1 # SPDX-License-Identifier: GPL-2.0
3 # General architecture dependent options
7 # Note: arch/$(SRCARCH)/Kconfig needs to be included first so that it can
8 # override the default values in this file.
10 source "arch/$(SRCARCH)/Kconfig"
12 menu "General architecture-dependent options"
34 tristate "OProfile system profiling"
36 depends on HAVE_OPROFILE
38 select RING_BUFFER_ALLOW_SWAP
40 OProfile is a profiling system capable of profiling the
41 whole system, include the kernel, kernel modules, libraries,
46 config OPROFILE_EVENT_MULTIPLEX
47 bool "OProfile multiplexing support (EXPERIMENTAL)"
49 depends on OPROFILE && X86
51 The number of hardware counters is limited. The multiplexing
52 feature enables OProfile to gather more events than counters
53 are provided by the hardware. This is realized by switching
54 between events at a user specified time interval.
61 config OPROFILE_NMI_TIMER
63 depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI && !PPC64
68 depends on HAVE_KPROBES
71 Kprobes allows you to trap at almost any kernel address and
72 execute a callback function. register_kprobe() establishes
73 a probepoint and specifies the callback. Kprobes is useful
74 for kernel debugging, non-intrusive instrumentation and testing.
78 bool "Optimize very unlikely/likely branches"
79 depends on HAVE_ARCH_JUMP_LABEL
80 depends on CC_HAS_ASM_GOTO
82 This option enables a transparent branch optimization that
83 makes certain almost-always-true or almost-always-false branch
84 conditions even cheaper to execute within the kernel.
86 Certain performance-sensitive kernel code, such as trace points,
87 scheduler functionality, networking code and KVM have such
88 branches and include support for this optimization technique.
90 If it is detected that the compiler has support for "asm goto",
91 the kernel will compile such branches with just a nop
92 instruction. When the condition flag is toggled to true, the
93 nop will be converted to a jump instruction to execute the
94 conditional block of instructions.
96 This technique lowers overhead and stress on the branch prediction
97 of the processor and generally makes the kernel faster. The update
98 of the condition is slower, but those are always very rare.
100 ( On 32-bit x86, the necessary options added to the compiler
101 flags may increase the size of the kernel slightly. )
103 config STATIC_KEYS_SELFTEST
104 bool "Static key selftest"
105 depends on JUMP_LABEL
107 Boot time self-test of the branch patching code.
109 config STATIC_CALL_SELFTEST
110 bool "Static call selftest"
111 depends on HAVE_STATIC_CALL
113 Boot time self-test of the call patching code.
117 depends on KPROBES && HAVE_OPTPROBES
118 select TASKS_RCU if PREEMPTION
120 config KPROBES_ON_FTRACE
122 depends on KPROBES && HAVE_KPROBES_ON_FTRACE
123 depends on DYNAMIC_FTRACE_WITH_REGS
125 If function tracer is enabled and the arch supports full
126 passing of pt_regs to function tracing, then kprobes can
127 optimize on top of function tracing.
131 depends on ARCH_SUPPORTS_UPROBES
133 Uprobes is the user-space counterpart to kprobes: they
134 enable instrumentation applications (such as 'perf probe')
135 to establish unintrusive probes in user-space binaries and
136 libraries, by executing handler functions when the probes
137 are hit by user-space applications.
139 ( These probes come in the form of single-byte breakpoints,
140 managed by the kernel and kept transparent to the probed
143 config HAVE_EFFICIENT_UNALIGNED_ACCESS
146 Some architectures are unable to perform unaligned accesses
147 without the use of get_unaligned/put_unaligned. Others are
148 unable to perform such accesses efficiently (e.g. trap on
149 unaligned access and require fixing it up in the exception
152 This symbol should be selected by an architecture if it can
153 perform unaligned accesses efficiently to allow different
154 code paths to be selected for these cases. Some network
155 drivers, for example, could opt to not fix up alignment
156 problems with received packets if doing so would not help
159 See Documentation/core-api/unaligned-memory-access.rst for more
160 information on the topic of unaligned memory accesses.
162 config ARCH_USE_BUILTIN_BSWAP
165 Modern versions of GCC (since 4.4) have builtin functions
166 for handling byte-swapping. Using these, instead of the old
167 inline assembler that the architecture code provides in the
168 __arch_bswapXX() macros, allows the compiler to see what's
169 happening and offers more opportunity for optimisation. In
170 particular, the compiler will be able to combine the byteswap
171 with a nearby load or store and use load-and-swap or
172 store-and-swap instructions if the architecture has them. It
173 should almost *never* result in code which is worse than the
174 hand-coded assembler in <asm/swab.h>. But just in case it
175 does, the use of the builtins is optional.
177 Any architecture with load-and-swap or store-and-swap
178 instructions should set this. And it shouldn't hurt to set it
179 on architectures that don't have such instructions.
183 depends on KPROBES && HAVE_KRETPROBES
185 config USER_RETURN_NOTIFIER
187 depends on HAVE_USER_RETURN_NOTIFIER
189 Provide a kernel-internal notification when a cpu is about to
192 config HAVE_IOREMAP_PROT
198 config HAVE_KRETPROBES
201 config HAVE_OPTPROBES
204 config HAVE_KPROBES_ON_FTRACE
207 config HAVE_FUNCTION_ERROR_INJECTION
214 # An arch should select this if it provides all these things:
216 # task_pt_regs() in asm/processor.h or asm/ptrace.h
217 # arch_has_single_step() if there is hardware single-step support
218 # arch_has_block_step() if there is hardware block-step support
219 # asm/syscall.h supplying asm-generic/syscall.h interface
220 # linux/regset.h user_regset interfaces
221 # CORE_DUMP_USE_REGSET #define'd in linux/elf.h
222 # TIF_SYSCALL_TRACE calls tracehook_report_syscall_{entry,exit}
223 # TIF_NOTIFY_RESUME calls tracehook_notify_resume()
224 # signal delivery calls tracehook_signal_handler()
226 config HAVE_ARCH_TRACEHOOK
229 config HAVE_DMA_CONTIGUOUS
232 config GENERIC_SMP_IDLE_THREAD
235 config GENERIC_IDLE_POLL_SETUP
238 config ARCH_HAS_FORTIFY_SOURCE
241 An architecture should select this when it can successfully
242 build and run with CONFIG_FORTIFY_SOURCE.
245 # Select if the arch provides a historic keepinit alias for the retain_initrd
246 # command line option
248 config ARCH_HAS_KEEPINITRD
251 # Select if arch has all set_memory_ro/rw/x/nx() functions in asm/cacheflush.h
252 config ARCH_HAS_SET_MEMORY
255 # Select if arch has all set_direct_map_invalid/default() functions
256 config ARCH_HAS_SET_DIRECT_MAP
260 # Select if the architecture provides the arch_dma_set_uncached symbol to
261 # either provide an uncached segement alias for a DMA allocation, or
262 # to remap the page tables in place.
264 config ARCH_HAS_DMA_SET_UNCACHED
268 # Select if the architectures provides the arch_dma_clear_uncached symbol
269 # to undo an in-place page table remap for uncached access.
271 config ARCH_HAS_DMA_CLEAR_UNCACHED
274 # Select if arch init_task must go in the __init_task_data section
275 config ARCH_TASK_STRUCT_ON_STACK
278 # Select if arch has its private alloc_task_struct() function
279 config ARCH_TASK_STRUCT_ALLOCATOR
282 config HAVE_ARCH_THREAD_STRUCT_WHITELIST
284 depends on !ARCH_TASK_STRUCT_ALLOCATOR
286 An architecture should select this to provide hardened usercopy
287 knowledge about what region of the thread_struct should be
288 whitelisted for copying to userspace. Normally this is only the
289 FPU registers. Specifically, arch_thread_struct_whitelist()
290 should be implemented. Without this, the entire thread_struct
291 field in task_struct will be left whitelisted.
293 # Select if arch has its private alloc_thread_stack() function
294 config ARCH_THREAD_STACK_ALLOCATOR
297 # Select if arch wants to size task_struct dynamically via arch_task_struct_size:
298 config ARCH_WANTS_DYNAMIC_TASK_STRUCT
301 config ARCH_32BIT_OFF_T
305 All new 32-bit architectures should have 64-bit off_t type on
306 userspace side which corresponds to the loff_t kernel type. This
307 is the requirement for modern ABIs. Some existing architectures
308 still support 32-bit off_t. This option is enabled for all such
309 architectures explicitly.
311 config HAVE_ASM_MODVERSIONS
314 This symbol should be selected by an architecure if it provides
315 <asm/asm-prototypes.h> to support the module versioning for symbols
316 exported from assembly code.
318 config HAVE_REGS_AND_STACK_ACCESS_API
321 This symbol should be selected by an architecure if it supports
322 the API needed to access registers and stack entries from pt_regs,
323 declared in asm/ptrace.h
324 For example the kprobes-based event tracer needs this API.
328 depends on HAVE_REGS_AND_STACK_ACCESS_API
330 This symbol should be selected by an architecture if it
331 supports an implementation of restartable sequences.
333 config HAVE_FUNCTION_ARG_ACCESS_API
336 This symbol should be selected by an architecure if it supports
337 the API needed to access function arguments from pt_regs,
338 declared in asm/ptrace.h
340 config HAVE_HW_BREAKPOINT
342 depends on PERF_EVENTS
344 config HAVE_MIXED_BREAKPOINTS_REGS
346 depends on HAVE_HW_BREAKPOINT
348 Depending on the arch implementation of hardware breakpoints,
349 some of them have separate registers for data and instruction
350 breakpoints addresses, others have mixed registers to store
351 them but define the access type in a control register.
352 Select this option if your arch implements breakpoints under the
355 config HAVE_USER_RETURN_NOTIFIER
358 config HAVE_PERF_EVENTS_NMI
361 System hardware can generate an NMI using the perf event
362 subsystem. Also has support for calculating CPU cycle events
363 to determine how many clock cycles in a given period.
365 config HAVE_HARDLOCKUP_DETECTOR_PERF
367 depends on HAVE_PERF_EVENTS_NMI
369 The arch chooses to use the generic perf-NMI-based hardlockup
370 detector. Must define HAVE_PERF_EVENTS_NMI.
372 config HAVE_NMI_WATCHDOG
376 The arch provides a low level NMI watchdog. It provides
377 asm/nmi.h, and defines its own arch_touch_nmi_watchdog().
379 config HAVE_HARDLOCKUP_DETECTOR_ARCH
381 select HAVE_NMI_WATCHDOG
383 The arch chooses to provide its own hardlockup detector, which is
384 a superset of the HAVE_NMI_WATCHDOG. It also conforms to config
385 interfaces and parameters provided by hardlockup detector subsystem.
387 config HAVE_PERF_REGS
390 Support selective register dumps for perf events. This includes
391 bit-mapping of each registers and a unique architecture id.
393 config HAVE_PERF_USER_STACK_DUMP
396 Support user stack dumps for perf event samples. This needs
397 access to the user stack pointer which is not unified across
400 config HAVE_ARCH_JUMP_LABEL
403 config HAVE_ARCH_JUMP_LABEL_RELATIVE
406 config MMU_GATHER_TABLE_FREE
409 config MMU_GATHER_RCU_TABLE_FREE
411 select MMU_GATHER_TABLE_FREE
413 config MMU_GATHER_PAGE_SIZE
416 config MMU_GATHER_NO_RANGE
419 config MMU_GATHER_NO_GATHER
421 depends on MMU_GATHER_TABLE_FREE
423 config ARCH_HAVE_NMI_SAFE_CMPXCHG
426 config HAVE_ALIGNED_STRUCT_PAGE
429 This makes sure that struct pages are double word aligned and that
430 e.g. the SLUB allocator can perform double word atomic operations
431 on a struct page for better performance. However selecting this
432 might increase the size of a struct page by a word.
434 config HAVE_CMPXCHG_LOCAL
437 config HAVE_CMPXCHG_DOUBLE
440 config ARCH_WEAK_RELEASE_ACQUIRE
443 config ARCH_WANT_IPC_PARSE_VERSION
446 config ARCH_WANT_COMPAT_IPC_PARSE_VERSION
449 config ARCH_WANT_OLD_COMPAT_IPC
450 select ARCH_WANT_COMPAT_IPC_PARSE_VERSION
453 config HAVE_ARCH_SECCOMP_FILTER
456 An arch should select this symbol if it provides all of these things:
458 - syscall_get_arguments()
460 - syscall_set_return_value()
461 - SIGSYS siginfo_t support
462 - secure_computing is called from a ptrace_event()-safe context
463 - secure_computing return value is checked and a return value of -1
464 results in the system call being skipped immediately.
465 - seccomp syscall wired up
467 config SECCOMP_FILTER
469 depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET
471 Enable tasks to build secure computing environments defined
472 in terms of Berkeley Packet Filter programs which implement
473 task-defined system call filtering polices.
475 See Documentation/userspace-api/seccomp_filter.rst for details.
477 config HAVE_ARCH_STACKLEAK
480 An architecture should select this if it has the code which
481 fills the used part of the kernel stack with the STACKLEAK_POISON
482 value before returning from system calls.
484 config HAVE_STACKPROTECTOR
487 An arch should select this symbol if:
488 - it has implemented a stack canary (e.g. __stack_chk_guard)
490 config STACKPROTECTOR
491 bool "Stack Protector buffer overflow detection"
492 depends on HAVE_STACKPROTECTOR
493 depends on $(cc-option,-fstack-protector)
496 This option turns on the "stack-protector" GCC feature. This
497 feature puts, at the beginning of functions, a canary value on
498 the stack just before the return address, and validates
499 the value just before actually returning. Stack based buffer
500 overflows (that need to overwrite this return address) now also
501 overwrite the canary, which gets detected and the attack is then
502 neutralized via a kernel panic.
504 Functions will have the stack-protector canary logic added if they
505 have an 8-byte or larger character array on the stack.
507 This feature requires gcc version 4.2 or above, or a distribution
508 gcc with the feature backported ("-fstack-protector").
510 On an x86 "defconfig" build, this feature adds canary checks to
511 about 3% of all kernel functions, which increases kernel code size
514 config STACKPROTECTOR_STRONG
515 bool "Strong Stack Protector"
516 depends on STACKPROTECTOR
517 depends on $(cc-option,-fstack-protector-strong)
520 Functions will have the stack-protector canary logic added in any
521 of the following conditions:
523 - local variable's address used as part of the right hand side of an
524 assignment or function argument
525 - local variable is an array (or union containing an array),
526 regardless of array type or length
527 - uses register local variables
529 This feature requires gcc version 4.9 or above, or a distribution
530 gcc with the feature backported ("-fstack-protector-strong").
532 On an x86 "defconfig" build, this feature adds canary checks to
533 about 20% of all kernel functions, which increases the kernel code
536 config ARCH_SUPPORTS_SHADOW_CALL_STACK
539 An architecture should select this if it supports Clang's Shadow
540 Call Stack and implements runtime support for shadow stack
543 config SHADOW_CALL_STACK
544 bool "Clang Shadow Call Stack"
545 depends on CC_IS_CLANG && ARCH_SUPPORTS_SHADOW_CALL_STACK
546 depends on DYNAMIC_FTRACE_WITH_REGS || !FUNCTION_GRAPH_TRACER
548 This option enables Clang's Shadow Call Stack, which uses a
549 shadow stack to protect function return addresses from being
550 overwritten by an attacker. More information can be found in
551 Clang's documentation:
553 https://clang.llvm.org/docs/ShadowCallStack.html
555 Note that security guarantees in the kernel differ from the
556 ones documented for user space. The kernel must store addresses
557 of shadow stacks in memory, which means an attacker capable of
558 reading and writing arbitrary memory may be able to locate them
559 and hijack control flow by modifying the stacks.
561 config HAVE_ARCH_WITHIN_STACK_FRAMES
564 An architecture should select this if it can walk the kernel stack
565 frames to determine if an object is part of either the arguments
566 or local variables (i.e. that it excludes saved return addresses,
567 and similar) by implementing an inline arch_within_stack_frames(),
568 which is used by CONFIG_HARDENED_USERCOPY.
570 config HAVE_CONTEXT_TRACKING
573 Provide kernel/user boundaries probes necessary for subsystems
574 that need it, such as userspace RCU extended quiescent state.
575 Syscalls need to be wrapped inside user_exit()-user_enter(), either
576 optimized behind static key or through the slow path using TIF_NOHZ
577 flag. Exceptions handlers must be wrapped as well. Irqs are already
578 protected inside rcu_irq_enter/rcu_irq_exit() but preemption or signal
579 handling on irq exit still need to be protected.
584 Arch relies on TIF_NOHZ and syscall slow path to implement context
585 tracking calls to user_enter()/user_exit().
587 config HAVE_VIRT_CPU_ACCOUNTING
590 config ARCH_HAS_SCALED_CPUTIME
593 config HAVE_VIRT_CPU_ACCOUNTING_GEN
597 With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit.
598 Before enabling this option, arch code must be audited
599 to ensure there are no races in concurrent read/write of
600 cputime_t. For example, reading/writing 64-bit cputime_t on
601 some 32-bit arches may require multiple accesses, so proper
602 locking is needed to protect against concurrent accesses.
605 config HAVE_IRQ_TIME_ACCOUNTING
608 Archs need to ensure they use a high enough resolution clock to
609 support irq time accounting and then call enable_sched_clock_irqtime().
614 Archs that select this are able to move page tables at the PMD level.
616 config HAVE_ARCH_TRANSPARENT_HUGEPAGE
619 config HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
622 config HAVE_ARCH_HUGE_VMAP
625 config ARCH_WANT_HUGE_PMD_SHARE
628 config HAVE_ARCH_SOFT_DIRTY
631 config HAVE_MOD_ARCH_SPECIFIC
634 The arch uses struct mod_arch_specific to store data. Many arches
635 just need a simple module loader without arch specific data - those
636 should not enable this.
638 config MODULES_USE_ELF_RELA
641 Modules only use ELF RELA relocations. Modules with ELF REL
642 relocations will give an error.
644 config MODULES_USE_ELF_REL
647 Modules only use ELF REL relocations. Modules with ELF RELA
648 relocations will give an error.
650 config HAVE_IRQ_EXIT_ON_IRQ_STACK
653 Architecture doesn't only execute the irq handler on the irq stack
654 but also irq_exit(). This way we can process softirqs on this irq
655 stack instead of switching to a new one when we call __do_softirq()
656 in the end of an hardirq.
657 This spares a stack switch and improves cache usage on softirq
660 config PGTABLE_LEVELS
664 config ARCH_HAS_ELF_RANDOMIZE
667 An architecture supports choosing randomized locations for
668 stack, mmap, brk, and ET_DYN. Defined functions:
670 - arch_randomize_brk()
672 config HAVE_ARCH_MMAP_RND_BITS
675 An arch should select this symbol if it supports setting a variable
676 number of bits for use in establishing the base address for mmap
677 allocations, has MMU enabled and provides values for both:
678 - ARCH_MMAP_RND_BITS_MIN
679 - ARCH_MMAP_RND_BITS_MAX
681 config HAVE_EXIT_THREAD
684 An architecture implements exit_thread.
686 config ARCH_MMAP_RND_BITS_MIN
689 config ARCH_MMAP_RND_BITS_MAX
692 config ARCH_MMAP_RND_BITS_DEFAULT
695 config ARCH_MMAP_RND_BITS
696 int "Number of bits to use for ASLR of mmap base address" if EXPERT
697 range ARCH_MMAP_RND_BITS_MIN ARCH_MMAP_RND_BITS_MAX
698 default ARCH_MMAP_RND_BITS_DEFAULT if ARCH_MMAP_RND_BITS_DEFAULT
699 default ARCH_MMAP_RND_BITS_MIN
700 depends on HAVE_ARCH_MMAP_RND_BITS
702 This value can be used to select the number of bits to use to
703 determine the random offset to the base address of vma regions
704 resulting from mmap allocations. This value will be bounded
705 by the architecture's minimum and maximum supported values.
707 This value can be changed after boot using the
708 /proc/sys/vm/mmap_rnd_bits tunable
710 config HAVE_ARCH_MMAP_RND_COMPAT_BITS
713 An arch should select this symbol if it supports running applications
714 in compatibility mode, supports setting a variable number of bits for
715 use in establishing the base address for mmap allocations, has MMU
716 enabled and provides values for both:
717 - ARCH_MMAP_RND_COMPAT_BITS_MIN
718 - ARCH_MMAP_RND_COMPAT_BITS_MAX
720 config ARCH_MMAP_RND_COMPAT_BITS_MIN
723 config ARCH_MMAP_RND_COMPAT_BITS_MAX
726 config ARCH_MMAP_RND_COMPAT_BITS_DEFAULT
729 config ARCH_MMAP_RND_COMPAT_BITS
730 int "Number of bits to use for ASLR of mmap base address for compatible applications" if EXPERT
731 range ARCH_MMAP_RND_COMPAT_BITS_MIN ARCH_MMAP_RND_COMPAT_BITS_MAX
732 default ARCH_MMAP_RND_COMPAT_BITS_DEFAULT if ARCH_MMAP_RND_COMPAT_BITS_DEFAULT
733 default ARCH_MMAP_RND_COMPAT_BITS_MIN
734 depends on HAVE_ARCH_MMAP_RND_COMPAT_BITS
736 This value can be used to select the number of bits to use to
737 determine the random offset to the base address of vma regions
738 resulting from mmap allocations for compatible applications This
739 value will be bounded by the architecture's minimum and maximum
742 This value can be changed after boot using the
743 /proc/sys/vm/mmap_rnd_compat_bits tunable
745 config HAVE_ARCH_COMPAT_MMAP_BASES
748 This allows 64bit applications to invoke 32-bit mmap() syscall
749 and vice-versa 32-bit applications to call 64-bit mmap().
750 Required for applications doing different bitness syscalls.
752 # This allows to use a set of generic functions to determine mmap base
753 # address by giving priority to top-down scheme only if the process
754 # is not in legacy mode (compat task, unlimited stack size or
755 # sysctl_legacy_va_layout).
756 # Architecture that selects this option can provide its own version of:
758 config ARCH_WANT_DEFAULT_TOPDOWN_MMAP_LAYOUT
761 select ARCH_HAS_ELF_RANDOMIZE
763 config HAVE_STACK_VALIDATION
766 Architecture supports the 'objtool check' host tool command, which
767 performs compile-time stack metadata validation.
769 config HAVE_RELIABLE_STACKTRACE
772 Architecture has either save_stack_trace_tsk_reliable() or
773 arch_stack_walk_reliable() function which only returns a stack trace
774 if it can guarantee the trace is reliable.
776 config HAVE_ARCH_HASH
780 If this is set, the architecture provides an <asm/hash.h>
781 file which provides platform-specific implementations of some
782 functions in <linux/hash.h> or fs/namei.c.
784 config HAVE_ARCH_NVRAM_OPS
793 config CLONE_BACKWARDS
796 Architecture has tls passed as the 4th argument of clone(2),
799 config CLONE_BACKWARDS2
802 Architecture has the first two arguments of clone(2) swapped.
804 config CLONE_BACKWARDS3
807 Architecture has tls passed as the 3rd argument of clone(2),
810 config ODD_RT_SIGACTION
813 Architecture has unusual rt_sigaction(2) arguments
815 config OLD_SIGSUSPEND
818 Architecture has old sigsuspend(2) syscall, of one-argument variety
820 config OLD_SIGSUSPEND3
823 Even weirder antique ABI - three-argument sigsuspend(2)
828 Architecture has old sigaction(2) syscall. Nope, not the same
829 as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2),
830 but fairly different variant of sigaction(2), thanks to OSF/1
833 config COMPAT_OLD_SIGACTION
836 config COMPAT_32BIT_TIME
837 bool "Provide system calls for 32-bit time_t"
838 default !64BIT || COMPAT
840 This enables 32 bit time_t support in addition to 64 bit time_t support.
841 This is relevant on all 32-bit architectures, and 64-bit architectures
842 as part of compat syscall handling.
844 config ARCH_NO_PREEMPT
847 config ARCH_SUPPORTS_RT
850 config CPU_NO_EFFICIENT_FFS
853 config HAVE_ARCH_VMAP_STACK
856 An arch should select this symbol if it can support kernel stacks
857 in vmalloc space. This means:
859 - vmalloc space must be large enough to hold many kernel stacks.
860 This may rule out many 32-bit architectures.
862 - Stacks in vmalloc space need to work reliably. For example, if
863 vmap page tables are created on demand, either this mechanism
864 needs to work while the stack points to a virtual address with
865 unpopulated page tables or arch code (switch_to() and switch_mm(),
866 most likely) needs to ensure that the stack's page table entries
867 are populated before running on a possibly unpopulated stack.
869 - If the stack overflows into a guard page, something reasonable
870 should happen. The definition of "reasonable" is flexible, but
871 instantly rebooting without logging anything would be unfriendly.
875 bool "Use a virtually-mapped stack"
876 depends on HAVE_ARCH_VMAP_STACK
877 depends on !KASAN || KASAN_VMALLOC
879 Enable this if you want the use virtually-mapped kernel stacks
880 with guard pages. This causes kernel stack overflows to be
881 caught immediately rather than causing difficult-to-diagnose
884 To use this with KASAN, the architecture must support backing
885 virtual mappings with real shadow memory, and KASAN_VMALLOC must
888 config ARCH_OPTIONAL_KERNEL_RWX
891 config ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
894 config ARCH_HAS_STRICT_KERNEL_RWX
897 config STRICT_KERNEL_RWX
898 bool "Make kernel text and rodata read-only" if ARCH_OPTIONAL_KERNEL_RWX
899 depends on ARCH_HAS_STRICT_KERNEL_RWX
900 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
902 If this is set, kernel text and rodata memory will be made read-only,
903 and non-text memory will be made non-executable. This provides
904 protection against certain security exploits (e.g. executing the heap
907 These features are considered standard security practice these days.
908 You should say Y here in almost all cases.
910 config ARCH_HAS_STRICT_MODULE_RWX
913 config STRICT_MODULE_RWX
914 bool "Set loadable kernel module data as NX and text as RO" if ARCH_OPTIONAL_KERNEL_RWX
915 depends on ARCH_HAS_STRICT_MODULE_RWX && MODULES
916 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
918 If this is set, module text and rodata memory will be made read-only,
919 and non-text memory will be made non-executable. This provides
920 protection against certain security exploits (e.g. writing to text)
922 # select if the architecture provides an asm/dma-direct.h header
923 config ARCH_HAS_PHYS_TO_DMA
926 config HAVE_ARCH_COMPILER_H
929 An architecture can select this if it provides an
930 asm/compiler.h header that should be included after
931 linux/compiler-*.h in order to override macro definitions that those
932 headers generally provide.
934 config HAVE_ARCH_PREL32_RELOCATIONS
937 May be selected by an architecture if it supports place-relative
938 32-bit relocations, both in the toolchain and in the module loader,
939 in which case relative references can be used in special sections
940 for PCI fixup, initcalls etc which are only half the size on 64 bit
941 architectures, and don't require runtime relocation on relocatable
944 config ARCH_USE_MEMREMAP_PROT
947 config LOCK_EVENT_COUNTS
948 bool "Locking event counts collection"
951 Enable light-weight counting of various locking related events
952 in the system with minimal performance impact. This reduces
953 the chance of application behavior change because of timing
954 differences. The counts are reported via debugfs.
956 # Select if the architecture has support for applying RELR relocations.
961 bool "Use RELR relocation packing"
962 depends on ARCH_HAS_RELR && TOOLS_SUPPORT_RELR
965 Store the kernel's dynamic relocations in the RELR relocation packing
966 format. Requires a compatible linker (LLD supports this feature), as
967 well as compatible NM and OBJCOPY utilities (llvm-nm and llvm-objcopy
970 config ARCH_HAS_MEM_ENCRYPT
973 config HAVE_SPARSE_SYSCALL_NR
976 An architecture should select this if its syscall numbering is sparse
977 to save space. For example, MIPS architecture has a syscall array with
978 entries at 4000, 5000 and 6000 locations. This option turns on syscall
979 related optimizations for a given architecture.
981 config ARCH_HAS_VDSO_DATA
984 config HAVE_STATIC_CALL
987 config HAVE_STATIC_CALL_INLINE
989 depends on HAVE_STATIC_CALL
991 source "kernel/gcov/Kconfig"
993 source "scripts/gcc-plugins/Kconfig"