xref: /openbmc/openbmc/poky/meta/lib/oe/reproducible.py (revision 56b44a98)
1#
2# Copyright OpenEmbedded Contributors
3#
4# SPDX-License-Identifier: GPL-2.0-only
5#
6import os
7import subprocess
8import bb
9
10# For reproducible builds, this code sets the default SOURCE_DATE_EPOCH in each
11# component's build environment. The format is number of seconds since the
12# system epoch.
13#
14# Upstream components (generally) respect this environment variable,
15# using it in place of the "current" date and time.
16# See https://reproducible-builds.org/specs/source-date-epoch/
17#
18# The default value of SOURCE_DATE_EPOCH comes from the function
19# get_source_date_epoch_value which reads from the SDE_FILE, or if the file
20# is not available will use the fallback of SOURCE_DATE_EPOCH_FALLBACK.
21#
22# The SDE_FILE is normally constructed from the function
23# create_source_date_epoch_stamp which is typically added as a postfuncs to
24# the do_unpack task.  If a recipe does NOT have do_unpack, it should be added
25# to a task that runs after the source is available and before the
26# do_deploy_source_date_epoch task is executed.
27#
28# If a recipe wishes to override the default behavior it should set it's own
29# SOURCE_DATE_EPOCH or override the do_deploy_source_date_epoch_stamp task
30# with recipe-specific functionality to write the appropriate
31# SOURCE_DATE_EPOCH into the SDE_FILE.
32#
33# SOURCE_DATE_EPOCH is intended to be a reproducible value.  This value should
34# be reproducible for anyone who builds the same revision from the same
35# sources.
36#
37# There are 4 ways the create_source_date_epoch_stamp function determines what
38# becomes SOURCE_DATE_EPOCH:
39#
40# 1. Use the value from __source_date_epoch.txt file if this file exists.
41#    This file was most likely created in the previous build by one of the
42#    following methods 2,3,4.
43#    Alternatively, it can be provided by a recipe via SRC_URI.
44#
45# If the file does not exist:
46#
47# 2. If there is a git checkout, use the last git commit timestamp.
48#    Git does not preserve file timestamps on checkout.
49#
50# 3. Use the mtime of "known" files such as NEWS, CHANGLELOG, ...
51#    This works for well-kept repositories distributed via tarball.
52#
53# 4. Use the modification time of the youngest file in the source tree, if
54#    there is one.
55#    This will be the newest file from the distribution tarball, if any.
56#
57# 5. Fall back to a fixed timestamp (SOURCE_DATE_EPOCH_FALLBACK).
58#
59# Once the value is determined, it is stored in the recipe's SDE_FILE.
60
61def get_source_date_epoch_from_known_files(d, sourcedir):
62    source_date_epoch = None
63    newest_file = None
64    known_files = set(["NEWS", "ChangeLog", "Changelog", "CHANGES"])
65    for file in known_files:
66        filepath = os.path.join(sourcedir, file)
67        if os.path.isfile(filepath):
68            mtime = int(os.lstat(filepath).st_mtime)
69            # There may be more than one "known_file" present, if so, use the youngest one
70            if not source_date_epoch or mtime > source_date_epoch:
71                source_date_epoch = mtime
72                newest_file = filepath
73    if newest_file:
74        bb.debug(1, "SOURCE_DATE_EPOCH taken from: %s" % newest_file)
75    return source_date_epoch
76
77def find_git_folder(d, sourcedir):
78    # First guess: WORKDIR/git
79    # This is the default git fetcher unpack path
80    workdir = d.getVar('WORKDIR')
81    gitpath = os.path.join(workdir, "git/.git")
82    if os.path.isdir(gitpath):
83        return gitpath
84
85    # Second guess: ${S}
86    gitpath = os.path.join(sourcedir, ".git")
87    if os.path.isdir(gitpath):
88        return gitpath
89
90    # Perhaps there was a subpath or destsuffix specified.
91    # Go looking in the WORKDIR
92    exclude = set(["build", "image", "license-destdir", "patches", "pseudo",
93                   "recipe-sysroot", "recipe-sysroot-native", "sysroot-destdir", "temp"])
94    for root, dirs, files in os.walk(workdir, topdown=True):
95        dirs[:] = [d for d in dirs if d not in exclude]
96        if '.git' in dirs:
97            return os.path.join(root, ".git")
98
99    bb.warn("Failed to find a git repository in WORKDIR: %s" % workdir)
100    return None
101
102def get_source_date_epoch_from_git(d, sourcedir):
103    if not "git://" in d.getVar('SRC_URI') and not "gitsm://" in d.getVar('SRC_URI'):
104        return None
105
106    gitpath = find_git_folder(d, sourcedir)
107    if not gitpath:
108        return None
109
110    # Check that the repository has a valid HEAD; it may not if subdir is used
111    # in SRC_URI
112    p = subprocess.run(['git', '--git-dir', gitpath, 'rev-parse', 'HEAD'], stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
113    if p.returncode != 0:
114        bb.debug(1, "%s does not have a valid HEAD: %s" % (gitpath, p.stdout.decode('utf-8')))
115        return None
116
117    bb.debug(1, "git repository: %s" % gitpath)
118    p = subprocess.run(['git', '-c', 'log.showSignature=false', '--git-dir', gitpath, 'log', '-1', '--pretty=%ct'],
119                       check=True, stdout=subprocess.PIPE)
120    return int(p.stdout.decode('utf-8'))
121
122def get_source_date_epoch_from_youngest_file(d, sourcedir):
123    if sourcedir == d.getVar('WORKDIR'):
124       # These sources are almost certainly not from a tarball
125       return None
126
127    # Do it the hard way: check all files and find the youngest one...
128    source_date_epoch = None
129    newest_file = None
130    for root, dirs, files in os.walk(sourcedir, topdown=True):
131        files = [f for f in files if not f[0] == '.']
132
133        for fname in files:
134            if fname == "singletask.lock":
135                 # Ignore externalsrc/devtool lockfile [YOCTO #14921]
136                 continue
137            filename = os.path.join(root, fname)
138            try:
139                mtime = int(os.lstat(filename).st_mtime)
140            except ValueError:
141                mtime = 0
142            if not source_date_epoch or mtime > source_date_epoch:
143                source_date_epoch = mtime
144                newest_file = filename
145
146    if newest_file:
147        bb.debug(1, "Newest file found: %s" % newest_file)
148    return source_date_epoch
149
150def fixed_source_date_epoch(d):
151    bb.debug(1, "No tarball or git repo found to determine SOURCE_DATE_EPOCH")
152    source_date_epoch = d.getVar('SOURCE_DATE_EPOCH_FALLBACK')
153    if source_date_epoch:
154        bb.debug(1, "Using SOURCE_DATE_EPOCH_FALLBACK")
155        return int(source_date_epoch)
156    return 0
157
158def get_source_date_epoch(d, sourcedir):
159    return (
160        get_source_date_epoch_from_git(d, sourcedir) or
161        get_source_date_epoch_from_youngest_file(d, sourcedir) or
162        fixed_source_date_epoch(d)       # Last resort
163    )
164
165def epochfile_read(epochfile, d):
166    cached, efile = d.getVar('__CACHED_SOURCE_DATE_EPOCH') or (None, None)
167    if cached and efile == epochfile:
168        return cached
169
170    if cached and epochfile != efile:
171        bb.debug(1, "Epoch file changed from %s to %s" % (efile, epochfile))
172
173    source_date_epoch = int(d.getVar('SOURCE_DATE_EPOCH_FALLBACK'))
174    try:
175        with open(epochfile, 'r') as f:
176            s = f.read()
177            try:
178                source_date_epoch = int(s)
179            except ValueError:
180                bb.warn("SOURCE_DATE_EPOCH value '%s' is invalid. Reverting to SOURCE_DATE_EPOCH_FALLBACK" % s)
181                source_date_epoch = int(d.getVar('SOURCE_DATE_EPOCH_FALLBACK'))
182        bb.debug(1, "SOURCE_DATE_EPOCH: %d" % source_date_epoch)
183    except FileNotFoundError:
184        bb.debug(1, "Cannot find %s. SOURCE_DATE_EPOCH will default to %d" % (epochfile, source_date_epoch))
185
186    d.setVar('__CACHED_SOURCE_DATE_EPOCH', (str(source_date_epoch), epochfile))
187    return str(source_date_epoch)
188
189def epochfile_write(source_date_epoch, epochfile, d):
190
191    bb.debug(1, "SOURCE_DATE_EPOCH: %d" % source_date_epoch)
192    bb.utils.mkdirhier(os.path.dirname(epochfile))
193
194    tmp_file = "%s.new" % epochfile
195    with open(tmp_file, 'w') as f:
196        f.write(str(source_date_epoch))
197    os.rename(tmp_file, epochfile)
198