]> git.proxmox.com Git - mirror_ubuntu-bionic-kernel.git/blob - init/Kconfig
perf: Add cgroup support
[mirror_ubuntu-bionic-kernel.git] / init / Kconfig
1 config ARCH
2 string
3 option env="ARCH"
4
5 config KERNELVERSION
6 string
7 option env="KERNELVERSION"
8
9 config DEFCONFIG_LIST
10 string
11 depends on !UML
12 option defconfig_list
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"
18
19 config CONSTRUCTORS
20 bool
21 depends on !UML
22 default y
23
24 config HAVE_IRQ_WORK
25 bool
26
27 config IRQ_WORK
28 bool
29 depends on HAVE_IRQ_WORK
30
31 menu "General setup"
32
33 config EXPERIMENTAL
34 bool "Prompt for development and/or incomplete code/drivers"
35 ---help---
36 Some of the various things that Linux supports (such as network
37 drivers, file systems, network protocols, etc.) can be in a state
38 of development where the functionality, stability, or the level of
39 testing is not yet high enough for general use. This is usually
40 known as the "alpha-test" phase among developers. If a feature is
41 currently in alpha-test, then the developers usually discourage
42 uninformed widespread use of this feature by the general public to
43 avoid "Why doesn't this work?" type mail messages. However, active
44 testing and use of these systems is welcomed. Just be aware that it
45 may not meet the normal level of reliability or it may fail to work
46 in some special cases. Detailed bug reports from people familiar
47 with the kernel internals are usually welcomed by the developers
48 (before submitting bug reports, please read the documents
49 <file:README>, <file:MAINTAINERS>, <file:REPORTING-BUGS>,
50 <file:Documentation/BUG-HUNTING>, and
51 <file:Documentation/oops-tracing.txt> in the kernel source).
52
53 This option will also make obsoleted drivers available. These are
54 drivers that have been replaced by something else, and/or are
55 scheduled to be removed in a future kernel release.
56
57 Unless you intend to help test and develop a feature or driver that
58 falls into this category, or you have a situation that requires
59 using these features, you should probably say N here, which will
60 cause the configurator to present you with fewer choices. If
61 you say Y here, you will be offered the choice of using features or
62 drivers that are currently considered to be in the alpha-test phase.
63
64 config BROKEN
65 bool
66
67 config BROKEN_ON_SMP
68 bool
69 depends on BROKEN || !SMP
70 default y
71
72 config LOCK_KERNEL
73 bool
74 depends on (SMP || PREEMPT) && BKL
75 default y
76
77 config INIT_ENV_ARG_LIMIT
78 int
79 default 32 if !UML
80 default 128 if UML
81 help
82 Maximum of each of the number of arguments and environment
83 variables passed to init from the kernel command line.
84
85
86 config CROSS_COMPILE
87 string "Cross-compiler tool prefix"
88 help
89 Same as running 'make CROSS_COMPILE=prefix-' but stored for
90 default make runs in this kernel build directory. You don't
91 need to set this unless you want the configured kernel build
92 directory to select the cross-compiler automatically.
93
94 config LOCALVERSION
95 string "Local version - append to kernel release"
96 help
97 Append an extra string to the end of your kernel version.
98 This will show up when you type uname, for example.
99 The string you set here will be appended after the contents of
100 any files with a filename matching localversion* in your
101 object and source tree, in that order. Your total string can
102 be a maximum of 64 characters.
103
104 config LOCALVERSION_AUTO
105 bool "Automatically append version information to the version string"
106 default y
107 help
108 This will try to automatically determine if the current tree is a
109 release tree by looking for git tags that belong to the current
110 top of tree revision.
111
112 A string of the format -gxxxxxxxx will be added to the localversion
113 if a git-based tree is found. The string generated by this will be
114 appended after any matching localversion* files, and after the value
115 set in CONFIG_LOCALVERSION.
116
117 (The actual string used here is the first eight characters produced
118 by running the command:
119
120 $ git rev-parse --verify HEAD
121
122 which is done within the script "scripts/setlocalversion".)
123
124 config HAVE_KERNEL_GZIP
125 bool
126
127 config HAVE_KERNEL_BZIP2
128 bool
129
130 config HAVE_KERNEL_LZMA
131 bool
132
133 config HAVE_KERNEL_XZ
134 bool
135
136 config HAVE_KERNEL_LZO
137 bool
138
139 choice
140 prompt "Kernel compression mode"
141 default KERNEL_GZIP
142 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO
143 help
144 The linux kernel is a kind of self-extracting executable.
145 Several compression algorithms are available, which differ
146 in efficiency, compression and decompression speed.
147 Compression speed is only relevant when building a kernel.
148 Decompression speed is relevant at each boot.
149
150 If you have any problems with bzip2 or lzma compressed
151 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
152 version of this functionality (bzip2 only), for 2.4, was
153 supplied by Christian Ludwig)
154
155 High compression options are mostly useful for users, who
156 are low on disk space (embedded systems), but for whom ram
157 size matters less.
158
159 If in doubt, select 'gzip'
160
161 config KERNEL_GZIP
162 bool "Gzip"
163 depends on HAVE_KERNEL_GZIP
164 help
165 The old and tried gzip compression. It provides a good balance
166 between compression ratio and decompression speed.
167
168 config KERNEL_BZIP2
169 bool "Bzip2"
170 depends on HAVE_KERNEL_BZIP2
171 help
172 Its compression ratio and speed is intermediate.
173 Decompression speed is slowest among the three. The kernel
174 size is about 10% smaller with bzip2, in comparison to gzip.
175 Bzip2 uses a large amount of memory. For modern kernels you
176 will need at least 8MB RAM or more for booting.
177
178 config KERNEL_LZMA
179 bool "LZMA"
180 depends on HAVE_KERNEL_LZMA
181 help
182 The most recent compression algorithm.
183 Its ratio is best, decompression speed is between the other
184 two. Compression is slowest. The kernel size is about 33%
185 smaller with LZMA in comparison to gzip.
186
187 config KERNEL_XZ
188 bool "XZ"
189 depends on HAVE_KERNEL_XZ
190 help
191 XZ uses the LZMA2 algorithm and instruction set specific
192 BCJ filters which can improve compression ratio of executable
193 code. The size of the kernel is about 30% smaller with XZ in
194 comparison to gzip. On architectures for which there is a BCJ
195 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
196 will create a few percent smaller kernel than plain LZMA.
197
198 The speed is about the same as with LZMA: The decompression
199 speed of XZ is better than that of bzip2 but worse than gzip
200 and LZO. Compression is slow.
201
202 config KERNEL_LZO
203 bool "LZO"
204 depends on HAVE_KERNEL_LZO
205 help
206 Its compression ratio is the poorest among the 4. The kernel
207 size is about 10% bigger than gzip; however its speed
208 (both compression and decompression) is the fastest.
209
210 endchoice
211
212 config SWAP
213 bool "Support for paging of anonymous memory (swap)"
214 depends on MMU && BLOCK
215 default y
216 help
217 This option allows you to choose whether you want to have support
218 for so called swap devices or swap files in your kernel that are
219 used to provide more virtual memory than the actual RAM present
220 in your computer. If unsure say Y.
221
222 config SYSVIPC
223 bool "System V IPC"
224 ---help---
225 Inter Process Communication is a suite of library functions and
226 system calls which let processes (running programs) synchronize and
227 exchange information. It is generally considered to be a good thing,
228 and some programs won't run unless you say Y here. In particular, if
229 you want to run the DOS emulator dosemu under Linux (read the
230 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
231 you'll need to say Y here.
232
233 You can find documentation about IPC with "info ipc" and also in
234 section 6.4 of the Linux Programmer's Guide, available from
235 <http://www.tldp.org/guides.html>.
236
237 config SYSVIPC_SYSCTL
238 bool
239 depends on SYSVIPC
240 depends on SYSCTL
241 default y
242
243 config POSIX_MQUEUE
244 bool "POSIX Message Queues"
245 depends on NET && EXPERIMENTAL
246 ---help---
247 POSIX variant of message queues is a part of IPC. In POSIX message
248 queues every message has a priority which decides about succession
249 of receiving it by a process. If you want to compile and run
250 programs written e.g. for Solaris with use of its POSIX message
251 queues (functions mq_*) say Y here.
252
253 POSIX message queues are visible as a filesystem called 'mqueue'
254 and can be mounted somewhere if you want to do filesystem
255 operations on message queues.
256
257 If unsure, say Y.
258
259 config POSIX_MQUEUE_SYSCTL
260 bool
261 depends on POSIX_MQUEUE
262 depends on SYSCTL
263 default y
264
265 config BSD_PROCESS_ACCT
266 bool "BSD Process Accounting"
267 help
268 If you say Y here, a user level program will be able to instruct the
269 kernel (via a special system call) to write process accounting
270 information to a file: whenever a process exits, information about
271 that process will be appended to the file by the kernel. The
272 information includes things such as creation time, owning user,
273 command name, memory usage, controlling terminal etc. (the complete
274 list is in the struct acct in <file:include/linux/acct.h>). It is
275 up to the user level program to do useful things with this
276 information. This is generally a good idea, so say Y.
277
278 config BSD_PROCESS_ACCT_V3
279 bool "BSD Process Accounting version 3 file format"
280 depends on BSD_PROCESS_ACCT
281 default n
282 help
283 If you say Y here, the process accounting information is written
284 in a new file format that also logs the process IDs of each
285 process and it's parent. Note that this file format is incompatible
286 with previous v0/v1/v2 file formats, so you will need updated tools
287 for processing it. A preliminary version of these tools is available
288 at <http://www.gnu.org/software/acct/>.
289
290 config TASKSTATS
291 bool "Export task/process statistics through netlink (EXPERIMENTAL)"
292 depends on NET
293 default n
294 help
295 Export selected statistics for tasks/processes through the
296 generic netlink interface. Unlike BSD process accounting, the
297 statistics are available during the lifetime of tasks/processes as
298 responses to commands. Like BSD accounting, they are sent to user
299 space on task exit.
300
301 Say N if unsure.
302
303 config TASK_DELAY_ACCT
304 bool "Enable per-task delay accounting (EXPERIMENTAL)"
305 depends on TASKSTATS
306 help
307 Collect information on time spent by a task waiting for system
308 resources like cpu, synchronous block I/O completion and swapping
309 in pages. Such statistics can help in setting a task's priorities
310 relative to other tasks for cpu, io, rss limits etc.
311
312 Say N if unsure.
313
314 config TASK_XACCT
315 bool "Enable extended accounting over taskstats (EXPERIMENTAL)"
316 depends on TASKSTATS
317 help
318 Collect extended task accounting data and send the data
319 to userland for processing over the taskstats interface.
320
321 Say N if unsure.
322
323 config TASK_IO_ACCOUNTING
324 bool "Enable per-task storage I/O accounting (EXPERIMENTAL)"
325 depends on TASK_XACCT
326 help
327 Collect information on the number of bytes of storage I/O which this
328 task has caused.
329
330 Say N if unsure.
331
332 config AUDIT
333 bool "Auditing support"
334 depends on NET
335 help
336 Enable auditing infrastructure that can be used with another
337 kernel subsystem, such as SELinux (which requires this for
338 logging of avc messages output). Does not do system-call
339 auditing without CONFIG_AUDITSYSCALL.
340
341 config AUDITSYSCALL
342 bool "Enable system-call auditing support"
343 depends on AUDIT && (X86 || PPC || S390 || IA64 || UML || SPARC64 || SUPERH)
344 default y if SECURITY_SELINUX
345 help
346 Enable low-overhead system-call auditing infrastructure that
347 can be used independently or with another kernel subsystem,
348 such as SELinux.
349
350 config AUDIT_WATCH
351 def_bool y
352 depends on AUDITSYSCALL
353 select FSNOTIFY
354
355 config AUDIT_TREE
356 def_bool y
357 depends on AUDITSYSCALL
358 select FSNOTIFY
359
360 source "kernel/irq/Kconfig"
361
362 menu "RCU Subsystem"
363
364 choice
365 prompt "RCU Implementation"
366 default TREE_RCU
367
368 config TREE_RCU
369 bool "Tree-based hierarchical RCU"
370 depends on !PREEMPT && SMP
371 help
372 This option selects the RCU implementation that is
373 designed for very large SMP system with hundreds or
374 thousands of CPUs. It also scales down nicely to
375 smaller systems.
376
377 config TREE_PREEMPT_RCU
378 bool "Preemptible tree-based hierarchical RCU"
379 depends on PREEMPT
380 help
381 This option selects the RCU implementation that is
382 designed for very large SMP systems with hundreds or
383 thousands of CPUs, but for which real-time response
384 is also required. It also scales down nicely to
385 smaller systems.
386
387 config TINY_RCU
388 bool "UP-only small-memory-footprint RCU"
389 depends on !SMP
390 help
391 This option selects the RCU implementation that is
392 designed for UP systems from which real-time response
393 is not required. This option greatly reduces the
394 memory footprint of RCU.
395
396 config TINY_PREEMPT_RCU
397 bool "Preemptible UP-only small-memory-footprint RCU"
398 depends on !SMP && PREEMPT
399 help
400 This option selects the RCU implementation that is designed
401 for real-time UP systems. This option greatly reduces the
402 memory footprint of RCU.
403
404 endchoice
405
406 config PREEMPT_RCU
407 def_bool ( TREE_PREEMPT_RCU || TINY_PREEMPT_RCU )
408 help
409 This option enables preemptible-RCU code that is common between
410 the TREE_PREEMPT_RCU and TINY_PREEMPT_RCU implementations.
411
412 config RCU_TRACE
413 bool "Enable tracing for RCU"
414 help
415 This option provides tracing in RCU which presents stats
416 in debugfs for debugging RCU implementation.
417
418 Say Y here if you want to enable RCU tracing
419 Say N if you are unsure.
420
421 config RCU_FANOUT
422 int "Tree-based hierarchical RCU fanout value"
423 range 2 64 if 64BIT
424 range 2 32 if !64BIT
425 depends on TREE_RCU || TREE_PREEMPT_RCU
426 default 64 if 64BIT
427 default 32 if !64BIT
428 help
429 This option controls the fanout of hierarchical implementations
430 of RCU, allowing RCU to work efficiently on machines with
431 large numbers of CPUs. This value must be at least the fourth
432 root of NR_CPUS, which allows NR_CPUS to be insanely large.
433 The default value of RCU_FANOUT should be used for production
434 systems, but if you are stress-testing the RCU implementation
435 itself, small RCU_FANOUT values allow you to test large-system
436 code paths on small(er) systems.
437
438 Select a specific number if testing RCU itself.
439 Take the default if unsure.
440
441 config RCU_FANOUT_EXACT
442 bool "Disable tree-based hierarchical RCU auto-balancing"
443 depends on TREE_RCU || TREE_PREEMPT_RCU
444 default n
445 help
446 This option forces use of the exact RCU_FANOUT value specified,
447 regardless of imbalances in the hierarchy. This is useful for
448 testing RCU itself, and might one day be useful on systems with
449 strong NUMA behavior.
450
451 Without RCU_FANOUT_EXACT, the code will balance the hierarchy.
452
453 Say N if unsure.
454
455 config RCU_FAST_NO_HZ
456 bool "Accelerate last non-dyntick-idle CPU's grace periods"
457 depends on TREE_RCU && NO_HZ && SMP
458 default n
459 help
460 This option causes RCU to attempt to accelerate grace periods
461 in order to allow the final CPU to enter dynticks-idle state
462 more quickly. On the other hand, this option increases the
463 overhead of the dynticks-idle checking, particularly on systems
464 with large numbers of CPUs.
465
466 Say Y if energy efficiency is critically important, particularly
467 if you have relatively few CPUs.
468
469 Say N if you are unsure.
470
471 config TREE_RCU_TRACE
472 def_bool RCU_TRACE && ( TREE_RCU || TREE_PREEMPT_RCU )
473 select DEBUG_FS
474 help
475 This option provides tracing for the TREE_RCU and
476 TREE_PREEMPT_RCU implementations, permitting Makefile to
477 trivially select kernel/rcutree_trace.c.
478
479 config RCU_BOOST
480 bool "Enable RCU priority boosting"
481 depends on RT_MUTEXES && TINY_PREEMPT_RCU
482 default n
483 help
484 This option boosts the priority of preempted RCU readers that
485 block the current preemptible RCU grace period for too long.
486 This option also prevents heavy loads from blocking RCU
487 callback invocation for all flavors of RCU.
488
489 Say Y here if you are working with real-time apps or heavy loads
490 Say N here if you are unsure.
491
492 config RCU_BOOST_PRIO
493 int "Real-time priority to boost RCU readers to"
494 range 1 99
495 depends on RCU_BOOST
496 default 1
497 help
498 This option specifies the real-time priority to which preempted
499 RCU readers are to be boosted. If you are working with CPU-bound
500 real-time applications, you should specify a priority higher then
501 the highest-priority CPU-bound application.
502
503 Specify the real-time priority, or take the default if unsure.
504
505 config RCU_BOOST_DELAY
506 int "Milliseconds to delay boosting after RCU grace-period start"
507 range 0 3000
508 depends on RCU_BOOST
509 default 500
510 help
511 This option specifies the time to wait after the beginning of
512 a given grace period before priority-boosting preempted RCU
513 readers blocking that grace period. Note that any RCU reader
514 blocking an expedited RCU grace period is boosted immediately.
515
516 Accept the default if unsure.
517
518 endmenu # "RCU Subsystem"
519
520 config IKCONFIG
521 tristate "Kernel .config support"
522 ---help---
523 This option enables the complete Linux kernel ".config" file
524 contents to be saved in the kernel. It provides documentation
525 of which kernel options are used in a running kernel or in an
526 on-disk kernel. This information can be extracted from the kernel
527 image file with the script scripts/extract-ikconfig and used as
528 input to rebuild the current kernel or to build another kernel.
529 It can also be extracted from a running kernel by reading
530 /proc/config.gz if enabled (below).
531
532 config IKCONFIG_PROC
533 bool "Enable access to .config through /proc/config.gz"
534 depends on IKCONFIG && PROC_FS
535 ---help---
536 This option enables access to the kernel configuration file
537 through /proc/config.gz.
538
539 config LOG_BUF_SHIFT
540 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
541 range 12 21
542 default 17
543 help
544 Select kernel log buffer size as a power of 2.
545 Examples:
546 17 => 128 KB
547 16 => 64 KB
548 15 => 32 KB
549 14 => 16 KB
550 13 => 8 KB
551 12 => 4 KB
552
553 #
554 # Architectures with an unreliable sched_clock() should select this:
555 #
556 config HAVE_UNSTABLE_SCHED_CLOCK
557 bool
558
559 menuconfig CGROUPS
560 boolean "Control Group support"
561 depends on EVENTFD
562 help
563 This option adds support for grouping sets of processes together, for
564 use with process control subsystems such as Cpusets, CFS, memory
565 controls or device isolation.
566 See
567 - Documentation/scheduler/sched-design-CFS.txt (CFS)
568 - Documentation/cgroups/ (features for grouping, isolation
569 and resource control)
570
571 Say N if unsure.
572
573 if CGROUPS
574
575 config CGROUP_DEBUG
576 bool "Example debug cgroup subsystem"
577 default n
578 help
579 This option enables a simple cgroup subsystem that
580 exports useful debugging information about the cgroups
581 framework.
582
583 Say N if unsure.
584
585 config CGROUP_NS
586 bool "Namespace cgroup subsystem"
587 help
588 Provides a simple namespace cgroup subsystem to
589 provide hierarchical naming of sets of namespaces,
590 for instance virtual servers and checkpoint/restart
591 jobs.
592
593 config CGROUP_FREEZER
594 bool "Freezer cgroup subsystem"
595 help
596 Provides a way to freeze and unfreeze all tasks in a
597 cgroup.
598
599 config CGROUP_DEVICE
600 bool "Device controller for cgroups"
601 help
602 Provides a cgroup implementing whitelists for devices which
603 a process in the cgroup can mknod or open.
604
605 config CPUSETS
606 bool "Cpuset support"
607 help
608 This option will let you create and manage CPUSETs which
609 allow dynamically partitioning a system into sets of CPUs and
610 Memory Nodes and assigning tasks to run only within those sets.
611 This is primarily useful on large SMP or NUMA systems.
612
613 Say N if unsure.
614
615 config PROC_PID_CPUSET
616 bool "Include legacy /proc/<pid>/cpuset file"
617 depends on CPUSETS
618 default y
619
620 config CGROUP_CPUACCT
621 bool "Simple CPU accounting cgroup subsystem"
622 help
623 Provides a simple Resource Controller for monitoring the
624 total CPU consumed by the tasks in a cgroup.
625
626 config RESOURCE_COUNTERS
627 bool "Resource counters"
628 help
629 This option enables controller independent resource accounting
630 infrastructure that works with cgroups.
631
632 config CGROUP_MEM_RES_CTLR
633 bool "Memory Resource Controller for Control Groups"
634 depends on RESOURCE_COUNTERS
635 select MM_OWNER
636 help
637 Provides a memory resource controller that manages both anonymous
638 memory and page cache. (See Documentation/cgroups/memory.txt)
639
640 Note that setting this option increases fixed memory overhead
641 associated with each page of memory in the system. By this,
642 20(40)bytes/PAGE_SIZE on 32(64)bit system will be occupied by memory
643 usage tracking struct at boot. Total amount of this is printed out
644 at boot.
645
646 Only enable when you're ok with these trade offs and really
647 sure you need the memory resource controller. Even when you enable
648 this, you can set "cgroup_disable=memory" at your boot option to
649 disable memory resource controller and you can avoid overheads.
650 (and lose benefits of memory resource controller)
651
652 This config option also selects MM_OWNER config option, which
653 could in turn add some fork/exit overhead.
654
655 config CGROUP_MEM_RES_CTLR_SWAP
656 bool "Memory Resource Controller Swap Extension"
657 depends on CGROUP_MEM_RES_CTLR && SWAP
658 help
659 Add swap management feature to memory resource controller. When you
660 enable this, you can limit mem+swap usage per cgroup. In other words,
661 when you disable this, memory resource controller has no cares to
662 usage of swap...a process can exhaust all of the swap. This extension
663 is useful when you want to avoid exhaustion swap but this itself
664 adds more overheads and consumes memory for remembering information.
665 Especially if you use 32bit system or small memory system, please
666 be careful about enabling this. When memory resource controller
667 is disabled by boot option, this will be automatically disabled and
668 there will be no overhead from this. Even when you set this config=y,
669 if boot option "noswapaccount" is set, swap will not be accounted.
670 Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page
671 size is 4096bytes, 512k per 1Gbytes of swap.
672 config CGROUP_MEM_RES_CTLR_SWAP_ENABLED
673 bool "Memory Resource Controller Swap Extension enabled by default"
674 depends on CGROUP_MEM_RES_CTLR_SWAP
675 default y
676 help
677 Memory Resource Controller Swap Extension comes with its price in
678 a bigger memory consumption. General purpose distribution kernels
679 which want to enable the feature but keep it disabled by default
680 and let the user enable it by swapaccount boot command line
681 parameter should have this option unselected.
682 For those who want to have the feature enabled by default should
683 select this option (if, for some reason, they need to disable it
684 then noswapaccount does the trick).
685
686 config CGROUP_PERF
687 bool "Enable perf_event per-cpu per-container group (cgroup) monitoring"
688 depends on PERF_EVENTS && CGROUPS
689 help
690 This option extends the per-cpu mode to restrict monitoring to
691 threads which belong to the cgroup specificied and run on the
692 designated cpu.
693
694 Say N if unsure.
695
696 menuconfig CGROUP_SCHED
697 bool "Group CPU scheduler"
698 depends on EXPERIMENTAL
699 default n
700 help
701 This feature lets CPU scheduler recognize task groups and control CPU
702 bandwidth allocation to such task groups. It uses cgroups to group
703 tasks.
704
705 if CGROUP_SCHED
706 config FAIR_GROUP_SCHED
707 bool "Group scheduling for SCHED_OTHER"
708 depends on CGROUP_SCHED
709 default CGROUP_SCHED
710
711 config RT_GROUP_SCHED
712 bool "Group scheduling for SCHED_RR/FIFO"
713 depends on EXPERIMENTAL
714 depends on CGROUP_SCHED
715 default n
716 help
717 This feature lets you explicitly allocate real CPU bandwidth
718 to task groups. If enabled, it will also make it impossible to
719 schedule realtime tasks for non-root users until you allocate
720 realtime bandwidth for them.
721 See Documentation/scheduler/sched-rt-group.txt for more information.
722
723 endif #CGROUP_SCHED
724
725 config BLK_CGROUP
726 tristate "Block IO controller"
727 depends on BLOCK
728 default n
729 ---help---
730 Generic block IO controller cgroup interface. This is the common
731 cgroup interface which should be used by various IO controlling
732 policies.
733
734 Currently, CFQ IO scheduler uses it to recognize task groups and
735 control disk bandwidth allocation (proportional time slice allocation)
736 to such task groups. It is also used by bio throttling logic in
737 block layer to implement upper limit in IO rates on a device.
738
739 This option only enables generic Block IO controller infrastructure.
740 One needs to also enable actual IO controlling logic/policy. For
741 enabling proportional weight division of disk bandwidth in CFQ seti
742 CONFIG_CFQ_GROUP_IOSCHED=y and for enabling throttling policy set
743 CONFIG_BLK_THROTTLE=y.
744
745 See Documentation/cgroups/blkio-controller.txt for more information.
746
747 config DEBUG_BLK_CGROUP
748 bool "Enable Block IO controller debugging"
749 depends on BLK_CGROUP
750 default n
751 ---help---
752 Enable some debugging help. Currently it exports additional stat
753 files in a cgroup which can be useful for debugging.
754
755 endif # CGROUPS
756
757 menuconfig NAMESPACES
758 bool "Namespaces support" if EXPERT
759 default !EXPERT
760 help
761 Provides the way to make tasks work with different objects using
762 the same id. For example same IPC id may refer to different objects
763 or same user id or pid may refer to different tasks when used in
764 different namespaces.
765
766 if NAMESPACES
767
768 config UTS_NS
769 bool "UTS namespace"
770 default y
771 help
772 In this namespace tasks see different info provided with the
773 uname() system call
774
775 config IPC_NS
776 bool "IPC namespace"
777 depends on (SYSVIPC || POSIX_MQUEUE)
778 default y
779 help
780 In this namespace tasks work with IPC ids which correspond to
781 different IPC objects in different namespaces.
782
783 config USER_NS
784 bool "User namespace (EXPERIMENTAL)"
785 depends on EXPERIMENTAL
786 default y
787 help
788 This allows containers, i.e. vservers, to use user namespaces
789 to provide different user info for different servers.
790 If unsure, say N.
791
792 config PID_NS
793 bool "PID Namespaces"
794 default y
795 help
796 Support process id namespaces. This allows having multiple
797 processes with the same pid as long as they are in different
798 pid namespaces. This is a building block of containers.
799
800 config NET_NS
801 bool "Network namespace"
802 depends on NET
803 default y
804 help
805 Allow user space to create what appear to be multiple instances
806 of the network stack.
807
808 endif # NAMESPACES
809
810 config SCHED_AUTOGROUP
811 bool "Automatic process group scheduling"
812 select EVENTFD
813 select CGROUPS
814 select CGROUP_SCHED
815 select FAIR_GROUP_SCHED
816 help
817 This option optimizes the scheduler for common desktop workloads by
818 automatically creating and populating task groups. This separation
819 of workloads isolates aggressive CPU burners (like build jobs) from
820 desktop applications. Task group autogeneration is currently based
821 upon task session.
822
823 config MM_OWNER
824 bool
825
826 config SYSFS_DEPRECATED
827 bool "enable deprecated sysfs features to support old userspace tools"
828 depends on SYSFS
829 default n
830 help
831 This option adds code that switches the layout of the "block" class
832 devices, to not show up in /sys/class/block/, but only in
833 /sys/block/.
834
835 This switch is only active when the sysfs.deprecated=1 boot option is
836 passed or the SYSFS_DEPRECATED_V2 option is set.
837
838 This option allows new kernels to run on old distributions and tools,
839 which might get confused by /sys/class/block/. Since 2007/2008 all
840 major distributions and tools handle this just fine.
841
842 Recent distributions and userspace tools after 2009/2010 depend on
843 the existence of /sys/class/block/, and will not work with this
844 option enabled.
845
846 Only if you are using a new kernel on an old distribution, you might
847 need to say Y here.
848
849 config SYSFS_DEPRECATED_V2
850 bool "enabled deprecated sysfs features by default"
851 default n
852 depends on SYSFS
853 depends on SYSFS_DEPRECATED
854 help
855 Enable deprecated sysfs by default.
856
857 See the CONFIG_SYSFS_DEPRECATED option for more details about this
858 option.
859
860 Only if you are using a new kernel on an old distribution, you might
861 need to say Y here. Even then, odds are you would not need it
862 enabled, you can always pass the boot option if absolutely necessary.
863
864 config RELAY
865 bool "Kernel->user space relay support (formerly relayfs)"
866 help
867 This option enables support for relay interface support in
868 certain file systems (such as debugfs).
869 It is designed to provide an efficient mechanism for tools and
870 facilities to relay large amounts of data from kernel space to
871 user space.
872
873 If unsure, say N.
874
875 config BLK_DEV_INITRD
876 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
877 depends on BROKEN || !FRV
878 help
879 The initial RAM filesystem is a ramfs which is loaded by the
880 boot loader (loadlin or lilo) and that is mounted as root
881 before the normal boot procedure. It is typically used to
882 load modules needed to mount the "real" root file system,
883 etc. See <file:Documentation/initrd.txt> for details.
884
885 If RAM disk support (BLK_DEV_RAM) is also included, this
886 also enables initial RAM disk (initrd) support and adds
887 15 Kbytes (more on some other architectures) to the kernel size.
888
889 If unsure say Y.
890
891 if BLK_DEV_INITRD
892
893 source "usr/Kconfig"
894
895 endif
896
897 config CC_OPTIMIZE_FOR_SIZE
898 bool "Optimize for size"
899 default y
900 help
901 Enabling this option will pass "-Os" instead of "-O2" to gcc
902 resulting in a smaller kernel.
903
904 If unsure, say Y.
905
906 config SYSCTL
907 bool
908
909 config ANON_INODES
910 bool
911
912 menuconfig EXPERT
913 bool "Configure standard kernel features (expert users)"
914 help
915 This option allows certain base kernel options and settings
916 to be disabled or tweaked. This is for specialized
917 environments which can tolerate a "non-standard" kernel.
918 Only use this if you really know what you are doing.
919
920 config EMBEDDED
921 bool "Embedded system"
922 select EXPERT
923 help
924 This option should be enabled if compiling the kernel for
925 an embedded system so certain expert options are available
926 for configuration.
927
928 config UID16
929 bool "Enable 16-bit UID system calls" if EXPERT
930 depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && COMPAT) || UML || (X86_64 && IA32_EMULATION)
931 default y
932 help
933 This enables the legacy 16-bit UID syscall wrappers.
934
935 config SYSCTL_SYSCALL
936 bool "Sysctl syscall support" if EXPERT
937 depends on PROC_SYSCTL
938 default y
939 select SYSCTL
940 ---help---
941 sys_sysctl uses binary paths that have been found challenging
942 to properly maintain and use. The interface in /proc/sys
943 using paths with ascii names is now the primary path to this
944 information.
945
946 Almost nothing using the binary sysctl interface so if you are
947 trying to save some space it is probably safe to disable this,
948 making your kernel marginally smaller.
949
950 If unsure say Y here.
951
952 config KALLSYMS
953 bool "Load all symbols for debugging/ksymoops" if EXPERT
954 default y
955 help
956 Say Y here to let the kernel print out symbolic crash information and
957 symbolic stack backtraces. This increases the size of the kernel
958 somewhat, as all symbols have to be loaded into the kernel image.
959
960 config KALLSYMS_ALL
961 bool "Include all symbols in kallsyms"
962 depends on DEBUG_KERNEL && KALLSYMS
963 help
964 Normally kallsyms only contains the symbols of functions, for nicer
965 OOPS messages. Some debuggers can use kallsyms for other
966 symbols too: say Y here to include all symbols, if you need them
967 and you don't care about adding 300k to the size of your kernel.
968
969 Say N.
970
971 config KALLSYMS_EXTRA_PASS
972 bool "Do an extra kallsyms pass"
973 depends on KALLSYMS
974 help
975 If kallsyms is not working correctly, the build will fail with
976 inconsistent kallsyms data. If that occurs, log a bug report and
977 turn on KALLSYMS_EXTRA_PASS which should result in a stable build.
978 Always say N here unless you find a bug in kallsyms, which must be
979 reported. KALLSYMS_EXTRA_PASS is only a temporary workaround while
980 you wait for kallsyms to be fixed.
981
982
983 config HOTPLUG
984 bool "Support for hot-pluggable devices" if EXPERT
985 default y
986 help
987 This option is provided for the case where no hotplug or uevent
988 capabilities is wanted by the kernel. You should only consider
989 disabling this option for embedded systems that do not use modules, a
990 dynamic /dev tree, or dynamic device discovery. Just say Y.
991
992 config PRINTK
993 default y
994 bool "Enable support for printk" if EXPERT
995 help
996 This option enables normal printk support. Removing it
997 eliminates most of the message strings from the kernel image
998 and makes the kernel more or less silent. As this makes it
999 very difficult to diagnose system problems, saying N here is
1000 strongly discouraged.
1001
1002 config BUG
1003 bool "BUG() support" if EXPERT
1004 default y
1005 help
1006 Disabling this option eliminates support for BUG and WARN, reducing
1007 the size of your kernel image and potentially quietly ignoring
1008 numerous fatal conditions. You should only consider disabling this
1009 option for embedded systems with no facilities for reporting errors.
1010 Just say Y.
1011
1012 config ELF_CORE
1013 default y
1014 bool "Enable ELF core dumps" if EXPERT
1015 help
1016 Enable support for generating core dumps. Disabling saves about 4k.
1017
1018 config PCSPKR_PLATFORM
1019 bool "Enable PC-Speaker support" if EXPERT
1020 depends on ALPHA || X86 || MIPS || PPC_PREP || PPC_CHRP || PPC_PSERIES
1021 default y
1022 help
1023 This option allows to disable the internal PC-Speaker
1024 support, saving some memory.
1025
1026 config BASE_FULL
1027 default y
1028 bool "Enable full-sized data structures for core" if EXPERT
1029 help
1030 Disabling this option reduces the size of miscellaneous core
1031 kernel data structures. This saves memory on small machines,
1032 but may reduce performance.
1033
1034 config FUTEX
1035 bool "Enable futex support" if EXPERT
1036 default y
1037 select RT_MUTEXES
1038 help
1039 Disabling this option will cause the kernel to be built without
1040 support for "fast userspace mutexes". The resulting kernel may not
1041 run glibc-based applications correctly.
1042
1043 config EPOLL
1044 bool "Enable eventpoll support" if EXPERT
1045 default y
1046 select ANON_INODES
1047 help
1048 Disabling this option will cause the kernel to be built without
1049 support for epoll family of system calls.
1050
1051 config SIGNALFD
1052 bool "Enable signalfd() system call" if EXPERT
1053 select ANON_INODES
1054 default y
1055 help
1056 Enable the signalfd() system call that allows to receive signals
1057 on a file descriptor.
1058
1059 If unsure, say Y.
1060
1061 config TIMERFD
1062 bool "Enable timerfd() system call" if EXPERT
1063 select ANON_INODES
1064 default y
1065 help
1066 Enable the timerfd() system call that allows to receive timer
1067 events on a file descriptor.
1068
1069 If unsure, say Y.
1070
1071 config EVENTFD
1072 bool "Enable eventfd() system call" if EXPERT
1073 select ANON_INODES
1074 default y
1075 help
1076 Enable the eventfd() system call that allows to receive both
1077 kernel notification (ie. KAIO) or userspace notifications.
1078
1079 If unsure, say Y.
1080
1081 config SHMEM
1082 bool "Use full shmem filesystem" if EXPERT
1083 default y
1084 depends on MMU
1085 help
1086 The shmem is an internal filesystem used to manage shared memory.
1087 It is backed by swap and manages resource limits. It is also exported
1088 to userspace as tmpfs if TMPFS is enabled. Disabling this
1089 option replaces shmem and tmpfs with the much simpler ramfs code,
1090 which may be appropriate on small systems without swap.
1091
1092 config AIO
1093 bool "Enable AIO support" if EXPERT
1094 default y
1095 help
1096 This option enables POSIX asynchronous I/O which may by used
1097 by some high performance threaded applications. Disabling
1098 this option saves about 7k.
1099
1100 config HAVE_PERF_EVENTS
1101 bool
1102 help
1103 See tools/perf/design.txt for details.
1104
1105 config PERF_USE_VMALLOC
1106 bool
1107 help
1108 See tools/perf/design.txt for details
1109
1110 menu "Kernel Performance Events And Counters"
1111
1112 config PERF_EVENTS
1113 bool "Kernel performance events and counters"
1114 default y if (PROFILING || PERF_COUNTERS)
1115 depends on HAVE_PERF_EVENTS
1116 select ANON_INODES
1117 select IRQ_WORK
1118 help
1119 Enable kernel support for various performance events provided
1120 by software and hardware.
1121
1122 Software events are supported either built-in or via the
1123 use of generic tracepoints.
1124
1125 Most modern CPUs support performance events via performance
1126 counter registers. These registers count the number of certain
1127 types of hw events: such as instructions executed, cachemisses
1128 suffered, or branches mis-predicted - without slowing down the
1129 kernel or applications. These registers can also trigger interrupts
1130 when a threshold number of events have passed - and can thus be
1131 used to profile the code that runs on that CPU.
1132
1133 The Linux Performance Event subsystem provides an abstraction of
1134 these software and hardware event capabilities, available via a
1135 system call and used by the "perf" utility in tools/perf/. It
1136 provides per task and per CPU counters, and it provides event
1137 capabilities on top of those.
1138
1139 Say Y if unsure.
1140
1141 config PERF_COUNTERS
1142 bool "Kernel performance counters (old config option)"
1143 depends on HAVE_PERF_EVENTS
1144 help
1145 This config has been obsoleted by the PERF_EVENTS
1146 config option - please see that one for details.
1147
1148 It has no effect on the kernel whether you enable
1149 it or not, it is a compatibility placeholder.
1150
1151 Say N if unsure.
1152
1153 config DEBUG_PERF_USE_VMALLOC
1154 default n
1155 bool "Debug: use vmalloc to back perf mmap() buffers"
1156 depends on PERF_EVENTS && DEBUG_KERNEL
1157 select PERF_USE_VMALLOC
1158 help
1159 Use vmalloc memory to back perf mmap() buffers.
1160
1161 Mostly useful for debugging the vmalloc code on platforms
1162 that don't require it.
1163
1164 Say N if unsure.
1165
1166 endmenu
1167
1168 config VM_EVENT_COUNTERS
1169 default y
1170 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1171 help
1172 VM event counters are needed for event counts to be shown.
1173 This option allows the disabling of the VM event counters
1174 on EXPERT systems. /proc/vmstat will only show page counts
1175 if VM event counters are disabled.
1176
1177 config PCI_QUIRKS
1178 default y
1179 bool "Enable PCI quirk workarounds" if EXPERT
1180 depends on PCI
1181 help
1182 This enables workarounds for various PCI chipset
1183 bugs/quirks. Disable this only if your target machine is
1184 unaffected by PCI quirks.
1185
1186 config SLUB_DEBUG
1187 default y
1188 bool "Enable SLUB debugging support" if EXPERT
1189 depends on SLUB && SYSFS
1190 help
1191 SLUB has extensive debug support features. Disabling these can
1192 result in significant savings in code size. This also disables
1193 SLUB sysfs support. /sys/slab will not exist and there will be
1194 no support for cache validation etc.
1195
1196 config COMPAT_BRK
1197 bool "Disable heap randomization"
1198 default y
1199 help
1200 Randomizing heap placement makes heap exploits harder, but it
1201 also breaks ancient binaries (including anything libc5 based).
1202 This option changes the bootup default to heap randomization
1203 disabled, and can be overridden at runtime by setting
1204 /proc/sys/kernel/randomize_va_space to 2.
1205
1206 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1207
1208 choice
1209 prompt "Choose SLAB allocator"
1210 default SLUB
1211 help
1212 This option allows to select a slab allocator.
1213
1214 config SLAB
1215 bool "SLAB"
1216 help
1217 The regular slab allocator that is established and known to work
1218 well in all environments. It organizes cache hot objects in
1219 per cpu and per node queues.
1220
1221 config SLUB
1222 bool "SLUB (Unqueued Allocator)"
1223 help
1224 SLUB is a slab allocator that minimizes cache line usage
1225 instead of managing queues of cached objects (SLAB approach).
1226 Per cpu caching is realized using slabs of objects instead
1227 of queues of objects. SLUB can use memory efficiently
1228 and has enhanced diagnostics. SLUB is the default choice for
1229 a slab allocator.
1230
1231 config SLOB
1232 depends on EXPERT
1233 bool "SLOB (Simple Allocator)"
1234 help
1235 SLOB replaces the stock allocator with a drastically simpler
1236 allocator. SLOB is generally more space efficient but
1237 does not perform as well on large systems.
1238
1239 endchoice
1240
1241 config MMAP_ALLOW_UNINITIALIZED
1242 bool "Allow mmapped anonymous memory to be uninitialized"
1243 depends on EXPERT && !MMU
1244 default n
1245 help
1246 Normally, and according to the Linux spec, anonymous memory obtained
1247 from mmap() has it's contents cleared before it is passed to
1248 userspace. Enabling this config option allows you to request that
1249 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1250 providing a huge performance boost. If this option is not enabled,
1251 then the flag will be ignored.
1252
1253 This is taken advantage of by uClibc's malloc(), and also by
1254 ELF-FDPIC binfmt's brk and stack allocator.
1255
1256 Because of the obvious security issues, this option should only be
1257 enabled on embedded devices where you control what is run in
1258 userspace. Since that isn't generally a problem on no-MMU systems,
1259 it is normally safe to say Y here.
1260
1261 See Documentation/nommu-mmap.txt for more information.
1262
1263 config PROFILING
1264 bool "Profiling support"
1265 help
1266 Say Y here to enable the extended profiling support mechanisms used
1267 by profilers such as OProfile.
1268
1269 #
1270 # Place an empty function call at each tracepoint site. Can be
1271 # dynamically changed for a probe function.
1272 #
1273 config TRACEPOINTS
1274 bool
1275
1276 source "arch/Kconfig"
1277
1278 endmenu # General setup
1279
1280 config HAVE_GENERIC_DMA_COHERENT
1281 bool
1282 default n
1283
1284 config SLABINFO
1285 bool
1286 depends on PROC_FS
1287 depends on SLAB || SLUB_DEBUG
1288 default y
1289
1290 config RT_MUTEXES
1291 boolean
1292
1293 config BASE_SMALL
1294 int
1295 default 0 if BASE_FULL
1296 default 1 if !BASE_FULL
1297
1298 menuconfig MODULES
1299 bool "Enable loadable module support"
1300 help
1301 Kernel modules are small pieces of compiled code which can
1302 be inserted in the running kernel, rather than being
1303 permanently built into the kernel. You use the "modprobe"
1304 tool to add (and sometimes remove) them. If you say Y here,
1305 many parts of the kernel can be built as modules (by
1306 answering M instead of Y where indicated): this is most
1307 useful for infrequently used options which are not required
1308 for booting. For more information, see the man pages for
1309 modprobe, lsmod, modinfo, insmod and rmmod.
1310
1311 If you say Y here, you will need to run "make
1312 modules_install" to put the modules under /lib/modules/
1313 where modprobe can find them (you may need to be root to do
1314 this).
1315
1316 If unsure, say Y.
1317
1318 if MODULES
1319
1320 config MODULE_FORCE_LOAD
1321 bool "Forced module loading"
1322 default n
1323 help
1324 Allow loading of modules without version information (ie. modprobe
1325 --force). Forced module loading sets the 'F' (forced) taint flag and
1326 is usually a really bad idea.
1327
1328 config MODULE_UNLOAD
1329 bool "Module unloading"
1330 help
1331 Without this option you will not be able to unload any
1332 modules (note that some modules may not be unloadable
1333 anyway), which makes your kernel smaller, faster
1334 and simpler. If unsure, say Y.
1335
1336 config MODULE_FORCE_UNLOAD
1337 bool "Forced module unloading"
1338 depends on MODULE_UNLOAD && EXPERIMENTAL
1339 help
1340 This option allows you to force a module to unload, even if the
1341 kernel believes it is unsafe: the kernel will remove the module
1342 without waiting for anyone to stop using it (using the -f option to
1343 rmmod). This is mainly for kernel developers and desperate users.
1344 If unsure, say N.
1345
1346 config MODVERSIONS
1347 bool "Module versioning support"
1348 help
1349 Usually, you have to use modules compiled with your kernel.
1350 Saying Y here makes it sometimes possible to use modules
1351 compiled for different kernels, by adding enough information
1352 to the modules to (hopefully) spot any changes which would
1353 make them incompatible with the kernel you are running. If
1354 unsure, say N.
1355
1356 config MODULE_SRCVERSION_ALL
1357 bool "Source checksum for all modules"
1358 help
1359 Modules which contain a MODULE_VERSION get an extra "srcversion"
1360 field inserted into their modinfo section, which contains a
1361 sum of the source files which made it. This helps maintainers
1362 see exactly which source was used to build a module (since
1363 others sometimes change the module source without updating
1364 the version). With this option, such a "srcversion" field
1365 will be created for all modules. If unsure, say N.
1366
1367 endif # MODULES
1368
1369 config INIT_ALL_POSSIBLE
1370 bool
1371 help
1372 Back when each arch used to define their own cpu_online_map and
1373 cpu_possible_map, some of them chose to initialize cpu_possible_map
1374 with all 1s, and others with all 0s. When they were centralised,
1375 it was better to provide this option than to break all the archs
1376 and have several arch maintainers pursuing me down dark alleys.
1377
1378 config STOP_MACHINE
1379 bool
1380 default y
1381 depends on (SMP && MODULE_UNLOAD) || HOTPLUG_CPU
1382 help
1383 Need stop_machine() primitive.
1384
1385 source "block/Kconfig"
1386
1387 config PREEMPT_NOTIFIERS
1388 bool
1389
1390 config PADATA
1391 depends on SMP
1392 bool
1393
1394 source "kernel/Kconfig.locks"