As I alluded to in my previous post, I am now back in Santa Barbara (or rather Goleta). This means I have some additional resources from my lab and professors, but also a lot of additional responsibilities from my classes, dealing with my apartment, etc. This means that I will have a smaller amount of time to work on GSoC, but I will make sure to work on the project as much as I can and continue to keep you all updated.

libcups Testing

Thankfully, libcups seems to finally be compiling and linking fine (besides the parts with mDNS Responder) after changing some Mbed TLS configs. I have also shifted the testing framework over from using stdio to Zephyr logging since I was receiving load-prohibited exceptions (like the one I mentioned in the previous post) from these functions. After modifying the tests to account for the filesystem, I was able to get testarray and testclock to pass, but now I am trying to tackle testcreds.

Currently, whenever I try to run testcreds, I get no output from my program’s logging, both with and without CONFIG_LOG_MODE_IMMEDIATE on. The last time I saw this was when I was testing PDFio using the md2pdf program, which likely caused a heap overflow or some memory overflow. I think that testcreds also uses quite a lot of stack memory as seen in the 30KB+ buffers for some of the TLS functions on top of the heap allocations, so I am suspecting a lack of memory here, despite any concrete evidence like an exception that would indicate a stack overflow. Therefore I will also work on reducing the memory footprint of these functions and the library as a whole. Lastly, I should also mention that there is a possibility that file i/o may be causing some issues here since file i/o is more difficult compared to on Linux and I do not have the CA certificate files on my board that a Linux system would have.

Other Thoughts and Next Steps

I currently do not have a board for JTAG debugging with my ESP32, but I should be getting one delivered relatively soon, which should make debugging easier (especially for testcreds). Otherwise, I will look toward tackling some of the other tests and seeing how I could implement mDNS Responder. The latter combined with my test suite’s already high memory usage makes me especially worried about memory constraints in the future, however. Since I do not have very much experience with memory optimization on the embedded level (CS classes only really seem to care about memory complexity in Leetcode), I will have to discuss this issue further throughout the porting process. Otherwise, you can find the current design diagram and proposal/timeline below.

Project Proposal:
https://docs.google.com/document/d/1cYL6S2JSkzY0ln1w_s3qwpm85-keB9jaVjGSwyPg5NM/edit?usp=sharing

Design Diagram:
https://excalidraw.com/#room=0a35b04c6340f86386ea,xwV6ErRli602L8FTUikm2w