1 #!/usr/bin/env python3
2 # SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause)
3 """Convert directories of JSON events to C code."""
4 import argparse
5 import csv
6 from functools import lru_cache
7 import json
8 import metric
9 import os
10 import sys
11 from typing import (Callable, Dict, Optional, Sequence, Set, Tuple)
12 import collections
13
14 # Global command line arguments.
15 _args = None
16 # List of regular event tables.
17 _event_tables = []
18 # List of event tables generated from "/sys" directories.
19 _sys_event_tables = []
20 # List of regular metric tables.
21 _metric_tables = []
22 # List of metric tables generated from "/sys" directories.
23 _sys_metric_tables = []
24 # Mapping between sys event table names and sys metric table names.
25 _sys_event_table_to_metric_table_mapping = {}
26 # Map from an event name to an architecture standard
27 # JsonEvent. Architecture standard events are in json files in the top
28 # f'{_args.starting_dir}/{_args.arch}' directory.
29 _arch_std_events = {}
30 # Events to write out when the table is closed
31 _pending_events = []
32 # Name of events table to be written out
33 _pending_events_tblname = None
34 # Metrics to write out when the table is closed
35 _pending_metrics = []
36 # Name of metrics table to be written out
37 _pending_metrics_tblname = None
38 # Global BigCString shared by all structures.
39 _bcs = None
40 # Map from the name of a metric group to a description of the group.
41 _metricgroups = {}
42 # Order specific JsonEvent attributes will be visited.
43 _json_event_attributes = [
44 # cmp_sevent related attributes.
45 'name', 'topic', 'desc',
46 # Seems useful, put it early.
47 'event',
48 # Short things in alphabetical order.
49 'compat', 'deprecated', 'perpkg', 'unit',
50 # Longer things (the last won't be iterated over during decompress).
51 'long_desc'
52 ]
53
54 # Attributes that are in pmu_metric rather than pmu_event.
55 _json_metric_attributes = [
56 'metric_name', 'metric_group', 'metric_expr', 'metric_threshold',
57 'desc', 'long_desc', 'unit', 'compat', 'metricgroup_no_group',
58 'default_metricgroup_name', 'aggr_mode', 'event_grouping'
59 ]
60 # Attributes that are bools or enum int values, encoded as '0', '1',...
61 _json_enum_attributes = ['aggr_mode', 'deprecated', 'event_grouping', 'perpkg']
62
63 def removesuffix(s: str, suffix: str) -> str:
64 """Remove the suffix from a string
65
66 The removesuffix function is added to str in Python 3.9. We aim for 3.6
67 compatibility and so provide our own function here.
68 """
69 return s[0:-len(suffix)] if s.endswith(suffix) else s
70
71
72 def file_name_to_table_name(prefix: str, parents: Sequence[str],
73 dirname: str) -> str:
74 """Generate a C table name from directory names."""
75 tblname = prefix
76 for p in parents:
77 tblname += '_' + p
78 tblname += '_' + dirname
79 return tblname.replace('-', '_')
80
81
82 def c_len(s: str) -> int:
83 """Return the length of s a C string
84
85 This doesn't handle all escape characters properly. It first assumes
86 all \\ are for escaping, it then adjusts as it will have over counted
87 \\. The code uses \000 rather than \0 as a terminator as an adjacent
88 number would be folded into a string of \0 (ie. "\0" + "5" doesn't
89 equal a terminator followed by the number 5 but the escape of
90 \05). The code adjusts for \000 but not properly for all octal, hex
91 or unicode values.
92 """
93 try:
94 utf = s.encode(encoding='utf-8',errors='strict')
95 except:
96 print(f'broken string {s}')
97 raise
98 return len(utf) - utf.count(b'\\') + utf.count(b'\\\\') - (utf.count(b'\\000') * 2)
99
100 class BigCString:
101 """A class to hold many strings concatenated together.
102
103 Generating a large number of stand-alone C strings creates a large
104 number of relocations in position independent code. The BigCString
105 is a helper for this case. It builds a single string which within it
106 are all the other C strings (to avoid memory issues the string
107 itself is held as a list of strings). The offsets within the big
108 string are recorded and when stored to disk these don't need
109 relocation. To reduce the size of the string further, identical
110 strings are merged. If a longer string ends-with the same value as a
111 shorter string, these entries are also merged.
112 """
113 strings: Set[str]
114 big_string: Sequence[str]
115 offsets: Dict[str, int]
116 insert_number: int
117 insert_point: Dict[str, int]
118 metrics: Set[str]
119
120 def __init__(self):
121 self.strings = set()
122 self.insert_number = 0;
123 self.insert_point = {}
124 self.metrics = set()
125
126 def add(self, s: str, metric: bool) -> None:
127 """Called to add to the big string."""
128 if s not in self.strings:
129 self.strings.add(s)
130 self.insert_point[s] = self.insert_number
131 self.insert_number += 1
132 if metric:
133 self.metrics.add(s)
134
135 def compute(self) -> None:
136 """Called once all strings are added to compute the string and offsets."""
137
138 folded_strings = {}
139 # Determine if two strings can be folded, ie. let 1 string use the
140 # end of another. First reverse all strings and sort them.
141 sorted_reversed_strings = sorted([x[::-1] for x in self.strings])
142
143 # Strings 'xyz' and 'yz' will now be [ 'zy', 'zyx' ]. Scan forward
144 # for each string to see if there is a better candidate to fold it
145 # into, in the example rather than using 'yz' we can use'xyz' at
146 # an offset of 1. We record which string can be folded into which
147 # in folded_strings, we don't need to record the offset as it is
148 # trivially computed from the string lengths.
149 for pos,s in enumerate(sorted_reversed_strings):
150 best_pos = pos
151 for check_pos in range(pos + 1, len(sorted_reversed_strings)):
152 if sorted_reversed_strings[check_pos].startswith(s):
153 best_pos = check_pos
154 else:
155 break
156 if pos != best_pos:
157 folded_strings[s[::-1]] = sorted_reversed_strings[best_pos][::-1]
158
159 # Compute reverse mappings for debugging.
160 fold_into_strings = collections.defaultdict(set)
161 for key, val in folded_strings.items():
162 if key != val:
163 fold_into_strings[val].add(key)
164
165 # big_string_offset is the current location within the C string
166 # being appended to - comments, etc. don't count. big_string is
167 # the string contents represented as a list. Strings are immutable
168 # in Python and so appending to one causes memory issues, while
169 # lists are mutable.
170 big_string_offset = 0
171 self.big_string = []
172 self.offsets = {}
173
174 def string_cmp_key(s: str) -> Tuple[bool, int, str]:
175 return (s in self.metrics, self.insert_point[s], s)
176
177 # Emit all strings that aren't folded in a sorted manner.
178 for s in sorted(self.strings, key=string_cmp_key):
179 if s not in folded_strings:
180 self.offsets[s] = big_string_offset
181 self.big_string.append(f'/* offset={big_string_offset} */ "')
182 self.big_string.append(s)
183 self.big_string.append('"')
184 if s in fold_into_strings:
185 self.big_string.append(' /* also: ' + ', '.join(fold_into_strings[s]) + ' */')
186 self.big_string.append('\n')
187 big_string_offset += c_len(s)
188 continue
189
190 # Compute the offsets of the folded strings.
191 for s in folded_strings.keys():
192 assert s not in self.offsets
193 folded_s = folded_strings[s]
194 self.offsets[s] = self.offsets[folded_s] + c_len(folded_s) - c_len(s)
195
196 _bcs = BigCString()
197
198 class JsonEvent:
199 """Representation of an event loaded from a json file dictionary."""
200
201 def __init__(self, jd: dict):
202 """Constructor passed the dictionary of parsed json values."""
203
204 def llx(x: int) -> str:
205 """Convert an int to a string similar to a printf modifier of %#llx."""
206 return str(x) if x >= 0 and x < 10 else hex(x)
207
208 def fixdesc(s: str) -> str:
209 """Fix formatting issue for the desc string."""
210 if s is None:
211 return None
212 return removesuffix(removesuffix(removesuffix(s, '. '),
213 '. '), '.').replace('\n', '\\n').replace(
214 '\"', '\\"').replace('\r', '\\r')
215
216 def convert_aggr_mode(aggr_mode: str) -> Optional[str]:
217 """Returns the aggr_mode_class enum value associated with the JSON string."""
218 if not aggr_mode:
219 return None
220 aggr_mode_to_enum = {
221 'PerChip': '1',
222 'PerCore': '2',
223 }
224 return aggr_mode_to_enum[aggr_mode]
225
226 def convert_metric_constraint(metric_constraint: str) -> Optional[str]:
227 """Returns the metric_event_groups enum value associated with the JSON string."""
228 if not metric_constraint:
229 return None
230 metric_constraint_to_enum = {
231 'NO_GROUP_EVENTS': '1',
232 'NO_GROUP_EVENTS_NMI': '2',
233 'NO_NMI_WATCHDOG': '2',
234 'NO_GROUP_EVENTS_SMT': '3',
235 }
236 return metric_constraint_to_enum[metric_constraint]
237
238 def lookup_msr(num: str) -> Optional[str]:
239 """Converts the msr number, or first in a list to the appropriate event field."""
240 if not num:
241 return None
242 msrmap = {
243 0x3F6: 'ldlat=',
244 0x1A6: 'offcore_rsp=',
245 0x1A7: 'offcore_rsp=',
246 0x3F7: 'frontend=',
247 }
248 return msrmap[int(num.split(',', 1)[0], 0)]
249
250 def real_event(name: str, event: str) -> Optional[str]:
251 """Convert well known event names to an event string otherwise use the event argument."""
252 fixed = {
253 'inst_retired.any': 'event=0xc0,period=2000003',
254 'inst_retired.any_p': 'event=0xc0,period=2000003',
255 'cpu_clk_unhalted.ref': 'event=0x0,umask=0x03,period=2000003',
256 'cpu_clk_unhalted.thread': 'event=0x3c,period=2000003',
257 'cpu_clk_unhalted.core': 'event=0x3c,period=2000003',
258 'cpu_clk_unhalted.thread_any': 'event=0x3c,any=1,period=2000003',
259 }
260 if not name:
261 return None
262 if name.lower() in fixed:
263 return fixed[name.lower()]
264 return event
265
266 def unit_to_pmu(unit: str) -> Optional[str]:
267 """Convert a JSON Unit to Linux PMU name."""
268 if not unit:
269 return 'default_core'
270 # Comment brought over from jevents.c:
271 # it's not realistic to keep adding these, we need something more scalable ...
272 table = {
273 'CBO': 'uncore_cbox',
274 'QPI LL': 'uncore_qpi',
275 'SBO': 'uncore_sbox',
276 'iMPH-U': 'uncore_arb',
277 'CPU-M-CF': 'cpum_cf',
278 'CPU-M-SF': 'cpum_sf',
279 'PAI-CRYPTO' : 'pai_crypto',
280 'PAI-EXT' : 'pai_ext',
281 'UPI LL': 'uncore_upi',
282 'hisi_sicl,cpa': 'hisi_sicl,cpa',
283 'hisi_sccl,ddrc': 'hisi_sccl,ddrc',
284 'hisi_sccl,hha': 'hisi_sccl,hha',
285 'hisi_sccl,l3c': 'hisi_sccl,l3c',
286 'imx8_ddr': 'imx8_ddr',
287 'imx9_ddr': 'imx9_ddr',
288 'L3PMC': 'amd_l3',
289 'DFPMC': 'amd_df',
290 'UMCPMC': 'amd_umc',
291 'cpu_core': 'cpu_core',
292 'cpu_atom': 'cpu_atom',
293 'ali_drw': 'ali_drw',
294 'arm_cmn': 'arm_cmn',
295 }
296 return table[unit] if unit in table else f'uncore_{unit.lower()}'
297
298 def is_zero(val: str) -> bool:
299 try:
300 if val.startswith('0x'):
301 return int(val, 16) == 0
302 else:
303 return int(val) == 0
304 except e:
305 return False
306
307 def canonicalize_value(val: str) -> str:
308 try:
309 if val.startswith('0x'):
310 return llx(int(val, 16))
311 return str(int(val))
312 except e:
313 return val
314
315 eventcode = 0
316 if 'EventCode' in jd:
317 eventcode = int(jd['EventCode'].split(',', 1)[0], 0)
318 if 'ExtSel' in jd:
319 eventcode |= int(jd['ExtSel']) << 8
320 configcode = int(jd['ConfigCode'], 0) if 'ConfigCode' in jd else None
321 eventidcode = int(jd['EventidCode'], 0) if 'EventidCode' in jd else None
322 self.name = jd['EventName'].lower() if 'EventName' in jd else None
323 self.topic = ''
324 self.compat = jd.get('Compat')
325 self.desc = fixdesc(jd.get('BriefDescription'))
326 self.long_desc = fixdesc(jd.get('PublicDescription'))
327 precise = jd.get('PEBS')
328 msr = lookup_msr(jd.get('MSRIndex'))
329 msrval = jd.get('MSRValue')
330 extra_desc = ''
331 if 'Data_LA' in jd:
332 extra_desc += ' Supports address when precise'
333 if 'Errata' in jd:
334 extra_desc += '.'
335 if 'Errata' in jd:
336 extra_desc += ' Spec update: ' + jd['Errata']
337 self.pmu = unit_to_pmu(jd.get('Unit'))
338 filter = jd.get('Filter')
339 self.unit = jd.get('ScaleUnit')
340 self.perpkg = jd.get('PerPkg')
341 self.aggr_mode = convert_aggr_mode(jd.get('AggregationMode'))
342 self.deprecated = jd.get('Deprecated')
343 self.metric_name = jd.get('MetricName')
344 self.metric_group = jd.get('MetricGroup')
345 self.metricgroup_no_group = jd.get('MetricgroupNoGroup')
346 self.default_metricgroup_name = jd.get('DefaultMetricgroupName')
347 self.event_grouping = convert_metric_constraint(jd.get('MetricConstraint'))
348 self.metric_expr = None
349 if 'MetricExpr' in jd:
350 self.metric_expr = metric.ParsePerfJson(jd['MetricExpr']).Simplify()
351 # Note, the metric formula for the threshold isn't parsed as the &
352 # and > have incorrect precedence.
353 self.metric_threshold = jd.get('MetricThreshold')
354
355 arch_std = jd.get('ArchStdEvent')
356 if precise and self.desc and '(Precise Event)' not in self.desc:
357 extra_desc += ' (Must be precise)' if precise == '2' else (' (Precise '
358 'event)')
359 event = None
360 if configcode is not None:
361 event = f'config={llx(configcode)}'
362 elif eventidcode is not None:
363 event = f'eventid={llx(eventidcode)}'
364 else:
365 event = f'event={llx(eventcode)}'
366 event_fields = [
367 ('AnyThread', 'any='),
368 ('PortMask', 'ch_mask='),
369 ('CounterMask', 'cmask='),
370 ('EdgeDetect', 'edge='),
371 ('FCMask', 'fc_mask='),
372 ('Invert', 'inv='),
373 ('SampleAfterValue', 'period='),
374 ('UMask', 'umask='),
375 ('NodeType', 'type='),
376 ('RdWrMask', 'rdwrmask='),
377 ('EnAllCores', 'enallcores='),
378 ('EnAllSlices', 'enallslices='),
379 ('SliceId', 'sliceid='),
380 ('ThreadMask', 'threadmask='),
381 ]
382 for key, value in event_fields:
383 if key in jd and not is_zero(jd[key]):
384 event += f',{value}{canonicalize_value(jd[key])}'
385 if filter:
386 event += f',{filter}'
387 if msr:
388 event += f',{msr}{msrval}'
389 if self.desc and extra_desc:
390 self.desc += extra_desc
391 if self.long_desc and extra_desc:
392 self.long_desc += extra_desc
393 if arch_std:
394 if arch_std.lower() in _arch_std_events:
395 event = _arch_std_events[arch_std.lower()].event
396 # Copy from the architecture standard event to self for undefined fields.
397 for attr, value in _arch_std_events[arch_std.lower()].__dict__.items():
398 if hasattr(self, attr) and not getattr(self, attr):
399 setattr(self, attr, value)
400 else:
401 raise argparse.ArgumentTypeError('Cannot find arch std event:', arch_std)
402
403 self.event = real_event(self.name, event)
404
405 def __repr__(self) -> str:
406 """String representation primarily for debugging."""
407 s = '{\n'
408 for attr, value in self.__dict__.items():
409 if value:
410 s += f'\t{attr} = {value},\n'
411 return s + '}'
412
413 def build_c_string(self, metric: bool) -> str:
414 s = ''
415 for attr in _json_metric_attributes if metric else _json_event_attributes:
416 x = getattr(self, attr)
417 if metric and x and attr == 'metric_expr':
418 # Convert parsed metric expressions into a string. Slashes
419 # must be doubled in the file.
420 x = x.ToPerfJson().replace('\\', '\\\\')
421 if metric and x and attr == 'metric_threshold':
422 x = x.replace('\\', '\\\\')
423 if attr in _json_enum_attributes:
424 s += x if x else '0'
425 else:
426 s += f'{x}\\000' if x else '\\000'
427 return s
428
429 def to_c_string(self, metric: bool) -> str:
430 """Representation of the event as a C struct initializer."""
431
432 s = self.build_c_string(metric)
433 return f'{{ { _bcs.offsets[s] } }}, /* {s} */\n'
434
435
436 @lru_cache(maxsize=None)
437 def read_json_events(path: str, topic: str) -> Sequence[JsonEvent]:
438 """Read json events from the specified file."""
439 try:
440 events = json.load(open(path), object_hook=JsonEvent)
441 except BaseException as err:
442 print(f"Exception processing {path}")
443 raise
444 metrics: list[Tuple[str, str, metric.Expression]] = []
445 for event in events:
446 event.topic = topic
447 if event.metric_name and '-' not in event.metric_name:
448 metrics.append((event.pmu, event.metric_name, event.metric_expr))
449 updates = metric.RewriteMetricsInTermsOfOthers(metrics)
450 if updates:
451 for event in events:
452 if event.metric_name in updates:
453 # print(f'Updated {event.metric_name} from\n"{event.metric_expr}"\n'
454 # f'to\n"{updates[event.metric_name]}"')
455 event.metric_expr = updates[event.metric_name]
456
457 return events
458
459 def preprocess_arch_std_files(archpath: str) -> None:
460 """Read in all architecture standard events."""
461 global _arch_std_events
462 for item in os.scandir(archpath):
463 if item.is_file() and item.name.endswith('.json'):
464 for event in read_json_events(item.path, topic=''):
465 if event.name:
466 _arch_std_events[event.name.lower()] = event
467 if event.metric_name:
468 _arch_std_events[event.metric_name.lower()] = event
469
470
471 def add_events_table_entries(item: os.DirEntry, topic: str) -> None:
472 """Add contents of file to _pending_events table."""
473 for e in read_json_events(item.path, topic):
474 if e.name:
475 _pending_events.append(e)
476 if e.metric_name:
477 _pending_metrics.append(e)
478
479
480 def print_pending_events() -> None:
481 """Optionally close events table."""
482
483 def event_cmp_key(j: JsonEvent) -> Tuple[str, str, bool, str, str]:
484 def fix_none(s: Optional[str]) -> str:
485 if s is None:
486 return ''
487 return s
488
489 return (fix_none(j.pmu).replace(',','_'), fix_none(j.name), j.desc is not None, fix_none(j.topic),
490 fix_none(j.metric_name))
491
492 global _pending_events
493 if not _pending_events:
494 return
495
496 global _pending_events_tblname
497 if _pending_events_tblname.endswith('_sys'):
498 global _sys_event_tables
499 _sys_event_tables.append(_pending_events_tblname)
500 else:
501 global event_tables
502 _event_tables.append(_pending_events_tblname)
503
504 first = True
505 last_pmu = None
506 last_name = None
507 pmus = set()
508 for event in sorted(_pending_events, key=event_cmp_key):
509 if last_pmu and last_pmu == event.pmu:
510 assert event.name != last_name, f"Duplicate event: {last_pmu}/{last_name}/ in {_pending_events_tblname}"
511 if event.pmu != last_pmu:
512 if not first:
513 _args.output_file.write('};\n')
514 pmu_name = event.pmu.replace(',', '_')
515 _args.output_file.write(
516 f'static const struct compact_pmu_event {_pending_events_tblname}_{pmu_name}[] = {{\n')
517 first = False
518 last_pmu = event.pmu
519 pmus.add((event.pmu, pmu_name))
520
521 _args.output_file.write(event.to_c_string(metric=False))
522 last_name = event.name
523 _pending_events = []
524
525 _args.output_file.write(f"""
526 }};
527
528 const struct pmu_table_entry {_pending_events_tblname}[] = {{
529 """)
530 for (pmu, tbl_pmu) in sorted(pmus):
531 pmu_name = f"{pmu}\\000"
532 _args.output_file.write(f"""{{
533 .entries = {_pending_events_tblname}_{tbl_pmu},
534 .num_entries = ARRAY_SIZE({_pending_events_tblname}_{tbl_pmu}),
535 .pmu_name = {{ {_bcs.offsets[pmu_name]} /* {pmu_name} */ }},
536 }},
537 """)
538 _args.output_file.write('};\n\n')
539
540 def print_pending_metrics() -> None:
541 """Optionally close metrics table."""
542
543 def metric_cmp_key(j: JsonEvent) -> Tuple[bool, str, str]:
544 def fix_none(s: Optional[str]) -> str:
545 if s is None:
546 return ''
547 return s
548
549 return (j.desc is not None, fix_none(j.pmu), fix_none(j.metric_name))
550
551 global _pending_metrics
552 if not _pending_metrics:
553 return
554
555 global _pending_metrics_tblname
556 if _pending_metrics_tblname.endswith('_sys'):
557 global _sys_metric_tables
558 _sys_metric_tables.append(_pending_metrics_tblname)
559 else:
560 global metric_tables
561 _metric_tables.append(_pending_metrics_tblname)
562
563 first = True
564 last_pmu = None
565 pmus = set()
566 for metric in sorted(_pending_metrics, key=metric_cmp_key):
567 if metric.pmu != last_pmu:
568 if not first:
569 _args.output_file.write('};\n')
570 pmu_name = metric.pmu.replace(',', '_')
571 _args.output_file.write(
572 f'static const struct compact_pmu_event {_pending_metrics_tblname}_{pmu_name}[] = {{\n')
573 first = False
574 last_pmu = metric.pmu
575 pmus.add((metric.pmu, pmu_name))
576
577 _args.output_file.write(metric.to_c_string(metric=True))
578 _pending_metrics = []
579
580 _args.output_file.write(f"""
581 }};
582
583 const struct pmu_table_entry {_pending_metrics_tblname}[] = {{
584 """)
585 for (pmu, tbl_pmu) in sorted(pmus):
586 pmu_name = f"{pmu}\\000"
587 _args.output_file.write(f"""{{
588 .entries = {_pending_metrics_tblname}_{tbl_pmu},
589 .num_entries = ARRAY_SIZE({_pending_metrics_tblname}_{tbl_pmu}),
590 .pmu_name = {{ {_bcs.offsets[pmu_name]} /* {pmu_name} */ }},
591 }},
592 """)
593 _args.output_file.write('};\n\n')
594
595 def get_topic(topic: str) -> str:
596 if topic.endswith('metrics.json'):
597 return 'metrics'
598 return removesuffix(topic, '.json').replace('-', ' ')
599
600 def preprocess_one_file(parents: Sequence[str], item: os.DirEntry) -> None:
601
602 if item.is_dir():
603 return
604
605 # base dir or too deep
606 level = len(parents)
607 if level == 0 or level > 4:
608 return
609
610 # Ignore other directories. If the file name does not have a .json
611 # extension, ignore it. It could be a readme.txt for instance.
612 if not item.is_file() or not item.name.endswith('.json'):
613 return
614
615 if item.name == 'metricgroups.json':
616 metricgroup_descriptions = json.load(open(item.path))
617 for mgroup in metricgroup_descriptions:
618 assert len(mgroup) > 1, parents
619 description = f"{metricgroup_descriptions[mgroup]}\\000"
620 mgroup = f"{mgroup}\\000"
621 _bcs.add(mgroup, metric=True)
622 _bcs.add(description, metric=True)
623 _metricgroups[mgroup] = description
624 return
625
626 topic = get_topic(item.name)
627 for event in read_json_events(item.path, topic):
628 pmu_name = f"{event.pmu}\\000"
629 if event.name:
630 _bcs.add(pmu_name, metric=False)
631 _bcs.add(event.build_c_string(metric=False), metric=False)
632 if event.metric_name:
633 _bcs.add(pmu_name, metric=True)
634 _bcs.add(event.build_c_string(metric=True), metric=True)
635
636 def process_one_file(parents: Sequence[str], item: os.DirEntry) -> None:
637 """Process a JSON file during the main walk."""
638 def is_leaf_dir_ignoring_sys(path: str) -> bool:
639 for item in os.scandir(path):
640 if item.is_dir() and item.name != 'sys':
641 return False
642 return True
643
644 # Model directories are leaves (ignoring possible sys
645 # directories). The FTW will walk into the directory next. Flush
646 # pending events and metrics and update the table names for the new
647 # model directory.
648 if item.is_dir() and is_leaf_dir_ignoring_sys(item.path):
649 print_pending_events()
650 print_pending_metrics()
651
652 global _pending_events_tblname
653 _pending_events_tblname = file_name_to_table_name('pmu_events_', parents, item.name)
654 global _pending_metrics_tblname
655 _pending_metrics_tblname = file_name_to_table_name('pmu_metrics_', parents, item.name)
656
657 if item.name == 'sys':
658 _sys_event_table_to_metric_table_mapping[_pending_events_tblname] = _pending_metrics_tblname
659 return
660
661 # base dir or too deep
662 level = len(parents)
663 if level == 0 or level > 4:
664 return
665
666 # Ignore other directories. If the file name does not have a .json
667 # extension, ignore it. It could be a readme.txt for instance.
668 if not item.is_file() or not item.name.endswith('.json') or item.name == 'metricgroups.json':
669 return
670
671 add_events_table_entries(item, get_topic(item.name))
672
673
674 def print_mapping_table(archs: Sequence[str]) -> None:
675 """Read the mapfile and generate the struct from cpuid string to event table."""
676 _args.output_file.write("""
677 /* Struct used to make the PMU event table implementation opaque to callers. */
678 struct pmu_events_table {
679 const struct pmu_table_entry *pmus;
680 uint32_t num_pmus;
681 };
682
683 /* Struct used to make the PMU metric table implementation opaque to callers. */
684 struct pmu_metrics_table {
685 const struct pmu_table_entry *pmus;
686 uint32_t num_pmus;
687 };
688
689 /*
690 * Map a CPU to its table of PMU events. The CPU is identified by the
691 * cpuid field, which is an arch-specific identifier for the CPU.
692 * The identifier specified in tools/perf/pmu-events/arch/xxx/mapfile
693 * must match the get_cpuid_str() in tools/perf/arch/xxx/util/header.c)
694 *
695 * The cpuid can contain any character other than the comma.
696 */
697 struct pmu_events_map {
698 const char *arch;
699 const char *cpuid;
700 struct pmu_events_table event_table;
701 struct pmu_metrics_table metric_table;
702 };
703
704 /*
705 * Global table mapping each known CPU for the architecture to its
706 * table of PMU events.
707 */
708 const struct pmu_events_map pmu_events_map[] = {
709 """)
710 for arch in archs:
711 if arch == 'test':
712 _args.output_file.write("""{
713 \t.arch = "testarch",
714 \t.cpuid = "testcpu",
715 \t.event_table = {
716 \t\t.pmus = pmu_events__test_soc_cpu,
717 \t\t.num_pmus = ARRAY_SIZE(pmu_events__test_soc_cpu),
718 \t},
719 \t.metric_table = {
720 \t\t.pmus = pmu_metrics__test_soc_cpu,
721 \t\t.num_pmus = ARRAY_SIZE(pmu_metrics__test_soc_cpu),
722 \t}
723 },
724 """)
725 else:
726 with open(f'{_args.starting_dir}/{arch}/mapfile.csv') as csvfile:
727 table = csv.reader(csvfile)
728 first = True
729 for row in table:
730 # Skip the first row or any row beginning with #.
731 if not first and len(row) > 0 and not row[0].startswith('#'):
732 event_tblname = file_name_to_table_name('pmu_events_', [], row[2].replace('/', '_'))
733 if event_tblname in _event_tables:
734 event_size = f'ARRAY_SIZE({event_tblname})'
735 else:
736 event_tblname = 'NULL'
737 event_size = '0'
738 metric_tblname = file_name_to_table_name('pmu_metrics_', [], row[2].replace('/', '_'))
739 if metric_tblname in _metric_tables:
740 metric_size = f'ARRAY_SIZE({metric_tblname})'
741 else:
742 metric_tblname = 'NULL'
743 metric_size = '0'
744 if event_size == '0' and metric_size == '0':
745 continue
746 cpuid = row[0].replace('\\', '\\\\')
747 _args.output_file.write(f"""{{
748 \t.arch = "{arch}",
749 \t.cpuid = "{cpuid}",
750 \t.event_table = {{
751 \t\t.pmus = {event_tblname},
752 \t\t.num_pmus = {event_size}
753 \t}},
754 \t.metric_table = {{
755 \t\t.pmus = {metric_tblname},
756 \t\t.num_pmus = {metric_size}
757 \t}}
758 }},
759 """)
760 first = False
761
762 _args.output_file.write("""{
763 \t.arch = 0,
764 \t.cpuid = 0,
765 \t.event_table = { 0, 0 },
766 \t.metric_table = { 0, 0 },
767 }
768 };
769 """)
770
771
772 def print_system_mapping_table() -> None:
773 """C struct mapping table array for tables from /sys directories."""
774 _args.output_file.write("""
775 struct pmu_sys_events {
776 \tconst char *name;
777 \tstruct pmu_events_table event_table;
778 \tstruct pmu_metrics_table metric_table;
779 };
780
781 static const struct pmu_sys_events pmu_sys_event_tables[] = {
782 """)
783 printed_metric_tables = []
784 for tblname in _sys_event_tables:
785 _args.output_file.write(f"""\t{{
786 \t\t.event_table = {{
787 \t\t\t.pmus = {tblname},
788 \t\t\t.num_pmus = ARRAY_SIZE({tblname})
789 \t\t}},""")
790 metric_tblname = _sys_event_table_to_metric_table_mapping[tblname]
791 if metric_tblname in _sys_metric_tables:
792 _args.output_file.write(f"""
793 \t\t.metric_table = {{
794 \t\t\t.pmus = {metric_tblname},
795 \t\t\t.num_pmus = ARRAY_SIZE({metric_tblname})
796 \t\t}},""")
797 printed_metric_tables.append(metric_tblname)
798 _args.output_file.write(f"""
799 \t\t.name = \"{tblname}\",
800 \t}},
801 """)
802 for tblname in _sys_metric_tables:
803 if tblname in printed_metric_tables:
804 continue
805 _args.output_file.write(f"""\t{{
806 \t\t.metric_table = {{
807 \t\t\t.pmus = {tblname},
808 \t\t\t.num_pmus = ARRAY_SIZE({tblname})
809 \t\t}},
810 \t\t.name = \"{tblname}\",
811 \t}},
812 """)
813 _args.output_file.write("""\t{
814 \t\t.event_table = { 0, 0 },
815 \t\t.metric_table = { 0, 0 },
816 \t},
817 };
818
819 static void decompress_event(int offset, struct pmu_event *pe)
820 {
821 \tconst char *p = &big_c_string[offset];
822 """)
823 for attr in _json_event_attributes:
824 _args.output_file.write(f'\n\tpe->{attr} = ')
825 if attr in _json_enum_attributes:
826 _args.output_file.write("*p - '0';\n")
827 else:
828 _args.output_file.write("(*p == '\\0' ? NULL : p);\n")
829 if attr == _json_event_attributes[-1]:
830 continue
831 if attr in _json_enum_attributes:
832 _args.output_file.write('\tp++;')
833 else:
834 _args.output_file.write('\twhile (*p++);')
835 _args.output_file.write("""}
836
837 static void decompress_metric(int offset, struct pmu_metric *pm)
838 {
839 \tconst char *p = &big_c_string[offset];
840 """)
841 for attr in _json_metric_attributes:
842 _args.output_file.write(f'\n\tpm->{attr} = ')
843 if attr in _json_enum_attributes:
844 _args.output_file.write("*p - '0';\n")
845 else:
846 _args.output_file.write("(*p == '\\0' ? NULL : p);\n")
847 if attr == _json_metric_attributes[-1]:
848 continue
849 if attr in _json_enum_attributes:
850 _args.output_file.write('\tp++;')
851 else:
852 _args.output_file.write('\twhile (*p++);')
853 _args.output_file.write("""}
854
855 static int pmu_events_table__for_each_event_pmu(const struct pmu_events_table *table,
856 const struct pmu_table_entry *pmu,
857 pmu_event_iter_fn fn,
858 void *data)
859 {
860 int ret;
861 struct pmu_event pe = {
862 .pmu = &big_c_string[pmu->pmu_name.offset],
863 };
864
865 for (uint32_t i = 0; i < pmu->num_entries; i++) {
866 decompress_event(pmu->entries[i].offset, &pe);
867 if (!pe.name)
868 continue;
869 ret = fn(&pe, table, data);
870 if (ret)
871 return ret;
872 }
873 return 0;
874 }
875
876 static int pmu_events_table__find_event_pmu(const struct pmu_events_table *table,
877 const struct pmu_table_entry *pmu,
878 const char *name,
879 pmu_event_iter_fn fn,
880 void *data)
881 {
882 struct pmu_event pe = {
883 .pmu = &big_c_string[pmu->pmu_name.offset],
884 };
885 int low = 0, high = pmu->num_entries - 1;
886
887 while (low <= high) {
888 int cmp, mid = (low + high) / 2;
889
890 decompress_event(pmu->entries[mid].offset, &pe);
891
892 if (!pe.name && !name)
893 goto do_call;
894
895 if (!pe.name && name) {
896 low = mid + 1;
897 continue;
898 }
899 if (pe.name && !name) {
900 high = mid - 1;
901 continue;
902 }
903
904 cmp = strcasecmp(pe.name, name);
905 if (cmp < 0) {
906 low = mid + 1;
907 continue;
908 }
909 if (cmp > 0) {
910 high = mid - 1;
911 continue;
912 }
913 do_call:
914 return fn ? fn(&pe, table, data) : 0;
915 }
916 return PMU_EVENTS__NOT_FOUND;
917 }
918
919 int pmu_events_table__for_each_event(const struct pmu_events_table *table,
920 struct perf_pmu *pmu,
921 pmu_event_iter_fn fn,
922 void *data)
923 {
924 for (size_t i = 0; i < table->num_pmus; i++) {
925 const struct pmu_table_entry *table_pmu = &table->pmus[i];
926 const char *pmu_name = &big_c_string[table_pmu->pmu_name.offset];
927 int ret;
928
929 if (pmu && !pmu__name_match(pmu, pmu_name))
930 continue;
931
932 ret = pmu_events_table__for_each_event_pmu(table, table_pmu, fn, data);
933 if (pmu || ret)
934 return ret;
935 }
936 return 0;
937 }
938
939 int pmu_events_table__find_event(const struct pmu_events_table *table,
940 struct perf_pmu *pmu,
941 const char *name,
942 pmu_event_iter_fn fn,
943 void *data)
944 {
945 for (size_t i = 0; i < table->num_pmus; i++) {
946 const struct pmu_table_entry *table_pmu = &table->pmus[i];
947 const char *pmu_name = &big_c_string[table_pmu->pmu_name.offset];
948 int ret;
949
950 if (!pmu__name_match(pmu, pmu_name))
951 continue;
952
953 ret = pmu_events_table__find_event_pmu(table, table_pmu, name, fn, data);
954 if (ret != PMU_EVENTS__NOT_FOUND)
955 return ret;
956 }
957 return PMU_EVENTS__NOT_FOUND;
958 }
959
960 size_t pmu_events_table__num_events(const struct pmu_events_table *table,
961 struct perf_pmu *pmu)
962 {
963 size_t count = 0;
964
965 for (size_t i = 0; i < table->num_pmus; i++) {
966 const struct pmu_table_entry *table_pmu = &table->pmus[i];
967 const char *pmu_name = &big_c_string[table_pmu->pmu_name.offset];
968
969 if (pmu__name_match(pmu, pmu_name))
970 count += table_pmu->num_entries;
971 }
972 return count;
973 }
974
975 static int pmu_metrics_table__for_each_metric_pmu(const struct pmu_metrics_table *table,
976 const struct pmu_table_entry *pmu,
977 pmu_metric_iter_fn fn,
978 void *data)
979 {
980 int ret;
981 struct pmu_metric pm = {
982 .pmu = &big_c_string[pmu->pmu_name.offset],
983 };
984
985 for (uint32_t i = 0; i < pmu->num_entries; i++) {
986 decompress_metric(pmu->entries[i].offset, &pm);
987 if (!pm.metric_expr)
988 continue;
989 ret = fn(&pm, table, data);
990 if (ret)
991 return ret;
992 }
993 return 0;
994 }
995
996 int pmu_metrics_table__for_each_metric(const struct pmu_metrics_table *table,
997 pmu_metric_iter_fn fn,
998 void *data)
999 {
1000 for (size_t i = 0; i < table->num_pmus; i++) {
1001 int ret = pmu_metrics_table__for_each_metric_pmu(table, &table->pmus[i],
1002 fn, data);
1003
1004 if (ret)
1005 return ret;
1006 }
1007 return 0;
1008 }
1009
1010 static const struct pmu_events_map *map_for_pmu(struct perf_pmu *pmu)
1011 {
1012 static struct {
1013 const struct pmu_events_map *map;
1014 struct perf_pmu *pmu;
1015 } last_result;
1016 static struct {
1017 const struct pmu_events_map *map;
1018 char *cpuid;
1019 } last_map_search;
1020 static bool has_last_result, has_last_map_search;
1021 const struct pmu_events_map *map = NULL;
1022 char *cpuid = NULL;
1023 size_t i;
1024
1025 if (has_last_result && last_result.pmu == pmu)
1026 return last_result.map;
1027
1028 cpuid = perf_pmu__getcpuid(pmu);
1029
1030 /*
1031 * On some platforms which uses cpus map, cpuid can be NULL for
1032 * PMUs other than CORE PMUs.
1033 */
1034 if (!cpuid)
1035 goto out_update_last_result;
1036
1037 if (has_last_map_search && !strcmp(last_map_search.cpuid, cpuid)) {
1038 map = last_map_search.map;
1039 free(cpuid);
1040 } else {
1041 i = 0;
1042 for (;;) {
1043 map = &pmu_events_map[i++];
1044
1045 if (!map->arch) {
1046 map = NULL;
1047 break;
1048 }
1049
1050 if (!strcmp_cpuid_str(map->cpuid, cpuid))
1051 break;
1052 }
1053 free(last_map_search.cpuid);
1054 last_map_search.cpuid = cpuid;
1055 last_map_search.map = map;
1056 has_last_map_search = true;
1057 }
1058 out_update_last_result:
1059 last_result.pmu = pmu;
1060 last_result.map = map;
1061 has_last_result = true;
1062 return map;
1063 }
1064
1065 const struct pmu_events_table *perf_pmu__find_events_table(struct perf_pmu *pmu)
1066 {
1067 const struct pmu_events_map *map = map_for_pmu(pmu);
1068
1069 if (!map)
1070 return NULL;
1071
1072 if (!pmu)
1073 return &map->event_table;
1074
1075 for (size_t i = 0; i < map->event_table.num_pmus; i++) {
1076 const struct pmu_table_entry *table_pmu = &map->event_table.pmus[i];
1077 const char *pmu_name = &big_c_string[table_pmu->pmu_name.offset];
1078
1079 if (pmu__name_match(pmu, pmu_name))
1080 return &map->event_table;
1081 }
1082 return NULL;
1083 }
1084
1085 const struct pmu_metrics_table *perf_pmu__find_metrics_table(struct perf_pmu *pmu)
1086 {
1087 const struct pmu_events_map *map = map_for_pmu(pmu);
1088
1089 if (!map)
1090 return NULL;
1091
1092 if (!pmu)
1093 return &map->metric_table;
1094
1095 for (size_t i = 0; i < map->metric_table.num_pmus; i++) {
1096 const struct pmu_table_entry *table_pmu = &map->metric_table.pmus[i];
1097 const char *pmu_name = &big_c_string[table_pmu->pmu_name.offset];
1098
1099 if (pmu__name_match(pmu, pmu_name))
1100 return &map->metric_table;
1101 }
1102 return NULL;
1103 }
1104
1105 const struct pmu_events_table *find_core_events_table(const char *arch, const char *cpuid)
1106 {
1107 for (const struct pmu_events_map *tables = &pmu_events_map[0];
1108 tables->arch;
1109 tables++) {
1110 if (!strcmp(tables->arch, arch) && !strcmp_cpuid_str(tables->cpuid, cpuid))
1111 return &tables->event_table;
1112 }
1113 return NULL;
1114 }
1115
1116 const struct pmu_metrics_table *find_core_metrics_table(const char *arch, const char *cpuid)
1117 {
1118 for (const struct pmu_events_map *tables = &pmu_events_map[0];
1119 tables->arch;
1120 tables++) {
1121 if (!strcmp(tables->arch, arch) && !strcmp_cpuid_str(tables->cpuid, cpuid))
1122 return &tables->metric_table;
1123 }
1124 return NULL;
1125 }
1126
1127 int pmu_for_each_core_event(pmu_event_iter_fn fn, void *data)
1128 {
1129 for (const struct pmu_events_map *tables = &pmu_events_map[0];
1130 tables->arch;
1131 tables++) {
1132 int ret = pmu_events_table__for_each_event(&tables->event_table,
1133 /*pmu=*/ NULL, fn, data);
1134
1135 if (ret)
1136 return ret;
1137 }
1138 return 0;
1139 }
1140
1141 int pmu_for_each_core_metric(pmu_metric_iter_fn fn, void *data)
1142 {
1143 for (const struct pmu_events_map *tables = &pmu_events_map[0];
1144 tables->arch;
1145 tables++) {
1146 int ret = pmu_metrics_table__for_each_metric(&tables->metric_table, fn, data);
1147
1148 if (ret)
1149 return ret;
1150 }
1151 return 0;
1152 }
1153
1154 const struct pmu_events_table *find_sys_events_table(const char *name)
1155 {
1156 for (const struct pmu_sys_events *tables = &pmu_sys_event_tables[0];
1157 tables->name;
1158 tables++) {
1159 if (!strcmp(tables->name, name))
1160 return &tables->event_table;
1161 }
1162 return NULL;
1163 }
1164
1165 int pmu_for_each_sys_event(pmu_event_iter_fn fn, void *data)
1166 {
1167 for (const struct pmu_sys_events *tables = &pmu_sys_event_tables[0];
1168 tables->name;
1169 tables++) {
1170 int ret = pmu_events_table__for_each_event(&tables->event_table,
1171 /*pmu=*/ NULL, fn, data);
1172
1173 if (ret)
1174 return ret;
1175 }
1176 return 0;
1177 }
1178
1179 int pmu_for_each_sys_metric(pmu_metric_iter_fn fn, void *data)
1180 {
1181 for (const struct pmu_sys_events *tables = &pmu_sys_event_tables[0];
1182 tables->name;
1183 tables++) {
1184 int ret = pmu_metrics_table__for_each_metric(&tables->metric_table, fn, data);
1185
1186 if (ret)
1187 return ret;
1188 }
1189 return 0;
1190 }
1191 """)
1192
1193 def print_metricgroups() -> None:
1194 _args.output_file.write("""
1195 static const int metricgroups[][2] = {
1196 """)
1197 for mgroup in sorted(_metricgroups):
1198 description = _metricgroups[mgroup]
1199 _args.output_file.write(
1200 f'\t{{ {_bcs.offsets[mgroup]}, {_bcs.offsets[description]} }}, /* {mgroup} => {description} */\n'
1201 )
1202 _args.output_file.write("""
1203 };
1204
1205 const char *describe_metricgroup(const char *group)
1206 {
1207 int low = 0, high = (int)ARRAY_SIZE(metricgroups) - 1;
1208
1209 while (low <= high) {
1210 int mid = (low + high) / 2;
1211 const char *mgroup = &big_c_string[metricgroups[mid][0]];
1212 int cmp = strcmp(mgroup, group);
1213
1214 if (cmp == 0) {
1215 return &big_c_string[metricgroups[mid][1]];
1216 } else if (cmp < 0) {
1217 low = mid + 1;
1218 } else {
1219 high = mid - 1;
1220 }
1221 }
1222 return NULL;
1223 }
1224 """)
1225
1226 def main() -> None:
1227 global _args
1228
1229 def dir_path(path: str) -> str:
1230 """Validate path is a directory for argparse."""
1231 if os.path.isdir(path):
1232 return path
1233 raise argparse.ArgumentTypeError(f'\'{path}\' is not a valid directory')
1234
1235 def ftw(path: str, parents: Sequence[str],
1236 action: Callable[[Sequence[str], os.DirEntry], None]) -> None:
1237 """Replicate the directory/file walking behavior of C's file tree walk."""
1238 for item in sorted(os.scandir(path), key=lambda e: e.name):
1239 if _args.model != 'all' and item.is_dir():
1240 # Check if the model matches one in _args.model.
1241 if len(parents) == _args.model.split(',')[0].count('/'):
1242 # We're testing the correct directory.
1243 item_path = '/'.join(parents) + ('/' if len(parents) > 0 else '') + item.name
1244 if 'test' not in item_path and item_path not in _args.model.split(','):
1245 continue
1246 action(parents, item)
1247 if item.is_dir():
1248 ftw(item.path, parents + [item.name], action)
1249
1250 ap = argparse.ArgumentParser()
1251 ap.add_argument('arch', help='Architecture name like x86')
1252 ap.add_argument('model', help='''Select a model such as skylake to
1253 reduce the code size. Normally set to "all". For architectures like
1254 ARM64 with an implementor/model, the model must include the implementor
1255 such as "arm/cortex-a34".''',
1256 default='all')
1257 ap.add_argument(
1258 'starting_dir',
1259 type=dir_path,
1260 help='Root of tree containing architecture directories containing json files'
1261 )
1262 ap.add_argument(
1263 'output_file', type=argparse.FileType('w', encoding='utf-8'), nargs='?', default=sys.stdout)
1264 _args = ap.parse_args()
1265
1266 _args.output_file.write(f"""
1267 /* SPDX-License-Identifier: GPL-2.0 */
1268 /* THIS FILE WAS AUTOGENERATED BY jevents.py arch={_args.arch} model={_args.model} ! */
1269 """)
1270 _args.output_file.write("""
1271 #include <pmu-events/pmu-events.h>
1272 #include "util/header.h"
1273 #include "util/pmu.h"
1274 #include <string.h>
1275 #include <stddef.h>
1276
1277 struct compact_pmu_event {
1278 int offset;
1279 };
1280
1281 struct pmu_table_entry {
1282 const struct compact_pmu_event *entries;
1283 uint32_t num_entries;
1284 struct compact_pmu_event pmu_name;
1285 };
1286
1287 """)
1288 archs = []
1289 for item in os.scandir(_args.starting_dir):
1290 if not item.is_dir():
1291 continue
1292 if item.name == _args.arch or _args.arch == 'all' or item.name == 'test':
1293 archs.append(item.name)
1294
1295 if len(archs) < 2 and _args.arch != 'none':
1296 raise IOError(f'Missing architecture directory \'{_args.arch}\'')
1297
1298 archs.sort()
1299 for arch in archs:
1300 arch_path = f'{_args.starting_dir}/{arch}'
1301 preprocess_arch_std_files(arch_path)
1302 ftw(arch_path, [], preprocess_one_file)
1303
1304 _bcs.compute()
1305 _args.output_file.write('static const char *const big_c_string =\n')
1306 for s in _bcs.big_string:
1307 _args.output_file.write(s)
1308 _args.output_file.write(';\n\n')
1309 for arch in archs:
1310 arch_path = f'{_args.starting_dir}/{arch}'
1311 ftw(arch_path, [], process_one_file)
1312 print_pending_events()
1313 print_pending_metrics()
1314
1315 print_mapping_table(archs)
1316 print_system_mapping_table()
1317 print_metricgroups()
1318
1319 if __name__ == '__main__':
1320 main()
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.