Good localtime,
I am the project manager assigned to CS/TSDL Project by Linaro. I hope this journey will be something pleasant and fruitful for all the participants of this project as well as for the community around ARM technologies and Linux.
So, let's start with an answer to "what?" question.
In order to be able to properly plan the scope of the project we need to define the first approximation of the project work item structure. Quite many of the people subscribed to this mailing list have practical experience implementing trace stream decoders for CoreSight. That increases our chances of defining a decent approximation of the project scope, which then will be refined gradually as project continues.
I need your help brainstorming this topic. So, please, throw your ideas about the potential work items into this thread and I'll be summarizing your ideas and putting them into a single "backlog", which will get prioritized according to needs and dependencies afterwards.
Usage of a hierarchical representation of the work items might be a good idea as it allows seeing the dependencies from the very beginning. So, something along the lines of the example below is proposed to be used (the structure below is just an example - I don't have enough technical knowledge for the area to propose a good initial structure):
1. Frontend 1.1. ... 2. Core 2.1. ... 2. Backends 2.1. ETMv3 2.1.1. ... 2.2. ETMv4 2.2.1. ... 2.3. STM 2.3.1. ... 3. ...
Also, please share your relevant DOs and DONTs from your past experience. This undertaking needs all your skills and knowledge to create a fast, reliable and easy to use decoding library.
PS. "CS/TSDL" in the subject stands for "CoreSight/Trace Stream Decoding Library". Proposals for a better name are welcome. :)