arch/Kconfig

   1 #
   2 # General architecture dependent options
   3 #
   4
   5 config OPROFILE
   6         tristate "OProfile system profiling"
   7         depends on PROFILING
   8         depends on HAVE_OPROFILE
   9         select RING_BUFFER
  10         select RING_BUFFER_ALLOW_SWAP
  11         help
  12           OProfile is a profiling system capable of profiling the
  13           whole system, include the kernel, kernel modules, libraries,
  14           and applications.
  15
  16           If unsure, say N.
  17
  18 config OPROFILE_EVENT_MULTIPLEX
  19         bool "OProfile multiplexing support (EXPERIMENTAL)"
  20         default n
  21         depends on OPROFILE && X86
  22         help
  23           The number of hardware counters is limited. The multiplexing
  24           feature enables OProfile to gather more events than counters
  25           are provided by the hardware. This is realized by switching
  26           between events at an user specified time interval.
  27
  28           If unsure, say N.
  29
  30 config HAVE_OPROFILE
  31         bool
  32
  33 config OPROFILE_NMI_TIMER
  34         def_bool y
  35         depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI
  36
  37 config KPROBES
  38         bool "Kprobes"
  39         depends on MODULES
  40         depends on HAVE_KPROBES
  41         select KALLSYMS
  42         help
  43           Kprobes allows you to trap at almost any kernel address and
  44           execute a callback function.  register_kprobe() establishes
  45           a probepoint and specifies the callback.  Kprobes is useful
  46           for kernel debugging, non-intrusive instrumentation and testing.
  47           If in doubt, say "N".
  48
  49 config JUMP_LABEL
  50        bool "Optimize very unlikely/likely branches"
  51        depends on HAVE_ARCH_JUMP_LABEL
  52        help
  53          This option enables a transparent branch optimization that
  54          makes certain almost-always-true or almost-always-false branch
  55          conditions even cheaper to execute within the kernel.
  56
  57          Certain performance-sensitive kernel code, such as trace points,
  58          scheduler functionality, networking code and KVM have such
  59          branches and include support for this optimization technique.
  60
  61          If it is detected that the compiler has support for "asm goto",
  62          the kernel will compile such branches with just a nop
  63          instruction. When the condition flag is toggled to true, the
  64          nop will be converted to a jump instruction to execute the
  65          conditional block of instructions.
  66
  67          This technique lowers overhead and stress on the branch prediction
  68          of the processor and generally makes the kernel faster. The update
  69          of the condition is slower, but those are always very rare.
  70
  71          ( On 32-bit x86, the necessary options added to the compiler
  72            flags may increase the size of the kernel slightly. )
  73
  74 config OPTPROBES
  75         def_bool y
  76         depends on KPROBES && HAVE_OPTPROBES
  77         depends on !PREEMPT
  78
  79 config KPROBES_ON_FTRACE
  80         def_bool y
  81         depends on KPROBES && HAVE_KPROBES_ON_FTRACE
  82         depends on DYNAMIC_FTRACE_WITH_REGS
  83         help
  84          If function tracer is enabled and the arch supports full
  85          passing of pt_regs to function tracing, then kprobes can
  86          optimize on top of function tracing.
  87
  88 config UPROBES
  89         def_bool n
  90         select PERCPU_RWSEM
  91         help
  92           Uprobes is the user-space counterpart to kprobes: they
  93           enable instrumentation applications (such as 'perf probe')
  94           to establish unintrusive probes in user-space binaries and
  95           libraries, by executing handler functions when the probes
  96           are hit by user-space applications.
  97
  98           ( These probes come in the form of single-byte breakpoints,
  99             managed by the kernel and kept transparent to the probed
 100             application. )
 101
 102 config HAVE_64BIT_ALIGNED_ACCESS
 103         def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS
 104         help
 105           Some architectures require 64 bit accesses to be 64 bit
 106           aligned, which also requires structs containing 64 bit values
 107           to be 64 bit aligned too. This includes some 32 bit
 108           architectures which can do 64 bit accesses, as well as 64 bit
 109           architectures without unaligned access.
 110
 111           This symbol should be selected by an architecture if 64 bit
 112           accesses are required to be 64 bit aligned in this way even
 113           though it is not a 64 bit architecture.
 114
 115           See Documentation/unaligned-memory-access.txt for more
 116           information on the topic of unaligned memory accesses.
 117
 118 config HAVE_EFFICIENT_UNALIGNED_ACCESS
 119         bool
 120         help
 121           Some architectures are unable to perform unaligned accesses
 122           without the use of get_unaligned/put_unaligned. Others are
 123           unable to perform such accesses efficiently (e.g. trap on
 124           unaligned access and require fixing it up in the exception
 125           handler.)
 126
 127           This symbol should be selected by an architecture if it can
 128           perform unaligned accesses efficiently to allow different
 129           code paths to be selected for these cases. Some network
 130           drivers, for example, could opt to not fix up alignment
 131           problems with received packets if doing so would not help
 132           much.
 133
 134           See Documentation/unaligned-memory-access.txt for more
 135           information on the topic of unaligned memory accesses.
 136
 137 config ARCH_USE_BUILTIN_BSWAP
 138        bool
 139        help
 140          Modern versions of GCC (since 4.4) have builtin functions
 141          for handling byte-swapping. Using these, instead of the old
 142          inline assembler that the architecture code provides in the
 143          __arch_bswapXX() macros, allows the compiler to see what's
 144          happening and offers more opportunity for optimisation. In
 145          particular, the compiler will be able to combine the byteswap
 146          with a nearby load or store and use load-and-swap or
 147          store-and-swap instructions if the architecture has them. It
 148          should almost *never* result in code which is worse than the
 149          hand-coded assembler in <asm/swab.h>.  But just in case it
 150          does, the use of the builtins is optional.
 151
 152          Any architecture with load-and-swap or store-and-swap
 153          instructions should set this. And it shouldn't hurt to set it
 154          on architectures that don't have such instructions.
 155
 156 config KRETPROBES
 157         def_bool y
 158         depends on KPROBES && HAVE_KRETPROBES
 159
 160 config USER_RETURN_NOTIFIER
 161         bool
 162         depends on HAVE_USER_RETURN_NOTIFIER
 163         help
 164           Provide a kernel-internal notification when a cpu is about to
 165           switch to user mode.
 166
 167 config HAVE_IOREMAP_PROT
 168         bool
 169
 170 config HAVE_KPROBES
 171         bool
 172
 173 config HAVE_KRETPROBES
 174         bool
 175
 176 config HAVE_OPTPROBES
 177         bool
 178
 179 config HAVE_KPROBES_ON_FTRACE
 180         bool
 181
 182 config HAVE_NMI_WATCHDOG
 183         bool
 184 #
 185 # An arch should select this if it provides all these things:
 186 #
 187 #       task_pt_regs()          in asm/processor.h or asm/ptrace.h
 188 #       arch_has_single_step()  if there is hardware single-step support
 189 #       arch_has_block_step()   if there is hardware block-step support
 190 #       asm/syscall.h           supplying asm-generic/syscall.h interface
 191 #       linux/regset.h          user_regset interfaces
 192 #       CORE_DUMP_USE_REGSET    #define'd in linux/elf.h
 193 #       TIF_SYSCALL_TRACE       calls tracehook_report_syscall_{entry,exit}
 194 #       TIF_NOTIFY_RESUME       calls tracehook_notify_resume()
 195 #       signal delivery         calls tracehook_signal_handler()
 196 #
 197 config HAVE_ARCH_TRACEHOOK
 198         bool
 199
 200 config HAVE_DMA_ATTRS
 201         bool
 202
 203 config HAVE_DMA_CONTIGUOUS
 204         bool
 205
 206 config GENERIC_SMP_IDLE_THREAD
 207        bool
 208
 209 config GENERIC_IDLE_POLL_SETUP
 210        bool
 211
 212 # Select if arch init_task initializer is different to init/init_task.c
 213 config ARCH_INIT_TASK
 214        bool
 215
 216 # Select if arch has its private alloc_task_struct() function
 217 config ARCH_TASK_STRUCT_ALLOCATOR
 218         bool
 219
 220 # Select if arch has its private alloc_thread_info() function
 221 config ARCH_THREAD_INFO_ALLOCATOR
 222         bool
 223
 224 config HAVE_REGS_AND_STACK_ACCESS_API
 225         bool
 226         help
 227           This symbol should be selected by an architecure if it supports
 228           the API needed to access registers and stack entries from pt_regs,
 229           declared in asm/ptrace.h
 230           For example the kprobes-based event tracer needs this API.
 231
 232 config HAVE_CLK
 233         bool
 234         help
 235           The <linux/clk.h> calls support software clock gating and
 236           thus are a key power management tool on many systems.
 237
 238 config HAVE_DMA_API_DEBUG
 239         bool
 240
 241 config HAVE_HW_BREAKPOINT
 242         bool
 243         depends on PERF_EVENTS
 244
 245 config HAVE_MIXED_BREAKPOINTS_REGS
 246         bool
 247         depends on HAVE_HW_BREAKPOINT
 248         help
 249           Depending on the arch implementation of hardware breakpoints,
 250           some of them have separate registers for data and instruction
 251           breakpoints addresses, others have mixed registers to store
 252           them but define the access type in a control register.
 253           Select this option if your arch implements breakpoints under the
 254           latter fashion.
 255
 256 config HAVE_USER_RETURN_NOTIFIER
 257         bool
 258
 259 config HAVE_PERF_EVENTS_NMI
 260         bool
 261         help
 262           System hardware can generate an NMI using the perf event
 263           subsystem.  Also has support for calculating CPU cycle events
 264           to determine how many clock cycles in a given period.
 265
 266 config HAVE_PERF_REGS
 267         bool
 268         help
 269           Support selective register dumps for perf events. This includes
 270           bit-mapping of each registers and a unique architecture id.
 271
 272 config HAVE_PERF_USER_STACK_DUMP
 273         bool
 274         help
 275           Support user stack dumps for perf event samples. This needs
 276           access to the user stack pointer which is not unified across
 277           architectures.
 278
 279 config HAVE_ARCH_JUMP_LABEL
 280         bool
 281
 282 config HAVE_RCU_TABLE_FREE
 283         bool
 284
 285 config ARCH_HAVE_NMI_SAFE_CMPXCHG
 286         bool
 287
 288 config HAVE_ALIGNED_STRUCT_PAGE
 289         bool
 290         help
 291           This makes sure that struct pages are double word aligned and that
 292           e.g. the SLUB allocator can perform double word atomic operations
 293           on a struct page for better performance. However selecting this
 294           might increase the size of a struct page by a word.
 295
 296 config HAVE_CMPXCHG_LOCAL
 297         bool
 298
 299 config HAVE_CMPXCHG_DOUBLE
 300         bool
 301
 302 config ARCH_WANT_IPC_PARSE_VERSION
 303         bool
 304
 305 config ARCH_WANT_COMPAT_IPC_PARSE_VERSION
 306         bool
 307
 308 config ARCH_WANT_OLD_COMPAT_IPC
 309         select ARCH_WANT_COMPAT_IPC_PARSE_VERSION
 310         bool
 311
 312 config HAVE_ARCH_SECCOMP_FILTER
 313         bool
 314         help
 315           An arch should select this symbol if it provides all of these things:
 316           - syscall_get_arch()
 317           - syscall_get_arguments()
 318           - syscall_rollback()
 319           - syscall_set_return_value()
 320           - SIGSYS siginfo_t support
 321           - secure_computing is called from a ptrace_event()-safe context
 322           - secure_computing return value is checked and a return value of -1
 323             results in the system call being skipped immediately.
 324           - seccomp syscall wired up
 325
 326           For best performance, an arch should use seccomp_phase1 and
 327           seccomp_phase2 directly.  It should call seccomp_phase1 for all
 328           syscalls if TIF_SECCOMP is set, but seccomp_phase1 does not
 329           need to be called from a ptrace-safe context.  It must then
 330           call seccomp_phase2 if seccomp_phase1 returns anything other
 331           than SECCOMP_PHASE1_OK or SECCOMP_PHASE1_SKIP.
 332
 333           As an additional optimization, an arch may provide seccomp_data
 334           directly to seccomp_phase1; this avoids multiple calls
 335           to the syscall_xyz helpers for every syscall.
 336
 337 config SECCOMP_FILTER
 338         def_bool y
 339         depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET
 340         help
 341           Enable tasks to build secure computing environments defined
 342           in terms of Berkeley Packet Filter programs which implement
 343           task-defined system call filtering polices.
 344
 345           See Documentation/prctl/seccomp_filter.txt for details.
 346
 347 config HAVE_CC_STACKPROTECTOR
 348         bool
 349         help
 350           An arch should select this symbol if:
 351           - its compiler supports the -fstack-protector option
 352           - it has implemented a stack canary (e.g. __stack_chk_guard)
 353
 354 config CC_STACKPROTECTOR
 355         def_bool n
 356         help
 357           Set when a stack-protector mode is enabled, so that the build
 358           can enable kernel-side support for the GCC feature.
 359
 360 choice
 361         prompt "Stack Protector buffer overflow detection"
 362         depends on HAVE_CC_STACKPROTECTOR
 363         default CC_STACKPROTECTOR_NONE
 364         help
 365           This option turns on the "stack-protector" GCC feature. This
 366           feature puts, at the beginning of functions, a canary value on
 367           the stack just before the return address, and validates
 368           the value just before actually returning.  Stack based buffer
 369           overflows (that need to overwrite this return address) now also
 370           overwrite the canary, which gets detected and the attack is then
 371           neutralized via a kernel panic.
 372
 373 config CC_STACKPROTECTOR_NONE
 374         bool "None"
 375         help
 376           Disable "stack-protector" GCC feature.
 377
 378 config CC_STACKPROTECTOR_REGULAR
 379         bool "Regular"
 380         select CC_STACKPROTECTOR
 381         help
 382           Functions will have the stack-protector canary logic added if they
 383           have an 8-byte or larger character array on the stack.
 384
 385           This feature requires gcc version 4.2 or above, or a distribution
 386           gcc with the feature backported ("-fstack-protector").
 387
 388           On an x86 "defconfig" build, this feature adds canary checks to
 389           about 3% of all kernel functions, which increases kernel code size
 390           by about 0.3%.
 391
 392 config CC_STACKPROTECTOR_STRONG
 393         bool "Strong"
 394         select CC_STACKPROTECTOR
 395         help
 396           Functions will have the stack-protector canary logic added in any
 397           of the following conditions:
 398
 399           - local variable's address used as part of the right hand side of an
 400             assignment or function argument
 401           - local variable is an array (or union containing an array),
 402             regardless of array type or length
 403           - uses register local variables
 404
 405           This feature requires gcc version 4.9 or above, or a distribution
 406           gcc with the feature backported ("-fstack-protector-strong").
 407
 408           On an x86 "defconfig" build, this feature adds canary checks to
 409           about 20% of all kernel functions, which increases the kernel code
 410           size by about 2%.
 411
 412 endchoice
 413
 414 config HAVE_CONTEXT_TRACKING
 415         bool
 416         help
 417           Provide kernel/user boundaries probes necessary for subsystems
 418           that need it, such as userspace RCU extended quiescent state.
 419           Syscalls need to be wrapped inside user_exit()-user_enter() through
 420           the slow path using TIF_NOHZ flag. Exceptions handlers must be
 421           wrapped as well. Irqs are already protected inside
 422           rcu_irq_enter/rcu_irq_exit() but preemption or signal handling on
 423           irq exit still need to be protected.
 424
 425 config HAVE_VIRT_CPU_ACCOUNTING
 426         bool
 427
 428 config HAVE_VIRT_CPU_ACCOUNTING_GEN
 429         bool
 430         default y if 64BIT
 431         help
 432           With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit.
 433           Before enabling this option, arch code must be audited
 434           to ensure there are no races in concurrent read/write of
 435           cputime_t. For example, reading/writing 64-bit cputime_t on
 436           some 32-bit arches may require multiple accesses, so proper
 437           locking is needed to protect against concurrent accesses.
 438
 439
 440 config HAVE_IRQ_TIME_ACCOUNTING
 441         bool
 442         help
 443           Archs need to ensure they use a high enough resolution clock to
 444           support irq time accounting and then call enable_sched_clock_irqtime().
 445
 446 config HAVE_ARCH_TRANSPARENT_HUGEPAGE
 447         bool
 448
 449 config HAVE_ARCH_SOFT_DIRTY
 450         bool
 451
 452 config HAVE_MOD_ARCH_SPECIFIC
 453         bool
 454         help
 455           The arch uses struct mod_arch_specific to store data.  Many arches
 456           just need a simple module loader without arch specific data - those
 457           should not enable this.
 458
 459 config MODULES_USE_ELF_RELA
 460         bool
 461         help
 462           Modules only use ELF RELA relocations.  Modules with ELF REL
 463           relocations will give an error.
 464
 465 config MODULES_USE_ELF_REL
 466         bool
 467         help
 468           Modules only use ELF REL relocations.  Modules with ELF RELA
 469           relocations will give an error.
 470
 471 config HAVE_UNDERSCORE_SYMBOL_PREFIX
 472         bool
 473         help
 474           Some architectures generate an _ in front of C symbols; things like
 475           module loading and assembly files need to know about this.
 476
 477 config HAVE_IRQ_EXIT_ON_IRQ_STACK
 478         bool
 479         help
 480           Architecture doesn't only execute the irq handler on the irq stack
 481           but also irq_exit(). This way we can process softirqs on this irq
 482           stack instead of switching to a new one when we call __do_softirq()
 483           in the end of an hardirq.
 484           This spares a stack switch and improves cache usage on softirq
 485           processing.
 486
 487 #
 488 # ABI hall of shame
 489 #
 490 config CLONE_BACKWARDS
 491         bool
 492         help
 493           Architecture has tls passed as the 4th argument of clone(2),
 494           not the 5th one.
 495
 496 config CLONE_BACKWARDS2
 497         bool
 498         help
 499           Architecture has the first two arguments of clone(2) swapped.
 500
 501 config CLONE_BACKWARDS3
 502         bool
 503         help
 504           Architecture has tls passed as the 3rd argument of clone(2),
 505           not the 5th one.
 506
 507 config ODD_RT_SIGACTION
 508         bool
 509         help
 510           Architecture has unusual rt_sigaction(2) arguments
 511
 512 config OLD_SIGSUSPEND
 513         bool
 514         help
 515           Architecture has old sigsuspend(2) syscall, of one-argument variety
 516
 517 config OLD_SIGSUSPEND3
 518         bool
 519         help
 520           Even weirder antique ABI - three-argument sigsuspend(2)
 521
 522 config OLD_SIGACTION
 523         bool
 524         help
 525           Architecture has old sigaction(2) syscall.  Nope, not the same
 526           as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2),
 527           but fairly different variant of sigaction(2), thanks to OSF/1
 528           compatibility...
 529
 530 config COMPAT_OLD_SIGACTION
 531         bool
 532
 533 source "kernel/gcov/Kconfig"