Commit Graph

654 Commits

Author SHA1 Message Date
Tomasz Miąsko
43cadc99b7 Extend comment in UsedLocals::visit_lhs 2021-03-20 00:00:00 +00:00
Dylan DPC
90e52a1ad2
Rollup merge of #83277 - spastorino:early_otherwise-opt-unsound, r=oli-obk
Mark early otherwise optimization unsound

r? `@oli-obk`
cc `@tmiasko`

Related to #78496 and #82905

Should I also bump this one to level 3 or 4 or given that is unsound it doesn't matter?.
Probably need to adjust some tests.
2021-03-19 23:01:40 +01:00
Santiago Pastorino
778e1978d5
Mark early otherwise optimization unsound 2021-03-18 20:57:44 -03:00
Mara Bos
cfb4ad4f2a Remove unwrap_none/expect_none from compiler/. 2021-03-18 14:25:54 +01:00
Erik Desjardins
93c1380e0b remove inhabitedness check 2021-03-17 20:56:38 -04:00
Erik Desjardins
1205e82578 remove unnecessary and ineffective caching 2021-03-17 20:24:29 -04:00
Dylan DPC
b688b694d0
Rollup merge of #83080 - tmiasko:inline-coverage, r=wesleywiser
Make source-based code coverage compatible with MIR inlining

When codegenning code coverage use the instance that coverage data was
originally generated for, to ensure basic level of compatibility with
MIR inlining.

Fixes #83061
2021-03-18 00:28:09 +01:00
Erik Desjardins
bca761dc5d move to separate pass, cache layouts 2021-03-16 21:34:31 -04:00
bors
e655fb6221 Auto merge of #82936 - oli-obk:valtree, r=RalfJung,lcnr,matthewjasper
Implement (but don't use) valtree and refactor in preparation of use

This PR does not cause any functional change. It refactors various things that are needed to make valtrees possible. This refactoring got big enough that I decided I'd want it reviewed as a PR instead of trying to make one huge PR with all the changes.

cc `@rust-lang/wg-const-eval` on the following commits:

* 2027184 implement valtree
* eeecea9 fallible Scalar -> ScalarInt
* 042f663 ScalarInt convenience methods

cc `@eddyb` on ef04a6d

cc `@rust-lang/wg-mir-opt` for cf1700c (`mir::Constant` can now represent either a `ConstValue` or a `ty::Const`, and it is totally possible to have two different representations for the same value)
2021-03-16 22:42:56 +00:00
Erik Desjardins
899836306f remove assignments to ZST places 2021-03-15 21:57:52 -04:00
Erik Desjardins
affb081fa0 revert changes from #83118 to mark zst return place as unused 2021-03-15 20:18:03 -04:00
Tomasz Miąsko
1796cc0e6c Make source-based code coverage compatible with MIR inlining
When codegenning code coverage use the instance that coverage data was
originally generated for, to ensure basic level of compatibility with
MIR inlining.
2021-03-15 23:26:03 +01:00
Oli Scherer
c30c1be1e6 s/ConstantSource/ConstantKind/ 2021-03-15 12:06:52 +00:00
Erik Desjardins
47f8bacc46 remove unnecessary condition
`_local` isn't visited in `_local = <rhs>` statements in the situation
we care about
2021-03-14 20:21:20 -04:00
Simon Vandel Sillesen
bb36e3c7e7 Move ZST check inside UsedLocals 2021-03-13 18:00:28 -05:00
Simon Vandel Sillesen
4e901be57d Extend SimplifyLocals to remove ZST writes 2021-03-13 17:58:11 -05:00
Oli Scherer
3127a9c60f Prepare mir::Constant for ty::Const only supporting valtrees 2021-03-12 12:43:54 +00:00
Cameron Steffen
b0514a6a0a Rename Option::get_or_insert_default 2021-03-10 09:07:16 -06:00
Yuki Okushi
1c3fea2f8c
Rollup merge of #82849 - camsteffen:option-get-or-default, r=joshtriplett
Add Option::get_or_default

Tracking issue: #82901

The original issue is #55042, which was closed, but for an invalid reason (see discussion there). Opening this to reconsider (I hope that's okay). It seems like the only gap for `Option` being "entry-like".

I ran into a need for this method where I had a `Vec<Option<MyData>>` and wanted to do `vec[n].get_or_default().my_data_method()`. Using an `Option` as an inner component of a data structure is probably where the need for this will normally arise.
2021-03-10 08:01:32 +09:00
Yuki Okushi
c46f948a80
Rollup merge of #79208 - LeSeulArtichaut:stable-unsafe_op_in_unsafe_fn, r=nikomatsakis
Stabilize `unsafe_op_in_unsafe_fn` lint

This makes it possible to override the level of the `unsafe_op_in_unsafe_fn`, as proposed in https://github.com/rust-lang/rust/issues/71668#issuecomment-729770896.

Tracking issue: #71668
r? ```@nikomatsakis``` cc ```@SimonSapin``` ```@RalfJung```

# Stabilization report

This is a stabilization report for `#![feature(unsafe_block_in_unsafe_fn)]`.

## Summary

Currently, the body of unsafe functions is an unsafe block, i.e. you can perform unsafe operations inside.

The `unsafe_op_in_unsafe_fn` lint, stabilized here, can be used to change this behavior, so performing unsafe operations in unsafe functions requires an unsafe block.

For now, the lint is allow-by-default, which means that this PR does not change anything without overriding the lint level.

For more information, see [RFC 2585](https://github.com/rust-lang/rfcs/blob/master/text/2585-unsafe-block-in-unsafe-fn.md)

### Example

```rust
// An `unsafe fn` for demonstration purposes.
// Calling this is an unsafe operation.
unsafe fn unsf() {}

// #[allow(unsafe_op_in_unsafe_fn)] by default,
// the behavior of `unsafe fn` is unchanged
unsafe fn allowed() {
    // Here, no `unsafe` block is needed to
    // perform unsafe operations...
    unsf();

    // ...and any `unsafe` block is considered
    // unused and is warned on by the compiler.
    unsafe {
        unsf();
    }
}

#[warn(unsafe_op_in_unsafe_fn)]
unsafe fn warned() {
    // Removing this `unsafe` block will
    // cause the compiler to emit a warning.
    // (Also, no "unused unsafe" warning will be emitted here.)
    unsafe {
        unsf();
    }
}

#[deny(unsafe_op_in_unsafe_fn)]
unsafe fn denied() {
    // Removing this `unsafe` block will
    // cause a compilation error.
    // (Also, no "unused unsafe" warning will be emitted here.)
    unsafe {
        unsf();
    }
}
```
2021-03-10 08:01:25 +09:00
kadmin
4bceb294f4 Clean up todos
Also add some span_bugs where it is unreachable
2021-03-09 19:31:31 +00:00
kadmin
217ff6b7ea Switch to changing cp_non_overlap in tform
It was suggested to lower this in MIR instead of ssa, so do that instead.
2021-03-09 16:54:14 +00:00
kadmin
982382dc03 Update cranelift 2021-03-09 16:54:14 +00:00
kadmin
37a6c04718 Update interpret step 2021-03-09 16:54:13 +00:00
kadmin
89f45ed9f3 Update match branches
This updates all places where match branches check on StatementKind or UseContext.
This doesn't properly implement them, but adds TODOs where they are, and also adds some best
guesses to what they should be in some cases.
2021-03-09 16:54:13 +00:00
Cameron Steffen
1cc8c4de6a Use Option::get_or_default 2021-03-08 09:24:11 -06:00
Dylan DPC
dd7a606804
Rollup merge of #82684 - tmiasko:dest-prop, r=jonas-schievink
Disable destination propagation on all mir-opt-levels

The new `// compile-flags: -Zunsound-mir-opts` are inserted without an extra newline to avoid introducing a large mir-opt diff.
2021-03-08 13:13:24 +01:00
bors
27885a94c6 Auto merge of #82727 - oli-obk:shrinkmem, r=pnkfelix
Test the effect of shrinking the size of Rvalue by 16 bytes

r? `@ghost`
2021-03-08 08:39:24 +00:00
Tomasz Miąsko
6f49aadabb Disable destination propagation on all mir-opt-levels 2021-03-06 00:00:00 +00:00
Santiago Pastorino
004465c620
Bump one missing mir_opt_level 2021-03-05 17:13:58 -03:00
Santiago Pastorino
782c7b04cf
Bump all mir_opt_level 2 to 3 2021-03-05 17:13:57 -03:00
Santiago Pastorino
f3b8920d4b
Bump all mir_opt_level 3 to 4 2021-03-05 17:13:56 -03:00
Santiago Pastorino
fde8d6ee81
For better consistency change mir_opt_level <= 1 to < 2 2021-03-05 17:13:56 -03:00
Santiago Pastorino
8152da22a1
Extract mir_opt_level to a method and use Option to be able to know if the value is provided or not 2021-03-05 17:13:56 -03:00
Oli Scherer
9a2362e5a9 Shrink the size of Rvalue by 16 bytes 2021-03-05 09:33:01 +00:00
bors
6f7673d077 Auto merge of #81114 - bugadani:generator, r=estebank
Box generator-related Body fields

Might save some memory on functions that aren't generators.
2021-03-04 00:23:42 +00:00
Yuki Okushi
ae5e024194
Rollup merge of #82376 - tmiasko:inline-options, r=oli-obk
Add option to enable MIR inlining independently of mir-opt-level

Add `-Zinline-mir` option that enables MIR inlining independently of the
current MIR opt level. The primary use-case is enabling MIR inlining on the
default MIR opt level.

Turn inlining thresholds into optional values to make it possible to configure
different defaults depending on the current mir-opt-level (although thresholds
are yet to be used in such a manner).
2021-03-02 21:23:14 +09:00
Guillaume Gomez
865cf0c3b6
Rollup merge of #80734 - abonander:ab/issue-66693, r=oli-obk
check that first arg to `panic!()` in const is `&str`

closes #66693

~~TODO: regression test~~

cc `@RalfJung` for error message wording
2021-03-02 00:50:04 +01:00
Austin Bonander
5a33f531cd check that first arg to panic!() in const is &str 2021-03-01 08:32:15 -08:00
Dániel Buga
b97eb23cd0 Box generator-related Body fields 2021-03-01 08:32:49 +01:00
Tomasz Miąsko
8b184ff1b8 Remove storage markers if they won't be used during code generation
The storage markers constitute a substantial portion of all MIR
statements. At the same time, for builds without any optimizations,
the storage markers have no further use during and after MIR
optimization phase.

If storage markers are not necessary for code generation, remove them.
2021-02-28 20:10:44 +01:00
Tomasz Miąsko
500aeccc5b Use optional values for inlining thresholds
Turn inlining threshold into optional values to make it possible to
configure different defaults depending on the current mir-opt-level.
2021-02-27 10:19:19 +01:00
Tomasz Miąsko
f895f1c35a Add option enabling MIR inlining independently of mir-opt-level 2021-02-27 10:18:06 +01:00
Dylan DPC
2942cf5a1f
Rollup merge of #82482 - tmiasko:small-cycles, r=varkor
Use small hash set in `mir_inliner_callees`

Use small hash set in `mir_inliner_callees` to avoid temporary
allocation when possible and quadratic behaviour for large number of
callees.
2021-02-27 02:34:32 +01:00
Guillaume Gomez
ee4129f70c
Rollup merge of #82491 - tmiasko:i, r=lcnr
Consider inexpensive inlining criteria first

Refactor inlining decisions so that inexpensive criteria are considered first:

1. Based on code generation attributes.
2. Based on MIR availability (examines call graph).
3. Based on MIR body.
2021-02-26 15:52:32 +01:00
bors
89d32eb1ea Auto merge of #82338 - RalfJung:interp-error-allocs, r=oli-obk
all InterpError allocate now, so adjust alloc-error-check

Cc https://github.com/rust-lang/rust/pull/82116#discussion_r578310770
r? `@oli-obk`
2021-02-25 08:27:09 +00:00
bors
1fdadbf13a Auto merge of #82159 - BoxyUwU:uwu, r=varkor
Use correct param_env in conservative_is_privately_uninhabited

cc `@lcnr`
r? `@varkor` since this is your FIXME that was removed ^^
2021-02-24 21:54:52 +00:00
bors
6b56603e35 Auto merge of #80475 - simonvandel:fix-77355, r=oli-obk
New mir-opt pass to simplify gotos with const values (reopening #77486)

Reopening PR #77486

Fixes #77355

This pass optimizes the following sequence
```rust
bb2: {
    _2 = const true;
    goto -> bb3;
}

bb3: {
    switchInt(_2) -> [false: bb4, otherwise: bb5];
}
```
into
```rust
bb2: {
    _2 = const true;
    goto -> bb5;
}
```
2021-02-24 07:23:54 +00:00
Tomasz Miąsko
ef731b31db Use small hash set in mir_inliner_callees
Use small hash set in `mir_inliner_callees` to avoid temporary
allocation when possible and quadratic behaviour for large number of
callees.
2021-02-24 00:00:00 +00:00
Tomasz Miąsko
6d5c0c1bcb Consider inexpensive inlining criteria first
Refactor inlining decisions so that inexpensive criteria are considered first:

1. Based on code generation attributes.
2. Based on MIR availability (examines call graph).
3. Based on MIR body.
2021-02-24 00:00:00 +00:00
Ellen
42cbfd6346 yeet 2021-02-23 23:35:59 +00:00
Dylan DPC
cc07061190
Rollup merge of #82091 - henryboisdequin:use-place-ref-more, r=RalfJung
use PlaceRef abstractions more consistently

Addresses this [comment](https://github.com/rust-lang/rust/pull/80865/files#r558978715)
Associated issue: #80647

r? ```@RalfJung```
2021-02-23 02:51:50 +01:00
Simon Vandel Sillesen
a6dccfeb23 New mir-opt pass to simplify gotos with const values
Fixes #77355
2021-02-22 21:03:57 +01:00
bors
15598a83db Auto merge of #77551 - simonvandel:extend-simplify-branch-same, r=oli-obk
MIR-OPT: Pass to deduplicate blocks

This pass finds basic blocks that are completely equal,
and replaces all uses with just one of them.

```bash
$ RUSTC_LOG=rustc_mir::transform::deduplicate_blocks ./x.py build --stage 2 | grep "SUCCESS: Replacing: " > log
...
$ cat log | wc -l
23875
```
2021-02-22 12:14:23 +00:00
Simon Vandel Sillesen
2d1e0adfe9 New pass to deduplicate blocks 2021-02-21 21:51:54 +01:00
Simon Vandel Sillesen
1e865709a6 Make MatchBranchSimplification clean up after itself 2021-02-21 21:22:04 +01:00
Simon Vandel Sillesen
ccecc4f6bf Drive-by formatting of comment 2021-02-21 21:22:04 +01:00
Matthias Krüger
da9a588d4f remove redundant wrapping of return types of allow_internal_unstable() and rustc_allow_const_fn_unstable() 2021-02-21 18:11:27 +01:00
Ralf Jung
d496bfc161 all InterpError allocate now, so adjust alloc-error-check 2021-02-20 19:01:25 +01:00
Henry Boisdequin
a9c6188889 make super_projection take a PlaceRef 2021-02-20 16:56:08 +05:30
bors
e7c23ab933 Auto merge of #82124 - tmiasko:op-ty-ref, r=oli-obk
Pass large interpreter types by reference, not value

r? `@ghost`
2021-02-20 10:20:42 +00:00
LeSeulArtichaut
ec20993c4d Stabilize unsafe_op_in_unsafe_fn lint 2021-02-18 17:12:15 +01:00
Eduard-Mihai Burtescu
6165d1cc72 Print -Ztime-passes (and misc stats/logs) on stderr, not stdout. 2021-02-18 14:13:38 +02:00
Tomasz Miąsko
e190f0d974 Reduce size of InterpErrorInfo to 8 bytes 2021-02-17 08:57:06 +01:00
Tomasz Miąsko
f2da425bff ./x.py fmt 2021-02-16 23:52:05 +01:00
Tomasz Miąsko
fe0c46d07e Pass PlaceTy by reference not value 2021-02-16 23:52:05 +01:00
Tomasz Miąsko
e915cf45dc Pass OpTy by reference not value 2021-02-16 23:52:05 +01:00
Henry Boisdequin
5ec4b060a7 make visit_projection take a PlaceRef 2021-02-16 14:20:36 +05:30
Henry Boisdequin
30c5125fbe update formating 2021-02-16 07:07:42 +05:30
Jonas Schievink
f02f7b05b2
Rollup merge of #81503 - henryboisdequin:fix-const-fn-arr-err-msg, r=estebank
Suggest to create a new `const` item if the `fn` in the array is a `const fn`

Fixes #73734. If the `fn` in the array repeat expression is a `const fn`, suggest creating a new `const` item. On nightly, suggest creating an inline `const` block. This PR also removes the `suggest_const_in_array_repeat_expressions` as it is no longer necessary.

Example:

```rust
fn main() {
    // Should not compile but hint to create a new const item (stable) or an inline const block (nightly)
    let strings: [String; 5] = [String::new(); 5];
    println!("{:?}", strings);
}

```

Gives this error:

```
error[E0277]: the trait bound `std::string::String: std::marker::Copy` is not satisfied
 --> $DIR/const-fn-in-vec.rs:3:32
  |
2 |     let strings: [String; 5] = [String::new(); 5];
  |                             ^^^^^^^^^^^^^^^^^^ the trait `std::marker::Copy` is not implemented for `String`
  |
  = note: the `Copy` trait is required because the repeated element will be copied
```

With this change, this is the error message:

```
error[E0277]: the trait bound `String: Copy` is not satisfied
  --> $DIR/const-fn-in-vec.rs:3:32
   |
LL |     let strings: [String; 5] = [String::new(); 5];
   |                                ^^^^^^^^^^^^^^^^^^ the trait `Copy` is not implemented for `String`
   |
   = help: moving the function call to a new `const` item will resolve the error
```
2021-02-15 16:06:47 +01:00
Henry Boisdequin
eace240ebe use option<PlaceRef<'tcx>> to clean up mir code a little 2021-02-14 14:39:24 +05:30
Tomasz Miąsko
361dcd5ca7 Use debug log level for developer oriented logs
The information logged here is of limited general interest, while at the
same times makes it impractical to simply enable logging and share the
resulting logs due to the amount of the output produced.

Reduce log level from info to debug for developer oriented information.

For example, when building cargo, this reduces the amount of logs
generated by `RUSTC_LOG=info cargo build` from 265 MB to 79 MB.

Continuation of changes from 81350.
2021-02-13 00:00:00 +00:00
bors
26e5bcd220 Auto merge of #81350 - tmiasko:instrument-debug, r=lcnr
Reduce log level used by tracing instrumentation from info to debug

Restore log level to debug to avoid make info log level overly verbose (the uses of instrument attribute modified there, were for the most part a replacement for `debug!`;  one use was novel).
2021-02-11 13:44:00 +00:00
Tomasz Miąsko
eb5e2d08c7 Never MIR inline functions with a different instruction set 2021-02-05 00:00:00 +00:00
Henry Boisdequin
c2e849c022 added a suggestion to create a const item if the fn in the array repeat expression is a const fn 2021-02-03 10:18:08 +05:30
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
kadmin
6946534d84 Remove const_in_array_rep_expr 2021-01-30 23:20:24 +00:00
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
kadmin
fe39653116 Check that value is explicitly none 2021-01-27 03:56:54 +00:00
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
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
Tomasz Miąsko
59457ab86e Reduce log level used by tracing instrumentation from info to debug 2021-01-24 00:00:00 +00:00
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
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
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
Ikko Ashimine
203df1764c
Fix typo in counters.rs
formating -> formatting
2021-01-19 23:42:18 +09:00
bors
47121d6d88 Auto merge of #81110 - LeSeulArtichaut:fix-unused-unsafe-label, r=RalfJung
Fix `unused_unsafe` label with `unsafe_block_in_unsafe_fn

Previously, the following code:

```rust
#![feature(unsafe_block_in_unsafe_fn)]

unsafe fn foo() {
    unsafe { unsf() }
}

unsafe fn unsf() {}
```

Would give the following warning:

```
warning: unnecessary `unsafe` block
 --> src/lib.rs:4:5
  |
4 |     unsafe { unsf() }
  |     ^^^^^^ unnecessary `unsafe` block
  |
  = note: `#[warn(unused_unsafe)]` on by default
```
which doesn't point out that the block is in an `unsafe fn`.

Tracking issue: #71668
cc #79208
2021-01-19 08:59:37 +00:00
bors
d98d2f57d9 Auto merge of #80707 - oli-obk:stability_hole_const_intrinsics, r=RalfJung
Stability oddity with const intrinsics

cc `@RalfJung`

In https://github.com/rust-lang/rust/pull/80699#discussion_r551495670 `@usbalbin` realized we accepted some intrinsics as `const` without a `#[rustc_const_(un)stable]` attribute. I did some digging, and that example works because intrinsics inherit their stability from their parents... including `#[rustc_const_(un)stable]` attributes. While we may want to fix that (not sure, wasn't there just a MCPed PR that caused this on purpose?), we definitely want tests for it, thus this PR adding tests and some fun tracing statements.
2021-01-18 20:54:36 +00:00
Tomasz Miąsko
508eec49e9 Combine instructions immediately
No functional changes intended
2021-01-18 13:15:27 +01:00
Tomasz Miąsko
a9292d871c Remove disabled transformation from instcombine 2021-01-18 13:15:27 +01:00
Ashley Mannix
e1d70bc697
Rollup merge of #81121 - tmiasko:simplify-cfg-no-dbg, r=jonas-schievink
Avoid logging the whole MIR body in SimplifyCfg
2021-01-18 21:53:31 +10:00
Ashley Mannix
064e47b99e
Rollup merge of #81116 - bugadani:body-span, r=wesleywiser
ConstProp: Copy body span instead of querying it
2021-01-18 21:53:30 +10:00
oli
5bac1c9229 Only inherit const stability for methods of impl const Trait blocks 2021-01-18 11:07:35 +00:00
bors
0677d97293 Auto merge of #80865 - oliviacrain:proj_based, r=RalfJung
Use PlaceRef projection abstractions more consistently in rustc_mir

PlaceRef contains abstractions for dealing with the `projections` array. This PR uses these abstractions more consistently within the `rustc_mir` crate.

See associated issue: rust-lang/rust#80647.

r? `@RalfJung`
2021-01-18 05:44:40 +00:00
Tomasz Miąsko
96e9562a7e Visit only terminators when removing landing pads
No functional changes intended
2021-01-18 00:00:00 +00:00
LeSeulArtichaut
2136a5cfad Fix unused_unsafe label with `unsafe_block_in_unsafe_fn 2021-01-17 16:42:27 +01:00
Tomasz Miąsko
c7bad7ba5d Avoid logging the whole MIR body in SimplifyCfg 2021-01-17 15:57:21 +01:00
Dániel Buga
428f948096 Copy body span instead of querying it 2021-01-17 14:50:47 +01:00
oli
949bdd8b79 Add regression test 2021-01-17 13:40:29 +00:00
oli
ad5aa2359d Remove an unnecessary field from a NonConstOp 2021-01-17 13:40:29 +00:00
Tomasz Miąsko
d0a756719f Visit only terminators when removing unneeded drops
No functional changes intended
2021-01-17 00:00:00 +00:00
Jack Huey
3dea68de1d Review changes 2021-01-16 18:56:37 -05:00
bors
492b83c697 Auto merge of #80290 - RalfJung:less-intrinsic-write, r=lcnr
implement ptr::write without dedicated intrinsic

This makes `ptr::write` more consistent with `ptr::write_unaligned`, `ptr::read`, `ptr::read_unaligned`, all of which are implemented in terms of `copy_nonoverlapping`.

This means we can also remove `move_val_init` implementations in codegen and Miri, and its special handling in the borrow checker.

Also see [this Zulip discussion](https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/ptr.3A.3Aread.20vs.20ptr.3A.3Awrite).
2021-01-16 17:28:32 +00:00
Olivia Crain
65b5e4386b Use PlaceRef more consistently in rustc_mir 2021-01-16 10:44:23 -06:00
Tomasz Miąsko
a65c050333 inline: Round word-size cost estimates up 2021-01-15 00:00:00 +00:00
LingMan
a56bffb4f9 Use Option::map_or instead of .map(..).unwrap_or(..) 2021-01-14 19:23:59 +01:00
bors
704e47f78b Auto merge of #78407 - oli-obk:ub_checkable_ctfe, r=RalfJung,pnkfelix
Make CTFE able to check for UB...

... by not doing any optimizations on the `const fn` MIR used in CTFE. This means we duplicate all `const fn`'s MIR now, once for CTFE, once for runtime. This PR is for checking the perf effect, so we have some data when talking about https://github.com/rust-lang/const-eval/blob/master/rfcs/0000-const-ub.md

To do this, we now have two queries for obtaining mir: `optimized_mir` and `mir_for_ctfe`. It is now illegal to invoke `optimized_mir` to obtain the MIR of a const/static item's initializer, an array length, an inline const expression or an enum discriminant initializer. For `const fn`, both `optimized_mir` and `mir_for_ctfe` work, the former returning the MIR that LLVM should use if the function is called at runtime. Similarly it is illegal to invoke `mir_for_ctfe` on regular functions.

This is all checked via appropriate assertions and I don't think it is easy to get wrong, as there should be no `mir_for_ctfe` calls outside the const evaluator or metadata encoding. Almost all rustc devs should keep using `optimized_mir` (or `instance_mir` for that matter).
2021-01-12 17:26:56 +00:00
oli
53e3a23572 Coverage computation needs access to the MIR, too 2021-01-12 15:12:03 +00:00
bors
fc9944fe84 Auto merge of #80499 - matthiaskrgr:red_clos, r=estebank
remove redundant closures (clippy::redundant_closure)
2021-01-12 11:20:47 +00:00
Nym Seddon
06fd212d6a
Add ABI argument to find_mir_or_eval_fn
Add ABI argument for called function in `find_mir_or_eval_fn` and
`call_extra_fn`. Useful for comparing with expected ABI in interpreters.

Related to [miri/1631](https://github.com/rust-lang/miri/issues/1631)
2021-01-10 15:12:50 +00:00
Yuki Okushi
3acd75dd25
Rollup merge of #80521 - richkadel:llvm-coverage-counters-2.4.0, r=wesleywiser
MIR Inline is incompatible with coverage

Fixes: #80060

Fixed by disabling inlining if `-Zinstrument-coverage` is set.

The PR also adds additional use cases to the coverage test for doctests.

r? `@wesleywiser`
cc: `@tmandry`
2021-01-08 02:06:03 +09:00
bors
bf5f30684a Auto merge of #80648 - Aaron1011:expn-data-private, r=petrochenkov
Make `ExpnData` fields `krate` and `orig_id` private

These fields are only used by hygiene serialized, and should not be
accessed by anything outside of `rustc_span`.
2021-01-07 08:25:39 +00:00
oli
65ee418e5c Do not run const prop on the mir_for_ctfe of const fn 2021-01-04 22:29:45 +00:00
oli
3af7989a7c No doc comments on expressions 2021-01-04 21:40:38 +00:00
oli
9eaec139d0 Small comment adjustments 2021-01-04 21:40:38 +00:00
oli
caeb3d525d Move MIR body loading to a machine function 2021-01-04 21:40:38 +00:00
oli
7202054800 Document all the things 2021-01-04 21:40:38 +00:00
oli
eb4e94b2e5 Simplify the optimize_mir query 2021-01-04 21:40:38 +00:00
oli
1f5fb3e056 Differentiate between the availability of ctfe MIR and runtime MIR 2021-01-04 21:40:38 +00:00
oli
cccd40f9b5 Keep an unoptimized duplicate of const fn around
This allows CTFE to reliably detect UB, as otherwise
optimizations may hide UB.
2021-01-04 21:40:38 +00:00
oli
68ff5f0a18 Stop optimizing promoteds 2021-01-04 21:40:38 +00:00
Rich Kadel
e4aa99fe7a Inlining enabled by -mir-opt-level > 1 is incompatible with coverage
Fixes: #80060

Also adds additional test cases for coverage of doctests.
2021-01-04 11:06:42 -08:00
oli
90b56b94f6 Stylistic fixes to diagnostic messages 2021-01-03 15:11:34 +00:00
oli
d3992f36ad Refactor the non-transient cell borrow error diagnostic 2021-01-03 14:46:49 +00:00
Oli Scherer
8968c8a103
Dangling pointers point to everything and nothing
Co-authored-by: Ralf Jung <post@ralfj.de>
2021-01-03 15:19:07 +01:00
Oli Scherer
e5e4a851c4
Grammar fixes
Co-authored-by: Ralf Jung <post@ralfj.de>
2021-01-03 15:15:23 +01:00
Aaron Hill
21b8f2ecde
Make ExpnData fields krate and orig_id private
These fields are only used by hygiene serialized, and should not be
accessed by anything outside of `rustc_span`.
2021-01-03 08:58:43 -05:00
oli
a137ff1706 Update now-more-precise operation with a preciser message 2021-01-03 13:45:16 +00:00
Oli Scherer
e5330a4f52 Apply suggestions from code review
comment nits

Co-authored-by: Ralf Jung <post@ralfj.de>
2021-01-03 13:32:56 +00:00
Matthias Krüger
e2272cdffc remove redundant closures (clippy::redundant_closure) 2021-01-03 13:34:24 +01:00
oli
aaee3f27ee Don't mix feature gates and hard errors, decide on one per op and stick with it 2021-01-01 16:59:12 +00:00
oli
3ed14033f7 Reinstate the error-code error over the feature gate error 2021-01-01 16:59:12 +00:00
oli
4158e58d79 Enhance some comments 2021-01-01 16:59:12 +00:00
oli
354e510f7d Fix cell checks in const fn 2021-01-01 16:59:12 +00:00
oli
3a44a20ed1 The proper name for the rule is "enclosing scope" 2021-01-01 16:59:12 +00:00
oli
0b841846ba Allow references to interior mutable data behind a feature gate 2021-01-01 16:59:12 +00:00
Ralf Jung
db03b58f23 remove move_val_init leftovers 2020-12-31 10:53:37 +01:00
Mara Bos
46111c1901
Rollup merge of #80458 - RalfJung:promotion-refactor, r=oli-obk
Some Promotion Refactoring

Clean up promotion a bit:
* factor out some common code
* more exhaustive matches

This *should* not break anything... the only potentially-breaking change is that `BorrowKind::Shallow | BorrowKind::Unique` are now rejected for internal references.

r? ``@oli-obk``
2020-12-30 20:56:52 +00:00
Matthias Krüger
bdc9291ed9 where possible, pass slices instead of &Vec or &String (clippy::ptr_arg) 2020-12-30 13:11:52 +01:00
Vadim Petrochenkov
4d2d0bad4e Remove compile-fail test suite 2020-12-29 23:39:56 +03:00