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
40 depends on BROKEN || !SMP
43 config INIT_ENV_ARG_LIMIT
48 Maximum of each of the number of arguments and environment
49 variables passed to init from the kernel command line.
53 string "Cross-compiler tool prefix"
55 Same as running 'make CROSS_COMPILE=prefix-' but stored for
56 default make runs in this kernel build directory. You don't
57 need to set this unless you want the configured kernel build
58 directory to select the cross-compiler automatically.
61 string "Local version - append to kernel release"
63 Append an extra string to the end of your kernel version.
64 This will show up when you type uname, for example.
65 The string you set here will be appended after the contents of
66 any files with a filename matching localversion* in your
67 object and source tree, in that order. Your total string can
68 be a maximum of 64 characters.
70 config LOCALVERSION_AUTO
71 bool "Automatically append version information to the version string"
74 This will try to automatically determine if the current tree is a
75 release tree by looking for git tags that belong to the current
78 A string of the format -gxxxxxxxx will be added to the localversion
79 if a git-based tree is found. The string generated by this will be
80 appended after any matching localversion* files, and after the value
81 set in CONFIG_LOCALVERSION.
83 (The actual string used here is the first eight characters produced
84 by running the command:
86 $ git rev-parse --verify HEAD
88 which is done within the script "scripts/setlocalversion".)
90 config HAVE_KERNEL_GZIP
93 config HAVE_KERNEL_BZIP2
96 config HAVE_KERNEL_LZMA
102 config HAVE_KERNEL_LZO
106 prompt "Kernel compression mode"
108 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO
110 The linux kernel is a kind of self-extracting executable.
111 Several compression algorithms are available, which differ
112 in efficiency, compression and decompression speed.
113 Compression speed is only relevant when building a kernel.
114 Decompression speed is relevant at each boot.
116 If you have any problems with bzip2 or lzma compressed
117 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
118 version of this functionality (bzip2 only), for 2.4, was
119 supplied by Christian Ludwig)
121 High compression options are mostly useful for users, who
122 are low on disk space (embedded systems), but for whom ram
125 If in doubt, select 'gzip'
129 depends on HAVE_KERNEL_GZIP
131 The old and tried gzip compression. It provides a good balance
132 between compression ratio and decompression speed.
136 depends on HAVE_KERNEL_BZIP2
138 Its compression ratio and speed is intermediate.
139 Decompression speed is slowest among the choices. The kernel
140 size is about 10% smaller with bzip2, in comparison to gzip.
141 Bzip2 uses a large amount of memory. For modern kernels you
142 will need at least 8MB RAM or more for booting.
146 depends on HAVE_KERNEL_LZMA
148 This compression algorithm's ratio is best. Decompression speed
149 is between gzip and bzip2. Compression is slowest.
150 The kernel size is about 33% smaller with LZMA in comparison to gzip.
154 depends on HAVE_KERNEL_XZ
156 XZ uses the LZMA2 algorithm and instruction set specific
157 BCJ filters which can improve compression ratio of executable
158 code. The size of the kernel is about 30% smaller with XZ in
159 comparison to gzip. On architectures for which there is a BCJ
160 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
161 will create a few percent smaller kernel than plain LZMA.
163 The speed is about the same as with LZMA: The decompression
164 speed of XZ is better than that of bzip2 but worse than gzip
165 and LZO. Compression is slow.
169 depends on HAVE_KERNEL_LZO
171 Its compression ratio is the poorest among the choices. The kernel
172 size is about 10% bigger than gzip; however its speed
173 (both compression and decompression) is the fastest.
177 config DEFAULT_HOSTNAME
178 string "Default hostname"
181 This option determines the default system hostname before userspace
182 calls sethostname(2). The kernel traditionally uses "(none)" here,
183 but you may wish to use a different default here to make a minimal
184 system more usable with less configuration.
187 bool "Support for paging of anonymous memory (swap)"
188 depends on MMU && BLOCK
191 This option allows you to choose whether you want to have support
192 for so called swap devices or swap files in your kernel that are
193 used to provide more virtual memory than the actual RAM present
194 in your computer. If unsure say Y.
199 Inter Process Communication is a suite of library functions and
200 system calls which let processes (running programs) synchronize and
201 exchange information. It is generally considered to be a good thing,
202 and some programs won't run unless you say Y here. In particular, if
203 you want to run the DOS emulator dosemu under Linux (read the
204 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
205 you'll need to say Y here.
207 You can find documentation about IPC with "info ipc" and also in
208 section 6.4 of the Linux Programmer's Guide, available from
209 <http://www.tldp.org/guides.html>.
211 config SYSVIPC_SYSCTL
218 bool "POSIX Message Queues"
221 POSIX variant of message queues is a part of IPC. In POSIX message
222 queues every message has a priority which decides about succession
223 of receiving it by a process. If you want to compile and run
224 programs written e.g. for Solaris with use of its POSIX message
225 queues (functions mq_*) say Y here.
227 POSIX message queues are visible as a filesystem called 'mqueue'
228 and can be mounted somewhere if you want to do filesystem
229 operations on message queues.
233 config POSIX_MQUEUE_SYSCTL
235 depends on POSIX_MQUEUE
240 bool "open by fhandle syscalls"
243 If you say Y here, a user level program will be able to map
244 file names to handle and then later use the handle for
245 different file system operations. This is useful in implementing
246 userspace file servers, which now track files using handles instead
247 of names. The handle would remain the same even if file names
248 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
252 bool "Auditing support"
255 Enable auditing infrastructure that can be used with another
256 kernel subsystem, such as SELinux (which requires this for
257 logging of avc messages output). Does not do system-call
258 auditing without CONFIG_AUDITSYSCALL.
261 bool "Enable system-call auditing support"
262 depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH || (ARM && AEABI && !OABI_COMPAT))
263 default y if SECURITY_SELINUX
265 Enable low-overhead system-call auditing infrastructure that
266 can be used independently or with another kernel subsystem,
271 depends on AUDITSYSCALL
276 depends on AUDITSYSCALL
279 config AUDIT_LOGINUID_IMMUTABLE
280 bool "Make audit loginuid immutable"
283 The config option toggles if a task setting its loginuid requires
284 CAP_SYS_AUDITCONTROL or if that task should require no special permissions
285 but should instead only allow setting its loginuid if it was never
286 previously set. On systems which use systemd or a similar central
287 process to restart login services this should be set to true. On older
288 systems in which an admin would typically have to directly stop and
289 start processes this should be set to false. Setting this to true allows
290 one to drop potentially dangerous capabilites from the login tasks,
291 but may not be backwards compatible with older init systems.
293 source "kernel/irq/Kconfig"
294 source "kernel/time/Kconfig"
296 menu "CPU/Task time and stats accounting"
298 config VIRT_CPU_ACCOUNTING
302 prompt "Cputime accounting"
303 default TICK_CPU_ACCOUNTING if !PPC64
304 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64
306 # Kind of a stub config for the pure tick based cputime accounting
307 config TICK_CPU_ACCOUNTING
308 bool "Simple tick based cputime accounting"
311 This is the basic tick based cputime accounting that maintains
312 statistics about user, system and idle time spent on per jiffies
317 config VIRT_CPU_ACCOUNTING_NATIVE
318 bool "Deterministic task and CPU time accounting"
319 depends on HAVE_VIRT_CPU_ACCOUNTING
320 select VIRT_CPU_ACCOUNTING
322 Select this option to enable more accurate task and CPU time
323 accounting. This is done by reading a CPU counter on each
324 kernel entry and exit and on transitions within the kernel
325 between system, softirq and hardirq state, so there is a
326 small performance impact. In the case of s390 or IBM POWER > 5,
327 this also enables accounting of stolen time on logically-partitioned
330 config VIRT_CPU_ACCOUNTING_GEN
331 bool "Full dynticks CPU time accounting"
332 depends on HAVE_CONTEXT_TRACKING && 64BIT
333 select VIRT_CPU_ACCOUNTING
334 select CONTEXT_TRACKING
336 Select this option to enable task and CPU time accounting on full
337 dynticks systems. This accounting is implemented by watching every
338 kernel-user boundaries using the context tracking subsystem.
339 The accounting is thus performed at the expense of some significant
342 For now this is only useful if you are working on the full
343 dynticks subsystem development.
347 config IRQ_TIME_ACCOUNTING
348 bool "Fine granularity task level IRQ time accounting"
349 depends on HAVE_IRQ_TIME_ACCOUNTING
351 Select this option to enable fine granularity task irq time
352 accounting. This is done by reading a timestamp on each
353 transitions between softirq and hardirq state, so there can be a
354 small performance impact.
356 If in doubt, say N here.
360 config BSD_PROCESS_ACCT
361 bool "BSD Process Accounting"
363 If you say Y here, a user level program will be able to instruct the
364 kernel (via a special system call) to write process accounting
365 information to a file: whenever a process exits, information about
366 that process will be appended to the file by the kernel. The
367 information includes things such as creation time, owning user,
368 command name, memory usage, controlling terminal etc. (the complete
369 list is in the struct acct in <file:include/linux/acct.h>). It is
370 up to the user level program to do useful things with this
371 information. This is generally a good idea, so say Y.
373 config BSD_PROCESS_ACCT_V3
374 bool "BSD Process Accounting version 3 file format"
375 depends on BSD_PROCESS_ACCT
378 If you say Y here, the process accounting information is written
379 in a new file format that also logs the process IDs of each
380 process and it's parent. Note that this file format is incompatible
381 with previous v0/v1/v2 file formats, so you will need updated tools
382 for processing it. A preliminary version of these tools is available
383 at <http://www.gnu.org/software/acct/>.
386 bool "Export task/process statistics through netlink"
390 Export selected statistics for tasks/processes through the
391 generic netlink interface. Unlike BSD process accounting, the
392 statistics are available during the lifetime of tasks/processes as
393 responses to commands. Like BSD accounting, they are sent to user
398 config TASK_DELAY_ACCT
399 bool "Enable per-task delay accounting"
402 Collect information on time spent by a task waiting for system
403 resources like cpu, synchronous block I/O completion and swapping
404 in pages. Such statistics can help in setting a task's priorities
405 relative to other tasks for cpu, io, rss limits etc.
410 bool "Enable extended accounting over taskstats"
413 Collect extended task accounting data and send the data
414 to userland for processing over the taskstats interface.
418 config TASK_IO_ACCOUNTING
419 bool "Enable per-task storage I/O accounting"
420 depends on TASK_XACCT
422 Collect information on the number of bytes of storage I/O which this
427 endmenu # "CPU/Task time and stats accounting"
432 prompt "RCU Implementation"
436 bool "Tree-based hierarchical RCU"
437 depends on !PREEMPT && SMP
439 This option selects the RCU implementation that is
440 designed for very large SMP system with hundreds or
441 thousands of CPUs. It also scales down nicely to
444 config TREE_PREEMPT_RCU
445 bool "Preemptible tree-based hierarchical RCU"
448 This option selects the RCU implementation that is
449 designed for very large SMP systems with hundreds or
450 thousands of CPUs, but for which real-time response
451 is also required. It also scales down nicely to
454 Select this option if you are unsure.
457 bool "UP-only small-memory-footprint RCU"
458 depends on !PREEMPT && !SMP
460 This option selects the RCU implementation that is
461 designed for UP systems from which real-time response
462 is not required. This option greatly reduces the
463 memory footprint of RCU.
465 config TINY_PREEMPT_RCU
466 bool "Preemptible UP-only small-memory-footprint RCU"
467 depends on PREEMPT && !SMP
469 This option selects the RCU implementation that is designed
470 for real-time UP systems. This option greatly reduces the
471 memory footprint of RCU.
476 def_bool ( TREE_PREEMPT_RCU || TINY_PREEMPT_RCU )
478 This option enables preemptible-RCU code that is common between
479 the TREE_PREEMPT_RCU and TINY_PREEMPT_RCU implementations.
481 config RCU_STALL_COMMON
482 def_bool ( TREE_RCU || TREE_PREEMPT_RCU || RCU_TRACE )
484 This option enables RCU CPU stall code that is common between
485 the TINY and TREE variants of RCU. The purpose is to allow
486 the tiny variants to disable RCU CPU stall warnings, while
487 making these warnings mandatory for the tree variants.
489 config CONTEXT_TRACKING
493 bool "Consider userspace as in RCU extended quiescent state"
494 depends on HAVE_CONTEXT_TRACKING && SMP
495 select CONTEXT_TRACKING
497 This option sets hooks on kernel / userspace boundaries and
498 puts RCU in extended quiescent state when the CPU runs in
499 userspace. It means that when a CPU runs in userspace, it is
500 excluded from the global RCU state machine and thus doesn't
501 try to keep the timer tick on for RCU.
503 Unless you want to hack and help the development of the full
504 dynticks mode, you shouldn't enable this option. It also
505 adds unnecessary overhead.
509 config CONTEXT_TRACKING_FORCE
510 bool "Force context tracking"
511 depends on CONTEXT_TRACKING
513 Probe on user/kernel boundaries by default in order to
514 test the features that rely on it such as userspace RCU extended
516 This test is there for debugging until we have a real user like the
520 int "Tree-based hierarchical RCU fanout value"
523 depends on TREE_RCU || TREE_PREEMPT_RCU
527 This option controls the fanout of hierarchical implementations
528 of RCU, allowing RCU to work efficiently on machines with
529 large numbers of CPUs. This value must be at least the fourth
530 root of NR_CPUS, which allows NR_CPUS to be insanely large.
531 The default value of RCU_FANOUT should be used for production
532 systems, but if you are stress-testing the RCU implementation
533 itself, small RCU_FANOUT values allow you to test large-system
534 code paths on small(er) systems.
536 Select a specific number if testing RCU itself.
537 Take the default if unsure.
539 config RCU_FANOUT_LEAF
540 int "Tree-based hierarchical RCU leaf-level fanout value"
541 range 2 RCU_FANOUT if 64BIT
542 range 2 RCU_FANOUT if !64BIT
543 depends on TREE_RCU || TREE_PREEMPT_RCU
546 This option controls the leaf-level fanout of hierarchical
547 implementations of RCU, and allows trading off cache misses
548 against lock contention. Systems that synchronize their
549 scheduling-clock interrupts for energy-efficiency reasons will
550 want the default because the smaller leaf-level fanout keeps
551 lock contention levels acceptably low. Very large systems
552 (hundreds or thousands of CPUs) will instead want to set this
553 value to the maximum value possible in order to reduce the
554 number of cache misses incurred during RCU's grace-period
555 initialization. These systems tend to run CPU-bound, and thus
556 are not helped by synchronized interrupts, and thus tend to
557 skew them, which reduces lock contention enough that large
558 leaf-level fanouts work well.
560 Select a specific number if testing RCU itself.
562 Select the maximum permissible value for large systems.
564 Take the default if unsure.
566 config RCU_FANOUT_EXACT
567 bool "Disable tree-based hierarchical RCU auto-balancing"
568 depends on TREE_RCU || TREE_PREEMPT_RCU
571 This option forces use of the exact RCU_FANOUT value specified,
572 regardless of imbalances in the hierarchy. This is useful for
573 testing RCU itself, and might one day be useful on systems with
574 strong NUMA behavior.
576 Without RCU_FANOUT_EXACT, the code will balance the hierarchy.
580 config RCU_FAST_NO_HZ
581 bool "Accelerate last non-dyntick-idle CPU's grace periods"
582 depends on NO_HZ && SMP
585 This option causes RCU to attempt to accelerate grace periods in
586 order to allow CPUs to enter dynticks-idle state more quickly.
587 On the other hand, this option increases the overhead of the
588 dynticks-idle checking, thus degrading scheduling latency.
590 Say Y if energy efficiency is critically important, and you don't
591 care about real-time response.
593 Say N if you are unsure.
595 config TREE_RCU_TRACE
596 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU )
599 This option provides tracing for the TREE_RCU and
600 TREE_PREEMPT_RCU implementations, permitting Makefile to
601 trivially select kernel/rcutree_trace.c.
604 bool "Enable RCU priority boosting"
605 depends on RT_MUTEXES && PREEMPT_RCU
608 This option boosts the priority of preempted RCU readers that
609 block the current preemptible RCU grace period for too long.
610 This option also prevents heavy loads from blocking RCU
611 callback invocation for all flavors of RCU.
613 Say Y here if you are working with real-time apps or heavy loads
614 Say N here if you are unsure.
616 config RCU_BOOST_PRIO
617 int "Real-time priority to boost RCU readers to"
622 This option specifies the real-time priority to which long-term
623 preempted RCU readers are to be boosted. If you are working
624 with a real-time application that has one or more CPU-bound
625 threads running at a real-time priority level, you should set
626 RCU_BOOST_PRIO to a priority higher then the highest-priority
627 real-time CPU-bound thread. The default RCU_BOOST_PRIO value
628 of 1 is appropriate in the common case, which is real-time
629 applications that do not have any CPU-bound threads.
631 Some real-time applications might not have a single real-time
632 thread that saturates a given CPU, but instead might have
633 multiple real-time threads that, taken together, fully utilize
634 that CPU. In this case, you should set RCU_BOOST_PRIO to
635 a priority higher than the lowest-priority thread that is
636 conspiring to prevent the CPU from running any non-real-time
637 tasks. For example, if one thread at priority 10 and another
638 thread at priority 5 are between themselves fully consuming
639 the CPU time on a given CPU, then RCU_BOOST_PRIO should be
640 set to priority 6 or higher.
642 Specify the real-time priority, or take the default if unsure.
644 config RCU_BOOST_DELAY
645 int "Milliseconds to delay boosting after RCU grace-period start"
650 This option specifies the time to wait after the beginning of
651 a given grace period before priority-boosting preempted RCU
652 readers blocking that grace period. Note that any RCU reader
653 blocking an expedited RCU grace period is boosted immediately.
655 Accept the default if unsure.
658 bool "Offload RCU callback processing from boot-selected CPUs"
659 depends on TREE_RCU || TREE_PREEMPT_RCU
662 Use this option to reduce OS jitter for aggressive HPC or
663 real-time workloads. It can also be used to offload RCU
664 callback invocation to energy-efficient CPUs in battery-powered
665 asymmetric multiprocessors.
667 This option offloads callback invocation from the set of
668 CPUs specified at boot time by the rcu_nocbs parameter.
669 For each such CPU, a kthread ("rcuoN") will be created to
670 invoke callbacks, where the "N" is the CPU being offloaded.
671 Nothing prevents this kthread from running on the specified
672 CPUs, but (1) the kthreads may be preempted between each
673 callback, and (2) affinity or cgroups can be used to force
674 the kthreads to run on whatever set of CPUs is desired.
676 Say Y here if you want reduced OS jitter on selected CPUs.
677 Say N here if you are unsure.
679 endmenu # "RCU Subsystem"
682 tristate "Kernel .config support"
684 This option enables the complete Linux kernel ".config" file
685 contents to be saved in the kernel. It provides documentation
686 of which kernel options are used in a running kernel or in an
687 on-disk kernel. This information can be extracted from the kernel
688 image file with the script scripts/extract-ikconfig and used as
689 input to rebuild the current kernel or to build another kernel.
690 It can also be extracted from a running kernel by reading
691 /proc/config.gz if enabled (below).
694 bool "Enable access to .config through /proc/config.gz"
695 depends on IKCONFIG && PROC_FS
697 This option enables access to the kernel configuration file
698 through /proc/config.gz.
701 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
705 Select kernel log buffer size as a power of 2.
715 # Architectures with an unreliable sched_clock() should select this:
717 config HAVE_UNSTABLE_SCHED_CLOCK
721 # For architectures that want to enable the support for NUMA-affine scheduler
724 config ARCH_SUPPORTS_NUMA_BALANCING
727 # For architectures that (ab)use NUMA to represent different memory regions
728 # all cpu-local but of different latencies, such as SuperH.
730 config ARCH_WANT_NUMA_VARIABLE_LOCALITY
734 # For architectures that are willing to define _PAGE_NUMA as _PAGE_PROTNONE
735 config ARCH_WANTS_PROT_NUMA_PROT_NONE
738 config ARCH_USES_NUMA_PROT_NONE
741 depends on ARCH_WANTS_PROT_NUMA_PROT_NONE
742 depends on NUMA_BALANCING
744 config NUMA_BALANCING_DEFAULT_ENABLED
745 bool "Automatically enable NUMA aware memory/task placement"
747 depends on NUMA_BALANCING
749 If set, autonumic NUMA balancing will be enabled if running on a NUMA
752 config NUMA_BALANCING
753 bool "Memory placement aware NUMA scheduler"
754 depends on ARCH_SUPPORTS_NUMA_BALANCING
755 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
756 depends on SMP && NUMA && MIGRATION
758 This option adds support for automatic NUMA aware memory/task placement.
759 The mechanism is quite primitive and is based on migrating memory when
760 it is references to the node the task is running on.
762 This system will be inactive on UMA systems.
765 boolean "Control Group support"
768 This option adds support for grouping sets of processes together, for
769 use with process control subsystems such as Cpusets, CFS, memory
770 controls or device isolation.
772 - Documentation/scheduler/sched-design-CFS.txt (CFS)
773 - Documentation/cgroups/ (features for grouping, isolation
774 and resource control)
781 bool "Example debug cgroup subsystem"
784 This option enables a simple cgroup subsystem that
785 exports useful debugging information about the cgroups
790 config CGROUP_FREEZER
791 bool "Freezer cgroup subsystem"
793 Provides a way to freeze and unfreeze all tasks in a
797 bool "Device controller for cgroups"
799 Provides a cgroup implementing whitelists for devices which
800 a process in the cgroup can mknod or open.
803 bool "Cpuset support"
805 This option will let you create and manage CPUSETs which
806 allow dynamically partitioning a system into sets of CPUs and
807 Memory Nodes and assigning tasks to run only within those sets.
808 This is primarily useful on large SMP or NUMA systems.
812 config PROC_PID_CPUSET
813 bool "Include legacy /proc/<pid>/cpuset file"
817 config CGROUP_CPUACCT
818 bool "Simple CPU accounting cgroup subsystem"
820 Provides a simple Resource Controller for monitoring the
821 total CPU consumed by the tasks in a cgroup.
823 config RESOURCE_COUNTERS
824 bool "Resource counters"
826 This option enables controller independent resource accounting
827 infrastructure that works with cgroups.
830 bool "Memory Resource Controller for Control Groups"
831 depends on RESOURCE_COUNTERS
834 Provides a memory resource controller that manages both anonymous
835 memory and page cache. (See Documentation/cgroups/memory.txt)
837 Note that setting this option increases fixed memory overhead
838 associated with each page of memory in the system. By this,
839 20(40)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory
840 usage tracking struct at boot. Total amount of this is printed out
843 Only enable when you're ok with these trade offs and really
844 sure you need the memory resource controller. Even when you enable
845 this, you can set "cgroup_disable=memory" at your boot option to
846 disable memory resource controller and you can avoid overheads.
847 (and lose benefits of memory resource controller)
849 This config option also selects MM_OWNER config option, which
850 could in turn add some fork/exit overhead.
853 bool "Memory Resource Controller Swap Extension"
854 depends on MEMCG && SWAP
856 Add swap management feature to memory resource controller. When you
857 enable this, you can limit mem+swap usage per cgroup. In other words,
858 when you disable this, memory resource controller has no cares to
859 usage of swap...a process can exhaust all of the swap. This extension
860 is useful when you want to avoid exhaustion swap but this itself
861 adds more overheads and consumes memory for remembering information.
862 Especially if you use 32bit system or small memory system, please
863 be careful about enabling this. When memory resource controller
864 is disabled by boot option, this will be automatically disabled and
865 there will be no overhead from this. Even when you set this config=y,
866 if boot option "swapaccount=0" is set, swap will not be accounted.
867 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page
868 size is 4096bytes, 512k per 1Gbytes of swap.
869 config MEMCG_SWAP_ENABLED
870 bool "Memory Resource Controller Swap Extension enabled by default"
871 depends on MEMCG_SWAP
874 Memory Resource Controller Swap Extension comes with its price in
875 a bigger memory consumption. General purpose distribution kernels
876 which want to enable the feature but keep it disabled by default
877 and let the user enable it by swapaccount boot command line
878 parameter should have this option unselected.
879 For those who want to have the feature enabled by default should
880 select this option (if, for some reason, they need to disable it
881 then swapaccount=0 does the trick).
883 bool "Memory Resource Controller Kernel Memory accounting"
885 depends on SLUB || SLAB
887 The Kernel Memory extension for Memory Resource Controller can limit
888 the amount of memory used by kernel objects in the system. Those are
889 fundamentally different from the entities handled by the standard
890 Memory Controller, which are page-based, and can be swapped. Users of
891 the kmem extension can use it to guarantee that no group of processes
892 will ever exhaust kernel resources alone.
894 config CGROUP_HUGETLB
895 bool "HugeTLB Resource Controller for Control Groups"
896 depends on RESOURCE_COUNTERS && HUGETLB_PAGE
899 Provides a cgroup Resource Controller for HugeTLB pages.
900 When you enable this, you can put a per cgroup limit on HugeTLB usage.
901 The limit is enforced during page fault. Since HugeTLB doesn't
902 support page reclaim, enforcing the limit at page fault time implies
903 that, the application will get SIGBUS signal if it tries to access
904 HugeTLB pages beyond its limit. This requires the application to know
905 beforehand how much HugeTLB pages it would require for its use. The
906 control group is tracked in the third page lru pointer. This means
907 that we cannot use the controller with huge page less than 3 pages.
910 bool "Enable perf_event per-cpu per-container group (cgroup) monitoring"
911 depends on PERF_EVENTS && CGROUPS
913 This option extends the per-cpu mode to restrict monitoring to
914 threads which belong to the cgroup specified and run on the
919 menuconfig CGROUP_SCHED
920 bool "Group CPU scheduler"
923 This feature lets CPU scheduler recognize task groups and control CPU
924 bandwidth allocation to such task groups. It uses cgroups to group
928 config FAIR_GROUP_SCHED
929 bool "Group scheduling for SCHED_OTHER"
930 depends on CGROUP_SCHED
934 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
935 depends on FAIR_GROUP_SCHED
938 This option allows users to define CPU bandwidth rates (limits) for
939 tasks running within the fair group scheduler. Groups with no limit
940 set are considered to be unconstrained and will run with no
942 See tip/Documentation/scheduler/sched-bwc.txt for more information.
944 config RT_GROUP_SCHED
945 bool "Group scheduling for SCHED_RR/FIFO"
946 depends on CGROUP_SCHED
949 This feature lets you explicitly allocate real CPU bandwidth
950 to task groups. If enabled, it will also make it impossible to
951 schedule realtime tasks for non-root users until you allocate
952 realtime bandwidth for them.
953 See Documentation/scheduler/sched-rt-group.txt for more information.
958 bool "Block IO controller"
962 Generic block IO controller cgroup interface. This is the common
963 cgroup interface which should be used by various IO controlling
966 Currently, CFQ IO scheduler uses it to recognize task groups and
967 control disk bandwidth allocation (proportional time slice allocation)
968 to such task groups. It is also used by bio throttling logic in
969 block layer to implement upper limit in IO rates on a device.
971 This option only enables generic Block IO controller infrastructure.
972 One needs to also enable actual IO controlling logic/policy. For
973 enabling proportional weight division of disk bandwidth in CFQ, set
974 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
975 CONFIG_BLK_DEV_THROTTLING=y.
977 See Documentation/cgroups/blkio-controller.txt for more information.
979 config DEBUG_BLK_CGROUP
980 bool "Enable Block IO controller debugging"
981 depends on BLK_CGROUP
984 Enable some debugging help. Currently it exports additional stat
985 files in a cgroup which can be useful for debugging.
989 config CHECKPOINT_RESTORE
990 bool "Checkpoint/restore support" if EXPERT
993 Enables additional kernel features in a sake of checkpoint/restore.
994 In particular it adds auxiliary prctl codes to setup process text,
995 data and heap segment sizes, and a few additional /proc filesystem
998 If unsure, say N here.
1000 menuconfig NAMESPACES
1001 bool "Namespaces support" if EXPERT
1004 Provides the way to make tasks work with different objects using
1005 the same id. For example same IPC id may refer to different objects
1006 or same user id or pid may refer to different tasks when used in
1007 different namespaces.
1012 bool "UTS namespace"
1015 In this namespace tasks see different info provided with the
1019 bool "IPC namespace"
1020 depends on (SYSVIPC || POSIX_MQUEUE)
1023 In this namespace tasks work with IPC ids which correspond to
1024 different IPC objects in different namespaces.
1027 bool "User namespace"
1028 depends on UIDGID_CONVERTED
1029 select UIDGID_STRICT_TYPE_CHECKS
1033 This allows containers, i.e. vservers, to use user namespaces
1034 to provide different user info for different servers.
1036 When user namespaces are enabled in the kernel it is
1037 recommended that the MEMCG and MEMCG_KMEM options also be
1038 enabled and that user-space use the memory control groups to
1039 limit the amount of memory a memory unprivileged users can
1045 bool "PID Namespaces"
1048 Support process id namespaces. This allows having multiple
1049 processes with the same pid as long as they are in different
1050 pid namespaces. This is a building block of containers.
1053 bool "Network namespace"
1057 Allow user space to create what appear to be multiple instances
1058 of the network stack.
1062 config UIDGID_CONVERTED
1063 # True if all of the selected software conmponents are known
1064 # to have uid_t and gid_t converted to kuid_t and kgid_t
1065 # where appropriate and are otherwise safe to use with
1066 # the user namespace.
1071 depends on XFS_FS = n
1073 config UIDGID_STRICT_TYPE_CHECKS
1074 bool "Require conversions between uid/gids and their internal representation"
1075 depends on UIDGID_CONVERTED
1078 While the nececessary conversions are being added to all subsystems this option allows
1079 the code to continue to build for unconverted subsystems.
1081 Say Y here if you want the strict type checking enabled
1083 config SCHED_AUTOGROUP
1084 bool "Automatic process group scheduling"
1088 select FAIR_GROUP_SCHED
1090 This option optimizes the scheduler for common desktop workloads by
1091 automatically creating and populating task groups. This separation
1092 of workloads isolates aggressive CPU burners (like build jobs) from
1093 desktop applications. Task group autogeneration is currently based
1099 config SYSFS_DEPRECATED
1100 bool "Enable deprecated sysfs features to support old userspace tools"
1104 This option adds code that switches the layout of the "block" class
1105 devices, to not show up in /sys/class/block/, but only in
1108 This switch is only active when the sysfs.deprecated=1 boot option is
1109 passed or the SYSFS_DEPRECATED_V2 option is set.
1111 This option allows new kernels to run on old distributions and tools,
1112 which might get confused by /sys/class/block/. Since 2007/2008 all
1113 major distributions and tools handle this just fine.
1115 Recent distributions and userspace tools after 2009/2010 depend on
1116 the existence of /sys/class/block/, and will not work with this
1119 Only if you are using a new kernel on an old distribution, you might
1122 config SYSFS_DEPRECATED_V2
1123 bool "Enable deprecated sysfs features by default"
1126 depends on SYSFS_DEPRECATED
1128 Enable deprecated sysfs by default.
1130 See the CONFIG_SYSFS_DEPRECATED option for more details about this
1133 Only if you are using a new kernel on an old distribution, you might
1134 need to say Y here. Even then, odds are you would not need it
1135 enabled, you can always pass the boot option if absolutely necessary.
1138 bool "Kernel->user space relay support (formerly relayfs)"
1140 This option enables support for relay interface support in
1141 certain file systems (such as debugfs).
1142 It is designed to provide an efficient mechanism for tools and
1143 facilities to relay large amounts of data from kernel space to
1148 config BLK_DEV_INITRD
1149 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
1150 depends on BROKEN || !FRV
1152 The initial RAM filesystem is a ramfs which is loaded by the
1153 boot loader (loadlin or lilo) and that is mounted as root
1154 before the normal boot procedure. It is typically used to
1155 load modules needed to mount the "real" root file system,
1156 etc. See <file:Documentation/initrd.txt> for details.
1158 If RAM disk support (BLK_DEV_RAM) is also included, this
1159 also enables initial RAM disk (initrd) support and adds
1160 15 Kbytes (more on some other architectures) to the kernel size.
1166 source "usr/Kconfig"
1170 config CC_OPTIMIZE_FOR_SIZE
1171 bool "Optimize for size"
1173 Enabling this option will pass "-Os" instead of "-O2" to gcc
1174 resulting in a smaller kernel.
1185 bool "Configure standard kernel features (expert users)"
1186 # Unhide debug options, to make the on-by-default options visible
1189 This option allows certain base kernel options and settings
1190 to be disabled or tweaked. This is for specialized
1191 environments which can tolerate a "non-standard" kernel.
1192 Only use this if you really know what you are doing.
1198 bool "Enable 16-bit UID system calls" if EXPERT
1199 depends on HAVE_UID16
1202 This enables the legacy 16-bit UID syscall wrappers.
1204 config SYSCTL_SYSCALL
1205 bool "Sysctl syscall support" if EXPERT
1206 depends on PROC_SYSCTL
1210 sys_sysctl uses binary paths that have been found challenging
1211 to properly maintain and use. The interface in /proc/sys
1212 using paths with ascii names is now the primary path to this
1215 Almost nothing using the binary sysctl interface so if you are
1216 trying to save some space it is probably safe to disable this,
1217 making your kernel marginally smaller.
1219 If unsure say N here.
1221 config SYSCTL_EXCEPTION_TRACE
1224 Enable support for /proc/sys/debug/exception-trace.
1226 config SYSCTL_ARCH_UNALIGN_NO_WARN
1229 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1230 Allows arch to define/use @no_unaligned_warning to possibly warn
1231 about unaligned access emulation going on under the hood.
1234 bool "Load all symbols for debugging/ksymoops" if EXPERT
1237 Say Y here to let the kernel print out symbolic crash information and
1238 symbolic stack backtraces. This increases the size of the kernel
1239 somewhat, as all symbols have to be loaded into the kernel image.
1242 bool "Include all symbols in kallsyms"
1243 depends on DEBUG_KERNEL && KALLSYMS
1245 Normally kallsyms only contains the symbols of functions for nicer
1246 OOPS messages and backtraces (i.e., symbols from the text and inittext
1247 sections). This is sufficient for most cases. And only in very rare
1248 cases (e.g., when a debugger is used) all symbols are required (e.g.,
1249 names of variables from the data sections, etc).
1251 This option makes sure that all symbols are loaded into the kernel
1252 image (i.e., symbols from all sections) in cost of increased kernel
1253 size (depending on the kernel configuration, it may be 300KiB or
1254 something like this).
1256 Say N unless you really need all symbols.
1263 bool "Enable support for printk" if EXPERT
1266 This option enables normal printk support. Removing it
1267 eliminates most of the message strings from the kernel image
1268 and makes the kernel more or less silent. As this makes it
1269 very difficult to diagnose system problems, saying N here is
1270 strongly discouraged.
1273 bool "BUG() support" if EXPERT
1276 Disabling this option eliminates support for BUG and WARN, reducing
1277 the size of your kernel image and potentially quietly ignoring
1278 numerous fatal conditions. You should only consider disabling this
1279 option for embedded systems with no facilities for reporting errors.
1285 bool "Enable ELF core dumps" if EXPERT
1287 Enable support for generating core dumps. Disabling saves about 4k.
1290 config PCSPKR_PLATFORM
1291 bool "Enable PC-Speaker support" if EXPERT
1292 depends on HAVE_PCSPKR_PLATFORM
1296 This option allows to disable the internal PC-Speaker
1297 support, saving some memory.
1299 config HAVE_PCSPKR_PLATFORM
1304 bool "Enable full-sized data structures for core" if EXPERT
1306 Disabling this option reduces the size of miscellaneous core
1307 kernel data structures. This saves memory on small machines,
1308 but may reduce performance.
1311 bool "Enable futex support" if EXPERT
1315 Disabling this option will cause the kernel to be built without
1316 support for "fast userspace mutexes". The resulting kernel may not
1317 run glibc-based applications correctly.
1320 bool "Enable eventpoll support" if EXPERT
1324 Disabling this option will cause the kernel to be built without
1325 support for epoll family of system calls.
1328 bool "Enable signalfd() system call" if EXPERT
1332 Enable the signalfd() system call that allows to receive signals
1333 on a file descriptor.
1338 bool "Enable timerfd() system call" if EXPERT
1342 Enable the timerfd() system call that allows to receive timer
1343 events on a file descriptor.
1348 bool "Enable eventfd() system call" if EXPERT
1352 Enable the eventfd() system call that allows to receive both
1353 kernel notification (ie. KAIO) or userspace notifications.
1358 bool "Use full shmem filesystem" if EXPERT
1362 The shmem is an internal filesystem used to manage shared memory.
1363 It is backed by swap and manages resource limits. It is also exported
1364 to userspace as tmpfs if TMPFS is enabled. Disabling this
1365 option replaces shmem and tmpfs with the much simpler ramfs code,
1366 which may be appropriate on small systems without swap.
1369 bool "Enable AIO support" if EXPERT
1372 This option enables POSIX asynchronous I/O which may by used
1373 by some high performance threaded applications. Disabling
1374 this option saves about 7k.
1377 bool "Embedded system"
1380 This option should be enabled if compiling the kernel for
1381 an embedded system so certain expert options are available
1384 config HAVE_PERF_EVENTS
1387 See tools/perf/design.txt for details.
1389 config PERF_USE_VMALLOC
1392 See tools/perf/design.txt for details
1394 menu "Kernel Performance Events And Counters"
1397 bool "Kernel performance events and counters"
1398 default y if PROFILING
1399 depends on HAVE_PERF_EVENTS
1403 Enable kernel support for various performance events provided
1404 by software and hardware.
1406 Software events are supported either built-in or via the
1407 use of generic tracepoints.
1409 Most modern CPUs support performance events via performance
1410 counter registers. These registers count the number of certain
1411 types of hw events: such as instructions executed, cachemisses
1412 suffered, or branches mis-predicted - without slowing down the
1413 kernel or applications. These registers can also trigger interrupts
1414 when a threshold number of events have passed - and can thus be
1415 used to profile the code that runs on that CPU.
1417 The Linux Performance Event subsystem provides an abstraction of
1418 these software and hardware event capabilities, available via a
1419 system call and used by the "perf" utility in tools/perf/. It
1420 provides per task and per CPU counters, and it provides event
1421 capabilities on top of those.
1425 config DEBUG_PERF_USE_VMALLOC
1427 bool "Debug: use vmalloc to back perf mmap() buffers"
1428 depends on PERF_EVENTS && DEBUG_KERNEL
1429 select PERF_USE_VMALLOC
1431 Use vmalloc memory to back perf mmap() buffers.
1433 Mostly useful for debugging the vmalloc code on platforms
1434 that don't require it.
1440 config VM_EVENT_COUNTERS
1442 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1444 VM event counters are needed for event counts to be shown.
1445 This option allows the disabling of the VM event counters
1446 on EXPERT systems. /proc/vmstat will only show page counts
1447 if VM event counters are disabled.
1451 bool "Enable PCI quirk workarounds" if EXPERT
1454 This enables workarounds for various PCI chipset
1455 bugs/quirks. Disable this only if your target machine is
1456 unaffected by PCI quirks.
1460 bool "Enable SLUB debugging support" if EXPERT
1461 depends on SLUB && SYSFS
1463 SLUB has extensive debug support features. Disabling these can
1464 result in significant savings in code size. This also disables
1465 SLUB sysfs support. /sys/slab will not exist and there will be
1466 no support for cache validation etc.
1469 bool "Disable heap randomization"
1472 Randomizing heap placement makes heap exploits harder, but it
1473 also breaks ancient binaries (including anything libc5 based).
1474 This option changes the bootup default to heap randomization
1475 disabled, and can be overridden at runtime by setting
1476 /proc/sys/kernel/randomize_va_space to 2.
1478 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1481 prompt "Choose SLAB allocator"
1484 This option allows to select a slab allocator.
1489 The regular slab allocator that is established and known to work
1490 well in all environments. It organizes cache hot objects in
1491 per cpu and per node queues.
1494 bool "SLUB (Unqueued Allocator)"
1496 SLUB is a slab allocator that minimizes cache line usage
1497 instead of managing queues of cached objects (SLAB approach).
1498 Per cpu caching is realized using slabs of objects instead
1499 of queues of objects. SLUB can use memory efficiently
1500 and has enhanced diagnostics. SLUB is the default choice for
1505 bool "SLOB (Simple Allocator)"
1507 SLOB replaces the stock allocator with a drastically simpler
1508 allocator. SLOB is generally more space efficient but
1509 does not perform as well on large systems.
1513 config MMAP_ALLOW_UNINITIALIZED
1514 bool "Allow mmapped anonymous memory to be uninitialized"
1515 depends on EXPERT && !MMU
1518 Normally, and according to the Linux spec, anonymous memory obtained
1519 from mmap() has it's contents cleared before it is passed to
1520 userspace. Enabling this config option allows you to request that
1521 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1522 providing a huge performance boost. If this option is not enabled,
1523 then the flag will be ignored.
1525 This is taken advantage of by uClibc's malloc(), and also by
1526 ELF-FDPIC binfmt's brk and stack allocator.
1528 Because of the obvious security issues, this option should only be
1529 enabled on embedded devices where you control what is run in
1530 userspace. Since that isn't generally a problem on no-MMU systems,
1531 it is normally safe to say Y here.
1533 See Documentation/nommu-mmap.txt for more information.
1536 bool "Profiling support"
1538 Say Y here to enable the extended profiling support mechanisms used
1539 by profilers such as OProfile.
1542 # Place an empty function call at each tracepoint site. Can be
1543 # dynamically changed for a probe function.
1548 source "arch/Kconfig"
1550 endmenu # General setup
1552 config HAVE_GENERIC_DMA_COHERENT
1559 depends on SLAB || SLUB_DEBUG
1567 default 0 if BASE_FULL
1568 default 1 if !BASE_FULL
1571 bool "Enable loadable module support"
1573 Kernel modules are small pieces of compiled code which can
1574 be inserted in the running kernel, rather than being
1575 permanently built into the kernel. You use the "modprobe"
1576 tool to add (and sometimes remove) them. If you say Y here,
1577 many parts of the kernel can be built as modules (by
1578 answering M instead of Y where indicated): this is most
1579 useful for infrequently used options which are not required
1580 for booting. For more information, see the man pages for
1581 modprobe, lsmod, modinfo, insmod and rmmod.
1583 If you say Y here, you will need to run "make
1584 modules_install" to put the modules under /lib/modules/
1585 where modprobe can find them (you may need to be root to do
1592 config MODULE_FORCE_LOAD
1593 bool "Forced module loading"
1596 Allow loading of modules without version information (ie. modprobe
1597 --force). Forced module loading sets the 'F' (forced) taint flag and
1598 is usually a really bad idea.
1600 config MODULE_UNLOAD
1601 bool "Module unloading"
1603 Without this option you will not be able to unload any
1604 modules (note that some modules may not be unloadable
1605 anyway), which makes your kernel smaller, faster
1606 and simpler. If unsure, say Y.
1608 config MODULE_FORCE_UNLOAD
1609 bool "Forced module unloading"
1610 depends on MODULE_UNLOAD
1612 This option allows you to force a module to unload, even if the
1613 kernel believes it is unsafe: the kernel will remove the module
1614 without waiting for anyone to stop using it (using the -f option to
1615 rmmod). This is mainly for kernel developers and desperate users.
1619 bool "Module versioning support"
1621 Usually, you have to use modules compiled with your kernel.
1622 Saying Y here makes it sometimes possible to use modules
1623 compiled for different kernels, by adding enough information
1624 to the modules to (hopefully) spot any changes which would
1625 make them incompatible with the kernel you are running. If
1628 config MODULE_SRCVERSION_ALL
1629 bool "Source checksum for all modules"
1631 Modules which contain a MODULE_VERSION get an extra "srcversion"
1632 field inserted into their modinfo section, which contains a
1633 sum of the source files which made it. This helps maintainers
1634 see exactly which source was used to build a module (since
1635 others sometimes change the module source without updating
1636 the version). With this option, such a "srcversion" field
1637 will be created for all modules. If unsure, say N.
1640 bool "Module signature verification"
1644 select ASYMMETRIC_KEY_TYPE
1645 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
1646 select PUBLIC_KEY_ALGO_RSA
1649 select X509_CERTIFICATE_PARSER
1651 Check modules for valid signatures upon load: the signature
1652 is simply appended to the module. For more information see
1653 Documentation/module-signing.txt.
1655 !!!WARNING!!! If you enable this option, you MUST make sure that the
1656 module DOES NOT get stripped after being signed. This includes the
1657 debuginfo strip done by some packagers (such as rpmbuild) and
1658 inclusion into an initramfs that wants the module size reduced.
1660 config MODULE_SIG_FORCE
1661 bool "Require modules to be validly signed"
1662 depends on MODULE_SIG
1664 Reject unsigned modules or signed modules for which we don't have a
1665 key. Without this, such modules will simply taint the kernel.
1667 config MODULE_SIG_ALL
1668 bool "Automatically sign all modules"
1670 depends on MODULE_SIG
1672 Sign all modules during make modules_install. Without this option,
1673 modules must be signed manually, using the scripts/sign-file tool.
1675 comment "Do not forget to sign required modules with scripts/sign-file"
1676 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL
1679 prompt "Which hash algorithm should modules be signed with?"
1680 depends on MODULE_SIG
1682 This determines which sort of hashing algorithm will be used during
1683 signature generation. This algorithm _must_ be built into the kernel
1684 directly so that signature verification can take place. It is not
1685 possible to load a signed module containing the algorithm to check
1686 the signature on that module.
1688 config MODULE_SIG_SHA1
1689 bool "Sign modules with SHA-1"
1692 config MODULE_SIG_SHA224
1693 bool "Sign modules with SHA-224"
1694 select CRYPTO_SHA256
1696 config MODULE_SIG_SHA256
1697 bool "Sign modules with SHA-256"
1698 select CRYPTO_SHA256
1700 config MODULE_SIG_SHA384
1701 bool "Sign modules with SHA-384"
1702 select CRYPTO_SHA512
1704 config MODULE_SIG_SHA512
1705 bool "Sign modules with SHA-512"
1706 select CRYPTO_SHA512
1710 config MODULE_SIG_HASH
1712 depends on MODULE_SIG
1713 default "sha1" if MODULE_SIG_SHA1
1714 default "sha224" if MODULE_SIG_SHA224
1715 default "sha256" if MODULE_SIG_SHA256
1716 default "sha384" if MODULE_SIG_SHA384
1717 default "sha512" if MODULE_SIG_SHA512
1721 config INIT_ALL_POSSIBLE
1724 Back when each arch used to define their own cpu_online_mask and
1725 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
1726 with all 1s, and others with all 0s. When they were centralised,
1727 it was better to provide this option than to break all the archs
1728 and have several arch maintainers pursuing me down dark alleys.
1733 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU
1735 Need stop_machine() primitive.
1737 source "block/Kconfig"
1739 config PREEMPT_NOTIFIERS
1746 # Can be selected by architectures with broken toolchains
1747 # that get confused by correct const<->read_only section
1749 config BROKEN_RODATA
1755 Build a simple ASN.1 grammar compiler that produces a bytecode output
1756 that can be interpreted by the ASN.1 stream decoder and used to
1757 inform it as to what tags are to be expected in a stream and what
1758 functions to call on what tags.
1760 source "kernel/Kconfig.locks"
projects / deliverable / linux.git / blob