depends on AUDITSYSCALL
376 source "kernel/irq/Kconfig"
381 prompt "RCU Implementation"
385 bool "Tree-based hierarchical RCU"
386 depends on !PREEMPT && SMP
388 This option selects the RCU implementation that is
389 designed for very large SMP system with hundreds or
390 thousands of CPUs. It also scales down nicely to
393 config TREE_PREEMPT_RCU
394 bool "Preemptible tree-based hierarchical RCU"
397 This option selects the RCU implementation that is
398 designed for very large SMP systems with hundreds or
399 thousands of CPUs, but for which real-time response
400 is also required. It also scales down nicely to
404 bool "UP-only small-memory-footprint RCU"
407 This option selects the RCU implementation that is
408 designed for UP systems from which real-time response
409 is not required. This option greatly reduces the
410 memory footprint of RCU.
412 config TINY_PREEMPT_RCU
413 bool "Preemptible UP-only small-memory-footprint RCU"
414 depends on !SMP && PREEMPT
416 This option selects the RCU implementation that is designed
417 for real-time UP systems. This option greatly reduces the
418 memory footprint of RCU.
423 def_bool ( TREE_PREEMPT_RCU || TINY_PREEMPT_RCU )
425 This option enables preemptible-RCU code that is common between
426 the TREE_PREEMPT_RCU and TINY_PREEMPT_RCU implementations.
429 bool "Enable tracing for RCU"
431 This option provides tracing in RCU which presents stats
432 in debugfs for debugging RCU implementation.
434 Say Y here if you want to enable RCU tracing
435 Say N if you are unsure.
438 int "Tree-based hierarchical RCU fanout value"
441 depends on TREE_RCU || TREE_PREEMPT_RCU
445 This option controls the fanout of hierarchical implementations
446 of RCU, allowing RCU to work efficiently on machines with
447 large numbers of CPUs. This value must be at least the fourth
448 root of NR_CPUS, which allows NR_CPUS to be insanely large.
449 The default value of RCU_FANOUT should be used for production
450 systems, but if you are stress-testing the RCU implementation
451 itself, small RCU_FANOUT values allow you to test large-system
452 code paths on small(er) systems.
454 Select a specific number if testing RCU itself.
455 Take the default if unsure.
457 config RCU_FANOUT_EXACT
458 bool "Disable tree-based hierarchical RCU auto-balancing"
459 depends on TREE_RCU || TREE_PREEMPT_RCU
462 This option forces use of the exact RCU_FANOUT value specified,
463 regardless of imbalances in the hierarchy. This is useful for
464 testing RCU itself, and might one day be useful on systems with
465 strong NUMA behavior.
467 Without RCU_FANOUT_EXACT, the code will balance the hierarchy.
471 config RCU_FAST_NO_HZ
472 bool "Accelerate last non-dyntick-idle CPU's grace periods"
473 depends on TREE_RCU && NO_HZ && SMP
476 This option causes RCU to attempt to accelerate grace periods
477 in order to allow the final CPU to enter dynticks-idle state
478 more quickly. On the other hand, this option increases the
479 overhead of the dynticks-idle checking, particularly on systems
480 with large numbers of CPUs.
482 Say Y if energy efficiency is critically important, particularly
483 if you have relatively few CPUs.
485 Say N if you are unsure.
487 config TREE_RCU_TRACE
488 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU )
491 This option provides tracing for the TREE_RCU and
492 TREE_PREEMPT_RCU implementations, permitting Makefile to
493 trivially select kernel/rcutree_trace.c.
496 bool "Enable RCU priority boosting"
497 depends on RT_MUTEXES && PREEMPT_RCU
500 This option boosts the priority of preempted RCU readers that
501 block the current preemptible RCU grace period for too long.
502 This option also prevents heavy loads from blocking RCU
503 callback invocation for all flavors of RCU.
505 Say Y here if you are working with real-time apps or heavy loads
506 Say N here if you are unsure.
508 config RCU_BOOST_PRIO
509 int "Real-time priority to boost RCU readers to"
514 This option specifies the real-time priority to which preempted
515 RCU readers are to be boosted. If you are working with CPU-bound
516 real-time applications, you should specify a priority higher then
517 the highest-priority CPU-bound application.
519 Specify the real-time priority, or take the default if unsure.
521 config RCU_BOOST_DELAY
522 int "Milliseconds to delay boosting after RCU grace-period start"
527 This option specifies the time to wait after the beginning of
528 a given grace period before priority-boosting preempted RCU
529 readers blocking that grace period. Note that any RCU reader
530 blocking an expedited RCU grace period is boosted immediately.
532 Accept the default if unsure.
534 endmenu # "RCU Subsystem"
537 tristate "Kernel .config support"
539 This option enables the complete Linux kernel ".config" file
540 contents to be saved in the kernel. It provides documentation
541 of which kernel options are used in a running kernel or in an
542 on-disk kernel. This information can be extracted from the kernel
543 image file with the script scripts/extract-ikconfig and used as
544 input to rebuild the current kernel or to build another kernel.
545 It can also be extracted from a running kernel by reading
546 /proc/config.gz if enabled (below).
549 bool "Enable access to .config through /proc/config.gz"
550 depends on IKCONFIG && PROC_FS
552 This option enables access to the kernel configuration file
553 through /proc/config.gz.
556 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
560 Select kernel log buffer size as a power of 2.
570 # Architectures with an unreliable sched_clock() should select this:
572 config HAVE_UNSTABLE_SCHED_CLOCK
576 boolean "Control Group support"
579 This option adds support for grouping sets of processes together, for
580 use with process control subsystems such as Cpusets, CFS, memory
581 controls or device isolation.
583 - Documentation/scheduler/sched-design-CFS.txt (CFS)
584 - Documentation/cgroups/ (features for grouping, isolation
585 and resource control)
592 bool "Example debug cgroup subsystem"
595 This option enables a simple cgroup subsystem that
596 exports useful debugging information about the cgroups
601 config CGROUP_FREEZER
602 bool "Freezer cgroup subsystem"
604 Provides a way to freeze and unfreeze all tasks in a
608 bool "Device controller for cgroups"
610 Provides a cgroup implementing whitelists for devices which
611 a process in the cgroup can mknod or open.
614 bool "Cpuset support"
616 This option will let you create and manage CPUSETs which
617 allow dynamically partitioning a system into sets of CPUs and
618 Memory Nodes and assigning tasks to run only within those sets.
619 This is primarily useful on large SMP or NUMA systems.
623 config PROC_PID_CPUSET
624 bool "Include legacy /proc/<pid>/cpuset file"
628 config CGROUP_CPUACCT
629 bool "Simple CPU accounting cgroup subsystem"
631 Provides a simple Resource Controller for monitoring the
632 total CPU consumed by the tasks in a cgroup.
634 config RESOURCE_COUNTERS
635 bool "Resource counters"
637 This option enables controller independent resource accounting
638 infrastructure that works with cgroups.
640 config CGROUP_MEM_RES_CTLR
641 bool "Memory Resource Controller for Control Groups"
642 depends on RESOURCE_COUNTERS
645 Provides a memory resource controller that manages both anonymous
646 memory and page cache. (See Documentation/cgroups/memory.txt)
648 Note that setting this option increases fixed memory overhead
649 associated with each page of memory in the system. By this,
650 20(40)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory
651 usage tracking struct at boot. Total amount of this is printed out
654 Only enable when you're ok with these trade offs and really
655 sure you need the memory resource controller. Even when you enable
656 this, you can set "cgroup_disable=memory" at your boot option to
657 disable memory resource controller and you can avoid overheads.
658 (and lose benefits of memory resource controller)
660 This config option also selects MM_OWNER config option, which
661 could in turn add some fork/exit overhead.
663 config CGROUP_MEM_RES_CTLR_SWAP
664 bool "Memory Resource Controller Swap Extension"
665 depends on CGROUP_MEM_RES_CTLR && SWAP
667 Add swap management feature to memory resource controller. When you
668 enable this, you can limit mem+swap usage per cgroup. In other words,
669 when you disable this, memory resource controller has no cares to
670 usage of swap...a process can exhaust all of the swap. This extension
671 is useful when you want to avoid exhaustion swap but this itself
672 adds more overheads and consumes memory for remembering information.
673 Especially if you use 32bit system or small memory system, please
674 be careful about enabling this. When memory resource controller
675 is disabled by boot option, this will be automatically disabled and
676 there will be no overhead from this. Even when you set this config=y,
677 if boot option "noswapaccount" is set, swap will not be accounted.
678 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page
679 size is 4096bytes, 512k per 1Gbytes of swap.
680 config CGROUP_MEM_RES_CTLR_SWAP_ENABLED
681 bool "Memory Resource Controller Swap Extension enabled by default"
682 depends on CGROUP_MEM_RES_CTLR_SWAP
685 Memory Resource Controller Swap Extension comes with its price in
686 a bigger memory consumption. General purpose distribution kernels
687 which want to enable the feature but keep it disabled by default
688 and let the user enable it by swapaccount boot command line
689 parameter should have this option unselected.
690 For those who want to have the feature enabled by default should
691 select this option (if, for some reason, they need to disable it
692 then noswapaccount does the trick).
695 bool "Enable perf_event per-cpu per-container group (cgroup) monitoring"
696 depends on PERF_EVENTS && CGROUPS
698 This option extends the per-cpu mode to restrict monitoring to
699 threads which belong to the cgroup specified and run on the
704 menuconfig CGROUP_SCHED
705 bool "Group CPU scheduler"
706 depends on EXPERIMENTAL
709 This feature lets CPU scheduler recognize task groups and control CPU
710 bandwidth allocation to such task groups. It uses cgroups to group
714 config FAIR_GROUP_SCHED
715 bool "Group scheduling for SCHED_OTHER"
716 depends on CGROUP_SCHED
719 config RT_GROUP_SCHED
720 bool "Group scheduling for SCHED_RR/FIFO"
721 depends on EXPERIMENTAL
722 depends on CGROUP_SCHED
725 This feature lets you explicitly allocate real CPU bandwidth
726 to task groups. If enabled, it will also make it impossible to
727 schedule realtime tasks for non-root users until you allocate
728 realtime bandwidth for them.
729 See Documentation/scheduler/sched-rt-group.txt for more information.
734 tristate "Block IO controller"
738 Generic block IO controller cgroup interface. This is the common
739 cgroup interface which should be used by various IO controlling
742 Currently, CFQ IO scheduler uses it to recognize task groups and
743 control disk bandwidth allocation (proportional time slice allocation)
744 to such task groups. It is also used by bio throttling logic in
745 block layer to implement upper limit in IO rates on a device.
747 This option only enables generic Block IO controller infrastructure.
748 One needs to also enable actual IO controlling logic/policy. For
749 enabling proportional weight division of disk bandwidth in CFQ, set
750 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
751 CONFIG_BLK_DEV_THROTTLING=y.
753 See Documentation/cgroups/blkio-controller.txt for more information.
755 config DEBUG_BLK_CGROUP
756 bool "Enable Block IO controller debugging"
757 depends on BLK_CGROUP
760 Enable some debugging help. Currently it exports additional stat
761 files in a cgroup which can be useful for debugging.
765 menuconfig NAMESPACES
766 bool "Namespaces support" if EXPERT
769 Provides the way to make tasks work with different objects using
770 the same id. For example same IPC id may refer to different objects
771 or same user id or pid may refer to different tasks when used in
772 different namespaces.
780 In this namespace tasks see different info provided with the
785 depends on (SYSVIPC || POSIX_MQUEUE)
788 In this namespace tasks work with IPC ids which correspond to
789 different IPC objects in different namespaces.
792 bool "User namespace (EXPERIMENTAL)"
793 depends on EXPERIMENTAL
796 This allows containers, i.e. vservers, to use user namespaces
797 to provide different user info for different servers.
801 bool "PID Namespaces"
804 Support process id namespaces. This allows having multiple
805 processes with the same pid as long as they are in different
806 pid namespaces. This is a building block of containers.
809 bool "Network namespace"
813 Allow user space to create what appear to be multiple instances
814 of the network stack.
818 config SCHED_AUTOGROUP
819 bool "Automatic process group scheduling"
823 select FAIR_GROUP_SCHED
825 This option optimizes the scheduler for common desktop workloads by
826 automatically creating and populating task groups. This separation
827 of workloads isolates aggressive CPU burners (like build jobs) from
828 desktop applications. Task group autogeneration is currently based
834 config SYSFS_DEPRECATED
835 bool "Enable deprecated sysfs features to support old userspace tools"
839 This option adds code that switches the layout of the "block" class
840 devices, to not show up in /sys/class/block/, but only in
843 This switch is only active when the sysfs.deprecated=1 boot option is
844 passed or the SYSFS_DEPRECATED_V2 option is set.
846 This option allows new kernels to run on old distributions and tools,
847 which might get confused by /sys/class/block/. Since 2007/2008 all
848 major distributions and tools handle this just fine.
850 Recent distributions and userspace tools after 2009/2010 depend on
851 the existence of /sys/class/block/, and will not work with this
854 Only if you are using a new kernel on an old distribution, you might
857 config SYSFS_DEPRECATED_V2
858 bool "Enable deprecated sysfs features by default"
861 depends on SYSFS_DEPRECATED
863 Enable deprecated sysfs by default.
865 See the CONFIG_SYSFS_DEPRECATED option for more details about this
868 Only if you are using a new kernel on an old distribution, you might
869 need to say Y here. Even then, odds are you would not need it
870 enabled, you can always pass the boot option if absolutely necessary.
873 bool "Kernel->user space relay support (formerly relayfs)"
875 This option enables support for relay interface support in
876 certain file systems (such as debugfs).
877 It is designed to provide an efficient mechanism for tools and
878 facilities to relay large amounts of data from kernel space to
883 config BLK_DEV_INITRD
884 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
885 depends on BROKEN || !FRV
887 The initial RAM filesystem is a ramfs which is loaded by the
888 boot loader (loadlin or lilo) and that is mounted as root
889 before the normal boot procedure. It is typically used to
890 load modules needed to mount the "real" root file system,
891 etc. See <file:Documentation/initrd.txt> for details.
893 If RAM disk support (BLK_DEV_RAM) is also included, this
894 also enables initial RAM disk (initrd) support and adds
895 15 Kbytes (more on some other architectures) to the kernel size.
905 config CC_OPTIMIZE_FOR_SIZE
906 bool "Optimize for size"
908 Enabling this option will pass "-Os" instead of "-O2" to gcc
909 resulting in a smaller kernel.
920 int "Default panic timeout"
923 Set default panic timeout.
926 bool "Configure standard kernel features (expert users)"
928 This option allows certain base kernel options and settings
929 to be disabled or tweaked. This is for specialized
930 environments which can tolerate a "non-standard" kernel.
931 Only use this if you really know what you are doing.
934 bool "Enable 16-bit UID system calls" if EXPERT
935 depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && COMPAT) || UML || (X86_64 && IA32_EMULATION)
938 This enables the legacy 16-bit UID syscall wrappers.
940 config SYSCTL_SYSCALL
941 bool "Sysctl syscall support" if EXPERT
942 depends on PROC_SYSCTL
946 sys_sysctl uses binary paths that have been found challenging
947 to properly maintain and use. The interface in /proc/sys
948 using paths with ascii names is now the primary path to this
951 Almost nothing using the binary sysctl interface so if you are
952 trying to save some space it is probably safe to disable this,
953 making your kernel marginally smaller.
955 If unsure say Y here.
958 bool "Load all symbols for debugging/ksymoops" if EXPERT
961 Say Y here to let the kernel print out symbolic crash information and
962 symbolic stack backtraces. This increases the size of the kernel
963 somewhat, as all symbols have to be loaded into the kernel image.
966 bool "Include all symbols in kallsyms"
967 depends on DEBUG_KERNEL && KALLSYMS
969 Normally kallsyms only contains the symbols of functions for nicer
970 OOPS messages and backtraces (i.e., symbols from the text and inittext
971 sections). This is sufficient for most cases. And only in very rare
972 cases (e.g., when a debugger is used) all symbols are required (e.g.,
973 names of variables from the data sections, etc).
975 This option makes sure that all symbols are loaded into the kernel
976 image (i.e., symbols from all sections) in cost of increased kernel
977 size (depending on the kernel configuration, it may be 300KiB or
978 something like this).
980 Say N unless you really need all symbols.
983 bool "Support for hot-pluggable devices" if EXPERT
986 This option is provided for the case where no hotplug or uevent
987 capabilities is wanted by the kernel. You should only consider
988 disabling this option for embedded systems that do not use modules, a
989 dynamic /dev tree, or dynamic device discovery. Just say Y.
993 bool "Enable support for printk" if EXPERT
995 This option enables normal printk support. Removing it
996 eliminates most of the message strings from the kernel image
997 and makes the kernel more or less silent. As this makes it
998 very difficult to diagnose system problems, saying N here is
999 strongly discouraged.
1002 bool "BUG() support" if EXPERT
1005 Disabling this option eliminates support for BUG and WARN, reducing
1006 the size of your kernel image and potentially quietly ignoring
1007 numerous fatal conditions. You should only consider disabling this
1008 option for embedded systems with no facilities for reporting errors.
1013 bool "Enable ELF core dumps" if EXPERT
1015 Enable support for generating core dumps. Disabling saves about 4k.
1017 config PCSPKR_PLATFORM
1018 bool "Enable PC-Speaker support" if EXPERT
1019 depends on ALPHA || X86 || MIPS || PPC_PREP || PPC_CHRP || PPC_PSERIES
1022 This option allows to disable the internal PC-Speaker
1023 support, saving some memory.
1027 bool "Enable full-sized data structures for core" if EXPERT
1029 Disabling this option reduces the size of miscellaneous core
1030 kernel data structures. This saves memory on small machines,
1031 but may reduce performance.
1034 bool "Enable futex support" if EXPERT
1038 Disabling this option will cause the kernel to be built without
1039 support for "fast userspace mutexes". The resulting kernel may not
1040 run glibc-based applications correctly.
1043 bool "Enable eventpoll support" if EXPERT
1047 Disabling this option will cause the kernel to be built without
1048 support for epoll family of system calls.
1051 bool "Enable signalfd() system call" if EXPERT
1055 Enable the signalfd() system call that allows to receive signals
1056 on a file descriptor.
1061 bool "Enable timerfd() system call" if EXPERT
1065 Enable the timerfd() system call that allows to receive timer
1066 events on a file descriptor.
1071 bool "Enable eventfd() system call" if EXPERT
1075 Enable the eventfd() system call that allows to receive both
1076 kernel notification (ie. KAIO) or userspace notifications.
1081 bool "Use full shmem filesystem" if EXPERT
1085 The shmem is an internal filesystem used to manage shared memory.
1086 It is backed by swap and manages resource limits. It is also exported
1087 to userspace as tmpfs if TMPFS is enabled. Disabling this
1088 option replaces shmem and tmpfs with the much simpler ramfs code,
1089 which may be appropriate on small systems without swap.
1092 bool "Enable the Anonymous Shared Memory Subsystem"
1094 depends on SHMEM || TINY_SHMEM
1096 The ashmem subsystem is a new shared memory allocator, similar to
1097 POSIX SHM but with different behavior and sporting a simpler
1101 bool "Enable AIO support" if EXPERT
1104 This option enables POSIX asynchronous I/O which may by used
1105 by some high performance threaded applications. Disabling
1106 this option saves about 7k.
1109 bool "Embedded system"
1112 This option should be enabled if compiling the kernel for
1113 an embedded system so certain expert options are available
1116 config HAVE_PERF_EVENTS
1119 See tools/perf/design.txt for details.
1121 config PERF_USE_VMALLOC
1124 See tools/perf/design.txt for details
1126 menu "Kernel Performance Events And Counters"
1129 bool "Kernel performance events and counters"
1130 default y if (PROFILING || PERF_COUNTERS)
1131 depends on HAVE_PERF_EVENTS
1135 Enable kernel support for various performance events provided
1136 by software and hardware.
1138 Software events are supported either built-in or via the
1139 use of generic tracepoints.
1141 Most modern CPUs support performance events via performance
1142 counter registers. These registers count the number of certain
1143 types of hw events: such as instructions executed, cachemisses
1144 suffered, or branches mis-predicted - without slowing down the
1145 kernel or applications. These registers can also trigger interrupts
1146 when a threshold number of events have passed - and can thus be
1147 used to profile the code that runs on that CPU.
1149 The Linux Performance Event subsystem provides an abstraction of
1150 these software and hardware event capabilities, available via a
1151 system call and used by the "perf" utility in tools/perf/. It
1152 provides per task and per CPU counters, and it provides event
1153 capabilities on top of those.
1157 config PERF_COUNTERS
1158 bool "Kernel performance counters (old config option)"
1159 depends on HAVE_PERF_EVENTS
1161 This config has been obsoleted by the PERF_EVENTS
1162 config option - please see that one for details.
1164 It has no effect on the kernel whether you enable
1165 it or not, it is a compatibility placeholder.
1169 config DEBUG_PERF_USE_VMALLOC
1171 bool "Debug: use vmalloc to back perf mmap() buffers"
1172 depends on PERF_EVENTS && DEBUG_KERNEL
1173 select PERF_USE_VMALLOC
1175 Use vmalloc memory to back perf mmap() buffers.
1177 Mostly useful for debugging the vmalloc code on platforms
1178 that don't require it.
1184 config VM_EVENT_COUNTERS
1186 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1188 VM event counters are needed for event counts to be shown.
1189 This option allows the disabling of the VM event counters
1190 on EXPERT systems. /proc/vmstat will only show page counts
1191 if VM event counters are disabled.
1195 bool "Enable PCI quirk workarounds" if EXPERT
1198 This enables workarounds for various PCI chipset
1199 bugs/quirks. Disable this only if your target machine is
1200 unaffected by PCI quirks.
1204 bool "Enable SLUB debugging support" if EXPERT
1205 depends on SLUB && SYSFS
1207 SLUB has extensive debug support features. Disabling these can
1208 result in significant savings in code size. This also disables
1209 SLUB sysfs support. /sys/slab will not exist and there will be
1210 no support for cache validation etc.
1213 bool "Disable heap randomization"
1216 Randomizing heap placement makes heap exploits harder, but it
1217 also breaks ancient binaries (including anything libc5 based).
1218 This option changes the bootup default to heap randomization
1219 disabled, and can be overridden at runtime by setting
1220 /proc/sys/kernel/randomize_va_space to 2.
1222 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1225 prompt "Choose SLAB allocator"
1228 This option allows to select a slab allocator.
1233 The regular slab allocator that is established and known to work
1234 well in all environments. It organizes cache hot objects in
1235 per cpu and per node queues.
1238 bool "SLUB (Unqueued Allocator)"
1240 SLUB is a slab allocator that minimizes cache line usage
1241 instead of managing queues of cached objects (SLAB approach).
1242 Per cpu caching is realized using slabs of objects instead
1243 of queues of objects. SLUB can use memory efficiently
1244 and has enhanced diagnostics. SLUB is the default choice for
1249 bool "SLOB (Simple Allocator)"
1251 SLOB replaces the stock allocator with a drastically simpler
1252 allocator. SLOB is generally more space efficient but
1253 does not perform as well on large systems.
1257 config MMAP_ALLOW_UNINITIALIZED
1258 bool "Allow mmapped anonymous memory to be uninitialized"
1259 depends on EXPERT && !MMU
1262 Normally, and according to the Linux spec, anonymous memory obtained
1263 from mmap() has it's contents cleared before it is passed to
1264 userspace. Enabling this config option allows you to request that
1265 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1266 providing a huge performance boost. If this option is not enabled,
1267 then the flag will be ignored.
1269 This is taken advantage of by uClibc's malloc(), and also by
1270 ELF-FDPIC binfmt's brk and stack allocator.
1272 Because of the obvious security issues, this option should only be
1273 enabled on embedded devices where you control what is run in
1274 userspace. Since that isn't generally a problem on no-MMU systems,
1275 it is normally safe to say Y here.
1277 See Documentation/nommu-mmap.txt for more information.
1280 bool "Profiling support"
1282 Say Y here to enable the extended profiling support mechanisms used
1283 by profilers such as OProfile.
1286 # Place an empty function call at each tracepoint site. Can be
1287 # dynamically changed for a probe function.
1292 source "arch/Kconfig"
1294 endmenu # General setup
1296 config HAVE_GENERIC_DMA_COHERENT
1303 depends on SLAB || SLUB_DEBUG
1311 default 0 if BASE_FULL
1312 default 1 if !BASE_FULL
1315 bool "Enable loadable module support"
1317 Kernel modules are small pieces of compiled code which can
1318 be inserted in the running kernel, rather than being
1319 permanently built into the kernel. You use the "modprobe"
1320 tool to add (and sometimes remove) them. If you say Y here,
1321 many parts of the kernel can be built as modules (by
1322 answering M instead of Y where indicated): this is most
1323 useful for infrequently used options which are not required
1324 for booting. For more information, see the man pages for
1325 modprobe, lsmod, modinfo, insmod and rmmod.
1327 If you say Y here, you will need to run "make
1328 modules_install" to put the modules under /lib/modules/
1329 where modprobe can find them (you may need to be root to do
1336 config MODULE_FORCE_LOAD
1337 bool "Forced module loading"
1340 Allow loading of modules without version information (ie. modprobe
1341 --force). Forced module loading sets the 'F' (forced) taint flag and
1342 is usually a really bad idea.
1344 config MODULE_UNLOAD
1345 bool "Module unloading"
1347 Without this option you will not be able to unload any
1348 modules (note that some modules may not be unloadable
1349 anyway), which makes your kernel smaller, faster
1350 and simpler. If unsure, say Y.
1352 config MODULE_FORCE_UNLOAD
1353 bool "Forced module unloading"
1354 depends on MODULE_UNLOAD && EXPERIMENTAL
1356 This option allows you to force a module to unload, even if the
1357 kernel believes it is unsafe: the kernel will remove the module
1358 without waiting for anyone to stop using it (using the -f option to
1359 rmmod). This is mainly for kernel developers and desperate users.
1363 bool "Module versioning support"
1365 Usually, you have to use modules compiled with your kernel.
1366 Saying Y here makes it sometimes possible to use modules
1367 compiled for different kernels, by adding enough information
1368 to the modules to (hopefully) spot any changes which would
1369 make them incompatible with the kernel you are running. If
1372 config MODULE_SRCVERSION_ALL
1373 bool "Source checksum for all modules"
1375 Modules which contain a MODULE_VERSION get an extra "srcversion"
1376 field inserted into their modinfo section, which contains a
1377 sum of the source files which made it. This helps maintainers
1378 see exactly which source was used to build a module (since
1379 others sometimes change the module source without updating
1380 the version). With this option, such a "srcversion" field
1381 will be created for all modules. If unsure, say N.
1385 config INIT_ALL_POSSIBLE
1388 Back when each arch used to define their own cpu_online_map and
1389 cpu_possible_map, some of them chose to initialize cpu_possible_map
1390 with all 1s, and others with all 0s. When they were centralised,
1391 it was better to provide this option than to break all the archs
1392 and have several arch maintainers pursuing me down dark alleys.
1397 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU
1399 Need stop_machine() primitive.
1401 source "block/Kconfig"
1403 config PREEMPT_NOTIFIERS
1410 source "kernel/Kconfig.locks"