Kconfig 15 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520
  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 HAVE_FTRACE_SYSCALLS
  28. bool
  29. config TRACER_MAX_TRACE
  30. bool
  31. config RING_BUFFER
  32. bool
  33. config FTRACE_NMI_ENTER
  34. bool
  35. depends on HAVE_FTRACE_NMI_ENTER
  36. default y
  37. config EVENT_TRACING
  38. select CONTEXT_SWITCH_TRACER
  39. bool
  40. config CONTEXT_SWITCH_TRACER
  41. select MARKERS
  42. bool
  43. config TRACING
  44. bool
  45. select DEBUG_FS
  46. select RING_BUFFER
  47. select STACKTRACE if STACKTRACE_SUPPORT
  48. select TRACEPOINTS
  49. select NOP_TRACER
  50. select BINARY_PRINTF
  51. select EVENT_TRACING
  52. #
  53. # Minimum requirements an architecture has to meet for us to
  54. # be able to offer generic tracing facilities:
  55. #
  56. config TRACING_SUPPORT
  57. bool
  58. # PPC32 has no irqflags tracing support, but it can use most of the
  59. # tracers anyway, they were tested to build and work. Note that new
  60. # exceptions to this list aren't welcomed, better implement the
  61. # irqflags tracing for your architecture.
  62. depends on TRACE_IRQFLAGS_SUPPORT || PPC32
  63. depends on STACKTRACE_SUPPORT
  64. default y
  65. if TRACING_SUPPORT
  66. menuconfig FTRACE
  67. bool "Tracers"
  68. default y if DEBUG_KERNEL
  69. help
  70. Enable the kernel tracing infrastructure.
  71. if FTRACE
  72. config FUNCTION_TRACER
  73. bool "Kernel Function Tracer"
  74. depends on HAVE_FUNCTION_TRACER
  75. select FRAME_POINTER
  76. select KALLSYMS
  77. select TRACING
  78. select CONTEXT_SWITCH_TRACER
  79. help
  80. Enable the kernel to trace every kernel function. This is done
  81. by using a compiler feature to insert a small, 5-byte No-Operation
  82. instruction to the beginning of every kernel function, which NOP
  83. sequence is then dynamically patched into a tracer call when
  84. tracing is enabled by the administrator. If it's runtime disabled
  85. (the bootup default), then the overhead of the instructions is very
  86. small and not measurable even in micro-benchmarks.
  87. config FUNCTION_GRAPH_TRACER
  88. bool "Kernel Function Graph Tracer"
  89. depends on HAVE_FUNCTION_GRAPH_TRACER
  90. depends on FUNCTION_TRACER
  91. default y
  92. help
  93. Enable the kernel to trace a function at both its return
  94. and its entry.
  95. Its first purpose is to trace the duration of functions and
  96. draw a call graph for each thread with some information like
  97. the return value. This is done by setting the current return
  98. address on the current task structure into a stack of calls.
  99. config IRQSOFF_TRACER
  100. bool "Interrupts-off Latency Tracer"
  101. default n
  102. depends on TRACE_IRQFLAGS_SUPPORT
  103. depends on GENERIC_TIME
  104. select TRACE_IRQFLAGS
  105. select TRACING
  106. select TRACER_MAX_TRACE
  107. help
  108. This option measures the time spent in irqs-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 preempt-off timing option can be
  116. used together or separately.)
  117. config PREEMPT_TRACER
  118. bool "Preemption-off Latency Tracer"
  119. default n
  120. depends on GENERIC_TIME
  121. depends on PREEMPT
  122. select TRACING
  123. select TRACER_MAX_TRACE
  124. help
  125. This option measures the time spent in preemption off critical
  126. sections, with microsecond accuracy.
  127. The default measurement method is a maximum search, which is
  128. disabled by default and can be runtime (re-)started
  129. via:
  130. echo 0 > /debugfs/tracing/tracing_max_latency
  131. (Note that kernel size and overhead increases with this option
  132. enabled. This option and the irqs-off timing option can be
  133. used together or separately.)
  134. config SYSPROF_TRACER
  135. bool "Sysprof Tracer"
  136. depends on X86
  137. select TRACING
  138. select CONTEXT_SWITCH_TRACER
  139. help
  140. This tracer provides the trace needed by the 'Sysprof' userspace
  141. tool.
  142. config SCHED_TRACER
  143. bool "Scheduling Latency Tracer"
  144. select TRACING
  145. select CONTEXT_SWITCH_TRACER
  146. select TRACER_MAX_TRACE
  147. help
  148. This tracer tracks the latency of the highest priority task
  149. to be scheduled in, starting from the point it has woken up.
  150. config ENABLE_CONTEXT_SWITCH_TRACER
  151. bool "Trace process context switches"
  152. select TRACING
  153. select CONTEXT_SWITCH_TRACER
  154. help
  155. This tracer gets called from the context switch and records
  156. all switching of tasks.
  157. config ENABLE_EVENT_TRACING
  158. bool "Trace various events in the kernel"
  159. select TRACING
  160. help
  161. This tracer hooks to various trace points in the kernel
  162. allowing the user to pick and choose which trace point they
  163. want to trace.
  164. Note, all tracers enable event tracing. This option is
  165. only a convenience to enable event tracing when no other
  166. tracers are selected.
  167. config FTRACE_SYSCALLS
  168. bool "Trace syscalls"
  169. depends on HAVE_FTRACE_SYSCALLS
  170. select TRACING
  171. select KALLSYMS
  172. help
  173. Basic tracer to catch the syscall entry and exit events.
  174. config BOOT_TRACER
  175. bool "Trace boot initcalls"
  176. select TRACING
  177. select CONTEXT_SWITCH_TRACER
  178. help
  179. This tracer helps developers to optimize boot times: it records
  180. the timings of the initcalls and traces key events and the identity
  181. of tasks that can cause boot delays, such as context-switches.
  182. Its aim is to be parsed by the /scripts/bootgraph.pl tool to
  183. produce pretty graphics about boot inefficiencies, giving a visual
  184. representation of the delays during initcalls - but the raw
  185. /debug/tracing/trace text output is readable too.
  186. You must pass in ftrace=initcall to the kernel command line
  187. to enable this on bootup.
  188. config TRACE_BRANCH_PROFILING
  189. bool
  190. select TRACING
  191. choice
  192. prompt "Branch Profiling"
  193. default BRANCH_PROFILE_NONE
  194. help
  195. The branch profiling is a software profiler. It will add hooks
  196. into the C conditionals to test which path a branch takes.
  197. The likely/unlikely profiler only looks at the conditions that
  198. are annotated with a likely or unlikely macro.
  199. The "all branch" profiler will profile every if statement in the
  200. kernel. This profiler will also enable the likely/unlikely
  201. profiler as well.
  202. Either of the above profilers add a bit of overhead to the system.
  203. If unsure choose "No branch profiling".
  204. config BRANCH_PROFILE_NONE
  205. bool "No branch profiling"
  206. help
  207. No branch profiling. Branch profiling adds a bit of overhead.
  208. Only enable it if you want to analyse the branching behavior.
  209. Otherwise keep it disabled.
  210. config PROFILE_ANNOTATED_BRANCHES
  211. bool "Trace likely/unlikely profiler"
  212. select TRACE_BRANCH_PROFILING
  213. help
  214. This tracer profiles all the the likely and unlikely macros
  215. in the kernel. It will display the results in:
  216. /debugfs/tracing/profile_annotated_branch
  217. Note: this will add a significant overhead, only turn this
  218. on if you need to profile the system's use of these macros.
  219. config PROFILE_ALL_BRANCHES
  220. bool "Profile all if conditionals"
  221. select TRACE_BRANCH_PROFILING
  222. help
  223. This tracer profiles all branch conditions. Every if ()
  224. taken in the kernel is recorded whether it hit or miss.
  225. The results will be displayed in:
  226. /debugfs/tracing/profile_branch
  227. This option also enables the likely/unlikely profiler.
  228. This configuration, when enabled, will impose a great overhead
  229. on the system. This should only be enabled when the system
  230. is to be analyzed
  231. endchoice
  232. config TRACING_BRANCHES
  233. bool
  234. help
  235. Selected by tracers that will trace the likely and unlikely
  236. conditions. This prevents the tracers themselves from being
  237. profiled. Profiling the tracing infrastructure can only happen
  238. when the likelys and unlikelys are not being traced.
  239. config BRANCH_TRACER
  240. bool "Trace likely/unlikely instances"
  241. depends on TRACE_BRANCH_PROFILING
  242. select TRACING_BRANCHES
  243. help
  244. This traces the events of likely and unlikely condition
  245. calls in the kernel. The difference between this and the
  246. "Trace likely/unlikely profiler" is that this is not a
  247. histogram of the callers, but actually places the calling
  248. events into a running trace buffer to see when and where the
  249. events happened, as well as their results.
  250. Say N if unsure.
  251. config POWER_TRACER
  252. bool "Trace power consumption behavior"
  253. depends on X86
  254. select TRACING
  255. help
  256. This tracer helps developers to analyze and optimize the kernels
  257. power management decisions, specifically the C-state and P-state
  258. behavior.
  259. config KSYM_TRACER
  260. bool "Trace read and write access on kernel memory locations"
  261. depends on HAVE_HW_BREAKPOINT
  262. select TRACING
  263. help
  264. This tracer helps find read and write operations on any given kernel
  265. symbol i.e. /proc/kallsyms.
  266. config PROFILE_KSYM_TRACER
  267. bool "Profile all kernel memory accesses on 'watched' variables"
  268. depends on KSYM_TRACER
  269. help
  270. This tracer profiles kernel accesses on variables watched through the
  271. ksym tracer ftrace plugin. Depending upon the hardware, all read
  272. and write operations on kernel variables can be monitored for
  273. accesses.
  274. The results will be displayed in:
  275. /debugfs/tracing/profile_ksym
  276. Say N if unsure.
  277. config STACK_TRACER
  278. bool "Trace max stack"
  279. depends on HAVE_FUNCTION_TRACER
  280. select FUNCTION_TRACER
  281. select STACKTRACE
  282. select KALLSYMS
  283. help
  284. This special tracer records the maximum stack footprint of the
  285. kernel and displays it in debugfs/tracing/stack_trace.
  286. This tracer works by hooking into every function call that the
  287. kernel executes, and keeping a maximum stack depth value and
  288. stack-trace saved. If this is configured with DYNAMIC_FTRACE
  289. then it will not have any overhead while the stack tracer
  290. is disabled.
  291. To enable the stack tracer on bootup, pass in 'stacktrace'
  292. on the kernel command line.
  293. The stack tracer can also be enabled or disabled via the
  294. sysctl kernel.stack_tracer_enabled
  295. Say N if unsure.
  296. config HW_BRANCH_TRACER
  297. depends on HAVE_HW_BRANCH_TRACER
  298. bool "Trace hw branches"
  299. select TRACING
  300. help
  301. This tracer records all branches on the system in a circular
  302. buffer giving access to the last N branches for each cpu.
  303. config KMEMTRACE
  304. bool "Trace SLAB allocations"
  305. select TRACING
  306. help
  307. kmemtrace provides tracing for slab allocator functions, such as
  308. kmalloc, kfree, kmem_cache_alloc, kmem_cache_free etc.. Collected
  309. data is then fed to the userspace application in order to analyse
  310. allocation hotspots, internal fragmentation and so on, making it
  311. possible to see how well an allocator performs, as well as debug
  312. and profile kernel code.
  313. This requires an userspace application to use. See
  314. Documentation/trace/kmemtrace.txt for more information.
  315. Saying Y will make the kernel somewhat larger and slower. However,
  316. if you disable kmemtrace at run-time or boot-time, the performance
  317. impact is minimal (depending on the arch the kernel is built for).
  318. If unsure, say N.
  319. config WORKQUEUE_TRACER
  320. bool "Trace workqueues"
  321. select TRACING
  322. help
  323. The workqueue tracer provides some statistical informations
  324. about each cpu workqueue thread such as the number of the
  325. works inserted and executed since their creation. It can help
  326. to evaluate the amount of work each of them have to perform.
  327. For example it can help a developer to decide whether he should
  328. choose a per cpu workqueue instead of a singlethreaded one.
  329. config BLK_DEV_IO_TRACE
  330. bool "Support for tracing block io actions"
  331. depends on SYSFS
  332. depends on BLOCK
  333. select RELAY
  334. select DEBUG_FS
  335. select TRACEPOINTS
  336. select TRACING
  337. select STACKTRACE
  338. help
  339. Say Y here if you want to be able to trace the block layer actions
  340. on a given queue. Tracing allows you to see any traffic happening
  341. on a block device queue. For more information (and the userspace
  342. support tools needed), fetch the blktrace tools from:
  343. git://git.kernel.dk/blktrace.git
  344. Tracing also is possible using the ftrace interface, e.g.:
  345. echo 1 > /sys/block/sda/sda1/trace/enable
  346. echo blk > /sys/kernel/debug/tracing/current_tracer
  347. cat /sys/kernel/debug/tracing/trace_pipe
  348. If unsure, say N.
  349. config DYNAMIC_FTRACE
  350. bool "enable/disable ftrace tracepoints dynamically"
  351. depends on FUNCTION_TRACER
  352. depends on HAVE_DYNAMIC_FTRACE
  353. default y
  354. help
  355. This option will modify all the calls to ftrace dynamically
  356. (will patch them out of the binary image and replaces them
  357. with a No-Op instruction) as they are called. A table is
  358. created to dynamically enable them again.
  359. This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but otherwise
  360. has native performance as long as no tracing is active.
  361. The changes to the code are done by a kernel thread that
  362. wakes up once a second and checks to see if any ftrace calls
  363. were made. If so, it runs stop_machine (stops all CPUS)
  364. and modifies the code to jump over the call to ftrace.
  365. config FUNCTION_PROFILER
  366. bool "Kernel function profiler"
  367. depends on FUNCTION_TRACER
  368. default n
  369. help
  370. This option enables the kernel function profiler. A file is created
  371. in debugfs called function_profile_enabled which defaults to zero.
  372. When a 1 is echoed into this file profiling begins, and when a
  373. zero is entered, profiling stops. A file in the trace_stats
  374. directory called functions, that show the list of functions that
  375. have been hit and their counters.
  376. If in doubt, say N
  377. config FTRACE_MCOUNT_RECORD
  378. def_bool y
  379. depends on DYNAMIC_FTRACE
  380. depends on HAVE_FTRACE_MCOUNT_RECORD
  381. config FTRACE_SELFTEST
  382. bool
  383. config FTRACE_STARTUP_TEST
  384. bool "Perform a startup test on ftrace"
  385. depends on TRACING
  386. select FTRACE_SELFTEST
  387. help
  388. This option performs a series of startup tests on ftrace. On bootup
  389. a series of tests are made to verify that the tracer is
  390. functioning properly. It will do tests on all the configured
  391. tracers of ftrace.
  392. config MMIOTRACE
  393. bool "Memory mapped IO tracing"
  394. depends on HAVE_MMIOTRACE_SUPPORT && PCI
  395. select TRACING
  396. help
  397. Mmiotrace traces Memory Mapped I/O access and is meant for
  398. debugging and reverse engineering. It is called from the ioremap
  399. implementation and works via page faults. Tracing is disabled by
  400. default and can be enabled at run-time.
  401. See Documentation/trace/mmiotrace.txt.
  402. If you are not helping to develop drivers, say N.
  403. config MMIOTRACE_TEST
  404. tristate "Test module for mmiotrace"
  405. depends on MMIOTRACE && m
  406. help
  407. This is a dumb module for testing mmiotrace. It is very dangerous
  408. as it will write garbage to IO memory starting at a given address.
  409. However, it should be safe to use on e.g. unused portion of VRAM.
  410. Say N, unless you absolutely know what you are doing.
  411. config RING_BUFFER_BENCHMARK
  412. tristate "Ring buffer benchmark stress tester"
  413. depends on RING_BUFFER
  414. help
  415. This option creates a test to stress the ring buffer and bench mark it.
  416. It creates its own ring buffer such that it will not interfer with
  417. any other users of the ring buffer (such as ftrace). It then creates
  418. a producer and consumer that will run for 10 seconds and sleep for
  419. 10 seconds. Each interval it will print out the number of events
  420. it recorded and give a rough estimate of how long each iteration took.
  421. It does not disable interrupts or raise its priority, so it may be
  422. affected by processes that are running.
  423. If unsure, say N
  424. endif # FTRACE
  425. endif # TRACING_SUPPORT