← Back to context

Comment by koverstreet

6 hours ago

I haven't been following become/TCO in Rust - but what I've usually seen is TCO getting flipped off because it interferes with backtraces and debugging.

So I think there's value in providing it as an explicit opt-in; that way when you're reading the code, you know to account for it when you're looking at backtraces.

Additionally, if you're relying on TCO it might be a major bug if the compiler isn't able to apply it - and optimizations that aren't applied are normally invisible. This might mean you could get an error if you're expecting TCO and you or the compiler screwed something up.

In a language like Rust where local variables are explicitly destroyed when scope ends a naive TCO is very annoying and `become` also helps fix that.

Suppose I have a recursive function f(n: u8) where f(0) is 0 and otherwise f(n) is n * bar(n) + f(n-1)

I might well write that with a local temporary to calculate bar(n) and then we do the sum, but this would inhibit TCO because that temporary should exist after we did the recursive calculation, even though it doesn't matter in practice.

A compiler could try to cleverly figure out whether it matters and destroy that local temporary earlier then apply TCO, but now your TCO is fragile because a seemingly minor code change might fool that "clever" logic, by ensuring it isn't correct to make this change and breaking your optimisation.

The `become` keyword is a claim by the programmer that we can drop all these locals and do TCO. So because the programmer claimed this should work they're giving the compiler permission to attempt the early drop and if it doesn't work and can't be TCO then complain that the program is wrong.