Tracking issue for RFC 2632 `?const` trait bound opt-out #67794

ecstatic-morse · 2020-01-02T00:38:25Z

This is the tracking issue for the ?const opt-out for trait bounds described in rust-lang/rfcs#2632. The primary tracking issue for that RFC is #67792. The feature gate for it is const_trait_bound_opt_out.

This RFC has not yet been accepted. It is being implemented on a provisional basis to evaluate the potential fallout.

cc #57563

Implementation steps:

Implement (Implement ?const opt-out for trait bounds #68140)
Forbid ?const outside const fn or impl const Trait

The text was updated successfully, but these errors were encountered:

@oli-obk

Parse the syntax described in RFC 2632 This adds support for both `impl const Trait for Ty` and `?const Trait` bound syntax from rust-lang/rfcs#2632 to the parser. For now, both modifiers end up in a newly-added `constness` field on `ast::TraitRef`, although this may change once the implementation is fleshed out. I was planning on using `delay_span_bug` when this syntax is encountered during lowering, but I can't write `should-ice` UI tests. I emit a normal error instead, which causes duplicates when the feature gate is not enabled (see the `.stderr` files for the feature gate tests). Not sure what the desired approach is; Maybe just do nothing when the syntax is encountered with the feature gate is enabled? @oli-obk I went with `const_trait_impl` and `const_trait_bound_opt_out` for the names of these features. Are these to your liking? cc rust-lang#67792 rust-lang#67794 r? @Centril

@oli-obk

Parse the syntax described in RFC 2632 This adds support for both `impl const Trait for Ty` and `?const Trait` bound syntax from rust-lang/rfcs#2632 to the parser. For now, both modifiers end up in a newly-added `constness` field on `ast::TraitRef`, although this may change once the implementation is fleshed out. I was planning on using `delay_span_bug` when this syntax is encountered during lowering, but I can't write `should-ice` UI tests. I emit a normal error instead, which causes duplicates when the feature gate is not enabled (see the `.stderr` files for the feature gate tests). Not sure what the desired approach is; Maybe just do nothing when the syntax is encountered with the feature gate is enabled? @oli-obk I went with `const_trait_impl` and `const_trait_bound_opt_out` for the names of these features. Are these to your liking? cc rust-lang#67792 rust-lang#67794 r? @Centril

@oli-obk

Parse the syntax described in RFC 2632 This adds support for both `impl const Trait for Ty` and `?const Trait` bound syntax from rust-lang/rfcs#2632 to the parser. For now, both modifiers end up in a newly-added `constness` field on `ast::TraitRef`, although this may change once the implementation is fleshed out. I was planning on using `delay_span_bug` when this syntax is encountered during lowering, but I can't write `should-ice` UI tests. I emit a normal error instead, which causes duplicates when the feature gate is not enabled (see the `.stderr` files for the feature gate tests). Not sure what the desired approach is; Maybe just do nothing when the syntax is encountered with the feature gate is enabled? @oli-obk I went with `const_trait_impl` and `const_trait_bound_opt_out` for the names of these features. Are these to your liking? cc rust-lang#67792 rust-lang#67794 r? @Centril

@oli-obk

Support `?const` opt-out for trait bounds For now, such bounds are treated exactly the same as unprefixed ones in all contexts. [RFC 2632](rust-lang/rfcs#2632) does not specify whether such bounds are forbidden outside of `const` contexts, so they are allowed at the moment. cc #67794 r? @oli-obk

@oli-obk

…i-obk Implement `?const` opt-out for trait bounds For now, such bounds are treated exactly the same as unprefixed ones in all contexts. [RFC 2632](rust-lang/rfcs#2632) does not specify whether such bounds are forbidden outside of `const` contexts, so they are allowed at the moment. Prior to this PR, the constness of a trait bound/impl was stored in `TraitRef`. Now, the constness of an `impl` is stored in `ast::ItemKind::Impl` and the constness of a bound in `ast::TraitBoundModifer`. Additionally, constness of trait bounds is now stored in an additional field of `ty::Predicate::Trait`, and the combination of the constness of the item along with any `TraitBoundModifier` determines the constness of the bound in accordance with the RFC. Encoding the constness of impls at the `ty` level is left for a later PR. After a discussion in \#wg-grammar on Discord, it was decided that the grammar should not encode the mutual exclusivity of trait bound modifiers. The grammar for trait bound modifiers remains `[?const] [?]`. To encode this, I add a dummy variant to `ast::TraitBoundModifier` that is used when the syntax `?const ?` appears. This variant causes an error in AST validation and disappears during HIR lowering. cc #67794 r? @oli-obk

@oli-obk

…t-out, r=oli-obk Implement `?const` opt-out for trait bounds For now, such bounds are treated exactly the same as unprefixed ones in all contexts. [RFC 2632](rust-lang/rfcs#2632) does not specify whether such bounds are forbidden outside of `const` contexts, so they are allowed at the moment. Prior to this PR, the constness of a trait bound/impl was stored in `TraitRef`. Now, the constness of an `impl` is stored in `ast::ItemKind::Impl` and the constness of a bound in `ast::TraitBoundModifer`. Additionally, constness of trait bounds is now stored in an additional field of `ty::Predicate::Trait`, and the combination of the constness of the item along with any `TraitBoundModifier` determines the constness of the bound in accordance with the RFC. Encoding the constness of impls at the `ty` level is left for a later PR. After a discussion in \#wg-grammar on Discord, it was decided that the grammar should not encode the mutual exclusivity of trait bound modifiers. The grammar for trait bound modifiers remains `[?const] [?]`. To encode this, I add a dummy variant to `ast::TraitBoundModifier` that is used when the syntax `?const ?` appears. This variant causes an error in AST validation and disappears during HIR lowering. cc rust-lang#67794 r? @oli-obk

clarfonthey · 2021-09-01T00:44:20Z

From what I understand, this was recently changed to ~const but none of the tilde-const PRs link back to this one, and the issue name/description should be updated?

clarfonthey · 2021-09-01T00:45:30Z

cc @oli-obk @fee1-dead since from what I understand you two know the most about this change

fee1-dead · 2021-09-01T02:52:31Z

The opt-out is no longer needed. The previous scheme was:

const fn a<T: /* inferred const bound */ Foo + /* opt-out normal trait bound */ ?const Bar>

And the current scheme is now:

const fn a<T: /* normal trait bound */ Foo + /* const bound if called in a const context */ ~const Bar>

This issue should be closed, but it can be reopened if somehow we decide to use ?const opt-out again.

jplatte · 2021-09-01T08:48:39Z

So where is ~const tracked then? #67792?

oli-obk · 2021-09-01T09:00:20Z

yea, const_trait_impl requires the opt in syntax, otherwise you can't write anything but trivial impls

clarfonthey · 2021-09-01T16:07:59Z

So... is there no dedicated tracking issue for ~const, then?

Mostly because I don't see any easy place someone can look to know what the new syntax means.

oli-obk · 2021-09-01T22:56:29Z

I updated the const_trait_impl tracking issue to link to the latest RFC text

ecstatic-morse mentioned this issue Jan 2, 2020

Deprecated tracking issue for closed RFC 2632, impl const Trait for Ty and ~const (tilde const) syntax (const_trait_impl) #67792

Open

13 tasks

ecstatic-morse added A-const-fn C-tracking-issue Category: An issue tracking the progress of sth. like the implementation of an RFC labels Jan 2, 2020

Centril added the T-lang Relevant to the language team, which will review and decide on the PR/issue. label Jan 2, 2020

ecstatic-morse mentioned this issue Jan 3, 2020

Parse the syntax described in RFC 2632 #67820

Merged

ecstatic-morse changed the title ~~Tracking issue for RFC 2362, ?const trait bound opt-out~~ Tracking issue for RFC 2632 ?const trait bound opt-out Jan 3, 2020

oli-obk added the F-const_trait_bound_opt_out label Jan 5, 2020

ecstatic-morse mentioned this issue Jan 12, 2020

Implement ?const opt-out for trait bounds #68140

Merged

jonas-schievink added the A-const-eval Area: Constant evaluation, covers all const contexts (static, const fn, ...) label Mar 28, 2020

jplatte mentioned this issue May 2, 2020

Tracking issue for RFC 2342, "Allow if and match in constants" #49146

Closed

4 tasks

jonas-schievink mentioned this issue Jun 7, 2020

Array range indexes don't work in const eval #73083

Closed

mcy mentioned this issue Aug 14, 2020

Building Tock with stable rustc tock/tock#1654

Open

29 tasks

Lokathor mentioned this issue Aug 23, 2020

Meta tracking issue for const fn #57563

Open

25 tasks

josephlr mentioned this issue Sep 29, 2020

Fix build error on latest nightly rust-osdev/x86_64#186

Merged

RalfJung mentioned this issue Apr 18, 2021

further split up const_fn feature flag #84310

Merged

oli-obk closed this as completed Sep 1, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Tracking issue for RFC 2632 `?const` trait bound opt-out #67794

Tracking issue for RFC 2632 `?const` trait bound opt-out #67794

ecstatic-morse commented Jan 2, 2020 •

edited by oli-obk

Loading

clarfonthey commented Sep 1, 2021

clarfonthey commented Sep 1, 2021

fee1-dead commented Sep 1, 2021

jplatte commented Sep 1, 2021

oli-obk commented Sep 1, 2021

clarfonthey commented Sep 1, 2021

oli-obk commented Sep 1, 2021

Tracking issue for RFC 2632 ?const trait bound opt-out #67794

Tracking issue for RFC 2632 ?const trait bound opt-out #67794

Comments

ecstatic-morse commented Jan 2, 2020 • edited by oli-obk Loading

clarfonthey commented Sep 1, 2021

clarfonthey commented Sep 1, 2021

fee1-dead commented Sep 1, 2021

jplatte commented Sep 1, 2021

oli-obk commented Sep 1, 2021

clarfonthey commented Sep 1, 2021

oli-obk commented Sep 1, 2021

Tracking issue for RFC 2632 `?const` trait bound opt-out #67794

Tracking issue for RFC 2632 `?const` trait bound opt-out #67794

ecstatic-morse commented Jan 2, 2020 •

edited by oli-obk

Loading