heavy revision on heap_breakdown's safety

[paperdave]

What does this PR do?

I wish we had some internal docs on heap breakdown. It is very cool. This PR makes it more safe for us to use.

Rename bun.HeapBreakdown to bun.heap_breakdown. Rename bun.is_heap_breakdown_enabled to bun.heap_breakdown.enabled.

To use the heap breakdown allocator functions directly, recommend using heap_breakdown.getZone(T).create(T, t). But it is better to simply use the bun.New mixin, which opts you into heap breakdown automatically.

Refactors the following: bun.new, bun.destroy, bun.create, bun.New, bun.NewRefCounted. They are much easier to read now.

Main motivation for this is to introduce a new assertion which ensures types that forcefully opt into heap breakdown (From bun.New) are impossible to create via standard library allocators.

[paperdave]

re: rename HeapBreakdown -> heap_breakdown

I want to slowly move parts of our codebase into proper casing.

[github-actions]

❌ @paperdave, your commit has failing tests :(

🪟💻 2 failing tests Windows x64 baseline

• test/cli/install/registry/bun-install-registry.test.ts 2 failing
• test/js/node/child_process/child_process.test.ts 1 failing

🪟💻 2 failing tests Windows x64

• test/cli/install/registry/bun-install-registry.test.ts 1 failing
• test/js/node/child_process/child_process.test.ts 1 failing

View logs