2 # Architectures that offer an FUNCTION_TRACER implementation should
3 # select HAVE_FUNCTION_TRACER:
6 config USER_STACKTRACE_SUPPORT
12 config HAVE_FTRACE_NMI_ENTER
15 config HAVE_FUNCTION_TRACER
18 config HAVE_FUNCTION_GRAPH_TRACER
21 config HAVE_FUNCTION_TRACE_MCOUNT_TEST
24 This gets selected when the arch tests the function_trace_stop
25 variable at the mcount call site. Otherwise, this variable
26 is tested by the called function.
28 config HAVE_DYNAMIC_FTRACE
31 config HAVE_FTRACE_MCOUNT_RECORD
34 config HAVE_HW_BRANCH_TRACER
37 config HAVE_FTRACE_SYSCALLS
40 config TRACER_MAX_TRACE
46 config FTRACE_NMI_ENTER
48 depends on HAVE_FTRACE_NMI_ENTER
52 select CONTEXT_SWITCH_TRACER
55 config CONTEXT_SWITCH_TRACER
59 # All tracer options should select GENERIC_TRACER. For those options that are
60 # enabled by all tracers (context switch and event tracer) they select TRACING.
61 # This allows those options to appear when no other tracer is selected. But the
62 # options do not appear when something else selects it. We need the two options
63 # GENERIC_TRACER and TRACING to avoid circular dependencies to accomplish the
64 # hidding of the automatic options options.
70 select STACKTRACE if STACKTRACE_SUPPORT
81 # Minimum requirements an architecture has to meet for us to
82 # be able to offer generic tracing facilities:
84 config TRACING_SUPPORT
86 # PPC32 has no irqflags tracing support, but it can use most of the
87 # tracers anyway, they were tested to build and work. Note that new
88 # exceptions to this list aren't welcomed, better implement the
89 # irqflags tracing for your architecture.
90 depends on TRACE_IRQFLAGS_SUPPORT || PPC32
91 depends on STACKTRACE_SUPPORT
98 default y if DEBUG_KERNEL
100 Enable the kernel tracing infrastructure.
104 config FUNCTION_TRACER
105 bool "Kernel Function Tracer"
106 depends on HAVE_FUNCTION_TRACER
109 select GENERIC_TRACER
110 select CONTEXT_SWITCH_TRACER
112 Enable the kernel to trace every kernel function. This is done
113 by using a compiler feature to insert a small, 5-byte No-Operation
114 instruction to the beginning of every kernel function, which NOP
115 sequence is then dynamically patched into a tracer call when
116 tracing is enabled by the administrator. If it's runtime disabled
117 (the bootup default), then the overhead of the instructions is very
118 small and not measurable even in micro-benchmarks.
120 config FUNCTION_GRAPH_TRACER
121 bool "Kernel Function Graph Tracer"
122 depends on HAVE_FUNCTION_GRAPH_TRACER
123 depends on FUNCTION_TRACER
126 Enable the kernel to trace a function at both its return
128 Its first purpose is to trace the duration of functions and
129 draw a call graph for each thread with some information like
130 the return value. This is done by setting the current return
131 address on the current task structure into a stack of calls.
134 config IRQSOFF_TRACER
135 bool "Interrupts-off Latency Tracer"
137 depends on TRACE_IRQFLAGS_SUPPORT
138 depends on GENERIC_TIME
139 select TRACE_IRQFLAGS
140 select GENERIC_TRACER
141 select TRACER_MAX_TRACE
143 This option measures the time spent in irqs-off critical
144 sections, with microsecond accuracy.
146 The default measurement method is a maximum search, which is
147 disabled by default and can be runtime (re-)started
150 echo 0 > /debugfs/tracing/tracing_max_latency
152 (Note that kernel size and overhead increases with this option
153 enabled. This option and the preempt-off timing option can be
154 used together or separately.)
156 config PREEMPT_TRACER
157 bool "Preemption-off Latency Tracer"
159 depends on GENERIC_TIME
161 select GENERIC_TRACER
162 select TRACER_MAX_TRACE
164 This option measures the time spent in preemption off critical
165 sections, with microsecond accuracy.
167 The default measurement method is a maximum search, which is
168 disabled by default and can be runtime (re-)started
171 echo 0 > /debugfs/tracing/tracing_max_latency
173 (Note that kernel size and overhead increases with this option
174 enabled. This option and the irqs-off timing option can be
175 used together or separately.)
177 config SYSPROF_TRACER
178 bool "Sysprof Tracer"
180 select GENERIC_TRACER
181 select CONTEXT_SWITCH_TRACER
183 This tracer provides the trace needed by the 'Sysprof' userspace
187 bool "Scheduling Latency Tracer"
188 select GENERIC_TRACER
189 select CONTEXT_SWITCH_TRACER
190 select TRACER_MAX_TRACE
192 This tracer tracks the latency of the highest priority task
193 to be scheduled in, starting from the point it has woken up.
195 config ENABLE_CONTEXT_SWITCH_TRACER
196 bool "Trace process context switches"
197 depends on !GENERIC_TRACER
199 select CONTEXT_SWITCH_TRACER
201 This tracer gets called from the context switch and records
202 all switching of tasks.
204 config ENABLE_EVENT_TRACING
205 bool "Trace various events in the kernel"
206 depends on !GENERIC_TRACER
209 This tracer hooks to various trace points in the kernel
210 allowing the user to pick and choose which trace point they
213 Note, all tracers enable event tracing. This option is
214 only a convenience to enable event tracing when no other
215 tracers are selected.
217 config FTRACE_SYSCALLS
218 bool "Trace syscalls"
219 depends on HAVE_FTRACE_SYSCALLS
220 select GENERIC_TRACER
223 Basic tracer to catch the syscall entry and exit events.
226 bool "Trace boot initcalls"
227 select GENERIC_TRACER
228 select CONTEXT_SWITCH_TRACER
230 This tracer helps developers to optimize boot times: it records
231 the timings of the initcalls and traces key events and the identity
232 of tasks that can cause boot delays, such as context-switches.
234 Its aim is to be parsed by the /scripts/bootgraph.pl tool to
235 produce pretty graphics about boot inefficiencies, giving a visual
236 representation of the delays during initcalls - but the raw
237 /debug/tracing/trace text output is readable too.
239 You must pass in ftrace=initcall to the kernel command line
240 to enable this on bootup.
242 config TRACE_BRANCH_PROFILING
244 select GENERIC_TRACER
247 prompt "Branch Profiling"
248 default BRANCH_PROFILE_NONE
250 The branch profiling is a software profiler. It will add hooks
251 into the C conditionals to test which path a branch takes.
253 The likely/unlikely profiler only looks at the conditions that
254 are annotated with a likely or unlikely macro.
256 The "all branch" profiler will profile every if statement in the
257 kernel. This profiler will also enable the likely/unlikely
260 Either of the above profilers add a bit of overhead to the system.
261 If unsure choose "No branch profiling".
263 config BRANCH_PROFILE_NONE
264 bool "No branch profiling"
266 No branch profiling. Branch profiling adds a bit of overhead.
267 Only enable it if you want to analyse the branching behavior.
268 Otherwise keep it disabled.
270 config PROFILE_ANNOTATED_BRANCHES
271 bool "Trace likely/unlikely profiler"
272 select TRACE_BRANCH_PROFILING
274 This tracer profiles all the the likely and unlikely macros
275 in the kernel. It will display the results in:
277 /debugfs/tracing/profile_annotated_branch
279 Note: this will add a significant overhead, only turn this
280 on if you need to profile the system's use of these macros.
282 config PROFILE_ALL_BRANCHES
283 bool "Profile all if conditionals"
284 select TRACE_BRANCH_PROFILING
286 This tracer profiles all branch conditions. Every if ()
287 taken in the kernel is recorded whether it hit or miss.
288 The results will be displayed in:
290 /debugfs/tracing/profile_branch
292 This option also enables the likely/unlikely profiler.
294 This configuration, when enabled, will impose a great overhead
295 on the system. This should only be enabled when the system
299 config TRACING_BRANCHES
302 Selected by tracers that will trace the likely and unlikely
303 conditions. This prevents the tracers themselves from being
304 profiled. Profiling the tracing infrastructure can only happen
305 when the likelys and unlikelys are not being traced.
308 bool "Trace likely/unlikely instances"
309 depends on TRACE_BRANCH_PROFILING
310 select TRACING_BRANCHES
312 This traces the events of likely and unlikely condition
313 calls in the kernel. The difference between this and the
314 "Trace likely/unlikely profiler" is that this is not a
315 histogram of the callers, but actually places the calling
316 events into a running trace buffer to see when and where the
317 events happened, as well as their results.
322 bool "Trace power consumption behavior"
324 select GENERIC_TRACER
326 This tracer helps developers to analyze and optimize the kernels
327 power management decisions, specifically the C-state and P-state
332 bool "Trace max stack"
333 depends on HAVE_FUNCTION_TRACER
334 select FUNCTION_TRACER
338 This special tracer records the maximum stack footprint of the
339 kernel and displays it in debugfs/tracing/stack_trace.
341 This tracer works by hooking into every function call that the
342 kernel executes, and keeping a maximum stack depth value and
343 stack-trace saved. If this is configured with DYNAMIC_FTRACE
344 then it will not have any overhead while the stack tracer
347 To enable the stack tracer on bootup, pass in 'stacktrace'
348 on the kernel command line.
350 The stack tracer can also be enabled or disabled via the
351 sysctl kernel.stack_tracer_enabled
355 config HW_BRANCH_TRACER
356 depends on HAVE_HW_BRANCH_TRACER
357 bool "Trace hw branches"
358 select GENERIC_TRACER
360 This tracer records all branches on the system in a circular
361 buffer giving access to the last N branches for each cpu.
364 bool "Trace SLAB allocations"
365 select GENERIC_TRACER
367 kmemtrace provides tracing for slab allocator functions, such as
368 kmalloc, kfree, kmem_cache_alloc, kmem_cache_free etc.. Collected
369 data is then fed to the userspace application in order to analyse
370 allocation hotspots, internal fragmentation and so on, making it
371 possible to see how well an allocator performs, as well as debug
372 and profile kernel code.
374 This requires an userspace application to use. See
375 Documentation/trace/kmemtrace.txt for more information.
377 Saying Y will make the kernel somewhat larger and slower. However,
378 if you disable kmemtrace at run-time or boot-time, the performance
379 impact is minimal (depending on the arch the kernel is built for).
383 config WORKQUEUE_TRACER
384 bool "Trace workqueues"
385 select GENERIC_TRACER
387 The workqueue tracer provides some statistical informations
388 about each cpu workqueue thread such as the number of the
389 works inserted and executed since their creation. It can help
390 to evaluate the amount of work each of them have to perform.
391 For example it can help a developer to decide whether he should
392 choose a per cpu workqueue instead of a singlethreaded one.
394 config BLK_DEV_IO_TRACE
395 bool "Support for tracing block io actions"
401 select GENERIC_TRACER
404 Say Y here if you want to be able to trace the block layer actions
405 on a given queue. Tracing allows you to see any traffic happening
406 on a block device queue. For more information (and the userspace
407 support tools needed), fetch the blktrace tools from:
409 git://git.kernel.dk/blktrace.git
411 Tracing also is possible using the ftrace interface, e.g.:
413 echo 1 > /sys/block/sda/sda1/trace/enable
414 echo blk > /sys/kernel/debug/tracing/current_tracer
415 cat /sys/kernel/debug/tracing/trace_pipe
419 config DYNAMIC_FTRACE
420 bool "enable/disable ftrace tracepoints dynamically"
421 depends on FUNCTION_TRACER
422 depends on HAVE_DYNAMIC_FTRACE
425 This option will modify all the calls to ftrace dynamically
426 (will patch them out of the binary image and replaces them
427 with a No-Op instruction) as they are called. A table is
428 created to dynamically enable them again.
430 This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but otherwise
431 has native performance as long as no tracing is active.
433 The changes to the code are done by a kernel thread that
434 wakes up once a second and checks to see if any ftrace calls
435 were made. If so, it runs stop_machine (stops all CPUS)
436 and modifies the code to jump over the call to ftrace.
438 config FUNCTION_PROFILER
439 bool "Kernel function profiler"
440 depends on FUNCTION_TRACER
443 This option enables the kernel function profiler. A file is created
444 in debugfs called function_profile_enabled which defaults to zero.
445 When a 1 is echoed into this file profiling begins, and when a
446 zero is entered, profiling stops. A file in the trace_stats
447 directory called functions, that show the list of functions that
448 have been hit and their counters.
452 config FTRACE_MCOUNT_RECORD
454 depends on DYNAMIC_FTRACE
455 depends on HAVE_FTRACE_MCOUNT_RECORD
457 config FTRACE_SELFTEST
460 config FTRACE_STARTUP_TEST
461 bool "Perform a startup test on ftrace"
462 depends on GENERIC_TRACER
463 select FTRACE_SELFTEST
465 This option performs a series of startup tests on ftrace. On bootup
466 a series of tests are made to verify that the tracer is
467 functioning properly. It will do tests on all the configured
471 bool "Memory mapped IO tracing"
472 depends on HAVE_MMIOTRACE_SUPPORT && PCI
473 select GENERIC_TRACER
475 Mmiotrace traces Memory Mapped I/O access and is meant for
476 debugging and reverse engineering. It is called from the ioremap
477 implementation and works via page faults. Tracing is disabled by
478 default and can be enabled at run-time.
480 See Documentation/trace/mmiotrace.txt.
481 If you are not helping to develop drivers, say N.
483 config MMIOTRACE_TEST
484 tristate "Test module for mmiotrace"
485 depends on MMIOTRACE && m
487 This is a dumb module for testing mmiotrace. It is very dangerous
488 as it will write garbage to IO memory starting at a given address.
489 However, it should be safe to use on e.g. unused portion of VRAM.
491 Say N, unless you absolutely know what you are doing.
493 config RING_BUFFER_BENCHMARK
494 tristate "Ring buffer benchmark stress tester"
495 depends on RING_BUFFER
497 This option creates a test to stress the ring buffer and bench mark it.
498 It creates its own ring buffer such that it will not interfer with
499 any other users of the ring buffer (such as ftrace). It then creates
500 a producer and consumer that will run for 10 seconds and sleep for
501 10 seconds. Each interval it will print out the number of events
502 it recorded and give a rough estimate of how long each iteration took.
504 It does not disable interrupts or raise its priority, so it may be
505 affected by processes that are running.
511 endif # TRACING_SUPPORT