1# SPDX-License-Identifier: GPL-2.0 2# 3# General architecture dependent options 4# 5 6# 7# Note: arch/$(SRCARCH)/Kconfig needs to be included first so that it can 8# override the default values in this file. 9# 10source "arch/$(SRCARCH)/Kconfig" 11 12menu "General architecture-dependent options" 13 14config CRASH_CORE 15 bool 16 17config KEXEC_CORE 18 select CRASH_CORE 19 bool 20 21config KEXEC_ELF 22 bool 23 24config HAVE_IMA_KEXEC 25 bool 26 27config HOTPLUG_SMT 28 bool 29 30config OPROFILE 31 tristate "OProfile system profiling" 32 depends on PROFILING 33 depends on HAVE_OPROFILE 34 select RING_BUFFER 35 select RING_BUFFER_ALLOW_SWAP 36 help 37 OProfile is a profiling system capable of profiling the 38 whole system, include the kernel, kernel modules, libraries, 39 and applications. 40 41 If unsure, say N. 42 43config OPROFILE_EVENT_MULTIPLEX 44 bool "OProfile multiplexing support (EXPERIMENTAL)" 45 default n 46 depends on OPROFILE && X86 47 help 48 The number of hardware counters is limited. The multiplexing 49 feature enables OProfile to gather more events than counters 50 are provided by the hardware. This is realized by switching 51 between events at a user specified time interval. 52 53 If unsure, say N. 54 55config HAVE_OPROFILE 56 bool 57 58config OPROFILE_NMI_TIMER 59 def_bool y 60 depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI && !PPC64 61 62config KPROBES 63 bool "Kprobes" 64 depends on MODULES 65 depends on HAVE_KPROBES 66 select KALLSYMS 67 help 68 Kprobes allows you to trap at almost any kernel address and 69 execute a callback function. register_kprobe() establishes 70 a probepoint and specifies the callback. Kprobes is useful 71 for kernel debugging, non-intrusive instrumentation and testing. 72 If in doubt, say "N". 73 74config JUMP_LABEL 75 bool "Optimize very unlikely/likely branches" 76 depends on HAVE_ARCH_JUMP_LABEL 77 depends on CC_HAS_ASM_GOTO 78 help 79 This option enables a transparent branch optimization that 80 makes certain almost-always-true or almost-always-false branch 81 conditions even cheaper to execute within the kernel. 82 83 Certain performance-sensitive kernel code, such as trace points, 84 scheduler functionality, networking code and KVM have such 85 branches and include support for this optimization technique. 86 87 If it is detected that the compiler has support for "asm goto", 88 the kernel will compile such branches with just a nop 89 instruction. When the condition flag is toggled to true, the 90 nop will be converted to a jump instruction to execute the 91 conditional block of instructions. 92 93 This technique lowers overhead and stress on the branch prediction 94 of the processor and generally makes the kernel faster. The update 95 of the condition is slower, but those are always very rare. 96 97 ( On 32-bit x86, the necessary options added to the compiler 98 flags may increase the size of the kernel slightly. ) 99 100config STATIC_KEYS_SELFTEST 101 bool "Static key selftest" 102 depends on JUMP_LABEL 103 help 104 Boot time self-test of the branch patching code. 105 106config OPTPROBES 107 def_bool y 108 depends on KPROBES && HAVE_OPTPROBES 109 select TASKS_RCU if PREEMPTION 110 111config KPROBES_ON_FTRACE 112 def_bool y 113 depends on KPROBES && HAVE_KPROBES_ON_FTRACE 114 depends on DYNAMIC_FTRACE_WITH_REGS 115 help 116 If function tracer is enabled and the arch supports full 117 passing of pt_regs to function tracing, then kprobes can 118 optimize on top of function tracing. 119 120config UPROBES 121 def_bool n 122 depends on ARCH_SUPPORTS_UPROBES 123 help 124 Uprobes is the user-space counterpart to kprobes: they 125 enable instrumentation applications (such as 'perf probe') 126 to establish unintrusive probes in user-space binaries and 127 libraries, by executing handler functions when the probes 128 are hit by user-space applications. 129 130 ( These probes come in the form of single-byte breakpoints, 131 managed by the kernel and kept transparent to the probed 132 application. ) 133 134config HAVE_64BIT_ALIGNED_ACCESS 135 def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS 136 help 137 Some architectures require 64 bit accesses to be 64 bit 138 aligned, which also requires structs containing 64 bit values 139 to be 64 bit aligned too. This includes some 32 bit 140 architectures which can do 64 bit accesses, as well as 64 bit 141 architectures without unaligned access. 142 143 This symbol should be selected by an architecture if 64 bit 144 accesses are required to be 64 bit aligned in this way even 145 though it is not a 64 bit architecture. 146 147 See Documentation/unaligned-memory-access.txt for more 148 information on the topic of unaligned memory accesses. 149 150config HAVE_EFFICIENT_UNALIGNED_ACCESS 151 bool 152 help 153 Some architectures are unable to perform unaligned accesses 154 without the use of get_unaligned/put_unaligned. Others are 155 unable to perform such accesses efficiently (e.g. trap on 156 unaligned access and require fixing it up in the exception 157 handler.) 158 159 This symbol should be selected by an architecture if it can 160 perform unaligned accesses efficiently to allow different 161 code paths to be selected for these cases. Some network 162 drivers, for example, could opt to not fix up alignment 163 problems with received packets if doing so would not help 164 much. 165 166 See Documentation/unaligned-memory-access.txt for more 167 information on the topic of unaligned memory accesses. 168 169config ARCH_USE_BUILTIN_BSWAP 170 bool 171 help 172 Modern versions of GCC (since 4.4) have builtin functions 173 for handling byte-swapping. Using these, instead of the old 174 inline assembler that the architecture code provides in the 175 __arch_bswapXX() macros, allows the compiler to see what's 176 happening and offers more opportunity for optimisation. In 177 particular, the compiler will be able to combine the byteswap 178 with a nearby load or store and use load-and-swap or 179 store-and-swap instructions if the architecture has them. It 180 should almost *never* result in code which is worse than the 181 hand-coded assembler in <asm/swab.h>. But just in case it 182 does, the use of the builtins is optional. 183 184 Any architecture with load-and-swap or store-and-swap 185 instructions should set this. And it shouldn't hurt to set it 186 on architectures that don't have such instructions. 187 188config KRETPROBES 189 def_bool y 190 depends on KPROBES && HAVE_KRETPROBES 191 192config USER_RETURN_NOTIFIER 193 bool 194 depends on HAVE_USER_RETURN_NOTIFIER 195 help 196 Provide a kernel-internal notification when a cpu is about to 197 switch to user mode. 198 199config HAVE_IOREMAP_PROT 200 bool 201 202config HAVE_KPROBES 203 bool 204 205config HAVE_KRETPROBES 206 bool 207 208config HAVE_OPTPROBES 209 bool 210 211config HAVE_KPROBES_ON_FTRACE 212 bool 213 214config HAVE_FUNCTION_ERROR_INJECTION 215 bool 216 217config HAVE_NMI 218 bool 219 220# 221# An arch should select this if it provides all these things: 222# 223# task_pt_regs() in asm/processor.h or asm/ptrace.h 224# arch_has_single_step() if there is hardware single-step support 225# arch_has_block_step() if there is hardware block-step support 226# asm/syscall.h supplying asm-generic/syscall.h interface 227# linux/regset.h user_regset interfaces 228# CORE_DUMP_USE_REGSET #define'd in linux/elf.h 229# TIF_SYSCALL_TRACE calls tracehook_report_syscall_{entry,exit} 230# TIF_NOTIFY_RESUME calls tracehook_notify_resume() 231# signal delivery calls tracehook_signal_handler() 232# 233config HAVE_ARCH_TRACEHOOK 234 bool 235 236config HAVE_DMA_CONTIGUOUS 237 bool 238 239config GENERIC_SMP_IDLE_THREAD 240 bool 241 242config GENERIC_IDLE_POLL_SETUP 243 bool 244 245config ARCH_HAS_FORTIFY_SOURCE 246 bool 247 help 248 An architecture should select this when it can successfully 249 build and run with CONFIG_FORTIFY_SOURCE. 250 251# 252# Select if the arch provides a historic keepinit alias for the retain_initrd 253# command line option 254# 255config ARCH_HAS_KEEPINITRD 256 bool 257 258# Select if arch has all set_memory_ro/rw/x/nx() functions in asm/cacheflush.h 259config ARCH_HAS_SET_MEMORY 260 bool 261 262# Select if arch has all set_direct_map_invalid/default() functions 263config ARCH_HAS_SET_DIRECT_MAP 264 bool 265 266# 267# Select if arch has an uncached kernel segment and provides the 268# uncached_kernel_address / cached_kernel_address symbols to use it 269# 270config ARCH_HAS_UNCACHED_SEGMENT 271 select ARCH_HAS_DMA_PREP_COHERENT 272 bool 273 274config ARCH_HAS_CPU_FINALIZE_INIT 275 bool 276 277# Select if arch init_task must go in the __init_task_data section 278config ARCH_TASK_STRUCT_ON_STACK 279 bool 280 281# Select if arch has its private alloc_task_struct() function 282config ARCH_TASK_STRUCT_ALLOCATOR 283 bool 284 285config HAVE_ARCH_THREAD_STRUCT_WHITELIST 286 bool 287 depends on !ARCH_TASK_STRUCT_ALLOCATOR 288 help 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. 295 296# Select if arch has its private alloc_thread_stack() function 297config ARCH_THREAD_STACK_ALLOCATOR 298 bool 299 300# Select if arch wants to size task_struct dynamically via arch_task_struct_size: 301config ARCH_WANTS_DYNAMIC_TASK_STRUCT 302 bool 303 304config ARCH_32BIT_OFF_T 305 bool 306 depends on !64BIT 307 help 308 All new 32-bit architectures should have 64-bit off_t type on 309 userspace side which corresponds to the loff_t kernel type. This 310 is the requirement for modern ABIs. Some existing architectures 311 still support 32-bit off_t. This option is enabled for all such 312 architectures explicitly. 313 314config HAVE_ASM_MODVERSIONS 315 bool 316 help 317 This symbol should be selected by an architecure if it provides 318 <asm/asm-prototypes.h> to support the module versioning for symbols 319 exported from assembly code. 320 321config HAVE_REGS_AND_STACK_ACCESS_API 322 bool 323 help 324 This symbol should be selected by an architecure if it supports 325 the API needed to access registers and stack entries from pt_regs, 326 declared in asm/ptrace.h 327 For example the kprobes-based event tracer needs this API. 328 329config HAVE_RSEQ 330 bool 331 depends on HAVE_REGS_AND_STACK_ACCESS_API 332 help 333 This symbol should be selected by an architecture if it 334 supports an implementation of restartable sequences. 335 336config HAVE_FUNCTION_ARG_ACCESS_API 337 bool 338 help 339 This symbol should be selected by an architecure if it supports 340 the API needed to access function arguments from pt_regs, 341 declared in asm/ptrace.h 342 343config HAVE_CLK 344 bool 345 help 346 The <linux/clk.h> calls support software clock gating and 347 thus are a key power management tool on many systems. 348 349config HAVE_HW_BREAKPOINT 350 bool 351 depends on PERF_EVENTS 352 353config HAVE_MIXED_BREAKPOINTS_REGS 354 bool 355 depends on HAVE_HW_BREAKPOINT 356 help 357 Depending on the arch implementation of hardware breakpoints, 358 some of them have separate registers for data and instruction 359 breakpoints addresses, others have mixed registers to store 360 them but define the access type in a control register. 361 Select this option if your arch implements breakpoints under the 362 latter fashion. 363 364config HAVE_USER_RETURN_NOTIFIER 365 bool 366 367config HAVE_PERF_EVENTS_NMI 368 bool 369 help 370 System hardware can generate an NMI using the perf event 371 subsystem. Also has support for calculating CPU cycle events 372 to determine how many clock cycles in a given period. 373 374config HAVE_HARDLOCKUP_DETECTOR_PERF 375 bool 376 depends on HAVE_PERF_EVENTS_NMI 377 help 378 The arch chooses to use the generic perf-NMI-based hardlockup 379 detector. Must define HAVE_PERF_EVENTS_NMI. 380 381config HAVE_NMI_WATCHDOG 382 depends on HAVE_NMI 383 bool 384 help 385 The arch provides a low level NMI watchdog. It provides 386 asm/nmi.h, and defines its own arch_touch_nmi_watchdog(). 387 388config HAVE_HARDLOCKUP_DETECTOR_ARCH 389 bool 390 select HAVE_NMI_WATCHDOG 391 help 392 The arch chooses to provide its own hardlockup detector, which is 393 a superset of the HAVE_NMI_WATCHDOG. It also conforms to config 394 interfaces and parameters provided by hardlockup detector subsystem. 395 396config HAVE_PERF_REGS 397 bool 398 help 399 Support selective register dumps for perf events. This includes 400 bit-mapping of each registers and a unique architecture id. 401 402config HAVE_PERF_USER_STACK_DUMP 403 bool 404 help 405 Support user stack dumps for perf event samples. This needs 406 access to the user stack pointer which is not unified across 407 architectures. 408 409config HAVE_ARCH_JUMP_LABEL 410 bool 411 412config HAVE_ARCH_JUMP_LABEL_RELATIVE 413 bool 414 415config HAVE_RCU_TABLE_FREE 416 bool 417 418config HAVE_MMU_GATHER_PAGE_SIZE 419 bool 420 421config MMU_GATHER_NO_RANGE 422 bool 423 424config HAVE_MMU_GATHER_NO_GATHER 425 bool 426 427config ARCH_WANT_IRQS_OFF_ACTIVATE_MM 428 bool 429 help 430 Temporary select until all architectures can be converted to have 431 irqs disabled over activate_mm. Architectures that do IPI based TLB 432 shootdowns should enable this. 433 434config ARCH_HAVE_NMI_SAFE_CMPXCHG 435 bool 436 437config HAVE_ALIGNED_STRUCT_PAGE 438 bool 439 help 440 This makes sure that struct pages are double word aligned and that 441 e.g. the SLUB allocator can perform double word atomic operations 442 on a struct page for better performance. However selecting this 443 might increase the size of a struct page by a word. 444 445config HAVE_CMPXCHG_LOCAL 446 bool 447 448config HAVE_CMPXCHG_DOUBLE 449 bool 450 451config ARCH_WEAK_RELEASE_ACQUIRE 452 bool 453 454config ARCH_WANT_IPC_PARSE_VERSION 455 bool 456 457config ARCH_WANT_COMPAT_IPC_PARSE_VERSION 458 bool 459 460config ARCH_WANT_OLD_COMPAT_IPC 461 select ARCH_WANT_COMPAT_IPC_PARSE_VERSION 462 bool 463 464config HAVE_ARCH_SECCOMP_FILTER 465 bool 466 help 467 An arch should select this symbol if it provides all of these things: 468 - syscall_get_arch() 469 - syscall_get_arguments() 470 - syscall_rollback() 471 - syscall_set_return_value() 472 - SIGSYS siginfo_t support 473 - secure_computing is called from a ptrace_event()-safe context 474 - secure_computing return value is checked and a return value of -1 475 results in the system call being skipped immediately. 476 - seccomp syscall wired up 477 478config SECCOMP_FILTER 479 def_bool y 480 depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET 481 help 482 Enable tasks to build secure computing environments defined 483 in terms of Berkeley Packet Filter programs which implement 484 task-defined system call filtering polices. 485 486 See Documentation/userspace-api/seccomp_filter.rst for details. 487 488config HAVE_ARCH_STACKLEAK 489 bool 490 help 491 An architecture should select this if it has the code which 492 fills the used part of the kernel stack with the STACKLEAK_POISON 493 value before returning from system calls. 494 495config HAVE_STACKPROTECTOR 496 bool 497 help 498 An arch should select this symbol if: 499 - it has implemented a stack canary (e.g. __stack_chk_guard) 500 501config CC_HAS_STACKPROTECTOR_NONE 502 def_bool $(cc-option,-fno-stack-protector) 503 504config STACKPROTECTOR 505 bool "Stack Protector buffer overflow detection" 506 depends on HAVE_STACKPROTECTOR 507 depends on $(cc-option,-fstack-protector) 508 default y 509 help 510 This option turns on the "stack-protector" GCC feature. This 511 feature puts, at the beginning of functions, a canary value on 512 the stack just before the return address, and validates 513 the value just before actually returning. Stack based buffer 514 overflows (that need to overwrite this return address) now also 515 overwrite the canary, which gets detected and the attack is then 516 neutralized via a kernel panic. 517 518 Functions will have the stack-protector canary logic added if they 519 have an 8-byte or larger character array on the stack. 520 521 This feature requires gcc version 4.2 or above, or a distribution 522 gcc with the feature backported ("-fstack-protector"). 523 524 On an x86 "defconfig" build, this feature adds canary checks to 525 about 3% of all kernel functions, which increases kernel code size 526 by about 0.3%. 527 528config STACKPROTECTOR_STRONG 529 bool "Strong Stack Protector" 530 depends on STACKPROTECTOR 531 depends on $(cc-option,-fstack-protector-strong) 532 default y 533 help 534 Functions will have the stack-protector canary logic added in any 535 of the following conditions: 536 537 - local variable's address used as part of the right hand side of an 538 assignment or function argument 539 - local variable is an array (or union containing an array), 540 regardless of array type or length 541 - uses register local variables 542 543 This feature requires gcc version 4.9 or above, or a distribution 544 gcc with the feature backported ("-fstack-protector-strong"). 545 546 On an x86 "defconfig" build, this feature adds canary checks to 547 about 20% of all kernel functions, which increases the kernel code 548 size by about 2%. 549 550config ARCH_SUPPORTS_SHADOW_CALL_STACK 551 bool 552 help 553 An architecture should select this if it supports Clang's Shadow 554 Call Stack, has asm/scs.h, and implements runtime support for shadow 555 stack switching. 556 557config SHADOW_CALL_STACK 558 bool "Clang Shadow Call Stack" 559 depends on ARCH_SUPPORTS_SHADOW_CALL_STACK 560 help 561 This option enables Clang's Shadow Call Stack, which uses a 562 shadow stack to protect function return addresses from being 563 overwritten by an attacker. More information can be found from 564 Clang's documentation: 565 566 https://clang.llvm.org/docs/ShadowCallStack.html 567 568 Note that security guarantees in the kernel differ from the ones 569 documented for user space. The kernel must store addresses of shadow 570 stacks used by other tasks and interrupt handlers in memory, which 571 means an attacker capable reading and writing arbitrary memory may 572 be able to locate them and hijack control flow by modifying shadow 573 stacks that are not currently in use. 574 575config SHADOW_CALL_STACK_VMAP 576 bool "Use virtually mapped shadow call stacks" 577 depends on SHADOW_CALL_STACK 578 default y 579 help 580 Use virtually mapped shadow call stacks. Selecting this option 581 provides better stack exhaustion protection, but increases per-thread 582 memory consumption as a full page is allocated for each shadow stack. 583 584config LTO 585 bool 586 587config ARCH_SUPPORTS_LTO_CLANG 588 bool 589 help 590 An architecture should select this option if it supports: 591 - compiling with Clang, 592 - compiling inline assembly with Clang's integrated assembler, 593 - and linking with LLD. 594 595config ARCH_SUPPORTS_THINLTO 596 bool 597 help 598 An architecture should select this if it supports Clang ThinLTO. 599 600config THINLTO 601 bool "Use Clang's ThinLTO (EXPERIMENTAL)" 602 depends on LTO_CLANG && ARCH_SUPPORTS_THINLTO 603 default y 604 help 605 Use ThinLTO to speed up Link Time Optimization. 606 607choice 608 prompt "Link-Time Optimization (LTO) (EXPERIMENTAL)" 609 default LTO_NONE 610 help 611 This option turns on Link-Time Optimization (LTO). 612 613config LTO_NONE 614 bool "None" 615 616config LTO_CLANG 617 bool "Use Clang's Link Time Optimization (LTO) (EXPERIMENTAL)" 618 depends on ARCH_SUPPORTS_LTO_CLANG 619 depends on !KASAN 620 depends on !FTRACE_MCOUNT_RECORD || HAVE_C_RECORDMCOUNT 621 depends on CC_IS_CLANG && CLANG_VERSION >= 100000 && LD_IS_LLD 622 select LTO 623 help 624 This option enables Clang's Link Time Optimization (LTO), which allows 625 the compiler to optimize the kernel globally at link time. If you 626 enable this option, the compiler generates LLVM IR instead of object 627 files, and the actual compilation from IR occurs at the LTO link step, 628 which may take several minutes. 629 630endchoice 631 632config CFI_CLANG 633 bool "Use Clang's Control Flow Integrity (CFI)" 634 depends on LTO_CLANG && KALLSYMS 635 depends on !GCOV_KERNEL 636 help 637 This option enables Clang's Control Flow Integrity (CFI), which adds 638 runtime checking for indirect function calls. 639 640config CFI_CLANG_SHADOW 641 bool "Use CFI shadow to speed up cross-module checks" 642 default y 643 depends on CFI_CLANG 644 help 645 If you select this option, the kernel builds a fast look-up table of 646 CFI check functions in loaded modules to reduce overhead. 647 648config CFI_PERMISSIVE 649 bool "Use CFI in permissive mode" 650 depends on CFI_CLANG 651 help 652 When selected, Control Flow Integrity (CFI) violations result in a 653 warning instead of a kernel panic. This option is useful for finding 654 CFI violations during development. 655 656config HAVE_ARCH_WITHIN_STACK_FRAMES 657 bool 658 help 659 An architecture should select this if it can walk the kernel stack 660 frames to determine if an object is part of either the arguments 661 or local variables (i.e. that it excludes saved return addresses, 662 and similar) by implementing an inline arch_within_stack_frames(), 663 which is used by CONFIG_HARDENED_USERCOPY. 664 665config HAVE_CONTEXT_TRACKING 666 bool 667 help 668 Provide kernel/user boundaries probes necessary for subsystems 669 that need it, such as userspace RCU extended quiescent state. 670 Syscalls need to be wrapped inside user_exit()-user_enter() through 671 the slow path using TIF_NOHZ flag. Exceptions handlers must be 672 wrapped as well. Irqs are already protected inside 673 rcu_irq_enter/rcu_irq_exit() but preemption or signal handling on 674 irq exit still need to be protected. 675 676config HAVE_VIRT_CPU_ACCOUNTING 677 bool 678 679config ARCH_HAS_SCALED_CPUTIME 680 bool 681 682config HAVE_VIRT_CPU_ACCOUNTING_GEN 683 bool 684 default y if 64BIT 685 help 686 With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit. 687 Before enabling this option, arch code must be audited 688 to ensure there are no races in concurrent read/write of 689 cputime_t. For example, reading/writing 64-bit cputime_t on 690 some 32-bit arches may require multiple accesses, so proper 691 locking is needed to protect against concurrent accesses. 692 693 694config HAVE_IRQ_TIME_ACCOUNTING 695 bool 696 help 697 Archs need to ensure they use a high enough resolution clock to 698 support irq time accounting and then call enable_sched_clock_irqtime(). 699 700config HAVE_MOVE_PMD 701 bool 702 help 703 Archs that select this are able to move page tables at the PMD level. 704 705config HAVE_ARCH_TRANSPARENT_HUGEPAGE 706 bool 707 708config HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD 709 bool 710 711config HAVE_ARCH_HUGE_VMAP 712 bool 713 714config ARCH_WANT_HUGE_PMD_SHARE 715 bool 716 717config HAVE_ARCH_SOFT_DIRTY 718 bool 719 720config HAVE_MOD_ARCH_SPECIFIC 721 bool 722 help 723 The arch uses struct mod_arch_specific to store data. Many arches 724 just need a simple module loader without arch specific data - those 725 should not enable this. 726 727config MODULES_USE_ELF_RELA 728 bool 729 help 730 Modules only use ELF RELA relocations. Modules with ELF REL 731 relocations will give an error. 732 733config MODULES_USE_ELF_REL 734 bool 735 help 736 Modules only use ELF REL relocations. Modules with ELF RELA 737 relocations will give an error. 738 739config HAVE_IRQ_EXIT_ON_IRQ_STACK 740 bool 741 help 742 Architecture doesn't only execute the irq handler on the irq stack 743 but also irq_exit(). This way we can process softirqs on this irq 744 stack instead of switching to a new one when we call __do_softirq() 745 in the end of an hardirq. 746 This spares a stack switch and improves cache usage on softirq 747 processing. 748 749config PGTABLE_LEVELS 750 int 751 default 2 752 753config ARCH_HAS_ELF_RANDOMIZE 754 bool 755 help 756 An architecture supports choosing randomized locations for 757 stack, mmap, brk, and ET_DYN. Defined functions: 758 - arch_mmap_rnd() 759 - arch_randomize_brk() 760 761config HAVE_ARCH_MMAP_RND_BITS 762 bool 763 help 764 An arch should select this symbol if it supports setting a variable 765 number of bits for use in establishing the base address for mmap 766 allocations, has MMU enabled and provides values for both: 767 - ARCH_MMAP_RND_BITS_MIN 768 - ARCH_MMAP_RND_BITS_MAX 769 770config HAVE_EXIT_THREAD 771 bool 772 help 773 An architecture implements exit_thread. 774 775config ARCH_MMAP_RND_BITS_MIN 776 int 777 778config ARCH_MMAP_RND_BITS_MAX 779 int 780 781config ARCH_MMAP_RND_BITS_DEFAULT 782 int 783 784config ARCH_MMAP_RND_BITS 785 int "Number of bits to use for ASLR of mmap base address" if EXPERT 786 range ARCH_MMAP_RND_BITS_MIN ARCH_MMAP_RND_BITS_MAX 787 default ARCH_MMAP_RND_BITS_DEFAULT if ARCH_MMAP_RND_BITS_DEFAULT 788 default ARCH_MMAP_RND_BITS_MIN 789 depends on HAVE_ARCH_MMAP_RND_BITS 790 help 791 This value can be used to select the number of bits to use to 792 determine the random offset to the base address of vma regions 793 resulting from mmap allocations. This value will be bounded 794 by the architecture's minimum and maximum supported values. 795 796 This value can be changed after boot using the 797 /proc/sys/vm/mmap_rnd_bits tunable 798 799config HAVE_ARCH_MMAP_RND_COMPAT_BITS 800 bool 801 help 802 An arch should select this symbol if it supports running applications 803 in compatibility mode, supports setting a variable number of bits for 804 use in establishing the base address for mmap allocations, has MMU 805 enabled and provides values for both: 806 - ARCH_MMAP_RND_COMPAT_BITS_MIN 807 - ARCH_MMAP_RND_COMPAT_BITS_MAX 808 809config ARCH_MMAP_RND_COMPAT_BITS_MIN 810 int 811 812config ARCH_MMAP_RND_COMPAT_BITS_MAX 813 int 814 815config ARCH_MMAP_RND_COMPAT_BITS_DEFAULT 816 int 817 818config ARCH_MMAP_RND_COMPAT_BITS 819 int "Number of bits to use for ASLR of mmap base address for compatible applications" if EXPERT 820 range ARCH_MMAP_RND_COMPAT_BITS_MIN ARCH_MMAP_RND_COMPAT_BITS_MAX 821 default ARCH_MMAP_RND_COMPAT_BITS_DEFAULT if ARCH_MMAP_RND_COMPAT_BITS_DEFAULT 822 default ARCH_MMAP_RND_COMPAT_BITS_MIN 823 depends on HAVE_ARCH_MMAP_RND_COMPAT_BITS 824 help 825 This value can be used to select the number of bits to use to 826 determine the random offset to the base address of vma regions 827 resulting from mmap allocations for compatible applications This 828 value will be bounded by the architecture's minimum and maximum 829 supported values. 830 831 This value can be changed after boot using the 832 /proc/sys/vm/mmap_rnd_compat_bits tunable 833 834config HAVE_ARCH_COMPAT_MMAP_BASES 835 bool 836 help 837 This allows 64bit applications to invoke 32-bit mmap() syscall 838 and vice-versa 32-bit applications to call 64-bit mmap(). 839 Required for applications doing different bitness syscalls. 840 841# This allows to use a set of generic functions to determine mmap base 842# address by giving priority to top-down scheme only if the process 843# is not in legacy mode (compat task, unlimited stack size or 844# sysctl_legacy_va_layout). 845# Architecture that selects this option can provide its own version of: 846# - STACK_RND_MASK 847config ARCH_WANT_DEFAULT_TOPDOWN_MMAP_LAYOUT 848 bool 849 depends on MMU 850 select ARCH_HAS_ELF_RANDOMIZE 851 852config HAVE_COPY_THREAD_TLS 853 bool 854 help 855 Architecture provides copy_thread_tls to accept tls argument via 856 normal C parameter passing, rather than extracting the syscall 857 argument from pt_regs. 858 859config HAVE_STACK_VALIDATION 860 bool 861 help 862 Architecture supports the 'objtool check' host tool command, which 863 performs compile-time stack metadata validation. 864 865config HAVE_RELIABLE_STACKTRACE 866 bool 867 help 868 Architecture has a save_stack_trace_tsk_reliable() function which 869 only returns a stack trace if it can guarantee the trace is reliable. 870 871config HAVE_ARCH_HASH 872 bool 873 default n 874 help 875 If this is set, the architecture provides an <asm/hash.h> 876 file which provides platform-specific implementations of some 877 functions in <linux/hash.h> or fs/namei.c. 878 879config HAVE_ARCH_NVRAM_OPS 880 bool 881 882config ISA_BUS_API 883 def_bool ISA 884 885# 886# ABI hall of shame 887# 888config CLONE_BACKWARDS 889 bool 890 help 891 Architecture has tls passed as the 4th argument of clone(2), 892 not the 5th one. 893 894config CLONE_BACKWARDS2 895 bool 896 help 897 Architecture has the first two arguments of clone(2) swapped. 898 899config CLONE_BACKWARDS3 900 bool 901 help 902 Architecture has tls passed as the 3rd argument of clone(2), 903 not the 5th one. 904 905config ODD_RT_SIGACTION 906 bool 907 help 908 Architecture has unusual rt_sigaction(2) arguments 909 910config OLD_SIGSUSPEND 911 bool 912 help 913 Architecture has old sigsuspend(2) syscall, of one-argument variety 914 915config OLD_SIGSUSPEND3 916 bool 917 help 918 Even weirder antique ABI - three-argument sigsuspend(2) 919 920config OLD_SIGACTION 921 bool 922 help 923 Architecture has old sigaction(2) syscall. Nope, not the same 924 as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2), 925 but fairly different variant of sigaction(2), thanks to OSF/1 926 compatibility... 927 928config COMPAT_OLD_SIGACTION 929 bool 930 931config 64BIT_TIME 932 def_bool y 933 help 934 This should be selected by all architectures that need to support 935 new system calls with a 64-bit time_t. This is relevant on all 32-bit 936 architectures, and 64-bit architectures as part of compat syscall 937 handling. 938 939config COMPAT_32BIT_TIME 940 def_bool !64BIT || COMPAT 941 help 942 This enables 32 bit time_t support in addition to 64 bit time_t support. 943 This is relevant on all 32-bit architectures, and 64-bit architectures 944 as part of compat syscall handling. 945 946config ARCH_NO_PREEMPT 947 bool 948 949config ARCH_SUPPORTS_RT 950 bool 951 952config CPU_NO_EFFICIENT_FFS 953 def_bool n 954 955config HAVE_ARCH_VMAP_STACK 956 def_bool n 957 help 958 An arch should select this symbol if it can support kernel stacks 959 in vmalloc space. This means: 960 961 - vmalloc space must be large enough to hold many kernel stacks. 962 This may rule out many 32-bit architectures. 963 964 - Stacks in vmalloc space need to work reliably. For example, if 965 vmap page tables are created on demand, either this mechanism 966 needs to work while the stack points to a virtual address with 967 unpopulated page tables or arch code (switch_to() and switch_mm(), 968 most likely) needs to ensure that the stack's page table entries 969 are populated before running on a possibly unpopulated stack. 970 971 - If the stack overflows into a guard page, something reasonable 972 should happen. The definition of "reasonable" is flexible, but 973 instantly rebooting without logging anything would be unfriendly. 974 975config VMAP_STACK 976 default y 977 bool "Use a virtually-mapped stack" 978 depends on HAVE_ARCH_VMAP_STACK && !KASAN 979 ---help--- 980 Enable this if you want the use virtually-mapped kernel stacks 981 with guard pages. This causes kernel stack overflows to be 982 caught immediately rather than causing difficult-to-diagnose 983 corruption. 984 985 This is presently incompatible with KASAN because KASAN expects 986 the stack to map directly to the KASAN shadow map using a formula 987 that is incorrect if the stack is in vmalloc space. 988 989config ARCH_OPTIONAL_KERNEL_RWX 990 def_bool n 991 992config ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 993 def_bool n 994 995config ARCH_HAS_STRICT_KERNEL_RWX 996 def_bool n 997 998config STRICT_KERNEL_RWX 999 bool "Make kernel text and rodata read-only" if ARCH_OPTIONAL_KERNEL_RWX 1000 depends on ARCH_HAS_STRICT_KERNEL_RWX 1001 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 1002 help 1003 If this is set, kernel text and rodata memory will be made read-only, 1004 and non-text memory will be made non-executable. This provides 1005 protection against certain security exploits (e.g. executing the heap 1006 or modifying text) 1007 1008 These features are considered standard security practice these days. 1009 You should say Y here in almost all cases. 1010 1011config ARCH_HAS_STRICT_MODULE_RWX 1012 def_bool n 1013 1014config STRICT_MODULE_RWX 1015 bool "Set loadable kernel module data as NX and text as RO" if ARCH_OPTIONAL_KERNEL_RWX 1016 depends on ARCH_HAS_STRICT_MODULE_RWX && MODULES 1017 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 1018 help 1019 If this is set, module text and rodata memory will be made read-only, 1020 and non-text memory will be made non-executable. This provides 1021 protection against certain security exploits (e.g. writing to text) 1022 1023# select if the architecture provides an asm/dma-direct.h header 1024config ARCH_HAS_PHYS_TO_DMA 1025 bool 1026 1027config HAVE_ARCH_COMPILER_H 1028 bool 1029 help 1030 An architecture can select this if it provides an 1031 asm/compiler.h header that should be included after 1032 linux/compiler-*.h in order to override macro definitions that those 1033 headers generally provide. 1034 1035config HAVE_ARCH_PREL32_RELOCATIONS 1036 bool 1037 help 1038 May be selected by an architecture if it supports place-relative 1039 32-bit relocations, both in the toolchain and in the module loader, 1040 in which case relative references can be used in special sections 1041 for PCI fixup, initcalls etc which are only half the size on 64 bit 1042 architectures, and don't require runtime relocation on relocatable 1043 kernels. 1044 1045config ARCH_USE_MEMREMAP_PROT 1046 bool 1047 1048config LOCK_EVENT_COUNTS 1049 bool "Locking event counts collection" 1050 depends on DEBUG_FS 1051 ---help--- 1052 Enable light-weight counting of various locking related events 1053 in the system with minimal performance impact. This reduces 1054 the chance of application behavior change because of timing 1055 differences. The counts are reported via debugfs. 1056 1057# Select if the architecture has support for applying RELR relocations. 1058config ARCH_HAS_RELR 1059 bool 1060 1061config RELR 1062 bool "Use RELR relocation packing" 1063 depends on ARCH_HAS_RELR && TOOLS_SUPPORT_RELR 1064 default y 1065 help 1066 Store the kernel's dynamic relocations in the RELR relocation packing 1067 format. Requires a compatible linker (LLD supports this feature), as 1068 well as compatible NM and OBJCOPY utilities (llvm-nm and llvm-objcopy 1069 are compatible). 1070 1071config ARCH_HAS_MEM_ENCRYPT 1072 bool 1073 1074source "kernel/gcov/Kconfig" 1075 1076source "scripts/gcc-plugins/Kconfig" 1077 1078endmenu 1079