7 option env="KERNELVERSION"
13 default "/lib/modules/$UNAME_RELEASE/.config"
14 default "/etc/kernel-config"
15 default "/boot/config-$UNAME_RELEASE"
16 default "$ARCH_DEFCONFIG"
17 default "arch/$ARCH/defconfig"
27 bool "Prompt for development and/or incomplete code/drivers"
29 Some of the various things that Linux supports (such as network
30 drivers, file systems, network protocols, etc.) can be in a state
31 of development where the functionality, stability, or the level of
32 testing is not yet high enough for general use. This is usually
33 known as the "alpha-test" phase among developers. If a feature is
34 currently in alpha-test, then the developers usually discourage
35 uninformed widespread use of this feature by the general public to
36 avoid "Why doesn't this work?" type mail messages. However, active
37 testing and use of these systems is welcomed. Just be aware that it
38 may not meet the normal level of reliability or it may fail to work
39 in some special cases. Detailed bug reports from people familiar
40 with the kernel internals are usually welcomed by the developers
41 (before submitting bug reports, please read the documents
42 <file:README>, <file:MAINTAINERS>, <file:REPORTING-BUGS>,
43 <file:Documentation/BUG-HUNTING>, and
44 <file:Documentation/oops-tracing.txt> in the kernel source).
46 This option will also make obsoleted drivers available. These are
47 drivers that have been replaced by something else, and/or are
48 scheduled to be removed in a future kernel release.
50 Unless you intend to help test and develop a feature or driver that
51 falls into this category, or you have a situation that requires
52 using these features, you should probably say N here, which will
53 cause the configurator to present you with fewer choices. If
54 you say Y here, you will be offered the choice of using features or
55 drivers that are currently considered to be in the alpha-test phase.
62 depends on BROKEN || !SMP
67 depends on SMP || PREEMPT
70 config INIT_ENV_ARG_LIMIT
75 Maximum of each of the number of arguments and environment
76 variables passed to init from the kernel command line.
80 string "Local version - append to kernel release"
82 Append an extra string to the end of your kernel version.
83 This will show up when you type uname, for example.
84 The string you set here will be appended after the contents of
85 any files with a filename matching localversion* in your
86 object and source tree, in that order. Your total string can
87 be a maximum of 64 characters.
89 config LOCALVERSION_AUTO
90 bool "Automatically append version information to the version string"
93 This will try to automatically determine if the current tree is a
94 release tree by looking for git tags that belong to the current
97 A string of the format -gxxxxxxxx will be added to the localversion
98 if a git-based tree is found. The string generated by this will be
99 appended after any matching localversion* files, and after the value
100 set in CONFIG_LOCALVERSION.
102 (The actual string used here is the first eight characters produced
103 by running the command:
105 $ git rev-parse --verify HEAD
107 which is done within the script "scripts/setlocalversion".)
109 config HAVE_KERNEL_GZIP
112 config HAVE_KERNEL_BZIP2
115 config HAVE_KERNEL_LZMA
118 config HAVE_KERNEL_LZO
122 prompt "Kernel compression mode"
123 default KERNEL_LZO if ARCH_RK29
125 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_LZO
127 The linux kernel is a kind of self-extracting executable.
128 Several compression algorithms are available, which differ
129 in efficiency, compression and decompression speed.
130 Compression speed is only relevant when building a kernel.
131 Decompression speed is relevant at each boot.
133 If you have any problems with bzip2 or lzma compressed
134 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
135 version of this functionality (bzip2 only), for 2.4, was
136 supplied by Christian Ludwig)
138 High compression options are mostly useful for users, who
139 are low on disk space (embedded systems), but for whom ram
142 If in doubt, select 'gzip'
146 depends on HAVE_KERNEL_GZIP
148 The old and tried gzip compression. It provides a good balance
149 between compression ratio and decompression speed.
153 depends on HAVE_KERNEL_BZIP2
155 Its compression ratio and speed is intermediate.
156 Decompression speed is slowest among the three. The kernel
157 size is about 10% smaller with bzip2, in comparison to gzip.
158 Bzip2 uses a large amount of memory. For modern kernels you
159 will need at least 8MB RAM or more for booting.
163 depends on HAVE_KERNEL_LZMA
165 The most recent compression algorithm.
166 Its ratio is best, decompression speed is between the other
167 two. Compression is slowest. The kernel size is about 33%
168 smaller with LZMA in comparison to gzip.
172 depends on HAVE_KERNEL_LZO
174 Its compression ratio is the poorest among the 4. The kernel
175 size is about about 10% bigger than gzip; however its speed
176 (both compression and decompression) is the fastest.
181 bool "Support for paging of anonymous memory (swap)"
182 depends on MMU && BLOCK
185 This option allows you to choose whether you want to have support
186 for so called swap devices or swap files in your kernel that are
187 used to provide more virtual memory than the actual RAM present
188 in your computer. If unsure say Y.
193 Inter Process Communication is a suite of library functions and
194 system calls which let processes (running programs) synchronize and
195 exchange information. It is generally considered to be a good thing,
196 and some programs won't run unless you say Y here. In particular, if
197 you want to run the DOS emulator dosemu under Linux (read the
198 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
199 you'll need to say Y here.
201 You can find documentation about IPC with "info ipc" and also in
202 section 6.4 of the Linux Programmer's Guide, available from
203 <http://www.tldp.org/guides.html>.
205 config SYSVIPC_SYSCTL
212 bool "POSIX Message Queues"
213 depends on NET && EXPERIMENTAL
215 POSIX variant of message queues is a part of IPC. In POSIX message
216 queues every message has a priority which decides about succession
217 of receiving it by a process. If you want to compile and run
218 programs written e.g. for Solaris with use of its POSIX message
219 queues (functions mq_*) say Y here.
221 POSIX message queues are visible as a filesystem called 'mqueue'
222 and can be mounted somewhere if you want to do filesystem
223 operations on message queues.
227 config POSIX_MQUEUE_SYSCTL
229 depends on POSIX_MQUEUE
233 config BSD_PROCESS_ACCT
234 bool "BSD Process Accounting"
236 If you say Y here, a user level program will be able to instruct the
237 kernel (via a special system call) to write process accounting
238 information to a file: whenever a process exits, information about
239 that process will be appended to the file by the kernel. The
240 information includes things such as creation time, owning user,
241 command name, memory usage, controlling terminal etc. (the complete
242 list is in the struct acct in <file:include/linux/acct.h>). It is
243 up to the user level program to do useful things with this
244 information. This is generally a good idea, so say Y.
246 config BSD_PROCESS_ACCT_V3
247 bool "BSD Process Accounting version 3 file format"
248 depends on BSD_PROCESS_ACCT
251 If you say Y here, the process accounting information is written
252 in a new file format that also logs the process IDs of each
253 process and it's parent. Note that this file format is incompatible
254 with previous v0/v1/v2 file formats, so you will need updated tools
255 for processing it. A preliminary version of these tools is available
256 at <http://www.gnu.org/software/acct/>.
259 bool "Export task/process statistics through netlink (EXPERIMENTAL)"
263 Export selected statistics for tasks/processes through the
264 generic netlink interface. Unlike BSD process accounting, the
265 statistics are available during the lifetime of tasks/processes as
266 responses to commands. Like BSD accounting, they are sent to user
271 config TASK_DELAY_ACCT
272 bool "Enable per-task delay accounting (EXPERIMENTAL)"
275 Collect information on time spent by a task waiting for system
276 resources like cpu, synchronous block I/O completion and swapping
277 in pages. Such statistics can help in setting a task's priorities
278 relative to other tasks for cpu, io, rss limits etc.
283 bool "Enable extended accounting over taskstats (EXPERIMENTAL)"
286 Collect extended task accounting data and send the data
287 to userland for processing over the taskstats interface.
291 config TASK_IO_ACCOUNTING
292 bool "Enable per-task storage I/O accounting (EXPERIMENTAL)"
293 depends on TASK_XACCT
295 Collect information on the number of bytes of storage I/O which this
301 bool "Auditing support"
304 Enable auditing infrastructure that can be used with another
305 kernel subsystem, such as SELinux (which requires this for
306 logging of avc messages output). Does not do system-call
307 auditing without CONFIG_AUDITSYSCALL.
310 bool "Enable system-call auditing support"
311 depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH)
312 default y if SECURITY_SELINUX
314 Enable low-overhead system-call auditing infrastructure that
315 can be used independently or with another kernel subsystem,
316 such as SELinux. To use audit's filesystem watch feature, please
317 ensure that INOTIFY is configured.
321 depends on AUDITSYSCALL
327 prompt "RCU Implementation"
331 bool "Tree-based hierarchical RCU"
333 This option selects the RCU implementation that is
334 designed for very large SMP system with hundreds or
335 thousands of CPUs. It also scales down nicely to
338 config TREE_PREEMPT_RCU
339 bool "Preemptable tree-based hierarchical RCU"
342 This option selects the RCU implementation that is
343 designed for very large SMP systems with hundreds or
344 thousands of CPUs, but for which real-time response
345 is also required. It also scales down nicely to
351 bool "Enable tracing for RCU"
352 depends on TREE_RCU || TREE_PREEMPT_RCU
354 This option provides tracing in RCU which presents stats
355 in debugfs for debugging RCU implementation.
357 Say Y here if you want to enable RCU tracing
358 Say N if you are unsure.
361 int "Tree-based hierarchical RCU fanout value"
364 depends on TREE_RCU || TREE_PREEMPT_RCU
368 This option controls the fanout of hierarchical implementations
369 of RCU, allowing RCU to work efficiently on machines with
370 large numbers of CPUs. This value must be at least the cube
371 root of NR_CPUS, which allows NR_CPUS up to 32,768 for 32-bit
372 systems and up to 262,144 for 64-bit systems.
374 Select a specific number if testing RCU itself.
375 Take the default if unsure.
377 config RCU_FANOUT_EXACT
378 bool "Disable tree-based hierarchical RCU auto-balancing"
379 depends on TREE_RCU || TREE_PREEMPT_RCU
382 This option forces use of the exact RCU_FANOUT value specified,
383 regardless of imbalances in the hierarchy. This is useful for
384 testing RCU itself, and might one day be useful on systems with
385 strong NUMA behavior.
387 Without RCU_FANOUT_EXACT, the code will balance the hierarchy.
391 config TREE_RCU_TRACE
392 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU )
395 This option provides tracing for the TREE_RCU and
396 TREE_PREEMPT_RCU implementations, permitting Makefile to
397 trivially select kernel/rcutree_trace.c.
399 endmenu # "RCU Subsystem"
402 tristate "Kernel .config support"
404 This option enables the complete Linux kernel ".config" file
405 contents to be saved in the kernel. It provides documentation
406 of which kernel options are used in a running kernel or in an
407 on-disk kernel. This information can be extracted from the kernel
408 image file with the script scripts/extract-ikconfig and used as
409 input to rebuild the current kernel or to build another kernel.
410 It can also be extracted from a running kernel by reading
411 /proc/config.gz if enabled (below).
414 bool "Enable access to .config through /proc/config.gz"
415 depends on IKCONFIG && PROC_FS
417 This option enables access to the kernel configuration file
418 through /proc/config.gz.
421 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
425 Select kernel log buffer size as a power of 2.
435 # Architectures with an unreliable sched_clock() should select this:
437 config HAVE_UNSTABLE_SCHED_CLOCK
441 boolean "Control Group support"
443 This option adds support for grouping sets of processes together, for
444 use with process control subsystems such as Cpusets, CFS, memory
445 controls or device isolation.
447 - Documentation/scheduler/sched-design-CFS.txt (CFS)
448 - Documentation/cgroups/ (features for grouping, isolation
449 and resource control)
456 bool "Example debug cgroup subsystem"
460 This option enables a simple cgroup subsystem that
461 exports useful debugging information about the cgroups
467 bool "Namespace cgroup subsystem"
470 Provides a simple namespace cgroup subsystem to
471 provide hierarchical naming of sets of namespaces,
472 for instance virtual servers and checkpoint/restart
475 config CGROUP_FREEZER
476 bool "Freezer cgroup subsystem"
479 Provides a way to freeze and unfreeze all tasks in a
483 bool "Device controller for cgroups"
484 depends on CGROUPS && EXPERIMENTAL
486 Provides a cgroup implementing whitelists for devices which
487 a process in the cgroup can mknod or open.
490 bool "Cpuset support"
493 This option will let you create and manage CPUSETs which
494 allow dynamically partitioning a system into sets of CPUs and
495 Memory Nodes and assigning tasks to run only within those sets.
496 This is primarily useful on large SMP or NUMA systems.
500 config PROC_PID_CPUSET
501 bool "Include legacy /proc/<pid>/cpuset file"
505 config CGROUP_CPUACCT
506 bool "Simple CPU accounting cgroup subsystem"
509 Provides a simple Resource Controller for monitoring the
510 total CPU consumed by the tasks in a cgroup.
512 config RESOURCE_COUNTERS
513 bool "Resource counters"
515 This option enables controller independent resource accounting
516 infrastructure that works with cgroups.
519 config CGROUP_MEM_RES_CTLR
520 bool "Memory Resource Controller for Control Groups"
521 depends on CGROUPS && RESOURCE_COUNTERS
524 Provides a memory resource controller that manages both anonymous
525 memory and page cache. (See Documentation/cgroups/memory.txt)
527 Note that setting this option increases fixed memory overhead
528 associated with each page of memory in the system. By this,
529 20(40)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory
530 usage tracking struct at boot. Total amount of this is printed out
533 Only enable when you're ok with these trade offs and really
534 sure you need the memory resource controller. Even when you enable
535 this, you can set "cgroup_disable=memory" at your boot option to
536 disable memory resource controller and you can avoid overheads.
537 (and lose benefits of memory resource controller)
539 This config option also selects MM_OWNER config option, which
540 could in turn add some fork/exit overhead.
542 config CGROUP_MEM_RES_CTLR_SWAP
543 bool "Memory Resource Controller Swap Extension(EXPERIMENTAL)"
544 depends on CGROUP_MEM_RES_CTLR && SWAP && EXPERIMENTAL
546 Add swap management feature to memory resource controller. When you
547 enable this, you can limit mem+swap usage per cgroup. In other words,
548 when you disable this, memory resource controller has no cares to
549 usage of swap...a process can exhaust all of the swap. This extension
550 is useful when you want to avoid exhaustion swap but this itself
551 adds more overheads and consumes memory for remembering information.
552 Especially if you use 32bit system or small memory system, please
553 be careful about enabling this. When memory resource controller
554 is disabled by boot option, this will be automatically disabled and
555 there will be no overhead from this. Even when you set this config=y,
556 if boot option "noswapaccount" is set, swap will not be accounted.
557 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page
558 size is 4096bytes, 512k per 1Gbytes of swap.
560 menuconfig CGROUP_SCHED
561 bool "Group CPU scheduler"
562 depends on EXPERIMENTAL && CGROUPS
565 This feature lets CPU scheduler recognize task groups and control CPU
566 bandwidth allocation to such task groups. It uses cgroups to group
570 config FAIR_GROUP_SCHED
571 bool "Group scheduling for SCHED_OTHER"
572 depends on CGROUP_SCHED
575 config RT_GROUP_SCHED
576 bool "Group scheduling for SCHED_RR/FIFO"
577 depends on EXPERIMENTAL
578 depends on CGROUP_SCHED
581 This feature lets you explicitly allocate real CPU bandwidth
582 to task groups. If enabled, it will also make it impossible to
583 schedule realtime tasks for non-root users until you allocate
584 realtime bandwidth for them.
585 See Documentation/scheduler/sched-rt-group.txt for more information.
594 config SYSFS_DEPRECATED
597 config SYSFS_DEPRECATED_V2
598 bool "enable deprecated sysfs features which may confuse old userspace tools"
601 select SYSFS_DEPRECATED
603 This option switches the layout of sysfs to the deprecated
604 version. Do not use it on recent distributions.
606 The current sysfs layout features a unified device tree at
607 /sys/devices/, which is able to express a hierarchy between
608 class devices. If the deprecated option is set to Y, the
609 unified device tree is split into a bus device tree at
610 /sys/devices/ and several individual class device trees at
611 /sys/class/. The class and bus devices will be connected by
612 "<subsystem>:<name>" and the "device" links. The "block"
613 class devices, will not show up in /sys/class/block/. Some
614 subsystems will suppress the creation of some devices which
615 depend on the unified device tree.
617 This option is not a pure compatibility option that can
618 be safely enabled on newer distributions. It will change the
619 layout of sysfs to the non-extensible deprecated version,
620 and disable some features, which can not be exported without
621 confusing older userspace tools. Since 2007/2008 all major
622 distributions do not enable this option, and ship no tools which
623 depend on the deprecated layout or this option.
625 If you are using a new kernel on an older distribution, or use
626 older userspace tools, you might need to say Y here. Do not say Y,
627 if the original kernel, that came with your distribution, has
628 this option set to N.
631 bool "Kernel->user space relay support (formerly relayfs)"
633 This option enables support for relay interface support in
634 certain file systems (such as debugfs).
635 It is designed to provide an efficient mechanism for tools and
636 facilities to relay large amounts of data from kernel space to
642 bool "Namespaces support" if EMBEDDED
645 Provides the way to make tasks work with different objects using
646 the same id. For example same IPC id may refer to different objects
647 or same user id or pid may refer to different tasks when used in
648 different namespaces.
652 depends on NAMESPACES
654 In this namespace tasks see different info provided with the
659 depends on NAMESPACES && (SYSVIPC || POSIX_MQUEUE)
661 In this namespace tasks work with IPC ids which correspond to
662 different IPC objects in different namespaces.
665 bool "User namespace (EXPERIMENTAL)"
666 depends on NAMESPACES && EXPERIMENTAL
668 This allows containers, i.e. vservers, to use user namespaces
669 to provide different user info for different servers.
673 bool "PID Namespaces (EXPERIMENTAL)"
675 depends on NAMESPACES && EXPERIMENTAL
677 Support process id namespaces. This allows having multiple
678 processes with the same pid as long as they are in different
679 pid namespaces. This is a building block of containers.
681 Unless you want to work with an experimental feature
685 bool "Network namespace"
687 depends on NAMESPACES && EXPERIMENTAL && NET
689 Allow user space to create what appear to be multiple instances
690 of the network stack.
692 config BLK_DEV_INITRD
693 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
694 depends on BROKEN || !FRV
696 The initial RAM filesystem is a ramfs which is loaded by the
697 boot loader (loadlin or lilo) and that is mounted as root
698 before the normal boot procedure. It is typically used to
699 load modules needed to mount the "real" root file system,
700 etc. See <file:Documentation/initrd.txt> for details.
702 If RAM disk support (BLK_DEV_RAM) is also included, this
703 also enables initial RAM disk (initrd) support and adds
704 15 Kbytes (more on some other architectures) to the kernel size.
714 config CC_OPTIMIZE_FOR_SIZE
715 bool "Optimize for size"
718 Enabling this option will pass "-Os" instead of "-O2" to gcc
719 resulting in a smaller kernel.
730 int "Default panic timeout"
733 Set default panic timeout.
736 bool "Configure standard kernel features (for small systems)"
738 This option allows certain base kernel options and settings
739 to be disabled or tweaked. This is for specialized
740 environments which can tolerate a "non-standard" kernel.
741 Only use this if you really know what you are doing.
744 bool "Enable 16-bit UID system calls" if EMBEDDED
745 depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && COMPAT) || UML || (X86_64 && IA32_EMULATION)
748 This enables the legacy 16-bit UID syscall wrappers.
750 config SYSCTL_SYSCALL
751 bool "Sysctl syscall support" if EMBEDDED
755 sys_sysctl uses binary paths that have been found challenging
756 to properly maintain and use. The interface in /proc/sys
757 using paths with ascii names is now the primary path to this
760 Almost nothing using the binary sysctl interface so if you are
761 trying to save some space it is probably safe to disable this,
762 making your kernel marginally smaller.
764 If unsure say Y here.
767 bool "Load all symbols for debugging/ksymoops" if EMBEDDED
770 Say Y here to let the kernel print out symbolic crash information and
771 symbolic stack backtraces. This increases the size of the kernel
772 somewhat, as all symbols have to be loaded into the kernel image.
775 bool "Include all symbols in kallsyms"
776 depends on DEBUG_KERNEL && KALLSYMS
778 Normally kallsyms only contains the symbols of functions, for nicer
779 OOPS messages. Some debuggers can use kallsyms for other
780 symbols too: say Y here to include all symbols, if you need them
781 and you don't care about adding 300k to the size of your kernel.
785 config KALLSYMS_EXTRA_PASS
786 bool "Do an extra kallsyms pass"
789 If kallsyms is not working correctly, the build will fail with
790 inconsistent kallsyms data. If that occurs, log a bug report and
791 turn on KALLSYMS_EXTRA_PASS which should result in a stable build.
792 Always say N here unless you find a bug in kallsyms, which must be
793 reported. KALLSYMS_EXTRA_PASS is only a temporary workaround while
794 you wait for kallsyms to be fixed.
798 bool "Support for hot-pluggable devices" if EMBEDDED
801 This option is provided for the case where no hotplug or uevent
802 capabilities is wanted by the kernel. You should only consider
803 disabling this option for embedded systems that do not use modules, a
804 dynamic /dev tree, or dynamic device discovery. Just say Y.
808 bool "Enable support for printk" if EMBEDDED
810 This option enables normal printk support. Removing it
811 eliminates most of the message strings from the kernel image
812 and makes the kernel more or less silent. As this makes it
813 very difficult to diagnose system problems, saying N here is
814 strongly discouraged.
817 bool "BUG() support" if EMBEDDED
820 Disabling this option eliminates support for BUG and WARN, reducing
821 the size of your kernel image and potentially quietly ignoring
822 numerous fatal conditions. You should only consider disabling this
823 option for embedded systems with no facilities for reporting errors.
828 bool "Enable ELF core dumps" if EMBEDDED
830 Enable support for generating core dumps. Disabling saves about 4k.
832 config PCSPKR_PLATFORM
833 bool "Enable PC-Speaker support" if EMBEDDED
834 depends on ALPHA || X86 || MIPS || PPC_PREP || PPC_CHRP || PPC_PSERIES
837 This option allows to disable the internal PC-Speaker
838 support, saving some memory.
842 bool "Enable full-sized data structures for core" if EMBEDDED
844 Disabling this option reduces the size of miscellaneous core
845 kernel data structures. This saves memory on small machines,
846 but may reduce performance.
849 bool "Enable futex support" if EMBEDDED
853 Disabling this option will cause the kernel to be built without
854 support for "fast userspace mutexes". The resulting kernel may not
855 run glibc-based applications correctly.
858 bool "Enable eventpoll support" if EMBEDDED
862 Disabling this option will cause the kernel to be built without
863 support for epoll family of system calls.
866 bool "Enable signalfd() system call" if EMBEDDED
870 Enable the signalfd() system call that allows to receive signals
871 on a file descriptor.
876 bool "Enable timerfd() system call" if EMBEDDED
880 Enable the timerfd() system call that allows to receive timer
881 events on a file descriptor.
886 bool "Enable eventfd() system call" if EMBEDDED
890 Enable the eventfd() system call that allows to receive both
891 kernel notification (ie. KAIO) or userspace notifications.
896 bool "Use full shmem filesystem" if EMBEDDED
900 The shmem is an internal filesystem used to manage shared memory.
901 It is backed by swap and manages resource limits. It is also exported
902 to userspace as tmpfs if TMPFS is enabled. Disabling this
903 option replaces shmem and tmpfs with the much simpler ramfs code,
904 which may be appropriate on small systems without swap.
907 bool "Enable the Anonymous Shared Memory Subsystem"
909 depends on SHMEM || TINY_SHMEM
911 The ashmem subsystem is a new shared memory allocator, similar to
912 POSIX SHM but with different behavior and sporting a simpler
916 bool "Enable AIO support" if EMBEDDED
919 This option enables POSIX asynchronous I/O which may by used
920 by some high performance threaded applications. Disabling
921 this option saves about 7k.
923 config HAVE_PERF_EVENTS
926 See tools/perf/design.txt for details.
928 config PERF_USE_VMALLOC
931 See tools/perf/design.txt for details
933 menu "Kernel Performance Events And Counters"
936 bool "Kernel performance events and counters"
937 default y if (PROFILING || PERF_COUNTERS)
938 depends on HAVE_PERF_EVENTS
941 Enable kernel support for various performance events provided
942 by software and hardware.
944 Software events are supported either built-in or via the
945 use of generic tracepoints.
947 Most modern CPUs support performance events via performance
948 counter registers. These registers count the number of certain
949 types of hw events: such as instructions executed, cachemisses
950 suffered, or branches mis-predicted - without slowing down the
951 kernel or applications. These registers can also trigger interrupts
952 when a threshold number of events have passed - and can thus be
953 used to profile the code that runs on that CPU.
955 The Linux Performance Event subsystem provides an abstraction of
956 these software and hardware event capabilities, available via a
957 system call and used by the "perf" utility in tools/perf/. It
958 provides per task and per CPU counters, and it provides event
959 capabilities on top of those.
964 bool "Tracepoint profiling sources"
965 depends on PERF_EVENTS && EVENT_TRACING
968 Allow the use of tracepoints as software performance events.
970 When this is enabled, you can create perf events based on
971 tracepoints using PERF_TYPE_TRACEPOINT and the tracepoint ID
972 found in debugfs://tracing/events/*/*/id. (The -e/--events
973 option to the perf tool can parse and interpret symbolic
974 tracepoints, in the subsystem:tracepoint_name format.)
977 bool "Kernel performance counters (old config option)"
978 depends on HAVE_PERF_EVENTS
980 This config has been obsoleted by the PERF_EVENTS
981 config option - please see that one for details.
983 It has no effect on the kernel whether you enable
984 it or not, it is a compatibility placeholder.
988 config DEBUG_PERF_USE_VMALLOC
990 bool "Debug: use vmalloc to back perf mmap() buffers"
991 depends on PERF_EVENTS && DEBUG_KERNEL
992 select PERF_USE_VMALLOC
994 Use vmalloc memory to back perf mmap() buffers.
996 Mostly useful for debugging the vmalloc code on platforms
997 that don't require it.
1003 config VM_EVENT_COUNTERS
1005 bool "Enable VM event counters for /proc/vmstat" if EMBEDDED
1007 VM event counters are needed for event counts to be shown.
1008 This option allows the disabling of the VM event counters
1009 on EMBEDDED systems. /proc/vmstat will only show page counts
1010 if VM event counters are disabled.
1014 bool "Enable PCI quirk workarounds" if EMBEDDED
1017 This enables workarounds for various PCI chipset
1018 bugs/quirks. Disable this only if your target machine is
1019 unaffected by PCI quirks.
1023 bool "Enable SLUB debugging support" if EMBEDDED
1024 depends on SLUB && SYSFS
1026 SLUB has extensive debug support features. Disabling these can
1027 result in significant savings in code size. This also disables
1028 SLUB sysfs support. /sys/slab will not exist and there will be
1029 no support for cache validation etc.
1032 bool "Disable heap randomization"
1035 Randomizing heap placement makes heap exploits harder, but it
1036 also breaks ancient binaries (including anything libc5 based).
1037 This option changes the bootup default to heap randomization
1038 disabled, and can be overridden at runtime by setting
1039 /proc/sys/kernel/randomize_va_space to 2.
1041 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1044 prompt "Choose SLAB allocator"
1047 This option allows to select a slab allocator.
1052 The regular slab allocator that is established and known to work
1053 well in all environments. It organizes cache hot objects in
1054 per cpu and per node queues.
1057 depends on BROKEN || NUMA || !DISCONTIGMEM
1058 bool "SLUB (Unqueued Allocator)"
1060 SLUB is a slab allocator that minimizes cache line usage
1061 instead of managing queues of cached objects (SLAB approach).
1062 Per cpu caching is realized using slabs of objects instead
1063 of queues of objects. SLUB can use memory efficiently
1064 and has enhanced diagnostics. SLUB is the default choice for
1069 bool "SLOB (Simple Allocator)"
1071 SLOB replaces the stock allocator with a drastically simpler
1072 allocator. SLOB is generally more space efficient but
1073 does not perform as well on large systems.
1078 bool "Profiling support (EXPERIMENTAL)"
1080 Say Y here to enable the extended profiling support mechanisms used
1081 by profilers such as OProfile.
1084 # Place an empty function call at each tracepoint site. Can be
1085 # dynamically changed for a probe function.
1090 source "arch/Kconfig"
1096 The slow work thread pool provides a number of dynamically allocated
1097 threads that can be used by the kernel to perform operations that
1098 take a relatively long time.
1100 An example of this would be CacheFiles doing a path lookup followed
1101 by a series of mkdirs and a create call, all of which have to touch
1104 See Documentation/slow-work.txt.
1106 config SLOW_WORK_DEBUG
1107 bool "Slow work debugging through debugfs"
1109 depends on SLOW_WORK && DEBUG_FS
1111 Display the contents of the slow work run queue through debugfs,
1112 including items currently executing.
1114 See Documentation/slow-work.txt.
1116 endmenu # General setup
1118 config HAVE_GENERIC_DMA_COHERENT
1125 depends on SLAB || SLUB_DEBUG
1133 default 0 if BASE_FULL
1134 default 1 if !BASE_FULL
1137 bool "Enable loadable module support"
1139 Kernel modules are small pieces of compiled code which can
1140 be inserted in the running kernel, rather than being
1141 permanently built into the kernel. You use the "modprobe"
1142 tool to add (and sometimes remove) them. If you say Y here,
1143 many parts of the kernel can be built as modules (by
1144 answering M instead of Y where indicated): this is most
1145 useful for infrequently used options which are not required
1146 for booting. For more information, see the man pages for
1147 modprobe, lsmod, modinfo, insmod and rmmod.
1149 If you say Y here, you will need to run "make
1150 modules_install" to put the modules under /lib/modules/
1151 where modprobe can find them (you may need to be root to do
1158 config MODULE_FORCE_LOAD
1159 bool "Forced module loading"
1162 Allow loading of modules without version information (ie. modprobe
1163 --force). Forced module loading sets the 'F' (forced) taint flag and
1164 is usually a really bad idea.
1166 config MODULE_UNLOAD
1167 bool "Module unloading"
1169 Without this option you will not be able to unload any
1170 modules (note that some modules may not be unloadable
1171 anyway), which makes your kernel smaller, faster
1172 and simpler. If unsure, say Y.
1174 config MODULE_FORCE_UNLOAD
1175 bool "Forced module unloading"
1176 depends on MODULE_UNLOAD && EXPERIMENTAL
1178 This option allows you to force a module to unload, even if the
1179 kernel believes it is unsafe: the kernel will remove the module
1180 without waiting for anyone to stop using it (using the -f option to
1181 rmmod). This is mainly for kernel developers and desperate users.
1185 bool "Module versioning support"
1187 Usually, you have to use modules compiled with your kernel.
1188 Saying Y here makes it sometimes possible to use modules
1189 compiled for different kernels, by adding enough information
1190 to the modules to (hopefully) spot any changes which would
1191 make them incompatible with the kernel you are running. If
1194 config MODULE_SRCVERSION_ALL
1195 bool "Source checksum for all modules"
1197 Modules which contain a MODULE_VERSION get an extra "srcversion"
1198 field inserted into their modinfo section, which contains a
1199 sum of the source files which made it. This helps maintainers
1200 see exactly which source was used to build a module (since
1201 others sometimes change the module source without updating
1202 the version). With this option, such a "srcversion" field
1203 will be created for all modules. If unsure, say N.
1207 config INIT_ALL_POSSIBLE
1210 Back when each arch used to define their own cpu_online_map and
1211 cpu_possible_map, some of them chose to initialize cpu_possible_map
1212 with all 1s, and others with all 0s. When they were centralised,
1213 it was better to provide this option than to break all the archs
1214 and have several arch maintainers pursuing me down dark alleys.
1219 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU
1221 Need stop_machine() primitive.
1223 source "block/Kconfig"
1225 config PREEMPT_NOTIFIERS