5 default "/lib/modules/$(shell,uname --release)/.config"
6 default "/etc/kernel-config"
7 default "/boot/config-$(shell,uname --release)"
9 default "arch/$(ARCH)/defconfig"
12 def_bool $(success,$(CC) --version | head -n 1 | grep -q gcc)
16 default $(shell,$(srctree)/scripts/gcc-version.sh -p $(CC) | sed 's/^0*//') if CC_IS_GCC
20 def_bool $(success,$(CC) --version | head -n 1 | grep -q clang)
24 default $(shell,$(srctree)/scripts/clang-version.sh $(CC))
33 config BUILDTIME_EXTABLE_SORT
36 config THREAD_INFO_IN_TASK
39 Select this to move thread_info off the stack into task_struct. To
40 make this work, an arch will need to remove all thread_info fields
41 except flags and fix any runtime bugs.
43 One subtle change that will be needed is to use try_get_task_stack()
44 and put_task_stack() in save_thread_stack_tsk() and get_wchan().
53 depends on BROKEN || !SMP
56 config INIT_ENV_ARG_LIMIT
61 Maximum of each of the number of arguments and environment
62 variables passed to init from the kernel command line.
65 bool "Compile also drivers which will not load"
69 Some drivers can be compiled on a different platform than they are
70 intended to be run on. Despite they cannot be loaded there (or even
71 when they load they cannot be used due to missing HW support),
72 developers still, opposing to distributors, might want to build such
73 drivers to compile-test them.
75 If you are a developer and want to build everything available, say Y
76 here. If you are a user/distributor, say N here to exclude useless
77 drivers to be distributed.
80 string "Local version - append to kernel release"
82 Append an extra string to the end of your kernel version.
83 This will show up when you type uname, for example.
84 The string you set here will be appended after the contents of
85 any files with a filename matching localversion* in your
86 object and source tree, in that order. Your total string can
87 be a maximum of 64 characters.
89 config LOCALVERSION_AUTO
90 bool "Automatically append version information to the version string"
92 depends on !COMPILE_TEST
94 This will try to automatically determine if the current tree is a
95 release tree by looking for git tags that belong to the current
98 A string of the format -gxxxxxxxx will be added to the localversion
99 if a git-based tree is found. The string generated by this will be
100 appended after any matching localversion* files, and after the value
101 set in CONFIG_LOCALVERSION.
103 (The actual string used here is the first eight characters produced
104 by running the command:
106 $ git rev-parse --verify HEAD
108 which is done within the script "scripts/setlocalversion".)
110 config HAVE_KERNEL_GZIP
113 config HAVE_KERNEL_BZIP2
116 config HAVE_KERNEL_LZMA
119 config HAVE_KERNEL_XZ
122 config HAVE_KERNEL_LZO
125 config HAVE_KERNEL_LZ4
128 config HAVE_KERNEL_UNCOMPRESSED
132 prompt "Kernel compression mode"
134 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_UNCOMPRESSED
136 The linux kernel is a kind of self-extracting executable.
137 Several compression algorithms are available, which differ
138 in efficiency, compression and decompression speed.
139 Compression speed is only relevant when building a kernel.
140 Decompression speed is relevant at each boot.
142 If you have any problems with bzip2 or lzma compressed
143 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older
144 version of this functionality (bzip2 only), for 2.4, was
145 supplied by Christian Ludwig)
147 High compression options are mostly useful for users, who
148 are low on disk space (embedded systems), but for whom ram
151 If in doubt, select 'gzip'
155 depends on HAVE_KERNEL_GZIP
157 The old and tried gzip compression. It provides a good balance
158 between compression ratio and decompression speed.
162 depends on HAVE_KERNEL_BZIP2
164 Its compression ratio and speed is intermediate.
165 Decompression speed is slowest among the choices. The kernel
166 size is about 10% smaller with bzip2, in comparison to gzip.
167 Bzip2 uses a large amount of memory. For modern kernels you
168 will need at least 8MB RAM or more for booting.
172 depends on HAVE_KERNEL_LZMA
174 This compression algorithm's ratio is best. Decompression speed
175 is between gzip and bzip2. Compression is slowest.
176 The kernel size is about 33% smaller with LZMA in comparison to gzip.
180 depends on HAVE_KERNEL_XZ
182 XZ uses the LZMA2 algorithm and instruction set specific
183 BCJ filters which can improve compression ratio of executable
184 code. The size of the kernel is about 30% smaller with XZ in
185 comparison to gzip. On architectures for which there is a BCJ
186 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ
187 will create a few percent smaller kernel than plain LZMA.
189 The speed is about the same as with LZMA: The decompression
190 speed of XZ is better than that of bzip2 but worse than gzip
191 and LZO. Compression is slow.
195 depends on HAVE_KERNEL_LZO
197 Its compression ratio is the poorest among the choices. The kernel
198 size is about 10% bigger than gzip; however its speed
199 (both compression and decompression) is the fastest.
203 depends on HAVE_KERNEL_LZ4
205 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
206 A preliminary version of LZ4 de/compression tool is available at
207 <https://code.google.com/p/lz4/>.
209 Its compression ratio is worse than LZO. The size of the kernel
210 is about 8% bigger than LZO. But the decompression speed is
213 config KERNEL_UNCOMPRESSED
215 depends on HAVE_KERNEL_UNCOMPRESSED
217 Produce uncompressed kernel image. This option is usually not what
218 you want. It is useful for debugging the kernel in slow simulation
219 environments, where decompressing and moving the kernel is awfully
220 slow. This option allows early boot code to skip the decompressor
221 and jump right at uncompressed kernel image.
225 config DEFAULT_HOSTNAME
226 string "Default hostname"
229 This option determines the default system hostname before userspace
230 calls sethostname(2). The kernel traditionally uses "(none)" here,
231 but you may wish to use a different default here to make a minimal
232 system more usable with less configuration.
235 bool "Support for paging of anonymous memory (swap)"
236 depends on MMU && BLOCK
239 This option allows you to choose whether you want to have support
240 for so called swap devices or swap files in your kernel that are
241 used to provide more virtual memory than the actual RAM present
242 in your computer. If unsure say Y.
247 Inter Process Communication is a suite of library functions and
248 system calls which let processes (running programs) synchronize and
249 exchange information. It is generally considered to be a good thing,
250 and some programs won't run unless you say Y here. In particular, if
251 you want to run the DOS emulator dosemu under Linux (read the
252 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
253 you'll need to say Y here.
255 You can find documentation about IPC with "info ipc" and also in
256 section 6.4 of the Linux Programmer's Guide, available from
257 <http://www.tldp.org/guides.html>.
259 config SYSVIPC_SYSCTL
266 bool "POSIX Message Queues"
269 POSIX variant of message queues is a part of IPC. In POSIX message
270 queues every message has a priority which decides about succession
271 of receiving it by a process. If you want to compile and run
272 programs written e.g. for Solaris with use of its POSIX message
273 queues (functions mq_*) say Y here.
275 POSIX message queues are visible as a filesystem called 'mqueue'
276 and can be mounted somewhere if you want to do filesystem
277 operations on message queues.
281 config POSIX_MQUEUE_SYSCTL
283 depends on POSIX_MQUEUE
287 config CROSS_MEMORY_ATTACH
288 bool "Enable process_vm_readv/writev syscalls"
292 Enabling this option adds the system calls process_vm_readv and
293 process_vm_writev which allow a process with the correct privileges
294 to directly read from or write to another process' address space.
295 See the man page for more details.
298 bool "uselib syscall"
299 def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION
301 This option enables the uselib syscall, a system call used in the
302 dynamic linker from libc5 and earlier. glibc does not use this
303 system call. If you intend to run programs built on libc5 or
304 earlier, you may need to enable this syscall. Current systems
305 running glibc can safely disable this.
308 bool "Auditing support"
311 Enable auditing infrastructure that can be used with another
312 kernel subsystem, such as SELinux (which requires this for
313 logging of avc messages output). System call auditing is included
314 on architectures which support it.
316 config HAVE_ARCH_AUDITSYSCALL
321 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL
325 depends on AUDITSYSCALL
330 depends on AUDITSYSCALL
333 source "kernel/irq/Kconfig"
334 source "kernel/time/Kconfig"
336 menu "CPU/Task time and stats accounting"
338 config VIRT_CPU_ACCOUNTING
342 prompt "Cputime accounting"
343 default TICK_CPU_ACCOUNTING if !PPC64
344 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64
346 # Kind of a stub config for the pure tick based cputime accounting
347 config TICK_CPU_ACCOUNTING
348 bool "Simple tick based cputime accounting"
349 depends on !S390 && !NO_HZ_FULL
351 This is the basic tick based cputime accounting that maintains
352 statistics about user, system and idle time spent on per jiffies
357 config VIRT_CPU_ACCOUNTING_NATIVE
358 bool "Deterministic task and CPU time accounting"
359 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL
360 select VIRT_CPU_ACCOUNTING
362 Select this option to enable more accurate task and CPU time
363 accounting. This is done by reading a CPU counter on each
364 kernel entry and exit and on transitions within the kernel
365 between system, softirq and hardirq state, so there is a
366 small performance impact. In the case of s390 or IBM POWER > 5,
367 this also enables accounting of stolen time on logically-partitioned
370 config VIRT_CPU_ACCOUNTING_GEN
371 bool "Full dynticks CPU time accounting"
372 depends on HAVE_CONTEXT_TRACKING
373 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN
374 select VIRT_CPU_ACCOUNTING
375 select CONTEXT_TRACKING
377 Select this option to enable task and CPU time accounting on full
378 dynticks systems. This accounting is implemented by watching every
379 kernel-user boundaries using the context tracking subsystem.
380 The accounting is thus performed at the expense of some significant
383 For now this is only useful if you are working on the full
384 dynticks subsystem development.
390 config IRQ_TIME_ACCOUNTING
391 bool "Fine granularity task level IRQ time accounting"
392 depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE
394 Select this option to enable fine granularity task irq time
395 accounting. This is done by reading a timestamp on each
396 transitions between softirq and hardirq state, so there can be a
397 small performance impact.
399 If in doubt, say N here.
401 config BSD_PROCESS_ACCT
402 bool "BSD Process Accounting"
405 If you say Y here, a user level program will be able to instruct the
406 kernel (via a special system call) to write process accounting
407 information to a file: whenever a process exits, information about
408 that process will be appended to the file by the kernel. The
409 information includes things such as creation time, owning user,
410 command name, memory usage, controlling terminal etc. (the complete
411 list is in the struct acct in <file:include/linux/acct.h>). It is
412 up to the user level program to do useful things with this
413 information. This is generally a good idea, so say Y.
415 config BSD_PROCESS_ACCT_V3
416 bool "BSD Process Accounting version 3 file format"
417 depends on BSD_PROCESS_ACCT
420 If you say Y here, the process accounting information is written
421 in a new file format that also logs the process IDs of each
422 process and it's parent. Note that this file format is incompatible
423 with previous v0/v1/v2 file formats, so you will need updated tools
424 for processing it. A preliminary version of these tools is available
425 at <http://www.gnu.org/software/acct/>.
428 bool "Export task/process statistics through netlink"
433 Export selected statistics for tasks/processes through the
434 generic netlink interface. Unlike BSD process accounting, the
435 statistics are available during the lifetime of tasks/processes as
436 responses to commands. Like BSD accounting, they are sent to user
441 config TASK_DELAY_ACCT
442 bool "Enable per-task delay accounting"
446 Collect information on time spent by a task waiting for system
447 resources like cpu, synchronous block I/O completion and swapping
448 in pages. Such statistics can help in setting a task's priorities
449 relative to other tasks for cpu, io, rss limits etc.
454 bool "Enable extended accounting over taskstats"
457 Collect extended task accounting data and send the data
458 to userland for processing over the taskstats interface.
462 config TASK_IO_ACCOUNTING
463 bool "Enable per-task storage I/O accounting"
464 depends on TASK_XACCT
466 Collect information on the number of bytes of storage I/O which this
471 endmenu # "CPU/Task time and stats accounting"
475 depends on SMP || COMPILE_TEST
478 Make sure that CPUs running critical tasks are not disturbed by
479 any source of "noise" such as unbound workqueues, timers, kthreads...
480 Unbound jobs get offloaded to housekeeping CPUs. This is driven by
481 the "isolcpus=" boot parameter.
485 source "kernel/rcu/Kconfig"
492 tristate "Kernel .config support"
495 This option enables the complete Linux kernel ".config" file
496 contents to be saved in the kernel. It provides documentation
497 of which kernel options are used in a running kernel or in an
498 on-disk kernel. This information can be extracted from the kernel
499 image file with the script scripts/extract-ikconfig and used as
500 input to rebuild the current kernel or to build another kernel.
501 It can also be extracted from a running kernel by reading
502 /proc/config.gz if enabled (below).
505 bool "Enable access to .config through /proc/config.gz"
506 depends on IKCONFIG && PROC_FS
508 This option enables access to the kernel configuration file
509 through /proc/config.gz.
512 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
517 Select the minimal kernel log buffer size as a power of 2.
518 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config
519 parameter, see below. Any higher size also might be forced
520 by "log_buf_len" boot parameter.
530 config LOG_CPU_MAX_BUF_SHIFT
531 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)"
534 default 12 if !BASE_SMALL
535 default 0 if BASE_SMALL
538 This option allows to increase the default ring buffer size
539 according to the number of CPUs. The value defines the contribution
540 of each CPU as a power of 2. The used space is typically only few
541 lines however it might be much more when problems are reported,
544 The increased size means that a new buffer has to be allocated and
545 the original static one is unused. It makes sense only on systems
546 with more CPUs. Therefore this value is used only when the sum of
547 contributions is greater than the half of the default kernel ring
548 buffer as defined by LOG_BUF_SHIFT. The default values are set
549 so that more than 64 CPUs are needed to trigger the allocation.
551 Also this option is ignored when "log_buf_len" kernel parameter is
552 used as it forces an exact (power of two) size of the ring buffer.
554 The number of possible CPUs is used for this computation ignoring
555 hotplugging making the computation optimal for the worst case
556 scenario while allowing a simple algorithm to be used from bootup.
558 Examples shift values and their meaning:
559 17 => 128 KB for each CPU
560 16 => 64 KB for each CPU
561 15 => 32 KB for each CPU
562 14 => 16 KB for each CPU
563 13 => 8 KB for each CPU
564 12 => 4 KB for each CPU
566 config PRINTK_SAFE_LOG_BUF_SHIFT
567 int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)"
572 Select the size of an alternate printk per-CPU buffer where messages
573 printed from usafe contexts are temporary stored. One example would
574 be NMI messages, another one - printk recursion. The messages are
575 copied to the main log buffer in a safe context to avoid a deadlock.
576 The value defines the size as a power of 2.
578 Those messages are rare and limited. The largest one is when
579 a backtrace is printed. It usually fits into 4KB. Select
580 8KB if you want to be on the safe side.
583 17 => 128 KB for each CPU
584 16 => 64 KB for each CPU
585 15 => 32 KB for each CPU
586 14 => 16 KB for each CPU
587 13 => 8 KB for each CPU
588 12 => 4 KB for each CPU
591 # Architectures with an unreliable sched_clock() should select this:
593 config HAVE_UNSTABLE_SCHED_CLOCK
596 config GENERIC_SCHED_CLOCK
600 # For architectures that want to enable the support for NUMA-affine scheduler
603 config ARCH_SUPPORTS_NUMA_BALANCING
607 # For architectures that prefer to flush all TLBs after a number of pages
608 # are unmapped instead of sending one IPI per page to flush. The architecture
609 # must provide guarantees on what happens if a clean TLB cache entry is
610 # written after the unmap. Details are in mm/rmap.c near the check for
611 # should_defer_flush. The architecture should also consider if the full flush
612 # and the refill costs are offset by the savings of sending fewer IPIs.
613 config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
617 # For architectures that know their GCC __int128 support is sound
619 config ARCH_SUPPORTS_INT128
622 # For architectures that (ab)use NUMA to represent different memory regions
623 # all cpu-local but of different latencies, such as SuperH.
625 config ARCH_WANT_NUMA_VARIABLE_LOCALITY
628 config NUMA_BALANCING
629 bool "Memory placement aware NUMA scheduler"
630 depends on ARCH_SUPPORTS_NUMA_BALANCING
631 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
632 depends on SMP && NUMA && MIGRATION
634 This option adds support for automatic NUMA aware memory/task placement.
635 The mechanism is quite primitive and is based on migrating memory when
636 it has references to the node the task is running on.
638 This system will be inactive on UMA systems.
640 config NUMA_BALANCING_DEFAULT_ENABLED
641 bool "Automatically enable NUMA aware memory/task placement"
643 depends on NUMA_BALANCING
645 If set, automatic NUMA balancing will be enabled if running on a NUMA
649 bool "Control Group support"
652 This option adds support for grouping sets of processes together, for
653 use with process control subsystems such as Cpusets, CFS, memory
654 controls or device isolation.
656 - Documentation/scheduler/sched-design-CFS.txt (CFS)
657 - Documentation/cgroup-v1/ (features for grouping, isolation
658 and resource control)
668 bool "Memory controller"
672 Provides control over the memory footprint of tasks in a cgroup.
675 bool "Swap controller"
676 depends on MEMCG && SWAP
678 Provides control over the swap space consumed by tasks in a cgroup.
680 config MEMCG_SWAP_ENABLED
681 bool "Swap controller enabled by default"
682 depends on MEMCG_SWAP
685 Memory Resource Controller Swap Extension comes with its price in
686 a bigger memory consumption. General purpose distribution kernels
687 which want to enable the feature but keep it disabled by default
688 and let the user enable it by swapaccount=1 boot command line
689 parameter should have this option unselected.
690 For those who want to have the feature enabled by default should
691 select this option (if, for some reason, they need to disable it
692 then swapaccount=0 does the trick).
699 Generic block IO controller cgroup interface. This is the common
700 cgroup interface which should be used by various IO controlling
703 Currently, CFQ IO scheduler uses it to recognize task groups and
704 control disk bandwidth allocation (proportional time slice allocation)
705 to such task groups. It is also used by bio throttling logic in
706 block layer to implement upper limit in IO rates on a device.
708 This option only enables generic Block IO controller infrastructure.
709 One needs to also enable actual IO controlling logic/policy. For
710 enabling proportional weight division of disk bandwidth in CFQ, set
711 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
712 CONFIG_BLK_DEV_THROTTLING=y.
714 See Documentation/cgroup-v1/blkio-controller.txt for more information.
716 config DEBUG_BLK_CGROUP
717 bool "IO controller debugging"
718 depends on BLK_CGROUP
721 Enable some debugging help. Currently it exports additional stat
722 files in a cgroup which can be useful for debugging.
724 config CGROUP_WRITEBACK
726 depends on MEMCG && BLK_CGROUP
729 menuconfig CGROUP_SCHED
730 bool "CPU controller"
733 This feature lets CPU scheduler recognize task groups and control CPU
734 bandwidth allocation to such task groups. It uses cgroups to group
738 config FAIR_GROUP_SCHED
739 bool "Group scheduling for SCHED_OTHER"
740 depends on CGROUP_SCHED
744 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
745 depends on FAIR_GROUP_SCHED
748 This option allows users to define CPU bandwidth rates (limits) for
749 tasks running within the fair group scheduler. Groups with no limit
750 set are considered to be unconstrained and will run with no
752 See Documentation/scheduler/sched-bwc.txt for more information.
754 config RT_GROUP_SCHED
755 bool "Group scheduling for SCHED_RR/FIFO"
756 depends on CGROUP_SCHED
759 This feature lets you explicitly allocate real CPU bandwidth
760 to task groups. If enabled, it will also make it impossible to
761 schedule realtime tasks for non-root users until you allocate
762 realtime bandwidth for them.
763 See Documentation/scheduler/sched-rt-group.txt for more information.
768 bool "PIDs controller"
770 Provides enforcement of process number limits in the scope of a
771 cgroup. Any attempt to fork more processes than is allowed in the
772 cgroup will fail. PIDs are fundamentally a global resource because it
773 is fairly trivial to reach PID exhaustion before you reach even a
774 conservative kmemcg limit. As a result, it is possible to grind a
775 system to halt without being limited by other cgroup policies. The
776 PIDs controller is designed to stop this from happening.
778 It should be noted that organisational operations (such as attaching
779 to a cgroup hierarchy will *not* be blocked by the PIDs controller),
780 since the PIDs limit only affects a process's ability to fork, not to
784 bool "RDMA controller"
786 Provides enforcement of RDMA resources defined by IB stack.
787 It is fairly easy for consumers to exhaust RDMA resources, which
788 can result into resource unavailability to other consumers.
789 RDMA controller is designed to stop this from happening.
790 Attaching processes with active RDMA resources to the cgroup
791 hierarchy is allowed even if can cross the hierarchy's limit.
793 config CGROUP_FREEZER
794 bool "Freezer controller"
796 Provides a way to freeze and unfreeze all tasks in a
799 This option affects the ORIGINAL cgroup interface. The cgroup2 memory
800 controller includes important in-kernel memory consumers per default.
802 If you're using cgroup2, say N.
804 config CGROUP_HUGETLB
805 bool "HugeTLB controller"
806 depends on HUGETLB_PAGE
810 Provides a cgroup controller for HugeTLB pages.
811 When you enable this, you can put a per cgroup limit on HugeTLB usage.
812 The limit is enforced during page fault. Since HugeTLB doesn't
813 support page reclaim, enforcing the limit at page fault time implies
814 that, the application will get SIGBUS signal if it tries to access
815 HugeTLB pages beyond its limit. This requires the application to know
816 beforehand how much HugeTLB pages it would require for its use. The
817 control group is tracked in the third page lru pointer. This means
818 that we cannot use the controller with huge page less than 3 pages.
821 bool "Cpuset controller"
824 This option will let you create and manage CPUSETs which
825 allow dynamically partitioning a system into sets of CPUs and
826 Memory Nodes and assigning tasks to run only within those sets.
827 This is primarily useful on large SMP or NUMA systems.
831 config PROC_PID_CPUSET
832 bool "Include legacy /proc/<pid>/cpuset file"
837 bool "Device controller"
839 Provides a cgroup controller implementing whitelists for
840 devices which a process in the cgroup can mknod or open.
842 config CGROUP_CPUACCT
843 bool "Simple CPU accounting controller"
845 Provides a simple controller for monitoring the
846 total CPU consumed by the tasks in a cgroup.
849 bool "Perf controller"
850 depends on PERF_EVENTS
852 This option extends the perf per-cpu mode to restrict monitoring
853 to threads which belong to the cgroup specified and run on the
859 bool "Support for eBPF programs attached to cgroups"
860 depends on BPF_SYSCALL
861 select SOCK_CGROUP_DATA
863 Allow attaching eBPF programs to a cgroup using the bpf(2)
864 syscall command BPF_PROG_ATTACH.
866 In which context these programs are accessed depends on the type
867 of attachment. For instance, programs that are attached using
868 BPF_CGROUP_INET_INGRESS will be executed on the ingress path of
872 bool "Debug controller"
874 depends on DEBUG_KERNEL
876 This option enables a simple controller that exports
877 debugging information about the cgroups framework. This
878 controller is for control cgroup debugging only. Its
879 interfaces are not stable.
883 config SOCK_CGROUP_DATA
889 menuconfig NAMESPACES
890 bool "Namespaces support" if EXPERT
894 Provides the way to make tasks work with different objects using
895 the same id. For example same IPC id may refer to different objects
896 or same user id or pid may refer to different tasks when used in
897 different namespaces.
905 In this namespace tasks see different info provided with the
910 depends on (SYSVIPC || POSIX_MQUEUE)
913 In this namespace tasks work with IPC ids which correspond to
914 different IPC objects in different namespaces.
917 bool "User namespace"
920 This allows containers, i.e. vservers, to use user namespaces
921 to provide different user info for different servers.
923 When user namespaces are enabled in the kernel it is
924 recommended that the MEMCG option also be enabled and that
925 user-space use the memory control groups to limit the amount
926 of memory a memory unprivileged users can use.
931 bool "PID Namespaces"
934 Support process id namespaces. This allows having multiple
935 processes with the same pid as long as they are in different
936 pid namespaces. This is a building block of containers.
939 bool "Network namespace"
943 Allow user space to create what appear to be multiple instances
944 of the network stack.
948 config SCHED_AUTOGROUP
949 bool "Automatic process group scheduling"
952 select FAIR_GROUP_SCHED
954 This option optimizes the scheduler for common desktop workloads by
955 automatically creating and populating task groups. This separation
956 of workloads isolates aggressive CPU burners (like build jobs) from
957 desktop applications. Task group autogeneration is currently based
960 config SYSFS_DEPRECATED
961 bool "Enable deprecated sysfs features to support old userspace tools"
965 This option adds code that switches the layout of the "block" class
966 devices, to not show up in /sys/class/block/, but only in
969 This switch is only active when the sysfs.deprecated=1 boot option is
970 passed or the SYSFS_DEPRECATED_V2 option is set.
972 This option allows new kernels to run on old distributions and tools,
973 which might get confused by /sys/class/block/. Since 2007/2008 all
974 major distributions and tools handle this just fine.
976 Recent distributions and userspace tools after 2009/2010 depend on
977 the existence of /sys/class/block/, and will not work with this
980 Only if you are using a new kernel on an old distribution, you might
983 config SYSFS_DEPRECATED_V2
984 bool "Enable deprecated sysfs features by default"
987 depends on SYSFS_DEPRECATED
989 Enable deprecated sysfs by default.
991 See the CONFIG_SYSFS_DEPRECATED option for more details about this
994 Only if you are using a new kernel on an old distribution, you might
995 need to say Y here. Even then, odds are you would not need it
996 enabled, you can always pass the boot option if absolutely necessary.
999 bool "Kernel->user space relay support (formerly relayfs)"
1002 This option enables support for relay interface support in
1003 certain file systems (such as debugfs).
1004 It is designed to provide an efficient mechanism for tools and
1005 facilities to relay large amounts of data from kernel space to
1010 config BLK_DEV_INITRD
1011 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
1013 The initial RAM filesystem is a ramfs which is loaded by the
1014 boot loader (loadlin or lilo) and that is mounted as root
1015 before the normal boot procedure. It is typically used to
1016 load modules needed to mount the "real" root file system,
1017 etc. See <file:Documentation/admin-guide/initrd.rst> for details.
1019 If RAM disk support (BLK_DEV_RAM) is also included, this
1020 also enables initial RAM disk (initrd) support and adds
1021 15 Kbytes (more on some other architectures) to the kernel size.
1027 source "usr/Kconfig"
1032 prompt "Compiler optimization level"
1033 default CC_OPTIMIZE_FOR_PERFORMANCE
1035 config CC_OPTIMIZE_FOR_PERFORMANCE
1036 bool "Optimize for performance"
1038 This is the default optimization level for the kernel, building
1039 with the "-O2" compiler flag for best performance and most
1040 helpful compile-time warnings.
1042 config CC_OPTIMIZE_FOR_SIZE
1043 bool "Optimize for size"
1045 Enabling this option will pass "-Os" instead of "-O2" to
1046 your compiler resulting in a smaller kernel.
1052 config HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1055 This requires that the arch annotates or otherwise protects
1056 its external entry points from being discarded. Linker scripts
1057 must also merge .text.*, .data.*, and .bss.* correctly into
1058 output sections. Care must be taken not to pull in unrelated
1059 sections (e.g., '.text.init'). Typically '.' in section names
1060 is used to distinguish them from label names / C identifiers.
1062 config LD_DEAD_CODE_DATA_ELIMINATION
1063 bool "Dead code and data elimination (EXPERIMENTAL)"
1064 depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1067 Enable this if you want to do dead code and data elimination with
1068 the linker by compiling with -ffunction-sections -fdata-sections,
1069 and linking with --gc-sections.
1071 This can reduce on disk and in-memory size of the kernel
1072 code and static data, particularly for small configs and
1073 on small systems. This has the possibility of introducing
1074 silently broken kernel if the required annotations are not
1075 present. This option is not well tested yet, so use at your
1087 config SYSCTL_EXCEPTION_TRACE
1090 Enable support for /proc/sys/debug/exception-trace.
1092 config SYSCTL_ARCH_UNALIGN_NO_WARN
1095 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1096 Allows arch to define/use @no_unaligned_warning to possibly warn
1097 about unaligned access emulation going on under the hood.
1099 config SYSCTL_ARCH_UNALIGN_ALLOW
1102 Enable support for /proc/sys/kernel/unaligned-trap
1103 Allows arches to define/use @unaligned_enabled to runtime toggle
1104 the unaligned access emulation.
1105 see arch/parisc/kernel/unaligned.c for reference
1107 config HAVE_PCSPKR_PLATFORM
1110 # interpreter that classic socket filters depend on
1115 bool "Configure standard kernel features (expert users)"
1116 # Unhide debug options, to make the on-by-default options visible
1119 This option allows certain base kernel options and settings
1120 to be disabled or tweaked. This is for specialized
1121 environments which can tolerate a "non-standard" kernel.
1122 Only use this if you really know what you are doing.
1125 bool "Enable 16-bit UID system calls" if EXPERT
1126 depends on HAVE_UID16 && MULTIUSER
1129 This enables the legacy 16-bit UID syscall wrappers.
1132 bool "Multiple users, groups and capabilities support" if EXPERT
1135 This option enables support for non-root users, groups and
1138 If you say N here, all processes will run with UID 0, GID 0, and all
1139 possible capabilities. Saying N here also compiles out support for
1140 system calls related to UIDs, GIDs, and capabilities, such as setuid,
1143 If unsure, say Y here.
1145 config SGETMASK_SYSCALL
1146 bool "sgetmask/ssetmask syscalls support" if EXPERT
1147 def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH
1149 sys_sgetmask and sys_ssetmask are obsolete system calls
1150 no longer supported in libc but still enabled by default in some
1153 If unsure, leave the default option here.
1155 config SYSFS_SYSCALL
1156 bool "Sysfs syscall support" if EXPERT
1159 sys_sysfs is an obsolete system call no longer supported in libc.
1160 Note that disabling this option is more secure but might break
1161 compatibility with some systems.
1163 If unsure say Y here.
1165 config SYSCTL_SYSCALL
1166 bool "Sysctl syscall support" if EXPERT
1167 depends on PROC_SYSCTL
1171 sys_sysctl uses binary paths that have been found challenging
1172 to properly maintain and use. The interface in /proc/sys
1173 using paths with ascii names is now the primary path to this
1176 Almost nothing using the binary sysctl interface so if you are
1177 trying to save some space it is probably safe to disable this,
1178 making your kernel marginally smaller.
1180 If unsure say N here.
1183 bool "open by fhandle syscalls" if EXPERT
1187 If you say Y here, a user level program will be able to map
1188 file names to handle and then later use the handle for
1189 different file system operations. This is useful in implementing
1190 userspace file servers, which now track files using handles instead
1191 of names. The handle would remain the same even if file names
1192 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
1196 bool "Posix Clocks & timers" if EXPERT
1199 This includes native support for POSIX timers to the kernel.
1200 Some embedded systems have no use for them and therefore they
1201 can be configured out to reduce the size of the kernel image.
1203 When this option is disabled, the following syscalls won't be
1204 available: timer_create, timer_gettime: timer_getoverrun,
1205 timer_settime, timer_delete, clock_adjtime, getitimer,
1206 setitimer, alarm. Furthermore, the clock_settime, clock_gettime,
1207 clock_getres and clock_nanosleep syscalls will be limited to
1208 CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only.
1214 bool "Enable support for printk" if EXPERT
1217 This option enables normal printk support. Removing it
1218 eliminates most of the message strings from the kernel image
1219 and makes the kernel more or less silent. As this makes it
1220 very difficult to diagnose system problems, saying N here is
1221 strongly discouraged.
1229 bool "BUG() support" if EXPERT
1232 Disabling this option eliminates support for BUG and WARN, reducing
1233 the size of your kernel image and potentially quietly ignoring
1234 numerous fatal conditions. You should only consider disabling this
1235 option for embedded systems with no facilities for reporting errors.
1241 bool "Enable ELF core dumps" if EXPERT
1243 Enable support for generating core dumps. Disabling saves about 4k.
1246 config PCSPKR_PLATFORM
1247 bool "Enable PC-Speaker support" if EXPERT
1248 depends on HAVE_PCSPKR_PLATFORM
1252 This option allows to disable the internal PC-Speaker
1253 support, saving some memory.
1257 bool "Enable full-sized data structures for core" if EXPERT
1259 Disabling this option reduces the size of miscellaneous core
1260 kernel data structures. This saves memory on small machines,
1261 but may reduce performance.
1264 bool "Enable futex support" if EXPERT
1268 Disabling this option will cause the kernel to be built without
1269 support for "fast userspace mutexes". The resulting kernel may not
1270 run glibc-based applications correctly.
1274 depends on FUTEX && RT_MUTEXES
1277 config HAVE_FUTEX_CMPXCHG
1281 Architectures should select this if futex_atomic_cmpxchg_inatomic()
1282 is implemented and always working. This removes a couple of runtime
1286 bool "Enable eventpoll support" if EXPERT
1290 Disabling this option will cause the kernel to be built without
1291 support for epoll family of system calls.
1294 bool "Enable signalfd() system call" if EXPERT
1298 Enable the signalfd() system call that allows to receive signals
1299 on a file descriptor.
1304 bool "Enable timerfd() system call" if EXPERT
1308 Enable the timerfd() system call that allows to receive timer
1309 events on a file descriptor.
1314 bool "Enable eventfd() system call" if EXPERT
1318 Enable the eventfd() system call that allows to receive both
1319 kernel notification (ie. KAIO) or userspace notifications.
1324 bool "Use full shmem filesystem" if EXPERT
1328 The shmem is an internal filesystem used to manage shared memory.
1329 It is backed by swap and manages resource limits. It is also exported
1330 to userspace as tmpfs if TMPFS is enabled. Disabling this
1331 option replaces shmem and tmpfs with the much simpler ramfs code,
1332 which may be appropriate on small systems without swap.
1335 bool "Enable AIO support" if EXPERT
1338 This option enables POSIX asynchronous I/O which may by used
1339 by some high performance threaded applications. Disabling
1340 this option saves about 7k.
1342 config ADVISE_SYSCALLS
1343 bool "Enable madvise/fadvise syscalls" if EXPERT
1346 This option enables the madvise and fadvise syscalls, used by
1347 applications to advise the kernel about their future memory or file
1348 usage, improving performance. If building an embedded system where no
1349 applications use these syscalls, you can disable this option to save
1353 bool "Enable membarrier() system call" if EXPERT
1356 Enable the membarrier() system call that allows issuing memory
1357 barriers across all running threads, which can be used to distribute
1358 the cost of user-space memory barriers asymmetrically by transforming
1359 pairs of memory barriers into pairs consisting of membarrier() and a
1364 config CHECKPOINT_RESTORE
1365 bool "Checkpoint/restore support" if EXPERT
1366 select PROC_CHILDREN
1369 Enables additional kernel features in a sake of checkpoint/restore.
1370 In particular it adds auxiliary prctl codes to setup process text,
1371 data and heap segment sizes, and a few additional /proc filesystem
1374 If unsure, say N here.
1377 bool "Load all symbols for debugging/ksymoops" if EXPERT
1380 Say Y here to let the kernel print out symbolic crash information and
1381 symbolic stack backtraces. This increases the size of the kernel
1382 somewhat, as all symbols have to be loaded into the kernel image.
1385 bool "Include all symbols in kallsyms"
1386 depends on DEBUG_KERNEL && KALLSYMS
1388 Normally kallsyms only contains the symbols of functions for nicer
1389 OOPS messages and backtraces (i.e., symbols from the text and inittext
1390 sections). This is sufficient for most cases. And only in very rare
1391 cases (e.g., when a debugger is used) all symbols are required (e.g.,
1392 names of variables from the data sections, etc).
1394 This option makes sure that all symbols are loaded into the kernel
1395 image (i.e., symbols from all sections) in cost of increased kernel
1396 size (depending on the kernel configuration, it may be 300KiB or
1397 something like this).
1399 Say N unless you really need all symbols.
1401 config KALLSYMS_ABSOLUTE_PERCPU
1404 default X86_64 && SMP
1406 config KALLSYMS_BASE_RELATIVE
1411 Instead of emitting them as absolute values in the native word size,
1412 emit the symbol references in the kallsyms table as 32-bit entries,
1413 each containing a relative value in the range [base, base + U32_MAX]
1414 or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either
1415 an absolute value in the range [0, S32_MAX] or a relative value in the
1416 range [base, base + S32_MAX], where base is the lowest relative symbol
1417 address encountered in the image.
1419 On 64-bit builds, this reduces the size of the address table by 50%,
1420 but more importantly, it results in entries whose values are build
1421 time constants, and no relocation pass is required at runtime to fix
1422 up the entries based on the runtime load address of the kernel.
1424 # end of the "standard kernel features (expert users)" menu
1426 # syscall, maps, verifier
1428 bool "Enable bpf() system call"
1434 Enable the bpf() system call that allows to manipulate eBPF
1435 programs and maps via file descriptors.
1437 config BPF_JIT_ALWAYS_ON
1438 bool "Permanently enable BPF JIT and remove BPF interpreter"
1439 depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT
1441 Enables BPF JIT and removes BPF interpreter to avoid
1442 speculative execution of BPF instructions by the interpreter
1445 bool "Enable userfaultfd() system call"
1449 Enable the userfaultfd() system call that allows to intercept and
1450 handle page faults in userland.
1452 config ARCH_HAS_MEMBARRIER_CALLBACKS
1455 config ARCH_HAS_MEMBARRIER_SYNC_CORE
1459 bool "Enable rseq() system call" if EXPERT
1461 depends on HAVE_RSEQ
1464 Enable the restartable sequences system call. It provides a
1465 user-space cache for the current CPU number value, which
1466 speeds up getting the current CPU number from user-space,
1467 as well as an ABI to speed up user-space operations on
1474 bool "Enabled debugging of rseq() system call" if EXPERT
1475 depends on RSEQ && DEBUG_KERNEL
1477 Enable extra debugging checks for the rseq system call.
1482 bool "Embedded system"
1483 option allnoconfig_y
1486 This option should be enabled if compiling the kernel for
1487 an embedded system so certain expert options are available
1490 config HAVE_PERF_EVENTS
1493 See tools/perf/design.txt for details.
1495 config PERF_USE_VMALLOC
1498 See tools/perf/design.txt for details
1501 bool "PC/104 support" if EXPERT
1503 Expose PC/104 form factor device drivers and options available for
1504 selection and configuration. Enable this option if your target
1505 machine has a PC/104 bus.
1507 menu "Kernel Performance Events And Counters"
1510 bool "Kernel performance events and counters"
1511 default y if PROFILING
1512 depends on HAVE_PERF_EVENTS
1517 Enable kernel support for various performance events provided
1518 by software and hardware.
1520 Software events are supported either built-in or via the
1521 use of generic tracepoints.
1523 Most modern CPUs support performance events via performance
1524 counter registers. These registers count the number of certain
1525 types of hw events: such as instructions executed, cachemisses
1526 suffered, or branches mis-predicted - without slowing down the
1527 kernel or applications. These registers can also trigger interrupts
1528 when a threshold number of events have passed - and can thus be
1529 used to profile the code that runs on that CPU.
1531 The Linux Performance Event subsystem provides an abstraction of
1532 these software and hardware event capabilities, available via a
1533 system call and used by the "perf" utility in tools/perf/. It
1534 provides per task and per CPU counters, and it provides event
1535 capabilities on top of those.
1539 config DEBUG_PERF_USE_VMALLOC
1541 bool "Debug: use vmalloc to back perf mmap() buffers"
1542 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC
1543 select PERF_USE_VMALLOC
1545 Use vmalloc memory to back perf mmap() buffers.
1547 Mostly useful for debugging the vmalloc code on platforms
1548 that don't require it.
1554 config VM_EVENT_COUNTERS
1556 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1558 VM event counters are needed for event counts to be shown.
1559 This option allows the disabling of the VM event counters
1560 on EXPERT systems. /proc/vmstat will only show page counts
1561 if VM event counters are disabled.
1565 bool "Enable SLUB debugging support" if EXPERT
1566 depends on SLUB && SYSFS
1568 SLUB has extensive debug support features. Disabling these can
1569 result in significant savings in code size. This also disables
1570 SLUB sysfs support. /sys/slab will not exist and there will be
1571 no support for cache validation etc.
1573 config SLUB_MEMCG_SYSFS_ON
1575 bool "Enable memcg SLUB sysfs support by default" if EXPERT
1576 depends on SLUB && SYSFS && MEMCG
1578 SLUB creates a directory under /sys/kernel/slab for each
1579 allocation cache to host info and debug files. If memory
1580 cgroup is enabled, each cache can have per memory cgroup
1581 caches. SLUB can create the same sysfs directories for these
1582 caches under /sys/kernel/slab/CACHE/cgroup but it can lead
1583 to a very high number of debug files being created. This is
1584 controlled by slub_memcg_sysfs boot parameter and this
1585 config option determines the parameter's default value.
1588 bool "Disable heap randomization"
1591 Randomizing heap placement makes heap exploits harder, but it
1592 also breaks ancient binaries (including anything libc5 based).
1593 This option changes the bootup default to heap randomization
1594 disabled, and can be overridden at runtime by setting
1595 /proc/sys/kernel/randomize_va_space to 2.
1597 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1600 prompt "Choose SLAB allocator"
1603 This option allows to select a slab allocator.
1607 select HAVE_HARDENED_USERCOPY_ALLOCATOR
1609 The regular slab allocator that is established and known to work
1610 well in all environments. It organizes cache hot objects in
1611 per cpu and per node queues.
1614 bool "SLUB (Unqueued Allocator)"
1615 select HAVE_HARDENED_USERCOPY_ALLOCATOR
1617 SLUB is a slab allocator that minimizes cache line usage
1618 instead of managing queues of cached objects (SLAB approach).
1619 Per cpu caching is realized using slabs of objects instead
1620 of queues of objects. SLUB can use memory efficiently
1621 and has enhanced diagnostics. SLUB is the default choice for
1626 bool "SLOB (Simple Allocator)"
1628 SLOB replaces the stock allocator with a drastically simpler
1629 allocator. SLOB is generally more space efficient but
1630 does not perform as well on large systems.
1634 config SLAB_MERGE_DEFAULT
1635 bool "Allow slab caches to be merged"
1638 For reduced kernel memory fragmentation, slab caches can be
1639 merged when they share the same size and other characteristics.
1640 This carries a risk of kernel heap overflows being able to
1641 overwrite objects from merged caches (and more easily control
1642 cache layout), which makes such heap attacks easier to exploit
1643 by attackers. By keeping caches unmerged, these kinds of exploits
1644 can usually only damage objects in the same cache. To disable
1645 merging at runtime, "slab_nomerge" can be passed on the kernel
1648 config SLAB_FREELIST_RANDOM
1650 depends on SLAB || SLUB
1651 bool "SLAB freelist randomization"
1653 Randomizes the freelist order used on creating new pages. This
1654 security feature reduces the predictability of the kernel slab
1655 allocator against heap overflows.
1657 config SLAB_FREELIST_HARDENED
1658 bool "Harden slab freelist metadata"
1661 Many kernel heap attacks try to target slab cache metadata and
1662 other infrastructure. This options makes minor performance
1663 sacrifies to harden the kernel slab allocator against common
1664 freelist exploit methods.
1666 config SLUB_CPU_PARTIAL
1668 depends on SLUB && SMP
1669 bool "SLUB per cpu partial cache"
1671 Per cpu partial caches accellerate objects allocation and freeing
1672 that is local to a processor at the price of more indeterminism
1673 in the latency of the free. On overflow these caches will be cleared
1674 which requires the taking of locks that may cause latency spikes.
1675 Typically one would choose no for a realtime system.
1677 config MMAP_ALLOW_UNINITIALIZED
1678 bool "Allow mmapped anonymous memory to be uninitialized"
1679 depends on EXPERT && !MMU
1682 Normally, and according to the Linux spec, anonymous memory obtained
1683 from mmap() has it's contents cleared before it is passed to
1684 userspace. Enabling this config option allows you to request that
1685 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1686 providing a huge performance boost. If this option is not enabled,
1687 then the flag will be ignored.
1689 This is taken advantage of by uClibc's malloc(), and also by
1690 ELF-FDPIC binfmt's brk and stack allocator.
1692 Because of the obvious security issues, this option should only be
1693 enabled on embedded devices where you control what is run in
1694 userspace. Since that isn't generally a problem on no-MMU systems,
1695 it is normally safe to say Y here.
1697 See Documentation/nommu-mmap.txt for more information.
1699 config SYSTEM_DATA_VERIFICATION
1701 select SYSTEM_TRUSTED_KEYRING
1705 select ASYMMETRIC_KEY_TYPE
1706 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
1709 select X509_CERTIFICATE_PARSER
1710 select PKCS7_MESSAGE_PARSER
1712 Provide PKCS#7 message verification using the contents of the system
1713 trusted keyring to provide public keys. This then can be used for
1714 module verification, kexec image verification and firmware blob
1718 bool "Profiling support"
1720 Say Y here to enable the extended profiling support mechanisms used
1721 by profilers such as OProfile.
1724 # Place an empty function call at each tracepoint site. Can be
1725 # dynamically changed for a probe function.
1730 source "arch/Kconfig"
1732 endmenu # General setup
1739 default 0 if BASE_FULL
1740 default 1 if !BASE_FULL
1743 bool "Enable loadable module support"
1746 Kernel modules are small pieces of compiled code which can
1747 be inserted in the running kernel, rather than being
1748 permanently built into the kernel. You use the "modprobe"
1749 tool to add (and sometimes remove) them. If you say Y here,
1750 many parts of the kernel can be built as modules (by
1751 answering M instead of Y where indicated): this is most
1752 useful for infrequently used options which are not required
1753 for booting. For more information, see the man pages for
1754 modprobe, lsmod, modinfo, insmod and rmmod.
1756 If you say Y here, you will need to run "make
1757 modules_install" to put the modules under /lib/modules/
1758 where modprobe can find them (you may need to be root to do
1765 config MODULE_FORCE_LOAD
1766 bool "Forced module loading"
1769 Allow loading of modules without version information (ie. modprobe
1770 --force). Forced module loading sets the 'F' (forced) taint flag and
1771 is usually a really bad idea.
1773 config MODULE_UNLOAD
1774 bool "Module unloading"
1776 Without this option you will not be able to unload any
1777 modules (note that some modules may not be unloadable
1778 anyway), which makes your kernel smaller, faster
1779 and simpler. If unsure, say Y.
1781 config MODULE_FORCE_UNLOAD
1782 bool "Forced module unloading"
1783 depends on MODULE_UNLOAD
1785 This option allows you to force a module to unload, even if the
1786 kernel believes it is unsafe: the kernel will remove the module
1787 without waiting for anyone to stop using it (using the -f option to
1788 rmmod). This is mainly for kernel developers and desperate users.
1792 bool "Module versioning support"
1794 Usually, you have to use modules compiled with your kernel.
1795 Saying Y here makes it sometimes possible to use modules
1796 compiled for different kernels, by adding enough information
1797 to the modules to (hopefully) spot any changes which would
1798 make them incompatible with the kernel you are running. If
1801 config MODULE_REL_CRCS
1803 depends on MODVERSIONS
1805 config MODULE_SRCVERSION_ALL
1806 bool "Source checksum for all modules"
1808 Modules which contain a MODULE_VERSION get an extra "srcversion"
1809 field inserted into their modinfo section, which contains a
1810 sum of the source files which made it. This helps maintainers
1811 see exactly which source was used to build a module (since
1812 others sometimes change the module source without updating
1813 the version). With this option, such a "srcversion" field
1814 will be created for all modules. If unsure, say N.
1817 bool "Module signature verification"
1819 select SYSTEM_DATA_VERIFICATION
1821 Check modules for valid signatures upon load: the signature
1822 is simply appended to the module. For more information see
1823 <file:Documentation/admin-guide/module-signing.rst>.
1825 Note that this option adds the OpenSSL development packages as a
1826 kernel build dependency so that the signing tool can use its crypto
1829 !!!WARNING!!! If you enable this option, you MUST make sure that the
1830 module DOES NOT get stripped after being signed. This includes the
1831 debuginfo strip done by some packagers (such as rpmbuild) and
1832 inclusion into an initramfs that wants the module size reduced.
1834 config MODULE_SIG_FORCE
1835 bool "Require modules to be validly signed"
1836 depends on MODULE_SIG
1838 Reject unsigned modules or signed modules for which we don't have a
1839 key. Without this, such modules will simply taint the kernel.
1841 config MODULE_SIG_ALL
1842 bool "Automatically sign all modules"
1844 depends on MODULE_SIG
1846 Sign all modules during make modules_install. Without this option,
1847 modules must be signed manually, using the scripts/sign-file tool.
1849 comment "Do not forget to sign required modules with scripts/sign-file"
1850 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL
1853 prompt "Which hash algorithm should modules be signed with?"
1854 depends on MODULE_SIG
1856 This determines which sort of hashing algorithm will be used during
1857 signature generation. This algorithm _must_ be built into the kernel
1858 directly so that signature verification can take place. It is not
1859 possible to load a signed module containing the algorithm to check
1860 the signature on that module.
1862 config MODULE_SIG_SHA1
1863 bool "Sign modules with SHA-1"
1866 config MODULE_SIG_SHA224
1867 bool "Sign modules with SHA-224"
1868 select CRYPTO_SHA256
1870 config MODULE_SIG_SHA256
1871 bool "Sign modules with SHA-256"
1872 select CRYPTO_SHA256
1874 config MODULE_SIG_SHA384
1875 bool "Sign modules with SHA-384"
1876 select CRYPTO_SHA512
1878 config MODULE_SIG_SHA512
1879 bool "Sign modules with SHA-512"
1880 select CRYPTO_SHA512
1884 config MODULE_SIG_HASH
1886 depends on MODULE_SIG
1887 default "sha1" if MODULE_SIG_SHA1
1888 default "sha224" if MODULE_SIG_SHA224
1889 default "sha256" if MODULE_SIG_SHA256
1890 default "sha384" if MODULE_SIG_SHA384
1891 default "sha512" if MODULE_SIG_SHA512
1893 config MODULE_COMPRESS
1894 bool "Compress modules on installation"
1898 Compresses kernel modules when 'make modules_install' is run; gzip or
1899 xz depending on "Compression algorithm" below.
1901 module-init-tools MAY support gzip, and kmod MAY support gzip and xz.
1903 Out-of-tree kernel modules installed using Kbuild will also be
1904 compressed upon installation.
1906 Note: for modules inside an initrd or initramfs, it's more efficient
1907 to compress the whole initrd or initramfs instead.
1909 Note: This is fully compatible with signed modules.
1914 prompt "Compression algorithm"
1915 depends on MODULE_COMPRESS
1916 default MODULE_COMPRESS_GZIP
1918 This determines which sort of compression will be used during
1919 'make modules_install'.
1921 GZIP (default) and XZ are supported.
1923 config MODULE_COMPRESS_GZIP
1926 config MODULE_COMPRESS_XZ
1931 config TRIM_UNUSED_KSYMS
1932 bool "Trim unused exported kernel symbols"
1933 depends on MODULES && !UNUSED_SYMBOLS
1935 The kernel and some modules make many symbols available for
1936 other modules to use via EXPORT_SYMBOL() and variants. Depending
1937 on the set of modules being selected in your kernel configuration,
1938 many of those exported symbols might never be used.
1940 This option allows for unused exported symbols to be dropped from
1941 the build. In turn, this provides the compiler more opportunities
1942 (especially when using LTO) for optimizing the code and reducing
1943 binary size. This might have some security advantages as well.
1945 If unsure, or if you need to build out-of-tree modules, say N.
1949 config MODULES_TREE_LOOKUP
1951 depends on PERF_EVENTS || TRACING
1953 config INIT_ALL_POSSIBLE
1956 Back when each arch used to define their own cpu_online_mask and
1957 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
1958 with all 1s, and others with all 0s. When they were centralised,
1959 it was better to provide this option than to break all the archs
1960 and have several arch maintainers pursuing me down dark alleys.
1962 source "block/Kconfig"
1964 config PREEMPT_NOTIFIERS
1974 Build a simple ASN.1 grammar compiler that produces a bytecode output
1975 that can be interpreted by the ASN.1 stream decoder and used to
1976 inform it as to what tags are to be expected in a stream and what
1977 functions to call on what tags.
1979 source "kernel/Kconfig.locks"
1981 config ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
1984 # It may be useful for an architecture to override the definitions of the
1985 # SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h>
1986 # and the COMPAT_ variants in <linux/compat.h>, in particular to use a
1987 # different calling convention for syscalls. They can also override the
1988 # macros for not-implemented syscalls in kernel/sys_ni.c and
1989 # kernel/time/posix-stubs.c. All these overrides need to be available in
1990 # <asm/syscall_wrapper.h>.
1991 config ARCH_HAS_SYSCALL_WRAPPER
projects / linux / blob