Garbage collection

[liquidev]

Some things simply cannot be collected when using reference counting, and even though GC is less predictable, that's not really something you worry about in high-level languages such as Mica.

[liquidev]

Turns out GC is way faster than RC, too. Making values not have an explicit Drop implementation prevents a lot of branching that would otherwise occur in a destructor.
(see #55)

> hyperfine --warmup 5 'target/release/mica code/functions.mi' 'target/release/mica-nodestructors code/functions.mi'
Benchmark 1: target/release/mica code/functions.mi
  Time (mean ± σ):     344.4 ms ±   5.1 ms    [User: 342.1 ms, System: 1.9 ms]
  Range (min … max):   338.4 ms … 356.3 ms    10 runs
 
Benchmark 2: target/release/mica-nodestructors code/functions.mi
  Time (mean ± σ):     288.2 ms ±   1.5 ms    [User: 286.8 ms, System: 1.4 ms]
  Range (min … max):   286.6 ms … 291.2 ms    10 runs
 
Summary
  'target/release/mica-nodestructors code/functions.mi' ran
    1.20 ± 0.02 times faster than 'target/release/mica code/functions.mi'

I'm adding this to 0.3.0.

[liquidev]

The GC should be controllable by the user, using a built-in global Gc. The following methods should be available:

impl Gc
  # Disables automatic garbage collection; collect/0 static must be called to prevent leaks.
  func disable() static

  # Makes the GC run on most allocations. In most cases this is not what you want, as this is only useful for
  # debugging the GC.
  func enable_always_run() static

  # Makes the GC run periodically every time an amount of bytes is allocated.
  # `next_run` then gets updated to `allocated_bytes * growth_factor / 256`.
  # This is the default, with `next_run = 65536` and `growth_factor = 384`.
  func enable_with_ceiling(next_run, growth_factor) static

  # Triggers a garbage collection now.
  func collect() static

  # Returns the amount of bytes currently allocated by the GC.
  func allocated_bytes() static
end

[liquidev]

I'm a little worried that maybe the GC doesn't fire often enough. Not all allocations are managed by it (foreign allocations in particular are not controlled by the GC). We'll see how it is in practice.

[liquidev]

I'll call this closed for now; future concerns about the GC's performance or other implementation stuff can be raised in new issues.