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