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 SET_FS 28 bool 29 30config HOTPLUG_SMT 31 bool 32 33config GENERIC_ENTRY 34 bool 35 36config KPROBES 37 bool "Kprobes" 38 depends on MODULES 39 depends on HAVE_KPROBES 40 select KALLSYMS 41 help 42 Kprobes allows you to trap at almost any kernel address and 43 execute a callback function. register_kprobe() establishes 44 a probepoint and specifies the callback. Kprobes is useful 45 for kernel debugging, non-intrusive instrumentation and testing. 46 If in doubt, say "N". 47 48config JUMP_LABEL 49 bool "Optimize very unlikely/likely branches" 50 depends on HAVE_ARCH_JUMP_LABEL 51 depends on CC_HAS_ASM_GOTO 52 help 53 This option enables a transparent branch optimization that 54 makes certain almost-always-true or almost-always-false branch 55 conditions even cheaper to execute within the kernel. 56 57 Certain performance-sensitive kernel code, such as trace points, 58 scheduler functionality, networking code and KVM have such 59 branches and include support for this optimization technique. 60 61 If it is detected that the compiler has support for "asm goto", 62 the kernel will compile such branches with just a nop 63 instruction. When the condition flag is toggled to true, the 64 nop will be converted to a jump instruction to execute the 65 conditional block of instructions. 66 67 This technique lowers overhead and stress on the branch prediction 68 of the processor and generally makes the kernel faster. The update 69 of the condition is slower, but those are always very rare. 70 71 ( On 32-bit x86, the necessary options added to the compiler 72 flags may increase the size of the kernel slightly. ) 73 74config STATIC_KEYS_SELFTEST 75 bool "Static key selftest" 76 depends on JUMP_LABEL 77 help 78 Boot time self-test of the branch patching code. 79 80config STATIC_CALL_SELFTEST 81 bool "Static call selftest" 82 depends on HAVE_STATIC_CALL 83 help 84 Boot time self-test of the call patching code. 85 86config OPTPROBES 87 def_bool y 88 depends on KPROBES && HAVE_OPTPROBES 89 select TASKS_RCU if PREEMPTION 90 91config KPROBES_ON_FTRACE 92 def_bool y 93 depends on KPROBES && HAVE_KPROBES_ON_FTRACE 94 depends on DYNAMIC_FTRACE_WITH_REGS 95 help 96 If function tracer is enabled and the arch supports full 97 passing of pt_regs to function tracing, then kprobes can 98 optimize on top of function tracing. 99 100config UPROBES 101 def_bool n 102 depends on ARCH_SUPPORTS_UPROBES 103 help 104 Uprobes is the user-space counterpart to kprobes: they 105 enable instrumentation applications (such as 'perf probe') 106 to establish unintrusive probes in user-space binaries and 107 libraries, by executing handler functions when the probes 108 are hit by user-space applications. 109 110 ( These probes come in the form of single-byte breakpoints, 111 managed by the kernel and kept transparent to the probed 112 application. ) 113 114config HAVE_64BIT_ALIGNED_ACCESS 115 def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS 116 help 117 Some architectures require 64 bit accesses to be 64 bit 118 aligned, which also requires structs containing 64 bit values 119 to be 64 bit aligned too. This includes some 32 bit 120 architectures which can do 64 bit accesses, as well as 64 bit 121 architectures without unaligned access. 122 123 This symbol should be selected by an architecture if 64 bit 124 accesses are required to be 64 bit aligned in this way even 125 though it is not a 64 bit architecture. 126 127 See Documentation/core-api/unaligned-memory-access.rst for 128 more information on the topic of unaligned memory accesses. 129 130config HAVE_EFFICIENT_UNALIGNED_ACCESS 131 bool 132 help 133 Some architectures are unable to perform unaligned accesses 134 without the use of get_unaligned/put_unaligned. Others are 135 unable to perform such accesses efficiently (e.g. trap on 136 unaligned access and require fixing it up in the exception 137 handler.) 138 139 This symbol should be selected by an architecture if it can 140 perform unaligned accesses efficiently to allow different 141 code paths to be selected for these cases. Some network 142 drivers, for example, could opt to not fix up alignment 143 problems with received packets if doing so would not help 144 much. 145 146 See Documentation/core-api/unaligned-memory-access.rst for more 147 information on the topic of unaligned memory accesses. 148 149config ARCH_USE_BUILTIN_BSWAP 150 bool 151 help 152 Modern versions of GCC (since 4.4) have builtin functions 153 for handling byte-swapping. Using these, instead of the old 154 inline assembler that the architecture code provides in the 155 __arch_bswapXX() macros, allows the compiler to see what's 156 happening and offers more opportunity for optimisation. In 157 particular, the compiler will be able to combine the byteswap 158 with a nearby load or store and use load-and-swap or 159 store-and-swap instructions if the architecture has them. It 160 should almost *never* result in code which is worse than the 161 hand-coded assembler in <asm/swab.h>. But just in case it 162 does, the use of the builtins is optional. 163 164 Any architecture with load-and-swap or store-and-swap 165 instructions should set this. And it shouldn't hurt to set it 166 on architectures that don't have such instructions. 167 168config KRETPROBES 169 def_bool y 170 depends on KPROBES && HAVE_KRETPROBES 171 172config USER_RETURN_NOTIFIER 173 bool 174 depends on HAVE_USER_RETURN_NOTIFIER 175 help 176 Provide a kernel-internal notification when a cpu is about to 177 switch to user mode. 178 179config HAVE_IOREMAP_PROT 180 bool 181 182config HAVE_KPROBES 183 bool 184 185config HAVE_KRETPROBES 186 bool 187 188config HAVE_OPTPROBES 189 bool 190 191config HAVE_KPROBES_ON_FTRACE 192 bool 193 194config HAVE_FUNCTION_ERROR_INJECTION 195 bool 196 197config HAVE_NMI 198 bool 199 200# 201# An arch should select this if it provides all these things: 202# 203# task_pt_regs() in asm/processor.h or asm/ptrace.h 204# arch_has_single_step() if there is hardware single-step support 205# arch_has_block_step() if there is hardware block-step support 206# asm/syscall.h supplying asm-generic/syscall.h interface 207# linux/regset.h user_regset interfaces 208# CORE_DUMP_USE_REGSET #define'd in linux/elf.h 209# TIF_SYSCALL_TRACE calls tracehook_report_syscall_{entry,exit} 210# TIF_NOTIFY_RESUME calls tracehook_notify_resume() 211# signal delivery calls tracehook_signal_handler() 212# 213config HAVE_ARCH_TRACEHOOK 214 bool 215 216config HAVE_DMA_CONTIGUOUS 217 bool 218 219config GENERIC_SMP_IDLE_THREAD 220 bool 221 222config GENERIC_IDLE_POLL_SETUP 223 bool 224 225config ARCH_HAS_FORTIFY_SOURCE 226 bool 227 help 228 An architecture should select this when it can successfully 229 build and run with CONFIG_FORTIFY_SOURCE. 230 231# 232# Select if the arch provides a historic keepinit alias for the retain_initrd 233# command line option 234# 235config ARCH_HAS_KEEPINITRD 236 bool 237 238# Select if arch has all set_memory_ro/rw/x/nx() functions in asm/cacheflush.h 239config ARCH_HAS_SET_MEMORY 240 bool 241 242# Select if arch has all set_direct_map_invalid/default() functions 243config ARCH_HAS_SET_DIRECT_MAP 244 bool 245 246# 247# Select if the architecture provides the arch_dma_set_uncached symbol to 248# either provide an uncached segment alias for a DMA allocation, or 249# to remap the page tables in place. 250# 251config ARCH_HAS_DMA_SET_UNCACHED 252 bool 253 254# 255# Select if the architectures provides the arch_dma_clear_uncached symbol 256# to undo an in-place page table remap for uncached access. 257# 258config ARCH_HAS_DMA_CLEAR_UNCACHED 259 bool 260 261# Select if arch init_task must go in the __init_task_data section 262config ARCH_TASK_STRUCT_ON_STACK 263 bool 264 265# Select if arch has its private alloc_task_struct() function 266config ARCH_TASK_STRUCT_ALLOCATOR 267 bool 268 269config HAVE_ARCH_THREAD_STRUCT_WHITELIST 270 bool 271 depends on !ARCH_TASK_STRUCT_ALLOCATOR 272 help 273 An architecture should select this to provide hardened usercopy 274 knowledge about what region of the thread_struct should be 275 whitelisted for copying to userspace. Normally this is only the 276 FPU registers. Specifically, arch_thread_struct_whitelist() 277 should be implemented. Without this, the entire thread_struct 278 field in task_struct will be left whitelisted. 279 280# Select if arch has its private alloc_thread_stack() function 281config ARCH_THREAD_STACK_ALLOCATOR 282 bool 283 284# Select if arch wants to size task_struct dynamically via arch_task_struct_size: 285config ARCH_WANTS_DYNAMIC_TASK_STRUCT 286 bool 287 288config ARCH_WANTS_NO_INSTR 289 bool 290 help 291 An architecture should select this if the noinstr macro is being used on 292 functions to denote that the toolchain should avoid instrumenting such 293 functions and is required for correctness. 294 295config ARCH_32BIT_OFF_T 296 bool 297 depends on !64BIT 298 help 299 All new 32-bit architectures should have 64-bit off_t type on 300 userspace side which corresponds to the loff_t kernel type. This 301 is the requirement for modern ABIs. Some existing architectures 302 still support 32-bit off_t. This option is enabled for all such 303 architectures explicitly. 304 305# Selected by 64 bit architectures which have a 32 bit f_tinode in struct ustat 306config ARCH_32BIT_USTAT_F_TINODE 307 bool 308 309config HAVE_ASM_MODVERSIONS 310 bool 311 help 312 This symbol should be selected by an architecture if it provides 313 <asm/asm-prototypes.h> to support the module versioning for symbols 314 exported from assembly code. 315 316config HAVE_REGS_AND_STACK_ACCESS_API 317 bool 318 help 319 This symbol should be selected by an architecture if it supports 320 the API needed to access registers and stack entries from pt_regs, 321 declared in asm/ptrace.h 322 For example the kprobes-based event tracer needs this API. 323 324config HAVE_RSEQ 325 bool 326 depends on HAVE_REGS_AND_STACK_ACCESS_API 327 help 328 This symbol should be selected by an architecture if it 329 supports an implementation of restartable sequences. 330 331config HAVE_FUNCTION_ARG_ACCESS_API 332 bool 333 help 334 This symbol should be selected by an architecture if it supports 335 the API needed to access function arguments from pt_regs, 336 declared in asm/ptrace.h 337 338config HAVE_HW_BREAKPOINT 339 bool 340 depends on PERF_EVENTS 341 342config HAVE_MIXED_BREAKPOINTS_REGS 343 bool 344 depends on HAVE_HW_BREAKPOINT 345 help 346 Depending on the arch implementation of hardware breakpoints, 347 some of them have separate registers for data and instruction 348 breakpoints addresses, others have mixed registers to store 349 them but define the access type in a control register. 350 Select this option if your arch implements breakpoints under the 351 latter fashion. 352 353config HAVE_USER_RETURN_NOTIFIER 354 bool 355 356config HAVE_PERF_EVENTS_NMI 357 bool 358 help 359 System hardware can generate an NMI using the perf event 360 subsystem. Also has support for calculating CPU cycle events 361 to determine how many clock cycles in a given period. 362 363config HAVE_HARDLOCKUP_DETECTOR_PERF 364 bool 365 depends on HAVE_PERF_EVENTS_NMI 366 help 367 The arch chooses to use the generic perf-NMI-based hardlockup 368 detector. Must define HAVE_PERF_EVENTS_NMI. 369 370config HAVE_NMI_WATCHDOG 371 depends on HAVE_NMI 372 bool 373 help 374 The arch provides a low level NMI watchdog. It provides 375 asm/nmi.h, and defines its own arch_touch_nmi_watchdog(). 376 377config HAVE_HARDLOCKUP_DETECTOR_ARCH 378 bool 379 select HAVE_NMI_WATCHDOG 380 help 381 The arch chooses to provide its own hardlockup detector, which is 382 a superset of the HAVE_NMI_WATCHDOG. It also conforms to config 383 interfaces and parameters provided by hardlockup detector subsystem. 384 385config HAVE_PERF_REGS 386 bool 387 help 388 Support selective register dumps for perf events. This includes 389 bit-mapping of each registers and a unique architecture id. 390 391config HAVE_PERF_USER_STACK_DUMP 392 bool 393 help 394 Support user stack dumps for perf event samples. This needs 395 access to the user stack pointer which is not unified across 396 architectures. 397 398config HAVE_ARCH_JUMP_LABEL 399 bool 400 401config HAVE_ARCH_JUMP_LABEL_RELATIVE 402 bool 403 404config MMU_GATHER_TABLE_FREE 405 bool 406 407config MMU_GATHER_RCU_TABLE_FREE 408 bool 409 select MMU_GATHER_TABLE_FREE 410 411config MMU_GATHER_PAGE_SIZE 412 bool 413 414config MMU_GATHER_NO_RANGE 415 bool 416 417config MMU_GATHER_NO_GATHER 418 bool 419 depends on MMU_GATHER_TABLE_FREE 420 421config ARCH_WANT_IRQS_OFF_ACTIVATE_MM 422 bool 423 help 424 Temporary select until all architectures can be converted to have 425 irqs disabled over activate_mm. Architectures that do IPI based TLB 426 shootdowns should enable this. 427 428config ARCH_HAVE_NMI_SAFE_CMPXCHG 429 bool 430 431config HAVE_ALIGNED_STRUCT_PAGE 432 bool 433 help 434 This makes sure that struct pages are double word aligned and that 435 e.g. the SLUB allocator can perform double word atomic operations 436 on a struct page for better performance. However selecting this 437 might increase the size of a struct page by a word. 438 439config HAVE_CMPXCHG_LOCAL 440 bool 441 442config HAVE_CMPXCHG_DOUBLE 443 bool 444 445config ARCH_WEAK_RELEASE_ACQUIRE 446 bool 447 448config ARCH_WANT_IPC_PARSE_VERSION 449 bool 450 451config ARCH_WANT_COMPAT_IPC_PARSE_VERSION 452 bool 453 454config ARCH_WANT_OLD_COMPAT_IPC 455 select ARCH_WANT_COMPAT_IPC_PARSE_VERSION 456 bool 457 458config HAVE_ARCH_SECCOMP 459 bool 460 help 461 An arch should select this symbol to support seccomp mode 1 (the fixed 462 syscall policy), and must provide an overrides for __NR_seccomp_sigreturn, 463 and compat syscalls if the asm-generic/seccomp.h defaults need adjustment: 464 - __NR_seccomp_read_32 465 - __NR_seccomp_write_32 466 - __NR_seccomp_exit_32 467 - __NR_seccomp_sigreturn_32 468 469config HAVE_ARCH_SECCOMP_FILTER 470 bool 471 select HAVE_ARCH_SECCOMP 472 help 473 An arch should select this symbol if it provides all of these things: 474 - all the requirements for HAVE_ARCH_SECCOMP 475 - syscall_get_arch() 476 - syscall_get_arguments() 477 - syscall_rollback() 478 - syscall_set_return_value() 479 - SIGSYS siginfo_t support 480 - secure_computing is called from a ptrace_event()-safe context 481 - secure_computing return value is checked and a return value of -1 482 results in the system call being skipped immediately. 483 - seccomp syscall wired up 484 - if !HAVE_SPARSE_SYSCALL_NR, have SECCOMP_ARCH_NATIVE, 485 SECCOMP_ARCH_NATIVE_NR, SECCOMP_ARCH_NATIVE_NAME defined. If 486 COMPAT is supported, have the SECCOMP_ARCH_COMPAT* defines too. 487 488config SECCOMP 489 prompt "Enable seccomp to safely execute untrusted bytecode" 490 def_bool y 491 depends on HAVE_ARCH_SECCOMP 492 help 493 This kernel feature is useful for number crunching applications 494 that may need to handle untrusted bytecode during their 495 execution. By using pipes or other transports made available 496 to the process as file descriptors supporting the read/write 497 syscalls, it's possible to isolate those applications in their 498 own address space using seccomp. Once seccomp is enabled via 499 prctl(PR_SET_SECCOMP) or the seccomp() syscall, it cannot be 500 disabled and the task is only allowed to execute a few safe 501 syscalls defined by each seccomp mode. 502 503 If unsure, say Y. 504 505config SECCOMP_FILTER 506 def_bool y 507 depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET 508 help 509 Enable tasks to build secure computing environments defined 510 in terms of Berkeley Packet Filter programs which implement 511 task-defined system call filtering polices. 512 513 See Documentation/userspace-api/seccomp_filter.rst for details. 514 515config SECCOMP_CACHE_DEBUG 516 bool "Show seccomp filter cache status in /proc/pid/seccomp_cache" 517 depends on SECCOMP_FILTER && !HAVE_SPARSE_SYSCALL_NR 518 depends on PROC_FS 519 help 520 This enables the /proc/pid/seccomp_cache interface to monitor 521 seccomp cache data. The file format is subject to change. Reading 522 the file requires CAP_SYS_ADMIN. 523 524 This option is for debugging only. Enabling presents the risk that 525 an adversary may be able to infer the seccomp filter logic. 526 527 If unsure, say N. 528 529config HAVE_ARCH_STACKLEAK 530 bool 531 help 532 An architecture should select this if it has the code which 533 fills the used part of the kernel stack with the STACKLEAK_POISON 534 value before returning from system calls. 535 536config HAVE_STACKPROTECTOR 537 bool 538 help 539 An arch should select this symbol if: 540 - it has implemented a stack canary (e.g. __stack_chk_guard) 541 542config STACKPROTECTOR 543 bool "Stack Protector buffer overflow detection" 544 depends on HAVE_STACKPROTECTOR 545 depends on $(cc-option,-fstack-protector) 546 default y 547 help 548 This option turns on the "stack-protector" GCC feature. This 549 feature puts, at the beginning of functions, a canary value on 550 the stack just before the return address, and validates 551 the value just before actually returning. Stack based buffer 552 overflows (that need to overwrite this return address) now also 553 overwrite the canary, which gets detected and the attack is then 554 neutralized via a kernel panic. 555 556 Functions will have the stack-protector canary logic added if they 557 have an 8-byte or larger character array on the stack. 558 559 This feature requires gcc version 4.2 or above, or a distribution 560 gcc with the feature backported ("-fstack-protector"). 561 562 On an x86 "defconfig" build, this feature adds canary checks to 563 about 3% of all kernel functions, which increases kernel code size 564 by about 0.3%. 565 566config STACKPROTECTOR_STRONG 567 bool "Strong Stack Protector" 568 depends on STACKPROTECTOR 569 depends on $(cc-option,-fstack-protector-strong) 570 default y 571 help 572 Functions will have the stack-protector canary logic added in any 573 of the following conditions: 574 575 - local variable's address used as part of the right hand side of an 576 assignment or function argument 577 - local variable is an array (or union containing an array), 578 regardless of array type or length 579 - uses register local variables 580 581 This feature requires gcc version 4.9 or above, or a distribution 582 gcc with the feature backported ("-fstack-protector-strong"). 583 584 On an x86 "defconfig" build, this feature adds canary checks to 585 about 20% of all kernel functions, which increases the kernel code 586 size by about 2%. 587 588config ARCH_SUPPORTS_SHADOW_CALL_STACK 589 bool 590 help 591 An architecture should select this if it supports Clang's Shadow 592 Call Stack and implements runtime support for shadow stack 593 switching. 594 595config SHADOW_CALL_STACK 596 bool "Clang Shadow Call Stack" 597 depends on CC_IS_CLANG && ARCH_SUPPORTS_SHADOW_CALL_STACK 598 depends on DYNAMIC_FTRACE_WITH_REGS || !FUNCTION_GRAPH_TRACER 599 help 600 This option enables Clang's Shadow Call Stack, which uses a 601 shadow stack to protect function return addresses from being 602 overwritten by an attacker. More information can be found in 603 Clang's documentation: 604 605 https://clang.llvm.org/docs/ShadowCallStack.html 606 607 Note that security guarantees in the kernel differ from the 608 ones documented for user space. The kernel must store addresses 609 of shadow stacks in memory, which means an attacker capable of 610 reading and writing arbitrary memory may be able to locate them 611 and hijack control flow by modifying the stacks. 612 613config LTO 614 bool 615 help 616 Selected if the kernel will be built using the compiler's LTO feature. 617 618config LTO_CLANG 619 bool 620 select LTO 621 help 622 Selected if the kernel will be built using Clang's LTO feature. 623 624config ARCH_SUPPORTS_LTO_CLANG 625 bool 626 help 627 An architecture should select this option if it supports: 628 - compiling with Clang, 629 - compiling inline assembly with Clang's integrated assembler, 630 - and linking with LLD. 631 632config ARCH_SUPPORTS_LTO_CLANG_THIN 633 bool 634 help 635 An architecture should select this option if it can support Clang's 636 ThinLTO mode. 637 638config HAS_LTO_CLANG 639 def_bool y 640 # Clang >= 11: https://github.com/ClangBuiltLinux/linux/issues/510 641 depends on CC_IS_CLANG && CLANG_VERSION >= 110000 && LD_IS_LLD && AS_IS_LLVM 642 depends on $(success,$(NM) --help | head -n 1 | grep -qi llvm) 643 depends on $(success,$(AR) --help | head -n 1 | grep -qi llvm) 644 depends on ARCH_SUPPORTS_LTO_CLANG 645 depends on !FTRACE_MCOUNT_USE_RECORDMCOUNT 646 depends on !KASAN || KASAN_HW_TAGS 647 depends on !GCOV_KERNEL 648 help 649 The compiler and Kconfig options support building with Clang's 650 LTO. 651 652choice 653 prompt "Link Time Optimization (LTO)" 654 default LTO_NONE 655 help 656 This option enables Link Time Optimization (LTO), which allows the 657 compiler to optimize binaries globally. 658 659 If unsure, select LTO_NONE. Note that LTO is very resource-intensive 660 so it's disabled by default. 661 662config LTO_NONE 663 bool "None" 664 help 665 Build the kernel normally, without Link Time Optimization (LTO). 666 667config LTO_CLANG_FULL 668 bool "Clang Full LTO (EXPERIMENTAL)" 669 depends on HAS_LTO_CLANG 670 depends on !COMPILE_TEST 671 select LTO_CLANG 672 help 673 This option enables Clang's full Link Time Optimization (LTO), which 674 allows the compiler to optimize the kernel globally. If you enable 675 this option, the compiler generates LLVM bitcode instead of ELF 676 object files, and the actual compilation from bitcode happens at 677 the LTO link step, which may take several minutes depending on the 678 kernel configuration. More information can be found from LLVM's 679 documentation: 680 681 https://llvm.org/docs/LinkTimeOptimization.html 682 683 During link time, this option can use a large amount of RAM, and 684 may take much longer than the ThinLTO option. 685 686config LTO_CLANG_THIN 687 bool "Clang ThinLTO (EXPERIMENTAL)" 688 depends on HAS_LTO_CLANG && ARCH_SUPPORTS_LTO_CLANG_THIN 689 select LTO_CLANG 690 help 691 This option enables Clang's ThinLTO, which allows for parallel 692 optimization and faster incremental compiles compared to the 693 CONFIG_LTO_CLANG_FULL option. More information can be found 694 from Clang's documentation: 695 696 https://clang.llvm.org/docs/ThinLTO.html 697 698 If unsure, say Y. 699endchoice 700 701config ARCH_SUPPORTS_CFI_CLANG 702 bool 703 help 704 An architecture should select this option if it can support Clang's 705 Control-Flow Integrity (CFI) checking. 706 707config CFI_CLANG 708 bool "Use Clang's Control Flow Integrity (CFI)" 709 depends on LTO_CLANG && ARCH_SUPPORTS_CFI_CLANG 710 # Clang >= 12: 711 # - https://bugs.llvm.org/show_bug.cgi?id=46258 712 # - https://bugs.llvm.org/show_bug.cgi?id=47479 713 depends on CLANG_VERSION >= 120000 714 select KALLSYMS 715 help 716 This option enables Clang’s forward-edge Control Flow Integrity 717 (CFI) checking, where the compiler injects a runtime check to each 718 indirect function call to ensure the target is a valid function with 719 the correct static type. This restricts possible call targets and 720 makes it more difficult for an attacker to exploit bugs that allow 721 the modification of stored function pointers. More information can be 722 found from Clang's documentation: 723 724 https://clang.llvm.org/docs/ControlFlowIntegrity.html 725 726config CFI_CLANG_SHADOW 727 bool "Use CFI shadow to speed up cross-module checks" 728 default y 729 depends on CFI_CLANG && MODULES 730 help 731 If you select this option, the kernel builds a fast look-up table of 732 CFI check functions in loaded modules to reduce performance overhead. 733 734 If unsure, say Y. 735 736config CFI_PERMISSIVE 737 bool "Use CFI in permissive mode" 738 depends on CFI_CLANG 739 help 740 When selected, Control Flow Integrity (CFI) violations result in a 741 warning instead of a kernel panic. This option should only be used 742 for finding indirect call type mismatches during development. 743 744 If unsure, say N. 745 746config HAVE_ARCH_WITHIN_STACK_FRAMES 747 bool 748 help 749 An architecture should select this if it can walk the kernel stack 750 frames to determine if an object is part of either the arguments 751 or local variables (i.e. that it excludes saved return addresses, 752 and similar) by implementing an inline arch_within_stack_frames(), 753 which is used by CONFIG_HARDENED_USERCOPY. 754 755config HAVE_CONTEXT_TRACKING 756 bool 757 help 758 Provide kernel/user boundaries probes necessary for subsystems 759 that need it, such as userspace RCU extended quiescent state. 760 Syscalls need to be wrapped inside user_exit()-user_enter(), either 761 optimized behind static key or through the slow path using TIF_NOHZ 762 flag. Exceptions handlers must be wrapped as well. Irqs are already 763 protected inside rcu_irq_enter/rcu_irq_exit() but preemption or signal 764 handling on irq exit still need to be protected. 765 766config HAVE_CONTEXT_TRACKING_OFFSTACK 767 bool 768 help 769 Architecture neither relies on exception_enter()/exception_exit() 770 nor on schedule_user(). Also preempt_schedule_notrace() and 771 preempt_schedule_irq() can't be called in a preemptible section 772 while context tracking is CONTEXT_USER. This feature reflects a sane 773 entry implementation where the following requirements are met on 774 critical entry code, ie: before user_exit() or after user_enter(): 775 776 - Critical entry code isn't preemptible (or better yet: 777 not interruptible). 778 - No use of RCU read side critical sections, unless rcu_nmi_enter() 779 got called. 780 - No use of instrumentation, unless instrumentation_begin() got 781 called. 782 783config HAVE_TIF_NOHZ 784 bool 785 help 786 Arch relies on TIF_NOHZ and syscall slow path to implement context 787 tracking calls to user_enter()/user_exit(). 788 789config HAVE_VIRT_CPU_ACCOUNTING 790 bool 791 792config HAVE_VIRT_CPU_ACCOUNTING_IDLE 793 bool 794 help 795 Architecture has its own way to account idle CPU time and therefore 796 doesn't implement vtime_account_idle(). 797 798config ARCH_HAS_SCALED_CPUTIME 799 bool 800 801config HAVE_VIRT_CPU_ACCOUNTING_GEN 802 bool 803 default y if 64BIT 804 help 805 With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit. 806 Before enabling this option, arch code must be audited 807 to ensure there are no races in concurrent read/write of 808 cputime_t. For example, reading/writing 64-bit cputime_t on 809 some 32-bit arches may require multiple accesses, so proper 810 locking is needed to protect against concurrent accesses. 811 812config HAVE_IRQ_TIME_ACCOUNTING 813 bool 814 help 815 Archs need to ensure they use a high enough resolution clock to 816 support irq time accounting and then call enable_sched_clock_irqtime(). 817 818config HAVE_MOVE_PUD 819 bool 820 help 821 Architectures that select this are able to move page tables at the 822 PUD level. If there are only 3 page table levels, the move effectively 823 happens at the PGD level. 824 825config HAVE_MOVE_PMD 826 bool 827 help 828 Archs that select this are able to move page tables at the PMD level. 829 830config HAVE_ARCH_TRANSPARENT_HUGEPAGE 831 bool 832 833config HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD 834 bool 835 836config HAVE_ARCH_HUGE_VMAP 837 bool 838 839# 840# Archs that select this would be capable of PMD-sized vmaps (i.e., 841# arch_vmap_pmd_supported() returns true), and they must make no assumptions 842# that vmalloc memory is mapped with PAGE_SIZE ptes. The VM_NO_HUGE_VMAP flag 843# can be used to prohibit arch-specific allocations from using hugepages to 844# help with this (e.g., modules may require it). 845# 846config HAVE_ARCH_HUGE_VMALLOC 847 depends on HAVE_ARCH_HUGE_VMAP 848 bool 849 850config ARCH_WANT_HUGE_PMD_SHARE 851 bool 852 853config HAVE_ARCH_SOFT_DIRTY 854 bool 855 856config HAVE_MOD_ARCH_SPECIFIC 857 bool 858 help 859 The arch uses struct mod_arch_specific to store data. Many arches 860 just need a simple module loader without arch specific data - those 861 should not enable this. 862 863config MODULES_USE_ELF_RELA 864 bool 865 help 866 Modules only use ELF RELA relocations. Modules with ELF REL 867 relocations will give an error. 868 869config MODULES_USE_ELF_REL 870 bool 871 help 872 Modules only use ELF REL relocations. Modules with ELF RELA 873 relocations will give an error. 874 875config HAVE_IRQ_EXIT_ON_IRQ_STACK 876 bool 877 help 878 Architecture doesn't only execute the irq handler on the irq stack 879 but also irq_exit(). This way we can process softirqs on this irq 880 stack instead of switching to a new one when we call __do_softirq() 881 in the end of an hardirq. 882 This spares a stack switch and improves cache usage on softirq 883 processing. 884 885config HAVE_SOFTIRQ_ON_OWN_STACK 886 bool 887 help 888 Architecture provides a function to run __do_softirq() on a 889 seperate stack. 890 891config PGTABLE_LEVELS 892 int 893 default 2 894 895config ARCH_HAS_ELF_RANDOMIZE 896 bool 897 help 898 An architecture supports choosing randomized locations for 899 stack, mmap, brk, and ET_DYN. Defined functions: 900 - arch_mmap_rnd() 901 - arch_randomize_brk() 902 903config HAVE_ARCH_MMAP_RND_BITS 904 bool 905 help 906 An arch should select this symbol if it supports setting a variable 907 number of bits for use in establishing the base address for mmap 908 allocations, has MMU enabled and provides values for both: 909 - ARCH_MMAP_RND_BITS_MIN 910 - ARCH_MMAP_RND_BITS_MAX 911 912config HAVE_EXIT_THREAD 913 bool 914 help 915 An architecture implements exit_thread. 916 917config ARCH_MMAP_RND_BITS_MIN 918 int 919 920config ARCH_MMAP_RND_BITS_MAX 921 int 922 923config ARCH_MMAP_RND_BITS_DEFAULT 924 int 925 926config ARCH_MMAP_RND_BITS 927 int "Number of bits to use for ASLR of mmap base address" if EXPERT 928 range ARCH_MMAP_RND_BITS_MIN ARCH_MMAP_RND_BITS_MAX 929 default ARCH_MMAP_RND_BITS_DEFAULT if ARCH_MMAP_RND_BITS_DEFAULT 930 default ARCH_MMAP_RND_BITS_MIN 931 depends on HAVE_ARCH_MMAP_RND_BITS 932 help 933 This value can be used to select the number of bits to use to 934 determine the random offset to the base address of vma regions 935 resulting from mmap allocations. This value will be bounded 936 by the architecture's minimum and maximum supported values. 937 938 This value can be changed after boot using the 939 /proc/sys/vm/mmap_rnd_bits tunable 940 941config HAVE_ARCH_MMAP_RND_COMPAT_BITS 942 bool 943 help 944 An arch should select this symbol if it supports running applications 945 in compatibility mode, supports setting a variable number of bits for 946 use in establishing the base address for mmap allocations, has MMU 947 enabled and provides values for both: 948 - ARCH_MMAP_RND_COMPAT_BITS_MIN 949 - ARCH_MMAP_RND_COMPAT_BITS_MAX 950 951config ARCH_MMAP_RND_COMPAT_BITS_MIN 952 int 953 954config ARCH_MMAP_RND_COMPAT_BITS_MAX 955 int 956 957config ARCH_MMAP_RND_COMPAT_BITS_DEFAULT 958 int 959 960config ARCH_MMAP_RND_COMPAT_BITS 961 int "Number of bits to use for ASLR of mmap base address for compatible applications" if EXPERT 962 range ARCH_MMAP_RND_COMPAT_BITS_MIN ARCH_MMAP_RND_COMPAT_BITS_MAX 963 default ARCH_MMAP_RND_COMPAT_BITS_DEFAULT if ARCH_MMAP_RND_COMPAT_BITS_DEFAULT 964 default ARCH_MMAP_RND_COMPAT_BITS_MIN 965 depends on HAVE_ARCH_MMAP_RND_COMPAT_BITS 966 help 967 This value can be used to select the number of bits to use to 968 determine the random offset to the base address of vma regions 969 resulting from mmap allocations for compatible applications This 970 value will be bounded by the architecture's minimum and maximum 971 supported values. 972 973 This value can be changed after boot using the 974 /proc/sys/vm/mmap_rnd_compat_bits tunable 975 976config HAVE_ARCH_COMPAT_MMAP_BASES 977 bool 978 help 979 This allows 64bit applications to invoke 32-bit mmap() syscall 980 and vice-versa 32-bit applications to call 64-bit mmap(). 981 Required for applications doing different bitness syscalls. 982 983# This allows to use a set of generic functions to determine mmap base 984# address by giving priority to top-down scheme only if the process 985# is not in legacy mode (compat task, unlimited stack size or 986# sysctl_legacy_va_layout). 987# Architecture that selects this option can provide its own version of: 988# - STACK_RND_MASK 989config ARCH_WANT_DEFAULT_TOPDOWN_MMAP_LAYOUT 990 bool 991 depends on MMU 992 select ARCH_HAS_ELF_RANDOMIZE 993 994config HAVE_STACK_VALIDATION 995 bool 996 help 997 Architecture supports the 'objtool check' host tool command, which 998 performs compile-time stack metadata validation. 999 1000config HAVE_RELIABLE_STACKTRACE 1001 bool 1002 help 1003 Architecture has either save_stack_trace_tsk_reliable() or 1004 arch_stack_walk_reliable() function which only returns a stack trace 1005 if it can guarantee the trace is reliable. 1006 1007config HAVE_ARCH_HASH 1008 bool 1009 default n 1010 help 1011 If this is set, the architecture provides an <asm/hash.h> 1012 file which provides platform-specific implementations of some 1013 functions in <linux/hash.h> or fs/namei.c. 1014 1015config HAVE_ARCH_NVRAM_OPS 1016 bool 1017 1018config ISA_BUS_API 1019 def_bool ISA 1020 1021# 1022# ABI hall of shame 1023# 1024config CLONE_BACKWARDS 1025 bool 1026 help 1027 Architecture has tls passed as the 4th argument of clone(2), 1028 not the 5th one. 1029 1030config CLONE_BACKWARDS2 1031 bool 1032 help 1033 Architecture has the first two arguments of clone(2) swapped. 1034 1035config CLONE_BACKWARDS3 1036 bool 1037 help 1038 Architecture has tls passed as the 3rd argument of clone(2), 1039 not the 5th one. 1040 1041config ODD_RT_SIGACTION 1042 bool 1043 help 1044 Architecture has unusual rt_sigaction(2) arguments 1045 1046config OLD_SIGSUSPEND 1047 bool 1048 help 1049 Architecture has old sigsuspend(2) syscall, of one-argument variety 1050 1051config OLD_SIGSUSPEND3 1052 bool 1053 help 1054 Even weirder antique ABI - three-argument sigsuspend(2) 1055 1056config OLD_SIGACTION 1057 bool 1058 help 1059 Architecture has old sigaction(2) syscall. Nope, not the same 1060 as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2), 1061 but fairly different variant of sigaction(2), thanks to OSF/1 1062 compatibility... 1063 1064config COMPAT_OLD_SIGACTION 1065 bool 1066 1067config COMPAT_32BIT_TIME 1068 bool "Provide system calls for 32-bit time_t" 1069 default !64BIT || COMPAT 1070 help 1071 This enables 32 bit time_t support in addition to 64 bit time_t support. 1072 This is relevant on all 32-bit architectures, and 64-bit architectures 1073 as part of compat syscall handling. 1074 1075config ARCH_NO_PREEMPT 1076 bool 1077 1078config ARCH_EPHEMERAL_INODES 1079 def_bool n 1080 help 1081 An arch should select this symbol if it doesn't keep track of inode 1082 instances on its own, but instead relies on something else (e.g. the 1083 host kernel for an UML kernel). 1084 1085config ARCH_SUPPORTS_RT 1086 bool 1087 1088config CPU_NO_EFFICIENT_FFS 1089 def_bool n 1090 1091config HAVE_ARCH_VMAP_STACK 1092 def_bool n 1093 help 1094 An arch should select this symbol if it can support kernel stacks 1095 in vmalloc space. This means: 1096 1097 - vmalloc space must be large enough to hold many kernel stacks. 1098 This may rule out many 32-bit architectures. 1099 1100 - Stacks in vmalloc space need to work reliably. For example, if 1101 vmap page tables are created on demand, either this mechanism 1102 needs to work while the stack points to a virtual address with 1103 unpopulated page tables or arch code (switch_to() and switch_mm(), 1104 most likely) needs to ensure that the stack's page table entries 1105 are populated before running on a possibly unpopulated stack. 1106 1107 - If the stack overflows into a guard page, something reasonable 1108 should happen. The definition of "reasonable" is flexible, but 1109 instantly rebooting without logging anything would be unfriendly. 1110 1111config VMAP_STACK 1112 default y 1113 bool "Use a virtually-mapped stack" 1114 depends on HAVE_ARCH_VMAP_STACK 1115 depends on !KASAN || KASAN_HW_TAGS || KASAN_VMALLOC 1116 help 1117 Enable this if you want the use virtually-mapped kernel stacks 1118 with guard pages. This causes kernel stack overflows to be 1119 caught immediately rather than causing difficult-to-diagnose 1120 corruption. 1121 1122 To use this with software KASAN modes, the architecture must support 1123 backing virtual mappings with real shadow memory, and KASAN_VMALLOC 1124 must be enabled. 1125 1126config HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET 1127 def_bool n 1128 help 1129 An arch should select this symbol if it can support kernel stack 1130 offset randomization with calls to add_random_kstack_offset() 1131 during syscall entry and choose_random_kstack_offset() during 1132 syscall exit. Careful removal of -fstack-protector-strong and 1133 -fstack-protector should also be applied to the entry code and 1134 closely examined, as the artificial stack bump looks like an array 1135 to the compiler, so it will attempt to add canary checks regardless 1136 of the static branch state. 1137 1138config RANDOMIZE_KSTACK_OFFSET_DEFAULT 1139 bool "Randomize kernel stack offset on syscall entry" 1140 depends on HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET 1141 help 1142 The kernel stack offset can be randomized (after pt_regs) by 1143 roughly 5 bits of entropy, frustrating memory corruption 1144 attacks that depend on stack address determinism or 1145 cross-syscall address exposures. This feature is controlled 1146 by kernel boot param "randomize_kstack_offset=on/off", and this 1147 config chooses the default boot state. 1148 1149config ARCH_OPTIONAL_KERNEL_RWX 1150 def_bool n 1151 1152config ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 1153 def_bool n 1154 1155config ARCH_HAS_STRICT_KERNEL_RWX 1156 def_bool n 1157 1158config STRICT_KERNEL_RWX 1159 bool "Make kernel text and rodata read-only" if ARCH_OPTIONAL_KERNEL_RWX 1160 depends on ARCH_HAS_STRICT_KERNEL_RWX 1161 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 1162 help 1163 If this is set, kernel text and rodata memory will be made read-only, 1164 and non-text memory will be made non-executable. This provides 1165 protection against certain security exploits (e.g. executing the heap 1166 or modifying text) 1167 1168 These features are considered standard security practice these days. 1169 You should say Y here in almost all cases. 1170 1171config ARCH_HAS_STRICT_MODULE_RWX 1172 def_bool n 1173 1174config STRICT_MODULE_RWX 1175 bool "Set loadable kernel module data as NX and text as RO" if ARCH_OPTIONAL_KERNEL_RWX 1176 depends on ARCH_HAS_STRICT_MODULE_RWX && MODULES 1177 default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 1178 help 1179 If this is set, module text and rodata memory will be made read-only, 1180 and non-text memory will be made non-executable. This provides 1181 protection against certain security exploits (e.g. writing to text) 1182 1183# select if the architecture provides an asm/dma-direct.h header 1184config ARCH_HAS_PHYS_TO_DMA 1185 bool 1186 1187config HAVE_ARCH_COMPILER_H 1188 bool 1189 help 1190 An architecture can select this if it provides an 1191 asm/compiler.h header that should be included after 1192 linux/compiler-*.h in order to override macro definitions that those 1193 headers generally provide. 1194 1195config HAVE_ARCH_PREL32_RELOCATIONS 1196 bool 1197 help 1198 May be selected by an architecture if it supports place-relative 1199 32-bit relocations, both in the toolchain and in the module loader, 1200 in which case relative references can be used in special sections 1201 for PCI fixup, initcalls etc which are only half the size on 64 bit 1202 architectures, and don't require runtime relocation on relocatable 1203 kernels. 1204 1205config ARCH_USE_MEMREMAP_PROT 1206 bool 1207 1208config LOCK_EVENT_COUNTS 1209 bool "Locking event counts collection" 1210 depends on DEBUG_FS 1211 help 1212 Enable light-weight counting of various locking related events 1213 in the system with minimal performance impact. This reduces 1214 the chance of application behavior change because of timing 1215 differences. The counts are reported via debugfs. 1216 1217# Select if the architecture has support for applying RELR relocations. 1218config ARCH_HAS_RELR 1219 bool 1220 1221config RELR 1222 bool "Use RELR relocation packing" 1223 depends on ARCH_HAS_RELR && TOOLS_SUPPORT_RELR 1224 default y 1225 help 1226 Store the kernel's dynamic relocations in the RELR relocation packing 1227 format. Requires a compatible linker (LLD supports this feature), as 1228 well as compatible NM and OBJCOPY utilities (llvm-nm and llvm-objcopy 1229 are compatible). 1230 1231config ARCH_HAS_MEM_ENCRYPT 1232 bool 1233 1234config HAVE_SPARSE_SYSCALL_NR 1235 bool 1236 help 1237 An architecture should select this if its syscall numbering is sparse 1238 to save space. For example, MIPS architecture has a syscall array with 1239 entries at 4000, 5000 and 6000 locations. This option turns on syscall 1240 related optimizations for a given architecture. 1241 1242config ARCH_HAS_VDSO_DATA 1243 bool 1244 1245config HAVE_STATIC_CALL 1246 bool 1247 1248config HAVE_STATIC_CALL_INLINE 1249 bool 1250 depends on HAVE_STATIC_CALL 1251 1252config HAVE_PREEMPT_DYNAMIC 1253 bool 1254 depends on HAVE_STATIC_CALL 1255 depends on GENERIC_ENTRY 1256 help 1257 Select this if the architecture support boot time preempt setting 1258 on top of static calls. It is strongly advised to support inline 1259 static call to avoid any overhead. 1260 1261config ARCH_WANT_LD_ORPHAN_WARN 1262 bool 1263 help 1264 An arch should select this symbol once all linker sections are explicitly 1265 included, size-asserted, or discarded in the linker scripts. This is 1266 important because we never want expected sections to be placed heuristically 1267 by the linker, since the locations of such sections can change between linker 1268 versions. 1269 1270config HAVE_ARCH_PFN_VALID 1271 bool 1272 1273config ARCH_SUPPORTS_DEBUG_PAGEALLOC 1274 bool 1275 1276config ARCH_SPLIT_ARG64 1277 bool 1278 help 1279 If a 32-bit architecture requires 64-bit arguments to be split into 1280 pairs of 32-bit arguments, select this option. 1281 1282config ARCH_HAS_ELFCORE_COMPAT 1283 bool 1284 1285source "kernel/gcov/Kconfig" 1286 1287source "scripts/gcc-plugins/Kconfig" 1288 1289endmenu 1290