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"
36 depends on HAVE_KPROBES
39 Kprobes allows you to trap at almost any kernel address and
40 execute a callback function. register_kprobe() establishes
41 a probepoint and specifies the callback. Kprobes is useful
42 for kernel debugging, non-intrusive instrumentation and testing.
46 bool "Optimize very unlikely/likely branches"
47 depends on HAVE_ARCH_JUMP_LABEL
48 depends on CC_HAS_ASM_GOTO
50 This option enables a transparent branch optimization that
51 makes certain almost-always-true or almost-always-false branch
52 conditions even cheaper to execute within the kernel.
54 Certain performance-sensitive kernel code, such as trace points,
55 scheduler functionality, networking code and KVM have such
56 branches and include support for this optimization technique.
58 If it is detected that the compiler has support for "asm goto",
59 the kernel will compile such branches with just a nop
60 instruction. When the condition flag is toggled to true, the
61 nop will be converted to a jump instruction to execute the
62 conditional block of instructions.
64 This technique lowers overhead and stress on the branch prediction
65 of the processor and generally makes the kernel faster. The update
66 of the condition is slower, but those are always very rare.
68 ( On 32-bit x86, the necessary options added to the compiler
69 flags may increase the size of the kernel slightly. )
71 config STATIC_KEYS_SELFTEST
72 bool "Static key selftest"
75 Boot time self-test of the branch patching code.
77 config STATIC_CALL_SELFTEST
78 bool "Static call selftest"
79 depends on HAVE_STATIC_CALL
81 Boot time self-test of the call patching code.
85 depends on KPROBES && HAVE_OPTPROBES
86 select TASKS_RCU if PREEMPTION
88 config KPROBES_ON_FTRACE
90 depends on KPROBES && HAVE_KPROBES_ON_FTRACE
91 depends on DYNAMIC_FTRACE_WITH_REGS
93 If function tracer is enabled and the arch supports full
94 passing of pt_regs to function tracing, then kprobes can
95 optimize on top of function tracing.
99 depends on ARCH_SUPPORTS_UPROBES
101 Uprobes is the user-space counterpart to kprobes: they
102 enable instrumentation applications (such as 'perf probe')
103 to establish unintrusive probes in user-space binaries and
104 libraries, by executing handler functions when the probes
105 are hit by user-space applications.
107 ( These probes come in the form of single-byte breakpoints,
108 managed by the kernel and kept transparent to the probed
111 config HAVE_64BIT_ALIGNED_ACCESS
112 def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS
114 Some architectures require 64 bit accesses to be 64 bit
115 aligned, which also requires structs containing 64 bit values
116 to be 64 bit aligned too. This includes some 32 bit
117 architectures which can do 64 bit accesses, as well as 64 bit
118 architectures without unaligned access.
120 This symbol should be selected by an architecture if 64 bit
121 accesses are required to be 64 bit aligned in this way even
122 though it is not a 64 bit architecture.
124 See Documentation/core-api/unaligned-memory-access.rst for
125 more information on the topic of unaligned memory accesses.
127 config HAVE_EFFICIENT_UNALIGNED_ACCESS
130 Some architectures are unable to perform unaligned accesses
131 without the use of get_unaligned/put_unaligned. Others are
132 unable to perform such accesses efficiently (e.g. trap on
133 unaligned access and require fixing it up in the exception
136 This symbol should be selected by an architecture if it can
137 perform unaligned accesses efficiently to allow different
138 code paths to be selected for these cases. Some network
139 drivers, for example, could opt to not fix up alignment
140 problems with received packets if doing so would not help
143 See Documentation/core-api/unaligned-memory-access.rst for more
144 information on the topic of unaligned memory accesses.
146 config ARCH_USE_BUILTIN_BSWAP
149 Modern versions of GCC (since 4.4) have builtin functions
150 for handling byte-swapping. Using these, instead of the old
151 inline assembler that the architecture code provides in the
152 __arch_bswapXX() macros, allows the compiler to see what's
153 happening and offers more opportunity for optimisation. In
154 particular, the compiler will be able to combine the byteswap
155 with a nearby load or store and use load-and-swap or
156 store-and-swap instructions if the architecture has them. It
157 should almost *never* result in code which is worse than the
158 hand-coded assembler in <asm/swab.h>. But just in case it
159 does, the use of the builtins is optional.
161 Any architecture with load-and-swap or store-and-swap
162 instructions should set this. And it shouldn't hurt to set it
163 on architectures that don't have such instructions.
167 depends on KPROBES && (HAVE_KRETPROBES || HAVE_RETHOOK)
169 config KRETPROBE_ON_RETHOOK
171 depends on HAVE_RETHOOK
172 depends on KRETPROBES
175 config USER_RETURN_NOTIFIER
177 depends on HAVE_USER_RETURN_NOTIFIER
179 Provide a kernel-internal notification when a cpu is about to
182 config HAVE_IOREMAP_PROT
188 config HAVE_KRETPROBES
191 config HAVE_OPTPROBES
194 config HAVE_KPROBES_ON_FTRACE
197 config ARCH_CORRECT_STACKTRACE_ON_KRETPROBE
200 Since kretprobes modifies return address on the stack, the
201 stacktrace may see the kretprobe trampoline address instead
202 of correct one. If the architecture stacktrace code and
203 unwinder can adjust such entries, select this configuration.
205 config HAVE_FUNCTION_ERROR_INJECTION
211 config HAVE_FUNCTION_DESCRIPTORS
214 config TRACE_IRQFLAGS_SUPPORT
218 # An arch should select this if it provides all these things:
220 # task_pt_regs() in asm/processor.h or asm/ptrace.h
221 # arch_has_single_step() if there is hardware single-step support
222 # arch_has_block_step() if there is hardware block-step support
223 # asm/syscall.h supplying asm-generic/syscall.h interface
224 # linux/regset.h user_regset interfaces
225 # CORE_DUMP_USE_REGSET #define'd in linux/elf.h
226 # TIF_SYSCALL_TRACE calls ptrace_report_syscall_{entry,exit}
227 # TIF_NOTIFY_RESUME calls resume_user_mode_work()
229 config HAVE_ARCH_TRACEHOOK
232 config HAVE_DMA_CONTIGUOUS
235 config GENERIC_SMP_IDLE_THREAD
238 config GENERIC_IDLE_POLL_SETUP
241 config ARCH_HAS_FORTIFY_SOURCE
244 An architecture should select this when it can successfully
245 build and run with CONFIG_FORTIFY_SOURCE.
248 # Select if the arch provides a historic keepinit alias for the retain_initrd
249 # command line option
251 config ARCH_HAS_KEEPINITRD
254 # Select if arch has all set_memory_ro/rw/x/nx() functions in asm/cacheflush.h
255 config ARCH_HAS_SET_MEMORY
258 # Select if arch has all set_direct_map_invalid/default() functions
259 config ARCH_HAS_SET_DIRECT_MAP
263 # Select if the architecture provides the arch_dma_set_uncached symbol to
264 # either provide an uncached segment alias for a DMA allocation, or
265 # to remap the page tables in place.
267 config ARCH_HAS_DMA_SET_UNCACHED
271 # Select if the architectures provides the arch_dma_clear_uncached symbol
272 # to undo an in-place page table remap for uncached access.
274 config ARCH_HAS_DMA_CLEAR_UNCACHED
277 # Select if arch init_task must go in the __init_task_data section
278 config ARCH_TASK_STRUCT_ON_STACK
281 # Select if arch has its private alloc_task_struct() function
282 config ARCH_TASK_STRUCT_ALLOCATOR
285 config HAVE_ARCH_THREAD_STRUCT_WHITELIST
287 depends on !ARCH_TASK_STRUCT_ALLOCATOR
289 An architecture should select this to provide hardened usercopy
290 knowledge about what region of the thread_struct should be
291 whitelisted for copying to userspace. Normally this is only the
292 FPU registers. Specifically, arch_thread_struct_whitelist()
293 should be implemented. Without this, the entire thread_struct
294 field in task_struct will be left whitelisted.
296 # Select if arch has its private alloc_thread_stack() function
297 config ARCH_THREAD_STACK_ALLOCATOR
300 # Select if arch wants to size task_struct dynamically via arch_task_struct_size:
301 config ARCH_WANTS_DYNAMIC_TASK_STRUCT
304 config ARCH_WANTS_NO_INSTR
307 An architecture should select this if the noinstr macro is being used on
308 functions to denote that the toolchain should avoid instrumenting such
309 functions and is required for correctness.
311 config ARCH_32BIT_OFF_T
315 All new 32-bit architectures should have 64-bit off_t type on
316 userspace side which corresponds to the loff_t kernel type. This
317 is the requirement for modern ABIs. Some existing architectures
318 still support 32-bit off_t. This option is enabled for all such
319 architectures explicitly.
321 # Selected by 64 bit architectures which have a 32 bit f_tinode in struct ustat
322 config ARCH_32BIT_USTAT_F_TINODE
325 config HAVE_ASM_MODVERSIONS
328 This symbol should be selected by an architecture if it provides
329 <asm/asm-prototypes.h> to support the module versioning for symbols
330 exported from assembly code.
332 config HAVE_REGS_AND_STACK_ACCESS_API
335 This symbol should be selected by an architecture if it supports
336 the API needed to access registers and stack entries from pt_regs,
337 declared in asm/ptrace.h
338 For example the kprobes-based event tracer needs this API.
342 depends on HAVE_REGS_AND_STACK_ACCESS_API
344 This symbol should be selected by an architecture if it
345 supports an implementation of restartable sequences.
347 config HAVE_FUNCTION_ARG_ACCESS_API
350 This symbol should be selected by an architecture if it supports
351 the API needed to access function arguments from pt_regs,
352 declared in asm/ptrace.h
354 config HAVE_HW_BREAKPOINT
356 depends on PERF_EVENTS
358 config HAVE_MIXED_BREAKPOINTS_REGS
360 depends on HAVE_HW_BREAKPOINT
362 Depending on the arch implementation of hardware breakpoints,
363 some of them have separate registers for data and instruction
364 breakpoints addresses, others have mixed registers to store
365 them but define the access type in a control register.
366 Select this option if your arch implements breakpoints under the
369 config HAVE_USER_RETURN_NOTIFIER
372 config HAVE_PERF_EVENTS_NMI
375 System hardware can generate an NMI using the perf event
376 subsystem. Also has support for calculating CPU cycle events
377 to determine how many clock cycles in a given period.
379 config HAVE_HARDLOCKUP_DETECTOR_PERF
381 depends on HAVE_PERF_EVENTS_NMI
383 The arch chooses to use the generic perf-NMI-based hardlockup
384 detector. Must define HAVE_PERF_EVENTS_NMI.
386 config HAVE_NMI_WATCHDOG
390 The arch provides a low level NMI watchdog. It provides
391 asm/nmi.h, and defines its own arch_touch_nmi_watchdog().
393 config HAVE_HARDLOCKUP_DETECTOR_ARCH
395 select HAVE_NMI_WATCHDOG
397 The arch chooses to provide its own hardlockup detector, which is
398 a superset of the HAVE_NMI_WATCHDOG. It also conforms to config
399 interfaces and parameters provided by hardlockup detector subsystem.
401 config HAVE_PERF_REGS
404 Support selective register dumps for perf events. This includes
405 bit-mapping of each registers and a unique architecture id.
407 config HAVE_PERF_USER_STACK_DUMP
410 Support user stack dumps for perf event samples. This needs
411 access to the user stack pointer which is not unified across
414 config HAVE_ARCH_JUMP_LABEL
417 config HAVE_ARCH_JUMP_LABEL_RELATIVE
420 config MMU_GATHER_TABLE_FREE
423 config MMU_GATHER_RCU_TABLE_FREE
425 select MMU_GATHER_TABLE_FREE
427 config MMU_GATHER_PAGE_SIZE
430 config MMU_GATHER_NO_RANGE
433 config MMU_GATHER_NO_GATHER
435 depends on MMU_GATHER_TABLE_FREE
437 config ARCH_WANT_IRQS_OFF_ACTIVATE_MM
440 Temporary select until all architectures can be converted to have
441 irqs disabled over activate_mm. Architectures that do IPI based TLB
442 shootdowns should enable this.
444 config ARCH_HAVE_NMI_SAFE_CMPXCHG
447 config HAVE_ALIGNED_STRUCT_PAGE
450 This makes sure that struct pages are double word aligned and that
451 e.g. the SLUB allocator can perform double word atomic operations
452 on a struct page for better performance. However selecting this
453 might increase the size of a struct page by a word.
455 config HAVE_CMPXCHG_LOCAL
458 config HAVE_CMPXCHG_DOUBLE
461 config ARCH_WEAK_RELEASE_ACQUIRE
464 config ARCH_WANT_IPC_PARSE_VERSION
467 config ARCH_WANT_COMPAT_IPC_PARSE_VERSION
470 config ARCH_WANT_OLD_COMPAT_IPC
471 select ARCH_WANT_COMPAT_IPC_PARSE_VERSION
474 config HAVE_ARCH_SECCOMP
477 An arch should select this symbol to support seccomp mode 1 (the fixed
478 syscall policy), and must provide an overrides for __NR_seccomp_sigreturn,
479 and compat syscalls if the asm-generic/seccomp.h defaults need adjustment:
480 - __NR_seccomp_read_32
481 - __NR_seccomp_write_32
482 - __NR_seccomp_exit_32
483 - __NR_seccomp_sigreturn_32
485 config HAVE_ARCH_SECCOMP_FILTER
487 select HAVE_ARCH_SECCOMP
489 An arch should select this symbol if it provides all of these things:
490 - all the requirements for HAVE_ARCH_SECCOMP
492 - syscall_get_arguments()
494 - syscall_set_return_value()
495 - SIGSYS siginfo_t support
496 - secure_computing is called from a ptrace_event()-safe context
497 - secure_computing return value is checked and a return value of -1
498 results in the system call being skipped immediately.
499 - seccomp syscall wired up
500 - if !HAVE_SPARSE_SYSCALL_NR, have SECCOMP_ARCH_NATIVE,
501 SECCOMP_ARCH_NATIVE_NR, SECCOMP_ARCH_NATIVE_NAME defined. If
502 COMPAT is supported, have the SECCOMP_ARCH_COMPAT* defines too.
505 prompt "Enable seccomp to safely execute untrusted bytecode"
507 depends on HAVE_ARCH_SECCOMP
509 This kernel feature is useful for number crunching applications
510 that may need to handle untrusted bytecode during their
511 execution. By using pipes or other transports made available
512 to the process as file descriptors supporting the read/write
513 syscalls, it's possible to isolate those applications in their
514 own address space using seccomp. Once seccomp is enabled via
515 prctl(PR_SET_SECCOMP) or the seccomp() syscall, it cannot be
516 disabled and the task is only allowed to execute a few safe
517 syscalls defined by each seccomp mode.
521 config SECCOMP_FILTER
523 depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET
525 Enable tasks to build secure computing environments defined
526 in terms of Berkeley Packet Filter programs which implement
527 task-defined system call filtering polices.
529 See Documentation/userspace-api/seccomp_filter.rst for details.
531 config SECCOMP_CACHE_DEBUG
532 bool "Show seccomp filter cache status in /proc/pid/seccomp_cache"
533 depends on SECCOMP_FILTER && !HAVE_SPARSE_SYSCALL_NR
536 This enables the /proc/pid/seccomp_cache interface to monitor
537 seccomp cache data. The file format is subject to change. Reading
538 the file requires CAP_SYS_ADMIN.
540 This option is for debugging only. Enabling presents the risk that
541 an adversary may be able to infer the seccomp filter logic.
545 config HAVE_ARCH_STACKLEAK
548 An architecture should select this if it has the code which
549 fills the used part of the kernel stack with the STACKLEAK_POISON
550 value before returning from system calls.
552 config HAVE_STACKPROTECTOR
555 An arch should select this symbol if:
556 - it has implemented a stack canary (e.g. __stack_chk_guard)
558 config STACKPROTECTOR
559 bool "Stack Protector buffer overflow detection"
560 depends on HAVE_STACKPROTECTOR
561 depends on $(cc-option,-fstack-protector)
564 This option turns on the "stack-protector" GCC feature. This
565 feature puts, at the beginning of functions, a canary value on
566 the stack just before the return address, and validates
567 the value just before actually returning. Stack based buffer
568 overflows (that need to overwrite this return address) now also
569 overwrite the canary, which gets detected and the attack is then
570 neutralized via a kernel panic.
572 Functions will have the stack-protector canary logic added if they
573 have an 8-byte or larger character array on the stack.
575 This feature requires gcc version 4.2 or above, or a distribution
576 gcc with the feature backported ("-fstack-protector").
578 On an x86 "defconfig" build, this feature adds canary checks to
579 about 3% of all kernel functions, which increases kernel code size
582 config STACKPROTECTOR_STRONG
583 bool "Strong Stack Protector"
584 depends on STACKPROTECTOR
585 depends on $(cc-option,-fstack-protector-strong)
588 Functions will have the stack-protector canary logic added in any
589 of the following conditions:
591 - local variable's address used as part of the right hand side of an
592 assignment or function argument
593 - local variable is an array (or union containing an array),
594 regardless of array type or length
595 - uses register local variables
597 This feature requires gcc version 4.9 or above, or a distribution
598 gcc with the feature backported ("-fstack-protector-strong").
600 On an x86 "defconfig" build, this feature adds canary checks to
601 about 20% of all kernel functions, which increases the kernel code
604 config ARCH_SUPPORTS_SHADOW_CALL_STACK
607 An architecture should select this if it supports the compiler's
608 Shadow Call Stack and implements runtime support for shadow stack
611 config SHADOW_CALL_STACK
612 bool "Shadow Call Stack"
613 depends on ARCH_SUPPORTS_SHADOW_CALL_STACK
614 depends on DYNAMIC_FTRACE_WITH_REGS || !FUNCTION_GRAPH_TRACER
616 This option enables the compiler's Shadow Call Stack, which
617 uses a shadow stack to protect function return addresses from
618 being overwritten by an attacker. More information can be found
619 in the compiler's documentation:
621 - Clang: https://clang.llvm.org/docs/ShadowCallStack.html
622 - GCC: https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html#Instrumentation-Options
624 Note that security guarantees in the kernel differ from the
625 ones documented for user space. The kernel must store addresses
626 of shadow stacks in memory, which means an attacker capable of
627 reading and writing arbitrary memory may be able to locate them
628 and hijack control flow by modifying the stacks.
633 Selected if the kernel will be built using the compiler's LTO feature.
639 Selected if the kernel will be built using Clang's LTO feature.
641 config ARCH_SUPPORTS_LTO_CLANG
644 An architecture should select this option if it supports:
645 - compiling with Clang,
646 - compiling inline assembly with Clang's integrated assembler,
647 - and linking with LLD.
649 config ARCH_SUPPORTS_LTO_CLANG_THIN
652 An architecture should select this option if it can support Clang's
657 depends on CC_IS_CLANG && LD_IS_LLD && AS_IS_LLVM
658 depends on $(success,$(NM) --help | head -n 1 | grep -qi llvm)
659 depends on $(success,$(AR) --help | head -n 1 | grep -qi llvm)
660 depends on ARCH_SUPPORTS_LTO_CLANG
661 depends on !FTRACE_MCOUNT_USE_RECORDMCOUNT
662 depends on !KASAN || KASAN_HW_TAGS
663 depends on !GCOV_KERNEL
665 The compiler and Kconfig options support building with Clang's
669 prompt "Link Time Optimization (LTO)"
672 This option enables Link Time Optimization (LTO), which allows the
673 compiler to optimize binaries globally.
675 If unsure, select LTO_NONE. Note that LTO is very resource-intensive
676 so it's disabled by default.
681 Build the kernel normally, without Link Time Optimization (LTO).
683 config LTO_CLANG_FULL
684 bool "Clang Full LTO (EXPERIMENTAL)"
685 depends on HAS_LTO_CLANG
686 depends on !COMPILE_TEST
689 This option enables Clang's full Link Time Optimization (LTO), which
690 allows the compiler to optimize the kernel globally. If you enable
691 this option, the compiler generates LLVM bitcode instead of ELF
692 object files, and the actual compilation from bitcode happens at
693 the LTO link step, which may take several minutes depending on the
694 kernel configuration. More information can be found from LLVM's
697 https://llvm.org/docs/LinkTimeOptimization.html
699 During link time, this option can use a large amount of RAM, and
700 may take much longer than the ThinLTO option.
702 config LTO_CLANG_THIN
703 bool "Clang ThinLTO (EXPERIMENTAL)"
704 depends on HAS_LTO_CLANG && ARCH_SUPPORTS_LTO_CLANG_THIN
707 This option enables Clang's ThinLTO, which allows for parallel
708 optimization and faster incremental compiles compared to the
709 CONFIG_LTO_CLANG_FULL option. More information can be found
710 from Clang's documentation:
712 https://clang.llvm.org/docs/ThinLTO.html
717 config ARCH_SUPPORTS_CFI_CLANG
720 An architecture should select this option if it can support Clang's
721 Control-Flow Integrity (CFI) checking.
724 bool "Use Clang's Control Flow Integrity (CFI)"
725 depends on LTO_CLANG && ARCH_SUPPORTS_CFI_CLANG
727 # - https://bugs.llvm.org/show_bug.cgi?id=46258
728 # - https://bugs.llvm.org/show_bug.cgi?id=47479
729 depends on CLANG_VERSION >= 120000
732 This option enables Clang’s forward-edge Control Flow Integrity
733 (CFI) checking, where the compiler injects a runtime check to each
734 indirect function call to ensure the target is a valid function with
735 the correct static type. This restricts possible call targets and
736 makes it more difficult for an attacker to exploit bugs that allow
737 the modification of stored function pointers. More information can be
738 found from Clang's documentation:
740 https://clang.llvm.org/docs/ControlFlowIntegrity.html
742 config CFI_CLANG_SHADOW
743 bool "Use CFI shadow to speed up cross-module checks"
745 depends on CFI_CLANG && MODULES
747 If you select this option, the kernel builds a fast look-up table of
748 CFI check functions in loaded modules to reduce performance overhead.
752 config CFI_PERMISSIVE
753 bool "Use CFI in permissive mode"
756 When selected, Control Flow Integrity (CFI) violations result in a
757 warning instead of a kernel panic. This option should only be used
758 for finding indirect call type mismatches during development.
762 config HAVE_ARCH_WITHIN_STACK_FRAMES
765 An architecture should select this if it can walk the kernel stack
766 frames to determine if an object is part of either the arguments
767 or local variables (i.e. that it excludes saved return addresses,
768 and similar) by implementing an inline arch_within_stack_frames(),
769 which is used by CONFIG_HARDENED_USERCOPY.
771 config HAVE_CONTEXT_TRACKING
774 Provide kernel/user boundaries probes necessary for subsystems
775 that need it, such as userspace RCU extended quiescent state.
776 Syscalls need to be wrapped inside user_exit()-user_enter(), either
777 optimized behind static key or through the slow path using TIF_NOHZ
778 flag. Exceptions handlers must be wrapped as well. Irqs are already
779 protected inside rcu_irq_enter/rcu_irq_exit() but preemption or signal
780 handling on irq exit still need to be protected.
782 config HAVE_CONTEXT_TRACKING_OFFSTACK
785 Architecture neither relies on exception_enter()/exception_exit()
786 nor on schedule_user(). Also preempt_schedule_notrace() and
787 preempt_schedule_irq() can't be called in a preemptible section
788 while context tracking is CONTEXT_USER. This feature reflects a sane
789 entry implementation where the following requirements are met on
790 critical entry code, ie: before user_exit() or after user_enter():
792 - Critical entry code isn't preemptible (or better yet:
794 - No use of RCU read side critical sections, unless rcu_nmi_enter()
796 - No use of instrumentation, unless instrumentation_begin() got
802 Arch relies on TIF_NOHZ and syscall slow path to implement context
803 tracking calls to user_enter()/user_exit().
805 config HAVE_VIRT_CPU_ACCOUNTING
808 config HAVE_VIRT_CPU_ACCOUNTING_IDLE
811 Architecture has its own way to account idle CPU time and therefore
812 doesn't implement vtime_account_idle().
814 config ARCH_HAS_SCALED_CPUTIME
817 config HAVE_VIRT_CPU_ACCOUNTING_GEN
821 With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit.
822 Before enabling this option, arch code must be audited
823 to ensure there are no races in concurrent read/write of
824 cputime_t. For example, reading/writing 64-bit cputime_t on
825 some 32-bit arches may require multiple accesses, so proper
826 locking is needed to protect against concurrent accesses.
828 config HAVE_IRQ_TIME_ACCOUNTING
831 Archs need to ensure they use a high enough resolution clock to
832 support irq time accounting and then call enable_sched_clock_irqtime().
837 Architectures that select this are able to move page tables at the
838 PUD level. If there are only 3 page table levels, the move effectively
839 happens at the PGD level.
844 Archs that select this are able to move page tables at the PMD level.
846 config HAVE_ARCH_TRANSPARENT_HUGEPAGE
849 config HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
852 config HAVE_ARCH_HUGE_VMAP
856 # Archs that select this would be capable of PMD-sized vmaps (i.e.,
857 # arch_vmap_pmd_supported() returns true), and they must make no assumptions
858 # that vmalloc memory is mapped with PAGE_SIZE ptes. The VM_NO_HUGE_VMAP flag
859 # can be used to prohibit arch-specific allocations from using hugepages to
860 # help with this (e.g., modules may require it).
862 config HAVE_ARCH_HUGE_VMALLOC
863 depends on HAVE_ARCH_HUGE_VMAP
866 config ARCH_WANT_HUGE_PMD_SHARE
869 config HAVE_ARCH_SOFT_DIRTY
872 config HAVE_MOD_ARCH_SPECIFIC
875 The arch uses struct mod_arch_specific to store data. Many arches
876 just need a simple module loader without arch specific data - those
877 should not enable this.
879 config MODULES_USE_ELF_RELA
882 Modules only use ELF RELA relocations. Modules with ELF REL
883 relocations will give an error.
885 config MODULES_USE_ELF_REL
888 Modules only use ELF REL relocations. Modules with ELF RELA
889 relocations will give an error.
891 config HAVE_IRQ_EXIT_ON_IRQ_STACK
894 Architecture doesn't only execute the irq handler on the irq stack
895 but also irq_exit(). This way we can process softirqs on this irq
896 stack instead of switching to a new one when we call __do_softirq()
897 in the end of an hardirq.
898 This spares a stack switch and improves cache usage on softirq
901 config HAVE_SOFTIRQ_ON_OWN_STACK
904 Architecture provides a function to run __do_softirq() on a
907 config ALTERNATE_USER_ADDRESS_SPACE
910 Architectures set this when the CPU uses separate address
911 spaces for kernel and user space pointers. In this case, the
912 access_ok() check on a __user pointer is skipped.
914 config PGTABLE_LEVELS
918 config ARCH_HAS_ELF_RANDOMIZE
921 An architecture supports choosing randomized locations for
922 stack, mmap, brk, and ET_DYN. Defined functions:
924 - arch_randomize_brk()
926 config HAVE_ARCH_MMAP_RND_BITS
929 An arch should select this symbol if it supports setting a variable
930 number of bits for use in establishing the base address for mmap
931 allocations, has MMU enabled and provides values for both:
932 - ARCH_MMAP_RND_BITS_MIN
933 - ARCH_MMAP_RND_BITS_MAX
935 config HAVE_EXIT_THREAD
938 An architecture implements exit_thread.
940 config ARCH_MMAP_RND_BITS_MIN
943 config ARCH_MMAP_RND_BITS_MAX
946 config ARCH_MMAP_RND_BITS_DEFAULT
949 config ARCH_MMAP_RND_BITS
950 int "Number of bits to use for ASLR of mmap base address" if EXPERT
951 range ARCH_MMAP_RND_BITS_MIN ARCH_MMAP_RND_BITS_MAX
952 default ARCH_MMAP_RND_BITS_DEFAULT if ARCH_MMAP_RND_BITS_DEFAULT
953 default ARCH_MMAP_RND_BITS_MIN
954 depends on HAVE_ARCH_MMAP_RND_BITS
956 This value can be used to select the number of bits to use to
957 determine the random offset to the base address of vma regions
958 resulting from mmap allocations. This value will be bounded
959 by the architecture's minimum and maximum supported values.
961 This value can be changed after boot using the
962 /proc/sys/vm/mmap_rnd_bits tunable
964 config HAVE_ARCH_MMAP_RND_COMPAT_BITS
967 An arch should select this symbol if it supports running applications
968 in compatibility mode, supports setting a variable number of bits for
969 use in establishing the base address for mmap allocations, has MMU
970 enabled and provides values for both:
971 - ARCH_MMAP_RND_COMPAT_BITS_MIN
972 - ARCH_MMAP_RND_COMPAT_BITS_MAX
974 config ARCH_MMAP_RND_COMPAT_BITS_MIN
977 config ARCH_MMAP_RND_COMPAT_BITS_MAX
980 config ARCH_MMAP_RND_COMPAT_BITS_DEFAULT
983 config ARCH_MMAP_RND_COMPAT_BITS
984 int "Number of bits to use for ASLR of mmap base address for compatible applications" if EXPERT
985 range ARCH_MMAP_RND_COMPAT_BITS_MIN ARCH_MMAP_RND_COMPAT_BITS_MAX
986 default ARCH_MMAP_RND_COMPAT_BITS_DEFAULT if ARCH_MMAP_RND_COMPAT_BITS_DEFAULT
987 default ARCH_MMAP_RND_COMPAT_BITS_MIN
988 depends on HAVE_ARCH_MMAP_RND_COMPAT_BITS
990 This value can be used to select the number of bits to use to
991 determine the random offset to the base address of vma regions
992 resulting from mmap allocations for compatible applications This
993 value will be bounded by the architecture's minimum and maximum
996 This value can be changed after boot using the
997 /proc/sys/vm/mmap_rnd_compat_bits tunable
999 config HAVE_ARCH_COMPAT_MMAP_BASES
1002 This allows 64bit applications to invoke 32-bit mmap() syscall
1003 and vice-versa 32-bit applications to call 64-bit mmap().
1004 Required for applications doing different bitness syscalls.
1006 config PAGE_SIZE_LESS_THAN_64KB
1008 depends on !ARM64_64K_PAGES
1009 depends on !IA64_PAGE_SIZE_64KB
1010 depends on !PAGE_SIZE_64KB
1011 depends on !PARISC_PAGE_SIZE_64KB
1012 depends on !PPC_64K_PAGES
1013 depends on PAGE_SIZE_LESS_THAN_256KB
1015 config PAGE_SIZE_LESS_THAN_256KB
1017 depends on !PPC_256K_PAGES
1018 depends on !PAGE_SIZE_256KB
1020 # This allows to use a set of generic functions to determine mmap base
1021 # address by giving priority to top-down scheme only if the process
1022 # is not in legacy mode (compat task, unlimited stack size or
1023 # sysctl_legacy_va_layout).
1024 # Architecture that selects this option can provide its own version of:
1026 config ARCH_WANT_DEFAULT_TOPDOWN_MMAP_LAYOUT
1029 select ARCH_HAS_ELF_RANDOMIZE
1031 config HAVE_STACK_VALIDATION
1034 Architecture supports the 'objtool check' host tool command, which
1035 performs compile-time stack metadata validation.
1037 config HAVE_RELIABLE_STACKTRACE
1040 Architecture has either save_stack_trace_tsk_reliable() or
1041 arch_stack_walk_reliable() function which only returns a stack trace
1042 if it can guarantee the trace is reliable.
1044 config HAVE_ARCH_HASH
1048 If this is set, the architecture provides an <asm/hash.h>
1049 file which provides platform-specific implementations of some
1050 functions in <linux/hash.h> or fs/namei.c.
1052 config HAVE_ARCH_NVRAM_OPS
1061 config CLONE_BACKWARDS
1064 Architecture has tls passed as the 4th argument of clone(2),
1067 config CLONE_BACKWARDS2
1070 Architecture has the first two arguments of clone(2) swapped.
1072 config CLONE_BACKWARDS3
1075 Architecture has tls passed as the 3rd argument of clone(2),
1078 config ODD_RT_SIGACTION
1081 Architecture has unusual rt_sigaction(2) arguments
1083 config OLD_SIGSUSPEND
1086 Architecture has old sigsuspend(2) syscall, of one-argument variety
1088 config OLD_SIGSUSPEND3
1091 Even weirder antique ABI - three-argument sigsuspend(2)
1093 config OLD_SIGACTION
1096 Architecture has old sigaction(2) syscall. Nope, not the same
1097 as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2),
1098 but fairly different variant of sigaction(2), thanks to OSF/1
1101 config COMPAT_OLD_SIGACTION
1104 config COMPAT_32BIT_TIME
1105 bool "Provide system calls for 32-bit time_t"
1106 default !64BIT || COMPAT
1108 This enables 32 bit time_t support in addition to 64 bit time_t support.
1109 This is relevant on all 32-bit architectures, and 64-bit architectures
1110 as part of compat syscall handling.
1112 config ARCH_NO_PREEMPT
1115 config ARCH_EPHEMERAL_INODES
1118 An arch should select this symbol if it doesn't keep track of inode
1119 instances on its own, but instead relies on something else (e.g. the
1120 host kernel for an UML kernel).
1122 config ARCH_SUPPORTS_RT
1125 config CPU_NO_EFFICIENT_FFS
1128 config HAVE_ARCH_VMAP_STACK
1131 An arch should select this symbol if it can support kernel stacks
1132 in vmalloc space. This means:
1134 - vmalloc space must be large enough to hold many kernel stacks.
1135 This may rule out many 32-bit architectures.
1137 - Stacks in vmalloc space need to work reliably. For example, if
1138 vmap page tables are created on demand, either this mechanism
1139 needs to work while the stack points to a virtual address with
1140 unpopulated page tables or arch code (switch_to() and switch_mm(),
1141 most likely) needs to ensure that the stack's page table entries
1142 are populated before running on a possibly unpopulated stack.
1144 - If the stack overflows into a guard page, something reasonable
1145 should happen. The definition of "reasonable" is flexible, but
1146 instantly rebooting without logging anything would be unfriendly.
1150 bool "Use a virtually-mapped stack"
1151 depends on HAVE_ARCH_VMAP_STACK
1152 depends on !KASAN || KASAN_HW_TAGS || KASAN_VMALLOC
1154 Enable this if you want the use virtually-mapped kernel stacks
1155 with guard pages. This causes kernel stack overflows to be
1156 caught immediately rather than causing difficult-to-diagnose
1159 To use this with software KASAN modes, the architecture must support
1160 backing virtual mappings with real shadow memory, and KASAN_VMALLOC
1163 config HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET
1166 An arch should select this symbol if it can support kernel stack
1167 offset randomization with calls to add_random_kstack_offset()
1168 during syscall entry and choose_random_kstack_offset() during
1169 syscall exit. Careful removal of -fstack-protector-strong and
1170 -fstack-protector should also be applied to the entry code and
1171 closely examined, as the artificial stack bump looks like an array
1172 to the compiler, so it will attempt to add canary checks regardless
1173 of the static branch state.
1175 config RANDOMIZE_KSTACK_OFFSET
1176 bool "Support for randomizing kernel stack offset on syscall entry" if EXPERT
1178 depends on HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET
1179 depends on INIT_STACK_NONE || !CC_IS_CLANG || CLANG_VERSION >= 140000
1181 The kernel stack offset can be randomized (after pt_regs) by
1182 roughly 5 bits of entropy, frustrating memory corruption
1183 attacks that depend on stack address determinism or
1184 cross-syscall address exposures.
1186 The feature is controlled via the "randomize_kstack_offset=on/off"
1187 kernel boot param, and if turned off has zero overhead due to its use
1188 of static branches (see JUMP_LABEL).
1192 config RANDOMIZE_KSTACK_OFFSET_DEFAULT
1193 bool "Default state of kernel stack offset randomization"
1194 depends on RANDOMIZE_KSTACK_OFFSET
1196 Kernel stack offset randomization is controlled by kernel boot param
1197 "randomize_kstack_offset=on/off", and this config chooses the default
1200 config ARCH_OPTIONAL_KERNEL_RWX
1203 config ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
1206 config ARCH_HAS_STRICT_KERNEL_RWX
1209 config STRICT_KERNEL_RWX
1210 bool "Make kernel text and rodata read-only" if ARCH_OPTIONAL_KERNEL_RWX
1211 depends on ARCH_HAS_STRICT_KERNEL_RWX
1212 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
1214 If this is set, kernel text and rodata memory will be made read-only,
1215 and non-text memory will be made non-executable. This provides
1216 protection against certain security exploits (e.g. executing the heap
1219 These features are considered standard security practice these days.
1220 You should say Y here in almost all cases.
1222 config ARCH_HAS_STRICT_MODULE_RWX
1225 config STRICT_MODULE_RWX
1226 bool "Set loadable kernel module data as NX and text as RO" if ARCH_OPTIONAL_KERNEL_RWX
1227 depends on ARCH_HAS_STRICT_MODULE_RWX && MODULES
1228 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
1230 If this is set, module text and rodata memory will be made read-only,
1231 and non-text memory will be made non-executable. This provides
1232 protection against certain security exploits (e.g. writing to text)
1234 # select if the architecture provides an asm/dma-direct.h header
1235 config ARCH_HAS_PHYS_TO_DMA
1238 config HAVE_ARCH_COMPILER_H
1241 An architecture can select this if it provides an
1242 asm/compiler.h header that should be included after
1243 linux/compiler-*.h in order to override macro definitions that those
1244 headers generally provide.
1246 config HAVE_ARCH_PREL32_RELOCATIONS
1249 May be selected by an architecture if it supports place-relative
1250 32-bit relocations, both in the toolchain and in the module loader,
1251 in which case relative references can be used in special sections
1252 for PCI fixup, initcalls etc which are only half the size on 64 bit
1253 architectures, and don't require runtime relocation on relocatable
1256 config ARCH_USE_MEMREMAP_PROT
1259 config LOCK_EVENT_COUNTS
1260 bool "Locking event counts collection"
1263 Enable light-weight counting of various locking related events
1264 in the system with minimal performance impact. This reduces
1265 the chance of application behavior change because of timing
1266 differences. The counts are reported via debugfs.
1268 # Select if the architecture has support for applying RELR relocations.
1269 config ARCH_HAS_RELR
1273 bool "Use RELR relocation packing"
1274 depends on ARCH_HAS_RELR && TOOLS_SUPPORT_RELR
1277 Store the kernel's dynamic relocations in the RELR relocation packing
1278 format. Requires a compatible linker (LLD supports this feature), as
1279 well as compatible NM and OBJCOPY utilities (llvm-nm and llvm-objcopy
1282 config ARCH_HAS_MEM_ENCRYPT
1285 config ARCH_HAS_CC_PLATFORM
1288 config HAVE_SPARSE_SYSCALL_NR
1291 An architecture should select this if its syscall numbering is sparse
1292 to save space. For example, MIPS architecture has a syscall array with
1293 entries at 4000, 5000 and 6000 locations. This option turns on syscall
1294 related optimizations for a given architecture.
1296 config ARCH_HAS_VDSO_DATA
1299 config HAVE_STATIC_CALL
1302 config HAVE_STATIC_CALL_INLINE
1304 depends on HAVE_STATIC_CALL
1306 config HAVE_PREEMPT_DYNAMIC
1309 config HAVE_PREEMPT_DYNAMIC_CALL
1311 depends on HAVE_STATIC_CALL
1312 select HAVE_PREEMPT_DYNAMIC
1314 An architecture should select this if it can handle the preemption
1315 model being selected at boot time using static calls.
1317 Where an architecture selects HAVE_STATIC_CALL_INLINE, any call to a
1318 preemption function will be patched directly.
1320 Where an architecture does not select HAVE_STATIC_CALL_INLINE, any
1321 call to a preemption function will go through a trampoline, and the
1322 trampoline will be patched.
1324 It is strongly advised to support inline static call to avoid any
1327 config HAVE_PREEMPT_DYNAMIC_KEY
1329 depends on HAVE_ARCH_JUMP_LABEL && CC_HAS_ASM_GOTO
1330 select HAVE_PREEMPT_DYNAMIC
1332 An architecture should select this if it can handle the preemption
1333 model being selected at boot time using static keys.
1335 Each preemption function will be given an early return based on a
1336 static key. This should have slightly lower overhead than non-inline
1337 static calls, as this effectively inlines each trampoline into the
1338 start of its callee. This may avoid redundant work, and may
1339 integrate better with CFI schemes.
1341 This will have greater overhead than using inline static calls as
1342 the call to the preemption function cannot be entirely elided.
1344 config ARCH_WANT_LD_ORPHAN_WARN
1347 An arch should select this symbol once all linker sections are explicitly
1348 included, size-asserted, or discarded in the linker scripts. This is
1349 important because we never want expected sections to be placed heuristically
1350 by the linker, since the locations of such sections can change between linker
1353 config HAVE_ARCH_PFN_VALID
1356 config ARCH_SUPPORTS_DEBUG_PAGEALLOC
1359 config ARCH_SUPPORTS_PAGE_TABLE_CHECK
1362 config ARCH_SPLIT_ARG64
1365 If a 32-bit architecture requires 64-bit arguments to be split into
1366 pairs of 32-bit arguments, select this option.
1368 config ARCH_HAS_ELFCORE_COMPAT
1371 config ARCH_HAS_PARANOID_L1D_FLUSH
1374 config DYNAMIC_SIGFRAME
1377 # Select, if arch has a named attribute group bound to NUMA device nodes.
1378 config HAVE_ARCH_NODE_DEV_GROUP
1381 source "kernel/gcov/Kconfig"
1383 source "scripts/gcc-plugins/Kconfig"