Michael Goulet comments

Results 300 comments of


                                            Michael Goulet

Stop storing a special inner body for the coroutine by-move body for async closures

Ah wait, mir-opt tests are gonna explode because of the new naming scheme.

Stop storing a special inner body for the coroutine by-move body for async closures

Anywhomst, cc @RalfJung if you're curious about what this ended up looking like.

Stop storing a special inner body for the coroutine by-move body for async closures

Let's get perf kicked off. I'll fix the codegen tests when I get home. @bors try @rust-timer queue

Stop storing a special inner body for the coroutine by-move body for async closures

If this doesn't help things, I'll investigate adding a new `DefKind::SyntheticBody` or something, to avoid needing to add `tcx.is_synthetic_mir` as a query.

Stop storing a special inner body for the coroutine by-move body for async closures

@bors try @rust-timer queue

Stop storing a special inner body for the coroutine by-move body for async closures

@bors try @rust-timer queue

Stop storing a special inner body for the coroutine by-move body for async closures

@bors try @rust-timer queue

Dont consider predicates that may hold as impossible in `is_impossible_associated_item`

I think the overhead is fine.

[perf] Tweak `GenericArgs`-related functions

r=me when ci is green or sooner i dont care much

`tower-sessions-core` crate rework

Hey 👋 I saw this cross-linked from one of my rust compiler PRs, so I'm just reaching out: > This will be solved with https://github.com/rust-lang/rust/issues/109417 which is hopefully close to...