xref: /openbmc/openbmc/poky/bitbake/lib/bb/siggen.py (revision 15ae2509)
1#
2# SPDX-License-Identifier: GPL-2.0-only
3#
4
5import hashlib
6import logging
7import os
8import re
9import tempfile
10import pickle
11import bb.data
12import difflib
13import simplediff
14from bb.checksum import FileChecksumCache
15
16logger = logging.getLogger('BitBake.SigGen')
17
18def init(d):
19    siggens = [obj for obj in globals().values()
20                      if type(obj) is type and issubclass(obj, SignatureGenerator)]
21
22    desired = d.getVar("BB_SIGNATURE_HANDLER") or "noop"
23    for sg in siggens:
24        if desired == sg.name:
25            return sg(d)
26            break
27    else:
28        logger.error("Invalid signature generator '%s', using default 'noop'\n"
29                     "Available generators: %s", desired,
30                     ', '.join(obj.name for obj in siggens))
31        return SignatureGenerator(d)
32
33class SignatureGenerator(object):
34    """
35    """
36    name = "noop"
37
38    def __init__(self, data):
39        self.basehash = {}
40        self.taskhash = {}
41        self.runtaskdeps = {}
42        self.file_checksum_values = {}
43        self.taints = {}
44
45    def finalise(self, fn, d, varient):
46        return
47
48    def get_unihash(self, task):
49        return self.taskhash[task]
50
51    def get_taskhash(self, fn, task, deps, dataCache):
52        return "0"
53
54    def writeout_file_checksum_cache(self):
55        """Write/update the file checksum cache onto disk"""
56        return
57
58    def stampfile(self, stampbase, file_name, taskname, extrainfo):
59        return ("%s.%s.%s" % (stampbase, taskname, extrainfo)).rstrip('.')
60
61    def stampcleanmask(self, stampbase, file_name, taskname, extrainfo):
62        return ("%s.%s.%s" % (stampbase, taskname, extrainfo)).rstrip('.')
63
64    def dump_sigtask(self, fn, task, stampbase, runtime):
65        return
66
67    def invalidate_task(self, task, d, fn):
68        bb.build.del_stamp(task, d, fn)
69
70    def dump_sigs(self, dataCache, options):
71        return
72
73    def get_taskdata(self):
74        return (self.runtaskdeps, self.taskhash, self.file_checksum_values, self.taints, self.basehash)
75
76    def set_taskdata(self, data):
77        self.runtaskdeps, self.taskhash, self.file_checksum_values, self.taints, self.basehash = data
78
79    def reset(self, data):
80        self.__init__(data)
81
82
83class SignatureGeneratorBasic(SignatureGenerator):
84    """
85    """
86    name = "basic"
87
88    def __init__(self, data):
89        self.basehash = {}
90        self.taskhash = {}
91        self.taskdeps = {}
92        self.runtaskdeps = {}
93        self.file_checksum_values = {}
94        self.taints = {}
95        self.gendeps = {}
96        self.lookupcache = {}
97        self.pkgnameextract = re.compile(r"(?P<fn>.*)\..*")
98        self.basewhitelist = set((data.getVar("BB_HASHBASE_WHITELIST") or "").split())
99        self.taskwhitelist = None
100        self.init_rundepcheck(data)
101        checksum_cache_file = data.getVar("BB_HASH_CHECKSUM_CACHE_FILE")
102        if checksum_cache_file:
103            self.checksum_cache = FileChecksumCache()
104            self.checksum_cache.init_cache(data, checksum_cache_file)
105        else:
106            self.checksum_cache = None
107
108    def init_rundepcheck(self, data):
109        self.taskwhitelist = data.getVar("BB_HASHTASK_WHITELIST") or None
110        if self.taskwhitelist:
111            self.twl = re.compile(self.taskwhitelist)
112        else:
113            self.twl = None
114
115    def _build_data(self, fn, d):
116
117        ignore_mismatch = ((d.getVar("BB_HASH_IGNORE_MISMATCH") or '') == '1')
118        tasklist, gendeps, lookupcache = bb.data.generate_dependencies(d)
119
120        taskdeps, basehash = bb.data.generate_dependency_hash(tasklist, gendeps, lookupcache, self.basewhitelist, fn)
121
122        for task in tasklist:
123            k = fn + "." + task
124            if not ignore_mismatch and k in self.basehash and self.basehash[k] != basehash[k]:
125                bb.error("When reparsing %s, the basehash value changed from %s to %s. The metadata is not deterministic and this needs to be fixed." % (k, self.basehash[k], basehash[k]))
126                bb.error("The following commands may help:")
127                cmd = "$ bitbake %s -c%s" % (d.getVar('PN'), task)
128                # Make sure sigdata is dumped before run printdiff
129                bb.error("%s -Snone" % cmd)
130                bb.error("Then:")
131                bb.error("%s -Sprintdiff\n" % cmd)
132            self.basehash[k] = basehash[k]
133
134        self.taskdeps[fn] = taskdeps
135        self.gendeps[fn] = gendeps
136        self.lookupcache[fn] = lookupcache
137
138        return taskdeps
139
140    def finalise(self, fn, d, variant):
141
142        mc = d.getVar("__BBMULTICONFIG", False) or ""
143        if variant or mc:
144            fn = bb.cache.realfn2virtual(fn, variant, mc)
145
146        try:
147            taskdeps = self._build_data(fn, d)
148        except bb.parse.SkipRecipe:
149            raise
150        except:
151            bb.warn("Error during finalise of %s" % fn)
152            raise
153
154        #Slow but can be useful for debugging mismatched basehashes
155        #for task in self.taskdeps[fn]:
156        #    self.dump_sigtask(fn, task, d.getVar("STAMP"), False)
157
158        for task in taskdeps:
159            d.setVar("BB_BASEHASH_task-%s" % task, self.basehash[fn + "." + task])
160
161    def rundep_check(self, fn, recipename, task, dep, depname, dataCache):
162        # Return True if we should keep the dependency, False to drop it
163        # We only manipulate the dependencies for packages not in the whitelist
164        if self.twl and not self.twl.search(recipename):
165            # then process the actual dependencies
166            if self.twl.search(depname):
167                return False
168        return True
169
170    def read_taint(self, fn, task, stampbase):
171        taint = None
172        try:
173            with open(stampbase + '.' + task + '.taint', 'r') as taintf:
174                taint = taintf.read()
175        except IOError:
176            pass
177        return taint
178
179    def get_taskhash(self, fn, task, deps, dataCache):
180
181        mc = ''
182        if fn.startswith('mc:'):
183            mc = fn.split(':')[1]
184        k = fn + "." + task
185
186        data = dataCache.basetaskhash[k]
187        self.basehash[k] = data
188        self.runtaskdeps[k] = []
189        self.file_checksum_values[k] = []
190        recipename = dataCache.pkg_fn[fn]
191        for dep in sorted(deps, key=clean_basepath):
192            pkgname = self.pkgnameextract.search(dep).group('fn')
193            if mc:
194                depmc = pkgname.split(':')[1]
195                if mc != depmc:
196                    continue
197            if dep.startswith("mc:") and not mc:
198                continue
199            depname = dataCache.pkg_fn[pkgname]
200            if not self.rundep_check(fn, recipename, task, dep, depname, dataCache):
201                continue
202            if dep not in self.taskhash:
203                bb.fatal("%s is not in taskhash, caller isn't calling in dependency order?" % dep)
204            data = data + self.get_unihash(dep)
205            self.runtaskdeps[k].append(dep)
206
207        if task in dataCache.file_checksums[fn]:
208            if self.checksum_cache:
209                checksums = self.checksum_cache.get_checksums(dataCache.file_checksums[fn][task], recipename)
210            else:
211                checksums = bb.fetch2.get_file_checksums(dataCache.file_checksums[fn][task], recipename)
212            for (f,cs) in checksums:
213                self.file_checksum_values[k].append((f,cs))
214                if cs:
215                    data = data + cs
216
217        taskdep = dataCache.task_deps[fn]
218        if 'nostamp' in taskdep and task in taskdep['nostamp']:
219            # Nostamp tasks need an implicit taint so that they force any dependent tasks to run
220            import uuid
221            taint = str(uuid.uuid4())
222            data = data + taint
223            self.taints[k] = "nostamp:" + taint
224
225        taint = self.read_taint(fn, task, dataCache.stamp[fn])
226        if taint:
227            data = data + taint
228            self.taints[k] = taint
229            logger.warning("%s is tainted from a forced run" % k)
230
231        h = hashlib.sha256(data.encode("utf-8")).hexdigest()
232        self.taskhash[k] = h
233        #d.setVar("BB_TASKHASH_task-%s" % task, taskhash[task])
234        return h
235
236    def writeout_file_checksum_cache(self):
237        """Write/update the file checksum cache onto disk"""
238        if self.checksum_cache:
239            self.checksum_cache.save_extras()
240            self.checksum_cache.save_merge()
241        else:
242            bb.fetch2.fetcher_parse_save()
243            bb.fetch2.fetcher_parse_done()
244
245    def dump_sigtask(self, fn, task, stampbase, runtime):
246
247        k = fn + "." + task
248        referencestamp = stampbase
249        if isinstance(runtime, str) and runtime.startswith("customfile"):
250            sigfile = stampbase
251            referencestamp = runtime[11:]
252        elif runtime and k in self.taskhash:
253            sigfile = stampbase + "." + task + ".sigdata" + "." + self.taskhash[k]
254        else:
255            sigfile = stampbase + "." + task + ".sigbasedata" + "." + self.basehash[k]
256
257        bb.utils.mkdirhier(os.path.dirname(sigfile))
258
259        data = {}
260        data['task'] = task
261        data['basewhitelist'] = self.basewhitelist
262        data['taskwhitelist'] = self.taskwhitelist
263        data['taskdeps'] = self.taskdeps[fn][task]
264        data['basehash'] = self.basehash[k]
265        data['gendeps'] = {}
266        data['varvals'] = {}
267        data['varvals'][task] = self.lookupcache[fn][task]
268        for dep in self.taskdeps[fn][task]:
269            if dep in self.basewhitelist:
270                continue
271            data['gendeps'][dep] = self.gendeps[fn][dep]
272            data['varvals'][dep] = self.lookupcache[fn][dep]
273
274        if runtime and k in self.taskhash:
275            data['runtaskdeps'] = self.runtaskdeps[k]
276            data['file_checksum_values'] = [(os.path.basename(f), cs) for f,cs in self.file_checksum_values[k]]
277            data['runtaskhashes'] = {}
278            for dep in data['runtaskdeps']:
279                data['runtaskhashes'][dep] = self.get_unihash(dep)
280            data['taskhash'] = self.taskhash[k]
281
282        taint = self.read_taint(fn, task, referencestamp)
283        if taint:
284            data['taint'] = taint
285
286        if runtime and k in self.taints:
287            if 'nostamp:' in self.taints[k]:
288                data['taint'] = self.taints[k]
289
290        computed_basehash = calc_basehash(data)
291        if computed_basehash != self.basehash[k]:
292            bb.error("Basehash mismatch %s versus %s for %s" % (computed_basehash, self.basehash[k], k))
293        if runtime and k in self.taskhash:
294            computed_taskhash = calc_taskhash(data)
295            if computed_taskhash != self.taskhash[k]:
296                bb.error("Taskhash mismatch %s versus %s for %s" % (computed_taskhash, self.taskhash[k], k))
297                sigfile = sigfile.replace(self.taskhash[k], computed_taskhash)
298
299        fd, tmpfile = tempfile.mkstemp(dir=os.path.dirname(sigfile), prefix="sigtask.")
300        try:
301            with os.fdopen(fd, "wb") as stream:
302                p = pickle.dump(data, stream, -1)
303                stream.flush()
304            os.chmod(tmpfile, 0o664)
305            os.rename(tmpfile, sigfile)
306        except (OSError, IOError) as err:
307            try:
308                os.unlink(tmpfile)
309            except OSError:
310                pass
311            raise err
312
313    def dump_sigfn(self, fn, dataCaches, options):
314        if fn in self.taskdeps:
315            for task in self.taskdeps[fn]:
316                tid = fn + ":" + task
317                (mc, _, _) = bb.runqueue.split_tid(tid)
318                k = fn + "." + task
319                if k not in self.taskhash:
320                    continue
321                if dataCaches[mc].basetaskhash[k] != self.basehash[k]:
322                    bb.error("Bitbake's cached basehash does not match the one we just generated (%s)!" % k)
323                    bb.error("The mismatched hashes were %s and %s" % (dataCaches[mc].basetaskhash[k], self.basehash[k]))
324                self.dump_sigtask(fn, task, dataCaches[mc].stamp[fn], True)
325
326class SignatureGeneratorBasicHash(SignatureGeneratorBasic):
327    name = "basichash"
328
329    def get_stampfile_hash(self, task):
330        if task in self.taskhash:
331            return self.taskhash[task]
332
333        # If task is not in basehash, then error
334        return self.basehash[task]
335
336    def stampfile(self, stampbase, fn, taskname, extrainfo, clean=False):
337        if taskname != "do_setscene" and taskname.endswith("_setscene"):
338            k = fn + "." + taskname[:-9]
339        else:
340            k = fn + "." + taskname
341        if clean:
342            h = "*"
343        else:
344            h = self.get_stampfile_hash(k)
345
346        return ("%s.%s.%s.%s" % (stampbase, taskname, h, extrainfo)).rstrip('.')
347
348    def stampcleanmask(self, stampbase, fn, taskname, extrainfo):
349        return self.stampfile(stampbase, fn, taskname, extrainfo, clean=True)
350
351    def invalidate_task(self, task, d, fn):
352        bb.note("Tainting hash to force rebuild of task %s, %s" % (fn, task))
353        bb.build.write_taint(task, d, fn)
354
355def dump_this_task(outfile, d):
356    import bb.parse
357    fn = d.getVar("BB_FILENAME")
358    task = "do_" + d.getVar("BB_CURRENTTASK")
359    referencestamp = bb.build.stamp_internal(task, d, None, True)
360    bb.parse.siggen.dump_sigtask(fn, task, outfile, "customfile:" + referencestamp)
361
362def init_colors(enable_color):
363    """Initialise colour dict for passing to compare_sigfiles()"""
364    # First set up the colours
365    colors = {'color_title':   '\033[1m',
366              'color_default': '\033[0m',
367              'color_add':     '\033[0;32m',
368              'color_remove':  '\033[0;31m',
369             }
370    # Leave all keys present but clear the values
371    if not enable_color:
372        for k in colors.keys():
373            colors[k] = ''
374    return colors
375
376def worddiff_str(oldstr, newstr, colors=None):
377    if not colors:
378        colors = init_colors(False)
379    diff = simplediff.diff(oldstr.split(' '), newstr.split(' '))
380    ret = []
381    for change, value in diff:
382        value = ' '.join(value)
383        if change == '=':
384            ret.append(value)
385        elif change == '+':
386            item = '{color_add}{{+{value}+}}{color_default}'.format(value=value, **colors)
387            ret.append(item)
388        elif change == '-':
389            item = '{color_remove}[-{value}-]{color_default}'.format(value=value, **colors)
390            ret.append(item)
391    whitespace_note = ''
392    if oldstr != newstr and ' '.join(oldstr.split()) == ' '.join(newstr.split()):
393        whitespace_note = ' (whitespace changed)'
394    return '"%s"%s' % (' '.join(ret), whitespace_note)
395
396def list_inline_diff(oldlist, newlist, colors=None):
397    if not colors:
398        colors = init_colors(False)
399    diff = simplediff.diff(oldlist, newlist)
400    ret = []
401    for change, value in diff:
402        value = ' '.join(value)
403        if change == '=':
404            ret.append("'%s'" % value)
405        elif change == '+':
406            item = '{color_add}+{value}{color_default}'.format(value=value, **colors)
407            ret.append(item)
408        elif change == '-':
409            item = '{color_remove}-{value}{color_default}'.format(value=value, **colors)
410            ret.append(item)
411    return '[%s]' % (', '.join(ret))
412
413def clean_basepath(a):
414    mc = None
415    if a.startswith("mc:"):
416        _, mc, a = a.split(":", 2)
417    b = a.rsplit("/", 2)[1] + '/' + a.rsplit("/", 2)[2]
418    if a.startswith("virtual:"):
419        b = b + ":" + a.rsplit(":", 1)[0]
420    if mc:
421        b = b + ":mc:" + mc
422    return b
423
424def clean_basepaths(a):
425    b = {}
426    for x in a:
427        b[clean_basepath(x)] = a[x]
428    return b
429
430def clean_basepaths_list(a):
431    b = []
432    for x in a:
433        b.append(clean_basepath(x))
434    return b
435
436def compare_sigfiles(a, b, recursecb=None, color=False, collapsed=False):
437    output = []
438
439    colors = init_colors(color)
440    def color_format(formatstr, **values):
441        """
442        Return colour formatted string.
443        NOTE: call with the format string, not an already formatted string
444        containing values (otherwise you could have trouble with { and }
445        characters)
446        """
447        if not formatstr.endswith('{color_default}'):
448            formatstr += '{color_default}'
449        # In newer python 3 versions you can pass both of these directly,
450        # but we only require 3.4 at the moment
451        formatparams = {}
452        formatparams.update(colors)
453        formatparams.update(values)
454        return formatstr.format(**formatparams)
455
456    with open(a, 'rb') as f:
457        p1 = pickle.Unpickler(f)
458        a_data = p1.load()
459    with open(b, 'rb') as f:
460        p2 = pickle.Unpickler(f)
461        b_data = p2.load()
462
463    def dict_diff(a, b, whitelist=set()):
464        sa = set(a.keys())
465        sb = set(b.keys())
466        common = sa & sb
467        changed = set()
468        for i in common:
469            if a[i] != b[i] and i not in whitelist:
470                changed.add(i)
471        added = sb - sa
472        removed = sa - sb
473        return changed, added, removed
474
475    def file_checksums_diff(a, b):
476        from collections import Counter
477        # Handle old siginfo format
478        if isinstance(a, dict):
479            a = [(os.path.basename(f), cs) for f, cs in a.items()]
480        if isinstance(b, dict):
481            b = [(os.path.basename(f), cs) for f, cs in b.items()]
482        # Compare lists, ensuring we can handle duplicate filenames if they exist
483        removedcount = Counter(a)
484        removedcount.subtract(b)
485        addedcount = Counter(b)
486        addedcount.subtract(a)
487        added = []
488        for x in b:
489            if addedcount[x] > 0:
490                addedcount[x] -= 1
491                added.append(x)
492        removed = []
493        changed = []
494        for x in a:
495            if removedcount[x] > 0:
496                removedcount[x] -= 1
497                for y in added:
498                    if y[0] == x[0]:
499                        changed.append((x[0], x[1], y[1]))
500                        added.remove(y)
501                        break
502                else:
503                    removed.append(x)
504        added = [x[0] for x in added]
505        removed = [x[0] for x in removed]
506        return changed, added, removed
507
508    if 'basewhitelist' in a_data and a_data['basewhitelist'] != b_data['basewhitelist']:
509        output.append(color_format("{color_title}basewhitelist changed{color_default} from '%s' to '%s'") % (a_data['basewhitelist'], b_data['basewhitelist']))
510        if a_data['basewhitelist'] and b_data['basewhitelist']:
511            output.append("changed items: %s" % a_data['basewhitelist'].symmetric_difference(b_data['basewhitelist']))
512
513    if 'taskwhitelist' in a_data and a_data['taskwhitelist'] != b_data['taskwhitelist']:
514        output.append(color_format("{color_title}taskwhitelist changed{color_default} from '%s' to '%s'") % (a_data['taskwhitelist'], b_data['taskwhitelist']))
515        if a_data['taskwhitelist'] and b_data['taskwhitelist']:
516            output.append("changed items: %s" % a_data['taskwhitelist'].symmetric_difference(b_data['taskwhitelist']))
517
518    if a_data['taskdeps'] != b_data['taskdeps']:
519        output.append(color_format("{color_title}Task dependencies changed{color_default} from:\n%s\nto:\n%s") % (sorted(a_data['taskdeps']), sorted(b_data['taskdeps'])))
520
521    if a_data['basehash'] != b_data['basehash'] and not collapsed:
522        output.append(color_format("{color_title}basehash changed{color_default} from %s to %s") % (a_data['basehash'], b_data['basehash']))
523
524    changed, added, removed = dict_diff(a_data['gendeps'], b_data['gendeps'], a_data['basewhitelist'] & b_data['basewhitelist'])
525    if changed:
526        for dep in changed:
527            output.append(color_format("{color_title}List of dependencies for variable %s changed from '{color_default}%s{color_title}' to '{color_default}%s{color_title}'") % (dep, a_data['gendeps'][dep], b_data['gendeps'][dep]))
528            if a_data['gendeps'][dep] and b_data['gendeps'][dep]:
529                output.append("changed items: %s" % a_data['gendeps'][dep].symmetric_difference(b_data['gendeps'][dep]))
530    if added:
531        for dep in added:
532            output.append(color_format("{color_title}Dependency on variable %s was added") % (dep))
533    if removed:
534        for dep in removed:
535            output.append(color_format("{color_title}Dependency on Variable %s was removed") % (dep))
536
537
538    changed, added, removed = dict_diff(a_data['varvals'], b_data['varvals'])
539    if changed:
540        for dep in changed:
541            oldval = a_data['varvals'][dep]
542            newval = b_data['varvals'][dep]
543            if newval and oldval and ('\n' in oldval or '\n' in newval):
544                diff = difflib.unified_diff(oldval.splitlines(), newval.splitlines(), lineterm='')
545                # Cut off the first two lines, since we aren't interested in
546                # the old/new filename (they are blank anyway in this case)
547                difflines = list(diff)[2:]
548                if color:
549                    # Add colour to diff output
550                    for i, line in enumerate(difflines):
551                        if line.startswith('+'):
552                            line = color_format('{color_add}{line}', line=line)
553                            difflines[i] = line
554                        elif line.startswith('-'):
555                            line = color_format('{color_remove}{line}', line=line)
556                            difflines[i] = line
557                output.append(color_format("{color_title}Variable {var} value changed:{color_default}\n{diff}", var=dep, diff='\n'.join(difflines)))
558            elif newval and oldval and (' ' in oldval or ' ' in newval):
559                output.append(color_format("{color_title}Variable {var} value changed:{color_default}\n{diff}", var=dep, diff=worddiff_str(oldval, newval, colors)))
560            else:
561                output.append(color_format("{color_title}Variable {var} value changed from '{color_default}{oldval}{color_title}' to '{color_default}{newval}{color_title}'{color_default}", var=dep, oldval=oldval, newval=newval))
562
563    if not 'file_checksum_values' in a_data:
564         a_data['file_checksum_values'] = {}
565    if not 'file_checksum_values' in b_data:
566         b_data['file_checksum_values'] = {}
567
568    changed, added, removed = file_checksums_diff(a_data['file_checksum_values'], b_data['file_checksum_values'])
569    if changed:
570        for f, old, new in changed:
571            output.append(color_format("{color_title}Checksum for file %s changed{color_default} from %s to %s") % (f, old, new))
572    if added:
573        for f in added:
574            output.append(color_format("{color_title}Dependency on checksum of file %s was added") % (f))
575    if removed:
576        for f in removed:
577            output.append(color_format("{color_title}Dependency on checksum of file %s was removed") % (f))
578
579    if not 'runtaskdeps' in a_data:
580         a_data['runtaskdeps'] = {}
581    if not 'runtaskdeps' in b_data:
582         b_data['runtaskdeps'] = {}
583
584    if not collapsed:
585        if len(a_data['runtaskdeps']) != len(b_data['runtaskdeps']):
586            changed = ["Number of task dependencies changed"]
587        else:
588            changed = []
589            for idx, task in enumerate(a_data['runtaskdeps']):
590                a = a_data['runtaskdeps'][idx]
591                b = b_data['runtaskdeps'][idx]
592                if a_data['runtaskhashes'][a] != b_data['runtaskhashes'][b] and not collapsed:
593                    changed.append("%s with hash %s\n changed to\n%s with hash %s" % (clean_basepath(a), a_data['runtaskhashes'][a], clean_basepath(b), b_data['runtaskhashes'][b]))
594
595        if changed:
596            clean_a = clean_basepaths_list(a_data['runtaskdeps'])
597            clean_b = clean_basepaths_list(b_data['runtaskdeps'])
598            if clean_a != clean_b:
599                output.append(color_format("{color_title}runtaskdeps changed:{color_default}\n%s") % list_inline_diff(clean_a, clean_b, colors))
600            else:
601                output.append(color_format("{color_title}runtaskdeps changed:"))
602            output.append("\n".join(changed))
603
604
605    if 'runtaskhashes' in a_data and 'runtaskhashes' in b_data:
606        a = a_data['runtaskhashes']
607        b = b_data['runtaskhashes']
608        changed, added, removed = dict_diff(a, b)
609        if added:
610            for dep in added:
611                bdep_found = False
612                if removed:
613                    for bdep in removed:
614                        if b[dep] == a[bdep]:
615                            #output.append("Dependency on task %s was replaced by %s with same hash" % (dep, bdep))
616                            bdep_found = True
617                if not bdep_found:
618                    output.append(color_format("{color_title}Dependency on task %s was added{color_default} with hash %s") % (clean_basepath(dep), b[dep]))
619        if removed:
620            for dep in removed:
621                adep_found = False
622                if added:
623                    for adep in added:
624                        if b[adep] == a[dep]:
625                            #output.append("Dependency on task %s was replaced by %s with same hash" % (adep, dep))
626                            adep_found = True
627                if not adep_found:
628                    output.append(color_format("{color_title}Dependency on task %s was removed{color_default} with hash %s") % (clean_basepath(dep), a[dep]))
629        if changed:
630            for dep in changed:
631                if not collapsed:
632                    output.append(color_format("{color_title}Hash for dependent task %s changed{color_default} from %s to %s") % (clean_basepath(dep), a[dep], b[dep]))
633                if callable(recursecb):
634                    recout = recursecb(dep, a[dep], b[dep])
635                    if recout:
636                        if collapsed:
637                            output.extend(recout)
638                        else:
639                            # If a dependent hash changed, might as well print the line above and then defer to the changes in
640                            # that hash since in all likelyhood, they're the same changes this task also saw.
641                            output = [output[-1]] + recout
642
643    a_taint = a_data.get('taint', None)
644    b_taint = b_data.get('taint', None)
645    if a_taint != b_taint:
646        if a_taint.startswith('nostamp:'):
647            a_taint = a_taint.replace('nostamp:', 'nostamp(uuid4):')
648        if b_taint.startswith('nostamp:'):
649            b_taint = b_taint.replace('nostamp:', 'nostamp(uuid4):')
650        output.append(color_format("{color_title}Taint (by forced/invalidated task) changed{color_default} from %s to %s") % (a_taint, b_taint))
651
652    return output
653
654
655def calc_basehash(sigdata):
656    task = sigdata['task']
657    basedata = sigdata['varvals'][task]
658
659    if basedata is None:
660        basedata = ''
661
662    alldeps = sigdata['taskdeps']
663    for dep in alldeps:
664        basedata = basedata + dep
665        val = sigdata['varvals'][dep]
666        if val is not None:
667            basedata = basedata + str(val)
668
669    return hashlib.sha256(basedata.encode("utf-8")).hexdigest()
670
671def calc_taskhash(sigdata):
672    data = sigdata['basehash']
673
674    for dep in sigdata['runtaskdeps']:
675        data = data + sigdata['runtaskhashes'][dep]
676
677    for c in sigdata['file_checksum_values']:
678        if c[1]:
679            data = data + c[1]
680
681    if 'taint' in sigdata:
682        if 'nostamp:' in sigdata['taint']:
683            data = data + sigdata['taint'][8:]
684        else:
685            data = data + sigdata['taint']
686
687    return hashlib.sha256(data.encode("utf-8")).hexdigest()
688
689
690def dump_sigfile(a):
691    output = []
692
693    with open(a, 'rb') as f:
694        p1 = pickle.Unpickler(f)
695        a_data = p1.load()
696
697    output.append("basewhitelist: %s" % (a_data['basewhitelist']))
698
699    output.append("taskwhitelist: %s" % (a_data['taskwhitelist']))
700
701    output.append("Task dependencies: %s" % (sorted(a_data['taskdeps'])))
702
703    output.append("basehash: %s" % (a_data['basehash']))
704
705    for dep in a_data['gendeps']:
706        output.append("List of dependencies for variable %s is %s" % (dep, a_data['gendeps'][dep]))
707
708    for dep in a_data['varvals']:
709        output.append("Variable %s value is %s" % (dep, a_data['varvals'][dep]))
710
711    if 'runtaskdeps' in a_data:
712        output.append("Tasks this task depends on: %s" % (a_data['runtaskdeps']))
713
714    if 'file_checksum_values' in a_data:
715        output.append("This task depends on the checksums of files: %s" % (a_data['file_checksum_values']))
716
717    if 'runtaskhashes' in a_data:
718        for dep in a_data['runtaskhashes']:
719            output.append("Hash for dependent task %s is %s" % (dep, a_data['runtaskhashes'][dep]))
720
721    if 'taint' in a_data:
722        if a_data['taint'].startswith('nostamp:'):
723            msg = a_data['taint'].replace('nostamp:', 'nostamp(uuid4):')
724        else:
725            msg = a_data['taint']
726        output.append("Tainted (by forced/invalidated task): %s" % msg)
727
728    if 'task' in a_data:
729        computed_basehash = calc_basehash(a_data)
730        output.append("Computed base hash is %s and from file %s" % (computed_basehash, a_data['basehash']))
731    else:
732        output.append("Unable to compute base hash")
733
734    computed_taskhash = calc_taskhash(a_data)
735    output.append("Computed task hash is %s" % computed_taskhash)
736
737    return output
738