1 ======================== 1 ========================
2 ftrace - Function Tracer 2 ftrace - Function Tracer
3 ======================== 3 ========================
4 4
5 Copyright 2008 Red Hat Inc. 5 Copyright 2008 Red Hat Inc.
6 6
7 :Author: Steven Rostedt <srostedt@redhat.com> 7 :Author: Steven Rostedt <srostedt@redhat.com>
8 :License: The GNU Free Documentation License, 8 :License: The GNU Free Documentation License, Version 1.2
9 (dual licensed under the GPL v2) 9 (dual licensed under the GPL v2)
10 :Original Reviewers: Elias Oltmanns, Randy Du 10 :Original Reviewers: Elias Oltmanns, Randy Dunlap, Andrew Morton,
11 John Kacur, and David Te 11 John Kacur, and David Teigland.
12 12
13 - Written for: 2.6.28-rc2 13 - Written for: 2.6.28-rc2
14 - Updated for: 3.10 14 - Updated for: 3.10
15 - Updated for: 4.13 - Copyright 2017 VMware In 15 - Updated for: 4.13 - Copyright 2017 VMware Inc. Steven Rostedt
16 - Converted to rst format - Changbin Du <changb 16 - Converted to rst format - Changbin Du <changbin.du@intel.com>
17 17
18 Introduction 18 Introduction
19 ------------ 19 ------------
20 20
21 Ftrace is an internal tracer designed to help 21 Ftrace is an internal tracer designed to help out developers and
22 designers of systems to find what is going on 22 designers of systems to find what is going on inside the kernel.
23 It can be used for debugging or analyzing late 23 It can be used for debugging or analyzing latencies and
24 performance issues that take place outside of 24 performance issues that take place outside of user-space.
25 25
26 Although ftrace is typically considered the fu 26 Although ftrace is typically considered the function tracer, it
27 is really a framework of several assorted trac 27 is really a framework of several assorted tracing utilities.
28 There's latency tracing to examine what occurs 28 There's latency tracing to examine what occurs between interrupts
29 disabled and enabled, as well as for preemptio 29 disabled and enabled, as well as for preemption and from a time
30 a task is woken to the task is actually schedu 30 a task is woken to the task is actually scheduled in.
31 31
32 One of the most common uses of ftrace is the e 32 One of the most common uses of ftrace is the event tracing.
33 Throughout the kernel is hundreds of static ev 33 Throughout the kernel is hundreds of static event points that
34 can be enabled via the tracefs file system to 34 can be enabled via the tracefs file system to see what is
35 going on in certain parts of the kernel. 35 going on in certain parts of the kernel.
36 36
37 See events.rst for more information. !! 37 See events.txt for more information.
38 38
39 39
40 Implementation Details 40 Implementation Details
41 ---------------------- 41 ----------------------
42 42
43 See Documentation/trace/ftrace-design.rst for !! 43 See :doc:`ftrace-design` for details for arch porters and such.
44 44
45 45
46 The File System 46 The File System
47 --------------- 47 ---------------
48 48
49 Ftrace uses the tracefs file system to hold th 49 Ftrace uses the tracefs file system to hold the control files as
50 well as the files to display output. 50 well as the files to display output.
51 51
52 When tracefs is configured into the kernel (wh 52 When tracefs is configured into the kernel (which selecting any ftrace
53 option will do) the directory /sys/kernel/trac 53 option will do) the directory /sys/kernel/tracing will be created. To mount
54 this directory, you can add to your /etc/fstab 54 this directory, you can add to your /etc/fstab file::
55 55
56 tracefs /sys/kernel/tracing trace 56 tracefs /sys/kernel/tracing tracefs defaults 0 0
57 57
58 Or you can mount it at run time with:: 58 Or you can mount it at run time with::
59 59
60 mount -t tracefs nodev /sys/kernel/tracing 60 mount -t tracefs nodev /sys/kernel/tracing
61 61
62 For quicker access to that directory you may w 62 For quicker access to that directory you may want to make a soft link to
63 it:: 63 it::
64 64
65 ln -s /sys/kernel/tracing /tracing 65 ln -s /sys/kernel/tracing /tracing
66 66
67 .. attention:: 67 .. attention::
68 68
69 Before 4.1, all ftrace tracing control files 69 Before 4.1, all ftrace tracing control files were within the debugfs
70 file system, which is typically located at / 70 file system, which is typically located at /sys/kernel/debug/tracing.
71 For backward compatibility, when mounting th 71 For backward compatibility, when mounting the debugfs file system,
72 the tracefs file system will be automaticall 72 the tracefs file system will be automatically mounted at:
73 73
74 /sys/kernel/debug/tracing 74 /sys/kernel/debug/tracing
75 75
76 All files located in the tracefs file system 76 All files located in the tracefs file system will be located in that
77 debugfs file system directory as well. 77 debugfs file system directory as well.
78 78
79 .. attention:: 79 .. attention::
80 80
81 Any selected ftrace option will also create 81 Any selected ftrace option will also create the tracefs file system.
82 The rest of the document will assume that yo 82 The rest of the document will assume that you are in the ftrace directory
83 (cd /sys/kernel/tracing) and will only conce 83 (cd /sys/kernel/tracing) and will only concentrate on the files within that
84 directory and not distract from the content 84 directory and not distract from the content with the extended
85 "/sys/kernel/tracing" path name. 85 "/sys/kernel/tracing" path name.
86 86
87 That's it! (assuming that you have ftrace conf 87 That's it! (assuming that you have ftrace configured into your kernel)
88 88
89 After mounting tracefs you will have access to 89 After mounting tracefs you will have access to the control and output files
90 of ftrace. Here is a list of some of the key f 90 of ftrace. Here is a list of some of the key files:
91 91
92 92
93 Note: all time values are in microseconds. 93 Note: all time values are in microseconds.
94 94
95 current_tracer: 95 current_tracer:
96 96
97 This is used to set or display the cur 97 This is used to set or display the current tracer
98 that is configured. Changing the curre 98 that is configured. Changing the current tracer clears
99 the ring buffer content as well as the 99 the ring buffer content as well as the "snapshot" buffer.
100 100
101 available_tracers: 101 available_tracers:
102 102
103 This holds the different types of trac 103 This holds the different types of tracers that
104 have been compiled into the kernel. Th 104 have been compiled into the kernel. The
105 tracers listed here can be configured 105 tracers listed here can be configured by
106 echoing their name into current_tracer 106 echoing their name into current_tracer.
107 107
108 tracing_on: 108 tracing_on:
109 109
110 This sets or displays whether writing 110 This sets or displays whether writing to the trace
111 ring buffer is enabled. Echo 0 into th 111 ring buffer is enabled. Echo 0 into this file to disable
112 the tracer or 1 to enable it. Note, th 112 the tracer or 1 to enable it. Note, this only disables
113 writing to the ring buffer, the tracin 113 writing to the ring buffer, the tracing overhead may
114 still be occurring. 114 still be occurring.
115 115
116 The kernel function tracing_off() can 116 The kernel function tracing_off() can be used within the
117 kernel to disable writing to the ring 117 kernel to disable writing to the ring buffer, which will
118 set this file to "0". User space can r 118 set this file to "0". User space can re-enable tracing by
119 echoing "1" into the file. 119 echoing "1" into the file.
120 120
121 Note, the function and event trigger " 121 Note, the function and event trigger "traceoff" will also
122 set this file to zero and stop tracing 122 set this file to zero and stop tracing. Which can also
123 be re-enabled by user space using this 123 be re-enabled by user space using this file.
124 124
125 trace: 125 trace:
126 126
127 This file holds the output of the trac 127 This file holds the output of the trace in a human
128 readable format (described below). Ope !! 128 readable format (described below). Note, tracing is temporarily
>> 129 disabled when the file is open for reading. Once all readers
>> 130 are closed, tracing is re-enabled. Opening this file for
129 writing with the O_TRUNC flag clears t 131 writing with the O_TRUNC flag clears the ring buffer content.
130 Note, this file is not a consumer. If <<
131 (no tracer running, or tracing_on is z <<
132 the same output each time it is read. <<
133 it may produce inconsistent results as <<
134 the entire buffer without consuming it <<
135 132
136 trace_pipe: 133 trace_pipe:
137 134
138 The output is the same as the "trace" 135 The output is the same as the "trace" file but this
139 file is meant to be streamed with live 136 file is meant to be streamed with live tracing.
140 Reads from this file will block until 137 Reads from this file will block until new data is
141 retrieved. Unlike the "trace" file, t 138 retrieved. Unlike the "trace" file, this file is a
142 consumer. This means reading from this 139 consumer. This means reading from this file causes
143 sequential reads to display more curre 140 sequential reads to display more current data. Once
144 data is read from this file, it is con 141 data is read from this file, it is consumed, and
145 will not be read again with a sequenti 142 will not be read again with a sequential read. The
146 "trace" file is static, and if the tra 143 "trace" file is static, and if the tracer is not
147 adding more data, it will display the 144 adding more data, it will display the same
148 information every time it is read. !! 145 information every time it is read. Unlike the
>> 146 "trace" file, opening this file for reading will not
>> 147 temporarily disable tracing.
149 148
150 trace_options: 149 trace_options:
151 150
152 This file lets the user control the am 151 This file lets the user control the amount of data
153 that is displayed in one of the above 152 that is displayed in one of the above output
154 files. Options also exist to modify ho 153 files. Options also exist to modify how a tracer
155 or events work (stack traces, timestam 154 or events work (stack traces, timestamps, etc).
156 155
157 options: 156 options:
158 157
159 This is a directory that has a file fo 158 This is a directory that has a file for every available
160 trace option (also in trace_options). 159 trace option (also in trace_options). Options may also be set
161 or cleared by writing a "1" or "0" res 160 or cleared by writing a "1" or "0" respectively into the
162 corresponding file with the option nam 161 corresponding file with the option name.
163 162
164 tracing_max_latency: 163 tracing_max_latency:
165 164
166 Some of the tracers record the max lat 165 Some of the tracers record the max latency.
167 For example, the maximum time that int 166 For example, the maximum time that interrupts are disabled.
168 The maximum time is saved in this file 167 The maximum time is saved in this file. The max trace will also be
169 stored, and displayed by "trace". A ne 168 stored, and displayed by "trace". A new max trace will only be
170 recorded if the latency is greater tha 169 recorded if the latency is greater than the value in this file
171 (in microseconds). 170 (in microseconds).
172 171
173 By echoing in a time into this file, n 172 By echoing in a time into this file, no latency will be recorded
174 unless it is greater than the time in 173 unless it is greater than the time in this file.
175 174
176 tracing_thresh: 175 tracing_thresh:
177 176
178 Some latency tracers will record a tra 177 Some latency tracers will record a trace whenever the
179 latency is greater than the number in 178 latency is greater than the number in this file.
180 Only active when the file contains a n 179 Only active when the file contains a number greater than 0.
181 (in microseconds) 180 (in microseconds)
182 181
183 buffer_percent: <<
184 <<
185 This is the watermark for how much the <<
186 before a waiter is woken up. That is, <<
187 blocking read syscall on one of the pe <<
188 will block until the given amount of d <<
189 is in the ring buffer before it wakes <<
190 controls how the splice system calls a <<
191 <<
192 0 - means to wake up as soon as th <<
193 50 - means to wake up when roughly <<
194 are full. <<
195 100 - means to block until the ring <<
196 about to start overwriting the <<
197 <<
198 buffer_size_kb: 182 buffer_size_kb:
199 183
200 This sets or displays the number of ki 184 This sets or displays the number of kilobytes each CPU
201 buffer holds. By default, the trace bu 185 buffer holds. By default, the trace buffers are the same size
202 for each CPU. The displayed number is 186 for each CPU. The displayed number is the size of the
203 CPU buffer and not total size of all b 187 CPU buffer and not total size of all buffers. The
204 trace buffers are allocated in pages ( 188 trace buffers are allocated in pages (blocks of memory
205 that the kernel uses for allocation, u 189 that the kernel uses for allocation, usually 4 KB in size).
206 A few extra pages may be allocated to 190 A few extra pages may be allocated to accommodate buffer management
207 meta-data. If the last page allocated 191 meta-data. If the last page allocated has room for more bytes
208 than requested, the rest of the page w 192 than requested, the rest of the page will be used,
209 making the actual allocation bigger th 193 making the actual allocation bigger than requested or shown.
210 ( Note, the size may not be a multiple 194 ( Note, the size may not be a multiple of the page size
211 due to buffer management meta-data. ) 195 due to buffer management meta-data. )
212 196
213 Buffer sizes for individual CPUs may v 197 Buffer sizes for individual CPUs may vary
214 (see "per_cpu/cpu0/buffer_size_kb" bel 198 (see "per_cpu/cpu0/buffer_size_kb" below), and if they do
215 this file will show "X". 199 this file will show "X".
216 200
217 buffer_total_size_kb: 201 buffer_total_size_kb:
218 202
219 This displays the total combined size 203 This displays the total combined size of all the trace buffers.
220 204
221 buffer_subbuf_size_kb: <<
222 <<
223 This sets or displays the sub buffer s <<
224 into several same size "sub buffers". <<
225 the size of the sub buffer. Normally, <<
226 architecture's page (4K on x86). The s <<
227 at the start which also limits the siz <<
228 the sub buffer is a page size, no even <<
229 size minus the sub buffer meta data. <<
230 <<
231 Note, the buffer_subbuf_size_kb is a w <<
232 minimum size of the subbuffer. The ker <<
233 implementation details, or simply fail <<
234 not handle the request. <<
235 <<
236 Changing the sub buffer size allows fo <<
237 page size. <<
238 <<
239 Note: When changing the sub-buffer siz <<
240 data in the ring buffer and the snapsh <<
241 <<
242 free_buffer: 205 free_buffer:
243 206
244 If a process is performing tracing, an 207 If a process is performing tracing, and the ring buffer should be
245 shrunk "freed" when the process is fin 208 shrunk "freed" when the process is finished, even if it were to be
246 killed by a signal, this file can be u 209 killed by a signal, this file can be used for that purpose. On close
247 of this file, the ring buffer will be 210 of this file, the ring buffer will be resized to its minimum size.
248 Having a process that is tracing also 211 Having a process that is tracing also open this file, when the process
249 exits its file descriptor for this fil 212 exits its file descriptor for this file will be closed, and in doing so,
250 the ring buffer will be "freed". 213 the ring buffer will be "freed".
251 214
252 It may also stop tracing if disable_on 215 It may also stop tracing if disable_on_free option is set.
253 216
254 tracing_cpumask: 217 tracing_cpumask:
255 218
256 This is a mask that lets the user only 219 This is a mask that lets the user only trace on specified CPUs.
257 The format is a hex string representin 220 The format is a hex string representing the CPUs.
258 221
259 set_ftrace_filter: 222 set_ftrace_filter:
260 223
261 When dynamic ftrace is configured in ( 224 When dynamic ftrace is configured in (see the
262 section below "dynamic ftrace"), the c 225 section below "dynamic ftrace"), the code is dynamically
263 modified (code text rewrite) to disabl 226 modified (code text rewrite) to disable calling of the
264 function profiler (mcount). This lets 227 function profiler (mcount). This lets tracing be configured
265 in with practically no overhead in per 228 in with practically no overhead in performance. This also
266 has a side effect of enabling or disab 229 has a side effect of enabling or disabling specific functions
267 to be traced. Echoing names of functio 230 to be traced. Echoing names of functions into this file
268 will limit the trace to only those fun 231 will limit the trace to only those functions.
269 This influences the tracers "function" 232 This influences the tracers "function" and "function_graph"
270 and thus also function profiling (see 233 and thus also function profiling (see "function_profile_enabled").
271 234
272 The functions listed in "available_fil 235 The functions listed in "available_filter_functions" are what
273 can be written into this file. 236 can be written into this file.
274 237
275 This interface also allows for command 238 This interface also allows for commands to be used. See the
276 "Filter commands" section for more det 239 "Filter commands" section for more details.
277 240
278 As a speed up, since processing string 241 As a speed up, since processing strings can be quite expensive
279 and requires a check of all functions 242 and requires a check of all functions registered to tracing, instead
280 an index can be written into this file 243 an index can be written into this file. A number (starting with "1")
281 written will instead select the same c 244 written will instead select the same corresponding at the line position
282 of the "available_filter_functions" fi 245 of the "available_filter_functions" file.
283 246
284 set_ftrace_notrace: 247 set_ftrace_notrace:
285 248
286 This has an effect opposite to that of 249 This has an effect opposite to that of
287 set_ftrace_filter. Any function that i 250 set_ftrace_filter. Any function that is added here will not
288 be traced. If a function exists in bot 251 be traced. If a function exists in both set_ftrace_filter
289 and set_ftrace_notrace, the function w 252 and set_ftrace_notrace, the function will _not_ be traced.
290 253
291 set_ftrace_pid: 254 set_ftrace_pid:
292 255
293 Have the function tracer only trace th 256 Have the function tracer only trace the threads whose PID are
294 listed in this file. 257 listed in this file.
295 258
296 If the "function-fork" option is set, 259 If the "function-fork" option is set, then when a task whose
297 PID is listed in this file forks, the 260 PID is listed in this file forks, the child's PID will
298 automatically be added to this file, a 261 automatically be added to this file, and the child will be
299 traced by the function tracer as well. 262 traced by the function tracer as well. This option will also
300 cause PIDs of tasks that exit to be re 263 cause PIDs of tasks that exit to be removed from the file.
301 264
302 set_ftrace_notrace_pid: <<
303 <<
304 Have the function tracer ignore thread <<
305 this file. <<
306 <<
307 If the "function-fork" option is set, <<
308 PID is listed in this file forks, the <<
309 automatically be added to this file, a <<
310 traced by the function tracer as well. <<
311 cause PIDs of tasks that exit to be re <<
312 <<
313 If a PID is in both this file and "set <<
314 file takes precedence, and the thread <<
315 <<
316 set_event_pid: 265 set_event_pid:
317 266
318 Have the events only trace a task with 267 Have the events only trace a task with a PID listed in this file.
319 Note, sched_switch and sched_wake_up w 268 Note, sched_switch and sched_wake_up will also trace events
320 listed in this file. 269 listed in this file.
321 270
322 To have the PIDs of children of tasks 271 To have the PIDs of children of tasks with their PID in this file
323 added on fork, enable the "event-fork" 272 added on fork, enable the "event-fork" option. That option will also
324 cause the PIDs of tasks to be removed 273 cause the PIDs of tasks to be removed from this file when the task
325 exits. 274 exits.
326 275
327 set_event_notrace_pid: <<
328 <<
329 Have the events not trace a task with <<
330 Note, sched_switch and sched_wakeup wi <<
331 in this file, even if a thread's PID i <<
332 sched_switch or sched_wakeup events al <<
333 be traced. <<
334 <<
335 To have the PIDs of children of tasks <<
336 added on fork, enable the "event-fork" <<
337 cause the PIDs of tasks to be removed <<
338 exits. <<
339 <<
340 set_graph_function: 276 set_graph_function:
341 277
342 Functions listed in this file will cau 278 Functions listed in this file will cause the function graph
343 tracer to only trace these functions a 279 tracer to only trace these functions and the functions that
344 they call. (See the section "dynamic f 280 they call. (See the section "dynamic ftrace" for more details).
345 Note, set_ftrace_filter and set_ftrace 281 Note, set_ftrace_filter and set_ftrace_notrace still affects
346 what functions are being traced. 282 what functions are being traced.
347 283
348 set_graph_notrace: 284 set_graph_notrace:
349 285
350 Similar to set_graph_function, but wil 286 Similar to set_graph_function, but will disable function graph
351 tracing when the function is hit until 287 tracing when the function is hit until it exits the function.
352 This makes it possible to ignore traci 288 This makes it possible to ignore tracing functions that are called
353 by a specific function. 289 by a specific function.
354 290
355 available_filter_functions: 291 available_filter_functions:
356 292
357 This lists the functions that ftrace h 293 This lists the functions that ftrace has processed and can trace.
358 These are the function names that you 294 These are the function names that you can pass to
359 "set_ftrace_filter", "set_ftrace_notra 295 "set_ftrace_filter", "set_ftrace_notrace",
360 "set_graph_function", or "set_graph_no 296 "set_graph_function", or "set_graph_notrace".
361 (See the section "dynamic ftrace" belo 297 (See the section "dynamic ftrace" below for more details.)
362 298
363 available_filter_functions_addrs: <<
364 <<
365 Similar to available_filter_functions, <<
366 for each function. The displayed addre <<
367 and can differ from /proc/kallsyms add <<
368 <<
369 dyn_ftrace_total_info: 299 dyn_ftrace_total_info:
370 300
371 This file is for debugging purposes. T 301 This file is for debugging purposes. The number of functions that
372 have been converted to nops and are av 302 have been converted to nops and are available to be traced.
373 303
374 enabled_functions: 304 enabled_functions:
375 305
376 This file is more for debugging ftrace 306 This file is more for debugging ftrace, but can also be useful
377 in seeing if any function has a callba 307 in seeing if any function has a callback attached to it.
378 Not only does the trace infrastructure 308 Not only does the trace infrastructure use ftrace function
379 trace utility, but other subsystems mi 309 trace utility, but other subsystems might too. This file
380 displays all functions that have a cal 310 displays all functions that have a callback attached to them
381 as well as the number of callbacks tha 311 as well as the number of callbacks that have been attached.
382 Note, a callback may also call multipl 312 Note, a callback may also call multiple functions which will
383 not be listed in this count. 313 not be listed in this count.
384 314
385 If the callback registered to be trace 315 If the callback registered to be traced by a function with
386 the "save regs" attribute (thus even m 316 the "save regs" attribute (thus even more overhead), a 'R'
387 will be displayed on the same line as 317 will be displayed on the same line as the function that
388 is returning registers. 318 is returning registers.
389 319
390 If the callback registered to be trace 320 If the callback registered to be traced by a function with
391 the "ip modify" attribute (thus the re 321 the "ip modify" attribute (thus the regs->ip can be changed),
392 an 'I' will be displayed on the same l 322 an 'I' will be displayed on the same line as the function that
393 can be overridden. 323 can be overridden.
394 324
395 If a non ftrace trampoline is attached <<
396 Note, normal ftrace trampolines can al <<
397 "direct" trampoline can be attached to <<
398 <<
399 Some architectures can not call direct <<
400 the ftrace ops function located above <<
401 such cases an 'O' will be displayed. <<
402 <<
403 If a function had either the "ip modif <<
404 it in the past, a 'M' will be shown. T <<
405 used to know if a function was every m <<
406 and can be used for debugging. <<
407 <<
408 If the architecture supports it, it wi 325 If the architecture supports it, it will also show what callback
409 is being directly called by the functi 326 is being directly called by the function. If the count is greater
410 than 1 it most likely will be ftrace_o 327 than 1 it most likely will be ftrace_ops_list_func().
411 328
412 If the callback of a function jumps to !! 329 If the callback of the function jumps to a trampoline that is
413 specific to the callback and which is !! 330 specific to a the callback and not the standard trampoline,
414 its address will be printed as well as 331 its address will be printed as well as the function that the
415 trampoline calls. 332 trampoline calls.
416 333
417 touched_functions: <<
418 <<
419 This file contains all the functions t <<
420 to it via the ftrace infrastructure. I <<
421 enabled_functions but shows all functi <<
422 traced. <<
423 <<
424 To see any function that has every bee <<
425 direct trampoline, one can perform the <<
426 <<
427 grep ' M ' /sys/kernel/tracing/touched <<
428 <<
429 function_profile_enabled: 334 function_profile_enabled:
430 335
431 When set it will enable all functions 336 When set it will enable all functions with either the function
432 tracer, or if configured, the function 337 tracer, or if configured, the function graph tracer. It will
433 keep a histogram of the number of func 338 keep a histogram of the number of functions that were called
434 and if the function graph tracer was c 339 and if the function graph tracer was configured, it will also keep
435 track of the time spent in those funct 340 track of the time spent in those functions. The histogram
436 content can be displayed in the files: 341 content can be displayed in the files:
437 342
438 trace_stat/function<cpu> ( function0, 343 trace_stat/function<cpu> ( function0, function1, etc).
439 344
440 trace_stat: 345 trace_stat:
441 346
442 A directory that holds different traci 347 A directory that holds different tracing stats.
443 348
444 kprobe_events: 349 kprobe_events:
445 350
446 Enable dynamic trace points. See kprob !! 351 Enable dynamic trace points. See kprobetrace.txt.
447 352
448 kprobe_profile: 353 kprobe_profile:
449 354
450 Dynamic trace points stats. See kprobe !! 355 Dynamic trace points stats. See kprobetrace.txt.
451 356
452 max_graph_depth: 357 max_graph_depth:
453 358
454 Used with the function graph tracer. T 359 Used with the function graph tracer. This is the max depth
455 it will trace into a function. Setting 360 it will trace into a function. Setting this to a value of
456 one will show only the first kernel fu 361 one will show only the first kernel function that is called
457 from user space. 362 from user space.
458 363
459 printk_formats: 364 printk_formats:
460 365
461 This is for tools that read the raw fo 366 This is for tools that read the raw format files. If an event in
462 the ring buffer references a string, o 367 the ring buffer references a string, only a pointer to the string
463 is recorded into the buffer and not th 368 is recorded into the buffer and not the string itself. This prevents
464 tools from knowing what that string wa 369 tools from knowing what that string was. This file displays the string
465 and address for the string allowing to 370 and address for the string allowing tools to map the pointers to what
466 the strings were. 371 the strings were.
467 372
468 saved_cmdlines: 373 saved_cmdlines:
469 374
470 Only the pid of the task is recorded i 375 Only the pid of the task is recorded in a trace event unless
471 the event specifically saves the task 376 the event specifically saves the task comm as well. Ftrace
472 makes a cache of pid mappings to comms 377 makes a cache of pid mappings to comms to try to display
473 comms for events. If a pid for a comm 378 comms for events. If a pid for a comm is not listed, then
474 "<...>" is displayed in the output. 379 "<...>" is displayed in the output.
475 380
476 If the option "record-cmd" is set to " 381 If the option "record-cmd" is set to "0", then comms of tasks
477 will not be saved during recording. By 382 will not be saved during recording. By default, it is enabled.
478 383
479 saved_cmdlines_size: 384 saved_cmdlines_size:
480 385
481 By default, 128 comms are saved (see " 386 By default, 128 comms are saved (see "saved_cmdlines" above). To
482 increase or decrease the amount of com 387 increase or decrease the amount of comms that are cached, echo
483 the number of comms to cache into this 388 the number of comms to cache into this file.
484 389
485 saved_tgids: 390 saved_tgids:
486 391
487 If the option "record-tgid" is set, on 392 If the option "record-tgid" is set, on each scheduling context switch
488 the Task Group ID of a task is saved i 393 the Task Group ID of a task is saved in a table mapping the PID of
489 the thread to its TGID. By default, th 394 the thread to its TGID. By default, the "record-tgid" option is
490 disabled. 395 disabled.
491 396
492 snapshot: 397 snapshot:
493 398
494 This displays the "snapshot" buffer an 399 This displays the "snapshot" buffer and also lets the user
495 take a snapshot of the current running 400 take a snapshot of the current running trace.
496 See the "Snapshot" section below for m 401 See the "Snapshot" section below for more details.
497 402
498 stack_max_size: 403 stack_max_size:
499 404
500 When the stack tracer is activated, th 405 When the stack tracer is activated, this will display the
501 maximum stack size it has encountered. 406 maximum stack size it has encountered.
502 See the "Stack Trace" section below. 407 See the "Stack Trace" section below.
503 408
504 stack_trace: 409 stack_trace:
505 410
506 This displays the stack back trace of 411 This displays the stack back trace of the largest stack
507 that was encountered when the stack tr 412 that was encountered when the stack tracer is activated.
508 See the "Stack Trace" section below. 413 See the "Stack Trace" section below.
509 414
510 stack_trace_filter: 415 stack_trace_filter:
511 416
512 This is similar to "set_ftrace_filter" 417 This is similar to "set_ftrace_filter" but it limits what
513 functions the stack tracer will check. 418 functions the stack tracer will check.
514 419
515 trace_clock: 420 trace_clock:
516 421
517 Whenever an event is recorded into the 422 Whenever an event is recorded into the ring buffer, a
518 "timestamp" is added. This stamp comes 423 "timestamp" is added. This stamp comes from a specified
519 clock. By default, ftrace uses the "lo 424 clock. By default, ftrace uses the "local" clock. This
520 clock is very fast and strictly per cp 425 clock is very fast and strictly per cpu, but on some
521 systems it may not be monotonic with r 426 systems it may not be monotonic with respect to other
522 CPUs. In other words, the local clocks 427 CPUs. In other words, the local clocks may not be in sync
523 with local clocks on other CPUs. 428 with local clocks on other CPUs.
524 429
525 Usual clocks for tracing:: 430 Usual clocks for tracing::
526 431
527 # cat trace_clock 432 # cat trace_clock
528 [local] global counter x86-tsc 433 [local] global counter x86-tsc
529 434
530 The clock with the square brackets aro 435 The clock with the square brackets around it is the one in effect.
531 436
532 local: 437 local:
533 Default clock, but may not be 438 Default clock, but may not be in sync across CPUs
534 439
535 global: 440 global:
536 This clock is in sync with all 441 This clock is in sync with all CPUs but may
537 be a bit slower than the local 442 be a bit slower than the local clock.
538 443
539 counter: 444 counter:
540 This is not a clock at all, bu 445 This is not a clock at all, but literally an atomic
541 counter. It counts up one by o 446 counter. It counts up one by one, but is in sync
542 with all CPUs. This is useful 447 with all CPUs. This is useful when you need to
543 know exactly the order events 448 know exactly the order events occurred with respect to
544 each other on different CPUs. 449 each other on different CPUs.
545 450
546 uptime: 451 uptime:
547 This uses the jiffies counter 452 This uses the jiffies counter and the time stamp
548 is relative to the time since 453 is relative to the time since boot up.
549 454
550 perf: 455 perf:
551 This makes ftrace use the same 456 This makes ftrace use the same clock that perf uses.
552 Eventually perf will be able t 457 Eventually perf will be able to read ftrace buffers
553 and this will help out in inte 458 and this will help out in interleaving the data.
554 459
555 x86-tsc: 460 x86-tsc:
556 Architectures may define their 461 Architectures may define their own clocks. For
557 example, x86 uses its own TSC 462 example, x86 uses its own TSC cycle clock here.
558 463
559 ppc-tb: 464 ppc-tb:
560 This uses the powerpc timebase 465 This uses the powerpc timebase register value.
561 This is in sync across CPUs an 466 This is in sync across CPUs and can also be used
562 to correlate events across hyp 467 to correlate events across hypervisor/guest if
563 tb_offset is known. 468 tb_offset is known.
564 469
565 mono: 470 mono:
566 This uses the fast monotonic c 471 This uses the fast monotonic clock (CLOCK_MONOTONIC)
567 which is monotonic and is subj 472 which is monotonic and is subject to NTP rate adjustments.
568 473
569 mono_raw: 474 mono_raw:
570 This is the raw monotonic cloc 475 This is the raw monotonic clock (CLOCK_MONOTONIC_RAW)
571 which is monotonic but is not 476 which is monotonic but is not subject to any rate adjustments
572 and ticks at the same rate as 477 and ticks at the same rate as the hardware clocksource.
573 478
574 boot: 479 boot:
575 This is the boot clock (CLOCK_ 480 This is the boot clock (CLOCK_BOOTTIME) and is based on the
576 fast monotonic clock, but also 481 fast monotonic clock, but also accounts for time spent in
577 suspend. Since the clock acces 482 suspend. Since the clock access is designed for use in
578 tracing in the suspend path, s 483 tracing in the suspend path, some side effects are possible
579 if clock is accessed after the 484 if clock is accessed after the suspend time is accounted before
580 the fast mono clock is updated 485 the fast mono clock is updated. In this case, the clock update
581 appears to happen slightly soo 486 appears to happen slightly sooner than it normally would have.
582 Also on 32-bit systems, it's p 487 Also on 32-bit systems, it's possible that the 64-bit boot offset
583 sees a partial update. These e 488 sees a partial update. These effects are rare and post
584 processing should be able to h 489 processing should be able to handle them. See comments in the
585 ktime_get_boot_fast_ns() funct 490 ktime_get_boot_fast_ns() function for more information.
586 491
587 tai: <<
588 This is the tai clock (CLOCK_T <<
589 clock time. However, this cloc <<
590 discontinuities and backwards <<
591 seconds. Since the clock acces <<
592 side effects are possible. The <<
593 readouts in case the internal <<
594 by setting the system time or <<
595 These effects are rare and pos <<
596 handle them. See comments in t <<
597 function for more information. <<
598 <<
599 To set a clock, simply echo the clock 492 To set a clock, simply echo the clock name into this file::
600 493
601 # echo global > trace_clock 494 # echo global > trace_clock
602 495
603 Setting a clock clears the ring buffer 496 Setting a clock clears the ring buffer content as well as the
604 "snapshot" buffer. 497 "snapshot" buffer.
605 498
606 trace_marker: 499 trace_marker:
607 500
608 This is a very useful file for synchro 501 This is a very useful file for synchronizing user space
609 with events happening in the kernel. W 502 with events happening in the kernel. Writing strings into
610 this file will be written into the ftr 503 this file will be written into the ftrace buffer.
611 504
612 It is useful in applications to open t 505 It is useful in applications to open this file at the start
613 of the application and just reference 506 of the application and just reference the file descriptor
614 for the file:: 507 for the file::
615 508
616 void trace_write(const char *f 509 void trace_write(const char *fmt, ...)
617 { 510 {
618 va_list ap; 511 va_list ap;
619 char buf[256]; 512 char buf[256];
620 int n; 513 int n;
621 514
622 if (trace_fd < 0) 515 if (trace_fd < 0)
623 return; 516 return;
624 517
625 va_start(ap, fmt); 518 va_start(ap, fmt);
626 n = vsnprintf(buf, 256 519 n = vsnprintf(buf, 256, fmt, ap);
627 va_end(ap); 520 va_end(ap);
628 521
629 write(trace_fd, buf, n 522 write(trace_fd, buf, n);
630 } 523 }
631 524
632 start:: 525 start::
633 526
634 trace_fd = open("trace_marker" !! 527 trace_fd = open("trace_marker", WR_ONLY);
635 528
636 Note: Writing into the trace_marker fi 529 Note: Writing into the trace_marker file can also initiate triggers
637 that are written into /sys/kerne 530 that are written into /sys/kernel/tracing/events/ftrace/print/trigger
638 See "Event triggers" in Document 531 See "Event triggers" in Documentation/trace/events.rst and an
639 example in Documentation/trace/h 532 example in Documentation/trace/histogram.rst (Section 3.)
640 533
641 trace_marker_raw: 534 trace_marker_raw:
642 535
643 This is similar to trace_marker above, !! 536 This is similar to trace_marker above, but is meant for for binary data
644 to be written to it, where a tool can 537 to be written to it, where a tool can be used to parse the data
645 from trace_pipe_raw. 538 from trace_pipe_raw.
646 539
647 uprobe_events: 540 uprobe_events:
648 541
649 Add dynamic tracepoints in programs. 542 Add dynamic tracepoints in programs.
650 See uprobetracer.rst !! 543 See uprobetracer.txt
651 544
652 uprobe_profile: 545 uprobe_profile:
653 546
654 Uprobe statistics. See uprobetrace.txt 547 Uprobe statistics. See uprobetrace.txt
655 548
656 instances: 549 instances:
657 550
658 This is a way to make multiple trace b 551 This is a way to make multiple trace buffers where different
659 events can be recorded in different bu 552 events can be recorded in different buffers.
660 See "Instances" section below. 553 See "Instances" section below.
661 554
662 events: 555 events:
663 556
664 This is the trace event directory. It 557 This is the trace event directory. It holds event tracepoints
665 (also known as static tracepoints) tha 558 (also known as static tracepoints) that have been compiled
666 into the kernel. It shows what event t 559 into the kernel. It shows what event tracepoints exist
667 and how they are grouped by system. Th 560 and how they are grouped by system. There are "enable"
668 files at various levels that can enabl 561 files at various levels that can enable the tracepoints
669 when a "1" is written to them. 562 when a "1" is written to them.
670 563
671 See events.rst for more information. !! 564 See events.txt for more information.
672 565
673 set_event: 566 set_event:
674 567
675 By echoing in the event into this file 568 By echoing in the event into this file, will enable that event.
676 569
677 See events.rst for more information. !! 570 See events.txt for more information.
678 571
679 available_events: 572 available_events:
680 573
681 A list of events that can be enabled i 574 A list of events that can be enabled in tracing.
682 575
683 See events.rst for more information. !! 576 See events.txt for more information.
684 577
685 timestamp_mode: 578 timestamp_mode:
686 579
687 Certain tracers may change the timesta 580 Certain tracers may change the timestamp mode used when
688 logging trace events into the event bu 581 logging trace events into the event buffer. Events with
689 different modes can coexist within a b 582 different modes can coexist within a buffer but the mode in
690 effect when an event is logged determi 583 effect when an event is logged determines which timestamp mode
691 is used for that event. The default t 584 is used for that event. The default timestamp mode is
692 'delta'. 585 'delta'.
693 586
694 Usual timestamp modes for tracing: 587 Usual timestamp modes for tracing:
695 588
696 # cat timestamp_mode 589 # cat timestamp_mode
697 [delta] absolute 590 [delta] absolute
698 591
699 The timestamp mode with the square b 592 The timestamp mode with the square brackets around it is the
700 one in effect. 593 one in effect.
701 594
702 delta: Default timestamp mode - time 595 delta: Default timestamp mode - timestamp is a delta against
703 a per-buffer timestamp. 596 a per-buffer timestamp.
704 597
705 absolute: The timestamp is a full ti 598 absolute: The timestamp is a full timestamp, not a delta
706 against some other value. As 599 against some other value. As such it takes up more
707 space and is less efficient. 600 space and is less efficient.
708 601
709 hwlat_detector: 602 hwlat_detector:
710 603
711 Directory for the Hardware Latency Det 604 Directory for the Hardware Latency Detector.
712 See "Hardware Latency Detector" sectio 605 See "Hardware Latency Detector" section below.
713 606
714 per_cpu: 607 per_cpu:
715 608
716 This is a directory that contains the 609 This is a directory that contains the trace per_cpu information.
717 610
718 per_cpu/cpu0/buffer_size_kb: 611 per_cpu/cpu0/buffer_size_kb:
719 612
720 The ftrace buffer is defined per_cpu. 613 The ftrace buffer is defined per_cpu. That is, there's a separate
721 buffer for each CPU to allow writes to 614 buffer for each CPU to allow writes to be done atomically,
722 and free from cache bouncing. These bu 615 and free from cache bouncing. These buffers may have different
723 size buffers. This file is similar to 616 size buffers. This file is similar to the buffer_size_kb
724 file, but it only displays or sets the 617 file, but it only displays or sets the buffer size for the
725 specific CPU. (here cpu0). 618 specific CPU. (here cpu0).
726 619
727 per_cpu/cpu0/trace: 620 per_cpu/cpu0/trace:
728 621
729 This is similar to the "trace" file, b 622 This is similar to the "trace" file, but it will only display
730 the data specific for the CPU. If writ 623 the data specific for the CPU. If written to, it only clears
731 the specific CPU buffer. 624 the specific CPU buffer.
732 625
733 per_cpu/cpu0/trace_pipe 626 per_cpu/cpu0/trace_pipe
734 627
735 This is similar to the "trace_pipe" fi 628 This is similar to the "trace_pipe" file, and is a consuming
736 read, but it will only display (and co 629 read, but it will only display (and consume) the data specific
737 for the CPU. 630 for the CPU.
738 631
739 per_cpu/cpu0/trace_pipe_raw 632 per_cpu/cpu0/trace_pipe_raw
740 633
741 For tools that can parse the ftrace ri 634 For tools that can parse the ftrace ring buffer binary format,
742 the trace_pipe_raw file can be used to 635 the trace_pipe_raw file can be used to extract the data
743 from the ring buffer directly. With th 636 from the ring buffer directly. With the use of the splice()
744 system call, the buffer data can be qu 637 system call, the buffer data can be quickly transferred to
745 a file or to the network where a serve 638 a file or to the network where a server is collecting the
746 data. 639 data.
747 640
748 Like trace_pipe, this is a consuming r 641 Like trace_pipe, this is a consuming reader, where multiple
749 reads will always produce different da 642 reads will always produce different data.
750 643
751 per_cpu/cpu0/snapshot: 644 per_cpu/cpu0/snapshot:
752 645
753 This is similar to the main "snapshot" 646 This is similar to the main "snapshot" file, but will only
754 snapshot the current CPU (if supported 647 snapshot the current CPU (if supported). It only displays
755 the content of the snapshot for a give 648 the content of the snapshot for a given CPU, and if
756 written to, only clears this CPU buffe 649 written to, only clears this CPU buffer.
757 650
758 per_cpu/cpu0/snapshot_raw: 651 per_cpu/cpu0/snapshot_raw:
759 652
760 Similar to the trace_pipe_raw, but wil 653 Similar to the trace_pipe_raw, but will read the binary format
761 from the snapshot buffer for the given 654 from the snapshot buffer for the given CPU.
762 655
763 per_cpu/cpu0/stats: 656 per_cpu/cpu0/stats:
764 657
765 This displays certain stats about the 658 This displays certain stats about the ring buffer:
766 659
767 entries: 660 entries:
768 The number of events that are 661 The number of events that are still in the buffer.
769 662
770 overrun: 663 overrun:
771 The number of lost events due 664 The number of lost events due to overwriting when
772 the buffer was full. 665 the buffer was full.
773 666
774 commit overrun: 667 commit overrun:
775 Should always be zero. 668 Should always be zero.
776 This gets set if so many event 669 This gets set if so many events happened within a nested
777 event (ring buffer is re-entra 670 event (ring buffer is re-entrant), that it fills the
778 buffer and starts dropping eve 671 buffer and starts dropping events.
779 672
780 bytes: 673 bytes:
781 Bytes actually read (not overw 674 Bytes actually read (not overwritten).
782 675
783 oldest event ts: 676 oldest event ts:
784 The oldest timestamp in the bu 677 The oldest timestamp in the buffer
785 678
786 now ts: 679 now ts:
787 The current timestamp 680 The current timestamp
788 681
789 dropped events: 682 dropped events:
790 Events lost due to overwrite o 683 Events lost due to overwrite option being off.
791 684
792 read events: 685 read events:
793 The number of events read. 686 The number of events read.
794 687
795 The Tracers 688 The Tracers
796 ----------- 689 -----------
797 690
798 Here is the list of current tracers that may b 691 Here is the list of current tracers that may be configured.
799 692
800 "function" 693 "function"
801 694
802 Function call tracer to trace all kern 695 Function call tracer to trace all kernel functions.
803 696
804 "function_graph" 697 "function_graph"
805 698
806 Similar to the function tracer except 699 Similar to the function tracer except that the
807 function tracer probes the functions o 700 function tracer probes the functions on their entry
808 whereas the function graph tracer trac 701 whereas the function graph tracer traces on both entry
809 and exit of the functions. It then pro 702 and exit of the functions. It then provides the ability
810 to draw a graph of function calls simi 703 to draw a graph of function calls similar to C code
811 source. 704 source.
812 705
813 "blk" 706 "blk"
814 707
815 The block tracer. The tracer used by t 708 The block tracer. The tracer used by the blktrace user
816 application. 709 application.
817 710
818 "hwlat" 711 "hwlat"
819 712
820 The Hardware Latency tracer is used to 713 The Hardware Latency tracer is used to detect if the hardware
821 produces any latency. See "Hardware La 714 produces any latency. See "Hardware Latency Detector" section
822 below. 715 below.
823 716
824 "irqsoff" 717 "irqsoff"
825 718
826 Traces the areas that disable interrup 719 Traces the areas that disable interrupts and saves
827 the trace with the longest max latency 720 the trace with the longest max latency.
828 See tracing_max_latency. When a new ma 721 See tracing_max_latency. When a new max is recorded,
829 it replaces the old trace. It is best 722 it replaces the old trace. It is best to view this
830 trace with the latency-format option e 723 trace with the latency-format option enabled, which
831 happens automatically when the tracer 724 happens automatically when the tracer is selected.
832 725
833 "preemptoff" 726 "preemptoff"
834 727
835 Similar to irqsoff but traces and reco 728 Similar to irqsoff but traces and records the amount of
836 time for which preemption is disabled. 729 time for which preemption is disabled.
837 730
838 "preemptirqsoff" 731 "preemptirqsoff"
839 732
840 Similar to irqsoff and preemptoff, but 733 Similar to irqsoff and preemptoff, but traces and
841 records the largest time for which irq 734 records the largest time for which irqs and/or preemption
842 is disabled. 735 is disabled.
843 736
844 "wakeup" 737 "wakeup"
845 738
846 Traces and records the max latency tha 739 Traces and records the max latency that it takes for
847 the highest priority task to get sched 740 the highest priority task to get scheduled after
848 it has been woken up. 741 it has been woken up.
849 Traces all tasks as an average develop 742 Traces all tasks as an average developer would expect.
850 743
851 "wakeup_rt" 744 "wakeup_rt"
852 745
853 Traces and records the max latency tha 746 Traces and records the max latency that it takes for just
854 RT tasks (as the current "wakeup" does 747 RT tasks (as the current "wakeup" does). This is useful
855 for those interested in wake up timing 748 for those interested in wake up timings of RT tasks.
856 749
857 "wakeup_dl" 750 "wakeup_dl"
858 751
859 Traces and records the max latency tha 752 Traces and records the max latency that it takes for
860 a SCHED_DEADLINE task to be woken (as 753 a SCHED_DEADLINE task to be woken (as the "wakeup" and
861 "wakeup_rt" does). 754 "wakeup_rt" does).
862 755
863 "mmiotrace" 756 "mmiotrace"
864 757
865 A special tracer that is used to trace 758 A special tracer that is used to trace binary module.
866 It will trace all the calls that a mod 759 It will trace all the calls that a module makes to the
867 hardware. Everything it writes and rea 760 hardware. Everything it writes and reads from the I/O
868 as well. 761 as well.
869 762
870 "branch" 763 "branch"
871 764
872 This tracer can be configured when tra 765 This tracer can be configured when tracing likely/unlikely
873 calls within the kernel. It will trace 766 calls within the kernel. It will trace when a likely and
874 unlikely branch is hit and if it was c 767 unlikely branch is hit and if it was correct in its prediction
875 of being correct. 768 of being correct.
876 769
877 "nop" 770 "nop"
878 771
879 This is the "trace nothing" tracer. To 772 This is the "trace nothing" tracer. To remove all
880 tracers from tracing simply echo "nop" 773 tracers from tracing simply echo "nop" into
881 current_tracer. 774 current_tracer.
882 775
883 Error conditions 776 Error conditions
884 ---------------- 777 ----------------
885 778
886 For most ftrace commands, failure modes are 779 For most ftrace commands, failure modes are obvious and communicated
887 using standard return codes. 780 using standard return codes.
888 781
889 For other more involved commands, extended e 782 For other more involved commands, extended error information may be
890 available via the tracing/error_log file. F 783 available via the tracing/error_log file. For the commands that
891 support it, reading the tracing/error_log fi 784 support it, reading the tracing/error_log file after an error will
892 display more detailed information about what 785 display more detailed information about what went wrong, if
893 information is available. The tracing/error 786 information is available. The tracing/error_log file is a circular
894 error log displaying a small number (current 787 error log displaying a small number (currently, 8) of ftrace errors
895 for the last (8) failed commands. 788 for the last (8) failed commands.
896 789
897 The extended error information and usage tak 790 The extended error information and usage takes the form shown in
898 this example:: 791 this example::
899 792
900 # echo xxx > /sys/kernel/tracing/events/sc !! 793 # echo xxx > /sys/kernel/debug/tracing/events/sched/sched_wakeup/trigger
901 echo: write error: Invalid argument 794 echo: write error: Invalid argument
902 795
903 # cat /sys/kernel/tracing/error_log !! 796 # cat /sys/kernel/debug/tracing/error_log
904 [ 5348.887237] location: error: Couldn't y 797 [ 5348.887237] location: error: Couldn't yyy: zzz
905 Command: xxx 798 Command: xxx
906 ^ 799 ^
907 [ 7517.023364] location: error: Bad rrr: s 800 [ 7517.023364] location: error: Bad rrr: sss
908 Command: ppp qqq 801 Command: ppp qqq
909 ^ 802 ^
910 803
911 To clear the error log, echo the empty strin 804 To clear the error log, echo the empty string into it::
912 805
913 # echo > /sys/kernel/tracing/error_log !! 806 # echo > /sys/kernel/debug/tracing/error_log
914 807
915 Examples of using the tracer 808 Examples of using the tracer
916 ---------------------------- 809 ----------------------------
917 810
918 Here are typical examples of using the tracers 811 Here are typical examples of using the tracers when controlling
919 them only with the tracefs interface (without 812 them only with the tracefs interface (without using any
920 user-land utilities). 813 user-land utilities).
921 814
922 Output format: 815 Output format:
923 -------------- 816 --------------
924 817
925 Here is an example of the output format of the 818 Here is an example of the output format of the file "trace"::
926 819
927 # tracer: function 820 # tracer: function
928 # 821 #
929 # entries-in-buffer/entries-written: 140080/ 822 # entries-in-buffer/entries-written: 140080/250280 #P:4
930 # 823 #
931 # _-----=> irqs 824 # _-----=> irqs-off
932 # / _----=> need 825 # / _----=> need-resched
933 # | / _---=> hard 826 # | / _---=> hardirq/softirq
934 # || / _--=> pree 827 # || / _--=> preempt-depth
935 # ||| / delay 828 # ||| / delay
936 # TASK-PID CPU# |||| TIMESTA 829 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
937 # | | | |||| | 830 # | | | |||| | |
938 bash-1977 [000] .... 17284.9936 831 bash-1977 [000] .... 17284.993652: sys_close <-system_call_fastpath
939 bash-1977 [000] .... 17284.9936 832 bash-1977 [000] .... 17284.993653: __close_fd <-sys_close
940 bash-1977 [000] .... 17284.9936 833 bash-1977 [000] .... 17284.993653: _raw_spin_lock <-__close_fd
941 sshd-1974 [003] .... 17284.9936 834 sshd-1974 [003] .... 17284.993653: __srcu_read_unlock <-fsnotify
942 bash-1977 [000] .... 17284.9936 835 bash-1977 [000] .... 17284.993654: add_preempt_count <-_raw_spin_lock
943 bash-1977 [000] ...1 17284.9936 836 bash-1977 [000] ...1 17284.993655: _raw_spin_unlock <-__close_fd
944 bash-1977 [000] ...1 17284.9936 837 bash-1977 [000] ...1 17284.993656: sub_preempt_count <-_raw_spin_unlock
945 bash-1977 [000] .... 17284.9936 838 bash-1977 [000] .... 17284.993657: filp_close <-__close_fd
946 bash-1977 [000] .... 17284.9936 839 bash-1977 [000] .... 17284.993657: dnotify_flush <-filp_close
947 sshd-1974 [003] .... 17284.9936 840 sshd-1974 [003] .... 17284.993658: sys_select <-system_call_fastpath
948 .... 841 ....
949 842
950 A header is printed with the tracer name that 843 A header is printed with the tracer name that is represented by
951 the trace. In this case the tracer is "functio 844 the trace. In this case the tracer is "function". Then it shows the
952 number of events in the buffer as well as the 845 number of events in the buffer as well as the total number of entries
953 that were written. The difference is the numbe 846 that were written. The difference is the number of entries that were
954 lost due to the buffer filling up (250280 - 14 847 lost due to the buffer filling up (250280 - 140080 = 110200 events
955 lost). 848 lost).
956 849
957 The header explains the content of the events. 850 The header explains the content of the events. Task name "bash", the task
958 PID "1977", the CPU that it was running on "00 851 PID "1977", the CPU that it was running on "000", the latency format
959 (explained below), the timestamp in <secs>.<us 852 (explained below), the timestamp in <secs>.<usecs> format, the
960 function name that was traced "sys_close" and 853 function name that was traced "sys_close" and the parent function that
961 called this function "system_call_fastpath". T 854 called this function "system_call_fastpath". The timestamp is the time
962 at which the function was entered. 855 at which the function was entered.
963 856
964 Latency trace format 857 Latency trace format
965 -------------------- 858 --------------------
966 859
967 When the latency-format option is enabled or w 860 When the latency-format option is enabled or when one of the latency
968 tracers is set, the trace file gives somewhat 861 tracers is set, the trace file gives somewhat more information to see
969 why a latency happened. Here is a typical trac 862 why a latency happened. Here is a typical trace::
970 863
971 # tracer: irqsoff 864 # tracer: irqsoff
972 # 865 #
973 # irqsoff latency trace v1.1.5 on 3.8.0-test 866 # irqsoff latency trace v1.1.5 on 3.8.0-test+
974 # ------------------------------------------ 867 # --------------------------------------------------------------------
975 # latency: 259 us, #4/4, CPU#2 | (M:preempt 868 # latency: 259 us, #4/4, CPU#2 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
976 # ----------------- 869 # -----------------
977 # | task: ps-6143 (uid:0 nice:0 policy:0 870 # | task: ps-6143 (uid:0 nice:0 policy:0 rt_prio:0)
978 # ----------------- 871 # -----------------
979 # => started at: __lock_task_sighand 872 # => started at: __lock_task_sighand
980 # => ended at: _raw_spin_unlock_irqrestor 873 # => ended at: _raw_spin_unlock_irqrestore
981 # 874 #
982 # 875 #
983 # _------=> CPU# 876 # _------=> CPU#
984 # / _-----=> irqs-off 877 # / _-----=> irqs-off
985 # | / _----=> need-resched 878 # | / _----=> need-resched
986 # || / _---=> hardirq/softirq 879 # || / _---=> hardirq/softirq
987 # ||| / _--=> preempt-depth 880 # ||| / _--=> preempt-depth
988 # |||| / delay 881 # |||| / delay
989 # cmd pid ||||| time | caller 882 # cmd pid ||||| time | caller
990 # \ / ||||| \ | / 883 # \ / ||||| \ | /
991 ps-6143 2d... 0us!: trace_hardir 884 ps-6143 2d... 0us!: trace_hardirqs_off <-__lock_task_sighand
992 ps-6143 2d..1 259us+: trace_hardir 885 ps-6143 2d..1 259us+: trace_hardirqs_on <-_raw_spin_unlock_irqrestore
993 ps-6143 2d..1 263us+: time_hardirq 886 ps-6143 2d..1 263us+: time_hardirqs_on <-_raw_spin_unlock_irqrestore
994 ps-6143 2d..1 306us : <stack trace 887 ps-6143 2d..1 306us : <stack trace>
995 => trace_hardirqs_on_caller 888 => trace_hardirqs_on_caller
996 => trace_hardirqs_on 889 => trace_hardirqs_on
997 => _raw_spin_unlock_irqrestore 890 => _raw_spin_unlock_irqrestore
998 => do_task_stat 891 => do_task_stat
999 => proc_tgid_stat 892 => proc_tgid_stat
1000 => proc_single_show 893 => proc_single_show
1001 => seq_read 894 => seq_read
1002 => vfs_read 895 => vfs_read
1003 => sys_read 896 => sys_read
1004 => system_call_fastpath 897 => system_call_fastpath
1005 898
1006 899
1007 This shows that the current tracer is "irqsof 900 This shows that the current tracer is "irqsoff" tracing the time
1008 for which interrupts were disabled. It gives 901 for which interrupts were disabled. It gives the trace version (which
1009 never changes) and the version of the kernel 902 never changes) and the version of the kernel upon which this was executed on
1010 (3.8). Then it displays the max latency in mi 903 (3.8). Then it displays the max latency in microseconds (259 us). The number
1011 of trace entries displayed and the total numb 904 of trace entries displayed and the total number (both are four: #4/4).
1012 VP, KP, SP, and HP are always zero and are re 905 VP, KP, SP, and HP are always zero and are reserved for later use.
1013 #P is the number of online CPUs (#P:4). 906 #P is the number of online CPUs (#P:4).
1014 907
1015 The task is the process that was running when 908 The task is the process that was running when the latency
1016 occurred. (ps pid: 6143). 909 occurred. (ps pid: 6143).
1017 910
1018 The start and stop (the functions in which th 911 The start and stop (the functions in which the interrupts were
1019 disabled and enabled respectively) that cause 912 disabled and enabled respectively) that caused the latencies:
1020 913
1021 - __lock_task_sighand is where the interrup 914 - __lock_task_sighand is where the interrupts were disabled.
1022 - _raw_spin_unlock_irqrestore is where they 915 - _raw_spin_unlock_irqrestore is where they were enabled again.
1023 916
1024 The next lines after the header are the trace 917 The next lines after the header are the trace itself. The header
1025 explains which is which. 918 explains which is which.
1026 919
1027 cmd: The name of the process in the trace. 920 cmd: The name of the process in the trace.
1028 921
1029 pid: The PID of that process. 922 pid: The PID of that process.
1030 923
1031 CPU#: The CPU which the process was running 924 CPU#: The CPU which the process was running on.
1032 925
1033 irqs-off: 'd' interrupts are disabled. '.' 926 irqs-off: 'd' interrupts are disabled. '.' otherwise.
1034 .. caution:: If the architecture does 927 .. caution:: If the architecture does not support a way to
1035 read the irq flags variable, 928 read the irq flags variable, an 'X' will always
1036 be printed here. 929 be printed here.
1037 930
1038 need-resched: 931 need-resched:
1039 - 'N' both TIF_NEED_RESCHED and PREEM 932 - 'N' both TIF_NEED_RESCHED and PREEMPT_NEED_RESCHED is set,
1040 - 'n' only TIF_NEED_RESCHED is set, 933 - 'n' only TIF_NEED_RESCHED is set,
1041 - 'p' only PREEMPT_NEED_RESCHED is se 934 - 'p' only PREEMPT_NEED_RESCHED is set,
1042 - '.' otherwise. 935 - '.' otherwise.
1043 936
1044 hardirq/softirq: 937 hardirq/softirq:
1045 - 'Z' - NMI occurred inside a hardirq 938 - 'Z' - NMI occurred inside a hardirq
1046 - 'z' - NMI is running 939 - 'z' - NMI is running
1047 - 'H' - hard irq occurred inside a so 940 - 'H' - hard irq occurred inside a softirq.
1048 - 'h' - hard irq is running 941 - 'h' - hard irq is running
1049 - 's' - soft irq is running 942 - 's' - soft irq is running
1050 - '.' - normal context. 943 - '.' - normal context.
1051 944
1052 preempt-depth: The level of preempt_disable 945 preempt-depth: The level of preempt_disabled
1053 946
1054 The above is mostly meaningful for kernel dev 947 The above is mostly meaningful for kernel developers.
1055 948
1056 time: 949 time:
1057 When the latency-format option is ena 950 When the latency-format option is enabled, the trace file
1058 output includes a timestamp relative 951 output includes a timestamp relative to the start of the
1059 trace. This differs from the output w 952 trace. This differs from the output when latency-format
1060 is disabled, which includes an absolu 953 is disabled, which includes an absolute timestamp.
1061 954
1062 delay: 955 delay:
1063 This is just to help catch your eye a 956 This is just to help catch your eye a bit better. And
1064 needs to be fixed to be only relative 957 needs to be fixed to be only relative to the same CPU.
1065 The marks are determined by the diffe 958 The marks are determined by the difference between this
1066 current trace and the next trace. 959 current trace and the next trace.
1067 960
1068 - '$' - greater than 1 second 961 - '$' - greater than 1 second
1069 - '@' - greater than 100 millisecon 962 - '@' - greater than 100 millisecond
1070 - '*' - greater than 10 millisecond 963 - '*' - greater than 10 millisecond
1071 - '#' - greater than 1000 microseco 964 - '#' - greater than 1000 microsecond
1072 - '!' - greater than 100 microsecon 965 - '!' - greater than 100 microsecond
1073 - '+' - greater than 10 microsecond 966 - '+' - greater than 10 microsecond
1074 - ' ' - less than or equal to 10 mi 967 - ' ' - less than or equal to 10 microsecond.
1075 968
1076 The rest is the same as the 'trace' file. 969 The rest is the same as the 'trace' file.
1077 970
1078 Note, the latency tracers will usually end 971 Note, the latency tracers will usually end with a back trace
1079 to easily find where the latency occurred. 972 to easily find where the latency occurred.
1080 973
1081 trace_options 974 trace_options
1082 ------------- 975 -------------
1083 976
1084 The trace_options file (or the options direct 977 The trace_options file (or the options directory) is used to control
1085 what gets printed in the trace output, or man 978 what gets printed in the trace output, or manipulate the tracers.
1086 To see what is available, simply cat the file 979 To see what is available, simply cat the file::
1087 980
1088 cat trace_options 981 cat trace_options
1089 print-parent 982 print-parent
1090 nosym-offset 983 nosym-offset
1091 nosym-addr 984 nosym-addr
1092 noverbose 985 noverbose
1093 noraw 986 noraw
1094 nohex 987 nohex
1095 nobin 988 nobin
1096 noblock 989 noblock
1097 nofields <<
1098 trace_printk 990 trace_printk
1099 annotate 991 annotate
1100 nouserstacktrace 992 nouserstacktrace
1101 nosym-userobj 993 nosym-userobj
1102 noprintk-msg-only 994 noprintk-msg-only
1103 context-info 995 context-info
1104 nolatency-format 996 nolatency-format
1105 record-cmd 997 record-cmd
1106 norecord-tgid 998 norecord-tgid
1107 overwrite 999 overwrite
1108 nodisable_on_free 1000 nodisable_on_free
1109 irq-info 1001 irq-info
1110 markers 1002 markers
1111 noevent-fork 1003 noevent-fork
1112 function-trace 1004 function-trace
1113 nofunction-fork 1005 nofunction-fork
1114 nodisplay-graph 1006 nodisplay-graph
1115 nostacktrace 1007 nostacktrace
1116 nobranch 1008 nobranch
1117 1009
1118 To disable one of the options, echo in the op 1010 To disable one of the options, echo in the option prepended with
1119 "no":: 1011 "no"::
1120 1012
1121 echo noprint-parent > trace_options 1013 echo noprint-parent > trace_options
1122 1014
1123 To enable an option, leave off the "no":: 1015 To enable an option, leave off the "no"::
1124 1016
1125 echo sym-offset > trace_options 1017 echo sym-offset > trace_options
1126 1018
1127 Here are the available options: 1019 Here are the available options:
1128 1020
1129 print-parent 1021 print-parent
1130 On function traces, display the calli 1022 On function traces, display the calling (parent)
1131 function as well as the function bein 1023 function as well as the function being traced.
1132 :: 1024 ::
1133 1025
1134 print-parent: 1026 print-parent:
1135 bash-4000 [01] 1477.606694: simp 1027 bash-4000 [01] 1477.606694: simple_strtoul <-kstrtoul
1136 1028
1137 noprint-parent: 1029 noprint-parent:
1138 bash-4000 [01] 1477.606694: simp 1030 bash-4000 [01] 1477.606694: simple_strtoul
1139 1031
1140 1032
1141 sym-offset 1033 sym-offset
1142 Display not only the function name, b 1034 Display not only the function name, but also the
1143 offset in the function. For example, 1035 offset in the function. For example, instead of
1144 seeing just "ktime_get", you will see 1036 seeing just "ktime_get", you will see
1145 "ktime_get+0xb/0x20". 1037 "ktime_get+0xb/0x20".
1146 :: 1038 ::
1147 1039
1148 sym-offset: 1040 sym-offset:
1149 bash-4000 [01] 1477.606694: simp 1041 bash-4000 [01] 1477.606694: simple_strtoul+0x6/0xa0
1150 1042
1151 sym-addr 1043 sym-addr
1152 This will also display the function a 1044 This will also display the function address as well
1153 as the function name. 1045 as the function name.
1154 :: 1046 ::
1155 1047
1156 sym-addr: 1048 sym-addr:
1157 bash-4000 [01] 1477.606694: simp 1049 bash-4000 [01] 1477.606694: simple_strtoul <c0339346>
1158 1050
1159 verbose 1051 verbose
1160 This deals with the trace file when t 1052 This deals with the trace file when the
1161 latency-format option is enabled. 1053 latency-format option is enabled.
1162 :: 1054 ::
1163 1055
1164 bash 4000 1 0 00000000 00010a95 1056 bash 4000 1 0 00000000 00010a95 [58127d26] 1720.415ms \
1165 (+0.000ms): simple_strtoul (kstrt 1057 (+0.000ms): simple_strtoul (kstrtoul)
1166 1058
1167 raw 1059 raw
1168 This will display raw numbers. This o 1060 This will display raw numbers. This option is best for
1169 use with user applications that can t 1061 use with user applications that can translate the raw
1170 numbers better than having it done in 1062 numbers better than having it done in the kernel.
1171 1063
1172 hex 1064 hex
1173 Similar to raw, but the numbers will 1065 Similar to raw, but the numbers will be in a hexadecimal format.
1174 1066
1175 bin 1067 bin
1176 This will print out the formats in ra 1068 This will print out the formats in raw binary.
1177 1069
1178 block 1070 block
1179 When set, reading trace_pipe will not 1071 When set, reading trace_pipe will not block when polled.
1180 1072
1181 fields <<
1182 Print the fields as described by thei <<
1183 option than using hex, bin or raw, as <<
1184 of the content of the event. <<
1185 <<
1186 trace_printk 1073 trace_printk
1187 Can disable trace_printk() from writi 1074 Can disable trace_printk() from writing into the buffer.
1188 1075
1189 trace_printk_dest <<
1190 Set to have trace_printk() and simila <<
1191 write into this instance. Note, only <<
1192 this set. By setting this flag, it cl <<
1193 of the instance that had it set previ <<
1194 level trace has this set, and will ge <<
1195 instance has it set then clears it. <<
1196 <<
1197 This flag cannot be cleared by the to <<
1198 default instance. The only way the to <<
1199 cleared, is by it being set in anothe <<
1200 <<
1201 annotate 1076 annotate
1202 It is sometimes confusing when the CP 1077 It is sometimes confusing when the CPU buffers are full
1203 and one CPU buffer had a lot of event 1078 and one CPU buffer had a lot of events recently, thus
1204 a shorter time frame, were another CP 1079 a shorter time frame, were another CPU may have only had
1205 a few events, which lets it have olde 1080 a few events, which lets it have older events. When
1206 the trace is reported, it shows the o 1081 the trace is reported, it shows the oldest events first,
1207 and it may look like only one CPU ran 1082 and it may look like only one CPU ran (the one with the
1208 oldest events). When the annotate opt 1083 oldest events). When the annotate option is set, it will
1209 display when a new CPU buffer started 1084 display when a new CPU buffer started::
1210 1085
1211 <idle>-0 [001] 1086 <idle>-0 [001] dNs4 21169.031481: wake_up_idle_cpu <-add_timer_on
1212 <idle>-0 [001] 1087 <idle>-0 [001] dNs4 21169.031482: _raw_spin_unlock_irqrestore <-add_timer_on
1213 <idle>-0 [001] 1088 <idle>-0 [001] .Ns4 21169.031484: sub_preempt_count <-_raw_spin_unlock_irqrestore
1214 ##### CPU 2 buffer started ## 1089 ##### CPU 2 buffer started ####
1215 <idle>-0 [002] 1090 <idle>-0 [002] .N.1 21169.031484: rcu_idle_exit <-cpu_idle
1216 <idle>-0 [001] 1091 <idle>-0 [001] .Ns3 21169.031484: _raw_spin_unlock <-clocksource_watchdog
1217 <idle>-0 [001] 1092 <idle>-0 [001] .Ns3 21169.031485: sub_preempt_count <-_raw_spin_unlock
1218 1093
1219 userstacktrace 1094 userstacktrace
1220 This option changes the trace. It rec 1095 This option changes the trace. It records a
1221 stacktrace of the current user space 1096 stacktrace of the current user space thread after
1222 each trace event. 1097 each trace event.
1223 1098
1224 sym-userobj 1099 sym-userobj
1225 when user stacktrace are enabled, loo 1100 when user stacktrace are enabled, look up which
1226 object the address belongs to, and pr 1101 object the address belongs to, and print a
1227 relative address. This is especially 1102 relative address. This is especially useful when
1228 ASLR is on, otherwise you don't get a 1103 ASLR is on, otherwise you don't get a chance to
1229 resolve the address to object/file/li 1104 resolve the address to object/file/line after
1230 the app is no longer running 1105 the app is no longer running
1231 1106
1232 The lookup is performed when you read 1107 The lookup is performed when you read
1233 trace,trace_pipe. Example:: 1108 trace,trace_pipe. Example::
1234 1109
1235 a.out-1623 [000] 40874.465 1110 a.out-1623 [000] 40874.465068: /root/a.out[+0x480] <-/root/a.out[+0
1236 x494] <- /root/a.out[+0x4a8 1111 x494] <- /root/a.out[+0x4a8] <- /lib/libc-2.7.so[+0x1e1a6]
1237 1112
1238 1113
1239 printk-msg-only 1114 printk-msg-only
1240 When set, trace_printk()s will only s 1115 When set, trace_printk()s will only show the format
1241 and not their parameters (if trace_bp 1116 and not their parameters (if trace_bprintk() or
1242 trace_bputs() was used to save the tr 1117 trace_bputs() was used to save the trace_printk()).
1243 1118
1244 context-info 1119 context-info
1245 Show only the event data. Hides the c 1120 Show only the event data. Hides the comm, PID,
1246 timestamp, CPU, and other useful data 1121 timestamp, CPU, and other useful data.
1247 1122
1248 latency-format 1123 latency-format
1249 This option changes the trace output. 1124 This option changes the trace output. When it is enabled,
1250 the trace displays additional informa 1125 the trace displays additional information about the
1251 latency, as described in "Latency tra 1126 latency, as described in "Latency trace format".
1252 1127
1253 pause-on-trace <<
1254 When set, opening the trace file for <<
1255 writing to the ring buffer (as if tra <<
1256 This simulates the original behavior <<
1257 When the file is closed, tracing will <<
1258 <<
1259 hash-ptr <<
1260 When set, "%p" in the event printk fo <<
1261 hashed pointer value instead of real <<
1262 This will be useful if you want to fi <<
1263 value is corresponding to the real va <<
1264 <<
1265 record-cmd 1128 record-cmd
1266 When any event or tracer is enabled, 1129 When any event or tracer is enabled, a hook is enabled
1267 in the sched_switch trace point to fi 1130 in the sched_switch trace point to fill comm cache
1268 with mapped pids and comms. But this 1131 with mapped pids and comms. But this may cause some
1269 overhead, and if you only care about 1132 overhead, and if you only care about pids, and not the
1270 name of the task, disabling this opti 1133 name of the task, disabling this option can lower the
1271 impact of tracing. See "saved_cmdline 1134 impact of tracing. See "saved_cmdlines".
1272 1135
1273 record-tgid 1136 record-tgid
1274 When any event or tracer is enabled, 1137 When any event or tracer is enabled, a hook is enabled
1275 in the sched_switch trace point to fi 1138 in the sched_switch trace point to fill the cache of
1276 mapped Thread Group IDs (TGID) mappin 1139 mapped Thread Group IDs (TGID) mapping to pids. See
1277 "saved_tgids". 1140 "saved_tgids".
1278 1141
1279 overwrite 1142 overwrite
1280 This controls what happens when the t 1143 This controls what happens when the trace buffer is
1281 full. If "1" (default), the oldest ev 1144 full. If "1" (default), the oldest events are
1282 discarded and overwritten. If "0", th 1145 discarded and overwritten. If "0", then the newest
1283 events are discarded. 1146 events are discarded.
1284 (see per_cpu/cpu0/stats for overrun a 1147 (see per_cpu/cpu0/stats for overrun and dropped)
1285 1148
1286 disable_on_free 1149 disable_on_free
1287 When the free_buffer is closed, traci 1150 When the free_buffer is closed, tracing will
1288 stop (tracing_on set to 0). 1151 stop (tracing_on set to 0).
1289 1152
1290 irq-info 1153 irq-info
1291 Shows the interrupt, preempt count, n 1154 Shows the interrupt, preempt count, need resched data.
1292 When disabled, the trace looks like:: 1155 When disabled, the trace looks like::
1293 1156
1294 # tracer: function 1157 # tracer: function
1295 # 1158 #
1296 # entries-in-buffer/entries-w 1159 # entries-in-buffer/entries-written: 144405/9452052 #P:4
1297 # 1160 #
1298 # TASK-PID CPU# 1161 # TASK-PID CPU# TIMESTAMP FUNCTION
1299 # | | | 1162 # | | | | |
1300 <idle>-0 [002] 1163 <idle>-0 [002] 23636.756054: ttwu_do_activate.constprop.89 <-try_to_wake_up
1301 <idle>-0 [002] 1164 <idle>-0 [002] 23636.756054: activate_task <-ttwu_do_activate.constprop.89
1302 <idle>-0 [002] 1165 <idle>-0 [002] 23636.756055: enqueue_task <-activate_task
1303 1166
1304 1167
1305 markers 1168 markers
1306 When set, the trace_marker is writabl 1169 When set, the trace_marker is writable (only by root).
1307 When disabled, the trace_marker will 1170 When disabled, the trace_marker will error with EINVAL
1308 on write. 1171 on write.
1309 1172
1310 event-fork 1173 event-fork
1311 When set, tasks with PIDs listed in s 1174 When set, tasks with PIDs listed in set_event_pid will have
1312 the PIDs of their children added to s 1175 the PIDs of their children added to set_event_pid when those
1313 tasks fork. Also, when tasks with PID 1176 tasks fork. Also, when tasks with PIDs in set_event_pid exit,
1314 their PIDs will be removed from the f 1177 their PIDs will be removed from the file.
1315 1178
1316 This affects PIDs listed in set_event <<
1317 <<
1318 function-trace 1179 function-trace
1319 The latency tracers will enable funct 1180 The latency tracers will enable function tracing
1320 if this option is enabled (default it 1181 if this option is enabled (default it is). When
1321 it is disabled, the latency tracers d 1182 it is disabled, the latency tracers do not trace
1322 functions. This keeps the overhead of 1183 functions. This keeps the overhead of the tracer down
1323 when performing latency tests. 1184 when performing latency tests.
1324 1185
1325 function-fork 1186 function-fork
1326 When set, tasks with PIDs listed in s 1187 When set, tasks with PIDs listed in set_ftrace_pid will
1327 have the PIDs of their children added 1188 have the PIDs of their children added to set_ftrace_pid
1328 when those tasks fork. Also, when tas 1189 when those tasks fork. Also, when tasks with PIDs in
1329 set_ftrace_pid exit, their PIDs will 1190 set_ftrace_pid exit, their PIDs will be removed from the
1330 file. 1191 file.
1331 1192
1332 This affects PIDs in set_ftrace_notra <<
1333 <<
1334 display-graph 1193 display-graph
1335 When set, the latency tracers (irqsof 1194 When set, the latency tracers (irqsoff, wakeup, etc) will
1336 use function graph tracing instead of 1195 use function graph tracing instead of function tracing.
1337 1196
1338 stacktrace 1197 stacktrace
1339 When set, a stack trace is recorded a 1198 When set, a stack trace is recorded after any trace event
1340 is recorded. 1199 is recorded.
1341 1200
1342 branch 1201 branch
1343 Enable branch tracing with the tracer 1202 Enable branch tracing with the tracer. This enables branch
1344 tracer along with the currently set t 1203 tracer along with the currently set tracer. Enabling this
1345 with the "nop" tracer is the same as 1204 with the "nop" tracer is the same as just enabling the
1346 "branch" tracer. 1205 "branch" tracer.
1347 1206
1348 .. tip:: Some tracers have their own options. 1207 .. tip:: Some tracers have their own options. They only appear in this
1349 file when the tracer is active. They a 1208 file when the tracer is active. They always appear in the
1350 options directory. 1209 options directory.
1351 1210
1352 1211
1353 Here are the per tracer options: 1212 Here are the per tracer options:
1354 1213
1355 Options for function tracer: 1214 Options for function tracer:
1356 1215
1357 func_stack_trace 1216 func_stack_trace
1358 When set, a stack trace is recorded a 1217 When set, a stack trace is recorded after every
1359 function that is recorded. NOTE! Limi 1218 function that is recorded. NOTE! Limit the functions
1360 that are recorded before enabling thi 1219 that are recorded before enabling this, with
1361 "set_ftrace_filter" otherwise the sys 1220 "set_ftrace_filter" otherwise the system performance
1362 will be critically degraded. Remember 1221 will be critically degraded. Remember to disable
1363 this option before clearing the funct 1222 this option before clearing the function filter.
1364 1223
1365 Options for function_graph tracer: 1224 Options for function_graph tracer:
1366 1225
1367 Since the function_graph tracer has a slight 1226 Since the function_graph tracer has a slightly different output
1368 it has its own options to control what is di 1227 it has its own options to control what is displayed.
1369 1228
1370 funcgraph-overrun 1229 funcgraph-overrun
1371 When set, the "overrun" of the graph 1230 When set, the "overrun" of the graph stack is
1372 displayed after each function traced. 1231 displayed after each function traced. The
1373 overrun, is when the stack depth of t 1232 overrun, is when the stack depth of the calls
1374 is greater than what is reserved for 1233 is greater than what is reserved for each task.
1375 Each task has a fixed array of functi 1234 Each task has a fixed array of functions to
1376 trace in the call graph. If the depth 1235 trace in the call graph. If the depth of the
1377 calls exceeds that, the function is n 1236 calls exceeds that, the function is not traced.
1378 The overrun is the number of function 1237 The overrun is the number of functions missed
1379 due to exceeding this array. 1238 due to exceeding this array.
1380 1239
1381 funcgraph-cpu 1240 funcgraph-cpu
1382 When set, the CPU number of the CPU w 1241 When set, the CPU number of the CPU where the trace
1383 occurred is displayed. 1242 occurred is displayed.
1384 1243
1385 funcgraph-overhead 1244 funcgraph-overhead
1386 When set, if the function takes longe 1245 When set, if the function takes longer than
1387 A certain amount, then a delay marker 1246 A certain amount, then a delay marker is
1388 displayed. See "delay" above, under t 1247 displayed. See "delay" above, under the
1389 header description. 1248 header description.
1390 1249
1391 funcgraph-proc 1250 funcgraph-proc
1392 Unlike other tracers, the process' co 1251 Unlike other tracers, the process' command line
1393 is not displayed by default, but inst 1252 is not displayed by default, but instead only
1394 when a task is traced in and out duri 1253 when a task is traced in and out during a context
1395 switch. Enabling this options has the 1254 switch. Enabling this options has the command
1396 of each process displayed at every li 1255 of each process displayed at every line.
1397 1256
1398 funcgraph-duration 1257 funcgraph-duration
1399 At the end of each function (the retu 1258 At the end of each function (the return)
1400 the duration of the amount of time in 1259 the duration of the amount of time in the
1401 function is displayed in microseconds 1260 function is displayed in microseconds.
1402 1261
1403 funcgraph-abstime 1262 funcgraph-abstime
1404 When set, the timestamp is displayed 1263 When set, the timestamp is displayed at each line.
1405 1264
1406 funcgraph-irqs 1265 funcgraph-irqs
1407 When disabled, functions that happen 1266 When disabled, functions that happen inside an
1408 interrupt will not be traced. 1267 interrupt will not be traced.
1409 1268
1410 funcgraph-tail 1269 funcgraph-tail
1411 When set, the return event will inclu 1270 When set, the return event will include the function
1412 that it represents. By default this i 1271 that it represents. By default this is off, and
1413 only a closing curly bracket "}" is d 1272 only a closing curly bracket "}" is displayed for
1414 the return of a function. 1273 the return of a function.
1415 1274
1416 funcgraph-retval <<
1417 When set, the return value of each tr <<
1418 will be printed after an equal sign " <<
1419 this is off. <<
1420 <<
1421 funcgraph-retval-hex <<
1422 When set, the return value will alway <<
1423 in hexadecimal format. If the option <<
1424 the return value is an error code, it <<
1425 in signed decimal format; otherwise i <<
1426 printed in hexadecimal format. By def <<
1427 is off. <<
1428 <<
1429 sleep-time 1275 sleep-time
1430 When running function graph tracer, t 1276 When running function graph tracer, to include
1431 the time a task schedules out in its 1277 the time a task schedules out in its function.
1432 When enabled, it will account time th 1278 When enabled, it will account time the task has been
1433 scheduled out as part of the function 1279 scheduled out as part of the function call.
1434 1280
1435 graph-time 1281 graph-time
1436 When running function profiler with f 1282 When running function profiler with function graph tracer,
1437 to include the time to call nested fu 1283 to include the time to call nested functions. When this is
1438 not set, the time reported for the fu 1284 not set, the time reported for the function will only
1439 include the time the function itself 1285 include the time the function itself executed for, not the
1440 time for functions that it called. 1286 time for functions that it called.
1441 1287
1442 Options for blk tracer: 1288 Options for blk tracer:
1443 1289
1444 blk_classic 1290 blk_classic
1445 Shows a more minimalistic output. 1291 Shows a more minimalistic output.
1446 1292
1447 1293
1448 irqsoff 1294 irqsoff
1449 ------- 1295 -------
1450 1296
1451 When interrupts are disabled, the CPU can not 1297 When interrupts are disabled, the CPU can not react to any other
1452 external event (besides NMIs and SMIs). This 1298 external event (besides NMIs and SMIs). This prevents the timer
1453 interrupt from triggering or the mouse interr 1299 interrupt from triggering or the mouse interrupt from letting
1454 the kernel know of a new mouse event. The res 1300 the kernel know of a new mouse event. The result is a latency
1455 with the reaction time. 1301 with the reaction time.
1456 1302
1457 The irqsoff tracer tracks the time for which 1303 The irqsoff tracer tracks the time for which interrupts are
1458 disabled. When a new maximum latency is hit, 1304 disabled. When a new maximum latency is hit, the tracer saves
1459 the trace leading up to that latency point so 1305 the trace leading up to that latency point so that every time a
1460 new maximum is reached, the old saved trace i 1306 new maximum is reached, the old saved trace is discarded and the
1461 new trace is saved. 1307 new trace is saved.
1462 1308
1463 To reset the maximum, echo 0 into tracing_max 1309 To reset the maximum, echo 0 into tracing_max_latency. Here is
1464 an example:: 1310 an example::
1465 1311
1466 # echo 0 > options/function-trace 1312 # echo 0 > options/function-trace
1467 # echo irqsoff > current_tracer 1313 # echo irqsoff > current_tracer
1468 # echo 1 > tracing_on 1314 # echo 1 > tracing_on
1469 # echo 0 > tracing_max_latency 1315 # echo 0 > tracing_max_latency
1470 # ls -ltr 1316 # ls -ltr
1471 [...] 1317 [...]
1472 # echo 0 > tracing_on 1318 # echo 0 > tracing_on
1473 # cat trace 1319 # cat trace
1474 # tracer: irqsoff 1320 # tracer: irqsoff
1475 # 1321 #
1476 # irqsoff latency trace v1.1.5 on 3.8.0-tes 1322 # irqsoff latency trace v1.1.5 on 3.8.0-test+
1477 # ----------------------------------------- 1323 # --------------------------------------------------------------------
1478 # latency: 16 us, #4/4, CPU#0 | (M:preempt 1324 # latency: 16 us, #4/4, CPU#0 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
1479 # ----------------- 1325 # -----------------
1480 # | task: swapper/0-0 (uid:0 nice:0 poli 1326 # | task: swapper/0-0 (uid:0 nice:0 policy:0 rt_prio:0)
1481 # ----------------- 1327 # -----------------
1482 # => started at: run_timer_softirq 1328 # => started at: run_timer_softirq
1483 # => ended at: run_timer_softirq 1329 # => ended at: run_timer_softirq
1484 # 1330 #
1485 # 1331 #
1486 # _------=> CPU# 1332 # _------=> CPU#
1487 # / _-----=> irqs-off 1333 # / _-----=> irqs-off
1488 # | / _----=> need-resched 1334 # | / _----=> need-resched
1489 # || / _---=> hardirq/softir 1335 # || / _---=> hardirq/softirq
1490 # ||| / _--=> preempt-depth 1336 # ||| / _--=> preempt-depth
1491 # |||| / delay 1337 # |||| / delay
1492 # cmd pid ||||| time | caller 1338 # cmd pid ||||| time | caller
1493 # \ / ||||| \ | / 1339 # \ / ||||| \ | /
1494 <idle>-0 0d.s2 0us+: _raw_spin_l 1340 <idle>-0 0d.s2 0us+: _raw_spin_lock_irq <-run_timer_softirq
1495 <idle>-0 0dNs3 17us : _raw_spin_u 1341 <idle>-0 0dNs3 17us : _raw_spin_unlock_irq <-run_timer_softirq
1496 <idle>-0 0dNs3 17us+: trace_hardi 1342 <idle>-0 0dNs3 17us+: trace_hardirqs_on <-run_timer_softirq
1497 <idle>-0 0dNs3 25us : <stack trac 1343 <idle>-0 0dNs3 25us : <stack trace>
1498 => _raw_spin_unlock_irq 1344 => _raw_spin_unlock_irq
1499 => run_timer_softirq 1345 => run_timer_softirq
1500 => __do_softirq 1346 => __do_softirq
1501 => call_softirq 1347 => call_softirq
1502 => do_softirq 1348 => do_softirq
1503 => irq_exit 1349 => irq_exit
1504 => smp_apic_timer_interrupt 1350 => smp_apic_timer_interrupt
1505 => apic_timer_interrupt 1351 => apic_timer_interrupt
1506 => rcu_idle_exit 1352 => rcu_idle_exit
1507 => cpu_idle 1353 => cpu_idle
1508 => rest_init 1354 => rest_init
1509 => start_kernel 1355 => start_kernel
1510 => x86_64_start_reservations 1356 => x86_64_start_reservations
1511 => x86_64_start_kernel 1357 => x86_64_start_kernel
1512 1358
1513 Here we see that we had a latency of 16 micro !! 1359 Here we see that that we had a latency of 16 microseconds (which is
1514 very good). The _raw_spin_lock_irq in run_tim 1360 very good). The _raw_spin_lock_irq in run_timer_softirq disabled
1515 interrupts. The difference between the 16 and 1361 interrupts. The difference between the 16 and the displayed
1516 timestamp 25us occurred because the clock was 1362 timestamp 25us occurred because the clock was incremented
1517 between the time of recording the max latency 1363 between the time of recording the max latency and the time of
1518 recording the function that had that latency. 1364 recording the function that had that latency.
1519 1365
1520 Note the above example had function-trace not 1366 Note the above example had function-trace not set. If we set
1521 function-trace, we get a much larger output:: 1367 function-trace, we get a much larger output::
1522 1368
1523 with echo 1 > options/function-trace 1369 with echo 1 > options/function-trace
1524 1370
1525 # tracer: irqsoff 1371 # tracer: irqsoff
1526 # 1372 #
1527 # irqsoff latency trace v1.1.5 on 3.8.0-tes 1373 # irqsoff latency trace v1.1.5 on 3.8.0-test+
1528 # ----------------------------------------- 1374 # --------------------------------------------------------------------
1529 # latency: 71 us, #168/168, CPU#3 | (M:pree 1375 # latency: 71 us, #168/168, CPU#3 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
1530 # ----------------- 1376 # -----------------
1531 # | task: bash-2042 (uid:0 nice:0 policy 1377 # | task: bash-2042 (uid:0 nice:0 policy:0 rt_prio:0)
1532 # ----------------- 1378 # -----------------
1533 # => started at: ata_scsi_queuecmd 1379 # => started at: ata_scsi_queuecmd
1534 # => ended at: ata_scsi_queuecmd 1380 # => ended at: ata_scsi_queuecmd
1535 # 1381 #
1536 # 1382 #
1537 # _------=> CPU# 1383 # _------=> CPU#
1538 # / _-----=> irqs-off 1384 # / _-----=> irqs-off
1539 # | / _----=> need-resched 1385 # | / _----=> need-resched
1540 # || / _---=> hardirq/softir 1386 # || / _---=> hardirq/softirq
1541 # ||| / _--=> preempt-depth 1387 # ||| / _--=> preempt-depth
1542 # |||| / delay 1388 # |||| / delay
1543 # cmd pid ||||| time | caller 1389 # cmd pid ||||| time | caller
1544 # \ / ||||| \ | / 1390 # \ / ||||| \ | /
1545 bash-2042 3d... 0us : _raw_spin_l 1391 bash-2042 3d... 0us : _raw_spin_lock_irqsave <-ata_scsi_queuecmd
1546 bash-2042 3d... 0us : add_preempt 1392 bash-2042 3d... 0us : add_preempt_count <-_raw_spin_lock_irqsave
1547 bash-2042 3d..1 1us : ata_scsi_fi 1393 bash-2042 3d..1 1us : ata_scsi_find_dev <-ata_scsi_queuecmd
1548 bash-2042 3d..1 1us : __ata_scsi_ 1394 bash-2042 3d..1 1us : __ata_scsi_find_dev <-ata_scsi_find_dev
1549 bash-2042 3d..1 2us : ata_find_de 1395 bash-2042 3d..1 2us : ata_find_dev.part.14 <-__ata_scsi_find_dev
1550 bash-2042 3d..1 2us : ata_qc_new_ 1396 bash-2042 3d..1 2us : ata_qc_new_init <-__ata_scsi_queuecmd
1551 bash-2042 3d..1 3us : ata_sg_init 1397 bash-2042 3d..1 3us : ata_sg_init <-__ata_scsi_queuecmd
1552 bash-2042 3d..1 4us : ata_scsi_rw 1398 bash-2042 3d..1 4us : ata_scsi_rw_xlat <-__ata_scsi_queuecmd
1553 bash-2042 3d..1 4us : ata_build_r 1399 bash-2042 3d..1 4us : ata_build_rw_tf <-ata_scsi_rw_xlat
1554 [...] 1400 [...]
1555 bash-2042 3d..1 67us : delay_tsc < 1401 bash-2042 3d..1 67us : delay_tsc <-__delay
1556 bash-2042 3d..1 67us : add_preempt 1402 bash-2042 3d..1 67us : add_preempt_count <-delay_tsc
1557 bash-2042 3d..2 67us : sub_preempt 1403 bash-2042 3d..2 67us : sub_preempt_count <-delay_tsc
1558 bash-2042 3d..1 67us : add_preempt 1404 bash-2042 3d..1 67us : add_preempt_count <-delay_tsc
1559 bash-2042 3d..2 68us : sub_preempt 1405 bash-2042 3d..2 68us : sub_preempt_count <-delay_tsc
1560 bash-2042 3d..1 68us+: ata_bmdma_s 1406 bash-2042 3d..1 68us+: ata_bmdma_start <-ata_bmdma_qc_issue
1561 bash-2042 3d..1 71us : _raw_spin_u 1407 bash-2042 3d..1 71us : _raw_spin_unlock_irqrestore <-ata_scsi_queuecmd
1562 bash-2042 3d..1 71us : _raw_spin_u 1408 bash-2042 3d..1 71us : _raw_spin_unlock_irqrestore <-ata_scsi_queuecmd
1563 bash-2042 3d..1 72us+: trace_hardi 1409 bash-2042 3d..1 72us+: trace_hardirqs_on <-ata_scsi_queuecmd
1564 bash-2042 3d..1 120us : <stack trac 1410 bash-2042 3d..1 120us : <stack trace>
1565 => _raw_spin_unlock_irqrestore 1411 => _raw_spin_unlock_irqrestore
1566 => ata_scsi_queuecmd 1412 => ata_scsi_queuecmd
1567 => scsi_dispatch_cmd 1413 => scsi_dispatch_cmd
1568 => scsi_request_fn 1414 => scsi_request_fn
1569 => __blk_run_queue_uncond 1415 => __blk_run_queue_uncond
1570 => __blk_run_queue 1416 => __blk_run_queue
1571 => blk_queue_bio 1417 => blk_queue_bio
1572 => submit_bio_noacct !! 1418 => generic_make_request
1573 => submit_bio 1419 => submit_bio
1574 => submit_bh 1420 => submit_bh
1575 => __ext3_get_inode_loc 1421 => __ext3_get_inode_loc
1576 => ext3_iget 1422 => ext3_iget
1577 => ext3_lookup 1423 => ext3_lookup
1578 => lookup_real 1424 => lookup_real
1579 => __lookup_hash 1425 => __lookup_hash
1580 => walk_component 1426 => walk_component
1581 => lookup_last 1427 => lookup_last
1582 => path_lookupat 1428 => path_lookupat
1583 => filename_lookup 1429 => filename_lookup
1584 => user_path_at_empty 1430 => user_path_at_empty
1585 => user_path_at 1431 => user_path_at
1586 => vfs_fstatat 1432 => vfs_fstatat
1587 => vfs_stat 1433 => vfs_stat
1588 => sys_newstat 1434 => sys_newstat
1589 => system_call_fastpath 1435 => system_call_fastpath
1590 1436
1591 1437
1592 Here we traced a 71 microsecond latency. But 1438 Here we traced a 71 microsecond latency. But we also see all the
1593 functions that were called during that time. 1439 functions that were called during that time. Note that by
1594 enabling function tracing, we incur an added 1440 enabling function tracing, we incur an added overhead. This
1595 overhead may extend the latency times. But ne 1441 overhead may extend the latency times. But nevertheless, this
1596 trace has provided some very helpful debuggin 1442 trace has provided some very helpful debugging information.
1597 1443
1598 If we prefer function graph output instead of 1444 If we prefer function graph output instead of function, we can set
1599 display-graph option:: 1445 display-graph option::
1600 1446
1601 with echo 1 > options/display-graph 1447 with echo 1 > options/display-graph
1602 1448
1603 # tracer: irqsoff 1449 # tracer: irqsoff
1604 # 1450 #
1605 # irqsoff latency trace v1.1.5 on 4.20.0-rc 1451 # irqsoff latency trace v1.1.5 on 4.20.0-rc6+
1606 # ----------------------------------------- 1452 # --------------------------------------------------------------------
1607 # latency: 3751 us, #274/274, CPU#0 | (M:de 1453 # latency: 3751 us, #274/274, CPU#0 | (M:desktop VP:0, KP:0, SP:0 HP:0 #P:4)
1608 # ----------------- 1454 # -----------------
1609 # | task: bash-1507 (uid:0 nice:0 policy 1455 # | task: bash-1507 (uid:0 nice:0 policy:0 rt_prio:0)
1610 # ----------------- 1456 # -----------------
1611 # => started at: free_debug_processing 1457 # => started at: free_debug_processing
1612 # => ended at: return_to_handler 1458 # => ended at: return_to_handler
1613 # 1459 #
1614 # 1460 #
1615 # _-- 1461 # _-----=> irqs-off
1616 # / _- 1462 # / _----=> need-resched
1617 # | / _ 1463 # | / _---=> hardirq/softirq
1618 # || / 1464 # || / _--=> preempt-depth
1619 # ||| / 1465 # ||| /
1620 # REL TIME CPU TASK/PID |||| 1466 # REL TIME CPU TASK/PID |||| DURATION FUNCTION CALLS
1621 # | | | | |||| 1467 # | | | | |||| | | | | | |
1622 0 us | 0) bash-1507 | d... 1468 0 us | 0) bash-1507 | d... | 0.000 us | _raw_spin_lock_irqsave();
1623 0 us | 0) bash-1507 | d..1 1469 0 us | 0) bash-1507 | d..1 | 0.378 us | do_raw_spin_trylock();
1624 1 us | 0) bash-1507 | d..2 1470 1 us | 0) bash-1507 | d..2 | | set_track() {
1625 2 us | 0) bash-1507 | d..2 1471 2 us | 0) bash-1507 | d..2 | | save_stack_trace() {
1626 2 us | 0) bash-1507 | d..2 1472 2 us | 0) bash-1507 | d..2 | | __save_stack_trace() {
1627 3 us | 0) bash-1507 | d..2 1473 3 us | 0) bash-1507 | d..2 | | __unwind_start() {
1628 3 us | 0) bash-1507 | d..2 1474 3 us | 0) bash-1507 | d..2 | | get_stack_info() {
1629 3 us | 0) bash-1507 | d..2 1475 3 us | 0) bash-1507 | d..2 | 0.351 us | in_task_stack();
1630 4 us | 0) bash-1507 | d..2 1476 4 us | 0) bash-1507 | d..2 | 1.107 us | }
1631 [...] 1477 [...]
1632 3750 us | 0) bash-1507 | d..1 1478 3750 us | 0) bash-1507 | d..1 | 0.516 us | do_raw_spin_unlock();
1633 3750 us | 0) bash-1507 | d..1 1479 3750 us | 0) bash-1507 | d..1 | 0.000 us | _raw_spin_unlock_irqrestore();
1634 3764 us | 0) bash-1507 | d..1 1480 3764 us | 0) bash-1507 | d..1 | 0.000 us | tracer_hardirqs_on();
1635 bash-1507 0d..1 3792us : <stack trac 1481 bash-1507 0d..1 3792us : <stack trace>
1636 => free_debug_processing 1482 => free_debug_processing
1637 => __slab_free 1483 => __slab_free
1638 => kmem_cache_free 1484 => kmem_cache_free
1639 => vm_area_free 1485 => vm_area_free
1640 => remove_vma 1486 => remove_vma
1641 => exit_mmap 1487 => exit_mmap
1642 => mmput 1488 => mmput
1643 => begin_new_exec !! 1489 => flush_old_exec
1644 => load_elf_binary 1490 => load_elf_binary
1645 => search_binary_handler 1491 => search_binary_handler
1646 => __do_execve_file.isra.32 1492 => __do_execve_file.isra.32
1647 => __x64_sys_execve 1493 => __x64_sys_execve
1648 => do_syscall_64 1494 => do_syscall_64
1649 => entry_SYSCALL_64_after_hwframe 1495 => entry_SYSCALL_64_after_hwframe
1650 1496
1651 preemptoff 1497 preemptoff
1652 ---------- 1498 ----------
1653 1499
1654 When preemption is disabled, we may be able t 1500 When preemption is disabled, we may be able to receive
1655 interrupts but the task cannot be preempted a 1501 interrupts but the task cannot be preempted and a higher
1656 priority task must wait for preemption to be 1502 priority task must wait for preemption to be enabled again
1657 before it can preempt a lower priority task. 1503 before it can preempt a lower priority task.
1658 1504
1659 The preemptoff tracer traces the places that 1505 The preemptoff tracer traces the places that disable preemption.
1660 Like the irqsoff tracer, it records the maxim 1506 Like the irqsoff tracer, it records the maximum latency for
1661 which preemption was disabled. The control of 1507 which preemption was disabled. The control of preemptoff tracer
1662 is much like the irqsoff tracer. 1508 is much like the irqsoff tracer.
1663 :: 1509 ::
1664 1510
1665 # echo 0 > options/function-trace 1511 # echo 0 > options/function-trace
1666 # echo preemptoff > current_tracer 1512 # echo preemptoff > current_tracer
1667 # echo 1 > tracing_on 1513 # echo 1 > tracing_on
1668 # echo 0 > tracing_max_latency 1514 # echo 0 > tracing_max_latency
1669 # ls -ltr 1515 # ls -ltr
1670 [...] 1516 [...]
1671 # echo 0 > tracing_on 1517 # echo 0 > tracing_on
1672 # cat trace 1518 # cat trace
1673 # tracer: preemptoff 1519 # tracer: preemptoff
1674 # 1520 #
1675 # preemptoff latency trace v1.1.5 on 3.8.0- 1521 # preemptoff latency trace v1.1.5 on 3.8.0-test+
1676 # ----------------------------------------- 1522 # --------------------------------------------------------------------
1677 # latency: 46 us, #4/4, CPU#1 | (M:preempt 1523 # latency: 46 us, #4/4, CPU#1 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
1678 # ----------------- 1524 # -----------------
1679 # | task: sshd-1991 (uid:0 nice:0 policy 1525 # | task: sshd-1991 (uid:0 nice:0 policy:0 rt_prio:0)
1680 # ----------------- 1526 # -----------------
1681 # => started at: do_IRQ 1527 # => started at: do_IRQ
1682 # => ended at: do_IRQ 1528 # => ended at: do_IRQ
1683 # 1529 #
1684 # 1530 #
1685 # _------=> CPU# 1531 # _------=> CPU#
1686 # / _-----=> irqs-off 1532 # / _-----=> irqs-off
1687 # | / _----=> need-resched 1533 # | / _----=> need-resched
1688 # || / _---=> hardirq/softir 1534 # || / _---=> hardirq/softirq
1689 # ||| / _--=> preempt-depth 1535 # ||| / _--=> preempt-depth
1690 # |||| / delay 1536 # |||| / delay
1691 # cmd pid ||||| time | caller 1537 # cmd pid ||||| time | caller
1692 # \ / ||||| \ | / 1538 # \ / ||||| \ | /
1693 sshd-1991 1d.h. 0us+: irq_enter < 1539 sshd-1991 1d.h. 0us+: irq_enter <-do_IRQ
1694 sshd-1991 1d..1 46us : irq_exit <- 1540 sshd-1991 1d..1 46us : irq_exit <-do_IRQ
1695 sshd-1991 1d..1 47us+: trace_preem 1541 sshd-1991 1d..1 47us+: trace_preempt_on <-do_IRQ
1696 sshd-1991 1d..1 52us : <stack trac 1542 sshd-1991 1d..1 52us : <stack trace>
1697 => sub_preempt_count 1543 => sub_preempt_count
1698 => irq_exit 1544 => irq_exit
1699 => do_IRQ 1545 => do_IRQ
1700 => ret_from_intr 1546 => ret_from_intr
1701 1547
1702 1548
1703 This has some more changes. Preemption was di 1549 This has some more changes. Preemption was disabled when an
1704 interrupt came in (notice the 'h'), and was e 1550 interrupt came in (notice the 'h'), and was enabled on exit.
1705 But we also see that interrupts have been dis 1551 But we also see that interrupts have been disabled when entering
1706 the preempt off section and leaving it (the ' 1552 the preempt off section and leaving it (the 'd'). We do not know if
1707 interrupts were enabled in the mean time or s 1553 interrupts were enabled in the mean time or shortly after this
1708 was over. 1554 was over.
1709 :: 1555 ::
1710 1556
1711 # tracer: preemptoff 1557 # tracer: preemptoff
1712 # 1558 #
1713 # preemptoff latency trace v1.1.5 on 3.8.0- 1559 # preemptoff latency trace v1.1.5 on 3.8.0-test+
1714 # ----------------------------------------- 1560 # --------------------------------------------------------------------
1715 # latency: 83 us, #241/241, CPU#1 | (M:pree 1561 # latency: 83 us, #241/241, CPU#1 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
1716 # ----------------- 1562 # -----------------
1717 # | task: bash-1994 (uid:0 nice:0 policy 1563 # | task: bash-1994 (uid:0 nice:0 policy:0 rt_prio:0)
1718 # ----------------- 1564 # -----------------
1719 # => started at: wake_up_new_task 1565 # => started at: wake_up_new_task
1720 # => ended at: task_rq_unlock 1566 # => ended at: task_rq_unlock
1721 # 1567 #
1722 # 1568 #
1723 # _------=> CPU# 1569 # _------=> CPU#
1724 # / _-----=> irqs-off 1570 # / _-----=> irqs-off
1725 # | / _----=> need-resched 1571 # | / _----=> need-resched
1726 # || / _---=> hardirq/softir 1572 # || / _---=> hardirq/softirq
1727 # ||| / _--=> preempt-depth 1573 # ||| / _--=> preempt-depth
1728 # |||| / delay 1574 # |||| / delay
1729 # cmd pid ||||| time | caller 1575 # cmd pid ||||| time | caller
1730 # \ / ||||| \ | / 1576 # \ / ||||| \ | /
1731 bash-1994 1d..1 0us : _raw_spin_l 1577 bash-1994 1d..1 0us : _raw_spin_lock_irqsave <-wake_up_new_task
1732 bash-1994 1d..1 0us : select_task 1578 bash-1994 1d..1 0us : select_task_rq_fair <-select_task_rq
1733 bash-1994 1d..1 1us : __rcu_read_ 1579 bash-1994 1d..1 1us : __rcu_read_lock <-select_task_rq_fair
1734 bash-1994 1d..1 1us : source_load 1580 bash-1994 1d..1 1us : source_load <-select_task_rq_fair
1735 bash-1994 1d..1 1us : source_load 1581 bash-1994 1d..1 1us : source_load <-select_task_rq_fair
1736 [...] 1582 [...]
1737 bash-1994 1d..1 12us : irq_enter < 1583 bash-1994 1d..1 12us : irq_enter <-smp_apic_timer_interrupt
1738 bash-1994 1d..1 12us : rcu_irq_ent 1584 bash-1994 1d..1 12us : rcu_irq_enter <-irq_enter
1739 bash-1994 1d..1 13us : add_preempt 1585 bash-1994 1d..1 13us : add_preempt_count <-irq_enter
1740 bash-1994 1d.h1 13us : exit_idle < 1586 bash-1994 1d.h1 13us : exit_idle <-smp_apic_timer_interrupt
1741 bash-1994 1d.h1 13us : hrtimer_int 1587 bash-1994 1d.h1 13us : hrtimer_interrupt <-smp_apic_timer_interrupt
1742 bash-1994 1d.h1 13us : _raw_spin_l 1588 bash-1994 1d.h1 13us : _raw_spin_lock <-hrtimer_interrupt
1743 bash-1994 1d.h1 14us : add_preempt 1589 bash-1994 1d.h1 14us : add_preempt_count <-_raw_spin_lock
1744 bash-1994 1d.h2 14us : ktime_get_u 1590 bash-1994 1d.h2 14us : ktime_get_update_offsets <-hrtimer_interrupt
1745 [...] 1591 [...]
1746 bash-1994 1d.h1 35us : lapic_next_ 1592 bash-1994 1d.h1 35us : lapic_next_event <-clockevents_program_event
1747 bash-1994 1d.h1 35us : irq_exit <- 1593 bash-1994 1d.h1 35us : irq_exit <-smp_apic_timer_interrupt
1748 bash-1994 1d.h1 36us : sub_preempt 1594 bash-1994 1d.h1 36us : sub_preempt_count <-irq_exit
1749 bash-1994 1d..2 36us : do_softirq 1595 bash-1994 1d..2 36us : do_softirq <-irq_exit
1750 bash-1994 1d..2 36us : __do_softir 1596 bash-1994 1d..2 36us : __do_softirq <-call_softirq
1751 bash-1994 1d..2 36us : __local_bh_ 1597 bash-1994 1d..2 36us : __local_bh_disable <-__do_softirq
1752 bash-1994 1d.s2 37us : add_preempt 1598 bash-1994 1d.s2 37us : add_preempt_count <-_raw_spin_lock_irq
1753 bash-1994 1d.s3 38us : _raw_spin_u 1599 bash-1994 1d.s3 38us : _raw_spin_unlock <-run_timer_softirq
1754 bash-1994 1d.s3 39us : sub_preempt 1600 bash-1994 1d.s3 39us : sub_preempt_count <-_raw_spin_unlock
1755 bash-1994 1d.s2 39us : call_timer_ 1601 bash-1994 1d.s2 39us : call_timer_fn <-run_timer_softirq
1756 [...] 1602 [...]
1757 bash-1994 1dNs2 81us : cpu_needs_a 1603 bash-1994 1dNs2 81us : cpu_needs_another_gp <-rcu_process_callbacks
1758 bash-1994 1dNs2 82us : __local_bh_ 1604 bash-1994 1dNs2 82us : __local_bh_enable <-__do_softirq
1759 bash-1994 1dNs2 82us : sub_preempt 1605 bash-1994 1dNs2 82us : sub_preempt_count <-__local_bh_enable
1760 bash-1994 1dN.2 82us : idle_cpu <- 1606 bash-1994 1dN.2 82us : idle_cpu <-irq_exit
1761 bash-1994 1dN.2 83us : rcu_irq_exi 1607 bash-1994 1dN.2 83us : rcu_irq_exit <-irq_exit
1762 bash-1994 1dN.2 83us : sub_preempt 1608 bash-1994 1dN.2 83us : sub_preempt_count <-irq_exit
1763 bash-1994 1.N.1 84us : _raw_spin_u 1609 bash-1994 1.N.1 84us : _raw_spin_unlock_irqrestore <-task_rq_unlock
1764 bash-1994 1.N.1 84us+: trace_preem 1610 bash-1994 1.N.1 84us+: trace_preempt_on <-task_rq_unlock
1765 bash-1994 1.N.1 104us : <stack trac 1611 bash-1994 1.N.1 104us : <stack trace>
1766 => sub_preempt_count 1612 => sub_preempt_count
1767 => _raw_spin_unlock_irqrestore 1613 => _raw_spin_unlock_irqrestore
1768 => task_rq_unlock 1614 => task_rq_unlock
1769 => wake_up_new_task 1615 => wake_up_new_task
1770 => do_fork 1616 => do_fork
1771 => sys_clone 1617 => sys_clone
1772 => stub_clone 1618 => stub_clone
1773 1619
1774 1620
1775 The above is an example of the preemptoff tra 1621 The above is an example of the preemptoff trace with
1776 function-trace set. Here we see that interrup 1622 function-trace set. Here we see that interrupts were not disabled
1777 the entire time. The irq_enter code lets us k 1623 the entire time. The irq_enter code lets us know that we entered
1778 an interrupt 'h'. Before that, the functions 1624 an interrupt 'h'. Before that, the functions being traced still
1779 show that it is not in an interrupt, but we c 1625 show that it is not in an interrupt, but we can see from the
1780 functions themselves that this is not the cas 1626 functions themselves that this is not the case.
1781 1627
1782 preemptirqsoff 1628 preemptirqsoff
1783 -------------- 1629 --------------
1784 1630
1785 Knowing the locations that have interrupts di 1631 Knowing the locations that have interrupts disabled or
1786 preemption disabled for the longest times is 1632 preemption disabled for the longest times is helpful. But
1787 sometimes we would like to know when either p 1633 sometimes we would like to know when either preemption and/or
1788 interrupts are disabled. 1634 interrupts are disabled.
1789 1635
1790 Consider the following code:: 1636 Consider the following code::
1791 1637
1792 local_irq_disable(); 1638 local_irq_disable();
1793 call_function_with_irqs_off(); 1639 call_function_with_irqs_off();
1794 preempt_disable(); 1640 preempt_disable();
1795 call_function_with_irqs_and_preemption_of 1641 call_function_with_irqs_and_preemption_off();
1796 local_irq_enable(); 1642 local_irq_enable();
1797 call_function_with_preemption_off(); 1643 call_function_with_preemption_off();
1798 preempt_enable(); 1644 preempt_enable();
1799 1645
1800 The irqsoff tracer will record the total leng 1646 The irqsoff tracer will record the total length of
1801 call_function_with_irqs_off() and 1647 call_function_with_irqs_off() and
1802 call_function_with_irqs_and_preemption_off(). 1648 call_function_with_irqs_and_preemption_off().
1803 1649
1804 The preemptoff tracer will record the total l 1650 The preemptoff tracer will record the total length of
1805 call_function_with_irqs_and_preemption_off() 1651 call_function_with_irqs_and_preemption_off() and
1806 call_function_with_preemption_off(). 1652 call_function_with_preemption_off().
1807 1653
1808 But neither will trace the time that interrup 1654 But neither will trace the time that interrupts and/or
1809 preemption is disabled. This total time is th 1655 preemption is disabled. This total time is the time that we can
1810 not schedule. To record this time, use the pr 1656 not schedule. To record this time, use the preemptirqsoff
1811 tracer. 1657 tracer.
1812 1658
1813 Again, using this trace is much like the irqs 1659 Again, using this trace is much like the irqsoff and preemptoff
1814 tracers. 1660 tracers.
1815 :: 1661 ::
1816 1662
1817 # echo 0 > options/function-trace 1663 # echo 0 > options/function-trace
1818 # echo preemptirqsoff > current_tracer 1664 # echo preemptirqsoff > current_tracer
1819 # echo 1 > tracing_on 1665 # echo 1 > tracing_on
1820 # echo 0 > tracing_max_latency 1666 # echo 0 > tracing_max_latency
1821 # ls -ltr 1667 # ls -ltr
1822 [...] 1668 [...]
1823 # echo 0 > tracing_on 1669 # echo 0 > tracing_on
1824 # cat trace 1670 # cat trace
1825 # tracer: preemptirqsoff 1671 # tracer: preemptirqsoff
1826 # 1672 #
1827 # preemptirqsoff latency trace v1.1.5 on 3. 1673 # preemptirqsoff latency trace v1.1.5 on 3.8.0-test+
1828 # ----------------------------------------- 1674 # --------------------------------------------------------------------
1829 # latency: 100 us, #4/4, CPU#3 | (M:preempt 1675 # latency: 100 us, #4/4, CPU#3 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
1830 # ----------------- 1676 # -----------------
1831 # | task: ls-2230 (uid:0 nice:0 policy:0 1677 # | task: ls-2230 (uid:0 nice:0 policy:0 rt_prio:0)
1832 # ----------------- 1678 # -----------------
1833 # => started at: ata_scsi_queuecmd 1679 # => started at: ata_scsi_queuecmd
1834 # => ended at: ata_scsi_queuecmd 1680 # => ended at: ata_scsi_queuecmd
1835 # 1681 #
1836 # 1682 #
1837 # _------=> CPU# 1683 # _------=> CPU#
1838 # / _-----=> irqs-off 1684 # / _-----=> irqs-off
1839 # | / _----=> need-resched 1685 # | / _----=> need-resched
1840 # || / _---=> hardirq/softir 1686 # || / _---=> hardirq/softirq
1841 # ||| / _--=> preempt-depth 1687 # ||| / _--=> preempt-depth
1842 # |||| / delay 1688 # |||| / delay
1843 # cmd pid ||||| time | caller 1689 # cmd pid ||||| time | caller
1844 # \ / ||||| \ | / 1690 # \ / ||||| \ | /
1845 ls-2230 3d... 0us+: _raw_spin_l 1691 ls-2230 3d... 0us+: _raw_spin_lock_irqsave <-ata_scsi_queuecmd
1846 ls-2230 3...1 100us : _raw_spin_u 1692 ls-2230 3...1 100us : _raw_spin_unlock_irqrestore <-ata_scsi_queuecmd
1847 ls-2230 3...1 101us+: trace_preem 1693 ls-2230 3...1 101us+: trace_preempt_on <-ata_scsi_queuecmd
1848 ls-2230 3...1 111us : <stack trac 1694 ls-2230 3...1 111us : <stack trace>
1849 => sub_preempt_count 1695 => sub_preempt_count
1850 => _raw_spin_unlock_irqrestore 1696 => _raw_spin_unlock_irqrestore
1851 => ata_scsi_queuecmd 1697 => ata_scsi_queuecmd
1852 => scsi_dispatch_cmd 1698 => scsi_dispatch_cmd
1853 => scsi_request_fn 1699 => scsi_request_fn
1854 => __blk_run_queue_uncond 1700 => __blk_run_queue_uncond
1855 => __blk_run_queue 1701 => __blk_run_queue
1856 => blk_queue_bio 1702 => blk_queue_bio
1857 => submit_bio_noacct !! 1703 => generic_make_request
1858 => submit_bio 1704 => submit_bio
1859 => submit_bh 1705 => submit_bh
1860 => ext3_bread 1706 => ext3_bread
1861 => ext3_dir_bread 1707 => ext3_dir_bread
1862 => htree_dirblock_to_tree 1708 => htree_dirblock_to_tree
1863 => ext3_htree_fill_tree 1709 => ext3_htree_fill_tree
1864 => ext3_readdir 1710 => ext3_readdir
1865 => vfs_readdir 1711 => vfs_readdir
1866 => sys_getdents 1712 => sys_getdents
1867 => system_call_fastpath 1713 => system_call_fastpath
1868 1714
1869 1715
1870 The trace_hardirqs_off_thunk is called from a 1716 The trace_hardirqs_off_thunk is called from assembly on x86 when
1871 interrupts are disabled in the assembly code. 1717 interrupts are disabled in the assembly code. Without the
1872 function tracing, we do not know if interrupt 1718 function tracing, we do not know if interrupts were enabled
1873 within the preemption points. We do see that 1719 within the preemption points. We do see that it started with
1874 preemption enabled. 1720 preemption enabled.
1875 1721
1876 Here is a trace with function-trace set:: 1722 Here is a trace with function-trace set::
1877 1723
1878 # tracer: preemptirqsoff 1724 # tracer: preemptirqsoff
1879 # 1725 #
1880 # preemptirqsoff latency trace v1.1.5 on 3. 1726 # preemptirqsoff latency trace v1.1.5 on 3.8.0-test+
1881 # ----------------------------------------- 1727 # --------------------------------------------------------------------
1882 # latency: 161 us, #339/339, CPU#3 | (M:pre 1728 # latency: 161 us, #339/339, CPU#3 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
1883 # ----------------- 1729 # -----------------
1884 # | task: ls-2269 (uid:0 nice:0 policy:0 1730 # | task: ls-2269 (uid:0 nice:0 policy:0 rt_prio:0)
1885 # ----------------- 1731 # -----------------
1886 # => started at: schedule 1732 # => started at: schedule
1887 # => ended at: mutex_unlock 1733 # => ended at: mutex_unlock
1888 # 1734 #
1889 # 1735 #
1890 # _------=> CPU# 1736 # _------=> CPU#
1891 # / _-----=> irqs-off 1737 # / _-----=> irqs-off
1892 # | / _----=> need-resched 1738 # | / _----=> need-resched
1893 # || / _---=> hardirq/softir 1739 # || / _---=> hardirq/softirq
1894 # ||| / _--=> preempt-depth 1740 # ||| / _--=> preempt-depth
1895 # |||| / delay 1741 # |||| / delay
1896 # cmd pid ||||| time | caller 1742 # cmd pid ||||| time | caller
1897 # \ / ||||| \ | / 1743 # \ / ||||| \ | /
1898 kworker/-59 3...1 0us : __schedule 1744 kworker/-59 3...1 0us : __schedule <-schedule
1899 kworker/-59 3d..1 0us : rcu_preempt 1745 kworker/-59 3d..1 0us : rcu_preempt_qs <-rcu_note_context_switch
1900 kworker/-59 3d..1 1us : add_preempt 1746 kworker/-59 3d..1 1us : add_preempt_count <-_raw_spin_lock_irq
1901 kworker/-59 3d..2 1us : deactivate_ 1747 kworker/-59 3d..2 1us : deactivate_task <-__schedule
1902 kworker/-59 3d..2 1us : dequeue_tas 1748 kworker/-59 3d..2 1us : dequeue_task <-deactivate_task
1903 kworker/-59 3d..2 2us : update_rq_c 1749 kworker/-59 3d..2 2us : update_rq_clock <-dequeue_task
1904 kworker/-59 3d..2 2us : dequeue_tas 1750 kworker/-59 3d..2 2us : dequeue_task_fair <-dequeue_task
1905 kworker/-59 3d..2 2us : update_curr 1751 kworker/-59 3d..2 2us : update_curr <-dequeue_task_fair
1906 kworker/-59 3d..2 2us : update_min_ 1752 kworker/-59 3d..2 2us : update_min_vruntime <-update_curr
1907 kworker/-59 3d..2 3us : cpuacct_cha 1753 kworker/-59 3d..2 3us : cpuacct_charge <-update_curr
1908 kworker/-59 3d..2 3us : __rcu_read_ 1754 kworker/-59 3d..2 3us : __rcu_read_lock <-cpuacct_charge
1909 kworker/-59 3d..2 3us : __rcu_read_ 1755 kworker/-59 3d..2 3us : __rcu_read_unlock <-cpuacct_charge
1910 kworker/-59 3d..2 3us : update_cfs_ 1756 kworker/-59 3d..2 3us : update_cfs_rq_blocked_load <-dequeue_task_fair
1911 kworker/-59 3d..2 4us : clear_buddi 1757 kworker/-59 3d..2 4us : clear_buddies <-dequeue_task_fair
1912 kworker/-59 3d..2 4us : account_ent 1758 kworker/-59 3d..2 4us : account_entity_dequeue <-dequeue_task_fair
1913 kworker/-59 3d..2 4us : update_min_ 1759 kworker/-59 3d..2 4us : update_min_vruntime <-dequeue_task_fair
1914 kworker/-59 3d..2 4us : update_cfs_ 1760 kworker/-59 3d..2 4us : update_cfs_shares <-dequeue_task_fair
1915 kworker/-59 3d..2 5us : hrtick_upda 1761 kworker/-59 3d..2 5us : hrtick_update <-dequeue_task_fair
1916 kworker/-59 3d..2 5us : wq_worker_s 1762 kworker/-59 3d..2 5us : wq_worker_sleeping <-__schedule
1917 kworker/-59 3d..2 5us : kthread_dat 1763 kworker/-59 3d..2 5us : kthread_data <-wq_worker_sleeping
1918 kworker/-59 3d..2 5us : put_prev_ta 1764 kworker/-59 3d..2 5us : put_prev_task_fair <-__schedule
1919 kworker/-59 3d..2 6us : pick_next_t 1765 kworker/-59 3d..2 6us : pick_next_task_fair <-pick_next_task
1920 kworker/-59 3d..2 6us : clear_buddi 1766 kworker/-59 3d..2 6us : clear_buddies <-pick_next_task_fair
1921 kworker/-59 3d..2 6us : set_next_en 1767 kworker/-59 3d..2 6us : set_next_entity <-pick_next_task_fair
1922 kworker/-59 3d..2 6us : update_stat 1768 kworker/-59 3d..2 6us : update_stats_wait_end <-set_next_entity
1923 ls-2269 3d..2 7us : finish_task 1769 ls-2269 3d..2 7us : finish_task_switch <-__schedule
1924 ls-2269 3d..2 7us : _raw_spin_u 1770 ls-2269 3d..2 7us : _raw_spin_unlock_irq <-finish_task_switch
1925 ls-2269 3d..2 8us : do_IRQ <-re 1771 ls-2269 3d..2 8us : do_IRQ <-ret_from_intr
1926 ls-2269 3d..2 8us : irq_enter < 1772 ls-2269 3d..2 8us : irq_enter <-do_IRQ
1927 ls-2269 3d..2 8us : rcu_irq_ent 1773 ls-2269 3d..2 8us : rcu_irq_enter <-irq_enter
1928 ls-2269 3d..2 9us : add_preempt 1774 ls-2269 3d..2 9us : add_preempt_count <-irq_enter
1929 ls-2269 3d.h2 9us : exit_idle < 1775 ls-2269 3d.h2 9us : exit_idle <-do_IRQ
1930 [...] 1776 [...]
1931 ls-2269 3d.h3 20us : sub_preempt 1777 ls-2269 3d.h3 20us : sub_preempt_count <-_raw_spin_unlock
1932 ls-2269 3d.h2 20us : irq_exit <- 1778 ls-2269 3d.h2 20us : irq_exit <-do_IRQ
1933 ls-2269 3d.h2 21us : sub_preempt 1779 ls-2269 3d.h2 21us : sub_preempt_count <-irq_exit
1934 ls-2269 3d..3 21us : do_softirq 1780 ls-2269 3d..3 21us : do_softirq <-irq_exit
1935 ls-2269 3d..3 21us : __do_softir 1781 ls-2269 3d..3 21us : __do_softirq <-call_softirq
1936 ls-2269 3d..3 21us+: __local_bh_ 1782 ls-2269 3d..3 21us+: __local_bh_disable <-__do_softirq
1937 ls-2269 3d.s4 29us : sub_preempt 1783 ls-2269 3d.s4 29us : sub_preempt_count <-_local_bh_enable_ip
1938 ls-2269 3d.s5 29us : sub_preempt 1784 ls-2269 3d.s5 29us : sub_preempt_count <-_local_bh_enable_ip
1939 ls-2269 3d.s5 31us : do_IRQ <-re 1785 ls-2269 3d.s5 31us : do_IRQ <-ret_from_intr
1940 ls-2269 3d.s5 31us : irq_enter < 1786 ls-2269 3d.s5 31us : irq_enter <-do_IRQ
1941 ls-2269 3d.s5 31us : rcu_irq_ent 1787 ls-2269 3d.s5 31us : rcu_irq_enter <-irq_enter
1942 [...] 1788 [...]
1943 ls-2269 3d.s5 31us : rcu_irq_ent 1789 ls-2269 3d.s5 31us : rcu_irq_enter <-irq_enter
1944 ls-2269 3d.s5 32us : add_preempt 1790 ls-2269 3d.s5 32us : add_preempt_count <-irq_enter
1945 ls-2269 3d.H5 32us : exit_idle < 1791 ls-2269 3d.H5 32us : exit_idle <-do_IRQ
1946 ls-2269 3d.H5 32us : handle_irq 1792 ls-2269 3d.H5 32us : handle_irq <-do_IRQ
1947 ls-2269 3d.H5 32us : irq_to_desc 1793 ls-2269 3d.H5 32us : irq_to_desc <-handle_irq
1948 ls-2269 3d.H5 33us : handle_fast 1794 ls-2269 3d.H5 33us : handle_fasteoi_irq <-handle_irq
1949 [...] 1795 [...]
1950 ls-2269 3d.s5 158us : _raw_spin_u 1796 ls-2269 3d.s5 158us : _raw_spin_unlock_irqrestore <-rtl8139_poll
1951 ls-2269 3d.s3 158us : net_rps_act 1797 ls-2269 3d.s3 158us : net_rps_action_and_irq_enable.isra.65 <-net_rx_action
1952 ls-2269 3d.s3 159us : __local_bh_ 1798 ls-2269 3d.s3 159us : __local_bh_enable <-__do_softirq
1953 ls-2269 3d.s3 159us : sub_preempt 1799 ls-2269 3d.s3 159us : sub_preempt_count <-__local_bh_enable
1954 ls-2269 3d..3 159us : idle_cpu <- 1800 ls-2269 3d..3 159us : idle_cpu <-irq_exit
1955 ls-2269 3d..3 159us : rcu_irq_exi 1801 ls-2269 3d..3 159us : rcu_irq_exit <-irq_exit
1956 ls-2269 3d..3 160us : sub_preempt 1802 ls-2269 3d..3 160us : sub_preempt_count <-irq_exit
1957 ls-2269 3d... 161us : __mutex_unl 1803 ls-2269 3d... 161us : __mutex_unlock_slowpath <-mutex_unlock
1958 ls-2269 3d... 162us+: trace_hardi 1804 ls-2269 3d... 162us+: trace_hardirqs_on <-mutex_unlock
1959 ls-2269 3d... 186us : <stack trac 1805 ls-2269 3d... 186us : <stack trace>
1960 => __mutex_unlock_slowpath 1806 => __mutex_unlock_slowpath
1961 => mutex_unlock 1807 => mutex_unlock
1962 => process_output 1808 => process_output
1963 => n_tty_write 1809 => n_tty_write
1964 => tty_write 1810 => tty_write
1965 => vfs_write 1811 => vfs_write
1966 => sys_write 1812 => sys_write
1967 => system_call_fastpath 1813 => system_call_fastpath
1968 1814
1969 This is an interesting trace. It started with 1815 This is an interesting trace. It started with kworker running and
1970 scheduling out and ls taking over. But as soo 1816 scheduling out and ls taking over. But as soon as ls released the
1971 rq lock and enabled interrupts (but not preem 1817 rq lock and enabled interrupts (but not preemption) an interrupt
1972 triggered. When the interrupt finished, it st 1818 triggered. When the interrupt finished, it started running softirqs.
1973 But while the softirq was running, another in 1819 But while the softirq was running, another interrupt triggered.
1974 When an interrupt is running inside a softirq 1820 When an interrupt is running inside a softirq, the annotation is 'H'.
1975 1821
1976 1822
1977 wakeup 1823 wakeup
1978 ------ 1824 ------
1979 1825
1980 One common case that people are interested in 1826 One common case that people are interested in tracing is the
1981 time it takes for a task that is woken to act 1827 time it takes for a task that is woken to actually wake up.
1982 Now for non Real-Time tasks, this can be arbi 1828 Now for non Real-Time tasks, this can be arbitrary. But tracing
1983 it nonetheless can be interesting. !! 1829 it none the less can be interesting.
1984 1830
1985 Without function tracing:: 1831 Without function tracing::
1986 1832
1987 # echo 0 > options/function-trace 1833 # echo 0 > options/function-trace
1988 # echo wakeup > current_tracer 1834 # echo wakeup > current_tracer
1989 # echo 1 > tracing_on 1835 # echo 1 > tracing_on
1990 # echo 0 > tracing_max_latency 1836 # echo 0 > tracing_max_latency
1991 # chrt -f 5 sleep 1 1837 # chrt -f 5 sleep 1
1992 # echo 0 > tracing_on 1838 # echo 0 > tracing_on
1993 # cat trace 1839 # cat trace
1994 # tracer: wakeup 1840 # tracer: wakeup
1995 # 1841 #
1996 # wakeup latency trace v1.1.5 on 3.8.0-test 1842 # wakeup latency trace v1.1.5 on 3.8.0-test+
1997 # ----------------------------------------- 1843 # --------------------------------------------------------------------
1998 # latency: 15 us, #4/4, CPU#3 | (M:preempt 1844 # latency: 15 us, #4/4, CPU#3 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
1999 # ----------------- 1845 # -----------------
2000 # | task: kworker/3:1H-312 (uid:0 nice:- 1846 # | task: kworker/3:1H-312 (uid:0 nice:-20 policy:0 rt_prio:0)
2001 # ----------------- 1847 # -----------------
2002 # 1848 #
2003 # _------=> CPU# 1849 # _------=> CPU#
2004 # / _-----=> irqs-off 1850 # / _-----=> irqs-off
2005 # | / _----=> need-resched 1851 # | / _----=> need-resched
2006 # || / _---=> hardirq/softir 1852 # || / _---=> hardirq/softirq
2007 # ||| / _--=> preempt-depth 1853 # ||| / _--=> preempt-depth
2008 # |||| / delay 1854 # |||| / delay
2009 # cmd pid ||||| time | caller 1855 # cmd pid ||||| time | caller
2010 # \ / ||||| \ | / 1856 # \ / ||||| \ | /
2011 <idle>-0 3dNs7 0us : 0:120: 1857 <idle>-0 3dNs7 0us : 0:120:R + [003] 312:100:R kworker/3:1H
2012 <idle>-0 3dNs7 1us+: ttwu_do_act 1858 <idle>-0 3dNs7 1us+: ttwu_do_activate.constprop.87 <-try_to_wake_up
2013 <idle>-0 3d..3 15us : __schedule 1859 <idle>-0 3d..3 15us : __schedule <-schedule
2014 <idle>-0 3d..3 15us : 0:120: 1860 <idle>-0 3d..3 15us : 0:120:R ==> [003] 312:100:R kworker/3:1H
2015 1861
2016 The tracer only traces the highest priority t 1862 The tracer only traces the highest priority task in the system
2017 to avoid tracing the normal circumstances. He 1863 to avoid tracing the normal circumstances. Here we see that
2018 the kworker with a nice priority of -20 (not 1864 the kworker with a nice priority of -20 (not very nice), took
2019 just 15 microseconds from the time it woke up 1865 just 15 microseconds from the time it woke up, to the time it
2020 ran. 1866 ran.
2021 1867
2022 Non Real-Time tasks are not that interesting. 1868 Non Real-Time tasks are not that interesting. A more interesting
2023 trace is to concentrate only on Real-Time tas 1869 trace is to concentrate only on Real-Time tasks.
2024 1870
2025 wakeup_rt 1871 wakeup_rt
2026 --------- 1872 ---------
2027 1873
2028 In a Real-Time environment it is very importa 1874 In a Real-Time environment it is very important to know the
2029 wakeup time it takes for the highest priority 1875 wakeup time it takes for the highest priority task that is woken
2030 up to the time that it executes. This is also 1876 up to the time that it executes. This is also known as "schedule
2031 latency". I stress the point that this is abo 1877 latency". I stress the point that this is about RT tasks. It is
2032 also important to know the scheduling latency 1878 also important to know the scheduling latency of non-RT tasks,
2033 but the average schedule latency is better fo 1879 but the average schedule latency is better for non-RT tasks.
2034 Tools like LatencyTop are more appropriate fo 1880 Tools like LatencyTop are more appropriate for such
2035 measurements. 1881 measurements.
2036 1882
2037 Real-Time environments are interested in the 1883 Real-Time environments are interested in the worst case latency.
2038 That is the longest latency it takes for some 1884 That is the longest latency it takes for something to happen,
2039 and not the average. We can have a very fast 1885 and not the average. We can have a very fast scheduler that may
2040 only have a large latency once in a while, bu 1886 only have a large latency once in a while, but that would not
2041 work well with Real-Time tasks. The wakeup_r 1887 work well with Real-Time tasks. The wakeup_rt tracer was designed
2042 to record the worst case wakeups of RT tasks. 1888 to record the worst case wakeups of RT tasks. Non-RT tasks are
2043 not recorded because the tracer only records 1889 not recorded because the tracer only records one worst case and
2044 tracing non-RT tasks that are unpredictable w 1890 tracing non-RT tasks that are unpredictable will overwrite the
2045 worst case latency of RT tasks (just run the 1891 worst case latency of RT tasks (just run the normal wakeup
2046 tracer for a while to see that effect). 1892 tracer for a while to see that effect).
2047 1893
2048 Since this tracer only deals with RT tasks, w 1894 Since this tracer only deals with RT tasks, we will run this
2049 slightly differently than we did with the pre 1895 slightly differently than we did with the previous tracers.
2050 Instead of performing an 'ls', we will run 's 1896 Instead of performing an 'ls', we will run 'sleep 1' under
2051 'chrt' which changes the priority of the task 1897 'chrt' which changes the priority of the task.
2052 :: 1898 ::
2053 1899
2054 # echo 0 > options/function-trace 1900 # echo 0 > options/function-trace
2055 # echo wakeup_rt > current_tracer 1901 # echo wakeup_rt > current_tracer
2056 # echo 1 > tracing_on 1902 # echo 1 > tracing_on
2057 # echo 0 > tracing_max_latency 1903 # echo 0 > tracing_max_latency
2058 # chrt -f 5 sleep 1 1904 # chrt -f 5 sleep 1
2059 # echo 0 > tracing_on 1905 # echo 0 > tracing_on
2060 # cat trace 1906 # cat trace
2061 # tracer: wakeup 1907 # tracer: wakeup
2062 # 1908 #
2063 # tracer: wakeup_rt 1909 # tracer: wakeup_rt
2064 # 1910 #
2065 # wakeup_rt latency trace v1.1.5 on 3.8.0-t 1911 # wakeup_rt latency trace v1.1.5 on 3.8.0-test+
2066 # ----------------------------------------- 1912 # --------------------------------------------------------------------
2067 # latency: 5 us, #4/4, CPU#3 | (M:preempt V 1913 # latency: 5 us, #4/4, CPU#3 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
2068 # ----------------- 1914 # -----------------
2069 # | task: sleep-2389 (uid:0 nice:0 polic 1915 # | task: sleep-2389 (uid:0 nice:0 policy:1 rt_prio:5)
2070 # ----------------- 1916 # -----------------
2071 # 1917 #
2072 # _------=> CPU# 1918 # _------=> CPU#
2073 # / _-----=> irqs-off 1919 # / _-----=> irqs-off
2074 # | / _----=> need-resched 1920 # | / _----=> need-resched
2075 # || / _---=> hardirq/softir 1921 # || / _---=> hardirq/softirq
2076 # ||| / _--=> preempt-depth 1922 # ||| / _--=> preempt-depth
2077 # |||| / delay 1923 # |||| / delay
2078 # cmd pid ||||| time | caller 1924 # cmd pid ||||| time | caller
2079 # \ / ||||| \ | / 1925 # \ / ||||| \ | /
2080 <idle>-0 3d.h4 0us : 0:120: 1926 <idle>-0 3d.h4 0us : 0:120:R + [003] 2389: 94:R sleep
2081 <idle>-0 3d.h4 1us+: ttwu_do_act 1927 <idle>-0 3d.h4 1us+: ttwu_do_activate.constprop.87 <-try_to_wake_up
2082 <idle>-0 3d..3 5us : __schedule 1928 <idle>-0 3d..3 5us : __schedule <-schedule
2083 <idle>-0 3d..3 5us : 0:120: 1929 <idle>-0 3d..3 5us : 0:120:R ==> [003] 2389: 94:R sleep
2084 1930
2085 1931
2086 Running this on an idle system, we see that i 1932 Running this on an idle system, we see that it only took 5 microseconds
2087 to perform the task switch. Note, since the 1933 to perform the task switch. Note, since the trace point in the schedule
2088 is before the actual "switch", we stop the tr 1934 is before the actual "switch", we stop the tracing when the recorded task
2089 is about to schedule in. This may change if w 1935 is about to schedule in. This may change if we add a new marker at the
2090 end of the scheduler. 1936 end of the scheduler.
2091 1937
2092 Notice that the recorded task is 'sleep' with 1938 Notice that the recorded task is 'sleep' with the PID of 2389
2093 and it has an rt_prio of 5. This priority is 1939 and it has an rt_prio of 5. This priority is user-space priority
2094 and not the internal kernel priority. The pol 1940 and not the internal kernel priority. The policy is 1 for
2095 SCHED_FIFO and 2 for SCHED_RR. 1941 SCHED_FIFO and 2 for SCHED_RR.
2096 1942
2097 Note, that the trace data shows the internal 1943 Note, that the trace data shows the internal priority (99 - rtprio).
2098 :: 1944 ::
2099 1945
2100 <idle>-0 3d..3 5us : 0:120:R 1946 <idle>-0 3d..3 5us : 0:120:R ==> [003] 2389: 94:R sleep
2101 1947
2102 The 0:120:R means idle was running with a nic 1948 The 0:120:R means idle was running with a nice priority of 0 (120 - 120)
2103 and in the running state 'R'. The sleep task 1949 and in the running state 'R'. The sleep task was scheduled in with
2104 2389: 94:R. That is the priority is the kerne 1950 2389: 94:R. That is the priority is the kernel rtprio (99 - 5 = 94)
2105 and it too is in the running state. 1951 and it too is in the running state.
2106 1952
2107 Doing the same with chrt -r 5 and function-tr 1953 Doing the same with chrt -r 5 and function-trace set.
2108 :: 1954 ::
2109 1955
2110 echo 1 > options/function-trace 1956 echo 1 > options/function-trace
2111 1957
2112 # tracer: wakeup_rt 1958 # tracer: wakeup_rt
2113 # 1959 #
2114 # wakeup_rt latency trace v1.1.5 on 3.8.0-t 1960 # wakeup_rt latency trace v1.1.5 on 3.8.0-test+
2115 # ----------------------------------------- 1961 # --------------------------------------------------------------------
2116 # latency: 29 us, #85/85, CPU#3 | (M:preemp 1962 # latency: 29 us, #85/85, CPU#3 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
2117 # ----------------- 1963 # -----------------
2118 # | task: sleep-2448 (uid:0 nice:0 polic 1964 # | task: sleep-2448 (uid:0 nice:0 policy:1 rt_prio:5)
2119 # ----------------- 1965 # -----------------
2120 # 1966 #
2121 # _------=> CPU# 1967 # _------=> CPU#
2122 # / _-----=> irqs-off 1968 # / _-----=> irqs-off
2123 # | / _----=> need-resched 1969 # | / _----=> need-resched
2124 # || / _---=> hardirq/softir 1970 # || / _---=> hardirq/softirq
2125 # ||| / _--=> preempt-depth 1971 # ||| / _--=> preempt-depth
2126 # |||| / delay 1972 # |||| / delay
2127 # cmd pid ||||| time | caller 1973 # cmd pid ||||| time | caller
2128 # \ / ||||| \ | / 1974 # \ / ||||| \ | /
2129 <idle>-0 3d.h4 1us+: 0:120: 1975 <idle>-0 3d.h4 1us+: 0:120:R + [003] 2448: 94:R sleep
2130 <idle>-0 3d.h4 2us : ttwu_do_act 1976 <idle>-0 3d.h4 2us : ttwu_do_activate.constprop.87 <-try_to_wake_up
2131 <idle>-0 3d.h3 3us : check_preem 1977 <idle>-0 3d.h3 3us : check_preempt_curr <-ttwu_do_wakeup
2132 <idle>-0 3d.h3 3us : resched_cur 1978 <idle>-0 3d.h3 3us : resched_curr <-check_preempt_curr
2133 <idle>-0 3dNh3 4us : task_woken_ 1979 <idle>-0 3dNh3 4us : task_woken_rt <-ttwu_do_wakeup
2134 <idle>-0 3dNh3 4us : _raw_spin_u 1980 <idle>-0 3dNh3 4us : _raw_spin_unlock <-try_to_wake_up
2135 <idle>-0 3dNh3 4us : sub_preempt 1981 <idle>-0 3dNh3 4us : sub_preempt_count <-_raw_spin_unlock
2136 <idle>-0 3dNh2 5us : ttwu_stat < 1982 <idle>-0 3dNh2 5us : ttwu_stat <-try_to_wake_up
2137 <idle>-0 3dNh2 5us : _raw_spin_u 1983 <idle>-0 3dNh2 5us : _raw_spin_unlock_irqrestore <-try_to_wake_up
2138 <idle>-0 3dNh2 6us : sub_preempt 1984 <idle>-0 3dNh2 6us : sub_preempt_count <-_raw_spin_unlock_irqrestore
2139 <idle>-0 3dNh1 6us : _raw_spin_l 1985 <idle>-0 3dNh1 6us : _raw_spin_lock <-__run_hrtimer
2140 <idle>-0 3dNh1 6us : add_preempt 1986 <idle>-0 3dNh1 6us : add_preempt_count <-_raw_spin_lock
2141 <idle>-0 3dNh2 7us : _raw_spin_u 1987 <idle>-0 3dNh2 7us : _raw_spin_unlock <-hrtimer_interrupt
2142 <idle>-0 3dNh2 7us : sub_preempt 1988 <idle>-0 3dNh2 7us : sub_preempt_count <-_raw_spin_unlock
2143 <idle>-0 3dNh1 7us : tick_progra 1989 <idle>-0 3dNh1 7us : tick_program_event <-hrtimer_interrupt
2144 <idle>-0 3dNh1 7us : clockevents 1990 <idle>-0 3dNh1 7us : clockevents_program_event <-tick_program_event
2145 <idle>-0 3dNh1 8us : ktime_get < 1991 <idle>-0 3dNh1 8us : ktime_get <-clockevents_program_event
2146 <idle>-0 3dNh1 8us : lapic_next_ 1992 <idle>-0 3dNh1 8us : lapic_next_event <-clockevents_program_event
2147 <idle>-0 3dNh1 8us : irq_exit <- 1993 <idle>-0 3dNh1 8us : irq_exit <-smp_apic_timer_interrupt
2148 <idle>-0 3dNh1 9us : sub_preempt 1994 <idle>-0 3dNh1 9us : sub_preempt_count <-irq_exit
2149 <idle>-0 3dN.2 9us : idle_cpu <- 1995 <idle>-0 3dN.2 9us : idle_cpu <-irq_exit
2150 <idle>-0 3dN.2 9us : rcu_irq_exi 1996 <idle>-0 3dN.2 9us : rcu_irq_exit <-irq_exit
2151 <idle>-0 3dN.2 10us : rcu_eqs_ent 1997 <idle>-0 3dN.2 10us : rcu_eqs_enter_common.isra.45 <-rcu_irq_exit
2152 <idle>-0 3dN.2 10us : sub_preempt 1998 <idle>-0 3dN.2 10us : sub_preempt_count <-irq_exit
2153 <idle>-0 3.N.1 11us : rcu_idle_ex 1999 <idle>-0 3.N.1 11us : rcu_idle_exit <-cpu_idle
2154 <idle>-0 3dN.1 11us : rcu_eqs_exi 2000 <idle>-0 3dN.1 11us : rcu_eqs_exit_common.isra.43 <-rcu_idle_exit
2155 <idle>-0 3.N.1 11us : tick_nohz_i 2001 <idle>-0 3.N.1 11us : tick_nohz_idle_exit <-cpu_idle
2156 <idle>-0 3dN.1 12us : menu_hrtime 2002 <idle>-0 3dN.1 12us : menu_hrtimer_cancel <-tick_nohz_idle_exit
2157 <idle>-0 3dN.1 12us : ktime_get < 2003 <idle>-0 3dN.1 12us : ktime_get <-tick_nohz_idle_exit
2158 <idle>-0 3dN.1 12us : tick_do_upd 2004 <idle>-0 3dN.1 12us : tick_do_update_jiffies64 <-tick_nohz_idle_exit
2159 <idle>-0 3dN.1 13us : cpu_load_up 2005 <idle>-0 3dN.1 13us : cpu_load_update_nohz <-tick_nohz_idle_exit
2160 <idle>-0 3dN.1 13us : _raw_spin_l 2006 <idle>-0 3dN.1 13us : _raw_spin_lock <-cpu_load_update_nohz
2161 <idle>-0 3dN.1 13us : add_preempt 2007 <idle>-0 3dN.1 13us : add_preempt_count <-_raw_spin_lock
2162 <idle>-0 3dN.2 13us : __cpu_load_ 2008 <idle>-0 3dN.2 13us : __cpu_load_update <-cpu_load_update_nohz
2163 <idle>-0 3dN.2 14us : sched_avg_u 2009 <idle>-0 3dN.2 14us : sched_avg_update <-__cpu_load_update
2164 <idle>-0 3dN.2 14us : _raw_spin_u 2010 <idle>-0 3dN.2 14us : _raw_spin_unlock <-cpu_load_update_nohz
2165 <idle>-0 3dN.2 14us : sub_preempt 2011 <idle>-0 3dN.2 14us : sub_preempt_count <-_raw_spin_unlock
2166 <idle>-0 3dN.1 15us : calc_load_n 2012 <idle>-0 3dN.1 15us : calc_load_nohz_stop <-tick_nohz_idle_exit
2167 <idle>-0 3dN.1 15us : touch_softl 2013 <idle>-0 3dN.1 15us : touch_softlockup_watchdog <-tick_nohz_idle_exit
2168 <idle>-0 3dN.1 15us : hrtimer_can 2014 <idle>-0 3dN.1 15us : hrtimer_cancel <-tick_nohz_idle_exit
2169 <idle>-0 3dN.1 15us : hrtimer_try 2015 <idle>-0 3dN.1 15us : hrtimer_try_to_cancel <-hrtimer_cancel
2170 <idle>-0 3dN.1 16us : lock_hrtime 2016 <idle>-0 3dN.1 16us : lock_hrtimer_base.isra.18 <-hrtimer_try_to_cancel
2171 <idle>-0 3dN.1 16us : _raw_spin_l 2017 <idle>-0 3dN.1 16us : _raw_spin_lock_irqsave <-lock_hrtimer_base.isra.18
2172 <idle>-0 3dN.1 16us : add_preempt 2018 <idle>-0 3dN.1 16us : add_preempt_count <-_raw_spin_lock_irqsave
2173 <idle>-0 3dN.2 17us : __remove_hr 2019 <idle>-0 3dN.2 17us : __remove_hrtimer <-remove_hrtimer.part.16
2174 <idle>-0 3dN.2 17us : hrtimer_for 2020 <idle>-0 3dN.2 17us : hrtimer_force_reprogram <-__remove_hrtimer
2175 <idle>-0 3dN.2 17us : tick_progra 2021 <idle>-0 3dN.2 17us : tick_program_event <-hrtimer_force_reprogram
2176 <idle>-0 3dN.2 18us : clockevents 2022 <idle>-0 3dN.2 18us : clockevents_program_event <-tick_program_event
2177 <idle>-0 3dN.2 18us : ktime_get < 2023 <idle>-0 3dN.2 18us : ktime_get <-clockevents_program_event
2178 <idle>-0 3dN.2 18us : lapic_next_ 2024 <idle>-0 3dN.2 18us : lapic_next_event <-clockevents_program_event
2179 <idle>-0 3dN.2 19us : _raw_spin_u 2025 <idle>-0 3dN.2 19us : _raw_spin_unlock_irqrestore <-hrtimer_try_to_cancel
2180 <idle>-0 3dN.2 19us : sub_preempt 2026 <idle>-0 3dN.2 19us : sub_preempt_count <-_raw_spin_unlock_irqrestore
2181 <idle>-0 3dN.1 19us : hrtimer_for 2027 <idle>-0 3dN.1 19us : hrtimer_forward <-tick_nohz_idle_exit
2182 <idle>-0 3dN.1 20us : ktime_add_s 2028 <idle>-0 3dN.1 20us : ktime_add_safe <-hrtimer_forward
2183 <idle>-0 3dN.1 20us : ktime_add_s 2029 <idle>-0 3dN.1 20us : ktime_add_safe <-hrtimer_forward
2184 <idle>-0 3dN.1 20us : hrtimer_sta 2030 <idle>-0 3dN.1 20us : hrtimer_start_range_ns <-hrtimer_start_expires.constprop.11
2185 <idle>-0 3dN.1 20us : __hrtimer_s 2031 <idle>-0 3dN.1 20us : __hrtimer_start_range_ns <-hrtimer_start_range_ns
2186 <idle>-0 3dN.1 21us : lock_hrtime 2032 <idle>-0 3dN.1 21us : lock_hrtimer_base.isra.18 <-__hrtimer_start_range_ns
2187 <idle>-0 3dN.1 21us : _raw_spin_l 2033 <idle>-0 3dN.1 21us : _raw_spin_lock_irqsave <-lock_hrtimer_base.isra.18
2188 <idle>-0 3dN.1 21us : add_preempt 2034 <idle>-0 3dN.1 21us : add_preempt_count <-_raw_spin_lock_irqsave
2189 <idle>-0 3dN.2 22us : ktime_add_s 2035 <idle>-0 3dN.2 22us : ktime_add_safe <-__hrtimer_start_range_ns
2190 <idle>-0 3dN.2 22us : enqueue_hrt 2036 <idle>-0 3dN.2 22us : enqueue_hrtimer <-__hrtimer_start_range_ns
2191 <idle>-0 3dN.2 22us : tick_progra 2037 <idle>-0 3dN.2 22us : tick_program_event <-__hrtimer_start_range_ns
2192 <idle>-0 3dN.2 23us : clockevents 2038 <idle>-0 3dN.2 23us : clockevents_program_event <-tick_program_event
2193 <idle>-0 3dN.2 23us : ktime_get < 2039 <idle>-0 3dN.2 23us : ktime_get <-clockevents_program_event
2194 <idle>-0 3dN.2 23us : lapic_next_ 2040 <idle>-0 3dN.2 23us : lapic_next_event <-clockevents_program_event
2195 <idle>-0 3dN.2 24us : _raw_spin_u 2041 <idle>-0 3dN.2 24us : _raw_spin_unlock_irqrestore <-__hrtimer_start_range_ns
2196 <idle>-0 3dN.2 24us : sub_preempt 2042 <idle>-0 3dN.2 24us : sub_preempt_count <-_raw_spin_unlock_irqrestore
2197 <idle>-0 3dN.1 24us : account_idl 2043 <idle>-0 3dN.1 24us : account_idle_ticks <-tick_nohz_idle_exit
2198 <idle>-0 3dN.1 24us : account_idl 2044 <idle>-0 3dN.1 24us : account_idle_time <-account_idle_ticks
2199 <idle>-0 3.N.1 25us : sub_preempt 2045 <idle>-0 3.N.1 25us : sub_preempt_count <-cpu_idle
2200 <idle>-0 3.N.. 25us : schedule <- 2046 <idle>-0 3.N.. 25us : schedule <-cpu_idle
2201 <idle>-0 3.N.. 25us : __schedule 2047 <idle>-0 3.N.. 25us : __schedule <-preempt_schedule
2202 <idle>-0 3.N.. 26us : add_preempt 2048 <idle>-0 3.N.. 26us : add_preempt_count <-__schedule
2203 <idle>-0 3.N.1 26us : rcu_note_co 2049 <idle>-0 3.N.1 26us : rcu_note_context_switch <-__schedule
2204 <idle>-0 3.N.1 26us : rcu_sched_q 2050 <idle>-0 3.N.1 26us : rcu_sched_qs <-rcu_note_context_switch
2205 <idle>-0 3dN.1 27us : rcu_preempt 2051 <idle>-0 3dN.1 27us : rcu_preempt_qs <-rcu_note_context_switch
2206 <idle>-0 3.N.1 27us : _raw_spin_l 2052 <idle>-0 3.N.1 27us : _raw_spin_lock_irq <-__schedule
2207 <idle>-0 3dN.1 27us : add_preempt 2053 <idle>-0 3dN.1 27us : add_preempt_count <-_raw_spin_lock_irq
2208 <idle>-0 3dN.2 28us : put_prev_ta 2054 <idle>-0 3dN.2 28us : put_prev_task_idle <-__schedule
2209 <idle>-0 3dN.2 28us : pick_next_t 2055 <idle>-0 3dN.2 28us : pick_next_task_stop <-pick_next_task
2210 <idle>-0 3dN.2 28us : pick_next_t 2056 <idle>-0 3dN.2 28us : pick_next_task_rt <-pick_next_task
2211 <idle>-0 3dN.2 29us : dequeue_pus 2057 <idle>-0 3dN.2 29us : dequeue_pushable_task <-pick_next_task_rt
2212 <idle>-0 3d..3 29us : __schedule 2058 <idle>-0 3d..3 29us : __schedule <-preempt_schedule
2213 <idle>-0 3d..3 30us : 0:120: 2059 <idle>-0 3d..3 30us : 0:120:R ==> [003] 2448: 94:R sleep
2214 2060
2215 This isn't that big of a trace, even with fun 2061 This isn't that big of a trace, even with function tracing enabled,
2216 so I included the entire trace. 2062 so I included the entire trace.
2217 2063
2218 The interrupt went off while when the system 2064 The interrupt went off while when the system was idle. Somewhere
2219 before task_woken_rt() was called, the NEED_R 2065 before task_woken_rt() was called, the NEED_RESCHED flag was set,
2220 this is indicated by the first occurrence of 2066 this is indicated by the first occurrence of the 'N' flag.
2221 2067
2222 Latency tracing and events 2068 Latency tracing and events
2223 -------------------------- 2069 --------------------------
2224 As function tracing can induce a much larger 2070 As function tracing can induce a much larger latency, but without
2225 seeing what happens within the latency it is 2071 seeing what happens within the latency it is hard to know what
2226 caused it. There is a middle ground, and that 2072 caused it. There is a middle ground, and that is with enabling
2227 events. 2073 events.
2228 :: 2074 ::
2229 2075
2230 # echo 0 > options/function-trace 2076 # echo 0 > options/function-trace
2231 # echo wakeup_rt > current_tracer 2077 # echo wakeup_rt > current_tracer
2232 # echo 1 > events/enable 2078 # echo 1 > events/enable
2233 # echo 1 > tracing_on 2079 # echo 1 > tracing_on
2234 # echo 0 > tracing_max_latency 2080 # echo 0 > tracing_max_latency
2235 # chrt -f 5 sleep 1 2081 # chrt -f 5 sleep 1
2236 # echo 0 > tracing_on 2082 # echo 0 > tracing_on
2237 # cat trace 2083 # cat trace
2238 # tracer: wakeup_rt 2084 # tracer: wakeup_rt
2239 # 2085 #
2240 # wakeup_rt latency trace v1.1.5 on 3.8.0-t 2086 # wakeup_rt latency trace v1.1.5 on 3.8.0-test+
2241 # ----------------------------------------- 2087 # --------------------------------------------------------------------
2242 # latency: 6 us, #12/12, CPU#2 | (M:preempt 2088 # latency: 6 us, #12/12, CPU#2 | (M:preempt VP:0, KP:0, SP:0 HP:0 #P:4)
2243 # ----------------- 2089 # -----------------
2244 # | task: sleep-5882 (uid:0 nice:0 polic 2090 # | task: sleep-5882 (uid:0 nice:0 policy:1 rt_prio:5)
2245 # ----------------- 2091 # -----------------
2246 # 2092 #
2247 # _------=> CPU# 2093 # _------=> CPU#
2248 # / _-----=> irqs-off 2094 # / _-----=> irqs-off
2249 # | / _----=> need-resched 2095 # | / _----=> need-resched
2250 # || / _---=> hardirq/softir 2096 # || / _---=> hardirq/softirq
2251 # ||| / _--=> preempt-depth 2097 # ||| / _--=> preempt-depth
2252 # |||| / delay 2098 # |||| / delay
2253 # cmd pid ||||| time | caller 2099 # cmd pid ||||| time | caller
2254 # \ / ||||| \ | / 2100 # \ / ||||| \ | /
2255 <idle>-0 2d.h4 0us : 0:120: 2101 <idle>-0 2d.h4 0us : 0:120:R + [002] 5882: 94:R sleep
2256 <idle>-0 2d.h4 0us : ttwu_do_act 2102 <idle>-0 2d.h4 0us : ttwu_do_activate.constprop.87 <-try_to_wake_up
2257 <idle>-0 2d.h4 1us : sched_wakeu 2103 <idle>-0 2d.h4 1us : sched_wakeup: comm=sleep pid=5882 prio=94 success=1 target_cpu=002
2258 <idle>-0 2dNh2 1us : hrtimer_exp 2104 <idle>-0 2dNh2 1us : hrtimer_expire_exit: hrtimer=ffff88007796feb8
2259 <idle>-0 2.N.2 2us : power_end: 2105 <idle>-0 2.N.2 2us : power_end: cpu_id=2
2260 <idle>-0 2.N.2 3us : cpu_idle: s 2106 <idle>-0 2.N.2 3us : cpu_idle: state=4294967295 cpu_id=2
2261 <idle>-0 2dN.3 4us : hrtimer_can 2107 <idle>-0 2dN.3 4us : hrtimer_cancel: hrtimer=ffff88007d50d5e0
2262 <idle>-0 2dN.3 4us : hrtimer_sta 2108 <idle>-0 2dN.3 4us : hrtimer_start: hrtimer=ffff88007d50d5e0 function=tick_sched_timer expires=34311211000000 softexpires=34311211000000
2263 <idle>-0 2.N.2 5us : rcu_utiliza 2109 <idle>-0 2.N.2 5us : rcu_utilization: Start context switch
2264 <idle>-0 2.N.2 5us : rcu_utiliza 2110 <idle>-0 2.N.2 5us : rcu_utilization: End context switch
2265 <idle>-0 2d..3 6us : __schedule 2111 <idle>-0 2d..3 6us : __schedule <-schedule
2266 <idle>-0 2d..3 6us : 0:120: 2112 <idle>-0 2d..3 6us : 0:120:R ==> [002] 5882: 94:R sleep
2267 2113
2268 2114
2269 Hardware Latency Detector 2115 Hardware Latency Detector
2270 ------------------------- 2116 -------------------------
2271 2117
2272 The hardware latency detector is executed by 2118 The hardware latency detector is executed by enabling the "hwlat" tracer.
2273 2119
2274 NOTE, this tracer will affect the performance 2120 NOTE, this tracer will affect the performance of the system as it will
2275 periodically make a CPU constantly busy with 2121 periodically make a CPU constantly busy with interrupts disabled.
2276 :: 2122 ::
2277 2123
2278 # echo hwlat > current_tracer 2124 # echo hwlat > current_tracer
2279 # sleep 100 2125 # sleep 100
2280 # cat trace 2126 # cat trace
2281 # tracer: hwlat 2127 # tracer: hwlat
2282 # 2128 #
2283 # entries-in-buffer/entries-written: 13/13 <<
2284 # <<
2285 # _-----=> irq 2129 # _-----=> irqs-off
2286 # / _----=> nee 2130 # / _----=> need-resched
2287 # | / _---=> har 2131 # | / _---=> hardirq/softirq
2288 # || / _--=> pre 2132 # || / _--=> preempt-depth
2289 # ||| / dela 2133 # ||| / delay
2290 # TASK-PID CPU# |||| TIMEST 2134 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
2291 # | | | |||| | 2135 # | | | |||| | |
2292 <...>-1729 [001] d... 678.473 !! 2136 <...>-3638 [001] d... 19452.055471: #1 inner/outer(us): 12/14 ts:1499801089.066141940
2293 <...>-1729 [004] d... 689.556 !! 2137 <...>-3638 [003] d... 19454.071354: #2 inner/outer(us): 11/9 ts:1499801091.082164365
2294 <...>-1729 [005] d... 714.756 !! 2138 <...>-3638 [002] dn.. 19461.126852: #3 inner/outer(us): 12/9 ts:1499801098.138150062
2295 <...>-1729 [001] d... 718.788 !! 2139 <...>-3638 [001] d... 19488.340960: #4 inner/outer(us): 8/12 ts:1499801125.354139633
2296 <...>-1729 [002] d... 719.796 !! 2140 <...>-3638 [003] d... 19494.388553: #5 inner/outer(us): 8/12 ts:1499801131.402150961
2297 <...>-1729 [006] d... 844.787 !! 2141 <...>-3638 [003] d... 19501.283419: #6 inner/outer(us): 0/12 ts:1499801138.297435289 nmi-total:4 nmi-count:1
2298 <...>-1729 [003] d... 849.827 <<
2299 <...>-1729 [007] d... 853.859 <<
2300 <...>-1729 [001] d... 855.874 <<
2301 <...>-1729 [001] d... 863.938 <<
2302 <...>-1729 [007] d... 878.050 <<
2303 <...>-1729 [007] d... 886.114 <<
2304 2142
2305 2143
2306 The above output is somewhat the same in the 2144 The above output is somewhat the same in the header. All events will have
2307 interrupts disabled 'd'. Under the FUNCTION t 2145 interrupts disabled 'd'. Under the FUNCTION title there is:
2308 2146
2309 #1 2147 #1
2310 This is the count of events recorded 2148 This is the count of events recorded that were greater than the
2311 tracing_threshold (See below). 2149 tracing_threshold (See below).
2312 2150
2313 inner/outer(us): 11/11 !! 2151 inner/outer(us): 12/14
2314 2152
2315 This shows two numbers as "inner latenc 2153 This shows two numbers as "inner latency" and "outer latency". The test
2316 runs in a loop checking a timestamp twi 2154 runs in a loop checking a timestamp twice. The latency detected within
2317 the two timestamps is the "inner latenc 2155 the two timestamps is the "inner latency" and the latency detected
2318 after the previous timestamp and the ne 2156 after the previous timestamp and the next timestamp in the loop is
2319 the "outer latency". 2157 the "outer latency".
2320 2158
2321 ts:1581527483.343962693 !! 2159 ts:1499801089.066141940
2322 2160
2323 The absolute timestamp that the first l !! 2161 The absolute timestamp that the event happened.
2324 2162
2325 count:6 !! 2163 nmi-total:4 nmi-count:1
2326 <<
2327 The number of times a latency was detec <<
2328 <<
2329 nmi-total:7 nmi-count:1 <<
2330 2164
2331 On architectures that support it, if an 2165 On architectures that support it, if an NMI comes in during the
2332 test, the time spent in NMI is reported 2166 test, the time spent in NMI is reported in "nmi-total" (in
2333 microseconds). 2167 microseconds).
2334 2168
2335 All architectures that have NMIs will s 2169 All architectures that have NMIs will show the "nmi-count" if an
2336 NMI comes in during the test. 2170 NMI comes in during the test.
2337 2171
2338 hwlat files: 2172 hwlat files:
2339 2173
2340 tracing_threshold 2174 tracing_threshold
2341 This gets automatically set to "10" t 2175 This gets automatically set to "10" to represent 10
2342 microseconds. This is the threshold o 2176 microseconds. This is the threshold of latency that
2343 needs to be detected before the trace 2177 needs to be detected before the trace will be recorded.
2344 2178
2345 Note, when hwlat tracer is finished ( 2179 Note, when hwlat tracer is finished (another tracer is
2346 written into "current_tracer"), the o 2180 written into "current_tracer"), the original value for
2347 tracing_threshold is placed back into 2181 tracing_threshold is placed back into this file.
2348 2182
2349 hwlat_detector/width 2183 hwlat_detector/width
2350 The length of time the test runs with 2184 The length of time the test runs with interrupts disabled.
2351 2185
2352 hwlat_detector/window 2186 hwlat_detector/window
2353 The length of time of the window whic 2187 The length of time of the window which the test
2354 runs. That is, the test will run for 2188 runs. That is, the test will run for "width"
2355 microseconds per "window" microsecond 2189 microseconds per "window" microseconds
2356 2190
2357 tracing_cpumask 2191 tracing_cpumask
2358 When the test is started. A kernel th 2192 When the test is started. A kernel thread is created that
2359 runs the test. This thread will alter 2193 runs the test. This thread will alternate between CPUs
2360 listed in the tracing_cpumask between 2194 listed in the tracing_cpumask between each period
2361 (one "window"). To limit the test to 2195 (one "window"). To limit the test to specific CPUs
2362 set the mask in this file to only the 2196 set the mask in this file to only the CPUs that the test
2363 should run on. 2197 should run on.
2364 2198
2365 function 2199 function
2366 -------- 2200 --------
2367 2201
2368 This tracer is the function tracer. Enabling 2202 This tracer is the function tracer. Enabling the function tracer
2369 can be done from the debug file system. Make 2203 can be done from the debug file system. Make sure the
2370 ftrace_enabled is set; otherwise this tracer 2204 ftrace_enabled is set; otherwise this tracer is a nop.
2371 See the "ftrace_enabled" section below. 2205 See the "ftrace_enabled" section below.
2372 :: 2206 ::
2373 2207
2374 # sysctl kernel.ftrace_enabled=1 2208 # sysctl kernel.ftrace_enabled=1
2375 # echo function > current_tracer 2209 # echo function > current_tracer
2376 # echo 1 > tracing_on 2210 # echo 1 > tracing_on
2377 # usleep 1 2211 # usleep 1
2378 # echo 0 > tracing_on 2212 # echo 0 > tracing_on
2379 # cat trace 2213 # cat trace
2380 # tracer: function 2214 # tracer: function
2381 # 2215 #
2382 # entries-in-buffer/entries-written: 24799/ 2216 # entries-in-buffer/entries-written: 24799/24799 #P:4
2383 # 2217 #
2384 # _-----=> irq 2218 # _-----=> irqs-off
2385 # / _----=> nee 2219 # / _----=> need-resched
2386 # | / _---=> har 2220 # | / _---=> hardirq/softirq
2387 # || / _--=> pre 2221 # || / _--=> preempt-depth
2388 # ||| / dela 2222 # ||| / delay
2389 # TASK-PID CPU# |||| TIMEST 2223 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
2390 # | | | |||| | 2224 # | | | |||| | |
2391 bash-1994 [002] .... 3082.063 2225 bash-1994 [002] .... 3082.063030: mutex_unlock <-rb_simple_write
2392 bash-1994 [002] .... 3082.063 2226 bash-1994 [002] .... 3082.063031: __mutex_unlock_slowpath <-mutex_unlock
2393 bash-1994 [002] .... 3082.063 2227 bash-1994 [002] .... 3082.063031: __fsnotify_parent <-fsnotify_modify
2394 bash-1994 [002] .... 3082.063 2228 bash-1994 [002] .... 3082.063032: fsnotify <-fsnotify_modify
2395 bash-1994 [002] .... 3082.063 2229 bash-1994 [002] .... 3082.063032: __srcu_read_lock <-fsnotify
2396 bash-1994 [002] .... 3082.063 2230 bash-1994 [002] .... 3082.063032: add_preempt_count <-__srcu_read_lock
2397 bash-1994 [002] ...1 3082.063 2231 bash-1994 [002] ...1 3082.063032: sub_preempt_count <-__srcu_read_lock
2398 bash-1994 [002] .... 3082.063 2232 bash-1994 [002] .... 3082.063033: __srcu_read_unlock <-fsnotify
2399 [...] 2233 [...]
2400 2234
2401 2235
2402 Note: function tracer uses ring buffers to st 2236 Note: function tracer uses ring buffers to store the above
2403 entries. The newest data may overwrite the ol 2237 entries. The newest data may overwrite the oldest data.
2404 Sometimes using echo to stop the trace is not 2238 Sometimes using echo to stop the trace is not sufficient because
2405 the tracing could have overwritten the data t 2239 the tracing could have overwritten the data that you wanted to
2406 record. For this reason, it is sometimes bett 2240 record. For this reason, it is sometimes better to disable
2407 tracing directly from a program. This allows 2241 tracing directly from a program. This allows you to stop the
2408 tracing at the point that you hit the part th 2242 tracing at the point that you hit the part that you are
2409 interested in. To disable the tracing directl 2243 interested in. To disable the tracing directly from a C program,
2410 something like following code snippet can be 2244 something like following code snippet can be used::
2411 2245
2412 int trace_fd; 2246 int trace_fd;
2413 [...] 2247 [...]
2414 int main(int argc, char *argv[]) { 2248 int main(int argc, char *argv[]) {
2415 [...] 2249 [...]
2416 trace_fd = open(tracing_file( 2250 trace_fd = open(tracing_file("tracing_on"), O_WRONLY);
2417 [...] 2251 [...]
2418 if (condition_hit()) { 2252 if (condition_hit()) {
2419 write(trace_fd, "0", 2253 write(trace_fd, "0", 1);
2420 } 2254 }
2421 [...] 2255 [...]
2422 } 2256 }
2423 2257
2424 2258
2425 Single thread tracing 2259 Single thread tracing
2426 --------------------- 2260 ---------------------
2427 2261
2428 By writing into set_ftrace_pid you can trace 2262 By writing into set_ftrace_pid you can trace a
2429 single thread. For example:: 2263 single thread. For example::
2430 2264
2431 # cat set_ftrace_pid 2265 # cat set_ftrace_pid
2432 no pid 2266 no pid
2433 # echo 3111 > set_ftrace_pid 2267 # echo 3111 > set_ftrace_pid
2434 # cat set_ftrace_pid 2268 # cat set_ftrace_pid
2435 3111 2269 3111
2436 # echo function > current_tracer 2270 # echo function > current_tracer
2437 # cat trace | head 2271 # cat trace | head
2438 # tracer: function 2272 # tracer: function
2439 # 2273 #
2440 # TASK-PID CPU# TIMESTAMP 2274 # TASK-PID CPU# TIMESTAMP FUNCTION
2441 # | | | | 2275 # | | | | |
2442 yum-updatesd-3111 [003] 1637.254676: 2276 yum-updatesd-3111 [003] 1637.254676: finish_task_switch <-thread_return
2443 yum-updatesd-3111 [003] 1637.254681: 2277 yum-updatesd-3111 [003] 1637.254681: hrtimer_cancel <-schedule_hrtimeout_range
2444 yum-updatesd-3111 [003] 1637.254682: 2278 yum-updatesd-3111 [003] 1637.254682: hrtimer_try_to_cancel <-hrtimer_cancel
2445 yum-updatesd-3111 [003] 1637.254683: 2279 yum-updatesd-3111 [003] 1637.254683: lock_hrtimer_base <-hrtimer_try_to_cancel
2446 yum-updatesd-3111 [003] 1637.254685: 2280 yum-updatesd-3111 [003] 1637.254685: fget_light <-do_sys_poll
2447 yum-updatesd-3111 [003] 1637.254686: 2281 yum-updatesd-3111 [003] 1637.254686: pipe_poll <-do_sys_poll
2448 # echo > set_ftrace_pid 2282 # echo > set_ftrace_pid
2449 # cat trace |head 2283 # cat trace |head
2450 # tracer: function 2284 # tracer: function
2451 # 2285 #
2452 # TASK-PID CPU# TIMESTAMP 2286 # TASK-PID CPU# TIMESTAMP FUNCTION
2453 # | | | | 2287 # | | | | |
2454 ##### CPU 3 buffer started #### 2288 ##### CPU 3 buffer started ####
2455 yum-updatesd-3111 [003] 1701.957688: 2289 yum-updatesd-3111 [003] 1701.957688: free_poll_entry <-poll_freewait
2456 yum-updatesd-3111 [003] 1701.957689: 2290 yum-updatesd-3111 [003] 1701.957689: remove_wait_queue <-free_poll_entry
2457 yum-updatesd-3111 [003] 1701.957691: 2291 yum-updatesd-3111 [003] 1701.957691: fput <-free_poll_entry
2458 yum-updatesd-3111 [003] 1701.957692: 2292 yum-updatesd-3111 [003] 1701.957692: audit_syscall_exit <-sysret_audit
2459 yum-updatesd-3111 [003] 1701.957693: 2293 yum-updatesd-3111 [003] 1701.957693: path_put <-audit_syscall_exit
2460 2294
2461 If you want to trace a function when executin 2295 If you want to trace a function when executing, you could use
2462 something like this simple program. 2296 something like this simple program.
2463 :: 2297 ::
2464 2298
2465 #include <stdio.h> 2299 #include <stdio.h>
2466 #include <stdlib.h> 2300 #include <stdlib.h>
2467 #include <sys/types.h> 2301 #include <sys/types.h>
2468 #include <sys/stat.h> 2302 #include <sys/stat.h>
2469 #include <fcntl.h> 2303 #include <fcntl.h>
2470 #include <unistd.h> 2304 #include <unistd.h>
2471 #include <string.h> 2305 #include <string.h>
2472 2306
2473 #define _STR(x) #x 2307 #define _STR(x) #x
2474 #define STR(x) _STR(x) 2308 #define STR(x) _STR(x)
2475 #define MAX_PATH 256 2309 #define MAX_PATH 256
2476 2310
2477 const char *find_tracefs(void) 2311 const char *find_tracefs(void)
2478 { 2312 {
2479 static char tracefs[MAX_PATH+1 2313 static char tracefs[MAX_PATH+1];
2480 static int tracefs_found; 2314 static int tracefs_found;
2481 char type[100]; 2315 char type[100];
2482 FILE *fp; 2316 FILE *fp;
2483 2317
2484 if (tracefs_found) 2318 if (tracefs_found)
2485 return tracefs; 2319 return tracefs;
2486 2320
2487 if ((fp = fopen("/proc/mounts" 2321 if ((fp = fopen("/proc/mounts","r")) == NULL) {
2488 perror("/proc/mounts") 2322 perror("/proc/mounts");
2489 return NULL; 2323 return NULL;
2490 } 2324 }
2491 2325
2492 while (fscanf(fp, "%*s %" 2326 while (fscanf(fp, "%*s %"
2493 STR(MAX_PATH) 2327 STR(MAX_PATH)
2494 "s %99s %*s %*d 2328 "s %99s %*s %*d %*d\n",
2495 tracefs, type) = 2329 tracefs, type) == 2) {
2496 if (strcmp(type, "trac 2330 if (strcmp(type, "tracefs") == 0)
2497 break; 2331 break;
2498 } 2332 }
2499 fclose(fp); 2333 fclose(fp);
2500 2334
2501 if (strcmp(type, "tracefs") != 2335 if (strcmp(type, "tracefs") != 0) {
2502 fprintf(stderr, "trace 2336 fprintf(stderr, "tracefs not mounted");
2503 return NULL; 2337 return NULL;
2504 } 2338 }
2505 2339
2506 strcat(tracefs, "/tracing/"); 2340 strcat(tracefs, "/tracing/");
2507 tracefs_found = 1; 2341 tracefs_found = 1;
2508 2342
2509 return tracefs; 2343 return tracefs;
2510 } 2344 }
2511 2345
2512 const char *tracing_file(const char * 2346 const char *tracing_file(const char *file_name)
2513 { 2347 {
2514 static char trace_file[MAX_PAT 2348 static char trace_file[MAX_PATH+1];
2515 snprintf(trace_file, MAX_PATH, 2349 snprintf(trace_file, MAX_PATH, "%s/%s", find_tracefs(), file_name);
2516 return trace_file; 2350 return trace_file;
2517 } 2351 }
2518 2352
2519 int main (int argc, char **argv) 2353 int main (int argc, char **argv)
2520 { 2354 {
2521 if (argc < 1) 2355 if (argc < 1)
2522 exit(-1); 2356 exit(-1);
2523 2357
2524 if (fork() > 0) { 2358 if (fork() > 0) {
2525 int fd, ffd; 2359 int fd, ffd;
2526 char line[64]; 2360 char line[64];
2527 int s; 2361 int s;
2528 2362
2529 ffd = open(tracing_fi 2363 ffd = open(tracing_file("current_tracer"), O_WRONLY);
2530 if (ffd < 0) 2364 if (ffd < 0)
2531 exit(-1); 2365 exit(-1);
2532 write(ffd, "nop", 3); 2366 write(ffd, "nop", 3);
2533 2367
2534 fd = open(tracing_fil 2368 fd = open(tracing_file("set_ftrace_pid"), O_WRONLY);
2535 s = sprintf(line, "%d 2369 s = sprintf(line, "%d\n", getpid());
2536 write(fd, line, s); 2370 write(fd, line, s);
2537 2371
2538 write(ffd, "function" 2372 write(ffd, "function", 8);
2539 2373
2540 close(fd); 2374 close(fd);
2541 close(ffd); 2375 close(ffd);
2542 2376
2543 execvp(argv[1], argv+ 2377 execvp(argv[1], argv+1);
2544 } 2378 }
2545 2379
2546 return 0; 2380 return 0;
2547 } 2381 }
2548 2382
2549 Or this simple script! 2383 Or this simple script!
2550 :: 2384 ::
2551 2385
2552 #!/bin/bash 2386 #!/bin/bash
2553 2387
2554 tracefs=`sed -ne 's/^tracefs \(.*\) tracefs 2388 tracefs=`sed -ne 's/^tracefs \(.*\) tracefs.*/\1/p' /proc/mounts`
2555 echo 0 > $tracefs/tracing_on !! 2389 echo nop > $tracefs/tracing/current_tracer
2556 echo $$ > $tracefs/set_ftrace_pid !! 2390 echo 0 > $tracefs/tracing/tracing_on
2557 echo function > $tracefs/current_tracer !! 2391 echo $$ > $tracefs/tracing/set_ftrace_pid
2558 echo 1 > $tracefs/tracing_on !! 2392 echo function > $tracefs/tracing/current_tracer
>> 2393 echo 1 > $tracefs/tracing/tracing_on
2559 exec "$@" 2394 exec "$@"
2560 2395
2561 2396
2562 function graph tracer 2397 function graph tracer
2563 --------------------------- 2398 ---------------------------
2564 2399
2565 This tracer is similar to the function tracer 2400 This tracer is similar to the function tracer except that it
2566 probes a function on its entry and its exit. 2401 probes a function on its entry and its exit. This is done by
2567 using a dynamically allocated stack of return 2402 using a dynamically allocated stack of return addresses in each
2568 task_struct. On function entry the tracer ove 2403 task_struct. On function entry the tracer overwrites the return
2569 address of each function traced to set a cust 2404 address of each function traced to set a custom probe. Thus the
2570 original return address is stored on the stac 2405 original return address is stored on the stack of return address
2571 in the task_struct. 2406 in the task_struct.
2572 2407
2573 Probing on both ends of a function leads to s 2408 Probing on both ends of a function leads to special features
2574 such as: 2409 such as:
2575 2410
2576 - measure of a function's time execution 2411 - measure of a function's time execution
2577 - having a reliable call stack to draw functi 2412 - having a reliable call stack to draw function calls graph
2578 2413
2579 This tracer is useful in several situations: 2414 This tracer is useful in several situations:
2580 2415
2581 - you want to find the reason of a strange ke 2416 - you want to find the reason of a strange kernel behavior and
2582 need to see what happens in detail on any a 2417 need to see what happens in detail on any areas (or specific
2583 ones). 2418 ones).
2584 2419
2585 - you are experiencing weird latencies but it 2420 - you are experiencing weird latencies but it's difficult to
2586 find its origin. 2421 find its origin.
2587 2422
2588 - you want to find quickly which path is take 2423 - you want to find quickly which path is taken by a specific
2589 function 2424 function
2590 2425
2591 - you just want to peek inside a working kern 2426 - you just want to peek inside a working kernel and want to see
2592 what happens there. 2427 what happens there.
2593 2428
2594 :: 2429 ::
2595 2430
2596 # tracer: function_graph 2431 # tracer: function_graph
2597 # 2432 #
2598 # CPU DURATION FUNCTION C 2433 # CPU DURATION FUNCTION CALLS
2599 # | | | | | | 2434 # | | | | | | |
2600 2435
2601 0) | sys_open() { 2436 0) | sys_open() {
2602 0) | do_sys_open() { 2437 0) | do_sys_open() {
2603 0) | getname() { 2438 0) | getname() {
2604 0) | kmem_cache_alloc 2439 0) | kmem_cache_alloc() {
2605 0) 1.382 us | __might_sleep( 2440 0) 1.382 us | __might_sleep();
2606 0) 2.478 us | } 2441 0) 2.478 us | }
2607 0) | strncpy_from_use 2442 0) | strncpy_from_user() {
2608 0) | might_fault() 2443 0) | might_fault() {
2609 0) 1.389 us | __might_slee 2444 0) 1.389 us | __might_sleep();
2610 0) 2.553 us | } 2445 0) 2.553 us | }
2611 0) 3.807 us | } 2446 0) 3.807 us | }
2612 0) 7.876 us | } 2447 0) 7.876 us | }
2613 0) | alloc_fd() { 2448 0) | alloc_fd() {
2614 0) 0.668 us | _spin_lock(); 2449 0) 0.668 us | _spin_lock();
2615 0) 0.570 us | expand_files(); 2450 0) 0.570 us | expand_files();
2616 0) 0.586 us | _spin_unlock(); 2451 0) 0.586 us | _spin_unlock();
2617 2452
2618 2453
2619 There are several columns that can be dynamic 2454 There are several columns that can be dynamically
2620 enabled/disabled. You can use every combinati 2455 enabled/disabled. You can use every combination of options you
2621 want, depending on your needs. 2456 want, depending on your needs.
2622 2457
2623 - The cpu number on which the function execut 2458 - The cpu number on which the function executed is default
2624 enabled. It is sometimes better to only tr 2459 enabled. It is sometimes better to only trace one cpu (see
2625 tracing_cpumask file) or you might sometime !! 2460 tracing_cpu_mask file) or you might sometimes see unordered
2626 function calls while cpu tracing switch. 2461 function calls while cpu tracing switch.
2627 2462
2628 - hide: echo nofuncgraph-cpu > trace_ 2463 - hide: echo nofuncgraph-cpu > trace_options
2629 - show: echo funcgraph-cpu > trace_op 2464 - show: echo funcgraph-cpu > trace_options
2630 2465
2631 - The duration (function's time of execution) 2466 - The duration (function's time of execution) is displayed on
2632 the closing bracket line of a function or o 2467 the closing bracket line of a function or on the same line
2633 than the current function in case of a leaf 2468 than the current function in case of a leaf one. It is default
2634 enabled. 2469 enabled.
2635 2470
2636 - hide: echo nofuncgraph-duration > t 2471 - hide: echo nofuncgraph-duration > trace_options
2637 - show: echo funcgraph-duration > tra 2472 - show: echo funcgraph-duration > trace_options
2638 2473
2639 - The overhead field precedes the duration fi 2474 - The overhead field precedes the duration field in case of
2640 reached duration thresholds. 2475 reached duration thresholds.
2641 2476
2642 - hide: echo nofuncgraph-overhead > t 2477 - hide: echo nofuncgraph-overhead > trace_options
2643 - show: echo funcgraph-overhead > tra 2478 - show: echo funcgraph-overhead > trace_options
2644 - depends on: funcgraph-duration 2479 - depends on: funcgraph-duration
2645 2480
2646 ie:: 2481 ie::
2647 2482
2648 3) # 1837.709 us | } /* __switch 2483 3) # 1837.709 us | } /* __switch_to */
2649 3) | finish_task_s 2484 3) | finish_task_switch() {
2650 3) 0.313 us | _raw_spin_u 2485 3) 0.313 us | _raw_spin_unlock_irq();
2651 3) 3.177 us | } 2486 3) 3.177 us | }
2652 3) # 1889.063 us | } /* __schedule 2487 3) # 1889.063 us | } /* __schedule */
2653 3) ! 140.417 us | } /* __schedule * 2488 3) ! 140.417 us | } /* __schedule */
2654 3) # 2034.948 us | } /* schedule */ 2489 3) # 2034.948 us | } /* schedule */
2655 3) * 33998.59 us | } /* schedule_preempt 2490 3) * 33998.59 us | } /* schedule_preempt_disabled */
2656 2491
2657 [...] 2492 [...]
2658 2493
2659 1) 0.260 us | msecs_to_ 2494 1) 0.260 us | msecs_to_jiffies();
2660 1) 0.313 us | __rcu_rea 2495 1) 0.313 us | __rcu_read_unlock();
2661 1) + 61.770 us | } 2496 1) + 61.770 us | }
2662 1) + 64.479 us | } 2497 1) + 64.479 us | }
2663 1) 0.313 us | rcu_bh_qs(); 2498 1) 0.313 us | rcu_bh_qs();
2664 1) 0.313 us | __local_bh_en 2499 1) 0.313 us | __local_bh_enable();
2665 1) ! 217.240 us | } 2500 1) ! 217.240 us | }
2666 1) 0.365 us | idle_cpu(); 2501 1) 0.365 us | idle_cpu();
2667 1) | rcu_irq_exit() 2502 1) | rcu_irq_exit() {
2668 1) 0.417 us | rcu_eqs_enter 2503 1) 0.417 us | rcu_eqs_enter_common.isra.47();
2669 1) 3.125 us | } 2504 1) 3.125 us | }
2670 1) ! 227.812 us | } 2505 1) ! 227.812 us | }
2671 1) ! 457.395 us | } 2506 1) ! 457.395 us | }
2672 1) @ 119760.2 us | } 2507 1) @ 119760.2 us | }
2673 2508
2674 [...] 2509 [...]
2675 2510
2676 2) | handle_IPI() { 2511 2) | handle_IPI() {
2677 1) 6.979 us | } 2512 1) 6.979 us | }
2678 2) 0.417 us | scheduler_ipi(); 2513 2) 0.417 us | scheduler_ipi();
2679 1) 9.791 us | } 2514 1) 9.791 us | }
2680 1) + 12.917 us | } 2515 1) + 12.917 us | }
2681 2) 3.490 us | } 2516 2) 3.490 us | }
2682 1) + 15.729 us | } 2517 1) + 15.729 us | }
2683 1) + 18.542 us | } 2518 1) + 18.542 us | }
2684 2) $ 3594274 us | } 2519 2) $ 3594274 us | }
2685 2520
2686 Flags:: 2521 Flags::
2687 2522
2688 + means that the function exceeded 10 usecs 2523 + means that the function exceeded 10 usecs.
2689 ! means that the function exceeded 100 usec 2524 ! means that the function exceeded 100 usecs.
2690 # means that the function exceeded 1000 use 2525 # means that the function exceeded 1000 usecs.
2691 * means that the function exceeded 10 msecs 2526 * means that the function exceeded 10 msecs.
2692 @ means that the function exceeded 100 msec 2527 @ means that the function exceeded 100 msecs.
2693 $ means that the function exceeded 1 sec. 2528 $ means that the function exceeded 1 sec.
2694 2529
2695 2530
2696 - The task/pid field displays the thread cmdl 2531 - The task/pid field displays the thread cmdline and pid which
2697 executed the function. It is default disabl 2532 executed the function. It is default disabled.
2698 2533
2699 - hide: echo nofuncgraph-proc > trace 2534 - hide: echo nofuncgraph-proc > trace_options
2700 - show: echo funcgraph-proc > trace_o 2535 - show: echo funcgraph-proc > trace_options
2701 2536
2702 ie:: 2537 ie::
2703 2538
2704 # tracer: function_graph 2539 # tracer: function_graph
2705 # 2540 #
2706 # CPU TASK/PID DURATION 2541 # CPU TASK/PID DURATION FUNCTION CALLS
2707 # | | | | | 2542 # | | | | | | | | |
2708 0) sh-4802 | | 2543 0) sh-4802 | | d_free() {
2709 0) sh-4802 | | 2544 0) sh-4802 | | call_rcu() {
2710 0) sh-4802 | | 2545 0) sh-4802 | | __call_rcu() {
2711 0) sh-4802 | 0.616 us | 2546 0) sh-4802 | 0.616 us | rcu_process_gp_end();
2712 0) sh-4802 | 0.586 us | 2547 0) sh-4802 | 0.586 us | check_for_new_grace_period();
2713 0) sh-4802 | 2.899 us | 2548 0) sh-4802 | 2.899 us | }
2714 0) sh-4802 | 4.040 us | 2549 0) sh-4802 | 4.040 us | }
2715 0) sh-4802 | 5.151 us | 2550 0) sh-4802 | 5.151 us | }
2716 0) sh-4802 | + 49.370 us | 2551 0) sh-4802 | + 49.370 us | }
2717 2552
2718 2553
2719 - The absolute time field is an absolute time 2554 - The absolute time field is an absolute timestamp given by the
2720 system clock since it started. A snapshot o 2555 system clock since it started. A snapshot of this time is
2721 given on each entry/exit of functions 2556 given on each entry/exit of functions
2722 2557
2723 - hide: echo nofuncgraph-abstime > tr 2558 - hide: echo nofuncgraph-abstime > trace_options
2724 - show: echo funcgraph-abstime > trac 2559 - show: echo funcgraph-abstime > trace_options
2725 2560
2726 ie:: 2561 ie::
2727 2562
2728 # 2563 #
2729 # TIME CPU DURATION 2564 # TIME CPU DURATION FUNCTION CALLS
2730 # | | | | 2565 # | | | | | | | |
2731 360.774522 | 1) 0.541 us | 2566 360.774522 | 1) 0.541 us | }
2732 360.774522 | 1) 4.663 us | 2567 360.774522 | 1) 4.663 us | }
2733 360.774523 | 1) 0.541 us | 2568 360.774523 | 1) 0.541 us | __wake_up_bit();
2734 360.774524 | 1) 6.796 us | 2569 360.774524 | 1) 6.796 us | }
2735 360.774524 | 1) 7.952 us | 2570 360.774524 | 1) 7.952 us | }
2736 360.774525 | 1) 9.063 us | 2571 360.774525 | 1) 9.063 us | }
2737 360.774525 | 1) 0.615 us | 2572 360.774525 | 1) 0.615 us | journal_mark_dirty();
2738 360.774527 | 1) 0.578 us | 2573 360.774527 | 1) 0.578 us | __brelse();
2739 360.774528 | 1) | 2574 360.774528 | 1) | reiserfs_prepare_for_journal() {
2740 360.774528 | 1) | 2575 360.774528 | 1) | unlock_buffer() {
2741 360.774529 | 1) | 2576 360.774529 | 1) | wake_up_bit() {
2742 360.774529 | 1) | 2577 360.774529 | 1) | bit_waitqueue() {
2743 360.774530 | 1) 0.594 us | 2578 360.774530 | 1) 0.594 us | __phys_addr();
2744 2579
2745 2580
2746 The function name is always displayed after t 2581 The function name is always displayed after the closing bracket
2747 for a function if the start of that function 2582 for a function if the start of that function is not in the
2748 trace buffer. 2583 trace buffer.
2749 2584
2750 Display of the function name after the closin 2585 Display of the function name after the closing bracket may be
2751 enabled for functions whose start is in the t 2586 enabled for functions whose start is in the trace buffer,
2752 allowing easier searching with grep for funct 2587 allowing easier searching with grep for function durations.
2753 It is default disabled. 2588 It is default disabled.
2754 2589
2755 - hide: echo nofuncgraph-tail > trace 2590 - hide: echo nofuncgraph-tail > trace_options
2756 - show: echo funcgraph-tail > trace_o 2591 - show: echo funcgraph-tail > trace_options
2757 2592
2758 Example with nofuncgraph-tail (default):: 2593 Example with nofuncgraph-tail (default)::
2759 2594
2760 0) | putname() { 2595 0) | putname() {
2761 0) | kmem_cache_free 2596 0) | kmem_cache_free() {
2762 0) 0.518 us | __phys_addr() 2597 0) 0.518 us | __phys_addr();
2763 0) 1.757 us | } 2598 0) 1.757 us | }
2764 0) 2.861 us | } 2599 0) 2.861 us | }
2765 2600
2766 Example with funcgraph-tail:: 2601 Example with funcgraph-tail::
2767 2602
2768 0) | putname() { 2603 0) | putname() {
2769 0) | kmem_cache_free 2604 0) | kmem_cache_free() {
2770 0) 0.518 us | __phys_addr() 2605 0) 0.518 us | __phys_addr();
2771 0) 1.757 us | } /* kmem_cache 2606 0) 1.757 us | } /* kmem_cache_free() */
2772 0) 2.861 us | } /* putname() */ 2607 0) 2.861 us | } /* putname() */
2773 2608
2774 The return value of each traced function can <<
2775 an equal sign "=". When encountering system c <<
2776 can be very helpful to quickly locate the fun <<
2777 returns an error code. <<
2778 <<
2779 - hide: echo nofuncgraph-retval > tra <<
2780 - show: echo funcgraph-retval > trace <<
2781 <<
2782 Example with funcgraph-retval:: <<
2783 <<
2784 1) | cgroup_migrate() { <<
2785 1) 0.651 us | cgroup_migrate_ad <<
2786 1) | cgroup_migrate_ex <<
2787 1) | cpu_cgroup_can_ <<
2788 1) | cgroup_taskse <<
2789 1) 0.732 us | cgroup_task <<
2790 1) 1.232 us | } /* cgroup_t <<
2791 1) 0.380 us | sched_rt_can_ <<
2792 1) 2.335 us | } /* cpu_cgroup <<
2793 1) 4.369 us | } /* cgroup_migra <<
2794 1) 7.143 us | } /* cgroup_migrate <<
2795 <<
2796 The above example shows that the function cpu <<
2797 returned the error code -22 firstly, then we <<
2798 of this function to get the root cause. <<
2799 <<
2800 When the option funcgraph-retval-hex is not s <<
2801 be displayed in a smart way. Specifically, if <<
2802 it will be printed in signed decimal format, <<
2803 printed in hexadecimal format. <<
2804 <<
2805 - smart: echo nofuncgraph-retval-hex <<
2806 - hexadecimal: echo funcgraph-retval- <<
2807 <<
2808 Example with funcgraph-retval-hex:: <<
2809 <<
2810 1) | cgroup_migrate() <<
2811 1) 0.651 us | cgroup_migrate_ <<
2812 1) | cgroup_migrate_ <<
2813 1) | cpu_cgroup_ca <<
2814 1) | cgroup_task <<
2815 1) 0.732 us | cgroup_ta <<
2816 1) 1.232 us | } /* cgroup <<
2817 1) 0.380 us | sched_rt_ca <<
2818 1) 2.335 us | } /* cpu_cgro <<
2819 1) 4.369 us | } /* cgroup_mig <<
2820 1) 7.143 us | } /* cgroup_migra <<
2821 <<
2822 At present, there are some limitations when u <<
2823 option, and these limitations will be elimina <<
2824 <<
2825 - Even if the function return type is void, a <<
2826 be printed, and you can just ignore it. <<
2827 <<
2828 - Even if return values are stored in multipl <<
2829 value contained in the first register will <<
2830 To illustrate, in the x86 architecture, eax <<
2831 a 64-bit return value, with the lower 32 bi <<
2832 upper 32 bits saved in edx. However, only t <<
2833 will be recorded and printed. <<
2834 <<
2835 - In certain procedure call standards, such a <<
2836 type is smaller than a GPR, it is the respo <<
2837 to perform the narrowing, and the upper bit <<
2838 Therefore, it is advisable to check the cod <<
2839 when using a u8 in a 64-bit GPR, bits [63:8 <<
2840 especially when larger types are truncated, <<
2841 Here are some specific cases to illustrate <<
2842 <<
2843 **Case One**: <<
2844 <<
2845 The function narrow_to_u8 is defined as fol <<
2846 <<
2847 u8 narrow_to_u8(u64 val) <<
2848 { <<
2849 // implicitly truncated <<
2850 return val; <<
2851 } <<
2852 <<
2853 It may be compiled to:: <<
2854 <<
2855 narrow_to_u8: <<
2856 < ... ftrace instrumentation <<
2857 RET <<
2858 <<
2859 If you pass 0x123456789abcdef to this funct <<
2860 it may be recorded as 0x123456789abcdef ins <<
2861 <<
2862 **Case Two**: <<
2863 <<
2864 The function error_if_not_4g_aligned is def <<
2865 <<
2866 int error_if_not_4g_aligned(u64 val) <<
2867 { <<
2868 if (val & GENMASK(31, 0)) <<
2869 return -EINVAL; <<
2870 <<
2871 return 0; <<
2872 } <<
2873 <<
2874 It could be compiled to:: <<
2875 <<
2876 error_if_not_4g_aligned: <<
2877 CBNZ w0, .Lnot_aligned <<
2878 RET // bi <<
2879 // [6 <<
2880 .Lnot_aligned: <<
2881 MOV x0, #-EINVAL <<
2882 RET <<
2883 <<
2884 When passing 0x2_0000_0000 to it, the retur <<
2885 0x2_0000_0000 instead of 0. <<
2886 <<
2887 You can put some comments on specific functio 2609 You can put some comments on specific functions by using
2888 trace_printk() For example, if you want to pu 2610 trace_printk() For example, if you want to put a comment inside
2889 the __might_sleep() function, you just have t 2611 the __might_sleep() function, you just have to include
2890 <linux/ftrace.h> and call trace_printk() insi 2612 <linux/ftrace.h> and call trace_printk() inside __might_sleep()::
2891 2613
2892 trace_printk("I'm a comment!\n") 2614 trace_printk("I'm a comment!\n")
2893 2615
2894 will produce:: 2616 will produce::
2895 2617
2896 1) | __might_sle 2618 1) | __might_sleep() {
2897 1) | /* I'm a 2619 1) | /* I'm a comment! */
2898 1) 1.449 us | } 2620 1) 1.449 us | }
2899 2621
2900 2622
2901 You might find other useful features for this 2623 You might find other useful features for this tracer in the
2902 following "dynamic ftrace" section such as tr 2624 following "dynamic ftrace" section such as tracing only specific
2903 functions or tasks. 2625 functions or tasks.
2904 2626
2905 dynamic ftrace 2627 dynamic ftrace
2906 -------------- 2628 --------------
2907 2629
2908 If CONFIG_DYNAMIC_FTRACE is set, the system w 2630 If CONFIG_DYNAMIC_FTRACE is set, the system will run with
2909 virtually no overhead when function tracing i 2631 virtually no overhead when function tracing is disabled. The way
2910 this works is the mcount function call (place 2632 this works is the mcount function call (placed at the start of
2911 every kernel function, produced by the -pg sw 2633 every kernel function, produced by the -pg switch in gcc),
2912 starts of pointing to a simple return. (Enabl 2634 starts of pointing to a simple return. (Enabling FTRACE will
2913 include the -pg switch in the compiling of th 2635 include the -pg switch in the compiling of the kernel.)
2914 2636
2915 At compile time every C file object is run th 2637 At compile time every C file object is run through the
2916 recordmcount program (located in the scripts 2638 recordmcount program (located in the scripts directory). This
2917 program will parse the ELF headers in the C o 2639 program will parse the ELF headers in the C object to find all
2918 the locations in the .text section that call 2640 the locations in the .text section that call mcount. Starting
2919 with gcc version 4.6, the -mfentry has been a 2641 with gcc version 4.6, the -mfentry has been added for x86, which
2920 calls "__fentry__" instead of "mcount". Which 2642 calls "__fentry__" instead of "mcount". Which is called before
2921 the creation of the stack frame. 2643 the creation of the stack frame.
2922 2644
2923 Note, not all sections are traced. They may b 2645 Note, not all sections are traced. They may be prevented by either
2924 a notrace, or blocked another way and all inl 2646 a notrace, or blocked another way and all inline functions are not
2925 traced. Check the "available_filter_functions 2647 traced. Check the "available_filter_functions" file to see what functions
2926 can be traced. 2648 can be traced.
2927 2649
2928 A section called "__mcount_loc" is created th 2650 A section called "__mcount_loc" is created that holds
2929 references to all the mcount/fentry call site 2651 references to all the mcount/fentry call sites in the .text section.
2930 The recordmcount program re-links this sectio 2652 The recordmcount program re-links this section back into the
2931 original object. The final linking stage of t 2653 original object. The final linking stage of the kernel will add all these
2932 references into a single table. 2654 references into a single table.
2933 2655
2934 On boot up, before SMP is initialized, the dy 2656 On boot up, before SMP is initialized, the dynamic ftrace code
2935 scans this table and updates all the location 2657 scans this table and updates all the locations into nops. It
2936 also records the locations, which are added t 2658 also records the locations, which are added to the
2937 available_filter_functions list. Modules are 2659 available_filter_functions list. Modules are processed as they
2938 are loaded and before they are executed. Whe 2660 are loaded and before they are executed. When a module is
2939 unloaded, it also removes its functions from 2661 unloaded, it also removes its functions from the ftrace function
2940 list. This is automatic in the module unload 2662 list. This is automatic in the module unload code, and the
2941 module author does not need to worry about it 2663 module author does not need to worry about it.
2942 2664
2943 When tracing is enabled, the process of modif 2665 When tracing is enabled, the process of modifying the function
2944 tracepoints is dependent on architecture. The 2666 tracepoints is dependent on architecture. The old method is to use
2945 kstop_machine to prevent races with the CPUs 2667 kstop_machine to prevent races with the CPUs executing code being
2946 modified (which can cause the CPU to do undes 2668 modified (which can cause the CPU to do undesirable things, especially
2947 if the modified code crosses cache (or page) 2669 if the modified code crosses cache (or page) boundaries), and the nops are
2948 patched back to calls. But this time, they do 2670 patched back to calls. But this time, they do not call mcount
2949 (which is just a function stub). They now cal 2671 (which is just a function stub). They now call into the ftrace
2950 infrastructure. 2672 infrastructure.
2951 2673
2952 The new method of modifying the function trac 2674 The new method of modifying the function tracepoints is to place
2953 a breakpoint at the location to be modified, 2675 a breakpoint at the location to be modified, sync all CPUs, modify
2954 the rest of the instruction not covered by th 2676 the rest of the instruction not covered by the breakpoint. Sync
2955 all CPUs again, and then remove the breakpoin 2677 all CPUs again, and then remove the breakpoint with the finished
2956 version to the ftrace call site. 2678 version to the ftrace call site.
2957 2679
2958 Some archs do not even need to monkey around 2680 Some archs do not even need to monkey around with the synchronization,
2959 and can just slap the new code on top of the 2681 and can just slap the new code on top of the old without any
2960 problems with other CPUs executing it at the 2682 problems with other CPUs executing it at the same time.
2961 2683
2962 One special side-effect to the recording of t 2684 One special side-effect to the recording of the functions being
2963 traced is that we can now selectively choose 2685 traced is that we can now selectively choose which functions we
2964 wish to trace and which ones we want the mcou 2686 wish to trace and which ones we want the mcount calls to remain
2965 as nops. 2687 as nops.
2966 2688
2967 Two files are used, one for enabling and one 2689 Two files are used, one for enabling and one for disabling the
2968 tracing of specified functions. They are: 2690 tracing of specified functions. They are:
2969 2691
2970 set_ftrace_filter 2692 set_ftrace_filter
2971 2693
2972 and 2694 and
2973 2695
2974 set_ftrace_notrace 2696 set_ftrace_notrace
2975 2697
2976 A list of available functions that you can ad 2698 A list of available functions that you can add to these files is
2977 listed in: 2699 listed in:
2978 2700
2979 available_filter_functions 2701 available_filter_functions
2980 2702
2981 :: 2703 ::
2982 2704
2983 # cat available_filter_functions 2705 # cat available_filter_functions
2984 put_prev_task_idle 2706 put_prev_task_idle
2985 kmem_cache_create 2707 kmem_cache_create
2986 pick_next_task_rt 2708 pick_next_task_rt
2987 cpus_read_lock !! 2709 get_online_cpus
2988 pick_next_task_fair 2710 pick_next_task_fair
2989 mutex_lock 2711 mutex_lock
2990 [...] 2712 [...]
2991 2713
2992 If I am only interested in sys_nanosleep and 2714 If I am only interested in sys_nanosleep and hrtimer_interrupt::
2993 2715
2994 # echo sys_nanosleep hrtimer_interrupt > se 2716 # echo sys_nanosleep hrtimer_interrupt > set_ftrace_filter
2995 # echo function > current_tracer 2717 # echo function > current_tracer
2996 # echo 1 > tracing_on 2718 # echo 1 > tracing_on
2997 # usleep 1 2719 # usleep 1
2998 # echo 0 > tracing_on 2720 # echo 0 > tracing_on
2999 # cat trace 2721 # cat trace
3000 # tracer: function 2722 # tracer: function
3001 # 2723 #
3002 # entries-in-buffer/entries-written: 5/5 2724 # entries-in-buffer/entries-written: 5/5 #P:4
3003 # 2725 #
3004 # _-----=> irq 2726 # _-----=> irqs-off
3005 # / _----=> nee 2727 # / _----=> need-resched
3006 # | / _---=> har 2728 # | / _---=> hardirq/softirq
3007 # || / _--=> pre 2729 # || / _--=> preempt-depth
3008 # ||| / dela 2730 # ||| / delay
3009 # TASK-PID CPU# |||| TIMEST 2731 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
3010 # | | | |||| | 2732 # | | | |||| | |
3011 usleep-2665 [001] .... 4186.475 2733 usleep-2665 [001] .... 4186.475355: sys_nanosleep <-system_call_fastpath
3012 <idle>-0 [001] d.h1 4186.475 2734 <idle>-0 [001] d.h1 4186.475409: hrtimer_interrupt <-smp_apic_timer_interrupt
3013 usleep-2665 [001] d.h1 4186.475 2735 usleep-2665 [001] d.h1 4186.475426: hrtimer_interrupt <-smp_apic_timer_interrupt
3014 <idle>-0 [003] d.h1 4186.475 2736 <idle>-0 [003] d.h1 4186.475426: hrtimer_interrupt <-smp_apic_timer_interrupt
3015 <idle>-0 [002] d.h1 4186.475 2737 <idle>-0 [002] d.h1 4186.475427: hrtimer_interrupt <-smp_apic_timer_interrupt
3016 2738
3017 To see which functions are being traced, you 2739 To see which functions are being traced, you can cat the file:
3018 :: 2740 ::
3019 2741
3020 # cat set_ftrace_filter 2742 # cat set_ftrace_filter
3021 hrtimer_interrupt 2743 hrtimer_interrupt
3022 sys_nanosleep 2744 sys_nanosleep
3023 2745
3024 2746
3025 Perhaps this is not enough. The filters also 2747 Perhaps this is not enough. The filters also allow glob(7) matching.
3026 2748
3027 ``<match>*`` 2749 ``<match>*``
3028 will match functions that begin with 2750 will match functions that begin with <match>
3029 ``*<match>`` 2751 ``*<match>``
3030 will match functions that end with <m 2752 will match functions that end with <match>
3031 ``*<match>*`` 2753 ``*<match>*``
3032 will match functions that have <match 2754 will match functions that have <match> in it
3033 ``<match1>*<match2>`` 2755 ``<match1>*<match2>``
3034 will match functions that begin with 2756 will match functions that begin with <match1> and end with <match2>
3035 2757
3036 .. note:: 2758 .. note::
3037 It is better to use quotes to enclose t 2759 It is better to use quotes to enclose the wild cards,
3038 otherwise the shell may expand the para 2760 otherwise the shell may expand the parameters into names
3039 of files in the local directory. 2761 of files in the local directory.
3040 2762
3041 :: 2763 ::
3042 2764
3043 # echo 'hrtimer_*' > set_ftrace_filter 2765 # echo 'hrtimer_*' > set_ftrace_filter
3044 2766
3045 Produces:: 2767 Produces::
3046 2768
3047 # tracer: function 2769 # tracer: function
3048 # 2770 #
3049 # entries-in-buffer/entries-written: 897/89 2771 # entries-in-buffer/entries-written: 897/897 #P:4
3050 # 2772 #
3051 # _-----=> irq 2773 # _-----=> irqs-off
3052 # / _----=> nee 2774 # / _----=> need-resched
3053 # | / _---=> har 2775 # | / _---=> hardirq/softirq
3054 # || / _--=> pre 2776 # || / _--=> preempt-depth
3055 # ||| / dela 2777 # ||| / delay
3056 # TASK-PID CPU# |||| TIMEST 2778 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
3057 # | | | |||| | 2779 # | | | |||| | |
3058 <idle>-0 [003] dN.1 4228.547 2780 <idle>-0 [003] dN.1 4228.547803: hrtimer_cancel <-tick_nohz_idle_exit
3059 <idle>-0 [003] dN.1 4228.547 2781 <idle>-0 [003] dN.1 4228.547804: hrtimer_try_to_cancel <-hrtimer_cancel
3060 <idle>-0 [003] dN.2 4228.547 2782 <idle>-0 [003] dN.2 4228.547805: hrtimer_force_reprogram <-__remove_hrtimer
3061 <idle>-0 [003] dN.1 4228.547 2783 <idle>-0 [003] dN.1 4228.547805: hrtimer_forward <-tick_nohz_idle_exit
3062 <idle>-0 [003] dN.1 4228.547 2784 <idle>-0 [003] dN.1 4228.547805: hrtimer_start_range_ns <-hrtimer_start_expires.constprop.11
3063 <idle>-0 [003] d..1 4228.547 2785 <idle>-0 [003] d..1 4228.547858: hrtimer_get_next_event <-get_next_timer_interrupt
3064 <idle>-0 [003] d..1 4228.547 2786 <idle>-0 [003] d..1 4228.547859: hrtimer_start <-__tick_nohz_idle_enter
3065 <idle>-0 [003] d..2 4228.547 2787 <idle>-0 [003] d..2 4228.547860: hrtimer_force_reprogram <-__rem
3066 2788
3067 Notice that we lost the sys_nanosleep. 2789 Notice that we lost the sys_nanosleep.
3068 :: 2790 ::
3069 2791
3070 # cat set_ftrace_filter 2792 # cat set_ftrace_filter
3071 hrtimer_run_queues 2793 hrtimer_run_queues
3072 hrtimer_run_pending 2794 hrtimer_run_pending
3073 hrtimer_init 2795 hrtimer_init
3074 hrtimer_cancel 2796 hrtimer_cancel
3075 hrtimer_try_to_cancel 2797 hrtimer_try_to_cancel
3076 hrtimer_forward 2798 hrtimer_forward
3077 hrtimer_start 2799 hrtimer_start
3078 hrtimer_reprogram 2800 hrtimer_reprogram
3079 hrtimer_force_reprogram 2801 hrtimer_force_reprogram
3080 hrtimer_get_next_event 2802 hrtimer_get_next_event
3081 hrtimer_interrupt 2803 hrtimer_interrupt
3082 hrtimer_nanosleep 2804 hrtimer_nanosleep
3083 hrtimer_wakeup 2805 hrtimer_wakeup
3084 hrtimer_get_remaining 2806 hrtimer_get_remaining
3085 hrtimer_get_res 2807 hrtimer_get_res
3086 hrtimer_init_sleeper 2808 hrtimer_init_sleeper
3087 2809
3088 2810
3089 This is because the '>' and '>>' act just lik 2811 This is because the '>' and '>>' act just like they do in bash.
3090 To rewrite the filters, use '>' 2812 To rewrite the filters, use '>'
3091 To append to the filters, use '>>' 2813 To append to the filters, use '>>'
3092 2814
3093 To clear out a filter so that all functions w 2815 To clear out a filter so that all functions will be recorded
3094 again:: 2816 again::
3095 2817
3096 # echo > set_ftrace_filter 2818 # echo > set_ftrace_filter
3097 # cat set_ftrace_filter 2819 # cat set_ftrace_filter
3098 # 2820 #
3099 2821
3100 Again, now we want to append. 2822 Again, now we want to append.
3101 2823
3102 :: 2824 ::
3103 2825
3104 # echo sys_nanosleep > set_ftrace_filter 2826 # echo sys_nanosleep > set_ftrace_filter
3105 # cat set_ftrace_filter 2827 # cat set_ftrace_filter
3106 sys_nanosleep 2828 sys_nanosleep
3107 # echo 'hrtimer_*' >> set_ftrace_filter 2829 # echo 'hrtimer_*' >> set_ftrace_filter
3108 # cat set_ftrace_filter 2830 # cat set_ftrace_filter
3109 hrtimer_run_queues 2831 hrtimer_run_queues
3110 hrtimer_run_pending 2832 hrtimer_run_pending
3111 hrtimer_init 2833 hrtimer_init
3112 hrtimer_cancel 2834 hrtimer_cancel
3113 hrtimer_try_to_cancel 2835 hrtimer_try_to_cancel
3114 hrtimer_forward 2836 hrtimer_forward
3115 hrtimer_start 2837 hrtimer_start
3116 hrtimer_reprogram 2838 hrtimer_reprogram
3117 hrtimer_force_reprogram 2839 hrtimer_force_reprogram
3118 hrtimer_get_next_event 2840 hrtimer_get_next_event
3119 hrtimer_interrupt 2841 hrtimer_interrupt
3120 sys_nanosleep 2842 sys_nanosleep
3121 hrtimer_nanosleep 2843 hrtimer_nanosleep
3122 hrtimer_wakeup 2844 hrtimer_wakeup
3123 hrtimer_get_remaining 2845 hrtimer_get_remaining
3124 hrtimer_get_res 2846 hrtimer_get_res
3125 hrtimer_init_sleeper 2847 hrtimer_init_sleeper
3126 2848
3127 2849
3128 The set_ftrace_notrace prevents those functio 2850 The set_ftrace_notrace prevents those functions from being
3129 traced. 2851 traced.
3130 :: 2852 ::
3131 2853
3132 # echo '*preempt*' '*lock*' > set_ftrace_no 2854 # echo '*preempt*' '*lock*' > set_ftrace_notrace
3133 2855
3134 Produces:: 2856 Produces::
3135 2857
3136 # tracer: function 2858 # tracer: function
3137 # 2859 #
3138 # entries-in-buffer/entries-written: 39608/ 2860 # entries-in-buffer/entries-written: 39608/39608 #P:4
3139 # 2861 #
3140 # _-----=> irq 2862 # _-----=> irqs-off
3141 # / _----=> nee 2863 # / _----=> need-resched
3142 # | / _---=> har 2864 # | / _---=> hardirq/softirq
3143 # || / _--=> pre 2865 # || / _--=> preempt-depth
3144 # ||| / dela 2866 # ||| / delay
3145 # TASK-PID CPU# |||| TIMEST 2867 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
3146 # | | | |||| | 2868 # | | | |||| | |
3147 bash-1994 [000] .... 4342.324 2869 bash-1994 [000] .... 4342.324896: file_ra_state_init <-do_dentry_open
3148 bash-1994 [000] .... 4342.324 2870 bash-1994 [000] .... 4342.324897: open_check_o_direct <-do_last
3149 bash-1994 [000] .... 4342.324 2871 bash-1994 [000] .... 4342.324897: ima_file_check <-do_last
3150 bash-1994 [000] .... 4342.324 2872 bash-1994 [000] .... 4342.324898: process_measurement <-ima_file_check
3151 bash-1994 [000] .... 4342.324 2873 bash-1994 [000] .... 4342.324898: ima_get_action <-process_measurement
3152 bash-1994 [000] .... 4342.324 2874 bash-1994 [000] .... 4342.324898: ima_match_policy <-ima_get_action
3153 bash-1994 [000] .... 4342.324 2875 bash-1994 [000] .... 4342.324899: do_truncate <-do_last
3154 bash-1994 [000] .... 4342.324 !! 2876 bash-1994 [000] .... 4342.324899: should_remove_suid <-do_truncate
3155 bash-1994 [000] .... 4342.324 2877 bash-1994 [000] .... 4342.324899: notify_change <-do_truncate
3156 bash-1994 [000] .... 4342.324 2878 bash-1994 [000] .... 4342.324900: current_fs_time <-notify_change
3157 bash-1994 [000] .... 4342.324 2879 bash-1994 [000] .... 4342.324900: current_kernel_time <-current_fs_time
3158 bash-1994 [000] .... 4342.324 2880 bash-1994 [000] .... 4342.324900: timespec_trunc <-current_fs_time
3159 2881
3160 We can see that there's no more lock or preem 2882 We can see that there's no more lock or preempt tracing.
3161 2883
3162 Selecting function filters via index 2884 Selecting function filters via index
3163 ------------------------------------ 2885 ------------------------------------
3164 2886
3165 Because processing of strings is expensive (t 2887 Because processing of strings is expensive (the address of the function
3166 needs to be looked up before comparing to the 2888 needs to be looked up before comparing to the string being passed in),
3167 an index can be used as well to enable functi 2889 an index can be used as well to enable functions. This is useful in the
3168 case of setting thousands of specific functio 2890 case of setting thousands of specific functions at a time. By passing
3169 in a list of numbers, no string processing wi 2891 in a list of numbers, no string processing will occur. Instead, the function
3170 at the specific location in the internal arra 2892 at the specific location in the internal array (which corresponds to the
3171 functions in the "available_filter_functions" 2893 functions in the "available_filter_functions" file), is selected.
3172 2894
3173 :: 2895 ::
3174 2896
3175 # echo 1 > set_ftrace_filter 2897 # echo 1 > set_ftrace_filter
3176 2898
3177 Will select the first function listed in "ava 2899 Will select the first function listed in "available_filter_functions"
3178 2900
3179 :: 2901 ::
3180 2902
3181 # head -1 available_filter_functions 2903 # head -1 available_filter_functions
3182 trace_initcall_finish_cb 2904 trace_initcall_finish_cb
3183 2905
3184 # cat set_ftrace_filter 2906 # cat set_ftrace_filter
3185 trace_initcall_finish_cb 2907 trace_initcall_finish_cb
3186 2908
3187 # head -50 available_filter_functions | tai 2909 # head -50 available_filter_functions | tail -1
3188 x86_pmu_commit_txn 2910 x86_pmu_commit_txn
3189 2911
3190 # echo 1 50 > set_ftrace_filter 2912 # echo 1 50 > set_ftrace_filter
3191 # cat set_ftrace_filter 2913 # cat set_ftrace_filter
3192 trace_initcall_finish_cb 2914 trace_initcall_finish_cb
3193 x86_pmu_commit_txn 2915 x86_pmu_commit_txn
3194 2916
3195 Dynamic ftrace with the function graph tracer 2917 Dynamic ftrace with the function graph tracer
3196 --------------------------------------------- 2918 ---------------------------------------------
3197 2919
3198 Although what has been explained above concer 2920 Although what has been explained above concerns both the
3199 function tracer and the function-graph-tracer 2921 function tracer and the function-graph-tracer, there are some
3200 special features only available in the functi 2922 special features only available in the function-graph tracer.
3201 2923
3202 If you want to trace only one function and al 2924 If you want to trace only one function and all of its children,
3203 you just have to echo its name into set_graph 2925 you just have to echo its name into set_graph_function::
3204 2926
3205 echo __do_fault > set_graph_function 2927 echo __do_fault > set_graph_function
3206 2928
3207 will produce the following "expanded" trace o 2929 will produce the following "expanded" trace of the __do_fault()
3208 function:: 2930 function::
3209 2931
3210 0) | __do_fault() { 2932 0) | __do_fault() {
3211 0) | filemap_fault() { 2933 0) | filemap_fault() {
3212 0) | find_lock_page() { 2934 0) | find_lock_page() {
3213 0) 0.804 us | find_get_page(); 2935 0) 0.804 us | find_get_page();
3214 0) | __might_sleep() 2936 0) | __might_sleep() {
3215 0) 1.329 us | } 2937 0) 1.329 us | }
3216 0) 3.904 us | } 2938 0) 3.904 us | }
3217 0) 4.979 us | } 2939 0) 4.979 us | }
3218 0) 0.653 us | _spin_lock(); 2940 0) 0.653 us | _spin_lock();
3219 0) 0.578 us | page_add_file_rmap() 2941 0) 0.578 us | page_add_file_rmap();
3220 0) 0.525 us | native_set_pte_at(); 2942 0) 0.525 us | native_set_pte_at();
3221 0) 0.585 us | _spin_unlock(); 2943 0) 0.585 us | _spin_unlock();
3222 0) | unlock_page() { 2944 0) | unlock_page() {
3223 0) 0.541 us | page_waitqueue(); 2945 0) 0.541 us | page_waitqueue();
3224 0) 0.639 us | __wake_up_bit(); 2946 0) 0.639 us | __wake_up_bit();
3225 0) 2.786 us | } 2947 0) 2.786 us | }
3226 0) + 14.237 us | } 2948 0) + 14.237 us | }
3227 0) | __do_fault() { 2949 0) | __do_fault() {
3228 0) | filemap_fault() { 2950 0) | filemap_fault() {
3229 0) | find_lock_page() { 2951 0) | find_lock_page() {
3230 0) 0.698 us | find_get_page(); 2952 0) 0.698 us | find_get_page();
3231 0) | __might_sleep() 2953 0) | __might_sleep() {
3232 0) 1.412 us | } 2954 0) 1.412 us | }
3233 0) 3.950 us | } 2955 0) 3.950 us | }
3234 0) 5.098 us | } 2956 0) 5.098 us | }
3235 0) 0.631 us | _spin_lock(); 2957 0) 0.631 us | _spin_lock();
3236 0) 0.571 us | page_add_file_rmap() 2958 0) 0.571 us | page_add_file_rmap();
3237 0) 0.526 us | native_set_pte_at(); 2959 0) 0.526 us | native_set_pte_at();
3238 0) 0.586 us | _spin_unlock(); 2960 0) 0.586 us | _spin_unlock();
3239 0) | unlock_page() { 2961 0) | unlock_page() {
3240 0) 0.533 us | page_waitqueue(); 2962 0) 0.533 us | page_waitqueue();
3241 0) 0.638 us | __wake_up_bit(); 2963 0) 0.638 us | __wake_up_bit();
3242 0) 2.793 us | } 2964 0) 2.793 us | }
3243 0) + 14.012 us | } 2965 0) + 14.012 us | }
3244 2966
3245 You can also expand several functions at once 2967 You can also expand several functions at once::
3246 2968
3247 echo sys_open > set_graph_function 2969 echo sys_open > set_graph_function
3248 echo sys_close >> set_graph_function 2970 echo sys_close >> set_graph_function
3249 2971
3250 Now if you want to go back to trace all funct 2972 Now if you want to go back to trace all functions you can clear
3251 this special filter via:: 2973 this special filter via::
3252 2974
3253 echo > set_graph_function 2975 echo > set_graph_function
3254 2976
3255 2977
3256 ftrace_enabled 2978 ftrace_enabled
3257 -------------- 2979 --------------
3258 2980
3259 Note, the proc sysctl ftrace_enable is a big 2981 Note, the proc sysctl ftrace_enable is a big on/off switch for the
3260 function tracer. By default it is enabled (wh 2982 function tracer. By default it is enabled (when function tracing is
3261 enabled in the kernel). If it is disabled, al 2983 enabled in the kernel). If it is disabled, all function tracing is
3262 disabled. This includes not only the function 2984 disabled. This includes not only the function tracers for ftrace, but
3263 also for any other uses (perf, kprobes, stack 2985 also for any other uses (perf, kprobes, stack tracing, profiling, etc). It
3264 cannot be disabled if there is a callback wit 2986 cannot be disabled if there is a callback with FTRACE_OPS_FL_PERMANENT set
3265 registered. 2987 registered.
3266 2988
3267 Please disable this with care. 2989 Please disable this with care.
3268 2990
3269 This can be disable (and enabled) with:: 2991 This can be disable (and enabled) with::
3270 2992
3271 sysctl kernel.ftrace_enabled=0 2993 sysctl kernel.ftrace_enabled=0
3272 sysctl kernel.ftrace_enabled=1 2994 sysctl kernel.ftrace_enabled=1
3273 2995
3274 or 2996 or
3275 2997
3276 echo 0 > /proc/sys/kernel/ftrace_enabled 2998 echo 0 > /proc/sys/kernel/ftrace_enabled
3277 echo 1 > /proc/sys/kernel/ftrace_enabled 2999 echo 1 > /proc/sys/kernel/ftrace_enabled
3278 3000
3279 3001
3280 Filter commands 3002 Filter commands
3281 --------------- 3003 ---------------
3282 3004
3283 A few commands are supported by the set_ftrac 3005 A few commands are supported by the set_ftrace_filter interface.
3284 Trace commands have the following format:: 3006 Trace commands have the following format::
3285 3007
3286 <function>:<command>:<parameter> 3008 <function>:<command>:<parameter>
3287 3009
3288 The following commands are supported: 3010 The following commands are supported:
3289 3011
3290 - mod: 3012 - mod:
3291 This command enables function filtering per 3013 This command enables function filtering per module. The
3292 parameter defines the module. For example, 3014 parameter defines the module. For example, if only the write*
3293 functions in the ext3 module are desired, r 3015 functions in the ext3 module are desired, run:
3294 3016
3295 echo 'write*:mod:ext3' > set_ftrace_filter 3017 echo 'write*:mod:ext3' > set_ftrace_filter
3296 3018
3297 This command interacts with the filter in t 3019 This command interacts with the filter in the same way as
3298 filtering based on function names. Thus, ad 3020 filtering based on function names. Thus, adding more functions
3299 in a different module is accomplished by ap 3021 in a different module is accomplished by appending (>>) to the
3300 filter file. Remove specific module functio 3022 filter file. Remove specific module functions by prepending
3301 '!':: 3023 '!'::
3302 3024
3303 echo '!writeback*:mod:ext3' >> set_ftrace_ 3025 echo '!writeback*:mod:ext3' >> set_ftrace_filter
3304 3026
3305 Mod command supports module globbing. Disab 3027 Mod command supports module globbing. Disable tracing for all
3306 functions except a specific module:: 3028 functions except a specific module::
3307 3029
3308 echo '!*:mod:!ext3' >> set_ftrace_filter 3030 echo '!*:mod:!ext3' >> set_ftrace_filter
3309 3031
3310 Disable tracing for all modules, but still 3032 Disable tracing for all modules, but still trace kernel::
3311 3033
3312 echo '!*:mod:*' >> set_ftrace_filter 3034 echo '!*:mod:*' >> set_ftrace_filter
3313 3035
3314 Enable filter only for kernel:: 3036 Enable filter only for kernel::
3315 3037
3316 echo '*write*:mod:!*' >> set_ftrace_filter 3038 echo '*write*:mod:!*' >> set_ftrace_filter
3317 3039
3318 Enable filter for module globbing:: 3040 Enable filter for module globbing::
3319 3041
3320 echo '*write*:mod:*snd*' >> set_ftrace_fil 3042 echo '*write*:mod:*snd*' >> set_ftrace_filter
3321 3043
3322 - traceon/traceoff: 3044 - traceon/traceoff:
3323 These commands turn tracing on and off when 3045 These commands turn tracing on and off when the specified
3324 functions are hit. The parameter determines 3046 functions are hit. The parameter determines how many times the
3325 tracing system is turned on and off. If uns 3047 tracing system is turned on and off. If unspecified, there is
3326 no limit. For example, to disable tracing w 3048 no limit. For example, to disable tracing when a schedule bug
3327 is hit the first 5 times, run:: 3049 is hit the first 5 times, run::
3328 3050
3329 echo '__schedule_bug:traceoff:5' > set_ftr 3051 echo '__schedule_bug:traceoff:5' > set_ftrace_filter
3330 3052
3331 To always disable tracing when __schedule_b 3053 To always disable tracing when __schedule_bug is hit::
3332 3054
3333 echo '__schedule_bug:traceoff' > set_ftrac 3055 echo '__schedule_bug:traceoff' > set_ftrace_filter
3334 3056
3335 These commands are cumulative whether or no 3057 These commands are cumulative whether or not they are appended
3336 to set_ftrace_filter. To remove a command, 3058 to set_ftrace_filter. To remove a command, prepend it by '!'
3337 and drop the parameter:: 3059 and drop the parameter::
3338 3060
3339 echo '!__schedule_bug:traceoff:0' > set_ft 3061 echo '!__schedule_bug:traceoff:0' > set_ftrace_filter
3340 3062
3341 The above removes the traceoff command for 3063 The above removes the traceoff command for __schedule_bug
3342 that have a counter. To remove commands wit 3064 that have a counter. To remove commands without counters::
3343 3065
3344 echo '!__schedule_bug:traceoff' > set_ftra 3066 echo '!__schedule_bug:traceoff' > set_ftrace_filter
3345 3067
3346 - snapshot: 3068 - snapshot:
3347 Will cause a snapshot to be triggered when 3069 Will cause a snapshot to be triggered when the function is hit.
3348 :: 3070 ::
3349 3071
3350 echo 'native_flush_tlb_others:snapshot' > 3072 echo 'native_flush_tlb_others:snapshot' > set_ftrace_filter
3351 3073
3352 To only snapshot once: 3074 To only snapshot once:
3353 :: 3075 ::
3354 3076
3355 echo 'native_flush_tlb_others:snapshot:1' 3077 echo 'native_flush_tlb_others:snapshot:1' > set_ftrace_filter
3356 3078
3357 To remove the above commands:: 3079 To remove the above commands::
3358 3080
3359 echo '!native_flush_tlb_others:snapshot' > 3081 echo '!native_flush_tlb_others:snapshot' > set_ftrace_filter
3360 echo '!native_flush_tlb_others:snapshot:0' 3082 echo '!native_flush_tlb_others:snapshot:0' > set_ftrace_filter
3361 3083
3362 - enable_event/disable_event: 3084 - enable_event/disable_event:
3363 These commands can enable or disable a trac 3085 These commands can enable or disable a trace event. Note, because
3364 function tracing callbacks are very sensiti 3086 function tracing callbacks are very sensitive, when these commands
3365 are registered, the trace point is activate 3087 are registered, the trace point is activated, but disabled in
3366 a "soft" mode. That is, the tracepoint will 3088 a "soft" mode. That is, the tracepoint will be called, but
3367 just will not be traced. The event tracepoi 3089 just will not be traced. The event tracepoint stays in this mode
3368 as long as there's a command that triggers 3090 as long as there's a command that triggers it.
3369 :: 3091 ::
3370 3092
3371 echo 'try_to_wake_up:enable_event:sched:sc 3093 echo 'try_to_wake_up:enable_event:sched:sched_switch:2' > \
3372 set_ftrace_filter 3094 set_ftrace_filter
3373 3095
3374 The format is:: 3096 The format is::
3375 3097
3376 <function>:enable_event:<system>:<event>[ 3098 <function>:enable_event:<system>:<event>[:count]
3377 <function>:disable_event:<system>:<event> 3099 <function>:disable_event:<system>:<event>[:count]
3378 3100
3379 To remove the events commands:: 3101 To remove the events commands::
3380 3102
3381 echo '!try_to_wake_up:enable_event:sched:s 3103 echo '!try_to_wake_up:enable_event:sched:sched_switch:0' > \
3382 set_ftrace_filter 3104 set_ftrace_filter
3383 echo '!schedule:disable_event:sched:sched_ 3105 echo '!schedule:disable_event:sched:sched_switch' > \
3384 set_ftrace_filter 3106 set_ftrace_filter
3385 3107
3386 - dump: 3108 - dump:
3387 When the function is hit, it will dump the 3109 When the function is hit, it will dump the contents of the ftrace
3388 ring buffer to the console. This is useful 3110 ring buffer to the console. This is useful if you need to debug
3389 something, and want to dump the trace when 3111 something, and want to dump the trace when a certain function
3390 is hit. Perhaps it's a function that is cal 3112 is hit. Perhaps it's a function that is called before a triple
3391 fault happens and does not allow you to get 3113 fault happens and does not allow you to get a regular dump.
3392 3114
3393 - cpudump: 3115 - cpudump:
3394 When the function is hit, it will dump the 3116 When the function is hit, it will dump the contents of the ftrace
3395 ring buffer for the current CPU to the cons 3117 ring buffer for the current CPU to the console. Unlike the "dump"
3396 command, it only prints out the contents of 3118 command, it only prints out the contents of the ring buffer for the
3397 CPU that executed the function that trigger 3119 CPU that executed the function that triggered the dump.
3398 3120
3399 - stacktrace: 3121 - stacktrace:
3400 When the function is hit, a stack trace is 3122 When the function is hit, a stack trace is recorded.
3401 3123
3402 trace_pipe 3124 trace_pipe
3403 ---------- 3125 ----------
3404 3126
3405 The trace_pipe outputs the same content as th 3127 The trace_pipe outputs the same content as the trace file, but
3406 the effect on the tracing is different. Every 3128 the effect on the tracing is different. Every read from
3407 trace_pipe is consumed. This means that subse 3129 trace_pipe is consumed. This means that subsequent reads will be
3408 different. The trace is live. 3130 different. The trace is live.
3409 :: 3131 ::
3410 3132
3411 # echo function > current_tracer 3133 # echo function > current_tracer
3412 # cat trace_pipe > /tmp/trace.out & 3134 # cat trace_pipe > /tmp/trace.out &
3413 [1] 4153 3135 [1] 4153
3414 # echo 1 > tracing_on 3136 # echo 1 > tracing_on
3415 # usleep 1 3137 # usleep 1
3416 # echo 0 > tracing_on 3138 # echo 0 > tracing_on
3417 # cat trace 3139 # cat trace
3418 # tracer: function 3140 # tracer: function
3419 # 3141 #
3420 # entries-in-buffer/entries-written: 0/0 3142 # entries-in-buffer/entries-written: 0/0 #P:4
3421 # 3143 #
3422 # _-----=> irq 3144 # _-----=> irqs-off
3423 # / _----=> nee 3145 # / _----=> need-resched
3424 # | / _---=> har 3146 # | / _---=> hardirq/softirq
3425 # || / _--=> pre 3147 # || / _--=> preempt-depth
3426 # ||| / dela 3148 # ||| / delay
3427 # TASK-PID CPU# |||| TIMEST 3149 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
3428 # | | | |||| | 3150 # | | | |||| | |
3429 3151
3430 # 3152 #
3431 # cat /tmp/trace.out 3153 # cat /tmp/trace.out
3432 bash-1994 [000] .... 5281.5689 3154 bash-1994 [000] .... 5281.568961: mutex_unlock <-rb_simple_write
3433 bash-1994 [000] .... 5281.5689 3155 bash-1994 [000] .... 5281.568963: __mutex_unlock_slowpath <-mutex_unlock
3434 bash-1994 [000] .... 5281.5689 3156 bash-1994 [000] .... 5281.568963: __fsnotify_parent <-fsnotify_modify
3435 bash-1994 [000] .... 5281.5689 3157 bash-1994 [000] .... 5281.568964: fsnotify <-fsnotify_modify
3436 bash-1994 [000] .... 5281.5689 3158 bash-1994 [000] .... 5281.568964: __srcu_read_lock <-fsnotify
3437 bash-1994 [000] .... 5281.5689 3159 bash-1994 [000] .... 5281.568964: add_preempt_count <-__srcu_read_lock
3438 bash-1994 [000] ...1 5281.5689 3160 bash-1994 [000] ...1 5281.568965: sub_preempt_count <-__srcu_read_lock
3439 bash-1994 [000] .... 5281.5689 3161 bash-1994 [000] .... 5281.568965: __srcu_read_unlock <-fsnotify
3440 bash-1994 [000] .... 5281.5689 3162 bash-1994 [000] .... 5281.568967: sys_dup2 <-system_call_fastpath
3441 3163
3442 3164
3443 Note, reading the trace_pipe file will block 3165 Note, reading the trace_pipe file will block until more input is
3444 added. This is contrary to the trace file. If 3166 added. This is contrary to the trace file. If any process opened
3445 the trace file for reading, it will actually 3167 the trace file for reading, it will actually disable tracing and
3446 prevent new entries from being added. The tra 3168 prevent new entries from being added. The trace_pipe file does
3447 not have this limitation. 3169 not have this limitation.
3448 3170
3449 trace entries 3171 trace entries
3450 ------------- 3172 -------------
3451 3173
3452 Having too much or not enough data can be tro 3174 Having too much or not enough data can be troublesome in
3453 diagnosing an issue in the kernel. The file b 3175 diagnosing an issue in the kernel. The file buffer_size_kb is
3454 used to modify the size of the internal trace 3176 used to modify the size of the internal trace buffers. The
3455 number listed is the number of entries that c 3177 number listed is the number of entries that can be recorded per
3456 CPU. To know the full size, multiply the numb 3178 CPU. To know the full size, multiply the number of possible CPUs
3457 with the number of entries. 3179 with the number of entries.
3458 :: 3180 ::
3459 3181
3460 # cat buffer_size_kb 3182 # cat buffer_size_kb
3461 1408 (units kilobytes) 3183 1408 (units kilobytes)
3462 3184
3463 Or simply read buffer_total_size_kb 3185 Or simply read buffer_total_size_kb
3464 :: 3186 ::
3465 3187
3466 # cat buffer_total_size_kb 3188 # cat buffer_total_size_kb
3467 5632 3189 5632
3468 3190
3469 To modify the buffer, simple echo in a number 3191 To modify the buffer, simple echo in a number (in 1024 byte segments).
3470 :: 3192 ::
3471 3193
3472 # echo 10000 > buffer_size_kb 3194 # echo 10000 > buffer_size_kb
3473 # cat buffer_size_kb 3195 # cat buffer_size_kb
3474 10000 (units kilobytes) 3196 10000 (units kilobytes)
3475 3197
3476 It will try to allocate as much as possible. 3198 It will try to allocate as much as possible. If you allocate too
3477 much, it can cause Out-Of-Memory to trigger. 3199 much, it can cause Out-Of-Memory to trigger.
3478 :: 3200 ::
3479 3201
3480 # echo 1000000000000 > buffer_size_kb 3202 # echo 1000000000000 > buffer_size_kb
3481 -bash: echo: write error: Cannot allocate m 3203 -bash: echo: write error: Cannot allocate memory
3482 # cat buffer_size_kb 3204 # cat buffer_size_kb
3483 85 3205 85
3484 3206
3485 The per_cpu buffers can be changed individual 3207 The per_cpu buffers can be changed individually as well:
3486 :: 3208 ::
3487 3209
3488 # echo 10000 > per_cpu/cpu0/buffer_size_kb 3210 # echo 10000 > per_cpu/cpu0/buffer_size_kb
3489 # echo 100 > per_cpu/cpu1/buffer_size_kb 3211 # echo 100 > per_cpu/cpu1/buffer_size_kb
3490 3212
3491 When the per_cpu buffers are not the same, th 3213 When the per_cpu buffers are not the same, the buffer_size_kb
3492 at the top level will just show an X 3214 at the top level will just show an X
3493 :: 3215 ::
3494 3216
3495 # cat buffer_size_kb 3217 # cat buffer_size_kb
3496 X 3218 X
3497 3219
3498 This is where the buffer_total_size_kb is use 3220 This is where the buffer_total_size_kb is useful:
3499 :: 3221 ::
3500 3222
3501 # cat buffer_total_size_kb 3223 # cat buffer_total_size_kb
3502 12916 3224 12916
3503 3225
3504 Writing to the top level buffer_size_kb will 3226 Writing to the top level buffer_size_kb will reset all the buffers
3505 to be the same again. 3227 to be the same again.
3506 3228
3507 Snapshot 3229 Snapshot
3508 -------- 3230 --------
3509 CONFIG_TRACER_SNAPSHOT makes a generic snapsh 3231 CONFIG_TRACER_SNAPSHOT makes a generic snapshot feature
3510 available to all non latency tracers. (Latenc 3232 available to all non latency tracers. (Latency tracers which
3511 record max latency, such as "irqsoff" or "wak 3233 record max latency, such as "irqsoff" or "wakeup", can't use
3512 this feature, since those are already using t 3234 this feature, since those are already using the snapshot
3513 mechanism internally.) 3235 mechanism internally.)
3514 3236
3515 Snapshot preserves a current trace buffer at 3237 Snapshot preserves a current trace buffer at a particular point
3516 in time without stopping tracing. Ftrace swap 3238 in time without stopping tracing. Ftrace swaps the current
3517 buffer with a spare buffer, and tracing conti 3239 buffer with a spare buffer, and tracing continues in the new
3518 current (=previous spare) buffer. 3240 current (=previous spare) buffer.
3519 3241
3520 The following tracefs files in "tracing" are 3242 The following tracefs files in "tracing" are related to this
3521 feature: 3243 feature:
3522 3244
3523 snapshot: 3245 snapshot:
3524 3246
3525 This is used to take a snapshot and t 3247 This is used to take a snapshot and to read the output
3526 of the snapshot. Echo 1 into this fil 3248 of the snapshot. Echo 1 into this file to allocate a
3527 spare buffer and to take a snapshot ( 3249 spare buffer and to take a snapshot (swap), then read
3528 the snapshot from this file in the sa 3250 the snapshot from this file in the same format as
3529 "trace" (described above in the secti 3251 "trace" (described above in the section "The File
3530 System"). Both reads snapshot and tra 3252 System"). Both reads snapshot and tracing are executable
3531 in parallel. When the spare buffer is 3253 in parallel. When the spare buffer is allocated, echoing
3532 0 frees it, and echoing else (positiv 3254 0 frees it, and echoing else (positive) values clear the
3533 snapshot contents. 3255 snapshot contents.
3534 More details are shown in the table b 3256 More details are shown in the table below.
3535 3257
3536 +--------------+------------+-------- 3258 +--------------+------------+------------+------------+
3537 |status\\input | 0 | 1 3259 |status\\input | 0 | 1 | else |
3538 +==============+============+======== 3260 +==============+============+============+============+
3539 |not allocated |(do nothing)| alloc+s 3261 |not allocated |(do nothing)| alloc+swap |(do nothing)|
3540 +--------------+------------+-------- 3262 +--------------+------------+------------+------------+
3541 |allocated | free | swap 3263 |allocated | free | swap | clear |
3542 +--------------+------------+-------- 3264 +--------------+------------+------------+------------+
3543 3265
3544 Here is an example of using the snapshot feat 3266 Here is an example of using the snapshot feature.
3545 :: 3267 ::
3546 3268
3547 # echo 1 > events/sched/enable 3269 # echo 1 > events/sched/enable
3548 # echo 1 > snapshot 3270 # echo 1 > snapshot
3549 # cat snapshot 3271 # cat snapshot
3550 # tracer: nop 3272 # tracer: nop
3551 # 3273 #
3552 # entries-in-buffer/entries-written: 71/71 3274 # entries-in-buffer/entries-written: 71/71 #P:8
3553 # 3275 #
3554 # _-----=> irq 3276 # _-----=> irqs-off
3555 # / _----=> nee 3277 # / _----=> need-resched
3556 # | / _---=> har 3278 # | / _---=> hardirq/softirq
3557 # || / _--=> pre 3279 # || / _--=> preempt-depth
3558 # ||| / dela 3280 # ||| / delay
3559 # TASK-PID CPU# |||| TIMEST 3281 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
3560 # | | | |||| | 3282 # | | | |||| | |
3561 <idle>-0 [005] d... 2440.603 3283 <idle>-0 [005] d... 2440.603828: sched_switch: prev_comm=swapper/5 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=snapshot-test-2 next_pid=2242 next_prio=120
3562 sleep-2242 [005] d... 2440.603 3284 sleep-2242 [005] d... 2440.603846: sched_switch: prev_comm=snapshot-test-2 prev_pid=2242 prev_prio=120 prev_state=R ==> next_comm=kworker/5:1 next_pid=60 next_prio=120
3563 [...] 3285 [...]
3564 <idle>-0 [002] d... 2440.70723 3286 <idle>-0 [002] d... 2440.707230: sched_switch: prev_comm=swapper/2 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=snapshot-test-2 next_pid=2229 next_prio=120
3565 3287
3566 # cat trace 3288 # cat trace
3567 # tracer: nop 3289 # tracer: nop
3568 # 3290 #
3569 # entries-in-buffer/entries-written: 77/77 3291 # entries-in-buffer/entries-written: 77/77 #P:8
3570 # 3292 #
3571 # _-----=> irq 3293 # _-----=> irqs-off
3572 # / _----=> nee 3294 # / _----=> need-resched
3573 # | / _---=> har 3295 # | / _---=> hardirq/softirq
3574 # || / _--=> pre 3296 # || / _--=> preempt-depth
3575 # ||| / dela 3297 # ||| / delay
3576 # TASK-PID CPU# |||| TIMEST 3298 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
3577 # | | | |||| | 3299 # | | | |||| | |
3578 <idle>-0 [007] d... 2440.707 3300 <idle>-0 [007] d... 2440.707395: sched_switch: prev_comm=swapper/7 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=snapshot-test-2 next_pid=2243 next_prio=120
3579 snapshot-test-2-2229 [002] d... 2440.707 3301 snapshot-test-2-2229 [002] d... 2440.707438: sched_switch: prev_comm=snapshot-test-2 prev_pid=2229 prev_prio=120 prev_state=S ==> next_comm=swapper/2 next_pid=0 next_prio=120
3580 [...] 3302 [...]
3581 3303
3582 3304
3583 If you try to use this snapshot feature when 3305 If you try to use this snapshot feature when current tracer is
3584 one of the latency tracers, you will get the 3306 one of the latency tracers, you will get the following results.
3585 :: 3307 ::
3586 3308
3587 # echo wakeup > current_tracer 3309 # echo wakeup > current_tracer
3588 # echo 1 > snapshot 3310 # echo 1 > snapshot
3589 bash: echo: write error: Device or resource 3311 bash: echo: write error: Device or resource busy
3590 # cat snapshot 3312 # cat snapshot
3591 cat: snapshot: Device or resource busy 3313 cat: snapshot: Device or resource busy
3592 3314
3593 3315
3594 Instances 3316 Instances
3595 --------- 3317 ---------
3596 In the tracefs tracing directory, there is a !! 3318 In the tracefs tracing directory is a directory called "instances".
3597 This directory can have new directories creat 3319 This directory can have new directories created inside of it using
3598 mkdir, and removing directories with rmdir. T 3320 mkdir, and removing directories with rmdir. The directory created
3599 with mkdir in this directory will already con 3321 with mkdir in this directory will already contain files and other
3600 directories after it is created. 3322 directories after it is created.
3601 :: 3323 ::
3602 3324
3603 # mkdir instances/foo 3325 # mkdir instances/foo
3604 # ls instances/foo 3326 # ls instances/foo
3605 buffer_size_kb buffer_total_size_kb event 3327 buffer_size_kb buffer_total_size_kb events free_buffer per_cpu
3606 set_event snapshot trace trace_clock tr 3328 set_event snapshot trace trace_clock trace_marker trace_options
3607 trace_pipe tracing_on 3329 trace_pipe tracing_on
3608 3330
3609 As you can see, the new directory looks simil 3331 As you can see, the new directory looks similar to the tracing directory
3610 itself. In fact, it is very similar, except t 3332 itself. In fact, it is very similar, except that the buffer and
3611 events are agnostic from the main directory, 3333 events are agnostic from the main directory, or from any other
3612 instances that are created. 3334 instances that are created.
3613 3335
3614 The files in the new directory work just like 3336 The files in the new directory work just like the files with the
3615 same name in the tracing directory except the 3337 same name in the tracing directory except the buffer that is used
3616 is a separate and new buffer. The files affec 3338 is a separate and new buffer. The files affect that buffer but do not
3617 affect the main buffer with the exception of 3339 affect the main buffer with the exception of trace_options. Currently,
3618 the trace_options affect all instances and th 3340 the trace_options affect all instances and the top level buffer
3619 the same, but this may change in future relea 3341 the same, but this may change in future releases. That is, options
3620 may become specific to the instance they resi 3342 may become specific to the instance they reside in.
3621 3343
3622 Notice that none of the function tracer files 3344 Notice that none of the function tracer files are there, nor is
3623 current_tracer and available_tracers. This is 3345 current_tracer and available_tracers. This is because the buffers
3624 can currently only have events enabled for th 3346 can currently only have events enabled for them.
3625 :: 3347 ::
3626 3348
3627 # mkdir instances/foo 3349 # mkdir instances/foo
3628 # mkdir instances/bar 3350 # mkdir instances/bar
3629 # mkdir instances/zoot 3351 # mkdir instances/zoot
3630 # echo 100000 > buffer_size_kb 3352 # echo 100000 > buffer_size_kb
3631 # echo 1000 > instances/foo/buffer_size_kb 3353 # echo 1000 > instances/foo/buffer_size_kb
3632 # echo 5000 > instances/bar/per_cpu/cpu1/bu 3354 # echo 5000 > instances/bar/per_cpu/cpu1/buffer_size_kb
3633 # echo function > current_trace 3355 # echo function > current_trace
3634 # echo 1 > instances/foo/events/sched/sched 3356 # echo 1 > instances/foo/events/sched/sched_wakeup/enable
3635 # echo 1 > instances/foo/events/sched/sched 3357 # echo 1 > instances/foo/events/sched/sched_wakeup_new/enable
3636 # echo 1 > instances/foo/events/sched/sched 3358 # echo 1 > instances/foo/events/sched/sched_switch/enable
3637 # echo 1 > instances/bar/events/irq/enable 3359 # echo 1 > instances/bar/events/irq/enable
3638 # echo 1 > instances/zoot/events/syscalls/e 3360 # echo 1 > instances/zoot/events/syscalls/enable
3639 # cat trace_pipe 3361 # cat trace_pipe
3640 CPU:2 [LOST 11745 EVENTS] 3362 CPU:2 [LOST 11745 EVENTS]
3641 bash-2044 [002] .... 10594.481 3363 bash-2044 [002] .... 10594.481032: _raw_spin_lock_irqsave <-get_page_from_freelist
3642 bash-2044 [002] d... 10594.481 3364 bash-2044 [002] d... 10594.481032: add_preempt_count <-_raw_spin_lock_irqsave
3643 bash-2044 [002] d..1 10594.481 3365 bash-2044 [002] d..1 10594.481032: __rmqueue <-get_page_from_freelist
3644 bash-2044 [002] d..1 10594.481 3366 bash-2044 [002] d..1 10594.481033: _raw_spin_unlock <-get_page_from_freelist
3645 bash-2044 [002] d..1 10594.481 3367 bash-2044 [002] d..1 10594.481033: sub_preempt_count <-_raw_spin_unlock
3646 bash-2044 [002] d... 10594.481 3368 bash-2044 [002] d... 10594.481033: get_pageblock_flags_group <-get_pageblock_migratetype
3647 bash-2044 [002] d... 10594.481 3369 bash-2044 [002] d... 10594.481034: __mod_zone_page_state <-get_page_from_freelist
3648 bash-2044 [002] d... 10594.481 3370 bash-2044 [002] d... 10594.481034: zone_statistics <-get_page_from_freelist
3649 bash-2044 [002] d... 10594.481 3371 bash-2044 [002] d... 10594.481034: __inc_zone_state <-zone_statistics
3650 bash-2044 [002] d... 10594.481 3372 bash-2044 [002] d... 10594.481034: __inc_zone_state <-zone_statistics
3651 bash-2044 [002] .... 10594.481 3373 bash-2044 [002] .... 10594.481035: arch_dup_task_struct <-copy_process
3652 [...] 3374 [...]
3653 3375
3654 # cat instances/foo/trace_pipe 3376 # cat instances/foo/trace_pipe
3655 bash-1998 [000] d..4 136.676 3377 bash-1998 [000] d..4 136.676759: sched_wakeup: comm=kworker/0:1 pid=59 prio=120 success=1 target_cpu=000
3656 bash-1998 [000] dN.4 136.676 3378 bash-1998 [000] dN.4 136.676760: sched_wakeup: comm=bash pid=1998 prio=120 success=1 target_cpu=000
3657 <idle>-0 [003] d.h3 136.676 3379 <idle>-0 [003] d.h3 136.676906: sched_wakeup: comm=rcu_preempt pid=9 prio=120 success=1 target_cpu=003
3658 <idle>-0 [003] d..3 136.676 3380 <idle>-0 [003] d..3 136.676909: sched_switch: prev_comm=swapper/3 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=rcu_preempt next_pid=9 next_prio=120
3659 rcu_preempt-9 [003] d..3 136.676 3381 rcu_preempt-9 [003] d..3 136.676916: sched_switch: prev_comm=rcu_preempt prev_pid=9 prev_prio=120 prev_state=S ==> next_comm=swapper/3 next_pid=0 next_prio=120
3660 bash-1998 [000] d..4 136.677 3382 bash-1998 [000] d..4 136.677014: sched_wakeup: comm=kworker/0:1 pid=59 prio=120 success=1 target_cpu=000
3661 bash-1998 [000] dN.4 136.677 3383 bash-1998 [000] dN.4 136.677016: sched_wakeup: comm=bash pid=1998 prio=120 success=1 target_cpu=000
3662 bash-1998 [000] d..3 136.677 3384 bash-1998 [000] d..3 136.677018: sched_switch: prev_comm=bash prev_pid=1998 prev_prio=120 prev_state=R+ ==> next_comm=kworker/0:1 next_pid=59 next_prio=120
3663 kworker/0:1-59 [000] d..4 136.677 3385 kworker/0:1-59 [000] d..4 136.677022: sched_wakeup: comm=sshd pid=1995 prio=120 success=1 target_cpu=001
3664 kworker/0:1-59 [000] d..3 136.677 3386 kworker/0:1-59 [000] d..3 136.677025: sched_switch: prev_comm=kworker/0:1 prev_pid=59 prev_prio=120 prev_state=S ==> next_comm=bash next_pid=1998 next_prio=120
3665 [...] 3387 [...]
3666 3388
3667 # cat instances/bar/trace_pipe 3389 # cat instances/bar/trace_pipe
3668 migration/1-14 [001] d.h3 138.732 3390 migration/1-14 [001] d.h3 138.732674: softirq_raise: vec=3 [action=NET_RX]
3669 <idle>-0 [001] dNh3 138.732 3391 <idle>-0 [001] dNh3 138.732725: softirq_raise: vec=3 [action=NET_RX]
3670 bash-1998 [000] d.h1 138.733 3392 bash-1998 [000] d.h1 138.733101: softirq_raise: vec=1 [action=TIMER]
3671 bash-1998 [000] d.h1 138.733 3393 bash-1998 [000] d.h1 138.733102: softirq_raise: vec=9 [action=RCU]
3672 bash-1998 [000] ..s2 138.733 3394 bash-1998 [000] ..s2 138.733105: softirq_entry: vec=1 [action=TIMER]
3673 bash-1998 [000] ..s2 138.733 3395 bash-1998 [000] ..s2 138.733106: softirq_exit: vec=1 [action=TIMER]
3674 bash-1998 [000] ..s2 138.733 3396 bash-1998 [000] ..s2 138.733106: softirq_entry: vec=9 [action=RCU]
3675 bash-1998 [000] ..s2 138.733 3397 bash-1998 [000] ..s2 138.733109: softirq_exit: vec=9 [action=RCU]
3676 sshd-1995 [001] d.h1 138.733 3398 sshd-1995 [001] d.h1 138.733278: irq_handler_entry: irq=21 name=uhci_hcd:usb4
3677 sshd-1995 [001] d.h1 138.733 3399 sshd-1995 [001] d.h1 138.733280: irq_handler_exit: irq=21 ret=unhandled
3678 sshd-1995 [001] d.h1 138.733 3400 sshd-1995 [001] d.h1 138.733281: irq_handler_entry: irq=21 name=eth0
3679 sshd-1995 [001] d.h1 138.733 3401 sshd-1995 [001] d.h1 138.733283: irq_handler_exit: irq=21 ret=handled
3680 [...] 3402 [...]
3681 3403
3682 # cat instances/zoot/trace 3404 # cat instances/zoot/trace
3683 # tracer: nop 3405 # tracer: nop
3684 # 3406 #
3685 # entries-in-buffer/entries-written: 18996/ 3407 # entries-in-buffer/entries-written: 18996/18996 #P:4
3686 # 3408 #
3687 # _-----=> irq 3409 # _-----=> irqs-off
3688 # / _----=> nee 3410 # / _----=> need-resched
3689 # | / _---=> har 3411 # | / _---=> hardirq/softirq
3690 # || / _--=> pre 3412 # || / _--=> preempt-depth
3691 # ||| / dela 3413 # ||| / delay
3692 # TASK-PID CPU# |||| TIMEST 3414 # TASK-PID CPU# |||| TIMESTAMP FUNCTION
3693 # | | | |||| | 3415 # | | | |||| | |
3694 bash-1998 [000] d... 140.733 3416 bash-1998 [000] d... 140.733501: sys_write -> 0x2
3695 bash-1998 [000] d... 140.733 3417 bash-1998 [000] d... 140.733504: sys_dup2(oldfd: a, newfd: 1)
3696 bash-1998 [000] d... 140.733 3418 bash-1998 [000] d... 140.733506: sys_dup2 -> 0x1
3697 bash-1998 [000] d... 140.733 3419 bash-1998 [000] d... 140.733508: sys_fcntl(fd: a, cmd: 1, arg: 0)
3698 bash-1998 [000] d... 140.733 3420 bash-1998 [000] d... 140.733509: sys_fcntl -> 0x1
3699 bash-1998 [000] d... 140.733 3421 bash-1998 [000] d... 140.733510: sys_close(fd: a)
3700 bash-1998 [000] d... 140.733 3422 bash-1998 [000] d... 140.733510: sys_close -> 0x0
3701 bash-1998 [000] d... 140.733 3423 bash-1998 [000] d... 140.733514: sys_rt_sigprocmask(how: 0, nset: 0, oset: 6e2768, sigsetsize: 8)
3702 bash-1998 [000] d... 140.733 3424 bash-1998 [000] d... 140.733515: sys_rt_sigprocmask -> 0x0
3703 bash-1998 [000] d... 140.733 3425 bash-1998 [000] d... 140.733516: sys_rt_sigaction(sig: 2, act: 7fff718846f0, oact: 7fff71884650, sigsetsize: 8)
3704 bash-1998 [000] d... 140.733 3426 bash-1998 [000] d... 140.733516: sys_rt_sigaction -> 0x0
3705 3427
3706 You can see that the trace of the top most tr 3428 You can see that the trace of the top most trace buffer shows only
3707 the function tracing. The foo instance displa 3429 the function tracing. The foo instance displays wakeups and task
3708 switches. 3430 switches.
3709 3431
3710 To remove the instances, simply delete their 3432 To remove the instances, simply delete their directories:
3711 :: 3433 ::
3712 3434
3713 # rmdir instances/foo 3435 # rmdir instances/foo
3714 # rmdir instances/bar 3436 # rmdir instances/bar
3715 # rmdir instances/zoot 3437 # rmdir instances/zoot
3716 3438
3717 Note, if a process has a trace file open in o 3439 Note, if a process has a trace file open in one of the instance
3718 directories, the rmdir will fail with EBUSY. 3440 directories, the rmdir will fail with EBUSY.
3719 3441
3720 3442
3721 Stack trace 3443 Stack trace
3722 ----------- 3444 -----------
3723 Since the kernel has a fixed sized stack, it 3445 Since the kernel has a fixed sized stack, it is important not to
3724 waste it in functions. A kernel developer mus !! 3446 waste it in functions. A kernel developer must be conscience of
3725 what they allocate on the stack. If they add 3447 what they allocate on the stack. If they add too much, the system
3726 can be in danger of a stack overflow, and cor 3448 can be in danger of a stack overflow, and corruption will occur,
3727 usually leading to a system panic. 3449 usually leading to a system panic.
3728 3450
3729 There are some tools that check this, usually 3451 There are some tools that check this, usually with interrupts
3730 periodically checking usage. But if you can p 3452 periodically checking usage. But if you can perform a check
3731 at every function call that will become very 3453 at every function call that will become very useful. As ftrace provides
3732 a function tracer, it makes it convenient to 3454 a function tracer, it makes it convenient to check the stack size
3733 at every function call. This is enabled via t 3455 at every function call. This is enabled via the stack tracer.
3734 3456
3735 CONFIG_STACK_TRACER enables the ftrace stack 3457 CONFIG_STACK_TRACER enables the ftrace stack tracing functionality.
3736 To enable it, write a '1' into /proc/sys/kern 3458 To enable it, write a '1' into /proc/sys/kernel/stack_tracer_enabled.
3737 :: 3459 ::
3738 3460
3739 # echo 1 > /proc/sys/kernel/stack_tracer_ena 3461 # echo 1 > /proc/sys/kernel/stack_tracer_enabled
3740 3462
3741 You can also enable it from the kernel comman 3463 You can also enable it from the kernel command line to trace
3742 the stack size of the kernel during boot up, 3464 the stack size of the kernel during boot up, by adding "stacktrace"
3743 to the kernel command line parameter. 3465 to the kernel command line parameter.
3744 3466
3745 After running it for a few minutes, the outpu 3467 After running it for a few minutes, the output looks like:
3746 :: 3468 ::
3747 3469
3748 # cat stack_max_size 3470 # cat stack_max_size
3749 2928 3471 2928
3750 3472
3751 # cat stack_trace 3473 # cat stack_trace
3752 Depth Size Location (18 ent 3474 Depth Size Location (18 entries)
3753 ----- ---- -------- 3475 ----- ---- --------
3754 0) 2928 224 update_sd_lb_stats+ 3476 0) 2928 224 update_sd_lb_stats+0xbc/0x4ac
3755 1) 2704 160 find_busiest_group+ 3477 1) 2704 160 find_busiest_group+0x31/0x1f1
3756 2) 2544 256 load_balance+0xd9/0 3478 2) 2544 256 load_balance+0xd9/0x662
3757 3) 2288 80 idle_balance+0xbb/0 3479 3) 2288 80 idle_balance+0xbb/0x130
3758 4) 2208 128 __schedule+0x26e/0x 3480 4) 2208 128 __schedule+0x26e/0x5b9
3759 5) 2080 16 schedule+0x64/0x66 3481 5) 2080 16 schedule+0x64/0x66
3760 6) 2064 128 schedule_timeout+0x 3482 6) 2064 128 schedule_timeout+0x34/0xe0
3761 7) 1936 112 wait_for_common+0x9 3483 7) 1936 112 wait_for_common+0x97/0xf1
3762 8) 1824 16 wait_for_completion 3484 8) 1824 16 wait_for_completion+0x1d/0x1f
3763 9) 1808 128 flush_work+0xfe/0x1 3485 9) 1808 128 flush_work+0xfe/0x119
3764 10) 1680 16 tty_flush_to_ldisc+ 3486 10) 1680 16 tty_flush_to_ldisc+0x1e/0x20
3765 11) 1664 48 input_available_p+0 3487 11) 1664 48 input_available_p+0x1d/0x5c
3766 12) 1616 48 n_tty_poll+0x6d/0x1 3488 12) 1616 48 n_tty_poll+0x6d/0x134
3767 13) 1568 64 tty_poll+0x64/0x7f 3489 13) 1568 64 tty_poll+0x64/0x7f
3768 14) 1504 880 do_select+0x31e/0x5 3490 14) 1504 880 do_select+0x31e/0x511
3769 15) 624 400 core_sys_select+0x1 3491 15) 624 400 core_sys_select+0x177/0x216
3770 16) 224 96 sys_select+0x91/0xb 3492 16) 224 96 sys_select+0x91/0xb9
3771 17) 128 128 system_call_fastpat 3493 17) 128 128 system_call_fastpath+0x16/0x1b
3772 3494
3773 Note, if -mfentry is being used by gcc, funct 3495 Note, if -mfentry is being used by gcc, functions get traced before
3774 they set up the stack frame. This means that 3496 they set up the stack frame. This means that leaf level functions
3775 are not tested by the stack tracer when -mfen 3497 are not tested by the stack tracer when -mfentry is used.
3776 3498
3777 Currently, -mfentry is used by gcc 4.6.0 and 3499 Currently, -mfentry is used by gcc 4.6.0 and above on x86 only.
3778 3500
3779 More 3501 More
3780 ---- 3502 ----
3781 More details can be found in the source code, 3503 More details can be found in the source code, in the `kernel/trace/*.c` files.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.