forked from micropython/micropython
-
Notifications
You must be signed in to change notification settings - Fork 0
/
run-multitests.py
executable file
·646 lines (550 loc) · 21.1 KB
/
run-multitests.py
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
#!/usr/bin/env python3
# This file is part of the MicroPython project, http://micropython.org/
# The MIT License (MIT)
# Copyright (c) 2020 Damien P. George
#
# run-multitests.py
# Runs a test suite that relies on two micropython instances/devices
# interacting in some way. Typically used to test networking / bluetooth etc.
import sys, os, time, re, select
import argparse
import itertools
import subprocess
import tempfile
test_dir = os.path.abspath(os.path.dirname(__file__))
if os.path.abspath(sys.path[0]) == test_dir:
# remove the micropython/tests dir from path to avoid
# accidentally importing tests like micropython/const.py
sys.path.pop(0)
sys.path.insert(0, test_dir + "/../tools")
import pyboard
if os.name == "nt":
CPYTHON3 = os.getenv("MICROPY_CPYTHON3", "python3.exe")
MICROPYTHON = os.path.abspath(
os.getenv(
"MICROPY_MICROPYTHON", test_dir + "/../ports/windows/build-standard/micropython.exe"
)
)
else:
CPYTHON3 = os.getenv("MICROPY_CPYTHON3", "python3")
MICROPYTHON = os.path.abspath(
os.getenv("MICROPY_MICROPYTHON", test_dir + "/../ports/unix/build-standard/micropython")
)
# For diff'ing test output
DIFF = os.getenv("MICROPY_DIFF", "diff -u")
PYTHON_TRUTH = CPYTHON3
INSTANCE_READ_TIMEOUT_S = 10
APPEND_CODE_TEMPLATE = """
import sys
class multitest:
@staticmethod
def flush():
try:
sys.stdout.flush()
except AttributeError:
pass
@staticmethod
def skip():
print("SKIP")
multitest.flush()
raise SystemExit
@staticmethod
def next():
print("NEXT")
multitest.flush()
@staticmethod
def broadcast(msg):
print("BROADCAST", msg)
multitest.flush()
@staticmethod
def wait(msg):
msg = "BROADCAST " + msg
while True:
if sys.stdin.readline().rstrip() == msg:
return
@staticmethod
def globals(**gs):
for g in gs:
print("SET {{}} = {{!r}}".format(g, gs[g]))
multitest.flush()
@staticmethod
def get_network_ip():
try:
ip = nic.ifconfig()[0]
except:
try:
import network
if hasattr(network, "WLAN"):
ip = network.WLAN().ifconfig()[0]
else:
ip = network.LAN().ifconfig()[0]
except:
ip = HOST_IP
return ip
@staticmethod
def expect_reboot(resume, delay_ms=0):
print("WAIT_FOR_REBOOT", resume, delay_ms)
@staticmethod
def output_metric(data):
print("OUTPUT_METRIC", data)
{}
instance{}()
multitest.flush()
"""
# The btstack implementation on Unix generates some spurious output that we
# can't control. Also other platforms may output certain warnings/errors that
# can be safely ignored.
IGNORE_OUTPUT_MATCHES = (
"libusb: error ", # It tries to open devices that it doesn't have access to (libusb prints unconditionally).
"hci_transport_h2_libusb.c", # Same issue. We enable LOG_ERROR in btstack.
"USB Path: ", # Hardcoded in btstack's libusb transport.
"hci_number_completed_packet", # Warning from btstack.
"lld_pdu_get_tx_flush_nb HCI packet count mismatch (", # From ESP-IDF, see https://github.com/espressif/esp-idf/issues/5105
)
def get_host_ip(_ip_cache=[]):
if not _ip_cache:
try:
import socket
s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
s.connect(("8.8.8.8", 80))
_ip_cache.append(s.getsockname()[0])
s.close()
except:
_ip_cache.append("127.0.0.1")
return _ip_cache[0]
class PyInstance:
def __init__(self):
pass
def close(self):
pass
def prepare_script_from_file(self, filename, prepend, append):
with open(filename, "rb") as f:
script = f.read()
if prepend:
script = bytes(prepend, "ascii") + b"\n" + script
if append:
script += b"\n" + bytes(append, "ascii")
return script
def run_file(self, filename, prepend="", append=""):
return self.run_script(self.prepare_script_from_file(filename, prepend, append))
def start_file(self, filename, prepend="", append=""):
return self.start_script(self.prepare_script_from_file(filename, prepend, append))
class PyInstanceSubProcess(PyInstance):
def __init__(self, argv, env=None):
self.argv = argv
self.cwd = None
self.env = {n: v for n, v in (i.split("=") for i in env)} if env else None
self.popen = None
self.finished = True
def __str__(self):
return self.argv[0].rsplit("/")[-1]
def prepare_script_from_file(self, filename, prepend, append):
# Make tests run in the directory of the test file, and in an isolated environment
# (i.e. `import io` would otherwise get the `tests/io` directory).
self.cwd = os.path.dirname(filename)
remove_cwd_from_sys_path = b"import sys\nsys.path.remove('')\n\n"
return remove_cwd_from_sys_path + super().prepare_script_from_file(
filename, prepend, append
)
def run_script(self, script):
output = b""
err = None
try:
p = subprocess.run(
self.argv,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
input=script,
cwd=self.cwd,
env=self.env,
)
output = p.stdout
except subprocess.CalledProcessError as er:
err = er
return str(output.strip(), "ascii"), err
def start_script(self, script):
self.popen = subprocess.Popen(
self.argv + ["-c", script],
stdin=subprocess.PIPE,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
cwd=self.cwd,
env=self.env,
)
self.finished = False
def stop(self):
if self.popen and self.popen.poll() is None:
self.popen.terminate()
def readline(self):
sel = select.select([self.popen.stdout.raw], [], [], 0.001)
if not sel[0]:
self.finished = self.popen.poll() is not None
return None, None
out = self.popen.stdout.raw.readline()
if out == b"":
self.finished = self.popen.poll() is not None
return None, None
else:
return str(out.rstrip(), "ascii"), None
def write(self, data):
self.popen.stdin.write(data)
self.popen.stdin.flush()
def is_finished(self):
return self.finished
def wait_finished(self):
self.popen.wait()
out = self.popen.stdout.read()
return str(out, "ascii"), ""
class PyInstancePyboard(PyInstance):
@staticmethod
def map_device_shortcut(device):
if device[0] == "a" and device[1:].isdigit():
return "/dev/ttyACM" + device[1:]
elif device[0] == "u" and device[1:].isdigit():
return "/dev/ttyUSB" + device[1:]
else:
return device
def __init__(self, device):
device = self.map_device_shortcut(device)
self.device = device
self.pyb = pyboard.Pyboard(device)
self.pyb.enter_raw_repl()
self.finished = True
def __str__(self):
return self.device.rsplit("/")[-1]
def close(self):
self.pyb.exit_raw_repl()
self.pyb.close()
def run_script(self, script):
output = b""
err = None
try:
self.pyb.enter_raw_repl()
output = self.pyb.exec_(script)
except pyboard.PyboardError as er:
err = er
return str(output.strip(), "ascii"), err
def start_script(self, script):
self.pyb.enter_raw_repl()
self.pyb.exec_raw_no_follow(script)
self.finished = False
def stop(self):
self.pyb.serial.write(b"\r\x03")
def readline(self):
if self.finished:
return None, None
if self.pyb.serial.inWaiting() == 0:
return None, None
out = self.pyb.read_until(1, (b"\r\n", b"\x04"))
if out.endswith(b"\x04"):
self.finished = True
out = out[:-1]
err = str(self.pyb.read_until(1, b"\x04"), "ascii")
err = err[:-1]
if not out and not err:
return None, None
else:
err = None
return str(out.rstrip(), "ascii"), err
def write(self, data):
self.pyb.serial.write(data)
def is_finished(self):
return self.finished
def wait_finished(self):
out, err = self.pyb.follow(10, None)
return str(out, "ascii"), str(err, "ascii")
def prepare_test_file_list(test_files):
test_files2 = []
for test_file in sorted(test_files):
num_instances = 0
with open(test_file) as f:
for line in f:
m = re.match(r"def instance([0-9]+)\(\):", line)
if m:
num_instances = max(num_instances, int(m.group(1)) + 1)
test_files2.append((test_file, num_instances))
return test_files2
def trace_instance_output(instance_idx, line):
if cmd_args.trace_output:
t_ms = round((time.time() - trace_t0) * 1000)
print("{:6} i{} :".format(t_ms, instance_idx), line)
sys.stdout.flush()
def run_test_on_instances(test_file, num_instances, instances):
global trace_t0
trace_t0 = time.time()
error = False
skip = False
injected_globals = ""
output = [[] for _ in range(num_instances)]
output_metrics = []
# If the test calls get_network_ip() then inject HOST_IP so that devices can know
# the IP address of the host. Do this lazily to not require a TCP/IP connection
# on the host if it's not needed.
with open(test_file, "rb") as f:
if b"get_network_ip" in f.read():
injected_globals += "HOST_IP = '" + get_host_ip() + "'\n"
if cmd_args.trace_output:
print("TRACE {}:".format("|".join(str(i) for i in instances)))
# Start all instances running, in order, waiting until they signal they are ready
for idx in range(num_instances):
append_code = APPEND_CODE_TEMPLATE.format(injected_globals, idx)
instance = instances[idx]
instance.start_file(test_file, append=append_code)
last_read_time = time.time()
while True:
if instance.is_finished():
break
out, err = instance.readline()
if out is None and err is None:
if time.time() > last_read_time + INSTANCE_READ_TIMEOUT_S:
output[idx].append("TIMEOUT")
error = True
break
time.sleep(0.1)
continue
last_read_time = time.time()
if out is not None and not any(m in out for m in IGNORE_OUTPUT_MATCHES):
trace_instance_output(idx, out)
if out.startswith("SET "):
injected_globals += out[4:] + "\n"
elif out == "SKIP":
skip = True
break
elif out == "NEXT":
break
else:
output[idx].append(out)
if err is not None:
trace_instance_output(idx, err)
output[idx].append(err)
error = True
if error or skip:
break
if not error and not skip:
# Capture output and wait for all instances to finish running
last_read_time = [time.time() for _ in range(num_instances)]
while True:
num_running = 0
num_output = 0
for idx in range(num_instances):
instance = instances[idx]
if instance.is_finished():
continue
num_running += 1
out, err = instance.readline()
if out is None and err is None:
if time.time() > last_read_time[idx] + INSTANCE_READ_TIMEOUT_S:
output[idx].append("TIMEOUT")
error = True
continue
num_output += 1
last_read_time[idx] = time.time()
if out is not None and not any(m in out for m in IGNORE_OUTPUT_MATCHES):
trace_instance_output(idx, out)
if out.startswith("WAIT_FOR_REBOOT"):
_, resume, delay_ms = out.split(" ")
if wait_for_reboot(instance, delay_ms):
# Restart the test code, resuming from requested instance block
if not resume.startswith("instance{}".format(idx)):
raise SystemExit(
'ERROR: resume function must start with "instance{}"'.format(
idx
)
)
append_code = APPEND_CODE_TEMPLATE.format(injected_globals, resume[8:])
instance.start_file(test_file, append=append_code)
last_read_time[idx] = time.time()
if out.startswith("BROADCAST "):
for instance2 in instances:
if instance2 is not instance:
instance2.write(bytes(out, "ascii") + b"\r\n")
elif out.startswith("OUTPUT_METRIC "):
output_metrics.append(out.split(" ", 1)[1])
else:
output[idx].append(out)
if err is not None:
trace_instance_output(idx, err)
output[idx].append(err)
error = True
if not num_output:
time.sleep(0.1)
if not num_running or error:
break
# Stop all instances
for idx in range(num_instances):
instances[idx].stop()
output_str = ""
for idx, lines in enumerate(output):
output_str += "--- instance{} ---\n".format(idx)
output_str += "\n".join(lines) + "\n"
return error, skip, output_str, output_metrics
def wait_for_reboot(instance, extra_timeout_ms=0):
# Monitor device responses for reboot banner, waiting for idle.
extra_timeout = float(extra_timeout_ms) * 1000
INITIAL_TIMEOUT = 1 + extra_timeout
FULL_TIMEOUT = 5 + extra_timeout
t_start = t_last_activity = time.monotonic()
while True:
t = time.monotonic()
out, err = instance.readline()
if err is not None:
print("Reboot: communication error", err)
return False
if out:
t_last_activity = t
# Check for reboot banner, see py/pyexec.c "reset friendly REPL"
if re.match(r"^MicroPython v\d+\.\d+\.\d+.* on .*; .* with .*$", out):
time.sleep(0.1)
break
if t_last_activity == t_start:
if t - t_start > INITIAL_TIMEOUT:
print("Reboot: missed initial Timeout")
return False
else:
if t - t_start > FULL_TIMEOUT:
print("Reboot: Timeout")
return False
instance.pyb.enter_raw_repl()
return True
def print_diff(a, b):
a_fd, a_path = tempfile.mkstemp(text=True)
b_fd, b_path = tempfile.mkstemp(text=True)
os.write(a_fd, a.encode())
os.write(b_fd, b.encode())
os.close(a_fd)
os.close(b_fd)
subprocess.run(DIFF.split(" ") + [a_path, b_path])
os.unlink(a_path)
os.unlink(b_path)
def run_tests(test_files, instances_truth, instances_test):
skipped_tests = []
passed_tests = []
failed_tests = []
for test_file, num_instances in test_files:
instances_str = "|".join(str(instances_test[i]) for i in range(num_instances))
print("{} on {}: ".format(test_file, instances_str), end="")
if cmd_args.show_output or cmd_args.trace_output:
print()
sys.stdout.flush()
# Run test on test instances
error, skip, output_test, output_metrics = run_test_on_instances(
test_file, num_instances, instances_test
)
if not skip:
# Check if truth exists in a file, and read it in
test_file_expected = test_file + ".exp"
if os.path.isfile(test_file_expected):
with open(test_file_expected) as f:
output_truth = f.read()
else:
# Run test on truth instances to get expected output
_, _, output_truth, _ = run_test_on_instances(
test_file, num_instances, instances_truth
)
if cmd_args.show_output:
print("### TEST ###")
print(output_test, end="")
if not skip:
print("### TRUTH ###")
print(output_truth, end="")
# Print result of test
if skip:
print("skip")
skipped_tests.append(test_file)
elif output_test == output_truth:
print("pass")
passed_tests.append(test_file)
else:
print("FAIL")
failed_tests.append(test_file)
if not cmd_args.show_output:
print("### TEST ###")
print(output_test, end="")
print("### TRUTH ###")
print(output_truth, end="")
print("### DIFF ###")
print_diff(output_truth, output_test)
# Print test output metrics, if there are any.
if output_metrics:
for metric in output_metrics:
print(test_file, ": ", metric, sep="")
if cmd_args.show_output:
print()
print("{} tests performed".format(len(skipped_tests) + len(passed_tests) + len(failed_tests)))
print("{} tests passed".format(len(passed_tests)))
if skipped_tests:
print("{} tests skipped: {}".format(len(skipped_tests), " ".join(skipped_tests)))
if failed_tests:
print("{} tests failed: {}".format(len(failed_tests), " ".join(failed_tests)))
return not failed_tests
def main():
global cmd_args
cmd_parser = argparse.ArgumentParser(
description="Run network tests for MicroPython",
formatter_class=argparse.RawTextHelpFormatter,
)
cmd_parser.add_argument(
"-s", "--show-output", action="store_true", help="show test output after running"
)
cmd_parser.add_argument(
"-t", "--trace-output", action="store_true", help="trace test output while running"
)
cmd_parser.add_argument(
"-i", "--instance", action="append", default=[], help="instance(s) to run the tests on"
)
cmd_parser.add_argument(
"-p",
"--permutations",
type=int,
default=1,
help="repeat the test with this many permutations of the instance order",
)
cmd_parser.epilog = (
"Supported instance types:\r\n"
" -i pyb:<port> physical device (eg. pyboard) on provided repl port.\n"
" -i micropython unix micropython instance, path customised with MICROPY_MICROPYTHON env.\n"
" -i cpython desktop python3 instance, path customised with MICROPY_CPYTHON3 env.\n"
" -i exec:<path> custom program run on provided path.\n"
"Each instance arg can optionally have custom env provided, eg. <cmd>,ENV=VAR,ENV=VAR...\n"
)
cmd_parser.add_argument("files", nargs="+", help="input test files")
cmd_args = cmd_parser.parse_args()
# clear search path to make sure tests use only builtin modules and those in extmod
os.environ["MICROPYPATH"] = os.pathsep.join((".frozen", "../extmod"))
test_files = prepare_test_file_list(cmd_args.files)
max_instances = max(t[1] for t in test_files)
instances_truth = [PyInstanceSubProcess([PYTHON_TRUTH]) for _ in range(max_instances)]
instances_test = []
for i in cmd_args.instance:
# Each instance arg is <cmd>,ENV=VAR,ENV=VAR...
i = i.split(",")
cmd = i[0]
env = i[1:]
if cmd.startswith("exec:"):
instances_test.append(PyInstanceSubProcess([cmd[len("exec:") :]], env))
elif cmd == "micropython":
instances_test.append(PyInstanceSubProcess([MICROPYTHON], env))
elif cmd == "cpython":
instances_test.append(PyInstanceSubProcess([CPYTHON3], env))
elif cmd.startswith("pyb:"):
instances_test.append(PyInstancePyboard(cmd[len("pyb:") :]))
else:
print("unknown instance string: {}".format(cmd), file=sys.stderr)
sys.exit(1)
for _ in range(max_instances - len(instances_test)):
instances_test.append(PyInstanceSubProcess([MICROPYTHON]))
all_pass = True
try:
for i, instances_test_permutation in enumerate(itertools.permutations(instances_test)):
if i >= cmd_args.permutations:
break
all_pass &= run_tests(test_files, instances_truth, instances_test_permutation)
finally:
for i in instances_truth:
i.close()
for i in instances_test:
i.close()
if not all_pass:
sys.exit(1)
if __name__ == "__main__":
main()