Switch to EarlyBinder for explicit_item_bounds
#110556
Conversation
rustbot
added
S-waiting-on-review
|
Some changes occurred to MIR optimizations cc @rust-lang/wg-mir-opt Some changes occurred in src/tools/clippy cc @rust-lang/clippy |
…icit_item_bounds usages; remove bound_explicit_item_bounds query
…lyBinder; use this to simplify some EarlyBinder noise around explicit_item_bounds calls
kylematsuda
force-pushed
the
earlybinder-explicit-item-bounds
branch
from
c7575e9
to
f8a95a7
Compare
| tcx.explicit_item_bounds(item_def_id).to_vec(), | ||
| tcx.explicit_item_bounds(item_def_id) | ||
| .subst_identity_iter_copied() | ||
| .collect::<Vec<_>>(), |
There was a problem hiding this comment.
is the turbofish necessary here? Alternatively, you could do something like subst_identity().to_vec(), right?
There was a problem hiding this comment.
This maybe ties into a larger question (sorry for the long comment below):
Right now, EarlyBinder<T>::subst_identity has the bound T: TypeFoldable, so it's not actually available here. I was considering getting rid of the TypeFoldable bound but wasn't sure how important it is to how EarlyBinder should work -- i.e., right now you can only call subst_identity on things that you could also call subst on. (Maybe this property would come in handy in the future if doing larger refactorings?)
The lack of subst_identity is what motivated me to add subst_identity_iter_copied in this PR, as an identity version of subst_iter_copied.
Haha but now I'm worried that I'm overthinking and over-complicating this. If it's alright to do subst_identity on any inner type, then we don't have to add new methods on EarlyBinder in this PR, and every subst_identity_iter_copied would just become subst_identity().iter() (or in the case above, subst_identity().to_vec() as you mentioned). What do you think?
There was a problem hiding this comment.
Hmmmmmmm.. maybe we could implement subst_identity on anything that implements TypeVisitable? We probably don't need it to be TypeFoldable, but I could potentially see us using some folder to make sure that it's actually valid to subst identity?
kylematsuda
force-pushed
the
earlybinder-explicit-item-bounds
branch
from
f8a95a7
to
5a69b5d
Compare
|
Thanks for your comments! I amended the last commit to change those Please see my response to your other comment, I had a question relating to it 😅 |
|
@bors r+ |
bors
added
S-waiting-on-bors
…iaskrgr Rollup of 6 pull requests Successful merges: - rust-lang#110556 (Switch to `EarlyBinder` for `explicit_item_bounds`) - rust-lang#110615 (Add `impl_tag!` macro to implement `Tag` for tagged pointer easily) - rust-lang#110649 (Fix no_global_oom_handling build) - rust-lang#110671 (Consider polarity in new solver) - rust-lang#110783 (Fix ICE on --print=... i/o errors) - rust-lang#110796 (Updating Wake example to use new 'pin!' macro) Failed merges: r? `@ghost` `@rustbot` modify labels: rollup
…tem-bounds, r=compiler-errors Switch to `EarlyBinder` for `explicit_item_bounds` Part of the work to finish rust-lang#105779. This PR adds `EarlyBinder` to the return type of the `explicit_item_bounds` query and removes `bound_explicit_item_bounds`. r? `@compiler-errors` (hope it's okay to request you, since you reviewed rust-lang#110299 and rust-lang#110498 😃)
Part of the work to finish #105779.
This PR adds
EarlyBinderto the return type of theexplicit_item_boundsquery and removesbound_explicit_item_bounds.r? @compiler-errors (hope it's okay to request you, since you reviewed #110299 and #110498 😃)