Don't check for alias bounds in liveness when aliases have escaping bound vars

[compiler-errors]

I actually have no idea how we should be treating aliases with escaping bound vars here... but the simplest behavior is just doing what we used to do before.

r? aliemjay

Fixes #117455

[tmandry]

Friendly ping, this is breaking our build and I'd rather see a fix forward land soon.

If it helps: I haven't seen the code where this visitor is used, but the diff looks reasonable and the strategy of "do what we did before" sounds safe. If you're confident it's not horribly broken you can r=me.

[aliemjay]

I have thought about "smarter" solutions but they all turned to be incorrect (unsound?) because of the way we handle well-formedness around higher ranked types. Now I'm fairly convinced that the approach of this PR is the only way handle escaping bound vars.

Accordingly I'd prefer removing the FIXME comment and adding the following test to make sure we are not trying to do anthing smarter in this inference but you can r=me without these changes.

trait Trait {
    type Gat<'a: 'b, 'b: 'c, 'c>: 'c;
}

fn get_func<'a, T: Trait>(_: &'a str) -> fn(T::Gat<'a, '_, 'static>) {
    loop {}
}

fn test<T: Trait>() {
    let func = get_func::<T>(&String::new()); //~ ERROR temporary value dropped
    drop(func);
}

[lcnr]

this test doesn't contain an opaque. Please add a comment referencing #117455 and one or two sentences explaining what this is testing

[compiler-errors]

Added aliemjay's test, added a comment and moved the original test cause it didn't have a TAIT anyways.

@bors r=aliemjay

[bors]

📌 Commit 4d5d763 has been approved by aliemjay

It is now in the queue for this repository.

[bors]

⌛ Testing commit 4d5d763 with merge b800c30...

[bors]

☀️ Test successful - checks-actions
Approved by: aliemjay
Pushing b800c30 to master...

[rust-timer]

Finished benchmarking commit (b800c30): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

This benchmark run did not return any relevant results for this metric.

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.9%	[-2.9%, -2.9%]	1
All ❌✅ (primary)	-	-	0

Cycles

This benchmark run did not return any relevant results for this metric.

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 639.068s -> 636.285s (-0.44%)
Artifact size: 304.47 MiB -> 304.50 MiB (0.01%)