forked from spdk/spdk
-
Notifications
You must be signed in to change notification settings - Fork 0
/
autorun_post.py
executable file
·223 lines (174 loc) · 8.42 KB
/
autorun_post.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
#!/usr/bin/python3
# SPDX-License-Identifier: BSD-3-Clause
# Copyright (C) 2017 Intel Corporation.
# All rights reserved.
import shutil
import subprocess
import argparse
import itertools
import os
import sys
import glob
import re
import pandas as pd
def generateTestCompletionTableByTest(output_dir, data_table):
columns_to_group = ['Domain', 'Test', 'Agent']
total_tests_number = len(data_table.groupby('Test'))
has_agent = data_table['Agent'] != 'None'
data_table_with_agent = data_table[has_agent]
executed_tests = len(data_table_with_agent.groupby('Test'))
tests_executions = len(data_table_with_agent.groupby(columns_to_group))
pivot_by_test = pd.pivot_table(data_table, index=columns_to_group, aggfunc=any)
output_file = os.path.join(output_dir, 'post_process', 'completions_table_by_test.html')
with open(output_file, 'w') as f:
table_row = '<tr><td>{}</td><td>{}</td>\n'
f.write('<table>\n')
f.write(table_row.format('Total number of tests', total_tests_number))
f.write(table_row.format('Tests executed', executed_tests))
f.write(table_row.format('Number of test executions', tests_executions))
f.write('</table>\n')
f.write(pivot_by_test.to_html(None))
def generateTestCompletionTables(output_dir, completion_table):
data_table = pd.DataFrame(completion_table, columns=["Agent", "Domain", "Test", "With Asan", "With UBsan"])
data_table.to_html(os.path.join(output_dir, 'completions_table.html'))
os.makedirs(os.path.join(output_dir, "post_process"), exist_ok=True)
pivot_by_agent = pd.pivot_table(data_table, index=["Agent", "Domain", "Test"], aggfunc=any)
pivot_by_agent.to_html(os.path.join(output_dir, "post_process", 'completions_table_by_agent.html'))
generateTestCompletionTableByTest(output_dir, data_table)
pivot_by_asan = pd.pivot_table(data_table, index=["Domain", "Test"], values=["With Asan"], aggfunc=any)
pivot_by_asan.to_html(os.path.join(output_dir, "post_process", 'completions_table_by_asan.html'))
pivot_by_ubsan = pd.pivot_table(data_table, index=["Domain", "Test"], values=["With UBsan"], aggfunc=any)
pivot_by_ubsan.to_html(os.path.join(output_dir, "post_process", 'completions_table_by_ubsan.html'))
def generateCoverageReport(output_dir, repo_dir):
coveragePath = os.path.join(output_dir, '**', 'cov_total.info')
covfiles = [os.path.abspath(p) for p in glob.glob(coveragePath, recursive=True)]
for f in covfiles:
print(f)
if len(covfiles) == 0:
return
lcov_opts = [
'--rc', 'lcov_branch_coverage=1',
'--rc', 'lcov_function_coverage=1',
'--rc', 'genhtml_branch_coverage=1',
'--rc', 'genhtml_function_coverage=1',
'--rc', 'genhtml_legend=1',
'--rc', 'geninfo_all_blocks=1',
]
# HACK: This is a workaround for some odd CI assumptions
details = '--show-details'
cov_total = os.path.abspath(os.path.join(output_dir, 'cov_total.info'))
coverage = os.path.join(output_dir, 'coverage')
lcov = ['lcov', *lcov_opts, '-q', *itertools.chain(*[('-a', f) for f in covfiles]), '-o', cov_total]
genhtml = ['genhtml', *lcov_opts, '-q', cov_total, '--legend', '-t', 'Combined', *details.split(), '-o', coverage]
try:
subprocess.check_call(lcov)
except subprocess.CalledProcessError as e:
print("lcov failed")
print(e)
return
with open(cov_total, 'r') as cov_total_file:
file_contents = cov_total_file.readlines()
replacement = "SF:" + repo_dir
os.remove(cov_total)
with open(cov_total, 'w+') as file:
for Line in file_contents:
Line = re.sub("^SF:.*/repo", replacement, Line)
file.write(Line + '\n')
try:
subprocess.check_call(genhtml)
except subprocess.CalledProcessError as e:
print("genhtml failed")
print(e)
for f in covfiles:
os.remove(f)
def collectOne(output_dir, dir_name):
dirs = glob.glob(os.path.join(output_dir, '*', dir_name))
dirs.sort()
if len(dirs) == 0:
return
# Collect first instance of dir_name and move it to the top level
collect_dir = dirs.pop(0)
shutil.move(collect_dir, os.path.join(output_dir, dir_name))
# Delete all other instances
for d in dirs:
shutil.rmtree(d)
def getCompletions(completionFile, test_list, test_completion_table):
agent_name = os.path.basename(os.path.dirname(completionFile))
with open(completionFile, 'r') as completionList:
completions = completionList.read()
asan_enabled = "asan" in completions
ubsan_enabled = "ubsan" in completions
for line in completions.splitlines():
try:
domain, test_name = line.strip().split()
test_list[test_name] = (True, asan_enabled | test_list[test_name][1], ubsan_enabled | test_list[test_name][2])
test_completion_table.append([agent_name, domain, test_name, asan_enabled, ubsan_enabled])
try:
test_completion_table.remove(["None", "None", test_name, False, False])
except ValueError:
continue
except KeyError:
continue
def printList(header, test_list, index, condition):
print("\n\n-----%s------" % header)
executed_tests = [x for x in sorted(test_list) if test_list[x][index] is condition]
print(*executed_tests, sep="\n")
def printListInformation(table_type, test_list):
printList("%s Executed in Build" % table_type, test_list, 0, True)
printList("%s Missing From Build" % table_type, test_list, 0, False)
printList("%s Missing ASAN" % table_type, test_list, 1, False)
printList("%s Missing UBSAN" % table_type, test_list, 2, False)
def getSkippedTests(repo_dir):
skipped_test_file = os.path.join(repo_dir, "test", "common", "skipped_tests.txt")
if not os.path.exists(skipped_test_file):
return []
with open(skipped_test_file, "r") as skipped_test_data:
return [x.strip() for x in skipped_test_data.readlines() if "#" not in x and x.strip() != '']
def confirmPerPatchTests(test_list, skiplist):
missing_tests = [x for x in sorted(test_list) if test_list[x][0] is False
and x not in skiplist]
if len(missing_tests) > 0:
print("Not all tests were run. Failing the build.")
print(missing_tests)
sys.exit(1)
def aggregateCompletedTests(output_dir, repo_dir, skip_confirm=False):
test_list = {}
test_completion_table = []
testFiles = glob.glob(os.path.join(output_dir, '**', 'all_tests.txt'), recursive=True)
completionFiles = glob.glob(os.path.join(output_dir, '**', 'test_completions.txt'), recursive=True)
if len(testFiles) == 0:
print("Unable to perform test completion aggregator. No input files.")
return 0
with open(testFiles[0], 'r') as raw_test_list:
for line in raw_test_list:
try:
test_name = line.strip()
except Exception:
print("Failed to parse a test type.")
return 1
test_list[test_name] = (False, False, False)
test_completion_table.append(["None", "None", test_name, False, False])
for completionFile in completionFiles:
getCompletions(completionFile, test_list, test_completion_table)
printListInformation("Tests", test_list)
generateTestCompletionTables(output_dir, test_completion_table)
skipped_tests = getSkippedTests(repo_dir)
if not skip_confirm:
confirmPerPatchTests(test_list, skipped_tests)
return 0
def main(output_dir, repo_dir, skip_confirm=False):
print("-----Begin Post Process Script------")
generateCoverageReport(output_dir, repo_dir)
collectOne(output_dir, 'doc')
collectOne(output_dir, 'ut_coverage')
aggregateCompletedTests(output_dir, repo_dir, skip_confirm)
if __name__ == "__main__":
parser = argparse.ArgumentParser(description="SPDK Coverage Processor")
parser.add_argument("-d", "--directory_location", type=str, required=True,
help="The location of your build's output directory")
parser.add_argument("-r", "--repo_directory", type=str, required=True,
help="The location of your spdk repository")
parser.add_argument("-s", "--skip_confirm", required=False, action="store_true",
help="Do not check if all autotest.sh tests were executed.")
args = parser.parse_args()
main(args.directory_location, args.repo_directory, args.skip_confirm)