This reverts commit b9d0753d2b.
Reason for revert: Re-land with MacOS workaround.
Test: Build (on Linux) and boot AOSP, with weak symbols enabled and disabled.
Change-Id: I5150cd90367178f3b039761dca3bccc9c2987df1
This reverts commit 9d3146af22.
Reason for revert: It appears that weak symbols don't work as expected on MacOS, breaking the MacOS aapt build.
Change-Id: Ica0955106485a7bf2e2c3f09ff7910e230eb4139
Move towards crashing if a normally configured RefBase object is
destroyed without ever incrementing the reference count. We've been
threatening to do this for a long time. The previously last known
violation had been fixed.
This also fixes stack trace printing from RefBase, which had previously
been broken, and which we found necessary to track down further
violations of this rule.
Unfortunately, we found several more violations with the aid of
that fix. After existing CLs are submitted, there are
still some failures, but they are no longer numerous. Thus this CL
doesn't actually crash in the event of a violation, but does log a
verbose stack trace if it encounters one.
Bugs have been filed against the remaining known RefBase client offenders.
We plan to enable crashing on usage violations once those are fixed.
The fix for the stack trace printing breakage unfortunately requires
the use of weak symbols in order to avoid a circular build dependency.
We expect to eventually replace this with execinfo.h functionality.
Some random reformatting, driven by consistency with current formatting
requirements.
Add missing include to BacktraceMap.h.
Bug: 79112958
Bug: 30292291
Test: Boot AOSP, Master
Change-Id: I8151c54560c3b6f75ffc4c48229f0388a2066958
Add test and benchmark.
Also switch libbacktrace, libunwindstack, libmemunreachable
to use libprocinfo for map file reading.
The benchmark shows using libprocinfo speeds up map file reading
in libbacktrace and libunwindstack 18% - 36% on walleye.
Bug: http://b/79118393
Test: run procinfo_test.
Test: run libunwindstack_test.
Test: run libbacktrace_test.
Test: run memunreachable_test.
Change-Id: Icf281c352f4103fc8d4ba6732c5c07b943330ca1
In cases where there might be a crash in the unwind library itself,
we need a method to avoid skipping these frames or we won't be able
to see the actual crash.
Added unit test for this behavior.
Bug: 74121887
Test: Ran unit tests on host and target.
Change-Id: I45825020c174016af39dd8ffdc67acb72a24ad4d
Changes:
- Add a new flag to the libbacktrace and libunwindstack map data.
- Modify the unwinder to handle this map to use the raw pc when stepping.
- Add new unit tests for this case.
Bug: http://b/73127105
Test: Run simpleperf to unwind through jit symfiles.
Test: Run new unit tests.
Test: Run 137-cfi test on host.
Change-Id: I10bc0410680accc6d35fe51e9f1098911f667e01
- Add a new function Backtrace::UnwindOffline that takes the stack data.
- Modify BacktraceMap::CreateOffline so it doesn't take the stack data.
This makes it easier to reuse the map object created this way.
Reusing the map object increases simpleperf speed (unwinds per second) by 50%.
Test: backtrace_test libunwindstack_test
Change-Id: I90cfbae9e50d95d8a0e3cd394b33ba36d65d45f7
libbbacktrace changes:
- Completely rewrite the BacktraceOffline class to use the new unwinder.
- Modify the test data to save ucontext_t data instead of unw_context data.
- Convert the previous tests from unw_context data to ucontext_t data.
Bug: 65682279
Test: New unit tests pass in libunwindstack.
Test: All offline tests continue to pass.
Change-Id: I540345c304b20199d46deeb0349a0638a0f3ab2f
Use the art dex file library to read the dex data.
Add unit tests for the UnwindDexFile code.
Bug: 72070049
Test: All unit tests continue to pass.
Test: Dumped the backtrace of the 137-cfi test while running in interpreter
Test: mode and verified that the stack trace is correct. Did this on host
Test: and for arm/arm64.
Change-Id: Ia6f343318c5dd6968a954015a7d59fdf101575b0
In order to support the offline unwinding properly, get rid of the
usage of non-fixed type uintptr_t from all API calls.
In addition, completely remove the old local and remote unwinding code
that used libunwind.
The next step will be to move the offline unwinding to the new unwinder.
Bug: 65682279
Test: Ran unit tests for libbacktrace/debuggerd.
Test: Ran debuggerd -b on a few arm and arm64 processes.
Test: Ran crasher and crasher64 and verified tombstones look correct.
Change-Id: Ib0c6cee3ad6785a102b74908a3d8e5e93e5c6b33
Changes:
- Change the register type from int16_t to uint32_t for the location data
and the Eval processing. This is because the special dex pc is > 65535.
- Add the ability for Dwarf register location information to point to a
register that is itself a Dwarf location register.
- Add dex_pc to the frame information.
- Modify the unwind tool to print the dex pc if non-zero.
This does not implement the printing of the dex information in anything
but the unwind tool. It's not the final form of this printing.
Bug: 72070049
Test: Ran new unit tests.
Test: Dumped stack while in interpreter running 137-cfi art test and
Test: verified dex pc is set to non-zero.
Change-Id: I6ce8a6b577fb4f92abacbd433b1f68977e272542
1. Extend MAX_BACKTRACE_FRAMES to 256, as we can have callchains
with length near 256 when recording Android applications.
2. .eh_frame and .ARM.exidx doesn't know how to unwind from instructions
setting up or destroying stack frames. It can lead to wrong callchains,
which may contain pcs outside executable mapping areas. Stop unwinding
once this is detected.
3. Some CIE entries in .eh_frame/.debug_frame suggest unwinder reading
personality routine from memory. However, that is not available in
offline unwinding, and leads to ACCESS_MEM_FAILED error. Work around
this by returning all zero data when failed to read data for some
addresses.
4. Some libraries have empty .eh_frame_hdr but valid .eh_frame.
Work around this by building eh_frame_hdr manually.
5. Prefer to use .debug_frame than .eh_frame. Because the former
can unwind instructions setting up and destroying stack frames.
Bug: http://b/69383534
Test: run backtrace_test.
Test: run simpleperf on some apps.
Change-Id: I62dddda3865dd567c6be150ce5527dc2d3516256
This is to help debugging different offline unwiding failures.
Bug: http://b/69383534
Test: run backtrace_test.
Change-Id: I5ed4837027a9f17d032925e97e9f5927161444b3
Add a static GetLoadBias method to the Elf object that only reads just
enough to get the load bias.
Add a method to MapInfo that gets the load bias. First attempt to get
it if the elf object already exists. If no elf object was created, use
the new static method to get the load bias.
In BacktraceMap, add a custom iterator so that when code dereferences
a map element, that's when the load bias will be retrieved if it hasn't
already been set.
Bug: 69871050
Test: New unit tests, verify tombstones have non-zero load bias values for
Test: libraries with a non-zero load bias.
Change-Id: I125f4abc827589957fce2f0df24b0f25d037d732
When doing a remote unwind, and there is a crash in the unwind stack
code or in libbacktrace, the old version would skip those frames.
This makes debugging crashes in those libraries a bit difficult.
Bug: 68820611
Test: Ran unit tests.
Test: kill -3 a zygote process and verify no frames in
Test: libunwindstack.so/libbacktrace.so.
Test: Forced a crash in libunwindstack.so and verified the tombstone
Test: contains the full crash.
Change-Id: I297e5b3775c87c0dddf57ebb423bd72ba1baf193
Update the backtrace leak tests to share a map since this is the most
common way it will be used, and it runs much faster.
Remove the CreateNew functions, and references to them.
Remove benchmarks of CreateNew functions.
Test: Builds, ran new unit tests, ran art tests.
Change-Id: I4a25a412f1a74c6d43aebbebbf32ead20ead8f94
Also, change the std::set parameters to std::vector. As jmgao points out,
a small std::set is not really the best choice for performance reasons.
Test: All unit tests pass, enabled the new unwinder and did a kill -3 on
Test: an android process.
Change-Id: I81227d7b79a9b7cf1d54fb0e3331d3cf4d4d3c4f
Tombstones (especially ones with lots of VMAs) are regularly truncated.
We can at least show the number of VMAs, though, for anyone interested
in knowing whether they got close to the default 64Ki limit.
Bug: http://b/66911122
Test: ran crasher, examined tombstone
Change-Id: I286db66f28f132307d573dbe5164efc969dc6ddc
Backtrace::FormatFrameData doesn't depend on anything from its
instance. Make it static so that callers that don't have a Backtrace*
available can format frames.
Test: mma
Change-Id: Id0ca93bbef19da488305374e58b25c7ca9d0e518
Also add a comment to the GetElf function to indicate that it never returns
nullptr.
Also needed to add this library to the a million and one places that the vndk
has hard-coded this data.
Bug: 23762183
Test: Built, nothing uses the new code.
Test: However, I did run backtrace_test using this code, and all tests pass.
Change-Id: Ib270665dcb7a7607075e36d88be76dbde6e2faa8
(cherry picked from commit dc4104b720)
Add NOLINT comment to work around clang-tidy
error in checking macro arguments used in
type expressions.
Bug: 28705665
Test: make with WITH_TIDY=1 WITH_TIDY_CHECKS=-*,misc-macro-* \
WITH_TIDY_FLAGS=-header-filter=system/core/.*
Change-Id: I7619978c1804e151a11a8b0477e80076bcf21cab
Also add a comment to the GetElf function to indicate that it never returns
nullptr.
Bug: 23762183
Test: Built, nothing uses the new code.
Test: However, I did run backtrace_test using this code, and all tests pass.
Change-Id: I252b9c2497e2d3d94347dd6e506170bf50cbfe16
- Change the field name load_base to load_bias (which is what it really is).
- Add a rel_pc field so that callers do not need to compute it themselves.
- Remove the BacktraceMap::GetRelativePc() since nobody should need to
compute this themselves.
Bug: 23762183
Test: Compiles and unit tests pass (debuggerd, libbacktrace).
Change-Id: I2cb579767120adf08c407a58f3c487ee3f2b45fc
