Esteban Küber [Mon, 23 Dec 2019 22:16:34 +0000 (14:16 -0800)]
Suggest type param when encountering `_` in fn defs
When encountering `_` type placeholder in fn arguments and return type,
suggest using generic type parameters.
Expand what counts as an inferable return type to slice, array and
tuples of `_`.
Wesley Wiser [Fri, 27 Dec 2019 19:33:22 +0000 (14:33 -0500)]
Resolve long compile times when evaluating always valid constants
This extends the existing logic which skips validating every integer or
floating point number type to also skip validating empty structs because
they are also trivially valid.
Dylan DPC [Sun, 29 Dec 2019 07:39:56 +0000 (13:09 +0530)]
Rollup merge of #67482 - ldm0:master, r=petrochenkov
Fix outdated comment
Logics in `libsyntax/ext/expand.rs:MacroExpander::expand()` have been moved to `libsyntax_expand/expand.rs:MacroExpander::fully_expand_fragment()`
This pull request fixs the dangling file path.
bors [Sat, 28 Dec 2019 23:02:09 +0000 (23:02 +0000)]
Auto merge of #67334 - estebank:ignore-triple, r=nikomatsakis
Teach `compiletest` to ignore platform triples
The UI tests are written assuming `--remap-path-prefix` is *not used* (`remap-debuginfo` in `config.toml`). The consequence is that the error messages may include paths and snippets into the standard library. When `remap-debuginfo` is enabled, these messages change in format and structure because `rustc` will not show paths and snippets into the standard library.
This normally isn't a problem for the "main" platforms (linux/macos/windows), because the CI infrastructure is set up so that the tests run without `remap-debuginfo`, but the `dist` artifacts are built separately with `remap-debuginfo` enabled. However, some of the lower-tier platforms perform both tests and distribution in a single step with `remap-debuginfo` enabled. This also affects developers and distributors who use `remap-debuginfo`.
To sidestep this problem, we add a way to ignore tests in specific platform triples, and update the overly broad `ignore-x86` rule in affected tests.
During MIR interpretation it may happen that a place containing
uninitialized bytes is copied. This would read the current
representation of these bytes and write it to the destination even
though they must, by definition, not matter to the execution.
This elides that representation change when no bytes are defined in such
a copy, saving some cpu cycles. In such a case, the memory of the target
allocation is not touched at all which also means that sometimes no
physical page backing the memory allocation of the representation needs
to be provided by the OS at all, reducing memory pressure on the system.
Oliver Scherer [Fri, 27 Dec 2019 23:36:14 +0000 (00:36 +0100)]
Rollup merge of #67635 - Mark-Simulacrum:path-doc-unsafe, r=dtolnay
Document safety of Path casting
I would personally feel more comfortable making the relevant (internal anyway) types repr(transparent) and then documenting that we can make these casts because of that, but I believe this is a more minimal PR, so posting it first.
Oliver Scherer [Fri, 27 Dec 2019 23:36:09 +0000 (00:36 +0100)]
Rollup merge of #67629 - kraai:remove-redundant-link-texts, r=steveklabnik
Remove redundant link texts
Most of these links are followed by a parenthesized expression. I think that the redundant link texts were added to prevent interpretation as an inline link. This is unnecessary since the closing square bracket and opening parenthesis are separated by whitespace.
Oliver Scherer [Fri, 27 Dec 2019 23:36:04 +0000 (00:36 +0100)]
Rollup merge of #67602 - xfix:use-issue-none-instead-of-0-in-intrinsics, r=Centril
Use issue = "none" instead of "0" in intrinsics
`issue = "0"` was incorrectly used in https://github.com/rust-lang/rust/commit/9ee16e14c51668309f446636e8960b7fbbac066d, the project is moving away from issue 0
Oliver Scherer [Fri, 27 Dec 2019 23:36:01 +0000 (00:36 +0100)]
Rollup merge of #67588 - Kixunil:nonnull-slice-iter, r=rkruppe
Use NonNull in slice::Iter and slice::IterMut.
`ptr` of `slice::Iter` and `slice::IterMut` can never be null, but this
fact wasn't exploited for layout optimizations. By changing `ptr` from
`*<mutability> T` to `NonNull<T>`, the compiler can now optimize layout
of `Option<Iter<'a, T>>`.
Oliver Scherer [Fri, 27 Dec 2019 23:35:58 +0000 (00:35 +0100)]
Rollup merge of #65244 - seanmonstar:into-future, r=seanmonstar
add IntoFuture trait and support for await
The [async-await RFC](https://rust-lang.github.io/rfcs/2394-async_await.html#the-await-compiler-built-in) mentions being able to `await` anything implementing `IntoFuture`. Somewhere along the way, it was left out.
We only want to return specializations when `Reveal::All` is passed, not
when `Reveal::UserFacing` is. Resolving this fixes several issues with
the `ConstProp`, `SimplifyBranches`, and `Inline` MIR optimization
passes.
bors [Fri, 27 Dec 2019 10:58:57 +0000 (10:58 +0000)]
Auto merge of #67437 - matthew-healy:skip-llvm-rebuild, r=Mark-Simulacrum
Add LLVM `skip-rebuild` option to `x.py`
This PR reimplements parts of @Walther's work from #65848, and closes #65612.
I decided not to implement the [arguments to override this setting](https://github.com/rust-lang/rust/issues/65612#issuecomment-544247546) in this PR. If there's strong feeling that this change shouldn't be merged without the overrides then I'm happy to close this until I've had a chance to add them in. Otherwise I'll aim to submit a second PR with those this weekend.
I'd have liked to have tested the change in `native.rs`, but there didn't seem to be any existing test infrastructure. I ran this a few times manually and it _worked on my machine_ though... 😬
bors [Fri, 27 Dec 2019 07:38:52 +0000 (07:38 +0000)]
Auto merge of #67192 - oli-obk:const_zst_addr, r=RalfJung,varkor
Various const eval and pattern matching ICE fixes
r? @RalfJung
cc @spastorino
This PR does not change existing behaviour anymore and just fixes a bunch of ICEs reachable from user code (sometimes even on stable via obscure union transmutes).
projects / rust.git / log