diff options
Diffstat (limited to 'build/valgrind')
| -rw-r--r-- | build/valgrind/__init__.py | 0 | ||||
| -rw-r--r-- | build/valgrind/cross-architecture.sup | 173 | ||||
| -rw-r--r-- | build/valgrind/i386-pc-linux-gnu.sup | 53 | ||||
| -rw-r--r-- | build/valgrind/mach_commands.py | 243 | ||||
| -rw-r--r-- | build/valgrind/output_handler.py | 119 | ||||
| -rw-r--r-- | build/valgrind/x86_64-pc-linux-gnu.sup | 1008 |
6 files changed, 1596 insertions, 0 deletions
diff --git a/build/valgrind/__init__.py b/build/valgrind/__init__.py new file mode 100644 index 0000000000..e69de29bb2 --- /dev/null +++ b/build/valgrind/__init__.py diff --git a/build/valgrind/cross-architecture.sup b/build/valgrind/cross-architecture.sup new file mode 100644 index 0000000000..d6b9adcf95 --- /dev/null +++ b/build/valgrind/cross-architecture.sup @@ -0,0 +1,173 @@ +# Full list is tracked through meta bug 793882 + +#################### +# Intended leaks # +#################### + +{ + PR_SetEnv requires its argument to be leaked, but does not appear on stacks. (See bug 793534.) + Memcheck:Leak + ... + fun:_ZN7mozilla9SaveToEnvEPKc + ... +} +{ + PR_SetEnv requires its argument to be leaked, but does not appear on stacks. (See bug 793534.) + Memcheck:Leak + ... + fun:SaveToEnv + ... +} +{ + PR_SetEnv requires its argument to be leaked, but does not appear on stacks. (See bug 793549.) + Memcheck:Leak + ... + fun:_ZL13SaveWordToEnvPKcRK12nsTSubstringIcE + ... +} +{ + PR_SetEnv requires its argument to be leaked, but does not appear on stacks. (See bug 793549.) + Memcheck:Leak + ... + fun:SaveWordToEnv + ... +} +{ + PR_SetEnv requires its argument to be leaked, but does not appear on stacks. (See bug 944133.) + Memcheck:Leak + ... + fun:_ZN13CrashReporter14SetRestartArgsEiPPc + ... +} +{ + PR_SetEnv requires its argument to be leaked, but does not appear on stacks. (See bug 793548.) + Memcheck:Leak + fun:malloc + ... + fun:_Z12ToNewCStringRK10nsACString + fun:_ZN13CrashReporter14SetupExtraDataEP7nsIFileRK10nsACString + ... +} +{ + PR_SetEnv requires its argument to be leaked, but does not appear on stacks. (See bug 1345153.) + Memcheck:Leak + fun:malloc + ... + fun:_ZN13CrashReporterL27SetupCrashReporterDirectoryEP7nsIFilePKcS3_PS1_ + ... +} +{ + We purposely leak the StatisticsReporter object + Memcheck:Leak + fun:malloc + fun:moz_xmalloc + fun:operator new + fun:_Z21XRE_CreateStatsObjectv + ... +} + +#################################### +# Leaks in third party libraries # +#################################### + +{ + See bug 793535 + Memcheck:Leak + fun:calloc + fun:_dlerror_run + fun:dlsym + ... +} +{ + See bug 793611 + Memcheck:Leak + fun:memalign + fun:tls_get_addr_tail + fun:__tls_get_addr + ... +} +{ + See bug 793611 + Memcheck:Leak + fun:memalign + fun:tls_get_addr_tail + fun:___tls_get_addr + ... +} + +################# +# Other leaks # +################# + +{ + Bug 794369 + Memcheck:Leak + fun:malloc + fun:moz_xmalloc + ... + fun:_ZN20mozJSComponentLoader10LoadModuleERN7mozilla12FileLocationE + ... +} +{ + Bug 794370 + Memcheck:Leak + fun:malloc + fun:moz_xmalloc + ... + fun:_ZN22nsComponentManagerImpl15RegisterFactoryERK4nsIDPKcS4_P10nsIFactory + ... +} +{ + Bug 794372 + Memcheck:Leak + fun:malloc + fun:moz_xmalloc + ... + fun:_ZN22nsComponentManagerImpl22RegisterCIDEntryLockedEPKN7mozilla6Module8CIDEntryEPNS_11KnownModuleE + ... +} +{ + Bug 794374 + Memcheck:Leak + fun:malloc + fun:moz_xmalloc + ... + fun:_ZN22nsComponentManagerImpl17ManifestComponentERNS_25ManifestProcessingContextEiPKPc + ... +} +{ + Bug 1017112 + Memcheck:Leak + fun:malloc + ... + fun:PK11_InitPin + fun:_ZN11nsPK11Token12InitPasswordEPKDs + ... +} +{ + We intentionally leak TLS data in the global servo thread-pool until we can free it consistently (https://github.com/rayon-rs/rayon/issues/688) + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:_ZN3std6thread5local4fast12Key$LT$T$GT$14try_initialize* + fun:_ZN5style5bloom19StyleBloom$LT$E$GT$3new* + ... +} +{ + We intentionally leak TLS data in the global servo thread-pool until we can free it consistently (https://github.com/rayon-rs/rayon/issues/688) + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:_ZN3std6thread5local4fast12Key$LT$T$GT$14try_initialize* + ... + fun:_ZN5style7sharing26StyleSharingCache$LT$E$GT$3new* + ... +} +{ + Leak in libfontconfig1 in Debian 8 and 9. See bug 1636003. + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:FcGetPrgname + ... +} diff --git a/build/valgrind/i386-pc-linux-gnu.sup b/build/valgrind/i386-pc-linux-gnu.sup new file mode 100644 index 0000000000..cfdc37bb00 --- /dev/null +++ b/build/valgrind/i386-pc-linux-gnu.sup @@ -0,0 +1,53 @@ +# Full list is tracked through meta bug 793882 + +#################################### +# Leaks in third party libraries # +#################################### + +{ + Bug 793537 + Memcheck:Leak + ... + obj:/usr/lib/libpango-1.0.so.0.2800.1 + ... +} +{ + Bug 793598 + Memcheck:Leak + ... + obj:/lib/libdbus-1.so.3.4.0 + ... +} +{ + Bug 794366 + Memcheck:Leak + ... + obj:/usr/lib/libgtk-x11-2.0.so.0.1800.9 + ... +} +{ + Bug 794368 + Memcheck:Leak + ... + obj:/usr/lib/libXrandr.so.2.2.0 + ... +} +{ + Bug 794373 + Memcheck:Leak + ... + obj:/lib/libgobject-2.0.so.0.2200.5 + ... +} + +################################### +# Leaks in short lived precesses # +################################### + +{ + Bug 984196 + Memcheck:Leak + ... + fun:glxtest + ... +} diff --git a/build/valgrind/mach_commands.py b/build/valgrind/mach_commands.py new file mode 100644 index 0000000000..408c0be28c --- /dev/null +++ b/build/valgrind/mach_commands.py @@ -0,0 +1,243 @@ +# This Source Code Form is subject to the terms of the Mozilla Public +# License, v. 2.0. If a copy of the MPL was not distributed with this +# file, You can obtain one at http://mozilla.org/MPL/2.0/. + +from __future__ import absolute_import, unicode_literals + +import json +import logging +import mozinfo +import os + +from mach.decorators import ( + Command, + CommandArgument, + CommandProvider, +) +from mozbuild.base import ( + MachCommandBase, + MachCommandConditions as conditions, + BinaryNotFoundException, +) + + +def is_valgrind_build(cls): + """Must be a build with --enable-valgrind and --disable-jemalloc.""" + defines = cls.config_environment.defines + return "MOZ_VALGRIND" in defines and "MOZ_MEMORY" not in defines + + +@CommandProvider +class MachCommands(MachCommandBase): + """ + Run Valgrind tests. + """ + + @Command( + "valgrind-test", + category="testing", + conditions=[conditions.is_firefox_or_thunderbird, is_valgrind_build], + description="Run the Valgrind test job (memory-related errors).", + ) + @CommandArgument( + "--suppressions", + default=[], + action="append", + metavar="FILENAME", + help="Specify a suppression file for Valgrind to use. Use " + "--suppression multiple times to specify multiple suppression " + "files.", + ) + def valgrind_test(self, suppressions): + + from mozfile import TemporaryDirectory + from mozhttpd import MozHttpd + from mozprofile import FirefoxProfile, Preferences + from mozprofile.permissions import ServerLocations + from mozrunner import FirefoxRunner + from mozrunner.utils import findInPath + from six import string_types + from valgrind.output_handler import OutputHandler + + build_dir = os.path.join(self.topsrcdir, "build") + + # XXX: currently we just use the PGO inputs for Valgrind runs. This may + # change in the future. + httpd = MozHttpd(docroot=os.path.join(build_dir, "pgo")) + httpd.start(block=False) + + with TemporaryDirectory() as profilePath: + # TODO: refactor this into mozprofile + profile_data_dir = os.path.join(self.topsrcdir, "testing", "profiles") + with open(os.path.join(profile_data_dir, "profiles.json"), "r") as fh: + base_profiles = json.load(fh)["valgrind"] + + prefpaths = [ + os.path.join(profile_data_dir, profile, "user.js") + for profile in base_profiles + ] + prefs = {} + for path in prefpaths: + prefs.update(Preferences.read_prefs(path)) + + interpolation = { + "server": "%s:%d" % httpd.httpd.server_address, + } + for k, v in prefs.items(): + if isinstance(v, string_types): + v = v.format(**interpolation) + prefs[k] = Preferences.cast(v) + + quitter = os.path.join( + self.topsrcdir, "tools", "quitter", "quitter@mozilla.org.xpi" + ) + + locations = ServerLocations() + locations.add_host( + host="127.0.0.1", port=httpd.httpd.server_port, options="primary" + ) + + profile = FirefoxProfile( + profile=profilePath, + preferences=prefs, + addons=[quitter], + locations=locations, + ) + + firefox_args = [httpd.get_url()] + + env = os.environ.copy() + env["G_SLICE"] = "always-malloc" + env["MOZ_CC_RUN_DURING_SHUTDOWN"] = "1" + env["MOZ_CRASHREPORTER_NO_REPORT"] = "1" + env["MOZ_DISABLE_NONLOCAL_CONNECTIONS"] = "1" + env["XPCOM_DEBUG_BREAK"] = "warn" + + outputHandler = OutputHandler(self.log) + kp_kwargs = { + "processOutputLine": [outputHandler], + "universal_newlines": True, + } + + valgrind = "valgrind" + if not os.path.exists(valgrind): + valgrind = findInPath(valgrind) + + valgrind_args = [ + valgrind, + "--sym-offsets=yes", + "--smc-check=all-non-file", + "--vex-iropt-register-updates=allregs-at-mem-access", + "--gen-suppressions=all", + "--num-callers=36", + "--leak-check=full", + "--show-possibly-lost=no", + "--track-origins=yes", + "--trace-children=yes", + "-v", # Enable verbosity to get the list of used suppressions + # Avoid excessive delays in the presence of spinlocks. + # See bug 1309851. + "--fair-sched=yes", + # Keep debuginfo after library unmap. See bug 1382280. + "--keep-debuginfo=yes", + # Reduce noise level on rustc and/or LLVM compiled code. + # See bug 1365915 + "--expensive-definedness-checks=yes", + # Compensate for the compiler inlining `new` but not `delete` + # or vice versa. + "--show-mismatched-frees=no", + ] + + for s in suppressions: + valgrind_args.append("--suppressions=" + s) + + supps_dir = os.path.join(build_dir, "valgrind") + supps_file1 = os.path.join(supps_dir, "cross-architecture.sup") + valgrind_args.append("--suppressions=" + supps_file1) + + if mozinfo.os == "linux": + machtype = { + "x86_64": "x86_64-pc-linux-gnu", + "x86": "i386-pc-linux-gnu", + }.get(mozinfo.processor) + if machtype: + supps_file2 = os.path.join(supps_dir, machtype + ".sup") + if os.path.isfile(supps_file2): + valgrind_args.append("--suppressions=" + supps_file2) + + exitcode = None + timeout = 1800 + binary_not_found_exception = None + try: + runner = FirefoxRunner( + profile=profile, + binary=self.get_binary_path(), + cmdargs=firefox_args, + env=env, + process_args=kp_kwargs, + ) + runner.start(debug_args=valgrind_args) + exitcode = runner.wait(timeout=timeout) + except BinaryNotFoundException as e: + binary_not_found_exception = e + finally: + errs = outputHandler.error_count + supps = outputHandler.suppression_count + if errs != supps: + status = 1 # turns the TBPL job orange + self.log( + logging.ERROR, + "valgrind-fail-parsing", + {"errs": errs, "supps": supps}, + "TEST-UNEXPECTED-FAIL | valgrind-test | error parsing: {errs} errors " + "seen, but {supps} generated suppressions seen", + ) + + elif errs == 0: + status = 0 + self.log( + logging.INFO, + "valgrind-pass", + {}, + "TEST-PASS | valgrind-test | valgrind found no errors", + ) + else: + status = 1 # turns the TBPL job orange + # We've already printed details of the errors. + + if binary_not_found_exception: + status = 2 # turns the TBPL job red + self.log( + logging.ERROR, + "valgrind-fail-errors", + {"error": str(binary_not_found_exception)}, + "TEST-UNEXPECTED-FAIL | valgrind-test | {error}", + ) + self.log( + logging.INFO, + "valgrind-fail-errors", + {"help": binary_not_found_exception.help()}, + "{help}", + ) + elif exitcode is None: + status = 2 # turns the TBPL job red + self.log( + logging.ERROR, + "valgrind-fail-timeout", + {"timeout": timeout}, + "TEST-UNEXPECTED-FAIL | valgrind-test | Valgrind timed out " + "(reached {timeout} second limit)", + ) + elif exitcode != 0: + status = 2 # turns the TBPL job red + self.log( + logging.ERROR, + "valgrind-fail-errors", + {"exitcode": exitcode}, + "TEST-UNEXPECTED-FAIL | valgrind-test | non-zero exit code " + "from Valgrind: {exitcode}", + ) + + httpd.stop() + + return status diff --git a/build/valgrind/output_handler.py b/build/valgrind/output_handler.py new file mode 100644 index 0000000000..d3a84d3444 --- /dev/null +++ b/build/valgrind/output_handler.py @@ -0,0 +1,119 @@ +# This Source Code Form is subject to the terms of the Mozilla Public +# License, v. 2.0. If a copy of the MPL was not distributed with this +# file, You can obtain one at http://mozilla.org/MPL/2.0/. + +from __future__ import print_function, unicode_literals + +import logging +import re + + +class OutputHandler(object): + """ + A class for handling Valgrind output. + + Valgrind errors look like this: + + ==60741== 40 (24 direct, 16 indirect) bytes in 1 blocks are definitely lost in loss record 2,746 of 5,235 + ==60741== at 0x4C26B43: calloc (vg_replace_malloc.c:593) + ==60741== by 0x63AEF65: PR_Calloc (prmem.c:443) + ==60741== by 0x69F236E: PORT_ZAlloc_Util (secport.c:117) + ==60741== by 0x69F1336: SECITEM_AllocItem_Util (secitem.c:28) + ==60741== by 0xA04280B: ffi_call_unix64 (in /builds/slave/m-in-l64-valgrind-000000000000/objdir/toolkit/library/libxul.so) + ==60741== by 0xA042443: ffi_call (ffi64.c:485) + + For each such error, this class extracts most or all of the first (error + kind) line, plus the function name in each of the first few stack entries. + With this data it constructs and prints a TEST-UNEXPECTED-FAIL message that + TBPL will highlight. + + It buffers these lines from which text is extracted so that the + TEST-UNEXPECTED-FAIL message can be printed before the full error. + + Parsing the Valgrind output isn't ideal, and it may break in the future if + Valgrind changes the format of the messages, or introduces new error kinds. + To protect against this, we also count how many lines containing + "<insert_a_suppression_name_here>" are seen. Thanks to the use of + --gen-suppressions=yes, exactly one of these lines is present per error. If + the count of these lines doesn't match the error count found during + parsing, then the parsing has missed one or more errors and we can fail + appropriately. + """ # NOQA: E501 + + def __init__(self, logger): + # The regexps in this list match all of Valgrind's errors. Note that + # Valgrind is English-only, so we don't have to worry about + # localization. + self.logger = logger + self.re_error = ( + r"==\d+== (" + + r"(Use of uninitialised value of size \d+)|" + + r"(Conditional jump or move depends on uninitialised value\(s\))|" + + r"(Syscall param .* contains uninitialised byte\(s\))|" + + r"(Syscall param .* points to (unaddressable|uninitialised) byte\(s\))|" + + r"((Unaddressable|Uninitialised) byte\(s\) found during client check request)|" + + r"(Invalid free\(\) / delete / delete\[\] / realloc\(\))|" + + r"(Mismatched free\(\) / delete / delete \[\])|" + + r"(Invalid (read|write) of size \d+)|" + + r"(Jump to the invalid address stated on the next line)|" + + r"(Source and destination overlap in .*)|" + + r"(.* bytes in .* blocks are .* lost)" + + r")" + ) + # Match identifer chars, plus ':' for namespaces, and '\?' in order to + # match "???" which Valgrind sometimes produces. + self.re_stack_entry = r"^==\d+==.*0x[A-Z0-9]+: ([A-Za-z0-9_:\?]+)" + self.re_suppression = r" *<insert_a_suppression_name_here>" + self.error_count = 0 + self.suppression_count = 0 + self.number_of_stack_entries_to_get = 0 + self.curr_error = None + self.curr_location = None + self.buffered_lines = None + + def log(self, line): + self.logger(logging.INFO, "valgrind-output", {"line": line}, "{line}") + + def __call__(self, line): + if self.number_of_stack_entries_to_get == 0: + # Look for the start of a Valgrind error. + m = re.search(self.re_error, line) + if m: + self.error_count += 1 + self.number_of_stack_entries_to_get = 4 + self.curr_error = m.group(1) + self.curr_location = "" + self.buffered_lines = [line] + else: + self.log(line) + + else: + # We've recently found a Valgrind error, and are now extracting + # details from the first few stack entries. + self.buffered_lines.append(line) + m = re.match(self.re_stack_entry, line) + if m: + self.curr_location += m.group(1) + else: + self.curr_location += "?!?" + + self.number_of_stack_entries_to_get -= 1 + if self.number_of_stack_entries_to_get != 0: + self.curr_location += " / " + else: + # We've finished getting the first few stack entries. Print the + # failure message and the buffered lines, and then reset state. + self.logger( + logging.ERROR, + "valgrind-error-msg", + {"error": self.curr_error, "location": self.curr_location}, + "TEST-UNEXPECTED-FAIL | valgrind-test | {error} at {location}", + ) + for b in self.buffered_lines: + self.log(b) + self.curr_error = None + self.curr_location = None + self.buffered_lines = None + + if re.match(self.re_suppression, line): + self.suppression_count += 1 diff --git a/build/valgrind/x86_64-pc-linux-gnu.sup b/build/valgrind/x86_64-pc-linux-gnu.sup new file mode 100644 index 0000000000..bc1aac6c79 --- /dev/null +++ b/build/valgrind/x86_64-pc-linux-gnu.sup @@ -0,0 +1,1008 @@ +# Full list is tracked through meta bug 793882 + +#################################### +# Leaks in third party libraries # +#################################### + +{ + Bug 793537 + Memcheck:Leak + ... + obj:/usr/lib64/libpango-1.0.so.0.2800.1 + ... +} +{ + Bug 793598 + Memcheck:Leak + ... + obj:/lib64/libdbus-1.so.3.4.0 + ... +} +# Fontconfig is going fancy with its cache structure and that confuses valgrind. +# https://bugs.freedesktop.org/show_bug.cgi?id=8215 +# https://bugs.freedesktop.org/show_bug.cgi?id=8428 +{ + Bug 1187649 + Memcheck:Leak + match-leak-kinds: definite + fun:realloc + fun:FcPatternObjectInsertElt + ... +} +{ + Bug 1187649 + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:FcPatternObjectInsertElt + ... +} +{ + Bug 1533462 (comment 9) + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:FcNameParseCharSet + fun:FcNameConvert + fun:FcNameParse + ... +} +# With older versions of fontconfig (e.g. 2.8.0 on taskcluster systems), +# there's an uninitialized memory usage and leak when loading app fonts. +{ + Bug 1231701 + Memcheck:Param + write(buf) + ... + fun:FcDirCacheWrite + fun:FcDirCacheScan + fun:FcConfigAddDirList + ... +} +{ + Bug 1231701 + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:FcDirScanConfig + fun:FcDirCacheScan + fun:FcConfigAddDirList + ... +} +{ + Bug 1231701 + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:FcDirScanConfig + fun:FcDirScanConfig + fun:FcDirCacheScan + fun:FcConfigAddDirList + ... +} +{ + Bug 1231701 + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:FcDirScanConfig + obj:/usr/lib/x86_64-linux-gnu/libfontconfig.so.1.5.0 + fun:FcDirCacheScan + fun:FcConfigAddDirList + ... +} +# Leaks due to either Gtk+3 or cairo, but Gecko is not directly involved with +# those cairo interactions. One suspected cause is Gecko not closing the +# display to work around a bug in old Gtk+3 versions. See also bug 1228724. +{ + Bug 1187649 + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:_cairo_freelist_alloc + fun:_cairo_xlib_display_queue_resource + fun:_cairo_xlib_surface_finish + ... +} +# The following leak is deep in Gtk+3, and it doesn't seem we're doing +# anything wrong on our end with the container objects. This suppression +# is purposefully verbose so as to avoid catching actual leaks due to +# Gecko code. +# Note: valgrind doesn't support more than 24 elements in a suppression stack. +{ + Bug 1187649 + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:g_malloc + fun:g_slice_alloc + fun:g_list_prepend + fun:gtk_combo_box_get_path_for_child + fun:gtk_container_get_path_for_child + fun:gtk_widget_get_path + fun:_gtk_widget_update_path + fun:reset_style_recurse + fun:gtk_widget_reset_style + fun:gtk_widget_set_parent + fun:gtk_combo_box_add + fun:g_cclosure_marshal_VOID__OBJECTv + fun:_g_closure_invoke_va + fun:g_signal_emit_valist + fun:g_signal_emit + fun:gtk_combo_box_constructor + fun:g_object_newv + fun:g_object_new_valist + fun:g_object_new + ... +} +# set_color() in gtkstyle.c of GTK version 3.4.4 only can leak GdkRGBA +# allocations when the theme has transparent colors: +# https://git.gnome.org/browse/gtk+/tree/gtk/deprecated/gtkstyle.c?h=3.4.4#n676 +{ + Bug 1250704 + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:g_malloc + fun:g_slice_alloc + fun:g_slice_copy + fun:boxed_proxy_lcopy_value + fun:gtk_style_context_get_valist + fun:gtk_style_context_get + fun:set_color + fun:gtk_style_update_from_context + fun:gtk_style_constructed + fun:g_object_newv + fun:g_object_new_valist + fun:g_object_new + ... +} +{ + Bug 794366 + Memcheck:Leak + ... + obj:/usr/lib64/libgtk-x11-2.0.so.0.1800.9 + ... +} +{ + Bug 794368 + Memcheck:Leak + ... + obj:/usr/lib64/libXrandr.so.2.2.0 + ... +} +{ + Bug 794373 + Memcheck:Leak + ... + obj:/lib64/libgobject-2.0.so.0.2200.5 + ... +} +{ + Bug 966673 + Memcheck:Leak + fun:malloc + obj:/lib/x86_64-linux-gnu/libresolv-2.13.so + ... + fun:gaih_inet + fun:getaddrinfo + fun:PR_GetAddrInfoByName + ... +} +{ + Bug 979242 + Memcheck:Leak + fun:calloc + fun:xcb_connect_to_fd + fun:xcb_connect_to_display_with_auth_info + fun:_XConnectXCB + fun:XOpenDisplay + fun:gdk_display_open + ... +} +{ + _g_io_module_get_default + Memcheck:Leak + ... + fun:g_io_module_new + fun:g_io_modules_scan_all_in_directory_with_scope + fun:_g_io_modules_ensure_loaded + fun:_g_io_module_get_default +} +{ + dlopen leak triggered by bug 1492121 + Memcheck:Leak + match-leak-kinds: definite + fun:malloc + fun:dl_open_worker + fun:_dl_catch_error + fun:_dl_open + fun:dlopen_doit + fun:_dl_catch_error + fun:_dlerror_run + fun:dlopen@@GLIBC_2.2.5 + ... +} + +################################### +# Leaks in short lived precesses # +################################### + +{ + Bug 984196 + Memcheck:Leak + ... + fun:childgltest + ... +} + +######################################### +# Uninitialised value false positives # +######################################### + +# This concerns a false positive pertaining to Memcheck's overly- +# conservative instrumentation of CPUID. See bug 1288618 comments +# 119 through 127. +{ + Bug 1288618 comments 119 through 127 + Memcheck:Cond + fun:_ZN6SkOptsL4initEv + fun:sk_once_no_arg_adaptor +} + +{ + Bug 1288618 comments 119 through 127 part 2 + Memcheck:Cond + fun:__get_cpuid + fun:cpuid + fun:_ZN6SkOptsL4initEv + fun:sk_once_no_arg_adaptor +} + +# More stuff to do with CPUID and Skia. Apparently we could get rid of +# these if we could patch our in-tree Skia, but that's not favoured. +# +# Conditional jump or move depends on uninitialised value(s) +# at 0xFDD1D97: SkCpu::CacheRuntimeFeatures() +# by 0xFE8A66E: SkGraphics::Init() +# by 0xE757308: gfxPlatform::Init() +# by 0xE75772C: gfxPlatform::GetPlatform() +{ + Skia and CPUID, Jan 2017, #1 + Memcheck:Cond + fun:_ZN5SkCpu20CacheRuntimeFeaturesEv + fun:_ZN10SkGraphics4InitEv + fun:_ZN11gfxPlatform4InitEv + fun:_ZN11gfxPlatform11GetPlatformEv +} + +# Conditional jump or move depends on uninitialised value(s) +# at 0xFD5B218: SkOpts::Init() +# by 0xE757308: gfxPlatform::Init() +# by 0xE75772C: gfxPlatform::GetPlatform() +# by 0xF1A3691: mozilla::dom::ContentProcess::Init() +# or mozilla::dom::ContentChild::RecvSetXPCOMProcessAttributes() +{ + Skia and CPUID, Jan 2017, #2 + Memcheck:Cond + fun:_ZN6SkOpts4InitEv + fun:_ZN11gfxPlatform4InitEv + fun:_ZN11gfxPlatform11GetPlatformEv + fun:_ZN7mozilla3dom*Content* +} + +# False positives triggered by rust 1.20.0 (at least) builds of stylo. +# See bug 1394696. The diagnosis is an llvm optimization transforming +# `if A && B` to `if B && A` if is can be proven that A is false +# whenever B is uninitialized. Confusing, but valid. +# +# Conditional jump or move depends on uninitialised value(s) +# at 0x113ED01E: selectors::matching::matches_complex_selector_internal (option.rs:421) +# by 0x113ECF19: selectors::matching::matches_complex_selector (matching.rs:501) +# by 0x113EBAC0: <style::selector_map::SelectorMap<style::stylist::Rule>>::get_matching_rules (matching.rs:397) +{ + Bug 1394696 Stylo selector, Sept 2017, part 1 + Memcheck:Cond + fun:_ZN9selectors8matching33matches_complex_selector_internal* + fun:_ZN9selectors8matching24matches_complex_selector* + ... + fun:_ZN69_$LT$style..selector_map..SelectorMap$LT$style..stylist..Rule$GT$$GT$18get_matching_rules* +} + +# Conditional jump or move depends on uninitialised value(s) +# at 0x113EFFDE: selectors::matching::matches_complex_selector_internal (option.rs:421) +# by 0x113EFED9: selectors::matching::matches_complex_selector (matching.rs:501) +# by 0x113DFE55: style::stylist::Stylist::match_revalidation_selectors::{{closure}} (matching.rs:397) +{ + Bug 1394696 Stylo selector, Sept 2017, part 2 + Memcheck:Cond + fun:_ZN9selectors8matching33matches_complex_selector_internal* + fun:_ZN9selectors8matching24matches_complex_selector* + ... + fun:_ZN5style9traversal13compute_style* + fun:recalc_style_at<style::gecko::wrapper::GeckoElement,style::gecko::traversal::RecalcStyleOnly,closure> +} + +# Similar issue triggered by rust 1.23.0 on builds of stylo. +# +# at 0x11819B3E: std::sync::once::Once::call_once::{{closure}} (raw_vec.rs:225) +# by 0x118FDCCC: std::sync::once::Once::call_inner (once.rs:341) +# by 0x1139C761: UnknownInlinedFun (once.rs:228) +{ + Bug 1418083 Servo::TraverseSubtree, January 2018 + Memcheck:Cond + fun:*ZN3std4sync4once4Once9call_once* + fun:*ZN3std4sync4once4Once10call_inner* + ... + fun:Servo_TraverseSubtree +} + +# Issue triggered by rust 1.23.0 on builds of stylo. +{ + Bug 1418083 SelectorList::parse, January 2018 + Memcheck:Cond + fun:_ZN36_$LT$smallvec..SmallVec* + fun:_ZN52_$LT$selectors..parser..SelectorList* +} + +# Conditional jump or move depends on uninitialised value(s) +# at 0x118720EA: <core::slice::Iter<'a, T> as core::iter::iterator::Iterator>::all::{{closure}} (mod.rs:1173) +# by 0x11871EF6: style::media_queries::MediaList::evaluate (mod.rs:1277) +{ + Bug 1430608 nsMediaFeatures, January 2018 + Memcheck:Cond + fun:_ZN91_$LT$core..slice..Iter$LT$$u27$a$C$$u20$T$GT$$u20$as$u20$core..iter..iterator..Iterator$GT$3all* + fun:_ZN5style13media_queries9MediaList8evaluate* +} + +# Several issues triggered by enabling stylo-chrome. + +# Conditional jump or move depends on uninitialised value(s) +# at 0x11F88693: style::properties::LonghandId::parse_value (heap.rs:0) +# by 0x1203BBAB: style::properties::UnparsedValue::substitute_variables (parser.rs:594) +# by 0x11CFCE33: style::properties::cascade (properties.rs:135672) +{ + Bug 1437796 LonghandId::parse_value, January 2018 + Memcheck:Cond + fun:_ZN5style10properties10LonghandId11parse_value* + fun:_ZN5style10properties13UnparsedValue20substitute_variables* + fun:_ZN5style10properties7cascade* +} + +# Conditional jump or move depends on uninitialised value(s) +# at 0x11DF7AF6: <smallvec::SmallVec<A>>::grow (raw_vec.rs:0) +# by 0x11DF6ED0: Servo_InvalidateStyleForDocStateChanges (lib.rs:475) +# by 0x10394B7E: mozilla::ServoStyleSet::InvalidateStyleForDocumentStateChanges(mozilla::EventStates) (ServoStyleSet.cpp:248) +{ + Bug 1437796 SmallVec::grow in Servo_InvalidateStyleForDocStateChanges, January 2018 + Memcheck:Cond + fun:_ZN36_$LT$smallvec..SmallVec$LT$A$GT$$GT$4grow* + fun:Servo_InvalidateStyleForDocStateChanges + fun:_ZN7mozilla13ServoStyleSet38InvalidateStyleForDocumentStateChangesENS_11EventStatesE +} + + +############################################################################## +## BEGIN suppressions for Stylo as compiled by rustc 1.25.0 +# Even more similar issues, resulting from transitioning to rustc 1.25.0. +# See bug 1447137. + + +# Suppressions rooted at *style*values*specified*color*Color*style*parser*Parse*parse* + +{ + bug1447137-01 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*SimpleShadow* + fun:*style*parse_value* + fun:*style*parse_into* +} + +{ + bug1447137-02 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*UrlPaintServer* + fun:*style*parse_value* + fun:*style*parse_into* +} + +{ + bug1447137-03 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_fallback* + fun:*style*UrlPaintServer* + fun:*style*parse_value* +} + +{ + bug1447137-04 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_border* + fun:*style*parse_into* + fun:*style*parse_value* +} + +{ + bug1447137-05 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_quirky* + fun:*style*parse_value* + fun:*style*parse_into* +} + +{ + bug1447137-06 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_value* + fun:*style*parse_into* + fun:*style*parse_value* +} + +{ + bug1447137-07 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_border* + fun:*style*parse_value* + fun:*style*parse_into* +} + +{ + bug1447137-08 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_quirky* + fun:*style*parse_value* + fun:*style*substitute_variables* +} + +{ + bug1447137-09 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*Gradient* + fun:*style*Parse* + fun:*style*parse_value* +} + +{ + bug1447137-10 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*SimpleShadow* + fun:*style*parse*closure* + fun:*style*parse_value* +} + +{ + bug1447137-11 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_border* + fun:*style*substitute_variables* + fun:*style*substitute_variables* +} + +{ + bug1447137-12 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_border* + fun:*style*parse_value* + fun:*style*substitute_variables* +} + +{ + bug1447137-13 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_value* + fun:*style*substitute_variables* + fun:*style*substitute_variables* +} + + +{ + bug1447137-14 + Memcheck:Cond + fun:*style*values*specified*color*Color*style*parser*Parse*parse* + fun:*style*parse_value* + fun:*style*parse_into* + fun:*geckoservo*parse_property_into* +} + +########################################## +# Suppressions rooted at *selectors*parser*SelectorList*Impl*parse* + +{ + bug1454511 + Memcheck:Cond + fun:*selectors*parser*SelectorList*Impl*parse* +} + +########################################## +# Suppressions rooted at *style*properties*shorthands* + +{ + bug1447137-17 + Memcheck:Cond + fun:*style*properties*shorthands* + fun:*style*properties*PropertyDeclaration*parse_into* + fun:*style*parse_value* + fun:*cssparser*Iterator*next* +} + +## END suppressions for Stylo as compiled by rustc 1.25.0 +############################################################################## + + +# Another suppression for Stylo, August 2018. See bug 1479055. +# +# Conditional jump or move depends on uninitialised value(s) +# at style::[..]::clip_path::cascade_property (gecko_properties.rs:17152) +# by style::properties::cascade_rules (properties.rs:83642) +# by [..]::cascade_style_and_visited (properties.rs:83129) +# by [..]::cascade_primary_style (style/style_resolver.rs:216) +{ + Bug 1479055: style::properties::longhands::clip_path::cascade_property + Memcheck:Cond + fun:_ZN5style10properties9longhands9clip_path16cascade_property* + fun:_ZN5style10properties13cascade_rules* + fun:_ZN109_$LT$style*style_resolver*cascade_style_and_visited* + fun:_ZN109_$LT$style*cascade_primary_style* +} + +# Another suppression for Stylo, October 2018. See bug 1496486. +# Conditional jump or move depends on uninitialised value(s) +# at 0x108E2931: style::properties::longhands::font_language_override::cascade_property+81 (font.rs:2095) +# by 0x107E95B4: style::properties::cascade::Cascade::apply_properties+580 (cascade.rs:463) +# by 0x107E82D8: style::properties::cascade::cascade_rules+2440 (cascade.rs:303) +# by 0x107E7206: <style::style_resolver::StyleResolverForElement<'a, 'ctx, 'le, E>>::cascade_style_and_visited+310 (cascade.rs:93) +# Uninitialised value was created by a stack allocation +# at 0x10A42A40: style::properties::shorthands::font::parse_value (font.rs:376) +{ + Bug 1496486: style::properties::longhands::font_language_override::cascade_property + Memcheck:Cond + fun:_ZN5style10properties9longhands22font_language_override16cascade_property* + fun:_ZN5style10properties7cascade7Cascade16apply_properties* + fun:_ZN5style10properties7cascade13cascade_rules* +} + +################################################### +# For valgrind-mochitest ("tc-M-V [tier 2]") runs on taskcluster. +# See bug 1248365. +# These are specific to Ubuntu 12.04.5, 64-bit. +################################################### + + +# Not sure what this is. Is it the well-known +# Memcheck-being-confused-by-zlib thing? I suspect so. +# +# Conditional jump or move depends on uninitialised value(s) +# at 0xB6154E0: inflateReset2 (in /lib/x86_64-linux-gnu/libz.so.1.2.3.4) +# by 0xB6155D8: inflateInit2_ (in /lib/x86_64-linux-gnu/libz.so.1.2.3.4) +# by 0xADDE253: png_create_read_struct_2 +# (in /lib/x86_64-linux-gnu/libpng12.so.0.46.0) +# Uninitialised value was created by a heap allocation +# at 0x4C2D027: malloc (coregrind/m_replacemalloc/vg_replace_malloc.c:298) +# by 0xADE960F: png_malloc (in /lib/x86_64-linux-gnu/libpng12.so.0.46.0) +# by 0xADD1B95: ??? (in /lib/x86_64-linux-gnu/libpng12.so.0.46.0) +# by 0xB6155B6: inflateInit2_ (in /lib/x86_64-linux-gnu/libz.so.1.2.3.4) +# by 0xADDE253: png_create_read_struct_2 +# (in /lib/x86_64-linux-gnu/libpng12.so.0.46.0) +# by 0x15707869: ??? (in /usr/lib/x86_64-linux-gnu/gdk-pixbuf-2.0/2.10.0/ +# loaders/libpixbufloader-png.so) +{ + Bug 1248365: mochitest-libz-1 + Memcheck:Cond + fun:inflateReset2 + fun:inflateInit2_ + fun:png_create_read_struct_2 +} + + +# I don't know what this is. +# +# Conditional jump or move depends on uninitialised value(s) +# at 0x39608257: ??? (in /usr/lib/x86_64-linux-gnu/librsvg-2.so.2.36.1) +# by 0x39608E96: rsvg_handle_get_pixbuf_sub (in /usr/lib/x86_64-linux-gnu/l +# by 0x37D00D2A: ??? (in /usr/lib/x86_64-linux-gnu/gdk-pixbuf-2.0/2.10.0/lo +# by 0x88FF740: gdk_pixbuf_loader_close (in /usr/lib/x86_64-linux-gnu/libgd +# by 0x88FBA48: ??? (in /usr/lib/x86_64-linux-gnu/libgdk_pixbuf-2.0.so.0.26 +# by 0x88FD290: gdk_pixbuf_new_from_stream_at_scale (in /usr/lib/x86_64-lin +# by 0x6EF96A7: ??? (in /usr/lib/x86_64-linux-gnu/libgtk-3.so.0.400.2) +# by 0x6EFC2B1: gtk_icon_info_load_icon (in /usr/lib/x86_64-linux-gnu/libgt +# Uninitialised value was created by a stack allocation +# at 0xAB786B0: ??? (in /usr/lib/x86_64-linux-gnu/libpixman-1.so.0.30.2) +{ + Bug 1248365: mochitest-librsvg-1 + Memcheck:Cond + obj:/*/librsvg-2.so.2.36* + fun:rsvg_handle_get_pixbuf_sub + obj:/*/libpixbufloader-svg.so* + fun:gdk_pixbuf_loader_close +} + + +# There now follow some suppressions to do with libpixman. There are two +# variants, one for errors in the system libpixman, and one for errors in +# our in-tree copy. I suspect that all of these sse2 compositing reports +# are Memcheck false positives, possibly to do with inaccurate +# instrumentation of the function combine1() in +# gfx/cairo/libpixman/src/pixman-sse2.c. +# +# Conditional jump or move depends on uninitialised value(s) +# at 0xAB93A10: ??? (in /usr/lib/x86_64-linux-gnu/libpixman-1.so.0.30.2) +# by 0xAB78927: ??? (in /usr/lib/x86_64-linux-gnu/libpixman-1.so.0.30.2) +# by 0xAB40B0B: pixman_image_composite32 (in /usr/lib/x86_64-linux-gnu/libp +# by 0x865C95C: ??? (in /usr/lib/x86_64-linux-gnu/libcairo.so.2.11000.2) +# by 0x865E3CB: ??? (in /usr/lib/x86_64-linux-gnu/libcairo.so.2.11000.2) +# by 0x865F6EA: ??? (in /usr/lib/x86_64-linux-gnu/libcairo.so.2.11000.2) +# by 0x865F968: ??? (in /usr/lib/x86_64-linux-gnu/libcairo.so.2.11000.2) +# by 0x867D7C3: ??? (in /usr/lib/x86_64-linux-gnu/libcairo.so.2.11000.2) +# Uninitialised value was created by a stack allocation +# at 0xAB786B0: ??? (in /usr/lib/x86_64-linux-gnu/libpixman-1.so.0.30.2) +{ + Bug 1248365: mochitest-libpixman-1 + Memcheck:Cond + obj:/*/libpixman-1.so.0.30* + obj:/*/libpixman-1.so.0.30* + fun:pixman_image_composite32 +} + + +# Conditional jump or move depends on uninitialised value(s) +# at 0xF9EA219: sse2_composite_over_8888_8888 +# by 0xF9F5B5F: _moz_pixman_image_composite32 +# by 0xF96E29E: _clip_and_composite_boxes +# by 0xF96F79D: _cairo_image_surface_fill +# by 0xF98790C: _cairo_surface_fill +# by 0xF96FE2E: _cairo_gstate_fill +# by 0xF98B3D9: _moz_cairo_fill_preserve +# by 0xE4CF383: mozilla::gfx::DrawTargetCairo::DrawPattern(mozilla::gfx: +# Uninitialised value was created by a stack allocation +# at 0xB8E46B0: ??? (in /usr/lib/x86_64-linux-gnu/libpixman-1.so.0.30.2) +{ + Bug 1248365: mochitest-libpixman-2 + Memcheck:Cond + fun:sse2_composite_over_8888_8888 + fun:_moz_pixman_image_composite32 + fun:_clip_and_composite_boxes + fun:_cairo_image_surface_fill +} + + +# Conditional jump or move depends on uninitialised value(s) +# at 0xF9D56AE: sse2_combine_over_u (in /builds/worker/workspace/build/applic +# by 0xF9D05D4: general_composite_rect (in /builds/worker/workspace/build/app +# by 0xF9F5B5F: _moz_pixman_image_composite32 (in /builds/worker/workspace/bu +# by 0xF96CF63: _clip_and_composite (in /builds/worker/workspace/build/applic +# by 0xF96D656: _clip_and_composite_boxes.part.32 (in /builds/worker/workspac +# by 0xF96E328: _clip_and_composite_boxes (in /builds/worker/workspace/build/ +# by 0xF96F79D: _cairo_image_surface_fill (in /builds/worker/workspace/build/ +# by 0xF98790C: _cairo_surface_fill (in /builds/worker/workspace/build/applic +# Uninitialised value was created by a stack allocation +# at 0xF9D024D: general_composite_rect (in /builds/worker/workspace/build/app +# +{ + Bug 1248365: mochitest-libpixman-3 + Memcheck:Cond + fun:sse2_combine_over_u + fun:general_composite_rect + fun:_moz_pixman_image_composite32 + fun:_clip_and_composite* +} + + +# Conditional jump or move depends on uninitialised value(s) +# at 0xE626A5C: mozilla::image::imgFrame::Optimize() (in /builds/worker/work +# by 0xE626C68: mozilla::image::imgFrame::UnlockImageData() (in /home/work +# by 0xE608E8F: mozilla::image::RawAccessFrameRef::~RawAccessFrameRef() (i +# by 0xE61F5E4: mozilla::image::Decoder::~Decoder() (in /builds/worker/works +# by 0xE630E32: mozilla::image::nsIconDecoder::~nsIconDecoder() (in /home/ +# by 0xE61A5B2: mozilla::image::Decoder::Release() (in /builds/worker/worksp +# by 0xE61DD73: mozilla::image::NotifyDecodeCompleteWorker::~NotifyDecodeC +# by 0xE61DD8F: mozilla::image::NotifyDecodeCompleteWorker::~NotifyDecodeC +# Uninitialised value was created by a stack allocation +# at 0xB8E46B0: ??? (in /usr/lib/x86_64-linux-gnu/libpixman-1.so.0.30.2) +{ + Bug 1248365: mochitest-libpixman-4 + Memcheck:Cond + fun:_ZN7mozilla5image8imgFrame8OptimizeEv + fun:_ZN7mozilla5image8imgFrame15UnlockImageDataEv + fun:_ZN7mozilla5image17RawAccessFrameRefD1Ev + fun:_ZN7mozilla5image7DecoderD1Ev +} + + +# Not sure what this. I can't reproduce it locally despite much trying. +# Syscall param sendmsg(msg.msg_iov[0]) points to uninitialised byte(s) +# at 0x4E4533D: ??? (syscall-template.S:82) +# by 0xE12C0A7: IPC::Channel::ChannelImpl::ProcessOutgoingMessages() (in /h +# by 0xE142FD0: RunnableMethod<IPC::Channel, bool (IPC::Channel::*)(IPC::Me +# by 0xE1240EA: MessageLoop::RunTask(Task*) (in /builds/worker/workspace/buil +# by 0xE128A46: MessageLoop::DeferOrRunPendingTask(MessageLoop::PendingTask +# by 0xE128B6D: MessageLoop::DoWork() (in /builds/worker/workspace/build/appl +# by 0xE12272C: base::MessagePumpLibevent::Run(base::MessagePump::Delegate* +# by 0xE124155: MessageLoop::Run() (in /builds/worker/workspace/build/applica +{ + Bug 1248365: mochitest-sendmsg-1 + Memcheck:Param + sendmsg(msg.msg_iov[0]) + obj:/lib/x86_64-linux-gnu/libpthread-2.15.so + fun:_ZN3IPC7Channel11ChannelImpl23ProcessOutgoingMessagesEv + fun:_ZN14RunnableMethodIN3IPC7ChannelEMS1_FbPNS0_7MessageEEN7mozilla5Tuple* +} + + +# I can't repro this either. +# Conditional jump or move depends on uninitialised value(s) +# at 0x418E7E7C: ??? (in /usr/lib/x86_64-linux-gnu/libavcodec.so.53.35.0) +# by 0x4192D620: ??? (in /usr/lib/x86_64-linux-gnu/libavcodec.so.53.35.0) +# by 0x4192E717: ??? (in /usr/lib/x86_64-linux-gnu/libavcodec.so.53.35.0) +# by 0x41711BC4: ??? (in /usr/lib/x86_64-linux-gnu/libavcodec.so.53.35.0) +# by 0x41B08B6A: avcodec_open2 (in /usr/lib/x86_64-linux-gnu/libavcodec.so. +# by 0xEEAD89C: mozilla::FFmpegDataDecoder<53>::InitDecoder() (in /home/wor +# by 0xEEAE42B: mozilla::FFmpegVideoDecoder<53>::Init() (in /builds/worker/wo +# by 0xEEA4C07: mozilla::H264Converter::Init() (in /builds/worker/workspace/b +# Uninitialised value was created by a heap allocation +# at 0x4C2D11F: realloc (vg_replace_malloc.c:785) +# by 0x406196: moz_xrealloc (in /builds/worker/workspace/build/application/fi +# by 0xDEB43AC: nsTArrayInfallibleAllocator::ResultTypeProxy nsTArray_base< +# by 0xEEAD850: mozilla::FFmpegDataDecoder<53>::InitDecoder() (in /home/wor +# by 0xEEAE42B: mozilla::FFmpegVideoDecoder<53>::Init() (in /builds/worker/wo +{ + Bug 1248365: mochitest-libavcodec-1-c + Memcheck:Cond + obj:/*/libavcodec.so.53* + obj:/*/libavcodec.so.53* + obj:/*/libavcodec.so.53* + obj:/*/libavcodec.so.53* +} +{ + Bug 1248365: mochitest-libavcodec-1-v8 + Memcheck:Value8 + obj:/*/libavcodec.so.53* + obj:/*/libavcodec.so.53* + obj:/*/libavcodec.so.53* + obj:/*/libavcodec.so.53* +} + + +# Not sure what this is, but I am inclined to think it is also probably a +# SSE2-induced false positive similar to mochitest-libpixman-2 above. +# Use of uninitialised value of size 8 +# at 0xE4F3E89: FastConvertYUVToRGB32Row (in /builds/worker/workspace/build/a +# by 0xE4F4A6D: mozilla::gfx::ConvertYCbCrToRGB32(unsigned char const*, uns +# by 0xE4F4B17: mozilla::gfx::ConvertYCbCrToRGB(mozilla::layers::PlanarYCbC +# by 0xE5227CB: mozilla::layers::PlanarYCbCrImage::GetAsSourceSurface() (in +# by 0xE5B2465: mozilla::layers::SharedPlanarYCbCrImage::GetAsSourceSurface +# by 0xE52FE44: mozilla::layers::BasicImageLayer::Paint(mozilla::gfx::DrawT +# by 0xE5618A1: mozilla::layers::BasicLayerManager::PaintSelfOrChildren(moz +# by 0xE560F83: mozilla::layers::BasicLayerManager::PaintLayer(gfxContext*, +# Uninitialised value was created by a stack allocation +# at 0x434B36B2: ??? (in /usr/lib/x86_64-linux-gnu/libavcodec.so.53.35.0) +{ + Bug 1248365: FastConvertYUVToRGB32Row-1 + Memcheck:Value8 + fun:FastConvertYUVToRGB32Row + fun:_ZN7mozilla3gfx19ConvertYCbCrToRGB32* + fun:_ZN7mozilla3gfx17ConvertYCbCrToRGB* + fun:_ZN7mozilla6layers16PlanarYCbCrImage18GetAsSourceSurface* +} + +# Similarly: +# Conditional jump or move depends on uninitialised value(s) +# at 0xFDAD1D1: sse41::blit_row_s32a_opaque(unsigned int*, unsigned int con +# by 0xFD60FA9: Sprite_D32_S32::blitRect(int, int, int, int) (in /home/work +# by 0xFEB9E0D: SkScan::FillIRect(SkIRect const&, SkRegion const*, SkBlitte +# by 0xFEBDDF3: SkScan::FillIRect(SkIRect const&, SkRasterClip const&, SkBl +{ + SKIA and SSE4, Jan 2017 + Memcheck:Cond + fun:_ZN5sse41L20blit_row_s32a_opaque* + fun:_ZN14Sprite_D32_S328blitRect* + fun:_ZN6SkScan9FillIRect* + fun:_ZN6SkScan9FillIRect* +} + +# This is probably a V false positive, due to an insufficiently accurate +# description of the ioctl(SIOCETHTOOL) behavior. +# Syscall param ioctl(SIOCETHTOOL) points to uninitialised byte(s) +# at 0x5D5CBF7: ioctl (syscall-template.S:82) +# by 0xF58EB67: nr_stun_get_addrs (in /builds/worker/workspace/build/applica +# by 0xF594791: nr_stun_find_local_addresses (in /builds/worker/workspace/bu +# by 0xF58A237: nr_ice_get_local_addresses (in /builds/worker/workspace/buil +# by 0xF58ADDE: nr_ice_gather (in /builds/worker/workspace/build/application +# by 0xE43F35F: mozilla::NrIceCtx::StartGathering() (in /builds/worker/works +# by 0xE419560: mozilla::PeerConnectionMedia::EnsureIceGathering_s() (in / +# by 0xE41A11C: mozilla::runnable_args_memfn<RefPtr<mozilla::PeerConnectio +# Address 0x1cc3fb48 is on thread 6's stack +# in frame #1, created by nr_stun_get_addrs (???:) +{ + Bug 1248365: mochitest-ioctl(SIOCETHTOOL)-1 + Memcheck:Param + ioctl(SIOCETHTOOL) + fun:ioctl + fun:nr_stun_get_addrs + fun:nr_stun_find_local_addresses +} + + +# This looks like uninitialised padding bytes in a structure owned +# by libfontconfig. So there's nothing we can do about it. +# Syscall param write(buf) points to uninitialised byte(s) +# at 0x4E44CCD: ??? (syscall-template.S:82) +# by 0x9F1FF56: ??? (in /usr/lib/x86_64-linux-gnu/libfontconfig.so.1.4.4) +# by 0x9F2679B: ??? (in /usr/lib/x86_64-linux-gnu/libfontconfig.so.1.4.4) +# by 0x9F22B98: ??? (in /usr/lib/x86_64-linux-gnu/libfontconfig.so.1.4.4) +# by 0x9F22C5F: FcConfigAppFontAddDir (in /usr/lib/x86_64-linux-gnu/libfon +# by 0xE850173: gfxFcPlatformFontList::ActivateBundledFonts() (in /home/wo +# by 0xE852258: gfxFcPlatformFontList::InitFontListForPlatform() (in /home +# by 0xE895E21: gfxPlatformFontList::InitFontList() (in /builds/worker/works +# Address 0x2316663c is 156 bytes inside a block of size 1,448 alloc'd +# at 0x4C2CF71: malloc (vg_replace_malloc.c:299) +# by 0x9F1FD1D: ??? (in /usr/lib/x86_64-linux-gnu/libfontconfig.so.1.4.4) +# by 0x9F26788: ??? (in /usr/lib/x86_64-linux-gnu/libfontconfig.so.1.4.4) +# by 0x9F22B98: ??? (in /usr/lib/x86_64-linux-gnu/libfontconfig.so.1.4.4) +# by 0x9F22C5F: FcConfigAppFontAddDir (in /usr/lib/x86_64-linux-gnu/libfon +# by 0xE850173: gfxFcPlatformFontList::ActivateBundledFonts() (in /home/wo +# by 0xE852258: gfxFcPlatformFontList::InitFontListForPlatform() (in /home +# by 0xE895E21: gfxPlatformFontList::InitFontList() (in /builds/worker/works +{ + Bug 1248365: libfontconfig-1 + Memcheck:Param + write(buf) + obj:/*/libpthread*.so* + obj:/*/libfontconfig.so* + ... + obj:/*/libfontconfig.so* + fun:FcConfigAppFontAddDir +} + + +# There's nothing we can do about these short of throwing in +# --show-mismatched-frees=no, but that's a bit drastic, so for now, +# just suppress them. A typical error is: +# +# Mismatched free() / delete / delete [] +# at 0x4C2BE97: free (vg_replace_malloc.c:530) +# by 0xFCD09EC: ots::ots_post_free(ots::Font*) (in /builds/worker/workspace/ +# by 0xFCC600E: ots::Font::~Font() (in /builds/worker/workspace/build/applic +# by 0xFCCBFA5: ots::OTSContext::Process(ots::OTSStream*, unsigned char co +# by 0xE7D7C8D: gfxUserFontEntry::SanitizeOpenTypeData(unsigned char const +# by 0xE7E371D: gfxUserFontEntry::LoadPlatformFont(unsigned char const*, u +# by 0xE7E48AA: gfxUserFontEntry::FontDataDownloadComplete(unsigned char c +# by 0xF49D25B: nsFontFaceLoader::OnStreamComplete(nsIStreamLoader*, nsISu +# Address 0x15671f00 is 0 bytes inside a block of size 490 alloc'd +# at 0x4C2CAEE: operator new(unsigned long) (vg_replace_malloc.c:332) +# by 0xF6AB737: std::vector<unsigned short, std::allocator<unsigned short> +# by 0xFCD0FDE: ots::ots_post_parse(ots::Font*, unsigned char const*, unsi +# by 0xFCCA3D9: (anonymous namespace)::ProcessGeneric(ots::OpenTypeFile*, +# by 0xFCCB17E: (anonymous namespace)::ProcessTTF(ots::OpenTypeFile*, ots: +# by 0xFCCBA54: ots::OTSContext::Process(ots::OTSStream*, unsigned char co +# by 0xE7D7C8D: gfxUserFontEntry::SanitizeOpenTypeData(unsigned char const +# by 0xE7E371D: gfxUserFontEntry::LoadPlatformFont(unsigned char const*, u +{ + Bug 1248365: ots::Font::~Font()-1 + Memcheck:Free + fun:free + fun:_ZN3ots13ots_post_free* + fun:_ZN3ots4FontD1Ev + fun:_ZN3ots10OTSContext7Process* +} + +# and various similar: +{ + ots mismatched frees, Jan 2017, #1 + Memcheck:Free + fun:_ZdlPv + fun:_ZN3ots14ots_glyf_parse* + fun:_ZN12_GLOBAL__N_114ProcessGenericEPN3ots12OpenTypeFile* + fun:_ZN12_GLOBAL__N_110ProcessTTFEPN3ots12OpenTypeFile* +} +{ + ots mismatched frees, Jan 2017, #2 + Memcheck:Free + fun:_ZdlPv + fun:_ZN3ots13ots_cff_parse* + fun:_ZN12_GLOBAL__N_114ProcessGenericEPN3ots12OpenTypeFile* + fun:_ZN3ots10OTSContext7ProcessEPNS_9OTSStream* +} +{ + ots mismatched frees, Jan 2017, #3 + Memcheck:Free + fun:_ZdlPv + fun:_ZN3ots13ots_cff_parse* + fun:_ZN12_GLOBAL__N_114ProcessGenericEPN3ots12OpenTypeFile* + fun:_ZN12_GLOBAL__N_110Process* +} +{ + ots mismatched frees, Jan 2017, #4 + Memcheck:Free + fun:_ZdlPv + fun:_ZN3ots12ots_cff_free* + fun:_ZN3ots4FontD1Ev* + fun:_ZN3ots10OTSContext7Process* +} +{ + ots mismatched frees, Jan 2017, #5 + Memcheck:Free + fun:_ZdlPv + fun:_ZN3ots13ots_loca_free* + fun:_ZN3ots4FontD1Ev* + fun:_ZN3ots10OTSContext7Process* +} +{ + ots mismatched frees, Jan 2017, #6 + Memcheck:Free + fun:_ZdlPv + fun:_ZN3ots14ots_glyf_parse* + fun:_ZN12_GLOBAL__N_114ProcessGenericEPN3ots12OpenTypeFile* + fun:_ZN3ots10OTSContext7ProcessEPNS_9OTSStream* +} + +{ + map_or<selectors::parser::Combinator,bool,closure> #1 (see bug 1365915) + Memcheck:Cond + fun:map_or<selectors::parser::Combinator,bool,closure> + fun:_ZN9selectors8matching33matches_complex_selector_internal* + fun:_ZN9selectors8matching24matches_complex_selector* + fun:matches_selector<style::gecko::wrapper::GeckoElement,closure> +} + +{ + map_or<selectors::parser::Combinator,bool,closure> #2 (see bug 1365915) + Memcheck:Cond + fun:map_or<selectors::parser::Combinator,bool,closure> + fun:_ZN9selectors8matching33matches_complex_selector_internal* + fun:_ZN9selectors8matching24matches_complex_selector* + fun:{{closure}}<closure> +} + +# more leaks in libLLVM-3.6-mesa.so, August 2017. See bug 1338651. +{ + static-object-leaks-in-libLLVM-3.6-mesa.so. See bug 1338651. + Memcheck:Leak + match-leak-kinds: definite + fun:_Znwm + obj:/*/lib*/libLLVM-3.6-mesa.so + obj:/*/lib*/libLLVM-3.6-mesa.so +} + +{ + bug 1632735 + Memcheck:Cond + fun:_ZN9selectors6parser25parse_one_simple_selector17h* + fun:_ZN9selectors6parser14parse_selector17h* + fun:_ZN9selectors6parser24SelectorList$LT$Impl$GT$16parse_with_state17h* +} + +{ + bug 1632735 + Memcheck:Cond + fun:_ZN9selectors6parser25parse_one_simple_selector17h* + fun:_ZN9selectors6parser29parse_functional_pseudo_class17h* + fun:_ZN9selectors6parser25parse_one_simple_selector17h* + fun:_ZN9selectors6parser14parse_selector17h* + fun:_ZN9selectors6parser24SelectorList$LT$Impl$GT$16parse_with_state17h* +} + +{ + bug 1632735 + Memcheck:Cond + fun:_ZN9selectors6parser25parse_one_simple_selector17h* + fun:_ZN9selectors6parser14parse_selector17h* + fun:_ZN9cssparser6parser18parse_until_before17h* + fun:_ZN5style5gecko15selector_parser94_$LT$impl$u20$selectors..parser..Parser$u20$for$u20$style..selector_parser..SelectorParser$GT$36parse_non_ts_functional_pseudo_class17h* + fun:_ZN9selectors6parser29parse_functional_pseudo_class17h* + fun:_ZN9selectors6parser25parse_one_simple_selector17* + fun:_ZN9selectors6parser14parse_selector17h* + fun:_ZN9selectors6parser24SelectorList$LT$Impl$GT$16parse_with_state17h* +} |