]>
Commit | Line | Data |
---|---|---|
fb32e03f MD |
1 | # |
2 | # General architecture dependent options | |
3 | # | |
125e5645 | 4 | |
692f66f2 HB |
5 | config CRASH_CORE |
6 | bool | |
7 | ||
2965faa5 | 8 | config KEXEC_CORE |
692f66f2 | 9 | select CRASH_CORE |
2965faa5 DY |
10 | bool |
11 | ||
467d2782 TJB |
12 | config HAVE_IMA_KEXEC |
13 | bool | |
14 | ||
125e5645 | 15 | config OPROFILE |
b309a294 | 16 | tristate "OProfile system profiling" |
125e5645 MD |
17 | depends on PROFILING |
18 | depends on HAVE_OPROFILE | |
d69d59f4 | 19 | select RING_BUFFER |
9a5963eb | 20 | select RING_BUFFER_ALLOW_SWAP |
125e5645 MD |
21 | help |
22 | OProfile is a profiling system capable of profiling the | |
23 | whole system, include the kernel, kernel modules, libraries, | |
24 | and applications. | |
25 | ||
26 | If unsure, say N. | |
27 | ||
4d4036e0 JY |
28 | config OPROFILE_EVENT_MULTIPLEX |
29 | bool "OProfile multiplexing support (EXPERIMENTAL)" | |
30 | default n | |
31 | depends on OPROFILE && X86 | |
32 | help | |
33 | The number of hardware counters is limited. The multiplexing | |
34 | feature enables OProfile to gather more events than counters | |
35 | are provided by the hardware. This is realized by switching | |
9332ef9d | 36 | between events at a user specified time interval. |
4d4036e0 JY |
37 | |
38 | If unsure, say N. | |
39 | ||
125e5645 | 40 | config HAVE_OPROFILE |
9ba16087 | 41 | bool |
125e5645 | 42 | |
dcfce4a0 RR |
43 | config OPROFILE_NMI_TIMER |
44 | def_bool y | |
af9feebe | 45 | depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI && !PPC64 |
dcfce4a0 | 46 | |
125e5645 MD |
47 | config KPROBES |
48 | bool "Kprobes" | |
05ed160e | 49 | depends on MODULES |
125e5645 | 50 | depends on HAVE_KPROBES |
05ed160e | 51 | select KALLSYMS |
125e5645 MD |
52 | help |
53 | Kprobes allows you to trap at almost any kernel address and | |
54 | execute a callback function. register_kprobe() establishes | |
55 | a probepoint and specifies the callback. Kprobes is useful | |
56 | for kernel debugging, non-intrusive instrumentation and testing. | |
57 | If in doubt, say "N". | |
58 | ||
45f81b1c | 59 | config JUMP_LABEL |
c5905afb | 60 | bool "Optimize very unlikely/likely branches" |
45f81b1c SR |
61 | depends on HAVE_ARCH_JUMP_LABEL |
62 | help | |
c5905afb IM |
63 | This option enables a transparent branch optimization that |
64 | makes certain almost-always-true or almost-always-false branch | |
65 | conditions even cheaper to execute within the kernel. | |
66 | ||
67 | Certain performance-sensitive kernel code, such as trace points, | |
68 | scheduler functionality, networking code and KVM have such | |
69 | branches and include support for this optimization technique. | |
70 | ||
45f81b1c | 71 | If it is detected that the compiler has support for "asm goto", |
c5905afb IM |
72 | the kernel will compile such branches with just a nop |
73 | instruction. When the condition flag is toggled to true, the | |
74 | nop will be converted to a jump instruction to execute the | |
75 | conditional block of instructions. | |
76 | ||
77 | This technique lowers overhead and stress on the branch prediction | |
78 | of the processor and generally makes the kernel faster. The update | |
79 | of the condition is slower, but those are always very rare. | |
45f81b1c | 80 | |
c5905afb IM |
81 | ( On 32-bit x86, the necessary options added to the compiler |
82 | flags may increase the size of the kernel slightly. ) | |
45f81b1c | 83 | |
1987c947 PZ |
84 | config STATIC_KEYS_SELFTEST |
85 | bool "Static key selftest" | |
86 | depends on JUMP_LABEL | |
87 | help | |
88 | Boot time self-test of the branch patching code. | |
89 | ||
afd66255 | 90 | config OPTPROBES |
5cc718b9 MH |
91 | def_bool y |
92 | depends on KPROBES && HAVE_OPTPROBES | |
afd66255 | 93 | depends on !PREEMPT |
afd66255 | 94 | |
e7dbfe34 MH |
95 | config KPROBES_ON_FTRACE |
96 | def_bool y | |
97 | depends on KPROBES && HAVE_KPROBES_ON_FTRACE | |
98 | depends on DYNAMIC_FTRACE_WITH_REGS | |
99 | help | |
100 | If function tracer is enabled and the arch supports full | |
101 | passing of pt_regs to function tracing, then kprobes can | |
102 | optimize on top of function tracing. | |
103 | ||
2b144498 | 104 | config UPROBES |
09294e31 | 105 | def_bool n |
e8f4aa60 | 106 | depends on ARCH_SUPPORTS_UPROBES |
2b144498 | 107 | help |
7b2d81d4 IM |
108 | Uprobes is the user-space counterpart to kprobes: they |
109 | enable instrumentation applications (such as 'perf probe') | |
110 | to establish unintrusive probes in user-space binaries and | |
111 | libraries, by executing handler functions when the probes | |
112 | are hit by user-space applications. | |
113 | ||
114 | ( These probes come in the form of single-byte breakpoints, | |
115 | managed by the kernel and kept transparent to the probed | |
116 | application. ) | |
2b144498 | 117 | |
c19fa94a JH |
118 | config HAVE_64BIT_ALIGNED_ACCESS |
119 | def_bool 64BIT && !HAVE_EFFICIENT_UNALIGNED_ACCESS | |
120 | help | |
121 | Some architectures require 64 bit accesses to be 64 bit | |
122 | aligned, which also requires structs containing 64 bit values | |
123 | to be 64 bit aligned too. This includes some 32 bit | |
124 | architectures which can do 64 bit accesses, as well as 64 bit | |
125 | architectures without unaligned access. | |
126 | ||
127 | This symbol should be selected by an architecture if 64 bit | |
128 | accesses are required to be 64 bit aligned in this way even | |
129 | though it is not a 64 bit architecture. | |
130 | ||
131 | See Documentation/unaligned-memory-access.txt for more | |
132 | information on the topic of unaligned memory accesses. | |
133 | ||
58340a07 | 134 | config HAVE_EFFICIENT_UNALIGNED_ACCESS |
9ba16087 | 135 | bool |
58340a07 JB |
136 | help |
137 | Some architectures are unable to perform unaligned accesses | |
138 | without the use of get_unaligned/put_unaligned. Others are | |
139 | unable to perform such accesses efficiently (e.g. trap on | |
140 | unaligned access and require fixing it up in the exception | |
141 | handler.) | |
142 | ||
143 | This symbol should be selected by an architecture if it can | |
144 | perform unaligned accesses efficiently to allow different | |
145 | code paths to be selected for these cases. Some network | |
146 | drivers, for example, could opt to not fix up alignment | |
147 | problems with received packets if doing so would not help | |
148 | much. | |
149 | ||
150 | See Documentation/unaligned-memory-access.txt for more | |
151 | information on the topic of unaligned memory accesses. | |
152 | ||
cf66bb93 DW |
153 | config ARCH_USE_BUILTIN_BSWAP |
154 | bool | |
155 | help | |
156 | Modern versions of GCC (since 4.4) have builtin functions | |
157 | for handling byte-swapping. Using these, instead of the old | |
158 | inline assembler that the architecture code provides in the | |
159 | __arch_bswapXX() macros, allows the compiler to see what's | |
160 | happening and offers more opportunity for optimisation. In | |
161 | particular, the compiler will be able to combine the byteswap | |
162 | with a nearby load or store and use load-and-swap or | |
163 | store-and-swap instructions if the architecture has them. It | |
164 | should almost *never* result in code which is worse than the | |
165 | hand-coded assembler in <asm/swab.h>. But just in case it | |
166 | does, the use of the builtins is optional. | |
167 | ||
168 | Any architecture with load-and-swap or store-and-swap | |
169 | instructions should set this. And it shouldn't hurt to set it | |
170 | on architectures that don't have such instructions. | |
171 | ||
9edddaa2 AM |
172 | config KRETPROBES |
173 | def_bool y | |
174 | depends on KPROBES && HAVE_KRETPROBES | |
175 | ||
7c68af6e AK |
176 | config USER_RETURN_NOTIFIER |
177 | bool | |
178 | depends on HAVE_USER_RETURN_NOTIFIER | |
179 | help | |
180 | Provide a kernel-internal notification when a cpu is about to | |
181 | switch to user mode. | |
182 | ||
28b2ee20 | 183 | config HAVE_IOREMAP_PROT |
9ba16087 | 184 | bool |
28b2ee20 | 185 | |
125e5645 | 186 | config HAVE_KPROBES |
9ba16087 | 187 | bool |
9edddaa2 AM |
188 | |
189 | config HAVE_KRETPROBES | |
9ba16087 | 190 | bool |
74bc7cee | 191 | |
afd66255 MH |
192 | config HAVE_OPTPROBES |
193 | bool | |
d314d74c | 194 | |
e7dbfe34 MH |
195 | config HAVE_KPROBES_ON_FTRACE |
196 | bool | |
197 | ||
42a0bb3f PM |
198 | config HAVE_NMI |
199 | bool | |
200 | ||
d314d74c | 201 | config HAVE_NMI_WATCHDOG |
42a0bb3f | 202 | depends on HAVE_NMI |
d314d74c | 203 | bool |
1f5a4ad9 RM |
204 | # |
205 | # An arch should select this if it provides all these things: | |
206 | # | |
207 | # task_pt_regs() in asm/processor.h or asm/ptrace.h | |
208 | # arch_has_single_step() if there is hardware single-step support | |
209 | # arch_has_block_step() if there is hardware block-step support | |
1f5a4ad9 RM |
210 | # asm/syscall.h supplying asm-generic/syscall.h interface |
211 | # linux/regset.h user_regset interfaces | |
212 | # CORE_DUMP_USE_REGSET #define'd in linux/elf.h | |
213 | # TIF_SYSCALL_TRACE calls tracehook_report_syscall_{entry,exit} | |
214 | # TIF_NOTIFY_RESUME calls tracehook_notify_resume() | |
215 | # signal delivery calls tracehook_signal_handler() | |
216 | # | |
217 | config HAVE_ARCH_TRACEHOOK | |
9ba16087 | 218 | bool |
1f5a4ad9 | 219 | |
c64be2bb MS |
220 | config HAVE_DMA_CONTIGUOUS |
221 | bool | |
222 | ||
29d5e047 TG |
223 | config GENERIC_SMP_IDLE_THREAD |
224 | bool | |
225 | ||
485cf5da KH |
226 | config GENERIC_IDLE_POLL_SETUP |
227 | bool | |
228 | ||
d2852a22 DB |
229 | # Select if arch has all set_memory_ro/rw/x/nx() functions in asm/cacheflush.h |
230 | config ARCH_HAS_SET_MEMORY | |
231 | bool | |
232 | ||
a6359d1e TG |
233 | # Select if arch init_task initializer is different to init/init_task.c |
234 | config ARCH_INIT_TASK | |
a4a2eb49 TG |
235 | bool |
236 | ||
f5e10287 TG |
237 | # Select if arch has its private alloc_task_struct() function |
238 | config ARCH_TASK_STRUCT_ALLOCATOR | |
239 | bool | |
240 | ||
b235beea LT |
241 | # Select if arch has its private alloc_thread_stack() function |
242 | config ARCH_THREAD_STACK_ALLOCATOR | |
f5e10287 TG |
243 | bool |
244 | ||
5aaeb5c0 IM |
245 | # Select if arch wants to size task_struct dynamically via arch_task_struct_size: |
246 | config ARCH_WANTS_DYNAMIC_TASK_STRUCT | |
247 | bool | |
248 | ||
f850c30c HC |
249 | config HAVE_REGS_AND_STACK_ACCESS_API |
250 | bool | |
e01292b1 HC |
251 | help |
252 | This symbol should be selected by an architecure if it supports | |
253 | the API needed to access registers and stack entries from pt_regs, | |
254 | declared in asm/ptrace.h | |
255 | For example the kprobes-based event tracer needs this API. | |
f850c30c | 256 | |
9483a578 | 257 | config HAVE_CLK |
9ba16087 | 258 | bool |
9483a578 DB |
259 | help |
260 | The <linux/clk.h> calls support software clock gating and | |
261 | thus are a key power management tool on many systems. | |
262 | ||
5ee00bd4 JR |
263 | config HAVE_DMA_API_DEBUG |
264 | bool | |
36cd3c9f | 265 | |
62a038d3 P |
266 | config HAVE_HW_BREAKPOINT |
267 | bool | |
99e8c5a3 | 268 | depends on PERF_EVENTS |
62a038d3 | 269 | |
0102752e FW |
270 | config HAVE_MIXED_BREAKPOINTS_REGS |
271 | bool | |
272 | depends on HAVE_HW_BREAKPOINT | |
273 | help | |
274 | Depending on the arch implementation of hardware breakpoints, | |
275 | some of them have separate registers for data and instruction | |
276 | breakpoints addresses, others have mixed registers to store | |
277 | them but define the access type in a control register. | |
278 | Select this option if your arch implements breakpoints under the | |
279 | latter fashion. | |
280 | ||
7c68af6e AK |
281 | config HAVE_USER_RETURN_NOTIFIER |
282 | bool | |
a1922ed6 | 283 | |
c01d4323 FW |
284 | config HAVE_PERF_EVENTS_NMI |
285 | bool | |
23637d47 FW |
286 | help |
287 | System hardware can generate an NMI using the perf event | |
288 | subsystem. Also has support for calculating CPU cycle events | |
289 | to determine how many clock cycles in a given period. | |
c01d4323 | 290 | |
c5e63197 JO |
291 | config HAVE_PERF_REGS |
292 | bool | |
293 | help | |
294 | Support selective register dumps for perf events. This includes | |
295 | bit-mapping of each registers and a unique architecture id. | |
296 | ||
c5ebcedb JO |
297 | config HAVE_PERF_USER_STACK_DUMP |
298 | bool | |
299 | help | |
300 | Support user stack dumps for perf event samples. This needs | |
301 | access to the user stack pointer which is not unified across | |
302 | architectures. | |
303 | ||
bf5438fc JB |
304 | config HAVE_ARCH_JUMP_LABEL |
305 | bool | |
306 | ||
26723911 PZ |
307 | config HAVE_RCU_TABLE_FREE |
308 | bool | |
309 | ||
df013ffb HY |
310 | config ARCH_HAVE_NMI_SAFE_CMPXCHG |
311 | bool | |
312 | ||
43570fd2 HC |
313 | config HAVE_ALIGNED_STRUCT_PAGE |
314 | bool | |
315 | help | |
316 | This makes sure that struct pages are double word aligned and that | |
317 | e.g. the SLUB allocator can perform double word atomic operations | |
318 | on a struct page for better performance. However selecting this | |
319 | might increase the size of a struct page by a word. | |
320 | ||
4156153c HC |
321 | config HAVE_CMPXCHG_LOCAL |
322 | bool | |
323 | ||
2565409f HC |
324 | config HAVE_CMPXCHG_DOUBLE |
325 | bool | |
326 | ||
77e58496 PM |
327 | config ARCH_WEAK_RELEASE_ACQUIRE |
328 | bool | |
329 | ||
c1d7e01d WD |
330 | config ARCH_WANT_IPC_PARSE_VERSION |
331 | bool | |
332 | ||
333 | config ARCH_WANT_COMPAT_IPC_PARSE_VERSION | |
334 | bool | |
335 | ||
48b25c43 | 336 | config ARCH_WANT_OLD_COMPAT_IPC |
c1d7e01d | 337 | select ARCH_WANT_COMPAT_IPC_PARSE_VERSION |
48b25c43 CM |
338 | bool |
339 | ||
e2cfabdf WD |
340 | config HAVE_ARCH_SECCOMP_FILTER |
341 | bool | |
342 | help | |
fb0fadf9 | 343 | An arch should select this symbol if it provides all of these things: |
bb6ea430 WD |
344 | - syscall_get_arch() |
345 | - syscall_get_arguments() | |
346 | - syscall_rollback() | |
347 | - syscall_set_return_value() | |
fb0fadf9 WD |
348 | - SIGSYS siginfo_t support |
349 | - secure_computing is called from a ptrace_event()-safe context | |
350 | - secure_computing return value is checked and a return value of -1 | |
351 | results in the system call being skipped immediately. | |
48dc92b9 | 352 | - seccomp syscall wired up |
e2cfabdf WD |
353 | |
354 | config SECCOMP_FILTER | |
355 | def_bool y | |
356 | depends on HAVE_ARCH_SECCOMP_FILTER && SECCOMP && NET | |
357 | help | |
358 | Enable tasks to build secure computing environments defined | |
359 | in terms of Berkeley Packet Filter programs which implement | |
360 | task-defined system call filtering polices. | |
361 | ||
362 | See Documentation/prctl/seccomp_filter.txt for details. | |
363 | ||
6b90bd4b ER |
364 | config HAVE_GCC_PLUGINS |
365 | bool | |
366 | help | |
367 | An arch should select this symbol if it supports building with | |
368 | GCC plugins. | |
369 | ||
370 | menuconfig GCC_PLUGINS | |
371 | bool "GCC plugins" | |
372 | depends on HAVE_GCC_PLUGINS | |
a519167e | 373 | depends on !COMPILE_TEST |
6b90bd4b ER |
374 | help |
375 | GCC plugins are loadable modules that provide extra features to the | |
376 | compiler. They are useful for runtime instrumentation and static analysis. | |
377 | ||
378 | See Documentation/gcc-plugins.txt for details. | |
379 | ||
0dae776c | 380 | config GCC_PLUGIN_CYC_COMPLEXITY |
215e2aa6 | 381 | bool "Compute the cyclomatic complexity of a function" if EXPERT |
0dae776c | 382 | depends on GCC_PLUGINS |
215e2aa6 | 383 | depends on !COMPILE_TEST |
0dae776c ER |
384 | help |
385 | The complexity M of a function's control flow graph is defined as: | |
386 | M = E - N + 2P | |
387 | where | |
388 | ||
389 | E = the number of edges | |
390 | N = the number of nodes | |
391 | P = the number of connected components (exit nodes). | |
392 | ||
215e2aa6 KC |
393 | Enabling this plugin reports the complexity to stderr during the |
394 | build. It mainly serves as a simple example of how to create a | |
395 | gcc plugin for the kernel. | |
396 | ||
543c37cb ER |
397 | config GCC_PLUGIN_SANCOV |
398 | bool | |
399 | depends on GCC_PLUGINS | |
400 | help | |
401 | This plugin inserts a __sanitizer_cov_trace_pc() call at the start of | |
402 | basic blocks. It supports all gcc versions with plugin support (from | |
403 | gcc-4.5 on). It is based on the commit "Add fuzzing coverage support" | |
404 | by Dmitry Vyukov <dvyukov@google.com>. | |
405 | ||
38addce8 ER |
406 | config GCC_PLUGIN_LATENT_ENTROPY |
407 | bool "Generate some entropy during boot and runtime" | |
408 | depends on GCC_PLUGINS | |
409 | help | |
410 | By saying Y here the kernel will instrument some kernel code to | |
411 | extract some entropy from both original and artificially created | |
412 | program state. This will help especially embedded systems where | |
413 | there is little 'natural' source of entropy normally. The cost | |
414 | is some slowdown of the boot process (about 0.5%) and fork and | |
415 | irq processing. | |
416 | ||
417 | Note that entropy extracted this way is not cryptographically | |
418 | secure! | |
419 | ||
420 | This plugin was ported from grsecurity/PaX. More information at: | |
421 | * https://grsecurity.net/ | |
422 | * https://pax.grsecurity.net/ | |
423 | ||
c61f13ea KC |
424 | config GCC_PLUGIN_STRUCTLEAK |
425 | bool "Force initialization of variables containing userspace addresses" | |
426 | depends on GCC_PLUGINS | |
427 | help | |
f136e090 | 428 | This plugin zero-initializes any structures containing a |
c61f13ea KC |
429 | __user attribute. This can prevent some classes of information |
430 | exposures. | |
431 | ||
432 | This plugin was ported from grsecurity/PaX. More information at: | |
433 | * https://grsecurity.net/ | |
434 | * https://pax.grsecurity.net/ | |
435 | ||
436 | config GCC_PLUGIN_STRUCTLEAK_VERBOSE | |
437 | bool "Report forcefully initialized variables" | |
438 | depends on GCC_PLUGIN_STRUCTLEAK | |
439 | depends on !COMPILE_TEST | |
440 | help | |
441 | This option will cause a warning to be printed each time the | |
442 | structleak plugin finds a variable it thinks needs to be | |
443 | initialized. Since not all existing initializers are detected | |
444 | by the plugin, this can produce false positive warnings. | |
445 | ||
313dd1b6 KC |
446 | config GCC_PLUGIN_RANDSTRUCT |
447 | bool "Randomize layout of sensitive kernel structures" | |
448 | depends on GCC_PLUGINS | |
449 | select MODVERSIONS if MODULES | |
450 | help | |
451 | If you say Y here, the layouts of structures explicitly | |
452 | marked by __randomize_layout will be randomized at | |
453 | compile-time. This can introduce the requirement of an | |
454 | additional information exposure vulnerability for exploits | |
455 | targeting these structure types. | |
456 | ||
457 | Enabling this feature will introduce some performance impact, | |
458 | slightly increase memory usage, and prevent the use of forensic | |
459 | tools like Volatility against the system (unless the kernel | |
460 | source tree isn't cleaned after kernel installation). | |
461 | ||
462 | The seed used for compilation is located at | |
463 | scripts/gcc-plgins/randomize_layout_seed.h. It remains after | |
464 | a make clean to allow for external modules to be compiled with | |
465 | the existing seed and will be removed by a make mrproper or | |
466 | make distclean. | |
467 | ||
468 | Note that the implementation requires gcc 4.7 or newer. | |
469 | ||
470 | This plugin was ported from grsecurity/PaX. More information at: | |
471 | * https://grsecurity.net/ | |
472 | * https://pax.grsecurity.net/ | |
473 | ||
474 | config GCC_PLUGIN_RANDSTRUCT_PERFORMANCE | |
475 | bool "Use cacheline-aware structure randomization" | |
476 | depends on GCC_PLUGIN_RANDSTRUCT | |
477 | depends on !COMPILE_TEST | |
478 | help | |
479 | If you say Y here, the RANDSTRUCT randomization will make a | |
480 | best effort at restricting randomization to cacheline-sized | |
481 | groups of elements. It will further not randomize bitfields | |
482 | in structures. This reduces the performance hit of RANDSTRUCT | |
483 | at the cost of weakened randomization. | |
484 | ||
19952a92 KC |
485 | config HAVE_CC_STACKPROTECTOR |
486 | bool | |
487 | help | |
488 | An arch should select this symbol if: | |
489 | - its compiler supports the -fstack-protector option | |
490 | - it has implemented a stack canary (e.g. __stack_chk_guard) | |
491 | ||
492 | config CC_STACKPROTECTOR | |
8779657d KC |
493 | def_bool n |
494 | help | |
495 | Set when a stack-protector mode is enabled, so that the build | |
496 | can enable kernel-side support for the GCC feature. | |
497 | ||
498 | choice | |
499 | prompt "Stack Protector buffer overflow detection" | |
19952a92 | 500 | depends on HAVE_CC_STACKPROTECTOR |
8779657d | 501 | default CC_STACKPROTECTOR_NONE |
19952a92 | 502 | help |
8779657d | 503 | This option turns on the "stack-protector" GCC feature. This |
19952a92 KC |
504 | feature puts, at the beginning of functions, a canary value on |
505 | the stack just before the return address, and validates | |
506 | the value just before actually returning. Stack based buffer | |
507 | overflows (that need to overwrite this return address) now also | |
508 | overwrite the canary, which gets detected and the attack is then | |
509 | neutralized via a kernel panic. | |
510 | ||
8779657d KC |
511 | config CC_STACKPROTECTOR_NONE |
512 | bool "None" | |
513 | help | |
514 | Disable "stack-protector" GCC feature. | |
515 | ||
516 | config CC_STACKPROTECTOR_REGULAR | |
517 | bool "Regular" | |
518 | select CC_STACKPROTECTOR | |
519 | help | |
520 | Functions will have the stack-protector canary logic added if they | |
521 | have an 8-byte or larger character array on the stack. | |
522 | ||
19952a92 | 523 | This feature requires gcc version 4.2 or above, or a distribution |
8779657d KC |
524 | gcc with the feature backported ("-fstack-protector"). |
525 | ||
526 | On an x86 "defconfig" build, this feature adds canary checks to | |
527 | about 3% of all kernel functions, which increases kernel code size | |
528 | by about 0.3%. | |
529 | ||
530 | config CC_STACKPROTECTOR_STRONG | |
531 | bool "Strong" | |
532 | select CC_STACKPROTECTOR | |
533 | help | |
534 | Functions will have the stack-protector canary logic added in any | |
535 | of the following conditions: | |
536 | ||
537 | - local variable's address used as part of the right hand side of an | |
538 | assignment or function argument | |
539 | - local variable is an array (or union containing an array), | |
540 | regardless of array type or length | |
541 | - uses register local variables | |
542 | ||
543 | This feature requires gcc version 4.9 or above, or a distribution | |
544 | gcc with the feature backported ("-fstack-protector-strong"). | |
545 | ||
546 | On an x86 "defconfig" build, this feature adds canary checks to | |
547 | about 20% of all kernel functions, which increases the kernel code | |
548 | size by about 2%. | |
549 | ||
550 | endchoice | |
19952a92 | 551 | |
a5967db9 | 552 | config THIN_ARCHIVES |
799c4341 | 553 | def_bool y |
a5967db9 SR |
554 | help |
555 | Select this if the architecture wants to use thin archives | |
556 | instead of ld -r to create the built-in.o files. | |
557 | ||
b67067f1 NP |
558 | config LD_DEAD_CODE_DATA_ELIMINATION |
559 | bool | |
560 | help | |
561 | Select this if the architecture wants to do dead code and | |
562 | data elimination with the linker by compiling with | |
563 | -ffunction-sections -fdata-sections and linking with | |
564 | --gc-sections. | |
565 | ||
566 | This requires that the arch annotates or otherwise protects | |
567 | its external entry points from being discarded. Linker scripts | |
568 | must also merge .text.*, .data.*, and .bss.* correctly into | |
0f4c4af0 NP |
569 | output sections. Care must be taken not to pull in unrelated |
570 | sections (e.g., '.text.init'). Typically '.' in section names | |
571 | is used to distinguish them from label names / C identifiers. | |
b67067f1 | 572 | |
0f60a8ef KC |
573 | config HAVE_ARCH_WITHIN_STACK_FRAMES |
574 | bool | |
575 | help | |
576 | An architecture should select this if it can walk the kernel stack | |
577 | frames to determine if an object is part of either the arguments | |
578 | or local variables (i.e. that it excludes saved return addresses, | |
579 | and similar) by implementing an inline arch_within_stack_frames(), | |
580 | which is used by CONFIG_HARDENED_USERCOPY. | |
581 | ||
91d1aa43 | 582 | config HAVE_CONTEXT_TRACKING |
2b1d5024 FW |
583 | bool |
584 | help | |
91d1aa43 FW |
585 | Provide kernel/user boundaries probes necessary for subsystems |
586 | that need it, such as userspace RCU extended quiescent state. | |
587 | Syscalls need to be wrapped inside user_exit()-user_enter() through | |
588 | the slow path using TIF_NOHZ flag. Exceptions handlers must be | |
589 | wrapped as well. Irqs are already protected inside | |
590 | rcu_irq_enter/rcu_irq_exit() but preemption or signal handling on | |
591 | irq exit still need to be protected. | |
2b1d5024 | 592 | |
b952741c FW |
593 | config HAVE_VIRT_CPU_ACCOUNTING |
594 | bool | |
595 | ||
40565b5a SG |
596 | config ARCH_HAS_SCALED_CPUTIME |
597 | bool | |
598 | ||
554b0004 KH |
599 | config HAVE_VIRT_CPU_ACCOUNTING_GEN |
600 | bool | |
601 | default y if 64BIT | |
602 | help | |
603 | With VIRT_CPU_ACCOUNTING_GEN, cputime_t becomes 64-bit. | |
604 | Before enabling this option, arch code must be audited | |
605 | to ensure there are no races in concurrent read/write of | |
606 | cputime_t. For example, reading/writing 64-bit cputime_t on | |
607 | some 32-bit arches may require multiple accesses, so proper | |
608 | locking is needed to protect against concurrent accesses. | |
609 | ||
610 | ||
fdf9c356 FW |
611 | config HAVE_IRQ_TIME_ACCOUNTING |
612 | bool | |
613 | help | |
614 | Archs need to ensure they use a high enough resolution clock to | |
615 | support irq time accounting and then call enable_sched_clock_irqtime(). | |
616 | ||
15626062 GS |
617 | config HAVE_ARCH_TRANSPARENT_HUGEPAGE |
618 | bool | |
619 | ||
a00cc7d9 MW |
620 | config HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD |
621 | bool | |
622 | ||
0ddab1d2 TK |
623 | config HAVE_ARCH_HUGE_VMAP |
624 | bool | |
625 | ||
0f8975ec PE |
626 | config HAVE_ARCH_SOFT_DIRTY |
627 | bool | |
628 | ||
786d35d4 DH |
629 | config HAVE_MOD_ARCH_SPECIFIC |
630 | bool | |
631 | help | |
632 | The arch uses struct mod_arch_specific to store data. Many arches | |
633 | just need a simple module loader without arch specific data - those | |
634 | should not enable this. | |
635 | ||
636 | config MODULES_USE_ELF_RELA | |
637 | bool | |
638 | help | |
639 | Modules only use ELF RELA relocations. Modules with ELF REL | |
640 | relocations will give an error. | |
641 | ||
642 | config MODULES_USE_ELF_REL | |
643 | bool | |
644 | help | |
645 | Modules only use ELF REL relocations. Modules with ELF RELA | |
646 | relocations will give an error. | |
647 | ||
b92021b0 RR |
648 | config HAVE_UNDERSCORE_SYMBOL_PREFIX |
649 | bool | |
650 | help | |
651 | Some architectures generate an _ in front of C symbols; things like | |
652 | module loading and assembly files need to know about this. | |
653 | ||
cc1f0274 FW |
654 | config HAVE_IRQ_EXIT_ON_IRQ_STACK |
655 | bool | |
656 | help | |
657 | Architecture doesn't only execute the irq handler on the irq stack | |
658 | but also irq_exit(). This way we can process softirqs on this irq | |
659 | stack instead of switching to a new one when we call __do_softirq() | |
660 | in the end of an hardirq. | |
661 | This spares a stack switch and improves cache usage on softirq | |
662 | processing. | |
663 | ||
235a8f02 KS |
664 | config PGTABLE_LEVELS |
665 | int | |
666 | default 2 | |
667 | ||
2b68f6ca KC |
668 | config ARCH_HAS_ELF_RANDOMIZE |
669 | bool | |
670 | help | |
671 | An architecture supports choosing randomized locations for | |
672 | stack, mmap, brk, and ET_DYN. Defined functions: | |
673 | - arch_mmap_rnd() | |
204db6ed | 674 | - arch_randomize_brk() |
2b68f6ca | 675 | |
d07e2259 DC |
676 | config HAVE_ARCH_MMAP_RND_BITS |
677 | bool | |
678 | help | |
679 | An arch should select this symbol if it supports setting a variable | |
680 | number of bits for use in establishing the base address for mmap | |
681 | allocations, has MMU enabled and provides values for both: | |
682 | - ARCH_MMAP_RND_BITS_MIN | |
683 | - ARCH_MMAP_RND_BITS_MAX | |
684 | ||
5f56a5df JS |
685 | config HAVE_EXIT_THREAD |
686 | bool | |
687 | help | |
688 | An architecture implements exit_thread. | |
689 | ||
d07e2259 DC |
690 | config ARCH_MMAP_RND_BITS_MIN |
691 | int | |
692 | ||
693 | config ARCH_MMAP_RND_BITS_MAX | |
694 | int | |
695 | ||
696 | config ARCH_MMAP_RND_BITS_DEFAULT | |
697 | int | |
698 | ||
699 | config ARCH_MMAP_RND_BITS | |
700 | int "Number of bits to use for ASLR of mmap base address" if EXPERT | |
701 | range ARCH_MMAP_RND_BITS_MIN ARCH_MMAP_RND_BITS_MAX | |
702 | default ARCH_MMAP_RND_BITS_DEFAULT if ARCH_MMAP_RND_BITS_DEFAULT | |
703 | default ARCH_MMAP_RND_BITS_MIN | |
704 | depends on HAVE_ARCH_MMAP_RND_BITS | |
705 | help | |
706 | This value can be used to select the number of bits to use to | |
707 | determine the random offset to the base address of vma regions | |
708 | resulting from mmap allocations. This value will be bounded | |
709 | by the architecture's minimum and maximum supported values. | |
710 | ||
711 | This value can be changed after boot using the | |
712 | /proc/sys/vm/mmap_rnd_bits tunable | |
713 | ||
714 | config HAVE_ARCH_MMAP_RND_COMPAT_BITS | |
715 | bool | |
716 | help | |
717 | An arch should select this symbol if it supports running applications | |
718 | in compatibility mode, supports setting a variable number of bits for | |
719 | use in establishing the base address for mmap allocations, has MMU | |
720 | enabled and provides values for both: | |
721 | - ARCH_MMAP_RND_COMPAT_BITS_MIN | |
722 | - ARCH_MMAP_RND_COMPAT_BITS_MAX | |
723 | ||
724 | config ARCH_MMAP_RND_COMPAT_BITS_MIN | |
725 | int | |
726 | ||
727 | config ARCH_MMAP_RND_COMPAT_BITS_MAX | |
728 | int | |
729 | ||
730 | config ARCH_MMAP_RND_COMPAT_BITS_DEFAULT | |
731 | int | |
732 | ||
733 | config ARCH_MMAP_RND_COMPAT_BITS | |
734 | int "Number of bits to use for ASLR of mmap base address for compatible applications" if EXPERT | |
735 | range ARCH_MMAP_RND_COMPAT_BITS_MIN ARCH_MMAP_RND_COMPAT_BITS_MAX | |
736 | default ARCH_MMAP_RND_COMPAT_BITS_DEFAULT if ARCH_MMAP_RND_COMPAT_BITS_DEFAULT | |
737 | default ARCH_MMAP_RND_COMPAT_BITS_MIN | |
738 | depends on HAVE_ARCH_MMAP_RND_COMPAT_BITS | |
739 | help | |
740 | This value can be used to select the number of bits to use to | |
741 | determine the random offset to the base address of vma regions | |
742 | resulting from mmap allocations for compatible applications This | |
743 | value will be bounded by the architecture's minimum and maximum | |
744 | supported values. | |
745 | ||
746 | This value can be changed after boot using the | |
747 | /proc/sys/vm/mmap_rnd_compat_bits tunable | |
748 | ||
1b028f78 DS |
749 | config HAVE_ARCH_COMPAT_MMAP_BASES |
750 | bool | |
751 | help | |
752 | This allows 64bit applications to invoke 32-bit mmap() syscall | |
753 | and vice-versa 32-bit applications to call 64-bit mmap(). | |
754 | Required for applications doing different bitness syscalls. | |
755 | ||
3033f14a JT |
756 | config HAVE_COPY_THREAD_TLS |
757 | bool | |
758 | help | |
759 | Architecture provides copy_thread_tls to accept tls argument via | |
760 | normal C parameter passing, rather than extracting the syscall | |
761 | argument from pt_regs. | |
762 | ||
b9ab5ebb JP |
763 | config HAVE_STACK_VALIDATION |
764 | bool | |
765 | help | |
766 | Architecture supports the 'objtool check' host tool command, which | |
767 | performs compile-time stack metadata validation. | |
768 | ||
af085d90 JP |
769 | config HAVE_RELIABLE_STACKTRACE |
770 | bool | |
771 | help | |
772 | Architecture has a save_stack_trace_tsk_reliable() function which | |
773 | only returns a stack trace if it can guarantee the trace is reliable. | |
774 | ||
468a9428 GS |
775 | config HAVE_ARCH_HASH |
776 | bool | |
777 | default n | |
778 | help | |
779 | If this is set, the architecture provides an <asm/hash.h> | |
780 | file which provides platform-specific implementations of some | |
781 | functions in <linux/hash.h> or fs/namei.c. | |
782 | ||
3a495511 WBG |
783 | config ISA_BUS_API |
784 | def_bool ISA | |
785 | ||
d2125043 AV |
786 | # |
787 | # ABI hall of shame | |
788 | # | |
789 | config CLONE_BACKWARDS | |
790 | bool | |
791 | help | |
792 | Architecture has tls passed as the 4th argument of clone(2), | |
793 | not the 5th one. | |
794 | ||
795 | config CLONE_BACKWARDS2 | |
796 | bool | |
797 | help | |
798 | Architecture has the first two arguments of clone(2) swapped. | |
799 | ||
dfa9771a MS |
800 | config CLONE_BACKWARDS3 |
801 | bool | |
802 | help | |
803 | Architecture has tls passed as the 3rd argument of clone(2), | |
804 | not the 5th one. | |
805 | ||
eaca6eae AV |
806 | config ODD_RT_SIGACTION |
807 | bool | |
808 | help | |
809 | Architecture has unusual rt_sigaction(2) arguments | |
810 | ||
0a0e8cdf AV |
811 | config OLD_SIGSUSPEND |
812 | bool | |
813 | help | |
814 | Architecture has old sigsuspend(2) syscall, of one-argument variety | |
815 | ||
816 | config OLD_SIGSUSPEND3 | |
817 | bool | |
818 | help | |
819 | Even weirder antique ABI - three-argument sigsuspend(2) | |
820 | ||
495dfbf7 AV |
821 | config OLD_SIGACTION |
822 | bool | |
823 | help | |
824 | Architecture has old sigaction(2) syscall. Nope, not the same | |
825 | as OLD_SIGSUSPEND | OLD_SIGSUSPEND3 - alpha has sigsuspend(2), | |
826 | but fairly different variant of sigaction(2), thanks to OSF/1 | |
827 | compatibility... | |
828 | ||
829 | config COMPAT_OLD_SIGACTION | |
830 | bool | |
831 | ||
0d4a619b CH |
832 | config ARCH_NO_COHERENT_DMA_MMAP |
833 | bool | |
834 | ||
fff7fb0b ZZ |
835 | config CPU_NO_EFFICIENT_FFS |
836 | def_bool n | |
837 | ||
ba14a194 AL |
838 | config HAVE_ARCH_VMAP_STACK |
839 | def_bool n | |
840 | help | |
841 | An arch should select this symbol if it can support kernel stacks | |
842 | in vmalloc space. This means: | |
843 | ||
844 | - vmalloc space must be large enough to hold many kernel stacks. | |
845 | This may rule out many 32-bit architectures. | |
846 | ||
847 | - Stacks in vmalloc space need to work reliably. For example, if | |
848 | vmap page tables are created on demand, either this mechanism | |
849 | needs to work while the stack points to a virtual address with | |
850 | unpopulated page tables or arch code (switch_to() and switch_mm(), | |
851 | most likely) needs to ensure that the stack's page table entries | |
852 | are populated before running on a possibly unpopulated stack. | |
853 | ||
854 | - If the stack overflows into a guard page, something reasonable | |
855 | should happen. The definition of "reasonable" is flexible, but | |
856 | instantly rebooting without logging anything would be unfriendly. | |
857 | ||
858 | config VMAP_STACK | |
859 | default y | |
860 | bool "Use a virtually-mapped stack" | |
861 | depends on HAVE_ARCH_VMAP_STACK && !KASAN | |
862 | ---help--- | |
863 | Enable this if you want the use virtually-mapped kernel stacks | |
864 | with guard pages. This causes kernel stack overflows to be | |
865 | caught immediately rather than causing difficult-to-diagnose | |
866 | corruption. | |
867 | ||
868 | This is presently incompatible with KASAN because KASAN expects | |
869 | the stack to map directly to the KASAN shadow map using a formula | |
870 | that is incorrect if the stack is in vmalloc space. | |
871 | ||
ad21fc4f LA |
872 | config ARCH_OPTIONAL_KERNEL_RWX |
873 | def_bool n | |
874 | ||
875 | config ARCH_OPTIONAL_KERNEL_RWX_DEFAULT | |
876 | def_bool n | |
877 | ||
878 | config ARCH_HAS_STRICT_KERNEL_RWX | |
879 | def_bool n | |
880 | ||
0f5bf6d0 | 881 | config STRICT_KERNEL_RWX |
ad21fc4f LA |
882 | bool "Make kernel text and rodata read-only" if ARCH_OPTIONAL_KERNEL_RWX |
883 | depends on ARCH_HAS_STRICT_KERNEL_RWX | |
884 | default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT | |
885 | help | |
886 | If this is set, kernel text and rodata memory will be made read-only, | |
887 | and non-text memory will be made non-executable. This provides | |
888 | protection against certain security exploits (e.g. executing the heap | |
889 | or modifying text) | |
890 | ||
891 | These features are considered standard security practice these days. | |
892 | You should say Y here in almost all cases. | |
893 | ||
894 | config ARCH_HAS_STRICT_MODULE_RWX | |
895 | def_bool n | |
896 | ||
0f5bf6d0 | 897 | config STRICT_MODULE_RWX |
ad21fc4f LA |
898 | bool "Set loadable kernel module data as NX and text as RO" if ARCH_OPTIONAL_KERNEL_RWX |
899 | depends on ARCH_HAS_STRICT_MODULE_RWX && MODULES | |
900 | default !ARCH_OPTIONAL_KERNEL_RWX || ARCH_OPTIONAL_KERNEL_RWX_DEFAULT | |
901 | help | |
902 | If this is set, module text and rodata memory will be made read-only, | |
903 | and non-text memory will be made non-executable. This provides | |
904 | protection against certain security exploits (e.g. writing to text) | |
905 | ||
1a8b6d76 MW |
906 | config ARCH_WANT_RELAX_ORDER |
907 | bool | |
908 | ||
fd25d19f KC |
909 | config REFCOUNT_FULL |
910 | bool "Perform full reference count validation at the expense of speed" | |
911 | help | |
912 | Enabling this switches the refcounting infrastructure from a fast | |
913 | unchecked atomic_t implementation to a fully state checked | |
914 | implementation, which can be (slightly) slower but provides protections | |
915 | against various use-after-free conditions that can be used in | |
916 | security flaw exploits. | |
917 | ||
2521f2c2 | 918 | source "kernel/gcov/Kconfig" |