The original code assumed that the load bias in the program headers
would be exactly the same as in eh_frame/eh_frame_hdr/debug_frame.
This isn't guaranteed, so add a section bias for use when creating
a DwarfSection. In addtion, make the load bias and section bias
a signed value. There is no reason that this value needs to be positive,
so don't force it to be.
Add a new offline test that has a different load bias in eh_frame than
in the executable load.
Add additional unit tests to verify the load bias values are set properly.
Clean up the tests in ElfInterfaceTest, making all tests names follow the
same convention.
Bug: 141888859
Bug: 142094469
Test: New units and old unit tests pass on host and taimen.
Change-Id: Ib878123ab5545f0f315c749cfe0d27b012d873ee
The load bias value set in ReadProgramHeaders is out of sync with the
algorithm used in the static GetLoadBias function.
Sync the two and add tests to verify that they stay in sync.
Test: Unit tests pass.
Change-Id: I20ac0104970a22a92a5314a41dcadad0c9c22e64
Avoid accessing nullptr of already deleted entry.
Add new unit tests that pass with the fix and fail without.
Test: fixes unwinding in ART gcstress tests
Test: All unit tests pass.
Change-Id: Ideb00e2adc899904dd6aeb5dad3fb6fad150322d
Created a special target, libunwindstack_unit_test, that doesn't
include the test that dlopen's a shared library. It appears atest
doesn't understand how to handle the require keyword.
Also, move the shared library into the libunwindstack_test directory
itself.
Test: Ran atest libunwindstack_unit_test.
Change-Id: I967919b1d74a08669b61d0363d80861685725609
Add a specific test that __libc_init is the last frame in a stack
when run on device. In addition, it verifies that the return address
register is marked as undefined given the unwind.
Bug: 140008396
Test: New unit test passes on arm/arm64 (taimen device).
Test: New unit test passes on x86 (cuttlefish).
Test: New unit test passes on x86_64 (modified bionic/tests/run-on-host.sh)
Change-Id: Iefc151a7dbf52ab083c2bb78bad3d38b4e9e1254
I was using the pc as the offest into the elf. That is obviously not
correct. Added an optional OFFSET argument like in unwind_info along
with this change.
Test: Verified that with no offest works, verified with a zero offset
Test: works, verified with a non-zero offset results in a bad elf
Test: on an elf without an offset.
Change-Id: I4b6d02609627288e9f8a0eb26988d03adf95cb1f
Modify the MapInfoCreateMemoryTest to work in the isolated mode.
Test: Ran unit tests on host/target.
Change-Id: I84e01d96e852acd813e0f203b4a207cfaf8ca556
The previous versions of the libc++ demangler crashed on bad input.
However, the new version passes a fuzzer and has a lot of tests. Since
it's more complete than the local demangler, use it instead.
Modified the expected output of an offline test since the new demangler
handles a case that didn't work before.
Verified that the time it takes for the check_for_leak tests did not
change after this.
Bug: 136138882
Test: Ran the unit tests.
Test: Verified the __cxa_demangle function passes the fuzzer when run for
Test: hours. Both the 32 bit and 64 bit version of __cxa_demangle were
Test: fuzzed using external/libcxxabi/fuzz.
Change-Id: I10c06b589d57c36d89dbecba020b1ef2da69634a
Found when I noticed that the unit tests segfaulted every once in a while.
Test: Ran all unit tests 1000 times. The crash usually happens before 100 runs.
Change-Id: I1c8d2e3637400dc80f273f3677f4c94b0dbeac9d
There are many subclasses of the Memory class and the overwhelming
majority of them don't need to be exposed externally. We move all of
them to internal headers except MemoryOfflineBuffer, which moves to a
separate header. This dramatically reduces the exposed API surface and
makes the code more modular.
Also, remove the Offline code from libbacktrace. It's not used any where.
Test: Unit tests pass, clean tree still builds
Change-Id: I55dacdf080daba0bfe65c1ad53a4b326bb482e83
When adding a dex frame, no offset is ever printed even if the dex data
is coming from a non-zero offset map. Fixed so an offset is printed
in this case.
Previous the line looked like this:
#17 pc 0000000000500d7a /product/app/GoogleCamera/GoogleCamera.apk (com.google.android.libraries.camera.async.AndroidPriorityThread.run+10)
Now looks like:
#17 pc 0000000000500d7a /product/app/GoogleCamera/GoogleCamera.apk (offset 0x11d0000) (com.google.android.libraries.camera.async.AndroidPriorityThread.run+10)
Add new unit test for this case.
Bug: 134420076
Test: Wrote new unit test, and verified a non-zero offset for dex file
Test: data results in a frame with a non-zero offset.
Change-Id: I58c134fda4fad5456ca0f1723192667a1ca5d509
In ART, some of the maps are /memfd:/jit-cache and it triggers the warning
about unreadable elf files. Do not set the elf from memory not file
flag in this case.
Bug: 131909548
Test: New unit tests pass.
Test: No warnings dumping stacks with this change done.
Change-Id: Ifba5e65da609525ded75430da173c614f6e4801e
Modify the unwinder library to indicate that at least one of the stack
frames contains an elf file that is unreadable.
Modify debuggerd to display a note about the unreadable frame and a possible
way to fix it.
Bug: 129769339
Test: New unit tests pass.
Test: Ran an app that crashes and has an unreadable file and verified the
Test: message is displayed. Then setenforce 0 and verify the message is
Test: not displayed.
Change-Id: Ibc4fe1d117e9b5840290454e90914ddc698d3cc2
This refactors the step function slightly to split it up into
distinct pieces since the code needs to handle a signal handler
versus normal step slightly differently.
Add a new error for an invalid elf.
Modify libbacktrace code to handle new error code.
Bug: 130302288
Test: libbacktrace/libunwindstack unit tests.
Change-Id: I3fb9b00c02d2cf2cc5911541bba0346c6f39b8e6
Update unit tests and add new build id displaying in offline unwinds.
Bug: 129873279
Test: All unit tests pass.
Test: Verify that debuggerd displays build id properly.
Change-Id: I97f4a204842447a20c812f535a458155b937d5e1
This is not full support for dwarf 5, this merely treats a Dwarf 5
version as Dwarf 4. There are new dwarf ops that are not supported yet,
but this minimally support should allow unwinding to work if those
ops are not present.
Bug: 127355724
Test: New Unit tests pass.
Change-Id: I35b24fbcb15a64acd49e1e0b6890dff4456ee6fd
Since there has never been a time that a sequential search has been
necessary, simply remove the sequential code and allow the DwarfEhFrame
object to be used in that case. If this does come up regularly, I'm not
sure that it wouldn't be more performant to use the DwarfEhFrame object
regardless.
Bug: 129550015
Test: libunwindstack unit tests, libbacktrace unit tests.
Change-Id: I738e81accb9bc1e182168cddc9c0c22823e3369b
Update the entries only when the list is modified by the runtime.
Check that the list wasn't concurrently modified when being read.
Bug: 124287208
Test: libunwindstack_test
Test: art/test.py -b --host -r -t 137-cfi
Change-Id: I87ba70322053a01b3d5be1fdf6310e1dc21bb084
Update debuggerd to print BuildId information by default.
Bug: 120975492
Test: New unit tests pass.
Test: debuggerd -b <PID> shows build id information.
Test: tombstones include build id information.
Change-Id: I019b031113d0b77385516223c63455b868924440
Changes:
- Change GetSoname to always returns a std::string.
- Added new unit tests for the soname printing.
- Modify the GetElf() function to save the same elf when we see rosegment
linkers that split the read-only and read-write across a map. This
avoids creating multiple elf objects for each map.
- Fixed a few offline unwind tests.
Bug: 29218999
Test: Unit tests pass.
Change-Id: Iad7c38b5c2957a8c5fd4ba94ebec335bafcad57d
The elf_start_offset field in a map_info was not set in one path.
Added new offline test for this, and update a few unit tests.
Bug: 128442792
Test: New unit tests pass.
Change-Id: I2c6ac1b5271a99aa0e8c4b6342a5970199185112
Currently, moving or copying a Maps object leads to double free of MapInfo.
Even moving a Maps object did not prevent this, as after a move
the object only has to be in an "unspecified but valid state", which can
be the original state for a vector of raw pointers (but not for a vector
of unique_ptrs).
Changing to unique_ptrs is the most failsafe way to make sure we never
accidentally destruct MapInfo.
Test: atest libuwindstack_test
Failed LocalUnwinderTest#unwind_after_dlopen which also fails at master.
Change-Id: Id1c9739b334da5c1ba532fd55366e115940a66d3
It should be dlopen'ed lazily by libdexfile_support now.
Also change debuggerd_test to not link libunwindstack and its dependencies
statically - the static libs can overlap with the dynamic ones.
Test: mmma system/core/debuggerd/
Test: atest debuggerd_test
Test: mmma system/core/{libunwindstack,libbacktrace}, run host gtests (cannot get atest to work)
Bug: 124827589
Bug: 123186083
Change-Id: I9e7bf9bcbae499af4e1be4c9854bce441e2a7b55
Adjust all callbacks to include the inode number as well.
Fixes: 123532375
Test: libprocinfo_test
Test: libdmabufinfo_test
Test: libmeminfo_test
Change-Id: I71fd75d90d5ee44585011c5ae654a1dd7e185458
Signed-off-by: Sandeep Patil <sspatil@google.com>
Added a unit test that fails before the change and passes afterwards.
Bug: 120606663
Test: All unit tests pass.
Change-Id: I054c7eac0c55abc3babe1d48a041f5819ad9db81
The rosegment changed the load segments and broke the unwind_info printing
of unwind information for the arm exidx frame.
Test: Fixed and verified by dumping the arm exidx frame.
Change-Id: I7270175e4dcb7fee2b96a16d168880ebd09f19b1
The GetBuildID() function returns the raw build id data, so add a function
to get the printable hex version of the data.
Bug: 120606663
Test: New unit tests pass.
Change-Id: Ia5aefc97457efb08bbd30ea96cbb2d47ae59f954
These tests are meant to replace the leak check tests from libbacktrace.
Bug: 120606663
Test: Ran tests on host and target.
Change-Id: I928b199304afc36b4bac78e9a2cd688b6f2910b9
There are currently two copies of DwarfOp callback tables in
data.rel.ro section, and that is 16KB of dirty memory per process on a
64-bit platform.
To reduce memory usage, we introduce a layer of indirection so that
all instances of DwarfOp can share a single callback table. We also
get rid of the pointers for op names in the callback table so that the
entire table can be placed in rodata section.
Test result shows 1MB saving on cuttlefish, and 750KB saving on a Go
device (32-bit ARM).
Bug: 112073665
Test: Boot on cuttlefish and a Go device.
Change-Id: Idf90a513f37534ab1ef30d5742a9d3ea779f3661
Actually make the destructor virtual so that objects in
UnwinderFromPid get freed.
Also, explicitly make the DexFiles and JitDebug destructors virtual.
Bug: 123099840
Test: Ran UnwindTest.local_from_pid thousands of times.
Change-Id: I8bf11c0eb3587fa82f5f14f4723ea5c91da0fd46
The signal tests that send a SIGUSR1 signal to a remote process, can
lead to a case where the code winds up in a location for which the
unwind information is not 100% accurate. This happens near the
end of the atomic functions which manipulate the stack right before
returning, but do not have separate cfi instructions for those last
few instructions.
Add a new test type for the from_context test that will wait explicitly
for the global g_finish and not piggy back on the remote test type.
Bug: 122902885
Test: Ran flaky tests 1000 times without failure on a taimen.
Change-Id: I7d0c9b016e3a3a2c0c0949cf74af11d7785a008c
Change the GetBuildID function to return a std::string.
Added benchmark to check how long it takes to get the build id from
a file versus an elf object.
Added a way to get an elf without passing in a valid process_memory and
added tests for this.
Test: New unit tests.
Change-Id: I3029019767e0181c758d611fe635bc1bf72d6e8e
This object is able to be easily used from other code and can be used
to replace the libbacktrace calls in other parts of the platform.
Also, demangle the function names when calling FormatFrame.
Bug: 120606663
Test: Unit tests pass, debuggerd using this code directly passes unit
Test: tests.
Change-Id: Ifd8cf9bdd89174c1736810711d20e9f37f29b1bf
Also remove some tests for implementation details in the libdexfile API that
are now covered by unit tests there (http://r.android.com/867370).
Test: mmma system/core/{libunwindstack,libbacktrace} and run host gtests
Bug: 119632407
Change-Id: Ie731bb929a564de56ce1c78385da12a43d0954ff
Test: m
Test: mmma system/core/{libunwindstack,libbacktrace} and run host gtests
Bug: 119632407
Bug: 120978655
Change-Id: I919586ab503be3617e8a0604dfe48db331e7af94
The CHECK(encoding != DW_EH_PE_aligned) can trip given the right
arguments. This check isn't necessary, since the code will return
false in that case.
Add new unit test that tries all values to make sure no CHECK fires.
Bug: 120968571
Test: Passes new unit test, and passes fuzzing that failed before.
Change-Id: I062bcd18508c75cd3a4ca9dd12f922e25aafda8a
This reverts commit 444e23d2fc.
The rest of the topic doesn't need to be reverted.
Reason for revert: Breaks renderscript on marlin and sailfish.
Test: Manual repro of http://b/121110092#comment1 on reported branch
Test: "atest CtsRenderscriptTestCases" on that branch
Test: mmma system/core/{libunwindstack,libbacktrace}, run host gtests
Test: Make image, flash, and reboot device.
Bug: 121110092, 119632407
Change-Id: If1976b19ce386c95bc5bd4fd6d523745c167de18
The linker was modified so that a shared library has a read-only
map, then a read-execute map to represent the whole shared library.
When backtraces are created, then the offsets are set incorrectly
for backtraces. For example, all backtraces wind up with an offset now,
and a shared library loaded out of an apk shows the wrong offset.
Changes:
- Fix the FormatFrame function which was putting the offset before the
map name.
- Refactor the Maps and MapInfo classes to keep track of the previous map
instead of all maps. This was the only map that was ever needed.
- Modify the unwind_for_offline tool to capture both the read-only and
read-execute map when using the read-only segment option.
Bug: 120981155
Test: Updated unit tests pass.
Test: Unwinds on device don't show the offsets everywhere any more.
Change-Id: I75b3727221be9c20316dfdcd7a1c6d242d7ce948
Pages in .data.rel.ro are always dirty. Move whatever we can to
.rodata so that we reduce memory pressure.
The size of rodata, text, and data.rel.ro sections of linker64 on
cuttlefish before/after this change:
rodata: 0x170e4 -> 0x17964 (+2176 bytes)
text: 0xec919 -> 0xec919 (no change)
data.rel.ro: 0x9cf0 -> 0x98f0 (-1024 bytes)
Test: Boot cuttlefish
Change-Id: Ib2fd2bb0afb50208d441731d8c31bc8f287272c3
If the first frame of an unwind is a totally invalid pc that's not in
any map, a speculative frame is added. Rather than deleting this frame
if no more unwinding is possible, leave it. This fixes a case where
the only frame you get is an invalid one, but the speculative frame
winds up in a shared library or somewhere else and gets removed.
Bug: 120505086
Test: New unit tests to catch this case pass.
Test: Verified original crashing program now emits two backtrace lines.
Change-Id: I088dff21c057386dcdaeb3fc2578b24322683bd0
When a shared library is loaded directly from an apk, the new way the
linker splits a shared library into a read-only and execute segment
broke unwinding. Modify the code to handle this case.
Other changes:
- Modify the algorithm for finding read-only map entries. Before, the code
would search the entire map for the closest offset. Now it simply looks
at the previous map. I did this because the old code was too lenient and
might still work even if the linker changes. I want this to break if the
linker behavior changes so that I can analyze the change.
- Update the tools to use PTRACE_SEIZE instead of PTRACE_ATTACH since
PTRACE_ATTACH doesn't work in all cases.
- Small refactor of the GetFileMemory function.
- Add new unit test cases and new offline unwind test cases.
Bug: 120618231
Test: Ran new unit tests, ran original failing test.
Change-Id: I4bade55cf33220d52f1d5e9b0cbbbcc8419669d4
This will be used by heapprofd to allow us to correlate build ids with
memory leaks in libraries and binaries.
Test: m
Test: host libunwindstack_test
Test: run unwind_info against my phone's libc.so (32/64) and compare
to readelf Build ID output.
Bug: 120186412
Change-Id: I3cefd6cce9a8733509bf35b7175eb0f967783477
Remove unit test that is not needed any more.
Slightly reorganize the code around the changed check.
Bug: 77958880
Test: Ran backtrace tests with new linker options.
Test: Ran libunwindstack unit tests.
Change-Id: Ie95707323fb51616a093cafb83e0343fa24e7e61
To avoid a case where a malicious app might try and trick the system to
create an elf and register object that mismatches, always verify that they
are the same arch.
Test: Ran unit tests.
Change-Id: I66978e9e02f8e4f396856912e7019528ead4838e
Every real call always sets the value to true to init the gnu_debugdata
interface, so remove this parameter.
Test: Builds, unit tests pass.
Change-Id: I07fb3adbbd5b65b51c0dc7608561e820a5095051
If multiple threads are unwinding at the same time, new maps that contain
the global variables for dex files and jit information are created. This
leads to threads creating more new maps that then get searched, then
more maps, then more searching until virtual address space exhaustion.
Fix this so that we only search maps that have a corresponding rw map that
could contain the global memory.
Small refactor to combine the code to search for global variables into
one class that both classes inherit from.
Modify unit tests for the new pattern checking.
Bug: 117761427
Test: Ran unit tests for libunwindstack/libbacktrace/simpleperf.
Test: Ran art 004-ThreadStress that used to fail.
Change-Id: I837ca6b9d0383100079de090bc7d019598e0cdfe
When adding a frame with a dex pc, two frames will be added total. However,
if there is only enough room for a single frame, two get added any way.
Only add a single frame in this case, and add a unit test for this case.
Test: Passes unit tests.
Change-Id: If320584b126967a042c623d8fdf3f51dbc1c2251
The function should have returned a bool, so now it does.
Also use a different lighter weight function for validating the elf
in one place that didn't need to keep the max size value.
Test: Unit tests pass.
Change-Id: Ibde674e608091ba04abf22fad2fdc4dbdb2c6e73
The new linker base map is read-only, but the global handling code
for jit and dex information assumed the base map was a read-execute
one. Relax that requirement to search read-only maps.
Adjust the unit tests for this case.
Bug: 117293117
Test: Passes unit tests.
Test: Passes art cfi tests with and without llvm linker.
Change-Id: I31a765d51847d0b8d778ace9cbaa29f42073f4d9
The rosegment linker option results in two maps containing the elf data
existing. One is an execute map where the code lives, and the other is the
read-only segment which contains the elf header information. If the file
backing a shared library in memory is not readable, then the new code
will attempt to find the read-only map that has the same name as the
current execute segment, and that is at offest zero in the file.
Add new unit tests for this functionality.
Add the missing MapInfoCreateMemoryTest.cpp to the list of tests.
Bug: 109657296
Test: Pass new unit tests.
Test: All unit libbacktrace/libunwindstack tests pass with rosegment enabled.
Change-Id: If8f69e4a067d77b3f2a7c31e2e5cd989a0702a8c
Add FALLTHROUGH_INTENDED for clang compiler.
Bug: 112564944
Test: build with global -Wimplicit-fallthrough.
Change-Id: I40f8bbf94e207c9dd90921e9b762ba51abab5777
In preparation for removal of ClassDataItemAccessor.
Bug: 79758018
Test: make
Test: test/run-test --host 137
Test: out/host/linux-x86/nativetest/libunwindstack_test/libunwindstack_test
Change-Id: I61a20fe7200a0d669b84573d3853b55877922602
No new test failures introduced.
atest libunwindstack_test on aosp_walleye-userdebug:
Without patch:
Total: 800, Passed: 781, Failed: 19
With patch:
Total: 800, Passed: 781, Failed: 19
Change-Id: I8853c3e1594799761b481464b75e2904b38c4830
Here is the allowable issues with an elf file that will not result in an error:
- The program headers/section headers offset points to unreadable memory.
- Allow missing program header and/or section headers.
- Allow a symbol table section header to point to invalid symbol table values.
There is no real reason to require the elf file be perfect. Everything in
the code has sane defaults, so any missing information won't cause any
problems.
This gets rid of the warning that occurs any time an elf is loaded
from memory. In memory elf files never contain all of the section headers,
and do not contain the symbol table data.
Update tests to test these new cases.
Test: Builds and unit tests all pass.
Change-Id: Iaefe2cd6b6c965a01ed425a112d6afae339f3b78
Create fast lookup table instead of iterating every single time.
This will create the cache as methods are searched for.
Test: 137-cfi
Change-Id: I4be190bb1a637fef5d385b993be6a7e2203a6814
In almost all cases, it is faster to read the entire structure rather
than do multiple reads using ReadField. The only case where it would be
slower is if doing a remote unwind and ptrace is the only way to read. In
all other cases, it's a single system call. In the ptrace call, it will be
multiple calls. Given that it is unusual to be forced to use ptrace,
it's better to avoid it.
It also reduces the code complexity to do a single read, and avoids
issues where the code forgets to read the field it needs.
Test: Unit tests pass on host and target.
Change-Id: I7b3875b2c85d0d88115b1776e1be28521dc0b932
There's still <cutils/atomic.h> in a test, but I don't understand why
that isn't just std::atomic.
Also add a shared tgkill wrapper to libbase.
Bug: N/A
Test: ran tests
Change-Id: Idd4baa1e1670a84b3a8f35803cc5ffe5aae008a6
Modify the code for the no header sections because it turns out that
it is not okay to assume that the fdes are non-overlapping. It's necessary
to read the fdes in order and match as you go.
Modify the code so that it only reads until it finds the given pc rather than
reading all of the cie/fde entries at once.
Rewrote the tests to verify the new behavior.
Bug: 68998033
Bug: 110235461
Test: Ran libbacktrace/libunwindstack unit tests.
Test: Unwind the mediaserver process on a walleye and verify it
Test: unwinds properly.
Change-Id: I7bb59d1db72c13fa34caa9735ec34c1a60e20ed2
Before, I was using p_vaddr to get the offset into the elf file where
the exidx frame starts. I changed that to use p_offset since this already
has the load bias offset in it and some elf files do not set p_vaddr
properly.
Also, use p_filesz instead of p_memsz, since again, some elf files do
not set p_memsz to the same as p_filesz.
Bug: 110704153
Test: All libbacktrace/libunwindstack unit tests pass.
Test: Randomly unwind process on a walleye.
Test: Verified that this properly dumps and unwinds the shared
Test: library that sets p_vaddr and p_memsz differently.
Change-Id: Ic7b1e5d07439f4636fa02cd884a8727a5737372b
It turns out that for the dwarf information, if a FDE indicates it's pc
relative, then pc has to be incremented by the load bias. If not, then
it should not be incremented.
Previously, the code always subtracted load bias values from pcs, and assumed
that all fdes were incremented by load bias values. The new code actually
reads the fdes and adjusted the pcs in the fde and in the eh frame hdr so
that load bias values are already handled properly.
In addition, add dumping of arm exidx values in unwind_reg_info. This allowed
verifying that the debug frame in those elf files was being handled properly.
Added a new unit test that only has a debug frame that has a non-zero load
bias and has fde entries that do not have pc relative encoding.
Fix a couple of other small bugs.
Bug: 109824792
Test: All libbacktrace/libunwindstack unit tests pass.
Test: Ran ART 137-cfi test and 004-ThreadStress.
Test: Verify that displaying the fde start and end pc actually match the
Test: real data for fde that have pc relative set, and that don't.
Test: Verified that the unwind information for arm exidx matches the
Test: debug frame data.
Change-Id: I707555286b5cb05df9f25489e8c5ede753cfe0fb
