[debuginfo] Emit usize and isize as typedefs

[Walnut356]

See: llvm/llvm-project#196812

TL;DR sidesteps a name collision issue that affects generics, e.g. S<u64> and S<usize> resolving as the same type.

This is accomplished by wrapping an existing primitive node in a usize/isize typedef instead.

I'd love to output typedefs for all the primitive types too, so LLDB displays the type names rust-style, but I've got a couple worries with that. Firstly, GDB doesn't seem to like it w.r.t. enums? IIRC it's because of a small difference in how GDB treats typedefs compared to LLDB. I'm like 80% sure it's hard-coded in GDB itself, so it's not something I want to touch atm.

Breakpoint 1, borrowed_enum::main () at tests/debuginfo/borrowed-enum.rs:67
67          zzz(); // #break
$1 = <error reading variable: Could not find active enum variant>
$2 = borrowed_enum::ABC::TheB(0, 286331153, 286331153)
$3 = borrowed_enum::Univariant::TheOnlyCase(4820353753753434)

Secondly, I'm not entirely sure it's a good idea to have both a type and a typedef with the exact same name in the first place. I could easily imagine that causing some similar issues to the one we're trying to solve.

This also includes a small extention to the Rc test to ensure we can read both Rc<u64> and Rc<usize>. Without this change, the second v command would fail to materialize any value:

...
(lldb) v rc_u64
(alloc::rc::Rc<unsigned long, alloc::alloc::Global>) rc_u64 = strong=1, weak=0 { value = 10 } 
(lldb) v rc_usize

error: Invalid type: Cannot determine size

(lldb) quit

[rustbot]

r? @chenyukang

rustbot has assigned @chenyukang.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler
• compiler expanded to 77 candidates
• Random selection from 18 candidates

[chenyukang]

@rustbot reroll

[Walnut356]

(#163098 means i need to modify the tests again before this gets merged)