Commit Graph

230026 Commits

Author SHA1 Message Date
Moulins
76c49aead6 support non-null pointer niches in CTFE 2023-07-21 03:31:45 +02:00
Moulins
3c05276866 restrict the valid range of references if -Z reference-niches is set
Note that this doesn't actually work at all, as many places in rustc
assume that references only have the null niche.
2023-07-21 03:31:45 +02:00
Moulins
8b847ef734 add crate-local -Z reference_niches unstable flag (does nothing for now) 2023-07-21 03:31:45 +02:00
Moulins
30ae640a3c properly handle arrays and wide pointers in naive_layout_of 2023-07-21 03:31:45 +02:00
Moulins
cb8b1d1bc9 add naive_layout_of query 2023-07-21 03:31:45 +02:00
bors
e2a7ba2771 Auto merge of #113858 - cjgillot:const-prop-pairs, r=oli-obk
Always const-prop scalars and scalar pairs

This removes some complexity from the pass.

The limitation to propagate ScalarPairs only for tuple comes from https://github.com/rust-lang/rust/pull/67015, when ScalarPair constant were modeled using `Rvalue::Aggregate`. Nowadays, we use `ConstValue::ByRef`, which does not care about the underlying type.

The justification for not propagating in all cases was perf. This seems not to be a clear cut any more: https://github.com/rust-lang/rust/pull/113858#issuecomment-1642396746
2023-07-20 22:22:31 +00:00
Camille GILLOT
4a177406ee Inline should_const_prop. 2023-07-20 21:30:51 +00:00
Camille GILLOT
3f708add2f Remove visit_terminator. 2023-07-20 21:30:51 +00:00
Camille GILLOT
ccfa9af29d Propagate ScalarPair for any type. 2023-07-20 21:30:51 +00:00
Camille GILLOT
895e2159f8 Also propagate ScalarPair operands. 2023-07-20 21:30:51 +00:00
Camille GILLOT
12a2edd149 Always propagate into operands. 2023-07-20 21:30:51 +00:00
bors
399b068235 Auto merge of #113856 - WaffleLapkin:vtablin', r=oli-obk
Refactor vtable encoding and optimize it for the case of multiple marker traits

This PR does two things
- Refactor `prepare_vtable_segments` (this was motivated by the other change, `prepare_vtable_segments` was quite hard to understand and while trying to edit it I've refactored it)
  - Mostly remove `loop`s labeled `break`s/`continue`s whenever there is a simpler solution
  - Also use `?`
- Make vtable format a bit more efficient wrt to marker traits
  - See the tests for an example

Fixes https://github.com/rust-lang/rust/issues/113840
cc `@crlf0710`

----

Review wise it's probably best to review each commit individually, as then it's more clear why the refactoring is correct.

I can split the last two commits (which change behavior) into a separate PR if it makes reviewing easier
2023-07-20 20:34:06 +00:00
bors
1554942cdc Auto merge of #113546 - cjgillot:unused-query, r=compiler-errors
Querify unused trait check.

This code transitively loads information for all bodies, and from resolutions. As it does not return a value, it should be beneficial to have it as a query.
2023-07-20 18:45:09 +00:00
bors
092e4f46be Auto merge of #113890 - matthiaskrgr:rollup-k1w2vii, r=matthiaskrgr
Rollup of 7 pull requests

Successful merges:

 - #110765 (rustdoc: fix position of `default` in method rendering)
 - #113529 (Permit pre-evaluated constants in simd_shuffle)
 - #113800 (Avoid another gha group nesting)
 - #113827 (Add Foreign, Never, FnDef, Closure and Generator tys to SMIR)
 - #113835 (new solver: don't consider blanket impls multiple times)
 - #113883 (Remove outdated Firefox-specific CSS for search's crate selector appearance)
 - #113884 (Don't translate compiler-internal bug messages)

r? `@ghost`
`@rustbot` modify labels: rollup
2023-07-20 16:57:05 +00:00
Matthias Krüger
464e02a267
Rollup merge of #113884 - oli-obk:delay_span_bug_detrans_late, r=davidtwco
Don't translate compiler-internal bug messages

These are not very useful to be translated, as

* translators would get really weird and bad english versions to start out from,
* compiler devs have to do some work for what is supposed to be dead code and just a sanity check,
* the target audience is other compiler devs.

r? `@davidtwco`
2023-07-20 17:19:34 +02:00
Matthias Krüger
8ac957aa03
Rollup merge of #113883 - steffahn:rustdoc-search-crate-selector-padding, r=GuillaumeGomez
Remove outdated Firefox-specific CSS for search's crate selector appearance

Remove adjustments that used to be necessary for search's crate selector appearance (padding) to look identical on Firefox. New versions of Firefox appear to have changed behavior to agree with Chrome.

As briefly discussed in https://github.com/rust-lang/rust/pull/98855#issuecomment-1624098112

r? ``@GuillaumeGomez``
2023-07-20 17:19:34 +02:00
Matthias Krüger
add8298aff
Rollup merge of #113835 - lcnr:assemble-candidates-considering-self-ty, r=compiler-errors
new solver: don't consider blanket impls multiple times

only consider candidates which rely on the self type in `assemble_candidates_after_normalizing_self_ty`.

r? ``@compiler-errors``
2023-07-20 17:19:33 +02:00
Matthias Krüger
1a748f0460
Rollup merge of #113827 - spastorino:smir-types-4, r=oli-obk
Add Foreign, Never, FnDef, Closure and Generator tys to SMIR

r? ``@oli-obk``
2023-07-20 17:19:33 +02:00
Matthias Krüger
6102785a29
Rollup merge of #113800 - oli-obk:gha_ci_cycle, r=jyn514
Avoid another gha group nesting

fixes https://github.com/rust-lang/rust/issues/113798 (`x test error_index_generator` did not work locally anymore)

r? ``@jyn514``
2023-07-20 17:19:32 +02:00
Matthias Krüger
8c17e0701e
Rollup merge of #113529 - oli-obk:simd_shuffle_evaluated, r=wesleywiser
Permit pre-evaluated constants in simd_shuffle

fixes https://github.com/rust-lang/rust/issues/113500
2023-07-20 17:19:32 +02:00
Matthias Krüger
80f749dfa0
Rollup merge of #110765 - wackbyte:fix-defaultness-position, r=fmease,GuillaumeGomez
rustdoc: fix position of `default` in method rendering

With the following code:
```rs
#![feature(specialization)]

pub trait A {
    unsafe fn a();
}

impl A for () {
    default unsafe fn a() {}
}
```
rustdoc would render the `impl` of `a` as
```rs
unsafe default fn a()
```
which is inconsistent with the actual position of `default`.
This PR fixes this issue.
2023-07-20 17:19:32 +02:00
bors
06a53ddc0b Auto merge of #113758 - cjgillot:move-dse, r=JakobDegen,oli-obk
Turn copy into moves during DSE.

Dead store elimination computes whether removing a direct store to an unborrowed place is allowed.
Where removing a store is allowed, writing `uninit` is too.

This means that we can use this pass to transform `copy` operands into `move` operands. This is only interesting in call terminators, so we only handle those.

Special care is taken for the `use_both(_1, _1)` case:
- moving the second argument is ok, as `_1` is not live after the call;
- moving the first argument is not, as the second argument reads `_1`.

Fixes #75993
Fixes https://github.com/rust-lang/rust/issues/108068

r? `@RalfJung`
cc `@JakobDegen`
2023-07-20 15:05:39 +00:00
bors
6b53175b5d Auto merge of #113861 - ibraheemdev:mpsc-tls-bug, r=Mark-Simulacrum
Avoid tls access while iterating through mpsc thread entries

Upstream fix: https://github.com/crossbeam-rs/crossbeam/pull/802. Possibly fixes https://github.com/rust-lang/rust/issues/113726.
2023-07-20 12:30:12 +00:00
lcnr
5c75bc5317 update doc comments 2023-07-20 12:01:34 +02:00
lcnr
2062f2ca82 review 2023-07-20 12:01:34 +02:00
Oli Scherer
d97ec97b94 Don't translate compiler-internal bug messages 2023-07-20 09:51:47 +00:00
Oli Scherer
7de9b654b7 Avoid another gha group nesting 2023-07-20 09:42:53 +00:00
lcnr
7c97a76b76 re-add comment 2023-07-20 11:05:52 +02:00
lcnr
aa28b77b5a add FIXME 2023-07-20 11:05:52 +02:00
lcnr
2d99f40ec5 assembly: only consider blanket impls once 2023-07-20 11:05:52 +02:00
lcnr
fdaec57a28 XSimplifiedType to SimplifiedType::X 2023-07-20 11:05:52 +02:00
bors
b14fd2359f Auto merge of #113695 - bjorn3:fix_rlib_cdylib_metadata_handling, r=pnkfelix,petrochenkov
Verify that all crate sources are in sync

This ensures that rustc will not attempt to link against a cdylib as if it is a rust dylib when an rlib for the same crate is available. Previously rustc didn't actually check if any further formats of a crate which has been loaded are of the same version and if they are actually valid. This caused a cdylib to be interpreted as rust dylib as soon as the corresponding rlib was loaded. As cdylibs don't export any rust symbols, linking would fail if rustc decides to link against the cdylib rather than the rlib.

Two crates depended on the previous behavior by separately compiling a test crate as both rlib and dylib. These have been changed to capture their original spirit to the best of my ability while still working when rustc verifies that all crates are in sync. It is unlikely that build systems depend on the current behavior and in any case we are taking a lot of measures to ensure that any change to either the source or the compilation options (including crate type) results in rustc rejecting it as incompatible. We merely didn't do this check here for now obsolete perf reasons.

Fixes https://github.com/rust-lang/rust/issues/10786
Fixes https://github.com/rust-lang/rust/issues/82151
Fixes https://github.com/rust-lang/rust/issues/82972
Closes https://github.com/bevy-cheatbook/bevy-cheatbook/issues/114
2023-07-20 09:00:10 +00:00
Oli Scherer
c7428d5052 Monomorphize constants before inspecting them 2023-07-20 08:53:09 +00:00
Frank Steffahn
2009b4a5cc Remove adjustments that used to be necessary for search's crate selector appearance (padding)
to look identical on Firefox. New versions of Firefox appear to have changed behavior to agree with Chrome.
2023-07-20 17:46:32 +09:00
bors
c67cb3e577 Auto merge of #113878 - matthiaskrgr:rollup-u0d3kzx, r=matthiaskrgr
Rollup of 5 pull requests

Successful merges:

 - #113710 (Fix rpath for libdir is specified)
 - #113787 (Update documentation for std::process::Command's new method)
 - #113795 (Properly document `lifetime_mapping` in `OpaqueTy`)
 - #113857 (Add tests for `--document-hidden-items` option)
 - #113871 (Use the correct span for displaying the line following a derive sugge…)

r? `@ghost`
`@rustbot` modify labels: rollup
2023-07-20 07:12:22 +00:00
Matthias Krüger
770c8d0667
Rollup merge of #113871 - clubby789:derive-sugg-span, r=compiler-errors
Use the correct span for displaying the line following a derive sugge…

`span` here is the main span of the diagnostic. In the linked issue's case, this belongs to `main.rs`. However, the line numbers (and line we are trying to display) are in `name.rs`, so using `span_to_lines` gives us the wrong `FileLines`.

Use `parts[0].span` (the span of the suggestion) here like the rest of the code does to get the right file.

Not sure if this needs a dedicated test because this fixes an existing error in the UI suite

Fixes #113844
2023-07-20 07:08:43 +02:00
Matthias Krüger
53580da72d
Rollup merge of #113857 - GuillaumeGomez:document-hidden-items-test, r=notriddle
Add tests for `--document-hidden-items` option

Since `--document-hidden-items` was greatly fixed/improved in https://github.com/rust-lang/rust/pull/113574, thought it might be worth adding some more tests for it to prevent new regressions.

As for the first commit, it allows to go from:

```
Traceback (most recent call last):
  File "/home/imperio/rust/rust/src/etc/htmldocck.py", line 706, in <module>
    check(sys.argv[1], get_commands(rust_test_path))
  File "/home/imperio/rust/rust/src/etc/htmldocck.py", line 689, in check
    for c in commands:
  File "/home/imperio/rust/rust/src/etc/htmldocck.py", line 274, in get_commands
    args = shlex.split(args)
  File "/usr/lib/python3.10/shlex.py", line 315, in split
    return list(lex)
  File "/usr/lib/python3.10/shlex.py", line 300, in __next__
    token = self.get_token()
  File "/usr/lib/python3.10/shlex.py", line 109, in get_token
    raw = self.read_token()
  File "/usr/lib/python3.10/shlex.py", line 191, in read_token
    raise ValueError("No closing quotation")
ValueError: No closing quotation
```

to:

```
Traceback (most recent call last):
  File "/home/imperio/rust/rust/src/etc/htmldocck.py", line 708, in <module>
    check(sys.argv[1], get_commands(rust_test_path))
  File "/home/imperio/rust/rust/src/etc/htmldocck.py", line 691, in check
    for c in commands:
  File "/home/imperio/rust/rust/src/etc/htmldocck.py", line 278, in get_commands
    raise Exception("line {}: {}".format(lineno + 1, exc)) from None
Exception: line 57: No closing quotation
```

Having the line where the error occurred is quite useful.

r? `@notriddle`
2023-07-20 07:08:43 +02:00
Matthias Krüger
b128750bc4
Rollup merge of #113795 - compiler-errors:doc, r=spastorino
Properly document `lifetime_mapping` in `OpaqueTy`

Also use an `Option` to signify that the value is actually present, instead of just no captured lifetimes.
2023-07-20 07:08:42 +02:00
Matthias Krüger
538dcdad31
Rollup merge of #113787 - sanchopanca:process-command-windows-docs, r=ChrisDenton
Update documentation for std::process::Command's new method

In the current documentation, it's not specified that when creating a Command, the .exe extension can be omitted for Windows executables. However, for other types of executable files like .bat or .cmd, the complete filename including the extension must be provided.

I encountered it by noticing that `Command::new("wt").spawn().unwrap()` succeeds on my machine while `Command::new("code").spawn().unwrap()` panics. Turns out VS Code's entrypoint is .cmd file.

`resolve_exe` method mentions this behaviour in [a comment](e7fda447e7/library/std/src/sys/windows/process.rs (L425)), but it makes sense to mention it at a more visible place.

I've added this clarification to the documentation, which should make it more accurate and helpful for Rust developers working on the Windows platform.
2023-07-20 07:08:42 +02:00
Matthias Krüger
8027dd95cf
Rollup merge of #113710 - loongarch-rs:fix-rpath, r=clubby789
Fix rpath for libdir is specified

## What does this PR try to resolve?

When building the Rust toolchain with `--libdir=lib64`, the executable tools such as `rustc` cannot find shared libraries.

```bash
./configure --prefix=/ --libdir=lib64
DESTDIR=/tmp/rust ./x.py install
```

```
$ /tmp/rust/bin/rustc
rustc: error while loading shared libraries: librustc_driver-13f1fd1bc7f7000d.so: cannot open shared object file: No such file or directory
```

This issue is caused by the link args `-Wl,rpath` being different from `--libdir`.

```
$ readelf -d /tmp/rust/bin/rustc | grep RUNPATH
 0x000000000000001d (RUNPATH)            Library runpath: [$ORIGIN/../lib]
```

## How to resolve?

When setting the rpath, get it from sysroot libdir relative path.

After this patch:

```
$ readelf -d /tmp/rust/bin/rustc | grep RUNPATH
 0x000000000000001d (RUNPATH)            Library runpath: [$ORIGIN/../lib64]
```
2023-07-20 07:08:41 +02:00
bors
0646a5d1aa Auto merge of #113622 - RickleAndMortimer:issue-113184-fix, r=oli-obk
add links to query documentation for E0391

This PR adds links to https://rustc-dev-guide.rust-lang.org/overview.html#queries and https://rustc-dev-guide.rust-lang.org/query.html for the rustc --explain E0391 and within the compiler error itself.

Fixes: #113184
2023-07-20 03:18:41 +00:00
bors
a6cdd81eff Auto merge of #108714 - estebank:ice_dump, r=oli-obk
On nightly, dump ICE backtraces to disk

Implement rust-lang/compiler-team#578.

When an ICE is encountered on nightly releases, the new rustc panic handler will also write the contents of the backtrace to disk. If any `delay_span_bug`s are encountered, their backtrace is also added to the file. The platform and rustc version will also be collected.

<img width="1032" alt="Screenshot 2023-03-03 at 2 13 25 PM" src="https://user-images.githubusercontent.com/1606434/222842420-8e039740-4042-4563-b31d-599677171acf.png">

The current behavior will *always* write to disk on nightly builds, regardless of whether the backtrace is printed to the terminal, unless the environment variable `RUSTC_ICE_DISK_DUMP` is set to `0`. This is a compromise and can be changed.
2023-07-20 01:29:17 +00:00
clubby789
20a3b9a215 Use the correct span for displaying the line following a derive suggestion 2023-07-19 23:48:43 +00:00
bors
39f42ad9e8 Auto merge of #113865 - Dylan-DPC:rollup-pt960bk, r=Dylan-DPC
Rollup of 7 pull requests

Successful merges:

 - #113444 (add tests for alias bound preference)
 - #113716 (Add the `no-builtins` attribute to functions when `no_builtins` is applied at the crate level.)
 - #113754 (Simplify native_libs query)
 - #113765 (Make it clearer that edition functions are `>=`, not `==`)
 - #113774 (Improve error message when closing bracket interpreted as formatting fill character)
 - #113785 (Fix invalid display of inlined re-export when both local and foreign items are inlined)
 - #113803 (Fix inline_const with interpolated block)

r? `@ghost`
`@rustbot` modify labels: rollup
2023-07-19 18:17:59 +00:00
Dylan DPC
6c3cbcd333
Rollup merge of #113803 - compiler-errors:const-interp-block, r=fee1-dead
Fix inline_const with interpolated block

Interpolation already worked when we had a `const $block` that wasn't a statement expr:

```
fn foo() {
  let _ = const $block;
}
```

But it was failing when the const block was in statement expr position:

```
fn foo() {
  const $block;
}
```

... because of a bug in a check for const items. This fixes that.

---

cc https://github.com/rust-lang/rust/pull/112953#issuecomment-1631354481, though I don't think this requires an FCP since it's already supported in exprs and seems to me to be fully a parser bug.
2023-07-19 22:37:08 +05:30
Dylan DPC
c7c89140b0
Rollup merge of #113785 - GuillaumeGomez:tests/rustdoc/issue-105735-fix, r=notriddle,aDotInTheVoid
Fix invalid display of inlined re-export when both local and foreign items are inlined

Fixes #105735.

The bug is actually quite interesting: at the `clean` pass, local inlined items have their `use` item removed, however foreign items don't have their `use` item removed because it's in the `clean` pass that we handle them. So when a `use` inlines both a local and a foreign item, it will work as expected for the foreign one, but not for the local as its `use` should not be around anymore.

To prevent this, I created a new `inlined_foreigns` field into the `Module` struct to allow to remove the `use` item early on for foreign items as well. Then we iterate it in the `clean` pass directly.

r? ``@notriddle``
2023-07-19 22:37:08 +05:30
Dylan DPC
444ac1a6df
Rollup merge of #113774 - compiler-errors:fill-expr-bracket, r=eholk
Improve error message when closing bracket interpreted as formatting fill character

Fixes #112732 by explaining why it's erroring in the way it is.
2023-07-19 22:37:07 +05:30
Dylan DPC
a47b7b013f
Rollup merge of #113765 - compiler-errors:at-least, r=oli-obk
Make it clearer that edition functions are `>=`, not `==`

r? `@Nilstrieb`

We could also perhaps derive `Ord` on `Edition` and use comparison operators.
2023-07-19 22:37:07 +05:30
Dylan DPC
dbb6b1ac31
Rollup merge of #113754 - cjgillot:simplify-foreign, r=petrochenkov
Simplify native_libs query

Drive-by cleanup I saw while implementing https://github.com/rust-lang/rust/pull/113734
2023-07-19 22:37:07 +05:30
Dylan DPC
c1d6d322f4
Rollup merge of #113716 - DianQK:add-no_builtins-to-function, r=pnkfelix
Add the `no-builtins` attribute to functions when `no_builtins` is applied at the crate level.

**When `no_builtins` is applied at the crate level, we should add the `no-builtins` attribute to each function to ensure it takes effect in LTO.**

This is also the reason why no_builtins does not take effect in LTO as mentioned in #35540.

Now, `#![no_builtins]` should be similar to `-fno-builtin` in clang/gcc, see https://clang.godbolt.org/z/z4j6Wsod5.

Next, we should make `#![no_builtins]` participate in LTO again. That makes sense, as LTO also takes into consideration function-level instruction optimizations, such as the MachineOutliner. More importantly, when a user writes a large `#![no_builtins]` crate, they would like this crate to participate in LTO as well.

We should also add a function-level no_builtins attribute to allow users to have more control over it. This is similar to Clang's `__attribute__((no_builtin))` feature, see https://clang.godbolt.org/z/Wod6KK6eq. Before implementing this feature, maybe we should discuss whether to support more fine-grained control, such as `__attribute__((no_builtin("memcpy")))`.

Related discussions:
- #109821
- #35540

Next (a separate pull request?):
- [ ] Revert #35637
- [ ] Add a function-level `no_builtin` attribute?
2023-07-19 22:37:06 +05:30