histogram-design.rst (89203B)
1.. SPDX-License-Identifier: GPL-2.0 2 3====================== 4Histogram Design Notes 5====================== 6 7:Author: Tom Zanussi <zanussi@kernel.org> 8 9This document attempts to provide a description of how the ftrace 10histograms work and how the individual pieces map to the data 11structures used to implement them in trace_events_hist.c and 12tracing_map.c. 13 14Note: All the ftrace histogram command examples assume the working 15 directory is the ftrace /tracing directory. For example:: 16 17 # cd /sys/kernel/debug/tracing 18 19Also, the histogram output displayed for those commands will be 20generally be truncated - only enough to make the point is displayed. 21 22'hist_debug' trace event files 23============================== 24 25If the kernel is compiled with CONFIG_HIST_TRIGGERS_DEBUG set, an 26event file named 'hist_debug' will appear in each event's 27subdirectory. This file can be read at any time and will display some 28of the hist trigger internals described in this document. Specific 29examples and output will be described in test cases below. 30 31Basic histograms 32================ 33 34First, basic histograms. Below is pretty much the simplest thing you 35can do with histograms - create one with a single key on a single 36event and cat the output:: 37 38 # echo 'hist:keys=pid' >> events/sched/sched_waking/trigger 39 40 # cat events/sched/sched_waking/hist 41 42 { pid: 18249 } hitcount: 1 43 { pid: 13399 } hitcount: 1 44 { pid: 17973 } hitcount: 1 45 { pid: 12572 } hitcount: 1 46 ... 47 { pid: 10 } hitcount: 921 48 { pid: 18255 } hitcount: 1444 49 { pid: 25526 } hitcount: 2055 50 { pid: 5257 } hitcount: 2055 51 { pid: 27367 } hitcount: 2055 52 { pid: 1728 } hitcount: 2161 53 54 Totals: 55 Hits: 21305 56 Entries: 183 57 Dropped: 0 58 59What this does is create a histogram on the sched_waking event using 60pid as a key and with a single value, hitcount, which even if not 61explicitly specified, exists for every histogram regardless. 62 63The hitcount value is a per-bucket value that's automatically 64incremented on every hit for the given key, which in this case is the 65pid. 66 67So in this histogram, there's a separate bucket for each pid, and each 68bucket contains a value for that bucket, counting the number of times 69sched_waking was called for that pid. 70 71Each histogram is represented by a hist_data struct. 72 73To keep track of each key and value field in the histogram, hist_data 74keeps an array of these fields named fields[]. The fields[] array is 75an array containing struct hist_field representations of each 76histogram val and key in the histogram (variables are also included 77here, but are discussed later). So for the above histogram we have one 78key and one value; in this case the one value is the hitcount value, 79which all histograms have, regardless of whether they define that 80value or not, which the above histogram does not. 81 82Each struct hist_field contains a pointer to the ftrace_event_field 83from the event's trace_event_file along with various bits related to 84that such as the size, offset, type, and a hist_field_fn_t function, 85which is used to grab the field's data from the ftrace event buffer 86(in most cases - some hist_fields such as hitcount don't directly map 87to an event field in the trace buffer - in these cases the function 88implementation gets its value from somewhere else). The flags field 89indicates which type of field it is - key, value, variable, variable 90reference, etc., with value being the default. 91 92The other important hist_data data structure in addition to the 93fields[] array is the tracing_map instance created for the histogram, 94which is held in the .map member. The tracing_map implements the 95lock-free hash table used to implement histograms (see 96kernel/trace/tracing_map.h for much more discussion about the 97low-level data structures implementing the tracing_map). For the 98purposes of this discussion, the tracing_map contains a number of 99buckets, each bucket corresponding to a particular tracing_map_elt 100object hashed by a given histogram key. 101 102Below is a diagram the first part of which describes the hist_data and 103associated key and value fields for the histogram described above. As 104you can see, there are two fields in the fields array, one val field 105for the hitcount and one key field for the pid key. 106 107Below that is a diagram of a run-time snapshot of what the tracing_map 108might look like for a given run. It attempts to show the 109relationships between the hist_data fields and the tracing_map 110elements for a couple hypothetical keys and values.:: 111 112 +------------------+ 113 | hist_data | 114 +------------------+ +----------------+ 115 | .fields[] |---->| val = hitcount |----------------------------+ 116 +----------------+ +----------------+ | 117 | .map | | .size | | 118 +----------------+ +--------------+ | 119 | .offset | | 120 +--------------+ | 121 | .fn() | | 122 +--------------+ | 123 . | 124 . | 125 . | 126 +----------------+ <--- n_vals | 127 | key = pid |----------------------------|--+ 128 +----------------+ | | 129 | .size | | | 130 +--------------+ | | 131 | .offset | | | 132 +--------------+ | | 133 | .fn() | | | 134 +----------------+ <--- n_fields | | 135 | unused | | | 136 +----------------+ | | 137 | | | | 138 +--------------+ | | 139 | | | | 140 +--------------+ | | 141 | | | | 142 +--------------+ | | 143 n_keys = n_fields - n_vals | | 144 145The hist_data n_vals and n_fields delineate the extent of the fields[] | | 146array and separate keys from values for the rest of the code. | | 147 148Below is a run-time representation of the tracing_map part of the | | 149histogram, with pointers from various parts of the fields[] array | | 150to corresponding parts of the tracing_map. | | 151 152The tracing_map consists of an array of tracing_map_entrys and a set | | 153of preallocated tracing_map_elts (abbreviated below as map_entry and | | 154map_elt). The total number of map_entrys in the hist_data.map array = | | 155map->max_elts (actually map->map_size but only max_elts of those are | | 156used. This is a property required by the map_insert() algorithm). | | 157 158If a map_entry is unused, meaning no key has yet hashed into it, its | | 159.key value is 0 and its .val pointer is NULL. Once a map_entry has | | 160been claimed, the .key value contains the key's hash value and the | | 161.val member points to a map_elt containing the full key and an entry | | 162for each key or value in the map_elt.fields[] array. There is an | | 163entry in the map_elt.fields[] array corresponding to each hist_field | | 164in the histogram, and this is where the continually aggregated sums | | 165corresponding to each histogram value are kept. | | 166 167The diagram attempts to show the relationship between the | | 168hist_data.fields[] and the map_elt.fields[] with the links drawn | | 169between diagrams:: 170 171 +-----------+ | | 172 | hist_data | | | 173 +-----------+ | | 174 | .fields | | | 175 +---------+ +-----------+ | | 176 | .map |---->| map_entry | | | 177 +---------+ +-----------+ | | 178 | .key |---> 0 | | 179 +---------+ | | 180 | .val |---> NULL | | 181 +-----------+ | | 182 | map_entry | | | 183 +-----------+ | | 184 | .key |---> pid = 999 | | 185 +---------+ +-----------+ | | 186 | .val |--->| map_elt | | | 187 +---------+ +-----------+ | | 188 . | .key |---> full key * | | 189 . +---------+ +---------------+ | | 190 . | .fields |--->| .sum (val) |<-+ | 191 +-----------+ +---------+ | 2345 | | | 192 | map_entry | +---------------+ | | 193 +-----------+ | .offset (key) |<----+ 194 | .key |---> 0 | 0 | | | 195 +---------+ +---------------+ | | 196 | .val |---> NULL . | | 197 +-----------+ . | | 198 | map_entry | . | | 199 +-----------+ +---------------+ | | 200 | .key | | .sum (val) or | | | 201 +---------+ +---------+ | .offset (key) | | | 202 | .val |--->| map_elt | +---------------+ | | 203 +-----------+ +---------+ | .sum (val) or | | | 204 | map_entry | | .offset (key) | | | 205 +-----------+ +---------------+ | | 206 | .key |---> pid = 4444 | | 207 +---------+ +-----------+ | | 208 | .val | | map_elt | | | 209 +---------+ +-----------+ | | 210 | .key |---> full key * | | 211 +---------+ +---------------+ | | 212 | .fields |--->| .sum (val) |<-+ | 213 +---------+ | 65523 | | 214 +---------------+ | 215 | .offset (key) |<----+ 216 | 0 | 217 +---------------+ 218 . 219 . 220 . 221 +---------------+ 222 | .sum (val) or | 223 | .offset (key) | 224 +---------------+ 225 | .sum (val) or | 226 | .offset (key) | 227 +---------------+ 228 229Abbreviations used in the diagrams:: 230 231 hist_data = struct hist_trigger_data 232 hist_data.fields = struct hist_field 233 fn = hist_field_fn_t 234 map_entry = struct tracing_map_entry 235 map_elt = struct tracing_map_elt 236 map_elt.fields = struct tracing_map_field 237 238Whenever a new event occurs and it has a hist trigger associated with 239it, event_hist_trigger() is called. event_hist_trigger() first deals 240with the key: for each subkey in the key (in the above example, there 241is just one subkey corresponding to pid), the hist_field that 242represents that subkey is retrieved from hist_data.fields[] and the 243hist_field_fn_t fn() associated with that field, along with the 244field's size and offset, is used to grab that subkey's data from the 245current trace record. 246 247Once the complete key has been retrieved, it's used to look that key 248up in the tracing_map. If there's no tracing_map_elt associated with 249that key, an empty one is claimed and inserted in the map for the new 250key. In either case, the tracing_map_elt associated with that key is 251returned. 252 253Once a tracing_map_elt available, hist_trigger_elt_update() is called. 254As the name implies, this updates the element, which basically means 255updating the element's fields. There's a tracing_map_field associated 256with each key and value in the histogram, and each of these correspond 257to the key and value hist_fields created when the histogram was 258created. hist_trigger_elt_update() goes through each value hist_field 259and, as for the keys, uses the hist_field's fn() and size and offset 260to grab the field's value from the current trace record. Once it has 261that value, it simply adds that value to that field's 262continually-updated tracing_map_field.sum member. Some hist_field 263fn()s, such as for the hitcount, don't actually grab anything from the 264trace record (the hitcount fn() just increments the counter sum by 1), 265but the idea is the same. 266 267Once all the values have been updated, hist_trigger_elt_update() is 268done and returns. Note that there are also tracing_map_fields for 269each subkey in the key, but hist_trigger_elt_update() doesn't look at 270them or update anything - those exist only for sorting, which can 271happen later. 272 273Basic histogram test 274-------------------- 275 276This is a good example to try. It produces 3 value fields and 2 key 277fields in the output:: 278 279 # echo 'hist:keys=common_pid,call_site.sym:values=bytes_req,bytes_alloc,hitcount' >> events/kmem/kmalloc/trigger 280 281To see the debug data, cat the kmem/kmalloc's 'hist_debug' file. It 282will show the trigger info of the histogram it corresponds to, along 283with the address of the hist_data associated with the histogram, which 284will become useful in later examples. It then displays the number of 285total hist_fields associated with the histogram along with a count of 286how many of those correspond to keys and how many correspond to values. 287 288It then goes on to display details for each field, including the 289field's flags and the position of each field in the hist_data's 290fields[] array, which is useful information for verifying that things 291internally appear correct or not, and which again will become even 292more useful in further examples:: 293 294 # cat events/kmem/kmalloc/hist_debug 295 296 # event histogram 297 # 298 # trigger info: hist:keys=common_pid,call_site.sym:vals=hitcount,bytes_req,bytes_alloc:sort=hitcount:size=2048 [active] 299 # 300 301 hist_data: 000000005e48c9a5 302 303 n_vals: 3 304 n_keys: 2 305 n_fields: 5 306 307 val fields: 308 309 hist_data->fields[0]: 310 flags: 311 VAL: HIST_FIELD_FL_HITCOUNT 312 type: u64 313 size: 8 314 is_signed: 0 315 316 hist_data->fields[1]: 317 flags: 318 VAL: normal u64 value 319 ftrace_event_field name: bytes_req 320 type: size_t 321 size: 8 322 is_signed: 0 323 324 hist_data->fields[2]: 325 flags: 326 VAL: normal u64 value 327 ftrace_event_field name: bytes_alloc 328 type: size_t 329 size: 8 330 is_signed: 0 331 332 key fields: 333 334 hist_data->fields[3]: 335 flags: 336 HIST_FIELD_FL_KEY 337 ftrace_event_field name: common_pid 338 type: int 339 size: 8 340 is_signed: 1 341 342 hist_data->fields[4]: 343 flags: 344 HIST_FIELD_FL_KEY 345 ftrace_event_field name: call_site 346 type: unsigned long 347 size: 8 348 is_signed: 0 349 350The commands below can be used to clean things up for the next test:: 351 352 # echo '!hist:keys=common_pid,call_site.sym:values=bytes_req,bytes_alloc,hitcount' >> events/kmem/kmalloc/trigger 353 354Variables 355========= 356 357Variables allow data from one hist trigger to be saved by one hist 358trigger and retrieved by another hist trigger. For example, a trigger 359on the sched_waking event can capture a timestamp for a particular 360pid, and later a sched_switch event that switches to that pid event 361can grab the timestamp and use it to calculate a time delta between 362the two events:: 363 364 # echo 'hist:keys=pid:ts0=common_timestamp.usecs' >> 365 events/sched/sched_waking/trigger 366 367 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0' >> 368 events/sched/sched_switch/trigger 369 370In terms of the histogram data structures, variables are implemented 371as another type of hist_field and for a given hist trigger are added 372to the hist_data.fields[] array just after all the val fields. To 373distinguish them from the existing key and val fields, they're given a 374new flag type, HIST_FIELD_FL_VAR (abbreviated FL_VAR) and they also 375make use of a new .var.idx field member in struct hist_field, which 376maps them to an index in a new map_elt.vars[] array added to the 377map_elt specifically designed to store and retrieve variable values. 378The diagram below shows those new elements and adds a new variable 379entry, ts0, corresponding to the ts0 variable in the sched_waking 380trigger above. 381 382sched_waking histogram 383----------------------:: 384 385 +------------------+ 386 | hist_data |<-------------------------------------------------------+ 387 +------------------+ +-------------------+ | 388 | .fields[] |-->| val = hitcount | | 389 +----------------+ +-------------------+ | 390 | .map | | .size | | 391 +----------------+ +-----------------+ | 392 | .offset | | 393 +-----------------+ | 394 | .fn() | | 395 +-----------------+ | 396 | .flags | | 397 +-----------------+ | 398 | .var.idx | | 399 +-------------------+ | 400 | var = ts0 | | 401 +-------------------+ | 402 | .size | | 403 +-----------------+ | 404 | .offset | | 405 +-----------------+ | 406 | .fn() | | 407 +-----------------+ | 408 | .flags & FL_VAR | | 409 +-----------------+ | 410 | .var.idx |----------------------------+-+ | 411 +-----------------+ | | | 412 . | | | 413 . | | | 414 . | | | 415 +-------------------+ <--- n_vals | | | 416 | key = pid | | | | 417 +-------------------+ | | | 418 | .size | | | | 419 +-----------------+ | | | 420 | .offset | | | | 421 +-----------------+ | | | 422 | .fn() | | | | 423 +-----------------+ | | | 424 | .flags & FL_KEY | | | | 425 +-----------------+ | | | 426 | .var.idx | | | | 427 +-------------------+ <--- n_fields | | | 428 | unused | | | | 429 +-------------------+ | | | 430 | | | | | 431 +-----------------+ | | | 432 | | | | | 433 +-----------------+ | | | 434 | | | | | 435 +-----------------+ | | | 436 | | | | | 437 +-----------------+ | | | 438 | | | | | 439 +-----------------+ | | | 440 n_keys = n_fields - n_vals | | | 441 | | | 442 443This is very similar to the basic case. In the above diagram, we can | | | 444see a new .flags member has been added to the struct hist_field | | | 445struct, and a new entry added to hist_data.fields representing the ts0 | | | 446variable. For a normal val hist_field, .flags is just 0 (modulo | | | 447modifier flags), but if the value is defined as a variable, the .flags | | | 448contains a set FL_VAR bit. | | | 449 450As you can see, the ts0 entry's .var.idx member contains the index | | | 451into the tracing_map_elts' .vars[] array containing variable values. | | | 452This idx is used whenever the value of the variable is set or read. | | | 453The map_elt.vars idx assigned to the given variable is assigned and | | | 454saved in .var.idx by create_tracing_map_fields() after it calls | | | 455tracing_map_add_var(). | | | 456 457Below is a representation of the histogram at run-time, which | | | 458populates the map, along with correspondence to the above hist_data and | | | 459hist_field data structures. | | | 460 461The diagram attempts to show the relationship between the | | | 462hist_data.fields[] and the map_elt.fields[] and map_elt.vars[] with | | | 463the links drawn between diagrams. For each of the map_elts, you can | | | 464see that the .fields[] members point to the .sum or .offset of a key | | | 465or val and the .vars[] members point to the value of a variable. The | | | 466arrows between the two diagrams show the linkages between those | | | 467tracing_map members and the field definitions in the corresponding | | | 468hist_data fields[] members.:: 469 470 +-----------+ | | | 471 | hist_data | | | | 472 +-----------+ | | | 473 | .fields | | | | 474 +---------+ +-----------+ | | | 475 | .map |---->| map_entry | | | | 476 +---------+ +-----------+ | | | 477 | .key |---> 0 | | | 478 +---------+ | | | 479 | .val |---> NULL | | | 480 +-----------+ | | | 481 | map_entry | | | | 482 +-----------+ | | | 483 | .key |---> pid = 999 | | | 484 +---------+ +-----------+ | | | 485 | .val |--->| map_elt | | | | 486 +---------+ +-----------+ | | | 487 . | .key |---> full key * | | | 488 . +---------+ +---------------+ | | | 489 . | .fields |--->| .sum (val) | | | | 490 . +---------+ | 2345 | | | | 491 . +--| .vars | +---------------+ | | | 492 . | +---------+ | .offset (key) | | | | 493 . | | 0 | | | | 494 . | +---------------+ | | | 495 . | . | | | 496 . | . | | | 497 . | . | | | 498 . | +---------------+ | | | 499 . | | .sum (val) or | | | | 500 . | | .offset (key) | | | | 501 . | +---------------+ | | | 502 . | | .sum (val) or | | | | 503 . | | .offset (key) | | | | 504 . | +---------------+ | | | 505 . | | | | 506 . +---------------->+---------------+ | | | 507 . | ts0 |<--+ | | 508 . | 113345679876 | | | | 509 . +---------------+ | | | 510 . | unused | | | | 511 . | | | | | 512 . +---------------+ | | | 513 . . | | | 514 . . | | | 515 . . | | | 516 . +---------------+ | | | 517 . | unused | | | | 518 . | | | | | 519 . +---------------+ | | | 520 . | unused | | | | 521 . | | | | | 522 . +---------------+ | | | 523 . | | | 524 +-----------+ | | | 525 | map_entry | | | | 526 +-----------+ | | | 527 | .key |---> pid = 4444 | | | 528 +---------+ +-----------+ | | | 529 | .val |--->| map_elt | | | | 530 +---------+ +-----------+ | | | 531 . | .key |---> full key * | | | 532 . +---------+ +---------------+ | | | 533 . | .fields |--->| .sum (val) | | | | 534 +---------+ | 2345 | | | | 535 +--| .vars | +---------------+ | | | 536 | +---------+ | .offset (key) | | | | 537 | | 0 | | | | 538 | +---------------+ | | | 539 | . | | | 540 | . | | | 541 | . | | | 542 | +---------------+ | | | 543 | | .sum (val) or | | | | 544 | | .offset (key) | | | | 545 | +---------------+ | | | 546 | | .sum (val) or | | | | 547 | | .offset (key) | | | | 548 | +---------------+ | | | 549 | | | | 550 | +---------------+ | | | 551 +---------------->| ts0 |<--+ | | 552 | 213499240729 | | | 553 +---------------+ | | 554 | unused | | | 555 | | | | 556 +---------------+ | | 557 . | | 558 . | | 559 . | | 560 +---------------+ | | 561 | unused | | | 562 | | | | 563 +---------------+ | | 564 | unused | | | 565 | | | | 566 +---------------+ | | 567 568For each used map entry, there's a map_elt pointing to an array of | | 569.vars containing the current value of the variables associated with | | 570that histogram entry. So in the above, the timestamp associated with | | 571pid 999 is 113345679876, and the timestamp variable in the same | | 572.var.idx for pid 4444 is 213499240729. | | 573 574sched_switch histogram | | 575---------------------- | | 576 577The sched_switch histogram paired with the above sched_waking | | 578histogram is shown below. The most important aspect of the | | 579sched_switch histogram is that it references a variable on the | | 580sched_waking histogram above. | | 581 582The histogram diagram is very similar to the others so far displayed, | | 583but it adds variable references. You can see the normal hitcount and | | 584key fields along with a new wakeup_lat variable implemented in the | | 585same way as the sched_waking ts0 variable, but in addition there's an | | 586entry with the new FL_VAR_REF (short for HIST_FIELD_FL_VAR_REF) flag. | | 587 588Associated with the new var ref field are a couple of new hist_field | | 589members, var.hist_data and var_ref_idx. For a variable reference, the | | 590var.hist_data goes with the var.idx, which together uniquely identify | | 591a particular variable on a particular histogram. The var_ref_idx is | | 592just the index into the var_ref_vals[] array that caches the values of | | 593each variable whenever a hist trigger is updated. Those resulting | | 594values are then finally accessed by other code such as trace action | | 595code that uses the var_ref_idx values to assign param values. | | 596 597The diagram below describes the situation for the sched_switch | | 598histogram referred to before:: 599 600 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0' >> | | 601 events/sched/sched_switch/trigger | | 602 | | 603 +------------------+ | | 604 | hist_data | | | 605 +------------------+ +-----------------------+ | | 606 | .fields[] |-->| val = hitcount | | | 607 +----------------+ +-----------------------+ | | 608 | .map | | .size | | | 609 +----------------+ +---------------------+ | | 610 +--| .var_refs[] | | .offset | | | 611 | +----------------+ +---------------------+ | | 612 | | .fn() | | | 613 | var_ref_vals[] +---------------------+ | | 614 | +-------------+ | .flags | | | 615 | | $ts0 |<---+ +---------------------+ | | 616 | +-------------+ | | .var.idx | | | 617 | | | | +---------------------+ | | 618 | +-------------+ | | .var.hist_data | | | 619 | | | | +---------------------+ | | 620 | +-------------+ | | .var_ref_idx | | | 621 | | | | +-----------------------+ | | 622 | +-------------+ | | var = wakeup_lat | | | 623 | . | +-----------------------+ | | 624 | . | | .size | | | 625 | . | +---------------------+ | | 626 | +-------------+ | | .offset | | | 627 | | | | +---------------------+ | | 628 | +-------------+ | | .fn() | | | 629 | | | | +---------------------+ | | 630 | +-------------+ | | .flags & FL_VAR | | | 631 | | +---------------------+ | | 632 | | | .var.idx | | | 633 | | +---------------------+ | | 634 | | | .var.hist_data | | | 635 | | +---------------------+ | | 636 | | | .var_ref_idx | | | 637 | | +---------------------+ | | 638 | | . | | 639 | | . | | 640 | | . | | 641 | | +-----------------------+ <--- n_vals | | 642 | | | key = pid | | | 643 | | +-----------------------+ | | 644 | | | .size | | | 645 | | +---------------------+ | | 646 | | | .offset | | | 647 | | +---------------------+ | | 648 | | | .fn() | | | 649 | | +---------------------+ | | 650 | | | .flags | | | 651 | | +---------------------+ | | 652 | | | .var.idx | | | 653 | | +-----------------------+ <--- n_fields | | 654 | | | unused | | | 655 | | +-----------------------+ | | 656 | | | | | | 657 | | +---------------------+ | | 658 | | | | | | 659 | | +---------------------+ | | 660 | | | | | | 661 | | +---------------------+ | | 662 | | | | | | 663 | | +---------------------+ | | 664 | | | | | | 665 | | +---------------------+ | | 666 | | n_keys = n_fields - n_vals | | 667 | | | | 668 | | | | 669 | | +-----------------------+ | | 670 +---------------------->| var_ref = $ts0 | | | 671 | +-----------------------+ | | 672 | | .size | | | 673 | +---------------------+ | | 674 | | .offset | | | 675 | +---------------------+ | | 676 | | .fn() | | | 677 | +---------------------+ | | 678 | | .flags & FL_VAR_REF | | | 679 | +---------------------+ | | 680 | | .var.idx |--------------------------+ | 681 | +---------------------+ | 682 | | .var.hist_data |----------------------------+ 683 | +---------------------+ 684 +---| .var_ref_idx | 685 +---------------------+ 686 687Abbreviations used in the diagrams:: 688 689 hist_data = struct hist_trigger_data 690 hist_data.fields = struct hist_field 691 fn = hist_field_fn_t 692 FL_KEY = HIST_FIELD_FL_KEY 693 FL_VAR = HIST_FIELD_FL_VAR 694 FL_VAR_REF = HIST_FIELD_FL_VAR_REF 695 696When a hist trigger makes use of a variable, a new hist_field is 697created with flag HIST_FIELD_FL_VAR_REF. For a VAR_REF field, the 698var.idx and var.hist_data take the same values as the referenced 699variable, as well as the referenced variable's size, type, and 700is_signed values. The VAR_REF field's .name is set to the name of the 701variable it references. If a variable reference was created using the 702explicit system.event.$var_ref notation, the hist_field's system and 703event_name variables are also set. 704 705So, in order to handle an event for the sched_switch histogram, 706because we have a reference to a variable on another histogram, we 707need to resolve all variable references first. This is done via the 708resolve_var_refs() calls made from event_hist_trigger(). What this 709does is grabs the var_refs[] array from the hist_data representing the 710sched_switch histogram. For each one of those, the referenced 711variable's var.hist_data along with the current key is used to look up 712the corresponding tracing_map_elt in that histogram. Once found, the 713referenced variable's var.idx is used to look up the variable's value 714using tracing_map_read_var(elt, var.idx), which yields the value of 715the variable for that element, ts0 in the case above. Note that both 716the hist_fields representing both the variable and the variable 717reference have the same var.idx, so this is straightforward. 718 719Variable and variable reference test 720------------------------------------ 721 722This example creates a variable on the sched_waking event, ts0, and 723uses it in the sched_switch trigger. The sched_switch trigger also 724creates its own variable, wakeup_lat, but nothing yet uses it:: 725 726 # echo 'hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 727 728 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0' >> events/sched/sched_switch/trigger 729 730Looking at the sched_waking 'hist_debug' output, in addition to the 731normal key and value hist_fields, in the val fields section we see a 732field with the HIST_FIELD_FL_VAR flag, which indicates that that field 733represents a variable. Note that in addition to the variable name, 734contained in the var.name field, it includes the var.idx, which is the 735index into the tracing_map_elt.vars[] array of the actual variable 736location. Note also that the output shows that variables live in the 737same part of the hist_data->fields[] array as normal values:: 738 739 # cat events/sched/sched_waking/hist_debug 740 741 # event histogram 742 # 743 # trigger info: hist:keys=pid:vals=hitcount:ts0=common_timestamp.usecs:sort=hitcount:size=2048:clock=global [active] 744 # 745 746 hist_data: 000000009536f554 747 748 n_vals: 2 749 n_keys: 1 750 n_fields: 3 751 752 val fields: 753 754 hist_data->fields[0]: 755 flags: 756 VAL: HIST_FIELD_FL_HITCOUNT 757 type: u64 758 size: 8 759 is_signed: 0 760 761 hist_data->fields[1]: 762 flags: 763 HIST_FIELD_FL_VAR 764 var.name: ts0 765 var.idx (into tracing_map_elt.vars[]): 0 766 type: u64 767 size: 8 768 is_signed: 0 769 770 key fields: 771 772 hist_data->fields[2]: 773 flags: 774 HIST_FIELD_FL_KEY 775 ftrace_event_field name: pid 776 type: pid_t 777 size: 8 778 is_signed: 1 779 780Moving on to the sched_switch trigger hist_debug output, in addition 781to the unused wakeup_lat variable, we see a new section displaying 782variable references. Variable references are displayed in a separate 783section because in addition to being logically separate from 784variables and values, they actually live in a separate hist_data 785array, var_refs[]. 786 787In this example, the sched_switch trigger has a reference to a 788variable on the sched_waking trigger, $ts0. Looking at the details, 789we can see that the var.hist_data value of the referenced variable 790matches the previously displayed sched_waking trigger, and the var.idx 791value matches the previously displayed var.idx value for that 792variable. Also displayed is the var_ref_idx value for that variable 793reference, which is where the value for that variable is cached for 794use when the trigger is invoked:: 795 796 # cat events/sched/sched_switch/hist_debug 797 798 # event histogram 799 # 800 # trigger info: hist:keys=next_pid:vals=hitcount:wakeup_lat=common_timestamp.usecs-$ts0:sort=hitcount:size=2048:clock=global [active] 801 # 802 803 hist_data: 00000000f4ee8006 804 805 n_vals: 2 806 n_keys: 1 807 n_fields: 3 808 809 val fields: 810 811 hist_data->fields[0]: 812 flags: 813 VAL: HIST_FIELD_FL_HITCOUNT 814 type: u64 815 size: 8 816 is_signed: 0 817 818 hist_data->fields[1]: 819 flags: 820 HIST_FIELD_FL_VAR 821 var.name: wakeup_lat 822 var.idx (into tracing_map_elt.vars[]): 0 823 type: u64 824 size: 0 825 is_signed: 0 826 827 key fields: 828 829 hist_data->fields[2]: 830 flags: 831 HIST_FIELD_FL_KEY 832 ftrace_event_field name: next_pid 833 type: pid_t 834 size: 8 835 is_signed: 1 836 837 variable reference fields: 838 839 hist_data->var_refs[0]: 840 flags: 841 HIST_FIELD_FL_VAR_REF 842 name: ts0 843 var.idx (into tracing_map_elt.vars[]): 0 844 var.hist_data: 000000009536f554 845 var_ref_idx (into hist_data->var_refs[]): 0 846 type: u64 847 size: 8 848 is_signed: 0 849 850The commands below can be used to clean things up for the next test:: 851 852 # echo '!hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0' >> events/sched/sched_switch/trigger 853 854 # echo '!hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 855 856Actions and Handlers 857==================== 858 859Adding onto the previous example, we will now do something with that 860wakeup_lat variable, namely send it and another field as a synthetic 861event. 862 863The onmatch() action below basically says that whenever we have a 864sched_switch event, if we have a matching sched_waking event, in this 865case if we have a pid in the sched_waking histogram that matches the 866next_pid field on this sched_switch event, we retrieve the 867variables specified in the wakeup_latency() trace action, and use 868them to generate a new wakeup_latency event into the trace stream. 869 870Note that the way the trace handlers such as wakeup_latency() (which 871could equivalently be written trace(wakeup_latency,$wakeup_lat,next_pid) 872are implemented, the parameters specified to the trace handler must be 873variables. In this case, $wakeup_lat is obviously a variable, but 874next_pid isn't, since it's just naming a field in the sched_switch 875trace event. Since this is something that almost every trace() and 876save() action does, a special shortcut is implemented to allow field 877names to be used directly in those cases. How it works is that under 878the covers, a temporary variable is created for the named field, and 879this variable is what is actually passed to the trace handler. In the 880code and documentation, this type of variable is called a 'field 881variable'. 882 883Fields on other trace event's histograms can be used as well. In that 884case we have to generate a new histogram and an unfortunately named 885'synthetic_field' (the use of synthetic here has nothing to do with 886synthetic events) and use that special histogram field as a variable. 887 888The diagram below illustrates the new elements described above in the 889context of the sched_switch histogram using the onmatch() handler and 890the trace() action. 891 892First, we define the wakeup_latency synthetic event:: 893 894 # echo 'wakeup_latency u64 lat; pid_t pid' >> synthetic_events 895 896Next, the sched_waking hist trigger as before:: 897 898 # echo 'hist:keys=pid:ts0=common_timestamp.usecs' >> 899 events/sched/sched_waking/trigger 900 901Finally, we create a hist trigger on the sched_switch event that 902generates a wakeup_latency() trace event. In this case we pass 903next_pid into the wakeup_latency synthetic event invocation, which 904means it will be automatically converted into a field variable:: 905 906 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0: \ 907 onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,next_pid)' >> 908 /sys/kernel/debug/tracing/events/sched/sched_switch/trigger 909 910The diagram for the sched_switch event is similar to previous examples 911but shows the additional field_vars[] array for hist_data and shows 912the linkages between the field_vars and the variables and references 913created to implement the field variables. The details are discussed 914below:: 915 916 +------------------+ 917 | hist_data | 918 +------------------+ +-----------------------+ 919 | .fields[] |-->| val = hitcount | 920 +----------------+ +-----------------------+ 921 | .map | | .size | 922 +----------------+ +---------------------+ 923 +---| .field_vars[] | | .offset | 924 | +----------------+ +---------------------+ 925 |+--| .var_refs[] | | .offset | 926 || +----------------+ +---------------------+ 927 || | .fn() | 928 || var_ref_vals[] +---------------------+ 929 || +-------------+ | .flags | 930 || | $ts0 |<---+ +---------------------+ 931 || +-------------+ | | .var.idx | 932 || | $next_pid |<-+ | +---------------------+ 933 || +-------------+ | | | .var.hist_data | 934 ||+>| $wakeup_lat | | | +---------------------+ 935 ||| +-------------+ | | | .var_ref_idx | 936 ||| | | | | +-----------------------+ 937 ||| +-------------+ | | | var = wakeup_lat | 938 ||| . | | +-----------------------+ 939 ||| . | | | .size | 940 ||| . | | +---------------------+ 941 ||| +-------------+ | | | .offset | 942 ||| | | | | +---------------------+ 943 ||| +-------------+ | | | .fn() | 944 ||| | | | | +---------------------+ 945 ||| +-------------+ | | | .flags & FL_VAR | 946 ||| | | +---------------------+ 947 ||| | | | .var.idx | 948 ||| | | +---------------------+ 949 ||| | | | .var.hist_data | 950 ||| | | +---------------------+ 951 ||| | | | .var_ref_idx | 952 ||| | | +---------------------+ 953 ||| | | . 954 ||| | | . 955 ||| | | . 956 ||| | | . 957 ||| +--------------+ | | . 958 +-->| field_var | | | . 959 || +--------------+ | | . 960 || | var | | | . 961 || +------------+ | | . 962 || | val | | | . 963 || +--------------+ | | . 964 || | field_var | | | . 965 || +--------------+ | | . 966 || | var | | | . 967 || +------------+ | | . 968 || | val | | | . 969 || +------------+ | | . 970 || . | | . 971 || . | | . 972 || . | | +-----------------------+ <--- n_vals 973 || +--------------+ | | | key = pid | 974 || | field_var | | | +-----------------------+ 975 || +--------------+ | | | .size | 976 || | var |--+| +---------------------+ 977 || +------------+ ||| | .offset | 978 || | val |-+|| +---------------------+ 979 || +------------+ ||| | .fn() | 980 || ||| +---------------------+ 981 || ||| | .flags | 982 || ||| +---------------------+ 983 || ||| | .var.idx | 984 || ||| +---------------------+ <--- n_fields 985 || ||| 986 || ||| n_keys = n_fields - n_vals 987 || ||| +-----------------------+ 988 || |+->| var = next_pid | 989 || | | +-----------------------+ 990 || | | | .size | 991 || | | +---------------------+ 992 || | | | .offset | 993 || | | +---------------------+ 994 || | | | .flags & FL_VAR | 995 || | | +---------------------+ 996 || | | | .var.idx | 997 || | | +---------------------+ 998 || | | | .var.hist_data | 999 || | | +-----------------------+ 1000 || +-->| val for next_pid | 1001 || | | +-----------------------+ 1002 || | | | .size | 1003 || | | +---------------------+ 1004 || | | | .offset | 1005 || | | +---------------------+ 1006 || | | | .fn() | 1007 || | | +---------------------+ 1008 || | | | .flags | 1009 || | | +---------------------+ 1010 || | | | | 1011 || | | +---------------------+ 1012 || | | 1013 || | | 1014 || | | +-----------------------+ 1015 +|------------------|-|>| var_ref = $ts0 | 1016 | | | +-----------------------+ 1017 | | | | .size | 1018 | | | +---------------------+ 1019 | | | | .offset | 1020 | | | +---------------------+ 1021 | | | | .fn() | 1022 | | | +---------------------+ 1023 | | | | .flags & FL_VAR_REF | 1024 | | | +---------------------+ 1025 | | +---| .var_ref_idx | 1026 | | +-----------------------+ 1027 | | | var_ref = $next_pid | 1028 | | +-----------------------+ 1029 | | | .size | 1030 | | +---------------------+ 1031 | | | .offset | 1032 | | +---------------------+ 1033 | | | .fn() | 1034 | | +---------------------+ 1035 | | | .flags & FL_VAR_REF | 1036 | | +---------------------+ 1037 | +-----| .var_ref_idx | 1038 | +-----------------------+ 1039 | | var_ref = $wakeup_lat | 1040 | +-----------------------+ 1041 | | .size | 1042 | +---------------------+ 1043 | | .offset | 1044 | +---------------------+ 1045 | | .fn() | 1046 | +---------------------+ 1047 | | .flags & FL_VAR_REF | 1048 | +---------------------+ 1049 +------------------------| .var_ref_idx | 1050 +---------------------+ 1051 1052As you can see, for a field variable, two hist_fields are created: one 1053representing the variable, in this case next_pid, and one to actually 1054get the value of the field from the trace stream, like a normal val 1055field does. These are created separately from normal variable 1056creation and are saved in the hist_data->field_vars[] array. See 1057below for how these are used. In addition, a reference hist_field is 1058also created, which is needed to reference the field variables such as 1059$next_pid variable in the trace() action. 1060 1061Note that $wakeup_lat is also a variable reference, referencing the 1062value of the expression common_timestamp-$ts0, and so also needs to 1063have a hist field entry representing that reference created. 1064 1065When hist_trigger_elt_update() is called to get the normal key and 1066value fields, it also calls update_field_vars(), which goes through 1067each field_var created for the histogram, and available from 1068hist_data->field_vars and calls val->fn() to get the data from the 1069current trace record, and then uses the var's var.idx to set the 1070variable at the var.idx offset in the appropriate tracing_map_elt's 1071variable at elt->vars[var.idx]. 1072 1073Once all the variables have been updated, resolve_var_refs() can be 1074called from event_hist_trigger(), and not only can our $ts0 and 1075$next_pid references be resolved but the $wakeup_lat reference as 1076well. At this point, the trace() action can simply access the values 1077assembled in the var_ref_vals[] array and generate the trace event. 1078 1079The same process occurs for the field variables associated with the 1080save() action. 1081 1082Abbreviations used in the diagram:: 1083 1084 hist_data = struct hist_trigger_data 1085 hist_data.fields = struct hist_field 1086 field_var = struct field_var 1087 fn = hist_field_fn_t 1088 FL_KEY = HIST_FIELD_FL_KEY 1089 FL_VAR = HIST_FIELD_FL_VAR 1090 FL_VAR_REF = HIST_FIELD_FL_VAR_REF 1091 1092trace() action field variable test 1093---------------------------------- 1094 1095This example adds to the previous test example by finally making use 1096of the wakeup_lat variable, but in addition also creates a couple of 1097field variables that then are all passed to the wakeup_latency() trace 1098action via the onmatch() handler. 1099 1100First, we create the wakeup_latency synthetic event:: 1101 1102 # echo 'wakeup_latency u64 lat; pid_t pid; char comm[16]' >> synthetic_events 1103 1104Next, the sched_waking trigger from previous examples:: 1105 1106 # echo 'hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 1107 1108Finally, as in the previous test example, we calculate and assign the 1109wakeup latency using the $ts0 reference from the sched_waking trigger 1110to the wakeup_lat variable, and finally use it along with a couple 1111sched_switch event fields, next_pid and next_comm, to generate a 1112wakeup_latency trace event. The next_pid and next_comm event fields 1113are automatically converted into field variables for this purpose:: 1114 1115 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,next_pid,next_comm)' >> /sys/kernel/debug/tracing/events/sched/sched_switch/trigger 1116 1117The sched_waking hist_debug output shows the same data as in the 1118previous test example:: 1119 1120 # cat events/sched/sched_waking/hist_debug 1121 1122 # event histogram 1123 # 1124 # trigger info: hist:keys=pid:vals=hitcount:ts0=common_timestamp.usecs:sort=hitcount:size=2048:clock=global [active] 1125 # 1126 1127 hist_data: 00000000d60ff61f 1128 1129 n_vals: 2 1130 n_keys: 1 1131 n_fields: 3 1132 1133 val fields: 1134 1135 hist_data->fields[0]: 1136 flags: 1137 VAL: HIST_FIELD_FL_HITCOUNT 1138 type: u64 1139 size: 8 1140 is_signed: 0 1141 1142 hist_data->fields[1]: 1143 flags: 1144 HIST_FIELD_FL_VAR 1145 var.name: ts0 1146 var.idx (into tracing_map_elt.vars[]): 0 1147 type: u64 1148 size: 8 1149 is_signed: 0 1150 1151 key fields: 1152 1153 hist_data->fields[2]: 1154 flags: 1155 HIST_FIELD_FL_KEY 1156 ftrace_event_field name: pid 1157 type: pid_t 1158 size: 8 1159 is_signed: 1 1160 1161The sched_switch hist_debug output shows the same key and value fields 1162as in the previous test example - note that wakeup_lat is still in the 1163val fields section, but that the new field variables are not there - 1164although the field variables are variables, they're held separately in 1165the hist_data's field_vars[] array. Although the field variables and 1166the normal variables are located in separate places, you can see that 1167the actual variable locations for those variables in the 1168tracing_map_elt.vars[] do have increasing indices as expected: 1169wakeup_lat takes the var.idx = 0 slot, while the field variables for 1170next_pid and next_comm have values var.idx = 1, and var.idx = 2. Note 1171also that those are the same values displayed for the variable 1172references corresponding to those variables in the variable reference 1173fields section. Since there are two triggers and thus two hist_data 1174addresses, those addresses also need to be accounted for when doing 1175the matching - you can see that the first variable refers to the 0 1176var.idx on the previous hist trigger (see the hist_data address 1177associated with that trigger), while the second variable refers to the 11780 var.idx on the sched_switch hist trigger, as do all the remaining 1179variable references. 1180 1181Finally, the action tracking variables section just shows the system 1182and event name for the onmatch() handler:: 1183 1184 # cat events/sched/sched_switch/hist_debug 1185 1186 # event histogram 1187 # 1188 # trigger info: hist:keys=next_pid:vals=hitcount:wakeup_lat=common_timestamp.usecs-$ts0:sort=hitcount:size=2048:clock=global:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,next_pid,next_comm) [active] 1189 # 1190 1191 hist_data: 0000000008f551b7 1192 1193 n_vals: 2 1194 n_keys: 1 1195 n_fields: 3 1196 1197 val fields: 1198 1199 hist_data->fields[0]: 1200 flags: 1201 VAL: HIST_FIELD_FL_HITCOUNT 1202 type: u64 1203 size: 8 1204 is_signed: 0 1205 1206 hist_data->fields[1]: 1207 flags: 1208 HIST_FIELD_FL_VAR 1209 var.name: wakeup_lat 1210 var.idx (into tracing_map_elt.vars[]): 0 1211 type: u64 1212 size: 0 1213 is_signed: 0 1214 1215 key fields: 1216 1217 hist_data->fields[2]: 1218 flags: 1219 HIST_FIELD_FL_KEY 1220 ftrace_event_field name: next_pid 1221 type: pid_t 1222 size: 8 1223 is_signed: 1 1224 1225 variable reference fields: 1226 1227 hist_data->var_refs[0]: 1228 flags: 1229 HIST_FIELD_FL_VAR_REF 1230 name: ts0 1231 var.idx (into tracing_map_elt.vars[]): 0 1232 var.hist_data: 00000000d60ff61f 1233 var_ref_idx (into hist_data->var_refs[]): 0 1234 type: u64 1235 size: 8 1236 is_signed: 0 1237 1238 hist_data->var_refs[1]: 1239 flags: 1240 HIST_FIELD_FL_VAR_REF 1241 name: wakeup_lat 1242 var.idx (into tracing_map_elt.vars[]): 0 1243 var.hist_data: 0000000008f551b7 1244 var_ref_idx (into hist_data->var_refs[]): 1 1245 type: u64 1246 size: 0 1247 is_signed: 0 1248 1249 hist_data->var_refs[2]: 1250 flags: 1251 HIST_FIELD_FL_VAR_REF 1252 name: next_pid 1253 var.idx (into tracing_map_elt.vars[]): 1 1254 var.hist_data: 0000000008f551b7 1255 var_ref_idx (into hist_data->var_refs[]): 2 1256 type: pid_t 1257 size: 4 1258 is_signed: 0 1259 1260 hist_data->var_refs[3]: 1261 flags: 1262 HIST_FIELD_FL_VAR_REF 1263 name: next_comm 1264 var.idx (into tracing_map_elt.vars[]): 2 1265 var.hist_data: 0000000008f551b7 1266 var_ref_idx (into hist_data->var_refs[]): 3 1267 type: char[16] 1268 size: 256 1269 is_signed: 0 1270 1271 field variables: 1272 1273 hist_data->field_vars[0]: 1274 1275 field_vars[0].var: 1276 flags: 1277 HIST_FIELD_FL_VAR 1278 var.name: next_pid 1279 var.idx (into tracing_map_elt.vars[]): 1 1280 1281 field_vars[0].val: 1282 ftrace_event_field name: next_pid 1283 type: pid_t 1284 size: 4 1285 is_signed: 1 1286 1287 hist_data->field_vars[1]: 1288 1289 field_vars[1].var: 1290 flags: 1291 HIST_FIELD_FL_VAR 1292 var.name: next_comm 1293 var.idx (into tracing_map_elt.vars[]): 2 1294 1295 field_vars[1].val: 1296 ftrace_event_field name: next_comm 1297 type: char[16] 1298 size: 256 1299 is_signed: 0 1300 1301 action tracking variables (for onmax()/onchange()/onmatch()): 1302 1303 hist_data->actions[0].match_data.event_system: sched 1304 hist_data->actions[0].match_data.event: sched_waking 1305 1306The commands below can be used to clean things up for the next test:: 1307 1308 # echo '!hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,next_pid,next_comm)' >> /sys/kernel/debug/tracing/events/sched/sched_switch/trigger 1309 1310 # echo '!hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 1311 1312 # echo '!wakeup_latency u64 lat; pid_t pid; char comm[16]' >> synthetic_events 1313 1314action_data and the trace() action 1315---------------------------------- 1316 1317As mentioned above, when the trace() action generates a synthetic 1318event, all the parameters to the synthetic event either already are 1319variables or are converted into variables (via field variables), and 1320finally all those variable values are collected via references to them 1321into a var_ref_vals[] array. 1322 1323The values in the var_ref_vals[] array, however, don't necessarily 1324follow the same ordering as the synthetic event params. To address 1325that, struct action_data contains another array, var_ref_idx[] that 1326maps the trace action params to the var_ref_vals[] values. Below is a 1327diagram illustrating that for the wakeup_latency() synthetic event:: 1328 1329 +------------------+ wakeup_latency() 1330 | action_data | event params var_ref_vals[] 1331 +------------------+ +-----------------+ +-----------------+ 1332 | .var_ref_idx[] |--->| $wakeup_lat idx |---+ | | 1333 +----------------+ +-----------------+ | +-----------------+ 1334 | .synth_event | | $next_pid idx |---|-+ | $wakeup_lat val | 1335 +----------------+ +-----------------+ | | +-----------------+ 1336 . | +->| $next_pid val | 1337 . | +-----------------+ 1338 . | . 1339 +-----------------+ | . 1340 | | | . 1341 +-----------------+ | +-----------------+ 1342 +--->| $wakeup_lat val | 1343 +-----------------+ 1344 1345Basically, how this ends up getting used in the synthetic event probe 1346function, trace_event_raw_event_synth(), is as follows:: 1347 1348 for each field i in .synth_event 1349 val_idx = .var_ref_idx[i] 1350 val = var_ref_vals[val_idx] 1351 1352action_data and the onXXX() handlers 1353------------------------------------ 1354 1355The hist trigger onXXX() actions other than onmatch(), such as onmax() 1356and onchange(), also make use of and internally create hidden 1357variables. This information is contained in the 1358action_data.track_data struct, and is also visible in the hist_debug 1359output as will be described in the example below. 1360 1361Typically, the onmax() or onchange() handlers are used in conjunction 1362with the save() and snapshot() actions. For example:: 1363 1364 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0: \ 1365 onmax($wakeup_lat).save(next_comm,prev_pid,prev_prio,prev_comm)' >> 1366 /sys/kernel/debug/tracing/events/sched/sched_switch/trigger 1367 1368or:: 1369 1370 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0: \ 1371 onmax($wakeup_lat).snapshot()' >> 1372 /sys/kernel/debug/tracing/events/sched/sched_switch/trigger 1373 1374save() action field variable test 1375--------------------------------- 1376 1377For this example, instead of generating a synthetic event, the save() 1378action is used to save field values whenever an onmax() handler 1379detects that a new max latency has been hit. As in the previous 1380example, the values being saved are also field values, but in this 1381case, are kept in a separate hist_data array named save_vars[]. 1382 1383As in previous test examples, we set up the sched_waking trigger:: 1384 1385 # echo 'hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 1386 1387In this case, however, we set up the sched_switch trigger to save some 1388sched_switch field values whenever we hit a new maximum latency. For 1389both the onmax() handler and save() action, variables will be created, 1390which we can use the hist_debug files to examine:: 1391 1392 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0:onmax($wakeup_lat).save(next_comm,prev_pid,prev_prio,prev_comm)' >> events/sched/sched_switch/trigger 1393 1394The sched_waking hist_debug output shows the same data as in the 1395previous test examples:: 1396 1397 # cat events/sched/sched_waking/hist_debug 1398 1399 # 1400 # trigger info: hist:keys=pid:vals=hitcount:ts0=common_timestamp.usecs:sort=hitcount:size=2048:clock=global [active] 1401 # 1402 1403 hist_data: 00000000e6290f48 1404 1405 n_vals: 2 1406 n_keys: 1 1407 n_fields: 3 1408 1409 val fields: 1410 1411 hist_data->fields[0]: 1412 flags: 1413 VAL: HIST_FIELD_FL_HITCOUNT 1414 type: u64 1415 size: 8 1416 is_signed: 0 1417 1418 hist_data->fields[1]: 1419 flags: 1420 HIST_FIELD_FL_VAR 1421 var.name: ts0 1422 var.idx (into tracing_map_elt.vars[]): 0 1423 type: u64 1424 size: 8 1425 is_signed: 0 1426 1427 key fields: 1428 1429 hist_data->fields[2]: 1430 flags: 1431 HIST_FIELD_FL_KEY 1432 ftrace_event_field name: pid 1433 type: pid_t 1434 size: 8 1435 is_signed: 1 1436 1437The output of the sched_switch trigger shows the same val and key 1438values as before, but also shows a couple new sections. 1439 1440First, the action tracking variables section now shows the 1441actions[].track_data information describing the special tracking 1442variables and references used to track, in this case, the running 1443maximum value. The actions[].track_data.var_ref member contains the 1444reference to the variable being tracked, in this case the $wakeup_lat 1445variable. In order to perform the onmax() handler function, there 1446also needs to be a variable that tracks the current maximum by getting 1447updated whenever a new maximum is hit. In this case, we can see that 1448an auto-generated variable named ' __max' has been created and is 1449visible in the actions[].track_data.track_var variable. 1450 1451Finally, in the new 'save action variables' section, we can see that 1452the 4 params to the save() function have resulted in 4 field variables 1453being created for the purposes of saving the values of the named 1454fields when the max is hit. These variables are kept in a separate 1455save_vars[] array off of hist_data, so are displayed in a separate 1456section:: 1457 1458 # cat events/sched/sched_switch/hist_debug 1459 1460 # event histogram 1461 # 1462 # trigger info: hist:keys=next_pid:vals=hitcount:wakeup_lat=common_timestamp.usecs-$ts0:sort=hitcount:size=2048:clock=global:onmax($wakeup_lat).save(next_comm,prev_pid,prev_prio,prev_comm) [active] 1463 # 1464 1465 hist_data: 0000000057bcd28d 1466 1467 n_vals: 2 1468 n_keys: 1 1469 n_fields: 3 1470 1471 val fields: 1472 1473 hist_data->fields[0]: 1474 flags: 1475 VAL: HIST_FIELD_FL_HITCOUNT 1476 type: u64 1477 size: 8 1478 is_signed: 0 1479 1480 hist_data->fields[1]: 1481 flags: 1482 HIST_FIELD_FL_VAR 1483 var.name: wakeup_lat 1484 var.idx (into tracing_map_elt.vars[]): 0 1485 type: u64 1486 size: 0 1487 is_signed: 0 1488 1489 key fields: 1490 1491 hist_data->fields[2]: 1492 flags: 1493 HIST_FIELD_FL_KEY 1494 ftrace_event_field name: next_pid 1495 type: pid_t 1496 size: 8 1497 is_signed: 1 1498 1499 variable reference fields: 1500 1501 hist_data->var_refs[0]: 1502 flags: 1503 HIST_FIELD_FL_VAR_REF 1504 name: ts0 1505 var.idx (into tracing_map_elt.vars[]): 0 1506 var.hist_data: 00000000e6290f48 1507 var_ref_idx (into hist_data->var_refs[]): 0 1508 type: u64 1509 size: 8 1510 is_signed: 0 1511 1512 hist_data->var_refs[1]: 1513 flags: 1514 HIST_FIELD_FL_VAR_REF 1515 name: wakeup_lat 1516 var.idx (into tracing_map_elt.vars[]): 0 1517 var.hist_data: 0000000057bcd28d 1518 var_ref_idx (into hist_data->var_refs[]): 1 1519 type: u64 1520 size: 0 1521 is_signed: 0 1522 1523 action tracking variables (for onmax()/onchange()/onmatch()): 1524 1525 hist_data->actions[0].track_data.var_ref: 1526 flags: 1527 HIST_FIELD_FL_VAR_REF 1528 name: wakeup_lat 1529 var.idx (into tracing_map_elt.vars[]): 0 1530 var.hist_data: 0000000057bcd28d 1531 var_ref_idx (into hist_data->var_refs[]): 1 1532 type: u64 1533 size: 0 1534 is_signed: 0 1535 1536 hist_data->actions[0].track_data.track_var: 1537 flags: 1538 HIST_FIELD_FL_VAR 1539 var.name: __max 1540 var.idx (into tracing_map_elt.vars[]): 1 1541 type: u64 1542 size: 8 1543 is_signed: 0 1544 1545 save action variables (save() params): 1546 1547 hist_data->save_vars[0]: 1548 1549 save_vars[0].var: 1550 flags: 1551 HIST_FIELD_FL_VAR 1552 var.name: next_comm 1553 var.idx (into tracing_map_elt.vars[]): 2 1554 1555 save_vars[0].val: 1556 ftrace_event_field name: next_comm 1557 type: char[16] 1558 size: 256 1559 is_signed: 0 1560 1561 hist_data->save_vars[1]: 1562 1563 save_vars[1].var: 1564 flags: 1565 HIST_FIELD_FL_VAR 1566 var.name: prev_pid 1567 var.idx (into tracing_map_elt.vars[]): 3 1568 1569 save_vars[1].val: 1570 ftrace_event_field name: prev_pid 1571 type: pid_t 1572 size: 4 1573 is_signed: 1 1574 1575 hist_data->save_vars[2]: 1576 1577 save_vars[2].var: 1578 flags: 1579 HIST_FIELD_FL_VAR 1580 var.name: prev_prio 1581 var.idx (into tracing_map_elt.vars[]): 4 1582 1583 save_vars[2].val: 1584 ftrace_event_field name: prev_prio 1585 type: int 1586 size: 4 1587 is_signed: 1 1588 1589 hist_data->save_vars[3]: 1590 1591 save_vars[3].var: 1592 flags: 1593 HIST_FIELD_FL_VAR 1594 var.name: prev_comm 1595 var.idx (into tracing_map_elt.vars[]): 5 1596 1597 save_vars[3].val: 1598 ftrace_event_field name: prev_comm 1599 type: char[16] 1600 size: 256 1601 is_signed: 0 1602 1603The commands below can be used to clean things up for the next test:: 1604 1605 # echo '!hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0:onmax($wakeup_lat).save(next_comm,prev_pid,prev_prio,prev_comm)' >> events/sched/sched_switch/trigger 1606 1607 # echo '!hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 1608 1609A couple special cases 1610====================== 1611 1612While the above covers the basics of the histogram internals, there 1613are a couple of special cases that should be discussed, since they 1614tend to create even more confusion. Those are field variables on other 1615histograms, and aliases, both described below through example tests 1616using the hist_debug files. 1617 1618Test of field variables on other histograms 1619------------------------------------------- 1620 1621This example is similar to the previous examples, but in this case, 1622the sched_switch trigger references a hist trigger field on another 1623event, namely the sched_waking event. In order to accomplish this, a 1624field variable is created for the other event, but since an existing 1625histogram can't be used, as existing histograms are immutable, a new 1626histogram with a matching variable is created and used, and we'll see 1627that reflected in the hist_debug output shown below. 1628 1629First, we create the wakeup_latency synthetic event. Note the 1630addition of the prio field:: 1631 1632 # echo 'wakeup_latency u64 lat; pid_t pid; int prio' >> synthetic_events 1633 1634As in previous test examples, we set up the sched_waking trigger:: 1635 1636 # echo 'hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 1637 1638Here we set up a hist trigger on sched_switch to send a wakeup_latency 1639event using an onmatch handler naming the sched_waking event. Note 1640that the third param being passed to the wakeup_latency() is prio, 1641which is a field name that needs to have a field variable created for 1642it. There isn't however any prio field on the sched_switch event so 1643it would seem that it wouldn't be possible to create a field variable 1644for it. The matching sched_waking event does have a prio field, so it 1645should be possible to make use of it for this purpose. The problem 1646with that is that it's not currently possible to define a new variable 1647on an existing histogram, so it's not possible to add a new prio field 1648variable to the existing sched_waking histogram. It is however 1649possible to create an additional new 'matching' sched_waking histogram 1650for the same event, meaning that it uses the same key and filters, and 1651define the new prio field variable on that. 1652 1653Here's the sched_switch trigger:: 1654 1655 # echo 'hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,next_pid,prio)' >> events/sched/sched_switch/trigger 1656 1657And here's the output of the hist_debug information for the 1658sched_waking hist trigger. Note that there are two histograms 1659displayed in the output: the first is the normal sched_waking 1660histogram we've seen in the previous examples, and the second is the 1661special histogram we created to provide the prio field variable. 1662 1663Looking at the second histogram below, we see a variable with the name 1664synthetic_prio. This is the field variable created for the prio field 1665on that sched_waking histogram:: 1666 1667 # cat events/sched/sched_waking/hist_debug 1668 1669 # event histogram 1670 # 1671 # trigger info: hist:keys=pid:vals=hitcount:ts0=common_timestamp.usecs:sort=hitcount:size=2048:clock=global [active] 1672 # 1673 1674 hist_data: 00000000349570e4 1675 1676 n_vals: 2 1677 n_keys: 1 1678 n_fields: 3 1679 1680 val fields: 1681 1682 hist_data->fields[0]: 1683 flags: 1684 VAL: HIST_FIELD_FL_HITCOUNT 1685 type: u64 1686 size: 8 1687 is_signed: 0 1688 1689 hist_data->fields[1]: 1690 flags: 1691 HIST_FIELD_FL_VAR 1692 var.name: ts0 1693 var.idx (into tracing_map_elt.vars[]): 0 1694 type: u64 1695 size: 8 1696 is_signed: 0 1697 1698 key fields: 1699 1700 hist_data->fields[2]: 1701 flags: 1702 HIST_FIELD_FL_KEY 1703 ftrace_event_field name: pid 1704 type: pid_t 1705 size: 8 1706 is_signed: 1 1707 1708 1709 # event histogram 1710 # 1711 # trigger info: hist:keys=pid:vals=hitcount:synthetic_prio=prio:sort=hitcount:size=2048 [active] 1712 # 1713 1714 hist_data: 000000006920cf38 1715 1716 n_vals: 2 1717 n_keys: 1 1718 n_fields: 3 1719 1720 val fields: 1721 1722 hist_data->fields[0]: 1723 flags: 1724 VAL: HIST_FIELD_FL_HITCOUNT 1725 type: u64 1726 size: 8 1727 is_signed: 0 1728 1729 hist_data->fields[1]: 1730 flags: 1731 HIST_FIELD_FL_VAR 1732 ftrace_event_field name: prio 1733 var.name: synthetic_prio 1734 var.idx (into tracing_map_elt.vars[]): 0 1735 type: int 1736 size: 4 1737 is_signed: 1 1738 1739 key fields: 1740 1741 hist_data->fields[2]: 1742 flags: 1743 HIST_FIELD_FL_KEY 1744 ftrace_event_field name: pid 1745 type: pid_t 1746 size: 8 1747 is_signed: 1 1748 1749Looking at the sched_switch histogram below, we can see a reference to 1750the synthetic_prio variable on sched_waking, and looking at the 1751associated hist_data address we see that it is indeed associated with 1752the new histogram. Note also that the other references are to a 1753normal variable, wakeup_lat, and to a normal field variable, next_pid, 1754the details of which are in the field variables section:: 1755 1756 # cat events/sched/sched_switch/hist_debug 1757 1758 # event histogram 1759 # 1760 # trigger info: hist:keys=next_pid:vals=hitcount:wakeup_lat=common_timestamp.usecs-$ts0:sort=hitcount:size=2048:clock=global:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,next_pid,prio) [active] 1761 # 1762 1763 hist_data: 00000000a73b67df 1764 1765 n_vals: 2 1766 n_keys: 1 1767 n_fields: 3 1768 1769 val fields: 1770 1771 hist_data->fields[0]: 1772 flags: 1773 VAL: HIST_FIELD_FL_HITCOUNT 1774 type: u64 1775 size: 8 1776 is_signed: 0 1777 1778 hist_data->fields[1]: 1779 flags: 1780 HIST_FIELD_FL_VAR 1781 var.name: wakeup_lat 1782 var.idx (into tracing_map_elt.vars[]): 0 1783 type: u64 1784 size: 0 1785 is_signed: 0 1786 1787 key fields: 1788 1789 hist_data->fields[2]: 1790 flags: 1791 HIST_FIELD_FL_KEY 1792 ftrace_event_field name: next_pid 1793 type: pid_t 1794 size: 8 1795 is_signed: 1 1796 1797 variable reference fields: 1798 1799 hist_data->var_refs[0]: 1800 flags: 1801 HIST_FIELD_FL_VAR_REF 1802 name: ts0 1803 var.idx (into tracing_map_elt.vars[]): 0 1804 var.hist_data: 00000000349570e4 1805 var_ref_idx (into hist_data->var_refs[]): 0 1806 type: u64 1807 size: 8 1808 is_signed: 0 1809 1810 hist_data->var_refs[1]: 1811 flags: 1812 HIST_FIELD_FL_VAR_REF 1813 name: wakeup_lat 1814 var.idx (into tracing_map_elt.vars[]): 0 1815 var.hist_data: 00000000a73b67df 1816 var_ref_idx (into hist_data->var_refs[]): 1 1817 type: u64 1818 size: 0 1819 is_signed: 0 1820 1821 hist_data->var_refs[2]: 1822 flags: 1823 HIST_FIELD_FL_VAR_REF 1824 name: next_pid 1825 var.idx (into tracing_map_elt.vars[]): 1 1826 var.hist_data: 00000000a73b67df 1827 var_ref_idx (into hist_data->var_refs[]): 2 1828 type: pid_t 1829 size: 4 1830 is_signed: 0 1831 1832 hist_data->var_refs[3]: 1833 flags: 1834 HIST_FIELD_FL_VAR_REF 1835 name: synthetic_prio 1836 var.idx (into tracing_map_elt.vars[]): 0 1837 var.hist_data: 000000006920cf38 1838 var_ref_idx (into hist_data->var_refs[]): 3 1839 type: int 1840 size: 4 1841 is_signed: 1 1842 1843 field variables: 1844 1845 hist_data->field_vars[0]: 1846 1847 field_vars[0].var: 1848 flags: 1849 HIST_FIELD_FL_VAR 1850 var.name: next_pid 1851 var.idx (into tracing_map_elt.vars[]): 1 1852 1853 field_vars[0].val: 1854 ftrace_event_field name: next_pid 1855 type: pid_t 1856 size: 4 1857 is_signed: 1 1858 1859 action tracking variables (for onmax()/onchange()/onmatch()): 1860 1861 hist_data->actions[0].match_data.event_system: sched 1862 hist_data->actions[0].match_data.event: sched_waking 1863 1864The commands below can be used to clean things up for the next test:: 1865 1866 # echo '!hist:keys=next_pid:wakeup_lat=common_timestamp.usecs-$ts0:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,next_pid,prio)' >> events/sched/sched_switch/trigger 1867 1868 # echo '!hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 1869 1870 # echo '!wakeup_latency u64 lat; pid_t pid; int prio' >> synthetic_events 1871 1872Alias test 1873---------- 1874 1875This example is very similar to previous examples, but demonstrates 1876the alias flag. 1877 1878First, we create the wakeup_latency synthetic event:: 1879 1880 # echo 'wakeup_latency u64 lat; pid_t pid; char comm[16]' >> synthetic_events 1881 1882Next, we create a sched_waking trigger similar to previous examples, 1883but in this case we save the pid in the waking_pid variable:: 1884 1885 # echo 'hist:keys=pid:waking_pid=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 1886 1887For the sched_switch trigger, instead of using $waking_pid directly in 1888the wakeup_latency synthetic event invocation, we create an alias of 1889$waking_pid named $woken_pid, and use that in the synthetic event 1890invocation instead:: 1891 1892 # echo 'hist:keys=next_pid:woken_pid=$waking_pid:wakeup_lat=common_timestamp.usecs-$ts0:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,$woken_pid,next_comm)' >> events/sched/sched_switch/trigger 1893 1894Looking at the sched_waking hist_debug output, in addition to the 1895normal fields, we can see the waking_pid variable:: 1896 1897 # cat events/sched/sched_waking/hist_debug 1898 1899 # event histogram 1900 # 1901 # trigger info: hist:keys=pid:vals=hitcount:waking_pid=pid,ts0=common_timestamp.usecs:sort=hitcount:size=2048:clock=global [active] 1902 # 1903 1904 hist_data: 00000000a250528c 1905 1906 n_vals: 3 1907 n_keys: 1 1908 n_fields: 4 1909 1910 val fields: 1911 1912 hist_data->fields[0]: 1913 flags: 1914 VAL: HIST_FIELD_FL_HITCOUNT 1915 type: u64 1916 size: 8 1917 is_signed: 0 1918 1919 hist_data->fields[1]: 1920 flags: 1921 HIST_FIELD_FL_VAR 1922 ftrace_event_field name: pid 1923 var.name: waking_pid 1924 var.idx (into tracing_map_elt.vars[]): 0 1925 type: pid_t 1926 size: 4 1927 is_signed: 1 1928 1929 hist_data->fields[2]: 1930 flags: 1931 HIST_FIELD_FL_VAR 1932 var.name: ts0 1933 var.idx (into tracing_map_elt.vars[]): 1 1934 type: u64 1935 size: 8 1936 is_signed: 0 1937 1938 key fields: 1939 1940 hist_data->fields[3]: 1941 flags: 1942 HIST_FIELD_FL_KEY 1943 ftrace_event_field name: pid 1944 type: pid_t 1945 size: 8 1946 is_signed: 1 1947 1948The sched_switch hist_debug output shows that a variable named 1949woken_pid has been created but that it also has the 1950HIST_FIELD_FL_ALIAS flag set. It also has the HIST_FIELD_FL_VAR flag 1951set, which is why it appears in the val field section. 1952 1953Despite that implementation detail, an alias variable is actually more 1954like a variable reference; in fact it can be thought of as a reference 1955to a reference. The implementation copies the var_ref->fn() from the 1956variable reference being referenced, in this case, the waking_pid 1957fn(), which is hist_field_var_ref() and makes that the fn() of the 1958alias. The hist_field_var_ref() fn() requires the var_ref_idx of the 1959variable reference it's using, so waking_pid's var_ref_idx is also 1960copied to the alias. The end result is that when the value of alias 1961is retrieved, in the end it just does the same thing the original 1962reference would have done and retrieves the same value from the 1963var_ref_vals[] array. You can verify this in the output by noting 1964that the var_ref_idx of the alias, in this case woken_pid, is the same 1965as the var_ref_idx of the reference, waking_pid, in the variable 1966reference fields section. 1967 1968Additionally, once it gets that value, since it is also a variable, it 1969then saves that value into its var.idx. So the var.idx of the 1970woken_pid alias is 0, which it fills with the value from var_ref_idx 0 1971when its fn() is called to update itself. You'll also notice that 1972there's a woken_pid var_ref in the variable refs section. That is the 1973reference to the woken_pid alias variable, and you can see that it 1974retrieves the value from the same var.idx as the woken_pid alias, 0, 1975and then in turn saves that value in its own var_ref_idx slot, 3, and 1976the value at this position is finally what gets assigned to the 1977$woken_pid slot in the trace event invocation:: 1978 1979 # cat events/sched/sched_switch/hist_debug 1980 1981 # event histogram 1982 # 1983 # trigger info: hist:keys=next_pid:vals=hitcount:woken_pid=$waking_pid,wakeup_lat=common_timestamp.usecs-$ts0:sort=hitcount:size=2048:clock=global:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,$woken_pid,next_comm) [active] 1984 # 1985 1986 hist_data: 0000000055d65ed0 1987 1988 n_vals: 3 1989 n_keys: 1 1990 n_fields: 4 1991 1992 val fields: 1993 1994 hist_data->fields[0]: 1995 flags: 1996 VAL: HIST_FIELD_FL_HITCOUNT 1997 type: u64 1998 size: 8 1999 is_signed: 0 2000 2001 hist_data->fields[1]: 2002 flags: 2003 HIST_FIELD_FL_VAR 2004 HIST_FIELD_FL_ALIAS 2005 var.name: woken_pid 2006 var.idx (into tracing_map_elt.vars[]): 0 2007 var_ref_idx (into hist_data->var_refs[]): 0 2008 type: pid_t 2009 size: 4 2010 is_signed: 1 2011 2012 hist_data->fields[2]: 2013 flags: 2014 HIST_FIELD_FL_VAR 2015 var.name: wakeup_lat 2016 var.idx (into tracing_map_elt.vars[]): 1 2017 type: u64 2018 size: 0 2019 is_signed: 0 2020 2021 key fields: 2022 2023 hist_data->fields[3]: 2024 flags: 2025 HIST_FIELD_FL_KEY 2026 ftrace_event_field name: next_pid 2027 type: pid_t 2028 size: 8 2029 is_signed: 1 2030 2031 variable reference fields: 2032 2033 hist_data->var_refs[0]: 2034 flags: 2035 HIST_FIELD_FL_VAR_REF 2036 name: waking_pid 2037 var.idx (into tracing_map_elt.vars[]): 0 2038 var.hist_data: 00000000a250528c 2039 var_ref_idx (into hist_data->var_refs[]): 0 2040 type: pid_t 2041 size: 4 2042 is_signed: 1 2043 2044 hist_data->var_refs[1]: 2045 flags: 2046 HIST_FIELD_FL_VAR_REF 2047 name: ts0 2048 var.idx (into tracing_map_elt.vars[]): 1 2049 var.hist_data: 00000000a250528c 2050 var_ref_idx (into hist_data->var_refs[]): 1 2051 type: u64 2052 size: 8 2053 is_signed: 0 2054 2055 hist_data->var_refs[2]: 2056 flags: 2057 HIST_FIELD_FL_VAR_REF 2058 name: wakeup_lat 2059 var.idx (into tracing_map_elt.vars[]): 1 2060 var.hist_data: 0000000055d65ed0 2061 var_ref_idx (into hist_data->var_refs[]): 2 2062 type: u64 2063 size: 0 2064 is_signed: 0 2065 2066 hist_data->var_refs[3]: 2067 flags: 2068 HIST_FIELD_FL_VAR_REF 2069 name: woken_pid 2070 var.idx (into tracing_map_elt.vars[]): 0 2071 var.hist_data: 0000000055d65ed0 2072 var_ref_idx (into hist_data->var_refs[]): 3 2073 type: pid_t 2074 size: 4 2075 is_signed: 1 2076 2077 hist_data->var_refs[4]: 2078 flags: 2079 HIST_FIELD_FL_VAR_REF 2080 name: next_comm 2081 var.idx (into tracing_map_elt.vars[]): 2 2082 var.hist_data: 0000000055d65ed0 2083 var_ref_idx (into hist_data->var_refs[]): 4 2084 type: char[16] 2085 size: 256 2086 is_signed: 0 2087 2088 field variables: 2089 2090 hist_data->field_vars[0]: 2091 2092 field_vars[0].var: 2093 flags: 2094 HIST_FIELD_FL_VAR 2095 var.name: next_comm 2096 var.idx (into tracing_map_elt.vars[]): 2 2097 2098 field_vars[0].val: 2099 ftrace_event_field name: next_comm 2100 type: char[16] 2101 size: 256 2102 is_signed: 0 2103 2104 action tracking variables (for onmax()/onchange()/onmatch()): 2105 2106 hist_data->actions[0].match_data.event_system: sched 2107 hist_data->actions[0].match_data.event: sched_waking 2108 2109The commands below can be used to clean things up for the next test:: 2110 2111 # echo '!hist:keys=next_pid:woken_pid=$waking_pid:wakeup_lat=common_timestamp.usecs-$ts0:onmatch(sched.sched_waking).wakeup_latency($wakeup_lat,$woken_pid,next_comm)' >> events/sched/sched_switch/trigger 2112 2113 # echo '!hist:keys=pid:ts0=common_timestamp.usecs' >> events/sched/sched_waking/trigger 2114 2115 # echo '!wakeup_latency u64 lat; pid_t pid; char comm[16]' >> synthetic_events