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 See Documentation/trace/ftrace-design.txt
17 config HAVE_FUNCTION_TRACER
20 See Documentation/trace/ftrace-design.txt
22 config HAVE_FUNCTION_GRAPH_TRACER
25 See Documentation/trace/ftrace-design.txt
27 config HAVE_FUNCTION_GRAPH_FP_TEST
30 See Documentation/trace/ftrace-design.txt
32 config HAVE_FUNCTION_TRACE_MCOUNT_TEST
35 See Documentation/trace/ftrace-design.txt
37 config HAVE_DYNAMIC_FTRACE
40 See Documentation/trace/ftrace-design.txt
42 config HAVE_DYNAMIC_FTRACE_WITH_REGS
45 config HAVE_FTRACE_MCOUNT_RECORD
48 See Documentation/trace/ftrace-design.txt
50 config HAVE_SYSCALL_TRACEPOINTS
53 See Documentation/trace/ftrace-design.txt
58 Arch supports the gcc options -pg with -mfentry
60 config HAVE_C_RECORDMCOUNT
63 C version of recordmcount available?
65 config TRACER_MAX_TRACE
75 config FTRACE_NMI_ENTER
77 depends on HAVE_FTRACE_NMI_ENTER
81 select CONTEXT_SWITCH_TRACER
84 config CONTEXT_SWITCH_TRACER
87 config RING_BUFFER_ALLOW_SWAP
90 Allow the use of ring_buffer_swap_cpu.
91 Adds a very slight overhead to tracing when enabled.
93 # All tracer options should select GENERIC_TRACER. For those options that are
94 # enabled by all tracers (context switch and event tracer) they select TRACING.
95 # This allows those options to appear when no other tracer is selected. But the
96 # options do not appear when something else selects it. We need the two options
97 # GENERIC_TRACER and TRACING to avoid circular dependencies to accomplish the
98 # hiding of the automatic options.
104 select STACKTRACE if STACKTRACE_SUPPORT
112 config GENERIC_TRACER
117 # Minimum requirements an architecture has to meet for us to
118 # be able to offer generic tracing facilities:
120 config TRACING_SUPPORT
122 # PPC32 has no irqflags tracing support, but it can use most of the
123 # tracers anyway, they were tested to build and work. Note that new
124 # exceptions to this list aren't welcomed, better implement the
125 # irqflags tracing for your architecture.
126 depends on TRACE_IRQFLAGS_SUPPORT || PPC32
127 depends on STACKTRACE_SUPPORT
134 default y if DEBUG_KERNEL
136 Enable the kernel tracing infrastructure.
140 config FUNCTION_TRACER
141 bool "Kernel Function Tracer"
142 depends on HAVE_FUNCTION_TRACER
144 select GENERIC_TRACER
145 select CONTEXT_SWITCH_TRACER
147 Enable the kernel to trace every kernel function. This is done
148 by using a compiler feature to insert a small, 5-byte No-Operation
149 instruction at the beginning of every kernel function, which NOP
150 sequence is then dynamically patched into a tracer call when
151 tracing is enabled by the administrator. If it's runtime disabled
152 (the bootup default), then the overhead of the instructions is very
153 small and not measurable even in micro-benchmarks.
155 config FUNCTION_GRAPH_TRACER
156 bool "Kernel Function Graph Tracer"
157 depends on HAVE_FUNCTION_GRAPH_TRACER
158 depends on FUNCTION_TRACER
159 depends on !X86_32 || !CC_OPTIMIZE_FOR_SIZE
162 Enable the kernel to trace a function at both its return
164 Its first purpose is to trace the duration of functions and
165 draw a call graph for each thread with some information like
166 the return value. This is done by setting the current return
167 address on the current task structure into a stack of calls.
170 config IRQSOFF_TRACER
171 bool "Interrupts-off Latency Tracer"
173 depends on TRACE_IRQFLAGS_SUPPORT
174 depends on !ARCH_USES_GETTIMEOFFSET
175 select TRACE_IRQFLAGS
176 select GENERIC_TRACER
177 select TRACER_MAX_TRACE
178 select RING_BUFFER_ALLOW_SWAP
180 This option measures the time spent in irqs-off critical
181 sections, with microsecond accuracy.
183 The default measurement method is a maximum search, which is
184 disabled by default and can be runtime (re-)started
187 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency
189 (Note that kernel size and overhead increase with this option
190 enabled. This option and the preempt-off timing option can be
191 used together or separately.)
193 config PREEMPT_TRACER
194 bool "Preemption-off Latency Tracer"
196 depends on !ARCH_USES_GETTIMEOFFSET
198 select GENERIC_TRACER
199 select TRACER_MAX_TRACE
200 select RING_BUFFER_ALLOW_SWAP
202 This option measures the time spent in preemption-off critical
203 sections, with microsecond accuracy.
205 The default measurement method is a maximum search, which is
206 disabled by default and can be runtime (re-)started
209 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency
211 (Note that kernel size and overhead increase with this option
212 enabled. This option and the irqs-off timing option can be
213 used together or separately.)
216 bool "Scheduling Latency Tracer"
217 select GENERIC_TRACER
218 select CONTEXT_SWITCH_TRACER
219 select TRACER_MAX_TRACE
221 This tracer tracks the latency of the highest priority task
222 to be scheduled in, starting from the point it has woken up.
224 config ENABLE_DEFAULT_TRACERS
225 bool "Trace process context switches and events"
226 depends on !GENERIC_TRACER
229 This tracer hooks to various trace points in the kernel,
230 allowing the user to pick and choose which trace point they
231 want to trace. It also includes the sched_switch tracer plugin.
233 config FTRACE_SYSCALLS
234 bool "Trace syscalls"
235 depends on HAVE_SYSCALL_TRACEPOINTS
236 select GENERIC_TRACER
239 Basic tracer to catch the syscall entry and exit events.
241 config TRACER_SNAPSHOT
242 bool "Create a snapshot trace buffer"
243 select TRACER_MAX_TRACE
245 Allow tracing users to take snapshot of the current buffer using the
246 ftrace interface, e.g.:
248 echo 1 > /sys/kernel/debug/tracing/snapshot
251 config TRACE_BRANCH_PROFILING
253 select GENERIC_TRACER
256 prompt "Branch Profiling"
257 default BRANCH_PROFILE_NONE
259 The branch profiling is a software profiler. It will add hooks
260 into the C conditionals to test which path a branch takes.
262 The likely/unlikely profiler only looks at the conditions that
263 are annotated with a likely or unlikely macro.
265 The "all branch" profiler will profile every if-statement in the
266 kernel. This profiler will also enable the likely/unlikely
269 Either of the above profilers adds a bit of overhead to the system.
270 If unsure, choose "No branch profiling".
272 config BRANCH_PROFILE_NONE
273 bool "No branch profiling"
275 No branch profiling. Branch profiling adds a bit of overhead.
276 Only enable it if you want to analyse the branching behavior.
277 Otherwise keep it disabled.
279 config PROFILE_ANNOTATED_BRANCHES
280 bool "Trace likely/unlikely profiler"
281 select TRACE_BRANCH_PROFILING
283 This tracer profiles all likely and unlikely macros
284 in the kernel. It will display the results in:
286 /sys/kernel/debug/tracing/trace_stat/branch_annotated
288 Note: this will add a significant overhead; only turn this
289 on if you need to profile the system's use of these macros.
291 config PROFILE_ALL_BRANCHES
292 bool "Profile all if conditionals"
293 select TRACE_BRANCH_PROFILING
295 This tracer profiles all branch conditions. Every if ()
296 taken in the kernel is recorded whether it hit or miss.
297 The results will be displayed in:
299 /sys/kernel/debug/tracing/trace_stat/branch_all
301 This option also enables the likely/unlikely profiler.
303 This configuration, when enabled, will impose a great overhead
304 on the system. This should only be enabled when the system
305 is to be analyzed in much detail.
308 config TRACING_BRANCHES
311 Selected by tracers that will trace the likely and unlikely
312 conditions. This prevents the tracers themselves from being
313 profiled. Profiling the tracing infrastructure can only happen
314 when the likelys and unlikelys are not being traced.
317 bool "Trace likely/unlikely instances"
318 depends on TRACE_BRANCH_PROFILING
319 select TRACING_BRANCHES
321 This traces the events of likely and unlikely condition
322 calls in the kernel. The difference between this and the
323 "Trace likely/unlikely profiler" is that this is not a
324 histogram of the callers, but actually places the calling
325 events into a running trace buffer to see when and where the
326 events happened, as well as their results.
331 bool "Trace max stack"
332 depends on HAVE_FUNCTION_TRACER
333 select FUNCTION_TRACER
337 This special tracer records the maximum stack footprint of the
338 kernel and displays it in /sys/kernel/debug/tracing/stack_trace.
340 This tracer works by hooking into every function call that the
341 kernel executes, and keeping a maximum stack depth value and
342 stack-trace saved. If this is configured with DYNAMIC_FTRACE
343 then it will not have any overhead while the stack tracer
346 To enable the stack tracer on bootup, pass in 'stacktrace'
347 on the kernel command line.
349 The stack tracer can also be enabled or disabled via the
350 sysctl kernel.stack_tracer_enabled
354 config BLK_DEV_IO_TRACE
355 bool "Support for tracing block IO actions"
361 select GENERIC_TRACER
364 Say Y here if you want to be able to trace the block layer actions
365 on a given queue. Tracing allows you to see any traffic happening
366 on a block device queue. For more information (and the userspace
367 support tools needed), fetch the blktrace tools from:
369 git://git.kernel.dk/blktrace.git
371 Tracing also is possible using the ftrace interface, e.g.:
373 echo 1 > /sys/block/sda/sda1/trace/enable
374 echo blk > /sys/kernel/debug/tracing/current_tracer
375 cat /sys/kernel/debug/tracing/trace_pipe
381 depends on HAVE_REGS_AND_STACK_ACCESS_API
382 bool "Enable kprobes-based dynamic events"
387 This allows the user to add tracing events (similar to tracepoints)
388 on the fly via the ftrace interface. See
389 Documentation/trace/kprobetrace.txt for more details.
391 Those events can be inserted wherever kprobes can probe, and record
392 various register and memory values.
394 This option is also required by perf-probe subcommand of perf tools.
395 If you want to use perf tools, this option is strongly recommended.
398 bool "Enable uprobes-based dynamic events"
399 depends on ARCH_SUPPORTS_UPROBES
406 This allows the user to add tracing events on top of userspace
407 dynamic events (similar to tracepoints) on the fly via the trace
408 events interface. Those events can be inserted wherever uprobes
409 can probe, and record various registers.
410 This option is required if you plan to use perf-probe subcommand
411 of perf tools on user space applications.
416 config DYNAMIC_FTRACE
417 bool "enable/disable function tracing dynamically"
418 depends on FUNCTION_TRACER
419 depends on HAVE_DYNAMIC_FTRACE
422 This option will modify all the calls to function tracing
423 dynamically (will patch them out of the binary image and
424 replace them with a No-Op instruction) on boot up. During
425 compile time, a table is made of all the locations that ftrace
426 can function trace, and this table is linked into the kernel
427 image. When this is enabled, functions can be individually
428 enabled, and the functions not enabled will not affect
429 performance of the system.
431 See the files in /sys/kernel/debug/tracing:
432 available_filter_functions
436 This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but
437 otherwise has native performance as long as no tracing is active.
439 config DYNAMIC_FTRACE_WITH_REGS
441 depends on DYNAMIC_FTRACE
442 depends on HAVE_DYNAMIC_FTRACE_WITH_REGS
444 config FUNCTION_PROFILER
445 bool "Kernel function profiler"
446 depends on FUNCTION_TRACER
449 This option enables the kernel function profiler. A file is created
450 in debugfs called function_profile_enabled which defaults to zero.
451 When a 1 is echoed into this file profiling begins, and when a
452 zero is entered, profiling stops. A "functions" file is created in
453 the trace_stats directory; this file shows the list of functions that
454 have been hit and their counters.
458 config FTRACE_MCOUNT_RECORD
460 depends on DYNAMIC_FTRACE
461 depends on HAVE_FTRACE_MCOUNT_RECORD
463 config FTRACE_SELFTEST
466 config FTRACE_STARTUP_TEST
467 bool "Perform a startup test on ftrace"
468 depends on GENERIC_TRACER
469 select FTRACE_SELFTEST
471 This option performs a series of startup tests on ftrace. On bootup
472 a series of tests are made to verify that the tracer is
473 functioning properly. It will do tests on all the configured
476 config EVENT_TRACE_TEST_SYSCALLS
477 bool "Run selftest on syscall events"
478 depends on FTRACE_STARTUP_TEST
480 This option will also enable testing every syscall event.
481 It only enables the event and disables it and runs various loads
482 with the event enabled. This adds a bit more time for kernel boot
483 up since it runs this on every system call defined.
485 TBD - enable a way to actually call the syscalls as we test their
489 bool "Memory mapped IO tracing"
490 depends on HAVE_MMIOTRACE_SUPPORT && PCI
491 select GENERIC_TRACER
493 Mmiotrace traces Memory Mapped I/O access and is meant for
494 debugging and reverse engineering. It is called from the ioremap
495 implementation and works via page faults. Tracing is disabled by
496 default and can be enabled at run-time.
498 See Documentation/trace/mmiotrace.txt.
499 If you are not helping to develop drivers, say N.
501 config MMIOTRACE_TEST
502 tristate "Test module for mmiotrace"
503 depends on MMIOTRACE && m
505 This is a dumb module for testing mmiotrace. It is very dangerous
506 as it will write garbage to IO memory starting at a given address.
507 However, it should be safe to use on e.g. unused portion of VRAM.
509 Say N, unless you absolutely know what you are doing.
511 config RING_BUFFER_BENCHMARK
512 tristate "Ring buffer benchmark stress tester"
513 depends on RING_BUFFER
515 This option creates a test to stress the ring buffer and benchmark it.
516 It creates its own ring buffer such that it will not interfere with
517 any other users of the ring buffer (such as ftrace). It then creates
518 a producer and consumer that will run for 10 seconds and sleep for
519 10 seconds. Each interval it will print out the number of events
520 it recorded and give a rough estimate of how long each iteration took.
522 It does not disable interrupts or raise its priority, so it may be
523 affected by processes that are running.
529 endif # TRACING_SUPPORT