feat: add casting of fundamental types to memory mapping

[Wodann]

Adds support for fundamental type conversions, when memory mapping. E.g. going from this struct:

struct Foo(
    u8,
    i16,
    u32,
    i64,
    f32,
)

to the following struct, is legal:

struct Foo(
    u16,
    i32,
    u64,
    i128,
    f64,
)

[Discussion] Non-guaranteed type conversions will result in a panic zero initialization. Is this desirable? Potentially we could at least attempt the conversion and only panic! zero initialize if the try_into call fails.

[Discussion] Non-legal or failed type conversions will result in a panic!. Is this desirable? Alternatives are:

1. Logging an error and initialising to zero.
2. Graciously fail the linking of the Assembly. Changes will have to be made to the source code before we re-evaluate the Assembly. This means that you can never "reuse" a field by doing an illegal conversion. This might force people to take a two-step hot reloading approach: i.e. remove first and then add the new field.

Edit: after discussion, we decided to never panic. If we cannot resolve the type conversion, we continue with the mapping - assuming zero initialization.

[codecov]

Codecov Report

Merging #140 into master will increase coverage by 0.26%.
The diff coverage is 96.15%.

@@            Coverage Diff             @@
##           master     #140      +/-   ##
==========================================
+ Coverage   81.80%   82.06%   +0.26%     
==========================================
  Files         159      160       +1     
  Lines       10672    10806     +134     
==========================================
+ Hits         8730     8868     +138     
+ Misses       1942     1938       -4     

Impacted Files	Coverage Δ
crates/mun/src/main.rs	0.00% <0.00%> (ø)
crates/mun_abi/src/lib.rs	92.85% <ø> (ø)
crates/mun_abi/src/static_type_map.rs	90.90% <ø> (ø)
crates/mun_runtime/src/reflection.rs	54.34% <ø> (ø)
crates/mun_runtime/src/lib.rs	71.73% <50.00%> (+0.77%)	⬆️
crates/mun_abi/src/type_info.rs	74.35% <77.77%> (ø)
crates/mun_abi/src/function_info.rs	100.00% <100.00%> (ø)
crates/mun_memory/src/cast.rs	100.00% <100.00%> (ø)
crates/mun_memory/src/gc/mark_sweep.rs	86.71% <100.00%> (+1.20%)	⬆️
crates/mun_memory/tests/diff/mod.rs	100.00% <100.00%> (ø)
... and 9 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 54580ca...6b4c47d. Read the comment docs.

[Wodann]

@baszalmstra any comments on the discussion topics in this description?

[baszalmstra]

[Discussion] Non-guaranteed type conversions will result in a panic. Is this desirable? Potentially we could at least attempt the conversion and only panic! if the try_into call fails.

I'd say that it should never panic! because I user can never catch that.

[Discussion] Non-legal or failed type conversions will result in a panic!. Is this desirable?

No, I don't think we should ever panic!. I think this is something that we have to play with before we can really make a decision. So far I think option number 1 is the most user friendly.

[Wodann]

I'll add a commit to init to zero, in that case 👌🏻

[Wodann]

I've updated the PR description to reflect that we decided to never panic! and added a separate commit to implement and test this behavior.

Could you review that commit please @baszalmstra ?