push to stable (#1734)

* afl++ -> AFL++

* update readme

* more debug

* slightly different weighting algo (#1719)

* better seed selection

* slightly different weighting calculation

* remove unnecessary memset

* Add "Hangs saved" to afl-whatsup (#1717)

The hangs could show long or infinite loops. This is important.

Co-authored-by: van Hauser <vh@thc.org>

* nits

* afl-showmap: Start a only a single fork server (#1718)

A forkserver is started by afl_fsrv_get_mapsize() when dynamically
finding the map size.  When an input directory option is specified a
second fork server was also started.  This commit re-arranges the inits
for several forkserver struct members so that we can re-use the server
started by the get_mapsize() call when not in coresight/qemu/unicorn
modes and just start the server otherwise.

* Source Code Coverage support for Nyx (Part 1) (#1720)

* Additional source code reformatting in afl-compiler-rt

* Add source code coverage support to afl-compiler-rt (for use with Nyx)

* doc, code format

* llvm 17 changes

* more llvm 17

* add frida mode tutorial

* fix effector map

* docs

* Should memset EFF_ALEN(len) of eff_map (#1722)

* fix reallocs

* fix afl-system-config for macos

* afl-fuzz.c: Document -i - in --help (#1725)

afl-fuzz.c: Document `-i -` in `--help`, to write that `-i` can be passed '-' to resume the prior fuzzing job. Also reference AFL_AUTORESUME so users know they can set that parameter to sidestep the issue entirely.

* tritondse custom mutator attempt

* tritondse fixes

* update libnyx (#1727)

* GNUmakefile: Update LLVM instructions (#1728)

Update LLVM instructions, because versions higher than 14 are supported and to be explicit that LLD is also required

* disable macos in the ci, works fine for me

* fix makefile

* better tritondse support

* next steps for tritondse

* qemuafl: Persistent mode for PPC32 targets

* update qemu_mode

* afl-clang-lto incomptable with -flto=thin

* add @responsefile support for afl-cc

---------

Co-authored-by: fxlb <devel.fx.lebail@orange.fr>
Co-authored-by: Nick Potenski <nick.potenski@garmin.com>
Co-authored-by: Christian Holler (:decoder) <choller@mozilla.com>
Co-authored-by: lazymio <mio@lazym.io>
Co-authored-by: Moshe Kaplan <me@moshekaplan.com>
Co-authored-by: Sergej Schumilo <sergej@schumilo.de>
Co-authored-by: Dominik Maier <domenukk@gmail.com>

1 files changed, 8 insertions, 0 deletions

diff --git a/docs/FAQ.md b/docs/FAQ.md
index 76350c79..8178db46 100644
--- a/docs/FAQ.md
+++ b/docs/FAQ.md
@@ -171,6 +171,14 @@ If you find an interesting or important question missing, submit it via
   The more "unstable" edges there are, the harder it is for AFL++ to identify
   valid new paths.
 
+  If you fuzz in persistent mode (`AFL_LOOP` or `LLVMFuzzerTestOneInput()`
+  harnesses, a large number of unstable edges can mean that the target keeps
+  internal state and therefore it is possible that crashes cannot be replayed.
+  In such a case do either **not** fuzz in persistent mode (remove `AFL_LOOP()`
+  from your harness or call `LLVMFuzzerTestOneInput()` harnesses with `@@`),
+  or set a low  `AFL_LOOP` value, e.g. 100, and enable `AFL_PERSISTENT_RECORD`
+  in `config.h` with the same value.
+
   A value above 90% is usually fine and a value above 80% is also still ok, and
   even a value above 20% can still result in successful finds of bugs. However,
   it is recommended that for values below 90% or 80% you should take