Open Source Summit North America 2022

PCIe is the most commonly used interface standard for connecting high-speed components. Today, every system has a number of PCIe slots you can use to add GPU, RAID cards, Wi-Fi cards, hard disks and SSD cards. As the PCIe standard is incrementing, we see a constant improvement like higher system bus throughput, lower I/O pin count and smaller physical footprint. This also increases the chances of PCIe failures like link instability, TLLP errors, Bad DLLP errors, low power state issues and many more. This presentation proposes a simple command line based diagnostic tool that can be loaded onto any linux system which will collect data and follow a streamlined debugging process for common PCIe issues taking DesignWare PCIe controller as reference. Common tools like setpci/lspci allow us to dump all data but it takes manual effort to understand where the discrepancy is happening. This script, built upon common tools, will make PCIe error solving much simpler and more efficient as it will automatically collect the required data and share the report and initial root-cause analysis with the user.