4 config ENTERPRISE_SUPPORT
5 bool "Enable enterprise support facility"
8 This feature enables the handling of the "supported" module flag.
9 This flag can be used to report unsupported module loads or even
10 refuse them entirely. It is useful when ensuring that the kernel
11 remains in a state that Novell Technical Services, or its
12 technical partners, is prepared to support.
14 Modules in the list of supported modules will be marked supported
15 on build. The default enforcement mode is to report, but not
16 deny, loading of unsupported modules.
18 If you aren't building a kernel for an enterprise distribution,
22 bool "Split the kernel package into multiple RPMs"
23 depends on SUSE_KERNEL && MODULES
25 This is an option used by the kernel packaging infrastructure
26 to split kernel modules into different packages. It isn't used
27 by the kernel itself, but allows the the packager to make
28 decisions on a per-config basis.
30 If you aren't packaging a kernel for distribution, it's safe to
34 bool "Kernel to suit desktop workloads"
36 This is an option used to tune kernel parameters to better suit
45 option env="KERNELVERSION"
51 default "/lib/modules/$UNAME_RELEASE/.config"
52 default "/etc/kernel-config"
53 default "/boot/config-$UNAME_RELEASE"
54 default "$ARCH_DEFCONFIG"
55 default "arch/$ARCH/defconfig"
66 depends on HAVE_IRQ_WORK
71 bool "Prompt for development and/or incomplete code/drivers"
73 Some of the various things that Linux supports (such as network
74 drivers, file systems, network protocols, etc.) can be in a state
75 of development where the functionality, stability, or the level of
76 testing is not yet high enough for general use. This is usually
77 known as the "alpha-test" phase among developers. If a feature is
78 currently in alpha-test, then the developers usually discourage
79 uninformed widespread use of this feature by the general public to
80 avoid "Why doesn't this work?" type mail messages. However, active
81 testing and use of these systems is welcomed. Just be aware that it
82 may not meet the normal level of reliability or it may fail to work
83 in some special cases. Detailed bug reports from people familiar
84 with the kernel internals are usually welcomed by the developers
85 (before submitting bug reports, please read the documents
86 <file:README>, <file:MAINTAINERS>, <file:REPORTING-BUGS>,
87 <file:Documentation/BUG-HUNTING>, and
88 <file:Documentation/oops-tracing.txt> in the kernel source).
90 This option will also make obsoleted drivers available. These are
91 drivers that have been replaced by something else, and/or are
92 scheduled to be removed in a future kernel release.
94 Unless you intend to help test and develop a feature or driver that
95 falls into this category, or you have a situation that requires
96 using these features, you should probably say N here, which will
97 cause the configurator to present you with fewer choices. If
98 you say Y here, you will be offered the choice of using features or
99 drivers that are currently considered to be in the alpha-test phase.
106 depends on BROKEN || !SMP
109 config INIT_ENV_ARG_LIMIT
114 Maximum of each of the number of arguments and environment
115 variables passed to init from the kernel command line.
119 string "Cross-compiler tool prefix"
121 Same as running 'make CROSS_COMPILE=prefix-' but stored for
122 default make runs in this kernel build directory. You don't
123 need to set this unless you want the configured kernel build
124 directory to select the cross-compiler automatically.
127 string "Local version - append to kernel release"
129 Append an extra string to the end of your kernel version.
130 This will show up when you type uname, for example.
131 The string you set here will be appended after the contents of
132 any files with a filename matching localversion* in your
133 object and source tree, in that order. Your total string can
134 be a maximum of 64 characters.
136 config LOCALVERSION_AUTO
137 bool "Automatically append version information to the version string"
140 This will try to automatically determine if the current tree is a
141 release tree by looking for git tags that belong to the current
142 top of tree revision.
144 A string of the format -gxxxxxxxx will be added to the localversion
145 if a git-based tree is found. The string generated by this will be
146 appended after any matching localversion* files, and after the value
147 set in CONFIG_LOCALVERSION.
149 (The actual string used here is the first eight characters produced
150 by running the command:
152 $ git rev-parse --verify HEAD
154 which is done within the script "scripts/setlocalversion".)
156 config HAVE_KERNEL_GZIP
159 config HAVE_KERNEL_BZIP2
162 config HAVE_KERNEL_LZMA
165 config HAVE_KERNEL_XZ
168 config HAVE_KERNEL_LZO
172 prompt "Kernel compression mode"
174 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO
176 The linux kernel is a kind of self-extracting executable.
177 Several compression algorithms are available, which differ
178 in efficiency, compression and decompression speed.
179 Compression speed is only relevant when building a kernel.
180 Decompression speed is relevant at each boot.
182 If you have any problems with bzip2 or lzma compressed
183 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
184 version of this functionality (bzip2 only), for 2.4, was
185 supplied by Christian Ludwig)
187 High compression options are mostly useful for users, who
188 are low on disk space (embedded systems), but for whom ram
191 If in doubt, select 'gzip'
195 depends on HAVE_KERNEL_GZIP
197 The old and tried gzip compression. It provides a good balance
198 between compression ratio and decompression speed.
202 depends on HAVE_KERNEL_BZIP2
204 Its compression ratio and speed is intermediate.
205 Decompression speed is slowest among the three. The kernel
206 size is about 10% smaller with bzip2, in comparison to gzip.
207 Bzip2 uses a large amount of memory. For modern kernels you
208 will need at least 8MB RAM or more for booting.
212 depends on HAVE_KERNEL_LZMA
214 The most recent compression algorithm.
215 Its ratio is best, decompression speed is between the other
216 two. Compression is slowest. The kernel size is about 33%
217 smaller with LZMA in comparison to gzip.
221 depends on HAVE_KERNEL_XZ
223 XZ uses the LZMA2 algorithm and instruction set specific
224 BCJ filters which can improve compression ratio of executable
225 code. The size of the kernel is about 30% smaller with XZ in
226 comparison to gzip. On architectures for which there is a BCJ
227 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
228 will create a few percent smaller kernel than plain LZMA.
230 The speed is about the same as with LZMA: The decompression
231 speed of XZ is better than that of bzip2 but worse than gzip
232 and LZO. Compression is slow.
236 depends on HAVE_KERNEL_LZO
238 Its compression ratio is the poorest among the 4. The kernel
239 size is about 10% bigger than gzip; however its speed
240 (both compression and decompression) is the fastest.
244 config DEFAULT_HOSTNAME
245 string "Default hostname"
248 This option determines the default system hostname before userspace
249 calls sethostname(2). The kernel traditionally uses "(none)" here,
250 but you may wish to use a different default here to make a minimal
251 system more usable with less configuration.
254 bool "Support for paging of anonymous memory (swap)"
255 depends on MMU && BLOCK
258 This option allows you to choose whether you want to have support
259 for so called swap devices or swap files in your kernel that are
260 used to provide more virtual memory than the actual RAM present
261 in your computer. If unsure say Y.
266 Inter Process Communication is a suite of library functions and
267 system calls which let processes (running programs) synchronize and
268 exchange information. It is generally considered to be a good thing,
269 and some programs won't run unless you say Y here. In particular, if
270 you want to run the DOS emulator dosemu under Linux (read the
271 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
272 you'll need to say Y here.
274 You can find documentation about IPC with "info ipc" and also in
275 section 6.4 of the Linux Programmer's Guide, available from
276 <http://www.tldp.org/guides.html>.
278 config SYSVIPC_SYSCTL
285 bool "POSIX Message Queues"
286 depends on NET && EXPERIMENTAL
288 POSIX variant of message queues is a part of IPC. In POSIX message
289 queues every message has a priority which decides about succession
290 of receiving it by a process. If you want to compile and run
291 programs written e.g. for Solaris with use of its POSIX message
292 queues (functions mq_*) say Y here.
294 POSIX message queues are visible as a filesystem called 'mqueue'
295 and can be mounted somewhere if you want to do filesystem
296 operations on message queues.
300 config POSIX_MQUEUE_SYSCTL
302 depends on POSIX_MQUEUE
306 config BSD_PROCESS_ACCT
307 bool "BSD Process Accounting"
309 If you say Y here, a user level program will be able to instruct the
310 kernel (via a special system call) to write process accounting
311 information to a file: whenever a process exits, information about
312 that process will be appended to the file by the kernel. The
313 information includes things such as creation time, owning user,
314 command name, memory usage, controlling terminal etc. (the complete
315 list is in the struct acct in <file:include/linux/acct.h>). It is
316 up to the user level program to do useful things with this
317 information. This is generally a good idea, so say Y.
319 config BSD_PROCESS_ACCT_V3
320 bool "BSD Process Accounting version 3 file format"
321 depends on BSD_PROCESS_ACCT
324 If you say Y here, the process accounting information is written
325 in a new file format that also logs the process IDs of each
326 process and it's parent. Note that this file format is incompatible
327 with previous v0/v1/v2 file formats, so you will need updated tools
328 for processing it. A preliminary version of these tools is available
329 at <http://www.gnu.org/software/acct/>.
332 bool "open by fhandle syscalls"
335 If you say Y here, a user level program will be able to map
336 file names to handle and then later use the handle for
337 different file system operations. This is useful in implementing
338 userspace file servers, which now track files using handles instead
339 of names. The handle would remain the same even if file names
340 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
344 bool "Export task/process statistics through netlink (EXPERIMENTAL)"
348 Export selected statistics for tasks/processes through the
349 generic netlink interface. Unlike BSD process accounting, the
350 statistics are available during the lifetime of tasks/processes as
351 responses to commands. Like BSD accounting, they are sent to user
356 config TASK_DELAY_ACCT
357 bool "Enable per-task delay accounting (EXPERIMENTAL)"
360 Collect information on time spent by a task waiting for system
361 resources like cpu, synchronous block I/O completion and swapping
362 in pages. Such statistics can help in setting a task's priorities
363 relative to other tasks for cpu, io, rss limits etc.
368 bool "Enable extended accounting over taskstats (EXPERIMENTAL)"
371 Collect extended task accounting data and send the data
372 to userland for processing over the taskstats interface.
376 config TASK_IO_ACCOUNTING
377 bool "Enable per-task storage I/O accounting (EXPERIMENTAL)"
378 depends on TASK_XACCT
380 Collect information on the number of bytes of storage I/O which this
386 bool "Auditing support"
389 Enable auditing infrastructure that can be used with another
390 kernel subsystem, such as SELinux (which requires this for
391 logging of avc messages output). Does not do system-call
392 auditing without CONFIG_AUDITSYSCALL.
395 bool "Enable system-call auditing support"
396 depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH)
397 default y if SECURITY_SELINUX
399 Enable low-overhead system-call auditing infrastructure that
400 can be used independently or with another kernel subsystem,
405 depends on AUDITSYSCALL
410 depends on AUDITSYSCALL
413 source "kernel/irq/Kconfig"
418 prompt "RCU Implementation"
422 bool "Tree-based hierarchical RCU"
423 depends on !PREEMPT && SMP
425 This option selects the RCU implementation that is
426 designed for very large SMP system with hundreds or
427 thousands of CPUs. It also scales down nicely to
430 config TREE_PREEMPT_RCU
431 bool "Preemptible tree-based hierarchical RCU"
432 depends on PREEMPT && SMP
434 This option selects the RCU implementation that is
435 designed for very large SMP systems with hundreds or
436 thousands of CPUs, but for which real-time response
437 is also required. It also scales down nicely to
441 bool "UP-only small-memory-footprint RCU"
442 depends on !PREEMPT && !SMP
444 This option selects the RCU implementation that is
445 designed for UP systems from which real-time response
446 is not required. This option greatly reduces the
447 memory footprint of RCU.
449 config TINY_PREEMPT_RCU
450 bool "Preemptible UP-only small-memory-footprint RCU"
451 depends on PREEMPT && !SMP
453 This option selects the RCU implementation that is designed
454 for real-time UP systems. This option greatly reduces the
455 memory footprint of RCU.
460 def_bool ( TREE_PREEMPT_RCU || TINY_PREEMPT_RCU )
462 This option enables preemptible-RCU code that is common between
463 the TREE_PREEMPT_RCU and TINY_PREEMPT_RCU implementations.
466 bool "Enable tracing for RCU"
468 This option provides tracing in RCU which presents stats
469 in debugfs for debugging RCU implementation.
471 Say Y here if you want to enable RCU tracing
472 Say N if you are unsure.
475 int "Tree-based hierarchical RCU fanout value"
478 depends on TREE_RCU || TREE_PREEMPT_RCU
482 This option controls the fanout of hierarchical implementations
483 of RCU, allowing RCU to work efficiently on machines with
484 large numbers of CPUs. This value must be at least the fourth
485 root of NR_CPUS, which allows NR_CPUS to be insanely large.
486 The default value of RCU_FANOUT should be used for production
487 systems, but if you are stress-testing the RCU implementation
488 itself, small RCU_FANOUT values allow you to test large-system
489 code paths on small(er) systems.
491 Select a specific number if testing RCU itself.
492 Take the default if unsure.
494 config RCU_FANOUT_EXACT
495 bool "Disable tree-based hierarchical RCU auto-balancing"
496 depends on TREE_RCU || TREE_PREEMPT_RCU
499 This option forces use of the exact RCU_FANOUT value specified,
500 regardless of imbalances in the hierarchy. This is useful for
501 testing RCU itself, and might one day be useful on systems with
502 strong NUMA behavior.
504 Without RCU_FANOUT_EXACT, the code will balance the hierarchy.
508 config RCU_FAST_NO_HZ
509 bool "Accelerate last non-dyntick-idle CPU's grace periods"
510 depends on TREE_RCU && NO_HZ && SMP
513 This option causes RCU to attempt to accelerate grace periods
514 in order to allow the final CPU to enter dynticks-idle state
515 more quickly. On the other hand, this option increases the
516 overhead of the dynticks-idle checking, particularly on systems
517 with large numbers of CPUs.
519 Say Y if energy efficiency is critically important, particularly
520 if you have relatively few CPUs.
522 Say N if you are unsure.
524 config TREE_RCU_TRACE
525 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU )
528 This option provides tracing for the TREE_RCU and
529 TREE_PREEMPT_RCU implementations, permitting Makefile to
530 trivially select kernel/rcutree_trace.c.
533 bool "Enable RCU priority boosting"
534 depends on RT_MUTEXES && PREEMPT_RCU
537 This option boosts the priority of preempted RCU readers that
538 block the current preemptible RCU grace period for too long.
539 This option also prevents heavy loads from blocking RCU
540 callback invocation for all flavors of RCU.
542 Say Y here if you are working with real-time apps or heavy loads
543 Say N here if you are unsure.
545 config RCU_BOOST_PRIO
546 int "Real-time priority to boost RCU readers to"
551 This option specifies the real-time priority to which preempted
552 RCU readers are to be boosted. If you are working with CPU-bound
553 real-time applications, you should specify a priority higher then
554 the highest-priority CPU-bound application.
556 Specify the real-time priority, or take the default if unsure.
558 config RCU_BOOST_DELAY
559 int "Milliseconds to delay boosting after RCU grace-period start"
564 This option specifies the time to wait after the beginning of
565 a given grace period before priority-boosting preempted RCU
566 readers blocking that grace period. Note that any RCU reader
567 blocking an expedited RCU grace period is boosted immediately.
569 Accept the default if unsure.
571 endmenu # "RCU Subsystem"
574 tristate "Kernel .config support"
576 This option enables the complete Linux kernel ".config" file
577 contents to be saved in the kernel. It provides documentation
578 of which kernel options are used in a running kernel or in an
579 on-disk kernel. This information can be extracted from the kernel
580 image file with the script scripts/extract-ikconfig and used as
581 input to rebuild the current kernel or to build another kernel.
582 It can also be extracted from a running kernel by reading
583 /proc/config.gz if enabled (below).
586 bool "Enable access to .config through /proc/config.gz"
587 depends on IKCONFIG && PROC_FS
589 This option enables access to the kernel configuration file
590 through /proc/config.gz.
593 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
597 Select kernel log buffer size as a power of 2.
607 # Architectures with an unreliable sched_clock() should select this:
609 config HAVE_UNSTABLE_SCHED_CLOCK
613 boolean "Control Group support"
615 default !KERNEL_DESKTOP
617 This option adds support for grouping sets of processes together, for
618 use with process control subsystems such as Cpusets, CFS, memory
619 controls or device isolation.
621 - Documentation/scheduler/sched-design-CFS.txt (CFS)
622 - Documentation/cgroups/ (features for grouping, isolation
623 and resource control)
630 bool "Example debug cgroup subsystem"
633 This option enables a simple cgroup subsystem that
634 exports useful debugging information about the cgroups
639 config CGROUP_FREEZER
640 bool "Freezer cgroup subsystem"
642 Provides a way to freeze and unfreeze all tasks in a
646 bool "Device controller for cgroups"
648 Provides a cgroup implementing whitelists for devices which
649 a process in the cgroup can mknod or open.
652 bool "Cpuset support"
654 This option will let you create and manage CPUSETs which
655 allow dynamically partitioning a system into sets of CPUs and
656 Memory Nodes and assigning tasks to run only within those sets.
657 This is primarily useful on large SMP or NUMA systems.
661 config PROC_PID_CPUSET
662 bool "Include legacy /proc/<pid>/cpuset file"
666 config CGROUP_CPUACCT
667 bool "Simple CPU accounting cgroup subsystem"
669 Provides a simple Resource Controller for monitoring the
670 total CPU consumed by the tasks in a cgroup.
672 config RESOURCE_COUNTERS
673 bool "Resource counters"
675 This option enables controller independent resource accounting
676 infrastructure that works with cgroups.
678 config CGROUP_MEM_RES_CTLR
679 bool "Memory Resource Controller for Control Groups"
680 depends on RESOURCE_COUNTERS
683 Provides a memory resource controller that manages both anonymous
684 memory and page cache. (See Documentation/cgroups/memory.txt)
686 Note that setting this option increases fixed memory overhead
687 associated with each page of memory in the system. By this,
688 20(40)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory
689 usage tracking struct at boot. Total amount of this is printed out
692 Only enable when you're ok with these trade offs and really
693 sure you need the memory resource controller. Even when you enable
694 this, you can set "cgroup_disable=memory" at your boot option to
695 disable memory resource controller and you can avoid overheads.
696 (and lose benefits of memory resource controller)
698 This config option also selects MM_OWNER config option, which
699 could in turn add some fork/exit overhead.
701 config CGROUP_MEM_RES_CTLR_SWAP
702 bool "Memory Resource Controller Swap Extension"
703 depends on CGROUP_MEM_RES_CTLR && SWAP
705 Add swap management feature to memory resource controller. When you
706 enable this, you can limit mem+swap usage per cgroup. In other words,
707 when you disable this, memory resource controller has no cares to
708 usage of swap...a process can exhaust all of the swap. This extension
709 is useful when you want to avoid exhaustion swap but this itself
710 adds more overheads and consumes memory for remembering information.
711 Especially if you use 32bit system or small memory system, please
712 be careful about enabling this. When memory resource controller
713 is disabled by boot option, this will be automatically disabled and
714 there will be no overhead from this. Even when you set this config=y,
715 if boot option "swapaccount=0" is set, swap will not be accounted.
716 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page
717 size is 4096bytes, 512k per 1Gbytes of swap.
718 config CGROUP_MEM_RES_CTLR_SWAP_ENABLED
719 bool "Memory Resource Controller Swap Extension enabled by default"
720 depends on CGROUP_MEM_RES_CTLR_SWAP
723 Memory Resource Controller Swap Extension comes with its price in
724 a bigger memory consumption. General purpose distribution kernels
725 which want to enable the feature but keep it disabled by default
726 and let the user enable it by swapaccount boot command line
727 parameter should have this option unselected.
728 For those who want to have the feature enabled by default should
729 select this option (if, for some reason, they need to disable it
730 then swapaccount=0 does the trick).
733 bool "Enable perf_event per-cpu per-container group (cgroup) monitoring"
734 depends on PERF_EVENTS && CGROUPS
736 This option extends the per-cpu mode to restrict monitoring to
737 threads which belong to the cgroup specified and run on the
742 menuconfig CGROUP_SCHED
743 bool "Group CPU scheduler"
744 depends on EXPERIMENTAL
745 default !KERNEL_DESKTOP
747 This feature lets CPU scheduler recognize task groups and control CPU
748 bandwidth allocation to such task groups. It uses cgroups to group
752 config FAIR_GROUP_SCHED
753 bool "Group scheduling for SCHED_OTHER"
754 depends on CGROUP_SCHED
758 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
759 depends on EXPERIMENTAL
760 depends on FAIR_GROUP_SCHED
763 This option allows users to define CPU bandwidth rates (limits) for
764 tasks running within the fair group scheduler. Groups with no limit
765 set are considered to be unconstrained and will run with no
767 See tip/Documentation/scheduler/sched-bwc.txt for more information.
769 config RT_GROUP_SCHED
770 bool "Group scheduling for SCHED_RR/FIFO"
771 depends on EXPERIMENTAL
772 depends on CGROUP_SCHED
775 This feature lets you explicitly allocate real CPU bandwidth
776 to task groups. If enabled, it will also make it impossible to
777 schedule realtime tasks for non-root users until you allocate
778 realtime bandwidth for them.
779 See Documentation/scheduler/sched-rt-group.txt for more information.
784 tristate "Block IO controller"
788 Generic block IO controller cgroup interface. This is the common
789 cgroup interface which should be used by various IO controlling
792 Currently, CFQ IO scheduler uses it to recognize task groups and
793 control disk bandwidth allocation (proportional time slice allocation)
794 to such task groups. It is also used by bio throttling logic in
795 block layer to implement upper limit in IO rates on a device.
797 This option only enables generic Block IO controller infrastructure.
798 One needs to also enable actual IO controlling logic/policy. For
799 enabling proportional weight division of disk bandwidth in CFQ, set
800 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
801 CONFIG_BLK_DEV_THROTTLING=y.
803 See Documentation/cgroups/blkio-controller.txt for more information.
805 config DEBUG_BLK_CGROUP
806 bool "Enable Block IO controller debugging"
807 depends on BLK_CGROUP
810 Enable some debugging help. Currently it exports additional stat
811 files in a cgroup which can be useful for debugging.
815 menuconfig NAMESPACES
816 bool "Namespaces support" if EXPERT
819 Provides the way to make tasks work with different objects using
820 the same id. For example same IPC id may refer to different objects
821 or same user id or pid may refer to different tasks when used in
822 different namespaces.
830 In this namespace tasks see different info provided with the
835 depends on (SYSVIPC || POSIX_MQUEUE)
838 In this namespace tasks work with IPC ids which correspond to
839 different IPC objects in different namespaces.
842 bool "User namespace (EXPERIMENTAL)"
843 depends on EXPERIMENTAL
846 This allows containers, i.e. vservers, to use user namespaces
847 to provide different user info for different servers.
851 bool "PID Namespaces"
854 Support process id namespaces. This allows having multiple
855 processes with the same pid as long as they are in different
856 pid namespaces. This is a building block of containers.
859 bool "Network namespace"
863 Allow user space to create what appear to be multiple instances
864 of the network stack.
868 config SCHED_AUTOGROUP
869 bool "Automatic process group scheduling"
873 select FAIR_GROUP_SCHED
875 This option optimizes the scheduler for common desktop workloads by
876 automatically creating and populating task groups. This separation
877 of workloads isolates aggressive CPU burners (like build jobs) from
878 desktop applications. Task group autogeneration is currently based
884 config SYSFS_DEPRECATED
885 bool "Enable deprecated sysfs features to support old userspace tools"
889 This option adds code that switches the layout of the "block" class
890 devices, to not show up in /sys/class/block/, but only in
893 This switch is only active when the sysfs.deprecated=1 boot option is
894 passed or the SYSFS_DEPRECATED_V2 option is set.
896 This option allows new kernels to run on old distributions and tools,
897 which might get confused by /sys/class/block/. Since 2007/2008 all
898 major distributions and tools handle this just fine.
900 Recent distributions and userspace tools after 2009/2010 depend on
901 the existence of /sys/class/block/, and will not work with this
904 Only if you are using a new kernel on an old distribution, you might
907 config SYSFS_DEPRECATED_V2
908 bool "Enable deprecated sysfs features by default"
911 depends on SYSFS_DEPRECATED
913 Enable deprecated sysfs by default.
915 See the CONFIG_SYSFS_DEPRECATED option for more details about this
918 Only if you are using a new kernel on an old distribution, you might
919 need to say Y here. Even then, odds are you would not need it
920 enabled, you can always pass the boot option if absolutely necessary.
923 bool "Kernel->user space relay support (formerly relayfs)"
925 This option enables support for relay interface support in
926 certain file systems (such as debugfs).
927 It is designed to provide an efficient mechanism for tools and
928 facilities to relay large amounts of data from kernel space to
933 config BLK_DEV_INITRD
934 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
935 depends on BROKEN || !FRV
937 The initial RAM filesystem is a ramfs which is loaded by the
938 boot loader (loadlin or lilo) and that is mounted as root
939 before the normal boot procedure. It is typically used to
940 load modules needed to mount the "real" root file system,
941 etc. See <file:Documentation/initrd.txt> for details.
943 If RAM disk support (BLK_DEV_RAM) is also included, this
944 also enables initial RAM disk (initrd) support and adds
945 15 Kbytes (more on some other architectures) to the kernel size.
955 config CC_OPTIMIZE_FOR_SIZE
956 bool "Optimize for size"
958 Enabling this option will pass "-Os" instead of "-O2" to gcc
959 resulting in a smaller kernel.
970 bool "Configure standard kernel features (expert users)"
971 # Unhide debug options, to make the on-by-default options visible
974 This option allows certain base kernel options and settings
975 to be disabled or tweaked. This is for specialized
976 environments which can tolerate a "non-standard" kernel.
977 Only use this if you really know what you are doing.
980 bool "Enable 16-bit UID system calls" if EXPERT
981 depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && COMPAT) || UML || (X86_64 && IA32_EMULATION)
984 This enables the legacy 16-bit UID syscall wrappers.
986 config SYSCTL_SYSCALL
987 bool "Sysctl syscall support" if EXPERT
988 depends on PROC_SYSCTL
992 sys_sysctl uses binary paths that have been found challenging
993 to properly maintain and use. The interface in /proc/sys
994 using paths with ascii names is now the primary path to this
997 Almost nothing using the binary sysctl interface so if you are
998 trying to save some space it is probably safe to disable this,
999 making your kernel marginally smaller.
1001 If unsure say N here.
1004 bool "Load all symbols for debugging/ksymoops" if EXPERT
1007 Say Y here to let the kernel print out symbolic crash information and
1008 symbolic stack backtraces. This increases the size of the kernel
1009 somewhat, as all symbols have to be loaded into the kernel image.
1012 bool "Include all symbols in kallsyms"
1013 depends on DEBUG_KERNEL && KALLSYMS
1015 Normally kallsyms only contains the symbols of functions for nicer
1016 OOPS messages and backtraces (i.e., symbols from the text and inittext
1017 sections). This is sufficient for most cases. And only in very rare
1018 cases (e.g., when a debugger is used) all symbols are required (e.g.,
1019 names of variables from the data sections, etc).
1021 This option makes sure that all symbols are loaded into the kernel
1022 image (i.e., symbols from all sections) in cost of increased kernel
1023 size (depending on the kernel configuration, it may be 300KiB or
1024 something like this).
1026 Say N unless you really need all symbols.
1029 bool "Support for hot-pluggable devices" if EXPERT
1032 This option is provided for the case where no hotplug or uevent
1033 capabilities is wanted by the kernel. You should only consider
1034 disabling this option for embedded systems that do not use modules, a
1035 dynamic /dev tree, or dynamic device discovery. Just say Y.
1039 bool "Enable support for printk" if EXPERT
1041 This option enables normal printk support. Removing it
1042 eliminates most of the message strings from the kernel image
1043 and makes the kernel more or less silent. As this makes it
1044 very difficult to diagnose system problems, saying N here is
1045 strongly discouraged.
1048 bool "BUG() support" if EXPERT
1051 Disabling this option eliminates support for BUG and WARN, reducing
1052 the size of your kernel image and potentially quietly ignoring
1053 numerous fatal conditions. You should only consider disabling this
1054 option for embedded systems with no facilities for reporting errors.
1059 bool "Enable ELF core dumps" if EXPERT
1061 Enable support for generating core dumps. Disabling saves about 4k.
1064 config PCSPKR_PLATFORM
1065 bool "Enable PC-Speaker support" if EXPERT
1066 depends on HAVE_PCSPKR_PLATFORM
1070 This option allows to disable the internal PC-Speaker
1071 support, saving some memory.
1073 config HAVE_PCSPKR_PLATFORM
1078 bool "Enable full-sized data structures for core" if EXPERT
1080 Disabling this option reduces the size of miscellaneous core
1081 kernel data structures. This saves memory on small machines,
1082 but may reduce performance.
1085 bool "Enable futex support" if EXPERT
1089 Disabling this option will cause the kernel to be built without
1090 support for "fast userspace mutexes". The resulting kernel may not
1091 run glibc-based applications correctly.
1094 bool "Enable eventpoll support" if EXPERT
1098 Disabling this option will cause the kernel to be built without
1099 support for epoll family of system calls.
1102 bool "Enable signalfd() system call" if EXPERT
1106 Enable the signalfd() system call that allows to receive signals
1107 on a file descriptor.
1112 bool "Enable timerfd() system call" if EXPERT
1116 Enable the timerfd() system call that allows to receive timer
1117 events on a file descriptor.
1122 bool "Enable eventfd() system call" if EXPERT
1126 Enable the eventfd() system call that allows to receive both
1127 kernel notification (ie. KAIO) or userspace notifications.
1132 bool "Use full shmem filesystem" if EXPERT
1136 The shmem is an internal filesystem used to manage shared memory.
1137 It is backed by swap and manages resource limits. It is also exported
1138 to userspace as tmpfs if TMPFS is enabled. Disabling this
1139 option replaces shmem and tmpfs with the much simpler ramfs code,
1140 which may be appropriate on small systems without swap.
1143 bool "Enable AIO support" if EXPERT
1146 This option enables POSIX asynchronous I/O which may by used
1147 by some high performance threaded applications. Disabling
1148 this option saves about 7k.
1151 bool "Embedded system"
1154 This option should be enabled if compiling the kernel for
1155 an embedded system so certain expert options are available
1158 config HAVE_PERF_EVENTS
1161 See tools/perf/design.txt for details.
1163 config PERF_USE_VMALLOC
1166 See tools/perf/design.txt for details
1168 menu "Kernel Performance Events And Counters"
1171 bool "Kernel performance events and counters"
1172 default y if (PROFILING || PERF_COUNTERS)
1173 depends on HAVE_PERF_EVENTS
1177 Enable kernel support for various performance events provided
1178 by software and hardware.
1180 Software events are supported either built-in or via the
1181 use of generic tracepoints.
1183 Most modern CPUs support performance events via performance
1184 counter registers. These registers count the number of certain
1185 types of hw events: such as instructions executed, cachemisses
1186 suffered, or branches mis-predicted - without slowing down the
1187 kernel or applications. These registers can also trigger interrupts
1188 when a threshold number of events have passed - and can thus be
1189 used to profile the code that runs on that CPU.
1191 The Linux Performance Event subsystem provides an abstraction of
1192 these software and hardware event capabilities, available via a
1193 system call and used by the "perf" utility in tools/perf/. It
1194 provides per task and per CPU counters, and it provides event
1195 capabilities on top of those.
1199 config PERF_COUNTERS
1200 bool "Kernel performance counters (old config option)"
1201 depends on HAVE_PERF_EVENTS
1203 This config has been obsoleted by the PERF_EVENTS
1204 config option - please see that one for details.
1206 It has no effect on the kernel whether you enable
1207 it or not, it is a compatibility placeholder.
1211 config DEBUG_PERF_USE_VMALLOC
1213 bool "Debug: use vmalloc to back perf mmap() buffers"
1214 depends on PERF_EVENTS && DEBUG_KERNEL
1215 select PERF_USE_VMALLOC
1217 Use vmalloc memory to back perf mmap() buffers.
1219 Mostly useful for debugging the vmalloc code on platforms
1220 that don't require it.
1226 config VM_EVENT_COUNTERS
1228 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1230 VM event counters are needed for event counts to be shown.
1231 This option allows the disabling of the VM event counters
1232 on EXPERT systems. /proc/vmstat will only show page counts
1233 if VM event counters are disabled.
1237 bool "Enable PCI quirk workarounds" if EXPERT
1240 This enables workarounds for various PCI chipset
1241 bugs/quirks. Disable this only if your target machine is
1242 unaffected by PCI quirks.
1246 bool "Enable SLUB debugging support" if EXPERT
1247 depends on SLUB && SYSFS
1249 SLUB has extensive debug support features. Disabling these can
1250 result in significant savings in code size. This also disables
1251 SLUB sysfs support. /sys/slab will not exist and there will be
1252 no support for cache validation etc.
1255 bool "Disable heap randomization"
1258 Randomizing heap placement makes heap exploits harder, but it
1259 also breaks ancient binaries (including anything libc5 based).
1260 This option changes the bootup default to heap randomization
1261 disabled, and can be overridden at runtime by setting
1262 /proc/sys/kernel/randomize_va_space to 2.
1264 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1267 prompt "Choose SLAB allocator"
1270 This option allows to select a slab allocator.
1275 The regular slab allocator that is established and known to work
1276 well in all environments. It organizes cache hot objects in
1277 per cpu and per node queues.
1280 bool "SLUB (Unqueued Allocator)"
1282 SLUB is a slab allocator that minimizes cache line usage
1283 instead of managing queues of cached objects (SLAB approach).
1284 Per cpu caching is realized using slabs of objects instead
1285 of queues of objects. SLUB can use memory efficiently
1286 and has enhanced diagnostics. SLUB is the default choice for
1291 bool "SLOB (Simple Allocator)"
1293 SLOB replaces the stock allocator with a drastically simpler
1294 allocator. SLOB is generally more space efficient but
1295 does not perform as well on large systems.
1299 config MMAP_ALLOW_UNINITIALIZED
1300 bool "Allow mmapped anonymous memory to be uninitialized"
1301 depends on EXPERT && !MMU
1304 Normally, and according to the Linux spec, anonymous memory obtained
1305 from mmap() has it's contents cleared before it is passed to
1306 userspace. Enabling this config option allows you to request that
1307 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1308 providing a huge performance boost. If this option is not enabled,
1309 then the flag will be ignored.
1311 This is taken advantage of by uClibc's malloc(), and also by
1312 ELF-FDPIC binfmt's brk and stack allocator.
1314 Because of the obvious security issues, this option should only be
1315 enabled on embedded devices where you control what is run in
1316 userspace. Since that isn't generally a problem on no-MMU systems,
1317 it is normally safe to say Y here.
1319 See Documentation/nommu-mmap.txt for more information.
1322 bool "Profiling support"
1324 Say Y here to enable the extended profiling support mechanisms used
1325 by profilers such as OProfile.
1328 # Place an empty function call at each tracepoint site. Can be
1329 # dynamically changed for a probe function.
1334 source "arch/Kconfig"
1336 endmenu # General setup
1338 config HAVE_GENERIC_DMA_COHERENT
1345 depends on SLAB || SLUB_DEBUG
1353 default 0 if BASE_FULL
1354 default 1 if !BASE_FULL
1357 bool "Enable loadable module support"
1359 Kernel modules are small pieces of compiled code which can
1360 be inserted in the running kernel, rather than being
1361 permanently built into the kernel. You use the "modprobe"
1362 tool to add (and sometimes remove) them. If you say Y here,
1363 many parts of the kernel can be built as modules (by
1364 answering M instead of Y where indicated): this is most
1365 useful for infrequently used options which are not required
1366 for booting. For more information, see the man pages for
1367 modprobe, lsmod, modinfo, insmod and rmmod.
1369 If you say Y here, you will need to run "make
1370 modules_install" to put the modules under /lib/modules/
1371 where modprobe can find them (you may need to be root to do
1378 config MODULE_FORCE_LOAD
1379 bool "Forced module loading"
1382 Allow loading of modules without version information (ie. modprobe
1383 --force). Forced module loading sets the 'F' (forced) taint flag and
1384 is usually a really bad idea.
1386 config MODULE_UNLOAD
1387 bool "Module unloading"
1389 Without this option you will not be able to unload any
1390 modules (note that some modules may not be unloadable
1391 anyway), which makes your kernel smaller, faster
1392 and simpler. If unsure, say Y.
1394 config MODULE_FORCE_UNLOAD
1395 bool "Forced module unloading"
1396 depends on MODULE_UNLOAD && EXPERIMENTAL
1398 This option allows you to force a module to unload, even if the
1399 kernel believes it is unsafe: the kernel will remove the module
1400 without waiting for anyone to stop using it (using the -f option to
1401 rmmod). This is mainly for kernel developers and desperate users.
1405 bool "Module versioning support"
1407 Usually, you have to use modules compiled with your kernel.
1408 Saying Y here makes it sometimes possible to use modules
1409 compiled for different kernels, by adding enough information
1410 to the modules to (hopefully) spot any changes which would
1411 make them incompatible with the kernel you are running. If
1414 config MODULE_SRCVERSION_ALL
1415 bool "Source checksum for all modules"
1417 Modules which contain a MODULE_VERSION get an extra "srcversion"
1418 field inserted into their modinfo section, which contains a
1419 sum of the source files which made it. This helps maintainers
1420 see exactly which source was used to build a module (since
1421 others sometimes change the module source without updating
1422 the version). With this option, such a "srcversion" field
1423 will be created for all modules. If unsure, say N.
1427 config INIT_ALL_POSSIBLE
1430 Back when each arch used to define their own cpu_online_map and
1431 cpu_possible_map, some of them chose to initialize cpu_possible_map
1432 with all 1s, and others with all 0s. When they were centralised,
1433 it was better to provide this option than to break all the archs
1434 and have several arch maintainers pursuing me down dark alleys.
1439 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU
1441 Need stop_machine() primitive.
1443 source "block/Kconfig"
1445 config PREEMPT_NOTIFIERS
1452 source "kernel/Kconfig.locks"