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 TRACER_MAX_TRACE
43 config FTRACE_NMI_ENTER
45 depends on HAVE_FTRACE_NMI_ENTER
52 select STACKTRACE if STACKTRACE_SUPPORT
57 # Minimum requirements an architecture has to meet for us to
58 # be able to offer generic tracing facilities:
60 config TRACING_SUPPORT
62 depends on TRACE_IRQFLAGS_SUPPORT
63 depends on STACKTRACE_SUPPORT
70 config FUNCTION_TRACER
71 bool "Kernel Function Tracer"
72 depends on HAVE_FUNCTION_TRACER
76 select CONTEXT_SWITCH_TRACER
78 Enable the kernel to trace every kernel function. This is done
79 by using a compiler feature to insert a small, 5-byte No-Operation
80 instruction to the beginning of every kernel function, which NOP
81 sequence is then dynamically patched into a tracer call when
82 tracing is enabled by the administrator. If it's runtime disabled
83 (the bootup default), then the overhead of the instructions is very
84 small and not measurable even in micro-benchmarks.
86 config FUNCTION_GRAPH_TRACER
87 bool "Kernel Function Graph Tracer"
88 depends on HAVE_FUNCTION_GRAPH_TRACER
89 depends on FUNCTION_TRACER
92 Enable the kernel to trace a function at both its return
94 It's first purpose is to trace the duration of functions and
95 draw a call graph for each thread with some informations like
97 This is done by setting the current return address on the current
98 task structure into a stack of calls.
100 config IRQSOFF_TRACER
101 bool "Interrupts-off Latency Tracer"
103 depends on TRACE_IRQFLAGS_SUPPORT
104 depends on GENERIC_TIME
105 select TRACE_IRQFLAGS
107 select TRACER_MAX_TRACE
109 This option measures the time spent in irqs-off critical
110 sections, with microsecond accuracy.
112 The default measurement method is a maximum search, which is
113 disabled by default and can be runtime (re-)started
116 echo 0 > /debugfs/tracing/tracing_max_latency
118 (Note that kernel size and overhead increases with this option
119 enabled. This option and the preempt-off timing option can be
120 used together or separately.)
122 config PREEMPT_TRACER
123 bool "Preemption-off Latency Tracer"
125 depends on GENERIC_TIME
128 select TRACER_MAX_TRACE
130 This option measures the time spent in preemption off critical
131 sections, with microsecond accuracy.
133 The default measurement method is a maximum search, which is
134 disabled by default and can be runtime (re-)started
137 echo 0 > /debugfs/tracing/tracing_max_latency
139 (Note that kernel size and overhead increases with this option
140 enabled. This option and the irqs-off timing option can be
141 used together or separately.)
143 config SYSPROF_TRACER
144 bool "Sysprof Tracer"
147 select CONTEXT_SWITCH_TRACER
149 This tracer provides the trace needed by the 'Sysprof' userspace
153 bool "Scheduling Latency Tracer"
155 select CONTEXT_SWITCH_TRACER
156 select TRACER_MAX_TRACE
158 This tracer tracks the latency of the highest priority task
159 to be scheduled in, starting from the point it has woken up.
161 config CONTEXT_SWITCH_TRACER
162 bool "Trace process context switches"
166 This tracer gets called from the context switch and records
167 all switching of tasks.
170 bool "Trace various events in the kernel"
173 This tracer hooks to various trace points in the kernel
174 allowing the user to pick and choose which trace point they
178 bool "Trace boot initcalls"
180 select CONTEXT_SWITCH_TRACER
182 This tracer helps developers to optimize boot times: it records
183 the timings of the initcalls and traces key events and the identity
184 of tasks that can cause boot delays, such as context-switches.
186 Its aim is to be parsed by the /scripts/bootgraph.pl tool to
187 produce pretty graphics about boot inefficiencies, giving a visual
188 representation of the delays during initcalls - but the raw
189 /debug/tracing/trace text output is readable too.
191 You must pass in ftrace=initcall to the kernel command line
192 to enable this on bootup.
194 config TRACE_BRANCH_PROFILING
195 bool "Trace likely/unlikely profiler"
198 This tracer profiles all the the likely and unlikely macros
199 in the kernel. It will display the results in:
201 /debugfs/tracing/profile_annotated_branch
203 Note: this will add a significant overhead, only turn this
204 on if you need to profile the system's use of these macros.
208 config PROFILE_ALL_BRANCHES
209 bool "Profile all if conditionals"
210 depends on TRACE_BRANCH_PROFILING
212 This tracer profiles all branch conditions. Every if ()
213 taken in the kernel is recorded whether it hit or miss.
214 The results will be displayed in:
216 /debugfs/tracing/profile_branch
218 This configuration, when enabled, will impose a great overhead
219 on the system. This should only be enabled when the system
224 config TRACING_BRANCHES
227 Selected by tracers that will trace the likely and unlikely
228 conditions. This prevents the tracers themselves from being
229 profiled. Profiling the tracing infrastructure can only happen
230 when the likelys and unlikelys are not being traced.
233 bool "Trace likely/unlikely instances"
234 depends on TRACE_BRANCH_PROFILING
235 select TRACING_BRANCHES
237 This traces the events of likely and unlikely condition
238 calls in the kernel. The difference between this and the
239 "Trace likely/unlikely profiler" is that this is not a
240 histogram of the callers, but actually places the calling
241 events into a running trace buffer to see when and where the
242 events happened, as well as their results.
247 bool "Trace power consumption behavior"
251 This tracer helps developers to analyze and optimize the kernels
252 power management decisions, specifically the C-state and P-state
257 bool "Trace max stack"
258 depends on HAVE_FUNCTION_TRACER
259 select FUNCTION_TRACER
263 This special tracer records the maximum stack footprint of the
264 kernel and displays it in debugfs/tracing/stack_trace.
266 This tracer works by hooking into every function call that the
267 kernel executes, and keeping a maximum stack depth value and
268 stack-trace saved. If this is configured with DYNAMIC_FTRACE
269 then it will not have any overhead while the stack tracer
272 To enable the stack tracer on bootup, pass in 'stacktrace'
273 on the kernel command line.
275 The stack tracer can also be enabled or disabled via the
276 sysctl kernel.stack_tracer_enabled
280 config HW_BRANCH_TRACER
281 depends on HAVE_HW_BRANCH_TRACER
282 bool "Trace hw branches"
285 This tracer records all branches on the system in a circular
286 buffer giving access to the last N branches for each cpu.
289 bool "Trace SLAB allocations"
292 kmemtrace provides tracing for slab allocator functions, such as
293 kmalloc, kfree, kmem_cache_alloc, kmem_cache_free etc.. Collected
294 data is then fed to the userspace application in order to analyse
295 allocation hotspots, internal fragmentation and so on, making it
296 possible to see how well an allocator performs, as well as debug
297 and profile kernel code.
299 This requires an userspace application to use. See
300 Documentation/vm/kmemtrace.txt for more information.
302 Saying Y will make the kernel somewhat larger and slower. However,
303 if you disable kmemtrace at run-time or boot-time, the performance
304 impact is minimal (depending on the arch the kernel is built for).
308 config WORKQUEUE_TRACER
309 bool "Trace workqueues"
312 The workqueue tracer provides some statistical informations
313 about each cpu workqueue thread such as the number of the
314 works inserted and executed since their creation. It can help
315 to evaluate the amount of work each of them have to perform.
316 For example it can help a developer to decide whether he should
317 choose a per cpu workqueue instead of a singlethreaded one.
319 config BLK_DEV_IO_TRACE
320 bool "Support for tracing block io actions"
329 Say Y here if you want to be able to trace the block layer actions
330 on a given queue. Tracing allows you to see any traffic happening
331 on a block device queue. For more information (and the userspace
332 support tools needed), fetch the blktrace tools from:
334 git://git.kernel.dk/blktrace.git
336 Tracing also is possible using the ftrace interface, e.g.:
338 echo 1 > /sys/block/sda/sda1/trace/enable
339 echo blk > /sys/kernel/debug/tracing/current_tracer
340 cat /sys/kernel/debug/tracing/trace_pipe
344 config DYNAMIC_FTRACE
345 bool "enable/disable ftrace tracepoints dynamically"
346 depends on FUNCTION_TRACER
347 depends on HAVE_DYNAMIC_FTRACE
350 This option will modify all the calls to ftrace dynamically
351 (will patch them out of the binary image and replaces them
352 with a No-Op instruction) as they are called. A table is
353 created to dynamically enable them again.
355 This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but otherwise
356 has native performance as long as no tracing is active.
358 The changes to the code are done by a kernel thread that
359 wakes up once a second and checks to see if any ftrace calls
360 were made. If so, it runs stop_machine (stops all CPUS)
361 and modifies the code to jump over the call to ftrace.
363 config FTRACE_MCOUNT_RECORD
365 depends on DYNAMIC_FTRACE
366 depends on HAVE_FTRACE_MCOUNT_RECORD
368 config FTRACE_SELFTEST
371 config FTRACE_STARTUP_TEST
372 bool "Perform a startup test on ftrace"
374 select FTRACE_SELFTEST
376 This option performs a series of startup tests on ftrace. On bootup
377 a series of tests are made to verify that the tracer is
378 functioning properly. It will do tests on all the configured
382 bool "Memory mapped IO tracing"
383 depends on HAVE_MMIOTRACE_SUPPORT && PCI
386 Mmiotrace traces Memory Mapped I/O access and is meant for
387 debugging and reverse engineering. It is called from the ioremap
388 implementation and works via page faults. Tracing is disabled by
389 default and can be enabled at run-time.
391 See Documentation/tracers/mmiotrace.txt.
392 If you are not helping to develop drivers, say N.
394 config MMIOTRACE_TEST
395 tristate "Test module for mmiotrace"
396 depends on MMIOTRACE && m
398 This is a dumb module for testing mmiotrace. It is very dangerous
399 as it will write garbage to IO memory starting at a given address.
400 However, it should be safe to use on e.g. unused portion of VRAM.
402 Say N, unless you absolutely know what you are doing.
406 endif # TRACING_SUPPORT