xref: /OK3568_Linux_fs/yocto/poky/meta/lib/oeqa/selftest/cases/reproducible.py (revision 4882a59341e53eb6f0b4789bf948001014eff981)
1#
2# SPDX-License-Identifier: MIT
3#
4# Copyright 2019-2020 by Garmin Ltd. or its subsidiaries
5
6from oeqa.selftest.case import OESelftestTestCase
7from oeqa.utils.commands import runCmd, bitbake, get_bb_var, get_bb_vars
8import bb.utils
9import functools
10import multiprocessing
11import textwrap
12import tempfile
13import shutil
14import stat
15import os
16import datetime
17
18exclude_packages = [
19	]
20
21def is_excluded(package):
22    package_name = os.path.basename(package)
23    for i in exclude_packages:
24        if package_name.startswith(i):
25            return i
26    return None
27
28MISSING = 'MISSING'
29DIFFERENT = 'DIFFERENT'
30SAME = 'SAME'
31
32@functools.total_ordering
33class CompareResult(object):
34    def __init__(self):
35        self.reference = None
36        self.test = None
37        self.status = 'UNKNOWN'
38
39    def __eq__(self, other):
40        return (self.status, self.test) == (other.status, other.test)
41
42    def __lt__(self, other):
43        return (self.status, self.test) < (other.status, other.test)
44
45class PackageCompareResults(object):
46    def __init__(self):
47        self.total = []
48        self.missing = []
49        self.different = []
50        self.different_excluded = []
51        self.same = []
52        self.active_exclusions = set()
53
54    def add_result(self, r):
55        self.total.append(r)
56        if r.status == MISSING:
57            self.missing.append(r)
58        elif r.status == DIFFERENT:
59            exclusion = is_excluded(r.reference)
60            if exclusion:
61                self.different_excluded.append(r)
62                self.active_exclusions.add(exclusion)
63            else:
64                self.different.append(r)
65        else:
66            self.same.append(r)
67
68    def sort(self):
69        self.total.sort()
70        self.missing.sort()
71        self.different.sort()
72        self.different_excluded.sort()
73        self.same.sort()
74
75    def __str__(self):
76        return 'same=%i different=%i different_excluded=%i missing=%i total=%i\nunused_exclusions=%s' % (len(self.same), len(self.different), len(self.different_excluded), len(self.missing), len(self.total), self.unused_exclusions())
77
78    def unused_exclusions(self):
79        return sorted(set(exclude_packages) - self.active_exclusions)
80
81def compare_file(reference, test, diffutils_sysroot):
82    result = CompareResult()
83    result.reference = reference
84    result.test = test
85
86    if not os.path.exists(reference):
87        result.status = MISSING
88        return result
89
90    r = runCmd(['cmp', '--quiet', reference, test], native_sysroot=diffutils_sysroot, ignore_status=True, sync=False)
91
92    if r.status:
93        result.status = DIFFERENT
94        return result
95
96    result.status = SAME
97    return result
98
99def run_diffoscope(a_dir, b_dir, html_dir, max_report_size=0, **kwargs):
100    return runCmd(['diffoscope', '--no-default-limits', '--max-report-size', str(max_report_size),
101                   '--exclude-directory-metadata', 'yes', '--html-dir', html_dir, a_dir, b_dir],
102                **kwargs)
103
104class DiffoscopeTests(OESelftestTestCase):
105    diffoscope_test_files = os.path.join(os.path.dirname(os.path.abspath(__file__)), "diffoscope")
106
107    def test_diffoscope(self):
108        bitbake("diffoscope-native -c addto_recipe_sysroot")
109        diffoscope_sysroot = get_bb_var("RECIPE_SYSROOT_NATIVE", "diffoscope-native")
110
111        # Check that diffoscope doesn't return an error when the files compare
112        # the same (a general check that diffoscope is working)
113        with tempfile.TemporaryDirectory() as tmpdir:
114            run_diffoscope('A', 'A', tmpdir,
115                native_sysroot=diffoscope_sysroot, cwd=self.diffoscope_test_files)
116
117        # Check that diffoscope generates an index.html file when the files are
118        # different
119        with tempfile.TemporaryDirectory() as tmpdir:
120            r = run_diffoscope('A', 'B', tmpdir,
121                native_sysroot=diffoscope_sysroot, ignore_status=True, cwd=self.diffoscope_test_files)
122
123            self.assertNotEqual(r.status, 0, msg="diffoscope was successful when an error was expected")
124            self.assertTrue(os.path.exists(os.path.join(tmpdir, 'index.html')), "HTML index not found!")
125
126class ReproducibleTests(OESelftestTestCase):
127    # Test the reproducibility of whatever is built between sstate_targets and targets
128
129    package_classes = ['deb', 'ipk', 'rpm']
130
131    # Maximum report size, in bytes
132    max_report_size = 250 * 1024 * 1024
133
134    # targets are the things we want to test the reproducibility of
135    targets = ['core-image-minimal', 'core-image-sato', 'core-image-full-cmdline', 'core-image-weston', 'world']
136    # sstate targets are things to pull from sstate to potentially cut build/debugging time
137    sstate_targets = []
138    save_results = False
139    if 'OEQA_DEBUGGING_SAVED_OUTPUT' in os.environ:
140        save_results = os.environ['OEQA_DEBUGGING_SAVED_OUTPUT']
141
142    # This variable controls if one of the test builds is allowed to pull from
143    # an sstate cache/mirror. The other build is always done clean as a point of
144    # comparison.
145    # If you know that your sstate archives are reproducible, enabling this
146    # will test that and also make the test run faster. If your sstate is not
147    # reproducible, disable this in your derived test class
148    build_from_sstate = True
149
150    def setUpLocal(self):
151        super().setUpLocal()
152        needed_vars = ['TOPDIR', 'TARGET_PREFIX', 'BB_NUMBER_THREADS']
153        bb_vars = get_bb_vars(needed_vars)
154        for v in needed_vars:
155            setattr(self, v.lower(), bb_vars[v])
156
157        self.extraresults = {}
158        self.extraresults.setdefault('reproducible.rawlogs', {})['log'] = ''
159        self.extraresults.setdefault('reproducible', {}).setdefault('files', {})
160
161    def append_to_log(self, msg):
162        self.extraresults['reproducible.rawlogs']['log'] += msg
163
164    def compare_packages(self, reference_dir, test_dir, diffutils_sysroot):
165        result = PackageCompareResults()
166
167        old_cwd = os.getcwd()
168        try:
169            file_result = {}
170            os.chdir(test_dir)
171            with multiprocessing.Pool(processes=int(self.bb_number_threads or 0)) as p:
172                for root, dirs, files in os.walk('.'):
173                    async_result = []
174                    for f in files:
175                        reference_path = os.path.join(reference_dir, root, f)
176                        test_path = os.path.join(test_dir, root, f)
177                        async_result.append(p.apply_async(compare_file, (reference_path, test_path, diffutils_sysroot)))
178
179                    for a in async_result:
180                        result.add_result(a.get())
181
182        finally:
183            os.chdir(old_cwd)
184
185        result.sort()
186        return result
187
188    def write_package_list(self, package_class, name, packages):
189        self.extraresults['reproducible']['files'].setdefault(package_class, {})[name] = [
190                {'reference': p.reference, 'test': p.test} for p in packages]
191
192    def copy_file(self, source, dest):
193        bb.utils.mkdirhier(os.path.dirname(dest))
194        shutil.copyfile(source, dest)
195
196    def do_test_build(self, name, use_sstate):
197        capture_vars = ['DEPLOY_DIR_' + c.upper() for c in self.package_classes]
198
199        tmpdir = os.path.join(self.topdir, name, 'tmp')
200        if os.path.exists(tmpdir):
201            bb.utils.remove(tmpdir, recurse=True)
202
203        config = textwrap.dedent('''\
204            PACKAGE_CLASSES = "{package_classes}"
205            INHIBIT_PACKAGE_STRIP = "1"
206            TMPDIR = "{tmpdir}"
207            LICENSE_FLAGS_ACCEPTED = "commercial"
208            DISTRO_FEATURES:append = ' systemd pam'
209            USERADDEXTENSION = "useradd-staticids"
210            USERADD_ERROR_DYNAMIC = "skip"
211            USERADD_UID_TABLES += "files/static-passwd"
212            USERADD_GID_TABLES += "files/static-group"
213            ''').format(package_classes=' '.join('package_%s' % c for c in self.package_classes),
214                        tmpdir=tmpdir)
215
216        if not use_sstate:
217            if self.sstate_targets:
218               self.logger.info("Building prebuild for %s (sstate allowed)..." % (name))
219               self.write_config(config)
220               bitbake(' '.join(self.sstate_targets))
221
222            # This config fragment will disable using shared and the sstate
223            # mirror, forcing a complete build from scratch
224            config += textwrap.dedent('''\
225                SSTATE_DIR = "${TMPDIR}/sstate"
226                SSTATE_MIRRORS = ""
227                ''')
228
229        self.logger.info("Building %s (sstate%s allowed)..." % (name, '' if use_sstate else ' NOT'))
230        self.write_config(config)
231        d = get_bb_vars(capture_vars)
232        # targets used to be called images
233        bitbake(' '.join(getattr(self, 'images', self.targets)))
234        return d
235
236    def test_reproducible_builds(self):
237        def strip_topdir(s):
238            if s.startswith(self.topdir):
239                return s[len(self.topdir):]
240            return s
241
242        # Build native utilities
243        self.write_config('')
244        bitbake("diffoscope-native diffutils-native jquery-native -c addto_recipe_sysroot")
245        diffutils_sysroot = get_bb_var("RECIPE_SYSROOT_NATIVE", "diffutils-native")
246        diffoscope_sysroot = get_bb_var("RECIPE_SYSROOT_NATIVE", "diffoscope-native")
247        jquery_sysroot = get_bb_var("RECIPE_SYSROOT_NATIVE", "jquery-native")
248
249        if self.save_results:
250            os.makedirs(self.save_results, exist_ok=True)
251            datestr = datetime.datetime.now().strftime('%Y%m%d')
252            save_dir = tempfile.mkdtemp(prefix='oe-reproducible-%s-' % datestr, dir=self.save_results)
253            os.chmod(save_dir, stat.S_IRWXU | stat.S_IRGRP | stat.S_IXGRP | stat.S_IROTH | stat.S_IXOTH)
254            self.logger.info('Non-reproducible packages will be copied to %s', save_dir)
255
256        vars_A = self.do_test_build('reproducibleA', self.build_from_sstate)
257
258        vars_B = self.do_test_build('reproducibleB', False)
259
260        # NOTE: The temp directories from the reproducible build are purposely
261        # kept after the build so it can be diffed for debugging.
262
263        fails = []
264
265        for c in self.package_classes:
266            with self.subTest(package_class=c):
267                package_class = 'package_' + c
268
269                deploy_A = vars_A['DEPLOY_DIR_' + c.upper()]
270                deploy_B = vars_B['DEPLOY_DIR_' + c.upper()]
271
272                self.logger.info('Checking %s packages for differences...' % c)
273                result = self.compare_packages(deploy_A, deploy_B, diffutils_sysroot)
274
275                self.logger.info('Reproducibility summary for %s: %s' % (c, result))
276
277                self.append_to_log('\n'.join("%s: %s" % (r.status, r.test) for r in result.total))
278
279                self.write_package_list(package_class, 'missing', result.missing)
280                self.write_package_list(package_class, 'different', result.different)
281                self.write_package_list(package_class, 'different_excluded', result.different_excluded)
282                self.write_package_list(package_class, 'same', result.same)
283
284                if self.save_results:
285                    for d in result.different:
286                        self.copy_file(d.reference, '/'.join([save_dir, 'packages', strip_topdir(d.reference)]))
287                        self.copy_file(d.test, '/'.join([save_dir, 'packages', strip_topdir(d.test)]))
288
289                    for d in result.different_excluded:
290                        self.copy_file(d.reference, '/'.join([save_dir, 'packages-excluded', strip_topdir(d.reference)]))
291                        self.copy_file(d.test, '/'.join([save_dir, 'packages-excluded', strip_topdir(d.test)]))
292
293                if result.missing or result.different:
294                    fails.append("The following %s packages are missing or different and not in exclusion list: %s" %
295                            (c, '\n'.join(r.test for r in (result.missing + result.different))))
296
297        # Clean up empty directories
298        if self.save_results:
299            if not os.listdir(save_dir):
300                os.rmdir(save_dir)
301            else:
302                self.logger.info('Running diffoscope')
303                package_dir = os.path.join(save_dir, 'packages')
304                package_html_dir = os.path.join(package_dir, 'diff-html')
305
306                # Copy jquery to improve the diffoscope output usability
307                self.copy_file(os.path.join(jquery_sysroot, 'usr/share/javascript/jquery/jquery.min.js'), os.path.join(package_html_dir, 'jquery.js'))
308
309                run_diffoscope('reproducibleA', 'reproducibleB', package_html_dir, max_report_size=self.max_report_size,
310                        native_sysroot=diffoscope_sysroot, ignore_status=True, cwd=package_dir)
311
312        if fails:
313            self.fail('\n'.join(fails))
314
315