Skip to content

AddressSanitizer

chefmax edited this page May 15, 2019 · 29 revisions

Introduction

AddressSanitizer (aka ASan) is a memory error detector for C/C++. It finds:

This tool is very fast. The average slowdown of the instrumented program is ~2x (see AddressSanitizerPerformanceNumbers).

The tool consists of a compiler instrumentation module (currently, an LLVM pass) and a run-time library which replaces the malloc function.

The tool works on x86, ARM, MIPS (both 32- and 64-bit versions of all architectures), PowerPC64. The supported operation systems are Linux, Darwin (OS X and iOS Simulator), FreeBSD, Android:

OS x86 x86_64 ARM ARM64 MIPS MIPS64 PowerPC PowerPC64
Linux yes yes yes yes yes yes
OS X yes yes
iOS Simulator yes yes
FreeBSD yes yes
Android yes yes yes yes

Other OS/arch combinations may work as well, but aren't actively developed/tested.

See also:

Getting AddressSanitizer

AddressSanitizer is a part of LLVM starting with version 3.1 and a part of GCC starting with version 4.8 If you prefer to build from source, see AddressSanitizerHowToBuild.

So far, AddressSanitizer has been tested only on Linux Ubuntu 12.04, 64-bit (it can run both 64- and 32-bit programs), Mac 10.6, 10.7 and 10.8, and AddressSanitizerOnAndroid 4.2+.

Using AddressSanitizer

In order to use AddressSanitizer you will need to compile and link your program using clang with the -fsanitize=address switch. To get a reasonable performance add -O1 or higher. To get nicer stack traces in error messages add -fno-omit-frame-pointer. Note: Clang 3.1 release uses another flag syntax.

% cat tests/use-after-free.c
#include <stdlib.h>
int main() {
  char *x = (char*)malloc(10 * sizeof(char*));
  free(x);
  return x[5];
}
% ../clang_build_Linux/Release+Asserts/bin/clang -fsanitize=address -O1 -fno-omit-frame-pointer -g   tests/use-after-free.c

Now, run the executable. AddressSanitizerCallStack page describes how to obtain symbolized stack traces.

% ./a.out
==9901==ERROR: AddressSanitizer: heap-use-after-free on address 0x60700000dfb5 at pc 0x45917b bp 0x7fff4490c700 sp 0x7fff4490c6f8
READ of size 1 at 0x60700000dfb5 thread T0
    #0 0x45917a in main use-after-free.c:5
    #1 0x7fce9f25e76c in __libc_start_main /build/buildd/eglibc-2.15/csu/libc-start.c:226
    #2 0x459074 in _start (a.out+0x459074)
0x60700000dfb5 is located 5 bytes inside of 80-byte region [0x60700000dfb0,0x60700000e000)
freed by thread T0 here:
    #0 0x4441ee in __interceptor_free projects/compiler-rt/lib/asan/asan_malloc_linux.cc:64
    #1 0x45914a in main use-after-free.c:4
    #2 0x7fce9f25e76c in __libc_start_main /build/buildd/eglibc-2.15/csu/libc-start.c:226
previously allocated by thread T0 here:
    #0 0x44436e in __interceptor_malloc projects/compiler-rt/lib/asan/asan_malloc_linux.cc:74
    #1 0x45913f in main use-after-free.c:3
    #2 0x7fce9f25e76c in __libc_start_main /build/buildd/eglibc-2.15/csu/libc-start.c:226
SUMMARY: AddressSanitizer: heap-use-after-free use-after-free.c:5 main

Interaction with other tools

gdb

See AddressSanitizerAndDebugger

ulimit -v

The ulimit -v command makes little sense with ASan-ified binaries because ASan consumes 20 terabytes of virtual memory (plus a bit).

You may try more sophisticated tools to limit your memory consumption, e.g. https://en.wikipedia.org/wiki/Cgroups

Flags

See the separate AddressSanitizerFlags page.

Call stack

See the separate AddressSanitizerCallStack page.

Incompatibility

Sometimes an AddressSanitizer build may behave differently than the regular one. See AddressSanitizerIncompatiblity for details.

Turning off instrumentation

In some cases a particular function should be ignored (not instrumented) by AddressSanitizer:

  • Ignore a very hot function known to be correct to speedup the app.
  • Ignore a function that does some low-level magic (e.g. walking through the thread's stack bypassing the frame boundaries).
  • Don't report a known problem. In either case, be very careful.

To ignore certain functions, one can use the no_sanitize_address attribute supported by Clang (3.3+) and GCC (4.8+). You can define the following macro:

#if defined(__clang__) || defined (__GNUC__)
# define ATTRIBUTE_NO_SANITIZE_ADDRESS __attribute__((no_sanitize_address))
#else
# define ATTRIBUTE_NO_SANITIZE_ADDRESS
#endif
...
ATTRIBUTE_NO_SANITIZE_ADDRESS
void ThisFunctionWillNotBeInstrumented() {...}

Clang 3.1 and 3.2 supported __attribute__((no_address_safety_analysis)) instead.

You may also ignore certain functions using a blacklist: create a file my_ignores.txt and pass it to AddressSanitizer at compile time using -fsanitize-blacklist=my_ignores.txt (This flag is new and is only supported by Clang now):

# Ignore exactly this function (the names are mangled)
fun:MyFooBar
# Ignore MyFooBar(void) if it is in C++:
fun:_Z8MyFooBarv
# Ignore all function containing MyFooBar
fun:*MyFooBar*

FAQ

  • Q: Can AddressSanitizer continue running after reporting first error?

  • A: Yes it can, AddressSanitizer has recently got continue-after-error mode. This is somewhat experimental so may not yet be as reliable as default setting (and not as timely supported). Also keep in mind that errors after the first one may actually be spurious. To enable continue-after-error, compile with -fsanitize-recover=address and then run your code with ASAN_OPTIONS=halt_on_error=0.

  • Q: Why didn't ASan report an obviously invalid memory access in my code?

  • A1: If your errors is too obvious, compiler might have already optimized it out by the time Asan runs.

  • A2: Another, C-only option is accesses to global common symbols which are not protected by Asan (you can use -fno-common to disable generation of common symbols and hopefully detect more bugs).

  • A3: If _FORTIFY_SOURCE is enabled, ASan may have false positives, see next question.

  • Q: I've compiled my code with -D_FORTIFY_SOURCE flag and ASan, or -D_FORTIFY_SOURCE is enabled by default in my distribution (most modern distributions). Now ASan misbehaves (either produces false warnings, or does not find some bugs).

  • A: Currently ASan (and other sanitizers) doesn't support source fortification, see https://github.com/google/sanitizers/issues/247. The fix should most likely be on the glibc side, see the (stalled) discussion here.

  • Q: When I link my shared library with -fsanitize=address, it fails due to some undefined ASan symbols (e.g. asan_init_v4)?

  • A: Most probably you link with -Wl,-z,defs or -Wl,--no-undefined. These flags don't work with ASan unless you also use -shared-libasan (which is the default mode for GCC, but not for Clang).

  • Q: My malloc stacktraces are too short or do not make sense?

  • A: Try to compile your code with -fno-omit-frame-pointer or set ASAN_OPTIONS=fast_unwind_on_malloc=0 (the latter would be a performance killer though unless you also specify malloc_context_size=2 or lower). Note that frame-pointer-based unwinding does not work on Thumb.

  • Q: My new() and delete() stacktraces are too short or do not make sense?

  • A: This may happen when the C++ standard library is linked statically. Prebuilt libstdc++/libc++ often do not use frame pointers, and it breaks fast (frame-pointer-based) unwinding. Either switch to the shared library with the -shared-libstdc++ flag, or use ASAN_OPTIONS=fast_unwind_on_malloc=0. The latter could be very slow.

  • Q: I'm using dynamic ASan runtime and my program crashes at start with "Shadow memory range interleaves with an existing memory mapping. ASan cannot proceed correctly.".

  • A1: If you are using shared ASan DSO, try LD_PRELOAD'ing Asan runtime into your program.

  • A2: Otherwise you are probably hitting a known limitation of dynamic runtime. Libasan is initialized at the end of program startup so if some preceeding library initializer did lots of memory allocations memory region required for ASan shadow memory could be occupied by unrelated mappings.

  • Q: The PC printed in ASan stack traces is consistently off by 1?

  • A: This is not a bug but rather a design choice. It is hard to compute exact size of preceding instruction on CISC platforms. So ASan just decrements 1 which is enough for tools like addr2line or readelf to symbolize addresses.

  • Q: I've ran with ASAN_OPTIONS=verbosity=1 and ASan tells something like

 ==30654== Parsed ASAN_OPTIONS: verbosity=1
 ==30654== AddressSanitizer: failed to intercept 'memcpy'
  • A: This warning is false (see https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58680 for details).

  • Q: I've built my main executable with ASan. Do I also need to build shared libraries?

  • A: ASan will work even if you rebuild just part of your program. But you'll have to rebuild all components to detect all errors.

  • Q: I've built my shared library with ASan. Can I run it with unsanitized executable?

  • A: Yes! You'll need to build your library with dynamic version of ASan and then run executable with LD_PRELOAD=path/to/asan/runtime/lib.

  • Q: On Linux I am seeings a crash at startup with something like this

ERROR: AddressSanitizer failed to allocate 0x400000000 (17179869184) bytes at address 67fff8000 (errno: 12)
  • A: Make sure you don't have 2 in /proc/sys/vm/overcommit_memory

  • Q: I'm working on a project that uses bare-metal OS with no pthread (TLS) support and no POSIX syscalls and want to use ASan, but its code depends on some stuff (e.g. dlsym) that is unavailable on my platform. Does ASan support bare-metal targets?

  • A: Out of the box we don't have support for your use case. The easiest for you would be to rip off everything you don't have and rebuild the ASan run-time. However, there have been many attempts in applying ASan to bare-metal and at least some were successful. E.g. http://events.linuxfoundation.org/sites/events/files/slides/Alexander_Popov-KASan_in_a_Bare-Metal_Hypervisor_0.pdf and also grep for "bare-metal" and similar stuff in https://groups.google.com/forum/#!forum/address-sanitizer group.

  • Q: Can I run AddressSanitizer with more aggressive diagnostics enabled?

  • A: Yes! In particular you may want to enable

 CFLAGS += -fsanitize-address-use-after-scope
 ASAN_OPTIONS=strict_string_checks=1:detect_stack_use_after_return=1:check_initialization_order=1:strict_init_order=1

check [Flags wiki] (https://github.com/google/sanitizers/wiki/AddressSanitizerFlags ) for more details on this.

  • Q: My library crashes with SIGABRT while calling free. What's going wrong?

  • A: Most probably you are dlopening your library with RTLD_DEEPBIND flag. ASan doesn't support RTLD_DEEPBIND, see issue #611 for details.

  • Q: I'm getting following error when building my code: gcc: error: cannot specify -static with -fsanitize=address.

  • A: ASan doesn't work with static linkage. You need to disable static linkage in order to use ASan on your code.

  • Q: What do 0xbebebebebebebebe and 0xbebebebe mean?

  • A: ASan, by default, writes 0xbe to newly allocated memory (see malloc_fill_byte). 0xbebebebebebebebe is (possibly) a 64-bit value that was allocated but not initialized. Similarly for 0xbebebebe.

  • Q: Can I mix code instrumented by Clang with code instrumented by GCC? Or can I at least compile code with Clang and link against GCC ASan runtime and vice versa?

  • A: No, you cannot, Clang and GCC have completely incompatible ASan implementations, you cannot mix them in any way.

Talks and papers

Comments?

Send comments to address-sanitizer@googlegroups.com or in Google+.

Clone this wiki locally