Permit perfect forwarding for Fn traits
#19884
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
@nikomatsakis There are some |
|
Just pushed two more commits that (I think) fix the lingering compile-fail tests, though my local build isn't 100% complete yet. I may revise the "WIP" commits, though I'll try to leave the others so as not to interfere with the review. |
|
@nikomatsakis And this needs a rebase :-) |
nikomatsakis
force-pushed
the
issue-19730-perfect-forwarding
branch
from
27cf881 to
649eab7
Compare
| want_hrtb::<AnyInt>() | ||
| } | ||
|
|
||
| // StaticInt only implements Foo<&'a int> for 'a, so it is an error. |
There was a problem hiding this comment.
As written it is not clear that the 'a being referred to is implicitly bound to one particular lifetime, namely 'static. This should be rewritten as "only implements Foo<&'a int> where 'a == 'static". Or "only implements Foo<&'static int>` if you prefer.
|
Okay, from what I can tell, this all seems to hang together. All of my notes above are largely nits/typos about comments. r=me after you rebase. |
nikomatsakis
force-pushed
the
issue-19730-perfect-forwarding
branch
4 times, most recently
from
30dc69e to
2e0cd29
Compare
|
Giving high priority because this is directly blocking progress on both unboxed closures and associated types. |
…gions and/or cause incorrect results.
have such a silly over-engineered interface.
impl for fn pointer types including bound regions, unfortunately.
issue-14039 test in a more natural way. Previously the "type we will cast to" was hidden unless it was an integer.
…ions bound in the impl before searching for `Copy` implements for all fields, leading to problems in the "copyability check". Basically the copyability check would wind up looking for an impl of `for<'tcx> Foo<&'tcx T>`. The impl that exists however is `impl<T> Copy for Foo<T>` and the current rules do not consider that a match (something I would like to revise in a later PR).
…han "skolemize" -- strictly speaking, this is not skolemization, because it is not discharging quantifiers. Also, the trait selection code will still be doing true skolemization, so it would be a confusing overlap of names.
…than having FnSig carry an implicit binding level. This means that we be more typesafe in general, since things that instantiate bound regions can drop the Binder to reflect that.
…ked code is in `higher_ranked.rs`.
read (`//!` is intrusive) and annoying to edit (must maintain a prefix on every line). Since the only purpose of a `doc.rs` file is to have a bunch of text, using `/*!` and `*/` without indentations seems appropriate.
lifetimes. This currently causes an ICE; it should (ideally) work, but failing that at least give a structured error. For the purposes of this PR, though, workaround is fine.
though it includes a `fn()`. This is really a more general problem but I wanted to ensures that `bytes` in particular remains working due to rust-lang#12677.
not dug deeply into what is going on here, although the errors ARE somewhat surprising.
nikomatsakis
force-pushed
the
issue-19730-perfect-forwarding
branch
from
2e0cd29 to
ebf1e4f
Compare
bors
added a commit
that referenced
this pull request
Dec 19, 2014
…ng, r=pnkfelix Rewrite how the HRTB algorithm matches impls against obligations. Instead of impls providing higher-ranked trait-references, impls now once again only have early-bound regions. The skolemization checks are thus moved out into trait matching itself. This allows to implement "perfect forwarding" impls like those described in #19730. This PR builds on a previous PR that was already reviewed by @pnkfelix. r? @pnkfelix Fixes #19730
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Rewrite how the HRTB algorithm matches impls against obligations. Instead of impls providing higher-ranked trait-references, impls now once again only have early-bound regions. The skolemization checks are thus moved out into trait matching itself. This allows to implement "perfect forwarding" impls like those described in #19730. This PR builds on a previous PR that was already reviewed by @pnkfelix.
r? @pnkfelix
Fixes #19730