Develop and enhance an automated flow (Open Debug (OD)) for the Periodic System Management Interrupt (PSMI) when a failure is captured by Open Debug (OD) using Python

Ng, Shin Huan (2023) Develop and enhance an automated flow (Open Debug (OD)) for the Periodic System Management Interrupt (PSMI) when a failure is captured by Open Debug (OD) using Python. Master dissertation/thesis, UTAR.

Preview

PDF Download (1982Kb) | Preview

Abstract

Consistency, reliability and accuracy of any implemented machine products are highly demanded by everyone in this world no matter which level of the workplace the user is. In order to achieve this goal, various advanced technologies are introduced with automated technique which is aimed to replace the traditional manual method – manpower. This applies the same in the post-Si environment. The debugger needs to understand the behaviour of that particular failure, hence PSMI tool is needed to capture it. However, this tool is needed to be used manually by every debugger. In order to automate it, this project will summarise the flow and that automated system will be inserted into a flow named Open Debug (OD). This has aimed to reduce the time taken and effort needed to capture the PSMI trace. In Literature Review, various fields are introduced with automated methods in order to achieve time and effort saving goal and at the same time increasing the efficiency and effectiveness of the results produced as compared to manual method. To design the automated way for PSMI capture, Python script is utilised. In Methodology, the design of the automated flow of the PSMI capture in the OD flow is explained in detailed as well as in the flow chart manner. For this developed Python script to be triggered, a failure has to be detected at the first stage. The trigger point used in the PSMI capture is mainly the memory trigger. This method is basically to detect the offset 0x28 of the error block address associated with the specific data depending on the error code. 4 projects are selected to test the developed script by comparing the manual method (the common way that all debuggers will use) and the automated way (using the implemented Python script that is inserted in the OD flow). The manual method time estimation for PSMI capture is computed based on survey from debuggers while the automated method time computation for PSMI capture is the average analysis from each of the project using the developed Python script. Based on the results obtained, using automated technique can greatly save 55% to 84% of the time used over the manual technique in capturing PSMI. In conclusion, by using automated approach for PSMI capture in the OD flow, it is proven that the time and effort can be effectively saved as compared to the manual approach.

Actions (login required)