Python Marshal

Python value serializationutilize is offered through the python marshal module. In other words, the module includes methods for binary-format writing and reading of Python objects. The format is unfortunately not defined, and Python maintainers may alter it in ways that are incompatible with previous Python versions. Other Python mutilise the marshal module internally, for instance, to read and write.py files that contain pseudo-compiled Python code. But you may also access this serialization technique using Python's open API.

The marshal module shouldn't be used with untrusted data, as demonstrated in this post, which also demonstrates how the module may be swiftly evaluated with a basic dumb fuzzer.

Because the marshal module is written in C, the easiest fuzzing objective is to simply search for common C programming errors like buffer overflows, use-after-free, null-pointer dereferences, etc. The excellent memory checker AddressSanitizer (ASan) might assist in locating such problems. Code is instrumented by AddressSanitizer during compilation. The utility provides a check for memory corruption problems and replaces the malloc and free routines. The program then tries to find memory corruptions during runtime and notify them right away with lots of helpful details. Python may be built with GCC 4.8+, which includes AddressSanitizer.

Building Python with AddressSanitizer

Python code (CPython) can be cloned with the following command:

hg clone https://hg.python.org/cpython

If you run ./configure --help, you can see that it has --with-address-sanitizer option which is supposed to enable AddressSanitizer. But for some reason it didn’t work for me, so I just used the following commands to build Python:

CFLAGS="-g -fsanitize=address -fno-omit-frame-pointer -O0" \
CPPFLAGS="-fsanitize=address -fno-omit-frame-pointer -O0" \
LDFLAGS="-fsanitize=address" \
    ./configure \
        --prefix=/home/artem/projects/fuzzing/python/build/ \
        --disable-ipv6
ASAN_OPTIONS="detect_leaks=0" make
ASAN_OPTIONS="detect_leaks=0" make install

Let me quickly explain what those options mean:

  1. CFLAGS, LDFLAGS, CPPFLAGS are standard enviroment variable which specify options for C/C++ compiler and linker.
  2. -fsanitize=address enables AddressSanitizer (it has to be passed to both compiler and linker)
  3. -g makes GCC produce debugging information.
  4. -O0 turns off compiler optimizations (but slows down execution).
  5. -fno-omit-frame-pointer is for nicer stack traces.
  6. ASAN_OPTIONS is an environment variable which contains parameters for AddressSanitizer at runtime.
  7. ASAN_OPTIONS="detect_leaks=0" turns off memory leaks checker which is part of AddressSanitizer.
  8. --prefix specifies a directory where it should put output binaries, libs, etc.
  9. --disable-ipv6 disables IPv6 (nothing surprising).

If the build runs smoothly, you can run python3.6 --version as a smoke test.



Comments

Post a Comment

Popular posts from this blog

Unleashing Creativity: The Latest Frontier of Animation AI

Image
 Introduction Animation AI is the result of the convergence of technology and creativity in the animation field. This groundbreaking fusion has not only transformed the animation landscape but also opened up exciting avenues for artists, storytellers, and tech enthusiasts alike. In this blog, we delve into the latest trends, advancements, and possibilities that Animation AI brings to the table. Lifelike Character Animation AI-driven character animation has taken realism to unprecedented levels. With machine learning algorithms studying human motion, characters move and interact in ways that mirror real-life actions. Expressions and emotions are portrayed authentically, immersing audiences in narratives that tug at heartstrings. Efficiency and Speed Automation is a hallmark of Animation AI. Inbetweening, traditionally a meticulous manual process, is now accelerated by AI, reducing production time. Creative minds can focus on refining the narrative while AI handles the technical intr
Read more

Tanh Activation Function

Image
With a difference in output range of -1 to 1, the tanh activation function is remarkably similar to the sigmoid/logistic activation function, and even has the same S-shape. The larger the input (more positive), the closer the output to 1.0, and the smaller the input (more negative), the closer the output to -1.0. Mathematically it can be represented as: The following are some of the benefits of using this activation function: We can simply map the output values as strongly negative, neutral, or very positive because the tanh activation function's output is Zero centred. Because its values range from -1 to, it's commonly utilised in hidden layers of neural networks. As a result, the hidden layer's mean is 0 or extremely close to it. It aids in data centering and makes learning the next layer much simpler. To understand the tanh activation function's limits, look at its gradient. As you can see, it, like the sigmoid activation function, has the problem of vanishing gradi
Read more

Unveiling the Hidden Gems: Exploring Data Mining Functionality

 Introduction:  In today's data-driven world, where information is abundant but insights are hidden, data mining functionality emerges as a powerful tool to unlock the hidden potential of vast datasets. By applying advanced algorithms and techniques, data mining allows us to extract valuable knowledge, patterns, and relationships from complex data. In this blog, we will delve into the functionality of data mining, uncovering its remarkable capabilities and highlighting its significance in various domains. Extracting Hidden Patterns: Data mining enables us to unearth hidden patterns and relationships within large datasets. By employing algorithms like association rule mining, we can discover interesting associations between different variables. These patterns can provide valuable insights for businesses, helping them understand customer behavior, market trends, and make informed decisions. Predictive Analytics: One of the key functionalities of data mining is predictive analytics
Read more