Adding the ARM RT sources to the CMake files, and enabling some
sanitizers to also build on ARM. This is far from supported or
production quality, but enabling it to build will get us errors
that we can actually fix.
Having said that, the Compiler-RT and the Asan libraries are
know to work on some variations of ARM.
llvm-svn: 200546
This change is a part of refactoring intended to have common signal handling behavior in all tools.
This particular CL moves InstallSignalHandlers() into sanitizer_common (making it InstallDeadlySignalHandlers()), but doesn't enable default signal handlers for any tool other than ASan.
llvm-svn: 200542
Also rename internal_sigaction() into internal_sigaction_norestorer(), as this function doesn't fully
implement the sigaction() functionality on Linux.
This change is a part of refactoring intended to have common signal handling behavior in all tools.
llvm-svn: 200535
Express the strto* interceptors though macros. This removes a lot of
duplicate code and fixes a couple of copypasto bugs (where "res" was declared of
a different type than the actual return type). Also, add a few more interceptors
for strto*_l.
llvm-svn: 200316
This change is a part of refactoring intended to have common signal handling behavior in all tools.
Note that this particular change doesn't enable use_sigaltstack support in every tool.
llvm-svn: 200310
allow SIGABRT to spoil errno, because some real programs
reset SIGABRT handler in the handler, re-raise SIGABRT and return from the handler
llvm-svn: 200304
It was redundant (since ARM mode is the default) and misleading since
(e.g.) Cortex-A15 would not satisfy the #ifdef but would be in ARM
mode regardless.
llvm-svn: 200043
remains ARM mode only, supporting thumb requires explicit it prefixes
for the predicted adds/subs and adjusting the offset computation for the
different block sizes.
llvm-svn: 200035
We left ignore_interceptors>0 when calling signal handlers
from blocking interceptors, this leads to missing synchronization in such signal handler.
llvm-svn: 200003
test-and-subtract loop and compute the initial block as address,
shaving off between 5% and 10% on Cortex A9 and 30%+ a Raspberry Pi.
Code written by Matt Thomas and Joerg Sonnenberger.
Differential Revision: http://llvm-reviews.chandlerc.com/D2595
llvm-svn: 200001
Currently correct programs can deadlock after fork, because atomic operations and async-signal-safe calls are not async-signal-safe under tsan.
With this change:
- if a single-threaded program forks, the child continues running with verification enabled (the tsan background thread is recreated as well)
- if a multi-threaded program forks, then the child runs with verification disabled (memory accesses, atomic operations and interceptors are disabled); it's expected that it will exec soon anyway
- if the child tries to create more threads after multi-threaded fork, the program aborts with error message
- die_after_fork flag is added that allows to continue running, but all bets are off
http://llvm-reviews.chandlerc.com/D2614
llvm-svn: 199993
warning: ISO C99 requires rest arguments to be used [enabled by default]
INTERCEPTOR(char *, dlerror) {
warning: invoking macro INTERCEPTOR argument 3: empty macro arguments are undefined in ISO C90 and ISO C++98 [enabled by default]
llvm-svn: 199873
New/delete implementations in system libraries almost always are built without
frame pointers. As we switched to frame pointer based unwinder on ARM, they no
longer work for us, resulting in broken allocation/deallocation stacks.
Note that this does not work with statically linked
libstdc++/libc++/libstlport.
llvm-svn: 199872
Clang now uses the *-apple-macho triples for embedded targets using MachO. The
environment makes less pretences to be running an actual Mach kernel with all
the support that would imply, so a few changes are needed to the build scripts:
+ The Apple way of detecting endianness applies in all cases so drop the
__MACH__ test.
+ A quick rename since "darwin_embedded" no longer makes any sense. If there's
no Mach going on, there's certainly no Darwin.
+ Change how we test for support for a platform. The functions we build don't
need any SDK support (stdio.h, ...) so we don't need to check for it.
llvm-svn: 199731
Intercept and sanitize arguments passed to printf functions in ASan and TSan
(don't do this in MSan for now). The checks are controlled by runtime flag
(off by default for now).
Patch http://llvm-reviews.chandlerc.com/D2480 by Yuri Gribov!
llvm-svn: 199729
This code works with AAPCS-style {fp, lr} frames. Unlike linux perf, it does
not unwind through code compiled with -mapcs-frame (which Clang does not
support anyway).
llvm-svn: 199725
In optimized hybrid execution we do not use DynamoRIO private loader, which
mangles TLS access, so we can access the application's TLS directly.
Patch by Qin Zhao.
llvm-svn: 199655
This change allows to compile sanitizer sources so that *san runtime
will attempt to use libbacktrace and/or libiberty for symbolization
(instead of communicating with llvm-symbolizer).
I've tested this patch by manually defining SANITIZER_LIBBACKTRACE and/or
SANITIZER_CP_DEMANGLE, linking with necessary libraries and verifying that
all tests from ASan test suite work.
Based on patches by Jakub Jelinek!
llvm-svn: 199384
This change adds ASAN_OPTIONS=start_deactivated=1 flag. When present, ASan will
start in "deactivated" mode, with no heap poisoning, no quarantine, no stack
trace gathering, and minimal redzones. All this features come back when
__asan_init is called for the constructor of an instrumented library.
The primary use case for this feature is Android. Code itself is not
Android-specific, and this patch includes a Linux test for it.
llvm-svn: 199377
struct stat defined in <asm/stat.h> is incorrect for x32. <asm/stat.h>
is included to get struct __old_kernel_stat. But struct __old_kernel_stat
isn't used for x86-64 nor x32. This patch includes <sys/stat.h> instead
of <asm/stat.h> and comments out size check of struct __old_kernel_stat
for x86-64.
Patch by H.J. Lu
llvm-svn: 199192
This unbreaks the configure build on darwin, where we build with the
stub SDK in -isysroot. Oddly, the cmake build doesn't have this
restriction, which probably means the cmake configs are out of date.
llvm-svn: 198676
Otherwise on (some) 64-bit systems, -Wformat will trigger a warning
because uint64_t is an 'unsigned long' not an 'unsigned long long'.
Consequently, PGOProfiling.c would fail to build if -Werror and
-Wformat are both enabled.
llvm-svn: 198644
This is fairly minimal support for instrumentation based PGO. The data
format is inefficient, and the output file name is hardcoded to
pgo-data.
llvm-svn: 198638
Currently fails with:
gotsan.cc:7686:40: error: the address of 'int __sanitizer_pthread_attr_getstack(void*, void**, size_t*)' will always evaluate as 'true' [-Werror=address]
llvm-svn: 198261
A helper function is a C++ function, and so even though one of the two
definitions is weak, it still technically triggers the ODR. Perhaps
these two definitions are ODR equivalent, but I'm not even confident in
that.
Instead, just define the function once, declare it as weak, and use
a wrapper that is clearly file-local. This avoids two definitions. Also
make the function extern "C" so that we can't even mess up the type
signature somehow or otherwise fail to match up the weak declaration
here with the interceptor defined elsewhere.
llvm-svn: 198253
Per Nick Kledzik (http://llvm.org/bugs/show_bug.cgi?id=17976):
"""
For MacOSX, the labels in __cstring section should be 'L' labels, and the labels
in the __const section should be non-L or 'l' labels. The later ('l') will cause
the linker to remove the label in the final executable.
"""
, so we need to ensure that the globals emitted by ASan have appropriate labels.
llvm-svn: 198022
This should fix http://llvm.org/bugs/show_bug.cgi?id=17976
Another test checking for the global variables' locations and prefixes on Darwin will be committed separately.
llvm-svn: 198017
This is intended to address the following problem.
Episodically we see CHECK-failures when recursive interceptors call back into user code. Effectively we are not "in_rtl" at this point, but it's very complicated and fragile to properly maintain in_rtl property. Instead get rid of it. It was used mostly for sanity CHECKs, which basically never uncover real problems.
Instead introduce ignore_interceptors flag, which is used in very few narrow places to disable recursive interceptors (e.g. during runtime initialization).
llvm-svn: 197979
Print the list of leaked objects after each leak report. Previously we
printed only a joint list of all leaked objects. As a bonus, suppressed objects
are no longer reported.
llvm-svn: 197977
Added the following custom wrappers:
strstr
strrchr
memchr
nanosleep
socketpair
Tweaked a couple of existing wrappers: if a wrapper returns a pointers derived
from an input pointer, then return the label of the input pointer
Sorted tests invocation
Differential Revision: D2354
llvm-svn: 197601
This code is not robust enough and triggers when simply linking with
libdynamorio.so, without any code translation at all. Disabling it is safe
(i.e. we may unpoison too much memory and see false negatives, but never false
positives).
llvm-svn: 197568
It conflicted with the verbosity flag we had in common flags. We don't need an
LSan-specific flag anyway.
Also, shift some logging levels and remove some unnecessary code.
llvm-svn: 197512
Introduce a flag to either always or never print matched suppressions.
Previously, matched suppressions were printed unconditionally if there were
unsuppressed leaks. Also, verbosity=1 no longer has the semantics of "always
print suppressions and summary".
llvm-svn: 197510
Add an interface for telling LSan that a region of memory is to be treated as a
source of live pointers. Useful for code which stores pointers in mapped memory.
llvm-svn: 197489
*h_errno is written not on success, but on failure.
In fact, it seems like it can be written even when return value signals
success, so we just unpoison it in all cases.
llvm-svn: 197383
We need to filter out architectures that the compiler hasn't been built to
target (most likely the ARM ones) before attemptint to build a version of
libcompiler_rt.
This can result in a completely empty library (e.g. soft-float doesn't have any
x86 variants), in which case we shouldn't even try the build
llvm-svn: 197028
Before we did it lazily on the first stack unwind in the thread.
It resulted in deadlock when the unwind was caused by memory allocation
inside pthread_getattr_np:
pthread_getattr_np <<< not reentable
GetThreadStackTopAndBottom
__interceptor_realloc
pthread_getattr_np
llvm-svn: 197026
it to the LLVM project through the appropriate channels.
This reverts:
r195837: "[Sanitizer] Add rudimentary support for using libbacktrace in ..."
llvm-svn: 196875
by their authors.
This may break builds where others added code relying on these patches,
but please *do not* revert this commit. Instead, we will prepare patches
which fix the failures.
Reverts the following commits:
r168306: "[asan] support x32 mode in the fast stack unwinder. Patch by H.J. Lu"
r168356: "[asan] more support for powerpc, patch by Peter Bergner"
r196489: "[sanitizer] fix the ppc32 build (patch by Jakub Jelinek)"
llvm-svn: 196802
Summary:
No more (potenital) false negatives due to red zones or fake stack
frames.
Reviewers: kcc, samsonov
Reviewed By: samsonov
CC: llvm-commits, samsonov
Differential Revision: http://llvm-reviews.chandlerc.com/D2359
llvm-svn: 196778
Summary:
Fix race on report_fd/report_fd_pid between the parent process and the
tracer task.
Reviewers: samsonov
Reviewed By: samsonov
CC: llvm-commits, kcc, dvyukov
Differential Revision: http://llvm-reviews.chandlerc.com/D2306
llvm-svn: 196385
More steps are needed to actually make it usable:
* sanitizer runtimes should be compiled with -DSANITIZER_LIBBACKTRACE.
* libbacktrace headers should be installed.
* user has to manually link in libbacktrace.a into the executable.
We can easily solve the first two problems in the build system, but
detecting/linking libbacktrace to all the tests we have and end-user programs
is more challenging (and will unlikely work w/o Driver support).
Based on the patch by Jakub Jelinek!
llvm-svn: 195837
1) Don't start external symbolizer subprocess until we actually try to
symbolize anything.
2) Allow to turn off external symbolizer by providing empty ?SAN_SYMBOLIZER_PATH
environment variable.
llvm-svn: 195771
No longer allow interceptors to be called during initialization, use the preinit
array (instead of initializing at the first call to an intercepted function) and
adopt the calloc() hack from ASan.
llvm-svn: 195642
Summary:
Implement TwoLevelByteMap and use it for the internal allocator on 64-bit.
This reduces bss on 64-bit by ~8Mb because we don't use FlatByteMap on 64-bits any more.
Dmitry, please check my understanding of atomics.
Reviewers: dvyukov
Reviewed By: dvyukov
CC: samsonov, llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D2259
llvm-svn: 195637
Performance improvement. Also, the allocator was using CompactSizeClassMap for
no good reason, so I switched it to DefaultSizeClassMap.
llvm-svn: 195570
Summary:
Call real pthread_attr_getstack instead of the interceptor
when we do intercept pthread_attr_getstack.
Reviewers: samsonov, eugenis
Reviewed By: samsonov
CC: llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D2237
llvm-svn: 195441
Some linux headers are broken on older kernels.
Instead of depending on the constants and types from such headers directly,
we provide our own definitions and then verify them with compile-time
assertions. This makes the dependency on the headers test-only and would allow
switching to some other way of testing on older kernels, or even disable the
tests as the last resort (after all, kernel interfaces are supposed to be
stable).
llvm-svn: 195427
- Introduce several new custom glibc wrappers
- Implement some of the not yet implemented wrappers
- Refactor and extend the tests
- Add script to make sure all declare custom glibc wrappers are implemented & tested.
Patch by Lorenzo Martignoni!
Differential Revision: http://llvm-reviews.chandlerc.com/D2234
llvm-svn: 195381
the tests produce episodic flakes
the hypothesis is that it's due to our "racy" race detection algorithm
sleeps should remove the flakes
llvm-svn: 195351
(-Wl,-syslibroot was accidentally overridden by -isysroot from Clang on OSX 10.9)
-isysroot is a Clang/LLVM-GCC-specific option, but hosting libsanitizer for LLVM
with GCC on Darwin shouldn't work anyway, because of the missing blocks support.
llvm-svn: 195132
- Add a bunch of glibc functions to the ABI list
- Group similar functions in the ABI
Patch by Lorenzo Martignoni!
Differential Revision: http://llvm-reviews.chandlerc.com/D2185
llvm-svn: 195110
Summary:
Definitions we use in public sanitizer headers may
slightly conflict with the ones we use in private sanitizer runtimes.
Moreover, we generally forbid to include any system headers (like <stdint.h>)
in sanitizer runtime headers. This leads to inevitable duplication of selected
interface function declarations, but we decided to live with it.
Reviewers: pcc
Reviewed By: pcc
CC: kcc, llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D2179
llvm-svn: 194955
This should build a separate set of embedded runtime libraries, supporting the
option product:
{PIC, static} x { Hard-float, soft-float }
The emphasis is on ARM platforms (Cortex-M4F, Cortex-M3, Cortex-M0) but X86
variants are also built where possible.
rdar://problem/10817205
llvm-svn: 194873
This header has not been supported at all for the last 2 major OS X releases.
Removed its include and the capture of related symbols.
<rdar://problem/15303348>
llvm-svn: 194841
I still don't know what is causing our bootstrapped LTO buildbots to fail,
but llvm r194701 seems to be OK and I can't imagine that these changes could
cause the problem.
llvm-svn: 194790
Apple's bootstrapped LTO builds have been failing, and these changes (along
with llvm 194701) are the only things on the blamelist. I will either reapply
these changes or help debug the problem, depending on whether this fixes the
buildbots.
llvm-svn: 194779
This reduces the number of "write" syscalls performed to print a single
stack frame description, and makes sanitizer output less intermixed with
program output. Also, add a number of unit tests.
llvm-svn: 194686
Invoke a fatal stack trace unwinder when ASan prints allocator-relevant
error reports (double-free, alloc-dealloc-mismatch, invalid-free).
Thus we'll be able to print complete stack trace even if allocation/free
stacks are not stored (malloc_context_size=0).
Based on the patch by Yuri Gribov!
llvm-svn: 194579
Fixes http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58994, which hadn't
manifested in LLVM because libclang_rt.asan_osx_dynamic.dylib used to depend on
the Foundation framework.
Without that dependency some interceptors may be called from the system
libraries before libSystem_initializer() is called, which lead to assertion
failures in sanitizer_mac.cc (_NSGetEnviron() returns NULL).
To fix the problem we fall back to the original functions in the common
libsanitizer interceptors and the __cxa_atexit() interceptor on Darwin.
This patch also prints a better error message in the case _NSGetEnviron()
returns NULL.
llvm-svn: 194573
ARM's UAL syntax allows the same assembly file to be compiled in both ARM and
Thumb mode. Conditional execution is handled by requiring the Thumb IT blocks,
but essentially ignoring them when compiling for ARM.
llvm-svn: 194429
CMake changes to build the ASan runtime for the iOS simulator. This is a universal library targeting the same architectures as the OSX ASan runtime does, thus the iossim version can't live in the same universal libclang_rt.asan_osx_dynamic.dylib
The difference between the OSX and iossim builds is in the -mios-simulator-version-min and -ios_simulator_version_min flags that tell Clang to compile and link iossim code.
The iossim runtime can only be built on a machine with both Xcode and the iOS Simulator SDK installed. If xcodebuild -version -sdk iphonesimulator Path returns a nonempty path, it is used when compiling and linking the iossim runtime.
llvm-svn: 194199
Alexander Potapenko