van Hauser

so is a fix to the grammar mutator needed?

In such a case you just compile it normally and use AFL++ qemu_mode or write a libafl qemu harness. but maybe it is not a user mode program? then libafl...

did you actually load the qasan module? read the readme

works fine for me: ``` $ clang -o qasan qasan.c # your example $ AFL_USE_QASAN=1 afl-fuzz -Q -i in -o out -- ./qasan @@ afl-fuzz++4.07a based on afl by Michal...

no that is not needed

@jonathanmetzman afl's makefile is not 100% portable either as it needs gnumake for several features. in afl++ I doubt we will switch to cmake. I dont see the benefit. And...

@jonathanmetzman yes, gnumake is nearly everywhere installed where a gcc is, cmake is not. this might change in the future though

There seems to be a bug in the code I tried https://github.com/hghwng/AFL/tree/skim against a target that has both 100% stability with google/afl and afl++. This however degrades to 85%- so...

Also I did not notice an impact on the execs_done when run for the same time. however that could be because of extra code paths because of falsely detected variable...

I can confirm this fixes the stability issue. my testing was more broad to see if it has a real impact on fuzzing. I believe you that it is always...