blob: 16cbeb2f695a9bf9c12c9c73f4902f012222ecdf (
plain) (
blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
|
Changelog
=========
https://github.com/vanhauser-thc/afl-dyninst
- added -x performance optimization options, before this afl-dyninst was meh,
now it is OK. It supports 3 levels: -x (+45%, -xx additional +45%,
-xxx additional ~3% but so far only on intel x64)
top speed is now ~110% on stock afl-dyninst and ~65% on stock afl-gcc
- -e option now also understands function names, not only 0x1234 addresses
- searches for multiple entrypoints now: main, init, start and _NAME variants
- afl-dyninst now works fine with AARCH64 and PPC :)
- more verbose output, -vv -vvv support
- fixed some typos
- renamed afl-fuzz.sh to afl-fuzz-dyninst.sh and make install
installs now the scripts
- added -E switch to force a program exit at specific addresses
- added -D switch which does not do basic block instrumentation
- the -S switch can now be specified several times
- checks the dyninst version and comments on the -f option
- added -S switch to skip instrumenting a specific function
- added make install target
- updated README
- Fix for programs that were unable to print to stdout after instrumentation
- added -f switch to fix a bug in dyninst where sometimes the edi/rdi
register is not saved which is used in the instrumentation function
when a function is using edi/rdi for parameter passing
- added -m switch to only instrument basic blocks of a minimum size
- code beautifications, more detailed output, spelling fixes
- added afl-dyninst.sh and afl-fuzz.sh helper scripts which presets
the necessary environment variables
- based on https://github.com/talos-vulndev/afl-dyninst
|