Kconfig 12 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408
  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. select BINARY_PRINTF
  43. #
  44. # Minimum requirements an architecture has to meet for us to
  45. # be able to offer generic tracing facilities:
  46. #
  47. config TRACING_SUPPORT
  48. bool
  49. depends on TRACE_IRQFLAGS_SUPPORT
  50. depends on STACKTRACE_SUPPORT
  51. default y
  52. if TRACING_SUPPORT
  53. menu "Tracers"
  54. config FUNCTION_TRACER
  55. bool "Kernel Function Tracer"
  56. depends on HAVE_FUNCTION_TRACER
  57. select FRAME_POINTER
  58. select KALLSYMS
  59. select TRACING
  60. select CONTEXT_SWITCH_TRACER
  61. help
  62. Enable the kernel to trace every kernel function. This is done
  63. by using a compiler feature to insert a small, 5-byte No-Operation
  64. instruction to the beginning of every kernel function, which NOP
  65. sequence is then dynamically patched into a tracer call when
  66. tracing is enabled by the administrator. If it's runtime disabled
  67. (the bootup default), then the overhead of the instructions is very
  68. small and not measurable even in micro-benchmarks.
  69. config FUNCTION_GRAPH_TRACER
  70. bool "Kernel Function Graph Tracer"
  71. depends on HAVE_FUNCTION_GRAPH_TRACER
  72. depends on FUNCTION_TRACER
  73. default y
  74. help
  75. Enable the kernel to trace a function at both its return
  76. and its entry.
  77. It's first purpose is to trace the duration of functions and
  78. draw a call graph for each thread with some informations like
  79. the return value.
  80. This is done by setting the current return address on the current
  81. task structure into a stack of calls.
  82. config IRQSOFF_TRACER
  83. bool "Interrupts-off Latency Tracer"
  84. default n
  85. depends on TRACE_IRQFLAGS_SUPPORT
  86. depends on GENERIC_TIME
  87. select TRACE_IRQFLAGS
  88. select TRACING
  89. select TRACER_MAX_TRACE
  90. help
  91. This option measures the time spent in irqs-off critical
  92. sections, with microsecond accuracy.
  93. The default measurement method is a maximum search, which is
  94. disabled by default and can be runtime (re-)started
  95. via:
  96. echo 0 > /debugfs/tracing/tracing_max_latency
  97. (Note that kernel size and overhead increases with this option
  98. enabled. This option and the preempt-off timing option can be
  99. used together or separately.)
  100. config PREEMPT_TRACER
  101. bool "Preemption-off Latency Tracer"
  102. default n
  103. depends on GENERIC_TIME
  104. depends on PREEMPT
  105. select TRACING
  106. select TRACER_MAX_TRACE
  107. help
  108. This option measures the time spent in preemption off critical
  109. sections, with microsecond accuracy.
  110. The default measurement method is a maximum search, which is
  111. disabled by default and can be runtime (re-)started
  112. via:
  113. echo 0 > /debugfs/tracing/tracing_max_latency
  114. (Note that kernel size and overhead increases with this option
  115. enabled. This option and the irqs-off timing option can be
  116. used together or separately.)
  117. config SYSPROF_TRACER
  118. bool "Sysprof Tracer"
  119. depends on X86
  120. select TRACING
  121. select CONTEXT_SWITCH_TRACER
  122. help
  123. This tracer provides the trace needed by the 'Sysprof' userspace
  124. tool.
  125. config SCHED_TRACER
  126. bool "Scheduling Latency Tracer"
  127. select TRACING
  128. select CONTEXT_SWITCH_TRACER
  129. select TRACER_MAX_TRACE
  130. help
  131. This tracer tracks the latency of the highest priority task
  132. to be scheduled in, starting from the point it has woken up.
  133. config CONTEXT_SWITCH_TRACER
  134. bool "Trace process context switches"
  135. select TRACING
  136. select MARKERS
  137. help
  138. This tracer gets called from the context switch and records
  139. all switching of tasks.
  140. config EVENT_TRACER
  141. bool "Trace various events in the kernel"
  142. select TRACING
  143. help
  144. This tracer hooks to various trace points in the kernel
  145. allowing the user to pick and choose which trace point they
  146. want to trace.
  147. config BOOT_TRACER
  148. bool "Trace boot initcalls"
  149. select TRACING
  150. select CONTEXT_SWITCH_TRACER
  151. help
  152. This tracer helps developers to optimize boot times: it records
  153. the timings of the initcalls and traces key events and the identity
  154. of tasks that can cause boot delays, such as context-switches.
  155. Its aim is to be parsed by the /scripts/bootgraph.pl tool to
  156. produce pretty graphics about boot inefficiencies, giving a visual
  157. representation of the delays during initcalls - but the raw
  158. /debug/tracing/trace text output is readable too.
  159. You must pass in ftrace=initcall to the kernel command line
  160. to enable this on bootup.
  161. config TRACE_BRANCH_PROFILING
  162. bool "Trace likely/unlikely profiler"
  163. select TRACING
  164. help
  165. This tracer profiles all the the likely and unlikely macros
  166. in the kernel. It will display the results in:
  167. /debugfs/tracing/profile_annotated_branch
  168. Note: this will add a significant overhead, only turn this
  169. on if you need to profile the system's use of these macros.
  170. Say N if unsure.
  171. config PROFILE_ALL_BRANCHES
  172. bool "Profile all if conditionals"
  173. depends on TRACE_BRANCH_PROFILING
  174. help
  175. This tracer profiles all branch conditions. Every if ()
  176. taken in the kernel is recorded whether it hit or miss.
  177. The results will be displayed in:
  178. /debugfs/tracing/profile_branch
  179. This configuration, when enabled, will impose a great overhead
  180. on the system. This should only be enabled when the system
  181. is to be analyzed
  182. Say N if unsure.
  183. config TRACING_BRANCHES
  184. bool
  185. help
  186. Selected by tracers that will trace the likely and unlikely
  187. conditions. This prevents the tracers themselves from being
  188. profiled. Profiling the tracing infrastructure can only happen
  189. when the likelys and unlikelys are not being traced.
  190. config BRANCH_TRACER
  191. bool "Trace likely/unlikely instances"
  192. depends on TRACE_BRANCH_PROFILING
  193. select TRACING_BRANCHES
  194. help
  195. This traces the events of likely and unlikely condition
  196. calls in the kernel. The difference between this and the
  197. "Trace likely/unlikely profiler" is that this is not a
  198. histogram of the callers, but actually places the calling
  199. events into a running trace buffer to see when and where the
  200. events happened, as well as their results.
  201. Say N if unsure.
  202. config POWER_TRACER
  203. bool "Trace power consumption behavior"
  204. depends on X86
  205. select TRACING
  206. help
  207. This tracer helps developers to analyze and optimize the kernels
  208. power management decisions, specifically the C-state and P-state
  209. behavior.
  210. config STACK_TRACER
  211. bool "Trace max stack"
  212. depends on HAVE_FUNCTION_TRACER
  213. select FUNCTION_TRACER
  214. select STACKTRACE
  215. select KALLSYMS
  216. help
  217. This special tracer records the maximum stack footprint of the
  218. kernel and displays it in debugfs/tracing/stack_trace.
  219. This tracer works by hooking into every function call that the
  220. kernel executes, and keeping a maximum stack depth value and
  221. stack-trace saved. If this is configured with DYNAMIC_FTRACE
  222. then it will not have any overhead while the stack tracer
  223. is disabled.
  224. To enable the stack tracer on bootup, pass in 'stacktrace'
  225. on the kernel command line.
  226. The stack tracer can also be enabled or disabled via the
  227. sysctl kernel.stack_tracer_enabled
  228. Say N if unsure.
  229. config HW_BRANCH_TRACER
  230. depends on HAVE_HW_BRANCH_TRACER
  231. bool "Trace hw branches"
  232. select TRACING
  233. help
  234. This tracer records all branches on the system in a circular
  235. buffer giving access to the last N branches for each cpu.
  236. config KMEMTRACE
  237. bool "Trace SLAB allocations"
  238. select TRACING
  239. help
  240. kmemtrace provides tracing for slab allocator functions, such as
  241. kmalloc, kfree, kmem_cache_alloc, kmem_cache_free etc.. Collected
  242. data is then fed to the userspace application in order to analyse
  243. allocation hotspots, internal fragmentation and so on, making it
  244. possible to see how well an allocator performs, as well as debug
  245. and profile kernel code.
  246. This requires an userspace application to use. See
  247. Documentation/vm/kmemtrace.txt for more information.
  248. Saying Y will make the kernel somewhat larger and slower. However,
  249. if you disable kmemtrace at run-time or boot-time, the performance
  250. impact is minimal (depending on the arch the kernel is built for).
  251. If unsure, say N.
  252. config WORKQUEUE_TRACER
  253. bool "Trace workqueues"
  254. select TRACING
  255. help
  256. The workqueue tracer provides some statistical informations
  257. about each cpu workqueue thread such as the number of the
  258. works inserted and executed since their creation. It can help
  259. to evaluate the amount of work each of them have to perform.
  260. For example it can help a developer to decide whether he should
  261. choose a per cpu workqueue instead of a singlethreaded one.
  262. config BLK_DEV_IO_TRACE
  263. bool "Support for tracing block io actions"
  264. depends on SYSFS
  265. depends on BLOCK
  266. select RELAY
  267. select DEBUG_FS
  268. select TRACEPOINTS
  269. select TRACING
  270. select STACKTRACE
  271. help
  272. Say Y here if you want to be able to trace the block layer actions
  273. on a given queue. Tracing allows you to see any traffic happening
  274. on a block device queue. For more information (and the userspace
  275. support tools needed), fetch the blktrace tools from:
  276. git://git.kernel.dk/blktrace.git
  277. Tracing also is possible using the ftrace interface, e.g.:
  278. echo 1 > /sys/block/sda/sda1/trace/enable
  279. echo blk > /sys/kernel/debug/tracing/current_tracer
  280. cat /sys/kernel/debug/tracing/trace_pipe
  281. If unsure, say N.
  282. config DYNAMIC_FTRACE
  283. bool "enable/disable ftrace tracepoints dynamically"
  284. depends on FUNCTION_TRACER
  285. depends on HAVE_DYNAMIC_FTRACE
  286. default y
  287. help
  288. This option will modify all the calls to ftrace dynamically
  289. (will patch them out of the binary image and replaces them
  290. with a No-Op instruction) as they are called. A table is
  291. created to dynamically enable them again.
  292. This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but otherwise
  293. has native performance as long as no tracing is active.
  294. The changes to the code are done by a kernel thread that
  295. wakes up once a second and checks to see if any ftrace calls
  296. were made. If so, it runs stop_machine (stops all CPUS)
  297. and modifies the code to jump over the call to ftrace.
  298. config FTRACE_MCOUNT_RECORD
  299. def_bool y
  300. depends on DYNAMIC_FTRACE
  301. depends on HAVE_FTRACE_MCOUNT_RECORD
  302. config FTRACE_SELFTEST
  303. bool
  304. config FTRACE_STARTUP_TEST
  305. bool "Perform a startup test on ftrace"
  306. depends on TRACING
  307. select FTRACE_SELFTEST
  308. help
  309. This option performs a series of startup tests on ftrace. On bootup
  310. a series of tests are made to verify that the tracer is
  311. functioning properly. It will do tests on all the configured
  312. tracers of ftrace.
  313. config MMIOTRACE
  314. bool "Memory mapped IO tracing"
  315. depends on HAVE_MMIOTRACE_SUPPORT && PCI
  316. select TRACING
  317. help
  318. Mmiotrace traces Memory Mapped I/O access and is meant for
  319. debugging and reverse engineering. It is called from the ioremap
  320. implementation and works via page faults. Tracing is disabled by
  321. default and can be enabled at run-time.
  322. See Documentation/tracers/mmiotrace.txt.
  323. If you are not helping to develop drivers, say N.
  324. config MMIOTRACE_TEST
  325. tristate "Test module for mmiotrace"
  326. depends on MMIOTRACE && m
  327. help
  328. This is a dumb module for testing mmiotrace. It is very dangerous
  329. as it will write garbage to IO memory starting at a given address.
  330. However, it should be safe to use on e.g. unused portion of VRAM.
  331. Say N, unless you absolutely know what you are doing.
  332. endmenu
  333. endif # TRACING_SUPPORT