1 # SPDX-License-Identifier: GPL-2.0-only 1 # SPDX-License-Identifier: GPL-2.0-only
2 # 2 #
3 # Architectures that offer an FUNCTION_TRACER 3 # Architectures that offer an FUNCTION_TRACER implementation should
4 # select HAVE_FUNCTION_TRACER: 4 # select HAVE_FUNCTION_TRACER:
5 # 5 #
6 6
7 config USER_STACKTRACE_SUPPORT 7 config USER_STACKTRACE_SUPPORT
8 bool 8 bool
9 9
10 config NOP_TRACER 10 config NOP_TRACER
11 bool 11 bool
12 12
13 config HAVE_RETHOOK <<
14 bool <<
15 <<
16 config RETHOOK <<
17 bool <<
18 depends on HAVE_RETHOOK <<
19 help <<
20 Enable generic return hooking featur <<
21 API, which will be used by other fun <<
22 features like fprobe and kprobes. <<
23 <<
24 config HAVE_FUNCTION_TRACER 13 config HAVE_FUNCTION_TRACER
25 bool 14 bool
26 help 15 help
27 See Documentation/trace/ftrace-desig 16 See Documentation/trace/ftrace-design.rst
28 17
29 config HAVE_FUNCTION_GRAPH_TRACER 18 config HAVE_FUNCTION_GRAPH_TRACER
30 bool 19 bool
31 help 20 help
32 See Documentation/trace/ftrace-desig 21 See Documentation/trace/ftrace-design.rst
33 22
34 config HAVE_FUNCTION_GRAPH_RETVAL <<
35 bool <<
36 <<
37 config HAVE_DYNAMIC_FTRACE 23 config HAVE_DYNAMIC_FTRACE
38 bool 24 bool
39 help 25 help
40 See Documentation/trace/ftrace-desig 26 See Documentation/trace/ftrace-design.rst
41 27
42 config HAVE_DYNAMIC_FTRACE_WITH_REGS 28 config HAVE_DYNAMIC_FTRACE_WITH_REGS
43 bool 29 bool
44 30
45 config HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS 31 config HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
46 bool 32 bool
47 33
48 config HAVE_DYNAMIC_FTRACE_WITH_CALL_OPS <<
49 bool <<
50 <<
51 config HAVE_DYNAMIC_FTRACE_WITH_ARGS 34 config HAVE_DYNAMIC_FTRACE_WITH_ARGS
52 bool 35 bool
53 help 36 help
54 If this is set, then arguments and st 37 If this is set, then arguments and stack can be found from
55 the ftrace_regs passed into the funct !! 38 the pt_regs passed into the function callback regs parameter
56 by default, even without setting the 39 by default, even without setting the REGS flag in the ftrace_ops.
57 This allows for use of ftrace_regs_ge !! 40 This allows for use of regs_get_kernel_argument() and
58 ftrace_regs_get_stack_pointer(). !! 41 kernel_stack_pointer().
59 <<
60 config HAVE_DYNAMIC_FTRACE_NO_PATCHABLE <<
61 bool <<
62 help <<
63 If the architecture generates __patc <<
64 but does not want them included in t <<
65 42
66 config HAVE_FTRACE_MCOUNT_RECORD 43 config HAVE_FTRACE_MCOUNT_RECORD
67 bool 44 bool
68 help 45 help
69 See Documentation/trace/ftrace-desig 46 See Documentation/trace/ftrace-design.rst
70 47
71 config HAVE_SYSCALL_TRACEPOINTS 48 config HAVE_SYSCALL_TRACEPOINTS
72 bool 49 bool
73 help 50 help
74 See Documentation/trace/ftrace-desig 51 See Documentation/trace/ftrace-design.rst
75 52
76 config HAVE_FENTRY 53 config HAVE_FENTRY
77 bool 54 bool
78 help 55 help
79 Arch supports the gcc options -pg wi 56 Arch supports the gcc options -pg with -mfentry
80 57
81 config HAVE_NOP_MCOUNT 58 config HAVE_NOP_MCOUNT
82 bool 59 bool
83 help 60 help
84 Arch supports the gcc options -pg wi 61 Arch supports the gcc options -pg with -mrecord-mcount and -nop-mcount
85 62
86 config HAVE_OBJTOOL_MCOUNT 63 config HAVE_OBJTOOL_MCOUNT
87 bool 64 bool
88 help 65 help
89 Arch supports objtool --mcount 66 Arch supports objtool --mcount
90 67
91 config HAVE_OBJTOOL_NOP_MCOUNT <<
92 bool <<
93 help <<
94 Arch supports the objtool options -- <<
95 An architecture can select this if i <<
96 of ftrace locations. <<
97 <<
98 config HAVE_C_RECORDMCOUNT 68 config HAVE_C_RECORDMCOUNT
99 bool 69 bool
100 help 70 help
101 C version of recordmcount available? 71 C version of recordmcount available?
102 72
103 config HAVE_BUILDTIME_MCOUNT_SORT <<
104 bool <<
105 help <<
106 An architecture selects this if it so <<
107 at build time. <<
108 <<
109 config BUILDTIME_MCOUNT_SORT <<
110 bool <<
111 default y <<
112 depends on HAVE_BUILDTIME_MCOUNT_SORT & <<
113 help <<
114 Sort the mcount_loc section at build <<
115 <<
116 config TRACER_MAX_TRACE 73 config TRACER_MAX_TRACE
117 bool 74 bool
118 75
119 config TRACE_CLOCK 76 config TRACE_CLOCK
120 bool 77 bool
121 78
122 config RING_BUFFER 79 config RING_BUFFER
123 bool 80 bool
124 select TRACE_CLOCK 81 select TRACE_CLOCK
125 select IRQ_WORK 82 select IRQ_WORK
126 83
127 config EVENT_TRACING 84 config EVENT_TRACING
128 select CONTEXT_SWITCH_TRACER 85 select CONTEXT_SWITCH_TRACER
129 select GLOB 86 select GLOB
130 bool 87 bool
131 88
132 config CONTEXT_SWITCH_TRACER 89 config CONTEXT_SWITCH_TRACER
133 bool 90 bool
134 91
135 config RING_BUFFER_ALLOW_SWAP 92 config RING_BUFFER_ALLOW_SWAP
136 bool 93 bool
137 help 94 help
138 Allow the use of ring_buffer_swap_cpu 95 Allow the use of ring_buffer_swap_cpu.
139 Adds a very slight overhead to tracin 96 Adds a very slight overhead to tracing when enabled.
140 97
141 config PREEMPTIRQ_TRACEPOINTS 98 config PREEMPTIRQ_TRACEPOINTS
142 bool 99 bool
143 depends on TRACE_PREEMPT_TOGGLE || TRA 100 depends on TRACE_PREEMPT_TOGGLE || TRACE_IRQFLAGS
144 select TRACING 101 select TRACING
145 default y 102 default y
146 help 103 help
147 Create preempt/irq toggle tracepoint 104 Create preempt/irq toggle tracepoints if needed, so that other parts
148 of the kernel can use them to genera 105 of the kernel can use them to generate or add hooks to them.
149 106
150 # All tracer options should select GENERIC_TRA 107 # All tracer options should select GENERIC_TRACER. For those options that are
151 # enabled by all tracers (context switch and e 108 # enabled by all tracers (context switch and event tracer) they select TRACING.
152 # This allows those options to appear when no 109 # This allows those options to appear when no other tracer is selected. But the
153 # options do not appear when something else se 110 # options do not appear when something else selects it. We need the two options
154 # GENERIC_TRACER and TRACING to avoid circular 111 # GENERIC_TRACER and TRACING to avoid circular dependencies to accomplish the
155 # hiding of the automatic options. 112 # hiding of the automatic options.
156 113
157 config TRACING 114 config TRACING
158 bool 115 bool
159 select RING_BUFFER 116 select RING_BUFFER
160 select STACKTRACE if STACKTRACE_SUPPOR 117 select STACKTRACE if STACKTRACE_SUPPORT
161 select TRACEPOINTS 118 select TRACEPOINTS
162 select NOP_TRACER 119 select NOP_TRACER
163 select BINARY_PRINTF 120 select BINARY_PRINTF
164 select EVENT_TRACING 121 select EVENT_TRACING
165 select TRACE_CLOCK 122 select TRACE_CLOCK
166 select NEED_TASKS_RCU <<
167 123
168 config GENERIC_TRACER 124 config GENERIC_TRACER
169 bool 125 bool
170 select TRACING 126 select TRACING
171 127
172 # 128 #
173 # Minimum requirements an architecture has to 129 # Minimum requirements an architecture has to meet for us to
174 # be able to offer generic tracing facilities: 130 # be able to offer generic tracing facilities:
175 # 131 #
176 config TRACING_SUPPORT 132 config TRACING_SUPPORT
177 bool 133 bool
178 depends on TRACE_IRQFLAGS_SUPPORT 134 depends on TRACE_IRQFLAGS_SUPPORT
179 depends on STACKTRACE_SUPPORT 135 depends on STACKTRACE_SUPPORT
180 default y 136 default y
181 137
182 menuconfig FTRACE 138 menuconfig FTRACE
183 bool "Tracers" 139 bool "Tracers"
184 depends on TRACING_SUPPORT 140 depends on TRACING_SUPPORT
185 default y if DEBUG_KERNEL 141 default y if DEBUG_KERNEL
186 help 142 help
187 Enable the kernel tracing infrastruc 143 Enable the kernel tracing infrastructure.
188 144
189 if FTRACE 145 if FTRACE
190 146
191 config BOOTTIME_TRACING 147 config BOOTTIME_TRACING
192 bool "Boot-time Tracing support" 148 bool "Boot-time Tracing support"
193 depends on TRACING 149 depends on TRACING
194 select BOOT_CONFIG 150 select BOOT_CONFIG
195 help 151 help
196 Enable developer to setup ftrace sub 152 Enable developer to setup ftrace subsystem via supplemental
197 kernel cmdline at boot time for debu 153 kernel cmdline at boot time for debugging (tracing) driver
198 initialization and boot process. 154 initialization and boot process.
199 155
200 config FUNCTION_TRACER 156 config FUNCTION_TRACER
201 bool "Kernel Function Tracer" 157 bool "Kernel Function Tracer"
202 depends on HAVE_FUNCTION_TRACER 158 depends on HAVE_FUNCTION_TRACER
203 select KALLSYMS 159 select KALLSYMS
204 select GENERIC_TRACER 160 select GENERIC_TRACER
205 select CONTEXT_SWITCH_TRACER 161 select CONTEXT_SWITCH_TRACER
206 select GLOB 162 select GLOB
207 select NEED_TASKS_RCU !! 163 select TASKS_RCU if PREEMPTION
208 select TASKS_RUDE_RCU 164 select TASKS_RUDE_RCU
209 help 165 help
210 Enable the kernel to trace every ker 166 Enable the kernel to trace every kernel function. This is done
211 by using a compiler feature to inser 167 by using a compiler feature to insert a small, 5-byte No-Operation
212 instruction at the beginning of ever 168 instruction at the beginning of every kernel function, which NOP
213 sequence is then dynamically patched 169 sequence is then dynamically patched into a tracer call when
214 tracing is enabled by the administra 170 tracing is enabled by the administrator. If it's runtime disabled
215 (the bootup default), then the overh 171 (the bootup default), then the overhead of the instructions is very
216 small and not measurable even in mic !! 172 small and not measurable even in micro-benchmarks.
217 x86, but may have impact on other ar <<
218 173
219 config FUNCTION_GRAPH_TRACER 174 config FUNCTION_GRAPH_TRACER
220 bool "Kernel Function Graph Tracer" 175 bool "Kernel Function Graph Tracer"
221 depends on HAVE_FUNCTION_GRAPH_TRACER 176 depends on HAVE_FUNCTION_GRAPH_TRACER
222 depends on FUNCTION_TRACER 177 depends on FUNCTION_TRACER
223 depends on !X86_32 || !CC_OPTIMIZE_FOR 178 depends on !X86_32 || !CC_OPTIMIZE_FOR_SIZE
224 default y 179 default y
225 help 180 help
226 Enable the kernel to trace a functio 181 Enable the kernel to trace a function at both its return
227 and its entry. 182 and its entry.
228 Its first purpose is to trace the du 183 Its first purpose is to trace the duration of functions and
229 draw a call graph for each thread wi 184 draw a call graph for each thread with some information like
230 the return value. This is done by se 185 the return value. This is done by setting the current return
231 address on the current task structur 186 address on the current task structure into a stack of calls.
232 187
233 config FUNCTION_GRAPH_RETVAL <<
234 bool "Kernel Function Graph Return Val <<
235 depends on HAVE_FUNCTION_GRAPH_RETVAL <<
236 depends on FUNCTION_GRAPH_TRACER <<
237 default n <<
238 help <<
239 Support recording and printing the f <<
240 using function graph tracer. It can <<
241 that return errors. This feature is <<
242 enable it via the trace option funcg <<
243 See Documentation/trace/ftrace.rst <<
244 <<
245 config DYNAMIC_FTRACE 188 config DYNAMIC_FTRACE
246 bool "enable/disable function tracing 189 bool "enable/disable function tracing dynamically"
247 depends on FUNCTION_TRACER 190 depends on FUNCTION_TRACER
248 depends on HAVE_DYNAMIC_FTRACE 191 depends on HAVE_DYNAMIC_FTRACE
249 default y 192 default y
250 help 193 help
251 This option will modify all the call 194 This option will modify all the calls to function tracing
252 dynamically (will patch them out of 195 dynamically (will patch them out of the binary image and
253 replace them with a No-Op instructio 196 replace them with a No-Op instruction) on boot up. During
254 compile time, a table is made of all 197 compile time, a table is made of all the locations that ftrace
255 can function trace, and this table i 198 can function trace, and this table is linked into the kernel
256 image. When this is enabled, functio 199 image. When this is enabled, functions can be individually
257 enabled, and the functions not enabl 200 enabled, and the functions not enabled will not affect
258 performance of the system. 201 performance of the system.
259 202
260 See the files in /sys/kernel/tracing !! 203 See the files in /sys/kernel/debug/tracing:
261 available_filter_functions 204 available_filter_functions
262 set_ftrace_filter 205 set_ftrace_filter
263 set_ftrace_notrace 206 set_ftrace_notrace
264 207
265 This way a CONFIG_FUNCTION_TRACER ke 208 This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but
266 otherwise has native performance as 209 otherwise has native performance as long as no tracing is active.
267 210
268 config DYNAMIC_FTRACE_WITH_REGS 211 config DYNAMIC_FTRACE_WITH_REGS
269 def_bool y 212 def_bool y
270 depends on DYNAMIC_FTRACE 213 depends on DYNAMIC_FTRACE
271 depends on HAVE_DYNAMIC_FTRACE_WITH_RE 214 depends on HAVE_DYNAMIC_FTRACE_WITH_REGS
272 215
273 config DYNAMIC_FTRACE_WITH_DIRECT_CALLS 216 config DYNAMIC_FTRACE_WITH_DIRECT_CALLS
274 def_bool y 217 def_bool y
275 depends on DYNAMIC_FTRACE_WITH_REGS || !! 218 depends on DYNAMIC_FTRACE_WITH_REGS
276 depends on HAVE_DYNAMIC_FTRACE_WITH_DI 219 depends on HAVE_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
277 220
278 config DYNAMIC_FTRACE_WITH_CALL_OPS <<
279 def_bool y <<
280 depends on HAVE_DYNAMIC_FTRACE_WITH_CA <<
281 <<
282 config DYNAMIC_FTRACE_WITH_ARGS 221 config DYNAMIC_FTRACE_WITH_ARGS
283 def_bool y 222 def_bool y
284 depends on DYNAMIC_FTRACE 223 depends on DYNAMIC_FTRACE
285 depends on HAVE_DYNAMIC_FTRACE_WITH_AR 224 depends on HAVE_DYNAMIC_FTRACE_WITH_ARGS
286 225
287 config FPROBE <<
288 bool "Kernel Function Probe (fprobe)" <<
289 depends on FUNCTION_TRACER <<
290 depends on DYNAMIC_FTRACE_WITH_REGS <<
291 depends on HAVE_RETHOOK <<
292 select RETHOOK <<
293 default n <<
294 help <<
295 This option enables kernel function <<
296 The fprobe is similar to kprobes, bu <<
297 entries and exits. This also can pro <<
298 fprobe. <<
299 <<
300 If unsure, say N. <<
301 <<
302 config FUNCTION_PROFILER 226 config FUNCTION_PROFILER
303 bool "Kernel function profiler" 227 bool "Kernel function profiler"
304 depends on FUNCTION_TRACER 228 depends on FUNCTION_TRACER
305 default n 229 default n
306 help 230 help
307 This option enables the kernel funct 231 This option enables the kernel function profiler. A file is created
308 in debugfs called function_profile_e 232 in debugfs called function_profile_enabled which defaults to zero.
309 When a 1 is echoed into this file pr 233 When a 1 is echoed into this file profiling begins, and when a
310 zero is entered, profiling stops. A 234 zero is entered, profiling stops. A "functions" file is created in
311 the trace_stat directory; this file 235 the trace_stat directory; this file shows the list of functions that
312 have been hit and their counters. 236 have been hit and their counters.
313 237
314 If in doubt, say N. 238 If in doubt, say N.
315 239
316 config STACK_TRACER 240 config STACK_TRACER
317 bool "Trace max stack" 241 bool "Trace max stack"
318 depends on HAVE_FUNCTION_TRACER 242 depends on HAVE_FUNCTION_TRACER
319 select FUNCTION_TRACER 243 select FUNCTION_TRACER
320 select STACKTRACE 244 select STACKTRACE
321 select KALLSYMS 245 select KALLSYMS
322 help 246 help
323 This special tracer records the maxi 247 This special tracer records the maximum stack footprint of the
324 kernel and displays it in /sys/kerne !! 248 kernel and displays it in /sys/kernel/debug/tracing/stack_trace.
325 249
326 This tracer works by hooking into ev 250 This tracer works by hooking into every function call that the
327 kernel executes, and keeping a maxim 251 kernel executes, and keeping a maximum stack depth value and
328 stack-trace saved. If this is confi 252 stack-trace saved. If this is configured with DYNAMIC_FTRACE
329 then it will not have any overhead w 253 then it will not have any overhead while the stack tracer
330 is disabled. 254 is disabled.
331 255
332 To enable the stack tracer on bootup 256 To enable the stack tracer on bootup, pass in 'stacktrace'
333 on the kernel command line. 257 on the kernel command line.
334 258
335 The stack tracer can also be enabled 259 The stack tracer can also be enabled or disabled via the
336 sysctl kernel.stack_tracer_enabled 260 sysctl kernel.stack_tracer_enabled
337 261
338 Say N if unsure. 262 Say N if unsure.
339 263
340 config TRACE_PREEMPT_TOGGLE 264 config TRACE_PREEMPT_TOGGLE
341 bool 265 bool
342 help 266 help
343 Enables hooks which will be called w 267 Enables hooks which will be called when preemption is first disabled,
344 and last enabled. 268 and last enabled.
345 269
346 config IRQSOFF_TRACER 270 config IRQSOFF_TRACER
347 bool "Interrupts-off Latency Tracer" 271 bool "Interrupts-off Latency Tracer"
348 default n 272 default n
349 depends on TRACE_IRQFLAGS_SUPPORT 273 depends on TRACE_IRQFLAGS_SUPPORT
350 select TRACE_IRQFLAGS 274 select TRACE_IRQFLAGS
351 select GENERIC_TRACER 275 select GENERIC_TRACER
352 select TRACER_MAX_TRACE 276 select TRACER_MAX_TRACE
353 select RING_BUFFER_ALLOW_SWAP 277 select RING_BUFFER_ALLOW_SWAP
354 select TRACER_SNAPSHOT 278 select TRACER_SNAPSHOT
355 select TRACER_SNAPSHOT_PER_CPU_SWAP 279 select TRACER_SNAPSHOT_PER_CPU_SWAP
356 help 280 help
357 This option measures the time spent 281 This option measures the time spent in irqs-off critical
358 sections, with microsecond accuracy. 282 sections, with microsecond accuracy.
359 283
360 The default measurement method is a 284 The default measurement method is a maximum search, which is
361 disabled by default and can be runti 285 disabled by default and can be runtime (re-)started
362 via: 286 via:
363 287
364 echo 0 > /sys/kernel/tracing/tra !! 288 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency
365 289
366 (Note that kernel size and overhead 290 (Note that kernel size and overhead increase with this option
367 enabled. This option and the preempt 291 enabled. This option and the preempt-off timing option can be
368 used together or separately.) 292 used together or separately.)
369 293
370 config PREEMPT_TRACER 294 config PREEMPT_TRACER
371 bool "Preemption-off Latency Tracer" 295 bool "Preemption-off Latency Tracer"
372 default n 296 default n
373 depends on PREEMPTION 297 depends on PREEMPTION
374 select GENERIC_TRACER 298 select GENERIC_TRACER
375 select TRACER_MAX_TRACE 299 select TRACER_MAX_TRACE
376 select RING_BUFFER_ALLOW_SWAP 300 select RING_BUFFER_ALLOW_SWAP
377 select TRACER_SNAPSHOT 301 select TRACER_SNAPSHOT
378 select TRACER_SNAPSHOT_PER_CPU_SWAP 302 select TRACER_SNAPSHOT_PER_CPU_SWAP
379 select TRACE_PREEMPT_TOGGLE 303 select TRACE_PREEMPT_TOGGLE
380 help 304 help
381 This option measures the time spent 305 This option measures the time spent in preemption-off critical
382 sections, with microsecond accuracy. 306 sections, with microsecond accuracy.
383 307
384 The default measurement method is a 308 The default measurement method is a maximum search, which is
385 disabled by default and can be runti 309 disabled by default and can be runtime (re-)started
386 via: 310 via:
387 311
388 echo 0 > /sys/kernel/tracing/tra !! 312 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency
389 313
390 (Note that kernel size and overhead 314 (Note that kernel size and overhead increase with this option
391 enabled. This option and the irqs-of 315 enabled. This option and the irqs-off timing option can be
392 used together or separately.) 316 used together or separately.)
393 317
394 config SCHED_TRACER 318 config SCHED_TRACER
395 bool "Scheduling Latency Tracer" 319 bool "Scheduling Latency Tracer"
396 select GENERIC_TRACER 320 select GENERIC_TRACER
397 select CONTEXT_SWITCH_TRACER 321 select CONTEXT_SWITCH_TRACER
398 select TRACER_MAX_TRACE 322 select TRACER_MAX_TRACE
399 select TRACER_SNAPSHOT 323 select TRACER_SNAPSHOT
400 help 324 help
401 This tracer tracks the latency of th 325 This tracer tracks the latency of the highest priority task
402 to be scheduled in, starting from th 326 to be scheduled in, starting from the point it has woken up.
403 327
404 config HWLAT_TRACER 328 config HWLAT_TRACER
405 bool "Tracer to detect hardware latenc 329 bool "Tracer to detect hardware latencies (like SMIs)"
406 select GENERIC_TRACER 330 select GENERIC_TRACER
407 select TRACER_MAX_TRACE <<
408 help 331 help
409 This tracer, when enabled will create 332 This tracer, when enabled will create one or more kernel threads,
410 depending on what the cpumask file is 333 depending on what the cpumask file is set to, which each thread
411 spinning in a loop looking for interr 334 spinning in a loop looking for interruptions caused by
412 something other than the kernel. For 335 something other than the kernel. For example, if a
413 System Management Interrupt (SMI) tak 336 System Management Interrupt (SMI) takes a noticeable amount of
414 time, this tracer will detect it. Thi 337 time, this tracer will detect it. This is useful for testing
415 if a system is reliable for Real Time 338 if a system is reliable for Real Time tasks.
416 339
417 Some files are created in the tracing 340 Some files are created in the tracing directory when this
418 is enabled: 341 is enabled:
419 342
420 hwlat_detector/width - time in us 343 hwlat_detector/width - time in usecs for how long to spin for
421 hwlat_detector/window - time in us 344 hwlat_detector/window - time in usecs between the start of each
422 iteration 345 iteration
423 346
424 A kernel thread is created that will 347 A kernel thread is created that will spin with interrupts disabled
425 for "width" microseconds in every "wi 348 for "width" microseconds in every "window" cycle. It will not spin
426 for "window - width" microseconds, wh 349 for "window - width" microseconds, where the system can
427 continue to operate. 350 continue to operate.
428 351
429 The output will appear in the trace a 352 The output will appear in the trace and trace_pipe files.
430 353
431 When the tracer is not running, it ha 354 When the tracer is not running, it has no affect on the system,
432 but when it is running, it can cause 355 but when it is running, it can cause the system to be
433 periodically non responsive. Do not r 356 periodically non responsive. Do not run this tracer on a
434 production system. 357 production system.
435 358
436 To enable this tracer, echo in "hwlat 359 To enable this tracer, echo in "hwlat" into the current_tracer
437 file. Every time a latency is greater 360 file. Every time a latency is greater than tracing_thresh, it will
438 be recorded into the ring buffer. 361 be recorded into the ring buffer.
439 362
440 config OSNOISE_TRACER 363 config OSNOISE_TRACER
441 bool "OS Noise tracer" 364 bool "OS Noise tracer"
442 select GENERIC_TRACER 365 select GENERIC_TRACER
443 select TRACER_MAX_TRACE <<
444 help 366 help
445 In the context of high-performance c 367 In the context of high-performance computing (HPC), the Operating
446 System Noise (osnoise) refers to the 368 System Noise (osnoise) refers to the interference experienced by an
447 application due to activities inside 369 application due to activities inside the operating system. In the
448 context of Linux, NMIs, IRQs, SoftIR 370 context of Linux, NMIs, IRQs, SoftIRQs, and any other system thread
449 can cause noise to the system. Moreo 371 can cause noise to the system. Moreover, hardware-related jobs can
450 also cause noise, for example, via S 372 also cause noise, for example, via SMIs.
451 373
452 The osnoise tracer leverages the hwl 374 The osnoise tracer leverages the hwlat_detector by running a similar
453 loop with preemption, SoftIRQs and I 375 loop with preemption, SoftIRQs and IRQs enabled, thus allowing all
454 the sources of osnoise during its ex 376 the sources of osnoise during its execution. The osnoise tracer takes
455 note of the entry and exit point of 377 note of the entry and exit point of any source of interferences,
456 increasing a per-cpu interference co 378 increasing a per-cpu interference counter. It saves an interference
457 counter for each source of interfere 379 counter for each source of interference. The interference counter for
458 NMI, IRQs, SoftIRQs, and threads is 380 NMI, IRQs, SoftIRQs, and threads is increased anytime the tool
459 observes these interferences' entry 381 observes these interferences' entry events. When a noise happens
460 without any interference from the op 382 without any interference from the operating system level, the
461 hardware noise counter increases, po 383 hardware noise counter increases, pointing to a hardware-related
462 noise. In this way, osnoise can acco 384 noise. In this way, osnoise can account for any source of
463 interference. At the end of the peri 385 interference. At the end of the period, the osnoise tracer prints
464 the sum of all noise, the max single 386 the sum of all noise, the max single noise, the percentage of CPU
465 available for the thread, and the co 387 available for the thread, and the counters for the noise sources.
466 388
467 In addition to the tracer, a set of 389 In addition to the tracer, a set of tracepoints were added to
468 facilitate the identification of the 390 facilitate the identification of the osnoise source.
469 391
470 The output will appear in the trace 392 The output will appear in the trace and trace_pipe files.
471 393
472 To enable this tracer, echo in "osno 394 To enable this tracer, echo in "osnoise" into the current_tracer
473 file. 395 file.
474 396
475 config TIMERLAT_TRACER 397 config TIMERLAT_TRACER
476 bool "Timerlat tracer" 398 bool "Timerlat tracer"
477 select OSNOISE_TRACER 399 select OSNOISE_TRACER
478 select GENERIC_TRACER 400 select GENERIC_TRACER
479 help 401 help
480 The timerlat tracer aims to help the 402 The timerlat tracer aims to help the preemptive kernel developers
481 to find sources of wakeup latencies 403 to find sources of wakeup latencies of real-time threads.
482 404
483 The tracer creates a per-cpu kernel 405 The tracer creates a per-cpu kernel thread with real-time priority.
484 The tracer thread sets a periodic ti 406 The tracer thread sets a periodic timer to wakeup itself, and goes
485 to sleep waiting for the timer to fi 407 to sleep waiting for the timer to fire. At the wakeup, the thread
486 then computes a wakeup latency value 408 then computes a wakeup latency value as the difference between
487 the current time and the absolute ti 409 the current time and the absolute time that the timer was set
488 to expire. 410 to expire.
489 411
490 The tracer prints two lines at every 412 The tracer prints two lines at every activation. The first is the
491 timer latency observed at the hardir 413 timer latency observed at the hardirq context before the
492 activation of the thread. The second 414 activation of the thread. The second is the timer latency observed
493 by the thread, which is the same lev 415 by the thread, which is the same level that cyclictest reports. The
494 ACTIVATION ID field serves to relate 416 ACTIVATION ID field serves to relate the irq execution to its
495 respective thread execution. 417 respective thread execution.
496 418
497 The tracer is build on top of osnois 419 The tracer is build on top of osnoise tracer, and the osnoise:
498 events can be used to trace the sour 420 events can be used to trace the source of interference from NMI,
499 IRQs and other threads. It also enab 421 IRQs and other threads. It also enables the capture of the
500 stacktrace at the IRQ context, which 422 stacktrace at the IRQ context, which helps to identify the code
501 path that can cause thread delay. 423 path that can cause thread delay.
502 424
503 config MMIOTRACE 425 config MMIOTRACE
504 bool "Memory mapped IO tracing" 426 bool "Memory mapped IO tracing"
505 depends on HAVE_MMIOTRACE_SUPPORT && P 427 depends on HAVE_MMIOTRACE_SUPPORT && PCI
506 select GENERIC_TRACER 428 select GENERIC_TRACER
507 help 429 help
508 Mmiotrace traces Memory Mapped I/O a 430 Mmiotrace traces Memory Mapped I/O access and is meant for
509 debugging and reverse engineering. I 431 debugging and reverse engineering. It is called from the ioremap
510 implementation and works via page fa 432 implementation and works via page faults. Tracing is disabled by
511 default and can be enabled at run-ti 433 default and can be enabled at run-time.
512 434
513 See Documentation/trace/mmiotrace.rs 435 See Documentation/trace/mmiotrace.rst.
514 If you are not helping to develop dr 436 If you are not helping to develop drivers, say N.
515 437
516 config ENABLE_DEFAULT_TRACERS 438 config ENABLE_DEFAULT_TRACERS
517 bool "Trace process context switches a 439 bool "Trace process context switches and events"
518 depends on !GENERIC_TRACER 440 depends on !GENERIC_TRACER
519 select TRACING 441 select TRACING
520 help 442 help
521 This tracer hooks to various trace p 443 This tracer hooks to various trace points in the kernel,
522 allowing the user to pick and choose 444 allowing the user to pick and choose which trace point they
523 want to trace. It also includes the 445 want to trace. It also includes the sched_switch tracer plugin.
524 446
525 config FTRACE_SYSCALLS 447 config FTRACE_SYSCALLS
526 bool "Trace syscalls" 448 bool "Trace syscalls"
527 depends on HAVE_SYSCALL_TRACEPOINTS 449 depends on HAVE_SYSCALL_TRACEPOINTS
528 select GENERIC_TRACER 450 select GENERIC_TRACER
529 select KALLSYMS 451 select KALLSYMS
530 help 452 help
531 Basic tracer to catch the syscall en 453 Basic tracer to catch the syscall entry and exit events.
532 454
533 config TRACER_SNAPSHOT 455 config TRACER_SNAPSHOT
534 bool "Create a snapshot trace buffer" 456 bool "Create a snapshot trace buffer"
535 select TRACER_MAX_TRACE 457 select TRACER_MAX_TRACE
536 help 458 help
537 Allow tracing users to take snapshot 459 Allow tracing users to take snapshot of the current buffer using the
538 ftrace interface, e.g.: 460 ftrace interface, e.g.:
539 461
540 echo 1 > /sys/kernel/tracing/sna !! 462 echo 1 > /sys/kernel/debug/tracing/snapshot
541 cat snapshot 463 cat snapshot
542 464
543 config TRACER_SNAPSHOT_PER_CPU_SWAP 465 config TRACER_SNAPSHOT_PER_CPU_SWAP
544 bool "Allow snapshot to swap per CPU" 466 bool "Allow snapshot to swap per CPU"
545 depends on TRACER_SNAPSHOT 467 depends on TRACER_SNAPSHOT
546 select RING_BUFFER_ALLOW_SWAP 468 select RING_BUFFER_ALLOW_SWAP
547 help 469 help
548 Allow doing a snapshot of a single C 470 Allow doing a snapshot of a single CPU buffer instead of a
549 full swap (all buffers). If this is 471 full swap (all buffers). If this is set, then the following is
550 allowed: 472 allowed:
551 473
552 echo 1 > /sys/kernel/tracing/per !! 474 echo 1 > /sys/kernel/debug/tracing/per_cpu/cpu2/snapshot
553 475
554 After which, only the tracing buffer 476 After which, only the tracing buffer for CPU 2 was swapped with
555 the main tracing buffer, and the oth 477 the main tracing buffer, and the other CPU buffers remain the same.
556 478
557 When this is enabled, this adds a li 479 When this is enabled, this adds a little more overhead to the
558 trace recording, as it needs to add 480 trace recording, as it needs to add some checks to synchronize
559 recording with swaps. But this does 481 recording with swaps. But this does not affect the performance
560 of the overall system. This is enabl 482 of the overall system. This is enabled by default when the preempt
561 or irq latency tracers are enabled, 483 or irq latency tracers are enabled, as those need to swap as well
562 and already adds the overhead (plus 484 and already adds the overhead (plus a lot more).
563 485
564 config TRACE_BRANCH_PROFILING 486 config TRACE_BRANCH_PROFILING
565 bool 487 bool
566 select GENERIC_TRACER 488 select GENERIC_TRACER
567 489
568 choice 490 choice
569 prompt "Branch Profiling" 491 prompt "Branch Profiling"
570 default BRANCH_PROFILE_NONE 492 default BRANCH_PROFILE_NONE
571 help 493 help
572 The branch profiling is a software pr 494 The branch profiling is a software profiler. It will add hooks
573 into the C conditionals to test which 495 into the C conditionals to test which path a branch takes.
574 496
575 The likely/unlikely profiler only loo 497 The likely/unlikely profiler only looks at the conditions that
576 are annotated with a likely or unlike 498 are annotated with a likely or unlikely macro.
577 499
578 The "all branch" profiler will profil 500 The "all branch" profiler will profile every if-statement in the
579 kernel. This profiler will also enabl 501 kernel. This profiler will also enable the likely/unlikely
580 profiler. 502 profiler.
581 503
582 Either of the above profilers adds a 504 Either of the above profilers adds a bit of overhead to the system.
583 If unsure, choose "No branch profilin 505 If unsure, choose "No branch profiling".
584 506
585 config BRANCH_PROFILE_NONE 507 config BRANCH_PROFILE_NONE
586 bool "No branch profiling" 508 bool "No branch profiling"
587 help 509 help
588 No branch profiling. Branch profilin 510 No branch profiling. Branch profiling adds a bit of overhead.
589 Only enable it if you want to analys 511 Only enable it if you want to analyse the branching behavior.
590 Otherwise keep it disabled. 512 Otherwise keep it disabled.
591 513
592 config PROFILE_ANNOTATED_BRANCHES 514 config PROFILE_ANNOTATED_BRANCHES
593 bool "Trace likely/unlikely profiler" 515 bool "Trace likely/unlikely profiler"
594 select TRACE_BRANCH_PROFILING 516 select TRACE_BRANCH_PROFILING
595 help 517 help
596 This tracer profiles all likely and 518 This tracer profiles all likely and unlikely macros
597 in the kernel. It will display the r 519 in the kernel. It will display the results in:
598 520
599 /sys/kernel/tracing/trace_stat/branc !! 521 /sys/kernel/debug/tracing/trace_stat/branch_annotated
600 522
601 Note: this will add a significant ov 523 Note: this will add a significant overhead; only turn this
602 on if you need to profile the system 524 on if you need to profile the system's use of these macros.
603 525
604 config PROFILE_ALL_BRANCHES 526 config PROFILE_ALL_BRANCHES
605 bool "Profile all if conditionals" if 527 bool "Profile all if conditionals" if !FORTIFY_SOURCE
606 select TRACE_BRANCH_PROFILING 528 select TRACE_BRANCH_PROFILING
607 help 529 help
608 This tracer profiles all branch cond 530 This tracer profiles all branch conditions. Every if ()
609 taken in the kernel is recorded whet 531 taken in the kernel is recorded whether it hit or miss.
610 The results will be displayed in: 532 The results will be displayed in:
611 533
612 /sys/kernel/tracing/trace_stat/branc !! 534 /sys/kernel/debug/tracing/trace_stat/branch_all
613 535
614 This option also enables the likely/ 536 This option also enables the likely/unlikely profiler.
615 537
616 This configuration, when enabled, wi 538 This configuration, when enabled, will impose a great overhead
617 on the system. This should only be e 539 on the system. This should only be enabled when the system
618 is to be analyzed in much detail. 540 is to be analyzed in much detail.
619 endchoice 541 endchoice
620 542
621 config TRACING_BRANCHES 543 config TRACING_BRANCHES
622 bool 544 bool
623 help 545 help
624 Selected by tracers that will trace 546 Selected by tracers that will trace the likely and unlikely
625 conditions. This prevents the tracer 547 conditions. This prevents the tracers themselves from being
626 profiled. Profiling the tracing infr 548 profiled. Profiling the tracing infrastructure can only happen
627 when the likelys and unlikelys are n 549 when the likelys and unlikelys are not being traced.
628 550
629 config BRANCH_TRACER 551 config BRANCH_TRACER
630 bool "Trace likely/unlikely instances" 552 bool "Trace likely/unlikely instances"
631 depends on TRACE_BRANCH_PROFILING 553 depends on TRACE_BRANCH_PROFILING
632 select TRACING_BRANCHES 554 select TRACING_BRANCHES
633 help 555 help
634 This traces the events of likely and 556 This traces the events of likely and unlikely condition
635 calls in the kernel. The difference 557 calls in the kernel. The difference between this and the
636 "Trace likely/unlikely profiler" is 558 "Trace likely/unlikely profiler" is that this is not a
637 histogram of the callers, but actual 559 histogram of the callers, but actually places the calling
638 events into a running trace buffer t 560 events into a running trace buffer to see when and where the
639 events happened, as well as their re 561 events happened, as well as their results.
640 562
641 Say N if unsure. 563 Say N if unsure.
642 564
643 config BLK_DEV_IO_TRACE 565 config BLK_DEV_IO_TRACE
644 bool "Support for tracing block IO act 566 bool "Support for tracing block IO actions"
645 depends on SYSFS 567 depends on SYSFS
646 depends on BLOCK 568 depends on BLOCK
647 select RELAY 569 select RELAY
648 select DEBUG_FS 570 select DEBUG_FS
649 select TRACEPOINTS 571 select TRACEPOINTS
650 select GENERIC_TRACER 572 select GENERIC_TRACER
651 select STACKTRACE 573 select STACKTRACE
652 help 574 help
653 Say Y here if you want to be able to 575 Say Y here if you want to be able to trace the block layer actions
654 on a given queue. Tracing allows you 576 on a given queue. Tracing allows you to see any traffic happening
655 on a block device queue. For more in 577 on a block device queue. For more information (and the userspace
656 support tools needed), fetch the blk 578 support tools needed), fetch the blktrace tools from:
657 579
658 git://git.kernel.dk/blktrace.git 580 git://git.kernel.dk/blktrace.git
659 581
660 Tracing also is possible using the f 582 Tracing also is possible using the ftrace interface, e.g.:
661 583
662 echo 1 > /sys/block/sda/sda1/trace 584 echo 1 > /sys/block/sda/sda1/trace/enable
663 echo blk > /sys/kernel/tracing/cur !! 585 echo blk > /sys/kernel/debug/tracing/current_tracer
664 cat /sys/kernel/tracing/trace_pipe !! 586 cat /sys/kernel/debug/tracing/trace_pipe
665 587
666 If unsure, say N. 588 If unsure, say N.
667 589
668 config FPROBE_EVENTS <<
669 depends on FPROBE <<
670 depends on HAVE_REGS_AND_STACK_ACCESS_ <<
671 bool "Enable fprobe-based dynamic even <<
672 select TRACING <<
673 select PROBE_EVENTS <<
674 select DYNAMIC_EVENTS <<
675 default y <<
676 help <<
677 This allows user to add tracing even <<
678 exit via ftrace interface. The synta <<
679 and the kprobe events on function en <<
680 transparently converted to this fpro <<
681 <<
682 config PROBE_EVENTS_BTF_ARGS <<
683 depends on HAVE_FUNCTION_ARG_ACCESS_AP <<
684 depends on FPROBE_EVENTS || KPROBE_EVE <<
685 depends on DEBUG_INFO_BTF && BPF_SYSCA <<
686 bool "Support BTF function arguments f <<
687 default y <<
688 help <<
689 The user can specify the arguments o <<
690 of the arguments of the probed funct <<
691 kernel function entry or a tracepoin <<
692 This is available only if BTF (BPF T <<
693 <<
694 config KPROBE_EVENTS 590 config KPROBE_EVENTS
695 depends on KPROBES 591 depends on KPROBES
696 depends on HAVE_REGS_AND_STACK_ACCESS_ 592 depends on HAVE_REGS_AND_STACK_ACCESS_API
697 bool "Enable kprobes-based dynamic eve 593 bool "Enable kprobes-based dynamic events"
698 select TRACING 594 select TRACING
699 select PROBE_EVENTS 595 select PROBE_EVENTS
700 select DYNAMIC_EVENTS 596 select DYNAMIC_EVENTS
701 default y 597 default y
702 help 598 help
703 This allows the user to add tracing 599 This allows the user to add tracing events (similar to tracepoints)
704 on the fly via the ftrace interface. 600 on the fly via the ftrace interface. See
705 Documentation/trace/kprobetrace.rst 601 Documentation/trace/kprobetrace.rst for more details.
706 602
707 Those events can be inserted whereve 603 Those events can be inserted wherever kprobes can probe, and record
708 various register and memory values. 604 various register and memory values.
709 605
710 This option is also required by perf 606 This option is also required by perf-probe subcommand of perf tools.
711 If you want to use perf tools, this 607 If you want to use perf tools, this option is strongly recommended.
712 608
713 config KPROBE_EVENTS_ON_NOTRACE 609 config KPROBE_EVENTS_ON_NOTRACE
714 bool "Do NOT protect notrace function 610 bool "Do NOT protect notrace function from kprobe events"
715 depends on KPROBE_EVENTS 611 depends on KPROBE_EVENTS
716 depends on DYNAMIC_FTRACE 612 depends on DYNAMIC_FTRACE
717 default n 613 default n
718 help 614 help
719 This is only for the developers who 615 This is only for the developers who want to debug ftrace itself
720 using kprobe events. 616 using kprobe events.
721 617
722 If kprobes can use ftrace instead of 618 If kprobes can use ftrace instead of breakpoint, ftrace related
723 functions are protected from kprobe- 619 functions are protected from kprobe-events to prevent an infinite
724 recursion or any unexpected executio 620 recursion or any unexpected execution path which leads to a kernel
725 crash. 621 crash.
726 622
727 This option disables such protection 623 This option disables such protection and allows you to put kprobe
728 events on ftrace functions for debug 624 events on ftrace functions for debugging ftrace by itself.
729 Note that this might let you shoot y 625 Note that this might let you shoot yourself in the foot.
730 626
731 If unsure, say N. 627 If unsure, say N.
732 628
733 config UPROBE_EVENTS 629 config UPROBE_EVENTS
734 bool "Enable uprobes-based dynamic eve 630 bool "Enable uprobes-based dynamic events"
735 depends on ARCH_SUPPORTS_UPROBES 631 depends on ARCH_SUPPORTS_UPROBES
736 depends on MMU 632 depends on MMU
737 depends on PERF_EVENTS 633 depends on PERF_EVENTS
738 select UPROBES 634 select UPROBES
739 select PROBE_EVENTS 635 select PROBE_EVENTS
740 select DYNAMIC_EVENTS 636 select DYNAMIC_EVENTS
741 select TRACING 637 select TRACING
742 default y 638 default y
743 help 639 help
744 This allows the user to add tracing 640 This allows the user to add tracing events on top of userspace
745 dynamic events (similar to tracepoin 641 dynamic events (similar to tracepoints) on the fly via the trace
746 events interface. Those events can b 642 events interface. Those events can be inserted wherever uprobes
747 can probe, and record various regist 643 can probe, and record various registers.
748 This option is required if you plan 644 This option is required if you plan to use perf-probe subcommand
749 of perf tools on user space applicat 645 of perf tools on user space applications.
750 646
751 config BPF_EVENTS 647 config BPF_EVENTS
752 depends on BPF_SYSCALL 648 depends on BPF_SYSCALL
753 depends on (KPROBE_EVENTS || UPROBE_EV 649 depends on (KPROBE_EVENTS || UPROBE_EVENTS) && PERF_EVENTS
754 bool 650 bool
755 default y 651 default y
756 help 652 help
757 This allows the user to attach BPF p 653 This allows the user to attach BPF programs to kprobe, uprobe, and
758 tracepoint events. 654 tracepoint events.
759 655
760 config DYNAMIC_EVENTS 656 config DYNAMIC_EVENTS
761 def_bool n 657 def_bool n
762 658
763 config PROBE_EVENTS 659 config PROBE_EVENTS
764 def_bool n 660 def_bool n
765 661
766 config BPF_KPROBE_OVERRIDE 662 config BPF_KPROBE_OVERRIDE
767 bool "Enable BPF programs to override 663 bool "Enable BPF programs to override a kprobed function"
768 depends on BPF_EVENTS 664 depends on BPF_EVENTS
769 depends on FUNCTION_ERROR_INJECTION 665 depends on FUNCTION_ERROR_INJECTION
770 default n 666 default n
771 help 667 help
772 Allows BPF to override the execution 668 Allows BPF to override the execution of a probed function and
773 set a different return value. This i 669 set a different return value. This is used for error injection.
774 670
775 config FTRACE_MCOUNT_RECORD 671 config FTRACE_MCOUNT_RECORD
776 def_bool y 672 def_bool y
777 depends on DYNAMIC_FTRACE 673 depends on DYNAMIC_FTRACE
778 depends on HAVE_FTRACE_MCOUNT_RECORD 674 depends on HAVE_FTRACE_MCOUNT_RECORD
779 675
780 config FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_EN 676 config FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
781 bool 677 bool
782 depends on FTRACE_MCOUNT_RECORD 678 depends on FTRACE_MCOUNT_RECORD
783 679
784 config FTRACE_MCOUNT_USE_CC 680 config FTRACE_MCOUNT_USE_CC
785 def_bool y 681 def_bool y
786 depends on $(cc-option,-mrecord-mcount 682 depends on $(cc-option,-mrecord-mcount)
787 depends on !FTRACE_MCOUNT_USE_PATCHABL 683 depends on !FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
788 depends on FTRACE_MCOUNT_RECORD 684 depends on FTRACE_MCOUNT_RECORD
789 685
790 config FTRACE_MCOUNT_USE_OBJTOOL 686 config FTRACE_MCOUNT_USE_OBJTOOL
791 def_bool y 687 def_bool y
792 depends on HAVE_OBJTOOL_MCOUNT 688 depends on HAVE_OBJTOOL_MCOUNT
793 depends on !FTRACE_MCOUNT_USE_PATCHABL 689 depends on !FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
794 depends on !FTRACE_MCOUNT_USE_CC 690 depends on !FTRACE_MCOUNT_USE_CC
795 depends on FTRACE_MCOUNT_RECORD 691 depends on FTRACE_MCOUNT_RECORD
796 select OBJTOOL <<
797 692
798 config FTRACE_MCOUNT_USE_RECORDMCOUNT 693 config FTRACE_MCOUNT_USE_RECORDMCOUNT
799 def_bool y 694 def_bool y
800 depends on !FTRACE_MCOUNT_USE_PATCHABL 695 depends on !FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY
801 depends on !FTRACE_MCOUNT_USE_CC 696 depends on !FTRACE_MCOUNT_USE_CC
802 depends on !FTRACE_MCOUNT_USE_OBJTOOL 697 depends on !FTRACE_MCOUNT_USE_OBJTOOL
803 depends on FTRACE_MCOUNT_RECORD 698 depends on FTRACE_MCOUNT_RECORD
804 699
805 config TRACING_MAP 700 config TRACING_MAP
806 bool 701 bool
807 depends on ARCH_HAVE_NMI_SAFE_CMPXCHG 702 depends on ARCH_HAVE_NMI_SAFE_CMPXCHG
808 help 703 help
809 tracing_map is a special-purpose loc 704 tracing_map is a special-purpose lock-free map for tracing,
810 separated out as a stand-alone facil 705 separated out as a stand-alone facility in order to allow it
811 to be shared between multiple tracer 706 to be shared between multiple tracers. It isn't meant to be
812 generally used outside of that conte 707 generally used outside of that context, and is normally
813 selected by tracers that use it. 708 selected by tracers that use it.
814 709
815 config SYNTH_EVENTS 710 config SYNTH_EVENTS
816 bool "Synthetic trace events" 711 bool "Synthetic trace events"
817 select TRACING 712 select TRACING
818 select DYNAMIC_EVENTS 713 select DYNAMIC_EVENTS
819 default n 714 default n
820 help 715 help
821 Synthetic events are user-defined tr 716 Synthetic events are user-defined trace events that can be
822 used to combine data from other trac 717 used to combine data from other trace events or in fact any
823 data source. Synthetic events can b 718 data source. Synthetic events can be generated indirectly
824 via the trace() action of histogram 719 via the trace() action of histogram triggers or directly
825 by way of an in-kernel API. 720 by way of an in-kernel API.
826 721
827 See Documentation/trace/events.rst o 722 See Documentation/trace/events.rst or
828 Documentation/trace/histogram.rst fo 723 Documentation/trace/histogram.rst for details and examples.
829 724
830 If in doubt, say N. 725 If in doubt, say N.
831 726
832 config USER_EVENTS <<
833 bool "User trace events" <<
834 select TRACING <<
835 select DYNAMIC_EVENTS <<
836 help <<
837 User trace events are user-defined t <<
838 can be used like an existing kernel <<
839 events are generated by writing to a <<
840 processes can determine if their tra <<
841 generated by registering a value and <<
842 that reflects when it is enabled or <<
843 <<
844 See Documentation/trace/user_events. <<
845 If in doubt, say N. <<
846 <<
847 config HIST_TRIGGERS 727 config HIST_TRIGGERS
848 bool "Histogram triggers" 728 bool "Histogram triggers"
849 depends on ARCH_HAVE_NMI_SAFE_CMPXCHG 729 depends on ARCH_HAVE_NMI_SAFE_CMPXCHG
850 select TRACING_MAP 730 select TRACING_MAP
851 select TRACING 731 select TRACING
852 select DYNAMIC_EVENTS 732 select DYNAMIC_EVENTS
853 select SYNTH_EVENTS 733 select SYNTH_EVENTS
854 default n 734 default n
855 help 735 help
856 Hist triggers allow one or more arbi 736 Hist triggers allow one or more arbitrary trace event fields
857 to be aggregated into hash tables an 737 to be aggregated into hash tables and dumped to stdout by
858 reading a debugfs/tracefs file. The 738 reading a debugfs/tracefs file. They're useful for
859 gathering quick and dirty (though pr 739 gathering quick and dirty (though precise) summaries of
860 event activity as an initial guide f 740 event activity as an initial guide for further investigation
861 using more advanced tools. 741 using more advanced tools.
862 742
863 Inter-event tracing of quantities su 743 Inter-event tracing of quantities such as latencies is also
864 supported using hist triggers under 744 supported using hist triggers under this option.
865 745
866 See Documentation/trace/histogram.rs 746 See Documentation/trace/histogram.rst.
867 If in doubt, say N. 747 If in doubt, say N.
868 748
869 config TRACE_EVENT_INJECT 749 config TRACE_EVENT_INJECT
870 bool "Trace event injection" 750 bool "Trace event injection"
871 depends on TRACING 751 depends on TRACING
872 help 752 help
873 Allow user-space to inject a specifi 753 Allow user-space to inject a specific trace event into the ring
874 buffer. This is mainly used for test 754 buffer. This is mainly used for testing purpose.
875 755
876 If unsure, say N. 756 If unsure, say N.
877 757
878 config TRACEPOINT_BENCHMARK 758 config TRACEPOINT_BENCHMARK
879 bool "Add tracepoint that benchmarks t 759 bool "Add tracepoint that benchmarks tracepoints"
880 help 760 help
881 This option creates the tracepoint "b 761 This option creates the tracepoint "benchmark:benchmark_event".
882 When the tracepoint is enabled, it ki 762 When the tracepoint is enabled, it kicks off a kernel thread that
883 goes into an infinite loop (calling c 763 goes into an infinite loop (calling cond_resched() to let other tasks
884 run), and calls the tracepoint. Each 764 run), and calls the tracepoint. Each iteration will record the time
885 it took to write to the tracepoint an 765 it took to write to the tracepoint and the next iteration that
886 data will be passed to the tracepoint 766 data will be passed to the tracepoint itself. That is, the tracepoint
887 will report the time it took to do th 767 will report the time it took to do the previous tracepoint.
888 The string written to the tracepoint 768 The string written to the tracepoint is a static string of 128 bytes
889 to keep the time the same. The initia 769 to keep the time the same. The initial string is simply a write of
890 "START". The second string records th 770 "START". The second string records the cold cache time of the first
891 write which is not added to the rest 771 write which is not added to the rest of the calculations.
892 772
893 As it is a tight loop, it benchmarks 773 As it is a tight loop, it benchmarks as hot cache. That's fine because
894 we care most about hot paths that are 774 we care most about hot paths that are probably in cache already.
895 775
896 An example of the output: 776 An example of the output:
897 777
898 START 778 START
899 first=3672 [COLD CACHED] 779 first=3672 [COLD CACHED]
900 last=632 first=3672 max=632 min= 780 last=632 first=3672 max=632 min=632 avg=316 std=446 std^2=199712
901 last=278 first=3672 max=632 min= 781 last=278 first=3672 max=632 min=278 avg=303 std=316 std^2=100337
902 last=277 first=3672 max=632 min= 782 last=277 first=3672 max=632 min=277 avg=296 std=258 std^2=67064
903 last=273 first=3672 max=632 min= 783 last=273 first=3672 max=632 min=273 avg=292 std=224 std^2=50411
904 last=273 first=3672 max=632 min= 784 last=273 first=3672 max=632 min=273 avg=288 std=200 std^2=40389
905 last=281 first=3672 max=632 min= 785 last=281 first=3672 max=632 min=273 avg=287 std=183 std^2=33666
906 786
907 787
908 config RING_BUFFER_BENCHMARK 788 config RING_BUFFER_BENCHMARK
909 tristate "Ring buffer benchmark stress 789 tristate "Ring buffer benchmark stress tester"
910 depends on RING_BUFFER 790 depends on RING_BUFFER
911 help 791 help
912 This option creates a test to stress 792 This option creates a test to stress the ring buffer and benchmark it.
913 It creates its own ring buffer such 793 It creates its own ring buffer such that it will not interfere with
914 any other users of the ring buffer ( 794 any other users of the ring buffer (such as ftrace). It then creates
915 a producer and consumer that will ru 795 a producer and consumer that will run for 10 seconds and sleep for
916 10 seconds. Each interval it will pr 796 10 seconds. Each interval it will print out the number of events
917 it recorded and give a rough estimat 797 it recorded and give a rough estimate of how long each iteration took.
918 798
919 It does not disable interrupts or ra 799 It does not disable interrupts or raise its priority, so it may be
920 affected by processes that are runni 800 affected by processes that are running.
921 801
922 If unsure, say N. 802 If unsure, say N.
923 803
924 config TRACE_EVAL_MAP_FILE 804 config TRACE_EVAL_MAP_FILE
925 bool "Show eval mappings for trace even 805 bool "Show eval mappings for trace events"
926 depends on TRACING 806 depends on TRACING
927 help 807 help
928 The "print fmt" of the trace events wi 808 The "print fmt" of the trace events will show the enum/sizeof names
929 instead of their values. This can caus 809 instead of their values. This can cause problems for user space tools
930 that use this string to parse the raw 810 that use this string to parse the raw data as user space does not know
931 how to convert the string to its value 811 how to convert the string to its value.
932 812
933 To fix this, there's a special macro i 813 To fix this, there's a special macro in the kernel that can be used
934 to convert an enum/sizeof into its val 814 to convert an enum/sizeof into its value. If this macro is used, then
935 the print fmt strings will be converte 815 the print fmt strings will be converted to their values.
936 816
937 If something does not get converted pr 817 If something does not get converted properly, this option can be
938 used to show what enums/sizeof the ker 818 used to show what enums/sizeof the kernel tried to convert.
939 819
940 This option is for debugging the conve 820 This option is for debugging the conversions. A file is created
941 in the tracing directory called "eval_ 821 in the tracing directory called "eval_map" that will show the
942 names matched with their values and wh 822 names matched with their values and what trace event system they
943 belong too. 823 belong too.
944 824
945 Normally, the mapping of the strings t 825 Normally, the mapping of the strings to values will be freed after
946 boot up or module load. With this opti 826 boot up or module load. With this option, they will not be freed, as
947 they are needed for the "eval_map" fil 827 they are needed for the "eval_map" file. Enabling this option will
948 increase the memory footprint of the r 828 increase the memory footprint of the running kernel.
949 829
950 If unsure, say N. 830 If unsure, say N.
951 831
952 config FTRACE_RECORD_RECURSION 832 config FTRACE_RECORD_RECURSION
953 bool "Record functions that recurse in 833 bool "Record functions that recurse in function tracing"
954 depends on FUNCTION_TRACER 834 depends on FUNCTION_TRACER
955 help 835 help
956 All callbacks that attach to the fun 836 All callbacks that attach to the function tracing have some sort
957 of protection against recursion. Eve 837 of protection against recursion. Even though the protection exists,
958 it adds overhead. This option will c 838 it adds overhead. This option will create a file in the tracefs
959 file system called "recursed_functio 839 file system called "recursed_functions" that will list the functions
960 that triggered a recursion. 840 that triggered a recursion.
961 841
962 This will add more overhead to cases 842 This will add more overhead to cases that have recursion.
963 843
964 If unsure, say N 844 If unsure, say N
965 845
966 config FTRACE_RECORD_RECURSION_SIZE 846 config FTRACE_RECORD_RECURSION_SIZE
967 int "Max number of recursed functions 847 int "Max number of recursed functions to record"
968 default 128 !! 848 default 128
969 depends on FTRACE_RECORD_RECURSION 849 depends on FTRACE_RECORD_RECURSION
970 help 850 help
971 This defines the limit of number of 851 This defines the limit of number of functions that can be
972 listed in the "recursed_functions" f 852 listed in the "recursed_functions" file, that lists all
973 the functions that caused a recursio 853 the functions that caused a recursion to happen.
974 This file can be reset, but the limi 854 This file can be reset, but the limit can not change in
975 size at runtime. 855 size at runtime.
976 856
977 config FTRACE_VALIDATE_RCU_IS_WATCHING <<
978 bool "Validate RCU is on during ftrace <<
979 depends on FUNCTION_TRACER <<
980 depends on ARCH_WANTS_NO_INSTR <<
981 help <<
982 All callbacks that attach to the fun <<
983 protection against recursion. This o <<
984 ftrace (and other users of ftrace_te <<
985 called outside of RCU, as if they ar <<
986 also has a noticeable overhead when <<
987 <<
988 If unsure, say N <<
989 <<
990 config RING_BUFFER_RECORD_RECURSION 857 config RING_BUFFER_RECORD_RECURSION
991 bool "Record functions that recurse in 858 bool "Record functions that recurse in the ring buffer"
992 depends on FTRACE_RECORD_RECURSION 859 depends on FTRACE_RECORD_RECURSION
993 # default y, because it is coupled wit 860 # default y, because it is coupled with FTRACE_RECORD_RECURSION
994 default y 861 default y
995 help 862 help
996 The ring buffer has its own internal 863 The ring buffer has its own internal recursion. Although when
997 recursion happens it won't cause har !! 864 recursion happens it wont cause harm because of the protection,
998 but it does cause unwanted overhead. !! 865 but it does cause an unwanted overhead. Enabling this option will
999 place where recursion was detected i 866 place where recursion was detected into the ftrace "recursed_functions"
1000 file. 867 file.
1001 868
1002 This will add more overhead to case 869 This will add more overhead to cases that have recursion.
1003 870
1004 config GCOV_PROFILE_FTRACE 871 config GCOV_PROFILE_FTRACE
1005 bool "Enable GCOV profiling on ftrace 872 bool "Enable GCOV profiling on ftrace subsystem"
1006 depends on GCOV_KERNEL 873 depends on GCOV_KERNEL
1007 help 874 help
1008 Enable GCOV profiling on ftrace sub 875 Enable GCOV profiling on ftrace subsystem for checking
1009 which functions/lines are tested. 876 which functions/lines are tested.
1010 877
1011 If unsure, say N. 878 If unsure, say N.
1012 879
1013 Note that on a kernel compiled with 880 Note that on a kernel compiled with this config, ftrace will
1014 run significantly slower. 881 run significantly slower.
1015 882
1016 config FTRACE_SELFTEST 883 config FTRACE_SELFTEST
1017 bool 884 bool
1018 885
1019 config FTRACE_STARTUP_TEST 886 config FTRACE_STARTUP_TEST
1020 bool "Perform a startup test on ftrac 887 bool "Perform a startup test on ftrace"
1021 depends on GENERIC_TRACER 888 depends on GENERIC_TRACER
1022 select FTRACE_SELFTEST 889 select FTRACE_SELFTEST
1023 help 890 help
1024 This option performs a series of st 891 This option performs a series of startup tests on ftrace. On bootup
1025 a series of tests are made to verif 892 a series of tests are made to verify that the tracer is
1026 functioning properly. It will do te 893 functioning properly. It will do tests on all the configured
1027 tracers of ftrace. 894 tracers of ftrace.
1028 895
1029 config EVENT_TRACE_STARTUP_TEST 896 config EVENT_TRACE_STARTUP_TEST
1030 bool "Run selftest on trace events" 897 bool "Run selftest on trace events"
1031 depends on FTRACE_STARTUP_TEST 898 depends on FTRACE_STARTUP_TEST
1032 default y 899 default y
1033 help 900 help
1034 This option performs a test on all 901 This option performs a test on all trace events in the system.
1035 It basically just enables each even 902 It basically just enables each event and runs some code that
1036 will trigger events (not necessaril 903 will trigger events (not necessarily the event it enables)
1037 This may take some time run as ther 904 This may take some time run as there are a lot of events.
1038 905
1039 config EVENT_TRACE_TEST_SYSCALLS 906 config EVENT_TRACE_TEST_SYSCALLS
1040 bool "Run selftest on syscall events" 907 bool "Run selftest on syscall events"
1041 depends on EVENT_TRACE_STARTUP_TEST 908 depends on EVENT_TRACE_STARTUP_TEST
1042 help 909 help
1043 This option will also enable testing 910 This option will also enable testing every syscall event.
1044 It only enables the event and disabl 911 It only enables the event and disables it and runs various loads
1045 with the event enabled. This adds a 912 with the event enabled. This adds a bit more time for kernel boot
1046 up since it runs this on every syste 913 up since it runs this on every system call defined.
1047 914
1048 TBD - enable a way to actually call 915 TBD - enable a way to actually call the syscalls as we test their
1049 events 916 events
1050 917
1051 config FTRACE_SORT_STARTUP_TEST <<
1052 bool "Verify compile time sorting of f <<
1053 depends on DYNAMIC_FTRACE <<
1054 depends on BUILDTIME_MCOUNT_SORT <<
1055 help <<
1056 Sorting of the mcount_loc sections t <<
1057 where the ftrace knows where to patc <<
1058 and other callbacks is done at compi <<
1059 is not done correctly, it will cause <<
1060 When this is set, the sorted section <<
1061 are in deed sorted and will warn if <<
1062 <<
1063 If unsure, say N <<
1064 <<
1065 config RING_BUFFER_STARTUP_TEST 918 config RING_BUFFER_STARTUP_TEST
1066 bool "Ring buffer startup self test" 919 bool "Ring buffer startup self test"
1067 depends on RING_BUFFER 920 depends on RING_BUFFER
1068 help 921 help
1069 Run a simple self test on the ring b 922 Run a simple self test on the ring buffer on boot up. Late in the
1070 kernel boot sequence, the test will 923 kernel boot sequence, the test will start that kicks off
1071 a thread per cpu. Each thread will w 924 a thread per cpu. Each thread will write various size events
1072 into the ring buffer. Another thread 925 into the ring buffer. Another thread is created to send IPIs
1073 to each of the threads, where the IP 926 to each of the threads, where the IPI handler will also write
1074 to the ring buffer, to test/stress t 927 to the ring buffer, to test/stress the nesting ability.
1075 If any anomalies are discovered, a w 928 If any anomalies are discovered, a warning will be displayed
1076 and all ring buffers will be disable 929 and all ring buffers will be disabled.
1077 930
1078 The test runs for 10 seconds. This w 931 The test runs for 10 seconds. This will slow your boot time
1079 by at least 10 more seconds. 932 by at least 10 more seconds.
1080 933
1081 At the end of the test, statistics a !! 934 At the end of the test, statics and more checks are done.
1082 It will output the stats of each per !! 935 It will output the stats of each per cpu buffer. What
1083 was written, the sizes, what was rea 936 was written, the sizes, what was read, what was lost, and
1084 other similar details. 937 other similar details.
1085 938
1086 If unsure, say N 939 If unsure, say N
1087 940
1088 config RING_BUFFER_VALIDATE_TIME_DELTAS 941 config RING_BUFFER_VALIDATE_TIME_DELTAS
1089 bool "Verify ring buffer time stamp d 942 bool "Verify ring buffer time stamp deltas"
1090 depends on RING_BUFFER 943 depends on RING_BUFFER
1091 help 944 help
1092 This will audit the time stamps on 945 This will audit the time stamps on the ring buffer sub
1093 buffer to make sure that all the ti 946 buffer to make sure that all the time deltas for the
1094 events on a sub buffer matches the 947 events on a sub buffer matches the current time stamp.
1095 This audit is performed for every e 948 This audit is performed for every event that is not
1096 interrupted, or interrupting anothe 949 interrupted, or interrupting another event. A check
1097 is also made when traversing sub bu 950 is also made when traversing sub buffers to make sure
1098 that all the deltas on the previous 951 that all the deltas on the previous sub buffer do not
1099 add up to be greater than the curre 952 add up to be greater than the current time stamp.
1100 953
1101 NOTE: This adds significant overhea 954 NOTE: This adds significant overhead to recording of events,
1102 and should only be used to test the 955 and should only be used to test the logic of the ring buffer.
1103 Do not use it on production systems 956 Do not use it on production systems.
1104 957
1105 Only say Y if you understand what t 958 Only say Y if you understand what this does, and you
1106 still want it enabled. Otherwise sa 959 still want it enabled. Otherwise say N
1107 960
1108 config MMIOTRACE_TEST 961 config MMIOTRACE_TEST
1109 tristate "Test module for mmiotrace" 962 tristate "Test module for mmiotrace"
1110 depends on MMIOTRACE && m 963 depends on MMIOTRACE && m
1111 help 964 help
1112 This is a dumb module for testing m 965 This is a dumb module for testing mmiotrace. It is very dangerous
1113 as it will write garbage to IO memo 966 as it will write garbage to IO memory starting at a given address.
1114 However, it should be safe to use o 967 However, it should be safe to use on e.g. unused portion of VRAM.
1115 968
1116 Say N, unless you absolutely know w 969 Say N, unless you absolutely know what you are doing.
1117 970
1118 config PREEMPTIRQ_DELAY_TEST 971 config PREEMPTIRQ_DELAY_TEST
1119 tristate "Test module to create a pre 972 tristate "Test module to create a preempt / IRQ disable delay thread to test latency tracers"
1120 depends on m 973 depends on m
1121 help 974 help
1122 Select this option to build a test 975 Select this option to build a test module that can help test latency
1123 tracers by executing a preempt or i 976 tracers by executing a preempt or irq disable section with a user
1124 configurable delay. The module busy 977 configurable delay. The module busy waits for the duration of the
1125 critical section. 978 critical section.
1126 979
1127 For example, the following invocati 980 For example, the following invocation generates a burst of three
1128 irq-disabled critical sections for 981 irq-disabled critical sections for 500us:
1129 modprobe preemptirq_delay_test test 982 modprobe preemptirq_delay_test test_mode=irq delay=500 burst_size=3
1130 983
1131 What's more, if you want to attach 984 What's more, if you want to attach the test on the cpu which the latency
1132 tracer is running on, specify cpu_a 985 tracer is running on, specify cpu_affinity=cpu_num at the end of the
1133 command. 986 command.
1134 987
1135 If unsure, say N 988 If unsure, say N
1136 989
1137 config SYNTH_EVENT_GEN_TEST 990 config SYNTH_EVENT_GEN_TEST
1138 tristate "Test module for in-kernel s 991 tristate "Test module for in-kernel synthetic event generation"
1139 depends on SYNTH_EVENTS && m !! 992 depends on SYNTH_EVENTS
1140 help 993 help
1141 This option creates a test module t 994 This option creates a test module to check the base
1142 functionality of in-kernel syntheti 995 functionality of in-kernel synthetic event definition and
1143 generation. 996 generation.
1144 997
1145 To test, insert the module, and the 998 To test, insert the module, and then check the trace buffer
1146 for the generated sample events. 999 for the generated sample events.
1147 1000
1148 If unsure, say N. 1001 If unsure, say N.
1149 1002
1150 config KPROBE_EVENT_GEN_TEST 1003 config KPROBE_EVENT_GEN_TEST
1151 tristate "Test module for in-kernel k 1004 tristate "Test module for in-kernel kprobe event generation"
1152 depends on KPROBE_EVENTS && m !! 1005 depends on KPROBE_EVENTS
1153 help 1006 help
1154 This option creates a test module t 1007 This option creates a test module to check the base
1155 functionality of in-kernel kprobe e 1008 functionality of in-kernel kprobe event definition.
1156 1009
1157 To test, insert the module, and the 1010 To test, insert the module, and then check the trace buffer
1158 for the generated kprobe events. 1011 for the generated kprobe events.
1159 1012
1160 If unsure, say N. 1013 If unsure, say N.
1161 1014
1162 config HIST_TRIGGERS_DEBUG 1015 config HIST_TRIGGERS_DEBUG
1163 bool "Hist trigger debug support" 1016 bool "Hist trigger debug support"
1164 depends on HIST_TRIGGERS 1017 depends on HIST_TRIGGERS
1165 help 1018 help
1166 Add "hist_debug" file for each even 1019 Add "hist_debug" file for each event, which when read will
1167 dump out a bunch of internal detail 1020 dump out a bunch of internal details about the hist triggers
1168 defined on that event. 1021 defined on that event.
1169 1022
1170 The hist_debug file serves a couple 1023 The hist_debug file serves a couple of purposes:
1171 1024
1172 - Helps developers verify that no 1025 - Helps developers verify that nothing is broken.
1173 1026
1174 - Provides educational informatio 1027 - Provides educational information to support the details
1175 of the hist trigger internals a 1028 of the hist trigger internals as described by
1176 Documentation/trace/histogram-d 1029 Documentation/trace/histogram-design.rst.
1177 1030
1178 The hist_debug output only covers t 1031 The hist_debug output only covers the data structures
1179 related to the histogram definition 1032 related to the histogram definitions themselves and doesn't
1180 display the internals of map bucket 1033 display the internals of map buckets or variable values of
1181 running histograms. 1034 running histograms.
1182 1035
1183 If unsure, say N. 1036 If unsure, say N.
1184 <<
1185 source "kernel/trace/rv/Kconfig" <<
1186 1037
1187 endif # FTRACE 1038 endif # FTRACE
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.