* Output afl-clang-fast stuffs only if necessary (#1912) * afl-cc header * afl-cc common declarations - Add afl-cc-state.c - Strip includes, find_object, debug/be_quiet/have_*/callname setting from afl-cc.c - Use debugf_args in main - Modify execvp stuffs to fit new aflcc struct * afl-cc show usage * afl-cc mode selecting 1. compiler_mode by callname in argv[0] 2. compiler_mode by env "AFL_CC_COMPILER" 3. compiler_mode/instrument_mode by command line options "--afl-..." 4. instrument_mode/compiler_mode by various env vars including "AFL_LLVM_INSTRUMENT" 5. final checking steps 6. print "... - mode: %s-%s\n" 7. determine real argv[0] according to compiler_mode * afl-cc macro defs * afl-cc linking behaviors * afl-cc fsanitize behaviors * afl-cc misc * afl-cc body update * afl-cc all-in-one formated with custom-format.py * nits --------- Co-authored-by: vanhauser-thc <vh@thc.org> * changelog * update grammar mutator * lto llvm 12+ * docs(custom_mutators): fix missing ':' (#1953) * Fix broken LTO mode and response file support (#1948) * Strip `-Wl,-no-undefined` during compilation (#1952) Make the compiler wrapper stripping `-Wl,-no-undefined` in addition to `-Wl,--no-undefined`. Both versions of the flag are accepted by clang and, therefore, used by building systems in the wild (e.g., samba will not build without this fix). * Remove dead code in write_to_testcase (#1955) The custom_mutators_count check in if case is duplicate with if condition. The else case is custom_mutators_count == 0, neither custom_mutator_list iteration nor sent check needed. Signed-off-by: Xeonacid <h.dwwwwww@gmail.com> * update qemuafl * WIP: Add ability to generate drcov trace using QEMU backend (#1956) * Document new drcov QEMU plugin * Add link to lightkeeper for QEMU drcov file loading --------- Co-authored-by: Jean-Romain Garnier <jean-romain.garnier@airbus.com> * code format * changelog * sleep on uid != 0 afl-system-config * fix segv about skip_next, warn on unsupported cases of linking options (#1958) * todos * ensure afl-cc only allows available compiler modes * update grammar mutator * disable aslr on apple * fix for arm64 * help selective instrumentation * typos * macos * add compiler test script * apple fixes * bump nyx submodules (#1963) * fix docs * update changelog * update grammar mutator * improve compiler test script * gcc asan workaround (#1966) * fix github merge fuckup * fix * Fix afl-cc (#1968) - Check if too many cmdline params here, each time before insert a new param. - Check if it is "-fsanitize=..." before we do sth. - Remove improper param_st transfer. * Avoid adding llvmnative instrumentation when linking rust sanitizer runtime (#1969) * Dynamic instrumentation filtering for LLVM native (#1971) * Add two dynamic instrumentation filter methods to runtime * Always use pc-table with native pcguard * Add make_symbol_list.py and README * changelog * todos * new forkserver check * fix * nyx test for CI * improve nyx docs * Fixes to afl-cc and documentation (#1974) * Always compile with -ldl when building for CODE_COVERAGE When building with CODE_COVERAGE, the afl runtime contains code that calls `dladdr` which requires -ldl. Under most circumstances, clang already adds this (e.g. when building with pc-table), but there are some circumstances where it isn't added automatically. * Add visibility declaration to __afl_connected When building with hidden visibility, the use of __AFL_LOOP inside such code can cause linker errors due to __afl_connected being declared "hidden". * Update docs to clarify that CODE_COVERAGE=1 is required for dynamic_covfilter * nits * nyx build script updates * test error output * debug ci * debug ci * Improve afl-cc (#1975) * update response file support - full support of rsp file - fix some segv issues * Improve afl-cc - remove dead code about allow/denylist options of sancov - missing `if (!aflcc->have_msan)` - add docs for each function - typo * enable nyx * debug ci * debug ci * debug ci * debug ci * debug ci * debug ci * debug ci * debug ci * fix ci * clean test script * NO_NYX * NO_NYX * fix ci * debug ci * fix ci * finalize ci fix * Enhancement on Deterministic stage (#1972) * fuzzer: init commit based on aflpp 60dc37a8cf09f8e9048e4b6a2204d6c90b27655a * fuzzers: adding the skip variables and initialize * log: profile the det/havoc finding * log: add profile log output * fuzzers: sperate log/skipdet module * fuzzers: add quick eff_map calc * fuzzers: add skip_eff_map in fuzz_one * fuzzers: mark whole input space in eff_map * fuzzers: add undet bit threshold to skip some seeds * fuzzers: fix one byte overflow * fuzzers: fix overflow * fix code format * add havoc only again * code format * remove log to INTROSPECTION, rename skipdet module * rename skipdet module * remove log to stats * clean redundant code * code format * remove redundant code format check * remove redundant doc * remove redundant objects * clean files * change -d to default skipdet * disable deterministic when using CUSTOM_MUTATOR * revert fix * final touches for skipdet * remove unused var * remove redundant eff struct (#1977) * update QEMU-Nyx submodule (#1978) * update QEMU-Nyx submodule (#1980) * Fix type in AFL_NOOPT env variable in afl-cc help message (#1982) * nits * 2024 v4.10c release * fixes --------- Signed-off-by: Xeonacid <h.dwwwwww@gmail.com> Co-authored-by: Sonic <50692172+SonicStark@users.noreply.github.com> Co-authored-by: Xeonacid <h.dwwwwww@gmail.com> Co-authored-by: Nils Bars <nils.bars@rub.de> Co-authored-by: Jean-Romain Garnier <7504819+JRomainG@users.noreply.github.com> Co-authored-by: Jean-Romain Garnier <jean-romain.garnier@airbus.com> Co-authored-by: Sergej Schumilo <sergej@schumilo.de> Co-authored-by: Christian Holler (:decoder) <choller@mozilla.com> Co-authored-by: Han Zheng <35988108+kdsjZh@users.noreply.github.com> Co-authored-by: Khaled Yakdan <yakdan@code-intelligence.com>
libdislocator, an abusive allocator
For the general instruction manual, see docs/README.md.
This is a companion library that can be used as a drop-in replacement for the libc allocator in the fuzzed binaries. It improves the odds of bumping into heap-related security bugs in several ways:
-
It allocates all buffers so that they are immediately adjacent to a subsequent PROT_NONE page, causing most off-by-one reads and writes to immediately segfault,
-
It adds a canary immediately below the allocated buffer, to catch writes to negative offsets (won't catch reads, though),
-
It sets the memory returned by malloc() to garbage values, improving the odds of crashing when the target accesses uninitialized data,
-
It sets freed memory to PROT_NONE and does not actually reuse it, causing most use-after-free bugs to segfault right away,
-
It forces all realloc() calls to return a new address - and sets PROT_NONE on the original block. This catches use-after-realloc bugs,
-
It checks for calloc() overflows and can cause soft or hard failures of alloc requests past a configurable memory limit (AFL_LD_LIMIT_MB, AFL_LD_HARD_FAIL).
-
Optionally, in platforms supporting it, huge pages can be used by passing
USEHUGEPAGE=1
to make. -
Optionally, in platforms supporting it,
named
pages can be used by passingUSENAMEDPAGE=1
to make. -
Size alignment to
max_align_t
can be enforced withAFL_ALIGNED_ALLOC=1
. In this case, a tail canary is inserted in the padding bytes at the end of the allocated zone. This reduces the ability of libdislocator to detect off-by-one bugs but also it makes libdislocator compliant to the C standard.
Basically, it is inspired by some of the non-default options available for the OpenBSD allocator - see malloc.conf(5) on that platform for reference. It is also somewhat similar to several other debugging libraries, such as gmalloc and DUMA - but is simple, plug-and-play, and designed specifically for fuzzing jobs.
Note that it does nothing for stack-based memory handling errors. The -fstack-protector-all setting for GCC / clang, enabled when using AFL_HARDEN, can catch some subset of that.
The allocator is slow and memory-intensive (even the tiniest allocation uses up 4 kB of physical memory and 8 kB of virtual mem), making it completely unsuitable for "production" uses; but it can be faster and more hassle-free than ASAN / MSAN when fuzzing small, self-contained binaries.
To use this library, run AFL++ like so:
AFL_PRELOAD=/path/to/libdislocator.so ./afl-fuzz [...other params...]
You have to specify path, even if it's just ./libdislocator.so or $PWD/libdislocator.so.
Similarly to afl-tmin, the library is not "proprietary" and can be used with other fuzzers or testing tools without the need for any code tweaks. It does not require AFL-instrumented binaries to work.
Note that the AFL_PRELOAD approach (which AFL++ internally maps to LD_PRELOAD or DYLD_INSERT_LIBRARIES, depending on the OS) works only if the target binary is dynamically linked. Otherwise, attempting to use the library will have no effect.