Commit Graph

807 Commits

Author SHA1 Message Date
Mara Bos
b263981fd7
Rollup merge of #81779 - geogriff:const-ptr-to-int-error, r=lcnr
improve error message for disallowed ptr-to-int casts in const eval

Improves an error message as [suggested](https://github.com/rust-lang/rust/issues/80875#issuecomment-762754580) in #80875.

Does the wording make enough sense? I tried to follow precedent for error message style while maintaining brevity.

It seems like the rest of the `ConstEvalErrKind::NeedsRfc` error messages could be improved as well. I could give that a go if this approach works.

Closes #80875
2021-02-08 19:28:17 +01:00
Jeffrey Griffin
ff315e34e8 improve error message for disallowed ptr-to-int casts in const eval 2021-02-05 00:52:19 -08:00
Tomasz Miąsko
eb5e2d08c7 Never MIR inline functions with a different instruction set 2021-02-05 00:00:00 +00:00
Aaron Hill
bc84e21107
Fix panic when emitting diagnostic for closure mutable binding error
Fixes #81700

The upvar borrow kind may be `ty::BorrowKind::UniqueImmBorrow`, which is
still a mutable borrow for the purposes of this diagnostic code.
2021-02-03 12:36:24 -05:00
Jonas Schievink
9e8753e2bb
Rollup merge of #81463 - matsujika:nll-ensure-c-case, r=varkor
Rename NLL* to Nll* accordingly to C-CASE

Given [C-CASE](https://rust-lang.github.io/api-guidelines/naming.html#casing-conforms-to-rfc-430-c-case), `NLLRegionVariableOrigin` and `NLL` are encouraged to be `NllRegionVariableOrigin` and `Nll` respectively.
2021-02-01 14:29:35 +01:00
Jonas Schievink
99f2f5a830
Rollup merge of #80404 - JulianKnodt:arr_ref, r=oli-obk
Remove const_in_array_repeat

Fixes #80371. Fixes #81315. Fixes #80767. Fixes #75682.

I thought there might be some issue with `Repeats(_, 0)`, but if you increase the items in the array it still ICEs. I'm not sure if this is the best fix but it does fix the given issue.
2021-01-31 16:36:42 +01:00
Jonas Schievink
7e3a8ec688
Rollup merge of #80092 - sexxi-goose:restrict_precision, r=nikomatsakis
2229: Fix issues with move closures and mutability

This PR fixes two issues when feature `capture_disjoint_fields` is used.

1. Can't mutate using a mutable reference
2. Move closures try to move value out through a reference.

To do so, we
1. Compute the mutability of the capture and store it as part of the `CapturedPlace`  that is written in TypeckResults
2. Restrict capture precision. Note this is temporary for now, to allow the feature to be used with move closures and ByValue captures and might change depending on discussions with the lang team.
    - No Derefs are captured for ByValue captures, since that will result in value behind a reference getting moved.
    - No projections are applied to raw pointers since these require unsafe blocks. We capture
    them completely.

r? `````@nikomatsakis`````
2021-01-31 16:36:41 +01:00
bors
9b32429822 Auto merge of #81327 - RalfJung:codegen-no-const-fail, r=oli-obk
codegen: assume constants cannot fail to evaluate

https://github.com/rust-lang/rust/pull/80579 landed, so we can finally remove this old hack from codegen and instead assume that consts never fail to evaluate. :)

r? `@oli-obk`
2021-01-31 07:03:09 +00:00
kadmin
6946534d84 Remove const_in_array_rep_expr 2021-01-30 23:20:24 +00:00
Ralf Jung
944237f6cd codegen: assume constants cannot fail to evaluate
also don't submit code to LLVM when the session has errors
2021-01-30 12:29:57 +01:00
Yuki Okushi
91ea1cbc17
Rollup merge of #80959 - jhpratt:unsigned_abs-stabilization, r=m-ou-se
Stabilize `unsigned_abs`

Resolves #74913.

This PR stabilizes the `i*::unsigned_abs()` method, which returns the absolute value of an integer _as its unsigned equivalent_. This has the advantage that it does not overflow on `i*::MIN`.

I have gone ahead and used this in a couple locations throughout the repository.
2021-01-30 13:36:44 +09:00
Aman Arora
fadf03ee1b Fix typos 2021-01-29 16:01:27 -05:00
Aman Arora
ffd53277dc Add fixme for precise path diagnostics 2021-01-29 15:37:44 -05:00
Aman Arora
c748f32ee4 Fix incorrect use mut diagnostics 2021-01-29 15:37:44 -05:00
Aman Arora
604cbdcfdd Fix unused 'mut' warning for capture's root variable 2021-01-29 15:37:42 -05:00
Aman Arora
3488082582 Compute mutability of closure captures
When `capture_disjoint_fields` is not enabled, checking if the root variable
binding is mutable would suffice.

However with the feature enabled, the captured place might be mutable
because it dereferences a mutable reference.

This PR computes the mutability of each capture after capture analysis
in rustc_typeck. We store this in `ty::CapturedPlace` and then use
`ty::CapturedPlace::mutability` in mir_build and borrow_check.
2021-01-29 15:37:40 -05:00
bors
b122908617 Auto merge of #81470 - tmiasko:remove-allocations, r=matthewjasper
Avoid memory allocation when removing dead blocks

Use `reachable_as_bitset` to reuse a bitset from the traversal rather
than allocating it seprately. Additionally check if there are any
unreachable blocks before proceeding.
2021-01-29 13:10:09 +00:00
bors
099f27b6cb Auto merge of #81440 - tmiasko:always-live-locals, r=matthewjasper
Visit only statements in always live locals

No functional changes intended.
2021-01-29 06:56:29 +00:00
Yuki Okushi
046a414c19
Rollup merge of #81333 - RalfJung:const-err-simplify, r=oli-obk
clean up some const error reporting around promoteds

These are some error reporting simplifications enabled by https://github.com/rust-lang/rust/pull/80579.

Further simplifications are possible but could be blocked on making `const_err` a hard error.

r? ``````@oli-obk``````
2021-01-29 09:17:38 +09:00
Yuki Okushi
4283623bc0
Rollup merge of #81158 - 1000teslas:issue-80313-fix, r=Aaron1011
Point to span of upvar making closure FnMut

For #80313.
2021-01-29 09:17:31 +09:00
bors
bf193d69fe Auto merge of #81441 - tmiasko:ctfe-inline, r=oli-obk
Try inlining trivial functions used by CTFE

r? `@ghost`
2021-01-28 15:19:16 +00:00
Hirochika Matsumoto
f8e0e78d75 Rename NLL* to Nll* accordingly to C-CASE 2021-01-28 16:18:25 +09:00
Yuki Okushi
446edd1e1a
Rollup merge of #79951 - LeSeulArtichaut:ty-ir, r=nikomatsakis
Refractor a few more types to `rustc_type_ir`

In the continuation of #79169, ~~blocked on that PR~~.

This PR:
 - moves `IntVarValue`, `FloatVarValue`, `InferTy` (and friends) and `Variance`
 - creates the `IntTy`, `UintTy` and `FloatTy` enums in `rustc_type_ir`, based on their `ast` and `chalk_ir` equilavents, and uses them for types in the rest of the compiler.

~~I will split up that commit to make this easier to review and to have a better commit history.~~
EDIT: done, I split the PR in commits of 200-ish lines each

r? `````@nikomatsakis````` cc `````@jackh726`````
2021-01-28 15:09:02 +09:00
Tomasz Miąsko
dc6ec3596b Avoid memory allocation when removing dead blocks
Use `reachable_as_bitset` to reuse a bitset from the traversal rather
than allocating it seprately. Additionally check if there are any
unreachable blocks before proceeding.
2021-01-28 00:00:00 +00:00
Tomasz Miąsko
20982b386f Inline MemPlace::offset 2021-01-27 16:13:58 +01:00
kadmin
fe39653116 Check that value is explicitly none 2021-01-27 03:56:54 +00:00
Tomasz Miąsko
56865936a7 Visit only statements in always live locals
No functional changes intended.
2021-01-27 00:00:00 +00:00
Yuki Okushi
c2c90bf548
Rollup merge of #80900 - camelid:readpointerasbytes-ice, r=oli-obk
Fix ICE with `ReadPointerAsBytes` validation error

Fixes #79690.

r? ``````@oli-obk``````
2021-01-27 04:43:16 +09:00
bors
f4eb5d9f71 Auto merge of #68828 - oli-obk:inline_cycle, r=wesleywiser
Prevent query cycles in the MIR inliner

r? `@eddyb` `@wesleywiser`

cc `@rust-lang/wg-mir-opt`

The general design is that we have a new query that is run on the `validated_mir` instead of on the `optimized_mir`. That query is forced before going into the optimization pipeline, so as to not try to read from a stolen MIR.

The query should not be cached cross crate, as you should never call it for items from other crates. By its very design calls into other crates can never cause query cycles.

This is a pessimistic approach to inlining, since we strictly have more calls in the `validated_mir` than we have in `optimized_mir`, but that's not a problem imo.
2021-01-25 19:03:37 +00:00
1000teslas
26b4baf46e Point to span of upvar making closure FnMut
Add expected error

Add comment

Tweak comment wording

Fix after rebase to updated master

Fix after rebase to updated master

Distinguish mutation in normal and move closures

Tweak error message

Fix error message for nested closures

Refactor code showing mutated upvar in closure

Remove debug assert

B
2021-01-25 16:53:27 +11:00
Jonas Schievink
d9c177f777
Rollup merge of #78578 - oli-obk:const_mut_refs, r=RalfJung
Permit mutable references in all const contexts

fixes #71212

cc `@rust-lang/wg-const-eval` `@christianpoveda`
2021-01-24 22:09:48 +01:00
Ralf Jung
48f9dbfd59 clean up some const error reporting around promoteds 2021-01-24 13:34:34 +01:00
Jonas Schievink
3a3470bf04
Rollup merge of #81243 - osa1:fix_80742_2, r=RalfJung
mir: Improve size_of handling when arg is unsized

As discussed on Zulip with `@RalfJung.`
2021-01-23 20:16:04 +01:00
Jonas Schievink
3382771dfd
Rollup merge of #81072 - RalfJung:place-ref-ty, r=oli-obk
PlaceRef::ty: use method call syntax
2021-01-23 20:15:59 +01:00
oli
209889ddc1 Leave some notes for future changes to the MIR opt level of mir inlining 2021-01-23 16:51:23 +00:00
oli
0491e74dd9 Make sure that const prop does not produce unsilenceable lints after inlining 2021-01-23 16:51:23 +00:00
oli
b8727e2d60 Prevent query cycles during inlining 2021-01-23 16:51:22 +00:00
bors
4d0dd02ee0 Auto merge of #80579 - RalfJung:no-fallible-promotion, r=oli-obk
avoid promoting division, modulo and indexing operations that could fail

For division, `x / y` will still be promoted if `y` is a non-zero integer literal; however, `1/(1+1)` will not be promoted any more.

While at it, also see if we can reject promoting floating-point arithmetic (which are [complicated](https://github.com/rust-lang/unsafe-code-guidelines/issues/237) so maybe we should not promote them).

This will need a crater run to see if there's code out there that relies on these things being promoted.

If we can land this, promoteds in `fn`/`const fn` cannot fail to evaluate any more, which should let us do some simplifications in codegen/Miri!

Cc https://github.com/rust-lang/rfcs/pull/3027
Fixes https://github.com/rust-lang/rust/issues/61821
r? `@oli-obk`
2021-01-23 13:19:04 +00:00
oli
14f39aa81a Do not allow arbitrary mutable references in static mut, just keep with the existing exceptions 2021-01-23 11:33:45 +00:00
oli
00e62fabf1 Adjust wording of a diagnostic 2021-01-23 11:33:45 +00:00
oli
d118021f8b Permit mutable references in all const contexts 2021-01-23 11:33:45 +00:00
bors
f2de221b00 Auto merge of #81101 - tmiasko:combine-now, r=nagisa
Combine instructions immediately
2021-01-22 13:10:48 +00:00
Ralf Jung
f62cecd807 do promote array indexing if we know it is in-bounds 2021-01-22 10:36:25 +01:00
Ralf Jung
5be27b7a70 avoid promoting division, modulo and indexing operations that could fail 2021-01-22 10:21:49 +01:00
Ömer Sinan Ağacan
e3faeb486a mir: Improve size_of handling when arg is unsized 2021-01-21 22:17:05 +03:00
bors
65767e5653 Auto merge of #81122 - tmiasko:no-drop, r=davidtwco
Visit only terminators when removing unneeded drops

No functional changes intended
2021-01-21 17:02:49 +00:00
Yuki Okushi
2ebc036220
Rollup merge of #81187 - eltociear:patch-6, r=jonas-schievink
Fix typo in counters.rs

formating -> formatting
2021-01-21 20:04:55 +09:00
Yuki Okushi
bc950c85c5
Rollup merge of #81185 - osa1:fix_80742, r=oli-obk
Fix ICE in mir when evaluating SizeOf on unsized type

Not quite ready yet. This tries to fix #80742 as discussed on [Zulip topic][1],
by using `delay_span_bug`.

I don't understand what `delay_span_bug` does. It seems like my error message
is never used. With this patch, in this program:

```rust
#![allow(incomplete_features)]
#![feature(const_evaluatable_checked)]
#![feature(const_generics)]

use std::fmt::Debug;
use std::marker::PhantomData;
use std::mem::size_of;

struct Inline<T>
where
    [u8; size_of::<T>() + 1]: ,
{
    _phantom: PhantomData<T>,
    buf: [u8; size_of::<T>() + 1],
}

impl<T> Inline<T>
where
    [u8; size_of::<T>() + 1]: ,
{
    pub fn new(val: T) -> Inline<T> {
        todo!()
    }
}

fn main() {
    let dst = Inline::<dyn Debug>::new(0); // line 27
}
```

these errors are printed, both for line 27 (annotated line above):

- "no function or associated item named `new` found for struct `Inline<dyn
  Debug>` in the current scope"
- "the size for values of type `dyn Debug` cannot be known at compilation time"

Second error makes sense, but I'm not sure about the first one and why it's
even printed.

Finally, I'm not sure about the span passing in `const_eval`.

[1]: https://rust-lang.zulipchat.com/#narrow/stream/269128-miri/topic/Help.20fixing.20.2380742
2021-01-21 20:04:53 +09:00
Yuki Okushi
cd0c54abb9
Rollup merge of #81178 - tmiasko:no-landing-pads, r=oli-obk
Visit only terminators when removing landing pads

No functional changes intended
2021-01-21 20:04:48 +09:00
Ömer Sinan Ağacan
3fb53c2c85 Fix ICE in mir when evaluating SizeOf on unsized type
Fixes #80742
2021-01-19 18:35:21 +03:00