Skip to main content
Online Training

macOS & iOS Kernel Internals for Security Researchers

Low-level macOS/iOS kernel internals for vulnerability analysis, security research, and modern debugging workflows.

Live online via Zoom
Next sessions
Upcoming dates by edition
EU/NA edition
2026-05-18 – 2026-05-22 - 5 days
APAC edition
2026-07-06 – 2026-07-10 - 5 days
Instructor
Stefan Esser
Stefan Esser
Antid0te

What you’ll be able to do

  • Understand how XNU is structured (Mach, BSD, IOKit) and how the major subsystems fit together.
  • Apply practical workflows for kernel crash triage and root-cause analysis (panic logs, call paths, and common failure modes).
  • Reason about kernel attack surface across IOKit and modern driver models, including DriverKit/SystemExtensions and EndpointSecurity.
  • Build a usable mental model of VM/page tables/permissions and how these concepts show up in real macOS/iOS security investigations.
  • Understand Apple proprietary ARM64 hardware-assisted security mitigations (e.g. CTRR/SPRR/GLx) and how they are used in practice.
  • Identify common kernel bug classes (lifetime, validation, memory corruption patterns) and where to look for them in code and interfaces.

Overview

With macOS Tahoe and iOS 26 Apple continues to harden the platform with ARM64-specific security mitigations and additional security boundaries. This training focuses on kernel internals from the perspective of a security researcher interested in vulnerability analysis, rootkit/malware analysis and detection, or general kernel security topics.

The goal is not just to list subsystems, but to build a practical mental model you can use when you debug crashes, audit attack surface, or reason about mitigations. We walk through how the major parts of XNU fit together (Mach, BSD, IOKit) and how data flows across kernel boundaries in real systems. Throughout the course we connect “what the kernel is doing” to the bug classes and failure modes that matter in practice: memory corruption, lifetime issues, validation mistakes, and security boundary assumptions.

The format is lecture-driven with daily hands-on tasks. After each session you work on exercises, and the next day starts with solution discussion and common pitfalls. That structure helps you build repeatable workflows instead of memorizing trivia.

Syllabus (PDF)

Download the syllabus PDF: Syllabus (PDF)

Training format

  • Live online via Zoom

  • Around 5 hours of live lecture per day plus breaks

  • After each day: tasks/exercises to work on

  • Next session starts with discussion of solutions and pitfalls

Related training
Deep Dive into SPTM / TXM / SK and Exclaves
Kernel Internals for Security Researchers includes a short overview of SPTM, TXM, SK and Exclaves. The Deep Dive covers these components in much more detail with research workflows and tooling. Ask for bundle pricing if you plan to take both trainings.

Topics

The following list of topics shows what will be covered by the course.

Introduction

  • Setting up a development and debugging environment

  • Developing kernel extensions, and how that compares to modern alternatives such as system extensions

  • Running kernel extensions in Apple virtual machines, including constraints and practical workarounds (details provided during the training)

Low-level x86_64 / ARM64 foundations

  • CPU details relevant to kernel work

  • Exception handling and control flow (as needed for debugging and analysis)

  • Physical memory management concepts

  • Hardware page tables and memory permissions

  • Apple and iOS specific system registers and concepts used in practice

  • PAN, PAC (Pointer Authentication) and MTE, and what they change for analysis and debugging

Kernel source code navigation

  • Structure of XNU sources and where to find specific subsystems

  • Reading kernel code with a security mindset

  • Where vulnerabilities tend to live, and why

  • How mitigations show up in code paths

Drivers and extensions

  • IOKit fundamentals and attack surface patterns

  • DriverKit and SystemExtensions, what changed and what it means for security work

  • EndpointSecurity and common inspection/control points

  • Code signing and trust decisions for drivers and related components

Kernel internals

  • Important kernel data structures you will encounter during analysis

  • Mach-O basics relevant to kernel and system components

  • Mach messages and IPC at a security-relevant level

  • Security subsystems overview: sandboxing, MAC policy hooks, code signing, kauth, socket filters

  • Filesystems and networking stack (security-relevant view)

Kernel debugging

  • Panic logs and practical crash triage

  • Built-in kernel debugging facilities and how to use them effectively

  • VM-based debugging

  • Debugging via Apple Virtualization framework using a GDB stub (details provided during the training)

  • A repeatable approach for inspection and root cause analysis

  • Kernel heap debugging and visualization techniques

Kernel heap

  • How kernel allocations work in practice

  • Modern kalloc zones and common bug classes

  • Typical lifetime issues, validation mistakes, and memory corruption patterns

Apple proprietary ARM64 hardware-assisted mitigations

  • KTRR / CTRR / XTRR and related concepts

  • APRR / SPRR style permission models (overview and practical implications)

  • PPL and GXF, and how they affect kernel behavior and analysis

  • How these mitigations are used by the kernel in real scenarios

Newer security components overview

  • High-level overview of SPTM, TXM, SK and Exclaves

  • How these components relate to the kernel and to platform security boundaries

Training takeaways

  • Training material (slides) is handed out in digital form.

  • Recordings are available for a limited time after the training.

  • Trainees receive a license for Antid0te tooling/scripts used during the training (usage allowed, redistribution not allowed).

Training requirements

Student requirements

  • Basic understanding of exploitation (note: not an exploitation training)

  • C and Python programming knowledge

  • Basic knowledge of x86_64 and/or ARM64 assembly

Hardware requirements

  • Apple Mac capable of running the latest macOS inside a virtual machine (needs to be able to boot in recovery mode)

  • Enough disk space to run VMs

Software requirements

  • Disassembler capable of understanding ARM64/x86_64 macOS/iOS binaries (IDA, Ghidra, Binary Ninja)

  • macOS Sequoia/Tahoe with Xcode and iOS SDK (or newer)

  • VirtualBuddy (ARM64) / VM-based tooling as required

  • Additional software will be provided during the training

Virtual venue

The sessions are held via Zoom. Trainees also get access to a Discord server used for information and exercise discussions. Recordings are made available for a limited time after the training.

Timezones

We offer this training in an EU/North America edition and an APAC edition. Live lecture blocks are followed by hands-on exercises trainees complete before the next day.

EU / North America edition
17:00 – 22:00        Berlin
16:00 – 21:00        London
08:00 – 13:00        Seattle / Vancouver
11:00 – 16:00        New York / Montreal
23:00 – 04:00 (+1)   Singapore
01:00 – 06:00 (+1)   Sydney
APAC edition
10:00–12:30 / 13:30–16:00   Singapore (SGT)
09:00–11:30 / 12:30–15:00   Bangkok / Jakarta
11:00–13:30 / 14:30–17:00   Tokyo / Seoul
12:00–14:30 / 15:30–18:00   Sydney / Melbourne
07:30–10:00 / 11:00–13:30   India
14:00–16:30 / 17:30–20:00   Auckland / Wellington

Pricing

We offer the following prices for the training.

Price per attendee
EUR 4500,- EUR
SGD 6500,- SGD
USD 5000,- USD

Payment is possible via international bank transfer or via credit card (Stripe). We usually charge in SGD, but can charge in EUR and USD if requested.

Register

If you have questions or want to register for this training, email us. Signup, billing and execution of the training is handled by Antid0te SG Pte. Ltd.

Private / in-house sessions

If none of the scheduled dates fit your timezone, or you want a private company session (remote or on-site), email us with your preferred time window, headcount, and topic focus.