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"
28 tristate "OProfile system profiling"
30 depends on HAVE_OPROFILE
32 select RING_BUFFER_ALLOW_SWAP
34 OProfile is a profiling system capable of profiling the
35 whole system, include the kernel, kernel modules, libraries,
40 config OPROFILE_EVENT_MULTIPLEX
41 bool "OProfile multiplexing support (EXPERIMENTAL)"
43 depends on OPROFILE && X86
45 The number of hardware counters is limited. The multiplexing
46 feature enables OProfile to gather more events than counters
47 are provided by the hardware. This is realized by switching
48 between events at a user specified time interval.
55 config OPROFILE_NMI_TIMER
57 depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI && !PPC64
62 depends on HAVE_KPROBES
65 Kprobes allows you to trap at almost any kernel address and
66 execute a callback function. register_kprobe() establishes
67 a probepoint and specifies the callback. Kprobes is useful
68 for kernel debugging, non-intrusive instrumentation and testing.
72 bool "Optimize very unlikely/likely branches"
73 depends on HAVE_ARCH_JUMP_LABEL
75 This option enables a transparent branch optimization that
76 makes certain almost-always-true or almost-always-false branch
77 conditions even cheaper to execute within the kernel.
79 Certain performance-sensitive kernel code, such as trace points,
80 scheduler functionality, networking code and KVM have such
81 branches and include support for this optimization technique.
83 If it is detected that the compiler has support for "asm goto",
84 the kernel will compile such branches with just a nop
85 instruction. When the condition flag is toggled to true, the
86 nop will be converted to a jump instruction to execute the
87 conditional block of instructions.
89 This technique lowers overhead and stress on the branch prediction
90 of the processor and generally makes the kernel faster. The update
91 of the condition is slower, but those are always very rare.
93 ( On 32-bit x86, the necessary options added to the compiler
94 flags may increase the size of the kernel slightly. )
96 config STATIC_KEYS_SELFTEST
97 bool "Static key selftest"
100 Boot time self-test of the branch patching code.
104 depends on KPROBES && HAVE_OPTPROBES
105 select TASKS_RCU if PREEMPT
107 config KPROBES_ON_FTRACE
109 depends on KPROBES && HAVE_KPROBES_ON_FTRACE
110 depends on DYNAMIC_FTRACE_WITH_REGS
112 If function tracer is enabled and the arch supports full
113 passing of pt_regs to function tracing, then kprobes can
114 optimize on top of function tracing.
118 depends on ARCH_SUPPORTS_UPROBES
120 Uprobes is the user-space counterpart to kprobes: they
121 enable instrumentation applications (such as 'perf probe')
122 to establish unintrusive probes in user-space binaries and
123 libraries, by executing handler functions when the probes
124 are hit by user-space applications.
126 ( These probes come in the form of single-byte breakpoints,
127 managed by the kernel and kept transparent to the probed
130 config HAVE_64BIT_ALIGNED_ACCESS
131 def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS
133 Some architectures require 64 bit accesses to be 64 bit
134 aligned, which also requires structs containing 64 bit values
135 to be 64 bit aligned too. This includes some 32 bit
136 architectures which can do 64 bit accesses, as well as 64 bit
137 architectures without unaligned access.
139 This symbol should be selected by an architecture if 64 bit
140 accesses are required to be 64 bit aligned in this way even
141 though it is not a 64 bit architecture.
143 See Documentation/unaligned-memory-access.txt for more
144 information on the topic of unaligned memory accesses.
146 config HAVE_EFFICIENT_UNALIGNED_ACCESS
149 Some architectures are unable to perform unaligned accesses
150 without the use of get_unaligned/put_unaligned. Others are
151 unable to perform such accesses efficiently (e.g. trap on
152 unaligned access and require fixing it up in the exception
155 This symbol should be selected by an architecture if it can
156 perform unaligned accesses efficiently to allow different
157 code paths to be selected for these cases. Some network
158 drivers, for example, could opt to not fix up alignment
159 problems with received packets if doing so would not help
162 See Documentation/unaligned-memory-access.txt for more
163 information on the topic of unaligned memory accesses.
165 config ARCH_USE_BUILTIN_BSWAP
168 Modern versions of GCC (since 4.4) have builtin functions
169 for handling byte-swapping. Using these, instead of the old
170 inline assembler that the architecture code provides in the
171 __arch_bswapXX() macros, allows the compiler to see what's
172 happening and offers more opportunity for optimisation. In
173 particular, the compiler will be able to combine the byteswap
174 with a nearby load or store and use load-and-swap or
175 store-and-swap instructions if the architecture has them. It
176 should almost *never* result in code which is worse than the
177 hand-coded assembler in <asm/swab.h>. But just in case it
178 does, the use of the builtins is optional.
180 Any architecture with load-and-swap or store-and-swap
181 instructions should set this. And it shouldn't hurt to set it
182 on architectures that don't have such instructions.
186 depends on KPROBES && HAVE_KRETPROBES
188 config USER_RETURN_NOTIFIER
190 depends on HAVE_USER_RETURN_NOTIFIER
192 Provide a kernel-internal notification when a cpu is about to
195 config HAVE_IOREMAP_PROT
201 config HAVE_KRETPROBES
204 config HAVE_OPTPROBES
207 config HAVE_KPROBES_ON_FTRACE
210 config HAVE_FUNCTION_ERROR_INJECTION
217 # An arch should select this if it provides all these things:
219 # task_pt_regs() in asm/processor.h or asm/ptrace.h
220 # arch_has_single_step() if there is hardware single-step support
221 # arch_has_block_step() if there is hardware block-step support
222 # asm/syscall.h supplying asm-generic/syscall.h interface
223 # linux/regset.h user_regset interfaces
224 # CORE_DUMP_USE_REGSET #define'd in linux/elf.h
225 # TIF_SYSCALL_TRACE calls tracehook_report_syscall_{entry,exit}
226 # TIF_NOTIFY_RESUME calls tracehook_notify_resume()
227 # signal delivery calls tracehook_signal_handler()
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.
247 # Select if arch has all set_memory_ro/rw/x/nx() functions in asm/cacheflush.h
248 config ARCH_HAS_SET_MEMORY
251 # Select if arch init_task must go in the __init_task_data section
252 config ARCH_TASK_STRUCT_ON_STACK
255 # Select if arch has its private alloc_task_struct() function
256 config ARCH_TASK_STRUCT_ALLOCATOR
259 config HAVE_ARCH_THREAD_STRUCT_WHITELIST
261 depends on !ARCH_TASK_STRUCT_ALLOCATOR
263 An architecture should select this to provide hardened usercopy
264 knowledge about what region of the thread_struct should be
265 whitelisted for copying to userspace. Normally this is only the
266 FPU registers. Specifically, arch_thread_struct_whitelist()
267 should be implemented. Without this, the entire thread_struct
268 field in task_struct will be left whitelisted.
270 # Select if arch has its private alloc_thread_stack() function
271 config ARCH_THREAD_STACK_ALLOCATOR
274 # Select if arch wants to size task_struct dynamically via arch_task_struct_size:
275 config ARCH_WANTS_DYNAMIC_TASK_STRUCT
278 config HAVE_REGS_AND_STACK_ACCESS_API
281 This symbol should be selected by an architecure if it supports
282 the API needed to access registers and stack entries from pt_regs,
283 declared in asm/ptrace.h
284 For example the kprobes-based event tracer needs this API.
288 depends on HAVE_REGS_AND_STACK_ACCESS_API
290 This symbol should be selected by an architecture if it
291 supports an implementation of restartable sequences.
293 config HAVE_FUNCTION_ARG_ACCESS_API
296 This symbol should be selected by an architecure if it supports
297 the API needed to access function arguments from pt_regs,
298 declared in asm/ptrace.h
303 The <linux/clk.h> calls support software clock gating and
304 thus are a key power management tool on many systems.
306 config HAVE_HW_BREAKPOINT
308 depends on PERF_EVENTS
310 config HAVE_MIXED_BREAKPOINTS_REGS
312 depends on HAVE_HW_BREAKPOINT
314 Depending on the arch implementation of hardware breakpoints,
315 some of them have separate registers for data and instruction
316 breakpoints addresses, others have mixed registers to store
317 them but define the access type in a control register.
318 Select this option if your arch implements breakpoints under the
321 config HAVE_USER_RETURN_NOTIFIER
324 config HAVE_PERF_EVENTS_NMI
327 System hardware can generate an NMI using the perf event
328 subsystem. Also has support for calculating CPU cycle events
329 to determine how many clock cycles in a given period.
331 config HAVE_HARDLOCKUP_DETECTOR_PERF
333 depends on HAVE_PERF_EVENTS_NMI
335 The arch chooses to use the generic perf-NMI-based hardlockup
336 detector. Must define HAVE_PERF_EVENTS_NMI.
338 config HAVE_NMI_WATCHDOG
342 The arch provides a low level NMI watchdog. It provides
343 asm/nmi.h, and defines its own arch_touch_nmi_watchdog().
345 config HAVE_HARDLOCKUP_DETECTOR_ARCH
347 select HAVE_NMI_WATCHDOG
349 The arch chooses to provide its own hardlockup detector, which is
350 a superset of the HAVE_NMI_WATCHDOG. It also conforms to config
351 interfaces and parameters provided by hardlockup detector subsystem.
353 config HAVE_PERF_REGS
356 Support selective register dumps for perf events. This includes
357 bit-mapping of each registers and a unique architecture id.
359 config HAVE_PERF_USER_STACK_DUMP
362 Support user stack dumps for perf event samples. This needs
363 access to the user stack pointer which is not unified across
366 config HAVE_ARCH_JUMP_LABEL
369 config HAVE_ARCH_JUMP_LABEL_RELATIVE
372 config HAVE_RCU_TABLE_FREE
375 config HAVE_RCU_TABLE_INVALIDATE
378 config ARCH_HAVE_NMI_SAFE_CMPXCHG
381 config HAVE_ALIGNED_STRUCT_PAGE
384 This makes sure that struct pages are double word aligned and that
385 e.g. the SLUB allocator can perform double word atomic operations
386 on a struct page for better performance. However selecting this
387 might increase the size of a struct page by a word.
389 config HAVE_CMPXCHG_LOCAL
392 config HAVE_CMPXCHG_DOUBLE
395 config ARCH_WEAK_RELEASE_ACQUIRE
398 config ARCH_WANT_IPC_PARSE_VERSION
401 config ARCH_WANT_COMPAT_IPC_PARSE_VERSION
404 config ARCH_WANT_OLD_COMPAT_IPC
405 select ARCH_WANT_COMPAT_IPC_PARSE_VERSION
408 config HAVE_ARCH_SECCOMP_FILTER
411 An arch should select this symbol if it provides all of these things:
413 - syscall_get_arguments()
415 - syscall_set_return_value()
416 - SIGSYS siginfo_t support
417 - secure_computing is called from a ptrace_event()-safe context
418 - secure_computing return value is checked and a return value of -1
419 results in the system call being skipped immediately.
420 - seccomp syscall wired up
422 config SECCOMP_FILTER
424 depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET
426 Enable tasks to build secure computing environments defined
427 in terms of Berkeley Packet Filter programs which implement
428 task-defined system call filtering polices.
430 See Documentation/userspace-api/seccomp_filter.rst for details.
432 config HAVE_ARCH_STACKLEAK
435 An architecture should select this if it has the code which
436 fills the used part of the kernel stack with the STACKLEAK_POISON
437 value before returning from system calls.
439 config HAVE_STACKPROTECTOR
442 An arch should select this symbol if:
443 - it has implemented a stack canary (e.g. __stack_chk_guard)
445 config CC_HAS_STACKPROTECTOR_NONE
446 def_bool $(cc-option,-fno-stack-protector)
448 config STACKPROTECTOR
449 bool "Stack Protector buffer overflow detection"
450 depends on HAVE_STACKPROTECTOR
451 depends on $(cc-option,-fstack-protector)
454 This option turns on the "stack-protector" GCC feature. This
455 feature puts, at the beginning of functions, a canary value on
456 the stack just before the return address, and validates
457 the value just before actually returning. Stack based buffer
458 overflows (that need to overwrite this return address) now also
459 overwrite the canary, which gets detected and the attack is then
460 neutralized via a kernel panic.
462 Functions will have the stack-protector canary logic added if they
463 have an 8-byte or larger character array on the stack.
465 This feature requires gcc version 4.2 or above, or a distribution
466 gcc with the feature backported ("-fstack-protector").
468 On an x86 "defconfig" build, this feature adds canary checks to
469 about 3% of all kernel functions, which increases kernel code size
472 config STACKPROTECTOR_STRONG
473 bool "Strong Stack Protector"
474 depends on STACKPROTECTOR
475 depends on $(cc-option,-fstack-protector-strong)
478 Functions will have the stack-protector canary logic added in any
479 of the following conditions:
481 - local variable's address used as part of the right hand side of an
482 assignment or function argument
483 - local variable is an array (or union containing an array),
484 regardless of array type or length
485 - uses register local variables
487 This feature requires gcc version 4.9 or above, or a distribution
488 gcc with the feature backported ("-fstack-protector-strong").
490 On an x86 "defconfig" build, this feature adds canary checks to
491 about 20% of all kernel functions, which increases the kernel code
494 config HAVE_ARCH_WITHIN_STACK_FRAMES
497 An architecture should select this if it can walk the kernel stack
498 frames to determine if an object is part of either the arguments
499 or local variables (i.e. that it excludes saved return addresses,
500 and similar) by implementing an inline arch_within_stack_frames(),
501 which is used by CONFIG_HARDENED_USERCOPY.
503 config HAVE_CONTEXT_TRACKING
506 Provide kernel/user boundaries probes necessary for subsystems
507 that need it, such as userspace RCU extended quiescent state.
508 Syscalls need to be wrapped inside user_exit()-user_enter() through
509 the slow path using TIF_NOHZ flag. Exceptions handlers must be
510 wrapped as well. Irqs are already protected inside
511 rcu_irq_enter/rcu_irq_exit() but preemption or signal handling on
512 irq exit still need to be protected.
514 config HAVE_VIRT_CPU_ACCOUNTING
517 config ARCH_HAS_SCALED_CPUTIME
520 config HAVE_VIRT_CPU_ACCOUNTING_GEN
524 With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit.
525 Before enabling this option, arch code must be audited
526 to ensure there are no races in concurrent read/write of
527 cputime_t. For example, reading/writing 64-bit cputime_t on
528 some 32-bit arches may require multiple accesses, so proper
529 locking is needed to protect against concurrent accesses.
532 config HAVE_IRQ_TIME_ACCOUNTING
535 Archs need to ensure they use a high enough resolution clock to
536 support irq time accounting and then call enable_sched_clock_irqtime().
538 config HAVE_ARCH_TRANSPARENT_HUGEPAGE
541 config HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
544 config HAVE_ARCH_HUGE_VMAP
547 config HAVE_ARCH_SOFT_DIRTY
550 config HAVE_MOD_ARCH_SPECIFIC
553 The arch uses struct mod_arch_specific to store data. Many arches
554 just need a simple module loader without arch specific data - those
555 should not enable this.
557 config MODULES_USE_ELF_RELA
560 Modules only use ELF RELA relocations. Modules with ELF REL
561 relocations will give an error.
563 config MODULES_USE_ELF_REL
566 Modules only use ELF REL relocations. Modules with ELF RELA
567 relocations will give an error.
569 config HAVE_IRQ_EXIT_ON_IRQ_STACK
572 Architecture doesn't only execute the irq handler on the irq stack
573 but also irq_exit(). This way we can process softirqs on this irq
574 stack instead of switching to a new one when we call __do_softirq()
575 in the end of an hardirq.
576 This spares a stack switch and improves cache usage on softirq
579 config PGTABLE_LEVELS
583 config ARCH_HAS_ELF_RANDOMIZE
586 An architecture supports choosing randomized locations for
587 stack, mmap, brk, and ET_DYN. Defined functions:
589 - arch_randomize_brk()
591 config HAVE_ARCH_MMAP_RND_BITS
594 An arch should select this symbol if it supports setting a variable
595 number of bits for use in establishing the base address for mmap
596 allocations, has MMU enabled and provides values for both:
597 - ARCH_MMAP_RND_BITS_MIN
598 - ARCH_MMAP_RND_BITS_MAX
600 config HAVE_EXIT_THREAD
603 An architecture implements exit_thread.
605 config ARCH_MMAP_RND_BITS_MIN
608 config ARCH_MMAP_RND_BITS_MAX
611 config ARCH_MMAP_RND_BITS_DEFAULT
614 config ARCH_MMAP_RND_BITS
615 int "Number of bits to use for ASLR of mmap base address" if EXPERT
616 range ARCH_MMAP_RND_BITS_MIN ARCH_MMAP_RND_BITS_MAX
617 default ARCH_MMAP_RND_BITS_DEFAULT if ARCH_MMAP_RND_BITS_DEFAULT
618 default ARCH_MMAP_RND_BITS_MIN
619 depends on HAVE_ARCH_MMAP_RND_BITS
621 This value can be used to select the number of bits to use to
622 determine the random offset to the base address of vma regions
623 resulting from mmap allocations. This value will be bounded
624 by the architecture's minimum and maximum supported values.
626 This value can be changed after boot using the
627 /proc/sys/vm/mmap_rnd_bits tunable
629 config HAVE_ARCH_MMAP_RND_COMPAT_BITS
632 An arch should select this symbol if it supports running applications
633 in compatibility mode, supports setting a variable number of bits for
634 use in establishing the base address for mmap allocations, has MMU
635 enabled and provides values for both:
636 - ARCH_MMAP_RND_COMPAT_BITS_MIN
637 - ARCH_MMAP_RND_COMPAT_BITS_MAX
639 config ARCH_MMAP_RND_COMPAT_BITS_MIN
642 config ARCH_MMAP_RND_COMPAT_BITS_MAX
645 config ARCH_MMAP_RND_COMPAT_BITS_DEFAULT
648 config ARCH_MMAP_RND_COMPAT_BITS
649 int "Number of bits to use for ASLR of mmap base address for compatible applications" if EXPERT
650 range ARCH_MMAP_RND_COMPAT_BITS_MIN ARCH_MMAP_RND_COMPAT_BITS_MAX
651 default ARCH_MMAP_RND_COMPAT_BITS_DEFAULT if ARCH_MMAP_RND_COMPAT_BITS_DEFAULT
652 default ARCH_MMAP_RND_COMPAT_BITS_MIN
653 depends on HAVE_ARCH_MMAP_RND_COMPAT_BITS
655 This value can be used to select the number of bits to use to
656 determine the random offset to the base address of vma regions
657 resulting from mmap allocations for compatible applications This
658 value will be bounded by the architecture's minimum and maximum
661 This value can be changed after boot using the
662 /proc/sys/vm/mmap_rnd_compat_bits tunable
664 config HAVE_ARCH_COMPAT_MMAP_BASES
667 This allows 64bit applications to invoke 32-bit mmap() syscall
668 and vice-versa 32-bit applications to call 64-bit mmap().
669 Required for applications doing different bitness syscalls.
671 config HAVE_COPY_THREAD_TLS
674 Architecture provides copy_thread_tls to accept tls argument via
675 normal C parameter passing, rather than extracting the syscall
676 argument from pt_regs.
678 config HAVE_STACK_VALIDATION
681 Architecture supports the 'objtool check' host tool command, which
682 performs compile-time stack metadata validation.
684 config HAVE_RELIABLE_STACKTRACE
687 Architecture has a save_stack_trace_tsk_reliable() function which
688 only returns a stack trace if it can guarantee the trace is reliable.
690 config HAVE_ARCH_HASH
694 If this is set, the architecture provides an <asm/hash.h>
695 file which provides platform-specific implementations of some
696 functions in <linux/hash.h> or fs/namei.c.
704 config CLONE_BACKWARDS
707 Architecture has tls passed as the 4th argument of clone(2),
710 config CLONE_BACKWARDS2
713 Architecture has the first two arguments of clone(2) swapped.
715 config CLONE_BACKWARDS3
718 Architecture has tls passed as the 3rd argument of clone(2),
721 config ODD_RT_SIGACTION
724 Architecture has unusual rt_sigaction(2) arguments
726 config OLD_SIGSUSPEND
729 Architecture has old sigsuspend(2) syscall, of one-argument variety
731 config OLD_SIGSUSPEND3
734 Even weirder antique ABI - three-argument sigsuspend(2)
739 Architecture has old sigaction(2) syscall. Nope, not the same
740 as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2),
741 but fairly different variant of sigaction(2), thanks to OSF/1
744 config COMPAT_OLD_SIGACTION
748 def_bool ARCH_HAS_64BIT_TIME
750 This should be selected by all architectures that need to support
751 new system calls with a 64-bit time_t. This is relevant on all 32-bit
752 architectures, and 64-bit architectures as part of compat syscall
755 config COMPAT_32BIT_TIME
756 def_bool (!64BIT && 64BIT_TIME) || COMPAT
758 This enables 32 bit time_t support in addition to 64 bit time_t support.
759 This is relevant on all 32-bit architectures, and 64-bit architectures
760 as part of compat syscall handling.
762 config ARCH_NO_COHERENT_DMA_MMAP
765 config ARCH_NO_PREEMPT
768 config CPU_NO_EFFICIENT_FFS
771 config HAVE_ARCH_VMAP_STACK
774 An arch should select this symbol if it can support kernel stacks
775 in vmalloc space. This means:
777 - vmalloc space must be large enough to hold many kernel stacks.
778 This may rule out many 32-bit architectures.
780 - Stacks in vmalloc space need to work reliably. For example, if
781 vmap page tables are created on demand, either this mechanism
782 needs to work while the stack points to a virtual address with
783 unpopulated page tables or arch code (switch_to() and switch_mm(),
784 most likely) needs to ensure that the stack's page table entries
785 are populated before running on a possibly unpopulated stack.
787 - If the stack overflows into a guard page, something reasonable
788 should happen. The definition of "reasonable" is flexible, but
789 instantly rebooting without logging anything would be unfriendly.
793 bool "Use a virtually-mapped stack"
794 depends on HAVE_ARCH_VMAP_STACK && !KASAN
796 Enable this if you want the use virtually-mapped kernel stacks
797 with guard pages. This causes kernel stack overflows to be
798 caught immediately rather than causing difficult-to-diagnose
801 This is presently incompatible with KASAN because KASAN expects
802 the stack to map directly to the KASAN shadow map using a formula
803 that is incorrect if the stack is in vmalloc space.
805 config ARCH_OPTIONAL_KERNEL_RWX
808 config ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
811 config ARCH_HAS_STRICT_KERNEL_RWX
814 config STRICT_KERNEL_RWX
815 bool "Make kernel text and rodata read-only" if ARCH_OPTIONAL_KERNEL_RWX
816 depends on ARCH_HAS_STRICT_KERNEL_RWX
817 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
819 If this is set, kernel text and rodata memory will be made read-only,
820 and non-text memory will be made non-executable. This provides
821 protection against certain security exploits (e.g. executing the heap
824 These features are considered standard security practice these days.
825 You should say Y here in almost all cases.
827 config ARCH_HAS_STRICT_MODULE_RWX
830 config STRICT_MODULE_RWX
831 bool "Set loadable kernel module data as NX and text as RO" if ARCH_OPTIONAL_KERNEL_RWX
832 depends on ARCH_HAS_STRICT_MODULE_RWX && MODULES
833 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
835 If this is set, module text and rodata memory will be made read-only,
836 and non-text memory will be made non-executable. This provides
837 protection against certain security exploits (e.g. writing to text)
839 # select if the architecture provides an asm/dma-direct.h header
840 config ARCH_HAS_PHYS_TO_DMA
843 config ARCH_HAS_REFCOUNT
846 An architecture selects this when it has implemented refcount_t
847 using open coded assembly primitives that provide an optimized
848 refcount_t implementation, possibly at the expense of some full
849 refcount state checks of CONFIG_REFCOUNT_FULL=y.
851 The refcount overflow check behavior, however, must be retained.
852 Catching overflows is the primary security concern for protecting
853 against bugs in reference counts.
856 bool "Perform full reference count validation at the expense of speed"
858 Enabling this switches the refcounting infrastructure from a fast
859 unchecked atomic_t implementation to a fully state checked
860 implementation, which can be (slightly) slower but provides protections
861 against various use-after-free conditions that can be used in
862 security flaw exploits.
864 config HAVE_ARCH_COMPILER_H
867 An architecture can select this if it provides an
868 asm/compiler.h header that should be included after
869 linux/compiler-*.h in order to override macro definitions that those
870 headers generally provide.
872 config HAVE_ARCH_PREL32_RELOCATIONS
875 May be selected by an architecture if it supports place-relative
876 32-bit relocations, both in the toolchain and in the module loader,
877 in which case relative references can be used in special sections
878 for PCI fixup, initcalls etc which are only half the size on 64 bit
879 architectures, and don't require runtime relocation on relocatable
882 source "kernel/gcov/Kconfig"
884 source "scripts/gcc-plugins/Kconfig"