Kconfig 12 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413
  1. #
  2. # Architectures that offer an FUNCTION_TRACER implementation should
  3. # select HAVE_FUNCTION_TRACER:
  4. #
  5. config USER_STACKTRACE_SUPPORT
  6. bool
  7. config NOP_TRACER
  8. bool
  9. config HAVE_FTRACE_NMI_ENTER
  10. bool
  11. config HAVE_FUNCTION_TRACER
  12. bool
  13. config HAVE_FUNCTION_GRAPH_TRACER
  14. bool
  15. config HAVE_FUNCTION_TRACE_MCOUNT_TEST
  16. bool
  17. help
  18. This gets selected when the arch tests the function_trace_stop
  19. variable at the mcount call site. Otherwise, this variable
  20. is tested by the called function.
  21. config HAVE_DYNAMIC_FTRACE
  22. bool
  23. config HAVE_FTRACE_MCOUNT_RECORD
  24. bool
  25. config HAVE_HW_BRANCH_TRACER
  26. bool
  27. config TRACER_MAX_TRACE
  28. bool
  29. config RING_BUFFER
  30. bool
  31. config FTRACE_NMI_ENTER
  32. bool
  33. depends on HAVE_FTRACE_NMI_ENTER
  34. default y
  35. config TRACING
  36. bool
  37. select DEBUG_FS
  38. select RING_BUFFER
  39. select STACKTRACE if STACKTRACE_SUPPORT
  40. select TRACEPOINTS
  41. select NOP_TRACER
  42. #
  43. # Minimum requirements an architecture has to meet for us to
  44. # be able to offer generic tracing facilities:
  45. #
  46. config TRACING_SUPPORT
  47. bool
  48. depends on TRACE_IRQFLAGS_SUPPORT
  49. depends on STACKTRACE_SUPPORT
  50. default y
  51. if TRACING_SUPPORT
  52. menu "Tracers"
  53. config FUNCTION_TRACER
  54. bool "Kernel Function Tracer"
  55. depends on HAVE_FUNCTION_TRACER
  56. select FRAME_POINTER
  57. select KALLSYMS
  58. select TRACING
  59. select CONTEXT_SWITCH_TRACER
  60. help
  61. Enable the kernel to trace every kernel function. This is done
  62. by using a compiler feature to insert a small, 5-byte No-Operation
  63. instruction to the beginning of every kernel function, which NOP
  64. sequence is then dynamically patched into a tracer call when
  65. tracing is enabled by the administrator. If it's runtime disabled
  66. (the bootup default), then the overhead of the instructions is very
  67. small and not measurable even in micro-benchmarks.
  68. config FUNCTION_GRAPH_TRACER
  69. bool "Kernel Function Graph Tracer"
  70. depends on HAVE_FUNCTION_GRAPH_TRACER
  71. depends on FUNCTION_TRACER
  72. default y
  73. help
  74. Enable the kernel to trace a function at both its return
  75. and its entry.
  76. It's first purpose is to trace the duration of functions and
  77. draw a call graph for each thread with some informations like
  78. the return value.
  79. This is done by setting the current return address on the current
  80. task structure into a stack of calls.
  81. config TRACE_BPRINTK
  82. bool "Binary printk for tracing"
  83. default y
  84. depends on TRACING
  85. select BINARY_PRINTF
  86. config IRQSOFF_TRACER
  87. bool "Interrupts-off Latency Tracer"
  88. default n
  89. depends on TRACE_IRQFLAGS_SUPPORT
  90. depends on GENERIC_TIME
  91. select TRACE_IRQFLAGS
  92. select TRACING
  93. select TRACER_MAX_TRACE
  94. help
  95. This option measures the time spent in irqs-off critical
  96. sections, with microsecond accuracy.
  97. The default measurement method is a maximum search, which is
  98. disabled by default and can be runtime (re-)started
  99. via:
  100. echo 0 > /debugfs/tracing/tracing_max_latency
  101. (Note that kernel size and overhead increases with this option
  102. enabled. This option and the preempt-off timing option can be
  103. used together or separately.)
  104. config PREEMPT_TRACER
  105. bool "Preemption-off Latency Tracer"
  106. default n
  107. depends on GENERIC_TIME
  108. depends on PREEMPT
  109. select TRACING
  110. select TRACER_MAX_TRACE
  111. help
  112. This option measures the time spent in preemption off critical
  113. sections, with microsecond accuracy.
  114. The default measurement method is a maximum search, which is
  115. disabled by default and can be runtime (re-)started
  116. via:
  117. echo 0 > /debugfs/tracing/tracing_max_latency
  118. (Note that kernel size and overhead increases with this option
  119. enabled. This option and the irqs-off timing option can be
  120. used together or separately.)
  121. config SYSPROF_TRACER
  122. bool "Sysprof Tracer"
  123. depends on X86
  124. select TRACING
  125. select CONTEXT_SWITCH_TRACER
  126. help
  127. This tracer provides the trace needed by the 'Sysprof' userspace
  128. tool.
  129. config SCHED_TRACER
  130. bool "Scheduling Latency Tracer"
  131. select TRACING
  132. select CONTEXT_SWITCH_TRACER
  133. select TRACER_MAX_TRACE
  134. help
  135. This tracer tracks the latency of the highest priority task
  136. to be scheduled in, starting from the point it has woken up.
  137. config CONTEXT_SWITCH_TRACER
  138. bool "Trace process context switches"
  139. select TRACING
  140. select MARKERS
  141. help
  142. This tracer gets called from the context switch and records
  143. all switching of tasks.
  144. config EVENT_TRACER
  145. bool "Trace various events in the kernel"
  146. select TRACING
  147. help
  148. This tracer hooks to various trace points in the kernel
  149. allowing the user to pick and choose which trace point they
  150. want to trace.
  151. config BOOT_TRACER
  152. bool "Trace boot initcalls"
  153. select TRACING
  154. select CONTEXT_SWITCH_TRACER
  155. help
  156. This tracer helps developers to optimize boot times: it records
  157. the timings of the initcalls and traces key events and the identity
  158. of tasks that can cause boot delays, such as context-switches.
  159. Its aim is to be parsed by the /scripts/bootgraph.pl tool to
  160. produce pretty graphics about boot inefficiencies, giving a visual
  161. representation of the delays during initcalls - but the raw
  162. /debug/tracing/trace text output is readable too.
  163. You must pass in ftrace=initcall to the kernel command line
  164. to enable this on bootup.
  165. config TRACE_BRANCH_PROFILING
  166. bool "Trace likely/unlikely profiler"
  167. select TRACING
  168. help
  169. This tracer profiles all the the likely and unlikely macros
  170. in the kernel. It will display the results in:
  171. /debugfs/tracing/profile_annotated_branch
  172. Note: this will add a significant overhead, only turn this
  173. on if you need to profile the system's use of these macros.
  174. Say N if unsure.
  175. config PROFILE_ALL_BRANCHES
  176. bool "Profile all if conditionals"
  177. depends on TRACE_BRANCH_PROFILING
  178. help
  179. This tracer profiles all branch conditions. Every if ()
  180. taken in the kernel is recorded whether it hit or miss.
  181. The results will be displayed in:
  182. /debugfs/tracing/profile_branch
  183. This configuration, when enabled, will impose a great overhead
  184. on the system. This should only be enabled when the system
  185. is to be analyzed
  186. Say N if unsure.
  187. config TRACING_BRANCHES
  188. bool
  189. help
  190. Selected by tracers that will trace the likely and unlikely
  191. conditions. This prevents the tracers themselves from being
  192. profiled. Profiling the tracing infrastructure can only happen
  193. when the likelys and unlikelys are not being traced.
  194. config BRANCH_TRACER
  195. bool "Trace likely/unlikely instances"
  196. depends on TRACE_BRANCH_PROFILING
  197. select TRACING_BRANCHES
  198. help
  199. This traces the events of likely and unlikely condition
  200. calls in the kernel. The difference between this and the
  201. "Trace likely/unlikely profiler" is that this is not a
  202. histogram of the callers, but actually places the calling
  203. events into a running trace buffer to see when and where the
  204. events happened, as well as their results.
  205. Say N if unsure.
  206. config POWER_TRACER
  207. bool "Trace power consumption behavior"
  208. depends on X86
  209. select TRACING
  210. help
  211. This tracer helps developers to analyze and optimize the kernels
  212. power management decisions, specifically the C-state and P-state
  213. behavior.
  214. config STACK_TRACER
  215. bool "Trace max stack"
  216. depends on HAVE_FUNCTION_TRACER
  217. select FUNCTION_TRACER
  218. select STACKTRACE
  219. select KALLSYMS
  220. help
  221. This special tracer records the maximum stack footprint of the
  222. kernel and displays it in debugfs/tracing/stack_trace.
  223. This tracer works by hooking into every function call that the
  224. kernel executes, and keeping a maximum stack depth value and
  225. stack-trace saved. If this is configured with DYNAMIC_FTRACE
  226. then it will not have any overhead while the stack tracer
  227. is disabled.
  228. To enable the stack tracer on bootup, pass in 'stacktrace'
  229. on the kernel command line.
  230. The stack tracer can also be enabled or disabled via the
  231. sysctl kernel.stack_tracer_enabled
  232. Say N if unsure.
  233. config HW_BRANCH_TRACER
  234. depends on HAVE_HW_BRANCH_TRACER
  235. bool "Trace hw branches"
  236. select TRACING
  237. help
  238. This tracer records all branches on the system in a circular
  239. buffer giving access to the last N branches for each cpu.
  240. config KMEMTRACE
  241. bool "Trace SLAB allocations"
  242. select TRACING
  243. help
  244. kmemtrace provides tracing for slab allocator functions, such as
  245. kmalloc, kfree, kmem_cache_alloc, kmem_cache_free etc.. Collected
  246. data is then fed to the userspace application in order to analyse
  247. allocation hotspots, internal fragmentation and so on, making it
  248. possible to see how well an allocator performs, as well as debug
  249. and profile kernel code.
  250. This requires an userspace application to use. See
  251. Documentation/vm/kmemtrace.txt for more information.
  252. Saying Y will make the kernel somewhat larger and slower. However,
  253. if you disable kmemtrace at run-time or boot-time, the performance
  254. impact is minimal (depending on the arch the kernel is built for).
  255. If unsure, say N.
  256. config WORKQUEUE_TRACER
  257. bool "Trace workqueues"
  258. select TRACING
  259. help
  260. The workqueue tracer provides some statistical informations
  261. about each cpu workqueue thread such as the number of the
  262. works inserted and executed since their creation. It can help
  263. to evaluate the amount of work each of them have to perform.
  264. For example it can help a developer to decide whether he should
  265. choose a per cpu workqueue instead of a singlethreaded one.
  266. config BLK_DEV_IO_TRACE
  267. bool "Support for tracing block io actions"
  268. depends on SYSFS
  269. depends on BLOCK
  270. select RELAY
  271. select DEBUG_FS
  272. select TRACEPOINTS
  273. select TRACING
  274. select STACKTRACE
  275. help
  276. Say Y here if you want to be able to trace the block layer actions
  277. on a given queue. Tracing allows you to see any traffic happening
  278. on a block device queue. For more information (and the userspace
  279. support tools needed), fetch the blktrace tools from:
  280. git://git.kernel.dk/blktrace.git
  281. Tracing also is possible using the ftrace interface, e.g.:
  282. echo 1 > /sys/block/sda/sda1/trace/enable
  283. echo blk > /sys/kernel/debug/tracing/current_tracer
  284. cat /sys/kernel/debug/tracing/trace_pipe
  285. If unsure, say N.
  286. config DYNAMIC_FTRACE
  287. bool "enable/disable ftrace tracepoints dynamically"
  288. depends on FUNCTION_TRACER
  289. depends on HAVE_DYNAMIC_FTRACE
  290. default y
  291. help
  292. This option will modify all the calls to ftrace dynamically
  293. (will patch them out of the binary image and replaces them
  294. with a No-Op instruction) as they are called. A table is
  295. created to dynamically enable them again.
  296. This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but otherwise
  297. has native performance as long as no tracing is active.
  298. The changes to the code are done by a kernel thread that
  299. wakes up once a second and checks to see if any ftrace calls
  300. were made. If so, it runs stop_machine (stops all CPUS)
  301. and modifies the code to jump over the call to ftrace.
  302. config FTRACE_MCOUNT_RECORD
  303. def_bool y
  304. depends on DYNAMIC_FTRACE
  305. depends on HAVE_FTRACE_MCOUNT_RECORD
  306. config FTRACE_SELFTEST
  307. bool
  308. config FTRACE_STARTUP_TEST
  309. bool "Perform a startup test on ftrace"
  310. depends on TRACING
  311. select FTRACE_SELFTEST
  312. help
  313. This option performs a series of startup tests on ftrace. On bootup
  314. a series of tests are made to verify that the tracer is
  315. functioning properly. It will do tests on all the configured
  316. tracers of ftrace.
  317. config MMIOTRACE
  318. bool "Memory mapped IO tracing"
  319. depends on HAVE_MMIOTRACE_SUPPORT && PCI
  320. select TRACING
  321. help
  322. Mmiotrace traces Memory Mapped I/O access and is meant for
  323. debugging and reverse engineering. It is called from the ioremap
  324. implementation and works via page faults. Tracing is disabled by
  325. default and can be enabled at run-time.
  326. See Documentation/tracers/mmiotrace.txt.
  327. If you are not helping to develop drivers, say N.
  328. config MMIOTRACE_TEST
  329. tristate "Test module for mmiotrace"
  330. depends on MMIOTRACE && m
  331. help
  332. This is a dumb module for testing mmiotrace. It is very dangerous
  333. as it will write garbage to IO memory starting at a given address.
  334. However, it should be safe to use on e.g. unused portion of VRAM.
  335. Say N, unless you absolutely know what you are doing.
  336. endmenu
  337. endif # TRACING_SUPPORT