IMF is a kernel API fuzzer that leverages an automated API model inferrence techinque proposed in our paper at CCS. IMF currently only supports macOS. To see how to configure and run IMF, see the followings.
- python2.7
- pypy
- clang
- Generate hooking library for APIs
$ ./gen-hook [output(hooking code) path]
$ clang -Wall -dynamiclib -framework IOKit -framework CoreFoundation -arch i386\
-arch x86_64 hook.c -o hook
- Collect logs
$ DYLD_INSERT_LIBRARIES=[hooking library path] [program path] [program args]
- Filter logs
$ ./filter-log [log dir] [output dir] [# of output(filtered log)] [# of core]
- Infer a model and generate a fuzzer.
$ ./gen-fuzz [filtered logs path] [output(fuzzer code) path] [# of core]
- Compile the fuzzer
$ clang -framework IOKit -framework CoreFoundation -arch i386 fuzz.c -o fuzz
- Run the fuzzer
$ ./fuzz -f [log path] -s [seed] -b [bitlen] -r [rate] -l [# of max loops]
- You may want to run the generated fuzzer within a while loop.
- CVE-2017-7159
This research project has been conducted by SoftSec Lab at KAIST.
To cite our paper (pdf):
@INPROCEEDINGS{han:ccs2017,
author = {HyungSeok Han and Sang Kil Cha},
title = {Inferred Model-based Fuzzing},
booktitle = {Proceedings of the ACM Conference on Computer and Communications Security},
year = {2017},
pages = {2345--2358}
}
The work was supported by Institute for Information & communications Technology Promotion (IITP) grant funded by the Korea government (MSIT).
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent