[7/n] rustc: desugar UFCS in HIR and don't use DefMap for associated resolutions. #37676

eddyb · 2016-11-09T21:50:50Z

This is part of a series (prev | next) of patches designed to rework rustc into an out-of-order on-demand pipeline model for both better feature support (e.g. MIR-based early constant evaluation) and incremental execution of compiler passes (e.g. type-checking), with beneficial consequences to IDE support as well.
If any motivation is unclear, please ask for additional PR description clarifications or code comments.

Previously, a path like T::Assoc::method, while equivalent to <<T>::Assoc>::method, wasn't desugared in any way at the HIR level and everything inspecting it had to either deal with knowing only T (before typeck) or knowing only the definition of method (after typeck).
Such a path also had only one NodeId and associated resolution during typeck modified DefMap, in a way that would be hard for incremental recompilation to track, and inconvenient for partial type conversions from HIR to Ty, which are required to break faux-cycles in on-demand type collection.

The desugarings performed by this PR are as follows:

use a::{b,c}; is flattened to use a as _; use a::b; use a::c;
- as resolution is complete, use a as _; doesn't do anything, except get checked for stability
Vec::new (an expression) becomes Vec<..>::new<..>, to distinguish it from <Vec>::new<..>
- the "infer all parameters" <..> form is internal and not even pretty-printed
- used when there are no type parameters at all, in an expression or pattern path segment
T::A::B becomes <<T>::A>::B in a type, and <<T<..>>::A<..>>::B<..> in an expression/pattern
- one additional hir::Ty node is created for each prefix, starting with the fully-resolved type (T) and extending it with each segment (e.g. <T>::A)
fully-resolved paths contain their Def in HIR, getting rid of the DefMap and absolving incremental recompilation of needing to manually look up nodes to handle that side information

Not keeping the DefMap around meant that associated resolutions had to be stored somewhere else:

expressions and patterns use a new NodeId -> Def map in ty::Tables
- compatible with the future per-body (constant / fn / closure) Tables
types are accessible via Ty and the usual per-item generics / predicates / type
- rustdoc and save-analysis are the only situations which insist on mapping syntactical types to semantical ones, or at least understand the resolution of associated types, therefore the type conversion cache, i.e. a NodeId -> Ty map, is exposed by typeck for this purpose
- stability had to be split into a pass that runs on HIR and checks the results of name resolution, and impromptu checks triggered by typeck for associated paths, methods, fields, etc.
- privacy using semantic types results in accurate reachability for impl Trait, which fixes Using impl trait across crates results in linker error #35870, and thorough introspection of associated types, which may allow relaxing private-in-public checking on bounds, while keeping the intended ban on projections with private type parameters

cc @petrochenkov

rust-highfive · 2016-11-09T21:51:02Z

r? @Aatch

(rust_highfive has picked a reviewer for you, use r? to override)

bors · 2016-11-09T23:15:30Z

☔ The latest upstream changes (presumably #37670) made this pull request unmergeable. Please resolve the merge conflicts.

eddyb · 2016-11-09T23:15:31Z

r? @nikomatsakis

@Manishearth

[6/n] rustc: transition HIR function bodies from Block to Expr. _This is part of a series ([prev](rust-lang#37408) | [next](rust-lang#37676)) of patches designed to rework rustc into an out-of-order on-demand pipeline model for both better feature support (e.g. [MIR-based](https://github.com/solson/miri) early constant evaluation) and incremental execution of compiler passes (e.g. type-checking), with beneficial consequences to IDE support as well. If any motivation is unclear, please ask for additional PR description clarifications or code comments._ <hr> The main change here is that functions and closures both use `Expr` instead of `Block` for their bodies. For closures this actually allows a honest representation of brace-less closure bodies, e.g. `|x| x + 1` is now distinguishable from `|x| { x + 1 }`, therefore this PR is `[syntax-breaking]` (cc @Manishearth). Using `Expr` allows more logic to be shared between constant bodies and function bodies, with some small such changes already part of this PR, and eventually easing rust-lang#35078 and per-body type tables. Incidentally, there used to be some corners cut here and there and as such I had to (re)write divergence tracking for type-checking so that it is capable of understanding basic structured control-flow: ``` rust fn a(x: bool) -> i32 { // match also works (as long as all arms diverge) if x { panic!("true") } else { return 1; } 0 // "unreachable expression" after this PR } ``` And since liveness' "not all control paths return a value" moved to type-checking we can have nice things: ``` rust // before & after: fn b() -> i32 { 0; } // help: consider removing this semicolon // only after this PR fn c() -> i32 { { 0; } } // help: consider removing this semicolon fn d() { let x: i32 = { 0; }; } // help: consider removing this semicolon fn e() { f({ 0; }); } // help: consider removing this semicolon ```

bors · 2016-11-10T05:28:05Z

☔ The latest upstream changes (presumably #37678) made this pull request unmergeable. Please resolve the merge conflicts.

eddyb · 2016-11-10T10:45:51Z

Crater report shows no user-facing regressions.

bors · 2016-11-10T13:11:11Z

☔ The latest upstream changes (presumably #37463) made this pull request unmergeable. Please resolve the merge conflicts.

petrochenkov · 2016-11-10T20:25:35Z

(I'll read this on Friday or Saturday.)

[8/n] rustc: clean up lookup_item_type and remove TypeScheme. _This is part of a series ([prev](rust-lang#37676) | [next]()) of patches designed to rework rustc into an out-of-order on-demand pipeline model for both better feature support (e.g. [MIR-based](https://github.com/solson/miri) early constant evaluation) and incremental execution of compiler passes (e.g. type-checking), with beneficial consequences to IDE support as well. If any motivation is unclear, please ask for additional PR description clarifications or code comments._ <hr> * `tcx.tcache` -> `tcx.item_types` * `TypeScheme` (grouping `Ty` and `ty::Generics`) is removed * `tcx.item_types` entries no longer duplicated in `tcx.tables.node_types` * `tcx.lookup_item_type(def_id).ty` -> `tcx.item_type(def_id)` * `tcx.lookup_item_type(def_id).generics` -> `tcx.item_generics(def_id)` * `tcx.lookup_generics(def_id)` -> `tcx.item_generics(def_id)` * `tcx.lookup_{super_,}predicates(def_id)` -> `tcx.item_{super_,}predicates(def_id)`

eddyb · 2016-11-28T03:24:59Z

@arielb1 Turns out your changes to the handling of closures means impl Fn just works now, heh.

eddyb · 2016-11-28T03:43:48Z

@bors r=nikomatsakis

bors · 2016-11-28T03:43:49Z

📌 Commit 372c6df has been approved by nikomatsakis

bors · 2016-11-28T05:32:57Z

⌛ Testing commit 372c6df with merge 0b399e5...

@petrochenkov

[7/n] rustc: desugar UFCS in HIR and don't use DefMap for associated resolutions. _This is part of a series ([prev](#37412) | [next](#37688)) of patches designed to rework rustc into an out-of-order on-demand pipeline model for both better feature support (e.g. [MIR-based](https://github.com/solson/miri) early constant evaluation) and incremental execution of compiler passes (e.g. type-checking), with beneficial consequences to IDE support as well. If any motivation is unclear, please ask for additional PR description clarifications or code comments._ <hr> Previously, a path like `T::Assoc::method`, while equivalent to `<<T>::Assoc>::method`, wasn't desugared in any way at the HIR level and everything inspecting it had to either deal with knowing only `T` (before typeck) or knowing only the definition of `method` (after typeck). Such a path also had only one `NodeId` and associated resolution during typeck modified `DefMap`, in a way that would be hard for incremental recompilation to track, and inconvenient for partial type conversions from HIR to `Ty`, which are required to break faux-cycles in on-demand type collection. The desugarings performed by this PR are as follows: * `use a::{b,c};` is flattened to `use a as _; use a::b; use a::c;` * as resolution is complete, `use a as _;` doesn't do anything, except get checked for stability * `Vec::new` (an expression) becomes `Vec<..>::new<..>`, to distinguish it from `<Vec>::new<..>` * the "infer all parameters" `<..>` form is internal and not even pretty-printed * used when there are no type parameters at all, in an expression or pattern path segment * `T::A::B` becomes `<<T>::A>::B` in a type, and `<<T<..>>::A<..>>::B<..>` in an expression/pattern * one additional `hir::Ty` node is created for each prefix, starting with the fully-resolved type (`T`) and extending it with each segment (e.g. `<T>::A`) * fully-resolved paths contain their `Def` in HIR, getting rid of the `DefMap` and absolving incremental recompilation of needing to manually look up nodes to handle that side information Not keeping the `DefMap` around meant that associated resolutions had to be stored somewhere else: * expressions and patterns use a new `NodeId -> Def` map in `ty::Tables` * compatible with the future per-body (constant / `fn` / closure) `Tables` * types are accessible via `Ty` and the usual per-item generics / predicates / type * `rustdoc` and `save-analysis` are the only situations which insist on mapping syntactical types to semantical ones, or at least understand the resolution of associated types, therefore the type conversion cache, i.e. a `NodeId -> Ty` map, is exposed by typeck for this purpose * stability had to be split into a pass that runs on HIR and checks the results of name resolution, and impromptu checks triggered by `typeck` for associated paths, methods, fields, etc. * privacy using semantic types results in accurate reachability for `impl Trait`, which fixes #35870, and thorough introspection of associated types, which may allow relaxing private-in-public checking on bounds, while keeping the intended ban on projections with private type parameters cc @petrochenkov

bors · 2016-11-28T13:03:41Z

This code was introduced in rust-lang#27565 to mark types in paths alive. It is now unnecessary since rust-lang#37676.

[8/n] rustc: clean up lookup_item_type and remove TypeScheme. _This is part of a series ([prev](rust-lang/rust#37676) | [next]()) of patches designed to rework rustc into an out-of-order on-demand pipeline model for both better feature support (e.g. [MIR-based](https://github.com/solson/miri) early constant evaluation) and incremental execution of compiler passes (e.g. type-checking), with beneficial consequences to IDE support as well. If any motivation is unclear, please ask for additional PR description clarifications or code comments._ <hr> * `tcx.tcache` -> `tcx.item_types` * `TypeScheme` (grouping `Ty` and `ty::Generics`) is removed * `tcx.item_types` entries no longer duplicated in `tcx.tables.node_types` * `tcx.lookup_item_type(def_id).ty` -> `tcx.item_type(def_id)` * `tcx.lookup_item_type(def_id).generics` -> `tcx.item_generics(def_id)` * `tcx.lookup_generics(def_id)` -> `tcx.item_generics(def_id)` * `tcx.lookup_{super_,}predicates(def_id)` -> `tcx.item_{super_,}predicates(def_id)`

@alexcrichton

Fix checking for missing stability annotations This was a regression from #37676 causing "unmarked API" ICEs like #43027. r? @alexcrichton

rust-highfive assigned Aatch Nov 9, 2016

This was referenced Nov 9, 2016

[6/n] rustc: transition HIR function bodies from Block to Expr. #37412

Merged

Tracking issue for private_in_public compatibility lint. #34537

Closed

rust-highfive assigned nikomatsakis and unassigned Aatch Nov 9, 2016

eddyb mentioned this pull request Nov 9, 2016

Warn unused type aliases #37631

Closed

eddyb force-pushed the lazy-7 branch 2 times, most recently from 869c0b3 to 0bf9c99 Compare November 10, 2016 01:14

eddyb force-pushed the lazy-7 branch 2 times, most recently from 16ee826 to 006c75d Compare November 10, 2016 03:01

eddyb force-pushed the lazy-7 branch from 006c75d to ff0bb52 Compare November 10, 2016 09:11

eddyb mentioned this pull request Nov 10, 2016

[8/n] rustc: clean up lookup_item_type and remove TypeScheme. #37688

Merged

eddyb force-pushed the lazy-7 branch 4 times, most recently from b2cde2a to 7449ef9 Compare November 10, 2016 17:08

alexcrichton added the T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. label Nov 10, 2016

eddyb force-pushed the lazy-7 branch from 7449ef9 to 74e72f3 Compare November 11, 2016 22:15

eddyb force-pushed the lazy-7 branch 2 times, most recently from a01d993 to 74e72f3 Compare November 12, 2016 11:16

eddyb added 5 commits November 28, 2016 04:18

rustc: rework stability to be on-demand for type-directed lookup.

9aaf26e

rustc_privacy: visit Ty instead of HIR types in EmbargoVisitor.

36d33d6

rustc: track fields in the HIR map.

fcdb4de

rustc_privacy: switch private-in-public checking to Ty.

548e681

rustc_typeck: don't record associated type resolutions.

372c6df

eddyb force-pushed the lazy-7 branch from 97bd232 to 372c6df Compare November 28, 2016 03:25

bors merged commit 372c6df into rust-lang:master Nov 28, 2016

This was referenced Nov 28, 2016

Separate function bodies from their signatures in HIR #37918

Merged

rustdoc: link to cross-crate sources directly. #37954

Merged

Refactor TraitObject to Slice<ExistentialPredicate> #37965

Merged

eddyb deleted the lazy-7 branch November 28, 2016 13:09

bors mentioned this pull request Nov 28, 2016

add hint to fix error for immutable ref in arg #37863

Merged

oli-obk mentioned this pull request Dec 1, 2016

WIP: rustup rust-lang/rust-clippy#1373

Merged

sanxiyn added a commit to sanxiyn/rust that referenced this pull request Dec 15, 2016

Remove now unnecessary code

a759098

This code was introduced in rust-lang#27565 to mark types in paths alive. It is now unnecessary since rust-lang#37676.

petrochenkov mentioned this pull request Dec 16, 2016

internal compiler error: ../src/librustc/middle/stability.rs:516: encountered unmarked API #38412

Closed

petrochenkov mentioned this pull request Jan 11, 2017

rustc panics on fn that returns associated struct tuple with Self #38919

Closed

petrochenkov mentioned this pull request Jul 16, 2017

Fix checking for missing stability annotations #43270

Merged

bors added a commit that referenced this pull request Jul 20, 2017

Auto merge of #43270 - petrochenkov:fixstab, r=alexcrichton

1edbc3d

Fix checking for missing stability annotations This was a regression from #37676 causing "unmarked API" ICEs like #43027. r? @alexcrichton

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[7/n] rustc: desugar UFCS in HIR and don't use DefMap for associated resolutions. #37676

[7/n] rustc: desugar UFCS in HIR and don't use DefMap for associated resolutions. #37676

eddyb commented Nov 9, 2016 •

edited

Loading

rust-highfive commented Nov 9, 2016

bors commented Nov 9, 2016

eddyb commented Nov 9, 2016

bors commented Nov 10, 2016

eddyb commented Nov 10, 2016

bors commented Nov 10, 2016

petrochenkov commented Nov 10, 2016

eddyb commented Nov 28, 2016

eddyb commented Nov 28, 2016

bors commented Nov 28, 2016

bors commented Nov 28, 2016

bors commented Nov 28, 2016

[7/n] rustc: desugar UFCS in HIR and don't use DefMap for associated resolutions. #37676

[7/n] rustc: desugar UFCS in HIR and don't use DefMap for associated resolutions. #37676

Conversation

eddyb commented Nov 9, 2016 • edited Loading

rust-highfive commented Nov 9, 2016

bors commented Nov 9, 2016

eddyb commented Nov 9, 2016

bors commented Nov 10, 2016

eddyb commented Nov 10, 2016

bors commented Nov 10, 2016

petrochenkov commented Nov 10, 2016

eddyb commented Nov 28, 2016

eddyb commented Nov 28, 2016

bors commented Nov 28, 2016

bors commented Nov 28, 2016

bors commented Nov 28, 2016

eddyb commented Nov 9, 2016 •

edited

Loading