eBPF 高级可观测性工程：fentry/fexit、BPF Trampoline 与 struct_ops 深度实战
深度解析eBPF三大高级特性——fentry/fexit零开销函数追踪、BPF Trampoline动态函数重定向引擎、struct_ops以BPF替换内核函数指针，涵盖内核源码级原理、完整代码示例和生产环境性能基准测试。
systems
eBPF Kernel Observability Deep Dive: Engineering Paradigm from Bytecode Injection to Zero-Overhead Tracing
eBPF revolutionizes Linux kernel observability. Covers virtual machine architecture, maps, kprobes, tracepoints, XDP, BTF, ring buffer, bpftrace, BCC, libbpf CO-RE, and four production-grade zero-overhead tracing case studies.
深入理解现代编译器优化PASS架构：从LLVM IR到生产级Pass设计实战
深入剖析LLVM Pass Manager架构与IR级优化Pass设计，涵盖New Pass Manager调度机制、经典Pass实现原理（MemCpyOpt/LoopUnswitch/SROA/GVN）、自定义Pass开发实战、Pass并行化与Cache复用策略，以及ML驱动Pass调度MLGO等前沿进展。
CFS Scheduler Internals: A Deep Dive into Linux Completely Fair Scheduler
Deep dive into the Linux Completely Fair Scheduler CFS internals: vruntime model, red-black tree task selection, group scheduling with cgroups, NUMA-aware load balancing, PELT tracking, and sched_ext extensibility in Linux 6.12+
