fix: many things

[OlivierBBB]

• the enter method of CallFrame now uses a CallDataInfo object
• CallSection's now produce the relevant CallDataInfo object, which the enter(...) method of the CallFrame re-uses
• a CallSection's callDataInfo and returnAtMemorySpan now implement the logic whereby if the size is zero we discard the (meaningless) offset and set the span to EMPTY
• copy paste of the test in @lorenzogentile404 PR + some fixes to it (it likely be replaced by the analoguous parametric tests on the HUB issue)

[OlivierBBB]

This was merged into the big HUB branch. I'm not sure what to do about this PR now 🤔

[OlivierBBB]

I'm closing it as this was already merged into the current HUB branch.