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"
26 config BUILDTIME_EXTABLE_SORT
36 depends on BROKEN || !SMP
39 config INIT_ENV_ARG_LIMIT
44 Maximum of each of the number of arguments and environment
45 variables passed to init from the kernel command line.
49 string "Cross-compiler tool prefix"
51 Same as running 'make CROSS_COMPILE=prefix-' but stored for
52 default make runs in this kernel build directory. You don't
53 need to set this unless you want the configured kernel build
54 directory to select the cross-compiler automatically.
57 bool "Compile also drivers which will not load"
60 Some drivers can be compiled on a different platform than they are
61 intended to be run on. Despite they cannot be loaded there (or even
62 when they load they cannot be used due to missing HW support),
63 developers still, opposing to distributors, might want to build such
64 drivers to compile-test them.
66 If you are a developer and want to build everything available, say Y
67 here. If you are a user/distributor, say N here to exclude useless
68 drivers to be distributed.
71 string "Local version - append to kernel release"
73 Append an extra string to the end of your kernel version.
74 This will show up when you type uname, for example.
75 The string you set here will be appended after the contents of
76 any files with a filename matching localversion* in your
77 object and source tree, in that order. Your total string can
78 be a maximum of 64 characters.
80 config LOCALVERSION_AUTO
81 bool "Automatically append version information to the version string"
84 This will try to automatically determine if the current tree is a
85 release tree by looking for git tags that belong to the current
88 A string of the format -gxxxxxxxx will be added to the localversion
89 if a git-based tree is found. The string generated by this will be
90 appended after any matching localversion* files, and after the value
91 set in CONFIG_LOCALVERSION.
93 (The actual string used here is the first eight characters produced
94 by running the command:
96 $ git rev-parse --verify HEAD
98 which is done within the script "scripts/setlocalversion".)
100 config HAVE_KERNEL_GZIP
103 config HAVE_KERNEL_BZIP2
106 config HAVE_KERNEL_LZMA
109 config HAVE_KERNEL_XZ
112 config HAVE_KERNEL_LZO
115 config HAVE_KERNEL_LZ4
119 prompt "Kernel compression mode"
121 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4
123 The linux kernel is a kind of self-extracting executable.
124 Several compression algorithms are available, which differ
125 in efficiency, compression and decompression speed.
126 Compression speed is only relevant when building a kernel.
127 Decompression speed is relevant at each boot.
129 If you have any problems with bzip2 or lzma compressed
130 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
131 version of this functionality (bzip2 only), for 2.4, was
132 supplied by Christian Ludwig)
134 High compression options are mostly useful for users, who
135 are low on disk space (embedded systems), but for whom ram
138 If in doubt, select 'gzip'
142 depends on HAVE_KERNEL_GZIP
144 The old and tried gzip compression. It provides a good balance
145 between compression ratio and decompression speed.
149 depends on HAVE_KERNEL_BZIP2
151 Its compression ratio and speed is intermediate.
152 Decompression speed is slowest among the choices. The kernel
153 size is about 10% smaller with bzip2, in comparison to gzip.
154 Bzip2 uses a large amount of memory. For modern kernels you
155 will need at least 8MB RAM or more for booting.
159 depends on HAVE_KERNEL_LZMA
161 This compression algorithm's ratio is best. Decompression speed
162 is between gzip and bzip2. Compression is slowest.
163 The kernel size is about 33% smaller with LZMA in comparison to gzip.
167 depends on HAVE_KERNEL_XZ
169 XZ uses the LZMA2 algorithm and instruction set specific
170 BCJ filters which can improve compression ratio of executable
171 code. The size of the kernel is about 30% smaller with XZ in
172 comparison to gzip. On architectures for which there is a BCJ
173 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
174 will create a few percent smaller kernel than plain LZMA.
176 The speed is about the same as with LZMA: The decompression
177 speed of XZ is better than that of bzip2 but worse than gzip
178 and LZO. Compression is slow.
182 depends on HAVE_KERNEL_LZO
184 Its compression ratio is the poorest among the choices. The kernel
185 size is about 10% bigger than gzip; however its speed
186 (both compression and decompression) is the fastest.
190 depends on HAVE_KERNEL_LZ4
192 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
193 A preliminary version of LZ4 de/compression tool is available at
194 <https://code.google.com/p/lz4/>.
196 Its compression ratio is worse than LZO. The size of the kernel
197 is about 8% bigger than LZO. But the decompression speed is
202 config DEFAULT_HOSTNAME
203 string "Default hostname"
206 This option determines the default system hostname before userspace
207 calls sethostname(2). The kernel traditionally uses "(none)" here,
208 but you may wish to use a different default here to make a minimal
209 system more usable with less configuration.
212 bool "Support for paging of anonymous memory (swap)"
213 depends on MMU && BLOCK
216 This option allows you to choose whether you want to have support
217 for so called swap devices or swap files in your kernel that are
218 used to provide more virtual memory than the actual RAM present
219 in your computer. If unsure say Y.
224 Inter Process Communication is a suite of library functions and
225 system calls which let processes (running programs) synchronize and
226 exchange information. It is generally considered to be a good thing,
227 and some programs won't run unless you say Y here. In particular, if
228 you want to run the DOS emulator dosemu under Linux (read the
229 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
230 you'll need to say Y here.
232 You can find documentation about IPC with "info ipc" and also in
233 section 6.4 of the Linux Programmer's Guide, available from
234 <http://www.tldp.org/guides.html>.
236 config SYSVIPC_SYSCTL
243 bool "POSIX Message Queues"
246 POSIX variant of message queues is a part of IPC. In POSIX message
247 queues every message has a priority which decides about succession
248 of receiving it by a process. If you want to compile and run
249 programs written e.g. for Solaris with use of its POSIX message
250 queues (functions mq_*) say Y here.
252 POSIX message queues are visible as a filesystem called 'mqueue'
253 and can be mounted somewhere if you want to do filesystem
254 operations on message queues.
258 config POSIX_MQUEUE_SYSCTL
260 depends on POSIX_MQUEUE
264 config CROSS_MEMORY_ATTACH
265 bool "Enable process_vm_readv/writev syscalls"
269 Enabling this option adds the system calls process_vm_readv and
270 process_vm_writev which allow a process with the correct privileges
271 to directly read from or write to another process' address space.
272 See the man page for more details.
275 bool "open by fhandle syscalls"
278 If you say Y here, a user level program will be able to map
279 file names to handle and then later use the handle for
280 different file system operations. This is useful in implementing
281 userspace file servers, which now track files using handles instead
282 of names. The handle would remain the same even if file names
283 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
287 bool "uselib syscall"
288 def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION
290 This option enables the uselib syscall, a system call used in the
291 dynamic linker from libc5 and earlier. glibc does not use this
292 system call. If you intend to run programs built on libc5 or
293 earlier, you may need to enable this syscall. Current systems
294 running glibc can safely disable this.
297 bool "Auditing support"
300 Enable auditing infrastructure that can be used with another
301 kernel subsystem, such as SELinux (which requires this for
302 logging of avc messages output). System call auditing is included
303 on architectures which support it.
305 config HAVE_ARCH_AUDITSYSCALL
310 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL
314 depends on AUDITSYSCALL
319 depends on AUDITSYSCALL
322 source "kernel/irq/Kconfig"
323 source "kernel/time/Kconfig"
325 menu "CPU/Task time and stats accounting"
327 config VIRT_CPU_ACCOUNTING
331 prompt "Cputime accounting"
332 default TICK_CPU_ACCOUNTING if !PPC64
333 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64
335 # Kind of a stub config for the pure tick based cputime accounting
336 config TICK_CPU_ACCOUNTING
337 bool "Simple tick based cputime accounting"
338 depends on !S390 && !NO_HZ_FULL
340 This is the basic tick based cputime accounting that maintains
341 statistics about user, system and idle time spent on per jiffies
346 config VIRT_CPU_ACCOUNTING_NATIVE
347 bool "Deterministic task and CPU time accounting"
348 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL
349 select VIRT_CPU_ACCOUNTING
351 Select this option to enable more accurate task and CPU time
352 accounting. This is done by reading a CPU counter on each
353 kernel entry and exit and on transitions within the kernel
354 between system, softirq and hardirq state, so there is a
355 small performance impact. In the case of s390 or IBM POWER > 5,
356 this also enables accounting of stolen time on logically-partitioned
359 config VIRT_CPU_ACCOUNTING_GEN
360 bool "Full dynticks CPU time accounting"
361 depends on HAVE_CONTEXT_TRACKING
362 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN
363 select VIRT_CPU_ACCOUNTING
364 select CONTEXT_TRACKING
366 Select this option to enable task and CPU time accounting on full
367 dynticks systems. This accounting is implemented by watching every
368 kernel-user boundaries using the context tracking subsystem.
369 The accounting is thus performed at the expense of some significant
372 For now this is only useful if you are working on the full
373 dynticks subsystem development.
377 config IRQ_TIME_ACCOUNTING
378 bool "Fine granularity task level IRQ time accounting"
379 depends on HAVE_IRQ_TIME_ACCOUNTING && !NO_HZ_FULL
381 Select this option to enable fine granularity task irq time
382 accounting. This is done by reading a timestamp on each
383 transitions between softirq and hardirq state, so there can be a
384 small performance impact.
386 If in doubt, say N here.
390 config BSD_PROCESS_ACCT
391 bool "BSD Process Accounting"
394 If you say Y here, a user level program will be able to instruct the
395 kernel (via a special system call) to write process accounting
396 information to a file: whenever a process exits, information about
397 that process will be appended to the file by the kernel. The
398 information includes things such as creation time, owning user,
399 command name, memory usage, controlling terminal etc. (the complete
400 list is in the struct acct in <file:include/linux/acct.h>). It is
401 up to the user level program to do useful things with this
402 information. This is generally a good idea, so say Y.
404 config BSD_PROCESS_ACCT_V3
405 bool "BSD Process Accounting version 3 file format"
406 depends on BSD_PROCESS_ACCT
409 If you say Y here, the process accounting information is written
410 in a new file format that also logs the process IDs of each
411 process and it's parent. Note that this file format is incompatible
412 with previous v0/v1/v2 file formats, so you will need updated tools
413 for processing it. A preliminary version of these tools is available
414 at <http://www.gnu.org/software/acct/>.
417 bool "Export task/process statistics through netlink"
422 Export selected statistics for tasks/processes through the
423 generic netlink interface. Unlike BSD process accounting, the
424 statistics are available during the lifetime of tasks/processes as
425 responses to commands. Like BSD accounting, they are sent to user
430 config TASK_DELAY_ACCT
431 bool "Enable per-task delay accounting"
435 Collect information on time spent by a task waiting for system
436 resources like cpu, synchronous block I/O completion and swapping
437 in pages. Such statistics can help in setting a task's priorities
438 relative to other tasks for cpu, io, rss limits etc.
443 bool "Enable extended accounting over taskstats"
446 Collect extended task accounting data and send the data
447 to userland for processing over the taskstats interface.
451 config TASK_IO_ACCOUNTING
452 bool "Enable per-task storage I/O accounting"
453 depends on TASK_XACCT
455 Collect information on the number of bytes of storage I/O which this
460 endmenu # "CPU/Task time and stats accounting"
466 default y if !PREEMPT && SMP
468 This option selects the RCU implementation that is
469 designed for very large SMP system with hundreds or
470 thousands of CPUs. It also scales down nicely to
477 This option selects the RCU implementation that is
478 designed for very large SMP systems with hundreds or
479 thousands of CPUs, but for which real-time response
480 is also required. It also scales down nicely to
483 Select this option if you are unsure.
487 default y if !PREEMPT && !SMP
489 This option selects the RCU implementation that is
490 designed for UP systems from which real-time response
491 is not required. This option greatly reduces the
492 memory footprint of RCU.
495 bool "Make expert-level adjustments to RCU configuration"
498 This option needs to be enabled if you wish to make
499 expert-level adjustments to RCU configuration. By default,
500 no such adjustments can be made, which has the often-beneficial
501 side-effect of preventing "make oldconfig" from asking you all
502 sorts of detailed questions about how you would like numerous
503 obscure RCU options to be set up.
505 Say Y if you need to make expert-level adjustments to RCU.
507 Say N if you are unsure.
512 This option selects the sleepable version of RCU. This version
513 permits arbitrary sleeping or blocking within RCU read-side critical
521 This option enables a task-based RCU implementation that uses
522 only voluntary context switch (not preemption!), idle, and
523 user-mode execution as quiescent states.
525 config RCU_STALL_COMMON
526 def_bool ( TREE_RCU || PREEMPT_RCU || RCU_TRACE )
528 This option enables RCU CPU stall code that is common between
529 the TINY and TREE variants of RCU. The purpose is to allow
530 the tiny variants to disable RCU CPU stall warnings, while
531 making these warnings mandatory for the tree variants.
533 config CONTEXT_TRACKING
536 config CONTEXT_TRACKING_FORCE
537 bool "Force context tracking"
538 depends on CONTEXT_TRACKING
539 default y if !NO_HZ_FULL
541 The major pre-requirement for full dynticks to work is to
542 support the context tracking subsystem. But there are also
543 other dependencies to provide in order to make the full
546 This option stands for testing when an arch implements the
547 context tracking backend but doesn't yet fullfill all the
548 requirements to make the full dynticks feature working.
549 Without the full dynticks, there is no way to test the support
550 for context tracking and the subsystems that rely on it: RCU
551 userspace extended quiescent state and tickless cputime
552 accounting. This option copes with the absence of the full
553 dynticks subsystem by forcing the context tracking on all
556 Say Y only if you're working on the development of an
557 architecture backend for the context tracking.
559 Say N otherwise, this option brings an overhead that you
560 don't want in production.
564 int "Tree-based hierarchical RCU fanout value"
567 depends on (TREE_RCU || PREEMPT_RCU) && RCU_EXPERT
571 This option controls the fanout of hierarchical implementations
572 of RCU, allowing RCU to work efficiently on machines with
573 large numbers of CPUs. This value must be at least the fourth
574 root of NR_CPUS, which allows NR_CPUS to be insanely large.
575 The default value of RCU_FANOUT should be used for production
576 systems, but if you are stress-testing the RCU implementation
577 itself, small RCU_FANOUT values allow you to test large-system
578 code paths on small(er) systems.
580 Select a specific number if testing RCU itself.
581 Take the default if unsure.
583 config RCU_FANOUT_LEAF
584 int "Tree-based hierarchical RCU leaf-level fanout value"
587 depends on (TREE_RCU || PREEMPT_RCU) && RCU_EXPERT
590 This option controls the leaf-level fanout of hierarchical
591 implementations of RCU, and allows trading off cache misses
592 against lock contention. Systems that synchronize their
593 scheduling-clock interrupts for energy-efficiency reasons will
594 want the default because the smaller leaf-level fanout keeps
595 lock contention levels acceptably low. Very large systems
596 (hundreds or thousands of CPUs) will instead want to set this
597 value to the maximum value possible in order to reduce the
598 number of cache misses incurred during RCU's grace-period
599 initialization. These systems tend to run CPU-bound, and thus
600 are not helped by synchronized interrupts, and thus tend to
601 skew them, which reduces lock contention enough that large
602 leaf-level fanouts work well.
604 Select a specific number if testing RCU itself.
606 Select the maximum permissible value for large systems.
608 Take the default if unsure.
610 config RCU_FAST_NO_HZ
611 bool "Accelerate last non-dyntick-idle CPU's grace periods"
612 depends on NO_HZ_COMMON && SMP && RCU_EXPERT
615 This option permits CPUs to enter dynticks-idle state even if
616 they have RCU callbacks queued, and prevents RCU from waking
617 these CPUs up more than roughly once every four jiffies (by
618 default, you can adjust this using the rcutree.rcu_idle_gp_delay
619 parameter), thus improving energy efficiency. On the other
620 hand, this option increases the duration of RCU grace periods,
621 for example, slowing down synchronize_rcu().
623 Say Y if energy efficiency is critically important, and you
624 don't care about increased grace-period durations.
626 Say N if you are unsure.
628 config TREE_RCU_TRACE
629 def_bool RCU_TRACE && ( TREE_RCU || PREEMPT_RCU )
632 This option provides tracing for the TREE_RCU and
633 PREEMPT_RCU implementations, permitting Makefile to
634 trivially select kernel/rcutree_trace.c.
637 bool "Enable RCU priority boosting"
638 depends on RT_MUTEXES && PREEMPT_RCU && RCU_EXPERT
641 This option boosts the priority of preempted RCU readers that
642 block the current preemptible RCU grace period for too long.
643 This option also prevents heavy loads from blocking RCU
644 callback invocation for all flavors of RCU.
646 Say Y here if you are working with real-time apps or heavy loads
647 Say N here if you are unsure.
649 config RCU_KTHREAD_PRIO
650 int "Real-time priority to use for RCU worker threads"
651 range 1 99 if RCU_BOOST
652 range 0 99 if !RCU_BOOST
653 default 1 if RCU_BOOST
654 default 0 if !RCU_BOOST
655 depends on RCU_EXPERT
657 This option specifies the SCHED_FIFO priority value that will be
658 assigned to the rcuc/n and rcub/n threads and is also the value
659 used for RCU_BOOST (if enabled). If you are working with a
660 real-time application that has one or more CPU-bound threads
661 running at a real-time priority level, you should set
662 RCU_KTHREAD_PRIO to a priority higher than the highest-priority
663 real-time CPU-bound application thread. The default RCU_KTHREAD_PRIO
664 value of 1 is appropriate in the common case, which is real-time
665 applications that do not have any CPU-bound threads.
667 Some real-time applications might not have a single real-time
668 thread that saturates a given CPU, but instead might have
669 multiple real-time threads that, taken together, fully utilize
670 that CPU. In this case, you should set RCU_KTHREAD_PRIO to
671 a priority higher than the lowest-priority thread that is
672 conspiring to prevent the CPU from running any non-real-time
673 tasks. For example, if one thread at priority 10 and another
674 thread at priority 5 are between themselves fully consuming
675 the CPU time on a given CPU, then RCU_KTHREAD_PRIO should be
676 set to priority 6 or higher.
678 Specify the real-time priority, or take the default if unsure.
680 config RCU_BOOST_DELAY
681 int "Milliseconds to delay boosting after RCU grace-period start"
686 This option specifies the time to wait after the beginning of
687 a given grace period before priority-boosting preempted RCU
688 readers blocking that grace period. Note that any RCU reader
689 blocking an expedited RCU grace period is boosted immediately.
691 Accept the default if unsure.
694 bool "Offload RCU callback processing from boot-selected CPUs"
695 depends on TREE_RCU || PREEMPT_RCU
696 depends on RCU_EXPERT || NO_HZ_FULL
699 Use this option to reduce OS jitter for aggressive HPC or
700 real-time workloads. It can also be used to offload RCU
701 callback invocation to energy-efficient CPUs in battery-powered
702 asymmetric multiprocessors.
704 This option offloads callback invocation from the set of
705 CPUs specified at boot time by the rcu_nocbs parameter.
706 For each such CPU, a kthread ("rcuox/N") will be created to
707 invoke callbacks, where the "N" is the CPU being offloaded,
708 and where the "x" is "b" for RCU-bh, "p" for RCU-preempt, and
709 "s" for RCU-sched. Nothing prevents this kthread from running
710 on the specified CPUs, but (1) the kthreads may be preempted
711 between each callback, and (2) affinity or cgroups can be used
712 to force the kthreads to run on whatever set of CPUs is desired.
714 Say Y here if you want to help to debug reduced OS jitter.
715 Say N here if you are unsure.
718 prompt "Build-forced no-CBs CPUs"
719 default RCU_NOCB_CPU_NONE
720 depends on RCU_NOCB_CPU
722 This option allows no-CBs CPUs (whose RCU callbacks are invoked
723 from kthreads rather than from softirq context) to be specified
724 at build time. Additional no-CBs CPUs may be specified by
725 the rcu_nocbs= boot parameter.
727 config RCU_NOCB_CPU_NONE
728 bool "No build_forced no-CBs CPUs"
730 This option does not force any of the CPUs to be no-CBs CPUs.
731 Only CPUs designated by the rcu_nocbs= boot parameter will be
732 no-CBs CPUs, whose RCU callbacks will be invoked by per-CPU
733 kthreads whose names begin with "rcuo". All other CPUs will
734 invoke their own RCU callbacks in softirq context.
736 Select this option if you want to choose no-CBs CPUs at
737 boot time, for example, to allow testing of different no-CBs
738 configurations without having to rebuild the kernel each time.
740 config RCU_NOCB_CPU_ZERO
741 bool "CPU 0 is a build_forced no-CBs CPU"
743 This option forces CPU 0 to be a no-CBs CPU, so that its RCU
744 callbacks are invoked by a per-CPU kthread whose name begins
745 with "rcuo". Additional CPUs may be designated as no-CBs
746 CPUs using the rcu_nocbs= boot parameter will be no-CBs CPUs.
747 All other CPUs will invoke their own RCU callbacks in softirq
750 Select this if CPU 0 needs to be a no-CBs CPU for real-time
751 or energy-efficiency reasons, but the real reason it exists
752 is to ensure that randconfig testing covers mixed systems.
754 config RCU_NOCB_CPU_ALL
755 bool "All CPUs are build_forced no-CBs CPUs"
757 This option forces all CPUs to be no-CBs CPUs. The rcu_nocbs=
758 boot parameter will be ignored. All CPUs' RCU callbacks will
759 be executed in the context of per-CPU rcuo kthreads created for
760 this purpose. Assuming that the kthreads whose names start with
761 "rcuo" are bound to "housekeeping" CPUs, this reduces OS jitter
762 on the remaining CPUs, but might decrease memory locality during
763 RCU-callback invocation, thus potentially degrading throughput.
765 Select this if all CPUs need to be no-CBs CPUs for real-time
766 or energy-efficiency reasons.
770 config RCU_EXPEDITE_BOOT
774 This option enables expedited grace periods at boot time,
775 as if rcu_expedite_gp() had been invoked early in boot.
776 The corresponding rcu_unexpedite_gp() is invoked from
777 rcu_end_inkernel_boot(), which is intended to be invoked
778 at the end of the kernel-only boot sequence, just before
781 Accept the default if unsure.
783 endmenu # "RCU Subsystem"
790 tristate "Kernel .config support"
793 This option enables the complete Linux kernel ".config" file
794 contents to be saved in the kernel. It provides documentation
795 of which kernel options are used in a running kernel or in an
796 on-disk kernel. This information can be extracted from the kernel
797 image file with the script scripts/extract-ikconfig and used as
798 input to rebuild the current kernel or to build another kernel.
799 It can also be extracted from a running kernel by reading
800 /proc/config.gz if enabled (below).
803 bool "Enable access to .config through /proc/config.gz"
804 depends on IKCONFIG && PROC_FS
806 This option enables access to the kernel configuration file
807 through /proc/config.gz.
810 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
815 Select the minimal kernel log buffer size as a power of 2.
816 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config
817 parameter, see below. Any higher size also might be forced
818 by "log_buf_len" boot parameter.
828 config LOG_CPU_MAX_BUF_SHIFT
829 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)"
832 default 12 if !BASE_SMALL
833 default 0 if BASE_SMALL
836 This option allows to increase the default ring buffer size
837 according to the number of CPUs. The value defines the contribution
838 of each CPU as a power of 2. The used space is typically only few
839 lines however it might be much more when problems are reported,
842 The increased size means that a new buffer has to be allocated and
843 the original static one is unused. It makes sense only on systems
844 with more CPUs. Therefore this value is used only when the sum of
845 contributions is greater than the half of the default kernel ring
846 buffer as defined by LOG_BUF_SHIFT. The default values are set
847 so that more than 64 CPUs are needed to trigger the allocation.
849 Also this option is ignored when "log_buf_len" kernel parameter is
850 used as it forces an exact (power of two) size of the ring buffer.
852 The number of possible CPUs is used for this computation ignoring
853 hotplugging making the compuation optimal for the the worst case
854 scenerio while allowing a simple algorithm to be used from bootup.
856 Examples shift values and their meaning:
857 17 => 128 KB for each CPU
858 16 => 64 KB for each CPU
859 15 => 32 KB for each CPU
860 14 => 16 KB for each CPU
861 13 => 8 KB for each CPU
862 12 => 4 KB for each CPU
865 # Architectures with an unreliable sched_clock() should select this:
867 config HAVE_UNSTABLE_SCHED_CLOCK
870 config GENERIC_SCHED_CLOCK
874 # For architectures that want to enable the support for NUMA-affine scheduler
877 config ARCH_SUPPORTS_NUMA_BALANCING
881 # For architectures that prefer to flush all TLBs after a number of pages
882 # are unmapped instead of sending one IPI per page to flush. The architecture
883 # must provide guarantees on what happens if a clean TLB cache entry is
884 # written after the unmap. Details are in mm/rmap.c near the check for
885 # should_defer_flush. The architecture should also consider if the full flush
886 # and the refill costs are offset by the savings of sending fewer IPIs.
887 config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
891 # For architectures that know their GCC __int128 support is sound
893 config ARCH_SUPPORTS_INT128
896 # For architectures that (ab)use NUMA to represent different memory regions
897 # all cpu-local but of different latencies, such as SuperH.
899 config ARCH_WANT_NUMA_VARIABLE_LOCALITY
902 config NUMA_BALANCING
903 bool "Memory placement aware NUMA scheduler"
904 depends on ARCH_SUPPORTS_NUMA_BALANCING
905 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
906 depends on SMP && NUMA && MIGRATION
908 This option adds support for automatic NUMA aware memory/task placement.
909 The mechanism is quite primitive and is based on migrating memory when
910 it has references to the node the task is running on.
912 This system will be inactive on UMA systems.
914 config NUMA_BALANCING_DEFAULT_ENABLED
915 bool "Automatically enable NUMA aware memory/task placement"
917 depends on NUMA_BALANCING
919 If set, automatic NUMA balancing will be enabled if running on a NUMA
923 bool "Control Group support"
926 This option adds support for grouping sets of processes together, for
927 use with process control subsystems such as Cpusets, CFS, memory
928 controls or device isolation.
930 - Documentation/scheduler/sched-design-CFS.txt (CFS)
931 - Documentation/cgroups/ (features for grouping, isolation
932 and resource control)
942 bool "Memory controller"
946 Provides control over the memory footprint of tasks in a cgroup.
949 bool "Swap controller"
950 depends on MEMCG && SWAP
952 Provides control over the swap space consumed by tasks in a cgroup.
954 config MEMCG_SWAP_ENABLED
955 bool "Swap controller enabled by default"
956 depends on MEMCG_SWAP
959 Memory Resource Controller Swap Extension comes with its price in
960 a bigger memory consumption. General purpose distribution kernels
961 which want to enable the feature but keep it disabled by default
962 and let the user enable it by swapaccount=1 boot command line
963 parameter should have this option unselected.
964 For those who want to have the feature enabled by default should
965 select this option (if, for some reason, they need to disable it
966 then swapaccount=0 does the trick).
973 Generic block IO controller cgroup interface. This is the common
974 cgroup interface which should be used by various IO controlling
977 Currently, CFQ IO scheduler uses it to recognize task groups and
978 control disk bandwidth allocation (proportional time slice allocation)
979 to such task groups. It is also used by bio throttling logic in
980 block layer to implement upper limit in IO rates on a device.
982 This option only enables generic Block IO controller infrastructure.
983 One needs to also enable actual IO controlling logic/policy. For
984 enabling proportional weight division of disk bandwidth in CFQ, set
985 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
986 CONFIG_BLK_DEV_THROTTLING=y.
988 See Documentation/cgroups/blkio-controller.txt for more information.
990 config DEBUG_BLK_CGROUP
991 bool "IO controller debugging"
992 depends on BLK_CGROUP
995 Enable some debugging help. Currently it exports additional stat
996 files in a cgroup which can be useful for debugging.
998 config CGROUP_WRITEBACK
1000 depends on MEMCG && BLK_CGROUP
1003 menuconfig CGROUP_SCHED
1004 bool "CPU controller"
1007 This feature lets CPU scheduler recognize task groups and control CPU
1008 bandwidth allocation to such task groups. It uses cgroups to group
1012 config FAIR_GROUP_SCHED
1013 bool "Group scheduling for SCHED_OTHER"
1014 depends on CGROUP_SCHED
1015 default CGROUP_SCHED
1017 config CFS_BANDWIDTH
1018 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
1019 depends on FAIR_GROUP_SCHED
1022 This option allows users to define CPU bandwidth rates (limits) for
1023 tasks running within the fair group scheduler. Groups with no limit
1024 set are considered to be unconstrained and will run with no
1026 See tip/Documentation/scheduler/sched-bwc.txt for more information.
1028 config RT_GROUP_SCHED
1029 bool "Group scheduling for SCHED_RR/FIFO"
1030 depends on CGROUP_SCHED
1033 This feature lets you explicitly allocate real CPU bandwidth
1034 to task groups. If enabled, it will also make it impossible to
1035 schedule realtime tasks for non-root users until you allocate
1036 realtime bandwidth for them.
1037 See Documentation/scheduler/sched-rt-group.txt for more information.
1042 bool "PIDs controller"
1044 Provides enforcement of process number limits in the scope of a
1045 cgroup. Any attempt to fork more processes than is allowed in the
1046 cgroup will fail. PIDs are fundamentally a global resource because it
1047 is fairly trivial to reach PID exhaustion before you reach even a
1048 conservative kmemcg limit. As a result, it is possible to grind a
1049 system to halt without being limited by other cgroup policies. The
1050 PIDs controller is designed to stop this from happening.
1052 It should be noted that organisational operations (such as attaching
1053 to a cgroup hierarchy will *not* be blocked by the PIDs controller),
1054 since the PIDs limit only affects a process's ability to fork, not to
1057 config CGROUP_FREEZER
1058 bool "Freezer controller"
1060 Provides a way to freeze and unfreeze all tasks in a
1063 This option affects the ORIGINAL cgroup interface. The cgroup2 memory
1064 controller includes important in-kernel memory consumers per default.
1066 If you're using cgroup2, say N.
1068 config CGROUP_HUGETLB
1069 bool "HugeTLB controller"
1070 depends on HUGETLB_PAGE
1074 Provides a cgroup controller for HugeTLB pages.
1075 When you enable this, you can put a per cgroup limit on HugeTLB usage.
1076 The limit is enforced during page fault. Since HugeTLB doesn't
1077 support page reclaim, enforcing the limit at page fault time implies
1078 that, the application will get SIGBUS signal if it tries to access
1079 HugeTLB pages beyond its limit. This requires the application to know
1080 beforehand how much HugeTLB pages it would require for its use. The
1081 control group is tracked in the third page lru pointer. This means
1082 that we cannot use the controller with huge page less than 3 pages.
1085 bool "Cpuset controller"
1087 This option will let you create and manage CPUSETs which
1088 allow dynamically partitioning a system into sets of CPUs and
1089 Memory Nodes and assigning tasks to run only within those sets.
1090 This is primarily useful on large SMP or NUMA systems.
1094 config PROC_PID_CPUSET
1095 bool "Include legacy /proc/<pid>/cpuset file"
1099 config CGROUP_DEVICE
1100 bool "Device controller"
1102 Provides a cgroup controller implementing whitelists for
1103 devices which a process in the cgroup can mknod or open.
1105 config CGROUP_CPUACCT
1106 bool "Simple CPU accounting controller"
1108 Provides a simple controller for monitoring the
1109 total CPU consumed by the tasks in a cgroup.
1112 bool "Perf controller"
1113 depends on PERF_EVENTS
1115 This option extends the perf per-cpu mode to restrict monitoring
1116 to threads which belong to the cgroup specified and run on the
1122 bool "Example controller"
1125 This option enables a simple controller that exports
1126 debugging information about the cgroups framework.
1132 config CHECKPOINT_RESTORE
1133 bool "Checkpoint/restore support" if EXPERT
1134 select PROC_CHILDREN
1137 Enables additional kernel features in a sake of checkpoint/restore.
1138 In particular it adds auxiliary prctl codes to setup process text,
1139 data and heap segment sizes, and a few additional /proc filesystem
1142 If unsure, say N here.
1144 menuconfig NAMESPACES
1145 bool "Namespaces support" if EXPERT
1146 depends on MULTIUSER
1149 Provides the way to make tasks work with different objects using
1150 the same id. For example same IPC id may refer to different objects
1151 or same user id or pid may refer to different tasks when used in
1152 different namespaces.
1157 bool "UTS namespace"
1160 In this namespace tasks see different info provided with the
1164 bool "IPC namespace"
1165 depends on (SYSVIPC || POSIX_MQUEUE)
1168 In this namespace tasks work with IPC ids which correspond to
1169 different IPC objects in different namespaces.
1172 bool "User namespace"
1175 This allows containers, i.e. vservers, to use user namespaces
1176 to provide different user info for different servers.
1178 When user namespaces are enabled in the kernel it is
1179 recommended that the MEMCG option also be enabled and that
1180 user-space use the memory control groups to limit the amount
1181 of memory a memory unprivileged users can use.
1186 bool "PID Namespaces"
1189 Support process id namespaces. This allows having multiple
1190 processes with the same pid as long as they are in different
1191 pid namespaces. This is a building block of containers.
1194 bool "Network namespace"
1198 Allow user space to create what appear to be multiple instances
1199 of the network stack.
1203 config SCHED_AUTOGROUP
1204 bool "Automatic process group scheduling"
1207 select FAIR_GROUP_SCHED
1209 This option optimizes the scheduler for common desktop workloads by
1210 automatically creating and populating task groups. This separation
1211 of workloads isolates aggressive CPU burners (like build jobs) from
1212 desktop applications. Task group autogeneration is currently based
1215 config SYSFS_DEPRECATED
1216 bool "Enable deprecated sysfs features to support old userspace tools"
1220 This option adds code that switches the layout of the "block" class
1221 devices, to not show up in /sys/class/block/, but only in
1224 This switch is only active when the sysfs.deprecated=1 boot option is
1225 passed or the SYSFS_DEPRECATED_V2 option is set.
1227 This option allows new kernels to run on old distributions and tools,
1228 which might get confused by /sys/class/block/. Since 2007/2008 all
1229 major distributions and tools handle this just fine.
1231 Recent distributions and userspace tools after 2009/2010 depend on
1232 the existence of /sys/class/block/, and will not work with this
1235 Only if you are using a new kernel on an old distribution, you might
1238 config SYSFS_DEPRECATED_V2
1239 bool "Enable deprecated sysfs features by default"
1242 depends on SYSFS_DEPRECATED
1244 Enable deprecated sysfs by default.
1246 See the CONFIG_SYSFS_DEPRECATED option for more details about this
1249 Only if you are using a new kernel on an old distribution, you might
1250 need to say Y here. Even then, odds are you would not need it
1251 enabled, you can always pass the boot option if absolutely necessary.
1254 bool "Kernel->user space relay support (formerly relayfs)"
1256 This option enables support for relay interface support in
1257 certain file systems (such as debugfs).
1258 It is designed to provide an efficient mechanism for tools and
1259 facilities to relay large amounts of data from kernel space to
1264 config BLK_DEV_INITRD
1265 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
1266 depends on BROKEN || !FRV
1268 The initial RAM filesystem is a ramfs which is loaded by the
1269 boot loader (loadlin or lilo) and that is mounted as root
1270 before the normal boot procedure. It is typically used to
1271 load modules needed to mount the "real" root file system,
1272 etc. See <file:Documentation/initrd.txt> for details.
1274 If RAM disk support (BLK_DEV_RAM) is also included, this
1275 also enables initial RAM disk (initrd) support and adds
1276 15 Kbytes (more on some other architectures) to the kernel size.
1282 source "usr/Kconfig"
1286 config CC_OPTIMIZE_FOR_SIZE
1287 bool "Optimize for size"
1289 Enabling this option will pass "-Os" instead of "-O2" to
1290 your compiler resulting in a smaller kernel.
1303 config SYSCTL_EXCEPTION_TRACE
1306 Enable support for /proc/sys/debug/exception-trace.
1308 config SYSCTL_ARCH_UNALIGN_NO_WARN
1311 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1312 Allows arch to define/use @no_unaligned_warning to possibly warn
1313 about unaligned access emulation going on under the hood.
1315 config SYSCTL_ARCH_UNALIGN_ALLOW
1318 Enable support for /proc/sys/kernel/unaligned-trap
1319 Allows arches to define/use @unaligned_enabled to runtime toggle
1320 the unaligned access emulation.
1321 see arch/parisc/kernel/unaligned.c for reference
1323 config HAVE_PCSPKR_PLATFORM
1326 # interpreter that classic socket filters depend on
1331 bool "Configure standard kernel features (expert users)"
1332 # Unhide debug options, to make the on-by-default options visible
1335 This option allows certain base kernel options and settings
1336 to be disabled or tweaked. This is for specialized
1337 environments which can tolerate a "non-standard" kernel.
1338 Only use this if you really know what you are doing.
1341 bool "Enable 16-bit UID system calls" if EXPERT
1342 depends on HAVE_UID16 && MULTIUSER
1345 This enables the legacy 16-bit UID syscall wrappers.
1348 bool "Multiple users, groups and capabilities support" if EXPERT
1351 This option enables support for non-root users, groups and
1354 If you say N here, all processes will run with UID 0, GID 0, and all
1355 possible capabilities. Saying N here also compiles out support for
1356 system calls related to UIDs, GIDs, and capabilities, such as setuid,
1359 If unsure, say Y here.
1361 config SGETMASK_SYSCALL
1362 bool "sgetmask/ssetmask syscalls support" if EXPERT
1363 def_bool PARISC || MN10300 || BLACKFIN || M68K || PPC || MIPS || X86 || SPARC || CRIS || MICROBLAZE || SUPERH
1365 sys_sgetmask and sys_ssetmask are obsolete system calls
1366 no longer supported in libc but still enabled by default in some
1369 If unsure, leave the default option here.
1371 config SYSFS_SYSCALL
1372 bool "Sysfs syscall support" if EXPERT
1375 sys_sysfs is an obsolete system call no longer supported in libc.
1376 Note that disabling this option is more secure but might break
1377 compatibility with some systems.
1379 If unsure say Y here.
1381 config SYSCTL_SYSCALL
1382 bool "Sysctl syscall support" if EXPERT
1383 depends on PROC_SYSCTL
1387 sys_sysctl uses binary paths that have been found challenging
1388 to properly maintain and use. The interface in /proc/sys
1389 using paths with ascii names is now the primary path to this
1392 Almost nothing using the binary sysctl interface so if you are
1393 trying to save some space it is probably safe to disable this,
1394 making your kernel marginally smaller.
1396 If unsure say N here.
1399 bool "Load all symbols for debugging/ksymoops" if EXPERT
1402 Say Y here to let the kernel print out symbolic crash information and
1403 symbolic stack backtraces. This increases the size of the kernel
1404 somewhat, as all symbols have to be loaded into the kernel image.
1407 bool "Include all symbols in kallsyms"
1408 depends on DEBUG_KERNEL && KALLSYMS
1410 Normally kallsyms only contains the symbols of functions for nicer
1411 OOPS messages and backtraces (i.e., symbols from the text and inittext
1412 sections). This is sufficient for most cases. And only in very rare
1413 cases (e.g., when a debugger is used) all symbols are required (e.g.,
1414 names of variables from the data sections, etc).
1416 This option makes sure that all symbols are loaded into the kernel
1417 image (i.e., symbols from all sections) in cost of increased kernel
1418 size (depending on the kernel configuration, it may be 300KiB or
1419 something like this).
1421 Say N unless you really need all symbols.
1423 config KALLSYMS_ABSOLUTE_PERCPU
1425 default X86_64 && SMP
1427 config KALLSYMS_BASE_RELATIVE
1430 default !IA64 && !(TILE && 64BIT)
1432 Instead of emitting them as absolute values in the native word size,
1433 emit the symbol references in the kallsyms table as 32-bit entries,
1434 each containing a relative value in the range [base, base + U32_MAX]
1435 or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either
1436 an absolute value in the range [0, S32_MAX] or a relative value in the
1437 range [base, base + S32_MAX], where base is the lowest relative symbol
1438 address encountered in the image.
1440 On 64-bit builds, this reduces the size of the address table by 50%,
1441 but more importantly, it results in entries whose values are build
1442 time constants, and no relocation pass is required at runtime to fix
1443 up the entries based on the runtime load address of the kernel.
1447 bool "Enable support for printk" if EXPERT
1450 This option enables normal printk support. Removing it
1451 eliminates most of the message strings from the kernel image
1452 and makes the kernel more or less silent. As this makes it
1453 very difficult to diagnose system problems, saying N here is
1454 strongly discouraged.
1457 bool "BUG() support" if EXPERT
1460 Disabling this option eliminates support for BUG and WARN, reducing
1461 the size of your kernel image and potentially quietly ignoring
1462 numerous fatal conditions. You should only consider disabling this
1463 option for embedded systems with no facilities for reporting errors.
1469 bool "Enable ELF core dumps" if EXPERT
1471 Enable support for generating core dumps. Disabling saves about 4k.
1474 config PCSPKR_PLATFORM
1475 bool "Enable PC-Speaker support" if EXPERT
1476 depends on HAVE_PCSPKR_PLATFORM
1480 This option allows to disable the internal PC-Speaker
1481 support, saving some memory.
1485 bool "Enable full-sized data structures for core" if EXPERT
1487 Disabling this option reduces the size of miscellaneous core
1488 kernel data structures. This saves memory on small machines,
1489 but may reduce performance.
1492 bool "Enable futex support" if EXPERT
1496 Disabling this option will cause the kernel to be built without
1497 support for "fast userspace mutexes". The resulting kernel may not
1498 run glibc-based applications correctly.
1500 config HAVE_FUTEX_CMPXCHG
1504 Architectures should select this if futex_atomic_cmpxchg_inatomic()
1505 is implemented and always working. This removes a couple of runtime
1509 bool "Enable eventpoll support" if EXPERT
1513 Disabling this option will cause the kernel to be built without
1514 support for epoll family of system calls.
1517 bool "Enable signalfd() system call" if EXPERT
1521 Enable the signalfd() system call that allows to receive signals
1522 on a file descriptor.
1527 bool "Enable timerfd() system call" if EXPERT
1531 Enable the timerfd() system call that allows to receive timer
1532 events on a file descriptor.
1537 bool "Enable eventfd() system call" if EXPERT
1541 Enable the eventfd() system call that allows to receive both
1542 kernel notification (ie. KAIO) or userspace notifications.
1546 # syscall, maps, verifier
1548 bool "Enable bpf() system call"
1553 Enable the bpf() system call that allows to manipulate eBPF
1554 programs and maps via file descriptors.
1557 bool "Use full shmem filesystem" if EXPERT
1561 The shmem is an internal filesystem used to manage shared memory.
1562 It is backed by swap and manages resource limits. It is also exported
1563 to userspace as tmpfs if TMPFS is enabled. Disabling this
1564 option replaces shmem and tmpfs with the much simpler ramfs code,
1565 which may be appropriate on small systems without swap.
1568 bool "Enable AIO support" if EXPERT
1571 This option enables POSIX asynchronous I/O which may by used
1572 by some high performance threaded applications. Disabling
1573 this option saves about 7k.
1575 config ADVISE_SYSCALLS
1576 bool "Enable madvise/fadvise syscalls" if EXPERT
1579 This option enables the madvise and fadvise syscalls, used by
1580 applications to advise the kernel about their future memory or file
1581 usage, improving performance. If building an embedded system where no
1582 applications use these syscalls, you can disable this option to save
1586 bool "Enable userfaultfd() system call"
1590 Enable the userfaultfd() system call that allows to intercept and
1591 handle page faults in userland.
1595 bool "Enable PCI quirk workarounds" if EXPERT
1598 This enables workarounds for various PCI chipset
1599 bugs/quirks. Disable this only if your target machine is
1600 unaffected by PCI quirks.
1603 bool "Enable membarrier() system call" if EXPERT
1606 Enable the membarrier() system call that allows issuing memory
1607 barriers across all running threads, which can be used to distribute
1608 the cost of user-space memory barriers asymmetrically by transforming
1609 pairs of memory barriers into pairs consisting of membarrier() and a
1615 bool "Embedded system"
1616 option allnoconfig_y
1619 This option should be enabled if compiling the kernel for
1620 an embedded system so certain expert options are available
1623 config HAVE_PERF_EVENTS
1626 See tools/perf/design.txt for details.
1628 config PERF_USE_VMALLOC
1631 See tools/perf/design.txt for details
1633 menu "Kernel Performance Events And Counters"
1636 bool "Kernel performance events and counters"
1637 default y if PROFILING
1638 depends on HAVE_PERF_EVENTS
1643 Enable kernel support for various performance events provided
1644 by software and hardware.
1646 Software events are supported either built-in or via the
1647 use of generic tracepoints.
1649 Most modern CPUs support performance events via performance
1650 counter registers. These registers count the number of certain
1651 types of hw events: such as instructions executed, cachemisses
1652 suffered, or branches mis-predicted - without slowing down the
1653 kernel or applications. These registers can also trigger interrupts
1654 when a threshold number of events have passed - and can thus be
1655 used to profile the code that runs on that CPU.
1657 The Linux Performance Event subsystem provides an abstraction of
1658 these software and hardware event capabilities, available via a
1659 system call and used by the "perf" utility in tools/perf/. It
1660 provides per task and per CPU counters, and it provides event
1661 capabilities on top of those.
1665 config DEBUG_PERF_USE_VMALLOC
1667 bool "Debug: use vmalloc to back perf mmap() buffers"
1668 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC
1669 select PERF_USE_VMALLOC
1671 Use vmalloc memory to back perf mmap() buffers.
1673 Mostly useful for debugging the vmalloc code on platforms
1674 that don't require it.
1680 config VM_EVENT_COUNTERS
1682 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1684 VM event counters are needed for event counts to be shown.
1685 This option allows the disabling of the VM event counters
1686 on EXPERT systems. /proc/vmstat will only show page counts
1687 if VM event counters are disabled.
1691 bool "Enable SLUB debugging support" if EXPERT
1692 depends on SLUB && SYSFS
1694 SLUB has extensive debug support features. Disabling these can
1695 result in significant savings in code size. This also disables
1696 SLUB sysfs support. /sys/slab will not exist and there will be
1697 no support for cache validation etc.
1700 bool "Disable heap randomization"
1703 Randomizing heap placement makes heap exploits harder, but it
1704 also breaks ancient binaries (including anything libc5 based).
1705 This option changes the bootup default to heap randomization
1706 disabled, and can be overridden at runtime by setting
1707 /proc/sys/kernel/randomize_va_space to 2.
1709 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1712 prompt "Choose SLAB allocator"
1715 This option allows to select a slab allocator.
1720 The regular slab allocator that is established and known to work
1721 well in all environments. It organizes cache hot objects in
1722 per cpu and per node queues.
1725 bool "SLUB (Unqueued Allocator)"
1727 SLUB is a slab allocator that minimizes cache line usage
1728 instead of managing queues of cached objects (SLAB approach).
1729 Per cpu caching is realized using slabs of objects instead
1730 of queues of objects. SLUB can use memory efficiently
1731 and has enhanced diagnostics. SLUB is the default choice for
1736 bool "SLOB (Simple Allocator)"
1738 SLOB replaces the stock allocator with a drastically simpler
1739 allocator. SLOB is generally more space efficient but
1740 does not perform as well on large systems.
1744 config SLUB_CPU_PARTIAL
1746 depends on SLUB && SMP
1747 bool "SLUB per cpu partial cache"
1749 Per cpu partial caches accellerate objects allocation and freeing
1750 that is local to a processor at the price of more indeterminism
1751 in the latency of the free. On overflow these caches will be cleared
1752 which requires the taking of locks that may cause latency spikes.
1753 Typically one would choose no for a realtime system.
1755 config MMAP_ALLOW_UNINITIALIZED
1756 bool "Allow mmapped anonymous memory to be uninitialized"
1757 depends on EXPERT && !MMU
1760 Normally, and according to the Linux spec, anonymous memory obtained
1761 from mmap() has it's contents cleared before it is passed to
1762 userspace. Enabling this config option allows you to request that
1763 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1764 providing a huge performance boost. If this option is not enabled,
1765 then the flag will be ignored.
1767 This is taken advantage of by uClibc's malloc(), and also by
1768 ELF-FDPIC binfmt's brk and stack allocator.
1770 Because of the obvious security issues, this option should only be
1771 enabled on embedded devices where you control what is run in
1772 userspace. Since that isn't generally a problem on no-MMU systems,
1773 it is normally safe to say Y here.
1775 See Documentation/nommu-mmap.txt for more information.
1777 config SYSTEM_DATA_VERIFICATION
1779 select SYSTEM_TRUSTED_KEYRING
1783 select ASYMMETRIC_KEY_TYPE
1784 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
1787 select X509_CERTIFICATE_PARSER
1788 select PKCS7_MESSAGE_PARSER
1790 Provide PKCS#7 message verification using the contents of the system
1791 trusted keyring to provide public keys. This then can be used for
1792 module verification, kexec image verification and firmware blob
1796 bool "Profiling support"
1798 Say Y here to enable the extended profiling support mechanisms used
1799 by profilers such as OProfile.
1802 # Place an empty function call at each tracepoint site. Can be
1803 # dynamically changed for a probe function.
1808 source "arch/Kconfig"
1810 endmenu # General setup
1812 config HAVE_GENERIC_DMA_COHERENT
1819 depends on SLAB || SLUB_DEBUG
1827 default 0 if BASE_FULL
1828 default 1 if !BASE_FULL
1831 bool "Enable loadable module support"
1834 Kernel modules are small pieces of compiled code which can
1835 be inserted in the running kernel, rather than being
1836 permanently built into the kernel. You use the "modprobe"
1837 tool to add (and sometimes remove) them. If you say Y here,
1838 many parts of the kernel can be built as modules (by
1839 answering M instead of Y where indicated): this is most
1840 useful for infrequently used options which are not required
1841 for booting. For more information, see the man pages for
1842 modprobe, lsmod, modinfo, insmod and rmmod.
1844 If you say Y here, you will need to run "make
1845 modules_install" to put the modules under /lib/modules/
1846 where modprobe can find them (you may need to be root to do
1853 config MODULE_FORCE_LOAD
1854 bool "Forced module loading"
1857 Allow loading of modules without version information (ie. modprobe
1858 --force). Forced module loading sets the 'F' (forced) taint flag and
1859 is usually a really bad idea.
1861 config MODULE_UNLOAD
1862 bool "Module unloading"
1864 Without this option you will not be able to unload any
1865 modules (note that some modules may not be unloadable
1866 anyway), which makes your kernel smaller, faster
1867 and simpler. If unsure, say Y.
1869 config MODULE_FORCE_UNLOAD
1870 bool "Forced module unloading"
1871 depends on MODULE_UNLOAD
1873 This option allows you to force a module to unload, even if the
1874 kernel believes it is unsafe: the kernel will remove the module
1875 without waiting for anyone to stop using it (using the -f option to
1876 rmmod). This is mainly for kernel developers and desperate users.
1880 bool "Module versioning support"
1882 Usually, you have to use modules compiled with your kernel.
1883 Saying Y here makes it sometimes possible to use modules
1884 compiled for different kernels, by adding enough information
1885 to the modules to (hopefully) spot any changes which would
1886 make them incompatible with the kernel you are running. If
1889 config MODULE_SRCVERSION_ALL
1890 bool "Source checksum for all modules"
1892 Modules which contain a MODULE_VERSION get an extra "srcversion"
1893 field inserted into their modinfo section, which contains a
1894 sum of the source files which made it. This helps maintainers
1895 see exactly which source was used to build a module (since
1896 others sometimes change the module source without updating
1897 the version). With this option, such a "srcversion" field
1898 will be created for all modules. If unsure, say N.
1901 bool "Module signature verification"
1903 select SYSTEM_DATA_VERIFICATION
1905 Check modules for valid signatures upon load: the signature
1906 is simply appended to the module. For more information see
1907 Documentation/module-signing.txt.
1909 Note that this option adds the OpenSSL development packages as a
1910 kernel build dependency so that the signing tool can use its crypto
1913 !!!WARNING!!! If you enable this option, you MUST make sure that the
1914 module DOES NOT get stripped after being signed. This includes the
1915 debuginfo strip done by some packagers (such as rpmbuild) and
1916 inclusion into an initramfs that wants the module size reduced.
1918 config MODULE_SIG_FORCE
1919 bool "Require modules to be validly signed"
1920 depends on MODULE_SIG
1922 Reject unsigned modules or signed modules for which we don't have a
1923 key. Without this, such modules will simply taint the kernel.
1925 config MODULE_SIG_ALL
1926 bool "Automatically sign all modules"
1928 depends on MODULE_SIG
1930 Sign all modules during make modules_install. Without this option,
1931 modules must be signed manually, using the scripts/sign-file tool.
1933 comment "Do not forget to sign required modules with scripts/sign-file"
1934 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL
1937 prompt "Which hash algorithm should modules be signed with?"
1938 depends on MODULE_SIG
1940 This determines which sort of hashing algorithm will be used during
1941 signature generation. This algorithm _must_ be built into the kernel
1942 directly so that signature verification can take place. It is not
1943 possible to load a signed module containing the algorithm to check
1944 the signature on that module.
1946 config MODULE_SIG_SHA1
1947 bool "Sign modules with SHA-1"
1950 config MODULE_SIG_SHA224
1951 bool "Sign modules with SHA-224"
1952 select CRYPTO_SHA256
1954 config MODULE_SIG_SHA256
1955 bool "Sign modules with SHA-256"
1956 select CRYPTO_SHA256
1958 config MODULE_SIG_SHA384
1959 bool "Sign modules with SHA-384"
1960 select CRYPTO_SHA512
1962 config MODULE_SIG_SHA512
1963 bool "Sign modules with SHA-512"
1964 select CRYPTO_SHA512
1968 config MODULE_SIG_HASH
1970 depends on MODULE_SIG
1971 default "sha1" if MODULE_SIG_SHA1
1972 default "sha224" if MODULE_SIG_SHA224
1973 default "sha256" if MODULE_SIG_SHA256
1974 default "sha384" if MODULE_SIG_SHA384
1975 default "sha512" if MODULE_SIG_SHA512
1977 config MODULE_COMPRESS
1978 bool "Compress modules on installation"
1982 Compresses kernel modules when 'make modules_install' is run; gzip or
1983 xz depending on "Compression algorithm" below.
1985 module-init-tools MAY support gzip, and kmod MAY support gzip and xz.
1987 Out-of-tree kernel modules installed using Kbuild will also be
1988 compressed upon installation.
1990 Note: for modules inside an initrd or initramfs, it's more efficient
1991 to compress the whole initrd or initramfs instead.
1993 Note: This is fully compatible with signed modules.
1998 prompt "Compression algorithm"
1999 depends on MODULE_COMPRESS
2000 default MODULE_COMPRESS_GZIP
2002 This determines which sort of compression will be used during
2003 'make modules_install'.
2005 GZIP (default) and XZ are supported.
2007 config MODULE_COMPRESS_GZIP
2010 config MODULE_COMPRESS_XZ
2017 config MODULES_TREE_LOOKUP
2019 depends on PERF_EVENTS || TRACING
2021 config INIT_ALL_POSSIBLE
2024 Back when each arch used to define their own cpu_online_mask and
2025 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
2026 with all 1s, and others with all 0s. When they were centralised,
2027 it was better to provide this option than to break all the archs
2028 and have several arch maintainers pursuing me down dark alleys.
2030 source "block/Kconfig"
2032 config PREEMPT_NOTIFIERS
2039 # Can be selected by architectures with broken toolchains
2040 # that get confused by correct const<->read_only section
2042 config BROKEN_RODATA
2048 Build a simple ASN.1 grammar compiler that produces a bytecode output
2049 that can be interpreted by the ASN.1 stream decoder and used to
2050 inform it as to what tags are to be expected in a stream and what
2051 functions to call on what tags.
2053 source "kernel/Kconfig.locks"