feat(vm): change UhyveVm struct parameters

[n0toose]

As I was moving some functionality away from new(...) so as to progress with my work on ASLR and some future work, I found it necessary to use certain parameters later. A particular example from UhyveVm's current structure would be params.thp and params.ksm, which pose one of the obstacles preventing us from initializing the memory later (e.g. in load_kernel or init_guest_mem, after loading the kernel and being able to establish a guest address), even though separate variables for those don't make much sense.

This change will mostly be useful for future work, but aims to establish a consistent convention now.

[n0toose]

P.S. This change was made before the latest changes to #725 - it seems like I forgot to push it before, but it still makes somewhat sense to me even after the changes. I'll gladly take care of the rebasing for the other PR (as well as explain some more examples via teleconference tomorrow) :)

It's one of those "architectural changes" that I'm unsure about. I'll fix up the clippy warnings and clean things up, but I'd be interested in an opinion.

[n0toose]

I made quite a few decisions that I wasn't very sure about because I felt that they'd bring me to the right direction. Happy to explain.

[n0toose]

memory_size might be useful if we move the memory initialization outside of new(...) (perhaps using a MaybeUninit - not quite sure).

[jounathaen]

But you can access the memory_size via mem already?

[n0toose]

I wasn't exactly sure whether doing this would get in the way of your work, given such a way to access this variable won't be possible after the replacement of the MmapMemory interface in #643.

There is some technical debt in #725 which does the exact same thing, but, as a trade-off, I assumed that a superfluous field would've been better than increasing that technical debt.

[codecov]

Codecov Report

Attention: Patch coverage is 70.37037% with 8 lines in your changes missing coverage. Please review.

Project coverage is 67.26%. Comparing base (5c9a2fd) to head (49b012c).
Report is 22 commits behind head on main.

Files with missing lines	Patch %	Lines
src/vm.rs	63.63%	8 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #755      +/-   ##
==========================================
+ Coverage   67.24%   67.26%   +0.01%     
==========================================
  Files          20       20              
  Lines        2351     2346       -5     
==========================================
- Hits         1581     1578       -3     
+ Misses        770      768       -2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[jounathaen]

It is a bit unfortunate, that we are simultaneously working on this bit of code.
So, I appreciate the work here, and I like some parts of it, but I tend to reject this one. I just think there is not too much value added by adding params to vm.
Give me some more time to finalize #725 and then we'll see how much of this kind of changes remain necessary.