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".)
111 string "Build ID Salt"
114 The build ID is used to link binaries and their debug info. Setting
115 this option will use the value in the calculation of the build id.
116 This is mostly useful for distributions which want to ensure the
117 build is unique between builds. It's safe to leave the default.
119 config HAVE_KERNEL_GZIP
122 config HAVE_KERNEL_BZIP2
125 config HAVE_KERNEL_LZMA
128 config HAVE_KERNEL_XZ
131 config HAVE_KERNEL_LZO
134 config HAVE_KERNEL_LZ4
137 config HAVE_KERNEL_UNCOMPRESSED
141 prompt "Kernel compression mode"
143 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_UNCOMPRESSED
145 The linux kernel is a kind of self-extracting executable.
146 Several compression algorithms are available, which differ
147 in efficiency, compression and decompression speed.
148 Compression speed is only relevant when building a kernel.
149 Decompression speed is relevant at each boot.
151 If you have any problems with bzip2 or lzma compressed
153 version of this functionality (bzip2 only), for 2.4, was
154 supplied by Christian Ludwig)
156 High compression options are mostly useful for users, who
157 are low on disk space (embedded systems), but for whom ram
160 If in doubt, select 'gzip'
164 depends on HAVE_KERNEL_GZIP
166 The old and tried gzip compression. It provides a good balance
167 between compression ratio and decompression speed.
171 depends on HAVE_KERNEL_BZIP2
173 Its compression ratio and speed is intermediate.
174 Decompression speed is slowest among the choices. The kernel
175 size is about 10% smaller with bzip2, in comparison to gzip.
176 Bzip2 uses a large amount of memory. For modern kernels you
177 will need at least 8MB RAM or more for booting.
181 depends on HAVE_KERNEL_LZMA
183 This compression algorithm's ratio is best. Decompression speed
184 is between gzip and bzip2. Compression is slowest.
185 The kernel size is about 33% smaller with LZMA in comparison to gzip.
189 depends on HAVE_KERNEL_XZ
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.
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.
204 depends on HAVE_KERNEL_LZO
206 Its compression ratio is the poorest among the choices. The kernel
207 size is about 10% bigger than gzip; however its speed
208 (both compression and decompression) is the fastest.
212 depends on HAVE_KERNEL_LZ4
214 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding.
215 A preliminary version of LZ4 de/compression tool is available at
216 <https://code.google.com/p/lz4/>.
218 Its compression ratio is worse than LZO. The size of the kernel
219 is about 8% bigger than LZO. But the decompression speed is
222 config KERNEL_UNCOMPRESSED
224 depends on HAVE_KERNEL_UNCOMPRESSED
226 Produce uncompressed kernel image. This option is usually not what
227 you want. It is useful for debugging the kernel in slow simulation
228 environments, where decompressing and moving the kernel is awfully
229 slow. This option allows early boot code to skip the decompressor
230 and jump right at uncompressed kernel image.
234 config DEFAULT_HOSTNAME
235 string "Default hostname"
238 This option determines the default system hostname before userspace
239 calls sethostname(2). The kernel traditionally uses "(none)" here,
240 but you may wish to use a different default here to make a minimal
241 system more usable with less configuration.
244 bool "Support for paging of anonymous memory (swap)"
245 depends on MMU && BLOCK
248 This option allows you to choose whether you want to have support
249 for so called swap devices or swap files in your kernel that are
250 used to provide more virtual memory than the actual RAM present
251 in your computer. If unsure say Y.
256 Inter Process Communication is a suite of library functions and
257 system calls which let processes (running programs) synchronize and
258 exchange information. It is generally considered to be a good thing,
259 and some programs won't run unless you say Y here. In particular, if
260 you want to run the DOS emulator dosemu under Linux (read the
261 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>),
262 you'll need to say Y here.
264 You can find documentation about IPC with "info ipc" and also in
265 section 6.4 of the Linux Programmer's Guide, available from
266 <http://www.tldp.org/guides.html>.
268 config SYSVIPC_SYSCTL
275 bool "POSIX Message Queues"
278 POSIX variant of message queues is a part of IPC. In POSIX message
279 queues every message has a priority which decides about succession
280 of receiving it by a process. If you want to compile and run
281 programs written e.g. for Solaris with use of its POSIX message
282 queues (functions mq_*) say Y here.
284 POSIX message queues are visible as a filesystem called 'mqueue'
285 and can be mounted somewhere if you want to do filesystem
286 operations on message queues.
290 config POSIX_MQUEUE_SYSCTL
292 depends on POSIX_MQUEUE
296 config CROSS_MEMORY_ATTACH
297 bool "Enable process_vm_readv/writev syscalls"
301 Enabling this option adds the system calls process_vm_readv and
302 process_vm_writev which allow a process with the correct privileges
303 to directly read from or write to another process' address space.
304 See the man page for more details.
307 bool "uselib syscall"
308 def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION
310 This option enables the uselib syscall, a system call used in the
311 dynamic linker from libc5 and earlier. glibc does not use this
312 system call. If you intend to run programs built on libc5 or
313 earlier, you may need to enable this syscall. Current systems
314 running glibc can safely disable this.
317 bool "Auditing support"
320 Enable auditing infrastructure that can be used with another
321 kernel subsystem, such as SELinux (which requires this for
322 logging of avc messages output). System call auditing is included
323 on architectures which support it.
325 config HAVE_ARCH_AUDITSYSCALL
330 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL
334 depends on AUDITSYSCALL
339 depends on AUDITSYSCALL
342 source "kernel/irq/Kconfig"
343 source "kernel/time/Kconfig"
345 menu "CPU/Task time and stats accounting"
347 config VIRT_CPU_ACCOUNTING
351 prompt "Cputime accounting"
352 default TICK_CPU_ACCOUNTING if !PPC64
353 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64
355 # Kind of a stub config for the pure tick based cputime accounting
356 config TICK_CPU_ACCOUNTING
357 bool "Simple tick based cputime accounting"
358 depends on !S390 && !NO_HZ_FULL
360 This is the basic tick based cputime accounting that maintains
361 statistics about user, system and idle time spent on per jiffies
366 config VIRT_CPU_ACCOUNTING_NATIVE
367 bool "Deterministic task and CPU time accounting"
368 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL
369 select VIRT_CPU_ACCOUNTING
371 Select this option to enable more accurate task and CPU time
372 accounting. This is done by reading a CPU counter on each
373 kernel entry and exit and on transitions within the kernel
374 between system, softirq and hardirq state, so there is a
375 small performance impact. In the case of s390 or IBM POWER > 5,
376 this also enables accounting of stolen time on logically-partitioned
379 config VIRT_CPU_ACCOUNTING_GEN
380 bool "Full dynticks CPU time accounting"
381 depends on HAVE_CONTEXT_TRACKING
382 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN
383 select VIRT_CPU_ACCOUNTING
384 select CONTEXT_TRACKING
386 Select this option to enable task and CPU time accounting on full
387 dynticks systems. This accounting is implemented by watching every
388 kernel-user boundaries using the context tracking subsystem.
389 The accounting is thus performed at the expense of some significant
392 For now this is only useful if you are working on the full
393 dynticks subsystem development.
399 config IRQ_TIME_ACCOUNTING
400 bool "Fine granularity task level IRQ time accounting"
401 depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE
403 Select this option to enable fine granularity task irq time
404 accounting. This is done by reading a timestamp on each
405 transitions between softirq and hardirq state, so there can be a
406 small performance impact.
408 If in doubt, say N here.
410 config BSD_PROCESS_ACCT
411 bool "BSD Process Accounting"
414 If you say Y here, a user level program will be able to instruct the
415 kernel (via a special system call) to write process accounting
416 information to a file: whenever a process exits, information about
417 that process will be appended to the file by the kernel. The
418 information includes things such as creation time, owning user,
419 command name, memory usage, controlling terminal etc. (the complete
420 list is in the struct acct in <file:include/linux/acct.h>). It is
421 up to the user level program to do useful things with this
422 information. This is generally a good idea, so say Y.
424 config BSD_PROCESS_ACCT_V3
425 bool "BSD Process Accounting version 3 file format"
426 depends on BSD_PROCESS_ACCT
429 If you say Y here, the process accounting information is written
430 in a new file format that also logs the process IDs of each
431 process and it's parent. Note that this file format is incompatible
432 with previous v0/v1/v2 file formats, so you will need updated tools
433 for processing it. A preliminary version of these tools is available
434 at <http://www.gnu.org/software/acct/>.
437 bool "Export task/process statistics through netlink"
442 Export selected statistics for tasks/processes through the
443 generic netlink interface. Unlike BSD process accounting, the
444 statistics are available during the lifetime of tasks/processes as
445 responses to commands. Like BSD accounting, they are sent to user
450 config TASK_DELAY_ACCT
451 bool "Enable per-task delay accounting"
455 Collect information on time spent by a task waiting for system
456 resources like cpu, synchronous block I/O completion and swapping
457 in pages. Such statistics can help in setting a task's priorities
458 relative to other tasks for cpu, io, rss limits etc.
463 bool "Enable extended accounting over taskstats"
466 Collect extended task accounting data and send the data
467 to userland for processing over the taskstats interface.
471 config TASK_IO_ACCOUNTING
472 bool "Enable per-task storage I/O accounting"
473 depends on TASK_XACCT
475 Collect information on the number of bytes of storage I/O which this
480 endmenu # "CPU/Task time and stats accounting"
484 depends on SMP || COMPILE_TEST
487 Make sure that CPUs running critical tasks are not disturbed by
488 any source of "noise" such as unbound workqueues, timers, kthreads...
489 Unbound jobs get offloaded to housekeeping CPUs. This is driven by
490 the "isolcpus=" boot parameter.
494 source "kernel/rcu/Kconfig"
501 tristate "Kernel .config support"
504 This option enables the complete Linux kernel ".config" file
505 contents to be saved in the kernel. It provides documentation
506 of which kernel options are used in a running kernel or in an
507 on-disk kernel. This information can be extracted from the kernel
508 image file with the script scripts/extract-ikconfig and used as
509 input to rebuild the current kernel or to build another kernel.
510 It can also be extracted from a running kernel by reading
511 /proc/config.gz if enabled (below).
514 bool "Enable access to .config through /proc/config.gz"
515 depends on IKCONFIG && PROC_FS
517 This option enables access to the kernel configuration file
518 through /proc/config.gz.
521 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)"
526 Select the minimal kernel log buffer size as a power of 2.
527 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config
528 parameter, see below. Any higher size also might be forced
529 by "log_buf_len" boot parameter.
539 config LOG_CPU_MAX_BUF_SHIFT
540 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)"
543 default 12 if !BASE_SMALL
544 default 0 if BASE_SMALL
547 This option allows to increase the default ring buffer size
548 according to the number of CPUs. The value defines the contribution
549 of each CPU as a power of 2. The used space is typically only few
550 lines however it might be much more when problems are reported,
553 The increased size means that a new buffer has to be allocated and
554 the original static one is unused. It makes sense only on systems
555 with more CPUs. Therefore this value is used only when the sum of
556 contributions is greater than the half of the default kernel ring
557 buffer as defined by LOG_BUF_SHIFT. The default values are set
558 so that more than 64 CPUs are needed to trigger the allocation.
560 Also this option is ignored when "log_buf_len" kernel parameter is
561 used as it forces an exact (power of two) size of the ring buffer.
563 The number of possible CPUs is used for this computation ignoring
564 hotplugging making the computation optimal for the worst case
565 scenario while allowing a simple algorithm to be used from bootup.
567 Examples shift values and their meaning:
568 17 => 128 KB for each CPU
569 16 => 64 KB for each CPU
570 15 => 32 KB for each CPU
571 14 => 16 KB for each CPU
572 13 => 8 KB for each CPU
573 12 => 4 KB for each CPU
575 config PRINTK_SAFE_LOG_BUF_SHIFT
576 int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)"
581 Select the size of an alternate printk per-CPU buffer where messages
582 printed from usafe contexts are temporary stored. One example would
583 be NMI messages, another one - printk recursion. The messages are
584 copied to the main log buffer in a safe context to avoid a deadlock.
585 The value defines the size as a power of 2.
587 Those messages are rare and limited. The largest one is when
588 a backtrace is printed. It usually fits into 4KB. Select
589 8KB if you want to be on the safe side.
592 17 => 128 KB for each CPU
593 16 => 64 KB for each CPU
594 15 => 32 KB for each CPU
595 14 => 16 KB for each CPU
596 13 => 8 KB for each CPU
597 12 => 4 KB for each CPU
600 # Architectures with an unreliable sched_clock() should select this:
602 config HAVE_UNSTABLE_SCHED_CLOCK
605 config GENERIC_SCHED_CLOCK
609 # For architectures that want to enable the support for NUMA-affine scheduler
612 config ARCH_SUPPORTS_NUMA_BALANCING
616 # For architectures that prefer to flush all TLBs after a number of pages
617 # are unmapped instead of sending one IPI per page to flush. The architecture
618 # must provide guarantees on what happens if a clean TLB cache entry is
619 # written after the unmap. Details are in mm/rmap.c near the check for
620 # should_defer_flush. The architecture should also consider if the full flush
621 # and the refill costs are offset by the savings of sending fewer IPIs.
622 config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH
626 # For architectures that know their GCC __int128 support is sound
628 config ARCH_SUPPORTS_INT128
631 # For architectures that (ab)use NUMA to represent different memory regions
632 # all cpu-local but of different latencies, such as SuperH.
634 config ARCH_WANT_NUMA_VARIABLE_LOCALITY
637 config NUMA_BALANCING
638 bool "Memory placement aware NUMA scheduler"
639 depends on ARCH_SUPPORTS_NUMA_BALANCING
640 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY
641 depends on SMP && NUMA && MIGRATION
643 This option adds support for automatic NUMA aware memory/task placement.
644 The mechanism is quite primitive and is based on migrating memory when
645 it has references to the node the task is running on.
647 This system will be inactive on UMA systems.
649 config NUMA_BALANCING_DEFAULT_ENABLED
650 bool "Automatically enable NUMA aware memory/task placement"
652 depends on NUMA_BALANCING
654 If set, automatic NUMA balancing will be enabled if running on a NUMA
658 bool "Control Group support"
661 This option adds support for grouping sets of processes together, for
662 use with process control subsystems such as Cpusets, CFS, memory
663 controls or device isolation.
665 - Documentation/scheduler/sched-design-CFS.txt (CFS)
666 - Documentation/cgroup-v1/ (features for grouping, isolation
667 and resource control)
677 bool "Memory controller"
681 Provides control over the memory footprint of tasks in a cgroup.
684 bool "Swap controller"
685 depends on MEMCG && SWAP
687 Provides control over the swap space consumed by tasks in a cgroup.
689 config MEMCG_SWAP_ENABLED
690 bool "Swap controller enabled by default"
691 depends on MEMCG_SWAP
694 Memory Resource Controller Swap Extension comes with its price in
695 a bigger memory consumption. General purpose distribution kernels
696 which want to enable the feature but keep it disabled by default
697 and let the user enable it by swapaccount=1 boot command line
698 parameter should have this option unselected.
699 For those who want to have the feature enabled by default should
700 select this option (if, for some reason, they need to disable it
701 then swapaccount=0 does the trick).
708 Generic block IO controller cgroup interface. This is the common
709 cgroup interface which should be used by various IO controlling
712 Currently, CFQ IO scheduler uses it to recognize task groups and
713 control disk bandwidth allocation (proportional time slice allocation)
714 to such task groups. It is also used by bio throttling logic in
715 block layer to implement upper limit in IO rates on a device.
717 This option only enables generic Block IO controller infrastructure.
718 One needs to also enable actual IO controlling logic/policy. For
719 enabling proportional weight division of disk bandwidth in CFQ, set
720 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set
721 CONFIG_BLK_DEV_THROTTLING=y.
723 See Documentation/cgroup-v1/blkio-controller.txt for more information.
725 config DEBUG_BLK_CGROUP
726 bool "IO controller debugging"
727 depends on BLK_CGROUP
730 Enable some debugging help. Currently it exports additional stat
731 files in a cgroup which can be useful for debugging.
733 config CGROUP_WRITEBACK
735 depends on MEMCG && BLK_CGROUP
738 menuconfig CGROUP_SCHED
739 bool "CPU controller"
742 This feature lets CPU scheduler recognize task groups and control CPU
743 bandwidth allocation to such task groups. It uses cgroups to group
747 config FAIR_GROUP_SCHED
748 bool "Group scheduling for SCHED_OTHER"
749 depends on CGROUP_SCHED
753 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED"
754 depends on FAIR_GROUP_SCHED
757 This option allows users to define CPU bandwidth rates (limits) for
758 tasks running within the fair group scheduler. Groups with no limit
759 set are considered to be unconstrained and will run with no
761 See Documentation/scheduler/sched-bwc.txt for more information.
763 config RT_GROUP_SCHED
764 bool "Group scheduling for SCHED_RR/FIFO"
765 depends on CGROUP_SCHED
768 This feature lets you explicitly allocate real CPU bandwidth
769 to task groups. If enabled, it will also make it impossible to
770 schedule realtime tasks for non-root users until you allocate
771 realtime bandwidth for them.
772 See Documentation/scheduler/sched-rt-group.txt for more information.
777 bool "PIDs controller"
779 Provides enforcement of process number limits in the scope of a
780 cgroup. Any attempt to fork more processes than is allowed in the
781 cgroup will fail. PIDs are fundamentally a global resource because it
782 is fairly trivial to reach PID exhaustion before you reach even a
783 conservative kmemcg limit. As a result, it is possible to grind a
784 system to halt without being limited by other cgroup policies. The
785 PIDs controller is designed to stop this from happening.
787 It should be noted that organisational operations (such as attaching
788 to a cgroup hierarchy will *not* be blocked by the PIDs controller),
789 since the PIDs limit only affects a process's ability to fork, not to
793 bool "RDMA controller"
795 Provides enforcement of RDMA resources defined by IB stack.
796 It is fairly easy for consumers to exhaust RDMA resources, which
797 can result into resource unavailability to other consumers.
798 RDMA controller is designed to stop this from happening.
799 Attaching processes with active RDMA resources to the cgroup
800 hierarchy is allowed even if can cross the hierarchy's limit.
802 config CGROUP_FREEZER
803 bool "Freezer controller"
805 Provides a way to freeze and unfreeze all tasks in a
808 This option affects the ORIGINAL cgroup interface. The cgroup2 memory
809 controller includes important in-kernel memory consumers per default.
811 If you're using cgroup2, say N.
813 config CGROUP_HUGETLB
814 bool "HugeTLB controller"
815 depends on HUGETLB_PAGE
819 Provides a cgroup controller for HugeTLB pages.
820 When you enable this, you can put a per cgroup limit on HugeTLB usage.
821 The limit is enforced during page fault. Since HugeTLB doesn't
822 support page reclaim, enforcing the limit at page fault time implies
823 that, the application will get SIGBUS signal if it tries to access
824 HugeTLB pages beyond its limit. This requires the application to know
825 beforehand how much HugeTLB pages it would require for its use. The
826 control group is tracked in the third page lru pointer. This means
827 that we cannot use the controller with huge page less than 3 pages.
830 bool "Cpuset controller"
833 This option will let you create and manage CPUSETs which
834 allow dynamically partitioning a system into sets of CPUs and
835 Memory Nodes and assigning tasks to run only within those sets.
836 This is primarily useful on large SMP or NUMA systems.
840 config PROC_PID_CPUSET
841 bool "Include legacy /proc/<pid>/cpuset file"
846 bool "Device controller"
848 Provides a cgroup controller implementing whitelists for
849 devices which a process in the cgroup can mknod or open.
851 config CGROUP_CPUACCT
852 bool "Simple CPU accounting controller"
854 Provides a simple controller for monitoring the
855 total CPU consumed by the tasks in a cgroup.
858 bool "Perf controller"
859 depends on PERF_EVENTS
861 This option extends the perf per-cpu mode to restrict monitoring
862 to threads which belong to the cgroup specified and run on the
868 bool "Support for eBPF programs attached to cgroups"
869 depends on BPF_SYSCALL
870 select SOCK_CGROUP_DATA
872 Allow attaching eBPF programs to a cgroup using the bpf(2)
873 syscall command BPF_PROG_ATTACH.
875 In which context these programs are accessed depends on the type
876 of attachment. For instance, programs that are attached using
877 BPF_CGROUP_INET_INGRESS will be executed on the ingress path of
881 bool "Debug controller"
883 depends on DEBUG_KERNEL
885 This option enables a simple controller that exports
886 debugging information about the cgroups framework. This
887 controller is for control cgroup debugging only. Its
888 interfaces are not stable.
892 config SOCK_CGROUP_DATA
898 menuconfig NAMESPACES
899 bool "Namespaces support" if EXPERT
903 Provides the way to make tasks work with different objects using
904 the same id. For example same IPC id may refer to different objects
905 or same user id or pid may refer to different tasks when used in
906 different namespaces.
914 In this namespace tasks see different info provided with the
919 depends on (SYSVIPC || POSIX_MQUEUE)
922 In this namespace tasks work with IPC ids which correspond to
923 different IPC objects in different namespaces.
926 bool "User namespace"
929 This allows containers, i.e. vservers, to use user namespaces
930 to provide different user info for different servers.
932 When user namespaces are enabled in the kernel it is
933 recommended that the MEMCG option also be enabled and that
934 user-space use the memory control groups to limit the amount
935 of memory a memory unprivileged users can use.
940 bool "PID Namespaces"
943 Support process id namespaces. This allows having multiple
944 processes with the same pid as long as they are in different
945 pid namespaces. This is a building block of containers.
948 bool "Network namespace"
952 Allow user space to create what appear to be multiple instances
953 of the network stack.
957 config SCHED_AUTOGROUP
958 bool "Automatic process group scheduling"
961 select FAIR_GROUP_SCHED
963 This option optimizes the scheduler for common desktop workloads by
964 automatically creating and populating task groups. This separation
965 of workloads isolates aggressive CPU burners (like build jobs) from
966 desktop applications. Task group autogeneration is currently based
969 config SYSFS_DEPRECATED
970 bool "Enable deprecated sysfs features to support old userspace tools"
974 This option adds code that switches the layout of the "block" class
975 devices, to not show up in /sys/class/block/, but only in
978 This switch is only active when the sysfs.deprecated=1 boot option is
979 passed or the SYSFS_DEPRECATED_V2 option is set.
981 This option allows new kernels to run on old distributions and tools,
982 which might get confused by /sys/class/block/. Since 2007/2008 all
983 major distributions and tools handle this just fine.
985 Recent distributions and userspace tools after 2009/2010 depend on
986 the existence of /sys/class/block/, and will not work with this
989 Only if you are using a new kernel on an old distribution, you might
992 config SYSFS_DEPRECATED_V2
993 bool "Enable deprecated sysfs features by default"
996 depends on SYSFS_DEPRECATED
998 Enable deprecated sysfs by default.
1000 See the CONFIG_SYSFS_DEPRECATED option for more details about this
1003 Only if you are using a new kernel on an old distribution, you might
1004 need to say Y here. Even then, odds are you would not need it
1005 enabled, you can always pass the boot option if absolutely necessary.
1008 bool "Kernel->user space relay support (formerly relayfs)"
1011 This option enables support for relay interface support in
1012 certain file systems (such as debugfs).
1013 It is designed to provide an efficient mechanism for tools and
1014 facilities to relay large amounts of data from kernel space to
1019 config BLK_DEV_INITRD
1020 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support"
1022 The initial RAM filesystem is a ramfs which is loaded by the
1023 boot loader (loadlin or lilo) and that is mounted as root
1024 before the normal boot procedure. It is typically used to
1025 load modules needed to mount the "real" root file system,
1026 etc. See <file:Documentation/admin-guide/initrd.rst> for details.
1028 If RAM disk support (BLK_DEV_RAM) is also included, this
1029 also enables initial RAM disk (initrd) support and adds
1030 15 Kbytes (more on some other architectures) to the kernel size.
1036 source "usr/Kconfig"
1041 prompt "Compiler optimization level"
1042 default CC_OPTIMIZE_FOR_PERFORMANCE
1044 config CC_OPTIMIZE_FOR_PERFORMANCE
1045 bool "Optimize for performance"
1047 This is the default optimization level for the kernel, building
1048 with the "-O2" compiler flag for best performance and most
1049 helpful compile-time warnings.
1051 config CC_OPTIMIZE_FOR_SIZE
1052 bool "Optimize for size"
1054 Enabling this option will pass "-Os" instead of "-O2" to
1055 your compiler resulting in a smaller kernel.
1061 config HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1064 This requires that the arch annotates or otherwise protects
1065 its external entry points from being discarded. Linker scripts
1066 must also merge .text.*, .data.*, and .bss.* correctly into
1067 output sections. Care must be taken not to pull in unrelated
1068 sections (e.g., '.text.init'). Typically '.' in section names
1069 is used to distinguish them from label names / C identifiers.
1071 config LD_DEAD_CODE_DATA_ELIMINATION
1072 bool "Dead code and data elimination (EXPERIMENTAL)"
1073 depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION
1076 Enable this if you want to do dead code and data elimination with
1077 the linker by compiling with -ffunction-sections -fdata-sections,
1078 and linking with --gc-sections.
1080 This can reduce on disk and in-memory size of the kernel
1081 code and static data, particularly for small configs and
1082 on small systems. This has the possibility of introducing
1083 silently broken kernel if the required annotations are not
1084 present. This option is not well tested yet, so use at your
1096 config SYSCTL_EXCEPTION_TRACE
1099 Enable support for /proc/sys/debug/exception-trace.
1101 config SYSCTL_ARCH_UNALIGN_NO_WARN
1104 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap
1105 Allows arch to define/use @no_unaligned_warning to possibly warn
1106 about unaligned access emulation going on under the hood.
1108 config SYSCTL_ARCH_UNALIGN_ALLOW
1111 Enable support for /proc/sys/kernel/unaligned-trap
1112 Allows arches to define/use @unaligned_enabled to runtime toggle
1113 the unaligned access emulation.
1114 see arch/parisc/kernel/unaligned.c for reference
1116 config HAVE_PCSPKR_PLATFORM
1119 # interpreter that classic socket filters depend on
1124 bool "Configure standard kernel features (expert users)"
1125 # Unhide debug options, to make the on-by-default options visible
1128 This option allows certain base kernel options and settings
1129 to be disabled or tweaked. This is for specialized
1130 environments which can tolerate a "non-standard" kernel.
1131 Only use this if you really know what you are doing.
1134 bool "Enable 16-bit UID system calls" if EXPERT
1135 depends on HAVE_UID16 && MULTIUSER
1138 This enables the legacy 16-bit UID syscall wrappers.
1141 bool "Multiple users, groups and capabilities support" if EXPERT
1144 This option enables support for non-root users, groups and
1147 If you say N here, all processes will run with UID 0, GID 0, and all
1148 possible capabilities. Saying N here also compiles out support for
1149 system calls related to UIDs, GIDs, and capabilities, such as setuid,
1152 If unsure, say Y here.
1154 config SGETMASK_SYSCALL
1155 bool "sgetmask/ssetmask syscalls support" if EXPERT
1156 def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH
1158 sys_sgetmask and sys_ssetmask are obsolete system calls
1159 no longer supported in libc but still enabled by default in some
1162 If unsure, leave the default option here.
1164 config SYSFS_SYSCALL
1165 bool "Sysfs syscall support" if EXPERT
1168 sys_sysfs is an obsolete system call no longer supported in libc.
1169 Note that disabling this option is more secure but might break
1170 compatibility with some systems.
1172 If unsure say Y here.
1174 config SYSCTL_SYSCALL
1175 bool "Sysctl syscall support" if EXPERT
1176 depends on PROC_SYSCTL
1180 sys_sysctl uses binary paths that have been found challenging
1181 to properly maintain and use. The interface in /proc/sys
1182 using paths with ascii names is now the primary path to this
1185 Almost nothing using the binary sysctl interface so if you are
1186 trying to save some space it is probably safe to disable this,
1187 making your kernel marginally smaller.
1189 If unsure say N here.
1192 bool "open by fhandle syscalls" if EXPERT
1196 If you say Y here, a user level program will be able to map
1197 file names to handle and then later use the handle for
1198 different file system operations. This is useful in implementing
1199 userspace file servers, which now track files using handles instead
1200 of names. The handle would remain the same even if file names
1201 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2)
1205 bool "Posix Clocks & timers" if EXPERT
1208 This includes native support for POSIX timers to the kernel.
1209 Some embedded systems have no use for them and therefore they
1210 can be configured out to reduce the size of the kernel image.
1212 When this option is disabled, the following syscalls won't be
1213 available: timer_create, timer_gettime: timer_getoverrun,
1214 timer_settime, timer_delete, clock_adjtime, getitimer,
1215 setitimer, alarm. Furthermore, the clock_settime, clock_gettime,
1216 clock_getres and clock_nanosleep syscalls will be limited to
1217 CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only.
1223 bool "Enable support for printk" if EXPERT
1226 This option enables normal printk support. Removing it
1227 eliminates most of the message strings from the kernel image
1228 and makes the kernel more or less silent. As this makes it
1229 very difficult to diagnose system problems, saying N here is
1230 strongly discouraged.
1238 bool "BUG() support" if EXPERT
1241 Disabling this option eliminates support for BUG and WARN, reducing
1242 the size of your kernel image and potentially quietly ignoring
1243 numerous fatal conditions. You should only consider disabling this
1244 option for embedded systems with no facilities for reporting errors.
1250 bool "Enable ELF core dumps" if EXPERT
1252 Enable support for generating core dumps. Disabling saves about 4k.
1255 config PCSPKR_PLATFORM
1256 bool "Enable PC-Speaker support" if EXPERT
1257 depends on HAVE_PCSPKR_PLATFORM
1261 This option allows to disable the internal PC-Speaker
1262 support, saving some memory.
1266 bool "Enable full-sized data structures for core" if EXPERT
1268 Disabling this option reduces the size of miscellaneous core
1269 kernel data structures. This saves memory on small machines,
1270 but may reduce performance.
1273 bool "Enable futex support" if EXPERT
1277 Disabling this option will cause the kernel to be built without
1278 support for "fast userspace mutexes". The resulting kernel may not
1279 run glibc-based applications correctly.
1283 depends on FUTEX && RT_MUTEXES
1286 config HAVE_FUTEX_CMPXCHG
1290 Architectures should select this if futex_atomic_cmpxchg_inatomic()
1291 is implemented and always working. This removes a couple of runtime
1295 bool "Enable eventpoll support" if EXPERT
1299 Disabling this option will cause the kernel to be built without
1300 support for epoll family of system calls.
1303 bool "Enable signalfd() system call" if EXPERT
1307 Enable the signalfd() system call that allows to receive signals
1308 on a file descriptor.
1313 bool "Enable timerfd() system call" if EXPERT
1317 Enable the timerfd() system call that allows to receive timer
1318 events on a file descriptor.
1323 bool "Enable eventfd() system call" if EXPERT
1327 Enable the eventfd() system call that allows to receive both
1328 kernel notification (ie. KAIO) or userspace notifications.
1333 bool "Use full shmem filesystem" if EXPERT
1337 The shmem is an internal filesystem used to manage shared memory.
1338 It is backed by swap and manages resource limits. It is also exported
1339 to userspace as tmpfs if TMPFS is enabled. Disabling this
1340 option replaces shmem and tmpfs with the much simpler ramfs code,
1341 which may be appropriate on small systems without swap.
1344 bool "Enable AIO support" if EXPERT
1347 This option enables POSIX asynchronous I/O which may by used
1348 by some high performance threaded applications. Disabling
1349 this option saves about 7k.
1351 config ADVISE_SYSCALLS
1352 bool "Enable madvise/fadvise syscalls" if EXPERT
1355 This option enables the madvise and fadvise syscalls, used by
1356 applications to advise the kernel about their future memory or file
1357 usage, improving performance. If building an embedded system where no
1358 applications use these syscalls, you can disable this option to save
1362 bool "Enable membarrier() system call" if EXPERT
1365 Enable the membarrier() system call that allows issuing memory
1366 barriers across all running threads, which can be used to distribute
1367 the cost of user-space memory barriers asymmetrically by transforming
1368 pairs of memory barriers into pairs consisting of membarrier() and a
1373 config CHECKPOINT_RESTORE
1374 bool "Checkpoint/restore support" if EXPERT
1375 select PROC_CHILDREN
1378 Enables additional kernel features in a sake of checkpoint/restore.
1379 In particular it adds auxiliary prctl codes to setup process text,
1380 data and heap segment sizes, and a few additional /proc filesystem
1383 If unsure, say N here.
1386 bool "Load all symbols for debugging/ksymoops" if EXPERT
1389 Say Y here to let the kernel print out symbolic crash information and
1390 symbolic stack backtraces. This increases the size of the kernel
1391 somewhat, as all symbols have to be loaded into the kernel image.
1394 bool "Include all symbols in kallsyms"
1395 depends on DEBUG_KERNEL && KALLSYMS
1397 Normally kallsyms only contains the symbols of functions for nicer
1398 OOPS messages and backtraces (i.e., symbols from the text and inittext
1399 sections). This is sufficient for most cases. And only in very rare
1400 cases (e.g., when a debugger is used) all symbols are required (e.g.,
1401 names of variables from the data sections, etc).
1403 This option makes sure that all symbols are loaded into the kernel
1404 image (i.e., symbols from all sections) in cost of increased kernel
1405 size (depending on the kernel configuration, it may be 300KiB or
1406 something like this).
1408 Say N unless you really need all symbols.
1410 config KALLSYMS_ABSOLUTE_PERCPU
1413 default X86_64 && SMP
1415 config KALLSYMS_BASE_RELATIVE
1420 Instead of emitting them as absolute values in the native word size,
1421 emit the symbol references in the kallsyms table as 32-bit entries,
1422 each containing a relative value in the range [base, base + U32_MAX]
1423 or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either
1424 an absolute value in the range [0, S32_MAX] or a relative value in the
1425 range [base, base + S32_MAX], where base is the lowest relative symbol
1426 address encountered in the image.
1428 On 64-bit builds, this reduces the size of the address table by 50%,
1429 but more importantly, it results in entries whose values are build
1430 time constants, and no relocation pass is required at runtime to fix
1431 up the entries based on the runtime load address of the kernel.
1433 # end of the "standard kernel features (expert users)" menu
1435 # syscall, maps, verifier
1437 bool "Enable bpf() system call"
1443 Enable the bpf() system call that allows to manipulate eBPF
1444 programs and maps via file descriptors.
1446 config BPF_JIT_ALWAYS_ON
1447 bool "Permanently enable BPF JIT and remove BPF interpreter"
1448 depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT
1450 Enables BPF JIT and removes BPF interpreter to avoid
1451 speculative execution of BPF instructions by the interpreter
1454 bool "Enable userfaultfd() system call"
1458 Enable the userfaultfd() system call that allows to intercept and
1459 handle page faults in userland.
1461 config ARCH_HAS_MEMBARRIER_CALLBACKS
1464 config ARCH_HAS_MEMBARRIER_SYNC_CORE
1468 bool "Enable rseq() system call" if EXPERT
1470 depends on HAVE_RSEQ
1473 Enable the restartable sequences system call. It provides a
1474 user-space cache for the current CPU number value, which
1475 speeds up getting the current CPU number from user-space,
1476 as well as an ABI to speed up user-space operations on
1483 bool "Enabled debugging of rseq() system call" if EXPERT
1484 depends on RSEQ && DEBUG_KERNEL
1486 Enable extra debugging checks for the rseq system call.
1491 bool "Embedded system"
1492 option allnoconfig_y
1495 This option should be enabled if compiling the kernel for
1496 an embedded system so certain expert options are available
1499 config HAVE_PERF_EVENTS
1502 See tools/perf/design.txt for details.
1504 config PERF_USE_VMALLOC
1507 See tools/perf/design.txt for details
1510 bool "PC/104 support" if EXPERT
1512 Expose PC/104 form factor device drivers and options available for
1513 selection and configuration. Enable this option if your target
1514 machine has a PC/104 bus.
1516 menu "Kernel Performance Events And Counters"
1519 bool "Kernel performance events and counters"
1520 default y if PROFILING
1521 depends on HAVE_PERF_EVENTS
1526 Enable kernel support for various performance events provided
1527 by software and hardware.
1529 Software events are supported either built-in or via the
1530 use of generic tracepoints.
1532 Most modern CPUs support performance events via performance
1533 counter registers. These registers count the number of certain
1534 types of hw events: such as instructions executed, cachemisses
1535 suffered, or branches mis-predicted - without slowing down the
1536 kernel or applications. These registers can also trigger interrupts
1537 when a threshold number of events have passed - and can thus be
1538 used to profile the code that runs on that CPU.
1540 The Linux Performance Event subsystem provides an abstraction of
1541 these software and hardware event capabilities, available via a
1542 system call and used by the "perf" utility in tools/perf/. It
1543 provides per task and per CPU counters, and it provides event
1544 capabilities on top of those.
1548 config DEBUG_PERF_USE_VMALLOC
1550 bool "Debug: use vmalloc to back perf mmap() buffers"
1551 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC
1552 select PERF_USE_VMALLOC
1554 Use vmalloc memory to back perf mmap() buffers.
1556 Mostly useful for debugging the vmalloc code on platforms
1557 that don't require it.
1563 config VM_EVENT_COUNTERS
1565 bool "Enable VM event counters for /proc/vmstat" if EXPERT
1567 VM event counters are needed for event counts to be shown.
1568 This option allows the disabling of the VM event counters
1569 on EXPERT systems. /proc/vmstat will only show page counts
1570 if VM event counters are disabled.
1574 bool "Enable SLUB debugging support" if EXPERT
1575 depends on SLUB && SYSFS
1577 SLUB has extensive debug support features. Disabling these can
1578 result in significant savings in code size. This also disables
1579 SLUB sysfs support. /sys/slab will not exist and there will be
1580 no support for cache validation etc.
1582 config SLUB_MEMCG_SYSFS_ON
1584 bool "Enable memcg SLUB sysfs support by default" if EXPERT
1585 depends on SLUB && SYSFS && MEMCG
1587 SLUB creates a directory under /sys/kernel/slab for each
1588 allocation cache to host info and debug files. If memory
1589 cgroup is enabled, each cache can have per memory cgroup
1590 caches. SLUB can create the same sysfs directories for these
1591 caches under /sys/kernel/slab/CACHE/cgroup but it can lead
1592 to a very high number of debug files being created. This is
1593 controlled by slub_memcg_sysfs boot parameter and this
1594 config option determines the parameter's default value.
1597 bool "Disable heap randomization"
1600 Randomizing heap placement makes heap exploits harder, but it
1601 also breaks ancient binaries (including anything libc5 based).
1602 This option changes the bootup default to heap randomization
1603 disabled, and can be overridden at runtime by setting
1604 /proc/sys/kernel/randomize_va_space to 2.
1606 On non-ancient distros (post-2000 ones) N is usually a safe choice.
1609 prompt "Choose SLAB allocator"
1612 This option allows to select a slab allocator.
1616 select HAVE_HARDENED_USERCOPY_ALLOCATOR
1618 The regular slab allocator that is established and known to work
1619 well in all environments. It organizes cache hot objects in
1620 per cpu and per node queues.
1623 bool "SLUB (Unqueued Allocator)"
1624 select HAVE_HARDENED_USERCOPY_ALLOCATOR
1626 SLUB is a slab allocator that minimizes cache line usage
1627 instead of managing queues of cached objects (SLAB approach).
1628 Per cpu caching is realized using slabs of objects instead
1629 of queues of objects. SLUB can use memory efficiently
1630 and has enhanced diagnostics. SLUB is the default choice for
1635 bool "SLOB (Simple Allocator)"
1637 SLOB replaces the stock allocator with a drastically simpler
1638 allocator. SLOB is generally more space efficient but
1639 does not perform as well on large systems.
1643 config SLAB_MERGE_DEFAULT
1644 bool "Allow slab caches to be merged"
1647 For reduced kernel memory fragmentation, slab caches can be
1648 merged when they share the same size and other characteristics.
1649 This carries a risk of kernel heap overflows being able to
1650 overwrite objects from merged caches (and more easily control
1651 cache layout), which makes such heap attacks easier to exploit
1652 by attackers. By keeping caches unmerged, these kinds of exploits
1653 can usually only damage objects in the same cache. To disable
1654 merging at runtime, "slab_nomerge" can be passed on the kernel
1657 config SLAB_FREELIST_RANDOM
1659 depends on SLAB || SLUB
1660 bool "SLAB freelist randomization"
1662 Randomizes the freelist order used on creating new pages. This
1663 security feature reduces the predictability of the kernel slab
1664 allocator against heap overflows.
1666 config SLAB_FREELIST_HARDENED
1667 bool "Harden slab freelist metadata"
1670 Many kernel heap attacks try to target slab cache metadata and
1671 other infrastructure. This options makes minor performance
1672 sacrifies to harden the kernel slab allocator against common
1673 freelist exploit methods.
1675 config SLUB_CPU_PARTIAL
1677 depends on SLUB && SMP
1678 bool "SLUB per cpu partial cache"
1680 Per cpu partial caches accellerate objects allocation and freeing
1681 that is local to a processor at the price of more indeterminism
1682 in the latency of the free. On overflow these caches will be cleared
1683 which requires the taking of locks that may cause latency spikes.
1684 Typically one would choose no for a realtime system.
1686 config MMAP_ALLOW_UNINITIALIZED
1687 bool "Allow mmapped anonymous memory to be uninitialized"
1688 depends on EXPERT && !MMU
1691 Normally, and according to the Linux spec, anonymous memory obtained
1692 from mmap() has it's contents cleared before it is passed to
1693 userspace. Enabling this config option allows you to request that
1694 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
1695 providing a huge performance boost. If this option is not enabled,
1696 then the flag will be ignored.
1698 This is taken advantage of by uClibc's malloc(), and also by
1699 ELF-FDPIC binfmt's brk and stack allocator.
1701 Because of the obvious security issues, this option should only be
1702 enabled on embedded devices where you control what is run in
1703 userspace. Since that isn't generally a problem on no-MMU systems,
1704 it is normally safe to say Y here.
1706 See Documentation/nommu-mmap.txt for more information.
1708 config SYSTEM_DATA_VERIFICATION
1710 select SYSTEM_TRUSTED_KEYRING
1714 select ASYMMETRIC_KEY_TYPE
1715 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE
1718 select X509_CERTIFICATE_PARSER
1719 select PKCS7_MESSAGE_PARSER
1721 Provide PKCS#7 message verification using the contents of the system
1722 trusted keyring to provide public keys. This then can be used for
1723 module verification, kexec image verification and firmware blob
1727 bool "Profiling support"
1729 Say Y here to enable the extended profiling support mechanisms used
1730 by profilers such as OProfile.
1733 # Place an empty function call at each tracepoint site. Can be
1734 # dynamically changed for a probe function.
1739 source "arch/Kconfig"
1741 endmenu # General setup
1748 default 0 if BASE_FULL
1749 default 1 if !BASE_FULL
1752 bool "Enable loadable module support"
1755 Kernel modules are small pieces of compiled code which can
1756 be inserted in the running kernel, rather than being
1757 permanently built into the kernel. You use the "modprobe"
1758 tool to add (and sometimes remove) them. If you say Y here,
1759 many parts of the kernel can be built as modules (by
1760 answering M instead of Y where indicated): this is most
1761 useful for infrequently used options which are not required
1762 for booting. For more information, see the man pages for
1763 modprobe, lsmod, modinfo, insmod and rmmod.
1765 If you say Y here, you will need to run "make
1766 modules_install" to put the modules under /lib/modules/
1767 where modprobe can find them (you may need to be root to do
1774 config MODULE_FORCE_LOAD
1775 bool "Forced module loading"
1778 Allow loading of modules without version information (ie. modprobe
1779 --force). Forced module loading sets the 'F' (forced) taint flag and
1780 is usually a really bad idea.
1782 config MODULE_UNLOAD
1783 bool "Module unloading"
1785 Without this option you will not be able to unload any
1786 modules (note that some modules may not be unloadable
1787 anyway), which makes your kernel smaller, faster
1788 and simpler. If unsure, say Y.
1790 config MODULE_FORCE_UNLOAD
1791 bool "Forced module unloading"
1792 depends on MODULE_UNLOAD
1794 This option allows you to force a module to unload, even if the
1795 kernel believes it is unsafe: the kernel will remove the module
1796 without waiting for anyone to stop using it (using the -f option to
1797 rmmod). This is mainly for kernel developers and desperate users.
1801 bool "Module versioning support"
1803 Usually, you have to use modules compiled with your kernel.
1804 Saying Y here makes it sometimes possible to use modules
1805 compiled for different kernels, by adding enough information
1806 to the modules to (hopefully) spot any changes which would
1807 make them incompatible with the kernel you are running. If
1810 config MODULE_REL_CRCS
1812 depends on MODVERSIONS
1814 config MODULE_SRCVERSION_ALL
1815 bool "Source checksum for all modules"
1817 Modules which contain a MODULE_VERSION get an extra "srcversion"
1818 field inserted into their modinfo section, which contains a
1819 sum of the source files which made it. This helps maintainers
1820 see exactly which source was used to build a module (since
1821 others sometimes change the module source without updating
1822 the version). With this option, such a "srcversion" field
1823 will be created for all modules. If unsure, say N.
1826 bool "Module signature verification"
1828 select SYSTEM_DATA_VERIFICATION
1830 Check modules for valid signatures upon load: the signature
1831 is simply appended to the module. For more information see
1832 <file:Documentation/admin-guide/module-signing.rst>.
1834 Note that this option adds the OpenSSL development packages as a
1835 kernel build dependency so that the signing tool can use its crypto
1838 !!!WARNING!!! If you enable this option, you MUST make sure that the
1839 module DOES NOT get stripped after being signed. This includes the
1840 debuginfo strip done by some packagers (such as rpmbuild) and
1841 inclusion into an initramfs that wants the module size reduced.
1843 config MODULE_SIG_FORCE
1844 bool "Require modules to be validly signed"
1845 depends on MODULE_SIG
1847 Reject unsigned modules or signed modules for which we don't have a
1848 key. Without this, such modules will simply taint the kernel.
1850 config MODULE_SIG_ALL
1851 bool "Automatically sign all modules"
1853 depends on MODULE_SIG
1855 Sign all modules during make modules_install. Without this option,
1856 modules must be signed manually, using the scripts/sign-file tool.
1858 comment "Do not forget to sign required modules with scripts/sign-file"
1859 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL
1862 prompt "Which hash algorithm should modules be signed with?"
1863 depends on MODULE_SIG
1865 This determines which sort of hashing algorithm will be used during
1866 signature generation. This algorithm _must_ be built into the kernel
1867 directly so that signature verification can take place. It is not
1868 possible to load a signed module containing the algorithm to check
1869 the signature on that module.
1871 config MODULE_SIG_SHA1
1872 bool "Sign modules with SHA-1"
1875 config MODULE_SIG_SHA224
1876 bool "Sign modules with SHA-224"
1877 select CRYPTO_SHA256
1879 config MODULE_SIG_SHA256
1880 bool "Sign modules with SHA-256"
1881 select CRYPTO_SHA256
1883 config MODULE_SIG_SHA384
1884 bool "Sign modules with SHA-384"
1885 select CRYPTO_SHA512
1887 config MODULE_SIG_SHA512
1888 bool "Sign modules with SHA-512"
1889 select CRYPTO_SHA512
1893 config MODULE_SIG_HASH
1895 depends on MODULE_SIG
1896 default "sha1" if MODULE_SIG_SHA1
1897 default "sha224" if MODULE_SIG_SHA224
1898 default "sha256" if MODULE_SIG_SHA256
1899 default "sha384" if MODULE_SIG_SHA384
1900 default "sha512" if MODULE_SIG_SHA512
1902 config MODULE_COMPRESS
1903 bool "Compress modules on installation"
1907 Compresses kernel modules when 'make modules_install' is run; gzip or
1908 xz depending on "Compression algorithm" below.
1910 module-init-tools MAY support gzip, and kmod MAY support gzip and xz.
1912 Out-of-tree kernel modules installed using Kbuild will also be
1913 compressed upon installation.
1915 Note: for modules inside an initrd or initramfs, it's more efficient
1916 to compress the whole initrd or initramfs instead.
1918 Note: This is fully compatible with signed modules.
1923 prompt "Compression algorithm"
1924 depends on MODULE_COMPRESS
1925 default MODULE_COMPRESS_GZIP
1927 This determines which sort of compression will be used during
1928 'make modules_install'.
1930 GZIP (default) and XZ are supported.
1932 config MODULE_COMPRESS_GZIP
1935 config MODULE_COMPRESS_XZ
1940 config TRIM_UNUSED_KSYMS
1941 bool "Trim unused exported kernel symbols"
1942 depends on MODULES && !UNUSED_SYMBOLS
1944 The kernel and some modules make many symbols available for
1945 other modules to use via EXPORT_SYMBOL() and variants. Depending
1946 on the set of modules being selected in your kernel configuration,
1947 many of those exported symbols might never be used.
1949 This option allows for unused exported symbols to be dropped from
1950 the build. In turn, this provides the compiler more opportunities
1951 (especially when using LTO) for optimizing the code and reducing
1952 binary size. This might have some security advantages as well.
1954 If unsure, or if you need to build out-of-tree modules, say N.
1958 config MODULES_TREE_LOOKUP
1960 depends on PERF_EVENTS || TRACING
1962 config INIT_ALL_POSSIBLE
1965 Back when each arch used to define their own cpu_online_mask and
1966 cpu_possible_mask, some of them chose to initialize cpu_possible_mask
1967 with all 1s, and others with all 0s. When they were centralised,
1968 it was better to provide this option than to break all the archs
1969 and have several arch maintainers pursuing me down dark alleys.
1971 source "block/Kconfig"
1973 config PREEMPT_NOTIFIERS
1983 Build a simple ASN.1 grammar compiler that produces a bytecode output
1984 that can be interpreted by the ASN.1 stream decoder and used to
1985 inform it as to what tags are to be expected in a stream and what
1986 functions to call on what tags.
1988 source "kernel/Kconfig.locks"
1990 config ARCH_HAS_SYNC_CORE_BEFORE_USERMODE
1993 # It may be useful for an architecture to override the definitions of the
1994 # SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h>
1995 # and the COMPAT_ variants in <linux/compat.h>, in particular to use a
1996 # different calling convention for syscalls. They can also override the
1997 # macros for not-implemented syscalls in kernel/sys_ni.c and
1998 # kernel/time/posix-stubs.c. All these overrides need to be available in
1999 # <asm/syscall_wrapper.h>.
2000 config ARCH_HAS_SYSCALL_WRAPPER
projects / linux.git / blob