doc: clarify allocator invariant

[lms0806]

Added base + size <= usize::MAX to the Allocator documentation.

Please refer to issue 153542.

Close #153542

[rustbot]

r? @scottmcm

rustbot has assigned @scottmcm.
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: @scottmcm, libs
• @scottmcm, libs expanded to 8 candidates
• Random selection from Mark-Simulacrum, joboet, scottmcm

[scottmcm]

I was pondering where best to put this, and whether it should be more related to the "currently allocated" conversations, but in the end I think here is good. If you wanted to be hyper-specific about a "proof" for why a offset didn't wrap, you'd want something in the safety section here, since the methods on the trait themselves can't impose postconditions that can be relied upon for soundness.

And I think this is "obviously" a requirement here, so as the trait is unstable I think I'm good to just 👍 it without involving an FCP from any team. They can look at it when contemplating stabilizing the trait.

@bors r+ rollup

[rust-bors]

📌 Commit 57fe818 has been approved by scottmcm

It is now in the queue for this repository.