2 # General architecture dependent options
6 tristate "OProfile system profiling"
8 depends on HAVE_OPROFILE
10 select RING_BUFFER_ALLOW_SWAP
12 OProfile is a profiling system capable of profiling the
13 whole system, include the kernel, kernel modules, libraries,
18 config OPROFILE_EVENT_MULTIPLEX
19 bool "OProfile multiplexing support (EXPERIMENTAL)"
21 depends on OPROFILE && X86
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.
33 config OPROFILE_NMI_TIMER
35 depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI
40 depends on HAVE_KPROBES
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.
50 bool "Optimize very unlikely/likely branches"
51 depends on HAVE_ARCH_JUMP_LABEL
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.
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.
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.
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.
71 ( On 32-bit x86, the necessary options added to the compiler
72 flags may increase the size of the kernel slightly. )
76 depends on KPROBES && HAVE_OPTPROBES
79 config KPROBES_ON_FTRACE
81 depends on KPROBES && HAVE_KPROBES_ON_FTRACE
82 depends on DYNAMIC_FTRACE_WITH_REGS
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.
89 bool "Transparent user-space probes (EXPERIMENTAL)"
90 depends on UPROBE_EVENT && PERF_EVENTS
94 Uprobes is the user-space counterpart to kprobes: they
95 enable instrumentation applications (such as 'perf probe')
96 to establish unintrusive probes in user-space binaries and
97 libraries, by executing handler functions when the probes
98 are hit by user-space applications.
100 ( These probes come in the form of single-byte breakpoints,
101 managed by the kernel and kept transparent to the probed
104 If in doubt, say "N".
106 config HAVE_64BIT_ALIGNED_ACCESS
107 def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS
109 Some architectures require 64 bit accesses to be 64 bit
110 aligned, which also requires structs containing 64 bit values
111 to be 64 bit aligned too. This includes some 32 bit
112 architectures which can do 64 bit accesses, as well as 64 bit
113 architectures without unaligned access.
115 This symbol should be selected by an architecture if 64 bit
116 accesses are required to be 64 bit aligned in this way even
117 though it is not a 64 bit architecture.
119 See Documentation/unaligned-memory-access.txt for more
120 information on the topic of unaligned memory accesses.
122 config HAVE_EFFICIENT_UNALIGNED_ACCESS
125 Some architectures are unable to perform unaligned accesses
126 without the use of get_unaligned/put_unaligned. Others are
127 unable to perform such accesses efficiently (e.g. trap on
128 unaligned access and require fixing it up in the exception
131 This symbol should be selected by an architecture if it can
132 perform unaligned accesses efficiently to allow different
133 code paths to be selected for these cases. Some network
134 drivers, for example, could opt to not fix up alignment
135 problems with received packets if doing so would not help
138 See Documentation/unaligned-memory-access.txt for more
139 information on the topic of unaligned memory accesses.
141 config ARCH_USE_BUILTIN_BSWAP
144 Modern versions of GCC (since 4.4) have builtin functions
145 for handling byte-swapping. Using these, instead of the old
146 inline assembler that the architecture code provides in the
147 __arch_bswapXX() macros, allows the compiler to see what's
148 happening and offers more opportunity for optimisation. In
149 particular, the compiler will be able to combine the byteswap
150 with a nearby load or store and use load-and-swap or
151 store-and-swap instructions if the architecture has them. It
152 should almost *never* result in code which is worse than the
153 hand-coded assembler in <asm/swab.h>. But just in case it
154 does, the use of the builtins is optional.
156 Any architecture with load-and-swap or store-and-swap
157 instructions should set this. And it shouldn't hurt to set it
158 on architectures that don't have such instructions.
162 depends on KPROBES && HAVE_KRETPROBES
164 config USER_RETURN_NOTIFIER
166 depends on HAVE_USER_RETURN_NOTIFIER
168 Provide a kernel-internal notification when a cpu is about to
171 config HAVE_IOREMAP_PROT
177 config HAVE_KRETPROBES
180 config HAVE_OPTPROBES
183 config HAVE_KPROBES_ON_FTRACE
186 config HAVE_NMI_WATCHDOG
189 # An arch should select this if it provides all these things:
191 # task_pt_regs() in asm/processor.h or asm/ptrace.h
192 # arch_has_single_step() if there is hardware single-step support
193 # arch_has_block_step() if there is hardware block-step support
194 # asm/syscall.h supplying asm-generic/syscall.h interface
195 # linux/regset.h user_regset interfaces
196 # CORE_DUMP_USE_REGSET #define'd in linux/elf.h
197 # TIF_SYSCALL_TRACE calls tracehook_report_syscall_{entry,exit}
198 # TIF_NOTIFY_RESUME calls tracehook_notify_resume()
199 # signal delivery calls tracehook_signal_handler()
201 config HAVE_ARCH_TRACEHOOK
204 config HAVE_DMA_ATTRS
207 config HAVE_DMA_CONTIGUOUS
210 config USE_GENERIC_SMP_HELPERS
213 config GENERIC_SMP_IDLE_THREAD
216 config GENERIC_IDLE_POLL_SETUP
219 # Select if arch init_task initializer is different to init/init_task.c
220 config ARCH_INIT_TASK
223 # Select if arch has its private alloc_task_struct() function
224 config ARCH_TASK_STRUCT_ALLOCATOR
227 # Select if arch has its private alloc_thread_info() function
228 config ARCH_THREAD_INFO_ALLOCATOR
231 config HAVE_REGS_AND_STACK_ACCESS_API
234 This symbol should be selected by an architecure if it supports
235 the API needed to access registers and stack entries from pt_regs,
236 declared in asm/ptrace.h
237 For example the kprobes-based event tracer needs this API.
242 The <linux/clk.h> calls support software clock gating and
243 thus are a key power management tool on many systems.
245 config HAVE_DMA_API_DEBUG
248 config HAVE_HW_BREAKPOINT
250 depends on PERF_EVENTS
252 config HAVE_MIXED_BREAKPOINTS_REGS
254 depends on HAVE_HW_BREAKPOINT
256 Depending on the arch implementation of hardware breakpoints,
257 some of them have separate registers for data and instruction
258 breakpoints addresses, others have mixed registers to store
259 them but define the access type in a control register.
260 Select this option if your arch implements breakpoints under the
263 config HAVE_USER_RETURN_NOTIFIER
266 config HAVE_PERF_EVENTS_NMI
269 System hardware can generate an NMI using the perf event
270 subsystem. Also has support for calculating CPU cycle events
271 to determine how many clock cycles in a given period.
273 config HAVE_PERF_REGS
276 Support selective register dumps for perf events. This includes
277 bit-mapping of each registers and a unique architecture id.
279 config HAVE_PERF_USER_STACK_DUMP
282 Support user stack dumps for perf event samples. This needs
283 access to the user stack pointer which is not unified across
286 config HAVE_ARCH_JUMP_LABEL
289 config HAVE_RCU_TABLE_FREE
292 config ARCH_HAVE_NMI_SAFE_CMPXCHG
295 config HAVE_ALIGNED_STRUCT_PAGE
298 This makes sure that struct pages are double word aligned and that
299 e.g. the SLUB allocator can perform double word atomic operations
300 on a struct page for better performance. However selecting this
301 might increase the size of a struct page by a word.
303 config HAVE_CMPXCHG_LOCAL
306 config HAVE_CMPXCHG_DOUBLE
309 config ARCH_WANT_IPC_PARSE_VERSION
312 config ARCH_WANT_COMPAT_IPC_PARSE_VERSION
315 config ARCH_WANT_OLD_COMPAT_IPC
316 select ARCH_WANT_COMPAT_IPC_PARSE_VERSION
319 config HAVE_ARCH_SECCOMP_FILTER
322 An arch should select this symbol if it provides all of these things:
324 - syscall_get_arguments()
326 - syscall_set_return_value()
327 - SIGSYS siginfo_t support
328 - secure_computing is called from a ptrace_event()-safe context
329 - secure_computing return value is checked and a return value of -1
330 results in the system call being skipped immediately.
332 config SECCOMP_FILTER
334 depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET
336 Enable tasks to build secure computing environments defined
337 in terms of Berkeley Packet Filter programs which implement
338 task-defined system call filtering polices.
340 See Documentation/prctl/seccomp_filter.txt for details.
342 config HAVE_CONTEXT_TRACKING
345 Provide kernel/user boundaries probes necessary for subsystems
346 that need it, such as userspace RCU extended quiescent state.
347 Syscalls need to be wrapped inside user_exit()-user_enter() through
348 the slow path using TIF_NOHZ flag. Exceptions handlers must be
349 wrapped as well. Irqs are already protected inside
350 rcu_irq_enter/rcu_irq_exit() but preemption or signal handling on
351 irq exit still need to be protected.
353 config HAVE_VIRT_CPU_ACCOUNTING
356 config HAVE_VIRT_CPU_ACCOUNTING_GEN
360 With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit.
361 Before enabling this option, arch code must be audited
362 to ensure there are no races in concurrent read/write of
363 cputime_t. For example, reading/writing 64-bit cputime_t on
364 some 32-bit arches may require multiple accesses, so proper
365 locking is needed to protect against concurrent accesses.
368 config HAVE_IRQ_TIME_ACCOUNTING
371 Archs need to ensure they use a high enough resolution clock to
372 support irq time accounting and then call enable_sched_clock_irqtime().
374 config HAVE_ARCH_TRANSPARENT_HUGEPAGE
377 config HAVE_ARCH_SOFT_DIRTY
380 config HAVE_MOD_ARCH_SPECIFIC
383 The arch uses struct mod_arch_specific to store data. Many arches
384 just need a simple module loader without arch specific data - those
385 should not enable this.
387 config MODULES_USE_ELF_RELA
390 Modules only use ELF RELA relocations. Modules with ELF REL
391 relocations will give an error.
393 config MODULES_USE_ELF_REL
396 Modules only use ELF REL relocations. Modules with ELF RELA
397 relocations will give an error.
399 config HAVE_UNDERSCORE_SYMBOL_PREFIX
402 Some architectures generate an _ in front of C symbols; things like
403 module loading and assembly files need to know about this.
405 config HAVE_IRQ_EXIT_ON_IRQ_STACK
408 Architecture doesn't only execute the irq handler on the irq stack
409 but also irq_exit(). This way we can process softirqs on this irq
410 stack instead of switching to a new one when we call __do_softirq()
411 in the end of an hardirq.
412 This spares a stack switch and improves cache usage on softirq
418 config CLONE_BACKWARDS
421 Architecture has tls passed as the 4th argument of clone(2),
424 config CLONE_BACKWARDS2
427 Architecture has the first two arguments of clone(2) swapped.
429 config CLONE_BACKWARDS3
432 Architecture has tls passed as the 3rd argument of clone(2),
435 config ODD_RT_SIGACTION
438 Architecture has unusual rt_sigaction(2) arguments
440 config OLD_SIGSUSPEND
443 Architecture has old sigsuspend(2) syscall, of one-argument variety
445 config OLD_SIGSUSPEND3
448 Even weirder antique ABI - three-argument sigsuspend(2)
453 Architecture has old sigaction(2) syscall. Nope, not the same
454 as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2),
455 but fairly different variant of sigaction(2), thanks to OSF/1
458 config COMPAT_OLD_SIGACTION
461 source "kernel/gcov/Kconfig"