Adding RAS Support for CXL Port Devices

Explore the challenges and solutions for implementing Reliability, Availability, and Serviceability (RAS) support for Compute Express Link (CXL) port devices in this Linux Plumbers Conference talk. Dive into the complexities of extending CXL error handling beyond endpoints to include CXL RCH down ports, root ports, USP/DSP switch ports, and CXL host bridges. Examine the limitations of existing endpoint driver handlers and the need for a new approach to accommodate CXL ports as separate PCIe devices. Consider the potential solution of introducing a CXL port service driver to provide common RAS handling across all CXL hardware entities. Analyze the constraints, requirements, and challenges associated with this approach, including binding to CXL port devices, registering error handlers, and refactoring existing CXL endpoint RAS handling for broader compatibility. Gain insights into the current state of CXL RAS, issues overcome, and future considerations for CXL 2.0 and port error handling.