1b2441318SGreg Kroah-Hartman# SPDX-License-Identifier: GPL-2.0 2fb32e03fSMathieu Desnoyers# 3fb32e03fSMathieu Desnoyers# General architecture dependent options 4fb32e03fSMathieu Desnoyers# 5125e5645SMathieu Desnoyers 61572497cSChristoph Hellwig# 71572497cSChristoph Hellwig# Note: arch/$(SRCARCH)/Kconfig needs to be included first so that it can 81572497cSChristoph Hellwig# override the default values in this file. 91572497cSChristoph Hellwig# 101572497cSChristoph Hellwigsource "arch/$(SRCARCH)/Kconfig" 111572497cSChristoph Hellwig 128292f4f8SSean Christophersonconfig ARCH_CONFIGURES_CPU_MITIGATIONS 138292f4f8SSean Christopherson bool 148292f4f8SSean Christopherson 158292f4f8SSean Christophersonif !ARCH_CONFIGURES_CPU_MITIGATIONS 168292f4f8SSean Christophersonconfig CPU_MITIGATIONS 178292f4f8SSean Christopherson def_bool y 188292f4f8SSean Christophersonendif 198292f4f8SSean Christopherson 2022471e13SRandy Dunlapmenu "General architecture-dependent options" 2122471e13SRandy Dunlap 22da32b581SCatalin Marinasconfig ARCH_HAS_SUBPAGE_FAULTS 23da32b581SCatalin Marinas bool 24da32b581SCatalin Marinas help 25da32b581SCatalin Marinas Select if the architecture can check permissions at sub-page 26da32b581SCatalin Marinas granularity (e.g. arm64 MTE). The probe_user_*() functions 27da32b581SCatalin Marinas must be implemented. 28da32b581SCatalin Marinas 2905736e4aSThomas Gleixnerconfig HOTPLUG_SMT 3005736e4aSThomas Gleixner bool 3105736e4aSThomas Gleixner 3238253464SMichael Ellermanconfig SMT_NUM_THREADS_DYNAMIC 3338253464SMichael Ellerman bool 3438253464SMichael Ellerman 356f062123SThomas Gleixner# Selected by HOTPLUG_CORE_SYNC_DEAD or HOTPLUG_CORE_SYNC_FULL 366f062123SThomas Gleixnerconfig HOTPLUG_CORE_SYNC 376f062123SThomas Gleixner bool 386f062123SThomas Gleixner 396f062123SThomas Gleixner# Basic CPU dead synchronization selected by architecture 406f062123SThomas Gleixnerconfig HOTPLUG_CORE_SYNC_DEAD 416f062123SThomas Gleixner bool 426f062123SThomas Gleixner select HOTPLUG_CORE_SYNC 436f062123SThomas Gleixner 446f062123SThomas Gleixner# Full CPU synchronization with alive state selected by architecture 456f062123SThomas Gleixnerconfig HOTPLUG_CORE_SYNC_FULL 466f062123SThomas Gleixner bool 476f062123SThomas Gleixner select HOTPLUG_CORE_SYNC_DEAD if HOTPLUG_CPU 486f062123SThomas Gleixner select HOTPLUG_CORE_SYNC 496f062123SThomas Gleixner 50a631be92SThomas Gleixnerconfig HOTPLUG_SPLIT_STARTUP 51a631be92SThomas Gleixner bool 52a631be92SThomas Gleixner select HOTPLUG_CORE_SYNC_FULL 53a631be92SThomas Gleixner 5418415f33SThomas Gleixnerconfig HOTPLUG_PARALLEL 5518415f33SThomas Gleixner bool 5618415f33SThomas Gleixner select HOTPLUG_SPLIT_STARTUP 5718415f33SThomas Gleixner 58142781e1SThomas Gleixnerconfig GENERIC_ENTRY 59142781e1SThomas Gleixner bool 60142781e1SThomas Gleixner 61125e5645SMathieu Desnoyersconfig KPROBES 62125e5645SMathieu Desnoyers bool "Kprobes" 6305ed160eSMasami Hiramatsu depends on MODULES 64125e5645SMathieu Desnoyers depends on HAVE_KPROBES 6505ed160eSMasami Hiramatsu select KALLSYMS 66835f14edSPaul E. McKenney select TASKS_RCU if PREEMPTION 67125e5645SMathieu Desnoyers help 68125e5645SMathieu Desnoyers Kprobes allows you to trap at almost any kernel address and 69125e5645SMathieu Desnoyers execute a callback function. register_kprobe() establishes 70125e5645SMathieu Desnoyers a probepoint and specifies the callback. Kprobes is useful 71125e5645SMathieu Desnoyers for kernel debugging, non-intrusive instrumentation and testing. 72125e5645SMathieu Desnoyers If in doubt, say "N". 73125e5645SMathieu Desnoyers 7445f81b1cSSteven Rostedtconfig JUMP_LABEL 75c5905afbSIngo Molnar bool "Optimize very unlikely/likely branches" 7645f81b1cSSteven Rostedt depends on HAVE_ARCH_JUMP_LABEL 774ab7674fSJosh Poimboeuf select OBJTOOL if HAVE_JUMP_LABEL_HACK 7845f81b1cSSteven Rostedt help 79c5905afbSIngo Molnar This option enables a transparent branch optimization that 80c5905afbSIngo Molnar makes certain almost-always-true or almost-always-false branch 81c5905afbSIngo Molnar conditions even cheaper to execute within the kernel. 8245f81b1cSSteven Rostedt 83c5905afbSIngo Molnar Certain performance-sensitive kernel code, such as trace points, 84c5905afbSIngo Molnar scheduler functionality, networking code and KVM have such 85c5905afbSIngo Molnar branches and include support for this optimization technique. 86c5905afbSIngo Molnar 87c5905afbSIngo Molnar If it is detected that the compiler has support for "asm goto", 88c5905afbSIngo Molnar the kernel will compile such branches with just a nop 89c5905afbSIngo Molnar instruction. When the condition flag is toggled to true, the 90c5905afbSIngo Molnar nop will be converted to a jump instruction to execute the 91c5905afbSIngo Molnar conditional block of instructions. 92c5905afbSIngo Molnar 93c5905afbSIngo Molnar This technique lowers overhead and stress on the branch prediction 94c5905afbSIngo Molnar of the processor and generally makes the kernel faster. The update 95c5905afbSIngo Molnar of the condition is slower, but those are always very rare. 96c5905afbSIngo Molnar 97c5905afbSIngo Molnar ( On 32-bit x86, the necessary options added to the compiler 98c5905afbSIngo Molnar flags may increase the size of the kernel slightly. ) 9945f81b1cSSteven Rostedt 1001987c947SPeter Zijlstraconfig STATIC_KEYS_SELFTEST 1011987c947SPeter Zijlstra bool "Static key selftest" 1021987c947SPeter Zijlstra depends on JUMP_LABEL 1031987c947SPeter Zijlstra help 1041987c947SPeter Zijlstra Boot time self-test of the branch patching code. 1051987c947SPeter Zijlstra 106f03c4129SPeter Zijlstraconfig STATIC_CALL_SELFTEST 107f03c4129SPeter Zijlstra bool "Static call selftest" 108f03c4129SPeter Zijlstra depends on HAVE_STATIC_CALL 109f03c4129SPeter Zijlstra help 110f03c4129SPeter Zijlstra Boot time self-test of the call patching code. 111f03c4129SPeter Zijlstra 112afd66255SMasami Hiramatsuconfig OPTPROBES 1135cc718b9SMasami Hiramatsu def_bool y 1145cc718b9SMasami Hiramatsu depends on KPROBES && HAVE_OPTPROBES 11501b1d88bSThomas Gleixner select TASKS_RCU if PREEMPTION 116afd66255SMasami Hiramatsu 117e7dbfe34SMasami Hiramatsuconfig KPROBES_ON_FTRACE 118e7dbfe34SMasami Hiramatsu def_bool y 119e7dbfe34SMasami Hiramatsu depends on KPROBES && HAVE_KPROBES_ON_FTRACE 120e7dbfe34SMasami Hiramatsu depends on DYNAMIC_FTRACE_WITH_REGS 121e7dbfe34SMasami Hiramatsu help 122e7dbfe34SMasami Hiramatsu If function tracer is enabled and the arch supports full 123e7dbfe34SMasami Hiramatsu passing of pt_regs to function tracing, then kprobes can 124e7dbfe34SMasami Hiramatsu optimize on top of function tracing. 125e7dbfe34SMasami Hiramatsu 1262b144498SSrikar Dronamrajuconfig UPROBES 12709294e31SDavid A. Long def_bool n 128e8f4aa60SAllen Pais depends on ARCH_SUPPORTS_UPROBES 1292b144498SSrikar Dronamraju help 1307b2d81d4SIngo Molnar Uprobes is the user-space counterpart to kprobes: they 1317b2d81d4SIngo Molnar enable instrumentation applications (such as 'perf probe') 1327b2d81d4SIngo Molnar to establish unintrusive probes in user-space binaries and 1337b2d81d4SIngo Molnar libraries, by executing handler functions when the probes 1347b2d81d4SIngo Molnar are hit by user-space applications. 1357b2d81d4SIngo Molnar 1367b2d81d4SIngo Molnar ( These probes come in the form of single-byte breakpoints, 1377b2d81d4SIngo Molnar managed by the kernel and kept transparent to the probed 1387b2d81d4SIngo Molnar application. ) 1392b144498SSrikar Dronamraju 140adab66b7SSteven Rostedt (VMware)config HAVE_64BIT_ALIGNED_ACCESS 141adab66b7SSteven Rostedt (VMware) def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS 142adab66b7SSteven Rostedt (VMware) help 143adab66b7SSteven Rostedt (VMware) Some architectures require 64 bit accesses to be 64 bit 144adab66b7SSteven Rostedt (VMware) aligned, which also requires structs containing 64 bit values 145adab66b7SSteven Rostedt (VMware) to be 64 bit aligned too. This includes some 32 bit 146adab66b7SSteven Rostedt (VMware) architectures which can do 64 bit accesses, as well as 64 bit 147adab66b7SSteven Rostedt (VMware) architectures without unaligned access. 148adab66b7SSteven Rostedt (VMware) 149adab66b7SSteven Rostedt (VMware) This symbol should be selected by an architecture if 64 bit 150adab66b7SSteven Rostedt (VMware) accesses are required to be 64 bit aligned in this way even 151adab66b7SSteven Rostedt (VMware) though it is not a 64 bit architecture. 152adab66b7SSteven Rostedt (VMware) 153ba1a297dSLukas Bulwahn See Documentation/core-api/unaligned-memory-access.rst for 154ba1a297dSLukas Bulwahn more information on the topic of unaligned memory accesses. 155adab66b7SSteven Rostedt (VMware) 15658340a07SJohannes Bergconfig HAVE_EFFICIENT_UNALIGNED_ACCESS 1579ba16087SJan Beulich bool 15858340a07SJohannes Berg help 15958340a07SJohannes Berg Some architectures are unable to perform unaligned accesses 16058340a07SJohannes Berg without the use of get_unaligned/put_unaligned. Others are 16158340a07SJohannes Berg unable to perform such accesses efficiently (e.g. trap on 16258340a07SJohannes Berg unaligned access and require fixing it up in the exception 16358340a07SJohannes Berg handler.) 16458340a07SJohannes Berg 16558340a07SJohannes Berg This symbol should be selected by an architecture if it can 16658340a07SJohannes Berg perform unaligned accesses efficiently to allow different 16758340a07SJohannes Berg code paths to be selected for these cases. Some network 16858340a07SJohannes Berg drivers, for example, could opt to not fix up alignment 16958340a07SJohannes Berg problems with received packets if doing so would not help 17058340a07SJohannes Berg much. 17158340a07SJohannes Berg 172c9b54d6fSMauro Carvalho Chehab See Documentation/core-api/unaligned-memory-access.rst for more 17358340a07SJohannes Berg information on the topic of unaligned memory accesses. 17458340a07SJohannes Berg 175cf66bb93SDavid Woodhouseconfig ARCH_USE_BUILTIN_BSWAP 176cf66bb93SDavid Woodhouse bool 177cf66bb93SDavid Woodhouse help 178cf66bb93SDavid Woodhouse Modern versions of GCC (since 4.4) have builtin functions 179cf66bb93SDavid Woodhouse for handling byte-swapping. Using these, instead of the old 180cf66bb93SDavid Woodhouse inline assembler that the architecture code provides in the 181cf66bb93SDavid Woodhouse __arch_bswapXX() macros, allows the compiler to see what's 182cf66bb93SDavid Woodhouse happening and offers more opportunity for optimisation. In 183cf66bb93SDavid Woodhouse particular, the compiler will be able to combine the byteswap 184cf66bb93SDavid Woodhouse with a nearby load or store and use load-and-swap or 185cf66bb93SDavid Woodhouse store-and-swap instructions if the architecture has them. It 186cf66bb93SDavid Woodhouse should almost *never* result in code which is worse than the 187cf66bb93SDavid Woodhouse hand-coded assembler in <asm/swab.h>. But just in case it 188cf66bb93SDavid Woodhouse does, the use of the builtins is optional. 189cf66bb93SDavid Woodhouse 190cf66bb93SDavid Woodhouse Any architecture with load-and-swap or store-and-swap 191cf66bb93SDavid Woodhouse instructions should set this. And it shouldn't hurt to set it 192cf66bb93SDavid Woodhouse on architectures that don't have such instructions. 193cf66bb93SDavid Woodhouse 1949edddaa2SAnanth N Mavinakayanahalliconfig KRETPROBES 1959edddaa2SAnanth N Mavinakayanahalli def_bool y 19673f9b911SMasami Hiramatsu depends on KPROBES && (HAVE_KRETPROBES || HAVE_RETHOOK) 19773f9b911SMasami Hiramatsu 19873f9b911SMasami Hiramatsuconfig KRETPROBE_ON_RETHOOK 19973f9b911SMasami Hiramatsu def_bool y 20073f9b911SMasami Hiramatsu depends on HAVE_RETHOOK 20173f9b911SMasami Hiramatsu depends on KRETPROBES 20273f9b911SMasami Hiramatsu select RETHOOK 2039edddaa2SAnanth N Mavinakayanahalli 2047c68af6eSAvi Kivityconfig USER_RETURN_NOTIFIER 2057c68af6eSAvi Kivity bool 2067c68af6eSAvi Kivity depends on HAVE_USER_RETURN_NOTIFIER 2077c68af6eSAvi Kivity help 2087c68af6eSAvi Kivity Provide a kernel-internal notification when a cpu is about to 2097c68af6eSAvi Kivity switch to user mode. 2107c68af6eSAvi Kivity 21128b2ee20SRik van Rielconfig HAVE_IOREMAP_PROT 2129ba16087SJan Beulich bool 21328b2ee20SRik van Riel 214125e5645SMathieu Desnoyersconfig HAVE_KPROBES 2159ba16087SJan Beulich bool 2169edddaa2SAnanth N Mavinakayanahalli 2179edddaa2SAnanth N Mavinakayanahalliconfig HAVE_KRETPROBES 2189ba16087SJan Beulich bool 21974bc7ceeSArthur Kepner 220afd66255SMasami Hiramatsuconfig HAVE_OPTPROBES 221afd66255SMasami Hiramatsu bool 222d314d74cSCong Wang 223e7dbfe34SMasami Hiramatsuconfig HAVE_KPROBES_ON_FTRACE 224e7dbfe34SMasami Hiramatsu bool 225e7dbfe34SMasami Hiramatsu 2261f6d3a8fSMasami Hiramatsuconfig ARCH_CORRECT_STACKTRACE_ON_KRETPROBE 2271f6d3a8fSMasami Hiramatsu bool 2281f6d3a8fSMasami Hiramatsu help 2291f6d3a8fSMasami Hiramatsu Since kretprobes modifies return address on the stack, the 2301f6d3a8fSMasami Hiramatsu stacktrace may see the kretprobe trampoline address instead 2311f6d3a8fSMasami Hiramatsu of correct one. If the architecture stacktrace code and 2321f6d3a8fSMasami Hiramatsu unwinder can adjust such entries, select this configuration. 2331f6d3a8fSMasami Hiramatsu 234540adea3SMasami Hiramatsuconfig HAVE_FUNCTION_ERROR_INJECTION 2359802d865SJosef Bacik bool 2369802d865SJosef Bacik 23742a0bb3fSPetr Mladekconfig HAVE_NMI 23842a0bb3fSPetr Mladek bool 23942a0bb3fSPetr Mladek 240a257caccSChristophe Leroyconfig HAVE_FUNCTION_DESCRIPTORS 241a257caccSChristophe Leroy bool 242a257caccSChristophe Leroy 2434aae683fSMasahiro Yamadaconfig TRACE_IRQFLAGS_SUPPORT 2444aae683fSMasahiro Yamada bool 2454aae683fSMasahiro Yamada 2464510bffbSMark Rutlandconfig TRACE_IRQFLAGS_NMI_SUPPORT 2474510bffbSMark Rutland bool 2484510bffbSMark Rutland 2491f5a4ad9SRoland McGrath# 2501f5a4ad9SRoland McGrath# An arch should select this if it provides all these things: 2511f5a4ad9SRoland McGrath# 2521f5a4ad9SRoland McGrath# task_pt_regs() in asm/processor.h or asm/ptrace.h 2531f5a4ad9SRoland McGrath# arch_has_single_step() if there is hardware single-step support 2541f5a4ad9SRoland McGrath# arch_has_block_step() if there is hardware block-step support 2551f5a4ad9SRoland McGrath# asm/syscall.h supplying asm-generic/syscall.h interface 2561f5a4ad9SRoland McGrath# linux/regset.h user_regset interfaces 2571f5a4ad9SRoland McGrath# CORE_DUMP_USE_REGSET #define'd in linux/elf.h 258153474baSEric W. Biederman# TIF_SYSCALL_TRACE calls ptrace_report_syscall_{entry,exit} 25903248addSEric W. Biederman# TIF_NOTIFY_RESUME calls resume_user_mode_work() 2601f5a4ad9SRoland McGrath# 2611f5a4ad9SRoland McGrathconfig HAVE_ARCH_TRACEHOOK 2629ba16087SJan Beulich bool 2631f5a4ad9SRoland McGrath 264c64be2bbSMarek Szyprowskiconfig HAVE_DMA_CONTIGUOUS 265c64be2bbSMarek Szyprowski bool 266c64be2bbSMarek Szyprowski 26729d5e047SThomas Gleixnerconfig GENERIC_SMP_IDLE_THREAD 26829d5e047SThomas Gleixner bool 26929d5e047SThomas Gleixner 270485cf5daSKevin Hilmanconfig GENERIC_IDLE_POLL_SETUP 271485cf5daSKevin Hilman bool 272485cf5daSKevin Hilman 2736974f0c4SDaniel Micayconfig ARCH_HAS_FORTIFY_SOURCE 2746974f0c4SDaniel Micay bool 2756974f0c4SDaniel Micay help 2766974f0c4SDaniel Micay An architecture should select this when it can successfully 2776974f0c4SDaniel Micay build and run with CONFIG_FORTIFY_SOURCE. 2786974f0c4SDaniel Micay 279d8ae8a37SChristoph Hellwig# 280d8ae8a37SChristoph Hellwig# Select if the arch provides a historic keepinit alias for the retain_initrd 281d8ae8a37SChristoph Hellwig# command line option 282d8ae8a37SChristoph Hellwig# 283d8ae8a37SChristoph Hellwigconfig ARCH_HAS_KEEPINITRD 284d8ae8a37SChristoph Hellwig bool 285d8ae8a37SChristoph Hellwig 286d2852a22SDaniel Borkmann# Select if arch has all set_memory_ro/rw/x/nx() functions in asm/cacheflush.h 287d2852a22SDaniel Borkmannconfig ARCH_HAS_SET_MEMORY 288d2852a22SDaniel Borkmann bool 289d2852a22SDaniel Borkmann 290d253ca0cSRick Edgecombe# Select if arch has all set_direct_map_invalid/default() functions 291d253ca0cSRick Edgecombeconfig ARCH_HAS_SET_DIRECT_MAP 292d253ca0cSRick Edgecombe bool 293d253ca0cSRick Edgecombe 294c30700dbSChristoph Hellwig# 295fa7e2247SChristoph Hellwig# Select if the architecture provides the arch_dma_set_uncached symbol to 296a86ecfa6SColin Ian King# either provide an uncached segment alias for a DMA allocation, or 297fa7e2247SChristoph Hellwig# to remap the page tables in place. 298c30700dbSChristoph Hellwig# 299fa7e2247SChristoph Hellwigconfig ARCH_HAS_DMA_SET_UNCACHED 300c30700dbSChristoph Hellwig bool 301c30700dbSChristoph Hellwig 302999a5d12SChristoph Hellwig# 303999a5d12SChristoph Hellwig# Select if the architectures provides the arch_dma_clear_uncached symbol 304999a5d12SChristoph Hellwig# to undo an in-place page table remap for uncached access. 305999a5d12SChristoph Hellwig# 306999a5d12SChristoph Hellwigconfig ARCH_HAS_DMA_CLEAR_UNCACHED 307f5e10287SThomas Gleixner bool 308f5e10287SThomas Gleixner 3097725acaaSThomas Gleixnerconfig ARCH_HAS_CPU_FINALIZE_INIT 3107725acaaSThomas Gleixner bool 3117725acaaSThomas Gleixner 312a6359d1eSThomas Gleixner# Select if arch init_task must go in the __init_task_data section 313a6359d1eSThomas Gleixnerconfig ARCH_TASK_STRUCT_ON_STACK 314a4a2eb49SThomas Gleixner bool 315a4a2eb49SThomas Gleixner 316f5e10287SThomas Gleixner# Select if arch has its private alloc_task_struct() function 317f5e10287SThomas Gleixnerconfig ARCH_TASK_STRUCT_ALLOCATOR 318f5e10287SThomas Gleixner bool 319f5e10287SThomas Gleixner 3205905429aSKees Cookconfig HAVE_ARCH_THREAD_STRUCT_WHITELIST 3215905429aSKees Cook bool 3225905429aSKees Cook depends on !ARCH_TASK_STRUCT_ALLOCATOR 3235905429aSKees Cook help 3245905429aSKees Cook An architecture should select this to provide hardened usercopy 3255905429aSKees Cook knowledge about what region of the thread_struct should be 3265905429aSKees Cook whitelisted for copying to userspace. Normally this is only the 3275905429aSKees Cook FPU registers. Specifically, arch_thread_struct_whitelist() 3285905429aSKees Cook should be implemented. Without this, the entire thread_struct 3295905429aSKees Cook field in task_struct will be left whitelisted. 3305905429aSKees Cook 331b235beeaSLinus Torvalds# Select if arch has its private alloc_thread_stack() function 332b235beeaSLinus Torvaldsconfig ARCH_THREAD_STACK_ALLOCATOR 333f5e10287SThomas Gleixner bool 334f5e10287SThomas Gleixner 3355aaeb5c0SIngo Molnar# Select if arch wants to size task_struct dynamically via arch_task_struct_size: 3365aaeb5c0SIngo Molnarconfig ARCH_WANTS_DYNAMIC_TASK_STRUCT 3375aaeb5c0SIngo Molnar bool 3385aaeb5c0SIngo Molnar 33951c2ee6dSNick Desaulniersconfig ARCH_WANTS_NO_INSTR 34051c2ee6dSNick Desaulniers bool 34151c2ee6dSNick Desaulniers help 34251c2ee6dSNick Desaulniers An architecture should select this if the noinstr macro is being used on 34351c2ee6dSNick Desaulniers functions to denote that the toolchain should avoid instrumenting such 34451c2ee6dSNick Desaulniers functions and is required for correctness. 34551c2ee6dSNick Desaulniers 346942fa985SYury Norovconfig ARCH_32BIT_OFF_T 347942fa985SYury Norov bool 348942fa985SYury Norov depends on !64BIT 349942fa985SYury Norov help 350942fa985SYury Norov All new 32-bit architectures should have 64-bit off_t type on 351942fa985SYury Norov userspace side which corresponds to the loff_t kernel type. This 352942fa985SYury Norov is the requirement for modern ABIs. Some existing architectures 353942fa985SYury Norov still support 32-bit off_t. This option is enabled for all such 354942fa985SYury Norov architectures explicitly. 355942fa985SYury Norov 35696c0a6a7SHeiko Carstens# Selected by 64 bit architectures which have a 32 bit f_tinode in struct ustat 35796c0a6a7SHeiko Carstensconfig ARCH_32BIT_USTAT_F_TINODE 35896c0a6a7SHeiko Carstens bool 35996c0a6a7SHeiko Carstens 3602ff2b7ecSMasahiro Yamadaconfig HAVE_ASM_MODVERSIONS 3612ff2b7ecSMasahiro Yamada bool 3622ff2b7ecSMasahiro Yamada help 363a86ecfa6SColin Ian King This symbol should be selected by an architecture if it provides 3642ff2b7ecSMasahiro Yamada <asm/asm-prototypes.h> to support the module versioning for symbols 3652ff2b7ecSMasahiro Yamada exported from assembly code. 3662ff2b7ecSMasahiro Yamada 367f850c30cSHeiko Carstensconfig HAVE_REGS_AND_STACK_ACCESS_API 368f850c30cSHeiko Carstens bool 369e01292b1SHeiko Carstens help 370a86ecfa6SColin Ian King This symbol should be selected by an architecture if it supports 371e01292b1SHeiko Carstens the API needed to access registers and stack entries from pt_regs, 372e01292b1SHeiko Carstens declared in asm/ptrace.h 373e01292b1SHeiko Carstens For example the kprobes-based event tracer needs this API. 374f850c30cSHeiko Carstens 375d7822b1eSMathieu Desnoyersconfig HAVE_RSEQ 376d7822b1eSMathieu Desnoyers bool 377d7822b1eSMathieu Desnoyers depends on HAVE_REGS_AND_STACK_ACCESS_API 378d7822b1eSMathieu Desnoyers help 379d7822b1eSMathieu Desnoyers This symbol should be selected by an architecture if it 380d7822b1eSMathieu Desnoyers supports an implementation of restartable sequences. 381d7822b1eSMathieu Desnoyers 3822f7ab126SMiguel Ojedaconfig HAVE_RUST 3832f7ab126SMiguel Ojeda bool 3842f7ab126SMiguel Ojeda help 3852f7ab126SMiguel Ojeda This symbol should be selected by an architecture if it 3862f7ab126SMiguel Ojeda supports Rust. 3872f7ab126SMiguel Ojeda 3883c88ee19SMasami Hiramatsuconfig HAVE_FUNCTION_ARG_ACCESS_API 3893c88ee19SMasami Hiramatsu bool 3903c88ee19SMasami Hiramatsu help 391a86ecfa6SColin Ian King This symbol should be selected by an architecture if it supports 3923c88ee19SMasami Hiramatsu the API needed to access function arguments from pt_regs, 3933c88ee19SMasami Hiramatsu declared in asm/ptrace.h 3943c88ee19SMasami Hiramatsu 39562a038d3SK.Prasadconfig HAVE_HW_BREAKPOINT 39662a038d3SK.Prasad bool 39799e8c5a3SFrederic Weisbecker depends on PERF_EVENTS 39862a038d3SK.Prasad 3990102752eSFrederic Weisbeckerconfig HAVE_MIXED_BREAKPOINTS_REGS 4000102752eSFrederic Weisbecker bool 4010102752eSFrederic Weisbecker depends on HAVE_HW_BREAKPOINT 4020102752eSFrederic Weisbecker help 4030102752eSFrederic Weisbecker Depending on the arch implementation of hardware breakpoints, 4040102752eSFrederic Weisbecker some of them have separate registers for data and instruction 4050102752eSFrederic Weisbecker breakpoints addresses, others have mixed registers to store 4060102752eSFrederic Weisbecker them but define the access type in a control register. 4070102752eSFrederic Weisbecker Select this option if your arch implements breakpoints under the 4080102752eSFrederic Weisbecker latter fashion. 4090102752eSFrederic Weisbecker 4107c68af6eSAvi Kivityconfig HAVE_USER_RETURN_NOTIFIER 4117c68af6eSAvi Kivity bool 412a1922ed6SIngo Molnar 413c01d4323SFrederic Weisbeckerconfig HAVE_PERF_EVENTS_NMI 414c01d4323SFrederic Weisbecker bool 41523637d47SFrederic Weisbecker help 41623637d47SFrederic Weisbecker System hardware can generate an NMI using the perf event 41723637d47SFrederic Weisbecker subsystem. Also has support for calculating CPU cycle events 41823637d47SFrederic Weisbecker to determine how many clock cycles in a given period. 419c01d4323SFrederic Weisbecker 42005a4a952SNicholas Pigginconfig HAVE_HARDLOCKUP_DETECTOR_PERF 42105a4a952SNicholas Piggin bool 42205a4a952SNicholas Piggin depends on HAVE_PERF_EVENTS_NMI 42305a4a952SNicholas Piggin help 42405a4a952SNicholas Piggin The arch chooses to use the generic perf-NMI-based hardlockup 42505a4a952SNicholas Piggin detector. Must define HAVE_PERF_EVENTS_NMI. 42605a4a952SNicholas Piggin 42705a4a952SNicholas Pigginconfig HAVE_HARDLOCKUP_DETECTOR_ARCH 42805a4a952SNicholas Piggin bool 42905a4a952SNicholas Piggin help 4301356d0b9SPetr Mladek The arch provides its own hardlockup detector implementation instead 4311356d0b9SPetr Mladek of the generic ones. 4321356d0b9SPetr Mladek 4331356d0b9SPetr Mladek It uses the same command line parameters, and sysctl interface, 4341356d0b9SPetr Mladek as the generic hardlockup detectors. 43505a4a952SNicholas Piggin 436c5e63197SJiri Olsaconfig HAVE_PERF_REGS 437c5e63197SJiri Olsa bool 438c5e63197SJiri Olsa help 439c5e63197SJiri Olsa Support selective register dumps for perf events. This includes 440c5e63197SJiri Olsa bit-mapping of each registers and a unique architecture id. 441c5e63197SJiri Olsa 442c5ebcedbSJiri Olsaconfig HAVE_PERF_USER_STACK_DUMP 443c5ebcedbSJiri Olsa bool 444c5ebcedbSJiri Olsa help 445c5ebcedbSJiri Olsa Support user stack dumps for perf event samples. This needs 446c5ebcedbSJiri Olsa access to the user stack pointer which is not unified across 447c5ebcedbSJiri Olsa architectures. 448c5ebcedbSJiri Olsa 449bf5438fcSJason Baronconfig HAVE_ARCH_JUMP_LABEL 450bf5438fcSJason Baron bool 451bf5438fcSJason Baron 45250ff18abSArd Biesheuvelconfig HAVE_ARCH_JUMP_LABEL_RELATIVE 45350ff18abSArd Biesheuvel bool 45450ff18abSArd Biesheuvel 4550d6e24d4SPeter Zijlstraconfig MMU_GATHER_TABLE_FREE 4560d6e24d4SPeter Zijlstra bool 4570d6e24d4SPeter Zijlstra 458ff2e6d72SPeter Zijlstraconfig MMU_GATHER_RCU_TABLE_FREE 45926723911SPeter Zijlstra bool 4600d6e24d4SPeter Zijlstra select MMU_GATHER_TABLE_FREE 46126723911SPeter Zijlstra 4623af4bd03SPeter Zijlstraconfig MMU_GATHER_PAGE_SIZE 463ed6a7935SPeter Zijlstra bool 464ed6a7935SPeter Zijlstra 46527796d03SPeter Zijlstraconfig MMU_GATHER_NO_RANGE 46627796d03SPeter Zijlstra bool 4671e9fdf21SPeter Zijlstra select MMU_GATHER_MERGE_VMAS 4681e9fdf21SPeter Zijlstra 4691e9fdf21SPeter Zijlstraconfig MMU_GATHER_NO_FLUSH_CACHE 4701e9fdf21SPeter Zijlstra bool 4711e9fdf21SPeter Zijlstra 4721e9fdf21SPeter Zijlstraconfig MMU_GATHER_MERGE_VMAS 4731e9fdf21SPeter Zijlstra bool 47427796d03SPeter Zijlstra 475580a586cSPeter Zijlstraconfig MMU_GATHER_NO_GATHER 476952a31c9SMartin Schwidefsky bool 4770d6e24d4SPeter Zijlstra depends on MMU_GATHER_TABLE_FREE 478952a31c9SMartin Schwidefsky 479d53c3dfbSNicholas Pigginconfig ARCH_WANT_IRQS_OFF_ACTIVATE_MM 480d53c3dfbSNicholas Piggin bool 481d53c3dfbSNicholas Piggin help 482d53c3dfbSNicholas Piggin Temporary select until all architectures can be converted to have 483d53c3dfbSNicholas Piggin irqs disabled over activate_mm. Architectures that do IPI based TLB 484d53c3dfbSNicholas Piggin shootdowns should enable this. 485d53c3dfbSNicholas Piggin 48688e3009bSNicholas Piggin# Use normal mm refcounting for MMU_LAZY_TLB kernel thread references. 48788e3009bSNicholas Piggin# MMU_LAZY_TLB_REFCOUNT=n can improve the scalability of context switching 48888e3009bSNicholas Piggin# to/from kernel threads when the same mm is running on a lot of CPUs (a large 48988e3009bSNicholas Piggin# multi-threaded application), by reducing contention on the mm refcount. 49088e3009bSNicholas Piggin# 49188e3009bSNicholas Piggin# This can be disabled if the architecture ensures no CPUs are using an mm as a 49288e3009bSNicholas Piggin# "lazy tlb" beyond its final refcount (i.e., by the time __mmdrop frees the mm 49388e3009bSNicholas Piggin# or its kernel page tables). This could be arranged by arch_exit_mmap(), or 49488e3009bSNicholas Piggin# final exit(2) TLB flush, for example. 49588e3009bSNicholas Piggin# 49688e3009bSNicholas Piggin# To implement this, an arch *must*: 49788e3009bSNicholas Piggin# Ensure the _lazy_tlb variants of mmgrab/mmdrop are used when manipulating 49888e3009bSNicholas Piggin# the lazy tlb reference of a kthread's ->active_mm (non-arch code has been 49988e3009bSNicholas Piggin# converted already). 50088e3009bSNicholas Pigginconfig MMU_LAZY_TLB_REFCOUNT 50188e3009bSNicholas Piggin def_bool y 5022655421aSNicholas Piggin depends on !MMU_LAZY_TLB_SHOOTDOWN 5032655421aSNicholas Piggin 5042655421aSNicholas Piggin# This option allows MMU_LAZY_TLB_REFCOUNT=n. It ensures no CPUs are using an 5052655421aSNicholas Piggin# mm as a lazy tlb beyond its last reference count, by shooting down these 5062655421aSNicholas Piggin# users before the mm is deallocated. __mmdrop() first IPIs all CPUs that may 5072655421aSNicholas Piggin# be using the mm as a lazy tlb, so that they may switch themselves to using 5082655421aSNicholas Piggin# init_mm for their active mm. mm_cpumask(mm) is used to determine which CPUs 5092655421aSNicholas Piggin# may be using mm as a lazy tlb mm. 5102655421aSNicholas Piggin# 5112655421aSNicholas Piggin# To implement this, an arch *must*: 5122655421aSNicholas Piggin# - At the time of the final mmdrop of the mm, ensure mm_cpumask(mm) contains 5132655421aSNicholas Piggin# at least all possible CPUs in which the mm is lazy. 5142655421aSNicholas Piggin# - It must meet the requirements for MMU_LAZY_TLB_REFCOUNT=n (see above). 5152655421aSNicholas Pigginconfig MMU_LAZY_TLB_SHOOTDOWN 5162655421aSNicholas Piggin bool 51788e3009bSNicholas Piggin 518df013ffbSHuang Yingconfig ARCH_HAVE_NMI_SAFE_CMPXCHG 519df013ffbSHuang Ying bool 520df013ffbSHuang Ying 5212e83b879SPaul E. McKenneyconfig ARCH_HAS_NMI_SAFE_THIS_CPU_OPS 5222e83b879SPaul E. McKenney bool 5232e83b879SPaul E. McKenney 52443570fd2SHeiko Carstensconfig HAVE_ALIGNED_STRUCT_PAGE 52543570fd2SHeiko Carstens bool 52643570fd2SHeiko Carstens help 52743570fd2SHeiko Carstens This makes sure that struct pages are double word aligned and that 52843570fd2SHeiko Carstens e.g. the SLUB allocator can perform double word atomic operations 52943570fd2SHeiko Carstens on a struct page for better performance. However selecting this 53043570fd2SHeiko Carstens might increase the size of a struct page by a word. 53143570fd2SHeiko Carstens 5324156153cSHeiko Carstensconfig HAVE_CMPXCHG_LOCAL 5334156153cSHeiko Carstens bool 5344156153cSHeiko Carstens 5352565409fSHeiko Carstensconfig HAVE_CMPXCHG_DOUBLE 5362565409fSHeiko Carstens bool 5372565409fSHeiko Carstens 53877e58496SPaul E. McKenneyconfig ARCH_WEAK_RELEASE_ACQUIRE 53977e58496SPaul E. McKenney bool 54077e58496SPaul E. McKenney 541c1d7e01dSWill Deaconconfig ARCH_WANT_IPC_PARSE_VERSION 542c1d7e01dSWill Deacon bool 543c1d7e01dSWill Deacon 544c1d7e01dSWill Deaconconfig ARCH_WANT_COMPAT_IPC_PARSE_VERSION 545c1d7e01dSWill Deacon bool 546c1d7e01dSWill Deacon 54748b25c43SChris Metcalfconfig ARCH_WANT_OLD_COMPAT_IPC 548c1d7e01dSWill Deacon select ARCH_WANT_COMPAT_IPC_PARSE_VERSION 54948b25c43SChris Metcalf bool 55048b25c43SChris Metcalf 551282a181bSYiFei Zhuconfig HAVE_ARCH_SECCOMP 552e2cfabdfSWill Drewry bool 553e2cfabdfSWill Drewry help 554282a181bSYiFei Zhu An arch should select this symbol to support seccomp mode 1 (the fixed 555282a181bSYiFei Zhu syscall policy), and must provide an overrides for __NR_seccomp_sigreturn, 556282a181bSYiFei Zhu and compat syscalls if the asm-generic/seccomp.h defaults need adjustment: 557282a181bSYiFei Zhu - __NR_seccomp_read_32 558282a181bSYiFei Zhu - __NR_seccomp_write_32 559282a181bSYiFei Zhu - __NR_seccomp_exit_32 560282a181bSYiFei Zhu - __NR_seccomp_sigreturn_32 561282a181bSYiFei Zhu 562282a181bSYiFei Zhuconfig HAVE_ARCH_SECCOMP_FILTER 563282a181bSYiFei Zhu bool 564282a181bSYiFei Zhu select HAVE_ARCH_SECCOMP 565282a181bSYiFei Zhu help 566fb0fadf9SWill Drewry An arch should select this symbol if it provides all of these things: 567282a181bSYiFei Zhu - all the requirements for HAVE_ARCH_SECCOMP 568bb6ea430SWill Drewry - syscall_get_arch() 569bb6ea430SWill Drewry - syscall_get_arguments() 570bb6ea430SWill Drewry - syscall_rollback() 571bb6ea430SWill Drewry - syscall_set_return_value() 572fb0fadf9SWill Drewry - SIGSYS siginfo_t support 573fb0fadf9SWill Drewry - secure_computing is called from a ptrace_event()-safe context 574fb0fadf9SWill Drewry - secure_computing return value is checked and a return value of -1 575fb0fadf9SWill Drewry results in the system call being skipped immediately. 57648dc92b9SKees Cook - seccomp syscall wired up 5770d8315ddSYiFei Zhu - if !HAVE_SPARSE_SYSCALL_NR, have SECCOMP_ARCH_NATIVE, 5780d8315ddSYiFei Zhu SECCOMP_ARCH_NATIVE_NR, SECCOMP_ARCH_NATIVE_NAME defined. If 5790d8315ddSYiFei Zhu COMPAT is supported, have the SECCOMP_ARCH_COMPAT* defines too. 580e2cfabdfSWill Drewry 581282a181bSYiFei Zhuconfig SECCOMP 582282a181bSYiFei Zhu prompt "Enable seccomp to safely execute untrusted bytecode" 583282a181bSYiFei Zhu def_bool y 584282a181bSYiFei Zhu depends on HAVE_ARCH_SECCOMP 585282a181bSYiFei Zhu help 586282a181bSYiFei Zhu This kernel feature is useful for number crunching applications 587282a181bSYiFei Zhu that may need to handle untrusted bytecode during their 588282a181bSYiFei Zhu execution. By using pipes or other transports made available 589282a181bSYiFei Zhu to the process as file descriptors supporting the read/write 590282a181bSYiFei Zhu syscalls, it's possible to isolate those applications in their 591282a181bSYiFei Zhu own address space using seccomp. Once seccomp is enabled via 592282a181bSYiFei Zhu prctl(PR_SET_SECCOMP) or the seccomp() syscall, it cannot be 593282a181bSYiFei Zhu disabled and the task is only allowed to execute a few safe 594282a181bSYiFei Zhu syscalls defined by each seccomp mode. 595282a181bSYiFei Zhu 596282a181bSYiFei Zhu If unsure, say Y. 597282a181bSYiFei Zhu 598e2cfabdfSWill Drewryconfig SECCOMP_FILTER 599e2cfabdfSWill Drewry def_bool y 600e2cfabdfSWill Drewry depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET 601e2cfabdfSWill Drewry help 602e2cfabdfSWill Drewry Enable tasks to build secure computing environments defined 603e2cfabdfSWill Drewry in terms of Berkeley Packet Filter programs which implement 604e2cfabdfSWill Drewry task-defined system call filtering polices. 605e2cfabdfSWill Drewry 6065fb94e9cSMauro Carvalho Chehab See Documentation/userspace-api/seccomp_filter.rst for details. 607e2cfabdfSWill Drewry 6080d8315ddSYiFei Zhuconfig SECCOMP_CACHE_DEBUG 6090d8315ddSYiFei Zhu bool "Show seccomp filter cache status in /proc/pid/seccomp_cache" 6100d8315ddSYiFei Zhu depends on SECCOMP_FILTER && !HAVE_SPARSE_SYSCALL_NR 6110d8315ddSYiFei Zhu depends on PROC_FS 6120d8315ddSYiFei Zhu help 6130d8315ddSYiFei Zhu This enables the /proc/pid/seccomp_cache interface to monitor 6140d8315ddSYiFei Zhu seccomp cache data. The file format is subject to change. Reading 6150d8315ddSYiFei Zhu the file requires CAP_SYS_ADMIN. 6160d8315ddSYiFei Zhu 6170d8315ddSYiFei Zhu This option is for debugging only. Enabling presents the risk that 6180d8315ddSYiFei Zhu an adversary may be able to infer the seccomp filter logic. 6190d8315ddSYiFei Zhu 6200d8315ddSYiFei Zhu If unsure, say N. 6210d8315ddSYiFei Zhu 622afaef01cSAlexander Popovconfig HAVE_ARCH_STACKLEAK 623afaef01cSAlexander Popov bool 624afaef01cSAlexander Popov help 625afaef01cSAlexander Popov An architecture should select this if it has the code which 626afaef01cSAlexander Popov fills the used part of the kernel stack with the STACKLEAK_POISON 627afaef01cSAlexander Popov value before returning from system calls. 628afaef01cSAlexander Popov 629d148eac0SMasahiro Yamadaconfig HAVE_STACKPROTECTOR 63019952a92SKees Cook bool 63119952a92SKees Cook help 63219952a92SKees Cook An arch should select this symbol if: 63319952a92SKees Cook - it has implemented a stack canary (e.g. __stack_chk_guard) 63419952a92SKees Cook 635050e9baaSLinus Torvaldsconfig STACKPROTECTOR 6362a61f474SMasahiro Yamada bool "Stack Protector buffer overflow detection" 637d148eac0SMasahiro Yamada depends on HAVE_STACKPROTECTOR 6382a61f474SMasahiro Yamada depends on $(cc-option,-fstack-protector) 6392a61f474SMasahiro Yamada default y 6408779657dSKees Cook help 6418779657dSKees Cook This option turns on the "stack-protector" GCC feature. This 64219952a92SKees Cook feature puts, at the beginning of functions, a canary value on 64319952a92SKees Cook the stack just before the return address, and validates 64419952a92SKees Cook the value just before actually returning. Stack based buffer 64519952a92SKees Cook overflows (that need to overwrite this return address) now also 64619952a92SKees Cook overwrite the canary, which gets detected and the attack is then 64719952a92SKees Cook neutralized via a kernel panic. 64819952a92SKees Cook 6498779657dSKees Cook Functions will have the stack-protector canary logic added if they 6508779657dSKees Cook have an 8-byte or larger character array on the stack. 6518779657dSKees Cook 65219952a92SKees Cook This feature requires gcc version 4.2 or above, or a distribution 6538779657dSKees Cook gcc with the feature backported ("-fstack-protector"). 6548779657dSKees Cook 6558779657dSKees Cook On an x86 "defconfig" build, this feature adds canary checks to 6568779657dSKees Cook about 3% of all kernel functions, which increases kernel code size 6578779657dSKees Cook by about 0.3%. 6588779657dSKees Cook 659050e9baaSLinus Torvaldsconfig STACKPROTECTOR_STRONG 6602a61f474SMasahiro Yamada bool "Strong Stack Protector" 661050e9baaSLinus Torvalds depends on STACKPROTECTOR 6622a61f474SMasahiro Yamada depends on $(cc-option,-fstack-protector-strong) 6632a61f474SMasahiro Yamada default y 6648779657dSKees Cook help 6658779657dSKees Cook Functions will have the stack-protector canary logic added in any 6668779657dSKees Cook of the following conditions: 6678779657dSKees Cook 6688779657dSKees Cook - local variable's address used as part of the right hand side of an 6698779657dSKees Cook assignment or function argument 6708779657dSKees Cook - local variable is an array (or union containing an array), 6718779657dSKees Cook regardless of array type or length 6728779657dSKees Cook - uses register local variables 6738779657dSKees Cook 6748779657dSKees Cook This feature requires gcc version 4.9 or above, or a distribution 6758779657dSKees Cook gcc with the feature backported ("-fstack-protector-strong"). 6768779657dSKees Cook 6778779657dSKees Cook On an x86 "defconfig" build, this feature adds canary checks to 6788779657dSKees Cook about 20% of all kernel functions, which increases the kernel code 6798779657dSKees Cook size by about 2%. 6808779657dSKees Cook 681d08b9f0cSSami Tolvanenconfig ARCH_SUPPORTS_SHADOW_CALL_STACK 682d08b9f0cSSami Tolvanen bool 683d08b9f0cSSami Tolvanen help 684afcf5441SDan Li An architecture should select this if it supports the compiler's 685afcf5441SDan Li Shadow Call Stack and implements runtime support for shadow stack 686aa7a65aeSWill Deacon switching. 687d08b9f0cSSami Tolvanen 688d08b9f0cSSami Tolvanenconfig SHADOW_CALL_STACK 689afcf5441SDan Li bool "Shadow Call Stack" 690afcf5441SDan Li depends on ARCH_SUPPORTS_SHADOW_CALL_STACK 69138792972SArd Biesheuvel depends on DYNAMIC_FTRACE_WITH_ARGS || DYNAMIC_FTRACE_WITH_REGS || !FUNCTION_GRAPH_TRACER 6928b5a4fe3SSamuel Holland depends on MMU 693d08b9f0cSSami Tolvanen help 694afcf5441SDan Li This option enables the compiler's Shadow Call Stack, which 695afcf5441SDan Li uses a shadow stack to protect function return addresses from 696afcf5441SDan Li being overwritten by an attacker. More information can be found 697afcf5441SDan Li in the compiler's documentation: 698d08b9f0cSSami Tolvanen 699afcf5441SDan Li - Clang: https://clang.llvm.org/docs/ShadowCallStack.html 700afcf5441SDan Li - GCC: https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html#Instrumentation-Options 701d08b9f0cSSami Tolvanen 702d08b9f0cSSami Tolvanen Note that security guarantees in the kernel differ from the 703d08b9f0cSSami Tolvanen ones documented for user space. The kernel must store addresses 704d08b9f0cSSami Tolvanen of shadow stacks in memory, which means an attacker capable of 705d08b9f0cSSami Tolvanen reading and writing arbitrary memory may be able to locate them 706d08b9f0cSSami Tolvanen and hijack control flow by modifying the stacks. 707d08b9f0cSSami Tolvanen 7089beccca0SArd Biesheuvelconfig DYNAMIC_SCS 7099beccca0SArd Biesheuvel bool 7109beccca0SArd Biesheuvel help 7119beccca0SArd Biesheuvel Set by the arch code if it relies on code patching to insert the 7129beccca0SArd Biesheuvel shadow call stack push and pop instructions rather than on the 7139beccca0SArd Biesheuvel compiler. 7149beccca0SArd Biesheuvel 715dc5723b0SSami Tolvanenconfig LTO 716dc5723b0SSami Tolvanen bool 717dc5723b0SSami Tolvanen help 718dc5723b0SSami Tolvanen Selected if the kernel will be built using the compiler's LTO feature. 719dc5723b0SSami Tolvanen 720dc5723b0SSami Tolvanenconfig LTO_CLANG 721dc5723b0SSami Tolvanen bool 722dc5723b0SSami Tolvanen select LTO 723dc5723b0SSami Tolvanen help 724dc5723b0SSami Tolvanen Selected if the kernel will be built using Clang's LTO feature. 725dc5723b0SSami Tolvanen 726dc5723b0SSami Tolvanenconfig ARCH_SUPPORTS_LTO_CLANG 727dc5723b0SSami Tolvanen bool 728dc5723b0SSami Tolvanen help 729dc5723b0SSami Tolvanen An architecture should select this option if it supports: 730dc5723b0SSami Tolvanen - compiling with Clang, 731dc5723b0SSami Tolvanen - compiling inline assembly with Clang's integrated assembler, 732dc5723b0SSami Tolvanen - and linking with LLD. 733dc5723b0SSami Tolvanen 734dc5723b0SSami Tolvanenconfig ARCH_SUPPORTS_LTO_CLANG_THIN 735dc5723b0SSami Tolvanen bool 736dc5723b0SSami Tolvanen help 737dc5723b0SSami Tolvanen An architecture should select this option if it can support Clang's 738dc5723b0SSami Tolvanen ThinLTO mode. 739dc5723b0SSami Tolvanen 740dc5723b0SSami Tolvanenconfig HAS_LTO_CLANG 741dc5723b0SSami Tolvanen def_bool y 7421e68a8afSNathan Chancellor depends on CC_IS_CLANG && LD_IS_LLD && AS_IS_LLVM 743dc5723b0SSami Tolvanen depends on $(success,$(NM) --help | head -n 1 | grep -qi llvm) 744dc5723b0SSami Tolvanen depends on $(success,$(AR) --help | head -n 1 | grep -qi llvm) 745dc5723b0SSami Tolvanen depends on ARCH_SUPPORTS_LTO_CLANG 746dc5723b0SSami Tolvanen depends on !FTRACE_MCOUNT_USE_RECORDMCOUNT 747349fde59SJakob Koschel # https://github.com/ClangBuiltLinux/linux/issues/1721 748349fde59SJakob Koschel depends on (!KASAN || KASAN_HW_TAGS || CLANG_VERSION >= 170000) || !DEBUG_INFO 749349fde59SJakob Koschel depends on (!KCOV || CLANG_VERSION >= 170000) || !DEBUG_INFO 750dc5723b0SSami Tolvanen depends on !GCOV_KERNEL 751dc5723b0SSami Tolvanen help 752dc5723b0SSami Tolvanen The compiler and Kconfig options support building with Clang's 753dc5723b0SSami Tolvanen LTO. 754dc5723b0SSami Tolvanen 755dc5723b0SSami Tolvanenchoice 756dc5723b0SSami Tolvanen prompt "Link Time Optimization (LTO)" 757dc5723b0SSami Tolvanen default LTO_NONE 758dc5723b0SSami Tolvanen help 759dc5723b0SSami Tolvanen This option enables Link Time Optimization (LTO), which allows the 760dc5723b0SSami Tolvanen compiler to optimize binaries globally. 761dc5723b0SSami Tolvanen 762dc5723b0SSami Tolvanen If unsure, select LTO_NONE. Note that LTO is very resource-intensive 763dc5723b0SSami Tolvanen so it's disabled by default. 764dc5723b0SSami Tolvanen 765dc5723b0SSami Tolvanenconfig LTO_NONE 766dc5723b0SSami Tolvanen bool "None" 767dc5723b0SSami Tolvanen help 768dc5723b0SSami Tolvanen Build the kernel normally, without Link Time Optimization (LTO). 769dc5723b0SSami Tolvanen 770dc5723b0SSami Tolvanenconfig LTO_CLANG_FULL 771dc5723b0SSami Tolvanen bool "Clang Full LTO (EXPERIMENTAL)" 772dc5723b0SSami Tolvanen depends on HAS_LTO_CLANG 773dc5723b0SSami Tolvanen depends on !COMPILE_TEST 774dc5723b0SSami Tolvanen select LTO_CLANG 775dc5723b0SSami Tolvanen help 776dc5723b0SSami Tolvanen This option enables Clang's full Link Time Optimization (LTO), which 777dc5723b0SSami Tolvanen allows the compiler to optimize the kernel globally. If you enable 778dc5723b0SSami Tolvanen this option, the compiler generates LLVM bitcode instead of ELF 779dc5723b0SSami Tolvanen object files, and the actual compilation from bitcode happens at 780dc5723b0SSami Tolvanen the LTO link step, which may take several minutes depending on the 781dc5723b0SSami Tolvanen kernel configuration. More information can be found from LLVM's 782dc5723b0SSami Tolvanen documentation: 783dc5723b0SSami Tolvanen 784dc5723b0SSami Tolvanen https://llvm.org/docs/LinkTimeOptimization.html 785dc5723b0SSami Tolvanen 786dc5723b0SSami Tolvanen During link time, this option can use a large amount of RAM, and 787dc5723b0SSami Tolvanen may take much longer than the ThinLTO option. 788dc5723b0SSami Tolvanen 789dc5723b0SSami Tolvanenconfig LTO_CLANG_THIN 790dc5723b0SSami Tolvanen bool "Clang ThinLTO (EXPERIMENTAL)" 791dc5723b0SSami Tolvanen depends on HAS_LTO_CLANG && ARCH_SUPPORTS_LTO_CLANG_THIN 792dc5723b0SSami Tolvanen select LTO_CLANG 793dc5723b0SSami Tolvanen help 794dc5723b0SSami Tolvanen This option enables Clang's ThinLTO, which allows for parallel 795dc5723b0SSami Tolvanen optimization and faster incremental compiles compared to the 796dc5723b0SSami Tolvanen CONFIG_LTO_CLANG_FULL option. More information can be found 797dc5723b0SSami Tolvanen from Clang's documentation: 798dc5723b0SSami Tolvanen 799dc5723b0SSami Tolvanen https://clang.llvm.org/docs/ThinLTO.html 800dc5723b0SSami Tolvanen 801dc5723b0SSami Tolvanen If unsure, say Y. 802dc5723b0SSami Tolvanenendchoice 803dc5723b0SSami Tolvanen 804cf68fffbSSami Tolvanenconfig ARCH_SUPPORTS_CFI_CLANG 805cf68fffbSSami Tolvanen bool 806cf68fffbSSami Tolvanen help 807cf68fffbSSami Tolvanen An architecture should select this option if it can support Clang's 808cf68fffbSSami Tolvanen Control-Flow Integrity (CFI) checking. 809cf68fffbSSami Tolvanen 81089245600SSami Tolvanenconfig ARCH_USES_CFI_TRAPS 81189245600SSami Tolvanen bool 81289245600SSami Tolvanen 813cf68fffbSSami Tolvanenconfig CFI_CLANG 814cf68fffbSSami Tolvanen bool "Use Clang's Control Flow Integrity (CFI)" 81589245600SSami Tolvanen depends on ARCH_SUPPORTS_CFI_CLANG 81689245600SSami Tolvanen depends on $(cc-option,-fsanitize=kcfi) 817cf68fffbSSami Tolvanen help 818cf68fffbSSami Tolvanen This option enables Clang’s forward-edge Control Flow Integrity 819cf68fffbSSami Tolvanen (CFI) checking, where the compiler injects a runtime check to each 820cf68fffbSSami Tolvanen indirect function call to ensure the target is a valid function with 821cf68fffbSSami Tolvanen the correct static type. This restricts possible call targets and 822cf68fffbSSami Tolvanen makes it more difficult for an attacker to exploit bugs that allow 823cf68fffbSSami Tolvanen the modification of stored function pointers. More information can be 824cf68fffbSSami Tolvanen found from Clang's documentation: 825cf68fffbSSami Tolvanen 826cf68fffbSSami Tolvanen https://clang.llvm.org/docs/ControlFlowIntegrity.html 827cf68fffbSSami Tolvanen 828cf68fffbSSami Tolvanenconfig CFI_PERMISSIVE 829cf68fffbSSami Tolvanen bool "Use CFI in permissive mode" 830cf68fffbSSami Tolvanen depends on CFI_CLANG 831cf68fffbSSami Tolvanen help 832cf68fffbSSami Tolvanen When selected, Control Flow Integrity (CFI) violations result in a 833cf68fffbSSami Tolvanen warning instead of a kernel panic. This option should only be used 834cf68fffbSSami Tolvanen for finding indirect call type mismatches during development. 835cf68fffbSSami Tolvanen 836cf68fffbSSami Tolvanen If unsure, say N. 837cf68fffbSSami Tolvanen 8380f60a8efSKees Cookconfig HAVE_ARCH_WITHIN_STACK_FRAMES 8390f60a8efSKees Cook bool 8400f60a8efSKees Cook help 8410f60a8efSKees Cook An architecture should select this if it can walk the kernel stack 8420f60a8efSKees Cook frames to determine if an object is part of either the arguments 8430f60a8efSKees Cook or local variables (i.e. that it excludes saved return addresses, 8440f60a8efSKees Cook and similar) by implementing an inline arch_within_stack_frames(), 8450f60a8efSKees Cook which is used by CONFIG_HARDENED_USERCOPY. 8460f60a8efSKees Cook 84724a9c541SFrederic Weisbeckerconfig HAVE_CONTEXT_TRACKING_USER 8482b1d5024SFrederic Weisbecker bool 8492b1d5024SFrederic Weisbecker help 85091d1aa43SFrederic Weisbecker Provide kernel/user boundaries probes necessary for subsystems 85191d1aa43SFrederic Weisbecker that need it, such as userspace RCU extended quiescent state. 852490f561bSFrederic Weisbecker Syscalls need to be wrapped inside user_exit()-user_enter(), either 853490f561bSFrederic Weisbecker optimized behind static key or through the slow path using TIF_NOHZ 854490f561bSFrederic Weisbecker flag. Exceptions handlers must be wrapped as well. Irqs are already 8556f0e6c15SFrederic Weisbecker protected inside ct_irq_enter/ct_irq_exit() but preemption or signal 856490f561bSFrederic Weisbecker handling on irq exit still need to be protected. 857490f561bSFrederic Weisbecker 85824a9c541SFrederic Weisbeckerconfig HAVE_CONTEXT_TRACKING_USER_OFFSTACK 85983c2da2eSFrederic Weisbecker bool 86083c2da2eSFrederic Weisbecker help 86183c2da2eSFrederic Weisbecker Architecture neither relies on exception_enter()/exception_exit() 86283c2da2eSFrederic Weisbecker nor on schedule_user(). Also preempt_schedule_notrace() and 86383c2da2eSFrederic Weisbecker preempt_schedule_irq() can't be called in a preemptible section 86483c2da2eSFrederic Weisbecker while context tracking is CONTEXT_USER. This feature reflects a sane 86583c2da2eSFrederic Weisbecker entry implementation where the following requirements are met on 86683c2da2eSFrederic Weisbecker critical entry code, ie: before user_exit() or after user_enter(): 86783c2da2eSFrederic Weisbecker 86883c2da2eSFrederic Weisbecker - Critical entry code isn't preemptible (or better yet: 86983c2da2eSFrederic Weisbecker not interruptible). 870493c1822SFrederic Weisbecker - No use of RCU read side critical sections, unless ct_nmi_enter() 87183c2da2eSFrederic Weisbecker got called. 87283c2da2eSFrederic Weisbecker - No use of instrumentation, unless instrumentation_begin() got 87383c2da2eSFrederic Weisbecker called. 87483c2da2eSFrederic Weisbecker 875490f561bSFrederic Weisbeckerconfig HAVE_TIF_NOHZ 876490f561bSFrederic Weisbecker bool 877490f561bSFrederic Weisbecker help 878490f561bSFrederic Weisbecker Arch relies on TIF_NOHZ and syscall slow path to implement context 879490f561bSFrederic Weisbecker tracking calls to user_enter()/user_exit(). 8802b1d5024SFrederic Weisbecker 881b952741cSFrederic Weisbeckerconfig HAVE_VIRT_CPU_ACCOUNTING 882b952741cSFrederic Weisbecker bool 883b952741cSFrederic Weisbecker 8842b91ec9fSFrederic Weisbeckerconfig HAVE_VIRT_CPU_ACCOUNTING_IDLE 8852b91ec9fSFrederic Weisbecker bool 8862b91ec9fSFrederic Weisbecker help 8872b91ec9fSFrederic Weisbecker Architecture has its own way to account idle CPU time and therefore 8882b91ec9fSFrederic Weisbecker doesn't implement vtime_account_idle(). 8892b91ec9fSFrederic Weisbecker 89040565b5aSStanislaw Gruszkaconfig ARCH_HAS_SCALED_CPUTIME 89140565b5aSStanislaw Gruszka bool 89240565b5aSStanislaw Gruszka 893554b0004SKevin Hilmanconfig HAVE_VIRT_CPU_ACCOUNTING_GEN 894554b0004SKevin Hilman bool 895554b0004SKevin Hilman default y if 64BIT 896554b0004SKevin Hilman help 897554b0004SKevin Hilman With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit. 898554b0004SKevin Hilman Before enabling this option, arch code must be audited 899554b0004SKevin Hilman to ensure there are no races in concurrent read/write of 900554b0004SKevin Hilman cputime_t. For example, reading/writing 64-bit cputime_t on 901554b0004SKevin Hilman some 32-bit arches may require multiple accesses, so proper 902554b0004SKevin Hilman locking is needed to protect against concurrent accesses. 903554b0004SKevin Hilman 904fdf9c356SFrederic Weisbeckerconfig HAVE_IRQ_TIME_ACCOUNTING 905fdf9c356SFrederic Weisbecker bool 906fdf9c356SFrederic Weisbecker help 907fdf9c356SFrederic Weisbecker Archs need to ensure they use a high enough resolution clock to 908fdf9c356SFrederic Weisbecker support irq time accounting and then call enable_sched_clock_irqtime(). 909fdf9c356SFrederic Weisbecker 910c49dd340SKalesh Singhconfig HAVE_MOVE_PUD 911c49dd340SKalesh Singh bool 912c49dd340SKalesh Singh help 913c49dd340SKalesh Singh Architectures that select this are able to move page tables at the 914c49dd340SKalesh Singh PUD level. If there are only 3 page table levels, the move effectively 915c49dd340SKalesh Singh happens at the PGD level. 916c49dd340SKalesh Singh 9172c91bd4aSJoel Fernandes (Google)config HAVE_MOVE_PMD 9182c91bd4aSJoel Fernandes (Google) bool 9192c91bd4aSJoel Fernandes (Google) help 9202c91bd4aSJoel Fernandes (Google) Archs that select this are able to move page tables at the PMD level. 9212c91bd4aSJoel Fernandes (Google) 92215626062SGerald Schaeferconfig HAVE_ARCH_TRANSPARENT_HUGEPAGE 92315626062SGerald Schaefer bool 92415626062SGerald Schaefer 925a00cc7d9SMatthew Wilcoxconfig HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD 926a00cc7d9SMatthew Wilcox bool 927a00cc7d9SMatthew Wilcox 9280ddab1d2SToshi Kaniconfig HAVE_ARCH_HUGE_VMAP 9290ddab1d2SToshi Kani bool 9300ddab1d2SToshi Kani 931121e6f32SNicholas Piggin# 932121e6f32SNicholas Piggin# Archs that select this would be capable of PMD-sized vmaps (i.e., 933559089e0SSong Liu# arch_vmap_pmd_supported() returns true). The VM_ALLOW_HUGE_VMAP flag 934559089e0SSong Liu# must be used to enable allocations to use hugepages. 935121e6f32SNicholas Piggin# 936121e6f32SNicholas Pigginconfig HAVE_ARCH_HUGE_VMALLOC 937121e6f32SNicholas Piggin depends on HAVE_ARCH_HUGE_VMAP 938121e6f32SNicholas Piggin bool 939121e6f32SNicholas Piggin 9403876d4a3SAlexandre Ghiticonfig ARCH_WANT_HUGE_PMD_SHARE 9413876d4a3SAlexandre Ghiti bool 9423876d4a3SAlexandre Ghiti 9432f0584f3SRick Edgecombe# Archs that want to use pmd_mkwrite on kernel memory need it defined even 9442f0584f3SRick Edgecombe# if there are no userspace memory management features that use it 9452f0584f3SRick Edgecombeconfig ARCH_WANT_KERNEL_PMD_MKWRITE 9462f0584f3SRick Edgecombe bool 9472f0584f3SRick Edgecombe 9482f0584f3SRick Edgecombeconfig ARCH_WANT_PMD_MKWRITE 9492f0584f3SRick Edgecombe def_bool TRANSPARENT_HUGEPAGE || ARCH_WANT_KERNEL_PMD_MKWRITE 9502f0584f3SRick Edgecombe 9510f8975ecSPavel Emelyanovconfig HAVE_ARCH_SOFT_DIRTY 9520f8975ecSPavel Emelyanov bool 9530f8975ecSPavel Emelyanov 954786d35d4SDavid Howellsconfig HAVE_MOD_ARCH_SPECIFIC 955786d35d4SDavid Howells bool 956786d35d4SDavid Howells help 957786d35d4SDavid Howells The arch uses struct mod_arch_specific to store data. Many arches 958786d35d4SDavid Howells just need a simple module loader without arch specific data - those 959786d35d4SDavid Howells should not enable this. 960786d35d4SDavid Howells 961786d35d4SDavid Howellsconfig MODULES_USE_ELF_RELA 962786d35d4SDavid Howells bool 963786d35d4SDavid Howells help 964786d35d4SDavid Howells Modules only use ELF RELA relocations. Modules with ELF REL 965786d35d4SDavid Howells relocations will give an error. 966786d35d4SDavid Howells 967786d35d4SDavid Howellsconfig MODULES_USE_ELF_REL 968786d35d4SDavid Howells bool 969786d35d4SDavid Howells help 970786d35d4SDavid Howells Modules only use ELF REL relocations. Modules with ELF RELA 971786d35d4SDavid Howells relocations will give an error. 972786d35d4SDavid Howells 97301dc0386SChristophe Leroyconfig ARCH_WANTS_MODULES_DATA_IN_VMALLOC 97401dc0386SChristophe Leroy bool 97501dc0386SChristophe Leroy help 97601dc0386SChristophe Leroy For architectures like powerpc/32 which have constraints on module 97701dc0386SChristophe Leroy allocation and need to allocate module data outside of module area. 97801dc0386SChristophe Leroy 979cc1f0274SFrederic Weisbeckerconfig HAVE_IRQ_EXIT_ON_IRQ_STACK 980cc1f0274SFrederic Weisbecker bool 981cc1f0274SFrederic Weisbecker help 982cc1f0274SFrederic Weisbecker Architecture doesn't only execute the irq handler on the irq stack 983cc1f0274SFrederic Weisbecker but also irq_exit(). This way we can process softirqs on this irq 984cc1f0274SFrederic Weisbecker stack instead of switching to a new one when we call __do_softirq() 985cc1f0274SFrederic Weisbecker in the end of an hardirq. 986cc1f0274SFrederic Weisbecker This spares a stack switch and improves cache usage on softirq 987cc1f0274SFrederic Weisbecker processing. 988cc1f0274SFrederic Weisbecker 989cd1a41ceSThomas Gleixnerconfig HAVE_SOFTIRQ_ON_OWN_STACK 990cd1a41ceSThomas Gleixner bool 991cd1a41ceSThomas Gleixner help 992cd1a41ceSThomas Gleixner Architecture provides a function to run __do_softirq() on a 993c226bc3cSColin Ian King separate stack. 994cd1a41ceSThomas Gleixner 9958cbb2b50SSebastian Andrzej Siewiorconfig SOFTIRQ_ON_OWN_STACK 9968cbb2b50SSebastian Andrzej Siewior def_bool HAVE_SOFTIRQ_ON_OWN_STACK && !PREEMPT_RT 9978cbb2b50SSebastian Andrzej Siewior 99812700c17SArnd Bergmannconfig ALTERNATE_USER_ADDRESS_SPACE 99912700c17SArnd Bergmann bool 100012700c17SArnd Bergmann help 100112700c17SArnd Bergmann Architectures set this when the CPU uses separate address 100212700c17SArnd Bergmann spaces for kernel and user space pointers. In this case, the 100312700c17SArnd Bergmann access_ok() check on a __user pointer is skipped. 100412700c17SArnd Bergmann 1005235a8f02SKirill A. Shutemovconfig PGTABLE_LEVELS 1006235a8f02SKirill A. Shutemov int 1007235a8f02SKirill A. Shutemov default 2 1008235a8f02SKirill A. Shutemov 10092b68f6caSKees Cookconfig ARCH_HAS_ELF_RANDOMIZE 10102b68f6caSKees Cook bool 10112b68f6caSKees Cook help 10122b68f6caSKees Cook An architecture supports choosing randomized locations for 10132b68f6caSKees Cook stack, mmap, brk, and ET_DYN. Defined functions: 10142b68f6caSKees Cook - arch_mmap_rnd() 1015204db6edSKees Cook - arch_randomize_brk() 10162b68f6caSKees Cook 1017d07e2259SDaniel Cashmanconfig HAVE_ARCH_MMAP_RND_BITS 1018d07e2259SDaniel Cashman bool 1019d07e2259SDaniel Cashman help 1020d07e2259SDaniel Cashman An arch should select this symbol if it supports setting a variable 1021d07e2259SDaniel Cashman number of bits for use in establishing the base address for mmap 1022d07e2259SDaniel Cashman allocations, has MMU enabled and provides values for both: 1023d07e2259SDaniel Cashman - ARCH_MMAP_RND_BITS_MIN 1024d07e2259SDaniel Cashman - ARCH_MMAP_RND_BITS_MAX 1025d07e2259SDaniel Cashman 10265f56a5dfSJiri Slabyconfig HAVE_EXIT_THREAD 10275f56a5dfSJiri Slaby bool 10285f56a5dfSJiri Slaby help 10295f56a5dfSJiri Slaby An architecture implements exit_thread. 10305f56a5dfSJiri Slaby 1031d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_BITS_MIN 1032d07e2259SDaniel Cashman int 1033d07e2259SDaniel Cashman 1034d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_BITS_MAX 1035d07e2259SDaniel Cashman int 1036d07e2259SDaniel Cashman 1037d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_BITS_DEFAULT 1038d07e2259SDaniel Cashman int 1039d07e2259SDaniel Cashman 1040d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_BITS 1041d07e2259SDaniel Cashman int "Number of bits to use for ASLR of mmap base address" if EXPERT 1042d07e2259SDaniel Cashman range ARCH_MMAP_RND_BITS_MIN ARCH_MMAP_RND_BITS_MAX 1043d07e2259SDaniel Cashman default ARCH_MMAP_RND_BITS_DEFAULT if ARCH_MMAP_RND_BITS_DEFAULT 1044d07e2259SDaniel Cashman default ARCH_MMAP_RND_BITS_MIN 1045d07e2259SDaniel Cashman depends on HAVE_ARCH_MMAP_RND_BITS 1046d07e2259SDaniel Cashman help 1047d07e2259SDaniel Cashman This value can be used to select the number of bits to use to 1048d07e2259SDaniel Cashman determine the random offset to the base address of vma regions 1049d07e2259SDaniel Cashman resulting from mmap allocations. This value will be bounded 1050d07e2259SDaniel Cashman by the architecture's minimum and maximum supported values. 1051d07e2259SDaniel Cashman 1052d07e2259SDaniel Cashman This value can be changed after boot using the 1053d07e2259SDaniel Cashman /proc/sys/vm/mmap_rnd_bits tunable 1054d07e2259SDaniel Cashman 1055d07e2259SDaniel Cashmanconfig HAVE_ARCH_MMAP_RND_COMPAT_BITS 1056d07e2259SDaniel Cashman bool 1057d07e2259SDaniel Cashman help 1058d07e2259SDaniel Cashman An arch should select this symbol if it supports running applications 1059d07e2259SDaniel Cashman in compatibility mode, supports setting a variable number of bits for 1060d07e2259SDaniel Cashman use in establishing the base address for mmap allocations, has MMU 1061d07e2259SDaniel Cashman enabled and provides values for both: 1062d07e2259SDaniel Cashman - ARCH_MMAP_RND_COMPAT_BITS_MIN 1063d07e2259SDaniel Cashman - ARCH_MMAP_RND_COMPAT_BITS_MAX 1064d07e2259SDaniel Cashman 1065d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_COMPAT_BITS_MIN 1066d07e2259SDaniel Cashman int 1067d07e2259SDaniel Cashman 1068d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_COMPAT_BITS_MAX 1069d07e2259SDaniel Cashman int 1070d07e2259SDaniel Cashman 1071d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_COMPAT_BITS_DEFAULT 1072d07e2259SDaniel Cashman int 1073d07e2259SDaniel Cashman 1074d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_COMPAT_BITS 1075d07e2259SDaniel Cashman int "Number of bits to use for ASLR of mmap base address for compatible applications" if EXPERT 1076d07e2259SDaniel Cashman range ARCH_MMAP_RND_COMPAT_BITS_MIN ARCH_MMAP_RND_COMPAT_BITS_MAX 1077d07e2259SDaniel Cashman default ARCH_MMAP_RND_COMPAT_BITS_DEFAULT if ARCH_MMAP_RND_COMPAT_BITS_DEFAULT 1078d07e2259SDaniel Cashman default ARCH_MMAP_RND_COMPAT_BITS_MIN 1079d07e2259SDaniel Cashman depends on HAVE_ARCH_MMAP_RND_COMPAT_BITS 1080d07e2259SDaniel Cashman help 1081d07e2259SDaniel Cashman This value can be used to select the number of bits to use to 1082d07e2259SDaniel Cashman determine the random offset to the base address of vma regions 1083d07e2259SDaniel Cashman resulting from mmap allocations for compatible applications This 1084d07e2259SDaniel Cashman value will be bounded by the architecture's minimum and maximum 1085d07e2259SDaniel Cashman supported values. 1086d07e2259SDaniel Cashman 1087d07e2259SDaniel Cashman This value can be changed after boot using the 1088d07e2259SDaniel Cashman /proc/sys/vm/mmap_rnd_compat_bits tunable 1089d07e2259SDaniel Cashman 10901b028f78SDmitry Safonovconfig HAVE_ARCH_COMPAT_MMAP_BASES 10911b028f78SDmitry Safonov bool 10921b028f78SDmitry Safonov help 10931b028f78SDmitry Safonov This allows 64bit applications to invoke 32-bit mmap() syscall 10941b028f78SDmitry Safonov and vice-versa 32-bit applications to call 64-bit mmap(). 10951b028f78SDmitry Safonov Required for applications doing different bitness syscalls. 10961b028f78SDmitry Safonov 10971f0e290cSGuenter Roeckconfig PAGE_SIZE_LESS_THAN_64KB 10981f0e290cSGuenter Roeck def_bool y 10991f0e290cSGuenter Roeck depends on !ARM64_64K_PAGES 11001f0e290cSGuenter Roeck depends on !IA64_PAGE_SIZE_64KB 11011f0e290cSGuenter Roeck depends on !PAGE_SIZE_64KB 11021f0e290cSGuenter Roeck depends on !PARISC_PAGE_SIZE_64KB 1103e4bbd20dSNathan Chancellor depends on PAGE_SIZE_LESS_THAN_256KB 1104e4bbd20dSNathan Chancellor 1105e4bbd20dSNathan Chancellorconfig PAGE_SIZE_LESS_THAN_256KB 1106e4bbd20dSNathan Chancellor def_bool y 11071f0e290cSGuenter Roeck depends on !PAGE_SIZE_256KB 11081f0e290cSGuenter Roeck 110967f3977fSAlexandre Ghiti# This allows to use a set of generic functions to determine mmap base 111067f3977fSAlexandre Ghiti# address by giving priority to top-down scheme only if the process 111167f3977fSAlexandre Ghiti# is not in legacy mode (compat task, unlimited stack size or 111267f3977fSAlexandre Ghiti# sysctl_legacy_va_layout). 111367f3977fSAlexandre Ghiti# Architecture that selects this option can provide its own version of: 111467f3977fSAlexandre Ghiti# - STACK_RND_MASK 111567f3977fSAlexandre Ghiticonfig ARCH_WANT_DEFAULT_TOPDOWN_MMAP_LAYOUT 111667f3977fSAlexandre Ghiti bool 111767f3977fSAlexandre Ghiti depends on MMU 1118e7142bf5SAlexandre Ghiti select ARCH_HAS_ELF_RANDOMIZE 111967f3977fSAlexandre Ghiti 112003f16cd0SJosh Poimboeufconfig HAVE_OBJTOOL 112103f16cd0SJosh Poimboeuf bool 112203f16cd0SJosh Poimboeuf 11234ab7674fSJosh Poimboeufconfig HAVE_JUMP_LABEL_HACK 11244ab7674fSJosh Poimboeuf bool 11254ab7674fSJosh Poimboeuf 112622102f45SJosh Poimboeufconfig HAVE_NOINSTR_HACK 112722102f45SJosh Poimboeuf bool 112822102f45SJosh Poimboeuf 1129489e355bSJosh Poimboeufconfig HAVE_NOINSTR_VALIDATION 1130489e355bSJosh Poimboeuf bool 1131489e355bSJosh Poimboeuf 11325f3da8c0SJosh Poimboeufconfig HAVE_UACCESS_VALIDATION 11335f3da8c0SJosh Poimboeuf bool 11345f3da8c0SJosh Poimboeuf select OBJTOOL 11355f3da8c0SJosh Poimboeuf 1136b9ab5ebbSJosh Poimboeufconfig HAVE_STACK_VALIDATION 1137b9ab5ebbSJosh Poimboeuf bool 1138b9ab5ebbSJosh Poimboeuf help 113903f16cd0SJosh Poimboeuf Architecture supports objtool compile-time frame pointer rule 114003f16cd0SJosh Poimboeuf validation. 1141b9ab5ebbSJosh Poimboeuf 1142af085d90SJosh Poimboeufconfig HAVE_RELIABLE_STACKTRACE 1143af085d90SJosh Poimboeuf bool 1144af085d90SJosh Poimboeuf help 1145140d7e88SMiroslav Benes Architecture has either save_stack_trace_tsk_reliable() or 1146140d7e88SMiroslav Benes arch_stack_walk_reliable() function which only returns a stack trace 1147140d7e88SMiroslav Benes if it can guarantee the trace is reliable. 1148af085d90SJosh Poimboeuf 1149468a9428SGeorge Spelvinconfig HAVE_ARCH_HASH 1150468a9428SGeorge Spelvin bool 1151468a9428SGeorge Spelvin default n 1152468a9428SGeorge Spelvin help 1153468a9428SGeorge Spelvin If this is set, the architecture provides an <asm/hash.h> 1154468a9428SGeorge Spelvin file which provides platform-specific implementations of some 1155468a9428SGeorge Spelvin functions in <linux/hash.h> or fs/namei.c. 1156468a9428SGeorge Spelvin 1157666047feSFinn Thainconfig HAVE_ARCH_NVRAM_OPS 1158666047feSFinn Thain bool 1159666047feSFinn Thain 11603a495511SWilliam Breathitt Grayconfig ISA_BUS_API 11613a495511SWilliam Breathitt Gray def_bool ISA 11623a495511SWilliam Breathitt Gray 1163d2125043SAl Viro# 1164d2125043SAl Viro# ABI hall of shame 1165d2125043SAl Viro# 1166d2125043SAl Viroconfig CLONE_BACKWARDS 1167d2125043SAl Viro bool 1168d2125043SAl Viro help 1169d2125043SAl Viro Architecture has tls passed as the 4th argument of clone(2), 1170d2125043SAl Viro not the 5th one. 1171d2125043SAl Viro 1172d2125043SAl Viroconfig CLONE_BACKWARDS2 1173d2125043SAl Viro bool 1174d2125043SAl Viro help 1175d2125043SAl Viro Architecture has the first two arguments of clone(2) swapped. 1176d2125043SAl Viro 1177dfa9771aSMichal Simekconfig CLONE_BACKWARDS3 1178dfa9771aSMichal Simek bool 1179dfa9771aSMichal Simek help 1180dfa9771aSMichal Simek Architecture has tls passed as the 3rd argument of clone(2), 1181dfa9771aSMichal Simek not the 5th one. 1182dfa9771aSMichal Simek 1183eaca6eaeSAl Viroconfig ODD_RT_SIGACTION 1184eaca6eaeSAl Viro bool 1185eaca6eaeSAl Viro help 1186eaca6eaeSAl Viro Architecture has unusual rt_sigaction(2) arguments 1187eaca6eaeSAl Viro 11880a0e8cdfSAl Viroconfig OLD_SIGSUSPEND 11890a0e8cdfSAl Viro bool 11900a0e8cdfSAl Viro help 11910a0e8cdfSAl Viro Architecture has old sigsuspend(2) syscall, of one-argument variety 11920a0e8cdfSAl Viro 11930a0e8cdfSAl Viroconfig OLD_SIGSUSPEND3 11940a0e8cdfSAl Viro bool 11950a0e8cdfSAl Viro help 11960a0e8cdfSAl Viro Even weirder antique ABI - three-argument sigsuspend(2) 11970a0e8cdfSAl Viro 1198495dfbf7SAl Viroconfig OLD_SIGACTION 1199495dfbf7SAl Viro bool 1200495dfbf7SAl Viro help 1201495dfbf7SAl Viro Architecture has old sigaction(2) syscall. Nope, not the same 1202495dfbf7SAl Viro as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2), 1203495dfbf7SAl Viro but fairly different variant of sigaction(2), thanks to OSF/1 1204495dfbf7SAl Viro compatibility... 1205495dfbf7SAl Viro 1206495dfbf7SAl Viroconfig COMPAT_OLD_SIGACTION 1207495dfbf7SAl Viro bool 1208495dfbf7SAl Viro 120917435e5fSDeepa Dinamaniconfig COMPAT_32BIT_TIME 1210942437c9SArnd Bergmann bool "Provide system calls for 32-bit time_t" 1211942437c9SArnd Bergmann default !64BIT || COMPAT 121217435e5fSDeepa Dinamani help 121317435e5fSDeepa Dinamani This enables 32 bit time_t support in addition to 64 bit time_t support. 121417435e5fSDeepa Dinamani This is relevant on all 32-bit architectures, and 64-bit architectures 121517435e5fSDeepa Dinamani as part of compat syscall handling. 121617435e5fSDeepa Dinamani 121787a4c375SChristoph Hellwigconfig ARCH_NO_PREEMPT 121887a4c375SChristoph Hellwig bool 121987a4c375SChristoph Hellwig 1220a50a3f4bSThomas Gleixnerconfig ARCH_SUPPORTS_RT 1221a50a3f4bSThomas Gleixner bool 1222a50a3f4bSThomas Gleixner 1223fff7fb0bSZhaoxiu Zengconfig CPU_NO_EFFICIENT_FFS 1224fff7fb0bSZhaoxiu Zeng def_bool n 1225fff7fb0bSZhaoxiu Zeng 1226ba14a194SAndy Lutomirskiconfig HAVE_ARCH_VMAP_STACK 1227ba14a194SAndy Lutomirski def_bool n 1228ba14a194SAndy Lutomirski help 1229ba14a194SAndy Lutomirski An arch should select this symbol if it can support kernel stacks 1230ba14a194SAndy Lutomirski in vmalloc space. This means: 1231ba14a194SAndy Lutomirski 1232ba14a194SAndy Lutomirski - vmalloc space must be large enough to hold many kernel stacks. 1233ba14a194SAndy Lutomirski This may rule out many 32-bit architectures. 1234ba14a194SAndy Lutomirski 1235ba14a194SAndy Lutomirski - Stacks in vmalloc space need to work reliably. For example, if 1236ba14a194SAndy Lutomirski vmap page tables are created on demand, either this mechanism 1237ba14a194SAndy Lutomirski needs to work while the stack points to a virtual address with 1238ba14a194SAndy Lutomirski unpopulated page tables or arch code (switch_to() and switch_mm(), 1239ba14a194SAndy Lutomirski most likely) needs to ensure that the stack's page table entries 1240ba14a194SAndy Lutomirski are populated before running on a possibly unpopulated stack. 1241ba14a194SAndy Lutomirski 1242ba14a194SAndy Lutomirski - If the stack overflows into a guard page, something reasonable 1243ba14a194SAndy Lutomirski should happen. The definition of "reasonable" is flexible, but 1244ba14a194SAndy Lutomirski instantly rebooting without logging anything would be unfriendly. 1245ba14a194SAndy Lutomirski 1246ba14a194SAndy Lutomirskiconfig VMAP_STACK 1247ba14a194SAndy Lutomirski default y 1248ba14a194SAndy Lutomirski bool "Use a virtually-mapped stack" 1249eafb149eSDaniel Axtens depends on HAVE_ARCH_VMAP_STACK 125038dd767dSAndrey Konovalov depends on !KASAN || KASAN_HW_TAGS || KASAN_VMALLOC 1251a7f7f624SMasahiro Yamada help 1252ba14a194SAndy Lutomirski Enable this if you want the use virtually-mapped kernel stacks 1253ba14a194SAndy Lutomirski with guard pages. This causes kernel stack overflows to be 1254ba14a194SAndy Lutomirski caught immediately rather than causing difficult-to-diagnose 1255ba14a194SAndy Lutomirski corruption. 1256ba14a194SAndy Lutomirski 125738dd767dSAndrey Konovalov To use this with software KASAN modes, the architecture must support 125838dd767dSAndrey Konovalov backing virtual mappings with real shadow memory, and KASAN_VMALLOC 125938dd767dSAndrey Konovalov must be enabled. 1260ba14a194SAndy Lutomirski 126139218ff4SKees Cookconfig HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET 126239218ff4SKees Cook def_bool n 126339218ff4SKees Cook help 126439218ff4SKees Cook An arch should select this symbol if it can support kernel stack 126539218ff4SKees Cook offset randomization with calls to add_random_kstack_offset() 126639218ff4SKees Cook during syscall entry and choose_random_kstack_offset() during 126739218ff4SKees Cook syscall exit. Careful removal of -fstack-protector-strong and 126839218ff4SKees Cook -fstack-protector should also be applied to the entry code and 126939218ff4SKees Cook closely examined, as the artificial stack bump looks like an array 127039218ff4SKees Cook to the compiler, so it will attempt to add canary checks regardless 127139218ff4SKees Cook of the static branch state. 127239218ff4SKees Cook 12738cb37a59SMarco Elverconfig RANDOMIZE_KSTACK_OFFSET 12748cb37a59SMarco Elver bool "Support for randomizing kernel stack offset on syscall entry" if EXPERT 12758cb37a59SMarco Elver default y 127639218ff4SKees Cook depends on HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET 1277efa90c11SMarco Elver depends on INIT_STACK_NONE || !CC_IS_CLANG || CLANG_VERSION >= 140000 127839218ff4SKees Cook help 127939218ff4SKees Cook The kernel stack offset can be randomized (after pt_regs) by 128039218ff4SKees Cook roughly 5 bits of entropy, frustrating memory corruption 128139218ff4SKees Cook attacks that depend on stack address determinism or 12828cb37a59SMarco Elver cross-syscall address exposures. 12838cb37a59SMarco Elver 12848cb37a59SMarco Elver The feature is controlled via the "randomize_kstack_offset=on/off" 12858cb37a59SMarco Elver kernel boot param, and if turned off has zero overhead due to its use 12868cb37a59SMarco Elver of static branches (see JUMP_LABEL). 12878cb37a59SMarco Elver 12888cb37a59SMarco Elver If unsure, say Y. 12898cb37a59SMarco Elver 12908cb37a59SMarco Elverconfig RANDOMIZE_KSTACK_OFFSET_DEFAULT 12918cb37a59SMarco Elver bool "Default state of kernel stack offset randomization" 12928cb37a59SMarco Elver depends on RANDOMIZE_KSTACK_OFFSET 12938cb37a59SMarco Elver help 12948cb37a59SMarco Elver Kernel stack offset randomization is controlled by kernel boot param 12958cb37a59SMarco Elver "randomize_kstack_offset=on/off", and this config chooses the default 12968cb37a59SMarco Elver boot state. 129739218ff4SKees Cook 1298ad21fc4fSLaura Abbottconfig ARCH_OPTIONAL_KERNEL_RWX 1299ad21fc4fSLaura Abbott def_bool n 1300ad21fc4fSLaura Abbott 1301ad21fc4fSLaura Abbottconfig ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 1302ad21fc4fSLaura Abbott def_bool n 1303ad21fc4fSLaura Abbott 1304ad21fc4fSLaura Abbottconfig ARCH_HAS_STRICT_KERNEL_RWX 1305ad21fc4fSLaura Abbott def_bool n 1306ad21fc4fSLaura Abbott 13070f5bf6d0SLaura Abbottconfig STRICT_KERNEL_RWX 1308ad21fc4fSLaura Abbott bool "Make kernel text and rodata read-only" if ARCH_OPTIONAL_KERNEL_RWX 1309ad21fc4fSLaura Abbott depends on ARCH_HAS_STRICT_KERNEL_RWX 1310ad21fc4fSLaura Abbott default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 1311ad21fc4fSLaura Abbott help 1312ad21fc4fSLaura Abbott If this is set, kernel text and rodata memory will be made read-only, 1313ad21fc4fSLaura Abbott and non-text memory will be made non-executable. This provides 1314ad21fc4fSLaura Abbott protection against certain security exploits (e.g. executing the heap 1315ad21fc4fSLaura Abbott or modifying text) 1316ad21fc4fSLaura Abbott 1317ad21fc4fSLaura Abbott These features are considered standard security practice these days. 1318ad21fc4fSLaura Abbott You should say Y here in almost all cases. 1319ad21fc4fSLaura Abbott 1320ad21fc4fSLaura Abbottconfig ARCH_HAS_STRICT_MODULE_RWX 1321ad21fc4fSLaura Abbott def_bool n 1322ad21fc4fSLaura Abbott 13230f5bf6d0SLaura Abbottconfig STRICT_MODULE_RWX 1324ad21fc4fSLaura Abbott bool "Set loadable kernel module data as NX and text as RO" if ARCH_OPTIONAL_KERNEL_RWX 1325ad21fc4fSLaura Abbott depends on ARCH_HAS_STRICT_MODULE_RWX && MODULES 1326ad21fc4fSLaura Abbott default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 1327ad21fc4fSLaura Abbott help 1328ad21fc4fSLaura Abbott If this is set, module text and rodata memory will be made read-only, 1329ad21fc4fSLaura Abbott and non-text memory will be made non-executable. This provides 1330ad21fc4fSLaura Abbott protection against certain security exploits (e.g. writing to text) 1331ad21fc4fSLaura Abbott 1332ea8c64acSChristoph Hellwig# select if the architecture provides an asm/dma-direct.h header 1333ea8c64acSChristoph Hellwigconfig ARCH_HAS_PHYS_TO_DMA 1334ea8c64acSChristoph Hellwig bool 1335ea8c64acSChristoph Hellwig 133604f264d3SPaul Burtonconfig HAVE_ARCH_COMPILER_H 133704f264d3SPaul Burton bool 133804f264d3SPaul Burton help 133904f264d3SPaul Burton An architecture can select this if it provides an 134004f264d3SPaul Burton asm/compiler.h header that should be included after 134104f264d3SPaul Burton linux/compiler-*.h in order to override macro definitions that those 134204f264d3SPaul Burton headers generally provide. 134304f264d3SPaul Burton 1344271ca788SArd Biesheuvelconfig HAVE_ARCH_PREL32_RELOCATIONS 1345271ca788SArd Biesheuvel bool 1346271ca788SArd Biesheuvel help 1347271ca788SArd Biesheuvel May be selected by an architecture if it supports place-relative 1348271ca788SArd Biesheuvel 32-bit relocations, both in the toolchain and in the module loader, 1349271ca788SArd Biesheuvel in which case relative references can be used in special sections 1350271ca788SArd Biesheuvel for PCI fixup, initcalls etc which are only half the size on 64 bit 1351271ca788SArd Biesheuvel architectures, and don't require runtime relocation on relocatable 1352271ca788SArd Biesheuvel kernels. 1353271ca788SArd Biesheuvel 1354ce9084baSArd Biesheuvelconfig ARCH_USE_MEMREMAP_PROT 1355ce9084baSArd Biesheuvel bool 1356ce9084baSArd Biesheuvel 1357fb346fd9SWaiman Longconfig LOCK_EVENT_COUNTS 1358fb346fd9SWaiman Long bool "Locking event counts collection" 1359fb346fd9SWaiman Long depends on DEBUG_FS 1360a7f7f624SMasahiro Yamada help 1361fb346fd9SWaiman Long Enable light-weight counting of various locking related events 1362fb346fd9SWaiman Long in the system with minimal performance impact. This reduces 1363fb346fd9SWaiman Long the chance of application behavior change because of timing 1364fb346fd9SWaiman Long differences. The counts are reported via debugfs. 1365fb346fd9SWaiman Long 13665cf896fbSPeter Collingbourne# Select if the architecture has support for applying RELR relocations. 13675cf896fbSPeter Collingbourneconfig ARCH_HAS_RELR 13685cf896fbSPeter Collingbourne bool 13695cf896fbSPeter Collingbourne 13705cf896fbSPeter Collingbourneconfig RELR 13715cf896fbSPeter Collingbourne bool "Use RELR relocation packing" 13725cf896fbSPeter Collingbourne depends on ARCH_HAS_RELR && TOOLS_SUPPORT_RELR 13735cf896fbSPeter Collingbourne default y 13745cf896fbSPeter Collingbourne help 13755cf896fbSPeter Collingbourne Store the kernel's dynamic relocations in the RELR relocation packing 13765cf896fbSPeter Collingbourne format. Requires a compatible linker (LLD supports this feature), as 13775cf896fbSPeter Collingbourne well as compatible NM and OBJCOPY utilities (llvm-nm and llvm-objcopy 13785cf896fbSPeter Collingbourne are compatible). 13795cf896fbSPeter Collingbourne 13800c9c1d56SThiago Jung Bauermannconfig ARCH_HAS_MEM_ENCRYPT 13810c9c1d56SThiago Jung Bauermann bool 13820c9c1d56SThiago Jung Bauermann 138346b49b12STom Lendackyconfig ARCH_HAS_CC_PLATFORM 138446b49b12STom Lendacky bool 138546b49b12STom Lendacky 13860e242208SHassan Naveedconfig HAVE_SPARSE_SYSCALL_NR 13870e242208SHassan Naveed bool 13880e242208SHassan Naveed help 13890e242208SHassan Naveed An architecture should select this if its syscall numbering is sparse 13900e242208SHassan Naveed to save space. For example, MIPS architecture has a syscall array with 13910e242208SHassan Naveed entries at 4000, 5000 and 6000 locations. This option turns on syscall 13920e242208SHassan Naveed related optimizations for a given architecture. 13930e242208SHassan Naveed 1394d60d7de3SSven Schnelleconfig ARCH_HAS_VDSO_DATA 1395d60d7de3SSven Schnelle bool 1396d60d7de3SSven Schnelle 1397115284d8SJosh Poimboeufconfig HAVE_STATIC_CALL 1398115284d8SJosh Poimboeuf bool 1399115284d8SJosh Poimboeuf 14009183c3f9SJosh Poimboeufconfig HAVE_STATIC_CALL_INLINE 14019183c3f9SJosh Poimboeuf bool 14029183c3f9SJosh Poimboeuf depends on HAVE_STATIC_CALL 140303f16cd0SJosh Poimboeuf select OBJTOOL 14049183c3f9SJosh Poimboeuf 14056ef869e0SMichal Hockoconfig HAVE_PREEMPT_DYNAMIC 14066ef869e0SMichal Hocko bool 140799cf983cSMark Rutland 140899cf983cSMark Rutlandconfig HAVE_PREEMPT_DYNAMIC_CALL 140999cf983cSMark Rutland bool 14106ef869e0SMichal Hocko depends on HAVE_STATIC_CALL 141199cf983cSMark Rutland select HAVE_PREEMPT_DYNAMIC 14126ef869e0SMichal Hocko help 141399cf983cSMark Rutland An architecture should select this if it can handle the preemption 141499cf983cSMark Rutland model being selected at boot time using static calls. 141599cf983cSMark Rutland 141699cf983cSMark Rutland Where an architecture selects HAVE_STATIC_CALL_INLINE, any call to a 141799cf983cSMark Rutland preemption function will be patched directly. 141899cf983cSMark Rutland 141999cf983cSMark Rutland Where an architecture does not select HAVE_STATIC_CALL_INLINE, any 142099cf983cSMark Rutland call to a preemption function will go through a trampoline, and the 142199cf983cSMark Rutland trampoline will be patched. 142299cf983cSMark Rutland 142399cf983cSMark Rutland It is strongly advised to support inline static call to avoid any 142499cf983cSMark Rutland overhead. 142599cf983cSMark Rutland 142699cf983cSMark Rutlandconfig HAVE_PREEMPT_DYNAMIC_KEY 142799cf983cSMark Rutland bool 1428a0a12c3eSNick Desaulniers depends on HAVE_ARCH_JUMP_LABEL 142999cf983cSMark Rutland select HAVE_PREEMPT_DYNAMIC 143099cf983cSMark Rutland help 143199cf983cSMark Rutland An architecture should select this if it can handle the preemption 143299cf983cSMark Rutland model being selected at boot time using static keys. 143399cf983cSMark Rutland 143499cf983cSMark Rutland Each preemption function will be given an early return based on a 143599cf983cSMark Rutland static key. This should have slightly lower overhead than non-inline 143699cf983cSMark Rutland static calls, as this effectively inlines each trampoline into the 143799cf983cSMark Rutland start of its callee. This may avoid redundant work, and may 143899cf983cSMark Rutland integrate better with CFI schemes. 143999cf983cSMark Rutland 144099cf983cSMark Rutland This will have greater overhead than using inline static calls as 144199cf983cSMark Rutland the call to the preemption function cannot be entirely elided. 14426ef869e0SMichal Hocko 144359612b24SNathan Chancellorconfig ARCH_WANT_LD_ORPHAN_WARN 144459612b24SNathan Chancellor bool 144559612b24SNathan Chancellor help 144659612b24SNathan Chancellor An arch should select this symbol once all linker sections are explicitly 144759612b24SNathan Chancellor included, size-asserted, or discarded in the linker scripts. This is 144859612b24SNathan Chancellor important because we never want expected sections to be placed heuristically 144959612b24SNathan Chancellor by the linker, since the locations of such sections can change between linker 145059612b24SNathan Chancellor versions. 145159612b24SNathan Chancellor 14524f5b0c17SMike Rapoportconfig HAVE_ARCH_PFN_VALID 14534f5b0c17SMike Rapoport bool 14544f5b0c17SMike Rapoport 14555d6ad668SMike Rapoportconfig ARCH_SUPPORTS_DEBUG_PAGEALLOC 14565d6ad668SMike Rapoport bool 14575d6ad668SMike Rapoport 1458df4e817bSPasha Tatashinconfig ARCH_SUPPORTS_PAGE_TABLE_CHECK 1459df4e817bSPasha Tatashin bool 1460df4e817bSPasha Tatashin 14612ca408d9SBrian Gerstconfig ARCH_SPLIT_ARG64 14622ca408d9SBrian Gerst bool 14632ca408d9SBrian Gerst help 14642ca408d9SBrian Gerst If a 32-bit architecture requires 64-bit arguments to be split into 14652ca408d9SBrian Gerst pairs of 32-bit arguments, select this option. 14662ca408d9SBrian Gerst 14677facdc42SAl Viroconfig ARCH_HAS_ELFCORE_COMPAT 14687facdc42SAl Viro bool 14697facdc42SAl Viro 147058e106e7SBalbir Singhconfig ARCH_HAS_PARANOID_L1D_FLUSH 147158e106e7SBalbir Singh bool 147258e106e7SBalbir Singh 1473d593d64fSPrasad Sodagudiconfig ARCH_HAVE_TRACE_MMIO_ACCESS 1474d593d64fSPrasad Sodagudi bool 1475d593d64fSPrasad Sodagudi 14761bdda24cSThomas Gleixnerconfig DYNAMIC_SIGFRAME 14771bdda24cSThomas Gleixner bool 14781bdda24cSThomas Gleixner 147950468e43SJarkko Sakkinen# Select, if arch has a named attribute group bound to NUMA device nodes. 148050468e43SJarkko Sakkinenconfig HAVE_ARCH_NODE_DEV_GROUP 148150468e43SJarkko Sakkinen bool 148250468e43SJarkko Sakkinen 1483eed9a328SYu Zhaoconfig ARCH_HAS_NONLEAF_PMD_YOUNG 1484eed9a328SYu Zhao bool 1485eed9a328SYu Zhao help 1486eed9a328SYu Zhao Architectures that select this option are capable of setting the 1487eed9a328SYu Zhao accessed bit in non-leaf PMD entries when using them as part of linear 1488eed9a328SYu Zhao address translations. Page table walkers that clear the accessed bit 1489eed9a328SYu Zhao may use this capability to reduce their search space. 1490eed9a328SYu Zhao 14912521f2c2SPeter Oberparleitersource "kernel/gcov/Kconfig" 149245332b1bSMasahiro Yamada 149345332b1bSMasahiro Yamadasource "scripts/gcc-plugins/Kconfig" 1494fa1b5d09SLinus Torvalds 1495d49a0626SPeter Zijlstraconfig FUNCTION_ALIGNMENT_4B 1496d49a0626SPeter Zijlstra bool 1497d49a0626SPeter Zijlstra 1498d49a0626SPeter Zijlstraconfig FUNCTION_ALIGNMENT_8B 1499d49a0626SPeter Zijlstra bool 1500d49a0626SPeter Zijlstra 1501d49a0626SPeter Zijlstraconfig FUNCTION_ALIGNMENT_16B 1502d49a0626SPeter Zijlstra bool 1503d49a0626SPeter Zijlstra 1504d49a0626SPeter Zijlstraconfig FUNCTION_ALIGNMENT_32B 1505d49a0626SPeter Zijlstra bool 1506d49a0626SPeter Zijlstra 1507d49a0626SPeter Zijlstraconfig FUNCTION_ALIGNMENT_64B 1508d49a0626SPeter Zijlstra bool 1509d49a0626SPeter Zijlstra 1510d49a0626SPeter Zijlstraconfig FUNCTION_ALIGNMENT 1511d49a0626SPeter Zijlstra int 1512d49a0626SPeter Zijlstra default 64 if FUNCTION_ALIGNMENT_64B 1513d49a0626SPeter Zijlstra default 32 if FUNCTION_ALIGNMENT_32B 1514d49a0626SPeter Zijlstra default 16 if FUNCTION_ALIGNMENT_16B 1515d49a0626SPeter Zijlstra default 8 if FUNCTION_ALIGNMENT_8B 1516d49a0626SPeter Zijlstra default 4 if FUNCTION_ALIGNMENT_4B 1517d49a0626SPeter Zijlstra default 0 1518d49a0626SPeter Zijlstra 151922471e13SRandy Dunlapendmenu 1520