Challenging testcases for fuzzers


Fuzzing introspection testcases

Test a fuzzer what fuzzing challenges it can solve. The goal is to have testcases that afl++ can all complete - once they are all implemented. Currently the different solving mechanisms have to be set as defined in src/afl-fuzz-redqueen.c. The comparisons to honggfuzz and libfuzzer are biased, as the testcases are how afl-fuzz sees things. The libfuzzer/ directories has libfuzzer's applicable own testcases, of which afl++ is only able to solve 64% (vs 86% honggfuzz and 92% libfuzzer).


The fuzzers and necessary compilers must be in the path: afl-clang-fast, afl-fuzz, hfuzz-clang, honggfuzz, clang.



TARGET can be (currently) one of:

  • afl++
  • afl++-qemu
  • honggfuzz
  • libfuzzer


  • test-u8 - several chained 8 bits checks
  • test-u16 - several chained 16 bits checks
  • test-u32 - several chained 32 bits checks
  • test-u64 - several chained 64 bits checks
  • test-u128 - several chained 128 bits checks
  • test-u32-cmp - several chained 32 bit lesser/greater checks
  • text-extint - llvm _ExtInt() tests
  • test-float - several chained float checks
  • test-double - several chained double checks
  • test-longdouble - several chained long double checks
  • test-memcmp - several chained memcmp checks
  • test-strcmp - several chained strncasecmp checks
  • test-transform - different transforming string checks
  • test-crc32 - several chained crc32 checks

The testcases are WIP - some might be further refined in the future and more added.


On success the time to solve is displayed.

On failure the generated corpus files are displayed (so you can see how many finds there were). There is also an output log called TESTCASE.log.

Test as of 21th January 2021

All from current repository state (afl++_vX is the -l option value X). Solve time: 120 seconds

testcase afl++_v3 afl++_v2 afl++_v1 afl++_QEMU_v3 honggfuzz-2.3 libfuzzer-12
test-crc32 0m5,596s 0m18,083s 0m38,924s 0m14,609s FAIL 0m55,987s
test-double 0m3,374s 0m2,414s 0m14,064s FAIL FAIL FAIL
test-extint 0m1,663s 0m1,051s 0m0,988s FAIL FAIL FAIL
test-float 0m1,650s 0m1,206s 0m15,023s FAIL FAIL FAIL
test-longdouble 0m1,241s 0m2,651s 0m15,159s FAIL FAIL FAIL
test-memcmp 0m0,622s 0m0,835s 0m13,031s 0m6,494s 0m1,023s 0m1,269s
test-strcmp 0m0,623s 0m0,624s 0m9,745s 0m5,727s 0m1,220s 0m1,469s
test-transform 0m4,334s FAIL FAIL FAIL FAIL FAIL
test-u128 0m0,418s 0m0,416s 0m1,272s FAIL FAIL FAIL
test-u16 0m1,233s 0m1,442s 0m27,811s 0m8,132s 0m1,425s 0m12,065s
test-u32 0m1,434s 0m1,641s 0m2,085s 0m5,185s 0m1,025s 0m6,984s
test-u32-cmp 1m29,184s 0m14,712s 0m53,016s 1m42,470s 0m1,824s 0m0,759s
test-u64 0m0,623s 0m0,625s 0m4,063s 0m3,844s 0m1,024s 0m10,346s
test-u8 0m1,449s 0m1,259s 0m43,906s 0m18,186s 0m1,026s 0m4,382s

afl++_v3 has the most solves, but due to the many solve attempts overall fuzzing performance is decreased, as can be seen at Interpretation: the slowest solver is the best in real-world fuzzing.

More testcases or fuzzers?

Just add them and send a PR.

You might also like...
  • test-longdouble: use sizeof(long double) instead of assuming packed 10b

    test-longdouble: use sizeof(long double) instead of assuming packed 10b

    Some platforms (wasm) use 128-bit values for long double, making the test impossible with the overlapping values in *(long double*) buf, *(long double*) buf+10, ... Note that this also marginally changes the test for x86 because sizeof(long double) is 16 even with 80-bit values as they are canonically padded from 10 to 16 bytes.

    opened by Mrmaxmeier 1
Advanced Fuzzing League ++
We want to make fuzzing better and better
Advanced Fuzzing League ++
A patched QEMU that exposes an interface for LibAFL-based fuzzers

QEMU LibAFL Bridge This is a patched QEMU that exposes an interface for LibAFL-based fuzzers. This raw interface is used in libafl_qemu that expose a

Advanced Fuzzing League ++ 29 Dec 14, 2022
A collection of user-space Linux kernel specific guided fuzzers based on LKL

kBdysch kBdysch is a collection of fast Linux kernel specific fuzzing harnesses supposed to be run in userspace in a guided fuzzing manner. It was des

Anatoly Trosinenko 63 Nov 25, 2022
Phan Sang 17 Dec 29, 2022
RRxIO - Robust Radar Visual/Thermal Inertial Odometry: Robust and accurate state estimation even in challenging visual conditions.

RRxIO - Robust Radar Visual/Thermal Inertial Odometry RRxIO offers robust and accurate state estimation even in challenging visual conditions. RRxIO c

Christopher Doer 63 Dec 20, 2022