Short-cut `T: Sized` trait selection for ADTs #33138

arielb1 · 2016-04-21T20:10:55Z

Basically avoids all nested obligations when checking whether an ADT is sized - this speeds up typeck by ~15%

The refactoring fixed #32963, but I also want to make Copy not object-safe (will commit that soon).

jonas-schievink · 2016-04-21T20:35:50Z

src/librustc/traits/select.rs

                span_bug!(
                    obligation.cause.span,
-                    "builtin bound for {:?} was ambig",
+                    "confiriming builtin impl for {:?} where none exists",


nit: s/confiriming/confirming/

arielb1 · 2016-04-21T23:36:42Z

I rationalized the WF behaviour for tuples - this needs a crater run.

nikomatsakis · 2016-04-22T00:58:15Z

crater run launched

nikomatsakis · 2016-04-22T08:45:31Z

Crater run failed because building this branch failed: https://tools.taskcluster.net/task-inspector/#eqYUkXetSja7miFf4nXmeQ/0

nikomatsakis · 2016-04-22T08:49:50Z

src/librustc/ty/mod.rs

+    ///       such.
+    ///     - a TyError, if a type contained itself. The representability
+    ///       check should catch this case.
+    fn calculate_sized_constraint_inner(&'tcx self, tcx: &ty::TyCtxt<'tcx>,


Total nit: I personally prefer the "main method" to be listed first, and the auxiliary helper methods to be listed below it. I find it easier to read (and indeed I was confused reading this patch for a bit, since the "flow" of code was interrupted by helper fns.) But feel free to ignore me if you would rather keep the current ordering.

I just rewrote this code about half-a-dozen times.

nikomatsakis · 2016-04-22T08:58:49Z

src/test/run-pass/issue-31299.rs

+//     - proving `Vec<M>: Front` requires
+//       - `M: Sized` <-- cycle!
+//
+// If we skip the normalization step, though, everything goes fine.


I would add to this comment to say:

"One fix would be lazy normalization. At the moment, we do still eagerly normalize, but we still skip the normalization step because, in PR #33138, we only consider the last field of Vec, which is a usize."

nikomatsakis · 2016-04-22T09:00:41Z

@arielb1 this is a beautiful patch.

I would r+ it, except that I think it violates my proposed stability policy. I think we ought to be making an effort to either issue warnings or at least issue targeted errors regarding the new changes. Moreover, this seems to be a case where we could readily do so.

I am happy to write up the "breaking change issue description" if you like.

In any case, you should fix the branch so we can do a crater run and test things.

arielb1 · 2016-04-22T09:51:35Z

@nikomatsakis

Are you talking about the tuple: Sized issue? I don't feel like introducing RFC1214 obligations again. But maybe we need such a framework in general for handling warnings?

nikomatsakis · 2016-04-22T10:27:04Z

@arielb1 I am talking about the Copy and Sized trait object-safety change but also the tuple: Sized issue, yeah. It seems like we have a specific obligation cause for "tuple-sized" requirements -- we should be able to just issue a warning in that case, no?

nikomatsakis · 2016-04-22T10:27:32Z

Failing that, we could issue an error, but at least direct people to some place that explains the fix. That may be sufficient, and doesn't require as much precision, since it's ok to direct people unnecessarily.

nikomatsakis · 2016-04-22T10:28:29Z

(I'd like to see the crater impact; if it is zero or very minimal, issuing an error and trying to direct people is probably adequate.)

arielb1 · 2016-04-22T10:28:50Z

@nikomatsakis

Reporting a warning would put us in caching hell. I will improve the error message.

nikomatsakis · 2016-04-22T10:29:50Z

@arielb1 ok, but let's get the crater run going too.

arielb1 · 2016-04-22T10:56:36Z

@nikomatsakis

Done.

nikomatsakis · 2016-04-22T12:35:24Z

@arielb1 crater run started

nikomatsakis · 2016-04-22T20:12:00Z

Crater run results: https://gist.github.com/nikomatsakis/abb0ed492d3cb785af858b87ad3b95fd

4668 crates tested: 3047 working / 1528 broken / 35 regressed / 2 fixed / 56 unknown.

I've not investigated these at all.

jonas-schievink · 2016-04-22T20:28:03Z

shapir is using Box<Error + Sized>
All other root regressions are due to trait methods returning Self in a tuple without a Self: Sized bound (I think)

nikomatsakis · 2016-04-22T20:32:18Z

Tricky. It seems like we can pull that change to tuples out into a distinct PR, and introduce as a warning first -- it may be that we can write some a kind of "warning post-pass" that checks all tuples types that occur in trait methods and issues warnings. This wouldn't cover all possible cases, but hopefully it would hit those that occur in practice (I think they are mostly tied to default methods?). Breaking 34 crates without a warning period seems ungood.

cc @rust-lang/core

arielb1 · 2016-04-22T20:59:32Z

The problem is that allowing tuples with unsized elements to be sized is too much of a footgun. I guess I should just disallow that.

bluss · 2016-04-22T23:07:42Z

@nikomatsakis Fixing rayon cleans up a majority of that list, so it's not so bad. ndarray has been fixed with 0.5.1.

this commit should be reverted after a release cycle

arielb1 · 2016-05-03T15:53:08Z

rebased

nikomatsakis · 2016-05-03T16:21:35Z

src/librustc_lint/lib.rs

+        FutureIncompatibleInfo {
+            id: LintId::of(UNSIZED_IN_TUPLE),
+            reference: "RFC PR 1592 <https://github.com/rust-lang/rfcs/pull/1592>",
+        }


Should we add info for OBJECT_UNSAFE_FRAGMENT too?

Also, can you point this at the summary issue that I see you are drafting? (#33242)

nikomatsakis · 2016-05-03T17:29:40Z

@bors r+

bors · 2016-05-03T17:29:41Z

📌 Commit 2d0fcc9 has been approved by nikomatsakis

arielb1 · 2016-05-04T22:28:19Z

@bors r=nikomatsakis

bors · 2016-05-04T22:28:20Z

📌 Commit ef581db has been approved by nikomatsakis

arielb1 · 2016-05-04T23:00:29Z

@bors r=nikomatsakis

bors · 2016-05-04T23:00:30Z

📌 Commit 238e4ee has been approved by nikomatsakis

bors · 2016-05-06T13:21:26Z

⌛ Testing commit 238e4ee with merge 5158f3b...

@nikomatsakis

Short-cut `T: Sized` trait selection for ADTs Basically avoids all nested obligations when checking whether an ADT is sized - this speeds up typeck by ~15% The refactoring fixed #32963, but I also want to make `Copy` not object-safe (will commit that soon). Fixes #33201 r? @nikomatsakis

bors · 2016-05-06T15:31:56Z

☀️ Test successful - auto-linux-32-nopt-t, auto-linux-32-opt, auto-linux-32cross-opt, auto-linux-64-cargotest, auto-linux-64-cross-armhf, auto-linux-64-cross-armsf, auto-linux-64-cross-freebsd, auto-linux-64-cross-netbsd, auto-linux-64-debug-opt, auto-linux-64-nopt-t, auto-linux-64-opt, auto-linux-64-opt-mir, auto-linux-64-opt-rustbuild, auto-linux-64-x-android-t, auto-linux-cross-opt, auto-linux-musl-64-opt, auto-mac-32-opt, auto-mac-64-nopt-t, auto-mac-64-opt, auto-mac-64-opt-rustbuild, auto-mac-cross-ios-opt, auto-win-gnu-32-nopt-t, auto-win-gnu-32-opt, auto-win-gnu-32-opt-rustbuild, auto-win-gnu-64-nopt-t, auto-win-gnu-64-opt, auto-win-msvc-32-cross-opt, auto-win-msvc-32-opt, auto-win-msvc-64-cargotest, auto-win-msvc-64-opt, auto-win-msvc-64-opt-mir, auto-win-msvc-64-opt-rustbuild

rust-highfive assigned nikomatsakis Apr 21, 2016

jonas-schievink reviewed Apr 21, 2016
View reviewed changes

arielb1 force-pushed the sized-shortcut branch from 5c4c9af to 3df303e Compare April 21, 2016 23:33

arielb1 added the S-waiting-on-crater Status: Waiting on a crater run to be completed. label Apr 21, 2016

arielb1 force-pushed the sized-shortcut branch from 3df303e to f885f7a Compare April 21, 2016 23:37

nikomatsakis reviewed Apr 22, 2016
View reviewed changes

nikomatsakis mentioned this pull request Apr 22, 2016

Normalize lazilly for sized #32844

Closed

nikomatsakis reviewed Apr 22, 2016
View reviewed changes

arielb1 mentioned this pull request Apr 23, 2016

Add Sized bounds to meet new WF requirements rayon-rs/rayon#76

Merged

arielb1 force-pushed the sized-shortcut branch from 85777ba to 645628a Compare April 27, 2016 13:25

arielb1 and others added 6 commits May 3, 2016 18:30

fix breaking changes

6fc19ad

address review comments

05f1a05

improve error message for WF Tuples

5876b4b

change the newly-added errors to warnings

6057a7f

this commit should be reverted after a release cycle

stop using commit_if_ok where no errors can happen

62db786

stop dropping impls from cause backtraces

6c88384

arielb1 force-pushed the sized-shortcut branch from 354d46d to 07ffa81 Compare May 3, 2016 15:53

nikomatsakis reviewed May 3, 2016
View reviewed changes

arielb1 force-pushed the sized-shortcut branch from 07ffa81 to 2d0fcc9 Compare May 3, 2016 17:28

nikomatsakis removed the S-waiting-on-crater Status: Waiting on a crater run to be completed. label May 4, 2016

arielb1 force-pushed the sized-shortcut branch from 2d0fcc9 to ef581db Compare May 4, 2016 22:27

fixes

238e4ee

arielb1 force-pushed the sized-shortcut branch from ef581db to 238e4ee Compare May 4, 2016 22:59

nikomatsakis mentioned this pull request May 6, 2016

Bogus error in beta and nightly: recursive type has infinite size #31299

Closed

bors merged commit 238e4ee into rust-lang:master May 6, 2016

bluss added the relnotes Marks issues that should be documented in the release notes of the next release. label May 6, 2016

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Short-cut `T: Sized` trait selection for ADTs #33138

Short-cut `T: Sized` trait selection for ADTs #33138

arielb1 commented Apr 21, 2016 •

edited

Loading

jonas-schievink Apr 21, 2016

arielb1 commented Apr 21, 2016

nikomatsakis commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

nikomatsakis Apr 22, 2016

arielb1 Apr 22, 2016

nikomatsakis Apr 22, 2016

nikomatsakis commented Apr 22, 2016

arielb1 commented Apr 22, 2016 •

edited

Loading

nikomatsakis commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

arielb1 commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

arielb1 commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

jonas-schievink commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

arielb1 commented Apr 22, 2016 •

edited

Loading

bluss commented Apr 22, 2016 •

edited

Loading

arielb1 commented May 3, 2016

nikomatsakis May 3, 2016

nikomatsakis commented May 3, 2016

bors commented May 3, 2016

arielb1 commented May 4, 2016

bors commented May 4, 2016

arielb1 commented May 4, 2016

bors commented May 4, 2016

bors commented May 6, 2016

bors commented May 6, 2016

Short-cut T: Sized trait selection for ADTs #33138

Short-cut T: Sized trait selection for ADTs #33138

Conversation

arielb1 commented Apr 21, 2016 • edited Loading

jonas-schievink Apr 21, 2016

Choose a reason for hiding this comment

arielb1 commented Apr 21, 2016

nikomatsakis commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

nikomatsakis Apr 22, 2016

Choose a reason for hiding this comment

arielb1 Apr 22, 2016

Choose a reason for hiding this comment

nikomatsakis Apr 22, 2016

Choose a reason for hiding this comment

nikomatsakis commented Apr 22, 2016

arielb1 commented Apr 22, 2016 • edited Loading

nikomatsakis commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

arielb1 commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

arielb1 commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

jonas-schievink commented Apr 22, 2016

nikomatsakis commented Apr 22, 2016

arielb1 commented Apr 22, 2016 • edited Loading

bluss commented Apr 22, 2016 • edited Loading

arielb1 commented May 3, 2016

nikomatsakis May 3, 2016

Choose a reason for hiding this comment

nikomatsakis commented May 3, 2016

bors commented May 3, 2016

arielb1 commented May 4, 2016

bors commented May 4, 2016

arielb1 commented May 4, 2016

bors commented May 4, 2016

bors commented May 6, 2016

bors commented May 6, 2016

Short-cut `T: Sized` trait selection for ADTs #33138

Short-cut `T: Sized` trait selection for ADTs #33138

arielb1 commented Apr 21, 2016 •

edited

Loading

arielb1 commented Apr 22, 2016 •

edited

Loading

arielb1 commented Apr 22, 2016 •

edited

Loading

bluss commented Apr 22, 2016 •

edited

Loading