Developing Continuous eBPF Profiler - From Kernel to Cloud
Explore eBPF profiling for interpreted languages and JIT runtimes, discussing challenges in scalability, uniformity, and integration with cloud-native ecosystems. Learn about edge cases and low-level observability implementation.
BPF Programmable Netdevice - Introduction to Netkit
Explore BPF networking advancements, including tcx fast path and the new "netkit" programmable netdevice. Learn about its integration with Cilium and its potential impact on network infrastructure.
Safe Sharing of the Network with eBPF
Explore eBPF-based transport tunings for improved network performance and reliability, covering intra-datacenter and inter-region solutions, custom TCP congestion control, and deployment challenges.
xprobes: Hybrid User/Kernel eBPF Probes for Cross-Layer Observability
Explore xprobes, a hybrid eBPF technology combining uprobes and kprobes for enhanced performance diagnostics. Learn how xprobes provide cross-layer visibility, enabling precise attribution of latency to specific functions and system events.
RTLA Requests and TODOs for Kernel Tracing Infrastructure - Linux Plumbers Conference
Exploring RTLA tool improvements for Linux tracing, including kernel-side histograms, dual tracer support, and enhanced library capabilities for better performance and functionality.
eBPF Shenanigans with Flux
Explore advanced eBPF techniques in Flux framework: data structures, memory allocation, and object-oriented programming for multicore schedulers on Google's Ghost kernel scheduling class.
Multi-KVM: Enabling Concurrent KVM Module Management
Proposal for Multi-KVM: enabling concurrent KVM modules on a Linux host for seamless upgrades, rollbacks, and A/B testing without disrupting running VMs. Refactoring KVM code for improved modularity.
TSCH@Zephyr: IEEE 802.15.4 SubG IIoT in the Making
Exploring TSCH protocol implementation in Zephyr for reliable, low-power industrial IoT networks. Discusses challenges, driver API changes, and precision distributed clocks for various timing applications.
Evolution of Direct Server Return Implementation for Containerized Applications
Exploring the evolution of Direct Server Return (DSR) for containerized apps, addressing challenges in multi-tenancy and network namespaces, and discussing solutions for efficient L4 load balancing.
Using Hardware Hints for Optimal Page Placement
Exploring hardware-provided memory access information for kernel optimization, focusing on NUMA balancing and hot page promotion using AMD EPYC's Instruction Based Sampling mechanism.
Breaking Barriers - Arduino Core API Advancements in Zephyr, Linux and IoT Systems
Explore Arduino Core API integration with Zephyr RTOS for simplified IoT development. Learn key features, advantages, and future possibilities in streamlining embedded systems programming.
Arm64 Live Patching - Current Status and Challenges
Explore the current state, challenges, and missing elements of Arm64 live patching in the Linux kernel architecture.
Linux-WPAN Updates and Future Developments
Explore recent features, challenges, and future developments in Linux-wpan, including team maintainership, pan coordinator services, and link-layer security.
BPF-Based Packet Filtering Framework: bpfilter - Current Capabilities and Development
Explore bpfilter's evolution into a userspace daemon for dynamic packet-filtering BPF programs. Learn about its capabilities, performance, and ongoing development efforts.
Extending AF_XDP with Hardware Metadata
Explore AF_XDP's recent advancements in NIC offloads, its impact on low-level networking, and future developments for enhanced userspace access to networking queues.
VSOCK: From Convenience to Performant VirtIO Communication
VSOCK's evolution from convenience to high-performance protocol for VM-host communication. Explores datagram support, performance advantages over UDP, and sockmap integration for enhanced eBPF capabilities.
Adaptive Userspace Spinlocks with RSEQ - Implementing Efficient Locks in Linux
Exploring efficient userspace spinlocks in Linux using restartable sequences, addressing challenges and proposing solutions for improved performance without system calls.
System Pressure on CPU Capacity and Feedback to Scheduler
Explore techniques to improve task placement and load balancing by reflecting CPU pressure in scheduler decisions, building on previous discussions and upcoming patchsets.
Extending Non-Repudiable Logs with eBPF
Explore how eBPF enhances TPM-based non-repudiable logging, enabling granular system integrity verification and container attestation for improved security and trust in Linux systems.
Improving CPU Isolation with Per-CPU Spinlocks - Performance Cost and Analysis
Explore CPU isolation improvements using per-CPU spinlocks. Analyze performance costs, cache coherence, and contention. Compare with current solutions and discuss strategies for optimizing time-sensitive task execution.
Improving Monitoring of Power Saving States in Linux
Exploring challenges and improvements for monitoring power-saving states in Linux, focusing on ChromeOS's approach to optimizing battery life and performance across various platforms.