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
18 depends on HAVE_RETHOOK
20 Enable generic return hooking feature. This is an internal
21 API, which will be used by other function-entry hooking
22 features like fprobe and kprobes.
24 config HAVE_FUNCTION_TRACER
27 See Documentation/trace/ftrace-design.rst
29 config HAVE_FUNCTION_GRAPH_TRACER
32 See Documentation/trace/ftrace-design.rst
34 config HAVE_DYNAMIC_FTRACE
37 See Documentation/trace/ftrace-design.rst
39 config HAVE_DYNAMIC_FTRACE_WITH_REGS
42 config HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
45 config HAVE_DYNAMIC_FTRACE_WITH_ARGS
48 If this is set, then arguments and stack can be found from
49 the pt_regs passed into the function callback regs parameter
50 by default, even without setting the REGS flag in the ftrace_ops.
51 This allows for use of regs_get_kernel_argument() and
52 kernel_stack_pointer().
54 config HAVE_DYNAMIC_FTRACE_NO_PATCHABLE
57 If the architecture generates __patchable_function_entries sections
58 but does not want them included in the ftrace locations.
60 config HAVE_FTRACE_MCOUNT_RECORD
63 See Documentation/trace/ftrace-design.rst
65 config HAVE_SYSCALL_TRACEPOINTS
68 See Documentation/trace/ftrace-design.rst
73 Arch supports the gcc options -pg with -mfentry
75 config HAVE_NOP_MCOUNT
78 Arch supports the gcc options -pg with -mrecord-mcount and -nop-mcount
80 config HAVE_OBJTOOL_MCOUNT
83 Arch supports objtool --mcount
85 config HAVE_C_RECORDMCOUNT
88 C version of recordmcount available?
90 config HAVE_BUILDTIME_MCOUNT_SORT
93 An architecture selects this if it sorts the mcount_loc section
96 config BUILDTIME_MCOUNT_SORT
99 depends on HAVE_BUILDTIME_MCOUNT_SORT && DYNAMIC_FTRACE
101 Sort the mcount_loc section at build time.
103 config TRACER_MAX_TRACE
115 select CONTEXT_SWITCH_TRACER
119 config CONTEXT_SWITCH_TRACER
122 config RING_BUFFER_ALLOW_SWAP
125 Allow the use of ring_buffer_swap_cpu.
126 Adds a very slight overhead to tracing when enabled.
128 config PREEMPTIRQ_TRACEPOINTS
130 depends on TRACE_PREEMPT_TOGGLE || TRACE_IRQFLAGS
134 Create preempt/irq toggle tracepoints if needed, so that other parts
135 of the kernel can use them to generate or add hooks to them.
137 # All tracer options should select GENERIC_TRACER. For those options that are
138 # enabled by all tracers (context switch and event tracer) they select TRACING.
139 # This allows those options to appear when no other tracer is selected. But the
140 # options do not appear when something else selects it. We need the two options
141 # GENERIC_TRACER and TRACING to avoid circular dependencies to accomplish the
142 # hiding of the automatic options.
147 select STACKTRACE if STACKTRACE_SUPPORT
153 select TASKS_RCU if PREEMPTION
155 config GENERIC_TRACER
160 # Minimum requirements an architecture has to meet for us to
161 # be able to offer generic tracing facilities:
163 config TRACING_SUPPORT
165 depends on TRACE_IRQFLAGS_SUPPORT
166 depends on STACKTRACE_SUPPORT
171 depends on TRACING_SUPPORT
172 default y if DEBUG_KERNEL
174 Enable the kernel tracing infrastructure.
178 config BOOTTIME_TRACING
179 bool "Boot-time Tracing support"
183 Enable developer to setup ftrace subsystem via supplemental
184 kernel cmdline at boot time for debugging (tracing) driver
185 initialization and boot process.
187 config FUNCTION_TRACER
188 bool "Kernel Function Tracer"
189 depends on HAVE_FUNCTION_TRACER
191 select GENERIC_TRACER
192 select CONTEXT_SWITCH_TRACER
194 select TASKS_RCU if PREEMPTION
195 select TASKS_RUDE_RCU
197 Enable the kernel to trace every kernel function. This is done
198 by using a compiler feature to insert a small, 5-byte No-Operation
199 instruction at the beginning of every kernel function, which NOP
200 sequence is then dynamically patched into a tracer call when
201 tracing is enabled by the administrator. If it's runtime disabled
202 (the bootup default), then the overhead of the instructions is very
203 small and not measurable even in micro-benchmarks (at least on
204 x86, but may have impact on other architectures).
206 config FUNCTION_GRAPH_TRACER
207 bool "Kernel Function Graph Tracer"
208 depends on HAVE_FUNCTION_GRAPH_TRACER
209 depends on FUNCTION_TRACER
210 depends on !X86_32 || !CC_OPTIMIZE_FOR_SIZE
213 Enable the kernel to trace a function at both its return
215 Its first purpose is to trace the duration of functions and
216 draw a call graph for each thread with some information like
217 the return value. This is done by setting the current return
218 address on the current task structure into a stack of calls.
220 config DYNAMIC_FTRACE
221 bool "enable/disable function tracing dynamically"
222 depends on FUNCTION_TRACER
223 depends on HAVE_DYNAMIC_FTRACE
226 This option will modify all the calls to function tracing
227 dynamically (will patch them out of the binary image and
228 replace them with a No-Op instruction) on boot up. During
229 compile time, a table is made of all the locations that ftrace
230 can function trace, and this table is linked into the kernel
231 image. When this is enabled, functions can be individually
232 enabled, and the functions not enabled will not affect
233 performance of the system.
235 See the files in /sys/kernel/debug/tracing:
236 available_filter_functions
240 This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but
241 otherwise has native performance as long as no tracing is active.
243 config DYNAMIC_FTRACE_WITH_REGS
245 depends on DYNAMIC_FTRACE
246 depends on HAVE_DYNAMIC_FTRACE_WITH_REGS
248 config DYNAMIC_FTRACE_WITH_DIRECT_CALLS
250 depends on DYNAMIC_FTRACE_WITH_REGS
251 depends on HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
253 config DYNAMIC_FTRACE_WITH_ARGS
255 depends on DYNAMIC_FTRACE
256 depends on HAVE_DYNAMIC_FTRACE_WITH_ARGS
259 bool "Kernel Function Probe (fprobe)"
260 depends on FUNCTION_TRACER
261 depends on DYNAMIC_FTRACE_WITH_REGS
262 depends on HAVE_RETHOOK
266 This option enables kernel function probe (fprobe) based on ftrace.
267 The fprobe is similar to kprobes, but probes only for kernel function
268 entries and exits. This also can probe multiple functions by one
273 config FUNCTION_PROFILER
274 bool "Kernel function profiler"
275 depends on FUNCTION_TRACER
278 This option enables the kernel function profiler. A file is created
279 in debugfs called function_profile_enabled which defaults to zero.
280 When a 1 is echoed into this file profiling begins, and when a
281 zero is entered, profiling stops. A "functions" file is created in
282 the trace_stat directory; this file shows the list of functions that
283 have been hit and their counters.
288 bool "Trace max stack"
289 depends on HAVE_FUNCTION_TRACER
290 select FUNCTION_TRACER
294 This special tracer records the maximum stack footprint of the
295 kernel and displays it in /sys/kernel/debug/tracing/stack_trace.
297 This tracer works by hooking into every function call that the
298 kernel executes, and keeping a maximum stack depth value and
299 stack-trace saved. If this is configured with DYNAMIC_FTRACE
300 then it will not have any overhead while the stack tracer
303 To enable the stack tracer on bootup, pass in 'stacktrace'
304 on the kernel command line.
306 The stack tracer can also be enabled or disabled via the
307 sysctl kernel.stack_tracer_enabled
311 config TRACE_PREEMPT_TOGGLE
314 Enables hooks which will be called when preemption is first disabled,
317 config IRQSOFF_TRACER
318 bool "Interrupts-off Latency Tracer"
320 depends on TRACE_IRQFLAGS_SUPPORT
321 select TRACE_IRQFLAGS
322 select GENERIC_TRACER
323 select TRACER_MAX_TRACE
324 select RING_BUFFER_ALLOW_SWAP
325 select TRACER_SNAPSHOT
326 select TRACER_SNAPSHOT_PER_CPU_SWAP
328 This option measures the time spent in irqs-off critical
329 sections, with microsecond accuracy.
331 The default measurement method is a maximum search, which is
332 disabled by default and can be runtime (re-)started
335 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency
337 (Note that kernel size and overhead increase with this option
338 enabled. This option and the preempt-off timing option can be
339 used together or separately.)
341 config PREEMPT_TRACER
342 bool "Preemption-off Latency Tracer"
344 depends on PREEMPTION
345 select GENERIC_TRACER
346 select TRACER_MAX_TRACE
347 select RING_BUFFER_ALLOW_SWAP
348 select TRACER_SNAPSHOT
349 select TRACER_SNAPSHOT_PER_CPU_SWAP
350 select TRACE_PREEMPT_TOGGLE
352 This option measures the time spent in preemption-off critical
353 sections, with microsecond accuracy.
355 The default measurement method is a maximum search, which is
356 disabled by default and can be runtime (re-)started
359 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency
361 (Note that kernel size and overhead increase with this option
362 enabled. This option and the irqs-off timing option can be
363 used together or separately.)
366 bool "Scheduling Latency Tracer"
367 select GENERIC_TRACER
368 select CONTEXT_SWITCH_TRACER
369 select TRACER_MAX_TRACE
370 select TRACER_SNAPSHOT
372 This tracer tracks the latency of the highest priority task
373 to be scheduled in, starting from the point it has woken up.
376 bool "Tracer to detect hardware latencies (like SMIs)"
377 select GENERIC_TRACER
378 select TRACER_MAX_TRACE
380 This tracer, when enabled will create one or more kernel threads,
381 depending on what the cpumask file is set to, which each thread
382 spinning in a loop looking for interruptions caused by
383 something other than the kernel. For example, if a
384 System Management Interrupt (SMI) takes a noticeable amount of
385 time, this tracer will detect it. This is useful for testing
386 if a system is reliable for Real Time tasks.
388 Some files are created in the tracing directory when this
391 hwlat_detector/width - time in usecs for how long to spin for
392 hwlat_detector/window - time in usecs between the start of each
395 A kernel thread is created that will spin with interrupts disabled
396 for "width" microseconds in every "window" cycle. It will not spin
397 for "window - width" microseconds, where the system can
400 The output will appear in the trace and trace_pipe files.
402 When the tracer is not running, it has no affect on the system,
403 but when it is running, it can cause the system to be
404 periodically non responsive. Do not run this tracer on a
407 To enable this tracer, echo in "hwlat" into the current_tracer
408 file. Every time a latency is greater than tracing_thresh, it will
409 be recorded into the ring buffer.
411 config OSNOISE_TRACER
412 bool "OS Noise tracer"
413 select GENERIC_TRACER
414 select TRACER_MAX_TRACE
416 In the context of high-performance computing (HPC), the Operating
417 System Noise (osnoise) refers to the interference experienced by an
418 application due to activities inside the operating system. In the
419 context of Linux, NMIs, IRQs, SoftIRQs, and any other system thread
420 can cause noise to the system. Moreover, hardware-related jobs can
421 also cause noise, for example, via SMIs.
423 The osnoise tracer leverages the hwlat_detector by running a similar
424 loop with preemption, SoftIRQs and IRQs enabled, thus allowing all
425 the sources of osnoise during its execution. The osnoise tracer takes
426 note of the entry and exit point of any source of interferences,
427 increasing a per-cpu interference counter. It saves an interference
428 counter for each source of interference. The interference counter for
429 NMI, IRQs, SoftIRQs, and threads is increased anytime the tool
430 observes these interferences' entry events. When a noise happens
431 without any interference from the operating system level, the
432 hardware noise counter increases, pointing to a hardware-related
433 noise. In this way, osnoise can account for any source of
434 interference. At the end of the period, the osnoise tracer prints
435 the sum of all noise, the max single noise, the percentage of CPU
436 available for the thread, and the counters for the noise sources.
438 In addition to the tracer, a set of tracepoints were added to
439 facilitate the identification of the osnoise source.
441 The output will appear in the trace and trace_pipe files.
443 To enable this tracer, echo in "osnoise" into the current_tracer
446 config TIMERLAT_TRACER
447 bool "Timerlat tracer"
448 select OSNOISE_TRACER
449 select GENERIC_TRACER
451 The timerlat tracer aims to help the preemptive kernel developers
452 to find sources of wakeup latencies of real-time threads.
454 The tracer creates a per-cpu kernel thread with real-time priority.
455 The tracer thread sets a periodic timer to wakeup itself, and goes
456 to sleep waiting for the timer to fire. At the wakeup, the thread
457 then computes a wakeup latency value as the difference between
458 the current time and the absolute time that the timer was set
461 The tracer prints two lines at every activation. The first is the
462 timer latency observed at the hardirq context before the
463 activation of the thread. The second is the timer latency observed
464 by the thread, which is the same level that cyclictest reports. The
465 ACTIVATION ID field serves to relate the irq execution to its
466 respective thread execution.
468 The tracer is build on top of osnoise tracer, and the osnoise:
469 events can be used to trace the source of interference from NMI,
470 IRQs and other threads. It also enables the capture of the
471 stacktrace at the IRQ context, which helps to identify the code
472 path that can cause thread delay.
475 bool "Memory mapped IO tracing"
476 depends on HAVE_MMIOTRACE_SUPPORT && PCI
477 select GENERIC_TRACER
479 Mmiotrace traces Memory Mapped I/O access and is meant for
480 debugging and reverse engineering. It is called from the ioremap
481 implementation and works via page faults. Tracing is disabled by
482 default and can be enabled at run-time.
484 See Documentation/trace/mmiotrace.rst.
485 If you are not helping to develop drivers, say N.
487 config ENABLE_DEFAULT_TRACERS
488 bool "Trace process context switches and events"
489 depends on !GENERIC_TRACER
492 This tracer hooks to various trace points in the kernel,
493 allowing the user to pick and choose which trace point they
494 want to trace. It also includes the sched_switch tracer plugin.
496 config FTRACE_SYSCALLS
497 bool "Trace syscalls"
498 depends on HAVE_SYSCALL_TRACEPOINTS
499 select GENERIC_TRACER
502 Basic tracer to catch the syscall entry and exit events.
504 config TRACER_SNAPSHOT
505 bool "Create a snapshot trace buffer"
506 select TRACER_MAX_TRACE
508 Allow tracing users to take snapshot of the current buffer using the
509 ftrace interface, e.g.:
511 echo 1 > /sys/kernel/debug/tracing/snapshot
514 config TRACER_SNAPSHOT_PER_CPU_SWAP
515 bool "Allow snapshot to swap per CPU"
516 depends on TRACER_SNAPSHOT
517 select RING_BUFFER_ALLOW_SWAP
519 Allow doing a snapshot of a single CPU buffer instead of a
520 full swap (all buffers). If this is set, then the following is
523 echo 1 > /sys/kernel/debug/tracing/per_cpu/cpu2/snapshot
525 After which, only the tracing buffer for CPU 2 was swapped with
526 the main tracing buffer, and the other CPU buffers remain the same.
528 When this is enabled, this adds a little more overhead to the
529 trace recording, as it needs to add some checks to synchronize
530 recording with swaps. But this does not affect the performance
531 of the overall system. This is enabled by default when the preempt
532 or irq latency tracers are enabled, as those need to swap as well
533 and already adds the overhead (plus a lot more).
535 config TRACE_BRANCH_PROFILING
537 select GENERIC_TRACER
540 prompt "Branch Profiling"
541 default BRANCH_PROFILE_NONE
543 The branch profiling is a software profiler. It will add hooks
544 into the C conditionals to test which path a branch takes.
546 The likely/unlikely profiler only looks at the conditions that
547 are annotated with a likely or unlikely macro.
549 The "all branch" profiler will profile every if-statement in the
550 kernel. This profiler will also enable the likely/unlikely
553 Either of the above profilers adds a bit of overhead to the system.
554 If unsure, choose "No branch profiling".
556 config BRANCH_PROFILE_NONE
557 bool "No branch profiling"
559 No branch profiling. Branch profiling adds a bit of overhead.
560 Only enable it if you want to analyse the branching behavior.
561 Otherwise keep it disabled.
563 config PROFILE_ANNOTATED_BRANCHES
564 bool "Trace likely/unlikely profiler"
565 select TRACE_BRANCH_PROFILING
567 This tracer profiles all likely and unlikely macros
568 in the kernel. It will display the results in:
570 /sys/kernel/debug/tracing/trace_stat/branch_annotated
572 Note: this will add a significant overhead; only turn this
573 on if you need to profile the system's use of these macros.
575 config PROFILE_ALL_BRANCHES
576 bool "Profile all if conditionals" if !FORTIFY_SOURCE
577 select TRACE_BRANCH_PROFILING
579 This tracer profiles all branch conditions. Every if ()
580 taken in the kernel is recorded whether it hit or miss.
581 The results will be displayed in:
583 /sys/kernel/debug/tracing/trace_stat/branch_all
585 This option also enables the likely/unlikely profiler.
587 This configuration, when enabled, will impose a great overhead
588 on the system. This should only be enabled when the system
589 is to be analyzed in much detail.
592 config TRACING_BRANCHES
595 Selected by tracers that will trace the likely and unlikely
596 conditions. This prevents the tracers themselves from being
597 profiled. Profiling the tracing infrastructure can only happen
598 when the likelys and unlikelys are not being traced.
601 bool "Trace likely/unlikely instances"
602 depends on TRACE_BRANCH_PROFILING
603 select TRACING_BRANCHES
605 This traces the events of likely and unlikely condition
606 calls in the kernel. The difference between this and the
607 "Trace likely/unlikely profiler" is that this is not a
608 histogram of the callers, but actually places the calling
609 events into a running trace buffer to see when and where the
610 events happened, as well as their results.
614 config BLK_DEV_IO_TRACE
615 bool "Support for tracing block IO actions"
621 select GENERIC_TRACER
624 Say Y here if you want to be able to trace the block layer actions
625 on a given queue. Tracing allows you to see any traffic happening
626 on a block device queue. For more information (and the userspace
627 support tools needed), fetch the blktrace tools from:
629 git://git.kernel.dk/blktrace.git
631 Tracing also is possible using the ftrace interface, e.g.:
633 echo 1 > /sys/block/sda/sda1/trace/enable
634 echo blk > /sys/kernel/debug/tracing/current_tracer
635 cat /sys/kernel/debug/tracing/trace_pipe
641 depends on HAVE_REGS_AND_STACK_ACCESS_API
642 bool "Enable kprobes-based dynamic events"
645 select DYNAMIC_EVENTS
648 This allows the user to add tracing events (similar to tracepoints)
649 on the fly via the ftrace interface. See
650 Documentation/trace/kprobetrace.rst for more details.
652 Those events can be inserted wherever kprobes can probe, and record
653 various register and memory values.
655 This option is also required by perf-probe subcommand of perf tools.
656 If you want to use perf tools, this option is strongly recommended.
658 config KPROBE_EVENTS_ON_NOTRACE
659 bool "Do NOT protect notrace function from kprobe events"
660 depends on KPROBE_EVENTS
661 depends on DYNAMIC_FTRACE
664 This is only for the developers who want to debug ftrace itself
667 If kprobes can use ftrace instead of breakpoint, ftrace related
668 functions are protected from kprobe-events to prevent an infinite
669 recursion or any unexpected execution path which leads to a kernel
672 This option disables such protection and allows you to put kprobe
673 events on ftrace functions for debugging ftrace by itself.
674 Note that this might let you shoot yourself in the foot.
679 bool "Enable uprobes-based dynamic events"
680 depends on ARCH_SUPPORTS_UPROBES
682 depends on PERF_EVENTS
685 select DYNAMIC_EVENTS
689 This allows the user to add tracing events on top of userspace
690 dynamic events (similar to tracepoints) on the fly via the trace
691 events interface. Those events can be inserted wherever uprobes
692 can probe, and record various registers.
693 This option is required if you plan to use perf-probe subcommand
694 of perf tools on user space applications.
697 depends on BPF_SYSCALL
698 depends on (KPROBE_EVENTS || UPROBE_EVENTS) && PERF_EVENTS
702 This allows the user to attach BPF programs to kprobe, uprobe, and
705 config DYNAMIC_EVENTS
711 config BPF_KPROBE_OVERRIDE
712 bool "Enable BPF programs to override a kprobed function"
713 depends on BPF_EVENTS
714 depends on FUNCTION_ERROR_INJECTION
717 Allows BPF to override the execution of a probed function and
718 set a different return value. This is used for error injection.
720 config FTRACE_MCOUNT_RECORD
722 depends on DYNAMIC_FTRACE
723 depends on HAVE_FTRACE_MCOUNT_RECORD
725 config FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
727 depends on FTRACE_MCOUNT_RECORD
729 config FTRACE_MCOUNT_USE_CC
731 depends on $(cc-option,-mrecord-mcount)
732 depends on !FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
733 depends on FTRACE_MCOUNT_RECORD
735 config FTRACE_MCOUNT_USE_OBJTOOL
737 depends on HAVE_OBJTOOL_MCOUNT
738 depends on !FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
739 depends on !FTRACE_MCOUNT_USE_CC
740 depends on FTRACE_MCOUNT_RECORD
743 config FTRACE_MCOUNT_USE_RECORDMCOUNT
745 depends on !FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
746 depends on !FTRACE_MCOUNT_USE_CC
747 depends on !FTRACE_MCOUNT_USE_OBJTOOL
748 depends on FTRACE_MCOUNT_RECORD
752 depends on ARCH_HAVE_NMI_SAFE_CMPXCHG
754 tracing_map is a special-purpose lock-free map for tracing,
755 separated out as a stand-alone facility in order to allow it
756 to be shared between multiple tracers. It isn't meant to be
757 generally used outside of that context, and is normally
758 selected by tracers that use it.
761 bool "Synthetic trace events"
763 select DYNAMIC_EVENTS
766 Synthetic events are user-defined trace events that can be
767 used to combine data from other trace events or in fact any
768 data source. Synthetic events can be generated indirectly
769 via the trace() action of histogram triggers or directly
770 by way of an in-kernel API.
772 See Documentation/trace/events.rst or
773 Documentation/trace/histogram.rst for details and examples.
778 bool "User trace events"
780 select DYNAMIC_EVENTS
781 depends on BROKEN || COMPILE_TEST # API needs to be straighten out
783 User trace events are user-defined trace events that
784 can be used like an existing kernel trace event. User trace
785 events are generated by writing to a tracefs file. User
786 processes can determine if their tracing events should be
787 generated by memory mapping a tracefs file and checking for
788 an associated byte being non-zero.
793 bool "Histogram triggers"
794 depends on ARCH_HAVE_NMI_SAFE_CMPXCHG
797 select DYNAMIC_EVENTS
801 Hist triggers allow one or more arbitrary trace event fields
802 to be aggregated into hash tables and dumped to stdout by
803 reading a debugfs/tracefs file. They're useful for
804 gathering quick and dirty (though precise) summaries of
805 event activity as an initial guide for further investigation
806 using more advanced tools.
808 Inter-event tracing of quantities such as latencies is also
809 supported using hist triggers under this option.
811 See Documentation/trace/histogram.rst.
814 config TRACE_EVENT_INJECT
815 bool "Trace event injection"
818 Allow user-space to inject a specific trace event into the ring
819 buffer. This is mainly used for testing purpose.
823 config TRACEPOINT_BENCHMARK
824 bool "Add tracepoint that benchmarks tracepoints"
826 This option creates the tracepoint "benchmark:benchmark_event".
827 When the tracepoint is enabled, it kicks off a kernel thread that
828 goes into an infinite loop (calling cond_resched() to let other tasks
829 run), and calls the tracepoint. Each iteration will record the time
830 it took to write to the tracepoint and the next iteration that
831 data will be passed to the tracepoint itself. That is, the tracepoint
832 will report the time it took to do the previous tracepoint.
833 The string written to the tracepoint is a static string of 128 bytes
834 to keep the time the same. The initial string is simply a write of
835 "START". The second string records the cold cache time of the first
836 write which is not added to the rest of the calculations.
838 As it is a tight loop, it benchmarks as hot cache. That's fine because
839 we care most about hot paths that are probably in cache already.
841 An example of the output:
844 first=3672 [COLD CACHED]
845 last=632 first=3672 max=632 min=632 avg=316 std=446 std^2=199712
846 last=278 first=3672 max=632 min=278 avg=303 std=316 std^2=100337
847 last=277 first=3672 max=632 min=277 avg=296 std=258 std^2=67064
848 last=273 first=3672 max=632 min=273 avg=292 std=224 std^2=50411
849 last=273 first=3672 max=632 min=273 avg=288 std=200 std^2=40389
850 last=281 first=3672 max=632 min=273 avg=287 std=183 std^2=33666
853 config RING_BUFFER_BENCHMARK
854 tristate "Ring buffer benchmark stress tester"
855 depends on RING_BUFFER
857 This option creates a test to stress the ring buffer and benchmark it.
858 It creates its own ring buffer such that it will not interfere with
859 any other users of the ring buffer (such as ftrace). It then creates
860 a producer and consumer that will run for 10 seconds and sleep for
861 10 seconds. Each interval it will print out the number of events
862 it recorded and give a rough estimate of how long each iteration took.
864 It does not disable interrupts or raise its priority, so it may be
865 affected by processes that are running.
869 config TRACE_EVAL_MAP_FILE
870 bool "Show eval mappings for trace events"
873 The "print fmt" of the trace events will show the enum/sizeof names
874 instead of their values. This can cause problems for user space tools
875 that use this string to parse the raw data as user space does not know
876 how to convert the string to its value.
878 To fix this, there's a special macro in the kernel that can be used
879 to convert an enum/sizeof into its value. If this macro is used, then
880 the print fmt strings will be converted to their values.
882 If something does not get converted properly, this option can be
883 used to show what enums/sizeof the kernel tried to convert.
885 This option is for debugging the conversions. A file is created
886 in the tracing directory called "eval_map" that will show the
887 names matched with their values and what trace event system they
890 Normally, the mapping of the strings to values will be freed after
891 boot up or module load. With this option, they will not be freed, as
892 they are needed for the "eval_map" file. Enabling this option will
893 increase the memory footprint of the running kernel.
897 config FTRACE_RECORD_RECURSION
898 bool "Record functions that recurse in function tracing"
899 depends on FUNCTION_TRACER
901 All callbacks that attach to the function tracing have some sort
902 of protection against recursion. Even though the protection exists,
903 it adds overhead. This option will create a file in the tracefs
904 file system called "recursed_functions" that will list the functions
905 that triggered a recursion.
907 This will add more overhead to cases that have recursion.
911 config FTRACE_RECORD_RECURSION_SIZE
912 int "Max number of recursed functions to record"
914 depends on FTRACE_RECORD_RECURSION
916 This defines the limit of number of functions that can be
917 listed in the "recursed_functions" file, that lists all
918 the functions that caused a recursion to happen.
919 This file can be reset, but the limit can not change in
922 config RING_BUFFER_RECORD_RECURSION
923 bool "Record functions that recurse in the ring buffer"
924 depends on FTRACE_RECORD_RECURSION
925 # default y, because it is coupled with FTRACE_RECORD_RECURSION
928 The ring buffer has its own internal recursion. Although when
929 recursion happens it wont cause harm because of the protection,
930 but it does cause an unwanted overhead. Enabling this option will
931 place where recursion was detected into the ftrace "recursed_functions"
934 This will add more overhead to cases that have recursion.
936 config GCOV_PROFILE_FTRACE
937 bool "Enable GCOV profiling on ftrace subsystem"
938 depends on GCOV_KERNEL
940 Enable GCOV profiling on ftrace subsystem for checking
941 which functions/lines are tested.
945 Note that on a kernel compiled with this config, ftrace will
946 run significantly slower.
948 config FTRACE_SELFTEST
951 config FTRACE_STARTUP_TEST
952 bool "Perform a startup test on ftrace"
953 depends on GENERIC_TRACER
954 select FTRACE_SELFTEST
956 This option performs a series of startup tests on ftrace. On bootup
957 a series of tests are made to verify that the tracer is
958 functioning properly. It will do tests on all the configured
961 config EVENT_TRACE_STARTUP_TEST
962 bool "Run selftest on trace events"
963 depends on FTRACE_STARTUP_TEST
966 This option performs a test on all trace events in the system.
967 It basically just enables each event and runs some code that
968 will trigger events (not necessarily the event it enables)
969 This may take some time run as there are a lot of events.
971 config EVENT_TRACE_TEST_SYSCALLS
972 bool "Run selftest on syscall events"
973 depends on EVENT_TRACE_STARTUP_TEST
975 This option will also enable testing every syscall event.
976 It only enables the event and disables it and runs various loads
977 with the event enabled. This adds a bit more time for kernel boot
978 up since it runs this on every system call defined.
980 TBD - enable a way to actually call the syscalls as we test their
983 config FTRACE_SORT_STARTUP_TEST
984 bool "Verify compile time sorting of ftrace functions"
985 depends on DYNAMIC_FTRACE
986 depends on BUILDTIME_MCOUNT_SORT
988 Sorting of the mcount_loc sections that is used to find the
989 where the ftrace knows where to patch functions for tracing
990 and other callbacks is done at compile time. But if the sort
991 is not done correctly, it will cause non-deterministic failures.
992 When this is set, the sorted sections will be verified that they
993 are in deed sorted and will warn if they are not.
997 config RING_BUFFER_STARTUP_TEST
998 bool "Ring buffer startup self test"
999 depends on RING_BUFFER
1001 Run a simple self test on the ring buffer on boot up. Late in the
1002 kernel boot sequence, the test will start that kicks off
1003 a thread per cpu. Each thread will write various size events
1004 into the ring buffer. Another thread is created to send IPIs
1005 to each of the threads, where the IPI handler will also write
1006 to the ring buffer, to test/stress the nesting ability.
1007 If any anomalies are discovered, a warning will be displayed
1008 and all ring buffers will be disabled.
1010 The test runs for 10 seconds. This will slow your boot time
1011 by at least 10 more seconds.
1013 At the end of the test, statics and more checks are done.
1014 It will output the stats of each per cpu buffer. What
1015 was written, the sizes, what was read, what was lost, and
1016 other similar details.
1020 config RING_BUFFER_VALIDATE_TIME_DELTAS
1021 bool "Verify ring buffer time stamp deltas"
1022 depends on RING_BUFFER
1024 This will audit the time stamps on the ring buffer sub
1025 buffer to make sure that all the time deltas for the
1026 events on a sub buffer matches the current time stamp.
1027 This audit is performed for every event that is not
1028 interrupted, or interrupting another event. A check
1029 is also made when traversing sub buffers to make sure
1030 that all the deltas on the previous sub buffer do not
1031 add up to be greater than the current time stamp.
1033 NOTE: This adds significant overhead to recording of events,
1034 and should only be used to test the logic of the ring buffer.
1035 Do not use it on production systems.
1037 Only say Y if you understand what this does, and you
1038 still want it enabled. Otherwise say N
1040 config MMIOTRACE_TEST
1041 tristate "Test module for mmiotrace"
1042 depends on MMIOTRACE && m
1044 This is a dumb module for testing mmiotrace. It is very dangerous
1045 as it will write garbage to IO memory starting at a given address.
1046 However, it should be safe to use on e.g. unused portion of VRAM.
1048 Say N, unless you absolutely know what you are doing.
1050 config PREEMPTIRQ_DELAY_TEST
1051 tristate "Test module to create a preempt / IRQ disable delay thread to test latency tracers"
1054 Select this option to build a test module that can help test latency
1055 tracers by executing a preempt or irq disable section with a user
1056 configurable delay. The module busy waits for the duration of the
1059 For example, the following invocation generates a burst of three
1060 irq-disabled critical sections for 500us:
1061 modprobe preemptirq_delay_test test_mode=irq delay=500 burst_size=3
1063 What's more, if you want to attach the test on the cpu which the latency
1064 tracer is running on, specify cpu_affinity=cpu_num at the end of the
1069 config SYNTH_EVENT_GEN_TEST
1070 tristate "Test module for in-kernel synthetic event generation"
1071 depends on SYNTH_EVENTS
1073 This option creates a test module to check the base
1074 functionality of in-kernel synthetic event definition and
1077 To test, insert the module, and then check the trace buffer
1078 for the generated sample events.
1082 config KPROBE_EVENT_GEN_TEST
1083 tristate "Test module for in-kernel kprobe event generation"
1084 depends on KPROBE_EVENTS
1086 This option creates a test module to check the base
1087 functionality of in-kernel kprobe event definition.
1089 To test, insert the module, and then check the trace buffer
1090 for the generated kprobe events.
1094 config HIST_TRIGGERS_DEBUG
1095 bool "Hist trigger debug support"
1096 depends on HIST_TRIGGERS
1098 Add "hist_debug" file for each event, which when read will
1099 dump out a bunch of internal details about the hist triggers
1100 defined on that event.
1102 The hist_debug file serves a couple of purposes:
1104 - Helps developers verify that nothing is broken.
1106 - Provides educational information to support the details
1107 of the hist trigger internals as described by
1108 Documentation/trace/histogram-design.rst.
1110 The hist_debug output only covers the data structures
1111 related to the histogram definitions themselves and doesn't
1112 display the internals of map buckets or variable values of
1117 source "kernel/trace/rv/Kconfig"