6 Repos
Tools for extracting detailed state and metadata from active kernel tasks.
Distinct from Kernel Debuggers: Focuses on inspecting the state of kernel tasks, distinct from general kernel debugging or construction.
Explore 6 awesome GitHub repositories matching operating systems & systems programming · Kernel Task Inspection. Refine with filters or upvote what's useful.
pwndbg is a GDB plugin and binary analysis framework designed for reverse engineering, exploit development, and low-level program analysis. It extends the core functionality of the debugger to provide advanced memory inspection and automation tools. The project distinguishes itself with specialized capabilities for heap analysis across glibc, jemalloc, and musl, as well as a comprehensive kernel debugging toolkit for inspecting Linux kernel tasks and slab allocators. It includes an integrated ROP gadget searcher for constructing exploit chains and an LLM-powered debugging assistant that provi
Enables inspection of kernel-level execution via system emulation to identify security vulnerabilities.
Falco is an eBPF runtime security monitor and cloud native detection engine that identifies abnormal behavior and security threats across hosts and containers. It functions as a Linux kernel event auditor, capturing system calls and kernel events in real-time to detect malicious activity. The system distinguishes itself through a rule-based threat detection model that evaluates system activity against a library of community-maintained rules and custom security definitions. It enriches raw kernel events with container and Kubernetes metadata to provide observability into isolated environments
Dumping of raw events from kernel drivers to a local interface for debugging and verification of data collection.
This project provides Rust bindings for the TensorFlow C API, serving as a tensor computation interface and machine learning library. It enables the construction and execution of machine learning models and neural networks by bridging a systems language to high-performance backends. The framework supports GPU-accelerated computing to increase the speed of model training and inference by offloading mathematical operations to graphics processing units. It offers both graph-based computation for defining static network architectures and an eager execution mode for immediate operation calls durin
Allows retrieval of metadata and definition buffers for all operations and kernels available in the environment.
Battery Historian ist ein Visualisierungs- und Profiling-Tool zur Analyse des Stromverbrauchs und der Batterieentladung auf Android-Geräten. Es fungiert als Bugreport-Viewer und Stromverbrauchsprofiler, der System-Logs parst, um Batteriestatistiken und Kernel-Daten in eine webbasierte Oberfläche zu extrahieren. Das Tool ist darauf spezialisiert, disparate Datenquellen auf einer synchronisierten chronologischen Zeitachse zu korrelieren. Es identifiziert stromverbrauchende Aktivitäten durch das Tracking von Userspace- und Kernel-Wakelock-Übergängen, das Mapping von Kernel-Wakeup-Quellen auf Echtzeit-Zeitstempel und das Überlagern von externen Hardware-Strommonitor-Logs auf Systemereignisse. Das System bietet Funktionen für vergleichende Analysen, die die Berechnung von Deltas zwischen mehreren Bugreports ermöglichen, um Änderungen im Stromverbrauchsverhalten zu messen. Zudem aggregiert es Metriken auf Anwendungsebene und Systemereignisse, um spezifische Software-Trigger zu identifizieren, die verhindern, dass ein Gerät in den Schlafmodus wechselt.
Logs kernel wakeup sources and activities to identify low-level system triggers for power consumption.
HyperDbg is a hardware-assisted kernel-mode debugging platform that leverages virtualization to monitor and control system execution. By utilizing hypervisor-level primitives, it enables deep system analysis and instrumentation without relying on standard operating system debugging interfaces. The framework provides a comprehensive environment for inspecting both kernel and user-mode processes, allowing for granular control over execution flow and system state. The project distinguishes itself through a transparent debugging layer designed to remain invisible to the target environment. It emp
Analyzes kernel-mode processes using a hypervisor to monitor memory and instruction flow without standard OS interfaces.
pwru is a tooling implementation for tracing, filtering, and debugging network packet movements and transformations within the Linux kernel. It functions as an eBPF network packet tracer and debugger used to analyze kernel state and identify where network packets are dropped or redirected. The project provides specialized capabilities for monitoring the packet lifecycle, including tracking packets through NAT transformations and tunnel decapsulation. It includes an eBPF traffic filter that reduces noise by restricting traced packets based on namespaces, interfaces, and kernel function names.
Extracts detailed packet metadata and L4 tuples from active kernel tasks and data structures.