Root/arch/Kconfig

1#
2# General architecture dependent options
3#
4
5config 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
18config 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
30config HAVE_OPROFILE
31    bool
32
33config OPROFILE_NMI_TIMER
34    def_bool y
35    depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI
36
37config 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
49config 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
74config OPTPROBES
75    def_bool y
76    depends on KPROBES && HAVE_OPTPROBES
77    depends on !PREEMPT
78
79config 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
88config 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
102config 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
118config 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
137config 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
156config KRETPROBES
157    def_bool y
158    depends on KPROBES && HAVE_KRETPROBES
159
160config 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
167config HAVE_IOREMAP_PROT
168    bool
169
170config HAVE_KPROBES
171    bool
172
173config HAVE_KRETPROBES
174    bool
175
176config HAVE_OPTPROBES
177    bool
178
179config HAVE_KPROBES_ON_FTRACE
180    bool
181
182config 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#
197config HAVE_ARCH_TRACEHOOK
198    bool
199
200config HAVE_DMA_ATTRS
201    bool
202
203config HAVE_DMA_CONTIGUOUS
204    bool
205
206config GENERIC_SMP_IDLE_THREAD
207       bool
208
209config GENERIC_IDLE_POLL_SETUP
210       bool
211
212# Select if arch init_task initializer is different to init/init_task.c
213config ARCH_INIT_TASK
214       bool
215
216# Select if arch has its private alloc_task_struct() function
217config ARCH_TASK_STRUCT_ALLOCATOR
218    bool
219
220# Select if arch has its private alloc_thread_info() function
221config ARCH_THREAD_INFO_ALLOCATOR
222    bool
223
224config 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
232config 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
238config HAVE_DMA_API_DEBUG
239    bool
240
241config HAVE_HW_BREAKPOINT
242    bool
243    depends on PERF_EVENTS
244
245config 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
256config HAVE_USER_RETURN_NOTIFIER
257    bool
258
259config 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
266config 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
272config 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
279config HAVE_ARCH_JUMP_LABEL
280    bool
281
282config HAVE_RCU_TABLE_FREE
283    bool
284
285config ARCH_HAVE_NMI_SAFE_CMPXCHG
286    bool
287
288config 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
296config HAVE_CMPXCHG_LOCAL
297    bool
298
299config HAVE_CMPXCHG_DOUBLE
300    bool
301
302config ARCH_WANT_IPC_PARSE_VERSION
303    bool
304
305config ARCH_WANT_COMPAT_IPC_PARSE_VERSION
306    bool
307
308config ARCH_WANT_OLD_COMPAT_IPC
309    select ARCH_WANT_COMPAT_IPC_PARSE_VERSION
310    bool
311
312config 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
337config 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
347config 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
354config 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
360choice
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
373config CC_STACKPROTECTOR_NONE
374    bool "None"
375    help
376      Disable "stack-protector" GCC feature.
377
378config 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
392config 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
412endchoice
413
414config 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
425config HAVE_VIRT_CPU_ACCOUNTING
426    bool
427
428config 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
440config 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
446config HAVE_ARCH_TRANSPARENT_HUGEPAGE
447    bool
448
449config HAVE_ARCH_SOFT_DIRTY
450    bool
451
452config 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
459config 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
465config 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
471config 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
477config 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#
490config CLONE_BACKWARDS
491    bool
492    help
493      Architecture has tls passed as the 4th argument of clone(2),
494      not the 5th one.
495
496config CLONE_BACKWARDS2
497    bool
498    help
499      Architecture has the first two arguments of clone(2) swapped.
500
501config CLONE_BACKWARDS3
502    bool
503    help
504      Architecture has tls passed as the 3rd argument of clone(2),
505      not the 5th one.
506
507config ODD_RT_SIGACTION
508    bool
509    help
510      Architecture has unusual rt_sigaction(2) arguments
511
512config OLD_SIGSUSPEND
513    bool
514    help
515      Architecture has old sigsuspend(2) syscall, of one-argument variety
516
517config OLD_SIGSUSPEND3
518    bool
519    help
520      Even weirder antique ABI - three-argument sigsuspend(2)
521
522config 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
530config COMPAT_OLD_SIGACTION
531    bool
532
533source "kernel/gcov/Kconfig"
534

Archive Download this file



interactive