bors [Thu, 11 Nov 2021 22:00:53 +0000 (22:00 +0000)]
Auto merge of #90489 - jyn514:load-all-extern-crates, r=petrochenkov
rustdoc: Go back to loading all external crates unconditionally
This *continues* to cause regressions. This code will be unnecessary
once access to the resolver happens fully before creating the tyctxt
(#83761), so load all crates unconditionally for now. To minimize churn, this leaves in the code for loading crates selectively.
bors [Thu, 11 Nov 2021 12:07:53 +0000 (12:07 +0000)]
Auto merge of #88798 - sunfishcode:sunfishcode/windows-null-handles, r=joshtriplett
Fix assertion failures in `OwnedHandle` with `windows_subsystem`.
As discussed in #88576, raw handle values in Windows can be null, such
as in `windows_subsystem` mode, or when consoles are detached from a
process. So, don't use `NonNull` to hold them, don't assert that they're
not null, and remove `OwnedHandle`'s `repr(transparent)`. Introduce a
new `HandleOrNull` type, similar to `HandleOrInvalid`, to cover the FFI
use case.
bors [Thu, 11 Nov 2021 09:13:22 +0000 (09:13 +0000)]
Auto merge of #90755 - scottmcm:spec-array-clone, r=jackh726
Specialize array cloning for Copy types
Because after PR 86041, the optimizer no longer load-merges at the LLVM IR level, which might be part of the perf loss. (I'll run perf and see if this makes a difference.)
Also I added a codegen test so this hopefully won't regress in future -- it passes on stable and with my change here, but not on the 2021-11-09 nightly.
Example on current nightly: <https://play.rust-lang.org/?version=nightly&mode=release&edition=2021&gist=1f52d46fb8fc3ca3ac9f097390085ffa>
```rust
type T = u8;
const N: usize = 3;
bors [Thu, 11 Nov 2021 02:52:32 +0000 (02:52 +0000)]
Auto merge of #83846 - torhovland:issue-10971, r=davidtwco
Added the --temps-dir option
Fixes #10971.
The new `--temps-dir` option puts intermediate files in a user-specified directory. This provides a fix for the issue where parallel invocations of rustc would overwrite each other's intermediate files.
No files are kept in the intermediate directory unless `-C save-temps=yes`.
If additional files are specifically requested using `--emit asm,llvm-bc,llvm-ir,obj,metadata,link,dep-info,mir`, these will be put in the output directory rather than the intermediate directory.
This is a backward-compatible change, i.e. if `--temps-dir` is not specified, the behavior is the same as before.
Matthias Krüger [Wed, 10 Nov 2021 22:04:26 +0000 (23:04 +0100)]
Rollup merge of #90736 - Lokathor:inline-asm-docs-updates, r=Amanieu
adjust documented inline-asm register constraints
This change more clearly specifies how `reg` and `reg_thumb` work with ARM, Thumb2, and Thumb1 code.
Based upon the [llvm documentation](https://llvm.org/docs/LangRef.html#supported-constraint-code-list) for register constraint codes.
To be clear, this just updates the docs to match what already happens with rustc/llvm.
No change in the compiler is required to make it match this new documentation.
Matthias Krüger [Wed, 10 Nov 2021 22:04:25 +0000 (23:04 +0100)]
Rollup merge of #89930 - cuviper:avoid-clone3, r=joshtriplett
Only use `clone3` when needed for pidfd
In #89522 we learned that `clone3` is interacting poorly with Gentoo's
`sandbox` tool. We only need that for the unstable pidfd extensions, so
otherwise avoid that and use a normal `fork`.
This is a re-application of beta #89924, now that we're aware that we need
more than just a temporary release fix. I also reverted 12fbabd27f700, as
that was just fallout from using `clone3` instead of `fork`.
Matthias Krüger [Wed, 10 Nov 2021 17:52:28 +0000 (18:52 +0100)]
Rollup merge of #90727 - GuillaumeGomez:remove-potential-useless-search-index-data, r=notriddle,camelid
Remove potential useless data for search index
I uncovered this case when working on https://github.com/rust-lang/rust/pull/90726 to debug https://github.com/rust-lang/rust/pull/90385.
Explanations: if we have a full generic, we check if it has generics then we do the following:
* If it has only one generic, we remove one nested level in order to not keep the "parent" generic (since it has empty name, it's useless after all).
* Otherwise we add it alongside its generics.
However, I didn't handle the case where a generic had no generics. Meaning that we were adding items with empty names in the search index. So basically useless data in the search index.
Matthias Krüger [Wed, 10 Nov 2021 17:52:26 +0000 (18:52 +0100)]
Rollup merge of #88447 - inquisitivecrystal:rustdoc-vis, r=jyn514
Use computed visibility in rustdoc
This PR changes `librustdoc` to use computed visibility instead of syntactic visibility. It was initially part of #88019, but was separated due to concerns that it might cause a regression somewhere we couldn't predict.
Scott McMurray [Wed, 10 Nov 2021 05:43:20 +0000 (21:43 -0800)]
Specialize array cloning for Copy types
Because after PR 86041, the optimizer no longer load-merges at the LLVM IR level, which might be part of the perf loss. (I'll run perf and see if this makes a difference.)
Also I added a codegen test so this hopefully won't regress in future -- it passes on stable and with my change here, but not on the 2021-11-09 nightly.
11 commits in fd9299792852c9a368cb236748781852f75cdac6..5c5dbc5b196c9564422b3193264f3288d2a051ce
2021-10-22 21:59:46 -0400 to 2021-11-09 19:30:43 -0500
- Fix constants link.
- Fix updated anchor
- Propagate edits to chapter 2 back
- Edits to nostarch's chapter 3 edits
- ch 3 from nostarch
- Fix Cargo.toml snippet about custom derive macros
- Snapshot of chapter 9 for nostarch
- Create tmp/src for converting quotes, not sure why this broke but ok
- Update question mark to better explain where it can be used
- Clarify sentence about Results in functions that don't return Result. Fixes rust-lang/book#2912.
- Merge pull request rust-lang/book#2913 from covariant/patch-1
13 commits in b06008731af0f7d07cd0614e820c8276dfed1c18..196ef69aa68f2cef44f37566ee7db37daf00301b
2021-10-21 15:13:09 -0500 to 2021-11-07 07:48:47 -0600
- Fix typo: [upv.rs_mentioned] -> [upvars_mentioned]
- Add note to emphasize replacing TARGET_TRIPLE (rust-lang/rustc-dev-guide#1250)
- Remove some legacy test suites.
- tiny capitalization fix
- Fix date
- Update some date-check comments
- Ensure date-check cron job is using latest stable Rust
- enhance subtree docs, link to clippy docs
- Edit introduction to bootstrapping
- Some minor adjustments to the diagnostic documentation
- Edit "About this guide" for semantic line feeds
- Fix `rustc_mir` related links (rust-lang/rustc-dev-guide#1228)
- Add documentation for LLVM CFI support
## edition-guide
3 commits in 7c0088ca744d293a5f4b1e2ac378e7c23d30fe55..27f4a84d3852e9416cae5861254fa53a825c56bd
2021-10-05 13:28:05 +0200 to 2021-11-08 10:13:20 -0500
- Add a missing period (rust-lang/edition-guide#271)
- Fix syntax error in code example (rust-lang/edition-guide#270)
- Fixed an example error of prelude.md (rust-lang/edition-guide#269)
Matthias Krüger [Wed, 10 Nov 2021 05:02:54 +0000 (06:02 +0100)]
Rollup merge of #90690 - solid-rs:fix-kmc-solid-asm-const, r=Mark-Simulacrum
kmc-solid: Avoid the use of `asm_const`
This PR removes the use of [the now-separated-out `asm_const` compiler feature][1] in `std::sys::solid` to fix the [`*-kmc-solid_*`](https://doc.rust-lang.org/nightly/rustc/platform-support/kmc-solid.html) Tier 3 targets.
bors [Wed, 10 Nov 2021 03:11:44 +0000 (03:11 +0000)]
Auto merge of #90663 - ken-matsui:support-early-stopping-too-old-tidy-for-macos, r=Mark-Simulacrum
Support early stopping too old pre-installed `tidy` command for macOS in the HTML checker
This PR brings early stopping the HTML checker before errors, which leave some macOS users confused, and suggesting installing a newer `tidy` command.
The pre-installed `tidy` command on macOS is too old, released on 31 October 2006. Additionally, I can see the same date at [StackOverflow](https://stackoverflow.com/questions/22283382/overwrite-osx-tidy
) seven years ago. The `tidy` does not support two indispensable options: `--mute-id` and `--mute`. So, the `./x.py test` command fails with a bunch of errors due not to muting them.
I could confirm the `./x.py test` command before installing a newer `tidy` failed and its command after the installation succeeded.
Noah Lev [Sat, 4 Sep 2021 23:07:20 +0000 (16:07 -0700)]
miri: Detect uninitialized integers and floats
Change the Miri engine to allow configuring whether to check
initialization of integers and floats. This allows the Miri tool to
optionally check for initialization if requested by the user.
bors [Tue, 9 Nov 2021 20:09:53 +0000 (20:09 +0000)]
Auto merge of #90734 - matthiaskrgr:rollup-e1euotp, r=matthiaskrgr
Rollup of 7 pull requests
Successful merges:
- #89561 (Type inference for inline consts)
- #90035 (implement rfc-2528 type_changing-struct-update)
- #90613 (Allow to run a specific rustdoc-js* test)
- #90683 (Make `compiler-docs` only control the default instead of being a hard off-switch)
- #90685 (x.py: remove fixme by deleting code)
- #90701 (Record more artifact sizes during self-profiling.)
- #90723 (Better document `Box` and `alloc::alloc::box_free` connection)
Matthias Krüger [Tue, 9 Nov 2021 18:00:46 +0000 (19:00 +0100)]
Rollup merge of #90723 - asquared31415:box_docs, r=jyn514
Better document `Box` and `alloc::alloc::box_free` connection
The internal `alloc::alloc::box_free` function requires that its signature matches the `owned_box` struct's declaration, but previously that connection was only documented on the `box_free` function.
This PR makes the documentation two-way to help anyone making theoretical changes to `Box` to see the connection, since changes are more likely to originate from `Box`.
Currently the identifiers emitted for these are hard-coded as string literals. Is it worth adding constants to https://github.com/rust-lang/measureme/blob/master/measureme/src/rustc.rs instead? We don't do that for query names and the like -- but artifact kinds might be more stable than query names.
Matthias Krüger [Tue, 9 Nov 2021 18:00:41 +0000 (19:00 +0100)]
Rollup merge of #90035 - SparrowLii:rfc2528, r=jackh726
implement rfc-2528 type_changing-struct-update
This PR implement rfc2528-type_changing-struct-update.
The main change process is as follows:
1. Move the processing part of `base_expr` into `check_expr_struct_fields` to avoid returning `remaining_fields` (a relatively complex hash table)
2. Before performing the type consistency check(`check_expr_has_type_or_error`), if the `type_changing_struct_update` feature is set, enter a different processing flow, otherwise keep the original flow
3. In the case of the same structure definition, check each field in `remaining_fields`. If the field in `base_expr` is not the suptype of the field in `adt_ty`, an error(`FeildMisMatch`) will be reported.
The MIR part does not need to be changed, because only the items contained in `remaining_fields` will be extracted from `base_expr` when MIR is generated. This means that fields with different types in `base_expr` will not be used
Updates #86618
cc `@nikomatsakis`
Perform type checking/inference of inline consts in the same context as the outer def, similar to what is currently done to closure.
Doing so would require `closure_base_def_id` of the inline const to return the outer def, and since `closure_base_def_id` can be called on non-local crate (and thus have no HIR available), a new `DefKind` is created for inline consts.
The type of the generated anon const can capture lifetime of outer def, so we couldn't just use the typeck result as the type of the inline const's def. Closure has a similar issue, and it uses extra type params `CK, CS, U` to capture closure kind, input/output signature and upvars. I use a similar approach for inline consts, letting it have an extra type param `R`, and then `typeof(InlineConst<[paremt generics], R>)` would just be `R`. In borrowck region requirements are also propagated to the outer MIR body just like it's currently done for closure.
With this PR, inline consts in expression position are quitely usable now; however the usage in pattern position is still incomplete -- since those does not remain in the MIR borrowck couldn't verify the lifetime there. I have left an ignored test as a FIXME.
Some disucssions can be found on [this Zulip thread](https://rust-lang.zulipchat.com/#narrow/stream/260443-project-const-generics/topic/inline.20consts.20typeck).
cc `````@spastorino````` `````@lcnr`````
r? `````@nikomatsakis`````
bors [Tue, 9 Nov 2021 14:12:54 +0000 (14:12 +0000)]
Auto merge of #90724 - JohnTitor:rollup-zg0kbm3, r=JohnTitor
Rollup of 6 pull requests
Successful merges:
- #87530 (Add comments regarding superfluous `!Sync` impls)
- #90591 (treat illumos like solaris in failing ui tests which need it)
- #90678 (Add some GATs-related regression tests)
- #90688 (enable `dotprod` target feature in arm)
- #90708 (Add a note about feature(explicit_generic_args_with_impl_trait) to the relevant error message)
- #90720 (Update cargo)
Yuki Okushi [Tue, 9 Nov 2021 13:02:26 +0000 (22:02 +0900)]
Rollup merge of #90720 - ehuss:update-cargo, r=ehuss
Update cargo
4 commits in 94ca096afbf25f670e76e07dca754fcfe27134be..2e2a16e983f597da62bc132eb191bc3276d4b1bb
2021-10-29 14:45:06 +0000 to 2021-11-08 15:13:38 +0000
- Fix debug panic on download with redirect body. (rust-lang/cargo#10048)
- no need to clone (rust-lang/cargo#10051)
- Update curl. (rust-lang/cargo#10040)
- Fix --scrape-examples-target-crate using package name (with dashes) instead of crate name (with underscores) (rust-lang/cargo#10037)
bors [Tue, 9 Nov 2021 08:21:10 +0000 (08:21 +0000)]
Auto merge of #87337 - jyn514:lint-error, r=oli-obk,flip1995
Don't abort compilation after giving a lint error
The only reason to use `abort_if_errors` is when the program is so broken that either:
1. later passes get confused and ICE
2. any diagnostics from later passes would be noise
This is never the case for lints, because the compiler has to be able to deal with `allow`-ed lints.
So it can continue to lint and compile even if there are lint errors.
Closes https://github.com/rust-lang/rust/issues/82761. This is a WIP because I have a feeling it will exit with 0 even if there were lint errors; I don't have a computer that can build rustc locally at the moment.
bors [Tue, 9 Nov 2021 05:33:16 +0000 (05:33 +0000)]
Auto merge of #90485 - camsteffen:fmt-args-less-bind, r=m-ou-se
Don't destructure args tuple in format_args!
This allows Clippy to parse the HIR more simply since `arg0` is changed to `_args.0`. (cc rust-lang/rust-clippy#7843). From rustc's perspective, I think this is something between a lateral move and a tiny improvement since there are fewer bindings.
bors [Mon, 8 Nov 2021 21:37:59 +0000 (21:37 +0000)]
Auto merge of #90443 - camelid:rustdoc-subst, r=GuillaumeGomez
Merge `DocContext.{ty,lt,ct}_substs` into one map
It should be impossible to have more than one entry with a particular
key across the three maps, so they should be one map. In addition to
making it impossible for multiple entries to exist, this should improve
memory usage since now only one map is allocated on the stack and heap.
Guillaume Gomez [Mon, 8 Nov 2021 14:15:24 +0000 (15:15 +0100)]
Rollup merge of #90657 - GuillaumeGomez:one-char-last-line-removed, r=jyn514
Fix bug with `#[doc]` string single-character last lines
Fixes #90618.
This is because `.iter().all(|c| c == '*')` returns `true` if there is no character checked. And in case the last line has only one character, it simply returns `true`, making the last line behind removed.
bors [Mon, 8 Nov 2021 13:38:08 +0000 (13:38 +0000)]
Auto merge of #90361 - Mark-Simulacrum:always-verify, r=michaelwoerister
Enable verification for 1/32th of queries loaded from disk
This is a limited enabling of incremental verification for query results loaded from disk, which previously did not run without -Zincremental-verify-ich. If enabled for all queries, we see a probably unacceptable hit of ~50% in the worst case, so this pairs back the verification to a more limited set based on the hash key.
Per collected [perf results](https://github.com/rust-lang/rust/pull/84227#issuecomment-953350582), this is a regression of at most 7% on coercions opt incr-unchanged, and typically less than 0.5% on other benchmarks (largely limited to incr-unchanged). I believe this is acceptable performance to land, and we can either ratchet it up or down fairly easily.
We have no real sense of whether this will lead to a large amount of assertions in the wild, but since those assertions may lead to miscompilations today, it seems potentially warranted. We have a good bit of lead time until the next stable release, though the holiday season will also start soon; we may wish to discuss the timing of enabling this and weigh the desire to prevent (possible) miscompilations against assertions.
bors [Mon, 8 Nov 2021 08:21:51 +0000 (08:21 +0000)]
Auto merge of #90675 - camelid:cleanup-impl, r=jyn514
rustdoc: Cleanup `clean::Impl` and other parts of `clean`
This PR cleans up and reduces the size of `clean::Impl`, makes some other small performance improvements, and removes some Clean impls that are either unnecessary or potentially confusing.
Joshua Nelson [Mon, 1 Nov 2021 23:01:39 +0000 (23:01 +0000)]
Go back to loading all external crates unconditionally
This *continues* to cause regressions. This code will be unnecessary
once access to the resolver happens fully before creating the tyctxt
(#83761), so load all crates unconditionally for now.
Joshua Nelson [Mon, 8 Nov 2021 01:45:10 +0000 (01:45 +0000)]
Add more missing methods to `IntraLinkCrateLoader`
This helps with (but does not fix)
https://github.com/rust-lang/rust/issues/84738. I tested on
https://github.com/jyn514/objr/commit/edcee7b8124abf0e4c63873e8422ff81beb11ebb
and still hit ICEs.
Joshua Nelson [Thu, 16 Sep 2021 05:35:23 +0000 (05:35 +0000)]
Use `has_errors_or_lint_errors` in rustdoc instead of `abort_if_errors()`
Rustdoc is special as usual and doesn't go through RunCompiler, so it
needs its own explicit checks. The rest of the tools go through
RunCompiler, so they should be fine.
Joshua Nelson [Wed, 21 Jul 2021 03:23:22 +0000 (23:23 -0400)]
Don't abort compilation after giving a lint error
The only reason to use `abort_if_errors` is when the program is so broken that either:
1. later passes get confused and ICE
2. any diagnostics from later passes would be noise
This is never the case for lints, because the compiler has to be able to deal with `allow`-ed lints.
So it can continue to lint and compile even if there are lint errors.