When fuzzing a program by feeding random inputs to it, we need a mechanism to tell when the program is doing unexpected things. Sanitizers help detect bugs in the program at runtime. They are usually used along with fuzzing to detect bugs in programs. The two roles of sanitizers: Detect incorrect program behaviour: like accessing memory that the program is not supposed to access Report incorrect behaviour: To be useful, the sanitizer needs to report useful information (like the stack trace and ) that makes it easier to understand and fix the bug.

In this post, we’ll see how fuzzing is used for finding different types of bugs in the Linux kernel. This post consists of my notes taken from the talk by Andrey Konavalov about Linux fuzzing. Operating systems kernels are complex. Testing kernels is of prime importance since any vulnerability in the kernel can lead to compromising the whole system. Fuzzing is a dynamic program analysis technique, used to find bugs in software.

Yesterday, I read an interesting research paper about reverse engineering TLBs using TLB desynchronization. In this post, I’ll write briefly about the key ideas and what I found very interesting in the paper. You can find the paper here: TLB;DR: Enhancing TLB-based Attacks with TLB Desynchronized Reverse Engineering. TLB;DR Source code Reverse engineering CPU internals In the subfield of hardware security that focuses on communicating (covert channels) or leaking (side channels) critical information using timing or storage channels, accurate information about the CPU internals helps create more efficient and reliable channels.

One of the hurdles in contributing to Linux kernel as a beginner is to find something to work on. If you are working as a kernel developer or are a part of a kernel mentorship program, this shouldn’t be a problem. But for others, the first step is to send a bunch of coding style fixes for warnings reported by checkpatch. In addition to checkpatch, there are many other static analysis tools like sparse, coccinelle and smatch, used to analyse source code and find possible bugs.

Compression is a common technique that allows us to reduce the size of data. We use it when we store or send files as zip, 7z or rar archives. It helps reduce the space occupied by files in the filesystem and also to save bandwidth while transferring over a network. In general, compression algorithms try to encode redundancy or regularity in data with lesser bits. There are many different software compression algorithms available.