Implement the new-style trait solver #56384

scalexm · 2018-11-30T18:52:04Z

Final PR of what I believe to be a minimally working implementation of the new-style trait solver.

The new trait solver can be used by providing the -Z chalk command line flag. It is currently used everywhere in rustc_typeck, and for everything relying on rustc::infer::canonical::query_response::enter_canonical_trait_query.

The trait solver is invoked in rustc by using the evaluate_goal canonical query. This is not optimal because each call to evaluate_goal creates a new chalk_engine::Forest, hence rustc cannot use answers to intermediate goals produced by the root goal. We'll need to change that but I guess that's ok for now.

Some next steps, I think, are:

handle region constraints: region constraints are computed but are completely ignored for now, I think we may need additional support from chalk_engine (as a side effect, types or trait references with outlive requirements cannot be proved well-formed)
deactivate eager normalization in the presence of -Z chalk in order to leverage the lazy normalization strategy of the new-style trait solver
add the remaining built-in impls (only Sized is supported currently)
transition the compiler to using generic goals instead of predicates that still refer to named type parameters etc

I added a few very simple tests to check that the new solver has the right behavior, they won't be needed anymore once it is mature enough. Additionally it shows off that we get implied bounds for free.

r? @nikomatsakis

bors · 2018-11-30T22:03:53Z

☔ The latest upstream changes (presumably #56381) made this pull request unmergeable. Please resolve the merge conflicts.

mark-i-m · 2018-12-01T21:54:57Z

The title of this PR is so epic

bors · 2018-12-08T06:17:17Z

☔ The latest upstream changes (presumably #56578) made this pull request unmergeable. Please resolve the merge conflicts.

nikomatsakis

Sorry for being so slow @scalexm -- this looks great, but I do have a few concerns:

We should decide how to handle subtyping before landing; a FIXME may suffice here
I'm a bit concerned about adding the def-id to param-env, it seems like this could affect queries in unintended ways, leading to less re-use than we would otherwise get

(Maybe we can only set the def-id if -Zchalk is used for now?)

nikomatsakis · 2018-12-17T18:42:11Z

src/librustc_traits/chalk_context/program_clauses.rs

+        ty::Infer(..) => {
+            // Everybody can find at least two types to unify against:
+            // general ty vars, int vars and float vars.
+            push_builtin_impl(tcx.types.i32, &[]);


Hmm, this worries me. I feel like there could be cases where chalk is able to try out all 4 of those possibilities and conclude that something is true since because it doesn't know of the other Sized impls. But I see the challenge here, of course, in that to generate the "full set of sized impls" would in some sense require iterating over all the types in the system.

(This problem, though, doesn't seem entirely unique to sized)

I wonder if we should think about tweaking how the engine walks around this case. It's probably ok to land the code as is for now, but we likely want to add some sort of FIXME issue here to revisit this question specifically.

This is definitely a hack indeed. We need to add a FIXME yes.

Also I believe this problem is indeed not unique to Sized, but at least it would only show for traits that have built-in impls generated by the compiler. Other traits would be fine since we would be able to enumerate all the (user-written) impls.

src/librustc_traits/chalk_context/program_clauses.rs

src/librustc_traits/chalk_context/mod.rs

src/test/compile-fail/chalkify/chalk_initial_program.rs

src/test/ui/chalkify/type_inference.rs

nikomatsakis · 2018-12-17T20:20:33Z

src/test/run-pass/chalkify/type_implied_bound.rs

+    // `Set<T>` is an input type of `take_a_set`, hence we know that
+    // `T` must implement `Hash`, and we know in turn that `T` must
+    // implement `Eq`.
+    only_eq::<T>()


nikomatsakis · 2018-12-19T17:05:43Z

Current status to my knowledge:

Waiting on @scalexm to rebase over the new version of chalk, at least, and perhaps to answer a few more of the questions above.

nikomatsakis · 2018-12-27T14:50:05Z

@bors r+

I think it's a good idea to land this so we can iterate in tree

bors · 2018-12-27T14:50:06Z

📌 Commit 9dab10dcb7a506351ea81af9751a40dac63309ae has been approved by nikomatsakis

bors · 2018-12-27T17:16:54Z

🔒 Merge conflict

This pull request and the master branch diverged in a way that cannot be automatically merged. Please rebase on top of the latest master branch, and let the reviewer approve again.

How do I rebase?

Assuming self is your fork and upstream is this repository, you can resolve the conflict following these steps:

git checkout chalk (switch to your branch)
git fetch upstream master (retrieve the latest master)
git rebase upstream/master -p (rebase on top of it)
Follow the on-screen instruction to resolve conflicts (check git status if you got lost).
git push self chalk --force-with-lease (update this PR)

You may also read Git Rebasing to Resolve Conflicts by Drew Blessing for a short tutorial.

Please avoid the "Resolve conflicts" button on GitHub. It uses git merge instead of git rebase which makes the PR commit history more difficult to read.

Sometimes step 4 will complete without asking for resolution. This is usually due to difference between how Cargo.lock conflict is handled during merge and rebase. This is normal, and you should still perform step 5 to update this PR.

Error message

warning: Cannot merge binary files: Cargo.lock (HEAD vs. heads/homu-tmp)
Auto-merging src/test/ui/issues/issue-44415.stderr
Auto-merging src/test/ui/consts/const-size_of-cycle.stderr
Auto-merging src/librustc_typeck/check/compare_method.rs
Auto-merging src/librustc_traits/type_op.rs
Auto-merging src/librustc_traits/lowering/mod.rs
Auto-merging src/librustc_traits/lowering/environment.rs
Auto-merging src/librustc_traits/lib.rs
Auto-merging src/librustc_traits/dropck_outlives.rs
Auto-merging src/librustc_traits/chalk_context/program_clauses.rs
Auto-merging src/librustc_traits/chalk_context/mod.rs
Auto-merging src/librustc/ty/structural_impls.rs
Auto-merging src/librustc/ty/query/plumbing.rs
Auto-merging src/librustc/ty/query/mod.rs
Auto-merging src/librustc/ty/query/config.rs
Auto-merging src/librustc/ty/mod.rs
Auto-merging src/librustc/ty/fold.rs
Auto-merging src/librustc/ty/context.rs
Auto-merging src/librustc/traits/structural_impls.rs
Auto-merging src/librustc/traits/select.rs
Auto-merging src/librustc/traits/mod.rs
Auto-merging src/librustc/traits/fulfill.rs
Auto-merging src/librustc/traits/error_reporting.rs
Auto-merging src/librustc/traits/engine.rs
Auto-merging src/librustc/traits/auto_trait.rs
Auto-merging src/librustc/infer/mod.rs
Auto-merging src/librustc/infer/canonical/query_response.rs
Auto-merging src/librustc/infer/canonical/mod.rs
Auto-merging src/librustc/infer/canonical/canonicalizer.rs
Auto-merging src/librustc/ich/impls_ty.rs
Auto-merging src/librustc/dep_graph/dep_node.rs
Auto-merging Cargo.lock
CONFLICT (content): Merge conflict in Cargo.lock
Automatic merge failed; fix conflicts and then commit the result.

Fixes rust-lang#55096.

Fixes rust-lang#55097.

The sub-visits were incorrectly combined with an `&&` instead of an `||`.

Multiple references to the same `BoundTy` were not using the same result.

scalexm · 2018-12-27T18:22:58Z

@bors r=nikomatsakis

bors · 2018-12-27T18:22:59Z

📌 Commit 993d213 has been approved by nikomatsakis

bors · 2018-12-27T19:48:03Z

⌛ Testing commit 993d213 with merge fb86d60...

@nikomatsakis

Implement the new-style trait solver Final PR of what I believe to be a minimally working implementation of the new-style trait solver. The new trait solver can be used by providing the `-Z chalk` command line flag. It is currently used everywhere in `rustc_typeck`, and for everything relying on `rustc::infer::canonical::query_response::enter_canonical_trait_query`. The trait solver is invoked in rustc by using the `evaluate_goal` canonical query. This is not optimal because each call to `evaluate_goal` creates a new `chalk_engine::Forest`, hence rustc cannot use answers to intermediate goals produced by the root goal. We'll need to change that but I guess that's ok for now. Some next steps, I think, are: * handle region constraints: region constraints are computed but are completely ignored for now, I think we may need additional support from `chalk_engine` (as a side effect, types or trait references with outlive requirements cannot be proved well-formed) * deactivate eager normalization in the presence of `-Z chalk` in order to leverage the lazy normalization strategy of the new-style trait solver * add the remaining built-in impls (only `Sized` is supported currently) * transition the compiler to using generic goals instead of predicates that still refer to named type parameters etc I added a few very simple tests to check that the new solver has the right behavior, they won't be needed anymore once it is mature enough. Additionally it shows off that we get [implied bounds](#44491) for free. r? @nikomatsakis

bors · 2018-12-27T22:27:21Z

☀️ Test successful - status-appveyor, status-travis
Approved by: nikomatsakis
Pushing fb86d60 to master...

rust-highfive assigned nikomatsakis Nov 30, 2018

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Nov 30, 2018

This comment has been minimized.

Sign in to view

scalexm force-pushed the chalk branch from 1cbf37f to eb4fa48 Compare November 30, 2018 21:34

scalexm force-pushed the chalk branch from eb4fa48 to ecc8604 Compare November 30, 2018 23:15

scalexm force-pushed the chalk branch from ecc8604 to f1b0eb4 Compare December 8, 2018 18:45

nikomatsakis requested changes Dec 17, 2018

View reviewed changes

nikomatsakis added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Dec 19, 2018

scalexm force-pushed the chalk branch from f1b0eb4 to 3926994 Compare December 20, 2018 18:56

This comment has been minimized.

Sign in to view

scalexm force-pushed the chalk branch from 3926994 to 9dab10d Compare December 20, 2018 20:05

nikomatsakis approved these changes Dec 27, 2018

View reviewed changes

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Dec 27, 2018

bors added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Dec 27, 2018

scalexm added 6 commits December 27, 2018 19:21

Implement is_coinductive

9b87f59

Fixes rust-lang#55096.

Implement "lifetime juggling" methods from chalk integration trait

69007bd

Fixes rust-lang#55097.

Add a def-id in ty::ParamEnv

50f8ae3

Fix visit_ex_clause_with

37ef0c3

The sub-visits were incorrectly combined with an `&&` instead of an `||`.

Correctly lift empty lists of clauses and goals

8cadd59

Return an instantiated environment instead of a generic one

ec73624

scalexm added 13 commits December 27, 2018 19:21

Add utilities for creating generic types

7b8da79

Rename functions giving WF clauses for builtin types

6f488b9

Add builtin impls for Sized in chalk_context::program_clauses

dfcad82

Fix Sized requirement for empty tuples in program_clauses

e5992e7

Fix replacement of escaping bound types

c0d7803

Multiple references to the same `BoundTy` were not using the same result.

Fix binding levels in implied bounds rules

ea4187a

Integrate chalk engine

ba6314a

Handle sub-typing in chalk-engine

da9467d

Use TraitEngine in enter_canonical_trait_query

26e0ea8

Add wf_clause_for_fn_def

8e2bdaa

Fix tcx.environment for inherent impls

3790f08

Add tests

81d6f9c

Set a def_id in ParamEnv only with -Z chalk

993d213

scalexm force-pushed the chalk branch from 9dab10d to 993d213 Compare December 27, 2018 18:21

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Dec 27, 2018

bors merged commit 993d213 into rust-lang:master Dec 27, 2018

bors mentioned this pull request Dec 27, 2018

NLL: User type annotations refactor, associated constant patterns and ref bindings. #55937

Merged

o01eg mentioned this pull request Dec 29, 2018

Rust ICEs with chalk enabled on println! macro #57193

Closed

scalexm deleted the chalk branch February 4, 2019 14:30

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Implement the new-style trait solver #56384

Implement the new-style trait solver #56384

scalexm commented Nov 30, 2018 •

edited

This comment has been minimized.

bors commented Nov 30, 2018

mark-i-m commented Dec 1, 2018

bors commented Dec 8, 2018

nikomatsakis left a comment

nikomatsakis Dec 17, 2018

scalexm Dec 17, 2018

scalexm Dec 17, 2018 •

edited

nikomatsakis Dec 17, 2018

nikomatsakis commented Dec 19, 2018

This comment has been minimized.

nikomatsakis commented Dec 27, 2018

bors commented Dec 27, 2018

bors commented Dec 27, 2018

scalexm commented Dec 27, 2018

bors commented Dec 27, 2018

bors commented Dec 27, 2018

bors commented Dec 27, 2018

Navigation Menu

Implement the new-style trait solver #56384

Implement the new-style trait solver #56384

Conversation

scalexm commented Nov 30, 2018 • edited

This comment has been minimized.

bors commented Nov 30, 2018

mark-i-m commented Dec 1, 2018

bors commented Dec 8, 2018

nikomatsakis left a comment

Choose a reason for hiding this comment

nikomatsakis Dec 17, 2018

Choose a reason for hiding this comment

scalexm Dec 17, 2018

Choose a reason for hiding this comment

scalexm Dec 17, 2018 • edited

Choose a reason for hiding this comment

nikomatsakis Dec 17, 2018

Choose a reason for hiding this comment

nikomatsakis commented Dec 19, 2018

This comment has been minimized.

nikomatsakis commented Dec 27, 2018

bors commented Dec 27, 2018

bors commented Dec 27, 2018

scalexm commented Dec 27, 2018

bors commented Dec 27, 2018

bors commented Dec 27, 2018

bors commented Dec 27, 2018

scalexm commented Nov 30, 2018 •

edited

scalexm Dec 17, 2018 •

edited