Repository navigation
Expand file tree
/
Copy pathmx_graalpython_python_benchmarks.py
More file actions
810 lines (707 loc) · 28.3 KB
/
Copy pathmx_graalpython_python_benchmarks.py
File metadata and controls
810 lines (707 loc) · 28.3 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
# Copyright (c) 2020, 2025, Oracle and/or its affiliates. All rights reserved.
# DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
#
# The Universal Permissive License (UPL), Version 1.0
#
# Subject to the condition set forth below, permission is hereby granted to any
# person obtaining a copy of this software, associated documentation and/or
# data (collectively the "Software"), free of charge and under any and all
# copyright rights in the Software, and any and all patent rights owned or
# freely licensable by each licensor hereunder covering either (i) the
# unmodified Software as contributed to or provided by such licensor, or (ii)
# the Larger Works (as defined below), to deal in both
#
# (a) the Software, and
#
# (b) any piece of software and/or hardware listed in the lrgrwrks.txt file if
# one is included with the Software each a "Larger Work" to which the Software
# is contributed by such licensors),
#
# without restriction, including without limitation the rights to copy, create
# derivative works of, display, perform, and distribute the Software and make,
# use, sell, offer for sale, import, export, have made, and have sold the
# Software and the Larger Work(s), and to sublicense the foregoing rights on
# either these or other terms.
#
# This license is subject to the following condition:
#
# The above copyright notice and either this complete permission notice or at a
# minimum a reference to the UPL must be included in all copies or substantial
# portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
# SOFTWARE.
import mx
import mx_benchmark
import glob
import json
import math
import os
import re
import shutil
import sys
from os.path import join, abspath
SUITE = None
python_vm_registry = None
# By default we disabled some benchmarks, both because some don't run and
# because we want to reduce the total runtime of the suites.
DEFAULT_NUMPY_BENCHMARKS = [
"bench_app",
"bench_core",
# "bench_function_base",
"bench_indexing",
# "bench_io",
"bench_linalg",
# "bench_ma",
# "bench_random",
"bench_reduce",
# "bench_shape_base",
# "bench_ufunc",
]
SKIPPED_NUMPY_BENCHMARKS = [
"bench_core.CountNonzero.time_count_nonzero(2, 1000000, <class 'str'>)", # Times out
"bench_core.CountNonzero.time_count_nonzero(3, 1000000, <class 'str'>)", # Times out
"bench_core.CountNonzero.time_count_nonzero_axis(2, 1000000, <class 'str'>)", # Times out
"bench_core.CountNonzero.time_count_nonzero_axis(3, 1000000, <class 'str'>)", # Times out
"bench_core.CountNonzero.time_count_nonzero_multi_axis(2, 1000000, <class 'str'>)", # Times out
"bench_core.CountNonzero.time_count_nonzero_multi_axis(3, 1000000, <class 'str'>)", # Times out
"bench_linalg.LinalgSmallArrays.time_det_small_array", # TODO fails with numpy.linalg.LinAlgError
]
DEFAULT_PANDAS_BENCHMARKS = [
"reshape",
"replace"
]
SKIPPED_PANDAS_BENCHMARKS = [
"replace.ReplaceDict.time_replace_series", # Times out
"replace.ReplaceList.time_replace_list", # OOM, WIP msimacek
"replace.ReplaceList.time_replace_list_one_match", # OOM, WIP msimacek
"reshape.Crosstab.time_crosstab_normalize_margins", # Times out
"reshape.Cut.peakmem_cut_interval", # Times out
"reshape.Cut.time_cut_interval", # Times out
"reshape.GetDummies.time_get_dummies_1d_sparse", # Times out
"reshape.PivotTable.time_pivot_table_margins", # Times out
"reshape.WideToLong.time_wide_to_long_big", # Times out
"reshape.Cut.time_qcut_datetime", # Transient failure GR-61245, exit code -11
"reshape.Explode.time_explode", # Transient failure GR-61245, exit code -11
]
DEFAULT_PYPERFORMANCE_BENCHMARKS = [
# "2to3",
# "chameleon",
"chaos",
# "crypto_pyaes",
# "django_template",
# "dulwich_log",
"fannkuch",
"float",
"go",
"hexiom",
# "html5lib",
"json_dumps",
"json_loads",
# "logging",
# "mako",
"meteor_contest",
"nbody",
"nqueens",
"pathlib",
"pickle",
"pickle_dict",
"pickle_list",
"pickle_pure_python",
"pidigits",
"pyflate",
"regex_compile",
"regex_dna",
"regex_effbot",
"regex_v8",
"richards",
"scimark",
"spectral_norm",
# "sqlalchemy_declarative",
# "sqlalchemy_imperative",
# "sqlite_synth",
# "sympy",
"telco",
# "tornado_http",
"unpack_sequence",
"unpickle",
"unpickle_list",
# "unpickle_pure_python",
# "xml_etree",
]
DEFAULT_PYPY_BENCHMARKS = [
"ai",
# "bm_chameleon",
# "bm_dulwich_log",
"bm_mako",
"bm_mdp",
"chaos",
# "cpython_doc",
"crypto_pyaes",
"deltablue",
"django",
"eparse",
"fannkuch",
"float",
"genshi_text",
"genshi_xml",
"go",
"hexiom2",
"html5lib",
"json_bench",
"meteor-contest",
"nbody_modified",
"nqueens",
"pidigits",
"pyflate-fast",
"pyxl_bench",
"raytrace-simple",
"richards",
"scimark_fft",
"scimark_lu",
"scimark_montecarlo",
"scimark_sor",
"scimark_sparsematmult",
"spectral-norm",
"spitfire2",
"spitfire_cstringio2",
# "sqlalchemy_declarative",
# "sqlalchemy_imperative",
# "sqlitesynth",
# "sympy_expand",
# "sympy_integrate",
# "sympy_str",
# "sympy_sum",
# "telco",
# "twisted_names",
# "twisted_pb",
# "twisted_tcp",
]
def create_asv_benchmark_selection(benchmarks, skipped=()):
regex = '|'.join(benchmarks)
if not skipped:
return regex
negative_lookaheads = [re.escape(skip) + (r'\b' if not skip.endswith(')') else '') for skip in skipped]
return '^(?!' + '|'.join(negative_lookaheads) + ')(' + regex + ')'
class PyPerfJsonRule(mx_benchmark.Rule):
"""Parses a JSON file produced by PyPerf and creates a measurement result."""
def __init__(self, filenames: str, suiteName: str):
self.filenames = filenames.split(",")
self.suiteName = suiteName
def parse(self, text: str) -> list:
r = []
for filename in self.filenames:
self._parse_file(r, filename)
return r
def _parse_file(self, r: list, filename: str):
with open(self._prepend_working_dir(filename)) as fp:
js = json.load(fp)
benchmarks = js["benchmarks"]
for benchmark in benchmarks:
name = benchmark.get("metadata", js["metadata"])["name"]
unit = benchmark.get("metadata", {}).get("unit") or js["metadata"]["unit"]
unit = {
"second": "s",
"byte": "B",
}[unit]
metric = {
"s": "time",
"B": "max-rss",
}[unit]
for run in benchmark["runs"]:
if values := run.get("values", None):
if metric == "time":
warmups = run.get("warmups", [])
for idx, warmup in enumerate(warmups):
r.append(
{
"bench-suite": self.suiteName,
"benchmark": name,
"metric.name": "warmup",
"metric.unit": unit,
"metric.score-function": "id",
"metric.better": "lower",
"metric.type": "numeric",
"metric.iteration": idx,
"metric.value": warmup[1], # 0 is inner_loop count
}
)
for value in values:
r.append(
{
"bench-suite": self.suiteName,
"benchmark": name,
"metric.name": metric,
"metric.unit": unit,
"metric.score-function": "id",
"metric.better": "lower",
"metric.type": "numeric",
"metric.iteration": 0,
"metric.value": value,
}
)
class AsvJsonRule(mx_benchmark.Rule):
"""Parses a JSON file produced by ASV (airspeed-velocity) and creates a measurement result."""
def __init__(self, filename: str, suiteName: str):
self.filename = filename
self.suiteName = suiteName
def parse(self, text: str) -> list:
import itertools
r = []
with open(self._prepend_working_dir(self.filename)) as fp:
js = json.load(fp)
columns = js["result_columns"] # type: list[str]
peak_idx = columns.index("result")
param_idx = columns.index("params")
try:
samples_idx = columns.index("samples")
except ValueError:
samples_idx = -1
for benchmark, result in js["results"].items():
param_combinations = itertools.product(*result[param_idx])
for run_idx, params in enumerate(param_combinations):
peak_values = result[peak_idx]
if not peak_values:
continue
value = peak_values[run_idx]
if not value or math.isnan(value):
continue
r.append(
{
"bench-suite": self.suiteName,
"benchmark": benchmark,
"metric.name": "time",
"metric.unit": "s",
"metric.score-function": "id",
"metric.better": "lower",
"metric.type": "numeric",
"metric.iteration": 0,
"metric.value": value,
"config.run-flags": " ".join(params),
}
)
# It may be that the samples are missing; so omit this step
if 0 <= samples_idx < len(result):
for iteration, value in enumerate(result[samples_idx][run_idx]):
r.append(
{
"bench-suite": self.suiteName,
"benchmark": benchmark,
"metric.name": "warmup",
"metric.unit": "s",
"metric.score-function": "id",
"metric.better": "lower",
"metric.type": "numeric",
"metric.iteration": iteration,
"metric.value": value,
"config.run-flags": " ".join(params),
}
)
return r
class PyPyJsonRule(mx_benchmark.Rule, mx_benchmark.AveragingBenchmarkMixin):
"""Parses a JSON file produced by the Unladen Swallow or PyPy benchmark harness and creates a measurement result."""
def __init__(self, filename: str, suiteName: str):
self.filename = filename
self.suiteName = suiteName
def parse(self, text: str) -> list:
r = []
with open(self._prepend_working_dir(self.filename)) as fp:
js = json.load(fp)
for result in js["results"]:
name = result[0]
if result[1] == "RawResult":
values = result[2]["base_times"]
elif result[1] == "SimpleComparisonResult":
values = [result[2]["base_time"]]
else:
mx.warn(f"No data found for {name} with {result[1]}")
continue
for iteration, value in enumerate(values):
r.append(
{
"bench-suite": self.suiteName,
"benchmark": name,
"metric.name": "warmup",
"metric.unit": "s",
"metric.score-function": "id",
"metric.better": "lower",
"metric.type": "numeric",
"metric.iteration": iteration,
"metric.value": value,
}
)
self.addAverageAcrossLatestResults(r)
return r
class WildcardList:
"""It is not easy to track for external suites which benchmarks are
available, so we just return a wildcard list and assume the caller knows
what they want to run"""
def __init__(self, benchmarks=None):
self.benchmarks = benchmarks
def __contains__(self, x):
return True
def __iter__(self):
if not self.benchmarks:
mx.abort(
"Cannot iterate over benchmark names in foreign benchmark suites. "
+ "Leave off the benchmark name part to run all, or name the benchmarks yourself."
)
else:
return iter(self.benchmarks)
class PySuite(mx_benchmark.TemporaryWorkdirMixin, mx_benchmark.VmBenchmarkSuite):
pass
class PyPerformanceSuite(PySuite):
VERSION = "1.0.6"
def name(self):
return "pyperformance-suite"
def group(self):
return "Graal"
def subgroup(self):
return "graalpython"
def benchmarkList(self, bmSuiteArgs):
return WildcardList(DEFAULT_PYPERFORMANCE_BENCHMARKS)
def rules(self, output, benchmarks, bmSuiteArgs):
return [PyPerfJsonRule(output, self.name())]
def createVmCommandLineArgs(self, benchmarks, runArgs):
return []
def get_vm_registry(self):
return python_vm_registry
def _vmRun(self, vm, workdir, command, benchmarks, bmSuiteArgs):
workdir = abspath(workdir)
vm_venv = f"{self.name()}-{vm.name()}-{vm.config_name()}"
_, _, vm_dims = vm.run(workdir, ["--version"])
if not hasattr(self, "prepared"):
self.prepared = True
vm.run(workdir, ["-m", "venv", join(workdir, vm_venv)])
mx.run(
[
join(vm_venv, "bin", "pip"),
"install",
f"pyperformance=={self.VERSION}",
],
cwd=workdir,
)
if benchmarks:
bms = ["-b", ",".join(benchmarks)]
else:
bms = ["-b", ",".join(DEFAULT_PYPERFORMANCE_BENCHMARKS)]
json_file = f"{vm_venv}.json"
retcode = mx.run(
[
join(vm_venv, "bin", "pyperformance"),
"run",
"--inherit-environ",
"PIP_INDEX_URL,PIP_EXTRA_INDEX_URL,PIP_TRUSTED_HOST,PIP_TIMEOUT,PIP_RETRIES,LD_LIBRARY_PATH,LIBRARY_PATH,CPATH,PATH,PYPY_GC_MAX,JAVA_OPTS,GRAAL_PYTHON_ARGS,GRAAL_PYTHON_VM_ARGS",
"-o",
json_file,
*bms,
],
cwd=workdir,
nonZeroIsFatal=False,
)
mx.log(f"Return code of benchmark harness: {retcode}")
# run again in single shot mode for memory measurements
json_file_memory = f"{vm_venv}_memory.json"
retcode = mx.run(
[
join(vm_venv, "bin", "pyperformance"),
"run",
"--debug-single-value",
"--track-memory",
"--inherit-environ",
"PIP_INDEX_URL,PIP_EXTRA_INDEX_URL,PIP_TRUSTED_HOST,PIP_TIMEOUT,PIP_RETRIES,LD_LIBRARY_PATH,LIBRARY_PATH,CPATH,PATH,PYPY_GC_MAX,JAVA_OPTS,GRAAL_PYTHON_ARGS,GRAAL_PYTHON_VM_ARGS",
"-o",
json_file_memory,
*bms,
],
cwd=workdir,
nonZeroIsFatal=False,
)
mx.log(f"Return code of benchmark harness: {retcode}")
shutil.copy(join(workdir, json_file), join(SUITE.dir, "raw_results.json"))
shutil.copy(join(workdir, json_file_memory), join(SUITE.dir, "raw_results_memory.json"))
return retcode, ",".join([join(workdir, json_file), join(workdir, json_file_memory)]), vm_dims
class PyPySuite(PySuite):
VERSION = "0324a252cf1a"
def name(self):
return "pypy-suite"
def group(self):
return "Graal"
def subgroup(self):
return "graalpython"
def benchmarkList(self, bmSuiteArgs):
return WildcardList(DEFAULT_PYPY_BENCHMARKS)
def rules(self, output, benchmarks, bmSuiteArgs):
return [PyPyJsonRule(output, self.name())]
def createVmCommandLineArgs(self, benchmarks, runArgs):
return []
def get_vm_registry(self):
return python_vm_registry
def _vmRun(self, vm, workdir, command, benchmarks, bmSuiteArgs):
workdir = abspath(workdir)
vm_venv = f"{self.name()}-{vm.name()}-{vm.config_name()}"
_, _, vm_dims = vm.run(workdir, ["--version"])
if not hasattr(self, "prepared"):
self.prepared = True
if artifact := os.environ.get("PYPY_BENCHMARKS_DIR"):
shutil.copytree(artifact, join(workdir, "benchmarks"))
else:
mx.warn("PYPY_BENCHMARKS_DIR is not set, cloning repository")
mx.run(
["hg", "clone", "https://foss.heptapod.net/pypy/benchmarks"],
cwd=workdir,
)
mx.run(
["hg", "up", "-C", self.VERSION], cwd=join(workdir, "benchmarks")
)
# workaround for pypy's benchmarks script issues
with open(join(workdir, "benchmarks", "nullpython.py")) as f:
content = f.read()
content = content.replace("/usr/bin/python", "/usr/bin/env python")
with open(join(workdir, "benchmarks", "nullpython.py"), "w") as f:
f.write(content)
with open(join(workdir, "benchmarks", "benchmarks.py")) as f:
content = f.read()
content = content.replace(
'float(line.split(b" ")[0])', "float(line.split()[0])"
)
with open(join(workdir, "benchmarks", "benchmarks.py"), "w") as f:
f.write(content)
vm.run(workdir, ["-m", "venv", join(workdir, vm_venv)])
json_file = f"{vm_venv}.json"
if benchmarks:
bms = ["-b", ",".join(benchmarks)]
else:
bms = ["-b", ",".join(DEFAULT_PYPY_BENCHMARKS)]
retcode = mx.run(
[
sys.executable,
join(workdir, "benchmarks", "run_local.py"),
f"{vm_venv}/bin/python",
"-o",
join(workdir, json_file),
*bms,
],
cwd=workdir,
nonZeroIsFatal=False,
)
shutil.copy(join(workdir, json_file), join(SUITE.dir, "raw_results.json"))
mx.log(f"Return code of benchmark harness: {retcode}")
return retcode, join(workdir, json_file), vm_dims
class NumPySuite(PySuite):
VERSION = "1.26.4"
BENCHMARK_REQ = [
"asv==0.5.1",
"setuptools==70.3.0",
"distlib==0.3.6",
"filelock==3.8.0",
"platformdirs==2.5.2",
"six==1.16.0",
"virtualenv==20.16.3",
"packaging==24.0",
f"numpy=={VERSION}",
]
def name(self):
return "numpy-suite"
def group(self):
return "Graal"
def subgroup(self):
return "graalpython"
def benchmarkList(self, bmSuiteArgs):
return WildcardList(DEFAULT_NUMPY_BENCHMARKS)
def rules(self, output, benchmarks, bmSuiteArgs):
return [AsvJsonRule(output, self.name())]
def createVmCommandLineArgs(self, benchmarks, runArgs):
return []
def get_vm_registry(self):
return python_vm_registry
def _vmRun(self, vm, workdir, command, benchmarks, bmSuiteArgs):
workdir = abspath(workdir)
benchdir = join(workdir, "numpy", "benchmarks")
vm_venv = f"{self.name()}-{vm.name()}-{vm.config_name()}"
_, _, vm_dims = vm.run(workdir, ["--version"])
if not hasattr(self, "prepared"):
self.prepared = True
npdir = join(workdir, "numpy")
if artifact := os.environ.get("NUMPY_BENCHMARKS_DIR"):
shutil.copytree(artifact, npdir)
else:
mx.warn("NUMPY_BENCHMARKS_DIR is not set, cloning numpy repository")
mx.run(
[
"git",
"clone",
"--depth",
"1",
"https://github.com/numpy/numpy.git",
"--branch",
f"v{self.VERSION}",
"--single-branch",
],
cwd=workdir,
)
shutil.rmtree(join(npdir, ".git"))
mx.run(["git", "init", "."], cwd=npdir)
mx.run(["git", "config", "user.email", "you@example.com"], cwd=npdir)
mx.run(["git", "config", "user.name", "YourName"], cwd=npdir)
mx.run(["git", "commit", "--allow-empty", "-m", "init"], cwd=npdir)
mx.run(["git", "branch", f"v{self.VERSION}"], cwd=npdir)
mx.run(["git", "branch", "main"], cwd=npdir, nonZeroIsFatal=False)
mx.run(["git", "branch", "master"], cwd=npdir, nonZeroIsFatal=False)
vm.run(workdir, ["-m", "venv", join(workdir, vm_venv)])
pip = join(workdir, vm_venv, "bin", "pip")
mx.run([pip, "install", *self.BENCHMARK_REQ], cwd=workdir)
mx.run(
[join(workdir, vm_venv, "bin", "asv"), "machine", "--yes"], cwd=benchdir
)
if not benchmarks:
benchmarks = DEFAULT_NUMPY_BENCHMARKS
retcode = mx.run(
[
join(workdir, vm_venv, "bin", "asv"),
"run",
"--strict",
"--record-samples",
"-e",
"--python=same",
"--set-commit-hash",
f"v{self.VERSION}",
"-b", create_asv_benchmark_selection(benchmarks, skipped=SKIPPED_NUMPY_BENCHMARKS),
],
cwd=benchdir,
nonZeroIsFatal=False,
)
json_file = glob.glob(join(benchdir, "results", "*", "*numpy*.json"))
mx.log(f"Return code of benchmark harness: {retcode}")
if json_file:
json_file = json_file[0]
shutil.copy(json_file, join(SUITE.dir, "raw_results.json"))
return retcode, json_file, vm_dims
else:
return -1, "", vm_dims
class PandasSuite(PySuite):
VERSION = "1.5.2"
VERSION_TAG = "v" + VERSION
BENCHMARK_REQ = [
"asv==0.5.1",
"setuptools==70.3.0",
"distlib==0.3.6",
"filelock==3.8.0",
"platformdirs==2.5.2",
"six==1.16.0",
"virtualenv==20.16.3",
"jinja2",
f"numpy=={NumPySuite.VERSION}",
f"pandas=={VERSION}",
]
def name(self):
return "pandas-suite"
def group(self):
return "Graal"
def subgroup(self):
return "graalpython"
def benchmarkList(self, bmSuiteArgs):
return WildcardList([
"reshape",
"replace",
])
def rules(self, output, benchmarks, bmSuiteArgs):
return [AsvJsonRule(output, self.name())]
def createVmCommandLineArgs(self, benchmarks, runArgs):
return []
def get_vm_registry(self):
return python_vm_registry
def _vmRun(self, vm, workdir, command, benchmarks, bmSuiteArgs):
workdir = abspath(workdir)
benchdir = join(workdir, "pandas", "asv_bench")
vm_venv = f"{self.name()}-{vm.name()}-{vm.config_name()}"
_, _, vm_dims = vm.run(workdir, ["--version"])
if not hasattr(self, "prepared"):
self.prepared = True
npdir = join(workdir, "pandas")
if artifact := os.environ.get("PANDAS_BENCHMARKS_DIR"):
shutil.copytree(artifact, npdir)
else:
mx.warn("PANDAS_BENCHMARKS_DIR is not set, cloning pandas repository")
repo_url = os.environ.get("PANDAS_REPO_URL", "https://github.com/pandas-dev/pandas.git")
mx.log("Cloning Pandas from " + repo_url)
mx.run(
[
"git",
"clone",
"--depth",
"1",
repo_url,
"--branch",
self.VERSION_TAG,
"--single-branch",
],
cwd=workdir,
)
shutil.rmtree(join(npdir, ".git"))
pandas_benchmarks_dir = join(npdir, "asv_bench", "benchmarks")
accepted = ["__init__", "pandas_vb_common"] + list(self.benchmarkList([]))
removed_files = []
for f in os.listdir(pandas_benchmarks_dir):
# Remove any file or directory that is not a benchmark suite we want to run.
# Keep all files starting with "_"
if os.path.splitext(f)[0] not in accepted:
removed_files.append(f)
f_path = join(pandas_benchmarks_dir, f)
if os.path.isdir(f_path):
shutil.rmtree(f_path)
else:
os.remove(f_path)
mx.log("Removed Pandas benchmark files: " + repr(removed_files))
mx.run(["git", "init", "."], cwd=npdir)
mx.run(["git", "config", "user.email", "you@example.com"], cwd=npdir)
mx.run(["git", "config", "user.name", "YourName"], cwd=npdir)
mx.run(["git", "commit", "--allow-empty", "-m", "init"], cwd=npdir)
mx.run(["git", "branch", self.VERSION_TAG], cwd=npdir)
mx.run(["git", "branch", "main"], cwd=npdir, nonZeroIsFatal=False)
mx.run(["git", "branch", "master"], cwd=npdir, nonZeroIsFatal=False)
vm.run(workdir, ["-m", "venv", join(workdir, vm_venv)])
pip = join(workdir, vm_venv, "bin", "pip")
mx.run([pip, "install", *self.BENCHMARK_REQ], cwd=workdir)
mx.run(
[join(workdir, vm_venv, "bin", "asv"), "machine", "--yes"], cwd=benchdir
)
if not benchmarks:
benchmarks = DEFAULT_PANDAS_BENCHMARKS
retcode = mx.run(
[
join(workdir, vm_venv, "bin", "asv"),
"run",
"--strict",
"--record-samples",
"-e",
"--python=same",
"--set-commit-hash",
self.VERSION_TAG,
"-b", create_asv_benchmark_selection(benchmarks, skipped=SKIPPED_PANDAS_BENCHMARKS),
],
cwd=benchdir,
nonZeroIsFatal=False,
)
json_file = glob.glob(join(benchdir, "results", "*", "*pandas*.json"))
mx.log(f"Return code of benchmark harness: {retcode}")
if json_file:
json_file = json_file[0]
shutil.copy(json_file, join(SUITE.dir, "raw_results.json"))
return retcode, json_file, vm_dims
else:
return -1, "", vm_dims
def register_python_benchmarks():
global python_vm_registry, SUITE
from mx_graalpython_benchmark import python_vm_registry as vm_registry
python_vm_registry = vm_registry
SUITE = mx.suite("graalpython")
mx_benchmark.add_bm_suite(PyPerformanceSuite())
mx_benchmark.add_bm_suite(PyPySuite())
mx_benchmark.add_bm_suite(NumPySuite())
mx_benchmark.add_bm_suite(PandasSuite())