/rust

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix exponential projection complexity on nested types #48296

Merged
merged 4 commits into from Feb 25, 2018
+133 −151

Conversation

Reviewers

@stjepang stjepang

@nikomatsakis nikomatsakis

Assignees

@nikomatsakis nikomatsakis

Labels
Projects
None yet
Milestone
No milestone
7 participants
@ishitatsuyuki @rust-highfive @kennytm @bors @Mark-Simulacrum @nikomatsakis @stjepang
@ishitatsuyuki
Member

ishitatsuyuki commented Feb 17, 2018 •
edited

This implements solution 1 from #38528 (comment).

The code quality was extremely poor, but it has improved during review.

Blocking issues:

  • we probably don't want a quadratic deduplication for obligations.
  • is there an alternative to deduplication?

Based on #48315.

Needs changelog. Noticable improvement on compile time is expected.

Fix #38528
Close #39684
Close #43757

@rust-highfive rust-highfive assigned nikomatsakis Feb 17, 2018

@rust-highfive

This comment has been minimized.

Sign in to view
Collaborator

rust-highfive commented Feb 17, 2018

r? @nikomatsakis

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

@rust-highfive rust-highfive added the S-waiting-on-review label Feb 17, 2018

@ishitatsuyuki ishitatsuyuki changed the title from [wip] (PoC) Fix exponential blowup on nested types to [wip] (PoC) Fix exponential projection complexity on nested types Feb 17, 2018

This was referenced Feb 17, 2018

Closed
Closed
@ishitatsuyuki

ishitatsuyuki reviewed Feb 17, 2018
View changes

Here are some self-reminders.

src/librustc/traits/project.rs Outdated
@@ -818,55 +822,74 @@ fn project_type<'cx, 'gcx, 'tcx>(
&obligation_trait_ref,
&mut candidates);
let decide_commit = |candidates: &mut ProjectionTyCandidateSet<'tcx>| {

This comment has been minimized.

Sign in to view
Copy link
@ishitatsuyuki

ishitatsuyuki Feb 17, 2018

Member

I have no way but to reorder this before the error handling. It looks non intuitive.
src/librustc/traits/project.rs Outdated
});
debug!("resulting candidate set: {:?}", candidates.vec);
if candidates.vec.len() != 1 {
candidates.ambiguous = true;

This comment has been minimized.

Sign in to view
Copy link
@ishitatsuyuki

ishitatsuyuki Feb 17, 2018

Member

This is nearly a hack.
src/librustc/traits/project.rs Outdated
}
};
match vtable {
match vtable.clone() {

This comment has been minimized.

Sign in to view
Copy link
@ishitatsuyuki

ishitatsuyuki Feb 17, 2018

Member

This copy can be avoided.

@kennytm kennytm added S-waiting-on-author and removed S-waiting-on-review labels Feb 17, 2018

@ishitatsuyuki ishitatsuyuki changed the title from [wip] (PoC) Fix exponential projection complexity on nested types to [wip] Fix exponential projection complexity on nested types Feb 18, 2018

@ishitatsuyuki

This comment has been minimized.

Sign in to view
Member

ishitatsuyuki commented Feb 18, 2018

The other part of the blowup (typeck related) is yet to be solved, but I think the code is ready for a round of review. I have removed whitespace changes and clarified the description.
@ishitatsuyuki

This comment has been minimized.

Sign in to view
Member

ishitatsuyuki commented Feb 18, 2018

@kennytm A reminder for you to change the review labels.

@kennytm kennytm added S-waiting-on-review and removed S-waiting-on-author labels Feb 18, 2018

@kennytm

This comment has been minimized.

Sign in to view
Member

kennytm commented Feb 18, 2018

@bors try

Let’s do a perf check while we’re waiting for review :)
@bors

This comment has been minimized.

Sign in to view
Contributor

bors commented Feb 18, 2018

⌛️ Trying commit 3d81698 with merge e35a327...

bors added a commit that referenced this pull request Feb 18, 2018

@bors
Auto merge of #48296 - ishitatsuyuki:exp-unblow, r=<try> 
[wip] Fix exponential projection complexity on nested types

This implements solution 1 from #38528 (comment).

The code quality is currently extremely poor, but we can improve them during review.

Based on #48315.

Needs changelog. Noticable improvement on compile time is expected.

Fix #38528 (in progress, another exponential blowup exist)
Close #39684
Close #43757
e35a327
@kennytm

This comment has been minimized.

Sign in to view
Member

kennytm commented Feb 18, 2018

@ishitatsuyuki please remove the [wip] from the PR title if it is ready for review.
@ishitatsuyuki

This comment has been minimized.

Sign in to view
Member

ishitatsuyuki commented Feb 18, 2018

Well, I'm currently tackling another part of the exponential issue (so this is WIP), but I need comments on improving the code organization.
@bors

This comment has been minimized.

Sign in to view
Contributor

bors commented Feb 18, 2018

☀️ Test successful - status-travis
State: approved= try=True

@ishitatsuyuki ishitatsuyuki changed the title from [wip] Fix exponential projection complexity on nested types to Fix exponential projection complexity on nested types Feb 18, 2018

@ishitatsuyuki

This comment has been minimized.

Sign in to view
Member

ishitatsuyuki commented Feb 18, 2018

By deduplicating the obligations the issue is completely resolved. I'm not sure if this approach is the best way though, and I copied the dumb deduplication becuase I didn't want to implement Ord everywhere. I will fix them later.
@ishitatsuyuki

This comment has been minimized.

Sign in to view
Member

ishitatsuyuki commented Feb 18, 2018

@kennytm I need another try.

@ishitatsuyuki ishitatsuyuki referenced this pull request Feb 18, 2018

Open

cache the results of type projection and normalization #20304

@Mark-Simulacrum Mark-Simulacrum added the relnotes label Feb 18, 2018

@Mark-Simulacrum

This comment has been minimized.

Sign in to view
Member

Mark-Simulacrum commented Feb 18, 2018

@bors try
@bors

This comment has been minimized.

Sign in to view
Contributor

bors commented Feb 18, 2018

⌛️ Trying commit 4a40c55 with merge 4ac2425...

bors added a commit that referenced this pull request Feb 18, 2018

@bors
Auto merge of #48296 - ishitatsuyuki:exp-unblow, r=<try> 
Fix exponential projection complexity on nested types

This implements solution 1 from #38528 (comment).

The code quality is currently extremely poor, but we can improve them during review.

Blocking issues:

- we probably don't want a quadratic deduplication for obligations.
- is there an alternative to deduplication?

Based on #48315.

Needs changelog. Noticable improvement on compile time is expected.

Fix #38528
Close #39684
Close #43757
4ac2425
@bors

This comment has been minimized.

Sign in to view
Contributor

bors commented Feb 18, 2018

☀️ Test successful - status-travis
State: approved= try=True
@ishitatsuyuki

This comment has been minimized.

Sign in to view
Member

ishitatsuyuki commented Feb 18, 2018

@Mark-Simulacrum can you queue a perf run?

@nikomatsakis nikomatsakis referenced this pull request Feb 21, 2018

Closed

Remove needless dedup from projection #48315

@nikomatsakis nikomatsakis added S-waiting-on-author and removed S-waiting-on-review labels Feb 22, 2018

@ishitatsuyuki

This comment has been minimized.

Sign in to view
Member

ishitatsuyuki commented Feb 22, 2018

Ready for review.
@ishitatsuyuki

This comment has been minimized.

Sign in to view
Member

ishitatsuyuki commented Feb 22, 2018

r? @nikomatsakis

Please change the label...

@kennytm kennytm added S-waiting-on-review and removed S-waiting-on-author labels Feb 22, 2018

@nikomatsakis

nikomatsakis approved these changes Feb 22, 2018
View changes

Looks great =)

src/librustc/traits/project.rs
} else {
Err(())
}
});

This comment has been minimized.

Sign in to view
Copy link
@nikomatsakis

nikomatsakis Feb 22, 2018

Contributor

I like the change to push errors into the candidate set. This is less confusing this way.

This comment has been minimized.

Sign in to view
Copy link
@ishitatsuyuki

ishitatsuyuki Feb 22, 2018

Member

I'm not sure if I understand. I already handle SelectionError above, and these Ok and Err is the return value to commit_if_ok.
src/librustc/traits/select.rs
// unknown. What we know is that the deduplication avoids exponential
// amount of predicates being propogated when processing deeply nested
// types.
let mut seen = FxHashSet();

This comment has been minimized.

Sign in to view
Copy link
@nikomatsakis

nikomatsakis Feb 22, 2018

Contributor

Oh, it seems like duplicates could come from many sources -- in particular, normalization might introduce the same obligation a few times.
@nikomatsakis

This comment has been minimized.

Sign in to view
Contributor

nikomatsakis commented Feb 22, 2018

@bors r+

@ishitatsuyuki -- it'd be great to add at least one of these exponential blow-up cases to the perf website so we can monitor for regressions. Perhaps @Mark-Simulacrum can walk you through it.
@bors

This comment has been minimized.

Sign in to view
Contributor

bors commented Feb 22, 2018

📌 Commit 5a2bec9 has been approved by nikomatsakis

@bors bors added S-waiting-on-bors and removed S-waiting-on-review labels Feb 22, 2018

Manishearth added a commit to Manishearth/rust that referenced this pull request Feb 24, 2018

@Manishearth
Rollup merge of rust-lang#48296 - ishitatsuyuki:exp-unblow, r=nikomat… 
…sakis

Fix exponential projection complexity on nested types

This implements solution 1 from rust-lang#38528 (comment).

The code quality is currently extremely poor, but we can improve them during review.

Blocking issues:

- we probably don't want a quadratic deduplication for obligations.
- is there an alternative to deduplication?

Based on rust-lang#48315.

Needs changelog. Noticable improvement on compile time is expected.

Fix rust-lang#38528
Close rust-lang#39684
Close rust-lang#43757
Verified
This commit was created on GitHub.com and signed with a verified signature using GitHub’s key.
GPG key ID: 4AEE18F83AFDEB23 Learn about signing commits
58af0c7

@Manishearth Manishearth referenced this pull request Feb 24, 2018

Merged

Rollup of 15 pull requests #48520

bors added a commit that referenced this pull request Feb 24, 2018

@bors
Auto merge of #48520 - Manishearth:rollup, r=Manishearth 
Rollup of 15 pull requests

- Successful merges: #47689, #48110, #48197, #48296, #48386, #48392, #48404, #48415, #48441, #48448, #48452, #48481, #48490, #48499, #48503
- Failed merges:
Loading status checks…
fe7075b

bors added a commit that referenced this pull request Feb 25, 2018

@bors
Auto merge of #48520 - Manishearth:rollup, r=Manishearth 
Rollup of 15 pull requests

- Successful merges: #47689, #48110, #48197, #48296, #48386, #48392, #48404, #48415, #48441, #48448, #48452, #48481, #48490, #48499, #48503
- Failed merges:
Loading status checks…
026339e

@bors bors merged commit 5a2bec9 into rust-lang:master Feb 25, 2018
1 check passed

1 check passed

continuous-integration/travis-ci/pr The Travis CI build passed
Details

@ishitatsuyuki ishitatsuyuki referenced this pull request Feb 26, 2018

Open

type checker takes O(~1.5^recursion_limit) time to reject simple-ish code #40353

@Antti Antti referenced this pull request Feb 26, 2018

Closed

Compiler panic while building alga #48542

@ishitatsuyuki ishitatsuyuki referenced this pull request Feb 26, 2018

Closed

Unexpected panic with cgmath #48551

@ishitatsuyuki ishitatsuyuki deleted the ishitatsuyuki:exp-unblow branch Mar 1, 2018

@ishitatsuyuki ishitatsuyuki referenced this pull request Mar 1, 2018

Merged

Add benchmark for deeply nested types #185

This was referenced Apr 1, 2018

Open
Closed
Closed
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment