A .dSYM may validly contain a __DWARF segment without any __debug_info
section. This can occur for Chromium Framework in a component build of
Chromium, because in that case, all of the code is in other libraries
that Chromium Framework depends on.
This was previously tested by an assertion, but the assertion did not
trigger in NDEBUG (release) builds. In NDEBUG builds, this condition
would lead to an out-of-bounds read, detected by AddressSanitizer.
Instead of an assertion, the check is now always done at runtime.
Instead of being fatal, it's now just a warning, because it's been
established that __DWARF without __debug_info can occur.
(In the Chromium case, it remains pointless to run dump_syms via the
"chrome_dump_syms" target on a component build, as it'll only attempt to
symbolize Chromium Framework, and not any of the libraries that Chromium
Framework depends on that actually contain the code.)
Bug: chromium:991206
Change-Id: I6c9c75f0be7901813e3eaae54aff38c1afe73ca9
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1741610
Reviewed-by: Robert Sesek <rsesek@chromium.org>
Since target_ptr is of type uint16_t* already, we don't need to scale
the byte count as the language does that for us. If it were void*, we
would need this code, but it's not.
In practice it's probably not a big deal due to how we preallocated
memory: when converting UTF8->UTF16, we'd reserve the same number of
code units, and UTF8 takes more code units per codepoint than UTF16,
so the UTF16 vector is always oversized.
When converting UTF32->UTF16, we also reserve the same number of
code units, but since one UTF32 code unit could require two UTF16
code units (for U+10000 codepoints and higher), we would probably
corrupt memory in the process. The APIs in this module don't seem
to take into account that range in general, so for now I'm only
fixing the memory corruption.
Bug: google-breakpad:768
Change-Id: Ibfaea4e866733ff8d99b505e72c500bd40d11a74
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1732888
Reviewed-by: Mark Mentovai <mark@chromium.org>
Pointer authentication codes are used to validate pointers against
accidental or malicious modification by storing a hash of the address
and a secret value in the pointer's unused, upper bits. The exact
bits used may vary by implementation and depend on the size of the
virtual address space of the target system, and whether other tagged
pointer features are in use.
Apple has implemented PACs in the Apple A12.
https://developer.apple.com/documentation/security/preparing_your_app_to_work_with_pointer_authentication
The documented method of stripping PACs from a pointer is to call
ptrauth_strip(), which ultimately emits an `xpaci` instruction, but
this option isn't available to the Breakpad processor not running on
the device. Instead, this patch selects likely address bits from
link register values by examining the address range of loaded modules.
Change-Id: I054bd1a03605719937fc85dcc8d8b9fe927f44be
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1713650
Reviewed-by: Mark Mentovai <mark@chromium.org>
Now decoding the OS name and exception codes for Fuchsia. Still not decoding exception flags (can be added later, if needed).
Change-Id: If66cb000828be18f0c1b35d1b1f52b3ca3e1fd67
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1699049
Reviewed-by: Mark Mentovai <mark@chromium.org>
- I think I previously fixed this for dump_syms, but must have missed
the symupload one because it didn't have a gn target.
Change-Id: Ibf4daa0dc874f329c2ee7c7b3d4de1ee6bc68d13
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1682717
Reviewed-by: Mark Mentovai <mark@chromium.org>
- This is a workaround to an issue with gn/clang build of Windows
symbol converter where dbghelp.dll is loaded from system32/syswow64
instead of alongside the process exe.
- Why do we care where dbghelp.dll is loaded from? Two considerations:
1. dbghelp.dll will only load symsrv.dll from the directory where it
resides.
2. symsrv.dll requires a file called "symsrv.yes" to be in the
directory where it resides in order to work with MS symbol stores.
Therefore if we load dbghelp.dll from syswow64, then we must also
ensure there is a symsrv.dll and symsrv.yes file in syswow64.
Change-Id: Ia283a2c11e276c855a48157aa7be77897af4b02e
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1680670
Reviewed-by: Mark Mentovai <mark@chromium.org>
Reviewed-by: Ivan Penkov <ivanpe@chromium.org>
- Only 64-bit PEs supported.
- Re-add some scripts that were missed in initial move of code.
- Change msdia120.dll dependency to msdia140.dll.
- Add tests for Intel, AMD, and NVidia Microsoft Symbol Stores.
- Windows symbol converter now attempts to fall back to PE-only metadata
when it fails to locate a PDB.
- Remove the 'binary' folder under converter_exe. Need to think more
about how a deployment should look and what tool(s) to use in creating
one.
Change-Id: I52e42cbe5e759874a25114c2483e8b50d73fdf77
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1670098
Reviewed-by: Ivan Penkov <ivanpe@chromium.org>
- Add a #define to testing.gyp to avoid warnings about TR1 deprecation.
- PESourceLineWriter now reads debug_file from CodeView record instead
of using code_file value.
- Updated PE-only MD reading unit test.
Change-Id: Ib4e6201df3e3fd651e160f310584b5a67b16c842
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1668347
Reviewed-by: Ivan Penkov <ivanpe@chromium.org>
- Rename two files from '.cpp' to '.cc' and insert missing comma in
between lines in source files list.
Change-Id: I169a738b565f52292cb0292b34c4397e9dc05370
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1666351
Reviewed-by: Ivan Penkov <ivanpe@chromium.org>
- First step, this is just enough to get it generating a msbuild
project with GYP, which in turn can build the executable.
- Tests need to be redesigned because there isn't an available server.
Change-Id: I45440fd32b3ede29666c127703bcd441f0e4288e
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1661134
Reviewed-by: Ivan Penkov <ivanpe@chromium.org>
Upstream googletest is dropping autotools support and the cmake
installer does not provide gtest-config.
BUG=chromium:940320
TEST=FEATURES=test emerge-lakitu google-breakpad
Change-Id: Icd51ffc48b39c24eebc27cc023e74a77e592de3e
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1660893
Reviewed-by: Mike Frysinger <vapier@chromium.org>
- Implement in common_windows_lib-- added class "PESourceLineWriter".
- Add command-line switch to tell dump_syms to use PESourceLineWriter.
Symbol data created this way will contain information to correlate the
module with ones found in minidumps, along with frame info that allows
much higher quality stack-walking in those minidumps.
- Significant refactor of PDBSourceLineWriter-- all code concerned with
extracting metadata from PE files has been moved into utility
functions. This is to allow sharing of this functionality with newly-
added PESourceLineWriter.
- Added a unit test to dump_syms for the PE-only scenario.
Change-Id: If0855f05d424d32d23f484995be5f34232179a37
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1525325
Reviewed-by: Ivan Penkov <ivanpe@chromium.org>
ELF modules are loaded in memory in several, possibly discontiguous,
segments. If the holes between segments are large enough, other things,
possibly other ELF modules may be mapped in that space. Crashpad
records the range of modules as the base address of the lowest mapped
segment to the high address of the highest mapped segment. This means
that when one module is mapped into a hole in another, it appears to
the Breakpad processor as overlapping modules. Module ranges are
relevant to the Breakpad processor during stackwalking for identifying
which module a particular program counter belongs to (i.e. mapping the
address to a module's text segment). This patch addresses this issue of
overlapping modules by truncating the range of the module with the
lower base address. A typical module's text segment is the first loaded
segment which would leave the text segment range unaffected. Module
producers can restrict the size of holes in their ELF modules with the
flag "-Wl,-z,max-page-size=4096", preventing other modules from being
mapped in their address range.
Properly contemplating ELF module address ranges would require
extensions to the minidump format to encode any holes.
crbug.com/crashpad/298
This patch also renames the concept of "shrinking down" (which
truncated the upper of two overlapping ranges) to "truncate upper".
Change-Id: I4599201f1e43918db036c390961f8b39e3af1849
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1646932
Reviewed-by: Mark Mentovai <mark@chromium.org>
The path NSCachesDirectory may change across app updates and sometimes
even across app launches. As a result, the Config-XXX files may end up
with an outdated path to the associated minidump file.
Change-Id: I0befde26b2ac406c154ce7c7e9be0063ee99892d
Bug:850379
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1592561
Reviewed-by: Mark Mentovai <mark@chromium.org>
This CL adds a result callback on report upload completion.
On failure, Breakpad deletes the configuration file and does retry to
upload a report.
Using this callback, the client will be able to log some metrics and to
act on upload failure.
Bug: 954175
Change-Id: I95a3264b65d4c06ba5d8dde8377440d23f1e2081
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1572661
Reviewed-by: Mark Mentovai <mark@chromium.org>
Chrome's test runner on Linux installs its own StackDumpSignalHandler
which swallows signals and doesn't re-raise them. This is sloppy, but
apparently there are reasons (https://crbug.com/551681). For
breakpad_unittests, it causes problems where a test process expects (via
waitpid()) to observe a child crash. Deal with those cases by
explicitly restoring the default signal handler.
In another case, Chrome's test runner seems to have been arriving at the
conclusion that it was to expect output from a child. Transitioning from
exit() to _exit() fixes this problem, and it's not necessarily a bad
idea to do this in post-fork() children without an execve() anyway.
Bug: chromium:949098
Change-Id: I5a6af0c2a09cd8eac9998358f6d5ea665288236f
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1575670
Reviewed-by: Mark Mentovai <mark@chromium.org>
Running this test on android-kitkat-arm-rel fails for some reason:
[ RUN ] LinuxCoreDumperTest.VerifyExceptionDetails
linux_core_dumper_unittest.cc:170: Failure
Expected: (0U) != (dumper.crash_address()), actual: 0 vs 0
linux_core_dumper_unittest.cc:178: Failure
Expected equality of these values:
2U
Which is: 2
info.size()
Which is: 0
[ FAILED ] LinuxCoreDumperTest.VerifyExceptionDetails (7 ms)
Disable it for now on Android until someone can look into it.
Bug: google-breakpad:791
Change-Id: I40a5e3dbeeb44e5eb0df187e61d55e07d8ad3613
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1546778
Reviewed-by: Mark Mentovai <mark@chromium.org>
Some of the fields we save might have signed types depending on the
system (e.g. `typedef int pid_t`). Depending on the toolchain, we
can trip -Werror=narrowing failures like:
src/client/linux/minidump_writer/linux_core_dumper.cc:248:66: error:
narrowing conversion of ‘(__pid_t)info->siginfo_t::_sifields.siginfo_t::<anonymous union>::_kill.siginfo_t::<anonymous union>::<anonymous struct>::si_pid’
from ‘__pid_t {aka int}’ to ‘long unsigned int’ inside { } [-Werror=narrowing]
set_crash_exception_info({info->si_pid, info->si_uid});
^^^^^^
src/client/linux/minidump_writer/linux_core_dumper.cc:252:71: error:
narrowing conversion of ‘(int)info->siginfo_t::_sifields.siginfo_t::<anonymous union>::_sigsys.siginfo_t::<anonymous union>::<anonymous struct>::_syscall’
from ‘int’ to ‘long unsigned int’ inside { } [-Werror=narrowing]
set_crash_exception_info({info->si_syscall, info->si_arch});
^^^^^^^^^^
Since the exception info fields are all uint64_t which should be large
enough to handle all the fields in the siginfo_t structure, add casts
for all the assignments to avoid these errors. We have implicit casts
even without them, so we aren't changing behavior.
Bug: google-breakpad:791
Bug: chromium:945653
Change-Id: Ib04e015998f08b857159ac13e9a065a66d228d49
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1544862
Reviewed-by: Mark Mentovai <mark@chromium.org>
Even 64-bit Mach-O (MH_MAGIC_64 = 0xfeedfacf) is not a fully 64-bit file
format. File offsets in sections are stored in 32-bit fields, with
Mach-O writers typically truncating offsets too large to fit to just
their low 32 bits. When a section begins at a file offset >= 4GB,
dump_syms would produce an error such as:
Google Chrome Framework.dSYM/Contents/Resources/DWARF/Google Chrome Framework: the section '__apple_names' in segment '__DWARF' claims its contents lie outside the segment's contents
As a workaround, this implements the strategy I first described in
https://crbug.com/940823#c22.
Segment file offsets are stored in 64-bit fields. Because segments
contain sections and must load contiguously, it’s possible to infer a
section’s actual offset by computing its load address relative to its
containing segment’s load address, and treating this as an offset into
the containing segment’s file offset. For safety, this is only done for
64-bit segments (LC_SEGMENT_64) where the 32-bit section offset stored
in the Mach-O file is equal to the low (truncated) 32 bits of the
section offset recomputed per the above strategy.
Beware that this does not provide full “large file” support for 64-bit
Mach-O files. There are other file offsets within Mach-O files aside
from section file offsets that are stored in 32-bit fields even in the
64-bit format, including offsets to symbol table data (LC_SYMTAB and
LC_DYSYMTAB). No attempt is made to recover correct file offsets for
such data because, at present, such data is always stored by dsymutil
near the beginning of .dSYM files, within the first 4GB. If it becomes
necessary to address these other offsets, it should be possible to
recover these offsets by reference to the __LINKEDIT segment that
normally contains them, provided that __LINKEDIT doesn’t span more than
4GB, according to the strategy discussed at the bottom of
https://crbug.com/940823#c22.
Although this is sufficient to allow dump_syms to interpret Chromium
.dSYM files that exceed 4GB, be warned that these Mach-O files are still
technically malformed, and most other tools that consume Mach-O files
will continue to have difficulties interpreting these large files.
As further warning, note that should any individual DWARF section exceed
4GB, internal section offsets will be truncated irrecoverably, unless
and until the toolchain implements support for DWARF64.
https://bugs.llvm.org/show_bug.cgi?id=14969
With this change, dump_syms is able to correctly recover file offsets
from and continue processing a .dSYM file with length 4530593528
(4321MB), whose largest section (__DWARF,__debug_info = .debug_info) has
size 0x8d64c0b8 (2262MB), and which contains four sections (starting
with __DWARF,__apple_names) beginning at file offsets >= 4GB.
Bug: chromium:940823, chromium:946404
Change-Id: I23f5f3b07773fa2f010204d5bb53b6fb1d4926f7
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1541830
Reviewed-by: Robert Sesek <rsesek@chromium.org>
Reviewed-by: Mike Frysinger <vapier@chromium.org>
- Baselines appeared to be stale. dump_syms now prints FUNC entries
with the full function signature, whereas the baselines only
contained the function name. The current state of the symbol file
docs
(https://chromium.googlesource.com/breakpad/breakpad/+/refs/heads/master/docs/symbol_files.md)
seem to agree with the new FUNC entries rather than the old ones.
Example of a name given in current docs:
"nsQueryInterfaceWithError::operator()(nsID const&, void**) const".
Change-Id: I9e01354cd82b7184b7cba31d132603e949a657ac
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1529133
Reviewed-by: Ivan Penkov <ivanpe@chromium.org>
If the kernel/C library headers are old, they might not have the
fields needed for SIGSYS decoding. Add ifdef checks for that and
skip the logic entirely. Easier than adding arch-specific siginfo
structs to the codebase.
Bug: google-breakpad:791
Change-Id: Ia473e3ffa61fce4c42cf4c1e73a9df044599bc5c
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1524447
Reviewed-by: Mark Mentovai <mark@chromium.org>
Many signals in Linux support additional metadata on a per-signal
basis. We can extract that from NT_SIGINFO and pass it through
in the exception_information fields.
The current core dumper logic doesn't set exception_information
at all, so this is an improvement.
Bug: google-breakpad:791
Change-Id: I38b78d6494e9bc682441750d98ac9be5b0656f5a
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1497662
Reviewed-by: Mark Mentovai <mark@chromium.org>
When building on an old system with outdated headers, this define
might not be available. Add a fallback to our existing elf header.
Bug: google-breakpad:790
Change-Id: I4dfe7a5cebd414cca3582a1a9cfc983503d5a779
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1507073
Reviewed-by: Mark Mentovai <mark@chromium.org>
The current failure message omits the underlying errno. This can
make diagnosing failures a bit difficult unless you run everything
through strace. For example:
$ core2md core /proc/self md
$ core2md core /proc/self md
Unable to generate minidump
Now we get the errno details:
Unable to generate minidump: File exists
Change-Id: I67f30879868ce4a726d5d888ee8c0a4a316b5186
Reviewed-on: https://chromium-review.googlesource.com/c/breakpad/breakpad/+/1497660
Reviewed-by: Ted Mielczarek <ted.mielczarek@gmail.com>
The current core dumper only parses NT_PRSTATUS notes. With signal
details, this note only includes three fields: signo, code, and errno.
We set exception_code to signo and exception_flag to code. The errno
value isn't set by the kernel, so there's no need to save it.
However, we never fill in exception_address which means all converted
crashes look like they happen at address 0. This implies a NULL jump
which is usually not the case, so it's just confusing. The prstatus
structure doesn't offer anything directly that tracks this.
Starting with linux-3.7, the kernel writes out the full siginfo
structure in the NT_SIGINFO note. So lets support that to pull out
si_addr which, for a bunch of common signals, is the value we want in
exception_address.
The size of the siginfo_t structure should be locked to 128 bytes at
build time for all architectures, so this should hopefully be stable.
Bug: google-breakpad:790
Change-Id: I458bad4787b1a8b73fad8fe068e9f23bec957599
Reviewed-on: https://chromium-review.googlesource.com/c/1497661
Reviewed-by: Mark Mentovai <mark@chromium.org>
TYPED_TEST_CASE is deprecated in modern googletest.
BUG=chromium:936654
Change-Id: I08004ffbb26089ebe17302934ed6d3268220d151
Reviewed-on: https://chromium-review.googlesource.com/c/1493423
Reviewed-by: Mark Mentovai <mark@chromium.org>
Matches current revision in Chromium.
BUG=chromium:936654
Change-Id: Ic856a805aadaf9901d906f11675f6b4c0c72e46f
Reviewed-on: https://chromium-review.googlesource.com/c/1493217
Reviewed-by: Mark Mentovai <mark@chromium.org>
