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 1222471e13SRandy Dunlapmenu "General architecture-dependent options" 1322471e13SRandy Dunlap 14692f66f2SHari Bathiniconfig CRASH_CORE 15692f66f2SHari Bathini bool 16692f66f2SHari Bathini 172965faa5SDave Youngconfig KEXEC_CORE 18692f66f2SHari Bathini select CRASH_CORE 192965faa5SDave Young bool 202965faa5SDave Young 21467d2782SThiago Jung Bauermannconfig HAVE_IMA_KEXEC 22467d2782SThiago Jung Bauermann bool 23467d2782SThiago Jung Bauermann 2405736e4aSThomas Gleixnerconfig HOTPLUG_SMT 2505736e4aSThomas Gleixner bool 2605736e4aSThomas Gleixner 27125e5645SMathieu Desnoyersconfig OPROFILE 28b309a294SRobert Richter tristate "OProfile system profiling" 29125e5645SMathieu Desnoyers depends on PROFILING 30125e5645SMathieu Desnoyers depends on HAVE_OPROFILE 31d69d59f4SIngo Molnar select RING_BUFFER 329a5963ebSChristian Borntraeger select RING_BUFFER_ALLOW_SWAP 33125e5645SMathieu Desnoyers help 34125e5645SMathieu Desnoyers OProfile is a profiling system capable of profiling the 35125e5645SMathieu Desnoyers whole system, include the kernel, kernel modules, libraries, 36125e5645SMathieu Desnoyers and applications. 37125e5645SMathieu Desnoyers 38125e5645SMathieu Desnoyers If unsure, say N. 39125e5645SMathieu Desnoyers 404d4036e0SJason Yehconfig OPROFILE_EVENT_MULTIPLEX 414d4036e0SJason Yeh bool "OProfile multiplexing support (EXPERIMENTAL)" 424d4036e0SJason Yeh default n 434d4036e0SJason Yeh depends on OPROFILE && X86 444d4036e0SJason Yeh help 454d4036e0SJason Yeh The number of hardware counters is limited. The multiplexing 464d4036e0SJason Yeh feature enables OProfile to gather more events than counters 474d4036e0SJason Yeh are provided by the hardware. This is realized by switching 489332ef9dSMasahiro Yamada between events at a user specified time interval. 494d4036e0SJason Yeh 504d4036e0SJason Yeh If unsure, say N. 514d4036e0SJason Yeh 52125e5645SMathieu Desnoyersconfig HAVE_OPROFILE 539ba16087SJan Beulich bool 54125e5645SMathieu Desnoyers 55dcfce4a0SRobert Richterconfig OPROFILE_NMI_TIMER 56dcfce4a0SRobert Richter def_bool y 57af9feebeSAnton Blanchard depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI && !PPC64 58dcfce4a0SRobert Richter 59125e5645SMathieu Desnoyersconfig KPROBES 60125e5645SMathieu Desnoyers bool "Kprobes" 6105ed160eSMasami Hiramatsu depends on MODULES 62125e5645SMathieu Desnoyers depends on HAVE_KPROBES 6305ed160eSMasami Hiramatsu select KALLSYMS 64125e5645SMathieu Desnoyers help 65125e5645SMathieu Desnoyers Kprobes allows you to trap at almost any kernel address and 66125e5645SMathieu Desnoyers execute a callback function. register_kprobe() establishes 67125e5645SMathieu Desnoyers a probepoint and specifies the callback. Kprobes is useful 68125e5645SMathieu Desnoyers for kernel debugging, non-intrusive instrumentation and testing. 69125e5645SMathieu Desnoyers If in doubt, say "N". 70125e5645SMathieu Desnoyers 7145f81b1cSSteven Rostedtconfig JUMP_LABEL 72c5905afbSIngo Molnar bool "Optimize very unlikely/likely branches" 7345f81b1cSSteven Rostedt depends on HAVE_ARCH_JUMP_LABEL 74e9666d10SMasahiro Yamada depends on CC_HAS_ASM_GOTO 7545f81b1cSSteven Rostedt help 76c5905afbSIngo Molnar This option enables a transparent branch optimization that 77c5905afbSIngo Molnar makes certain almost-always-true or almost-always-false branch 78c5905afbSIngo Molnar conditions even cheaper to execute within the kernel. 7945f81b1cSSteven Rostedt 80c5905afbSIngo Molnar Certain performance-sensitive kernel code, such as trace points, 81c5905afbSIngo Molnar scheduler functionality, networking code and KVM have such 82c5905afbSIngo Molnar branches and include support for this optimization technique. 83c5905afbSIngo Molnar 84c5905afbSIngo Molnar If it is detected that the compiler has support for "asm goto", 85c5905afbSIngo Molnar the kernel will compile such branches with just a nop 86c5905afbSIngo Molnar instruction. When the condition flag is toggled to true, the 87c5905afbSIngo Molnar nop will be converted to a jump instruction to execute the 88c5905afbSIngo Molnar conditional block of instructions. 89c5905afbSIngo Molnar 90c5905afbSIngo Molnar This technique lowers overhead and stress on the branch prediction 91c5905afbSIngo Molnar of the processor and generally makes the kernel faster. The update 92c5905afbSIngo Molnar of the condition is slower, but those are always very rare. 93c5905afbSIngo Molnar 94c5905afbSIngo Molnar ( On 32-bit x86, the necessary options added to the compiler 95c5905afbSIngo Molnar flags may increase the size of the kernel slightly. ) 9645f81b1cSSteven Rostedt 971987c947SPeter Zijlstraconfig STATIC_KEYS_SELFTEST 981987c947SPeter Zijlstra bool "Static key selftest" 991987c947SPeter Zijlstra depends on JUMP_LABEL 1001987c947SPeter Zijlstra help 1011987c947SPeter Zijlstra Boot time self-test of the branch patching code. 1021987c947SPeter Zijlstra 103afd66255SMasami Hiramatsuconfig OPTPROBES 1045cc718b9SMasami Hiramatsu def_bool y 1055cc718b9SMasami Hiramatsu depends on KPROBES && HAVE_OPTPROBES 106a30b85dfSMasami Hiramatsu select TASKS_RCU if PREEMPT 107afd66255SMasami Hiramatsu 108e7dbfe34SMasami Hiramatsuconfig KPROBES_ON_FTRACE 109e7dbfe34SMasami Hiramatsu def_bool y 110e7dbfe34SMasami Hiramatsu depends on KPROBES && HAVE_KPROBES_ON_FTRACE 111e7dbfe34SMasami Hiramatsu depends on DYNAMIC_FTRACE_WITH_REGS 112e7dbfe34SMasami Hiramatsu help 113e7dbfe34SMasami Hiramatsu If function tracer is enabled and the arch supports full 114e7dbfe34SMasami Hiramatsu passing of pt_regs to function tracing, then kprobes can 115e7dbfe34SMasami Hiramatsu optimize on top of function tracing. 116e7dbfe34SMasami Hiramatsu 1172b144498SSrikar Dronamrajuconfig UPROBES 11809294e31SDavid A. Long def_bool n 119e8f4aa60SAllen Pais depends on ARCH_SUPPORTS_UPROBES 1202b144498SSrikar Dronamraju help 1217b2d81d4SIngo Molnar Uprobes is the user-space counterpart to kprobes: they 1227b2d81d4SIngo Molnar enable instrumentation applications (such as 'perf probe') 1237b2d81d4SIngo Molnar to establish unintrusive probes in user-space binaries and 1247b2d81d4SIngo Molnar libraries, by executing handler functions when the probes 1257b2d81d4SIngo Molnar are hit by user-space applications. 1267b2d81d4SIngo Molnar 1277b2d81d4SIngo Molnar ( These probes come in the form of single-byte breakpoints, 1287b2d81d4SIngo Molnar managed by the kernel and kept transparent to the probed 1297b2d81d4SIngo Molnar application. ) 1302b144498SSrikar Dronamraju 131c19fa94aSJames Hoganconfig HAVE_64BIT_ALIGNED_ACCESS 132c19fa94aSJames Hogan def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS 133c19fa94aSJames Hogan help 134c19fa94aSJames Hogan Some architectures require 64 bit accesses to be 64 bit 135c19fa94aSJames Hogan aligned, which also requires structs containing 64 bit values 136c19fa94aSJames Hogan to be 64 bit aligned too. This includes some 32 bit 137c19fa94aSJames Hogan architectures which can do 64 bit accesses, as well as 64 bit 138c19fa94aSJames Hogan architectures without unaligned access. 139c19fa94aSJames Hogan 140c19fa94aSJames Hogan This symbol should be selected by an architecture if 64 bit 141c19fa94aSJames Hogan accesses are required to be 64 bit aligned in this way even 142c19fa94aSJames Hogan though it is not a 64 bit architecture. 143c19fa94aSJames Hogan 144c19fa94aSJames Hogan See Documentation/unaligned-memory-access.txt for more 145c19fa94aSJames Hogan information on the topic of unaligned memory accesses. 146c19fa94aSJames Hogan 14758340a07SJohannes Bergconfig HAVE_EFFICIENT_UNALIGNED_ACCESS 1489ba16087SJan Beulich bool 14958340a07SJohannes Berg help 15058340a07SJohannes Berg Some architectures are unable to perform unaligned accesses 15158340a07SJohannes Berg without the use of get_unaligned/put_unaligned. Others are 15258340a07SJohannes Berg unable to perform such accesses efficiently (e.g. trap on 15358340a07SJohannes Berg unaligned access and require fixing it up in the exception 15458340a07SJohannes Berg handler.) 15558340a07SJohannes Berg 15658340a07SJohannes Berg This symbol should be selected by an architecture if it can 15758340a07SJohannes Berg perform unaligned accesses efficiently to allow different 15858340a07SJohannes Berg code paths to be selected for these cases. Some network 15958340a07SJohannes Berg drivers, for example, could opt to not fix up alignment 16058340a07SJohannes Berg problems with received packets if doing so would not help 16158340a07SJohannes Berg much. 16258340a07SJohannes Berg 16358340a07SJohannes Berg See Documentation/unaligned-memory-access.txt for more 16458340a07SJohannes Berg information on the topic of unaligned memory accesses. 16558340a07SJohannes Berg 166cf66bb93SDavid Woodhouseconfig ARCH_USE_BUILTIN_BSWAP 167cf66bb93SDavid Woodhouse bool 168cf66bb93SDavid Woodhouse help 169cf66bb93SDavid Woodhouse Modern versions of GCC (since 4.4) have builtin functions 170cf66bb93SDavid Woodhouse for handling byte-swapping. Using these, instead of the old 171cf66bb93SDavid Woodhouse inline assembler that the architecture code provides in the 172cf66bb93SDavid Woodhouse __arch_bswapXX() macros, allows the compiler to see what's 173cf66bb93SDavid Woodhouse happening and offers more opportunity for optimisation. In 174cf66bb93SDavid Woodhouse particular, the compiler will be able to combine the byteswap 175cf66bb93SDavid Woodhouse with a nearby load or store and use load-and-swap or 176cf66bb93SDavid Woodhouse store-and-swap instructions if the architecture has them. It 177cf66bb93SDavid Woodhouse should almost *never* result in code which is worse than the 178cf66bb93SDavid Woodhouse hand-coded assembler in <asm/swab.h>. But just in case it 179cf66bb93SDavid Woodhouse does, the use of the builtins is optional. 180cf66bb93SDavid Woodhouse 181cf66bb93SDavid Woodhouse Any architecture with load-and-swap or store-and-swap 182cf66bb93SDavid Woodhouse instructions should set this. And it shouldn't hurt to set it 183cf66bb93SDavid Woodhouse on architectures that don't have such instructions. 184cf66bb93SDavid Woodhouse 1859edddaa2SAnanth N Mavinakayanahalliconfig KRETPROBES 1869edddaa2SAnanth N Mavinakayanahalli def_bool y 1879edddaa2SAnanth N Mavinakayanahalli depends on KPROBES && HAVE_KRETPROBES 1889edddaa2SAnanth N Mavinakayanahalli 1897c68af6eSAvi Kivityconfig USER_RETURN_NOTIFIER 1907c68af6eSAvi Kivity bool 1917c68af6eSAvi Kivity depends on HAVE_USER_RETURN_NOTIFIER 1927c68af6eSAvi Kivity help 1937c68af6eSAvi Kivity Provide a kernel-internal notification when a cpu is about to 1947c68af6eSAvi Kivity switch to user mode. 1957c68af6eSAvi Kivity 19628b2ee20SRik van Rielconfig HAVE_IOREMAP_PROT 1979ba16087SJan Beulich bool 19828b2ee20SRik van Riel 199125e5645SMathieu Desnoyersconfig HAVE_KPROBES 2009ba16087SJan Beulich bool 2019edddaa2SAnanth N Mavinakayanahalli 2029edddaa2SAnanth N Mavinakayanahalliconfig HAVE_KRETPROBES 2039ba16087SJan Beulich bool 20474bc7ceeSArthur Kepner 205afd66255SMasami Hiramatsuconfig HAVE_OPTPROBES 206afd66255SMasami Hiramatsu bool 207d314d74cSCong Wang 208e7dbfe34SMasami Hiramatsuconfig HAVE_KPROBES_ON_FTRACE 209e7dbfe34SMasami Hiramatsu bool 210e7dbfe34SMasami Hiramatsu 211540adea3SMasami Hiramatsuconfig HAVE_FUNCTION_ERROR_INJECTION 2129802d865SJosef Bacik bool 2139802d865SJosef Bacik 21442a0bb3fSPetr Mladekconfig HAVE_NMI 21542a0bb3fSPetr Mladek bool 21642a0bb3fSPetr Mladek 2171f5a4ad9SRoland McGrath# 2181f5a4ad9SRoland McGrath# An arch should select this if it provides all these things: 2191f5a4ad9SRoland McGrath# 2201f5a4ad9SRoland McGrath# task_pt_regs() in asm/processor.h or asm/ptrace.h 2211f5a4ad9SRoland McGrath# arch_has_single_step() if there is hardware single-step support 2221f5a4ad9SRoland McGrath# arch_has_block_step() if there is hardware block-step support 2231f5a4ad9SRoland McGrath# asm/syscall.h supplying asm-generic/syscall.h interface 2241f5a4ad9SRoland McGrath# linux/regset.h user_regset interfaces 2251f5a4ad9SRoland McGrath# CORE_DUMP_USE_REGSET #define'd in linux/elf.h 2261f5a4ad9SRoland McGrath# TIF_SYSCALL_TRACE calls tracehook_report_syscall_{entry,exit} 2271f5a4ad9SRoland McGrath# TIF_NOTIFY_RESUME calls tracehook_notify_resume() 2281f5a4ad9SRoland McGrath# signal delivery calls tracehook_signal_handler() 2291f5a4ad9SRoland McGrath# 2301f5a4ad9SRoland McGrathconfig HAVE_ARCH_TRACEHOOK 2319ba16087SJan Beulich bool 2321f5a4ad9SRoland McGrath 233c64be2bbSMarek Szyprowskiconfig HAVE_DMA_CONTIGUOUS 234c64be2bbSMarek Szyprowski bool 235c64be2bbSMarek Szyprowski 23629d5e047SThomas Gleixnerconfig GENERIC_SMP_IDLE_THREAD 23729d5e047SThomas Gleixner bool 23829d5e047SThomas Gleixner 239485cf5daSKevin Hilmanconfig GENERIC_IDLE_POLL_SETUP 240485cf5daSKevin Hilman bool 241485cf5daSKevin Hilman 2426974f0c4SDaniel Micayconfig ARCH_HAS_FORTIFY_SOURCE 2436974f0c4SDaniel Micay bool 2446974f0c4SDaniel Micay help 2456974f0c4SDaniel Micay An architecture should select this when it can successfully 2466974f0c4SDaniel Micay build and run with CONFIG_FORTIFY_SOURCE. 2476974f0c4SDaniel Micay 248d2852a22SDaniel Borkmann# Select if arch has all set_memory_ro/rw/x/nx() functions in asm/cacheflush.h 249d2852a22SDaniel Borkmannconfig ARCH_HAS_SET_MEMORY 250d2852a22SDaniel Borkmann bool 251d2852a22SDaniel Borkmann 2520500871fSDavid Howells# Select if arch init_task must go in the __init_task_data section 2530500871fSDavid Howellsconfig ARCH_TASK_STRUCT_ON_STACK 254a4a2eb49SThomas Gleixner bool 255a4a2eb49SThomas Gleixner 256f5e10287SThomas Gleixner# Select if arch has its private alloc_task_struct() function 257f5e10287SThomas Gleixnerconfig ARCH_TASK_STRUCT_ALLOCATOR 258f5e10287SThomas Gleixner bool 259f5e10287SThomas Gleixner 2605905429aSKees Cookconfig HAVE_ARCH_THREAD_STRUCT_WHITELIST 2615905429aSKees Cook bool 2625905429aSKees Cook depends on !ARCH_TASK_STRUCT_ALLOCATOR 2635905429aSKees Cook help 2645905429aSKees Cook An architecture should select this to provide hardened usercopy 2655905429aSKees Cook knowledge about what region of the thread_struct should be 2665905429aSKees Cook whitelisted for copying to userspace. Normally this is only the 2675905429aSKees Cook FPU registers. Specifically, arch_thread_struct_whitelist() 2685905429aSKees Cook should be implemented. Without this, the entire thread_struct 2695905429aSKees Cook field in task_struct will be left whitelisted. 2705905429aSKees Cook 271b235beeaSLinus Torvalds# Select if arch has its private alloc_thread_stack() function 272b235beeaSLinus Torvaldsconfig ARCH_THREAD_STACK_ALLOCATOR 273f5e10287SThomas Gleixner bool 274f5e10287SThomas Gleixner 2755aaeb5c0SIngo Molnar# Select if arch wants to size task_struct dynamically via arch_task_struct_size: 2765aaeb5c0SIngo Molnarconfig ARCH_WANTS_DYNAMIC_TASK_STRUCT 2775aaeb5c0SIngo Molnar bool 2785aaeb5c0SIngo Molnar 279942fa985SYury Norovconfig ARCH_32BIT_OFF_T 280942fa985SYury Norov bool 281942fa985SYury Norov depends on !64BIT 282942fa985SYury Norov help 283942fa985SYury Norov All new 32-bit architectures should have 64-bit off_t type on 284942fa985SYury Norov userspace side which corresponds to the loff_t kernel type. This 285942fa985SYury Norov is the requirement for modern ABIs. Some existing architectures 286942fa985SYury Norov still support 32-bit off_t. This option is enabled for all such 287942fa985SYury Norov architectures explicitly. 288942fa985SYury Norov 289f850c30cSHeiko Carstensconfig HAVE_REGS_AND_STACK_ACCESS_API 290f850c30cSHeiko Carstens bool 291e01292b1SHeiko Carstens help 292e01292b1SHeiko Carstens This symbol should be selected by an architecure if it supports 293e01292b1SHeiko Carstens the API needed to access registers and stack entries from pt_regs, 294e01292b1SHeiko Carstens declared in asm/ptrace.h 295e01292b1SHeiko Carstens For example the kprobes-based event tracer needs this API. 296f850c30cSHeiko Carstens 297d7822b1eSMathieu Desnoyersconfig HAVE_RSEQ 298d7822b1eSMathieu Desnoyers bool 299d7822b1eSMathieu Desnoyers depends on HAVE_REGS_AND_STACK_ACCESS_API 300d7822b1eSMathieu Desnoyers help 301d7822b1eSMathieu Desnoyers This symbol should be selected by an architecture if it 302d7822b1eSMathieu Desnoyers supports an implementation of restartable sequences. 303d7822b1eSMathieu Desnoyers 3043c88ee19SMasami Hiramatsuconfig HAVE_FUNCTION_ARG_ACCESS_API 3053c88ee19SMasami Hiramatsu bool 3063c88ee19SMasami Hiramatsu help 3073c88ee19SMasami Hiramatsu This symbol should be selected by an architecure if it supports 3083c88ee19SMasami Hiramatsu the API needed to access function arguments from pt_regs, 3093c88ee19SMasami Hiramatsu declared in asm/ptrace.h 3103c88ee19SMasami Hiramatsu 3119483a578SDavid Brownellconfig HAVE_CLK 3129ba16087SJan Beulich bool 3139483a578SDavid Brownell help 3149483a578SDavid Brownell The <linux/clk.h> calls support software clock gating and 3159483a578SDavid Brownell thus are a key power management tool on many systems. 3169483a578SDavid Brownell 31762a038d3SK.Prasadconfig HAVE_HW_BREAKPOINT 31862a038d3SK.Prasad bool 31999e8c5a3SFrederic Weisbecker depends on PERF_EVENTS 32062a038d3SK.Prasad 3210102752eSFrederic Weisbeckerconfig HAVE_MIXED_BREAKPOINTS_REGS 3220102752eSFrederic Weisbecker bool 3230102752eSFrederic Weisbecker depends on HAVE_HW_BREAKPOINT 3240102752eSFrederic Weisbecker help 3250102752eSFrederic Weisbecker Depending on the arch implementation of hardware breakpoints, 3260102752eSFrederic Weisbecker some of them have separate registers for data and instruction 3270102752eSFrederic Weisbecker breakpoints addresses, others have mixed registers to store 3280102752eSFrederic Weisbecker them but define the access type in a control register. 3290102752eSFrederic Weisbecker Select this option if your arch implements breakpoints under the 3300102752eSFrederic Weisbecker latter fashion. 3310102752eSFrederic Weisbecker 3327c68af6eSAvi Kivityconfig HAVE_USER_RETURN_NOTIFIER 3337c68af6eSAvi Kivity bool 334a1922ed6SIngo Molnar 335c01d4323SFrederic Weisbeckerconfig HAVE_PERF_EVENTS_NMI 336c01d4323SFrederic Weisbecker bool 33723637d47SFrederic Weisbecker help 33823637d47SFrederic Weisbecker System hardware can generate an NMI using the perf event 33923637d47SFrederic Weisbecker subsystem. Also has support for calculating CPU cycle events 34023637d47SFrederic Weisbecker to determine how many clock cycles in a given period. 341c01d4323SFrederic Weisbecker 34205a4a952SNicholas Pigginconfig HAVE_HARDLOCKUP_DETECTOR_PERF 34305a4a952SNicholas Piggin bool 34405a4a952SNicholas Piggin depends on HAVE_PERF_EVENTS_NMI 34505a4a952SNicholas Piggin help 34605a4a952SNicholas Piggin The arch chooses to use the generic perf-NMI-based hardlockup 34705a4a952SNicholas Piggin detector. Must define HAVE_PERF_EVENTS_NMI. 34805a4a952SNicholas Piggin 34905a4a952SNicholas Pigginconfig HAVE_NMI_WATCHDOG 35005a4a952SNicholas Piggin depends on HAVE_NMI 35105a4a952SNicholas Piggin bool 35205a4a952SNicholas Piggin help 35305a4a952SNicholas Piggin The arch provides a low level NMI watchdog. It provides 35405a4a952SNicholas Piggin asm/nmi.h, and defines its own arch_touch_nmi_watchdog(). 35505a4a952SNicholas Piggin 35605a4a952SNicholas Pigginconfig HAVE_HARDLOCKUP_DETECTOR_ARCH 35705a4a952SNicholas Piggin bool 35805a4a952SNicholas Piggin select HAVE_NMI_WATCHDOG 35905a4a952SNicholas Piggin help 36005a4a952SNicholas Piggin The arch chooses to provide its own hardlockup detector, which is 36105a4a952SNicholas Piggin a superset of the HAVE_NMI_WATCHDOG. It also conforms to config 36205a4a952SNicholas Piggin interfaces and parameters provided by hardlockup detector subsystem. 36305a4a952SNicholas Piggin 364c5e63197SJiri Olsaconfig HAVE_PERF_REGS 365c5e63197SJiri Olsa bool 366c5e63197SJiri Olsa help 367c5e63197SJiri Olsa Support selective register dumps for perf events. This includes 368c5e63197SJiri Olsa bit-mapping of each registers and a unique architecture id. 369c5e63197SJiri Olsa 370c5ebcedbSJiri Olsaconfig HAVE_PERF_USER_STACK_DUMP 371c5ebcedbSJiri Olsa bool 372c5ebcedbSJiri Olsa help 373c5ebcedbSJiri Olsa Support user stack dumps for perf event samples. This needs 374c5ebcedbSJiri Olsa access to the user stack pointer which is not unified across 375c5ebcedbSJiri Olsa architectures. 376c5ebcedbSJiri Olsa 377bf5438fcSJason Baronconfig HAVE_ARCH_JUMP_LABEL 378bf5438fcSJason Baron bool 379bf5438fcSJason Baron 38050ff18abSArd Biesheuvelconfig HAVE_ARCH_JUMP_LABEL_RELATIVE 38150ff18abSArd Biesheuvel bool 38250ff18abSArd Biesheuvel 38326723911SPeter Zijlstraconfig HAVE_RCU_TABLE_FREE 38426723911SPeter Zijlstra bool 38526723911SPeter Zijlstra 386d86564a2SPeter Zijlstraconfig HAVE_RCU_TABLE_INVALIDATE 387d86564a2SPeter Zijlstra bool 388d86564a2SPeter Zijlstra 389df013ffbSHuang Yingconfig ARCH_HAVE_NMI_SAFE_CMPXCHG 390df013ffbSHuang Ying bool 391df013ffbSHuang Ying 39243570fd2SHeiko Carstensconfig HAVE_ALIGNED_STRUCT_PAGE 39343570fd2SHeiko Carstens bool 39443570fd2SHeiko Carstens help 39543570fd2SHeiko Carstens This makes sure that struct pages are double word aligned and that 39643570fd2SHeiko Carstens e.g. the SLUB allocator can perform double word atomic operations 39743570fd2SHeiko Carstens on a struct page for better performance. However selecting this 39843570fd2SHeiko Carstens might increase the size of a struct page by a word. 39943570fd2SHeiko Carstens 4004156153cSHeiko Carstensconfig HAVE_CMPXCHG_LOCAL 4014156153cSHeiko Carstens bool 4024156153cSHeiko Carstens 4032565409fSHeiko Carstensconfig HAVE_CMPXCHG_DOUBLE 4042565409fSHeiko Carstens bool 4052565409fSHeiko Carstens 40677e58496SPaul E. McKenneyconfig ARCH_WEAK_RELEASE_ACQUIRE 40777e58496SPaul E. McKenney bool 40877e58496SPaul E. McKenney 409c1d7e01dSWill Deaconconfig ARCH_WANT_IPC_PARSE_VERSION 410c1d7e01dSWill Deacon bool 411c1d7e01dSWill Deacon 412c1d7e01dSWill Deaconconfig ARCH_WANT_COMPAT_IPC_PARSE_VERSION 413c1d7e01dSWill Deacon bool 414c1d7e01dSWill Deacon 41548b25c43SChris Metcalfconfig ARCH_WANT_OLD_COMPAT_IPC 416c1d7e01dSWill Deacon select ARCH_WANT_COMPAT_IPC_PARSE_VERSION 41748b25c43SChris Metcalf bool 41848b25c43SChris Metcalf 419e2cfabdfSWill Drewryconfig HAVE_ARCH_SECCOMP_FILTER 420e2cfabdfSWill Drewry bool 421e2cfabdfSWill Drewry help 422fb0fadf9SWill Drewry An arch should select this symbol if it provides all of these things: 423bb6ea430SWill Drewry - syscall_get_arch() 424bb6ea430SWill Drewry - syscall_get_arguments() 425bb6ea430SWill Drewry - syscall_rollback() 426bb6ea430SWill Drewry - syscall_set_return_value() 427fb0fadf9SWill Drewry - SIGSYS siginfo_t support 428fb0fadf9SWill Drewry - secure_computing is called from a ptrace_event()-safe context 429fb0fadf9SWill Drewry - secure_computing return value is checked and a return value of -1 430fb0fadf9SWill Drewry results in the system call being skipped immediately. 43148dc92b9SKees Cook - seccomp syscall wired up 432e2cfabdfSWill Drewry 433e2cfabdfSWill Drewryconfig SECCOMP_FILTER 434e2cfabdfSWill Drewry def_bool y 435e2cfabdfSWill Drewry depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET 436e2cfabdfSWill Drewry help 437e2cfabdfSWill Drewry Enable tasks to build secure computing environments defined 438e2cfabdfSWill Drewry in terms of Berkeley Packet Filter programs which implement 439e2cfabdfSWill Drewry task-defined system call filtering polices. 440e2cfabdfSWill Drewry 4415fb94e9cSMauro Carvalho Chehab See Documentation/userspace-api/seccomp_filter.rst for details. 442e2cfabdfSWill Drewry 443afaef01cSAlexander Popovconfig HAVE_ARCH_STACKLEAK 444afaef01cSAlexander Popov bool 445afaef01cSAlexander Popov help 446afaef01cSAlexander Popov An architecture should select this if it has the code which 447afaef01cSAlexander Popov fills the used part of the kernel stack with the STACKLEAK_POISON 448afaef01cSAlexander Popov value before returning from system calls. 449afaef01cSAlexander Popov 450d148eac0SMasahiro Yamadaconfig HAVE_STACKPROTECTOR 45119952a92SKees Cook bool 45219952a92SKees Cook help 45319952a92SKees Cook An arch should select this symbol if: 45419952a92SKees Cook - it has implemented a stack canary (e.g. __stack_chk_guard) 45519952a92SKees Cook 4562a61f474SMasahiro Yamadaconfig CC_HAS_STACKPROTECTOR_NONE 4572a61f474SMasahiro Yamada def_bool $(cc-option,-fno-stack-protector) 4582a61f474SMasahiro Yamada 459050e9baaSLinus Torvaldsconfig STACKPROTECTOR 4602a61f474SMasahiro Yamada bool "Stack Protector buffer overflow detection" 461d148eac0SMasahiro Yamada depends on HAVE_STACKPROTECTOR 4622a61f474SMasahiro Yamada depends on $(cc-option,-fstack-protector) 4632a61f474SMasahiro Yamada default y 4648779657dSKees Cook help 4658779657dSKees Cook This option turns on the "stack-protector" GCC feature. This 46619952a92SKees Cook feature puts, at the beginning of functions, a canary value on 46719952a92SKees Cook the stack just before the return address, and validates 46819952a92SKees Cook the value just before actually returning. Stack based buffer 46919952a92SKees Cook overflows (that need to overwrite this return address) now also 47019952a92SKees Cook overwrite the canary, which gets detected and the attack is then 47119952a92SKees Cook neutralized via a kernel panic. 47219952a92SKees Cook 4738779657dSKees Cook Functions will have the stack-protector canary logic added if they 4748779657dSKees Cook have an 8-byte or larger character array on the stack. 4758779657dSKees Cook 47619952a92SKees Cook This feature requires gcc version 4.2 or above, or a distribution 4778779657dSKees Cook gcc with the feature backported ("-fstack-protector"). 4788779657dSKees Cook 4798779657dSKees Cook On an x86 "defconfig" build, this feature adds canary checks to 4808779657dSKees Cook about 3% of all kernel functions, which increases kernel code size 4818779657dSKees Cook by about 0.3%. 4828779657dSKees Cook 483050e9baaSLinus Torvaldsconfig STACKPROTECTOR_STRONG 4842a61f474SMasahiro Yamada bool "Strong Stack Protector" 485050e9baaSLinus Torvalds depends on STACKPROTECTOR 4862a61f474SMasahiro Yamada depends on $(cc-option,-fstack-protector-strong) 4872a61f474SMasahiro Yamada default y 4888779657dSKees Cook help 4898779657dSKees Cook Functions will have the stack-protector canary logic added in any 4908779657dSKees Cook of the following conditions: 4918779657dSKees Cook 4928779657dSKees Cook - local variable's address used as part of the right hand side of an 4938779657dSKees Cook assignment or function argument 4948779657dSKees Cook - local variable is an array (or union containing an array), 4958779657dSKees Cook regardless of array type or length 4968779657dSKees Cook - uses register local variables 4978779657dSKees Cook 4988779657dSKees Cook This feature requires gcc version 4.9 or above, or a distribution 4998779657dSKees Cook gcc with the feature backported ("-fstack-protector-strong"). 5008779657dSKees Cook 5018779657dSKees Cook On an x86 "defconfig" build, this feature adds canary checks to 5028779657dSKees Cook about 20% of all kernel functions, which increases the kernel code 5038779657dSKees Cook size by about 2%. 5048779657dSKees Cook 5050f60a8efSKees Cookconfig HAVE_ARCH_WITHIN_STACK_FRAMES 5060f60a8efSKees Cook bool 5070f60a8efSKees Cook help 5080f60a8efSKees Cook An architecture should select this if it can walk the kernel stack 5090f60a8efSKees Cook frames to determine if an object is part of either the arguments 5100f60a8efSKees Cook or local variables (i.e. that it excludes saved return addresses, 5110f60a8efSKees Cook and similar) by implementing an inline arch_within_stack_frames(), 5120f60a8efSKees Cook which is used by CONFIG_HARDENED_USERCOPY. 5130f60a8efSKees Cook 51491d1aa43SFrederic Weisbeckerconfig HAVE_CONTEXT_TRACKING 5152b1d5024SFrederic Weisbecker bool 5162b1d5024SFrederic Weisbecker help 51791d1aa43SFrederic Weisbecker Provide kernel/user boundaries probes necessary for subsystems 51891d1aa43SFrederic Weisbecker that need it, such as userspace RCU extended quiescent state. 51991d1aa43SFrederic Weisbecker Syscalls need to be wrapped inside user_exit()-user_enter() through 52091d1aa43SFrederic Weisbecker the slow path using TIF_NOHZ flag. Exceptions handlers must be 52191d1aa43SFrederic Weisbecker wrapped as well. Irqs are already protected inside 52291d1aa43SFrederic Weisbecker rcu_irq_enter/rcu_irq_exit() but preemption or signal handling on 52391d1aa43SFrederic Weisbecker irq exit still need to be protected. 5242b1d5024SFrederic Weisbecker 525b952741cSFrederic Weisbeckerconfig HAVE_VIRT_CPU_ACCOUNTING 526b952741cSFrederic Weisbecker bool 527b952741cSFrederic Weisbecker 52840565b5aSStanislaw Gruszkaconfig ARCH_HAS_SCALED_CPUTIME 52940565b5aSStanislaw Gruszka bool 53040565b5aSStanislaw Gruszka 531554b0004SKevin Hilmanconfig HAVE_VIRT_CPU_ACCOUNTING_GEN 532554b0004SKevin Hilman bool 533554b0004SKevin Hilman default y if 64BIT 534554b0004SKevin Hilman help 535554b0004SKevin Hilman With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit. 536554b0004SKevin Hilman Before enabling this option, arch code must be audited 537554b0004SKevin Hilman to ensure there are no races in concurrent read/write of 538554b0004SKevin Hilman cputime_t. For example, reading/writing 64-bit cputime_t on 539554b0004SKevin Hilman some 32-bit arches may require multiple accesses, so proper 540554b0004SKevin Hilman locking is needed to protect against concurrent accesses. 541554b0004SKevin Hilman 542554b0004SKevin Hilman 543fdf9c356SFrederic Weisbeckerconfig HAVE_IRQ_TIME_ACCOUNTING 544fdf9c356SFrederic Weisbecker bool 545fdf9c356SFrederic Weisbecker help 546fdf9c356SFrederic Weisbecker Archs need to ensure they use a high enough resolution clock to 547fdf9c356SFrederic Weisbecker support irq time accounting and then call enable_sched_clock_irqtime(). 548fdf9c356SFrederic Weisbecker 5492c91bd4aSJoel Fernandes (Google)config HAVE_MOVE_PMD 5502c91bd4aSJoel Fernandes (Google) bool 5512c91bd4aSJoel Fernandes (Google) help 5522c91bd4aSJoel Fernandes (Google) Archs that select this are able to move page tables at the PMD level. 5532c91bd4aSJoel Fernandes (Google) 55415626062SGerald Schaeferconfig HAVE_ARCH_TRANSPARENT_HUGEPAGE 55515626062SGerald Schaefer bool 55615626062SGerald Schaefer 557a00cc7d9SMatthew Wilcoxconfig HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD 558a00cc7d9SMatthew Wilcox bool 559a00cc7d9SMatthew Wilcox 5600ddab1d2SToshi Kaniconfig HAVE_ARCH_HUGE_VMAP 5610ddab1d2SToshi Kani bool 5620ddab1d2SToshi Kani 5630f8975ecSPavel Emelyanovconfig HAVE_ARCH_SOFT_DIRTY 5640f8975ecSPavel Emelyanov bool 5650f8975ecSPavel Emelyanov 566786d35d4SDavid Howellsconfig HAVE_MOD_ARCH_SPECIFIC 567786d35d4SDavid Howells bool 568786d35d4SDavid Howells help 569786d35d4SDavid Howells The arch uses struct mod_arch_specific to store data. Many arches 570786d35d4SDavid Howells just need a simple module loader without arch specific data - those 571786d35d4SDavid Howells should not enable this. 572786d35d4SDavid Howells 573786d35d4SDavid Howellsconfig MODULES_USE_ELF_RELA 574786d35d4SDavid Howells bool 575786d35d4SDavid Howells help 576786d35d4SDavid Howells Modules only use ELF RELA relocations. Modules with ELF REL 577786d35d4SDavid Howells relocations will give an error. 578786d35d4SDavid Howells 579786d35d4SDavid Howellsconfig MODULES_USE_ELF_REL 580786d35d4SDavid Howells bool 581786d35d4SDavid Howells help 582786d35d4SDavid Howells Modules only use ELF REL relocations. Modules with ELF RELA 583786d35d4SDavid Howells relocations will give an error. 584786d35d4SDavid Howells 585cc1f0274SFrederic Weisbeckerconfig HAVE_IRQ_EXIT_ON_IRQ_STACK 586cc1f0274SFrederic Weisbecker bool 587cc1f0274SFrederic Weisbecker help 588cc1f0274SFrederic Weisbecker Architecture doesn't only execute the irq handler on the irq stack 589cc1f0274SFrederic Weisbecker but also irq_exit(). This way we can process softirqs on this irq 590cc1f0274SFrederic Weisbecker stack instead of switching to a new one when we call __do_softirq() 591cc1f0274SFrederic Weisbecker in the end of an hardirq. 592cc1f0274SFrederic Weisbecker This spares a stack switch and improves cache usage on softirq 593cc1f0274SFrederic Weisbecker processing. 594cc1f0274SFrederic Weisbecker 595235a8f02SKirill A. Shutemovconfig PGTABLE_LEVELS 596235a8f02SKirill A. Shutemov int 597235a8f02SKirill A. Shutemov default 2 598235a8f02SKirill A. Shutemov 5992b68f6caSKees Cookconfig ARCH_HAS_ELF_RANDOMIZE 6002b68f6caSKees Cook bool 6012b68f6caSKees Cook help 6022b68f6caSKees Cook An architecture supports choosing randomized locations for 6032b68f6caSKees Cook stack, mmap, brk, and ET_DYN. Defined functions: 6042b68f6caSKees Cook - arch_mmap_rnd() 605204db6edSKees Cook - arch_randomize_brk() 6062b68f6caSKees Cook 607d07e2259SDaniel Cashmanconfig HAVE_ARCH_MMAP_RND_BITS 608d07e2259SDaniel Cashman bool 609d07e2259SDaniel Cashman help 610d07e2259SDaniel Cashman An arch should select this symbol if it supports setting a variable 611d07e2259SDaniel Cashman number of bits for use in establishing the base address for mmap 612d07e2259SDaniel Cashman allocations, has MMU enabled and provides values for both: 613d07e2259SDaniel Cashman - ARCH_MMAP_RND_BITS_MIN 614d07e2259SDaniel Cashman - ARCH_MMAP_RND_BITS_MAX 615d07e2259SDaniel Cashman 6165f56a5dfSJiri Slabyconfig HAVE_EXIT_THREAD 6175f56a5dfSJiri Slaby bool 6185f56a5dfSJiri Slaby help 6195f56a5dfSJiri Slaby An architecture implements exit_thread. 6205f56a5dfSJiri Slaby 621d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_BITS_MIN 622d07e2259SDaniel Cashman int 623d07e2259SDaniel Cashman 624d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_BITS_MAX 625d07e2259SDaniel Cashman int 626d07e2259SDaniel Cashman 627d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_BITS_DEFAULT 628d07e2259SDaniel Cashman int 629d07e2259SDaniel Cashman 630d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_BITS 631d07e2259SDaniel Cashman int "Number of bits to use for ASLR of mmap base address" if EXPERT 632d07e2259SDaniel Cashman range ARCH_MMAP_RND_BITS_MIN ARCH_MMAP_RND_BITS_MAX 633d07e2259SDaniel Cashman default ARCH_MMAP_RND_BITS_DEFAULT if ARCH_MMAP_RND_BITS_DEFAULT 634d07e2259SDaniel Cashman default ARCH_MMAP_RND_BITS_MIN 635d07e2259SDaniel Cashman depends on HAVE_ARCH_MMAP_RND_BITS 636d07e2259SDaniel Cashman help 637d07e2259SDaniel Cashman This value can be used to select the number of bits to use to 638d07e2259SDaniel Cashman determine the random offset to the base address of vma regions 639d07e2259SDaniel Cashman resulting from mmap allocations. This value will be bounded 640d07e2259SDaniel Cashman by the architecture's minimum and maximum supported values. 641d07e2259SDaniel Cashman 642d07e2259SDaniel Cashman This value can be changed after boot using the 643d07e2259SDaniel Cashman /proc/sys/vm/mmap_rnd_bits tunable 644d07e2259SDaniel Cashman 645d07e2259SDaniel Cashmanconfig HAVE_ARCH_MMAP_RND_COMPAT_BITS 646d07e2259SDaniel Cashman bool 647d07e2259SDaniel Cashman help 648d07e2259SDaniel Cashman An arch should select this symbol if it supports running applications 649d07e2259SDaniel Cashman in compatibility mode, supports setting a variable number of bits for 650d07e2259SDaniel Cashman use in establishing the base address for mmap allocations, has MMU 651d07e2259SDaniel Cashman enabled and provides values for both: 652d07e2259SDaniel Cashman - ARCH_MMAP_RND_COMPAT_BITS_MIN 653d07e2259SDaniel Cashman - ARCH_MMAP_RND_COMPAT_BITS_MAX 654d07e2259SDaniel Cashman 655d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_COMPAT_BITS_MIN 656d07e2259SDaniel Cashman int 657d07e2259SDaniel Cashman 658d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_COMPAT_BITS_MAX 659d07e2259SDaniel Cashman int 660d07e2259SDaniel Cashman 661d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_COMPAT_BITS_DEFAULT 662d07e2259SDaniel Cashman int 663d07e2259SDaniel Cashman 664d07e2259SDaniel Cashmanconfig ARCH_MMAP_RND_COMPAT_BITS 665d07e2259SDaniel Cashman int "Number of bits to use for ASLR of mmap base address for compatible applications" if EXPERT 666d07e2259SDaniel Cashman range ARCH_MMAP_RND_COMPAT_BITS_MIN ARCH_MMAP_RND_COMPAT_BITS_MAX 667d07e2259SDaniel Cashman default ARCH_MMAP_RND_COMPAT_BITS_DEFAULT if ARCH_MMAP_RND_COMPAT_BITS_DEFAULT 668d07e2259SDaniel Cashman default ARCH_MMAP_RND_COMPAT_BITS_MIN 669d07e2259SDaniel Cashman depends on HAVE_ARCH_MMAP_RND_COMPAT_BITS 670d07e2259SDaniel Cashman help 671d07e2259SDaniel Cashman This value can be used to select the number of bits to use to 672d07e2259SDaniel Cashman determine the random offset to the base address of vma regions 673d07e2259SDaniel Cashman resulting from mmap allocations for compatible applications This 674d07e2259SDaniel Cashman value will be bounded by the architecture's minimum and maximum 675d07e2259SDaniel Cashman supported values. 676d07e2259SDaniel Cashman 677d07e2259SDaniel Cashman This value can be changed after boot using the 678d07e2259SDaniel Cashman /proc/sys/vm/mmap_rnd_compat_bits tunable 679d07e2259SDaniel Cashman 6801b028f78SDmitry Safonovconfig HAVE_ARCH_COMPAT_MMAP_BASES 6811b028f78SDmitry Safonov bool 6821b028f78SDmitry Safonov help 6831b028f78SDmitry Safonov This allows 64bit applications to invoke 32-bit mmap() syscall 6841b028f78SDmitry Safonov and vice-versa 32-bit applications to call 64-bit mmap(). 6851b028f78SDmitry Safonov Required for applications doing different bitness syscalls. 6861b028f78SDmitry Safonov 6873033f14aSJosh Triplettconfig HAVE_COPY_THREAD_TLS 6883033f14aSJosh Triplett bool 6893033f14aSJosh Triplett help 6903033f14aSJosh Triplett Architecture provides copy_thread_tls to accept tls argument via 6913033f14aSJosh Triplett normal C parameter passing, rather than extracting the syscall 6923033f14aSJosh Triplett argument from pt_regs. 6933033f14aSJosh Triplett 694b9ab5ebbSJosh Poimboeufconfig HAVE_STACK_VALIDATION 695b9ab5ebbSJosh Poimboeuf bool 696b9ab5ebbSJosh Poimboeuf help 697b9ab5ebbSJosh Poimboeuf Architecture supports the 'objtool check' host tool command, which 698b9ab5ebbSJosh Poimboeuf performs compile-time stack metadata validation. 699b9ab5ebbSJosh Poimboeuf 700af085d90SJosh Poimboeufconfig HAVE_RELIABLE_STACKTRACE 701af085d90SJosh Poimboeuf bool 702af085d90SJosh Poimboeuf help 703af085d90SJosh Poimboeuf Architecture has a save_stack_trace_tsk_reliable() function which 704af085d90SJosh Poimboeuf only returns a stack trace if it can guarantee the trace is reliable. 705af085d90SJosh Poimboeuf 706468a9428SGeorge Spelvinconfig HAVE_ARCH_HASH 707468a9428SGeorge Spelvin bool 708468a9428SGeorge Spelvin default n 709468a9428SGeorge Spelvin help 710468a9428SGeorge Spelvin If this is set, the architecture provides an <asm/hash.h> 711468a9428SGeorge Spelvin file which provides platform-specific implementations of some 712468a9428SGeorge Spelvin functions in <linux/hash.h> or fs/namei.c. 713468a9428SGeorge Spelvin 714666047feSFinn Thainconfig HAVE_ARCH_NVRAM_OPS 715666047feSFinn Thain bool 716666047feSFinn Thain 7173a495511SWilliam Breathitt Grayconfig ISA_BUS_API 7183a495511SWilliam Breathitt Gray def_bool ISA 7193a495511SWilliam Breathitt Gray 720d2125043SAl Viro# 721d2125043SAl Viro# ABI hall of shame 722d2125043SAl Viro# 723d2125043SAl Viroconfig CLONE_BACKWARDS 724d2125043SAl Viro bool 725d2125043SAl Viro help 726d2125043SAl Viro Architecture has tls passed as the 4th argument of clone(2), 727d2125043SAl Viro not the 5th one. 728d2125043SAl Viro 729d2125043SAl Viroconfig CLONE_BACKWARDS2 730d2125043SAl Viro bool 731d2125043SAl Viro help 732d2125043SAl Viro Architecture has the first two arguments of clone(2) swapped. 733d2125043SAl Viro 734dfa9771aSMichal Simekconfig CLONE_BACKWARDS3 735dfa9771aSMichal Simek bool 736dfa9771aSMichal Simek help 737dfa9771aSMichal Simek Architecture has tls passed as the 3rd argument of clone(2), 738dfa9771aSMichal Simek not the 5th one. 739dfa9771aSMichal Simek 740eaca6eaeSAl Viroconfig ODD_RT_SIGACTION 741eaca6eaeSAl Viro bool 742eaca6eaeSAl Viro help 743eaca6eaeSAl Viro Architecture has unusual rt_sigaction(2) arguments 744eaca6eaeSAl Viro 7450a0e8cdfSAl Viroconfig OLD_SIGSUSPEND 7460a0e8cdfSAl Viro bool 7470a0e8cdfSAl Viro help 7480a0e8cdfSAl Viro Architecture has old sigsuspend(2) syscall, of one-argument variety 7490a0e8cdfSAl Viro 7500a0e8cdfSAl Viroconfig OLD_SIGSUSPEND3 7510a0e8cdfSAl Viro bool 7520a0e8cdfSAl Viro help 7530a0e8cdfSAl Viro Even weirder antique ABI - three-argument sigsuspend(2) 7540a0e8cdfSAl Viro 755495dfbf7SAl Viroconfig OLD_SIGACTION 756495dfbf7SAl Viro bool 757495dfbf7SAl Viro help 758495dfbf7SAl Viro Architecture has old sigaction(2) syscall. Nope, not the same 759495dfbf7SAl Viro as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2), 760495dfbf7SAl Viro but fairly different variant of sigaction(2), thanks to OSF/1 761495dfbf7SAl Viro compatibility... 762495dfbf7SAl Viro 763495dfbf7SAl Viroconfig COMPAT_OLD_SIGACTION 764495dfbf7SAl Viro bool 765495dfbf7SAl Viro 766d4703ddaSDeepa Dinamaniconfig 64BIT_TIME 767d4703ddaSDeepa Dinamani def_bool ARCH_HAS_64BIT_TIME 768d4703ddaSDeepa Dinamani help 769d4703ddaSDeepa Dinamani This should be selected by all architectures that need to support 770d4703ddaSDeepa Dinamani new system calls with a 64-bit time_t. This is relevant on all 32-bit 771d4703ddaSDeepa Dinamani architectures, and 64-bit architectures as part of compat syscall 772d4703ddaSDeepa Dinamani handling. 773d4703ddaSDeepa Dinamani 77417435e5fSDeepa Dinamaniconfig COMPAT_32BIT_TIME 77500bf25d6SArnd Bergmann def_bool !64BIT || COMPAT 77617435e5fSDeepa Dinamani help 77717435e5fSDeepa Dinamani This enables 32 bit time_t support in addition to 64 bit time_t support. 77817435e5fSDeepa Dinamani This is relevant on all 32-bit architectures, and 64-bit architectures 77917435e5fSDeepa Dinamani as part of compat syscall handling. 78017435e5fSDeepa Dinamani 7810d4a619bSChristoph Hellwigconfig ARCH_NO_COHERENT_DMA_MMAP 7820d4a619bSChristoph Hellwig bool 7830d4a619bSChristoph Hellwig 78487a4c375SChristoph Hellwigconfig ARCH_NO_PREEMPT 78587a4c375SChristoph Hellwig bool 78687a4c375SChristoph Hellwig 787fff7fb0bSZhaoxiu Zengconfig CPU_NO_EFFICIENT_FFS 788fff7fb0bSZhaoxiu Zeng def_bool n 789fff7fb0bSZhaoxiu Zeng 790ba14a194SAndy Lutomirskiconfig HAVE_ARCH_VMAP_STACK 791ba14a194SAndy Lutomirski def_bool n 792ba14a194SAndy Lutomirski help 793ba14a194SAndy Lutomirski An arch should select this symbol if it can support kernel stacks 794ba14a194SAndy Lutomirski in vmalloc space. This means: 795ba14a194SAndy Lutomirski 796ba14a194SAndy Lutomirski - vmalloc space must be large enough to hold many kernel stacks. 797ba14a194SAndy Lutomirski This may rule out many 32-bit architectures. 798ba14a194SAndy Lutomirski 799ba14a194SAndy Lutomirski - Stacks in vmalloc space need to work reliably. For example, if 800ba14a194SAndy Lutomirski vmap page tables are created on demand, either this mechanism 801ba14a194SAndy Lutomirski needs to work while the stack points to a virtual address with 802ba14a194SAndy Lutomirski unpopulated page tables or arch code (switch_to() and switch_mm(), 803ba14a194SAndy Lutomirski most likely) needs to ensure that the stack's page table entries 804ba14a194SAndy Lutomirski are populated before running on a possibly unpopulated stack. 805ba14a194SAndy Lutomirski 806ba14a194SAndy Lutomirski - If the stack overflows into a guard page, something reasonable 807ba14a194SAndy Lutomirski should happen. The definition of "reasonable" is flexible, but 808ba14a194SAndy Lutomirski instantly rebooting without logging anything would be unfriendly. 809ba14a194SAndy Lutomirski 810ba14a194SAndy Lutomirskiconfig VMAP_STACK 811ba14a194SAndy Lutomirski default y 812ba14a194SAndy Lutomirski bool "Use a virtually-mapped stack" 813ba14a194SAndy Lutomirski depends on HAVE_ARCH_VMAP_STACK && !KASAN 814ba14a194SAndy Lutomirski ---help--- 815ba14a194SAndy Lutomirski Enable this if you want the use virtually-mapped kernel stacks 816ba14a194SAndy Lutomirski with guard pages. This causes kernel stack overflows to be 817ba14a194SAndy Lutomirski caught immediately rather than causing difficult-to-diagnose 818ba14a194SAndy Lutomirski corruption. 819ba14a194SAndy Lutomirski 820ba14a194SAndy Lutomirski This is presently incompatible with KASAN because KASAN expects 821ba14a194SAndy Lutomirski the stack to map directly to the KASAN shadow map using a formula 822ba14a194SAndy Lutomirski that is incorrect if the stack is in vmalloc space. 823ba14a194SAndy Lutomirski 824ad21fc4fSLaura Abbottconfig ARCH_OPTIONAL_KERNEL_RWX 825ad21fc4fSLaura Abbott def_bool n 826ad21fc4fSLaura Abbott 827ad21fc4fSLaura Abbottconfig ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 828ad21fc4fSLaura Abbott def_bool n 829ad21fc4fSLaura Abbott 830ad21fc4fSLaura Abbottconfig ARCH_HAS_STRICT_KERNEL_RWX 831ad21fc4fSLaura Abbott def_bool n 832ad21fc4fSLaura Abbott 8330f5bf6d0SLaura Abbottconfig STRICT_KERNEL_RWX 834ad21fc4fSLaura Abbott bool "Make kernel text and rodata read-only" if ARCH_OPTIONAL_KERNEL_RWX 835ad21fc4fSLaura Abbott depends on ARCH_HAS_STRICT_KERNEL_RWX 836ad21fc4fSLaura Abbott default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 837ad21fc4fSLaura Abbott help 838ad21fc4fSLaura Abbott If this is set, kernel text and rodata memory will be made read-only, 839ad21fc4fSLaura Abbott and non-text memory will be made non-executable. This provides 840ad21fc4fSLaura Abbott protection against certain security exploits (e.g. executing the heap 841ad21fc4fSLaura Abbott or modifying text) 842ad21fc4fSLaura Abbott 843ad21fc4fSLaura Abbott These features are considered standard security practice these days. 844ad21fc4fSLaura Abbott You should say Y here in almost all cases. 845ad21fc4fSLaura Abbott 846ad21fc4fSLaura Abbottconfig ARCH_HAS_STRICT_MODULE_RWX 847ad21fc4fSLaura Abbott def_bool n 848ad21fc4fSLaura Abbott 8490f5bf6d0SLaura Abbottconfig STRICT_MODULE_RWX 850ad21fc4fSLaura Abbott bool "Set loadable kernel module data as NX and text as RO" if ARCH_OPTIONAL_KERNEL_RWX 851ad21fc4fSLaura Abbott depends on ARCH_HAS_STRICT_MODULE_RWX && MODULES 852ad21fc4fSLaura Abbott default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT 853ad21fc4fSLaura Abbott help 854ad21fc4fSLaura Abbott If this is set, module text and rodata memory will be made read-only, 855ad21fc4fSLaura Abbott and non-text memory will be made non-executable. This provides 856ad21fc4fSLaura Abbott protection against certain security exploits (e.g. writing to text) 857ad21fc4fSLaura Abbott 858ea8c64acSChristoph Hellwig# select if the architecture provides an asm/dma-direct.h header 859ea8c64acSChristoph Hellwigconfig ARCH_HAS_PHYS_TO_DMA 860ea8c64acSChristoph Hellwig bool 861ea8c64acSChristoph Hellwig 8627a46ec0eSKees Cookconfig ARCH_HAS_REFCOUNT 8637a46ec0eSKees Cook bool 8647a46ec0eSKees Cook help 8657a46ec0eSKees Cook An architecture selects this when it has implemented refcount_t 8667a46ec0eSKees Cook using open coded assembly primitives that provide an optimized 8677a46ec0eSKees Cook refcount_t implementation, possibly at the expense of some full 8687a46ec0eSKees Cook refcount state checks of CONFIG_REFCOUNT_FULL=y. 8697a46ec0eSKees Cook 8707a46ec0eSKees Cook The refcount overflow check behavior, however, must be retained. 8717a46ec0eSKees Cook Catching overflows is the primary security concern for protecting 8727a46ec0eSKees Cook against bugs in reference counts. 8737a46ec0eSKees Cook 874fd25d19fSKees Cookconfig REFCOUNT_FULL 875fd25d19fSKees Cook bool "Perform full reference count validation at the expense of speed" 876fd25d19fSKees Cook help 877fd25d19fSKees Cook Enabling this switches the refcounting infrastructure from a fast 878fd25d19fSKees Cook unchecked atomic_t implementation to a fully state checked 879fd25d19fSKees Cook implementation, which can be (slightly) slower but provides protections 880fd25d19fSKees Cook against various use-after-free conditions that can be used in 881fd25d19fSKees Cook security flaw exploits. 882fd25d19fSKees Cook 88304f264d3SPaul Burtonconfig HAVE_ARCH_COMPILER_H 88404f264d3SPaul Burton bool 88504f264d3SPaul Burton help 88604f264d3SPaul Burton An architecture can select this if it provides an 88704f264d3SPaul Burton asm/compiler.h header that should be included after 88804f264d3SPaul Burton linux/compiler-*.h in order to override macro definitions that those 88904f264d3SPaul Burton headers generally provide. 89004f264d3SPaul Burton 891271ca788SArd Biesheuvelconfig HAVE_ARCH_PREL32_RELOCATIONS 892271ca788SArd Biesheuvel bool 893271ca788SArd Biesheuvel help 894271ca788SArd Biesheuvel May be selected by an architecture if it supports place-relative 895271ca788SArd Biesheuvel 32-bit relocations, both in the toolchain and in the module loader, 896271ca788SArd Biesheuvel in which case relative references can be used in special sections 897271ca788SArd Biesheuvel for PCI fixup, initcalls etc which are only half the size on 64 bit 898271ca788SArd Biesheuvel architectures, and don't require runtime relocation on relocatable 899271ca788SArd Biesheuvel kernels. 900271ca788SArd Biesheuvel 901ce9084baSArd Biesheuvelconfig ARCH_USE_MEMREMAP_PROT 902ce9084baSArd Biesheuvel bool 903ce9084baSArd Biesheuvel 904fb346fd9SWaiman Longconfig LOCK_EVENT_COUNTS 905fb346fd9SWaiman Long bool "Locking event counts collection" 906fb346fd9SWaiman Long depends on DEBUG_FS 907fb346fd9SWaiman Long depends on QUEUED_SPINLOCKS 908fb346fd9SWaiman Long ---help--- 909fb346fd9SWaiman Long Enable light-weight counting of various locking related events 910fb346fd9SWaiman Long in the system with minimal performance impact. This reduces 911fb346fd9SWaiman Long the chance of application behavior change because of timing 912fb346fd9SWaiman Long differences. The counts are reported via debugfs. 913fb346fd9SWaiman Long 9142521f2c2SPeter Oberparleitersource "kernel/gcov/Kconfig" 91545332b1bSMasahiro Yamada 91645332b1bSMasahiro Yamadasource "scripts/gcc-plugins/Kconfig" 917fa1b5d09SLinus Torvalds 91822471e13SRandy Dunlapendmenu 919