Shorten drop scope of temporaries in conditions

[azizghuloum]

During AST->HIR lowering, instead of transforming

if e1 && e2 ... { --- }

to

if DropTemps(e1' && e2' ...) { --- }

now it gets transformed to

if DropTemps(e1') && DropTemps(e2') ... { --- }

leading to subsequent passes (in MIR) to recognize the && pattern in if e1 && e2 which was previously obscured by unneeded Use markers.

Note that the && itself never constructs temporaries and thus doesn't need to be wrapped by DropTemps.

Fixes #111583
Replaces #111644

[rustbot]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @davidtwco (or someone else) soon.

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

• @rustbot author: the review is finished, PR author should check the comments and take action accordingly
• @rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

[azizghuloum]

@cjgillot @dingxiangfei2009 does this address the concerns with the temporaries? thanks in advance.

[rustbot]

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

[cjgillot]

Nominating this issue for lang team discussion.

This PR makes 2 user-visible changes to the language:

1. Temporaries in && conditions are dropped earlier: DropTemps(e1 && e2) becomes DropTemps(e1) && DropTemps(e2). This behaviour already exists with let chains, when one of the conditions is a let.

2. The MIR built for a if e1 && e2 { is now equivalent to the MIR for if e1 { if e2 {, which changes how borrowck understands that code, see tests/ui/rfc-2497-if-let-chains/chains-without-let.rs.

Point 1 may create unexpected bugs for users that rely on drop order for synchronization (for instance). A crater run may be useful, but I'm not sure it will find such bugs.

[cjgillot]

Why is the test in this function unchanged?

[azizghuloum]

Sorry. I don't understand the question. It is changed. I removed the ~^ (meaning, it's no an error) and the output in tests/ui/rfc-2497-if-let-chains/chains-without-let.stderr shows that.

[azizghuloum]

Or do you mean I need to merge master to resolve the conflict?

[cjgillot]

I mean in the other function and_chain_2.

[azizghuloum]

Because so far I'm only doing the &&s inside the test of if, not && as expression.

I agree that that too should be fixed, along with the || cases.

If this branch gets merged, I will fix the other cases.

[nikomatsakis]

Discussed in the rust-lang/lang meeting.

We felt this change would require an edition. @dingxiangfei2009 and I are exploring some changes to temporary lifetimes for rust 2024 so perhaps you can reach out on Zulip and we can discuss there.