For each of the months of January, February and March, I did a webinar on JTAG-based debugging. Since each of the video recordings are about 45 minutes long, I thought it would be helpful to point out the highlights of each, if you don’t have time to sit through the full durations. At the risk of sounding immodest, there are some real gems of information within the demos of each webinar.
Beginning with Microsoft Azure's Project Olympus, and now a standard within the Open Compute Project, many datacenter servers are now optionally equipped with hardware connectivity between the platform BMC and CPU scan chain. The BMC can thus act as an autonomous JTAG-based embedded out-of-band debug agent, provide low-level triage of system events, such as crashes and hangs. Other use cases, such as hardware validation, manufacturing test, and forensics telemetry are also enabled by this technology.
In Part 3 of this series, we did a code review of “ltloop”, a utility firmware application that uses the BMC to do out-of-band stress tests of PCI Express ports. In this article, we begin to examine a more general-purpose application that uses JTAG to extract register, memory and IO contents of the target. This On-Target Diagnostic (OTD), called “libtest”, is used by ASSET to test the functionality of run-control on new targets.
In the last article on this topic, we did a dive into the main routine of the lt_loop JTAG-based On-Target Diagnostic, seeing the overall flow of the program. In this article, we’ll look at the routine that does the heavy lifting for retraining the PCI Express link and checking for errors.
In my previous blog, I did a walkthrough of the source code for main() within the ltloop JTAG-based on-target diagnostic. This article covers main() in more detail, and provides insight into some of the operations of the utility functions and data structures.
In my webinar with the UEFI Forum, I demonstrated some of the utility of using JTAG functionality within BMCs to perform out-of-band debug. This is a tutorial on the coding practices to use the SED API.
This past week, I did a webinar in collaboration with the UEFI Forum on JTAG-based UEFI Debug and Trace. This reviewed some of the often-used tools for low-level triage of difficult-to-diagnose, intermittent bugs. Near the end, I demonstrated the usage of technology running directly down on a BMC to perform low-level functions not achievable with firmware or OS-based applications.