This article link: https://www.ebpf.top/en/post/ebpf_and_kernel_feature BPF kfuncs Bloom Filter Map: 5.16 Compile Once – Run Everywhere: Linux 5.17 [Kernel Space] bpf_loop() Helper Function: 5.17 BPF_LINK_TYPE_KPROBE_MULTI: 5.18 Dynamic Pointers and Type Pointers: 5.19 USDT: 5.19 BPF panic: 6.1 BPF Memory Allocator, Linked Lists: 6.1 User Ring Buffer: 6.1 User-defined linked list support: 6.2 User-defined rbtree support: 6.3 BPF Generic Iterator: 6.4 References: In 2022, the Linux kernel primarily released versions 5.16-5.19, 6.0, and 6.1, each introducing numerous new features for eBPF. This article briefly introduces these new features. For more in-depth information, please refer to the provided links. Overall, eBPF remains one of the most active modules in the kernel, with its functionality continuously evolving at a rapid pace. In a sense, eBPF is rapidly evolving towards a comprehensive kernel-state programmable interface. BPF kfuncs The BPF subsystem exposes many aspects of kernel……

This article is available at: https://www.ebpf.top/post/xdp_lb_demo Author: Qiu Kang Creating network environment SLB Cluster Routing NAT Mode Load Balancing Algorithms With the progress of eBPF, we can now deploy eBPF/XDP programs directly on regular servers to achieve load balancing, saving the need for dedicated machines for LVS deployment. The previous article shared how to use xdp/ebpf to replace LVS for SLB. It adopted the independent machine deployment mode for SLB and loaded xdp program using bpftool and hardcoded configuration, which was version 0.1. Version 0.2 modified the 0.1 version to a programmatic loading mode based on BPF skeleton. To experience this workflow easily without changing the overall deployment mode of version 0.1, you can check out https://github.com/MageekChiu/xdp4slb/tree/dev-0.2 Version 0.3 added support for dynamically loading SLB configurations in the form of configuration files and command-line parameters. This article belongs to version 0.……

This article can be found at: https://www.ebpf.top/post/get_start_chatgpt 1. Introduction 2. Direct Experience 3. Official Account Registration 3.1 Account Registration 3.2 International SMS Verification 3.3 Testing the Effect 4. Summary References 1. Introduction ChatGPT is likely familiar to most people, sweeping the entire Internet and becoming a hot technology pursued by tech companies and a topic of great interest in the IT industry. ChatGPT is an AI chatbot program developed by OpenAI, launched in November 2022. ChatGPT is a large language model developed and trained by OpenAI, using a large language model based on the GPT-3.5 architecture and trained through reinforcement learning. It can generate high-quality text and performs well in various language tasks such as conversation generation, text summarization, text classification, etc. Some say it’s a career killer for coders, a firefighter of inspiration for poets, a dream crusher for “town problem solvers”… Give it a simple command, and it can code, write poetry, or research papers.……

This article can be found at: https://www.ebpf.top/post/ebpf_rust_aya 1. Introduction 2. Setting Up Rust Development Environment 2.1 Create a VM Virtual Machine 2.2 Install Rust Development Environment 2.3 Install Dependencies for bpf-linker and bpftool 3. Aya Guide to Create eBPF Programs 3.1 Creating a Project Using the Guide 3.2 Compile eBPF Program 3.3 Run User-space Program 4. Conclusion References 1. Introduction A significant change in Linux Kernel version 6.1 is the introduction of support for the Rust programming language. Rust is a system programming language that offers robust compile-time guarantees and precise control over memory lifetimes. Integrating Rust language into kernel development will bring additional safety measures to the early stages of kernel development. eBPF is a technology in the kernel that enables running user-defined programs based on events, with a validator mechanism ensuring the security of eBPF programs running in……

Site link: https://www.ebpf.top/post/bpf2pbpf_tail_call Author: Richard Li (Original author’s permission obtained) Original article link: https://blog.csdn.net/weixin_43705457/article/details/123474244 1. Introduction 2. Tail Call 3. BPF to BPF Calls 4. CO-RE Sample 5. Tail Call Costs in eBPF 6. Summary 7. References 1. Introduction This article first introduces the general restrictions and usage of tail calls, compares them with BPF to BPF calls, and finally provides a modified version I made of the tail call sample in the kernel source code (using CO-RE). (When learning about tail calls, I struggled with not having a simple and understandable example that could run, so I ended up creating one myself. I believe this version is the most beginner-friendly and logically clear among all examples currently available). 2. Tail Call BPF provides a capability to safely inject code when kernel events and user program events occur, allowing non-kernel developers to control the kernel.……

This article can be found at: https://www.ebpf.top/post/ftrace_kernel_dynamic 1. Basic Knowledge 1.1 Default Compilation 1.2 Using the -pg Option 1.3 Using the -pg and -mfentry Options 1.4 Kernel Verification 2. Validation of kprobe tracing mechanism in ftrace 3. Validation using a Kernel Module 3.1 Using kallsyms_lookup_name 3.2 Using Kernel Function Addresses Directly (Pitfalls, Optional) 4. Verification using gdb + qemu Reference In the Geek Time’s “Mastering Container Battles” by Teacher Li Chengyuan, a pondering question was left in the extra session 04 | Understanding ftrace(2): How to Understand the Technology Behind ftrace Tracepoint and kprobe?: Consider this, how can we observe that the first instruction of the corresponding kernel function has been replaced after we register a probe with kprobe for it? Kprobe is a mechanism for dynamic tracking of kernel functions. By using this mechanism, almost all kernel functions can be tracked (excluding those annotated with __kprobes/nokprobe_inline and those marked with NOKPROBE_SYMBOL).……

This article can be found at: https://www.ebpf.top/post/no_space_left_on_devices 1. Understanding “no space left on device” Error 2. Problem Analysis and Localization 2.1 Preliminary Identification of Problematic Function 2.2 Locating the Root Cause of the Issue 2.3 Identifying the Root Cause of the Issue 3. Analysis of Discrepancies Between Code Flow and Tracing Process References Recently, there have been cases of failures in creating containers with the error “no space left on device” in the production environment. However, during the investigation, it was found that disk space and inodes were quite normal. In cases where conventional troubleshooting methods have failed, is there a quick and universal approach to pinpointing the root cause of the problem? This article records the analysis and troubleshooting process using eBPF + Ftrace in a separate environment. Considering the general applicability of this approach, it has been organized in the hope of serving as a stepping stone for further exploration.……