1 perf-intel-pt(1) 1 perf-intel-pt(1)
2 ================ 2 ================
3 3
4 NAME 4 NAME
5 ---- 5 ----
6 perf-intel-pt - Support for Intel Processor Tr 6 perf-intel-pt - Support for Intel Processor Trace within perf tools
7 7
8 SYNOPSIS 8 SYNOPSIS
9 -------- 9 --------
10 [verse] 10 [verse]
11 'perf record' -e intel_pt// 11 'perf record' -e intel_pt//
12 12
13 DESCRIPTION 13 DESCRIPTION
14 ----------- 14 -----------
15 15
16 Intel Processor Trace (Intel PT) is an extensi 16 Intel Processor Trace (Intel PT) is an extension of Intel Architecture that
17 collects information about software execution 17 collects information about software execution such as control flow, execution
18 modes and timings and formats it into highly c 18 modes and timings and formats it into highly compressed binary packets.
19 Technical details are documented in the Intel 19 Technical details are documented in the Intel 64 and IA-32 Architectures
20 Software Developer Manuals, Chapter 36 Intel P 20 Software Developer Manuals, Chapter 36 Intel Processor Trace.
21 21
22 Intel PT is first supported in Intel Core M an 22 Intel PT is first supported in Intel Core M and 5th generation Intel Core
23 processors that are based on the Intel micro-a 23 processors that are based on the Intel micro-architecture code name Broadwell.
24 24
25 Trace data is collected by 'perf record' and s 25 Trace data is collected by 'perf record' and stored within the perf.data file.
26 See below for options to 'perf record'. 26 See below for options to 'perf record'.
27 27
28 Trace data must be 'decoded' which involves wa 28 Trace data must be 'decoded' which involves walking the object code and matching
29 the trace data packets. For example a TNT pack 29 the trace data packets. For example a TNT packet only tells whether a
30 conditional branch was taken or not taken, so 30 conditional branch was taken or not taken, so to make use of that packet the
31 decoder must know precisely which instruction 31 decoder must know precisely which instruction was being executed.
32 32
33 Decoding is done on-the-fly. The decoder outp 33 Decoding is done on-the-fly. The decoder outputs samples in the same format as
34 samples output by perf hardware events, for ex 34 samples output by perf hardware events, for example as though the "instructions"
35 or "branches" events had been recorded. Prese 35 or "branches" events had been recorded. Presently 3 tools support this:
36 'perf script', 'perf report' and 'perf inject' 36 'perf script', 'perf report' and 'perf inject'. See below for more information
37 on using those tools. 37 on using those tools.
38 38
39 The main distinguishing feature of Intel PT is 39 The main distinguishing feature of Intel PT is that the decoder can determine
40 the exact flow of software execution. Intel P 40 the exact flow of software execution. Intel PT can be used to understand why
41 and how did software get to a certain point, o 41 and how did software get to a certain point, or behave a certain way. The
42 software does not have to be recompiled, so In 42 software does not have to be recompiled, so Intel PT works with debug or release
43 builds, however the executed images are needed 43 builds, however the executed images are needed - which makes use in JIT-compiled
44 environments, or with self-modified code, a ch 44 environments, or with self-modified code, a challenge. Also symbols need to be
45 provided to make sense of addresses. 45 provided to make sense of addresses.
46 46
47 A limitation of Intel PT is that it produces h 47 A limitation of Intel PT is that it produces huge amounts of trace data
48 (hundreds of megabytes per second per core) wh 48 (hundreds of megabytes per second per core) which takes a long time to decode,
49 for example two or three orders of magnitude l 49 for example two or three orders of magnitude longer than it took to collect.
50 Another limitation is the performance impact o 50 Another limitation is the performance impact of tracing, something that will
51 vary depending on the use-case and architectur 51 vary depending on the use-case and architecture.
52 52
53 53
54 Quickstart 54 Quickstart
55 ---------- 55 ----------
56 56
57 It is important to start small. That is becau 57 It is important to start small. That is because it is easy to capture vastly
58 more data than can possibly be processed. 58 more data than can possibly be processed.
59 59
60 The simplest thing to do with Intel PT is user 60 The simplest thing to do with Intel PT is userspace profiling of small programs.
61 Data is captured with 'perf record' e.g. to tr 61 Data is captured with 'perf record' e.g. to trace 'ls' userspace-only:
62 62
63 perf record -e intel_pt//u ls 63 perf record -e intel_pt//u ls
64 64
65 And profiled with 'perf report' e.g. 65 And profiled with 'perf report' e.g.
66 66
67 perf report 67 perf report
68 68
69 To also trace kernel space presents a problem, 69 To also trace kernel space presents a problem, namely kernel self-modifying
70 code. A fairly good kernel image is available 70 code. A fairly good kernel image is available in /proc/kcore but to get an
71 accurate image a copy of /proc/kcore needs to 71 accurate image a copy of /proc/kcore needs to be made under the same conditions
72 as the data capture. 'perf record' can make a 72 as the data capture. 'perf record' can make a copy of /proc/kcore if the option
73 --kcore is used, but access to /proc/kcore is 73 --kcore is used, but access to /proc/kcore is restricted e.g.
74 74
75 sudo perf record -o pt_ls --kcore -e i 75 sudo perf record -o pt_ls --kcore -e intel_pt// -- ls
76 76
77 which will create a directory named 'pt_ls' an 77 which will create a directory named 'pt_ls' and put the perf.data file (named
78 simply 'data') and copies of /proc/kcore, /pro 78 simply 'data') and copies of /proc/kcore, /proc/kallsyms and /proc/modules into
79 it. The other tools understand the directory 79 it. The other tools understand the directory format, so to use 'perf report'
80 becomes: 80 becomes:
81 81
82 sudo perf report -i pt_ls 82 sudo perf report -i pt_ls
83 83
84 Because samples are synthesized after-the-fact 84 Because samples are synthesized after-the-fact, the sampling period can be
85 selected for reporting. e.g. sample every micr 85 selected for reporting. e.g. sample every microsecond
86 86
87 sudo perf report pt_ls --itrace=i1usge 87 sudo perf report pt_ls --itrace=i1usge
88 88
89 See the sections below for more information ab 89 See the sections below for more information about the --itrace option.
90 90
91 Beware the smaller the period, the more sample 91 Beware the smaller the period, the more samples that are produced, and the
92 longer it takes to process them. 92 longer it takes to process them.
93 93
94 Also note that the coarseness of Intel PT timi 94 Also note that the coarseness of Intel PT timing information will start to
95 distort the statistical value of the sampling 95 distort the statistical value of the sampling as the sampling period becomes
96 smaller. 96 smaller.
97 97
98 To represent software control flow, "branches" 98 To represent software control flow, "branches" samples are produced. By default
99 a branch sample is synthesized for every singl 99 a branch sample is synthesized for every single branch. To get an idea what
100 data is available you can use the 'perf script 100 data is available you can use the 'perf script' tool with all itrace sampling
101 options, which will list all the samples. 101 options, which will list all the samples.
102 102
103 perf record -e intel_pt//u ls 103 perf record -e intel_pt//u ls
104 perf script --itrace=iybxwpe !! 104 perf script --itrace=ibxwpe
105 105
106 An interesting field that is not printed by de 106 An interesting field that is not printed by default is 'flags' which can be
107 displayed as follows: 107 displayed as follows:
108 108
109 perf script --itrace=iybxwpe -F+flags !! 109 perf script --itrace=ibxwpe -F+flags
110 110
111 The flags are "bcrosyiABExghDt" which stand fo !! 111 The flags are "bcrosyiABEx" which stand for branch, call, return, conditional,
112 system, asynchronous, interrupt, transaction a !! 112 system, asynchronous, interrupt, transaction abort, trace begin, trace end, and
113 in transaction, VM-entry, VM-exit, interrupt d !! 113 in transaction, respectively.
114 toggle respectively. <<
115 <<
116 perf script also supports higher level ways to <<
117 <<
118 perf script --insn-trace=disasm <<
119 <<
120 or to use the xed disassembler, which requires <<
121 (see XED below): <<
122 <<
123 perf script --insn-trace --xed <<
124 <<
125 Dumping all instructions in a long trace can b <<
126 to start with higher level decoding, like <<
127 <<
128 perf script --call-trace <<
129 <<
130 or <<
131 <<
132 perf script --call-ret-trace <<
133 <<
134 and then select a time range of interest. The <<
135 in detail with <<
136 <<
137 perf script --time starttime,stoptime <<
138 <<
139 While examining the trace it's also useful to <<
140 the -C option <<
141 <<
142 perf script --time starttime,stoptime <<
143 <<
144 Dump all instructions in time range on CPU 1. <<
145 114
146 Another interesting field that is not printed 115 Another interesting field that is not printed by default is 'ipc' which can be
147 displayed as follows: 116 displayed as follows:
148 117
149 perf script --itrace=be -F+ipc 118 perf script --itrace=be -F+ipc
150 119
151 There are two ways that instructions-per-cycle 120 There are two ways that instructions-per-cycle (IPC) can be calculated depending
152 on the recording. 121 on the recording.
153 122
154 If the 'cyc' config term (see config terms sec !! 123 If the 'cyc' config term (see config terms section below) was used, then IPC is
155 and cycle events are calculated using the cycl !! 124 calculated using the cycle count from CYC packets, otherwise MTC packets are
156 MTC packets are used - refer to the 'mtc' conf !! 125 used - refer to the 'mtc' config term. When MTC is used, however, the values
157 the values are less accurate because the timin !! 126 are less accurate because the timing is less accurate.
158 127
159 Because Intel PT does not update the cycle cou 128 Because Intel PT does not update the cycle count on every branch or instruction,
160 the values will often be zero. When there are 129 the values will often be zero. When there are values, they will be the number
161 of instructions and number of cycles since the 130 of instructions and number of cycles since the last update, and thus represent
162 the average IPC cycle count since the last IPC !! 131 the average IPC since the last IPC for that event type. Note IPC for "branches"
163 Note IPC for "branches" events is calculated s !! 132 events is calculated separately from IPC for "instructions" events.
164 events. <<
165 <<
166 Even with the 'cyc' config term, it is possibl <<
167 every change of timestamp, but at the expense <<
168 specifying the itrace 'A' option. Due to the <<
169 actual number of cycles increases even though <<
170 The number of instructions is known, but if IP <<
171 low and so IPC is too high. Note that inaccur <<
172 sampling increases i.e. if the number of cycle <<
173 that becomes less significant if the number of <<
174 useful to use the 'A' option in conjunction wi <<
175 provide higher granularity cycle information. <<
176 133
177 Also note that the IPC instruction count may o 134 Also note that the IPC instruction count may or may not include the current
178 instruction. If the cycle count is associated 135 instruction. If the cycle count is associated with an asynchronous branch
179 (e.g. page fault or interrupt), then the instr 136 (e.g. page fault or interrupt), then the instruction count does not include the
180 current instruction, otherwise it does. That 137 current instruction, otherwise it does. That is consistent with whether or not
181 that instruction has retired when the cycle co 138 that instruction has retired when the cycle count is updated.
182 139
183 Another note, in the case of "branches" events 140 Another note, in the case of "branches" events, non-taken branches are not
184 presently sampled, so IPC values for them do n 141 presently sampled, so IPC values for them do not appear e.g. a CYC packet with a
185 TNT packet that starts with a non-taken branch 142 TNT packet that starts with a non-taken branch. To see every possible IPC
186 value, "instructions" events can be used e.g. 143 value, "instructions" events can be used e.g. --itrace=i0ns
187 144
188 While it is possible to create scripts to anal 145 While it is possible to create scripts to analyze the data, an alternative
189 approach is available to export the data to a 146 approach is available to export the data to a sqlite or postgresql database.
190 Refer to script export-to-sqlite.py or export- 147 Refer to script export-to-sqlite.py or export-to-postgresql.py for more details,
191 and to script exported-sql-viewer.py for an ex 148 and to script exported-sql-viewer.py for an example of using the database.
192 149
193 There is also script intel-pt-events.py which 150 There is also script intel-pt-events.py which provides an example of how to
194 unpack the raw data for power events and PTWRI !! 151 unpack the raw data for power events and PTWRITE.
195 branches, and supports 2 additional modes sele <<
196 <<
197 - --insn-trace - instruction trace <<
198 - --src-trace - source trace <<
199 <<
200 The intel-pt-events.py script also has options <<
201 <<
202 - --all-switch-events - display all switch ev <<
203 - --interleave [<n>] - interleave sample outp <<
204 no more than n samples for a CPU are displaye <<
205 Note this only affects the order of output, a <<
206 same. <<
207 152
208 As mentioned above, it is easy to capture too 153 As mentioned above, it is easy to capture too much data. One way to limit the
209 data captured is to use 'snapshot' mode which 154 data captured is to use 'snapshot' mode which is explained further below.
210 Refer to 'new snapshot option' and 'Intel PT m 155 Refer to 'new snapshot option' and 'Intel PT modes of operation' further below.
211 156
212 Another problem that will be experienced is de 157 Another problem that will be experienced is decoder errors. They can be caused
213 by inability to access the executed image, sel 158 by inability to access the executed image, self-modified or JIT-ed code, or the
214 inability to match side-band information (such 159 inability to match side-band information (such as context switches and mmaps)
215 which results in the decoder not knowing what 160 which results in the decoder not knowing what code was executed.
216 161
217 There is also the problem of perf not being ab 162 There is also the problem of perf not being able to copy the data fast enough,
218 resulting in data lost because the buffer was 163 resulting in data lost because the buffer was full. See 'Buffer handling' below
219 for more details. 164 for more details.
220 165
221 166
222 perf record 167 perf record
223 ----------- 168 -----------
224 169
225 new event 170 new event
226 ~~~~~~~~~ 171 ~~~~~~~~~
227 172
228 The Intel PT kernel driver creates a new PMU f 173 The Intel PT kernel driver creates a new PMU for Intel PT. PMU events are
229 selected by providing the PMU name followed by 174 selected by providing the PMU name followed by the "config" separated by slashes.
230 An enhancement has been made to allow default 175 An enhancement has been made to allow default "config" e.g. the option
231 176
232 -e intel_pt// 177 -e intel_pt//
233 178
234 will use a default config value. Currently th 179 will use a default config value. Currently that is the same as
235 180
236 -e intel_pt/tsc,noretcomp=0/ 181 -e intel_pt/tsc,noretcomp=0/
237 182
238 which is the same as 183 which is the same as
239 184
240 -e intel_pt/tsc=1,noretcomp=0/ 185 -e intel_pt/tsc=1,noretcomp=0/
241 186
242 Note there are now new config terms - see sect 187 Note there are now new config terms - see section 'config terms' further below.
243 188
244 The config terms are listed in /sys/devices/in 189 The config terms are listed in /sys/devices/intel_pt/format. They are bit
245 fields within the config member of the struct 190 fields within the config member of the struct perf_event_attr which is
246 passed to the kernel by the perf_event_open sy 191 passed to the kernel by the perf_event_open system call. They correspond to bit
247 fields in the IA32_RTIT_CTL MSR. Here is a li 192 fields in the IA32_RTIT_CTL MSR. Here is a list of them and their definitions:
248 193
249 $ grep -H . /sys/bus/event_source/devi 194 $ grep -H . /sys/bus/event_source/devices/intel_pt/format/*
250 /sys/bus/event_source/devices/intel_pt 195 /sys/bus/event_source/devices/intel_pt/format/cyc:config:1
251 /sys/bus/event_source/devices/intel_pt 196 /sys/bus/event_source/devices/intel_pt/format/cyc_thresh:config:19-22
252 /sys/bus/event_source/devices/intel_pt 197 /sys/bus/event_source/devices/intel_pt/format/mtc:config:9
253 /sys/bus/event_source/devices/intel_pt 198 /sys/bus/event_source/devices/intel_pt/format/mtc_period:config:14-17
254 /sys/bus/event_source/devices/intel_pt 199 /sys/bus/event_source/devices/intel_pt/format/noretcomp:config:11
255 /sys/bus/event_source/devices/intel_pt 200 /sys/bus/event_source/devices/intel_pt/format/psb_period:config:24-27
256 /sys/bus/event_source/devices/intel_pt 201 /sys/bus/event_source/devices/intel_pt/format/tsc:config:10
257 202
258 Note that the default config must be overridde 203 Note that the default config must be overridden for each term i.e.
259 204
260 -e intel_pt/noretcomp=0/ 205 -e intel_pt/noretcomp=0/
261 206
262 is the same as: 207 is the same as:
263 208
264 -e intel_pt/tsc=1,noretcomp=0/ 209 -e intel_pt/tsc=1,noretcomp=0/
265 210
266 So, to disable TSC packets use: 211 So, to disable TSC packets use:
267 212
268 -e intel_pt/tsc=0/ 213 -e intel_pt/tsc=0/
269 214
270 It is also possible to specify the config valu 215 It is also possible to specify the config value explicitly:
271 216
272 -e intel_pt/config=0x400/ 217 -e intel_pt/config=0x400/
273 218
274 Note that, as with all events, the event is su 219 Note that, as with all events, the event is suffixed with event modifiers:
275 220
276 u userspace 221 u userspace
277 k kernel 222 k kernel
278 h hypervisor 223 h hypervisor
279 G guest 224 G guest
280 H host 225 H host
281 p precise ip 226 p precise ip
282 227
283 'h', 'G' and 'H' are for virtualization which !! 228 'h', 'G' and 'H' are for virtualization which is not supported by Intel PT.
284 'p' is also not relevant to Intel PT. So only 229 'p' is also not relevant to Intel PT. So only options 'u' and 'k' are
285 meaningful for Intel PT. 230 meaningful for Intel PT.
286 231
287 perf_event_attr is displayed if the -vv option 232 perf_event_attr is displayed if the -vv option is used e.g.
288 233
289 -------------------------------------- 234 ------------------------------------------------------------
290 perf_event_attr: 235 perf_event_attr:
291 type 6 236 type 6
292 size 112 237 size 112
293 config 0x400 238 config 0x400
294 { sample_period, sample_freq } 1 239 { sample_period, sample_freq } 1
295 sample_type IP|TI 240 sample_type IP|TID|TIME|CPU|IDENTIFIER
296 read_format ID 241 read_format ID
297 disabled 1 242 disabled 1
298 inherit 1 243 inherit 1
299 exclude_kernel 1 244 exclude_kernel 1
300 exclude_hv 1 245 exclude_hv 1
301 enable_on_exec 1 246 enable_on_exec 1
302 sample_id_all 1 247 sample_id_all 1
303 -------------------------------------- 248 ------------------------------------------------------------
304 sys_perf_event_open: pid 31104 cpu 0 249 sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
305 sys_perf_event_open: pid 31104 cpu 1 250 sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
306 sys_perf_event_open: pid 31104 cpu 2 251 sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
307 sys_perf_event_open: pid 31104 cpu 3 252 sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
308 -------------------------------------- 253 ------------------------------------------------------------
309 254
310 255
311 config terms 256 config terms
312 ~~~~~~~~~~~~ 257 ~~~~~~~~~~~~
313 258
314 The June 2015 version of Intel 64 and IA-32 Ar 259 The June 2015 version of Intel 64 and IA-32 Architectures Software Developer
315 Manuals, Chapter 36 Intel Processor Trace, def 260 Manuals, Chapter 36 Intel Processor Trace, defined new Intel PT features.
316 Some of the features are reflect in new config 261 Some of the features are reflect in new config terms. All the config terms are
317 described below. 262 described below.
318 263
319 tsc Always supported. Produces TS 264 tsc Always supported. Produces TSC timestamp packets to provide
320 timing information. In some c 265 timing information. In some cases it is possible to decode
321 without timing information, fo 266 without timing information, for example a per-thread context
322 that does not overlap executab 267 that does not overlap executable memory maps.
323 268
324 The default config selects tsc 269 The default config selects tsc (i.e. tsc=1).
325 270
326 noretcomp Always supported. Disables "r 271 noretcomp Always supported. Disables "return compression" so a TIP packet
327 is produced when a function re 272 is produced when a function returns. Causes more packets to be
328 produced but might make decodi 273 produced but might make decoding more reliable.
329 274
330 The default config does not se 275 The default config does not select noretcomp (i.e. noretcomp=0).
331 276
332 psb_period Allows the frequency of PSB pa 277 psb_period Allows the frequency of PSB packets to be specified.
333 278
334 The PSB packet is a synchroniz 279 The PSB packet is a synchronization packet that provides a
335 starting point for decoding or 280 starting point for decoding or recovery from errors.
336 281
337 Support for psb_period is indi 282 Support for psb_period is indicated by:
338 283
339 /sys/bus/event_source/ 284 /sys/bus/event_source/devices/intel_pt/caps/psb_cyc
340 285
341 which contains "1" if the feat 286 which contains "1" if the feature is supported and "0"
342 otherwise. 287 otherwise.
343 288
344 Valid values are given by: 289 Valid values are given by:
345 290
346 /sys/bus/event_source/ 291 /sys/bus/event_source/devices/intel_pt/caps/psb_periods
347 292
348 which contains a hexadecimal v 293 which contains a hexadecimal value, the bits of which represent
349 valid values e.g. bit 2 set me 294 valid values e.g. bit 2 set means value 2 is valid.
350 295
351 The psb_period value is conver 296 The psb_period value is converted to the approximate number of
352 trace bytes between PSB packet 297 trace bytes between PSB packets as:
353 298
354 2 ^ (value + 11) 299 2 ^ (value + 11)
355 300
356 e.g. value 3 means 16KiB bytes 301 e.g. value 3 means 16KiB bytes between PSBs
357 302
358 If an invalid value is entered 303 If an invalid value is entered, the error message
359 will give a list of valid valu 304 will give a list of valid values e.g.
360 305
361 $ perf record -e intel 306 $ perf record -e intel_pt/psb_period=15/u uname
362 Invalid psb_period for 307 Invalid psb_period for intel_pt. Valid values are: 0-5
363 308
364 If MTC packets are selected, t 309 If MTC packets are selected, the default config selects a value
365 of 3 (i.e. psb_period=3) or th 310 of 3 (i.e. psb_period=3) or the nearest lower value that is
366 supported (0 is always support 311 supported (0 is always supported). Otherwise the default is 0.
367 312
368 If decoding is expected to be 313 If decoding is expected to be reliable and the buffer is large
369 then a large PSB period can be 314 then a large PSB period can be used.
370 315
371 Because a TSC packet is produc 316 Because a TSC packet is produced with PSB, the PSB period can
372 also affect the granularity to 317 also affect the granularity to timing information in the absence
373 of MTC or CYC. 318 of MTC or CYC.
374 319
375 mtc Produces MTC timing packets. 320 mtc Produces MTC timing packets.
376 321
377 MTC packets provide finer grai 322 MTC packets provide finer grain timestamp information than TSC
378 packets. MTC packets record t 323 packets. MTC packets record time using the hardware crystal
379 clock (CTC) which is related t 324 clock (CTC) which is related to TSC packets using a TMA packet.
380 325
381 Support for this feature is in 326 Support for this feature is indicated by:
382 327
383 /sys/bus/event_source/ 328 /sys/bus/event_source/devices/intel_pt/caps/mtc
384 329
385 which contains "1" if the feat 330 which contains "1" if the feature is supported and
386 "0" otherwise. 331 "0" otherwise.
387 332
388 The frequency of MTC packets c 333 The frequency of MTC packets can also be specified - see
389 mtc_period below. 334 mtc_period below.
390 335
391 mtc_period Specifies how frequently MTC p 336 mtc_period Specifies how frequently MTC packets are produced - see mtc
392 above for how to determine if 337 above for how to determine if MTC packets are supported.
393 338
394 Valid values are given by: 339 Valid values are given by:
395 340
396 /sys/bus/event_source/ 341 /sys/bus/event_source/devices/intel_pt/caps/mtc_periods
397 342
398 which contains a hexadecimal v 343 which contains a hexadecimal value, the bits of which represent
399 valid values e.g. bit 2 set me 344 valid values e.g. bit 2 set means value 2 is valid.
400 345
401 The mtc_period value is conver 346 The mtc_period value is converted to the MTC frequency as:
402 347
403 CTC-frequency / (2 ^ v 348 CTC-frequency / (2 ^ value)
404 349
405 e.g. value 3 means one eighth 350 e.g. value 3 means one eighth of CTC-frequency
406 351
407 Where CTC is the hardware crys 352 Where CTC is the hardware crystal clock, the frequency of which
408 can be related to TSC via valu 353 can be related to TSC via values provided in cpuid leaf 0x15.
409 354
410 If an invalid value is entered 355 If an invalid value is entered, the error message
411 will give a list of valid valu 356 will give a list of valid values e.g.
412 357
413 $ perf record -e intel 358 $ perf record -e intel_pt/mtc_period=15/u uname
414 Invalid mtc_period for 359 Invalid mtc_period for intel_pt. Valid values are: 0,3,6,9
415 360
416 The default value is 3 or the 361 The default value is 3 or the nearest lower value
417 that is supported (0 is always 362 that is supported (0 is always supported).
418 363
419 cyc Produces CYC timing packets. 364 cyc Produces CYC timing packets.
420 365
421 CYC packets provide even finer 366 CYC packets provide even finer grain timestamp information than
422 MTC and TSC packets. A CYC pa 367 MTC and TSC packets. A CYC packet contains the number of CPU
423 cycles since the last CYC pack 368 cycles since the last CYC packet. Unlike MTC and TSC packets,
424 CYC packets are only sent when 369 CYC packets are only sent when another packet is also sent.
425 370
426 Support for this feature is in 371 Support for this feature is indicated by:
427 372
428 /sys/bus/event_source/ 373 /sys/bus/event_source/devices/intel_pt/caps/psb_cyc
429 374
430 which contains "1" if the feat 375 which contains "1" if the feature is supported and
431 "0" otherwise. 376 "0" otherwise.
432 377
433 The number of CYC packets prod 378 The number of CYC packets produced can be reduced by specifying
434 a threshold - see cyc_thresh b 379 a threshold - see cyc_thresh below.
435 380
436 cyc_thresh Specifies how frequently CYC p 381 cyc_thresh Specifies how frequently CYC packets are produced - see cyc
437 above for how to determine if 382 above for how to determine if CYC packets are supported.
438 383
439 Valid cyc_thresh values are gi 384 Valid cyc_thresh values are given by:
440 385
441 /sys/bus/event_source/ 386 /sys/bus/event_source/devices/intel_pt/caps/cycle_thresholds
442 387
443 which contains a hexadecimal v 388 which contains a hexadecimal value, the bits of which represent
444 valid values e.g. bit 2 set me 389 valid values e.g. bit 2 set means value 2 is valid.
445 390
446 The cyc_thresh value represent 391 The cyc_thresh value represents the minimum number of CPU cycles
447 that must have passed before a 392 that must have passed before a CYC packet can be sent. The
448 number of CPU cycles is: 393 number of CPU cycles is:
449 394
450 2 ^ (value - 1) 395 2 ^ (value - 1)
451 396
452 e.g. value 4 means 8 CPU cycle 397 e.g. value 4 means 8 CPU cycles must pass before a CYC packet
453 can be sent. Note a CYC packe 398 can be sent. Note a CYC packet is still only sent when another
454 packet is sent, not at, e.g. e 399 packet is sent, not at, e.g. every 8 CPU cycles.
455 400
456 If an invalid value is entered 401 If an invalid value is entered, the error message
457 will give a list of valid valu 402 will give a list of valid values e.g.
458 403
459 $ perf record -e intel 404 $ perf record -e intel_pt/cyc,cyc_thresh=15/u uname
460 Invalid cyc_thresh for 405 Invalid cyc_thresh for intel_pt. Valid values are: 0-12
461 406
462 CYC packets are not requested 407 CYC packets are not requested by default.
463 408
464 pt Specifies pass-through which e 409 pt Specifies pass-through which enables the 'branch' config term.
465 410
466 The default config selects 'pt 411 The default config selects 'pt' if it is available, so a user will
467 never need to specify this ter 412 never need to specify this term.
468 413
469 branch Enable branch tracing. Branch 414 branch Enable branch tracing. Branch tracing is enabled by default so to
470 disable branch tracing use 'br 415 disable branch tracing use 'branch=0'.
471 416
472 The default config selects 'br 417 The default config selects 'branch' if it is available.
473 418
474 ptw Enable PTWRITE packets which a 419 ptw Enable PTWRITE packets which are produced when a ptwrite instruction
475 is executed. 420 is executed.
476 421
477 Support for this feature is in 422 Support for this feature is indicated by:
478 423
479 /sys/bus/event_source/ 424 /sys/bus/event_source/devices/intel_pt/caps/ptwrite
480 425
481 which contains "1" if the feat 426 which contains "1" if the feature is supported and
482 "0" otherwise. 427 "0" otherwise.
483 428
484 As an alternative, refer to "E <<
485 <<
486 fup_on_ptw Enable a FUP packet to follow 429 fup_on_ptw Enable a FUP packet to follow the PTWRITE packet. The FUP packet
487 provides the address of the pt 430 provides the address of the ptwrite instruction. In the absence of
488 fup_on_ptw, the decoder will u 431 fup_on_ptw, the decoder will use the address of the previous branch
489 if branch tracing is enabled, 432 if branch tracing is enabled, otherwise the address will be zero.
490 Note that fup_on_ptw will work 433 Note that fup_on_ptw will work even when branch tracing is disabled.
491 434
492 pwr_evt Enable power events. The powe 435 pwr_evt Enable power events. The power events provide information about
493 changes to the CPU C-state. 436 changes to the CPU C-state.
494 437
495 Support for this feature is in 438 Support for this feature is indicated by:
496 439
497 /sys/bus/event_source/ 440 /sys/bus/event_source/devices/intel_pt/caps/power_event_trace
498 441
499 which contains "1" if the feat 442 which contains "1" if the feature is supported and
500 "0" otherwise. 443 "0" otherwise.
501 444
502 event Enable Event Trace. The event <<
503 events. <<
504 <<
505 Support for this feature is in <<
506 <<
507 /sys/bus/event_source/ <<
508 <<
509 which contains "1" if the feat <<
510 "0" otherwise. <<
511 <<
512 notnt Disable TNT packets. Without <<
513 executable code to reconstruct <<
514 and TIP.PGD packets still indi <<
515 return compression is disabled <<
516 The advantage of eliminating T <<
517 trace and corresponding tracin <<
518 <<
519 Support for this feature is in <<
520 <<
521 /sys/bus/event_source/ <<
522 <<
523 which contains "1" if the feat <<
524 "0" otherwise. <<
525 <<
526 445
527 AUX area sampling option 446 AUX area sampling option
528 ~~~~~~~~~~~~~~~~~~~~~~~~ 447 ~~~~~~~~~~~~~~~~~~~~~~~~
529 448
530 To select Intel PT "sampling" the AUX area sam 449 To select Intel PT "sampling" the AUX area sampling option can be used:
531 450
532 --aux-sample 451 --aux-sample
533 452
534 Optionally it can be followed by the sample si 453 Optionally it can be followed by the sample size in bytes e.g.
535 454
536 --aux-sample=8192 455 --aux-sample=8192
537 456
538 In addition, the Intel PT event to sample must 457 In addition, the Intel PT event to sample must be defined e.g.
539 458
540 -e intel_pt//u 459 -e intel_pt//u
541 460
542 Samples on other events will be created contai 461 Samples on other events will be created containing Intel PT data e.g. the
543 following will create Intel PT samples on the 462 following will create Intel PT samples on the branch-misses event, note the
544 events must be grouped using {}: 463 events must be grouped using {}:
545 464
546 perf record --aux-sample -e '{intel_pt 465 perf record --aux-sample -e '{intel_pt//u,branch-misses:u}'
547 466
548 An alternative to '--aux-sample' is to add the 467 An alternative to '--aux-sample' is to add the config term 'aux-sample-size' to
549 events. In this case, the grouping is implied 468 events. In this case, the grouping is implied e.g.
550 469
551 perf record -e intel_pt//u -e branch-m 470 perf record -e intel_pt//u -e branch-misses/aux-sample-size=8192/u
552 471
553 is the same as: 472 is the same as:
554 473
555 perf record -e '{intel_pt//u,branch-mi 474 perf record -e '{intel_pt//u,branch-misses/aux-sample-size=8192/u}'
556 475
557 but allows for also using an address filter e. 476 but allows for also using an address filter e.g.:
558 477
559 perf record -e intel_pt//u --filter 'f 478 perf record -e intel_pt//u --filter 'filter * @/bin/ls' -e branch-misses/aux-sample-size=8192/u -- ls
560 479
561 It is important to select a sample size that i 480 It is important to select a sample size that is big enough to contain at least
562 one PSB packet. If not a warning will be disp 481 one PSB packet. If not a warning will be displayed:
563 482
564 Intel PT sample size (%zu) may be too 483 Intel PT sample size (%zu) may be too small for PSB period (%zu)
565 484
566 The calculation used for that is: if sample_si 485 The calculation used for that is: if sample_size <= psb_period + 256 display the
567 warning. When sampling is used, psb_period de 486 warning. When sampling is used, psb_period defaults to 0 (2KiB).
568 487
569 The default sample size is 4KiB. 488 The default sample size is 4KiB.
570 489
571 The sample size is passed in aux_sample_size i 490 The sample size is passed in aux_sample_size in struct perf_event_attr. The
572 sample size is limited by the maximum event si 491 sample size is limited by the maximum event size which is 64KiB. It is
573 difficult to know how big the event might be w 492 difficult to know how big the event might be without the trace sample attached,
574 but the tool validates that the sample size is 493 but the tool validates that the sample size is not greater than 60KiB.
575 494
576 495
577 new snapshot option 496 new snapshot option
578 ~~~~~~~~~~~~~~~~~~~ 497 ~~~~~~~~~~~~~~~~~~~
579 498
580 The difference between full trace and snapshot 499 The difference between full trace and snapshot from the kernel's perspective is
581 that in full trace we don't overwrite trace da 500 that in full trace we don't overwrite trace data that the user hasn't collected
582 yet (and indicated that by advancing aux_tail) 501 yet (and indicated that by advancing aux_tail), whereas in snapshot mode we let
583 the trace run and overwrite older data in the 502 the trace run and overwrite older data in the buffer so that whenever something
584 interesting happens, we can stop it and grab a 503 interesting happens, we can stop it and grab a snapshot of what was going on
585 around that interesting moment. 504 around that interesting moment.
586 505
587 To select snapshot mode a new option has been 506 To select snapshot mode a new option has been added:
588 507
589 -S 508 -S
590 509
591 Optionally it can be followed by the snapshot 510 Optionally it can be followed by the snapshot size e.g.
592 511
593 -S0x100000 512 -S0x100000
594 513
595 The default snapshot size is the auxtrace mmap 514 The default snapshot size is the auxtrace mmap size. If neither auxtrace mmap size
596 nor snapshot size is specified, then the defau 515 nor snapshot size is specified, then the default is 4MiB for privileged users
597 (or if /proc/sys/kernel/perf_event_paranoid < 516 (or if /proc/sys/kernel/perf_event_paranoid < 0), 128KiB for unprivileged users.
598 If an unprivileged user does not specify mmap 517 If an unprivileged user does not specify mmap pages, the mmap pages will be
599 reduced as described in the 'new auxtrace mmap 518 reduced as described in the 'new auxtrace mmap size option' section below.
600 519
601 The snapshot size is displayed if the option - 520 The snapshot size is displayed if the option -vv is used e.g.
602 521
603 Intel PT snapshot size: %zu 522 Intel PT snapshot size: %zu
604 523
605 524
606 new auxtrace mmap size option 525 new auxtrace mmap size option
607 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 526 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
608 527
609 Intel PT buffer size is specified by an additi 528 Intel PT buffer size is specified by an addition to the -m option e.g.
610 529
611 -m,16 530 -m,16
612 531
613 selects a buffer size of 16 pages i.e. 64KiB. 532 selects a buffer size of 16 pages i.e. 64KiB.
614 533
615 Note that the existing functionality of -m is 534 Note that the existing functionality of -m is unchanged. The auxtrace mmap size
616 is specified by the optional addition of a com 535 is specified by the optional addition of a comma and the value.
617 536
618 The default auxtrace mmap size for Intel PT is 537 The default auxtrace mmap size for Intel PT is 4MiB/page_size for privileged users
619 (or if /proc/sys/kernel/perf_event_paranoid < 538 (or if /proc/sys/kernel/perf_event_paranoid < 0), 128KiB for unprivileged users.
620 If an unprivileged user does not specify mmap 539 If an unprivileged user does not specify mmap pages, the mmap pages will be
621 reduced from the default 512KiB/page_size to 2 540 reduced from the default 512KiB/page_size to 256KiB/page_size, otherwise the
622 user is likely to get an error as they exceed 541 user is likely to get an error as they exceed their mlock limit (Max locked
623 memory as shown in /proc/self/limits). Note t 542 memory as shown in /proc/self/limits). Note that perf does not count the first
624 512KiB (actually /proc/sys/kernel/perf_event_m 543 512KiB (actually /proc/sys/kernel/perf_event_mlock_kb minus 1 page) per cpu
625 against the mlock limit so an unprivileged use 544 against the mlock limit so an unprivileged user is allowed 512KiB per cpu plus
626 their mlock limit (which defaults to 64KiB but 545 their mlock limit (which defaults to 64KiB but is not multiplied by the number
627 of cpus). 546 of cpus).
628 547
629 In full-trace mode, powers of two are allowed 548 In full-trace mode, powers of two are allowed for buffer size, with a minimum
630 size of 2 pages. In snapshot mode or sampling 549 size of 2 pages. In snapshot mode or sampling mode, it is the same but the
631 minimum size is 1 page. 550 minimum size is 1 page.
632 551
633 The mmap size and auxtrace mmap size are displ 552 The mmap size and auxtrace mmap size are displayed if the -vv option is used e.g.
634 553
635 mmap length 528384 554 mmap length 528384
636 auxtrace mmap length 4198400 555 auxtrace mmap length 4198400
637 556
638 557
639 Intel PT modes of operation 558 Intel PT modes of operation
640 ~~~~~~~~~~~~~~~~~~~~~~~~~~~ 559 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
641 560
642 Intel PT can be used in 3 modes: !! 561 Intel PT can be used in 2 modes:
643 full-trace mode 562 full-trace mode
644 sample mode 563 sample mode
645 snapshot mode 564 snapshot mode
646 565
647 Full-trace mode traces continuously e.g. 566 Full-trace mode traces continuously e.g.
648 567
649 perf record -e intel_pt//u uname 568 perf record -e intel_pt//u uname
650 569
651 Sample mode attaches a Intel PT sample to othe 570 Sample mode attaches a Intel PT sample to other events e.g.
652 571
653 perf record --aux-sample -e intel_pt// 572 perf record --aux-sample -e intel_pt//u -e branch-misses:u
654 573
655 Snapshot mode captures the available data when !! 574 Snapshot mode captures the available data when a signal is sent e.g.
656 control command is issued. e.g. using a signal <<
657 575
658 perf record -v -e intel_pt//u -S ./loo 576 perf record -v -e intel_pt//u -S ./loopy 1000000000 &
659 [1] 11435 577 [1] 11435
660 kill -USR2 11435 578 kill -USR2 11435
661 Recording AUX area tracing snapshot 579 Recording AUX area tracing snapshot
662 580
663 Note that the signal sent is SIGUSR2. 581 Note that the signal sent is SIGUSR2.
664 Note that "Recording AUX area tracing snapshot 582 Note that "Recording AUX area tracing snapshot" is displayed because the -v
665 option is used. 583 option is used.
666 584
667 The advantage of using "snapshot" control comm !! 585 The 2 modes cannot be used together.
668 controlled by access to a FIFO e.g. <<
669 <<
670 $ mkfifo perf.control <<
671 $ mkfifo perf.ack <<
672 $ cat perf.ack & <<
673 [1] 15235 <<
674 $ sudo ~/bin/perf record --control fif <<
675 [2] 15243 <<
676 $ ps -e | grep perf <<
677 15244 pts/1 00:00:00 perf <<
678 $ kill -USR2 15244 <<
679 bash: kill: (15244) - Operation not pe <<
680 $ echo snapshot > perf.control <<
681 ack <<
682 <<
683 The 3 Intel PT modes of operation cannot be us <<
684 586
685 587
686 Buffer handling 588 Buffer handling
687 ~~~~~~~~~~~~~~~ 589 ~~~~~~~~~~~~~~~
688 590
689 There may be buffer limitations (i.e. single T 591 There may be buffer limitations (i.e. single ToPa entry) which means that actual
690 buffer sizes are limited to powers of 2 up to !! 592 buffer sizes are limited to powers of 2 up to 4MiB (MAX_ORDER). In order to
691 provide other sizes, and in particular an arbi 593 provide other sizes, and in particular an arbitrarily large size, multiple
692 buffers are logically concatenated. However a 594 buffers are logically concatenated. However an interrupt must be used to switch
693 between buffers. That has two potential probl 595 between buffers. That has two potential problems:
694 a) the interrupt may not be handled in 596 a) the interrupt may not be handled in time so that the current buffer
695 becomes full and some trace data is lo 597 becomes full and some trace data is lost.
696 b) the interrupts may slow the system 598 b) the interrupts may slow the system and affect the performance
697 results. 599 results.
698 600
699 If trace data is lost, the driver sets 'trunca 601 If trace data is lost, the driver sets 'truncated' in the PERF_RECORD_AUX event
700 which the tools report as an error. 602 which the tools report as an error.
701 603
702 In full-trace mode, the driver waits for data 604 In full-trace mode, the driver waits for data to be copied out before allowing
703 the (logical) buffer to wrap-around. If data 605 the (logical) buffer to wrap-around. If data is not copied out quickly enough,
704 again 'truncated' is set in the PERF_RECORD_AU 606 again 'truncated' is set in the PERF_RECORD_AUX event. If the driver has to
705 wait, the intel_pt event gets disabled. Becau 607 wait, the intel_pt event gets disabled. Because it is difficult to know when
706 that happens, perf tools always re-enable the 608 that happens, perf tools always re-enable the intel_pt event after copying out
707 data. 609 data.
708 610
709 611
710 Intel PT and build ids 612 Intel PT and build ids
711 ~~~~~~~~~~~~~~~~~~~~~~ 613 ~~~~~~~~~~~~~~~~~~~~~~
712 614
713 By default "perf record" post-processes the ev 615 By default "perf record" post-processes the event stream to find all build ids
714 for executables for all addresses sampled. De 616 for executables for all addresses sampled. Deliberately, Intel PT is not
715 decoded for that purpose (it would take too lo 617 decoded for that purpose (it would take too long). Instead the build ids for
716 all executables encountered (due to mmap, comm 618 all executables encountered (due to mmap, comm or task events) are included
717 in the perf.data file. 619 in the perf.data file.
718 620
719 To see buildids included in the perf.data file 621 To see buildids included in the perf.data file use the command:
720 622
721 perf buildid-list 623 perf buildid-list
722 624
723 If the perf.data file contains Intel PT data, 625 If the perf.data file contains Intel PT data, that is the same as:
724 626
725 perf buildid-list --with-hits 627 perf buildid-list --with-hits
726 628
727 629
728 Snapshot mode and event disabling 630 Snapshot mode and event disabling
729 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 631 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
730 632
731 In order to make a snapshot, the intel_pt even 633 In order to make a snapshot, the intel_pt event is disabled using an IOCTL,
732 namely PERF_EVENT_IOC_DISABLE. However doing 634 namely PERF_EVENT_IOC_DISABLE. However doing that can also disable the
733 collection of side-band information. In order 635 collection of side-band information. In order to prevent that, a dummy
734 software event has been introduced that permit 636 software event has been introduced that permits tracking events (like mmaps) to
735 continue to be recorded while intel_pt is disa 637 continue to be recorded while intel_pt is disabled. That is important to ensure
736 there is complete side-band information to all 638 there is complete side-band information to allow the decoding of subsequent
737 snapshots. 639 snapshots.
738 640
739 A test has been created for that. To find the 641 A test has been created for that. To find the test:
740 642
741 perf test list 643 perf test list
742 ... 644 ...
743 23: Test using a dummy software event 645 23: Test using a dummy software event to keep tracking
744 646
745 To run the test: 647 To run the test:
746 648
747 perf test 23 649 perf test 23
748 23: Test using a dummy software event 650 23: Test using a dummy software event to keep tracking : Ok
749 651
750 652
751 perf record modes (nothing new here) 653 perf record modes (nothing new here)
752 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 654 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
753 655
754 perf record essentially operates in one of thr 656 perf record essentially operates in one of three modes:
755 per thread 657 per thread
756 per cpu 658 per cpu
757 workload only 659 workload only
758 660
759 "per thread" mode is selected by -t or by --pe 661 "per thread" mode is selected by -t or by --per-thread (with -p or -u or just a
760 workload). 662 workload).
761 "per cpu" is selected by -C or -a. 663 "per cpu" is selected by -C or -a.
762 "workload only" mode is selected by not using 664 "workload only" mode is selected by not using the other options but providing a
763 command to run (i.e. the workload). 665 command to run (i.e. the workload).
764 666
765 In per-thread mode an exact list of threads is 667 In per-thread mode an exact list of threads is traced. There is no inheritance.
766 Each thread has its own event buffer. 668 Each thread has its own event buffer.
767 669
768 In per-cpu mode all processes (or processes fr 670 In per-cpu mode all processes (or processes from the selected cgroup i.e. -G
769 option, or processes selected with -p or -u) a 671 option, or processes selected with -p or -u) are traced. Each cpu has its own
770 buffer. Inheritance is allowed. 672 buffer. Inheritance is allowed.
771 673
772 In workload-only mode, the workload is traced 674 In workload-only mode, the workload is traced but with per-cpu buffers.
773 Inheritance is allowed. Note that you can now 675 Inheritance is allowed. Note that you can now trace a workload in per-thread
774 mode by using the --per-thread option. 676 mode by using the --per-thread option.
775 677
776 678
777 Privileged vs non-privileged users 679 Privileged vs non-privileged users
778 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 680 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
779 681
780 Unless /proc/sys/kernel/perf_event_paranoid is 682 Unless /proc/sys/kernel/perf_event_paranoid is set to -1, unprivileged users
781 have memory limits imposed upon them. That af 683 have memory limits imposed upon them. That affects what buffer sizes they can
782 have as outlined above. 684 have as outlined above.
783 685
784 The v4.2 kernel introduced support for a conte 686 The v4.2 kernel introduced support for a context switch metadata event,
785 PERF_RECORD_SWITCH, which allows unprivileged 687 PERF_RECORD_SWITCH, which allows unprivileged users to see when their processes
786 are scheduled out and in, just not by whom, wh 688 are scheduled out and in, just not by whom, which is left for the
787 PERF_RECORD_SWITCH_CPU_WIDE, that is only acce 689 PERF_RECORD_SWITCH_CPU_WIDE, that is only accessible in system wide context,
788 which in turn requires CAP_PERFMON or CAP_SYS_ 690 which in turn requires CAP_PERFMON or CAP_SYS_ADMIN.
789 691
790 Please see the 45ac1403f564 ("perf: Add PERF_R 692 Please see the 45ac1403f564 ("perf: Add PERF_RECORD_SWITCH to indicate context
791 switches") commit, that introduces these metad 693 switches") commit, that introduces these metadata events for further info.
792 694
793 When working with kernels < v4.2, the followin 695 When working with kernels < v4.2, the following considerations must be taken,
794 as the sched:sched_switch tracepoints will be 696 as the sched:sched_switch tracepoints will be used to receive such information:
795 697
796 Unless /proc/sys/kernel/perf_event_paranoid is 698 Unless /proc/sys/kernel/perf_event_paranoid is set to -1, unprivileged users are
797 not permitted to use tracepoints which means t 699 not permitted to use tracepoints which means there is insufficient side-band
798 information to decode Intel PT in per-cpu mode 700 information to decode Intel PT in per-cpu mode, and potentially workload-only
799 mode too if the workload creates new processes 701 mode too if the workload creates new processes.
800 702
801 Note also, that to use tracepoints, read-acces 703 Note also, that to use tracepoints, read-access to debugfs is required. So if
802 debugfs is not mounted or the user does not ha 704 debugfs is not mounted or the user does not have read-access, it will again not
803 be possible to decode Intel PT in per-cpu mode 705 be possible to decode Intel PT in per-cpu mode.
804 706
805 707
806 sched_switch tracepoint 708 sched_switch tracepoint
807 ~~~~~~~~~~~~~~~~~~~~~~~ 709 ~~~~~~~~~~~~~~~~~~~~~~~
808 710
809 The sched_switch tracepoint is used to provide 711 The sched_switch tracepoint is used to provide side-band data for Intel PT
810 decoding in kernels where the PERF_RECORD_SWIT 712 decoding in kernels where the PERF_RECORD_SWITCH metadata event isn't
811 available. 713 available.
812 714
813 The sched_switch events are automatically adde 715 The sched_switch events are automatically added. e.g. the second event shown
814 below: 716 below:
815 717
816 $ perf record -vv -e intel_pt//u uname 718 $ perf record -vv -e intel_pt//u uname
817 -------------------------------------- 719 ------------------------------------------------------------
818 perf_event_attr: 720 perf_event_attr:
819 type 6 721 type 6
820 size 112 722 size 112
821 config 0x400 723 config 0x400
822 { sample_period, sample_freq } 1 724 { sample_period, sample_freq } 1
823 sample_type IP|TI 725 sample_type IP|TID|TIME|CPU|IDENTIFIER
824 read_format ID 726 read_format ID
825 disabled 1 727 disabled 1
826 inherit 1 728 inherit 1
827 exclude_kernel 1 729 exclude_kernel 1
828 exclude_hv 1 730 exclude_hv 1
829 enable_on_exec 1 731 enable_on_exec 1
830 sample_id_all 1 732 sample_id_all 1
831 -------------------------------------- 733 ------------------------------------------------------------
832 sys_perf_event_open: pid 31104 cpu 0 734 sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
833 sys_perf_event_open: pid 31104 cpu 1 735 sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
834 sys_perf_event_open: pid 31104 cpu 2 736 sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
835 sys_perf_event_open: pid 31104 cpu 3 737 sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
836 -------------------------------------- 738 ------------------------------------------------------------
837 perf_event_attr: 739 perf_event_attr:
838 type 2 740 type 2
839 size 112 741 size 112
840 config 0x108 742 config 0x108
841 { sample_period, sample_freq } 1 743 { sample_period, sample_freq } 1
842 sample_type IP|TI 744 sample_type IP|TID|TIME|CPU|PERIOD|RAW|IDENTIFIER
843 read_format ID 745 read_format ID
844 inherit 1 746 inherit 1
845 sample_id_all 1 747 sample_id_all 1
846 exclude_guest 1 748 exclude_guest 1
847 -------------------------------------- 749 ------------------------------------------------------------
848 sys_perf_event_open: pid -1 cpu 0 gr 750 sys_perf_event_open: pid -1 cpu 0 group_fd -1 flags 0x8
849 sys_perf_event_open: pid -1 cpu 1 gr 751 sys_perf_event_open: pid -1 cpu 1 group_fd -1 flags 0x8
850 sys_perf_event_open: pid -1 cpu 2 gr 752 sys_perf_event_open: pid -1 cpu 2 group_fd -1 flags 0x8
851 sys_perf_event_open: pid -1 cpu 3 gr 753 sys_perf_event_open: pid -1 cpu 3 group_fd -1 flags 0x8
852 -------------------------------------- 754 ------------------------------------------------------------
853 perf_event_attr: 755 perf_event_attr:
854 type 1 756 type 1
855 size 112 757 size 112
856 config 0x9 758 config 0x9
857 { sample_period, sample_freq } 1 759 { sample_period, sample_freq } 1
858 sample_type IP|TI 760 sample_type IP|TID|TIME|IDENTIFIER
859 read_format ID 761 read_format ID
860 disabled 1 762 disabled 1
861 inherit 1 763 inherit 1
862 exclude_kernel 1 764 exclude_kernel 1
863 exclude_hv 1 765 exclude_hv 1
864 mmap 1 766 mmap 1
865 comm 1 767 comm 1
866 enable_on_exec 1 768 enable_on_exec 1
867 task 1 769 task 1
868 sample_id_all 1 770 sample_id_all 1
869 mmap2 1 771 mmap2 1
870 comm_exec 1 772 comm_exec 1
871 -------------------------------------- 773 ------------------------------------------------------------
872 sys_perf_event_open: pid 31104 cpu 0 774 sys_perf_event_open: pid 31104 cpu 0 group_fd -1 flags 0x8
873 sys_perf_event_open: pid 31104 cpu 1 775 sys_perf_event_open: pid 31104 cpu 1 group_fd -1 flags 0x8
874 sys_perf_event_open: pid 31104 cpu 2 776 sys_perf_event_open: pid 31104 cpu 2 group_fd -1 flags 0x8
875 sys_perf_event_open: pid 31104 cpu 3 777 sys_perf_event_open: pid 31104 cpu 3 group_fd -1 flags 0x8
876 mmap size 528384B 778 mmap size 528384B
877 AUX area mmap length 4194304 779 AUX area mmap length 4194304
878 perf event ring buffer mmapped per cpu 780 perf event ring buffer mmapped per cpu
879 Synthesizing auxtrace information 781 Synthesizing auxtrace information
880 Linux 782 Linux
881 [ perf record: Woken up 1 times to wri 783 [ perf record: Woken up 1 times to write data ]
882 [ perf record: Captured and wrote 0.04 784 [ perf record: Captured and wrote 0.042 MB perf.data ]
883 785
884 Note, the sched_switch event is only added if 786 Note, the sched_switch event is only added if the user is permitted to use it
885 and only in per-cpu mode. 787 and only in per-cpu mode.
886 788
887 Note also, the sched_switch event is only adde 789 Note also, the sched_switch event is only added if TSC packets are requested.
888 That is because, in the absence of timing info 790 That is because, in the absence of timing information, the sched_switch events
889 cannot be matched against the Intel PT trace. 791 cannot be matched against the Intel PT trace.
890 792
891 793
892 perf script 794 perf script
893 ----------- 795 -----------
894 796
895 By default, perf script will decode trace data 797 By default, perf script will decode trace data found in the perf.data file.
896 This can be further controlled by new option - 798 This can be further controlled by new option --itrace.
897 799
898 800
899 New --itrace option 801 New --itrace option
900 ~~~~~~~~~~~~~~~~~~~ 802 ~~~~~~~~~~~~~~~~~~~
901 803
902 Having no option is the same as 804 Having no option is the same as
903 805
904 --itrace 806 --itrace
905 807
906 which, in turn, is the same as 808 which, in turn, is the same as
907 809
908 --itrace=cepwxy !! 810 --itrace=cepwx
909 811
910 The letters are: 812 The letters are:
911 813
912 i synthesize "instructions" even 814 i synthesize "instructions" events
913 y synthesize "cycles" events <<
914 b synthesize "branches" events 815 b synthesize "branches" events
915 x synthesize "transactions" even 816 x synthesize "transactions" events
916 w synthesize "ptwrite" events 817 w synthesize "ptwrite" events
917 p synthesize "power" events (inc !! 818 p synthesize "power" events
918 c synthesize branches events (ca 819 c synthesize branches events (calls only)
919 r synthesize branches events (re 820 r synthesize branches events (returns only)
920 o synthesize PEBS-via-PT events <<
921 I synthesize Event Trace events <<
922 e synthesize tracing error event 821 e synthesize tracing error events
923 d create a debug log 822 d create a debug log
924 g synthesize a call chain (use w 823 g synthesize a call chain (use with i or x)
925 G synthesize a call chain on exi 824 G synthesize a call chain on existing event records
926 l synthesize last branch entries 825 l synthesize last branch entries (use with i or x)
927 L synthesize last branch entries 826 L synthesize last branch entries on existing event records
928 s skip initial number of events 827 s skip initial number of events
929 q quicker (less detailed) decodi <<
930 A approximate IPC <<
931 Z prefer to ignore timestamps (s <<
932 828
933 "Instructions" events look like they were reco 829 "Instructions" events look like they were recorded by "perf record -e
934 instructions". 830 instructions".
935 831
936 "Cycles" events look like they were recorded b <<
937 (ie., the default). Note that even with CYC pa <<
938 these are not fully accurate, since CYC packet <<
939 instruction, only when some other event (like <<
940 TNT packet representing multiple branches) hap <<
941 be emitted. Thus, it is more effective for att <<
942 (and possibly basic blocks) than to individual <<
943 is not even perfect for functions (although it <<
944 option is active). <<
945 <<
946 "Branches" events look like they were recorded 832 "Branches" events look like they were recorded by "perf record -e branches". "c"
947 and "r" can be combined to get calls and retur 833 and "r" can be combined to get calls and returns.
948 834
949 "Transactions" events correspond to the start 835 "Transactions" events correspond to the start or end of transactions. The
950 'flags' field can be used in perf script to de 836 'flags' field can be used in perf script to determine whether the event is a
951 transaction start, commit or abort. !! 837 tranasaction start, commit or abort.
952 838
953 Note that "instructions", "cycles", "branches" !! 839 Note that "instructions", "branches" and "transactions" events depend on code
954 depend on code flow packets which can be disab !! 840 flow packets which can be disabled by using the config term "branch=0". Refer
955 "branch=0". Refer to the config terms section !! 841 to the config terms section above.
956 842
957 "ptwrite" events record the payload of the ptw 843 "ptwrite" events record the payload of the ptwrite instruction and whether
958 "fup_on_ptw" was used. "ptwrite" events depen 844 "fup_on_ptw" was used. "ptwrite" events depend on PTWRITE packets which are
959 recorded only if the "ptw" config term was use 845 recorded only if the "ptw" config term was used. Refer to the config terms
960 section above. perf script "synth" field disp 846 section above. perf script "synth" field displays "ptwrite" information like
961 this: "ip: 0 payload: 0x123456789abcdef0" whe 847 this: "ip: 0 payload: 0x123456789abcdef0" where "ip" is 1 if "fup_on_ptw" was
962 used. 848 used.
963 849
964 "Power" events correspond to power event packe 850 "Power" events correspond to power event packets and CBR (core-to-bus ratio)
965 packets. While CBR packets are always recorde 851 packets. While CBR packets are always recorded when tracing is enabled, power
966 event packets are recorded only if the "pwr_ev 852 event packets are recorded only if the "pwr_evt" config term was used. Refer to
967 the config terms section above. The power eve 853 the config terms section above. The power events record information about
968 C-state changes, whereas CBR is indicative of 854 C-state changes, whereas CBR is indicative of CPU frequency. perf script
969 "event,synth" fields display information like 855 "event,synth" fields display information like this:
970 <<
971 cbr: cbr: 22 freq: 2189 MHz (200%) 856 cbr: cbr: 22 freq: 2189 MHz (200%)
972 mwait: hints: 0x60 extensions: 0x1 857 mwait: hints: 0x60 extensions: 0x1
973 pwre: hw: 0 cstate: 2 sub-cstate: 0 858 pwre: hw: 0 cstate: 2 sub-cstate: 0
974 exstop: ip: 1 859 exstop: ip: 1
975 pwrx: deepest cstate: 2 last cstate: 860 pwrx: deepest cstate: 2 last cstate: 2 wake reason: 0x4
976 <<
977 Where: 861 Where:
978 <<
979 "cbr" includes the frequency and the p 862 "cbr" includes the frequency and the percentage of maximum non-turbo
980 "mwait" shows mwait hints and extensio 863 "mwait" shows mwait hints and extensions
981 "pwre" shows C-state transitions (to a 864 "pwre" shows C-state transitions (to a C-state deeper than C0) and
982 whether initiated by hardware 865 whether initiated by hardware
983 "exstop" indicates execution stopped a 866 "exstop" indicates execution stopped and whether the IP was recorded
984 exactly, 867 exactly,
985 "pwrx" indicates return to C0 868 "pwrx" indicates return to C0
986 <<
987 For more details refer to the Intel 64 and IA- 869 For more details refer to the Intel 64 and IA-32 Architectures Software
988 Developer Manuals. 870 Developer Manuals.
989 871
990 PSB events show when a PSB+ occurred and also <<
991 Emitting a PSB+ can cause a CPU a slight delay <<
992 of code with Intel PT, it is useful to know if <<
993 by Intel PT or not. <<
994 <<
995 Error events show where the decoder lost the t 872 Error events show where the decoder lost the trace. Error events
996 are quite important. Users must know if what 873 are quite important. Users must know if what they are seeing is a complete
997 picture or not. The "e" option may be followed !! 874 picture or not.
998 will or will not be reported. Each flag must <<
999 The flags supported by Intel PT are: <<
1000 <<
1001 -o Suppress overflow err <<
1002 -l Suppress trace data l <<
1003 <<
1004 For example, for errors but not overflow or d <<
1005 <<
1006 --itrace=e-o-l <<
1007 875
1008 The "d" option will cause the creation of a f 876 The "d" option will cause the creation of a file "intel_pt.log" containing all
1009 decoded packets and instructions. Note that 877 decoded packets and instructions. Note that this option slows down the decoder
1010 and that the resulting file may be very large !! 878 and that the resulting file may be very large.
1011 by flags which affect what debug messages wil <<
1012 must be preceded by either '+' or '-'. The fl <<
1013 <<
1014 -a Suppress logging of p <<
1015 +a Log all perf events <<
1016 +e Output only on decodi <<
1017 +o Output to stdout inst <<
1018 <<
1019 By default, logged perf events are filtered b <<
1020 flag +a overrides that. The +e flag can be u <<
1021 default, the log size in that case is 16384 b <<
1022 linkperf:perf-config[1] e.g. perf config itra <<
1023 879
1024 In addition, the period of the "instructions" 880 In addition, the period of the "instructions" event can be specified. e.g.
1025 881
1026 --itrace=i10us 882 --itrace=i10us
1027 883
1028 sets the period to 10us i.e. one instruction 884 sets the period to 10us i.e. one instruction sample is synthesized for each 10
1029 microseconds of trace. Alternatives to "us" 885 microseconds of trace. Alternatives to "us" are "ms" (milliseconds),
1030 "ns" (nanoseconds), "t" (TSC ticks) or "i" (i 886 "ns" (nanoseconds), "t" (TSC ticks) or "i" (instructions).
1031 887
1032 "ms", "us" and "ns" are converted to TSC tick 888 "ms", "us" and "ns" are converted to TSC ticks.
1033 889
1034 The timing information included with Intel PT 890 The timing information included with Intel PT does not give the time of every
1035 instruction. Consequently, for the purpose o 891 instruction. Consequently, for the purpose of sampling, the decoder estimates
1036 the time since the last timing packet based o 892 the time since the last timing packet based on 1 tick per instruction. The time
1037 on the sample is *not* adjusted and reflects 893 on the sample is *not* adjusted and reflects the last known value of TSC.
1038 894
1039 For Intel PT, the default period is 100us. 895 For Intel PT, the default period is 100us.
1040 896
1041 Setting it to a zero period means "as often a 897 Setting it to a zero period means "as often as possible".
1042 898
1043 In the case of Intel PT that is the same as a 899 In the case of Intel PT that is the same as a period of 1 and a unit of
1044 'instructions' (i.e. --itrace=i1i). 900 'instructions' (i.e. --itrace=i1i).
1045 901
1046 Also the call chain size (default 16, max. 10 902 Also the call chain size (default 16, max. 1024) for instructions or
1047 transactions events can be specified. e.g. 903 transactions events can be specified. e.g.
1048 904
1049 --itrace=ig32 905 --itrace=ig32
1050 --itrace=xg32 906 --itrace=xg32
1051 907
1052 Also the number of last branch entries (defau 908 Also the number of last branch entries (default 64, max. 1024) for instructions or
1053 transactions events can be specified. e.g. 909 transactions events can be specified. e.g.
1054 910
1055 --itrace=il10 911 --itrace=il10
1056 --itrace=xl10 912 --itrace=xl10
1057 913
1058 Note that last branch entries are cleared for 914 Note that last branch entries are cleared for each sample, so there is no overlap
1059 from one sample to the next. 915 from one sample to the next.
1060 916
1061 The G and L options are designed in particula 917 The G and L options are designed in particular for sample mode, and work much
1062 like g and l but add call chain and branch st 918 like g and l but add call chain and branch stack to the other selected events
1063 instead of synthesized events. For example, t 919 instead of synthesized events. For example, to record branch-misses events for
1064 'ls' and then add a call chain derived from t 920 'ls' and then add a call chain derived from the Intel PT trace:
1065 921
1066 perf record --aux-sample -e '{intel_p 922 perf record --aux-sample -e '{intel_pt//u,branch-misses:u}' -- ls
1067 perf report --itrace=Ge 923 perf report --itrace=Ge
1068 924
1069 Although in fact G is a default for perf repo 925 Although in fact G is a default for perf report, so that is the same as just:
1070 926
1071 perf report 927 perf report
1072 928
1073 One caveat with the G and L options is that t 929 One caveat with the G and L options is that they work poorly with "Large PEBS".
1074 Large PEBS means PEBS records will be accumul 930 Large PEBS means PEBS records will be accumulated by hardware and the written
1075 into the event buffer in one go. That reduce 931 into the event buffer in one go. That reduces interrupts, but can give very
1076 late timestamps. Because the Intel PT trace 932 late timestamps. Because the Intel PT trace is synchronized by timestamps,
1077 the PEBS events do not match the trace. Curr 933 the PEBS events do not match the trace. Currently, Large PEBS is used only in
1078 certain circumstances: 934 certain circumstances:
1079 - hardware supports it 935 - hardware supports it
1080 - PEBS is used 936 - PEBS is used
1081 - event period is specified, instead 937 - event period is specified, instead of frequency
1082 - the sample type is limited to the f 938 - the sample type is limited to the following flags:
1083 PERF_SAMPLE_IP | PERF_SAMPLE_ 939 PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_ADDR |
1084 PERF_SAMPLE_ID | PERF_SAMPLE_ 940 PERF_SAMPLE_ID | PERF_SAMPLE_CPU | PERF_SAMPLE_STREAM_ID |
1085 PERF_SAMPLE_DATA_SRC | PERF_S 941 PERF_SAMPLE_DATA_SRC | PERF_SAMPLE_IDENTIFIER |
1086 PERF_SAMPLE_TRANSACTION | PER 942 PERF_SAMPLE_TRANSACTION | PERF_SAMPLE_PHYS_ADDR |
1087 PERF_SAMPLE_REGS_INTR | PERF_ 943 PERF_SAMPLE_REGS_INTR | PERF_SAMPLE_REGS_USER |
1088 PERF_SAMPLE_PERIOD (and somet 944 PERF_SAMPLE_PERIOD (and sometimes) | PERF_SAMPLE_TIME
1089 Because Intel PT sample mode uses a different 945 Because Intel PT sample mode uses a different sample type to the list above,
1090 Large PEBS is not used with Intel PT sample m 946 Large PEBS is not used with Intel PT sample mode. To avoid Large PEBS in other
1091 cases, avoid specifying the event period i.e. 947 cases, avoid specifying the event period i.e. avoid the 'perf record' -c option,
1092 --count option, or 'period' config term. 948 --count option, or 'period' config term.
1093 949
1094 To disable trace decoding entirely, use the o 950 To disable trace decoding entirely, use the option --no-itrace.
1095 951
1096 It is also possible to skip events generated 952 It is also possible to skip events generated (instructions, branches, transactions)
1097 at the beginning. This is useful to ignore in 953 at the beginning. This is useful to ignore initialization code.
1098 954
1099 --itrace=i0nss1000000 955 --itrace=i0nss1000000
1100 956
1101 skips the first million instructions. 957 skips the first million instructions.
1102 958
1103 The q option changes the way the trace is dec <<
1104 but much less detailed. Specifically, with t <<
1105 decode TNT packets, and does not walk object <<
1106 TIP packets. The q option can be used with t <<
1107 is not used. The q option decodes more quick <<
1108 control flow of interest is represented or in <<
1109 TIP.PGD packets (refer below). However the q <<
1110 ranges that could then be decoded fully using <<
1111 <<
1112 What will *not* be decoded with the (single) <<
1113 <<
1114 - direct calls and jmps <<
1115 - conditional branches <<
1116 - non-branch instructions <<
1117 <<
1118 What *will* be decoded with the (single) q op <<
1119 <<
1120 - asynchronous branches such as inter <<
1121 - indirect branches <<
1122 - function return target address *if* <<
1123 config terms section) was used <<
1124 - start of (control-flow) tracing <<
1125 - end of (control-flow) tracing, if i <<
1126 - power events, ptwrite, transaction <<
1127 - instruction pointer associated with <<
1128 <<
1129 Note the q option does not specify what event <<
1130 option must be used also to show power events <<
1131 <<
1132 Repeating the q option (double-q i.e. qq) res <<
1133 less detail. The decoder decodes only extend <<
1134 instruction pointer if there is a FUP packet <<
1135 PSBEND). Note PSB packets occur regularly in <<
1136 config term (refer config terms section). Th <<
1137 PSB+ occurs while control flow is being trace <<
1138 <<
1139 What will *not* be decoded with the qq option <<
1140 <<
1141 - everything except instruction point <<
1142 <<
1143 What *will* be decoded with the qq option: <<
1144 <<
1145 - instruction pointer associated with <<
1146 <<
1147 The Z option is equivalent to having recorded <<
1148 (i.e. config term tsc=0). It can be useful to <<
1149 decoding a trace of a virtual machine. <<
1150 <<
1151 <<
1152 dlfilter-show-cycles.so <<
1153 ~~~~~~~~~~~~~~~~~~~~~~~ <<
1154 <<
1155 Cycles can be displayed using dlfilter-show-c <<
1156 option can be useful to provide higher granul <<
1157 <<
1158 perf script --itrace=A --call-trace - <<
1159 <<
1160 To see a list of dlfilters: <<
1161 <<
1162 perf script -v --list-dlfilters <<
1163 <<
1164 See also linkperf:perf-dlfilters[1] <<
1165 <<
1166 <<
1167 dump option 959 dump option
1168 ~~~~~~~~~~~ 960 ~~~~~~~~~~~
1169 961
1170 perf script has an option (-D) to "dump" the 962 perf script has an option (-D) to "dump" the events i.e. display the binary
1171 data. 963 data.
1172 964
1173 When -D is used, Intel PT packets are display 965 When -D is used, Intel PT packets are displayed. The packet decoder does not
1174 pay attention to PSB packets, but just decode 966 pay attention to PSB packets, but just decodes the bytes - so the packets seen
1175 by the actual decoder may not be identical in 967 by the actual decoder may not be identical in places where the data is corrupt.
1176 One example of that would be when the buffer- 968 One example of that would be when the buffer-switching interrupt has been too
1177 slow, and the buffer has been filled complete 969 slow, and the buffer has been filled completely. In that case, the last packet
1178 in the buffer might be truncated and immediat 970 in the buffer might be truncated and immediately followed by a PSB as the trace
1179 continues in the next buffer. 971 continues in the next buffer.
1180 972
1181 To disable the display of Intel PT packets, c 973 To disable the display of Intel PT packets, combine the -D option with
1182 --no-itrace. 974 --no-itrace.
1183 975
1184 976
1185 perf report 977 perf report
1186 ----------- 978 -----------
1187 979
1188 By default, perf report will decode trace dat 980 By default, perf report will decode trace data found in the perf.data file.
1189 This can be further controlled by new option 981 This can be further controlled by new option --itrace exactly the same as
1190 perf script, with the exception that the defa 982 perf script, with the exception that the default is --itrace=igxe.
1191 983
1192 984
1193 perf inject 985 perf inject
1194 ----------- 986 -----------
1195 987
1196 perf inject also accepts the --itrace option 988 perf inject also accepts the --itrace option in which case tracing data is
1197 removed and replaced with the synthesized eve 989 removed and replaced with the synthesized events. e.g.
1198 990
1199 perf inject --itrace -i perf.data -o 991 perf inject --itrace -i perf.data -o perf.data.new
1200 992
1201 Below is an example of using Intel PT with au 993 Below is an example of using Intel PT with autofdo. It requires autofdo
1202 (https://github.com/google/autofdo) and gcc v 994 (https://github.com/google/autofdo) and gcc version 5. The bubble
1203 sort example is from the AutoFDO tutorial (ht 995 sort example is from the AutoFDO tutorial (https://gcc.gnu.org/wiki/AutoFDO/Tutorial)
1204 amended to take the number of elements as a p 996 amended to take the number of elements as a parameter.
1205 997
1206 $ gcc-5 -O3 sort.c -o sort_optimized 998 $ gcc-5 -O3 sort.c -o sort_optimized
1207 $ ./sort_optimized 30000 999 $ ./sort_optimized 30000
1208 Bubble sorting array of 30000 element 1000 Bubble sorting array of 30000 elements
1209 2254 ms 1001 2254 ms
1210 1002
1211 $ cat ~/.perfconfig 1003 $ cat ~/.perfconfig
1212 [intel-pt] 1004 [intel-pt]
1213 mispred-all = on 1005 mispred-all = on
1214 1006
1215 $ perf record -e intel_pt//u ./sort 3 1007 $ perf record -e intel_pt//u ./sort 3000
1216 Bubble sorting array of 3000 elements 1008 Bubble sorting array of 3000 elements
1217 58 ms 1009 58 ms
1218 [ perf record: Woken up 2 times to wr 1010 [ perf record: Woken up 2 times to write data ]
1219 [ perf record: Captured and wrote 3.9 1011 [ perf record: Captured and wrote 3.939 MB perf.data ]
1220 $ perf inject -i perf.data -o inj --i 1012 $ perf inject -i perf.data -o inj --itrace=i100usle --strip
1221 $ ./create_gcov --binary=./sort --pro 1013 $ ./create_gcov --binary=./sort --profile=inj --gcov=sort.gcov -gcov_version=1
1222 $ gcc-5 -O3 -fauto-profile=sort.gcov 1014 $ gcc-5 -O3 -fauto-profile=sort.gcov sort.c -o sort_autofdo
1223 $ ./sort_autofdo 30000 1015 $ ./sort_autofdo 30000
1224 Bubble sorting array of 30000 element 1016 Bubble sorting array of 30000 elements
1225 2155 ms 1017 2155 ms
1226 1018
1227 Note there is currently no advantage to using 1019 Note there is currently no advantage to using Intel PT instead of LBR, but
1228 that may change in the future if greater use 1020 that may change in the future if greater use is made of the data.
1229 1021
1230 1022
1231 PEBS via Intel PT 1023 PEBS via Intel PT
1232 ----------------- 1024 -----------------
1233 1025
1234 Some hardware has the feature to redirect PEB 1026 Some hardware has the feature to redirect PEBS records to the Intel PT trace.
1235 Recording is selected by using the aux-output 1027 Recording is selected by using the aux-output config term e.g.
1236 1028
1237 perf record -c 10000 -e '{intel_pt/br 1029 perf record -c 10000 -e '{intel_pt/branch=0/,cycles/aux-output/ppp}' uname
1238 1030
1239 Originally, software only supported redirecti !! 1031 Note that currently, software only supports redirecting at most one PEBS event.
1240 was not able to differentiate one event from <<
1241 kernels and perf tools add support for the PE <<
1242 To check for the presence of that event in a <<
1243 <<
1244 perf script -D --no-itrace | grep PER <<
1245 1032
1246 To display PEBS events from the Intel PT trac 1033 To display PEBS events from the Intel PT trace, use the itrace 'o' option e.g.
1247 1034
1248 perf script --itrace=oe 1035 perf script --itrace=oe
1249 <<
1250 XED <<
1251 --- <<
1252 <<
1253 include::build-xed.txt[] <<
1254 <<
1255 <<
1256 Tracing Virtual Machines (kernel only) <<
1257 -------------------------------------- <<
1258 <<
1259 Currently, kernel tracing is supported with e <<
1260 (i.e. no TSC timestamps) or VM Time Correlati <<
1261 using 'perf inject' and requires unchanging V <<
1262 <<
1263 Other limitations and caveats <<
1264 <<
1265 VMX controls may suppress packets needed for <<
1266 VMX controls may block the perf NMI to the h <<
1267 Guest kernel self-modifying code (e.g. jump <<
1268 Guest thread information is unknown <<
1269 Guest VCPU is unknown but may be able to be <<
1270 Callchains are not supported <<
1271 <<
1272 Example using "timeless" decoding <<
1273 <<
1274 Start VM <<
1275 <<
1276 $ sudo virsh start kubuntu20.04 <<
1277 Domain kubuntu20.04 started <<
1278 <<
1279 Mount the guest file system. Note sshfs need <<
1280 <<
1281 $ mkdir vm0 <<
1282 $ sshfs -o direct_io root@vm0:/ vm0 <<
1283 <<
1284 Copy the guest /proc/kallsyms, /proc/modules <<
1285 <<
1286 $ perf buildid-cache -v --kcore vm0/proc/kco <<
1287 kcore added to build-id cache directory /hom <<
1288 $ KALLSYMS=/home/user/.debug/[kernel.kcore]/ <<
1289 <<
1290 Find the VM process <<
1291 <<
1292 $ ps -eLl | grep 'KVM\|PID' <<
1293 F S UID PID PPID LWP C PRI NI <<
1294 3 S 64055 1430 1 1440 1 80 0 <<
1295 3 S 64055 1430 1 1441 1 80 0 <<
1296 3 S 64055 1430 1 1442 1 80 0 <<
1297 3 S 64055 1430 1 1443 2 80 0 <<
1298 <<
1299 Start an open-ended perf record, tracing the <<
1300 TSC is not supported and tsc=0 must be specif <<
1301 However, IPC can still be determined, hence c <<
1302 Only kernel decoding is supported, so 'k' mus <<
1303 Intel PT traces both the host and the guest s <<
1304 Without timestamps, --per-thread must be spec <<
1305 <<
1306 $ sudo perf kvm --guest --host --guestkallsy <<
1307 ^C <<
1308 [ perf record: Woken up 1 times to write dat <<
1309 [ perf record: Captured and wrote 5.829 MB ] <<
1310 <<
1311 perf script can be used to provide an instruc <<
1312 <<
1313 $ perf script --guestkallsyms $KALLSYMS --in <<
1314 CPU 0/KVM 1440 ffffffff82133cdd __vm <<
1315 CPU 0/KVM 1440 ffffffff82133ce1 __vm <<
1316 CPU 0/KVM 1440 ffffffff82133ce5 __vm <<
1317 CPU 0/KVM 1440 ffffffff82133ce9 __vm <<
1318 CPU 0/KVM 1440 ffffffff82133ced __vm <<
1319 CPU 0/KVM 1440 ffffffff82133cf1 __vm <<
1320 CPU 0/KVM 1440 ffffffff82133cf5 __vm <<
1321 CPU 0/KVM 1440 ffffffff82133cf9 __vm <<
1322 CPU 0/KVM 1440 ffffffff82133cfc __vm <<
1323 CPU 0/KVM 1440 ffffffff82133c40 vmx_ <<
1324 CPU 0/KVM 1440 ffffffff82133c42 vmx_ <<
1325 :1440 1440 ffffffffbb678b06 nati <<
1326 :1440 1440 ffffffffbb678b0b nati <<
1327 :1440 1440 ffffffffbb666646 lapi <<
1328 :1440 1440 ffffffffbb666648 lapi <<
1329 :1440 1440 ffffffffbb66664a lapi <<
1330 :1440 1440 ffffffffbb66664b lapi <<
1331 :1440 1440 ffffffffbb74607f cloc <<
1332 :1440 1440 ffffffffbb746081 cloc <<
1333 :1440 1440 ffffffffbb74603c cloc <<
1334 :1440 1440 ffffffffbb74603d cloc <<
1335 <<
1336 Example using VM Time Correlation <<
1337 <<
1338 Start VM <<
1339 <<
1340 $ sudo virsh start kubuntu20.04 <<
1341 Domain kubuntu20.04 started <<
1342 <<
1343 Mount the guest file system. Note sshfs need <<
1344 <<
1345 $ mkdir -p vm0 <<
1346 $ sshfs -o direct_io root@vm0:/ vm0 <<
1347 <<
1348 Copy the guest /proc/kallsyms, /proc/modules <<
1349 <<
1350 $ perf buildid-cache -v --kcore vm0/proc/kco <<
1351 same kcore found in /home/user/.debug/[kerne <<
1352 $ KALLSYMS=/home/user/.debug/\[kernel.kcore\ <<
1353 <<
1354 Find the VM process <<
1355 <<
1356 $ ps -eLl | grep 'KVM\|PID' <<
1357 F S UID PID PPID LWP C PRI NI <<
1358 3 S 64055 16998 1 17005 13 80 0 <<
1359 3 S 64055 16998 1 17006 4 80 0 <<
1360 3 S 64055 16998 1 17007 3 80 0 <<
1361 3 S 64055 16998 1 17008 4 80 0 <<
1362 <<
1363 Start an open-ended perf record, tracing the <<
1364 IPC can be determined, hence cyc=1 can be add <<
1365 Only kernel decoding is supported, so 'k' mus <<
1366 Intel PT traces both the host and the guest s <<
1367 <<
1368 $ sudo perf kvm --guest --host --guestkallsy <<
1369 ^C[ perf record: Woken up 1 times to write d <<
1370 [ perf record: Captured and wrote 9.041 MB p <<
1371 <<
1372 Now 'perf inject' can be used to determine th <<
1373 only 7-bytes, so the TSC Offset might differ <<
1374 have no effect i.e. the resulting timestamps <<
1375 <<
1376 $ perf inject -i perf.data.kvm --vm-time-cor <<
1377 ERROR: Unknown TSC Offset for VMCS 0x1bff6a <<
1378 VMCS: 0x1bff6a TSC Offset 0xffffe42722c64c4 <<
1379 ERROR: Unknown TSC Offset for VMCS 0x1cbc08 <<
1380 VMCS: 0x1cbc08 TSC Offset 0xffffe42722c64c4 <<
1381 ERROR: Unknown TSC Offset for VMCS 0x1c3ce8 <<
1382 VMCS: 0x1c3ce8 TSC Offset 0xffffe42722c64c4 <<
1383 ERROR: Unknown TSC Offset for VMCS 0x1cbce9 <<
1384 VMCS: 0x1cbce9 TSC Offset 0xffffe42722c64c4 <<
1385 <<
1386 Each virtual CPU has a different Virtual Mach <<
1387 shown above with the calculated TSC Offset. F <<
1388 they should all be the same for the same virt <<
1389 <<
1390 Now that the TSC Offset is known, it can be p <<
1391 <<
1392 $ perf inject -i perf.data.kvm --vm-time-cor <<
1393 <<
1394 Note the options for 'perf inject' --vm-time- <<
1395 <<
1396 [ dry-run ] [ <TSC Offset> [ : <VMCS> [ , <V <<
1397 <<
1398 So it is possible to specify different TSC Of <<
1399 The option "dry-run" will cause the file to b <<
1400 Note it is also possible to get a intel_pt.lo <<
1401 <<
1402 There were no errors so, do it for real <<
1403 <<
1404 $ perf inject -i perf.data.kvm --vm-time-cor <<
1405 <<
1406 'perf script' can be used to see if there are <<
1407 <<
1408 $ perf script -i perf.data.kvm --guestkallsy <<
1409 <<
1410 There were none. <<
1411 <<
1412 'perf script' can be used to provide an instr <<
1413 <<
1414 $ perf script -i perf.data.kvm --guestkallsy <<
1415 CPU 1/KVM 17006 [001] 11500.262865593: <<
1416 CPU 1/KVM 17006 [001] 11500.262865593: <<
1417 CPU 1/KVM 17006 [001] 11500.262865593: <<
1418 CPU 1/KVM 17006 [001] 11500.262865593: <<
1419 CPU 1/KVM 17006 [001] 11500.262865593: <<
1420 CPU 1/KVM 17006 [001] 11500.262865593: <<
1421 CPU 1/KVM 17006 [001] 11500.262865593: <<
1422 CPU 1/KVM 17006 [001] 11500.262865593: <<
1423 CPU 1/KVM 17006 [001] 11500.262865593: <<
1424 CPU 1/KVM 17006 [001] 11500.262865593: <<
1425 CPU 1/KVM 17006 [001] 11500.262866075: <<
1426 :17006 17006 [001] 11500.262869216: <<
1427 :17006 17006 [001] 11500.262869216: <<
1428 :17006 17006 [001] 11500.262869216: <<
1429 :17006 17006 [001] 11500.262869216: <<
1430 :17006 17006 [001] 11500.262869216: <<
1431 :17006 17006 [001] 11500.262869216: <<
1432 :17006 17006 [001] 11500.262869216: <<
1433 :17006 17006 [001] 11500.262869216: <<
1434 :17006 17006 [001] 11500.262869216: <<
1435 :17006 17006 [001] 11500.262869216: <<
1436 <<
1437 <<
1438 Tracing Virtual Machines (including user spac <<
1439 --------------------------------------------- <<
1440 <<
1441 It is possible to use perf record to record s <<
1442 Sideband events from the guest perf.data file <<
1443 <<
1444 Here is an example of the steps needed: <<
1445 <<
1446 On the guest machine: <<
1447 <<
1448 Check that no-kvmclock kernel command line op <<
1449 <<
1450 Note, this is essential to enable time correl <<
1451 <<
1452 $ cat /proc/cmdline <<
1453 BOOT_IMAGE=/boot/vmlinuz-5.10.0-16-amd64 roo <<
1454 <<
1455 There is no BPF support at present so, if pos <<
1456 <<
1457 $ echo 0 | sudo tee /proc/sys/net/core/bpf_j <<
1458 0 <<
1459 <<
1460 Start perf record to collect sideband events: <<
1461 <<
1462 $ sudo perf record -o guest-sideband-testing <<
1463 <<
1464 On the host machine: <<
1465 <<
1466 Start perf record to collect Intel PT trace: <<
1467 <<
1468 Note, the host trace will get very big, very <<
1469 <<
1470 $ sudo perf record -o guest-sideband-testing <<
1471 <<
1472 On the guest machine: <<
1473 <<
1474 Run a small test case, just 'uname' in this e <<
1475 <<
1476 $ uname <<
1477 Linux <<
1478 <<
1479 On the host machine: <<
1480 <<
1481 Stop the Intel PT trace: <<
1482 <<
1483 ^C <<
1484 [ perf record: Woken up 1 times to write dat <<
1485 [ perf record: Captured and wrote 76.122 MB <<
1486 <<
1487 On the guest machine: <<
1488 <<
1489 Stop the Intel PT trace: <<
1490 <<
1491 ^C <<
1492 [ perf record: Woken up 1 times to write dat <<
1493 [ perf record: Captured and wrote 1.247 MB g <<
1494 <<
1495 And then copy guest-sideband-testing-guest-pe <<
1496 <<
1497 On the host machine: <<
1498 <<
1499 With the 2 perf.data recordings, and with the <<
1500 <<
1501 Identify the TSC Offset: <<
1502 <<
1503 $ perf inject -i guest-sideband-testing-host <<
1504 VMCS: 0x103fc6 TSC Offset 0xfffffa6ae070cb2 <<
1505 VMCS: 0x103ff2 TSC Offset 0xfffffa6ae070cb2 <<
1506 VMCS: 0x10fdaa TSC Offset 0xfffffa6ae070cb2 <<
1507 VMCS: 0x24d57c TSC Offset 0xfffffa6ae070cb2 <<
1508 <<
1509 Correct Intel PT TSC timestamps for the guest <<
1510 <<
1511 $ perf inject -i guest-sideband-testing-host <<
1512 <<
1513 Identify the guest machine PID: <<
1514 <<
1515 $ perf script -i guest-sideband-testing-host <<
1516 CPU 0/KVM 0 [000] 0.000000: PE <<
1517 CPU 1/KVM 0 [000] 0.000000: PE <<
1518 CPU 2/KVM 0 [000] 0.000000: PE <<
1519 CPU 3/KVM 0 [000] 0.000000: PE <<
1520 <<
1521 Note, the QEMU option -name debug-threads=on <<
1522 can be used to determine which thread is runn <<
1523 <<
1524 Create a guestmount, assuming the guest machi <<
1525 <<
1526 $ mkdir -p ~/guestmount/13376 <<
1527 $ sshfs -o direct_io vm_to_test:/ ~/guestmou <<
1528 <<
1529 Inject the guest perf.data file into the host <<
1530 <<
1531 Note, due to the guestmount option, guest obj <<
1532 If needed, VDSO can be copied manually in a f <<
1533 <<
1534 $ perf inject -i guest-sideband-testing-host <<
1535 <<
1536 Show an excerpt from the result. In this cas <<
1537 <<
1538 Notes: <<
1539 <<
1540 - the CPU displayed, [002] in this ca <<
1541 - events happening in the virtual mac <<
1542 - only calls and errors are displayed <<
1543 - branches entering and exiting the v <<
1544 <<
1545 $ perf script -i inj --itrace=ce -F+machine_ <<
1546 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1547 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1548 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1549 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1550 VM:13376 VCPU:003 uname 3404/340 <<
1551 VM:13376 VCPU:003 uname 3404/340 <<
1552 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1553 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1554 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1555 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1556 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1557 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1558 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1559 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1560 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1561 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1562 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1563 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1564 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1565 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1566 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1567 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1568 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1569 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1570 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1571 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1572 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1573 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1574 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1575 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1576 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1577 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1578 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1579 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1580 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1581 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1582 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1583 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1584 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1585 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1586 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1587 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1588 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1589 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1590 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1591 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1592 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1593 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1594 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1595 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1596 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1597 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1598 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1599 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1600 CPU 3/KVM 13376/13384 [002] 7919.4088 <<
1601 VM:13376 VCPU:003 uname 3404/340 <<
1602 VM:13376 VCPU:003 uname 3404/340 <<
1603 VM:13376 VCPU:003 uname 3404/340 <<
1604 VM:13376 VCPU:003 uname 3404/340 <<
1605 VM:13376 VCPU:003 uname 3404/340 <<
1606 VM:13376 VCPU:003 uname 3404/340 <<
1607 VM:13376 VCPU:003 uname 3404/340 <<
1608 VM:13376 VCPU:003 uname 3404/340 <<
1609 <<
1610 <<
1611 Tracing Virtual Machines - Guest Code <<
1612 ------------------------------------- <<
1613 <<
1614 A common case for KVM test programs is that t <<
1615 hypervisor, creating, running and destroying <<
1616 providing the guest object code from its own <<
1617 the VM is not running an OS, but only the fun <<
1618 hypervisor test program, and conveniently, lo <<
1619 addresses. To support that, option "--guest-c <<
1620 and perf kvm report. <<
1621 <<
1622 Here is an example tracing a test program fro <<
1623 <<
1624 # perf record --kcore -e intel_pt/cyc/ -- to <<
1625 [ perf record: Woken up 1 times to write dat <<
1626 [ perf record: Captured and wrote 0.280 MB p <<
1627 # perf script --guest-code --itrace=bep --ns <<
1628 [SNIP] <<
1629 tsc_msrs_test 18436 [007] 10897.962087733: <<
1630 tsc_msrs_test 18436 [007] 10897.962087733: <<
1631 tsc_msrs_test 18436 [007] 10897.962087733: <<
1632 tsc_msrs_test 18436 [007] 10897.962087836: <<
1633 [guest/18436] 18436 [007] 10897.962087836: <<
1634 [guest/18436] 18436 [007] 10897.962087836: <<
1635 [guest/18436] 18436 [007] 10897.962088248: <<
1636 tsc_msrs_test 18436 [007] 10897.962088248: <<
1637 tsc_msrs_test 18436 [007] 10897.962088248: <<
1638 tsc_msrs_test 18436 [007] 10897.962088256: <<
1639 tsc_msrs_test 18436 [007] 10897.962088270: <<
1640 [SNIP] <<
1641 tsc_msrs_test 18436 [007] 10897.962089321: <<
1642 tsc_msrs_test 18436 [007] 10897.962089321: <<
1643 tsc_msrs_test 18436 [007] 10897.962089321: <<
1644 tsc_msrs_test 18436 [007] 10897.962089424: <<
1645 [guest/18436] 18436 [007] 10897.962089424: <<
1646 [guest/18436] 18436 [007] 10897.962089701: <<
1647 [guest/18436] 18436 [007] 10897.962089701: <<
1648 [guest/18436] 18436 [007] 10897.962089701: <<
1649 [guest/18436] 18436 [007] 10897.962089701: <<
1650 [guest/18436] 18436 [007] 10897.962089878: <<
1651 tsc_msrs_test 18436 [007] 10897.962089878: <<
1652 tsc_msrs_test 18436 [007] 10897.962089878: <<
1653 tsc_msrs_test 18436 [007] 10897.962089887: <<
1654 tsc_msrs_test 18436 [007] 10897.962089901: <<
1655 [SNIP] <<
1656 <<
1657 # perf kvm --guest-code --guest --host repor <<
1658 <<
1659 # To display the perf.data header info, plea <<
1660 # <<
1661 # <<
1662 # Total Lost Samples: 0 <<
1663 # <<
1664 # Samples: 12 of event 'instructions' <<
1665 # Event count (approx.): 2274583 <<
1666 # <<
1667 # Children Self Command Shared <<
1668 # ........ ........ ............. ....... <<
1669 # <<
1670 54.70% 0.00% tsc_msrs_test [kernel. <<
1671 | <<
1672 ---entry_SYSCALL_64_after_hwframe <<
1673 do_syscall_64 <<
1674 | <<
1675 |--29.44%--syscall_exit_to_use <<
1676 | exit_to_user_mode_p <<
1677 | task_work_run <<
1678 | __fput <<
1679 <<
1680 <<
1681 Event Trace <<
1682 ----------- <<
1683 <<
1684 Event Trace records information about asynchr <<
1685 faults, VM exits and entries. The informatio <<
1686 and also the Interrupt Flag is recorded on th <<
1687 contains a type field to identify one of the <<
1688 <<
1689 1 INTR interrupt, fa <<
1690 2 IRET interrupt ret <<
1691 3 SMI system manage <<
1692 4 RSM resume from s <<
1693 5 SIPI startup inter <<
1694 6 INIT INIT signal <<
1695 7 VMENTRY VM-Entry <<
1696 8 VMEXIT VM-Entry <<
1697 9 VMEXIT_INTR VM-Exit due t <<
1698 10 SHUTDOWN Shutdown <<
1699 <<
1700 For more details, refer to the Intel 64 and I <<
1701 Developer Manuals (version 076 or later). <<
1702 <<
1703 The capability to do Event Trace is indicated <<
1704 /sys/bus/event_source/devices/intel_pt/caps/e <<
1705 <<
1706 Event trace is selected for recording using t <<
1707 <<
1708 perf record -e intel_pt/event/u uname <<
1709 <<
1710 Event trace events are output using the --itr <<
1711 <<
1712 perf script --itrace=Ie <<
1713 <<
1714 perf script displays events containing CFE ty <<
1715 in the form: <<
1716 <<
1717 evt: hw int (t) cfe: <<
1718 <<
1719 The IP flag indicates if the event binds to a <<
1720 flow control packet generation is enabled, as <<
1721 set. <<
1722 <<
1723 perf script displays events containing change <<
1724 <<
1725 iflag: t IFLAG <<
1726 <<
1727 where "via branch" indicates a branch (interr <<
1728 "non branch" indicates an instruction such as <<
1729 <<
1730 In addition, the current state of the interru <<
1731 or absence of the "D" (interrupt disabled) pe <<
1732 flag is changed, then the "t" flag is also in <<
1733 <<
1734 no flag, interrupts enabled I <<
1735 t interrupts become disabled IF <<
1736 D interrupts are disabled IF=0 <<
1737 Dt interrupts become enabled IF <<
1738 <<
1739 The intel-pt-events.py script illustrates how <<
1740 using a Python script. <<
1741 <<
1742 <<
1743 TNT Disable <<
1744 ----------- <<
1745 <<
1746 TNT packets are disabled using the "notnt" co <<
1747 <<
1748 perf record -e intel_pt/notnt/u uname <<
1749 <<
1750 In that case the --itrace q option is forced <<
1751 to reconstruct the control flow is not possib <<
1752 <<
1753 <<
1754 Emulated PTWRITE <<
1755 ---------------- <<
1756 <<
1757 Later perf tools support a method to emulate <<
1758 can be useful if hardware does not support th <<
1759 <<
1760 Instead of using the ptwrite instruction, a f <<
1761 a trace that encodes the payload data into TN <<
1762 of the function: <<
1763 <<
1764 #include <stdint.h> <<
1765 <<
1766 void perf_emulate_ptwrite(uint64_t x) <<
1767 __attribute__((externally_visible, noipa, no <<
1768 <<
1769 #define PERF_EMULATE_PTWRITE_8_BITS \ <<
1770 "1: shl %rax\n" \ <<
1771 " jc 1f\n" \ <<
1772 "1: shl %rax\n" \ <<
1773 " jc 1f\n" \ <<
1774 "1: shl %rax\n" \ <<
1775 " jc 1f\n" \ <<
1776 "1: shl %rax\n" \ <<
1777 " jc 1f\n" \ <<
1778 "1: shl %rax\n" \ <<
1779 " jc 1f\n" \ <<
1780 "1: shl %rax\n" \ <<
1781 " jc 1f\n" \ <<
1782 "1: shl %rax\n" \ <<
1783 " jc 1f\n" \ <<
1784 "1: shl %rax\n" \ <<
1785 " jc 1f\n" <<
1786 <<
1787 /* Undefined instruction */ <<
1788 #define PERF_EMULATE_PTWRITE_UD2 ".by <<
1789 <<
1790 #define PERF_EMULATE_PTWRITE_MAGIC PE <<
1791 <<
1792 void perf_emulate_ptwrite(uint64_t x __attri <<
1793 { <<
1794 /* Assumes SysV ABI : x passed in r <<
1795 __asm__ volatile ( <<
1796 "jmp 1f\n" <<
1797 PERF_EMULATE_PTWRITE_MAGIC <<
1798 "1: mov %rdi, %rax\n" <<
1799 PERF_EMULATE_PTWRITE_8_BITS <<
1800 PERF_EMULATE_PTWRITE_8_BITS <<
1801 PERF_EMULATE_PTWRITE_8_BITS <<
1802 PERF_EMULATE_PTWRITE_8_BITS <<
1803 PERF_EMULATE_PTWRITE_8_BITS <<
1804 PERF_EMULATE_PTWRITE_8_BITS <<
1805 PERF_EMULATE_PTWRITE_8_BITS <<
1806 PERF_EMULATE_PTWRITE_8_BITS <<
1807 "1: ret\n" <<
1808 ); <<
1809 } <<
1810 <<
1811 For example, a test program with the function <<
1812 <<
1813 #include <stdio.h> <<
1814 #include <stdint.h> <<
1815 #include <stdlib.h> <<
1816 <<
1817 #include "perf_emulate_ptwrite.h" <<
1818 <<
1819 int main(int argc, char *argv[]) <<
1820 { <<
1821 uint64_t x = 0; <<
1822 <<
1823 if (argc > 1) <<
1824 x = strtoull(argv[1], NULL, <<
1825 perf_emulate_ptwrite(x); <<
1826 return 0; <<
1827 } <<
1828 <<
1829 Can be compiled and traced: <<
1830 <<
1831 $ gcc -Wall -Wextra -O3 -g -o eg_ptw eg_ptw. <<
1832 $ perf record -e intel_pt//u ./eg_ptw 0x1234 <<
1833 [ perf record: Woken up 1 times to write dat <<
1834 [ perf record: Captured and wrote 0.017 MB p <<
1835 $ perf script --itrace=ew <<
1836 eg_ptw 19875 [007] 8061.235912: <<
1837 $ <<
1838 <<
1839 <<
1840 Pipe mode <<
1841 --------- <<
1842 Pipe mode is a problem for Intel PT and possi <<
1843 It's not recommended to use a pipe as data ou <<
1844 of the following reason. <<
1845 <<
1846 Essentially the auxtrace buffers do not behav <<
1847 event buffers. That is because the head and <<
1848 software, but in the auxtrace case the data i <<
1849 So the head and tail do not get updated as da <<
1850 <<
1851 In the Intel PT case, the head and tail are u <<
1852 is disabled by software, for example: <<
1853 - full-trace, system wide : when buffer p <<
1854 - full-trace, not system-wide : when buff <<
1855 context s <<
1856 - snapshot mode : as above but also when <<
1857 - sample mode : as above but also when a <<
1858 <<
1859 That means finished-round ordering doesn't wo <<
1860 can turn up that has data that extends back i <<
1861 very beginning of tracing. <<
1862 <<
1863 For a perf.data file, that problem is solved <<
1864 and queuing up the auxtrace buffers in advanc <<
1865 <<
1866 For pipe mode, the order of events and timest <<
1867 be messed up. <<
1868 <<
1869 <<
1870 EXAMPLE <<
1871 ------- <<
1872 <<
1873 Examples can be found on perf wiki page "Perf <<
1874 <<
1875 https://perf.wiki.kernel.org/index.php/Perf_t <<
1876 1036
1877 1037
1878 SEE ALSO 1038 SEE ALSO
1879 -------- 1039 --------
1880 1040
1881 linkperf:perf-record[1], linkperf:perf-script 1041 linkperf:perf-record[1], linkperf:perf-script[1], linkperf:perf-report[1],
1882 linkperf:perf-inject[1] 1042 linkperf:perf-inject[1]
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.