/usr/lib/python2.7/dist-packages/apport/report.py is in python-apport 2.14.1-0ubuntu3.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 | '''Representation of and data collection for a problem report.'''
# Copyright (C) 2006 - 2009 Canonical Ltd.
# Author: Martin Pitt <martin.pitt@ubuntu.com>
#
# This program is free software; you can redistribute it and/or modify it
# under the terms of the GNU General Public License as published by the
# Free Software Foundation; either version 2 of the License, or (at your
# option) any later version. See http://www.gnu.org/copyleft/gpl.html for
# the full text of the license.
import subprocess, tempfile, os.path, re, pwd, grp, os, time
import fnmatch, glob, traceback, errno, sys, atexit, locale
import xml.dom, xml.dom.minidom
from xml.parsers.expat import ExpatError
if sys.version > '3':
_python2 = False
from urllib.error import URLError
from urllib.request import urlopen
(urlopen, URLError) # pyflakes
else:
_python2 = True
from urllib import urlopen
URLError = IOError
import problem_report
import apport
import apport.fileutils
from apport.packaging_impl import impl as packaging
_data_dir = os.environ.get('APPORT_DATA_DIR', '/usr/share/apport')
_hook_dir = '%s/package-hooks/' % (_data_dir)
_common_hook_dir = '%s/general-hooks/' % (_data_dir)
_opt_dir = '/opt'
# path of the ignore file
_ignore_file = '~/.apport-ignore.xml'
# system-wide blacklist
_blacklist_dir = '/etc/apport/blacklist.d'
_whitelist_dir = '/etc/apport/whitelist.d'
# programs that we consider interpreters
interpreters = ['sh', 'bash', 'dash', 'csh', 'tcsh', 'python*',
'ruby*', 'php', 'perl*', 'mono*', 'awk']
#
# helper functions
#
def _transitive_dependencies(package, depends_set):
'''Recursively add dependencies of package to depends_set.'''
try:
packaging.get_version(package)
except ValueError:
return
for d in packaging.get_dependencies(package):
if not d in depends_set:
depends_set.add(d)
_transitive_dependencies(d, depends_set)
def _read_file(path):
'''Read file content.
Return its content, or return a textual error if it failed.
'''
try:
with open(path, 'rb') as fd:
return fd.read().strip().decode('UTF-8', errors='replace')
except (OSError, IOError) as e:
return 'Error: ' + str(e)
def _read_maps(pid):
'''Read /proc/pid/maps.
Since /proc/$pid/maps may become unreadable unless we are ptracing the
process, detect this, and attempt to attach/detach.
'''
maps = 'Error: unable to read /proc maps file'
try:
with open('/proc/%d/maps' % pid) as fd:
maps = fd.read().strip()
except (OSError, IOError) as e:
return 'Error: ' + str(e)
return maps
def _command_output(command, input=None, stderr=subprocess.STDOUT):
'''Run command and capture its output.
Try to execute given command (argv list) and return its stdout, or return
a textual error if it failed.
'''
sp = subprocess.Popen(command, stdout=subprocess.PIPE, stderr=stderr)
(out, err) = sp.communicate(input)
if sp.returncode == 0:
return out
else:
if err:
err = err.decode('UTF-8', errors='replace')
else:
err = ''
raise OSError('Error: command %s failed with exit code %i: %s' % (
str(command), sp.returncode, err))
def _check_bug_pattern(report, pattern):
'''Check if given report matches the given bug pattern XML DOM node.
Return the bug URL on match, otherwise None.
'''
if _python2:
if not pattern.attributes.has_key('url'):
return None
else:
if 'url' not in pattern.attributes:
return None
for c in pattern.childNodes:
# regular expression condition
if c.nodeType == xml.dom.Node.ELEMENT_NODE and c.nodeName == 're':
try:
key = c.attributes['key'].nodeValue
except KeyError:
continue
if key not in report:
return None
c.normalize()
if c.hasChildNodes() and c.childNodes[0].nodeType == xml.dom.Node.TEXT_NODE:
regexp = c.childNodes[0].nodeValue
if _python2:
regexp = regexp.encode('UTF-8')
v = report[key]
if isinstance(v, problem_report.CompressedValue):
v = v.get_value()
if not _python2:
regexp = regexp.encode('UTF-8')
elif isinstance(v, bytes):
if not _python2:
regexp = regexp.encode('UTF-8')
try:
re_c = re.compile(regexp)
except:
continue
if not re_c.search(v):
return None
if _python2:
return pattern.attributes['url'].nodeValue.encode('UTF-8')
else:
return pattern.attributes['url'].nodeValue
def _check_bug_patterns(report, patterns):
try:
if _python2:
patterns = patterns.encode('UTF-8')
dom = xml.dom.minidom.parseString(patterns)
except (ExpatError, UnicodeEncodeError):
return None
for pattern in dom.getElementsByTagName('pattern'):
url = _check_bug_pattern(report, pattern)
if url:
return url
return None
def _dom_remove_space(node):
'''Recursively remove whitespace from given XML DOM node.'''
for c in node.childNodes:
if c.nodeType == xml.dom.Node.TEXT_NODE and c.nodeValue.strip() == '':
c.unlink()
node.removeChild(c)
else:
_dom_remove_space(c)
def _run_hook(report, ui, hook):
if not os.path.exists(hook):
return False
symb = {}
try:
with open(hook) as fd:
exec(compile(fd.read(), hook, 'exec'), symb)
try:
symb['add_info'](report, ui)
except TypeError as e:
if str(e).startswith('add_info()'):
# older versions of apport did not pass UI, and hooks that
# do not require it don't need to take it
symb['add_info'](report)
else:
raise
except StopIteration:
return True
except:
hookname = os.path.splitext(os.path.basename(hook))[0].replace('-', '_')
report['HookError_' + hookname] = traceback.format_exc()
apport.error('hook %s crashed:', hook)
traceback.print_exc()
return False
#
# Report class
#
class Report(problem_report.ProblemReport):
'''A problem report specific to apport (crash or bug).
This class wraps a standard ProblemReport and adds methods for collecting
standard debugging data.'''
def __init__(self, type='Crash', date=None):
'''Initialize a fresh problem report.
date is the desired date/time string; if None (default), the current
local time is used.
If the report is attached to a process ID, this should be set in
self.pid, so that e. g. hooks can use it to collect additional data.
'''
problem_report.ProblemReport.__init__(self, type, date)
self.pid = None
self._proc_maps_cache = None
def _customized_package_suffix(self, package):
'''Return a string suitable for appending to Package/Dependencies.
If package has only unmodified files, return the empty string. If not,
return ' [modified: ...]' with a list of modified files.
'''
suffix = ''
mod = packaging.get_modified_files(package)
if mod:
suffix += ' [modified: %s]' % ' '.join(mod)
try:
if not packaging.is_distro_package(package):
origin = packaging.get_package_origin(package)
if origin:
suffix += ' [origin: %s]' % origin
else:
suffix += ' [origin: unknown]'
except ValueError:
# no-op for nonexisting packages
pass
return suffix
def add_package_info(self, package=None):
'''Add packaging information.
If package is not given, the report must have ExecutablePath.
This adds:
- Package: package name and installed version
- SourcePackage: source package name (if possible to determine)
- PackageArchitecture: processor architecture this package was built
for
- Dependencies: package names and versions of all dependencies and
pre-dependencies; this also checks if the files are unmodified and
appends a list of all modified files
'''
if not package:
# the kernel does not have a executable path but a package
if not 'ExecutablePath' in self and self['ProblemType'] == 'KernelCrash':
package = self['Package']
else:
package = apport.fileutils.find_file_package(self['ExecutablePath'])
if not package:
return
try:
version = packaging.get_version(package)
except ValueError:
# package not installed
version = None
self['Package'] = '%s %s%s' % (package, version or '(not installed)',
self._customized_package_suffix(package))
if version or 'SourcePackage' not in self:
try:
self['SourcePackage'] = packaging.get_source(package)
except ValueError:
# might not exist for non-free third-party packages
pass
if not version:
return
self['PackageArchitecture'] = packaging.get_architecture(package)
# get set of all transitive dependencies
dependencies = set([])
_transitive_dependencies(package, dependencies)
# get dependency versions
self['Dependencies'] = ''
for dep in sorted(dependencies):
try:
v = packaging.get_version(dep)
except ValueError:
# can happen with uninstalled alternate dependencies
continue
if self['Dependencies']:
self['Dependencies'] += '\n'
self['Dependencies'] += '%s %s%s' % (
dep, v, self._customized_package_suffix(dep))
def add_os_info(self):
'''Add operating system information.
This adds:
- DistroRelease: NAME and VERSION from /etc/os-release, or
'lsb_release -sir' output
- Architecture: system architecture in distro specific notation
- Uname: uname -srm output
- NonfreeKernelModules: loaded kernel modules which are not free (if
there are none, this field will not be present)
'''
if 'DistroRelease' not in self:
self['DistroRelease'] = '%s %s' % apport.packaging.get_os_version()
if 'Uname' not in self:
u = os.uname()
self['Uname'] = '%s %s %s' % (u[0], u[2], u[4])
if 'Architecture' not in self:
self['Architecture'] = packaging.get_system_architecture()
def add_user_info(self):
'''Add information about the user.
This adds:
- UserGroups: system groups the user is in
'''
user = pwd.getpwuid(os.getuid())[0]
groups = [name for name, p, gid, memb in grp.getgrall()
if user in memb and gid < 1000]
groups.sort()
self['UserGroups'] = ' '.join(groups)
def _check_interpreted(self):
'''Check if process is a script.
Use ExecutablePath, ProcStatus and ProcCmdline to determine if
process is an interpreted script. If so, set InterpreterPath
accordingly.
'''
if 'ExecutablePath' not in self:
return
exebasename = os.path.basename(self['ExecutablePath'])
# check if we consider ExecutablePath an interpreter; we have to do
# this, otherwise 'gedit /tmp/foo.txt' would be detected as interpreted
# script as well
if not any(filter(lambda i: fnmatch.fnmatch(exebasename, i), interpreters)):
return
# first, determine process name
name = None
for l in self['ProcStatus'].splitlines():
try:
(k, v) = l.split('\t', 1)
except ValueError:
continue
if k == 'Name:':
name = v
break
if not name:
return
cmdargs = self['ProcCmdline'].split('\0')
bindirs = ['/bin/', '/sbin/', '/usr/bin/', '/usr/sbin/']
# filter out interpreter options
while len(cmdargs) >= 2 and cmdargs[1].startswith('-'):
# check for -m
if name.startswith('python') and cmdargs[1] == '-m' and len(cmdargs) >= 3:
path = self._python_module_path(cmdargs[2])
if path:
self['InterpreterPath'] = self['ExecutablePath']
self['ExecutablePath'] = path
else:
self['UnreportableReason'] = 'Cannot determine path of python module %s' % cmdargs[2]
return
del cmdargs[1]
# catch scripts explicitly called with interpreter
if len(cmdargs) >= 2:
# ensure that cmdargs[1] is an absolute path
if cmdargs[1].startswith('.') and 'ProcCwd' in self:
cmdargs[1] = os.path.join(self['ProcCwd'], cmdargs[1])
if os.access(cmdargs[1], os.R_OK):
self['InterpreterPath'] = self['ExecutablePath']
self['ExecutablePath'] = os.path.realpath(cmdargs[1])
# catch directly executed scripts
if 'InterpreterPath' not in self and name != exebasename:
for p in bindirs:
if os.access(p + cmdargs[0], os.R_OK):
argvexe = p + cmdargs[0]
if os.path.basename(os.path.realpath(argvexe)) == name:
self['InterpreterPath'] = self['ExecutablePath']
self['ExecutablePath'] = argvexe
break
# special case: crashes from twistd are usually the fault of the
# launched program
if 'InterpreterPath' in self and os.path.basename(self['ExecutablePath']) == 'twistd':
self['InterpreterPath'] = self['ExecutablePath']
exe = self._twistd_executable()
if exe:
self['ExecutablePath'] = exe
else:
self['UnreportableReason'] = 'Cannot determine twistd client program'
def _twistd_executable(self):
'''Determine the twistd client program from ProcCmdline.'''
args = self['ProcCmdline'].split('\0')[2:]
# search for a -f/--file, -y/--python or -s/--source argument
while args:
arg = args[0].split('=', 1)
if arg[0].startswith('--file') or arg[0].startswith('--python') or arg[0].startswith('--source'):
if len(arg) == 2:
return arg[1]
else:
return args[1]
elif len(arg[0]) > 1 and arg[0][0] == '-' and arg[0][1] != '-':
opts = arg[0][1:]
if 'f' in opts or 'y' in opts or 's' in opts:
return args[1]
args.pop(0)
return None
@classmethod
def _python_module_path(klass, module):
'''Determine path of given Python module'''
module = module.replace('/', '.')
try:
m = __import__(module)
m
except:
return None
# chop off the first component, as it's already covered by m
submodule = module.split('.')[1:]
if submodule:
path = eval('m.%s.__file__' % '.'.join(submodule))
else:
path = m.__file__
if path.endswith('.pyc'):
path = path[:-1]
return path
def add_proc_info(self, pid=None, extraenv=[]):
'''Add /proc/pid information.
If neither pid nor self.pid are given, it defaults to the process'
current pid and sets self.pid.
This adds the following fields:
- ExecutablePath: /proc/pid/exe contents; if the crashed process is
interpreted, this contains the script path instead
- InterpreterPath: /proc/pid/exe contents if the crashed process is
interpreted; otherwise this key does not exist
- ExecutableTimestamp: time stamp of ExecutablePath, for comparing at
report time
- ProcEnviron: A subset of the process' environment (only some standard
variables that do not disclose potentially sensitive information, plus
the ones mentioned in extraenv)
- ProcCmdline: /proc/pid/cmdline contents
- ProcStatus: /proc/pid/status contents
- ProcMaps: /proc/pid/maps contents
- ProcAttrCurrent: /proc/pid/attr/current contents, if not "unconfined"
- CurrentDesktop: Value of $XDG_CURRENT_DESKTOP, if present
- _LogindSession: logind cgroup path, if present (Used for filtering
out crashes that happened in a session that is not running any more)
'''
if not pid:
pid = self.pid or os.getpid()
if not self.pid:
self.pid = int(pid)
pid = str(pid)
try:
self['ProcCwd'] = os.readlink('/proc/' + pid + '/cwd')
except OSError:
pass
self.add_proc_environ(pid, extraenv)
self['ProcStatus'] = _read_file('/proc/' + pid + '/status')
self['ProcCmdline'] = _read_file('/proc/' + pid + '/cmdline').rstrip('\0')
self['ProcMaps'] = _read_maps(int(pid))
try:
self['ExecutablePath'] = os.readlink('/proc/' + pid + '/exe')
except OSError as e:
if e.errno == errno.ENOENT:
raise ValueError('invalid process')
else:
raise
for p in ('rofs', 'rwfs', 'squashmnt', 'persistmnt'):
if self['ExecutablePath'].startswith('/%s/' % p):
self['ExecutablePath'] = self['ExecutablePath'][len('/%s' % p):]
break
if not os.path.exists(self['ExecutablePath']):
raise ValueError('%s does not exist' % self['ExecutablePath'])
# check if we have an interpreted program
self._check_interpreted()
self['ExecutableTimestamp'] = str(int(os.stat(self['ExecutablePath']).st_mtime))
# make ProcCmdline ASCII friendly, do shell escaping
self['ProcCmdline'] = self['ProcCmdline'].replace('\\', '\\\\').replace(' ', '\\ ').replace('\0', ' ')
# grab AppArmor or SELinux context
# If no LSM is loaded, reading will return -EINVAL
try:
# On Linux 2.6.28+, 'current' is world readable, but read() gives
# EPERM; Python 2.5.3+ crashes on that (LP: #314065)
if os.getuid() == 0:
with open('/proc/' + pid + '/attr/current') as fd:
val = fd.read().strip()
if val != 'unconfined':
self['ProcAttrCurrent'] = val
except (IOError, OSError):
pass
ret = self.get_logind_session(pid)
if ret:
self['_LogindSession'] = ret[0]
def add_proc_environ(self, pid=None, extraenv=[]):
'''Add environment information.
If pid is not given, it defaults to the process' current pid.
This adds the following fields:
- ProcEnviron: A subset of the process' environment (only some standard
variables that do not disclose potentially sensitive information, plus
the ones mentioned in extraenv)
- CurrentDesktop: Value of $XDG_CURRENT_DESKTOP, if present
'''
safe_vars = ['SHELL', 'TERM', 'LANGUAGE', 'LANG', 'LC_CTYPE',
'LC_COLLATE', 'LC_TIME', 'LC_NUMERIC', 'LC_MONETARY',
'LC_MESSAGES', 'LC_PAPER', 'LC_NAME', 'LC_ADDRESS',
'LC_TELEPHONE', 'LC_MEASUREMENT', 'LC_IDENTIFICATION',
'LOCPATH'] + extraenv
if not pid:
pid = os.getpid()
pid = str(pid)
self['ProcEnviron'] = ''
env = _read_file('/proc/' + pid + '/environ').replace('\n', '\\n')
if env.startswith('Error:'):
self['ProcEnviron'] = env
else:
for l in env.split('\0'):
if l.split('=', 1)[0] in safe_vars:
if self['ProcEnviron']:
self['ProcEnviron'] += '\n'
self['ProcEnviron'] += l
elif l.startswith('PATH='):
p = l.split('=', 1)[1]
if '/home' in p or '/tmp' in p:
if self['ProcEnviron']:
self['ProcEnviron'] += '\n'
self['ProcEnviron'] += 'PATH=(custom, user)'
elif p != '/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games':
if self['ProcEnviron']:
self['ProcEnviron'] += '\n'
self['ProcEnviron'] += 'PATH=(custom, no user)'
elif l.startswith('XDG_RUNTIME_DIR='):
if self['ProcEnviron']:
self['ProcEnviron'] += '\n'
self['ProcEnviron'] += 'XDG_RUNTIME_DIR=<set>'
elif l.startswith('LD_PRELOAD='):
if self['ProcEnviron']:
self['ProcEnviron'] += '\n'
self['ProcEnviron'] += 'LD_PRELOAD=<set>'
elif l.startswith('LD_LIBRARY_PATH='):
if self['ProcEnviron']:
self['ProcEnviron'] += '\n'
self['ProcEnviron'] += 'LD_LIBRARY_PATH=<set>'
elif l.startswith('XDG_CURRENT_DESKTOP='):
self['CurrentDesktop'] = l.split('=', 1)[1]
def add_kernel_crash_info(self, debugdir=None):
'''Add information from kernel crash.
This needs a VmCore in the Report.
'''
if 'VmCore' not in self:
return
unlink_core = False
ret = False
try:
if hasattr(self['VmCore'], 'find'):
(fd, core) = tempfile.mkstemp()
os.write(fd, self['VmCore'])
os.close(fd)
unlink_core = True
kver = self['Uname'].split()[1]
command = ['crash',
'/usr/lib/debug/boot/vmlinux-%s' % kver,
core,
]
try:
p = subprocess.Popen(command,
stdin=subprocess.PIPE,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT)
except OSError:
return False
p.stdin.write('bt -a -f\n')
p.stdin.write('ps\n')
p.stdin.write('runq\n')
p.stdin.write('quit\n')
# FIXME: split it up nicely etc
out = p.stdout.read()
ret = (p.wait() == 0)
if ret:
self['Stacktrace'] = out
finally:
if unlink_core:
os.unlink(core)
return ret
def add_gdb_info(self, rootdir=None):
'''Add information from gdb.
This requires that the report has a CoreDump and an
ExecutablePath. This adds the following fields:
- Registers: Output of gdb's 'info registers' command
- Disassembly: Output of gdb's 'x/16i $pc' command
- Stacktrace: Output of gdb's 'bt full' command
- ThreadStacktrace: Output of gdb's 'thread apply all bt full' command
- StacktraceTop: simplified stacktrace (topmost 5 functions) for inline
inclusion into bug reports and easier processing
- AssertionMessage: Value of __abort_msg, __glib_assert_msg, or
__nih_abort_msg if present
The optional rootdir can specify a root directory which has the
executable, libraries, and debug symbols. This does not require
chroot() or root privileges, it just instructs gdb to search for the
files there.
'''
if 'CoreDump' not in self or 'ExecutablePath' not in self:
return
gdb_reports = {'Registers': 'info registers',
'Disassembly': 'x/16i $pc',
'Stacktrace': 'bt full',
'ThreadStacktrace': 'thread apply all bt full',
'AssertionMessage': 'print __abort_msg->msg',
'GLibAssertionMessage': 'print __glib_assert_msg',
'NihAssertionMessage': 'print (char*) __nih_abort_msg'}
gdb_cmd = self.gdb_command(rootdir)
# limit maximum backtrace depth (to avoid looped stacks)
gdb_cmd += ['--batch', '--ex', 'set backtrace limit 2000']
value_keys = []
# append the actual commands and something that acts as a separator
for name, cmd in gdb_reports.items():
value_keys.append(name)
gdb_cmd += ['--ex', 'p -99', '--ex', cmd]
# call gdb
try:
out = _command_output(gdb_cmd).decode('UTF-8', errors='replace')
except OSError:
return
# split the output into the various fields
part_re = re.compile('^\$\d+\s*=\s*-99$', re.MULTILINE)
parts = part_re.split(out)
# drop the gdb startup text prior to first separator
parts.pop(0)
for part in parts:
self[value_keys.pop(0)] = part.replace('\n\n', '\n.\n').strip()
# glib's assertion has precedence, since it internally uses
# abort(), and then glib's __abort_msg is bogus
if 'GLibAssertionMessage' in self:
if '"ERROR:' in self['GLibAssertionMessage']:
self['AssertionMessage'] = self['GLibAssertionMessage']
del self['GLibAssertionMessage']
# same reason for libnih's assertion messages
if 'NihAssertionMessage' in self:
if self['NihAssertionMessage'].startswith('$'):
self['AssertionMessage'] = self['NihAssertionMessage']
del self['NihAssertionMessage']
# clean up AssertionMessage
if 'AssertionMessage' in self:
# chop off "$n = 0x...." prefix, drop empty ones
m = re.match('^\$\d+\s+=\s+0x[0-9a-fA-F]+\s+"(.*)"\s*$',
self['AssertionMessage'])
if m:
self['AssertionMessage'] = m.group(1)
if self['AssertionMessage'].endswith('\\n'):
self['AssertionMessage'] = self['AssertionMessage'][0:-2]
else:
del self['AssertionMessage']
if 'Stacktrace' in self:
self._gen_stacktrace_top()
addr_signature = self.crash_signature_addresses()
if addr_signature:
self['StacktraceAddressSignature'] = addr_signature
def _gen_stacktrace_top(self):
'''Build field StacktraceTop as the top five functions of Stacktrace.
Signal handler invocations and related functions are skipped since they
are generally not useful for triaging and duplicate detection.
'''
unwind_functions = set(['g_logv', 'g_log', 'IA__g_log', 'IA__g_logv',
'g_assert_warning', 'IA__g_assert_warning',
'__GI_abort', '_XError'])
toptrace = [''] * 5
depth = 0
unwound = False
unwinding = False
unwinding_xerror = False
bt_fn_re = re.compile('^#(\d+)\s+(?:0x(?:\w+)\s+in\s+\*?(.*)|(<signal handler called>)\s*)$')
bt_fn_noaddr_re = re.compile('^#(\d+)\s+(?:(.*)|(<signal handler called>)\s*)$')
# some internal functions like the SSE stubs cause unnecessary jitter
ignore_functions_re = re.compile('^(__.*_s?sse\d+(?:_\w+)?|__kernel_vsyscall)$')
for line in self['Stacktrace'].splitlines():
m = bt_fn_re.match(line)
if not m:
m = bt_fn_noaddr_re.match(line)
if not m:
continue
if not unwound or unwinding:
if m.group(2):
fn = m.group(2).split()[0].split('(')[0]
else:
fn = None
# handle XErrors
if unwinding_xerror:
if fn.startswith('_X') or fn in ['handle_response', 'handle_error', 'XWindowEvent']:
continue
else:
unwinding_xerror = False
if m.group(3) or fn in unwind_functions:
unwinding = True
depth = 0
toptrace = [''] * 5
if m.group(3):
# we stop unwinding when we found a <signal handler>,
# but we continue unwinding otherwise, as e. g. a glib
# abort is usually sitting on top of an XError
unwound = True
if fn == '_XError':
unwinding_xerror = True
continue
else:
unwinding = False
frame = m.group(2) or m.group(3)
function = frame.split()[0]
if depth < len(toptrace) and not ignore_functions_re.match(function):
toptrace[depth] = frame
depth += 1
self['StacktraceTop'] = '\n'.join(toptrace).strip()
def add_hooks_info(self, ui, package=None, srcpackage=None):
'''Run hook script for collecting package specific data.
A hook script needs to be in _hook_dir/<Package>.py or in
_common_hook_dir/*.py and has to contain a function 'add_info(report,
ui)' that takes and modifies a Report, and gets an UserInterface
reference for interactivity.
return True if the hook requested to stop the report filing process,
False otherwise.
'''
# determine package names, unless already given as arguments
if not package:
package = self.get('Package')
if package:
package = package.split()[0]
if not srcpackage:
srcpackage = self.get('SourcePackage')
if srcpackage:
srcpackage = srcpackage.split()[0]
hook_dirs = [_hook_dir]
# also search hooks in /opt, when program is from there
opt_path = None
if self.get('ExecutablePath', '').startswith(_opt_dir):
opt_path = self.get('ExecutablePath', '')
elif package:
# check package contents
try:
for f in apport.packaging.get_files(package):
if f.startswith(_opt_dir) and os.path.isfile(f):
opt_path = f
break
except ValueError:
# uninstalled package
pass
if opt_path:
while len(opt_path) >= len(_opt_dir):
hook_dirs.append(os.path.join(opt_path, 'share', 'apport', 'package-hooks'))
opt_path = os.path.dirname(opt_path)
# common hooks
for hook in glob.glob(_common_hook_dir + '/*.py'):
if _run_hook(self, ui, hook):
return True
# binary package hook
if package:
for hook_dir in hook_dirs:
if _run_hook(self, ui, os.path.join(hook_dir, package + '.py')):
return True
# source package hook
if srcpackage:
for hook_dir in hook_dirs:
if _run_hook(self, ui, os.path.join(hook_dir, 'source_%s.py' % srcpackage)):
return True
return False
def search_bug_patterns(self, url):
'''Check bug patterns loaded from the specified url.
Return bug URL on match, or None otherwise.
The url must refer to a valid XML document with the following syntax:
root element := <patterns>
patterns := <pattern url="http://bug.url"> *
pattern := <re key="report_key">regular expression*</re> +
For example:
<?xml version="1.0"?>
<patterns>
<pattern url="http://bugtracker.net/bugs/1">
<re key="Foo">ba.*r</re>
</pattern>
<pattern url="http://bugtracker.net/bugs/2">
<re key="Package">^\S* 1-2$</re> <!-- test for a particular version -->
<re key="Foo">write_(hello|goodbye)</re>
</pattern>
</patterns>
'''
# some distros might not want to support these
if not url:
return
try:
f = urlopen(url)
patterns = f.read().decode('UTF-8', errors='replace')
f.close()
except (IOError, URLError):
# doesn't exist or failed to load
return
if '<title>404 Not Found' in patterns:
return
url = _check_bug_patterns(self, patterns)
if url:
return url
return None
def _get_ignore_dom(self):
'''Read ignore list XML file and return a DOM tree.
Return an empty DOM tree if file does not exist.
Raises ValueError if the file exists but is invalid XML.
'''
orig_home = os.getenv('HOME')
if orig_home is not None:
del os.environ['HOME']
ifpath = os.path.expanduser(_ignore_file)
if orig_home is not None:
os.environ['HOME'] = orig_home
if not os.access(ifpath, os.R_OK) or os.path.getsize(ifpath) == 0:
# create a document from scratch
dom = xml.dom.getDOMImplementation().createDocument(None, 'apport', None)
else:
try:
dom = xml.dom.minidom.parse(ifpath)
except ExpatError as e:
raise ValueError('%s has invalid format: %s' % (_ignore_file, str(e)))
# remove whitespace so that writing back the XML does not accumulate
# whitespace
dom.documentElement.normalize()
_dom_remove_space(dom.documentElement)
return dom
def check_ignored(self):
'''Check if current report should not be presented.
Reports can be suppressed by per-user blacklisting in
~/.apport-ignore.xml (in the real UID's home) and
/etc/apport/blacklist.d/. For environments where you are only
interested in crashes of some programs, you can also create a whitelist
in /etc/apport/whitelist.d/, everything which does not match gets
ignored then.
This requires the ExecutablePath attribute. Throws a ValueError if the
file has an invalid format.
'''
assert 'ExecutablePath' in self
# check blacklist
try:
for f in os.listdir(_blacklist_dir):
try:
with open(os.path.join(_blacklist_dir, f)) as fd:
for line in fd:
if line.strip() == self['ExecutablePath']:
return True
except IOError:
continue
except OSError:
pass
# check whitelist
try:
whitelist = set()
for f in os.listdir(_whitelist_dir):
try:
with open(os.path.join(_whitelist_dir, f)) as fd:
for line in fd:
whitelist.add(line.strip())
except IOError:
continue
if whitelist and self['ExecutablePath'] not in whitelist:
return True
except OSError:
pass
try:
dom = self._get_ignore_dom()
except (ValueError, KeyError):
apport.error('Could not get ignore file:')
traceback.print_exc()
return False
try:
cur_mtime = int(os.stat(self['ExecutablePath']).st_mtime)
except OSError:
# if it does not exist any more, do nothing
return False
# search for existing entry and update it
for ignore in dom.getElementsByTagName('ignore'):
if ignore.getAttribute('program') == self['ExecutablePath']:
if float(ignore.getAttribute('mtime')) >= cur_mtime:
return True
return False
def mark_ignore(self):
'''Ignore future crashes of this executable.
Add a ignore list entry for this report to ~/.apport-ignore.xml, so
that future reports for this ExecutablePath are not presented to the
user any more.
Throws a ValueError if the file already exists and has an invalid
format.
'''
assert 'ExecutablePath' in self
dom = self._get_ignore_dom()
try:
mtime = str(int(os.stat(self['ExecutablePath']).st_mtime))
except OSError as e:
# file went away underneath us, ignore
if e.errno == errno.ENOENT:
return
else:
raise
# search for existing entry and update it
for ignore in dom.getElementsByTagName('ignore'):
if ignore.getAttribute('program') == self['ExecutablePath']:
ignore.setAttribute('mtime', mtime)
break
else:
# none exists yet, create new ignore node if none exists yet
e = dom.createElement('ignore')
e.setAttribute('program', self['ExecutablePath'])
e.setAttribute('mtime', mtime)
dom.documentElement.appendChild(e)
# write back file; temporarily unset $HOME, as this gets the wrong home
# dir for e. g. sudo
orig_home = os.getenv('HOME')
if orig_home is not None:
del os.environ['HOME']
ignore_file_path = os.path.expanduser(_ignore_file)
if orig_home is not None:
os.environ['HOME'] = orig_home
with open(ignore_file_path, 'w') as fd:
dom.writexml(fd, addindent=' ', newl='\n')
dom.unlink()
def has_useful_stacktrace(self):
'''Check whether StackTrace can be considered 'useful'.
The current heuristic is to consider it useless if it either is shorter
than three lines and has any unknown function, or for longer traces, a
minority of known functions.
'''
if not self.get('StacktraceTop'):
return False
unknown_fn = [f.startswith('??') for f in self['StacktraceTop'].splitlines()]
if len(unknown_fn) < 3:
return unknown_fn.count(True) == 0
return unknown_fn.count(True) <= len(unknown_fn) / 2.
def stacktrace_top_function(self):
'''Return topmost function in StacktraceTop'''
for l in self.get('StacktraceTop', '').splitlines():
fname = l.split('(')[0].strip()
if fname != '??':
return fname
return None
def standard_title(self):
'''Create an appropriate title for a crash database entry.
This contains the topmost function name from the stack trace and the
signal (for signal crashes) or the Python exception (for unhandled
Python exceptions).
Return None if the report is not a crash or a default title could not
be generated.
'''
# assertion failure
if self.get('Signal') == '6' and \
'ExecutablePath' in self and \
'AssertionMessage' in self:
return '%s assert failure: %s' % (
os.path.basename(self['ExecutablePath']),
self['AssertionMessage'])
# signal crash
if 'Signal' in self and 'ExecutablePath' in self and 'StacktraceTop' in self:
signal_names = {
'4': 'SIGILL',
'6': 'SIGABRT',
'8': 'SIGFPE',
'11': 'SIGSEGV',
'13': 'SIGPIPE'}
fn = self.stacktrace_top_function()
if fn:
fn = ' in %s()' % fn
else:
fn = ''
arch_mismatch = ''
if 'Architecture' in self and 'PackageArchitecture' in self and self['Architecture'] != self['PackageArchitecture'] and self['PackageArchitecture'] != 'all':
arch_mismatch = ' [non-native %s package]' % self['PackageArchitecture']
return '%s crashed with %s%s%s' % (
os.path.basename(self['ExecutablePath']),
signal_names.get(self.get('Signal'), 'signal ' + self.get('Signal')),
fn, arch_mismatch
)
# Python exception
if 'Traceback' in self and 'ExecutablePath' in self:
trace = self['Traceback'].splitlines()
if len(trace) < 1:
return None
if len(trace) < 3:
return '%s crashed with %s' % (
os.path.basename(self['ExecutablePath']),
trace[0])
trace_re = re.compile('^\s*File\s*"(\S+)".* in (.+)$')
i = len(trace) - 1
function = 'unknown'
while i >= 0:
m = trace_re.match(trace[i])
if m:
module_path = m.group(1)
function = m.group(2)
break
i -= 1
path = os.path.basename(self['ExecutablePath'])
last_line = trace[-1]
exception = last_line.split(':')[0]
m = re.match('^%s: (.+)$' % re.escape(exception), last_line)
if m:
message = m.group(1)
else:
message = None
if function == '<module>':
if module_path == self['ExecutablePath']:
context = '__main__'
else:
# Maybe use os.path.basename?
context = module_path
else:
context = '%s()' % function
title = '%s crashed with %s in %s' % (
path,
exception,
context
)
if message:
title += ': %s' % message
return title
# package problem
if self.get('ProblemType') == 'Package' and 'Package' in self:
title = 'package %s failed to install/upgrade' % \
self['Package']
if self.get('ErrorMessage'):
title += ': ' + self['ErrorMessage'].splitlines()[-1]
return title
if self.get('ProblemType') == 'KernelOops' and 'OopsText' in self:
oops = self['OopsText']
if oops.startswith('------------[ cut here ]------------'):
title = oops.split('\n', 2)[1]
else:
title = oops.split('\n', 1)[0]
return title
if self.get('ProblemType') == 'KernelOops' and 'Failure' in self:
# Title the report with suspend or hibernate as appropriate,
# and mention any non-free modules loaded up front.
title = ''
if 'MachineType' in self:
title += '[' + self['MachineType'] + '] '
title += self['Failure'] + ' failure'
if 'NonfreeKernelModules' in self:
title += ' [non-free: ' + self['NonfreeKernelModules'] + ']'
title += '\n'
return title
return None
def obsolete_packages(self):
'''Return list of obsolete packages in Package and Dependencies.'''
obsolete = []
for l in (self.get('Package', '') + '\n' + self.get('Dependencies', '')).splitlines():
if not l:
continue
pkg, ver = l.split()[:2]
avail = packaging.get_available_version(pkg)
if ver is not None and ver != 'None' and avail is not None and packaging.compare_versions(ver, avail) < 0:
obsolete.append(pkg)
return obsolete
def crash_signature(self):
'''Get a signature string for a crash.
This is suitable for identifying duplicates.
For signal crashes this the concatenation of ExecutablePath, Signal
number, and StacktraceTop function names, separated by a colon. If
StacktraceTop has unknown functions or the report lacks any of those
fields, return None. In this case, you can use
crash_signature_addresses() to get a less precise duplicate signature
based on addresses instead of symbol names.
For assertion failures, it is the concatenation of ExecutablePath
and assertion message, separated by colons.
For Python crashes, this concatenates the ExecutablePath, exception
name, and Traceback function names, again separated by a colon.
For suspend/resume failures, this concatenates whether it was a suspend
or resume failure with the hardware identifier and the BIOS version, if
it exists.
'''
if 'ExecutablePath' not in self:
if not self['ProblemType'] in ('KernelCrash', 'KernelOops'):
return None
# kernel crash
if 'Stacktrace' in self and self['ProblemType'] == 'KernelCrash':
sig = 'kernel'
regex = re.compile('^\s*\#\d+\s\[\w+\]\s(\w+)')
for line in self['Stacktrace'].splitlines():
m = regex.match(line)
if m:
sig += ':' + (m.group(1))
return sig
# assertion failures
if self.get('Signal') == '6' and 'AssertionMessage' in self:
sig = self['ExecutablePath'] + ':' + self['AssertionMessage']
# filter out addresses, to help match duplicates more sanely
return re.sub(r'0x[0-9a-f]{6,}', 'ADDR', sig)
# signal crashes
if 'StacktraceTop' in self and 'Signal' in self:
sig = '%s:%s' % (self['ExecutablePath'], self['Signal'])
bt_fn_re = re.compile('^(?:([\w:~]+).*|(<signal handler called>)\s*)$')
lines = self['StacktraceTop'].splitlines()
if len(lines) < 2:
return None
for line in lines:
m = bt_fn_re.match(line)
if m:
sig += ':' + (m.group(1) or m.group(2))
else:
# this will also catch ??
return None
return sig
# Python crashes
if 'Traceback' in self:
trace = self['Traceback'].splitlines()
sig = ''
if len(trace) == 1:
# sometimes, Python exceptions do not have file references
m = re.match('(\w+): ', trace[0])
if m:
return self['ExecutablePath'] + ':' + m.group(1)
else:
return None
elif len(trace) < 3:
return None
loc_re = re.compile('^\s+File "([^"]+).*line (\d+).*\sin (.*)$')
for l in trace:
m = loc_re.match(l)
if m:
# if we have a function name, use this; for a a crash
# outside of a function/method, fall back to the source
# file location
if m.group(3) != '<module>':
sig += ':' + m.group(3)
else:
# resolve symlinks for more stable signatures
f = m.group(1)
if os.path.islink(f):
f = os.path.realpath(f)
sig += ':%s@%s' % (f, m.group(2))
return self['ExecutablePath'] + ':' + trace[-1].split(':')[0] + sig
if self['ProblemType'] == 'KernelOops' and 'Failure' in self:
if 'suspend' in self['Failure'] or 'resume' in self['Failure']:
# Suspend / resume failure
sig = self['Failure']
if self.get('MachineType'):
sig += ':%s' % self['MachineType']
if self.get('dmi.bios.version'):
sig += ':%s' % self['dmi.bios.version']
return sig
# KernelOops crashes
if 'OopsText' in self:
in_trace_body = False
parts = []
for line in self['OopsText'].split('\n'):
if line.startswith('BUG: unable to handle'):
parsed = re.search('^BUG: unable to handle (.*) at ', line)
if parsed:
match = parsed.group(1)
assert match, 'could not parse expected problem type line: %s' % line
parts.append(match)
if line.startswith('IP: '):
match = self._extract_function_and_address(line)
if match:
parts.append(match)
elif line.startswith('Call Trace:'):
in_trace_body = True
elif in_trace_body:
match = None
if line and line[0] == ' ':
match = self._extract_function_and_address(line)
if match:
parts.append(match)
else:
in_trace_body = False
if parts:
return ':'.join(parts)
return None
def _extract_function_and_address(self, line):
parsed = re.search('\[.*\] (.*)$', line)
if parsed:
match = parsed.group(1)
assert match, 'could not parse expected call trace line: %s' % line
if match[0] != '?':
return match
return None
def crash_signature_addresses(self):
'''Compute heuristic duplicate signature for a signal crash.
This should be used if crash_signature() fails, i. e. Stacktrace does
not have enough symbols.
This approach only uses addresses in the stack trace and does not rely
on symbol resolution. As we can't unwind these stack traces, we cannot
limit them to the top five frames and thus will end up with several or
many different signatures for a particular crash. But these can be
computed and synchronously checked with a crash database at the client
side, which avoids having to upload and process the full report. So on
the server-side crash database we will only have to deal with all the
equivalence classes (i. e. same crash producing a number of possible
signatures) instead of every single report.
Return None when signature cannot be determined.
'''
if not 'ProcMaps' in self or not 'Stacktrace' in self or not 'Signal' in self:
return None
stack = []
failed = 0
for line in self['Stacktrace'].splitlines():
if line.startswith('#'):
addr = line.split()[1]
if not addr.startswith('0x'):
continue
addr = int(addr, 16) # we do want to know about ValueErrors here, so don't catch
# ignore impossibly low addresses; these are usually artifacts
# from gdb when not having debug symbols
if addr < 0x1000:
continue
offset = self._address_to_offset(addr)
if offset:
# avoid ':' in ELF paths, we use that as separator
stack.append(offset.replace(':', '..'))
else:
failed += 1
# stack unwinding chops off ~ 5 functions, and we need some more
# accuracy because we do not have symbols; but beyond a depth of 15
# we get too much noise, so we can abort there
if len(stack) >= 15:
break
# we only accept a small minority (< 20%) of failed resolutions, otherwise we
# discard
if failed > 0 and len(stack) / failed < 4:
return None
# we also discard if the trace is too short
if (failed == 0 and len(stack) < 3) or (failed > 0 and len(stack) < 6):
return None
return '%s:%s:%s:%s' % (
self['ExecutablePath'],
self['Signal'],
os.uname()[4],
':'.join(stack))
def anonymize(self):
'''Remove user identifying strings from the report.
This particularly removes the user name, host name, and IPs
from attributes which contain data read from the environment, and
removes the ProcCwd attribute completely.
'''
replacements = []
if (os.getuid() > 0):
# do not replace "root"
p = pwd.getpwuid(os.getuid())
if len(p[0]) >= 2:
replacements.append((re.compile(r'\b%s\b' % re.escape(p[0])), 'username'))
replacements.append((re.compile(r'\b%s\b' % re.escape(p[5])), '/home/username'))
for s in p[4].split(','):
s = s.strip()
if len(s) > 2:
replacements.append((re.compile(r'(\b|\s)%s\b' % re.escape(s)), r'\1User Name'))
hostname = os.uname()[1]
if len(hostname) >= 2:
replacements.append((re.compile(r'\b%s\b' % re.escape(hostname)), 'hostname'))
try:
del self['ProcCwd']
except KeyError:
pass
for k in self:
is_proc_field = k.startswith('Proc') and not k in [
'ProcCpuinfo', 'ProcMaps', 'ProcStatus', 'ProcInterrupts', 'ProcModules']
if is_proc_field or 'Stacktrace' in k or k in ['Traceback', 'PythonArgs', 'Title']:
if not hasattr(self[k], 'isspace'):
continue
for (pattern, repl) in replacements:
if type(self[k]) == bytes:
self[k] = pattern.sub(repl, self[k].decode('UTF-8', errors='replace')).encode('UTF-8')
else:
self[k] = pattern.sub(repl, self[k])
def gdb_command(self, sandbox):
'''Build gdb command for this report.
This builds a gdb command for processing the given report, by setting
the file to the ExecutablePath/InterpreterPath, unpacking the core dump
and pointing "core-file" to it (if the report has a core dump), and
setting up the paths when calling gdb in a package sandbox.
When available, this calls "gdb-multiarch" instead of "gdb", for
processing crash reports from foreign architectures.
Return argv list.
'''
assert 'ExecutablePath' in self
executable = self.get('InterpreterPath', self['ExecutablePath'])
command = ['gdb']
if 'Architecture' in self and self['Architecture'] != packaging.get_system_architecture():
# check if we have gdb-multiarch
which = subprocess.Popen(['which', 'gdb-multiarch'],
stdout=subprocess.PIPE)
which.communicate()
if which.returncode == 0:
command = ['gdb-multiarch']
else:
sys.stderr.write(
'WARNING: Please install gdb-multiarch for processing '
'reports from foreign architectures. Results with "gdb" '
'will be very poor.\n')
# check for foreign architecture
arch = self.get('Uname', 'none').split()[-1]
if 'arm' in arch:
command += ['--ex', 'set architecture arm', '--ex', 'set gnutarget elf32-littlearm']
elif 'ppc' in arch:
command += ['--ex', 'set architecture powerpc:common', '--ex', 'set gnutarget elf32-powerpc']
# note, i386 vs. x86_64 is auto-detected just fine
if sandbox:
command += ['--ex', 'set debug-file-directory %s/usr/lib/debug' % sandbox,
'--ex', 'set solib-absolute-prefix ' + sandbox]
executable = sandbox + '/' + executable
assert os.path.exists(executable)
command += ['--ex', 'file "%s"' % executable]
if 'CoreDump' in self:
if hasattr(self['CoreDump'], 'find'):
(fd, core) = tempfile.mkstemp(prefix='apport_core_')
atexit.register(os.unlink, core)
os.write(fd, self['CoreDump'])
os.close(fd)
elif hasattr(self['CoreDump'], 'gzipvalue'):
(fd, core) = tempfile.mkstemp(prefix='apport_core_')
atexit.register(os.unlink, core)
os.close(fd)
with open(core, 'wb') as f:
self['CoreDump'].write(f)
else:
# value is a file path
core = self['CoreDump'][0]
command += ['--ex', 'core-file ' + core]
return command
def _address_to_offset(self, addr):
'''Resolve a memory address to an ELF name and offset.
This can be used for building duplicate signatures from non-symbolic
stack traces. These often do not have enough symbols available to
resolve function names, but taking the raw addresses also is not
suitable due to ASLR. But the offsets within a library should be
constant between crashes (assuming the same version of all libraries).
This needs and uses the "ProcMaps" field to resolve addresses.
Return 'path+offset' when found, or None if address is not in any
mapped range.
'''
self._build_proc_maps_cache()
for (start, end, elf) in self._proc_maps_cache:
if start <= addr and end >= addr:
return '%s+%x' % (elf, addr - start)
return None
def _build_proc_maps_cache(self):
'''Generate self._proc_maps_cache from ProcMaps field.
This only gets done once.
'''
if self._proc_maps_cache:
return
assert 'ProcMaps' in self
self._proc_maps_cache = []
# library paths might have spaces, so we need to make some assumptions
# about the intermediate fields. But we know that in between the pre-last
# data field and the path there are many spaces, while between the
# other data fields there is only one. So we take 2 or more spaces as
# the separator of the last data field and the path.
fmt = re.compile('^([0-9a-fA-F]+)-([0-9a-fA-F]+).*\s{2,}(\S.*$)')
fmt_unknown = re.compile('^([0-9a-fA-F]+)-([0-9a-fA-F]+)\s')
for line in self['ProcMaps'].splitlines():
if not line.strip():
continue
m = fmt.match(line)
if not m:
# ignore lines with unknown ELF
if fmt_unknown.match(line):
continue
# but complain otherwise, as this means we encounter an
# architecture or new kernel version where the format changed
assert m, 'cannot parse ProcMaps line: ' + line
self._proc_maps_cache.append((int(m.group(1), 16),
int(m.group(2), 16), m.group(3)))
@classmethod
def get_logind_session(klass, pid):
'''Get logind session path and start time.
Return (path, session_start_timestamp) if pid is in a logind session,
or None otherwise.
'''
# determine cgroup
try:
with open('/proc/%s/cgroup' % pid) as f:
for l in f:
if 'name=systemd:' in l:
my_cgroup = l.split('systemd:', 1)[1].strip()
if len(my_cgroup) < 2:
return None
break
else:
return None
# determine cgroup creation time
session_start_time = os.stat('/sys/fs/cgroup/systemd/' + my_cgroup).st_mtime
except (IOError, OSError):
return None
return (my_cgroup, session_start_time)
def get_timestamp(self):
'''Get timestamp (seconds since epoch) from Date field
Return None if it is not present.
'''
# report time is from asctime(), not in locale representation
orig_ctime = locale.getlocale(locale.LC_TIME)
try:
try:
locale.setlocale(locale.LC_TIME, 'C')
return time.mktime(time.strptime(self['Date']))
except KeyError:
return None
finally:
locale.setlocale(locale.LC_TIME, orig_ctime)
except locale.Error:
return None
|