MoonLight-SteinsGate
learning_engine.py seems to hang when fuzzing mp3gain
I'm using LearnAFL to fuzz mp3gain 1.6.2, after running 1.5 minutes, the fuzzer hangs.
Probably learning_engine.py stucks into an infinite loop?
# ps aux|grep python
root 961 99.8 0.0 25052 9608 ? R+ 05:09 6:07 python ./learning_engine.py /work/output/learnafl/knowledge/2019-9-12-3-9-40-25-knowledge.txt /work/output/learnafl/models/0 /work/output/learnafl/information_of_time
Whole afl output folder with mp3gain binary: learnafl_issue.zip
Hi! When learning_engine.py deals with some complex test cases, it may take some time to implement the longest common substring searching algorithm. I have set some time conditions to prevent the python script stuck (e.g., the 261, 266 lines in learning_engine.py). Generally, it takes about 5s~3min for LearnAFL to execute learning_engine.py one time. How long is learning_engine.py stuck?
Hi! When learning_engine.py deals with some complex test cases, it may take some time to implement the longest common substring searching algorithm. I have set some time conditions to prevent the python script stuck (e.g., the 261, 266 lines in learning_engine.py). Generally, it takes about 5s~3min for LearnAFL to execute learning_engine.py one time. How long is learning_engine.py stuck?
hi, @MoonLight-SteinsGate . It stuck half one hour when fuzzed base64.
Hi! When learning_engine.py deals with some complex test cases, it may take some time to implement the longest common substring searching algorithm. I have set some time conditions to prevent the python script stuck (e.g., the 261, 266 lines in learning_engine.py). Generally, it takes about 5s~3min for LearnAFL to execute learning_engine.py one time. How long is learning_engine.py stuck?
hi, @MoonLight-SteinsGate . It stuck half one hour when fuzzed base64.
It seems that there are some bugs in our learning_engine.py. Can you please provide the knowledge and model file which the script need to run?