Various cleanup

[micprog]

• Update build flow/testbench
• Fix testbench for badly addressed bursts
• Add types to all parameters
• Replace AXI interfaces with structs

[micprog]

I can also split this PR up a bit, there are quite a few changes coming together here...

[micprog]

There were some more issues due to bad testing on my part, sorry. Should be fixed now (with some speed improvements and no more warnings). Might also be worth it to ensure the CI works properly and reports fails when it should...

[yvantor]

There were some more issues due to bad testing on my part, sorry. Should be fixed now (with some speed improvements and no more warnings). Might also be worth it to ensure the CI works properly and reports fails when it should...

I think the problem is that the CI greps for a fatal or some error message (linked to the returned value I think) to understand if a failed, and if the grep does not succeed (as in the case of a test that did not even start) it just returns success. I agree that this should be improved. Also making the regressions' outputs more homogeneous (like all of them printing either "Success!" or "Failed!" at the end) could maybe help.

[yvantor]

@OttG as soon as @micprog merges [#44] and the CI completes, this can also be merged.

[OttG]

Everything seems fine now, CI is passing. Merging now.