bors [Fri, 6 Dec 2019 23:35:00 +0000 (23:35 +0000)]
Auto merge of #67104 - Centril:rollup-07vahh9, r=Centril
Rollup of 10 pull requests
Successful merges:
- #66606 (Add feature gate for mut refs in const fn)
- #66841 (Add `{f32,f64}::approx_unchecked_to<Int>` unsafe methods)
- #67009 (Emit coercion suggestions in more places)
- #67052 (Ditch `parse_in_attr`)
- #67071 (Do not ICE on closure typeck)
- #67078 (accept union inside enum if not followed by identifier)
- #67090 (Change "either" to "any" in Layout::from_size_align's docs)
- #67092 (Fix comment typos in src/libcore/alloc.rs)
- #67094 (get rid of __ in field names)
- #67102 (Add note to src/ci/docker/README.md about multiple docker images)
Failed merges:
- #67101 (use `#[allow(unused_attributes)]` to paper over incr.comp problem)
Rollup merge of #67102 - Aaron1011:patch-3, r=Mark-Simulacrum
Add note to src/ci/docker/README.md about multiple docker images
I spent a while debugging a strage linker error about an outdated `glibc` version, only to discover that it was caused by a stale `obj` directory. It wasn't obviously to be that using the same obj dir with multiple Docker images (for the same target triple) could be a problem.
This commit adds a note to the README, which should hopefully be helpful to anyone else who runs into this issue.
Rollup merge of #67009 - Aaron1011:fix/coerce-suggestion, r=Centril
Emit coercion suggestions in more places
Fixes #66910
We have several different kinds of suggestions we can try to make when
type coercion fails. However, we were previously only emitting these
suggestions from `demand_coerce_diag`. This resulted in the compiler
failing to emit applicable suggestions in several different cases, such
as when the implicit return value of a function had the wrong type.
This commit adds a new `emit_coerce_suggestions` method, which tries to
emit a number of related suggestions. This method is called from both
`demand_coerce_diag` and `CoerceMany::coerce_inner`, which covers a much
wider range of cases than before.
We now suggest using `.await` in more cases where it is applicable,
among other improvements.
I'm not happy about disabling the `issue-59756`, but from what I can tell, the suggestion infrastructure in rustc lacks any way of indicating mutually exclusive suggestions (and compiletest lacks a way to only apply a subset of available suggestions).
As discussed in https://github.com/rust-lang/rust/issues/10184
Currently, casting a floating point number to an integer with `as` is Undefined Behavior if the value is out of range. `-Z saturating-float-casts` fixes this soundness hole by making `as` “saturate” to the maximum or minimum value of the integer type (or zero for `NaN`), but has measurable negative performance impact in some benchmarks. There is some consensus in that thread for enabling saturation by default anyway, but provide an `unsafe fn` alternative for users who know through some other mean that their values are in range.
<del>The “fit” wording is copied from https://llvm.org/docs/LangRef.html#fptoui-to-instruction, but I’m not certain what it means exactly. Presumably this is after rounding towards zero, and the doc-test with `i8::MIN` seems to confirm this.</del> Clang presumably uses those LLVM intrinsics to implement C and C++ casts, whose respective standard specify that the value *after truncating to keep its integral part* must be representable in the target type.
Aaron Hill [Fri, 6 Dec 2019 21:38:57 +0000 (16:38 -0500)]
Add note to src/ci/docker/README.md about multiple docker images
I spent a while debugging a strage linker error about an outdated `glibc` version, only to discover that it was caused by a stale `obj` directory. It wasn't obviously to be that using the same obj dir with multiple Docker images (for the same target triple) could be a problem.
This commit adds a note to the README, which should hopefully be helpful to anyone else who runs into this issue.
bors [Fri, 6 Dec 2019 19:14:51 +0000 (19:14 +0000)]
Auto merge of #65195 - varkor:to_option, r=Centril
Rename `bool::then_*` to `bool::to_option_*` and use where appropriate
Name change following https://github.com/rust-lang/rfcs/pull/2757. Also try it out throughout the compiler in places I think makes the code more readable.
Yuki Okushi [Fri, 6 Dec 2019 15:10:05 +0000 (00:10 +0900)]
Rollup merge of #67054 - RalfJung:set-discriminant-unreachable, r=oli-obk
codegen "unreachable" for invalid SetDiscriminant
Follow-up from https://github.com/rust-lang/rust/pull/66960. I also realized I don't understand our policy for using `abort` vs `unreachable`. AFAIK `abort` is safe to call and just aborts the process, while `unreachable` is UB. But sometimes we use both, like here
Yuki Okushi [Fri, 6 Dec 2019 15:10:02 +0000 (00:10 +0900)]
Rollup merge of #67033 - cuviper:ValueName2, r=rkruppe
Migrate to LLVM{Get,Set}ValueName2
The deprecated `LLVM{Get,Set}ValueName` only work with NUL-terminated
strings, but the `2` variants use explicit lengths, which fits better
with Rust strings and slices. We now use these in new helper functions
`llvm::{get,set}_value_name` that convert to/from `&[u8]`.
Yuki Okushi [Fri, 6 Dec 2019 15:10:00 +0000 (00:10 +0900)]
Rollup merge of #67023 - jethrogb:jb/bootstrap-target-linker, r=alexcrichton
SGX: Fix target linker used by bootstrap
Bootstrap, for some reason, overrides the target linker. This is not correct for x86_64-fortanix-unknown-sgx. Add such targets to the list of exceptions.
Before going any further (the issue seems to be linked to metadata as far as I can tell). Do you think this is the good place to do it or should it be done before?
As discussed in https://github.com/rust-lang/rust/issues/10184
Currently, casting a floating point number to an integer with `as` is Undefined Behavior if the value is out of range. `-Z saturating-float-casts` fixes this soundness hole by making `as` “saturate” to the maximum or minimum value of the integer type (or zero for `NaN`), but has measurable negative performance impact in some benchmarks. There is some consensus in that thread for enabling saturation by default anyway, but provide an `unsafe fn` alternative for users who know through some other mean that their values are in range.
bors [Fri, 6 Dec 2019 07:43:56 +0000 (07:43 +0000)]
Auto merge of #67080 - JohnTitor:rollup-2t6fm3u, r=JohnTitor
Rollup of 10 pull requests
Successful merges:
- #66649 (VxWorks: fix issues in accessing environment variables)
- #66764 (Tweak wording of `collect()` on bad target type)
- #66900 (Clean up error codes)
- #66974 ([CI] fix the `! isCI` check in src/ci/run.sh)
- #66979 (Add long error for E0631 and update ui tests.)
- #67017 (cleanup long error explanations)
- #67021 (Fix docs for formatting delegations)
- #67041 (add ExitStatusExt into prelude)
- #67065 (Fix fetching arguments on the wasm32-wasi target)
- #67066 (Update the revision of wasi-libc used in wasm32-wasi)
Yuki Okushi [Fri, 6 Dec 2019 06:37:14 +0000 (15:37 +0900)]
Rollup merge of #67066 - alexcrichton:update-wasi-libc, r=Mark-Simulacrum
Update the revision of wasi-libc used in wasm32-wasi
This commit updates the `wasi-libc` repository used to build the
wasm32-wasi target's libstd to ensure that both libstd and libc are
using the same wasi snapshot version.
Yuki Okushi [Fri, 6 Dec 2019 06:37:13 +0000 (15:37 +0900)]
Rollup merge of #67065 - alexcrichton:update-wasi, r=sfackler
Fix fetching arguments on the wasm32-wasi target
Fixes an error introduced in #66750 where wasi executables always think
they have zero arguments because one of the vectors returned here
accidentally thought it was length 0.
Yuki Okushi [Fri, 6 Dec 2019 06:37:10 +0000 (15:37 +0900)]
Rollup merge of #67021 - elichai:2019-12-fmt, r=QuietMisdreavus
Fix docs for formatting delegations
If you use the example in the docs right now it breaks all the options Formatters have to offer.
i.e. https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=214392ecc6eff73b4789c32568395f72 this should've padded the output with 4 zeros but didn't.
with the new example it does work: https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=3bdfb29f395230c5129c5f56dcfcb2a9
The only thing i'm not quite sure about is what's the right way to do it in a loop (altough non of the docs talk about it people are doing it in the wild and there were a couple of attempts to include in libcore)
i.e. https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=4c4dca3c90ba36779ecd014f3899ab9c
Yuki Okushi [Fri, 6 Dec 2019 06:37:06 +0000 (15:37 +0900)]
Rollup merge of #66979 - reese:E0631-long-error, r=GuillaumeGomez
Add long error for E0631 and update ui tests.
This PR adds a long error for `E0631`, which covers errors where closure argument types are mismatched. It also updates UI tests where this error is applicable.
Yuki Okushi [Fri, 6 Dec 2019 06:37:04 +0000 (15:37 +0900)]
Rollup merge of #66974 - cuviper:not-isCI, r=alexcrichton
[CI] fix the `! isCI` check in src/ci/run.sh
Using `if [ ! isCI ] || ...` doesn't run any command, just tests `isCI`
as a string, whereas `if ! isCI || ...` will actually run the `isCI`
command and negate its exit status.
bors [Fri, 6 Dec 2019 04:30:51 +0000 (04:30 +0000)]
Auto merge of #66835 - AviKozokin:master, r=alexcrichton
std:win: avoid WSA_FLAG_NO_INHERIT flag and don't use SetHandleInformation on UWP
This flag is not supported on Windows 7 before SP1, and on windows server 2008 SP2. This breaks Socket creation & duplication.
This was fixed in a previous PR. cc #26658
This PR: cc #60260 reuses this flag to support UWP, and makes an attempt to handle the potential error.
This version still fails to create a socket, as the error returned by WSA on this case is WSAEINVAL (invalid argument). and not WSAEPROTOTYPE.
MSDN page for WSASocketW (that states the platform support for WSA_FLAG_NO_HANDLE_INHERIT): https://docs.microsoft.com/en-us/windows/win32/api/winsock2/nf-winsock2-wsasocketw
Alex Crichton [Thu, 5 Dec 2019 19:36:35 +0000 (11:36 -0800)]
Update the revision of wasi-libc used in wasm32-wasi
This commit updates the `wasi-libc` repository used to build the
wasm32-wasi target's libstd to ensure that both libstd and libc are
using the same wasi snapshot version.
Alex Crichton [Thu, 5 Dec 2019 19:01:26 +0000 (11:01 -0800)]
Fix fetching arguments on the wasm32-wasi target
Fixes an error introduced in #66750 where wasi executables always think
they have zero arguments because one of the vectors returned here
accidentally thought it was length 0.
Rollup merge of #67011 - Aaron1011:fix/expected-found-span, r=Dylan-DPC
Include a span in more `expected...found` notes
In most places, we use a span when emitting `expected...found` errors.
However, there were a couple of places where we didn't use any span,
resulting in hard-to-interpret error messages.
This commit attaches the relevant span to these notes, and additionally
switches over to using `note_expected_found` instead of manually
formatting the message
Rollup merge of #66710 - vorner:weak-into-raw-null-docs, r=dtolnay
weak-into-raw: Clarify some details in Safety
Clarify it is OK to pass a pointer that never owned a weak count (one
from Weak::new) back into it as it was created from it. Relates to
discussion in #60728.
@CAD97 Do you want to have a look at the new docs?
bors [Thu, 5 Dec 2019 17:45:31 +0000 (17:45 +0000)]
Auto merge of #66828 - GuillaumeGomez:less-minification, r=kinnison
Less minification
The goal of this PR is to remove the minification process on the `search-index.js` file. It provides great result in term of space reduction but the computation time is far too long. I'll work on this issue and will put it back once it's fast enough.
bors [Thu, 5 Dec 2019 14:39:11 +0000 (14:39 +0000)]
Auto merge of #66815 - mark-i-m:simplify-borrow_check-errors, r=Dylan-DPC
Reorganize borrow check diagnostic code
Currently borrow checker diagnostics are split across many different modules in different places in the `librustc_mir` crate. This moves them all to a `diagnostics` module. This also reduces the nesting of the modules a bit (sooo much nesting).
I am also thinking of moving stuff out of the `nll` module since we only have one borrow checker now (:tada:), and maybe it even makes sense to split out all of this stuff to a `librustc_borrow_check`, but those are for the future. Feel free to ping me here or on zulip and let me know what you think...
Clarify it is OK to pass a pointer that never owned a weak count (one
from Weak::new) back into it as it was created from it. Relates to
discussion in #60728.
Josh Stone [Wed, 4 Dec 2019 20:00:28 +0000 (12:00 -0800)]
Migrate to LLVM{Get,Set}ValueName2
The deprecated `LLVM{Get,Set}ValueName` only work with NUL-terminated
strings, but the `2` variants use explicit lengths, which fits better
with Rust strings and slices. We now use these in new helper functions
`llvm::{get,set}_value_name` that convert to/from `&[u8]`.