1 # SPDX-License-Identifier: GPL-2.0-only
3 # Architectures that offer an FUNCTION_TRACER implementation should
4 # select HAVE_FUNCTION_TRACER:
7 config USER_STACKTRACE_SUPPORT
13 config HAVE_FUNCTION_TRACER
16 See Documentation/trace/ftrace-design.rst
18 config HAVE_FUNCTION_GRAPH_TRACER
21 See Documentation/trace/ftrace-design.rst
23 config HAVE_DYNAMIC_FTRACE
26 See Documentation/trace/ftrace-design.rst
28 config HAVE_DYNAMIC_FTRACE_WITH_REGS
31 config HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
34 config HAVE_FTRACE_MCOUNT_RECORD
37 See Documentation/trace/ftrace-design.rst
39 config HAVE_SYSCALL_TRACEPOINTS
42 See Documentation/trace/ftrace-design.rst
47 Arch supports the gcc options -pg with -mfentry
49 config HAVE_NOP_MCOUNT
52 Arch supports the gcc options -pg with -mrecord-mcount and -nop-mcount
54 config HAVE_C_RECORDMCOUNT
57 C version of recordmcount available?
59 config TRACER_MAX_TRACE
71 select CONTEXT_SWITCH_TRACER
75 config CONTEXT_SWITCH_TRACER
78 config RING_BUFFER_ALLOW_SWAP
81 Allow the use of ring_buffer_swap_cpu.
82 Adds a very slight overhead to tracing when enabled.
84 config PREEMPTIRQ_TRACEPOINTS
86 depends on TRACE_PREEMPT_TOGGLE || TRACE_IRQFLAGS
90 Create preempt/irq toggle tracepoints if needed, so that other parts
91 of the kernel can use them to generate or add hooks to them.
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.
103 select STACKTRACE if STACKTRACE_SUPPORT
110 config GENERIC_TRACER
115 # Minimum requirements an architecture has to meet for us to
116 # be able to offer generic tracing facilities:
118 config TRACING_SUPPORT
120 depends on TRACE_IRQFLAGS_SUPPORT
121 depends on STACKTRACE_SUPPORT
128 default y if DEBUG_KERNEL
130 Enable the kernel tracing infrastructure.
134 config BOOTTIME_TRACING
135 bool "Boot-time Tracing support"
139 Enable developer to setup ftrace subsystem via supplemental
140 kernel cmdline at boot time for debugging (tracing) driver
141 initialization and boot process.
143 config FUNCTION_TRACER
144 bool "Kernel Function Tracer"
145 depends on HAVE_FUNCTION_TRACER
147 select GENERIC_TRACER
148 select CONTEXT_SWITCH_TRACER
150 select TASKS_RCU if PREEMPTION
151 select TASKS_RUDE_RCU
153 Enable the kernel to trace every kernel function. This is done
154 by using a compiler feature to insert a small, 5-byte No-Operation
155 instruction at the beginning of every kernel function, which NOP
156 sequence is then dynamically patched into a tracer call when
157 tracing is enabled by the administrator. If it's runtime disabled
158 (the bootup default), then the overhead of the instructions is very
159 small and not measurable even in micro-benchmarks.
161 config FUNCTION_GRAPH_TRACER
162 bool "Kernel Function Graph Tracer"
163 depends on HAVE_FUNCTION_GRAPH_TRACER
164 depends on FUNCTION_TRACER
165 depends on !X86_32 || !CC_OPTIMIZE_FOR_SIZE
168 Enable the kernel to trace a function at both its return
170 Its first purpose is to trace the duration of functions and
171 draw a call graph for each thread with some information like
172 the return value. This is done by setting the current return
173 address on the current task structure into a stack of calls.
175 config DYNAMIC_FTRACE
176 bool "enable/disable function tracing dynamically"
177 depends on FUNCTION_TRACER
178 depends on HAVE_DYNAMIC_FTRACE
181 This option will modify all the calls to function tracing
182 dynamically (will patch them out of the binary image and
183 replace them with a No-Op instruction) on boot up. During
184 compile time, a table is made of all the locations that ftrace
185 can function trace, and this table is linked into the kernel
186 image. When this is enabled, functions can be individually
187 enabled, and the functions not enabled will not affect
188 performance of the system.
190 See the files in /sys/kernel/debug/tracing:
191 available_filter_functions
195 This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but
196 otherwise has native performance as long as no tracing is active.
198 config DYNAMIC_FTRACE_WITH_REGS
200 depends on DYNAMIC_FTRACE
201 depends on HAVE_DYNAMIC_FTRACE_WITH_REGS
203 config DYNAMIC_FTRACE_WITH_DIRECT_CALLS
205 depends on DYNAMIC_FTRACE_WITH_REGS
206 depends on HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
208 config FUNCTION_PROFILER
209 bool "Kernel function profiler"
210 depends on FUNCTION_TRACER
213 This option enables the kernel function profiler. A file is created
214 in debugfs called function_profile_enabled which defaults to zero.
215 When a 1 is echoed into this file profiling begins, and when a
216 zero is entered, profiling stops. A "functions" file is created in
217 the trace_stat directory; this file shows the list of functions that
218 have been hit and their counters.
223 bool "Trace max stack"
224 depends on HAVE_FUNCTION_TRACER
225 select FUNCTION_TRACER
229 This special tracer records the maximum stack footprint of the
230 kernel and displays it in /sys/kernel/debug/tracing/stack_trace.
232 This tracer works by hooking into every function call that the
233 kernel executes, and keeping a maximum stack depth value and
234 stack-trace saved. If this is configured with DYNAMIC_FTRACE
235 then it will not have any overhead while the stack tracer
238 To enable the stack tracer on bootup, pass in 'stacktrace'
239 on the kernel command line.
241 The stack tracer can also be enabled or disabled via the
242 sysctl kernel.stack_tracer_enabled
246 config TRACE_PREEMPT_TOGGLE
249 Enables hooks which will be called when preemption is first disabled,
252 config IRQSOFF_TRACER
253 bool "Interrupts-off Latency Tracer"
255 depends on TRACE_IRQFLAGS_SUPPORT
256 select TRACE_IRQFLAGS
257 select GENERIC_TRACER
258 select TRACER_MAX_TRACE
259 select RING_BUFFER_ALLOW_SWAP
260 select TRACER_SNAPSHOT
261 select TRACER_SNAPSHOT_PER_CPU_SWAP
263 This option measures the time spent in irqs-off critical
264 sections, with microsecond accuracy.
266 The default measurement method is a maximum search, which is
267 disabled by default and can be runtime (re-)started
270 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency
272 (Note that kernel size and overhead increase with this option
273 enabled. This option and the preempt-off timing option can be
274 used together or separately.)
276 config PREEMPT_TRACER
277 bool "Preemption-off Latency Tracer"
279 depends on PREEMPTION
280 select GENERIC_TRACER
281 select TRACER_MAX_TRACE
282 select RING_BUFFER_ALLOW_SWAP
283 select TRACER_SNAPSHOT
284 select TRACER_SNAPSHOT_PER_CPU_SWAP
285 select TRACE_PREEMPT_TOGGLE
287 This option measures the time spent in preemption-off critical
288 sections, with microsecond accuracy.
290 The default measurement method is a maximum search, which is
291 disabled by default and can be runtime (re-)started
294 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency
296 (Note that kernel size and overhead increase with this option
297 enabled. This option and the irqs-off timing option can be
298 used together or separately.)
301 bool "Scheduling Latency Tracer"
302 select GENERIC_TRACER
303 select CONTEXT_SWITCH_TRACER
304 select TRACER_MAX_TRACE
305 select TRACER_SNAPSHOT
307 This tracer tracks the latency of the highest priority task
308 to be scheduled in, starting from the point it has woken up.
311 bool "Tracer to detect hardware latencies (like SMIs)"
312 select GENERIC_TRACER
314 This tracer, when enabled will create one or more kernel threads,
315 depending on what the cpumask file is set to, which each thread
316 spinning in a loop looking for interruptions caused by
317 something other than the kernel. For example, if a
318 System Management Interrupt (SMI) takes a noticeable amount of
319 time, this tracer will detect it. This is useful for testing
320 if a system is reliable for Real Time tasks.
322 Some files are created in the tracing directory when this
325 hwlat_detector/width - time in usecs for how long to spin for
326 hwlat_detector/window - time in usecs between the start of each
329 A kernel thread is created that will spin with interrupts disabled
330 for "width" microseconds in every "window" cycle. It will not spin
331 for "window - width" microseconds, where the system can
334 The output will appear in the trace and trace_pipe files.
336 When the tracer is not running, it has no affect on the system,
337 but when it is running, it can cause the system to be
338 periodically non responsive. Do not run this tracer on a
341 To enable this tracer, echo in "hwlat" into the current_tracer
342 file. Every time a latency is greater than tracing_thresh, it will
343 be recorded into the ring buffer.
346 bool "Memory mapped IO tracing"
347 depends on HAVE_MMIOTRACE_SUPPORT && PCI
348 select GENERIC_TRACER
350 Mmiotrace traces Memory Mapped I/O access and is meant for
351 debugging and reverse engineering. It is called from the ioremap
352 implementation and works via page faults. Tracing is disabled by
353 default and can be enabled at run-time.
355 See Documentation/trace/mmiotrace.rst.
356 If you are not helping to develop drivers, say N.
358 config ENABLE_DEFAULT_TRACERS
359 bool "Trace process context switches and events"
360 depends on !GENERIC_TRACER
363 This tracer hooks to various trace points in the kernel,
364 allowing the user to pick and choose which trace point they
365 want to trace. It also includes the sched_switch tracer plugin.
367 config FTRACE_SYSCALLS
368 bool "Trace syscalls"
369 depends on HAVE_SYSCALL_TRACEPOINTS
370 select GENERIC_TRACER
373 Basic tracer to catch the syscall entry and exit events.
375 config TRACER_SNAPSHOT
376 bool "Create a snapshot trace buffer"
377 select TRACER_MAX_TRACE
379 Allow tracing users to take snapshot of the current buffer using the
380 ftrace interface, e.g.:
382 echo 1 > /sys/kernel/debug/tracing/snapshot
385 config TRACER_SNAPSHOT_PER_CPU_SWAP
386 bool "Allow snapshot to swap per CPU"
387 depends on TRACER_SNAPSHOT
388 select RING_BUFFER_ALLOW_SWAP
390 Allow doing a snapshot of a single CPU buffer instead of a
391 full swap (all buffers). If this is set, then the following is
394 echo 1 > /sys/kernel/debug/tracing/per_cpu/cpu2/snapshot
396 After which, only the tracing buffer for CPU 2 was swapped with
397 the main tracing buffer, and the other CPU buffers remain the same.
399 When this is enabled, this adds a little more overhead to the
400 trace recording, as it needs to add some checks to synchronize
401 recording with swaps. But this does not affect the performance
402 of the overall system. This is enabled by default when the preempt
403 or irq latency tracers are enabled, as those need to swap as well
404 and already adds the overhead (plus a lot more).
406 config TRACE_BRANCH_PROFILING
408 select GENERIC_TRACER
411 prompt "Branch Profiling"
412 default BRANCH_PROFILE_NONE
414 The branch profiling is a software profiler. It will add hooks
415 into the C conditionals to test which path a branch takes.
417 The likely/unlikely profiler only looks at the conditions that
418 are annotated with a likely or unlikely macro.
420 The "all branch" profiler will profile every if-statement in the
421 kernel. This profiler will also enable the likely/unlikely
424 Either of the above profilers adds a bit of overhead to the system.
425 If unsure, choose "No branch profiling".
427 config BRANCH_PROFILE_NONE
428 bool "No branch profiling"
430 No branch profiling. Branch profiling adds a bit of overhead.
431 Only enable it if you want to analyse the branching behavior.
432 Otherwise keep it disabled.
434 config PROFILE_ANNOTATED_BRANCHES
435 bool "Trace likely/unlikely profiler"
436 select TRACE_BRANCH_PROFILING
438 This tracer profiles all likely and unlikely macros
439 in the kernel. It will display the results in:
441 /sys/kernel/debug/tracing/trace_stat/branch_annotated
443 Note: this will add a significant overhead; only turn this
444 on if you need to profile the system's use of these macros.
446 config PROFILE_ALL_BRANCHES
447 bool "Profile all if conditionals" if !FORTIFY_SOURCE
448 select TRACE_BRANCH_PROFILING
450 This tracer profiles all branch conditions. Every if ()
451 taken in the kernel is recorded whether it hit or miss.
452 The results will be displayed in:
454 /sys/kernel/debug/tracing/trace_stat/branch_all
456 This option also enables the likely/unlikely profiler.
458 This configuration, when enabled, will impose a great overhead
459 on the system. This should only be enabled when the system
460 is to be analyzed in much detail.
463 config TRACING_BRANCHES
466 Selected by tracers that will trace the likely and unlikely
467 conditions. This prevents the tracers themselves from being
468 profiled. Profiling the tracing infrastructure can only happen
469 when the likelys and unlikelys are not being traced.
472 bool "Trace likely/unlikely instances"
473 depends on TRACE_BRANCH_PROFILING
474 select TRACING_BRANCHES
476 This traces the events of likely and unlikely condition
477 calls in the kernel. The difference between this and the
478 "Trace likely/unlikely profiler" is that this is not a
479 histogram of the callers, but actually places the calling
480 events into a running trace buffer to see when and where the
481 events happened, as well as their results.
485 config BLK_DEV_IO_TRACE
486 bool "Support for tracing block IO actions"
492 select GENERIC_TRACER
495 Say Y here if you want to be able to trace the block layer actions
496 on a given queue. Tracing allows you to see any traffic happening
497 on a block device queue. For more information (and the userspace
498 support tools needed), fetch the blktrace tools from:
500 git://git.kernel.dk/blktrace.git
502 Tracing also is possible using the ftrace interface, e.g.:
504 echo 1 > /sys/block/sda/sda1/trace/enable
505 echo blk > /sys/kernel/debug/tracing/current_tracer
506 cat /sys/kernel/debug/tracing/trace_pipe
512 depends on HAVE_REGS_AND_STACK_ACCESS_API
513 bool "Enable kprobes-based dynamic events"
516 select DYNAMIC_EVENTS
519 This allows the user to add tracing events (similar to tracepoints)
520 on the fly via the ftrace interface. See
521 Documentation/trace/kprobetrace.rst for more details.
523 Those events can be inserted wherever kprobes can probe, and record
524 various register and memory values.
526 This option is also required by perf-probe subcommand of perf tools.
527 If you want to use perf tools, this option is strongly recommended.
529 config KPROBE_EVENTS_ON_NOTRACE
530 bool "Do NOT protect notrace function from kprobe events"
531 depends on KPROBE_EVENTS
532 depends on KPROBES_ON_FTRACE
535 This is only for the developers who want to debug ftrace itself
538 If kprobes can use ftrace instead of breakpoint, ftrace related
539 functions are protected from kprobe-events to prevent an infinit
540 recursion or any unexpected execution path which leads to a kernel
543 This option disables such protection and allows you to put kprobe
544 events on ftrace functions for debugging ftrace by itself.
545 Note that this might let you shoot yourself in the foot.
550 bool "Enable uprobes-based dynamic events"
551 depends on ARCH_SUPPORTS_UPROBES
553 depends on PERF_EVENTS
556 select DYNAMIC_EVENTS
560 This allows the user to add tracing events on top of userspace
561 dynamic events (similar to tracepoints) on the fly via the trace
562 events interface. Those events can be inserted wherever uprobes
563 can probe, and record various registers.
564 This option is required if you plan to use perf-probe subcommand
565 of perf tools on user space applications.
568 depends on BPF_SYSCALL
569 depends on (KPROBE_EVENTS || UPROBE_EVENTS) && PERF_EVENTS
573 This allows the user to attach BPF programs to kprobe, uprobe, and
576 config DYNAMIC_EVENTS
582 config BPF_KPROBE_OVERRIDE
583 bool "Enable BPF programs to override a kprobed function"
584 depends on BPF_EVENTS
585 depends on FUNCTION_ERROR_INJECTION
588 Allows BPF to override the execution of a probed function and
589 set a different return value. This is used for error injection.
591 config FTRACE_MCOUNT_RECORD
593 depends on DYNAMIC_FTRACE
594 depends on HAVE_FTRACE_MCOUNT_RECORD
598 depends on ARCH_HAVE_NMI_SAFE_CMPXCHG
600 tracing_map is a special-purpose lock-free map for tracing,
601 separated out as a stand-alone facility in order to allow it
602 to be shared between multiple tracers. It isn't meant to be
603 generally used outside of that context, and is normally
604 selected by tracers that use it.
607 bool "Synthetic trace events"
609 select DYNAMIC_EVENTS
612 Synthetic events are user-defined trace events that can be
613 used to combine data from other trace events or in fact any
614 data source. Synthetic events can be generated indirectly
615 via the trace() action of histogram triggers or directly
616 by way of an in-kernel API.
618 See Documentation/trace/events.rst or
619 Documentation/trace/histogram.rst for details and examples.
624 bool "Histogram triggers"
625 depends on ARCH_HAVE_NMI_SAFE_CMPXCHG
628 select DYNAMIC_EVENTS
632 Hist triggers allow one or more arbitrary trace event fields
633 to be aggregated into hash tables and dumped to stdout by
634 reading a debugfs/tracefs file. They're useful for
635 gathering quick and dirty (though precise) summaries of
636 event activity as an initial guide for further investigation
637 using more advanced tools.
639 Inter-event tracing of quantities such as latencies is also
640 supported using hist triggers under this option.
642 See Documentation/trace/histogram.rst.
645 config TRACE_EVENT_INJECT
646 bool "Trace event injection"
649 Allow user-space to inject a specific trace event into the ring
650 buffer. This is mainly used for testing purpose.
654 config TRACEPOINT_BENCHMARK
655 bool "Add tracepoint that benchmarks tracepoints"
657 This option creates the tracepoint "benchmark:benchmark_event".
658 When the tracepoint is enabled, it kicks off a kernel thread that
659 goes into an infinite loop (calling cond_sched() to let other tasks
660 run), and calls the tracepoint. Each iteration will record the time
661 it took to write to the tracepoint and the next iteration that
662 data will be passed to the tracepoint itself. That is, the tracepoint
663 will report the time it took to do the previous tracepoint.
664 The string written to the tracepoint is a static string of 128 bytes
665 to keep the time the same. The initial string is simply a write of
666 "START". The second string records the cold cache time of the first
667 write which is not added to the rest of the calculations.
669 As it is a tight loop, it benchmarks as hot cache. That's fine because
670 we care most about hot paths that are probably in cache already.
672 An example of the output:
675 first=3672 [COLD CACHED]
676 last=632 first=3672 max=632 min=632 avg=316 std=446 std^2=199712
677 last=278 first=3672 max=632 min=278 avg=303 std=316 std^2=100337
678 last=277 first=3672 max=632 min=277 avg=296 std=258 std^2=67064
679 last=273 first=3672 max=632 min=273 avg=292 std=224 std^2=50411
680 last=273 first=3672 max=632 min=273 avg=288 std=200 std^2=40389
681 last=281 first=3672 max=632 min=273 avg=287 std=183 std^2=33666
684 config RING_BUFFER_BENCHMARK
685 tristate "Ring buffer benchmark stress tester"
686 depends on RING_BUFFER
688 This option creates a test to stress the ring buffer and benchmark it.
689 It creates its own ring buffer such that it will not interfere with
690 any other users of the ring buffer (such as ftrace). It then creates
691 a producer and consumer that will run for 10 seconds and sleep for
692 10 seconds. Each interval it will print out the number of events
693 it recorded and give a rough estimate of how long each iteration took.
695 It does not disable interrupts or raise its priority, so it may be
696 affected by processes that are running.
700 config TRACE_EVAL_MAP_FILE
701 bool "Show eval mappings for trace events"
704 The "print fmt" of the trace events will show the enum/sizeof names
705 instead of their values. This can cause problems for user space tools
706 that use this string to parse the raw data as user space does not know
707 how to convert the string to its value.
709 To fix this, there's a special macro in the kernel that can be used
710 to convert an enum/sizeof into its value. If this macro is used, then
711 the print fmt strings will be converted to their values.
713 If something does not get converted properly, this option can be
714 used to show what enums/sizeof the kernel tried to convert.
716 This option is for debugging the conversions. A file is created
717 in the tracing directory called "eval_map" that will show the
718 names matched with their values and what trace event system they
721 Normally, the mapping of the strings to values will be freed after
722 boot up or module load. With this option, they will not be freed, as
723 they are needed for the "eval_map" file. Enabling this option will
724 increase the memory footprint of the running kernel.
728 config GCOV_PROFILE_FTRACE
729 bool "Enable GCOV profiling on ftrace subsystem"
730 depends on GCOV_KERNEL
732 Enable GCOV profiling on ftrace subsystem for checking
733 which functions/lines are tested.
737 Note that on a kernel compiled with this config, ftrace will
738 run significantly slower.
740 config FTRACE_SELFTEST
743 config FTRACE_STARTUP_TEST
744 bool "Perform a startup test on ftrace"
745 depends on GENERIC_TRACER
746 select FTRACE_SELFTEST
748 This option performs a series of startup tests on ftrace. On bootup
749 a series of tests are made to verify that the tracer is
750 functioning properly. It will do tests on all the configured
753 config EVENT_TRACE_STARTUP_TEST
754 bool "Run selftest on trace events"
755 depends on FTRACE_STARTUP_TEST
758 This option performs a test on all trace events in the system.
759 It basically just enables each event and runs some code that
760 will trigger events (not necessarily the event it enables)
761 This may take some time run as there are a lot of events.
763 config EVENT_TRACE_TEST_SYSCALLS
764 bool "Run selftest on syscall events"
765 depends on EVENT_TRACE_STARTUP_TEST
767 This option will also enable testing every syscall event.
768 It only enables the event and disables it and runs various loads
769 with the event enabled. This adds a bit more time for kernel boot
770 up since it runs this on every system call defined.
772 TBD - enable a way to actually call the syscalls as we test their
775 config RING_BUFFER_STARTUP_TEST
776 bool "Ring buffer startup self test"
777 depends on RING_BUFFER
779 Run a simple self test on the ring buffer on boot up. Late in the
780 kernel boot sequence, the test will start that kicks off
781 a thread per cpu. Each thread will write various size events
782 into the ring buffer. Another thread is created to send IPIs
783 to each of the threads, where the IPI handler will also write
784 to the ring buffer, to test/stress the nesting ability.
785 If any anomalies are discovered, a warning will be displayed
786 and all ring buffers will be disabled.
788 The test runs for 10 seconds. This will slow your boot time
789 by at least 10 more seconds.
791 At the end of the test, statics and more checks are done.
792 It will output the stats of each per cpu buffer. What
793 was written, the sizes, what was read, what was lost, and
794 other similar details.
798 config MMIOTRACE_TEST
799 tristate "Test module for mmiotrace"
800 depends on MMIOTRACE && m
802 This is a dumb module for testing mmiotrace. It is very dangerous
803 as it will write garbage to IO memory starting at a given address.
804 However, it should be safe to use on e.g. unused portion of VRAM.
806 Say N, unless you absolutely know what you are doing.
808 config PREEMPTIRQ_DELAY_TEST
809 tristate "Test module to create a preempt / IRQ disable delay thread to test latency tracers"
812 Select this option to build a test module that can help test latency
813 tracers by executing a preempt or irq disable section with a user
814 configurable delay. The module busy waits for the duration of the
817 For example, the following invocation generates a burst of three
818 irq-disabled critical sections for 500us:
819 modprobe preemptirq_delay_test test_mode=irq delay=500 burst_size=3
823 config SYNTH_EVENT_GEN_TEST
824 tristate "Test module for in-kernel synthetic event generation"
825 depends on SYNTH_EVENTS
827 This option creates a test module to check the base
828 functionality of in-kernel synthetic event definition and
831 To test, insert the module, and then check the trace buffer
832 for the generated sample events.
836 config KPROBE_EVENT_GEN_TEST
837 tristate "Test module for in-kernel kprobe event generation"
838 depends on KPROBE_EVENTS
840 This option creates a test module to check the base
841 functionality of in-kernel kprobe event definition.
843 To test, insert the module, and then check the trace buffer
844 for the generated kprobe events.
848 config HIST_TRIGGERS_DEBUG
849 bool "Hist trigger debug support"
850 depends on HIST_TRIGGERS
852 Add "hist_debug" file for each event, which when read will
853 dump out a bunch of internal details about the hist triggers
854 defined on that event.
856 The hist_debug file serves a couple of purposes:
858 - Helps developers verify that nothing is broken.
860 - Provides educational information to support the details
861 of the hist trigger internals as described by
862 Documentation/trace/histogram-design.rst.
864 The hist_debug output only covers the data structures
865 related to the histogram definitions themselves and doesn't
866 display the internals of map buckets or variable values of
873 endif # TRACING_SUPPORT