600b7e5945
- Adds new LLVM-based instrumentation support via afl-clang-fast.
- Experimental support for cgroup management via afl-cgroup, to
mitigate OOM issues when using afl with address sanitizer.
Signed-off-by: Austin Seipp <aseipp@pobox.com>
20 lines
777 B
Markdown
20 lines
777 B
Markdown
Updating the QEMU patches
|
|
=========================
|
|
|
|
When updating to the latest American Fuzzy Lop, make sure to check for
|
|
any new patches to qemu for binary fuzzing support:
|
|
|
|
https://github.com/mirrorer/afl/tree/master/qemu_mode
|
|
|
|
Be sure to check the build script and make sure it's also using the
|
|
right QEMU version and options in `qemu.nix`:
|
|
|
|
https://github.com/mirrorer/afl/blob/master/qemu_mode/build_qemu_support.sh
|
|
|
|
`afl-config.h`, `afl-types.h`, and `afl-qemu-cpu-inl.h` are part of
|
|
the afl source code, and copied from `config.h`, `types.h` and
|
|
`afl-qemu-cpu-inl.h` appropriately. These files and the QEMU patches
|
|
need to be slightly adjusted to fix their `#include`s (the patches
|
|
try to otherwise include files like `../../config.h` which causes the
|
|
build to fail).
|