Make all coercion valid casts and add trivial cast lints by nrc · Pull Request #23630 · rust-lang/rust

nrc · 2015-03-23T04:51:17Z

See notes on the first commit

petrochenkov · 2015-03-23T06:22:23Z

The lint names don't follow the conventions. trivial_cast -> trivial_casts, trivial_numeric_cast -> trivial_numeric_casts?

eddyb · 2015-03-23T07:30:16Z

This seems to not interfere too much with my own DST branch (where I did a similar change to coercions to have both a and b shallow resolutions in the same place).

nikomatsakis · 2015-03-23T19:49:50Z

I'm wondering if we want the lints to default to warn, in the absence of type ascription? I guess it is very rare to need to convert to a trait object, and one can typically use an intermediate variable, and that seems like the only case where one might require a cast today (other than the trivial numeric case, which is already separated). I agree that the lint names seem wrong.

nrc · 2015-03-23T19:59:38Z

Yeah, should be &x in the examples

nrc · 2015-03-23T20:01:05Z

Type ascription doesn't let you do any more, it is just more convenient, given there's an implementation that just needs rebasing and an accepted RFC, I would expect it to land very soon in any case.

nikomatsakis · 2015-03-23T20:43:52Z

r+

nrc · 2015-03-23T22:24:03Z

@bors r=nikomatsakis c2fde5a

bors · 2015-03-23T22:24:04Z

🙀 c2fde5a is not a valid commit SHA. Please try again with c4f3029.

nrc · 2015-03-23T22:24:34Z

@bors r=nikomatsakis c4f3029

nrc · 2015-03-23T23:02:19Z

@bors: try

bors · 2015-03-23T23:02:22Z

⌛ Trying commit c4f3029 with merge d8f41c5...

bors · 2015-03-23T23:41:24Z

💔 Test failed - try-bsd

nrc · 2015-03-24T00:51:06Z

@bors: try

bors · 2015-03-24T00:51:08Z

⌛ Trying commit 48287fe with merge 31738c0...

bors · 2015-03-24T00:56:09Z

💔 Test failed - try-mac

nrc · 2015-03-24T01:11:46Z

@bors: try

bors · 2015-03-24T01:11:47Z

⌛ Trying commit fc7e9b3 with merge 8381980...

bors · 2015-03-24T01:36:17Z

💔 Test failed - try-win-64

nrc · 2015-03-24T01:51:30Z

@bors: try

bors · 2015-03-24T01:51:32Z

⌛ Trying commit 7315e0a with merge 2782e72...

bors · 2015-03-24T02:18:17Z

💔 Test failed - try-win-64

nrc · 2015-03-24T02:32:41Z

@bors: try

@nikomatsakis

See notes on the first commit Closes#18601 r? @nikomatsakis cc @eddyb

bors · 2015-03-24T02:32:43Z

⌛ Trying commit a7b6af9 with merge ce76bff...

bors · 2015-03-24T03:41:18Z

💔 Test failed - try-mac

nrc · 2015-03-24T04:01:13Z

@bors: try

nrc · 2015-03-24T08:24:51Z

@bors r=nikomatsakis a7b6af9

nrc · 2015-03-24T20:10:35Z

@bors retry

bors · 2015-03-24T20:10:38Z

⌛ Trying commit a7b6af9 with merge 3f8058d...

bors · 2015-03-24T21:02:51Z

☔ The latest upstream changes (presumably #23654) made this pull request unmergeable. Please resolve the merge conflicts.

This permits all coercions to be performed in casts, but adds lints to warn in those cases. Part of this patch moves cast checking to a later stage of type checking. We acquire obligations to check casts as part of type checking where we previously checked them. Once we have type checked a function or module, then we check any cast obligations which have been acquired. That means we have more type information available to check casts (this was crucial to making coercions work properly in place of some casts), but it means that casts cannot feed input into type inference. [breaking change] * Adds two new lints for trivial casts and trivial numeric casts, these are warn by default, but can cause errors if you build with warnings as errors. Previously, trivial numeric casts and casts to trait objects were allowed. * The unused casts lint has gone. * Interactions between casting and type inference have changed in subtle ways. Two ways this might manifest are: - You may need to 'direct' casts more with extra type information, for example, in some cases where `foo as _ as T` succeeded, you may now need to specify the type for `_` - Casts do not influence inference of integer types. E.g., the following used to type check: ``` let x = 42; let y = &x as *const u32; ``` Because the cast would inform inference that `x` must have type `u32`. This no longer applies and the compiler will fallback to `i32` for `x` and thus there will be a type error in the cast. The solution is to add more type information: ``` let x: u32 = 42; let y = &x as *const u32; ```

nrc · 2015-03-24T21:38:10Z

@bors: r=nikomatsakis 7e3ee02

@nikomatsakis

See notes on the first commit Closesrust-lang#18601 r? @nikomatsakis cc @eddyb

alexcrichton · 2015-03-26T16:33:17Z

@nrc how would you feel about making the trivial_numeric_casts lint allow-by-default? Right now if you're using a libc type you don't actually know what the definition is (e.g. libc::time_t) so it's casted into an i64 so we know what it is. This conversion, though, is a trivial cast on some platforms (but not necessarily on all platforms).

lilyball · 2015-03-26T17:25:20Z

I just filed #23739 requesting that trivial_numeric_casts ignore casts between types that are type aliases. This was motivated by libc types.

bluss · 2015-03-26T18:55:24Z

Ah, breaking change needs to be exactly [breaking-change] (with hyphen) to show up in the live list, so this didn't.

rust-highfive assigned nikomatsakis Mar 23, 2015

petrochenkov reviewed Mar 23, 2015
View reviewed changes

nrc force-pushed the coerce-tidy branch from f441fbd to c2fde5aCompare March 23, 2015 22:14

nrc force-pushed the coerce-tidy branch from 48287fe to fc7e9b3Compare March 24, 2015 01:11

nrc force-pushed the coerce-tidy branch from fc7e9b3 to 7315e0aCompare March 24, 2015 01:51

nrc force-pushed the coerce-tidy branch from 7315e0a to a7b6af9Compare March 24, 2015 02:31

bors added a commit that referenced this pull request Mar 24, 2015
Auto merge of#23630- nrc:coerce-tidy, r=<try>
ce76bff
See notes on the first commit Closes#18601 r? @nikomatsakis cc @eddyb

Nick Cameron added 6 commits March 25, 2015 10:03

Minor refactoring in coercion.rs
9374c21

Remove the UnusedCasts lint
088cd56

Add tests
dc206a9

Change lint names to plurals
e7122a5

Bug fixes
7e3ee02

nrc force-pushed the coerce-tidy branch from a7b6af9 to 7e3ee02Compare March 24, 2015 21:37

alexcrichton added a commit to alexcrichton/rust that referenced this pull request Mar 24, 2015
rollup merge ofrust-lang#23630: nrc/coerce-tidy
a1d2e62
See notes on the first commit Closesrust-lang#18601 r? @nikomatsakis cc @eddyb

bors merged commit 7e3ee02 into rust-lang:masterMar 25, 2015

Ms2ger mentioned this pull request Mar 25, 2015
Coercion to an unspecified but easily inferrable raw pointer (as *const _) has regressed #23699
Closed

lilyball mentioned this pull request Mar 26, 2015
trivial_numeric_casts warning incorrect when casting between type aliases #23739
Closed

bluss mentioned this pull request Mar 26, 2015
Trivial cast warning for cast that can't be removed #23742
Closed

eddyb mentioned this pull request Mar 26, 2015
Implement type ascription. #21836
Closed

This was referenced Mar 26, 2015
Move the trivial_numeric_casts lint to allow-by-default rust-lang/rfcs#1020
Closed
trivial_numeric_cast makes working with type aliases impossible #23770
Closed

Uh oh!

Uh oh!

Conversation

nrc commented Mar 23, 2015

Uh oh!

petrochenkovMar 23, 2015

Choose a reason for hiding this comment

Uh oh!

eddyb commented Mar 23, 2015

Uh oh!

nikomatsakis commented Mar 23, 2015

Uh oh!

nrc commented Mar 23, 2015

Uh oh!

nrc commented Mar 23, 2015

Uh oh!

nikomatsakis commented Mar 23, 2015

Uh oh!

nrc commented Mar 23, 2015

Uh oh!

bors commented Mar 23, 2015

Uh oh!

nrc commented Mar 23, 2015

Uh oh!

nrc commented Mar 23, 2015

Uh oh!

bors commented Mar 23, 2015

Uh oh!

bors commented Mar 23, 2015

Uh oh!

nrc commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

nrc commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

nrc commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

nrc commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

nrc commented Mar 24, 2015

Uh oh!

nrc commented Mar 24, 2015

Uh oh!

nrc commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

bors commented Mar 24, 2015

Uh oh!

nrc commented Mar 24, 2015

Uh oh!

alexcrichton commented Mar 26, 2015

Uh oh!

lilyball commented Mar 26, 2015

Uh oh!

bluss commented Mar 26, 2015

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

8 participants